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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
730850ed-f196-4b91-94f7-a2f5a2fcde0c | a-joint-training-framework-for-open-world | null | null | https://openreview.net/forum?id=HozL9BGbnKr | https://openreview.net/pdf?id=HozL9BGbnKr | A Joint Training Framework for Open-World Knowledge Graph Embeddings | Knowledge Graphs(KGs) represent factual information as graphs of entities connected by relations. Knowledge graph embeddings have emerged as a popular approach to encode this information for various downstream tasks like natural language inference, question answering and dialogue generation. As knowledge bases expand, we are presented with newer (open-world) entities, often with textual descriptions. We require techniques to embed new entities as they arrive using the textual information at hand. This task of open-world KG completion has received some attention in recent years. However, we find that existing approaches suffer from one or more of four drawbacks – 1) They are not modular with respect to the choice of the KG embedding model 2) They ignore best practices for aligning two embedding spaces 3) They do not account for differences in training strategy needed when presented with datasets with different description sizes and 4) They do not produce entity embeddings for use by downstream tasks. To address these problems, we propose FOlK (Framework for Open-World KG embeddings) - a technique that jointly learns embeddings for KG entities from descriptions and KG structure for open-world knowledge graph completion. Additionally, we modify existing data sources and make available YAGO3-10- Open and WN18RR-Open two datasets that are well suited for demonstrating the efficacy of open-world KG completion approaches. Finally, we empirically demonstrate the effectiveness of our model in improving upon state-of-the-art baselines on several tasks resulting in performance increases of up to 72% on mean reciprocal rank.
| ['Balaraman Ravindran', 'Mitesh M Khapra', 'Beethika Tripathi', 'Karthik V'] | 2021-06-22 | null | null | null | akbc-2021-10 | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [-2.81592906e-01 6.65491283e-01 -3.70746017e-01 -1.23953290e-01
-6.09631658e-01 -7.66991735e-01 8.49622965e-01 7.51471400e-01
-6.72011554e-01 7.89534271e-01 7.23493934e-01 -2.01157317e-01
-3.92536879e-01 -1.02653885e+00 -6.90510988e-01 3.69405709e-02
-2.93032348e-01 7.30165064e-01 2.73946494e-01 -4.94657099e-01
3.38057540e-02 9.80969295e-02 -1.31406128e+00 1.77565441e-01
9.30200577e-01 6.58695877e-01 -1.51643887e-01 5.04401207e-01
-3.49278510e-01 7.53405809e-01 -3.65686208e-01 -9.91095006e-01
9.96907279e-02 1.45767763e-01 -1.28612828e+00 -3.52912068e-01
4.67577487e-01 -2.65043467e-01 -7.15730667e-01 6.58778131e-01
4.66596156e-01 3.67178321e-01 7.92850077e-01 -1.33123469e+00
-1.31697428e+00 9.03032839e-01 -1.73457652e-01 8.03993121e-02
5.69430113e-01 -1.22797750e-01 1.60295367e+00 -1.02103698e+00
1.11580539e+00 1.10598612e+00 7.33310819e-01 4.69834030e-01
-1.33904576e+00 -2.65332371e-01 8.78837034e-02 4.88558024e-01
-1.42663622e+00 -3.78147274e-01 3.10233504e-01 -2.57179290e-01
1.48990059e+00 2.08605617e-01 5.33357620e-01 1.04951560e+00
-3.11011702e-01 5.93614757e-01 5.58113337e-01 -4.84596491e-01
-4.79559079e-02 2.61916578e-01 3.64879549e-01 8.07486236e-01
6.94737732e-01 -3.30854774e-01 -5.65390825e-01 -3.20489913e-01
4.53608543e-01 -2.99339980e-01 -4.60708976e-01 -5.74257851e-01
-1.34573889e+00 1.00428498e+00 7.02563703e-01 2.88281471e-01
-2.59298503e-01 2.91001022e-01 4.73757207e-01 3.52874845e-01
5.36152244e-01 9.59819138e-01 -6.16242051e-01 -9.66382399e-02
-3.54624689e-01 5.62808931e-01 1.17185509e+00 1.13852787e+00
8.44551623e-01 -4.04550225e-01 -1.19451366e-01 8.48913312e-01
2.46382385e-01 6.72026724e-02 4.72354561e-01 -6.66249871e-01
8.09762716e-01 9.04963434e-01 1.83456898e-01 -1.07625461e+00
-3.78698230e-01 -1.22473687e-01 -2.42425397e-01 -4.14031178e-01
5.28084874e-01 -1.28391683e-01 -6.96442842e-01 1.67033935e+00
4.75871533e-01 6.48292080e-02 4.49920356e-01 5.53951144e-01
1.40427840e+00 5.30650139e-01 1.77858889e-01 3.48813951e-01
1.58618164e+00 -8.79176259e-01 -7.38972962e-01 -2.00167552e-01
9.42973018e-01 -3.41540307e-01 1.00098789e+00 -1.21266179e-01
-7.36945808e-01 -1.99148446e-01 -1.04211390e+00 -5.43396413e-01
-1.22607291e+00 -2.34974071e-01 8.78716886e-01 4.55207944e-01
-1.12376928e+00 5.72088838e-01 -5.93273461e-01 -6.25533938e-01
2.57222503e-01 2.57154256e-01 -8.74710262e-01 -2.77627081e-01
-1.60658026e+00 1.18897355e+00 8.36912751e-01 -2.40012750e-01
-3.21454942e-01 -1.00538254e+00 -1.29743099e+00 2.55469292e-01
6.01318479e-01 -9.07282472e-01 9.09987330e-01 -1.68655276e-01
-8.87451649e-01 8.29735875e-01 1.84473068e-01 -4.83274192e-01
1.82624042e-01 -3.24397385e-01 -5.28564274e-01 9.02755465e-03
1.82906166e-01 7.75451183e-01 4.06511754e-01 -1.01017654e+00
-4.49013084e-01 -2.51573533e-01 6.38807118e-01 3.56832802e-01
-6.53830171e-01 -3.46557617e-01 -6.07627511e-01 -5.41657507e-01
-9.04452279e-02 -7.05429375e-01 1.75758302e-02 -1.17616102e-01
-4.62922752e-01 -5.30919433e-01 6.41153157e-01 -6.73074245e-01
1.37973106e+00 -1.87034202e+00 1.88312888e-01 4.69435967e-04
6.83197856e-01 4.32409912e-01 -1.77643776e-01 1.07323992e+00
-2.32836399e-02 5.66985428e-01 1.33629635e-01 -3.26634645e-01
3.16081762e-01 3.66060525e-01 -3.40543479e-01 3.89557630e-02
3.60481828e-01 1.31722403e+00 -1.15350401e+00 -3.63811493e-01
-1.29244616e-02 3.64517272e-01 -6.03166044e-01 7.99423605e-02
-3.53459746e-01 -1.90409899e-01 -4.43443835e-01 3.16861331e-01
3.08155537e-01 -4.17379707e-01 4.05020475e-01 -4.29653913e-01
2.67092675e-01 4.59757447e-01 -1.26041210e+00 1.68596697e+00
-4.36462492e-01 5.20553768e-01 -4.94051278e-01 -6.87989533e-01
5.66157043e-01 4.68218118e-01 7.38957003e-02 -3.66332918e-01
-1.75464481e-01 2.14416713e-01 -2.25651965e-01 -5.36435485e-01
9.84698117e-01 1.84949525e-02 -1.57918781e-01 4.32490677e-01
4.81059551e-01 -7.17166066e-02 5.42374849e-01 7.60331869e-01
1.44209921e+00 -4.35376987e-02 4.16041464e-01 3.12533751e-02
1.63388550e-01 1.31376758e-01 2.88461983e-01 5.83344996e-01
9.61600319e-02 2.19517067e-01 7.06789672e-01 -3.38736683e-01
-8.59269261e-01 -1.17076325e+00 -6.64321333e-02 1.02763057e+00
4.34159525e-02 -8.89331460e-01 -2.51347750e-01 -9.09412563e-01
4.80854958e-01 7.71159947e-01 -8.03108692e-01 -1.16951294e-01
-3.66915524e-01 -3.93158525e-01 8.54719996e-01 7.01572657e-01
1.35716617e-01 -9.90472674e-01 -1.22695595e-01 3.02349985e-01
-2.91263133e-01 -1.51764822e+00 -3.36675674e-01 -1.55215561e-02
-6.82203531e-01 -1.37694526e+00 -5.11542499e-01 -7.59785175e-01
6.93933189e-01 7.10201040e-02 1.41509974e+00 4.77431081e-02
-2.29979381e-01 7.97015786e-01 -6.40995264e-01 -2.08185777e-01
-3.07110131e-01 2.33253106e-01 2.94471029e-02 -2.14323044e-01
4.54913437e-01 -5.56690514e-01 -4.16243464e-01 -5.86820655e-02
-1.06712651e+00 -3.51039483e-03 4.04571831e-01 8.38540375e-01
2.44893700e-01 -9.16100815e-02 9.85097229e-01 -1.26333022e+00
1.15026319e+00 -7.09759831e-01 -3.49844277e-01 4.70674217e-01
-7.73862183e-01 4.05742645e-01 5.19275367e-01 -2.14484543e-01
-8.20090294e-01 -5.09558141e-01 5.08841276e-02 -2.91708916e-01
1.63222253e-01 1.01435983e+00 1.48703977e-01 9.46524553e-03
7.99220145e-01 -2.39579737e-01 -1.68073654e-01 -3.70623440e-01
1.16026723e+00 5.44700325e-01 4.95880485e-01 -7.27007449e-01
9.40785348e-01 2.12412864e-01 -2.30388910e-01 -6.60006225e-01
-7.38529921e-01 -6.18353724e-01 -4.92901295e-01 2.20970035e-01
7.92373478e-01 -9.91280854e-01 -6.10773385e-01 2.39711273e-02
-1.05354750e+00 -1.07015364e-01 -4.17555392e-01 3.67744088e-01
-2.22195044e-01 5.32819450e-01 -6.58368587e-01 -3.45442981e-01
-4.14516866e-01 -6.52147949e-01 7.94066012e-01 1.95334032e-01
-2.94707686e-01 -1.43594635e+00 1.06954984e-01 4.92830396e-01
4.98590440e-01 3.29919606e-01 1.31577420e+00 -1.04559982e+00
-6.14791632e-01 -3.32596511e-01 -4.27576244e-01 1.09237783e-01
1.71729475e-01 -1.26843259e-01 -8.38179052e-01 -1.47646949e-01
-9.45100486e-01 -6.75464213e-01 7.86507785e-01 -2.17597008e-01
6.29504442e-01 -5.42342365e-01 -4.87988353e-01 2.91669577e-01
1.46117496e+00 -5.17377257e-01 5.72208643e-01 4.07956839e-01
8.63425910e-01 7.57566869e-01 4.15424883e-01 3.40112805e-01
9.74853337e-01 6.41358733e-01 2.60219097e-01 1.24148734e-01
-2.10118815e-01 -7.24733531e-01 9.03584957e-02 8.05522799e-01
6.26452118e-02 -4.21195716e-01 -9.46012497e-01 1.08460701e+00
-1.93348980e+00 -8.30221534e-01 -1.04279555e-01 2.12515306e+00
1.09734845e+00 -3.04572601e-02 -1.02481149e-01 -1.67664856e-01
3.53159845e-01 2.34835893e-01 -3.23052347e-01 -2.92906255e-01
6.80997921e-03 3.53809774e-01 5.14321029e-01 4.88687158e-01
-9.11880732e-01 1.11157405e+00 5.53757334e+00 5.86498976e-01
-5.75342894e-01 8.47530290e-02 -6.52520806e-02 2.29162917e-01
-5.79605579e-01 2.21972898e-01 -8.53254676e-01 1.61160514e-01
7.78463542e-01 -4.97546017e-01 3.92805994e-01 6.79841220e-01
-4.55727547e-01 9.94959921e-02 -1.45776772e+00 8.97262931e-01
-3.94045971e-02 -1.59677947e+00 1.87922820e-01 -3.12404819e-02
5.67317843e-01 1.18380696e-01 -1.94424868e-01 8.32102656e-01
6.73978090e-01 -1.07999861e+00 4.10294831e-01 1.94200203e-01
8.71062756e-01 -3.93766195e-01 6.37963116e-01 1.71947390e-01
-1.25339174e+00 1.26102328e-01 -3.72894168e-01 3.64657231e-02
1.53333530e-01 3.45780700e-01 -1.29752886e+00 1.02740014e+00
4.41542417e-01 5.51598549e-01 -7.36547172e-01 7.12360024e-01
-6.20266736e-01 3.26935172e-01 -2.60386854e-01 -1.10067159e-01
2.52091467e-01 1.73534676e-01 3.69214416e-01 1.18993437e+00
2.98199710e-02 -9.81124938e-02 -8.78646038e-03 9.52288926e-01
-5.88744402e-01 2.20611826e-01 -9.40911531e-01 -5.50515175e-01
7.67764866e-01 1.29919672e+00 -4.33086336e-01 -3.34545672e-01
-6.68891430e-01 7.78760135e-01 8.93200159e-01 4.18013573e-01
-6.06580257e-01 -7.09069073e-01 7.23680437e-01 1.39512360e-01
4.44217443e-01 -2.49327555e-01 2.83518583e-01 -1.37385821e+00
1.95067659e-01 -6.61414385e-01 7.91149735e-01 -8.28235388e-01
-1.56835282e+00 4.73279446e-01 2.82792687e-01 -6.95990682e-01
-3.82061571e-01 -6.09852374e-01 -2.54258841e-01 8.04299057e-01
-1.64662170e+00 -1.33010089e+00 -1.99462041e-01 4.48726237e-01
1.56169951e-01 1.79819465e-01 1.15084136e+00 4.19402212e-01
-2.26090029e-01 5.74789703e-01 -1.03376093e-04 4.59034443e-01
7.37744033e-01 -1.50275326e+00 8.41631413e-01 6.21602476e-01
5.30561268e-01 9.35594499e-01 5.30565441e-01 -7.08420634e-01
-1.52390206e+00 -9.32780564e-01 1.32852590e+00 -8.84984255e-01
9.98621166e-01 -4.78158742e-01 -1.00089252e+00 1.00514889e+00
1.57602161e-01 3.20185035e-01 7.90408731e-01 7.79187441e-01
-6.91379130e-01 1.26367822e-01 -9.94348109e-01 7.37380803e-01
1.19035530e+00 -6.44191802e-01 -9.40704823e-01 3.44052970e-01
1.01310861e+00 -4.67060298e-01 -1.31424057e+00 2.87978888e-01
3.88717711e-01 -4.45895255e-01 9.03623283e-01 -9.69947755e-01
3.25107157e-01 -3.07740986e-01 -1.65877849e-01 -1.50108242e+00
-2.78587848e-01 -4.42219347e-01 -5.38240314e-01 1.44132471e+00
6.93765283e-01 -8.58055413e-01 5.96715927e-01 7.82290697e-01
-3.62539887e-02 -9.18249667e-01 -7.51325667e-01 -7.00222492e-01
-1.31541580e-01 -1.71873331e-01 5.66603541e-01 1.27225733e+00
4.91299152e-01 6.34366632e-01 5.17424662e-03 2.05228850e-01
2.13459581e-01 -2.22387817e-02 9.38372850e-01 -1.26902509e+00
-1.65622145e-01 -6.09888770e-02 -6.76468968e-01 -8.95999730e-01
2.54989415e-01 -1.14684129e+00 -4.22731191e-01 -2.15127897e+00
1.57615051e-01 -5.69849372e-01 -1.88936338e-01 8.30948532e-01
-3.87100816e-01 -3.70048098e-02 1.58200357e-02 5.10496367e-03
-6.81427479e-01 6.28153324e-01 8.21585298e-01 -9.77393314e-02
-8.25170502e-02 -6.58491790e-01 -1.00627387e+00 3.24320138e-01
3.91176552e-01 -3.62690598e-01 -7.24329770e-01 -5.93269169e-01
6.59620821e-01 -1.26501799e-01 3.77714515e-01 -6.87714100e-01
2.44338006e-01 1.32748604e-01 1.64386304e-03 -1.29566759e-01
4.22539443e-01 -5.71377218e-01 8.61416459e-02 2.03854553e-02
-2.99660563e-01 1.78773060e-01 3.23970139e-01 8.66166651e-01
-2.22679496e-01 -2.59042263e-01 1.08578928e-01 -5.12385629e-02
-1.13729954e+00 2.07124352e-01 1.54662495e-02 6.47214592e-01
8.98152888e-01 -1.39204174e-01 -8.69508266e-01 -3.90654117e-01
-6.35419071e-01 4.80303884e-01 3.75848353e-01 7.62467682e-01
5.22817910e-01 -1.30496573e+00 -6.53283358e-01 -2.00514331e-01
6.92918956e-01 2.88663339e-02 4.90391180e-02 5.83955586e-01
-5.56927264e-01 5.28526902e-01 1.37081556e-02 2.56708618e-02
-1.00061011e+00 5.66076398e-01 -1.23623302e-02 -7.47456193e-01
-6.76703215e-01 8.74432981e-01 -4.27479222e-02 -8.04427922e-01
8.15510005e-02 -4.06294376e-01 -3.33305746e-01 2.18779176e-01
2.56874174e-01 2.98558891e-01 2.93710381e-01 -4.77918357e-01
-3.60076159e-01 1.10173263e-01 -3.44959587e-01 3.50150338e-04
1.51285744e+00 5.95097430e-02 -4.07810286e-02 2.32481420e-01
1.20981061e+00 2.03365177e-01 -6.62747741e-01 -5.42362690e-01
2.26616994e-01 -4.56734449e-01 -1.82429582e-01 -7.93264270e-01
-6.08924091e-01 3.55190217e-01 5.23422509e-02 3.72897834e-01
5.31597912e-01 2.59284437e-01 8.09568405e-01 7.27803409e-01
4.33181554e-01 -9.46558177e-01 7.34033734e-02 5.69225132e-01
8.41652930e-01 -1.12819743e+00 2.33273849e-01 -4.60800171e-01
-6.74807310e-01 1.02861845e+00 5.39881349e-01 1.73404098e-01
4.61435437e-01 -2.74759531e-01 -9.38476548e-02 -6.92573547e-01
-9.58498299e-01 -5.35548985e-01 4.43108022e-01 7.80769348e-01
4.60175872e-01 -4.76073893e-03 -2.90067375e-01 5.82564712e-01
-1.58021361e-01 -1.14453070e-01 4.58980560e-01 9.50826526e-01
-1.53676182e-01 -1.21223116e+00 1.88775048e-01 6.58067405e-01
-4.13372964e-01 -3.69813234e-01 -4.27325875e-01 1.07058120e+00
-1.10959575e-01 8.84309471e-01 -3.21735710e-01 -3.05619538e-01
5.02126455e-01 2.84514517e-01 4.79395837e-01 -1.03177011e+00
-3.71122330e-01 -8.86687636e-01 6.88999474e-01 -3.97576958e-01
-1.19619526e-01 -2.94353813e-01 -1.24264491e+00 -4.00522113e-01
-6.23711169e-01 2.63322115e-01 4.32463288e-01 7.58137167e-01
7.55941868e-01 4.02310967e-01 -2.12761145e-02 -3.63325775e-01
-5.86296856e-01 -9.12189662e-01 -6.88634872e-01 6.76736653e-01
-5.07166497e-02 -8.41538131e-01 -1.13710195e-01 9.33262892e-03] | [9.101128578186035, 8.08977222442627] |
d934773b-a41d-4e02-bf3f-379f8e6f1f7c | deep-quality-a-deep-no-reference-quality | 1609.07170 | null | http://arxiv.org/abs/1609.07170v1 | http://arxiv.org/pdf/1609.07170v1.pdf | Deep Quality: A Deep No-reference Quality Assessment System | Image quality assessment (IQA) continues to garner great interest in the
research community, particularly given the tremendous rise in consumer video
capture and streaming. Despite significant research effort in IQA in the past
few decades, the area of no-reference image quality assessment remains a great
challenge and is largely unsolved. In this paper, we propose a novel
no-reference image quality assessment system called Deep Quality, which
leverages the power of deep learning to model the complex relationship between
visual content and the perceived quality. Deep Quality consists of a novel
multi-scale deep convolutional neural network, trained to learn to assess image
quality based on training samples consisting of different distortions and
degradations such as blur, Gaussian noise, and compression artifacts.
Preliminary results using the CSIQ benchmark image quality dataset showed that
Deep Quality was able to achieve strong quality prediction performance (89%
patch-level and 98% image-level prediction accuracy), being able to achieve
similar performance as full-reference IQA methods. | ['Alexander Wong', 'Prajna Paramita Dash', 'Akshaya Mishra'] | 2016-09-22 | null | null | null | null | ['no-reference-image-quality-assessment'] | ['computer-vision'] | [ 3.14563870e-01 -5.54515660e-01 7.56502599e-02 -3.88110071e-01
-1.08495784e+00 -2.15531588e-01 4.47918087e-01 2.42844131e-02
-2.43387967e-02 3.97774369e-01 4.50814337e-01 -1.96807280e-01
-2.03406766e-01 -8.01805198e-01 -4.95449752e-01 -5.01671851e-01
-3.36217463e-01 -2.95087278e-01 9.45833791e-03 -3.09656739e-01
3.51831704e-01 3.02349150e-01 -1.59087956e+00 5.12752056e-01
9.93599296e-01 1.67723501e+00 -6.49701208e-02 1.07962966e+00
4.71675962e-01 9.76789296e-01 -9.18606102e-01 -5.11779189e-01
4.44239020e-01 -5.86626649e-01 -6.78365707e-01 1.95327342e-01
8.95431280e-01 -8.54447544e-01 -8.27973664e-01 1.03124881e+00
9.50728834e-01 -1.27560139e-01 2.72147954e-01 -1.20681274e+00
-8.69125307e-01 6.29432350e-02 -3.45825970e-01 6.91378355e-01
4.72457856e-01 5.62692463e-01 1.15619445e+00 -6.09888375e-01
2.12785840e-01 1.11806297e+00 7.41766393e-01 7.01336414e-02
-1.32983673e+00 -5.18064320e-01 -5.64219356e-01 7.08167911e-01
-1.07366383e+00 -5.52071810e-01 5.29257298e-01 -3.67652386e-01
9.17815089e-01 -1.21305034e-01 7.14170694e-01 8.43193948e-01
7.39544213e-01 6.20351374e-01 1.16100347e+00 -1.10355243e-01
3.02579075e-01 -4.59504098e-01 -3.37950617e-01 4.08585817e-01
-1.02241874e-01 3.89676690e-01 -5.96870124e-01 1.52612656e-01
8.31980348e-01 -2.46915162e-01 -4.87812102e-01 -2.89094746e-01
-1.08710539e+00 5.54391861e-01 6.52322531e-01 2.20253989e-01
-3.65396708e-01 2.86857039e-01 4.34186012e-01 6.73436224e-01
3.99168849e-01 4.78676885e-01 -2.04658896e-01 -5.21746218e-01
-1.34612024e+00 3.51097643e-01 4.59420592e-01 4.06254292e-01
4.24845695e-01 3.96552682e-01 -3.25451881e-01 9.66697514e-01
6.95682094e-02 7.32280970e-01 3.08034122e-01 -1.47171068e+00
3.08174551e-01 3.64525229e-01 2.73500860e-01 -1.41848278e+00
-2.66882688e-01 -6.53329074e-01 -1.10749376e+00 8.78052533e-01
3.21833193e-01 2.35830858e-01 -7.05691695e-01 1.31121647e+00
-1.86463684e-01 1.12664670e-01 -2.33982671e-02 1.13388455e+00
7.27173150e-01 8.35614026e-01 -2.35037416e-01 -5.71274571e-02
1.00656724e+00 -8.48285019e-01 -6.68637276e-01 8.79658610e-02
-2.20871828e-02 -8.89824331e-01 9.71051872e-01 8.54263067e-01
-1.70925784e+00 -1.13746262e+00 -1.44705403e+00 -9.67447460e-02
6.86220452e-02 -1.47390902e-01 3.32797706e-01 7.58007646e-01
-1.54876494e+00 8.90284479e-01 -6.11540735e-01 -2.43987031e-02
7.42364347e-01 2.77679205e-01 -3.25925887e-01 -5.65639913e-01
-1.23009086e+00 6.81620181e-01 -2.50259240e-04 -8.33965465e-02
-1.24240112e+00 -1.01195157e+00 -6.74259663e-01 3.28213632e-01
1.50875345e-01 -7.92275369e-01 1.14302886e+00 -1.30220807e+00
-1.54532826e+00 6.59223557e-01 8.09624940e-02 -5.67024350e-01
4.33075935e-01 -2.88909763e-01 -9.41079021e-01 5.23718297e-01
1.31869374e-03 5.44671774e-01 9.42680418e-01 -1.06863129e+00
-7.19069600e-01 -3.59618515e-01 3.13301772e-01 8.92274454e-02
-5.83330467e-02 1.57575399e-01 -4.95250911e-01 -7.27763057e-01
6.52506649e-02 -5.25491357e-01 1.16391607e-01 3.75135332e-01
9.13985744e-02 8.61462429e-02 5.15266895e-01 -8.62588406e-01
1.33782625e+00 -2.10224557e+00 -1.02514639e-01 -8.80259797e-02
4.79425639e-01 5.15025377e-01 -5.58896184e-01 2.77735233e-01
-8.33328217e-02 1.46267042e-01 -2.39531592e-01 5.71006276e-02
-6.96057007e-02 -1.68070719e-01 -2.98671052e-02 4.75495845e-01
4.84077096e-01 9.45796311e-01 -1.03246415e+00 -1.49275959e-01
3.54554594e-01 6.60212398e-01 -5.61878860e-01 5.32973170e-01
6.39823675e-02 2.07415208e-01 1.11618176e-01 8.17470670e-01
8.83995354e-01 -5.25233984e-01 -2.66682357e-01 -5.51880896e-01
1.09310456e-01 1.64768144e-01 -1.08970642e+00 1.75688577e+00
-5.35783887e-01 9.75774586e-01 2.07707003e-01 -6.79205358e-01
7.86848366e-01 4.08582628e-01 7.33156145e-01 -1.65622854e+00
7.29735121e-02 2.82812446e-01 2.33950034e-01 -5.84945202e-01
5.17692506e-01 -8.05970505e-02 2.31537670e-01 2.74162203e-01
9.90105793e-02 -1.36290818e-01 1.26047254e-01 2.53008693e-01
1.43096268e+00 -3.45669612e-02 6.31651953e-02 1.83788082e-03
4.73826915e-01 -3.74876171e-01 3.41948837e-01 5.56381404e-01
-7.40671456e-01 9.62500930e-01 3.42398643e-01 -5.49375653e-01
-1.47280836e+00 -1.31448841e+00 -1.29771307e-01 8.15064788e-01
2.82709122e-01 -2.36090273e-01 -5.47801852e-01 -1.60347641e-01
-1.65951297e-01 1.46200895e-01 -4.66831625e-01 -3.16735744e-01
-4.20396894e-01 -5.37692368e-01 3.42147201e-01 3.59314203e-01
9.75304067e-01 -1.22944915e+00 -5.68148077e-01 4.43196833e-01
-3.37410420e-01 -1.06419456e+00 -3.21211606e-01 -2.88513988e-01
-7.89004982e-01 -1.07923400e+00 -9.89925385e-01 -4.33735520e-01
-1.81330338e-01 3.57724011e-01 1.80691755e+00 9.95785594e-02
-3.81612569e-01 2.47002840e-01 -4.03570205e-01 1.71343699e-01
-6.61118805e-01 -3.46933693e-01 -2.58602470e-01 -1.18304249e-02
-1.06702475e-02 -6.77786350e-01 -1.31380868e+00 3.36087108e-01
-1.14614046e+00 -2.27542475e-01 6.65118098e-01 8.52129281e-01
4.16799188e-01 5.05706429e-01 6.87490284e-01 -8.52859914e-02
7.19609559e-01 -3.54852021e-01 -4.55990046e-01 -7.90155083e-02
-8.18186760e-01 -3.54259312e-01 5.06757736e-01 5.32559715e-02
-8.72286797e-01 -7.43877292e-01 -5.80852211e-01 -3.02474678e-01
-1.79437339e-01 4.19915706e-01 -2.93975681e-01 -2.15075657e-01
7.45194852e-01 1.78332388e-01 -3.43111306e-02 -2.13930875e-01
2.26130769e-01 6.35332048e-01 8.59811604e-01 -1.48157433e-01
5.85206985e-01 3.69698286e-01 3.69994044e-02 -5.10171652e-01
-5.70901394e-01 -4.07838613e-01 -3.13902646e-01 -4.38891917e-01
7.40760922e-01 -1.14046609e+00 -6.57447517e-01 8.22646856e-01
-8.05660605e-01 -3.97136897e-01 -1.13203540e-01 2.56240696e-01
-6.13013506e-01 5.01666605e-01 -8.23055923e-01 -4.05306190e-01
-5.53887248e-01 -1.27781701e+00 1.01365805e+00 1.44096151e-01
2.17046633e-01 -8.48638892e-01 3.19931768e-02 5.05592227e-01
9.71383691e-01 3.58257979e-01 8.80908430e-01 2.80491322e-01
-8.66244376e-01 -1.11789644e-01 -7.02426493e-01 8.17810535e-01
2.05706283e-01 -1.37188554e-01 -8.89568985e-01 -5.38289547e-01
-8.89079049e-02 -6.06090307e-01 6.94551468e-01 6.11820877e-01
1.14694142e+00 -6.67815655e-02 4.68434542e-01 8.14775109e-01
1.69397259e+00 2.33855978e-01 1.26612341e+00 4.11778897e-01
4.15131956e-01 1.13604893e-03 3.36907327e-01 5.75560272e-01
1.48061872e-01 9.59023476e-01 6.93699718e-01 -3.49806935e-01
-6.64270282e-01 2.96865031e-02 2.10330009e-01 6.81397796e-01
4.97066975e-02 -3.39266300e-01 -8.31715703e-01 5.03708065e-01
-1.16118872e+00 -1.20026481e+00 8.57050642e-02 1.99656105e+00
7.06683695e-01 2.14454114e-01 9.28783864e-02 5.72988510e-01
3.78568590e-01 2.59708434e-01 -7.60323524e-01 -3.97587925e-01
-4.06560481e-01 3.15705776e-01 7.97230154e-02 1.29513815e-01
-9.47919548e-01 4.23286408e-01 7.05830526e+00 7.31374919e-01
-1.29729784e+00 -1.38744280e-01 1.10653019e+00 -4.05898131e-02
-9.67050195e-02 -5.88923037e-01 -7.09295645e-02 5.87931514e-01
1.32845378e+00 3.27730291e-02 6.39156342e-01 4.67227787e-01
6.17348433e-01 -1.59184560e-01 -8.78307343e-01 1.36741042e+00
1.41943365e-01 -1.26969743e+00 7.16553107e-02 2.87157353e-02
9.24947023e-01 2.45678082e-01 5.18411100e-01 1.87061638e-01
9.75664407e-02 -1.27421308e+00 6.14453018e-01 4.61246341e-01
1.13662863e+00 -6.28933191e-01 8.88264120e-01 -1.28331274e-01
-1.01622045e+00 -2.29871929e-01 -4.41415310e-01 -1.45970792e-01
2.22144201e-01 7.18095839e-01 -1.67915612e-01 5.25590360e-01
1.12540531e+00 9.31191742e-01 -8.18211138e-01 1.37046325e+00
1.60597697e-01 7.58837044e-01 2.88740158e-01 6.89166069e-01
1.14431925e-01 -1.85830835e-02 2.25067601e-01 1.00256789e+00
5.21514237e-01 8.36229175e-02 -1.12934314e-01 6.52646065e-01
-3.22911561e-01 5.81626315e-03 -1.44852102e-01 1.94459986e-02
-1.22986399e-01 1.03959453e+00 -2.34496608e-01 -3.08184624e-01
-5.35454273e-01 1.18549073e+00 -1.19160235e-01 3.65772933e-01
-8.18696439e-01 -2.28372663e-01 8.97472143e-01 2.21038163e-01
3.28190714e-01 -1.03623919e-01 -3.55089784e-01 -1.04404402e+00
2.12568007e-02 -1.38158286e+00 5.81843033e-02 -1.18224490e+00
-1.26854730e+00 6.88455403e-01 -5.36860764e-01 -1.42971694e+00
2.28454117e-02 -5.56609452e-01 -4.16475624e-01 1.01677382e+00
-1.80202806e+00 -6.40057921e-01 -6.63353384e-01 6.09731734e-01
4.90846574e-01 -5.73129021e-02 6.22512460e-01 6.88020170e-01
-9.95127261e-02 7.57219315e-01 2.83345103e-01 1.93374455e-01
7.78977036e-01 -1.18770361e+00 4.70281631e-01 8.54287148e-01
-1.02426358e-01 1.82541013e-01 6.10278130e-01 -1.86861694e-01
-1.49189210e+00 -9.45695519e-01 5.07875741e-01 -1.31575271e-01
5.93492985e-01 1.36944830e-01 -9.45053339e-01 -1.76520348e-01
5.81937611e-01 3.26682627e-01 5.66450238e-01 -2.71424443e-01
-6.20647013e-01 -6.39422178e-01 -1.27000260e+00 3.43472511e-01
7.76287854e-01 -7.80128837e-01 -1.31199002e-01 -1.57100499e-01
6.68222487e-01 -2.03745320e-01 -1.29602098e+00 6.58913374e-01
7.25095212e-01 -1.62189651e+00 1.14144564e+00 5.18504381e-02
1.07235646e+00 -3.36149693e-01 -3.61895800e-01 -1.39757383e+00
-6.82109118e-01 -2.92042464e-01 -2.89371729e-01 9.08453703e-01
1.92634179e-03 4.14644703e-02 6.76931262e-01 2.86962092e-01
-1.98750138e-01 -7.67182291e-01 -8.59644294e-01 -6.08847201e-01
5.89785464e-02 -3.80366951e-01 5.42432010e-01 4.69298512e-01
-3.26016247e-01 2.42330387e-01 -5.61296940e-01 9.00616869e-02
6.16057634e-01 9.43149403e-02 6.48588717e-01 -9.16366935e-01
-5.64306259e-01 -7.87436783e-01 -8.33126366e-01 -1.18642557e+00
-4.03907955e-01 -4.74212080e-01 -7.86556229e-02 -1.89014423e+00
1.85175985e-01 -2.93258041e-01 -7.38034427e-01 -5.96003458e-02
-3.13191026e-01 8.58808100e-01 1.68645427e-01 1.12271622e-01
-9.39355850e-01 6.62309349e-01 1.44239891e+00 -5.78898251e-01
1.14887282e-01 -1.20742872e-01 -6.06878638e-01 1.32588178e-01
6.77620590e-01 5.27187809e-02 -3.02327156e-01 -6.59164369e-01
4.74500656e-01 5.50342143e-01 4.14476663e-01 -1.70652282e+00
-1.84028745e-01 1.38943136e-01 6.53078496e-01 -3.96684527e-01
2.41536736e-01 -8.12095940e-01 2.14481637e-01 3.83557230e-01
-4.27973062e-01 1.88817501e-01 1.69234842e-01 4.66135025e-01
-6.68822527e-01 3.30459028e-01 1.06209660e+00 1.54408682e-02
-9.33316052e-01 5.57977557e-01 -3.52424830e-02 8.95620212e-02
4.51171309e-01 -3.13888311e-01 -3.05311680e-01 -7.29849398e-01
-5.74688315e-01 -1.61764979e-01 6.34231627e-01 4.86090571e-01
9.53170240e-01 -1.46317804e+00 -9.92411375e-01 3.56445797e-02
2.56654710e-01 -5.67329109e-01 7.10536838e-01 4.99119401e-01
-6.11681104e-01 2.15234131e-01 -6.07172728e-01 -7.08946645e-01
-9.75157022e-01 6.80812836e-01 5.03840983e-01 -2.68834114e-01
-4.85820442e-01 5.52711368e-01 -1.33579075e-01 2.59560317e-01
1.27970293e-01 -1.98376432e-01 -2.00194374e-01 -3.59879524e-01
1.03121686e+00 6.70831740e-01 3.58236343e-01 -7.87277758e-01
9.79582742e-02 4.05636668e-01 3.26132834e-01 8.14598873e-02
1.23936749e+00 -4.62858230e-01 1.78340197e-01 2.63316005e-01
1.53024948e+00 -3.43837708e-01 -1.64074564e+00 -2.79129446e-01
-2.97603846e-01 -8.20847571e-01 4.82002527e-01 -1.40212011e+00
-1.37328887e+00 1.18188071e+00 1.41821563e+00 3.80540639e-01
1.68684185e+00 -3.98312598e-01 1.04848659e+00 -1.72230437e-01
4.45093364e-01 -7.55580127e-01 6.39721036e-01 2.37315655e-01
9.27294254e-01 -1.59974444e+00 -1.11598410e-02 9.70160961e-02
-1.98681638e-01 9.98514235e-01 3.08841288e-01 -1.63126841e-01
5.81517398e-01 3.35248783e-02 3.18338186e-01 4.29436006e-02
-6.83274567e-01 2.42129322e-02 4.12716687e-01 6.81742251e-01
6.20766878e-01 -5.28008975e-02 1.03820801e-01 9.53809172e-02
-4.01502363e-02 4.18023348e-01 4.93627191e-01 4.27860856e-01
-4.49925572e-01 -8.13443720e-01 -3.11082423e-01 5.22918463e-01
-8.82720888e-01 -2.85537720e-01 3.81468922e-01 2.63069689e-01
3.17414850e-01 1.36685288e+00 4.01253328e-02 -4.39644128e-01
4.24417734e-01 -5.10579169e-01 4.47440535e-01 -4.93518226e-02
-6.33834004e-01 -1.80298202e-02 -2.80335575e-01 -1.15204895e+00
-5.71614683e-01 -2.90245593e-01 -6.62638068e-01 -6.61511600e-01
7.76352510e-02 -3.00040066e-01 6.54243350e-01 6.64983392e-01
3.70657414e-01 8.02745104e-01 7.89261043e-01 -9.05915022e-01
-3.54911298e-01 -8.47740889e-01 -7.09677100e-01 8.73247147e-01
6.53555572e-01 -3.80918890e-01 -2.56436914e-01 3.04084539e-01] | [11.806092262268066, -1.8198254108428955] |
0709f91f-d435-4fbe-ad30-b60656221aa4 | cot-mote-exploring-contextual-masked-auto | 2304.10195 | null | https://arxiv.org/abs/2304.10195v1 | https://arxiv.org/pdf/2304.10195v1.pdf | CoT-MoTE: Exploring ConTextual Masked Auto-Encoder Pre-training with Mixture-of-Textual-Experts for Passage Retrieval | Passage retrieval aims to retrieve relevant passages from large collections of the open-domain corpus. Contextual Masked Auto-Encoding has been proven effective in representation bottleneck pre-training of a monolithic dual-encoder for passage retrieval. Siamese or fully separated dual-encoders are often adopted as basic retrieval architecture in the pre-training and fine-tuning stages for encoding queries and passages into their latent embedding spaces. However, simply sharing or separating the parameters of the dual-encoder results in an imbalanced discrimination of the embedding spaces. In this work, we propose to pre-train Contextual Masked Auto-Encoder with Mixture-of-Textual-Experts (CoT-MoTE). Specifically, we incorporate textual-specific experts for individually encoding the distinct properties of queries and passages. Meanwhile, a shared self-attention layer is still kept for unified attention modeling. Results on large-scale passage retrieval benchmarks show steady improvement in retrieval performances. The quantitive analysis also shows a more balanced discrimination of the latent embedding spaces. | ['Songlin Hu', 'Peng Wang', 'Xing Wu', 'Guangyuan Ma'] | 2023-04-20 | null | null | null | null | ['passage-retrieval'] | ['natural-language-processing'] | [-1.81668013e-01 -4.58231419e-01 -7.70688653e-02 -1.34154350e-01
-1.74624753e+00 -7.18122900e-01 7.74688125e-01 3.82973313e-01
-6.13734722e-01 5.93935966e-01 6.13303363e-01 -5.77754639e-02
-1.43179864e-01 -6.68090999e-01 -6.74134910e-01 -5.54312468e-01
7.33388215e-02 6.36088729e-01 7.14137927e-02 -4.46002811e-01
2.22818583e-01 1.08894698e-01 -1.34017944e+00 6.50812864e-01
8.87292862e-01 9.44102228e-01 2.65752673e-01 1.06962490e+00
-2.96513855e-01 5.60564101e-01 -7.70531237e-01 -4.80375379e-01
2.66330630e-01 -5.24067402e-01 -1.11848545e+00 -3.21463674e-01
5.43077230e-01 -7.12895215e-01 -9.21862304e-01 6.38464749e-01
7.70390749e-01 5.34178436e-01 1.13923919e+00 -6.11416459e-01
-1.30774593e+00 5.60218930e-01 -2.19367683e-01 6.54303730e-01
4.50167626e-01 -8.42240751e-02 1.56821287e+00 -8.27138364e-01
3.76976877e-01 9.49992776e-01 1.72391802e-01 4.11392093e-01
-1.01874113e+00 -3.02610636e-01 -5.64017333e-02 4.66529965e-01
-1.51895440e+00 -4.58661377e-01 6.10793769e-01 -2.48445660e-01
1.36098921e+00 6.13597512e-01 4.40500557e-01 1.35479510e+00
2.36816760e-02 1.22310090e+00 1.43985689e-01 -4.76256728e-01
-1.06378973e-01 2.96143442e-01 3.31119061e-01 1.65164158e-01
1.41400704e-02 -2.80591369e-01 -2.74926722e-01 -2.90557146e-01
5.31703413e-01 2.72314280e-01 -4.31116313e-01 -3.72090675e-02
-8.96464527e-01 8.27716053e-01 5.88575900e-01 3.81251127e-01
-3.27988267e-01 -1.08122446e-01 8.58204603e-01 6.81909263e-01
5.23309588e-01 7.12484539e-01 -1.63655266e-01 5.70927598e-02
-1.23621714e+00 4.71475929e-01 5.18473983e-01 1.18358397e+00
4.56098258e-01 -4.11092848e-01 -1.06091213e+00 1.09797931e+00
2.62361288e-01 3.38598937e-01 9.10000503e-01 -5.35407662e-01
7.43040681e-01 5.11106730e-01 2.97638237e-01 -7.62523830e-01
5.19352667e-02 -8.71850967e-01 -7.87042975e-01 -7.14334786e-01
-2.37858091e-02 2.45988369e-01 -8.29385936e-01 1.44162261e+00
-2.35847626e-02 -1.03427716e-01 2.92569757e-01 1.13649058e+00
6.57576740e-01 9.97033119e-01 8.97417367e-02 2.25651771e-01
1.49478316e+00 -1.65942597e+00 -7.42299378e-01 -1.81696698e-01
6.23187542e-01 -9.09523785e-01 1.20949709e+00 -2.76858330e-01
-1.44540823e+00 -7.67151952e-01 -1.23360085e+00 -8.86451542e-01
-6.14918172e-01 4.20208663e-01 6.87587634e-03 7.32651353e-02
-9.20106113e-01 3.39977831e-01 -5.92810273e-01 -2.00713739e-01
3.20728481e-01 1.24654338e-01 -2.59281188e-01 -1.66314289e-01
-1.53497899e+00 6.97833955e-01 3.48467946e-01 1.39771670e-01
-8.91049981e-01 -7.35618830e-01 -6.87748730e-01 7.24084854e-01
-1.97030589e-01 -1.14172852e+00 1.14660549e+00 -5.96415460e-01
-1.34800458e+00 8.61902952e-01 3.14152837e-02 -4.27926153e-01
3.83971512e-01 -6.11449301e-01 -4.30611342e-01 5.95354974e-01
2.21259639e-01 5.10515869e-01 1.05957639e+00 -8.47419798e-01
-2.51330763e-01 -1.47537753e-01 3.45502377e-01 5.18799305e-01
-7.95778811e-01 1.25577450e-01 -8.61510456e-01 -8.18798721e-01
-2.23430127e-01 -6.59349799e-01 5.28018037e-03 -3.07342261e-01
-3.63950074e-01 -4.98529345e-01 5.32097518e-01 -8.04961860e-01
1.66450310e+00 -2.21045780e+00 2.21943170e-01 -1.96753815e-01
1.92432404e-01 4.03829366e-01 -6.15799189e-01 8.87804568e-01
1.54773965e-01 2.69062594e-02 1.50089785e-01 -7.59388745e-01
2.92997956e-01 -8.53741169e-02 -7.51965940e-01 2.65242517e-01
3.93658400e-01 1.15130568e+00 -9.45488870e-01 -6.44061148e-01
-4.35666852e-02 6.10031962e-01 -7.42428720e-01 6.20890677e-01
-1.26016289e-01 -3.35359760e-02 -6.59837186e-01 2.15630010e-01
4.31195676e-01 -5.39425850e-01 -3.61214966e-01 -4.29520756e-02
4.63918716e-01 8.43507648e-01 -4.91353482e-01 2.17250967e+00
-6.48826301e-01 7.88860381e-01 -3.04324150e-01 -7.75698960e-01
6.36312366e-01 6.11122072e-01 2.23828778e-01 -9.82166052e-01
1.68806434e-01 2.63263732e-01 -3.34531724e-01 -6.40118539e-01
1.06269228e+00 -4.76884469e-03 -2.79981673e-01 8.27784896e-01
4.64443237e-01 2.74059594e-01 4.39839363e-01 4.87222791e-01
1.14821482e+00 3.15731764e-02 -1.57276839e-01 -1.66043192e-02
8.47796023e-01 -1.37930408e-01 -8.51440802e-02 9.27870631e-01
7.45101050e-02 1.12933791e+00 5.75862639e-02 -1.31103605e-01
-1.19735539e+00 -1.05041122e+00 -4.71629053e-02 1.38869572e+00
1.77622244e-01 -4.63766426e-01 -5.36234498e-01 -5.42237639e-01
-5.20342551e-02 4.51200426e-01 -6.54495716e-01 -6.94995165e-01
-5.55372238e-01 -4.22352165e-01 7.60655940e-01 6.81291938e-01
6.51813149e-02 -1.02889383e+00 -4.47328776e-01 3.77352059e-01
-4.68156517e-01 -1.00518084e+00 -8.27945054e-01 1.06842004e-01
-5.77924609e-01 -7.38246977e-01 -1.36404526e+00 -8.26814711e-01
4.66053158e-01 4.53723431e-01 1.29363227e+00 1.73814937e-01
-2.70794153e-01 3.57023478e-01 -6.87659383e-01 6.01851474e-03
-6.12842552e-02 6.91766858e-01 -2.45436639e-01 -7.18458518e-02
5.62038124e-01 -4.73991483e-01 -1.10027504e+00 1.11610860e-01
-1.16739154e+00 -3.85672152e-01 5.45544565e-01 1.08897698e+00
4.13025588e-01 -4.48024660e-01 5.86432457e-01 -4.06175733e-01
1.17989695e+00 -7.52099574e-01 -1.69927925e-01 7.28332698e-01
-2.49625936e-01 1.98347822e-01 8.54934990e-01 -4.00843620e-01
-8.38582456e-01 -8.62475753e-01 -1.15973152e-01 -7.58965671e-01
6.68352619e-02 4.54521567e-01 1.69207126e-01 6.10564530e-01
5.73487759e-01 4.92309332e-01 -3.55872363e-01 -7.16777980e-01
4.37622756e-01 1.04240370e+00 3.07939112e-01 -6.52287006e-01
6.02237523e-01 1.43190309e-01 -6.96061671e-01 -5.53585768e-01
-8.36803257e-01 -1.02538955e+00 -4.70529467e-01 2.75994301e-01
7.32227981e-01 -1.10738218e+00 -1.25498921e-01 -1.55606763e-02
-1.28319573e+00 -9.46063921e-02 -4.32180732e-01 3.57158244e-01
-4.80788529e-01 3.40167403e-01 -9.25870895e-01 -4.79151011e-01
-8.94829035e-01 -1.08982348e+00 1.54029620e+00 1.21495068e-01
-1.54473051e-01 -1.00114942e+00 5.56467593e-01 5.03465533e-01
7.55417228e-01 -5.60642719e-01 9.50311959e-01 -1.18035710e+00
-4.93439168e-01 -7.17859566e-01 -4.21832860e-01 3.49326521e-01
-8.01277384e-02 -1.94820598e-01 -1.20733202e+00 -5.85575461e-01
-2.34740198e-01 -7.27585375e-01 1.09266734e+00 -8.42767730e-02
1.07964969e+00 -3.90548021e-01 -1.51551262e-01 4.72679883e-01
1.54723215e+00 -3.45000066e-02 6.09908104e-01 4.24399197e-01
2.83800066e-01 4.03106362e-01 2.78564036e-01 2.90029317e-01
2.10427061e-01 5.86493909e-01 -1.10607408e-01 1.49882630e-01
-1.03062734e-01 -4.39394951e-01 3.35828900e-01 1.31617451e+00
3.09992403e-01 -7.37579167e-01 -5.32332242e-01 8.36456537e-01
-1.59307587e+00 -1.05666316e+00 4.74951953e-01 1.93097699e+00
9.55319405e-01 -7.47602955e-02 -4.93559241e-02 -1.28102124e-01
2.65153080e-01 3.85080010e-01 -3.82891446e-01 -3.45692217e-01
-5.65918759e-02 2.49716640e-01 -2.92845685e-02 5.44630885e-01
-1.10938513e+00 5.64311266e-01 5.89526367e+00 1.22557259e+00
-9.30901527e-01 6.45541102e-02 3.33144844e-01 -5.81836283e-01
-4.53675002e-01 -1.82754382e-01 -8.36248815e-01 4.40644294e-01
1.05343246e+00 -3.94207001e-01 6.29113615e-02 5.70592046e-01
-2.74207503e-01 4.74775523e-01 -1.30652428e+00 8.00697505e-01
2.87430942e-01 -1.30573988e+00 5.20885706e-01 -2.25466564e-01
7.74183989e-01 3.11434120e-01 2.33301595e-01 9.70009148e-01
-2.44506836e-01 -7.07161605e-01 5.96281528e-01 5.79453170e-01
7.06729770e-01 -4.30778652e-01 8.50738764e-01 2.47668952e-01
-1.01433468e+00 -1.15758076e-01 -7.53998697e-01 2.29492843e-01
1.09571517e-01 -7.56811276e-02 -5.13808310e-01 9.09849763e-01
5.80860198e-01 6.46633863e-01 -5.87814808e-01 1.08164310e+00
2.29678825e-01 3.18632066e-01 -2.03975752e-01 6.30375817e-02
5.99338889e-01 -2.33941481e-01 4.90370870e-01 1.56488812e+00
7.86334276e-02 -1.20709427e-01 -1.35218158e-01 9.38195407e-01
-2.14032009e-01 2.71779120e-01 -3.48975688e-01 -4.45973605e-01
1.75436720e-01 9.12580788e-01 -1.09708689e-01 -5.64306796e-01
-5.96890748e-01 1.27234435e+00 5.39148211e-01 7.87728548e-01
-7.62506485e-01 -7.52259374e-01 7.24707723e-01 -6.23374879e-02
6.30463660e-01 1.66120663e-01 1.77898407e-01 -1.44002187e+00
2.19385982e-01 -7.07635462e-01 5.32213211e-01 -6.35196328e-01
-1.51168728e+00 9.36569750e-01 -4.84778136e-02 -1.47389126e+00
-4.01513755e-01 -2.54092872e-01 -5.90979755e-01 1.26492786e+00
-1.99134529e+00 -9.32139039e-01 4.16873954e-02 5.72351992e-01
9.80199099e-01 -9.95449126e-02 1.10747600e+00 7.68233418e-01
-7.59863973e-01 1.05675578e+00 6.87064648e-01 5.06120145e-01
8.60383689e-01 -1.17585015e+00 2.20578298e-01 5.24533927e-01
3.22911114e-01 1.05783832e+00 4.24566388e-01 -5.43382317e-02
-1.21436274e+00 -8.06224108e-01 9.93560195e-01 -5.00345767e-01
6.87405586e-01 -3.99940044e-01 -1.27437186e+00 5.30049562e-01
7.75256395e-01 -3.88023257e-02 1.00867343e+00 2.03903373e-02
-5.05402148e-01 -7.32665136e-02 -5.05837679e-01 6.41338766e-01
5.03472745e-01 -1.30235004e+00 -8.90662491e-01 4.34537590e-01
1.02212024e+00 -2.12540790e-01 -9.53752339e-01 1.17616747e-02
6.21825278e-01 -7.65207052e-01 1.36471772e+00 -1.02289486e+00
7.53199518e-01 1.15071155e-01 -1.93786740e-01 -1.00949216e+00
-3.92380238e-01 -4.33789104e-01 -4.35570896e-01 1.08169997e+00
3.06817710e-01 -1.59882843e-01 5.23909569e-01 3.58155280e-01
-2.01945350e-01 -9.93606389e-01 -6.62877321e-01 -6.25253618e-01
3.89179915e-01 3.00179124e-02 5.50327361e-01 5.90766311e-01
1.98285207e-01 7.99867749e-01 -4.58452031e-02 -2.77655367e-02
2.21823722e-01 4.00782287e-01 6.44349635e-01 -7.09590018e-01
-4.43017215e-01 -7.76572227e-01 -1.79650426e-01 -1.75516415e+00
1.83094963e-01 -1.05089319e+00 -1.06986642e-01 -1.64942825e+00
3.31658542e-01 -2.20507368e-01 -8.86733532e-01 -3.52901295e-02
-5.40450275e-01 1.53747141e-01 -1.06901355e-01 3.49737793e-01
-1.10688496e+00 1.11207819e+00 1.25542688e+00 -5.31229913e-01
2.12829863e-03 -1.75805986e-01 -5.75059474e-01 -8.58228356e-02
5.70631325e-01 -4.65005279e-01 -6.96783543e-01 -1.00749028e+00
3.38737726e-01 1.95231065e-01 2.36415491e-01 -8.23250353e-01
4.66823846e-01 3.65291923e-01 2.13454634e-01 -7.23360479e-01
5.40128469e-01 -7.84220755e-01 -4.25511718e-01 -2.68360488e-02
-9.04754698e-01 8.01115707e-02 3.63400817e-01 7.19184697e-01
-8.16651821e-01 -4.98209357e-01 1.20671168e-01 -6.64256811e-02
-2.14758962e-01 4.13649946e-01 -1.76602989e-01 3.92320096e-01
3.31588864e-01 -5.65285459e-02 -2.76257575e-01 -3.75990868e-01
-5.27623892e-01 5.44736266e-01 2.05205828e-01 6.30455256e-01
6.49636626e-01 -1.33986938e+00 -9.51254964e-01 3.53783786e-01
4.18750197e-01 2.12282427e-02 7.53531098e-01 6.25431001e-01
-5.07634580e-01 9.96844828e-01 -1.06726214e-01 -3.43059123e-01
-9.00672436e-01 6.57215714e-01 2.54375935e-01 -7.96719551e-01
-5.52346110e-01 1.08042490e+00 2.06317052e-01 -1.90754011e-01
3.19235146e-01 -1.80674955e-01 -3.22383225e-01 3.59307647e-01
7.90530920e-01 2.54714549e-01 2.04557851e-01 -4.71993148e-01
5.93062863e-02 4.59291428e-01 -7.21473634e-01 -6.30912706e-02
1.26844573e+00 -4.89818901e-01 -3.14689875e-02 3.27268481e-01
2.06222486e+00 -1.78076833e-01 -9.33325887e-01 -4.12655979e-01
-1.38530225e-01 -3.30288321e-01 2.58393921e-02 -3.51415515e-01
-7.18271673e-01 1.29237688e+00 1.55012429e-01 3.25633138e-01
1.10339105e+00 -1.35487109e-01 1.17395568e+00 6.62973106e-01
-5.39667672e-03 -9.60605800e-01 1.68110237e-01 4.67400402e-01
1.04264021e+00 -1.16361105e+00 -1.56331778e-01 2.23331615e-01
-4.02848125e-01 9.86802042e-01 5.15647411e-01 -4.89922792e-01
5.03901064e-01 -3.59669149e-01 3.36481072e-02 -2.03406498e-01
-1.09047270e+00 -8.81963670e-02 8.40845108e-01 -2.49271858e-02
4.77494001e-01 -3.29100132e-01 -1.82330027e-01 6.34986818e-01
2.11392082e-02 -3.51731092e-01 -1.02341197e-01 8.50439847e-01
-1.39663666e-01 -9.58999693e-01 -3.20709497e-02 4.22437876e-01
-5.21221578e-01 -3.71854037e-01 -4.98289727e-02 4.44792390e-01
-5.94448805e-01 6.75603151e-01 2.41814345e-01 -6.85374960e-02
2.61765659e-01 3.53460819e-01 4.87030119e-01 -6.14337087e-01
-9.22964633e-01 1.71689183e-01 -2.29417920e-01 -2.84006417e-01
-4.10504453e-02 -2.07321018e-01 -7.37928808e-01 2.42627576e-01
-5.66895008e-01 8.06438208e-01 1.98310718e-01 7.69003212e-01
5.89233637e-01 6.73164904e-01 6.47917211e-01 -5.96321702e-01
-1.01176775e+00 -1.26798940e+00 -4.15372044e-01 6.09692276e-01
7.53898084e-01 -3.57760489e-01 -5.27773499e-01 -2.00965956e-01] | [11.460837364196777, 7.698537349700928] |
d06d07d4-8549-421b-8db2-48a3f162a9cb | cheating-off-your-neighbors-improving | 2306.06078 | null | https://arxiv.org/abs/2306.06078v1 | https://arxiv.org/pdf/2306.06078v1.pdf | Cheating off your neighbors: Improving activity recognition through corroboration | Understanding the complexity of human activities solely through an individual's data can be challenging. However, in many situations, surrounding individuals are likely performing similar activities, while existing human activity recognition approaches focus almost exclusively on individual measurements and largely ignore the context of the activity. Consider two activities: attending a small group meeting and working at an office desk. From solely an individual's perspective, it can be difficult to differentiate between these activities as they may appear very similar, even though they are markedly different. Yet, by observing others nearby, it can be possible to distinguish between these activities. In this paper, we propose an approach to enhance the prediction accuracy of an individual's activities by incorporating insights from surrounding individuals. We have collected a real-world dataset from 20 participants with over 58 hours of data including activities such as attending lectures, having meetings, working in the office, and eating together. Compared to observing a single person in isolation, our proposed approach significantly improves accuracy. We regard this work as a first step in collaborative activity recognition, opening new possibilities for understanding human activity in group settings. | ['Christine Julien', 'Edison Thomaz', 'Evan King', 'Jingyi An', 'Haoxiang Yu'] | 2023-05-27 | null | null | null | null | ['activity-recognition', 'human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'computer-vision', 'time-series'] | [ 5.76330245e-01 -1.11257628e-01 -7.82384425e-02 -3.01628262e-01
-3.67036492e-01 -5.31722009e-01 6.71281695e-01 5.93937635e-01
-2.66543746e-01 5.44944048e-01 5.98068833e-01 1.31771013e-01
-1.94359928e-01 -7.28564382e-01 -2.21527100e-01 -7.43351698e-01
-1.82310387e-01 1.78302571e-01 1.70586929e-01 1.27498075e-01
5.93733713e-02 4.39462423e-01 -1.53111362e+00 2.31982619e-01
5.21657705e-01 5.91636837e-01 2.65526101e-02 8.54662716e-01
2.86531776e-01 1.02958715e+00 -1.09075832e+00 2.75575519e-01
2.09190026e-01 -6.66741848e-01 -7.09400833e-01 7.29256630e-01
5.58346152e-01 -2.80054361e-01 -2.64237136e-01 4.34948057e-01
3.66468340e-01 6.21678531e-01 3.02827299e-01 -1.22305465e+00
-1.87557563e-01 1.80293933e-01 -3.24489266e-01 3.43455523e-01
1.22336376e+00 5.66859953e-02 9.08873260e-01 -3.49919856e-01
1.51485562e-01 7.66813755e-01 7.23371804e-01 2.20116958e-01
-1.18346334e+00 -3.30995917e-01 4.66917843e-01 1.08775809e-01
-1.43158841e+00 -5.86763382e-01 7.91741610e-01 -4.57298785e-01
8.42160344e-01 4.72426623e-01 1.04794824e+00 1.18280292e+00
-2.57370472e-01 9.55118537e-01 8.45433712e-01 -4.59254622e-01
2.28040218e-01 -2.24367324e-02 3.62301558e-01 4.07384038e-01
3.38597059e-01 -3.27353984e-01 -9.34597731e-01 -5.58522701e-01
5.72200358e-01 7.10039437e-01 -4.17631745e-01 -4.17104393e-01
-1.79924381e+00 2.32311755e-01 1.68439597e-02 5.00479400e-01
-4.52834189e-01 -7.95250833e-02 -7.78431520e-02 1.17276050e-01
1.65866449e-01 3.06160778e-01 -1.11808397e-01 -6.97278619e-01
-8.04654896e-01 3.62726867e-01 1.12816501e+00 6.93754315e-01
6.97942376e-01 -4.82231081e-01 -1.87711075e-01 6.98932052e-01
1.24940068e-01 6.23882003e-02 2.74929941e-01 -1.04198217e+00
5.58875084e-01 8.20609987e-01 6.50933325e-01 -1.17656219e+00
-3.74152929e-01 -1.44410923e-01 -6.14670277e-01 -3.04887239e-02
1.01644206e+00 -4.65938747e-01 -3.85571688e-01 1.50867581e+00
4.43566591e-01 4.03436720e-01 -1.97517678e-01 7.96150923e-01
3.41408581e-01 2.65856922e-01 -1.00280933e-01 -3.70905042e-01
1.27578807e+00 -1.21339130e+00 -9.36771452e-01 -3.42924356e-01
7.71652997e-01 -4.95082796e-01 7.27351248e-01 4.57587123e-01
-9.92132246e-01 -7.84031808e-01 -9.54762757e-01 3.54404420e-01
-2.89186567e-01 -6.34824531e-03 7.29163289e-01 7.74133861e-01
-7.13491321e-01 8.26448262e-01 -1.25844717e+00 -9.51544046e-01
5.63058257e-02 2.99580634e-01 -5.16225934e-01 1.20905615e-01
-6.55721962e-01 5.39912522e-01 -1.61155909e-01 2.04946265e-01
-6.00333631e-01 -2.39948526e-01 -7.57305861e-01 -1.74036305e-02
6.92565620e-01 -4.01646733e-01 1.25057387e+00 -8.61833513e-01
-1.17394352e+00 4.61890370e-01 -5.15520275e-01 -1.34326428e-01
5.90607226e-01 -4.52558935e-01 -6.87510192e-01 -2.18641430e-01
-1.99331865e-02 -5.24925962e-02 3.22188586e-01 -8.83352101e-01
-7.92594969e-01 -5.76995313e-01 3.74996603e-01 5.43705523e-01
-4.34305519e-01 1.26921870e-02 -3.44967395e-01 -5.59238017e-01
1.51632354e-01 -1.07857049e+00 -8.07588845e-02 1.01936236e-01
-1.86304510e-01 -1.14362724e-01 8.14097762e-01 -5.53092480e-01
1.47546840e+00 -2.18510675e+00 -1.35520458e-01 2.85364866e-01
2.82356441e-01 6.26845285e-02 2.31970772e-01 8.99451196e-01
3.28565568e-01 -9.72670391e-02 -5.06234653e-02 -5.03420174e-01
-1.18522502e-01 2.56290019e-01 3.61404628e-01 5.88370502e-01
-2.80414879e-01 6.30943358e-01 -1.06974244e+00 -3.85094762e-01
2.83979386e-01 4.18699145e-01 -3.31035435e-01 2.49526650e-01
4.36022490e-01 7.17813015e-01 -3.91170591e-01 7.02279627e-01
1.99191436e-01 -3.12872261e-01 5.91727078e-01 4.38234299e-01
1.85923036e-02 2.35666960e-01 -1.68120897e+00 1.54856443e+00
-1.59281611e-01 9.40923512e-01 6.82864040e-02 -1.24978435e+00
7.08890915e-01 5.01498699e-01 7.81644225e-01 -1.85899362e-01
-2.53893256e-01 -2.80187696e-01 1.42626375e-01 -4.50104326e-01
4.61711049e-01 2.62554526e-01 -4.87707593e-02 1.10209954e+00
-3.62997353e-01 6.35631979e-01 3.71190280e-01 -2.76620686e-02
1.66997993e+00 7.26292026e-04 8.23042989e-01 1.44247636e-01
3.82034689e-01 -3.96719128e-01 5.85507751e-01 1.02417254e+00
-8.86861265e-01 6.06128275e-01 1.77957907e-01 -5.06872535e-01
-3.77176911e-01 -1.21651185e+00 3.48039180e-01 1.21959865e+00
2.86749959e-01 -7.96431363e-01 -5.42203963e-01 -6.52088702e-01
-9.51272845e-02 4.54733931e-02 -7.11113095e-01 6.55044168e-02
-7.57649124e-01 -4.93272364e-01 4.82806861e-01 7.09003687e-01
7.55669177e-01 -8.09364021e-01 -7.66816139e-01 3.27431113e-01
-6.62061036e-01 -1.10091496e+00 -6.65670693e-01 -4.00313772e-02
-9.04713929e-01 -1.14464903e+00 -5.88093758e-01 -5.32569885e-01
7.90461063e-01 8.77662063e-01 1.05309629e+00 4.10901457e-01
-2.45670468e-01 7.44957268e-01 -3.16632032e-01 -3.96414816e-01
1.26938060e-01 -8.06194916e-02 3.20514321e-01 4.96109813e-01
7.40643024e-01 -8.52654040e-01 -8.13365698e-01 7.92349279e-01
-2.93122828e-01 -2.46648833e-01 2.31446415e-01 3.00494760e-01
1.61829263e-01 3.86883378e-01 5.06289527e-02 -7.79627800e-01
5.13166428e-01 -5.91634154e-01 2.97655791e-01 3.66225690e-01
-3.47362936e-01 -5.01913488e-01 3.47362399e-01 -7.97434092e-01
-1.07513893e+00 1.22004576e-01 5.28549016e-01 1.67957410e-01
-7.95895696e-01 1.69737428e-01 -3.07619631e-01 1.82465702e-01
6.59680605e-01 1.73214659e-01 -2.31330961e-01 -5.05842745e-01
-3.98589730e-01 8.12679112e-01 3.17817986e-01 -5.02044678e-01
6.31488383e-01 7.36100852e-01 -2.62050241e-01 -1.26528597e+00
-7.90526211e-01 -9.93743777e-01 -9.79983509e-01 -4.23153311e-01
6.40834749e-01 -8.70337665e-01 -9.95223582e-01 5.25155544e-01
-6.12877607e-01 -4.65458155e-01 -1.02807052e-01 5.46858251e-01
-1.99034154e-01 6.98469698e-01 -1.78332418e-01 -1.10156488e+00
3.82667303e-01 -4.99826223e-01 9.88493502e-01 3.50799769e-01
-1.03092647e+00 -1.32551515e+00 2.68539727e-01 1.07393241e+00
1.38109997e-01 4.71393943e-01 -1.47918940e-01 -7.17664063e-01
-3.65092546e-01 -4.46570694e-01 3.84472936e-01 2.01798659e-02
1.00690854e+00 -9.38341841e-02 -8.68197978e-01 -3.93623799e-01
-2.41466746e-01 1.41803876e-01 1.67770594e-01 1.24752231e-01
8.90429199e-01 -2.11223483e-01 -6.62258625e-01 1.27109215e-01
7.14016616e-01 3.30200553e-01 4.94081706e-01 1.60667330e-01
6.94773376e-01 7.39655972e-01 6.55879319e-01 8.49223375e-01
4.47345018e-01 9.73500371e-01 -6.68157786e-02 4.06877585e-02
1.30757838e-01 -2.41003901e-01 5.85004687e-01 3.08742046e-01
-7.46558130e-01 -4.68073159e-01 -8.49557459e-01 4.77204114e-01
-2.12576580e+00 -1.45428002e+00 -7.74031729e-02 2.41498208e+00
2.53905416e-01 -2.95057260e-02 5.05104840e-01 4.69684303e-01
6.64176166e-01 4.93848175e-01 -3.00732374e-01 1.44674644e-01
2.01318488e-01 -2.86119580e-01 7.33344108e-02 3.05834115e-01
-1.05161321e+00 8.03031325e-02 6.64011621e+00 2.46676713e-01
-4.98565406e-01 -7.02616423e-02 1.47404954e-01 -6.17996395e-01
2.91349590e-01 2.77850069e-02 -8.30200911e-01 5.64732909e-01
8.86840522e-01 1.45202771e-01 5.27889550e-01 6.38459206e-01
5.04701078e-01 -6.01006687e-01 -1.61819518e+00 1.11728525e+00
2.95344234e-01 -6.96887732e-01 -4.07002032e-01 5.18008173e-01
5.22958696e-01 -4.68332082e-01 -4.66636479e-01 2.90312797e-01
6.63853064e-02 -9.52215135e-01 4.08375561e-01 6.03832603e-01
-1.94801122e-01 -4.79870319e-01 3.60340595e-01 9.33463514e-01
-1.39757228e+00 -1.47826076e-01 3.77520233e-01 -1.00197315e+00
2.79434204e-01 3.31144422e-01 -7.64742732e-01 1.52243495e-01
9.08156395e-01 8.08489203e-01 -5.78236282e-01 6.80701733e-01
7.53596053e-02 4.91224110e-01 -3.61741573e-01 -1.58613145e-01
-1.60762802e-01 -3.73923451e-01 3.65142882e-01 1.00536585e+00
3.92085612e-01 4.12602961e-01 6.73493147e-01 3.44871551e-01
1.61185652e-01 -2.44638205e-01 -8.02057862e-01 -3.05608839e-01
3.93692583e-01 1.06036556e+00 -6.80720568e-01 -3.39052022e-01
-7.00368226e-01 1.20052254e+00 3.21650088e-01 4.43814069e-01
-8.01400781e-01 -3.99116039e-01 9.95677292e-01 4.50049847e-01
-5.79326972e-02 -6.12138569e-01 8.89056623e-02 -1.28760064e+00
5.29471517e-01 -8.38578045e-01 4.57586855e-01 -5.95193148e-01
-8.34300399e-01 -7.77597502e-02 -1.02798358e-01 -1.55988348e+00
-2.62491196e-01 -1.87371209e-01 -8.54320705e-01 5.10740995e-01
-7.37084210e-01 -8.55166316e-01 -9.30781007e-01 7.10528195e-01
5.62053859e-01 1.50821596e-01 7.41703928e-01 2.69493312e-01
-6.17267609e-01 4.67174113e-01 -2.49907542e-02 3.61342996e-01
8.12728167e-01 -1.17536068e+00 1.03883080e-01 9.47762489e-01
6.91744149e-01 1.05695844e+00 5.90720534e-01 -5.40326715e-01
-1.18080854e+00 -5.88444471e-01 1.19679296e+00 -9.59640682e-01
4.38931286e-01 -5.65734744e-01 -9.04620349e-01 9.23662305e-01
4.16198373e-03 8.00116882e-02 1.46292734e+00 5.49076438e-01
8.39117393e-02 6.59656823e-02 -1.00942028e+00 7.76078522e-01
1.65380263e+00 -5.54004252e-01 -7.03431010e-01 3.99960756e-01
-1.00431517e-01 -2.18363166e-01 -1.07682753e+00 8.96912664e-02
9.39382017e-01 -1.25443602e+00 1.01320982e+00 -4.38261330e-01
-1.20344445e-01 -4.60776240e-01 -5.18936813e-02 -1.24391377e+00
-2.90377975e-01 -8.22887301e-01 -5.44804454e-01 1.12048459e+00
-9.68899280e-02 -6.17984116e-01 1.16966784e+00 1.07538486e+00
2.75187939e-01 -4.19036061e-01 -4.96448904e-01 -8.86422157e-01
-7.79810965e-01 -4.18837637e-01 6.78759277e-01 9.17259037e-01
5.26617289e-01 9.14990455e-02 -5.88913739e-01 2.40638927e-01
4.71092522e-01 -4.51290458e-02 1.35842276e+00 -1.50695467e+00
-7.55927563e-01 -1.24912942e-02 -4.74761277e-01 -1.41139448e+00
-2.29771167e-01 -2.13443965e-01 -1.45491868e-01 -1.50932097e+00
3.40599775e-01 -2.19828859e-01 -3.25352311e-01 4.58502650e-01
-1.06820337e-01 1.50326893e-01 -4.28612493e-02 3.69395941e-01
-8.41683686e-01 8.36035889e-03 9.78792727e-01 -5.97545989e-02
-4.60961670e-01 4.99911457e-01 -7.29053736e-01 8.93683910e-01
7.32083976e-01 -3.84774096e-02 -4.21767473e-01 -2.01217741e-01
-1.46571994e-01 5.50597794e-02 3.89250964e-01 -1.39657426e+00
4.68735367e-01 -4.85264659e-01 4.23091978e-01 -2.49673173e-01
6.57142401e-01 -9.54545557e-01 4.47123230e-01 4.40827370e-01
-2.46674329e-01 -2.76459217e-01 -3.54365826e-01 1.01539290e+00
-2.98909962e-01 8.59242752e-02 3.21448207e-01 -4.13875610e-01
-5.75636148e-01 2.12256700e-01 -6.79233074e-01 -2.28135541e-01
1.35433447e+00 -9.49923992e-01 -2.31544495e-01 -6.94706321e-01
-1.12307060e+00 1.98285431e-01 5.97923338e-01 4.70614046e-01
2.58270741e-01 -1.31801343e+00 -2.62393087e-01 3.36513191e-01
4.66871858e-01 -2.67582864e-01 3.05349976e-01 1.00755596e+00
-2.33059943e-01 2.15086818e-01 -5.72936982e-02 -5.41281939e-01
-1.82202792e+00 1.77684247e-01 2.15121165e-01 -2.74881929e-01
-4.59823966e-01 3.60637426e-01 -3.84390466e-02 -4.33066666e-01
5.50844848e-01 -3.23335648e-01 -3.54954243e-01 1.96013749e-01
8.08683813e-01 9.56662834e-01 -7.63376132e-02 -5.73261797e-01
-4.34029490e-01 4.53105986e-01 1.82069287e-01 -4.87321690e-02
9.77369726e-01 -4.03177172e-01 3.15520078e-01 8.10545504e-01
8.33164871e-01 2.50864774e-01 -1.24933016e+00 -4.50540662e-01
4.25578505e-02 -8.12070429e-01 -4.54410464e-01 -4.63238895e-01
-5.61488330e-01 7.48343766e-01 3.74539196e-01 5.29854000e-01
9.18160021e-01 -1.37802020e-01 7.03058302e-01 8.21193755e-01
6.69848680e-01 -1.36557353e+00 4.18791145e-01 1.82098404e-01
3.58190536e-01 -1.54957247e+00 1.75504059e-01 -3.63415748e-01
-5.33043146e-01 8.33291709e-01 5.70483744e-01 2.10910589e-01
5.99406362e-01 -1.29760161e-01 -6.64372668e-02 -4.79169041e-02
-7.53312349e-01 -1.50344118e-01 1.21512808e-01 7.86359549e-01
6.99887991e-01 2.25098193e-01 5.25305280e-03 2.45059535e-01
-1.16677759e-02 -3.14477570e-02 7.15365589e-01 1.51869512e+00
-4.10928309e-01 -1.00144994e+00 -6.30937278e-01 5.69512367e-01
-4.36459959e-01 5.16808093e-01 -6.04719877e-01 6.63881063e-01
1.34457186e-01 1.47954571e+00 2.13077649e-01 -3.02391142e-01
5.56521416e-01 1.65844619e-01 4.87647265e-01 -8.96525323e-01
-6.17843926e-01 -2.29659051e-01 3.73330384e-01 -7.29400516e-01
-8.66636395e-01 -1.21382463e+00 -8.45154583e-01 -5.93042970e-01
1.07871212e-01 1.31422326e-01 2.02874124e-01 1.04116833e+00
2.86700070e-01 2.97597796e-01 6.12609565e-01 -8.50391567e-01
-1.10619634e-01 -9.95153010e-01 -9.07227695e-01 6.87836945e-01
4.70195651e-01 -6.98913574e-01 -5.27414560e-01 4.01953518e-01] | [7.9137282371521, 0.5964123606681824] |
2c85cdf0-94f3-4048-a52f-46af5677ba08 | srcd-semantic-reasoning-with-compound-domains | 2307.01750 | null | https://arxiv.org/abs/2307.01750v2 | https://arxiv.org/pdf/2307.01750v2.pdf | SRCD: Semantic Reasoning with Compound Domains for Single-Domain Generalized Object Detection | This paper provides a novel framework for single-domain generalized object detection (i.e., Single-DGOD), where we are interested in learning and maintaining the semantic structures of self-augmented compound cross-domain samples to enhance the model's generalization ability. Different from DGOD trained on multiple source domains, Single-DGOD is far more challenging to generalize well to multiple target domains with only one single source domain. Existing methods mostly adopt a similar treatment from DGOD to learn domain-invariant features by decoupling or compressing the semantic space. However, there may have two potential limitations: 1) pseudo attribute-label correlation, due to extremely scarce single-domain data; and 2) the semantic structural information is usually ignored, i.e., we found the affinities of instance-level semantic relations in samples are crucial to model generalization. In this paper, we introduce Semantic Reasoning with Compound Domains (SRCD) for Single-DGOD. Specifically, our SRCD contains two main components, namely, the texture-based self-augmentation (TBSA) module, and the local-global semantic reasoning (LGSR) module. TBSA aims to eliminate the effects of irrelevant attributes associated with labels, such as light, shadow, color, etc., at the image level by a light-yet-efficient self-augmentation. Moreover, LGSR is used to further model the semantic relationships on instance features to uncover and maintain the intrinsic semantic structures. Extensive experiments on multiple benchmarks demonstrate the effectiveness of the proposed SRCD. | ['Song Guo', 'Xinghao Ding', 'Yue Huang', 'Luyao Tang', 'Jingcai Guo', 'Zhijie Rao'] | 2023-07-04 | null | null | null | null | ['object-detection'] | ['computer-vision'] | [ 3.32631111e-01 -5.69384806e-02 -3.69531244e-01 -3.87806654e-01
-5.41604280e-01 -3.28385741e-01 4.89833772e-01 4.38659862e-02
1.41155660e-01 5.99826515e-01 -1.18916407e-02 9.21971872e-02
-2.23307803e-01 -9.77400184e-01 -7.64919937e-01 -7.99377263e-01
3.05590034e-01 1.97614849e-01 5.38073421e-01 -2.54064649e-01
5.77116245e-03 3.77167642e-01 -1.72369230e+00 4.92448807e-01
1.13738108e+00 1.28470027e+00 3.12815905e-01 -2.68371046e-01
-3.85203451e-01 5.49865425e-01 -5.83981693e-01 -1.28271848e-01
2.86518961e-01 -4.79912728e-01 -6.00795209e-01 4.89383519e-01
4.62717772e-01 -2.05432251e-01 -3.24223489e-01 1.26135266e+00
1.66853040e-01 1.84322894e-01 5.98751068e-01 -1.51448965e+00
-9.05678511e-01 1.26006052e-01 -6.28230870e-01 -8.21052939e-02
7.85625055e-02 8.51686299e-02 8.32878828e-01 -9.62776721e-01
6.77839100e-01 1.58270693e+00 4.07181859e-01 5.57341814e-01
-1.32150495e+00 -7.76641667e-01 5.35485864e-01 3.23683858e-01
-1.40450740e+00 -9.01744962e-02 1.02641511e+00 -3.41265559e-01
5.08474648e-01 7.49244913e-03 3.52983087e-01 1.33250678e+00
-7.07950741e-02 9.12821233e-01 1.44981706e+00 -3.48700106e-01
4.66211826e-01 2.85184979e-01 1.63939074e-01 5.78167796e-01
3.68487239e-01 -5.55360503e-02 -4.72718477e-01 -7.57530332e-02
8.52265120e-01 1.05319522e-01 -9.96140391e-02 -8.59848499e-01
-1.09889281e+00 6.46640599e-01 5.35313249e-01 2.50560433e-01
-2.53632888e-02 -4.65563387e-01 4.46923792e-01 3.59652698e-01
6.22826338e-01 1.93010151e-01 -5.63408256e-01 4.05982465e-01
-3.79937559e-01 2.08043203e-01 3.30020547e-01 1.29231060e+00
1.09338915e+00 -7.39121437e-02 -1.91151232e-01 1.16500008e+00
-2.11074110e-02 5.21906912e-01 6.19228423e-01 -6.22058690e-01
4.86916035e-01 1.06830549e+00 -2.19722554e-01 -1.11944401e+00
-3.12863797e-01 -7.03474462e-01 -9.62736487e-01 1.31727755e-01
2.78608590e-01 2.71912128e-01 -9.77862656e-01 2.00785875e+00
4.86029685e-01 2.28601992e-01 1.77135020e-01 8.91203821e-01
9.37393010e-01 2.92821079e-01 1.08737722e-01 5.58886081e-02
1.31880558e+00 -1.05176997e+00 -5.62278330e-01 -5.10190308e-01
6.71393752e-01 -5.16578019e-01 1.40282702e+00 2.65387148e-01
-6.10413492e-01 -7.49867201e-01 -1.01627815e+00 -5.68276197e-02
-6.74793243e-01 2.21432820e-01 6.06700778e-01 4.28630888e-01
-5.18194020e-01 2.42938176e-01 -4.88515556e-01 -4.39263135e-01
7.36386716e-01 2.08967447e-01 -4.62689251e-01 -5.31492412e-01
-1.16556215e+00 4.68584180e-01 6.49499834e-01 -3.11004639e-01
-8.78506064e-01 -7.56316662e-01 -9.72809613e-01 -7.38890246e-02
6.99884593e-01 -5.12863219e-01 7.73791194e-01 -1.17597127e+00
-1.15004456e+00 9.29850340e-01 -2.62943894e-01 6.46716803e-02
2.60174066e-01 -3.17993127e-02 -6.49971247e-01 7.48849511e-02
4.80845481e-01 5.26253223e-01 9.28085446e-01 -1.53875768e+00
-6.47327840e-01 -7.24891007e-01 1.23255059e-01 2.01382026e-01
-7.52255440e-01 -3.39645833e-01 -6.18394792e-01 -1.03031063e+00
5.08944094e-01 -8.63785684e-01 6.93962052e-02 2.56859243e-01
-4.03992742e-01 -2.14092374e-01 1.12154782e+00 -4.99000609e-01
9.19503272e-01 -2.60661435e+00 4.86577190e-02 1.94392368e-01
1.12206355e-01 2.71254718e-01 -4.96534586e-01 7.90516788e-04
-1.42948285e-01 -2.10265577e-01 -3.90473098e-01 -1.65001377e-01
-7.95078501e-02 4.11366969e-01 -2.89127380e-01 1.10335365e-01
5.54820418e-01 8.29285860e-01 -8.91764283e-01 -4.88478124e-01
9.86014456e-02 2.49354571e-01 -3.61515224e-01 2.27318499e-02
-3.98483574e-01 4.37108636e-01 -7.71215618e-01 9.49560523e-01
1.18796253e+00 -3.47294897e-01 2.59621859e-01 -3.40227485e-01
2.76924074e-01 1.93739831e-01 -1.44858813e+00 1.80417752e+00
-4.14919943e-01 -1.46817714e-02 1.44901825e-02 -1.34224248e+00
1.27284265e+00 -2.14908779e-01 3.80831540e-01 -1.11628354e+00
-1.04059555e-01 3.79266858e-01 -3.12698811e-01 -2.31240943e-01
1.84280127e-01 -2.15148881e-01 -1.32649869e-01 1.95256963e-01
1.43122911e-01 2.74431348e-01 4.08341363e-03 1.28993496e-01
9.19482648e-01 1.92317814e-01 1.60857886e-01 -5.11374116e-01
6.62883818e-01 1.61850691e-01 9.31982219e-01 5.51012278e-01
-2.31972877e-02 5.46569347e-01 4.67454582e-01 -1.29731894e-01
-8.24916840e-01 -1.18827128e+00 -3.56207609e-01 1.03053236e+00
8.87325585e-01 -1.42564565e-01 -5.93345463e-01 -9.34921563e-01
2.86684692e-01 5.30262887e-01 -6.61872149e-01 -6.30781710e-01
-2.54799396e-01 -9.06199574e-01 2.72887409e-01 8.05212736e-01
9.92944121e-01 -7.25287020e-01 -1.59334857e-02 1.19873933e-01
-1.18498504e-01 -1.37421191e+00 -2.95903295e-01 1.43530518e-01
-9.66673434e-01 -9.94296372e-01 -6.00583851e-01 -9.20464694e-01
8.29963148e-01 7.10362077e-01 8.98273289e-01 -1.19847350e-01
-1.51965901e-01 2.09825486e-01 -5.49883842e-01 -2.68661112e-01
-1.67453051e-01 -7.21205920e-02 -5.73810078e-02 3.65099549e-01
5.15576541e-01 -4.54187691e-01 -4.32433158e-01 6.64098382e-01
-1.05035675e+00 9.40692574e-02 8.11539710e-01 1.10036373e+00
8.22700560e-01 3.63841593e-01 7.46656239e-01 -1.06893551e+00
3.16275924e-01 -5.72305679e-01 -4.09648269e-01 4.43043530e-01
-6.00168943e-01 -2.96773631e-02 7.06288874e-01 -6.50701880e-01
-1.42348289e+00 -9.81090441e-02 2.95256615e-01 -4.88159686e-01
-3.48312467e-01 2.01375112e-01 -8.84127021e-01 -7.64215738e-02
5.85005164e-01 4.53302681e-01 6.70090243e-02 -6.95342541e-01
1.12961873e-01 5.47573924e-01 4.68243510e-01 -9.58968997e-01
8.15788627e-01 7.63992548e-01 1.92163080e-01 -5.54537296e-01
-1.09990680e+00 -4.73030835e-01 -6.29217923e-01 2.77121872e-01
5.28794587e-01 -1.10457826e+00 -1.66606948e-01 7.79026628e-01
-8.07632089e-01 -2.42511943e-01 -4.29776788e-01 2.96770960e-01
-3.29290271e-01 4.35356498e-01 -3.62689972e-01 -3.97625118e-01
1.03638731e-01 -1.00145936e+00 1.17696190e+00 2.12677866e-01
2.63236433e-01 -1.00194931e+00 -4.43207562e-01 4.44888890e-01
1.91723749e-01 2.01541871e-01 1.25617933e+00 -7.37989545e-01
-5.62355876e-01 1.97639719e-01 -6.39785826e-01 7.69548118e-01
5.26689470e-01 -5.04532278e-01 -1.02197146e+00 -2.63011068e-01
3.48225385e-02 -3.99283230e-01 8.33261788e-01 6.79458901e-02
1.47589624e+00 -8.50510523e-02 -4.91080046e-01 5.91619074e-01
1.43527102e+00 1.46868691e-01 5.13056278e-01 4.94467556e-01
7.53342211e-01 7.43020415e-01 1.06004369e+00 4.35468465e-01
3.34150046e-01 7.68034279e-01 3.53351951e-01 -4.21229154e-01
-4.80326682e-01 -3.04366261e-01 3.08539718e-01 3.23961765e-01
2.18459740e-01 -6.70086741e-02 -6.87316835e-01 5.20771742e-01
-1.88369179e+00 -4.62129414e-01 -1.65745422e-01 2.14472675e+00
7.34704912e-01 9.05882195e-02 1.84580348e-02 8.84936303e-02
8.72745216e-01 -9.99080762e-02 -8.34514081e-01 1.33606210e-01
-6.85463190e-01 1.87789470e-01 2.27114916e-01 6.37815669e-02
-1.17099857e+00 9.22163844e-01 4.62001944e+00 1.21316445e+00
-9.70684409e-01 1.65603518e-01 4.78812963e-01 2.07496569e-01
-3.21623504e-01 1.38673196e-02 -9.20468152e-01 5.13157725e-01
-3.89441587e-02 1.15909122e-01 2.90681541e-01 1.01350653e+00
-2.50458747e-01 1.14745386e-02 -8.64322007e-01 7.44883060e-01
7.92407766e-02 -9.43056643e-01 3.83241951e-01 -2.29769778e-02
8.81187975e-01 -4.65195328e-01 1.38240844e-01 4.11209315e-01
9.28911269e-02 -5.00889301e-01 6.55638278e-01 1.84777796e-01
9.77981448e-01 -6.19392276e-01 6.65064812e-01 2.27392867e-01
-1.26887584e+00 -3.44452411e-01 -5.56179285e-01 2.10077107e-01
-4.20449048e-01 8.02209079e-01 -4.55819815e-01 9.72889245e-01
8.58628571e-01 9.78135049e-01 -7.79181302e-01 6.21541739e-01
-1.49691716e-01 2.16245592e-01 -1.84520692e-01 3.93743932e-01
1.53495565e-01 -1.35791957e-01 3.59743088e-01 8.84526730e-01
3.10111761e-01 -3.05823255e-02 3.22469115e-01 9.94464338e-01
8.89306515e-02 1.05532803e-01 -3.88707310e-01 2.28701904e-01
4.48573142e-01 1.01051581e+00 -6.06100500e-01 -3.71074617e-01
-7.56254494e-01 1.00694644e+00 1.72980353e-01 5.64015448e-01
-6.88879907e-01 -2.08881229e-01 8.04099858e-01 2.21346453e-01
4.25854653e-01 8.92039668e-03 -4.33232367e-01 -1.38979340e+00
3.68723333e-01 -1.02511930e+00 5.99001825e-01 -5.88026702e-01
-1.70816970e+00 3.46566856e-01 3.96549515e-02 -1.44519937e+00
4.43701357e-01 -7.52772868e-01 -2.40908399e-01 8.79291534e-01
-1.81011808e+00 -1.55279291e+00 -7.62991726e-01 1.07929122e+00
4.58552122e-01 -2.70290852e-01 7.16158688e-01 4.75159973e-01
-6.52698040e-01 7.68920720e-01 2.61177421e-01 -3.16860527e-03
8.16353858e-01 -1.03766525e+00 -8.99861455e-02 6.67063534e-01
-2.00240448e-01 5.36670446e-01 2.76092917e-01 -7.03930318e-01
-1.15633762e+00 -1.38121557e+00 4.12693799e-01 -8.62124562e-02
6.01935506e-01 -6.48942113e-01 -1.35528135e+00 5.85197985e-01
-5.28198659e-01 3.83635044e-01 4.27041829e-01 2.48896345e-01
-7.32760310e-01 -5.29518664e-01 -1.22182715e+00 4.29418564e-01
1.51816189e+00 -5.02126694e-01 -5.22438288e-01 3.25351238e-01
6.26744628e-01 -2.22988307e-01 -7.28122413e-01 7.54641056e-01
1.79116338e-01 -8.65120769e-01 1.01177371e+00 -4.99028265e-01
4.02462631e-01 -4.93888766e-01 -4.14068997e-01 -1.14909315e+00
-4.21244949e-01 7.59874284e-02 -2.63103587e-03 1.61946571e+00
-1.68696523e-03 -8.40212643e-01 7.15356171e-01 3.14611554e-01
-2.58611947e-01 -5.88965058e-01 -9.55999970e-01 -1.35061216e+00
1.29571661e-01 -1.16460890e-01 1.01851380e+00 1.15941620e+00
-4.71167684e-01 2.78104872e-01 -7.72655085e-02 3.82503301e-01
6.71934664e-01 5.32966793e-01 6.96896255e-01 -1.40249455e+00
-1.83131546e-01 -2.42567942e-01 -4.73433465e-01 -1.21308017e+00
2.82605171e-01 -9.90423381e-01 -1.89324975e-01 -1.16121280e+00
4.43093777e-01 -8.14726233e-01 -6.40511394e-01 6.19827926e-01
-1.96274176e-01 7.81016797e-02 -8.24136883e-02 3.35239321e-01
-5.86375058e-01 9.30363417e-01 1.62962878e+00 -3.12885582e-01
-1.11423060e-01 -1.20353751e-01 -9.98707652e-01 5.58358371e-01
6.75836802e-01 -4.41392750e-01 -7.34786093e-01 -3.51914287e-01
-2.97883719e-01 -4.19551820e-01 8.91330838e-01 -1.02475905e+00
-4.72546592e-02 -3.68600428e-01 3.60426873e-01 -3.26499552e-01
3.44032735e-01 -8.89939249e-01 -1.43146902e-01 2.25829050e-01
-1.15898170e-01 -5.45222044e-01 3.45299900e-01 7.24600613e-01
-5.58267713e-01 2.02462319e-02 9.37002957e-01 -3.66787575e-02
-1.21919227e+00 2.46078685e-01 1.75942630e-01 1.49071917e-01
1.11565399e+00 -3.77896905e-01 -5.74938297e-01 1.32522747e-01
-5.70882678e-01 8.93858969e-02 6.47239447e-01 6.90187097e-01
5.57910621e-01 -1.59729910e+00 -4.88045275e-01 4.76868749e-01
7.37352073e-01 2.75208414e-01 5.60592830e-01 6.14785254e-01
1.30120620e-01 2.52144694e-01 -2.92534322e-01 -7.59723663e-01
-1.01586962e+00 8.21855307e-01 1.95482507e-01 -1.26742706e-01
-7.46430695e-01 7.50113428e-01 1.02357531e+00 -6.27860248e-01
-2.91681290e-03 -1.76917866e-01 9.65112224e-02 -4.07072380e-02
3.02310973e-01 1.75866574e-01 1.74936041e-01 -4.75538522e-01
-4.85427469e-01 7.03683197e-01 -1.77605733e-01 4.10719246e-01
1.23749232e+00 -1.34188995e-01 -1.83269143e-01 2.49201030e-01
1.11157799e+00 -2.72765428e-01 -1.36303198e+00 -7.96548903e-01
-2.45263614e-02 -6.87439859e-01 3.79997157e-02 -8.97605479e-01
-1.05515456e+00 8.73086751e-01 5.90588689e-01 -2.00406253e-01
1.56404245e+00 2.45213389e-01 7.87871897e-01 1.86442509e-01
4.33924079e-01 -1.16643751e+00 4.47602600e-01 3.70346814e-01
7.69810200e-01 -1.24132574e+00 -5.96118569e-02 -9.85349000e-01
-6.63523078e-01 9.51336741e-01 1.13147545e+00 1.53379425e-01
5.55104911e-01 3.10621019e-02 -1.73427895e-01 -7.07164779e-02
-4.16767895e-01 -2.59084195e-01 3.57664943e-01 8.62689674e-01
-7.69541860e-02 4.89588939e-02 -1.98184401e-02 9.56275702e-01
3.56198221e-01 -8.42478941e-04 7.23536611e-02 8.99539173e-01
-3.23471427e-01 -1.23314488e+00 -4.20461088e-01 3.18315506e-01
6.61947429e-02 -1.27232494e-02 -3.60677510e-01 9.03027356e-01
5.80799937e-01 7.03527808e-01 5.50097451e-02 -4.52825934e-01
5.70450068e-01 -9.36910659e-02 3.43778610e-01 -7.86310196e-01
-1.36251021e-02 -7.55129978e-02 -1.99826345e-01 -5.37719965e-01
-4.14016813e-01 -6.18351698e-01 -1.18887520e+00 -9.26975086e-02
-1.68328270e-01 -2.37206861e-01 3.00929219e-01 8.76258194e-01
5.88868439e-01 5.36230326e-01 6.86233103e-01 -1.53218687e-01
-5.66166401e-01 -6.81062043e-01 -9.62732255e-01 9.03806210e-01
1.70312151e-01 -1.17983103e+00 -2.00282753e-01 -7.81268179e-02] | [10.075669288635254, 2.4394288063049316] |
386d0396-36df-4d1c-9c94-5cdcc60d0a9c | megacrn-meta-graph-convolutional-recurrent | 2212.05989 | null | https://arxiv.org/abs/2212.05989v2 | https://arxiv.org/pdf/2212.05989v2.pdf | MegaCRN: Meta-Graph Convolutional Recurrent Network for Spatio-Temporal Modeling | Spatio-temporal modeling as a canonical task of multivariate time series forecasting has been a significant research topic in AI community. To address the underlying heterogeneity and non-stationarity implied in the graph streams, in this study, we propose Spatio-Temporal Meta-Graph Learning as a novel Graph Structure Learning mechanism on spatio-temporal data. Specifically, we implement this idea into Meta-Graph Convolutional Recurrent Network (MegaCRN) by plugging the Meta-Graph Learner powered by a Meta-Node Bank into GCRN encoder-decoder. We conduct a comprehensive evaluation on two benchmark datasets (METR-LA and PEMS-BAY) and a large-scale spatio-temporal dataset that contains a variaty of non-stationary phenomena. Our model outperformed the state-of-the-arts to a large degree on all three datasets (over 27% MAE and 34% RMSE). Besides, through a series of qualitative evaluations, we demonstrate that our model can explicitly disentangle locations and time slots with different patterns and be robustly adaptive to different anomalous situations. Codes and datasets are available at https://github.com/deepkashiwa20/MegaCRN. | ['Shintaro Fukushima', 'Toyotaro Suzumura', 'Xuan Song', 'Yasumasa Kobayashi', 'Quanjun Chen', 'Puneet Jeph', 'Jiawei Yong', 'Zhaonan Wang', 'Renhe Jiang'] | 2022-12-12 | null | null | null | null | ['graph-structure-learning'] | ['graphs'] | [-1.68560833e-01 -1.87939167e-01 -3.73647839e-01 -1.65493235e-01
-5.79481542e-01 -4.43307817e-01 7.89424181e-01 2.57208496e-01
1.00580089e-01 4.42996621e-01 4.34785634e-01 -6.57538414e-01
-2.42224261e-01 -8.81442249e-01 -8.09686065e-01 -3.98724616e-01
-7.89804995e-01 2.09706485e-01 1.25999749e-01 -4.63839710e-01
-7.36557543e-02 3.21087748e-01 -1.09898782e+00 3.50123793e-01
7.39404917e-01 7.76142478e-01 -4.47675586e-02 8.80578518e-01
1.36552423e-01 1.24557412e+00 -4.23865944e-01 -2.88535297e-01
3.26380618e-02 -3.34272832e-01 -3.94035846e-01 2.16383506e-02
2.43051276e-02 -1.03729852e-02 -9.43539858e-01 6.16611242e-01
3.42289388e-01 1.26121238e-01 5.65459907e-01 -1.45629513e+00
-8.03865612e-01 7.97633171e-01 -8.00680637e-01 7.99937725e-01
3.29887420e-02 3.00707221e-01 1.11845458e+00 -5.26312470e-01
2.43517280e-01 1.18017352e+00 7.85891950e-01 -9.49843824e-02
-1.02771378e+00 -7.61871696e-01 5.31190336e-01 3.36793035e-01
-1.43660808e+00 -3.91360402e-01 1.03320098e+00 -4.29387391e-01
1.16357899e+00 2.14272693e-01 5.77237308e-01 1.38069618e+00
5.06413817e-01 6.95120215e-01 5.99022985e-01 1.04863457e-01
7.92098120e-02 -5.29098332e-01 -1.23652510e-01 5.28725028e-01
1.08947784e-01 1.25215173e-01 -4.55205679e-01 -1.86682567e-01
6.15570962e-01 3.99290979e-01 -1.08569413e-01 9.26710218e-02
-1.22555876e+00 7.19741464e-01 6.85868859e-01 4.58345830e-01
-5.62038481e-01 5.40144742e-01 4.76028770e-01 3.32550168e-01
1.16998661e+00 1.26523331e-01 -3.21089923e-01 -1.45494506e-01
-7.93430924e-01 1.19503893e-01 6.31769001e-01 9.66725051e-01
3.84421289e-01 5.65854907e-01 -1.41728178e-01 5.30524433e-01
2.42404684e-01 5.91033816e-01 2.72395670e-01 -4.11343396e-01
9.65861797e-01 6.59852147e-01 -1.91858605e-01 -1.62461066e+00
-8.24569583e-01 -9.32116807e-01 -1.28623426e+00 -4.95860726e-01
1.02314934e-01 -4.86685872e-01 -9.13660645e-01 1.65963840e+00
2.40302369e-01 1.16319203e+00 -2.47708708e-01 7.44301975e-01
7.86476731e-01 1.08746433e+00 -5.69692738e-02 -1.79885983e-01
8.97182286e-01 -9.55951631e-01 -6.22063994e-01 -2.27397189e-01
8.35584044e-01 -2.26245463e-01 7.33489752e-01 7.00234100e-02
-7.83499956e-01 -3.85001987e-01 -7.88990736e-01 2.22933680e-01
-4.95693177e-01 -1.00585401e-01 6.83541775e-01 4.16552387e-02
-1.05835819e+00 5.47632575e-01 -1.05559742e+00 -2.68144965e-01
3.65526229e-01 -5.36096655e-02 -9.30777565e-02 1.02059752e-01
-1.40695286e+00 4.44453984e-01 3.33437145e-01 2.66219914e-01
-9.02420819e-01 -9.43799615e-01 -8.42565238e-01 9.81024057e-02
5.24341404e-01 -6.34053290e-01 7.93837130e-01 -5.59250176e-01
-1.13090897e+00 3.18230778e-01 -1.85413942e-01 -7.29339480e-01
4.40946788e-01 5.15202358e-02 -1.17422688e+00 -6.58742562e-02
7.84272552e-02 -8.16686824e-02 7.54672647e-01 -8.96896541e-01
-5.55710196e-01 -3.14182431e-01 -1.71673298e-01 -8.38959366e-02
-1.77983999e-01 -3.39590937e-01 -4.62934583e-01 -1.06645262e+00
-2.73287874e-02 -8.11254740e-01 -3.16780150e-01 -8.02362740e-01
-5.58205664e-01 -2.37276599e-01 6.59006536e-01 -6.97142661e-01
1.84753788e+00 -2.09188819e+00 1.43299654e-01 2.87469149e-01
5.87231994e-01 8.63844156e-02 -3.22980523e-01 1.11066568e+00
-9.78988633e-02 2.29460880e-01 -2.48565599e-01 -4.87198740e-01
-1.19602196e-01 1.12333395e-01 -6.64122164e-01 7.74814427e-01
1.18413024e-01 1.25772846e+00 -1.10725486e+00 -1.02702126e-01
1.68301761e-01 6.07279480e-01 -2.56366611e-01 1.37591697e-02
-3.47746909e-01 5.57011783e-01 -4.54904228e-01 5.59259236e-01
3.93317789e-01 -9.38553095e-01 2.87619770e-01 1.22553647e-01
-7.60814361e-03 3.63345921e-01 -8.44636917e-01 1.54408324e+00
-3.13252062e-01 6.96508884e-01 -4.98606980e-01 -9.99698043e-01
8.24605286e-01 3.57013762e-01 7.72653580e-01 -9.34303284e-01
-5.10751233e-02 1.08113259e-01 -5.50888665e-02 -3.44378918e-01
6.56083167e-01 3.96617472e-01 -6.01544268e-02 4.36527401e-01
-1.67633727e-01 4.32176739e-01 6.78102970e-02 3.23975921e-01
1.29273129e+00 -3.42518270e-01 9.50575471e-02 -1.29182279e-01
2.37968951e-01 -9.84651223e-02 4.76949990e-01 6.51485622e-01
9.42404270e-02 4.83309656e-01 5.86457133e-01 -6.75751925e-01
-9.34767842e-01 -7.81193376e-01 3.33160579e-01 1.09824562e+00
-1.08184561e-01 -7.10301280e-01 -2.86826454e-02 -4.39186752e-01
2.25554064e-01 8.18079114e-01 -8.14346790e-01 -9.21958312e-02
-5.71549237e-01 -1.05260015e+00 6.50834799e-01 4.14477319e-01
3.01556855e-01 -8.24178576e-01 -2.08323970e-01 4.09815609e-01
-1.38234884e-01 -1.19874740e+00 -5.12652576e-01 -2.29594365e-01
-6.98117018e-01 -8.64732623e-01 -5.16528904e-01 -2.53653198e-01
2.82160431e-01 6.40116692e-01 1.24423790e+00 3.72371147e-03
2.11620685e-02 2.86606163e-01 -5.57074726e-01 -2.09023163e-01
-3.09584945e-01 6.50100410e-01 -2.07605287e-01 3.44787091e-01
1.93280086e-01 -1.12444937e+00 -7.57769763e-01 2.25261748e-01
-8.15168917e-01 3.13744724e-01 1.88243747e-01 4.00293499e-01
4.38075542e-01 1.92325652e-01 9.10742283e-01 -9.65718806e-01
7.02692330e-01 -1.33285880e+00 -6.95623934e-01 1.61587805e-01
-5.82233548e-01 -2.18070224e-01 7.97537029e-01 -3.04194152e-01
-4.46307212e-01 -4.21292275e-01 2.81696677e-01 -7.89422750e-01
7.77337253e-02 1.02135336e+00 2.51937181e-01 4.47299480e-01
4.34731603e-01 3.57252896e-01 -3.27684104e-01 -3.46760899e-01
4.26322639e-01 4.19164538e-01 4.17328417e-01 -1.40628964e-01
8.56822610e-01 5.06951213e-01 1.24563701e-01 -7.73551524e-01
-7.53633499e-01 -5.59958816e-01 -3.78708780e-01 -2.64543504e-01
6.62246704e-01 -1.23738337e+00 -5.93115449e-01 5.11491001e-01
-1.08061504e+00 -8.65746975e-01 6.46458641e-02 4.14188683e-01
-1.75018758e-01 1.56577617e-01 -5.86076856e-01 -7.89758682e-01
-3.37458491e-01 -6.08950734e-01 1.14484441e+00 -2.30586350e-01
8.72756541e-02 -1.61458158e+00 2.43206829e-01 -3.61063704e-02
5.90587437e-01 8.34725559e-01 7.37299263e-01 -5.76424956e-01
-5.57064712e-01 -2.44896308e-01 -3.02149087e-01 -3.00807357e-01
4.96839613e-01 2.31905784e-02 -8.92934442e-01 -5.44325352e-01
-4.17137355e-01 1.26494467e-01 1.11455011e+00 4.81397510e-01
1.26472032e+00 -5.95730722e-01 -4.13980782e-01 1.00644553e+00
1.16323197e+00 9.76462215e-02 3.86474997e-01 1.22800425e-01
1.36083055e+00 3.33742887e-01 1.37042016e-01 6.92755461e-01
1.10349751e+00 4.38145578e-01 6.35860801e-01 -1.06225163e-01
2.02657860e-02 -4.60014284e-01 2.70124912e-01 1.25629556e+00
-1.90947667e-01 -9.73983169e-01 -1.42238367e+00 8.06706190e-01
-2.30057549e+00 -1.16220069e+00 -4.85837877e-01 1.79141617e+00
2.30878264e-01 1.49582192e-01 4.64180827e-01 3.72663210e-03
6.01470530e-01 1.06012416e+00 -7.20222533e-01 3.58302938e-03
-2.71355242e-01 -3.20404947e-01 8.87685835e-01 4.63772714e-01
-1.21464574e+00 1.02204978e+00 5.68731546e+00 6.90017700e-01
-1.39932752e+00 5.12550473e-02 7.61436045e-01 -1.76117361e-01
-5.39197683e-01 -1.65702194e-01 -3.88553649e-01 6.66137755e-01
1.48264909e+00 -3.90454739e-01 5.75475216e-01 4.01821017e-01
4.14803505e-01 5.36177039e-01 -8.87770236e-01 9.53992486e-01
-1.23750657e-01 -1.72861826e+00 9.09890682e-02 1.19702064e-01
8.36854279e-01 8.67870688e-01 1.98967174e-01 2.82185316e-01
4.62310076e-01 -1.15305316e+00 4.52590436e-01 6.07307374e-01
6.48626089e-01 -7.49087989e-01 3.61919016e-01 3.90186876e-01
-1.70818889e+00 -4.26802263e-02 -1.66108444e-01 -1.06494211e-01
3.17330301e-01 7.30696380e-01 -7.24600554e-01 1.09757996e+00
5.82942188e-01 1.67566943e+00 -6.26571894e-01 7.48898566e-01
8.69631693e-02 1.06942475e+00 -2.07713306e-01 1.71627969e-01
4.15531516e-01 -1.37971565e-01 7.45240927e-01 1.31054413e+00
3.93740624e-01 1.06306970e-01 1.34707421e-01 7.06110418e-01
-2.36909851e-01 -8.44376832e-02 -8.75947714e-01 -3.41353744e-01
5.27143538e-01 1.01049042e+00 -4.50280398e-01 -2.85855383e-01
-5.35156012e-01 6.61476552e-01 3.64759594e-01 7.59033144e-01
-1.03479874e+00 -3.75981815e-02 4.24962848e-01 1.82747945e-01
3.71954978e-01 -5.55843234e-01 -4.69892956e-02 -1.44587421e+00
-9.20158401e-02 -8.54736924e-01 6.53668463e-01 -7.07705438e-01
-1.50884879e+00 8.43005419e-01 4.99997549e-02 -1.46361482e+00
-4.12705183e-01 -1.65614724e-01 -8.97820175e-01 6.84919953e-01
-1.68357933e+00 -1.35977232e+00 -3.09040129e-01 7.25612402e-01
2.94558316e-01 -3.56565118e-01 3.83533835e-01 5.12112677e-01
-9.04599369e-01 4.13716346e-01 1.04610831e-01 2.43305042e-01
2.83959568e-01 -1.05390978e+00 1.26719427e+00 1.08881366e+00
3.17963868e-01 1.99126035e-01 6.16997361e-01 -8.42576742e-01
-1.73041284e+00 -1.78105342e+00 7.66950309e-01 -3.75674605e-01
1.24309087e+00 -4.56946909e-01 -1.18251896e+00 9.93421078e-01
2.01122642e-01 2.53682852e-01 3.56086433e-01 1.58990145e-01
-4.10954118e-01 -1.58859089e-01 -3.99498105e-01 6.10702276e-01
1.25448489e+00 -5.99006474e-01 -3.31925154e-02 5.39747894e-01
1.16648769e+00 -4.77383226e-01 -9.06945407e-01 4.69369262e-01
1.63505971e-01 -6.92240834e-01 6.55118048e-01 -7.85782993e-01
3.81612480e-01 -1.51917219e-01 -2.00948074e-01 -1.53230524e+00
-6.31353021e-01 -8.83579910e-01 -5.19260526e-01 1.01331425e+00
5.70181072e-01 -9.39743280e-01 5.28086782e-01 8.05661678e-02
-2.40364671e-01 -7.40127981e-01 -8.61917973e-01 -7.94936895e-01
-6.94207549e-02 -8.67149055e-01 1.01701427e+00 1.28469217e+00
-1.90926418e-01 3.20899338e-01 -8.04277182e-01 6.10380650e-01
4.55728292e-01 2.77989417e-01 8.80927265e-01 -1.00510108e+00
-4.43966597e-01 -5.49174368e-01 -3.20282042e-01 -9.22896683e-01
2.27909878e-01 -1.05519557e+00 -6.06340826e-01 -1.52744162e+00
-3.37513715e-01 -3.14878851e-01 -5.74831963e-01 4.51440573e-01
-1.10618388e-02 -1.26612723e-01 -9.12601128e-02 3.45560700e-01
-8.11651289e-01 7.06385851e-01 1.09752607e+00 -2.09922910e-01
-3.15592736e-01 1.16939142e-01 -4.17780399e-01 2.87350684e-01
1.08990407e+00 -4.96595770e-01 -6.84759974e-01 -7.11169600e-01
6.79056227e-01 4.84000325e-01 5.22409201e-01 -7.94220388e-01
3.44772667e-01 -2.67875373e-01 -5.45802414e-02 -8.08241665e-01
1.43441036e-01 -7.24707365e-01 4.66406882e-01 2.52644688e-01
-2.65710592e-01 7.56581545e-01 3.50010902e-01 1.01529157e+00
-1.43205628e-01 9.45023656e-01 -2.27993838e-02 2.28271231e-01
-7.43096828e-01 8.72079253e-01 -1.90470129e-01 3.35680574e-01
8.74427676e-01 1.44236997e-01 -7.49328554e-01 -8.26212943e-01
-3.47022086e-01 6.50851369e-01 1.78364888e-01 7.24702835e-01
5.80878854e-01 -1.38645458e+00 -9.69994843e-01 1.42181456e-01
2.94376433e-01 -2.01550219e-02 4.98627543e-01 1.07673430e+00
-3.70305777e-01 4.75003213e-01 2.97903776e-01 -6.74504697e-01
-5.75169563e-01 6.15365505e-01 4.85440224e-01 -3.27592701e-01
-8.49752367e-01 3.48002613e-01 -3.50090526e-02 -3.03923100e-01
-2.52601050e-04 -6.94573939e-01 -7.56563842e-02 -1.04181226e-02
2.75057018e-01 5.53439915e-01 -1.15552517e-02 -8.02028060e-01
-4.91785765e-01 2.54231244e-01 3.04179311e-01 1.44462049e-01
1.76357615e+00 -2.49939397e-01 1.13575822e-02 1.03352392e+00
1.31472397e+00 -7.39480928e-02 -1.36758721e+00 -4.81282502e-01
1.63797975e-01 -2.02337876e-01 5.85498661e-02 -4.33601499e-01
-1.46904182e+00 6.95605695e-01 8.45805481e-02 8.32658768e-01
1.06373382e+00 3.00043169e-02 7.68571734e-01 1.36662632e-01
2.38592520e-01 -5.86411119e-01 -8.60877782e-02 5.32291114e-01
9.41282094e-01 -1.27721500e+00 -1.91305857e-02 -1.40970603e-01
-7.75472522e-01 8.47521842e-01 2.42768720e-01 -4.16725725e-01
1.15402400e+00 6.75721746e-03 -1.65527582e-01 -6.37179971e-01
-1.31878150e+00 -7.82534108e-02 7.80864954e-01 1.74992934e-01
4.26531017e-01 4.41266656e-01 2.83834875e-01 3.41777176e-01
-2.01861575e-01 -1.76222503e-01 3.56483012e-01 5.89821279e-01
5.58741689e-02 -5.50189435e-01 3.06814134e-01 4.86566722e-01
-4.19268161e-01 -1.10015161e-01 -2.37033721e-02 8.77170146e-01
-4.05998915e-01 1.22005916e+00 2.41674468e-01 -8.05482328e-01
2.93117881e-01 -4.52556670e-01 -1.95121616e-01 -3.95716459e-01
-5.75849473e-01 1.49341211e-01 5.30783981e-02 -7.49550879e-01
-4.48088348e-01 -5.30951381e-01 -9.91634011e-01 -6.67270303e-01
1.42710477e-01 -4.12888899e-02 2.06151515e-01 8.54772151e-01
8.78991127e-01 1.06568241e+00 8.72941017e-01 -6.31043375e-01
-2.85351705e-02 -9.51822579e-01 -6.95133328e-01 2.01882511e-01
7.64354348e-01 -3.78724664e-01 -4.15132940e-01 -3.72620344e-01] | [6.737337112426758, 2.699059247970581] |
3bd00cef-9210-4d32-8f80-245db6e93daf | learning-3d-human-pose-estimation-from-dozens | 2212.14474 | null | https://arxiv.org/abs/2212.14474v1 | https://arxiv.org/pdf/2212.14474v1.pdf | Learning 3D Human Pose Estimation from Dozens of Datasets using a Geometry-Aware Autoencoder to Bridge Between Skeleton Formats | Deep learning-based 3D human pose estimation performs best when trained on large amounts of labeled data, making combined learning from many datasets an important research direction. One obstacle to this endeavor are the different skeleton formats provided by different datasets, i.e., they do not label the same set of anatomical landmarks. There is little prior research on how to best supervise one model with such discrepant labels. We show that simply using separate output heads for different skeletons results in inconsistent depth estimates and insufficient information sharing across skeletons. As a remedy, we propose a novel affine-combining autoencoder (ACAE) method to perform dimensionality reduction on the number of landmarks. The discovered latent 3D points capture the redundancy among skeletons, enabling enhanced information sharing when used for consistency regularization. Our approach scales to an extreme multi-dataset regime, where we use 28 3D human pose datasets to supervise one model, which outperforms prior work on a range of benchmarks, including the challenging 3D Poses in the Wild (3DPW) dataset. Our code and models are available for research purposes. | ['Bastian Leibe', 'Alexander Hermans', 'István Sárándi'] | 2022-12-29 | null | null | null | null | ['3d-human-pose-estimation'] | ['computer-vision'] | [ 7.85622671e-02 1.67147309e-01 -4.41757143e-01 -4.42283422e-01
-9.05811191e-01 -4.75938946e-01 3.86333764e-01 -1.37573332e-01
-6.52015984e-01 4.57777023e-01 6.22368634e-01 1.87964261e-01
-1.16296090e-01 -3.97235006e-01 -7.65856266e-01 -6.31620467e-01
1.35483503e-01 9.25402164e-01 4.42101844e-02 1.51848635e-02
2.61015370e-02 4.40815479e-01 -1.25613487e+00 1.91955671e-01
1.75255209e-01 8.32689762e-01 -1.75933674e-01 3.92433465e-01
1.98514253e-01 3.82795662e-01 -4.60953534e-01 -4.99410391e-01
5.65222383e-01 -2.06165060e-01 -6.96814418e-01 6.53936639e-02
9.74376023e-01 -7.25452721e-01 -3.51196200e-01 6.35655642e-01
1.00612855e+00 -1.23470008e-01 7.62478292e-01 -1.16516232e+00
-1.20740034e-01 4.27680463e-01 -7.18469024e-01 -4.63199839e-02
1.87082440e-01 -3.16974558e-02 1.02468121e+00 -6.08240485e-01
8.06948423e-01 1.17620277e+00 8.82569313e-01 6.43685639e-01
-1.34325266e+00 -4.95163798e-01 -2.88188439e-02 -1.00772314e-01
-1.27902770e+00 -3.66069704e-01 8.60703707e-01 -4.28015769e-01
6.64936543e-01 1.03039816e-01 8.23413253e-01 1.48116422e+00
3.75767171e-01 7.73168445e-01 9.51305449e-01 -2.63477921e-01
2.74519950e-01 -4.23346758e-01 -1.10781696e-02 8.58894408e-01
3.63434732e-01 -4.26576333e-03 -8.32803488e-01 -2.59073824e-01
1.00555301e+00 2.33214960e-01 7.10880617e-03 -8.66492987e-01
-1.42810333e+00 8.56879652e-01 4.08383965e-01 -3.86879556e-02
-2.85182893e-01 1.82555825e-01 4.53301460e-01 2.18832567e-01
3.26478630e-01 3.74411136e-01 -7.47638285e-01 -4.50094510e-03
-8.03454876e-01 5.41772485e-01 6.79637969e-01 8.33597064e-01
6.57269657e-01 -4.76989150e-01 1.10199980e-01 9.09911513e-01
3.85684192e-01 2.84907728e-01 6.20467782e-01 -1.32559538e+00
4.10848677e-01 5.46801507e-01 -2.75771767e-01 -9.58453953e-01
-9.12118554e-01 -4.11979854e-01 -6.85949266e-01 3.40937853e-01
7.93738544e-01 -1.24690413e-01 -9.36117113e-01 1.90544558e+00
5.74687421e-01 -2.19776798e-02 -3.35413009e-01 1.05612707e+00
8.47956896e-01 -3.57170915e-03 -2.47422624e-02 3.80795896e-01
1.37637830e+00 -9.65414286e-01 -3.95053744e-01 -3.78515601e-01
8.32439125e-01 -5.15563667e-01 8.10008943e-01 4.55776542e-01
-9.46817756e-01 -3.80299658e-01 -9.52509582e-01 -3.99304479e-01
-1.86526582e-01 5.48121333e-03 8.41550529e-01 5.12392521e-01
-8.45375478e-01 6.00695848e-01 -1.31849635e+00 -2.12129667e-01
5.54817915e-01 5.98308206e-01 -8.74725938e-01 -1.02582537e-01
-7.86075056e-01 8.95259440e-01 2.48756036e-01 -1.59494635e-02
-7.04760909e-01 -6.61198914e-01 -9.42578256e-01 -5.26163340e-01
4.38146353e-01 -9.70179677e-01 1.05173278e+00 -4.06295031e-01
-1.30747914e+00 1.17217624e+00 9.08448398e-02 -3.04959357e-01
7.98573017e-01 -6.15570486e-01 1.61026627e-01 1.49860993e-01
1.91260397e-01 1.02396524e+00 8.00520003e-01 -1.26185632e+00
-1.86600149e-01 -1.02245545e+00 -2.47463062e-01 3.18835348e-01
-3.43992531e-01 -3.38460505e-01 -8.11596036e-01 -7.36633241e-01
6.78979218e-01 -1.23649919e+00 -2.55189627e-01 4.02244985e-01
-4.58028436e-01 -6.77440092e-02 5.63130379e-01 -6.27058506e-01
9.94470835e-01 -1.89724076e+00 6.71023667e-01 2.25320786e-01
5.17662704e-01 -1.94371432e-01 6.95340857e-02 1.07058808e-01
-6.49450645e-02 -1.40920311e-01 -2.64727563e-01 -8.35319877e-01
-2.10709870e-02 5.59989691e-01 2.52854079e-01 8.43665779e-01
-6.19530268e-02 8.58876050e-01 -6.59752965e-01 -8.08797061e-01
2.82272816e-01 5.97333550e-01 -8.53145301e-01 9.44193918e-03
1.01033248e-01 7.96669364e-01 -4.09226328e-01 8.45135450e-01
5.48878253e-01 -3.51169169e-01 8.49673674e-02 -4.54133779e-01
4.32313710e-01 8.50083306e-02 -1.22187650e+00 2.53960514e+00
-3.57253373e-01 1.75138026e-01 -1.47816166e-01 -6.89229488e-01
6.72642589e-01 1.67094871e-01 1.07059669e+00 -4.47148263e-01
3.25257897e-01 3.40336800e-01 -5.67915365e-02 -3.06871563e-01
4.68272120e-02 -1.15407988e-01 -2.03102276e-01 6.22447789e-01
3.96719515e-01 -3.69263217e-02 -1.52831197e-01 2.15038378e-02
1.31521297e+00 3.39223921e-01 2.95207262e-01 5.82223944e-02
-4.28982824e-02 -2.19117031e-01 6.44242346e-01 6.09424949e-01
-4.85403836e-01 1.05206597e+00 4.29112315e-01 -8.34446371e-01
-1.25264180e+00 -1.17329848e+00 -3.89096379e-01 1.00896740e+00
-5.59065007e-02 -5.91359913e-01 -7.62611985e-01 -8.74225914e-01
2.47848198e-01 6.77042466e-04 -8.71957421e-01 -4.72624972e-02
-6.65143192e-01 -6.36604071e-01 7.25995719e-01 7.12828994e-01
1.71869144e-01 -8.09932411e-01 -8.77730072e-01 4.18614931e-02
-1.66674495e-01 -1.01507390e+00 -2.40886450e-01 4.61267948e-01
-1.08108163e+00 -1.09674656e+00 -1.05953789e+00 -5.57188869e-01
6.61871433e-01 -1.47265583e-01 1.11343801e+00 -8.62598270e-02
-2.35984340e-01 5.30161858e-01 -1.90558940e-01 -2.44549423e-01
-9.37924013e-02 3.41593921e-01 1.72001764e-01 -1.90022424e-01
3.76716524e-01 -7.06407011e-01 -7.23882556e-01 4.97655332e-01
-6.96519792e-01 -2.67629456e-02 7.64860570e-01 9.85457599e-01
8.67032290e-01 -3.23542804e-01 1.75447717e-01 -8.68016243e-01
2.66884178e-01 -4.08766955e-01 -4.92245331e-02 5.35667315e-02
-3.56346399e-01 2.79909223e-01 2.54338264e-01 -3.81869704e-01
-7.51262009e-01 4.53215897e-01 -2.89221793e-01 -5.99689424e-01
-3.45651299e-01 1.88486382e-01 -1.27157703e-01 -1.13536678e-01
7.94136107e-01 -1.36387646e-01 4.42650437e-01 -7.06474125e-01
1.50947660e-01 3.96192193e-01 6.78404808e-01 -6.92811310e-01
5.32659054e-01 8.31324995e-01 3.61803770e-01 -4.82169926e-01
-9.55565572e-01 -4.97146219e-01 -1.14973617e+00 -9.57358852e-02
8.53700638e-01 -1.13137937e+00 -3.39449614e-01 4.29358363e-01
-9.42927361e-01 -2.13893056e-01 -1.23297393e-01 6.92897558e-01
-8.69753003e-01 4.64872330e-01 -6.48419440e-01 -2.42128447e-01
-2.19812602e-01 -1.39222956e+00 1.68471348e+00 -2.13264599e-01
-6.73552811e-01 -8.69096398e-01 4.57385838e-01 6.44214392e-01
-1.38370261e-01 5.79474568e-01 5.49231172e-01 -7.53236353e-01
-1.89109251e-01 -4.44910854e-01 2.42487773e-01 2.10823715e-01
-3.41203287e-02 -3.91044587e-01 -8.49603474e-01 -3.32609564e-01
1.35657182e-02 -7.36590266e-01 8.52304757e-01 5.16863823e-01
1.14370620e+00 5.17537296e-02 -2.79402345e-01 8.50883305e-01
1.03945541e+00 -5.28340280e-01 3.88313800e-01 3.61795276e-01
9.97271717e-01 7.37015367e-01 3.17624211e-01 5.05786717e-01
5.17678499e-01 8.67367446e-01 4.82080847e-01 -2.39368267e-02
-1.33903667e-01 -2.88686097e-01 7.27119893e-02 8.02469015e-01
-2.92892247e-01 2.32309371e-01 -1.00422180e+00 2.21465483e-01
-1.77537179e+00 -5.16079009e-01 2.39244193e-01 2.14483953e+00
7.68989861e-01 1.41509429e-01 2.57097185e-01 1.59090713e-01
3.64621848e-01 2.09361017e-01 -7.03007102e-01 1.64882705e-01
5.12270406e-02 2.66115874e-01 5.75890243e-01 1.77309662e-01
-1.32332921e+00 6.51241541e-01 6.19215822e+00 5.77252567e-01
-8.32347572e-01 1.42552003e-01 4.70992893e-01 -3.77617866e-01
-6.23853467e-02 -1.15924679e-01 -7.88416386e-01 4.07080948e-01
6.72408044e-01 4.96658534e-01 9.94757637e-02 9.83396232e-01
-4.96482179e-02 -6.38357103e-02 -1.26314652e+00 1.27584231e+00
2.88897395e-01 -9.80993211e-01 -6.93827420e-02 3.17038119e-01
5.74175596e-01 1.72238201e-01 4.85205986e-02 -2.62123160e-03
1.59263924e-01 -9.00948286e-01 6.07879698e-01 5.17802596e-01
5.49388647e-01 -6.76570177e-01 6.20272040e-01 3.74955595e-01
-8.45672965e-01 -2.20266525e-02 -3.86560977e-01 9.09346342e-02
1.72685370e-01 4.06627029e-01 -6.38094962e-01 3.80457252e-01
8.49474728e-01 7.66729712e-01 -7.00400174e-01 9.11796927e-01
-1.14097513e-01 1.81271568e-01 -7.05799580e-01 3.25781554e-01
-8.62316415e-02 2.00476930e-01 3.05588782e-01 7.51706183e-01
3.89958825e-03 -3.04367751e-01 1.97872281e-01 5.11651397e-01
-5.23442812e-02 8.58765468e-02 -5.22528648e-01 3.91693830e-01
2.37882003e-01 1.17171299e+00 -7.68599987e-01 1.18529657e-02
-4.99446720e-01 1.14598989e+00 4.66192216e-01 -4.72816750e-02
-7.39392936e-01 2.07226589e-01 4.74675208e-01 7.23996609e-02
-4.28802371e-02 -4.32941705e-01 -5.45233369e-01 -1.22577810e+00
1.91028848e-01 -8.70569468e-01 7.17338443e-01 -4.07078624e-01
-1.44005394e+00 1.57849908e-01 1.78699031e-01 -1.32172740e+00
-3.31178010e-01 -7.06286788e-01 -1.00873128e-01 4.69217539e-01
-1.07613707e+00 -1.25284433e+00 -3.30253005e-01 6.07599258e-01
3.60054404e-01 -9.16554257e-02 9.11872625e-01 4.40920770e-01
-4.50351655e-01 6.77787185e-01 -2.18338937e-01 3.24039191e-01
1.19387364e+00 -1.21011281e+00 3.35006654e-01 3.04689795e-01
5.35615206e-01 7.54963756e-01 4.44896609e-01 -6.37342095e-01
-1.57367575e+00 -6.45268440e-01 5.79991639e-01 -9.51493502e-01
2.93486804e-01 -3.44739079e-01 -6.38897598e-01 9.03654099e-01
-2.97029495e-01 3.63892257e-01 1.11768675e+00 3.72431040e-01
-4.35040832e-01 1.79382786e-01 -9.65927601e-01 5.75389206e-01
1.35250103e+00 -3.15217137e-01 -6.59084320e-01 2.94629693e-01
4.93603200e-01 -8.29263926e-01 -1.27540982e+00 5.51248610e-01
1.04201448e+00 -1.03667212e+00 1.22534311e+00 -6.05822444e-01
6.42790854e-01 -2.45633140e-01 -3.22217345e-01 -1.10876203e+00
-1.06088489e-01 -1.30416036e-01 -4.12948668e-01 6.39526963e-01
1.94646448e-01 -2.54901171e-01 1.22609282e+00 7.90130317e-01
-1.36790738e-01 -1.05201399e+00 -1.15311432e+00 -5.59708238e-01
2.73610532e-01 -4.33117539e-01 3.49530548e-01 8.43293607e-01
-3.89334381e-01 1.32863998e-01 -6.17874324e-01 -1.25600740e-01
8.34620535e-01 -1.99375957e-01 1.22764611e+00 -1.27783954e+00
-5.05179524e-01 -1.70309260e-01 -7.39526153e-01 -1.14853406e+00
2.28816986e-01 -9.71826851e-01 -2.78616905e-01 -1.19592535e+00
1.38203040e-01 -3.60868067e-01 -2.14805976e-01 7.70855010e-01
5.50005697e-02 4.97854054e-01 8.51278231e-02 4.00099188e-01
-7.14084566e-01 5.25688529e-01 1.30088055e+00 -9.61899234e-04
-8.81249979e-02 -6.75719604e-02 -5.88840246e-01 9.59660411e-01
5.62567234e-01 -5.38144529e-01 -3.05079341e-01 -7.31463373e-01
2.28902593e-01 -1.51902407e-01 5.74913919e-01 -1.28980339e+00
3.50127488e-01 1.83797479e-01 9.58275616e-01 -6.77563012e-01
4.97649640e-01 -9.72335041e-01 1.73991531e-01 2.90031374e-01
-3.71816218e-01 5.32233436e-03 -1.50745049e-01 6.89708769e-01
-1.88022666e-02 -2.19184496e-02 6.48678601e-01 -4.93429840e-01
-6.02904201e-01 6.03531539e-01 2.04754680e-01 2.09516749e-01
8.97283971e-01 -2.70606548e-01 1.08029835e-01 -2.49325559e-01
-1.04014504e+00 5.21895438e-02 6.80868506e-01 3.78186345e-01
5.12260973e-01 -1.47087991e+00 -6.37439489e-01 2.41248116e-01
2.61817783e-01 4.74885374e-01 4.10508424e-01 8.37654173e-01
-5.40456176e-01 1.36856109e-01 -5.11222243e-01 -1.01485491e+00
-1.01963401e+00 1.59411564e-01 3.43170106e-01 -3.14735919e-01
-8.78460586e-01 8.35311949e-01 -6.07305653e-02 -7.84276009e-01
4.10506725e-01 6.09451085e-02 9.79954675e-02 1.70653716e-01
3.77716541e-01 3.76296222e-01 2.61605740e-01 -7.57400632e-01
-4.08270597e-01 8.88839245e-01 -2.54771918e-01 -1.38357341e-01
1.50006020e+00 3.47655788e-02 1.38576493e-01 4.85115230e-01
1.63208771e+00 -6.99990243e-02 -1.52492011e+00 -2.95801342e-01
-2.35165462e-01 -4.96992350e-01 2.50347368e-02 -6.48744583e-01
-1.19583952e+00 7.41622031e-01 7.38075018e-01 -4.36646879e-01
7.99348116e-01 2.40977421e-01 8.58128250e-01 5.11869311e-01
6.66639924e-01 -1.26257598e+00 1.76995873e-01 2.83752531e-01
7.74795592e-01 -1.55446851e+00 4.63631988e-01 -1.18807785e-01
-7.91593194e-01 1.08395028e+00 6.04544759e-01 -1.94527730e-01
6.65369213e-01 3.27557683e-01 1.92857832e-01 -3.93124282e-01
-4.19142276e-01 -3.54898348e-03 2.88407624e-01 6.18851602e-01
3.91581982e-01 -2.93524266e-04 -2.26146564e-01 5.58277130e-01
-4.36921269e-01 -7.31026828e-02 2.83558629e-02 1.14250684e+00
-6.79494962e-02 -1.30255795e+00 -4.86303627e-01 5.69259524e-01
-5.18305182e-01 3.06083471e-01 -4.30167228e-01 7.87205994e-01
1.30184874e-01 2.80923337e-01 9.04984213e-03 -4.54618454e-01
2.29111999e-01 1.34260505e-01 7.24748552e-01 -6.13375604e-01
-3.00492913e-01 2.41712496e-01 -2.76772678e-01 -9.20982301e-01
-5.32418370e-01 -8.54102552e-01 -1.20668912e+00 -9.75762457e-02
-1.24701083e-01 -4.95592117e-01 6.65143609e-01 8.45293105e-01
4.07240421e-01 2.80482650e-01 2.05601737e-01 -1.13376415e+00
-6.69499397e-01 -7.16448247e-01 -6.40113473e-01 8.14146876e-01
2.12606177e-01 -1.31124067e+00 -3.90199065e-01 -9.93362889e-02] | [6.926694393157959, -0.9942705631256104] |
5d475081-4ffc-442d-82ad-a733d10474c4 | optimality-of-thompson-sampling-with | 2302.01544 | null | https://arxiv.org/abs/2302.01544v1 | https://arxiv.org/pdf/2302.01544v1.pdf | Optimality of Thompson Sampling with Noninformative Priors for Pareto Bandits | In the stochastic multi-armed bandit problem, a randomized probability matching policy called Thompson sampling (TS) has shown excellent performance in various reward models. In addition to the empirical performance, TS has been shown to achieve asymptotic problem-dependent lower bounds in several models. However, its optimality has been mainly addressed under light-tailed or one-parameter models that belong to exponential families. In this paper, we consider the optimality of TS for the Pareto model that has a heavy tail and is parameterized by two unknown parameters. Specifically, we discuss the optimality of TS with probability matching priors that include the Jeffreys prior and the reference priors. We first prove that TS with certain probability matching priors can achieve the optimal regret bound. Then, we show the suboptimality of TS with other priors, including the Jeffreys and the reference priors. Nevertheless, we find that TS with the Jeffreys and reference priors can achieve the asymptotic lower bound if one uses a truncation procedure. These results suggest carefully choosing noninformative priors to avoid suboptimality and show the effectiveness of truncation procedures in TS-based policies. | ['Masashi Sugiyama', 'Chao-Kai Chiang', 'Junya Honda', 'Jongyeong Lee'] | 2023-02-03 | null | null | null | null | ['thompson-sampling'] | ['methodology'] | [-7.62051195e-02 -1.96786504e-02 -7.55467594e-01 -1.06536247e-01
-1.10799479e+00 -6.25509202e-01 3.01000625e-01 -3.12456023e-02
-4.81049091e-01 1.12476993e+00 9.57120061e-02 -7.28335559e-01
-7.05281734e-01 -6.64070070e-01 -8.11754823e-01 -8.95049930e-01
1.43644303e-01 5.91588378e-01 6.63620383e-02 1.08690098e-01
2.90068626e-01 3.33246171e-01 -1.30725944e+00 -2.64327526e-01
9.23079848e-01 1.24260056e+00 2.79359281e-01 5.51923454e-01
7.00650886e-02 6.45402193e-01 -5.56361198e-01 -5.59695184e-01
5.97854257e-01 -2.28723392e-01 -4.70871180e-01 -2.62463003e-01
-2.25259751e-01 -6.04296625e-01 -7.85965323e-02 1.11482930e+00
4.34157908e-01 2.49675453e-01 8.25547814e-01 -1.30402505e+00
-4.65531081e-01 6.28010333e-01 -9.30481255e-01 2.96681166e-01
4.78666509e-03 -2.47800991e-01 1.20225930e+00 -1.36741191e-01
3.61646228e-02 1.37208724e+00 4.62048441e-01 2.90391922e-01
-1.01191306e+00 -4.59790707e-01 2.66307771e-01 1.59615040e-01
-1.18119097e+00 -1.91641867e-01 6.11920416e-01 -1.66956395e-01
2.91477084e-01 3.68116438e-01 2.88740724e-01 9.59992409e-01
1.13405049e-01 1.22502148e+00 1.29322636e+00 -6.55499101e-01
5.09438753e-01 1.56770408e-01 2.95021504e-01 1.27129391e-01
7.09329367e-01 3.90122294e-01 -1.55049145e-01 -7.50563264e-01
7.70205438e-01 5.05461842e-02 -1.93589136e-01 -2.93956876e-01
-5.32152951e-01 1.11720932e+00 -2.08126500e-01 -1.00884363e-01
-8.39534044e-01 5.04837036e-01 1.12931237e-01 1.36919752e-01
5.39819241e-01 6.53043166e-02 -4.87123013e-01 -2.03233227e-01
-8.95597398e-01 2.98602104e-01 9.37140584e-01 1.21950901e+00
3.19875717e-01 -4.25348058e-02 -9.89852846e-01 8.14156711e-01
2.40812048e-01 7.99918711e-01 -3.65955420e-02 -1.16068530e+00
6.95183635e-01 -3.96588922e-01 1.15247786e+00 -5.24124026e-01
-2.35881910e-01 -8.62623274e-01 -3.43780845e-01 -1.50347114e-01
8.68448973e-01 -3.68805349e-01 -5.46624541e-01 1.86869574e+00
2.34618068e-01 -1.57593384e-01 9.63485800e-03 8.62969339e-01
-1.16389342e-01 5.22106111e-01 -9.34724957e-02 -6.68605387e-01
1.20517242e+00 -7.09777296e-01 -9.11593676e-01 -1.17086820e-01
5.26185036e-01 -6.55933380e-01 1.02155900e+00 5.74846804e-01
-1.18837941e+00 2.29161069e-01 -7.46384025e-01 4.94646251e-01
3.85861069e-01 1.17659926e-01 4.96645540e-01 1.16858733e+00
-5.48951447e-01 5.26623309e-01 -6.46913946e-01 -1.45932794e-01
3.34359139e-01 3.00487597e-02 5.55776477e-01 -2.86660910e-01
-8.67751420e-01 9.14274633e-01 9.62279588e-02 -5.73386364e-02
-9.38469827e-01 -4.38464403e-01 -2.22236633e-01 4.38850224e-01
8.63637328e-01 -6.78211570e-01 1.61789072e+00 -7.15670586e-01
-1.78432202e+00 1.71190485e-01 -2.54090667e-01 -5.84580958e-01
8.42056572e-01 -3.15240115e-01 7.99350142e-02 2.13914961e-01
2.77587660e-02 -3.69858444e-01 7.95778394e-01 -1.15619278e+00
-7.98418283e-01 -2.77798802e-01 3.31622541e-01 2.02429876e-01
-2.34949633e-01 3.08761537e-01 -2.05231518e-01 -7.02958882e-01
-8.09766650e-02 -1.03793335e+00 -2.98674792e-01 -4.72453296e-01
-4.00108278e-01 -2.25721687e-01 7.13450238e-02 -3.90433341e-01
1.20509744e+00 -2.06706214e+00 -4.04938519e-01 2.84028530e-01
-4.66564357e-01 -1.78409696e-01 9.42113921e-02 4.44924861e-01
3.02654177e-01 1.32912129e-01 7.89879858e-02 -1.19324528e-01
3.99389148e-01 2.68702000e-01 -3.52381021e-01 7.96513259e-01
-5.15676141e-01 3.81869137e-01 -7.04289615e-01 -3.25296998e-01
-1.19556926e-01 -2.43680239e-01 -5.94527304e-01 4.90893982e-02
-2.64754802e-01 1.01609372e-01 -9.34712827e-01 4.24780369e-01
8.00073802e-01 -4.58959132e-01 1.31156236e-01 2.61456639e-01
-1.01740539e-01 -3.36038955e-02 -1.06759953e+00 7.84306228e-01
-4.79589611e-01 1.98789816e-02 5.33483684e-01 -1.11882687e+00
5.71331441e-01 2.78488636e-01 3.76789898e-01 -2.87036240e-01
3.98162574e-01 2.79010952e-01 -1.07466787e-01 -3.34683895e-01
4.63188082e-01 -5.85072517e-01 9.33056884e-03 7.17102468e-01
-3.41176122e-01 2.19041109e-02 6.62754253e-02 -7.95106590e-03
9.29872096e-01 -8.53079706e-02 4.21237677e-01 -4.76073444e-01
-6.06916938e-03 -3.25547874e-01 6.93120599e-01 1.42931604e+00
-3.26217651e-01 3.26932013e-01 7.09943473e-01 1.20031245e-01
-7.71793365e-01 -8.61500859e-01 -2.73143977e-01 1.36143208e+00
2.07083970e-01 1.86318949e-01 -3.23580176e-01 -5.20797253e-01
2.74742097e-01 1.38554609e+00 -6.51277959e-01 -4.73807864e-02
-1.63244065e-02 -1.16229928e+00 1.74352810e-01 2.91898996e-01
3.41365784e-01 -2.33935744e-01 -6.25416696e-01 3.09049368e-01
-2.78143764e-01 -9.09137666e-01 -5.16013205e-01 4.03099328e-01
-7.71768332e-01 -9.85276222e-01 -1.20468533e+00 2.12103307e-01
3.93251657e-01 4.11256462e-01 6.68602884e-01 -4.64498371e-01
4.75940794e-01 7.05188453e-01 -4.99408931e-01 -6.92565262e-01
-4.18007001e-02 -1.99853152e-01 -6.46185428e-02 1.43146724e-01
1.12020783e-01 -2.23018259e-01 -6.75955474e-01 5.48441291e-01
-7.91585565e-01 -4.07618016e-01 4.76931572e-01 8.77762318e-01
3.40816200e-01 -3.75915989e-02 7.54130542e-01 -5.87828636e-01
7.81633973e-01 -7.04134524e-01 -8.39685142e-01 6.08689785e-01
-5.02251923e-01 2.05685392e-01 5.00291586e-01 -5.44530153e-01
-1.38662338e+00 -4.45303947e-01 2.38156825e-01 -4.89849150e-01
1.40488282e-01 5.23680091e-01 3.82252559e-02 1.71628535e-01
4.31307256e-01 5.98437227e-02 -1.72521487e-01 -6.18182421e-01
1.14072226e-01 7.55478740e-01 3.50809805e-02 -1.22517133e+00
2.80418456e-01 5.07268369e-01 1.21999152e-01 -4.89967942e-01
-1.40856671e+00 -2.32207909e-01 2.91774869e-01 -1.34762749e-01
3.88532370e-01 -5.83448470e-01 -1.00806963e+00 -6.44837171e-02
-8.33437324e-01 -2.60899901e-01 -2.55232811e-01 1.01058161e+00
-1.22481656e+00 4.54999834e-01 -5.30707061e-01 -1.85967243e+00
-9.68678668e-02 -9.49168444e-01 5.60874641e-01 2.25839674e-01
1.44011676e-01 -7.74525881e-01 -1.91484496e-01 3.74967933e-01
3.97850275e-01 -5.58458120e-02 8.37311447e-01 -8.76161993e-01
-3.49174500e-01 -2.94454038e-01 -1.75116554e-01 2.26759598e-01
-9.30695161e-02 -3.26843798e-01 -4.83118951e-01 -4.71899956e-01
2.29045168e-01 -3.14141810e-02 7.46613503e-01 1.01747549e+00
1.29452491e+00 -5.39979815e-01 -2.97370315e-01 4.46639955e-01
1.29007220e+00 6.37591183e-01 1.57906085e-01 6.99026465e-01
-2.15995789e-01 5.69338918e-01 9.00956929e-01 1.19528556e+00
1.96210131e-01 4.71355766e-01 6.48453534e-01 4.94443506e-01
6.25030816e-01 -1.12859480e-01 2.06739828e-01 5.10323271e-02
-1.51061773e-01 -5.21108985e-01 -5.77170789e-01 6.79587722e-01
-2.25035334e+00 -1.03398991e+00 -7.24999830e-02 2.81946564e+00
7.77052104e-01 7.66165853e-02 3.28552932e-01 -1.20365180e-01
1.07279718e+00 -1.06121816e-01 -7.45941699e-01 -4.84508425e-01
-2.02506423e-01 -2.15864871e-02 1.24381709e+00 3.48730117e-01
-7.29532123e-01 5.98527551e-01 7.15356541e+00 1.33049238e+00
-3.43271106e-01 3.14305693e-01 6.41927660e-01 -4.16199356e-01
-3.76065254e-01 1.45800084e-01 -9.75788593e-01 7.52552450e-01
9.10653651e-01 -6.76985919e-01 4.23200041e-01 9.39053774e-01
4.63215500e-01 -4.86198515e-01 -8.86007667e-01 7.15170622e-01
-4.09474134e-01 -9.42454815e-01 -3.74833584e-01 2.68654019e-01
8.56070042e-01 -2.60512412e-01 1.69145912e-01 2.77615726e-01
1.01974344e+00 -6.07487500e-01 9.71441627e-01 6.38341188e-01
4.16505486e-01 -1.13447070e+00 8.42869937e-01 6.59477770e-01
-4.80007499e-01 -4.81682271e-01 -4.16029096e-01 -2.20138207e-01
2.76077241e-01 8.69261861e-01 -4.56643313e-01 7.64290631e-01
5.59771419e-01 4.69298773e-02 3.33315611e-01 1.58006215e+00
-1.02925643e-01 9.19224501e-01 -6.90439999e-01 -3.32902819e-01
4.12792295e-01 -4.72235709e-01 6.12281024e-01 8.56130421e-01
6.64969265e-01 2.00617835e-01 1.15084231e-01 5.58652997e-01
3.16420645e-01 1.05362378e-01 -7.85196722e-02 1.51537672e-01
6.94902241e-01 6.15279078e-01 -5.21860361e-01 -3.82468045e-01
-3.71625900e-01 3.32413346e-01 1.76290259e-01 7.51399994e-01
-9.61955369e-01 -2.39283547e-01 4.81770128e-01 -1.11199066e-01
7.04976737e-01 2.56319586e-02 -4.02991384e-01 -8.13437104e-01
1.49186524e-02 -3.66742522e-01 6.93763494e-01 -4.19936299e-01
-1.37525952e+00 -4.40466553e-02 5.97071767e-01 -9.29560900e-01
-1.90451220e-01 -4.04807121e-01 -4.48366433e-01 9.14640427e-01
-1.57149100e+00 -5.63237071e-01 4.96334881e-01 5.96947014e-01
7.17946067e-02 8.97788554e-02 2.54030168e-01 -6.50821030e-02
-6.62691236e-01 5.23126125e-01 1.07146192e+00 -4.86014813e-01
5.66675365e-01 -1.12112164e+00 -5.78039169e-01 5.48710585e-01
-5.47707856e-01 4.90604192e-01 1.29539108e+00 -4.50576842e-01
-1.19194853e+00 -7.54555583e-01 3.13316256e-01 1.80562317e-01
7.25928068e-01 1.83631942e-01 -4.55122054e-01 7.93911755e-01
1.02606721e-01 -1.73840955e-01 6.94887042e-01 2.78525144e-01
-8.81005675e-02 -6.74951896e-02 -1.38846362e+00 4.89630103e-01
6.95672572e-01 4.08605151e-02 -4.33197618e-01 6.43403590e-01
4.38374221e-01 -1.44644231e-01 -6.73638165e-01 3.76434594e-01
7.52142668e-01 -8.98524106e-01 5.98857403e-01 -7.41587400e-01
-3.05246767e-02 1.61838189e-01 -4.96183932e-01 -1.37752080e+00
-3.84603322e-01 -9.41072464e-01 -1.46703541e-01 1.02102923e+00
3.23448926e-01 -9.84375417e-01 7.28412688e-01 8.25049281e-01
1.27619982e-01 -7.27979064e-01 -1.29290950e+00 -1.49973440e+00
3.90808076e-01 -5.43374777e-01 5.79499960e-01 4.19967890e-01
-2.50130668e-02 -2.02163562e-01 -7.51287937e-01 1.03034340e-01
9.54409003e-01 4.11458105e-01 5.76678634e-01 -1.02345133e+00
-5.54011345e-01 -4.47969317e-01 5.02945840e-01 -1.62479007e+00
1.86186746e-01 -3.31089437e-01 1.05239987e-01 -1.45054400e+00
4.92533326e-01 -4.83255774e-01 -3.68532896e-01 1.84660122e-01
-1.46408379e-01 -5.41621506e-01 2.60839909e-01 2.21666932e-01
-5.90709984e-01 9.01068807e-01 1.27901864e+00 1.64102465e-01
-6.33983463e-02 7.48020709e-01 -8.81369948e-01 6.47716284e-01
9.15967941e-01 -6.90650165e-01 -2.79245347e-01 -1.61089137e-01
3.82846504e-01 6.37325943e-01 1.35826364e-01 -2.57954597e-01
5.70785366e-02 -7.54746556e-01 -1.48736760e-01 -9.18424070e-01
3.75621289e-01 -8.51666093e-01 5.12869470e-02 3.54605675e-01
-4.07489181e-01 -3.96921337e-01 -2.06053909e-02 1.13386405e+00
2.76535094e-01 -7.48111308e-01 8.08205724e-01 -1.26889169e-01
4.02241647e-01 1.81019291e-01 -6.16568327e-01 1.39855042e-01
1.05390310e+00 -8.65001902e-02 -3.55856180e-01 -9.39130962e-01
-7.51533628e-01 4.78571564e-01 6.79449365e-02 -1.56431258e-01
1.87363818e-01 -1.15398049e+00 -6.14697576e-01 -6.12265348e-01
-2.80059934e-01 -5.32722294e-01 2.30794385e-01 1.19674611e+00
1.37010008e-01 6.54392064e-01 2.38062009e-01 -3.07435215e-01
-8.60772789e-01 7.65339553e-01 1.62490875e-01 -4.70942616e-01
-1.10705368e-01 4.79919165e-01 2.22799569e-01 8.15597922e-02
3.49189341e-01 -2.50807405e-01 5.30871674e-02 -4.62196730e-02
3.70278984e-01 8.40913057e-01 -2.27901235e-01 -5.78273423e-02
-2.24498019e-01 2.14435130e-01 3.09910579e-03 -4.57901776e-01
1.18871737e+00 -4.17209566e-01 1.95624143e-01 4.10746485e-01
7.33811855e-01 9.42196175e-02 -1.38998806e+00 -4.90704328e-01
1.85252771e-01 -7.99349308e-01 2.76238233e-01 -7.39400566e-01
-8.59741271e-01 4.56722677e-01 5.97901680e-02 5.76769590e-01
8.97291660e-01 -1.22318864e-01 4.37813491e-01 2.18100026e-01
8.11142623e-01 -1.19982743e+00 -4.51489598e-01 5.51533937e-01
7.44190574e-01 -9.51792359e-01 1.37340307e-01 -2.51967371e-01
-6.22514307e-01 8.43675852e-01 7.42321238e-02 -1.50735244e-01
7.55667627e-01 6.64959550e-02 -5.78898728e-01 4.17916596e-01
-8.60630035e-01 -5.39581597e-01 1.19853765e-01 3.72018367e-01
3.43193039e-02 4.08002496e-01 -9.87178802e-01 1.06763780e+00
-1.90636531e-01 9.42092016e-03 6.02923453e-01 9.28290248e-01
-6.62532985e-01 -8.61652315e-01 -8.40744019e-01 7.76591957e-01
-1.09220040e+00 -1.28262341e-01 -1.76827479e-02 5.68814397e-01
-6.59859836e-01 1.38328254e+00 -1.93297952e-01 2.53502458e-01
2.18151078e-01 -8.94706994e-02 7.18311667e-01 -2.13595331e-01
-2.98733354e-01 5.00683665e-01 3.28116238e-01 -2.78944224e-01
-4.29400802e-01 -8.91216338e-01 -5.92408478e-01 -4.77940351e-01
-6.69883013e-01 5.24326384e-01 5.01306355e-01 1.07135940e+00
2.23256916e-01 1.02382265e-01 8.33098710e-01 -4.13465112e-01
-1.65445960e+00 -1.16386735e+00 -1.09372723e+00 -2.04634905e-01
2.67437279e-01 -1.03281891e+00 -7.61849225e-01 -7.27052271e-01] | [4.529558181762695, 3.2811362743377686] |
86a283d1-ebb8-41ef-919e-51b81a0eb083 | heterogeneous-temporal-graph-transformer-an | null | null | https://dl.acm.org/doi/abs/10.1145/3447548.3467168 | https://dl.acm.org/doi/pdf/10.1145/3447548.3467168 | heterogeneous temporal graph transformer: an intelligent system for evolving android malware detection | The explosive growth and increasing sophistication of Android malware call for new defensive techniques to protect mobile users against novel threats. To address this challenge, in this paper, we propose and develop an intelligent system named Dr.Droid to jointly model malware propagation and evolution for their detection at the first attempt. In Dr.Droid, we first exploit higher-level semantic and social relations within the ecosystem (e.g., app-market, app-developer, market-developer relations etc.) to characterize app propagation patterns; and then we present a structured heterogeneous graph to model the complex relations among different types of entities. To capture malware evolution, we further consider the temporal dependence and introduce a heterogeneous temporal graph to jointly model malware propagation and evolution by considering heterogeneous spatial dependencies with temporal dimensions. Afterwards, we propose a novel heterogeneous temporal graph transformer framework (denoted as HTGT) to integrate both spatial and temporal dependencies while preserving the heterogeneity to learn node representations for malware detection. Specifically, in our proposed HTGT, to preserve the heterogeneity, we devise a heterogeneous spatial transformer to derive heterogeneous attentions over each node and edge to learn dedicated representations for different types of entities and relations; to model temporal dependencies, we design a temporal transformer into the HTGT to attentively aggregate its historical sequences of a given node (e.g., app); the two transformers work in an iterative manner for representation learning. Promising experimental results based on the large-scale sample collections from anti-malware industry demonstrate the performance of Dr.Droid, by comparison with state-of-the-art baselines and popular mobile security products. | ['Qi Xiong', 'Yinming Mei', 'Kui Wang', 'Wenqiang Wan', 'Yanfang Ye', 'Shifu Hou', 'Mingxuan Ju', 'Yujie Fan'] | 2021-08-14 | null | null | null | kdd-2021-8 | ['android-malware-detection', 'mobile-security'] | ['miscellaneous', 'miscellaneous'] | [-5.97785041e-02 -3.82407218e-01 -5.59829175e-01 1.42074540e-01
-6.46818206e-02 -7.26953566e-01 7.46926308e-01 -7.09817559e-02
1.31159440e-01 2.64577180e-01 1.02140814e-01 -5.84904075e-01
-2.46510729e-01 -9.23160672e-01 -6.20423734e-01 -3.37816328e-01
-3.78381997e-01 5.34095243e-02 5.62617421e-01 -2.90041089e-01
-2.47351099e-02 3.57043028e-01 -1.19695425e+00 1.56496197e-01
8.93554151e-01 1.10555017e+00 7.48266429e-02 4.77369457e-01
-3.24162059e-02 1.05222666e+00 -5.93523800e-01 -7.34307826e-01
-4.41882089e-02 -2.63749748e-01 -6.57582760e-01 -1.40386209e-01
-1.37214541e-01 -1.79051161e-01 -6.79795027e-01 1.15736377e+00
1.03359856e-01 -1.27768815e-01 6.66657746e-01 -1.60675704e+00
-1.02576470e+00 7.57182360e-01 -9.60798323e-01 6.63665712e-01
1.94423750e-01 1.84690237e-01 1.10822880e+00 -3.79491031e-01
7.83535182e-01 1.11730921e+00 9.85649467e-01 5.31197488e-01
-9.51508939e-01 -7.15356648e-01 8.98499191e-01 4.33758199e-01
-1.35579169e+00 -1.27607854e-02 1.03849256e+00 -5.51091909e-01
8.50706875e-01 3.23421687e-01 6.40809059e-01 1.95065594e+00
5.02723694e-01 6.15401149e-01 5.71331084e-01 4.65959460e-01
-4.50335555e-02 1.11667059e-01 6.40040755e-01 8.09229195e-01
3.26395363e-01 -1.23944484e-01 -1.43223360e-01 -4.66657013e-01
2.58911431e-01 5.38268745e-01 -2.23972797e-01 -2.20320463e-01
-4.36290503e-01 7.74815559e-01 7.01696694e-01 5.18689930e-01
-3.57488036e-01 1.22749329e-01 6.97624147e-01 5.15193008e-02
5.39991498e-01 7.20866164e-03 -5.28487504e-01 -9.75564718e-02
-5.31817496e-01 -1.82949025e-02 6.65872455e-01 8.39986980e-01
5.52001953e-01 2.34538335e-02 -3.26803364e-02 7.08727717e-01
5.32914102e-01 2.41719663e-01 7.96087444e-01 -1.77821815e-01
5.11622727e-01 1.07541859e+00 -5.40257454e-01 -1.60615540e+00
-3.58032107e-01 -7.03584135e-01 -9.30779219e-01 -5.19705892e-01
-3.37490737e-01 -6.79555163e-02 -8.61742496e-01 1.89716756e+00
3.52682739e-01 1.04448473e+00 -1.37233123e-01 1.21171959e-01
6.03939831e-01 7.64031947e-01 1.48555085e-01 -3.63393009e-01
1.43592763e+00 -8.78505349e-01 -6.00449324e-01 -2.27873266e-01
7.45144427e-01 3.63502018e-02 1.01618278e+00 7.29543567e-02
-5.24947166e-01 -2.80047953e-01 -9.73606884e-01 3.50858569e-01
-7.99166620e-01 -2.09550977e-01 4.31668729e-01 7.30030477e-01
-8.30263674e-01 4.10708189e-01 -1.04116046e+00 -2.12299168e-01
6.26042068e-01 2.86305338e-01 -8.53440389e-02 1.60786450e-01
-1.39464390e+00 3.14711809e-01 1.71609998e-01 3.74645926e-02
-1.25746155e+00 -7.20802546e-01 -9.77340996e-01 2.21779873e-03
7.66450286e-01 -5.69614768e-01 9.09403861e-01 -6.44182742e-01
-1.12893629e+00 3.98195893e-01 3.54502089e-02 -5.39208770e-01
4.33212481e-02 -1.02116413e-01 -8.94675791e-01 -1.82151899e-01
1.21118156e-02 -1.84946463e-01 1.03789270e+00 -1.41064370e+00
-4.04719830e-01 -6.13722861e-01 3.48914742e-01 -1.60853535e-01
-9.50054467e-01 1.89319044e-01 -8.74752700e-01 -7.36983895e-01
-5.06711602e-01 -1.11585820e+00 -9.60669592e-02 -6.68524027e-01
-7.33136773e-01 -2.00602546e-01 1.48386216e+00 -1.01089168e+00
2.00857043e+00 -2.25039840e+00 5.09587288e-01 8.85137245e-02
6.95012212e-01 5.08428156e-01 -1.77390128e-01 2.39908338e-01
6.75516650e-02 5.67604184e-01 -4.57799375e-01 -6.18963301e-01
-1.45361483e-01 3.20280701e-01 -3.33443880e-01 1.87166646e-01
1.92314565e-01 1.20336080e+00 -9.10540938e-01 -3.17131519e-01
-3.46427888e-01 8.69984746e-01 -5.81036329e-01 -7.88337439e-02
-3.68227899e-01 1.37503669e-01 -8.50682259e-01 8.03803325e-01
6.87778592e-01 -5.92759371e-01 4.22163010e-01 -1.09076202e-01
2.80506849e-01 2.90455937e-01 -5.46506166e-01 1.17159498e+00
-6.41183436e-01 3.25021356e-01 1.77719489e-01 -7.61439860e-01
6.28326535e-01 -4.83140051e-02 6.83824360e-01 -5.83764195e-01
2.79040843e-01 -1.12955369e-01 1.51047364e-01 -4.09655571e-01
4.54138786e-01 5.73445082e-01 -1.87567994e-01 4.43768293e-01
-6.06068373e-02 5.80218375e-01 -1.31749749e-01 5.41545093e-01
1.50713122e+00 -1.21813968e-01 9.85600650e-02 -1.03126699e-02
8.98812830e-01 -3.72950315e-01 7.60471284e-01 2.33212680e-01
-2.00116128e-01 -8.03049207e-02 9.58722115e-01 -3.02406251e-01
-3.64514232e-01 -9.75271404e-01 3.86414051e-01 1.16410613e+00
4.33615685e-01 -1.11439383e+00 -8.26706529e-01 -1.57490563e+00
1.97692290e-02 2.28886127e-01 -1.01138949e+00 -6.64979517e-01
-8.32006991e-01 -9.94915426e-01 5.86184442e-01 4.71406758e-01
5.33152163e-01 -8.69391501e-01 -7.22792819e-02 6.53534085e-02
4.71285619e-02 -1.06119549e+00 -6.61723316e-01 -1.57169446e-01
-3.53435636e-01 -1.18635428e+00 -2.70116746e-01 -4.91647184e-01
3.74709189e-01 2.69852817e-01 9.49603796e-01 2.30379879e-01
-1.76472411e-01 4.73968267e-01 -5.10726392e-01 4.45003994e-03
-1.97846323e-01 5.51581442e-01 1.56291232e-01 3.79191160e-01
2.52284631e-02 -8.45653176e-01 -5.19614577e-01 4.72022891e-01
-1.01866877e+00 -5.81551015e-01 3.11485171e-01 6.25899196e-01
4.56342965e-01 5.93785167e-01 3.46072853e-01 -1.22084141e+00
9.59688306e-01 -1.25274765e+00 -4.32749748e-01 3.11236382e-01
-6.06926680e-01 -2.61534959e-01 8.07955384e-01 -1.04575300e+00
-8.73469472e-01 -4.08764094e-01 -7.15564787e-02 -7.39780784e-01
5.47369003e-01 7.80608296e-01 -5.68739891e-01 -1.03109829e-01
4.99015868e-01 3.56770605e-01 -2.42610693e-01 -3.59395206e-01
4.37981099e-01 5.52660644e-01 2.55410880e-01 -3.65950197e-01
1.11857176e+00 2.26570204e-01 -1.95899680e-01 -6.27834797e-01
-4.01147127e-01 -1.62900016e-01 -1.82280391e-01 6.35725930e-02
8.72423470e-01 -6.69660985e-01 -6.74271166e-01 7.42608190e-01
-1.08219254e+00 -3.71997476e-01 1.05977841e-01 -1.09686688e-01
-5.03575690e-02 5.07770896e-01 -7.28928685e-01 -6.67465746e-01
-3.19677144e-01 -1.41290307e+00 1.08723450e+00 1.01781942e-01
4.13324535e-02 -1.36836624e+00 1.95519224e-01 1.14470854e-01
3.74636054e-01 2.52109408e-01 9.67140675e-01 -8.19835067e-01
-6.23318315e-01 -7.88576454e-02 -2.97337830e-01 3.57900351e-01
6.06163740e-01 2.12410375e-01 -5.77080011e-01 -2.77761579e-01
4.40498479e-02 2.59617686e-01 8.96873653e-01 -1.06721697e-03
1.37655389e+00 -8.14094007e-01 -9.72058773e-01 8.45783770e-01
1.04061270e+00 3.40345442e-01 5.33591807e-01 7.76761174e-02
1.35797560e+00 5.96208334e-01 2.40993470e-01 2.91967183e-01
7.42181182e-01 9.20999229e-01 6.14960551e-01 4.16617632e-01
-9.78000462e-03 -4.53010857e-01 6.60537124e-01 9.73470032e-01
2.72509083e-02 -5.20722210e-01 -8.93117011e-01 2.68155783e-01
-1.87088716e+00 -8.77480388e-01 4.87808771e-02 1.80810022e+00
3.94168198e-01 1.99490979e-01 5.02029479e-01 -1.51919007e-01
6.97857797e-01 5.80366433e-01 -6.98825359e-01 -3.00427645e-01
1.20224491e-01 -9.68900844e-02 3.71256053e-01 4.67792571e-01
-1.04439640e+00 1.09313619e+00 5.43177414e+00 1.08527076e+00
-1.24117780e+00 5.29771924e-01 5.25429845e-01 2.49764353e-01
-6.20784581e-01 -2.51865119e-01 -9.38989460e-01 1.03644550e+00
1.12100899e+00 -3.25309306e-01 6.95026934e-01 8.46069515e-01
-2.66462296e-01 8.34885895e-01 -5.26590586e-01 9.22877133e-01
1.71634912e-01 -1.25185966e+00 1.34348556e-01 5.45389414e-01
5.46603441e-01 -1.38481120e-02 4.63536531e-01 6.66768074e-01
4.76500064e-01 -7.83075988e-01 2.87971765e-01 3.56247365e-01
5.31963766e-01 -8.02917838e-01 3.73666614e-01 2.68491507e-01
-1.88232017e+00 -6.79989874e-01 1.54374346e-01 4.91548091e-01
8.53157267e-02 1.84234038e-01 -5.79627156e-01 1.01493537e+00
7.18115091e-01 1.39892411e+00 -9.73003447e-01 4.42245424e-01
-7.48221502e-02 6.68756723e-01 5.29297628e-02 7.12714791e-02
2.08138093e-01 -2.10380837e-01 5.81065118e-01 9.79171693e-01
4.53719676e-01 -3.14954102e-01 5.85752055e-02 7.99320400e-01
-3.60698462e-01 -2.35149171e-02 -9.76187170e-01 -5.24735212e-01
6.01417184e-01 1.38178802e+00 -5.40928543e-01 -1.78710774e-01
-4.67647761e-01 1.07705748e+00 4.40813154e-01 4.28608179e-01
-1.24657583e+00 -1.12138905e-01 1.03755867e+00 3.75005513e-01
3.02476972e-01 -2.75353938e-01 1.49268106e-01 -1.26951468e+00
1.00180194e-01 -1.10114610e+00 5.85305691e-01 -3.32110077e-01
-1.37581038e+00 1.20012736e+00 7.97804147e-02 -1.00969863e+00
-3.40578914e-01 -6.19391322e-01 -8.53257775e-01 5.04550993e-01
-1.06697941e+00 -1.45215380e+00 -1.99749604e-01 8.74179780e-01
3.79298955e-01 -3.47313374e-01 2.81709224e-01 4.16826993e-01
-1.20368385e+00 8.12347114e-01 -3.23554486e-01 3.00431754e-02
-2.35500075e-02 -9.27653372e-01 8.68673086e-01 9.38756824e-01
2.41787478e-01 1.01246190e+00 1.32007003e-01 -1.25189483e+00
-1.42807555e+00 -1.48937511e+00 3.52899641e-01 -6.48572743e-01
1.28256011e+00 -7.18388379e-01 -1.13243401e+00 9.89059865e-01
8.86331350e-02 1.35946482e-01 4.77360427e-01 2.61682302e-01
-9.29782331e-01 -2.52312005e-01 -1.05611444e+00 8.10630798e-01
1.67420685e+00 -8.81057441e-01 2.35836394e-02 2.88930446e-01
1.43369269e+00 3.97678651e-02 -6.42558157e-01 7.66126335e-01
3.64697844e-01 -6.02684438e-01 1.14329541e+00 -6.36306345e-01
4.58524823e-02 -2.08678856e-01 -1.53782940e-03 -1.13185441e+00
-4.78459775e-01 -9.19042826e-01 -1.10300267e+00 1.43058395e+00
4.43736881e-01 -8.58175218e-01 7.81182885e-01 -8.84258002e-02
-1.23392321e-01 -1.07742298e+00 -8.05191100e-01 -9.76111948e-01
-2.02359021e-01 -4.57319260e-01 9.21073914e-01 9.70752597e-01
-2.20828176e-01 4.62303102e-01 -5.75314403e-01 3.88148189e-01
3.76276404e-01 -3.17562491e-01 5.24106622e-01 -1.34140122e+00
-6.52352631e-01 -7.46006191e-01 -6.09034061e-01 -8.44443619e-01
5.35883546e-01 -8.26947987e-01 -6.13138855e-01 -1.11282158e+00
1.93108723e-01 -3.70919764e-01 -5.21334648e-01 3.30551893e-01
-2.56427944e-01 1.21368207e-01 1.89619977e-02 1.81386873e-01
-5.75954080e-01 5.12571871e-01 9.23091233e-01 -6.36166096e-01
-4.65141654e-01 2.25435257e-01 -1.13171065e+00 5.80522656e-01
6.41176224e-01 -3.06399345e-01 -1.04869723e+00 -2.63857394e-01
3.27629566e-01 -3.81704807e-01 2.08734706e-01 -7.66479552e-01
1.12229995e-01 2.62495764e-02 -2.07440272e-01 -3.58276695e-01
4.80160207e-01 -7.29031086e-01 3.87591898e-01 4.46491539e-01
7.81284571e-02 4.11986172e-01 1.38896108e-01 1.11787415e+00
-3.21640484e-02 1.78205058e-01 3.97319138e-01 1.64580673e-01
-5.91139376e-01 1.00280225e+00 -2.27484867e-01 -8.33725333e-02
1.35642278e+00 1.56599879e-01 -6.16316617e-01 -2.46290311e-01
-6.87707901e-01 1.96232393e-01 4.29161042e-01 9.48661327e-01
6.33395851e-01 -1.38830364e+00 -2.00605944e-01 1.49405926e-01
1.10427648e-01 -4.99618024e-01 5.26857138e-01 8.56560767e-01
-5.87989092e-02 -1.69186983e-02 6.80760741e-02 -3.90500337e-01
-1.46459079e+00 1.11212432e+00 3.41288865e-01 -9.55645263e-01
-4.65827584e-01 8.20421994e-01 4.10385370e-01 -2.74530381e-01
1.21317409e-01 -1.88220739e-01 -5.28138340e-01 1.84527695e-01
5.84314704e-01 3.00110847e-01 -1.10253781e-01 -1.01252830e+00
-7.37703860e-01 5.96032917e-01 -4.49996471e-01 3.05621028e-01
1.12236392e+00 4.74570245e-02 -2.76814401e-01 1.20430999e-01
1.40715599e+00 3.47579837e-01 -9.45683539e-01 -1.98040813e-01
1.60413414e-01 -2.67599016e-01 -1.84029877e-01 -5.73292255e-01
-1.49192691e+00 6.95307851e-01 3.80703747e-01 7.55122006e-01
1.21916211e+00 1.16806097e-01 1.11294270e+00 -4.69584800e-02
3.90000701e-01 -3.84119004e-01 3.98850828e-01 5.47840953e-01
6.20593548e-01 -6.80157959e-01 -2.41909429e-01 -7.25828290e-01
-6.74946845e-01 4.23354030e-01 8.16854417e-01 -3.63874622e-02
1.04067588e+00 2.94963270e-01 -4.57080722e-01 -4.25116271e-01
-6.76848173e-01 -1.07789613e-01 4.16838437e-01 8.68317306e-01
-5.78579232e-02 8.65700915e-02 -1.53276503e-01 1.03008378e+00
1.81453094e-01 -5.10078192e-01 1.97092012e-01 7.34571159e-01
4.75095697e-02 -1.44150138e+00 1.64727792e-01 4.48144555e-01
-4.45917338e-01 -8.82921591e-02 -9.12071764e-01 6.66742086e-01
3.82010758e-01 8.44955862e-01 -1.74476132e-01 -1.40416563e+00
1.80762842e-01 -2.95214742e-01 3.80207598e-02 -6.82181716e-01
-8.26969445e-01 -2.55472153e-01 2.26392895e-02 -5.83187163e-01
1.48197532e-01 -5.43673813e-01 -9.22010124e-01 -1.99528173e-01
2.11654194e-02 6.87297201e-03 4.43110943e-01 5.71061134e-01
9.33891296e-01 9.66352820e-01 9.67852116e-01 -5.89119613e-01
-2.71269172e-01 -7.90347397e-01 -4.21047062e-01 3.29393983e-01
3.86104614e-01 -9.75372314e-01 -3.68844301e-01 -3.72726947e-01] | [14.384057998657227, 9.661163330078125] |
82e24f91-70b3-4581-a10b-4b5aa2cd9fa7 | efficient-inference-in-phylogenetic-indel | null | null | http://papers.nips.cc/paper/3406-efficient-inference-in-phylogenetic-indel-trees | http://papers.nips.cc/paper/3406-efficient-inference-in-phylogenetic-indel-trees.pdf | Efficient Inference in Phylogenetic InDel Trees | Accurate and efficient inference in evolutionary trees is a central problem in computational biology. Realistic models require tracking insertions and deletions along the phylogenetic tree, making inference challenging. We propose new sampling techniques that speed up inference and improve the quality of the samples. We compare our method to previous approaches and show performance improvement on metrics evaluating multiple sequence alignment and reconstruction of ancestral sequences. | ['Michael. I. Jordan', 'Alexandre Bouchard-Côté', 'Dan Klein'] | 2008-12-01 | null | null | null | neurips-2008-12 | ['multiple-sequence-alignment'] | ['medical'] | [ 6.91442072e-01 -4.81092066e-01 -3.12936902e-01 -4.69002932e-01
-5.11883795e-01 -9.26598549e-01 7.92515054e-02 5.43663144e-01
-7.25306571e-01 1.24084353e+00 -1.31374430e-02 -5.00011563e-01
-2.14665115e-01 -5.81202388e-01 -7.38568544e-01 -8.09333563e-01
-1.13493674e-01 1.00033987e+00 5.31813204e-01 1.07723765e-01
6.29632533e-01 6.06002450e-01 -1.28847909e+00 1.28436700e-01
9.10973251e-01 1.33381575e-01 4.50206906e-01 1.28476143e+00
-1.57850415e-01 -1.93896115e-01 -7.25036860e-01 -6.97736740e-01
4.39536339e-03 -7.47562170e-01 -8.91873598e-01 -4.04847175e-01
1.49527684e-01 -1.83650479e-01 2.20691293e-01 9.89364207e-01
7.57192969e-01 -7.43924230e-02 5.95364809e-01 -1.03681946e+00
-3.40453982e-02 6.75720513e-01 -5.54058015e-01 4.94891226e-01
1.03832781e-01 1.80452630e-01 1.04349852e+00 -3.11466753e-01
1.15040565e+00 1.23958158e+00 1.11183798e+00 3.33338439e-01
-1.44504511e+00 -3.06858450e-01 -2.58386761e-01 4.32473332e-01
-1.26106310e+00 -2.71058500e-01 -2.59709228e-02 -2.33825952e-01
1.06663108e+00 3.88430357e-01 8.02510798e-01 8.28594804e-01
2.60047764e-01 4.94157195e-01 5.99324405e-01 -1.56999007e-01
5.02552271e-01 -7.50165224e-01 -1.04788385e-01 5.62371969e-01
5.83669841e-01 -2.01265976e-01 -3.61607522e-01 -8.84344757e-01
5.34512997e-01 -6.66572377e-02 -7.17477426e-02 -8.80740508e-02
-1.02375960e+00 7.85588324e-01 -1.59997568e-02 -1.16429240e-01
-5.34201562e-01 2.48520210e-01 5.37040234e-01 -2.61415362e-01
1.39801338e-01 4.44422364e-01 -6.05105400e-01 -7.00733960e-01
-1.05293167e+00 2.37720355e-01 6.75466418e-01 7.37441242e-01
3.77594739e-01 -3.43845695e-01 4.47234064e-01 8.97299767e-01
-1.71744958e-01 2.74264067e-01 1.63888276e-01 -1.27087069e+00
-1.52817011e-01 3.05059314e-01 1.54563770e-01 -6.63889766e-01
-1.54849216e-01 -2.61178106e-01 -5.79879522e-01 -1.23985209e-01
4.96363193e-01 1.90306097e-01 -7.16089308e-01 1.73577988e+00
1.01337922e+00 3.91105652e-01 -2.87624896e-01 1.40581056e-01
5.99323306e-03 6.58613622e-01 -4.85769361e-02 -7.94788837e-01
9.20092046e-01 -4.90200400e-01 -5.82610369e-01 8.08979198e-03
2.61250079e-01 -8.32186043e-01 5.24054050e-01 5.15417337e-01
-9.95807827e-01 -7.07915351e-02 -8.19211662e-01 -8.50322694e-02
2.09412709e-01 -4.90701765e-01 7.62614846e-01 6.80317998e-01
-7.70471454e-01 1.10766053e+00 -1.04331970e+00 -3.79096240e-01
5.17639339e-01 2.73479164e-01 -2.66158521e-01 -4.00819853e-02
-3.37418288e-01 8.77793610e-01 6.54473782e-01 5.89531548e-02
-5.85914850e-01 -5.95141470e-01 -2.40790218e-01 -9.55185592e-02
2.75479943e-01 -6.90095663e-01 1.20468771e+00 -2.32143506e-01
-1.14166439e+00 8.28283727e-01 -6.89705193e-01 -4.84893054e-01
4.36200589e-01 -2.22529680e-01 2.49073610e-01 9.09759104e-02
-1.24048822e-01 4.75571781e-01 1.92895412e-01 -6.96949661e-01
-5.90158045e-01 -4.50994521e-01 -5.90431154e-01 -1.47590399e-01
3.41401517e-01 1.69131890e-01 -3.07007104e-01 -5.43496668e-01
6.10188134e-02 -9.94556606e-01 -6.86520219e-01 2.87569284e-01
-1.81800529e-01 -1.07768640e-01 3.61402750e-01 -1.12575829e+00
1.01250112e+00 -1.78588617e+00 6.67635202e-01 -5.10126315e-02
-6.31993562e-02 4.21970874e-01 -8.35034475e-02 6.34911895e-01
1.29797682e-01 1.68705076e-01 -8.29129279e-01 6.60796165e-02
-1.95100695e-01 8.57466400e-01 -1.29627600e-01 6.44285142e-01
-6.03661984e-02 6.56601608e-01 -1.00697470e+00 -6.62995160e-01
-4.47463728e-02 4.28240716e-01 -8.78840327e-01 3.19280297e-01
-4.89057362e-01 4.59986329e-01 -8.32230672e-02 5.99322200e-01
7.27420807e-01 -2.78905958e-01 7.05454648e-01 2.65187889e-01
-2.07207277e-01 6.05149329e-01 -8.14858735e-01 1.56000566e+00
-1.87290385e-02 4.13471371e-01 2.59853862e-02 -9.29280937e-01
7.01168954e-01 -8.46478045e-02 2.17669010e-01 1.73985988e-01
1.14998277e-02 3.06812704e-01 4.38148171e-01 -1.84603706e-01
2.63975859e-01 -1.54384300e-01 2.11166322e-01 3.17738771e-01
-1.01222381e-01 -3.66323501e-01 3.20236534e-01 -4.98957410e-02
9.95811999e-01 4.55981046e-01 9.49954212e-01 -9.17921439e-02
1.38960674e-01 1.37683526e-01 1.20602775e+00 7.64764309e-01
-1.76797390e-01 2.10970804e-01 5.77304244e-01 -5.87792635e-01
-1.63373864e+00 -1.08102858e+00 -4.15494382e-01 1.10359871e+00
-2.48324454e-01 -5.56777298e-01 -8.22243154e-01 -3.84543896e-01
4.97511849e-02 7.53066957e-01 -5.23276389e-01 -7.63237998e-02
-7.00454175e-01 -1.21247113e+00 8.75839531e-01 3.80872846e-01
-1.67046115e-01 -6.82974815e-01 -9.17218328e-01 3.97899419e-01
-5.70606649e-01 -7.01880395e-01 -2.83879995e-01 2.84514189e-01
-1.27534544e+00 -1.25908291e+00 -4.91534203e-01 -4.18990344e-01
4.33737010e-01 -1.44215122e-01 9.06405389e-01 3.37139666e-01
-1.04033041e+00 -5.37982047e-01 -2.11165190e-01 -3.36555541e-01
-6.70683682e-01 2.79768519e-02 1.76741570e-01 -7.13891566e-01
2.64710873e-01 -8.84542167e-01 -3.19793522e-01 4.20983464e-01
-6.94594264e-01 -5.84107563e-02 2.97858953e-01 1.21956336e+00
7.84085274e-01 -1.66923106e-01 1.22595526e-01 -7.64965832e-01
2.34102011e-01 -4.20505047e-01 -1.03466570e+00 7.76859820e-01
-2.36936957e-01 3.15055579e-01 4.86154437e-01 -4.13287818e-01
-8.26118112e-01 2.00700331e-02 -8.10173452e-01 8.41953829e-02
7.23404437e-02 2.44745016e-01 -2.17036027e-02 -1.22936573e-02
4.68586385e-01 4.20318574e-01 -2.79335063e-02 -6.55858815e-01
9.60076898e-02 6.56238317e-01 6.87969029e-01 -7.44983673e-01
1.71401188e-01 4.03228670e-01 4.25598174e-01 -1.17869961e+00
-4.63600278e-01 -3.04423958e-01 -9.73770559e-01 1.19363390e-01
3.36114138e-01 -2.50548899e-01 -1.12966728e+00 2.73183376e-01
-1.05540466e+00 -1.73453599e-01 4.58896346e-02 4.58521813e-01
-6.14634752e-01 9.35025752e-01 -6.16850615e-01 -9.37586844e-01
-3.43182087e-01 -1.06987560e+00 9.89671588e-01 1.18444577e-01
-4.80075568e-01 -8.47683311e-01 7.35977173e-01 9.26256552e-02
1.49258524e-01 5.35470963e-01 1.09432054e+00 -4.05268043e-01
-3.81897688e-01 1.00225821e-01 2.46054813e-01 -2.34738722e-01
1.59219205e-01 9.00125146e-01 -4.89363372e-01 -1.99328557e-01
-2.38516301e-01 -3.48763950e-02 8.04582834e-01 5.64125776e-01
1.28570557e+00 -2.59957671e-01 -6.46037817e-01 9.58715975e-01
1.25496387e+00 1.72823071e-01 5.56685686e-01 2.00781330e-01
9.09873396e-02 7.31999338e-01 8.80161047e-01 6.73788011e-01
-1.54938430e-01 7.61534989e-01 1.39656335e-01 5.84718287e-01
3.59765559e-01 -3.93087298e-01 -1.23202205e-01 7.46317923e-01
-1.93000615e-01 -2.91479021e-01 -1.06796777e+00 7.57393301e-01
-1.81415021e+00 -1.19431269e+00 -1.49327844e-01 2.23321247e+00
1.33179855e+00 -2.21450746e-01 3.78769279e-01 1.27012089e-01
8.92190993e-01 -3.73120487e-01 -1.04100418e+00 -6.54054880e-01
-1.80342406e-01 5.25036991e-01 5.89839697e-01 7.32508540e-01
-5.12113035e-01 9.32982743e-01 8.92461967e+00 7.12121189e-01
-6.64481461e-01 -3.15979332e-01 4.59490955e-01 -3.96447122e-01
-4.05868828e-01 2.41016209e-01 -9.57328498e-01 4.94633198e-01
1.12351179e+00 -2.86390960e-01 3.68043751e-01 8.00290227e-01
-9.69013572e-03 -2.26539299e-01 -8.82750809e-01 5.78620374e-01
-3.68936360e-01 -1.43223596e+00 -1.11696348e-01 2.04580650e-01
3.32753599e-01 1.58544555e-01 -3.90479684e-01 -5.54310918e-01
1.09901488e+00 -8.55937362e-01 1.18003450e-01 2.74939239e-01
6.17500663e-01 -1.15197802e+00 5.96367657e-01 3.99238288e-01
-6.75575018e-01 2.62579739e-01 -9.36812639e-01 -8.12107790e-03
6.92112505e-01 7.53885448e-01 -1.33703446e+00 3.27818036e-01
6.42531753e-01 5.27000010e-01 -4.70056981e-01 1.42167473e+00
-5.10394514e-01 8.89509976e-01 -7.48456597e-01 -3.33673388e-01
-6.92903027e-02 -4.95727003e-01 5.22334576e-01 1.26444948e+00
4.12384152e-01 6.74976632e-02 5.47760203e-02 4.54289794e-01
2.21164808e-01 -7.19537362e-02 -3.65430534e-01 -3.39266628e-01
9.59348500e-01 7.32533455e-01 -9.04166341e-01 -4.36259836e-01
3.48995566e-01 1.10563326e+00 5.06410182e-01 -2.32341528e-01
-9.15234387e-01 -2.81950653e-01 1.23756015e+00 -4.95339543e-01
5.95333338e-01 -3.39505404e-01 1.03970751e-01 -8.60760093e-01
-1.96277231e-01 -9.53981042e-01 7.84737825e-01 -3.05911243e-01
-1.12211478e+00 1.30284041e-01 1.70682538e-02 -4.04530108e-01
-5.99996924e-01 -4.98306811e-01 -1.58810869e-01 7.00806081e-01
-9.65229273e-01 -5.36379158e-01 2.19129831e-01 -2.24570334e-01
5.33914447e-01 4.90854859e-01 8.12400162e-01 -3.33685540e-02
-7.97471106e-01 5.97189903e-01 6.34590328e-01 -4.76721674e-01
4.40737575e-01 -1.11295033e+00 1.21967697e+00 1.03098643e+00
-8.64114687e-02 1.10889673e+00 9.39847112e-01 -1.08401990e+00
-1.39444125e+00 -6.19477808e-01 6.85924351e-01 -4.30650972e-02
4.48234826e-01 -4.03172046e-01 -1.15573752e+00 7.67050326e-01
-1.24655984e-01 -5.40498495e-01 1.17838442e+00 2.62260914e-01
-6.27436697e-01 3.90412539e-01 -1.30948794e+00 5.45514345e-01
1.37811995e+00 -2.44580522e-01 -5.80730498e-01 2.68939793e-01
5.53601801e-01 -1.67090505e-01 -9.96062517e-01 3.15842569e-01
1.16172767e+00 -7.59442449e-01 1.26922131e+00 -1.06300056e+00
1.02579236e-01 -5.05465329e-01 -1.49840236e-01 -9.49299455e-01
-3.56754839e-01 -6.83310390e-01 1.53308749e-01 8.97666633e-01
2.24951997e-01 -4.80511189e-01 7.13882267e-01 -2.25170311e-02
1.21891350e-01 -1.74122751e-01 -1.07236695e+00 -9.21436131e-01
8.63123462e-02 6.06784560e-02 5.66296101e-01 9.91148651e-01
-7.97084868e-02 2.21331164e-01 -5.84354639e-01 1.95996419e-01
1.03008401e+00 2.09055409e-01 8.02695215e-01 -1.21563256e+00
-6.13795936e-01 -4.66457218e-01 -4.88348752e-01 -6.78630590e-01
-1.39427558e-01 -5.72545171e-01 2.04555452e-01 -1.30609524e+00
4.24055010e-01 -1.45525917e-01 3.65497589e-01 2.83684790e-01
-7.78976202e-01 2.10335664e-02 -3.27812493e-01 2.42260154e-02
-2.65338421e-01 3.16813380e-01 3.62326831e-01 4.62875158e-01
4.38441038e-01 -2.53314435e-01 -1.15331076e-01 6.93087578e-01
9.22106504e-01 -8.59373569e-01 6.63105696e-02 -3.31374586e-01
4.00340050e-01 3.09715997e-02 -1.69591814e-01 -5.78745961e-01
1.63439065e-02 -6.28090620e-01 2.95114815e-01 -1.03659856e+00
3.21662426e-01 -2.66425341e-01 8.75320971e-01 1.33826256e+00
-2.46351272e-01 4.70633686e-01 1.69679761e-01 5.56725562e-01
2.27813274e-01 -6.79898918e-01 1.12827623e+00 -2.57520080e-01
-4.05215800e-01 9.90443975e-02 -2.99291462e-01 -5.86337261e-02
9.37191486e-01 -5.98491505e-02 -3.13746423e-01 3.14888246e-02
-3.68964821e-01 1.21158630e-01 1.09271884e+00 -1.97227076e-02
5.64623177e-01 -5.32723427e-01 -9.70678866e-01 -2.40360290e-01
-1.99970417e-02 -1.60028934e-01 2.03003008e-02 5.39421022e-01
-1.37060440e+00 4.35926206e-02 -5.33984780e-01 -8.02159846e-01
-2.12108064e+00 7.42028713e-01 9.80737433e-02 -4.48652264e-03
-4.23538953e-01 1.07913566e+00 -2.81927198e-01 -7.67606199e-01
9.97422114e-02 2.89191753e-01 2.42704183e-01 -1.75583929e-01
8.20744872e-01 6.29392803e-01 -4.14487302e-01 -3.45859230e-01
-5.40647626e-01 5.63033760e-01 -2.20975935e-01 -9.22443345e-02
1.73155141e+00 5.79283908e-02 -7.35128462e-01 5.51821351e-01
8.83270264e-01 -2.37290859e-02 -9.21999574e-01 -1.05835460e-01
4.96844828e-01 -9.46407914e-01 -7.15410650e-01 -5.25213897e-01
-5.69681451e-02 9.91110802e-01 1.69132218e-01 -4.62234020e-01
8.67342889e-01 -1.94869310e-01 5.72028339e-01 7.18509316e-01
3.94255072e-01 -8.55814993e-01 -4.06552374e-01 2.59509057e-01
3.12618464e-01 -6.97462022e-01 4.94163394e-01 -6.37485683e-01
-2.01764740e-02 9.67075169e-01 3.97074781e-02 3.48791420e-01
-2.60880798e-01 5.48702896e-01 -3.57091099e-01 1.10502347e-01
-1.10278177e+00 9.09461677e-02 -3.21379274e-01 4.82598305e-01
4.29950237e-01 -5.76361045e-02 -7.01452196e-01 -3.81628931e-01
-4.22868609e-01 -8.01190808e-02 4.08506393e-01 1.20236421e+00
-7.90492117e-01 -1.66850328e+00 -4.04659361e-01 1.97934061e-01
-7.04041839e-01 -1.88334942e-01 -8.83728862e-01 3.79607052e-01
-2.14884251e-01 5.30275524e-01 1.26567706e-01 8.94503891e-02
-2.32675135e-01 3.26843411e-01 8.84669662e-01 -2.27173433e-01
-4.95651454e-01 2.05216110e-01 3.86445969e-01 -2.88746595e-01
-1.89770222e-01 -1.05040312e+00 -9.52343881e-01 -8.09955776e-01
-6.72354817e-01 5.59098780e-01 9.52520370e-01 5.89019835e-01
3.87525111e-01 2.73677945e-01 4.39899117e-01 -3.45786393e-01
-6.45581603e-01 -8.15204382e-01 -5.59896715e-02 1.69188544e-01
2.01018944e-01 -4.24786508e-01 -1.61284506e-01 4.33323607e-02] | [4.868743419647217, 5.155880451202393] |
cbbb5cbc-1053-4f62-9155-e17caa503d79 | fmg-net-and-w-net-multigrid-inspired-deep | 2304.02725 | null | https://arxiv.org/abs/2304.02725v1 | https://arxiv.org/pdf/2304.02725v1.pdf | FMG-Net and W-Net: Multigrid Inspired Deep Learning Architectures For Medical Imaging Segmentation | Accurate medical imaging segmentation is critical for precise and effective medical interventions. However, despite the success of convolutional neural networks (CNNs) in medical image segmentation, they still face challenges in handling fine-scale features and variations in image scales. These challenges are particularly evident in complex and challenging segmentation tasks, such as the BraTS multi-label brain tumor segmentation challenge. In this task, accurately segmenting the various tumor sub-components, which vary significantly in size and shape, remains a significant challenge, with even state-of-the-art methods producing substantial errors. Therefore, we propose two architectures, FMG-Net and W-Net, that incorporate the principles of geometric multigrid methods for solving linear systems of equations into CNNs to address these challenges. Our experiments on the BraTS 2020 dataset demonstrate that both FMG-Net and W-Net outperform the widely used U-Net architecture regarding tumor subcomponent segmentation accuracy and training efficiency. These findings highlight the potential of incorporating the principles of multigrid methods into CNNs to improve the accuracy and efficiency of medical imaging segmentation. | ['David Fuentes', 'Beatrice Riviere', 'Adrian Celaya'] | 2023-04-05 | null | null | null | null | ['tumor-segmentation', 'brain-tumor-segmentation'] | ['computer-vision', 'medical'] | [ 1.86166137e-01 1.04659162e-01 -2.35791802e-01 -3.26651514e-01
-1.02924740e+00 -4.14642543e-01 1.23447157e-01 2.68706977e-01
-5.87905705e-01 6.38564050e-01 -1.81845829e-01 -4.92382824e-01
-8.88394341e-02 -7.52075076e-01 -4.07459855e-01 -7.59226620e-01
-8.57864693e-02 6.57683372e-01 3.34575117e-01 -1.60892397e-01
-1.65592104e-01 8.39092493e-01 -7.89622664e-01 3.45666379e-01
9.09232140e-01 1.26759720e+00 1.18314192e-01 7.51434207e-01
-2.89383918e-01 7.41661668e-01 -2.49864012e-01 5.44647649e-02
2.93667108e-01 -1.94888160e-01 -1.18176186e+00 1.73137002e-02
6.60692513e-01 -4.91217762e-01 -2.19413429e-01 1.08566868e+00
5.06608844e-01 1.81956422e-02 7.29509354e-01 -8.08067858e-01
-2.50135154e-01 3.55066508e-01 -6.19649768e-01 4.63554353e-01
-4.28036451e-01 1.08010486e-01 5.95158756e-01 -3.41943383e-01
5.35084724e-01 8.36634696e-01 1.28892815e+00 6.21792138e-01
-1.04793429e+00 -5.53988576e-01 -1.61926717e-01 -1.90872401e-01
-1.24974263e+00 -1.56428874e-01 2.46704474e-01 -5.94337165e-01
9.40087616e-01 4.06621665e-01 6.16682529e-01 4.19082850e-01
7.19187319e-01 8.09179842e-01 9.98840511e-01 -9.63084176e-02
8.43427796e-03 -5.43548584e-01 2.38732368e-01 1.05307519e+00
2.14174256e-01 -1.60530925e-01 1.30675852e-01 -1.04649484e-01
1.14519453e+00 1.39589846e-01 -2.79094428e-01 -7.24824220e-02
-1.27639174e+00 9.46075022e-01 9.41452026e-01 5.48723280e-01
-3.44761282e-01 5.44691682e-01 5.36856353e-01 -2.29290679e-01
8.12683403e-01 5.75565517e-01 -3.88208449e-01 9.94704738e-02
-1.26600099e+00 1.77528620e-01 4.23231691e-01 7.18747377e-01
4.39566433e-01 -8.79428908e-02 -3.65024894e-01 7.66782820e-01
3.41112241e-02 8.94671157e-02 6.22263610e-01 -8.30841899e-01
-5.94272688e-02 6.93117023e-01 -3.70342553e-01 -7.50736892e-01
-1.09154773e+00 -7.79944599e-01 -1.22054350e+00 2.80019164e-01
6.40318990e-01 -1.69381589e-01 -1.54492247e+00 1.34721053e+00
5.54150403e-01 2.08744809e-01 -2.45304465e-01 8.39494824e-01
1.22520840e+00 1.75754011e-01 3.84056568e-01 9.38231423e-02
1.47064459e+00 -1.16333067e+00 -5.91320038e-01 -2.22855508e-01
9.77171838e-01 -8.17163408e-01 5.26915252e-01 3.24325659e-03
-1.33360982e+00 -2.98375964e-01 -8.69090796e-01 -2.44301885e-01
-3.15517455e-01 -1.18367923e-02 9.30612743e-01 6.79588497e-01
-1.30745816e+00 7.93424010e-01 -1.21202540e+00 -1.77889928e-01
1.10695481e+00 7.19287574e-01 -8.17398056e-02 -9.82814133e-02
-9.20081794e-01 8.73279095e-01 4.04953212e-01 8.99880379e-02
-8.13808799e-01 -1.32879114e+00 -7.20982194e-01 -5.57510592e-02
1.70704246e-01 -7.17123151e-01 1.49550486e+00 -7.68069148e-01
-1.15019310e+00 9.68105674e-01 1.84735745e-01 -5.27144372e-01
8.37700963e-01 2.96826094e-01 7.18446821e-02 4.63363200e-01
1.81869060e-01 1.03453147e+00 3.39100152e-01 -9.22795057e-01
-6.00051701e-01 -4.13491100e-01 -2.11978346e-01 1.61969334e-01
-2.82658283e-02 -1.10987164e-01 -5.20333827e-01 -6.59246266e-01
2.73245573e-01 -1.04699481e+00 -8.34139407e-01 2.73027390e-01
-4.12571341e-01 -7.50335380e-02 9.52634692e-01 -7.60466397e-01
6.29700601e-01 -1.77885532e+00 -7.82666281e-02 1.27760872e-01
7.37526178e-01 2.44441777e-01 7.56697878e-02 -4.68272477e-01
9.93563533e-02 2.54970849e-01 -3.63450855e-01 -2.77527332e-01
-4.82955247e-01 3.77255321e-01 4.94684488e-01 6.17308080e-01
3.07073332e-02 1.51995504e+00 -7.88182199e-01 -7.42283762e-01
2.79362470e-01 4.53858793e-01 -5.25281608e-01 8.22763592e-02
-2.21430197e-01 6.61308885e-01 -5.67930341e-01 8.68494034e-01
5.88869810e-01 -6.54793322e-01 -5.57559729e-02 -4.45556849e-01
-4.14399877e-02 -2.98085093e-01 -4.72386360e-01 1.75143433e+00
-3.86515290e-01 5.55246055e-01 4.89607364e-01 -1.12952459e+00
2.25710660e-01 4.27937508e-01 1.23532355e+00 -6.86140239e-01
5.31787455e-01 4.68062282e-01 2.32673362e-01 -4.02782679e-01
3.56356233e-01 -4.32858080e-01 2.14293540e-01 1.99703321e-01
3.03344466e-02 -5.16345918e-01 2.11304218e-01 -9.00011975e-03
1.20016193e+00 -3.36169928e-01 1.51646405e-01 -6.79726422e-01
1.82534724e-01 4.28950518e-01 4.64366227e-01 7.26633310e-01
-5.73478103e-01 9.04174626e-01 4.45162535e-01 -8.52538764e-01
-9.39308405e-01 -7.41666138e-01 -4.82946277e-01 7.57224083e-01
-7.71678239e-02 1.16010569e-01 -1.13933492e+00 -7.43844867e-01
-1.35806696e-02 -7.48750642e-02 -9.24629807e-01 1.72865853e-01
-7.45770156e-01 -1.24805880e+00 6.86386228e-01 7.10878193e-01
6.68400168e-01 -9.78315532e-01 -9.21364844e-01 5.01552343e-01
-2.49003351e-01 -1.42507339e+00 -5.68703413e-01 3.03705573e-01
-1.09355700e+00 -1.26935101e+00 -1.16268110e+00 -8.99669945e-01
9.39309001e-01 7.58460015e-02 1.30338454e+00 4.88947600e-01
-9.54725564e-01 1.83582053e-01 -1.26320973e-01 -2.73875952e-01
-5.63088655e-01 4.89417344e-01 -5.88610470e-01 -4.72801208e-01
-1.69440329e-01 -1.88042313e-01 -9.28305507e-01 2.94732511e-01
-1.05996323e+00 5.04606128e-01 6.06258392e-01 8.36135566e-01
9.15315568e-01 -1.64132286e-02 3.66840690e-01 -1.18275917e+00
5.82309902e-01 -3.20097774e-01 -2.98122793e-01 2.95318872e-01
-3.85069132e-01 -3.34405541e-01 4.72588450e-01 -2.69141585e-01
-5.44503212e-01 1.44619301e-01 -4.98221070e-01 -2.68909544e-01
-1.95874274e-01 6.13739431e-01 5.66917479e-01 -8.88023257e-01
5.36346316e-01 -1.66257229e-02 3.38515282e-01 8.72236192e-02
7.96129704e-02 1.51419386e-01 5.41041017e-01 -4.91323709e-01
1.48449987e-01 7.59631097e-01 4.39889878e-01 -8.08292925e-01
-1.03929090e+00 -5.74974418e-01 -6.42424583e-01 -3.89163822e-01
1.41341221e+00 -5.89387119e-01 -5.21135986e-01 8.29074502e-01
-1.09856117e+00 -7.48375416e-01 -4.34809476e-01 4.85238284e-02
-5.49943209e-01 1.81575164e-01 -1.11002970e+00 3.32848132e-02
-7.94014275e-01 -1.88812518e+00 1.19155121e+00 3.54282856e-01
-1.51803717e-01 -1.37767863e+00 -3.15035373e-01 5.15332401e-01
9.24643219e-01 8.24099839e-01 1.05785513e+00 -4.65707898e-01
-5.05502224e-01 -1.56885013e-01 -6.56376719e-01 1.99242860e-01
3.01496625e-01 -2.19830781e-01 -6.78112745e-01 -3.77388030e-01
-1.25228703e-01 -4.25552070e-01 8.72424185e-01 1.23798752e+00
1.69344580e+00 9.35155600e-02 -5.73354959e-01 1.18168044e+00
1.52316284e+00 1.28640324e-01 3.02200288e-01 1.25041142e-01
9.47007835e-01 1.28590629e-01 -4.52127494e-02 -5.02555892e-02
3.44899118e-01 4.05792177e-01 6.03842020e-01 -9.48387802e-01
-4.69672889e-01 6.08625531e-01 -6.61237001e-01 8.02944481e-01
2.04148889e-02 8.20390955e-02 -1.34150100e+00 6.29543662e-01
-1.59449542e+00 -4.20977145e-01 -3.57681215e-01 1.44471407e+00
9.75906849e-01 -9.32807848e-03 -1.79194868e-01 -2.56336212e-01
4.16694552e-01 5.95561648e-03 -7.77351916e-01 -1.23336725e-01
1.49599686e-01 4.70396638e-01 8.89857292e-01 3.95318598e-01
-1.36977303e+00 8.76750112e-01 6.96064377e+00 1.12588227e+00
-1.30045807e+00 4.28015620e-01 1.58556902e+00 8.65449682e-02
1.45975202e-01 -6.58265531e-01 -5.75187504e-01 1.23888865e-01
7.17292309e-01 2.48323783e-01 1.87043890e-01 5.12469709e-01
3.42213549e-02 -2.48396412e-01 -8.08478713e-01 7.27109492e-01
-1.34605825e-01 -1.90020490e+00 -1.39085606e-01 1.29961088e-01
1.06846976e+00 5.23809016e-01 -1.43250898e-02 7.33201057e-02
3.90066773e-01 -1.44988763e+00 3.54512066e-01 1.96308434e-01
1.01310718e+00 -6.13981426e-01 8.53347242e-01 3.93175483e-02
-1.17738831e+00 4.29998606e-01 -3.90047520e-01 4.86891747e-01
1.62873819e-01 7.15228975e-01 -7.89175928e-01 4.74847972e-01
5.83303273e-01 5.76420784e-01 -5.17279029e-01 1.23088646e+00
4.12175059e-01 4.91671354e-01 -2.97906667e-01 2.79730022e-01
6.81792676e-01 -3.47889811e-02 -2.69945674e-02 1.38680136e+00
2.74409413e-01 2.54515857e-01 5.11467099e-01 8.66699219e-01
-3.05787295e-01 7.41550606e-03 -2.41932925e-02 1.07973605e-01
-2.70819098e-01 1.77379870e+00 -1.52540302e+00 -5.09621978e-01
-2.46563420e-01 6.01148605e-01 2.77461618e-01 2.19913498e-01
-7.85548210e-01 1.39574841e-01 7.12863147e-01 9.60947424e-02
-1.14066983e-02 -3.14209878e-01 -6.70245111e-01 -7.85109580e-01
-5.10507822e-01 -8.41430366e-01 4.44433838e-01 -4.01043773e-01
-1.24508429e+00 8.34740520e-01 -2.31220782e-01 -7.44077027e-01
2.04626441e-01 -8.04867983e-01 -5.06204486e-01 7.90708959e-01
-1.78051436e+00 -1.50114059e+00 -7.03439951e-01 5.91809154e-01
5.98161221e-01 2.24573538e-01 8.13809216e-01 3.81984800e-01
-3.11057776e-01 3.75012934e-01 8.36162791e-02 4.06965196e-01
2.56010741e-01 -1.29012942e+00 3.68111074e-01 4.10859704e-01
-4.72838461e-01 1.63763404e-01 2.97776312e-01 -6.86900854e-01
-1.03591418e+00 -1.43081653e+00 1.91626459e-01 -8.09330568e-02
6.11165643e-01 3.90854524e-03 -8.91749799e-01 5.34893334e-01
1.98456511e-01 7.88308144e-01 7.75492668e-01 -3.14866781e-01
1.96582153e-01 2.92748511e-01 -1.46363640e+00 4.31922704e-01
7.26337373e-01 -2.17901945e-01 6.66590035e-02 7.82023787e-01
6.81178451e-01 -1.24632931e+00 -1.19874537e+00 7.33450651e-01
4.45214212e-01 -7.37671793e-01 1.17047966e+00 -5.67132592e-01
6.78055167e-01 2.21426204e-01 2.60586709e-01 -1.46351933e+00
-3.47906679e-01 -2.27859855e-01 1.87022194e-01 4.02199149e-01
2.92003065e-01 -5.03938556e-01 1.14565480e+00 7.23970592e-01
-6.02046967e-01 -1.38364482e+00 -1.24285626e+00 -4.65947360e-01
7.50015199e-01 -3.16322744e-01 3.99133146e-01 9.96557236e-01
-5.82806766e-01 -3.89027536e-01 1.80506021e-01 -1.73670158e-01
7.20368683e-01 -1.10490046e-01 1.25695214e-01 -1.08677101e+00
1.09978048e-02 -9.02359843e-01 -3.07157218e-01 -6.78221226e-01
-1.24817155e-02 -1.06590950e+00 9.42786634e-02 -1.89621317e+00
1.41672537e-01 -7.21304655e-01 -3.10303539e-01 5.44363678e-01
-2.03617111e-01 7.38592088e-01 -4.58116494e-02 1.36861324e-01
-5.36996722e-01 8.63782465e-02 1.92180252e+00 -4.55952257e-01
1.96425617e-01 -1.26628593e-01 -4.63427246e-01 8.31823289e-01
8.41691315e-01 -3.40146840e-01 -1.05062820e-01 -6.37455940e-01
-2.09304154e-01 2.36902356e-01 3.93149108e-01 -1.28094304e+00
4.05736804e-01 -1.95195749e-01 5.92785358e-01 -4.76289064e-01
4.08204384e-02 -7.74505615e-01 -9.69249755e-02 8.12154710e-01
-3.27092588e-01 1.81489870e-01 7.51796663e-01 -7.44230580e-04
-1.05972998e-01 -6.72668144e-02 1.23394966e+00 -6.28999174e-01
-2.74878323e-01 8.80180895e-01 -4.90521729e-01 2.68432796e-01
1.15472984e+00 -2.55675942e-01 -4.18590695e-01 1.65355653e-01
-8.59163582e-01 3.42797041e-01 1.29263654e-01 -1.74593162e-02
4.97798175e-01 -1.14566398e+00 -6.34000361e-01 9.82571095e-02
-2.11385086e-01 6.26433730e-01 5.68092346e-01 1.19742262e+00
-1.25550365e+00 3.68023872e-01 -3.87017220e-01 -8.00374746e-01
-1.04888952e+00 -1.16122372e-01 1.12514532e+00 -8.15131426e-01
-6.31389856e-01 1.09997499e+00 4.11502749e-01 -4.89897162e-01
-1.25057213e-02 -9.55426395e-01 -5.47152460e-02 -1.90485716e-01
2.99879640e-01 1.79941416e-01 4.12741154e-01 -5.10468543e-01
-2.23554730e-01 5.04096746e-01 -3.96700323e-01 4.16139364e-01
1.43371284e+00 2.39339247e-01 -4.54560995e-01 -1.81892574e-01
1.23215842e+00 -6.93666577e-01 -1.23918772e+00 -9.90571156e-02
-1.39173940e-01 4.15850244e-02 6.14932418e-01 -9.31261957e-01
-1.75366592e+00 8.55658889e-01 7.93755949e-01 2.52251118e-01
9.41295922e-01 -8.27325061e-02 1.26672292e+00 -9.56564993e-02
1.25779852e-01 -9.51786876e-01 -7.19197765e-02 4.77433980e-01
6.07469499e-01 -1.39507020e+00 4.77102436e-02 -5.77839136e-01
-6.21321648e-02 1.30595291e+00 8.19475532e-01 -1.03582434e-01
8.57154131e-01 6.38859808e-01 3.50726575e-01 -6.11324668e-01
-1.80066466e-01 2.67223604e-02 4.44794774e-01 2.04076290e-01
7.27044404e-01 2.96506047e-01 -7.11321160e-02 9.87869874e-02
-7.86069632e-02 1.85429290e-01 3.42776626e-01 9.28387165e-01
-2.38438815e-01 -7.57234335e-01 -2.30636790e-01 1.01186121e+00
-8.34860563e-01 -1.40382841e-01 1.89157333e-02 7.82147110e-01
2.19083428e-01 4.43305522e-01 7.28713349e-02 1.09315775e-01
7.98349008e-02 -3.38336587e-01 4.88799632e-01 -6.10087514e-01
-1.12651467e+00 7.95322731e-02 -3.50894392e-01 -5.78589737e-01
-3.86504918e-01 -4.80621964e-01 -1.54055727e+00 -2.17728302e-01
-1.29882634e-01 -1.35324240e-01 7.73367763e-01 1.16679740e+00
6.87294975e-02 1.26562309e+00 -1.14164986e-01 -1.03262532e+00
-3.92669648e-01 -7.32367992e-01 -4.20847148e-01 3.95637214e-01
4.03174073e-01 -3.39387357e-01 1.01397499e-01 -1.46932974e-01] | [14.499258995056152, -2.530097246170044] |
b99b97c5-434a-4d7f-b857-94c06dbdac06 | improving-the-neural-network-based-machine | null | null | https://aclanthology.org/Y18-1038 | https://aclanthology.org/Y18-1038.pdf | Improving the neural network-based machine transliteration for low-resourced language pair | null | ['Fatiha Sadat', 'Ngoc Tan Le'] | null | null | null | null | paclic-2018-12 | ['transliteration'] | ['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] | [-7.351063251495361, 3.6975409984588623] |
5f56a25c-ab80-4b25-9bef-5929b35df4fe | entity-relation-and-event-extraction-with | 1909.03546 | null | https://arxiv.org/abs/1909.03546v2 | https://arxiv.org/pdf/1909.03546v2.pdf | Entity, Relation, and Event Extraction with Contextualized Span Representations | We examine the capabilities of a unified, multi-task framework for three information extraction tasks: named entity recognition, relation extraction, and event extraction. Our framework (called DyGIE++) accomplishes all tasks by enumerating, refining, and scoring text spans designed to capture local (within-sentence) and global (cross-sentence) context. Our framework achieves state-of-the-art results across all tasks, on four datasets from a variety of domains. We perform experiments comparing different techniques to construct span representations. Contextualized embeddings like BERT perform well at capturing relationships among entities in the same or adjacent sentences, while dynamic span graph updates model long-range cross-sentence relationships. For instance, propagating span representations via predicted coreference links can enable the model to disambiguate challenging entity mentions. Our code is publicly available at https://github.com/dwadden/dygiepp and can be easily adapted for new tasks or datasets. | ['Hannaneh Hajishirzi', 'Yi Luan', 'Ulme Wennberg', 'David Wadden'] | 2019-09-08 | entity-relation-and-event-extraction-with-1 | https://aclanthology.org/D19-1585 | https://aclanthology.org/D19-1585.pdf | ijcnlp-2019-11 | ['joint-entity-and-relation-extraction'] | ['natural-language-processing'] | [-1.70485526e-01 1.68898657e-01 -3.85228723e-01 -4.13079888e-01
-1.10974383e+00 -7.55477846e-01 5.92236757e-01 9.22072649e-01
-4.45638150e-01 7.35987961e-01 8.72642338e-01 -1.93472207e-01
-1.94119856e-01 -7.56078959e-01 -4.31422621e-01 6.87445849e-02
-3.96930546e-01 4.92227763e-01 2.87244260e-01 -1.44544750e-01
9.21098515e-02 4.58485514e-01 -9.72940087e-01 5.40558934e-01
5.62951267e-01 5.65452456e-01 -1.50743440e-01 5.26841462e-01
-4.11215514e-01 6.52576864e-01 -6.01887107e-01 -7.82554686e-01
-1.64775059e-01 1.06805928e-01 -1.30567575e+00 -5.95308065e-01
4.10695970e-01 1.25383094e-01 -6.39258206e-01 6.68490529e-01
6.87151790e-01 2.44489387e-01 3.75907928e-01 -1.09054804e+00
-8.38540256e-01 1.21049047e+00 -4.18881744e-01 7.39242494e-01
7.07469940e-01 -1.09553784e-01 1.52738881e+00 -1.10107195e+00
9.00167465e-01 1.10666883e+00 9.34320271e-01 3.46599787e-01
-1.31053090e+00 -5.65060496e-01 2.28412896e-01 3.37464184e-01
-1.37412190e+00 -6.53680384e-01 5.43831587e-01 -2.77504653e-01
1.72208488e+00 2.31785074e-01 2.16598094e-01 1.42360520e+00
2.85921633e-01 7.45008588e-01 5.51778555e-01 -2.79397219e-01
-1.92725062e-01 -3.03949773e-01 8.21102440e-01 6.12564802e-01
4.40461844e-01 -1.54635668e-01 -8.48561049e-01 -5.50147593e-01
5.32499626e-02 -2.01512516e-01 -3.26709539e-01 8.95610824e-02
-1.19962311e+00 6.56168878e-01 2.82970279e-01 6.30966544e-01
-4.39063668e-01 6.09698333e-02 6.94322884e-01 1.24016672e-01
5.88638604e-01 7.71827102e-01 -9.40178216e-01 -2.20778033e-01
-8.19577157e-01 4.07268703e-01 1.08942890e+00 9.68931019e-01
5.61490357e-01 -5.02006054e-01 -4.92100924e-01 8.53027701e-01
-8.31243861e-03 -1.13047786e-01 2.84813583e-01 -6.45629585e-01
8.89961421e-01 5.31872392e-01 5.04899733e-02 -9.68527198e-01
-8.51881921e-01 -2.32137635e-01 -3.25857580e-01 -5.85072100e-01
3.85119975e-01 -4.39965069e-01 -6.42798901e-01 1.91811633e+00
4.28820282e-01 3.86583030e-01 5.46061806e-02 4.39120233e-01
1.25441158e+00 5.74723125e-01 5.26619911e-01 -5.71489632e-02
1.90162313e+00 -8.58277261e-01 -7.93699265e-01 -5.71727872e-01
8.57781649e-01 -7.03374684e-01 7.02510178e-01 -2.37821877e-01
-1.05290616e+00 -2.57866263e-01 -7.09264338e-01 -5.32473564e-01
-6.08292341e-01 -1.29655361e-01 5.27193248e-01 6.36653528e-02
-7.26238787e-01 6.48492992e-01 -9.83526945e-01 -4.64248896e-01
2.74081141e-01 1.56981181e-02 -6.39858484e-01 -3.90939340e-02
-1.77141321e+00 1.26611686e+00 6.34407222e-01 -2.30776802e-01
-4.21671122e-01 -1.11030328e+00 -1.29790008e+00 3.55494052e-01
3.69191974e-01 -7.23790288e-01 1.03628600e+00 -4.31264639e-02
-6.16160572e-01 9.89847243e-01 -3.83843690e-01 -6.48539782e-01
-1.38014138e-01 -5.96629024e-01 -8.39449942e-01 8.16647708e-02
2.85815418e-01 4.22580570e-01 1.57113299e-02 -6.85675681e-01
-2.67327845e-01 -3.54901612e-01 -5.03338315e-02 4.39264439e-02
-3.14689010e-01 5.33442736e-01 -2.96156198e-01 -7.31333613e-01
-2.94185072e-01 -5.92289686e-01 -8.60017166e-02 -3.59804094e-01
-6.68820918e-01 -6.94783092e-01 5.43996334e-01 -8.93317878e-01
1.84979498e+00 -2.01374722e+00 2.60084607e-02 -2.33592927e-01
2.87794769e-01 1.62586078e-01 -3.60015571e-01 9.42835987e-01
-3.00195009e-01 3.16036075e-01 -2.30403811e-01 -5.09358525e-01
1.03336222e-01 -8.96966085e-02 -1.07651114e-01 9.19559970e-02
6.96499705e-01 1.06371689e+00 -1.00841415e+00 -6.84566915e-01
-2.30393216e-01 4.76390988e-01 -1.53393015e-01 -5.85072720e-03
-1.06874760e-02 3.65562066e-02 -4.08595473e-01 5.35558760e-01
3.15482020e-01 -4.48847055e-01 3.35047364e-01 -3.33316714e-01
8.29572827e-02 1.08051670e+00 -9.85815048e-01 1.86949456e+00
-5.31231701e-01 6.43171370e-01 -2.25050762e-01 -7.83748329e-01
7.99455822e-01 3.98517787e-01 4.65924531e-01 -2.97132075e-01
-1.19235694e-01 -1.46209568e-01 -2.43382350e-01 -6.21026218e-01
7.18124628e-01 2.79395193e-01 -5.56804538e-01 4.96675819e-01
5.26446998e-01 3.88722181e-01 5.31832039e-01 5.24384916e-01
1.59106183e+00 -3.78695386e-03 8.23470712e-01 -2.83311866e-02
1.33045107e-01 -1.03544645e-01 8.87017190e-01 4.10584837e-01
-2.67669022e-01 2.76770085e-01 6.59340382e-01 -3.85051876e-01
-8.06618392e-01 -1.17667818e+00 -4.10105526e-01 1.30391991e+00
-8.05841982e-02 -9.36780334e-01 -2.24434927e-01 -9.88273919e-01
3.07059377e-01 9.52541947e-01 -6.84951067e-01 -6.85424432e-02
-7.50004590e-01 -6.20778441e-01 9.17491972e-01 9.03488696e-01
8.74407366e-02 -1.14880383e+00 -2.66623825e-01 5.81300378e-01
-4.85403627e-01 -1.25839436e+00 -8.05860519e-01 2.94807017e-01
-5.10746956e-01 -1.20680022e+00 -2.14852437e-01 -7.43704736e-01
1.00709289e-01 -2.63400078e-01 1.68418944e+00 -1.48737639e-01
-4.99183923e-01 3.72347027e-01 -3.51214468e-01 -2.15990573e-01
-9.97957215e-02 5.20272791e-01 -1.38259411e-01 -4.96621668e-01
8.69041324e-01 -6.18764579e-01 -3.97279024e-01 -3.66299301e-02
-6.59376144e-01 -4.86824065e-01 2.82737762e-01 7.49684095e-01
3.70137900e-01 -4.78659064e-01 8.80613327e-01 -1.12586808e+00
7.73121297e-01 -9.54776764e-01 -1.72191560e-01 5.27203739e-01
-5.50330102e-01 2.07896441e-01 2.38738224e-01 -3.10670167e-01
-1.06216359e+00 -2.17218027e-01 -2.14076638e-01 -4.88494597e-02
-3.62133592e-01 8.20641041e-01 -5.89861795e-02 5.74536026e-01
6.56766951e-01 -1.76709384e-01 -6.41820550e-01 -5.70744097e-01
6.75614417e-01 4.89126205e-01 4.85640973e-01 -7.35481918e-01
5.02096415e-01 1.06998943e-01 -4.08140182e-01 -4.45687920e-01
-1.06749272e+00 -6.60829484e-01 -7.34213829e-01 2.61753887e-01
8.35662246e-01 -9.11423564e-01 -2.98182279e-01 4.83045205e-02
-1.43776965e+00 -4.56669815e-02 -2.81553030e-01 3.45531106e-01
2.35435013e-02 3.63738626e-01 -1.05886698e+00 -3.66445124e-01
-7.17936933e-01 -5.39065897e-01 1.09171569e+00 2.66974568e-01
-7.82878578e-01 -1.29598820e+00 4.21967894e-01 2.32231364e-01
1.57591760e-01 3.71189594e-01 9.72433746e-01 -1.28551984e+00
-1.21722117e-01 -1.07598178e-01 -1.41109720e-01 -2.99684733e-01
9.13526416e-02 2.42759865e-02 -7.57599413e-01 -1.64842810e-02
-6.32895291e-01 -3.62584412e-01 1.07018137e+00 9.59643945e-02
9.50879753e-01 -4.41628307e-01 -7.91502059e-01 4.37969953e-01
1.22162426e+00 -1.81124777e-01 4.08511609e-01 4.22504455e-01
4.54820246e-01 5.41588724e-01 5.35782754e-01 5.57329774e-01
7.95319736e-01 6.38893485e-01 -2.46656798e-02 3.20855349e-01
-2.15810135e-01 -2.60451883e-01 2.84544468e-01 6.45941973e-01
2.82541066e-01 -3.77795696e-01 -1.17533731e+00 1.04958069e+00
-1.72054589e+00 -1.22241759e+00 -1.13885239e-01 1.74178600e+00
1.27189565e+00 2.19823554e-01 -1.16299326e-02 -3.07880461e-01
8.44262540e-01 5.73590159e-01 -4.32098001e-01 -5.92814267e-01
-4.31680493e-02 4.55716938e-01 2.37881303e-01 4.38715905e-01
-1.33877146e+00 1.00881219e+00 6.00298357e+00 5.36952972e-01
-6.80440009e-01 2.70433545e-01 3.34949344e-01 -1.44813687e-01
-3.96030873e-01 3.41918692e-02 -1.18430698e+00 4.11816239e-01
1.28769183e+00 -5.54235399e-01 -1.76386517e-02 6.28400445e-01
-2.04576612e-01 1.89538851e-01 -1.25952530e+00 4.85551238e-01
2.16120016e-02 -1.45817387e+00 -2.56765038e-01 -2.93337077e-01
4.19038832e-01 2.76166409e-01 -3.16275865e-01 5.17819941e-01
5.47549963e-01 -7.47756600e-01 2.96542555e-01 5.39203107e-01
7.89056122e-01 -6.31807864e-01 7.37989128e-01 -4.36991304e-02
-1.69047928e+00 8.96131396e-02 -1.90913789e-02 3.46138269e-01
6.35121226e-01 9.23843145e-01 -8.37041676e-01 8.65072727e-01
6.29800379e-01 8.34920049e-01 -7.33710110e-01 9.54773545e-01
-3.55341166e-01 5.35741508e-01 -2.95338362e-01 3.94212864e-02
-2.10028123e-02 3.47069353e-01 7.43056238e-01 1.81898141e+00
1.11232154e-01 1.22957796e-01 1.65124789e-01 7.43033230e-01
-4.33370739e-01 1.05678730e-01 -6.18691504e-01 -2.28901669e-01
1.24335325e+00 1.31067014e+00 -4.27653313e-01 -3.77736539e-01
-5.13076782e-01 7.92144775e-01 9.18620884e-01 2.41923347e-01
-9.10532951e-01 -9.09447432e-01 9.60501373e-01 -2.40488529e-01
4.22733784e-01 -2.37009898e-01 -4.65143204e-01 -1.34624934e+00
1.22865001e-02 -5.02026021e-01 1.01862824e+00 -5.13687968e-01
-1.57027388e+00 6.61291122e-01 1.52113035e-01 -6.92428946e-01
-1.49489969e-01 -3.56093258e-01 -8.76572728e-01 6.17771626e-01
-1.35322452e+00 -1.04331958e+00 -8.82680994e-03 4.08722371e-01
4.06494319e-01 1.44557059e-01 1.16220796e+00 3.99454534e-01
-7.78965771e-01 7.25862920e-01 -2.65103102e-01 7.18545437e-01
1.01753271e+00 -1.26165938e+00 8.46582294e-01 9.26440835e-01
4.70697612e-01 8.03841889e-01 4.59055752e-01 -7.24833369e-01
-9.21122670e-01 -1.24461627e+00 1.70041335e+00 -8.03698838e-01
1.08648622e+00 -3.58849287e-01 -1.10678554e+00 1.19468546e+00
4.81308699e-01 1.25277147e-01 1.16528440e+00 9.28373098e-01
-9.08542514e-01 6.35990798e-02 -8.86166394e-01 3.00303102e-01
1.41702962e+00 -6.76602900e-01 -1.15841985e+00 2.98731774e-01
9.86083508e-01 -4.30880815e-01 -1.41013265e+00 3.97135586e-01
2.56118536e-01 -4.50540423e-01 9.46489811e-01 -1.07185423e+00
5.58640897e-01 1.10266529e-01 1.15964457e-01 -1.38527787e+00
-4.41168696e-01 -5.82023859e-01 -7.45442212e-01 1.68160582e+00
9.58026826e-01 -6.55886114e-01 2.84537613e-01 6.91410542e-01
-3.08333397e-01 -9.14346814e-01 -8.85454357e-01 -7.09706068e-01
6.79540038e-02 -3.26433867e-01 7.76325762e-01 1.31221104e+00
5.41473687e-01 6.79527402e-01 8.32817778e-02 4.94391739e-01
3.91945809e-01 3.34232181e-01 1.67304039e-01 -1.12557352e+00
-1.73599049e-01 -4.44470495e-01 -2.67225057e-01 -6.59863114e-01
6.06707871e-01 -1.12314379e+00 -2.45520324e-01 -1.79422867e+00
1.71623990e-01 -4.16597843e-01 -5.15270650e-01 9.20511961e-01
-6.88458085e-01 -2.40517050e-01 1.67478144e-01 -8.07290375e-02
-7.97335088e-01 3.20373893e-01 5.20615757e-01 -2.69270334e-02
-1.56373367e-01 -2.87090063e-01 -8.34314406e-01 4.43002969e-01
8.53776693e-01 -7.26232648e-01 3.64585482e-02 -7.50334024e-01
1.82481587e-01 1.84876904e-01 4.31766622e-02 -7.86383450e-01
3.50137442e-01 6.40512705e-02 2.61211455e-01 -4.88838106e-01
1.91112608e-01 -2.74427563e-01 6.73092753e-02 1.11851133e-01
-7.00703740e-01 4.24818188e-01 3.73365402e-01 5.01188755e-01
-2.75434166e-01 -2.11269066e-01 4.65638191e-01 -6.20837919e-02
-7.15694129e-01 3.25036585e-01 3.45412754e-02 7.31727540e-01
9.63873148e-01 4.47791159e-01 -7.90261865e-01 -2.70017087e-02
-1.02744937e+00 5.12436926e-01 1.91268325e-02 6.21307671e-01
5.27423441e-01 -1.38312316e+00 -8.73691678e-01 -3.05816114e-01
3.86784226e-01 -1.33707106e-01 1.93750545e-01 6.69589400e-01
-1.87412743e-02 3.88895541e-01 1.20707683e-01 -1.43033005e-02
-1.44178319e+00 5.71616948e-01 8.33580270e-02 -9.08725739e-01
-6.41401649e-01 1.05996060e+00 -3.40640664e-01 -4.70682353e-01
9.08813104e-02 -3.86697590e-01 -3.52845043e-01 4.56953913e-01
6.89024031e-01 4.25523549e-01 1.23920135e-01 -3.52103651e-01
-7.68303216e-01 2.40123063e-01 -4.57317263e-01 1.20830931e-01
1.50338459e+00 -2.37622596e-02 -1.44543082e-01 4.10047293e-01
1.31925845e+00 1.83509067e-01 -7.63537407e-01 -6.03577018e-01
6.82259619e-01 -8.95897858e-03 -2.16921598e-01 -7.82320559e-01
-6.36712968e-01 5.82594812e-01 -1.50442764e-01 2.72263288e-01
8.35761011e-01 4.67624187e-01 1.09609497e+00 3.04626226e-01
3.38809133e-01 -8.49686384e-01 -9.71933082e-02 7.36944437e-01
7.68831968e-01 -1.06901169e+00 4.28840220e-02 -3.57168138e-01
-8.59353244e-01 1.11022520e+00 6.54089689e-01 -1.56846475e-02
7.83607960e-01 5.26552916e-01 -2.79162198e-01 -3.49888265e-01
-1.20647454e+00 -3.32758307e-01 4.00604546e-01 3.96657974e-01
8.91528845e-01 7.32749850e-02 -4.91741329e-01 8.96846890e-01
-9.38391015e-02 -2.71317512e-01 2.43883342e-01 8.56185317e-01
-2.85545856e-01 -1.20945072e+00 2.22753789e-02 5.12275398e-01
-7.10613608e-01 -4.99618232e-01 -4.02123690e-01 6.97977126e-01
1.38728190e-02 8.90393615e-01 1.97383463e-01 -1.79392099e-01
4.76771683e-01 6.49176300e-01 2.00362608e-01 -9.15193021e-01
-7.72916913e-01 -3.70121419e-01 8.88086438e-01 -7.12327361e-01
-3.00802618e-01 -8.92934442e-01 -1.39239490e+00 -2.22058341e-01
-2.29342401e-01 2.79101819e-01 1.77025035e-01 7.92966425e-01
9.02367055e-01 6.18724525e-01 3.09269905e-01 -2.50587612e-01
-3.79290253e-01 -1.21795738e+00 -2.21146822e-01 5.87855041e-01
1.35762557e-01 -6.20208681e-01 -1.79038107e-01 -1.48862660e-01] | [9.314842224121094, 9.08706283569336] |
83d7db60-32c0-43ca-a36d-dc1845b112a7 | 3d-pose-estimation-and-future-motion | 2111.13285 | null | https://arxiv.org/abs/2111.13285v1 | https://arxiv.org/pdf/2111.13285v1.pdf | 3D Pose Estimation and Future Motion Prediction from 2D Images | This paper considers to jointly tackle the highly correlated tasks of estimating 3D human body poses and predicting future 3D motions from RGB image sequences. Based on Lie algebra pose representation, a novel self-projection mechanism is proposed that naturally preserves human motion kinematics. This is further facilitated by a sequence-to-sequence multi-task architecture based on an encoder-decoder topology, which enables us to tap into the common ground shared by both tasks. Finally, a global refinement module is proposed to boost the performance of our framework. The effectiveness of our approach, called PoseMoNet, is demonstrated by ablation tests and empirical evaluations on Human3.6M and HumanEva-I benchmark, where competitive performance is obtained comparing to the state-of-the-arts. | ['Li Cheng', 'Minglun Gong', 'Sen Wang', 'Xinxin Zuo', 'Youdong Ma', 'Ji Yang'] | 2021-11-26 | null | null | null | null | ['3d-pose-estimation'] | ['computer-vision'] | [ 4.50609140e-02 9.23634171e-02 -8.12739059e-02 -1.56695545e-01
-7.22144127e-01 -1.18318334e-01 5.82970798e-01 -5.56107879e-01
-6.57747626e-01 4.79631484e-01 5.46207666e-01 3.32998931e-01
3.23218286e-01 -1.31945550e-01 -7.80543804e-01 -4.51010704e-01
-1.73627347e-01 3.32361400e-01 1.99849233e-01 -3.33854914e-01
-2.12349415e-01 -2.89559364e-02 -1.39796925e+00 1.05733626e-01
4.57888216e-01 8.41500401e-01 1.97812989e-01 6.98186159e-01
5.20892978e-01 8.14917564e-01 -1.65507421e-01 -5.15696347e-01
4.51005369e-01 -4.68099087e-01 -8.62441421e-01 3.06110024e-01
3.90269876e-01 -3.79502326e-01 -4.58498657e-01 7.71738291e-01
7.25663900e-01 2.45045260e-01 3.63885015e-01 -9.91472304e-01
-1.48052096e-01 1.23126127e-01 -5.49592674e-01 -7.31620193e-02
7.29338527e-01 3.70836407e-01 1.00691080e+00 -8.62886608e-01
7.98663020e-01 1.32590759e+00 7.27493763e-01 6.49337351e-01
-1.09417140e+00 -3.91774476e-01 -7.56560452e-03 3.25511694e-01
-1.49712813e+00 -3.82927120e-01 7.15876281e-01 -4.03156906e-01
1.06279659e+00 2.56850161e-02 1.02012455e+00 1.52609456e+00
5.12087226e-01 1.00952983e+00 7.42767274e-01 -1.05052650e-01
-1.97125785e-02 -5.81270576e-01 -3.60714138e-01 9.65298116e-01
-1.79260119e-03 2.25579202e-01 -1.07342553e+00 1.71796098e-01
9.16468561e-01 -1.42810985e-01 -3.46929997e-01 -7.26120710e-01
-1.62371135e+00 4.17482793e-01 5.00834882e-01 -8.85678828e-02
-6.32132292e-01 4.90670055e-01 6.50041461e-01 3.61490250e-03
3.21075797e-01 5.79283573e-02 -2.57007569e-01 -3.85650575e-01
-9.64683056e-01 5.80389738e-01 7.21678674e-01 1.11637676e+00
3.36727858e-01 -1.79550052e-02 -2.57872343e-01 3.91592681e-01
7.10797310e-01 4.34825957e-01 3.62820715e-01 -8.60082865e-01
6.02915049e-01 4.91157025e-01 1.53334914e-02 -8.23758125e-01
-7.34720826e-01 -4.01858777e-01 -9.61903334e-01 5.50772063e-02
2.07980543e-01 1.06171846e-01 -8.47365022e-01 1.79847217e+00
6.14895582e-01 3.69974464e-01 9.46188942e-02 1.52501130e+00
5.21939933e-01 3.50245476e-01 1.03516258e-01 1.46274909e-01
1.40334058e+00 -1.21918511e+00 -5.50024033e-01 -3.84509981e-01
3.94105077e-01 -2.77480543e-01 9.71818805e-01 3.72512132e-01
-1.19163525e+00 -9.37080264e-01 -1.13422167e+00 -3.38812500e-01
3.35646927e-01 1.55425400e-01 4.84617114e-01 3.63036454e-01
-8.95093024e-01 6.55367076e-01 -1.41969371e+00 -4.25905108e-01
1.64896309e-01 3.83616686e-01 -6.76756501e-01 2.92156875e-01
-1.06258416e+00 1.01838887e+00 5.49472392e-01 3.83443713e-01
-9.52598870e-01 -4.76495117e-01 -1.19537306e+00 -3.58761579e-01
3.81660253e-01 -1.42139137e+00 1.21074927e+00 -4.58909184e-01
-1.98479903e+00 9.62247849e-01 -4.45296541e-02 -7.07224190e-01
9.37834084e-01 -9.07821059e-01 1.01792011e-02 3.41263056e-01
5.22172526e-02 8.45345855e-01 8.73459280e-01 -7.50795066e-01
-5.67088842e-01 -5.60398817e-01 -2.91742891e-01 6.17305994e-01
-8.39972049e-02 -6.93103299e-02 -1.05271649e+00 -6.87018812e-01
2.21753150e-01 -1.38597560e+00 -5.08577168e-01 -6.46724552e-02
-4.58216667e-01 8.48699510e-02 2.68416673e-01 -9.76565063e-01
1.03938591e+00 -1.92498696e+00 1.18748641e+00 2.36700699e-01
1.67120203e-01 7.71337077e-02 1.09136827e-01 2.75822639e-01
1.37395551e-02 -5.17388284e-01 -3.00039858e-01 -8.33155274e-01
1.30356759e-01 1.83165178e-01 -2.86461990e-02 8.80170345e-01
1.07038334e-01 1.17803144e+00 -6.15772128e-01 -4.03709024e-01
3.71051282e-01 5.36516249e-01 -5.80310106e-01 3.84939790e-01
-1.01744615e-01 1.03368127e+00 -2.42111757e-01 5.24860263e-01
2.24491969e-01 -4.51778471e-01 2.52304465e-01 -2.87997127e-01
1.33095846e-01 2.60618865e-01 -1.12743247e+00 2.86422968e+00
-2.03015611e-01 -1.18052691e-01 -1.07726097e-01 -6.32073641e-01
7.70277321e-01 5.35061300e-01 6.36923075e-01 -6.39391840e-01
3.09719861e-01 9.69948340e-03 -6.86197132e-02 -4.29571658e-01
4.67422634e-01 -4.51238528e-02 -3.20076644e-01 6.88565969e-02
2.43503302e-01 9.17632505e-02 -3.65884230e-02 -4.09862250e-02
1.00651610e+00 1.15727913e+00 6.19947553e-01 -3.02158087e-01
7.30888605e-01 -8.31602737e-02 6.64467514e-01 2.37639368e-01
-3.41674596e-01 7.93117642e-01 9.66001973e-02 -5.28115392e-01
-1.18705642e+00 -1.26209450e+00 3.82860571e-01 8.64365101e-01
2.56007731e-01 -7.30079651e-01 -7.45667338e-01 -4.87675369e-01
-1.50478661e-01 2.17373595e-01 -6.44198120e-01 -2.85202324e-01
-8.67719471e-01 -4.13974881e-01 6.48643911e-01 8.25905085e-01
6.05570912e-01 -8.68343234e-01 -1.31206465e+00 3.34507793e-01
-5.81160724e-01 -1.55054700e+00 -6.28994823e-01 -1.63820416e-01
-8.94592464e-01 -1.04220355e+00 -9.16497588e-01 -4.19888586e-01
1.30622581e-01 -1.50048137e-01 1.01061594e+00 -3.20387959e-01
-3.40668708e-01 5.23204863e-01 -3.88633549e-01 2.33516078e-02
-2.36907408e-01 1.94549590e-01 3.54444385e-01 5.45720048e-02
9.42065641e-02 -7.66780317e-01 -7.82472193e-01 2.84159660e-01
-4.57038939e-01 5.64872861e-01 7.23611057e-01 7.43254483e-01
6.54039085e-01 -8.20476949e-01 9.21317115e-02 -1.80120140e-01
5.32665814e-04 -1.44635513e-01 -3.29798102e-01 1.98710859e-01
-2.27377981e-01 2.92779654e-01 2.59667128e-01 -3.16098124e-01
-9.52529728e-01 7.11087644e-01 -2.97296166e-01 -8.30497563e-01
-3.04433219e-02 2.35337093e-01 -3.38676423e-01 1.00032613e-01
4.64789748e-01 4.18504357e-01 2.94458210e-01 -6.10177934e-01
5.56040287e-01 1.76272258e-01 1.17771935e+00 -5.14954388e-01
7.76202261e-01 6.74803376e-01 2.78502405e-01 -6.03336871e-01
-5.33021629e-01 -6.33928895e-01 -1.11388421e+00 -2.88881809e-01
1.28910124e+00 -1.45859921e+00 -9.35296416e-01 5.36232710e-01
-1.26538348e+00 -2.47425154e-01 -9.35178474e-02 7.61422813e-01
-1.20586979e+00 5.50927818e-01 -7.96856284e-01 -6.44262433e-01
-6.32707655e-01 -1.18851292e+00 1.47884953e+00 -2.02338248e-01
-6.23385072e-01 -7.12861419e-01 3.61947984e-01 4.37663496e-01
-1.27920404e-01 4.95011866e-01 2.81539321e-01 -3.79451782e-01
-5.75831234e-01 3.87474597e-02 1.46000206e-01 1.74474686e-01
-2.48095766e-01 -6.83300257e-01 -6.54121816e-01 -4.87779498e-01
9.59060388e-04 -5.00913262e-01 7.84062684e-01 1.36057451e-01
6.91794336e-01 -8.32213089e-02 -2.06179887e-01 1.00342155e+00
1.02669215e+00 -4.29600507e-01 8.87465537e-01 3.91166985e-01
1.16094935e+00 5.74794412e-01 6.77316666e-01 6.26782715e-01
8.78954530e-01 1.14915824e+00 3.04612339e-01 6.18326738e-02
-3.30564111e-01 -5.16290128e-01 6.47579372e-01 1.10164046e+00
-4.71725225e-01 2.56828874e-01 -9.23664808e-01 2.99573332e-01
-2.14690137e+00 -6.16432071e-01 -4.95050587e-02 1.98830760e+00
7.26820290e-01 2.40585264e-02 4.10447001e-01 -5.07523306e-02
3.03217977e-01 2.84548461e-01 -5.41332662e-01 1.21782832e-01
5.79840913e-02 3.21325421e-01 3.74090403e-01 2.48797700e-01
-1.30008888e+00 1.12696218e+00 6.03588867e+00 4.66017783e-01
-8.10550451e-01 1.76604182e-01 2.03050929e-03 -3.28587860e-01
3.09086978e-01 -1.87765315e-01 -7.29408622e-01 2.17111319e-01
7.98394263e-01 -4.14191969e-02 2.35448763e-01 7.26511240e-01
1.70544162e-01 1.39844000e-01 -1.33914721e+00 1.27689719e+00
3.88499260e-01 -9.26088393e-01 -8.11449736e-02 1.40419468e-01
4.49146032e-01 1.17121384e-01 -1.35121182e-01 1.11228697e-01
4.84764241e-02 -7.81727612e-01 1.23361802e+00 7.38226295e-01
7.32444942e-01 -6.84801996e-01 3.82329047e-01 3.91857922e-01
-1.50247967e+00 1.36862025e-01 3.57188880e-02 -2.54193425e-01
5.82278192e-01 1.67734265e-01 -6.51529312e-01 1.07470083e+00
8.40710998e-01 1.07026768e+00 -4.87998217e-01 8.53383243e-01
-5.44012129e-01 2.60707498e-01 -2.33164355e-01 3.07186007e-01
-6.71725022e-04 -9.89229381e-02 7.32412279e-01 1.07900310e+00
2.60187119e-01 8.60600397e-02 5.20249963e-01 6.85917616e-01
1.89776391e-01 2.74394024e-02 -6.00073874e-01 3.97439629e-01
-2.24575415e-01 1.12155652e+00 -3.63161564e-01 -1.58149615e-01
-3.99253190e-01 1.69911849e+00 2.59064406e-01 1.87837705e-01
-1.02663398e+00 1.54304028e-01 8.41990411e-01 -1.62300333e-01
5.01493156e-01 -6.68896914e-01 -1.12108976e-01 -1.37414229e+00
2.68364877e-01 -9.94010627e-01 3.44452769e-01 -4.93082821e-01
-9.86308098e-01 4.62261736e-01 -1.51062226e-02 -1.46413910e+00
-7.32449532e-01 -4.58521515e-01 -1.11420028e-01 7.42439866e-01
-1.01855266e+00 -1.55733299e+00 -2.11268336e-01 7.93271244e-01
4.03680593e-01 6.73144534e-02 7.61959791e-01 2.06773013e-01
-4.41775739e-01 5.40554464e-01 -5.09273767e-01 8.16879869e-02
5.99012196e-01 -1.16021991e+00 1.11259031e+00 1.02052450e+00
2.80957460e-01 4.47434366e-01 6.89058483e-01 -7.93940008e-01
-1.79297554e+00 -1.06900382e+00 6.81839943e-01 -6.97078645e-01
4.90014523e-01 -6.49882734e-01 -7.44625926e-01 7.86887646e-01
-3.56793664e-02 3.68527323e-03 5.74249148e-01 5.01010083e-02
-3.62893701e-01 2.16067225e-01 -6.20209932e-01 6.73919797e-01
1.55739629e+00 -5.27844310e-01 -8.41127098e-01 1.06634066e-01
6.19371116e-01 -8.88290584e-01 -9.93707955e-01 5.55951715e-01
8.34585488e-01 -9.42976296e-01 1.18136334e+00 -6.54052556e-01
5.33965647e-01 -3.64432901e-01 -2.50031948e-01 -1.12022531e+00
-3.32069248e-01 -8.82558823e-01 -6.15284860e-01 6.32487476e-01
-3.38558666e-02 1.10439993e-01 9.50536370e-01 3.02128345e-01
-6.91849813e-02 -5.91457605e-01 -1.18259299e+00 -8.79678309e-01
-2.59824216e-01 -4.65815008e-01 1.91553667e-01 3.65048409e-01
1.75479427e-02 5.05918026e-01 -1.12900174e+00 1.78973243e-01
8.09874415e-01 1.23498598e-02 1.26940262e+00 -9.30348575e-01
-6.82867408e-01 -6.83946311e-02 -7.83234060e-01 -1.47860444e+00
2.80130863e-01 -8.49508107e-01 3.70291531e-01 -1.15807486e+00
2.30267167e-01 5.02853096e-01 -1.69410437e-01 4.75077093e-01
-3.19554210e-01 4.33405340e-01 6.39311671e-01 3.24525952e-01
-1.07253861e+00 1.12010968e+00 1.18611073e+00 2.63556927e-01
-2.24679381e-01 -1.23786725e-01 -2.21648380e-01 8.67157996e-01
3.01139951e-01 -2.20011428e-01 -1.59865409e-01 -4.73388970e-01
-1.86913423e-02 2.99801975e-01 7.97229767e-01 -1.48664892e+00
3.18818092e-01 3.44446599e-01 4.00888622e-01 -6.42446458e-01
6.95209861e-01 -6.32806540e-01 2.75719821e-01 7.23204851e-01
-1.01366796e-01 2.53662467e-01 1.17511451e-01 7.43503809e-01
-1.14617087e-01 6.44640982e-01 5.28875887e-01 -9.96843129e-02
-1.01106834e+00 4.32312906e-01 1.19038587e-02 -5.86632453e-03
9.26288545e-01 -2.22163811e-01 4.43765819e-01 -2.32235551e-01
-8.80473852e-01 3.02536011e-01 3.99682760e-01 7.23338664e-01
7.31076300e-01 -1.54965854e+00 -9.17156577e-01 1.66721433e-01
3.55904669e-01 -4.14072275e-02 2.53144145e-01 1.11861038e+00
-4.65717256e-01 4.46644425e-01 -4.11107033e-01 -8.98150861e-01
-1.27999258e+00 3.81463230e-01 2.47108415e-01 -4.29455042e-01
-1.12563276e+00 6.94689572e-01 1.32104620e-01 -4.53728318e-01
3.54372650e-01 -3.51259172e-01 1.78919770e-02 -4.14309353e-01
4.88542438e-01 4.46619689e-01 -1.08280480e-01 -1.04091656e+00
-6.70987189e-01 5.81247687e-01 2.43588582e-01 -4.95225191e-01
1.17897582e+00 -3.31406325e-01 1.20532565e-01 5.47356784e-01
1.14133823e+00 -4.81408805e-01 -1.72691667e+00 -2.88733602e-01
1.34558275e-01 -3.51761669e-01 -4.21313345e-01 -5.82843423e-01
-7.47792363e-01 8.83121610e-01 6.08550012e-01 -8.68481398e-01
1.10010087e+00 -4.93258163e-02 9.26862419e-01 4.40730393e-01
7.26489782e-01 -9.27106082e-01 2.21289352e-01 6.38471305e-01
1.04323816e+00 -1.15495980e+00 7.43296146e-02 -3.70378286e-01
-8.65741193e-01 8.32781196e-01 6.05565727e-01 -2.34719738e-01
2.88051099e-01 9.69594195e-02 -8.04662928e-02 -6.46437556e-02
-6.67481780e-01 -4.33492690e-01 7.53565192e-01 4.19796526e-01
3.61235529e-01 -5.76714613e-02 -3.96453887e-01 6.98396742e-01
-3.10317546e-01 2.78937191e-01 -1.45690888e-01 8.56824219e-01
-2.61857081e-02 -1.08027601e+00 -3.72126728e-01 -2.78477669e-01
-3.64482582e-01 1.61919698e-01 -2.12427169e-01 7.73842931e-01
4.89005726e-03 5.30875444e-01 -8.13858360e-02 -7.81626046e-01
5.02451301e-01 1.26202971e-01 8.37931454e-01 -4.68085557e-01
-7.11671472e-01 2.83502251e-01 -3.40500940e-03 -1.34950137e+00
-6.10766053e-01 -7.03146398e-01 -1.23290634e+00 -5.71275130e-02
2.94634789e-01 -3.09951633e-01 5.04274070e-01 9.53257382e-01
3.73954028e-01 6.14439905e-01 9.02379081e-02 -1.24366307e+00
-8.00468922e-01 -9.69493449e-01 -4.31753665e-01 7.37242579e-01
1.56648889e-01 -6.93959057e-01 2.91419327e-01 2.05432698e-01] | [7.1775641441345215, -0.6116892099380493] |
1ae6c26c-da89-4594-b43a-5737c126ad2f | domain-generalization-for-mammographic-image | 2304.10226 | null | https://arxiv.org/abs/2304.10226v4 | https://arxiv.org/pdf/2304.10226v4.pdf | Domain Generalization for Mammographic Image Analysis with Contrastive Learning | The deep learning technique has been shown to be effectively addressed several image analysis tasks in the computer-aided diagnosis scheme for mammography. The training of an efficacious deep learning model requires large data with diverse styles and qualities. The diversity of data often comes from the use of various scanners of vendors. But, in practice, it is impractical to collect a sufficient amount of diverse data for training. To this end, a novel contrastive learning is developed to equip the deep learning models with better style generalization capability. Specifically, the multi-style and multi-view unsupervised self-learning scheme is carried out to seek robust feature embedding against style diversity as a pretrained model. Afterward, the pretrained network is further fine-tuned to the downstream tasks, e.g., mass detection, matching, BI-RADS rating, and breast density classification. The proposed method has been evaluated extensively and rigorously with mammograms from various vendor style domains and several public datasets. The experimental results suggest that the proposed domain generalization method can effectively improve performance of four mammographic image tasks on the data from both seen and unseen domains, and outperform many state-of-the-art (SOTA) generalization methods. | ['Jie-Zhi Cheng', 'Dinggang Shen', 'Chunling Liu', 'Zaiyi Liu', 'Yajia Gu', 'Xiangyu Zhao', 'Dongdong Chen', 'Xi Ouyang', 'Chenjin Lei', 'Sheng Wang', 'Lichi Zhang', 'Zhiming Cui', 'Zheren Li'] | 2023-04-20 | null | null | null | null | ['style-generalization', 'breast-density-classification', 'self-learning'] | ['computer-vision', 'medical', 'natural-language-processing'] | [ 3.10728788e-01 6.25399798e-02 -1.69073865e-01 -8.76994431e-01
-7.92537868e-01 -8.41748416e-02 2.96686769e-01 5.27530946e-02
-2.93044090e-01 4.60087925e-01 6.72653392e-02 -3.83959591e-01
-2.51898497e-01 -6.68989062e-01 -5.50062835e-01 -9.90456343e-01
1.24855831e-01 4.67203945e-01 8.23672563e-02 -2.47036219e-01
-1.08305708e-01 5.71002901e-01 -1.08049726e+00 4.18239683e-01
7.62503207e-01 1.33518183e+00 3.92954499e-01 4.02002901e-01
1.63193550e-02 6.09590471e-01 -2.41943449e-01 -5.52911818e-01
3.35047394e-01 -4.32787120e-01 -4.76219833e-01 4.08486962e-01
3.13638598e-01 -5.27158797e-01 -4.79735583e-01 1.08586001e+00
8.96520019e-01 -2.79411197e-01 9.36316490e-01 -7.63305426e-01
-9.52008903e-01 3.33804995e-01 -9.30530488e-01 6.53369844e-01
-3.23710620e-01 -5.86836375e-02 3.82061779e-01 -9.37469602e-01
4.74531561e-01 1.17485332e+00 6.54774368e-01 7.96278477e-01
-8.03939700e-01 -9.26548243e-01 -2.71602154e-01 3.23917530e-02
-1.14525950e+00 -3.20668340e-01 9.69223917e-01 -3.47292513e-01
2.65196443e-01 -1.11014768e-01 2.77516663e-01 1.44289827e+00
6.97830200e-01 6.99288368e-01 1.13977456e+00 -2.29859918e-01
2.71590084e-01 6.56898618e-01 -8.03801045e-02 6.93242967e-01
4.47703779e-01 5.84124252e-02 -2.01916113e-01 -1.46169901e-01
8.57042313e-01 2.16578618e-01 3.62407677e-02 -6.96694851e-01
-9.51110363e-01 8.94261539e-01 7.26308405e-01 3.93850982e-01
-3.28701586e-01 -5.19101381e-01 9.16893601e-01 4.23403710e-01
5.54138541e-01 1.53893918e-01 -4.12765682e-01 4.21076953e-01
-6.45013332e-01 -1.15436636e-01 4.38490748e-01 1.16777551e+00
3.13458025e-01 1.23294927e-01 1.09162778e-01 1.04901743e+00
1.44772872e-01 4.96853322e-01 1.05311990e+00 -2.37404406e-01
5.86875319e-01 7.46784806e-01 -3.09615195e-01 -1.18009758e+00
-5.84425807e-01 -8.03451955e-01 -1.54656196e+00 -9.13310722e-02
4.17441092e-02 -8.14294815e-02 -1.24806225e+00 1.40097451e+00
4.59865779e-01 -1.06915899e-01 4.81417418e-01 7.61853933e-01
1.11233139e+00 4.42428082e-01 4.41246599e-01 -2.26342291e-01
1.35826123e+00 -6.62166655e-01 -5.79500079e-01 -2.57406712e-01
7.79364347e-01 -6.27413154e-01 9.64531660e-01 2.30366498e-01
-8.41902852e-01 -1.00681174e+00 -1.38713384e+00 8.68997201e-02
-1.90789565e-01 2.77710557e-01 5.96051931e-01 8.02102983e-01
-6.83403969e-01 2.61469603e-01 -1.03067005e+00 -5.14793038e-01
1.00702369e+00 3.41843724e-01 -5.56997895e-01 -6.16775274e-01
-1.09105921e+00 6.56250894e-01 5.34328699e-01 1.40644357e-01
-9.65528488e-01 -6.19062304e-01 -8.02336395e-01 -2.16730013e-01
1.88718475e-02 -7.13286161e-01 1.14529419e+00 -1.39591014e+00
-9.55731332e-01 1.13306463e+00 3.51337463e-01 -5.17458975e-01
5.57849586e-01 1.45719990e-01 -7.85103440e-01 1.94151044e-01
2.87419379e-01 5.80957353e-01 9.25506473e-01 -1.08906364e+00
-6.34951890e-01 -7.96989620e-01 -2.71516085e-01 3.78125906e-01
-9.30694699e-01 -2.56427050e-01 -2.52189606e-01 -8.09153795e-01
4.21587199e-01 -8.85411799e-01 -2.47649819e-01 1.41332865e-01
-2.83977520e-02 -9.87935290e-02 9.39590633e-01 -6.99421823e-01
1.02385378e+00 -2.48146296e+00 7.10738823e-02 1.06405117e-01
1.01695277e-01 8.48929286e-02 4.27866168e-02 -2.43912250e-01
-2.45948851e-01 -1.34856150e-01 -1.63112983e-01 1.25458539e-02
-4.76129621e-01 3.76407832e-01 3.09782475e-01 5.82142591e-01
4.02225703e-01 7.02149808e-01 -8.16336751e-01 -8.93425286e-01
2.91892532e-02 9.48552042e-02 -4.71525967e-01 2.72071242e-01
1.28308833e-01 5.38029552e-01 -8.91274393e-01 9.44007754e-01
9.58962798e-01 -6.00033462e-01 1.80693388e-01 -6.23572707e-01
6.00276709e-01 -3.75449538e-01 -9.00914788e-01 1.92277837e+00
-5.48748076e-01 1.72382638e-01 -2.78271884e-01 -1.45052576e+00
9.60161388e-01 1.81538865e-01 5.76450169e-01 -8.15749347e-01
2.60221541e-01 5.26239038e-01 3.50966811e-01 -8.64295125e-01
1.94315016e-01 -3.94504249e-01 -2.32216746e-01 1.18526638e-01
3.22052062e-01 1.88291848e-01 -2.02009544e-01 1.37060463e-01
9.04523730e-01 -2.67924011e-01 4.46256340e-01 -5.48555851e-01
5.74736476e-01 -1.14304721e-02 5.87870777e-01 5.70336699e-01
-2.73539096e-01 4.16877925e-01 9.55436155e-02 -5.40279210e-01
-1.26912165e+00 -1.00992525e+00 -6.30072534e-01 1.05399370e+00
1.99592844e-01 4.82654899e-01 -4.18320626e-01 -9.75263834e-01
4.21812758e-03 1.67148128e-01 -9.62466896e-01 -4.89516079e-01
-4.60778296e-01 -9.40819740e-01 4.67304051e-01 1.04322684e+00
9.68466520e-01 -8.90625238e-01 -4.43534464e-01 1.40810773e-01
1.78563610e-01 -9.47918713e-01 -3.05232435e-01 2.79254019e-01
-1.21570218e+00 -8.32795858e-01 -1.03111398e+00 -1.28234613e+00
8.16632032e-01 1.24492079e-01 9.93720889e-01 -2.21889004e-01
-4.00707871e-01 3.44280154e-01 -4.66936260e-01 -5.16327262e-01
-7.98839927e-01 2.18251422e-01 6.21099770e-02 3.83471325e-02
6.27664089e-01 -3.47174913e-01 -8.14223528e-01 4.01822627e-01
-1.27012014e+00 -7.79634574e-03 1.13452661e+00 1.49595106e+00
6.88269377e-01 1.68838382e-01 9.73373055e-01 -1.22020519e+00
3.91257912e-01 -7.91247904e-01 -1.06320553e-01 2.78805435e-01
-4.61058497e-01 -1.24425314e-01 5.20144045e-01 -6.48834109e-01
-1.38887501e+00 2.24437080e-02 -1.43719912e-01 -1.94629669e-01
-1.95335194e-01 5.85306823e-01 -3.91924322e-01 -1.42029971e-01
1.01647317e+00 3.79157901e-01 8.86410326e-02 -2.75073797e-01
-7.95845166e-02 1.07583952e+00 5.26057065e-01 -2.42670894e-01
6.10214591e-01 4.23123121e-01 4.23979945e-02 -6.91826880e-01
-8.32708001e-01 -5.31329572e-01 -4.82652277e-01 4.80589718e-02
8.16035390e-01 -1.17976797e+00 2.03921229e-01 5.75202703e-01
-5.85353911e-01 7.98796788e-02 2.23045107e-02 5.05697608e-01
-3.83556455e-01 3.13142329e-01 -5.73842704e-01 -3.57351065e-01
-4.91009861e-01 -1.25260782e+00 9.73286033e-01 5.34528673e-01
1.88720539e-01 -9.86882389e-01 -4.49778795e-01 3.17137152e-01
4.50924128e-01 1.76747784e-01 1.15905917e+00 -8.92645240e-01
-5.16026840e-02 -2.41815642e-01 -3.98405761e-01 7.16147065e-01
6.90003991e-01 -6.22715533e-01 -9.63003516e-01 -6.46361351e-01
5.73821783e-01 -7.24005580e-01 6.56520605e-01 4.34508413e-01
1.64235055e+00 5.85567057e-02 -5.44727266e-01 9.10886288e-01
1.41356850e+00 4.47795957e-01 5.01198888e-01 5.33286750e-01
4.83809590e-01 3.36607456e-01 6.81849062e-01 3.97883654e-01
1.10272497e-01 1.14160791e-01 1.93550199e-01 -4.70675170e-01
5.77184185e-03 -1.50452107e-01 -2.86961496e-01 6.88085735e-01
2.55067766e-01 -1.61895469e-01 -8.88525367e-01 5.12885451e-01
-1.38657022e+00 -6.45713627e-01 4.92443383e-01 1.61755240e+00
7.41200805e-01 3.11433703e-01 -1.91654578e-01 6.50314838e-02
5.69023371e-01 -1.20333865e-01 -1.05187738e+00 -1.77389532e-01
-1.49320781e-01 1.01359837e-01 6.44088089e-01 -4.02896792e-01
-1.40711582e+00 4.47052866e-01 5.29640341e+00 9.25044894e-01
-1.33343554e+00 1.04559787e-01 1.31964731e+00 3.03032041e-01
-1.23042027e-02 -7.41146564e-01 -6.22970343e-01 3.76007318e-01
7.05995023e-01 8.54629744e-03 -3.91080052e-01 1.25538123e+00
-1.01355001e-01 -3.34138819e-03 -1.28072858e+00 1.21619654e+00
2.89053351e-01 -1.21376526e+00 2.68990993e-01 -1.92279622e-01
8.82070720e-01 -2.04795554e-01 5.51285803e-01 5.24043441e-01
-3.22263330e-01 -9.65684295e-01 3.88483815e-02 1.92687541e-01
1.05123556e+00 -5.81002355e-01 1.02519691e+00 3.36898804e-01
-8.41344357e-01 -3.88175964e-01 -7.08559394e-01 7.26809621e-01
-2.90954083e-01 2.82005429e-01 -9.28080916e-01 8.14112008e-01
9.13361669e-01 7.66503572e-01 -8.84332776e-01 8.05257440e-01
6.74803793e-01 4.96964782e-01 -8.75647832e-03 6.02359138e-02
3.55903804e-01 3.16197574e-01 1.85842589e-02 1.09562051e+00
3.97279710e-01 3.20908844e-01 1.61292404e-01 1.27100945e-01
-1.72463030e-01 2.55605698e-01 -6.06505871e-01 4.71489206e-02
2.41446659e-01 1.21306014e+00 -6.20655596e-01 -3.28543633e-01
-5.65399289e-01 9.01968360e-01 -4.00321372e-03 1.36447996e-01
-7.22958207e-01 -1.14170834e-02 -3.68751772e-02 2.77678102e-01
2.91739196e-01 2.27256611e-01 -2.54264772e-01 -8.27798069e-01
-5.11649959e-02 -1.28418672e+00 7.73814559e-01 -5.19873321e-01
-1.90900481e+00 9.37068522e-01 9.52416435e-02 -1.53595042e+00
-2.04885587e-01 -9.03179407e-01 -3.82931978e-01 7.85294592e-01
-1.47463965e+00 -1.35552919e+00 -5.88726878e-01 8.56197238e-01
8.41149688e-01 -9.90865111e-01 7.92622447e-01 5.38013220e-01
-2.98064440e-01 1.07926011e+00 4.18667555e-01 2.85686523e-01
8.80086184e-01 -1.01590586e+00 -1.96243711e-02 4.29782629e-01
-4.89756972e-01 3.11684400e-01 4.05007660e-01 -5.13075650e-01
-1.42624581e+00 -1.34190214e+00 2.36885641e-02 5.46344044e-03
5.46684623e-01 -6.19392358e-02 -9.91671562e-01 4.33617473e-01
-3.82995531e-02 4.85250980e-01 9.18715715e-01 -3.06205362e-01
9.34981555e-03 -4.83381867e-01 -1.39180636e+00 1.82205036e-01
9.05486643e-01 -1.89305589e-01 -4.76508647e-01 4.12933201e-01
3.68134409e-01 -6.98642969e-01 -1.01448441e+00 7.93923855e-01
5.22166789e-01 -7.30541468e-01 8.45346272e-01 -8.81545603e-01
7.34985590e-01 2.10889563e-01 -3.81415814e-01 -1.05719805e+00
-4.11035389e-01 1.62048027e-01 1.77861333e-01 1.02014875e+00
4.14927006e-01 -4.57459569e-01 1.04623210e+00 3.56845140e-01
-2.62433112e-01 -1.12564862e+00 -7.78338075e-01 -5.51303983e-01
3.60051244e-01 1.05601221e-01 3.23610097e-01 9.60383654e-01
-5.29901683e-01 2.46798202e-01 -1.78157970e-01 3.05827916e-01
6.03373706e-01 2.38883141e-02 5.61704874e-01 -9.48639035e-01
-4.45694417e-01 1.90491080e-02 -7.73984134e-01 -1.03349721e+00
-1.47977024e-01 -8.97837996e-01 -2.36916512e-01 -1.19164884e+00
4.96544898e-01 -8.62773180e-01 -6.22903585e-01 -3.91490990e-03
-1.76994041e-01 1.04339965e-01 -5.24081171e-01 1.78157568e-01
-3.26588094e-01 3.78847152e-01 1.66365063e+00 -6.03668153e-01
-9.63829309e-02 2.54772514e-01 -9.75729644e-01 7.10701942e-01
8.13771427e-01 -3.85609120e-01 -7.61251330e-01 -6.01501942e-01
-2.49761760e-01 2.28386089e-01 2.07072869e-01 -9.92118120e-01
2.93849576e-02 -2.82692593e-02 1.13462353e+00 -6.18999898e-01
8.22775811e-02 -8.61550927e-01 -5.43868914e-02 5.35927296e-01
-3.71688187e-01 1.22748036e-02 3.71642679e-01 8.90626550e-01
-2.70738751e-01 -1.94426686e-01 1.05924761e+00 -2.93586820e-01
-9.08766031e-01 6.07065797e-01 2.11806089e-01 7.97176268e-03
1.03582191e+00 -3.67895514e-01 1.66899770e-01 7.44507983e-02
-8.71538341e-01 8.26594308e-02 1.39672205e-01 4.81148988e-01
7.53146350e-01 -1.54190445e+00 -8.74228239e-01 3.83929342e-01
5.19540131e-01 2.59814501e-01 6.16171777e-01 6.16327465e-01
-5.18483043e-01 -2.26709526e-02 -5.14113367e-01 -1.04020393e+00
-1.05999660e+00 6.32865429e-01 4.83628839e-01 -3.95031512e-01
-5.39817572e-01 9.12572205e-01 5.82047462e-01 -1.63141429e-01
3.10238004e-01 -3.39608192e-01 -1.95769846e-01 -2.51955152e-01
4.37566072e-01 -7.14420453e-02 3.32679421e-01 -4.52987790e-01
-3.40490580e-01 3.21238786e-01 -6.28615618e-01 4.42917377e-01
1.40078700e+00 -8.07375982e-02 3.85820866e-01 2.13438630e-01
1.49492824e+00 -6.11690044e-01 -9.58953083e-01 -4.92323995e-01
-1.69664338e-01 -3.21812600e-01 2.65972558e-02 -6.44414961e-01
-1.15600455e+00 9.22350466e-01 1.52949011e+00 -4.34984304e-02
1.49801254e+00 4.61733229e-02 6.79923713e-01 5.47190785e-01
2.99492836e-01 -1.10555732e+00 4.99528050e-01 -1.28562167e-01
8.77293229e-01 -1.96038079e+00 5.15240319e-02 -4.38342765e-02
-9.07355428e-01 1.08879328e+00 1.04389751e+00 -1.48601055e-01
9.24317062e-01 2.13258043e-01 1.90351456e-01 -1.00099690e-01
-2.75977492e-01 5.60536802e-01 2.84485012e-01 6.22487724e-01
4.98450786e-01 1.43803526e-02 -3.05374205e-01 8.28707457e-01
1.02825060e-01 8.54058415e-02 2.58559495e-01 1.00632572e+00
-3.72517109e-01 -8.03049505e-01 -2.63185054e-01 9.07606602e-01
-5.17105043e-01 2.68590689e-01 1.59723327e-01 9.44626093e-01
2.83329546e-01 5.07830977e-01 -1.14904799e-01 -3.17509860e-01
4.64601189e-01 -2.88476199e-01 4.88733649e-01 -7.08042085e-01
-1.83692962e-01 -3.11990902e-02 -2.17632949e-01 -3.93688977e-02
-5.68255961e-01 -4.64117289e-01 -7.26481259e-01 1.41082421e-01
-3.69093865e-01 -2.34672859e-01 2.75990576e-01 7.43755758e-01
-6.08883873e-02 6.32991970e-01 9.70743954e-01 -5.54295897e-01
-1.12830997e+00 -1.17807150e+00 -6.25480533e-01 8.08704972e-01
1.70606881e-01 -8.11749041e-01 2.74290722e-02 2.33936459e-01] | [14.813976287841797, -2.0026071071624756] |
fcd53fa7-83e4-4529-af82-e239c01aafee | advances-in-hyperspectral-image | 1310.5107 | null | http://arxiv.org/abs/1310.5107v1 | http://arxiv.org/pdf/1310.5107v1.pdf | Advances in Hyperspectral Image Classification: Earth monitoring with statistical learning methods | Hyperspectral images show similar statistical properties to natural grayscale
or color photographic images. However, the classification of hyperspectral
images is more challenging because of the very high dimensionality of the
pixels and the small number of labeled examples typically available for
learning. These peculiarities lead to particular signal processing problems,
mainly characterized by indetermination and complex manifolds. The framework of
statistical learning has gained popularity in the last decade. New methods have
been presented to account for the spatial homogeneity of images, to include
user's interaction via active learning, to take advantage of the manifold
structure with semisupervised learning, to extract and encode invariances, or
to adapt classifiers and image representations to unseen yet similar scenes.
This tutuorial reviews the main advances for hyperspectral remote sensing image
classification through illustrative examples. | ['Jón Atli Benediktsson', 'Gustavo Camps-Valls', 'Devis Tuia', 'Lorenzo Bruzzone'] | 2013-10-18 | null | null | null | null | ['classification-of-hyperspectral-images', 'remote-sensing-image-classification'] | ['computer-vision', 'miscellaneous'] | [ 6.96173310e-01 -9.66399908e-02 -1.41945094e-01 -4.71897542e-01
-4.05720919e-01 -6.40059769e-01 5.61039090e-01 2.44340394e-02
-2.13935032e-01 7.94873059e-01 -3.28968853e-01 1.38757601e-02
-7.46924996e-01 -6.95236742e-01 -3.12676467e-02 -1.20579410e+00
-3.66810769e-01 2.21775874e-01 -2.41422012e-01 -3.67141142e-03
2.88905233e-01 9.11696792e-01 -1.91715610e+00 -3.85162979e-02
1.22706163e+00 8.76939476e-01 3.27791214e-01 4.90402550e-01
-2.44880825e-01 4.26145732e-01 -2.74798512e-01 2.88858533e-01
4.12523419e-01 -4.68423516e-01 -6.14615619e-01 8.73504043e-01
4.84132290e-01 1.01774372e-01 -1.09509043e-01 1.46091425e+00
3.28045458e-01 4.09605891e-01 9.65972900e-01 -1.24796236e+00
-6.49372339e-01 3.60067308e-01 -4.60482419e-01 1.80308476e-01
-1.42958075e-01 -1.03321001e-01 7.57325768e-01 -9.52577055e-01
3.78380209e-01 8.68709624e-01 8.09895769e-02 2.83784598e-01
-1.38315046e+00 -1.63250834e-01 1.95390061e-02 6.24109030e-01
-1.50042653e+00 -5.81854522e-01 8.87705982e-01 -6.26276195e-01
4.47580576e-01 5.34488916e-01 4.69644904e-01 4.57264453e-01
-4.65319902e-01 6.09591782e-01 1.36314452e+00 -6.53071880e-01
5.61702728e-01 3.85475725e-01 2.98012346e-01 4.62795049e-01
3.59116554e-01 6.56474084e-02 -2.21306801e-01 -1.34403333e-01
5.53808272e-01 1.68213829e-01 -5.11244595e-01 -8.09417248e-01
-9.34349179e-01 7.82972634e-01 4.36744750e-01 5.96375883e-01
-4.59710091e-01 -7.62102902e-01 -9.73048881e-02 3.20411384e-01
5.99514782e-01 8.40914190e-01 -3.86300087e-01 4.80439574e-01
-8.68072331e-01 -5.00543296e-01 5.53927362e-01 5.42378783e-01
1.31196618e+00 2.47402266e-01 3.56903762e-01 9.02347565e-01
1.42969668e-01 8.09986711e-01 4.35637176e-01 -6.66159451e-01
1.17117733e-01 7.71679401e-01 -6.04300946e-02 -1.01414609e+00
-4.91083145e-01 -3.14592242e-01 -1.09969938e+00 6.92775488e-01
2.15553552e-01 -4.38750051e-02 -8.59597266e-01 1.10729516e+00
1.60044983e-01 -1.23594463e-01 1.90138146e-01 9.31682467e-01
4.67997432e-01 7.70731390e-01 -1.62398664e-03 -6.58070505e-01
5.55034816e-01 -5.20911932e-01 -6.64234102e-01 -2.06400290e-01
5.22679448e-01 -4.89749730e-01 9.39719558e-01 5.45349240e-01
-5.56331336e-01 -4.57185954e-01 -1.09465885e+00 4.75515485e-01
-7.80152082e-01 3.10304403e-01 5.24850786e-01 7.41236627e-01
-7.39844680e-01 7.11712658e-01 -8.42942715e-01 -3.25296789e-01
5.79425037e-01 4.93670017e-01 -4.45070773e-01 8.55846237e-03
-7.81472206e-01 7.48346865e-01 6.98006511e-01 2.95827389e-01
-3.25789303e-01 -4.19042617e-01 -9.19092119e-01 3.91535051e-02
2.41833761e-01 1.14719654e-02 4.86707300e-01 -1.62129188e+00
-1.56233692e+00 6.68183565e-01 1.22204430e-01 6.20792061e-02
1.52450845e-01 1.31144688e-01 -5.81255734e-01 4.48629022e-01
-3.00021440e-01 1.71531141e-01 1.22707367e+00 -1.08214617e+00
-2.76040345e-01 -7.36013234e-01 -2.44735956e-01 2.94368625e-01
-6.78213179e-01 -1.09327316e-01 2.78307289e-01 -3.26393276e-01
7.22202063e-01 -9.67319906e-01 -3.81504416e-01 -6.44479692e-02
-4.52711672e-01 1.01321347e-01 1.29856062e+00 -3.60438079e-01
5.96696496e-01 -2.47592330e+00 3.40768725e-01 5.48403382e-01
4.92908098e-02 4.36031520e-01 -2.16557369e-01 3.07496458e-01
-6.15782917e-01 -6.48825914e-02 -5.61559260e-01 1.03928365e-01
-2.27372363e-01 4.89855669e-02 -2.69312769e-01 8.26474726e-01
3.72113854e-01 5.65228760e-01 -8.83469284e-01 -4.12456155e-01
5.84504962e-01 2.79787928e-01 1.14191130e-01 1.73168883e-01
3.31809334e-02 7.61210978e-01 -4.06747699e-01 6.05600119e-01
8.65843952e-01 -5.06725371e-01 9.64594334e-02 -2.68768251e-01
-3.46010119e-01 -3.26116681e-01 -1.27961278e+00 1.33070087e+00
-8.71910453e-02 6.65361822e-01 3.12704116e-01 -1.60572028e+00
9.16553795e-01 3.93642157e-01 8.48956347e-01 -2.60045946e-01
-2.68582795e-02 2.03253672e-01 -1.90814450e-01 -6.98153973e-01
3.21336359e-01 -4.62332554e-02 5.49571693e-01 3.06179851e-01
-5.37380725e-02 -4.03426349e-01 2.13527873e-01 -1.75820291e-01
3.75083834e-01 -3.32773179e-02 5.68528056e-01 -4.34707344e-01
6.82344556e-01 6.72095791e-02 2.22654447e-01 3.55983943e-01
-5.88378124e-02 3.98292720e-01 -9.82575193e-02 -2.00709105e-01
-7.31210947e-01 -8.46099973e-01 -4.76129264e-01 7.52203763e-01
6.93661943e-02 2.66380101e-01 -3.48488122e-01 -5.37794113e-01
-1.95607200e-01 3.77044022e-01 -4.28183585e-01 -5.01642004e-03
-1.13726683e-01 -1.20458198e+00 -1.43844396e-01 2.00658329e-02
6.27683997e-01 -7.84882784e-01 -2.95694321e-01 -5.22888228e-02
1.54050246e-01 -9.39443886e-01 3.73414904e-01 3.08104157e-01
-1.27766871e+00 -1.20561361e+00 -6.64214969e-01 -5.69150686e-01
9.94654417e-01 6.12037539e-01 6.21921360e-01 -3.69699091e-01
-6.06561422e-01 6.48515821e-01 -5.25697410e-01 -5.16184747e-01
-3.65478367e-01 5.36584146e-02 1.70877367e-01 4.39451545e-01
1.82359442e-01 -5.64507663e-01 -3.37328225e-01 4.98382330e-01
-1.25805295e+00 -2.58271217e-01 8.03484201e-01 8.63394678e-01
6.04050517e-01 7.07152784e-01 4.09518033e-01 -9.21633542e-01
3.05667520e-02 -3.36213797e-01 -9.55707014e-01 4.30753648e-01
-6.10624313e-01 -8.26883093e-02 7.38897443e-01 -3.44546229e-01
-1.18823075e+00 5.70449471e-01 5.24617553e-01 -1.11368299e-01
-5.68881273e-01 6.07482076e-01 -3.48778427e-01 -5.09488821e-01
1.17276216e+00 2.60273695e-01 3.29410404e-01 -1.95731044e-01
3.32497388e-01 9.57438946e-01 1.91440955e-01 -8.12726840e-02
1.19888818e+00 6.01195216e-01 2.28831887e-01 -1.70651186e+00
-7.96666741e-01 -6.65984690e-01 -1.22876370e+00 -2.82638490e-01
6.58880115e-01 -7.08468854e-01 2.93571372e-02 4.68952864e-01
-5.58641911e-01 -2.85814613e-01 -4.63748962e-01 8.53738844e-01
-5.26422620e-01 7.28048623e-01 -1.08226821e-01 -1.06283987e+00
-7.78038800e-02 -8.66477132e-01 5.96266925e-01 2.81677365e-01
1.69199243e-01 -1.21900511e+00 1.65467076e-02 2.49307543e-01
4.58313763e-01 1.98525861e-01 1.09624839e+00 -5.26406825e-01
-6.45221889e-01 -2.34756038e-01 -1.90346539e-01 6.30308807e-01
8.00723195e-01 1.18910089e-01 -1.38646984e+00 -4.26288038e-01
2.43817329e-01 -2.57201195e-01 7.11499751e-01 4.40971136e-01
1.21052873e+00 -7.89725929e-02 -1.33538753e-01 5.50162315e-01
1.56675601e+00 2.56540447e-01 6.90923572e-01 1.13304429e-01
4.53609884e-01 8.39486182e-01 5.62261045e-01 2.35737249e-01
-5.68360567e-01 2.66259670e-01 4.73428875e-01 -4.07096386e-01
3.44126523e-01 4.17995811e-01 3.23097073e-02 8.03798020e-01
-3.41363072e-01 1.92996144e-01 -7.50091851e-01 9.74126011e-02
-1.56915450e+00 -1.11687040e+00 -2.04415470e-01 2.50995612e+00
3.80896717e-01 -3.84177238e-01 -1.95098639e-01 6.67968929e-01
9.89236116e-01 2.42853507e-01 -6.23207331e-01 3.13563198e-01
-6.81080163e-01 2.98260674e-02 5.44000983e-01 2.57922947e-01
-1.49278641e+00 4.34809834e-01 6.73987055e+00 5.61658561e-01
-1.59374845e+00 -2.26064458e-01 4.96868938e-01 4.32521880e-01
8.47100243e-02 1.57988053e-02 -1.91412121e-01 4.71870080e-02
4.08166617e-01 -3.02208588e-02 5.39400220e-01 6.87649548e-01
2.41520986e-01 -9.95756015e-02 -8.86311471e-01 1.27330172e+00
1.20196693e-01 -8.96739721e-01 1.90821826e-01 1.04141690e-01
9.94178355e-01 -4.28095087e-02 2.60944009e-01 -1.54255152e-01
-1.86203092e-01 -8.64523947e-01 1.03150278e-01 6.68962896e-01
6.55207872e-01 -4.16873753e-01 4.44323361e-01 5.09583771e-01
-7.99956679e-01 -4.04909819e-01 -8.01471233e-01 -8.28612000e-02
-3.18953514e-01 6.82017088e-01 -7.15543509e-01 6.75406814e-01
3.86068761e-01 9.89313066e-01 -7.20194817e-01 1.26083243e+00
-4.50941660e-02 4.94168788e-01 -3.06696475e-01 -4.59382646e-02
6.88543022e-02 -9.55805182e-01 7.21006691e-01 7.60433674e-01
4.58974481e-01 3.44835162e-01 9.17945430e-02 8.36153507e-01
5.18062413e-01 6.28859580e-01 -7.83511877e-01 -5.42529583e-01
1.86058179e-01 1.53313243e+00 -8.96063745e-01 -2.71515608e-01
-3.42423201e-01 9.47576582e-01 -7.53484070e-02 6.46132827e-01
-1.97408393e-01 -3.35363328e-01 2.68866777e-01 -5.48953116e-02
3.75397876e-02 -5.66873074e-01 -5.23675717e-02 -1.32098210e+00
-3.71829182e-01 -6.89408183e-01 4.48701829e-01 -8.08471799e-01
-1.34627306e+00 6.15193367e-01 1.78587660e-01 -1.61414897e+00
-1.03480689e-01 -9.39586699e-01 -4.43550527e-01 6.91618383e-01
-1.59579825e+00 -7.08763182e-01 -5.33527315e-01 8.59111547e-01
2.55520791e-01 -5.31983674e-01 1.20246625e+00 6.47945255e-02
-6.39495373e-01 -6.33912981e-02 7.03357518e-01 5.36517724e-02
5.42653203e-01 -1.38525581e+00 -4.80210632e-01 7.93466270e-01
2.17933506e-01 3.23208153e-01 5.50116599e-01 -1.89001262e-01
-1.43678176e+00 -1.07894278e+00 2.80155271e-01 2.38489117e-02
5.78345418e-01 -1.12642338e-02 -8.95364225e-01 2.65420675e-01
5.10521792e-02 7.36277848e-02 1.10593164e+00 -8.21402222e-02
-2.28588223e-01 -3.44108671e-01 -1.05178809e+00 4.61795360e-01
5.68184495e-01 -5.51186919e-01 -3.97765487e-01 6.24676764e-01
-9.02330726e-02 3.40559185e-01 -6.75911784e-01 4.04979914e-01
-3.48280407e-02 -8.52195978e-01 7.64093816e-01 -5.78565955e-01
-1.47392392e-01 -2.47527048e-01 -1.80738002e-01 -1.54623044e+00
-6.11987352e-01 -4.67113674e-01 4.75895047e-01 8.11776876e-01
6.33877456e-01 -8.89275193e-01 8.17935944e-01 4.87083495e-01
-6.87058493e-02 1.37567613e-02 -6.35529518e-01 -7.88045466e-01
-2.44578809e-01 1.55506566e-01 2.47915551e-01 1.43563807e+00
1.85677901e-01 3.29046756e-01 -1.66185331e-02 5.02611220e-01
1.03807247e+00 5.08222699e-01 4.71413821e-01 -1.82139075e+00
-1.36622369e-01 -3.70418131e-01 -8.19180846e-01 -6.04844987e-01
2.09661469e-01 -8.68148923e-01 -3.32339071e-02 -1.23201084e+00
6.32747710e-02 -4.07628566e-01 -2.53714561e-01 3.99859071e-01
-8.84180516e-02 4.15516794e-01 -1.83299910e-02 5.18540382e-01
-3.22362512e-01 4.73813653e-01 1.14040411e+00 -4.70518470e-01
-5.09379089e-01 2.03506514e-01 -3.42219800e-01 7.05251813e-01
9.54453588e-01 -2.53335126e-02 -6.38987243e-01 -1.73397064e-01
2.67178267e-01 -1.97822973e-01 2.11906955e-01 -1.02893162e+00
8.32911655e-02 -4.19992894e-01 5.12238026e-01 -3.40903670e-01
4.01490182e-01 -1.22105515e+00 2.45617107e-01 1.83207557e-01
-2.09601209e-01 -5.31368613e-01 -7.11869225e-02 5.39147019e-01
-5.04885316e-01 -5.05401790e-01 1.18623686e+00 -2.59232014e-01
-9.76858377e-01 4.75352973e-01 -6.01271510e-01 -4.61378366e-01
1.00135064e+00 -4.10296738e-01 -9.31177810e-02 -2.50266463e-01
-1.06420517e+00 -3.90657306e-01 3.48144710e-01 3.77782062e-02
6.25403762e-01 -1.04032195e+00 -7.34124362e-01 4.73138571e-01
4.19531375e-01 -2.27487043e-01 3.09502631e-01 8.37188125e-01
-5.20615220e-01 2.28698656e-01 -5.11980593e-01 -8.74516726e-01
-1.38481390e+00 6.31047308e-01 4.95545149e-01 4.32216823e-01
-3.62454057e-01 3.40814084e-01 8.69654492e-02 -3.78820062e-01
-1.56066403e-01 -1.72095932e-03 -6.88862085e-01 3.46097350e-01
4.97773230e-01 6.34481668e-01 2.07193524e-01 -7.45108247e-01
5.38076414e-03 6.22004330e-01 3.77481848e-01 1.19265310e-01
1.44774067e+00 -2.46588275e-01 -5.46597183e-01 5.79925716e-01
1.22907782e+00 -5.07345982e-02 -1.08261395e+00 -4.74649161e-01
1.83300421e-01 -6.21080816e-01 3.03950846e-01 -4.25923854e-01
-9.81258273e-01 7.91943610e-01 9.15953994e-01 7.18072176e-01
1.36528468e+00 -2.91424543e-01 -4.76050943e-01 9.43960130e-01
4.06394422e-01 -1.15677023e+00 -1.25239342e-01 3.16696167e-01
7.45341957e-01 -1.42525995e+00 2.15435982e-01 -8.28688800e-01
-4.02916640e-01 1.57694304e+00 4.12193388e-02 8.02792683e-02
9.51579571e-01 -1.93051621e-01 4.38565046e-01 -1.46701381e-01
-2.00787876e-02 -3.47970963e-01 4.00768936e-01 8.18966866e-01
3.05589020e-01 3.70017104e-02 -8.15604776e-02 -2.53251791e-01
7.85954520e-02 -3.86757612e-01 5.02093256e-01 7.82350719e-01
-7.40040720e-01 -8.98662627e-01 -6.57108486e-01 4.92726535e-01
-2.32353695e-02 1.49407253e-01 -4.78926063e-01 5.64449191e-01
-5.41560650e-02 1.19369686e+00 -6.76816031e-02 -1.52636589e-02
-2.65194457e-02 2.06892505e-01 5.65038979e-01 -7.83544123e-01
2.70733863e-01 8.70976821e-02 -2.59148806e-01 -1.24199793e-01
-9.89590228e-01 -7.12565720e-01 -9.41551149e-01 3.03270340e-01
-7.18615830e-01 5.35092413e-01 8.42922091e-01 8.20114255e-01
6.12895638e-02 -1.09944157e-01 1.18053734e+00 -1.02347839e+00
-7.52969623e-01 -1.11960781e+00 -1.42439592e+00 4.78211313e-01
2.85897940e-01 -4.96503383e-01 -8.60675275e-01 2.76564360e-01] | [9.993253707885742, -1.9149678945541382] |
ac81425e-bc7d-4d31-9481-c3c38b87a7a2 | multi-microphone-speaker-separation-by | 2303.07143 | null | https://arxiv.org/abs/2303.07143v1 | https://arxiv.org/pdf/2303.07143v1.pdf | Multi-Microphone Speaker Separation by Spatial Regions | We consider the task of region-based source separation of reverberant multi-microphone recordings. We assume pre-defined spatial regions with a single active source per region. The objective is to estimate the signals from the individual spatial regions as captured by a reference microphone while retaining a correspondence between signals and spatial regions. We propose a data-driven approach using a modified version of a state-of-the-art network, where different layers model spatial and spectro-temporal information. The network is trained to enforce a fixed mapping of regions to network outputs. Using speech from LibriMix, we construct a data set specifically designed to contain the region information. Additionally, we train the network with permutation invariant training. We show that both training methods result in a fixed mapping of regions to network outputs, achieve comparable performance, and that the networks exploit spatial information. The proposed network outperforms a baseline network by 1.5 dB in scale-invariant signal-to-distortion ratio. | ['Emanuël A. P. Habets', 'Wolfgang Mack', 'Srikanth Raj Chetupalli', 'Julian Wechsler'] | 2023-03-13 | null | null | null | null | ['speaker-separation'] | ['speech'] | [ 5.01856387e-01 -2.41272956e-01 2.67113388e-01 -3.79076183e-01
-1.43541539e+00 -1.00343752e+00 4.14370596e-01 -1.26233637e-01
-4.23844665e-01 4.22326118e-01 5.33590853e-01 -9.28461626e-02
-3.18496585e-01 -2.43055880e-01 -9.74368572e-01 -7.79744148e-01
-3.16978693e-01 -2.95017153e-01 1.69890746e-01 1.78414851e-01
1.98109984e-01 7.20096052e-01 -1.27045965e+00 6.02608442e-01
4.57077175e-01 1.20515871e+00 2.87676960e-01 1.22498727e+00
4.48116094e-01 6.79149210e-01 -9.55338478e-01 5.99849820e-01
4.70599651e-01 -5.61366320e-01 -3.84976357e-01 -1.03750177e-01
6.79587483e-01 2.29939252e-01 -3.84426445e-01 8.54667962e-01
9.07126069e-01 7.31584728e-01 5.59439182e-01 -7.62684345e-01
-3.19771439e-01 6.53175235e-01 -3.17885607e-01 4.84585464e-01
1.85564339e-01 -4.64449674e-01 8.56920421e-01 -1.19345033e+00
3.71141061e-02 8.48208070e-01 7.88578749e-01 3.48080754e-01
-1.17455196e+00 -6.12807393e-01 2.45162949e-01 -1.25894114e-01
-1.76029444e+00 -1.13328111e+00 1.02212584e+00 -3.11351955e-01
9.19713557e-01 6.73696816e-01 8.45963601e-03 9.37483251e-01
-5.22157773e-02 2.37430334e-01 9.90473211e-01 -5.52770913e-01
4.78053242e-01 6.15691878e-02 -4.10862118e-02 1.33793607e-01
-5.40890992e-01 1.54789016e-01 -9.72110391e-01 -1.13066785e-01
7.71931469e-01 -2.36475259e-01 -7.75337398e-01 -3.28697711e-01
-1.21786153e+00 4.53544825e-01 5.88249564e-01 6.88967824e-01
-2.61661887e-01 9.49378386e-02 6.14127629e-02 2.24813908e-01
8.47205698e-01 5.35222888e-01 -5.44200778e-01 1.12755597e-01
-1.38047206e+00 7.61877075e-02 8.25864971e-01 8.54843318e-01
5.18725693e-01 3.59115124e-01 -8.93679857e-02 1.18018115e+00
2.52509624e-01 4.99536455e-01 4.40313458e-01 -1.16725755e+00
6.25989974e-01 -1.98191077e-01 4.25060280e-02 -9.04649317e-01
-4.46097344e-01 -1.08682585e+00 -9.63249862e-01 3.47451121e-02
4.00649071e-01 -3.91352832e-01 -7.93666184e-01 1.81515515e+00
1.12965964e-01 6.91370547e-01 5.08889109e-02 9.65113580e-01
4.36093092e-01 8.27370524e-01 -8.45770895e-01 -2.76694715e-01
7.77667403e-01 -1.13279605e+00 -5.61240971e-01 -4.33473021e-01
-4.49751616e-02 -6.92172348e-01 7.43118227e-01 5.29132962e-01
-1.32816005e+00 -8.54419708e-01 -1.09870064e+00 2.45657057e-01
-4.80861753e-01 2.38496363e-01 -3.66694599e-01 7.67787516e-01
-1.40699184e+00 4.77514386e-01 -4.66836363e-01 7.52898008e-02
1.66290682e-02 3.05830449e-01 -3.20566565e-01 2.31471494e-01
-9.65070426e-01 3.11450601e-01 1.18230274e-02 3.25104475e-01
-1.16200042e+00 -1.00144494e+00 -8.92858267e-01 1.62325390e-02
-5.65060824e-02 -1.86180308e-01 1.22402465e+00 -1.15024090e+00
-1.57522428e+00 4.74452674e-01 -5.67775786e-01 -4.36692446e-01
1.15712956e-01 -2.45217621e-01 -7.94444859e-01 3.59256148e-01
1.79493561e-01 2.52610356e-01 1.16079414e+00 -1.50917780e+00
-6.74516559e-01 -1.45105228e-01 -1.87572464e-01 2.19398335e-01
-3.47820014e-01 4.45027739e-01 -4.53651011e-01 -1.07381392e+00
2.47884139e-01 -6.87641144e-01 -1.36899203e-01 -1.60905451e-01
-5.50952375e-01 3.46429527e-01 4.94562775e-01 -9.16204274e-01
1.18499899e+00 -2.47796369e+00 1.31731823e-01 4.89148378e-01
1.52116135e-01 -8.49207863e-02 -4.01121557e-01 6.97587952e-02
-4.06607807e-01 -2.01881286e-02 -6.01663113e-01 -8.64148021e-01
-1.64703846e-01 -2.17786267e-01 -5.51874399e-01 1.01375449e+00
-5.34331203e-02 2.73754209e-01 -6.64467037e-01 -5.34619801e-02
1.77353799e-01 6.60788298e-01 -6.28728569e-01 5.17099798e-01
2.91507006e-01 8.02607298e-01 1.95993006e-01 3.14515054e-01
8.34014654e-01 2.83097267e-01 -1.24079615e-01 9.54217277e-03
-3.93934429e-01 5.26098788e-01 -1.43500066e+00 1.86371493e+00
-8.46815526e-01 9.47715104e-01 7.49921918e-01 -1.10092318e+00
9.20031428e-01 6.10116124e-01 4.42058831e-01 -4.62582290e-01
-2.09102735e-01 3.59795451e-01 -1.14250503e-01 -1.17887624e-01
3.95899445e-01 -6.24596141e-02 -6.14876784e-02 4.85520810e-01
1.45452380e-01 6.87421709e-02 -4.27869678e-01 -3.22514921e-01
1.15618956e+00 -2.12229684e-01 1.06306732e-01 -3.62705261e-01
4.31447893e-01 -6.80432677e-01 5.43417573e-01 1.02708328e+00
-9.37786475e-02 1.35114479e+00 1.55386655e-02 3.82468142e-02
-7.61089385e-01 -1.39643717e+00 -4.53167371e-02 1.25065124e+00
-1.22783609e-01 1.41022608e-01 -8.02845120e-01 -3.97159994e-01
-3.62716705e-01 6.68604791e-01 -5.64923465e-01 1.35788336e-01
-1.13698435e+00 -2.39804819e-01 8.64482224e-01 5.63819468e-01
5.16954422e-01 -8.09682965e-01 -3.36094469e-01 1.54778838e-01
-2.26530612e-01 -1.07935154e+00 -1.00671673e+00 5.91073036e-01
-4.66279238e-01 -5.29757082e-01 -1.01542032e+00 -9.67258811e-01
5.88908732e-01 5.14545679e-01 1.04357481e+00 -5.94885588e-01
2.66440988e-01 2.65681237e-01 -1.94874376e-01 -3.35048825e-01
-2.08848611e-01 1.22044928e-01 2.78883070e-01 5.01457036e-01
-2.45251894e-01 -1.00037432e+00 -6.40296519e-01 4.95447785e-01
-7.38137901e-01 -4.16422695e-01 1.65933460e-01 4.79757935e-01
5.52118599e-01 3.34963292e-01 7.77936578e-01 -2.54371047e-01
4.32566702e-01 -4.63587373e-01 -3.65673274e-01 -9.96650308e-02
4.38259616e-02 -2.79957920e-01 1.02603197e+00 -5.15309095e-01
-1.03638613e+00 1.79371804e-01 -3.05984914e-02 -5.21551371e-01
-5.13911307e-01 8.09033737e-02 -3.52112830e-01 -1.51820123e-01
9.19407129e-01 2.28881523e-01 -5.29009283e-01 -7.87969828e-01
4.55718338e-01 9.73404944e-01 8.22223544e-01 -2.91937113e-01
7.55772829e-01 5.58645070e-01 -4.51951832e-01 -8.52443576e-01
-7.26640105e-01 -1.02716374e+00 -7.26838887e-01 -1.52119681e-01
6.17228508e-01 -1.00589049e+00 -1.58391416e-01 3.25556993e-01
-1.44553375e+00 -4.70503271e-01 -4.27620441e-01 6.31946087e-01
-5.62300026e-01 -1.07007705e-01 -2.63404191e-01 -1.07714570e+00
5.39275818e-02 -6.95593715e-01 1.16052032e+00 -1.34565353e-01
-1.45635992e-01 -1.20381439e+00 3.59085172e-01 1.75504163e-02
5.20372748e-01 3.54062840e-02 3.07185709e-01 -8.00454140e-01
-3.63369286e-01 1.10332072e-02 3.69760506e-02 6.69342220e-01
4.87963229e-01 -5.98156214e-01 -1.77693653e+00 -3.61717224e-01
5.92588842e-01 2.61794090e-01 9.32983577e-01 8.30307961e-01
1.16491342e+00 -6.26432180e-01 -1.14419535e-01 9.30660903e-01
1.24011600e+00 4.32517290e-01 2.49187946e-01 -6.28327429e-02
6.97059333e-01 7.53611565e-01 2.30377764e-02 1.98958218e-01
-1.11145385e-01 7.15644121e-01 2.53356665e-01 -4.34073567e-01
-3.23669910e-01 -9.38778967e-02 4.89283621e-01 9.28574562e-01
3.87220949e-01 -5.11706352e-01 -7.23429978e-01 1.16551137e+00
-1.61217630e+00 -9.43570673e-01 3.63053828e-01 2.32437277e+00
8.48468959e-01 -1.17304347e-01 2.10990950e-01 3.86073530e-01
8.65403891e-01 6.04347467e-01 -3.34227383e-01 -3.05622160e-01
-3.34628880e-01 3.11161250e-01 6.53024912e-01 1.05820334e+00
-1.20425177e+00 3.45420003e-01 6.90669584e+00 7.63008475e-01
-1.26735008e+00 2.99047977e-01 4.84421790e-01 -6.06163561e-01
-1.97103769e-01 -4.80490357e-01 -4.21612084e-01 2.19676092e-01
1.38201308e+00 2.13037312e-01 8.49094629e-01 3.84935260e-01
6.92213714e-01 2.52813488e-01 -1.27039790e+00 8.43754828e-01
4.69637871e-01 -1.13511014e+00 -5.67496121e-01 -1.35490865e-01
7.98303068e-01 1.28595307e-01 2.84607172e-01 -2.22138390e-01
-3.51928212e-02 -1.26082516e+00 1.28797424e+00 5.72904706e-01
7.68016100e-01 -6.97804213e-01 2.67576426e-01 4.77009386e-01
-1.44812465e+00 -2.31996939e-01 -2.10106447e-01 -1.13710411e-01
1.73730806e-01 7.31931269e-01 -9.59136903e-01 4.62115169e-01
8.54682982e-01 5.79342663e-01 -2.12200373e-01 1.09776628e+00
3.30725238e-02 8.62332404e-01 -3.72040033e-01 3.64934802e-01
-3.55900563e-02 5.80882542e-02 8.55077982e-01 1.61698437e+00
5.28607666e-01 -3.34388584e-01 9.60253458e-03 8.78992975e-01
-1.78223401e-01 -1.18260019e-01 -8.52198839e-01 5.70745230e-01
6.81637645e-01 9.17262316e-01 -4.70435739e-01 -6.90812171e-02
-2.46763766e-01 1.04866517e+00 5.48228621e-02 9.00641561e-01
-7.09960938e-01 -6.98437274e-01 7.46630847e-01 -3.59906536e-03
5.66322505e-01 -3.04974735e-01 -5.21933198e-01 -6.53194427e-01
2.86154658e-01 -8.99805129e-01 -7.37847239e-02 -7.64952898e-01
-1.02414846e+00 7.64316678e-01 3.23664211e-02 -1.32450294e+00
-3.02468985e-01 -4.46974993e-01 -7.01180816e-01 1.33421469e+00
-1.55488193e+00 -8.37317348e-01 6.70688087e-03 6.65625572e-01
6.34483397e-01 -1.88703373e-01 7.84723222e-01 4.18920219e-01
-3.87041450e-01 6.88735425e-01 4.74570602e-01 2.87914276e-01
8.11519206e-01 -1.43124056e+00 5.41947722e-01 1.28024709e+00
5.38542449e-01 7.65547693e-01 6.38021052e-01 -5.15303500e-02
-8.41138721e-01 -1.39723170e+00 9.99405444e-01 -5.65603018e-01
4.24724936e-01 -9.73496735e-01 -7.47449219e-01 6.72139645e-01
3.67035389e-01 3.77391160e-01 6.18214726e-01 -8.64588246e-02
-4.95159566e-01 -5.14092982e-01 -9.37913179e-01 2.96797872e-01
9.79685366e-01 -1.21793163e+00 -6.36561990e-01 3.22538972e-01
9.28343296e-01 -4.92028147e-01 -5.19847512e-01 1.01135306e-01
4.10608470e-01 -1.02167702e+00 1.19175851e+00 -2.04264522e-01
-4.37460728e-02 -5.76451957e-01 -5.81842721e-01 -1.70845962e+00
-2.87114888e-01 -9.91493762e-01 8.04966874e-03 1.28880680e+00
8.01723838e-01 -7.07186043e-01 4.39765394e-01 3.96818249e-03
-5.75204730e-01 -4.45110649e-01 -1.43760824e+00 -1.01337194e+00
1.72353566e-01 -7.35820293e-01 4.32276338e-01 6.84863567e-01
-1.66985810e-01 2.01240420e-01 -3.70598823e-01 8.23157966e-01
4.40926969e-01 -4.59989384e-02 3.17521036e-01 -7.01433718e-01
-5.71737409e-01 -4.13174361e-01 2.01040030e-01 -1.46948445e+00
2.83629775e-01 -6.33230031e-01 6.92143857e-01 -1.30199862e+00
-3.97224545e-01 -4.05874312e-01 -7.18623698e-01 1.89478442e-01
1.11787148e-01 4.27972883e-01 -4.77735437e-02 -3.73432338e-02
-4.28429306e-01 1.88442335e-01 6.90326035e-01 -1.15925878e-01
-5.09097040e-01 3.61973077e-01 -7.46883750e-01 6.66771650e-01
9.86309588e-01 -5.26421309e-01 -5.38932502e-01 -6.05359316e-01
5.19181788e-03 5.17571755e-02 4.80738372e-01 -1.25023746e+00
5.61392307e-01 5.40905520e-02 2.87640810e-01 -4.31219280e-01
7.29056001e-01 -9.54841316e-01 6.25986308e-02 -2.53560930e-01
-8.03408086e-01 1.56424884e-02 3.33059132e-01 5.59772730e-01
-4.24915344e-01 4.14874926e-02 8.74898911e-01 1.52468696e-01
-1.40451357e-01 -1.43662974e-01 -3.82014513e-01 5.85198961e-02
5.46312571e-01 -2.11845323e-01 -3.22733596e-02 -6.64466321e-01
-5.13304353e-01 -4.22716975e-01 2.39757635e-02 2.88136095e-01
6.63567185e-01 -1.35116172e+00 -7.69535542e-01 5.52279174e-01
-1.37492210e-01 6.80858046e-02 2.04975769e-01 6.52638555e-01
-1.70790583e-01 5.45041800e-01 3.17203343e-01 -6.42390847e-01
-1.13644135e+00 3.20903987e-01 9.48599458e-01 1.18262470e-01
-3.29390198e-01 1.25416183e+00 5.01270056e-01 -6.52620494e-01
5.27365804e-01 -6.88152432e-01 -1.17713735e-01 -7.27872178e-02
7.27414787e-01 5.61716795e-01 2.90379912e-01 -8.32118869e-01
-6.12761021e-01 7.72565305e-01 5.60009122e-01 -9.12943542e-01
1.16215205e+00 -2.87573755e-01 -7.62131959e-02 8.06093574e-01
1.71848023e+00 8.09162378e-01 -1.41762698e+00 -3.98977727e-01
-3.37775409e-01 -3.75115216e-01 2.46037364e-01 -8.38806629e-01
-1.00588787e+00 1.00439370e+00 6.68372393e-01 4.22910929e-01
1.56945884e+00 -1.36111051e-01 4.14561570e-01 9.90463719e-02
9.73608717e-03 -9.02113795e-01 5.70012219e-02 4.89168525e-01
1.11226213e+00 -7.34398782e-01 -4.76992041e-01 -1.60384700e-01
-2.57710479e-02 8.17638993e-01 5.61687276e-02 -1.86622754e-01
9.46460605e-01 5.65577328e-01 4.26777929e-01 2.42094621e-01
-4.75532889e-01 1.47130936e-01 7.34802067e-01 8.93311203e-01
4.09872800e-01 9.44863446e-03 7.54329205e-01 5.72754383e-01
-3.83603573e-01 -6.15558803e-01 5.45957759e-02 5.72264075e-01
-5.47574878e-01 -6.00698948e-01 -9.07246113e-01 -1.24084214e-02
-7.74817586e-01 -3.89568001e-01 -3.46998483e-01 2.33447462e-01
1.37532040e-01 1.62782717e+00 2.87860483e-01 -2.83429593e-01
5.05331278e-01 -9.01825055e-02 5.69071360e-02 -6.23682022e-01
-8.58642280e-01 5.27588785e-01 6.05752394e-02 -4.51986998e-01
-4.49000210e-01 -5.84539056e-01 -9.38800275e-01 2.55439401e-01
-1.85882434e-01 3.36083829e-01 7.29290485e-01 7.60049224e-01
4.60142702e-01 1.00975680e+00 1.26260042e+00 -1.08412278e+00
-2.17934966e-01 -1.11980641e+00 -7.53389478e-01 -1.17646053e-01
1.35828805e+00 7.01623857e-02 -7.56298840e-01 3.20864916e-01] | [15.111133575439453, 5.769516468048096] |
1b1d3600-b2d1-45a0-a1fa-e1ca8fc112a1 | reasoning-over-different-types-of-knowledge | 2212.05767 | null | https://arxiv.org/abs/2212.05767v6 | https://arxiv.org/pdf/2212.05767v6.pdf | A Survey of Knowledge Graph Reasoning on Graph Types: Static, Dynamic, and Multimodal | Knowledge graph reasoning (KGR), aiming to deduce new facts from existing facts based on mined logic rules underlying knowledge graphs (KGs), has become a fast-growing research direction. It has been proven to significantly benefit the usage of KGs in many AI applications, such as question answering and recommendation systems, etc. According to the graph types, the existing KGR models can be roughly divided into three categories, i.e., static models, temporal models, and multi-modal models. The early works in this domain mainly focus on static KGR and tend to directly apply general knowledge graph embedding models to the reasoning task. However, these models are not suitable for more complex but practical tasks, such as inductive static KGR, temporal KGR, and multi-modal KGR. To this end, multiple works have been developed recently, but no survey papers and open-source repositories comprehensively summarize and discuss models in this important direction. To fill the gap, we conduct a survey for knowledge graph reasoning tracing from static to temporal and then to multi-modal KGs. Concretely, the preliminaries, summaries of KGR models, and typical datasets are introduced and discussed consequently. Moreover, we discuss the challenges and potential opportunities. The corresponding open-source repository is shared on GitHub: https://github.com/LIANGKE23/Awesome-Knowledge-Graph-Reasoning. | ['Fuchun Sun', 'Xinwang Liu', 'Sihang Zhou', 'Siwei Wang', 'Wenxuan Tu', 'Yue Liu', 'Meng Liu', 'Lingyuan Meng', 'Ke Liang'] | 2022-12-12 | null | null | null | null | ['knowledge-graph-embedding', 'general-knowledge'] | ['graphs', 'miscellaneous'] | [-3.69607836e-01 6.46863461e-01 -7.78911233e-01 -1.33508340e-01
-1.44428894e-01 -5.60185075e-01 4.71010774e-01 2.37522185e-01
1.54762715e-01 8.13408554e-01 1.44963652e-01 -6.40915632e-01
-6.78851008e-01 -1.35127866e+00 -5.81269443e-01 -3.24824870e-01
1.16783539e-02 5.05815029e-01 5.75208485e-01 -4.22182530e-01
-5.24069741e-02 1.68740049e-01 -1.54559553e+00 2.26777971e-01
7.24799156e-01 7.14831054e-01 -3.34726363e-01 1.40392959e-01
-4.07167464e-01 1.37812543e+00 -4.79321927e-02 -1.14767122e+00
-1.01443574e-01 -2.21574992e-01 -1.22010911e+00 -4.44436878e-01
-1.55147091e-01 -1.25327511e-02 -9.58715975e-01 1.21803129e+00
1.60892919e-01 1.82757348e-01 3.06622058e-01 -1.77099872e+00
-1.06692243e+00 1.17657411e+00 -2.18594402e-01 3.88987921e-02
6.42435789e-01 -3.09778482e-01 1.32731211e+00 -5.92666805e-01
7.42713869e-01 1.15966904e+00 6.16853118e-01 5.97500443e-01
-4.94172424e-01 -3.96800518e-01 4.69025701e-01 1.24981034e+00
-1.53856421e+00 -7.67873675e-02 1.04692340e+00 -2.42436662e-01
8.46382201e-01 4.29572672e-01 8.87748301e-01 7.74181426e-01
-9.70718563e-02 1.07689750e+00 9.51451480e-01 -5.60519993e-01
4.57874388e-02 3.02821517e-01 4.98302102e-01 9.27044630e-01
6.58258438e-01 -1.50649503e-01 -7.24198520e-01 -7.59866089e-02
5.96900344e-01 -9.03750062e-02 -4.88054633e-01 -5.46341658e-01
-9.84877825e-01 7.67761707e-01 3.01877886e-01 5.73436618e-01
-1.53990671e-01 5.32231554e-02 4.07881886e-01 3.18533123e-01
2.58910447e-01 1.37592569e-01 -6.76810801e-01 7.67730176e-02
-4.78848845e-01 4.04598296e-01 1.03849041e+00 1.21323144e+00
6.35821581e-01 3.70498523e-02 1.40696064e-01 6.46879017e-01
5.60784459e-01 2.36911178e-01 3.09323102e-01 -6.70932591e-01
3.75228673e-01 1.09480774e+00 -1.60816982e-01 -1.30890894e+00
-3.45075071e-01 -1.06367253e-01 -6.18483007e-01 -4.35237259e-01
1.09805241e-01 1.00421719e-01 -4.97341901e-01 1.40187550e+00
7.31068134e-01 2.15285629e-01 4.56935376e-01 6.44906223e-01
1.66272414e+00 4.10870194e-01 1.05379242e-02 -1.88822374e-01
1.43161654e+00 -9.83425677e-01 -9.88409460e-01 4.17483822e-02
7.00170100e-01 -9.22256485e-02 7.03374147e-01 2.45765477e-01
-1.02917969e+00 -7.10869357e-02 -9.54286575e-01 -2.91894346e-01
-1.16066408e+00 -1.28300235e-01 1.15017045e+00 4.33379203e-01
-1.07721782e+00 2.55050868e-01 -6.08220696e-01 -5.18718898e-01
3.54040474e-01 7.87368324e-03 -3.28005999e-01 -4.71762568e-01
-2.02753210e+00 1.00240457e+00 9.72668529e-01 2.15152457e-01
-3.52153957e-01 -8.06477487e-01 -9.81234312e-01 -7.34392703e-02
1.26148582e+00 -1.07529235e+00 9.44821477e-01 -1.70196816e-01
-1.25133836e+00 7.34340847e-01 -4.45083156e-02 -4.48607028e-01
3.99852604e-01 -1.06607392e-01 -1.02927065e+00 1.00761928e-01
-1.49505705e-01 -5.61849661e-02 3.42981637e-01 -1.22708774e+00
-5.86196244e-01 -4.75075573e-01 7.03842759e-01 1.24962710e-01
-3.58010590e-01 -1.02909856e-01 -7.61122704e-01 -3.85931432e-01
-1.48010431e-02 -5.24837315e-01 2.90173683e-02 -4.36222762e-01
-4.72397655e-01 -6.19966984e-01 7.51464427e-01 -6.97718322e-01
1.76000953e+00 -1.71013582e+00 2.26611525e-01 2.37655804e-01
4.46918428e-01 2.73934692e-01 5.58187962e-01 7.63378382e-01
-4.09202129e-02 1.32387415e-01 1.18635576e-02 5.80043912e-01
3.44229281e-01 4.06733721e-01 -5.49408674e-01 -4.99824025e-02
-2.68040210e-01 1.60245991e+00 -1.21430337e+00 -7.33620584e-01
3.93801183e-01 2.02884883e-01 -1.48678690e-01 -2.38059491e-01
-5.59674025e-01 -9.07606781e-02 -6.26007259e-01 8.80553961e-01
2.12688237e-01 -5.55830717e-01 6.23798013e-01 -5.45092523e-01
2.61022240e-01 3.10973525e-02 -1.13664842e+00 1.34210479e+00
-1.93096161e-01 2.17573643e-01 -4.27353024e-01 -1.20786476e+00
7.00981259e-01 4.72329527e-01 3.81813616e-01 -3.14762086e-01
9.42650139e-02 1.89153552e-01 -1.39737830e-01 -7.03693986e-01
5.82058907e-01 -3.14567059e-01 -5.18339872e-02 5.59747918e-03
-9.98237636e-03 2.96906177e-02 3.92496645e-01 4.97092158e-01
1.08805871e+00 2.10823089e-01 7.72044003e-01 8.27115178e-02
7.01200902e-01 3.38833779e-01 5.29004216e-01 4.30716395e-01
-1.34228179e-02 -2.48639196e-01 5.77278733e-01 -5.06046414e-01
-2.70547897e-01 -9.92158353e-01 2.24681981e-02 5.50856233e-01
4.64096636e-01 -9.47987080e-01 -2.40984902e-01 -1.01401484e+00
2.43245035e-01 9.82332945e-01 -6.45890296e-01 -2.98072845e-01
-2.12495491e-01 -6.38755679e-01 7.08580017e-01 7.38171697e-01
6.75720751e-01 -1.14450717e+00 -1.06034026e-01 4.33093868e-02
-4.14855123e-01 -1.21165407e+00 3.71339262e-01 -4.08561587e-01
-7.94865370e-01 -1.61006808e+00 -9.46799666e-02 -6.33767664e-01
4.94505197e-01 1.13047406e-01 1.21518779e+00 2.71351784e-01
-1.32444680e-01 9.32841241e-01 -7.42344320e-01 -4.82422709e-01
4.07200344e-02 -1.36317030e-01 4.67958115e-02 -1.66586846e-01
6.68488920e-01 -5.97250044e-01 -2.99461067e-01 2.70152688e-01
-1.03546154e+00 1.68764517e-01 4.48120922e-01 5.30939102e-01
7.58324981e-01 7.83241570e-01 7.86821067e-01 -1.28628862e+00
5.71074247e-01 -7.07320154e-01 -4.28505957e-01 9.14670289e-01
-9.32979286e-01 6.92073703e-02 6.31691456e-01 -2.71776300e-02
-1.11240280e+00 -6.27117097e-01 -4.38226387e-03 -4.54105377e-01
1.44565672e-01 1.21957898e+00 -3.81616920e-01 1.61360484e-02
3.30013663e-01 3.14585239e-01 -4.84990954e-01 -2.81772703e-01
8.41155291e-01 4.66322839e-01 4.34405118e-01 -8.95386457e-01
1.04873013e+00 4.34982508e-01 1.20507210e-01 -4.47515517e-01
-1.06072581e+00 -3.94719005e-01 -3.91117334e-01 -4.41398501e-01
5.91230690e-01 -5.72326720e-01 -7.79726982e-01 2.23031431e-01
-7.31506467e-01 -4.20833558e-01 -4.83120888e-01 3.23122859e-01
-4.87654001e-01 5.11420965e-01 -4.88565087e-01 -7.50180662e-01
-3.71649832e-01 -5.13295531e-01 4.56151932e-01 2.66525626e-01
1.72274839e-02 -1.49794865e+00 -1.26454890e-01 7.63805389e-01
1.54631808e-01 3.81974936e-01 1.39995003e+00 -6.31692827e-01
-6.54935718e-01 -1.82657450e-01 -1.40100405e-01 2.17334591e-02
2.36563578e-01 -4.90281247e-02 -5.10922551e-01 1.14602603e-01
-3.96522820e-01 -2.37717882e-01 6.43273234e-01 -4.27964479e-02
1.23687005e+00 -3.56006384e-01 -6.25041127e-01 1.33864194e-01
1.45047808e+00 2.43921295e-01 7.48044431e-01 3.70061845e-01
8.25250506e-01 4.57471848e-01 8.24768305e-01 4.70924973e-02
1.04029584e+00 5.52422822e-01 2.86418766e-01 5.95070601e-01
-5.80226555e-02 -5.48743248e-01 1.43980965e-01 1.08652782e+00
-6.84902966e-01 -1.61348179e-01 -1.06248009e+00 7.14936197e-01
-2.40721774e+00 -1.10696173e+00 -1.76530138e-01 1.82185173e+00
9.29301858e-01 -2.69063096e-02 -1.18823871e-01 3.10233325e-01
7.38661230e-01 2.26773173e-01 -6.18665457e-01 -7.97036886e-02
-1.87891707e-01 2.72536203e-02 9.71843228e-02 5.62347114e-01
-6.79505229e-01 1.24875259e+00 5.33656311e+00 8.83917391e-01
-5.04154205e-01 1.12639166e-01 -1.91068158e-01 2.26535842e-01
-7.03527689e-01 4.04641867e-01 -7.44889736e-01 1.21677279e-01
6.38732195e-01 -7.99297214e-01 6.15794182e-01 9.84779239e-01
-4.50763315e-01 1.48196504e-01 -8.81059349e-01 1.09106934e+00
1.59001425e-01 -1.51938391e+00 4.13830519e-01 -2.39181787e-01
5.50781965e-01 -4.50662255e-01 -3.09060901e-01 7.96406865e-01
6.57207549e-01 -7.06009328e-01 4.31902140e-01 7.12338567e-01
5.23127496e-01 -6.83258951e-01 7.95570552e-01 2.27523297e-01
-1.50830376e+00 -2.34138351e-02 -2.24756107e-01 -3.47031862e-03
3.24937433e-01 8.20265114e-01 -3.91695321e-01 1.81776547e+00
8.29721987e-01 1.04004920e+00 -5.10093689e-01 6.80480123e-01
-7.83437312e-01 4.26658303e-01 9.78362095e-03 4.88069747e-03
-2.34211117e-01 -4.16404381e-02 3.64682943e-01 7.70300388e-01
9.57422182e-02 3.41737837e-01 -7.60166198e-02 6.59661770e-01
-9.38902199e-02 1.15475848e-01 -6.21182382e-01 -2.44102925e-01
5.07516563e-01 1.18707657e+00 -5.19511163e-01 -5.08736432e-01
-7.20527172e-01 4.87053424e-01 4.94219244e-01 3.59746277e-01
-1.04924548e+00 -4.06310529e-01 2.30650380e-01 8.10344517e-02
1.56661347e-01 -2.81345695e-02 1.08626604e-01 -1.54688466e+00
1.88125551e-01 -7.75417328e-01 1.13170242e+00 -1.05200243e+00
-1.48316216e+00 2.83313155e-01 6.02665544e-01 -7.73649096e-01
-6.98307380e-02 -7.18652189e-01 -2.58035004e-01 3.85254383e-01
-1.72404635e+00 -1.35034287e+00 -4.05056030e-01 9.14901555e-01
8.77983123e-03 2.77146567e-02 7.42252946e-01 3.08173567e-01
-5.36006093e-01 4.36781675e-01 -4.08983439e-01 1.88766554e-01
2.89319843e-01 -1.25682688e+00 -4.78528477e-02 6.95325255e-01
3.32599938e-01 8.88484538e-01 3.15314949e-01 -9.00543511e-01
-1.70281601e+00 -1.29792416e+00 9.60162640e-01 -6.67074680e-01
1.08466482e+00 5.60509041e-02 -1.05959773e+00 1.32109475e+00
-1.24071576e-02 1.80609211e-01 6.68325007e-01 4.98056173e-01
-6.19637668e-01 -2.25418985e-01 -9.63965416e-01 9.50193822e-01
1.45059764e+00 -4.39225823e-01 -8.98665905e-01 3.97945791e-01
7.51051247e-01 -4.98982877e-01 -1.43224585e+00 7.52562106e-01
4.04761493e-01 -7.06456900e-01 9.02982175e-01 -8.70538652e-01
2.23626241e-01 -6.72204733e-01 -6.90582320e-02 -9.37464178e-01
-3.77082914e-01 -3.96580249e-01 -1.11628127e+00 1.24065745e+00
3.30011666e-01 -9.54489708e-01 6.10831797e-01 5.92997015e-01
-1.24757469e-01 -1.22042668e+00 -7.47566044e-01 -8.61473203e-01
-2.12994799e-01 -6.52210057e-01 8.22695315e-01 1.24826610e+00
6.15282834e-01 3.89431685e-01 -1.39079273e-01 3.81997019e-01
4.97474223e-01 6.38443232e-01 5.50729692e-01 -1.32086873e+00
-3.23895425e-01 -3.12896192e-01 -6.25910163e-01 -7.80194402e-01
3.26145589e-01 -1.29212296e+00 -5.85387886e-01 -2.28380585e+00
2.06783384e-01 -2.75729686e-01 -3.59454423e-01 1.05212605e+00
-1.40581861e-01 -1.96173146e-01 -1.76216438e-01 3.88897508e-02
-8.50303650e-01 3.55458915e-01 1.14930427e+00 -1.97555408e-01
5.87649923e-03 -2.68718719e-01 -1.02868974e+00 7.94550717e-01
8.92992377e-01 1.28827374e-02 -9.71088052e-01 -1.14748232e-01
9.75569069e-01 3.99928598e-04 5.84808886e-01 -5.38350046e-01
3.75535011e-01 -4.15462673e-01 -2.04391077e-01 -4.82050270e-01
1.12997510e-01 -8.55842292e-01 5.81239462e-01 1.47066355e-01
4.98737134e-02 -1.83442757e-01 2.08412379e-01 7.25536644e-01
-4.48554754e-01 -2.38546193e-01 1.40558317e-01 -2.39232257e-01
-1.35968840e+00 4.96468842e-01 8.33489299e-02 2.64260709e-01
1.23634100e+00 -1.36435786e-02 -6.18279517e-01 -3.24508816e-01
-8.79505336e-01 4.70733374e-01 2.80026998e-02 5.98978221e-01
6.59660041e-01 -1.50137484e+00 -3.82492959e-01 -4.52609777e-01
3.86177659e-01 1.56665742e-01 6.19840205e-01 9.79063869e-01
-1.70092314e-01 6.74505055e-01 2.81879485e-01 1.21194981e-01
-1.04798949e+00 1.08853185e+00 1.29416674e-01 -5.18492460e-01
-8.16714585e-01 3.77116829e-01 -1.66007847e-01 -3.86308461e-01
1.53035950e-02 -2.81789482e-01 -5.01420319e-01 -1.78744063e-01
4.13488686e-01 5.59516013e-01 8.66243616e-02 -3.21152538e-01
-6.14225626e-01 4.96294647e-01 2.16498412e-02 3.77486348e-01
1.20771694e+00 -2.69006770e-02 -4.84818608e-01 7.04200864e-01
5.39049387e-01 -4.56553400e-02 -2.27707177e-01 -5.33915639e-01
3.45325060e-02 -1.28005028e-01 -1.62018463e-01 -8.57862294e-01
-1.14931929e+00 3.37618113e-01 -3.60231191e-01 5.77022731e-01
9.64095712e-01 5.12560070e-01 6.77969873e-01 5.71618617e-01
9.05295312e-01 -9.88457501e-01 -2.75381148e-01 4.64394063e-01
9.02363181e-01 -7.76910603e-01 2.01455370e-01 -8.18175554e-01
-5.18187940e-01 9.93547797e-01 6.13709748e-01 3.65081191e-01
9.70080197e-01 2.06158366e-02 -2.76023418e-01 -6.64445460e-01
-6.70191646e-01 -4.01112169e-01 2.25738287e-01 6.27587259e-01
9.69742537e-02 1.25664830e-01 -3.26227307e-01 1.09164655e+00
-4.55900311e-01 4.48360413e-01 2.37479761e-01 7.85190463e-01
-5.73199056e-02 -1.15792286e+00 4.38404866e-02 5.92910171e-01
-9.95274633e-02 -1.11394413e-01 -4.56350029e-01 1.16873705e+00
4.38093953e-02 9.45566356e-01 -6.88557506e-01 -5.26993155e-01
5.63949108e-01 3.97502303e-01 7.95404851e-01 -3.21438819e-01
1.36048393e-02 -9.76387024e-01 4.42143410e-01 -3.33100677e-01
-7.60418355e-01 -4.30110097e-01 -1.29949903e+00 -5.07448554e-01
-4.61959988e-01 3.92530620e-01 1.82285234e-02 9.08484042e-01
2.16129750e-01 6.19614422e-01 -9.22021940e-02 2.24619880e-01
-1.10417604e-02 -4.82096642e-01 -7.51273274e-01 2.08230272e-01
-4.39698488e-01 -9.49614346e-01 -2.63204277e-01 2.36411206e-02] | [8.863030433654785, 7.893924236297607] |
d05483ec-faaf-4939-90e3-2f662ae45133 | camera-based-image-forgery-localization-using | 1808.09714 | null | http://arxiv.org/abs/1808.09714v1 | http://arxiv.org/pdf/1808.09714v1.pdf | Camera-based Image Forgery Localization using Convolutional Neural Networks | Camera fingerprints are precious tools for a number of image forensics tasks.
A well-known example is the photo response non-uniformity (PRNU) noise pattern,
a powerful device fingerprint. Here, to address the image forgery localization
problem, we rely on noiseprint, a recently proposed CNN-based camera model
fingerprint. The CNN is trained to minimize the distance between same-model
patches, and maximize the distance otherwise. As a result, the noiseprint
accounts for model-related artifacts just like the PRNU accounts for
device-related non-uniformities. However, unlike the PRNU, it is only mildly
affected by residuals of high-level scene content. The experiments show that
the proposed noiseprint-based forgery localization method improves over the
PRNU-based reference. | ['Luisa Verdoliva', 'Davide Cozzolino'] | 2018-08-29 | null | null | null | null | ['image-forensics'] | ['computer-vision'] | [ 2.44771436e-01 -5.52878976e-01 9.28922147e-02 1.14726778e-02
-7.83878446e-01 -6.72773957e-01 3.11975896e-01 -2.00587347e-01
4.50964607e-02 3.57395172e-01 -5.01052327e-02 -1.79533467e-01
8.05070996e-02 -6.11698925e-01 -1.07372308e+00 -8.02161336e-01
4.18718159e-01 -4.22972977e-01 1.81013852e-01 3.18461031e-01
5.87180495e-01 6.07354462e-01 -1.24108768e+00 2.58320868e-01
6.29240215e-01 1.31071353e+00 7.42569119e-02 2.37644821e-01
3.55440766e-01 8.14175725e-01 -6.61339641e-01 -6.21815801e-01
3.87750685e-01 -3.55879545e-01 -1.69174910e-01 -8.83102976e-03
8.98078799e-01 -7.61072993e-01 -7.83643425e-01 1.64702189e+00
3.69953930e-01 -1.43627271e-01 3.22518885e-01 -1.08032322e+00
-9.34664130e-01 1.25842854e-01 -6.60046875e-01 2.11967036e-01
1.33842036e-01 2.58824915e-01 2.09726527e-01 -8.46541584e-01
5.63229263e-01 1.17646098e+00 1.24645078e+00 3.09235424e-01
-1.03932571e+00 -8.74426067e-01 -7.57390440e-01 3.89515936e-01
-1.54116082e+00 -3.21057767e-01 1.12977111e+00 -2.43656844e-01
5.57770729e-01 1.18657269e-01 2.00961828e-02 1.32552779e+00
5.85204780e-01 4.52351332e-01 1.14122653e+00 -4.53238070e-01
8.15385729e-02 2.07090881e-02 1.94238365e-01 5.31328321e-01
6.46159470e-01 4.71079409e-01 -4.19791311e-01 -3.25632870e-01
1.15488434e+00 3.12135518e-01 -6.40820026e-01 -8.41316581e-02
-7.03906357e-01 3.55580032e-01 4.33010191e-01 3.77130419e-01
-1.01260006e-01 4.75764662e-01 4.45450470e-02 1.64771184e-01
6.89869002e-02 6.74340904e-01 -3.15967277e-02 -2.18886644e-01
-1.14287829e+00 4.99531738e-02 4.17179197e-01 1.06530666e+00
1.10196972e+00 2.31857032e-01 -2.66794145e-01 8.58751297e-01
-1.51994333e-01 4.65070158e-01 2.44556710e-01 -1.04251838e+00
3.92343163e-01 3.92384708e-01 3.30536515e-01 -1.61108053e+00
-3.73752415e-03 -6.02648079e-01 -1.02624106e+00 2.19533399e-01
5.14689088e-01 2.01088145e-01 -8.07272434e-01 1.39320219e+00
-1.41097739e-01 4.66435343e-01 -6.48460627e-01 8.36948752e-01
4.31171596e-01 2.18790442e-01 -4.05833155e-01 2.08165906e-02
1.04605806e+00 -9.38875973e-01 -6.79265857e-01 -4.05550711e-02
1.42099336e-01 -1.05917418e+00 9.23584342e-01 6.67516112e-01
-6.57484174e-01 -7.99821556e-01 -1.08430541e+00 3.59133258e-02
-9.03945267e-02 3.19546402e-01 1.28599644e-01 1.03627706e+00
-9.94444013e-01 9.93037522e-01 -3.37659061e-01 -1.56302482e-01
4.78022277e-01 2.40916312e-02 -4.99028772e-01 -4.99433756e-01
-7.67818570e-01 7.84274757e-01 -8.57290775e-02 1.64419800e-01
-8.05812836e-01 -5.87490916e-01 -4.59094584e-01 8.90141949e-02
3.11760575e-01 -8.47859867e-03 8.60195816e-01 -7.67282903e-01
-1.42402112e+00 6.97945297e-01 -9.84733552e-02 -2.91430235e-01
8.12265992e-01 -7.23346025e-02 -6.63155675e-01 1.86364979e-01
4.80217934e-02 -8.10344964e-02 1.16988063e+00 -1.46183550e+00
1.54511286e-02 -2.34920621e-01 -3.42179686e-01 -6.97077692e-01
-3.29768330e-01 1.81939766e-01 -8.15845311e-01 -9.22692299e-01
8.18664581e-02 -6.65179849e-01 7.13974088e-02 1.90846160e-01
-7.50665665e-01 4.37901437e-01 1.05130553e+00 -9.20035839e-01
1.18966556e+00 -2.33838773e+00 -9.31596577e-01 1.68563575e-01
2.11345285e-01 3.83316427e-01 -2.84629762e-01 3.88734490e-01
-6.97017163e-02 1.19978428e-01 -2.88226455e-01 -3.67791355e-01
-1.60958081e-01 -2.20823959e-01 -3.88440341e-01 7.76505291e-01
2.11966828e-01 9.17441010e-01 -7.42648602e-01 -8.72215163e-03
3.85415971e-01 4.14822280e-01 -2.02618882e-01 4.65064049e-02
7.21554309e-02 3.71014655e-01 -1.37817264e-01 9.09145355e-01
1.52356184e+00 -1.12370089e-01 -1.99039176e-01 -6.19716465e-01
-6.86638430e-02 -3.12186658e-01 -8.90575528e-01 1.47978115e+00
-3.65845233e-01 9.89551544e-01 1.19720921e-01 -3.99103224e-01
9.55143750e-01 -9.37586874e-02 1.69990659e-01 -9.08887029e-01
2.81428725e-01 4.21968549e-01 -2.35787436e-01 -6.39094710e-01
7.78502584e-01 1.47323459e-01 2.33256727e-01 2.71677554e-01
-5.47446832e-02 5.03998935e-01 -6.11287296e-01 -1.09939434e-01
1.47102332e+00 -1.07578658e-01 -7.63761476e-02 -2.76071578e-01
1.71689585e-01 -4.60465610e-01 7.02888787e-01 1.06303668e+00
-2.61755019e-01 1.34241831e+00 3.99842083e-01 -3.29680502e-01
-1.27048016e+00 -8.88910472e-01 -1.85986564e-01 -2.47598719e-02
6.26485288e-01 -2.08140090e-01 -1.00668871e+00 -5.93903005e-01
7.52587244e-02 4.22806352e-01 -6.36134624e-01 -1.84542775e-01
-7.17112243e-01 -3.24763566e-01 9.66271222e-01 5.23711562e-01
8.48518252e-01 -6.68288291e-01 -1.77338600e-01 1.50842577e-01
-1.13802001e-01 -1.51122797e+00 -8.61963093e-01 -2.21285582e-01
-5.55058777e-01 -1.35247838e+00 -6.74582481e-01 -4.07768816e-01
6.64976120e-01 7.72017777e-01 8.04876328e-01 3.33534569e-01
-4.56563622e-01 2.45227456e-01 -1.63937241e-01 9.80621725e-02
-4.15042162e-01 -6.38197541e-01 -6.49138466e-02 6.00697100e-01
2.50645250e-01 -6.04485393e-01 -8.41372311e-01 6.10115826e-01
-9.56980348e-01 -2.95869261e-01 5.59151053e-01 9.08971190e-01
6.05542541e-01 3.15223843e-01 2.25640729e-01 -6.97766900e-01
5.12451172e-01 -3.14128637e-01 -6.18131638e-01 2.59067744e-01
-6.71341300e-01 -1.12376116e-01 7.70131648e-01 -8.40958595e-01
-9.01482105e-01 -1.50420278e-01 2.80204237e-01 -1.28437006e+00
-1.87355697e-01 9.77514163e-02 -6.72463298e-01 -7.87880898e-01
6.59670174e-01 2.99159378e-01 -4.09912825e-01 -8.08126152e-01
3.70524824e-02 6.52797878e-01 1.11890471e+00 -4.31964159e-01
9.07449126e-01 5.95161557e-01 5.28791547e-02 -9.36904073e-01
-2.78168559e-01 -5.70486486e-01 -1.51896685e-01 -2.93779999e-01
5.53046882e-01 -6.98136330e-01 -6.52023852e-01 1.04272425e+00
-1.66154027e+00 -2.38298178e-01 4.02934812e-02 2.32180864e-01
-2.81840801e-01 1.05487514e+00 -6.42274499e-01 -9.58591759e-01
-1.87291697e-01 -1.17668819e+00 1.00174737e+00 3.28044802e-01
1.57475784e-01 -5.93498349e-01 -1.51296735e-01 2.11730510e-01
6.28008723e-01 4.84178454e-01 8.64631951e-01 -2.02970892e-01
-9.47967708e-01 -8.05814385e-01 -6.47488892e-01 7.58939266e-01
1.71008781e-01 -1.55355975e-01 -1.27992558e+00 -2.57958055e-01
5.90661705e-01 6.26444221e-02 7.46440709e-01 1.96621180e-01
1.45792997e+00 -1.84204414e-01 -1.81281269e-01 9.34595466e-01
1.88179564e+00 2.18118995e-01 1.27091694e+00 3.68811518e-01
9.69421208e-01 1.90949440e-01 1.98713928e-01 3.03203166e-01
-2.28597432e-01 9.94487405e-01 5.06964982e-01 -1.87461227e-01
-4.77505863e-01 -5.51773369e-01 1.73538521e-01 3.16584021e-01
1.76874727e-01 -5.51641226e-01 -7.33084857e-01 4.09146070e-01
-1.64618587e+00 -1.00159967e+00 -5.77465892e-01 2.42225957e+00
2.42191166e-01 -2.37739116e-01 -2.79417604e-01 -6.99433237e-02
1.24404490e+00 1.27827048e-01 -5.87481976e-01 3.22775096e-02
-5.89577496e-01 2.95132697e-01 1.02077985e+00 1.58859313e-01
-8.73830318e-01 7.39927649e-01 6.12510586e+00 1.22643316e+00
-1.11905348e+00 4.15779024e-01 5.51903903e-01 3.08281839e-01
-1.14938140e-01 2.59782560e-02 -4.48310077e-01 1.00174522e+00
5.46490788e-01 3.88956696e-01 4.93807316e-01 7.73190141e-01
2.87771314e-01 -2.14941114e-01 -8.92961323e-01 1.65164852e+00
4.39711004e-01 -1.41543961e+00 -1.52924150e-01 2.34212205e-01
8.53198409e-01 -2.86670715e-01 3.55346888e-01 -3.25783819e-01
-1.84410810e-01 -9.84934211e-01 9.38191056e-01 6.37871921e-01
1.06329250e+00 -6.34399593e-01 8.23431373e-01 2.17614084e-01
-9.68827069e-01 -2.55114399e-02 -7.06952691e-01 2.81865418e-01
1.53537676e-01 1.00376928e+00 -2.05226868e-01 5.14524281e-01
6.70359313e-01 3.82522643e-01 -5.05527496e-01 1.45451844e+00
-2.32558772e-01 6.33793473e-01 -1.73133239e-01 5.35986245e-01
-7.95868263e-02 -1.70888782e-01 5.32348931e-01 1.23907590e+00
7.40830421e-01 -2.35479400e-01 -2.99224526e-01 1.38319957e+00
-5.41730523e-01 -4.00791645e-01 -6.06810629e-01 2.14119241e-01
6.20185852e-01 1.09904981e+00 -5.91411412e-01 6.45435825e-02
-2.17729241e-01 1.48719001e+00 -5.50122038e-02 5.22011817e-01
-8.22703183e-01 -5.09553730e-01 5.90059996e-01 2.90024906e-01
3.70135516e-01 -3.50660384e-02 -3.42424095e-01 -1.31632304e+00
4.90586579e-01 -8.77058804e-01 -3.36263657e-01 -1.01989579e+00
-1.52424300e+00 4.20900732e-01 -6.01317048e-01 -1.70135915e+00
1.64572671e-01 -6.97351336e-01 -9.62957501e-01 8.82098496e-01
-1.45834446e+00 -1.21205974e+00 -4.84122455e-01 4.90003288e-01
2.53535092e-01 -8.62165727e-03 4.57662672e-01 6.15541577e-01
-5.88449657e-01 1.00399601e+00 5.00213742e-01 3.74029934e-01
8.80022407e-01 -6.36052132e-01 5.42829633e-01 1.28340840e+00
-9.41943303e-02 8.47098351e-01 5.06291568e-01 -9.75428402e-01
-1.48191726e+00 -1.19893503e+00 4.38853353e-01 -5.21363139e-01
4.60632890e-01 -3.10169518e-01 -1.09056163e+00 6.34178668e-02
6.51860610e-02 2.74093330e-01 2.60582834e-01 -5.90610027e-01
-9.33648705e-01 -1.13561079e-01 -1.46206784e+00 2.12046713e-01
8.06515336e-01 -9.88343835e-01 1.30379602e-01 -2.79159453e-02
4.11021620e-01 -2.80564725e-01 -5.04050612e-01 1.90712303e-01
6.76395774e-01 -1.48928225e+00 9.06936347e-01 5.34993634e-02
4.58200365e-01 -3.61172974e-01 -2.59392023e-01 -8.07189107e-01
-3.74134213e-01 -9.74640012e-01 -1.91293240e-01 1.46431041e+00
-3.05384095e-03 -3.69713098e-01 9.43088531e-01 5.06666064e-01
-1.91224515e-01 -2.41875127e-01 -1.03220177e+00 -1.41764891e+00
-2.02324107e-01 -4.91241693e-01 6.79229558e-01 1.08748686e+00
-6.36913598e-01 -3.84686500e-01 -8.93449962e-01 4.19129074e-01
1.08796239e+00 -9.54294503e-02 7.16299415e-01 -8.13579261e-01
-6.24785781e-01 -2.04243660e-01 -5.27903557e-01 -9.69834983e-01
-2.39302248e-01 -2.62389123e-01 1.70665622e-01 -9.03535664e-01
3.54143232e-01 -1.88893706e-01 -3.23177814e-01 3.26556712e-02
-1.17695048e-01 5.71479082e-01 2.58085817e-01 4.59178030e-01
-3.00157160e-01 2.26681307e-01 9.90172565e-01 -3.02156299e-01
1.77880421e-01 1.42400982e-02 -4.72328961e-01 5.87343216e-01
6.73132718e-01 -7.86450326e-01 -5.63823953e-02 -5.55405915e-01
5.25307767e-02 1.56269222e-02 8.23627830e-01 -1.40879083e+00
4.27182108e-01 1.09929681e-01 4.19620603e-01 -5.24095058e-01
2.54772127e-01 -8.04973960e-01 5.47559023e-01 2.48181060e-01
1.30173177e-01 1.01178005e-01 1.53330803e-01 8.38421226e-01
-3.80785227e-01 -3.48274559e-01 9.54521358e-01 -6.54762834e-02
-3.96911412e-01 3.08699697e-01 -3.21177244e-01 -3.37353915e-01
4.09111261e-01 -7.46636927e-01 -8.14763844e-01 -4.71302688e-01
7.86423758e-02 -6.14889801e-01 1.02274132e+00 1.73970655e-01
7.98939288e-01 -1.40470815e+00 -1.67031929e-01 2.28619620e-01
1.89413130e-01 -5.09025753e-01 5.66172063e-01 6.94605112e-01
-5.54955423e-01 1.06677309e-01 -1.55422479e-01 -3.24510962e-01
-9.62291598e-01 7.56649852e-01 4.66013134e-01 -1.14188530e-01
-4.31887239e-01 7.37071216e-01 3.90190989e-01 -6.54948205e-02
1.63869098e-01 -3.16041321e-01 5.01611233e-01 -6.75771117e-01
7.66272902e-01 7.41738677e-01 2.79103190e-01 -6.08525157e-01
-1.52143717e-01 7.94449687e-01 2.22313836e-01 5.93524687e-02
1.17557585e+00 -8.46744478e-02 -2.17420727e-01 -1.79884389e-01
1.50214493e+00 1.91635713e-01 -1.48678899e+00 -1.76885337e-01
-1.03904773e-02 -8.98356080e-01 2.27729842e-01 -8.11309218e-01
-1.25572431e+00 1.02308869e+00 1.01231325e+00 5.50886802e-02
1.13826275e+00 -5.10093689e-01 1.02291858e+00 4.78563122e-02
6.85664833e-01 -9.21578050e-01 2.97405392e-01 2.19520956e-01
7.69256890e-01 -9.92107749e-01 -1.02383278e-01 -4.23285961e-01
-1.66312337e-01 1.22447145e+00 4.20294046e-01 -2.16142699e-01
4.54953372e-01 3.18916351e-01 -9.82056186e-02 -2.73098536e-02
1.58876747e-01 2.55664021e-01 9.56418738e-02 8.79471898e-01
-1.81222171e-01 -1.40203461e-01 1.73643559e-01 7.95106351e-01
9.58148614e-02 -1.56888366e-03 5.88444710e-01 6.30064011e-01
-3.85538451e-02 -1.11968589e+00 -6.49148881e-01 3.74922812e-01
-4.47084963e-01 -1.43581599e-01 -5.49672008e-01 5.03754556e-01
4.31697369e-01 1.16446376e+00 -1.05727270e-01 -9.59123969e-01
3.52628887e-01 -4.46573198e-01 4.85330969e-01 -6.29195794e-02
-7.76350677e-01 -2.43192469e-03 -2.38468021e-01 -9.15812850e-01
-5.73157892e-02 -2.93580472e-01 -4.79665220e-01 -6.15918815e-01
-7.88617373e-01 -4.34071243e-01 6.40349686e-01 6.71459377e-01
5.05826294e-01 5.56118600e-02 7.36666024e-01 -8.85154128e-01
-4.98234898e-01 -8.24313998e-01 -8.18644345e-01 6.58993423e-01
4.79784906e-01 -5.88818908e-01 -5.99205852e-01 -1.38466507e-01] | [12.3809814453125, 0.9952765703201294] |
152ea822-e34f-4e83-885a-521c07c7fae8 | a-new-baseline-for-greenai-finding-the | 2302.10798 | null | https://arxiv.org/abs/2302.10798v2 | https://arxiv.org/pdf/2302.10798v2.pdf | Lightweight Parameter Pruning for Energy-Efficient Deep Learning: A Binarized Gating Module Approach | The subject of green AI has been gaining attention within the deep learning community given the recent trend of ever larger and more complex neural network models. Existing solutions for reducing the computational load of training at inference time usually involve pruning the network parameters. Pruning schemes often create extra overhead either by iterative training and fine-tuning for static pruning or repeated computation of a dynamic pruning graph. We propose a new parameter pruning strategy for learning a lighter-weight sub-network that minimizes the energy cost while maintaining comparable performance to the fully parameterised network on given downstream tasks. Our proposed pruning scheme is green-oriented, as it only requires a one-off training to discover the optimal static sub-networks by dynamic pruning methods. The pruning scheme consists of a binary gating module and a novel loss function to uncover sub-networks with user-defined sparsity. Our method enables pruning and training simultaneously, which saves energy in both the training and inference phases and avoids extra computational overhead from gating modules at inference time. Our results on CIFAR-10 and CIFAR-100 suggest that our scheme can remove 50% of connections in deep networks with 1% reduction in classification accuracy. Compared to other related pruning methods, our method demonstrates a lower drop in accuracy for equivalent reductions in computational cost. | ['Sean Moran', 'Ruibo Shi', 'Fran Silavong', 'Pheobe Sun', 'Varun Babbar', 'Xiaoying Zhi'] | 2023-02-17 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [ 4.71283734e-01 3.61462116e-01 -8.28359574e-02 -5.91207027e-01
-1.95792884e-01 -2.38888800e-01 -1.07678249e-01 8.46172050e-02
-9.13670540e-01 7.24904656e-01 -5.65008163e-01 -6.57156229e-01
-3.82736802e-01 -1.05656660e+00 -7.30986834e-01 -6.22183800e-01
-1.45039037e-01 2.41742343e-01 4.02362198e-01 1.79227620e-01
-6.81795122e-04 6.06033981e-01 -1.54367077e+00 2.48090297e-01
7.86997914e-01 1.35020196e+00 3.02462637e-01 4.82637525e-01
5.93570471e-02 7.05342829e-01 -6.85587764e-01 -6.02517068e-01
6.09375000e-01 -3.08515012e-01 -7.93609738e-01 -5.44856787e-01
5.84594846e-01 -1.30408153e-01 -4.60563600e-01 1.16369832e+00
7.00445771e-01 3.25576365e-01 1.26487732e-01 -1.00395858e+00
-6.09081313e-02 1.01582694e+00 -5.13147295e-01 6.61215782e-01
-7.51598001e-01 -1.33834615e-01 1.08933079e+00 -4.51397270e-01
2.83004284e-01 1.07191181e+00 8.11047435e-01 7.73946166e-01
-1.47305250e+00 -1.15565300e+00 3.06650460e-01 3.41905564e-01
-1.37379789e+00 -6.58352971e-01 7.28086114e-01 6.67324737e-02
1.39129686e+00 2.50617832e-01 9.11918402e-01 7.74379373e-01
1.24922246e-02 5.25741041e-01 4.59383279e-01 -5.06095707e-01
4.99856472e-01 -5.71290813e-02 3.33744109e-01 1.11217761e+00
7.04960406e-01 -1.21964617e-02 -5.01055896e-01 -2.90016551e-02
5.06397724e-01 1.68467894e-01 -1.06744759e-01 -4.65899035e-02
-3.96423310e-01 8.15346241e-01 5.50605595e-01 1.20529838e-01
-5.34653440e-02 7.49241114e-01 4.31817204e-01 3.77104312e-01
3.78969103e-01 4.08141643e-01 -7.55798936e-01 -7.48542398e-02
-1.36234796e+00 -1.14601164e-03 8.91484559e-01 7.20102787e-01
7.41147757e-01 3.06170672e-01 -2.13658996e-02 9.95669782e-01
1.30941480e-01 -3.07346620e-02 3.72832894e-01 -1.04224837e+00
6.66139185e-01 8.93005610e-01 -5.77014208e-01 -7.53163755e-01
-4.88861144e-01 -1.07712209e+00 -9.30201113e-01 3.25103492e-01
1.49193019e-01 -5.04612327e-01 -1.04158115e+00 1.93878710e+00
1.69503868e-01 1.71349853e-01 -7.86437541e-02 4.56722528e-01
6.22970998e-01 3.28654051e-01 1.05497479e-01 -6.57798126e-02
1.23110938e+00 -1.07525694e+00 -3.41806769e-01 -5.09480000e-01
8.56713593e-01 -2.30158478e-01 7.32859015e-01 4.17705894e-01
-1.19910908e+00 -3.30759704e-01 -1.36008453e+00 -2.57228553e-01
-3.69433910e-01 4.45121497e-01 8.40798795e-01 9.30600882e-01
-9.50099170e-01 1.10384023e+00 -9.92101789e-01 1.01453304e-01
1.19809818e+00 1.05341935e+00 4.95445244e-02 5.79813635e-03
-7.84133017e-01 7.36476839e-01 7.29063570e-01 2.36318752e-01
-9.06166136e-01 -9.67044413e-01 -6.30755305e-01 7.58120179e-01
4.53529269e-01 -5.96777916e-01 1.12245274e+00 -1.04771399e+00
-1.49202430e+00 4.81034517e-01 -3.92358154e-02 -9.70346093e-01
2.14106351e-01 -3.83492470e-01 -2.60242373e-01 6.31958842e-02
-4.14621323e-01 1.05301428e+00 8.24716687e-01 -5.64706087e-01
-8.10616672e-01 -1.91996872e-01 2.58744746e-01 -1.32969888e-02
-6.06292486e-01 -3.31188083e-01 -3.92544001e-01 -7.90421784e-01
1.84099153e-01 -8.59142184e-01 -3.01551759e-01 4.06621173e-02
-3.54550183e-01 -1.80963024e-01 1.06334531e+00 -3.13829452e-01
1.44031155e+00 -1.90563548e+00 -1.08340852e-01 2.71071196e-01
5.41977227e-01 5.19067287e-01 -1.72315851e-01 -3.36397797e-01
-1.06104456e-01 2.30363756e-01 -3.23942572e-01 -3.89971435e-01
-3.89175266e-01 3.46283674e-01 -5.37315048e-02 2.76234210e-01
2.85950184e-01 7.72030652e-01 -7.36363709e-01 -4.32459354e-01
-1.56953648e-01 5.90417027e-01 -9.82689857e-01 -3.54096204e-01
-8.43648016e-02 -2.01478988e-01 -1.19595282e-01 5.39961040e-01
4.72895682e-01 -2.47790173e-01 4.37746316e-01 -2.67820686e-01
1.50714323e-01 6.62581384e-01 -9.16788697e-01 1.53422678e+00
-3.99120867e-01 8.61148179e-01 8.32019225e-02 -1.46412027e+00
8.18974674e-01 7.24337548e-02 2.50083983e-01 -7.71498919e-01
4.16418314e-01 2.08840534e-01 5.85703790e-01 4.45831679e-02
-1.64305046e-02 1.62594423e-01 3.25120777e-01 4.17970151e-01
4.16136712e-01 5.09090722e-01 4.20233876e-01 -5.38562946e-02
1.50231230e+00 -1.06850319e-01 -9.84566659e-02 -3.67289126e-01
2.33728841e-01 -2.38370642e-01 9.73440766e-01 6.92776620e-01
4.10409123e-02 -1.78018823e-01 5.71120203e-01 -5.72726130e-01
-7.61504352e-01 -5.76972365e-01 -1.46801084e-01 1.32730019e+00
-9.18992907e-02 -4.70783532e-01 -8.40345442e-01 -8.71654093e-01
-1.24434389e-01 5.91252863e-01 -4.88354504e-01 -6.19661331e-01
-9.39439118e-01 -9.79148924e-01 8.19863558e-01 5.49549818e-01
8.72008860e-01 -8.99287403e-01 -1.31747425e+00 1.98435426e-01
2.65143603e-01 -9.33452606e-01 -2.21921559e-02 1.08618569e+00
-1.78833032e+00 -1.12736297e+00 -2.85289705e-01 -8.82621825e-01
9.76663113e-01 5.92339374e-02 1.06484270e+00 4.32125390e-01
-6.21748745e-01 -4.62835819e-01 5.07812761e-02 -5.08980989e-01
1.29637197e-01 4.90686357e-01 -2.21665919e-01 -4.24695551e-01
4.51169610e-01 -9.35922742e-01 -7.96509743e-01 2.83724274e-02
-5.15232623e-01 2.70759016e-01 7.88521171e-01 8.43048930e-01
6.62617683e-01 3.46418917e-01 5.79618096e-01 -1.10569847e+00
4.40520465e-01 -3.31043839e-01 -8.19573939e-01 1.39209941e-01
-9.30456102e-01 4.05699730e-01 7.33110368e-01 -4.58991140e-01
-8.78988743e-01 2.93168604e-01 2.61069220e-02 -5.54198265e-01
3.26106668e-01 2.49962822e-01 1.41697610e-02 -2.87505418e-01
6.85231268e-01 -1.82223663e-01 -4.69037801e-01 -7.90176809e-01
1.50620686e-02 1.30413752e-02 3.67449135e-01 -4.15331751e-01
5.61611474e-01 2.95364052e-01 3.99595082e-01 -4.73208517e-01
-9.63833153e-01 -3.02911159e-02 -5.14044344e-01 1.24716610e-01
3.60256642e-01 -6.77154243e-01 -8.71071339e-01 -1.46438535e-02
-9.57093656e-01 -4.25630152e-01 -5.58853328e-01 3.58995825e-01
1.72992870e-01 -1.27992988e-01 -5.38752735e-01 -6.74577534e-01
-9.66134667e-01 -9.60999548e-01 5.77588856e-01 3.06010664e-01
-4.70355488e-02 -7.26475418e-01 -2.72835463e-01 -1.76653042e-02
3.38748664e-01 -5.83815798e-02 1.31818795e+00 -5.65955102e-01
-6.72249913e-01 -1.38823748e-01 -3.63144159e-01 5.37022889e-01
-1.43970340e-01 -2.76214898e-01 -1.15287673e+00 -3.26424658e-01
-1.69168770e-01 -2.16990590e-01 1.40125310e+00 5.15707016e-01
1.82776260e+00 -4.67960954e-01 -5.04980803e-01 1.01013064e+00
1.61750984e+00 4.82907027e-01 3.38933170e-01 9.45101678e-02
6.07172310e-01 2.15618014e-01 -3.55468430e-02 2.21386418e-01
-2.99731761e-01 2.48552918e-01 5.63198626e-01 -6.26711058e-04
-1.62419423e-01 2.46967785e-02 1.06009088e-01 6.44066632e-01
-2.27489352e-01 -2.22476199e-01 -8.56620252e-01 4.58171546e-01
-1.68599939e+00 -9.28892970e-01 3.64634484e-01 2.04706526e+00
9.85321999e-01 5.36332667e-01 -3.26906174e-01 5.94261825e-01
3.98227364e-01 -2.66612340e-02 -9.81635571e-01 -6.60369515e-01
1.97428972e-01 1.00873387e+00 7.65267968e-01 3.11313748e-01
-9.82050121e-01 8.16465378e-01 6.35347366e+00 9.00169432e-01
-1.12948394e+00 2.71915346e-01 9.74362135e-01 -9.37966943e-01
3.42624709e-02 -5.57074845e-02 -1.20073915e+00 1.59148186e-01
1.16644311e+00 1.09762102e-01 4.97953981e-01 1.11392844e+00
7.10463971e-02 -1.24143668e-01 -1.35329449e+00 1.00708282e+00
-3.11483920e-01 -1.69388938e+00 -1.62671637e-02 -4.85531501e-02
5.80891609e-01 3.33584338e-01 -3.07255507e-01 4.32586849e-01
3.37887555e-01 -1.07313514e+00 6.23812556e-01 1.74563348e-01
7.55168438e-01 -1.08420706e+00 6.34522438e-01 3.53263766e-01
-1.10720658e+00 -4.67086434e-01 -5.95790565e-01 -1.16367154e-01
-2.63058096e-01 1.00401855e+00 -5.64322948e-01 -3.88480090e-02
1.05857158e+00 4.63300258e-01 -6.10713005e-01 1.17215431e+00
-1.14039317e-01 8.88892531e-01 -6.18844390e-01 -6.31900951e-02
1.35111660e-01 -1.02802806e-01 2.59928346e-01 1.30238676e+00
1.63724035e-01 -1.74148530e-02 -1.01829261e-01 8.45814407e-01
-7.67601788e-01 -3.03151727e-01 -1.85588673e-01 -1.04762517e-01
7.63513505e-01 1.24642324e+00 -9.82151806e-01 -3.57060790e-01
-5.62749133e-02 6.04830682e-01 7.44797230e-01 2.22609431e-01
-8.76953125e-01 -8.15606594e-01 5.86960971e-01 -4.68914658e-02
5.58212698e-01 -9.86239314e-03 -6.50886595e-01 -6.87977135e-01
3.06621380e-03 -3.75971407e-01 3.45724225e-01 -5.43653630e-02
-4.61408675e-01 5.98528922e-01 -1.74494550e-01 -5.86937487e-01
1.74772039e-01 -4.98413622e-01 -6.28415465e-01 6.24750495e-01
-1.60698318e+00 -4.90302265e-01 -2.63250291e-01 3.37804198e-01
5.98428726e-01 -2.60042757e-01 7.64531791e-01 7.13834405e-01
-1.10107243e+00 1.03584361e+00 -8.04183856e-02 1.14507332e-01
7.35445768e-02 -8.41079295e-01 1.87472954e-01 8.84050250e-01
1.98884487e-01 6.52270555e-01 1.60614267e-01 -4.21564311e-01
-1.04600871e+00 -1.31681645e+00 7.88293660e-01 3.16965073e-01
1.91055506e-01 -4.92220342e-01 -7.61632621e-01 5.83049536e-01
-7.89456889e-02 1.71781704e-01 6.78157032e-01 2.77979732e-01
-2.51800418e-01 -6.05427861e-01 -1.13595009e+00 4.94326204e-01
1.50970268e+00 -8.57821181e-02 -2.72025585e-01 1.84381470e-01
6.75715744e-01 -1.96865782e-01 -5.73161125e-01 5.10443270e-01
5.36911547e-01 -6.97242320e-01 8.12557995e-01 -4.27406132e-01
2.32633874e-01 2.65807025e-02 1.23994209e-01 -7.63662577e-01
-3.10005218e-01 -7.27213740e-01 -8.01619649e-01 7.85939336e-01
6.22982919e-01 -3.84266287e-01 1.34478986e+00 4.23862785e-01
-2.13171378e-01 -1.44787717e+00 -1.13928962e+00 -6.85006618e-01
-3.02122623e-01 -5.62552333e-01 3.62545460e-01 5.04863441e-01
-4.41802472e-01 6.01759970e-01 6.03681877e-02 1.52228745e-02
6.01687372e-01 -6.49911314e-02 1.94192275e-01 -1.57592356e+00
-3.82610947e-01 -8.14428985e-01 -2.37149447e-01 -9.66759264e-01
3.08988686e-03 -1.02786553e+00 -9.36215073e-02 -1.33088160e+00
8.36115181e-02 -7.81186163e-01 -5.29072762e-01 1.17058897e+00
1.38851181e-01 2.57070541e-01 -4.19933051e-02 3.74241732e-02
-4.71773177e-01 3.58023435e-01 7.88675606e-01 -1.78477749e-01
-3.33853334e-01 -3.71733084e-02 -6.36177957e-01 1.00579071e+00
9.42950010e-01 -1.06499898e+00 -8.97504449e-01 -7.74463773e-01
4.98765379e-01 -5.29771864e-01 3.37085783e-01 -1.36371922e+00
2.53815591e-01 2.21039712e-01 5.23123622e-01 -5.54449618e-01
4.00534600e-01 -9.57795382e-01 -1.00003570e-01 9.60377812e-01
-3.25548440e-01 -2.61051077e-02 5.67410111e-01 6.10626817e-01
5.41256927e-02 -5.85804760e-01 1.15536487e+00 -9.39723775e-02
-5.11781693e-01 2.95371503e-01 -1.14708774e-01 1.00395963e-01
9.30174589e-01 -5.05245388e-01 -2.85459608e-01 3.32880229e-01
-8.00552428e-01 1.84083074e-01 -3.46444845e-01 4.13997434e-02
5.49277902e-01 -8.96899462e-01 -2.00642377e-01 1.55971125e-01
-6.91442311e-01 3.98519337e-01 2.10987013e-02 6.62612855e-01
-6.49864256e-01 5.11252761e-01 -1.66584149e-01 -2.82038033e-01
-1.42156458e+00 -2.95254234e-02 6.68712795e-01 -6.67778432e-01
-8.92115057e-01 1.52998459e+00 -1.05131395e-01 2.26875730e-02
6.94481432e-01 -5.19715905e-01 7.71774948e-02 -2.73216758e-02
2.74731785e-01 7.41658092e-01 4.60688591e-01 1.75666541e-01
-3.65755349e-01 8.93168524e-02 -3.03542614e-01 2.73990870e-01
1.70834875e+00 3.91491890e-01 -5.27350679e-02 8.48896205e-02
1.27185726e+00 -4.54260796e-01 -1.26427126e+00 -1.38833359e-01
8.34355429e-02 -4.87891883e-02 7.44661093e-01 -9.44230258e-01
-1.87468183e+00 9.68821108e-01 7.90524900e-01 -6.94907829e-02
1.50001693e+00 -3.11764270e-01 9.52845752e-01 9.98794317e-01
9.93519723e-02 -1.39810979e+00 -2.00123951e-01 5.41967452e-01
4.32253271e-01 -8.47315133e-01 3.12029988e-01 -3.13452691e-01
6.27381653e-02 1.15356898e+00 9.00267839e-01 -2.61202514e-01
9.00119185e-01 5.26889086e-01 -5.88272452e-01 -3.52584302e-01
-1.06411469e+00 1.40850663e-01 1.94506988e-01 1.06760800e-01
3.32594872e-01 -1.98289141e-01 -1.98517874e-01 7.21981168e-01
-3.21330547e-01 8.95296484e-02 -1.28943831e-01 1.13266397e+00
-6.75633967e-01 -1.03105891e+00 2.60228574e-01 1.05877221e+00
-7.45436609e-01 -4.69035238e-01 -1.41951710e-01 6.28697276e-01
5.68781435e-01 7.08828986e-01 3.72294545e-01 -4.14042056e-01
2.18863592e-01 3.18378955e-01 6.38759077e-01 -8.05135369e-01
-8.78711522e-01 -8.14916790e-02 2.01186255e-01 -6.70323372e-01
-3.64399105e-01 -1.10238269e-01 -1.37721384e+00 -3.98152947e-01
-6.95500433e-01 -1.30692485e-03 6.51592910e-01 9.24261630e-01
7.25952506e-01 1.01881468e+00 8.48712102e-02 -7.86300182e-01
-3.60772789e-01 -7.60686278e-01 -3.23829383e-01 -3.23259979e-01
1.64201394e-01 -6.52724028e-01 -5.91954626e-02 -3.05445716e-02] | [8.55786418914795, 3.1786952018737793] |
3b3ff4f5-ba39-4f7e-9dbb-374c2d40af8b | higher-order-graph-attention-network-for | 2306.15526 | null | https://arxiv.org/abs/2306.15526v1 | https://arxiv.org/pdf/2306.15526v1.pdf | Higher-order Graph Attention Network for Stock Selection with Joint Analysis | Stock selection is important for investors to construct profitable portfolios. Graph neural networks (GNNs) are increasingly attracting researchers for stock prediction due to their strong ability of relation modelling and generalisation. However, the existing GNN methods only focus on simple pairwise stock relation and do not capture complex higher-order structures modelling relations more than two nodes. In addition, they only consider factors of technical analysis and overlook factors of fundamental analysis that can affect the stock trend significantly. Motivated by them, we propose higher-order graph attention network with joint analysis (H-GAT). H-GAT is able to capture higher-order structures and jointly incorporate factors of fundamental analysis with factors of technical analysis. Specifically, the sequential layer of H-GAT take both types of factors as the input of a long-short term memory model. The relation embedding layer of H-GAT constructs a higher-order graph and learn node embedding with GAT. We then predict the ranks of stock return. Extensive experiments demonstrate the superiority of our H-GAT method on the profitability test and Sharp ratio over both NSDAQ and NYSE datasets | ['Yan Ge', 'Zheng Li', 'Xiang Li', 'Yiping Xia', 'Yang Qiao'] | 2023-06-27 | null | null | null | null | ['graph-attention', 'stock-prediction'] | ['graphs', 'time-series'] | [-8.34471643e-01 3.72776687e-02 -5.11198163e-01 -6.95097670e-02
2.32080162e-01 -6.20350301e-01 6.02863133e-01 4.20496054e-02
3.17046903e-02 3.28202665e-01 5.31875908e-01 -8.42294633e-01
-4.88554716e-01 -1.31156027e+00 -6.44752145e-01 -3.07502747e-01
-4.99695927e-01 4.13686544e-01 7.83374235e-02 -4.35179770e-01
2.90468156e-01 4.34206665e-01 -8.39779615e-01 -1.12470560e-01
6.66372061e-01 1.23855698e+00 -3.06357294e-01 3.20740789e-01
-2.82793880e-01 1.26484585e+00 -2.64410585e-01 -1.14769244e+00
6.37579799e-01 -3.00059348e-01 -2.78511018e-01 5.12266718e-03
9.03496593e-02 -5.33931971e-01 -8.04723263e-01 1.02635300e+00
2.55247384e-01 -9.35338289e-02 8.10924351e-01 -1.32076585e+00
-1.28897202e+00 1.38251328e+00 -7.31767833e-01 8.06814432e-01
-2.34680831e-01 1.03478804e-01 2.06822252e+00 -7.77580142e-01
2.28304461e-01 1.04875541e+00 6.69510782e-01 -1.50118917e-01
-1.15197408e+00 -7.09059179e-01 6.48503184e-01 4.21576723e-02
-9.20158029e-01 2.26846844e-01 1.10998094e+00 -6.73691213e-01
1.16816270e+00 3.49300075e-03 1.28938007e+00 5.72051346e-01
6.37847424e-01 5.93302667e-01 6.01018429e-01 1.39762864e-01
-1.26625702e-01 -1.00070603e-01 4.82017159e-01 5.93647063e-01
7.87528872e-01 2.01845825e-01 -6.65641308e-01 -4.06807996e-02
1.17909098e+00 4.19637799e-01 -1.71690300e-01 -1.55783549e-01
-1.14801705e+00 1.10799015e+00 9.14217830e-01 3.54358196e-01
-7.36297607e-01 4.08544600e-01 5.81614021e-03 5.34975529e-01
5.39254785e-01 7.02299237e-01 -3.49520653e-01 2.00304702e-01
-7.67386556e-01 2.09625795e-01 1.00689340e+00 9.11103964e-01
5.73711097e-01 5.94960928e-01 -2.55942643e-02 3.24656755e-01
6.16687417e-01 2.24709868e-01 5.79375267e-01 -4.12112236e-01
6.33339524e-01 1.06283450e+00 -4.32214618e-01 -1.39055657e+00
-4.74727422e-01 -1.52191710e+00 -9.60750639e-01 -5.97601496e-02
-7.08469152e-02 -8.61714855e-02 -5.41954637e-01 1.24691081e+00
-3.59845251e-01 4.83682513e-01 7.72521123e-02 6.91392362e-01
8.00055861e-01 5.14171064e-01 -1.71972618e-01 -4.82781641e-02
1.23838949e+00 -1.14377224e+00 -8.41127992e-01 -3.42800230e-01
6.43682659e-01 -2.29255453e-01 3.61588567e-01 7.15297386e-02
-1.10594440e+00 -4.09252435e-01 -1.20838499e+00 9.91569236e-02
-4.46518242e-01 -2.34379068e-01 9.83565211e-01 2.15334296e-01
-1.14983916e+00 6.91820860e-01 -4.25960004e-01 6.14138603e-01
5.70414722e-01 4.55525070e-01 6.19995445e-02 4.85151261e-01
-1.68708241e+00 6.78779244e-01 2.39134729e-01 4.77356702e-01
-5.35651982e-01 -9.24222648e-01 -7.34824955e-01 5.51240027e-01
5.42918444e-01 -8.56056511e-01 8.39125216e-01 -4.87791389e-01
-9.96739745e-01 3.35125417e-01 5.50957799e-01 -8.04940641e-01
2.10196421e-01 -1.73706800e-01 -5.36003709e-01 -2.38316953e-01
-1.23800464e-01 -2.01113839e-02 7.00785100e-01 -6.72046304e-01
-4.09079075e-01 -3.82480681e-01 3.56189430e-01 3.09920069e-02
-4.59101945e-01 -3.44035953e-01 -1.41375646e-01 -1.06101954e+00
1.66104838e-01 -5.52244008e-01 -3.33639890e-01 -6.25119805e-01
-4.80488956e-01 -2.61991322e-01 2.80871034e-01 -8.37006450e-01
1.84725678e+00 -1.85156679e+00 2.92593539e-01 4.28129256e-01
9.19032693e-01 1.10595906e-02 -4.26548608e-02 5.23841918e-01
-2.75375992e-01 3.79538447e-01 1.20489322e-01 -1.12565681e-01
4.97295558e-01 -1.51889890e-01 -5.08409262e-01 1.63423061e-01
3.71391356e-01 1.78528392e+00 -3.99877816e-01 2.99095605e-02
-1.91898167e-01 2.38386169e-01 -5.33190727e-01 4.85914685e-02
-1.58583537e-01 -3.96910250e-01 -5.31878710e-01 6.96375370e-01
5.06318510e-01 -6.20181799e-01 1.17401727e-01 -2.12535888e-01
4.26754653e-02 5.10093987e-01 -9.51772690e-01 7.14651227e-01
5.07632084e-02 3.69608670e-01 -3.96215498e-01 -8.13357592e-01
1.13303149e+00 -1.22991338e-01 2.02568471e-01 -6.77370012e-01
2.36633107e-01 3.73549134e-01 5.78800917e-01 5.62623218e-02
5.21963120e-01 -3.78336102e-01 2.21932307e-02 5.40101767e-01
7.26609258e-03 2.15307996e-01 1.21539272e-01 5.74809313e-01
9.96726811e-01 -3.13063771e-01 3.02463114e-01 -3.81442726e-01
2.83490717e-01 -6.08303308e-01 6.97263122e-01 2.51620263e-01
-3.33892256e-02 8.09851959e-02 1.35290670e+00 -5.65519094e-01
-8.26442420e-01 -8.20201039e-01 2.45722011e-01 7.81928301e-01
-2.84080595e-01 -4.81459290e-01 -6.59720898e-02 -7.40666747e-01
7.96074331e-01 6.82174385e-01 -7.94867992e-01 -2.23891214e-01
-4.10252303e-01 -8.06245148e-01 2.04819366e-01 1.01448631e+00
3.83005679e-01 -9.23402190e-01 -1.11644022e-01 3.36057842e-01
4.85689640e-01 -8.89814496e-01 -5.93292236e-01 2.62612373e-01
-9.87015188e-01 -9.95277464e-01 -7.34167039e-01 -7.03373671e-01
2.68942505e-01 -2.36693863e-02 1.44005334e+00 3.76982391e-01
2.05273241e-01 9.82878581e-02 -2.08746139e-02 -6.13887727e-01
1.35672763e-01 2.34706983e-01 -1.07487947e-01 3.70466709e-01
6.19428217e-01 -8.00631046e-01 -6.55794799e-01 1.67401776e-01
-6.74533606e-01 -2.65045077e-01 8.45605850e-01 5.76693535e-01
2.36261889e-01 3.65747541e-01 7.94332504e-01 -9.93680954e-01
9.01709080e-01 -7.78620183e-01 -9.35061336e-01 1.32628545e-01
-1.16493511e+00 9.61530283e-02 1.80672482e-01 -2.66372055e-01
-5.55609167e-01 -7.76938200e-01 3.91407758e-01 -6.65506721e-01
9.63151813e-01 1.40500402e+00 -9.60961953e-02 1.63658768e-01
-1.15492865e-01 2.06817970e-01 4.36231084e-02 -3.12074810e-01
6.53236285e-02 9.81758386e-02 6.90096617e-02 3.59819978e-02
1.13086462e+00 6.67461827e-02 5.02465248e-01 -3.03896040e-01
-7.20100343e-01 -5.44535890e-02 -6.38244629e-01 -1.72404036e-01
6.76090002e-01 -1.08064663e+00 -6.79304421e-01 6.32747889e-01
-5.28097391e-01 -2.84830868e-01 -1.39865687e-03 6.75639808e-01
-6.50715753e-02 1.17448494e-01 -9.80500102e-01 -9.35705125e-01
-1.89313293e-01 -7.76732385e-01 6.66467428e-01 4.17523421e-02
1.33532956e-01 -1.61732459e+00 -1.19105294e-01 2.09122524e-01
5.04787982e-01 1.40383676e-01 1.16513836e+00 -9.68831182e-01
-1.23611093e+00 -4.05708373e-01 -4.36144441e-01 2.54400611e-01
4.30170111e-02 -1.97471216e-01 -3.96337092e-01 -3.07270326e-03
1.37346555e-02 1.96331784e-01 1.31988132e+00 6.52901471e-01
4.93170977e-01 -5.14469743e-01 3.31909284e-02 6.49810076e-01
1.46558166e+00 1.70818672e-01 6.48032367e-01 4.71939296e-01
1.06427073e+00 8.50829899e-01 1.42485887e-01 3.61309439e-01
7.23943233e-01 2.46876121e-01 4.40162808e-01 1.85884699e-01
2.46256694e-01 -4.51288253e-01 4.51281011e-01 1.35755849e+00
-3.55327368e-01 -2.68331915e-01 -9.60960209e-01 5.44464052e-01
-1.80742502e+00 -8.76371026e-01 -2.47870818e-01 1.69754457e+00
3.17898244e-01 7.79985428e-01 3.37291479e-01 1.44147471e-01
5.60135484e-01 5.98324478e-01 -6.08414292e-01 -5.81826270e-02
-3.74993056e-01 -1.95332825e-01 8.03132057e-01 4.28401917e-01
-8.41398180e-01 6.77913785e-01 5.62675095e+00 4.15357739e-01
-7.37646282e-01 -4.72343951e-01 7.81509697e-01 -7.75373653e-02
-1.09948206e+00 8.70521530e-04 -1.08461714e+00 3.88590485e-01
9.97914851e-01 -4.31407392e-01 2.92570204e-01 5.86775064e-01
-3.17706972e-01 5.95661342e-01 -1.03838027e+00 7.00817823e-01
-2.49257050e-02 -1.53232312e+00 3.65925074e-01 6.22669935e-01
6.87629104e-01 -2.21443281e-01 4.18690771e-01 6.27990127e-01
2.99210668e-01 -1.17547035e+00 7.92808354e-01 9.35140908e-01
1.76363572e-01 -9.23550844e-01 9.86384034e-01 8.63641128e-02
-1.52208698e+00 -4.32121247e-01 -4.21951354e-01 -4.61089253e-01
2.81726062e-01 6.50647283e-01 -2.81286865e-01 7.45670736e-01
4.39015031e-01 1.21790445e+00 -6.81170106e-01 6.70644522e-01
-1.97675899e-01 6.53664947e-01 9.07491073e-02 -3.33408326e-01
5.25547266e-01 -6.23721957e-01 3.28035057e-01 5.56122124e-01
4.29867446e-01 1.95730165e-01 -2.15438768e-01 9.72422004e-01
-4.52281147e-01 -2.48879269e-02 -6.29123628e-01 -8.56857002e-01
1.22776486e-01 9.53334510e-01 -5.97316444e-01 -1.41521439e-01
-7.62271345e-01 2.05907255e-01 4.09756571e-01 3.99223328e-01
-6.63906932e-01 -3.52784634e-01 6.14373684e-01 5.53924382e-01
8.39001060e-01 -2.23334640e-01 -3.75055254e-01 -1.26680422e+00
8.71866122e-02 -7.49611378e-01 5.63565195e-01 -6.11168325e-01
-1.65536582e+00 5.63989401e-01 -3.91155332e-01 -1.08740044e+00
-9.84947458e-02 -8.46568644e-01 -6.42890155e-01 1.14122272e+00
-1.88237524e+00 -1.06401217e+00 2.26220965e-01 2.16858506e-01
2.40111016e-02 -7.75943100e-01 1.49651200e-01 2.12984785e-01
-6.89162731e-01 5.10400176e-01 -1.64837196e-01 6.07017398e-01
-2.22867392e-02 -1.46863341e+00 7.83163071e-01 8.22292089e-01
3.82678360e-01 7.87392437e-01 2.41368100e-01 -1.24644566e+00
-1.40716672e+00 -1.03656483e+00 1.02356017e+00 -4.80285585e-01
1.51478589e+00 -3.61941367e-01 -1.04321575e+00 1.19012845e+00
2.66259640e-01 -1.26719117e-01 6.26073778e-01 3.24653685e-01
-5.73278308e-01 -3.09568673e-01 -3.15815389e-01 4.65086162e-01
1.08340621e+00 -5.42220294e-01 -6.66422367e-01 8.63484368e-02
1.16768873e+00 5.23136631e-02 -1.31211829e+00 3.47542763e-01
3.39963824e-01 -1.02032411e+00 9.78269279e-01 -7.32999206e-01
7.29549527e-01 1.57821521e-01 -6.38767779e-02 -1.23969114e+00
-1.00111008e+00 -6.73973858e-01 -6.29185855e-01 1.12091231e+00
7.90742576e-01 -1.26781178e+00 7.81150937e-01 4.40837890e-01
-3.27812023e-02 -1.06688130e+00 -5.92195809e-01 -7.75114477e-01
1.30075499e-01 -1.77130684e-01 1.26831520e+00 1.03694725e+00
-7.51207918e-02 6.11411512e-01 -3.50517958e-01 7.30251819e-02
8.12500536e-01 5.46991646e-01 4.36109036e-01 -1.57819378e+00
-7.48553157e-01 -1.05793381e+00 -6.83445036e-01 -9.35209870e-01
2.33091369e-01 -1.03773618e+00 -7.65748024e-01 -1.62390769e+00
1.35943949e-01 -1.78710744e-01 -9.34582829e-01 1.80700183e-01
-3.87416124e-01 -5.58519885e-02 2.15194046e-01 3.58993888e-01
-1.95470318e-01 7.09960878e-01 1.41441596e+00 -1.60220221e-01
-3.32298763e-02 2.02201501e-01 -1.25543320e+00 4.57855910e-01
6.55782402e-01 -1.05177872e-01 -4.76472259e-01 -2.00873017e-01
1.08350539e+00 1.58100814e-01 3.35438371e-01 -5.67108095e-01
1.13584340e-01 3.49247828e-02 2.74474055e-01 -7.11570799e-01
2.47310176e-02 -6.74985707e-01 1.53789595e-01 5.64474702e-01
-2.81451792e-01 6.10595584e-01 -9.66670439e-02 9.04067278e-01
-4.77926403e-01 -1.60660967e-02 -2.94020325e-02 2.44799219e-02
-2.93928921e-01 7.04967856e-01 3.01756915e-02 -6.14584126e-02
7.93857038e-01 2.62814127e-02 -5.57016313e-01 -7.57617235e-01
-5.05473256e-01 4.60599124e-01 4.36630566e-03 4.37079340e-01
6.43932879e-01 -1.61004722e+00 -8.20769966e-01 3.06751430e-01
-1.57931253e-01 -1.08385585e-01 4.07479852e-02 9.50180411e-01
-5.65118968e-01 7.28123009e-01 -8.70486051e-02 1.23800393e-02
-6.00946903e-01 8.17361534e-01 2.78606206e-01 -8.01116943e-01
-5.35003841e-01 1.01868916e+00 5.79130411e-01 -4.62000705e-02
4.36975025e-02 -6.91740632e-01 -7.50967741e-01 5.94359815e-01
2.96693772e-01 3.90359372e-01 -1.47657692e-01 -6.81958973e-01
-9.31218714e-02 6.08968139e-01 -4.81965020e-02 1.52405659e-02
1.82105446e+00 1.35619760e-01 -5.47855139e-01 7.08591878e-01
9.86597717e-01 1.25584140e-01 -9.43827748e-01 -4.40243691e-01
2.62297511e-01 -2.85379499e-01 3.29984367e-01 -4.14170712e-01
-1.72975910e+00 8.35194767e-01 -2.92699605e-01 7.92114139e-01
1.02359295e+00 -1.44244522e-01 8.87417376e-01 2.26806656e-01
8.46312344e-02 -7.24830687e-01 1.95056736e-01 3.94534349e-01
1.10716414e+00 -7.29069054e-01 3.06663424e-01 -3.08028430e-01
-7.65843511e-01 1.04398286e+00 5.87603211e-01 -4.68712687e-01
1.29116940e+00 1.44319803e-01 1.29184993e-02 -7.11596608e-01
-9.17192221e-01 -9.31357965e-02 8.14062774e-01 -9.23067983e-03
6.75430670e-02 2.06693336e-02 -5.68678454e-02 1.20548773e+00
-5.35364449e-01 -4.41548228e-01 6.41890228e-01 5.87338507e-01
-2.25487858e-01 -7.78276801e-01 9.83064175e-02 1.06602931e+00
-6.10669911e-01 -4.26895857e-01 -5.11357248e-01 8.78581822e-01
-4.55954641e-01 2.65134484e-01 3.48174334e-01 -7.20534742e-01
3.85930091e-01 1.11137733e-01 -2.26803441e-02 -4.76826787e-01
-8.94887686e-01 4.36440468e-01 3.17449979e-02 -3.21983546e-01
-2.18019694e-01 -7.70099938e-01 -9.58168209e-01 -5.61865866e-01
-3.72278094e-01 2.11099580e-01 1.41469702e-01 6.00468516e-01
3.35066497e-01 1.10073423e+00 6.19557977e-01 -4.13268298e-01
-6.68603420e-01 -1.06546676e+00 -1.36574984e+00 1.39369160e-01
4.36443239e-01 -7.46016085e-01 -7.72264659e-01 -4.81309146e-01] | [4.3458662033081055, 4.324032306671143] |
4128e61e-75cf-4fe6-b140-d09529cbb740 | a-data-dependent-multiscale-model-for | 1808.01047 | null | https://arxiv.org/abs/1808.01047v4 | https://arxiv.org/pdf/1808.01047v4.pdf | A Data Dependent Multiscale Model for Hyperspectral Unmixing With Spectral Variability | Spectral variability in hyperspectral images can result from factors including environmental, illumination, atmospheric and temporal changes. Its occurrence may lead to the propagation of significant estimation errors in the unmixing process. To address this issue, extended linear mixing models have been proposed which lead to large scale nonsmooth ill-posed inverse problems. Furthermore, the regularization strategies used to obtain meaningful results have introduced interdependencies among abundance solutions that further increase the complexity of the resulting optimization problem. In this paper we present a novel data dependent multiscale model for hyperspectral unmixing accounting for spectral variability. The new method incorporates spatial contextual information to the abundances in extended linear mixing models by using a multiscale transform based on superpixels. The proposed method results in a fast algorithm that solves the abundance estimation problem only once in each scale during each iteration. Simulation results using synthetic and real images compare the performances, both in accuracy and execution time, of the proposed algorithm and other state-of-the-art solutions. | ['José Carlos Moreira Bermudez', 'Tales Imbiriba', 'Ricardo Augusto Borsoi'] | 2018-08-02 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 7.21179724e-01 -8.03539634e-01 5.52804708e-01 -1.12237133e-01
-4.43357527e-01 -5.45468032e-01 4.24212813e-01 4.32420410e-02
-3.94494563e-01 9.15914893e-01 -5.85175604e-02 1.49693445e-03
-4.71867323e-01 -7.42094576e-01 -3.89469326e-01 -1.11535418e+00
2.46265605e-01 3.70131999e-01 1.89416949e-02 -2.52756663e-02
1.82764843e-01 5.54738164e-01 -1.56504357e+00 -2.54186131e-02
1.39377427e+00 7.58164287e-01 5.47336876e-01 6.17417991e-01
-1.29579768e-01 5.25945902e-01 -3.19926888e-01 4.11486983e-01
5.59417427e-01 -5.48575103e-01 -1.43219799e-01 7.24165440e-01
4.71459210e-01 -8.46716687e-02 3.10821235e-01 1.49011075e+00
3.31766874e-01 4.51411873e-01 7.91872501e-01 -9.94716346e-01
-9.47602838e-02 3.04506242e-01 -1.23939145e+00 1.39785230e-01
-3.65088731e-01 2.83601265e-02 6.12182975e-01 -9.85192835e-01
2.46686190e-01 1.03364837e+00 7.34925985e-01 -3.97553980e-01
-1.51685286e+00 -5.77981472e-01 -2.19554767e-01 1.62589714e-01
-1.62028623e+00 -2.77316272e-01 7.17497528e-01 -6.32802665e-01
6.31198585e-01 4.69737321e-01 6.74352646e-01 1.32789478e-01
-4.98276055e-02 -6.32848516e-02 1.67052305e+00 -5.83671093e-01
1.39676839e-01 -6.01617359e-02 1.78894356e-01 4.47823912e-01
7.92932034e-01 1.18291676e-02 -1.60707638e-01 -2.66639709e-01
5.91838300e-01 -6.75313473e-02 -3.69055122e-01 -2.34026775e-01
-1.00668931e+00 6.59366906e-01 3.69783342e-01 3.92610073e-01
-8.28943431e-01 -1.53870851e-01 -1.78845391e-01 7.53364563e-02
6.98364556e-01 2.57180959e-01 -1.88486621e-01 5.05221605e-01
-1.30653632e+00 1.30039945e-01 3.37812036e-01 3.23482811e-01
1.15255344e+00 1.69399396e-01 1.25103667e-01 8.65690172e-01
5.81468225e-01 9.88533974e-01 2.32529402e-01 -8.50687027e-01
3.55587512e-01 5.90869665e-01 5.02268910e-01 -1.26727569e+00
-5.50090551e-01 -8.24588537e-01 -1.03778243e+00 4.26013201e-01
3.94898832e-01 -1.40207276e-01 -8.84931505e-01 1.43816173e+00
5.86472809e-01 4.54010010e-01 2.10432768e-01 9.47334111e-01
1.66459143e-01 1.18426192e+00 1.28823027e-01 -7.01779544e-01
1.10827291e+00 -8.78128231e-01 -9.97555017e-01 -3.22518736e-01
1.68434069e-01 -1.29317117e+00 5.52615047e-01 4.15057689e-01
-7.99023628e-01 -4.71369147e-01 -1.24197161e+00 3.62614363e-01
-3.80566120e-01 4.23597246e-01 1.16241626e-01 5.73071301e-01
-7.80572116e-01 5.89826167e-01 -8.21047843e-01 -5.17028093e-01
-1.23981975e-01 2.79938191e-01 1.96436215e-02 8.20140392e-02
-7.51747966e-01 1.00065291e+00 4.04056102e-01 5.90177178e-01
-4.16648060e-01 -4.22733366e-01 -3.98800671e-01 -6.61549866e-02
2.21289247e-01 -5.57705462e-01 6.43001437e-01 -1.39231169e+00
-1.23250175e+00 4.36815947e-01 -3.10039282e-01 -1.77761972e-01
4.95328903e-01 -4.64292169e-02 -5.05651295e-01 1.80954129e-01
2.61043996e-01 1.47133842e-01 9.80361938e-01 -1.34713137e+00
-5.78999639e-01 -5.43724120e-01 -5.34937859e-01 4.60015535e-01
-2.37607524e-01 -1.55601382e-01 5.79041913e-02 -7.24852264e-01
6.84016228e-01 -1.13272703e+00 -4.14403945e-01 -9.98857021e-02
-4.22363281e-02 6.48317099e-01 7.95016885e-01 -1.02363515e+00
1.00886929e+00 -2.08976507e+00 3.54867131e-01 3.74316275e-01
-2.24319145e-01 1.91612825e-01 -3.36506441e-02 3.20981055e-01
-1.27224460e-01 -1.94041997e-01 -9.13944185e-01 -1.25005081e-01
-3.69678468e-01 1.37977749e-01 1.58875763e-01 8.67238700e-01
-7.51968175e-02 1.16554871e-01 -5.46934366e-01 -2.14780450e-01
5.70442021e-01 6.75846875e-01 -1.21705979e-01 1.16262965e-01
-1.96060494e-01 7.20696747e-01 -1.31415054e-01 4.04307842e-01
1.26369381e+00 -1.17139675e-01 1.25952631e-01 -5.93902349e-01
-6.51211619e-01 -4.03349400e-01 -1.78139472e+00 1.39577711e+00
-4.36976671e-01 3.82127643e-01 7.87892103e-01 -1.11169648e+00
6.33793056e-01 3.33508015e-01 4.71989095e-01 6.10733731e-03
6.92047030e-02 6.05481088e-01 3.47609892e-02 -4.46094006e-01
5.33951104e-01 -6.37242317e-01 7.39995897e-01 2.53058523e-01
-3.35212827e-01 -3.68113995e-01 2.96454132e-01 -2.55470097e-01
2.08865508e-01 1.17116623e-01 4.41910207e-01 -6.68172121e-01
9.78473127e-01 2.94312358e-01 4.16783690e-01 5.28030694e-01
2.54421812e-02 3.62227499e-01 -8.80900025e-02 -7.85206631e-02
-1.00943995e+00 -6.57311738e-01 -2.47550368e-01 5.10190845e-01
2.30496779e-01 3.80627722e-01 -6.19681954e-01 2.64635026e-01
-2.14903355e-01 5.23395717e-01 -3.09856862e-01 3.22967559e-01
-2.36701831e-01 -1.71769834e+00 3.18787247e-02 -1.31227940e-01
8.96853507e-01 -5.28004885e-01 -4.85995412e-01 5.36243916e-01
-4.61049080e-01 -1.08402836e+00 -1.33565679e-01 8.95181578e-03
-1.36333430e+00 -9.40650344e-01 -7.00056970e-01 -3.59215051e-01
7.93915331e-01 5.18354475e-01 6.48123801e-01 -2.74003446e-01
-4.48706537e-01 5.26274070e-02 -3.24166685e-01 -2.88077086e-01
-5.15825570e-01 -2.44478017e-01 -6.02816567e-02 6.38958454e-01
4.51635420e-02 -6.08702183e-01 -5.50822437e-01 3.02655548e-01
-1.30849099e+00 9.88709182e-02 5.88757575e-01 6.54451430e-01
5.89383364e-01 5.60825527e-01 2.29984865e-01 -6.90246105e-01
4.14788932e-01 -2.96445370e-01 -1.23829591e+00 2.98267871e-01
-6.98806703e-01 8.55942667e-02 2.54200757e-01 -2.49166176e-01
-1.37505460e+00 3.91190410e-01 4.02062416e-01 -5.99973053e-02
-1.41016826e-01 6.96819663e-01 -9.53106880e-02 -4.82028067e-01
6.75011039e-01 4.28379536e-01 1.79175183e-01 -5.67283273e-01
2.78310031e-01 7.06340671e-01 3.24512631e-01 -2.52405316e-01
1.09137833e+00 8.04301023e-01 6.04721546e-01 -1.39520049e+00
-5.49167693e-01 -7.69197404e-01 -6.30099177e-01 -2.39192441e-01
9.28806961e-01 -1.04232860e+00 -5.18345051e-02 9.06503499e-01
-9.85646486e-01 7.49757662e-02 2.42463782e-01 6.40523195e-01
-1.05040587e-01 7.47177303e-01 -4.67394769e-01 -1.11616015e+00
-3.54298651e-01 -1.20272100e+00 4.72738415e-01 2.24615932e-01
1.35603502e-01 -1.02960515e+00 7.16865482e-03 4.94898647e-01
7.22007692e-01 4.95547652e-01 8.04279447e-01 1.57511383e-01
-6.89408183e-01 1.15033522e-01 -3.74570370e-01 4.74836618e-01
4.67320740e-01 2.07735449e-01 -8.58016729e-01 -2.23561198e-01
4.56355095e-01 2.07475096e-01 7.87957966e-01 7.97228158e-01
3.60555053e-01 -2.31314689e-01 -4.05527353e-02 5.94338655e-01
2.14980531e+00 1.52094647e-01 3.22124064e-01 4.37525302e-01
6.05835199e-01 7.10976124e-01 6.85366511e-01 5.95096707e-01
-3.17386061e-01 4.14251715e-01 6.03593290e-01 -4.64981437e-01
9.42502990e-02 4.62749302e-01 5.29115126e-02 7.06656218e-01
-3.15462708e-01 -1.51898295e-01 -7.13854313e-01 5.34315169e-01
-1.88372099e+00 -9.89960790e-01 -9.29169059e-01 2.39310789e+00
5.21159828e-01 -5.58201373e-01 -3.25972736e-01 3.30811620e-01
9.24882412e-01 3.84843618e-01 -3.04915279e-01 1.17411323e-01
-5.07108212e-01 1.32552192e-01 1.03952277e+00 1.06177771e+00
-1.20304894e+00 6.22045279e-01 5.63977957e+00 5.10027885e-01
-1.29241824e+00 3.44130903e-01 2.72724092e-01 8.02413002e-02
-1.79139171e-02 2.44375002e-02 -3.22184414e-01 1.46927938e-01
6.77012682e-01 -8.80690441e-02 5.92801511e-01 7.80569613e-02
7.63617098e-01 -5.95989645e-01 -1.09723248e-01 9.09227550e-01
1.98087506e-02 -8.07889402e-01 4.01231088e-02 1.52881201e-02
1.25165534e+00 3.85871567e-02 -3.72596644e-02 -6.20877683e-01
-4.34772149e-02 -6.33476138e-01 4.59723324e-01 4.86592203e-01
2.13628992e-01 -5.93198299e-01 6.03312552e-01 3.50360960e-01
-1.25380433e+00 -5.23345135e-02 -4.40815389e-01 -2.69361377e-01
2.35088974e-01 1.04902685e+00 -5.09544730e-01 6.73486590e-01
2.45910332e-01 5.43360651e-01 -4.09645647e-01 1.14901721e+00
-1.17797948e-01 4.88602787e-01 -6.88882351e-01 2.85463065e-01
3.82966787e-01 -1.12093472e+00 7.46514380e-01 9.90382433e-01
8.98647130e-01 3.09960485e-01 -6.35567009e-02 9.48425293e-01
4.78860140e-01 5.65653145e-01 -3.03328365e-01 -1.83528587e-01
-1.32955154e-02 1.35649562e+00 -8.49622965e-01 -4.46199566e-01
-4.53980476e-01 9.92607594e-01 -3.97518128e-01 6.45282745e-01
-6.98492706e-01 1.09485619e-01 6.11683309e-01 2.73713052e-01
2.43659258e-01 -4.11797106e-01 -3.04844141e-01 -1.12792218e+00
-1.03979215e-01 -8.70865881e-01 3.57566327e-01 -7.41095603e-01
-9.82297838e-01 5.05361080e-01 8.02292582e-03 -1.23711360e+00
2.81972021e-01 -6.87171340e-01 -2.38630474e-01 1.24324918e+00
-1.68705630e+00 -9.96582806e-01 -6.05582535e-01 5.03396630e-01
4.63144660e-01 6.57559708e-02 7.07105696e-01 3.53033543e-01
-6.08497560e-01 -5.48210204e-01 7.13365078e-01 -6.18421674e-01
6.21251702e-01 -1.03306782e+00 -4.22780573e-01 1.37283337e+00
-1.38853461e-01 2.45706484e-01 1.16990864e+00 -7.00599968e-01
-9.59739506e-01 -1.07397556e+00 5.98917544e-01 3.97476822e-01
8.27630579e-01 3.18247646e-01 -8.10842991e-01 2.78382570e-01
2.86542863e-01 -2.89086178e-02 7.13348985e-01 -5.38083434e-01
7.96310529e-02 -3.78059387e-01 -1.26244485e+00 2.57152259e-01
3.30568582e-01 -2.45173529e-01 -2.60429561e-01 4.59531635e-01
1.26898974e-01 -1.93962324e-02 -5.85764289e-01 6.36642337e-01
3.81312996e-01 -1.08129621e+00 8.86730433e-01 1.33495837e-01
1.21302396e-01 -9.31244493e-01 -2.06007436e-01 -1.30843532e+00
-4.82461721e-01 -3.71655524e-01 3.89462531e-01 9.58824456e-01
6.36255741e-01 -8.78761947e-01 3.47858787e-01 4.30457413e-01
2.72637218e-01 1.57882825e-01 -6.09713078e-01 -6.03389204e-01
-2.85366565e-01 6.54224679e-02 2.65232950e-01 1.04962575e+00
-5.12213409e-01 5.31778000e-02 -4.54054326e-01 9.96520221e-01
1.18968904e+00 3.33354920e-01 3.63235742e-01 -1.43313611e+00
-4.23676431e-01 -2.00951844e-01 -4.01481539e-02 -4.96662647e-01
-1.52717978e-01 -4.77525711e-01 1.19610392e-01 -1.58752608e+00
2.00614199e-01 -3.04854333e-01 -3.28715414e-01 9.45426524e-02
-4.51167107e-01 4.25722808e-01 6.16216697e-02 3.94981891e-01
3.07116508e-01 2.63433188e-01 1.00846446e+00 -3.39483827e-01
-3.74850243e-01 -1.86870500e-01 -1.06395222e-01 7.00234950e-01
9.51877475e-01 -5.85107565e-01 -5.35650194e-01 -5.37551403e-01
1.31390020e-01 3.65156941e-02 1.05625443e-01 -1.39384067e+00
4.10368592e-02 -3.53928238e-01 2.23762661e-01 -4.68773335e-01
5.20867467e-01 -1.22262025e+00 8.83149266e-01 5.76418221e-01
6.94425479e-02 -1.04502559e-01 3.73794168e-01 5.33290207e-01
-3.69917452e-01 -5.89072287e-01 1.17725432e+00 -1.92967042e-01
-4.81997609e-01 6.29607439e-02 -5.51732659e-01 -7.06475794e-01
8.75962913e-01 -1.15985699e-01 -3.80242504e-02 -3.31904233e-01
-9.02978659e-01 -2.81059563e-01 5.43244720e-01 -2.23432034e-01
3.54734100e-02 -8.78454745e-01 -8.20372641e-01 -3.01283784e-02
-1.76986739e-01 -2.62378871e-01 3.19105208e-01 1.14890718e+00
-1.12986660e+00 1.71679333e-01 -2.50744313e-01 -4.44357038e-01
-1.54619682e+00 3.99958163e-01 6.38879538e-01 -2.73989439e-01
1.77814234e-02 6.11846447e-01 9.21914577e-02 -2.78612345e-01
-2.86883861e-01 -3.57733637e-01 -1.81038246e-01 2.37543002e-01
3.44246835e-01 6.57549977e-01 -3.31332944e-02 -9.67684388e-01
-8.92748237e-02 1.02442658e+00 7.15182960e-01 -5.37645102e-01
1.21774232e+00 -4.18006480e-01 -8.50234747e-01 5.20759284e-01
1.00329947e+00 1.52682975e-01 -1.02336192e+00 -2.24988386e-01
-1.15542352e-01 -4.95752066e-01 5.44214368e-01 -8.16745460e-01
-9.84478652e-01 8.31045926e-01 9.88910854e-01 1.73925951e-01
1.48238266e+00 -8.80953074e-01 2.48412564e-01 1.79781467e-01
5.14153484e-03 -1.06076419e+00 -4.69594598e-01 9.65296552e-02
7.74761438e-01 -1.33240545e+00 4.68911380e-01 -8.80621314e-01
-2.01813281e-01 1.09978974e+00 2.82519609e-01 1.67084813e-01
6.52350187e-01 3.00662238e-02 5.11706471e-01 1.50467530e-01
-4.30387147e-02 -3.97627801e-01 3.82344127e-02 2.12940037e-01
4.83890474e-01 2.59705722e-01 -7.61471570e-01 -1.40491828e-01
4.03698385e-01 4.83404286e-02 5.60370207e-01 7.27307498e-01
-7.75108278e-01 -1.09631145e+00 -1.20828307e+00 1.76921368e-01
-3.85304660e-01 -2.23270357e-01 -3.90714072e-02 3.58767718e-01
4.02918249e-01 1.24591935e+00 -2.16154546e-01 2.31603712e-01
-3.32457721e-02 2.44976789e-01 3.31166267e-01 -2.37538859e-01
-2.39022419e-01 6.51517689e-01 -9.18604657e-02 -3.20923738e-02
-1.04462564e+00 -6.92687452e-01 -9.71905708e-01 8.55262429e-02
-6.41621351e-01 4.23799455e-01 1.02876341e+00 7.27665782e-01
5.58890812e-02 3.33539754e-01 5.21839321e-01 -9.66177464e-01
-5.52705765e-01 -1.40275240e+00 -1.05632591e+00 2.91464508e-01
3.74538630e-01 -4.72445458e-01 -7.80322731e-01 2.25693509e-01] | [10.078707695007324, -2.089751720428467] |
ad17a78c-9733-452f-bf25-3eb249a0b868 | silveralign-mt-based-silver-data-algorithm | 2210.06207 | null | https://arxiv.org/abs/2210.06207v2 | https://arxiv.org/pdf/2210.06207v2.pdf | SilverAlign: MT-Based Silver Data Algorithm For Evaluating Word Alignment | Word alignments are essential for a variety of NLP tasks. Therefore, choosing the best approaches for their creation is crucial. However, the scarce availability of gold evaluation data makes the choice difficult. We propose SilverAlign, a new method to automatically create silver data for the evaluation of word aligners by exploiting machine translation and minimal pairs. We show that performance on our silver data correlates well with gold benchmarks for 9 language pairs, making our approach a valid resource for evaluation of different domains and languages when gold data are not available. This addresses the important scenario of missing gold data alignments for low-resource languages. | ['Hinrich Schütze', 'Silvia Severini', 'Abdullatif Köksal'] | 2022-10-12 | null | null | null | null | ['word-alignment'] | ['natural-language-processing'] | [ 2.71911155e-02 -2.69070119e-01 -3.79194528e-01 -2.89133161e-01
-1.31335044e+00 -9.88614440e-01 6.46177709e-01 2.18806982e-01
-9.41263974e-01 1.04150724e+00 2.09995866e-01 -5.81056893e-01
3.45827699e-01 -5.31737149e-01 -3.86241049e-01 -4.35574710e-01
4.33867961e-01 1.15421975e+00 2.01785803e-01 -6.46137536e-01
1.82397574e-01 1.90913901e-01 -1.09440017e+00 3.98502231e-01
1.29318464e+00 1.19267620e-01 4.96095806e-01 3.76406699e-01
-4.23914313e-01 -3.70673575e-02 -4.86175478e-01 -8.89033735e-01
3.56612086e-01 -4.31535989e-01 -1.12777364e+00 -6.49846673e-01
5.54075003e-01 2.09150463e-01 4.71508771e-01 1.07459092e+00
7.22745538e-01 -1.67547181e-01 4.55102324e-01 -1.02924967e+00
-4.22524512e-01 8.53997290e-01 -1.42539605e-01 4.51136291e-01
5.58557034e-01 2.66138494e-01 1.52645051e+00 -1.10068846e+00
9.71730471e-01 8.66773486e-01 7.12148726e-01 4.93308961e-01
-1.17647803e+00 -4.64915723e-01 -2.53712118e-01 2.79099137e-01
-1.39106953e+00 -4.64101344e-01 1.43324524e-01 -3.54436666e-01
1.63411939e+00 3.06371510e-01 3.46934140e-01 1.30029857e+00
1.56467319e-01 2.03876227e-01 1.27696741e+00 -1.06323469e+00
-3.51326704e-01 9.12383758e-03 2.75475308e-02 3.82297277e-01
4.40364718e-01 -1.94523260e-01 -5.83031952e-01 -2.90467709e-01
1.12796485e-01 -6.00429654e-01 -9.45468433e-03 7.29460195e-02
-1.45953095e+00 6.44816637e-01 -1.26757130e-01 8.13649535e-01
-3.09399337e-01 -4.21274036e-01 2.88231581e-01 7.58098781e-01
6.03965044e-01 1.09529293e+00 -6.01462781e-01 -4.08225447e-01
-9.30895567e-01 3.85174781e-01 9.46496665e-01 8.03436041e-01
6.92756414e-01 -2.43994728e-01 6.49181753e-02 9.60261762e-01
8.14397112e-02 8.78582954e-01 6.15480781e-01 -4.41965371e-01
1.02516723e+00 5.14658630e-01 3.70496094e-01 -7.29593217e-01
-3.50347251e-01 -1.88543037e-01 -3.91430020e-01 -1.12622254e-01
7.46998191e-01 -6.59641325e-02 -5.29061317e-01 1.78332579e+00
1.59202576e-01 -3.29373419e-01 2.54254073e-01 7.31242776e-01
7.29330659e-01 4.39476281e-01 5.63058965e-02 -1.76146433e-01
1.38351715e+00 -1.00559723e+00 -6.15683794e-01 -5.61064065e-01
9.74332571e-01 -1.49603021e+00 1.55412722e+00 1.67954579e-01
-1.04667091e+00 -3.92646492e-01 -9.26747262e-01 -1.16200984e-01
-4.25337732e-01 -1.27993882e-01 2.53252715e-01 4.13746744e-01
-1.08955288e+00 6.08256459e-01 -6.66354418e-01 -6.80689633e-01
-2.70922691e-01 3.31128329e-01 -7.09152102e-01 -2.02001542e-01
-1.43154395e+00 1.65400422e+00 3.38998139e-01 -1.64807767e-01
-2.49689639e-01 -3.12035501e-01 -5.35543144e-01 -3.97177845e-01
-4.50927466e-02 -5.79049170e-01 1.26062822e+00 -7.51767993e-01
-1.12448525e+00 1.46518040e+00 -5.33663295e-02 -3.83808374e-01
5.51626503e-01 -4.09113854e-01 -6.01780772e-01 -4.93041873e-01
3.18977594e-01 3.91143560e-01 2.28730470e-01 -5.39664507e-01
-5.89829564e-01 -1.63332134e-01 -2.05138072e-01 3.02998960e-01
-1.55691147e-01 6.92467332e-01 -2.67363071e-01 -5.18517375e-01
-4.40458596e-01 -9.65454876e-01 -5.03083348e-01 -8.89142156e-01
-2.43013069e-01 -3.49267870e-01 5.46288863e-03 -1.07548177e+00
1.36799049e+00 -1.85303175e+00 1.84107378e-01 2.03627169e-01
-2.75292784e-01 5.67223966e-01 -7.13792086e-01 9.44028080e-01
9.19069722e-02 2.72330612e-01 -1.90294877e-01 -3.99043918e-01
-6.88913837e-02 3.51701170e-01 -2.31642529e-01 1.64094880e-01
2.49833941e-01 8.45461667e-01 -1.05706012e+00 -7.24445641e-01
-4.65208180e-02 2.27728784e-01 -3.88125062e-01 4.32078391e-01
-2.57628590e-01 5.99307477e-01 -3.28422248e-01 3.96933913e-01
3.67178738e-01 9.22525022e-03 4.51716930e-01 7.83110261e-02
-4.63449746e-01 1.07452619e+00 -9.76380289e-01 1.79180944e+00
-6.81529045e-01 4.86648768e-01 -2.80239046e-01 -4.96742129e-01
1.10434639e+00 3.46256316e-01 1.70157403e-01 -8.30841243e-01
2.03674454e-02 9.27947283e-01 4.62653428e-01 -3.98585528e-01
7.50731289e-01 -6.85236081e-02 -1.77388847e-01 7.03354180e-01
3.49034995e-01 -4.35044020e-01 7.20342815e-01 -2.01173455e-01
1.27944982e+00 1.65998459e-01 6.95617139e-01 -6.04307950e-01
4.19421971e-01 4.48037207e-01 5.35210729e-01 4.52635229e-01
6.97278380e-02 5.58098733e-01 -6.13175295e-02 -6.58579588e-01
-1.42762828e+00 -8.77561212e-01 -5.46542667e-02 1.14182794e+00
-1.76183105e-01 -6.71113014e-01 -8.14308047e-01 -6.38000131e-01
-4.61234748e-01 6.49289250e-01 -1.05237901e-01 2.61747897e-01
-1.09756780e+00 -8.80091190e-01 7.42306054e-01 2.59757433e-02
3.78135554e-02 -1.29259610e+00 -2.63553232e-01 4.72181946e-01
-9.15741205e-01 -1.51870215e+00 -4.53696966e-01 1.34860277e-01
-6.02783561e-01 -1.20535624e+00 -5.42640924e-01 -9.87218618e-01
4.08334315e-01 2.07842156e-01 1.86543298e+00 1.50584221e-01
-6.01382852e-02 -2.02530891e-01 -5.74745238e-01 -3.38392287e-01
-9.28110600e-01 8.99635792e-01 1.44908816e-01 -5.42681575e-01
8.81403685e-01 -5.79763591e-01 -2.53248125e-01 4.06393558e-01
-5.34166574e-01 4.96152751e-02 6.20685935e-01 7.54183531e-01
5.74436009e-01 -8.32527876e-01 3.10009927e-01 -9.35781181e-01
9.61405993e-01 -2.93074936e-01 -5.95145047e-01 5.68056941e-01
-6.96354687e-01 3.86786550e-01 7.18551397e-01 -1.75733015e-01
-6.72706604e-01 -2.74644941e-01 -5.73989630e-01 4.74185079e-01
-2.00431850e-02 4.46314961e-01 -1.34802729e-01 -4.23006155e-02
1.07230067e+00 -2.71968186e-01 -2.81240493e-01 -8.78371716e-01
2.59108156e-01 8.00410151e-01 3.88327360e-01 -7.00999677e-01
6.68695688e-01 -1.56636760e-01 -3.56323719e-01 -5.98401368e-01
-7.72751391e-01 -7.13696063e-01 -8.61126900e-01 3.09292704e-01
8.46891165e-01 -7.85257757e-01 2.09527880e-01 1.24019347e-01
-1.56873584e+00 -3.07286948e-01 -1.61190018e-01 6.14181459e-01
-3.44446123e-01 3.52749437e-01 -3.41681600e-01 -3.97457182e-02
-9.36864316e-01 -1.30932796e+00 9.66151118e-01 -2.95457691e-01
-1.00374460e+00 -9.61413980e-01 1.01135921e+00 4.03446585e-01
3.83458644e-01 -3.96686643e-02 8.22440088e-01 -1.12655711e+00
-1.82578728e-01 -2.02578351e-01 1.09617874e-01 3.49938929e-01
3.03968459e-01 1.29705593e-01 -6.08841181e-01 -9.62036923e-02
-4.55252290e-01 -3.42564642e-01 3.43788087e-01 -1.02164209e-01
3.37932616e-01 -3.01120311e-01 -1.89783141e-01 3.95964414e-01
1.32016826e+00 -1.81986228e-01 5.65381527e-01 7.84378529e-01
5.56085825e-01 6.37711585e-01 8.33672166e-01 3.55178677e-02
4.25907344e-01 1.24777925e+00 7.38055483e-02 -1.89168990e-01
-5.20453677e-02 1.89978685e-02 4.62048888e-01 1.55087841e+00
-4.97697175e-01 -3.81844372e-01 -1.48435175e+00 7.42351353e-01
-1.57514739e+00 -7.49063551e-01 -4.08483297e-01 2.34028411e+00
1.37728572e+00 -5.36757261e-02 6.96367547e-02 -2.73042083e-01
5.38094342e-01 9.32901576e-02 2.52808809e-01 -8.16927671e-01
-4.54678655e-01 8.05154681e-01 4.22938645e-01 7.04131186e-01
-8.70104790e-01 1.41339481e+00 6.96426439e+00 8.38913798e-01
-9.63906348e-01 4.31111693e-01 -4.92796674e-03 2.15171099e-01
-5.98967075e-01 1.17558129e-01 -1.07745600e+00 5.84092140e-01
1.12198615e+00 -3.04844916e-01 2.70051688e-01 5.87563932e-01
4.50780392e-02 1.58270419e-01 -1.09192348e+00 8.40696990e-01
-7.18128607e-02 -1.04014277e+00 -1.98922195e-02 -1.59448832e-01
8.88737679e-01 8.10762525e-01 -2.98090726e-01 -2.52351258e-02
1.02492237e+00 -1.07777154e+00 3.89606535e-01 3.32270097e-03
8.98459613e-01 -4.84845638e-01 1.02482188e+00 2.68662244e-01
-7.93121696e-01 8.00353587e-01 -3.75772715e-01 2.02522539e-02
4.37051415e-01 5.70603967e-01 -1.20086145e+00 5.89987934e-01
4.34619963e-01 3.55960995e-01 -5.52026093e-01 9.55323398e-01
-4.62951094e-01 5.61532915e-01 -3.85194033e-01 -1.01801828e-01
3.41722071e-01 -4.17445034e-01 6.47105396e-01 1.64032292e+00
6.61352813e-01 -5.05029202e-01 2.99078047e-01 2.22965196e-01
-7.96748474e-02 9.29492593e-01 -8.85539591e-01 -1.36786267e-01
5.96092105e-01 1.24961567e+00 -4.44049150e-01 -4.11110930e-02
-5.01336873e-01 9.71427023e-01 7.67708719e-01 -1.13605291e-01
-5.18074989e-01 -1.89682290e-01 8.51970077e-01 -2.24715238e-03
-2.42928475e-01 -5.71956217e-01 -9.26611796e-02 -1.17116153e+00
2.47353822e-01 -1.67495382e+00 6.46074116e-01 -4.04656947e-01
-1.26222479e+00 1.25895870e+00 -1.01185814e-01 -1.38906550e+00
-7.84059286e-01 -7.52201021e-01 -4.27802294e-01 1.10082150e+00
-1.42579901e+00 -1.33881783e+00 -1.09706603e-01 2.62504667e-01
4.47812676e-01 -3.40799630e-01 1.31274998e+00 4.61409211e-01
-2.51473159e-01 7.05594778e-01 1.39659107e-01 -4.49943393e-02
1.30376852e+00 -1.10527968e+00 1.19880271e+00 1.01762867e+00
7.68188000e-01 5.48729479e-01 9.34051275e-01 -5.50788462e-01
-8.54418993e-01 -7.37497091e-01 1.88329756e+00 -7.57831156e-01
9.56176758e-01 -2.76946843e-01 -9.73027766e-01 5.16866505e-01
5.68617761e-01 -5.04146874e-01 1.10346818e+00 4.72710222e-01
-4.64264691e-01 1.14226840e-01 -7.41005063e-01 7.67720222e-01
1.25373828e+00 -4.95214820e-01 -8.58367443e-01 5.65677941e-01
4.79522467e-01 -4.76909399e-01 -7.77210414e-01 3.66316885e-01
4.82797265e-01 -4.92231846e-01 7.74978280e-01 -8.99766982e-01
4.07818198e-01 -1.32247388e-01 -2.49980152e-01 -1.85654330e+00
3.03216069e-03 -7.54272699e-01 6.21266544e-01 1.33719420e+00
1.00166345e+00 -7.01074958e-01 3.24846327e-01 3.06224823e-01
-1.93170547e-01 -2.82455057e-01 -1.03635454e+00 -1.00707996e+00
3.86476547e-01 -3.84698033e-01 8.41261923e-01 1.28517437e+00
7.34532326e-02 5.62938929e-01 -3.80155832e-01 -3.24526280e-01
1.37441203e-01 1.53079450e-01 9.73061502e-01 -1.11709094e+00
-2.97933757e-01 -4.14228588e-01 -2.34396860e-01 -5.37086904e-01
3.28420937e-01 -1.08894026e+00 1.52405366e-01 -1.54199100e+00
1.36816338e-01 -7.26708233e-01 -1.45655409e-01 5.66263795e-01
-4.47716624e-01 4.63347137e-01 -4.05024029e-02 3.80545437e-01
-4.23221081e-01 1.28449246e-01 9.36219037e-01 8.69632326e-03
-5.74871115e-02 -1.46193966e-01 -3.94929975e-01 6.47626817e-01
1.11746168e+00 -6.59439087e-01 9.21419784e-02 -1.01145411e+00
8.36963832e-01 -7.13771403e-01 -3.58352870e-01 -7.88197279e-01
-3.19031328e-02 -4.78573769e-01 -4.40599322e-01 -2.74953991e-01
-1.65470988e-02 -5.88509500e-01 2.56895125e-01 2.77958184e-01
-6.22916184e-02 9.40158486e-01 1.31643385e-01 -3.32478434e-01
-4.89263803e-01 -4.91512001e-01 4.85378355e-01 -2.81685591e-01
-7.78486848e-01 1.18309364e-01 -1.12993099e-01 7.08816290e-01
5.72566628e-01 2.63819415e-02 -3.09601575e-01 7.37859160e-02
-2.37204924e-01 5.01549579e-02 6.94334865e-01 5.85438967e-01
4.40194085e-02 -1.32820296e+00 -1.32145166e+00 1.86708644e-02
4.61471111e-01 -2.62570620e-01 -4.31829870e-01 6.60530388e-01
-8.20374310e-01 3.83248001e-01 -5.75361848e-01 -2.81655163e-01
-1.63440001e+00 9.11494419e-02 8.62179250e-02 -1.01818681e+00
-3.44819933e-01 7.88890719e-01 -3.23023438e-01 -6.81095660e-01
-3.49929869e-01 3.88987400e-02 -2.72188336e-01 6.54341578e-02
5.05893886e-01 1.75860420e-01 6.58517957e-01 -9.85668480e-01
-4.42955315e-01 4.81315136e-01 -1.56659961e-01 -3.41274321e-01
1.38129926e+00 3.13089825e-02 -4.72707242e-01 3.51322800e-01
8.31393778e-01 3.80414754e-01 -1.69371903e-01 -3.67059082e-01
5.74567437e-01 -5.29714108e-01 -6.90696299e-01 -4.98106897e-01
-3.42753112e-01 9.17456090e-01 2.43444592e-01 -1.35942549e-01
7.39207208e-01 -4.04603854e-02 7.94959307e-01 5.90431809e-01
6.78104341e-01 -1.30043399e+00 -1.31266862e-01 9.96474385e-01
8.84166896e-01 -1.30201411e+00 -2.03244761e-01 -4.98624146e-01
-4.58617121e-01 8.63145173e-01 5.42884469e-01 1.25633270e-01
5.19358478e-02 4.12542880e-01 5.45929134e-01 -5.51671647e-02
-5.63569963e-01 -4.52502668e-01 5.14439642e-01 7.41971254e-01
8.79499674e-01 1.65835425e-01 -1.00979257e+00 4.70632583e-01
-6.84061527e-01 -3.94136518e-01 1.36237830e-01 6.05636418e-01
-1.94746792e-01 -2.18749571e+00 -1.36694700e-01 1.30901739e-01
-9.54783976e-01 -8.47839117e-01 -8.07452619e-01 9.40592527e-01
6.20549060e-02 8.24283421e-01 2.98827514e-02 -1.80744827e-01
3.19631547e-01 3.72580618e-01 6.63570821e-01 -1.00427246e+00
-7.94652343e-01 -4.06920642e-01 8.32699776e-01 -4.10657436e-01
-4.03596401e-01 -6.64218843e-01 -7.18380511e-01 -4.30270374e-01
-2.83425212e-01 4.94838119e-01 5.14899552e-01 1.16013479e+00
2.60919422e-01 -1.89163238e-01 3.80824864e-01 -4.17135149e-01
-5.83889484e-01 -1.15251017e+00 9.16040316e-02 6.88367069e-01
-1.80196792e-01 -3.66290867e-01 -5.49766049e-02 7.50660971e-02] | [11.235960960388184, 10.187694549560547] |
0febbf12-be27-44f7-aae0-559af9b3f324 | bcot-a-markerless-high-precision-3d-object | 2203.13437 | null | https://arxiv.org/abs/2203.13437v1 | https://arxiv.org/pdf/2203.13437v1.pdf | BCOT: A Markerless High-Precision 3D Object Tracking Benchmark | Template-based 3D object tracking still lacks a high-precision benchmark of real scenes due to the difficulty of annotating the accurate 3D poses of real moving video objects without using markers. In this paper, we present a multi-view approach to estimate the accurate 3D poses of real moving objects, and then use binocular data to construct a new benchmark for monocular textureless 3D object tracking. The proposed method requires no markers, and the cameras only need to be synchronous, relatively fixed as cross-view and calibrated. Based on our object-centered model, we jointly optimize the object pose by minimizing shape re-projection constraints in all views, which greatly improves the accuracy compared with the single-view approach, and is even more accurate than the depth-based method. Our new benchmark dataset contains 20 textureless objects, 22 scenes, 404 video sequences and 126K images captured in real scenes. The annotation error is guaranteed to be less than 2mm, according to both theoretical analysis and validation experiments. We re-evaluate the state-of-the-art 3D object tracking methods with our dataset, reporting their performance ranking in real scenes. Our BCOT benchmark and code can be found at https://ar3dv.github.io/BCOT-Benchmark/. | ['Xueying Qin', 'Jason Gu', 'Te Li', 'Wenxuan Chen', 'Fan Zhong', 'Xin Cao', 'Shiqiang Zhu', 'Bin Wang', 'Jiachen Li'] | 2022-03-25 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Li_BCOT_A_Markerless_High-Precision_3D_Object_Tracking_Benchmark_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Li_BCOT_A_Markerless_High-Precision_3D_Object_Tracking_Benchmark_CVPR_2022_paper.pdf | cvpr-2022-1 | ['3d-object-tracking'] | ['computer-vision'] | [-2.72459567e-01 -5.20972729e-01 -1.46388197e-02 5.99207468e-02
-5.82611799e-01 -9.29511011e-01 4.44204271e-01 -5.39920151e-01
-3.75398666e-01 2.93666780e-01 -4.26622748e-01 -9.90485549e-02
2.35217661e-01 -2.08762422e-01 -8.77218008e-01 -7.26698518e-01
1.96576431e-01 7.23619223e-01 1.00874352e+00 1.92424461e-01
7.88875371e-02 7.21034110e-01 -1.53456581e+00 -3.15183699e-01
3.07146251e-01 1.17307544e+00 4.98631686e-01 7.29487419e-01
1.50195330e-01 4.56183106e-01 -4.95718569e-01 -4.72261727e-01
6.46836996e-01 1.50107905e-01 -2.10191563e-01 3.85663509e-01
9.39926028e-01 -6.47880793e-01 -5.12956917e-01 1.14349174e+00
6.10140324e-01 -1.58285037e-01 2.44165450e-01 -1.30448318e+00
-3.55781078e-01 -2.18345612e-01 -7.83980310e-01 1.35552451e-01
5.57625651e-01 9.67921503e-03 3.93590063e-01 -1.08437991e+00
8.84528875e-01 1.07706225e+00 7.60875583e-01 7.45867729e-01
-7.76178181e-01 -6.94433153e-01 3.42163742e-01 4.52165306e-02
-1.58574939e+00 -5.25817811e-01 6.57115817e-01 -6.41232550e-01
5.57706714e-01 3.57875347e-01 1.00156415e+00 8.25409770e-01
3.17770272e-01 5.36184549e-01 7.69846499e-01 -2.15038300e-01
-1.77478328e-01 1.68289363e-01 -1.63298413e-01 7.23830044e-01
7.00649738e-01 2.84506679e-01 -3.42442065e-01 -1.07973345e-01
1.11435378e+00 3.66107851e-01 -4.25000548e-01 -1.17559206e+00
-1.56620932e+00 2.05761045e-01 1.08092241e-01 2.56442488e-03
1.03052765e-01 2.01984763e-01 2.51703322e-01 -9.71564054e-02
6.15355372e-01 -2.61865467e-01 -5.32508612e-01 -1.32395685e-01
-3.56823951e-01 2.69818872e-01 3.67528349e-01 1.60420883e+00
4.13656414e-01 -1.93880871e-02 1.21745922e-01 4.90291595e-01
6.13427162e-01 1.18393111e+00 5.35905138e-02 -1.11192000e+00
4.88447487e-01 5.08185387e-01 5.83570957e-01 -9.58242774e-01
-2.72590190e-01 -2.52017558e-01 -4.97243464e-01 4.13775504e-01
6.26798391e-01 1.10248320e-01 -7.95348942e-01 1.21090329e+00
9.15359139e-01 3.21338654e-01 -1.85818955e-01 1.33498383e+00
1.01893890e+00 3.74856919e-01 -7.07061410e-01 -3.08473706e-01
1.41268003e+00 -9.91512418e-01 -7.36320853e-01 -4.42310311e-02
4.55655783e-01 -1.10283113e+00 6.41980231e-01 4.21016991e-01
-1.03323615e+00 -5.92638075e-01 -9.00746584e-01 8.04947913e-02
3.67348990e-03 2.74107069e-01 4.62446809e-01 8.22303176e-01
-8.12102616e-01 1.10251755e-02 -9.24074471e-01 -2.28103235e-01
1.87772751e-01 3.90301883e-01 -4.55920815e-01 -8.58069882e-02
-7.58388877e-01 1.02441776e+00 2.45714366e-01 2.83754647e-01
-9.21158433e-01 -7.38140643e-01 -7.12203801e-01 -6.16298914e-01
7.01767206e-01 -8.14830184e-01 1.25985813e+00 -3.13077301e-01
-1.62989295e+00 1.32172990e+00 -1.99018016e-01 2.65374575e-02
7.82997847e-01 -5.37195027e-01 -2.96763033e-01 1.16858356e-01
3.19553576e-02 3.23529094e-01 6.69169903e-01 -1.49960697e+00
-6.40123963e-01 -6.28253937e-01 1.33136868e-01 3.57014209e-01
3.33703160e-02 1.12372100e-01 -1.37972116e+00 -2.85861939e-01
3.14325720e-01 -1.18696070e+00 -1.07677102e-01 5.77624321e-01
-1.74168900e-01 -4.51198444e-02 1.27430701e+00 -2.55045563e-01
8.05118680e-01 -2.08237886e+00 3.69921513e-02 -3.02360713e-01
3.44026953e-01 3.19784403e-01 2.45156899e-01 -1.47355661e-01
4.23995346e-01 -4.75046098e-01 4.26252902e-01 -5.24486601e-01
-1.81220606e-01 2.91824583e-02 -8.70287493e-02 9.00580466e-01
-5.43330193e-01 6.77405417e-01 -8.68609428e-01 -6.35737300e-01
6.46790028e-01 4.99436826e-01 -2.50112057e-01 1.93583727e-01
-1.44258112e-01 5.63316941e-01 -5.38917661e-01 8.84205401e-01
1.17678905e+00 -2.86686689e-01 -1.21341810e-01 -3.73513460e-01
-3.52354288e-01 -1.38088897e-01 -1.52832007e+00 1.86917806e+00
-1.33132979e-01 6.28116488e-01 1.06855616e-01 -3.84946257e-01
1.02684104e+00 2.66885191e-01 8.02783132e-01 -4.19574499e-01
1.94926783e-01 2.34679297e-01 -3.85244071e-01 -3.93267781e-01
3.79750967e-01 3.01258892e-01 1.51458308e-01 1.19081952e-01
-2.38213807e-01 -3.32302272e-01 8.10901728e-03 -3.92904207e-02
6.73950434e-01 5.38427055e-01 1.27809525e-01 -7.15474486e-02
5.59558749e-01 3.51694077e-02 7.16574728e-01 5.77169597e-01
-2.35470548e-01 8.04445744e-01 -5.85867688e-02 -6.42536998e-01
-1.02645087e+00 -1.18183339e+00 -3.86097193e-01 4.39664721e-01
9.55712140e-01 -2.84938812e-01 -4.40523267e-01 -5.90626359e-01
2.29900435e-01 7.18061402e-02 -4.32071745e-01 2.67875493e-01
-6.14748061e-01 -3.59082907e-01 2.33769283e-01 4.17504787e-01
3.99526238e-01 -2.19453827e-01 -1.01860619e+00 -8.63389000e-02
-8.59090313e-02 -1.62972283e+00 -7.59988487e-01 -2.56156802e-01
-9.68964577e-01 -1.32843494e+00 -9.55646157e-01 -5.32859325e-01
6.36624515e-01 9.80459750e-01 1.05749595e+00 3.49351857e-03
-1.27238989e-01 5.83391011e-01 -2.25689977e-01 -5.49156129e-01
-7.62828216e-02 -3.90190125e-01 4.66042489e-01 -1.26293033e-01
2.68837065e-01 -5.72737940e-02 -7.13857710e-01 1.05199265e+00
-3.62156749e-01 2.44836345e-01 1.12335563e-01 3.41039926e-01
8.78125966e-01 -3.38527679e-01 -3.00410986e-01 -2.88526028e-01
-5.13701022e-01 1.86964065e-01 -1.54206479e+00 2.47386068e-01
-2.09822685e-01 -5.90001106e-01 1.12118848e-01 -7.21750915e-01
-7.90810704e-01 5.50551832e-01 2.29199499e-01 -1.23514760e+00
1.06692299e-01 -3.04877132e-01 -3.25029612e-01 -5.70279956e-01
3.26196283e-01 1.51110590e-01 2.56121997e-02 -5.23902416e-01
9.57372263e-02 4.53043550e-01 5.38888633e-01 -4.16588515e-01
1.02372789e+00 9.12989080e-01 9.28123742e-02 -6.70311987e-01
-8.51825178e-01 -7.04111099e-01 -8.24503779e-01 -7.63748229e-01
7.67983496e-01 -1.18893969e+00 -1.18799102e+00 6.58679008e-01
-1.36869907e+00 -1.79965660e-01 1.73910372e-02 9.63845789e-01
-6.41315341e-01 4.99879658e-01 -3.14961582e-01 -7.88025260e-01
-2.15939835e-01 -1.28677845e+00 1.45343423e+00 1.77801087e-01
4.49285597e-01 -7.76461065e-01 1.12657656e-03 3.18557143e-01
8.92371833e-02 3.25859129e-01 -2.85567297e-03 6.84142299e-03
-1.09609449e+00 -2.53377229e-01 -1.05983458e-01 -1.66016035e-02
1.09567784e-01 3.83523665e-02 -8.38576436e-01 -3.81479561e-01
1.71293721e-01 1.45164698e-01 3.33816469e-01 6.37561560e-01
8.61865163e-01 2.55390406e-01 -6.86452985e-01 8.54908168e-01
1.43256533e+00 4.76523280e-01 4.41284746e-01 4.23917592e-01
9.80089366e-01 1.54173851e-01 1.10839999e+00 4.07166034e-01
4.04680699e-01 1.49019313e+00 6.35686159e-01 6.68247342e-02
-8.02042261e-02 4.34559323e-02 3.83613914e-01 7.85962522e-01
-4.55563813e-01 -2.43176550e-01 -9.28200006e-01 2.12133020e-01
-1.81676066e+00 -8.78527343e-01 -7.45593429e-01 2.42656684e+00
4.07686412e-01 1.29424214e-01 2.26890326e-01 -2.68051624e-01
8.86006117e-01 2.92086601e-03 -5.90639412e-01 4.97887582e-01
5.86420186e-02 -5.12123466e-01 8.50908577e-01 4.50178266e-01
-1.16485643e+00 8.08423817e-01 5.75196218e+00 5.74168563e-01
-1.03453183e+00 3.54323775e-01 -2.42475439e-02 -3.10537755e-01
2.35966340e-01 -1.19551912e-01 -1.54999578e+00 4.83453095e-01
3.00572872e-01 -1.75055414e-02 5.20632379e-02 8.89125705e-01
4.43811603e-02 -1.42334327e-01 -1.02551675e+00 1.40568197e+00
2.30101675e-01 -1.33005440e+00 -2.52376825e-01 1.76102906e-01
6.89849913e-01 1.90546155e-01 -9.16762277e-02 -2.91079074e-01
-5.27009740e-02 -2.87254959e-01 1.13250387e+00 5.57339072e-01
8.04220259e-01 -2.96698719e-01 7.23475456e-01 3.78548086e-01
-1.61002624e+00 3.74009162e-01 -7.28469253e-01 1.84677735e-01
3.67545336e-01 5.07240415e-01 -4.78673428e-01 7.25037694e-01
1.04356527e+00 7.80087769e-01 -4.81516749e-01 1.30509400e+00
1.48800761e-01 -1.57269001e-01 -4.65159386e-01 5.60966879e-02
-2.51590729e-01 -1.75308406e-01 7.20022023e-01 8.45924377e-01
5.37140667e-01 2.10043997e-01 2.90304959e-01 4.06792372e-01
2.34526336e-01 -3.02254140e-01 -7.22660005e-01 5.63787997e-01
4.52305019e-01 1.20052040e+00 -7.40772963e-01 -2.96853542e-01
-6.65425062e-01 8.50794375e-01 -1.73363060e-01 1.15713045e-01
-1.22153783e+00 7.07275933e-04 7.02179909e-01 2.19592497e-01
6.47234917e-01 -6.00782931e-01 1.44672051e-01 -1.57337880e+00
3.47971559e-01 -5.83315909e-01 8.41187537e-02 -1.05307424e+00
-8.73846591e-01 5.98023176e-01 2.26214156e-01 -2.12198067e+00
5.81200756e-02 -8.18342745e-01 -2.69465838e-02 5.05623102e-01
-1.19992554e+00 -1.12384403e+00 -7.52850652e-01 7.57679701e-01
4.42059457e-01 1.07484059e-02 4.87725049e-01 6.05466008e-01
-2.82510340e-01 4.27396417e-01 2.43619397e-01 7.07987249e-02
8.37123752e-01 -8.18233609e-01 3.43284607e-01 7.40166426e-01
1.47435054e-01 3.49447489e-01 4.77240354e-01 -5.82256913e-01
-2.02850103e+00 -9.39682961e-01 5.19853294e-01 -1.22789574e+00
4.06133085e-01 -8.05443525e-01 -6.62393749e-01 7.10438728e-01
-1.67007670e-01 6.87779844e-01 1.35136917e-01 -3.71731341e-01
-5.04214130e-02 -2.75208920e-01 -8.22179496e-01 3.12327057e-01
1.63595700e+00 -5.51414490e-02 -3.25780511e-01 3.76432151e-01
7.84805775e-01 -1.45969093e+00 -9.09420192e-01 5.19159615e-01
9.40206945e-01 -9.57969189e-01 1.41085839e+00 -5.75709902e-03
-3.59576762e-01 -1.00335348e+00 -2.32016653e-01 -7.40251720e-01
-1.12210132e-01 -5.27868032e-01 -4.11034942e-01 9.96668279e-01
-1.08117498e-01 -6.19268179e-01 9.85463440e-01 2.83024013e-01
-2.37729281e-01 -5.09405553e-01 -1.16772413e+00 -1.24862409e+00
-6.53599679e-01 -3.85267824e-01 4.58852738e-01 7.17810631e-01
-7.18333185e-01 -3.29693332e-02 -5.47341764e-01 5.94262183e-01
1.08182275e+00 4.67084467e-01 1.39838684e+00 -1.31645799e+00
-1.91793263e-01 -9.78401080e-02 -7.91651189e-01 -1.59077895e+00
-1.73745617e-01 -5.05163491e-01 -5.26126586e-02 -1.15128767e+00
3.70726109e-01 -4.56845611e-01 1.83629751e-01 -1.43915368e-02
2.29269881e-02 5.82892120e-01 3.73098642e-01 3.75429153e-01
-9.32143152e-01 4.76998866e-01 1.56507528e+00 6.07071072e-02
1.90507602e-02 2.66693980e-01 -9.05522779e-02 9.57639396e-01
3.18827033e-01 -5.86269200e-01 -8.99623856e-02 -8.70165646e-01
-8.27352852e-02 2.67051935e-01 7.62121677e-01 -9.71472204e-01
3.46839368e-01 -1.84857130e-01 6.76789761e-01 -1.27308488e+00
6.85226798e-01 -1.33159244e+00 6.34959936e-01 6.21166766e-01
4.44723696e-01 2.23737389e-01 3.26692760e-01 4.43149447e-01
6.12847582e-02 -1.34422928e-01 6.33912385e-01 -3.80516313e-02
-7.64823258e-01 7.07373857e-01 1.16640493e-01 -1.66542679e-02
1.28684092e+00 -7.33656526e-01 -3.82492244e-01 -1.28416970e-01
-6.07362807e-01 2.03266695e-01 9.86162841e-01 5.28084993e-01
6.96770966e-01 -1.56184828e+00 -5.51124394e-01 1.78837255e-01
2.41504490e-01 2.24594876e-01 2.07086623e-01 1.05162346e+00
-6.91139698e-01 6.22081399e-01 -2.55864859e-02 -1.42884159e+00
-1.62461662e+00 7.07256615e-01 4.77804482e-01 1.53089374e-01
-8.35283458e-01 4.12568539e-01 4.55692858e-01 -3.86227131e-01
3.30307811e-01 -3.79813582e-01 1.85522273e-01 -4.26042974e-01
5.53207517e-01 3.53460819e-01 7.08163306e-02 -7.15152383e-01
-7.56994367e-01 1.48116744e+00 1.10780835e-01 1.26661822e-01
1.01884735e+00 -2.69949257e-01 2.86883324e-01 3.43860090e-01
1.03565598e+00 1.85285047e-01 -1.61935842e+00 -2.25142658e-01
-4.19004112e-01 -1.15794003e+00 -2.35538512e-01 -1.47180349e-01
-1.13442123e+00 6.43543065e-01 9.83569622e-01 -5.66815883e-02
9.20717537e-01 2.91420579e-01 4.63525295e-01 3.51404756e-01
8.93196821e-01 -6.71417594e-01 1.04526676e-01 5.84784091e-01
7.41506934e-01 -1.29682696e+00 2.82389909e-01 -7.17622340e-01
-3.41068059e-01 1.07275045e+00 9.05947030e-01 -2.01050229e-02
5.18955410e-01 5.02966523e-01 1.88669175e-01 -1.90177888e-01
-4.95356113e-01 -9.07555819e-02 3.98689449e-01 6.34674847e-01
1.55324772e-01 -2.36449167e-01 4.47390601e-02 -4.21735458e-03
2.26022959e-01 -8.65659118e-02 2.82639712e-01 8.36387098e-01
-2.23590016e-01 -8.38310242e-01 -7.67401636e-01 1.02553992e-02
-3.54578137e-01 5.01176000e-01 -8.42487589e-02 1.13411117e+00
2.89927498e-02 6.25761151e-01 1.46303490e-01 -3.28310370e-01
6.87543988e-01 -4.51660424e-01 9.42202151e-01 -3.88289720e-01
-8.51121098e-02 4.37128574e-01 -1.06345110e-01 -7.63627231e-01
-8.87582004e-01 -8.20045590e-01 -9.27842557e-01 -3.04341376e-01
-7.87221551e-01 -1.62760645e-01 6.49424434e-01 4.99318898e-01
3.73054415e-01 3.28323357e-02 6.49930060e-01 -1.30474734e+00
-2.86000729e-01 -4.65946823e-01 -3.32615316e-01 1.78268239e-01
2.90591389e-01 -1.14896393e+00 -2.91902930e-01 1.66233346e-01] | [6.869147300720215, -2.154954671859741] |
63377b83-9894-4598-aa4e-b062fa1de5bc | compressive-self-localization-using-relative | 2208.08863 | null | https://arxiv.org/abs/2208.08863v1 | https://arxiv.org/pdf/2208.08863v1.pdf | Compressive Self-localization Using Relative Attribute Embedding | The use of relative attribute (e.g., beautiful, safe, convenient) -based image embeddings in visual place recognition, as a domain-adaptive compact image descriptor that is orthogonal to the typical approach of absolute attribute (e.g., color, shape, texture) -based image embeddings, is explored in this paper. | ['Kanji Tanaka', 'Ryogo Yamamoto'] | 2022-08-03 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [-1.63294598e-01 -8.47755596e-02 -3.86493146e-01 -4.42396134e-01
-6.58270717e-02 -5.84707201e-01 1.09841990e+00 6.39907956e-01
-8.17794919e-01 5.85594594e-01 3.52334946e-01 5.82745522e-02
-3.55168253e-01 -6.33715987e-01 -2.96823353e-01 -6.85436428e-01
-7.67008141e-02 -5.78547455e-02 -2.24357411e-01 -2.23734409e-01
7.23247468e-01 9.92422402e-01 -1.64153230e+00 -4.61469948e-01
2.98377842e-01 1.14478314e+00 -3.17151874e-01 3.55256796e-01
-1.48603499e-01 4.70591545e-01 -4.17565346e-01 2.79385261e-02
-1.42035574e-01 -1.17582835e-01 -6.00713909e-01 1.29080012e-01
4.15726155e-01 -4.86028679e-02 -2.01661184e-01 1.21510422e+00
1.52320936e-01 4.40967798e-01 1.38869250e+00 -1.30272818e+00
-1.83600676e+00 -1.91225246e-01 -2.90863842e-01 1.42374977e-01
7.98076987e-01 -3.42893779e-01 8.48740578e-01 -1.33810902e+00
7.28328764e-01 1.32856429e+00 4.37212676e-01 4.20535177e-01
-1.52450824e+00 -1.53085142e-01 5.67075163e-02 1.73625857e-01
-2.03319716e+00 -1.64505512e-01 1.14793265e+00 -4.51190799e-01
5.37582457e-01 4.29120034e-01 8.18213046e-01 1.09720516e+00
5.28438270e-01 6.98989034e-01 1.38345683e+00 -5.08446336e-01
4.38475460e-01 6.25730276e-01 -3.02116334e-01 4.88758475e-01
4.34741139e-01 2.31285021e-01 -3.04130822e-01 -3.59528184e-01
8.01129818e-01 4.13082808e-01 -1.25425950e-01 -1.01587296e+00
-1.28293622e+00 1.08430624e+00 8.21931601e-01 4.08783972e-01
-3.39482099e-01 2.31834203e-01 3.86878550e-01 1.53234139e-01
5.97663045e-01 8.81988883e-01 -9.10972953e-02 -2.59176075e-01
-6.25148356e-01 2.55973727e-01 5.32637000e-01 1.27829242e+00
9.91038799e-01 3.39923233e-01 -7.81581402e-02 6.55366898e-01
6.68228567e-01 5.79946220e-01 8.06970298e-01 -5.15632272e-01
-6.04673624e-01 5.66906631e-01 1.10628055e-02 -1.39914870e+00
-3.04264277e-01 4.31860149e-01 -5.25939822e-01 4.75493968e-01
-2.44159728e-01 6.48978651e-01 -8.27620268e-01 1.47306585e+00
1.18074790e-01 -1.01253331e-01 9.05646831e-02 9.69593644e-01
8.68483663e-01 4.45252925e-01 4.00941253e-01 3.88989329e-01
1.33683062e+00 -7.39222884e-01 -6.92612886e-01 -9.83943697e-03
3.97483170e-01 -6.53060734e-01 1.06586921e+00 -5.78493550e-02
-5.20973146e-01 -3.53432626e-01 -1.30239165e+00 -1.94217280e-01
-1.45495820e+00 -8.82449076e-02 6.89008057e-01 7.60646641e-01
-1.23558486e+00 5.40052712e-01 -2.81161755e-01 -8.18727434e-01
3.39598566e-01 1.78697452e-01 -1.20352960e+00 -8.72309785e-03
-8.71600389e-01 1.27076399e+00 6.05392992e-01 -4.28775162e-01
-5.01025975e-01 -2.09058881e-01 -1.71393764e+00 -3.98418307e-01
-3.76177698e-01 2.17180345e-02 6.47348821e-01 -9.09918547e-01
-1.52559638e+00 1.15388322e+00 9.76742581e-02 -1.53968140e-01
2.26458572e-02 2.97800694e-02 -7.02370763e-01 1.87519282e-01
-8.76613110e-02 1.12016547e+00 1.26464689e+00 -1.12577462e+00
-1.03197716e-01 -4.25541610e-01 -1.58749267e-01 2.03644916e-01
-5.39945662e-01 5.97500987e-02 5.20830810e-01 -9.22725379e-01
2.46009022e-01 -8.62966418e-01 -2.07003489e-01 8.29200983e-01
4.19543730e-03 -4.08413798e-01 9.66573238e-01 -4.24075127e-01
8.12477946e-01 -2.84121060e+00 3.64442736e-01 3.76285821e-01
2.90819049e-01 2.03199554e-02 -3.79459537e-03 5.74381471e-01
-1.50799006e-01 3.49159300e-01 -3.87658566e-01 -4.50285450e-02
3.85632247e-01 6.47193253e-01 1.56672329e-01 1.16334617e+00
5.04323184e-01 8.80745828e-01 -1.14226794e+00 -6.19631469e-01
5.66774726e-01 8.36110950e-01 -1.07484818e-01 -2.87673716e-02
3.70559305e-01 2.83455729e-01 -6.72990859e-01 1.00668693e+00
5.62737346e-01 3.78737479e-01 -3.98952454e-01 6.48464337e-02
-5.76365590e-01 -9.34951156e-02 -9.00701284e-01 1.50346684e+00
-4.08778399e-01 1.10804236e+00 -4.15274024e-01 -6.51513934e-01
1.52781415e+00 3.49124044e-01 1.43594280e-01 -8.38641763e-01
1.71716452e-01 2.60358036e-01 -4.76710200e-01 -1.77627355e-01
1.01149356e+00 -1.63952246e-01 -3.52236778e-01 2.25702092e-01
1.87812626e-01 -4.91049923e-02 -3.25073123e-01 -3.51819307e-01
4.43790942e-01 3.43149543e-01 1.14762056e+00 -9.86215174e-01
5.54544985e-01 -2.35804275e-01 -9.74864978e-03 1.00027323e-01
-8.04584205e-01 7.43533075e-01 7.61495680e-02 -6.17050946e-01
-1.51019239e+00 -1.23356044e+00 -5.03722727e-01 1.02551329e+00
4.49316859e-01 -2.92101920e-01 -4.11282331e-01 -1.57736272e-01
4.62599218e-01 5.75628996e-01 -1.06936109e+00 -7.04681516e-01
-2.79735439e-02 2.88456231e-01 4.30612206e-01 9.84879613e-01
1.47967905e-01 -1.12105691e+00 -7.65455782e-01 1.73148438e-01
6.01376832e-01 -5.44120789e-01 -5.77105880e-01 2.48103559e-01
-6.52284205e-01 -5.07432520e-01 -9.84486639e-01 -8.73857081e-01
8.54695797e-01 2.02596590e-01 9.85980630e-01 -1.71335429e-01
-3.24159056e-01 9.93233025e-01 -9.79757369e-01 -4.01802808e-01
2.37852395e-01 -5.63446462e-01 3.77307624e-01 5.35524249e-01
5.91916442e-01 -3.47191453e-01 -6.20606542e-01 5.12910411e-02
-1.17893386e+00 -5.69228828e-01 5.28632045e-01 1.06885755e+00
9.25782323e-01 -5.22164881e-01 -2.41774227e-02 -4.43289965e-01
7.82346904e-01 -3.19939971e-01 8.94474089e-02 1.01854593e-01
-7.19663084e-01 2.16225475e-01 5.22770524e-01 -8.76876533e-01
-4.31297719e-01 -7.64020160e-02 2.07086295e-01 -6.31029785e-01
-5.73105097e-01 5.51981509e-01 1.18198276e-01 -8.09676528e-01
7.89286971e-01 7.00558007e-01 1.13161504e-01 -3.18918079e-01
7.03693569e-01 7.47457504e-01 3.98343712e-01 -4.72964853e-01
7.06302643e-01 5.30779779e-01 4.19664234e-02 -1.24415076e+00
6.42476678e-02 -7.01077163e-01 -1.07092810e+00 -6.07631393e-02
1.18476284e+00 -6.57183886e-01 -2.66421944e-01 3.39972451e-02
-9.81690526e-01 2.43839264e-01 -5.46801329e-01 7.47777641e-01
-9.24839735e-01 1.37438983e-01 -1.12985158e-02 -9.93965805e-01
3.34960893e-02 -5.87440908e-01 1.14329624e+00 3.86660874e-01
-4.61256385e-01 -1.26662290e+00 1.65106449e-02 -6.61653578e-01
6.68495476e-01 8.73949349e-01 7.09069490e-01 -8.12992692e-01
-8.67552161e-02 -4.19274688e-01 -3.07938606e-01 3.42624485e-01
3.99484932e-01 1.38641998e-01 -8.30375791e-01 -1.66469947e-01
-1.23460047e-01 -2.53045619e-01 5.67965567e-01 -3.16762291e-02
8.84238541e-01 -5.39562464e-01 -2.52921462e-01 7.18787611e-01
1.59082484e+00 4.44801062e-01 7.70603478e-01 7.05928028e-01
5.15747726e-01 3.27481002e-01 8.51567686e-01 6.51128113e-01
2.81573832e-01 6.18423223e-01 1.29898742e-01 -1.80153504e-01
7.33343959e-02 -3.19303483e-01 2.26235643e-01 3.12150538e-01
-1.13957658e-01 2.61089116e-01 -7.30138302e-01 6.77217424e-01
-1.49441588e+00 -6.63451195e-01 6.07083738e-01 2.02859664e+00
4.27456915e-01 -1.57062471e-01 6.26665950e-02 1.99442714e-01
6.97364271e-01 4.03113782e-01 -9.92749482e-02 -1.29890013e+00
-1.53238654e-01 1.82334706e-01 6.29710138e-01 8.26833844e-02
-1.44918334e+00 1.01003957e+00 6.51924610e+00 6.67661428e-01
-1.32662988e+00 1.91427302e-02 3.47673118e-01 6.62964940e-01
-5.18479228e-01 -1.32076636e-01 5.34381680e-02 4.05560285e-01
6.29679382e-01 -5.14544189e-01 5.30853033e-01 1.46597958e+00
-3.22179377e-01 1.16347060e-01 -1.58850157e+00 1.30563021e+00
5.90496659e-01 -9.44601059e-01 7.96778351e-02 -3.52240913e-02
2.21708775e-01 -6.61957324e-01 7.12664545e-01 4.81627882e-02
-4.11544770e-01 -1.41501188e+00 1.32574701e+00 4.90097374e-01
1.10871375e+00 -7.72688329e-01 4.92743373e-01 -3.92273307e-01
-1.29388058e+00 9.91650149e-02 -4.91924614e-01 -2.57183939e-01
-1.98171124e-01 -3.14552456e-01 -4.82105672e-01 3.56259853e-01
7.79732168e-01 1.08620417e+00 -7.56033421e-01 1.04787886e+00
-1.55745894e-01 -3.05680819e-02 -8.21912512e-02 -5.21601260e-01
7.09943771e-01 -3.76845837e-01 6.25144839e-01 1.39524651e+00
2.64025927e-01 1.72846747e-04 4.64689955e-02 9.07535851e-01
4.31966960e-01 4.08269316e-01 -1.60595012e+00 -4.12539095e-01
3.81665587e-01 1.07670641e+00 -6.63849652e-01 -1.50149390e-01
-4.19954926e-01 1.15499592e+00 2.34918535e-01 2.58106440e-01
-4.94271278e-01 -6.85665309e-01 1.00375700e+00 -9.62653011e-03
2.98544168e-01 -7.33114064e-01 -2.25499168e-01 -5.05061865e-01
-3.02736580e-01 -2.69859463e-01 1.87263489e-01 -8.87835920e-01
-1.19089782e+00 7.44643450e-01 -3.56789231e-02 -1.64940262e+00
1.28136203e-01 -1.15108120e+00 -5.30788124e-01 7.76998162e-01
-1.34396303e+00 -1.20261002e+00 -7.39836469e-02 8.43714654e-01
-5.13829663e-03 -1.32020339e-01 1.31558871e+00 1.45294964e-01
2.10496128e-01 7.14405239e-01 3.84971648e-01 3.07928503e-01
5.34257174e-01 -1.51750386e+00 2.15470493e-01 5.97012825e-02
3.79766762e-01 8.79338324e-01 6.10801756e-01 -7.26092830e-02
-1.68393242e+00 -9.40472364e-01 9.04028237e-01 -6.49657786e-01
8.02886307e-01 -2.06905261e-01 -7.01966703e-01 5.33684671e-01
2.83767004e-02 2.32606545e-01 8.77631128e-01 -9.05029029e-02
-7.93755651e-01 -3.65273631e-03 -1.62212026e+00 9.36265767e-01
5.47768176e-01 -1.01517630e+00 -7.62229085e-01 -8.96101892e-02
7.20794976e-01 -7.09418207e-02 -1.26281714e+00 -7.66996220e-02
5.83325446e-01 -3.26681882e-01 1.17081940e+00 -7.21027493e-01
-1.22512981e-01 -3.08257461e-01 -4.73191649e-01 -1.48960149e+00
-9.58828032e-01 -1.05070516e-01 1.00106619e-01 9.21652794e-01
2.52159864e-01 -9.40802336e-01 1.96408585e-01 4.00119185e-01
-2.03042060e-01 -6.11566842e-01 -1.43020689e+00 -1.22486389e+00
6.82736933e-02 2.09423780e-01 7.68837154e-01 1.03008986e+00
4.06505585e-01 -2.08775803e-01 -3.67964655e-01 -1.63363233e-01
2.98443377e-01 -5.02629802e-02 1.15042552e-01 -1.38904452e+00
5.46374798e-01 -7.20353007e-01 -1.74080598e+00 -4.38836634e-01
1.49221480e-01 -8.46991181e-01 -6.00372851e-02 -1.32652235e+00
-6.86036050e-02 -3.69673222e-01 -8.55681181e-01 3.56453061e-01
2.10576311e-01 5.19362032e-01 1.87446460e-01 -3.60416956e-02
-7.40801930e-01 1.00538468e+00 1.24723053e+00 -5.16424716e-01
3.15080909e-03 -6.36562824e-01 -7.04276443e-01 4.07629699e-01
7.98963726e-01 -4.15914237e-01 -5.76232001e-02 -3.64083424e-02
-8.60539302e-02 -4.98184323e-01 5.13615131e-01 -9.37647641e-01
6.34710789e-02 -3.13551873e-01 5.98759532e-01 2.12516844e-01
6.82179332e-01 -1.00364232e+00 -1.40866190e-01 5.04190922e-01
-2.93528110e-01 5.02130687e-01 3.30138877e-02 7.65188873e-01
-6.07185781e-01 -2.66037792e-01 6.85940087e-01 -4.92655784e-02
-1.46336377e+00 3.92431378e-01 -4.61996853e-01 -2.25237951e-01
1.27561712e+00 -1.11632371e+00 -8.62482786e-02 -1.76255107e-01
-4.49323505e-01 -4.85761195e-01 9.07491028e-01 7.65022695e-01
1.29430997e+00 -2.18056726e+00 -5.46378851e-01 2.66159177e-01
1.18329191e+00 -8.21192503e-01 -2.29738578e-01 5.78348577e-01
-7.35534310e-01 4.20271814e-01 -9.09498394e-01 -4.74394530e-01
-8.99429679e-01 1.27869904e+00 -7.09462389e-02 7.15531766e-01
-8.56672704e-01 6.66830957e-01 4.86260019e-02 -2.96162248e-01
-7.47512057e-02 -2.31714383e-01 -5.39848030e-01 5.70499785e-02
4.79875654e-01 3.65537256e-01 -4.00667429e-01 -1.29454601e+00
-9.49912488e-01 7.45065391e-01 2.47053474e-01 -1.93922386e-01
9.57483649e-01 1.18602149e-01 -4.23815727e-01 9.85263884e-01
1.47026002e+00 -3.62696528e-01 -8.94938231e-01 -1.01806641e-01
3.05842668e-01 -1.01320755e+00 4.88599576e-03 -3.87935311e-01
-4.09132272e-01 7.12689161e-01 9.93766248e-01 2.99277872e-01
7.40170360e-01 1.80837959e-01 -3.05516616e-04 3.82024944e-01
6.55433774e-01 -1.28690302e+00 9.06995609e-02 3.93396407e-01
1.30223143e+00 -1.31028438e+00 -2.01150015e-01 3.55241708e-02
-6.82198763e-01 1.10675681e+00 2.97185272e-01 -2.24019617e-01
1.05202854e+00 -1.85147956e-01 3.56510840e-02 -2.11261481e-01
6.25581741e-02 -2.16698006e-01 6.79528058e-01 1.22596431e+00
5.60798347e-01 7.89228916e-01 -4.80917543e-01 6.52333423e-02
2.58338004e-02 -5.37110806e-01 3.73666078e-01 1.28174722e+00
-5.16255796e-01 -7.01825857e-01 -4.56776589e-01 2.13572770e-01
-6.00792319e-02 -7.41181225e-02 -5.51700711e-01 8.61201406e-01
1.83370396e-01 6.35206521e-01 4.23969358e-01 -4.43976909e-01
3.28322023e-01 -6.79678470e-02 6.24375284e-01 -5.33569217e-01
-6.47648796e-02 -5.95505953e-01 -3.78578901e-01 -4.94574219e-01
-4.65004295e-01 -5.51111996e-01 -9.77785766e-01 -3.58785450e-01
1.64951324e-01 -9.46254432e-02 7.58979976e-01 6.18443251e-01
7.98072368e-02 -1.08967870e-01 5.44127762e-01 -1.17809737e+00
-2.18446046e-01 -9.58992302e-01 -1.16247082e+00 5.74763894e-01
5.64348459e-01 -1.24745524e+00 -1.54785126e-01 -2.69647479e-01] | [7.720747947692871, -1.5958683490753174] |
8d71ef41-7c7f-4cb0-8222-a7d29993cc41 | real-time-hand-gesture-identification-in | 2303.02321 | null | https://arxiv.org/abs/2303.02321v1 | https://arxiv.org/pdf/2303.02321v1.pdf | Real-Time Hand Gesture Identification in Thermal Images | Hand gesture-based human-computer interaction is an important problem that is well explored using color camera data. In this work we proposed a hand gesture detection system using thermal images. Our system is capable of handling multiple hand regions in a frame and process it fast for real-time applications. Our system performs a series of steps including background subtraction-based hand mask generation, k-means based hand region identification, hand segmentation to remove the forearm region, and a Convolutional Neural Network (CNN) based gesture classification. Our work introduces two novel algorithms, bubble growth and bubble search, for faster hand segmentation. We collected a new thermal image data set with 10 gestures and reported an end-to-end hand gesture recognition accuracy of 97%. | ['Soumyabrata Dey', 'James Ballow'] | 2023-03-04 | null | null | null | null | ['hand-gesture-recognition', 'hand-gesture-recognition-1', 'gesture-recognition', 'hand-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 4.14802432e-01 -8.06895375e-01 -1.66605264e-02 -2.13257074e-01
-2.89786696e-01 -6.74547672e-01 1.97959796e-01 -5.67366540e-01
-1.05483437e+00 3.19356263e-01 -1.55402005e-01 -4.61349696e-01
4.06919003e-01 -4.17067528e-01 -5.98103367e-03 -9.53349769e-01
4.25938983e-03 5.39483845e-01 6.81720197e-01 2.14863166e-01
5.51273763e-01 1.01129448e+00 -1.16111279e+00 3.63069952e-01
1.08392172e-01 6.99148536e-01 1.33448318e-01 1.54501307e+00
-3.21098268e-01 5.70118904e-01 -7.49370694e-01 1.37120783e-01
6.22594893e-01 -4.26922560e-01 -9.00385499e-01 3.88211869e-02
2.70959407e-01 -9.61528242e-01 -3.14338416e-01 5.32482028e-01
1.11987269e+00 3.83274049e-01 4.83682424e-01 -1.10729337e+00
2.67593432e-02 2.66206443e-01 -9.83394444e-01 4.27554190e-01
3.15376699e-01 5.86163700e-01 1.40828669e-01 -7.53439486e-01
6.56767428e-01 1.20195878e+00 3.42922211e-01 1.08960330e+00
-6.49851620e-01 -9.89180088e-01 -6.40755668e-02 1.52109489e-01
-1.50755596e+00 5.44268824e-02 6.15919352e-01 -3.29283297e-01
1.29415500e+00 4.38641876e-01 8.53634715e-01 9.48556423e-01
-1.48283306e-03 1.01240444e+00 1.12123823e+00 -8.96772325e-01
1.97338745e-01 -5.66672981e-01 4.07662600e-01 8.08948159e-01
-2.36180171e-01 2.62975901e-01 -4.25016344e-01 5.56924939e-02
1.57474518e+00 3.12786460e-01 -1.61033243e-01 2.02962339e-01
-1.16899931e+00 2.05518335e-01 3.59444499e-01 5.37937999e-01
-4.56693381e-01 4.55952823e-01 4.60106313e-01 -1.33361295e-01
-1.59548789e-01 -4.48538601e-01 -3.19142222e-01 -4.72426951e-01
-1.10948896e+00 1.00774691e-01 8.67796898e-01 8.40301633e-01
-1.11789048e-01 -2.68345773e-01 -3.99743199e-01 3.80805105e-01
4.63461190e-01 6.49617553e-01 3.85247856e-01 -5.74788749e-01
4.50460494e-01 6.18205428e-01 1.44679502e-01 -3.31205994e-01
-5.32811999e-01 5.69046736e-01 -9.84963715e-01 1.01110280e+00
1.09959829e+00 -3.22314322e-01 -1.71604443e+00 8.62003744e-01
3.29093993e-01 -2.42537875e-02 -4.79167104e-01 1.44658899e+00
7.11027801e-01 3.80529106e-01 3.84045482e-01 -4.31672931e-02
1.53357637e+00 -9.86198425e-01 -6.90383911e-01 2.45566964e-01
1.91536138e-03 -9.41534519e-01 1.14021325e+00 8.33923042e-01
-7.92186737e-01 -3.94864738e-01 -6.63057148e-01 5.74016683e-02
-3.84810388e-01 3.00672323e-01 6.99554205e-01 1.29040444e+00
-9.55672383e-01 1.56160876e-01 -1.32781410e+00 -7.03814507e-01
4.72472310e-01 8.82483304e-01 -1.57613426e-01 2.63676673e-01
-2.57560432e-01 4.50286239e-01 4.76370841e-01 5.29231906e-01
-3.95604044e-01 -1.13659957e-02 -1.45101622e-01 -2.23893702e-01
2.74647623e-01 -3.18541974e-01 1.18121457e+00 -9.39522207e-01
-1.99693465e+00 9.06451344e-01 -5.01281321e-01 2.30487302e-01
8.43985438e-01 -2.85138309e-01 -2.54620127e-02 3.40730339e-01
-8.83643568e-01 5.16753554e-01 1.16457140e+00 -1.00875330e+00
-8.14322174e-01 -8.08227420e-01 -5.19228756e-01 1.28131002e-01
7.24151060e-02 8.13421309e-01 -8.57306778e-01 -6.57525659e-01
3.29451561e-01 -8.74971032e-01 -1.30701378e-01 -1.22757200e-02
-3.59777510e-01 -4.53003347e-01 1.08955777e+00 -1.11488652e+00
1.18986189e+00 -1.65149701e+00 -7.64083788e-02 7.18997359e-01
1.36237070e-01 7.98570514e-01 -1.20401315e-01 -1.75216421e-02
1.26260966e-01 1.03419766e-01 -1.13790587e-01 -5.36993071e-02
-4.19539362e-01 8.66657123e-02 -5.24077229e-02 2.03951210e-01
-1.69450596e-01 1.00084627e+00 -6.06632352e-01 -7.76206434e-01
7.77210772e-01 7.13447809e-01 2.52000224e-02 3.73111218e-01
4.71024886e-02 7.08566904e-01 -2.66323566e-01 1.22484720e+00
6.26790881e-01 3.20022225e-01 1.11467816e-01 1.99414082e-02
-1.68097943e-01 -4.93811995e-01 -1.47833669e+00 1.54571033e+00
-2.25398429e-02 8.14487159e-01 3.08955371e-01 -9.07511488e-02
6.54798687e-01 3.74378204e-01 4.58367735e-01 -2.91654527e-01
6.83197021e-01 1.48247465e-01 4.60427888e-02 -7.85339594e-01
3.11279774e-01 2.57984489e-01 6.43240154e-01 1.09790456e+00
-3.73835325e-01 1.40001893e-01 -1.40419260e-01 -2.37671420e-01
8.99963558e-01 2.78007925e-01 1.10342488e-01 1.36956170e-01
3.44207555e-01 3.03722769e-01 -6.28629774e-02 8.51248145e-01
-7.51659155e-01 8.63043427e-01 -1.17422596e-01 -7.78655291e-01
-8.94988298e-01 -7.77697921e-01 4.54573154e-01 1.37149787e+00
2.70531084e-02 1.95392221e-01 -1.04852402e+00 -6.38027012e-01
-3.92367750e-01 -3.67035344e-02 -3.68137628e-01 7.95864463e-01
-1.34955263e+00 -6.95465028e-01 5.80784500e-01 1.06906664e+00
8.51803303e-01 -1.64216113e+00 -1.44053376e+00 1.73473537e-01
2.23003596e-01 -6.75640225e-01 -7.60010183e-01 2.25904569e-01
-9.46269274e-01 -1.43378031e+00 -1.33222759e+00 -1.03607118e+00
7.57524610e-01 1.03156656e-01 4.65859324e-01 5.06646156e-01
-1.10996926e+00 3.43725353e-01 -6.76824629e-01 -3.51648837e-01
7.06959292e-02 1.25954896e-01 -2.52719164e-01 -3.69081438e-01
8.37524295e-01 -1.27456069e-01 -1.04430938e+00 3.11458230e-01
-7.53890991e-01 -2.03676417e-01 5.88217258e-01 6.19925201e-01
2.68519998e-01 -3.13702285e-01 -5.65334916e-01 -6.64318651e-02
7.08183289e-01 6.48361325e-01 -4.39408690e-01 7.17466474e-01
-1.66183591e-01 -1.65959597e-01 -1.57608047e-01 -8.20806563e-01
-1.22460544e+00 7.73755074e-01 -9.27238315e-02 -2.43102595e-01
-6.70500398e-01 -2.92141050e-01 9.98808220e-02 -4.00854588e-01
5.04383266e-01 3.31164032e-01 9.77529027e-03 -5.62455952e-01
4.62388575e-01 1.34328580e+00 6.63280308e-01 -4.32163894e-01
3.50811064e-01 5.39826989e-01 -3.07198614e-01 -9.78838861e-01
4.05208737e-01 -9.52547133e-01 -1.59993744e+00 -6.82645202e-01
1.19128466e+00 -1.65048674e-01 -1.17830431e+00 1.30920553e+00
-1.43127012e+00 -8.64081740e-01 2.75286764e-01 5.15100479e-01
-7.63025731e-02 3.42298001e-01 -7.70474494e-01 -1.60984159e+00
-9.28789318e-01 -8.61617565e-01 1.06863785e+00 5.23714125e-01
-2.11605996e-01 -4.59819704e-01 -2.87619177e-02 6.09347485e-02
3.01145077e-01 2.80843258e-01 4.06981558e-01 -3.29557925e-01
-7.44938910e-01 -2.90984184e-01 -5.22038519e-01 -2.10516751e-02
2.44919673e-01 4.04303730e-01 -8.90016496e-01 -2.43856356e-01
-4.46045339e-01 3.28420140e-02 1.01805401e+00 5.82018018e-01
1.15509784e+00 1.97399095e-01 -5.34722626e-01 4.10295963e-01
1.26693380e+00 8.62382352e-01 8.21256697e-01 1.41127959e-01
1.00192487e+00 3.42834383e-01 3.86239678e-01 4.68063265e-01
-2.89130330e-01 5.61771691e-01 7.76693551e-03 -1.53471142e-01
-3.34268481e-01 4.20435697e-01 -1.87860876e-02 2.30090171e-01
-1.09124899e+00 -3.00679892e-01 -1.23388922e+00 2.86637276e-01
-1.70231462e+00 -8.36026371e-01 -3.64574999e-01 1.94406021e+00
5.81806064e-01 -4.24368441e-01 8.20975363e-01 3.98862183e-01
8.36912513e-01 -4.41141486e-01 -5.35084188e-01 -2.04196274e-01
2.84730792e-01 1.07989383e+00 6.97134793e-01 5.72343647e-01
-1.26814342e+00 1.43804622e+00 6.64534760e+00 4.29394037e-01
-1.38810933e+00 9.56721976e-02 4.10356671e-01 -2.15851158e-01
8.26130748e-01 -6.41568184e-01 -5.68535030e-01 -2.02136487e-02
1.12142384e-01 8.09695542e-01 6.44121826e-01 5.75817108e-01
2.01563299e-01 -6.11220956e-01 -8.09048533e-01 1.33256590e+00
9.70072970e-02 -7.57913947e-01 -4.83339317e-02 -7.22181872e-02
2.08238170e-01 -2.27378741e-01 -3.49337041e-01 -3.79161537e-01
1.66220874e-01 -1.26126122e+00 6.26477778e-01 2.80221432e-01
9.57163215e-01 -4.71932709e-01 6.15621448e-01 1.69835806e-01
-1.46559930e+00 -6.27809996e-03 2.88028061e-01 -2.84309685e-01
1.60545036e-01 -2.90221393e-01 -1.03373277e+00 -2.05718040e-01
9.49569583e-01 -2.54734755e-01 -3.11780035e-01 1.09604728e+00
-5.45663908e-02 5.80163479e-01 -6.73500419e-01 -5.63481331e-01
-1.35325862e-03 -4.97172074e-03 2.42676914e-01 1.61028981e+00
-3.92048210e-02 9.39945221e-01 1.35592045e-02 6.27886355e-01
4.77036506e-01 3.09846271e-02 1.88319296e-01 -7.57984966e-02
2.71928698e-01 1.05035031e+00 -1.85560632e+00 -4.84395117e-01
8.72079469e-03 1.74732435e+00 -3.17256331e-01 5.53859890e-01
-4.19441432e-01 -8.57647538e-01 2.65542209e-01 -2.42207453e-01
2.33729899e-01 -9.24963593e-01 -8.12864125e-01 -8.23563755e-01
1.42735228e-01 -5.06241441e-01 4.49061990e-01 -5.17539084e-01
-7.39445567e-01 3.71630549e-01 -2.10673362e-01 -6.38604581e-01
-2.07160473e-01 -1.27720606e+00 -7.35734701e-01 1.12355149e+00
-8.89764726e-01 -1.19705999e+00 -9.38935161e-01 1.03585088e+00
7.63349116e-01 -2.09251195e-02 7.99435496e-01 5.67145683e-02
-6.58193588e-01 5.76695025e-01 -3.13642263e-01 9.24776256e-01
5.65247118e-01 -1.20807302e+00 4.81833696e-01 9.69583690e-01
9.98589918e-02 7.63290524e-01 1.62149161e-01 -9.43380415e-01
-1.37565327e+00 -5.22024512e-01 3.48353088e-01 -5.89687169e-01
-4.40568514e-02 -1.35736987e-01 -5.57437181e-01 5.65733612e-01
2.02920482e-01 -1.91981941e-01 4.24907684e-01 -2.84615487e-01
-1.49408773e-01 3.19314599e-01 -1.43538833e+00 5.12593627e-01
1.15005195e+00 -2.24766016e-01 -4.27117944e-01 1.06256858e-01
-9.67385918e-02 -7.69001961e-01 -3.10148537e-01 -6.44006431e-02
1.41806018e+00 -8.03878367e-01 1.05165899e+00 -4.65509266e-01
-2.28961840e-01 -2.73468852e-01 2.41366386e-01 -4.89772379e-01
1.52198691e-02 -6.86736166e-01 -8.36277828e-02 8.30759645e-01
-1.03061080e-01 -2.76614100e-01 1.28768241e+00 1.30558217e+00
6.73197150e-01 -1.35171741e-01 -9.11164343e-01 -5.76685429e-01
-1.08144075e-01 -5.69484293e-01 3.79506767e-01 3.17052066e-01
1.07536606e-01 -3.92298311e-01 -3.89903784e-03 2.99680587e-02
7.54280269e-01 -1.03435740e-02 7.70193636e-01 -9.57440138e-01
2.37092331e-01 -8.75027239e-01 -2.04928517e-01 -8.71050835e-01
-5.27402282e-01 -1.10358782e-01 2.41056621e-01 -1.43560708e+00
3.79978150e-01 -3.19099575e-01 -7.59656206e-02 7.25691617e-01
-2.33310629e-02 6.07076049e-01 3.90667528e-01 3.72298896e-01
-1.26920417e-01 -4.56100881e-01 1.19785547e+00 -7.30042309e-02
-9.49020684e-01 8.19662288e-02 4.14361864e-01 4.73666638e-01
8.95865798e-01 -1.32245990e-02 2.68443435e-01 -2.54182249e-01
-5.92626333e-01 -1.12515472e-01 5.66885591e-01 -8.31581712e-01
5.28712749e-01 -4.57624078e-01 8.69243085e-01 -9.89012957e-01
4.46979478e-02 -9.80226755e-01 -1.45043463e-01 1.14824498e+00
-1.45025760e-01 -1.15058690e-01 -2.96949856e-02 1.58806607e-01
2.48248279e-01 -5.26916869e-02 5.67968488e-01 -4.76477534e-01
-1.23645270e+00 4.50929962e-02 -6.40119255e-01 -6.14464700e-01
1.18354428e+00 -6.74461365e-01 6.54384494e-02 3.06829251e-02
-9.27714169e-01 -1.75929978e-01 2.95352377e-03 4.70383316e-01
9.54829991e-01 -8.94358277e-01 -5.00862300e-01 4.82460737e-01
-3.25927973e-01 3.54219824e-02 -1.96583420e-01 7.69226193e-01
-1.44104505e+00 4.62663382e-01 -4.37199682e-01 -7.93436527e-01
-2.37116051e+00 1.25116736e-01 2.36033157e-01 1.57458737e-01
-8.37065518e-01 1.25969744e+00 -7.45296597e-01 4.66733240e-02
9.29915249e-01 -9.82173741e-01 3.83536629e-02 -2.77710319e-01
1.08100307e+00 7.80103385e-01 -1.42707855e-01 -5.04806817e-01
-5.67755938e-01 9.66951728e-01 2.59453803e-01 -7.07202137e-01
7.36417413e-01 1.19318135e-01 -2.11102441e-01 -1.37642007e-02
7.23997712e-01 -4.69555736e-01 -1.24641907e+00 2.91623361e-02
8.73098820e-02 -8.98771644e-01 1.27645493e-01 -1.30867171e+00
-1.14553869e+00 1.14768529e+00 1.47066307e+00 -2.82373339e-01
1.30139947e+00 -9.59769040e-02 9.24507320e-01 5.66289544e-01
5.02683699e-01 -1.30747199e+00 6.06297925e-02 6.16526425e-01
7.24579275e-01 -1.29025853e+00 -2.04017013e-01 -4.68812883e-01
-3.84057105e-01 1.70525074e+00 7.11406052e-01 -5.16256783e-04
5.80046892e-01 7.25837708e-01 6.42001152e-01 -9.49598923e-02
2.50122637e-01 -6.71877563e-01 3.57736081e-01 8.82980943e-01
3.70082438e-01 3.16905737e-01 -4.15441602e-01 3.25746834e-01
1.31728485e-01 5.36523223e-01 -7.13986605e-02 1.47964966e+00
-5.34649849e-01 -1.05233419e+00 -9.68728542e-01 2.24120289e-01
-2.78676659e-01 6.36343798e-03 -8.62691104e-01 8.83715272e-01
-4.85575497e-02 8.45847785e-01 9.48665217e-02 -6.90113127e-01
-9.29066092e-02 5.84427714e-01 1.14252293e+00 -2.80128270e-01
-1.01348114e+00 3.51469547e-01 -6.43998384e-01 -7.19428480e-01
-5.46282828e-01 -5.51467538e-01 -1.64619541e+00 -3.15726519e-01
-3.04031640e-01 -6.25868797e-01 1.01687658e+00 9.82428908e-01
-2.17543960e-01 2.74462402e-01 -2.08871420e-02 -1.55693448e+00
-6.05916530e-02 -1.25976872e+00 -5.60695946e-01 1.43784419e-01
3.66426319e-01 -2.51237005e-01 1.69480219e-01 4.57809865e-01] | [6.497348308563232, -0.35542652010917664] |
7152836e-a683-4373-a572-3ce2da5d5a4b | don-t-freeze-finetune-encoders-for-better | 2307.01168 | null | https://arxiv.org/abs/2307.01168v1 | https://arxiv.org/pdf/2307.01168v1.pdf | Don't freeze: Finetune encoders for better Self-Supervised HAR | Recently self-supervised learning has been proposed in the field of human activity recognition as a solution to the labelled data availability problem. The idea being that by using pretext tasks such as reconstruction or contrastive predictive coding, useful representations can be learned that then can be used for classification. Those approaches follow the pretrain, freeze and fine-tune procedure. In this paper we will show how a simple change - not freezing the representation - leads to substantial performance gains across pretext tasks. The improvement was found in all four investigated datasets and across all four pretext tasks and is inversely proportional to amount of labelled data. Moreover the effect is present whether the pretext task is carried on the Capture24 dataset or directly in unlabelled data of the target dataset. | ['Paul Lukowicz', 'Dominique Nshimyimana', 'Vitor Fortes Rey'] | 2023-07-03 | null | null | null | null | ['self-supervised-learning', 'activity-recognition', 'human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision', 'time-series'] | [ 7.16237605e-01 2.72091389e-01 -3.52277756e-01 -3.55552673e-01
-5.10713458e-01 -3.98936749e-01 1.04797626e+00 3.56121719e-01
-6.77445233e-01 9.71557021e-01 2.54596770e-01 1.09848902e-01
-4.74781901e-01 -5.17141521e-01 -6.07750237e-01 -5.90620637e-01
-1.02301925e-01 6.17284894e-01 4.48878437e-01 -1.08885430e-01
2.96137422e-01 3.59702021e-01 -1.95730472e+00 6.54160976e-01
3.99589539e-01 8.58493626e-01 1.91741899e-01 6.49941742e-01
-2.38078475e-01 1.04358351e+00 -5.93383074e-01 -4.21733595e-02
3.36574793e-01 -6.59378350e-01 -1.17811549e+00 1.34985179e-01
6.89642876e-02 3.39658797e-01 -1.17547661e-01 3.80532265e-01
4.20228899e-01 2.67406434e-01 7.57201910e-01 -1.15066803e+00
1.94332004e-01 4.12815809e-01 -2.27272436e-01 3.18217158e-01
6.68427825e-01 -8.41637552e-02 6.54571295e-01 -5.95688105e-01
7.35188544e-01 8.93006384e-01 5.57982624e-01 4.31089997e-01
-1.35792720e+00 -2.68758118e-01 -2.87151396e-01 4.12689775e-01
-1.05669785e+00 -6.37297332e-01 5.62472582e-01 -6.54150903e-01
1.23034847e+00 3.63582432e-01 6.06141031e-01 1.20844471e+00
1.78764388e-01 5.26577175e-01 1.45757127e+00 -9.28672493e-01
4.65851337e-01 3.34855229e-01 1.61086842e-01 3.67424309e-01
2.41161197e-01 2.53736556e-01 -7.51984656e-01 -4.78583807e-03
4.79481310e-01 1.14686945e-02 -1.23618450e-02 -3.92624646e-01
-1.17454672e+00 6.72555149e-01 2.14074686e-01 7.97779918e-01
-4.23456579e-01 -6.58957437e-02 8.72341156e-01 5.12605190e-01
4.10658687e-01 5.50224602e-01 -7.73394108e-01 -4.19046164e-01
-1.03324473e+00 -2.78171245e-02 7.82392800e-01 4.71764803e-01
7.88483799e-01 -3.86352912e-02 -1.53053164e-01 7.38971591e-01
-2.07806262e-03 -5.04635572e-02 1.10419321e+00 -7.08017170e-01
3.38848621e-01 8.43678534e-01 -1.53880849e-01 -4.44627017e-01
-4.35079873e-01 -3.05428416e-01 -7.98786044e-01 1.98061898e-01
4.33789998e-01 -2.17157360e-02 -9.98051822e-01 1.50372851e+00
2.07392260e-01 1.15205914e-01 2.07514510e-01 4.57281619e-01
5.64593554e-01 3.49052519e-01 9.42177549e-02 -4.69224364e-01
1.05203342e+00 -8.80652189e-01 -7.27011323e-01 -1.72581196e-01
8.42673063e-01 -5.10721624e-01 9.25073326e-01 7.42093742e-01
-7.91119397e-01 -8.14538658e-01 -1.12642050e+00 7.17243776e-02
-6.72278047e-01 -7.34742209e-02 4.61645424e-01 6.09550059e-01
-9.92823958e-01 8.88689518e-01 -6.19538963e-01 -6.56517804e-01
3.57015312e-01 5.13939083e-01 -6.64270520e-01 8.88925977e-03
-9.55653727e-01 1.03595352e+00 9.42865789e-01 -2.70282239e-01
-7.44804204e-01 -2.22192109e-01 -6.20812178e-01 -4.58162837e-02
3.87805700e-01 -4.04341608e-01 9.38157558e-01 -1.32708764e+00
-1.53467023e+00 9.49920654e-01 2.84458082e-02 -7.67767847e-01
7.80820847e-01 -1.43152669e-01 -3.98970783e-01 -1.61677390e-01
-1.01060361e-01 5.81024349e-01 8.59752476e-01 -8.80687118e-01
-4.57119644e-01 -3.06179285e-01 -3.92091602e-01 1.73430949e-01
5.41321235e-03 -1.04400121e-01 -1.47354826e-01 -3.85651648e-01
-4.22922112e-02 -9.96673405e-01 -7.42819533e-02 -4.42927450e-01
-6.28195256e-02 -2.84111232e-01 7.35708833e-01 -4.33103949e-01
9.98797774e-01 -2.06098819e+00 3.74507904e-01 2.25855425e-01
-1.29820123e-01 4.43797350e-01 2.49757580e-02 5.90260148e-01
-6.35517299e-01 -1.85766771e-01 -2.06295446e-01 -9.93654951e-02
-1.89825922e-01 6.37899816e-01 -5.72411753e-02 4.06507581e-01
5.23559339e-02 7.39890873e-01 -5.37783265e-01 -4.77548212e-01
4.16814595e-01 2.75913090e-01 -8.19184557e-02 2.46904776e-01
-2.08047554e-01 4.74087417e-01 -1.08728833e-01 9.37833041e-02
2.92997897e-01 -8.51885006e-02 2.98000336e-01 1.87846869e-01
-4.33392115e-02 2.45413736e-01 -1.21578324e+00 1.91929221e+00
-4.64202225e-01 6.43819392e-01 -6.18969858e-01 -1.29806936e+00
1.10156333e+00 4.26156342e-01 5.13055205e-01 -8.60165417e-01
7.36908019e-02 6.92503080e-02 -6.34511048e-03 -4.66074198e-01
3.89224946e-01 -2.31111228e-01 -7.92630576e-03 3.06135952e-01
4.83957797e-01 2.78174192e-01 3.69097322e-01 -1.14857599e-01
1.33151448e+00 4.70646918e-01 6.64310634e-01 -2.21516564e-01
6.83060646e-01 -2.09793821e-02 2.99169153e-01 3.93987268e-01
3.56437340e-02 5.62333822e-01 5.49966693e-01 -5.21270752e-01
-1.01551032e+00 -5.12064099e-01 -9.69585869e-03 1.30463445e+00
-3.24444950e-01 -5.79032898e-01 -5.02339840e-01 -9.65881526e-01
-3.14072818e-01 6.90673709e-01 -9.11864460e-01 -3.00302505e-01
-3.97322029e-01 -4.41100568e-01 3.95633012e-01 3.96790057e-01
5.19725978e-01 -1.50176334e+00 -1.13074958e+00 2.44088426e-01
1.03844129e-01 -7.22671807e-01 3.54324371e-01 9.70491230e-01
-1.29811013e+00 -1.19648027e+00 -4.46106970e-01 -4.37683403e-01
3.13179582e-01 -1.02708802e-01 1.04440320e+00 6.89816624e-02
-3.55538517e-01 3.01608384e-01 -5.62427819e-01 -4.77894187e-01
-5.57869077e-01 3.02350760e-01 -1.21290043e-01 -1.18168248e-02
2.50167459e-01 -5.54808438e-01 -7.96820894e-02 1.81193769e-01
-8.93896997e-01 -1.03980377e-01 6.52101636e-01 8.52406740e-01
7.11779594e-01 1.77892104e-01 4.59769994e-01 -1.20217764e+00
3.69707793e-01 -4.40279394e-01 -3.59986335e-01 2.38270462e-01
-9.12434518e-01 5.30224800e-01 4.70503211e-01 -2.91228473e-01
-8.51464927e-01 5.34027517e-01 -9.21446681e-02 -3.99547786e-01
-4.06474710e-01 3.60175371e-01 -5.03516942e-02 7.59142861e-02
1.05064976e+00 3.25180471e-01 -2.31025666e-02 -5.59384823e-01
1.08283266e-01 5.33656657e-01 3.47321600e-01 -1.51181415e-01
3.71897310e-01 1.38138324e-01 1.54948369e-01 -7.99223542e-01
-5.15835822e-01 -6.16926074e-01 -9.72211540e-01 -2.56632864e-01
6.96521938e-01 -6.51785612e-01 -3.22891474e-01 2.79758990e-01
-7.46730983e-01 -5.07692516e-01 -9.04628575e-01 3.91290367e-01
-8.47319782e-01 3.11478376e-01 5.69964480e-03 -9.01432455e-01
-2.78108448e-01 -8.00811887e-01 7.49095321e-01 -1.70338023e-02
-5.48242390e-01 -9.63932157e-01 4.73853678e-01 2.73736387e-01
2.66031474e-01 3.26330394e-01 6.77474201e-01 -1.04566789e+00
4.17355485e-02 -4.33486581e-01 1.84286058e-01 3.95787805e-01
1.84785649e-01 -4.21049565e-01 -1.14886570e+00 -3.15012574e-01
1.50151281e-02 -5.45658946e-01 9.85839665e-01 -1.86975785e-02
1.05696011e+00 -3.38946611e-01 -3.24054718e-01 1.57296017e-01
1.35392988e+00 3.34958881e-02 1.09210634e+00 5.26672900e-01
3.51909727e-01 6.91628635e-01 5.55907726e-01 3.87989402e-01
-7.09758773e-02 8.11323524e-01 2.38108441e-01 5.79780042e-02
-9.16554183e-02 -1.66120321e-01 2.12614715e-01 4.57220584e-01
-2.13712946e-01 -2.02595279e-01 -9.01685596e-01 5.04767120e-01
-1.92009437e+00 -1.07767439e+00 -2.70277590e-01 2.35779357e+00
6.69860899e-01 3.99794340e-01 3.15449804e-01 8.67325425e-01
4.69662875e-01 -1.77449299e-04 -2.70101786e-01 -6.46933615e-01
4.02056500e-02 4.26482141e-01 5.41043460e-01 2.66412675e-01
-1.02113962e+00 5.49108684e-01 6.81781483e+00 7.16974735e-01
-1.08128798e+00 3.47550392e-01 4.35488909e-01 -6.92353994e-02
3.29895318e-01 -3.42805870e-02 -3.93124700e-01 5.77379405e-01
1.48456419e+00 8.65185410e-02 2.77613401e-01 7.36757398e-01
-2.11445056e-02 -7.89720774e-01 -1.26574492e+00 1.05629802e+00
1.23657078e-01 -1.09278989e+00 -1.33312166e-01 2.40613982e-01
5.14887333e-01 -3.92103493e-02 -3.52656126e-01 3.98612380e-01
7.02899024e-02 -1.01552618e+00 5.58400929e-01 8.18164110e-01
7.42911339e-01 -8.22845280e-01 8.63029480e-01 7.67801583e-01
-8.68659496e-01 -2.44659856e-01 -8.84544253e-02 -2.11581275e-01
-1.88330814e-01 3.53992701e-01 -9.50802028e-01 7.19613135e-01
6.11842096e-01 6.84441566e-01 -8.70073617e-01 1.12978482e+00
-1.02500513e-01 7.58364558e-01 -2.38328338e-01 2.02597916e-01
1.34060740e-01 1.12077989e-01 2.15346441e-01 1.41117179e+00
7.07958490e-02 -2.37882569e-01 -3.50425467e-02 1.74829021e-01
3.54573607e-01 1.38684228e-01 -7.15853572e-01 8.94846618e-02
-2.43275911e-02 9.67874408e-01 -7.37164080e-01 -4.96708304e-01
-5.16004264e-02 1.27394390e+00 3.91127437e-01 -4.11980040e-02
-4.68748897e-01 -2.12345675e-01 -1.43602401e-01 5.27126133e-01
4.46047425e-01 1.14390597e-01 -1.11942388e-01 -8.16012204e-01
-1.30601883e-01 -7.24759221e-01 6.48620486e-01 -7.45886147e-01
-8.24037075e-01 6.65563285e-01 3.00304532e-01 -1.09186327e+00
-8.27737689e-01 -3.90983433e-01 -5.20169914e-01 5.74393094e-01
-1.03567982e+00 -8.82760704e-01 -2.93647200e-01 6.67707920e-01
5.82258165e-01 -4.58845705e-01 1.10033953e+00 1.66781709e-01
-4.03305054e-01 2.66532153e-01 1.08110681e-01 -2.48631403e-01
5.36387920e-01 -1.29558265e+00 -2.10580871e-01 6.03738010e-01
4.80538726e-01 2.06331402e-01 7.72428572e-01 -4.21304196e-01
-9.90340412e-01 -9.31556404e-01 1.02127957e+00 -5.34445286e-01
2.41302744e-01 -2.43871883e-01 -1.00832951e+00 5.09989917e-01
4.05839533e-01 1.48294255e-01 7.80465484e-01 6.38743341e-02
-1.88972607e-01 -1.05558485e-01 -1.15572286e+00 -2.79432088e-01
9.37042415e-01 -4.69065368e-01 -8.10038805e-01 2.96888977e-01
1.30535156e-01 -2.28671879e-01 -7.85519958e-01 1.78268358e-01
3.55887055e-01 -1.01771867e+00 6.26350462e-01 -5.99517703e-01
2.47272342e-01 -1.42541334e-01 -3.90309878e-02 -1.12810576e+00
-2.46406883e-01 -5.95634282e-01 -3.19301754e-01 1.12774181e+00
3.73249501e-01 -4.31335151e-01 8.50276470e-01 2.03205705e-01
1.26988903e-01 -3.49921405e-01 -1.27470076e+00 -8.19237351e-01
-1.97908625e-01 -4.15282547e-01 2.39224792e-01 9.74971175e-01
3.52723710e-02 6.95692062e-01 -5.23972154e-01 -6.88405156e-01
2.99930573e-01 -2.38226533e-01 8.12595069e-01 -1.62389016e+00
-4.56070095e-01 2.37181634e-02 -7.62479484e-01 -3.36064428e-01
-1.98830734e-03 -9.95636404e-01 -2.76281275e-02 -1.32873440e+00
1.14367425e-01 -2.91179895e-01 -5.43251038e-01 8.09866846e-01
2.87862808e-01 2.26546481e-01 5.84780350e-02 5.40630758e-01
-8.45176518e-01 2.23652765e-01 8.21162403e-01 8.79422873e-02
-3.78120989e-01 1.94416586e-02 -3.58155549e-01 4.04812187e-01
9.56075490e-01 -7.26512015e-01 -6.01391971e-01 9.98655930e-02
1.32173821e-01 5.35847433e-02 5.05063176e-01 -1.48589325e+00
9.00402591e-02 1.25747785e-01 6.22561395e-01 -6.23127818e-01
3.66200060e-01 -9.91384804e-01 4.80612695e-01 5.66405952e-01
-6.21232986e-01 -3.79710570e-02 3.50047499e-01 8.17495942e-01
-1.27437845e-01 -5.36799729e-01 7.36419201e-01 -2.14857429e-01
-7.34520137e-01 -3.18970770e-01 -4.02005047e-01 -6.99170604e-02
1.20380723e+00 -6.00404918e-01 2.30554432e-01 -2.01142982e-01
-1.08527052e+00 -2.50233620e-01 3.51781905e-01 3.40995222e-01
1.91889882e-01 -1.13959825e+00 -5.84942639e-01 2.27601498e-01
3.75676095e-01 -1.50379002e-01 1.13151692e-01 7.97478914e-01
-2.92131782e-01 5.20795345e-01 -5.60418606e-01 -6.11713171e-01
-1.39679360e+00 6.95077121e-01 3.13581854e-01 -6.24401987e-01
-5.63778341e-01 3.37990105e-01 -4.09339190e-01 -1.44562840e-01
2.33153820e-01 -2.83838362e-01 -4.58753496e-01 4.13479239e-01
4.67886865e-01 5.42556882e-01 4.81173605e-01 -6.80148184e-01
-4.36189055e-01 2.25393802e-01 -1.81857854e-01 -1.58483803e-01
1.54640293e+00 1.63962230e-01 2.65452087e-01 8.32497656e-01
1.07356870e+00 -4.97030467e-01 -1.14101398e+00 -1.86805502e-01
4.08220053e-01 -2.99475342e-01 6.90378025e-02 -1.00267708e+00
-8.67805362e-01 8.23543668e-01 1.13804543e+00 3.58675092e-01
1.13461208e+00 8.43274519e-02 1.00156572e-02 5.08728743e-01
5.51342964e-01 -1.22866416e+00 1.40450403e-01 3.16436470e-01
9.69012022e-01 -1.13969362e+00 2.81682700e-01 -1.23219760e-02
-8.61593246e-01 9.88034189e-01 3.61242384e-01 -1.16378210e-01
3.54171902e-01 8.38145763e-02 -4.28717703e-01 -2.68617511e-01
-8.39004815e-01 -2.30442524e-01 2.82604605e-01 8.50655615e-01
4.80824858e-01 -1.12567239e-01 -4.78977919e-01 3.61923575e-01
5.58041036e-02 4.17451739e-01 4.24499601e-01 1.26175630e+00
-4.17841524e-01 -1.32821870e+00 -3.71525168e-01 6.94484890e-01
-1.33063242e-01 2.70968050e-01 -7.01156437e-01 8.83777142e-01
4.06927675e-01 7.29123175e-01 -2.18126997e-02 -4.23941553e-01
4.94051367e-01 6.40873492e-01 5.61272740e-01 -7.56567180e-01
-9.34205115e-01 -1.04852580e-01 1.64412707e-01 -5.80500484e-01
-7.51228154e-01 -7.92397738e-01 -1.11642015e+00 -1.00751864e-02
-4.76302534e-01 1.64107099e-01 4.81968313e-01 1.05744684e+00
3.82221520e-01 5.94753087e-01 5.01864254e-01 -8.73499215e-01
-2.90921241e-01 -1.26828039e+00 -5.11389911e-01 3.82018328e-01
1.55593225e-04 -8.61013710e-01 -2.38833025e-01 2.49160066e-01] | [9.880827903747559, 3.045367956161499] |
3fccc9d4-5a4a-4165-a0a2-1c2bc96ae32a | learning-from-what-is-already-out-there-few | 2301.03769 | null | https://arxiv.org/abs/2301.03769v1 | https://arxiv.org/pdf/2301.03769v1.pdf | Learning from What is Already Out There: Few-shot Sign Language Recognition with Online Dictionaries | Today's sign language recognition models require large training corpora of laboratory-like videos, whose collection involves an extensive workforce and financial resources. As a result, only a handful of such systems are publicly available, not to mention their limited localization capabilities for less-populated sign languages. Utilizing online text-to-video dictionaries, which inherently hold annotated data of various attributes and sign languages, and training models in a few-shot fashion hence poses a promising path for the democratization of this technology. In this work, we collect and open-source the UWB-SL-Wild few-shot dataset, the first of its kind training resource consisting of dictionary-scraped videos. This dataset represents the actual distribution and characteristics of available online sign language data. We select glosses that directly overlap with the already existing datasets WLASL100 and ASLLVD and share their class mappings to allow for transfer learning experiments. Apart from providing baseline results on a pose-based architecture, we introduce a novel approach to training sign language recognition models in a few-shot scenario, resulting in state-of-the-art results on ASLLVD-Skeleton and ASLLVD-Skeleton-20 datasets with top-1 accuracy of $30.97~\%$ and $95.45~\%$, respectively. | ['Marek Hrúz', 'Matyáš Boháček'] | 2023-01-10 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 1.23301029e-01 -4.17379498e-01 -5.37288129e-01 -4.35862154e-01
-1.07548523e+00 -3.94622117e-01 4.78713393e-01 -7.24825740e-01
-7.42908478e-01 3.75202209e-01 3.41333330e-01 8.24132338e-02
6.43406287e-02 -4.30507272e-01 -6.19354129e-01 -6.00226879e-01
-7.69133791e-02 5.78146696e-01 5.29166222e-01 -3.23090643e-01
4.53864448e-02 5.11427462e-01 -1.93709528e+00 3.37831199e-01
4.34611857e-01 9.93418872e-01 2.84069199e-02 7.14582086e-01
-1.46773085e-01 9.50119674e-01 -2.97892183e-01 -7.02364683e-01
5.35496056e-01 -4.09494340e-01 -4.10198897e-01 8.50136485e-03
1.47647285e+00 -6.30825281e-01 -8.71315658e-01 7.79411733e-01
1.02264893e+00 4.56991605e-02 5.04454136e-01 -1.07049382e+00
-4.61787760e-01 4.07236725e-01 -3.41307342e-01 1.27316236e-01
4.25789446e-01 6.94390118e-01 1.08614874e+00 -9.77348864e-01
1.32805979e+00 8.42288971e-01 7.32703507e-01 1.02427423e+00
-6.97283387e-01 -7.55855501e-01 4.73839156e-02 2.91745543e-01
-1.24336493e+00 -7.05991924e-01 7.82498300e-01 -4.07071710e-01
1.05726564e+00 -1.48291528e-01 1.01968241e+00 1.60005355e+00
-3.84493887e-01 1.11719704e+00 9.91845369e-01 -6.23196840e-01
5.37944064e-02 -3.29688370e-01 -4.15270822e-03 1.04714024e+00
2.88974226e-01 2.38631785e-01 -1.06871319e+00 2.59645671e-01
6.64080024e-01 -2.35394984e-01 -1.11087322e-01 -6.66560650e-01
-1.24426746e+00 4.56722766e-01 1.97657004e-01 3.18275154e-01
-8.63503739e-02 4.70979780e-01 6.00193799e-01 4.71290112e-01
-1.11364305e-01 -1.11896368e-02 -3.57510328e-01 -5.97956717e-01
-9.85852003e-01 2.66924858e-01 7.21653461e-01 1.38047159e+00
4.13836032e-01 2.35406682e-01 8.19295868e-02 8.56335640e-01
3.11603874e-01 8.43817413e-01 6.27842367e-01 -7.25564718e-01
6.71122491e-01 3.55547398e-01 -2.03769296e-01 -4.12811548e-01
-1.92492098e-01 -1.58065945e-01 -3.80225509e-01 3.10922444e-01
9.84777033e-01 1.06458321e-01 -1.49594605e+00 1.63557088e+00
-1.67851206e-02 3.35517526e-01 -9.12783593e-02 1.02323925e+00
9.38374519e-01 9.85129774e-02 1.48606956e-01 2.89245844e-01
1.20557976e+00 -8.84872794e-01 -2.94666260e-01 -1.44598082e-01
6.20677948e-01 -7.15629518e-01 1.48360407e+00 4.75942135e-01
-7.69988894e-01 -3.01197767e-01 -8.04858804e-01 -1.35321677e-01
-4.21365976e-01 3.60624522e-01 9.51056421e-01 8.27713370e-01
-8.21717083e-01 3.40105087e-01 -1.01110578e+00 -8.77933562e-01
7.78939724e-01 7.30134323e-02 -5.26918948e-01 -5.83925605e-01
-6.45337105e-01 8.11626971e-01 -3.50019895e-02 6.13880791e-02
-8.81637573e-01 -7.40288913e-01 -7.98013449e-01 -6.40904546e-01
2.38176137e-01 -2.61826992e-01 1.08825827e+00 -5.05564868e-01
-1.63188505e+00 1.38880706e+00 1.10525578e-01 -3.06495249e-01
1.01200485e+00 -1.82760969e-01 -6.87703848e-01 1.50592208e-01
-1.89625502e-01 6.95604503e-01 8.49702895e-01 -8.90083849e-01
-7.65830100e-01 -1.88060760e-01 -8.82582441e-02 -1.99055269e-01
-3.21282536e-01 3.47822636e-01 -7.90416420e-01 -7.06441700e-01
8.44833106e-02 -8.68380129e-01 -1.96474381e-02 7.38294423e-01
7.26692379e-02 -2.51831971e-02 6.11080110e-01 -7.36230075e-01
9.73317444e-01 -2.04383278e+00 -3.66246421e-03 1.88620031e-01
-1.08107470e-01 5.85809350e-01 -5.03044665e-01 5.26594579e-01
3.00462306e-01 -2.24699020e-01 -2.75409490e-01 -2.96129048e-01
2.20681980e-01 3.90321553e-01 -3.07875693e-01 6.21305466e-01
1.30395204e-01 8.28280807e-01 -9.27640200e-01 -7.43093133e-01
4.29362893e-01 3.86665583e-01 -7.01354086e-01 -9.46127102e-02
-2.89120138e-01 2.27149367e-01 -2.70269752e-01 1.39842021e+00
3.15528929e-01 5.65259531e-02 6.18946180e-02 -3.86630118e-01
-1.13213085e-01 -1.82501003e-01 -1.15921104e+00 2.33515930e+00
-3.87439787e-01 9.15022910e-01 -1.03787966e-01 -7.44657338e-01
7.74978638e-01 1.64734811e-01 7.70540953e-01 -7.28386045e-01
2.92991400e-01 7.46948600e-01 7.42738172e-02 -8.17878604e-01
9.71280038e-02 -1.25428632e-01 -1.70965955e-01 2.96363980e-01
6.84731424e-01 -1.21305563e-01 6.83372557e-01 -1.39293065e-02
1.34325182e+00 3.73999894e-01 1.45032525e-01 2.38058582e-01
2.38155156e-01 8.09937716e-02 2.91168034e-01 7.12572455e-01
-5.57971358e-01 7.69613624e-01 -4.31684181e-02 -4.26360875e-01
-1.13026083e+00 -1.25058925e+00 -2.16522500e-01 1.01660204e+00
-1.59693077e-01 -2.76059866e-01 -3.36866200e-01 -6.15399301e-01
1.00580733e-02 3.63596112e-01 -3.73643160e-01 1.96001381e-01
-9.55404043e-01 -3.11057121e-01 1.01272333e+00 7.27241158e-01
5.76310694e-01 -1.03533101e+00 -9.85950351e-01 -5.42536266e-02
1.69742703e-01 -1.28508568e+00 -5.12984157e-01 -1.28579766e-01
-6.02557361e-01 -1.35962927e+00 -1.15141022e+00 -1.10062540e+00
5.26404142e-01 -2.02883780e-01 8.95552754e-01 -1.09528691e-01
-8.35247159e-01 6.71861827e-01 -6.00514472e-01 -3.74131322e-01
-2.24207595e-01 -1.80414364e-01 7.55109787e-02 -5.47906943e-02
5.54701746e-01 -6.26691043e-01 -4.70021695e-01 2.71229208e-01
-5.33960581e-01 -2.38651589e-01 8.30246270e-01 1.05850530e+00
4.74716812e-01 -7.77238488e-01 1.25356063e-01 -2.83397436e-01
1.10325463e-01 9.69109610e-02 -6.26617432e-01 3.45395386e-01
-1.64735645e-01 6.38783723e-02 3.17478806e-01 -7.43424594e-01
-8.81622672e-01 3.13132226e-01 -3.40448529e-01 -7.43240118e-01
-2.81310767e-01 2.52916962e-01 -5.17883226e-02 -4.43280280e-01
7.62286425e-01 4.71658558e-01 -4.91318107e-03 -7.26376891e-01
5.95269740e-01 6.34092808e-01 9.65865076e-01 -5.36085904e-01
8.60424519e-01 7.33120322e-01 -1.54280305e-01 -1.00048232e+00
-6.06070042e-01 -6.93246365e-01 -8.38294923e-01 -5.78886151e-01
5.97437024e-01 -8.52634430e-01 -4.44138199e-01 8.77949655e-01
-6.82265520e-01 -3.86357129e-01 -7.15157092e-01 7.16593087e-01
-9.11574185e-01 3.15887630e-01 -3.20298731e-01 -5.95781922e-01
-9.31523889e-02 -8.18381190e-01 1.10363209e+00 -6.90414710e-03
1.74184085e-03 -4.87428665e-01 3.13372970e-01 3.83296579e-01
4.29014623e-01 3.48248243e-01 4.65254307e-01 -3.94750684e-01
-7.22265244e-01 -6.12205684e-01 -2.07306415e-01 4.67393637e-01
-7.57980794e-02 -2.57751253e-02 -1.04260325e+00 -3.15171510e-01
-8.17833126e-01 -8.48028123e-01 9.96428549e-01 1.75097138e-01
7.86081672e-01 9.73286703e-02 -2.12049067e-01 9.08022165e-01
1.40686953e+00 -1.16050690e-02 6.61528230e-01 1.23420186e-01
5.80094159e-01 1.87493682e-01 5.39404392e-01 5.28009832e-01
2.58656830e-01 8.60642195e-01 9.79637206e-02 1.95616916e-01
-9.65658784e-01 -7.31088400e-01 4.41791862e-01 1.08577979e+00
-5.86492479e-01 6.69858977e-02 -1.02086580e+00 7.38143504e-01
-1.46011615e+00 -1.19186962e+00 2.85709411e-01 2.00718975e+00
8.19506109e-01 -1.17989920e-01 3.37179422e-01 1.39738843e-01
2.77193904e-01 3.61817777e-01 -6.13467813e-01 2.16125831e-01
-4.51140612e-01 5.91417789e-01 6.63415849e-01 5.78266382e-02
-1.19623029e+00 1.20842564e+00 6.26026058e+00 7.35461950e-01
-1.35657001e+00 1.45179003e-01 -2.94174820e-01 -4.79131073e-01
2.55925626e-01 -2.06684336e-01 -8.59894812e-01 3.49598795e-01
6.89606428e-01 6.78077713e-02 4.02120650e-01 1.01148152e+00
-6.77421615e-02 1.12170324e-01 -1.04242718e+00 1.36166835e+00
4.89902079e-01 -1.37501061e+00 1.10810801e-01 -1.51636656e-02
6.03599370e-01 6.36051714e-01 -1.17521726e-01 2.63091743e-01
2.87760586e-01 -7.66308308e-01 9.69653606e-01 6.46444559e-01
1.50845492e+00 -1.45656198e-01 4.94916886e-01 1.25303105e-01
-1.44844794e+00 -1.74712151e-01 -2.03965500e-01 2.78953999e-01
5.26833296e-01 -1.03435084e-01 -4.00181085e-01 2.13973328e-01
7.36942828e-01 1.06053436e+00 -4.97323275e-01 1.36771905e+00
-2.46330827e-01 6.57783926e-01 -6.79855585e-01 -2.76669562e-01
1.14641033e-01 8.84739310e-02 5.31824410e-01 1.26779521e+00
4.35035408e-01 -3.71943824e-02 1.54256403e-01 3.46767098e-01
-1.66781515e-01 9.17578042e-02 -6.33773565e-01 -2.61363566e-01
2.66565323e-01 5.97233474e-01 -4.67861563e-01 -3.01668048e-01
-6.97340786e-01 9.68777537e-01 9.24773514e-02 1.16562940e-01
-8.20754409e-01 -3.97018880e-01 8.52455080e-01 7.76368380e-02
4.47025836e-01 -5.35867095e-01 1.40270412e-01 -1.58921266e+00
3.49192351e-01 -1.03188884e+00 4.74507213e-01 -5.07780194e-01
-1.44300258e+00 3.57430577e-01 -4.61815931e-02 -1.73931265e+00
-3.13044280e-01 -1.18903983e+00 -3.22324365e-01 3.64157885e-01
-1.67575574e+00 -1.75034070e+00 -6.40051007e-01 8.22439134e-01
6.69959843e-01 -6.84133530e-01 8.65392864e-01 8.36757600e-01
-1.75442636e-01 9.87100840e-01 9.12495703e-02 6.30538821e-01
9.32651281e-01 -7.67681062e-01 3.50268483e-01 9.44800138e-01
7.97752619e-01 3.30242485e-01 4.08780694e-01 -4.04697806e-01
-1.80137181e+00 -7.66275287e-01 7.38385201e-01 -6.13563538e-01
9.26585257e-01 -3.28277797e-01 -3.18295002e-01 6.08848989e-01
-4.26878691e-01 5.51651657e-01 4.17364866e-01 -2.25488022e-01
-6.86189950e-01 -2.28188291e-01 -9.65284050e-01 6.09918833e-01
2.07118011e+00 -7.05539763e-01 -6.13663077e-01 3.66992354e-01
1.47016570e-02 -6.35893643e-01 -7.83999085e-01 2.25810096e-01
1.24766886e+00 -8.01025391e-01 1.01407337e+00 -8.35847199e-01
3.22836876e-01 -2.06742212e-01 -5.44147551e-01 -8.62543583e-01
3.13081205e-01 -4.74639565e-01 -2.82853305e-01 9.20092762e-01
2.71951556e-01 -3.46396267e-01 1.16526210e+00 2.98850387e-01
-2.04914272e-01 -6.06714666e-01 -1.29899228e+00 -1.22228587e+00
-1.74316213e-01 -8.75521541e-01 1.76940575e-01 6.47495389e-01
-1.79724485e-01 -4.47077394e-01 -6.10566795e-01 -3.24205548e-01
9.92110610e-01 1.21959560e-01 1.16117442e+00 -9.38110709e-01
-3.00423056e-01 -5.90200186e-01 -1.03453028e+00 -1.08392680e+00
8.33891854e-02 -9.67212975e-01 9.59278643e-02 -1.39292216e+00
-1.77823231e-01 -3.52840334e-01 -1.37974620e-01 6.67624295e-01
6.04151130e-01 5.51674843e-01 3.16567570e-01 2.13989779e-01
-4.80066121e-01 3.86371762e-01 1.13036668e+00 -3.31387848e-01
9.31672752e-02 -2.23643988e-01 1.43334260e-02 8.27736080e-01
3.50482792e-01 -1.95375949e-01 -1.63202256e-01 -5.31749189e-01
-3.03522170e-01 -3.71326417e-01 5.33232868e-01 -1.40402985e+00
3.83072138e-01 -9.59600061e-02 -1.38684660e-01 -5.14531493e-01
6.83502913e-01 -6.93648160e-01 -1.69309542e-01 5.39509773e-01
-1.99906275e-01 -4.04913098e-01 -4.15971130e-02 5.30854642e-01
-3.46009791e-01 7.08240643e-02 9.56799865e-01 -2.16888919e-01
-1.65043771e+00 5.99003971e-01 -1.37122916e-02 4.04099852e-01
8.75118971e-01 -5.76004505e-01 -3.00584555e-01 -2.63064265e-01
-5.91415882e-01 -4.56357487e-02 3.81076545e-01 6.95399940e-01
7.54426181e-01 -1.20010853e+00 -7.01613367e-01 4.31083113e-01
7.54394412e-01 -4.02362883e-01 2.45461375e-01 6.41608655e-01
-8.38492990e-01 3.74461442e-01 -7.02907741e-01 -5.72551310e-01
-1.35226917e+00 1.32514825e-02 3.15515667e-01 6.42020553e-02
-1.14854324e+00 1.11947203e+00 -7.40182042e-01 -4.32912469e-01
5.07248342e-01 -5.48811853e-01 2.91938782e-01 -1.55253056e-02
4.76485759e-01 3.33623499e-01 -6.66122213e-02 -8.26425552e-01
-4.93862480e-01 9.85798180e-01 3.37733150e-01 -2.40965188e-01
1.45777392e+00 5.01927316e-01 4.68149304e-01 3.83880794e-01
1.05060339e+00 3.42571512e-02 -1.26600409e+00 -3.32054943e-01
1.03443675e-02 -8.12466323e-01 -2.21743181e-01 -9.87768888e-01
-1.11740911e+00 9.86392140e-01 9.16351199e-01 -8.06894720e-01
1.00947547e+00 2.63725489e-01 8.26303840e-01 6.31318152e-01
1.04602122e+00 -1.29710710e+00 1.65074944e-01 4.96132523e-01
1.00660110e+00 -1.35277224e+00 -1.10012561e-01 -1.28133848e-01
-4.71768886e-01 9.89144325e-01 4.53124374e-01 -2.40813583e-01
7.15590537e-01 3.05411637e-01 3.83738160e-01 -1.64097488e-01
-3.21078598e-01 -6.21519744e-01 3.05323958e-01 9.64664102e-01
1.90375462e-01 -7.53196925e-02 -3.43423605e-01 4.63817388e-01
-2.80140936e-01 5.43569505e-01 9.33850482e-02 1.15276504e+00
-1.85609490e-01 -1.27091551e+00 -4.57473807e-02 5.58049679e-01
1.23797394e-01 4.10678573e-02 -2.84841657e-01 1.12852204e+00
2.68934339e-01 3.74941826e-01 -1.43891454e-01 -3.41852009e-01
6.92657828e-01 3.23997498e-01 9.67306614e-01 -3.93935919e-01
-1.59372985e-01 -2.68664420e-01 4.30393785e-01 -8.02489102e-01
-6.05432510e-01 -8.90369177e-01 -1.13907301e+00 -7.66472891e-02
-1.01142138e-01 -7.12064266e-01 6.60388350e-01 8.42217147e-01
5.66097982e-02 1.45669878e-01 -1.20501012e-01 -1.06669080e+00
-8.75455439e-01 -9.02986109e-01 -6.94272399e-01 7.00231791e-01
9.23252329e-02 -9.85471547e-01 -2.37098224e-02 2.61154354e-01] | [9.146193504333496, -6.467928409576416] |
bf1acdf1-6633-45e7-b47e-3988176ea50a | prix-lm-pretraining-for-multilingual | 2110.08443 | null | https://arxiv.org/abs/2110.08443v2 | https://arxiv.org/pdf/2110.08443v2.pdf | Prix-LM: Pretraining for Multilingual Knowledge Base Construction | Knowledge bases (KBs) contain plenty of structured world and commonsense knowledge. As such, they often complement distributional text-based information and facilitate various downstream tasks. Since their manual construction is resource- and time-intensive, recent efforts have tried leveraging large pretrained language models (PLMs) to generate additional monolingual knowledge facts for KBs. However, such methods have not been attempted for building and enriching multilingual KBs. Besides wider application, such multilingual KBs can provide richer combined knowledge than monolingual (e.g., English) KBs. Knowledge expressed in different languages may be complementary and unequally distributed: this implies that the knowledge available in high-resource languages can be transferred to low-resource ones. To achieve this, it is crucial to represent multilingual knowledge in a shared/unified space. To this end, we propose a unified representation model, Prix-LM, for multilingual KB construction and completion. We leverage two types of knowledge, monolingual triples and cross-lingual links, extracted from existing multilingual KBs, and tune a multilingual language encoder XLM-R via a causal language modeling objective. Prix-LM integrates useful multilingual and KB-based factual knowledge into a single model. Experiments on standard entity-related tasks, such as link prediction in multiple languages, cross-lingual entity linking and bilingual lexicon induction, demonstrate its effectiveness, with gains reported over strong task-specialised baselines. | ['Muhao Chen', 'Nigel Collier', 'Ivan Vulić', 'Fangyu Liu', 'Wenxuan Zhou'] | 2021-10-16 | null | https://aclanthology.org/2022.acl-long.371 | https://aclanthology.org/2022.acl-long.371.pdf | acl-2022-5 | ['cross-lingual-entity-linking'] | ['natural-language-processing'] | [-6.20127320e-01 2.08124325e-01 -9.40706074e-01 -2.10796788e-01
-1.05943680e+00 -7.17959523e-01 6.07758045e-01 3.92465174e-01
-6.20166421e-01 1.51780760e+00 6.56196654e-01 -4.57236916e-01
1.39848337e-01 -9.05861437e-01 -1.14106572e+00 -2.02583708e-02
2.19449192e-01 6.51587844e-01 1.06323116e-01 -6.46346092e-01
-4.28300768e-01 2.96008829e-02 -8.13531339e-01 4.05694515e-01
1.23461914e+00 4.67305750e-01 2.99449772e-01 -4.01316509e-02
-5.60366094e-01 1.18043327e+00 -3.92106563e-01 -1.22316790e+00
-1.65227607e-01 -2.21790433e-01 -1.10949087e+00 -6.67589068e-01
1.14543460e-01 8.71729553e-02 -3.09464544e-01 8.94638300e-01
4.22812343e-01 -1.75514802e-01 6.22647643e-01 -9.98374224e-01
-1.13194394e+00 1.62678516e+00 -4.66761887e-01 1.85853288e-01
4.33415681e-01 -1.91088051e-01 1.33745611e+00 -9.16379094e-01
1.00633907e+00 1.33102989e+00 7.20282018e-01 2.00569764e-01
-1.20122826e+00 -9.24063027e-01 2.07256183e-01 5.63812315e-01
-1.60366488e+00 -3.75660449e-01 6.03412867e-01 -3.51296276e-01
1.33160770e+00 -1.60513893e-01 5.19882083e-01 1.37457502e+00
-1.27594054e-01 9.12082374e-01 1.03291464e+00 -6.70360744e-01
-5.56344390e-01 4.62593615e-01 1.27596781e-01 7.22959101e-01
6.11543357e-01 -2.21664310e-01 -7.26630509e-01 1.40080690e-01
5.10882020e-01 -5.86879611e-01 -3.65715116e-01 4.08577360e-02
-1.52557480e+00 7.40847170e-01 4.19034272e-01 5.50192058e-01
-4.08889830e-01 2.21121665e-02 5.77285945e-01 2.62434453e-01
4.70553219e-01 5.10643661e-01 -9.79638040e-01 1.66106164e-01
-5.52516878e-01 -1.14848921e-02 8.82354856e-01 1.29393756e+00
1.00073266e+00 -2.40327135e-01 1.25638610e-02 1.15647304e+00
2.57985950e-01 7.50589073e-01 6.28175914e-01 -3.07660222e-01
9.63331461e-01 6.18221045e-01 -6.79706633e-02 -8.60563099e-01
-3.39887589e-01 -2.93917298e-01 -7.80741692e-01 -7.79697359e-01
1.08901046e-01 -2.64221549e-01 -6.05740666e-01 2.03238177e+00
3.28940153e-01 1.18043326e-01 6.92518473e-01 5.39154649e-01
1.08089805e+00 6.20333731e-01 4.70834017e-01 -1.31390259e-01
1.50404358e+00 -8.04425657e-01 -8.85803640e-01 -2.70494103e-01
9.78910625e-01 -7.30476022e-01 1.11166537e+00 -7.12209046e-02
-7.61817276e-01 -3.17682087e-01 -9.02229965e-01 -4.45258826e-01
-9.27747190e-01 2.37524047e-01 7.96057761e-01 2.91740388e-01
-6.36887729e-01 6.79411143e-02 -4.62014794e-01 -2.22849622e-01
1.54985711e-01 -1.18345968e-01 -6.25590086e-01 -4.08219844e-01
-2.25629115e+00 1.39103699e+00 1.22136319e+00 1.35630861e-01
-5.93527615e-01 -1.03576314e+00 -1.19920945e+00 -2.92323917e-01
7.01507211e-01 -8.95319462e-01 8.75317037e-01 -5.05740821e-01
-1.13361990e+00 8.76653314e-01 4.97781225e-02 -3.58933002e-01
3.29125486e-02 -4.20312762e-01 -8.90051901e-01 -3.53718877e-01
5.52806914e-01 5.47829628e-01 1.63982317e-01 -1.29340804e+00
-7.59060919e-01 -8.80121663e-02 1.96924627e-01 3.65882128e-01
-4.67564136e-01 1.74659759e-01 -7.04168916e-01 -7.75763690e-01
-5.47841012e-01 -6.90369546e-01 4.29413840e-02 -9.44461763e-01
-7.89931774e-01 -2.41739452e-01 2.60460794e-01 -1.06680655e+00
1.46577644e+00 -1.51976430e+00 2.28296787e-01 1.70159176e-01
-1.19563036e-01 3.75201851e-01 -1.45760357e-01 4.81646538e-01
-6.16905279e-02 2.66403556e-01 4.75362204e-02 5.51117538e-03
1.60503775e-01 7.12656021e-01 -5.05653560e-01 -1.65471643e-01
4.14523751e-01 1.54836929e+00 -1.28642654e+00 -7.83048093e-01
-1.13801360e-01 4.71731216e-01 -4.07305092e-01 -8.27002823e-02
-2.68843442e-01 4.75594401e-01 -3.47192287e-01 6.40050948e-01
8.49655364e-04 -1.51930124e-01 6.24309897e-01 -6.60713494e-01
9.95973274e-02 8.05452108e-01 -9.81133103e-01 1.82011712e+00
-1.00882471e+00 2.49768272e-01 -5.47482491e-01 -7.48196006e-01
5.65374076e-01 4.94986057e-01 1.07644714e-01 -7.59450138e-01
-1.27971441e-01 5.24989843e-01 -1.53184578e-01 -5.49233913e-01
6.86323941e-01 -5.69312453e-01 -4.28809285e-01 1.99824676e-01
5.58975697e-01 -2.96115084e-03 4.94371325e-01 5.05228162e-01
7.79956400e-01 2.81036139e-01 7.76521385e-01 -4.81520221e-02
6.23313189e-01 2.16290519e-01 6.46930456e-01 2.22151175e-01
4.14157540e-01 -3.57245654e-01 2.62123495e-01 -2.60061286e-02
-8.22031558e-01 -9.46354926e-01 -2.84842193e-01 1.10427892e+00
6.49911389e-02 -7.52046287e-01 -7.69121423e-02 -9.76229608e-01
2.08561033e-01 9.06323850e-01 -4.55071926e-01 -8.33637044e-02
-7.15218246e-01 -9.63306367e-01 1.10566258e+00 6.24255717e-01
2.63935059e-01 -1.02890432e+00 2.85503477e-01 5.59283495e-01
-7.03243971e-01 -1.63875210e+00 -1.97981700e-01 2.19114229e-01
-2.53522456e-01 -1.17521441e+00 -3.58012885e-01 -6.06457591e-01
3.35937470e-01 -1.31007031e-01 1.72686422e+00 -4.73333180e-01
4.29655053e-03 1.98718429e-01 -4.98726279e-01 -3.69655013e-01
-5.48522651e-01 2.69736111e-01 3.42285097e-01 -3.68170917e-01
5.90510190e-01 -4.50227231e-01 6.98011518e-02 1.08406786e-02
-8.90927851e-01 1.33561447e-01 7.18390584e-01 9.90939617e-01
7.68387973e-01 1.30183965e-01 1.05397987e+00 -1.08139670e+00
6.97996497e-01 -7.69635379e-01 -3.70040029e-01 6.72432065e-01
-3.75197977e-01 4.16522264e-01 7.07053483e-01 -4.00321990e-01
-1.41053081e+00 -4.62007821e-01 -5.28686084e-02 -4.35162298e-02
2.09242508e-01 1.40738308e+00 -6.04519486e-01 2.94863760e-01
6.92588389e-01 3.31319943e-02 -7.12541044e-01 -4.74770963e-01
1.27292502e+00 4.75941420e-01 6.31884277e-01 -1.21263969e+00
6.93705857e-01 -9.97260734e-02 -4.26389396e-01 -4.88764703e-01
-1.33294570e+00 -4.69252229e-01 -8.77333820e-01 1.52882695e-01
7.65728951e-01 -1.43039751e+00 -1.74736619e-01 1.29954606e-01
-1.32219577e+00 -3.68830860e-01 -2.10606307e-01 5.62896371e-01
-6.58765510e-02 6.75873309e-02 -7.10150361e-01 -1.80097386e-01
-1.78606421e-01 -7.35570133e-01 8.96751404e-01 -9.84727964e-02
-1.97397962e-01 -1.53431153e+00 4.21340853e-01 4.55376446e-01
-1.28907198e-02 -8.79950821e-03 1.28045952e+00 -6.74063563e-01
-3.41110229e-01 2.84578279e-02 -3.99909079e-01 3.98629278e-01
2.74008632e-01 -2.73876935e-01 -5.54301620e-01 4.39405143e-02
-8.22279096e-01 -9.82094646e-01 7.62334943e-01 -1.26663268e-01
4.76711363e-01 -4.74808842e-01 -5.17861724e-01 3.49877030e-01
1.46593618e+00 -2.29695484e-01 3.06910962e-01 4.09557939e-01
1.17477798e+00 5.57977080e-01 4.90713030e-01 -1.56358108e-02
1.34457457e+00 6.99962854e-01 -1.91610992e-01 -1.41225487e-01
-4.15183127e-01 -6.71727836e-01 6.29786730e-01 1.58865190e+00
-1.81798831e-01 -4.33867844e-03 -1.30835128e+00 9.11063612e-01
-1.80350423e+00 -8.43536317e-01 -1.99774012e-01 1.89368582e+00
1.94619179e+00 -4.32662889e-02 -2.38299578e-01 -4.34330612e-01
4.70025390e-01 -2.04292580e-01 -3.10555726e-01 1.61989689e-01
-7.34316170e-01 1.82570249e-01 5.26448846e-01 6.07019186e-01
-9.54840958e-01 1.71947110e+00 4.92325115e+00 1.21275389e+00
-9.33458984e-01 4.75057781e-01 1.45366386e-01 1.96138084e-01
-6.84908509e-01 1.77940741e-01 -1.33364701e+00 3.26678485e-01
8.74415040e-01 -4.84180719e-01 1.30939648e-01 5.59183121e-01
-3.34928542e-01 1.10694997e-01 -1.00168788e+00 9.03026819e-01
2.67621465e-02 -1.39148831e+00 3.54120642e-01 -2.04441622e-01
9.26625490e-01 3.86907935e-01 -3.39274287e-01 1.04081368e+00
1.10551262e+00 -9.98536468e-01 6.65523708e-01 4.25171345e-01
1.06871283e+00 -7.76466668e-01 8.42751324e-01 2.54114658e-01
-1.39725518e+00 3.92706335e-01 -4.32402819e-01 4.09183174e-01
5.95613778e-01 9.63756621e-01 -8.25895488e-01 1.30202758e+00
4.86430258e-01 9.36203480e-01 -6.05670869e-01 4.19915944e-01
-9.31766331e-01 5.77217400e-01 -2.17408940e-01 2.49499351e-01
2.16078758e-01 1.50165424e-01 3.06558579e-01 1.70181549e+00
1.26523310e-02 5.97788431e-02 3.26697379e-01 7.43315220e-01
-6.18540347e-01 6.42737925e-01 -7.00778127e-01 -2.89104134e-01
6.27314925e-01 1.25243235e+00 -8.01483542e-02 -6.22326732e-01
-7.34529376e-01 8.07751417e-01 1.03391230e+00 3.88687193e-01
-8.35409939e-01 -2.32098922e-01 4.78120923e-01 -3.95989865e-01
-8.75935480e-02 -1.72370449e-01 1.56982228e-01 -1.76148868e+00
3.63908410e-02 -8.51313710e-01 6.06398046e-01 -6.10480428e-01
-1.83611715e+00 6.25107765e-01 1.60214648e-01 -7.44254351e-01
-4.56940711e-01 -6.00048006e-01 2.31861964e-01 1.13763332e+00
-2.17921638e+00 -1.92339742e+00 2.66838163e-01 7.72588253e-01
1.19678520e-01 -3.08811933e-01 8.44382405e-01 7.08318293e-01
-6.39106393e-01 7.23989248e-01 -2.40965635e-01 5.19515038e-01
1.17022049e+00 -1.34064794e+00 7.00670257e-02 8.30336690e-01
5.12711823e-01 9.93659079e-01 2.27519527e-01 -1.12513793e+00
-1.20817566e+00 -1.47255313e+00 1.38330972e+00 -7.87916601e-01
1.38790619e+00 -1.74294189e-01 -1.17018294e+00 1.11278391e+00
2.72151977e-01 -4.33552116e-02 1.17534029e+00 9.49102998e-01
-8.99685204e-01 2.13857368e-02 -3.72501135e-01 7.35836506e-01
9.76512313e-01 -9.25154865e-01 -9.45952892e-01 2.87331700e-01
8.36899579e-01 -3.36589009e-01 -1.40209734e+00 5.85850954e-01
3.17826718e-01 -2.24167421e-01 1.08162093e+00 -1.06835735e+00
6.04736567e-01 -3.55542183e-01 -2.86302477e-01 -1.82202375e+00
-1.08200334e-01 -2.17814103e-01 -2.68456608e-01 1.71333301e+00
9.03855324e-01 -5.20572603e-01 1.73618887e-02 2.13570654e-01
-1.39013454e-01 -5.65851748e-01 -7.45924771e-01 -9.89794314e-01
3.97910655e-01 -6.93577349e-01 5.75366557e-01 1.73078859e+00
5.04742801e-01 1.11528635e+00 -4.29548413e-01 3.55267406e-01
3.42981309e-01 2.55104080e-02 6.87788188e-01 -1.06949687e+00
-1.80144385e-01 -4.12118852e-01 -1.98904276e-01 -7.18607008e-01
1.00076425e+00 -1.56018150e+00 -1.72867224e-01 -1.56193411e+00
4.12000507e-01 -1.01774013e+00 -3.80242556e-01 1.18076277e+00
-7.48131573e-01 1.37681499e-01 -1.64754540e-01 1.23092070e-01
-6.43398762e-01 7.94592083e-01 1.06292105e+00 -9.66608450e-02
-1.17504410e-01 -8.89683306e-01 -8.52174461e-01 7.01100290e-01
4.90082890e-01 -3.98755610e-01 -4.00542319e-01 -6.77772403e-01
9.83230650e-01 -5.01403548e-02 1.80702254e-01 -3.55806112e-01
3.09006006e-01 -2.11609662e-01 1.56764537e-01 -2.66031355e-01
-2.66300775e-02 -5.89791596e-01 1.48193479e-01 -1.44435436e-01
-2.41698816e-01 -5.96082769e-02 4.58735108e-01 4.73890960e-01
-6.09225452e-01 -9.13591757e-02 2.50997812e-01 -1.53413624e-01
-1.00232172e+00 1.92973107e-01 2.33063832e-01 6.51437104e-01
7.49036133e-01 4.28351164e-01 -6.17450356e-01 -1.41358534e-02
-5.28433800e-01 4.48774934e-01 1.66970029e-01 6.39983714e-01
2.02084377e-01 -1.72013068e+00 -9.98788774e-01 -2.53068238e-01
6.39369369e-01 -6.31198585e-02 -1.40289739e-02 9.06556666e-01
1.73949432e-02 8.00316393e-01 -3.96460947e-03 -4.68393713e-02
-6.89725339e-01 6.87417626e-01 1.02761574e-01 -7.61652350e-01
-3.51306349e-01 7.55973101e-01 8.60930458e-02 -8.62649560e-01
-2.27282286e-01 -1.64333001e-01 -3.75382513e-01 3.01166385e-01
3.68566424e-01 -2.89618932e-02 1.41723260e-01 -8.82927716e-01
-3.73241156e-01 2.64095962e-01 -1.06351383e-01 -1.83343515e-01
1.12955368e+00 -2.84999818e-01 -5.70815206e-01 6.37762308e-01
8.34519267e-01 6.88669443e-01 -4.23126519e-01 -7.86773980e-01
4.27163392e-01 4.70698997e-02 -4.32898365e-02 -1.19823205e+00
-8.12921584e-01 5.75905561e-01 -5.03460526e-01 -3.28301847e-01
8.05768132e-01 4.01985019e-01 8.49634528e-01 5.11823535e-01
7.58315861e-01 -1.03617430e+00 -2.47909293e-01 9.89900410e-01
8.76255631e-01 -1.38651931e+00 -1.50681660e-01 -5.44203341e-01
-8.51935446e-01 6.61447346e-01 6.90462291e-01 5.34337699e-01
5.89689255e-01 3.73555869e-01 1.69907555e-01 -2.39562467e-01
-7.50168085e-01 -6.29834950e-01 6.98126912e-01 5.49091399e-01
7.89093852e-01 3.93132687e-01 -3.72280866e-01 1.17947483e+00
-4.13253933e-01 -9.96982828e-02 6.69917241e-02 4.90980595e-01
8.99431035e-02 -1.51528311e+00 -6.05863072e-02 2.40654513e-01
-5.16376555e-01 -8.64517868e-01 -1.50530845e-01 9.24435675e-01
5.31363189e-01 6.25380933e-01 -5.74716032e-01 -2.81807929e-01
2.69106567e-01 3.70331585e-01 5.41954756e-01 -7.90339530e-01
-3.35898280e-01 -3.11356187e-01 6.75909996e-01 -3.12078834e-01
-6.74956620e-01 -3.62049758e-01 -1.17910933e+00 -8.21892396e-02
-3.19321036e-01 1.55303597e-01 3.95944893e-01 1.18966305e+00
2.18892470e-01 6.59570694e-01 -9.87274572e-02 -2.15162575e-01
7.11192116e-02 -8.92708480e-01 -3.20357323e-01 3.15547228e-01
-2.32747674e-01 -7.80507803e-01 2.67189682e-01 2.85140067e-01] | [9.541332244873047, 8.808162689208984] |
141a7dda-ff71-42ab-9f01-eedba263bec4 | multi-label-meta-weighting-for-long-tailed | 2306.10122 | null | https://arxiv.org/abs/2306.10122v1 | https://arxiv.org/pdf/2306.10122v1.pdf | Multi-Label Meta Weighting for Long-Tailed Dynamic Scene Graph Generation | This paper investigates the problem of scene graph generation in videos with the aim of capturing semantic relations between subjects and objects in the form of $\langle$subject, predicate, object$\rangle$ triplets. Recognizing the predicate between subject and object pairs is imbalanced and multi-label in nature, ranging from ubiquitous interactions such as spatial relationships (\eg \emph{in front of}) to rare interactions such as \emph{twisting}. In widely-used benchmarks such as Action Genome and VidOR, the imbalance ratio between the most and least frequent predicates reaches 3,218 and 3,408, respectively, surpassing even benchmarks specifically designed for long-tailed recognition. Due to the long-tailed distributions and label co-occurrences, recent state-of-the-art methods predominantly focus on the most frequently occurring predicate classes, ignoring those in the long tail. In this paper, we analyze the limitations of current approaches for scene graph generation in videos and identify a one-to-one correspondence between predicate frequency and recall performance. To make the step towards unbiased scene graph generation in videos, we introduce a multi-label meta-learning framework to deal with the biased predicate distribution. Our meta-learning framework learns a meta-weight network for each training sample over all possible label losses. We evaluate our approach on the Action Genome and VidOR benchmarks by building upon two current state-of-the-art methods for each benchmark. The experiments demonstrate that the multi-label meta-weight network improves the performance for predicates in the long tail without compromising performance for head classes, resulting in better overall performance and favorable generalizability. Code: \url{https://github.com/shanshuo/ML-MWN}. | ['Cees G. M. Snoek', 'Pascal Mettes', 'Yingjun Du', 'Shuo Chen'] | 2023-06-16 | null | null | null | null | ['scene-graph-generation', 'unbiased-scene-graph-generation', 'meta-learning'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 4.62854683e-01 8.24774243e-03 -4.09177810e-01 -4.48494375e-01
-8.69191468e-01 -5.70624173e-01 5.05696177e-01 5.77605255e-02
-2.44765967e-01 6.72145188e-01 3.30609888e-01 -5.90898842e-02
-2.32349232e-01 -7.91157544e-01 -1.18492353e+00 -7.12014019e-01
-1.83902025e-01 5.91764987e-01 2.23854110e-01 -9.97711495e-02
-7.77327642e-02 -2.07133647e-02 -2.01213288e+00 7.77666986e-01
6.05208099e-01 1.10204327e+00 2.08323486e-02 5.40766001e-01
-1.81697179e-02 1.00532615e+00 -6.35396719e-01 -8.12456369e-01
1.57403067e-01 -5.38115799e-01 -7.98107505e-01 2.10778221e-01
1.02369034e+00 -1.24158233e-01 -5.84599674e-01 1.05810225e+00
6.90538764e-01 1.59743518e-01 8.07566643e-01 -1.59536874e+00
-4.47433054e-01 7.16957867e-01 -7.72390902e-01 2.73735583e-01
5.44366598e-01 1.15940645e-01 1.53377461e+00 -7.41753161e-01
7.62524843e-01 1.21298325e+00 6.36069357e-01 4.09388900e-01
-9.49916899e-01 -7.83525944e-01 5.02091885e-01 5.34890532e-01
-1.40988123e+00 -1.57994807e-01 8.26586604e-01 -4.76757258e-01
8.81679296e-01 4.74341571e-01 6.38529778e-01 1.49540377e+00
8.88414029e-03 9.54931498e-01 9.89761353e-01 -2.01580048e-01
-5.54741099e-02 -1.59250170e-01 -3.46034975e-03 7.49985754e-01
4.00049597e-01 -1.77719906e-01 -8.84097099e-01 -7.42298886e-02
4.51355994e-01 -1.81444973e-01 -3.17731321e-01 -4.30343628e-01
-1.32727444e+00 6.18511856e-01 2.48519436e-01 9.54178944e-02
-1.25198200e-01 4.75341499e-01 4.39906478e-01 -7.39587024e-02
5.37325859e-01 2.12444723e-01 -5.31352043e-01 -2.32688338e-01
-7.19791472e-01 6.61383927e-01 6.96690679e-01 1.25629807e+00
6.95710361e-01 -3.72829527e-01 -6.33853078e-01 9.24426079e-01
1.09322399e-01 3.36313158e-01 3.20233017e-01 -9.14935887e-01
7.82148719e-01 6.88095033e-01 -3.12157348e-03 -1.15166783e+00
-5.11171401e-01 -5.33992290e-01 -6.17896914e-01 -1.42504811e-01
6.00527346e-01 6.77500963e-02 -7.67638147e-01 2.03237629e+00
4.67402041e-01 1.66946307e-01 -2.08965391e-01 7.42543221e-01
1.29749668e+00 4.80424404e-01 4.26871926e-01 -1.09110855e-01
1.64294875e+00 -1.13056862e+00 -3.50098550e-01 -2.58935422e-01
7.65289485e-01 -6.44594312e-01 1.34897399e+00 1.96572766e-01
-9.42787707e-01 -5.35801768e-01 -6.05554640e-01 -1.37477741e-01
-4.02772516e-01 3.75360698e-02 6.70632362e-01 3.39069337e-01
-7.18668103e-01 4.04223561e-01 -2.78866857e-01 -3.02648276e-01
6.83115721e-01 2.47420728e-01 -3.85783762e-01 -3.70849341e-01
-1.18140626e+00 4.89446342e-01 5.01423717e-01 -2.89796054e-01
-8.64177763e-01 -1.00620079e+00 -9.13100600e-01 1.04115065e-02
7.50968575e-01 -7.78661966e-01 1.04595554e+00 -1.02131104e+00
-7.97081649e-01 1.27169871e+00 -1.13183118e-01 -1.83483720e-01
6.64281547e-01 -1.50350526e-01 -4.08179104e-01 7.64733180e-02
2.35727474e-01 8.70184660e-01 6.69732809e-01 -1.27255034e+00
-8.93949151e-01 -2.07122445e-01 4.09572005e-01 6.18606806e-01
-2.49516547e-01 -1.49588525e-01 -7.09181488e-01 -8.41196358e-01
-7.20049022e-03 -9.91021395e-01 2.64587879e-01 -3.34880203e-01
-7.07565069e-01 -4.61173385e-01 5.96987844e-01 -3.51065457e-01
1.25477350e+00 -2.12395239e+00 1.09359846e-01 -1.42981842e-01
2.28148654e-01 -1.00798011e-01 -1.63825408e-01 4.80436027e-01
-4.04727519e-01 -6.82964996e-02 6.49075434e-02 -2.05058411e-01
1.23135395e-01 1.80654481e-01 -1.81342691e-01 5.28923094e-01
-4.42768112e-02 9.56255436e-01 -1.11938655e+00 -6.36228859e-01
8.15993473e-02 3.39097053e-01 -6.87974453e-01 1.88882142e-01
-5.46179771e-01 3.16003293e-01 -2.68970430e-01 8.27919841e-01
3.91115904e-01 -6.39846265e-01 1.74818784e-01 -7.60512471e-01
2.53347129e-01 2.02261478e-01 -1.07381284e+00 1.55413485e+00
-1.24817103e-01 3.16042781e-01 -3.73201966e-01 -9.73180056e-01
4.33115721e-01 1.84532300e-01 8.48460495e-01 -7.84160435e-01
1.70538798e-02 3.05925477e-02 -7.31608719e-02 -7.41897225e-01
4.79992807e-01 -6.45527849e-03 -3.45180631e-01 1.47525623e-01
2.11051181e-01 -3.00204009e-02 6.46834373e-01 3.18115592e-01
1.25631380e+00 4.09191966e-01 2.08825707e-01 -1.62693769e-01
2.92507738e-01 -2.33845823e-02 5.70016921e-01 8.36311519e-01
-1.37370601e-01 6.67607188e-01 7.61015177e-01 -3.93233478e-01
-7.87748814e-01 -1.00421512e+00 6.01392835e-02 1.61201000e+00
5.02280891e-01 -5.15264452e-01 -7.00001121e-01 -9.39621806e-01
1.53261319e-01 7.28016317e-01 -6.28151178e-01 -2.13485003e-01
-4.84849155e-01 -9.57910657e-01 5.28673232e-01 3.57604891e-01
5.10700643e-01 -1.25016034e+00 -5.38440645e-01 1.54171493e-02
-5.50142825e-01 -1.38115144e+00 -4.54139441e-01 -1.96632021e-03
-3.63711596e-01 -1.39426374e+00 -5.51444292e-01 -6.65605485e-01
7.11642683e-01 6.93391263e-02 1.74128687e+00 -7.11920531e-03
-3.62637252e-01 4.85248238e-01 -4.66067255e-01 -3.37475330e-01
2.48032603e-02 -9.58766267e-02 -9.65807512e-02 5.87356985e-02
2.93876022e-01 -4.38996673e-01 -7.02856302e-01 3.67876679e-01
-8.02967429e-01 2.62766361e-01 3.17946762e-01 7.57007837e-01
7.40947962e-01 3.31083566e-01 2.84167886e-01 -1.21447515e+00
1.36593968e-01 -6.10304058e-01 -3.29999596e-01 3.34561825e-01
-2.43797958e-01 -2.54777610e-01 3.02644551e-01 -5.68898916e-01
-9.21943486e-01 3.66184488e-02 -3.69028002e-02 -3.46838921e-01
-1.42261595e-01 2.42935583e-01 -4.90998745e-01 6.68783635e-02
5.15989006e-01 9.29793045e-02 -4.91159230e-01 -3.52061801e-02
4.75543290e-01 8.51446763e-02 5.72877347e-01 -8.23993623e-01
4.51086760e-01 5.35447776e-01 -5.50531372e-02 -6.14618778e-01
-1.31400073e+00 -5.23590922e-01 -2.16633394e-01 -5.17469347e-01
9.02122140e-01 -1.15361333e+00 -8.04741681e-01 5.90030074e-01
-9.10929859e-01 -5.45320868e-01 -6.36321366e-01 2.61564791e-01
-9.54852462e-01 1.56190708e-01 -4.30742174e-01 -4.09247339e-01
4.31223139e-02 -1.11585343e+00 1.33659661e+00 1.90015603e-02
-2.89638370e-01 -8.98440242e-01 -3.85408774e-02 7.06424594e-01
2.91840285e-02 4.15869951e-01 1.16054666e+00 -6.19396508e-01
-6.49810970e-01 4.42239232e-02 -3.49238276e-01 1.02969922e-01
1.22452013e-01 -1.74184844e-01 -9.20130253e-01 -2.09777281e-01
-4.34320033e-01 -4.95329022e-01 9.35137033e-01 4.47027504e-01
1.49986029e+00 -3.36320609e-01 -4.24536139e-01 6.82111025e-01
1.38349104e+00 -4.64424491e-02 5.98144770e-01 2.09049568e-01
1.04605269e+00 7.18829691e-01 7.18971074e-01 4.78771895e-01
8.73791695e-01 7.91383505e-01 6.88483775e-01 4.28512059e-02
-5.14837325e-01 -3.90522420e-01 2.97021359e-01 3.13531846e-01
-3.24409492e-02 -7.50135124e-01 -8.13543558e-01 6.96230412e-01
-1.95631742e+00 -1.13751793e+00 -8.14535320e-02 2.00741887e+00
9.44397748e-01 1.12819187e-01 1.63444579e-01 1.63819864e-02
6.88476384e-01 4.68481600e-01 -4.32572991e-01 1.64020196e-01
-3.72349471e-01 -1.44466087e-01 5.52666306e-01 1.76007450e-01
-1.44219780e+00 1.06324279e+00 5.66020584e+00 1.14995086e+00
-8.40904236e-01 2.28658110e-01 8.50038171e-01 -4.41998243e-01
-3.74403536e-01 -2.53723830e-01 -1.09966195e+00 5.64967334e-01
5.51279008e-01 1.03147656e-01 3.36008251e-01 7.48660386e-01
-1.23176068e-01 -2.13519797e-01 -1.21578288e+00 1.29294658e+00
3.86895835e-01 -1.31179500e+00 2.11926818e-01 -8.11883584e-02
9.60936666e-01 4.97735944e-03 1.16209537e-01 4.55256730e-01
3.51150036e-01 -8.74344468e-01 1.12763703e+00 2.83909589e-01
9.66119647e-01 -4.27351296e-01 4.23918903e-01 1.88715369e-01
-1.27366424e+00 -1.65117443e-01 -7.89552107e-02 9.09049958e-02
2.01877803e-01 8.89692545e-01 -4.27742183e-01 4.76737648e-01
1.03363168e+00 7.10508466e-01 -5.66573024e-01 8.14550042e-01
-1.83873564e-01 4.28822756e-01 -1.55758545e-01 8.17491785e-02
-1.67067943e-03 1.13707066e-01 5.34609497e-01 1.23309422e+00
2.49275908e-01 9.27695706e-02 2.84602821e-01 5.90807796e-01
-4.54422295e-01 6.79363385e-02 -6.78581536e-01 -1.25266314e-02
3.37326676e-01 1.22835612e+00 -8.03315878e-01 -4.39220399e-01
-4.43473428e-01 7.53295481e-01 3.99283558e-01 4.40474361e-01
-1.29418552e+00 1.23328760e-01 5.64990878e-01 2.81655580e-01
4.55452740e-01 1.07545406e-01 -2.07720608e-01 -1.05593276e+00
2.07684755e-01 -1.02146637e+00 8.02695811e-01 -6.82312191e-01
-1.27277243e+00 3.74517918e-01 4.54826832e-01 -1.11787355e+00
-8.69352296e-02 -4.83865559e-01 -2.65027612e-01 4.07840401e-01
-1.39611828e+00 -1.41525543e+00 -5.33365667e-01 7.01686502e-01
6.42878413e-01 9.10997167e-02 5.35988688e-01 6.79728925e-01
-4.60861892e-01 7.15673149e-01 -5.07063568e-01 -2.59843431e-02
7.42601454e-01 -1.14663720e+00 1.05215244e-01 5.57062089e-01
3.85265172e-01 1.23842210e-01 7.38623202e-01 -7.39160180e-01
-1.18875253e+00 -1.34425426e+00 9.41348374e-01 -6.63671732e-01
4.64671075e-01 -4.07148153e-01 -4.80113059e-01 7.72548020e-01
-5.82003593e-02 3.26434165e-01 6.57648146e-01 1.26224309e-01
-6.45380914e-01 -1.68684706e-01 -1.03313529e+00 6.75491929e-01
1.62300098e+00 -3.43436003e-01 -1.62141189e-01 8.54990482e-01
6.22887015e-01 -6.83002949e-01 -7.33405888e-01 7.98448205e-01
5.05232513e-01 -1.22112381e+00 1.07279980e+00 -6.10017478e-01
7.13191330e-01 -2.38223121e-01 -4.58412647e-01 -1.00825548e+00
-3.01297158e-01 -2.91900218e-01 -4.74146366e-01 1.27164149e+00
4.15771216e-01 -2.16653779e-01 1.00539637e+00 2.99161941e-01
-1.32332653e-01 -9.66292143e-01 -8.85722935e-01 -6.09340310e-01
-3.49804074e-01 -6.24474347e-01 4.65746403e-01 8.51563096e-01
-3.83971393e-01 1.92918763e-01 -4.44887698e-01 6.96481913e-02
7.37159908e-01 2.29342520e-01 7.37725556e-01 -8.36927354e-01
-4.75152940e-01 -4.41344708e-01 -4.55242246e-01 -9.31275964e-01
2.53269404e-01 -8.66566479e-01 4.34729308e-02 -1.52326763e+00
4.95406926e-01 -5.26116192e-01 -2.36402929e-01 6.09912932e-01
-3.18511486e-01 6.86328053e-01 1.31284207e-01 -5.27060032e-02
-1.11403823e+00 3.56212944e-01 1.20606494e+00 -3.05590689e-01
2.12035328e-01 -1.47894502e-01 -8.46601367e-01 9.05155122e-01
4.91557986e-01 -5.15143335e-01 -5.70860565e-01 -4.00732577e-01
6.07424617e-01 -2.52899140e-01 5.61525881e-01 -1.02467942e+00
-2.01712511e-02 -1.46398485e-01 2.87008941e-01 -6.33395016e-01
5.21053910e-01 -7.64832377e-01 3.55402887e-01 1.41838923e-01
-3.36803466e-01 6.55702800e-02 5.33297472e-02 6.34872496e-01
-1.55445784e-01 1.24602444e-01 5.33703208e-01 -3.09811711e-01
-8.80935013e-01 4.51974571e-01 2.14743629e-01 5.71656108e-01
1.27193141e+00 -2.30327308e-01 -6.39798582e-01 -4.45516050e-01
-5.71041942e-01 1.77611962e-01 3.40097338e-01 6.38257861e-01
4.02539283e-01 -1.30792093e+00 -6.69508457e-01 6.29600286e-02
4.46692437e-01 8.92204940e-02 5.81890941e-01 8.42641890e-01
-3.29130590e-01 5.56378216e-02 1.02894567e-02 -6.28846288e-01
-1.30202079e+00 3.84241641e-01 2.64131665e-01 -3.93561035e-01
-4.24189091e-01 1.38281786e+00 5.47218919e-01 -3.14968735e-01
4.91607636e-01 -2.46338859e-01 -1.13994218e-01 3.33091944e-01
3.02143395e-01 3.50024343e-01 -1.52460486e-02 -9.90195453e-01
-4.65986848e-01 6.55418813e-01 7.03030676e-02 3.56047660e-01
1.15172136e+00 1.28651425e-01 -7.78748170e-02 4.37558234e-01
1.19922602e+00 2.67006867e-02 -1.27960563e+00 -4.65696491e-02
-2.28021413e-01 -5.15205979e-01 -4.61587787e-01 -7.90820479e-01
-1.16922879e+00 3.40639979e-01 4.09875184e-01 1.80291221e-01
1.09821594e+00 4.59209621e-01 6.55377209e-01 -4.83851843e-02
4.12509412e-01 -1.07865834e+00 3.28321040e-01 3.11714590e-01
7.63840079e-01 -1.27984810e+00 1.72274232e-01 -8.06083024e-01
-7.93065488e-01 5.34600139e-01 6.99168205e-01 -4.49977294e-02
5.27453959e-01 4.84136045e-02 -2.11045057e-01 -5.66855371e-01
-7.32887447e-01 -2.33589068e-01 5.12096167e-01 4.79935467e-01
4.87381995e-01 2.22832784e-01 -3.81354690e-01 3.62933099e-01
-2.84024924e-01 -1.96505144e-01 1.73967019e-01 8.04439306e-01
-9.97433439e-02 -9.94130552e-01 -6.54290393e-02 8.23652804e-01
-5.98732352e-01 -1.89091638e-01 -1.51584968e-01 8.23873222e-01
6.60008490e-01 7.45865405e-01 1.86212435e-01 -3.80002499e-01
5.70451736e-01 3.42925563e-02 7.22056925e-01 -5.33736825e-01
-5.64757347e-01 -1.11907832e-01 4.14319009e-01 -9.26245391e-01
-7.67903566e-01 -7.01443255e-01 -1.10015273e+00 -6.27495795e-02
-1.69271663e-01 -2.31588319e-01 2.85345078e-01 8.07937801e-01
2.96247900e-01 7.51576245e-01 2.10799277e-01 -7.52294898e-01
-3.18079472e-01 -7.54012704e-01 -5.95275939e-01 1.09260631e+00
2.06778385e-02 -1.07736766e+00 -3.70647162e-01 1.88353807e-01] | [10.313591957092285, 1.7061268091201782] |
71baf2d4-0bd2-43d3-a608-995fcf65743d | visual-prompt-based-personalized-federated | 2303.08678 | null | https://arxiv.org/abs/2303.08678v1 | https://arxiv.org/pdf/2303.08678v1.pdf | Visual Prompt Based Personalized Federated Learning | As a popular paradigm of distributed learning, personalized federated learning (PFL) allows personalized models to improve generalization ability and robustness by utilizing knowledge from all distributed clients. Most existing PFL algorithms tackle personalization in a model-centric way, such as personalized layer partition, model regularization, and model interpolation, which all fail to take into account the data characteristics of distributed clients. In this paper, we propose a novel PFL framework for image classification tasks, dubbed pFedPT, that leverages personalized visual prompts to implicitly represent local data distribution information of clients and provides that information to the aggregation model to help with classification tasks. Specifically, in each round of pFedPT training, each client generates a local personalized prompt related to local data distribution. Then, the local model is trained on the input composed of raw data and a visual prompt to learn the distribution information contained in the prompt. During model testing, the aggregated model obtains prior knowledge of the data distributions based on the prompts, which can be seen as an adaptive fine-tuning of the aggregation model to improve model performances on different clients. Furthermore, the visual prompt can be added as an orthogonal method to implement personalization on the client for existing FL methods to boost their performance. Experiments on the CIFAR10 and CIFAR100 datasets show that pFedPT outperforms several state-of-the-art (SOTA) PFL algorithms by a large margin in various settings. | ['DaCheng Tao', 'Baoyuan Wu', 'Li Shen', 'Yan Sun', 'Wansen Wu', 'Guanghao Li'] | 2023-03-15 | null | null | null | null | ['personalized-federated-learning'] | ['methodology'] | [-4.07151163e-01 -1.77113220e-01 -5.74443281e-01 -6.87287569e-01
-7.57261515e-01 -4.21628654e-01 2.41968155e-01 -7.53115043e-02
-1.14074498e-01 6.36530995e-01 6.90276846e-02 -1.45349562e-01
-1.93905622e-01 -6.07957423e-01 -8.19857299e-01 -9.97499466e-01
1.57076329e-01 7.86529958e-01 1.64651811e-01 3.49090695e-01
-8.05336758e-02 5.63583136e-01 -1.32439888e+00 8.11421514e-01
9.55625951e-01 1.41487670e+00 2.93098807e-01 4.56101418e-01
-3.46747160e-01 1.05983877e+00 -5.58941126e-01 -3.23368222e-01
4.30617146e-02 -2.58810278e-02 -6.92692935e-01 1.19851261e-01
4.37125146e-01 -6.11605406e-01 -2.90981799e-01 5.90117455e-01
5.39307833e-01 1.82314605e-01 3.72822613e-01 -1.32357109e+00
-8.01132321e-01 9.10105169e-01 -4.30073261e-01 1.51143357e-01
-3.22385252e-01 4.57650304e-01 8.43488038e-01 -5.96069574e-01
3.30911100e-01 1.28525603e+00 5.43742597e-01 6.59041584e-01
-1.57358980e+00 -7.32199550e-01 7.58509636e-01 5.04963040e-01
-1.23152065e+00 -2.68529087e-01 7.75140226e-01 -3.39572698e-01
5.73173463e-01 3.48789215e-01 3.28936458e-01 9.71686244e-01
-1.49208248e-01 1.12784576e+00 8.37899208e-01 -8.06416795e-02
6.22425497e-01 5.06541550e-01 3.89171243e-01 3.87987465e-01
-2.13962555e-01 -2.09372848e-01 -8.36678505e-01 -6.21434391e-01
4.54630256e-01 3.52594018e-01 -2.94590473e-01 -4.84598309e-01
-7.09175229e-01 6.54960334e-01 7.11563230e-01 -8.85811001e-02
-7.64530063e-01 7.72682056e-02 4.36939269e-01 1.37067929e-01
6.39432788e-01 -1.16380960e-01 -8.47755015e-01 1.86871082e-01
-6.38444245e-01 2.63504624e-01 6.81144178e-01 6.38997197e-01
1.34031630e+00 -1.34353682e-01 -8.36172998e-01 8.42741668e-01
3.74707580e-01 4.57002759e-01 6.75261497e-01 -9.82616544e-01
5.90698957e-01 8.36178601e-01 -3.10967136e-02 -5.50002217e-01
-5.49844094e-02 -6.18082583e-01 -6.41783178e-01 -5.50702736e-02
1.02339663e-01 -2.18611985e-01 -6.67883039e-01 1.69224846e+00
8.39999616e-01 4.54629362e-01 -1.13412909e-01 6.92351878e-01
4.58621174e-01 5.77363849e-01 2.83352733e-01 7.05092177e-02
8.73257399e-01 -1.20394564e+00 -2.41392195e-01 2.42049228e-02
5.54610133e-01 -3.53785425e-01 1.19867754e+00 5.26783526e-01
-6.28712654e-01 -3.88507366e-01 -4.45953518e-01 2.23181307e-01
5.67519218e-02 3.63186486e-02 6.44123554e-01 4.37519521e-01
-8.76664162e-01 4.72298473e-01 -8.80683661e-01 -1.10609770e-01
8.94784272e-01 2.96558797e-01 -1.02363892e-01 -3.63769948e-01
-5.75276911e-01 2.87205845e-01 2.56411463e-01 -3.39341074e-01
-1.21608782e+00 -1.22255170e+00 -3.15874785e-01 4.92452294e-01
3.14312577e-01 -8.58362079e-01 1.54304826e+00 -1.10881388e+00
-1.54972017e+00 2.82761127e-01 -1.89378619e-01 -5.86933672e-01
6.22838736e-01 -8.02460965e-03 -5.47677390e-02 3.61200832e-02
-2.65401512e-01 3.58289391e-01 9.95697677e-01 -1.45328498e+00
-9.09320712e-01 -5.07215977e-01 2.44979821e-02 -3.60515676e-02
-7.24042356e-01 -2.35659286e-01 -6.31181538e-01 -3.55537325e-01
-3.13211769e-01 -6.47250295e-01 -1.92547187e-01 1.05144262e-01
-3.32511008e-01 -6.37361646e-01 1.19380736e+00 -4.70809609e-01
1.26777768e+00 -2.22625375e+00 -1.38627186e-01 3.61063421e-01
3.33730906e-01 3.17434192e-01 -3.32713574e-01 3.52046162e-01
2.60530025e-01 -1.68421660e-02 1.65592566e-01 -8.37015688e-01
3.11552942e-01 4.11795944e-01 -4.29431379e-01 2.05280438e-01
-3.16656530e-01 8.27623785e-01 -5.59935331e-01 -4.68556941e-01
-1.00229591e-01 6.67885721e-01 -8.65516961e-01 4.22500402e-01
-8.02592158e-01 7.03773797e-01 -7.42327988e-01 3.34139466e-01
8.41217041e-01 -7.46502995e-01 3.54926735e-01 -2.63398379e-01
1.89751208e-01 1.18285790e-01 -1.00484145e+00 1.39224875e+00
-6.95235252e-01 9.87229273e-02 2.20801666e-01 -9.51040268e-01
7.55209208e-01 3.64624351e-01 6.35548890e-01 -5.52158117e-01
-2.12411061e-01 -1.44027308e-01 -5.43757677e-01 -4.70304340e-01
-3.15690786e-02 2.84932107e-01 3.82260740e-01 9.42189932e-01
-2.38123443e-02 7.57036865e-01 -1.89092815e-01 2.67666876e-01
9.89896536e-01 -1.16927996e-01 -2.62444079e-01 1.12767778e-01
3.78304541e-01 -3.36300880e-01 7.61236727e-01 8.99574518e-01
-1.20380260e-02 1.81084335e-01 2.87096024e-01 -6.83739245e-01
-6.53821111e-01 -9.83437002e-01 1.29282266e-01 1.82939696e+00
-1.38521358e-01 -4.80526716e-01 -7.05904305e-01 -1.17212439e+00
3.67461056e-01 7.10633576e-01 -6.39905512e-01 -2.01131165e-01
-3.47834527e-01 -6.55438185e-01 7.62057677e-02 6.08699262e-01
4.57443446e-01 -9.02211428e-01 -4.45371777e-01 1.78996623e-01
-2.29782477e-01 -7.42972493e-01 -7.34274149e-01 1.68718204e-01
-8.35706055e-01 -9.49230433e-01 -5.08431494e-01 -5.13221741e-01
7.41292775e-01 2.76878029e-01 8.37855756e-01 -1.54541815e-02
-5.52564114e-02 5.23847580e-01 -2.15025648e-01 -2.90515274e-01
-2.30878294e-01 1.93049163e-01 -2.12054461e-01 7.64620185e-01
1.48788989e-01 -7.62672961e-01 -7.87207484e-01 4.31806266e-01
-9.57131088e-01 7.73014501e-02 5.24904788e-01 8.18135798e-01
8.26675653e-01 -1.37064457e-01 6.90543830e-01 -1.37137079e+00
4.95111346e-01 -7.95893371e-01 -4.57193047e-01 6.62402868e-01
-8.47129047e-01 1.88041046e-01 9.55069423e-01 -9.08275366e-01
-1.46390355e+00 3.09453774e-02 2.65557647e-01 -9.71359670e-01
-9.64633003e-02 6.59086108e-01 -4.14711833e-01 7.12775886e-02
7.67183840e-01 2.26272643e-01 -1.35083690e-01 -1.11131477e+00
4.77299482e-01 8.67585599e-01 4.63166505e-01 -8.58538151e-01
4.54191744e-01 4.81578201e-01 -5.21808863e-01 -4.57548499e-02
-8.12935472e-01 -2.50929207e-01 -2.85738140e-01 -2.87372842e-02
3.15774798e-01 -8.01995456e-01 -9.59639847e-01 5.75487971e-01
-9.40569580e-01 -9.20718610e-01 -3.95218015e-01 2.59825826e-01
-3.82091671e-01 2.72622611e-02 -5.68383873e-01 -7.12409496e-01
-5.90686381e-01 -1.03701210e+00 7.93766141e-01 3.94890904e-01
2.88678437e-01 -1.13377762e+00 -2.29026210e-02 6.07496858e-01
8.42165828e-01 -3.28861654e-01 1.27387667e+00 -9.34084058e-01
-8.50056112e-01 -2.98872173e-01 -1.79548070e-01 4.98125821e-01
5.64702749e-02 -1.46700084e-01 -1.36456919e+00 -4.08824891e-01
-2.24517554e-01 -4.29827899e-01 7.49143124e-01 1.79343641e-01
1.87863731e+00 -9.40201044e-01 -4.63978767e-01 8.61426055e-01
1.22685492e+00 -1.54032456e-02 1.99723169e-01 5.12923449e-02
7.98200309e-01 2.38563091e-01 7.77233541e-02 1.04106569e+00
6.78239584e-01 6.84240043e-01 5.48785448e-01 1.94413483e-01
-1.79470405e-01 -4.12535548e-01 2.17355728e-01 3.03447843e-01
4.24459517e-01 -2.02806488e-01 -7.59699285e-01 3.08485866e-01
-2.24210835e+00 -7.14597464e-01 4.25571829e-01 2.28858519e+00
9.09853399e-01 -5.41575551e-01 1.53189555e-01 -4.04286504e-01
6.54008925e-01 -3.43064889e-02 -1.26056910e+00 -3.96716073e-02
1.60075426e-01 -2.27768734e-01 2.91802108e-01 1.34189919e-01
-8.44151914e-01 7.62508690e-01 5.15509462e+00 9.86902654e-01
-1.52529860e+00 4.22617763e-01 8.48548651e-01 -2.64958382e-01
-4.83304560e-01 -9.18005183e-02 -8.89712870e-01 7.51646578e-01
8.86472940e-01 -3.46310139e-01 7.98483729e-01 1.16475189e+00
1.53282970e-01 1.49930894e-01 -1.27136374e+00 9.63387787e-01
-1.40200198e-01 -1.49083269e+00 2.18072116e-01 1.04417883e-01
9.00238872e-01 4.42140549e-01 3.97512674e-01 3.98879498e-01
6.09816611e-01 -6.72337413e-01 5.28994322e-01 9.46035683e-01
4.23764944e-01 -8.37679744e-01 3.88713717e-01 7.90699303e-01
-9.10245657e-01 -5.70302308e-01 -4.11193848e-01 4.42006707e-01
-1.87208220e-01 5.86236656e-01 -9.56190407e-01 4.76585388e-01
1.05303764e+00 4.60705549e-01 -6.15996778e-01 1.25900388e+00
-1.56880785e-02 9.62774038e-01 -3.95939887e-01 3.60296011e-01
-1.30519569e-01 1.78006664e-01 2.11807489e-01 8.62993538e-01
1.27645051e-02 -2.34285191e-01 4.30788249e-01 8.01185489e-01
-3.30031514e-01 2.64074534e-01 -2.78682671e-02 3.86540413e-01
7.39260435e-01 1.28189933e+00 3.15262489e-02 -5.12027681e-01
-3.75646383e-01 7.11503625e-01 8.42727363e-01 7.28703260e-01
-6.23189270e-01 1.00149408e-01 7.49898136e-01 -1.53321195e-02
6.40163660e-01 4.25697535e-01 -1.49838328e-01 -1.15080392e+00
-8.14249099e-04 -9.02137101e-01 8.23807180e-01 -5.81302643e-01
-1.93628633e+00 6.66539907e-01 -1.11122288e-01 -8.52263987e-01
-3.21935982e-01 -1.26279533e-01 -8.90701056e-01 1.02069986e+00
-1.48311222e+00 -1.52858663e+00 -4.29386139e-01 1.05374277e+00
1.91069633e-01 -2.17009753e-01 7.75063694e-01 3.00833821e-01
-9.29020882e-01 1.03775740e+00 5.15326500e-01 -1.57893211e-01
8.57470632e-01 -9.61615026e-01 -1.20884024e-01 4.31544393e-01
1.29119726e-03 4.15963978e-01 3.90883714e-01 -5.14886975e-01
-1.30139029e+00 -1.63957548e+00 6.19631052e-01 -7.31093585e-02
4.51627403e-01 -1.45700857e-01 -1.23457742e+00 1.01547277e+00
-3.29074226e-02 5.57590365e-01 8.21834147e-01 2.31534004e-01
-6.92844748e-01 -8.70410442e-01 -1.20827186e+00 2.41348684e-01
6.87102079e-01 -4.60949689e-01 1.91166878e-01 6.12604856e-01
6.98648334e-01 -2.07091272e-01 -7.61570632e-01 5.58044054e-02
2.94059396e-01 -9.17031586e-01 7.42930472e-01 -8.86236429e-01
-2.42838800e-01 2.71734726e-02 -2.49089614e-01 -1.38620424e+00
-3.65428984e-01 -6.87174141e-01 -4.78810847e-01 1.54532361e+00
1.89181119e-01 -9.87517476e-01 9.51698184e-01 1.09570289e+00
-2.05924809e-01 -9.63363111e-01 -8.58168960e-01 -5.71051061e-01
-8.21648464e-02 -4.97428298e-01 1.24125421e+00 5.80698311e-01
-2.05182493e-01 9.80042964e-02 -3.32905948e-01 3.66351873e-01
7.48688042e-01 4.47667450e-01 1.11578941e+00 -1.16567373e+00
-7.05241561e-01 -3.69425565e-01 1.39522195e-01 -1.19898272e+00
4.35545832e-01 -9.96702194e-01 -2.50513911e-01 -1.41886187e+00
4.91431117e-01 -9.13486063e-01 -5.51403046e-01 1.03138030e+00
-2.81765670e-01 -1.31850243e-01 4.24881071e-01 5.34759104e-01
-1.03311241e+00 6.93633735e-01 1.01697493e+00 -2.21844062e-01
-4.12304491e-01 4.31547016e-01 -7.32919157e-01 4.08236951e-01
6.14879429e-01 -4.98317391e-01 -6.13515437e-01 -7.12394953e-01
-1.63367599e-01 -5.99189214e-02 5.41716337e-01 -7.49237180e-01
5.89650273e-01 -4.31854695e-01 4.20726359e-01 -4.44214463e-01
1.18245631e-01 -8.74025583e-01 3.04343253e-01 -1.65818795e-03
-5.42736411e-01 -2.64824927e-01 2.31149253e-02 6.54285848e-01
1.65881813e-02 1.90090314e-01 7.42734909e-01 1.16180554e-01
-4.48952138e-01 8.76706302e-01 1.25413584e-02 -1.46145776e-01
9.79965091e-01 3.27707559e-01 -6.73662901e-01 -4.05858248e-01
-6.84751511e-01 6.27230048e-01 3.21205616e-01 6.48733079e-02
3.16571146e-01 -1.23975563e+00 -5.54636717e-01 4.10790980e-01
8.04462060e-02 3.02042902e-01 9.30346906e-01 9.41752553e-01
4.01741154e-02 2.67437428e-01 1.11808449e-01 -7.01515079e-01
-1.07961929e+00 7.64656425e-01 5.18649936e-01 -4.18561608e-01
-5.72562456e-01 8.41033340e-01 7.50815094e-01 -3.94168347e-01
6.63713396e-01 -9.42432508e-02 9.16851908e-02 -7.35658705e-02
9.12231743e-01 3.64636064e-01 2.34934300e-01 -1.85528323e-01
-2.59735435e-01 1.27807390e-02 -4.19678003e-01 2.62679458e-01
1.53840125e+00 -3.19635123e-01 -3.33907153e-03 4.29968476e-01
1.22743607e+00 -2.09609300e-01 -1.98647642e+00 -9.16947901e-01
-2.11049482e-01 -5.05924165e-01 9.67244729e-02 -1.28743768e+00
-1.60251427e+00 6.77178144e-01 6.06655359e-01 -1.12077937e-01
1.26826119e+00 2.80753225e-01 7.26017177e-01 2.90107608e-01
3.67479563e-01 -7.32252419e-01 8.68451074e-02 3.11143577e-01
6.73906744e-01 -9.60931480e-01 -4.40140069e-01 9.93118063e-02
-8.58860672e-01 7.96893835e-01 8.33823860e-01 2.49889925e-01
7.74125397e-01 5.93014807e-02 2.94580251e-01 2.13143215e-01
-1.32198453e+00 3.00927520e-01 4.27505910e-01 6.73678517e-01
-2.04362497e-01 3.57116461e-02 4.66949046e-01 1.09117866e+00
3.02295357e-01 2.25308210e-01 -3.13137800e-01 7.79479265e-01
-3.32537681e-01 -1.23525155e+00 -3.61417323e-01 5.47964931e-01
-2.47491181e-01 2.53370345e-01 3.92340161e-02 1.04163229e-01
1.91751182e-01 7.42718279e-01 1.09904274e-01 -3.68800640e-01
1.97965488e-01 2.33092681e-01 1.28450319e-01 -5.55531561e-01
-6.93268955e-01 7.65149444e-02 -4.72731471e-01 -7.01352894e-01
1.19666457e-01 -4.73776072e-01 -1.21088934e+00 -4.48805481e-01
-4.32890980e-03 2.95395732e-01 5.94521224e-01 9.78370368e-01
9.35047567e-01 3.27280164e-01 9.93871033e-01 -9.78405833e-01
-9.45290327e-01 -8.92033160e-01 -7.27173209e-01 4.56398904e-01
3.80302191e-01 -2.84422249e-01 -3.42247635e-01 9.78619382e-02] | [5.8199262619018555, 6.2872796058654785] |
6ea15585-5125-4f0b-9a11-a7635f1d0221 | joint-biomedical-entity-and-relation | 2105.13456 | null | https://arxiv.org/abs/2105.13456v2 | https://arxiv.org/pdf/2105.13456v2.pdf | Joint Biomedical Entity and Relation Extraction with Knowledge-Enhanced Collective Inference | Compared to the general news domain, information extraction (IE) from biomedical text requires much broader domain knowledge. However, many previous IE methods do not utilize any external knowledge during inference. Due to the exponential growth of biomedical publications, models that do not go beyond their fixed set of parameters will likely fall behind. Inspired by how humans look up relevant information to comprehend a scientific text, we present a novel framework that utilizes external knowledge for joint entity and relation extraction named KECI (Knowledge-Enhanced Collective Inference). Given an input text, KECI first constructs an initial span graph representing its initial understanding of the text. It then uses an entity linker to form a knowledge graph containing relevant background knowledge for the the entity mentions in the text. To make the final predictions, KECI fuses the initial span graph and the knowledge graph into a more refined graph using an attention mechanism. KECI takes a collective approach to link mention spans to entities by integrating global relational information into local representations using graph convolutional networks. Our experimental results show that the framework is highly effective, achieving new state-of-the-art results in two different benchmark datasets: BioRelEx (binding interaction detection) and ADE (adverse drug event extraction). For example, KECI achieves absolute improvements of 4.59% and 4.91% in F1 scores over the state-of-the-art on the BioRelEx entity and relation extraction tasks. | ['Quan Hung Tran', 'ChengXiang Zhai', 'Heng Ji', 'Tuan Lai'] | 2021-05-27 | null | https://aclanthology.org/2021.acl-long.488 | https://aclanthology.org/2021.acl-long.488.pdf | acl-2021-5 | ['joint-entity-and-relation-extraction'] | ['natural-language-processing'] | [ 6.52044639e-02 7.34552801e-01 -4.51176524e-01 -2.32131481e-01
-6.47096097e-01 -3.47904533e-01 3.62223089e-01 9.13764715e-01
-3.88165414e-01 9.89138901e-01 3.09101820e-01 -3.56870711e-01
-2.89115071e-01 -1.14809322e+00 -9.81873274e-01 -4.07490134e-01
-9.24462220e-04 6.68069124e-01 2.53666252e-01 -6.79501593e-02
-1.03074029e-01 2.37606838e-01 -7.43664861e-01 2.53620297e-01
1.00802791e+00 7.50542521e-01 9.27485228e-02 2.95904905e-01
-2.57127017e-01 9.46303070e-01 -5.73427022e-01 -5.45236707e-01
-3.83328259e-01 -1.76438928e-01 -1.27254927e+00 -4.88248825e-01
-1.95977718e-01 1.12583667e-01 -4.65300143e-01 9.89550889e-01
4.92470860e-01 -2.46307328e-02 6.13369286e-01 -8.25334549e-01
-9.15309608e-01 1.15607822e+00 -4.37291741e-01 2.04185039e-01
3.97773832e-01 -1.75359413e-01 1.27392685e+00 -9.23963785e-01
1.05409777e+00 1.01111531e+00 5.83058357e-01 3.67705345e-01
-1.07745957e+00 -6.62197530e-01 2.92841703e-01 2.47355163e-01
-1.55201006e+00 -1.53173581e-01 4.24870133e-01 -3.36081505e-01
1.41870165e+00 1.24651976e-01 4.22214508e-01 1.01050138e+00
4.86107200e-01 6.06338739e-01 5.54518163e-01 -1.66479200e-01
1.39195502e-01 -5.22237495e-02 6.54195726e-01 9.32388723e-01
6.91678762e-01 -4.45115030e-01 -4.77685183e-01 -1.68998405e-01
4.00375992e-01 -3.75680178e-02 -4.35118318e-01 2.36352444e-01
-1.28025699e+00 6.51956797e-01 7.96303749e-01 3.22280705e-01
-6.58191323e-01 -3.78428772e-02 3.98717284e-01 -1.35726094e-01
4.63934153e-01 7.78450370e-01 -7.65766680e-01 3.87841493e-01
-5.54714441e-01 3.37719917e-02 1.11630297e+00 9.24607158e-01
5.31152010e-01 -6.82440341e-01 -3.03421855e-01 6.01271808e-01
3.27458590e-01 9.42591578e-02 2.68881708e-01 -1.77516520e-01
5.12391806e-01 1.08418798e+00 -1.97986498e-01 -1.19007444e+00
-8.29365969e-01 -9.26496267e-01 -9.75569963e-01 -6.20164454e-01
1.32388830e-01 -3.16278040e-01 -1.14270997e+00 1.74422479e+00
4.89337295e-01 3.70400548e-01 2.70063043e-01 5.85241556e-01
1.59784019e+00 4.86817777e-01 5.26802599e-01 -7.52841532e-02
1.76427543e+00 -7.59848237e-01 -1.03398466e+00 -2.19281986e-01
6.62979126e-01 -3.50391120e-01 1.95731610e-01 1.89129621e-01
-8.52565765e-01 -2.87477970e-01 -1.03304422e+00 -3.16160262e-01
-8.41218948e-01 2.80189188e-03 7.19218969e-01 -5.19063510e-02
-7.37745702e-01 4.81087476e-01 -7.96718359e-01 -4.23037916e-01
6.79810405e-01 4.95303780e-01 -6.73507154e-01 -6.38810769e-02
-1.73683393e+00 1.10648727e+00 7.25837529e-01 1.71284229e-01
-5.99958181e-01 -1.17778778e+00 -9.56995428e-01 3.33596528e-01
7.81426609e-01 -1.23457968e+00 6.14727437e-01 -1.02690846e-01
-9.97639954e-01 7.86484957e-01 -1.38285577e-01 -5.30973196e-01
-2.47308221e-02 -2.78995872e-01 -4.01665628e-01 1.78968012e-01
1.47391662e-01 4.34615493e-01 7.24421516e-02 -8.52838755e-01
-3.54987770e-01 -5.16125679e-01 1.59870788e-01 -9.56892967e-04
-9.09552425e-02 -1.38262644e-01 -1.06560135e+00 -4.48483348e-01
-6.18513152e-02 -6.47869110e-01 -3.69885445e-01 -5.36291957e-01
-9.21825409e-01 -6.19580150e-01 4.99841422e-01 -8.63075912e-01
1.54058671e+00 -1.52638447e+00 5.99034607e-01 2.65587002e-01
9.10576522e-01 2.85599917e-01 -5.14631793e-02 5.10294497e-01
-2.58173794e-01 4.20668751e-01 4.79712673e-02 7.14421272e-02
-1.56615421e-01 1.01513281e-01 6.36180639e-02 1.50563881e-01
4.78536755e-01 1.34766626e+00 -1.16871512e+00 -5.41696787e-01
-2.64694840e-01 6.45197630e-01 -5.72571039e-01 5.69741195e-03
-4.36171263e-01 2.59213805e-01 -8.76656294e-01 7.00073779e-01
4.52769905e-01 -1.00972009e+00 5.07354677e-01 -6.80551112e-01
3.94317150e-01 4.00768697e-01 -6.46221399e-01 1.56412756e+00
-1.24393925e-01 1.88055575e-01 -2.47214735e-01 -1.20387137e+00
7.81737328e-01 3.11553687e-01 6.10215843e-01 -2.56851643e-01
1.50743142e-01 -1.91454738e-02 -2.09314842e-02 -5.74678838e-01
2.42935065e-02 6.00703247e-03 -1.40941754e-01 3.10796918e-03
4.41774637e-01 4.65080947e-01 1.31826535e-01 7.54102051e-01
1.56640232e+00 -2.32623052e-02 7.50353575e-01 -2.08229005e-01
6.66597366e-01 -7.74349645e-02 7.94235528e-01 5.70721269e-01
2.86200315e-01 1.60878658e-01 8.54827642e-01 -3.48753333e-01
-3.92136902e-01 -8.22886646e-01 -9.53263417e-02 7.29413629e-01
1.10137998e-03 -9.14084435e-01 -7.15352118e-01 -1.13225627e+00
1.06559686e-01 4.30848956e-01 -9.84811664e-01 -2.99435884e-01
-2.54172057e-01 -1.07093954e+00 5.82998514e-01 6.37534201e-01
4.54183996e-01 -9.16936338e-01 3.11108660e-02 4.02210087e-01
-3.79435539e-01 -1.45992982e+00 -2.61329293e-01 1.99230194e-01
-5.57646990e-01 -1.28674066e+00 -3.49676341e-01 -6.31838560e-01
8.34715486e-01 -3.93757731e-01 1.20394754e+00 1.18359834e-01
-4.64622229e-01 7.54855275e-02 -2.46943742e-01 -6.80070639e-01
-2.17724487e-01 3.68291855e-01 -2.39089489e-01 -2.23556623e-01
7.02830076e-01 -2.47259170e-01 -7.48748660e-01 -7.53953084e-02
-8.45873475e-01 2.42521703e-01 7.75636911e-01 1.05677986e+00
7.76145399e-01 -3.83984782e-02 9.79178190e-01 -1.37162948e+00
4.99547422e-01 -9.10447001e-01 -3.36259305e-01 4.51968819e-01
-6.65649533e-01 2.68897653e-01 4.95582461e-01 -1.61822349e-01
-9.81125116e-01 -8.20710734e-02 -2.50075877e-01 -8.73158406e-03
4.89887334e-02 1.46121347e+00 -3.35675359e-01 3.55229795e-01
5.23625672e-01 -1.74031466e-01 -1.92422554e-01 -4.46204305e-01
3.32640320e-01 5.87966204e-01 4.94303584e-01 -4.86740172e-01
3.39227349e-01 8.20869207e-02 1.62821949e-01 -5.45587301e-01
-1.37464428e+00 -4.67481196e-01 -5.76168716e-01 2.13606656e-01
1.25417948e+00 -9.44033504e-01 -1.01675093e+00 1.58041775e-01
-1.17539489e+00 -1.79422600e-03 -1.51254963e-02 4.31332022e-01
-5.54919578e-02 2.66305625e-01 -7.28710532e-01 -1.79353237e-01
-7.61743307e-01 -9.90704417e-01 1.10454345e+00 2.89598942e-01
-2.29947507e-01 -1.25088370e+00 5.43772578e-02 4.99473363e-01
4.50953580e-02 5.29845595e-01 1.33900583e+00 -1.09445846e+00
-6.50017321e-01 -3.34857911e-01 -4.27442700e-01 -1.91195667e-01
3.66172731e-01 -2.74159104e-01 -7.76934028e-01 -6.19785786e-02
-6.10994995e-01 -2.17996761e-02 1.03801477e+00 1.37981951e-01
1.19767404e+00 -3.16733778e-01 -1.05946076e+00 5.35284877e-01
1.36280727e+00 -2.56564990e-02 6.88013434e-01 1.89770743e-01
1.01931095e+00 4.72325921e-01 3.91726702e-01 2.09546268e-01
6.57065392e-01 5.31861663e-01 2.64663488e-01 -3.28450769e-01
-9.89530981e-02 -3.67714673e-01 -2.03602742e-02 8.07636976e-01
-1.74033403e-01 -4.79787022e-01 -1.10612237e+00 6.19537890e-01
-2.00812840e+00 -6.26074195e-01 -1.99086796e-02 1.80468583e+00
1.28705132e+00 7.40420297e-02 -3.72996181e-01 -2.70983070e-01
6.60367727e-01 -2.82946736e-01 -8.09448004e-01 4.52166945e-02
-3.77465598e-02 4.30272907e-01 5.36329925e-01 3.16362739e-01
-1.04763818e+00 1.15221798e+00 5.41672993e+00 7.88897216e-01
-7.56285965e-01 -3.11201643e-02 7.21418142e-01 3.26216102e-01
-2.00149849e-01 -3.18025872e-02 -1.17027664e+00 1.81707487e-01
1.19429946e+00 -3.42598855e-01 1.39213996e-02 3.56495917e-01
-5.26033379e-02 9.44455564e-02 -1.17011130e+00 6.04995191e-01
-1.39689505e-01 -1.85110223e+00 2.93267190e-01 1.81789264e-01
5.29603601e-01 2.23096669e-01 -4.16973680e-01 4.55595016e-01
6.67699277e-01 -1.22059572e+00 -2.67515704e-02 9.99833226e-01
7.18105435e-01 -6.85903311e-01 1.18024683e+00 2.57750154e-01
-1.15402305e+00 1.88455403e-01 -4.09229696e-01 4.82587606e-01
-3.29341441e-02 9.63075221e-01 -1.16388452e+00 1.23710299e+00
5.67694008e-01 9.54925835e-01 -6.22728765e-01 7.84325361e-01
-4.53625679e-01 6.06198311e-01 -1.43399775e-01 -4.16863374e-02
5.32252043e-02 1.46892846e-01 3.55871558e-01 1.43704307e+00
5.89763699e-03 4.30015206e-01 2.08011463e-01 9.48645175e-01
-7.10090339e-01 3.34703714e-01 -5.06919086e-01 -4.73436981e-01
3.19334656e-01 1.36263573e+00 -6.41169965e-01 -6.35989428e-01
-4.53888059e-01 6.76454842e-01 7.39349127e-01 3.96134257e-01
-8.90552223e-01 -7.35653043e-01 4.80398864e-01 -2.15574056e-01
3.85789961e-01 2.77220249e-01 -4.83519621e-02 -1.15071023e+00
-2.73364902e-01 -7.01087534e-01 8.46032858e-01 -6.04542792e-01
-1.53845537e+00 6.21951342e-01 -3.94105352e-02 -5.31789303e-01
7.37048909e-02 -5.58866024e-01 -2.36645520e-01 9.47894514e-01
-1.45493782e+00 -1.27914548e+00 -1.33972526e-01 2.94555485e-01
2.51040429e-01 3.25418115e-02 1.02959001e+00 1.98419765e-01
-7.73335159e-01 6.49564445e-01 -2.73744613e-01 5.15430212e-01
7.08643734e-01 -1.32725823e+00 4.39880311e-01 3.52114081e-01
-2.07362354e-01 1.28240073e+00 3.48390192e-01 -1.16248167e+00
-1.55976415e+00 -1.36391866e+00 1.18095553e+00 -6.48359656e-01
8.82003248e-01 -2.39715919e-01 -1.21816111e+00 8.56902957e-01
1.80729359e-01 1.64106920e-01 9.84846413e-01 3.98581058e-01
-4.16762948e-01 1.78161547e-01 -9.43070292e-01 4.56646711e-01
1.19939196e+00 -3.81086469e-01 -7.99530149e-01 4.85606194e-01
1.04366159e+00 -5.31978369e-01 -1.56101143e+00 6.06751978e-01
3.82871151e-01 -4.92461212e-03 1.08205795e+00 -1.16648376e+00
4.61255193e-01 -3.36341798e-01 2.47549608e-01 -1.15099931e+00
-5.51893234e-01 -5.13738751e-01 -4.12614226e-01 1.05781090e+00
9.16674495e-01 -7.00192153e-01 4.58077848e-01 5.54380417e-01
-1.19594656e-01 -1.14090645e+00 -5.31922102e-01 -3.71041059e-01
4.41275872e-02 -1.35610132e-02 6.60539508e-01 1.14099371e+00
5.49438357e-01 1.07456315e+00 -8.64705071e-03 4.69204694e-01
4.50850904e-01 1.23955518e-01 4.04891908e-01 -1.54828262e+00
-2.21803829e-01 -1.70315400e-01 -5.01254916e-01 -7.33150363e-01
3.25218260e-01 -1.39461184e+00 -2.03001708e-01 -2.17753553e+00
6.42624855e-01 -2.03088760e-01 -7.27710366e-01 9.24833536e-01
-7.26116061e-01 -1.46097771e-03 -2.49794319e-01 -1.78830713e-01
-8.81980777e-01 2.54646361e-01 1.16083264e+00 -3.18863034e-01
-5.23710698e-02 -4.19551581e-01 -1.22996199e+00 5.79446197e-01
5.92994928e-01 -4.10001397e-01 -3.27858001e-01 -1.35546625e-01
6.42035544e-01 1.09158687e-01 1.09264240e-01 -5.73752284e-01
5.62010527e-01 3.94907333e-02 5.73290884e-01 -5.58519244e-01
-6.16200902e-02 -4.62550491e-01 3.12827170e-01 3.64713460e-01
-4.39708591e-01 -3.00452828e-01 3.91586453e-01 9.32451904e-01
-1.22258186e-01 2.28890367e-02 3.80553842e-01 -1.23138957e-01
-3.53415161e-01 4.85502452e-01 -8.96382183e-02 6.42015934e-02
7.94767022e-01 3.18515390e-01 -8.58871698e-01 -3.57376598e-02
-1.18961191e+00 5.28050900e-01 -6.73499107e-02 3.75105053e-01
6.88064694e-01 -9.43195760e-01 -7.54655898e-01 -1.02030262e-01
2.35195130e-01 1.51541784e-01 2.52966762e-01 1.03797174e+00
-2.54787624e-01 6.99093461e-01 2.11787909e-01 -3.49405617e-01
-1.25071478e+00 8.27021182e-01 2.07350761e-01 -8.89753878e-01
-6.82050467e-01 7.98895836e-01 6.71756506e-01 -3.61725986e-01
9.71567482e-02 -4.16593194e-01 -5.79196095e-01 3.55581045e-02
5.05912781e-01 -1.81103917e-03 3.41898143e-01 -3.77168059e-01
-6.64522946e-01 3.85846257e-01 -4.97444957e-01 4.60881025e-01
1.61687994e+00 1.75616562e-01 -4.45395052e-01 7.42396414e-02
1.10460401e+00 7.20273405e-02 -5.43340981e-01 -4.34683591e-01
2.74972647e-01 2.00305909e-01 1.30080879e-01 -1.27199996e+00
-1.10970247e+00 5.72975934e-01 -7.96283036e-02 -1.39132828e-01
9.20021892e-01 4.48587686e-01 7.71242559e-01 6.17131710e-01
1.58724546e-01 -7.86210120e-01 -2.10549966e-01 6.10989630e-01
9.59889531e-01 -1.11970544e+00 3.53033274e-01 -8.61670375e-01
-4.51467842e-01 9.94419932e-01 5.56237519e-01 2.97174424e-01
8.38836133e-01 3.40060323e-01 -3.30629468e-01 -8.30384552e-01
-1.02588189e+00 -2.45805919e-01 6.97813332e-01 3.82973999e-01
5.73590577e-01 5.72193265e-02 -4.14215684e-01 1.24077141e+00
1.96781337e-01 2.53936827e-01 1.63838178e-01 6.29184306e-01
-2.25365981e-01 -1.03697801e+00 1.32456943e-01 6.68569863e-01
-9.11166072e-01 -3.35204452e-01 -5.49976230e-01 7.02660382e-01
7.22213462e-02 8.35323274e-01 -3.64312500e-01 -4.34122682e-01
3.09475094e-01 1.98046014e-01 2.15492114e-01 -8.98160100e-01
-7.00882077e-01 1.18488975e-01 3.68441641e-01 -5.03127098e-01
-2.81342834e-01 -2.27647811e-01 -1.82190299e+00 -1.76301245e-02
-5.00180125e-01 3.39002460e-01 2.99042493e-01 1.00045800e+00
8.53214681e-01 1.18344164e+00 -3.45889479e-02 -2.66762208e-02
1.01890206e-01 -9.37818289e-01 -1.55636206e-01 2.34866485e-01
6.24344498e-02 -6.28414869e-01 8.43599364e-02 7.26565570e-02] | [8.636580467224121, 8.70535945892334] |
cdf18f97-e3cf-4206-b587-1c82beef07eb | starss22-a-dataset-of-spatial-recordings-of | 2206.01948 | null | https://arxiv.org/abs/2206.01948v2 | https://arxiv.org/pdf/2206.01948v2.pdf | STARSS22: A dataset of spatial recordings of real scenes with spatiotemporal annotations of sound events | This report presents the Sony-TAu Realistic Spatial Soundscapes 2022 (STARS22) dataset for sound event localization and detection, comprised of spatial recordings of real scenes collected in various interiors of two different sites. The dataset is captured with a high resolution spherical microphone array and delivered in two 4-channel formats, first-order Ambisonics and tetrahedral microphone array. Sound events in the dataset belonging to 13 target sound classes are annotated both temporally and spatially through a combination of human annotation and optical tracking. The dataset serves as the development and evaluation dataset for the Task 3 of the DCASE2022 Challenge on Sound Event Localization and Detection and introduces significant new challenges for the task compared to the previous iterations, which were based on synthetic spatialized sound scene recordings. Dataset specifications are detailed including recording and annotation process, target classes and their presence, and details on the development and evaluation splits. Additionally, the report presents the baseline system that accompanies the dataset in the challenge with emphasis on the differences with the baseline of the previous iterations; namely, introduction of the multi-ACCDOA representation to handle multiple simultaneous occurences of events of the same class, and support for additional improved input features for the microphone array format. Results of the baseline indicate that with a suitable training strategy a reasonable detection and localization performance can be achieved on real sound scene recordings. The dataset is available in https://zenodo.org/record/6387880. | ['Tuomas Virtanen', 'Yuki Mitsufuji', 'Shusuke Takahashi', 'Naoya Takahashi', 'Yuichiro Koyama', 'Daniel Krause', 'Sharath Adavanne', 'Parthasaarathy Sudarsanam', 'Kazuki Shimada', 'Archontis Politis'] | 2022-06-04 | null | null | null | null | ['sound-event-localization-and-detection'] | ['audio'] | [ 2.33147621e-01 -5.73920012e-01 8.65104139e-01 -8.35925341e-02
-1.61569989e+00 -9.48229551e-01 6.11475289e-01 7.43535981e-02
-3.45237225e-01 3.18891048e-01 5.43541968e-01 1.20015934e-01
-8.15876648e-02 -2.75855631e-01 -4.70377505e-01 -6.25667810e-01
-3.51106197e-01 -3.32131013e-02 5.98722100e-01 2.08348408e-01
2.57008430e-02 3.93597633e-01 -1.96693504e+00 6.63771689e-01
-3.30831021e-01 1.32414007e+00 3.35383385e-01 1.36373389e+00
2.79838324e-01 7.72665143e-01 -9.04892743e-01 4.35027719e-01
4.03307825e-01 -2.75093764e-01 -3.66663367e-01 -4.16919351e-01
9.72815990e-01 -8.75049978e-02 -2.87448943e-01 5.86327195e-01
1.35299468e+00 6.57203555e-01 1.93652377e-01 -1.16546154e+00
3.49923462e-01 2.92925566e-01 1.18609875e-01 7.58209050e-01
9.78570044e-01 1.37538910e-02 9.45586205e-01 -1.43981922e+00
2.82788306e-01 9.18778360e-01 1.17543674e+00 2.13199347e-01
-9.76543665e-01 -8.51272047e-01 -3.75385463e-01 1.40510559e-01
-1.58669293e+00 -1.04223120e+00 4.81155664e-01 -5.63545883e-01
1.43177009e+00 8.34601045e-01 5.11449039e-01 1.33708513e+00
-5.52248001e-01 4.41627115e-01 7.91961491e-01 -4.80593860e-01
5.22408605e-01 -8.43872353e-02 -1.70798406e-01 1.07127996e-02
-3.19047511e-01 4.22533274e-01 -1.00346601e+00 -5.04469156e-01
6.18002117e-01 -6.13989055e-01 -4.40508723e-01 2.39092961e-01
-1.42642987e+00 2.83850461e-01 -5.12776747e-02 3.53709221e-01
-3.99958730e-01 2.69484520e-01 6.84961081e-01 -7.28594288e-02
5.65838933e-01 4.36130732e-01 -5.90167046e-01 -7.05151141e-01
-1.08934414e+00 4.14009929e-01 7.89817274e-01 8.93348277e-01
1.30006984e-01 4.08028424e-01 -7.00154230e-02 1.05499458e+00
3.82804751e-01 6.82114005e-01 2.81735986e-01 -1.22077847e+00
5.25216520e-01 -3.33547860e-01 3.35926294e-01 -1.08877122e+00
-6.16064012e-01 -2.77159929e-01 -1.58643469e-01 7.68486187e-02
4.21043664e-01 -3.50692540e-01 -6.66985810e-01 1.49096656e+00
3.88799429e-01 9.31820631e-01 -1.83784723e-01 8.96223187e-01
1.15921235e+00 1.10201848e+00 -9.74077210e-02 -2.01800257e-01
1.26546109e+00 -6.89517617e-01 -8.86252820e-01 2.14270502e-02
1.59430131e-01 -1.05454993e+00 9.31633055e-01 7.14727879e-01
-1.11035156e+00 -7.71071017e-01 -6.87820077e-01 2.27925420e-01
-4.74834532e-01 3.12946625e-02 3.11108023e-01 7.12013960e-01
-1.11070168e+00 -1.03820406e-01 -9.00375009e-01 -5.21229446e-01
-6.83382004e-02 -5.08322045e-02 -1.80702671e-01 4.24309582e-01
-1.09836960e+00 1.68651640e-01 -8.30199197e-02 1.11027941e-01
-1.20141625e+00 -1.02891016e+00 -8.27109516e-01 -2.05862015e-01
6.72341809e-02 1.89162716e-02 1.85325778e+00 -3.53603840e-01
-1.22886038e+00 4.93934453e-01 -9.78221446e-02 -3.28218043e-01
3.29994291e-01 -3.93758833e-01 -1.03953743e+00 4.38285291e-01
5.15045345e-01 3.63093644e-01 6.31275654e-01 -1.08394027e+00
-1.12101805e+00 1.73727497e-01 -3.02771509e-01 1.99832574e-01
1.14841172e-02 8.64541650e-01 -3.78874630e-01 -9.10950422e-01
-5.45497686e-02 -6.10569715e-01 4.18997463e-03 -2.06113487e-01
-4.22555864e-01 2.14159593e-01 8.28285515e-01 -6.98762476e-01
1.25803399e+00 -2.58155799e+00 -4.92659479e-01 6.69432208e-02
-2.41368949e-01 6.08782433e-02 -1.70872897e-01 6.61120474e-01
-4.79638606e-01 -1.46945298e-01 -9.71274525e-02 -7.44668603e-01
7.29036108e-02 -2.24569857e-01 -9.21634793e-01 5.07899046e-01
-3.16188067e-01 -3.03526223e-02 -1.02757275e+00 -1.28593907e-01
5.22607207e-01 4.64426130e-01 -5.62687993e-01 3.23482722e-01
1.54231489e-01 6.69011235e-01 -8.53039399e-02 5.71380258e-01
4.83822316e-01 4.10439223e-01 -6.76217496e-01 -1.82835445e-01
-5.83242655e-01 6.81544304e-01 -1.92038000e+00 1.63406098e+00
-6.19445682e-01 9.70556021e-01 5.09954095e-01 -3.76681775e-01
6.46966815e-01 9.85288680e-01 6.20861411e-01 -4.54291195e-01
-3.19391042e-01 4.48818356e-01 -5.03377616e-01 -8.09321821e-01
4.90658015e-01 1.53995836e-02 -1.28859252e-01 2.75836259e-01
5.83821833e-02 -5.40917218e-01 1.36890322e-01 1.94320343e-02
1.56896162e+00 -4.18166667e-02 9.39524174e-02 -7.61039555e-03
3.53297025e-01 -1.06641687e-01 2.78278083e-01 9.76893544e-01
-1.76979870e-01 1.39978456e+00 -2.45949671e-01 -2.96161830e-01
-6.87362790e-01 -1.28591502e+00 -4.58668351e-01 1.20281243e+00
-2.71148920e-01 -7.05098033e-01 -5.28110802e-01 -2.05833331e-01
-3.62232924e-01 8.53999794e-01 -3.22763324e-01 2.59606093e-01
-6.19576514e-01 -5.54523170e-01 1.29995918e+00 5.65198898e-01
4.28011179e-01 -1.28368080e+00 -9.60288048e-01 2.66008139e-01
-4.58576977e-01 -1.20852840e+00 -3.60606521e-01 3.47193897e-01
-9.96369272e-02 -9.77769971e-01 -3.18682760e-01 -7.19895899e-01
-1.03663005e-01 2.04584211e-01 9.40722525e-01 -4.86088753e-01
-5.86060882e-01 1.13067424e+00 -6.36703193e-01 -6.32956922e-01
-1.21002078e-01 -5.04047990e-01 4.44589347e-01 8.59587491e-02
4.32387516e-02 -7.16493368e-01 -6.32138968e-01 4.03603673e-01
-6.15247190e-01 -4.90921587e-01 -2.85538137e-01 3.44152510e-01
5.17437875e-01 2.35200807e-01 6.14570618e-01 -1.26073465e-01
2.09992856e-01 -7.08257735e-01 -5.11896193e-01 -4.85228091e-01
1.91583633e-01 -1.05780816e+00 5.51588595e-01 -3.05611730e-01
-1.14719725e+00 3.37718368e-01 -4.92976248e-01 -4.02999103e-01
-7.56076097e-01 1.01987615e-01 -8.18893462e-02 1.37938768e-01
9.67680097e-01 9.16868150e-02 -6.98088706e-01 -8.61371040e-01
3.87393608e-04 1.08014297e+00 9.39870775e-01 -4.04138327e-01
3.00098807e-01 5.90556085e-01 -3.54665756e-01 -1.22779560e+00
-5.97338140e-01 -9.28409278e-01 -2.79548705e-01 -3.60129237e-01
7.80456364e-01 -1.25443554e+00 -2.42600232e-01 7.66794980e-01
-1.32660711e+00 -3.97933006e-01 -6.66352272e-01 8.19229364e-01
-5.94758451e-01 -1.28040999e-01 -4.08044755e-01 -1.20621121e+00
7.84660056e-02 -8.22345853e-01 1.53891885e+00 -1.67384624e-01
-4.22817260e-01 -8.68844628e-01 6.42467320e-01 1.54117987e-01
3.84499580e-01 2.98040152e-01 8.71768370e-02 -9.39036429e-01
-2.42437333e-01 -3.50998223e-01 2.41936579e-01 3.95374119e-01
1.50536850e-01 -1.17376726e-02 -1.73435974e+00 -5.96631132e-02
2.70383835e-01 -1.56075224e-01 7.13417232e-01 5.40194869e-01
8.19660306e-01 -1.79192647e-01 -8.55020434e-02 5.69187462e-01
1.03118670e+00 5.67735553e-01 3.85708451e-01 2.01200455e-01
4.92903322e-01 4.91887391e-01 6.34536684e-01 8.25158656e-01
1.52045578e-01 1.03469408e+00 5.24966896e-01 -1.31874001e-02
-6.69258952e-01 -2.40519866e-01 4.75632071e-01 7.69373000e-01
3.01387787e-01 -4.72847790e-01 -1.08355105e+00 9.66098666e-01
-1.36188960e+00 -1.15176487e+00 -5.86228907e-01 2.16464686e+00
5.73807120e-01 -3.90124708e-01 1.91473201e-01 5.81214845e-01
8.48343253e-01 3.81424248e-01 6.04383573e-02 -2.32409716e-01
-1.81595594e-01 3.57218146e-01 6.09103367e-02 6.28833711e-01
-1.38913822e+00 3.99459660e-01 7.33505630e+00 8.63912165e-01
-9.70673144e-01 4.24775302e-01 -5.12778573e-02 -5.23902118e-01
1.84655294e-01 -4.12029654e-01 -7.07626402e-01 5.92268884e-01
1.43312907e+00 3.88943881e-01 5.39044917e-01 5.41835785e-01
6.84388459e-01 -2.61211783e-01 -9.64769363e-01 1.22306359e+00
1.67744040e-01 -1.22073269e+00 -6.03852391e-01 -3.66055459e-01
5.41856170e-01 3.85539144e-01 7.42021948e-02 1.33203015e-01
-1.42125234e-01 -6.43102407e-01 1.38074219e+00 4.37055051e-01
9.87774134e-01 -3.78532112e-01 4.24071193e-01 1.11348584e-01
-1.56169689e+00 -1.82419151e-01 5.53931035e-02 -3.53577346e-01
6.10319138e-01 4.47737753e-01 -1.03384578e+00 2.62373090e-01
1.46884000e+00 6.35189295e-01 -3.86930048e-01 1.50234950e+00
-1.33251585e-02 1.27140570e+00 -9.68036175e-01 2.38801003e-01
-7.00003430e-02 3.36406112e-01 1.31287992e+00 1.71145988e+00
7.36079514e-01 7.43358163e-03 6.39870465e-02 4.07136768e-01
3.47687006e-01 4.68005426e-03 -7.96028316e-01 4.19429392e-01
1.00631642e+00 1.16682303e+00 -4.96048450e-01 -8.92317221e-02
-2.92316437e-01 4.98363703e-01 -5.44697702e-01 4.29480135e-01
-9.58185554e-01 -5.58586538e-01 6.09608173e-01 1.94095105e-01
3.95501763e-01 -2.84500215e-02 -1.85517073e-01 -5.76289058e-01
1.51505619e-02 -7.05912888e-01 4.77094293e-01 -1.28792453e+00
-9.02160645e-01 6.66438401e-01 2.72174031e-01 -1.52004564e+00
-3.03965539e-01 -3.69285733e-01 -8.17573905e-01 5.63981593e-01
-7.88382232e-01 -9.34143364e-01 -2.92560071e-01 7.13645697e-01
7.53439605e-01 -1.57665431e-01 1.09882557e+00 9.44577098e-01
-4.23057526e-01 3.54507178e-01 1.74078077e-01 -9.46657583e-02
8.41805041e-01 -1.36534739e+00 4.49231058e-01 9.43168640e-01
6.12611294e-01 1.75855875e-01 9.25351918e-01 -4.82122809e-01
-9.54454720e-01 -1.41596901e+00 8.67727518e-01 -8.39804411e-01
8.66545916e-01 -7.47781813e-01 -4.35640752e-01 7.68681467e-01
5.20199277e-02 4.27247912e-01 6.11603200e-01 -4.04650390e-01
-3.90314721e-02 -7.73930997e-02 -1.10940599e+00 5.64422160e-02
9.35323238e-01 -8.26726794e-01 -5.06833017e-01 6.09584510e-01
6.30767703e-01 -7.93244779e-01 -6.32655859e-01 2.86394715e-01
3.63822550e-01 -8.93793523e-01 1.07984018e+00 -3.73149887e-02
-2.71413643e-02 -7.43112683e-01 -7.56745756e-01 -1.11014175e+00
1.07561104e-01 -9.00677919e-01 -4.78276089e-02 1.62232888e+00
2.97058910e-01 -4.36762303e-01 1.79314330e-01 -4.41368632e-02
-6.70352340e-01 -3.88847105e-02 -1.63785028e+00 -7.63900161e-01
-5.12705982e-01 -1.60741687e+00 3.25436652e-01 7.19888449e-01
-2.63441741e-01 7.72163421e-02 -2.56131381e-01 6.70108438e-01
3.50694358e-01 -3.61695230e-01 7.21467435e-01 -9.07910764e-01
-3.30165982e-01 4.45385929e-03 -4.98651803e-01 -9.27453637e-01
-4.41680074e-01 -5.34701824e-01 6.33129478e-01 -1.20775318e+00
-5.10208845e-01 -4.71627235e-01 -2.43802145e-01 5.02001822e-01
3.31200033e-01 6.94437504e-01 -5.94175160e-02 4.94820960e-02
-7.16347277e-01 2.20810771e-01 2.89703846e-01 3.06798905e-01
-1.65619984e-01 3.49323064e-01 -1.11575253e-01 9.29426908e-01
4.99298394e-01 -6.04828656e-01 -1.68325931e-01 -4.88468945e-01
3.76570404e-01 1.71903908e-01 9.22156990e-01 -1.47517979e+00
6.18618309e-01 3.01352829e-01 6.25106320e-02 -7.56888211e-01
8.30892503e-01 -9.50521588e-01 5.43624222e-01 -2.03577101e-01
-3.77855867e-01 2.95364726e-02 6.72523320e-01 5.09126067e-01
-3.86769593e-01 -2.83541083e-02 5.85314333e-01 -5.52673498e-03
-7.82339096e-01 -2.25616992e-01 -7.72386134e-01 2.87097007e-01
8.57486308e-01 -8.87828246e-02 -1.48188084e-01 -5.24518490e-01
-9.78289902e-01 -2.62112826e-01 -5.93678281e-02 5.20060897e-01
4.69443321e-01 -1.43242896e+00 -7.28375077e-01 2.95742840e-01
-1.79688931e-02 1.54430764e-02 4.45554793e-01 8.14253211e-01
-5.21470785e-01 3.19487154e-01 2.21820742e-01 -6.90110564e-01
-1.14854896e+00 3.83872129e-02 6.70077324e-01 4.13325101e-01
-5.43976963e-01 1.27481043e+00 1.34963736e-01 -5.20726740e-01
6.18458331e-01 -5.39572716e-01 -1.60538092e-01 2.01449275e-01
9.19020116e-01 9.74174917e-01 4.16217238e-01 -8.76001716e-01
-7.14719594e-01 4.09677744e-01 6.97084725e-01 -6.75566614e-01
1.33053231e+00 -3.99620742e-01 1.37894139e-01 8.24967504e-01
1.19381225e+00 6.91547275e-01 -1.04370081e+00 5.78624979e-02
-2.78725326e-01 -4.27184135e-01 2.56121099e-01 -1.15978634e+00
-7.08258033e-01 9.07303333e-01 9.85992551e-01 6.48243666e-01
1.12280214e+00 1.32408217e-01 6.55118167e-01 1.20855600e-01
2.78593063e-01 -8.06596458e-01 -1.31697521e-01 6.14042401e-01
1.26099145e+00 -7.59106636e-01 -4.49181497e-01 -2.81169474e-01
-2.90334344e-01 8.94910097e-01 4.31122929e-01 1.73172038e-02
8.27201843e-01 7.69182861e-01 2.86919177e-01 -3.25577766e-01
-5.80962420e-01 7.27103949e-02 8.00066963e-02 7.73402870e-01
5.33274859e-02 -1.51383191e-01 5.77509761e-01 8.75296295e-01
-4.41346675e-01 -5.07940352e-01 2.77004898e-01 8.44730616e-01
-2.59180605e-01 -3.84973705e-01 -8.95567417e-01 1.03420630e-01
-7.11373866e-01 -3.42889309e-01 -1.88331187e-01 5.23509979e-01
4.43867713e-01 1.47717738e+00 2.50012904e-01 -4.16435033e-01
8.74078035e-01 -3.09880171e-02 -2.08811928e-03 -8.51168633e-01
-9.92220581e-01 4.27694440e-01 5.21365225e-01 -7.81605721e-01
-3.02435160e-01 -1.20842385e+00 -1.25681174e+00 4.54974324e-01
-2.76264489e-01 2.33684883e-01 6.64286673e-01 5.10586500e-01
4.24139112e-01 8.13431144e-01 7.79472232e-01 -1.38342094e+00
-2.52767056e-01 -1.10253263e+00 -7.31537282e-01 6.61081672e-02
8.08896124e-01 -5.09645462e-01 -1.02891541e+00 3.74994993e-01] | [15.125121116638184, 5.2570719718933105] |
83de367b-b4e7-4dd5-8063-7b0b1a1726a8 | trust-your-nabla-gradient-based-intervention | 2211.13715 | null | https://arxiv.org/abs/2211.13715v2 | https://arxiv.org/pdf/2211.13715v2.pdf | Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery | Inferring causal structure from data is a challenging task of fundamental importance in science. Observational data are often insufficient to identify a system's causal structure uniquely. While conducting interventions (i.e., experiments) can improve the identifiability, such samples are usually challenging and expensive to obtain. Hence, experimental design approaches for causal discovery aim to minimize the number of interventions by estimating the most informative intervention target. In this work, we propose a novel Gradient-based Intervention Targeting method, abbreviated GIT, that 'trusts' the gradient estimator of a gradient-based causal discovery framework to provide signals for the intervention acquisition function. We provide extensive experiments in simulated and real-world datasets and demonstrate that GIT performs on par with competitive baselines, surpassing them in the low-data regime. | ['Piotr Miłoś', 'Łukasz Kuciński', 'Stefan Bauer', 'Yashas Annadani', 'Nino Scherrer', 'Aleksandra Nowak', 'Michał Zając', 'Mateusz Olko'] | 2022-11-24 | null | null | null | null | ['causal-discovery'] | ['knowledge-base'] | [ 4.57367986e-01 5.99281639e-02 -8.81526411e-01 -1.72957599e-01
-6.14535511e-01 -5.33176899e-01 6.72761738e-01 2.73998290e-01
-1.89696506e-01 1.00524807e+00 3.31156939e-01 -9.25364375e-01
-6.52011395e-01 -5.48841298e-01 -1.11256969e+00 -5.49795151e-01
-4.41044927e-01 4.27617580e-01 -9.84899998e-02 3.77242833e-01
3.62923384e-01 1.70730546e-01 -1.15256035e+00 -9.39070955e-02
1.09335983e+00 2.62062162e-01 -4.24437858e-02 5.30990481e-01
5.03578842e-01 5.44635177e-01 -6.71883225e-02 -3.67339104e-02
3.07299532e-02 -8.03896010e-01 -4.75843549e-01 -2.74380684e-01
8.50916579e-02 -3.88885349e-01 -2.36176834e-01 8.48528802e-01
5.88364601e-01 -1.95230424e-01 7.83773541e-01 -1.22066164e+00
-2.49795318e-01 9.53634918e-01 -8.89776826e-01 3.40416282e-01
3.43132287e-01 3.27222198e-01 1.20982325e+00 -5.50860763e-01
4.83060449e-01 1.71001971e+00 5.27763665e-01 2.46106178e-01
-1.95253158e+00 -6.84577107e-01 4.46239620e-01 -5.78129329e-02
-1.12994814e+00 -6.57414019e-01 6.73155963e-01 -5.35005271e-01
3.39111716e-01 2.42857635e-01 3.94380957e-01 1.39414489e+00
1.17904030e-01 8.04136753e-01 1.34641290e+00 -2.79799759e-01
6.01705551e-01 -2.55317181e-01 7.41056874e-02 6.39068902e-01
6.75432682e-01 8.40363979e-01 -7.30055511e-01 -6.65264487e-01
8.02580297e-01 -1.37247387e-02 -4.30863619e-01 -5.28293073e-01
-1.10051501e+00 1.11641192e+00 3.14574331e-01 -7.59945363e-02
-5.91144383e-01 5.19349098e-01 4.04406190e-01 3.72286677e-01
4.73468512e-01 6.21066809e-01 -5.37622690e-01 -1.32714078e-01
-6.08998120e-01 5.01325369e-01 6.24901354e-01 6.02031946e-01
2.59013325e-01 -2.08250418e-01 -5.67218661e-01 3.66358399e-01
3.49711150e-01 6.08592689e-01 -2.99421370e-01 -6.45921052e-01
4.00462389e-01 6.29970312e-01 5.75426579e-01 -8.50527704e-01
-5.01729548e-01 -3.09095979e-01 -8.81548107e-01 -1.39488041e-01
8.07710171e-01 -3.96623969e-01 -6.39880776e-01 1.93395054e+00
6.91394508e-01 6.52498245e-01 -5.48119366e-01 8.85596395e-01
1.44294262e-01 3.38492244e-01 3.21129352e-01 -6.89371347e-01
1.12952065e+00 -1.39919072e-01 -5.68531394e-01 -2.26645768e-01
4.41624671e-01 -3.57630491e-01 1.34452903e+00 2.95391411e-01
-7.65765250e-01 2.76211053e-02 -5.76876402e-01 5.96867383e-01
2.56869316e-01 1.51595036e-02 9.59398627e-01 6.81075811e-01
-5.19693851e-01 7.04319417e-01 -9.99231040e-01 -1.63475156e-01
7.65532434e-01 3.56534749e-01 4.25362512e-02 -1.81846619e-01
-1.05129755e+00 4.31749701e-01 1.32449538e-01 -3.42176557e-02
-1.62740529e+00 -1.57883155e+00 -4.24961090e-01 2.17948064e-01
8.79901469e-01 -8.04461241e-01 1.15617394e+00 -3.51887792e-01
-1.33174062e+00 1.80183902e-01 -2.14708254e-01 -5.77660739e-01
5.44923604e-01 -2.57011920e-01 1.94507226e-01 -1.28466249e-01
1.46805063e-01 1.56051680e-01 9.31978226e-01 -1.22279203e+00
-7.60154068e-01 -4.58213627e-01 6.79368898e-02 -1.99929506e-01
-1.04822114e-01 1.15174115e-01 1.39975861e-01 -4.04173970e-01
-1.57117739e-01 -8.21734369e-01 -6.26801789e-01 -2.82073170e-01
-8.21648002e-01 -2.25659639e-01 6.85774088e-01 -4.47157174e-01
1.54243350e+00 -1.74602222e+00 5.05924337e-02 2.82719463e-01
4.06228989e-01 -1.15302071e-01 1.09632283e-01 5.35257578e-01
-1.36279196e-01 3.13890725e-01 -3.13109159e-01 2.04102173e-01
4.44415025e-02 -3.57715450e-02 -3.44709575e-01 8.21925581e-01
2.59445369e-01 8.45128000e-01 -1.19546926e+00 -2.38058493e-01
1.60919443e-01 1.30997524e-01 -8.22119892e-01 3.08303833e-01
-3.39854062e-01 9.11710560e-01 -8.29164743e-01 3.55301768e-01
2.54171163e-01 -3.76992643e-01 6.56184137e-01 1.23422153e-01
-1.91430509e-01 3.13023239e-01 -1.13841748e+00 9.07381713e-01
-5.29090047e-01 2.42918864e-01 2.17297184e-03 -1.72318614e+00
3.77966851e-01 3.36414129e-01 5.83353043e-01 -4.93974000e-01
7.51971528e-02 9.33053158e-03 3.51410657e-01 -5.42077303e-01
-3.79951954e-01 -1.31464660e-01 -2.89569199e-01 4.90421146e-01
-3.44999999e-01 2.18087003e-01 3.19621377e-02 6.74637184e-02
1.50594175e+00 -1.44628599e-01 6.95454836e-01 -7.03697205e-01
1.50415115e-02 -1.36578575e-01 7.65150189e-01 1.26870608e+00
1.86505727e-03 -9.92994457e-02 1.10403705e+00 -2.60415584e-01
-9.68278944e-01 -1.15235198e+00 -2.15021625e-01 8.12282443e-01
-1.87684223e-01 -4.49017175e-02 -6.11999393e-01 -8.34329486e-01
3.93942952e-01 8.10954154e-01 -8.63048434e-01 -4.01542753e-01
-4.45822626e-01 -1.18301499e+00 2.23890617e-01 2.82961935e-01
1.23073786e-01 -5.18285751e-01 -6.20257080e-01 3.84432703e-01
1.26020228e-02 -6.10328853e-01 -5.76689899e-01 2.18174875e-01
-8.56566608e-01 -1.48426938e+00 -3.18115860e-01 -1.06960729e-01
7.11480975e-01 1.88778743e-01 1.19640028e+00 -2.59751678e-01
-2.30416998e-01 1.46141008e-01 -1.89682379e-01 -5.98901749e-01
-3.85371596e-01 -1.46093056e-01 2.41913438e-01 -2.05942281e-02
1.74322993e-01 -7.31134117e-01 -8.86194289e-01 3.26040387e-01
-5.02334237e-01 -2.65462816e-01 6.39795899e-01 1.22882891e+00
3.34430605e-01 2.40185320e-01 9.98873472e-01 -1.20568645e+00
8.15776765e-01 -7.57512450e-01 -1.21214938e+00 1.45983920e-01
-1.03461421e+00 2.25482091e-01 7.50009239e-01 -7.33518720e-01
-1.07473373e+00 -8.31464976e-02 5.29267311e-01 -1.20056726e-01
-2.73656603e-02 8.39356422e-01 -2.35540390e-01 1.82255909e-01
8.12652051e-01 -4.57676560e-01 -2.83299387e-01 -4.53849286e-01
3.47120404e-01 3.41620833e-01 1.39114872e-01 -9.71379399e-01
6.30673230e-01 3.76873791e-01 3.19499582e-01 -4.82829720e-01
-8.50250244e-01 -2.97126591e-01 -1.74566001e-01 -2.31262133e-01
3.80526662e-01 -7.35810101e-01 -1.30838525e+00 -9.32498276e-02
-7.72401571e-01 -5.73936582e-01 -1.56248465e-01 7.82346964e-01
-4.40348685e-01 -1.42812848e-01 -2.75501132e-01 -1.06115377e+00
8.58117267e-03 -1.03480732e+00 1.05406141e+00 1.20059058e-01
-2.54541099e-01 -1.10232425e+00 2.74183095e-01 -6.97202310e-02
4.79770824e-02 3.69779319e-01 1.04725683e+00 -3.15048635e-01
-5.13024807e-01 -1.45858511e-01 -2.87630618e-01 -4.27689433e-01
2.90178835e-01 5.23043387e-02 -7.36084640e-01 -3.44250262e-01
-1.16368920e-01 -1.06611758e-01 6.99118912e-01 1.22003007e+00
1.20145202e+00 -6.25413775e-01 -6.88969314e-01 1.53091744e-01
1.13289928e+00 3.11297446e-01 3.26253265e-01 -1.63652658e-01
6.07649744e-01 5.57762027e-01 7.59259641e-01 8.42722595e-01
1.54742599e-01 7.56629407e-01 2.11710259e-01 -1.39117867e-01
2.86388725e-01 -7.91941524e-01 2.99122125e-01 1.87299575e-03
5.36181368e-02 -2.20949054e-01 -9.01300669e-01 5.03986478e-01
-2.04748464e+00 -8.88072729e-01 -4.79867071e-01 2.71835542e+00
1.21480775e+00 2.18043864e-01 5.06080627e-01 -1.14561833e-01
6.64041758e-01 -3.21151465e-01 -9.69581902e-01 -1.64511636e-01
2.42819190e-01 1.00510173e-01 8.71601820e-01 5.23153305e-01
-9.64414716e-01 6.00863636e-01 7.02773094e+00 6.41478062e-01
-7.04785168e-01 1.04910724e-01 7.92416215e-01 9.75876451e-02
-3.62127423e-01 3.58603626e-01 -7.18296468e-01 6.04243934e-01
1.40420794e+00 -5.02192557e-01 5.56804121e-01 4.31781352e-01
1.23620641e+00 -3.81008953e-01 -1.44640851e+00 3.52792531e-01
-8.94166291e-01 -1.17139339e+00 -3.01774144e-01 2.26872265e-01
7.82095850e-01 -3.24935019e-01 -4.18254063e-02 2.57058311e-02
1.12995744e+00 -1.03188109e+00 3.44705969e-01 4.56243813e-01
6.22247040e-01 -6.67920649e-01 3.99956495e-01 3.88013214e-01
-8.53970349e-01 -3.32953244e-01 -1.28312469e-01 -2.77979016e-01
2.59619709e-02 1.29283953e+00 -9.22661483e-01 3.13814372e-01
6.18073702e-01 6.60573304e-01 -1.05517611e-01 1.01298773e+00
-4.08043146e-01 1.46023214e+00 -3.64042938e-01 -1.15744673e-01
-1.66738015e-02 -3.50056261e-01 5.91852188e-01 7.77919590e-01
1.75986513e-01 1.28071979e-01 2.59097248e-01 1.11219275e+00
-1.00665860e-01 -1.80003986e-01 -8.64962757e-01 -4.16644633e-01
7.06004858e-01 6.53530896e-01 -6.01894259e-01 -1.39182121e-01
-1.96887881e-01 3.45992237e-01 4.13977243e-02 3.96335334e-01
-9.75589514e-01 2.45207652e-01 5.66697776e-01 -1.64643526e-02
4.93927114e-02 9.64400694e-02 -3.73339325e-01 -7.88480043e-01
-2.50202239e-01 -1.17859626e+00 6.32601678e-01 -7.69946799e-02
-1.18678319e+00 -5.94224930e-01 4.80427414e-01 -8.22817028e-01
-1.77958384e-01 -2.92957753e-01 -4.90723044e-01 7.34685659e-01
-9.72879767e-01 -5.88589609e-01 2.96240538e-01 2.57548660e-01
2.47098193e-01 4.34175283e-01 4.18793201e-01 1.99130818e-01
-1.00679505e+00 4.95678186e-01 3.72760475e-01 -3.94748867e-01
5.60602665e-01 -1.47106028e+00 1.51008070e-01 7.30126739e-01
-2.15450421e-01 6.65682018e-01 1.17839086e+00 -9.35291052e-01
-1.79147673e+00 -8.78825605e-01 3.48666638e-01 -1.96026102e-01
1.08536017e+00 -5.34868777e-01 -7.40230322e-01 5.65223873e-01
-1.47673458e-01 -3.21831375e-01 3.24904174e-01 7.39332974e-01
-1.90135032e-01 -2.35887006e-01 -1.04810226e+00 9.28027868e-01
1.26998258e+00 -2.38773480e-01 -1.47112653e-01 5.67337155e-01
6.78845286e-01 3.34321000e-02 -9.32049692e-01 3.88770878e-01
4.43059206e-01 -5.53476036e-01 1.05798090e+00 -9.89335716e-01
4.99228150e-01 -1.43992335e-01 2.41653562e-01 -1.58402205e+00
-1.46077916e-01 -8.95190477e-01 -1.46052673e-01 1.18312061e+00
4.29477781e-01 -7.51717269e-01 6.88723505e-01 5.00771046e-01
3.65167022e-01 -4.58934635e-01 -8.62287939e-01 -7.90732861e-01
1.03646785e-01 -2.20193669e-01 3.78039747e-01 1.10693038e+00
1.97674837e-02 7.19493330e-01 -5.38937509e-01 5.00759006e-01
1.14581823e+00 1.99430227e-01 7.62887836e-01 -1.25436628e+00
-4.13851470e-01 -5.29440403e-01 -1.46995357e-03 -6.45294189e-01
1.88549772e-01 -3.82998347e-01 1.58848554e-01 -9.87804949e-01
5.62401175e-01 -5.99278867e-01 -8.85379538e-02 3.10307294e-01
-8.04255009e-01 -5.52003741e-01 -5.48034787e-01 -1.48173779e-01
-9.02835876e-02 6.13362789e-01 1.00461400e+00 -2.08474591e-01
-4.74738359e-01 2.46110097e-01 -8.07510555e-01 5.95394552e-01
7.91485608e-01 -8.32934141e-01 -8.28295231e-01 1.32115215e-01
2.69426435e-01 5.44797838e-01 6.94477379e-01 -2.78308690e-01
-1.06513433e-01 -7.26435184e-01 -3.59618403e-02 -2.70180404e-01
-5.46742022e-01 -6.16322219e-01 3.31454515e-01 8.62395942e-01
-7.14390397e-01 -2.00089261e-01 1.27550974e-01 1.03462017e+00
2.57789969e-01 1.22331614e-02 5.52295327e-01 2.55442828e-01
-9.38093886e-02 2.84309894e-01 -2.33443081e-01 1.89533561e-01
7.09933639e-01 6.44892931e-01 -4.82931525e-01 -3.66707027e-01
-2.25761488e-01 5.81668317e-01 2.53073759e-02 8.64404142e-02
4.07842755e-01 -1.14267886e+00 -9.53802347e-01 -3.77617300e-01
1.49331836e-03 -2.45206952e-01 1.03293151e-01 1.27781153e+00
2.99272358e-01 4.88393426e-01 5.00405550e-01 -6.33080482e-01
-1.03711820e+00 9.40666080e-01 2.34338641e-01 -4.94156867e-01
-6.39092088e-01 5.51881254e-01 7.99828351e-01 -3.51799935e-01
-7.17518060e-03 -2.62590051e-01 9.08360779e-02 -1.16334975e-01
6.21666610e-01 4.46369916e-01 -3.00769418e-01 3.56145024e-01
-3.24412465e-01 -1.02044456e-01 4.43323031e-02 -4.79655042e-02
1.28248155e+00 -1.18905753e-01 4.07699682e-03 5.10571539e-01
9.09581304e-01 -1.76733971e-01 -1.50910652e+00 -1.85139611e-01
5.01466513e-01 -6.85745060e-01 4.27443415e-01 -7.72678375e-01
-7.44416952e-01 5.12493610e-01 6.21493042e-01 4.27604944e-01
1.03744245e+00 1.33913327e-02 -5.92769347e-02 1.88070863e-01
3.33832175e-01 -7.53107250e-01 -1.16376027e-01 -2.92460412e-01
8.35078955e-01 -1.29590690e+00 2.15643123e-01 -4.04988617e-01
-4.78284284e-02 5.47103703e-01 3.15135777e-01 -2.68239938e-02
6.38302743e-01 3.12975854e-01 -5.81146896e-01 -4.43454117e-01
-1.02225113e+00 -9.81183946e-02 1.60158053e-01 4.86178398e-01
3.38067740e-01 4.55127478e-01 -6.58220470e-01 3.82559687e-01
3.22123349e-01 1.42974600e-01 4.46306497e-01 6.60871327e-01
-1.71554238e-01 -9.83770370e-01 -6.49057150e-01 9.55232024e-01
-4.32066739e-01 -1.62373215e-01 -3.79632264e-01 9.53221440e-01
-5.45945287e-01 1.25075388e+00 -2.80643731e-01 -6.23281896e-02
5.06381571e-01 -3.91802430e-01 3.24711680e-01 -3.28327596e-01
-1.27372861e-01 2.71571785e-01 2.67940551e-01 -8.59480143e-01
-4.94742632e-01 -1.19395578e+00 -7.53347814e-01 -5.77458620e-01
-4.11699027e-01 2.75283784e-01 2.43243933e-01 1.00931835e+00
3.40038955e-01 7.67312527e-01 1.20550609e+00 -4.65773404e-01
-8.40681076e-01 -7.80402601e-01 -4.73878562e-01 5.65238409e-02
3.20767909e-01 -1.09970570e+00 -5.20095289e-01 4.26088162e-02] | [7.86415958404541, 5.243180274963379] |
c2ec22ae-a2c1-4c16-8536-78e8d1df2eab | matrix-completion-with-heterogonous-cost | 2203.12120 | null | https://arxiv.org/abs/2203.12120v3 | https://arxiv.org/pdf/2203.12120v3.pdf | Matrix Completion with Heterogonous Cost | The matrix completion problem has been studied broadly under many underlying conditions. The problem has been explored under adaptive or non-adaptive, exact or estimation, single-phase or multi-phase, and many other categories. In most of these cases, the observation cost of each entry is uniform and has the same cost across the columns. However, in many real-life scenarios, we could expect elements from distinct columns or distinct positions to have a different cost. In this paper, we explore this generalization under adaptive conditions. We approach the problem under two different cost models. The first one is that entries from different columns have different observation costs, but, within the same column, each entry has a uniform cost. The second one is any two entry has different observation cost, despite being the same or different columns. We provide complexity analysis of our algorithms and provide tightness guarantees. | ['Ilqar Ramazanli'] | 2022-03-23 | null | null | null | null | ['matrix-completion'] | ['methodology'] | [ 4.19345796e-01 5.42996824e-02 -3.09779018e-01 1.17183469e-01
-5.60460091e-01 -9.56966579e-01 1.67015091e-01 4.70083177e-01
-2.44449854e-01 8.13863158e-01 8.38375092e-02 -2.45733514e-01
-4.89703447e-01 -6.39606595e-01 -7.81064868e-01 -9.67002571e-01
-4.65513259e-01 7.25560069e-01 1.57973275e-01 -3.85881141e-02
3.03124338e-01 2.51871258e-01 -8.09332430e-01 -1.73439130e-01
5.58846295e-01 9.05665636e-01 2.07337797e-01 6.59219027e-01
1.18492313e-01 1.84959590e-01 -3.89882863e-01 -2.67003924e-01
9.51957405e-01 -3.60520840e-01 -5.34638584e-01 3.28606188e-01
1.38795882e-01 2.09217779e-02 -3.26422483e-01 1.34990478e+00
3.93215775e-01 1.26955444e-02 4.79471534e-01 -1.39972186e+00
-1.09146126e-01 6.81243598e-01 -1.16126215e+00 -3.08562275e-02
5.18737376e-01 -3.23622435e-01 9.74053621e-01 -7.70073771e-01
5.16779959e-01 6.99438512e-01 5.38797975e-01 4.72744461e-03
-1.56849265e+00 -4.80679661e-01 2.63148844e-01 8.23805705e-02
-1.59204590e+00 -4.47830170e-01 4.96739864e-01 -2.64442801e-01
1.12289693e-02 3.28252316e-01 4.58328456e-01 5.99184215e-01
2.13959038e-01 4.59521264e-01 1.22046697e+00 -5.56174338e-01
4.49007183e-01 1.38126567e-01 -1.22435749e-01 2.83779442e-01
1.05687010e+00 -1.97824880e-01 -6.09176457e-01 -5.37610710e-01
3.83824974e-01 2.43052408e-01 -6.48837566e-01 -8.31993043e-01
-1.35424542e+00 7.47026742e-01 3.39160673e-02 3.31892133e-01
-3.94672841e-01 1.61447138e-01 1.02693386e-01 4.24790382e-01
-2.53202468e-01 2.29781628e-01 -4.18606579e-01 6.25727847e-02
-9.77666914e-01 3.37141156e-01 1.19503081e+00 1.35522294e+00
9.64388013e-01 -2.86729604e-01 -7.35239163e-02 3.99496824e-01
1.00918971e-01 6.26645148e-01 1.28043950e-01 -7.80921996e-01
1.00458014e+00 2.94299964e-02 4.67474788e-01 -1.19774973e+00
-5.09062231e-01 -5.74259818e-01 -1.14929187e+00 -1.24438040e-01
6.41216397e-01 -4.65021670e-01 -4.95038271e-01 1.98927474e+00
2.45086953e-01 1.83600143e-01 5.80734238e-02 8.74231994e-01
-1.36720866e-01 5.63524842e-01 -6.23360157e-01 -7.83171713e-01
1.31103683e+00 -7.93146789e-01 -6.79324329e-01 -5.43236017e-01
3.75309795e-01 -9.29270744e-01 2.82355517e-01 5.72477043e-01
-1.12107980e+00 4.93593253e-02 -1.28983462e+00 6.30990148e-01
-1.05955703e-02 -1.00437449e-02 4.17802572e-01 7.30284095e-01
-8.79294515e-01 3.08509886e-01 -5.81362367e-01 -1.73279881e-01
-3.02846104e-01 4.04041350e-01 -4.00580585e-01 -5.57140946e-01
-8.16544294e-01 3.41704100e-01 1.49110943e-01 9.53502730e-02
-6.28183186e-01 -4.64956671e-01 -6.68493450e-01 2.67171860e-01
8.72173011e-01 -4.59843129e-01 1.05083728e+00 -7.96665907e-01
-8.08446050e-01 4.98054147e-01 -3.10172915e-01 -3.87649924e-01
5.03926456e-01 -5.82093038e-02 -2.91970551e-01 -3.33041511e-02
4.05600667e-01 -5.94494939e-02 6.76044762e-01 -1.11814964e+00
-5.02952695e-01 -6.23543978e-01 6.61125481e-02 3.50267380e-01
-1.15006126e-01 -2.94420481e-01 -7.87869155e-01 -6.53806210e-01
7.22514808e-01 -1.41897368e+00 -6.71485603e-01 -4.68239523e-02
-8.07731748e-01 4.64535683e-01 3.20827186e-01 -4.04411322e-03
1.28931391e+00 -2.23633218e+00 3.79064471e-01 6.88501954e-01
3.09222251e-01 -3.98065895e-01 1.22127317e-01 9.89497066e-01
-4.13836390e-02 -1.30735477e-02 -4.52919453e-01 -2.73405284e-01
6.85340092e-02 2.10647196e-01 -1.95567802e-01 8.21367860e-01
-3.59626323e-01 7.28080571e-02 -8.36499870e-01 -1.44440845e-01
-2.03140929e-01 -5.36466204e-02 -6.65891647e-01 -1.77534893e-01
2.93191254e-01 2.80823410e-01 -4.53433394e-01 4.06542957e-01
1.20013547e+00 -3.53877485e-01 9.73325014e-01 -1.51507109e-01
-5.48951030e-02 2.93271318e-02 -2.32244802e+00 1.35997283e+00
-2.53992707e-01 5.90286553e-01 4.82026279e-01 -1.15108311e+00
1.74347430e-01 4.50353950e-01 7.19898403e-01 -5.91647625e-01
-1.97793409e-01 4.32419866e-01 1.80724517e-01 -1.65534616e-01
8.50421488e-01 -2.95843016e-02 -3.44019741e-01 7.42860556e-01
-4.18716073e-01 4.00414437e-01 4.04166996e-01 5.69339156e-01
1.33481598e+00 -5.46550333e-01 5.42976081e-01 -3.09324235e-01
2.88331926e-01 -2.60832936e-01 9.58480895e-01 1.27765560e+00
-1.62172720e-01 8.18437696e-01 5.34056842e-01 5.27906306e-02
-9.14050281e-01 -1.01040149e+00 -1.19622238e-01 5.31772137e-01
7.71418691e-01 -6.25286281e-01 -3.30552429e-01 -2.51341611e-01
2.27204375e-02 6.21876977e-02 -4.42207128e-01 1.41050503e-01
-4.25663769e-01 -8.18177164e-01 -7.93756843e-02 2.08581850e-01
3.15144300e-01 -2.77103394e-01 -5.42237878e-01 5.03555059e-01
-3.06897402e-01 -1.17663383e+00 -7.44655788e-01 5.66159308e-01
-6.87486589e-01 -1.21753860e+00 -6.87781572e-01 -4.01933283e-01
8.28067362e-01 8.76770973e-01 8.73815715e-01 4.26151641e-02
-7.50509202e-02 4.59576726e-01 -4.37189817e-01 -1.98461354e-01
2.59026051e-01 -5.97618222e-02 4.73797709e-01 2.34445140e-01
-4.31238711e-02 -3.53858888e-01 -8.23578417e-01 3.88630897e-01
-9.35809135e-01 -3.57788742e-01 4.76489335e-01 6.56352162e-01
8.67208958e-01 4.21949804e-01 2.48322904e-01 -1.06694019e+00
3.32385540e-01 -8.60699117e-01 -6.70070648e-01 1.94792286e-01
-7.01426983e-01 2.28442118e-01 6.34069502e-01 -3.88485312e-01
-5.58334470e-01 9.80492756e-02 3.69020909e-01 -2.85647482e-01
1.91182315e-01 6.37781560e-01 -3.26241553e-01 -2.40929518e-02
2.76298940e-01 2.66521960e-01 -3.17025453e-01 -3.74373615e-01
2.12750509e-01 1.90164164e-01 2.73677945e-01 -6.12972975e-01
1.19507694e+00 6.01660967e-01 4.38600510e-01 -6.87587619e-01
-5.01948833e-01 -6.45783663e-01 -2.90837348e-01 1.03649512e-01
3.90189737e-01 -9.79532182e-01 -4.98425126e-01 2.55871594e-01
-8.05361569e-01 -9.40362141e-02 -2.00596735e-01 6.01149619e-01
-3.62659305e-01 6.71133697e-01 -1.46448031e-01 -9.00372684e-01
1.24499999e-01 -1.40848005e+00 7.29311168e-01 2.78521985e-01
7.37694800e-02 -8.07339549e-01 1.24120094e-01 -2.49811053e-01
3.16208661e-01 1.65921286e-01 5.69222808e-01 -6.97271407e-01
-9.31668282e-01 -3.33742350e-01 6.24933839e-02 -2.90672421e-01
1.05310552e-01 -3.55781585e-01 -1.42306805e-01 -7.64704525e-01
-3.12553123e-02 5.14580756e-02 3.99947435e-01 4.89828318e-01
1.06619716e+00 -6.11274481e-01 -7.15725958e-01 4.68346119e-01
1.76873255e+00 1.40979782e-01 5.89210570e-01 3.96917909e-01
2.65652895e-01 5.41407347e-01 7.08235145e-01 7.59989083e-01
2.37676442e-01 1.02316427e+00 4.88031954e-01 1.15978755e-01
5.48600912e-01 2.06043914e-01 2.10497409e-01 6.31819785e-01
1.57772452e-01 -6.31563723e-01 -5.13082147e-01 6.24509037e-01
-1.92153108e+00 -9.82724071e-01 -4.05254394e-01 2.75699282e+00
4.04447019e-01 1.56222567e-01 -2.69163735e-02 2.98078418e-01
8.17575872e-01 1.77919492e-01 -4.49327707e-01 -1.90275535e-01
-2.85552680e-01 4.27915901e-02 1.13376713e+00 2.69590735e-01
-6.42910242e-01 4.61802408e-02 6.74468136e+00 5.99183142e-01
-6.04694664e-01 -1.31505709e-02 3.87317330e-01 -2.28086501e-01
-2.22275764e-01 5.31409621e-01 -8.61544549e-01 6.53054476e-01
6.02935255e-01 -3.32280308e-01 4.70473558e-01 6.54991388e-01
-1.23387821e-01 -5.40887892e-01 -1.27805948e+00 8.85982931e-01
1.09027341e-01 -1.09146833e+00 -2.92401314e-01 5.80808699e-01
9.27536011e-01 -1.68181419e-01 -3.27649117e-02 -4.50924411e-02
2.06778690e-01 -4.88713801e-01 6.96457207e-01 2.10677579e-01
7.96258032e-01 -7.50797451e-01 5.50530493e-01 3.83562803e-01
-1.47893000e+00 -1.61779568e-01 -3.11376214e-01 -1.83857545e-01
3.12093586e-01 9.55302477e-01 -2.44761720e-01 1.07145298e+00
5.40678501e-01 1.39457434e-01 -1.27431527e-01 1.39531493e+00
1.57241121e-01 4.58001614e-01 -6.24478340e-01 2.66319871e-01
1.11334860e-01 -6.41058683e-01 7.53943026e-01 8.22387636e-01
6.47550583e-01 1.41044065e-01 6.18502617e-01 7.53167272e-02
-8.87948200e-02 1.77655164e-02 -6.84789240e-01 2.10289091e-01
1.00463080e+00 1.10900211e+00 -7.13196039e-01 -2.35876337e-01
-8.38277340e-01 8.69862914e-01 9.83567685e-02 4.30538893e-01
-6.83841705e-01 -4.92710471e-01 6.32808566e-01 2.31715545e-01
4.52107280e-01 -3.00071597e-01 -2.65059888e-01 -1.17732632e+00
1.94708824e-01 -7.15699017e-01 5.24133980e-01 -2.73946732e-01
-1.08717370e+00 1.76606864e-01 1.18487608e-02 -1.45311749e+00
-1.60488531e-01 -4.59470779e-01 -4.62961167e-01 7.26993859e-01
-1.17302537e+00 -3.10093880e-01 -5.73180504e-02 5.45438826e-01
2.95607865e-01 -1.25820026e-01 4.33316201e-01 5.37330210e-01
-6.68557227e-01 6.27880335e-01 6.28644764e-01 -8.83364305e-02
7.12952793e-01 -1.17575157e+00 -1.95600241e-01 1.07390678e+00
2.50506938e-01 6.86111212e-01 9.74784791e-01 -4.08124655e-01
-1.84098387e+00 -8.16016436e-01 6.62081420e-01 -1.59434192e-02
5.71220398e-01 -2.77812034e-01 -4.97566432e-01 9.22627032e-01
2.86105543e-01 -7.54233599e-02 7.17912138e-01 1.82540417e-01
-2.10142076e-01 -4.94334251e-02 -1.01971769e+00 5.55407345e-01
7.44599104e-01 -2.94579208e-01 -4.21996191e-02 5.71271777e-01
1.99866086e-01 -4.48563755e-01 -6.77136064e-01 4.74383026e-01
5.07433832e-01 -9.82211113e-01 9.52269435e-01 -3.24390322e-01
4.34301607e-02 -6.46156907e-01 -6.38188958e-01 -1.22774208e+00
-3.93992871e-01 -5.87866545e-01 -3.19740981e-01 9.18205321e-01
5.39316833e-01 -6.86700106e-01 8.65923524e-01 5.68729520e-01
5.75399175e-02 -5.66194177e-01 -1.10305262e+00 -1.05865455e+00
-2.96357691e-01 -2.60537893e-01 5.60037911e-01 7.84398198e-01
-2.40618140e-02 4.57714319e-01 -7.20472574e-01 5.28366506e-01
7.71142244e-01 5.37658870e-01 9.34702873e-01 -1.05789745e+00
-7.89773226e-01 -2.68568456e-01 -2.11697862e-01 -1.38917029e+00
-4.91229296e-01 -6.45065367e-01 3.28282043e-02 -1.35419190e+00
5.35850048e-01 -8.50120664e-01 -6.52370751e-02 1.91189926e-02
-1.86209723e-01 2.57366985e-01 2.81838119e-01 3.55761558e-01
-5.53004265e-01 7.01239984e-03 7.05400407e-01 -1.22458495e-01
-1.25788286e-01 2.94646055e-01 -7.89944351e-01 5.05609632e-01
6.35632217e-01 -7.27971911e-01 -4.93811399e-01 -3.08708876e-01
5.86342454e-01 5.66332400e-01 -1.13109581e-01 -1.03443313e+00
4.65056807e-01 -1.96982160e-01 2.07370251e-01 -6.78634584e-01
3.90027195e-01 -1.21999252e+00 6.39889181e-01 6.62549675e-01
1.26515895e-01 2.76104391e-01 -2.49937519e-01 1.19056261e+00
2.51037385e-02 -6.09946787e-01 6.31234109e-01 2.08172258e-02
-2.09819376e-01 5.39286911e-01 -3.85212243e-01 1.84319913e-01
1.23130441e+00 -4.60219115e-01 -1.38340548e-01 -6.01915777e-01
-7.42273927e-01 3.72938663e-01 6.41515434e-01 1.26431230e-02
2.00546697e-01 -1.33583665e+00 -5.57371378e-01 -5.36261313e-02
2.87501305e-01 -1.99786294e-02 3.68977487e-01 1.30208397e+00
-1.80169389e-01 2.61753380e-01 2.47111812e-01 -6.30311072e-01
-9.88395989e-01 7.52657533e-01 8.04111455e-03 -5.28940618e-01
-3.24769557e-01 4.42184895e-01 3.67716551e-01 -2.01649703e-02
5.57243451e-02 2.82183945e-01 1.53678179e-01 5.68420393e-03
5.39913237e-01 4.13687825e-01 3.76736224e-02 -5.85521758e-01
-2.38279372e-01 6.57593191e-01 -1.97531953e-01 -3.66562992e-01
9.19647157e-01 -4.16465282e-01 1.12453252e-02 3.92242074e-01
8.40464234e-01 7.29077995e-01 -9.02374983e-01 -5.28626025e-01
-4.27840427e-02 -7.41366327e-01 -4.27562028e-01 -2.98127383e-01
-1.20343351e+00 2.77302504e-01 3.66678417e-01 5.58718681e-01
1.16629863e+00 -1.89183727e-01 3.78048837e-01 3.81820679e-01
1.04788733e+00 -1.10720754e+00 -8.11792016e-02 4.03376281e-01
4.26630199e-01 -1.10719728e+00 2.99725682e-01 -6.11855447e-01
-4.04513955e-01 8.44386697e-01 3.85724574e-01 -1.30491897e-01
5.47007620e-01 3.12384427e-01 -5.70156276e-01 -2.10922018e-01
-6.32215559e-01 1.69696531e-03 -9.44818407e-02 4.12607163e-01
2.88346559e-01 2.25555688e-01 -6.25325739e-01 2.91890204e-01
1.50826618e-01 -5.16503036e-01 1.10021198e+00 1.16684020e+00
-2.41280302e-01 -1.38870835e+00 -7.65069067e-01 5.91698110e-01
-4.75014985e-01 -1.35253713e-01 -1.03347018e-01 6.47686839e-01
-1.96737275e-01 1.03736746e+00 -4.19426709e-02 -1.34038761e-01
2.85001278e-01 -2.49431148e-01 5.65580249e-01 -6.98965490e-01
-1.42810270e-01 3.34173024e-01 6.11446276e-02 -4.10712183e-01
-3.28789860e-01 -9.12598372e-01 -1.02791893e+00 -4.73515004e-01
-4.09406424e-01 4.56782222e-01 6.17725432e-01 7.64282286e-01
3.75819057e-01 2.20008001e-01 1.02782071e+00 -5.41466951e-01
-6.72146142e-01 -7.66691446e-01 -1.08876383e+00 -2.26209443e-02
3.65775108e-01 -7.20474839e-01 -4.50319827e-01 -2.66386360e-01] | [6.936349391937256, 4.711359977722168] |
f840cb58-1287-44a5-8f11-8d3837d16737 | spanproto-a-two-stage-span-based-prototypical | 2210.09049 | null | https://arxiv.org/abs/2210.09049v2 | https://arxiv.org/pdf/2210.09049v2.pdf | SpanProto: A Two-stage Span-based Prototypical Network for Few-shot Named Entity Recognition | Few-shot Named Entity Recognition (NER) aims to identify named entities with very little annotated data. Previous methods solve this problem based on token-wise classification, which ignores the information of entity boundaries, and inevitably the performance is affected by the massive non-entity tokens. To this end, we propose a seminal span-based prototypical network (SpanProto) that tackles few-shot NER via a two-stage approach, including span extraction and mention classification. In the span extraction stage, we transform the sequential tags into a global boundary matrix, enabling the model to focus on the explicit boundary information. For mention classification, we leverage prototypical learning to capture the semantic representations for each labeled span and make the model better adapt to novel-class entities. To further improve the model performance, we split out the false positives generated by the span extractor but not labeled in the current episode set, and then present a margin-based loss to separate them from each prototype region. Experiments over multiple benchmarks demonstrate that our model outperforms strong baselines by a large margin. | ['Chengyu Wang', 'Chengcheng Han', 'Ming Gao', 'Jun Huang', 'Songfang Huang', 'Minghui Qiu', 'Chuanqi Tan', 'Jianing Wang'] | 2022-10-17 | null | null | null | null | ['few-shot-ner'] | ['natural-language-processing'] | [-2.27989912e-01 1.58130437e-01 -4.48465109e-01 -3.88426363e-01
-1.13086677e+00 -6.85641170e-01 3.40529174e-01 5.14004648e-01
-6.51432574e-01 7.13163912e-01 4.04390037e-01 3.99003848e-02
3.18549186e-01 -8.27347517e-01 -5.21430433e-01 -5.13648510e-01
-9.55278948e-02 4.35929596e-01 5.44075370e-01 1.19744621e-01
5.63695095e-02 3.37111235e-01 -8.87229860e-01 3.45695883e-01
9.86809254e-01 9.83938634e-01 -6.75684139e-02 1.21378668e-01
-6.71904206e-01 9.10910070e-01 -6.35251641e-01 -5.72804451e-01
7.08017200e-02 -2.40099698e-01 -9.37615633e-01 -2.55091846e-01
2.93411702e-01 4.57728431e-02 -2.51997650e-01 9.45457220e-01
6.22097075e-01 4.47644562e-01 5.74075520e-01 -8.69434953e-01
-5.26610851e-01 1.06080329e+00 -5.58075011e-01 1.69934109e-01
3.74980457e-02 -2.69561401e-03 1.44688845e+00 -1.34541023e+00
8.80255163e-01 9.08152819e-01 9.76546586e-01 5.21322668e-01
-1.04598773e+00 -7.94276536e-01 4.57540214e-01 1.69590056e-01
-1.64285421e+00 -4.90731627e-01 5.93989968e-01 -2.74636507e-01
1.05854321e+00 -1.36993369e-02 2.46922731e-01 9.78291631e-01
-3.17455679e-01 9.76460218e-01 4.68879730e-01 -2.89966971e-01
4.94775891e-01 -1.49440439e-02 5.31547785e-01 4.58324611e-01
2.38042086e-01 -2.82379776e-01 -3.14146787e-01 -1.44139200e-01
2.57700592e-01 8.93964898e-03 -1.84983209e-01 -1.73446238e-01
-1.03790581e+00 7.40595758e-01 5.92898190e-01 6.45645440e-01
-4.66802478e-01 -1.44390628e-01 6.96497560e-01 -8.94180033e-03
5.99135101e-01 6.15418792e-01 -7.90034592e-01 -2.50131879e-02
-1.36310768e+00 -1.03571393e-01 1.03218615e+00 9.67322469e-01
9.70971525e-01 -3.12777460e-01 -6.62868798e-01 9.52574015e-01
5.01452014e-03 5.40389568e-02 4.95847970e-01 -3.30430746e-01
6.66447580e-01 8.47299635e-01 9.92546976e-02 -7.00598896e-01
-5.91711640e-01 -4.42915738e-01 -6.33026421e-01 -5.13983667e-01
3.05149615e-01 -5.01383066e-01 -1.11711681e+00 1.61804056e+00
6.62389755e-01 4.66034770e-01 -6.26352578e-02 7.31558383e-01
9.46647644e-01 5.28317869e-01 5.28181136e-01 -3.40132117e-02
1.52361047e+00 -1.12633097e+00 -5.95330954e-01 -3.90342444e-01
1.11977756e+00 -3.87841731e-01 7.82201231e-01 -1.97614506e-01
-6.00523651e-01 -1.66377753e-01 -8.51928413e-01 -1.93715721e-01
-6.30546153e-01 2.90164798e-01 3.75133574e-01 3.18999559e-01
-3.72183472e-01 6.25726402e-01 -8.22325826e-01 -3.13322425e-01
4.81206656e-01 1.20060481e-01 -3.83254439e-01 -6.93978965e-02
-1.71352875e+00 8.14519107e-01 9.45927918e-01 5.67759760e-02
-5.53438723e-01 -1.00989079e+00 -1.22234166e+00 3.81516069e-01
7.26753354e-01 -3.86260986e-01 1.09817636e+00 -5.09069800e-01
-9.34995532e-01 7.48953223e-01 -3.10783386e-01 -5.91795564e-01
2.79287249e-01 -2.62288213e-01 -5.57579279e-01 4.73681092e-02
4.01684135e-01 3.83980542e-01 2.89483696e-01 -1.10879397e+00
-7.62794733e-01 -1.11983642e-01 4.62938845e-02 -4.41443101e-02
-2.51561940e-01 4.83050570e-02 -5.13510168e-01 -6.95522010e-01
-5.07161655e-02 -6.45000577e-01 -3.60360742e-01 -5.48044860e-01
-8.07670236e-01 -5.42257726e-01 4.49615717e-01 -5.93734801e-01
1.52018189e+00 -2.27093863e+00 -3.13054621e-01 9.69151855e-02
3.29992801e-01 3.99105817e-01 -1.64769128e-01 5.04409909e-01
-7.34226853e-02 2.36714184e-01 -1.46494120e-01 -5.46377540e-01
1.50213644e-01 4.52484787e-02 -3.48390907e-01 2.63538301e-01
4.18424278e-01 1.00512326e+00 -1.19737136e+00 -7.56987989e-01
-2.23562121e-01 1.29558489e-01 -3.26013535e-01 2.27206349e-01
-1.18736036e-01 9.59743708e-02 -5.25461197e-01 6.47898734e-01
6.13191068e-01 -1.95363775e-01 2.65649021e-01 -2.88961291e-01
-1.59776315e-01 5.63439548e-01 -1.27215946e+00 1.72645056e+00
-4.61562037e-01 1.48981392e-01 -1.46099702e-01 -7.80436635e-01
9.33306634e-01 4.31386024e-01 4.54734832e-01 -4.50049430e-01
-3.44559215e-02 2.76501477e-01 -3.27993691e-01 -3.60959351e-01
5.90124011e-01 -2.72486687e-01 -5.47402442e-01 2.42271423e-01
4.11593020e-01 6.11170590e-01 3.81532580e-01 3.61363828e-01
1.39781916e+00 -8.05924553e-03 4.82532471e-01 7.61690512e-02
3.38810831e-01 9.69934389e-02 1.31033516e+00 7.38173544e-01
-3.60139817e-01 5.37284315e-01 4.86884415e-01 -3.94832015e-01
-9.05645251e-01 -9.61918831e-01 -1.98740922e-02 1.16216898e+00
2.24946037e-01 -4.91045415e-01 -7.32847750e-01 -1.33394563e+00
-3.68702933e-02 8.36820543e-01 -7.69972086e-01 -1.86102882e-01
-7.29073763e-01 -5.23404479e-01 8.66172135e-01 8.19839060e-01
3.86302143e-01 -1.15898752e+00 -2.48558670e-01 4.69229072e-01
-2.98642367e-01 -1.18901134e+00 -7.89037883e-01 3.87455016e-01
-5.14589667e-01 -9.03333485e-01 -7.11741328e-01 -8.99224043e-01
4.99429882e-01 -1.16162531e-01 1.10021913e+00 -7.96243325e-02
-1.79287732e-01 -1.57210708e-01 -5.99591374e-01 -1.51131719e-01
1.13411248e-01 7.13969052e-01 -1.12899765e-01 -7.37973768e-03
7.84197390e-01 -4.01479185e-01 -5.28536022e-01 2.03251272e-01
-5.24390757e-01 -3.59735459e-01 7.69824922e-01 1.04179537e+00
5.32997072e-01 -2.23944098e-01 9.61565495e-01 -1.28354228e+00
2.22759724e-01 -8.86204660e-01 -1.79947168e-01 5.42536974e-01
-4.58171964e-01 1.52481258e-01 8.03854644e-01 -5.42741716e-01
-1.32757139e+00 1.66218668e-01 -1.73837543e-01 -4.42430913e-01
-2.69139469e-01 6.17435992e-01 -6.09993219e-01 3.46093684e-01
6.81773126e-01 1.57765567e-01 -9.29774523e-01 -7.45553970e-01
5.89490294e-01 6.31098330e-01 5.17372191e-01 -5.01300275e-01
7.31944382e-01 2.05990359e-01 -5.91337919e-01 -5.49143970e-01
-1.47341919e+00 -1.01732397e+00 -8.14333439e-01 1.35077000e-01
8.55932832e-01 -9.81722653e-01 -2.69422501e-01 1.49846002e-01
-1.26512551e+00 -8.94643813e-02 -6.20813489e-01 5.19477487e-01
2.33698986e-03 2.89925516e-01 -9.65603769e-01 -7.73497581e-01
-5.13071060e-01 -5.15291214e-01 1.02246821e+00 4.48327512e-01
-2.75241375e-01 -9.20919836e-01 3.31878245e-01 7.95105193e-03
9.88906175e-02 2.56830398e-02 7.68024564e-01 -1.62863827e+00
-2.84311831e-01 -4.05468583e-01 -3.03919584e-01 -3.61651033e-02
-7.61274099e-02 -3.60175818e-01 -9.82610047e-01 2.00304277e-02
-2.62872279e-01 -2.64599532e-01 1.23073804e+00 -1.69979528e-01
7.73507774e-01 -3.94303352e-01 -6.52917862e-01 6.39656186e-01
1.42171860e+00 -1.44877136e-02 4.12685990e-01 4.14171070e-01
9.04748321e-01 5.01644611e-01 8.09171200e-01 4.03825253e-01
4.87240821e-01 4.00958270e-01 6.06098250e-02 -5.63957393e-02
-3.77311707e-02 -5.91804266e-01 1.75474063e-01 7.65495777e-01
2.92429060e-01 -2.12036997e-01 -1.01682186e+00 1.02734447e+00
-1.81161821e+00 -1.07591939e+00 1.29242033e-01 2.00965142e+00
1.11790538e+00 2.21542224e-01 3.77221629e-02 -2.67374337e-01
1.16455984e+00 3.57582301e-01 -6.67217135e-01 4.40535061e-02
3.44995670e-02 1.54633805e-01 5.67661941e-01 2.95232553e-02
-1.42354405e+00 1.37722278e+00 5.46615410e+00 1.12231541e+00
-8.70280504e-01 2.28268310e-01 3.69986922e-01 6.72663450e-02
-1.70073897e-01 2.56223261e-01 -1.31505060e+00 5.53207517e-01
1.00663054e+00 -2.11994797e-01 -1.89184561e-01 1.08774030e+00
-1.42081216e-01 1.20426111e-01 -1.02761459e+00 5.53585172e-01
-5.17978668e-02 -1.24884987e+00 -1.91465780e-01 -2.15819493e-01
8.51434112e-01 1.72031403e-01 -5.20642698e-01 8.22446644e-01
4.74336356e-01 -5.79522669e-01 5.83006263e-01 4.86280531e-01
6.47273123e-01 -8.84481609e-01 9.72618997e-01 4.80723083e-01
-1.62525809e+00 -7.71733224e-02 -4.65961784e-01 3.18268299e-01
4.23961103e-01 9.73222315e-01 -8.88484418e-01 6.85456872e-01
4.01846409e-01 6.15621567e-01 -2.54172623e-01 1.42779684e+00
-4.47129071e-01 8.08651805e-01 -3.28130960e-01 -4.03447729e-03
3.68217468e-01 1.80625379e-01 5.10052741e-01 1.75096619e+00
1.47967875e-01 1.42742380e-01 3.97868723e-01 9.21455503e-01
-6.45720541e-01 4.33146209e-01 -2.75685370e-01 -4.73248400e-02
1.01356339e+00 1.48607779e+00 -8.26695561e-01 -5.71000516e-01
-5.23144484e-01 1.06025183e+00 9.26653683e-01 2.77936637e-01
-8.41117144e-01 -9.85200644e-01 5.39526761e-01 -1.80499941e-01
6.76088750e-01 8.32354277e-02 -3.78925174e-01 -1.50664353e+00
5.64374663e-02 -2.94977814e-01 7.52334714e-01 -8.69298279e-02
-1.59348714e+00 5.32768905e-01 -3.43472898e-01 -1.11679208e+00
-2.20556874e-02 -1.11565910e-01 -9.44579840e-01 6.54922068e-01
-1.59571028e+00 -1.23015094e+00 -1.85711272e-02 1.99860886e-01
5.45120537e-01 2.64976650e-01 6.93239212e-01 5.79387963e-01
-9.20192957e-01 1.01659882e+00 5.03698401e-02 9.74241912e-01
1.01603854e+00 -1.39251757e+00 6.78730309e-01 9.92234409e-01
2.08564743e-01 8.64413083e-01 3.17189038e-01 -8.88195097e-01
-5.95544934e-01 -1.57202625e+00 1.41363776e+00 -3.96631896e-01
7.79884756e-01 -4.67962176e-01 -1.18041718e+00 7.99514890e-01
-2.23957896e-01 4.32591647e-01 8.95564914e-01 5.55745244e-01
-7.73183286e-01 1.42276481e-01 -1.04317605e+00 4.51570451e-01
1.05641663e+00 -5.72001338e-01 -1.10743845e+00 5.34432530e-02
9.49699879e-01 -1.86246008e-01 -9.53577280e-01 2.36571640e-01
3.34589660e-01 -4.11699593e-01 7.30805814e-01 -8.31377089e-01
1.46312952e-01 -3.02508354e-01 2.66079485e-01 -1.34966874e+00
-3.47456753e-01 -3.42445016e-01 -3.43411714e-01 1.93134260e+00
7.93087304e-01 -3.70060444e-01 7.15149343e-01 7.53946006e-01
-2.22591668e-01 -6.91927791e-01 -1.00267458e+00 -9.50062752e-01
7.01672807e-02 -1.10437535e-01 7.05568075e-01 1.30345333e+00
4.27018613e-01 6.47855520e-01 -4.15375799e-01 2.60275632e-01
4.87833887e-01 2.27424458e-01 3.41612846e-01 -1.18497336e+00
-2.02032048e-02 -1.40228570e-01 -3.60354304e-01 -1.08912420e+00
3.83957952e-01 -9.34123516e-01 5.51025093e-01 -1.50293076e+00
3.80728424e-01 -6.72229588e-01 -5.81976950e-01 8.47336948e-01
-6.39120102e-01 -1.78253315e-02 1.14038348e-01 3.40942860e-01
-1.24278665e+00 5.32136500e-01 6.56305552e-01 -7.60622472e-02
-3.28342259e-01 -3.09214413e-01 -5.96098006e-01 7.05391765e-01
5.75337827e-01 -9.67684567e-01 7.76883364e-02 -1.90236822e-01
1.88970435e-02 -2.04258174e-01 -9.22610238e-02 -9.64927316e-01
6.03850901e-01 -1.15727611e-01 3.66946489e-01 -5.23161948e-01
-1.03778400e-01 -5.80220282e-01 -2.49222085e-01 4.82075363e-02
-4.71785545e-01 -3.92522633e-01 -1.43375129e-01 6.61786497e-01
-2.75097936e-01 -4.35107976e-01 7.77328372e-01 -1.77839890e-01
-1.03949440e+00 5.06506622e-01 4.99818958e-02 6.82887137e-01
1.09609282e+00 6.46274835e-02 -3.80531579e-01 2.22842649e-01
-8.66765380e-01 4.98758942e-01 3.82197052e-01 2.14393124e-01
1.75949708e-01 -1.39756751e+00 -5.62762201e-01 -2.16268510e-01
4.41923648e-01 2.17616320e-01 5.50399244e-01 8.84824634e-01
-1.09800249e-01 1.35757074e-01 3.26849878e-01 -5.40148094e-02
-8.10544372e-01 8.28772426e-01 2.75234431e-01 -7.35086560e-01
-8.45225334e-01 9.66230452e-01 2.70938754e-01 -4.96903390e-01
2.28085741e-01 -5.69195338e-02 -4.27573502e-01 4.20541793e-01
7.39859998e-01 2.74014682e-01 5.22995852e-02 -5.90334713e-01
-5.45783639e-01 2.68778205e-01 -4.30793405e-01 1.46917000e-01
1.29406881e+00 -1.22840345e-01 1.81951329e-01 5.98893762e-01
1.15799868e+00 2.32603416e-01 -1.07341743e+00 -5.51670790e-01
7.27699161e-01 -8.77543613e-02 -2.44370893e-01 -6.91812098e-01
-1.04115069e+00 7.49007404e-01 1.51692871e-02 5.67575824e-03
7.07487643e-01 1.73259497e-01 1.34490013e+00 4.24363852e-01
2.60164648e-01 -1.27849996e+00 -2.72267610e-01 8.78447235e-01
1.21657670e-01 -1.07996273e+00 -3.74206543e-01 -4.40454245e-01
-7.26605117e-01 8.35888922e-01 7.87297130e-01 -8.10307786e-02
4.98810500e-01 3.37187260e-01 3.71484668e-03 -6.07702956e-02
-7.12159574e-01 -4.96318012e-01 2.29290679e-01 3.82749856e-01
3.54995161e-01 -1.16873354e-01 -4.50436860e-01 1.23971617e+00
3.08518380e-01 -2.02628486e-02 5.18185310e-02 7.89396524e-01
-7.10241258e-01 -9.86049294e-01 1.02835642e-02 4.54470545e-01
-7.69024909e-01 -4.88144726e-01 -3.55268985e-01 5.32642365e-01
3.38197410e-01 7.28131235e-01 -7.24336645e-03 -3.03696096e-01
5.59581101e-01 5.21886587e-01 -2.34718159e-01 -9.35189188e-01
-8.33042502e-01 -2.08503589e-01 3.48706633e-01 -3.81979167e-01
-3.11000645e-02 -5.18085361e-01 -1.70911062e+00 8.32130536e-02
-6.85458064e-01 7.65903831e-01 2.35145643e-01 1.12908697e+00
5.88947833e-01 4.43367660e-01 6.33105993e-01 -5.49549043e-01
-7.58600056e-01 -1.07139158e+00 -7.44102359e-01 5.36999643e-01
1.28474802e-01 -6.56717241e-01 -4.71303880e-01 -1.90054044e-01] | [9.566177368164062, 9.400917053222656] |
4cea5fe2-4407-4f14-821a-7ec622abee62 | dcl-net-deep-correspondence-learning-network | 2210.05232 | null | https://arxiv.org/abs/2210.05232v1 | https://arxiv.org/pdf/2210.05232v1.pdf | DCL-Net: Deep Correspondence Learning Network for 6D Pose Estimation | Establishment of point correspondence between camera and object coordinate systems is a promising way to solve 6D object poses. However, surrogate objectives of correspondence learning in 3D space are a step away from the true ones of object pose estimation, making the learning suboptimal for the end task. In this paper, we address this shortcoming by introducing a new method of Deep Correspondence Learning Network for direct 6D object pose estimation, shortened as DCL-Net. Specifically, DCL-Net employs dual newly proposed Feature Disengagement and Alignment (FDA) modules to establish, in the feature space, partial-to-partial correspondence and complete-to-complete one for partial object observation and its complete CAD model, respectively, which result in aggregated pose and match feature pairs from two coordinate systems; these two FDA modules thus bring complementary advantages. The match feature pairs are used to learn confidence scores for measuring the qualities of deep correspondence, while the pose feature pairs are weighted by confidence scores for direct object pose regression. A confidence-based pose refinement network is also proposed to further improve pose precision in an iterative manner. Extensive experiments show that DCL-Net outperforms existing methods on three benchmarking datasets, including YCB-Video, LineMOD, and Oclussion-LineMOD; ablation studies also confirm the efficacy of our novel designs. | ['Kui Jia', 'Jiehong Lin', 'Hongyang Li'] | 2022-10-11 | null | null | null | null | ['6d-pose-estimation-1', '6d-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-2.85123199e-01 -1.02172710e-01 -3.21541697e-01 -4.17649746e-01
-9.82564270e-01 -4.39196140e-01 5.02673924e-01 -1.35286823e-01
-1.87718272e-01 3.82301390e-01 -5.40481769e-02 3.05628508e-01
-4.24797982e-01 -5.00666261e-01 -9.80967879e-01 -4.93251979e-01
1.54612675e-01 7.42268562e-01 5.11017442e-02 1.43640980e-01
4.51762766e-01 9.55331624e-01 -1.29710937e+00 -2.31356218e-01
6.08155191e-01 1.27659953e+00 8.40008259e-02 1.71851784e-01
1.80286005e-01 2.97336102e-01 -3.15048248e-01 -1.79421484e-01
5.75679362e-01 -5.43574197e-03 -3.23657423e-01 3.64142150e-01
8.27582836e-01 -6.10394001e-01 -1.51447177e-01 9.66283381e-01
4.59181190e-01 -9.75631624e-02 6.53031409e-01 -1.39501381e+00
-3.43411714e-01 -9.90577880e-03 -6.95947647e-01 -3.14166993e-01
5.66554427e-01 9.70051140e-02 1.02274203e+00 -1.32093692e+00
5.35130024e-01 1.27819324e+00 8.51388574e-01 1.77895427e-01
-1.04535878e+00 -8.33450675e-01 5.69224358e-02 1.94275111e-01
-1.62588143e+00 -2.78367311e-01 1.00101173e+00 -4.38859314e-01
7.96141982e-01 1.44863902e-02 7.91381061e-01 8.51380229e-01
1.73180446e-01 7.39945173e-01 8.15356374e-01 -2.19383821e-01
-2.23028157e-02 -1.00821126e-02 -1.36923760e-01 6.86096907e-01
3.29331070e-01 2.81729430e-01 -6.31681442e-01 8.97326972e-03
1.26476693e+00 1.26770034e-01 -3.71537298e-01 -1.20341718e+00
-1.33207107e+00 7.45196998e-01 7.35018730e-01 -1.25432745e-01
-1.74718320e-01 4.68409173e-02 -3.72724831e-02 3.00287232e-02
3.84859383e-01 8.17716122e-01 -5.42963088e-01 2.28710324e-02
-5.06755829e-01 4.28976566e-01 5.70903957e-01 1.41866100e+00
8.17317009e-01 -2.01877832e-01 -1.33076176e-01 4.56365764e-01
5.86135626e-01 6.01090610e-01 1.00073218e-01 -1.04752445e+00
5.77291429e-01 8.38534474e-01 2.58311808e-01 -1.37010682e+00
-6.24628067e-01 -7.70019531e-01 -8.42942595e-01 1.17211156e-01
3.54839206e-01 1.67706668e-01 -6.13660991e-01 1.54674888e+00
5.49559534e-01 3.54321361e-01 -2.49384671e-01 1.14618790e+00
7.25085080e-01 4.14259642e-01 -5.17428398e-01 -1.15668431e-01
1.00818431e+00 -8.15890789e-01 -4.10844773e-01 -5.92239201e-02
5.60940742e-01 -8.39653611e-01 7.14916170e-01 2.98577547e-01
-1.10885739e+00 -7.84444988e-01 -1.25195372e+00 -8.72042999e-02
1.33282151e-02 4.66261029e-01 4.48669732e-01 1.14139862e-01
-5.59740484e-01 4.90913093e-01 -7.80888081e-01 1.99472010e-02
5.01327991e-01 5.96469700e-01 -5.86363375e-01 8.02456215e-02
-5.94079316e-01 1.05124104e+00 3.16446334e-01 4.47350413e-01
-5.73044300e-01 -8.98033142e-01 -9.78307068e-01 -2.38024406e-02
6.46847248e-01 -7.16373265e-01 1.00358891e+00 -1.85718581e-01
-1.41940558e+00 8.54958713e-01 1.38190135e-01 -5.02791703e-02
7.96255171e-01 -5.76883972e-01 1.16379028e-02 -8.77223760e-02
1.79722115e-01 6.59279227e-01 8.75891566e-01 -1.38471937e+00
-6.55601561e-01 -6.91661119e-01 9.10820365e-02 4.07964200e-01
-1.14397943e-01 -4.12432820e-01 -8.05684865e-01 -3.67669255e-01
7.79112756e-01 -9.07486677e-01 -1.47289947e-01 6.22783899e-01
-4.58072484e-01 -4.05748069e-01 8.44977438e-01 -3.10842007e-01
6.78705633e-01 -2.22880673e+00 3.62464905e-01 3.45362902e-01
4.49269593e-01 7.36228824e-02 -2.34757289e-01 -6.92071244e-02
-1.09146968e-01 -2.95119703e-01 1.60348043e-01 -3.95419419e-01
-4.25511040e-02 -7.07402006e-02 7.57683665e-02 8.79219770e-01
5.83557248e-01 1.04546928e+00 -7.88068950e-01 -3.96887183e-01
4.91318554e-01 4.84820813e-01 -5.77295303e-01 3.71088207e-01
-8.67448449e-02 4.23075169e-01 -5.92332542e-01 5.92300296e-01
8.84245217e-01 -2.19182596e-01 -2.23571837e-01 -8.30278039e-01
-2.71696318e-02 -3.10188960e-02 -1.44058907e+00 1.82241476e+00
-3.80514145e-01 3.36973459e-01 -2.26154312e-01 -7.74386168e-01
1.30575871e+00 2.96528079e-02 6.98277414e-01 -4.80566084e-01
2.95650363e-01 2.03595117e-01 -1.86564505e-01 -2.12253779e-01
2.84405231e-01 1.83162048e-01 -2.48643905e-02 1.79883480e-01
1.03608064e-01 -6.46821141e-01 -1.64798483e-01 -2.96636760e-01
4.51482832e-01 4.25231695e-01 4.71544325e-01 2.94133350e-02
6.06978118e-01 -2.02243030e-01 6.12165153e-01 4.54558909e-01
-1.78783283e-01 9.62843299e-01 4.22175944e-01 -4.64504033e-01
-1.12470031e+00 -1.13595712e+00 -2.92516381e-01 1.38159111e-01
5.93043506e-01 -3.75568330e-01 -3.46513867e-01 -8.99003506e-01
3.01529199e-01 3.34838301e-01 -3.29432696e-01 -3.87920558e-01
-8.59199286e-01 -3.23272377e-01 -3.42388302e-02 8.33625793e-01
4.71477807e-01 -6.68589711e-01 -3.89624119e-01 2.34275553e-02
-7.59304129e-03 -1.24815965e+00 -5.99017262e-01 1.11453220e-01
-9.29519117e-01 -1.30728590e+00 -5.74926555e-01 -7.32258558e-01
7.70097911e-01 2.84525782e-01 8.09932590e-01 -9.51170027e-02
-1.22410946e-01 1.58319026e-01 -1.29435882e-01 -1.59432188e-01
-3.80453877e-02 7.28443712e-02 2.60285050e-01 -9.00001675e-02
3.27747077e-01 -4.71476018e-01 -6.10926569e-01 6.97528005e-01
-3.74545217e-01 9.25485417e-02 7.96657145e-01 8.47464919e-01
7.50690281e-01 -2.22334251e-01 2.71081179e-01 -2.55758405e-01
1.47970572e-01 -1.00625567e-01 -1.00200057e+00 1.10965848e-01
-4.91167575e-01 8.26439727e-03 2.64973164e-01 -4.29135829e-01
-5.68549871e-01 5.79860806e-01 -1.47501960e-01 -8.64151716e-01
-4.48860675e-02 3.86959076e-01 -3.61068517e-01 -2.95966804e-01
3.92542809e-01 -1.45739864e-03 1.93588823e-01 -4.38216835e-01
1.65652469e-01 2.46021673e-01 6.04615629e-01 -4.41674918e-01
1.28096747e+00 2.72395581e-01 3.03523153e-01 -4.04973775e-01
-1.00787127e+00 -5.40432870e-01 -1.12575018e+00 -2.67373770e-01
7.39698470e-01 -1.00209141e+00 -9.49168861e-01 4.25307304e-01
-1.26987457e+00 3.15619946e-01 -3.07506710e-01 7.99759030e-01
-7.16453969e-01 2.99829602e-01 -2.39655003e-01 -3.53348643e-01
-7.88520053e-02 -1.29727018e+00 1.50196815e+00 1.05440229e-01
-1.23551846e-01 -5.34595966e-01 -1.82732061e-01 2.78000146e-01
-1.57447204e-01 3.49601567e-01 6.61772132e-01 -4.25065428e-01
-9.42017198e-01 -6.25066042e-01 -4.60008562e-01 3.34710926e-01
1.95357233e-01 -8.32285136e-02 -7.38379776e-01 -5.08748114e-01
1.91871405e-01 -3.18030387e-01 3.00865591e-01 4.64002639e-01
1.06411004e+00 1.17267147e-01 -3.43975812e-01 8.77683103e-01
1.36636305e+00 -4.40920740e-02 3.01011205e-01 3.92299592e-01
9.68902886e-01 3.25151384e-01 1.10020125e+00 5.85133076e-01
2.57004052e-01 1.14579880e+00 7.92539299e-01 -5.13036400e-02
-8.15445557e-02 -3.34032357e-01 7.84874037e-02 8.77639174e-01
-7.25647528e-03 1.84233025e-01 -6.37523055e-01 3.24955508e-02
-1.83636403e+00 -2.70992815e-01 -2.22120881e-01 2.21609569e+00
5.16671598e-01 3.37280035e-01 1.55074773e-02 -5.21657392e-02
6.83802485e-01 -4.39506881e-02 -7.15078294e-01 4.90746111e-01
5.19677922e-02 -7.74250329e-02 2.08967239e-01 3.27946931e-01
-1.09445572e+00 8.21915984e-01 5.16879272e+00 7.01273739e-01
-9.76778805e-01 -1.76272765e-01 2.87742317e-01 8.16307217e-02
1.00372329e-01 -5.68532869e-02 -1.13145971e+00 9.32733938e-02
-1.58251170e-02 6.43606484e-02 -6.94269910e-02 1.05352926e+00
1.07911844e-02 1.17358498e-01 -1.64312983e+00 1.32835591e+00
1.62126660e-01 -1.37586164e+00 7.35647604e-02 7.15101212e-02
6.58198118e-01 -2.22028911e-01 8.92622955e-03 1.36709124e-01
-2.40030155e-01 -6.61685884e-01 9.19695437e-01 5.26565671e-01
8.15481722e-01 -8.57090473e-01 8.90010118e-01 4.34082061e-01
-1.21088564e+00 4.24618423e-02 -4.37168628e-01 6.53868020e-02
1.97471008e-01 3.97552729e-01 -8.55590403e-01 8.52468312e-01
7.21761286e-01 1.05328405e+00 -7.17043340e-01 1.20116234e+00
-8.51071551e-02 -1.91990599e-01 -4.22244608e-01 -8.60196818e-03
7.41977543e-02 -1.10305883e-01 7.25486875e-01 5.68456233e-01
5.06690443e-01 -1.72759183e-02 2.00466290e-01 1.10093343e+00
7.02847987e-02 -1.93582162e-01 -3.89269561e-01 3.24028403e-01
5.40163517e-01 1.21755207e+00 -6.82548106e-01 1.23114930e-02
-1.73073590e-01 6.59249425e-01 2.74212092e-01 2.78795194e-02
-9.15158927e-01 -8.49017501e-02 6.43123865e-01 8.91114846e-02
2.73779958e-01 -4.23087656e-01 -2.46107966e-01 -1.25324178e+00
3.45947713e-01 -7.68111229e-01 -2.41431650e-02 -8.84872973e-01
-1.33871067e+00 3.68814349e-01 1.48096457e-01 -1.71750522e+00
-1.82135791e-01 -8.04902554e-01 -2.55245179e-01 6.20515943e-01
-1.31233239e+00 -1.19230950e+00 -5.82566440e-01 5.05729973e-01
6.43383682e-01 -1.37108654e-01 5.35227776e-01 1.58415824e-01
-3.27296317e-01 7.24689543e-01 -1.69803143e-01 1.09860063e-01
7.54606426e-01 -1.00531542e+00 2.91901439e-01 4.14957255e-01
2.07930520e-01 4.68528837e-01 3.09202880e-01 -5.26931822e-01
-1.60208201e+00 -1.04509842e+00 5.12103498e-01 -7.52088368e-01
2.17113182e-01 -5.98010659e-01 -6.72405839e-01 5.54376543e-01
-4.60227758e-01 2.47960404e-01 -5.69476048e-03 -4.77329195e-02
-2.38936871e-01 -3.61615360e-01 -9.64973450e-01 4.99670655e-01
1.10329878e+00 -3.37479591e-01 -5.17210245e-01 3.72084588e-01
8.18772256e-01 -9.01763201e-01 -9.49534833e-01 8.47955763e-01
7.90172100e-01 -8.63986850e-01 1.23238528e+00 -2.50963390e-01
3.97564739e-01 -5.84554434e-01 -8.18322599e-02 -1.07391179e+00
-4.23524529e-01 -3.70230705e-01 -2.28461325e-01 1.16648793e+00
1.00807607e-01 -3.85162026e-01 1.00582194e+00 1.93797231e-01
-3.02815467e-01 -1.03273201e+00 -1.00826681e+00 -8.00658047e-01
-1.16651893e-01 -4.56796736e-01 8.21751833e-01 8.25229526e-01
-5.81016660e-01 4.53684747e-01 -2.88010120e-01 4.92290735e-01
7.36236513e-01 2.61024982e-01 1.10709608e+00 -1.37505853e+00
-1.23092018e-01 -5.17687082e-01 -7.76383936e-01 -1.52272010e+00
1.78895026e-01 -6.72805250e-01 1.18677624e-01 -1.10256493e+00
1.67167649e-01 -5.63707888e-01 -7.26196729e-03 1.99877709e-01
3.75284217e-02 3.37782890e-01 2.05903485e-01 2.77234286e-01
-4.65246022e-01 8.38299274e-01 1.57232809e+00 -3.14953295e-03
-1.83853179e-01 3.76202196e-01 -4.04254168e-01 7.85991728e-01
3.52845371e-01 -4.06122625e-01 -1.68717638e-01 -3.88502419e-01
1.59301266e-01 1.47199109e-01 5.67195535e-01 -1.11720300e+00
3.01931947e-01 1.99509673e-02 7.82673836e-01 -1.20911539e+00
5.86570680e-01 -1.21786594e+00 5.16204834e-02 3.92106086e-01
-2.04358011e-01 -9.85154603e-03 -6.89297318e-02 5.04993975e-01
-1.49327502e-01 -1.49682835e-01 6.79306865e-01 1.54731914e-01
-5.62988400e-01 6.26121402e-01 3.27666521e-01 -2.88194805e-01
1.09655869e+00 -6.17647886e-01 1.91149861e-01 -2.89340824e-01
-6.93566263e-01 3.70780140e-01 3.98547083e-01 6.63873494e-01
8.76003981e-01 -1.67780209e+00 -6.53417170e-01 5.17106593e-01
3.93227428e-01 7.18983948e-01 -1.32811368e-01 8.70948434e-01
-4.63339448e-01 4.96224940e-01 -9.17482078e-02 -1.31406760e+00
-1.10882425e+00 4.05678034e-01 2.87397742e-01 -9.79316011e-02
-6.01560473e-01 8.69834244e-01 3.27483833e-01 -8.29611778e-01
4.52593207e-01 -6.12651527e-01 -1.00853883e-01 -1.09370321e-01
4.36970405e-02 3.49927038e-01 1.50940806e-01 -6.80312276e-01
-3.51461828e-01 1.20922720e+00 -1.29245266e-01 2.95629948e-01
1.52438998e+00 -1.57590851e-01 -2.89041530e-02 1.98984668e-01
1.44920814e+00 -2.45352075e-01 -1.63252985e+00 -3.46833080e-01
-2.13662419e-03 -7.38586187e-01 -4.81827557e-02 -4.06802416e-01
-1.10682905e+00 7.21606851e-01 7.61281192e-01 -3.50731820e-01
8.80405605e-01 1.96916327e-01 5.04783750e-01 5.69488645e-01
3.48784000e-01 -7.79965281e-01 5.58381081e-01 5.19440055e-01
1.21545994e+00 -1.45468903e+00 2.95868903e-01 -7.02991128e-01
-1.12765826e-01 1.36696649e+00 1.09782052e+00 -4.02223408e-01
6.08097434e-01 1.93654597e-01 -2.24518806e-01 -2.97961503e-01
-2.03204766e-01 1.81205347e-01 7.44159818e-01 5.34181118e-01
1.02694646e-01 -2.79389411e-01 1.22766234e-01 4.84273434e-01
-2.69737720e-01 -2.74582148e-01 2.08394062e-02 7.97641397e-01
-1.31258160e-01 -7.45501101e-01 -4.53803122e-01 2.29148626e-01
1.42357647e-01 3.73424530e-01 -3.65154296e-01 1.17924130e+00
9.47423130e-02 4.41307008e-01 3.12535495e-01 -4.21758890e-01
5.91530561e-01 -4.58460957e-01 8.49101603e-01 -6.18902743e-01
-2.20669568e-01 2.96843737e-01 -2.19324797e-01 -8.11322093e-01
-4.30410594e-01 -5.90475023e-01 -1.01959741e+00 4.36118394e-02
-1.01184356e+00 -1.71438083e-01 5.41276157e-01 1.04817688e+00
1.85452580e-01 2.78054327e-01 8.71373594e-01 -1.35448575e+00
-8.48252714e-01 -7.86095381e-01 -3.17108721e-01 1.90196529e-01
4.32113826e-01 -1.03561616e+00 -3.22769970e-01 -2.46503308e-01] | [7.500960350036621, -2.651764392852783] |
1ef2348f-15c3-4c74-8eed-39dbd1821cf6 | a-fast-successive-qp-algorithm-for-general | 2212.06983 | null | https://arxiv.org/abs/2212.06983v1 | https://arxiv.org/pdf/2212.06983v1.pdf | A Fast Successive QP Algorithm for General Mean-Variance Portfolio Optimization | The mean and variance of portfolio returns are the standard quantities to measure the expected return and risk of a portfolio. Efficient portfolios that provide optimal trade-offs between mean and variance warrant consideration. To express a preference among these efficient portfolios, investors have put forward many mean-variance portfolio (MVP) formulations which date back to the classical Markowitz portfolio. However, most existing algorithms are highly specialized to particular formulations and cannot be generalized for broader applications. Therefore, a fast and unified algorithm would be extremely beneficial. In this paper, we first introduce a general MVP problem formulation that can fit most existing cases by exploring their commonalities. Then, we propose a widely applicable and provably convergent successive quadratic programming algorithm (SCQP) for the general formulation. The proposed algorithm can be implemented based on only the QP solvers and thus is computationally efficient. In addition, a fast implementation is considered to accelerate the algorithm. The numerical results show that our proposed algorithm significantly outperforms the state-of-the-art ones in terms of convergence speed and scalability. | ['Daniel P. Palomar', 'Xiwen Wang', 'Shengjie Xiu'] | 2022-12-14 | null | null | null | null | ['portfolio-optimization'] | ['time-series'] | [-1.65421307e-01 -3.65384072e-01 -2.32214943e-01 -2.05372080e-01
-6.77932799e-01 -7.24537730e-01 1.50327399e-01 -1.24442548e-01
-1.76715910e-01 8.87012303e-01 -3.80192280e-01 -4.85694379e-01
-8.64303112e-01 -9.28144217e-01 -3.29184920e-01 -8.03430974e-01
5.97652718e-02 3.42822999e-01 8.56200010e-02 -1.37064219e-01
6.12260044e-01 4.64909911e-01 -1.17321551e+00 -2.31933862e-01
1.09137225e+00 1.40658700e+00 1.52390793e-01 -3.95544246e-02
-2.63677806e-01 4.75948542e-01 -2.47728467e-01 -7.15795457e-01
7.39094496e-01 -4.09278154e-01 -2.77603507e-01 1.36331350e-01
-2.07213089e-01 -2.71601915e-01 8.16615149e-02 1.23006701e+00
4.64373261e-01 2.25971147e-01 5.00282884e-01 -1.17602241e+00
-6.01657033e-01 6.46451950e-01 -9.13727224e-01 3.30276430e-01
-3.55832954e-03 -1.82142407e-01 1.35821819e+00 -8.42276394e-01
5.05492613e-02 8.89506996e-01 4.48099196e-01 9.37023610e-02
-9.25544500e-01 -5.42938352e-01 3.24853987e-01 7.79835880e-02
-1.21277821e+00 -2.73372866e-02 8.92392874e-01 -3.87448996e-01
5.25470734e-01 6.43698335e-01 7.22946763e-01 3.01141322e-01
3.49528849e-01 6.97757840e-01 1.23369730e+00 -1.07994623e-01
3.23528200e-01 3.96313190e-01 3.67515115e-03 1.61790296e-01
7.11439967e-01 1.28229976e-01 -2.12883681e-01 -3.68112475e-01
6.13752186e-01 1.00575231e-01 -5.42555153e-01 -4.49234873e-01
-7.91054249e-01 9.69888628e-01 7.73510262e-02 1.81905776e-02
-5.20028234e-01 -7.22955018e-02 4.49035913e-02 2.56230116e-01
4.04670596e-01 4.26080167e-01 -1.40855804e-01 -1.68595046e-01
-1.02802253e+00 7.50724316e-01 1.01894581e+00 8.55378330e-01
3.57907623e-01 5.11666872e-02 -2.35699192e-01 6.40236080e-01
4.88653898e-01 5.70233703e-01 1.14564694e-01 -9.82961059e-01
7.58610129e-01 4.61307436e-01 4.76235300e-01 -1.32262909e+00
-1.27494529e-01 -7.67390966e-01 -5.24014711e-01 2.85559952e-01
2.58845061e-01 -3.41823786e-01 3.50186750e-02 1.25859904e+00
2.40620956e-01 1.44987836e-01 7.16561675e-02 8.73983800e-01
5.91103509e-02 8.58153880e-01 -6.04224622e-01 -7.52606332e-01
1.09715462e+00 -9.86713409e-01 -6.86773241e-01 -1.32896245e-01
-1.93604082e-02 -9.09279406e-01 5.54825842e-01 4.49594826e-01
-1.30725324e+00 -1.03419079e-02 -1.14347267e+00 5.40490806e-01
1.59794286e-01 -6.48407042e-02 7.37352133e-01 9.22588646e-01
-7.99747288e-01 8.45567882e-01 -5.10766149e-01 3.15577865e-01
2.68548518e-01 3.49140286e-01 3.22780669e-01 3.30585629e-01
-9.67954457e-01 8.08092177e-01 5.65147877e-01 4.97677088e-01
-3.73978734e-01 -8.14989626e-01 -2.56108969e-01 3.40612620e-01
6.41141295e-01 -6.43998682e-01 1.35439813e+00 -7.89438963e-01
-1.82640398e+00 3.45961690e-01 5.73583320e-02 -3.97528201e-01
8.54450226e-01 -1.68121323e-01 1.60396062e-02 7.96527565e-02
-1.90267444e-01 -3.29547614e-01 6.38628900e-01 -8.29573631e-01
-8.78844798e-01 -4.16639186e-02 2.96608150e-01 2.45237932e-01
-6.19951367e-01 3.83534074e-01 -4.18939590e-02 -8.06863725e-01
3.47908810e-02 -7.36665964e-01 -3.64000142e-01 -5.26249781e-02
-1.55013785e-01 1.50265485e-01 2.25876436e-01 -4.64347959e-01
1.55190909e+00 -1.84917474e+00 6.65664673e-02 3.10564548e-01
-2.15495363e-01 9.69898328e-02 3.74703944e-01 7.34302402e-01
-7.94565454e-02 -6.15233108e-02 -3.90078902e-01 4.99064103e-02
3.53390217e-01 -9.68592837e-02 -6.82916582e-01 6.58191323e-01
-1.13075562e-01 7.15175629e-01 -6.97195768e-01 -2.80899644e-01
-2.45625526e-02 2.03077178e-02 -5.10099053e-01 3.00290823e-01
-6.79435674e-03 1.00729845e-01 -8.52227390e-01 5.32165945e-01
1.00861609e+00 -1.40368432e-01 1.38597414e-01 6.35531396e-02
-5.52407026e-01 -2.67645437e-02 -1.58705568e+00 8.40252340e-01
-2.29334623e-01 1.53394505e-01 1.37004450e-01 -1.23732293e+00
1.13228965e+00 1.88396677e-01 5.80458343e-01 -2.97980040e-01
1.82922944e-01 6.81748688e-01 -7.88662285e-02 -2.51155794e-01
5.35898864e-01 -5.76585889e-01 3.16349827e-02 5.58862746e-01
-6.94126487e-01 1.53703913e-01 3.77135843e-01 -3.38207096e-01
5.72109699e-01 -4.55008149e-02 5.08675039e-01 -5.40641606e-01
8.17762136e-01 -2.00988874e-01 8.58812511e-01 6.15797460e-01
-4.88517992e-02 3.34618866e-01 8.38305712e-01 -9.79285315e-02
-8.15267324e-01 -1.09937918e+00 -2.74604350e-01 5.14127195e-01
2.02193052e-01 9.77963060e-02 -4.37739730e-01 -2.11659148e-01
2.19051242e-01 7.38565445e-01 -1.26803994e-01 3.19292754e-01
-4.24347967e-01 -1.14354277e+00 -8.33692327e-02 3.52330714e-01
4.93264943e-01 -7.68330336e-01 -6.36742711e-01 4.91111100e-01
2.81749189e-01 -7.22194135e-01 -3.75588804e-01 -1.90648109e-01
-9.98100460e-01 -9.63768721e-01 -1.21645331e+00 -4.25453901e-01
4.51915443e-01 4.12087947e-01 7.98249900e-01 -2.64916956e-01
1.60857230e-01 1.61138818e-01 -3.25983107e-01 -6.05792642e-01
1.19364093e-04 -2.32688934e-01 1.51585834e-02 4.43968713e-01
-4.10981774e-02 -5.84776402e-01 -6.19647443e-01 2.68350899e-01
-7.45975852e-01 -1.50804117e-01 7.53948271e-01 7.75661588e-01
7.01612651e-01 2.85600334e-01 8.37356329e-01 -8.42824936e-01
9.05531108e-01 -5.15095651e-01 -1.32153523e+00 6.34674788e-01
-9.23146605e-01 4.14146855e-02 7.04399884e-01 -1.11525513e-01
-1.33300054e+00 -3.75360221e-01 1.65676147e-01 -3.01434904e-01
6.99963093e-01 6.82436764e-01 -2.25245327e-01 -3.49369347e-01
-2.12004244e-01 4.21102375e-01 -2.56047428e-01 -4.88588840e-01
9.62185636e-02 6.21775329e-01 2.23772883e-01 -7.28954732e-01
8.54142249e-01 3.47907484e-01 3.56932193e-01 -3.13934118e-01
-9.60870326e-01 -3.18293482e-01 -5.93202114e-02 -5.13801090e-02
3.62649143e-01 -3.88994366e-01 -8.07363868e-01 4.81917173e-01
-8.85482430e-01 3.20954740e-01 -4.42787744e-02 5.45373678e-01
-8.38722765e-01 5.89574218e-01 -5.01622975e-01 -1.28744316e+00
-5.45255244e-01 -1.17885482e+00 4.15537238e-01 4.58338380e-01
2.25368589e-01 -1.18814075e+00 1.80249855e-01 1.57301784e-01
5.06883323e-01 4.48621273e-01 6.41152322e-01 -6.58042431e-01
-7.92032897e-01 -1.95578292e-01 -3.27519268e-01 3.60805959e-01
3.79675478e-02 -7.07552508e-02 -5.14881730e-01 -4.57455456e-01
6.00076258e-01 1.96503848e-01 4.75455046e-01 4.50265229e-01
1.00792313e+00 -6.22298717e-01 -1.51861802e-01 8.69835138e-01
1.82131541e+00 5.70057034e-01 2.85225749e-01 8.42207611e-01
1.12070635e-01 6.66905999e-01 1.05568206e+00 9.48014438e-01
1.46934297e-02 4.72621292e-01 3.97529572e-01 3.90058100e-01
1.02365780e+00 1.02955841e-01 3.86545181e-01 9.41412449e-01
-1.36876583e-01 -2.11896360e-01 -6.41288221e-01 4.20599759e-01
-2.01418614e+00 -1.08220613e+00 -1.17126100e-01 2.37288809e+00
6.04677796e-01 2.15240374e-01 -2.02240590e-02 5.81330471e-02
7.91569054e-01 2.87786365e-01 -4.78967607e-01 -5.75927258e-01
-4.92767058e-02 1.40518531e-01 7.98096836e-01 1.64851159e-01
-1.12437093e+00 2.66659349e-01 6.30796909e+00 9.81898427e-01
-9.49366987e-01 -1.53551430e-01 5.89358211e-01 -3.22982579e-01
-7.02822268e-01 2.31161579e-01 -1.02532613e+00 6.77020133e-01
8.24180841e-01 -1.06451714e+00 3.29710960e-01 1.13454366e+00
3.71751189e-01 1.98833242e-01 -8.27726662e-01 1.02126718e+00
-2.80449182e-01 -1.31040335e+00 -2.25686237e-01 2.06596196e-01
9.18898284e-01 -6.82475865e-01 3.16697419e-01 1.87249348e-01
-2.71721601e-01 -7.58665860e-01 7.26938605e-01 7.91561306e-01
1.49444297e-01 -1.14933527e+00 1.04620016e+00 4.42921430e-01
-1.08022141e+00 -3.18399280e-01 -8.57227623e-01 -1.83289155e-01
5.30466080e-01 6.86197221e-01 -2.28077263e-01 9.46349144e-01
5.61324120e-01 5.59160650e-01 -5.32868318e-02 1.53111899e+00
-1.23422757e-01 3.28809589e-01 -3.83603871e-01 -3.71362418e-01
4.06375915e-01 -8.76482487e-01 6.70278370e-01 7.73511648e-01
8.31378281e-01 1.33709133e-01 1.67523734e-02 9.20640469e-01
2.38267720e-01 5.69231153e-01 -2.66985625e-01 -1.36038378e-01
4.69121009e-01 1.09041798e+00 -5.90381205e-01 -7.26155788e-02
-7.62165964e-01 4.83606249e-01 1.79046661e-01 2.26746589e-01
-9.68401134e-01 -5.20266593e-01 6.36797190e-01 -2.89968271e-02
5.38022101e-01 1.00332431e-01 -6.22711599e-01 -1.06354535e+00
4.86440748e-01 -7.39998996e-01 3.97452265e-01 -1.85106814e-01
-1.49604547e+00 3.58172089e-01 2.39843577e-01 -1.58108878e+00
-1.35356173e-01 -7.24500775e-01 -7.84448266e-01 1.09558642e+00
-1.84651971e+00 -4.37264115e-01 1.37579693e-02 1.31438360e-01
2.04715714e-01 -3.88072014e-01 3.80950928e-01 3.23620200e-01
-8.58994544e-01 5.94298244e-01 9.39982653e-01 -3.44290316e-01
2.57539988e-01 -1.25074828e+00 1.70407258e-02 1.10796261e+00
-2.90734410e-01 7.23581314e-01 8.68879437e-01 -4.64277774e-01
-1.39953113e+00 -8.12824130e-01 6.72565937e-01 1.68319345e-01
1.10764694e+00 2.42459193e-01 -8.34949672e-01 5.07401764e-01
1.44394651e-01 -4.09062237e-01 7.20299602e-01 -3.09735447e-01
2.96167135e-01 -5.86961627e-01 -9.40342188e-01 6.18869126e-01
5.02175689e-01 5.46634831e-02 -7.32848346e-01 2.20595196e-01
3.30235392e-01 -2.77245492e-01 -1.00940418e+00 3.58891666e-01
6.17897749e-01 -1.27345240e+00 9.03265059e-01 -1.36528537e-01
1.98773965e-01 -1.45830855e-01 -9.24182013e-02 -9.35437381e-01
-2.89407521e-01 -1.08758557e+00 -1.79244205e-01 1.41851079e+00
3.01566333e-01 -1.24712586e+00 6.85790122e-01 7.93173790e-01
9.04690921e-02 -1.32219112e+00 -1.05340421e+00 -1.10841107e+00
2.37072870e-01 -2.91500449e-01 8.60762179e-01 4.94553506e-01
1.23767126e-02 -3.11470360e-01 -5.60829461e-01 1.13747098e-01
1.02143228e+00 9.36758637e-01 3.77592862e-01 -1.01857376e+00
-7.73180783e-01 -8.26465607e-01 -1.44555762e-01 -1.16173208e+00
-6.37769550e-02 -7.89903998e-01 -2.03351259e-01 -1.33351350e+00
3.65922838e-01 -4.79253620e-01 -5.87444723e-01 -6.98326677e-02
-3.84723246e-01 -2.89583802e-01 3.09166580e-01 3.14603895e-01
-3.03185374e-01 7.10159361e-01 1.17819154e+00 7.60468394e-02
-1.69741154e-01 6.24200344e-01 -1.00189674e+00 7.58651018e-01
9.38541532e-01 -5.89021921e-01 -5.42553306e-01 -1.53211802e-01
4.87021685e-01 3.99554282e-01 -3.25806066e-02 -6.63119078e-01
1.85172692e-01 -7.66252398e-01 -1.04909465e-01 -8.18984747e-01
1.02907740e-01 -7.19922245e-01 4.21323538e-01 4.19839323e-01
-9.10835788e-02 1.38748959e-01 -1.49043053e-01 6.26727641e-01
-4.33892697e-01 -1.04420972e+00 7.60966361e-01 -2.32619300e-01
-3.63945514e-01 5.62601089e-01 -1.12824664e-01 1.19358376e-01
1.42648244e+00 -1.82762176e-01 -1.14675730e-01 -3.58407259e-01
-1.55059665e-01 4.62874502e-01 3.97981197e-01 1.26813263e-01
7.37904191e-01 -1.26822865e+00 -7.44701207e-01 -2.88350731e-01
-3.96163881e-01 -2.97792494e-01 1.23553082e-01 9.14872289e-01
-8.98362696e-01 8.88158262e-01 -5.81401400e-02 -1.71771720e-02
-7.75649130e-01 7.03099251e-01 4.59718376e-01 -5.74679017e-01
-4.30094212e-01 7.40393758e-01 4.11419123e-01 1.36669040e-01
-3.40273883e-03 -2.09061503e-01 -2.69794583e-01 3.96646634e-02
7.30646312e-01 7.34229565e-01 -3.81391197e-01 -2.55795598e-01
-3.45517486e-01 8.23114038e-01 2.55948603e-01 -1.42528355e-01
1.42394722e+00 -2.28984490e-01 -4.58860964e-01 4.06344622e-01
1.06444478e+00 1.84698999e-01 -1.22221267e+00 3.13716494e-02
3.26597959e-01 -7.71028340e-01 -1.20727651e-01 -9.02599841e-02
-1.19803047e+00 7.46117890e-01 1.85697839e-01 2.79140830e-01
1.41038966e+00 -5.98539531e-01 8.70088339e-01 3.45803618e-01
6.87816441e-01 -1.02128685e+00 -3.29077721e-01 1.97753474e-01
9.92208958e-01 -7.79719472e-01 4.98377830e-01 -5.94855130e-01
-6.46245718e-01 1.35537326e+00 2.65403599e-01 -4.61080045e-01
7.48847902e-01 1.24444701e-01 -1.63858682e-01 2.10202083e-01
-6.05828941e-01 1.40039846e-01 3.38328362e-01 1.51709005e-01
2.63720572e-01 -3.12851323e-03 -1.14319611e+00 1.15694880e+00
-2.17435136e-01 -4.31535952e-02 4.99360025e-01 8.29327047e-01
-5.68507493e-01 -1.28249454e+00 -5.34266174e-01 4.19422269e-01
-9.90159333e-01 -2.72234692e-03 5.27772754e-02 6.65202975e-01
-2.27081403e-01 9.38735902e-01 -2.24980980e-01 2.25697413e-01
4.42385465e-01 -3.18371296e-01 3.23361546e-01 -3.00451547e-01
-4.95243013e-01 1.50702491e-01 -7.49647617e-02 -3.07614684e-01
-3.82099807e-01 -8.43712330e-01 -8.53435040e-01 -3.93679291e-01
-4.13218319e-01 4.58579481e-01 2.94796139e-01 7.97479987e-01
-1.56104475e-01 3.19092870e-01 1.01691806e+00 -4.85400021e-01
-1.51028788e+00 -4.76537913e-01 -1.12221420e+00 -2.36049131e-01
-2.30543748e-01 -8.08258474e-01 -4.68467534e-01 -4.69753563e-01] | [5.036386489868164, 3.95029354095459] |
bac47a84-096f-4f34-be69-5768f8fab8e5 | bsp-net-generating-compact-meshes-via-binary | 1911.06971 | null | https://arxiv.org/abs/1911.06971v6 | https://arxiv.org/pdf/1911.06971v6.pdf | BSP-Net: Generating Compact Meshes via Binary Space Partitioning | Polygonal meshes are ubiquitous in the digital 3D domain, yet they have only played a minor role in the deep learning revolution. Leading methods for learning generative models of shapes rely on implicit functions, and generate meshes only after expensive iso-surfacing routines. To overcome these challenges, we are inspired by a classical spatial data structure from computer graphics, Binary Space Partitioning (BSP), to facilitate 3D learning. The core ingredient of BSP is an operation for recursive subdivision of space to obtain convex sets. By exploiting this property, we devise BSP-Net, a network that learns to represent a 3D shape via convex decomposition. Importantly, BSP-Net is unsupervised since no convex shape decompositions are needed for training. The network is trained to reconstruct a shape using a set of convexes obtained from a BSP-tree built on a set of planes. The convexes inferred by BSP-Net can be easily extracted to form a polygon mesh, without any need for iso-surfacing. The generated meshes are compact (i.e., low-poly) and well suited to represent sharp geometry; they are guaranteed to be watertight and can be easily parameterized. We also show that the reconstruction quality by BSP-Net is competitive with state-of-the-art methods while using much fewer primitives. Code is available at https://github.com/czq142857/BSP-NET-original. | ['Hao Zhang', 'Andrea Tagliasacchi', 'Zhiqin Chen'] | 2019-11-16 | bsp-net-generating-compact-meshes-via-binary-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Chen_BSP-Net_Generating_Compact_Meshes_via_Binary_Space_Partitioning_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Chen_BSP-Net_Generating_Compact_Meshes_via_Binary_Space_Partitioning_CVPR_2020_paper.pdf | cvpr-2020-6 | ['3d-shape-representation'] | ['computer-vision'] | [-1.31994560e-01 5.12033284e-01 7.39458278e-02 -9.13469940e-02
-7.18750894e-01 -7.27733314e-01 6.27758086e-01 -9.67678055e-02
2.27240831e-01 4.57810014e-01 -6.55309930e-02 -2.81278491e-01
2.52461713e-02 -1.48766160e+00 -1.24346387e+00 -5.95294297e-01
2.32633986e-02 9.90473211e-01 2.26563364e-01 -2.04329550e-01
-2.32786629e-02 1.01207185e+00 -1.51327479e+00 3.39758217e-01
9.37396288e-01 9.22356367e-01 2.66869878e-03 5.18676221e-01
-1.62262976e-01 1.04104862e-01 6.57983571e-02 -4.27622736e-01
4.59932774e-01 -9.68686566e-02 -7.22983241e-01 8.11895207e-02
3.98254484e-01 -3.88985336e-01 -2.67307341e-01 6.59666121e-01
2.81398088e-01 -6.71076700e-02 1.00346017e+00 -1.07331657e+00
-3.93287271e-01 1.82202473e-01 -4.13638502e-01 -6.33417606e-01
1.00264862e-01 -1.82142660e-01 9.11557376e-01 -1.44731903e+00
7.68152475e-01 1.24580455e+00 1.09841835e+00 4.30758774e-01
-1.70698118e+00 -5.07135928e-01 -1.89935818e-01 -5.38556635e-01
-1.45715547e+00 -2.52962768e-01 1.04157162e+00 -5.06087899e-01
5.83875656e-01 5.56726873e-01 1.15339351e+00 8.09367895e-01
1.27403691e-01 8.04191351e-01 8.77335072e-01 -1.22622527e-01
3.77706349e-01 -2.21586198e-01 -2.66622931e-01 1.00415659e+00
1.46425813e-01 -3.11080404e-02 -1.93757311e-01 -3.89714688e-01
1.68574798e+00 7.24296793e-02 -2.05192208e-01 -9.14584339e-01
-9.41422284e-01 8.47162843e-01 6.88515186e-01 -6.47711828e-02
-2.61571974e-01 4.02180433e-01 1.45104885e-01 1.24759875e-01
8.47778499e-01 2.53089219e-01 -4.15377378e-01 3.77986394e-02
-9.49082494e-01 5.69433630e-01 9.61522102e-01 1.20300186e+00
9.20035243e-01 7.03827068e-02 3.94059211e-01 7.61408687e-01
3.44686896e-01 5.64772427e-01 -3.44025463e-01 -1.21123672e+00
2.34194428e-01 6.67968571e-01 -2.81203561e-03 -1.25071859e+00
-3.44648570e-01 -2.16336578e-01 -1.22162330e+00 5.01783669e-01
2.77498931e-01 4.20476720e-02 -9.36702549e-01 1.22730720e+00
5.60943305e-01 1.98021621e-01 -3.92201185e-01 5.61175823e-01
9.13237989e-01 8.77608180e-01 -5.49091637e-01 2.85541862e-01
1.02905285e+00 -5.50893307e-01 9.54205170e-02 3.45431983e-01
5.27541876e-01 -2.82556534e-01 9.75776196e-01 5.85241497e-01
-1.43339348e+00 -3.95543069e-01 -8.86280179e-01 -2.89592773e-01
-1.30531818e-01 -1.24908701e-01 7.04856217e-01 4.01910514e-01
-1.12214208e+00 9.77316558e-01 -1.12811279e+00 1.52315229e-01
9.81395304e-01 3.34820300e-01 -3.05744380e-01 5.16605377e-02
-6.11952782e-01 4.48780268e-01 1.39952749e-01 1.37332007e-02
-9.45098758e-01 -1.09270799e+00 -9.75844383e-01 3.89660173e-03
1.28619552e-01 -1.05868781e+00 1.05987442e+00 -7.90785849e-01
-1.60069370e+00 1.02044415e+00 -8.30904692e-02 -1.99533641e-01
7.71281898e-01 -4.52630818e-02 3.12665492e-01 1.94823205e-01
-1.61125675e-01 6.92540884e-01 1.04011965e+00 -1.78514254e+00
-1.93342298e-01 -2.57372916e-01 8.47756714e-02 8.78214762e-02
1.17500372e-01 -6.20857120e-01 -5.45532763e-01 -7.62420952e-01
5.94827175e-01 -8.89533937e-01 -3.35374624e-01 3.83765787e-01
-6.04054332e-01 -2.99304843e-01 8.16431344e-01 -5.84209263e-01
7.76422083e-01 -1.91296470e+00 3.85169178e-01 5.87670088e-01
6.14828289e-01 -4.21233103e-02 1.24000832e-01 3.98586214e-01
2.27544755e-02 4.57266510e-01 -7.02192962e-01 -6.53668642e-01
9.90931317e-02 3.62224519e-01 -4.65572298e-01 5.71625352e-01
2.83393264e-01 9.34555113e-01 -8.48211467e-01 -2.96232790e-01
3.51397783e-01 7.69257724e-01 -9.37306702e-01 6.11828975e-02
-5.96828699e-01 6.01912737e-01 -4.31765437e-01 7.28775799e-01
9.64294851e-01 -1.36988044e-01 3.27638239e-02 -1.87873095e-01
-2.07319304e-01 2.41428688e-01 -1.24649811e+00 1.70955229e+00
-6.01021051e-01 2.75810212e-01 3.49022895e-01 -1.00770199e+00
9.20715928e-01 4.12564397e-01 5.39030373e-01 -2.48531476e-01
1.43495813e-01 4.92025733e-01 -6.04458690e-01 6.03354536e-02
4.13153097e-02 -1.92419305e-01 4.10396792e-02 2.65989780e-01
-1.34199157e-01 -9.60572302e-01 -1.88719854e-01 -2.50292700e-02
7.36366928e-01 5.94808280e-01 -5.42781800e-02 -4.36183006e-01
1.96277857e-01 -5.75940944e-02 5.91800690e-01 3.59342992e-01
6.74699545e-01 1.00035107e+00 5.64239383e-01 -7.17962861e-01
-1.37880087e+00 -1.45740438e+00 -5.33735156e-01 6.96390629e-01
6.80487081e-02 -4.38909829e-01 -8.71163785e-01 -3.52412105e-01
2.27651954e-01 4.71160442e-01 -6.15347624e-01 3.33280236e-01
-9.26239133e-01 -2.46704966e-01 3.25498790e-01 5.44011772e-01
2.37042457e-01 -9.83834684e-01 -4.45419967e-01 1.89877659e-01
3.36350262e-01 -6.85765505e-01 -2.54414290e-01 7.32858106e-02
-1.36201656e+00 -1.07553530e+00 -8.36399794e-01 -9.39335525e-01
1.03906989e+00 1.21463865e-01 1.38281000e+00 3.29610795e-01
-4.39098142e-02 4.35914099e-01 1.60895120e-02 -3.13154817e-01
-4.61979091e-01 2.14389205e-01 -1.24632128e-01 -1.64966565e-02
-4.23371971e-01 -1.32507181e+00 -5.61549783e-01 2.41082087e-01
-8.43949974e-01 7.35585928e-01 2.05360055e-01 6.15442693e-01
1.24567592e+00 4.07216772e-02 2.70530850e-01 -9.97577369e-01
1.11772045e-01 -4.47113603e-01 -7.87356257e-01 -1.89796045e-01
1.64087396e-02 -1.29262432e-02 7.93236792e-01 -1.15677826e-01
-7.27657378e-01 2.05538511e-01 -4.02149528e-01 -6.81257367e-01
-1.20641842e-01 3.56639326e-01 -1.64594457e-01 -1.84199303e-01
4.40120846e-01 2.51354307e-01 -1.07244052e-01 -6.72617316e-01
5.57893872e-01 1.64760098e-01 3.70564550e-01 -9.97736931e-01
1.12914813e+00 8.45580459e-01 2.94356436e-01 -1.22874177e+00
-6.57127678e-01 -1.84097011e-02 -9.31811512e-01 -1.28552586e-01
7.19202340e-01 -7.43975699e-01 -6.77194417e-01 4.96457100e-01
-1.25130165e+00 -7.40485787e-01 -4.63311195e-01 -1.08085744e-01
-9.77706015e-01 7.08069280e-02 -6.32567942e-01 -6.39342964e-01
-3.64247620e-01 -9.09914970e-01 1.26074469e+00 -7.37865046e-02
-1.85823068e-01 -1.04258227e+00 1.55844577e-02 8.36845189e-02
9.64139104e-02 7.60444164e-01 1.14256477e+00 -5.36304973e-02
-8.27019095e-01 -1.22437589e-02 -2.83072125e-02 1.73185036e-01
1.06504574e-01 2.96777844e-01 -8.46415043e-01 -2.17173293e-01
-1.62158918e-04 -2.04667509e-01 6.73005819e-01 4.31422621e-01
1.75562131e+00 -4.76735353e-01 -3.83216798e-01 1.29909289e+00
1.34637654e+00 -1.60326734e-01 5.89665413e-01 -3.65580320e-02
1.16975689e+00 4.02278721e-01 -1.83840618e-01 4.05136377e-01
5.22694051e-01 4.41925913e-01 6.57180071e-01 -2.62598664e-01
-1.45912170e-01 -6.00189924e-01 -3.45666818e-02 9.85384405e-01
-5.44784546e-01 2.00793445e-01 -1.11654842e+00 3.48147392e-01
-1.56967711e+00 -6.60886407e-01 -3.94314051e-01 2.22693086e+00
9.93462503e-01 1.53315440e-01 2.11529374e-01 1.46326989e-01
3.70508671e-01 1.58708207e-02 -5.43252468e-01 -3.01618785e-01
3.20011145e-03 6.22475922e-01 3.27255547e-01 7.10648894e-01
-1.02524662e+00 9.59318936e-01 5.47188711e+00 8.54738295e-01
-1.12911999e+00 -8.66374522e-02 7.39898086e-01 1.67335257e-01
-7.73815036e-01 -2.15910003e-01 -4.91077632e-01 2.26910800e-01
4.68640745e-01 -6.65868521e-02 3.50268424e-01 9.91566300e-01
7.66142979e-02 2.75134683e-01 -1.27265120e+00 9.82742131e-01
-2.90516943e-01 -1.86882544e+00 2.20681340e-01 1.55380994e-01
9.14763153e-01 1.19974203e-01 2.72583459e-02 1.26986355e-01
3.90298247e-01 -1.32602608e+00 9.86044049e-01 4.89616483e-01
1.12099898e+00 -9.47207630e-01 1.18623152e-01 6.32239819e-01
-1.37977946e+00 5.29556215e-01 -5.75075984e-01 1.07855223e-01
7.55697340e-02 7.74016917e-01 -5.54126740e-01 6.20001256e-01
5.77242255e-01 8.22273850e-01 8.91499780e-03 9.21200156e-01
-2.28510007e-01 6.45133734e-01 -7.04566181e-01 3.33514005e-01
6.38552010e-02 -6.89351439e-01 6.60112143e-01 1.04990637e+00
4.83820438e-01 3.12464535e-01 1.93317711e-01 1.29268873e+00
-4.45204258e-01 1.73757542e-02 -8.85519862e-01 2.50990123e-01
4.57188368e-01 1.08607411e+00 -9.91240442e-01 -2.15436086e-01
-1.59072667e-01 7.77984083e-01 4.36965376e-01 2.36708313e-01
-7.45340109e-01 -1.01690993e-01 5.46341598e-01 6.61804855e-01
5.13474524e-01 -5.97090662e-01 -7.53290236e-01 -9.70776320e-01
4.55204546e-02 -6.77311003e-01 -1.61606342e-01 -6.45399928e-01
-1.34668434e+00 5.73841810e-01 -2.44436855e-03 -1.21828377e+00
2.94406444e-01 -6.35981083e-01 -6.41139328e-01 7.64329970e-01
-1.19144225e+00 -1.34868240e+00 -3.11090022e-01 4.70783651e-01
4.94735926e-01 3.15072298e-01 9.42423046e-01 -7.59492293e-02
-2.00772077e-01 2.24838227e-01 1.13169082e-01 1.60333931e-01
-3.67070474e-02 -1.37903082e+00 7.91863739e-01 3.72309893e-01
1.89179197e-01 5.30266285e-01 2.79690325e-01 -7.40581691e-01
-1.68459928e+00 -1.20333397e+00 3.34537238e-01 -5.93668878e-01
1.90349668e-01 -7.80116677e-01 -1.08076572e+00 7.12566853e-01
-3.44821215e-01 2.74207145e-01 3.57978374e-01 -3.80136296e-02
-4.91824567e-01 1.40559033e-01 -1.15215743e+00 7.86367953e-01
1.28859258e+00 -3.76472741e-01 -5.10002553e-01 3.08269203e-01
7.77382493e-01 -9.12503958e-01 -9.99600232e-01 3.73998553e-01
5.06883502e-01 -1.03684473e+00 1.29110253e+00 -3.90819728e-01
7.06662118e-01 -2.33287588e-01 -7.80638531e-02 -1.32303047e+00
-3.63789380e-01 -7.83563614e-01 -4.06086862e-01 7.32453108e-01
3.55483025e-01 -7.10165083e-01 1.11122990e+00 4.49181408e-01
-5.79628885e-01 -1.38717031e+00 -9.75495100e-01 -7.06542075e-01
6.27123773e-01 -5.40101349e-01 6.47020102e-01 1.05088234e+00
-4.66631830e-01 -9.51540098e-02 1.81301013e-02 7.59970471e-02
7.70434916e-01 3.56804997e-01 9.22875166e-01 -1.65180707e+00
-2.35248297e-01 -5.92038155e-01 -5.11875413e-02 -1.43676567e+00
7.32349902e-02 -1.26959026e+00 -9.77483764e-02 -1.78640962e+00
-2.64417231e-01 -1.04368544e+00 4.39305127e-01 5.36194861e-01
3.57810020e-01 4.55219090e-01 2.60938238e-02 2.54333764e-01
-1.59170926e-02 8.65652621e-01 1.61874890e+00 9.46608186e-02
-4.76502985e-01 1.45093948e-01 -3.90349716e-01 1.25197065e+00
8.19982767e-01 -3.54383647e-01 -3.03600073e-01 -5.73526621e-01
4.93152440e-01 1.95023596e-01 4.35938954e-01 -8.00542951e-01
-2.36161556e-02 -9.76463035e-02 5.10082483e-01 -9.08318162e-01
5.16477525e-01 -8.44717264e-01 4.81462657e-01 2.81163871e-01
1.29726082e-01 -1.21966973e-01 2.62956351e-01 3.31959456e-01
2.25589529e-01 -2.20928192e-02 7.74343908e-01 -2.80941546e-01
9.57533643e-02 8.52213442e-01 -2.94721127e-02 1.83179513e-01
8.32214355e-01 -4.33981955e-01 1.91246703e-01 -1.35032922e-01
-7.23624885e-01 3.43513885e-03 9.39288378e-01 1.45082725e-02
8.13865423e-01 -1.64102983e+00 -7.30752885e-01 2.49403968e-01
-3.95625353e-01 1.20547569e+00 1.30711809e-01 4.62476611e-01
-1.13534594e+00 7.92028978e-02 7.49001727e-02 -8.32989752e-01
-8.15651298e-01 1.44699961e-01 5.13096035e-01 -5.46664223e-02
-1.33584261e+00 9.46524441e-01 5.31878293e-01 -8.51384163e-01
1.23353265e-01 -6.87705278e-01 3.70647430e-01 -7.40439966e-02
1.05046630e-01 3.41776609e-01 2.05603048e-01 -4.05046493e-01
-1.00272469e-01 7.82675147e-01 4.73731101e-01 1.17001999e-02
1.86509490e+00 5.29036522e-01 -3.62089187e-01 4.47642565e-01
1.14528823e+00 1.27468154e-01 -1.53454137e+00 -1.40747547e-01
-3.36124539e-01 -3.49433452e-01 -1.81066319e-02 -1.85756460e-01
-1.16157901e+00 8.32914650e-01 -2.47460276e-01 3.82975042e-01
7.11994410e-01 1.25863418e-01 9.60335314e-01 3.03097188e-01
6.46297157e-01 -6.55463874e-01 -4.21071472e-03 7.51481771e-01
1.47202671e+00 -8.04870486e-01 2.16080602e-02 -8.79129589e-01
1.30087556e-03 1.36526763e+00 2.06864625e-01 -6.65795982e-01
1.05972850e+00 4.09228146e-01 -4.29762363e-01 -4.87371325e-01
-2.84108490e-01 2.70850450e-01 4.98537093e-01 6.54629827e-01
2.80277073e-01 2.14224368e-01 2.88320899e-01 5.57218373e-01
-6.73864543e-01 -2.99764276e-01 2.94075102e-01 8.52942169e-01
-3.56664419e-01 -9.67918158e-01 -3.49076539e-01 5.08497715e-01
-1.52778870e-03 3.37260924e-02 -2.92267472e-01 7.21379876e-01
1.21857964e-01 1.04966320e-01 2.98670083e-01 -1.16951324e-01
3.19398671e-01 -1.45751581e-01 8.13890636e-01 -8.19532692e-01
-2.69690603e-01 -1.66290980e-02 -1.63187385e-01 -6.66840494e-01
-1.04645051e-01 -6.34643137e-01 -1.41990864e+00 -5.40950835e-01
1.40910655e-01 -6.10789433e-02 4.75512236e-01 5.34216940e-01
4.10545290e-01 2.98833281e-01 7.43085861e-01 -1.65681612e+00
-2.23575637e-01 -4.26287144e-01 -4.53157395e-01 1.91866890e-01
1.82860568e-01 -6.64414167e-01 -3.12449396e-01 7.33019859e-02] | [8.671662330627441, -3.6393113136291504] |
fd4b5232-eac9-4386-8146-4303dfa194c7 | extended-fastslam-using-cellular-multipath | 2301.07560 | null | https://arxiv.org/abs/2301.07560v2 | https://arxiv.org/pdf/2301.07560v2.pdf | Extended FastSLAM Using Cellular Multipath Component Delays and Angular Information | Opportunistic navigation using cellular signals is appealing for scenarios where other navigation technologies face challenges. In this paper, long-term evolution (LTE) downlink signals from two neighboring commercial base stations (BS) are received by a massive antenna array mounted on a passenger vehicle. Multipath component (MPC) delays and angle-of-arrival (AOA) extracted from the received signals are used to jointly estimate the positions of the vehicle, transmitters, and virtual transmitters (VT) with an extended fast simultaneous localization and mapping (FastSLAM) algorithm. The results show that the algorithm can accurately estimate the positions of the vehicle and the transmitters (and virtual transmitters). The vehicle's horizontal position error of SLAM fused with proprioception is less than 6 meters after a traversed distance of 530 meters, whereas un-aided proprioception results in a horizontal error of 15 meters. | ['Fredrik Tufvesson', 'Russ Whiton', 'Junshi Chen'] | 2023-01-18 | null | null | null | null | ['simultaneous-localization-and-mapping'] | ['computer-vision'] | [-3.69392663e-01 1.28177935e-02 1.40369073e-01 -1.61519591e-02
-8.16708624e-01 -7.67433345e-01 3.52728009e-01 -4.46974665e-01
-5.93124628e-01 1.20296681e+00 -5.17871827e-02 -6.95109248e-01
-8.51999447e-02 -5.05472600e-01 -7.15645373e-01 -9.05275881e-01
-3.57149452e-01 2.82366931e-01 -2.26116423e-02 -2.10936308e-01
1.59173548e-01 4.81538683e-01 -9.96409178e-01 -4.47226346e-01
8.91198218e-01 1.31736374e+00 2.90488005e-01 8.05678308e-01
1.06339306e-01 2.34181494e-01 -6.25628412e-01 1.29166851e-02
2.50933766e-01 -7.33693764e-02 1.01320311e-01 -3.97872478e-01
1.16543204e-01 -1.81104168e-01 -6.14463747e-01 6.41823947e-01
6.27549946e-01 -1.15375035e-02 5.56268871e-01 -1.40766215e+00
5.06215692e-01 -1.53096393e-01 -4.01385814e-01 -9.08850785e-03
3.37417394e-01 -2.48864844e-01 2.75847554e-01 -8.23164582e-01
3.97654235e-01 5.98410070e-01 1.11003888e+00 -5.67454472e-02
-5.22062063e-01 -9.30062354e-01 -2.35361800e-01 4.57923859e-02
-1.81965911e+00 -7.44434118e-01 1.37907758e-01 -3.98611665e-01
2.44862333e-01 4.84294146e-01 5.69725752e-01 5.96310377e-01
1.03513348e+00 1.55509099e-01 3.68033201e-01 -2.44213641e-01
4.19415444e-01 3.77966106e-01 -2.68742532e-01 6.03956163e-01
6.74881041e-01 1.64071470e-01 -6.02296591e-01 -3.20498288e-01
5.52070618e-01 -4.44886714e-01 -5.32070160e-01 -5.62882543e-01
-1.23463750e+00 4.91202146e-01 4.01163876e-01 1.68142825e-01
-8.84805858e-01 5.85601330e-01 -4.89878863e-01 5.18875659e-01
5.79094887e-02 1.25799283e-01 -3.83923620e-01 -3.68731529e-01
-9.41616595e-01 -1.20837934e-01 7.14489162e-01 1.68196452e+00
5.95853925e-01 4.76107687e-01 1.07360929e-01 2.39285484e-01
9.47150469e-01 1.36512268e+00 8.92596319e-02 -9.69841361e-01
5.07014334e-01 -4.10679787e-01 7.46160865e-01 -1.03522635e+00
-8.07696164e-01 -1.81894100e+00 -4.96795267e-01 8.96041654e-03
4.70333457e-01 -1.24014652e+00 -3.97661209e-01 1.49800861e+00
2.81701446e-01 5.78911006e-01 4.57717031e-01 7.00797141e-01
8.96775126e-02 3.22538644e-01 -7.52583265e-01 -3.91138852e-01
9.90212977e-01 -8.32609177e-01 -8.57274652e-01 -6.61242664e-01
7.98725128e-01 -7.54671991e-01 -3.41441035e-01 5.25283039e-01
-8.06007564e-01 -3.73924643e-01 -1.40239573e+00 8.31629515e-01
1.22468762e-01 1.90535307e-01 3.69905472e-01 1.03747225e+00
-1.20145726e+00 -4.34596092e-01 -4.75215524e-01 -2.85514265e-01
-8.82290378e-02 3.83106440e-01 1.56139741e-02 -1.34672523e-01
-1.13110948e+00 7.59643734e-01 -5.19993007e-01 3.67558300e-01
-3.28329086e-01 -6.05015635e-01 -5.59683263e-01 -3.13879550e-01
-1.66694839e-02 -7.44611025e-01 1.03333306e+00 -6.00023210e-01
-1.18102252e+00 -8.78148600e-02 -8.05852711e-01 -6.13850296e-01
4.53725815e-01 -8.68356004e-02 -1.10407245e+00 -2.61335939e-01
4.98044699e-01 3.92125137e-02 2.03395024e-01 -1.35429847e+00
-1.40288913e+00 -3.12962025e-01 -4.93877798e-01 2.93637872e-01
3.61558408e-01 -7.61811018e-01 -5.66487551e-01 -1.11523196e-01
1.00200951e+00 -1.44939315e+00 -3.51490855e-01 -1.47250086e-01
-3.11317146e-01 9.14680898e-01 3.59916270e-01 -5.32343864e-01
9.47347105e-01 -2.23774362e+00 -3.14422250e-01 5.95620871e-01
-1.71874002e-01 -5.70533752e-01 2.30635896e-01 4.68036056e-01
6.23642921e-01 -5.91237485e-01 4.19565707e-01 -2.78468251e-01
-1.74129769e-01 8.09002370e-02 -3.52558911e-01 8.39472950e-01
-1.12318575e+00 5.78179896e-01 -9.43516135e-01 3.62103641e-01
2.95562614e-02 4.14647192e-01 -3.50946307e-01 -4.29935366e-01
7.18368471e-01 8.10998023e-01 -4.02375668e-01 5.69483578e-01
1.09351742e+00 4.47571456e-01 1.45392969e-01 5.36292419e-02
-7.89545298e-01 -5.71707599e-02 -1.08386707e+00 1.45135128e+00
-6.66346133e-01 1.08459008e+00 5.29480040e-01 -4.10617702e-02
6.49366796e-01 5.25060594e-01 4.56250101e-01 -9.46397543e-01
1.68183431e-01 8.27552736e-01 -8.72956738e-02 -4.54328693e-02
5.93819261e-01 6.08690418e-02 -4.42228615e-01 1.12782978e-01
-2.24160239e-01 3.00691634e-01 -4.34449404e-01 8.56173560e-02
1.23680294e+00 -3.87796313e-01 3.03151190e-01 -2.36374661e-01
5.67721307e-01 -1.83226112e-02 7.04752207e-01 1.06493986e+00
1.69360905e-03 -6.80431575e-02 -2.98734680e-02 1.77268922e-01
-3.87809932e-01 -9.41025734e-01 -1.54386535e-01 4.14418310e-01
8.13275576e-01 1.48965016e-01 -6.74961805e-02 -1.98813215e-01
6.29262865e-01 9.93011057e-01 -1.68095887e-01 1.28651351e-01
-2.33854339e-01 -4.64257210e-01 4.71077651e-01 1.27567574e-01
3.66356343e-01 3.57796460e-01 -9.78858024e-02 5.94703197e-01
-3.20887893e-01 -1.06690156e+00 -2.98525870e-01 -6.31727874e-02
-3.35054666e-01 -5.31350553e-01 -5.46528816e-01 -4.69662786e-01
6.12424254e-01 1.05397022e+00 2.09400997e-01 -4.21649098e-01
7.73111880e-01 4.94313449e-01 1.09775707e-01 -4.33926851e-01
1.66737467e-01 -3.74787837e-01 6.59885168e-01 3.83196324e-01
-8.31183866e-02 -4.57918942e-01 -6.23093188e-01 9.20602441e-01
5.42127073e-01 -1.20841496e-01 6.50664806e-01 2.06617579e-01
3.03360164e-01 3.23229134e-01 5.33339679e-01 -1.92176282e-01
2.63713524e-02 -1.01436031e+00 -8.89964879e-01 -2.64511496e-01
-5.18191874e-01 -4.63552475e-01 1.29965410e-01 2.41745085e-01
-1.01335740e+00 -4.99782562e-02 -2.15520814e-01 3.81178677e-01
1.36198327e-01 6.47672653e-01 -1.79996297e-01 -9.31721687e-01
3.87094140e-01 5.52725077e-01 -2.32755005e-01 -6.13380075e-02
2.13977918e-01 1.02646184e+00 7.19685555e-01 4.19658661e-01
9.74928916e-01 9.45403039e-01 2.96762019e-01 -9.69324827e-01
-1.23264983e-01 -6.79702342e-01 -3.78571600e-01 -6.56110525e-01
1.08833700e-01 -1.61641693e+00 -5.67454100e-01 2.62447625e-01
-9.27387297e-01 8.55812132e-02 6.68147504e-01 1.16862941e+00
-2.39677176e-01 1.54193655e-01 7.00727776e-02 -1.16993988e+00
1.44884601e-01 -7.79439986e-01 5.24648666e-01 1.57415792e-01
-3.43953818e-01 -9.07256901e-01 -1.05619214e-01 3.51324648e-01
1.02400541e+00 3.72860767e-02 2.24856567e-02 4.29474078e-02
-1.04063594e+00 -7.16166377e-01 1.46109253e-01 -6.68394089e-01
7.90136680e-03 -9.55505788e-01 -6.10908687e-01 -5.82093358e-01
4.63243499e-02 8.85814488e-01 1.03880450e-01 9.51277137e-01
-1.63051978e-01 -2.07847729e-01 -1.27110624e+00 9.15890276e-01
1.40074337e+00 8.36288035e-01 5.42504072e-01 5.39126456e-01
3.11197340e-01 -3.96035472e-03 1.06726444e+00 4.49685007e-01
6.21064365e-01 9.53345120e-01 7.60622203e-01 3.54160696e-01
1.64056525e-01 3.09208065e-01 2.57825255e-01 4.35726196e-01
1.90106213e-01 -9.64851558e-01 -6.52744830e-01 3.09592485e-01
-1.59018636e+00 -6.38925552e-01 -8.71481836e-01 2.37843251e+00
-2.10544765e-01 1.33091748e-01 -6.74185693e-01 -3.63350660e-01
3.90729815e-01 -1.53261960e-01 -4.03632849e-01 3.43817323e-02
-2.72275619e-02 -7.47177064e-01 1.76240778e+00 1.06798315e+00
-5.46487868e-01 3.33989024e-01 5.35953999e+00 6.48298979e-01
-1.08502722e+00 1.88975334e-01 -3.78556699e-01 -3.29262838e-02
-6.20213091e-01 -2.57516336e-02 -8.68946135e-01 7.15578139e-01
1.18842924e+00 -1.24847405e-02 -4.63829525e-02 6.79577291e-01
5.51754236e-01 -8.01495373e-01 -5.34853041e-01 1.02756619e+00
-1.94127616e-02 -1.40364695e+00 -8.24731052e-01 7.08874166e-01
6.67157173e-01 4.79892284e-01 2.57321268e-01 3.51757318e-01
-1.27421901e-01 -2.27142289e-01 1.03676236e+00 9.61181819e-01
5.52479565e-01 -9.82950747e-01 1.12734115e+00 4.45625037e-01
-1.31236315e+00 -4.75869775e-01 3.10281315e-03 -2.05556571e-01
8.92232001e-01 7.89219081e-01 -1.10732639e+00 8.47337365e-01
4.20508869e-02 1.96203068e-02 1.18414015e-01 1.78043056e+00
-9.32411477e-02 3.81784797e-01 -3.98972243e-01 -8.88127014e-02
4.82599527e-01 -1.79422870e-01 1.18261182e+00 8.00401330e-01
1.26688302e+00 -1.26623854e-01 -3.91187578e-01 5.66100441e-02
2.26329893e-01 -2.01711178e-01 -5.53600490e-01 8.40381444e-01
1.06781030e+00 9.05955076e-01 -2.88674146e-01 9.22920778e-02
-4.22682405e-01 9.32588398e-01 -8.23878706e-01 8.71377230e-01
-8.68573546e-01 -6.57894909e-01 9.63901341e-01 1.32600020e-04
3.50501418e-01 -8.84717405e-01 -5.32559395e-01 -4.80658799e-01
-1.03831865e-01 6.78770989e-02 -4.83625889e-01 -8.08647871e-01
-2.11656228e-01 3.70113581e-01 -1.02133071e+00 -1.98737788e+00
-3.85973662e-01 7.51844794e-03 -2.61771649e-01 1.08670700e+00
-1.31915236e+00 -7.44272411e-01 -4.75964040e-01 -6.17718149e-04
1.23694658e-01 -6.29856050e-01 6.53883100e-01 6.31860793e-01
-1.20978564e-01 5.21504164e-01 1.25319815e+00 -2.21677601e-01
3.94950867e-01 -7.42908835e-01 3.25107038e-01 8.57452393e-01
-3.09162438e-01 5.36013544e-01 1.24113476e+00 -6.97311401e-01
-1.79922760e+00 -1.06506693e+00 1.16633368e+00 -3.88203979e-01
3.28756988e-01 -3.59173268e-01 2.00661555e-01 7.16017067e-01
5.02164997e-02 -2.12869033e-01 7.40058541e-01 -1.64243445e-01
5.11833012e-01 -1.97583973e-01 -1.06674194e+00 7.66001284e-01
7.79218674e-01 -2.51500793e-02 2.73670852e-01 9.75373015e-02
1.32318825e-01 -6.75918400e-01 -1.85660273e-01 1.35811672e-01
1.20963550e+00 -7.03260839e-01 8.96434546e-01 4.34412450e-01
-1.20832515e+00 -6.57052934e-01 -5.81384063e-01 -1.34755874e+00
-4.08904076e-01 -7.05910742e-01 2.30556298e-02 7.77517915e-01
6.72259748e-01 -1.17978394e+00 8.11825395e-01 1.15123600e-01
-3.01720500e-01 -2.05343187e-01 -1.38281763e+00 -8.23060989e-01
-7.07942426e-01 -9.41649020e-01 3.29924464e-01 4.89924163e-01
-1.82584912e-01 3.87441903e-01 -6.53256774e-01 1.30511320e+00
8.91135037e-01 -2.21577868e-01 1.12981939e+00 -1.15699136e+00
-2.01277006e-02 -2.64966600e-02 -4.94396716e-01 -1.97784245e+00
-2.34513029e-01 -4.83165532e-01 1.92965284e-01 -1.68735504e+00
-7.70074129e-01 -8.74484479e-01 -8.34836736e-02 -3.49632263e-01
7.15093374e-01 1.83786005e-01 -5.42997956e-01 -7.61481002e-02
-4.39231962e-01 4.70498115e-01 5.96656859e-01 1.39779478e-01
-2.16919422e-01 1.09059572e+00 -6.21060371e-01 6.15411460e-01
4.87157315e-01 -4.07185763e-01 -3.67408246e-01 -5.76116562e-01
3.99407268e-01 8.46018195e-01 1.45563185e-01 -1.56783760e+00
8.60438585e-01 2.14281783e-01 3.99120152e-01 -1.09587979e+00
6.62493646e-01 -1.13298333e+00 8.54801357e-01 8.46594870e-01
5.05359232e-01 -5.54013968e-01 2.90026963e-01 1.08913529e+00
4.21870090e-02 7.33499825e-02 3.10029387e-01 5.98511934e-01
-8.21096241e-01 1.10855289e-01 -1.38438988e+00 -7.75642574e-01
1.20083725e+00 -4.95060146e-01 -1.69851050e-01 -1.29691589e+00
-3.61502409e-01 6.07782722e-01 2.35417455e-01 1.35452792e-01
3.76353532e-01 -1.34993196e+00 -3.83959681e-01 3.55581075e-01
-1.21522091e-01 -8.65162611e-01 4.31449443e-01 1.45154774e+00
-5.68586230e-01 1.48993194e+00 6.10060953e-02 -6.70995533e-01
-1.27961385e+00 -3.52483779e-01 6.03268862e-01 7.46355951e-01
4.00286503e-02 1.20238125e+00 -1.69005152e-02 -2.32044950e-01
6.12558983e-02 2.14515030e-01 -1.69405475e-01 7.97724444e-03
4.49607730e-01 5.28709352e-01 -1.16132135e-02 -1.10053873e+00
-7.44305849e-01 7.99100459e-01 3.91538501e-01 -7.17654884e-01
6.25371993e-01 -1.24000740e+00 3.15852076e-01 1.41956478e-01
1.01476300e+00 1.17885518e+00 -1.08469784e+00 -2.75111228e-01
-2.34336574e-02 -5.35519540e-01 3.43920708e-01 -6.08108163e-01
-6.96438074e-01 1.57856077e-01 9.05096054e-01 -1.34225562e-01
4.93372411e-01 -3.04476142e-01 7.89962173e-01 4.38827693e-01
1.56393433e+00 -5.80705643e-01 -6.12932205e-01 8.50824714e-01
3.70109200e-01 -5.22402704e-01 -3.10575694e-01 -2.57460982e-01
-2.90754020e-01 9.52181399e-01 2.82574832e-01 4.75365818e-01
7.26364493e-01 2.34269083e-01 5.04421830e-01 1.58125665e-02
-3.58739883e-01 -2.70583749e-01 5.39875031e-02 5.91141462e-01
-9.16927606e-02 2.17652749e-02 -2.24154666e-01 6.38173878e-01
-4.98457313e-01 -5.60846508e-01 6.71368182e-01 9.29972053e-01
-1.04587376e+00 -6.70841277e-01 -7.10295618e-01 2.49703839e-01
6.83177859e-02 9.32850614e-02 5.50205588e-01 6.24697924e-01
3.24313223e-01 1.06080627e+00 1.70954034e-01 -5.81999719e-01
4.06764895e-01 -2.82025576e-01 1.27416477e-01 -6.29323646e-02
4.90482926e-01 2.57219762e-01 6.15959525e-01 -5.96030056e-01
3.21123689e-01 -7.90849507e-01 -1.41633654e+00 -1.81718394e-01
-3.32099319e-01 7.11301923e-01 1.04427612e+00 1.03313076e+00
6.43128097e-01 6.51540577e-01 9.22700346e-01 -7.91649222e-01
3.38814288e-01 -4.15725440e-01 -9.91434634e-01 -1.04258716e+00
7.31365740e-01 -7.18386114e-01 -6.42831862e-01 -7.41307199e-01] | [6.266557216644287, 1.0977281332015991] |
04780bec-f26f-40a8-bfa5-ca4a9e4ac5ec | open-source-hamnosys-parser-for-multilingual | 2204.06924 | null | https://arxiv.org/abs/2204.06924v3 | https://arxiv.org/pdf/2204.06924v3.pdf | Handling sign language transcription system with the computer-friendly numerical multilabels | This paper presents our recent developments in the automatic processing of sign language corpora using the Hamburg Sign Language Annotation System (HamNoSys). We designed an automated tool to convert HamNoSys annotations into numerical labels for defined initial features of body and hand positions. Our proposed numerical multilabels greatly simplify annotations' structure without significant loss of gloss meaning. These numerical multilabels can potentially be used to feed the machine learning models, which would accelerate the development of vision-based sign language recognition. In addition, this tool can assist experts in the annotation process and help identify semantic errors. The code and sample annotations are publicly available at \url{https://github.com/hearai/parse-hamnosys}. | ['Milena Olech', 'Marta Plantykow', 'Sylwia Majchrowska'] | 2022-04-14 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [-2.06530094e-01 9.01124850e-02 -7.23618045e-02 -6.86264336e-01
-8.38034570e-01 -6.78541899e-01 4.28079337e-01 -1.66104227e-01
-6.50466442e-01 8.79805624e-01 4.65013236e-01 5.57884062e-03
9.06978250e-02 -2.85619289e-01 -1.51309490e-01 -6.15060508e-01
3.30528587e-01 6.32900059e-01 3.98249596e-01 -1.51584983e-01
1.08064562e-01 3.48273963e-01 -1.59182084e+00 3.38275313e-01
5.92316508e-01 7.94341326e-01 1.17095195e-01 8.02895129e-01
-3.33969921e-01 8.25579405e-01 -4.50420022e-01 -5.26520252e-01
3.15979689e-01 -3.00869554e-01 -8.41931462e-01 -5.20289600e-01
6.95628643e-01 -3.95926595e-01 7.29438150e-03 1.25672054e+00
7.22798765e-01 -1.34127781e-01 5.73196590e-01 -1.05993009e+00
-3.37255478e-01 7.11603820e-01 3.40975188e-02 -3.05237144e-01
5.63085675e-01 1.12017438e-01 9.98383880e-01 -7.37819493e-01
1.03270650e+00 9.82743323e-01 8.02302420e-01 8.02494645e-01
-3.37604940e-01 -6.89719737e-01 -2.02738032e-01 2.95776844e-01
-1.29296374e+00 -3.01080853e-01 6.10817730e-01 -7.34138668e-01
6.30406976e-01 3.38166356e-01 8.90814602e-01 6.78161979e-01
-4.44362074e-01 9.85001624e-01 1.11708546e+00 -9.06241536e-01
-1.80890962e-01 -1.71629310e-01 5.79426706e-01 1.25993586e+00
2.38726959e-01 -5.08635975e-02 -5.73506296e-01 -1.44280111e-02
8.00935328e-01 -5.11891842e-01 -9.25539955e-02 -2.42168829e-01
-1.09851158e+00 3.67214650e-01 4.92624231e-02 5.62635660e-01
-2.13183716e-01 4.11762506e-01 6.17646039e-01 1.18683055e-01
-1.42099082e-01 -3.51182148e-02 -3.93313617e-01 -5.19751549e-01
-6.44872963e-01 2.67868698e-01 5.62003553e-01 1.08797336e+00
3.20544720e-01 -3.01920958e-02 -1.63611323e-02 1.04519105e+00
6.16976798e-01 6.60098791e-01 4.95630592e-01 -1.24007714e+00
3.38598698e-01 5.78988731e-01 3.32281888e-01 -3.46240938e-01
-6.12963378e-01 3.80308360e-01 -1.38857022e-01 7.24880159e-01
1.06471741e+00 -1.07737184e-01 -1.15504348e+00 1.24262309e+00
2.66519397e-01 -5.28170168e-01 -9.71062705e-02 1.12407470e+00
1.22155094e+00 -1.00019045e-01 3.87372345e-01 2.50287652e-01
1.57188284e+00 -8.20418417e-01 -1.01352549e+00 4.04741019e-01
9.55850840e-01 -1.09585488e+00 1.26069498e+00 4.05007988e-01
-8.65911245e-01 -2.37052709e-01 -7.06858754e-01 -4.38167006e-01
-6.41862333e-01 7.85716355e-01 6.26887739e-01 7.90879607e-01
-8.94322634e-01 9.27456021e-02 -8.79561663e-01 -6.09319389e-01
3.28187823e-01 3.56007457e-01 -3.62667620e-01 1.69998407e-01
-1.13323450e+00 1.08620560e+00 3.72758865e-01 3.12120706e-01
-2.01039508e-01 3.85476835e-02 -8.24313641e-01 -6.76107407e-01
1.73403695e-01 -4.10014033e-01 1.78425050e+00 -9.59002495e-01
-1.49203372e+00 1.40200186e+00 -2.71049559e-01 -2.55012006e-01
8.75039577e-01 -4.21604663e-01 -4.75126028e-01 2.74704903e-01
2.59760451e-02 6.91749275e-01 5.32561898e-01 -9.85623181e-01
-9.00019705e-01 -1.61583915e-01 -4.14550416e-02 1.18753485e-01
1.56847149e-01 6.41510844e-01 -5.21090806e-01 -8.05888772e-01
9.00474712e-02 -8.19323599e-01 1.74998119e-01 5.42237580e-01
-8.74317959e-02 -3.52544218e-01 4.34023350e-01 -1.23619354e+00
1.11525142e+00 -1.96818805e+00 -2.09644303e-01 3.40070099e-01
-3.46932970e-02 8.34063113e-01 8.59821290e-02 2.06472948e-01
2.19622627e-01 -6.53338283e-02 -1.52755082e-01 -1.88802466e-01
1.99217841e-01 5.13591230e-01 1.40215099e-01 1.54960230e-01
-3.88362855e-01 9.60268974e-01 -1.04012656e+00 -8.31144333e-01
6.52473927e-01 4.80926603e-01 -1.10847577e-01 -2.73794364e-02
-2.09643707e-01 5.15559375e-01 -3.26493621e-01 9.97591376e-01
2.89535403e-01 3.11499059e-01 5.29951043e-02 -1.95683479e-01
-2.93182760e-01 1.26808688e-01 -1.16270041e+00 1.63136649e+00
-3.73871356e-01 7.69550145e-01 1.40803203e-01 -4.57528800e-01
6.76544309e-01 5.08445740e-01 5.54918826e-01 -2.90928960e-01
5.58637738e-01 5.12431085e-01 -1.34046197e-01 -9.36435759e-01
3.59074593e-01 -3.33824800e-03 -9.71270576e-02 1.38702646e-01
3.48070264e-02 -1.25585109e-01 5.55611014e-01 -1.16155535e-01
6.02733076e-01 5.26839137e-01 7.17340112e-01 1.85473084e-01
6.75412118e-01 1.87400818e-01 5.15717447e-01 3.51211935e-01
-5.04432857e-01 3.99602354e-01 1.03052244e-01 -5.33096135e-01
-7.70211995e-01 -1.04154551e+00 -1.64647013e-01 9.44742024e-01
-2.93087274e-01 -5.57053506e-01 -7.47439444e-01 -6.39876008e-01
-1.76560074e-01 4.55383390e-01 -1.77987248e-01 6.52429581e-01
-5.93385935e-01 6.94321617e-02 1.12637413e+00 8.11209023e-01
4.43777174e-01 -1.37258410e+00 -9.35108840e-01 -5.58179356e-02
-3.08168650e-01 -1.05528474e+00 -4.91751194e-01 -1.36947751e-01
-4.28104967e-01 -1.41250324e+00 -9.80854273e-01 -1.17592740e+00
8.46681118e-01 -5.60309291e-01 4.40920115e-01 7.16002509e-02
-5.75458825e-01 6.81829810e-01 -8.43317926e-01 -6.59702122e-01
-5.45628488e-01 -2.62988925e-01 -1.26741692e-01 -5.51712751e-01
7.03201294e-01 -2.99626678e-01 -3.05448651e-01 2.99318761e-01
-6.81125879e-01 2.64058597e-02 1.95202112e-01 7.81853914e-01
4.23784286e-01 -6.97993755e-01 1.88776061e-01 -5.97884119e-01
6.18395567e-01 3.85130465e-01 -9.34628189e-01 3.72893810e-01
-3.86080205e-01 1.91829607e-01 1.72025025e-01 -9.30131376e-02
-1.13466668e+00 5.32974362e-01 -6.80904329e-01 9.72527340e-02
-5.88499129e-01 3.06268424e-01 -1.89978212e-01 -4.07359928e-01
4.28503990e-01 -5.31701855e-02 -4.38346155e-02 -6.59107745e-01
7.98834860e-01 1.09570825e+00 8.74666870e-01 -4.65984404e-01
3.20246160e-01 3.51248533e-01 -8.31533968e-02 -9.28283691e-01
-4.87157375e-01 -8.55687082e-01 -1.15580034e+00 -6.30785167e-01
9.60477889e-01 -5.40608644e-01 -6.95151329e-01 8.24178278e-01
-1.15659535e+00 -3.76133889e-01 -3.65271866e-01 5.22742033e-01
-8.16778183e-01 7.93455124e-01 -4.69462395e-01 -8.72080684e-01
-4.75687921e-01 -7.81096518e-01 9.58930135e-01 2.07992554e-01
-7.33488679e-01 -8.64467978e-01 1.19624205e-01 7.48700082e-01
-1.07075423e-01 1.87896624e-01 6.42307639e-01 -5.11674285e-01
-1.72715575e-01 -4.96785671e-01 -1.44247249e-01 7.19120562e-01
6.60487115e-02 8.21451657e-03 -7.80084133e-01 3.35371047e-01
-8.62746596e-01 -3.92534792e-01 5.77953160e-01 4.01994795e-01
6.46785676e-01 -1.29312485e-01 -5.05872518e-02 3.70442897e-01
1.11336875e+00 3.10693204e-01 5.12346148e-01 2.23959908e-01
6.91750467e-01 5.54979503e-01 9.74747121e-01 4.77394670e-01
4.16428894e-01 9.31941330e-01 9.43600461e-02 2.11426646e-01
-7.76039958e-01 -2.71064699e-01 1.95216998e-01 8.84949327e-01
-8.49008501e-01 1.42060161e-01 -1.22057748e+00 5.65522969e-01
-2.00227547e+00 -9.66935039e-01 -5.37473321e-01 1.84867918e+00
9.19228315e-01 -3.00376743e-01 2.68622309e-01 2.59021908e-01
5.39042234e-01 -1.49637192e-01 1.43142799e-02 -4.74863797e-01
-1.19315647e-01 3.97862464e-01 6.97960496e-01 8.78954589e-01
-1.23940361e+00 1.44852388e+00 6.21972895e+00 4.10403669e-01
-9.20300186e-01 3.38091016e-01 -6.19403601e-01 4.76403534e-02
2.85517275e-01 -1.57995299e-01 -9.47747111e-01 3.70944053e-01
4.88140166e-01 1.08423427e-01 1.76535040e-01 8.49494398e-01
4.80152696e-01 -2.30515420e-01 -6.96638703e-01 1.15851581e+00
1.72259018e-01 -1.03185225e+00 3.43356468e-02 -3.25325072e-01
3.41334224e-01 1.70393363e-01 -4.82907295e-01 -1.44504547e-01
5.75078011e-01 -3.80283207e-01 9.56363320e-01 9.14473057e-01
1.10859537e+00 -1.70034617e-01 9.10372913e-01 -1.16267793e-01
-1.39436877e+00 1.18640542e-01 9.38985422e-02 -1.06887095e-01
6.75182760e-01 1.40170187e-01 -9.76381838e-01 3.82146448e-01
4.44392890e-01 3.28055263e-01 -4.81016397e-01 1.47839737e+00
-9.63705838e-01 5.47354043e-01 -5.84498584e-01 -2.71800995e-01
-3.30692567e-02 -6.53045624e-02 6.97156191e-01 1.34608448e+00
2.20097587e-01 8.76350924e-02 2.25882120e-02 2.20449314e-01
3.10988933e-01 5.86531401e-01 -2.67055124e-01 -4.43590134e-02
2.68882185e-01 7.65739143e-01 -7.98432887e-01 -4.14059013e-01
-3.88131797e-01 1.04120946e+00 -4.73579951e-02 2.31315151e-01
-6.99213505e-01 -6.68332338e-01 5.15083134e-01 2.55012631e-01
1.03971750e-01 -4.06166106e-01 -4.27486002e-01 -1.04114664e+00
4.30742085e-01 -5.96183896e-01 6.29052520e-01 -1.04316306e+00
-8.15235317e-01 2.97135293e-01 2.14029644e-02 -1.41783249e+00
-5.01230597e-01 -1.04015958e+00 -1.44410402e-01 6.49414897e-01
-1.14642227e+00 -1.70154023e+00 -4.55402017e-01 6.20146811e-01
4.21629727e-01 -1.16826236e-01 1.15965104e+00 5.32381415e-01
2.65558660e-02 7.55150616e-01 -1.13348223e-01 7.59983838e-01
9.99388814e-01 -1.18330324e+00 -4.58706357e-02 6.85179353e-01
3.34545851e-01 4.62527782e-01 5.65145254e-01 -8.49569082e-01
-3.05411011e-01 -7.03536808e-01 1.34576654e+00 -5.56848109e-01
6.42967820e-01 1.66843325e-01 -1.46524385e-01 7.01079726e-01
2.65185479e-02 -7.09605664e-02 7.22175300e-01 -1.95903435e-01
-3.86387110e-01 2.25687206e-01 -1.09082699e+00 5.10146916e-01
1.18441319e+00 -5.53255379e-01 -1.03936052e+00 3.80316257e-01
-2.92005036e-02 -5.39240122e-01 -7.35158563e-01 2.36960098e-01
1.31022370e+00 -4.58939195e-01 5.49587488e-01 -5.27545750e-01
-1.23711206e-01 -4.95660901e-01 -2.68932153e-02 -8.66709173e-01
1.96250007e-01 -3.82308453e-01 1.72321782e-01 1.13176370e+00
4.58204299e-01 -6.06380522e-01 6.07925355e-01 8.03799629e-01
-2.70247519e-01 -1.34150520e-01 -1.08906245e+00 -8.02918851e-01
-3.10629278e-01 -8.31234992e-01 1.24014832e-01 6.98587060e-01
4.97425884e-01 -4.17270660e-01 -4.69585866e-01 8.18830132e-02
5.85165083e-01 -3.60800892e-01 6.05669618e-01 -1.30539382e+00
1.39318004e-01 -5.38377583e-01 -8.81488979e-01 -7.16593325e-01
1.31740034e-01 -9.78210270e-01 1.28409341e-01 -1.86746335e+00
-4.61070418e-01 -3.12812001e-01 -8.61431360e-02 1.17498314e+00
3.36106479e-01 4.36053962e-01 3.60918820e-01 1.65831149e-01
-5.18951654e-01 -2.54119467e-02 8.99863660e-01 1.29287347e-01
-1.25826418e-01 8.43589380e-02 -2.47200355e-02 1.26542628e+00
1.12492585e+00 -4.12802994e-01 1.40893281e-01 -4.00458217e-01
3.36759612e-02 -4.35232550e-01 5.34031093e-01 -1.07056117e+00
2.17685774e-01 1.27744615e-01 -1.69761360e-01 -6.53836250e-01
1.97959647e-01 -8.99821579e-01 -1.36212241e-02 4.79980379e-01
-3.57440859e-01 -2.78817385e-01 -7.90841430e-02 -1.37333106e-02
-2.94161201e-01 -5.65039456e-01 6.10884070e-01 -2.65709132e-01
-1.11815763e+00 -2.88735002e-01 -7.06872582e-01 -1.30317584e-01
9.18161809e-01 -4.04955000e-01 -1.41461328e-01 -3.89686972e-01
-1.34198034e+00 2.62309074e-01 3.28606695e-01 4.08386022e-01
5.25759697e-01 -1.29948878e+00 -2.12822706e-01 1.19997226e-01
3.38946700e-01 -2.44605735e-01 -1.60867684e-02 7.82509685e-01
-1.27736008e+00 5.57943225e-01 -6.35726750e-01 -2.03880131e-01
-2.08925915e+00 -1.32973403e-01 1.99580088e-01 4.67265882e-02
-6.17161274e-01 8.40014338e-01 -7.14285970e-01 -5.14926672e-01
7.06820548e-01 -8.10165584e-01 -3.71760339e-01 1.03112385e-01
4.67127979e-01 5.80841959e-01 -1.85294569e-01 -9.89997327e-01
-3.70224148e-01 1.03878236e+00 3.87902081e-01 -4.98510540e-01
8.67494583e-01 -1.27967298e-01 -7.58205950e-02 4.53919768e-01
5.58628559e-01 1.99933112e-01 -6.07278705e-01 -4.84144092e-02
2.99743265e-01 -3.56635451e-01 -2.63037652e-01 -1.24117279e+00
-6.49528682e-01 7.20139265e-01 1.01306367e+00 -5.39473057e-01
9.77058828e-01 1.79415733e-01 7.25199580e-01 5.21573603e-01
6.80319428e-01 -1.49074340e+00 -6.52879834e-01 4.74682271e-01
9.49515760e-01 -1.05714595e+00 -1.61048859e-01 -5.51782370e-01
-7.58264184e-01 1.21075559e+00 2.39911214e-01 2.33111292e-01
7.75676489e-01 4.92067367e-01 9.28938508e-01 -1.96380079e-01
1.12050861e-01 -7.85609663e-01 4.05948341e-01 9.07525301e-01
5.77406943e-01 4.21024889e-01 -1.31662023e+00 7.25982964e-01
-3.28568459e-01 5.91510415e-01 2.42331743e-01 1.10336304e+00
-4.83326554e-01 -1.52499521e+00 -4.29734856e-01 3.31726640e-01
-2.94352710e-01 -6.89393580e-02 -6.97350860e-01 7.41730809e-01
7.26002932e-01 5.31061709e-01 -5.62546372e-01 -1.02059655e-01
5.44442415e-01 7.90360510e-01 6.64485872e-01 -4.76930082e-01
-3.52749586e-01 1.29184080e-02 6.93957210e-01 -5.22996366e-01
-9.34422493e-01 -8.86287868e-01 -1.82801700e+00 3.96069288e-01
-2.29763445e-02 -7.87597988e-03 6.42232120e-01 9.37834501e-01
-2.66628534e-01 6.08443469e-02 -5.43663681e-01 -5.87828279e-01
-3.50904644e-01 -1.08996272e+00 -4.80385542e-01 4.06466722e-01
5.80103556e-03 -7.61044085e-01 -9.11574662e-02 7.02930570e-01] | [9.129755973815918, -6.428381443023682] |
dc48ac40-ca61-46f6-9c72-82a6b045c6f5 | reconstruction-guided-attention-improves-the | 2209.13620 | null | https://arxiv.org/abs/2209.13620v2 | https://arxiv.org/pdf/2209.13620v2.pdf | Reconstruction-guided attention improves the robustness and shape processing of neural networks | Many visual phenomena suggest that humans use top-down generative or reconstructive processes to create visual percepts (e.g., imagery, object completion, pareidolia), but little is known about the role reconstruction plays in robust object recognition. We built an iterative encoder-decoder network that generates an object reconstruction and used it as top-down attentional feedback to route the most relevant spatial and feature information to feed-forward object recognition processes. We tested this model using the challenging out-of-distribution digit recognition dataset, MNIST-C, where 15 different types of transformation and corruption are applied to handwritten digit images. Our model showed strong generalization performance against various image perturbations, on average outperforming all other models including feedforward CNNs and adversarially trained networks. Our model is particularly robust to blur, noise, and occlusion corruptions, where shape perception plays an important role. Ablation studies further reveal two complementary roles of spatial and feature-based attention in robust object recognition, with the former largely consistent with spatial masking benefits in the attention literature (the reconstruction serves as a mask) and the latter mainly contributing to the model's inference speed (i.e., number of time steps to reach a certain confidence threshold) by reducing the space of possible object hypotheses. We also observed that the model sometimes hallucinates a non-existing pattern out of noise, leading to highly interpretable human-like errors. Our study shows that modeling reconstruction-based feedback endows AI systems with a powerful attention mechanism, which can help us understand the role of generating perception in human visual processing. | ['Gregory J. Zelinsky', 'Hossein Adeli', 'Seoyoung Ahn'] | 2022-09-27 | null | null | null | null | ['object-reconstruction'] | ['computer-vision'] | [ 6.13933682e-01 1.27711589e-03 3.28095108e-01 -1.66951627e-01
-4.62807506e-01 -5.85404217e-01 8.36329103e-01 -1.42932862e-01
-5.13904691e-01 4.68276113e-01 3.49525928e-01 -3.97081494e-01
-5.85263371e-02 -6.56195462e-01 -1.08867073e+00 -8.15061927e-01
1.19159743e-01 3.51393409e-02 5.88616021e-02 -9.87179279e-02
6.04818225e-01 8.10198784e-01 -1.57071364e+00 5.41389346e-01
5.77505469e-01 8.48941207e-01 6.06478989e-01 6.66392982e-01
3.17187726e-01 8.23182881e-01 -7.92608619e-01 -3.34469467e-01
3.06391418e-01 -6.38538301e-01 -5.15367389e-01 2.12263495e-01
5.43105304e-01 -5.15476763e-01 -6.37361050e-01 1.04864252e+00
5.25879502e-01 8.04437026e-02 6.86590075e-01 -8.17823112e-01
-1.38726771e+00 2.81506479e-01 -3.62339050e-01 5.67383587e-01
2.79046118e-01 8.51653695e-01 4.83952612e-01 -1.13730073e+00
5.59064746e-01 1.36044312e+00 3.42094272e-01 5.22177398e-01
-1.67245793e+00 -4.46835816e-01 1.10358402e-01 1.67339653e-01
-1.29571187e+00 -6.35794282e-01 4.80129004e-01 -5.00612795e-01
1.14083779e+00 3.85994613e-01 7.91766465e-01 1.29819334e+00
4.17367041e-01 5.48917294e-01 1.36198044e+00 -3.71976167e-01
1.63974732e-01 -1.62576050e-01 -2.71524251e-01 4.17779624e-01
3.59400004e-01 6.67466998e-01 -5.05318284e-01 1.10820748e-01
1.29278469e+00 1.35402441e-01 -5.15276313e-01 1.47354528e-01
-1.29008174e+00 4.46106493e-01 8.92815888e-01 2.17254192e-01
-6.73569858e-01 2.60611832e-01 -1.52822495e-01 1.55927092e-01
-6.07685447e-02 9.28430378e-01 4.13987264e-02 2.67858803e-01
-7.10301697e-01 1.38777435e-01 1.52770996e-01 4.39267576e-01
4.93193775e-01 5.38018525e-01 -4.59603399e-01 7.74388015e-01
2.29209870e-01 4.77846712e-01 4.92076576e-01 -9.79444325e-01
2.98492491e-01 3.97432745e-01 9.12209898e-02 -9.19806600e-01
-2.46243194e-01 -7.84159541e-01 -7.69254386e-01 8.32043469e-01
6.45210922e-01 1.30813539e-01 -1.22291160e+00 1.98856485e+00
-2.71328807e-01 -1.79011971e-01 -8.14149827e-02 1.31597590e+00
3.71656805e-01 3.61586392e-01 1.51610583e-01 3.73838320e-02
1.35744596e+00 -4.69245821e-01 -4.61722434e-01 -7.96157837e-01
-4.38602567e-02 -6.60196662e-01 1.15338182e+00 3.52485597e-01
-1.08232951e+00 -6.30970597e-01 -1.14217806e+00 -1.14744022e-01
-2.14779422e-01 1.03182364e-02 6.58965707e-01 2.93851793e-01
-1.02622867e+00 7.74135947e-01 -7.12433636e-01 -2.66835272e-01
9.65993524e-01 1.67107657e-01 -5.32546639e-01 -3.36793572e-01
-8.17807853e-01 1.22681999e+00 3.27286899e-01 3.35548788e-01
-1.44628978e+00 -4.45855707e-01 -8.27750564e-01 1.15333200e-01
2.74860375e-02 -9.78027344e-01 7.95224309e-01 -1.13328123e+00
-1.04605448e+00 7.81877816e-01 -2.45324418e-01 -4.69982624e-01
2.21158192e-01 -1.09551772e-01 -1.74677104e-01 2.39996478e-01
-6.59130812e-02 9.72290695e-01 1.23867774e+00 -1.34850430e+00
7.68060684e-02 -5.75630903e-01 -1.67053074e-01 2.75963485e-01
1.94148898e-01 -7.29268864e-02 -1.04817003e-02 -9.79468167e-01
3.10241461e-01 -6.17459059e-01 -1.58595651e-01 2.89413393e-01
-2.73834050e-01 1.65025309e-01 4.50375795e-01 -8.11763406e-01
5.50071359e-01 -2.22748709e+00 1.66735962e-01 -4.50626388e-02
1.77867800e-01 3.42572123e-01 -5.58053732e-01 4.42245185e-01
-4.18511480e-01 1.93246201e-01 -4.12595928e-01 -1.20406128e-01
-2.06902459e-01 5.84012158e-02 -7.97668815e-01 5.29080451e-01
6.63614571e-01 1.34440660e+00 -6.22188866e-01 3.34182709e-01
2.10869268e-01 6.24062777e-01 -4.82906431e-01 2.50183195e-02
-2.57818222e-01 6.56570256e-01 1.69968545e-01 5.96154988e-01
5.53946197e-01 -1.44497678e-01 -5.49501404e-02 -1.35633782e-01
-1.03781596e-02 4.10395086e-01 -7.35272884e-01 1.52133679e+00
-2.32036546e-01 8.15434754e-01 -1.22141436e-01 -8.01335514e-01
6.05969071e-01 9.58294347e-02 -4.17699426e-01 -9.94916916e-01
1.62340552e-01 9.83883590e-02 7.19514430e-01 -3.06620985e-01
4.73380014e-02 -2.28074700e-01 3.95334929e-01 5.46638370e-01
7.34127834e-02 -8.30582231e-02 -1.86685205e-01 2.06109479e-01
1.08264470e+00 2.10054398e-01 1.89776391e-01 -3.74623910e-02
-9.18123722e-02 -1.78948313e-01 2.07873940e-01 1.06416714e+00
3.65482233e-02 9.76685166e-01 4.94700253e-01 -2.71940827e-01
-1.11004162e+00 -1.23899722e+00 -1.68230742e-01 8.89814675e-01
-6.86442945e-03 1.74737766e-01 -4.97439474e-01 -2.16004372e-01
1.10319778e-01 1.16639495e+00 -8.88880908e-01 -7.83204854e-01
-3.84367526e-01 -6.40682638e-01 6.67819142e-01 6.81208014e-01
5.12726188e-01 -1.61377907e+00 -9.43607032e-01 9.38706398e-02
-3.66910244e-03 -7.52927601e-01 -3.20557684e-01 2.36527011e-01
-7.98573136e-01 -9.85929608e-01 -7.94325948e-01 -5.70157707e-01
1.11767089e+00 2.95443952e-01 7.86964595e-01 7.92770609e-02
-6.52208030e-01 1.49257630e-01 7.75406882e-02 -2.63541698e-01
-3.64055574e-01 -6.25140011e-01 -6.48944974e-02 -1.10802673e-01
-2.54048649e-02 -6.15267456e-01 -7.36467719e-01 3.42636049e-01
-1.09062672e+00 1.93322271e-01 9.48657513e-01 1.02585125e+00
3.59114498e-01 -2.05804020e-01 6.06961906e-01 -4.66060311e-01
7.73402035e-01 -2.48960167e-01 -4.23207372e-01 -6.72833920e-02
-2.66810656e-01 2.05660746e-01 4.31142747e-01 -6.51591480e-01
-1.06770384e+00 -2.26566270e-01 -6.94797486e-02 -4.98321772e-01
-3.48078668e-01 4.08386141e-01 -2.01803848e-01 -1.74955130e-01
1.14161658e+00 7.41112828e-01 1.15699247e-01 -2.58629739e-01
3.19764227e-01 1.24366451e-02 6.49455249e-01 -3.06680381e-01
8.32206428e-01 4.46538448e-01 -6.93786368e-02 -6.54172897e-01
-5.76655746e-01 2.39204168e-01 -4.01484102e-01 -6.03975952e-02
8.05292428e-01 -7.03333557e-01 -8.78653705e-01 6.34606063e-01
-1.23045897e+00 -5.10204613e-01 -3.73041540e-01 5.31552255e-01
-4.76522535e-01 -2.92004589e-02 -6.83474422e-01 -7.59973168e-01
1.23004988e-02 -1.06825423e+00 7.34459937e-01 5.92031106e-02
-2.67803580e-01 -4.82204825e-01 -3.51078451e-01 1.72495097e-01
5.74461162e-01 1.00376628e-01 1.08553672e+00 -3.79269361e-01
-1.01627564e+00 -8.14925879e-02 -4.22286808e-01 3.56802702e-01
-1.10447872e-02 -4.59364802e-01 -1.29076695e+00 -1.72928333e-01
2.36093789e-01 -5.13284028e-01 1.36236477e+00 4.25192922e-01
1.27732027e+00 -4.89135116e-01 -1.86139882e-01 6.16916299e-01
1.22571158e+00 3.56252849e-01 1.15642476e+00 3.15807462e-02
4.31291491e-01 4.64976609e-01 1.11970820e-01 1.04539748e-02
-1.45861268e-01 4.99029815e-01 5.67699134e-01 -1.76973805e-01
-5.44433594e-01 -4.32344377e-01 2.78165638e-01 -9.07924026e-02
-1.29968807e-01 -1.10923231e-01 -8.30891609e-01 5.16758800e-01
-1.40315866e+00 -1.18760443e+00 2.34549567e-01 2.37836123e+00
8.57713521e-01 4.08921331e-01 -3.52691680e-01 6.98863864e-02
5.62787950e-01 6.04823604e-02 -8.34390163e-01 -1.63391158e-01
-3.65944475e-01 2.96541512e-01 2.94207692e-01 3.72505248e-01
-6.62205398e-01 9.00110602e-01 6.79154825e+00 3.70662510e-01
-1.27712476e+00 -1.32958367e-01 6.08898759e-01 -1.32636040e-01
-2.85505474e-01 -1.03306964e-01 -3.93832505e-01 2.10987329e-01
6.64422393e-01 1.16328418e-01 9.42889810e-01 4.26888943e-01
1.31249622e-01 -4.07526195e-01 -1.21639657e+00 8.45064938e-01
2.53942877e-01 -1.36181128e+00 2.08318383e-01 2.57831484e-01
3.90775532e-01 -3.74223217e-02 5.77351034e-01 1.27398774e-01
3.69012058e-01 -1.53894496e+00 9.73171592e-01 7.14625359e-01
1.03095078e+00 -3.36064279e-01 3.51942956e-01 5.24291337e-01
-4.93016720e-01 -1.01382598e-01 -3.83222699e-01 -4.13378537e-01
8.16195458e-02 3.25586796e-01 -8.80519986e-01 -9.39265937e-02
3.65707695e-01 3.23023170e-01 -6.88751519e-01 1.02239561e+00
-5.59823513e-01 4.51326132e-01 -9.44740102e-02 2.70020723e-01
-1.63278490e-01 1.91210121e-01 6.96526527e-01 8.41937661e-01
1.46916121e-01 2.81473011e-01 -4.46205378e-01 1.55522943e+00
-1.30290672e-01 -5.13893962e-01 -6.93137348e-01 -1.14286639e-01
4.63689744e-01 9.70710397e-01 -6.79341495e-01 -1.47715509e-01
-5.89300878e-02 1.27545130e+00 5.58210850e-01 7.77149677e-01
-6.17913902e-01 -2.74078369e-01 4.04043257e-01 2.16906831e-01
4.41556543e-01 -3.36487442e-01 -6.28698230e-01 -9.14258361e-01
4.84088846e-02 -9.43970382e-01 -1.64507464e-01 -1.26006246e+00
-1.09097159e+00 5.86907983e-01 -2.79599279e-01 -7.24311590e-01
-8.35057274e-02 -7.09460974e-01 -4.92251009e-01 1.32799959e+00
-1.16895199e+00 -8.69249821e-01 -2.00490206e-01 5.27143598e-01
4.86347497e-01 -2.46539600e-02 8.28749478e-01 -2.55418178e-02
-3.18014055e-01 3.96560252e-01 -4.78168190e-01 2.20040709e-01
6.03142500e-01 -9.31676686e-01 6.12754226e-01 1.10499096e+00
4.08077151e-01 1.18687177e+00 5.64895749e-01 -7.89131761e-01
-1.29361224e+00 -1.04382181e+00 4.26622093e-01 -6.43179417e-01
1.40486568e-01 -5.81675649e-01 -1.08371902e+00 7.86995411e-01
1.38498694e-01 1.74896330e-01 4.00453568e-01 -1.55739725e-01
-7.24108815e-01 6.18879087e-02 -1.15832114e+00 7.85718322e-01
1.19458747e+00 -5.84785581e-01 -8.12548280e-01 1.07017875e-01
4.43968594e-01 -2.24406555e-01 -2.91293472e-01 1.17508017e-01
4.74663258e-01 -9.97347355e-01 1.02253938e+00 -6.91425741e-01
5.43025196e-01 -4.17348772e-01 -1.57925218e-01 -1.57299876e+00
-8.06438923e-01 -1.39467284e-01 4.98736575e-02 9.09329772e-01
5.01901150e-01 -7.82113731e-01 8.04209486e-02 5.69414258e-01
-2.75156170e-01 -4.05583292e-01 -7.73219168e-01 -6.37821078e-01
-3.62358131e-02 -4.20571029e-01 2.27458000e-01 6.54124856e-01
-3.59938025e-01 1.97559148e-01 -1.43324658e-01 1.32410407e-01
5.84716439e-01 -1.01942047e-01 3.79940093e-01 -8.03905487e-01
-4.38814938e-01 -5.42700410e-01 -3.43392134e-01 -9.02257919e-01
-5.34202829e-02 -1.06868672e+00 4.36236598e-02 -1.61325967e+00
1.34446919e-01 -1.70396775e-01 -1.62265509e-01 7.37765670e-01
-1.18878677e-01 3.99419904e-01 4.74209875e-01 4.38487828e-01
-2.26062443e-02 5.03324270e-01 1.42311788e+00 1.02261361e-02
5.11709638e-02 -1.29747510e-01 -1.21503615e+00 6.13100946e-01
5.64897478e-01 -4.03360546e-01 -2.48860642e-01 -6.12497151e-01
9.03978422e-02 7.24513689e-03 1.11092436e+00 -9.38985646e-01
5.15995137e-02 -9.68803018e-02 1.17964625e+00 -1.31113693e-01
4.13917303e-01 -6.56619668e-01 1.58599332e-01 6.88801885e-01
-4.77447420e-01 -1.16518162e-01 5.79071701e-01 5.61478913e-01
6.94383755e-02 -4.01460230e-02 9.78602588e-01 -3.31890047e-01
-6.76395595e-01 -6.94315583e-02 -6.11320853e-01 -1.44261364e-02
7.92785645e-01 -3.25207531e-01 -7.43014455e-01 -3.37486535e-01
-8.94678116e-01 -3.83594841e-01 4.50451940e-01 4.38908726e-01
1.00520706e+00 -1.32697845e+00 -8.56053710e-01 5.76552272e-01
-5.07497415e-02 -3.94463122e-01 2.21788019e-01 7.48897076e-01
-2.87067384e-01 3.41035366e-01 -4.53895748e-01 -5.75257242e-01
-6.77008033e-01 7.76695371e-01 5.18551946e-01 4.20032978e-01
-6.89517140e-01 1.05052722e+00 5.73505402e-01 6.86344951e-02
8.07348043e-02 -1.38752505e-01 1.55025959e-01 -2.87745327e-01
7.18552589e-01 -6.25256896e-02 -7.17652142e-02 -3.59963238e-01
-4.56321329e-01 2.43091136e-01 -6.94978237e-02 -3.47681671e-01
1.03278255e+00 1.47156313e-01 -6.02168553e-02 2.95913994e-01
6.65282905e-01 -1.45659924e-01 -1.60436416e+00 -1.05667092e-01
-3.57728034e-01 -6.73417866e-01 -2.13598404e-02 -1.37305260e+00
-8.31899285e-01 1.06641519e+00 5.09429693e-01 -2.07236737e-01
1.07976556e+00 1.27436668e-01 1.72669664e-01 2.07941741e-01
3.63030106e-01 -6.02938175e-01 3.91169012e-01 3.33654046e-01
1.59974074e+00 -1.05395389e+00 -1.53307408e-01 -4.47645858e-02
-7.94602811e-01 7.39379585e-01 8.05308700e-01 -2.91206658e-01
1.32526129e-01 1.89683214e-01 -2.04439014e-01 -1.25076801e-01
-9.11829770e-01 -2.79502928e-01 5.17512619e-01 9.11811352e-01
2.55164564e-01 -1.64117798e-01 1.77621052e-01 5.45629144e-01
-2.22936913e-01 -1.27904668e-01 2.98360676e-01 6.51197970e-01
-5.81053972e-01 -5.44898629e-01 -6.25129402e-01 4.78109598e-01
-2.31132776e-01 -3.74389976e-01 -3.11894596e-01 6.10710800e-01
1.31003320e-01 6.71234429e-01 2.73865879e-01 -1.60659209e-01
2.28324577e-01 1.69182256e-01 9.22590077e-01 -7.73890138e-01
-4.13827807e-01 9.84648839e-02 -1.79953843e-01 -4.32057559e-01
1.13631129e-01 -7.23184466e-01 -1.03952658e+00 9.57268476e-02
-8.89436826e-02 -6.18453503e-01 5.32261670e-01 8.53963077e-01
4.88992274e-01 8.63884270e-01 9.76886973e-02 -9.16935742e-01
-5.54145694e-01 -1.12665582e+00 -4.09018278e-01 5.37852108e-01
3.92383039e-01 -7.12733507e-01 -4.07278359e-01 9.10593569e-02] | [10.124659538269043, 2.3653786182403564] |
5293e48f-7790-423d-878c-cc5e89358761 | educational-multi-question-generation-for | null | null | https://aclanthology.org/2022.bea-1.26 | https://aclanthology.org/2022.bea-1.26.pdf | Educational Multi-Question Generation for Reading Comprehension | Automated question generation has made great advances with the help of large NLP generation models. However, typically only one question is generated for each intended answer. We propose a new task, Multi-Question Generation, aimed at generating multiple semantically similar but lexically diverse questions assessing the same concept. We develop an evaluation framework based on desirable qualities of the resulting questions. Results comparing multiple question generation approaches in the two-question generation condition show a trade-off between question answerability and lexical diversity between the two questions. We also report preliminary results from sampling multiple questions from our model, to explore generating more than two questions. Our task can be used to further explore the educational impact of showing multiple distinct question wordings to students. | ['Katherine Stasaski', 'Tony Tu', 'Manav Rathod'] | null | null | null | null | naacl-bea-2022-7 | ['question-generation'] | ['natural-language-processing'] | [ 2.94442862e-01 7.54607022e-01 4.49259311e-01 -2.19180137e-01
-1.53530908e+00 -9.92326736e-01 6.62635088e-01 5.43625712e-01
-2.06124142e-01 1.01667142e+00 6.17712796e-01 -5.86149216e-01
-2.88291067e-01 -9.44596887e-01 -3.48070830e-01 1.37935475e-01
5.45337677e-01 6.88369215e-01 4.68453020e-01 -6.35636568e-01
5.16380608e-01 -2.56770737e-02 -1.94950247e+00 6.70464456e-01
1.60289884e+00 3.30449164e-01 4.11748201e-01 9.87539172e-01
-9.82695282e-01 9.86117244e-01 -1.30024040e+00 -6.78997040e-01
-1.78904459e-01 -1.14797771e+00 -1.36723423e+00 -5.90965785e-02
1.05382502e+00 -1.27999052e-01 3.75815749e-01 7.41104424e-01
5.91538966e-01 4.40641195e-01 5.46238363e-01 -1.04266822e+00
-1.12068987e+00 6.96883619e-01 3.40465307e-01 2.43608877e-01
1.36059487e+00 -7.97221735e-02 1.37195587e+00 -7.65269458e-01
6.94261193e-01 1.22634292e+00 4.01528776e-01 7.40615606e-01
-1.10828578e+00 -5.28860033e-01 -2.21070200e-01 3.00497293e-01
-1.03227901e+00 -2.27063462e-01 4.56392914e-01 -3.06619376e-01
8.33284736e-01 6.60416543e-01 6.50145829e-01 7.90851414e-01
2.33312413e-01 5.19236863e-01 1.32358670e+00 -7.19833791e-01
2.01178819e-01 4.08891350e-01 1.09462433e-01 5.51310599e-01
4.31792110e-01 -2.65455276e-01 -3.87198210e-01 -3.97387028e-01
3.26428860e-01 -8.02044034e-01 -2.43937045e-01 1.80844560e-01
-7.79810488e-01 1.19282687e+00 -1.82058662e-01 4.87961233e-01
-6.48862198e-02 -1.42492339e-01 -9.58447978e-02 5.96628785e-01
1.67065814e-01 1.53863657e+00 -4.89437878e-01 -3.49239856e-01
-9.12948310e-01 1.13513494e+00 1.44910908e+00 1.13911438e+00
5.50518036e-01 -2.81646520e-01 -7.72059381e-01 8.56833518e-01
1.94135621e-01 3.40425402e-01 8.69852602e-01 -1.34144986e+00
4.26883101e-01 7.17118442e-01 3.08056384e-01 -9.92356062e-01
-1.40891403e-01 -1.52030125e-01 2.27879565e-02 -1.68299675e-01
5.95507383e-01 -4.95082259e-01 -5.60314894e-01 1.68340731e+00
3.10432315e-01 -4.33084935e-01 1.90774143e-01 2.80282497e-01
1.56263363e+00 6.49599910e-01 2.95290470e-01 2.29055379e-02
1.80940378e+00 -8.78359258e-01 -8.87136936e-01 -1.15510561e-01
8.05077374e-01 -1.16040564e+00 1.45943522e+00 1.37764052e-01
-1.48158860e+00 -6.82401896e-01 -6.11579239e-01 -2.41153613e-01
-2.98175603e-01 -1.20887622e-01 8.91812518e-02 1.16689754e+00
-1.14953113e+00 1.15819074e-01 2.29257569e-01 -4.72978145e-01
-2.10181419e-02 -9.05069411e-02 1.83461607e-01 -2.27892429e-01
-1.33911002e+00 9.93530512e-01 4.98073511e-02 -8.32327485e-01
-4.18291777e-01 -1.02063477e+00 -9.33054090e-01 2.26089001e-01
1.27872005e-01 -9.76600587e-01 1.71312642e+00 -6.45574689e-01
-1.42026055e+00 7.18348980e-01 -2.23587722e-01 9.45234224e-02
5.88505529e-02 -1.07413478e-01 -2.82743245e-01 3.35896194e-01
6.81995988e-01 9.74435747e-01 3.37307125e-01 -1.23062015e+00
-6.19903326e-01 1.98756903e-01 4.90862608e-01 5.46070218e-01
-1.99994132e-01 2.27123201e-01 3.68003398e-01 -7.61500597e-01
-9.43757519e-02 -5.58473229e-01 -2.19244614e-01 -4.38485056e-01
-1.67915151e-01 -7.48346150e-01 2.18837157e-01 -7.40143538e-01
1.19100010e+00 -1.25386453e+00 -4.41585720e-01 -1.81062415e-01
9.99999195e-02 -2.03609141e-03 -6.42803371e-01 8.11688423e-01
-1.19060636e-01 4.96517211e-01 6.16770387e-02 1.93785116e-01
2.28315368e-01 -1.80305794e-01 -4.32105064e-01 -6.31937623e-01
2.95415491e-01 9.64295089e-01 -1.23564255e+00 -7.03984678e-01
-4.23905581e-01 -2.37530366e-01 -6.85663462e-01 6.50371492e-01
-5.99373043e-01 8.46429169e-02 -5.45709133e-01 2.81518936e-01
1.76858723e-01 -1.48188815e-01 -2.08738312e-01 3.79076511e-01
1.56748593e-01 9.02176440e-01 -1.06903410e+00 1.52014351e+00
-7.26142228e-01 4.97715324e-01 -3.92086267e-01 -2.27601916e-01
1.07169843e+00 6.07540667e-01 -8.33519589e-05 -7.29897738e-01
-1.42294988e-01 2.64746875e-01 1.37116671e-01 -9.99393225e-01
9.71345663e-01 -6.01842821e-01 -2.28306085e-01 8.96914542e-01
2.89698899e-01 -1.06993973e+00 8.71810138e-01 4.52450991e-01
1.14376485e+00 1.94605112e-01 3.75725627e-01 -5.42926192e-01
5.16134024e-01 2.59196997e-01 -2.63570338e-01 1.13206220e+00
-4.09669913e-02 6.88469827e-01 3.80482733e-01 2.38807574e-01
-5.76157033e-01 -1.11727273e+00 2.79130936e-02 1.21685076e+00
-1.71910435e-01 -5.89836121e-01 -8.06497574e-01 -8.06349218e-01
-2.16464087e-01 1.47727239e+00 -4.43096399e-01 -2.14826316e-02
-3.71131599e-01 -2.15624303e-01 7.71441936e-01 2.28938982e-01
3.65739167e-02 -1.23885703e+00 -6.47882700e-01 2.96075344e-01
-7.85593390e-01 -9.27127898e-01 -5.12297034e-01 -3.30119073e-01
-6.24314785e-01 -1.00616324e+00 -7.50716448e-01 -1.09594119e+00
4.88115817e-01 1.87683150e-01 1.78714931e+00 2.81269491e-01
-7.72928298e-02 9.80831623e-01 -6.15459681e-01 -6.21522129e-01
-8.51672828e-01 2.29780316e-01 -6.34829879e-01 -7.64854670e-01
3.74213815e-01 -3.93187165e-01 -4.30753618e-01 -1.82117894e-02
-1.33272219e+00 -1.63955148e-02 2.86150783e-01 4.67550606e-01
1.19776838e-01 -2.43520990e-01 1.28625929e+00 -8.68059218e-01
1.75589395e+00 -7.48582184e-01 -1.93887547e-01 5.99372685e-01
-3.36211950e-01 3.24239492e-01 5.33301532e-01 -4.11832124e-01
-1.21845973e+00 -5.85950255e-01 -6.03362381e-01 5.21029055e-01
-4.49204236e-01 5.65998316e-01 -1.43030658e-01 -6.65784925e-02
9.94141996e-01 1.57703459e-01 -1.30952328e-01 -1.73301086e-01
8.04516613e-01 3.30321699e-01 3.23480205e-03 -9.67777431e-01
7.58238614e-01 -3.70200664e-01 -3.16481292e-01 -8.55666459e-01
-9.84884739e-01 -3.54773104e-01 2.71657575e-02 -3.97312284e-01
6.79386139e-01 -5.88149905e-01 -4.35856491e-01 -9.79639292e-02
-1.31902885e+00 -1.13454461e-01 -9.65239227e-01 3.75547223e-02
-5.32554865e-01 3.22527647e-01 -4.77220207e-01 -5.36792397e-01
-2.65096217e-01 -8.41664076e-01 7.89596498e-01 6.32219434e-01
-8.92615676e-01 -1.20509243e+00 5.86723983e-01 9.50969636e-01
4.80643868e-01 -4.37669270e-02 1.41721737e+00 -1.30651987e+00
-4.17031914e-01 -3.97276394e-02 9.22436267e-02 1.48050368e-01
3.55912209e-01 -1.95639327e-01 -5.49343050e-01 2.23766148e-01
3.70789319e-01 -8.27131927e-01 3.09386373e-01 -1.15862407e-01
7.55427301e-01 -6.74714267e-01 2.60971814e-01 -4.15487140e-01
1.30475605e+00 -6.69706659e-03 6.01277649e-01 1.85977891e-01
3.12593013e-01 1.29130542e+00 5.27092159e-01 1.69241633e-02
8.62651587e-01 2.89304465e-01 -2.89383139e-02 5.46317816e-01
-4.87196714e-01 -4.27947551e-01 2.41934210e-01 1.04132175e+00
6.01224005e-01 -4.69485670e-01 -9.10594046e-01 1.02716005e+00
-9.49852645e-01 -1.05714369e+00 -4.89579618e-01 1.91677558e+00
1.27394760e+00 -2.53885686e-01 4.56863083e-02 -2.12341592e-01
2.99652219e-01 -1.54443774e-02 1.37398496e-01 -7.68134832e-01
2.17511575e-03 1.00293517e+00 -3.18979353e-01 8.61978173e-01
-1.12965047e-01 6.97194993e-01 7.24058104e+00 8.81490171e-01
-2.08674312e-01 -3.75450077e-03 5.11053979e-01 9.91853252e-02
-1.41443169e+00 1.79020405e-01 -9.41826880e-01 2.67183930e-01
1.13968360e+00 -7.57103026e-01 -2.08871365e-01 3.76711875e-01
2.13586092e-02 -2.77153075e-01 -8.52132022e-01 2.13753909e-01
5.46830773e-01 -1.12132847e+00 8.06026995e-01 -2.80215561e-01
1.05283749e+00 -7.85606742e-01 -2.05197617e-01 5.09922206e-01
7.10486710e-01 -1.12137914e+00 4.58168447e-01 4.05031085e-01
4.73619878e-01 -7.69769728e-01 5.05830705e-01 3.55353922e-01
-1.00143635e+00 6.92944005e-02 -2.10046753e-01 -2.33392701e-01
8.04763809e-02 2.29735538e-01 -1.09217513e+00 4.39685553e-01
2.65812099e-01 -4.04439688e-01 -1.15021110e+00 1.11754513e+00
-5.15459657e-01 7.12770343e-01 -6.59211129e-02 -7.26843596e-01
1.47076905e-01 -1.24416485e-01 3.84976923e-01 9.34198916e-01
7.22477794e-01 4.82991934e-01 7.92031288e-02 9.85736847e-01
-5.28945513e-02 7.27854311e-01 -4.77072895e-01 8.79644901e-02
6.24228835e-01 1.22117507e+00 -4.51965660e-01 -4.44371879e-01
-3.65037322e-01 6.24831140e-01 2.62109965e-01 2.54095227e-01
-3.73299658e-01 -7.21024513e-01 3.39844763e-01 3.35424066e-01
-3.07844095e-02 6.37335479e-02 -3.95910472e-01 -8.53511333e-01
-2.37882864e-02 -1.37930906e+00 3.15956175e-01 -1.09625018e+00
-1.38811767e+00 3.94461840e-01 1.59284502e-01 -7.44888306e-01
-6.73155785e-01 -2.47629255e-01 -8.95762026e-01 1.23892677e+00
-1.27737343e+00 -6.40124619e-01 -3.51583481e-01 2.58892775e-01
8.45156550e-01 1.16731457e-01 1.06561947e+00 -1.55006931e-03
4.58603464e-02 6.35607183e-01 -3.63786727e-01 -3.32725257e-01
8.76915157e-01 -1.67120910e+00 4.15958345e-01 6.67219102e-01
1.75041035e-01 5.53059101e-01 8.35496604e-01 -8.02348018e-01
-7.24450111e-01 -8.83323491e-01 1.68575394e+00 -7.97035277e-01
5.70297539e-01 9.72815007e-02 -1.06676829e+00 2.55704880e-01
9.25080836e-01 -9.10444498e-01 1.48259759e+00 -6.43600523e-02
-9.18603837e-02 4.92589712e-01 -1.25649786e+00 8.12403500e-01
5.94955385e-01 -3.43901038e-01 -1.38130796e+00 4.40356016e-01
1.04881632e+00 -1.67456210e-01 -9.39439893e-01 8.63405839e-02
1.44675463e-01 -9.21531498e-01 7.19726503e-01 -8.28640759e-01
8.93525958e-01 -1.85218938e-02 -3.67465522e-03 -1.37405527e+00
-1.68796927e-01 -6.17815554e-01 5.77283977e-03 1.58842814e+00
6.38222158e-01 -4.26812023e-01 8.15901160e-01 8.42379630e-01
-1.43764392e-01 -6.51953280e-01 -8.48125160e-01 -6.62166715e-01
7.30162740e-01 -1.68927565e-01 7.92607903e-01 8.28266680e-01
1.40068680e-01 8.61221850e-01 3.99176031e-01 -2.41623133e-01
2.58279741e-01 2.00915843e-01 5.60286343e-01 -1.12654841e+00
-3.10918778e-01 -5.98311961e-01 6.98554441e-02 -1.18035936e+00
-5.99930389e-03 -8.50054145e-01 3.53012264e-01 -1.99147463e+00
5.30988947e-02 -4.87218536e-02 3.68703604e-01 5.58120152e-03
-1.03974676e+00 6.46097809e-02 2.62225330e-01 -4.89127368e-01
-5.08284867e-01 6.07251942e-01 1.69356716e+00 2.85098940e-01
-5.81865804e-03 -3.49664837e-02 -1.32818139e+00 4.67423588e-01
8.76625657e-01 -5.50708950e-01 -8.13069761e-01 -3.50067914e-01
5.58523059e-01 1.75501928e-01 7.62675405e-02 -1.01207542e+00
1.06019273e-01 -3.36265981e-01 1.93757281e-01 -3.61712754e-01
-8.20421278e-02 -3.20053071e-01 -3.91412407e-01 2.85588950e-01
-9.53487515e-01 5.39497435e-01 4.07502174e-01 6.46976456e-02
-3.30226779e-01 -1.13643777e+00 3.72639865e-01 -4.59520012e-01
-1.69837788e-01 -2.82429606e-01 -9.00568068e-01 1.04231787e+00
7.85545349e-01 -2.48795182e-01 -2.99676865e-01 -8.61908019e-01
-4.67899084e-01 3.05487573e-01 9.04187635e-02 6.60626233e-01
7.69982159e-01 -1.43951297e+00 -1.23218024e+00 -2.55504906e-01
3.09627146e-01 -2.28071615e-01 1.88075632e-01 -1.41176984e-01
-5.70373237e-01 6.43421292e-01 -6.71665817e-02 1.31240338e-01
-1.33930373e+00 9.38245282e-02 3.10731709e-01 -7.03778863e-01
1.43564165e-01 1.05877030e+00 -5.76129071e-02 -8.41325462e-01
-2.43496910e-01 -3.25541317e-01 -7.55971789e-01 4.16414112e-01
7.17274547e-01 4.60279167e-01 1.19297110e-01 -1.61143810e-01
3.76780599e-01 1.91487595e-01 2.13874895e-02 -5.73703110e-01
5.81480861e-01 -2.26913512e-01 -1.50538668e-01 2.77125537e-01
9.35116053e-01 4.13617045e-01 -3.74014348e-01 7.57234991e-02
2.10453331e-01 -3.36294413e-01 -6.67294323e-01 -1.08437204e+00
-2.65671581e-01 3.72529656e-01 2.93442547e-01 7.89369643e-01
9.25626218e-01 1.10085085e-01 8.54333103e-01 4.89327043e-01
-3.15763652e-02 -7.14928031e-01 6.05324805e-01 6.91302717e-01
1.16040564e+00 -7.94624686e-01 -2.35676780e-01 -4.47602987e-01
-3.87735993e-01 1.07535553e+00 1.07772470e+00 1.74470156e-01
9.70743746e-02 -2.84863651e-01 2.77758539e-01 -2.92216063e-01
-9.08436537e-01 -2.46131167e-01 5.35054266e-01 6.93851471e-01
9.79284763e-01 -2.94474214e-01 -7.77806520e-01 7.52725661e-01
-1.07890940e+00 -2.49949008e-01 1.12029421e+00 1.01176572e+00
-8.07519138e-01 -1.53147674e+00 -5.29146314e-01 7.54158795e-01
-5.49910724e-01 -5.96722543e-01 -8.76392424e-01 6.05489194e-01
-1.11875638e-01 1.57210338e+00 -2.28482425e-01 3.60788591e-02
6.13321126e-01 4.79695916e-01 6.21659398e-01 -1.39051890e+00
-1.19060695e+00 -8.09360027e-01 4.89934027e-01 -1.64638199e-02
-4.19067368e-02 -5.99474967e-01 -8.90985191e-01 -1.24230430e-01
-5.27869642e-01 8.56489658e-01 3.76665950e-01 1.02761793e+00
3.73053491e-01 5.57006299e-01 3.49290520e-01 1.08239092e-01
-8.72795343e-01 -1.16137314e+00 -1.70640036e-01 4.56600964e-01
9.54155102e-02 -7.69541785e-02 -2.51815557e-01 -1.37051241e-02] | [11.520038604736328, 8.101999282836914] |
69cb75ca-bc59-4707-b74c-c28c841319d1 | fine-grained-action-detection-with-rgb-and | 2302.02755 | null | https://arxiv.org/abs/2302.02755v1 | https://arxiv.org/pdf/2302.02755v1.pdf | Fine-Grained Action Detection with RGB and Pose Information using Two Stream Convolutional Networks | As participants of the MediaEval 2022 Sport Task, we propose a two-stream network approach for the classification and detection of table tennis strokes. Each stream is a succession of 3D Convolutional Neural Network (CNN) blocks using attention mechanisms. Each stream processes different 4D inputs. Our method utilizes raw RGB data and pose information computed from MMPose toolbox. The pose information is treated as an image by applying the pose either on a black background or on the original RGB frame it has been computed from. Best performance is obtained by feeding raw RGB data to one stream, Pose + RGB (PRGB) information to the other stream and applying late fusion on the features. The approaches were evaluated on the provided TTStroke-21 data sets. We can report an improvement in stroke classification, reaching 87.3% of accuracy, while the detection does not outperform the baseline but still reaches an IoU of 0.349 and mAP of 0.110. | ['Pierre-Etienne Martin', 'Finn Bartels', 'Leonard Hacker'] | 2023-02-06 | null | null | null | null | ['fine-grained-action-detection', 'action-classification', 'stroke-classification'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 2.32320458e-01 -1.90999657e-01 -1.04649328e-01 -8.31913948e-03
-6.04274869e-01 -6.16117597e-01 8.04157674e-01 -1.35164589e-01
-1.05957818e+00 3.37957710e-01 1.54084608e-01 -5.50823398e-02
3.42036545e-01 -7.69480109e-01 -8.92347217e-01 -5.24241447e-01
1.08885460e-01 4.14840192e-01 6.70780778e-01 -1.61465377e-01
3.55634034e-01 9.62449670e-01 -1.83455348e+00 6.52954817e-01
7.98055083e-02 1.34143412e+00 -3.35645564e-02 1.30185986e+00
-9.19245183e-02 1.01993477e+00 -9.89907622e-01 -3.53777081e-01
6.56562567e-01 -2.78401136e-01 -4.73338187e-01 -4.66755509e-01
1.00413108e+00 -7.17784226e-01 -9.62214589e-01 5.60078204e-01
8.88157427e-01 2.37372488e-01 5.32563567e-01 -1.21005929e+00
-3.68212424e-02 2.81644076e-01 -5.81282675e-01 5.16548872e-01
2.52370685e-01 3.76649052e-01 4.71586257e-01 -1.10293758e+00
8.18517625e-01 1.21529138e+00 6.13505363e-01 6.23150527e-01
-8.68769765e-01 -5.44955254e-01 -5.84164634e-02 4.32180583e-01
-9.65417206e-01 -1.27521560e-01 2.79317319e-01 -2.68213362e-01
9.96576667e-01 4.17697251e-01 1.00513339e+00 1.28129721e+00
4.48604859e-02 1.30876362e+00 1.09899616e+00 -1.73081502e-01
2.00722739e-01 -6.14491463e-01 2.30343968e-01 4.58745807e-01
6.43308833e-02 2.32504621e-01 -1.34640241e+00 3.72973949e-01
1.00748229e+00 -2.62100785e-03 -1.84948102e-01 -2.67325521e-01
-1.23961854e+00 3.43718857e-01 8.02887201e-01 -3.22742045e-01
-6.23634636e-01 4.69004512e-01 5.34479737e-01 1.23664089e-01
2.55301856e-02 -9.69142690e-02 -3.87882739e-01 -7.21513093e-01
-1.00279558e+00 5.45189977e-01 6.32470131e-01 9.18921173e-01
6.16306514e-02 -2.78017037e-02 -6.69500768e-01 1.97588652e-01
-3.32336932e-01 8.18714082e-01 1.20714098e-01 -9.74188745e-01
7.35870540e-01 6.00336909e-01 1.99445397e-01 -6.83275819e-01
-5.73495150e-01 -3.17513794e-01 -5.63685000e-01 1.01013887e+00
1.14731717e+00 -3.18745524e-01 -1.36277533e+00 1.13705432e+00
7.28729367e-02 1.91790164e-01 -5.68280555e-02 1.53919697e+00
1.15252841e+00 4.33335602e-01 2.24867407e-02 5.53934693e-01
1.29224527e+00 -1.01950824e+00 -4.72726554e-01 -1.92435950e-01
1.90973163e-01 -6.57667875e-01 1.04623961e+00 7.69992054e-01
-1.15098512e+00 -8.03203464e-01 -1.27933657e+00 -3.82956594e-01
-5.48444271e-01 6.53415382e-01 3.47083569e-01 4.96400774e-01
-7.55938947e-01 8.58731151e-01 -1.03680038e+00 -2.78314501e-01
8.04288030e-01 1.51162162e-01 -3.02973002e-01 1.47956178e-01
-6.80131257e-01 9.90104258e-01 8.32397193e-02 4.76759791e-01
-9.41139221e-01 -6.91355586e-01 -4.60443914e-01 -1.18930824e-01
1.67203024e-01 -3.76107097e-01 1.25500107e+00 -8.01774144e-01
-1.65369725e+00 8.02915514e-01 1.46472603e-01 -5.88647604e-01
1.40443158e+00 -9.89205956e-01 1.24470100e-01 4.35274422e-01
-7.29080364e-02 7.11665154e-01 5.99368989e-01 -8.92881513e-01
-9.56685483e-01 -4.53118145e-01 4.68509607e-02 4.28270251e-01
-4.60951589e-02 2.27977941e-03 -8.56885910e-01 -5.83512545e-01
1.63265914e-01 -9.36562479e-01 9.18286294e-02 5.42682886e-01
-4.57847804e-01 9.01497900e-02 8.63621950e-01 -9.51357901e-01
7.29483843e-01 -1.79183424e+00 3.70085061e-01 1.27997294e-01
7.07753971e-02 4.83192980e-01 7.78241679e-02 6.90655634e-02
1.08761087e-01 -3.46521795e-01 5.01700230e-02 -3.71255159e-01
-2.06941545e-01 7.22100362e-02 -2.92871773e-01 5.47636032e-01
2.30553791e-01 1.01204026e+00 -7.77802289e-01 -2.21513972e-01
5.03597379e-01 6.89266086e-01 -1.28493533e-01 2.01671764e-01
1.03013322e-01 1.55730665e-01 -2.05696300e-01 5.97066522e-01
5.76284587e-01 4.17787373e-01 -2.69382328e-01 -3.09374630e-01
-3.16263676e-01 8.38140994e-02 -1.27398324e+00 2.00715804e+00
2.93053109e-02 1.17266738e+00 -7.21376687e-02 -5.44380426e-01
7.18786478e-01 1.73552167e-02 3.12233478e-01 -7.40579724e-01
4.97628480e-01 -2.08724067e-01 -3.15709412e-02 -3.43090951e-01
4.61145699e-01 3.67809802e-01 -1.16441645e-01 3.03483903e-01
1.50691167e-01 2.18643308e-01 1.85990483e-01 2.04005048e-01
1.42370367e+00 7.73647487e-01 -5.15395701e-01 2.98665464e-02
1.55470192e-01 2.91268915e-01 2.66280353e-01 1.12382925e+00
-3.62840235e-01 1.13457274e+00 8.04741323e-01 -6.92519963e-01
-1.11367166e+00 -1.23912776e+00 2.62659132e-01 1.11613917e+00
2.71582186e-01 -4.23020422e-01 -8.37056577e-01 -6.20116711e-01
1.80344552e-01 3.63766462e-01 -1.04400694e+00 -1.90821350e-01
-9.57890928e-01 -4.24743682e-01 8.31411541e-01 1.26915193e+00
8.68847549e-01 -1.33927047e+00 -1.56859875e+00 -6.66210502e-02
5.62935695e-02 -1.05967128e+00 -7.89707229e-02 3.87723237e-01
-6.76199198e-01 -1.28027797e+00 -9.66259956e-01 -3.26103061e-01
1.49967581e-01 5.86736808e-03 9.23860073e-01 -9.09741968e-02
-6.06938362e-01 1.07699342e-01 -4.43443447e-01 -6.61353052e-01
1.05021074e-01 3.16754907e-01 -1.23956241e-01 -3.89335975e-02
3.52315992e-01 -7.31702298e-02 -8.02407384e-01 -1.17808124e-02
-3.64453733e-01 2.34220251e-01 6.41511977e-01 5.37273109e-01
4.67537314e-01 -8.09755921e-01 1.42693212e-02 -3.03126782e-01
3.59054029e-01 3.20464641e-01 -4.29133654e-01 -6.34050593e-02
-4.57144575e-04 -1.17788747e-01 2.30584100e-01 -5.15739918e-01
-8.17682743e-01 5.59910357e-01 7.00022131e-02 -7.15449691e-01
-2.61651397e-01 -1.01872593e-01 2.42170356e-02 1.17331408e-01
9.39486504e-01 -6.15296792e-03 1.05714895e-01 -6.23116195e-01
5.11338174e-01 6.24610245e-01 1.31551504e+00 -3.91386300e-01
5.79312503e-01 5.33999562e-01 3.46060209e-02 -6.56159341e-01
-7.08590508e-01 -2.82000542e-01 -1.09203362e+00 -8.38543594e-01
1.14713275e+00 -6.92106426e-01 -1.02862692e+00 1.02744162e+00
-1.13683414e+00 -6.58876419e-01 -3.07760507e-01 3.67725253e-01
-5.06413460e-01 -2.87493207e-02 -7.74791420e-01 -8.14680755e-01
-4.82986718e-01 -8.51426363e-01 1.07386434e+00 5.96471250e-01
-7.37558529e-02 2.40650568e-02 -1.88276514e-01 1.37100354e-01
2.05470443e-01 6.28473461e-01 2.67345518e-01 -3.91277105e-01
-3.11009467e-01 -7.62022376e-01 -4.88964528e-01 2.11484894e-01
-2.80433714e-01 -7.56080300e-02 -1.16526532e+00 -1.37174595e-02
-7.66799629e-01 -3.92150909e-01 1.06239140e+00 2.68958777e-01
1.28306794e+00 5.72457373e-01 -5.79150245e-02 5.74097455e-01
1.04880190e+00 -1.56039968e-01 7.25947380e-01 5.38185835e-01
8.40169668e-01 2.64487118e-01 3.89897466e-01 3.58685374e-01
5.02208322e-02 6.58077538e-01 5.94806314e-01 -3.15859646e-01
-8.15351009e-01 -1.51943281e-01 3.65976185e-01 -1.24306791e-01
-7.32319415e-01 -3.95533442e-02 -1.05437160e+00 6.69017285e-02
-1.92584360e+00 -1.09232974e+00 -4.68435079e-01 2.31755543e+00
5.72458982e-01 6.84350133e-01 5.62036037e-01 3.20813686e-01
6.57761157e-01 -9.09560025e-02 -8.11262846e-01 -2.90988058e-01
-4.27895099e-01 6.02312624e-01 1.02483749e+00 2.15540111e-01
-1.21763456e+00 1.11237109e+00 6.26493454e+00 4.04958516e-01
-1.15679550e+00 -1.78008482e-01 4.80090886e-01 -8.08537900e-01
7.56483257e-01 -3.95803005e-01 -6.75744593e-01 2.97385663e-01
7.12379396e-01 2.37445340e-01 3.22039306e-01 7.54554093e-01
9.85183120e-02 -5.14820874e-01 -8.46739471e-01 1.07864320e+00
1.47963539e-01 -1.26673973e+00 -1.14547662e-01 -1.84872612e-01
2.45066926e-01 3.11894000e-01 -2.06692070e-01 3.94489497e-01
5.52617386e-02 -1.10147440e+00 1.21824896e+00 1.01836181e+00
8.39618802e-01 -9.35619473e-01 6.89808011e-01 3.91862839e-01
-1.04618084e+00 5.68748303e-02 -1.64814010e-01 -4.75187987e-01
-2.87430994e-02 -1.01497948e-01 -5.27245343e-01 2.33816326e-01
1.22068655e+00 6.85372412e-01 -8.90538871e-01 1.20910311e+00
-4.29847956e-01 5.85660338e-01 -5.23344874e-01 -2.28884891e-01
1.07205048e-01 1.22280024e-01 4.73060280e-01 1.38001931e+00
-1.40072517e-02 4.24414761e-02 -7.10663050e-02 4.57716316e-01
7.08117429e-03 -2.93562114e-01 -2.52770990e-01 4.46212918e-01
1.51410490e-01 1.24671400e+00 -9.55219924e-01 -8.18002343e-01
-1.67887226e-01 1.40658653e+00 1.62349895e-01 2.68650502e-01
-8.75133812e-01 -8.03731203e-01 6.71543062e-01 3.78974751e-02
4.98456508e-01 -3.12080085e-01 -8.19378972e-01 -8.63940299e-01
1.40548661e-01 -5.15465677e-01 3.18664432e-01 -1.08117092e+00
-8.10212433e-01 5.48271120e-01 -9.44639891e-02 -9.78006840e-01
-7.64426484e-04 -1.16085374e+00 -6.47938073e-01 9.24468517e-01
-9.75112140e-01 -9.48565125e-01 -9.74471092e-01 4.67651725e-01
5.46554804e-01 -1.93957180e-01 6.66454852e-01 7.90632963e-02
-6.84903085e-01 6.37825549e-01 -2.37842426e-01 7.00767100e-01
7.37628102e-01 -1.67605400e+00 7.23643064e-01 9.02155638e-01
-3.21677886e-02 -1.03083976e-01 5.57057261e-01 -7.10966527e-01
-1.32734549e+00 -7.16227293e-01 4.66778338e-01 -8.80691826e-01
2.91711092e-01 -5.05146205e-01 -6.51722252e-01 4.28674608e-01
1.63723767e-01 1.82005808e-01 1.96800828e-01 -2.86141455e-01
-3.27613801e-01 1.07111521e-01 -8.51060688e-01 6.16345406e-01
1.28530276e+00 -1.91085130e-01 -6.17596447e-01 1.32572185e-02
6.80768490e-02 -1.08850276e+00 -5.69599748e-01 7.00962171e-02
1.07660532e+00 -1.00662971e+00 1.07137775e+00 -1.00352192e+00
7.28825688e-01 -3.75778586e-01 5.54537820e-03 -9.18071151e-01
1.78676918e-01 -3.44551086e-01 -3.67863953e-01 6.23851478e-01
1.32318571e-01 1.02608964e-01 1.33655095e+00 6.70259595e-01
-1.12098791e-01 -7.02866673e-01 -1.11298382e+00 -5.89558780e-01
2.51958132e-01 -7.37264276e-01 2.92770654e-01 8.02211538e-02
-1.70169771e-01 1.02713406e-01 -3.23573679e-01 -8.74736011e-02
6.95250034e-01 -2.49823518e-02 1.25095820e+00 -9.26702678e-01
-3.47598307e-02 -7.48400331e-01 -7.63047338e-01 -1.13491368e+00
-2.71276593e-01 -6.82750404e-01 1.89889044e-01 -1.52626383e+00
-9.59286243e-02 1.48104250e-01 -2.96781808e-01 7.50850856e-01
-1.63031936e-01 6.50709748e-01 7.57916331e-01 6.45186603e-02
-3.74068975e-01 1.89120024e-01 1.09563041e+00 -4.45952229e-02
-1.53656766e-01 5.51059768e-02 1.86826810e-02 7.12576807e-01
9.17951465e-01 -1.66963562e-01 2.79515624e-01 -5.11422932e-01
-1.05855167e-01 4.03284468e-02 9.81107593e-01 -1.59479642e+00
3.64747673e-01 3.39935333e-01 1.21267903e+00 -1.16330123e+00
7.07353592e-01 -3.97424906e-01 -2.66217977e-01 6.94430351e-01
-4.68657494e-01 2.28587940e-01 3.26395124e-01 4.11940426e-01
4.88235712e-01 2.40627021e-01 5.00133634e-01 6.91349804e-02
-8.95151496e-01 1.05862401e-01 -5.21664619e-01 -6.37296960e-02
8.25159669e-01 -4.25349832e-01 -1.14310354e-01 -3.34232688e-01
-1.00336826e+00 1.31116569e-01 1.12082765e-01 7.67742038e-01
5.65424263e-01 -1.49221456e+00 -7.66873062e-01 2.45809421e-01
-7.46290833e-02 1.68198459e-02 7.14582717e-03 6.31535411e-01
-9.20918584e-01 5.50639071e-02 -7.46439219e-01 -6.66898012e-01
-1.28080487e+00 1.53599873e-01 5.52493215e-01 2.92079687e-01
-1.20254934e+00 9.23977494e-01 -5.19465983e-01 -2.33563483e-02
8.37348044e-01 -3.50906521e-01 -2.00216159e-01 2.11496800e-01
8.11820447e-01 7.43365884e-01 1.81313857e-01 -6.72940254e-01
-4.00545418e-01 6.11027539e-01 2.70176530e-01 -5.54063141e-01
1.26050949e+00 6.28622949e-01 5.83100259e-01 4.35666084e-01
8.87471914e-01 -4.08821881e-01 -1.99617088e+00 7.76423588e-02
-2.09607825e-01 -6.06691778e-01 1.34130657e-01 -1.16555929e+00
-1.32908177e+00 1.17981684e+00 1.00386381e+00 -3.36654514e-01
8.35957706e-01 -2.32695431e-01 5.77559948e-01 4.97889251e-01
-3.02013650e-04 -1.36038744e+00 8.61238688e-02 5.01347482e-01
1.11335909e+00 -9.79146898e-01 -7.02135339e-02 3.62535380e-02
-6.13558412e-01 1.48049271e+00 8.21494341e-01 -6.30804062e-01
3.46471816e-01 4.90453511e-01 3.53517503e-01 -2.06076697e-01
-2.44749188e-01 -5.29252291e-01 2.61444628e-01 5.90462148e-01
2.19088212e-01 -1.36780888e-01 -6.41419291e-02 6.97758496e-01
-4.18817818e-01 1.78962395e-01 3.68725032e-01 1.13547480e+00
-3.08518559e-01 -6.26613081e-01 -4.67525154e-01 6.34700000e-01
-1.93674698e-01 2.47216329e-01 -8.02710772e-01 9.78353202e-01
1.34137720e-01 5.50584674e-01 2.67660946e-01 -5.85543692e-01
1.00209403e+00 1.23905495e-01 6.78696334e-01 -2.16848001e-01
-9.89665568e-01 -1.33803576e-01 7.66980136e-03 -8.28390718e-01
-1.55457228e-01 -9.35584605e-01 -1.37206841e+00 -2.84513205e-01
5.89722097e-02 -5.04814565e-01 6.70712411e-01 7.67515540e-01
1.99373722e-01 7.66005039e-01 9.24352743e-03 -1.54269218e+00
-3.54825437e-01 -1.01861918e+00 -1.54788122e-01 3.77435356e-01
1.42369032e-01 -7.88892150e-01 -7.32754767e-02 1.65739041e-02] | [7.867064476013184, 0.0860719308257103] |
a69dce0e-73e8-46fb-ba80-ddedea1a70d9 | session-based-sequential-skip-prediction-via | 1902.04743 | null | http://arxiv.org/abs/1902.04743v1 | http://arxiv.org/pdf/1902.04743v1.pdf | Session-based Sequential Skip Prediction via Recurrent Neural Networks | The focus of WSDM cup 2019 is session-based sequential skip prediction, i.e.
predicting whether users will skip tracks, given their immediately preceding
interactions in their listening session. This paper provides the solution of
our team \textbf{ekffar} to this challenge. We focus on
recurrent-neural-network-based deep learning approaches which have previously
been shown to perform well on session-based recommendation problems. We show
that by choosing an appropriate recurrent architecture that properly accounts
for the given information such as user interaction features and song metadata,
a single neural network could achieve a Mean Average Accuracy (AA) score of
0.648 on the withheld test data. Meanwhile, by ensembling several variants of
the core model, the overall recommendation accuracy can be improved even
further. By using the proposed approach, our team was able to attain the 1st
place in the competition. We have open-sourced our implementation at GitHub. | ['Lin Zhu', 'Yihong Chen'] | 2019-02-13 | null | null | null | null | ['sequential-skip-prediction'] | ['time-series'] | [ 1.67713821e-01 -7.57936090e-02 -3.17224503e-01 -4.37985659e-01
-8.75165701e-01 -3.83231431e-01 3.12717408e-01 -2.08521113e-01
-3.75657588e-01 5.66815615e-01 6.28403842e-01 -3.38278115e-01
-2.97326863e-01 -3.54378313e-01 -6.91122711e-01 -2.91368425e-01
-1.76450029e-01 4.50767308e-01 1.42024130e-01 -3.16488385e-01
1.20272778e-01 8.74837581e-03 -1.89345539e+00 7.24483371e-01
3.28445047e-01 1.12884116e+00 2.52983093e-01 8.42941225e-01
1.71615735e-01 9.10000205e-01 -3.55054259e-01 -1.85614705e-01
2.30359823e-01 -4.74236518e-01 -9.38475609e-01 -4.57701057e-01
5.75009704e-01 -4.20559138e-01 -4.84459013e-01 3.12481135e-01
7.66599834e-01 6.37898862e-01 2.43093595e-01 -8.84763420e-01
-1.87750265e-01 1.34835374e+00 -1.66020408e-01 2.60184109e-01
5.19805372e-01 -2.11213112e-01 1.62887621e+00 -7.83793330e-01
3.18253487e-01 5.64317226e-01 8.10764909e-01 6.50210381e-01
-1.18742454e+00 -6.56889498e-01 3.55102867e-01 4.74442095e-01
-1.18356597e+00 -7.61526465e-01 5.75954854e-01 5.07846735e-02
1.17228842e+00 5.54407716e-01 5.61281145e-01 1.45408082e+00
-2.63386756e-01 1.11839664e+00 5.33625484e-01 -3.24910581e-01
-3.13070379e-02 -4.23581600e-02 4.40875262e-01 9.66866836e-02
-6.90740794e-02 2.00237017e-02 -8.71656179e-01 8.95290673e-02
3.12782139e-01 1.20851569e-01 -3.15735370e-01 -1.58116259e-02
-1.19067061e+00 4.48885024e-01 4.08429354e-01 6.75558686e-01
-2.22148538e-01 1.51394054e-01 4.54583347e-01 5.06280243e-01
5.33808887e-01 6.16617918e-01 -8.64276767e-01 -5.00615180e-01
-1.46419811e+00 4.58900422e-01 1.05877292e+00 7.08540142e-01
1.31958008e-01 -5.32092750e-02 -5.01981199e-01 1.01667643e+00
6.06224164e-02 7.08876699e-02 5.86913049e-01 -9.57484424e-01
2.41465971e-01 1.03280336e-01 2.79687881e-01 -5.39534628e-01
-8.08323026e-01 -9.22680736e-01 -6.32181048e-01 -3.95714641e-01
4.88253176e-01 -3.02832276e-01 -7.07538605e-01 1.83226871e+00
-2.58980721e-01 6.48418903e-01 -3.75057966e-01 8.65224600e-01
1.11255836e+00 4.53718394e-01 -1.24885686e-01 -3.14925015e-01
8.76897991e-01 -1.21748960e+00 -5.32886386e-01 -1.41596794e-02
6.94381952e-01 -7.01296449e-01 1.24421561e+00 7.27324188e-01
-1.27236474e+00 -7.85384119e-01 -1.23012805e+00 1.64198905e-01
-1.05716912e-02 3.64957273e-01 7.10964620e-01 6.13778174e-01
-1.29181385e+00 1.01293910e+00 -9.15436506e-01 -6.69205248e-01
5.01513816e-02 7.37535417e-01 5.65827340e-02 2.10675612e-01
-1.15112758e+00 5.01546443e-01 5.32794334e-02 2.45363995e-01
-7.55809546e-01 -7.96889603e-01 -7.21454695e-02 3.44730645e-01
3.61146986e-01 -6.55083835e-01 1.70482767e+00 -8.49642575e-01
-1.70369518e+00 6.86612308e-01 -3.39862496e-01 -8.07694376e-01
3.35580528e-01 -5.50902784e-01 -6.97699666e-01 -4.34353143e-01
-2.82898426e-01 2.58380622e-01 3.24410260e-01 -7.85782337e-01
-8.50259602e-01 -1.74874485e-01 2.40673617e-01 2.85176039e-01
-7.17974424e-01 1.35254115e-02 -5.79792976e-01 -4.95274782e-01
-1.00551523e-01 -1.12055838e+00 -1.82473689e-01 -6.56125724e-01
-3.73086959e-01 -3.58428091e-01 2.12620080e-01 -6.28964901e-01
1.71150851e+00 -2.00544786e+00 1.25862285e-01 1.42370269e-01
1.56359151e-02 3.01950037e-01 -1.99958518e-01 5.00983894e-01
6.00311607e-02 -2.14260886e-03 3.98948073e-01 -7.63743401e-01
1.71657085e-01 -1.08230293e-01 -4.96950448e-01 1.26813740e-01
-5.64279020e-01 7.68391013e-01 -4.81393605e-01 1.95474535e-01
-1.38663957e-02 3.57525617e-01 -7.42904425e-01 2.55044788e-01
-2.63173014e-01 3.78011584e-01 -2.71425128e-01 1.39179036e-01
3.91723156e-01 -4.51796770e-01 5.03766060e-01 -3.61851305e-02
-4.32293937e-02 1.03706634e+00 -1.11296344e+00 2.08445215e+00
-8.26834381e-01 7.28172481e-01 -1.53361082e-01 -7.78757036e-01
5.89687765e-01 4.65184987e-01 6.54799938e-01 -7.08433390e-01
2.80098058e-02 9.40658301e-02 7.80085102e-02 -2.09460437e-01
7.60565102e-01 1.98750392e-01 2.61223260e-02 7.83755362e-01
2.38022253e-01 7.56919563e-01 1.03013657e-01 1.95711389e-01
1.34316111e+00 3.78981680e-01 -1.40013769e-01 -1.35506660e-01
4.45476949e-01 -4.93491322e-01 2.35729694e-01 1.24310446e+00
1.05971783e-01 8.61725628e-01 1.06279008e-01 -5.14626384e-01
-8.64076138e-01 -7.82581210e-01 2.42796298e-02 1.92274690e+00
-2.77527601e-01 -8.50401700e-01 -5.17568350e-01 -5.37440181e-01
-4.03239042e-01 8.86649072e-01 -6.82500899e-01 -9.13510025e-02
-5.57252467e-01 -5.80440998e-01 4.72048968e-01 5.04103720e-01
1.73615962e-01 -1.22202539e+00 -4.62098829e-02 4.51562881e-01
-4.14287060e-01 -8.97682250e-01 -4.87779051e-01 3.77827317e-01
-5.95372200e-01 -5.40468216e-01 -7.07236171e-01 -5.45639217e-01
-1.66089386e-01 5.04783273e-01 1.35270536e+00 2.15337604e-01
1.96659341e-01 1.40458137e-01 -6.06611490e-01 -4.17304263e-02
1.53755665e-01 9.10036683e-01 2.21944481e-01 -1.82962250e-02
4.63185728e-01 -1.01886642e+00 -8.31785917e-01 3.65689009e-01
-4.68799859e-01 4.43711504e-02 1.88008934e-01 6.39030635e-01
1.94342658e-01 -3.44623774e-01 6.70080185e-01 -1.13818681e+00
5.29329002e-01 -4.30955827e-01 7.85335246e-03 1.74616203e-01
-7.48680055e-01 -2.32271537e-01 6.78292155e-01 -2.83990264e-01
-8.86879325e-01 -7.61389434e-02 -7.24305332e-01 -6.19684756e-02
-2.37627566e-01 6.26137257e-01 1.38223007e-01 3.44766140e-01
7.64138579e-01 2.82255769e-01 -5.71828246e-01 -8.55605900e-01
2.02350661e-01 7.32710600e-01 2.35224083e-01 -3.16464037e-01
2.47199178e-01 9.27547887e-02 -5.01310349e-01 -4.49524760e-01
-1.45073247e+00 -6.87643230e-01 -4.50796247e-01 -1.65056616e-01
3.87924433e-01 -1.05172479e+00 -1.13151729e+00 2.07183406e-01
-7.69955158e-01 -5.45674205e-01 -1.97657153e-01 5.73656917e-01
-5.28209567e-01 1.00577511e-02 -6.26148939e-01 -8.18460464e-01
-5.87476134e-01 -6.37588084e-01 7.04239130e-01 1.59456894e-01
-6.08505189e-01 -7.88141251e-01 2.54794300e-01 5.11442959e-01
6.63448513e-01 -3.86091948e-01 4.54178691e-01 -1.17607081e+00
-3.72253418e-01 -3.49582404e-01 2.00485229e-01 1.38627067e-01
-1.38909638e-01 -2.86253124e-01 -1.23253775e+00 -4.16114211e-01
-3.90923381e-01 -2.70968646e-01 1.16745770e+00 6.64091110e-01
1.53805888e+00 -2.08235383e-01 -2.96276540e-01 5.87922990e-01
1.02413809e+00 -8.21929052e-02 4.86394227e-01 2.66582102e-01
5.84651709e-01 1.88727796e-01 3.29230934e-01 5.80623627e-01
4.68502909e-01 1.07267654e+00 2.69428581e-01 1.66191608e-01
-2.58363783e-01 -3.67475361e-01 4.37782168e-01 6.73445761e-01
-4.60774273e-01 -6.92488968e-01 -5.31434834e-01 4.48562473e-01
-2.16847253e+00 -1.18285012e+00 -2.12618396e-01 2.33797908e+00
6.12676799e-01 4.09707308e-01 6.08740747e-01 1.70692250e-01
3.12124491e-01 1.60160214e-01 -4.63219792e-01 -3.02403808e-01
5.04780300e-02 4.09819722e-01 3.16390246e-01 3.93103570e-01
-1.15896547e+00 8.26540053e-01 6.74259186e+00 9.20071006e-01
-1.02301574e+00 9.68015715e-02 3.85389894e-01 -8.49696279e-01
-3.77378315e-01 -2.54963666e-01 -1.14108968e+00 5.01009643e-01
1.54846036e+00 -2.82513313e-02 8.86303544e-01 5.48489630e-01
4.07119364e-01 1.74286291e-01 -1.32709432e+00 7.36881554e-01
1.55471399e-01 -1.45108521e+00 -2.75885254e-01 7.67774060e-02
8.18096280e-01 5.13984621e-01 3.07516068e-01 7.66383827e-01
4.14515436e-01 -1.15470707e+00 5.04346788e-01 6.15153134e-01
5.43462694e-01 -6.62925780e-01 7.73065984e-01 4.04109538e-01
-1.02577662e+00 -2.28624880e-01 -1.14469349e-01 -3.58724236e-01
4.04599160e-02 5.22341251e-01 -8.49485636e-01 5.94359279e-01
9.07097995e-01 1.05101335e+00 -4.48031962e-01 1.26841414e+00
2.99672782e-02 1.18107152e+00 -4.42946523e-01 -9.56274644e-02
-3.18166688e-02 1.93077713e-01 4.10125852e-01 1.16843748e+00
6.10754788e-01 -1.21613219e-01 -1.74459908e-02 3.29004228e-01
-3.73567134e-01 1.86727300e-01 -1.92908615e-01 1.08989045e-01
2.12944895e-01 1.30664492e+00 -5.10339439e-01 4.60183844e-02
-3.36837411e-01 9.26833808e-01 3.38964760e-01 3.47086132e-01
-7.41414309e-01 -9.97332558e-02 6.21727407e-01 2.21002296e-01
6.79950953e-01 -4.16757353e-02 -4.04170305e-01 -1.33758771e+00
-1.80048883e-01 -7.68841028e-01 4.69498754e-01 -6.11999452e-01
-1.17387354e+00 8.09369981e-01 -5.97514153e-01 -1.31060803e+00
-3.74503851e-01 -3.21123004e-01 -7.43350983e-01 7.38066256e-01
-9.76904988e-01 -1.24356711e+00 8.62264726e-03 4.38695163e-01
6.39990032e-01 -4.12460893e-01 1.15690196e+00 7.38880992e-01
-6.82318807e-01 1.16384101e+00 4.36968625e-01 -2.70128787e-01
8.54783475e-01 -1.22867870e+00 3.71939689e-01 5.49269140e-01
7.04886019e-01 7.68226564e-01 1.04014194e+00 -3.99408378e-02
-1.07922435e+00 -8.34912717e-01 1.20795107e+00 -6.37676954e-01
6.19998395e-01 -4.65862781e-01 -5.02222538e-01 9.83677685e-01
1.72679648e-01 -3.55122566e-01 1.13789451e+00 1.24913549e+00
-1.76838607e-01 -3.83195311e-01 -4.40242261e-01 6.04807079e-01
1.48049045e+00 -4.95957524e-01 -2.45301753e-01 3.28153223e-01
7.26049542e-01 -5.24390519e-01 -7.04568207e-01 3.14130992e-01
1.04003632e+00 -9.93014455e-01 9.75926399e-01 -8.90282035e-01
5.28889358e-01 3.93112600e-02 -2.66214848e-01 -1.17306995e+00
-5.17321587e-01 -7.89562166e-01 -5.25250912e-01 1.15095925e+00
8.17841053e-01 6.29967228e-02 1.02758944e+00 4.31498975e-01
-1.80740044e-01 -7.47011185e-01 -7.90438950e-01 -4.24177229e-01
-1.56099781e-01 -7.60571361e-01 5.07076144e-01 5.86491168e-01
2.11613566e-01 5.94533324e-01 -1.06941378e+00 -1.92037836e-01
2.14064956e-01 3.38205755e-01 8.04372132e-01 -1.33809924e+00
-7.27168024e-01 -4.55615461e-01 1.74362257e-01 -1.51351917e+00
5.95786534e-02 -1.08157444e+00 -7.29335546e-02 -1.52133954e+00
1.22081853e-01 -3.38358968e-01 -1.11330175e+00 5.97615063e-01
7.43440390e-02 5.72563946e-01 1.87924117e-01 1.40393555e-01
-1.22105157e+00 2.84411132e-01 9.09847200e-01 1.32651538e-01
-4.52088743e-01 8.03786099e-01 -1.06837916e+00 3.95980597e-01
7.52534211e-01 -3.06678623e-01 -3.43927056e-01 -3.21256489e-01
6.02126420e-01 1.51096627e-01 6.60057515e-02 -1.14458895e+00
2.91792363e-01 1.79572165e-01 3.79589349e-01 -7.62875915e-01
5.41809082e-01 -7.10082531e-01 1.67767316e-01 1.16594657e-01
-1.12977922e+00 -5.15727758e-01 1.87749833e-01 5.83400190e-01
1.19703919e-01 -4.66486141e-02 5.33934422e-02 9.65970084e-02
-4.96795177e-01 3.40869039e-01 -4.76812005e-01 -3.21390748e-01
3.74221891e-01 -9.00362711e-03 6.43572025e-03 -8.09352279e-01
-1.37813640e+00 1.61836147e-01 -8.75031799e-02 6.82787001e-01
1.85911745e-01 -1.28187716e+00 -6.22900188e-01 -1.29183028e-02
2.11452916e-02 -7.98661768e-01 6.22297287e-01 8.85236800e-01
1.56134903e-01 6.60078228e-01 -1.28061756e-01 -3.18009675e-01
-1.25901544e+00 9.00083184e-02 3.72152269e-01 -6.72003567e-01
-4.71534580e-01 1.15515316e+00 -2.75260925e-01 -5.55348396e-01
7.05923259e-01 -1.03153594e-01 -5.66023767e-01 3.37515883e-02
6.72841132e-01 2.92510986e-01 4.49011266e-01 -1.33674443e-01
-2.41193876e-01 1.43916190e-01 -4.35715288e-01 -9.12570134e-02
1.69539583e+00 -2.51969308e-01 2.87974328e-01 7.51058221e-01
1.10670090e+00 1.09693713e-01 -1.05460048e+00 -5.02269745e-01
-7.08038807e-02 -1.04042880e-01 3.73775423e-01 -1.20927298e+00
-1.00554252e+00 6.54675126e-01 6.42465770e-01 4.97096807e-01
1.07545125e+00 -1.80952474e-01 1.02788103e+00 6.02997005e-01
3.32783550e-01 -1.01195693e+00 -2.60346025e-01 6.78559124e-01
6.29813135e-01 -1.19772112e+00 -2.50898395e-02 1.75555512e-01
-4.61752295e-01 9.66419756e-01 4.88497734e-01 -2.30332866e-01
8.88389885e-01 -5.81353642e-02 -2.47375056e-01 1.17548436e-01
-1.17401326e+00 -2.74381846e-01 6.06184125e-01 2.00521082e-01
1.02375472e+00 2.44292021e-01 -3.35291654e-01 1.24754643e+00
-3.85154635e-01 3.61085653e-01 3.89840007e-01 3.86225253e-01
-3.94712538e-01 -1.31479216e+00 3.65410775e-01 8.82027090e-01
-7.62279391e-01 -3.25171322e-01 -1.21118993e-01 2.18284383e-01
-7.84955323e-02 1.07315171e+00 -4.47982140e-02 -9.63445961e-01
4.02211398e-01 1.27913103e-01 4.04238343e-01 -6.37694478e-01
-1.07100391e+00 1.27191991e-01 5.27200520e-01 -7.44396329e-01
-3.33751380e-01 -7.18482614e-01 -8.62274647e-01 -5.97716510e-01
-1.22870333e-01 2.60806561e-01 4.35319245e-01 1.03340232e+00
4.89183366e-01 7.60781705e-01 6.99674249e-01 -9.38386261e-01
-3.74366760e-01 -1.24215257e+00 -6.83423162e-01 9.04778168e-02
4.44456577e-01 -2.76521385e-01 -2.25003332e-01 -1.91870421e-01] | [15.61312198638916, 5.190537452697754] |
6ff8cf64-f2d2-40bb-8509-e86eadf7b831 | self-supervised-contrastive-attributed-graph | 2110.08264 | null | https://arxiv.org/abs/2110.08264v1 | https://arxiv.org/pdf/2110.08264v1.pdf | Self-supervised Contrastive Attributed Graph Clustering | Attributed graph clustering, which learns node representation from node attribute and topological graph for clustering, is a fundamental but challenging task for graph analysis. Recently, methods based on graph contrastive learning (GCL) have obtained impressive clustering performance on this task. Yet, we observe that existing GCL-based methods 1) fail to benefit from imprecise clustering labels; 2) require a post-processing operation to get clustering labels; 3) cannot solve out-of-sample (OOS) problem. To address these issues, we propose a novel attributed graph clustering network, namely Self-supervised Contrastive Attributed Graph Clustering (SCAGC). In SCAGC, by leveraging inaccurate clustering labels, a self-supervised contrastive loss, which aims to maximize the similarities of intra-cluster nodes while minimizing the similarities of inter-cluster nodes, are designed for node representation learning. Meanwhile, a clustering module is built to directly output clustering labels by contrasting the representation of different clusters. Thus, for the OOS nodes, SCAGC can directly calculate their clustering labels. Extensive experimental results on four benchmark datasets have shown that SCAGC consistently outperforms 11 competitive clustering methods. | ['Xinbo Gao', 'Ming Yang', 'Quanxue Gao', 'Wei Xia'] | 2021-10-15 | null | null | null | null | ['graph-clustering'] | ['graphs'] | [ 2.80186273e-02 1.25773609e-01 -2.21635446e-01 -5.11178792e-01
-6.17703557e-01 -3.75894487e-01 4.89729375e-01 6.15624905e-01
5.78309521e-02 2.31291935e-01 -2.20079720e-01 -8.73026848e-02
-3.66805971e-01 -7.79263854e-01 -4.39855874e-01 -9.53542829e-01
-4.82450068e-01 5.01448154e-01 1.92185566e-01 2.61935562e-01
1.56368241e-01 8.84533450e-02 -1.18146646e+00 5.33154830e-02
1.12314725e+00 6.81571960e-01 1.03496209e-01 2.51335025e-01
-2.08656222e-01 8.40852559e-01 -3.08379412e-01 -2.23430581e-02
6.58540353e-02 -6.07539475e-01 -6.09106958e-01 2.92318761e-01
1.60344899e-01 2.97942102e-01 -3.57525259e-01 1.23255157e+00
2.14508787e-01 8.03906396e-02 9.79997396e-01 -1.86129367e+00
-6.53577626e-01 8.88250411e-01 -8.27260315e-01 -2.24810570e-01
6.29914328e-02 3.43844593e-02 1.36393011e+00 -7.02205122e-01
3.87747556e-01 1.34928787e+00 5.20824969e-01 1.98621899e-01
-1.50169718e+00 -6.56611204e-01 4.72277015e-01 1.15291826e-01
-1.75501633e+00 7.31763765e-02 1.19403362e+00 -4.93914306e-01
9.63969901e-02 3.22540820e-01 4.66615170e-01 6.22854590e-01
-3.61592859e-01 7.26289749e-01 9.19620872e-01 -2.39413574e-01
3.56158704e-01 -2.96479780e-02 2.89831042e-01 8.65539312e-01
1.77008107e-01 -1.84048012e-01 -9.15542915e-02 -2.25147959e-02
2.57925153e-01 2.15408102e-01 -2.98825324e-01 -8.22645128e-01
-1.07028985e+00 7.77791381e-01 1.07969475e+00 2.72122383e-01
-1.51532426e-01 1.61090493e-01 4.19298589e-01 2.49399513e-01
5.67724228e-01 -8.28701886e-04 -1.25286698e-01 5.56651533e-01
-8.07843268e-01 -2.16509283e-01 3.90464902e-01 1.18381977e+00
1.23013699e+00 -1.58290267e-01 -1.03045650e-01 7.03128397e-01
6.02928579e-01 1.93521962e-01 8.82072151e-02 -7.20424235e-01
5.43670833e-01 1.27566922e+00 -4.80714798e-01 -1.66673100e+00
-4.53658730e-01 -4.45475489e-01 -1.33042228e+00 -1.25886902e-01
1.19798288e-01 2.78968632e-01 -8.27496707e-01 1.68399096e+00
4.81517732e-01 4.65243787e-01 -1.49926513e-01 8.00832152e-01
1.03731775e+00 5.91128767e-01 2.29409873e-01 -2.08708078e-01
9.12572801e-01 -8.16433191e-01 -5.58088481e-01 8.59374478e-02
8.32956731e-01 -2.77734041e-01 1.18311834e+00 1.03983186e-01
-7.89003432e-01 -4.42722976e-01 -1.03254080e+00 2.95763433e-01
-3.31111908e-01 -3.99574488e-02 4.84968752e-01 5.66661775e-01
-9.95713711e-01 5.02837777e-01 -8.73521090e-01 -1.52099177e-01
5.26812077e-01 4.18405086e-01 -2.68345296e-01 -1.83821216e-01
-7.30986893e-01 8.31435993e-02 7.46366143e-01 9.15571228e-02
-6.39451206e-01 -4.70872968e-01 -9.89904284e-01 2.85331339e-01
6.59057856e-01 -2.69412398e-01 3.92355412e-01 -8.47652972e-01
-8.44412625e-01 7.79459834e-01 5.19412383e-02 7.69400820e-02
2.46781826e-01 4.96693343e-01 -5.51150501e-01 2.83013314e-01
3.86928916e-01 7.03432679e-01 7.95522928e-01 -1.70429397e+00
-4.79453981e-01 -5.17682374e-01 -3.63265067e-01 1.37520164e-01
-5.88314414e-01 -5.72782278e-01 -6.73351049e-01 -7.10202456e-01
7.02608705e-01 -9.14158046e-01 -3.88040811e-01 -2.30443835e-01
-9.10322368e-01 -6.43155873e-01 9.28434730e-01 -2.50188023e-01
1.43847072e+00 -2.33184600e+00 2.15646937e-01 7.64570415e-01
8.20871830e-01 1.03765065e-02 -1.49784222e-01 3.82093430e-01
-9.46695581e-02 3.44197243e-01 -5.07782757e-01 -2.81671733e-01
4.25395966e-02 -5.97349443e-02 2.27044299e-01 6.33342564e-01
1.63863286e-01 7.65239060e-01 -1.18503356e+00 -9.68467593e-01
4.20417130e-01 2.51910537e-01 -4.99488235e-01 3.16420287e-01
-1.92157682e-02 5.50432086e-01 -4.41215485e-01 6.42914712e-01
7.85972118e-01 -6.50120437e-01 6.36939585e-01 -3.44402432e-01
3.18612725e-01 -3.81590761e-02 -1.21835160e+00 1.38726282e+00
-1.30214691e-01 2.14653283e-01 1.16549805e-01 -1.31272244e+00
1.18293846e+00 8.49056523e-03 6.04110479e-01 -4.02518243e-01
7.50847682e-02 4.58305329e-02 -1.37303963e-01 -2.14137420e-01
-3.92238311e-02 2.95745522e-01 -2.31641471e-01 3.87103260e-01
-2.47932702e-01 2.65196770e-01 2.35225245e-01 8.32097292e-01
1.10608566e+00 -2.70998240e-01 -1.37616649e-01 -4.45864856e-01
8.06930780e-01 -8.66542831e-02 9.01429355e-01 2.98346817e-01
-2.39328668e-01 6.44995749e-01 8.60023022e-01 -2.24352479e-01
-5.97125173e-01 -1.20536780e+00 3.44218999e-01 9.50037599e-01
5.49234033e-01 -7.40134120e-01 -7.95678556e-01 -9.93385136e-01
-2.07519308e-02 2.63874680e-01 -4.66672629e-01 -3.15757036e-01
-6.03890061e-01 -7.29984522e-01 2.81566620e-01 5.18988192e-01
2.70951033e-01 -9.64033723e-01 1.37495309e-01 6.36191070e-02
-1.51802987e-01 -9.43342686e-01 -7.57264495e-01 5.37166931e-02
-7.70485044e-01 -1.47558928e+00 -3.48395169e-01 -1.45090222e+00
1.27215469e+00 5.88222742e-01 1.02381682e+00 6.41226828e-01
-1.87161252e-01 1.59248725e-01 -4.69151855e-01 3.01038802e-01
-3.45160812e-01 3.13133568e-01 -6.08119927e-02 2.36497417e-01
2.62584895e-01 -8.38328004e-01 -5.64928114e-01 3.78421187e-01
-7.55612850e-01 -6.45426847e-03 5.79560399e-01 7.34533966e-01
7.44178295e-01 4.57206994e-01 7.11847484e-01 -1.39278066e+00
4.24332201e-01 -7.14081287e-01 -7.12918520e-01 4.08245355e-01
-9.57459927e-01 6.02596700e-02 9.85818923e-01 -1.78150639e-01
-6.07632339e-01 1.63321704e-01 3.49930465e-01 -6.87790811e-01
2.37136949e-02 5.60736239e-01 -5.36710382e-01 1.73748329e-01
2.86445558e-01 1.12026386e-01 1.69894695e-01 -3.37756842e-01
5.68318188e-01 6.00466311e-01 6.19238079e-01 -4.05461550e-01
1.09574163e+00 4.72181976e-01 2.45402187e-01 -5.11535525e-01
-5.70201516e-01 -7.34920382e-01 -7.52105594e-01 -3.15247506e-01
7.15467453e-01 -8.76824558e-01 -9.43568766e-01 2.48244748e-01
-5.54883063e-01 -2.75175244e-01 2.00846732e-01 2.32536107e-01
-1.74217790e-01 7.35113740e-01 -5.83933294e-01 -6.03103518e-01
-1.16291523e-01 -1.07722449e+00 8.24909806e-01 1.23669647e-01
2.49517821e-02 -1.31291914e+00 -2.91639864e-01 1.83714136e-01
-6.59349188e-02 6.35097325e-01 1.31211936e+00 -5.73122740e-01
-5.36041439e-01 -1.33511007e-01 -4.37321037e-01 8.79181027e-02
6.00172758e-01 6.96922168e-02 -5.48276842e-01 -7.10195303e-01
-5.86748004e-01 -4.58917439e-01 7.92229056e-01 2.33076394e-01
1.42454386e+00 -3.74237835e-01 -7.28856027e-01 6.08590722e-01
1.32540131e+00 -1.37337342e-01 3.25623095e-01 -1.68575421e-01
1.37107539e+00 9.18188274e-01 5.62432528e-01 1.25518113e-01
7.31713176e-01 4.07308817e-01 4.93783683e-01 -1.32070035e-01
-1.74526602e-01 -5.70219934e-01 1.28109485e-01 1.23221171e+00
3.22149038e-01 -1.69293448e-01 -8.86282563e-01 4.07823473e-01
-1.98696804e+00 -4.86446321e-01 -7.89678097e-01 2.15711117e+00
5.05917728e-01 1.94815576e-01 2.63682425e-01 2.46237189e-01
1.37886584e+00 2.07783505e-01 -6.78159654e-01 3.43601108e-01
-6.37414679e-02 -2.38781109e-01 1.48276925e-01 4.63980377e-01
-1.24855494e+00 9.59098816e-01 5.26142216e+00 8.73255253e-01
-8.32789123e-01 -1.54536098e-01 8.55212510e-01 4.94659036e-01
-4.99690801e-01 2.69926280e-01 -2.65947849e-01 8.39126050e-01
4.95981753e-01 -1.11837246e-01 2.51140714e-01 8.01525891e-01
-1.01380125e-02 3.99354160e-01 -1.09440327e+00 9.67938662e-01
-1.65173516e-01 -9.65641499e-01 1.02206610e-01 1.00557640e-01
8.23828459e-01 -3.01666170e-01 2.04630047e-02 3.37245226e-01
6.22384846e-01 -9.60055232e-01 1.54505298e-01 1.63698316e-01
7.17491269e-01 -9.42063153e-01 5.30061424e-01 4.26225007e-01
-1.80067110e+00 -3.29197472e-04 -3.74876022e-01 2.71061599e-01
-2.74932802e-01 6.73760474e-01 -7.16168642e-01 6.19905710e-01
7.24769771e-01 1.08698094e+00 -8.29286397e-01 8.29083145e-01
-2.86261588e-01 8.11230719e-01 -9.05987397e-02 3.22831832e-02
3.42786372e-01 -5.02609611e-01 3.15900803e-01 8.90855193e-01
-7.10228756e-02 -1.07657083e-03 7.34299123e-01 1.06455255e+00
-2.41081104e-01 2.32051745e-01 -4.05298769e-01 5.22479117e-02
1.07114041e+00 1.62876964e+00 -1.27450585e+00 -1.86181635e-01
-1.06076293e-01 7.47651994e-01 6.99561179e-01 3.26009095e-01
-5.62108338e-01 -5.66741109e-01 1.98982105e-01 1.06509350e-01
2.33730331e-01 -1.07055731e-01 -1.51101932e-01 -8.41949344e-01
4.09063399e-02 -6.14180982e-01 7.60605574e-01 -3.63979250e-01
-1.54293287e+00 5.19863367e-01 -1.13936640e-01 -1.50689232e+00
1.70831546e-01 -1.18364878e-01 -8.88468266e-01 3.35145056e-01
-1.28550720e+00 -1.29726577e+00 -5.59238970e-01 6.26276612e-01
8.20059255e-02 -1.65648594e-01 4.02772248e-01 3.47949743e-01
-7.88705468e-01 6.72090411e-01 6.48044571e-02 4.66365457e-01
6.00037396e-01 -1.53460467e+00 2.09008023e-01 6.49385273e-01
7.20537752e-02 4.92492735e-01 2.34910578e-01 -6.60241365e-01
-1.13761008e+00 -1.58913875e+00 6.29238546e-01 -1.33312076e-01
3.16560090e-01 -5.54473877e-01 -1.19701838e+00 6.04817569e-01
-1.44153073e-01 3.46281022e-01 6.70768559e-01 1.64104462e-01
-5.87296426e-01 -3.85429502e-01 -1.07249856e+00 4.97766554e-01
1.07688105e+00 -3.39073479e-01 4.20472138e-02 4.58042413e-01
8.70089412e-01 1.64540127e-01 -1.00935352e+00 4.80019599e-01
-9.87038910e-02 -1.00159848e+00 8.73214602e-01 -3.79664660e-01
1.36263758e-01 -7.89921165e-01 6.83915764e-02 -1.41835201e+00
-5.72430789e-01 -4.41549182e-01 2.00906079e-02 1.60352421e+00
1.84486553e-01 -4.96095926e-01 1.01079249e+00 2.01813534e-01
-1.90157309e-01 -6.28363550e-01 -5.13954759e-01 -8.25023174e-01
-1.02353901e-01 -7.11079985e-02 6.47223353e-01 1.36131060e+00
8.22373852e-02 6.99136496e-01 -1.15949892e-01 4.04936373e-01
1.27403021e+00 1.85931772e-01 7.11396337e-01 -1.63123167e+00
1.33150786e-01 -4.82847989e-01 -5.30119121e-01 -6.80565298e-01
4.68279541e-01 -1.41617346e+00 1.18411735e-01 -1.63074005e+00
2.92418510e-01 -1.13486207e+00 -3.35159183e-01 5.30940950e-01
-6.38637304e-01 2.39683717e-01 9.22381356e-02 5.47100127e-01
-1.19782889e+00 6.49796546e-01 1.08263183e+00 -4.36864644e-01
-2.73809850e-01 -1.82338819e-01 -6.96710527e-01 6.98075116e-01
7.12940037e-01 -6.55183733e-01 -6.88694239e-01 -4.08717021e-02
9.29570794e-02 -4.56865132e-02 2.34768301e-01 -9.03501332e-01
5.01587510e-01 4.09335718e-02 1.14789553e-01 -8.23434889e-01
-3.28886844e-02 -8.28826785e-01 8.20379779e-02 5.90446711e-01
-3.65429133e-01 -7.16512650e-02 -5.88816047e-01 9.32499111e-01
-2.54168719e-01 2.18437329e-01 7.29077578e-01 1.66870039e-02
-6.39113069e-01 6.01591229e-01 -1.04257219e-01 4.19082761e-01
1.15722072e+00 -6.88143447e-02 -9.56786722e-02 -2.72094250e-01
-7.71697104e-01 9.04325664e-01 4.24966127e-01 1.89790666e-01
7.04331517e-01 -1.62207413e+00 -7.97234237e-01 7.94066712e-02
3.86651129e-01 5.06462872e-01 2.85917670e-01 7.01666653e-01
-3.86690140e-01 4.06829221e-03 2.10327432e-01 -9.93817806e-01
-1.31945252e+00 8.51545691e-01 -9.24639404e-02 -2.40315244e-01
-6.27066910e-01 7.37697840e-01 2.39038438e-01 -9.02867377e-01
3.68765146e-01 7.40939155e-02 -1.82225540e-01 -2.69536544e-02
3.23951021e-02 6.21868670e-01 -6.86533749e-02 -7.05989063e-01
-4.08575624e-01 5.27360797e-01 -2.30085403e-01 4.64795530e-01
1.17041624e+00 -2.68013030e-01 -3.44093651e-01 3.77044350e-01
1.51245499e+00 -2.69358903e-01 -1.25618970e+00 -3.97445470e-01
4.69032168e-01 -4.09488231e-01 -1.76561363e-02 -2.21646622e-01
-1.43405247e+00 8.04080963e-01 2.84194887e-01 4.98266965e-01
1.19694829e+00 2.42214337e-01 5.94398081e-01 3.26030195e-01
2.33279213e-01 -8.95853400e-01 4.04348165e-01 -1.49236992e-01
3.46864730e-01 -1.34032524e+00 1.69967055e-01 -9.95293736e-01
-6.57437265e-01 7.65880346e-01 8.74377429e-01 -2.19216615e-01
8.64505827e-01 -1.42984673e-01 1.66070368e-02 -5.08018732e-01
-4.99373525e-01 -1.54338896e-01 2.49713272e-01 6.48848414e-01
2.51094580e-01 3.18116188e-01 -4.77957167e-02 3.63734156e-01
9.16412994e-02 -7.24929154e-01 2.15742216e-01 6.29378617e-01
-2.89049327e-01 -1.11283076e+00 -1.66095346e-01 5.38197577e-01
-2.27185665e-03 2.93218289e-02 -7.09008336e-01 7.17847288e-01
-7.72219598e-02 1.07507777e+00 1.13960087e-01 -6.61939681e-01
-2.96566705e-03 -5.07209361e-01 7.97985345e-02 -6.13363981e-01
-3.81253600e-01 7.35430941e-02 -3.94199908e-01 -2.46757179e-01
-5.64069688e-01 -3.53177011e-01 -1.62643933e+00 -3.53560150e-01
-4.82155621e-01 5.35760522e-01 1.38681158e-01 6.08677804e-01
4.12423819e-01 5.10524333e-01 1.18290114e+00 -5.26506364e-01
-2.31259674e-01 -7.43386567e-01 -9.12543178e-01 8.12522888e-01
1.57508850e-01 -5.02006352e-01 -6.16310000e-01 -1.14771545e-01] | [7.324424743652344, 5.968616485595703] |
a5ccfd77-6fae-4ab3-aba3-52b5f967d451 | reflection-removal-using-a-dual-pixel-sensor | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Punnappurath_Reflection_Removal_Using_a_Dual-Pixel_Sensor_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Punnappurath_Reflection_Removal_Using_a_Dual-Pixel_Sensor_CVPR_2019_paper.pdf | Reflection Removal Using a Dual-Pixel Sensor | Reflection removal is the challenging problem of removing unwanted reflections that occur when imaging a scene that is behind a pane of glass. In this paper, we show that most cameras have an overlooked mechanism that can greatly simplify this task. Specifically, modern DLSR and smartphone cameras use dual pixel (DP) sensors that have two photodiodes per pixel to provide two sub-aperture views of the scene from a single captured image. "Defocus-disparity" cues, which are natural by-products of the DP sensor encoded within these two sub-aperture views, can be used to distinguish between image gradients belonging to the in-focus background and those caused by reflection interference. This gradient information can then be incorporated into an optimization framework to recover the background layer with higher accuracy than currently possible from the single captured image. As part of this work, we provide the first image dataset for reflection removal consisting of the sub-aperture views from the DP sensor.
| [' Michael S. Brown', 'Abhijith Punnappurath'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['reflection-removal'] | ['computer-vision'] | [ 1.15631068e+00 -1.40059248e-01 6.13523483e-01 -1.40819564e-01
-6.44178867e-01 -4.47631180e-01 3.24934393e-01 -4.45442349e-01
-3.40876132e-01 5.41755974e-01 1.51947320e-01 -1.29219085e-01
2.72019684e-01 -6.31592035e-01 -5.67969739e-01 -1.19423544e+00
6.58736944e-01 -2.48109981e-01 4.82008010e-01 1.52948678e-01
3.61465365e-01 5.11049926e-01 -1.86481905e+00 6.17159188e-01
5.66444218e-01 9.62670922e-01 5.56510031e-01 6.30431056e-01
1.09180272e-01 7.67750561e-01 -5.44756889e-01 3.22504039e-03
5.15450060e-01 -3.56707036e-01 1.40299127e-01 2.64439881e-01
1.02546930e+00 -8.69341075e-01 -3.12029243e-01 1.26061797e+00
1.88328758e-01 -1.27176255e-01 3.35053861e-01 -6.22157633e-01
1.08597446e-02 -1.58651739e-01 -7.36964345e-01 -1.44623304e-02
7.15471268e-01 2.58761108e-01 4.45792198e-01 -9.75135088e-01
3.72125417e-01 8.97161365e-01 5.07545769e-01 3.29040557e-01
-1.16008401e+00 -4.62097436e-01 -1.68313161e-01 -1.14329547e-01
-9.82492566e-01 -7.89969862e-01 1.03095555e+00 -4.42659914e-01
7.25539505e-01 4.06420261e-01 5.56903005e-01 8.28708291e-01
4.55278128e-01 2.44543895e-01 1.49704754e+00 -4.62774992e-01
-2.32463460e-02 4.35421258e-01 1.08136348e-01 4.86445069e-01
6.11715376e-01 2.86760300e-01 -4.84985471e-01 1.11189894e-02
6.42519534e-01 3.84418279e-01 -1.09184921e+00 -3.03562939e-01
-8.03463578e-01 2.02667594e-01 1.04639597e-01 1.35370448e-01
-2.97924310e-01 -2.11529300e-01 -5.47298312e-01 -4.55234423e-02
3.29638302e-01 4.71419930e-01 2.71082809e-03 1.96269140e-01
-8.02940845e-01 -1.33446857e-01 6.75721109e-01 4.88193274e-01
8.96161199e-01 -2.49619260e-01 7.16361701e-02 7.02249765e-01
6.07467413e-01 9.93667066e-01 7.56089985e-02 -1.01751494e+00
4.68959361e-01 6.35127485e-01 5.55247009e-01 -7.91119635e-01
-1.13138638e-01 -2.40756750e-01 -4.61472481e-01 6.46589279e-01
6.03838563e-01 -1.32450610e-01 -8.73229086e-01 1.09715629e+00
4.39904511e-01 1.55883402e-01 2.66116321e-01 1.18240786e+00
7.09411800e-01 5.68133295e-01 -1.09282231e+00 -4.70859736e-01
1.34354937e+00 -7.04905391e-01 -6.34544373e-01 -6.67052329e-01
-5.41124158e-02 -1.08354592e+00 6.24676108e-01 8.89880538e-01
-1.33924341e+00 -2.45775163e-01 -1.27888715e+00 -7.73977712e-02
1.71739593e-01 3.59083951e-01 6.13724440e-02 7.43647337e-01
-6.96465909e-01 3.12187135e-01 -6.96683884e-01 1.36973888e-01
-3.19943242e-02 9.25821438e-02 -8.37108344e-02 -7.32422531e-01
-4.93411571e-01 6.24863148e-01 -3.60305637e-01 2.13587463e-01
-5.74406326e-01 -8.55465233e-01 -6.03880525e-01 -1.88795820e-01
6.18495941e-01 -3.51142734e-01 7.57435203e-01 -8.10449839e-01
-1.67541349e+00 9.85623479e-01 -4.80962545e-01 -1.60032690e-01
2.92486966e-01 -3.94653648e-01 -2.43824273e-01 7.75912464e-01
-9.60035920e-02 -2.83331454e-01 1.27577639e+00 -1.54205382e+00
-5.28663337e-01 -6.28675759e-01 -7.53383618e-03 2.93501973e-01
2.65622199e-01 -4.94727939e-02 -4.79543358e-01 4.64680679e-02
3.60273749e-01 -7.74733841e-01 -3.10144722e-02 -2.03312963e-01
-5.13116360e-01 8.79744589e-01 1.02584827e+00 -5.44885397e-01
8.24568391e-01 -2.25759673e+00 -8.60583857e-02 -2.54067242e-01
3.28509241e-01 5.73040664e-01 -2.94636935e-02 4.16623533e-01
1.18968666e-01 -5.92193127e-01 -3.09597552e-01 -3.36738706e-01
-7.93902576e-01 -1.22424014e-01 -1.98889926e-01 6.64394200e-01
-3.25344950e-02 2.97429591e-01 -6.46897197e-01 1.26344219e-01
5.26003480e-01 5.89680374e-01 -2.33924016e-01 3.37946475e-01
-1.03373736e-01 7.01964617e-01 -2.13111877e-01 5.11143684e-01
1.30312085e+00 1.14866979e-01 1.83572635e-01 -4.86136109e-01
-6.38666511e-01 1.29441902e-01 -1.21891904e+00 1.08694041e+00
-4.36091989e-01 6.96351469e-01 5.31325877e-01 -5.97458661e-01
9.56596673e-01 6.39693812e-02 4.89934981e-01 -8.34136486e-01
3.20296511e-02 3.35721314e-01 -2.53548443e-01 -7.05103636e-01
5.49755275e-01 -2.86278307e-01 3.50238800e-01 2.89102823e-01
-4.43905950e-01 -2.86033452e-01 1.77858621e-02 -2.36953676e-01
1.12236238e+00 -1.14551909e-01 3.05231288e-02 2.05566175e-02
7.27119148e-01 -2.61867315e-01 5.36595821e-01 8.66958201e-01
2.80010611e-01 1.20503569e+00 1.18200600e-01 -2.50720024e-01
-9.41797197e-01 -9.73369181e-01 -4.11845371e-02 -4.04293500e-02
3.40395123e-01 -5.64828366e-02 -7.69900262e-01 -3.65919247e-02
-1.41458929e-01 5.32171488e-01 -8.18749741e-02 6.74351072e-03
-5.88249385e-01 -6.96565151e-01 -1.08158812e-01 -1.27087489e-01
6.22915268e-01 -4.09799874e-01 -1.31839967e+00 1.13320880e-01
-4.37686563e-01 -1.63684046e+00 2.91294176e-02 7.00226566e-03
-6.62141979e-01 -1.64632869e+00 -6.97851062e-01 -1.15638852e-01
6.91837609e-01 1.06397319e+00 1.10525620e+00 -1.77205071e-01
-7.47146666e-01 7.33272135e-01 -2.23351568e-01 -2.98928887e-01
-1.71960250e-01 -8.63609850e-01 -3.01779538e-01 7.09542632e-01
2.80844033e-01 -5.17523766e-01 -9.21900868e-01 2.87040293e-01
-9.01150405e-01 4.06998724e-01 5.56444108e-01 5.70705354e-01
3.84313613e-01 -1.35514783e-02 -4.41385806e-01 -6.95581675e-01
1.63016655e-02 -2.97542531e-02 -1.17731202e+00 -1.33789390e-01
-2.47999653e-01 -2.81547755e-01 4.97178614e-01 -1.67006239e-01
-1.59360290e+00 1.18911214e-01 2.76159793e-01 -4.54000890e-01
-3.22290778e-01 -1.67729221e-02 -2.36844450e-01 -1.59622759e-01
4.06205684e-01 2.99978882e-01 7.85363168e-02 -3.89048874e-01
-1.96143508e-01 6.72476649e-01 4.78299886e-01 1.15943616e-02
4.04213518e-01 1.26368332e+00 3.75347406e-01 -1.62538040e+00
-1.13704312e+00 -8.94057155e-01 -2.27034479e-01 -4.25643265e-01
9.63202715e-01 -9.92189407e-01 -7.75474787e-01 8.77695084e-01
-1.10374928e+00 -2.62872994e-01 -3.00948042e-02 6.75597489e-01
-5.72369918e-02 4.93606240e-01 -4.76050049e-01 -1.14190733e+00
1.87177062e-02 -1.17197216e+00 1.34413159e+00 6.53487980e-01
5.32784164e-01 -6.59261823e-01 2.87737604e-02 7.95113981e-01
4.36429709e-01 -1.77988131e-02 4.22198653e-01 4.72592533e-01
-1.07442391e+00 -1.14196114e-01 -2.19632402e-01 7.82543600e-01
1.68166265e-01 6.34595705e-03 -1.45362639e+00 -2.43960932e-01
9.19091940e-01 1.19605415e-01 1.02279663e+00 8.08102429e-01
6.34866178e-01 7.20955059e-02 -4.15911555e-01 9.38131690e-01
1.97128069e+00 2.80838102e-01 1.05090237e+00 1.45633250e-01
6.54576123e-01 6.22567892e-01 7.67584205e-01 3.19435328e-01
-3.51337083e-02 8.69895995e-01 7.43740499e-01 -3.19004238e-01
-3.67513388e-01 2.97367960e-01 4.67769980e-01 1.06925502e-01
-5.63397296e-02 -6.31803811e-01 -5.54721653e-01 3.22908163e-01
-1.43209803e+00 -9.62994039e-01 -9.01679933e-01 2.52082801e+00
2.66592920e-01 -3.44679952e-01 -5.04789472e-01 6.38901442e-02
5.90867639e-01 1.14411213e-01 -3.20584267e-01 1.46699369e-01
-4.12529469e-01 2.18758419e-01 7.67719865e-01 1.05293405e+00
-7.38287985e-01 3.72207791e-01 5.92084265e+00 9.80782136e-02
-1.36551917e+00 -3.08725387e-01 1.97554037e-01 -2.84142077e-01
-4.76737410e-01 4.24056873e-02 -1.22004628e+00 6.65729702e-01
5.53675294e-01 5.58183432e-01 3.40001643e-01 5.03229238e-02
4.01651829e-01 -1.05685365e+00 -1.01278424e+00 1.30273950e+00
3.84381950e-01 -6.29415095e-01 -3.27965856e-01 3.03132147e-01
6.48993790e-01 5.35055175e-02 3.00620764e-01 -7.87875831e-01
-2.13197544e-01 -5.65266728e-01 7.90740401e-02 8.49461138e-01
5.26181102e-01 -3.13267499e-01 4.90377247e-01 2.93706924e-01
-6.58888876e-01 -1.83259025e-01 -3.51773232e-01 -8.43394622e-02
4.03677285e-01 1.34660947e+00 -7.35099554e-01 4.48735893e-01
7.08106101e-01 7.07539439e-01 -1.59416765e-01 8.44641566e-01
-3.88298601e-01 1.99860886e-01 -4.22424525e-01 4.50774252e-01
-4.72860783e-02 -1.03401637e+00 1.07699704e+00 7.23432064e-01
6.43662393e-01 2.54437208e-01 -1.15762971e-01 1.13083625e+00
3.43554199e-01 -7.35616922e-01 -7.99279451e-01 2.90165037e-01
-9.75704566e-02 1.42423820e+00 -4.69631284e-01 -1.38472781e-01
-9.06901717e-01 1.08364868e+00 -3.32097113e-01 7.28570700e-01
-4.76810575e-01 -3.81157756e-01 9.75103498e-01 5.55562079e-01
4.54055250e-01 -1.93641946e-01 1.22155279e-01 -1.49607563e+00
4.72824454e-01 -6.54944658e-01 -2.84116954e-01 -1.25869203e+00
-8.67045581e-01 1.40274227e-01 -1.97841719e-01 -1.07936180e+00
2.81722784e-01 -9.42998469e-01 -4.57710743e-01 1.11342883e+00
-1.96960998e+00 -7.94303119e-01 -6.93725348e-01 6.94158137e-01
1.27975136e-01 4.26756740e-01 6.20805621e-01 4.56365198e-01
-2.95040101e-01 -2.62392998e-01 5.83098590e-01 -1.55363575e-01
6.21985316e-01 -8.86013806e-01 -2.85978258e-01 1.30767298e+00
-2.29083210e-01 3.80434632e-01 7.03167856e-01 -5.07816374e-01
-1.74362826e+00 -5.22482216e-01 4.38679546e-01 -2.79332459e-01
-8.09058547e-02 -3.73692960e-01 -7.00957119e-01 4.37564731e-01
1.38649270e-01 5.73176332e-02 4.99088168e-01 -5.38207650e-01
-9.82798934e-02 -3.91235471e-01 -1.04791784e+00 3.85818481e-01
6.60527766e-01 -4.80775088e-01 -4.28281933e-01 3.77857909e-02
1.25787899e-01 -5.47842264e-01 -2.56305099e-01 4.25508231e-01
7.18967497e-01 -1.68342805e+00 1.18728232e+00 3.19901407e-01
4.17854160e-01 -5.17305851e-01 -2.26976216e-01 -1.08159161e+00
3.23221385e-01 -5.92577696e-01 2.67640859e-01 1.07057941e+00
2.15091952e-03 -1.09764862e+00 9.63848889e-01 4.22454834e-01
-2.01107353e-01 -1.11984968e-01 -3.49634260e-01 -4.56650823e-01
-6.50791466e-01 -1.65241793e-01 -1.87271368e-02 4.54135418e-01
-4.88057762e-01 3.31385672e-01 -3.01928490e-01 5.28048158e-01
1.23531306e+00 4.48688954e-01 8.73944581e-01 -1.21270740e+00
-4.38776731e-01 9.58378017e-02 -1.44385785e-01 -1.48613572e+00
-3.67761999e-01 -3.06141466e-01 2.38218263e-01 -1.34498930e+00
1.93616122e-01 -2.49003038e-01 2.66916901e-01 -3.29128236e-01
-1.60227001e-01 3.59027505e-01 1.55205712e-01 1.33409142e-01
-5.96225448e-02 1.21367224e-01 1.03759801e+00 1.89252511e-01
-3.25730562e-01 1.68741807e-01 -5.50110638e-01 9.27433193e-01
5.04530430e-01 -4.20597345e-01 -2.88497597e-01 -5.80974162e-01
4.30619985e-01 2.55579561e-01 6.22200072e-01 -1.06501734e+00
1.97509840e-01 2.81803440e-02 3.81459922e-01 -6.58576250e-01
9.00007725e-01 -1.08758843e+00 4.72444087e-01 2.75824189e-01
2.84037352e-01 -6.74153388e-01 -4.91971150e-02 6.82796180e-01
-2.84284532e-01 -5.62628031e-01 9.96149063e-01 -5.01430333e-01
-1.24405339e-01 -2.58864999e-01 -6.58287346e-01 1.62240956e-02
7.74853051e-01 -3.51222992e-01 -6.62510455e-01 -2.51854718e-01
-3.12473059e-01 -3.09681088e-01 9.60542619e-01 -2.33688593e-01
9.54686880e-01 -6.65211618e-01 -4.27633584e-01 4.78577375e-01
-1.94769815e-01 1.66547656e-01 4.82528538e-01 1.10714936e+00
-7.90046632e-01 4.15274262e-01 -9.84478649e-03 -7.87225664e-01
-1.81671739e+00 3.15634757e-01 6.57113016e-01 -1.32396072e-01
-8.82925868e-01 6.76534891e-01 6.03176534e-01 1.15635663e-01
-1.49794534e-01 -3.25649381e-01 -7.79869109e-02 -2.31480896e-01
1.02821505e+00 4.74302143e-01 1.81254253e-01 -4.34509754e-01
-2.02098951e-01 1.04230881e+00 1.22898385e-01 -2.52467692e-01
1.42371655e+00 -5.24401844e-01 -2.36171037e-01 5.54993451e-01
1.16288078e+00 6.67911530e-01 -1.46191132e+00 -1.79947585e-01
-6.51246786e-01 -9.91605878e-01 3.00087005e-01 -5.06225884e-01
-1.08274209e+00 1.11250389e+00 4.20639575e-01 2.32714817e-01
1.31326067e+00 -3.24092716e-01 5.61410010e-01 2.71507017e-02
1.78324074e-01 -1.00893736e+00 6.05555400e-02 1.11552991e-01
4.29129601e-01 -1.08127463e+00 1.33622393e-01 -8.51464987e-01
-2.58576870e-01 1.41617525e+00 2.11604312e-01 -6.28698617e-02
3.63964081e-01 5.14236748e-01 1.99955165e-01 -3.35368782e-01
-5.02532184e-01 -3.19688827e-01 1.45546451e-01 6.45787120e-01
2.11295232e-01 -4.08540010e-01 -2.94483267e-03 -3.31460945e-02
3.87080252e-01 -1.02646835e-01 1.08212304e+00 7.75027931e-01
-6.31511331e-01 -8.65910888e-01 -7.28923619e-01 3.22776020e-01
-6.25662446e-01 -7.60247782e-02 -2.29410112e-01 3.50311756e-01
1.50556844e-02 1.30775404e+00 -3.61103266e-02 2.28060797e-01
4.33120519e-01 -1.62178889e-01 9.17119741e-01 -6.50843143e-01
-2.82975793e-01 4.38684851e-01 -2.91263536e-02 -8.00356925e-01
-7.10699916e-01 -7.17416406e-01 -7.08514094e-01 2.91549385e-01
-3.37498635e-01 -2.48798609e-01 7.48310745e-01 5.56360364e-01
1.67965099e-01 2.68422544e-01 6.28088415e-01 -1.05254805e+00
-1.74739957e-01 -5.21775424e-01 -9.46238816e-01 3.43021661e-01
8.59255970e-01 -4.29605067e-01 -1.09349716e+00 -1.21259749e-01] | [10.117599487304688, -2.8109686374664307] |
483de1cd-8084-48e2-9d77-fe68d292dd5b | unified-gradient-reweighting-for-model | 2010.13228 | null | https://arxiv.org/abs/2010.13228v1 | https://arxiv.org/pdf/2010.13228v1.pdf | Unified Gradient Reweighting for Model Biasing with Applications to Source Separation | Recent deep learning approaches have shown great improvement in audio source separation tasks. However, the vast majority of such work is focused on improving average separation performance, often neglecting to examine or control the distribution of the results. In this paper, we propose a simple, unified gradient reweighting scheme, with a lightweight modification to bias the learning process of a model and steer it towards a certain distribution of results. More specifically, we reweight the gradient updates of each batch, using a user-specified probability distribution. We apply this method to various source separation tasks, in order to shift the operating point of the models towards different objectives. We demonstrate different parameterizations of our unified reweighting scheme can be used towards addressing several real-world problems, such as unreliable separation estimates. Our framework enables the user to control a robustness trade-off between worst and average performance. Moreover, we experimentally show that our unified reweighting scheme can also be used in order to shift the focus of the model towards being more accurate for user-specified sound classes or even towards easier examples in order to enable faster convergence. | ['Paris Smaragdis', 'Dimitrios Bralios', 'Efthymios Tzinis'] | 2020-10-25 | null | null | null | null | ['audio-source-separation'] | ['audio'] | [ 2.11502746e-01 -2.29674980e-01 7.14680180e-02 -3.80998284e-01
-7.26417065e-01 -5.22328615e-01 4.47396278e-01 1.06709458e-01
-4.84554976e-01 4.67648536e-01 1.41233400e-01 -2.37538040e-01
-3.51260960e-01 -3.38640630e-01 -4.01806861e-01 -9.52752769e-01
-1.41057283e-01 2.56235480e-01 3.70578259e-01 -1.12982243e-01
2.35650659e-01 5.49841285e-01 -1.46371436e+00 1.38878793e-01
7.31640160e-01 7.64260650e-01 3.15620273e-01 8.55489075e-01
2.05517024e-01 3.03963065e-01 -8.02901804e-01 -1.28050327e-01
2.32873380e-01 -5.44956803e-01 -5.86253941e-01 6.12244047e-02
3.38408321e-01 -4.56537306e-03 1.42082080e-01 1.08868325e+00
8.62131417e-01 2.86875606e-01 6.61790848e-01 -1.09414577e+00
-1.33651823e-01 8.34839702e-01 -4.78612036e-01 4.19291466e-01
7.75903091e-02 4.69276831e-02 9.73735690e-01 -5.03448308e-01
4.67947684e-02 1.15057623e+00 7.90964186e-01 4.65625018e-01
-1.49965906e+00 -8.40339601e-01 3.94878149e-01 1.93868876e-01
-1.30045569e+00 -8.06296945e-01 9.47984040e-01 -3.87273431e-01
5.95097840e-01 3.07419986e-01 3.19211811e-01 9.78836179e-01
-9.52027366e-02 7.17270851e-01 8.73079479e-01 -6.28425717e-01
4.86682266e-01 2.36840844e-01 1.38914689e-01 1.93072647e-01
3.11511904e-01 -1.49003342e-01 -5.65753162e-01 -3.33228976e-01
5.12153089e-01 -3.60224038e-01 -7.50745833e-01 -6.74808145e-01
-9.84638631e-01 7.03403115e-01 1.62766919e-01 5.08642793e-01
-1.69527963e-01 1.40009612e-01 3.87427151e-01 4.07481939e-01
6.39387250e-01 6.87580645e-01 -5.29587328e-01 -2.37297922e-01
-1.11732483e+00 3.71382236e-01 8.47810209e-01 3.95956099e-01
4.26025987e-01 1.09714054e-01 -4.07673270e-01 1.18972313e+00
2.90187836e-01 1.39047816e-01 7.96626985e-01 -8.33235204e-01
4.14999872e-01 -1.88674964e-02 3.58579606e-01 -8.73133898e-01
-5.19733131e-01 -7.95168281e-01 -4.93315697e-01 4.69711781e-01
6.87411845e-01 -4.90437299e-01 -8.53107393e-01 1.85043502e+00
2.17015862e-01 2.63317734e-01 -1.94126874e-01 1.13286078e+00
1.80425450e-01 5.37797272e-01 -8.97995606e-02 -1.02731884e-01
1.04133177e+00 -8.43339920e-01 -4.83850002e-01 -3.29036713e-01
4.79508638e-01 -7.88887084e-01 1.23927093e+00 7.24253595e-01
-1.12539756e+00 -5.15055001e-01 -1.24670982e+00 3.58926386e-01
-5.63681498e-02 3.51265579e-01 1.98888779e-01 8.99080992e-01
-9.10986006e-01 9.39945936e-01 -9.45365429e-01 -1.47622764e-01
1.31648466e-01 5.21058440e-01 5.86781688e-02 3.26954544e-01
-1.01428425e+00 7.23927200e-01 4.80523676e-01 1.28948361e-01
-6.37391329e-01 -8.45011652e-01 -7.12490857e-01 3.04543436e-01
2.66709030e-01 -5.75494051e-01 1.19374967e+00 -1.09858406e+00
-1.83479595e+00 5.06450772e-01 5.84659316e-02 -5.53656876e-01
6.36427164e-01 -4.20411736e-01 -2.60254204e-01 -2.70262778e-01
-3.35248023e-01 4.31097329e-01 1.29483104e+00 -1.28842556e+00
-5.47150791e-01 -1.67933792e-01 -1.58653900e-01 1.96506009e-01
-7.00816095e-01 7.40109608e-02 -4.04315352e-01 -8.52799952e-01
-1.62706554e-01 -1.01076281e+00 -1.76536649e-01 -2.25967139e-01
-1.63761958e-01 -5.69253415e-02 5.29238880e-01 -3.06363195e-01
1.30872738e+00 -2.43665743e+00 5.40951908e-01 2.40743935e-01
4.16602232e-02 4.99251783e-01 -2.46064112e-01 2.50616848e-01
-2.02973604e-01 -1.67869240e-01 -4.74859148e-01 -7.76456952e-01
1.66801419e-02 1.58002481e-01 -4.62775767e-01 6.81115985e-01
2.59182781e-01 1.03245676e-01 -9.41907465e-01 -1.02126688e-01
5.31365089e-02 6.92990541e-01 -7.22928405e-01 4.41248715e-01
1.13468878e-01 4.85640198e-01 -6.16963916e-02 -9.09649804e-02
6.79572701e-01 1.39465600e-01 1.23545595e-01 -8.49575773e-02
1.87764615e-02 4.19968158e-01 -1.74486756e+00 1.43197572e+00
-8.21680665e-01 6.82668447e-01 4.92418319e-01 -1.06166339e+00
7.53462911e-01 1.68109879e-01 2.29394868e-01 -1.32305816e-01
1.98386967e-01 4.16721106e-02 3.78067285e-01 -1.86544538e-01
4.48025584e-01 -1.48433462e-01 4.46430355e-01 5.03888130e-01
8.10761303e-02 -2.18456447e-01 2.55543232e-01 -1.52631909e-01
6.77659035e-01 -3.80162187e-02 1.27050923e-02 -5.14048994e-01
6.74694419e-01 -6.83992267e-01 3.30455065e-01 9.01746452e-01
-1.31174624e-01 1.00660586e+00 4.76075977e-01 5.14985844e-02
-6.98643506e-01 -8.92390907e-01 -3.84897321e-01 1.41264105e+00
-4.40991260e-02 -3.18223476e-01 -8.01829875e-01 -5.66760778e-01
-4.37829010e-02 7.43414581e-01 -4.76096809e-01 -3.66163045e-01
-6.34380877e-01 -1.03558064e+00 5.09156883e-01 3.85065198e-01
1.15897425e-01 -8.90706182e-01 -8.54387343e-01 2.62788773e-01
-2.16653198e-02 -8.18102062e-01 -5.43124318e-01 5.11071265e-01
-8.16147864e-01 -7.23002970e-01 -1.01328421e+00 -4.81506497e-01
5.09487689e-01 1.18263207e-01 1.00893068e+00 -9.42791700e-02
1.59011647e-01 3.90343845e-01 -4.33326364e-01 -3.63233477e-01
-4.49360788e-01 4.60420877e-01 2.02444509e-01 3.70241255e-01
-4.70538959e-02 -6.44418895e-01 -5.37240803e-01 2.39017710e-01
-1.03964806e+00 -3.97492379e-01 1.99911714e-01 7.60454059e-01
3.92753296e-02 2.12843910e-01 6.85050607e-01 -7.00227261e-01
1.19668472e+00 -4.85764086e-01 -5.12108684e-01 6.98237717e-02
-7.49985576e-01 4.51873481e-01 7.19490945e-01 -9.09224510e-01
-8.36985886e-01 -1.70224503e-01 -3.33960503e-01 -3.32791090e-01
-7.15837404e-02 2.55054504e-01 -8.89150426e-02 -7.82959443e-03
5.78607500e-01 4.72931787e-02 -1.70145407e-01 -6.49316311e-01
2.90715247e-01 6.97482288e-01 3.69334340e-01 -6.79485857e-01
5.95050514e-01 2.19026521e-01 -2.77523071e-01 -5.33813715e-01
-6.77153051e-01 -5.12431681e-01 -4.87699896e-01 -5.62929660e-02
3.01121920e-01 -6.19227231e-01 -4.41336721e-01 5.28390706e-01
-8.13629985e-01 -6.40093148e-01 -1.72370672e-01 5.22756875e-01
-4.86475855e-01 2.54633248e-01 -3.75342190e-01 -8.29606473e-01
-2.67941147e-01 -1.20175850e+00 1.05006754e+00 1.38400644e-01
-4.27563667e-01 -1.16853714e+00 3.65145594e-01 -1.92567676e-01
6.31015837e-01 -3.57088715e-01 6.55595779e-01 -8.36556673e-01
2.80667901e-01 -3.03770974e-02 -3.96952070e-02 5.46847582e-01
2.56260842e-01 8.84270743e-02 -1.27996719e+00 -5.44688165e-01
1.23654716e-01 -7.81833753e-02 1.06958485e+00 4.28227395e-01
1.16813409e+00 -3.57874036e-02 -2.59966433e-01 6.08798444e-01
8.75793934e-01 6.45948797e-02 3.98027807e-01 3.93115968e-01
3.55959475e-01 5.52493036e-01 3.62343788e-01 4.10391241e-01
-7.63577148e-02 9.57628131e-01 3.09305817e-01 -6.28306493e-02
-1.51835173e-01 1.69137701e-01 3.91687125e-01 4.80764508e-01
3.44783030e-02 -2.39086851e-01 -7.99393177e-01 6.48240805e-01
-1.85146117e+00 -6.87940419e-01 2.80610085e-01 2.64797330e+00
1.05469179e+00 4.80489939e-01 4.01679605e-01 5.19030333e-01
6.18990541e-01 3.40250045e-01 -3.19231868e-01 -3.84163737e-01
4.59904611e-01 2.38295883e-01 3.29470694e-01 7.35670328e-01
-1.00986814e+00 5.76519728e-01 6.64894724e+00 9.00012195e-01
-1.58668590e+00 7.85386749e-03 2.73324311e-01 -4.00924146e-01
-1.49356768e-01 -1.82065025e-01 -5.81355333e-01 5.60389459e-01
8.45006168e-01 -8.75114347e-04 4.16233242e-01 6.03495836e-01
2.50421762e-01 4.58699763e-02 -1.16609132e+00 9.74206865e-01
1.63664073e-02 -8.57598662e-01 -2.31714442e-01 -1.71473771e-01
3.94694299e-01 -9.20561180e-02 1.36715621e-01 2.64852971e-01
1.85081065e-01 -8.10223699e-01 8.46510530e-01 2.73567230e-01
2.47937724e-01 -8.54620218e-01 5.67651212e-01 2.67040551e-01
-7.97076821e-01 -2.70394653e-01 -1.86757743e-01 -1.52939364e-01
1.03491485e-01 8.57977390e-01 -7.79743314e-01 3.29252213e-01
7.57406175e-01 2.54848093e-01 -4.88004029e-01 1.23703969e+00
-4.28660996e-02 8.93297493e-01 -3.13365489e-01 -3.28041054e-02
1.64844513e-01 -1.82472631e-01 7.33959615e-01 1.63697636e+00
1.60873011e-01 -5.49886286e-01 8.42914358e-02 5.40831923e-01
7.78542757e-02 9.03533027e-02 -1.78053588e-01 2.81485379e-01
3.86545449e-01 1.20496893e+00 -7.86920488e-01 -1.76539883e-01
-2.89109871e-02 7.75043368e-01 5.05920410e-01 4.49618429e-01
-9.87728775e-01 -7.29000211e-01 8.92905891e-01 1.83361158e-01
6.86927974e-01 -3.56838197e-01 -2.68687993e-01 -1.11077309e+00
-1.06994204e-01 -8.90124083e-01 1.58088893e-01 -5.94454110e-01
-9.77255464e-01 7.47688115e-01 1.98374346e-01 -9.17083740e-01
-2.29097903e-01 -5.19853115e-01 -8.07583392e-01 9.08538282e-01
-1.44584334e+00 -4.01254177e-01 -3.47726680e-02 3.58421683e-01
5.99697053e-01 -1.15918308e-01 4.78387743e-01 4.50318962e-01
-7.20131874e-01 7.09923387e-01 1.36415854e-01 -1.27372190e-01
1.06807637e+00 -1.35835505e+00 1.06461167e-01 9.29882467e-01
4.25429255e-01 5.70595980e-01 1.31731021e+00 -4.80693579e-02
-7.92766273e-01 -8.05818021e-01 4.49022889e-01 -3.58719558e-01
8.10327888e-01 -4.45972055e-01 -1.06113851e+00 3.10818583e-01
1.93843693e-01 -1.41633570e-01 5.64857900e-01 4.70189005e-01
-2.42729858e-01 -4.78984922e-01 -9.44489777e-01 6.73713744e-01
6.37414634e-01 -2.31066108e-01 -4.39127386e-01 1.21793091e-01
3.74594957e-01 -4.73884106e-01 -4.58992541e-01 4.30574149e-01
5.70061028e-01 -1.06654346e+00 9.59903061e-01 -5.45133829e-01
5.65687232e-02 -3.64020348e-01 8.72025043e-02 -1.79749286e+00
-4.49515760e-01 -5.97238302e-01 -2.32921615e-01 1.44292009e+00
4.13218856e-01 -7.00847089e-01 4.30996537e-01 3.84789169e-01
-8.70488510e-02 -6.70310974e-01 -9.39063728e-01 -6.80670857e-01
8.08184370e-02 -6.27160549e-01 4.97263163e-01 5.30311763e-01
-1.19407341e-01 1.69690371e-01 -5.22266090e-01 3.47154796e-01
3.56834829e-01 6.32010028e-02 6.88840449e-01 -9.59382474e-01
-6.52800918e-01 -8.29347074e-01 -2.16178969e-01 -1.12588263e+00
1.27679244e-01 -5.01899183e-01 2.52959162e-01 -9.39586043e-01
-1.72404394e-01 -5.10382533e-01 -6.91313446e-01 3.44347149e-01
-4.49607790e-01 3.51863682e-01 3.87896627e-01 9.67506692e-03
-3.65486145e-01 4.68516141e-01 7.15405166e-01 -8.87743384e-02
-5.64048350e-01 4.64461595e-01 -8.73905361e-01 6.65984809e-01
9.15646315e-01 -6.78397775e-01 -6.51136875e-01 -5.37730336e-01
1.84405595e-01 -4.30553198e-01 2.47755170e-01 -1.20327556e+00
1.97869450e-01 -8.01286381e-03 1.67903543e-01 2.70742066e-02
2.53191501e-01 -7.93079138e-01 -1.23752147e-01 2.85837710e-01
-5.72300792e-01 -2.81264752e-01 3.66956294e-01 3.85034680e-01
-1.52876675e-01 -6.63334429e-01 8.83361518e-01 3.23968232e-01
-1.34207323e-01 -1.62346944e-01 -3.11432809e-01 4.32692617e-02
6.21187866e-01 -1.03510367e-02 2.76545674e-01 -5.38783610e-01
-8.69788587e-01 2.36422732e-01 2.36057758e-01 4.74025011e-01
2.02667132e-01 -1.11115396e+00 -6.68679655e-01 3.52517068e-01
-3.24430801e-02 -2.02519387e-01 -5.76828606e-02 8.65632474e-01
-1.71568632e-01 1.60254493e-01 6.78769201e-02 -6.70427799e-01
-1.27792656e+00 5.36058068e-01 6.08881295e-01 -1.70041442e-01
-4.86094564e-01 8.24918449e-01 6.33456707e-02 -1.54563010e-01
6.23032451e-01 -5.47107160e-01 -2.92231381e-01 2.94362754e-01
6.74420476e-01 3.14953089e-01 3.97893488e-01 -3.27463180e-01
-4.60964054e-01 4.76631522e-01 -8.56951401e-02 -1.59929186e-01
1.31916165e+00 -1.91162825e-01 2.31090277e-01 5.81641972e-01
9.37001526e-01 5.07259309e-01 -1.59300256e+00 4.96177189e-03
-1.05524128e-02 -4.21240598e-01 2.91405559e-01 -6.16098821e-01
-1.14398086e+00 8.16165805e-01 6.86167479e-01 5.90554476e-01
1.26471841e+00 -1.32543996e-01 4.72394198e-01 1.35075971e-01
1.28548250e-01 -8.60410392e-01 -1.90344110e-01 3.22818846e-01
8.16732824e-01 -1.10034454e+00 -6.17115349e-02 3.00107785e-02
-5.57641089e-01 1.23455715e+00 2.45923430e-01 -2.98500210e-01
7.07592785e-01 4.67245579e-01 2.43133217e-01 1.97865501e-01
-5.71227491e-01 -1.04829013e-01 4.42880362e-01 6.26293361e-01
6.50511384e-01 -1.75763533e-01 -1.93187371e-01 5.05088389e-01
-1.92885369e-01 -1.92486942e-01 4.17890042e-01 6.25800788e-01
-3.98197412e-01 -1.33717501e+00 -5.98345578e-01 1.68702722e-01
-6.51305318e-01 -1.00552760e-01 -1.34354562e-01 4.49608386e-01
4.31114845e-02 1.07850122e+00 1.23037115e-01 -4.76384833e-02
3.94441754e-01 1.13085106e-01 5.10696113e-01 -5.79487145e-01
-5.48799813e-01 2.29797289e-01 1.10281110e-01 -1.96342766e-01
-4.90349293e-01 -6.15918040e-01 -8.42949986e-01 -3.33730206e-02
-5.05992949e-01 3.90120417e-01 4.99057531e-01 1.03801835e+00
2.96578705e-01 7.18002737e-01 5.59220076e-01 -1.19375312e+00
-7.76280642e-01 -1.08051193e+00 -4.99169767e-01 3.40170503e-01
6.69977248e-01 -7.95937479e-01 -5.85048795e-01 1.38640194e-03] | [15.408809661865234, 5.601019859313965] |
0d0f0627-c7e3-4939-a09a-3490d0080665 | updet-universal-multi-agent-reinforcement | 2101.08001 | null | https://arxiv.org/abs/2101.08001v3 | https://arxiv.org/pdf/2101.08001v3.pdf | UPDeT: Universal Multi-agent Reinforcement Learning via Policy Decoupling with Transformers | Recent advances in multi-agent reinforcement learning have been largely limited in training one model from scratch for every new task. The limitation is due to the restricted model architecture related to fixed input and output dimensions. This hinders the experience accumulation and transfer of the learned agent over tasks with diverse levels of difficulty (e.g. 3 vs 3 or 5 vs 6 multi-agent games). In this paper, we make the first attempt to explore a universal multi-agent reinforcement learning pipeline, designing one single architecture to fit tasks with the requirement of different observation and action configurations. Unlike previous RNN-based models, we utilize a transformer-based model to generate a flexible policy by decoupling the policy distribution from the intertwined input observation with an importance weight measured by the merits of the self-attention mechanism. Compared to a standard transformer block, the proposed model, named as Universal Policy Decoupling Transformer (UPDeT), further relaxes the action restriction and makes the multi-agent task's decision process more explainable. UPDeT is general enough to be plugged into any multi-agent reinforcement learning pipeline and equip them with strong generalization abilities that enables the handling of multiple tasks at a time. Extensive experiments on large-scale SMAC multi-agent competitive games demonstrate that the proposed UPDeT-based multi-agent reinforcement learning achieves significant results relative to state-of-the-art approaches, demonstrating advantageous transfer capability in terms of both performance and training speed (10 times faster). | ['Xiaodan Liang', 'Xiaojun Chang', 'Fengda Zhu', 'Siyi Hu'] | 2021-01-20 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-1.34242907e-01 1.62419468e-01 -1.54671654e-01 1.44339323e-01
-3.63652349e-01 -4.54277217e-01 6.58258677e-01 6.70209620e-03
-8.85684133e-01 8.56861353e-01 -3.26317586e-02 -2.58946776e-01
-3.48509192e-01 -6.39822066e-01 -6.90965354e-01 -7.65508413e-01
5.74603900e-02 7.88868368e-01 4.61977243e-01 -6.10648036e-01
1.24518849e-01 4.05396372e-02 -1.52934992e+00 1.33277178e-01
1.13427508e+00 6.41987085e-01 8.94311428e-01 5.05435288e-01
-3.24063189e-02 9.21391308e-01 -7.83137619e-01 -2.98039168e-01
3.29434663e-01 -5.20153642e-01 -7.68167973e-01 -3.89246806e-03
-1.46004066e-01 -5.81487417e-01 -1.78856462e-01 7.73360789e-01
6.73628271e-01 4.31979120e-01 4.83930618e-01 -1.44392872e+00
-8.72316718e-01 7.55673110e-01 -6.58358872e-01 1.74056619e-01
-1.88399758e-02 3.83383870e-01 1.08556962e+00 -3.01833719e-01
2.85496682e-01 1.19120395e+00 1.45661294e-01 1.04187179e+00
-1.06039000e+00 -5.06171823e-01 7.21562743e-01 3.29953760e-01
-6.79117799e-01 1.23195000e-01 6.62571073e-01 -1.19435795e-01
1.29934537e+00 -6.09694533e-02 7.97456145e-01 1.31107962e+00
2.62036264e-01 1.05724895e+00 1.25492573e+00 -2.93404579e-01
4.76229519e-01 -3.25911865e-02 -2.21164361e-01 6.10781074e-01
7.86682665e-02 2.56401986e-01 -3.18313390e-01 6.08863905e-02
1.09451222e+00 1.27590775e-01 -6.17757672e-03 -6.18567109e-01
-1.28283954e+00 9.46709991e-01 5.39031982e-01 2.46615887e-01
-6.71626568e-01 3.96511585e-01 5.99874258e-01 6.28360212e-01
2.30854586e-01 7.41082430e-01 -6.46788836e-01 -2.62308240e-01
-3.74202251e-01 4.69319314e-01 5.96946359e-01 7.92755723e-01
6.21147454e-01 4.75944996e-01 -1.37729764e-01 8.73958409e-01
1.77421063e-01 3.98231775e-01 8.85249019e-01 -8.94646585e-01
4.70800787e-01 5.83005846e-01 7.04372451e-02 -3.31673384e-01
-6.83983326e-01 -7.29356110e-01 -6.45973563e-01 7.81413019e-01
2.72572637e-01 -5.61148822e-01 -7.98738956e-01 1.97069132e+00
4.00778115e-01 2.27741361e-01 4.63964045e-01 9.25824344e-01
4.89200443e-01 4.75076318e-01 3.47625464e-01 -1.10960998e-01
1.41691256e+00 -1.47697031e+00 -4.86224532e-01 -6.69099629e-01
3.36574495e-01 -3.20383638e-01 1.14950323e+00 3.05359215e-01
-1.07495975e+00 -6.42074645e-01 -1.08126509e+00 3.16219479e-01
-2.64586121e-01 -2.50341684e-01 7.70532131e-01 2.14200154e-01
-1.02995479e+00 4.06560779e-01 -8.01065862e-01 -8.13926011e-02
1.38634622e-01 6.15729451e-01 -1.44290969e-01 1.73841998e-01
-1.21570277e+00 1.20040596e+00 6.44967973e-01 -1.99102372e-01
-1.05825818e+00 -5.61164200e-01 -7.72210896e-01 3.39304119e-01
7.34517336e-01 -1.03976798e+00 1.56193590e+00 -1.13916266e+00
-2.11538482e+00 9.76452157e-02 3.71004015e-01 -5.58665931e-01
5.58547735e-01 -1.65307239e-01 -8.37315619e-02 4.06067409e-02
-3.85115966e-02 8.38765085e-01 9.77636814e-01 -1.24860346e+00
-1.02337825e+00 -6.09350316e-02 7.32299447e-01 8.23222518e-01
-3.70928675e-01 -1.45278484e-01 -1.87505230e-01 -6.61631167e-01
-5.87822139e-01 -1.03487718e+00 -4.74385649e-01 -6.27966762e-01
3.03345650e-01 -4.27172720e-01 6.14217520e-01 -9.69569534e-02
9.11124647e-01 -1.97688055e+00 6.53229773e-01 -3.40986639e-01
3.49927664e-01 4.10513788e-01 -7.05391884e-01 5.04239917e-01
4.85399179e-02 -2.72831738e-01 1.26132771e-01 -4.92417246e-01
2.88014293e-01 5.13068020e-01 1.92961842e-02 6.29639775e-02
2.27444455e-01 1.07788026e+00 -1.20803618e+00 -1.60432592e-01
3.24723065e-01 1.22044250e-01 -7.67173171e-01 3.86678785e-01
-5.51727295e-01 4.92117941e-01 -6.53854549e-01 1.40931398e-01
4.03533638e-01 -3.23952347e-01 3.10337782e-01 3.96073312e-01
-1.40931169e-02 1.74304977e-01 -1.12539816e+00 1.68037117e+00
-8.12262297e-01 4.20027077e-02 4.04345021e-02 -9.07912850e-01
7.91836917e-01 5.32014191e-01 4.55593139e-01 -1.00090647e+00
5.13934381e-02 1.31769672e-01 6.00028515e-01 -2.31447354e-01
6.30067885e-01 -1.41466722e-01 -1.50065601e-01 7.25809336e-01
4.24966216e-01 -1.31468764e-02 3.78208220e-01 -8.17968026e-02
1.04724157e+00 3.58777225e-01 3.94949526e-01 -1.71172082e-01
3.33568037e-01 -8.22519958e-02 7.01753080e-01 9.17866051e-01
-2.80267447e-01 2.03778580e-01 3.81697327e-01 -4.40655142e-01
-1.06066394e+00 -8.38082194e-01 4.82219696e-01 1.73543119e+00
2.18224287e-01 -2.55806446e-01 -5.50832272e-01 -9.91846085e-01
1.10730939e-01 5.72774649e-01 -6.75067484e-01 -2.78440535e-01
-7.57249892e-01 -8.91521454e-01 1.45826012e-01 6.13556027e-01
5.88965774e-01 -1.75740218e+00 -1.06535411e+00 5.22476256e-01
8.27269852e-02 -9.91793096e-01 -3.86253804e-01 4.78475928e-01
-6.81314111e-01 -8.58719528e-01 -7.56980836e-01 -8.62535775e-01
3.94108981e-01 2.97282964e-01 9.89008248e-01 7.61247724e-02
2.45707840e-01 4.17602390e-01 -5.37898779e-01 -4.45355564e-01
-4.40617085e-01 3.51462513e-01 2.43035153e-01 -2.71775872e-01
7.43266344e-02 -6.64798379e-01 -7.31431425e-01 3.45399112e-01
-9.71618652e-01 2.32377067e-01 9.09365058e-01 1.18961179e+00
1.36299983e-01 -3.17586549e-02 1.11084008e+00 -6.78406835e-01
1.07349813e+00 -5.46178818e-01 -5.58965802e-01 1.62313119e-01
-7.67205119e-01 1.55758038e-01 9.94068265e-01 -8.79780054e-01
-1.07906914e+00 -2.43795946e-01 -5.73102571e-02 -2.28172168e-01
-8.61820355e-02 4.69535291e-01 2.42621943e-01 1.42624751e-01
4.29680496e-01 3.51142406e-01 3.79650712e-01 -1.73781425e-01
3.89598817e-01 3.38623405e-01 1.04869887e-01 -6.71058655e-01
5.48539042e-01 -5.02076410e-02 -2.82802165e-01 -3.06432456e-01
-4.71936584e-01 -1.66947156e-01 -2.86206305e-01 -1.59085274e-01
7.38130093e-01 -1.04362667e+00 -1.10014629e+00 6.47891998e-01
-9.26097929e-01 -9.31219816e-01 -3.50509584e-01 5.59300601e-01
-8.61571431e-01 -7.36254593e-03 -8.10353220e-01 -5.79694211e-01
-3.12042326e-01 -1.43994689e+00 6.57071888e-01 4.97550994e-01
7.04806298e-02 -1.15654743e+00 1.73339352e-01 2.31366485e-01
7.03012228e-01 -2.27360502e-01 9.75040674e-01 -8.37400675e-01
-3.78608167e-01 3.83203030e-01 4.69909012e-02 2.49513000e-01
1.63518742e-01 -4.96991724e-01 -6.16067469e-01 -5.60573578e-01
-8.10657069e-02 -7.36516237e-01 6.76549435e-01 3.09846789e-01
8.14108729e-01 -1.58952713e-01 7.03409538e-02 2.09029719e-01
1.40837681e+00 2.91867286e-01 4.96185094e-01 9.24536169e-01
5.23535490e-01 1.89384863e-01 6.54449582e-01 5.84540129e-01
6.51884735e-01 6.54829979e-01 8.93864751e-01 -2.17773184e-01
-1.21080680e-02 1.43422028e-02 6.37147248e-01 7.51410127e-01
-3.13131213e-01 -3.18462104e-01 -5.68076015e-01 4.03204590e-01
-2.31927872e+00 -9.44411576e-01 5.70620060e-01 1.90756726e+00
7.37929583e-01 2.12409183e-01 4.19289082e-01 -3.62791330e-01
4.45793748e-01 2.91691452e-01 -9.34728742e-01 -5.83575487e-01
1.53553858e-01 2.43838951e-01 2.76693285e-01 4.63226110e-01
-8.96738291e-01 1.22505856e+00 6.17554283e+00 8.73044610e-01
-1.02324140e+00 2.20543399e-01 1.94250017e-01 -2.85614401e-01
-5.61116524e-02 -3.51425618e-01 -7.00669229e-01 2.92103350e-01
7.77264893e-01 -1.92394331e-01 8.66967678e-01 8.08081388e-01
9.91961360e-02 4.23970222e-02 -9.32427168e-01 5.90549409e-01
-2.22698838e-01 -1.06056428e+00 1.34657174e-01 1.00735538e-01
6.78582191e-01 2.92558402e-01 2.50797838e-01 9.83718693e-01
8.90983462e-01 -6.93644166e-01 8.63837302e-01 4.62934598e-02
3.77852201e-01 -8.39677870e-01 7.09286392e-01 6.38756514e-01
-9.76973951e-01 -5.10963142e-01 -4.39030588e-01 -3.95576656e-01
-2.12913398e-02 -3.26420248e-01 -7.90704429e-01 6.92335069e-01
5.17974257e-01 4.11869377e-01 -3.79083842e-01 7.84162462e-01
-2.50911981e-01 2.15166718e-01 5.02310842e-02 -3.12909722e-01
7.22704351e-01 -2.80664057e-01 3.74428779e-01 7.88631678e-01
1.48215383e-01 -1.34126931e-01 5.54406285e-01 5.85888505e-01
6.27054423e-02 2.52254009e-02 -4.37814146e-01 1.89759374e-01
4.75496173e-01 1.41864336e+00 -5.76709330e-01 -1.99639633e-01
-6.31538987e-01 1.06783426e+00 9.88967538e-01 4.10493314e-01
-1.03916919e+00 -4.49255407e-02 8.02260995e-01 -3.32658708e-01
7.30610549e-01 -2.42915228e-01 1.63945198e-01 -1.03637505e+00
-1.57646388e-01 -1.36940408e+00 3.23430598e-01 -4.83723611e-01
-1.36086833e+00 9.59819257e-01 -5.19042239e-02 -1.12834716e+00
-6.89270437e-01 -7.24532008e-01 -6.52245820e-01 7.61021554e-01
-1.71829093e+00 -1.28609419e+00 -7.48052727e-03 6.58696711e-01
8.86794865e-01 -6.17982268e-01 9.85924840e-01 2.91192364e-02
-7.10279822e-01 5.43170214e-01 1.11414909e-01 -2.01244354e-01
6.10966027e-01 -1.66649365e+00 2.95365125e-01 5.25550902e-01
-2.08374247e-01 3.82053196e-01 5.50636470e-01 -4.11925852e-01
-1.32467461e+00 -8.56585979e-01 7.03952760e-02 -1.70132890e-01
8.88656378e-01 -3.16755325e-01 -7.35768020e-01 6.53906643e-01
7.41809011e-01 -2.20298633e-01 5.77591777e-01 3.09130251e-01
-8.90764967e-02 -1.00852214e-01 -8.76227021e-01 8.53029191e-01
8.53531897e-01 -1.24195509e-01 -7.36633480e-01 3.94565091e-02
8.59631002e-01 -5.22006333e-01 -7.70386577e-01 1.80659100e-01
3.66032988e-01 -8.58664393e-01 8.60850036e-01 -8.53357315e-01
5.04198551e-01 -2.33590782e-01 1.76965535e-01 -1.97692668e+00
-8.66915166e-01 -6.22477353e-01 -2.55555451e-01 8.73569429e-01
4.64905590e-01 -9.19578314e-01 5.28528273e-01 3.39171648e-01
-3.84861171e-01 -9.93988335e-01 -8.35420430e-01 -7.77135372e-01
2.44730130e-01 9.18338373e-02 7.39460111e-01 7.28324533e-01
1.84210360e-01 6.66318059e-01 -5.93148649e-01 1.08463973e-01
2.34101459e-01 2.59712994e-01 7.22360492e-01 -1.01684713e+00
-9.28926528e-01 -6.24293447e-01 -7.55386753e-03 -1.28376245e+00
2.82253593e-01 -7.21898437e-01 4.35648952e-03 -1.55862379e+00
2.06242099e-01 -6.47991061e-01 -7.28988290e-01 6.87181354e-01
-4.95750874e-01 -1.33143529e-01 6.30787015e-01 4.31979932e-02
-8.90163541e-01 9.31951225e-01 1.60866833e+00 -1.35099003e-02
-3.88465315e-01 2.24414878e-02 -8.65782082e-01 5.82741916e-01
1.04388905e+00 -3.36407781e-01 -7.62076259e-01 -7.76484907e-01
7.60783330e-02 -3.25266086e-03 1.75023213e-01 -9.96214807e-01
2.04224303e-01 -3.99850339e-01 1.68398604e-01 1.12585470e-01
2.89652497e-01 -8.75903070e-01 -1.74416497e-01 5.84617674e-01
-2.16201887e-01 5.51052213e-01 3.76529455e-01 5.98110795e-01
-5.09704165e-02 -3.59618425e-01 6.24306738e-01 -3.55367184e-01
-8.99201572e-01 2.68265396e-01 -5.41455090e-01 -2.75077559e-02
1.28941441e+00 8.09099749e-02 -6.71532273e-01 -2.32922882e-01
-6.33858562e-01 6.13642812e-01 2.08076656e-01 7.00674415e-01
5.16386509e-01 -1.22581351e+00 -7.67414033e-01 6.04838319e-02
-3.93378325e-02 -8.88170972e-02 4.35541660e-01 5.70302904e-01
-1.07935697e-01 2.16597140e-01 -8.24545383e-01 -3.14715952e-01
-9.88222957e-01 7.90915787e-01 3.87710780e-01 -9.51554418e-01
-7.19163954e-01 3.50961030e-01 3.88625234e-01 -5.44214129e-01
8.00887942e-02 -2.07579643e-01 -4.07506436e-01 -5.48292175e-02
5.09064615e-01 1.20467708e-01 -1.37092978e-01 -7.46608227e-02
-3.18074115e-02 1.92352578e-01 -5.49620509e-01 -1.84244886e-01
1.61736631e+00 -2.03204453e-02 2.28008673e-01 3.06600362e-01
4.63909596e-01 -3.56235981e-01 -1.76941574e+00 -2.34147191e-01
-2.48500094e-01 -1.02708757e-01 -8.20671394e-02 -1.04617393e+00
-9.95120645e-01 4.51458722e-01 3.29222649e-01 3.50657940e-01
1.02502346e+00 -5.09365946e-02 5.55035472e-01 4.41956997e-01
6.12005651e-01 -1.21543872e+00 5.41945875e-01 8.53073776e-01
9.81934130e-01 -1.27593327e+00 -2.34231830e-01 1.58219844e-01
-1.14892423e+00 8.78010333e-01 1.18388617e+00 -3.36226046e-01
1.77558213e-01 1.88052356e-01 1.26660481e-01 -1.02156624e-01
-1.15139532e+00 -4.73134696e-01 -1.22051220e-03 7.27853000e-01
1.55721322e-01 9.23257172e-02 -2.09647462e-01 5.47858715e-01
-2.09260229e-02 -1.26884073e-01 5.07239044e-01 9.21781898e-01
-5.03452420e-01 -1.37722278e+00 3.96223739e-02 2.88852185e-01
-2.22949281e-01 -1.61004364e-02 1.84584811e-01 1.11436033e+00
3.12198866e-02 8.34088504e-01 8.04613158e-02 -2.85089135e-01
3.16486686e-01 -1.05460413e-01 7.58085787e-01 -7.07809150e-01
-9.83222485e-01 8.82021636e-02 -1.29466385e-01 -4.09342289e-01
-3.81325454e-01 -3.60317439e-01 -1.38302886e+00 -1.58375442e-01
-1.88981831e-01 2.01182440e-01 3.67811084e-01 1.00840914e+00
5.38482845e-01 1.09748685e+00 5.57010829e-01 -8.99939418e-01
-1.03285229e+00 -1.14997756e+00 -6.63604021e-01 3.60386580e-01
3.72707665e-01 -9.53937948e-01 9.36839357e-02 -4.09419298e-01] | [3.860304594039917, 1.7740404605865479] |
a5893ef2-58f6-47c7-bc5d-0e6a5d276386 | comprehensive-privacy-analysis-on-federated | 2205.11857 | null | https://arxiv.org/abs/2205.11857v2 | https://arxiv.org/pdf/2205.11857v2.pdf | Comprehensive Privacy Analysis on Federated Recommender System against Attribute Inference Attacks | In recent years, recommender systems are crucially important for the delivery of personalized services that satisfy users' preferences. With personalized recommendation services, users can enjoy a variety of recommendations such as movies, books, ads, restaurants, and more. Despite the great benefits, personalized recommendations typically require the collection of personal data for user modelling and analysis, which can make users susceptible to attribute inference attacks. Specifically, the vulnerability of existing centralized recommenders under attribute inference attacks leaves malicious attackers a backdoor to infer users' private attributes, as the systems remember information of their training data (i.e., interaction data and side information). An emerging practice is to implement recommender systems in the federated setting, which enables all user devices to collaboratively learn a shared global recommender while keeping all the training data on device. However, the privacy issues in federated recommender systems have been rarely explored. In this paper, we first design a novel attribute inference attacker to perform a comprehensive privacy analysis of the state-of-the-art federated recommender models. The experimental results show that the vulnerability of each model component against attribute inference attack is varied, highlighting the need for new defense approaches. Therefore, we propose a novel adaptive privacy-preserving approach to protect users' sensitive data in the presence of attribute inference attacks and meanwhile maximize the recommendation accuracy. Extensive experimental results on two real-world datasets validate the superior performance of our model on both recommendation effectiveness and resistance to inference attacks. | ['Hongzhi Yin', 'Wei Yuan', 'Shijie Zhang'] | 2022-05-24 | null | null | null | null | ['inference-attack'] | ['adversarial'] | [-1.20397275e-02 -1.06793620e-01 -4.61780876e-01 -5.80002785e-01
-3.02894831e-01 -1.22751772e+00 3.13163429e-01 -5.10941111e-02
1.37693286e-01 3.94809127e-01 1.79776445e-01 -4.69902426e-01
-3.39861959e-01 -1.01051307e+00 -5.17891586e-01 -7.49464035e-01
-1.12938480e-02 -3.29504609e-02 -8.81275088e-02 -2.40149468e-01
7.28721637e-03 4.61977988e-01 -1.36956966e+00 3.55755985e-01
7.97417760e-01 1.25898850e+00 -3.94996494e-01 1.67134583e-01
2.54934490e-01 2.71266162e-01 -4.12526906e-01 -1.09608448e+00
5.69316268e-01 -2.47981790e-02 -3.31360042e-01 -4.00879472e-01
1.18037701e-01 -8.00874412e-01 -5.42275608e-01 1.21650088e+00
4.91627008e-01 1.13514215e-01 7.72530213e-02 -1.51095176e+00
-8.01802099e-01 8.83548617e-01 -3.06940172e-02 -2.76707917e-01
4.37590212e-01 -1.64763182e-01 9.87388670e-01 -2.69945920e-01
2.46817276e-01 6.82070851e-01 3.63247991e-01 6.06190801e-01
-9.97056127e-01 -1.02545416e+00 4.95880276e-01 1.50420396e-02
-1.48952115e+00 -4.54790443e-01 4.41980600e-01 -6.21056445e-02
2.84215987e-01 1.13694692e+00 3.76244456e-01 1.29041708e+00
-1.26837539e-02 5.87226987e-01 7.61495948e-01 3.33259910e-01
4.55971867e-01 7.92561173e-01 2.69253194e-01 1.80583090e-01
6.50312841e-01 1.79438964e-01 -2.47879237e-01 -1.04151523e+00
2.95782506e-01 7.38575578e-01 -5.30938208e-01 -5.77410161e-01
-6.99066699e-01 5.86794496e-01 8.30163434e-02 -1.19504146e-01
-2.61404932e-01 -4.56114739e-01 3.88989359e-01 4.03999209e-01
1.02712229e-01 2.16598645e-01 -7.07993627e-01 2.01737896e-01
-2.89142847e-01 1.59966692e-01 1.15618503e+00 1.20108497e+00
4.58547831e-01 -3.11480045e-01 7.48953223e-02 2.94527948e-01
4.73168701e-01 5.87840736e-01 3.15154403e-01 -6.47293031e-01
2.55923599e-01 5.72446465e-01 3.03784013e-01 -1.24418354e+00
2.64689922e-01 -4.74119246e-01 -9.03134406e-01 -3.41636986e-01
2.66659796e-01 -4.69989002e-01 1.30284131e-01 1.65922379e+00
5.41438341e-01 4.33114380e-01 1.28153935e-01 9.03342009e-01
3.43765944e-01 3.63181144e-01 2.12267451e-02 -2.97329724e-01
1.45976400e+00 -4.53758717e-01 -7.05394864e-01 4.28055137e-01
4.49226141e-01 -3.10663193e-01 8.90647531e-01 5.23942828e-01
-7.31663525e-01 -1.11048505e-01 -1.03558898e+00 3.36588860e-01
-4.16822165e-01 -1.00271724e-01 9.48214471e-01 1.43127441e+00
-3.11922342e-01 4.69450444e-01 -6.77554965e-01 -1.36030704e-01
4.64335442e-01 7.82412171e-01 -3.11207235e-01 1.38224795e-01
-1.32667303e+00 5.75303100e-02 -2.00856879e-01 -1.47526830e-01
-6.81518376e-01 -9.09105420e-01 -3.53716195e-01 4.94228184e-01
5.68669438e-01 -6.83607161e-01 1.13402021e+00 -3.77752393e-01
-1.53213072e+00 2.31838956e-01 3.16834778e-01 -4.22656596e-01
3.66544604e-01 -3.16357225e-01 -1.05203509e+00 -2.86841184e-01
-4.71743107e-01 -7.11635292e-01 8.29243004e-01 -1.13410318e+00
-8.34553123e-01 -8.29459190e-01 6.33552372e-01 -5.40534668e-02
-9.33270752e-01 9.11392570e-02 -1.71844661e-01 -4.59105462e-01
-2.03521386e-01 -1.10322726e+00 -2.14563876e-01 -1.98731199e-01
-4.92302328e-01 2.42674485e-01 1.17977321e+00 -3.09718788e-01
1.45375979e+00 -2.49280119e+00 -2.97652692e-01 7.22881436e-01
2.69859403e-01 4.15400177e-01 4.19468880e-01 5.06694317e-01
2.77777165e-01 3.57013106e-01 2.46547788e-01 -1.94644213e-01
2.99225748e-01 1.49844617e-01 -9.24338341e-01 6.22234643e-01
-8.98258924e-01 5.38287461e-01 -5.81112444e-01 1.57389775e-01
1.65508747e-01 6.47661746e-01 -8.00612152e-01 3.22656989e-01
-1.36975750e-01 4.11511600e-01 -1.04920805e+00 5.16920507e-01
9.40844297e-01 -1.39215693e-01 7.75822520e-01 -2.46717706e-01
1.73644014e-02 3.24920923e-01 -1.19270802e+00 1.29972601e+00
-4.09691930e-01 -2.81686515e-01 5.32159865e-01 -3.57610822e-01
7.43669510e-01 4.56988573e-01 4.59041685e-01 -2.59875476e-01
3.54015142e-01 -1.69104934e-01 -8.46045017e-02 -1.19939007e-01
4.53412533e-01 5.24746060e-01 -3.72122407e-01 9.24729705e-01
-4.93909061e-01 8.98838878e-01 -7.17903018e-01 4.33738679e-01
7.92953134e-01 -2.79053003e-01 2.10710734e-01 -1.34356305e-01
7.88308024e-01 -7.91610658e-01 6.64313853e-01 8.53228211e-01
-2.23452561e-02 8.90489072e-02 3.59672979e-02 -3.55236351e-01
-3.50042194e-01 -9.61080968e-01 -6.63115904e-02 1.26356328e+00
4.75613654e-01 -9.83878434e-01 -6.49042070e-01 -1.11480224e+00
2.84067303e-01 8.70576739e-01 -5.17555118e-01 -4.72779572e-01
-3.03426504e-01 -4.59692478e-01 5.97335875e-01 2.69422114e-01
4.07693595e-01 -3.14445734e-01 -1.76744044e-01 -8.15762654e-02
-1.24822482e-01 -8.65443349e-01 -8.97269785e-01 -4.20231134e-01
-4.84793544e-01 -9.57216918e-01 1.93631247e-01 -1.68017238e-01
7.25754738e-01 7.26167321e-01 3.84456724e-01 3.43310088e-01
2.53294796e-01 5.08172631e-01 -2.22264946e-01 -9.37830359e-02
-2.63283253e-01 2.16345936e-01 5.13914466e-01 7.38217294e-01
4.88084078e-01 -9.25582409e-01 -9.11887884e-01 8.10309291e-01
-7.39610374e-01 -5.93198240e-01 6.40904754e-02 3.55017424e-01
4.69033957e-01 1.80436581e-01 5.96801758e-01 -1.40661848e+00
7.52915502e-01 -8.03104699e-01 -6.51523530e-01 4.45872635e-01
-8.67875338e-01 -3.79955292e-01 1.29102433e+00 -6.99194252e-01
-1.07366145e+00 -1.15057260e-01 -1.36142775e-01 -1.29506275e-01
-1.65132567e-01 5.38380966e-02 -1.12289155e+00 -3.57074469e-01
3.45818520e-01 2.07155809e-01 -1.24560699e-01 -6.82728112e-01
6.52726531e-01 1.06607771e+00 3.64152461e-01 -5.56590199e-01
7.09489942e-01 6.79089189e-01 -2.42861003e-01 -4.50898051e-01
-5.51159561e-01 -3.65115523e-01 -1.44167719e-02 7.49004707e-02
2.30303288e-01 -6.94204628e-01 -1.48144412e+00 1.79677412e-01
-5.35321772e-01 4.48901772e-01 -8.45156536e-02 3.97359699e-01
-2.63356447e-01 6.30502999e-01 -5.44705689e-01 -7.16011822e-01
-7.52499044e-01 -1.12100828e+00 3.25738311e-01 2.81779021e-01
-1.09316528e-01 -7.45123625e-01 -2.66232580e-01 4.59587783e-01
6.85631514e-01 -1.03269085e-01 6.81756496e-01 -1.37216365e+00
-7.08900273e-01 -5.80378234e-01 1.75479844e-01 6.63641319e-02
3.92523050e-01 -2.81621337e-01 -9.21733260e-01 -5.85776508e-01
3.63728344e-01 2.98931301e-01 -4.72797677e-02 -1.69501901e-01
1.65170932e+00 -1.19458556e+00 -5.18029392e-01 9.48614180e-01
1.04277134e+00 2.50059158e-01 6.84442043e-01 -6.67887181e-02
6.45937204e-01 2.55294949e-01 6.90350473e-01 9.17996585e-01
3.10700893e-01 7.15392530e-01 6.30363464e-01 5.06624281e-01
7.25719571e-01 -6.16853178e-01 1.45764455e-01 3.37044775e-01
1.01784199e-01 -3.97425443e-01 1.11899465e-01 1.71129254e-03
-1.80925667e+00 -1.04370892e+00 1.07206702e-01 2.83088756e+00
4.61297601e-01 -1.75929695e-01 2.14220852e-01 -6.80149347e-02
5.78911901e-01 -9.46283489e-02 -7.98745990e-01 -3.30939591e-01
2.12168545e-01 -1.23961806e-01 8.47902775e-01 2.89817676e-02
-9.24274802e-01 5.75695753e-01 5.08214283e+00 3.32257092e-01
-9.59137261e-01 2.02291548e-01 3.72437030e-01 -2.98249424e-01
-5.94707727e-01 2.29134932e-01 -8.09808552e-01 7.95862973e-01
9.66897309e-01 -5.18707871e-01 7.81529903e-01 1.20947373e+00
-7.77896196e-02 5.96960247e-01 -1.03754795e+00 8.32158566e-01
-1.47389516e-01 -1.22200537e+00 -2.48597283e-03 6.53660655e-01
4.38693613e-01 -3.38497311e-01 3.87706399e-01 1.67128325e-01
4.85659212e-01 -5.10363996e-01 1.10139735e-01 4.67295200e-01
6.08913422e-01 -1.22287107e+00 3.35573107e-01 4.87430125e-01
-8.43764365e-01 -5.00571072e-01 -5.22581577e-01 4.35704067e-02
5.88150024e-02 2.84580112e-01 -4.13092673e-01 6.78995609e-01
6.80837631e-01 3.39281827e-01 -2.78952658e-01 6.64690673e-01
-1.70661286e-02 7.16675937e-01 -3.97523254e-01 -7.53947124e-02
-4.38455045e-01 -5.54696798e-01 3.74303877e-01 5.16977191e-01
3.73314023e-01 5.74959934e-01 5.89431003e-02 4.56930995e-01
-4.14088428e-01 3.42527270e-01 -5.71393609e-01 -3.19329882e-03
1.05946541e+00 1.43603575e+00 -3.02675664e-02 9.52628534e-03
-4.02898818e-01 1.01657033e+00 2.71505583e-02 2.04835460e-01
-8.11713398e-01 -1.32938877e-01 1.47005928e+00 3.40284318e-01
1.97480336e-01 -1.31717287e-02 -4.93590124e-02 -1.31066787e+00
-1.54212505e-01 -1.22831035e+00 7.27436244e-01 -5.57776503e-02
-1.35300612e+00 3.19042772e-01 -4.06797200e-01 -1.16573119e+00
-2.00600266e-01 5.87916821e-02 -4.57645953e-01 5.90898097e-01
-8.23360562e-01 -1.15442657e+00 1.35364616e-03 1.12736750e+00
-3.34791481e-01 -3.67601246e-01 1.24145699e+00 5.31526625e-01
-6.14771724e-01 1.54550529e+00 4.84385043e-01 7.52128065e-02
6.49365664e-01 -5.80490291e-01 3.71549755e-01 7.13565648e-01
1.86077759e-01 1.48455191e+00 4.28191513e-01 -6.42507613e-01
-2.13996506e+00 -9.97675478e-01 3.21739286e-01 -5.53836703e-01
4.07555282e-01 -7.55151987e-01 -9.19853210e-01 9.66963649e-01
-2.35350475e-01 1.30031735e-01 1.57926476e+00 2.11339146e-01
-7.60131896e-01 -4.15151209e-01 -1.62861371e+00 7.40935922e-01
9.71888244e-01 -7.66329408e-01 1.28007546e-01 1.52900591e-01
8.79043818e-01 1.65227952e-03 -1.16555750e+00 1.71588689e-01
1.05087769e+00 -9.81151700e-01 9.24593210e-01 -1.03047168e+00
-5.48026502e-01 -3.68390799e-01 -4.00261462e-01 -8.61061871e-01
-3.03542435e-01 -1.17595530e+00 -7.45468020e-01 1.65475833e+00
1.62444502e-01 -1.05046308e+00 9.15741682e-01 1.33353543e+00
4.14262652e-01 -3.88288558e-01 -5.52898705e-01 -4.80679780e-01
-4.35254514e-01 -2.41626620e-01 1.54139638e+00 1.13493538e+00
1.91748738e-01 -3.39418389e-02 -8.17349315e-01 8.90992224e-01
6.59682453e-01 3.57018858e-01 9.96603429e-01 -1.27665603e+00
-5.41432023e-01 1.48226559e-01 -1.99563101e-01 -9.87221539e-01
4.13217442e-03 -6.94347143e-01 -8.08387995e-01 -7.71490037e-01
-1.42726853e-01 -8.70665252e-01 -5.53741157e-01 3.17081571e-01
2.89754540e-01 -1.54363945e-01 8.37324560e-02 3.14971358e-01
-6.23933017e-01 3.13939095e-01 7.17752457e-01 1.38095334e-01
-2.55480081e-01 9.36846197e-01 -1.51820898e+00 4.28979427e-01
1.00187087e+00 -4.24929559e-01 -9.85186160e-01 3.57260481e-02
2.82587349e-01 6.14031963e-02 -6.94518089e-02 -5.58771193e-01
4.02631879e-01 -3.22389245e-01 -2.15366483e-02 -2.78915703e-01
1.65324166e-01 -1.55664885e+00 5.78293860e-01 1.00596629e-01
-2.61326551e-01 -2.17309847e-01 -3.49414676e-01 7.58060336e-01
4.45000589e-01 2.19844595e-01 3.08659792e-01 1.88033223e-01
-8.07479993e-02 8.49078655e-01 -6.30178303e-02 -3.80364984e-01
1.17994881e+00 -2.98471004e-02 -5.13566136e-01 -6.18090987e-01
-6.30689621e-01 2.15092435e-01 8.75390112e-01 6.15875959e-01
4.25985664e-01 -9.75396752e-01 -2.68403381e-01 6.66023672e-01
1.51407095e-02 -5.24176419e-01 5.95721960e-01 4.15237784e-01
3.61235678e-01 2.38148365e-02 -2.18771379e-02 1.50919199e-01
-1.45193720e+00 1.16466272e+00 1.68526620e-01 2.29875267e-01
-6.12388372e-01 3.73770386e-01 2.13737294e-01 -4.51056004e-01
4.64463562e-01 2.69330174e-01 -4.59066546e-03 -2.42154494e-01
1.10893559e+00 4.75880355e-01 9.88274142e-02 -4.76473957e-01
-2.96129614e-01 -2.19746277e-01 -4.97369796e-01 3.71763498e-01
8.69230807e-01 -6.21307969e-01 -2.51488201e-02 -3.83906573e-01
1.11449766e+00 8.19440722e-01 -9.30213630e-01 -2.81017900e-01
-5.58262646e-01 -1.04319179e+00 -6.39431551e-02 -8.45970213e-01
-1.41742516e+00 3.40007424e-01 4.92011160e-01 5.01572609e-01
1.07970512e+00 -3.83198768e-01 1.14749670e+00 2.26547912e-01
7.78604567e-01 -6.37837946e-01 -7.27265239e-01 -1.58314437e-01
1.57613277e-01 -6.65144682e-01 7.32240453e-02 -7.25208342e-01
-6.18886530e-01 5.50409734e-01 5.39382696e-01 1.39031097e-01
1.07513809e+00 2.75391519e-01 -1.42851755e-01 1.44045785e-01
-6.54154420e-01 5.64918041e-01 -8.69302601e-02 7.16649950e-01
-3.53580229e-02 3.57677162e-01 -2.44823501e-01 1.73106289e+00
-1.98274717e-01 -1.62451670e-01 5.63752651e-01 7.83870161e-01
1.25438673e-02 -1.56231427e+00 2.63763983e-02 5.51436067e-01
-9.32824910e-01 1.38887227e-01 -3.87518346e-01 1.29913092e-01
-1.49457082e-01 1.16747224e+00 -3.61660510e-01 -8.17366183e-01
3.28523964e-01 -1.17850609e-01 -1.83323741e-01 -2.19851553e-01
-1.02848518e+00 -3.23321193e-01 7.79447630e-02 -7.05337703e-01
1.66275233e-01 -6.40999198e-01 -1.08606648e+00 -7.38289833e-01
-5.07761478e-01 5.90762794e-01 7.39422739e-01 5.38625658e-01
9.48170602e-01 -5.21450080e-02 1.45136523e+00 3.26124877e-02
-1.06424570e+00 -9.21357349e-02 -8.93214762e-01 3.50095063e-01
5.25403172e-02 -3.93218488e-01 -4.01438892e-01 -4.24663633e-01] | [5.902075290679932, 6.760597229003906] |
532c1619-0dc9-4076-9063-68e092bec0fa | 3d-pictorial-structures-revisited-multiple | null | null | https://ieeexplore.ieee.org/document/7360209/authors#authors | http://campar.in.tum.de/pub/belagiannis2016pami/belagiannis2016pami.pdf | 3D Pictorial Structures Revisited: Multiple Human Pose Estimation | We address the problem of 3D pose estimation of multiple humans from multiple views. The transition from single to multiple human pose estimation and from the 2D to 3D space is challenging due to a much larger state space, occlusions and across-view ambiguities when not knowing the identity of the humans in advance. To address these problems, we first create a reduced state space by triangulation of corresponding pairs of body parts obtained by part detectors for each camera view. In order to resolve ambiguities of wrong and mixed parts of multiple humans after triangulation and also those coming from false positive detections, we introduce a 3D pictorial structures (3DPS) model. Our model builds on multi-view unary potentials, while a prior model is integrated into pairwise and ternary potential functions. To balance the potentials' influence, the model parameters are learnt using a Structured SVM (SSVM). The model is generic and applicable to both single and multiple human pose estimation. To evaluate our model on single and multiple human pose estimation, we rely on four different datasets. We first analyse the contribution of the potentials and then compare our results with related work where we demonstrate superior performance. | ['Slobodan Ilic', 'Mykhaylo Andriluka', 'Vasileios Belagiannis', 'Nassir Navab', 'Sikandar Amin', 'Bernt Schiele'] | 2016-10-01 | null | null | null | null | ['3d-multi-person-pose-estimation'] | ['computer-vision'] | [ 2.60486484e-01 1.81198403e-01 1.81155831e-01 -2.22110674e-01
-4.81679499e-01 -5.74453235e-01 6.30333483e-01 -9.14750472e-02
-6.49638295e-01 7.43050218e-01 3.84177850e-03 4.49437708e-01
9.19912979e-02 -1.67515725e-01 -8.26988816e-01 -4.42628741e-01
-3.56283225e-02 1.07973468e+00 6.16499126e-01 -7.78918192e-02
-3.53236161e-02 4.16445702e-01 -1.52063477e+00 1.36772126e-01
3.70523930e-01 7.56060660e-01 7.53523335e-02 6.68378770e-01
5.47588646e-01 1.27421588e-01 -6.64154053e-01 -3.80545318e-01
6.66216910e-01 -1.56262591e-01 -6.90301120e-01 5.64757288e-01
7.90934861e-01 -2.66690046e-01 -1.92266405e-02 9.40568864e-01
5.00291169e-01 1.03557348e-01 6.36824310e-01 -1.60828960e+00
3.15340698e-01 -1.14171572e-01 -7.82908380e-01 7.83696547e-02
9.15978968e-01 1.73592567e-01 7.77907848e-01 -8.83662462e-01
8.79755199e-01 1.58230841e+00 8.63415301e-01 5.27442038e-01
-1.36219728e+00 -1.91193014e-01 4.54736650e-01 1.42391130e-01
-1.20545340e+00 -3.84671241e-01 6.01276696e-01 -7.29816020e-01
8.91112983e-01 1.74046293e-01 1.03984785e+00 1.42551112e+00
3.56134772e-01 8.35146248e-01 1.18612921e+00 -2.96354473e-01
2.51923432e-03 1.90733507e-01 9.44141150e-02 7.63188541e-01
4.89357799e-01 1.14078462e-01 -6.00037158e-01 -2.39494935e-01
8.89023006e-01 -1.52027979e-01 -9.21195820e-02 -1.33251202e+00
-1.37398350e+00 5.54259956e-01 2.69155830e-01 -2.85335362e-01
-4.42529082e-01 -1.61897123e-01 1.74963534e-01 -1.29714295e-01
4.51806374e-02 2.76598632e-01 -5.22671163e-01 1.65279746e-01
-7.72775888e-01 6.03466690e-01 7.51409471e-01 9.35027957e-01
4.70453054e-01 -3.64541829e-01 -1.14740375e-02 6.10110819e-01
4.84951615e-01 5.74981332e-01 9.13594812e-02 -9.02341962e-01
7.32775867e-01 6.21577561e-01 4.68411922e-01 -9.31321561e-01
-6.70484781e-01 -2.54317969e-01 -6.80146635e-01 2.94807523e-01
6.27067864e-01 -9.17214826e-02 -1.00968039e+00 1.76735771e+00
8.14945340e-01 -3.18425208e-01 -2.28414670e-01 1.19280398e+00
4.45132077e-01 2.92461634e-01 -1.59421086e-01 -1.14466876e-01
1.54081202e+00 -1.05965316e+00 -6.10178828e-01 -8.51494193e-01
2.09090561e-02 -4.02244747e-01 2.79715508e-01 5.41004717e-01
-1.21204400e+00 -7.27804303e-01 -1.07658482e+00 5.19699641e-02
-2.95393676e-01 2.87282079e-01 1.87301889e-01 4.42646980e-01
-7.97004640e-01 5.64644039e-01 -1.04206038e+00 -6.01331234e-01
7.39296600e-02 6.86864495e-01 -8.49071145e-01 1.49045855e-01
-1.10369098e+00 1.61342776e+00 3.49057198e-01 4.63919401e-01
-5.60356438e-01 -3.67716178e-02 -1.22972810e+00 -4.23308283e-01
7.75184572e-01 -1.26419413e+00 1.03325260e+00 -6.38170063e-01
-1.19094968e+00 9.98534381e-01 -4.63759340e-02 -3.31198007e-01
9.34859335e-01 -5.48542023e-01 6.69122189e-02 1.79944128e-01
1.41512901e-01 7.34309614e-01 8.75540018e-01 -1.45654404e+00
-4.08921778e-01 -8.18847597e-01 -8.77855048e-02 7.49949753e-01
5.64752400e-01 -1.83083579e-01 -6.53974056e-01 -3.06840122e-01
4.40718293e-01 -1.35020149e+00 -4.16517615e-01 1.23377092e-01
-5.52124262e-01 2.63998508e-01 4.48802054e-01 -8.98037434e-01
7.60838568e-01 -1.66942787e+00 9.85062599e-01 1.79006487e-01
1.70658320e-01 1.44753948e-01 1.18830606e-01 2.92770803e-01
2.57485397e-02 -4.96445864e-01 -1.85801968e-01 -5.92463732e-01
1.80331618e-02 4.56356466e-01 2.69886702e-01 7.38536179e-01
3.05250466e-01 7.13517725e-01 -5.84402740e-01 -6.55316114e-01
3.61591995e-01 4.96049494e-01 -3.91990185e-01 2.38869488e-01
1.35789230e-01 6.30850792e-01 -3.53671759e-01 5.41274965e-01
7.16249764e-01 -2.43460789e-01 3.52161855e-01 -3.06561381e-01
4.67382707e-02 5.12253493e-02 -1.87097049e+00 1.58992362e+00
-1.84591152e-02 -1.86314419e-01 3.68904144e-01 -7.61429131e-01
4.22708660e-01 3.38664860e-01 4.89391804e-01 -1.73072904e-01
1.78202435e-01 -7.41752237e-02 -5.16624860e-02 -2.40763053e-01
4.25394863e-01 -2.73601621e-01 -3.45852345e-01 5.99696226e-02
4.40887332e-01 -3.42030786e-02 1.48532405e-01 -2.40031388e-02
7.62792289e-01 6.42447114e-01 7.68964052e-01 -3.56828198e-02
5.74299216e-01 -1.22295424e-01 6.70529008e-01 5.25348127e-01
-5.80198824e-01 7.39107430e-01 4.41397041e-01 -5.41521430e-01
-8.24992239e-01 -1.32683575e+00 5.25212437e-02 6.63286269e-01
4.61489081e-01 -3.72945666e-01 -6.97828650e-01 -8.45925391e-01
1.98722214e-01 1.87401474e-01 -6.43836558e-01 -4.63540554e-02
-7.46174216e-01 -6.27341330e-01 2.19637714e-02 5.47797620e-01
1.90407410e-01 -7.38458455e-01 -1.24072850e+00 -1.10858986e-02
-5.06492972e-01 -1.33632386e+00 -4.27139610e-01 2.43963912e-01
-8.36714447e-01 -1.27797782e+00 -7.34420955e-01 -4.37798411e-01
6.55698001e-01 3.07023451e-02 1.10516894e+00 -4.35006708e-01
-4.34706360e-01 7.61250079e-01 -6.69377595e-02 -3.50772917e-01
-2.64818013e-01 -2.62228668e-01 6.85401261e-01 -5.40959975e-03
3.34561281e-02 -2.38625914e-01 -3.77378762e-01 5.69285929e-01
-2.72572696e-01 1.62454173e-01 6.28570080e-01 8.52055013e-01
6.31528318e-01 -3.33177447e-01 -2.17895865e-01 -5.50969899e-01
-5.10711269e-03 -7.49980146e-03 -7.37215340e-01 3.53446782e-01
-1.51171803e-01 8.79606977e-02 -1.12061277e-02 -5.49421489e-01
-8.69592071e-01 7.69640565e-01 1.80854514e-01 -4.60352510e-01
-4.03565705e-01 -2.13982686e-02 -2.53948450e-01 -4.58604023e-02
3.78807873e-01 -9.66798067e-02 2.36615792e-01 -4.64272290e-01
2.02891782e-01 1.19618885e-01 5.24967968e-01 -3.55608672e-01
8.93523157e-01 4.02828842e-01 3.33916426e-01 -8.26948583e-01
-8.54094923e-01 -6.41500354e-01 -1.45343089e+00 -4.54232872e-01
9.78191137e-01 -1.06594634e+00 -5.92061400e-01 5.17750442e-01
-1.35211957e+00 1.07002661e-01 -1.23372190e-01 5.22399068e-01
-6.97611094e-01 7.03859866e-01 -4.97394264e-01 -1.08831739e+00
5.50638437e-02 -1.32561195e+00 1.56154335e+00 -2.00481694e-02
-5.14036834e-01 -7.38487780e-01 2.97796637e-01 4.46786553e-01
-3.05856705e-01 5.03878415e-01 2.54368633e-01 -5.14120460e-01
-4.11929727e-01 -3.21022570e-01 2.70424962e-01 8.15367252e-02
-2.44958624e-01 -3.98726434e-01 -8.03874314e-01 -5.68419695e-01
1.14586666e-01 -2.13646650e-01 6.73120558e-01 4.67791080e-01
1.87650710e-01 -1.20177932e-01 -7.36206710e-01 2.81588912e-01
1.04138386e+00 -1.61494419e-01 2.10639626e-01 3.77555370e-01
8.00819218e-01 9.42220569e-01 7.60205209e-01 4.71053958e-01
6.39607728e-01 1.21158314e+00 6.20870173e-01 1.31549969e-01
1.10550240e-01 -3.34903777e-01 4.82953370e-01 4.00272340e-01
-3.08952481e-01 -6.25274330e-02 -8.63640666e-01 2.24337205e-01
-1.87716329e+00 -8.56736124e-01 -5.99298142e-02 2.50944972e+00
2.64601767e-01 2.17324242e-01 6.07979417e-01 1.25557572e-01
8.54765594e-01 -7.18901446e-03 -4.69088405e-01 1.88626692e-01
1.52079418e-01 -1.92673132e-01 3.97120863e-01 6.00462019e-01
-1.46862614e+00 7.55212307e-01 6.57083464e+00 3.31041291e-02
-5.44011652e-01 -9.30339321e-02 3.10759265e-02 -2.80744731e-01
3.85169029e-01 -7.34381517e-03 -1.14648068e+00 7.30568469e-02
2.87076145e-01 5.00334740e-01 1.12115778e-01 5.48684955e-01
-2.33789951e-01 -4.08951133e-01 -1.38115394e+00 1.17437541e+00
3.59294474e-01 -4.88922685e-01 -4.24076281e-02 2.48583630e-01
4.14660245e-01 -2.16168538e-01 -1.86564475e-01 2.77906656e-01
-1.44863371e-02 -6.30721450e-01 1.10862648e+00 5.61525643e-01
2.69173622e-01 -3.59245539e-01 7.48113811e-01 7.66699851e-01
-1.24648261e+00 -5.58882132e-02 -1.32034510e-01 -1.76540837e-01
5.22222698e-01 1.44617975e-01 -6.82522237e-01 7.32106924e-01
6.13628745e-01 6.18067622e-01 -6.95182741e-01 9.75741386e-01
-2.78164774e-01 -3.00412327e-01 -5.53565085e-01 2.82716304e-01
-2.40666196e-01 -2.33692303e-01 9.06610906e-01 7.80070901e-01
1.23486802e-01 3.81214321e-02 5.76137125e-01 5.98067224e-01
6.34505570e-01 -3.21269453e-01 -4.82803494e-01 4.82210577e-01
-3.17474753e-02 1.16554940e+00 -8.08047831e-01 -2.29258657e-01
-2.40062565e-01 1.44502008e+00 3.74702305e-01 1.45599425e-01
-9.81119931e-01 1.60845652e-01 4.74797964e-01 2.82492220e-01
4.60110873e-01 -4.22601104e-01 1.62320673e-01 -1.60067201e+00
5.36367536e-01 -8.45875263e-01 5.59583306e-01 -7.18949854e-01
-1.17429221e+00 5.02276123e-01 5.01878142e-01 -1.33015096e+00
-4.46879774e-01 -7.98901081e-01 -4.90705036e-02 7.45016515e-01
-8.74185741e-01 -1.24532557e+00 -1.49083938e-02 3.01295370e-01
4.04500127e-01 2.99506605e-01 6.60203636e-01 -2.79113371e-02
-4.71864134e-01 2.72401690e-01 -6.59073174e-01 -1.10669307e-01
7.93554127e-01 -1.37205100e+00 6.19965315e-01 8.64798009e-01
1.00712612e-01 5.87371826e-01 8.58605266e-01 -9.22639251e-01
-1.35220373e+00 -6.00673914e-01 8.56232524e-01 -1.25584555e+00
1.18692182e-01 -7.15539098e-01 -6.04735851e-01 9.70809221e-01
-2.92276800e-01 5.61903045e-02 2.84115911e-01 1.35631889e-01
-1.61382794e-01 3.14869940e-01 -1.15212619e+00 3.92787158e-01
1.14713848e+00 -9.60363299e-02 -1.00600266e+00 1.07798725e-01
1.42960623e-01 -8.19461584e-01 -7.34635055e-01 5.62689185e-01
9.09458518e-01 -1.15458632e+00 1.35607731e+00 -4.72996742e-01
-1.25578448e-01 -3.93539697e-01 -8.68860483e-02 -1.21369743e+00
-2.55567968e-01 -2.58792907e-01 -2.81100571e-01 5.44204891e-01
1.15893818e-01 -4.40548360e-01 7.58410633e-01 5.12244821e-01
2.65213400e-01 -5.03160000e-01 -1.14676166e+00 -7.95947075e-01
-3.05866957e-01 -1.28269717e-02 3.50326709e-02 5.08526266e-01
-9.43050906e-03 6.50745749e-01 -6.97276235e-01 6.22766912e-01
9.68856215e-01 9.61857587e-02 1.03062642e+00 -1.53150523e+00
-6.41588449e-01 -1.11654416e-01 -6.81550741e-01 -1.24663460e+00
-1.25763118e-01 -4.35777932e-01 3.15411031e-01 -1.38010931e+00
4.82476562e-01 2.36838683e-01 1.74737081e-01 2.67290860e-01
-3.06888103e-01 1.31389067e-01 5.19672930e-01 3.26823778e-02
-8.61702263e-01 4.95545000e-01 1.19117498e+00 1.29817322e-01
-3.59850787e-02 3.16971928e-01 -1.56926885e-01 1.07768512e+00
2.24661916e-01 -4.00912195e-01 -8.55753478e-03 -3.05571973e-01
1.52841151e-01 4.87667054e-01 8.59511077e-01 -1.22158015e+00
1.80993870e-01 1.25824496e-01 8.20029497e-01 -8.73188257e-01
9.85760868e-01 -8.57210696e-01 3.72765929e-01 6.81536853e-01
-2.76106130e-02 1.79546729e-01 9.80951414e-02 8.78797054e-01
-2.66564880e-02 -4.36224975e-02 8.23779404e-01 -5.30040383e-01
-5.26706696e-01 2.71241665e-01 -2.75607914e-01 1.27620250e-02
9.89246368e-01 -5.04411221e-01 3.93321782e-01 -2.61846125e-01
-1.25782812e+00 3.54311943e-01 6.47672474e-01 5.10336399e-01
4.71819878e-01 -1.22784424e+00 -3.12474042e-01 4.96131480e-01
1.78016797e-01 3.62448357e-02 2.95882910e-01 1.01505733e+00
-1.88120738e-01 4.36183363e-01 -5.08567154e-01 -9.14315403e-01
-1.67469811e+00 6.42816067e-01 3.77814323e-01 -4.56721187e-01
-5.00529587e-01 6.90255463e-01 2.00656623e-01 -8.41156244e-01
3.00878584e-01 -4.08332139e-01 -3.54265958e-01 5.25571518e-02
2.63695002e-01 6.46105826e-01 -1.76624477e-01 -1.32557750e+00
-6.98702812e-01 1.06135225e+00 2.09427878e-01 -3.30078512e-01
1.01061833e+00 -2.96024799e-01 1.68137074e-01 5.79273880e-01
8.15115094e-01 -1.25146165e-01 -1.47275269e+00 -8.85240063e-02
-1.79856569e-01 -3.37860256e-01 -6.05449140e-01 -7.31160760e-01
-4.51262534e-01 8.96636724e-01 6.81107163e-01 -2.27110237e-01
6.07929468e-01 4.45382893e-02 4.46126878e-01 2.70757169e-01
7.67516553e-01 -1.05883777e+00 1.18102983e-01 5.01036346e-01
9.36656713e-01 -1.37237704e+00 3.82449597e-01 -6.44434810e-01
-8.38413358e-01 8.66095304e-01 7.86590338e-01 -3.21281105e-02
4.34292585e-01 3.45407948e-02 8.98146536e-03 -2.11738542e-01
-5.38687468e-01 -2.48229697e-01 6.12168670e-01 6.53374910e-01
-4.78414223e-02 -2.99840700e-02 8.71099811e-03 5.14195204e-01
-1.27988353e-01 -3.12900841e-01 2.84457892e-01 1.03211939e+00
-2.49651968e-01 -1.20229399e+00 -8.38499606e-01 5.14380820e-02
-1.83424011e-01 5.47076762e-01 -6.25697136e-01 9.95096922e-01
3.55553865e-01 6.86303139e-01 -1.04035266e-01 -3.71336788e-01
8.61115158e-01 2.98093289e-01 9.95584726e-01 -6.67551398e-01
-4.02219802e-01 3.40701520e-01 1.00201994e-01 -8.12412024e-01
-6.19211197e-01 -1.18840909e+00 -7.78749108e-01 3.72282982e-01
-5.05280137e-01 -1.71027407e-01 4.27787870e-01 1.06200194e+00
7.10298568e-02 3.07788879e-01 6.92203874e-03 -1.55223095e+00
-9.18660045e-01 -8.15777361e-01 -6.04807138e-01 4.55079764e-01
3.19077313e-01 -1.26357162e+00 -2.85361260e-01 -6.44539073e-02] | [7.045414924621582, -0.9893955588340759] |
6cc269d3-295b-445d-9cd3-37b9551f36be | does-constituency-analysis-enhance-domain | 2112.02955 | null | https://arxiv.org/abs/2112.02955v1 | https://arxiv.org/pdf/2112.02955v1.pdf | Does constituency analysis enhance domain-specific pre-trained BERT models for relation extraction? | Recently many studies have been conducted on the topic of relation extraction. The DrugProt track at BioCreative VII provides a manually-annotated corpus for the purpose of the development and evaluation of relation extraction systems, in which interactions between chemicals and genes are studied. We describe the ensemble system that we used for our submission, which combines predictions of fine-tuned bioBERT, sciBERT and const-bioBERT models by majority voting. We specifically tested the contribution of syntactic information to relation extraction with BERT. We observed that adding constituentbased syntactic information to BERT improved precision, but decreased recall, since relations rarely seen in the train set were less likely to be predicted by BERT models in which the syntactic information is infused. Our code is available online [https://github.com/Maple177/drugprot-relation-extraction]. | ['Claire Nédellec', 'Pierre Zweigenbaum', 'Robert Bossy', 'Louise Deléger', 'Anfu Tang'] | 2021-11-25 | null | null | null | null | ['drugprot'] | ['natural-language-processing'] | [ 2.12139487e-01 5.96881151e-01 -4.30905163e-01 -4.20082927e-01
-5.62233210e-01 -7.37164974e-01 6.38376772e-01 8.33365560e-01
-9.41566974e-02 1.38094473e+00 3.59877646e-01 -6.58598721e-01
-2.53360808e-01 -7.25292027e-01 -6.69934332e-01 -5.63811123e-01
-1.32790981e-02 7.28612721e-01 2.00737447e-01 -2.74083406e-01
-2.48123571e-01 3.65885913e-01 -6.98693037e-01 7.07060814e-01
6.65448725e-01 3.32390934e-01 -1.72080174e-01 5.44741452e-01
1.57705545e-01 9.16150391e-01 -4.25112218e-01 -5.93657136e-01
-9.57336873e-02 -6.39069438e-01 -1.04503489e+00 -7.46495485e-01
-2.15768397e-01 3.71875912e-01 -3.07500780e-01 7.00787604e-01
5.06537616e-01 -3.60295355e-01 8.01110804e-01 -1.00378048e+00
-6.33888617e-02 1.12832642e+00 -2.18690634e-01 1.86488196e-01
6.96261108e-01 5.72911985e-02 1.18501699e+00 -6.41746461e-01
1.23439121e+00 8.53958189e-01 6.28278434e-01 5.29036701e-01
-1.26987398e+00 -7.48652339e-01 -3.56615126e-01 2.80415148e-01
-1.44413495e+00 -5.85978091e-01 2.34364331e-01 -5.44260621e-01
1.83137739e+00 5.67994058e-01 4.90619451e-01 8.70133400e-01
6.37267470e-01 4.64001656e-01 9.72564697e-01 -3.43626022e-01
3.92084569e-02 1.18443906e-01 3.01313430e-01 6.87630117e-01
5.63880324e-01 8.87807533e-02 -4.77975905e-01 -6.42974913e-01
-1.25942659e-02 -4.89869505e-01 -3.16445649e-01 2.75619477e-01
-7.75894761e-01 5.64736187e-01 2.97069937e-01 4.32517171e-01
-4.84241962e-01 -2.63891071e-01 4.22146529e-01 9.07586962e-02
7.29413033e-01 9.40014660e-01 -1.27030575e+00 1.62201747e-02
-5.06428301e-01 9.82807204e-02 1.36501670e+00 1.06577682e+00
2.26433203e-01 -7.50342131e-01 -1.12013690e-01 6.96364522e-01
2.52470702e-01 -2.47357771e-01 4.63606268e-01 -4.10529792e-01
3.12447608e-01 8.65478873e-01 -7.61622414e-02 -5.51479697e-01
-7.27841496e-01 -4.63757366e-02 -4.63282168e-01 -4.81580377e-01
3.81978810e-01 -5.12801647e-01 -8.84245396e-01 1.30909812e+00
4.45128858e-01 1.18296184e-02 2.82744020e-01 4.03911382e-01
1.47162509e+00 1.62053466e-01 7.85245121e-01 -3.85421783e-01
1.44999707e+00 -5.51536143e-01 -9.70476091e-01 2.11930439e-01
1.22103333e+00 -8.57107580e-01 -3.37857381e-02 4.01866615e-01
-7.62789547e-01 1.23140916e-01 -1.11348855e+00 -9.02413055e-02
-9.04185772e-01 -6.12675734e-02 9.28417742e-01 4.02817518e-01
-6.36263311e-01 1.03668344e+00 -9.05979276e-01 -5.64738572e-01
5.30767083e-01 6.86248183e-01 -7.77719080e-01 1.16845854e-01
-1.32907617e+00 1.37331653e+00 5.43332100e-01 -2.35838190e-01
-3.47146541e-01 -6.21953547e-01 -6.64076567e-01 -1.53875738e-01
3.92503977e-01 -6.40870214e-01 1.10689664e+00 -9.25173536e-02
-1.02015960e+00 7.63546526e-01 -3.50893676e-01 -4.97699052e-01
1.40585005e-02 -1.45344570e-01 -4.52864885e-01 -1.67606369e-01
-5.74169122e-02 4.00070459e-01 -4.55933273e-01 -6.92337990e-01
-5.12268484e-01 -4.44022566e-01 -1.47253454e-01 -2.71487236e-02
2.04307914e-01 2.85444528e-01 -1.62999943e-01 -1.24306820e-01
-2.48675168e-01 -7.68478394e-01 -3.02021235e-01 -6.58645093e-01
-9.48671103e-01 -5.45372307e-01 3.60571176e-01 -8.31988037e-01
1.16626382e+00 -1.54199970e+00 9.19657201e-02 2.69792467e-01
3.94601464e-01 4.03902113e-01 -8.11621360e-03 1.07147145e+00
-7.28058219e-01 6.04024053e-01 1.43133895e-02 3.61198127e-01
-3.91812921e-01 5.53524382e-02 1.34962365e-01 3.19924414e-01
4.81589258e-01 1.01278210e+00 -9.39682364e-01 -5.74058950e-01
-1.21528730e-02 5.12701929e-01 -3.04671407e-01 -2.40004137e-02
-5.09392083e-01 2.19725877e-01 -6.14647150e-01 8.84341955e-01
2.72544026e-01 -2.85552055e-01 8.58393431e-01 -4.15838271e-01
-1.68419499e-02 8.10121059e-01 -4.15094465e-01 1.09223306e+00
2.01000303e-01 2.34540164e-01 -4.76860255e-01 -5.98801017e-01
7.30089366e-01 7.51833260e-01 7.80279517e-01 -1.49730325e-01
3.30544949e-01 1.61393791e-01 4.58360314e-01 -4.88846451e-01
-9.04285014e-02 1.50564313e-01 2.99544990e-01 -2.88220365e-02
3.36934716e-01 -1.44258812e-01 5.02788544e-01 2.00840503e-01
1.75094867e+00 4.96155977e-01 1.07639503e+00 -1.53005272e-01
3.23479891e-01 5.83788574e-01 7.80870497e-01 1.40155703e-01
1.23618677e-01 5.34748510e-02 7.29226172e-01 -2.71648109e-01
-6.29205942e-01 -2.78520048e-01 -6.49352252e-01 5.26528418e-01
-3.39142829e-01 -1.17782962e+00 -5.57874322e-01 -1.11897814e+00
-2.36799736e-02 7.74557173e-01 -5.56514025e-01 9.90610383e-03
-2.36732170e-01 -1.42539418e+00 7.27476716e-01 -1.45007866e-02
-2.02031478e-01 -7.75753081e-01 1.02814972e-01 4.69530076e-01
-2.82189697e-02 -1.03901291e+00 2.96411179e-02 8.63438308e-01
-5.67131221e-01 -1.80380464e+00 7.59166703e-02 -3.25821191e-01
4.70148325e-01 -5.83704650e-01 9.42814589e-01 6.35018945e-02
-4.72033143e-01 -3.98981810e-01 -4.69593078e-01 -1.05536032e+00
-5.02887309e-01 1.33353680e-01 -2.96656340e-01 -8.58678639e-01
1.03088748e+00 -3.90523463e-01 -3.22225749e-01 2.49442145e-01
-4.40949023e-01 1.29683286e-01 4.20442909e-01 8.93976748e-01
7.22387969e-01 -9.88456756e-02 6.91659868e-01 -1.54253995e+00
5.39783955e-01 -8.48470926e-01 -3.33283365e-01 3.63737792e-01
-7.93881774e-01 -1.20589109e-02 2.02053472e-01 -1.31000295e-01
-5.92555523e-01 6.06660485e-01 -5.07210493e-01 5.42656720e-01
-2.83822566e-01 8.80706251e-01 -3.35841209e-01 7.09843240e-04
7.14317143e-01 -3.44049960e-01 -2.61651605e-01 -4.44455683e-01
2.18245789e-01 7.46900260e-01 -1.26823410e-01 -4.72751290e-01
5.05790450e-02 -1.91927329e-01 3.11639905e-01 -4.89352584e-01
-5.52221298e-01 -5.05297780e-01 -6.16292715e-01 3.92262787e-01
7.81582475e-01 -8.01258385e-01 -8.13856006e-01 5.33128381e-02
-1.13612771e+00 -3.14527154e-01 -9.73365381e-02 5.75691223e-01
-1.78447962e-01 1.51226101e-02 -9.10612047e-01 -4.12346810e-01
-7.46408761e-01 -1.06192708e+00 8.26659799e-01 -7.27746561e-02
-8.95909786e-01 -8.38503122e-01 3.75640780e-01 2.28095159e-01
-3.50996643e-01 3.75108272e-01 1.23550129e+00 -1.45576358e+00
-1.63765818e-01 -3.30574304e-01 5.40054869e-03 -2.46122524e-01
6.11438990e-01 2.81689733e-01 -9.29760039e-01 3.39257240e-01
-6.37502372e-01 -1.72442533e-02 6.73640132e-01 1.80987149e-01
7.89165497e-01 -4.29155052e-01 -1.20310926e+00 2.95940012e-01
1.31899810e+00 5.67898929e-01 6.94116712e-01 1.36862263e-01
6.23687804e-01 6.84466898e-01 6.03749216e-01 1.70875371e-01
1.92252934e-01 7.21601725e-01 -2.51786634e-02 -6.51520193e-02
-1.01420015e-01 -2.46920407e-01 -2.10306495e-01 3.07083815e-01
-4.18763936e-01 -5.97666800e-01 -1.11899745e+00 2.75821924e-01
-1.74705780e+00 -5.94950974e-01 -6.26762211e-01 1.61430061e+00
1.62545931e+00 1.42531535e-02 3.24001722e-02 3.22575308e-02
2.50211805e-01 -7.25172937e-01 -3.19606125e-01 -5.09777606e-01
5.36857918e-02 6.28521144e-01 6.24100327e-01 5.70989609e-01
-1.09698534e+00 1.08884609e+00 6.47608137e+00 8.17989409e-01
-5.35227239e-01 -7.00325742e-02 7.24767148e-01 1.34384677e-01
-2.18688734e-02 1.17931597e-01 -1.07735360e+00 2.71652102e-01
1.45145953e+00 -3.32138568e-01 -1.79893762e-01 4.00226086e-01
3.31561297e-01 -1.86792925e-01 -1.42646694e+00 1.57039046e-01
-4.43277866e-01 -1.46697462e+00 -2.68933684e-01 3.28535557e-01
4.33286577e-01 3.00730348e-01 -7.15484798e-01 2.25784481e-02
8.69680822e-01 -1.12735462e+00 -1.15693614e-01 5.11136889e-01
4.56801265e-01 -4.99949276e-01 1.17924619e+00 7.32867941e-02
-8.18377495e-01 3.01746309e-01 -2.01898545e-01 1.38434693e-01
-1.35647103e-01 1.09539831e+00 -1.67260313e+00 9.12558913e-01
3.02457541e-01 9.53454912e-01 -5.63526154e-01 1.03820550e+00
-5.64571619e-01 8.03566754e-01 -2.53647566e-01 -1.92752600e-01
-3.38250458e-01 7.79288411e-02 3.58733237e-01 1.38372707e+00
-6.78110942e-02 6.27691865e-01 2.16193143e-02 4.56721067e-01
-1.05714321e-01 5.11201560e-01 -6.63174033e-01 -3.91695350e-01
4.43407655e-01 1.23252201e+00 -5.86377263e-01 -3.65765393e-01
-2.73975402e-01 4.43844795e-01 2.15249538e-01 1.13605298e-01
-4.76330101e-01 -3.22832286e-01 5.31128585e-01 4.97781076e-02
-1.18920110e-01 3.55290741e-01 -3.77643079e-01 -7.67333448e-01
-5.91012597e-01 -9.45401013e-01 7.18243361e-01 -7.65512943e-01
-1.18702686e+00 7.00774550e-01 1.44438118e-01 -7.77618408e-01
-3.35465848e-01 -6.21688485e-01 -9.48790610e-02 9.04508471e-01
-7.69308686e-01 -1.11609650e+00 3.70412230e-01 -1.00184560e-01
2.29084909e-01 4.71609123e-02 1.50813806e+00 2.91890830e-01
-9.00186718e-01 4.48251635e-01 -9.84281600e-02 8.59289095e-02
9.69497144e-01 -1.20620191e+00 4.31185633e-01 5.10415025e-02
-6.08195364e-02 9.18224752e-01 8.71754408e-01 -1.30401087e+00
-9.94631290e-01 -1.11739421e+00 1.61347973e+00 -8.91662061e-01
7.47078061e-01 -2.53370911e-01 -6.86488926e-01 8.45582664e-01
2.66282976e-01 -3.53606880e-01 1.31461751e+00 3.21888477e-01
-2.28915796e-01 4.57300872e-01 -1.41635752e+00 3.29413891e-01
9.08728004e-01 -1.22290865e-01 -4.63901639e-01 8.79079640e-01
5.77834547e-01 -4.86457378e-01 -1.37515438e+00 6.09563470e-01
4.76794690e-01 -1.21228270e-01 7.76751220e-01 -9.81569350e-01
6.08659744e-01 -2.26178706e-01 1.37244061e-01 -1.34102547e+00
-3.38250756e-01 -3.64919841e-01 -1.96501464e-01 1.08205914e+00
1.44069457e+00 -6.64655983e-01 5.65809667e-01 6.98173761e-01
5.39091742e-03 -1.20847213e+00 -5.04622579e-01 -3.80766004e-01
-2.97997911e-02 7.93338642e-02 5.41928947e-01 1.07181072e+00
9.18248475e-01 7.13845074e-01 1.09203741e-01 -1.92759447e-02
7.02960491e-02 -2.85420746e-01 2.75399953e-01 -1.21285999e+00
-4.98917460e-01 -5.19520789e-02 -5.15248239e-01 -1.13736004e-01
-1.09682828e-01 -1.15619493e+00 -5.45110106e-02 -1.70100987e+00
5.23717880e-01 -4.31418531e-02 -1.56240135e-01 1.33306122e+00
-2.71001458e-01 2.42750287e-01 -4.09645110e-01 4.49851193e-02
-1.90387413e-01 -9.99295413e-02 9.95662391e-01 -1.54000238e-01
-2.26605371e-01 -1.43239379e-01 -8.72136056e-01 5.67248225e-01
9.34740543e-01 -7.91154444e-01 -2.06227213e-01 2.78097689e-01
3.31312150e-01 -4.00390364e-02 -1.77735001e-01 -3.35370243e-01
1.67172030e-01 -7.61854947e-02 6.71265125e-01 -5.52731872e-01
3.12565565e-01 -6.18416846e-01 7.81244576e-01 5.69760263e-01
-5.15465736e-01 -2.69358546e-01 5.87252498e-01 1.96651995e-01
4.14962918e-02 -2.07651958e-01 1.95525810e-01 -1.51337981e-01
2.41847057e-03 1.37412876e-01 -4.29308087e-01 -4.83490795e-01
9.48601186e-01 2.13244885e-01 -7.65956581e-01 -1.68306418e-02
-1.09817719e+00 9.55013782e-02 5.08321747e-02 4.30754945e-03
2.11888522e-01 -7.10560858e-01 -8.49193275e-01 -4.44819123e-01
2.23635226e-01 -3.52053165e-01 -2.91371107e-01 1.11500823e+00
-3.50721419e-01 7.34966099e-01 1.64127052e-01 -5.44244424e-02
-1.81658173e+00 4.77229863e-01 4.54348147e-01 -5.42952776e-01
-2.42622063e-01 8.81879866e-01 -9.29671153e-02 -4.26658064e-01
-1.42268181e-01 -1.62328824e-01 -5.21844149e-01 1.71928648e-02
4.10308838e-01 1.60020486e-01 5.11057436e-01 -3.13746572e-01
-6.73240423e-01 -4.47233692e-02 -3.95562410e-01 3.55253100e-01
1.61052930e+00 5.33371150e-01 -4.70010728e-01 8.02466199e-02
8.97425234e-01 8.64570122e-03 -1.73949823e-01 1.08801208e-01
4.91438478e-01 2.15385124e-01 -1.18371397e-01 -1.65936399e+00
-6.74891889e-01 1.39852479e-01 1.89522326e-01 -7.15982988e-02
6.79703057e-01 2.77297825e-01 2.09919602e-01 4.60197926e-01
3.24807763e-01 -5.70676684e-01 -9.22982275e-01 4.50060129e-01
8.15368652e-01 -9.34964299e-01 6.93120658e-01 -1.23764181e+00
-3.40139419e-01 7.76726782e-01 5.54361343e-01 2.32395262e-01
7.87330687e-01 6.89811826e-01 -2.20840558e-01 -6.43485248e-01
-1.26378310e+00 -2.20280513e-01 4.81710464e-01 5.60761094e-01
1.28128839e+00 3.21199507e-01 -1.17665827e+00 1.01047719e+00
-1.76973686e-01 2.80620694e-01 2.75934756e-01 9.92055595e-01
1.78836454e-02 -1.82671010e+00 1.14272557e-01 8.07860613e-01
-9.14281487e-01 -6.49933457e-01 -1.55885470e+00 7.68995702e-01
2.73801863e-01 1.14845848e+00 -6.55695200e-01 -6.50443733e-01
3.79752547e-01 4.30049866e-01 5.74224651e-01 -1.03143263e+00
-9.37000453e-01 3.41604948e-01 1.15818918e+00 -4.94709045e-01
-3.92802328e-01 -6.33178592e-01 -1.50811446e+00 -5.72475120e-02
-6.77073359e-01 6.74735069e-01 5.71828663e-01 9.84163344e-01
5.85482359e-01 8.18342507e-01 1.42553439e-02 -5.11845499e-02
1.95736349e-01 -1.37249494e+00 -3.40530068e-01 -2.12078035e-01
-3.21022779e-01 -3.46512318e-01 2.98512392e-02 3.02548647e-01] | [8.469704627990723, 8.7819242477417] |
e6c60932-d0bc-4ddf-8642-7188eb1e7c0b | noise-estimation-for-generative-diffusion | 2104.02600 | null | https://arxiv.org/abs/2104.02600v2 | https://arxiv.org/pdf/2104.02600v2.pdf | Noise Estimation for Generative Diffusion Models | Generative diffusion models have emerged as leading models in speech and image generation. However, in order to perform well with a small number of denoising steps, a costly tuning of the set of noise parameters is needed. In this work, we present a simple and versatile learning scheme that can step-by-step adjust those noise parameters, for any given number of steps, while the previous work needs to retune for each number separately. Furthermore, without modifying the weights of the diffusion model, we are able to significantly improve the synthesis results, for a small number of steps. Our approach comes at a negligible computation cost. | ['Lior Wolf', 'Eliya Nachmani', 'Robin San-Roman'] | 2021-04-06 | null | null | null | null | ['noise-estimation'] | ['medical'] | [ 9.16314274e-02 -2.05145534e-02 4.06266868e-01 -7.96986520e-02
-7.37536907e-01 -4.76564080e-01 7.14829266e-01 2.59091765e-01
-4.68044788e-01 6.48921609e-01 -2.19157830e-01 -2.33929724e-01
-9.69296917e-02 -8.89423370e-01 -3.76978964e-01 -8.42285872e-01
1.96191698e-01 3.88605058e-01 4.98860955e-01 -4.17650402e-01
2.17454925e-01 6.16384327e-01 -1.29635513e+00 1.58978142e-02
6.90869808e-01 7.46431589e-01 3.80930692e-01 9.98588979e-01
-2.71881744e-02 6.31293654e-01 -5.91658294e-01 -4.21927840e-01
1.09814011e-01 -8.91087651e-01 -4.78943259e-01 3.26108158e-01
7.34436885e-02 -1.25862092e-01 -2.65782267e-01 1.07718444e+00
8.15332234e-01 2.31746241e-01 6.51388824e-01 -5.73054254e-01
-2.97273219e-01 7.49187946e-01 -2.87251830e-01 1.98616594e-01
-1.31671444e-01 1.16599850e-01 5.84474623e-01 -8.50121081e-01
5.71618497e-01 1.00740445e+00 5.09895921e-01 6.51324451e-01
-1.41091645e+00 -5.39844871e-01 1.09904058e-01 6.27909154e-02
-1.44675303e+00 -7.49855340e-01 9.77684617e-01 -2.53143638e-01
7.99689651e-01 7.67616183e-02 7.30820954e-01 1.03723371e+00
-6.61400706e-02 2.50925750e-01 9.68428731e-01 -6.29951596e-01
4.97288972e-01 1.25908822e-01 -2.59620070e-01 7.65327454e-01
1.57811325e-02 -1.35763660e-01 -2.06869960e-01 -6.13274286e-03
9.17979538e-01 -3.72384936e-01 -9.59890559e-02 -1.99505582e-01
-7.76394069e-01 1.02857327e+00 2.89360106e-01 7.49556899e-01
-3.19592983e-01 4.34049249e-01 1.20549276e-01 4.58474874e-01
6.71976447e-01 3.15845251e-01 -1.42464325e-01 -2.34442115e-01
-1.11219954e+00 1.60349071e-01 9.00414169e-01 4.77090150e-01
8.22625637e-01 3.23496163e-01 -6.17086440e-02 1.01588976e+00
1.11016994e-02 4.17894512e-01 4.69945222e-01 -9.90348935e-01
1.75263360e-01 5.79868220e-02 1.04867302e-01 -1.10122895e+00
-2.66942799e-01 -5.84313095e-01 -1.02493072e+00 3.61159861e-01
5.31298459e-01 -3.47556859e-01 -1.14732432e+00 1.58420300e+00
3.79400522e-01 1.48319170e-01 -7.57844448e-02 5.67638516e-01
2.13072360e-01 7.26443768e-01 -2.82636821e-01 -4.27106142e-01
9.83226955e-01 -1.03352284e+00 -7.85250962e-01 -2.77672887e-01
3.58443439e-01 -1.17521918e+00 9.48109269e-01 6.09229326e-01
-1.46370840e+00 -4.38602984e-01 -9.87536907e-01 9.06220078e-02
-1.15085497e-01 1.90992177e-01 5.40123463e-01 8.60196710e-01
-1.35397685e+00 9.58358645e-01 -9.36696827e-01 -2.06151932e-01
2.86792696e-01 5.22489667e-01 4.60445508e-02 7.84116611e-02
-8.67363572e-01 9.53217268e-01 1.67712763e-01 1.97585002e-01
-8.53305936e-01 -3.70126188e-01 -5.27940571e-01 2.08131239e-01
5.07368922e-01 -7.11906791e-01 1.20473778e+00 -7.77306437e-01
-2.13555217e+00 4.12173957e-01 -2.98031986e-01 -5.38255870e-01
8.92520487e-01 5.50797507e-02 -5.16105704e-02 1.42365709e-01
-4.54972714e-01 6.28172636e-01 1.39399052e+00 -1.21596348e+00
-2.19509527e-01 -5.44999773e-03 6.25532717e-02 9.35190544e-02
-4.03463006e-01 -5.12324907e-02 -8.16009462e-01 -8.34212899e-01
2.17703074e-01 -1.11773038e+00 -6.26545310e-01 1.11983947e-01
-8.75476375e-02 4.55776937e-02 5.50781488e-01 -5.72888792e-01
1.08383644e+00 -2.29691434e+00 3.80167693e-01 2.43153587e-01
2.29111820e-01 4.35632944e-01 -1.04347907e-01 2.94857442e-01
1.31208271e-01 1.37235492e-01 -4.54669416e-01 -6.74327195e-01
-2.86377132e-01 2.27667898e-01 -3.38307209e-02 2.82277435e-01
2.16305822e-01 5.59005618e-01 -7.00939775e-01 -2.70040095e-01
3.29739124e-01 8.17753494e-01 -7.88016260e-01 -4.47051367e-03
-1.15663700e-01 3.51103574e-01 -2.06943661e-01 -1.75652876e-02
5.26834249e-01 -2.49552019e-02 1.84536815e-01 -5.07585071e-02
1.64693315e-02 2.53723174e-01 -1.51573300e+00 1.49922812e+00
-7.86651850e-01 5.82427263e-01 2.51298785e-01 -1.09685075e+00
8.41620326e-01 3.31275374e-01 3.22721303e-01 -2.76281476e-01
3.16879332e-01 4.71071661e-01 1.58867404e-01 -1.82143122e-01
2.00126216e-01 -4.79766786e-01 2.35294089e-01 4.62988734e-01
2.12782882e-02 -5.02978504e-01 5.64399421e-01 3.96472104e-02
1.00953281e+00 -3.50653738e-01 1.87756509e-01 2.67462134e-02
6.55101001e-01 -2.91266978e-01 1.46626100e-01 5.84093153e-01
1.04102351e-01 8.30893040e-01 6.04025960e-01 -3.06857735e-01
-1.22836185e+00 -6.33782566e-01 2.63984740e-01 7.79477596e-01
-1.93622023e-01 -2.41349056e-01 -1.07012892e+00 -3.53318185e-01
-4.57178295e-01 6.07349277e-01 -5.02144277e-01 -9.88699719e-02
-8.91397774e-01 -9.51637506e-01 5.04796505e-01 2.33166426e-01
2.36108348e-01 -1.04446328e+00 -3.11208725e-01 4.39958066e-01
1.63008243e-01 -9.66956198e-01 -6.52924955e-01 2.95078099e-01
-9.59461212e-01 -5.66775620e-01 -1.04218638e+00 -7.72063196e-01
8.25193524e-01 1.56964645e-01 7.65807986e-01 2.75210410e-01
-1.40850350e-01 -4.40001562e-02 -2.20547453e-01 -4.18719947e-02
-7.99417436e-01 2.83631563e-01 -2.50478059e-01 6.29975721e-02
-3.21190804e-01 -8.52132142e-01 -4.70026255e-01 1.49813816e-01
-9.70736921e-01 -8.24606419e-03 6.16684139e-01 7.67973840e-01
4.36166346e-01 2.97470987e-01 3.47986221e-01 -8.17484498e-01
8.30460310e-01 -9.66443568e-02 -7.22297788e-01 -6.02975115e-02
-6.98931634e-01 3.63861769e-01 7.20320463e-01 -7.51584709e-01
-9.31631148e-01 1.87159136e-01 -5.32003164e-01 -3.32640737e-01
7.18012871e-03 3.38453829e-01 1.52092874e-01 -3.39212358e-01
6.50829852e-01 2.03372404e-01 -2.63228342e-02 -6.43360138e-01
5.80081463e-01 3.46520156e-01 1.71114564e-01 -2.19462961e-01
8.49161208e-01 3.24568897e-01 2.92790346e-02 -9.16024268e-01
-5.24642944e-01 -7.61746764e-02 -6.29862130e-01 -1.07211411e-01
5.00218809e-01 -5.78500330e-01 -3.26638758e-01 6.04945660e-01
-1.12299788e+00 -4.57435101e-01 -4.16684866e-01 3.30996245e-01
-4.18632925e-01 3.26414615e-01 -6.97257817e-01 -6.86351895e-01
-2.15892911e-01 -1.23514700e+00 6.07262552e-01 1.60692949e-02
-1.86443269e-01 -9.64204192e-01 2.68914755e-02 -1.46389930e-02
7.07778573e-01 -1.23681314e-01 8.20539534e-01 -2.81206101e-01
-5.66696644e-01 -2.47753143e-01 1.36503531e-02 4.86681223e-01
2.35346943e-01 2.85361782e-02 -7.57575274e-01 -3.11448395e-01
3.04297984e-01 1.51617564e-02 1.01263499e+00 4.30120528e-01
1.02751935e+00 -3.04947078e-01 1.10865263e-02 6.59050703e-01
1.26453030e+00 3.79841417e-01 6.28433704e-01 8.99374858e-02
5.01691759e-01 2.77192503e-01 -6.66670799e-02 3.68043274e-01
-6.86654821e-02 6.77790999e-01 -9.29935202e-02 -1.95680559e-01
-5.54495990e-01 -2.12063324e-02 1.52210176e-01 1.04842770e+00
-1.89129427e-01 -3.91165018e-01 -6.81499660e-01 6.21877134e-01
-1.54677856e+00 -8.23484540e-01 1.37072682e-01 2.20674205e+00
9.91862416e-01 3.20694208e-01 4.18515727e-02 4.92465585e-01
5.04647791e-01 3.45844090e-01 -3.59180212e-01 -4.55184668e-01
-5.55317067e-02 5.52232921e-01 4.56782311e-01 1.02170730e+00
-7.41225600e-01 1.05770600e+00 7.42574692e+00 1.10589159e+00
-1.33366323e+00 9.62709859e-02 4.11061257e-01 -3.45762968e-01
-5.03514767e-01 6.90784752e-02 -5.39416671e-01 4.14548814e-01
1.01062000e+00 3.57145653e-03 8.02301407e-01 5.97332776e-01
3.78247529e-01 -1.86794132e-01 -6.78150058e-01 1.03633857e+00
8.89611989e-02 -1.26812100e+00 1.23744011e-01 -1.32506862e-01
7.21073508e-01 -2.87714660e-01 5.22277988e-02 -1.18043020e-01
6.18963949e-02 -9.08264160e-01 5.38631320e-01 3.08106452e-01
1.77415401e-01 -9.50054705e-01 5.22474468e-01 3.26242685e-01
-7.28652358e-01 -2.08040904e-02 -5.01260936e-01 6.75956532e-02
4.22901928e-01 1.04350460e+00 -4.24541175e-01 1.14153460e-01
3.78296137e-01 1.52184546e-01 -2.70319849e-01 1.04024577e+00
-3.85187656e-01 7.47087479e-01 -4.74561185e-01 -1.06766984e-01
2.51657844e-01 -3.80043149e-01 4.76124555e-01 1.25206780e+00
7.60698795e-01 5.94391339e-02 -1.17802382e-01 6.66374207e-01
-1.98016107e-01 1.41465217e-01 -2.99998820e-01 -2.56163813e-02
2.85178125e-01 1.16464746e+00 -9.80109334e-01 -3.57967079e-01
-1.25512537e-02 1.26570749e+00 3.71513784e-01 2.97140241e-01
-7.26881564e-01 -5.19745290e-01 3.98048341e-01 2.97589630e-01
7.57252991e-01 -7.16746747e-01 -3.14926714e-01 -1.00133657e+00
-8.13498348e-02 -8.44857395e-01 -1.34795085e-01 -4.15778548e-01
-7.48380661e-01 7.15014637e-01 -2.42363781e-01 -7.78652370e-01
-4.11422849e-01 -1.90533325e-01 -3.89004737e-01 8.75802815e-01
-1.42859769e+00 -8.09040070e-01 -1.89500675e-02 4.03150916e-01
6.49387658e-01 4.35043089e-02 7.01267719e-01 4.65485007e-01
-5.35277128e-01 4.28791434e-01 2.14034081e-01 -2.29709476e-01
6.30731642e-01 -1.02769947e+00 5.01631856e-01 8.16364288e-01
3.35927963e-01 5.63358605e-01 1.10126615e+00 -3.95138770e-01
-8.92209351e-01 -5.64996302e-01 1.21173429e+00 1.20425195e-01
7.21664786e-01 -4.36758727e-01 -9.27500665e-01 1.82671532e-01
3.24345231e-01 -1.59806147e-01 3.24319452e-01 -5.67288809e-02
5.17775640e-02 -2.36464947e-01 -1.02477241e+00 7.54224598e-01
8.56351554e-01 -4.22934145e-01 -1.14818253e-01 3.40158969e-01
3.74321967e-01 -5.85330606e-01 -6.09081745e-01 -5.61490059e-02
2.52967358e-01 -9.45351899e-01 1.01677036e+00 1.47439027e-02
2.10492551e-01 -4.35978740e-01 1.76785529e-01 -1.55579746e+00
-4.04947728e-01 -1.05430698e+00 -1.62027687e-01 1.21696913e+00
7.02074945e-01 -5.62905073e-01 7.23693788e-01 3.96439880e-01
1.28728911e-01 -6.13600552e-01 -9.25554276e-01 -6.58586085e-01
3.97089906e-02 -4.01954800e-01 3.18589538e-01 3.87011319e-01
-4.68446702e-01 4.10191447e-01 -7.36784399e-01 4.48874719e-02
5.01660705e-01 -1.98674530e-01 5.80706060e-01 -1.01910758e+00
-5.51405251e-01 -5.03597438e-01 -7.18292147e-02 -1.10434353e+00
-2.47056827e-01 -3.87836099e-01 1.30902693e-01 -1.38770854e+00
-1.43022433e-01 -4.39203680e-01 -2.19020471e-01 3.24052751e-01
-3.19109321e-01 5.39860427e-01 3.34625691e-01 5.43686859e-02
-4.76366319e-02 5.71988285e-01 1.20044160e+00 -1.15688048e-01
-4.03164059e-01 2.29212061e-01 -6.64455950e-01 8.29531252e-01
9.92301643e-01 -7.60297835e-01 -6.09380305e-01 -6.48907244e-01
1.65589780e-01 -1.19115673e-01 4.36899439e-02 -1.05309570e+00
3.66930187e-01 1.34882042e-02 2.03049839e-01 -2.64315784e-01
6.35815382e-01 -5.47533631e-01 1.82137579e-01 4.48013723e-01
-3.02877486e-01 -2.98314150e-02 1.84497222e-01 4.20032799e-01
-2.13162392e-01 -7.76321769e-01 1.19189036e+00 -1.12461649e-01
-1.85347632e-01 -3.92548228e-03 -6.96328342e-01 -2.62377888e-01
8.14785838e-01 -6.37939349e-02 2.67429799e-01 -5.80253839e-01
-1.02149510e+00 -4.04429406e-01 3.61087799e-01 1.81747437e-01
3.72953355e-01 -9.83326137e-01 -5.52005410e-01 1.75490007e-01
-6.21322751e-01 -9.05727297e-02 3.17860633e-01 6.55216217e-01
-5.47748089e-01 6.26268908e-02 1.18762970e-01 -3.03735912e-01
-1.15351915e+00 5.23303330e-01 4.03387338e-01 -4.29505080e-01
-7.14306533e-01 9.49997783e-01 -3.17980558e-01 -3.21434811e-02
9.99490395e-02 -3.39983910e-01 -1.77497894e-01 3.64066690e-01
5.17568409e-01 4.02219951e-01 2.18245506e-01 -3.84028316e-01
-9.60508212e-02 9.02699590e-01 -2.22848989e-02 -4.80517179e-01
1.39979601e+00 -2.33868733e-01 -5.30156679e-03 2.39013299e-01
9.63114679e-01 2.41291299e-01 -1.33473432e+00 -8.92953053e-02
-1.41672805e-01 -3.64948332e-01 4.51911181e-01 -4.60757434e-01
-1.32024169e+00 8.91185522e-01 6.30800009e-01 3.14604700e-01
1.26637709e+00 -2.32034534e-01 8.39953780e-01 2.81274527e-01
1.60210386e-01 -1.02798557e+00 4.89128716e-02 4.65483695e-01
7.31123686e-01 -9.56587911e-01 6.42002150e-02 -4.75917816e-01
-4.38881129e-01 1.11946034e+00 -1.09799914e-01 -2.58237541e-01
7.01679468e-01 6.14320993e-01 2.13788733e-01 9.36801583e-02
-6.05332792e-01 -2.48079583e-01 1.62210122e-01 5.08607745e-01
4.35517907e-01 -1.86716840e-01 -5.43181539e-01 -2.54071392e-02
-1.60740986e-01 -7.79234394e-02 4.77662623e-01 7.41035223e-01
-7.19691098e-01 -1.68013358e+00 -2.50886261e-01 5.45733646e-02
-5.04611373e-01 -2.54982173e-01 -2.55470157e-01 4.87356991e-01
2.10936610e-02 1.07967424e+00 -2.17097655e-01 -1.16078682e-01
3.08076352e-01 1.07025332e-03 6.37369692e-01 -2.94197261e-01
-6.42809689e-01 5.65966547e-01 4.58546802e-02 -3.81163567e-01
-4.83067304e-01 -4.83809918e-01 -1.03098118e+00 -4.31185335e-01
-3.33151042e-01 1.36420116e-01 6.71550989e-01 9.96423781e-01
1.81140095e-01 5.92206299e-01 6.45819604e-01 -8.45976591e-01
-5.86020172e-01 -9.57050085e-01 -4.46177781e-01 -1.45663870e-02
2.42374644e-01 -3.13200295e-01 -4.19065386e-01 2.01530576e-01] | [15.123665809631348, 5.917774677276611] |
cf8c677f-4ffc-4e7a-8f5a-072f0a53c7a5 | cell-detection-on-image-based-immunoassays | 1810.09707 | null | http://arxiv.org/abs/1810.09707v1 | http://arxiv.org/pdf/1810.09707v1.pdf | Cell detection on image-based immunoassays | Cell detection and counting in the image-based ELISPOT and Fluorospot
immunoassays is considered a bottleneck. The task has remained hard to
automatize, and biomedical researchers often have to rely on results that are
not accurate. Previously proposed solutions are heuristic, and data-based
solutions are subject to a lack of objective ground truth data. In this paper,
we analyze a partial differential equations model for ELISPOT, Fluorospot, and
assays of similar design. This leads us to a mathematical observation model for
the images generated by these assays. We use this model to motivate a
methodology for cell detection. Finally, we provide a real-data example that
suggests that this cell detection methodology and a human expert perform
comparably. | [] | 2018-10-23 | null | null | null | null | ['cell-detection'] | ['computer-vision'] | [-7.32088313e-02 -6.59468174e-01 1.98763207e-01 -3.29466201e-02
-4.80600148e-01 -8.51163805e-01 3.40468943e-01 5.46880126e-01
-6.28580749e-01 1.22292471e+00 -6.48449719e-01 -2.68597245e-01
4.79948632e-02 -6.70081556e-01 -3.72691602e-01 -1.02997792e+00
1.01644590e-01 1.04967260e+00 2.10313782e-01 2.08479583e-01
3.68431598e-01 1.01720846e+00 -1.10506511e+00 1.56255141e-01
4.07659888e-01 9.90009367e-01 -1.04403690e-01 1.23440015e+00
-3.11136693e-01 8.51684570e-01 -8.13823700e-01 -1.60792768e-01
2.05653459e-01 -5.65175891e-01 -5.42556226e-01 1.00801773e-01
-2.29346305e-01 -4.71475393e-01 2.84872949e-01 5.75832546e-01
6.52577937e-01 -4.65810537e-01 1.12144887e+00 -1.50984669e+00
-2.68447012e-01 -5.39261639e-01 -3.58906567e-01 2.69297659e-01
3.37728441e-01 2.38542050e-01 -2.74647586e-02 -8.54281664e-01
9.43389475e-01 1.10908055e+00 7.01615691e-01 4.65056181e-01
-1.47884512e+00 -2.70988762e-01 -5.09052157e-01 -4.04313862e-01
-1.59381688e+00 -6.80550411e-02 6.61036819e-02 -7.22351491e-01
7.14314461e-01 5.37508667e-01 1.00256383e+00 6.76151574e-01
1.03778112e+00 6.07833385e-01 1.75314677e+00 -2.54675478e-01
5.48558295e-01 4.82440323e-01 6.42087311e-02 6.50818646e-01
7.57011235e-01 1.28040239e-01 -1.24895208e-01 -3.02063942e-01
1.13466489e+00 4.44416016e-01 1.39561385e-01 -4.46485877e-01
-6.35416508e-01 9.53574359e-01 -1.99427769e-01 4.88656640e-01
-3.82632971e-01 -8.98768157e-02 3.37848999e-02 1.54239252e-01
7.39294589e-01 6.24674261e-01 -1.04605719e-01 2.74782479e-01
-9.99378800e-01 5.09224832e-01 1.06872666e+00 5.96600294e-01
5.50223529e-01 -3.08525294e-01 -2.24727243e-01 4.18741256e-01
2.10108995e-01 8.48490536e-01 7.62586892e-02 -8.87373149e-01
-8.13473463e-01 6.58651650e-01 7.56756186e-01 -9.48216259e-01
-6.57926440e-01 -2.69162416e-01 -5.74202180e-01 3.65713567e-01
8.82567227e-01 -3.34262043e-01 -8.08111489e-01 8.49457324e-01
5.26007175e-01 -1.36545356e-02 -8.21554363e-02 1.04872096e+00
3.31699401e-01 7.25734651e-01 -1.79463819e-01 -7.95099378e-01
1.28493357e+00 -5.33354700e-01 -8.36068451e-01 5.37319005e-01
7.00532436e-01 -9.40040529e-01 5.29449761e-01 5.00596941e-01
-9.66868341e-01 3.75862792e-02 -8.05211544e-01 1.92191437e-01
-6.46146119e-01 -3.47453982e-01 4.34029818e-01 6.97640717e-01
-1.06362581e+00 1.35107622e-01 -7.69869506e-01 -8.09373975e-01
1.84894696e-01 5.48906803e-01 8.83151591e-02 7.02484027e-02
-5.09051919e-01 1.00575614e+00 -2.07423940e-01 -4.91858795e-02
-7.09143221e-01 -5.50202847e-01 -1.82308942e-01 -4.26898450e-01
-6.68032244e-02 -7.27421880e-01 1.07785833e+00 -3.46556753e-01
-1.52051663e+00 1.23881686e+00 -2.53389448e-01 -5.52464724e-01
6.39317751e-01 3.40030521e-01 -1.77104771e-01 4.13123429e-01
1.83328949e-02 5.13138175e-01 1.43343508e-01 -1.65386450e+00
-8.23006868e-01 -4.51101571e-01 -2.16634765e-01 -2.76796728e-01
3.64410967e-01 3.09277743e-01 3.75747867e-02 -1.69562295e-01
-5.12774959e-02 -6.38083398e-01 -3.48997384e-01 8.90216678e-02
-1.21397592e-01 5.55656031e-02 7.02539921e-01 -1.77490667e-01
7.74192572e-01 -1.54622340e+00 -6.89037442e-01 1.91355020e-01
3.53655338e-01 1.93968326e-01 1.26298398e-01 6.40482664e-01
7.67767251e-01 1.25316009e-01 1.78610638e-03 -5.40149994e-02
-2.43733764e-01 8.82671401e-02 -1.52813673e-01 7.76567698e-01
-1.73077043e-02 9.94606674e-01 -1.07102752e+00 -1.01247168e+00
3.65581691e-01 7.29321480e-01 -1.19526930e-01 3.20252270e-01
3.10407188e-02 7.08814681e-01 -5.72520196e-01 1.00909066e+00
9.08384919e-01 -4.70652431e-01 1.42777562e-01 1.01381317e-01
-5.69448173e-01 -5.66011071e-01 -7.73966968e-01 6.76554680e-01
-2.81910039e-02 3.51071090e-01 2.75558114e-01 -7.38733590e-01
9.46313560e-01 2.63633162e-01 9.22332585e-01 -6.98330462e-01
6.68214083e-01 5.73737741e-01 -2.32330114e-01 -5.38827300e-01
1.65335566e-01 -8.88629079e-01 2.98371941e-01 6.37562990e-01
-2.20272109e-01 -3.75547111e-01 3.00605923e-01 9.70577002e-02
1.09885383e+00 -3.00140947e-01 3.11091542e-01 -7.06145108e-01
5.65377295e-01 6.69948936e-01 2.49749452e-01 8.56533527e-01
-4.65695888e-01 6.17357969e-01 4.27804768e-01 -7.59320140e-01
-1.06345141e+00 -1.13784862e+00 -4.62106586e-01 2.75953621e-01
4.21613932e-01 3.30901116e-01 -1.06975102e+00 -1.26497597e-01
2.99581647e-01 1.62961945e-01 -6.20058119e-01 4.50673193e-01
-1.42648891e-01 -1.17551613e+00 5.07099152e-01 1.70214117e-01
-5.01805432e-02 -5.90093493e-01 -9.42723095e-01 5.96364796e-01
3.93433534e-02 -1.20490265e+00 -5.65155670e-02 7.00546894e-03
-9.86798704e-01 -1.30280387e+00 -1.14927435e+00 -6.01533711e-01
9.72231567e-01 2.15413034e-01 8.57435822e-01 4.14776891e-01
-6.37396276e-01 6.10392272e-01 -1.67871147e-01 -7.83251524e-01
-4.26655293e-01 -4.78171945e-01 1.54767588e-01 2.57374644e-01
7.08177507e-01 -6.12431206e-02 -1.01422095e+00 5.52053869e-01
-8.06728959e-01 -1.58929601e-01 4.36979204e-01 7.41970181e-01
1.06068981e+00 -4.05031204e-01 5.52325130e-01 -1.07240784e+00
6.55337036e-01 -2.73835450e-01 -9.29201484e-01 2.44159266e-01
-9.11060929e-01 -2.38256395e-01 7.40969479e-01 -3.17716956e-01
-8.30746889e-01 1.59358412e-01 1.45701051e-01 -2.21860260e-01
-1.36372104e-01 2.20597923e-01 5.40636599e-01 -5.06154895e-01
6.15520477e-01 1.27730310e-01 5.07555842e-01 -1.98829830e-01
-2.62922019e-01 7.73119509e-01 2.36993209e-01 -3.85331005e-01
6.89113513e-02 1.30973721e+00 4.38363910e-01 -9.41923618e-01
-4.20593202e-01 -7.18048036e-01 -2.49222726e-01 -4.11883980e-01
6.52383506e-01 -4.29742396e-01 -1.10421431e+00 7.18259156e-01
-1.28712142e+00 -4.03182894e-01 -2.83061087e-01 6.50159359e-01
-7.05569804e-01 7.75597570e-03 -9.66732144e-01 -1.54850328e+00
-1.50736451e-01 -1.04201591e+00 1.16201162e+00 1.95494562e-01
-1.47964254e-01 -1.19394541e+00 4.24091429e-01 7.70981386e-02
5.82504869e-01 6.68936312e-01 3.68637502e-01 -4.88578498e-01
-7.68324077e-01 -4.81521696e-01 -2.91589648e-01 -4.02328938e-01
1.46114621e-02 3.00608218e-01 -1.07551706e+00 -2.73021668e-01
5.41768372e-01 -1.02846526e-01 4.21669275e-01 9.24230576e-01
6.66799843e-01 -1.40715867e-01 -9.06868935e-01 3.96209776e-01
1.65353763e+00 6.24004841e-01 6.01856828e-01 -1.62763119e-01
2.26671979e-01 9.88701105e-01 6.39485419e-01 5.26650310e-01
8.37521777e-02 3.99647266e-01 1.17760887e-02 -3.50718558e-01
3.54524016e-01 1.11725349e-02 -2.58791327e-01 6.17057502e-01
-9.96038169e-02 -7.20346451e-01 -8.47725153e-01 5.37566721e-01
-1.71464312e+00 -6.80607855e-01 -7.01340914e-01 2.09113932e+00
4.58477646e-01 -3.68677527e-01 5.12375772e-01 -1.35476202e-01
6.92818463e-01 -7.13816106e-01 -3.42262864e-01 -4.52751815e-01
-3.34764034e-01 1.46435380e-01 7.27464199e-01 7.10617602e-01
-5.90089321e-01 4.13711190e-01 8.21637440e+00 7.57671773e-01
-1.33919203e+00 1.17165044e-01 1.10197592e+00 -1.72691569e-01
-1.30510464e-01 1.57043543e-02 -6.98270559e-01 5.57106376e-01
9.64707553e-01 -2.61514366e-01 2.32984498e-01 4.49094400e-02
4.75581229e-01 -9.63265836e-01 -1.11730802e+00 1.06646311e+00
-3.16434614e-02 -1.36700177e+00 -4.12275270e-02 3.94486874e-01
6.00020289e-01 -4.22724277e-01 -1.91682622e-01 -3.24754536e-01
1.95137128e-01 -1.08633816e+00 3.09861451e-01 7.43124604e-01
6.56647146e-01 -3.60525817e-01 1.05624485e+00 4.19847667e-01
-7.33593464e-01 3.76609713e-01 -3.53805780e-01 -2.59767890e-01
2.86463410e-01 8.57543349e-01 -1.20039499e+00 5.45895584e-02
2.45356202e-01 -1.75576389e-01 -1.37253448e-01 1.06797242e+00
4.08267528e-01 1.81810379e-01 -4.55095738e-01 -7.79211044e-01
2.43741013e-02 -3.19463074e-01 2.76854932e-01 1.36915302e+00
3.75549406e-01 2.17417166e-01 -1.21492408e-01 1.00752449e+00
4.89360899e-01 3.09048265e-01 -7.38668859e-01 -1.57229647e-01
1.82994664e-01 1.50896561e+00 -1.56708920e+00 -4.77585822e-01
-4.39106613e-01 7.30209649e-01 -4.06686543e-03 5.56643784e-01
-5.89648068e-01 8.49185809e-02 3.08506459e-01 7.09846735e-01
-3.55272025e-01 -5.62680513e-02 -4.71561879e-01 -7.85880148e-01
-5.09901166e-01 -2.86314368e-01 3.51296693e-01 -7.73055613e-01
-1.29646409e+00 5.95417023e-02 -6.75854832e-03 -1.04569507e+00
-1.64809227e-01 -1.11940706e+00 -1.23999305e-01 7.95980692e-01
-1.14862752e+00 -6.91218853e-01 -3.46718216e-03 1.55122980e-01
5.63851893e-02 2.40430892e-01 6.71726406e-01 2.04278871e-01
-1.50627986e-01 -2.16564555e-02 4.07182157e-01 -1.65851876e-01
4.52413261e-01 -1.05536652e+00 -2.42963135e-01 4.93220687e-01
-4.51149642e-01 7.06667125e-01 1.01804209e+00 -7.57859945e-01
-1.56275392e+00 -7.76861787e-01 1.01111817e+00 -8.28689158e-01
4.63175774e-01 -3.11844379e-01 -6.21710539e-01 3.10578406e-01
4.08489630e-02 1.83170125e-01 7.03112543e-01 -4.66645837e-01
5.48297882e-01 -1.70070499e-01 -1.68586123e+00 4.70229179e-01
5.45983553e-01 -3.62151504e-01 3.99190485e-02 4.16648954e-01
-3.51545542e-01 -4.95259136e-01 -6.16829813e-01 -1.20571507e-02
1.16130447e+00 -9.46607888e-01 6.18614078e-01 -4.80436146e-01
-1.60174295e-01 -4.59065884e-01 1.60254896e-01 -9.10456002e-01
-7.33124791e-03 -4.58536327e-01 1.34718403e-01 6.84538960e-01
1.29524738e-01 -1.02191341e+00 8.53410244e-01 5.23381054e-01
4.87977087e-01 -9.80352819e-01 -8.55897188e-01 -1.29763830e+00
1.46583036e-01 2.30995998e-01 4.62998629e-01 8.05079043e-01
2.92217165e-01 3.53882551e-01 -2.46398915e-02 -2.20102444e-01
8.04715157e-01 1.65565103e-01 8.11919630e-01 -1.37010026e+00
1.76680595e-01 -2.43520260e-01 -5.19154012e-01 -7.67842412e-01
-5.29066443e-01 -3.53059590e-01 1.93639815e-01 -1.47332406e+00
5.85075736e-01 -2.89015919e-01 -1.79833233e-01 -3.19406062e-01
1.09294966e-01 4.71657991e-01 -1.91549852e-01 2.36294180e-01
-5.13412952e-01 -2.89185852e-01 1.45773995e+00 -3.88864707e-03
2.82339249e-02 -4.15554523e-01 -3.86170983e-01 4.66589391e-01
8.37132215e-01 -7.93389440e-01 -1.27355427e-01 -7.28748320e-03
3.23789835e-01 1.87338501e-01 7.72091031e-01 -6.55787945e-01
4.57293361e-01 -3.71546149e-01 6.17825985e-01 -6.71822846e-01
1.34785771e-01 -7.56372571e-01 6.02035344e-01 9.50343788e-01
2.69861460e-01 7.81243742e-02 -7.38439262e-02 2.78476119e-01
-1.11197174e-01 -2.07885921e-01 1.04288268e+00 -2.65338004e-01
-3.55446711e-02 2.18201071e-01 -1.01166117e+00 1.13847274e-02
1.21570277e+00 -6.79832101e-01 -5.84436417e-01 -2.72284746e-01
-4.82621938e-01 3.57696325e-01 8.45351219e-01 -4.80693847e-01
5.46978593e-01 -1.11197615e+00 -5.32574058e-01 1.70799360e-01
-3.66032943e-02 -2.73392618e-01 2.80092925e-01 1.33047891e+00
-1.37992024e+00 7.21910179e-01 -3.99258852e-01 -1.03494310e+00
-8.80348563e-01 9.07188475e-01 6.94535553e-01 -3.42585862e-01
1.42593384e-02 6.38119638e-01 5.49841225e-01 -1.99255750e-01
-4.77505982e-01 -3.68553966e-01 7.71917403e-02 -1.34160683e-01
4.82713133e-01 7.38226235e-01 -9.03892666e-02 -4.99056220e-01
-5.23160696e-01 6.57865047e-01 8.33836198e-02 -3.53644043e-01
8.50656807e-01 -2.97207624e-01 -1.91133060e-02 6.43843055e-01
9.02454555e-01 -2.01988369e-01 -1.00033927e+00 3.84782016e-01
-2.88511187e-01 -4.89982665e-01 -2.34366402e-01 -7.34080553e-01
-6.11361861e-01 9.34372842e-01 6.97020769e-01 7.53099442e-01
9.06091809e-01 -8.57513919e-02 5.94408393e-01 5.84543608e-02
5.40329099e-01 -1.32997179e+00 -1.35520294e-01 8.08441639e-02
7.14815438e-01 -1.01167917e+00 1.78662702e-01 -6.80756450e-01
6.33447105e-03 8.50593328e-01 4.08799052e-01 -1.94565162e-01
6.15809083e-01 8.96127164e-01 4.75255668e-01 -5.13318360e-01
-5.76118290e-01 -1.54345736e-01 -4.20619667e-01 7.53153563e-01
5.23038387e-01 8.76825005e-02 -9.38723266e-01 3.97607744e-01
4.09810603e-01 7.25554645e-01 5.22793949e-01 1.11245930e+00
-5.18697798e-01 -8.06353390e-01 -8.23412180e-01 5.74950457e-01
-6.81405187e-01 4.88478273e-01 -7.82754898e-01 9.66642857e-01
3.36114764e-02 1.15737295e+00 1.62139744e-01 4.60050851e-02
2.26658970e-01 -8.87927413e-02 8.33896816e-01 -1.45096496e-01
-2.26501003e-01 5.64102054e-01 -3.72279644e-01 -3.58317733e-01
-6.19944334e-01 -6.61972940e-01 -1.42751527e+00 -7.60332227e-01
-3.53613764e-01 2.59079993e-01 3.01077425e-01 9.58981514e-01
3.32730263e-01 6.93001226e-02 5.85790873e-01 -4.21687126e-01
3.41626517e-02 -6.24739051e-01 -1.33275998e+00 1.32469863e-01
2.78887779e-01 -6.83948159e-01 -3.52810591e-01 2.38023564e-01] | [14.169873237609863, -3.164111852645874] |
386bef42-f868-40f5-bb11-ade76b249b66 | when-fair-classification-meets-noisy | 2307.03306 | null | https://arxiv.org/abs/2307.03306v2 | https://arxiv.org/pdf/2307.03306v2.pdf | When Fair Classification Meets Noisy Protected Attributes | The operationalization of algorithmic fairness comes with several practical challenges, not the least of which is the availability or reliability of protected attributes in datasets. In real-world contexts, practical and legal impediments may prevent the collection and use of demographic data, making it difficult to ensure algorithmic fairness. While initial fairness algorithms did not consider these limitations, recent proposals aim to achieve algorithmic fairness in classification by incorporating noisiness in protected attributes or not using protected attributes at all. To the best of our knowledge, this is the first head-to-head study of fair classification algorithms to compare attribute-reliant, noise-tolerant and attribute-blind algorithms along the dual axes of predictivity and fairness. We evaluated these algorithms via case studies on four real-world datasets and synthetic perturbations. Our study reveals that attribute-blind and noise-tolerant fair classifiers can potentially achieve similar level of performance as attribute-reliant algorithms, even when protected attributes are noisy. However, implementing them in practice requires careful nuance. Our study provides insights into the practical implications of using fair classification algorithms in scenarios where protected attributes are noisy or partially available. | ['Christo Wilson', 'Pablo Kvitca', 'Avijit Ghosh'] | 2023-07-06 | null | null | null | null | ['fairness', 'classification-1', 'fairness'] | ['computer-vision', 'methodology', 'miscellaneous'] | [ 2.58964241e-01 -4.85834517e-02 -3.26413482e-01 -8.69674504e-01
-8.02378237e-01 -7.35052884e-01 4.40298527e-01 5.05880058e-01
-7.88448513e-01 1.18927228e+00 1.98512927e-01 -5.95096231e-01
-4.21900392e-01 -7.68060327e-01 -2.77565330e-01 -6.12815499e-01
-2.01915260e-02 4.28246051e-01 -5.27137876e-01 -9.50751174e-03
3.45170617e-01 4.68361437e-01 -1.86798513e+00 -1.05868384e-01
1.07143927e+00 9.95033920e-01 -8.03772628e-01 4.22124296e-01
9.51150060e-02 6.92214608e-01 -6.43236578e-01 -1.00448334e+00
7.90699363e-01 -1.51652858e-01 -7.77630985e-01 -3.22507977e-01
5.33584297e-01 -5.94912708e-01 4.69494313e-02 9.68637943e-01
8.29332769e-01 1.12557113e-01 6.09168410e-01 -1.67775106e+00
-9.16137278e-01 8.57847154e-01 -4.10781533e-01 6.61424361e-03
2.85488427e-01 2.18571514e-01 1.15865791e+00 -1.90362602e-01
1.26363605e-01 1.18783665e+00 8.39612544e-01 5.77803612e-01
-1.64801800e+00 -9.80085254e-01 -7.08666369e-02 -2.63943002e-02
-1.45238626e+00 -9.68202591e-01 2.95772195e-01 -2.52944589e-01
3.55960518e-01 8.80113244e-01 1.87319085e-01 9.44618762e-01
-9.86001417e-02 2.97876894e-01 1.56342304e+00 -3.40758026e-01
5.44197738e-01 2.63719410e-01 2.88345248e-01 8.20260867e-02
8.82897496e-01 3.75517100e-01 -5.49328864e-01 -9.86050367e-01
-3.97035144e-02 -1.38735935e-01 -2.11183757e-01 -2.58654743e-01
-8.43361616e-01 8.59969795e-01 1.42646492e-01 -2.65573442e-01
-4.77202624e-01 -1.00233093e-01 5.88027894e-01 5.33610404e-01
5.69171846e-01 5.41635990e-01 -3.58079940e-01 -1.17961295e-01
-9.72144604e-01 4.90716547e-01 8.17918420e-01 5.99241078e-01
6.11711919e-01 -1.14131458e-01 -1.41064078e-01 6.25732124e-01
-1.21325115e-02 4.69821662e-01 1.94233969e-01 -1.31334591e+00
3.83018434e-01 1.67421192e-01 4.95498508e-01 -1.04099107e+00
-2.02981323e-01 -2.81216592e-01 -9.26277876e-01 4.69529569e-01
8.47836137e-01 -1.46342153e-02 -5.71948946e-01 2.05657268e+00
3.17053199e-01 -3.19790125e-01 7.69116133e-02 1.04233444e+00
2.97570944e-01 5.22478335e-02 8.28235865e-01 -4.19828624e-01
1.33615208e+00 1.01888485e-01 -7.24016011e-01 -4.27051671e-02
3.64730269e-01 -7.42308259e-01 1.08098221e+00 2.35195026e-01
-1.03221571e+00 -1.21718384e-01 -8.77608120e-01 -1.14349507e-01
-3.20240110e-01 -4.76718366e-01 8.84959280e-01 1.59897709e+00
-7.12013304e-01 5.27122080e-01 -3.17855507e-01 -3.49901050e-01
7.53191471e-01 1.81152016e-01 -6.65500462e-01 -2.63656043e-02
-1.37368071e+00 8.19805503e-01 -5.80056645e-02 -1.75431728e-01
-4.00810182e-01 -9.27344263e-01 -5.23690224e-01 1.62500352e-01
3.70469511e-01 -7.43079126e-01 9.78112280e-01 -1.18700385e+00
-6.63538933e-01 1.10024023e+00 -2.10704599e-02 -5.16721785e-01
1.05377388e+00 -8.62998143e-03 -3.00052434e-01 -3.54496986e-01
3.61673474e-01 1.43207684e-01 3.38939309e-01 -1.36983478e+00
-6.07299745e-01 -6.79838359e-01 4.22115587e-02 2.17349336e-01
-4.28562850e-01 1.99072301e-01 5.87068975e-01 -6.44298494e-01
-4.70382236e-02 -5.68393409e-01 -2.34843567e-01 5.03567457e-01
-4.16840792e-01 1.08636305e-01 2.73041070e-01 -5.25308847e-01
1.00620341e+00 -2.13443351e+00 -7.35327363e-01 4.11000758e-01
2.83583224e-01 1.67377323e-01 1.26082212e-01 2.29838192e-01
-2.80428259e-03 6.86911106e-01 -3.87076497e-01 -3.27357441e-01
2.23250747e-01 2.75539309e-01 -3.52334559e-01 8.12310100e-01
-2.74386823e-01 4.16095495e-01 -6.61697090e-01 -5.62721670e-01
-1.72870740e-01 1.35949552e-01 -4.34368044e-01 7.17710191e-03
2.65550911e-01 -3.38591188e-02 -2.34594807e-01 7.98228562e-01
1.01085305e+00 3.42716306e-01 2.63782412e-01 1.30921587e-01
-2.42609046e-02 3.36802453e-01 -1.30205572e+00 9.64491606e-01
-9.19984952e-02 2.06893504e-01 4.21677053e-01 -9.54957485e-01
7.22418725e-01 2.27162004e-01 4.18272167e-01 -8.50536644e-01
4.99532893e-02 3.10402572e-01 2.25936957e-02 -1.04051098e-01
7.52519548e-01 -6.11810744e-01 -2.85005659e-01 1.02918994e+00
-6.35755479e-01 1.11872628e-01 -2.35263303e-01 1.49475887e-01
8.14946413e-01 -3.76642019e-01 5.43748975e-01 -5.58367848e-01
9.66797099e-02 -1.60402119e-01 1.03144240e+00 1.13713777e+00
-8.78981709e-01 7.27200747e-01 3.66041392e-01 -5.98127127e-01
-1.13618565e+00 -1.10128713e+00 -4.45033938e-01 1.36335588e+00
-4.18730304e-02 -1.01854816e-01 -6.39799953e-01 -6.03474975e-01
6.49761379e-01 9.67724025e-01 -7.96293557e-01 -2.28482798e-01
3.70792709e-02 -1.05950761e+00 1.01346207e+00 2.26710603e-01
3.35987061e-01 -4.04433429e-01 -8.51714909e-01 3.68832201e-02
-2.79631466e-01 -7.82327831e-01 -3.09398204e-01 -4.23132293e-02
-4.70567942e-01 -1.15033662e+00 -1.59011737e-01 1.91972136e-01
5.83188713e-01 1.83742687e-01 1.13593388e+00 4.00263369e-01
-3.66394579e-01 2.40373284e-01 -8.71641561e-02 -7.10910380e-01
-4.45136964e-01 -1.97241604e-01 3.54630530e-01 4.07331996e-02
5.43565333e-01 -7.09146142e-01 -5.45739591e-01 5.27664244e-01
-8.23669970e-01 -4.44681466e-01 1.39713213e-01 6.92755342e-01
3.00987273e-01 -5.16467839e-02 7.69091427e-01 -1.13952351e+00
9.27701354e-01 -5.37603378e-01 -3.29130858e-01 3.66173297e-01
-1.10513210e+00 -1.01540729e-01 6.88652456e-01 -1.57883763e-01
-1.02697003e+00 -1.73284248e-01 8.78648683e-02 2.10099116e-01
-2.79739916e-01 1.29455075e-01 -3.73304248e-01 4.96803783e-02
1.17083502e+00 -1.98027164e-01 2.00119913e-01 -3.54001313e-01
3.15300912e-01 1.05861568e+00 4.90400612e-01 -1.11858714e+00
6.20196164e-01 6.06609464e-01 8.59688818e-02 -2.58162916e-01
-6.83733284e-01 1.42439893e-02 -2.92196482e-01 5.57833649e-02
1.75553039e-01 -7.83374965e-01 -1.05519724e+00 3.52650791e-01
-5.02729177e-01 1.43742552e-02 -5.58769584e-01 2.62002736e-01
-4.97842461e-01 5.56379914e-01 -2.82981247e-01 -1.19338775e+00
-5.07751524e-01 -8.47915113e-01 4.47435766e-01 -3.81033309e-02
-5.75607240e-01 -4.86449629e-01 -3.10613960e-01 6.12737060e-01
6.39397800e-01 4.11026478e-01 8.05394292e-01 -9.61491108e-01
-1.18298031e-01 -5.35699546e-01 -2.56487489e-01 1.96689561e-01
3.30501854e-01 7.67515153e-02 -1.43454814e+00 -2.22946420e-01
-2.09903315e-01 -6.53052449e-01 5.44559717e-01 1.98326319e-01
1.18984306e+00 -6.16075754e-01 7.80743137e-02 5.06093204e-01
1.26840854e+00 -7.58949518e-02 5.93312204e-01 3.39118004e-01
2.47069418e-01 9.11588550e-01 6.47436678e-01 9.88671184e-01
7.10695028e-01 4.77741748e-01 3.79178882e-01 5.35077713e-02
3.49129081e-01 -6.07875213e-02 -3.02423239e-01 -2.68393964e-01
-2.25696191e-01 -1.11175708e-01 -9.85703051e-01 5.29466689e-01
-1.60221088e+00 -1.19452131e+00 -1.38649652e-02 2.86589766e+00
1.05160940e+00 2.16613407e-03 3.55111271e-01 5.35612583e-01
7.65716732e-01 -9.83556211e-02 -5.82440257e-01 -7.11368918e-01
-2.74796486e-01 -1.40808821e-01 6.98806465e-01 3.96908075e-01
-9.90416944e-01 2.82135814e-01 6.48146486e+00 4.44008350e-01
-7.30503380e-01 5.20346574e-02 1.21223152e+00 -3.96927208e-01
-5.13058424e-01 1.82154492e-01 -1.83452398e-01 5.95532417e-01
8.74072134e-01 -6.58957243e-01 4.53091830e-01 7.22932041e-01
1.90811887e-01 -1.91992432e-01 -1.04730666e+00 7.31398463e-01
-2.07923770e-01 -7.78079093e-01 -1.67428613e-01 1.82416737e-01
2.98192739e-01 -4.67837512e-01 1.16468869e-01 -1.17834479e-01
7.74385989e-01 -1.28818691e+00 1.02367592e+00 5.09270608e-01
8.84660721e-01 -1.01320159e+00 7.03252673e-01 2.27388009e-01
-5.13013601e-01 -3.30403835e-01 -4.11690861e-01 -3.86652231e-01
-1.24075979e-01 9.00647640e-01 -2.68571407e-01 5.88313162e-01
8.04765224e-01 -1.35093760e-02 -4.15701330e-01 9.63591099e-01
2.82595605e-01 6.60046339e-01 -3.69336337e-01 2.82267094e-01
-2.31072858e-01 2.16051310e-01 2.35194549e-01 8.65780354e-01
2.01246813e-01 4.09318358e-01 8.03648308e-03 7.26575077e-01
-1.67058513e-01 1.83034942e-01 -4.59218383e-01 1.89394325e-01
1.10281658e+00 1.02313006e+00 -3.91856700e-01 -1.46579623e-01
-1.61895916e-01 4.79832798e-01 2.13917896e-01 1.67255044e-01
-4.41095144e-01 -2.21600264e-01 1.23257506e+00 3.38183522e-01
-5.94691038e-01 1.51698679e-01 -1.03649497e+00 -8.06109965e-01
1.11387767e-01 -1.18458867e+00 9.72859740e-01 -4.01304007e-01
-1.75975025e+00 6.74884394e-02 -1.51684791e-01 -9.78657484e-01
6.50214329e-02 -5.56769595e-02 -2.00704947e-01 1.38592422e+00
-1.31013906e+00 -1.03926158e+00 -2.61046946e-01 5.93806565e-01
-3.42777759e-01 -1.93502575e-01 1.18501318e+00 2.64362335e-01
-2.88403600e-01 1.19886386e+00 1.68302000e-01 -2.68368749e-03
1.11821270e+00 -1.03255665e+00 4.24975932e-01 7.85532415e-01
-3.97028804e-01 9.91993368e-01 1.04098308e+00 -5.19914865e-01
-1.04643917e+00 -8.38896215e-01 9.60479736e-01 -5.26357710e-01
1.81565806e-01 -4.01959985e-01 -9.81849313e-01 4.36863124e-01
-1.99272528e-01 2.40155712e-01 1.17401147e+00 4.04679656e-01
-9.08233762e-01 -3.74110192e-01 -1.93604803e+00 4.62218940e-01
1.00938046e+00 -5.78208029e-01 -2.05862477e-01 1.14065269e-02
2.32761309e-01 -6.57855272e-02 -8.31979930e-01 2.36023843e-01
1.04303777e+00 -1.28439045e+00 9.19705629e-01 -7.90641367e-01
7.61217773e-02 -1.12285972e-01 -4.80393112e-01 -1.11686027e+00
-3.10982287e-01 -6.04246438e-01 6.87078416e-01 1.65305007e+00
2.95521706e-01 -1.03312302e+00 6.76041961e-01 1.66006887e+00
4.35851604e-01 -1.21752933e-01 -1.13463354e+00 -6.39027953e-01
3.87751728e-01 -5.91394246e-01 1.24735439e+00 1.45191383e+00
-1.57848760e-01 -4.65012968e-01 -5.98156631e-01 1.10310808e-01
1.04624772e+00 6.52650148e-02 7.21125782e-01 -1.52143228e+00
4.41988297e-02 -3.40035856e-01 -5.23083568e-01 1.79636091e-01
2.97037274e-01 -5.81104636e-01 -1.43775001e-01 -1.03635466e+00
4.25158262e-01 -1.20750105e+00 -2.21744463e-01 8.88152540e-01
-5.18539846e-01 1.75059766e-01 2.89618164e-01 2.87117690e-01
-5.93868606e-02 2.18893737e-01 4.76361752e-01 -6.43293560e-02
8.75001177e-02 1.96515650e-01 -1.60000718e+00 9.21436697e-02
1.01870894e+00 -6.77524388e-01 -3.89374465e-01 -1.65699378e-01
1.58364311e-01 -1.80076703e-01 5.25049686e-01 -7.54439592e-01
1.81295574e-01 -7.35726714e-01 3.58385712e-01 2.04635382e-01
5.32533936e-02 -9.62835908e-01 4.01122987e-01 2.44581506e-01
-6.89781010e-01 2.93953046e-02 9.28878933e-02 4.12881970e-01
2.18748957e-01 4.59556989e-02 8.79924595e-01 -1.38585269e-01
-1.99388176e-01 2.04656854e-01 -2.35217705e-01 2.88542211e-01
1.11023927e+00 -3.52225989e-01 -7.69794881e-01 -5.65514207e-01
-2.88969904e-01 1.91065222e-01 8.62877965e-01 8.66794959e-02
1.80534452e-01 -1.15166950e+00 -1.08563817e+00 2.07041517e-01
2.53024638e-01 -5.30658424e-01 1.88842297e-01 3.36665124e-01
-1.75791204e-01 1.38635347e-02 -5.32880247e-01 -4.10566851e-02
-1.38834858e+00 6.66723788e-01 3.92457813e-01 3.12563479e-01
-8.75017792e-02 6.05723023e-01 -1.39889941e-01 -6.11903071e-01
2.94653058e-01 4.15599793e-01 3.72935653e-01 2.36583367e-01
6.41766429e-01 6.50496721e-01 1.62283957e-01 -6.43988967e-01
-5.89275479e-01 6.87672868e-02 5.08324876e-02 -2.41547748e-01
9.83320117e-01 -3.85668993e-01 -2.62323678e-01 8.18445757e-02
6.10081613e-01 1.67681873e-01 -9.22122002e-01 -1.56352341e-01
-6.92360774e-02 -1.08813286e+00 -9.80642587e-02 -1.11262691e+00
-8.92058313e-01 6.76844835e-01 6.08170152e-01 4.48840141e-01
1.34598088e+00 -6.38614535e-01 2.38080844e-01 1.67383373e-01
5.67602098e-01 -1.15279877e+00 -8.99837017e-01 -2.93659508e-01
5.03757894e-01 -1.24932945e+00 4.23981220e-01 -2.69651979e-01
-7.83572674e-01 5.43067575e-01 3.49762022e-01 4.69900966e-01
5.13852239e-01 2.81659335e-01 3.33505303e-01 3.35784763e-01
-7.27046013e-01 5.66740707e-03 -1.31065771e-01 8.92951548e-01
4.91147399e-01 5.18605232e-01 -7.33106911e-01 7.54709065e-01
-4.19516861e-01 -2.64699403e-02 7.60064304e-01 1.03701568e+00
-1.45878240e-01 -1.14242983e+00 -7.31300652e-01 8.19703102e-01
-1.00695920e+00 -1.01302162e-01 -4.55169261e-01 5.71739554e-01
7.29454458e-02 1.34961569e+00 9.48489234e-02 -1.78581119e-01
3.88520151e-01 2.12051213e-01 -1.43643254e-02 -2.06531763e-01
-7.21238136e-01 -6.94571555e-01 4.85019863e-01 -5.42269945e-01
-3.74529004e-01 -9.41400170e-01 -8.30094874e-01 -1.18965054e+00
-5.81099726e-02 4.06038612e-01 6.30497217e-01 6.18800938e-01
5.04478097e-01 -2.48868957e-01 7.81922579e-01 -2.73732096e-01
-1.16786802e+00 -3.99120837e-01 -9.00315821e-01 7.73991406e-01
4.93412793e-01 -4.54168171e-01 -5.79952657e-01 -1.98243812e-01] | [8.844497680664062, 5.304582118988037] |
9870d5f0-4653-4d05-a11a-f565f79eb877 | mattnet-modular-attention-network-for | 1801.08186 | null | http://arxiv.org/abs/1801.08186v3 | http://arxiv.org/pdf/1801.08186v3.pdf | MAttNet: Modular Attention Network for Referring Expression Comprehension | In this paper, we address referring expression comprehension: localizing an
image region described by a natural language expression. While most recent work
treats expressions as a single unit, we propose to decompose them into three
modular components related to subject appearance, location, and relationship to
other objects. This allows us to flexibly adapt to expressions containing
different types of information in an end-to-end framework. In our model, which
we call the Modular Attention Network (MAttNet), two types of attention are
utilized: language-based attention that learns the module weights as well as
the word/phrase attention that each module should focus on; and visual
attention that allows the subject and relationship modules to focus on relevant
image components. Module weights combine scores from all three modules
dynamically to output an overall score. Experiments show that MAttNet
outperforms previous state-of-art methods by a large margin on both
bounding-box-level and pixel-level comprehension tasks. Demo and code are
provided. | ['Jimei Yang', 'Xiaohui Shen', 'Xin Lu', 'Tamara L. Berg', 'Licheng Yu', 'Zhe Lin', 'Mohit Bansal'] | 2018-01-24 | mattnet-modular-attention-network-for-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Yu_MAttNet_Modular_Attention_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Yu_MAttNet_Modular_Attention_CVPR_2018_paper.pdf | cvpr-2018-6 | ['generalized-referring-expression-segmentation', 'referring-expression-segmentation'] | ['computer-vision', 'computer-vision'] | [ 2.90803134e-01 2.58866429e-01 -8.52545723e-02 -6.17287040e-01
-5.05229950e-01 -3.88622642e-01 4.58878636e-01 2.45347455e-01
-4.80130672e-01 2.63749808e-01 2.93104887e-01 -5.84178865e-02
4.09594268e-01 -7.19377041e-01 -7.62070537e-01 -4.14510131e-01
1.92678437e-01 1.52134657e-01 5.53938806e-01 -1.49498805e-01
1.80426210e-01 3.73259664e-01 -1.29228497e+00 5.19767463e-01
7.41201580e-01 1.00317883e+00 4.77786571e-01 4.06202525e-01
-4.29365307e-01 9.87245739e-01 -6.52623177e-01 -4.55903530e-01
-1.71018630e-01 -3.67988467e-01 -1.01721883e+00 2.81593412e-01
6.23368740e-01 -2.05029249e-01 9.49917063e-02 1.18808877e+00
3.81643236e-01 1.91574901e-01 6.43947899e-01 -9.22215462e-01
-9.38069344e-01 5.53199232e-01 -1.02559364e+00 3.09157997e-01
4.32207048e-01 1.37642965e-01 1.09962654e+00 -9.25519884e-01
5.61446905e-01 1.58066964e+00 1.20900981e-01 5.50902069e-01
-1.39380801e+00 -4.04380888e-01 7.61468947e-01 3.01714033e-01
-1.20975304e+00 -1.52401015e-01 7.56613493e-01 -4.78626430e-01
1.11042869e+00 1.57565381e-02 4.30312961e-01 7.42990613e-01
1.97650984e-01 9.55906630e-01 9.19527769e-01 -4.72379088e-01
1.76482573e-01 1.54569447e-01 1.10631026e-01 7.70045519e-01
-1.27905712e-01 -5.49513638e-01 -4.78537560e-01 7.39317983e-02
8.58400762e-01 -7.66941085e-02 -3.16178262e-01 -6.41949296e-01
-1.25196123e+00 6.69107080e-01 9.05279815e-01 2.77532786e-01
-6.13476336e-01 4.97782111e-01 2.97577143e-01 5.83959036e-02
3.20503205e-01 2.75271863e-01 -4.37515557e-01 2.89515346e-01
-5.27999163e-01 7.22760111e-02 3.00620794e-01 1.00460029e+00
9.52747107e-01 -2.63867646e-01 -5.31861424e-01 9.34119582e-01
3.58545840e-01 1.75410897e-01 3.10194939e-01 -9.20843601e-01
4.31506813e-01 9.84167039e-01 -1.00906141e-01 -8.98654938e-01
-4.34738845e-01 -4.32815522e-01 -6.52003706e-01 3.71718407e-01
4.52753678e-02 -1.25745595e-01 -9.86851037e-01 2.10435390e+00
2.43003160e-01 -2.83231258e-01 -2.49439329e-01 9.73639071e-01
1.04902160e+00 5.79422891e-01 6.89278126e-01 2.63344616e-01
1.81548679e+00 -1.45188618e+00 -7.36929834e-01 -6.52248144e-01
4.31169897e-01 -5.22530735e-01 1.25015593e+00 -1.76700260e-02
-1.56276250e+00 -6.54072344e-01 -9.49455202e-01 -5.26265383e-01
-4.84183431e-01 4.95638326e-02 3.96266311e-01 1.22610047e-01
-1.30484498e+00 -2.53646933e-02 -6.12935543e-01 -5.80696344e-01
4.91858274e-01 3.08277309e-01 -4.11075443e-01 8.16629529e-02
-1.05297804e+00 1.07512689e+00 1.17439024e-01 -1.12541981e-01
-7.02852488e-01 -5.06124377e-01 -1.06363618e+00 4.90814447e-01
3.20583254e-01 -1.28313458e+00 1.49059904e+00 -1.55925596e+00
-1.28129137e+00 1.28530324e+00 -5.21918654e-01 -2.05512479e-01
2.02381551e-01 -1.64802626e-01 9.24618319e-02 4.05616522e-01
1.37566626e-01 1.21490741e+00 8.28772366e-01 -1.39267385e+00
-5.90401590e-01 -3.53547513e-01 4.64902014e-01 2.91683435e-01
-2.46941820e-01 3.06131423e-01 -9.77602720e-01 -6.82034731e-01
1.05053380e-01 -5.96666932e-01 -2.33369485e-01 5.68765402e-01
-1.60492346e-01 -4.01728868e-01 6.18895352e-01 -5.68295360e-01
1.25793040e+00 -2.06070375e+00 6.92691505e-01 -1.34967729e-01
2.15329096e-01 -1.16390020e-01 -2.54463553e-01 1.74267143e-01
-3.95414203e-01 8.48849565e-02 -4.34784710e-01 -5.78531981e-01
-1.01608790e-01 1.11344144e-01 -8.60029757e-02 2.24788845e-01
6.57756865e-01 1.16974401e+00 -9.23836350e-01 -6.32269263e-01
1.40941888e-01 4.91732389e-01 -5.22485793e-01 3.04309517e-01
-4.55831259e-01 2.48277366e-01 -3.97129536e-01 7.54131854e-01
6.36696339e-01 -4.91473526e-01 -5.47206737e-02 -2.82588631e-01
1.14378870e-01 2.09943596e-02 -8.01206648e-01 1.90451646e+00
-6.81143522e-01 6.21962786e-01 3.53688538e-01 -7.87206531e-01
9.10737872e-01 5.22177666e-02 1.34549990e-01 -6.93706930e-01
1.67304307e-01 -3.28918189e-01 5.32671669e-03 -7.02719629e-01
4.47539747e-01 1.10909633e-01 -2.82555968e-01 4.62986708e-01
2.80233622e-01 1.34460792e-01 1.29640937e-01 4.31381762e-01
1.00222945e+00 3.24963868e-01 5.05762815e-01 -2.89803863e-01
7.88167298e-01 -2.63375252e-01 3.78874779e-01 4.90529954e-01
-2.94374079e-01 7.21332967e-01 7.22947955e-01 -2.61194855e-01
-6.65271878e-01 -1.16221929e+00 9.00526494e-02 1.74262428e+00
3.22158039e-01 -4.11335498e-01 -8.64428580e-01 -6.18615627e-01
-2.38308474e-01 6.05340242e-01 -1.06482482e+00 -3.63257416e-02
-4.65827137e-01 -3.62239212e-01 -4.96994779e-02 9.27396536e-01
5.99286795e-01 -1.38953161e+00 -9.66375172e-01 -2.56929640e-02
-1.43808261e-01 -1.00428796e+00 -6.96506381e-01 1.96758524e-01
-6.55210614e-01 -8.54914308e-01 -1.02478039e+00 -1.10735023e+00
1.13547873e+00 1.65163323e-01 1.49588192e+00 1.57502159e-01
-2.56415397e-01 5.99472046e-01 -4.04426605e-01 -4.04249221e-01
-5.71414493e-02 1.59228563e-01 -6.82163954e-01 2.34925464e-01
2.50435591e-01 -4.68898743e-01 -7.10343301e-01 2.29307681e-01
-1.09136820e+00 3.11095923e-01 8.65313530e-01 6.37538731e-01
6.06796205e-01 -5.95616519e-01 2.84659177e-01 -8.02233696e-01
6.18699789e-01 -3.56328070e-01 -4.97326612e-01 6.43311799e-01
-7.15509653e-02 1.34564236e-01 1.94894537e-01 -4.28351313e-01
-1.13726330e+00 3.98763448e-01 1.41507918e-02 -3.06081653e-01
-1.58114642e-01 3.96356314e-01 -4.94399250e-01 1.39282152e-01
4.17240083e-01 1.51914850e-01 -2.58768946e-01 -2.70816177e-01
6.38334274e-01 3.38530809e-01 6.43321276e-01 -6.59543753e-01
4.85973686e-01 3.73910874e-01 -2.68411964e-01 -4.55384761e-01
-9.41439807e-01 -3.89294147e-01 -9.26494360e-01 -2.46298611e-01
1.28378201e+00 -8.95036697e-01 -6.35617435e-01 2.78467387e-01
-1.47936845e+00 -4.25210834e-01 -1.66549399e-01 -1.88747957e-01
-5.46748698e-01 1.88984379e-01 -4.17150915e-01 -5.86126447e-01
-4.56277549e-01 -1.31048441e+00 1.40854108e+00 4.00218934e-01
-3.83112848e-01 -1.04674232e+00 -1.77447960e-01 1.89030290e-01
5.44899106e-01 2.00361282e-01 1.08772993e+00 -1.44724160e-01
-6.55811071e-01 1.07687535e-02 -5.94352663e-01 8.14508572e-02
-1.47405118e-01 -6.66565597e-02 -1.01306474e+00 -1.58804446e-01
-1.87800810e-01 -3.00817609e-01 1.12477398e+00 3.36590976e-01
1.45687020e+00 -1.25440434e-01 -3.96758348e-01 6.78144217e-01
1.27266824e+00 -5.79771213e-03 7.81110466e-01 3.53946745e-01
5.11288762e-01 7.94403434e-01 4.00313109e-01 1.73373625e-01
6.35753930e-01 8.13700438e-01 6.42341673e-01 -6.06705606e-01
-1.84540540e-01 -1.16435535e-01 3.88115227e-01 2.41896108e-01
1.62629709e-01 -2.31982246e-01 -7.64318585e-01 6.41260982e-01
-2.07515478e+00 -8.28638136e-01 8.86698812e-02 1.87168491e+00
7.76171327e-01 1.20961322e-02 6.85259625e-02 -5.81639349e-01
5.68727076e-01 2.47133270e-01 -6.21026754e-01 -6.89477682e-01
-2.68360764e-01 1.23447485e-01 1.61867723e-01 5.28701723e-01
-1.14548361e+00 1.06976545e+00 6.17501307e+00 5.68470538e-01
-1.09008515e+00 2.08228186e-01 8.62954438e-01 8.14011320e-02
-3.56558174e-01 4.40956000e-03 -4.40836012e-01 2.09126204e-01
2.44347245e-01 -1.23263679e-01 8.45673755e-02 7.11017191e-01
2.76894663e-02 -4.57848936e-01 -1.11865056e+00 8.96640480e-01
4.01257306e-01 -1.02732563e+00 2.51488090e-01 -2.72724509e-01
6.01208925e-01 -1.92224279e-01 9.13579389e-02 2.91925400e-01
1.28024384e-01 -1.10990512e+00 8.87337983e-01 6.66211307e-01
6.23197317e-01 -4.31364059e-01 6.12484932e-01 6.79054707e-02
-1.37741399e+00 2.03630328e-02 -1.77139103e-01 1.38741238e-02
2.68633634e-01 3.25178169e-02 -4.18821484e-01 2.95999438e-01
7.97687948e-01 5.48896551e-01 -8.91121328e-01 9.91925895e-01
-7.67209589e-01 2.75379688e-01 7.11649805e-02 1.17610302e-02
2.55839348e-01 1.77432820e-02 2.70017177e-01 1.37135136e+00
-4.21192311e-02 2.06775777e-02 5.23767993e-02 1.36149931e+00
-1.66890740e-01 6.83472976e-02 -2.42495298e-01 5.19120812e-01
2.08687544e-01 1.50679207e+00 -5.39583743e-01 -4.99366581e-01
-6.95016205e-01 1.43535256e+00 6.97660267e-01 6.18981838e-01
-8.66462231e-01 -4.67473030e-01 6.86171174e-01 2.44214356e-01
5.17655134e-01 -7.20254257e-02 -3.15737695e-01 -1.10571718e+00
7.79105574e-02 -4.83951211e-01 5.46931982e-01 -1.40094459e+00
-1.14129901e+00 7.48229802e-01 1.61068350e-01 -9.64002907e-01
-2.13720035e-02 -7.68570185e-01 -9.39067066e-01 1.16300428e+00
-1.61422586e+00 -1.42489374e+00 -5.51963806e-01 5.16261518e-01
6.80146158e-01 1.89374641e-01 9.47649479e-01 1.63304922e-03
-6.01413429e-01 6.38907611e-01 -5.47867179e-01 2.87133068e-01
7.72574723e-01 -1.31215715e+00 3.16020191e-01 7.35693395e-01
5.16577102e-02 8.26891184e-01 5.49391627e-01 -2.12475032e-01
-8.63384008e-01 -9.36312854e-01 9.36314106e-01 -3.54434609e-01
5.39977729e-01 -4.30022001e-01 -1.05726683e+00 8.16738248e-01
7.22694218e-01 8.77702702e-03 5.72478235e-01 1.23024635e-01
-6.52435303e-01 -6.74326494e-02 -9.32502031e-01 7.51540363e-01
7.94912994e-01 -4.27837402e-01 -6.47425652e-01 1.29809991e-01
6.81422949e-01 -4.27803814e-01 -6.03151977e-01 2.20939949e-01
3.21288526e-01 -9.13946390e-01 8.66490245e-01 -5.69550812e-01
6.45617783e-01 -4.53812867e-01 -3.22409719e-02 -1.08656347e+00
-6.33033574e-01 -2.94232398e-01 2.49970220e-02 1.22409689e+00
6.53913498e-01 -3.03471476e-01 3.97943020e-01 7.94451296e-01
-1.62116960e-02 -9.39962506e-01 -7.30151594e-01 -2.17770174e-01
1.44567147e-01 -2.06125289e-01 4.02997285e-01 6.90655410e-01
2.52085477e-01 8.70720506e-01 -8.03195089e-02 1.69923708e-01
2.50150770e-01 2.73363501e-01 7.32626438e-01 -7.95671940e-01
-2.30668992e-01 -9.20082450e-01 -4.85460937e-01 -1.44144976e+00
3.13322008e-01 -8.46974730e-01 -5.52318571e-03 -1.76022112e+00
7.18514740e-01 1.11550011e-01 -3.38630438e-01 7.11319566e-01
-5.72358429e-01 6.25749826e-02 3.93197894e-01 -3.09156440e-02
-1.01371384e+00 5.64512730e-01 1.21870863e+00 -4.21510935e-01
-2.94879191e-02 -2.85800308e-01 -9.28170621e-01 7.95098901e-01
6.18948877e-01 -8.97758678e-02 -1.78913638e-01 -9.25738454e-01
4.01945747e-02 -2.10210487e-01 5.27706861e-01 -8.46094787e-01
2.94053227e-01 -5.66093586e-02 5.73200524e-01 -4.30476367e-01
3.25696021e-01 -7.54350364e-01 -3.21762353e-01 1.37892008e-01
-6.20446026e-01 3.54984224e-01 3.80540162e-01 3.50711614e-01
-3.91804785e-01 -9.78998616e-02 8.83236349e-01 -2.84862101e-01
-1.00076604e+00 1.66042000e-01 -4.90082771e-01 -1.89930409e-01
1.14715457e+00 -1.80394426e-01 -2.61052758e-01 -6.06911600e-01
-8.18734646e-01 6.40709817e-01 6.12200320e-01 5.37462652e-01
6.88108385e-01 -1.22376513e+00 -6.56934142e-01 8.78695548e-02
5.16248941e-01 1.53485358e-01 4.05966461e-01 7.62465000e-01
-3.44540775e-01 2.48099163e-01 -1.71840131e-01 -6.83556497e-01
-1.37700009e+00 6.72005892e-01 6.29561126e-01 -2.47222215e-01
-3.37487161e-01 1.08873832e+00 9.14171636e-01 -2.05156624e-01
5.16496301e-01 -4.70372736e-01 -3.39342237e-01 -8.20866078e-02
6.26320541e-01 -3.07387054e-01 -2.82330066e-01 -9.17164862e-01
-4.01414186e-01 9.29348052e-01 -1.19147673e-01 -7.73813501e-02
1.03761744e+00 -2.84528613e-01 -3.60386521e-01 3.90340686e-01
1.04412746e+00 -2.25457966e-01 -1.30663884e+00 -4.76375520e-01
-1.29151806e-01 -2.68344462e-01 6.06887229e-03 -1.11066639e+00
-1.15084267e+00 1.21993744e+00 5.81847310e-01 -1.70485854e-01
1.54929090e+00 5.26832640e-01 3.43331277e-01 9.05323923e-02
6.45676553e-02 -7.80512989e-01 3.66082013e-01 4.75875556e-01
1.31763089e+00 -1.15606272e+00 -6.82007521e-02 -3.55869353e-01
-7.86716700e-01 9.92640138e-01 1.14459395e+00 2.65342072e-02
4.49555427e-01 2.32321441e-01 2.07831860e-01 -2.43253857e-01
-9.54861283e-01 -4.65085387e-01 4.57431942e-01 5.67418933e-01
8.11894238e-01 -1.05996214e-01 -1.28415704e-01 6.77520573e-01
3.20958495e-01 -2.86210090e-01 1.04760207e-01 8.86364877e-01
-6.68491542e-01 -1.03755951e+00 -4.31760311e-01 1.05557501e-01
-3.75985771e-01 -9.02360529e-02 -6.76286399e-01 6.13642097e-01
2.51702577e-01 6.89762771e-01 4.45484251e-01 4.96121729e-03
6.28093958e-01 -7.79627487e-02 5.18214762e-01 -7.84155130e-01
-7.56929636e-01 1.04772948e-01 -2.48697937e-01 -5.81060350e-01
-5.58530629e-01 -4.40129906e-01 -1.41496575e+00 1.78626418e-01
-5.98905832e-02 -2.96774924e-01 3.12722653e-01 6.67759955e-01
4.23408866e-01 9.27428305e-01 1.29294291e-01 -9.58869219e-01
-3.01334970e-02 -1.14254212e+00 -1.67935312e-01 6.53417230e-01
3.41329366e-01 -4.79767770e-01 6.12579808e-02 2.91532397e-01] | [10.458418846130371, 1.392624020576477] |
71eda82f-9fc0-401b-98d6-1ac6e9772a5a | idea-interpretable-dynamic-ensemble | 2201.05336 | null | https://arxiv.org/abs/2201.05336v1 | https://arxiv.org/pdf/2201.05336v1.pdf | IDEA: Interpretable Dynamic Ensemble Architecture for Time Series Prediction | We enhance the accuracy and generalization of univariate time series point prediction by an explainable ensemble on the fly. We propose an Interpretable Dynamic Ensemble Architecture (IDEA), in which interpretable base learners give predictions independently with sparse communication as a group. The model is composed of several sequentially stacked groups connected by group backcast residuals and recurrent input competition. Ensemble driven by end-to-end training both horizontally and vertically brings state-of-the-art (SOTA) performances. Forecast accuracy improves by 2.6% over the best statistical benchmark on the TOURISM dataset and 2% over the best deep learning benchmark on the M4 dataset. The architecture enjoys several advantages, being applicable to time series from various domains, explainable to users with specialized modular structure and robust to changes in task distribution. | ['Tong Zhang', 'Kani Chen', 'Mengyue Zha'] | 2022-01-14 | null | null | null | null | ['time-series-prediction'] | ['time-series'] | [ 3.75419781e-02 4.55029607e-01 8.03880841e-02 -7.54159451e-01
-6.40975714e-01 -5.90945423e-01 7.16602564e-01 -2.69170552e-01
1.67844981e-01 9.06885028e-01 5.20548224e-01 -5.01197934e-01
-4.02224302e-01 -3.53265136e-01 -9.76299882e-01 -6.68773949e-01
-8.33117664e-01 7.66924977e-01 -5.12215436e-01 -6.84203744e-01
-1.02344573e-01 1.11088166e-02 -1.58946323e+00 7.45459557e-01
1.26065826e+00 1.37234116e+00 -1.59828395e-01 4.78455245e-01
1.15972519e-01 1.10413969e+00 -6.09261394e-01 -2.88397849e-01
2.01559559e-01 -2.66398728e-01 -5.29578745e-01 -1.19735070e-01
1.88724592e-01 -7.51985516e-03 2.41505150e-02 1.58240944e-01
3.04871053e-01 3.63143742e-01 6.98167801e-01 -1.32586873e+00
-1.06782424e+00 1.09597170e+00 -2.13927135e-01 4.75648195e-01
-1.89178377e-01 -6.56702137e-03 1.21818030e+00 -1.10708058e+00
2.34481856e-01 1.07174659e+00 9.39839423e-01 5.39327800e-01
-1.60994840e+00 -6.93980753e-01 8.56913090e-01 3.53504568e-01
-7.32509673e-01 -2.37468228e-01 3.42350662e-01 -4.04020995e-01
1.35619915e+00 7.45281696e-01 4.22148019e-01 1.61628628e+00
6.04453683e-01 7.03414023e-01 8.89481544e-01 1.80391967e-01
4.69901621e-01 -2.53475785e-01 3.26102853e-01 1.96461260e-01
-2.09811360e-01 2.34558076e-01 -6.62109673e-01 3.10507584e-02
2.85914361e-01 3.84525299e-01 -2.00953737e-01 2.53524780e-01
-1.42335212e+00 7.83033490e-01 7.78100729e-01 -1.83061324e-02
-7.74447799e-01 2.75385797e-01 4.22663659e-01 7.57636845e-01
1.36368608e+00 7.70860434e-01 -1.17471147e+00 3.34557262e-03
-7.41112769e-01 1.84664473e-01 8.62193465e-01 5.30726969e-01
2.91391253e-01 7.67715454e-01 2.71978620e-02 3.70921910e-01
-8.29809681e-02 7.07665145e-01 7.65066624e-01 -7.18209505e-01
5.64893723e-01 4.21475440e-01 -2.01580934e-02 -6.99899018e-01
-6.90753818e-01 -1.27629614e+00 -1.39158630e+00 1.63940281e-01
-7.34109506e-02 -6.58204556e-01 -1.11881733e+00 1.59285808e+00
-2.35552877e-01 6.98361695e-01 1.58550441e-01 7.28844523e-01
8.98155868e-01 1.10191607e+00 1.25270233e-01 -9.79469121e-02
9.38879311e-01 -1.21366549e+00 -5.34532547e-01 -3.35355133e-01
6.01068556e-01 -1.65433571e-01 5.71454942e-01 6.72169268e-01
-8.28966856e-01 -9.29464459e-01 -8.49069297e-01 1.91576198e-01
-4.58539516e-01 8.42618197e-02 7.26299107e-01 6.96831718e-02
-1.29881668e+00 8.11821818e-01 -1.15041101e+00 6.87979981e-02
2.02529088e-01 7.18881965e-01 -4.45756227e-01 4.69646692e-01
-1.10101068e+00 9.18552160e-01 4.71153438e-01 3.40538412e-01
-9.81114149e-01 -1.10722375e+00 -5.86379230e-01 2.45697543e-01
-1.42075926e-01 -1.16618586e+00 1.18632197e+00 -1.40962529e+00
-1.51722693e+00 1.67853460e-01 -4.68657225e-01 -1.21864903e+00
5.53586066e-01 -7.29030550e-01 -8.54639888e-01 -5.94044268e-01
-1.22296982e-01 5.56761861e-01 1.01664281e+00 -1.04257655e+00
-7.66621232e-01 -5.06694436e-01 -4.24321055e-01 2.67196111e-02
-7.75931105e-02 -4.68576938e-01 4.65692580e-01 -9.06199872e-01
1.22060932e-01 -1.15707922e+00 -5.92200696e-01 -9.05490994e-01
-3.91850144e-01 -3.41366887e-01 9.65166450e-01 -9.60316479e-01
1.25600684e+00 -1.96286690e+00 5.95360696e-01 1.36350185e-01
4.04591113e-01 -3.02497119e-01 -3.77088279e-01 5.77538013e-01
-6.61899030e-01 1.72741696e-01 -5.26726097e-02 -9.44405735e-01
6.11248501e-02 3.84501219e-01 -1.37954152e+00 -5.55690527e-02
2.88904727e-01 1.00464892e+00 -7.01176882e-01 6.95598662e-01
5.69294468e-02 5.53959072e-01 -3.71869475e-01 1.97573960e-01
-3.74357849e-01 1.04448032e+00 -3.56108665e-01 2.79335916e-01
4.10547435e-01 -7.46197522e-01 5.80576546e-02 1.61585599e-01
-6.99877292e-02 3.01274925e-01 -6.63916886e-01 1.29702020e+00
-7.37110198e-01 7.18112409e-01 -5.25839984e-01 -1.06232381e+00
1.17528141e+00 4.48401004e-01 3.15783054e-01 -4.87135291e-01
-3.44217509e-01 4.85557824e-01 -1.30126914e-02 -9.69024375e-02
6.14069879e-01 1.12438723e-01 -4.47630100e-02 4.98514116e-01
2.19041705e-01 3.79059166e-01 -2.94349253e-01 -1.49068639e-01
8.69509101e-01 1.38257205e-01 1.10858649e-01 -4.93484646e-01
3.49964082e-01 -2.64743745e-01 4.03699815e-01 7.47874200e-01
6.81116760e-01 5.61643898e-01 2.90784150e-01 -1.38538861e+00
-9.71117020e-01 -1.03404570e+00 -8.87434334e-02 1.79772425e+00
-2.36479953e-01 -4.25778985e-01 -2.28607222e-01 -6.74333811e-01
1.83544993e-01 1.35918629e+00 -1.17757380e+00 -8.18648413e-02
-4.81765747e-01 -6.67955995e-01 1.84277743e-02 9.52675164e-01
3.26030642e-01 -1.02924109e+00 -2.38636062e-01 5.06323457e-01
-3.01542997e-01 -8.20266724e-01 -3.68675333e-04 6.57609046e-01
-1.32805240e+00 -2.56905437e-01 -4.35697526e-01 -1.92746535e-01
3.87809038e-01 -5.78763112e-02 1.59945381e+00 -1.78489715e-01
5.11867404e-01 1.85871989e-01 -3.63648653e-01 -9.69056189e-01
-8.16593766e-02 6.56912744e-01 2.71881998e-01 3.33963096e-01
-1.94865186e-02 -8.88132632e-01 -7.08188713e-01 6.10081434e-01
-4.59391266e-01 2.11534187e-01 1.51718080e-01 1.01281941e+00
4.95965511e-01 -2.98880577e-01 9.37885702e-01 -6.82422578e-01
4.83915716e-01 -1.14226556e+00 -3.05168480e-01 7.30323493e-02
-8.31959784e-01 3.32398176e-01 1.05498433e+00 -1.60166934e-01
-9.74366486e-01 -2.67322272e-01 2.18368769e-01 -3.47751230e-01
-9.31110233e-02 6.75920367e-01 6.17291272e-01 3.84651691e-01
8.75033438e-01 4.95990932e-01 1.38099670e-01 -5.33102393e-01
3.89785588e-01 3.04899365e-01 5.47321737e-01 -3.20828944e-01
5.75501740e-01 4.48330075e-01 1.63571443e-02 -3.12420636e-01
-1.16562188e+00 2.69164741e-02 -4.57720965e-01 -1.57202154e-01
6.21582270e-01 -1.39446926e+00 -6.67402565e-01 9.78310034e-02
-1.10989928e+00 -4.30734515e-01 -1.74867794e-01 5.32622814e-01
-6.09812140e-01 -5.11430621e-01 -3.74741465e-01 -7.15362430e-01
-9.09453392e-01 -8.41382980e-01 1.31525302e+00 -1.76322013e-01
-6.10952556e-01 -1.44861615e+00 -7.15774521e-02 3.45816910e-02
8.66314352e-01 6.46282852e-01 6.04003251e-01 -1.17975545e+00
-6.34812713e-01 -1.75402328e-01 3.26613098e-01 3.95690084e-01
-1.16021089e-01 -1.90986872e-01 -1.12313676e+00 -2.53172308e-01
-1.35062054e-01 -4.55923006e-02 1.36587131e+00 6.40949130e-01
1.69157636e+00 -5.24746001e-01 -4.04536813e-01 8.07870924e-01
7.92501569e-01 -9.95152257e-03 4.10127431e-01 3.06342185e-01
5.42765260e-01 5.83995223e-01 5.26129343e-02 6.03173316e-01
5.87081134e-01 2.48448730e-01 8.69179249e-01 -5.09103015e-02
3.91584903e-01 -1.97018549e-01 5.88737667e-01 1.01977253e+00
-9.40214455e-01 -5.56486130e-01 -8.97330880e-01 2.27580860e-01
-2.36867976e+00 -1.36697996e+00 -3.37924451e-01 1.94811821e+00
1.75605655e-01 4.50620130e-02 1.06732145e-01 -4.52861860e-02
3.27358484e-01 3.03139776e-01 -8.60096753e-01 -3.84040505e-01
-4.86462921e-01 1.21946819e-01 4.01806772e-01 2.91151553e-01
-1.14659572e+00 4.02357399e-01 6.81632328e+00 5.66457808e-01
-1.36339009e+00 -3.80652212e-02 1.21559596e+00 -3.11445206e-01
-5.99370956e-01 -2.11715743e-01 -6.98428810e-01 5.77693403e-01
1.55295658e+00 -4.04270798e-01 5.48596680e-01 9.28532720e-01
2.59560585e-01 7.34977424e-01 -1.44992495e+00 7.96432853e-01
-1.14328265e-01 -1.69733548e+00 5.38046807e-02 2.00976551e-01
1.17729545e+00 7.41973400e-01 8.69675815e-01 8.44941020e-01
4.40548122e-01 -1.53504550e+00 8.75214398e-01 9.46357548e-01
4.68820065e-01 -7.46233165e-01 9.09139156e-01 5.38432956e-01
-9.70235765e-01 -5.40304124e-01 -1.09861217e-01 -4.45027858e-01
1.31567031e-01 3.92234743e-01 -6.74358904e-01 8.13060999e-01
9.10982490e-01 1.39355981e+00 -4.68441099e-01 6.62939072e-01
-1.76186450e-02 1.17634416e+00 -3.16952884e-01 1.11348137e-01
4.78055596e-01 -2.41825543e-02 6.08296096e-01 8.81306469e-01
8.44074488e-01 1.51065633e-01 -3.30343060e-02 4.52933043e-01
-5.95356226e-02 -2.17565373e-01 -6.14873409e-01 4.28241104e-01
1.11411579e-01 1.05111134e+00 -1.31750748e-01 -4.18574303e-01
-1.17237046e-02 8.68416846e-01 3.59118789e-01 6.54940665e-01
-8.12695920e-01 2.60127962e-01 6.86140716e-01 -1.05236925e-01
6.46005869e-01 4.54342254e-02 -7.05366433e-01 -1.25423467e+00
-1.86473623e-01 -9.54120755e-01 6.38234854e-01 -1.01751685e+00
-1.79033864e+00 1.32035136e+00 -2.94971734e-01 -1.63292515e+00
-7.98220515e-01 -7.02080429e-01 -8.88154089e-01 9.09227371e-01
-1.12029874e+00 -1.24445951e+00 -3.71497720e-01 3.09671760e-01
8.70036364e-01 -6.40842795e-01 1.30452609e+00 -2.44087383e-01
-4.50779200e-01 4.53407854e-01 6.74613953e-01 -3.68764192e-01
4.48601902e-01 -1.47977293e+00 9.87094760e-01 3.79255503e-01
1.94146574e-01 4.95243013e-01 9.50249970e-01 -2.78516531e-01
-1.00252080e+00 -1.49621844e+00 9.12377834e-01 -1.14587581e+00
7.67631888e-01 -2.85649598e-01 -1.14225733e+00 1.24353337e+00
6.68559611e-01 1.28692137e-02 9.50579226e-01 9.35251117e-01
-5.08130133e-01 -5.28091490e-01 -5.92163980e-01 3.91569525e-01
1.03990173e+00 -2.12466985e-01 -6.56973362e-01 5.58672905e-01
8.84956241e-01 -6.33556962e-01 -9.94189620e-01 4.52118874e-01
5.81110895e-01 -9.94530082e-01 7.36874282e-01 -1.22952855e+00
8.43660116e-01 -4.24563922e-02 -1.52072340e-01 -2.15911198e+00
-6.68859601e-01 -9.49910462e-01 -5.26346326e-01 4.46448207e-01
1.01110053e+00 -1.12473094e+00 3.65012050e-01 5.77540159e-01
-6.86926305e-01 -8.61182690e-01 -7.76234627e-01 -8.89214218e-01
3.91454101e-02 -6.68766499e-01 1.23382080e+00 8.28641355e-01
-6.70258002e-03 6.21648788e-01 -7.56769300e-01 3.15387517e-01
5.80065906e-01 3.73370916e-01 7.54946768e-01 -1.73312533e+00
-3.79102081e-01 -4.57912952e-01 -2.57891446e-01 -1.02216780e+00
3.98311526e-01 -1.01204562e+00 -4.57676440e-01 -1.12451792e+00
-3.50488275e-01 1.29562160e-02 -7.97849596e-01 5.13996005e-01
-2.00816900e-01 2.95417048e-02 2.03531995e-01 4.54754114e-01
-8.28734994e-01 8.87902975e-01 8.53079736e-01 2.03226116e-02
-2.13000000e-01 4.59977716e-01 -7.30671644e-01 5.82084656e-01
8.40039074e-01 -2.72137642e-01 -4.34637249e-01 -8.80839407e-01
2.97121555e-01 3.12940806e-01 5.12204885e-01 -9.61046576e-01
2.09202934e-02 2.66315103e-01 6.21188521e-01 -8.10192347e-01
3.14282894e-01 -7.83805072e-01 6.27145529e-01 3.51388395e-01
-6.50871813e-01 6.88820481e-01 4.22940373e-01 9.58981276e-01
-4.31699038e-01 7.08279252e-01 6.76080510e-02 1.41181946e-01
-8.59490216e-01 4.79205579e-01 -3.41966599e-02 -1.93638697e-01
9.00634944e-01 -9.42612290e-02 -4.77816910e-01 -8.54027927e-01
-1.28529775e+00 5.24378836e-01 -4.61157560e-01 7.80827105e-01
2.65259296e-01 -1.39572656e+00 -1.37545097e+00 3.29261869e-01
1.13781869e-01 -2.08369553e-01 4.92393494e-01 9.60107148e-01
-5.64902499e-02 6.06683195e-01 -1.13479190e-01 -8.73346388e-01
-7.16519654e-01 1.84860036e-01 5.06160378e-01 -3.58629525e-01
-6.06394649e-01 9.71311510e-01 3.74978065e-01 -7.09589839e-01
-9.75095667e-03 -6.34680688e-01 -2.47330606e-01 2.68966239e-02
6.88645303e-01 5.63346148e-01 2.26983905e-01 -3.59860718e-01
-8.71969312e-02 2.59243727e-01 -6.58623055e-02 2.56439030e-01
1.99569190e+00 -1.09233126e-01 1.72654223e-02 8.28772664e-01
9.52243209e-01 -5.49117684e-01 -1.33262110e+00 -1.72094211e-01
-1.03111476e-01 2.35149801e-01 7.76124420e-03 -1.22024727e+00
-1.00345027e+00 7.42785394e-01 3.60839576e-01 7.54919052e-01
1.16974914e+00 -1.98170125e-01 5.07945120e-01 5.79816043e-01
2.07434967e-01 -6.28359079e-01 -1.79710299e-01 9.12983477e-01
1.59638846e+00 -1.38767135e+00 -2.75592089e-01 2.49003381e-01
-1.05831861e+00 1.38361561e+00 3.56955111e-01 -4.02516454e-01
9.14199829e-01 -9.82089639e-02 5.91297373e-02 -7.11005852e-02
-1.81032932e+00 2.44546399e-01 9.17231262e-01 3.84012967e-01
4.65894669e-01 4.02573049e-01 2.64584780e-01 9.72525418e-01
-5.98081231e-01 -1.69044852e-01 -4.52156663e-02 -6.31514043e-02
-1.45259663e-01 -3.39467496e-01 -1.94931284e-01 7.09613740e-01
-3.76374930e-01 -1.54083624e-01 -7.59146214e-02 7.39341736e-01
-9.61675793e-02 1.02515888e+00 4.75174099e-01 -6.60597861e-01
4.25284386e-01 2.79564619e-01 -3.04815382e-01 -1.97186783e-01
-1.05744612e+00 -2.29961485e-01 1.76241845e-01 -6.33948863e-01
-5.34323715e-02 -6.34978712e-01 -9.83201325e-01 -5.31593680e-01
1.67335600e-01 2.61208415e-01 3.73535097e-01 9.95951176e-01
1.07155359e+00 8.40244830e-01 9.70722973e-01 -1.09623396e+00
-8.67278695e-01 -1.25626755e+00 -4.34360355e-01 2.43788362e-01
8.11485648e-01 -3.47142071e-01 -6.68400347e-01 2.02658191e-01] | [6.9954986572265625, 3.0992963314056396] |
1b73c128-a9d0-4841-a6e2-52ba4cce3c4f | captioning-near-future-activity-sequences | 1908.00943 | null | https://arxiv.org/abs/1908.00943v5 | https://arxiv.org/pdf/1908.00943v5.pdf | Prediction and Description of Near-Future Activities in Video | Most of the existing works on human activity analysis focus on recognition or early recognition of the activity labels from complete or partial observations. Similarly, almost all of the existing video captioning approaches focus on the observed events in videos. Predicting the labels and the captions of future activities where no frames of the predicted activities have been observed is a challenging problem, with important applications that require anticipatory response. In this work, we propose a system that can infer the labels and the captions of a sequence of future activities. Our proposed network for label prediction of a future activity sequence has three branches where the first branch takes visual features from the objects present in the scene, the second branch takes observed sequential activity features, and the third branch captures the last observed activity features. The predicted labels and the observed scene context are then mapped to meaningful captions using a sequence-to-sequence learning-based method. Experiments on four challenging activity analysis datasets and a video description dataset demonstrate that our label prediction approach achieves comparable performance with the state-of-the-arts and our captioning framework outperform the state-of-the-arts. | ['Amit K. Roy-Chowdhury', 'Mahmudul Hasan', 'Tahmida Mahmud', 'Mohammad Billah'] | 2019-08-02 | null | null | null | null | ['video-description'] | ['computer-vision'] | [ 9.24061358e-01 -3.69403698e-02 -6.34909749e-01 -6.24481916e-01
-4.93797988e-01 -4.19216543e-01 8.22432816e-01 7.29614720e-02
-1.72964856e-01 8.07834685e-01 6.67401910e-01 1.63372755e-01
2.39594921e-01 -1.38318256e-01 -8.33383441e-01 -4.58590060e-01
-3.24938655e-01 2.59233057e-01 6.70341969e-01 3.21004719e-01
3.21459740e-01 2.33757690e-01 -1.58188093e+00 8.40682685e-01
1.76997378e-01 1.21563613e+00 3.10603142e-01 7.88903773e-01
-2.05445483e-01 1.77479231e+00 -3.87742072e-01 1.57562748e-01
2.83144936e-02 -8.76061499e-01 -8.81450474e-01 5.89789033e-01
4.10681069e-01 -4.86636370e-01 -6.10131621e-01 5.72005272e-01
-6.79224879e-02 2.71061361e-01 5.68784893e-01 -1.70424354e+00
-3.31999600e-01 3.14197153e-01 -2.70070881e-01 5.50340533e-01
9.42263067e-01 1.48954034e-01 8.61087501e-01 -7.88108230e-01
7.46564507e-01 8.53277445e-01 3.48183751e-01 6.03088796e-01
-6.99302733e-01 -4.96816337e-01 5.46343565e-01 7.07505822e-01
-1.34007347e+00 -4.70659167e-01 7.77976215e-01 -8.98374200e-01
8.97218406e-01 2.60195080e-02 7.41259158e-01 1.24750364e+00
-1.90383345e-01 1.01246119e+00 6.56328797e-01 -3.03341240e-01
3.56648237e-01 -2.80591577e-01 -5.00528365e-02 7.70742297e-01
-2.14873195e-01 -1.42363042e-01 -9.12231028e-01 -1.53458521e-01
6.47363424e-01 4.63554233e-01 -4.03544694e-01 -3.92140806e-01
-1.66806972e+00 3.27893049e-01 4.78477366e-02 1.50208265e-01
-6.73631310e-01 5.05996287e-01 5.57647526e-01 -1.58384800e-01
3.96910757e-01 1.41149282e-01 -4.79703605e-01 -4.54521090e-01
-1.09366369e+00 3.87075394e-02 7.31058240e-01 1.02661347e+00
7.50199199e-01 -2.54203767e-01 -5.10609627e-01 3.88112754e-01
2.72453874e-01 1.27004012e-01 4.00226474e-01 -1.21513653e+00
6.73178911e-01 6.44376040e-01 4.47229356e-01 -8.94602239e-01
-1.80825859e-01 4.36955206e-02 -2.40032315e-01 -3.61787289e-01
5.01258373e-01 -9.62963179e-02 -8.33900690e-01 1.52793288e+00
2.14288816e-01 1.06137180e+00 2.72651948e-03 9.56435442e-01
5.97488225e-01 1.10710645e+00 3.76166940e-01 -4.73316103e-01
1.20919907e+00 -1.38088870e+00 -8.55223238e-01 -6.04269147e-01
7.38192201e-01 -4.66680527e-01 6.25922799e-01 1.38783216e-01
-6.35852695e-01 -7.26957321e-01 -9.22823131e-01 2.13638201e-01
-9.82175767e-03 3.23351741e-01 4.38466132e-01 -8.75658020e-02
-8.65124524e-01 4.73842919e-01 -1.02742779e+00 -5.89385390e-01
4.11597788e-01 1.10277124e-01 -3.68021280e-01 -1.38059497e-01
-7.50772595e-01 6.12739623e-01 7.11873710e-01 -2.51310840e-02
-1.43432832e+00 -3.29345196e-01 -9.56012070e-01 1.48954213e-01
7.27323294e-01 -2.09157363e-01 1.42673278e+00 -1.43180287e+00
-1.21094668e+00 7.10995972e-01 -5.74626744e-01 -5.64923346e-01
2.66004503e-01 -3.06963086e-01 -5.17483056e-01 4.72744733e-01
1.27757534e-01 8.51179600e-01 4.80314881e-01 -1.06225586e+00
-1.25808680e+00 6.07486404e-02 -2.05769931e-04 4.23727095e-01
-2.22331136e-01 1.69446751e-01 -5.74159443e-01 -3.87440830e-01
-1.62483260e-01 -1.04451370e+00 4.24282327e-02 1.77476957e-01
-1.52562231e-01 -4.30390209e-01 9.73511875e-01 -7.68938780e-01
1.19165027e+00 -2.23648715e+00 -7.54552856e-02 -1.93812400e-01
-7.99026340e-02 9.12647843e-02 -8.90825167e-02 6.77459717e-01
-1.33675173e-01 -3.05611253e-01 3.20705362e-02 -1.97414517e-01
-1.89812779e-01 3.09693366e-01 -5.57632446e-01 4.72989857e-01
2.05830917e-01 7.03261137e-01 -1.25274873e+00 -8.70845020e-01
2.74166614e-01 2.24064827e-01 -2.05689177e-01 8.35943818e-01
-5.52756906e-01 8.21632266e-01 -5.56272984e-01 4.40397680e-01
-1.82284817e-01 -6.70279205e-01 4.33326542e-01 -3.69154066e-02
-1.16530210e-01 2.15188369e-01 -8.77731621e-01 1.89212370e+00
-5.38446195e-02 9.11768019e-01 -6.50387287e-01 -1.11030531e+00
6.58458710e-01 7.16820776e-01 9.21268463e-01 -3.53857458e-01
-1.58374891e-01 6.84164539e-02 -3.42625886e-01 -1.02718031e+00
3.72665197e-01 3.23800176e-01 -6.81704208e-02 6.05775297e-01
1.63490742e-01 7.59928048e-01 6.02822959e-01 2.10241213e-01
1.35171115e+00 6.79266512e-01 4.91820425e-01 4.70643193e-01
7.16879785e-01 1.44264057e-01 7.76412487e-01 6.66839719e-01
-5.48987806e-01 5.51992655e-01 5.17729104e-01 -7.59702682e-01
-9.66413617e-01 -7.86570787e-01 7.18552113e-01 1.53480828e+00
2.95460045e-01 -5.00474632e-01 -7.34699249e-01 -1.05690145e+00
-6.08159125e-01 5.82887113e-01 -6.50947273e-01 4.09987457e-02
-7.92695999e-01 2.86685340e-02 2.77096778e-01 7.60505438e-01
5.38321853e-01 -1.63823175e+00 -8.32181156e-01 1.23516835e-01
-8.52009237e-01 -1.53480184e+00 -7.16053009e-01 -1.40951663e-01
-6.53267145e-01 -1.19840956e+00 -5.84548950e-01 -1.07049692e+00
7.04564095e-01 2.10753053e-01 1.05171871e+00 1.10604607e-01
8.53427574e-02 4.99820054e-01 -5.45920372e-01 -1.65850282e-01
-2.95152903e-01 -2.18048006e-01 -9.05179754e-02 5.24034798e-01
5.05825579e-01 -3.99460912e-01 -7.19883442e-01 4.82224792e-01
-6.56795681e-01 4.68902051e-01 4.83948231e-01 3.37512255e-01
7.77070820e-01 -3.53754252e-01 3.67377371e-01 -5.72104156e-01
-4.25048284e-02 -5.52731752e-01 -1.80969343e-01 5.74132025e-01
-1.88809022e-01 4.49964292e-02 5.28013766e-01 -6.67835951e-01
-1.08570194e+00 7.96745956e-01 3.12816799e-01 -4.62257415e-01
-5.91178298e-01 3.62935752e-01 -7.18777776e-02 4.94433671e-01
4.44695294e-01 4.37210828e-01 -4.22800869e-01 -3.42287540e-01
7.21848309e-02 7.11836874e-01 6.93316042e-01 -3.12987119e-01
2.76666701e-01 6.97065830e-01 -1.11625560e-01 -5.37825882e-01
-1.39242864e+00 -9.66522515e-01 -7.21106470e-01 -8.47797275e-01
1.19438648e+00 -1.09871864e+00 -4.97128040e-01 2.34108225e-01
-1.41343498e+00 -3.67066830e-01 -3.06029767e-01 6.05017722e-01
-1.09403408e+00 4.62608367e-01 -4.61574048e-01 -7.95468390e-01
-5.21729700e-03 -9.64260042e-01 1.20339847e+00 3.73344690e-01
-5.08018553e-01 -7.52644420e-01 2.62517214e-01 5.60561836e-01
-1.10671543e-01 3.69112015e-01 4.54228193e-01 -8.74917030e-01
-7.46135414e-01 -1.98797315e-01 -1.16426945e-01 3.36759202e-02
2.04087928e-01 -2.23243371e-01 -7.51883090e-01 -4.40504029e-02
-2.97142297e-01 -3.05794001e-01 4.51014876e-01 3.19253176e-01
1.14004087e+00 -4.65351343e-01 -5.57889819e-01 2.47213706e-01
1.14661908e+00 5.42533934e-01 6.61581814e-01 2.26727113e-01
6.85479164e-01 5.99803448e-01 1.07655311e+00 6.07143402e-01
3.06275010e-01 7.58451343e-01 5.48392951e-01 -8.89460964e-04
-8.88089389e-02 -6.29675567e-01 6.72203898e-01 5.54125726e-01
-6.68164045e-02 -4.91447955e-01 -9.00318146e-01 8.28278065e-01
-2.51462245e+00 -1.38262963e+00 -1.90377459e-02 1.88032508e+00
5.10828078e-01 3.79479267e-02 1.82192937e-01 8.54526064e-04
1.11018825e+00 4.47939992e-01 -6.30269408e-01 -1.25130406e-02
4.04614300e-01 -4.85293180e-01 2.92621434e-01 -5.76213188e-02
-1.35112286e+00 7.47006834e-01 6.16102171e+00 3.70155424e-01
-7.07252264e-01 1.97758719e-01 4.55031097e-01 -1.87890336e-01
4.91530240e-01 2.18567550e-01 -6.57358527e-01 5.77053368e-01
1.05236471e+00 2.90921107e-02 2.22966641e-01 9.00927305e-01
5.78168452e-01 -3.30655813e-01 -1.69378161e+00 1.10891759e+00
5.29109776e-01 -1.34625649e+00 6.67899940e-03 -3.10939271e-02
7.37381399e-01 2.73945294e-02 -4.67537791e-01 1.54004633e-01
1.42290944e-03 -6.98604524e-01 1.01299465e+00 9.62107360e-01
6.66360736e-01 -4.36627686e-01 6.83763385e-01 6.52318954e-01
-1.35608900e+00 -3.04459840e-01 2.37294696e-02 -4.61819172e-01
7.32562423e-01 1.44898340e-01 -1.17246389e+00 6.58325031e-02
5.51805496e-01 1.27245641e+00 -6.08113050e-01 1.13225365e+00
-4.46210623e-01 9.53195393e-01 1.14863694e-01 -1.42430276e-01
2.23235905e-01 -2.66171806e-02 2.14228928e-01 1.33536291e+00
3.38615865e-01 1.28815919e-01 7.30910957e-01 3.36454272e-01
-8.00385997e-02 -7.30758607e-02 -5.99463224e-01 -3.22064072e-01
2.62835652e-01 9.89390790e-01 -8.49285245e-01 -7.17733145e-01
-5.67503929e-01 1.09752095e+00 2.01727033e-01 3.60489875e-01
-1.04445899e+00 -6.06841128e-03 4.02596027e-01 1.95297077e-01
4.92527634e-01 -1.19114712e-01 2.02095002e-01 -1.21579611e+00
1.34008765e-01 -5.73816001e-01 6.57545626e-01 -1.18645036e+00
-9.39034522e-01 3.33175600e-01 6.09484455e-03 -1.53222096e+00
-5.65093338e-01 -2.40250930e-01 -4.98790979e-01 2.36226290e-01
-1.22850466e+00 -1.12298155e+00 -5.76547027e-01 4.06515390e-01
1.02191281e+00 -5.12970239e-02 6.13396585e-01 1.97029158e-01
-4.20552820e-01 -1.95464805e-01 -1.72117531e-01 5.09730279e-01
5.69744051e-01 -8.44390750e-01 2.17357516e-01 9.89549756e-01
5.13723493e-01 -8.11707526e-02 7.85954595e-01 -8.51537108e-01
-8.92025650e-01 -1.24262023e+00 1.20461643e+00 -4.93621737e-01
6.38205826e-01 -3.62581283e-01 -8.14759910e-01 1.06422174e+00
2.38109484e-01 2.76023924e-01 8.33307087e-01 -4.00274903e-01
-1.14759691e-01 6.67604432e-02 -5.98150671e-01 3.78060967e-01
1.37716317e+00 -4.81298298e-01 -7.32081592e-01 5.99731982e-01
5.11717498e-01 -2.26070419e-01 -5.31800687e-01 1.55785516e-01
7.30628490e-01 -6.64328575e-01 7.67498016e-01 -8.28606963e-01
6.83983564e-01 -6.54940426e-01 -9.90718305e-02 -8.68069172e-01
-2.79811054e-01 -3.58087450e-01 -5.40136755e-01 1.18233132e+00
2.57213116e-01 3.13480496e-01 1.02165627e+00 5.04006326e-01
-1.90391660e-01 -3.83622527e-01 -7.63430178e-01 -6.34983361e-01
-1.07651162e+00 -2.94438630e-01 4.07827377e-01 8.94060194e-01
1.88913837e-01 3.58386368e-01 -8.63802493e-01 9.59585905e-02
3.29878360e-01 1.66375965e-01 7.61449516e-01 -9.53705370e-01
-2.65079468e-01 2.12904125e-01 -6.76603615e-01 -1.31299114e+00
3.60847920e-01 -7.05064654e-01 5.40529668e-01 -1.80745971e+00
6.07672513e-01 2.35640287e-01 -3.42970997e-01 6.58409953e-01
3.27586867e-02 2.05324486e-01 3.19678932e-01 5.01666725e-01
-1.59097385e+00 3.57062131e-01 8.83542597e-01 -1.81253031e-01
-1.38141975e-01 -1.57215938e-01 -1.55610433e-02 8.72904778e-01
6.03824317e-01 -7.67042279e-01 -5.12144327e-01 -3.30081969e-01
5.67304380e-02 4.78060544e-01 3.01102161e-01 -1.40952039e+00
3.13263655e-01 -6.31218791e-01 2.31272429e-01 -7.18946159e-01
5.27832448e-01 -9.19072449e-01 3.46271932e-01 4.15589869e-01
-8.06658626e-01 -1.10235274e-01 -3.29858094e-01 1.09516239e+00
-2.81052053e-01 -1.68046787e-01 3.76090646e-01 -1.48569524e-01
-1.26315856e+00 4.29621220e-01 -7.53535211e-01 8.72358680e-03
1.55096161e+00 -4.48836744e-01 -2.72450179e-01 -6.79060459e-01
-8.18472326e-01 2.68745720e-01 3.06713670e-01 6.79570258e-01
6.75083220e-01 -1.32009721e+00 -5.93690872e-01 -2.25121811e-01
5.66249251e-01 -2.54139811e-01 1.02621742e-01 7.02653944e-01
-4.80387568e-01 4.77931112e-01 -2.78257102e-01 -6.89705014e-01
-1.36584711e+00 7.15737522e-01 1.39112383e-01 -4.38829988e-01
-5.25702298e-01 4.53149945e-01 4.73667622e-01 2.80401975e-01
4.21125412e-01 -5.23367040e-02 -6.44856036e-01 -1.12828158e-01
7.75871098e-01 4.31200773e-01 -4.65931535e-01 -1.18587649e+00
-4.46897924e-01 2.16623902e-01 1.00698970e-01 -2.27073412e-02
1.30265737e+00 -2.12385401e-01 1.93996713e-01 7.51033962e-01
1.07613158e+00 -6.36876285e-01 -1.91546071e+00 -3.71637613e-01
3.85009140e-01 -4.28570628e-01 -3.61753732e-01 -7.20433414e-01
-8.97085309e-01 7.85412014e-01 5.82244337e-01 -6.67829588e-02
1.10110641e+00 3.17894787e-01 8.65276694e-01 3.75253201e-01
4.92509753e-01 -1.14750051e+00 5.55139005e-01 4.39661503e-01
5.46727121e-01 -1.09986889e+00 1.00621125e-02 -4.23469633e-01
-1.01273680e+00 9.09223258e-01 7.32733786e-01 1.31323367e-01
2.66636729e-01 -1.54511809e-01 -5.64263016e-02 -2.06815422e-01
-9.22064126e-01 -2.42469609e-01 3.09647024e-01 6.13564253e-01
2.38979355e-01 -1.66552320e-01 -1.60421908e-01 3.62604409e-01
5.00376284e-01 5.03970027e-01 6.47694528e-01 1.29231727e+00
-6.56373680e-01 -6.97552919e-01 -3.28606814e-01 4.88206983e-01
-5.15842140e-01 2.11682379e-01 -5.40450633e-01 2.56676793e-01
2.82987148e-01 1.10212708e+00 3.87446672e-01 -2.98919827e-01
2.40721181e-01 4.77399617e-01 2.45414063e-01 -8.43696773e-01
-2.57483512e-01 -7.77146742e-02 2.34154433e-01 -8.44830155e-01
-1.09643817e+00 -9.66153920e-01 -1.29017210e+00 2.46259660e-01
-1.30773589e-01 1.90743700e-01 4.70604300e-01 1.31877851e+00
4.66235638e-01 3.46026570e-01 4.61530179e-01 -9.32921290e-01
1.54020876e-01 -1.01457572e+00 -3.93798620e-01 8.13617527e-01
2.84651816e-01 -6.20146215e-01 -2.02733174e-01 1.07773173e+00] | [8.500168800354004, 0.6015111804008484] |
06775c4f-f626-4f2c-be73-62882290f698 | carbon-aware-ev-charging | 2209.12373 | null | https://arxiv.org/abs/2209.12373v1 | https://arxiv.org/pdf/2209.12373v1.pdf | Carbon-Aware EV Charging | This paper examines the problem of optimizing the charging pattern of electric vehicles (EV) by taking real-time electricity grid carbon intensity into consideration. The objective of the proposed charging scheme is to minimize the carbon emissions contributed by EV charging events, while simultaneously satisfying constraints posed by EV user's charging schedules, charging station transformer limits, and battery physical constraints. Using real-world EV charging data and California electricity generation records, this paper shows that our carbon-aware real-time charging scheme saves an average of 3.81% of carbon emission while delivering satisfactory amount of energy. Furthermore, by using an adaptive balanced factor, we can reduce 26.00% of carbon emission on average while compromising 12.61% of total energy delivered. | ['Yize Chen', 'Yuanyuan Shi', 'Yuexin Bian', 'Kai-Wen Cheng'] | 2022-09-26 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [-6.96912035e-02 -1.06412144e-02 -2.44399667e-01 -3.60181630e-01
-5.32804787e-01 -9.27505612e-01 5.94228446e-01 2.83475071e-01
-3.19506377e-01 9.08020198e-01 -1.87977239e-01 -5.94353795e-01
-3.67044091e-01 -1.32808232e+00 -5.49161434e-01 -8.37940991e-01
1.58329695e-01 4.68827844e-01 -3.82573634e-01 3.95852700e-02
3.50136995e-01 6.23515189e-01 -1.34630013e+00 -6.67377830e-01
1.46904278e+00 1.07501876e+00 5.89306414e-01 1.86837651e-02
-1.54726412e-02 -2.30130672e-01 -5.88694155e-01 1.05459370e-01
3.00227880e-01 -5.81718273e-02 -4.28043991e-01 -5.33464611e-01
-5.76199412e-01 -3.51773500e-01 2.51511484e-01 1.45714188e+00
1.70847714e-01 1.52639285e-01 6.03840649e-01 -1.97071397e+00
-2.22667560e-01 5.81850708e-01 -4.08312380e-01 8.33433419e-02
-5.14499009e-01 -5.98886609e-02 7.36780286e-01 -2.79022068e-01
-1.52686864e-01 6.53299332e-01 -3.93811055e-02 6.04080915e-01
-1.16184378e+00 -1.16089976e+00 -6.83292150e-02 3.79500151e-01
-1.35114574e+00 1.10325795e-02 8.27720821e-01 -7.93895721e-02
1.33545756e+00 8.12053800e-01 1.15545750e+00 -1.30430371e-01
5.77221632e-01 2.79551297e-01 9.14801300e-01 -2.43558437e-01
5.44807971e-01 5.96823841e-02 4.80053239e-02 -4.86414462e-01
8.51675928e-01 -1.10296801e-01 6.07348859e-01 -2.83828676e-01
-5.96926033e-01 7.23420549e-03 -1.84026167e-01 8.36171284e-02
-3.58559251e-01 5.53593695e-01 1.70783494e-02 3.34964573e-01
-4.57053065e-01 3.08722287e-01 2.12857738e-01 -3.97609264e-01
4.09170508e-01 3.47753704e-01 -5.12078643e-01 2.43259668e-02
-8.31197977e-01 2.44853184e-01 3.34131122e-01 1.19639575e+00
6.93619907e-01 3.52956533e-01 7.93559924e-02 3.68022472e-01
4.09981906e-01 1.34319234e+00 -2.03556448e-01 -1.11336458e+00
3.91770184e-01 2.97236919e-01 8.40493023e-01 -3.81466597e-01
-3.39882493e-01 -1.60164401e-01 -5.36106527e-01 4.23545396e-04
-3.20112199e-01 -4.33160812e-01 -6.26327634e-01 1.83242452e+00
-6.65137172e-02 -2.78503209e-01 -6.32143170e-02 3.47437620e-01
-9.82234180e-02 1.06557870e+00 5.91925263e-01 -9.25067902e-01
1.15137148e+00 -3.72584492e-01 -1.35200918e+00 3.07680130e-01
2.33427882e-01 -2.51939029e-01 4.67192799e-01 1.29443720e-01
-1.52570462e+00 1.16154522e-01 -9.42296982e-01 5.45948505e-01
-7.16267705e-01 -3.11984152e-01 2.02548161e-01 6.83916450e-01
-9.73321259e-01 3.51860374e-01 -5.34319222e-01 1.36456430e-01
2.09500208e-01 3.33762288e-01 5.41009367e-01 2.18531027e-01
-1.43754077e+00 1.18864918e+00 4.62163985e-01 9.78037715e-02
-8.44913542e-01 -1.29701924e+00 -5.22538066e-01 3.75126809e-01
1.58622697e-01 -7.91104790e-03 1.37700891e+00 -2.09003687e-01
-8.13078344e-01 1.20768070e-01 4.09343559e-03 -3.02862376e-01
4.14163858e-01 1.75420761e-01 -8.58494520e-01 2.90649142e-02
-1.24612814e-02 3.30986172e-01 -7.07441866e-02 -1.43224478e+00
-1.08397961e+00 -3.28439355e-01 -2.48817667e-01 1.89059153e-01
-6.09171748e-01 -1.27261549e-01 1.71120256e-01 -8.65871087e-02
-9.37656581e-01 -1.14167690e+00 -2.49498054e-01 -7.60542750e-01
-2.75779426e-01 -9.18872118e-01 1.25809002e+00 -8.64172995e-01
1.19063783e+00 -1.83389342e+00 -3.29012215e-01 7.89985478e-01
-5.58590889e-01 1.64041743e-01 2.32980788e-01 4.00791496e-01
-5.39068319e-02 3.63485813e-01 -3.01836818e-01 -1.14456572e-01
4.67515975e-01 5.65688550e-01 6.94988621e-03 4.73402649e-01
-1.94674224e-01 7.60127664e-01 -1.16617942e+00 1.12673149e-01
8.24886322e-01 4.92464662e-01 -1.48184255e-01 1.31309196e-01
-2.99651295e-01 -3.75245363e-02 -3.98755223e-01 5.46656132e-01
1.32331884e+00 4.61279631e-01 4.62666869e-01 -2.15795636e-01
-6.09378517e-01 -1.25781596e-01 -8.32283556e-01 1.02589917e+00
-7.23662078e-01 5.80727518e-01 2.68341124e-01 -9.79946077e-01
4.82082337e-01 1.74340427e-01 1.01742232e+00 -1.44602931e+00
1.07377462e-01 2.72471100e-01 -2.99602002e-01 1.40430275e-02
6.42675817e-01 7.21976906e-03 -8.96994472e-02 3.48329395e-01
-6.18792653e-01 -1.85634747e-01 3.40628177e-01 1.27967268e-01
5.37432194e-01 -5.21664858e-01 -4.29001510e-01 -1.34966385e+00
4.36530203e-01 -1.20142726e-02 7.18354106e-01 -2.61971653e-01
8.07659030e-02 -6.25818372e-01 -1.34740278e-01 2.56401241e-01
-1.33675742e+00 -9.92600918e-01 -5.32024920e-01 4.79485214e-01
6.21105731e-01 1.46575853e-01 -9.81282055e-01 -1.26855880e-01
3.81536901e-01 2.05503178e+00 -2.40253672e-01 -2.35573128e-01
-7.63558686e-01 -1.05206060e+00 -8.70357305e-02 5.76291621e-01
2.55936116e-01 -3.56683612e-01 -1.07544637e+00 2.00479925e-01
-4.25027817e-01 -7.48071015e-01 -5.73707402e-01 2.41554767e-01
-2.48049185e-01 -9.94008899e-01 -4.17612404e-01 -2.33369663e-01
1.05002177e+00 4.01592702e-01 1.01933837e+00 1.06347233e-01
-2.33853340e-01 2.32416108e-01 2.37743407e-01 -9.52044070e-01
-1.01917014e-01 -3.78701180e-01 1.95169762e-01 -4.99236792e-01
4.64049876e-01 -1.91098779e-01 -8.09810102e-01 2.37751290e-01
-7.66305089e-01 2.60543320e-02 -3.57895672e-01 8.82635191e-02
8.35264981e-01 1.23459113e+00 1.05853379e+00 -2.28346035e-01
8.29917073e-01 -7.23903060e-01 -1.32944405e+00 2.51459122e-01
-1.81830895e+00 -2.38708317e-01 7.53073037e-01 1.21235147e-01
-1.31541979e+00 -2.93059677e-01 -5.63122286e-03 -1.90351591e-01
2.78297514e-01 -1.61728919e-01 -8.22868049e-01 1.82986438e-01
-8.04321468e-01 1.03070512e-01 -3.51808846e-01 -5.52205145e-01
8.77582654e-03 5.03842652e-01 7.86810040e-01 -3.60437781e-01
9.86159980e-01 3.31836015e-01 2.53633022e-01 -6.80203289e-02
-5.64477919e-03 -1.99436739e-01 9.58758071e-02 -5.86693764e-01
1.01934457e+00 -8.89601409e-01 -1.33495450e+00 3.82942975e-01
-7.21671581e-01 -2.54055828e-01 -1.15638182e-01 3.09184134e-01
-3.31420273e-01 1.06258467e-01 1.32591620e-01 -1.21417272e+00
-8.11693668e-01 -9.66339529e-01 2.42079258e-01 3.35691839e-01
1.56773522e-01 -7.54170597e-01 -1.36029109e-01 1.14248972e-02
7.75372922e-01 6.66751862e-01 1.20245242e+00 6.27953336e-02
-3.49051923e-01 9.03476849e-02 -1.55798435e-01 6.01199985e-01
4.87100244e-01 2.45628189e-02 -4.84228343e-01 -9.92908061e-01
-3.03042740e-01 4.67777848e-01 2.28850543e-01 3.12370449e-01
1.71999323e+00 -4.97872829e-01 -6.60182416e-01 2.36855373e-01
2.37805009e+00 1.09965014e+00 5.59825420e-01 3.37668151e-01
1.73487186e-01 3.61593902e-01 9.70363557e-01 7.86305547e-01
7.23349571e-01 5.36687791e-01 1.27563298e+00 -8.17048401e-02
4.89647835e-01 -9.48429033e-02 -2.57565290e-01 5.67418933e-01
-1.64977629e-02 -6.09811127e-01 -7.91894257e-01 1.01591218e+00
-1.40112364e+00 -8.94676030e-01 -4.27028060e-01 2.19721699e+00
6.53488874e-01 -2.17783283e-02 -5.96501529e-02 3.66988868e-01
7.51286089e-01 -4.20926690e-01 -6.89669311e-01 -9.33035433e-01
3.71790193e-02 -8.55069160e-02 1.46014881e+00 3.98059309e-01
-1.74956560e-01 -1.69538587e-01 6.19108820e+00 6.70438647e-01
-6.73655689e-01 1.97880894e-01 4.51560408e-01 -4.18878168e-01
-1.29472303e+00 -8.71848911e-02 -6.17184103e-01 1.28444171e+00
1.57697654e+00 -1.28581250e+00 8.01842093e-01 3.47218245e-01
1.03157604e+00 -3.19171727e-01 -9.18642163e-01 4.95671421e-01
-3.33302230e-01 -1.21390629e+00 -1.78544521e-01 5.00072241e-01
9.72196043e-01 1.67213961e-01 -3.88275176e-01 -3.20941955e-02
4.52628225e-01 -7.15882063e-01 9.44024384e-01 6.96504235e-01
8.32643211e-01 -2.02191806e+00 6.86339736e-01 4.40800309e-01
-1.40502763e+00 -3.94056261e-01 -2.17026621e-02 3.75662327e-01
7.28996336e-01 8.15100074e-01 -1.98280513e-01 7.87129939e-01
1.12980616e+00 -7.19396695e-02 9.43001285e-02 7.01260448e-01
8.11546817e-02 4.74804938e-01 -5.49692869e-01 -9.17466171e-03
2.09512010e-01 -6.84830785e-01 1.07432783e-01 1.14522111e+00
4.14029270e-01 5.76611817e-01 -1.72319636e-01 9.35925901e-01
-1.98880926e-01 -3.80213320e-01 -2.08832711e-01 9.65584964e-02
1.13330185e+00 1.34587061e+00 -3.38214695e-01 -4.72195357e-01
-2.18537867e-01 3.16597968e-01 -7.09762692e-01 2.79104471e-01
-1.21412218e+00 -7.26455986e-01 8.98463011e-01 6.59993961e-02
-1.98882446e-02 5.14211655e-02 -5.19417703e-01 -3.72150898e-01
-1.18287541e-01 2.96571374e-01 5.50161079e-02 -8.63493443e-01
-9.34095025e-01 -2.02947572e-01 5.02550066e-01 -6.68522358e-01
-2.11727768e-01 -1.42488778e-01 -1.10305870e+00 1.32739615e+00
-2.39652777e+00 -5.20702899e-01 -3.06347340e-01 5.84061742e-01
5.06078362e-01 2.46751443e-01 6.86832488e-01 8.69485319e-01
-6.30880952e-01 2.75467366e-01 8.13777566e-01 -5.39044976e-01
-1.71285361e-01 -1.33482230e+00 8.58749151e-02 7.76536644e-01
-1.16961503e+00 4.92506362e-02 1.06344438e+00 -5.33929646e-01
-1.89078355e+00 -1.43952584e+00 1.04764521e+00 1.42026693e-01
5.18961489e-01 -1.55165866e-01 -5.62402546e-01 2.00095922e-01
1.07086778e+00 -5.82542658e-01 4.56686288e-01 -8.38950813e-01
4.04790699e-01 -3.40068430e-01 -2.02753687e+00 3.40552419e-01
6.98718011e-01 -1.49032891e-01 -2.05592930e-01 3.49573433e-01
4.27376509e-01 1.64935082e-01 -1.22423899e+00 6.74412072e-01
3.47036362e-01 8.89246762e-02 5.28694212e-01 -1.42339006e-01
-4.60938722e-01 -3.75683784e-01 -2.04594597e-01 -1.79472113e+00
-4.37564254e-01 -4.51115668e-01 8.14905763e-02 1.56742787e+00
3.43682021e-01 -7.45591819e-01 1.59112617e-01 1.19646788e+00
-5.08331835e-01 -5.68213999e-01 -1.32184482e+00 -9.12995398e-01
4.39742327e-01 -3.67925614e-01 1.39357603e+00 1.00912178e+00
-1.22339688e-02 -4.34000015e-01 -1.04129955e-01 4.01376724e-01
1.44378006e+00 5.16655445e-02 -2.81000137e-02 -1.04304886e+00
7.25853622e-01 -4.67671752e-01 4.32423949e-01 2.02037707e-01
5.80765307e-01 -8.96824598e-01 6.43147388e-03 -2.14751625e+00
5.24897099e-01 -4.20402586e-01 -8.96373391e-01 7.47311175e-01
3.69212836e-01 1.54566690e-01 4.24746454e-01 -2.68357605e-01
4.99676652e-02 7.57748485e-01 6.29954159e-01 -5.18049598e-01
3.70428354e-01 -2.03090683e-01 -6.48526549e-01 5.14020979e-01
1.22402346e+00 -4.14439112e-01 -7.37457573e-01 -5.08708239e-01
2.50635803e-01 -3.63674402e-01 7.00102672e-02 -4.15783703e-01
-7.46025741e-02 -9.30089116e-01 -2.15033647e-02 -1.34058750e+00
-1.46386683e-01 -1.81593657e+00 1.09458995e+00 9.37816262e-01
2.36535504e-01 3.93510252e-01 3.17249328e-01 4.69241261e-01
3.56381983e-01 -8.38256702e-02 9.18836117e-01 2.73863137e-01
-5.87166667e-01 5.40522970e-02 -6.30214095e-01 -4.38760072e-01
1.77786231e+00 -6.49202615e-02 -4.69797015e-01 2.34292388e-01
-1.43568441e-01 1.40656400e+00 5.27717948e-01 7.65972853e-01
2.40090024e-02 -1.62056935e+00 -4.33591664e-01 -2.09904224e-01
-1.70428127e-01 -3.53271425e-01 1.85779616e-01 2.02873871e-01
1.20264426e-01 7.67218769e-01 -3.02941322e-01 -3.02900672e-01
-1.11962712e+00 5.24249554e-01 4.82483327e-01 1.97972953e-01
-4.52695847e-01 -4.36004661e-02 -3.25517178e-01 5.67455292e-02
-9.46649089e-02 -1.85393184e-01 7.71167874e-02 1.45228311e-01
4.36316788e-01 1.05974400e+00 4.65896904e-01 -3.51685494e-01
-7.22404659e-01 3.89236659e-01 6.51217222e-01 3.29900146e-01
1.31892955e+00 -4.41994011e-01 -2.17958763e-01 -1.35636404e-01
1.24882925e+00 -5.21780849e-02 -9.66581821e-01 5.56288660e-01
-7.53974393e-02 -5.32320142e-01 4.17237699e-01 -1.29885828e+00
-1.45338333e+00 1.78213701e-01 7.80509472e-01 5.62811196e-01
1.64401388e+00 -2.34918639e-01 1.04053259e+00 -1.57466114e-01
7.25024462e-01 -1.94374979e+00 -9.49869990e-01 -6.41084090e-02
5.19358873e-01 -6.92355454e-01 -1.81320339e-01 8.14960375e-02
-3.81239578e-02 4.59100187e-01 5.14514208e-01 9.45699215e-02
7.51438856e-01 6.60787106e-01 -6.91622436e-01 2.07738474e-01
-7.01294780e-01 3.72283220e-01 -4.30437446e-01 2.98214048e-01
-3.03454906e-01 8.80975306e-01 -1.02763987e+00 4.50296998e-01
2.14465350e-01 -1.13600023e-01 5.46836734e-01 8.99879813e-01
-6.14150107e-01 -7.90994883e-01 -1.05778106e-01 7.11786389e-01
-6.03899658e-01 1.12819672e-01 3.13494205e-01 3.65886062e-01
5.33543408e-01 1.19234121e+00 5.84510863e-01 1.56480789e-01
8.80263150e-01 -2.08265543e-01 3.94304916e-02 2.25861236e-01
-1.37372464e-01 -1.30729347e-01 2.55060107e-01 -2.85633177e-01
-3.77796561e-01 -8.02163482e-01 -2.11388040e+00 -8.52280915e-01
-5.70519030e-01 1.03699839e+00 1.50602806e+00 8.77949953e-01
2.23163933e-01 7.29667485e-01 1.52712893e+00 -5.74014246e-01
-5.41110992e-01 -5.19120872e-01 -9.55126047e-01 7.16952756e-02
1.36182770e-01 -5.79818010e-01 -7.69338429e-01 -3.68176222e-01] | [5.6335673332214355, 2.359947443008423] |
1a97aa92-d067-4e52-8bdc-04ca3409d322 | structuring-representation-geometry-with | 2306.13924 | null | https://arxiv.org/abs/2306.13924v1 | https://arxiv.org/pdf/2306.13924v1.pdf | Structuring Representation Geometry with Rotationally Equivariant Contrastive Learning | Self-supervised learning converts raw perceptual data such as images to a compact space where simple Euclidean distances measure meaningful variations in data. In this paper, we extend this formulation by adding additional geometric structure to the embedding space by enforcing transformations of input space to correspond to simple (i.e., linear) transformations of embedding space. Specifically, in the contrastive learning setting, we introduce an equivariance objective and theoretically prove that its minima forces augmentations on input space to correspond to rotations on the spherical embedding space. We show that merely combining our equivariant loss with a non-collapse term results in non-trivial representations, without requiring invariance to data augmentations. Optimal performance is achieved by also encouraging approximate invariance, where input augmentations correspond to small rotations. Our method, CARE: Contrastive Augmentation-induced Rotational Equivariance, leads to improved performance on downstream tasks, and ensures sensitivity in embedding space to important variations in data (e.g., color) that standard contrastive methods do not achieve. Code is available at https://github.com/Sharut/CARE. | ['Stefanie Jegelka', 'Soledad Villar', 'Derek Lim', 'Joshua Robinson', 'Sharut Gupta'] | 2023-06-24 | null | null | null | null | ['contrastive-learning', 'self-supervised-learning', 'contrastive-learning'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 2.08129361e-01 2.18410626e-01 -4.38799337e-02 -4.30962026e-01
-5.89599609e-01 -1.01533473e+00 8.28652143e-01 8.00887868e-02
-5.13577938e-01 2.60058939e-01 4.97605324e-01 4.45033535e-02
9.14630014e-03 -6.21057153e-01 -8.83159220e-01 -5.45875847e-01
8.79541859e-02 -6.13734052e-02 -9.99359414e-02 -1.66093260e-01
1.25322476e-01 7.18742251e-01 -1.22350776e+00 3.24276052e-02
5.97035646e-01 4.89305943e-01 -1.92942873e-01 6.87488735e-01
3.57012272e-01 3.57530802e-01 -1.86194673e-01 -2.94244677e-01
7.32392192e-01 -4.78616595e-01 -8.24276805e-01 1.24666005e-01
9.12026763e-01 -2.44140416e-01 -6.48977280e-01 1.19508696e+00
2.56468594e-01 3.79823774e-01 7.99393237e-01 -1.24472070e+00
-1.23586130e+00 2.38524243e-01 -5.90025365e-01 1.84118122e-01
-2.60053799e-02 7.88163096e-02 1.37864149e+00 -1.16706073e+00
6.38872206e-01 1.38061798e+00 4.48944509e-01 5.14133155e-01
-1.84290648e+00 -1.94058239e-01 5.10432981e-02 -7.97288772e-03
-1.27674901e+00 -5.96688688e-01 1.08755207e+00 -4.29397762e-01
6.31854773e-01 5.24604976e-01 3.96427065e-01 7.38461792e-01
1.09441780e-01 5.22392631e-01 1.04457057e+00 -5.69243491e-01
2.16724351e-01 1.17436625e-01 -7.45708197e-02 7.47327209e-01
2.47239485e-01 1.50267007e-02 -1.84943244e-01 2.78441429e-01
9.77685392e-01 2.90454507e-01 -2.38345399e-01 -9.29304540e-01
-1.32906795e+00 9.74651694e-01 1.00019670e+00 1.21722557e-01
-3.05227581e-02 7.03677014e-02 1.76854879e-01 3.38955581e-01
5.34263670e-01 7.85343885e-01 -2.42351040e-01 2.69330531e-01
-4.85330254e-01 -2.08039265e-02 3.60515535e-01 8.02764356e-01
9.14280593e-01 7.53976777e-02 -1.20864427e-02 7.60925293e-01
4.38056365e-02 4.99959797e-01 4.91582721e-01 -1.37620497e+00
1.99550137e-01 5.04483044e-01 -1.50624320e-01 -1.10023963e+00
-5.48504949e-01 -4.72278953e-01 -7.47038007e-01 4.77962822e-01
5.59554935e-01 9.44214463e-02 -8.39540064e-01 2.28651786e+00
6.05698489e-03 3.13792974e-02 3.72383110e-02 9.31065381e-01
4.58124995e-01 5.18054426e-01 -1.27992913e-01 1.08443893e-01
1.25192142e+00 -8.61813247e-01 -4.00335163e-01 -3.19556892e-01
6.24605834e-01 -8.36760283e-01 1.77793229e+00 -1.99770811e-03
-1.18333459e+00 -3.39654416e-01 -1.32440615e+00 -5.38990378e-01
-3.98962170e-01 -1.45743415e-02 4.09280300e-01 3.26562196e-01
-1.12237191e+00 6.71880722e-01 -9.77154255e-01 -4.21071440e-01
2.77823985e-01 2.15441599e-01 -6.55612230e-01 2.76795011e-02
-8.30060124e-01 7.88113952e-01 5.76546378e-02 -1.92768857e-01
-3.94783467e-01 -8.72439086e-01 -1.15168333e+00 7.35937431e-03
1.52867600e-01 -6.65377140e-01 9.62735116e-01 -1.15273106e+00
-1.26879418e+00 9.61325765e-01 -1.07779212e-01 -2.89681792e-01
5.75725436e-01 -2.93711156e-01 5.81556931e-04 3.17547709e-01
-1.56417340e-02 9.74985242e-01 9.72482800e-01 -1.22597873e+00
-4.66670729e-02 -3.37067544e-01 3.92022491e-01 4.31858718e-01
-7.19485283e-01 -1.66945547e-01 -3.64337653e-01 -9.31861281e-01
3.46717924e-01 -1.36486268e+00 -1.08406842e-01 6.13820910e-01
-3.41969073e-01 2.85002291e-01 7.70316720e-01 -6.24718130e-01
6.98031366e-01 -2.44329381e+00 3.53876472e-01 2.54928589e-01
5.62970400e-01 -1.45206675e-01 -4.78459477e-01 2.75701374e-01
-4.02785510e-01 1.29261613e-01 -5.19261360e-01 -2.92643309e-01
8.42731744e-02 2.99525380e-01 -3.28240007e-01 8.62517893e-01
5.57194769e-01 9.99335825e-01 -9.37248647e-01 -3.03307563e-01
4.12707895e-01 7.03361690e-01 -9.75963414e-01 -1.09241270e-02
9.70109329e-02 3.76097858e-01 6.40314945e-04 1.93649784e-01
6.11695349e-01 -1.70113862e-01 -1.71356708e-01 -5.16462862e-01
-3.23826335e-02 2.41716698e-01 -9.98111844e-01 1.77589929e+00
-6.05909586e-01 9.13933456e-01 -4.88671549e-02 -1.04185581e+00
6.31189942e-01 -2.88059056e-01 3.20156991e-01 -4.72583205e-01
-4.00171094e-02 -1.03357427e-01 1.71293586e-01 6.39240742e-02
3.31838459e-01 -9.58415121e-02 5.61909266e-02 4.10510749e-01
7.02105686e-02 -2.46036932e-01 -1.55673902e-02 2.98269719e-01
8.92763555e-01 8.51313844e-02 2.21654728e-01 -4.09160167e-01
2.83571720e-01 -3.55972797e-01 4.89438027e-01 3.77093166e-01
-2.68742859e-01 8.51403892e-01 5.32022774e-01 -1.43967792e-01
-1.45505440e+00 -1.55344212e+00 -3.17798406e-01 1.03946877e+00
2.44974956e-01 -4.07163560e-01 -8.13644409e-01 -5.84980965e-01
-9.09163675e-04 4.38391656e-01 -7.68273115e-01 -6.97589815e-01
-5.50567508e-01 -5.49301088e-01 3.66217643e-01 6.51690125e-01
4.61607337e-01 -6.45341456e-01 -3.51444095e-01 -3.01865458e-01
1.02635041e-01 -9.82406914e-01 -1.00826418e+00 2.07640216e-01
-9.72960293e-01 -8.27503204e-01 -7.53035367e-01 -8.20910931e-01
1.11906254e+00 4.10314530e-01 7.66941071e-01 -2.26705864e-01
-4.95690584e-01 5.29836595e-01 -1.72255442e-01 -6.54560402e-02
-1.57561719e-01 -1.97559312e-01 2.15926304e-01 8.41630250e-02
1.04117990e-01 -5.96404731e-01 -8.10534537e-01 2.49148622e-01
-1.08026481e+00 6.73902482e-02 4.93651599e-01 8.77201557e-01
7.49210060e-01 -3.04400563e-01 1.50118411e-01 -7.19208121e-01
4.84124213e-01 -6.75185546e-02 -4.53592837e-01 -8.15781355e-02
-6.41507089e-01 6.60198748e-01 7.92752147e-01 -4.20647830e-01
-6.93727732e-01 6.83012791e-03 2.65183210e-01 -6.67160988e-01
2.50305887e-02 1.81305885e-01 -7.95745775e-02 -2.50673145e-01
1.02170169e+00 -1.67771690e-02 2.30265170e-01 -4.09299046e-01
8.70398521e-01 1.62800029e-01 7.14969099e-01 -5.38547277e-01
1.33653522e+00 7.38380611e-01 2.36814484e-01 -7.18575418e-01
-7.43828118e-01 -2.39097953e-01 -1.02506638e+00 1.47665814e-01
7.51357019e-01 -8.30932438e-01 -2.62453079e-01 1.67897940e-01
-7.23300755e-01 -3.77296388e-01 -7.20395923e-01 6.20413661e-01
-7.10028052e-01 4.13035631e-01 -6.45049095e-01 -2.15066597e-01
-5.12660332e-02 -9.19783950e-01 8.23267281e-01 -2.84657832e-02
-2.81642377e-01 -1.21132791e+00 2.02171549e-01 4.75217588e-03
3.16408485e-01 3.01353246e-01 8.58192801e-01 -4.22361463e-01
-3.60517740e-01 -1.49574161e-01 -3.44292521e-01 6.61931694e-01
3.81209254e-01 1.12187698e-01 -1.03258574e+00 -4.10629064e-01
-2.59205401e-01 -2.80761778e-01 1.08067989e+00 3.21950406e-01
1.18703008e+00 -5.35779834e-01 2.51769096e-01 1.07256699e+00
1.25256753e+00 -2.49632210e-01 3.96739990e-01 2.97248691e-01
9.05579925e-01 6.52462304e-01 2.03698352e-01 1.86445862e-01
1.56700656e-01 7.03366816e-01 3.46397877e-01 -5.86560249e-01
-3.68683785e-01 -4.25171643e-01 4.04985160e-01 7.31919348e-01
1.12810619e-02 2.68566728e-01 -6.51880383e-01 3.74147892e-01
-1.56838226e+00 -9.16152835e-01 2.87703156e-01 2.34838128e+00
9.78278995e-01 2.14090258e-01 1.31291494e-01 4.13021855e-02
6.30380094e-01 3.52321833e-01 -8.14653039e-01 -4.90315020e-01
-2.73180157e-01 2.45573059e-01 5.85892618e-01 9.80059564e-01
-1.30697048e+00 1.02152264e+00 5.70947027e+00 3.74544203e-01
-1.35181260e+00 5.55858351e-02 5.55579245e-01 -3.34091663e-01
-6.06999934e-01 -4.87067327e-02 -7.90171251e-02 2.64361143e-01
5.77457011e-01 -4.33592826e-01 7.09656537e-01 7.15801001e-01
2.21028030e-01 5.13559103e-01 -1.36481607e+00 9.30068016e-01
1.99726388e-01 -1.24687111e+00 2.16121450e-01 1.03210814e-01
7.64622033e-01 4.57552150e-02 6.09138548e-01 1.69277005e-02
3.82948607e-01 -8.92807901e-01 7.12200820e-01 2.70014763e-01
9.78423953e-01 -8.03299308e-01 2.97998518e-01 -1.39475226e-01
-9.81714368e-01 3.54092345e-02 -5.17756641e-01 1.63471047e-02
-2.49223977e-01 3.40055645e-01 -5.83839536e-01 1.05537148e-02
3.80746514e-01 8.62402737e-01 -7.38796175e-01 7.96148360e-01
-2.31410488e-01 3.43917847e-01 -2.50440687e-01 3.21540117e-01
1.85010254e-01 -5.13948023e-01 6.04927182e-01 1.23944747e+00
1.56765375e-02 2.38865297e-02 7.65580386e-02 9.79703963e-01
-5.00013769e-01 2.17692062e-01 -8.99871409e-01 1.34193227e-01
4.11410481e-01 1.27853954e+00 -6.01898611e-01 -1.39959499e-01
-5.34436405e-01 1.33900630e+00 4.93974328e-01 6.36306405e-01
-5.89447439e-01 -6.65328085e-01 1.07852566e+00 1.41917467e-01
1.92893855e-02 -4.05215085e-01 -4.36493456e-01 -1.39761806e+00
1.85535401e-01 -6.30759954e-01 3.16068530e-01 -8.41227531e-01
-1.10033810e+00 2.81982243e-01 -2.22410727e-02 -1.30250823e+00
-1.00043528e-01 -7.88215339e-01 -7.40710199e-01 7.88757443e-01
-1.36128938e+00 -1.11736369e+00 -2.63334304e-01 7.22267449e-01
3.50887567e-01 5.49085662e-02 6.40584111e-01 5.31978123e-02
-4.00848567e-01 9.96756315e-01 2.96734720e-01 2.94417173e-01
1.12079382e+00 -1.64738476e+00 3.89976948e-01 1.06028283e+00
4.95045304e-01 8.80637944e-01 7.97634125e-01 -2.02438012e-02
-1.30812418e+00 -1.19809568e+00 4.82761770e-01 -6.22697949e-01
9.22233045e-01 -5.88579059e-01 -8.90408754e-01 9.74067807e-01
3.96098085e-02 5.22576451e-01 7.19509065e-01 5.15709780e-02
-1.02081418e+00 -2.32027620e-01 -1.19039881e+00 1.05995107e+00
1.08387601e+00 -9.60071504e-01 -4.59520549e-01 2.64725059e-01
8.77054513e-01 -1.42441779e-01 -8.51107895e-01 2.49723136e-01
4.45294112e-01 -4.62343216e-01 1.26345146e+00 -1.09568608e+00
5.99333107e-01 -2.86808461e-01 -3.41204911e-01 -1.59699428e+00
-7.12940991e-01 -6.70407057e-01 1.16251968e-02 8.54543030e-01
4.28268433e-01 -8.57834458e-01 5.47759235e-01 6.34214401e-01
-1.77003846e-01 -5.63864708e-01 -7.63472736e-01 -9.13250029e-01
5.80303907e-01 -1.98546667e-02 1.58063918e-01 1.24446583e+00
2.23228112e-01 3.32765847e-01 -1.06227860e-01 2.32937753e-01
8.32925081e-01 -1.66055892e-04 5.59719801e-01 -1.00351238e+00
-2.98167586e-01 -6.93994761e-01 -7.81267524e-01 -9.64895070e-01
2.63923138e-01 -1.26850450e+00 -2.70490527e-01 -1.17485678e+00
3.14610422e-01 -2.65959024e-01 -6.23725295e-01 6.79081917e-01
-2.01863199e-01 7.63015389e-01 4.22606707e-01 3.51195395e-01
-3.32884252e-01 6.63074434e-01 1.27016413e+00 -2.11096957e-01
-3.33859086e-01 -3.86651665e-01 -1.05242729e+00 8.16633582e-01
9.52632308e-01 -1.41519308e-01 -6.06360674e-01 -6.12661541e-01
5.90204932e-02 -5.54913878e-01 5.22389412e-01 -6.96975529e-01
-7.80584142e-02 -1.12490967e-01 5.68733156e-01 2.50436336e-01
3.64268810e-01 -5.06504714e-01 -3.69792551e-01 4.14139867e-01
-8.80495012e-01 1.32355139e-01 2.05126241e-01 3.53569955e-01
-7.09545910e-02 -1.06131189e-01 1.21726322e+00 2.61779577e-01
-4.22077805e-01 3.73002052e-01 1.40110165e-01 4.33668345e-01
8.45650136e-01 -1.37658417e-01 -4.63446617e-01 -4.61963028e-01
-9.28764224e-01 -1.34520367e-01 8.79795790e-01 4.25266981e-01
5.14329135e-01 -1.63724875e+00 -7.06640363e-01 2.84091949e-01
2.17152476e-01 -1.94721669e-01 -1.32281572e-01 7.18428373e-01
-5.97829163e-01 2.76967704e-01 -4.30687040e-01 -4.50748712e-01
-1.31508327e+00 6.37927473e-01 3.90328169e-01 1.45199761e-01
-8.62210572e-01 7.97909319e-01 7.81963170e-01 -5.72346330e-01
7.94016123e-02 -4.52090323e-01 5.83780967e-02 -1.31129533e-01
3.85726660e-01 2.36072332e-01 -3.84792499e-02 -6.91818416e-01
-4.04564112e-01 7.88956463e-01 -1.50675490e-01 -3.80071133e-01
1.27175915e+00 -7.36128092e-02 7.97525197e-02 3.89461994e-01
1.80400360e+00 3.52725297e-01 -1.62487531e+00 -5.32092035e-01
-2.76921749e-01 -5.92909873e-01 2.38799974e-01 -4.66183037e-01
-1.08419371e+00 9.67722952e-01 7.08833933e-01 1.13064229e-01
9.95989740e-01 2.40719318e-01 3.40207577e-01 6.55805111e-01
-1.38116494e-01 -8.38712275e-01 3.69126171e-01 3.92177165e-01
1.17715049e+00 -1.29085362e+00 4.07859311e-02 -2.03391731e-01
-4.71810430e-01 9.97130513e-01 5.71613073e-01 -5.10197937e-01
6.17918074e-01 6.52858242e-02 9.24322158e-02 1.00176930e-01
-4.73703593e-01 -1.13859750e-01 4.82961893e-01 5.64074218e-01
4.38837886e-01 1.57248691e-01 -8.78488421e-02 8.26542825e-02
-5.58395863e-01 -7.09095001e-01 6.25982285e-01 6.56456947e-01
-4.50402141e-01 -8.57770801e-01 -3.37698489e-01 2.84902960e-01
-1.78205013e-01 -2.17317149e-01 -5.46504736e-01 7.68184125e-01
-2.38497287e-01 5.33105969e-01 5.42538404e-01 -2.20661357e-01
4.07903433e-01 -1.72816768e-01 6.93836153e-01 -6.01655543e-01
1.69647351e-01 5.66231832e-02 -3.45041335e-01 -5.98443210e-01
-1.01669967e-01 -7.46203721e-01 -1.42439127e+00 -3.88278127e-01
1.57670960e-01 -1.96380824e-01 4.56916362e-01 5.01898348e-01
3.96311104e-01 3.77507836e-01 9.03550088e-01 -8.57043386e-01
-6.39608502e-01 -6.33967817e-01 -4.07708973e-01 8.54069650e-01
7.30303645e-01 -6.02272987e-01 -7.29316652e-01 3.57549340e-01] | [9.043534278869629, 2.8060855865478516] |
a481a658-92fa-471c-b06b-5b3456cb6ab2 | task-driven-and-experience-based-question | null | null | https://aclanthology.org/2022.lrec-1.670 | https://aclanthology.org/2022.lrec-1.670.pdf | Task-Driven and Experience-Based Question Answering Corpus for In-Home Robot Application in the House3D Virtual Environment | At present, more and more work has begun to pay attention to the long-term housekeeping robot scene. Naturally, we wonder whether the robot can answer the questions raised by the owner according to the actual situation at home. These questions usually do not have a clear text context, are directly related to the actual scene, and it is difficult to find the answer from the general knowledge base (such as Wikipedia). Therefore, the experience accumulated from the task seems to be a more natural choice. We present a corpus called TEQA (task-driven and experience-based question answering) in the long-term household task. Based on a popular in-house virtual environment (AI2-THOR) and agent task experiences of ALFRED, we design six types of questions along with answering including 24 question templates, 37 answer templates, and nearly 10k different question answering pairs. Our corpus aims at investigating the ability of task experience understanding of agents for the daily question answering scenario on the ALFRED dataset. | ['Yang Liu', 'Liubo Ouyang', 'Zhuoqun Xu'] | null | null | null | null | lrec-2022-6 | ['general-knowledge'] | ['miscellaneous'] | [-4.09983128e-01 3.59068990e-01 5.01886010e-01 -4.89959598e-01
-6.39026225e-01 -6.23814285e-01 5.30229867e-01 2.66463578e-01
-5.90247631e-01 7.93842912e-01 4.73278224e-01 -3.22238058e-01
-1.36731520e-01 -7.20594406e-01 -5.06215215e-01 -4.53026593e-02
1.14097670e-01 9.29904878e-01 3.32160354e-01 -9.01382089e-01
-2.85068788e-02 -1.33161485e-01 -1.50982058e+00 2.68496782e-01
9.64390516e-01 7.45127141e-01 1.16366422e+00 6.33617997e-01
-2.67099291e-01 1.65175128e+00 -9.28282559e-01 -6.68987989e-01
-5.33731915e-02 -4.20762807e-01 -1.68799829e+00 1.65088505e-01
1.13524370e-01 -7.94022679e-01 -4.49142158e-01 7.87866354e-01
4.42276388e-01 4.91200060e-01 4.65067983e-01 -1.46985197e+00
-9.42293286e-01 3.35874766e-01 3.28856736e-01 1.58176050e-01
1.20236218e+00 3.76915723e-01 1.07387602e+00 -7.00268686e-01
1.07651925e+00 1.16162574e+00 4.89182055e-01 3.30562741e-01
-5.48362076e-01 -2.69283094e-02 5.43984994e-02 9.31430161e-01
-1.25714922e+00 -1.44422978e-01 4.93921638e-01 -3.98287684e-01
1.30203426e+00 1.08586885e-01 5.32608986e-01 1.13299441e+00
1.25576034e-01 9.50782120e-01 6.83782816e-01 -3.47249955e-01
3.43355864e-01 2.95534223e-01 2.96156257e-01 6.55513287e-01
-1.14564233e-01 -2.33005032e-01 -3.75019103e-01 -2.33446360e-01
6.61886096e-01 -2.99405605e-01 -5.67795336e-01 -6.79337859e-01
-1.33804870e+00 9.25402403e-01 5.09506404e-01 4.80040461e-01
-7.00529695e-01 -7.64666423e-02 6.74800813e-01 8.36935461e-01
-7.99349993e-02 1.00609112e+00 -7.39304066e-01 -3.71037662e-01
2.98700780e-01 8.32227647e-01 1.39137173e+00 1.35809159e+00
1.03782940e+00 -5.45730591e-01 -1.88379794e-01 6.32774591e-01
3.01812202e-01 4.71884221e-01 3.89034539e-01 -1.47837305e+00
6.38541520e-01 7.19130158e-01 7.96446860e-01 -1.08717895e+00
-4.94662583e-01 2.40577698e-01 -2.15297624e-01 -7.31662959e-02
7.06181586e-01 -4.02242601e-01 -4.35218215e-01 1.52524185e+00
5.09904087e-01 -5.53415596e-01 5.49554348e-01 9.78247643e-01
1.46048605e+00 6.64749622e-01 2.88016826e-01 1.38236955e-02
1.82088685e+00 -1.28106177e+00 -1.22952950e+00 -5.49143851e-01
1.09857821e+00 -4.31591898e-01 1.36217415e+00 -4.71042134e-02
-7.28735745e-01 -6.05424345e-01 -8.96618962e-01 -5.49425244e-01
-6.26250923e-01 -3.53924893e-02 6.47258937e-01 8.53965282e-02
-1.09619701e+00 -8.38455707e-02 -2.53534913e-01 -1.17414272e+00
-4.96294916e-01 3.78633998e-02 -4.47445244e-01 -4.34871554e-01
-1.43164515e+00 1.55946577e+00 3.22772652e-01 1.41069740e-01
-9.16089356e-01 -1.17033504e-01 -1.29694283e+00 -2.03841448e-01
7.35687256e-01 -1.02571929e+00 1.93121898e+00 -7.33167112e-01
-1.45597041e+00 9.27478135e-01 -1.69713497e-01 -3.64466339e-01
2.06735894e-01 -3.83807421e-01 -2.75641829e-01 2.12125421e-01
5.03781676e-01 7.13335574e-01 4.70574647e-01 -1.16949308e+00
-6.64645195e-01 -4.75416213e-01 1.05611980e+00 7.21904695e-01
2.73819625e-01 7.03935921e-02 -1.76019758e-01 -1.59637943e-01
-2.82828867e-01 -7.43547797e-01 -3.10402006e-01 -2.26371139e-01
1.29698709e-01 -7.31245279e-01 6.03774071e-01 -9.85847235e-01
6.56727314e-01 -2.09427142e+00 1.54317945e-01 -5.21535635e-01
1.72387317e-01 -3.57603021e-02 -3.47117901e-01 8.33503067e-01
3.05846959e-01 -3.05973530e-01 1.43690437e-01 5.98274097e-02
3.39366496e-01 4.45335478e-01 -1.92158267e-01 1.50422044e-02
-1.03784747e-01 1.12544131e+00 -1.37534380e+00 -5.73739767e-01
2.03504771e-01 -9.94549692e-03 -4.24134910e-01 7.37989962e-01
-6.26628101e-01 2.60184139e-01 -9.19874609e-01 3.69978875e-01
1.70631975e-01 -3.66515726e-01 1.50031522e-01 5.99080175e-02
2.41156325e-01 2.51937032e-01 -7.68431842e-01 2.18665957e+00
-6.75295055e-01 7.23729372e-01 2.82516360e-01 -3.30655158e-01
9.21862483e-01 6.22004986e-01 -1.52989196e-05 -9.35921073e-01
1.64583415e-01 1.05073154e-01 -6.07344732e-02 -1.36018109e+00
9.83950496e-01 8.35051984e-02 -3.51424605e-01 2.35530630e-01
1.84637696e-01 -7.97670245e-01 2.87138522e-01 1.76846519e-01
1.46528435e+00 1.58659682e-01 6.57359660e-01 -2.03679115e-01
3.46657038e-01 6.40182316e-01 9.15851370e-02 7.75844932e-01
-6.66439772e-01 3.51815552e-01 2.04263851e-01 -7.37881362e-01
-7.48096526e-01 -8.36034238e-01 2.65986204e-01 1.25455630e+00
4.66595501e-01 -6.37468696e-01 -8.54868948e-01 -6.66541517e-01
-4.23108101e-01 1.03038836e+00 -5.32048106e-01 3.55129898e-01
-3.96540016e-01 9.98435840e-02 1.55203447e-01 3.02196175e-01
9.37516630e-01 -1.76166832e+00 -1.33978689e+00 4.01100069e-01
-8.07634473e-01 -1.26367593e+00 -2.36078396e-01 1.54401913e-01
-3.51858914e-01 -1.17732906e+00 -7.12274969e-01 -1.18742192e+00
3.89947653e-01 2.96856403e-01 1.68890595e+00 3.55832875e-02
2.07437173e-01 1.13269162e+00 -1.00574565e+00 -4.42783117e-01
-3.27598035e-01 1.52425617e-01 -3.20987225e-01 -5.96183956e-01
7.22670317e-01 -2.15048611e-01 -4.15987968e-01 6.34382367e-01
-5.29909194e-01 -3.20296198e-01 3.20514292e-01 5.11295259e-01
-1.44733936e-01 -2.29943339e-02 5.82929432e-01 -5.15376568e-01
1.17010236e+00 -6.23185515e-01 -1.52983572e-02 4.58851337e-01
2.17089057e-01 -7.96185210e-02 3.27372551e-01 -1.69222653e-01
-1.39059794e+00 -7.36560598e-02 -1.56388611e-01 6.60342053e-02
-5.86086154e-01 6.03936672e-01 -3.44558984e-01 5.24752259e-01
9.46358681e-01 -7.75361806e-02 -2.10606486e-01 -1.14073612e-01
4.51597989e-01 5.67359567e-01 6.88662708e-01 -7.49095976e-01
4.46846008e-01 2.67332681e-02 -8.56618464e-01 -1.01482737e+00
-5.64922690e-01 -7.77671874e-01 -4.51161176e-01 -3.94036919e-01
1.36621368e+00 -8.50951254e-01 -8.62558722e-01 4.39058930e-01
-1.84925616e+00 -9.39202189e-01 -3.29525411e-01 2.11596340e-01
-1.01690018e+00 1.95979238e-01 -7.15220690e-01 -7.87784159e-01
-4.86674644e-02 -1.30081534e+00 9.74624157e-01 1.00517914e-01
-8.23179364e-01 -8.33964825e-01 1.93112567e-01 5.31152129e-01
3.23974490e-01 -1.82386197e-03 1.11987770e+00 -6.39606297e-01
-5.28920293e-01 1.65736362e-01 -1.57813117e-01 -2.33838901e-01
2.28690282e-02 -7.77185738e-01 -7.03620017e-01 6.09998517e-02
4.24307257e-01 -8.26567411e-01 5.92486747e-02 -3.94179448e-02
3.57980341e-01 -2.24604011e-01 -1.13323413e-01 -3.30581009e-01
1.11193287e+00 3.75192702e-01 1.00088000e+00 6.30918503e-01
3.71036351e-01 1.17642200e+00 1.00337148e+00 3.82667691e-01
1.17320359e+00 7.14918435e-01 4.29045737e-01 3.39443237e-01
2.54053473e-01 -3.72890145e-01 3.12561303e-01 5.78514338e-01
3.63585465e-02 -4.25934881e-01 -8.80411744e-01 8.82020473e-01
-2.11345506e+00 -8.11825275e-01 -4.37674165e-01 1.55212593e+00
4.49615330e-01 -4.33328241e-01 -7.15430528e-02 -2.73709595e-01
4.87669587e-01 2.38174632e-01 -6.34156227e-01 -2.86091775e-01
1.90175638e-01 -4.57722843e-01 -1.50579154e-01 3.65623236e-01
-8.89115453e-01 8.78708661e-01 6.25863743e+00 4.77938592e-01
-1.77249089e-02 2.20203191e-01 2.96090245e-01 6.14316165e-01
-2.16462780e-02 -1.14730269e-01 -1.40279055e-01 -1.17153050e-02
6.18221104e-01 -1.34850264e-01 3.34941059e-01 1.13345957e+00
-1.37892808e-03 -6.46535575e-01 -1.41280437e+00 1.19255900e+00
1.92264482e-01 -6.39016271e-01 -2.07557380e-01 -2.12356344e-01
2.57794023e-01 -6.24707825e-02 -2.68298954e-01 9.74393129e-01
7.33261287e-01 -1.00804484e+00 8.57610881e-01 3.74389410e-01
2.32059911e-01 -3.96641344e-01 1.18051517e+00 7.72245347e-01
-1.23756897e+00 -1.62967667e-01 -4.26711798e-01 -5.20520329e-01
4.92089063e-01 -2.10864797e-01 -1.12307513e+00 4.90298480e-01
1.12256503e+00 2.47627899e-01 -4.92666155e-01 6.66492224e-01
-3.57657075e-01 -1.74744293e-01 -1.18127823e-01 -4.83112305e-01
2.93935001e-01 -1.05986729e-01 4.30877954e-01 4.90463406e-01
2.04248413e-01 7.39969373e-01 1.68064326e-01 5.14049292e-01
1.22946188e-01 3.00284088e-01 -9.98519957e-01 1.66541338e-01
3.16625535e-01 9.90428686e-01 -4.64401722e-01 -2.00002730e-01
-6.02356911e-01 1.37936008e+00 5.26952744e-01 5.19978404e-01
-5.39600730e-01 -4.58017379e-01 5.60347378e-01 -1.71385910e-02
1.25983700e-01 -2.96487182e-01 3.13466907e-01 -7.62144327e-01
1.87342241e-01 -1.31011510e+00 4.44408447e-01 -1.62677670e+00
-1.30697358e+00 8.83595705e-01 2.11905362e-03 -8.20145428e-01
-6.76378667e-01 -7.29989827e-01 -4.25320327e-01 4.45206583e-01
-1.33051157e+00 -9.47289348e-01 -6.90093338e-01 5.96961379e-01
9.73100841e-01 1.33887723e-01 1.09422946e+00 9.75384265e-02
-5.34286350e-03 -8.82625505e-02 -3.45813602e-01 4.37277853e-02
7.59382367e-01 -1.17741227e+00 3.18922251e-01 3.03479377e-02
-2.07634106e-01 4.24975187e-01 1.10922384e+00 -5.96349418e-01
-1.56374967e+00 -6.38147473e-01 1.00428903e+00 -1.19658077e+00
8.11800003e-01 -1.04798853e-01 -1.01588404e+00 9.56695378e-01
7.37296820e-01 -6.61172092e-01 3.76505554e-01 3.94300148e-02
9.22524463e-03 3.60852689e-01 -1.29043472e+00 7.47701943e-01
1.26683939e+00 -8.00389647e-01 -1.37187421e+00 4.73568380e-01
1.12452602e+00 -4.17672426e-01 -5.68608582e-01 2.30658561e-01
1.31612301e-01 -9.19031560e-01 7.15158641e-01 -4.10097212e-01
5.40996075e-01 -3.60033542e-01 -4.99002278e-01 -1.48196316e+00
-1.99605152e-01 -4.71370965e-01 2.12610036e-01 1.00942373e+00
5.89634180e-01 -4.94455278e-01 4.50178921e-01 1.18080580e+00
-1.25539124e-01 -3.04153025e-01 -6.47033513e-01 -5.41500211e-01
-3.33257765e-01 -4.02987957e-01 8.08735788e-01 8.33426654e-01
4.84741509e-01 8.34160447e-01 -3.47472765e-02 2.34830722e-01
-9.41171572e-02 -1.51764154e-01 1.14044404e+00 -1.06573546e+00
-2.69898802e-01 1.26786679e-01 -2.58198202e-01 -1.61988103e+00
3.32884133e-01 -5.09411514e-01 5.86175919e-01 -2.17711353e+00
6.87223449e-02 -6.11533262e-02 5.79981267e-01 1.83690898e-02
-1.53933302e-01 -4.25146699e-01 3.72897089e-01 8.42655748e-02
-1.13694882e+00 8.78569901e-01 1.49112689e+00 -2.14748383e-01
-2.51529127e-01 -4.18620229e-01 -4.64932352e-01 9.51041698e-01
6.45018756e-01 2.23308336e-03 -6.52135134e-01 -9.04023886e-01
6.36908054e-01 4.92729336e-01 3.00103277e-01 -8.89081120e-01
5.29115736e-01 1.69310030e-02 -8.02890118e-03 -5.05637884e-01
5.92463911e-01 -1.24342144e+00 1.52299657e-01 3.47744264e-02
-1.69586316e-01 4.93360817e-01 -1.91956069e-02 6.10486269e-01
-3.50591362e-01 -7.42741227e-01 -9.83716026e-02 -6.08216941e-01
-1.55603123e+00 -5.93032837e-02 -8.04321468e-01 1.52620375e-01
1.18954968e+00 -1.98225006e-01 -4.99782324e-01 -1.08671641e+00
-6.70815587e-01 9.55160320e-01 5.03948390e-01 5.28958321e-01
5.85237503e-01 -1.13890064e+00 -6.29782438e-01 -5.69365263e-01
6.62784159e-01 2.58214682e-01 4.01393175e-01 1.97095886e-01
-5.63851714e-01 5.39378464e-01 -4.13768589e-02 -2.20098853e-01
-7.43805468e-01 7.13526785e-01 3.85367781e-01 -3.08734059e-01
-5.70132375e-01 6.12320483e-01 6.13009214e-01 -9.08960760e-01
2.11245000e-01 -3.38062704e-01 -7.35574841e-01 1.31217688e-01
6.16794705e-01 2.99016207e-01 -1.17668755e-01 -7.34075665e-01
-1.05554201e-01 3.03006262e-01 1.40494689e-01 -1.51949584e-01
1.07375693e+00 -4.54619527e-01 9.77595523e-03 5.28018296e-01
1.00719714e+00 -5.65531015e-01 -9.90073383e-01 -2.07139239e-01
3.44119728e-01 -2.34343693e-01 -5.51640332e-01 -6.73962712e-01
-4.20882434e-01 4.57491130e-01 2.10636184e-01 5.84357381e-01
6.46976948e-01 6.45721614e-01 9.02186334e-01 1.45482397e+00
1.17996633e+00 -1.20908761e+00 4.68658447e-01 9.95509207e-01
1.59149325e+00 -1.31128049e+00 -4.67544854e-01 -4.78054374e-01
-9.67953444e-01 7.02326477e-01 1.05499148e+00 2.57486433e-01
2.94952214e-01 -1.99395031e-01 4.34653997e-01 -6.92586899e-01
-6.96041882e-01 -5.14499187e-01 -4.33502734e-01 1.19825625e+00
6.60759062e-02 2.30770651e-02 7.47118741e-02 6.90742850e-01
-6.79625094e-01 -6.77804872e-02 6.55903578e-01 1.13628221e+00
-5.46585441e-01 -6.08040690e-01 -2.99254209e-01 2.54829794e-01
2.76768450e-02 2.28967860e-01 -6.43027425e-01 8.30329239e-01
-1.42497793e-01 1.68104362e+00 -2.52115298e-02 -1.44392788e-01
9.54536438e-01 1.51516274e-01 4.16612387e-01 -8.15977216e-01
-4.73943740e-01 -9.59903955e-01 7.04470813e-01 -5.46658218e-01
-5.46590209e-01 -4.64007676e-01 -1.26017094e+00 -1.52065262e-01
-1.92978531e-01 4.06797647e-01 3.40551853e-01 1.13369739e+00
2.48021722e-01 1.50880903e-01 -9.70619824e-03 -7.45532274e-01
-2.67245024e-01 -1.21142030e+00 -5.14321208e-01 5.63474894e-01
2.43287832e-01 -6.61988080e-01 -2.76181549e-01 1.69734713e-02] | [4.446498394012451, 0.6592406630516052] |
fea6a145-393d-4c1d-8c68-14be81a99b11 | logo-net-large-scale-deep-logo-detection-and | 1511.02462 | null | http://arxiv.org/abs/1511.02462v2 | http://arxiv.org/pdf/1511.02462v2.pdf | LOGO-Net: Large-scale Deep Logo Detection and Brand Recognition with Deep Region-based Convolutional Networks | Logo detection from images has many applications, particularly for brand
recognition and intellectual property protection. Most existing studies for
logo recognition and detection are based on small-scale datasets which are not
comprehensive enough when exploring emerging deep learning techniques. In this
paper, we introduce "LOGO-Net", a large-scale logo image database for logo
detection and brand recognition from real-world product images. To facilitate
research, LOGO-Net has two datasets: (i)"logos-18" consists of 18 logo classes,
10 brands, and 16,043 logo objects, and (ii) "logos-160" consists of 160 logo
classes, 100 brands, and 130,608 logo objects. We describe the ideas and
challenges for constructing such a large-scale database. Another key
contribution of this work is to apply emerging deep learning techniques for
logo detection and brand recognition tasks, and conduct extensive experiments
by exploring several state-of-the-art deep region-based convolutional networks
techniques for object detection tasks. The LOGO-net will be released at
http://logo-net.org/ | ['Hui Xue', 'Steven C. H. Hoi', 'Qiang Wu', 'Hantang Liu', 'Yue Wu', 'Xiongwei Wu', 'Huiqiong Wang'] | 2015-11-08 | null | null | null | null | ['logo-recognition'] | ['computer-vision'] | [ 1.57291852e-02 -4.52348113e-01 -6.75234914e-01 -2.60803074e-01
-4.07086492e-01 -5.48667490e-01 3.14676821e-01 2.28420809e-01
3.08065444e-01 -1.53309733e-01 -3.09478551e-01 -2.66973078e-01
2.54042417e-01 -1.15639699e+00 -8.93856049e-01 -3.97963166e-01
-2.04761103e-01 4.90782708e-01 -5.99796064e-02 -1.99006364e-01
3.02025259e-01 9.26364958e-01 -1.47752190e+00 3.43155026e-01
2.20316201e-01 2.09988379e+00 -1.77479729e-01 3.78712565e-01
-4.61648637e-03 7.44881094e-01 -6.30697548e-01 -2.24775627e-01
5.81526160e-01 -1.91254720e-01 -3.24886560e-01 5.53609014e-01
9.32274818e-01 -4.85681295e-01 -6.48048699e-01 1.24436963e+00
5.60045600e-01 -3.47387463e-01 6.92285001e-01 -1.25884569e+00
-1.33886266e+00 4.87604469e-01 -7.24259377e-01 3.88789475e-01
7.82403350e-02 3.14885974e-01 1.08332908e+00 -9.82175887e-01
5.67212582e-01 1.42934728e+00 8.50242496e-01 -8.27341229e-02
-8.51013362e-01 -1.16241014e+00 -1.59882218e-01 4.50778484e-01
-1.71465874e+00 -2.12851077e-01 8.91168773e-01 -6.69159770e-01
1.13128352e+00 -2.84051206e-02 7.56351173e-01 9.02260900e-01
6.87434196e-01 1.06953537e+00 9.12709653e-01 -3.47844034e-01
-5.17507317e-04 2.45926574e-01 2.99469471e-01 7.13052452e-01
7.70457208e-01 3.99052858e-01 -8.90082896e-01 1.55566782e-01
9.64520812e-01 2.10715175e-01 4.85976458e-01 -2.01067850e-01
-5.69815636e-01 1.19938135e+00 8.43527853e-01 2.12266341e-01
-3.42601180e-01 1.72532007e-01 5.94584465e-01 4.10606116e-01
4.92851108e-01 1.56719416e-01 -8.48631188e-02 4.26981956e-01
-4.44034845e-01 2.01699436e-01 6.06816649e-01 1.29084027e+00
7.33674526e-01 1.95799068e-01 3.93834203e-01 1.00796700e+00
5.16775429e-01 4.85399425e-01 4.96968508e-01 -4.01543468e-01
2.70781785e-01 8.32716286e-01 -8.10048804e-02 -1.04150569e+00
-3.37116182e-01 -3.67900699e-01 -5.79789579e-01 5.69617450e-01
7.21375644e-02 4.70777392e-01 -1.13153255e+00 5.71220458e-01
-4.79742400e-02 -2.55912632e-01 -3.66216451e-01 5.49947560e-01
1.26341844e+00 5.93780935e-01 -3.00629020e-01 4.64319080e-01
2.00811267e+00 -1.05253553e+00 -7.08499610e-01 -7.91504025e-01
4.36974436e-01 -1.09247935e+00 8.63970697e-01 5.38001955e-01
-4.39670712e-01 -7.32311666e-01 -1.49743974e+00 7.76973665e-02
-9.31627870e-01 1.84352338e-01 1.06773055e+00 8.34558368e-01
-4.91608113e-01 2.75093108e-01 -4.54571217e-01 -5.35010695e-01
1.12036312e+00 3.06737214e-01 -1.42978445e-01 -2.89936393e-01
-9.10945535e-01 4.80121762e-01 5.90893209e-01 2.74986237e-01
-1.49302852e+00 -1.95662379e-01 -9.73852158e-01 3.62745821e-01
5.01327217e-01 -8.53129998e-02 1.17271900e+00 -9.00392175e-01
-8.81935537e-01 1.29973137e+00 3.56846064e-01 -4.13953811e-01
1.75654553e-02 -2.81012475e-01 -9.26185787e-01 8.64242539e-02
2.85751611e-01 5.58428407e-01 9.99093175e-01 -1.09938574e+00
-7.89363265e-01 -5.17900765e-01 -7.51444101e-02 -3.59268606e-01
-3.52427393e-01 3.21583986e-01 -2.71587223e-01 -7.59926677e-01
4.10675444e-02 -7.54546106e-01 1.06509347e-02 5.90752549e-02
-6.64096057e-01 -5.92292428e-01 6.50304139e-01 -3.33911717e-01
1.08827019e+00 -2.30435514e+00 -1.18950200e+00 -4.29316610e-02
5.51493526e-01 3.86326075e-01 -4.16383237e-01 3.71528387e-01
-2.40596890e-01 -2.25535005e-01 4.29832429e-01 -2.81680729e-02
3.31850082e-01 -1.85078099e-01 -1.60864547e-01 5.68610668e-01
1.97011188e-01 1.25310886e+00 -4.51474845e-01 -4.06853497e-01
2.72237360e-01 2.35166445e-01 -1.41495913e-01 -1.62695721e-01
-7.50825852e-02 -2.63928503e-01 -6.22228563e-01 1.89746511e+00
9.86483157e-01 -4.21036869e-01 -3.92609864e-01 -2.46890783e-01
2.12197617e-01 1.34928152e-01 -1.04524052e+00 1.07842171e+00
-2.42902532e-01 1.00610840e+00 -1.30100638e-01 -9.44063306e-01
1.35697961e+00 -8.21312591e-02 6.37124062e-01 -8.90356719e-01
6.73371732e-01 5.37333488e-01 -4.90233541e-01 -3.89194638e-01
5.26575863e-01 -1.17773421e-01 -7.72336274e-02 1.25727564e-01
7.10173622e-02 3.43687117e-01 2.51736194e-01 -1.70878217e-01
1.00583375e+00 -3.52796137e-01 4.42434192e-01 -6.84014410e-02
1.87375396e-01 2.44794473e-01 5.19574583e-01 1.01738596e+00
-5.59137821e-01 4.35768366e-01 5.37456274e-01 -8.03148508e-01
-8.15920115e-01 -1.25654292e+00 -5.59103847e-01 1.08336568e+00
2.34901950e-01 -6.06261864e-02 -3.87925893e-01 -5.51057220e-01
6.75695837e-01 1.94279194e-01 -8.24919224e-01 -2.17317462e-01
-6.15699172e-01 -7.47394383e-01 7.13966906e-01 1.00626421e+00
6.77764177e-01 -1.58316994e+00 -2.09615782e-01 3.05169910e-01
4.92375344e-01 -9.39945817e-01 -9.64641213e-01 4.01974857e-01
-8.19551766e-01 -1.44966197e+00 -5.14374256e-01 -1.53816521e+00
3.09581012e-01 3.78304482e-01 1.23273540e+00 -5.05566716e-01
-8.53678226e-01 -1.13339312e-01 -2.29103312e-01 -8.50180745e-01
-3.43204558e-01 -1.21571392e-01 -1.78116158e-01 3.97634834e-01
1.21107161e+00 -1.64372876e-01 -6.00249052e-01 6.24081135e-01
-9.36520040e-01 -7.29613841e-01 9.82036829e-01 7.42030501e-01
7.35002339e-01 1.86641157e-01 6.64258480e-01 -8.67038429e-01
3.98911983e-01 -4.01596129e-01 -9.98864353e-01 2.22443864e-02
-1.08367443e+00 -5.60647726e-01 2.54743099e-01 -7.32949197e-01
-6.18841946e-01 3.32626045e-01 -2.57271975e-01 -4.64227021e-01
-1.22725300e-01 5.55310190e-01 -3.93921614e-01 -1.74104661e-01
7.87376046e-01 1.40702665e-01 2.28876770e-02 -9.42598104e-01
2.80231744e-01 1.30297148e+00 5.90396047e-01 7.36594051e-02
6.74712360e-01 6.06577754e-01 -2.45194390e-01 -7.19031751e-01
-9.82261121e-01 -1.05552244e+00 -4.45838958e-01 -2.05781702e-02
6.87474668e-01 -1.00514233e+00 -1.17503679e+00 1.18980193e+00
-8.66538346e-01 -6.73277080e-02 -4.21930552e-01 2.17434689e-01
-3.55482906e-01 1.12747289e-01 -9.76922333e-01 -4.03808266e-01
-4.40784603e-01 -1.00562441e+00 1.27614641e+00 2.27092355e-01
3.23702216e-01 -7.75100231e-01 -1.91407531e-01 7.14079976e-01
4.37863559e-01 3.53934109e-01 4.91830707e-01 -6.74565732e-01
-8.96158040e-01 -1.11158931e+00 -7.18758702e-01 5.02494037e-01
3.67058963e-01 -6.42785013e-01 -1.00780678e+00 -3.47668082e-01
-4.52262759e-02 -5.51216900e-01 1.13749063e+00 4.39243227e-01
9.36668336e-01 1.82980895e-02 -4.28588033e-01 6.83644414e-01
1.38005245e+00 6.78754032e-01 7.06669569e-01 8.22236657e-01
1.10186017e+00 2.87949979e-01 7.65692115e-01 4.96241391e-01
3.57190706e-02 5.31078458e-01 7.21767366e-01 -2.17433363e-01
-1.78730965e-01 -1.91548347e-01 4.73001361e-01 3.45400274e-01
3.97456378e-01 -3.97979617e-01 -7.19517231e-01 6.01673841e-01
-1.42211473e+00 -6.49779081e-01 -9.42473412e-02 1.78670633e+00
3.05209607e-01 3.64317507e-01 3.54764104e-01 -8.54984522e-02
8.70032966e-01 4.98039991e-01 -1.02172792e+00 -2.06479356e-01
-2.60838658e-01 2.06362441e-01 1.08820617e+00 -5.84942140e-02
-1.52588868e+00 8.95166755e-01 6.71337318e+00 1.07404780e+00
-1.08354414e+00 2.50656992e-01 6.76638424e-01 1.54892309e-02
3.20512921e-01 -3.44929338e-01 -1.29632032e+00 2.64212817e-01
5.76289535e-01 -8.23278502e-02 3.32868636e-01 1.57853019e+00
-1.36162817e-01 -1.11959390e-02 -1.10875344e+00 1.23878396e+00
4.26745474e-01 -1.55655897e+00 -1.33360699e-01 6.68693125e-01
6.57202542e-01 3.10947567e-01 2.03754246e-01 3.70589763e-01
4.19241637e-02 -1.39871192e+00 9.73313510e-01 -2.11234882e-01
8.86760592e-01 -7.65256166e-01 9.86470520e-01 -3.30717981e-01
-1.86946404e+00 -3.06954056e-01 -5.00719488e-01 4.14806813e-01
1.34849921e-01 6.65964603e-01 -5.27022123e-01 1.05208091e-01
1.05818558e+00 1.23200738e+00 -9.05682743e-01 1.00217974e+00
2.02492461e-01 5.53248405e-01 -4.64600623e-01 -2.49160841e-01
4.15038764e-01 1.07923515e-01 1.19696885e-01 1.13113856e+00
2.20218748e-02 -7.69282937e-01 3.58893603e-01 1.03141105e+00
-4.69126016e-01 7.19901696e-02 -6.55879676e-01 -6.36385918e-01
2.59274960e-01 1.15780938e+00 -1.01878238e+00 -6.33929074e-01
-7.54293382e-01 8.57779264e-01 -3.25143099e-01 5.43787479e-02
-6.95323527e-01 -8.00529838e-01 8.13083947e-01 5.37535369e-01
7.40184009e-01 -8.38961825e-02 -3.89337391e-01 -8.02336991e-01
2.12955132e-01 -7.46533155e-01 6.17085695e-01 -4.35857624e-01
-1.70620680e+00 3.67310584e-01 -7.44635165e-02 -1.41646552e+00
4.57974762e-01 -1.42306900e+00 -5.36435902e-01 4.50964123e-01
-1.56895125e+00 -1.53228593e+00 -6.00381315e-01 1.05171122e-01
6.94732845e-01 -7.14297295e-01 1.75102383e-01 8.99636388e-01
-4.93127406e-01 7.91875839e-01 3.87979180e-01 5.50343812e-01
6.18817687e-01 -1.00520349e+00 8.17880452e-01 2.89770961e-01
8.68803784e-02 4.99187171e-01 2.41074488e-01 -7.41291702e-01
-1.51651275e+00 -1.31122100e+00 8.37290764e-01 -4.85806167e-01
9.79534805e-01 -5.21135747e-01 -7.18182743e-01 1.04255021e+00
-8.14909115e-03 3.32320809e-01 4.70902234e-01 -1.19564392e-01
-7.07412541e-01 -5.67595661e-01 -1.11520326e+00 -1.67199224e-01
1.00830245e+00 -5.63081384e-01 -1.18462339e-01 7.55743563e-01
3.42226714e-01 -1.80925474e-01 -1.07632196e+00 5.45317270e-02
7.98517704e-01 -6.67491376e-01 1.11489344e+00 -1.52482897e-01
2.22920075e-01 -2.08138376e-01 -1.77151114e-01 -5.94554603e-01
-4.84600872e-01 -3.17946345e-01 -1.11577816e-01 1.37827742e+00
-2.76987012e-02 -1.10167491e+00 1.31963944e+00 -5.22774220e-01
-2.88453907e-01 -6.53685927e-01 -9.38233852e-01 -1.34629560e+00
-1.94431860e-02 -4.29185450e-01 8.24413955e-01 7.23484397e-01
-7.90330350e-01 4.18209523e-01 -3.52349281e-01 9.37335044e-02
6.27878249e-01 3.78696531e-01 4.71720964e-01 -1.10063744e+00
-4.34675336e-01 -6.88825905e-01 -9.04866457e-01 -1.23553228e+00
-3.33044976e-01 -8.23256195e-01 5.48128970e-02 -1.22934628e+00
1.95715070e-01 -3.43082130e-01 -6.40239596e-01 5.58480263e-01
2.57445186e-01 9.18304026e-01 8.12312495e-03 4.74371821e-01
-4.63910908e-01 3.73119593e-01 1.11566532e+00 -8.59952450e-01
-1.25290886e-01 2.81217117e-02 -7.01080918e-01 5.63370645e-01
8.77029598e-01 -5.42063296e-01 1.82084978e-01 8.63577649e-02
5.17004766e-02 -5.40875971e-01 3.18854779e-01 -7.31206179e-01
-2.88903803e-01 4.35761437e-02 4.24815536e-01 -1.05893755e+00
2.51908839e-01 -4.64003742e-01 8.10682178e-02 8.55096459e-01
1.44024596e-01 -5.80175668e-02 1.90363973e-01 5.73867917e-01
-4.73548949e-01 -4.48137552e-01 9.39131081e-01 -3.13058138e-01
-1.29519534e+00 3.07329178e-01 -5.54298937e-01 -4.15757924e-01
1.12414253e+00 -6.82339311e-01 -4.21614408e-01 -2.62261536e-02
-5.08257985e-01 -1.21106696e-03 2.02049583e-01 7.44227231e-01
5.86310089e-01 -1.81220412e+00 -4.85397786e-01 3.08702648e-01
9.81802702e-01 -2.90261507e-01 -2.18308210e-01 3.79933000e-01
-1.20856643e+00 8.28404963e-01 -4.37919408e-01 -5.22721291e-01
-1.29677427e+00 9.53004599e-01 3.46578211e-01 -1.88805476e-01
-6.14053726e-01 7.67734587e-01 6.06836796e-01 -1.78208560e-01
4.20271903e-01 -2.62033761e-01 -1.53755382e-01 3.50045741e-01
6.19021773e-01 3.58273178e-01 3.13526809e-01 -6.42484963e-01
-5.75197935e-01 7.07115233e-01 -2.46044949e-01 6.97705448e-01
1.28031290e+00 6.72009736e-02 1.67600304e-01 2.83241570e-01
1.40590727e+00 -3.50182056e-01 -1.09307218e+00 -3.08887988e-01
9.61930975e-02 -7.01947272e-01 -2.62978836e-03 -5.26875556e-01
-1.44766772e+00 8.49139154e-01 1.44717550e+00 2.97473848e-01
1.05712521e+00 5.27366102e-01 1.33441210e+00 3.36330861e-01
4.09695059e-01 -1.11130321e+00 6.85984015e-01 2.31749639e-01
8.07360470e-01 -1.58778942e+00 -4.90651140e-03 -7.34817982e-01
-2.52436399e-01 1.13199222e+00 4.66969192e-01 -3.24537754e-01
7.77777553e-01 6.02522828e-02 4.35641497e-01 -8.24702621e-01
-1.06856814e-02 -4.13860410e-01 3.10756743e-01 6.68917775e-01
3.00234705e-01 3.26676220e-01 1.40267849e-01 4.34187621e-01
-5.39700948e-02 -3.69182602e-02 3.84217769e-01 1.32713544e+00
-7.12029397e-01 -1.03033113e+00 -7.74102807e-01 6.38885200e-01
-3.30323368e-01 2.78227590e-02 -5.57310402e-01 7.65707970e-01
3.44298273e-01 7.77550757e-01 1.43502891e-01 -4.37660873e-01
3.50198418e-01 -3.89210403e-01 2.13288724e-01 -6.20474756e-01
-5.13476670e-01 4.15592521e-01 9.73995477e-02 -4.96930957e-01
3.17648858e-01 -5.25870562e-01 -6.87851667e-01 -5.75237274e-01
-6.33247852e-01 -5.10835588e-01 6.70278072e-01 5.36394298e-01
2.49377459e-01 6.23438656e-01 8.40498745e-01 -9.10722852e-01
-5.86918592e-01 -1.00440967e+00 -1.43381917e+00 4.67314243e-01
9.10503939e-02 -8.08864117e-01 -7.11691380e-02 1.99390441e-01] | [9.32280158996582, 1.3059922456741333] |
f77ec3f8-03e2-4b12-86fb-5c926e95889c | visual-knowledge-tracing | 2207.10157 | null | https://arxiv.org/abs/2207.10157v2 | https://arxiv.org/pdf/2207.10157v2.pdf | Visual Knowledge Tracing | Each year, thousands of people learn new visual categorization tasks -- radiologists learn to recognize tumors, birdwatchers learn to distinguish similar species, and crowd workers learn how to annotate valuable data for applications like autonomous driving. As humans learn, their brain updates the visual features it extracts and attend to, which ultimately informs their final classification decisions. In this work, we propose a novel task of tracing the evolving classification behavior of human learners as they engage in challenging visual classification tasks. We propose models that jointly extract the visual features used by learners as well as predicting the classification functions they utilize. We collect three challenging new datasets from real human learners in order to evaluate the performance of different visual knowledge tracing methods. Our results show that our recurrent models are able to predict the classification behavior of human learners on three challenging medical image and species identification tasks. | ['Oisin Mac Aodha', 'Pietro Perona', 'Neehar Kondapaneni'] | 2022-07-20 | null | null | null | null | ['classification'] | ['methodology'] | [-4.65078512e-03 -1.62349001e-01 -1.83471918e-01 -3.54431719e-01
-7.03893676e-02 -6.58005536e-01 6.53367937e-01 4.36519057e-01
-9.25823569e-01 6.16761029e-01 -1.17976978e-01 -1.92772001e-01
2.21566990e-01 -3.72070760e-01 -4.60066110e-01 -5.10221541e-01
-1.23499550e-01 5.21715224e-01 3.19736242e-01 1.98565740e-02
2.38941550e-01 4.69823688e-01 -1.73784244e+00 5.08332431e-01
7.24394619e-01 8.57604086e-01 3.10672909e-01 1.03299820e+00
-5.98146655e-02 1.45519638e+00 -6.74227238e-01 -3.40535611e-01
-1.01571542e-03 -1.71191115e-02 -9.05265391e-01 5.02834879e-02
3.87816399e-01 -8.26883391e-02 -5.25683999e-01 8.90855730e-01
8.81094411e-02 2.25059226e-01 9.00855362e-01 -1.30564940e+00
-8.30516636e-01 5.67906260e-01 -5.87788463e-01 9.22554433e-01
2.82398406e-02 4.83576715e-01 6.24037325e-01 -6.73701346e-01
8.59824002e-01 9.50124443e-01 5.91153979e-01 9.22026873e-01
-9.31238174e-01 -7.39113271e-01 3.49119544e-01 6.11753941e-01
-1.16004360e+00 -4.67173338e-01 4.30814058e-01 -9.65135455e-01
5.04217207e-01 1.00998402e-01 1.03765166e+00 1.27333510e+00
2.32748151e-01 9.61906135e-01 1.09493017e+00 2.36928109e-02
1.96653128e-01 6.30018175e-01 6.01890147e-01 1.22049487e+00
2.09654635e-03 1.53262243e-01 -9.88532424e-01 -3.14498544e-02
1.60841897e-01 3.48047882e-01 -2.03797281e-01 -3.08671266e-01
-1.33002877e+00 6.90534055e-01 8.17565143e-01 4.02922928e-01
-2.83012956e-01 3.75535667e-01 2.57316440e-01 2.54749417e-01
2.19822168e-01 4.75702256e-01 -7.23478645e-02 7.25851133e-02
-4.90609676e-01 -4.07013856e-02 4.90494221e-01 6.81162953e-01
7.75700092e-01 -1.11502036e-01 -6.39689863e-01 7.42466331e-01
1.71086490e-01 2.90767014e-01 9.85320866e-01 -6.75464630e-01
-1.72073141e-01 6.94850445e-01 -1.21012993e-01 -6.25233352e-01
-6.43428624e-01 -7.92988539e-01 -5.59822023e-01 1.03128523e-01
4.62396890e-01 1.32137462e-01 -1.20902860e+00 1.41898274e+00
2.47740537e-01 1.50106594e-01 5.68134757e-03 6.61236644e-01
9.64895606e-01 3.45879346e-01 3.00986260e-01 5.89551553e-02
1.44494748e+00 -9.66878951e-01 -3.30758631e-01 -3.25535297e-01
4.70948875e-01 -1.28794834e-01 8.26114237e-01 5.80915034e-01
-6.94401145e-01 -7.66336799e-01 -8.02137256e-01 1.30973801e-01
-3.37126672e-01 1.84450089e-03 7.86250770e-01 4.23781186e-01
-9.83161449e-01 2.29562059e-01 -8.96356225e-01 -4.51187670e-01
1.15529752e+00 9.59866047e-02 -1.39226213e-01 -4.26661298e-02
-5.53382814e-01 8.50904763e-01 2.14117259e-01 -1.63461655e-01
-1.71415722e+00 -1.03933394e+00 -4.97402310e-01 -3.36407833e-02
1.60485163e-01 -7.27536023e-01 1.43868065e+00 -1.15230393e+00
-8.74474168e-01 1.38955510e+00 -1.26548275e-01 -6.69690788e-01
6.96719766e-01 2.55079389e-01 -1.88329503e-01 -2.86459737e-02
-1.15481829e-02 9.16366160e-01 1.01182437e+00 -1.26141083e+00
-1.12971902e+00 -4.70660150e-01 -1.12931110e-01 7.79702067e-02
-5.54419041e-01 -5.37732661e-01 -2.20207661e-01 -3.76815349e-01
-5.12663007e-01 -1.01556540e+00 -1.99971199e-01 2.18883872e-01
-2.41781667e-01 -4.75462973e-01 6.31860137e-01 -3.61195534e-01
7.37016559e-01 -2.27472234e+00 4.21848506e-01 1.77960157e-01
1.01205957e+00 5.57259060e-02 -1.01821475e-01 -2.27962956e-01
4.01209742e-01 -1.03948034e-01 -1.57261178e-01 -1.88423112e-01
-3.72493356e-01 2.58987457e-01 -2.53003895e-01 4.56017554e-01
1.76977888e-02 1.13867104e+00 -1.09360647e+00 -4.53187734e-01
-3.53720300e-02 2.76844442e-01 -1.67769611e-01 4.20306325e-01
-1.08038560e-01 6.70672119e-01 -4.95207220e-01 6.66659176e-01
2.00054664e-02 -6.69531405e-01 -1.40425852e-02 1.77802876e-01
1.03272736e-01 -4.34844762e-01 -2.33788475e-01 1.47575927e+00
-3.99922520e-01 1.13716650e+00 -2.07110718e-01 -1.13518035e+00
8.15072298e-01 -2.73166507e-01 2.90147156e-01 -8.95625532e-01
2.23249763e-01 -1.67303726e-01 2.22122371e-01 -6.28258765e-01
2.85315275e-01 -1.48531929e-01 -4.47568186e-02 4.62186128e-01
2.08550215e-01 1.05888188e-01 2.25776941e-01 3.84189725e-01
1.30771780e+00 -4.21459079e-01 4.01294321e-01 -3.38556021e-01
5.42183816e-01 2.28404671e-01 3.42384249e-01 9.07804012e-01
-6.49337053e-01 -8.31572637e-02 2.44825974e-01 -1.05661416e+00
-6.16746306e-01 -1.23562443e+00 6.36156723e-02 1.65147352e+00
1.93054780e-01 1.56787261e-01 -6.24800503e-01 -9.84849870e-01
2.84634829e-01 6.18889987e-01 -1.30402577e+00 -4.20865148e-01
-2.00054079e-01 -5.06662846e-01 4.15997863e-01 5.06761432e-01
2.31497064e-01 -1.33886707e+00 -1.30015373e+00 -7.44348317e-02
1.75232083e-01 -8.78593683e-01 -5.25727808e-01 3.91117930e-01
-4.06020671e-01 -1.43022692e+00 -7.82457054e-01 -9.24139917e-01
9.28761184e-01 3.35488766e-01 1.00929070e+00 4.98239666e-01
-1.19212794e+00 9.78767693e-01 -1.23012066e-01 -8.75829518e-01
-5.25676012e-01 2.23393977e-01 6.08836934e-02 3.38523597e-01
3.29467982e-01 -3.50486279e-01 -4.96120214e-01 2.25521415e-01
-6.26666605e-01 1.04641788e-01 4.80616331e-01 7.78459668e-01
5.51729858e-01 -4.44535643e-01 4.76927102e-01 -1.15450060e+00
6.03791416e-01 -6.78370535e-01 -4.72401589e-01 7.28551567e-01
-3.91112566e-01 3.81367117e-01 6.44273818e-01 -7.90173471e-01
-7.39873052e-01 3.13850939e-01 5.35938025e-01 -5.92288435e-01
-1.68845505e-01 2.04731047e-01 7.58320570e-01 -1.83563918e-01
1.00151682e+00 4.66956884e-01 -2.04378918e-01 -4.60925288e-02
2.42545530e-01 5.14092684e-01 9.11714792e-01 -2.42428005e-01
8.59611094e-01 6.73544705e-01 -3.07443947e-01 -1.12446332e+00
-1.19139969e+00 -5.78883410e-01 -7.34234214e-01 -8.33007097e-01
1.08448386e+00 -9.18457031e-01 -1.30305386e+00 4.80925590e-01
-9.05347526e-01 -6.35394871e-01 -4.31403011e-01 2.98761815e-01
-2.82065898e-01 -3.72640602e-02 -2.21995950e-01 -7.88274944e-01
-2.13993222e-01 -9.26262319e-01 7.01839447e-01 6.79639161e-01
-2.13098988e-01 -1.07806039e+00 2.01662302e-01 6.21209741e-01
4.25493479e-01 -3.30807036e-03 9.85232413e-01 -5.74155569e-01
-6.01646900e-01 3.84238988e-01 -1.60162255e-01 -2.02885121e-01
2.97963750e-02 -2.01141044e-01 -1.22670531e+00 -4.41271424e-01
-4.76363063e-01 -5.82201958e-01 1.36871040e+00 1.13262966e-01
1.62057233e+00 -5.28559722e-02 -6.58030748e-01 8.04949462e-01
7.72401035e-01 2.07684159e-01 -2.51693148e-02 -4.48638834e-02
7.41916418e-01 7.69037426e-01 1.18745089e-01 2.96567112e-01
5.43804526e-01 1.81284055e-01 4.10431743e-01 1.41594559e-01
-2.16658682e-01 -1.32031307e-01 3.40119600e-01 7.53276527e-01
-2.18603566e-01 -1.34526625e-01 -1.35197544e+00 6.04105055e-01
-1.75084507e+00 -8.15834641e-01 3.14950496e-01 1.82429767e+00
7.64145494e-01 -1.41550032e-02 1.94095984e-01 -3.19444895e-01
4.71716821e-01 -1.66916028e-01 -1.18201804e+00 -1.03565708e-01
2.01118231e-01 1.00597829e-01 5.49243450e-01 2.86379009e-01
-9.99392688e-01 9.28756297e-01 6.31648636e+00 2.31819138e-01
-1.40612531e+00 2.95925856e-01 9.36107278e-01 -1.38768360e-01
-1.10778220e-01 -5.19254982e-01 -6.51991487e-01 2.10762218e-01
7.73158014e-01 -4.03046727e-01 5.39219081e-01 7.48620570e-01
-4.57696706e-01 -2.42923364e-01 -1.30823469e+00 1.45246840e+00
4.57923412e-01 -1.43225360e+00 1.10434825e-02 5.66989034e-02
7.02701330e-01 2.73850054e-01 5.37722051e-01 4.11863357e-01
7.82385409e-01 -1.34089148e+00 7.79145420e-01 1.12265670e+00
4.03762937e-01 -5.35724342e-01 2.76750803e-01 4.76829916e-01
-9.60493386e-01 -6.04801476e-01 -1.61273226e-01 1.15203880e-01
-4.74754363e-01 7.02735260e-02 -1.43133688e+00 -1.49642453e-01
9.00821269e-01 9.62773204e-01 -1.23150778e+00 1.17171633e+00
-1.13222688e-01 5.64265013e-01 1.77654251e-01 -3.74129713e-01
1.49735764e-01 4.93879884e-01 2.89719492e-01 1.01793098e+00
6.73927292e-02 -5.34568243e-02 3.55139643e-01 9.13422942e-01
-3.92498165e-01 -1.00996494e-01 -4.49797213e-01 -2.05675855e-01
2.02846766e-01 1.38399446e+00 -9.78281677e-01 -2.67173797e-01
-6.17144182e-02 9.22153771e-01 8.31562161e-01 4.05205607e-01
-6.91818178e-01 1.05011299e-01 6.85290039e-01 1.24696642e-01
2.30938047e-01 -6.33467212e-02 9.24977649e-04 -1.05994475e+00
-3.54712486e-01 -6.54824734e-01 6.10049427e-01 -7.38742650e-01
-1.43194044e+00 8.04666281e-01 -3.22598696e-01 -8.61726820e-01
-1.61892712e-01 -7.87370741e-01 -5.80664575e-01 1.67578325e-01
-1.62878132e+00 -9.44501400e-01 -9.05069828e-01 8.10467839e-01
6.17117345e-01 -7.16630101e-01 3.31265509e-01 -1.21420398e-01
-4.81514990e-01 5.41580200e-01 -1.21837325e-01 3.48435134e-01
5.88503122e-01 -9.89121437e-01 1.45367980e-01 4.23981160e-01
8.20368171e-01 1.02362849e-01 3.81854951e-01 -3.69520247e-01
-1.31371772e+00 -1.08109212e+00 2.35529810e-01 -7.02681363e-01
5.43545187e-01 -4.13441896e-01 -1.00317395e+00 5.30917883e-01
-1.07455038e-01 4.43665266e-01 7.67468452e-01 -4.30759303e-02
-5.79852462e-01 -1.89308792e-01 -9.04144824e-01 2.05649346e-01
1.17422926e+00 -6.77419007e-01 -5.28961837e-01 3.87918115e-01
4.56343889e-01 -6.92537129e-02 -3.90207201e-01 -4.65540253e-02
6.24517083e-01 -7.83247650e-01 8.13568354e-01 -1.19025910e+00
7.30725303e-02 -1.41800448e-01 1.31913587e-01 -1.44887543e+00
-2.76199579e-01 -1.84501842e-01 -1.44898593e-01 5.85355818e-01
4.50612247e-01 -4.90665942e-01 7.78299093e-01 1.57424003e-01
1.45871148e-01 -5.25156677e-01 -8.37959409e-01 -3.35740775e-01
-1.72999486e-01 -1.23530202e-01 1.45042017e-01 1.15007353e+00
-2.11946905e-01 3.98599893e-01 -9.69332457e-02 -1.74619518e-02
8.26542735e-01 1.03405339e-03 7.31221259e-01 -1.61656606e+00
-3.24984401e-01 -6.09651148e-01 -6.21713877e-01 -4.91749644e-01
3.74080628e-01 -1.38304698e+00 1.94182098e-02 -1.11274350e+00
6.03731275e-01 -4.12849456e-01 -5.86329043e-01 7.05211699e-01
-2.84760535e-01 2.87296325e-01 3.19168568e-01 1.73900411e-01
-9.43368137e-01 2.23964334e-01 1.04524136e+00 -7.73621380e-01
-1.47590712e-01 3.78451310e-02 -6.87481344e-01 7.41817653e-01
5.24102569e-01 -5.46981394e-01 -3.25706571e-01 -4.56814915e-01
2.50898540e-01 -1.96746051e-01 7.38909304e-01 -1.17837727e+00
6.13415778e-01 -2.19030723e-01 8.07693660e-01 -3.36002797e-01
6.01213006e-03 -4.67915088e-01 -2.18325973e-01 1.06982219e+00
-6.17985308e-01 -7.42508378e-03 2.02124730e-01 8.07432234e-01
1.38097405e-01 -1.18688628e-01 9.82569814e-01 -2.86294907e-01
-1.24101770e+00 5.53357065e-01 -5.78104734e-01 3.31902117e-01
1.39492285e+00 1.53695777e-01 -5.34592628e-01 -2.90584743e-01
-9.98827219e-01 6.65118992e-01 1.70520470e-01 7.90303349e-01
7.71449566e-01 -8.94586504e-01 -9.42115843e-01 1.20785274e-01
5.65235972e-01 -3.88572663e-01 3.73305023e-01 6.66006446e-01
-4.81510371e-01 8.08060169e-02 -5.21471143e-01 -9.91835833e-01
-1.43118250e+00 8.93494070e-01 6.32504642e-01 -2.33940542e-01
-3.60480607e-01 1.14540780e+00 6.95201159e-01 -3.01816046e-01
5.30755460e-01 -2.30716333e-01 -5.06485164e-01 3.18722248e-01
8.12124074e-01 2.46813297e-01 -3.02853584e-01 -4.60964620e-01
-3.35568786e-01 2.93608248e-01 -4.27397162e-01 1.89453036e-01
1.30225980e+00 1.42583057e-01 3.13420206e-01 7.33416736e-01
9.23836410e-01 -4.16998982e-01 -1.27819419e+00 -5.07602155e-01
1.29356891e-01 -1.54438764e-01 -2.80790329e-01 -8.56775284e-01
-1.11204624e+00 1.30187511e+00 9.39899385e-01 -9.79930833e-02
8.59022439e-01 4.73149270e-01 3.39505255e-01 7.63569057e-01
4.21642989e-01 -1.05813122e+00 5.41592538e-01 3.46960187e-01
6.86406076e-01 -1.36608601e+00 -1.05946280e-01 1.45646676e-01
-8.62363696e-01 9.91008937e-01 7.74117291e-01 1.41402915e-01
5.39078653e-01 7.52726793e-02 1.52254984e-01 -3.32752168e-01
-1.15454757e+00 -3.99004847e-01 5.93110621e-01 8.72259319e-01
-1.14483638e-02 1.75963521e-01 5.21864593e-01 4.11868721e-01
-2.49156505e-01 -1.38406590e-01 5.36768079e-01 6.53814495e-01
-7.27348447e-01 -5.39165199e-01 -1.68901592e-01 7.43779182e-01
2.65942849e-02 2.46670187e-01 -7.62034297e-01 4.01636571e-01
2.67178804e-01 4.92122084e-01 4.27928507e-01 -4.14004862e-01
1.31024674e-01 2.31394365e-01 6.38196111e-01 -6.56456590e-01
-7.59875655e-01 -8.48671556e-01 -3.96172762e-01 -3.63689959e-01
-4.54648733e-01 -7.46014237e-01 -1.25721610e+00 3.08257658e-02
3.10465574e-01 9.81894806e-02 4.68748480e-01 9.35990691e-01
-5.52269444e-02 9.29847419e-01 6.20862722e-01 -4.13810700e-01
-3.52633715e-01 -6.66741908e-01 -3.49576861e-01 5.18611729e-01
5.84330618e-01 -6.89541161e-01 -1.72351241e-01 3.76771003e-01] | [9.77791690826416, 2.083139419555664] |
8d82b9cd-cff6-4c26-9f88-1bf30e14e2b0 | zoho-at-semeval-2019-task-9-semi-supervised | 1902.10623 | null | http://arxiv.org/abs/1902.10623v2 | http://arxiv.org/pdf/1902.10623v2.pdf | Zoho at SemEval-2019 Task 9: Semi-supervised Domain Adaptation using Tri-training for Suggestion Mining | This paper describes our submission for the SemEval-2019 Suggestion Mining
task. A simple Convolutional Neural Network (CNN) classifier with contextual
word representations from a pre-trained language model was used for sentence
classification. The model is trained using tri-training, a semi-supervised
bootstrapping mechanism for labelling unseen data. Tri-training proved to be an
effective technique to accommodate domain shift for cross-domain suggestion
mining (Subtask B) where there is no hand labelled training data. For in-domain
evaluation (Subtask A), we use the same technique to augment the training set.
Our system ranks thirteenth in Subtask A with an $F_1$-score of 68.07 and third
in Subtask B with an $F_1$-score of 81.94. | ['Sri Ananda Seelan', 'Sai Prasanna'] | 2019-02-27 | zoho-at-semeval-2019-task-9-semi-supervised-1 | https://aclanthology.org/S19-2225 | https://aclanthology.org/S19-2225.pdf | semeval-2019-6 | ['suggestion-mining'] | ['natural-language-processing'] | [ 1.30671173e-01 4.49429035e-01 -2.43727013e-01 -6.77010417e-01
-7.70692110e-01 -4.33302432e-01 5.11211932e-01 5.12441456e-01
-8.99502099e-01 1.14285982e+00 2.31546193e-01 -8.75273347e-01
1.19603530e-01 -4.61091667e-01 -5.24331808e-01 -7.32786283e-02
-8.79398361e-02 6.39340818e-01 2.38769814e-01 -6.80644393e-01
4.45772588e-01 4.39934097e-02 -1.27995372e+00 9.92445588e-01
5.36491692e-01 8.83209050e-01 1.25422508e-01 6.30517185e-01
-4.99971062e-01 4.51865464e-01 -9.80252922e-01 -3.24791431e-01
6.31744936e-02 -2.63428658e-01 -1.32555068e+00 -1.23473004e-01
4.05004084e-01 -2.61898078e-02 7.93867782e-02 6.64916635e-01
4.73486632e-01 2.25301981e-01 5.75097084e-01 -7.61229634e-01
-4.75598931e-01 8.67550552e-01 -3.02043438e-01 5.17195046e-01
5.55286109e-01 -2.69228727e-01 7.44926989e-01 -1.25788438e+00
6.67820752e-01 9.53340232e-01 6.52834892e-01 7.53592312e-01
-1.33429193e+00 -7.88865507e-01 2.89508188e-03 6.43862709e-02
-1.18847454e+00 -2.27427766e-01 6.92725420e-01 -3.25091928e-01
1.68223011e+00 9.88335013e-02 1.23117261e-01 1.22262788e+00
-1.07012875e-03 5.36618590e-01 1.21958458e+00 -8.93050194e-01
4.94800657e-01 5.26008844e-01 3.99033904e-01 2.52951533e-01
-4.71467748e-02 5.20966351e-02 -6.41348481e-01 -3.05595249e-01
3.24277192e-01 -6.15656614e-01 6.16986342e-02 1.41119137e-01
-8.54064941e-01 1.00217032e+00 1.03272215e-01 4.91309941e-01
-3.04466069e-01 -4.38865572e-01 9.60055828e-01 8.65233243e-01
8.16717744e-01 8.89798701e-01 -1.26771080e+00 -2.97373179e-02
-1.09228408e+00 3.74554515e-01 7.53621280e-01 7.19716132e-01
6.36938214e-01 1.22176498e-01 -2.44711474e-01 1.30861866e+00
-1.36466950e-01 -2.99395561e-01 9.64011669e-01 -6.54577315e-01
5.51600277e-01 5.02597094e-01 1.29000947e-01 -3.54665458e-01
-6.60061598e-01 -5.70654631e-01 -7.65213549e-01 5.05169667e-02
3.58289778e-01 -3.72825950e-01 -7.72454977e-01 1.68364692e+00
-9.62239131e-02 -3.17618579e-01 2.12498039e-01 5.17829239e-01
1.19637620e+00 3.39612126e-01 3.17014396e-01 -3.10349703e-01
1.39559698e+00 -8.14305842e-01 -2.05025762e-01 -3.62551719e-01
1.07796657e+00 -7.39535093e-01 1.09711015e+00 5.44971943e-01
-7.53398955e-01 -9.45021272e-01 -1.08101881e+00 2.33397365e-01
-4.87655908e-01 2.74204344e-01 2.53680199e-01 6.72786653e-01
-1.29180765e+00 4.31333542e-01 -3.13809887e-03 -5.67750573e-01
2.21700117e-01 6.08831942e-01 -5.80466390e-01 -6.96542636e-02
-1.63010836e+00 1.03662002e+00 9.91826057e-01 -6.74393713e-01
-3.80329311e-01 -4.83057946e-01 -8.72666538e-01 -7.73851052e-02
1.70048952e-01 -2.65384346e-01 1.37833941e+00 -1.18236864e+00
-1.35799277e+00 1.35830891e+00 -1.39427319e-01 -8.55568469e-01
2.45641202e-01 -1.01800375e-01 -8.28320384e-01 -1.62776783e-01
4.65244979e-01 8.43159258e-01 9.07247603e-01 -9.47988391e-01
-5.54301500e-01 -1.96015283e-01 4.26105708e-02 1.30026385e-01
-3.30469042e-01 4.23225045e-01 1.57565698e-01 -7.42048025e-01
-3.03560533e-02 -5.88190198e-01 -4.01981205e-01 -8.79711807e-01
-4.24295187e-01 -8.28766823e-01 6.54417515e-01 -6.15870774e-01
1.22516525e+00 -1.80325055e+00 -6.65661573e-01 7.58346096e-02
1.68124571e-01 4.53566670e-01 -1.70749485e-01 3.21469456e-01
-6.55185521e-01 1.93301022e-01 -4.98101488e-02 -3.45005065e-01
-2.55141467e-01 3.26327533e-02 -2.59415150e-01 -5.61411083e-02
5.79694808e-01 5.73666573e-01 -9.10178781e-01 -2.55062580e-01
5.10880947e-02 -1.65249363e-01 -3.42547923e-01 1.43709585e-01
-2.30594218e-01 2.58821577e-01 4.17695083e-02 2.60455936e-01
7.61548638e-01 -1.41483709e-01 3.05775911e-01 4.36470538e-01
-1.29363621e-02 7.85550952e-01 -9.47606146e-01 1.62993598e+00
-3.51404190e-01 6.62292123e-01 -1.21365771e-01 -1.35984695e+00
1.51997483e+00 3.38904828e-01 1.12404115e-01 -8.32347214e-01
-3.80253345e-02 3.45047385e-01 3.53439488e-02 -3.63510519e-01
8.52654219e-01 -3.06520641e-01 -3.48550856e-01 5.84585845e-01
3.35082740e-01 2.08010785e-02 1.73639596e-01 3.19097996e-01
1.40618193e+00 -2.72943079e-02 4.85124826e-01 -5.14712274e-01
7.92497575e-01 1.98616445e-01 3.87899667e-01 8.40786815e-01
-1.48958445e-01 4.92839992e-01 3.55038524e-01 -8.58817518e-01
-1.06480062e+00 -6.58195317e-01 -3.01369041e-01 1.50859261e+00
-5.64884305e-01 -4.39528674e-01 -5.34863830e-01 -1.23095882e+00
-2.86077082e-01 1.05991542e+00 -4.58007634e-01 -4.04012725e-02
-5.53074181e-01 -5.49610853e-01 6.23637438e-01 5.86704254e-01
4.51106548e-01 -1.40874290e+00 -4.84076947e-01 3.95751685e-01
-2.70553857e-01 -1.03023481e+00 -9.79332030e-02 1.02046597e+00
-9.15300667e-01 -6.48781419e-01 -5.66442192e-01 -1.26325917e+00
5.73358297e-01 -1.49894163e-01 1.57774913e+00 7.22054243e-02
5.28445318e-02 -1.92152217e-01 -5.74395001e-01 -5.40234923e-01
-5.64512551e-01 2.51039416e-01 2.82934785e-01 -6.71393633e-01
1.18135345e+00 -8.24351758e-02 -2.84968883e-01 1.69032767e-01
-7.05826998e-01 -1.76500976e-01 3.78090978e-01 1.31633222e+00
3.51350605e-01 -1.54758915e-01 1.20444655e+00 -9.46087539e-01
1.24132097e+00 -6.56624854e-01 -1.67767599e-01 -9.21052173e-02
-8.01938772e-01 -1.56672180e-01 7.88587451e-01 -4.90016282e-01
-7.30539680e-01 1.24234110e-01 -4.80930239e-01 6.81230351e-02
-8.01076472e-01 6.39705360e-01 1.47575006e-01 4.43501681e-01
1.26907086e+00 2.12622825e-02 -2.22742587e-01 -5.91812015e-01
1.79234207e-01 1.03241611e+00 5.45080125e-01 -3.98980945e-01
3.24459016e-01 -1.89687803e-01 -5.63879788e-01 -8.16676080e-01
-9.70648766e-01 -6.36711657e-01 -8.17932487e-01 5.64112850e-02
4.99197274e-01 -7.41688013e-01 -8.31277519e-02 1.80893997e-03
-1.16446412e+00 -4.11220968e-01 -3.60830992e-01 3.62948567e-01
-2.87293941e-01 3.05905700e-01 -1.99890420e-01 -6.40367210e-01
-5.43088377e-01 -6.58030272e-01 6.63167119e-01 5.71560413e-02
-8.66951942e-01 -9.39373493e-01 1.93897933e-01 4.20706213e-01
3.61456424e-01 -3.14708203e-01 9.48792636e-01 -1.69735491e+00
5.21806300e-01 -3.59254569e-01 -1.74637556e-01 5.35579860e-01
-7.62761161e-02 -8.07066560e-01 -1.21541238e+00 -2.00034767e-01
-1.86115839e-02 -8.43631387e-01 9.11144972e-01 2.38450453e-01
1.50535810e+00 -1.30302608e-01 -1.06791846e-01 -6.90585375e-02
8.80059958e-01 1.62995294e-01 4.73762602e-01 8.23601425e-01
-1.48598373e-01 7.00480402e-01 7.07914770e-01 2.90945679e-01
1.84312046e-01 6.55691445e-01 -5.76255098e-02 1.20843332e-02
7.94108436e-02 -1.40640810e-01 3.23359698e-01 4.89767492e-01
5.07688403e-01 -1.87425151e-01 -1.27220619e+00 8.08686972e-01
-1.55224419e+00 -8.28342557e-01 -1.56709448e-01 2.01956391e+00
9.62007463e-01 7.39730895e-01 1.96154013e-01 2.61397421e-01
6.46844566e-01 -2.56077498e-01 -2.21136704e-01 -1.14357650e+00
-1.69400126e-01 7.51030505e-01 2.73176342e-01 3.09049875e-01
-1.40338254e+00 1.06999075e+00 6.39523458e+00 9.12833810e-01
-9.63480413e-01 9.87919942e-02 1.00347269e+00 2.23320469e-01
-4.48553674e-02 -1.53593346e-02 -7.76449502e-01 6.38046205e-01
1.41017342e+00 1.56597555e-01 -3.23756278e-01 9.45825815e-01
1.10158175e-01 -3.79481465e-01 -9.96313870e-01 6.01751447e-01
2.83779830e-01 -1.46620154e+00 -2.02050045e-01 -2.49951392e-01
6.25525475e-01 2.70985603e-01 -9.17902663e-02 9.07646716e-01
3.16556245e-01 -1.15971804e+00 1.71668395e-01 -7.58751929e-02
1.00110030e+00 -8.29420507e-01 1.30913174e+00 6.46393001e-01
-6.40466571e-01 1.19671032e-01 -5.35037875e-01 -4.37438875e-01
-1.01970069e-01 6.41972125e-01 -1.59900630e+00 3.80345047e-01
8.95743787e-01 2.13792384e-01 -5.06263852e-01 8.34463894e-01
-3.08127791e-01 8.83193791e-01 -2.20330939e-01 -1.65453166e-01
1.50882557e-01 2.96564221e-01 4.54081327e-01 1.66315222e+00
-9.80050638e-02 -2.44017303e-01 2.74131626e-01 4.59973812e-01
-2.99345106e-01 2.72859901e-01 -6.39125407e-01 2.89883971e-01
2.95826763e-01 1.00973189e+00 -5.95148683e-01 -5.28423429e-01
-3.21810663e-01 9.42308962e-01 4.90495741e-01 1.21566519e-01
-2.94680536e-01 -4.72383261e-01 1.19942643e-01 5.54632023e-02
1.95860624e-01 2.91563701e-02 -5.96509576e-01 -8.70646060e-01
-2.09503829e-01 -7.34008074e-01 5.17096996e-01 -5.76408207e-01
-1.53142214e+00 9.17174578e-01 8.81095305e-02 -1.17047930e+00
-6.06433988e-01 -6.96074069e-01 -8.16738009e-01 1.19543016e+00
-1.30793273e+00 -9.32781756e-01 1.90480754e-01 4.06982511e-01
8.84162724e-01 -8.57963860e-01 1.29005075e+00 -8.14547166e-02
-1.48571625e-01 9.67223763e-01 7.56048411e-02 2.87990868e-01
9.43572819e-01 -1.43636787e+00 4.26842302e-01 4.42563504e-01
3.67741585e-02 7.92519748e-01 6.86958790e-01 -7.10105538e-01
-3.60035479e-01 -1.17191398e+00 1.66269016e+00 -4.23521727e-01
5.57354629e-01 -2.83281058e-01 -1.06445241e+00 3.09226662e-01
4.18310493e-01 -2.68764406e-01 1.29283452e+00 5.59920371e-01
-3.18620145e-01 2.60573477e-01 -1.39353800e+00 2.65726000e-01
5.28558373e-01 -5.87010622e-01 -1.14793646e+00 4.17609483e-01
8.38642597e-01 -1.78551659e-01 -7.66443133e-01 2.84763068e-01
2.30876684e-01 -5.70172846e-01 8.59569669e-01 -9.90504920e-01
6.53092623e-01 -1.09267957e-03 -1.57984152e-01 -1.28482366e+00
-2.05970719e-01 -3.51329088e-01 1.26997605e-01 9.94112074e-01
9.35222089e-01 -4.28577691e-01 9.62707162e-01 4.02335465e-01
-1.20603934e-01 -9.03627336e-01 -1.18999910e+00 -8.64482582e-01
6.11092508e-01 -8.41865420e-01 3.06507587e-01 1.15476894e+00
4.71465975e-01 8.49531054e-01 -9.52345058e-02 -4.91210759e-01
1.08777359e-01 -1.86990976e-01 5.20482719e-01 -1.22686994e+00
-1.56970844e-01 -3.45075130e-01 -4.64737192e-02 -9.01007295e-01
4.29457605e-01 -9.94829416e-01 2.79220399e-02 -1.25496018e+00
1.07312217e-01 -3.14368725e-01 -6.41207576e-01 5.23102224e-01
4.94679585e-02 2.62431741e-01 -2.40720198e-01 -9.06988084e-02
-6.25817597e-01 1.48049414e-01 5.27878165e-01 -1.73960134e-01
-3.26861352e-01 1.59217760e-01 -8.72823775e-01 6.68884158e-01
1.38149333e+00 -4.96129304e-01 -2.61554033e-01 -1.66482896e-01
1.47662580e-01 -2.27181926e-01 1.05818193e-02 -6.46422148e-01
-5.16341766e-03 1.04378778e-02 6.37998879e-01 -8.75234187e-01
1.41294450e-01 -3.37408006e-01 -6.59302711e-01 3.42055738e-01
-7.97396004e-01 1.66500673e-01 5.46118021e-01 3.72283250e-01
-2.63764173e-01 -6.18804812e-01 6.33505106e-01 -3.65023732e-01
-8.78564060e-01 -1.99102908e-01 -8.68020236e-01 1.31260306e-01
6.21949553e-01 -2.72226661e-01 -3.58605653e-01 -3.23716283e-01
-8.57223392e-01 1.27008781e-01 1.16546191e-01 5.60082734e-01
9.32231426e-01 -1.09758615e+00 -8.27223361e-01 3.68003339e-01
5.23019075e-01 -5.50372005e-02 -4.01586667e-03 4.39267427e-01
-1.98713601e-01 6.22301936e-01 -1.48435563e-01 -3.89933646e-01
-1.25343251e+00 1.67738959e-01 1.93724800e-02 -6.27056897e-01
-3.26953858e-01 1.36328995e+00 -2.43478730e-01 -8.66314054e-01
2.47336015e-01 -3.63627225e-02 -6.47056282e-01 -7.46894777e-02
5.27663827e-01 5.27291186e-02 5.90537012e-01 -4.41012233e-01
-5.20906329e-01 -1.83198348e-01 -5.66779137e-01 -3.32376182e-01
1.29001021e+00 9.46014225e-02 1.37048876e-02 2.92911053e-01
1.10480571e+00 -1.96308315e-01 -4.34537143e-01 -3.79650265e-01
6.49023533e-01 -2.98926025e-03 1.42052934e-01 -1.35075247e+00
-9.79163721e-02 8.56478274e-01 5.40190041e-01 3.36823702e-01
8.51241946e-01 4.62732017e-02 5.05504429e-01 6.16924644e-01
3.21709901e-01 -1.74607182e+00 2.86945015e-01 1.28739393e+00
1.03894317e+00 -1.65027046e+00 -7.48359486e-02 2.60908306e-01
-8.29012513e-01 1.28683186e+00 1.07119823e+00 -1.15554355e-01
7.35822618e-01 8.33290368e-02 -6.65960228e-03 -2.02622175e-01
-7.84568429e-01 1.38833299e-01 3.67294073e-01 7.57272959e-01
8.45719099e-01 2.48963565e-01 -5.30182540e-01 1.05964565e+00
-4.54417825e-01 -3.09172664e-02 3.36529672e-01 9.21744585e-01
-6.15911245e-01 -1.29711497e+00 -1.59349158e-01 8.57188821e-01
-5.69162011e-01 -3.38696778e-01 -7.96598494e-01 5.24872839e-01
3.71251911e-01 1.18724954e+00 1.36548832e-01 -7.26086676e-01
2.21631713e-02 6.17494583e-01 -7.85928741e-02 -1.11030829e+00
-1.04980898e+00 -1.83251575e-01 6.71131790e-01 -1.33864060e-01
-2.09221750e-01 -5.89717329e-01 -1.12647879e+00 -8.51356909e-02
-3.71096253e-01 3.30404997e-01 6.42583787e-01 1.12588978e+00
2.36895308e-01 3.25828135e-01 4.54048961e-01 -3.19577068e-01
-5.38882315e-01 -1.58645558e+00 -2.83111811e-01 2.23217547e-01
2.21359760e-01 -4.88116860e-01 -5.87789388e-03 -1.92413151e-01] | [10.86599349975586, 7.60838508605957] |
b913126a-bff9-44eb-985c-cda2e6dee8be | lagrange-motion-analysis-and-view-embeddings | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Chai_Lagrange_Motion_Analysis_and_View_Embeddings_for_Improved_Gait_Recognition_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Chai_Lagrange_Motion_Analysis_and_View_Embeddings_for_Improved_Gait_Recognition_CVPR_2022_paper.pdf | Lagrange Motion Analysis and View Embeddings for Improved Gait Recognition | Gait is considered the walking pattern of human body, which includes both shape and motion cues. However, the main-stream appearance-based methods for gait recognition rely on the shape of silhouette. It is unclear whether motion can be explicitly represented in the gait sequence modeling. In this paper, we analyzed human walking using the Lagrange's equation and come to the conclusion that second-order information in the temporal dimension is necessary for identification. We designed a second-order motion extraction module based on the conclusions drawn. Also, a light weight view-embedding module is designed by analyzing the problem that current methods to cross-view task do not take view itself into consideration explicitly. Experiments on CASIA-B and OU-MVLP datasets show the effectiveness of our method and some visualization for extracted motion are done to show the interpretability of our motion extraction module. | ['Yunhong Wang', 'Zilong Li', 'Shaoxiong Zhang', 'Annan Li', 'Tianrui Chai'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['gait-recognition'] | ['computer-vision'] | [-1.73286766e-01 -1.12044781e-01 -1.81243479e-01 -2.87764426e-02
9.78436545e-02 -2.15866357e-01 4.19949323e-01 -2.76411593e-01
-2.93643028e-01 5.55243671e-01 3.82084161e-01 3.89124639e-02
-3.27614322e-02 -7.72935390e-01 -1.62476242e-01 -7.39141524e-01
-4.59858567e-01 5.01293875e-02 4.50006276e-01 -2.25837231e-01
1.91379413e-01 3.63206238e-01 -1.58984971e+00 2.57456978e-03
3.78266603e-01 5.50013006e-01 5.33595048e-02 8.42874467e-01
2.47756079e-01 6.56279087e-01 -3.97263587e-01 -1.97995037e-01
2.05322742e-01 -4.99296635e-01 -6.81787848e-01 3.27420056e-01
2.31638789e-01 -5.60741544e-01 -3.88850302e-01 7.11006582e-01
6.68379247e-01 -1.18685504e-02 7.22521305e-01 -1.37372673e+00
-4.86248136e-01 5.88126145e-02 -7.76068926e-01 3.34527582e-01
5.96076608e-01 3.29752743e-01 8.42448890e-01 -1.01027024e+00
9.01903450e-01 1.33037341e+00 6.89105570e-01 4.75775003e-01
-8.73762965e-01 -1.42259210e-01 6.57774359e-02 6.43301070e-01
-1.14461863e+00 -8.03779215e-02 1.18838286e+00 -5.45828044e-01
7.61453927e-01 4.24182594e-01 1.15576220e+00 1.09067225e+00
3.54841620e-01 8.80439162e-01 9.12337363e-01 -5.10765076e-01
-1.46774482e-02 -1.81111470e-01 3.81440818e-01 1.09365702e+00
6.92246377e-01 1.12423152e-01 -4.95788306e-01 8.00592676e-02
7.65098333e-01 -2.53976732e-01 -3.22588205e-01 -5.74800849e-01
-1.27061284e+00 6.06718421e-01 -8.75566807e-03 4.03678030e-01
3.59440073e-02 2.28511676e-01 5.89773059e-01 -1.55797964e-02
3.84932995e-01 -1.22769125e-01 9.38781872e-02 -3.12922627e-01
-8.98494840e-01 6.59472495e-02 5.33126473e-01 6.58124089e-01
5.91502666e-01 3.82092983e-01 8.15310106e-02 4.51342016e-01
4.27445412e-01 6.41138077e-01 4.64387268e-01 -1.02624118e+00
3.27781558e-01 7.95448899e-01 1.12542650e-02 -1.55755305e+00
-5.86047232e-01 -8.83755907e-02 -9.04985964e-01 4.98670429e-01
5.88010430e-01 -1.77826863e-02 -7.35412121e-01 1.39197028e+00
4.06080753e-01 1.14996791e-01 -1.26509875e-01 1.20948207e+00
6.90369248e-01 3.04362208e-01 -1.57300591e-01 -1.09191015e-01
1.57888377e+00 -8.75445485e-01 -1.05515838e+00 -3.23504349e-03
7.90801644e-01 -5.93127131e-01 1.00219321e+00 2.57688820e-01
-8.04347873e-01 -8.09745848e-01 -1.42821288e+00 -2.55395789e-02
-5.07511795e-01 4.84187722e-01 4.54733223e-01 7.91821182e-01
-7.34883726e-01 8.48202169e-01 -9.94141340e-01 -6.17681742e-01
-1.66927069e-01 2.44698927e-01 -6.56824410e-01 3.05032879e-01
-1.13628209e+00 8.98886085e-01 3.05134147e-01 3.22184235e-01
-1.18490525e-01 -2.19939157e-01 -1.00360525e+00 -2.53137201e-01
5.66752441e-02 -1.14168131e+00 5.49982905e-01 -6.01004481e-01
-1.31561434e+00 8.27442944e-01 -3.70161027e-01 -3.16555142e-01
9.26911354e-01 -1.55335739e-01 -4.86236036e-01 6.61198139e-01
3.64560708e-02 3.18570942e-01 8.00812662e-01 -1.33485067e+00
-5.02366841e-01 -3.99067551e-01 -2.02265028e-02 8.86637121e-02
-2.91562587e-01 -4.34623808e-01 -4.35022652e-01 -8.89592230e-01
2.42797628e-01 -9.45724308e-01 8.26510228e-03 1.45915970e-01
-3.32923144e-01 8.93046409e-02 1.23664165e+00 -1.21848619e+00
1.75660026e+00 -2.13729119e+00 2.08090246e-01 2.91463345e-01
2.59696007e-01 1.39209121e-01 3.70590210e-01 5.44705212e-01
-1.54064149e-01 2.33764872e-01 -2.10330278e-01 -2.46920362e-01
3.38575542e-02 2.88402677e-01 3.18822712e-02 7.55902350e-01
1.67005271e-01 5.55755973e-01 -8.02524984e-01 -9.66296852e-01
4.91454959e-01 3.24770600e-01 -5.65365851e-01 -6.08029142e-02
4.38579172e-01 3.73422295e-01 -3.34741384e-01 6.87738597e-01
5.81468761e-01 -8.23482946e-02 1.68527320e-01 -7.03790009e-01
-5.80795445e-02 -2.35254139e-01 -1.41242921e+00 1.51299679e+00
-3.90954465e-02 6.52185619e-01 -2.32061148e-01 -8.43316078e-01
6.28405631e-01 3.75217915e-01 8.49520266e-01 -3.72428924e-01
6.01601638e-02 1.20007463e-01 1.71194330e-01 -9.58455741e-01
4.80137616e-01 1.78909123e-01 1.54196590e-01 1.49929643e-01
-9.18406323e-02 5.08462548e-01 5.50190330e-01 -3.37778078e-03
8.49332273e-01 6.35943592e-01 5.71493745e-01 -2.83400327e-01
7.41341233e-01 1.21193863e-01 5.54256320e-01 1.77410975e-01
-3.90036911e-01 8.77056062e-01 1.73419207e-01 -5.24909556e-01
-1.13376474e+00 -8.80939901e-01 1.51883557e-01 3.27191681e-01
1.71345696e-01 -7.05942869e-01 -7.51671433e-01 -5.66923022e-01
-3.83745022e-02 8.69869143e-02 -7.52736509e-01 -6.35987073e-02
-9.80099022e-01 -7.77922988e-01 4.93179291e-01 7.25552738e-01
5.98473012e-01 -7.83047020e-01 -1.15072751e+00 1.91694066e-01
-4.81782615e-01 -1.15703976e+00 -6.29864752e-01 -3.64597797e-01
-1.04036832e+00 -1.21981359e+00 -9.32952940e-01 -5.96950829e-01
6.92467153e-01 1.99497685e-01 6.45094812e-01 1.46602526e-01
-6.04276419e-01 6.06640339e-01 -3.40016425e-01 1.31251216e-01
-1.74130797e-01 -2.27217734e-01 1.63624987e-01 1.55287579e-01
2.64643908e-01 -7.49543369e-01 -9.54047024e-01 3.21201533e-01
-4.24147248e-01 2.66955763e-01 3.11443418e-01 7.40742207e-01
1.64019406e-01 -1.43844590e-01 1.43390864e-01 -4.01556969e-01
3.30250055e-01 -4.93161194e-02 -7.37595335e-02 2.03199938e-01
-4.05939549e-01 2.01386258e-01 2.35401660e-01 -4.86293226e-01
-8.09605658e-01 1.66723832e-01 -3.54600102e-02 -6.33269370e-01
-7.63650760e-02 4.15028006e-01 -3.35376978e-01 1.23384804e-01
1.72505572e-01 3.89640361e-01 2.08423421e-01 -4.94192809e-01
3.06719393e-01 2.53844231e-01 4.08262312e-01 -4.15875524e-01
8.78202438e-01 9.31158423e-01 3.65288824e-01 -1.16197848e+00
3.68215516e-02 -6.52008593e-01 -1.05183494e+00 -7.09375858e-01
1.17020202e+00 -6.68838561e-01 -9.51163113e-01 2.98852801e-01
-1.16817772e+00 2.33228400e-01 -1.67562053e-01 6.18328273e-01
-8.46432924e-01 1.15205228e+00 -6.80021286e-01 -1.16802990e+00
-2.85374969e-01 -9.48657751e-01 9.82159317e-01 -9.77291092e-02
-5.48240721e-01 -1.17905223e+00 1.51695251e-01 2.66018450e-01
-1.64403901e-01 5.30013919e-01 8.18366230e-01 -8.68935883e-02
-4.18733567e-01 5.24189556e-03 3.75307561e-03 -5.12097627e-02
3.10567021e-01 3.77021521e-01 -9.41588223e-01 -1.10649504e-01
-1.12088755e-01 3.97577316e-01 7.70182312e-01 4.79966044e-01
5.94642699e-01 -3.23483348e-01 -3.70654941e-01 5.23024440e-01
1.47137415e+00 1.70898676e-01 7.67055094e-01 5.20748198e-01
8.99697304e-01 9.27455306e-01 8.14776182e-01 5.80038428e-01
2.75532514e-01 9.66994703e-01 2.60290146e-01 -1.21979833e-01
-4.15241361e-01 -4.16632891e-01 6.50652409e-01 1.24322355e+00
-6.49684966e-01 1.77235063e-02 -7.84124613e-01 5.11856496e-01
-2.01887536e+00 -1.27112627e+00 -4.82637942e-01 2.04266143e+00
3.41289401e-01 2.11582705e-01 6.21158063e-01 7.11817384e-01
5.62315702e-01 7.06927255e-02 -9.38537866e-02 -1.63348511e-01
-1.32660300e-01 -3.68200600e-01 5.30622244e-01 4.88796026e-01
-1.09608412e+00 5.24027586e-01 6.05765724e+00 6.68935120e-01
-1.08368444e+00 -1.41014129e-01 6.06149137e-02 3.29262525e-01
-3.21957618e-02 1.75930634e-02 -6.86225176e-01 5.25765538e-01
4.61285591e-01 -1.72995999e-01 -2.85033822e-01 6.49062514e-01
5.38222551e-01 -1.68478325e-01 -8.89363408e-01 1.18104839e+00
1.38534874e-01 -1.02082348e+00 2.20469281e-01 3.50181073e-01
-7.05190897e-02 -8.62159669e-01 -1.90696403e-01 -2.47981101e-01
-2.05794126e-01 -6.17181957e-01 7.24703848e-01 5.54448962e-01
4.85942930e-01 -5.49856067e-01 6.37753904e-01 2.10905015e-01
-1.83823073e+00 8.29940960e-02 -1.43479347e-01 -1.41198799e-01
5.91491938e-01 4.22777772e-01 -7.32235909e-01 1.10275304e+00
5.70990682e-01 1.00928485e+00 -8.43813479e-01 1.08063853e+00
9.74389464e-02 6.01047277e-01 -3.15203667e-01 1.28018975e-01
-8.19807779e-03 -4.49526548e-01 7.29272187e-01 1.33789575e+00
5.60984135e-01 -2.65307315e-02 1.41138256e-01 7.18326986e-01
8.73092115e-01 1.22540012e-01 -1.07478094e+00 -3.64273898e-02
-1.45750031e-01 1.06848025e+00 -1.09655213e+00 -2.49165758e-01
-4.74692434e-01 1.08057499e+00 -2.30638415e-01 2.31330216e-01
-8.67788792e-01 -1.26280889e-01 5.64390182e-01 3.90534222e-01
1.63061276e-01 -6.46230161e-01 -4.02339906e-01 -1.36238575e+00
2.06975043e-01 -5.23519516e-01 4.83094484e-01 -6.76269352e-01
-1.06311488e+00 2.31017798e-01 2.20682472e-01 -1.89014125e+00
-3.79961401e-01 -8.67472708e-01 -5.94638050e-01 5.86915374e-01
-9.10821855e-01 -1.39297366e+00 -1.09432749e-01 4.54112291e-01
6.25750363e-01 1.66350931e-01 6.96371436e-01 5.87889850e-01
-6.12087846e-01 3.57987165e-01 -3.46237004e-01 4.32848364e-01
5.91978669e-01 -1.27672195e+00 3.28429788e-01 9.87586558e-01
9.05513912e-02 5.96112072e-01 9.89905536e-01 -8.43894005e-01
-1.50018239e+00 -5.69679976e-01 8.24584782e-01 -4.98898417e-01
4.59288836e-01 -1.03117540e-01 -7.89133370e-01 4.56239820e-01
1.75020263e-01 -1.81122363e-01 6.21128142e-01 -3.25617582e-01
4.71006967e-02 1.38875395e-01 -8.21907103e-01 8.29379201e-01
1.26761627e+00 -2.61000186e-01 -7.72059619e-01 -2.39339232e-01
2.77813584e-01 4.28094864e-02 -7.09465563e-01 5.26357234e-01
8.78073394e-01 -8.74670386e-01 1.35652614e+00 -5.08918524e-01
3.73022974e-01 -5.93254983e-01 -1.41279742e-01 -9.45644498e-01
-3.42206150e-01 -1.39423162e-01 -4.59144682e-01 1.03966916e+00
-6.91681355e-03 -3.43042195e-01 9.57322896e-01 1.58393875e-01
2.21702799e-01 -5.96204758e-01 -8.93140256e-01 -1.04254341e+00
-3.00317377e-01 -1.63872421e-01 -5.40582873e-02 1.07480919e+00
4.12329912e-01 2.89391339e-01 -9.08478677e-01 1.45899594e-01
9.07581091e-01 8.63353983e-02 9.41570401e-01 -1.16580093e+00
-3.04982424e-01 -2.00976864e-01 -8.67384791e-01 -7.69854486e-01
-1.97575673e-01 -4.48620409e-01 -4.38618273e-01 -1.63442624e+00
-4.71997149e-02 2.34520316e-01 7.74521083e-02 1.60795584e-01
-1.76435322e-01 2.25844309e-01 4.12285417e-01 1.68891922e-01
-1.12453021e-01 5.10568142e-01 1.44344723e+00 -2.11337343e-01
7.82542229e-02 -2.22983330e-01 1.77232221e-01 1.03378582e+00
6.87599242e-01 -2.17105910e-01 -5.26371777e-01 3.27196680e-02
4.56610695e-02 1.21203937e-01 4.62685496e-01 -1.18175769e+00
7.16230422e-02 -1.04930483e-01 4.03183997e-01 -9.49591279e-01
5.98640800e-01 -9.85526502e-01 4.28162664e-01 9.67449605e-01
2.41154939e-01 5.05949140e-01 1.69908591e-02 5.81954181e-01
-2.43405282e-01 -9.41906720e-02 4.79735792e-01 -2.07487702e-01
-1.03788996e+00 -3.47022787e-02 -6.74740851e-01 -1.43069074e-01
8.30364883e-01 -1.03248370e+00 4.16073157e-03 -4.58054125e-01
-1.15660238e+00 4.03910391e-02 4.38324332e-01 4.63488400e-01
8.24603796e-01 -1.58077192e+00 -4.61824626e-01 1.56655192e-01
1.31155610e-01 -6.79249942e-01 2.86189348e-01 1.04806209e+00
-8.34439516e-01 2.19744101e-01 -5.69889486e-01 -6.78176463e-01
-1.74479127e+00 6.66462541e-01 1.73394471e-01 -2.35666424e-01
-1.00168812e+00 4.07271553e-03 1.16323009e-01 3.18894424e-02
-1.99227054e-02 -4.40221995e-01 -6.19781196e-01 2.83224612e-01
3.52740020e-01 8.70520592e-01 -2.87087232e-01 -9.78970766e-01
-5.03239691e-01 1.15300333e+00 5.09572148e-01 -3.70335013e-01
9.61023986e-01 -4.91418630e-01 1.34474218e-01 8.54767025e-01
1.14698374e+00 -1.83915701e-02 -9.87349212e-01 2.94316202e-01
2.29013726e-01 -4.43221837e-01 -5.42808533e-01 -2.80854046e-01
-8.94943416e-01 1.27310944e+00 9.42775130e-01 5.57685606e-02
9.59857643e-01 -6.19313359e-01 7.35144794e-01 6.16677366e-02
5.46812952e-01 -1.17415500e+00 2.34751731e-01 1.57472804e-01
8.33996892e-01 -1.07631862e+00 1.14515938e-01 -8.61379504e-01
-5.30812263e-01 1.42388833e+00 6.67477846e-01 -8.87762457e-02
7.39930093e-01 2.71139592e-01 -9.96522699e-03 -1.36410352e-02
-4.39880669e-01 -3.40035111e-01 4.60067660e-01 6.49272382e-01
3.77467334e-01 -6.88657239e-02 -8.11300695e-01 4.49414164e-01
-9.58314091e-02 1.43053845e-01 5.18219233e-01 1.32321954e+00
-4.28252965e-01 -1.13656592e+00 -5.27695954e-01 -1.11888565e-01
-4.97643381e-01 2.82370746e-01 -2.76563764e-01 1.13005424e+00
2.64931917e-01 6.82883978e-01 -2.86390185e-01 -7.17415094e-01
3.71546239e-01 2.44935393e-01 6.44812942e-01 -2.40276694e-01
-1.60863847e-01 2.26210758e-01 3.28930944e-01 -6.21937394e-01
-6.88006759e-01 -5.38441122e-01 -1.19100583e+00 -3.92259151e-01
-1.23412706e-01 6.29755631e-02 2.66826332e-01 7.74089932e-01
1.39973491e-01 4.69611466e-01 1.17304794e-01 -9.98630762e-01
-2.03464746e-01 -7.13587165e-01 -6.97419107e-01 6.68599010e-01
4.20324862e-01 -8.69840503e-01 -5.15403092e-01 6.97531223e-01] | [14.227476119995117, 1.436474323272705] |
624cceb2-9ad6-4841-83ee-9d4f892d28cb | parametric-reshaping-of-portraits-in-videos | 2205.02538 | null | https://arxiv.org/abs/2205.02538v1 | https://arxiv.org/pdf/2205.02538v1.pdf | Parametric Reshaping of Portraits in Videos | Sharing short personalized videos to various social media networks has become quite popular in recent years. This raises the need for digital retouching of portraits in videos. However, applying portrait image editing directly on portrait video frames cannot generate smooth and stable video sequences. To this end, we present a robust and easy-to-use parametric method to reshape the portrait in a video to produce smooth retouched results. Given an input portrait video, our method consists of two main stages: stabilized face reconstruction, and continuous video reshaping. In the first stage, we start by estimating face rigid pose transformations across video frames. Then we jointly optimize multiple frames to reconstruct an accurate face identity, followed by recovering face expressions over the entire video. In the second stage, we first reshape the reconstructed 3D face using a parametric reshaping model reflecting the weight change of the face, and then utilize the reshaped 3D face to guide the warping of video frames. We develop a novel signed distance function based dense mapping method for the warping between face contours before and after reshaping, resulting in stable warped video frames with minimum distortions. In addition, we use the 3D structure of the face to correct the dense mapping to achieve temporal consistency. We generate the final result by minimizing the background distortion through optimizing a content-aware warping mesh. Extensive experiments show that our method is able to create visually pleasing results by adjusting a simple reshaping parameter, which facilitates portrait video editing for social media and visual effects. | ['Xiaogang Jin', 'Yong-Liang Yang', 'Wenxin Sun', 'Xiangjun Tang'] | 2022-05-05 | null | null | null | null | ['face-reconstruction'] | ['computer-vision'] | [ 3.72135133e-01 -1.87211540e-02 2.99588311e-02 -3.58731002e-01
-2.98078805e-01 -6.91306829e-01 3.46860141e-01 -6.46655679e-01
-9.62656364e-02 5.65801203e-01 3.00814897e-01 4.40590650e-01
9.42119360e-02 -7.02554464e-01 -8.38301361e-01 -7.64033437e-01
3.88311148e-01 -9.38796252e-02 2.03547582e-01 -1.78102627e-01
9.73719507e-02 7.58602560e-01 -1.38522208e+00 7.38922656e-02
8.12173605e-01 7.09032714e-01 1.15989402e-01 5.22588372e-01
-4.42408621e-02 4.25277650e-01 -2.04141945e-01 -6.69182360e-01
4.00848776e-01 -6.72036886e-01 -5.09775579e-01 8.20291698e-01
5.59398115e-01 -5.83360434e-01 -5.14234662e-01 1.18788648e+00
2.28433818e-01 3.06342453e-01 3.33705306e-01 -1.23682845e+00
-7.52934039e-01 2.85470456e-01 -1.14030647e+00 -1.58141389e-01
6.05401516e-01 -6.94570690e-02 5.34651935e-01 -1.00693893e+00
1.05387735e+00 1.44197392e+00 6.26097262e-01 6.65278137e-01
-1.32178211e+00 -8.55931044e-01 2.19088465e-01 2.24535670e-02
-1.74170864e+00 -7.03161418e-01 1.16941905e+00 -3.05918455e-01
1.17060862e-01 5.48514009e-01 9.87672687e-01 8.74883950e-01
1.26094058e-01 4.56006885e-01 7.39155710e-01 -2.26637006e-01
-1.61537886e-01 -1.07952721e-01 -5.86353421e-01 8.18927824e-01
-1.98625535e-01 -3.76601458e-01 -3.65068316e-01 -1.37330472e-01
1.18937969e+00 3.41559708e-01 -5.86896837e-01 -1.78660646e-01
-1.24643981e+00 4.99955297e-01 2.37593681e-01 3.57143253e-01
-3.66119564e-01 -4.74964902e-02 1.19608482e-02 2.05273122e-01
8.22705209e-01 -8.73923451e-02 1.84033692e-01 2.17415959e-01
-1.17415988e+00 2.45701283e-01 4.03775662e-01 8.02724004e-01
8.39449942e-01 1.07377265e-02 -5.17537929e-02 1.11419559e+00
2.71708548e-01 5.88547289e-01 1.86123878e-01 -1.26056290e+00
4.03911434e-02 2.72095114e-01 4.92946757e-03 -1.35224938e+00
1.63380310e-01 2.63124019e-01 -7.61664331e-01 9.32674948e-03
2.61610568e-01 -1.74752876e-01 -6.69091582e-01 1.59996557e+00
7.27751911e-01 6.64228439e-01 -3.22798699e-01 9.72395182e-01
5.01396716e-01 9.27573383e-01 -1.93394989e-01 -6.77489281e-01
1.08691764e+00 -6.26302779e-01 -1.13754702e+00 9.78126377e-02
-5.35317250e-02 -9.27833200e-01 7.69540787e-01 7.52688646e-02
-1.35120988e+00 -3.63114089e-01 -8.20922315e-01 -2.52899766e-01
4.13201690e-01 -2.71177113e-01 1.25990901e-03 3.94671738e-01
-1.04718626e+00 6.29219413e-01 -7.78127193e-01 -3.28526884e-01
5.28019965e-01 3.68596315e-01 -6.57085598e-01 -1.00226440e-01
-9.05474603e-01 4.26884204e-01 5.79372868e-02 4.57271114e-02
-6.27841234e-01 -8.90854001e-01 -8.95246148e-01 -2.07441404e-01
4.19841141e-01 -6.13132358e-01 8.27140510e-01 -1.70965838e+00
-1.97574985e+00 1.08431160e+00 -2.97135681e-01 1.22410655e-01
7.52053618e-01 -8.10138658e-02 -4.99414712e-01 1.45817921e-01
-1.25898778e-01 7.35537052e-01 1.41542220e+00 -1.40444851e+00
-3.09361309e-01 -2.68796921e-01 -2.41509050e-01 3.79790634e-01
-6.45455539e-01 1.58736572e-01 -1.04573905e+00 -1.06160593e+00
8.88777059e-03 -1.12576461e+00 2.19856918e-01 6.68499410e-01
-2.43574828e-01 3.26812148e-01 1.24787188e+00 -1.03356099e+00
1.19639409e+00 -2.45531011e+00 6.21518731e-01 3.13023835e-01
2.12787017e-01 1.21790402e-01 -2.90038019e-01 9.13372189e-02
-1.56872615e-01 2.17944104e-02 -4.89736259e-01 -5.01161098e-01
-5.39475381e-01 2.32504293e-01 -2.34190390e-01 6.92949951e-01
1.65033922e-01 7.93734133e-01 -9.01530206e-01 -6.26364708e-01
1.52029172e-01 9.60707366e-01 -7.36833692e-01 2.34242380e-01
-7.62324780e-02 6.05061352e-01 -3.97283345e-01 5.33422887e-01
1.01053286e+00 1.48179248e-01 3.14080745e-01 -4.44573849e-01
-2.53578454e-01 -5.75950503e-01 -1.06431913e+00 1.84769201e+00
-1.22596920e-01 6.78993523e-01 4.82047945e-01 -5.55993438e-01
9.57962513e-01 1.51159927e-01 9.03898835e-01 -2.91301548e-01
3.31563026e-01 -1.81312178e-04 -5.42868018e-01 -5.84780872e-01
4.25233811e-01 -4.18485254e-01 3.03197324e-01 4.19221103e-01
-1.85750723e-01 -3.43954057e-01 -2.23181143e-01 -8.34009871e-02
5.05136371e-01 2.38448769e-01 -1.72886532e-02 -1.84431523e-01
6.14296496e-01 -5.00763953e-01 6.44403517e-01 -3.68388176e-01
3.29065174e-02 9.77436960e-01 2.51832575e-01 -4.25943643e-01
-1.17752659e+00 -8.88580680e-01 -2.68566571e-02 7.34198391e-01
4.18990701e-01 -2.01326862e-01 -1.29331970e+00 -4.33291078e-01
-1.59126803e-01 3.69868726e-01 -5.78554928e-01 -7.58844912e-02
-8.16936910e-01 -3.86821479e-01 2.34052688e-01 6.16693422e-02
8.50965142e-01 -8.02549005e-01 7.06972703e-02 9.46042687e-02
-4.45199460e-01 -1.08106577e+00 -1.45874596e+00 -9.66201246e-01
-6.66815937e-01 -9.10685241e-01 -1.11663795e+00 -1.06222498e+00
9.61263776e-01 6.38187647e-01 3.69567782e-01 1.88407451e-01
-8.49786028e-02 4.40191865e-01 -1.14988029e-01 2.31812790e-01
-4.82249588e-01 -3.31973195e-01 2.31106251e-01 8.46068203e-01
-3.58264208e-01 -7.34475434e-01 -4.73046660e-01 6.95663333e-01
-1.26237702e+00 2.99018562e-01 -7.94927329e-02 4.36346561e-01
7.04206944e-01 6.07444867e-02 1.59999564e-01 -5.90728819e-01
3.32520813e-01 -3.36816669e-01 -4.81027573e-01 3.74245971e-01
-3.93699594e-02 -2.66564608e-01 5.16681612e-01 -8.50700259e-01
-1.20493627e+00 9.92317870e-02 -1.66580513e-01 -1.04680979e+00
4.07704532e-01 -5.14731593e-02 -3.43334258e-01 -4.74561393e-01
3.33322972e-01 1.96166098e-01 5.46883404e-01 -2.19891354e-01
3.86511236e-01 4.43795860e-01 5.90700984e-01 -2.55715251e-01
1.17478585e+00 8.25177014e-01 -1.98711067e-01 -1.10651219e+00
-4.27695841e-01 1.98210403e-01 -6.79024875e-01 -7.46789575e-01
1.05454111e+00 -7.43863821e-01 -6.64008319e-01 6.67388737e-01
-9.99643326e-01 -2.80703872e-01 -2.73501724e-01 3.11839551e-01
-3.86795074e-01 7.27297485e-01 -4.75400954e-01 -3.92904490e-01
-1.98708028e-01 -1.00771356e+00 9.62331295e-01 2.84447223e-01
-2.07336009e-01 -9.78667259e-01 1.59861639e-01 3.48750353e-01
2.31115878e-01 6.36901975e-01 5.51970780e-01 2.91689873e-01
-6.17146790e-01 4.87228446e-02 -7.46398717e-02 3.65015835e-01
5.02086222e-01 4.19003665e-01 -7.87820935e-01 -3.46369892e-01
8.01202059e-02 1.17666714e-01 3.60322177e-01 3.89151633e-01
1.11775064e+00 -5.76873481e-01 -2.93225080e-01 9.67803657e-01
9.81045544e-01 2.22895488e-01 8.33317339e-01 3.20257284e-02
9.69556630e-01 6.86951399e-01 5.01384318e-01 3.45213711e-01
2.10590735e-01 7.60128558e-01 2.07182780e-01 -7.99235329e-02
-2.28276432e-01 -5.28706014e-01 6.91006720e-01 8.79920661e-01
-4.18709546e-01 -1.76699907e-01 -3.89315128e-01 2.63031274e-01
-1.65978038e+00 -1.08062530e+00 3.56443137e-01 2.45620012e+00
8.45856667e-01 -4.29572225e-01 1.72724761e-02 -8.03780258e-02
1.21704853e+00 3.14234644e-01 -6.51233792e-01 5.68829179e-02
-7.49803931e-02 -1.35787040e-01 2.43388757e-01 9.05445337e-01
-6.97734535e-01 9.73611176e-01 5.63210058e+00 8.63627136e-01
-1.39050734e+00 5.19143902e-02 6.53794885e-01 -3.66925329e-01
-7.46807694e-01 -1.18986413e-01 -3.47325236e-01 5.09804070e-01
4.57473695e-01 -5.09774148e-01 8.51789773e-01 4.20551449e-01
4.46110189e-01 3.07636350e-01 -7.23497331e-01 1.26768160e+00
3.70390564e-01 -1.51919234e+00 3.10004413e-01 4.07166593e-02
8.87022913e-01 -6.25326276e-01 6.12098491e-04 -2.74413854e-01
-2.60251760e-01 -7.08989024e-01 9.14804697e-01 7.09534943e-01
1.28407800e+00 -8.96763206e-01 4.48982045e-02 -8.54960978e-02
-1.25369585e+00 1.80518851e-01 -1.29363164e-01 3.94714177e-01
3.96129370e-01 3.51427436e-01 -3.11796129e-01 3.26982856e-01
5.23921609e-01 1.07046616e+00 -8.77884850e-02 6.80655956e-01
2.19750665e-02 8.20897669e-02 -1.93988428e-01 4.84641939e-01
-2.50313938e-01 -6.99467242e-01 9.10105586e-01 8.02283943e-01
5.93783319e-01 5.35616934e-01 2.05556955e-02 7.00313210e-01
-5.03243923e-01 2.15946436e-01 -6.22208536e-01 4.68544438e-02
3.57312709e-01 1.26053774e+00 -7.23163247e-01 -1.19827956e-01
-3.01037967e-01 1.51695013e+00 5.31640127e-02 4.29765970e-01
-1.06843841e+00 4.96934680e-03 9.03086185e-01 6.37286663e-01
-2.02407520e-02 -1.20375752e-01 2.81328142e-01 -1.44780481e+00
7.38408118e-02 -7.95038760e-01 1.46786636e-02 -8.36614907e-01
-1.06353247e+00 7.98128366e-01 3.87990177e-02 -1.25641918e+00
7.22514689e-02 -4.77700122e-02 -6.88162506e-01 5.12127519e-01
-1.16433513e+00 -1.13261116e+00 -5.65932810e-01 9.53551292e-01
7.67604828e-01 2.13469967e-01 4.12414134e-01 4.08495486e-01
-7.40601242e-01 6.11472964e-01 -3.96168306e-02 -9.32010040e-02
1.03645635e+00 -4.05411243e-01 2.89770961e-01 7.72077084e-01
-1.37813538e-01 4.13685352e-01 5.90191782e-01 -8.61225188e-01
-1.51718628e+00 -1.26000130e+00 5.35832226e-01 -1.19280793e-01
4.09920782e-01 -2.82405585e-01 -1.11251521e+00 8.09003651e-01
2.48863578e-01 9.69726741e-02 5.73869884e-01 -6.35952950e-01
-1.27684891e-01 -2.32129693e-01 -1.49279308e+00 8.32240343e-01
1.28360927e+00 -5.26433885e-01 -2.30488077e-01 2.12570101e-01
6.39762998e-01 -4.87064868e-01 -9.49740350e-01 1.18188396e-01
8.50091159e-01 -6.84144378e-01 9.87404704e-01 -5.77858463e-02
3.72017413e-01 -5.00670612e-01 6.41759560e-02 -1.22094333e+00
-3.78053129e-01 -1.41329455e+00 2.35813797e-01 1.48963559e+00
-9.41299368e-04 -5.18154263e-01 6.48650765e-01 8.94863546e-01
5.89763336e-02 -4.55222934e-01 -7.45448172e-01 -3.10042173e-01
-2.72633076e-01 -1.00585535e-01 5.98273873e-01 1.28778040e+00
-1.73761845e-01 -1.41945928e-01 -7.62327194e-01 7.42701292e-02
7.01333225e-01 -9.10114944e-02 7.48451769e-01 -9.46136117e-01
9.96399298e-02 -1.90382645e-01 -2.48382926e-01 -8.64230096e-01
3.52816254e-01 -7.81906486e-01 -2.31846511e-01 -9.37105179e-01
2.68429190e-01 -1.87766373e-01 3.62966597e-01 2.58340210e-01
-1.14452831e-01 6.51866496e-01 3.33327919e-01 4.10534322e-01
-2.18623001e-02 7.25277841e-01 1.80312657e+00 3.55567038e-02
-4.23765212e-01 -2.58292317e-01 -6.24189079e-01 8.46039653e-01
4.35040504e-01 -3.40085894e-01 -4.86963451e-01 -5.14909267e-01
1.03147617e-02 2.04544768e-01 2.98560023e-01 -6.23653710e-01
-1.78209934e-02 -4.90515143e-01 4.65609431e-01 -5.37048802e-02
5.32498121e-01 -9.19281006e-01 7.98798382e-01 1.74071416e-01
-1.27924532e-01 -2.33289640e-04 2.63134569e-01 5.18946290e-01
-1.16736069e-01 2.77954508e-02 1.26280403e+00 2.50088930e-01
-2.93003350e-01 7.20726073e-01 -1.98668674e-01 -2.50673562e-01
1.51699340e+00 -4.67019022e-01 2.57640630e-01 -4.55721974e-01
-7.27920949e-01 7.02303872e-02 1.06489992e+00 6.84954882e-01
7.58448541e-01 -1.56329548e+00 -8.03432524e-01 6.02632344e-01
-4.50143695e-01 -8.51629153e-02 5.87705851e-01 6.26090348e-01
-7.80811191e-01 -4.50104147e-01 -5.15345633e-01 -4.66775179e-01
-1.66147411e+00 5.14650881e-01 3.53554279e-01 4.58168209e-01
-8.36153030e-01 6.91757917e-01 3.75054032e-01 7.56785274e-02
-4.00574431e-02 -1.37094244e-01 -8.76062736e-02 2.44920850e-01
8.10000598e-01 3.72977555e-01 -3.45955282e-01 -1.01323855e+00
-2.65113324e-01 1.38441634e+00 -8.69552568e-02 -2.28129491e-01
1.31435418e+00 -5.08743882e-01 -2.31200963e-01 2.01808035e-01
1.41789627e+00 3.99165720e-01 -1.72048461e+00 -1.18719973e-01
-7.48219788e-01 -1.09094930e+00 9.47792530e-02 4.69302908e-02
-1.55967677e+00 3.22688431e-01 3.45421523e-01 -4.52162027e-02
1.33677614e+00 -2.86785722e-01 1.20761037e+00 -2.41023764e-01
2.03699708e-01 -7.81568289e-01 1.18187331e-01 3.93533446e-02
1.22215307e+00 -7.61202216e-01 1.18786834e-01 -7.02764630e-01
-7.21659541e-01 1.21860754e+00 2.88547665e-01 -2.81824261e-01
6.68170631e-01 9.45795625e-02 -1.38294458e-01 7.14368969e-02
-1.86734676e-01 4.73361075e-01 3.44014108e-01 3.31227660e-01
6.46214038e-02 -2.22148851e-01 -2.13123128e-01 6.93261176e-02
-1.27902431e-02 1.95480481e-01 5.06845176e-01 6.65799022e-01
-2.17492327e-01 -9.94993329e-01 -7.29791343e-01 -4.71459217e-02
-3.44511420e-01 1.94871649e-01 -4.11648601e-01 5.67413449e-01
1.28889278e-01 6.89624786e-01 1.75137296e-01 -3.96111041e-01
3.34222674e-01 -5.85601293e-02 7.34878421e-01 -3.65336329e-01
-3.07407111e-01 3.67522538e-01 -3.09575498e-01 -6.25603080e-01
-6.04107440e-01 -7.40965843e-01 -1.14002168e+00 -9.22295451e-01
-1.39498159e-01 -1.01779856e-01 3.65720302e-01 6.64290428e-01
3.68966460e-01 1.88321218e-01 9.94308829e-01 -1.35213661e+00
2.18501627e-01 -4.70761806e-01 -6.34498715e-01 5.40690362e-01
4.34694827e-01 -5.63787758e-01 -3.64983886e-01 6.44190729e-01] | [12.781636238098145, -0.2935020327568054] |
3cfb4e7e-c838-47eb-969f-44586af45d23 | integrated-triaging-for-fast-reading | 1909.13128 | null | https://arxiv.org/abs/1909.13128v1 | https://arxiv.org/pdf/1909.13128v1.pdf | Integrated Triaging for Fast Reading Comprehension | Although according to several benchmarks automatic machine reading comprehension (MRC) systems have recently reached super-human performance, less attention has been paid to their computational efficiency. However, efficiency is of crucial importance for training and deployment in real world applications. This paper introduces Integrated Triaging, a framework that prunes almost all context in early layers of a network, leaving the remaining (deep) layers to scan only a tiny fraction of the full corpus. This pruning drastically increases the efficiency of MRC models and further prevents the later layers from overfitting to prevalent short paragraphs in the training set. Our framework is extremely flexible and naturally applicable to a wide variety of models. Our experiment on doc-SQuAD and TriviaQA tasks demonstrates its effectiveness in consistently improving both speed and quality of several diverse MRC models. | ['Boyi Li', 'Kilian Q. Weinberger', 'Ni Lao', 'John Blitzer', 'Lequn Wang', 'Felix Wu'] | 2019-09-28 | null | null | null | null | ['triviaqa'] | ['miscellaneous'] | [ 4.83955383e-01 3.55868042e-01 1.57027543e-02 -3.87428224e-01
-8.31093490e-01 -6.65291250e-01 4.67475742e-01 4.52607334e-01
-7.93785870e-01 7.20417380e-01 3.00379097e-01 -7.60299325e-01
-3.02283168e-02 -7.06488252e-01 -7.09920883e-01 -3.33211392e-01
1.01024866e-01 5.65097809e-01 4.87030804e-01 -4.11068976e-01
2.36501932e-01 1.91588104e-01 -1.54326618e+00 5.57392955e-01
1.15896451e+00 6.91382527e-01 4.39984798e-01 9.72414970e-01
-5.11143133e-02 9.65355754e-01 -7.69897163e-01 -6.61699593e-01
-1.38161212e-01 -2.77718991e-01 -1.18118632e+00 -5.51921010e-01
8.35188389e-01 -3.29422861e-01 -1.51134640e-01 5.81835628e-01
5.44368744e-01 2.25521877e-01 3.10559094e-01 -4.94674832e-01
-5.57658195e-01 9.53752160e-01 -3.86093616e-01 5.34014046e-01
4.08328056e-01 -9.20035243e-02 1.35052991e+00 -6.43722713e-01
3.21018726e-01 9.97928321e-01 5.85297465e-01 6.31727219e-01
-8.76981556e-01 -2.18650863e-01 3.92972976e-01 3.19663644e-01
-8.40548337e-01 -6.06231809e-01 2.97514260e-01 1.07119106e-01
1.49104416e+00 4.05526549e-01 5.31713963e-01 9.73101139e-01
1.87009141e-01 1.20243430e+00 8.23119521e-01 -7.24063754e-01
-9.41936683e-04 -2.37959936e-01 6.95313334e-01 7.43552029e-01
2.06062585e-01 -4.30263042e-01 -5.33319831e-01 7.73328096e-02
2.80085385e-01 -4.63277519e-01 -5.04008472e-01 -1.02176115e-01
-9.17344093e-01 6.17077351e-01 2.68660069e-01 2.87793577e-01
-1.84707254e-01 -9.87762399e-03 4.07114327e-01 5.34208655e-01
3.53547513e-01 7.43146539e-01 -9.07746673e-01 -4.13819253e-01
-9.50590134e-01 2.47852415e-01 9.83206153e-01 9.21582043e-01
3.24102581e-01 -2.13111430e-01 -1.15305409e-01 1.12239349e+00
-1.89297438e-01 3.22208643e-01 4.26205158e-01 -8.18864226e-01
8.51874352e-01 7.77525067e-01 -1.10938989e-01 -5.94136894e-01
-5.62162876e-01 -9.50305939e-01 -8.32976341e-01 -1.34355262e-01
6.92140520e-01 -6.16077557e-02 -8.67285848e-01 1.80677760e+00
1.68712921e-02 -3.68423641e-01 6.55781403e-02 5.09531856e-01
7.22295880e-01 6.86052203e-01 2.34605342e-01 -2.83831288e-03
1.31165195e+00 -1.15085959e+00 -3.13254565e-01 -7.85151780e-01
7.34609663e-01 -6.63034558e-01 1.51306033e+00 6.99904621e-01
-1.62723851e+00 -5.15581250e-01 -1.13904917e+00 -6.39533281e-01
-3.33121598e-01 -5.54456655e-03 6.20777249e-01 3.99350345e-01
-1.10450149e+00 5.45541644e-01 -6.31067395e-01 -2.37938210e-01
6.26851141e-01 4.39567000e-01 -1.76683336e-01 -5.59260845e-01
-1.10954058e+00 1.13857579e+00 4.89398956e-01 9.77817029e-02
-5.26915669e-01 -7.12764919e-01 -6.58329606e-01 5.35357654e-01
4.19697702e-01 -8.30727875e-01 1.86026990e+00 -8.30333114e-01
-1.36126387e+00 6.01698279e-01 -2.61462718e-01 -6.30725980e-01
4.15629625e-01 -7.95083702e-01 -1.58777997e-01 9.34167951e-02
-3.10645193e-01 7.12744594e-01 5.91234565e-01 -7.28310704e-01
-8.95684242e-01 -2.38277584e-01 3.24037939e-01 4.99198735e-01
-3.49087119e-01 5.01370206e-02 -4.64544147e-01 -3.56557190e-01
8.38942453e-03 -5.66735148e-01 -2.05845442e-02 -5.47495186e-01
-2.92137563e-01 -4.99232113e-01 2.14990094e-01 -8.49156022e-01
1.44110262e+00 -1.72600961e+00 3.31980586e-01 -1.26027495e-01
2.07870036e-01 6.60365403e-01 -3.89021635e-01 4.17806089e-01
6.12224825e-02 8.28233138e-02 -3.64821613e-01 -3.61273378e-01
-1.03704907e-01 2.76046693e-01 -3.37052196e-01 9.34580117e-02
2.08840102e-01 1.08877265e+00 -1.04265845e+00 -2.65735447e-01
-1.04605116e-01 2.39528745e-01 -5.46156287e-01 2.26639196e-01
-5.76173902e-01 -4.28197570e-02 -2.28004485e-01 4.25219834e-01
3.57077003e-01 -3.79016459e-01 5.94816497e-03 4.87754434e-01
7.10379630e-02 9.07051682e-01 -5.93337834e-01 1.68411863e+00
-6.63276017e-01 7.90653408e-01 -2.26895064e-01 -6.27068162e-01
4.44207758e-01 1.30352780e-01 -1.89768463e-01 -1.14404655e+00
1.27398118e-01 1.64701432e-01 3.66167724e-01 -4.72228587e-01
8.02341580e-01 2.66480118e-01 6.41618595e-02 5.81966877e-01
-6.18329868e-02 2.24878877e-01 4.25913155e-01 4.00891393e-01
1.28574276e+00 1.60307344e-02 3.08691800e-01 -1.64672315e-01
6.06484950e-01 8.76579061e-02 2.24756628e-01 9.85001564e-01
1.54071540e-01 6.01644993e-01 6.06344581e-01 -3.51322263e-01
-1.14919543e+00 -9.20167089e-01 7.43688494e-02 1.55127263e+00
-2.78495163e-01 -4.58173811e-01 -9.91254807e-01 -9.02467430e-01
-3.28922927e-01 9.58800316e-01 -5.80165744e-01 -1.64695293e-01
-1.11941743e+00 -7.49036133e-01 7.10035622e-01 6.69313133e-01
6.75198674e-01 -1.19457006e+00 -8.71998787e-01 3.79644454e-01
-3.51935178e-01 -1.03323388e+00 -7.39081055e-02 2.46522591e-01
-9.69818115e-01 -1.18650377e+00 -6.29638553e-01 -8.01692963e-01
5.66974163e-01 3.37673277e-01 1.76946378e+00 5.16710103e-01
1.75014697e-03 -5.60537465e-02 -5.54369271e-01 -5.50158620e-01
-5.26039422e-01 8.95655155e-01 -5.48523247e-01 -5.55084407e-01
6.77892983e-01 -2.65249670e-01 -5.31136394e-01 -1.71464123e-02
-6.87370777e-01 2.96753317e-01 7.46542096e-01 7.14797378e-01
1.70710459e-01 -1.08333685e-01 6.99406207e-01 -1.07889974e+00
9.76984799e-01 -3.19809526e-01 -3.80453467e-01 5.08980632e-01
-5.72483182e-01 -2.97890455e-02 9.28834617e-01 -3.56266797e-01
-9.97612655e-01 -3.64084542e-01 -3.90065670e-01 3.20309520e-01
-1.67859346e-01 6.07708752e-01 -4.06537317e-02 3.52351159e-01
7.46665120e-01 1.12970471e-01 -2.17446968e-01 -8.14048886e-01
4.21568394e-01 5.64579904e-01 6.99912488e-01 -4.99790013e-01
5.06038010e-01 -2.31256157e-01 -4.17747796e-01 -8.51847351e-01
-1.44754553e+00 -4.15368468e-01 -5.28373122e-01 1.82333246e-01
4.86724675e-01 -7.38284647e-01 -4.34360653e-01 4.71551597e-01
-1.09787524e+00 -4.75064903e-01 -2.88789243e-01 9.68063250e-02
-2.20849156e-01 3.99178207e-01 -7.34778881e-01 -4.73077238e-01
-7.19846010e-01 -9.60499108e-01 7.37905383e-01 3.56157422e-01
-4.20646727e-01 -9.75658894e-01 9.68411751e-03 5.61007798e-01
4.24290568e-01 -2.82918543e-01 1.41295779e+00 -7.52112389e-01
-4.64838684e-01 -1.35230377e-01 -1.85071260e-01 6.50456786e-01
-3.70507777e-01 -2.29782015e-01 -1.06460726e+00 -2.76477277e-01
-1.55261472e-01 -5.46985030e-01 1.21417320e+00 1.46854192e-01
1.20485246e+00 -1.71542853e-01 -6.99919909e-02 2.94621408e-01
1.16587424e+00 -6.08792678e-02 7.87246883e-01 5.15164793e-01
5.92712998e-01 6.47139311e-01 5.13864994e-01 -2.45816320e-01
4.27017510e-01 2.39011973e-01 3.50205809e-01 -3.90015058e-02
-1.55801117e-01 -3.18221241e-01 2.90850818e-01 1.11528909e+00
-1.44340947e-01 -7.74009109e-01 -1.13247120e+00 7.94464707e-01
-1.61906302e+00 -8.01851630e-01 -2.24391386e-01 2.00375843e+00
1.02524400e+00 4.03501451e-01 -1.76724285e-01 3.44020516e-01
2.36488268e-01 1.36714414e-01 -5.49356878e-01 -8.57230783e-01
-2.77946800e-01 6.89939499e-01 2.66017497e-01 5.35842419e-01
-8.66464317e-01 1.14635086e+00 6.75094748e+00 7.21838772e-01
-5.69064856e-01 2.40574982e-02 7.36189008e-01 -7.08931237e-02
-1.72563583e-01 -1.88617140e-01 -8.33962083e-01 9.67047177e-03
1.20162117e+00 1.49728432e-01 3.20360988e-01 7.25874901e-01
-9.05373245e-02 -4.25636142e-01 -1.25626290e+00 4.52919096e-01
4.90883999e-02 -1.21630704e+00 1.39541537e-01 -1.58458620e-01
6.36722803e-01 3.25688690e-01 -2.78470758e-02 5.56062579e-01
3.53946149e-01 -1.35004461e+00 5.76356411e-01 3.50865364e-01
7.14305520e-01 -9.48240459e-01 9.30350006e-01 7.33061969e-01
-5.16688228e-01 -2.71083891e-01 -6.68607891e-01 -2.16890141e-01
8.03514123e-02 3.79309475e-01 -8.59829545e-01 2.99523503e-01
5.99515915e-01 1.44486949e-01 -9.49598908e-01 1.13151145e+00
-6.32515073e-01 1.02791822e+00 -3.05205435e-01 -2.62426287e-01
4.65284884e-01 2.06480742e-01 2.87544936e-01 1.39904892e+00
-7.86740780e-02 1.91674501e-01 -3.55405241e-01 2.14761138e-01
-3.95785600e-01 2.14288980e-01 -1.97753236e-01 1.39500111e-01
4.04918015e-01 9.24316704e-01 -4.17988718e-01 -3.58660638e-01
-6.59036338e-01 7.37449348e-01 8.97359252e-01 2.06931978e-01
-6.27757132e-01 -5.79842567e-01 3.66292626e-01 1.26545548e-01
2.92430490e-01 -2.69365609e-01 -4.22489315e-01 -1.13846827e+00
2.46221974e-01 -1.33255982e+00 5.02507806e-01 -6.75081193e-01
-9.02942479e-01 7.59506643e-01 -1.43691048e-01 -5.16179383e-01
-2.11795002e-01 -6.52951241e-01 -5.91504931e-01 9.82607245e-01
-1.81028390e+00 -9.41898346e-01 -4.15039837e-01 3.49787086e-01
8.51219177e-01 -9.78116244e-02 8.35022151e-01 1.91018671e-01
-7.18951821e-01 8.24281812e-01 1.78293347e-01 2.78392574e-03
4.71878469e-01 -1.50403190e+00 9.16465580e-01 9.32639480e-01
1.67900175e-01 8.39857697e-01 6.91203952e-01 -3.19975138e-01
-1.11910093e+00 -6.95807934e-01 1.44447100e+00 -6.43084526e-01
4.81181920e-01 -4.14312452e-01 -1.29575837e+00 6.16004765e-01
5.36657691e-01 -7.34796882e-01 5.17636836e-01 4.89539057e-01
-4.40338701e-01 -1.89669766e-02 -6.48922622e-01 7.02367663e-01
9.83461499e-01 -4.01862144e-01 -1.08025718e+00 2.77793765e-01
6.97729945e-01 -4.93153870e-01 -6.43735647e-01 3.65413904e-01
5.88264644e-01 -9.71585453e-01 7.69133091e-01 -8.46816838e-01
7.03153014e-01 1.42195374e-01 1.59748435e-01 -1.22670233e+00
-2.69338369e-01 -6.11956120e-01 -4.90905523e-01 9.22325253e-01
7.38824129e-01 -3.22385073e-01 7.89020300e-01 4.60618883e-01
-3.88522953e-01 -1.09224057e+00 -6.79247499e-01 -3.09788883e-01
5.69641888e-01 -4.42471176e-01 6.58500314e-01 6.14324033e-01
-9.86969247e-02 7.67121613e-01 -5.11380751e-03 2.65529305e-02
3.08695853e-01 -3.44634764e-02 6.83847249e-01 -1.26877773e+00
-3.55880678e-01 -4.72331822e-01 2.59384811e-01 -1.63821828e+00
-7.36548081e-02 -7.95546651e-01 -1.26023199e-02 -1.78676438e+00
1.06975392e-01 -4.91361439e-01 -3.03687781e-01 5.98629057e-01
-4.60569143e-01 1.52752921e-01 -1.28486101e-02 -2.53820475e-02
-8.17184329e-01 2.49337345e-01 1.26284599e+00 1.50093630e-01
-4.23530936e-02 1.12391099e-01 -9.76938844e-01 6.28140688e-01
1.13875031e+00 -2.83197552e-01 -6.82659745e-01 -1.16007638e+00
6.20453656e-01 -1.40033275e-01 1.10844433e-01 -1.03954530e+00
2.21872568e-01 1.73500776e-01 3.94962370e-01 -6.29724801e-01
-4.10183296e-02 -3.92768443e-01 -5.56112945e-01 1.58045724e-01
-6.62160993e-01 3.20715338e-01 3.19426239e-01 2.83639222e-01
-6.93433359e-02 -5.48861384e-01 7.95500100e-01 -2.33903676e-01
-5.27415395e-01 6.58005709e-03 -4.19353396e-01 5.59918165e-01
4.68585432e-01 -2.38530919e-01 -6.45145595e-01 -1.93982914e-01
-4.94022667e-01 4.41444725e-01 3.72314274e-01 4.85673249e-01
3.97039860e-01 -4.36920166e-01 -7.46350646e-01 1.39722183e-01
-6.06643073e-02 5.51219583e-01 1.43682718e-01 5.02865136e-01
-7.81712234e-01 7.97363877e-01 5.60181551e-02 -4.30034488e-01
-1.36868727e+00 1.72672719e-01 3.50592107e-01 -5.87863505e-01
-8.48637402e-01 1.21276665e+00 -1.00971058e-01 -2.40289405e-01
5.47097147e-01 -4.18741643e-01 -5.09409904e-01 -8.68189111e-02
8.34473908e-01 4.39946175e-01 4.59672272e-01 -2.85376012e-01
2.95208991e-02 2.22366840e-01 -7.51726508e-01 2.14834958e-01
1.38406754e+00 -4.70545180e-02 -7.10877478e-02 1.79837972e-01
8.92902493e-01 -9.68935434e-03 -1.12885976e+00 -2.31524631e-01
3.38908106e-01 8.60577300e-02 1.35000736e-01 -1.34255731e+00
-6.33202672e-01 1.15923083e+00 -2.30197664e-02 1.96135983e-01
1.25923622e+00 -2.16001034e-01 1.04102898e+00 9.20034409e-01
2.42496416e-01 -1.08933997e+00 -1.52335003e-01 9.81985450e-01
6.69682145e-01 -9.47570026e-01 9.55134183e-02 -1.95465952e-01
-4.39276844e-01 1.04762900e+00 6.62345409e-01 4.59078979e-03
6.25423938e-02 1.16178468e-01 4.24264222e-02 -4.60469648e-02
-1.12575173e+00 4.79684547e-02 2.95807183e-01 3.37377250e-01
5.44470191e-01 -2.30520204e-01 -2.96918571e-01 4.57885355e-01
-6.20367229e-01 -2.28182673e-01 6.46181107e-01 1.06165731e+00
-8.79834950e-01 -1.15741622e+00 1.90495029e-01 5.69237709e-01
-8.48802388e-01 -4.88854706e-01 -3.68680716e-01 1.04482007e+00
-1.15331565e-03 1.02294433e+00 1.79169234e-02 1.07877649e-01
5.24151027e-01 3.37841839e-01 6.44199491e-01 -7.35023141e-01
-1.04627967e+00 -2.56852061e-01 5.11043370e-01 -4.76182550e-01
-9.65058804e-02 -5.74293613e-01 -1.12060857e+00 -3.41079533e-01
-4.40919310e-01 3.06198925e-01 5.20866275e-01 1.03515208e+00
2.48375133e-01 6.14587426e-01 7.88768195e-03 -1.78405195e-01
-8.27261925e-01 -1.17370057e+00 -1.09535758e-03 -3.93079370e-02
4.06428486e-01 -1.40839189e-01 -1.72336742e-01 -2.07311600e-01] | [11.174077033996582, 8.301637649536133] |
4a5abd84-53c7-499e-98ff-2f816ffb1e76 | learning-to-view-decision-transformers-for | 2301.09544 | null | https://arxiv.org/abs/2301.09544v1 | https://arxiv.org/pdf/2301.09544v1.pdf | Learning to View: Decision Transformers for Active Object Detection | Active perception describes a broad class of techniques that couple planning and perception systems to move the robot in a way to give the robot more information about the environment. In most robotic systems, perception is typically independent of motion planning. For example, traditional object detection is passive: it operates only on the images it receives. However, we have a chance to improve the results if we allow planning to consume detection signals and move the robot to collect views that maximize the quality of the results. In this paper, we use reinforcement learning (RL) methods to control the robot in order to obtain images that maximize the detection quality. Specifically, we propose using a Decision Transformer with online fine-tuning, which first optimizes the policy with a pre-collected expert dataset and then improves the learned policy by exploring better solutions in the environment. We evaluate the performance of proposed method on an interactive dataset collected from an indoor scenario simulator. Experimental results demonstrate that our method outperforms all baselines, including expert policy and pure offline RL methods. We also provide exhaustive analyses of the reward distribution and observation space. | ['Arnie Sen', 'Rajasimman Madhivanan', 'Ding Zhao', 'Xuewei Qi', 'Mohit Deshpande', 'Nathalie Majcherczyk', 'Wenhao Ding'] | 2023-01-23 | null | null | null | null | ['active-object-detection', 'motion-planning'] | ['computer-vision', 'robots'] | [ 3.31689626e-01 5.84724784e-01 -1.57356963e-01 -3.92168254e-01
-6.42538428e-01 -7.14381516e-01 4.04636353e-01 6.60983399e-02
-8.35204601e-01 7.01365292e-01 1.85087062e-02 -6.10529270e-04
-8.46352130e-02 -8.19500089e-01 -1.00914991e+00 -9.48301375e-01
-2.08164424e-01 4.08818096e-01 4.30768639e-01 -1.34234428e-01
5.62476218e-01 3.94455731e-01 -1.38293767e+00 -2.45158583e-01
6.77937210e-01 9.78878081e-01 1.21944869e+00 8.18348944e-01
4.38626319e-01 1.11595154e+00 -3.93771350e-01 4.34194595e-01
5.26016593e-01 -2.47137193e-02 -7.87031114e-01 4.28348571e-01
-3.42163444e-01 -6.43862188e-01 -3.66116017e-02 9.90436137e-01
5.16656816e-01 4.62219119e-01 4.11987305e-01 -1.12997842e+00
-2.34292135e-01 5.52667141e-01 -4.95985538e-01 2.65247002e-02
5.16359925e-01 5.40300131e-01 6.73877001e-01 -4.46113199e-01
6.03951395e-01 1.48655307e+00 -2.64927864e-01 4.56500113e-01
-8.42576921e-01 -2.86621213e-01 5.47357202e-01 2.66629219e-01
-8.02801788e-01 -4.65772063e-01 6.60886526e-01 -1.38503581e-01
9.49625790e-01 -2.55875945e-01 6.68440342e-01 9.40355659e-01
3.49150985e-01 9.38733816e-01 1.09664285e+00 -4.99498516e-01
6.37182295e-01 1.41377226e-01 -4.24377024e-01 8.06659579e-01
8.43645409e-02 4.45798039e-01 -5.12892187e-01 1.65771395e-01
7.44712353e-01 -2.59785559e-02 -2.46998966e-01 -9.17223334e-01
-1.51521409e+00 6.30261600e-01 6.49125159e-01 -7.00130612e-02
-9.89157438e-01 2.85021067e-01 -6.20229580e-02 2.03237727e-01
-2.13485822e-01 6.38323665e-01 -2.25482777e-01 -1.19146444e-02
-2.83809245e-01 2.54031032e-01 6.30653381e-01 9.37761068e-01
8.92752349e-01 -2.25223023e-02 -1.27875060e-01 6.04051054e-01
5.16873896e-01 6.88012064e-01 3.66434366e-01 -1.61508799e+00
4.65343893e-01 3.76910448e-01 7.81340003e-01 -8.46592069e-01
-3.79301429e-01 -5.33084236e-02 -6.72305971e-02 8.37688923e-01
-3.77108417e-02 -4.33921784e-01 -9.42143738e-01 1.48203027e+00
5.09532452e-01 -2.99899220e-01 5.01906097e-01 1.19000638e+00
1.73421696e-01 7.22471356e-01 -7.94374272e-02 -3.10750365e-01
7.84543455e-01 -1.16340685e+00 -5.69021165e-01 -7.87720025e-01
1.80830777e-01 -4.88871038e-01 8.60916197e-01 5.92467368e-01
-9.39336061e-01 -6.30208910e-01 -1.21065974e+00 3.69175851e-01
8.03684294e-02 2.80652970e-01 3.50773782e-01 1.04226738e-01
-1.04462552e+00 5.79278708e-01 -1.20215321e+00 -6.07932627e-01
1.58935487e-01 4.38172787e-01 -2.26474568e-01 -1.00240745e-01
-6.95537984e-01 1.03505039e+00 6.61479235e-01 1.02059662e-01
-1.74582875e+00 1.39360219e-01 -7.35668182e-01 -1.52588874e-01
1.05226231e+00 -4.91319805e-01 1.71223283e+00 -9.49703276e-01
-2.00332236e+00 3.49677414e-01 7.84584656e-02 -6.95720673e-01
3.24152380e-01 -4.88391042e-01 3.56446594e-01 2.79203475e-01
1.11857422e-01 1.11994898e+00 7.75184989e-01 -1.52929294e+00
-1.12444162e+00 -3.57399791e-01 5.48561871e-01 8.75359416e-01
1.10427074e-01 -5.79365134e-01 -3.72545213e-01 1.54297799e-01
2.00624079e-01 -1.09855568e+00 -6.96013689e-01 -7.79502317e-02
-2.11001888e-01 -1.33134827e-01 5.85135460e-01 -6.44806921e-02
2.38516599e-01 -1.99128115e+00 2.09800586e-01 -4.02610451e-02
-1.91775456e-01 -2.89579302e-01 -1.15842238e-01 4.74890679e-01
6.26142263e-01 -4.27003354e-01 8.38885680e-02 -3.00684035e-01
-1.28573611e-01 5.66124499e-01 -3.19386214e-01 4.63181734e-01
-2.14340794e-03 6.11534476e-01 -1.10209882e+00 -4.06032294e-01
4.09905642e-01 -1.60977259e-01 -4.96498376e-01 6.31007314e-01
-4.64758873e-01 7.75791883e-01 -8.83992910e-01 4.58888888e-01
3.03645670e-01 -5.99315725e-02 4.02099818e-01 2.31834620e-01
-4.58625108e-01 1.89917669e-01 -1.08159363e+00 1.84613168e+00
-4.52913791e-01 1.92780226e-01 3.86041105e-01 -1.10353136e+00
8.66642773e-01 7.38633797e-02 5.59331179e-01 -8.57483566e-01
2.21862331e-01 -4.86738607e-03 2.54968517e-02 -7.19051480e-01
5.31741381e-01 4.14436698e-01 -6.54756278e-02 1.20055526e-01
-9.43273082e-02 -1.87151656e-01 -3.14907432e-02 7.25842193e-02
1.34048307e+00 5.96430480e-01 4.11320716e-01 2.39109963e-01
1.96907580e-01 3.19292337e-01 5.97734153e-01 1.14942276e+00
-3.46436948e-01 -5.63066527e-02 1.39474096e-02 5.93496151e-02
-7.48774230e-01 -1.08751559e+00 5.29872179e-01 1.06298196e+00
8.49752486e-01 4.86487716e-01 -4.10470456e-01 -5.65284491e-01
-3.53911936e-01 8.79534841e-01 -4.89479631e-01 -1.31736964e-01
-5.50104856e-01 -4.32271749e-01 -2.78572023e-01 3.36466044e-01
7.51379788e-01 -1.59502280e+00 -1.72705567e+00 3.55090171e-01
-1.86523378e-01 -1.19240773e+00 2.79479772e-02 5.58326244e-01
-7.95867264e-01 -9.14226234e-01 -3.68019879e-01 -7.25739181e-01
7.91127324e-01 5.98223925e-01 6.61353290e-01 -2.93799073e-01
6.16549067e-02 7.91049540e-01 -6.92703784e-01 -5.08148849e-01
-3.20208102e-01 -1.74308106e-01 -3.98533307e-02 -2.92533606e-01
-2.85297841e-01 -3.30650270e-01 -7.02186584e-01 5.22637554e-02
-3.16449851e-01 4.34987098e-02 1.36622000e+00 5.38886130e-01
6.02119744e-01 2.15217307e-01 2.77705997e-01 -4.68172789e-01
5.24901927e-01 -3.38540822e-01 -9.12979364e-01 4.35733981e-02
-6.61631703e-01 3.11398268e-01 3.78822476e-01 -5.23391962e-01
-1.24229825e+00 7.32800066e-01 4.37978625e-01 -3.39695126e-01
-1.78950235e-01 2.03391328e-01 -1.06730729e-01 -1.37602864e-02
5.52419543e-01 3.41519386e-01 -1.22756183e-01 -6.70412555e-02
3.66418958e-01 4.09703791e-01 5.99217534e-01 -5.12777865e-01
4.59670395e-01 5.34245074e-01 -5.35412095e-02 -5.12375295e-01
-5.28955698e-01 -5.88087797e-01 -4.64098364e-01 -5.35380125e-01
9.42353427e-01 -7.72746921e-01 -1.10759664e+00 9.80514586e-02
-1.03500605e+00 -6.13885105e-01 -2.12782785e-01 7.17259824e-01
-1.08870494e+00 7.46721253e-02 -1.17011014e-02 -1.36741126e+00
-5.74077480e-03 -1.31695354e+00 1.00899279e+00 4.16524619e-01
2.39775300e-01 -4.68483388e-01 1.12160586e-01 4.93211560e-02
8.18909183e-02 2.23911032e-01 3.59278500e-01 -3.17006946e-01
-1.11230469e+00 2.57888377e-01 1.45718932e-01 -5.52519150e-02
8.44592676e-02 -5.32629073e-01 -7.98734009e-01 -3.42092305e-01
7.29937153e-03 -5.75596452e-01 7.00209379e-01 5.09761512e-01
7.99634516e-01 -5.36277890e-01 -7.02420890e-01 3.99701297e-02
1.45355141e+00 8.42368424e-01 4.16260898e-01 7.45563924e-01
9.70588773e-02 6.97613180e-01 1.51066887e+00 5.23218095e-01
4.80629236e-01 6.07857466e-01 1.17851639e+00 2.09516466e-01
3.85071576e-01 -2.47682393e-01 6.93671644e-01 1.22850150e-01
4.11638711e-03 -3.35142642e-01 -5.59444487e-01 5.20036817e-01
-2.21630669e+00 -8.44708025e-01 6.72295451e-01 2.15150046e+00
5.04749358e-01 3.70025218e-01 -1.44529231e-02 -1.26352429e-01
4.82933789e-01 -5.14483713e-02 -9.61615086e-01 -1.47751927e-01
3.75060707e-01 -3.90502423e-01 9.11521137e-01 6.82136655e-01
-9.21902001e-01 1.06649661e+00 5.95535994e+00 1.29631013e-01
-1.10446906e+00 -3.44529538e-03 8.31515640e-02 -6.35956228e-02
2.98024714e-01 1.60763204e-01 -6.09884560e-01 1.39748216e-01
5.80565929e-01 1.25562921e-01 6.64466381e-01 1.21871114e+00
5.28275013e-01 -9.32450891e-01 -1.25473166e+00 6.92151368e-01
-3.56467813e-01 -8.58760536e-01 -4.13673043e-01 1.29274085e-01
3.85394901e-01 1.59774408e-01 -7.20661506e-02 4.53641057e-01
8.15008402e-01 -6.39309049e-01 1.09440482e+00 7.51324534e-01
-1.38660237e-01 -6.92483842e-01 5.45055270e-01 1.15064251e+00
-7.31139302e-01 -6.50499225e-01 -5.42752326e-01 -1.45377770e-01
2.01336637e-01 5.68597466e-02 -1.65036011e+00 4.14276838e-01
6.65887892e-01 4.21918631e-01 -2.50663429e-01 1.09508431e+00
-3.82064760e-01 2.09131211e-01 -1.70108512e-01 -5.70548892e-01
2.86573946e-01 -1.22154027e-01 7.40485430e-01 6.28891706e-01
2.75825143e-01 2.23352119e-01 8.80308032e-01 6.96146786e-01
3.86558384e-01 -2.46992245e-01 -6.44773126e-01 1.01759642e-01
7.42642701e-01 1.15496504e+00 -7.26759851e-01 -1.39404073e-01
-4.46598530e-02 1.08562863e+00 5.63965917e-01 3.50492924e-01
-5.74576855e-01 5.13653830e-02 -1.70072056e-02 -2.73179829e-01
3.83036494e-01 -4.38869894e-01 2.83336788e-01 -5.06178796e-01
-1.57162458e-01 -5.16272128e-01 -3.34947780e-02 -1.04227650e+00
-5.40442407e-01 4.06604201e-01 1.83700584e-02 -1.14697313e+00
-6.11963212e-01 -4.53868300e-01 -1.01344526e-01 4.84296739e-01
-1.50076950e+00 -7.28694916e-01 -3.48058641e-01 2.82337129e-01
9.26266968e-01 -1.07514411e-01 5.16353190e-01 -4.13240641e-01
-6.09662607e-02 -1.35149047e-01 -8.83612558e-02 -1.62710682e-01
5.55980146e-01 -1.25595939e+00 -1.52076542e-01 7.26678073e-01
-6.83675110e-02 2.35581100e-01 9.82902527e-01 -6.21123254e-01
-1.61015069e+00 -7.80144155e-01 -2.65624881e-01 -1.12020858e-01
1.44947499e-01 -6.21348210e-02 -2.42295131e-01 6.61902845e-01
2.92354822e-01 -2.58006781e-01 -1.43844157e-01 -3.90693337e-01
5.02979338e-01 -2.12874621e-01 -1.28583419e+00 4.25716758e-01
7.17055202e-01 2.07548156e-01 -6.39890909e-01 4.65011150e-02
8.13892722e-01 -4.93335962e-01 -2.83044457e-01 4.68460232e-01
5.32581985e-01 -9.41052198e-01 8.19632351e-01 -1.54331490e-01
-5.18044084e-02 -4.63560313e-01 -3.92212033e-01 -1.54023826e+00
-3.49044353e-01 -3.83026391e-01 -5.63678914e-04 6.85432553e-01
4.04004544e-01 -4.69598711e-01 7.40675867e-01 2.35652611e-01
-2.37464905e-01 -5.66641152e-01 -6.70851171e-01 -6.30838931e-01
-5.79494417e-01 -2.38940626e-01 1.30535811e-01 2.58503765e-01
-4.80865221e-03 3.39307547e-01 -3.23902935e-01 7.78156698e-01
7.46925592e-01 2.40608603e-01 9.13946807e-01 -7.11397529e-01
-4.81139064e-01 1.60095021e-01 -1.86185345e-01 -1.49988556e+00
-5.96869253e-02 -3.07086200e-01 1.02646446e+00 -1.89126873e+00
1.84134483e-01 -4.87114906e-01 -1.36049569e-01 5.35786092e-01
1.69831470e-01 -5.77509224e-01 3.42469096e-01 3.89020979e-01
-1.02605927e+00 5.85568190e-01 1.68197107e+00 -1.41047642e-01
-4.29673523e-01 2.14317292e-01 -3.96331787e-01 7.55118132e-01
9.25175846e-01 -3.73914301e-01 -7.54125357e-01 -4.69737768e-01
1.02455013e-01 6.13312185e-01 3.66082907e-01 -1.21960843e+00
4.49942470e-01 -6.65148199e-01 4.79659051e-01 -7.53883779e-01
7.03669012e-01 -1.11868644e+00 -3.46731752e-01 7.63137817e-01
-5.88136017e-01 -1.91941127e-01 -1.96712434e-01 1.03386056e+00
2.07433879e-01 -4.12931114e-01 7.01472461e-01 -4.96131927e-01
-1.11902916e+00 1.62550062e-02 -7.32803226e-01 -3.64451170e-01
1.24437463e+00 -1.36777878e-01 -3.22378501e-02 -6.56678081e-01
-7.55552232e-01 6.80851817e-01 3.30143690e-01 4.77690130e-01
7.84989297e-01 -8.71229172e-01 -1.94392234e-01 -2.83279449e-01
-4.58435751e-02 3.18658412e-01 -1.92700207e-01 6.06053591e-01
-1.85199723e-01 4.49876577e-01 -3.86056840e-01 -6.86591744e-01
-9.06349659e-01 7.80469120e-01 3.28310579e-01 -1.00272022e-01
-4.38828826e-01 4.73133475e-01 3.15811276e-01 -4.70823616e-01
5.33588052e-01 -4.13115919e-01 -4.85363066e-01 -3.83746386e-01
2.37156034e-01 3.68320733e-01 -3.63532454e-01 -1.15695000e-01
-2.83432811e-01 1.63968295e-01 8.54396299e-02 -6.30258918e-01
1.17085099e+00 -6.65288031e-01 3.75838488e-01 5.77257991e-01
5.26611745e-01 -1.02422006e-01 -1.87055576e+00 -1.21963322e-01
2.76680738e-02 -4.18896407e-01 1.77290216e-01 -1.00908399e+00
-6.19407237e-01 4.82151866e-01 9.56663430e-01 9.73969102e-02
1.11860800e+00 9.01857093e-02 2.53024459e-01 9.90646899e-01
1.04000616e+00 -1.34140182e+00 7.34910071e-01 4.52372521e-01
8.64327908e-01 -1.47573924e+00 -4.55440432e-02 -1.14703640e-01
-9.72304642e-01 9.02866125e-01 9.43484247e-01 -3.76755267e-01
1.97542682e-01 1.43976748e-01 5.98444045e-02 -7.43301511e-02
-9.12812948e-01 -4.77299452e-01 -2.65206784e-01 7.21652687e-01
-3.32900256e-01 6.65023401e-02 1.72538832e-01 8.01363513e-02
-1.72587529e-01 -1.82975512e-02 6.66247189e-01 1.27179658e+00
-1.40740514e+00 -7.97811627e-01 -5.94794571e-01 4.42941040e-02
-5.73562086e-02 5.62437117e-01 -3.56512994e-01 6.42986953e-01
1.06305443e-01 1.28564060e+00 6.31522713e-03 -1.52737334e-01
2.94141710e-01 -5.07742047e-01 7.18866646e-01 -8.32788110e-01
1.57601655e-01 2.36699507e-01 3.70976031e-02 -9.13605571e-01
-6.60865724e-01 -6.46948099e-01 -1.83954704e+00 4.12588358e-01
-2.36311138e-01 3.42035368e-02 8.97984087e-01 9.74339604e-01
2.13107497e-01 6.68192804e-01 9.63534474e-01 -1.01105070e+00
-7.74814904e-01 -9.69176114e-01 -2.68057227e-01 -2.55896002e-01
6.04265749e-01 -9.69271481e-01 -1.60984606e-01 -3.11292987e-02] | [4.61674690246582, 0.827681303024292] |
f93876a9-2d5e-4a8f-9eec-96ca255ba5ae | ear-u-net-efficientnet-and-attention-based | 2110.01014 | null | https://arxiv.org/abs/2110.01014v1 | https://arxiv.org/pdf/2110.01014v1.pdf | EAR-U-Net: EfficientNet and attention-based residual U-Net for automatic liver segmentation in CT | Purpose: This paper proposes a new network framework called EAR-U-Net, which leverages EfficientNetB4, attention gate, and residual learning techniques to achieve automatic and accurate liver segmentation. Methods: The proposed method is based on the U-Net framework. First, we use EfficientNetB4 as the encoder to extract more feature information during the encoding stage. Then, an attention gate is introduced in the skip connection to eliminate irrelevant regions and highlight features of a specific segmentation task. Finally, to alleviate the problem of gradient vanishment, we replace the traditional convolution of the decoder with a residual block to improve the segmentation accuracy. Results: We verified the proposed method on the LiTS17 and SLiver07 datasets and compared it with classical networks such as FCN, U-Net, Attention U-Net, and Attention Res-U-Net. In the Sliver07 evaluation, the proposed method achieved the best segmentation performance on all five standard metrics. Meanwhile, in the LiTS17 assessment, the best performance is obtained except for a slight inferior on RVD. Moreover, we also participated in the MICCIA-LiTS17 challenge, and the Dice per case score was 0.952. Conclusion: The proposed method's qualitative and quantitative results demonstrated its applicability in liver segmentation and proved its good prospect in computer-assisted liver segmentation. | ['Haiying Wang', 'Lubiao Zhou', 'Peiqing Lv', 'Xiangyang Zhang', 'Jinke Wang'] | 2021-10-03 | null | null | null | null | ['liver-segmentation'] | ['medical'] | [-1.27585456e-01 3.99668887e-02 -2.98991889e-01 -1.25871375e-01
-6.76683486e-01 -1.89759731e-01 3.16243649e-01 1.09219290e-01
-5.33425868e-01 6.97917163e-01 4.30481851e-01 -3.00110430e-01
2.10090168e-02 -6.23231590e-01 -4.03586894e-01 -7.74803340e-01
-9.53761935e-02 -5.19318655e-02 2.66512662e-01 8.97394046e-02
-2.82655880e-02 4.35659885e-01 -5.46834350e-01 7.80074000e-02
1.10915411e+00 1.14052057e+00 2.84376264e-01 4.52854723e-01
2.47745719e-02 8.37489128e-01 -3.68170559e-01 -2.07420394e-01
2.50303745e-01 -6.93909168e-01 -9.55328047e-01 -1.13844775e-01
9.66439694e-02 -5.20585895e-01 -4.64812756e-01 1.08517551e+00
9.62108910e-01 5.27140172e-03 5.65441608e-01 -7.47657776e-01
-4.01160210e-01 8.24038267e-01 -6.02863193e-01 5.78465104e-01
3.58278118e-02 5.05143702e-01 6.91855311e-01 -7.46174753e-01
5.87748945e-01 7.50388801e-01 9.70968366e-01 5.44953167e-01
-8.62024248e-01 -6.36487842e-01 -3.00907493e-01 1.24184877e-01
-1.36875129e+00 -7.57381618e-02 4.28526610e-01 -3.28370661e-01
7.22863555e-01 2.68118113e-01 9.26114559e-01 7.05802321e-01
4.51662987e-01 1.14796305e+00 9.83173072e-01 -2.22890735e-01
-9.53432918e-02 -9.60091725e-02 2.26650596e-01 9.49421823e-01
5.39078601e-02 1.45753771e-01 -7.79305026e-03 3.46216202e-01
8.10578823e-01 5.48501778e-03 -6.94480538e-01 -9.05368626e-02
-1.46046889e+00 8.09508681e-01 1.10893559e+00 6.64368033e-01
-6.93443239e-01 1.55717552e-01 6.46369398e-01 3.64182554e-02
5.61796069e-01 3.69266063e-01 -2.75748223e-01 2.03772336e-01
-1.17313969e+00 -3.44095856e-01 6.18929684e-01 7.80261695e-01
1.67827621e-01 8.59842747e-02 -9.14404273e-01 5.30440092e-01
2.38441721e-01 7.30289891e-02 8.22195709e-01 -4.63244349e-01
-4.47073672e-03 5.56272924e-01 -4.05379742e-01 -6.43412292e-01
-6.94361925e-01 -9.98307347e-01 -1.37439072e+00 -3.31719294e-02
4.34384227e-01 -3.30749780e-01 -1.08062577e+00 1.47142959e+00
1.64100423e-01 4.85750735e-01 -6.22199960e-02 1.33324933e+00
1.39903188e+00 2.94099748e-01 2.86243200e-01 -5.52827232e-02
1.44160986e+00 -1.37492251e+00 -7.05585897e-01 3.43552113e-01
7.90754259e-01 -8.25857460e-01 7.77995169e-01 9.63883474e-02
-1.13400078e+00 -3.33072275e-01 -9.40311193e-01 1.14433445e-01
-1.60590619e-01 3.84515524e-01 6.96866214e-01 7.06308067e-01
-1.34783423e+00 5.75339258e-01 -7.69062221e-01 -3.69805694e-01
6.91794097e-01 4.44790572e-01 -1.52523294e-01 -1.60924569e-01
-1.18416989e+00 9.48064208e-01 5.62592864e-01 2.69153446e-01
-1.05981624e+00 -9.54975009e-01 -8.32033157e-01 3.19927037e-01
4.08542931e-01 -9.39732671e-01 1.13105762e+00 -7.17578530e-01
-1.41770232e+00 5.45602858e-01 2.40238309e-01 -8.30667496e-01
7.15826631e-01 2.12123573e-01 -6.88579008e-02 4.17438567e-01
-1.35856301e-01 1.01704085e+00 2.56996304e-01 -6.50113940e-01
-3.32082301e-01 -3.45158800e-02 -6.84465468e-02 2.15819985e-01
-1.20851122e-01 -1.21844508e-01 -6.43726170e-01 -9.20175791e-01
-1.63842123e-02 -5.37456512e-01 -3.30905944e-01 -3.07320170e-02
-3.93199444e-01 -1.72705501e-01 6.79765165e-01 -1.11937809e+00
1.17098725e+00 -2.10264373e+00 -8.57909843e-02 2.49921262e-01
3.71300161e-01 4.90058154e-01 -2.00492308e-01 -4.55858946e-01
-1.63497671e-01 1.96870312e-01 -3.66296172e-01 -5.70880473e-02
-3.19799930e-01 -3.17648314e-02 4.51285332e-01 7.31932938e-01
-5.60781397e-02 1.33224773e+00 -1.08377123e+00 -7.93325186e-01
4.69833046e-01 6.30205512e-01 -8.42949688e-01 1.83406398e-01
1.27906963e-01 6.37827456e-01 -1.85959473e-01 7.42045939e-01
5.30090332e-01 -2.21705377e-01 -1.98203586e-02 -4.41823900e-01
-1.58254221e-01 -4.41712476e-02 -6.89286232e-01 1.95940340e+00
-5.36174297e-01 4.68716592e-01 2.10785180e-01 -8.79326642e-01
6.80972219e-01 5.48265755e-01 8.59238863e-01 -8.89221609e-01
4.91792977e-01 3.24853688e-01 2.68346339e-01 -4.74995166e-01
1.48376569e-01 5.50881699e-02 3.06117952e-01 1.14619881e-01
3.25181782e-01 -1.29842609e-01 3.32787842e-01 2.16650695e-01
8.38516533e-01 9.87576470e-02 4.95515347e-01 -7.05691397e-01
8.69599164e-01 -1.73021615e-01 5.52672088e-01 7.45878577e-01
-5.69409072e-01 6.50149345e-01 5.78925848e-01 -5.00491083e-01
-4.90996659e-01 -6.65498674e-01 -2.98190296e-01 5.22517562e-01
2.29218856e-01 -3.94485205e-01 -9.18974996e-01 -1.21688390e+00
-3.59125316e-01 5.37174046e-01 -7.31217325e-01 8.01277757e-02
-4.71613705e-01 -1.11157095e+00 6.07549429e-01 5.20859182e-01
8.77513766e-01 -1.19191551e+00 -7.84225404e-01 3.55787396e-01
-3.41698438e-01 -7.95574367e-01 -9.39412534e-01 4.33041677e-02
-8.39002132e-01 -1.40619862e+00 -1.36453104e+00 -8.70880902e-01
7.65143037e-01 8.21928754e-02 1.12548268e+00 3.19899201e-01
-5.59885383e-01 1.34098887e-01 -3.18251878e-01 -9.24239382e-02
-1.67488188e-01 3.40771884e-01 -5.40927052e-01 -2.82302648e-01
-7.99691081e-02 -1.25762075e-01 -1.18159807e+00 3.08785945e-01
-7.19322205e-01 1.41233668e-01 9.60623920e-01 1.18614042e+00
5.29511929e-01 -1.40123814e-01 5.41071117e-01 -7.13741779e-01
4.03441370e-01 -4.37958479e-01 -5.62137246e-01 1.88534096e-01
-6.79299772e-01 -2.60088265e-01 5.75475037e-01 -2.46674977e-02
-7.63969779e-01 -3.07795480e-02 -4.85494822e-01 -3.67387563e-01
2.21364215e-01 4.79843140e-01 9.43461433e-02 -3.79760683e-01
3.14500540e-01 3.14170688e-01 1.67353317e-01 -3.09696436e-01
6.98375180e-02 3.48067492e-01 4.72473472e-01 -6.08406775e-02
1.50739416e-01 1.25337839e-01 -9.12312642e-02 -6.90919101e-01
-4.85078692e-01 -4.58054543e-01 -3.80780250e-01 -2.77472228e-01
1.10925591e+00 -7.65446663e-01 -7.07059562e-01 5.72993636e-01
-1.07683504e+00 -5.21018386e-01 -5.03786266e-01 6.92588568e-01
-2.39633739e-01 3.65729749e-01 -1.02613032e+00 -1.87500864e-01
-1.01183343e+00 -1.87973452e+00 6.62149727e-01 5.87827504e-01
-2.25267112e-02 -1.13256323e+00 -3.97249043e-01 4.95103747e-03
9.40232396e-01 1.91034496e-01 7.32562184e-01 -7.74188221e-01
-5.23573220e-01 1.45337731e-02 -7.75294006e-01 4.13271695e-01
1.44566866e-02 -3.80270511e-01 -5.29874861e-01 -5.05740881e-01
-1.79932147e-01 -8.04618523e-02 1.03585625e+00 7.98777938e-01
1.58872783e+00 -2.42559433e-01 -2.03014966e-02 1.12024188e+00
1.50698113e+00 2.04914242e-01 6.77689254e-01 8.35999250e-02
4.39029098e-01 -2.14685430e-03 5.57614900e-02 1.74396858e-01
2.03785211e-01 4.14011061e-01 6.64086342e-01 -7.29481399e-01
-8.14854443e-01 7.80520067e-02 1.05339130e-02 8.44714940e-01
2.55363081e-02 -6.21767417e-02 -9.33956623e-01 6.57533824e-01
-1.42105162e+00 -5.13685763e-01 -4.60947827e-02 1.92667925e+00
6.32754683e-01 -1.05258942e-01 -1.09817190e-02 -2.03537017e-01
5.96726775e-01 7.73464516e-02 -2.61202842e-01 -6.98536411e-02
-1.43325822e-02 3.47704440e-01 6.42558396e-01 4.68338192e-01
-1.20231354e+00 7.29998231e-01 6.17824650e+00 9.74255800e-01
-1.37177575e+00 4.58788604e-01 1.12883925e+00 9.75756422e-02
-3.33415493e-02 -4.25087363e-01 -3.65712970e-01 5.45596540e-01
4.74805683e-01 -2.79258862e-02 1.44445300e-01 4.74368781e-01
1.23890989e-01 -2.91773498e-01 -5.76805413e-01 8.71212721e-01
1.70239776e-01 -1.57381499e+00 -8.60861018e-02 -5.44540770e-02
7.54830778e-01 2.10862786e-01 -8.44137147e-02 4.50214893e-01
7.29251206e-02 -1.09821582e+00 1.35017753e-01 5.32065094e-01
1.15648854e+00 -7.45003104e-01 1.43285489e+00 3.81431845e-03
-9.95458126e-01 2.62970954e-01 -1.77622676e-01 4.11835283e-01
1.86698511e-02 7.50612438e-01 -1.22490072e+00 7.99675405e-01
4.45699751e-01 7.91557014e-01 -5.59555650e-01 1.60915101e+00
-3.08467746e-01 7.15562165e-01 -1.76056102e-01 1.29257247e-01
6.32358253e-01 -1.65449083e-01 6.15488768e-01 1.53928840e+00
4.05609787e-01 1.26783371e-01 1.54909477e-01 6.58330917e-01
-4.52629805e-01 4.54059422e-01 -1.56802252e-01 4.22075152e-01
-1.56217515e-01 1.67437720e+00 -1.17635524e+00 -5.86074412e-01
-4.35638726e-01 8.42972040e-01 -3.03571939e-01 1.93548203e-01
-9.59255278e-01 -3.24643672e-01 -7.78336525e-02 5.02577275e-02
2.00268045e-01 3.84021878e-01 -3.05253416e-01 -1.11905313e+00
-5.37403345e-01 -6.82723165e-01 5.70928931e-01 -5.91422558e-01
-7.78269172e-01 7.98827052e-01 -3.65429282e-01 -1.04467165e+00
8.29383731e-02 -2.96539187e-01 -8.27811241e-01 9.58570421e-01
-1.87955236e+00 -1.12886691e+00 -4.74754691e-01 5.52763343e-01
5.40552735e-01 -2.12953672e-01 6.07058465e-01 5.20238578e-01
-5.69574714e-01 7.21233368e-01 -4.34412733e-02 5.17322779e-01
5.85433900e-01 -1.36292768e+00 -9.13533419e-02 9.65187967e-01
-3.70964408e-01 4.26662594e-01 2.47855887e-01 -5.90544999e-01
-1.12719524e+00 -1.23159611e+00 6.17598593e-01 3.94140214e-01
3.64785641e-01 1.65281355e-01 -5.71621358e-01 4.69784111e-01
6.85490072e-01 5.68967640e-01 4.71364468e-01 -4.50156510e-01
2.10645840e-01 1.93405226e-02 -1.54891443e+00 3.11475843e-01
5.25895000e-01 6.05334044e-02 -4.15429741e-01 3.05205822e-01
6.94219589e-01 -8.46050680e-01 -1.12269604e+00 5.17935097e-01
5.41692317e-01 -8.80170584e-01 9.33354616e-01 -2.72313744e-01
5.84570527e-01 -1.84140399e-01 1.56124026e-01 -1.56276584e+00
-3.79677176e-01 -5.43585777e-01 5.45935482e-02 7.67913640e-01
3.14919353e-01 -5.28048337e-01 6.81410372e-01 1.05647771e-02
-5.61753035e-01 -1.19171786e+00 -9.53361809e-01 -3.27944010e-01
1.22403637e-01 -1.01354718e-02 5.58480084e-01 9.06263888e-01
-2.84752816e-01 4.17994186e-02 -1.46405697e-01 -2.55799741e-01
6.64765418e-01 -1.47528514e-01 2.62030423e-01 -7.01470613e-01
-1.20864868e-01 -7.83364773e-01 -2.69494414e-01 -9.92229640e-01
-5.52503951e-02 -1.30835366e+00 -8.51063803e-02 -1.65535069e+00
4.03469026e-01 -3.55979443e-01 -6.63062751e-01 4.46871102e-01
-2.77785748e-01 6.27102375e-01 4.84106570e-01 -5.22462726e-02
-6.12789989e-01 3.72085840e-01 1.81671226e+00 -3.84601235e-01
-9.97279882e-02 3.33589055e-02 -6.69982791e-01 6.77132130e-01
7.50521660e-01 -3.95229533e-02 -4.09739017e-02 -2.78334677e-01
-4.15094107e-01 3.36562991e-01 4.15773928e-01 -1.05401051e+00
4.46832687e-01 4.23084617e-01 6.99866474e-01 -7.26534903e-01
-3.66608500e-01 -9.17498887e-01 -2.48517707e-01 1.10526454e+00
-1.25580803e-01 -1.41208202e-01 1.89722374e-01 -7.76333883e-02
-2.85893232e-01 -6.01501316e-02 1.07211041e+00 -1.98004678e-01
-6.53625906e-01 5.61715841e-01 -2.25167021e-01 4.18093381e-03
1.08181357e+00 5.67191876e-02 -6.02346249e-02 -9.05391797e-02
-9.11952138e-01 5.66166997e-01 2.31758989e-02 -8.02166164e-02
6.72351837e-01 -1.15938187e+00 -9.77178395e-01 3.88071507e-01
-2.95022070e-01 2.00188622e-01 4.76842582e-01 1.70010734e+00
-9.51533675e-01 5.82619190e-01 -1.15504131e-01 -7.20831871e-01
-9.99342918e-01 3.41553777e-01 8.65364015e-01 -7.47564673e-01
-1.02649748e+00 8.46451998e-01 3.50038499e-01 -3.05988938e-01
2.42416546e-01 -5.73159516e-01 -2.50761777e-01 1.04298860e-01
5.60194314e-01 2.99241126e-01 2.31888980e-01 -5.22103965e-01
-3.48017573e-01 1.72393471e-01 -7.04287142e-02 3.39198440e-01
1.30888009e+00 -5.43536106e-03 -2.75429040e-01 -3.59005243e-01
1.16694999e+00 -2.05677524e-01 -1.13451421e+00 -1.78896740e-01
-1.89102471e-01 -1.62611797e-01 6.13631904e-01 -1.33242047e+00
-1.87044477e+00 7.95303404e-01 8.22342992e-01 -2.12038845e-01
1.39366806e+00 -4.00712073e-01 1.04170012e+00 -3.60985368e-01
-4.67829704e-02 -5.54631889e-01 -2.49295712e-01 5.92806041e-01
7.39051819e-01 -1.15421379e+00 -2.99791582e-02 -2.64356971e-01
-4.30250376e-01 1.14235318e+00 6.66145325e-01 -6.20403774e-02
5.93327761e-01 3.91371816e-01 1.71059147e-02 -7.88272545e-02
-1.88948169e-01 -1.33884683e-01 3.26646030e-01 1.99645996e-01
6.67044461e-01 8.12026113e-02 -6.84418380e-01 6.52891219e-01
1.21742986e-01 3.59410554e-01 4.92152333e-01 3.61018986e-01
-1.06909908e-01 -5.86982965e-01 -1.55046687e-01 6.16027594e-01
-8.32600296e-01 -3.59717637e-01 1.95200801e-01 1.04982150e+00
1.46343887e-01 3.51701051e-01 -1.34418160e-01 -1.69881545e-02
2.47598991e-01 -6.85584471e-02 3.69735807e-01 -1.53367594e-01
-1.24067175e+00 3.94903690e-01 -3.27981055e-01 -7.23199129e-01
-3.39934886e-01 -3.63038540e-01 -1.11224794e+00 -2.93083280e-01
-2.49629185e-01 2.67324477e-01 6.08626842e-01 7.83294678e-01
5.73068149e-02 1.19827282e+00 5.06082535e-01 -6.56406343e-01
-4.73784953e-01 -1.05335712e+00 -4.30157572e-01 3.29889715e-01
2.65989393e-01 -3.62704128e-01 -2.55031914e-01 -3.60502303e-01] | [14.558320045471191, -2.6657490730285645] |
59671223-f87c-4150-9997-d15f6b594053 | stay-on-topic-with-classifier-free-guidance | 2306.17806 | null | https://arxiv.org/abs/2306.17806v1 | https://arxiv.org/pdf/2306.17806v1.pdf | Stay on topic with Classifier-Free Guidance | Classifier-Free Guidance (CFG) has recently emerged in text-to-image generation as a lightweight technique to encourage prompt-adherence in generations. In this work, we demonstrate that CFG can be used broadly as an inference-time technique in pure language modeling. We show that CFG (1) improves the performance of Pythia, GPT-2 and LLaMA-family models across an array of tasks: Q\&A, reasoning, code generation, and machine translation, achieving SOTA on LAMBADA with LLaMA-7B over PaLM-540B; (2) brings improvements equivalent to a model with twice the parameter-count; (3) can stack alongside other inference-time methods like Chain-of-Thought and Self-Consistency, yielding further improvements in difficult tasks; (4) can be used to increase the faithfulness and coherence of assistants in challenging form-driven and content-driven prompts: in a human evaluation we show a 75\% preference for GPT4All using CFG over baseline. | ['Stella Biderman', 'Pawan Sasanka Ammanamanchi', 'Elad Levi', 'Alexander Spangher', 'Honglu Fan', 'Guillaume Sanchez'] | 2023-06-30 | null | null | null | null | ['code-generation', 'image-generation', 'zero-shot-learning', 'text-generation', 'machine-translation', 'lambada', 'common-sense-reasoning'] | ['computer-code', 'computer-vision', 'methodology', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'reasoning'] | [ 2.84770995e-01 6.48387730e-01 -1.48714662e-01 -3.55741918e-01
-1.11218262e+00 -4.84875113e-01 1.01329923e+00 7.17419386e-02
-1.66442230e-01 8.70971799e-01 5.33163548e-01 -6.62388623e-01
1.46205276e-01 -3.48864377e-01 -8.27909946e-01 -1.02005228e-01
1.72172382e-01 7.41046369e-01 -2.04994544e-01 -3.38614285e-01
3.14593166e-01 -1.60349414e-01 -1.26865053e+00 6.39441669e-01
1.04120338e+00 4.98507172e-01 1.56708121e-01 1.00917435e+00
4.62925620e-02 1.30657351e+00 -5.65682948e-01 -7.97146022e-01
6.76352084e-02 -4.29332435e-01 -1.19713557e+00 -5.34202121e-02
1.02399385e+00 -4.87415284e-01 1.99514907e-02 5.86189926e-01
8.13713074e-01 -1.11008808e-01 6.29191160e-01 -1.05553377e+00
-8.89852524e-01 1.11903679e+00 -6.84050500e-01 -7.25092217e-02
7.89667845e-01 5.79279006e-01 8.95171106e-01 -5.06305456e-01
7.32054174e-01 1.72829175e+00 7.67269194e-01 9.52329576e-01
-1.41797113e+00 -6.62873149e-01 7.65863294e-03 -1.58380091e-01
-9.43991661e-01 -6.77392483e-01 1.14670612e-01 -2.81296998e-01
1.43671799e+00 3.77487689e-01 6.06542051e-01 1.57522154e+00
6.06721006e-02 8.63997817e-01 1.30539680e+00 -9.00857747e-01
1.03341423e-01 2.76761919e-01 -8.95294324e-02 9.88235950e-01
-1.01963684e-01 -7.98270945e-03 -1.15749609e+00 -5.72804153e-01
4.64918941e-01 -7.02999294e-01 -2.99017783e-02 5.16486824e-01
-1.42107701e+00 8.81576598e-01 7.75467604e-02 -9.63662043e-02
-2.30284795e-01 7.66262591e-01 1.60024717e-01 1.19769014e-01
6.54276311e-01 9.72302735e-01 -1.70604065e-01 -7.96621144e-01
-9.56199944e-01 1.11243200e+00 1.17929101e+00 1.57211566e+00
4.80605513e-01 -8.20057690e-02 -7.87844598e-01 9.74170744e-01
2.82559305e-01 8.65246356e-01 2.85846025e-01 -1.54933965e+00
5.11421442e-01 2.54237920e-01 4.05802578e-01 -5.73054194e-01
-4.44916040e-02 -2.31225386e-01 -5.01085818e-01 -6.99575469e-02
4.54032838e-01 -3.35591197e-01 -8.83522272e-01 1.85169137e+00
4.25056033e-02 -4.26971793e-01 -2.19340429e-01 4.06835258e-01
5.59328854e-01 5.94404399e-01 3.14310670e-01 -1.49042144e-01
1.44614136e+00 -1.10776412e+00 -6.16204679e-01 -6.29207790e-01
7.67904997e-01 -1.08208191e+00 1.42986214e+00 6.36775792e-01
-1.52239251e+00 -5.41063488e-01 -7.06943810e-01 -2.74039507e-01
7.81118274e-02 4.75957207e-02 1.03952742e+00 7.06496239e-01
-1.70592701e+00 3.97680938e-01 -5.02654612e-01 -7.96607018e-01
3.32558632e-01 3.78476270e-02 -1.61282331e-01 -2.84466743e-01
-8.20395470e-01 9.08894360e-01 9.91494581e-02 -4.25420463e-01
-8.96450877e-01 -7.91913807e-01 -3.81494612e-01 -3.16081971e-01
3.61086398e-01 -1.28071177e+00 1.85557580e+00 -5.18966675e-01
-1.52035463e+00 8.98612857e-01 -2.21582964e-01 -4.59722221e-01
7.37770140e-01 -6.95924938e-01 1.58848435e-01 -1.34581774e-01
4.73934948e-01 1.44873536e+00 8.43461931e-01 -8.85464668e-01
-6.69660628e-01 -2.28667585e-03 1.09147884e-01 3.39611471e-01
-3.86486173e-01 2.00581953e-01 -3.19010109e-01 -6.54275000e-01
-4.27201599e-01 -1.08764672e+00 -2.56738275e-01 7.54689574e-02
-6.57185912e-01 -5.81301689e-01 3.41786265e-01 -8.46096575e-01
1.16111672e+00 -1.82073367e+00 3.71726081e-02 -2.11154401e-01
1.69244021e-01 7.38483965e-02 -2.44459540e-01 6.97448373e-01
3.40569705e-01 4.51231211e-01 -4.97605838e-02 -4.40717608e-01
2.61563510e-01 -2.33046035e-03 -4.54812258e-01 -1.63721889e-01
3.02811056e-01 1.08017266e+00 -9.55662012e-01 -7.85682917e-01
-2.86964923e-01 1.86089188e-01 -9.47272420e-01 3.95755798e-01
-8.29705715e-01 1.30055487e-01 -2.07294717e-01 4.82105255e-01
1.24884151e-01 -4.86991823e-01 -1.71176102e-02 4.63129848e-01
-1.23302788e-02 3.91801775e-01 -6.05434716e-01 1.95380974e+00
-5.67354918e-01 4.77192402e-01 -1.49724975e-01 -1.63452588e-02
7.93640435e-01 3.19817811e-01 -2.89050698e-01 -6.04783893e-01
-3.43739808e-01 3.61184388e-01 -8.69342610e-02 -6.46709740e-01
8.40914130e-01 2.14952096e-01 -9.12780017e-02 9.25518274e-01
-7.02226087e-02 -8.44255984e-01 4.69509125e-01 7.10067451e-01
1.29316223e+00 4.24714118e-01 -3.48831527e-02 -4.90425944e-01
-4.29250784e-02 1.74236540e-02 -1.18756533e-01 1.33068478e+00
2.84916103e-01 3.60241055e-01 4.29837704e-01 -3.61356586e-01
-1.27042282e+00 -6.99853659e-01 2.09166825e-01 1.43391979e+00
-5.35345078e-01 -8.61769915e-01 -1.12266600e+00 -4.59998131e-01
-6.42575026e-02 1.32806706e+00 -4.51021850e-01 1.64897414e-03
-5.85614860e-01 -5.52574396e-01 8.88898194e-01 4.68301982e-01
6.15706086e-01 -1.15502632e+00 -5.58106303e-01 3.97442520e-01
-5.74004650e-01 -1.15536082e+00 -6.39526665e-01 6.50207102e-02
-7.45172262e-01 -4.51413244e-01 -7.56503642e-01 -4.57310200e-01
5.25588274e-01 5.97198531e-02 1.49753416e+00 3.07531327e-01
-7.02976659e-02 5.37513137e-01 -4.90235984e-01 -6.66662216e-01
-9.26074803e-01 3.06307912e-01 -5.97035959e-02 -7.50305057e-01
1.58262879e-01 -6.43741786e-01 -4.16217923e-01 -1.98065713e-02
-4.46483135e-01 6.19411707e-01 6.01570070e-01 8.49297166e-01
-8.85447934e-02 -6.64870203e-01 3.37506294e-01 -1.08515835e+00
1.12802124e+00 8.91185999e-02 -2.69622236e-01 3.52845669e-01
-9.61529016e-01 1.24060355e-01 2.98000216e-01 -4.03709352e-01
-1.17833829e+00 -3.86752307e-01 1.32207349e-01 1.26306266e-01
1.10342875e-01 2.41772428e-01 3.76289874e-01 1.51329622e-01
1.18219995e+00 1.74089685e-01 3.81645896e-02 -2.55838603e-01
8.73955786e-01 1.00269806e+00 5.36218822e-01 -1.22764778e+00
6.39315009e-01 6.97415397e-02 -4.06477988e-01 -5.86577713e-01
-8.07756782e-01 8.59265495e-03 -5.56631796e-02 -8.04166198e-02
6.70115530e-01 -1.05824673e+00 -7.12096035e-01 4.61575210e-01
-1.60394776e+00 -9.38301206e-01 -3.01246315e-01 1.33017451e-02
-7.67997801e-01 2.94515371e-01 -7.98570573e-01 -1.05038071e+00
-8.07703197e-01 -9.29482341e-01 1.45687604e+00 2.92709142e-01
-1.05713415e+00 -7.56747365e-01 2.75981873e-02 7.68695891e-01
5.47020257e-01 -1.70210317e-01 1.17609966e+00 -2.46991485e-01
-5.74323475e-01 5.38751148e-02 -3.42290908e-01 1.00340962e-01
-3.36569399e-01 1.58199415e-01 -1.07609558e+00 -2.87911117e-01
-4.06593293e-01 -7.57944405e-01 4.48101789e-01 1.06450126e-01
8.76589060e-01 -8.20422888e-01 -3.21700275e-01 4.60347176e-01
7.05333769e-01 -1.20329492e-01 4.47061211e-01 2.16106072e-01
5.76091707e-01 6.79781914e-01 5.90438783e-01 5.89591503e-01
5.35328209e-01 7.96233356e-01 3.70770693e-04 3.98081727e-03
-3.43602329e-01 -5.90815544e-01 5.55940211e-01 6.68328524e-01
-2.15555608e-01 -4.28436458e-01 -1.04653907e+00 5.44084430e-01
-2.00398970e+00 -7.83173740e-01 -1.87839329e-01 1.66360343e+00
1.42908883e+00 5.52150682e-02 2.87726581e-01 -3.41884255e-01
3.68593633e-01 5.17973192e-02 -1.96085870e-01 -7.86313236e-01
1.49192452e-01 3.49385977e-01 2.38335118e-01 5.26771128e-01
-4.92173880e-01 1.10653591e+00 7.36562729e+00 9.09978449e-01
-7.31250465e-01 1.98155597e-01 9.02568758e-01 -1.53830051e-01
-6.06860340e-01 6.54187053e-02 -1.02473736e+00 4.72225785e-01
1.08452523e+00 -2.24042788e-01 8.20827901e-01 8.66509557e-01
6.39436767e-02 -1.79817736e-01 -1.51598048e+00 7.86935449e-01
1.69804335e-01 -1.27027011e+00 1.60192490e-01 2.07957804e-01
9.77409303e-01 -2.35844150e-01 1.03735574e-01 4.62505102e-01
9.28358436e-01 -1.28654838e+00 1.22568536e+00 3.92538399e-01
1.06641006e+00 -3.76144350e-01 4.18751299e-01 6.17259026e-01
-3.71974438e-01 5.46704105e-04 -2.70482212e-01 -2.20781952e-01
1.04730271e-01 4.34241384e-01 -1.42333555e+00 2.84838617e-01
7.56976247e-01 3.01667333e-01 -8.30609024e-01 4.86169696e-01
-4.34703231e-01 8.34877670e-01 -3.00697982e-01 -3.85361016e-01
2.10675076e-01 2.96472639e-01 4.15557504e-01 1.47395265e+00
3.01218271e-01 -8.14180672e-02 9.83413905e-02 1.17749608e+00
-3.27428095e-02 -3.89775187e-02 -5.01904666e-01 -2.74049878e-01
6.58959746e-01 1.14011812e+00 -2.66689003e-01 -3.94560277e-01
4.31483053e-02 9.83735561e-01 3.85382384e-01 2.16787577e-01
-7.64830112e-01 -6.02861494e-02 1.03995420e-01 4.45484221e-01
3.99273373e-02 4.44418471e-03 -1.77467674e-01 -6.18543863e-01
1.82779096e-02 -1.55774927e+00 -2.04540640e-02 -1.35196412e+00
-1.23939359e+00 6.97993338e-01 2.99383819e-01 -6.88042045e-01
-8.67080927e-01 -4.33854252e-01 -3.80209625e-01 1.03581905e+00
-1.12282717e+00 -1.55983484e+00 -3.12675267e-01 1.82263672e-01
5.87098718e-01 -3.03880721e-01 1.11541069e+00 -1.27846688e-01
-1.20580882e-01 7.12119281e-01 -5.55403948e-01 -2.99997628e-01
1.02206123e+00 -1.55832362e+00 9.40497756e-01 8.86649787e-01
1.51589289e-01 1.07739604e+00 8.42845619e-01 -8.10542226e-01
-1.27663589e+00 -9.81701851e-01 1.20268464e+00 -9.36635554e-01
4.88199115e-01 -3.99705797e-01 -3.44812483e-01 7.21588969e-01
6.47108436e-01 -4.95929360e-01 4.29611415e-01 3.91418874e-01
-5.95581651e-01 4.23655100e-02 -8.10460269e-01 8.15537930e-01
1.34582949e+00 -6.12802744e-01 -2.48278603e-01 9.68418717e-01
1.04105699e+00 -6.15391731e-01 -6.63813174e-01 -8.76651984e-03
5.96379757e-01 -9.28886771e-01 5.63042283e-01 -4.67872679e-01
9.22694027e-01 1.01086393e-01 -3.59401964e-02 -1.09497952e+00
-4.51589078e-01 -1.70971560e+00 -1.41303092e-01 1.44085038e+00
4.87167776e-01 -4.12303388e-01 6.17331445e-01 7.64707565e-01
-1.53424129e-01 -6.09453142e-01 -5.87034106e-01 -4.27383274e-01
9.08573642e-02 -3.99021268e-01 6.16493225e-01 4.22043949e-01
2.08122328e-01 5.27630985e-01 -4.90325928e-01 -5.75830817e-01
4.63767409e-01 -4.19272453e-01 1.19774282e+00 -9.57716346e-01
-8.54973614e-01 -3.25942129e-01 4.10843045e-01 -1.16692138e+00
-7.03381598e-02 -8.68620098e-01 2.57767588e-01 -1.53919172e+00
5.16279578e-01 -5.23985445e-01 6.09661758e-01 8.14000010e-01
-4.87786829e-01 1.26467735e-01 3.94170254e-01 3.05002093e-01
-6.02907956e-01 7.52011240e-02 1.30363965e+00 -1.00147977e-01
6.10371009e-02 -1.90947235e-01 -1.27825356e+00 4.49243069e-01
3.26084167e-01 -3.06290329e-01 -4.76771891e-01 -5.84717572e-01
7.95471430e-01 1.22607790e-01 4.27855164e-01 -9.03039932e-01
4.57853526e-02 -3.01684532e-02 1.14419222e-01 -1.63765952e-01
1.02804646e-01 -1.90540686e-01 1.79420561e-01 3.65753353e-01
-7.20262766e-01 2.13670179e-01 2.51087129e-01 3.94605964e-01
5.11922300e-01 -3.64407629e-01 2.15290084e-01 -3.73791218e-01
-2.47179762e-01 -2.06483290e-01 -4.05960023e-01 2.43598297e-01
4.64784890e-01 -7.90339857e-02 -7.95527101e-01 -7.46019721e-01
1.39498068e-02 3.18266392e-01 4.87254411e-01 3.33252102e-01
2.84438014e-01 -1.09044981e+00 -1.15214241e+00 1.63265038e-02
2.64558285e-01 1.62848487e-01 -1.57187492e-01 6.97244108e-01
-6.96923494e-01 4.59159017e-01 4.83191051e-02 -8.15165639e-01
-1.24203587e+00 2.50349138e-02 6.12191297e-03 -5.95712960e-01
-4.02885497e-01 1.45947909e+00 -1.82640091e-01 -3.65332603e-01
3.12067270e-01 1.75454910e-03 3.27841222e-01 -3.76116067e-01
5.62109470e-01 2.45709106e-01 -1.00361705e-01 -7.48747401e-03
-1.39516905e-01 1.53573900e-01 -6.32429719e-01 -8.51491332e-01
1.12723219e+00 2.20404819e-01 -3.16030473e-01 2.90855408e-01
3.81073922e-01 -2.02492578e-03 -1.23450232e+00 5.68241924e-02
-1.00798421e-01 -3.06936920e-01 -2.04169482e-01 -1.45488870e+00
-2.19912097e-01 8.18138480e-01 2.58936256e-01 1.59160689e-01
6.68045282e-01 1.37512460e-02 7.78767467e-01 4.95722055e-01
7.01332092e-01 -9.43343878e-01 4.71146584e-01 3.31009448e-01
1.09492517e+00 -9.76712465e-01 2.39487305e-01 -1.87197924e-01
-8.07470143e-01 1.00365365e+00 6.71964884e-01 4.10051793e-01
-1.11552462e-01 3.85545522e-01 9.28392448e-03 -1.64142430e-01
-1.38628042e+00 3.32048506e-01 5.39247394e-02 8.66628766e-01
9.35277224e-01 1.28434718e-01 -7.60996789e-02 2.27266863e-01
-7.32159138e-01 2.60489464e-01 4.86853898e-01 8.42276156e-01
-2.00531080e-01 -1.23174417e+00 -2.36213788e-01 4.49184537e-01
-3.74111742e-01 -5.00564635e-01 -5.69760025e-01 6.63601339e-01
1.04242042e-01 1.24109912e+00 -2.27642894e-01 -2.57795632e-01
7.77608948e-03 1.27455950e-01 8.42816651e-01 -9.49521065e-01
-1.07946217e+00 1.86986908e-01 6.55791759e-01 -5.67812383e-01
-2.31029272e-01 -6.52737498e-01 -6.57835066e-01 -7.01089442e-01
-4.28062677e-01 1.40544638e-01 4.72639143e-01 6.99646711e-01
5.65099716e-01 4.18053895e-01 5.12567163e-02 -6.21851742e-01
-9.21239078e-01 -1.45932019e+00 -8.08517262e-02 2.77652234e-01
-2.34279424e-01 -2.10416056e-02 -2.43665248e-01 4.53574479e-01] | [11.538152694702148, 8.770659446716309] |
fa984957-44f6-40a6-8961-5915c9e5de42 | adversarial-machine-learning-based | 2208.05073 | null | https://arxiv.org/abs/2208.05073v1 | https://arxiv.org/pdf/2208.05073v1.pdf | Adversarial Machine Learning-Based Anticipation of Threats Against Vehicle-to-Microgrid Services | In this paper, we study the expanding attack surface of Adversarial Machine Learning (AML) and the potential attacks against Vehicle-to-Microgrid (V2M) services. We present an anticipatory study of a multi-stage gray-box attack that can achieve a comparable result to a white-box attack. Adversaries aim to deceive the targeted Machine Learning (ML) classifier at the network edge to misclassify the incoming energy requests from microgrids. With an inference attack, an adversary can collect real-time data from the communication between smart microgrids and a 5G gNodeB to train a surrogate (i.e., shadow) model of the targeted classifier at the edge. To anticipate the associated impact of an adversary's capability to collect real-time data instances, we study five different cases, each representing different amounts of real-time data instances collected by an adversary. Out of six ML models trained on the complete dataset, K-Nearest Neighbour (K-NN) is selected as the surrogate model, and through simulations, we demonstrate that the multi-stage gray-box attack is able to mislead the ML classifier and cause an Evasion Increase Rate (EIR) up to 73.2% using 40% less data than what a white-box attack needs to achieve a similar EIR. | ['Burak Kantarci', 'Ahmed Omara'] | 2022-08-09 | null | null | null | null | ['inference-attack'] | ['adversarial'] | [-9.47460830e-02 4.00271088e-01 -5.00484146e-02 5.66336885e-02
-9.06627119e-01 -1.11272991e+00 6.15084827e-01 -7.42299706e-02
1.58669706e-02 7.30491996e-01 -5.92853010e-01 -9.15699959e-01
3.81261349e-01 -1.01844025e+00 -1.02032614e+00 -1.27853286e+00
-6.86750531e-01 2.10225776e-01 -6.00854792e-02 -3.38526890e-02
-1.53299600e-01 9.04342651e-01 -7.10817873e-01 9.84846279e-02
6.38141572e-01 1.17891312e+00 -4.72612262e-01 6.91648543e-01
7.49071956e-01 6.07190967e-01 -1.55752385e+00 -3.03453833e-01
5.93642116e-01 -2.26057395e-01 -2.59020537e-01 -3.75153571e-01
-2.10288912e-01 -2.37246737e-01 -6.31739378e-01 1.43415093e+00
5.22500098e-01 -3.57043892e-01 5.08746564e-01 -2.38760519e+00
4.84928526e-02 6.67336524e-01 -1.90063134e-01 9.33661163e-02
4.53642383e-02 6.76656961e-01 1.52237728e-01 -1.82845309e-01
-1.27031617e-02 9.29359019e-01 6.18639946e-01 8.07688773e-01
-1.05939436e+00 -1.35133266e+00 -1.20231912e-01 3.47863972e-01
-1.57862997e+00 -1.99501500e-01 8.98495495e-01 -4.36982550e-02
8.54880035e-01 6.13512099e-01 3.41406584e-01 1.32916617e+00
7.64145255e-01 5.80342233e-01 1.21598458e+00 -3.75318117e-02
8.01708817e-01 4.54746813e-01 -2.06612393e-01 2.41198406e-01
4.10313398e-01 8.13348591e-01 1.63450167e-02 -8.07156265e-01
-1.50264427e-01 -1.78107083e-01 -4.03285354e-01 8.05844888e-02
-5.56567609e-01 9.17535841e-01 6.19165659e-01 -1.24222435e-01
-3.96583438e-01 5.83515286e-01 4.56987679e-01 3.93642336e-01
3.21587801e-01 2.34835759e-01 -6.07005537e-01 2.85162657e-01
-6.42714083e-01 -2.55713463e-01 1.08265805e+00 8.28365088e-01
4.31261331e-01 8.97596359e-01 1.23384364e-01 -5.88629365e-01
2.19459668e-01 9.99001443e-01 1.41866133e-01 -2.07171321e-01
3.01275760e-01 3.21267962e-01 1.39411896e-01 -5.93361259e-01
-5.94314575e-01 -4.86517578e-01 -9.01335835e-01 8.11800897e-01
3.37533087e-01 -9.53299999e-01 -5.84391892e-01 1.55371118e+00
2.03813046e-01 6.28159583e-01 5.22720993e-01 6.97885275e-01
3.13369066e-01 9.38060284e-01 1.52380347e-01 -2.78571427e-01
1.16925752e+00 -5.24684370e-01 -4.99310553e-01 -1.58901095e-01
9.39104795e-01 -7.16902092e-02 3.94426942e-01 4.50281292e-01
-7.38862574e-01 -1.04431197e-01 -1.65621281e+00 1.16444981e+00
-9.17856097e-01 -2.69180655e-01 1.90835342e-01 1.30632830e+00
-6.92670465e-01 1.59461737e-01 -7.60570049e-01 -5.52697852e-02
4.25015718e-01 7.77236640e-01 -1.01038046e-01 2.23612875e-01
-1.48754859e+00 1.16906083e+00 1.93433404e-01 -2.91298665e-02
-1.81859696e+00 -8.49516630e-01 -6.78785324e-01 -2.27512047e-01
2.31870875e-01 -2.09521964e-01 6.80838406e-01 -9.04519618e-01
-1.05142546e+00 3.69933367e-01 2.97934294e-01 -1.05788600e+00
6.02892160e-01 3.86884928e-01 -1.04770684e+00 3.83741893e-02
-4.84111398e-01 -1.20381527e-01 8.54941368e-01 -1.60760403e+00
-6.28735781e-01 -4.37521577e-01 9.08450484e-02 -3.82832021e-01
4.53125015e-02 -5.79558052e-02 6.18329465e-01 -6.94421411e-01
-6.80419743e-01 -1.09299088e+00 -1.34405032e-01 -6.94901288e-01
-7.11119056e-01 -1.13985091e-02 1.60116625e+00 -5.22691250e-01
8.49781096e-01 -2.05511951e+00 -4.84509856e-01 6.32964790e-01
-1.69086009e-01 3.51931989e-01 1.10571161e-01 4.88441139e-01
-1.13888562e-01 1.32161930e-01 -1.28982574e-01 -2.44575620e-01
1.15494609e-01 3.46848905e-01 -7.95701563e-01 1.09015119e+00
-1.08635463e-02 9.98407245e-01 -8.41239512e-01 4.46334571e-01
1.45731479e-01 3.84145111e-01 9.45203304e-02 4.17808384e-01
-7.07126632e-02 4.38250810e-01 -4.43017840e-01 6.18630707e-01
8.62938583e-01 2.15656713e-01 1.94206059e-01 -2.84712851e-01
5.31278312e-01 -1.15714528e-01 -7.25114882e-01 4.03638899e-01
-5.32822371e-01 7.75554359e-01 2.10973680e-01 -9.29146290e-01
8.12500894e-01 4.15289879e-01 -3.91737446e-02 -5.86646140e-01
4.67448741e-01 1.43992901e-01 -6.47945283e-03 1.99401341e-02
-1.45128533e-01 -7.59638241e-03 -6.93179190e-01 5.48783898e-01
-3.47824365e-01 -4.69307490e-02 -6.03438675e-01 1.18462861e-01
1.37875235e+00 -6.53125048e-01 9.71256644e-02 -5.93810320e-01
6.46857679e-01 1.30074143e-01 4.21625972e-01 7.88772464e-01
-2.66673297e-01 -3.49192619e-01 4.59338546e-01 -4.04593706e-01
-5.98054588e-01 -1.22640061e+00 8.27907026e-02 2.66902477e-01
3.56315583e-01 1.85008764e-01 -9.42510962e-01 -1.34692252e+00
3.42463493e-01 1.75056577e+00 -6.91282630e-01 -8.50006580e-01
-3.84800732e-01 -1.16767430e+00 1.28333116e+00 3.91744435e-01
4.72692043e-01 -6.03250384e-01 -5.79400420e-01 4.70233671e-02
3.63438368e-01 -1.16831768e+00 -3.14987957e-01 7.20920265e-01
-1.27335504e-01 -1.29135394e+00 1.46511674e-01 -3.13377202e-01
1.15231860e+00 -6.19506277e-02 8.12442243e-01 2.79704351e-02
2.35205237e-02 2.11857215e-01 -2.28704602e-01 -6.19184911e-01
-1.33159709e+00 -2.95589507e-01 5.15837967e-01 1.26511350e-01
5.51925182e-01 -3.72298002e-01 -3.85788649e-01 5.87337136e-01
-7.80943155e-01 -6.12709105e-01 2.30322868e-01 5.56512058e-01
3.67909595e-02 7.32077122e-01 1.00960445e+00 -6.06849134e-01
3.32665235e-01 -9.03212130e-01 -1.06045127e+00 1.47324353e-01
-8.13854694e-01 -3.16908449e-01 1.32694221e+00 -7.78689206e-01
-3.76379550e-01 -1.43497214e-01 6.56188875e-02 -5.20752430e-01
-2.02810422e-01 -1.09035566e-01 -7.81221569e-01 -8.05524468e-01
3.66897076e-01 4.44158345e-01 -1.45094365e-01 -2.13027522e-02
2.89140522e-01 7.79374063e-01 6.20403290e-01 -8.99955854e-02
1.72470200e+00 5.77665865e-01 3.08708340e-01 -4.56091642e-01
-1.73873365e-01 4.17039514e-01 4.15093601e-02 -4.07828659e-01
6.62602961e-01 -9.31180000e-01 -1.07525861e+00 7.35500991e-01
-1.05253828e+00 -6.67027593e-01 -5.12824692e-02 3.07142258e-01
-1.58327132e-01 -1.23041689e-01 -3.71371269e-01 -9.93951917e-01
-3.83901149e-01 -1.22625697e+00 5.68138003e-01 1.98692530e-01
1.33275002e-01 -1.16028035e+00 -5.55842221e-01 1.94019049e-01
4.50214416e-01 8.30599964e-01 9.78658319e-01 -1.49413967e+00
-4.91128922e-01 -8.25848699e-01 4.07538235e-01 3.82822871e-01
4.84273657e-02 -2.82855541e-01 -1.16632497e+00 -1.03513932e+00
5.49271524e-01 -1.21129580e-01 -1.03288107e-01 -6.72894940e-02
9.70625579e-01 -1.07819533e+00 -5.91557980e-01 8.25574756e-01
1.49437666e+00 8.39089096e-01 5.86523652e-01 1.86491519e-01
4.16253775e-01 9.58171338e-02 4.62300867e-01 2.61346549e-01
3.47779453e-01 5.86065650e-01 1.15678680e+00 -2.29399383e-01
4.77683187e-01 -2.78095424e-01 8.77194881e-01 -3.01216450e-03
7.91805804e-01 -6.92952454e-01 -6.05972946e-01 1.79406315e-01
-1.10432780e+00 -6.88753903e-01 -2.53253412e-02 2.33442283e+00
4.74302292e-01 6.22685850e-01 1.10140190e-01 3.21736664e-01
5.99764645e-01 8.67602006e-02 -8.47526073e-01 -5.51848114e-01
-2.71366715e-01 -2.46096447e-01 1.30380499e+00 5.75232089e-01
-9.75500762e-01 5.62823117e-01 5.72585392e+00 1.07334745e+00
-1.20916617e+00 2.92039633e-01 8.64121854e-01 -9.60621238e-02
-7.15610832e-02 -1.28166601e-01 -7.89661467e-01 9.39791560e-01
1.54862726e+00 -6.26261234e-01 6.03483737e-01 8.69716287e-01
4.89486575e-01 -2.61204001e-02 -1.29330742e+00 4.77302819e-01
2.42977053e-01 -1.05460787e+00 -1.41200706e-01 4.38529968e-01
7.41923273e-01 1.69295266e-01 -1.42981812e-01 3.39972526e-01
8.03027093e-01 -1.26928079e+00 5.75418115e-01 2.54156381e-01
5.50527334e-01 -1.40742803e+00 1.07750010e+00 7.09691703e-01
-1.00884104e+00 -3.59055459e-01 1.29326820e-01 2.03818873e-01
5.58049120e-02 3.52531880e-01 -9.35819507e-01 6.91725731e-01
3.95566940e-01 -3.60615879e-01 -4.40613598e-01 4.61681098e-01
-3.22116166e-01 1.09539437e+00 -5.27624846e-01 -1.96820349e-01
2.97768772e-01 1.64340854e-01 9.33834434e-01 1.07499218e+00
1.59322249e-03 9.76488143e-02 8.14819708e-02 8.18086684e-01
-4.79573496e-02 -8.01807880e-01 -6.60668731e-01 3.79056036e-01
8.58371556e-01 1.34812570e+00 -4.57664549e-01 -3.13041568e-01
-1.36296824e-01 8.38578761e-01 -6.26443624e-01 6.42667532e-01
-1.25244975e+00 -3.95636052e-01 8.34525168e-01 -1.06142476e-01
5.36056142e-03 2.74828747e-02 -3.64131719e-01 -4.67658639e-01
-2.73561329e-01 -9.57464755e-01 4.28997964e-01 -6.83613181e-01
-1.37386394e+00 6.87482059e-01 -9.91732404e-02 -1.35373926e+00
-4.48555857e-01 -3.18175077e-01 -1.20753527e+00 9.81873035e-01
-1.42346680e+00 -1.19126487e+00 -7.43975043e-02 7.23030329e-01
8.78138915e-02 -4.58436847e-01 8.79622459e-01 9.59935412e-02
-7.26954341e-01 1.28882647e+00 3.73116046e-01 4.45324272e-01
-2.61181332e-02 -1.03136551e+00 7.03124583e-01 1.19854057e+00
-2.42248952e-01 -9.19445455e-02 9.85902250e-01 -6.62787437e-01
-2.02722120e+00 -1.69300008e+00 5.36264200e-03 -4.80527997e-01
8.41144204e-01 -6.89627707e-01 -5.20949662e-01 8.50631237e-01
1.70626059e-01 3.09060991e-01 5.30681133e-01 -1.18432522e+00
-1.61642447e-01 -3.65365922e-01 -1.98610759e+00 6.08027756e-01
1.93563566e-01 -5.66903532e-01 -2.99891710e-01 1.50766611e-01
5.98824561e-01 -1.19769908e-01 -8.28163564e-01 5.65047443e-01
-4.79684360e-02 -3.64833802e-01 7.02501416e-01 -6.09108627e-01
-8.06791425e-01 -6.79527640e-01 -3.72144103e-01 -1.82369924e+00
4.27256793e-01 -1.19793987e+00 -5.82300723e-01 1.05627477e+00
3.10830444e-01 -1.27204001e+00 7.89441943e-01 2.74855554e-01
6.57579750e-02 -4.95094538e-01 -1.44585323e+00 -1.16562498e+00
4.20376718e-01 -5.13646781e-01 9.71681714e-01 8.79665375e-01
1.34151340e-01 -2.95739532e-01 -2.10386023e-01 1.29534054e+00
1.10799527e+00 -2.07796097e-01 8.82240951e-01 -5.04606009e-01
-5.94898798e-02 -2.66311139e-01 -6.33060038e-01 -5.30662954e-01
2.35009462e-01 -8.84284079e-01 -8.79056081e-02 -6.75641835e-01
-5.02447486e-01 -1.97032213e-01 -2.86338478e-01 4.60671484e-01
1.42411828e-01 4.39232439e-01 2.79775321e-01 -2.45921105e-01
-3.95855904e-02 3.08664590e-01 2.05940694e-01 -6.33580625e-01
1.53566763e-01 5.36898553e-01 -3.48607510e-01 6.38933718e-01
1.10684657e+00 -7.95170903e-01 -2.91279286e-01 3.08268964e-01
-2.94828087e-01 3.05208176e-01 7.88947582e-01 -9.83650863e-01
4.02438939e-01 -4.25406769e-02 4.15334642e-01 -6.01019859e-01
1.92394927e-02 -1.30208480e+00 5.82422495e-01 1.10488498e+00
2.77381182e-01 5.29617406e-02 5.35274684e-01 4.93236303e-01
2.56965965e-01 1.08454183e-01 8.92112851e-01 4.09187645e-01
-3.22176695e-01 1.27334699e-01 -8.62374246e-01 -2.51777828e-01
1.75437284e+00 3.11755966e-02 -9.32504117e-01 -5.16424179e-01
-5.25754571e-01 5.98452985e-01 5.79890132e-01 4.86872852e-01
3.23814631e-01 -1.30166876e+00 -6.97157085e-01 4.40717846e-01
-7.15667605e-02 -5.80351889e-01 -1.11797512e-01 5.04915297e-01
-1.15107456e-02 2.38628507e-01 1.09784320e-01 -2.86180735e-01
-1.01726055e+00 8.84074032e-01 1.03305256e+00 -9.25364792e-02
-2.84982383e-01 3.01366538e-01 1.87667012e-01 -1.14254363e-01
4.04731870e-01 6.00108802e-02 2.80327350e-01 -2.20789522e-01
4.79890317e-01 6.60681903e-01 3.04531515e-01 -6.77698076e-01
-7.03934908e-01 4.78404611e-02 5.21632671e-01 3.16362739e-01
6.25523388e-01 6.71972334e-02 1.60993159e-01 1.21747129e-01
1.32780802e+00 1.75904527e-01 -1.27489448e+00 1.88542143e-01
-1.18205912e-01 -8.12215060e-02 7.01168254e-02 -1.22275388e+00
-1.53775799e+00 4.00600433e-01 7.90614963e-01 8.66591692e-01
1.19503891e+00 -6.80008382e-02 7.90191352e-01 4.40880544e-02
8.92999828e-01 -5.73606014e-01 -4.32157129e-01 -2.64400542e-01
5.52681148e-01 -8.18864584e-01 -2.89313644e-01 -5.91957234e-02
-4.86181140e-01 8.64644766e-01 5.52387893e-01 -2.91785359e-01
8.44297111e-01 1.10600424e+00 2.23060265e-01 -8.12314451e-02
-8.08973968e-01 5.24488270e-01 -1.56000659e-01 9.33162153e-01
-9.78616834e-01 4.58355069e-01 3.49643379e-01 9.10510898e-01
-1.78387627e-01 -5.03262401e-01 9.26318109e-01 5.44354141e-01
-2.08637584e-02 -5.18116653e-01 -6.86774850e-01 3.63691837e-01
-5.97927511e-01 2.26687744e-01 -2.18320921e-01 8.22958469e-01
-2.63885595e-03 1.53494656e+00 1.52413115e-01 -8.26850355e-01
2.40861863e-01 -7.02385604e-02 -2.43135363e-01 1.76141679e-01
-8.22413862e-01 -4.16670322e-01 5.43560460e-02 -5.94041288e-01
2.72320807e-01 -4.39198077e-01 -1.32682824e+00 -6.88529968e-01
-3.02750200e-01 5.27771771e-01 7.94063091e-01 9.58568811e-01
9.35973600e-02 4.89144385e-01 1.74257708e+00 -8.26970696e-01
-9.63472247e-01 -8.01740348e-01 -8.17566216e-01 -5.44546992e-02
5.34319818e-01 -3.25653821e-01 -1.41878355e+00 -5.06767929e-01] | [5.560283660888672, 7.491052150726318] |
c96e7945-0585-4a08-b054-00cfa4c71424 | countering-the-influence-of-essay-length-in | null | null | https://aclanthology.org/2021.sustainlp-1.4 | https://aclanthology.org/2021.sustainlp-1.4.pdf | Countering the Influence of Essay Length in Neural Essay Scoring | Previous work has shown that automated essay scoring systems, in particular machine learning-based systems, are not capable of assessing the quality of essays, but are relying on essay length, a factor irrelevant to writing proficiency. In this work, we first show that state-of-the-art systems, recent neural essay scoring systems, might be also influenced by the correlation between essay length and scores in a standard dataset. In our evaluation, a very simple neural model shows the state-of-the-art performance on the standard dataset. To consider essay content without taking essay length into account, we introduce a simple neural model assessing the similarity of content between an input essay and essays assigned different scores. This neural model achieves performance comparable to the state of the art on a standard dataset as well as on a second dataset. Our findings suggest that neural essay scoring systems should consider the characteristics of datasets to focus on text quality. | ['Michael Strube', 'Sungho Jeon'] | null | null | null | null | emnlp-sustainlp-2021-11 | ['automated-essay-scoring'] | ['natural-language-processing'] | [-1.43448487e-01 -5.32712676e-02 -1.43455118e-01 -6.43251181e-01
-4.84296113e-01 -5.83310187e-01 4.39500242e-01 4.82361794e-01
-7.88156271e-01 7.18151808e-01 2.36392036e-01 -3.56298864e-01
-4.47327942e-01 -9.56167758e-01 -4.23728563e-02 3.05757709e-02
7.78060257e-01 6.35417879e-01 1.23148337e-01 -4.19444740e-01
9.55986500e-01 6.53079525e-02 -1.39983523e+00 2.55451679e-01
1.06035364e+00 6.74241602e-01 -1.75622627e-01 1.05037403e+00
-4.17497337e-01 1.02609789e+00 -1.18454421e+00 -7.73726344e-01
1.84185319e-02 -7.14723885e-01 -8.68806243e-01 -4.74703729e-01
1.08355486e+00 -5.48470020e-01 -3.44534069e-01 9.20434415e-01
2.70206541e-01 2.77539790e-01 9.83727217e-01 -7.32401371e-01
-1.32395351e+00 6.96697772e-01 -6.58033714e-02 3.78042310e-01
5.59765399e-01 1.06920451e-01 1.37647009e+00 -8.85687470e-01
3.87272388e-01 7.12066293e-01 1.01201999e+00 6.09570920e-01
-8.92237484e-01 -4.67329085e-01 -7.20729381e-02 3.94932985e-01
-5.51852226e-01 -2.64930904e-01 6.19833410e-01 -5.49077511e-01
7.56092548e-01 1.95096329e-01 6.37704194e-01 9.17157054e-01
3.84878069e-01 8.62149477e-01 1.33819199e+00 -7.27355957e-01
5.70525005e-02 5.28610907e-02 1.25080073e+00 9.43621099e-01
4.46552962e-01 -1.10541232e-01 -9.11790073e-01 6.33891597e-02
4.25192237e-01 -3.98439705e-01 -1.29647061e-01 2.77182251e-01
-9.99447227e-01 1.01507795e+00 -1.23308651e-01 5.19265056e-01
-5.92308082e-02 -8.68153125e-02 4.03015226e-01 8.01246822e-01
5.00654161e-01 1.22107863e+00 -4.29540932e-01 -5.34271657e-01
-1.60217214e+00 4.10810322e-01 1.19070959e+00 2.17283592e-01
3.50484043e-01 5.83081692e-02 -7.49213457e-01 9.60778117e-01
1.90232143e-01 1.05272889e-01 1.11872721e+00 -1.05315459e+00
4.87304449e-01 9.07160759e-01 -7.37558911e-03 -8.69035542e-01
-7.35288620e-01 -4.12747949e-01 -3.29415053e-01 4.36734676e-01
1.16503882e+00 -1.68930963e-01 -3.04315925e-01 1.57002306e+00
-3.86660993e-01 -6.29133403e-01 -1.55232191e-01 7.83069432e-01
1.33142328e+00 1.94002479e-01 -2.57738858e-01 -2.94132996e-02
1.20261216e+00 -1.07328451e+00 -8.20392430e-01 -1.00144923e-01
9.08665180e-01 -7.09353447e-01 1.60922968e+00 7.11121857e-01
-1.62990189e+00 -8.06350291e-01 -1.41617429e+00 -4.14094329e-01
-4.40959483e-01 5.79989910e-01 3.25286835e-01 1.01319027e+00
-1.29720700e+00 9.56710458e-01 -3.26449513e-01 -2.97265321e-01
6.63075298e-02 2.45183289e-01 -2.43278503e-01 3.77108574e-01
-1.33014095e+00 1.32405782e+00 7.67473653e-02 -3.08524013e-01
-1.03484124e-01 -7.17523873e-01 -5.53195119e-01 4.60541338e-01
-1.59746245e-01 -4.59960580e-01 1.67901218e+00 -1.00049472e+00
-1.86251616e+00 9.91769612e-01 1.97745338e-02 -1.99772418e-01
3.97580683e-01 -1.08277805e-01 -1.13978334e-01 -1.65647998e-01
-6.72130808e-02 1.78771958e-01 2.60038793e-01 -4.60706413e-01
-4.37999696e-01 -4.93731111e-01 1.53175533e-01 1.61154658e-01
-9.25262094e-01 -5.94435213e-03 1.25635847e-01 -4.62441146e-01
-1.63801521e-01 -6.05951190e-01 3.49423319e-01 -7.15907440e-02
4.38066721e-02 -7.73633718e-01 2.45596766e-01 -6.80054009e-01
1.76801407e+00 -1.47818995e+00 -2.18239248e-01 -1.57407466e-02
2.63337016e-01 3.86170000e-01 -2.98340172e-01 1.88816771e-01
3.07191998e-01 1.72624260e-01 1.78900316e-01 -4.31083381e-01
4.39313650e-01 -3.33211660e-01 1.86526179e-01 3.11411709e-01
-1.01315916e-01 1.08774447e+00 -8.12689483e-01 -3.91198099e-01
-2.96336710e-01 -1.39200374e-01 -3.24462712e-01 3.05548400e-01
-8.38637911e-03 -2.88730502e-01 -2.22609371e-01 4.56216723e-01
1.46701813e-01 -3.05226862e-01 -9.23678726e-02 4.01776463e-01
-1.48814827e-01 7.64307976e-01 -6.17500305e-01 1.55492079e+00
-3.34316343e-01 1.31083500e+00 -4.75677997e-01 -4.57299262e-01
1.32842720e+00 2.54977077e-01 -1.43646047e-01 -9.04439270e-01
2.25048602e-01 3.45040262e-01 5.19826055e-01 -3.96627069e-01
1.12453115e+00 5.09427190e-02 -3.01980823e-02 1.49178517e+00
3.16463143e-01 -3.25356424e-01 7.06060290e-01 1.35765284e-01
1.24874949e+00 -8.42979625e-02 4.55600321e-01 -4.03088301e-01
6.05031073e-01 -1.92027718e-01 -1.54396638e-01 1.02878273e+00
-4.55499738e-01 7.06988454e-01 7.72807240e-01 -4.85186756e-01
-1.26326621e+00 -8.52759719e-01 -1.24063864e-01 1.44041920e+00
-5.37416875e-01 -2.84403652e-01 -1.09885085e+00 -8.50625813e-01
-3.78817916e-02 7.87961006e-01 -1.03181148e+00 -1.89695671e-01
-3.91899794e-01 -7.66904593e-01 9.54512477e-01 5.00810683e-01
1.45658806e-01 -1.34482229e+00 -7.32152224e-01 3.37143838e-01
-6.35644943e-02 -5.02947807e-01 -4.19845015e-01 2.26164013e-01
-1.00488889e+00 -1.04107285e+00 -8.31855714e-01 -4.23408806e-01
3.53671581e-01 -8.82665366e-02 1.67844558e+00 5.80007255e-01
2.79991180e-01 3.38914692e-01 -2.45152757e-01 -5.92955828e-01
-4.98744309e-01 7.62207925e-01 -7.07016811e-02 -5.03788948e-01
9.49009717e-01 -2.95272589e-01 -2.83372581e-01 1.48897758e-02
-6.82866573e-01 -2.79319018e-01 4.39360350e-01 1.04975867e+00
-2.06703514e-01 -6.69497728e-01 9.16932464e-01 -9.40066755e-01
1.74807441e+00 -2.76959538e-01 -1.84078127e-01 3.83126616e-01
-1.16973186e+00 9.29416046e-02 8.31714749e-01 -3.87110651e-01
-8.34330261e-01 -4.71028894e-01 -1.19022615e-01 3.18315983e-01
-1.67884842e-01 8.12287867e-01 5.23976266e-01 -7.14451596e-02
1.09804046e+00 1.62064597e-01 1.37468651e-01 -3.34931642e-01
-1.62742361e-01 7.30662823e-01 5.33113897e-01 -6.08157158e-01
2.10503370e-01 -2.17221171e-01 4.73709404e-03 -3.16070437e-01
-1.31675756e+00 -4.68585670e-01 -1.05475581e+00 -5.07782996e-01
2.99665213e-01 -3.82837385e-01 -7.82474875e-01 5.79037309e-01
-1.23339617e+00 -4.54843253e-01 -2.72316545e-01 4.36395049e-01
-6.04090214e-01 3.41353029e-01 -9.89760995e-01 -6.21075273e-01
-5.99633634e-01 -8.15382600e-01 4.21369970e-01 5.63146710e-01
-9.36452985e-01 -1.34041488e+00 7.30547905e-01 5.61763287e-01
4.43820745e-01 -3.81101876e-01 1.00955844e+00 -1.14509225e+00
1.09130770e-01 -3.22040886e-01 -2.81461865e-01 3.46789122e-01
-2.28165388e-01 2.37668931e-01 -9.61091816e-01 1.97860673e-01
6.95303036e-03 -7.58989751e-01 9.75541830e-01 3.23000669e-01
1.04986346e+00 -7.95294642e-02 6.74167871e-01 2.50325143e-01
1.04412723e+00 -1.99768379e-01 5.49251735e-01 7.46168256e-01
3.46371979e-01 6.90920949e-01 3.63169432e-01 3.46321851e-01
5.11750519e-01 3.47848386e-01 -2.96437945e-02 2.64058053e-01
-1.20720178e-01 4.20509204e-02 4.29975688e-01 1.30405128e+00
-1.32694408e-01 -3.19774210e-01 -9.26625431e-01 5.74241996e-01
-1.80276585e+00 -9.96159434e-01 -6.80569530e-01 1.96784580e+00
1.12345338e+00 2.13247567e-01 2.97634751e-01 6.07331336e-01
1.39824376e-01 2.11426705e-01 -2.85834789e-01 -1.12607515e+00
-9.14200321e-02 6.59155250e-01 1.32153273e-01 5.60081840e-01
-6.75911129e-01 5.41722894e-01 7.42214680e+00 4.83015686e-01
-8.35368156e-01 1.14053398e-01 5.27242422e-01 -3.75687718e-01
-1.88568264e-01 -7.02882111e-01 -7.17559218e-01 4.71641541e-01
1.40202117e+00 -3.76212567e-01 1.22271463e-01 5.75343251e-01
1.94447916e-02 -2.62877136e-01 -1.29924333e+00 3.79357904e-01
5.28311789e-01 -1.02813578e+00 -9.78814587e-02 -2.10192889e-01
8.67405951e-01 -3.05861115e-01 2.30207890e-01 7.06786096e-01
2.87361413e-01 -1.33658922e+00 8.48692358e-01 9.14147139e-01
5.49964726e-01 -6.89625382e-01 1.09963715e+00 3.88358891e-01
-3.72753114e-01 2.09854040e-02 -5.94602168e-01 -8.25087190e-01
-4.05833572e-01 5.48094332e-01 -5.96195698e-01 -9.65998247e-02
3.29841465e-01 5.05703986e-01 -1.18414426e+00 9.78073716e-01
-4.01220918e-01 8.59447837e-01 2.10695356e-01 -8.21000159e-01
4.21712369e-01 2.63173170e-02 3.06063630e-02 1.35185516e+00
3.98840874e-01 7.19334558e-02 -3.93135697e-01 8.95627260e-01
-3.59528452e-01 3.88770550e-01 -2.86900461e-01 -6.40835017e-02
1.59057513e-01 1.19412613e+00 -4.16944772e-01 -3.49901497e-01
-5.27512312e-01 7.04444051e-01 8.11447740e-01 -5.00878990e-02
-4.36823577e-01 -5.85147500e-01 2.55292892e-01 -2.92852763e-02
-2.93020189e-01 -5.89299910e-02 -1.31804454e+00 -9.77319181e-01
7.01519623e-02 -9.15473342e-01 3.04192930e-01 -8.57738018e-01
-1.51798415e+00 3.31725717e-01 -7.27596819e-01 -7.68113971e-01
-3.92165571e-01 -1.06032765e+00 -1.02425122e+00 1.12157953e+00
-1.42105699e+00 -4.49472576e-01 -5.48095226e-01 2.11073741e-01
6.05549335e-01 -6.21713459e-01 9.30864751e-01 -3.03657334e-02
-3.55912566e-01 1.04475284e+00 3.37509155e-01 3.69289517e-01
1.23929822e+00 -1.69973099e+00 4.00926977e-01 4.27343071e-01
2.64966905e-01 6.24725282e-01 4.96800095e-01 -3.84551883e-01
-6.15563810e-01 -4.76114959e-01 1.40083754e+00 -9.72445965e-01
7.93136001e-01 1.87413946e-01 -1.22948945e+00 2.57847369e-01
3.62822413e-01 -6.53750181e-01 1.08692944e+00 6.87331975e-01
-4.98811662e-01 3.03539157e-01 -8.83465409e-01 4.80303973e-01
5.50546885e-01 -7.64851093e-01 -1.03383970e+00 -3.56815313e-03
1.50402889e-01 -3.26567113e-01 -9.07945275e-01 1.25893071e-01
1.30099964e+00 -1.37905884e+00 6.01306617e-01 -8.54723036e-01
1.38498652e+00 3.24104667e-01 3.64282250e-01 -1.60854018e+00
-7.20086038e-01 7.12382272e-02 -3.91283929e-01 7.47013211e-01
6.71475232e-01 -1.39843762e-01 1.20172858e+00 7.78579295e-01
-3.59674282e-02 -1.14418459e+00 -6.07794702e-01 -6.20524347e-01
8.87942553e-01 -1.10611863e-01 6.08564258e-01 1.10049403e+00
6.26884818e-01 4.47358578e-01 -1.97341833e-02 -6.02686584e-01
2.43293583e-01 -1.33991957e-01 5.73027909e-01 -1.96428561e+00
-1.17281348e-01 -1.45389581e+00 -1.78953871e-01 -5.87964356e-01
4.40386951e-01 -7.84003496e-01 1.77869239e-04 -1.60480595e+00
4.60690230e-01 1.09387068e-02 -4.21831071e-01 2.40478247e-01
-6.34825349e-01 5.29280543e-01 3.00305694e-01 1.87540218e-01
-6.78859055e-01 1.81733251e-01 1.39584005e+00 -2.65684038e-01
-6.16519563e-02 -7.40930885e-02 -8.04941893e-01 7.64223158e-01
9.92332518e-01 -4.94711727e-01 -1.66533753e-01 -4.78876173e-01
7.17894614e-01 9.42314193e-02 -5.21206185e-02 -1.11563575e+00
5.57259917e-01 -1.11631557e-01 6.19714618e-01 -4.27419692e-01
5.67716882e-02 -5.49302250e-02 -7.27339983e-01 3.29086751e-01
-1.00990593e+00 3.45644355e-01 1.50179602e-02 -1.49865583e-01
-2.60344684e-01 -1.16326213e+00 6.74422979e-01 -3.18722188e-01
-1.48014188e-01 -1.04426056e-01 -7.10774422e-01 1.24271616e-01
3.02843601e-01 -4.83591884e-01 -6.89790785e-01 -5.85516810e-01
-3.96355033e-01 -3.95419374e-02 3.66562963e-01 4.50810879e-01
4.21657592e-01 -1.18909097e+00 -1.10223281e+00 -2.39337385e-02
8.89900774e-02 -7.15329051e-01 -1.55818630e-02 6.41523957e-01
-6.52635932e-01 9.20908451e-01 -6.74790740e-01 -2.41410956e-02
-1.46139753e+00 -1.45900816e-01 4.13135022e-01 -9.36838388e-01
1.56775102e-01 7.23743737e-01 -2.75573552e-01 -9.29809809e-01
1.37347504e-01 -2.61419028e-01 -7.66604245e-01 3.86424333e-01
8.29560339e-01 7.36889482e-01 4.32968676e-01 -1.86404929e-01
3.91316086e-01 2.68371493e-01 -1.28751442e-01 -2.34355256e-01
1.00237215e+00 2.33965993e-01 -1.12147052e-02 9.28227842e-01
6.36223495e-01 7.55231231e-02 -5.93466938e-01 -4.27375175e-02
2.66825080e-01 -3.95478159e-01 7.44462237e-02 -1.29458892e+00
-6.80031955e-01 1.13833320e+00 9.25670937e-02 3.30989957e-01
5.47597468e-01 -7.72682667e-01 6.80957437e-01 9.75675702e-01
7.45505542e-02 -1.48941135e+00 4.73707348e-01 1.18217599e+00
6.63630664e-01 -1.30650055e+00 2.09190428e-01 4.63695228e-01
-3.15972239e-01 1.76475263e+00 9.40634549e-01 -2.38482237e-01
2.25615472e-01 5.98358251e-02 3.14373344e-01 -1.38730168e-01
-9.70914125e-01 2.20289171e-01 8.01062405e-01 2.42169619e-01
1.22719407e+00 7.96507969e-02 -8.67380202e-01 1.22480726e+00
-7.89414167e-01 2.24019498e-01 1.34011149e+00 4.58487898e-01
-7.46390879e-01 -1.05768573e+00 -5.13641119e-01 1.10117400e+00
-8.50708485e-01 -3.37215811e-01 -1.02742517e+00 6.36679471e-01
-2.45823026e-01 8.47598672e-01 9.21213478e-02 -2.66056925e-01
3.20836276e-01 7.31480479e-01 7.55107522e-01 -7.87540734e-01
-1.40020633e+00 -1.04434335e+00 9.07152444e-02 -1.03427947e-01
-2.01076388e-01 -7.48059988e-01 -7.41344333e-01 -7.28632748e-01
-5.27124941e-01 1.35318547e-01 6.98827624e-01 1.11555386e+00
-2.34273955e-01 6.11748040e-01 3.03252432e-02 -4.82348770e-01
-9.18651283e-01 -1.44374084e+00 -7.64151692e-01 2.68170267e-01
2.02661827e-01 -3.68770391e-01 -4.10703599e-01 -2.56961197e-01] | [11.285841941833496, 9.352557182312012] |
2775557a-5dc6-4ac6-9fd1-61235f64d964 | a-dynamic-mode-decomposition-approach-for | 2203.00004 | null | https://arxiv.org/abs/2203.00004v2 | https://arxiv.org/pdf/2203.00004v2.pdf | A Dynamic Mode Decomposition Approach for Decentralized Spectral Clustering of Graphs | We propose a novel robust decentralized graph clustering algorithm that is provably equivalent to the popular spectral clustering approach. Our proposed method uses the existing wave equation clustering algorithm that is based on propagating waves through the graph. However, instead of using a fast Fourier transform (FFT) computation at every node, our proposed approach exploits the Koopman operator framework. Specifically, we show that propagating waves in the graph followed by a local dynamic mode decomposition (DMD) computation at every node is capable of retrieving the eigenvalues and the local eigenvector components of the graph Laplacian, thereby providing local cluster assignments for all nodes. We demonstrate that the DMD computation is more robust than the existing FFT based approach and requires 20 times fewer steps of the wave equation to accurately recover the clustering information and reduces the relative error by orders of magnitude. We demonstrate the decentralized approach on a range of graph clustering problems. | ['Tuhin Sahai', 'Stefan Klus', 'Hongyu Zhu'] | 2022-02-26 | null | null | null | null | ['graph-clustering'] | ['graphs'] | [ 5.07857017e-02 -4.68549505e-02 2.67260134e-01 2.86515057e-01
-5.76541007e-01 -8.20089042e-01 3.01716477e-01 3.97376567e-01
-2.52185374e-01 2.01578736e-01 -2.65948117e-01 -4.19423252e-01
-5.92047155e-01 -9.64243054e-01 -4.28283036e-01 -1.12718904e+00
-6.36911213e-01 4.69457299e-01 3.00288886e-01 -1.50241882e-01
2.46800959e-01 5.62924087e-01 -9.63341534e-01 -4.05074805e-01
7.32426882e-01 6.56494379e-01 -6.91641346e-02 9.78237629e-01
2.11389929e-01 6.33746266e-01 -3.39238584e-01 8.72762501e-02
4.01550382e-01 -6.42881215e-01 -8.79223466e-01 3.77994061e-01
5.81361055e-02 1.41522586e-02 -3.68698567e-01 1.17641032e+00
4.57713783e-01 4.57260579e-01 5.82991421e-01 -1.32607269e+00
-1.62964225e-01 5.82149088e-01 -1.01005375e+00 1.22411758e-01
5.61041236e-01 -2.83610344e-01 8.60579014e-01 -5.86389184e-01
7.11915076e-01 7.50556707e-01 9.71942961e-01 -8.03516731e-02
-1.51638508e+00 -5.02680659e-01 -2.43016556e-01 2.00651929e-01
-1.97033179e+00 -2.76334733e-01 1.14978695e+00 -3.21532935e-01
5.44468105e-01 3.10623854e-01 6.21543586e-01 1.20427758e-01
-6.47311583e-02 -4.95236479e-02 8.88980806e-01 -5.29416621e-01
4.90338832e-01 -5.34004152e-01 1.58987403e-01 8.96045387e-01
3.16259354e-01 -2.28561595e-01 -4.21863586e-01 -6.33813024e-01
5.33711135e-01 -9.63888466e-02 -3.91989380e-01 -6.50694907e-01
-1.06426752e+00 8.00127387e-01 3.85440290e-01 3.75027090e-01
-4.98212010e-01 5.48728585e-01 -5.17991779e-04 3.88818949e-01
2.96834201e-01 -8.25522169e-02 1.90589949e-01 9.16253328e-02
-1.21315813e+00 4.30172943e-02 1.15465581e+00 7.49260843e-01
1.28317213e+00 -2.03948230e-01 4.28604215e-01 1.09258562e-01
4.25923139e-01 5.40295541e-01 -1.99609652e-01 -1.32883191e+00
1.55958027e-01 6.84808910e-01 -1.28380647e-02 -1.73661911e+00
-5.63299656e-01 -4.06006157e-01 -1.16944802e+00 -9.40836221e-02
2.83661097e-01 -4.81264979e-01 -3.28605771e-01 1.52323759e+00
6.59707963e-01 5.97539008e-01 -4.25815620e-02 7.05585063e-01
2.31778637e-01 7.38488138e-01 -5.22192717e-01 -7.30163634e-01
7.61710942e-01 -5.82717538e-01 -4.72985923e-01 3.81736219e-01
6.08693779e-01 -7.48305500e-01 3.18976820e-01 2.46488720e-01
-1.02285945e+00 1.26439864e-02 -9.97833490e-01 3.67858589e-01
2.83305421e-02 -3.71617824e-01 3.94695401e-01 7.71721840e-01
-1.65654182e+00 6.15375042e-01 -1.11836112e+00 -4.80526626e-01
-2.11342826e-01 5.12892425e-01 -2.03543827e-01 -4.64855060e-02
-5.26505649e-01 1.37755737e-01 8.69591907e-02 1.32133603e-01
-4.92991388e-01 -6.90871060e-01 -5.99584758e-01 2.18235537e-01
2.72880584e-01 -5.69343984e-01 6.19491994e-01 -7.18968511e-01
-1.32638490e+00 3.85077894e-01 -4.59349960e-01 -3.25724989e-01
3.20270449e-01 4.88730609e-01 -2.45741177e-02 9.12638187e-01
5.26045561e-02 -1.50538281e-01 9.52282965e-01 -1.40894079e+00
-6.79103136e-02 -2.27177501e-01 -1.47594646e-01 7.54580796e-02
-4.31839526e-01 -1.28608614e-01 -3.26866150e-01 -3.58238727e-01
6.40590072e-01 -1.07899320e+00 -5.30103028e-01 -2.63557076e-01
-4.46288705e-01 1.58438347e-02 9.10218000e-01 -5.35237551e-01
1.29679573e+00 -2.18912840e+00 3.77822042e-01 1.08997846e+00
9.09070790e-01 -3.77692819e-01 9.07133892e-02 1.21874356e+00
8.27858448e-02 1.17323361e-01 -5.21823883e-01 -1.07711032e-01
-1.78638607e-01 5.48313148e-02 7.24583566e-02 1.13140774e+00
-4.64995593e-01 4.27807957e-01 -9.58332479e-01 -4.17287201e-01
7.08729774e-02 5.35614014e-01 -6.80662215e-01 -2.86150843e-01
3.60432327e-01 4.39444125e-01 -3.80206317e-01 1.94187358e-01
1.07884550e+00 -3.98214638e-01 7.20330715e-01 -1.64523274e-01
-2.22396001e-01 -1.49379000e-01 -1.60940957e+00 1.61984742e+00
-1.30349740e-01 4.31104153e-01 8.49663317e-01 -1.43734109e+00
5.96147358e-01 4.84277308e-01 1.16346133e+00 -7.91761279e-02
5.28755225e-02 6.28820807e-02 -1.81636468e-01 -6.59757927e-02
8.04182589e-02 -4.64528352e-02 1.55686261e-02 8.99199009e-01
-1.81308970e-01 1.91583410e-01 4.32267368e-01 8.72981608e-01
1.69998252e+00 -4.45807576e-01 -9.51082725e-03 -6.49444640e-01
5.15324235e-01 3.13045770e-01 3.65405738e-01 5.41710854e-01
-1.18239306e-01 1.54896647e-01 3.99610400e-01 -5.11572920e-02
-6.81892931e-01 -1.07098770e+00 2.40550473e-01 6.84092641e-01
5.47768891e-01 -8.98585200e-01 -9.20823872e-01 -1.54882997e-01
-1.30955577e-01 1.28330989e-02 -2.69994646e-01 1.21860886e-02
-4.24279720e-01 -8.54131043e-01 4.19411331e-01 8.02405179e-02
4.06171948e-01 -2.17953697e-01 -4.75315720e-01 2.00761154e-01
-3.19159746e-01 -1.03020573e+00 -4.88631576e-01 -1.84244335e-01
-9.92528260e-01 -1.33300602e+00 -3.05666506e-01 -9.16627645e-01
1.11479414e+00 7.52035737e-01 6.75509989e-01 5.24486959e-01
-2.37631828e-01 8.65893245e-01 -4.22271639e-01 5.15378714e-01
-2.90830195e-01 -2.35952754e-02 9.71201956e-02 3.93736333e-01
-9.40976217e-02 -1.14666200e+00 -6.44153655e-01 -4.93338592e-02
-7.77321100e-01 -1.55463114e-01 2.94142902e-01 3.98932636e-01
4.28498983e-01 8.29921663e-01 1.51423678e-01 -7.71720111e-01
7.30016351e-01 -6.43037796e-01 -7.62973130e-01 -2.79658772e-02
-6.26036644e-01 1.28188338e-02 7.01715231e-01 -8.69871974e-02
-6.63741112e-01 4.60792094e-01 3.35450530e-01 -2.90697485e-01
1.51729077e-01 5.82936943e-01 4.75993007e-01 -8.05742264e-01
5.08669615e-01 3.42231929e-01 -6.71706125e-02 -4.47758377e-01
5.01312494e-01 2.84417659e-01 5.63647926e-01 -5.19558549e-01
1.54344022e+00 9.22631562e-01 6.21815145e-01 -1.12666690e+00
1.25037715e-01 -8.74725699e-01 -8.32753479e-01 -4.94605154e-01
7.12757707e-01 -6.59551799e-01 -1.24039233e+00 2.17217967e-01
-9.51178253e-01 -1.31538287e-01 2.15398937e-01 3.87635678e-01
-3.40650141e-01 7.91738749e-01 -7.58827984e-01 -9.13846612e-01
-1.96890444e-01 -6.88461125e-01 7.35269427e-01 -2.02119827e-01
-1.51235657e-02 -1.13062680e+00 3.18027049e-01 1.00437507e-01
4.22890931e-01 7.13087857e-01 8.66725504e-01 -6.36361688e-02
-7.08145201e-01 -3.23599905e-01 -1.24585710e-01 -3.75701368e-01
2.10287441e-02 1.07502177e-01 -3.57404917e-01 -6.67967856e-01
1.04792684e-01 2.89398432e-01 5.32550812e-01 3.37573767e-01
4.36195850e-01 -5.24654686e-01 -4.71382052e-01 7.44776070e-01
1.92073905e+00 -2.66655296e-01 2.41499841e-01 -1.80230409e-01
8.76826167e-01 4.54594225e-01 -1.94635019e-01 5.75889826e-01
5.33475459e-01 4.65459406e-01 1.52482301e-01 2.40134373e-02
8.62139016e-02 3.87166068e-02 3.64446193e-01 1.29371631e+00
-3.90833437e-01 4.27305438e-02 -1.14873326e+00 6.71001256e-01
-2.02018762e+00 -1.02707195e+00 -5.99092960e-01 2.03679109e+00
4.57369238e-01 -4.20234889e-01 3.30838352e-01 4.91374493e-01
8.83611739e-01 3.02952044e-02 -5.58628105e-02 -1.08758971e-01
3.06178182e-01 2.59246051e-01 7.97082067e-01 8.83071721e-01
-7.25288808e-01 5.33571601e-01 6.77402735e+00 4.74341035e-01
-8.14255834e-01 2.03172311e-01 -1.63105473e-01 1.91837445e-01
-1.64869830e-01 4.46178079e-01 4.93312590e-02 1.08005568e-01
8.53780806e-01 -6.45525396e-01 9.20964181e-01 4.60898966e-01
5.46677828e-01 -1.40447393e-01 -6.02609813e-01 9.38474119e-01
-1.58126310e-01 -1.30176270e+00 -2.51625776e-01 3.88336807e-01
8.98135841e-01 -1.17518187e-01 -3.37448567e-01 -6.01222098e-01
5.28954804e-01 -6.67906225e-01 1.99390024e-01 3.19775730e-01
3.84353071e-01 -1.05893254e+00 2.38421336e-01 4.43130583e-01
-1.81841266e+00 7.89441913e-02 -1.06102765e-01 -2.88261771e-01
1.59943312e-01 8.84055912e-01 -5.23561478e-01 8.16322923e-01
6.68465912e-01 5.95341980e-01 -3.82188946e-01 9.34548140e-01
1.23128116e-01 8.52124095e-01 -7.78945029e-01 2.96695381e-01
2.63030261e-01 -7.84020960e-01 8.70361686e-01 9.50457871e-01
4.24520552e-01 2.83369452e-01 5.55231035e-01 7.34690428e-01
-9.84832495e-02 2.29037121e-01 -4.40806061e-01 8.21557716e-02
5.73506832e-01 1.46912098e+00 -1.29179609e+00 -1.33531570e-01
-3.36420894e-01 1.09973300e+00 1.58198833e-01 7.05095828e-01
-3.70838732e-01 -5.70708156e-01 2.86743551e-01 1.84880480e-01
5.18746495e-01 -8.00689757e-01 2.41786372e-02 -9.52780902e-01
2.57474603e-03 -5.38869560e-01 4.25165236e-01 -2.82092243e-01
-9.69331682e-01 1.83999971e-01 -9.70954821e-02 -9.15616691e-01
-2.51907825e-01 -5.29688373e-02 -9.03035045e-01 5.49978495e-01
-1.15086472e+00 -9.24022436e-01 -3.80473375e-01 1.15881395e+00
-3.59706849e-01 5.54548502e-01 7.05193400e-01 3.22854549e-01
-5.33871412e-01 8.01478550e-02 4.01640803e-01 2.70676106e-01
3.37752491e-01 -1.19324183e+00 -1.71897650e-01 1.21535921e+00
-1.47829938e-03 7.76279688e-01 8.34660828e-01 -6.21639550e-01
-1.99280941e+00 -8.80980313e-01 7.32795298e-01 9.02713984e-02
9.68175173e-01 -3.47739458e-01 -5.05645514e-01 6.45066202e-01
3.27787399e-01 -1.41736388e-01 7.38134205e-01 -2.44995341e-01
-1.37536362e-01 -1.24174297e-01 -1.19565010e+00 3.24974626e-01
9.83112931e-01 -4.68556970e-01 4.07812968e-02 4.49528307e-01
2.90155888e-01 -2.84591522e-02 -1.05070376e+00 1.02874778e-01
1.50524229e-01 -8.92935753e-01 8.31602573e-01 1.68571174e-01
-2.84008741e-01 -7.56518304e-01 -1.55622542e-01 -1.08150911e+00
-5.07680058e-01 -1.39244330e+00 -1.33493856e-01 9.73882496e-01
2.21460283e-01 -7.38758504e-01 7.31281519e-01 2.46412918e-01
2.60590523e-01 -2.32423022e-01 -9.70111728e-01 -6.71426117e-01
-2.64531761e-01 -4.05417025e-01 6.66237399e-02 1.18890345e+00
4.95687187e-01 1.66192666e-01 -1.89917423e-02 8.90595436e-01
1.32987416e+00 4.18132156e-01 6.58654690e-01 -1.35077488e+00
-5.12742460e-01 -2.87796944e-01 -5.57145774e-01 -8.36325824e-01
8.43852162e-02 -1.13809025e+00 -8.59954879e-02 -1.60716236e+00
2.03726515e-01 -3.13187182e-01 1.18733525e-01 2.19572023e-01
2.04406470e-01 5.97442985e-01 1.15382493e-01 3.77901018e-01
-6.01683080e-01 7.79292583e-02 6.87728882e-01 9.15410295e-02
-2.70851165e-01 -1.02297045e-01 -4.36290383e-01 6.87602401e-01
5.94650149e-01 -6.57715976e-01 -3.89718473e-01 -1.87740549e-01
4.70368713e-01 1.54893562e-01 4.72207069e-01 -1.08534420e+00
9.36900020e-01 8.02274719e-02 -1.23113513e-01 -4.47940081e-01
8.84054974e-02 -9.16288435e-01 3.74767393e-01 8.30492198e-01
2.12533474e-01 2.76047617e-01 -6.54930472e-02 9.78808165e-01
1.20700449e-02 1.63167194e-01 6.63211405e-01 9.27409828e-02
-3.80386353e-01 1.86275259e-01 -8.16189289e-01 -1.93234041e-01
1.13939142e+00 -7.37540945e-02 -5.03036492e-02 -7.26901591e-01
-9.10551071e-01 2.51953930e-01 7.33683467e-01 -4.75722522e-01
4.96354848e-01 -1.16665459e+00 -6.51772141e-01 1.95306204e-02
-4.47767109e-01 -1.74192116e-01 9.93787646e-02 1.44150269e+00
-9.09502685e-01 -7.57481828e-02 3.45830113e-01 -6.29438460e-01
-1.34654081e+00 5.10873139e-01 3.21150750e-01 -1.18553862e-01
-7.87353337e-01 4.51959550e-01 -3.54418129e-01 -1.79626778e-01
-1.00142136e-01 2.03109875e-01 3.48191082e-01 -1.23601682e-01
2.19618350e-01 9.03604627e-01 -1.21225446e-01 -8.75144243e-01
-6.01452172e-01 1.10045338e+00 6.30856037e-01 -4.65255708e-01
1.33173323e+00 -5.63733399e-01 -8.77193332e-01 2.50587729e-03
1.27459967e+00 5.18980086e-01 -7.92637467e-01 -6.50648102e-02
4.84047756e-02 -9.97921526e-02 2.90757805e-01 5.20048924e-02
-1.23674393e+00 5.23883939e-01 1.88054577e-01 7.39740670e-01
1.44518912e+00 1.75875910e-02 7.77335525e-01 4.99940634e-01
6.05288386e-01 -9.04481053e-01 -3.15238327e-01 6.48949295e-02
1.93718880e-01 -4.61740643e-01 1.97357833e-01 -9.93135095e-01
9.39502493e-02 1.12913525e+00 -1.56448096e-01 -5.91471910e-01
1.05833125e+00 5.17167568e-01 -6.79295063e-02 -4.73252863e-01
-4.11193877e-01 -2.25950345e-01 -8.68985876e-02 6.15124345e-01
6.25281259e-02 -1.41970180e-02 -4.22461241e-01 -2.34297384e-02
-9.41089094e-02 -4.54929799e-01 7.07003653e-01 9.17706370e-01
-4.76356804e-01 -1.09167278e+00 -6.26402378e-01 1.13009347e-03
-1.72414243e-01 -3.10874898e-02 -7.51461089e-01 5.36260724e-01
-1.91464260e-01 1.43894517e+00 -1.07493334e-01 -4.09102201e-01
-1.76289394e-01 4.18889225e-02 4.51873779e-01 -4.43268478e-01
-2.62800336e-01 2.44652405e-01 -2.50467062e-01 -6.69318020e-01
-8.83228362e-01 -4.94967014e-01 -1.75393927e+00 -9.12670076e-01
-2.32199520e-01 5.93163848e-01 5.48192203e-01 8.47196817e-01
4.90114331e-01 2.03352079e-01 9.86971617e-01 -9.72650290e-01
-6.71487860e-03 -4.73730862e-01 -8.61625433e-01 2.13170037e-01
3.46372575e-01 -3.61083955e-01 -7.43128479e-01 1.42500982e-01] | [7.076850891113281, 5.071138381958008] |
d9f087a8-295b-42af-af8d-4a85433857bb | r-2-range-regularization-for-model | 2303.08253 | null | https://arxiv.org/abs/2303.08253v1 | https://arxiv.org/pdf/2303.08253v1.pdf | R^2: Range Regularization for Model Compression and Quantization | Model parameter regularization is a widely used technique to improve generalization, but also can be used to shape the weight distributions for various purposes. In this work, we shed light on how weight regularization can assist model quantization and compression techniques, and then propose range regularization (R^2) to further boost the quality of model optimization by focusing on the outlier prevention. By effectively regulating the minimum and maximum weight values from a distribution, we mold the overall distribution into a tight shape so that model compression and quantization techniques can better utilize their limited numeric representation powers. We introduce L-inf regularization, its extension margin regularization and a new soft-min-max regularization to be used as a regularization loss during full-precision model training. Coupled with state-of-the-art quantization and compression techniques, models trained with R^2 perform better on an average, specifically at lower bit weights with 16x compression ratio. We also demonstrate that R^2 helps parameter constrained models like MobileNetV1 achieve significant improvement of around 8% for 2 bit quantization and 7% for 1 bit compression. | ['Saurabh Adya', 'Minsik Cho', 'Srijan Mishra', 'Chungkuk Yoo', 'Arnav Kundu'] | 2023-03-14 | null | null | null | null | ['model-compression'] | ['methodology'] | [ 2.89800018e-01 -4.36348841e-02 -1.07431984e+00 -5.60535312e-01
-8.46550703e-01 -2.10380912e-01 2.27867514e-01 2.59541154e-01
-5.98090887e-01 4.21407074e-01 3.75954092e-01 -5.80351889e-01
-3.35377790e-02 -6.68116510e-01 -7.92795241e-01 -4.38516945e-01
-9.88555104e-02 1.77216396e-01 1.34030178e-01 -2.32673779e-01
3.72846395e-01 3.24264735e-01 -1.39076793e+00 4.26930606e-01
1.00212383e+00 1.24460399e+00 1.06417373e-01 5.56629658e-01
1.05266040e-02 7.50046074e-01 -3.84358019e-01 -3.96741271e-01
4.55695540e-01 3.22720036e-02 -5.37610888e-01 -1.52649567e-01
4.23975259e-01 -5.53606927e-01 -5.73049843e-01 1.10067117e+00
3.43116999e-01 2.75182664e-01 8.39038372e-01 -7.60512292e-01
-6.47333205e-01 9.06726062e-01 -5.79259813e-01 3.42987508e-01
-5.54146320e-02 1.94590077e-01 1.09170353e+00 -7.30832040e-01
1.97984233e-01 1.37125015e+00 7.20842421e-01 5.59866488e-01
-1.31146193e+00 -8.33178103e-01 8.80306363e-02 2.82203048e-01
-1.68598318e+00 -7.82260180e-01 4.82717544e-01 -8.08450803e-02
1.30222535e+00 2.18785495e-01 2.75358349e-01 7.26130247e-01
1.96496546e-01 6.18494749e-01 4.19046402e-01 -5.08191347e-01
3.74677122e-01 4.26127464e-02 7.96843097e-02 5.09135008e-01
3.85835141e-01 4.13311236e-02 -6.21447086e-01 -1.45446867e-01
7.28516638e-01 -1.13870911e-01 -3.29702497e-01 -3.99801731e-02
-6.40678108e-01 1.09588349e+00 5.18508434e-01 2.35151798e-02
-1.16638102e-01 8.05172980e-01 5.51773965e-01 2.75236905e-01
4.57849711e-01 2.87049592e-01 -3.77330244e-01 -4.65781331e-01
-1.31772137e+00 2.03262776e-01 5.50455153e-01 1.08511341e+00
6.31192505e-01 3.88091147e-01 -5.10736406e-01 1.19527900e+00
8.29241216e-01 4.77576524e-01 6.18151546e-01 -1.17893493e+00
8.86990309e-01 3.57749879e-01 -1.29232809e-01 -7.98453212e-01
-6.55945614e-02 -3.34772944e-01 -8.45988572e-01 -1.62194297e-02
7.31626004e-02 2.27011859e-01 -1.19542491e+00 1.70565403e+00
-3.00644904e-01 4.26289365e-02 -5.90645559e-02 7.72505999e-01
2.34320387e-01 6.94826245e-01 2.61356086e-01 -1.98122442e-01
1.23502445e+00 -8.93914044e-01 -8.00566971e-01 -3.86511683e-01
1.07020617e+00 -6.52908444e-01 1.46834815e+00 4.76235241e-01
-1.26969588e+00 -5.81746399e-01 -1.36005628e+00 -3.59869003e-01
-1.11954868e-01 -3.89588103e-02 5.79586089e-01 9.12608564e-01
-8.57710660e-01 9.59022105e-01 -1.09395075e+00 1.37941226e-01
7.58880675e-01 5.29249609e-01 9.44216847e-02 -2.89322436e-01
-1.11486411e+00 7.68846571e-01 5.09111643e-01 -3.14695209e-01
-4.04270589e-01 -8.87539029e-01 -8.87980402e-01 4.30861652e-01
9.73134115e-02 -3.41483653e-01 1.02997303e+00 -2.26605833e-01
-1.33373427e+00 5.64308941e-01 -2.80884415e-01 -1.18561733e+00
1.47936210e-01 -3.76641482e-01 -5.00439584e-01 1.53963178e-01
-3.39319229e-01 8.54710162e-01 9.55617011e-01 -9.62724388e-01
-2.12966383e-01 -2.60734022e-01 -3.81601691e-01 5.82997762e-02
-9.55639899e-01 -2.41719335e-01 -8.24672580e-01 -9.21164215e-01
2.58092612e-01 -6.44620419e-01 -3.00955534e-01 2.95896590e-01
-1.93631947e-01 3.61482403e-03 8.32882226e-01 -6.01847053e-01
1.99541938e+00 -2.11208153e+00 -1.08083963e-01 5.83052635e-01
1.49195164e-01 6.45580947e-01 -1.43777668e-01 1.10421553e-02
3.31366718e-01 6.18040323e-01 -3.37612659e-01 -7.33310401e-01
1.44809052e-01 7.19134808e-01 -4.11386758e-01 3.50390702e-01
1.04399949e-01 7.90736318e-01 -6.28422797e-01 -3.91955823e-01
1.23246796e-01 8.05508077e-01 -1.31020629e+00 -1.40384296e-02
-3.74253094e-01 -1.02024078e-01 -2.07128718e-01 5.45942307e-01
8.19616079e-01 -3.66058588e-01 1.51445180e-01 -3.79411966e-01
3.90241563e-01 3.82643104e-01 -1.16407073e+00 1.66167462e+00
-5.89785516e-01 6.27320409e-01 8.95794928e-02 -9.57320452e-01
1.05888450e+00 -8.84536654e-02 2.73804247e-01 -7.82951891e-01
1.93699554e-01 3.11290234e-01 7.90481046e-02 -1.68483421e-01
9.13613081e-01 8.31929296e-02 2.55767256e-01 1.84099227e-01
-9.29861441e-02 -3.03066224e-01 3.63349557e-01 1.23035222e-01
6.93286061e-01 -2.64327824e-01 3.73143405e-02 -2.49726608e-01
4.17870730e-01 -6.52785540e-01 5.70045888e-01 6.93268895e-01
-7.06050843e-02 5.79591811e-01 3.73010904e-01 -1.84254825e-01
-1.30051363e+00 -7.12683260e-01 -5.42667031e-01 1.25642371e+00
-6.89539826e-03 -8.69903684e-01 -5.27761817e-01 -3.02490860e-01
3.36204022e-01 1.06378114e+00 -2.85644233e-01 -6.99300289e-01
-6.05084658e-01 -8.81883562e-01 8.57653320e-01 7.98732519e-01
5.86907864e-01 -5.54413497e-01 -3.21786523e-01 1.82362914e-01
5.81394657e-02 -1.00778687e+00 -7.91325748e-01 6.65311337e-01
-1.20871437e+00 -4.69462544e-01 -2.82900065e-01 -4.72701669e-01
4.10109758e-01 -1.62802413e-01 1.15822649e+00 4.47437793e-01
1.11155704e-01 -1.17945537e-01 -3.66294265e-01 -2.44488224e-01
-6.65066183e-01 3.42530966e-01 3.37088972e-01 -4.13201094e-01
4.47676629e-01 -8.55945587e-01 -6.83549106e-01 3.26539785e-01
-1.21108079e+00 -4.41900194e-01 3.93105268e-01 8.23970735e-01
8.97131741e-01 -1.97563455e-01 4.33321118e-01 -8.09516788e-01
8.07880759e-01 -2.72815466e-01 -4.52262461e-01 5.03774872e-03
-1.23862398e+00 4.96391177e-01 6.81945920e-01 -5.10521710e-01
-5.17014265e-01 -4.27122712e-01 -4.41703469e-01 -9.35833275e-01
4.87763971e-01 4.77627128e-01 5.34447208e-02 -2.51398683e-01
9.73510742e-01 7.14315102e-03 7.55743161e-02 -5.64251721e-01
5.85881293e-01 8.55055094e-01 3.90643358e-01 -6.38526201e-01
5.85811675e-01 1.86537638e-01 -2.94461586e-02 -6.81300700e-01
-6.60725892e-01 -3.33523989e-01 -4.58935767e-01 4.20883447e-01
6.05307400e-01 -1.01298916e+00 -4.52477008e-01 -1.09607823e-01
-7.35041916e-01 -6.46667063e-01 -6.14873409e-01 3.81410420e-01
-6.36332035e-01 5.01322627e-01 -6.40455902e-01 -6.84955597e-01
-5.32878578e-01 -1.27845812e+00 9.41742480e-01 -8.79769772e-02
-1.31159678e-01 -1.08526468e+00 -3.05664122e-01 2.00648203e-01
7.85008252e-01 -2.88528621e-01 1.15257227e+00 -5.99027693e-01
-3.02919060e-01 -1.23747915e-01 -3.73450309e-01 7.70733595e-01
-7.50292316e-02 -1.70895413e-01 -8.39878917e-01 -6.09123647e-01
-4.83818464e-02 -4.95158404e-01 1.32571244e+00 4.50938672e-01
1.77333975e+00 -4.32897985e-01 -2.00083867e-01 1.29223311e+00
1.30134034e+00 -7.93914646e-02 9.10371006e-01 9.70947817e-02
6.89682484e-01 -1.86194673e-01 3.49369824e-01 7.04953015e-01
-2.39255875e-02 6.94210172e-01 4.40793067e-01 1.91898644e-01
-2.72714376e-01 -3.00072521e-01 2.18004927e-01 1.04031658e+00
7.36980420e-03 -3.23762447e-01 -7.75917768e-01 1.13112465e-01
-1.48348606e+00 -5.49771428e-01 3.17017794e-01 2.38831449e+00
1.19962823e+00 6.59158826e-01 -1.81179225e-01 3.95048708e-01
3.76898885e-01 4.12979305e-01 -6.95742011e-01 -6.38507307e-01
-2.62123011e-02 3.06144744e-01 1.07461214e+00 6.79262996e-01
-9.57091749e-01 9.61777031e-01 7.12197638e+00 1.48456967e+00
-1.08695018e+00 -1.33918554e-01 8.66256773e-01 -3.33244383e-01
-4.75637794e-01 -1.91692948e-01 -1.14050341e+00 6.83865964e-01
1.45743537e+00 -3.84318866e-02 6.83566809e-01 1.04522026e+00
2.56299764e-01 2.24274427e-01 -1.19303465e+00 1.16456103e+00
-1.30619943e-01 -1.53117907e+00 5.58136523e-01 3.58679175e-01
7.56676972e-01 1.67929366e-01 4.37604845e-01 5.21878779e-01
-1.10960238e-01 -1.39653385e+00 5.20701170e-01 2.87619680e-01
1.24412835e+00 -9.16380823e-01 6.18242562e-01 1.98316485e-01
-1.25703657e+00 -4.33207542e-01 -9.88976002e-01 9.42663774e-02
3.11706036e-01 5.43808699e-01 -6.58916473e-01 5.71653657e-02
3.83546263e-01 7.40918815e-01 -5.61609566e-01 8.50556135e-01
4.10852619e-02 9.13378835e-01 -6.08923614e-01 2.42347613e-01
1.98994830e-01 -1.67330518e-01 2.79996663e-01 1.38029504e+00
2.46147931e-01 -3.95045988e-02 1.95457846e-01 7.78448939e-01
-5.62216282e-01 -1.43057527e-02 -1.75804988e-01 9.18736588e-03
9.03156161e-01 5.81643283e-01 -3.63096446e-01 -3.79344016e-01
-1.75087616e-01 6.84049964e-01 2.99958169e-01 2.76377439e-01
-9.51868594e-01 -4.03971434e-01 7.53782749e-01 3.64930868e-01
5.89089751e-01 -3.42603058e-01 -5.88922143e-01 -1.09292936e+00
-7.38162920e-02 -8.81079972e-01 1.39368877e-01 -2.72992462e-01
-9.38869417e-01 4.05422091e-01 6.62242472e-02 -1.21199894e+00
-3.25034499e-01 -6.55053198e-01 -3.56416076e-01 7.29263306e-01
-1.83752418e+00 -6.05700314e-01 -1.09028727e-01 3.91099870e-01
4.89358753e-01 -2.81302959e-01 8.27947140e-01 6.07822716e-01
-3.75777870e-01 1.40825427e+00 4.88551140e-01 -1.71018034e-01
6.24562740e-01 -8.89452040e-01 4.32012886e-01 6.23202741e-01
1.50214732e-01 8.18708777e-01 5.32884181e-01 -5.22337794e-01
-1.29399109e+00 -1.32845736e+00 4.18101281e-01 -6.27113953e-02
4.77586359e-01 -2.67995656e-01 -1.28348064e+00 5.72303891e-01
-2.19306961e-01 3.44707996e-01 6.38453603e-01 -6.98074996e-02
-6.53587341e-01 -3.09756193e-02 -1.47897029e+00 5.80176473e-01
1.01815581e+00 -5.91440082e-01 -1.12342976e-01 4.88817036e-01
1.09965277e+00 -5.27837455e-01 -1.15316093e+00 5.44271588e-01
3.87013882e-01 -7.88850009e-01 1.21796703e+00 -4.43169773e-01
3.96660209e-01 4.72410768e-02 -8.06119442e-01 -8.97484243e-01
-2.64118224e-01 -7.48531342e-01 -9.00461257e-01 9.35373306e-01
4.30431217e-01 -3.42891514e-01 1.09403610e+00 6.43479288e-01
-2.55376548e-01 -1.17039740e+00 -1.09529400e+00 -9.24596012e-01
4.91158843e-01 -9.00163412e-01 4.98870820e-01 3.86311978e-01
-8.34582224e-02 -1.38836190e-01 -4.65344548e-01 -1.12816304e-01
4.47260499e-01 -6.91174686e-01 3.54099631e-01 -7.95951128e-01
-4.00563270e-01 -5.67639649e-01 -4.48160112e-01 -1.83935547e+00
3.93848903e-02 -1.04550803e+00 -1.73863247e-01 -9.73872721e-01
-1.04686402e-01 -6.66866422e-01 -4.04530883e-01 3.97645026e-01
-2.26356853e-02 5.24493158e-01 2.61588991e-01 4.38936859e-01
-5.32979548e-01 8.93363595e-01 1.17217803e+00 -1.62151426e-01
-4.36855048e-01 -1.08076677e-01 -7.64098108e-01 7.26196289e-01
6.67223871e-01 -3.25331658e-01 -5.97348809e-01 -6.61205709e-01
7.85602555e-02 -3.31827015e-01 -1.69109553e-01 -1.25626576e+00
1.68226317e-01 5.88338785e-02 3.73578042e-01 -4.89825308e-01
4.68811631e-01 -7.11768687e-01 -4.72606301e-01 3.52046162e-01
-6.20763540e-01 -4.91928495e-02 3.58752608e-01 6.02754653e-01
-1.02071285e-01 -4.07706797e-01 1.06302464e+00 1.92400843e-01
-4.22715336e-01 2.96809614e-01 -2.35106394e-01 2.41182253e-01
3.99272412e-01 -1.84409782e-01 -1.01003997e-01 -4.98698682e-01
-7.27999270e-01 3.66297513e-01 4.54337746e-01 1.35283738e-01
7.94222772e-01 -1.45311487e+00 -4.72842246e-01 6.17785275e-01
-1.07629960e-02 -2.44281162e-02 -6.34521917e-02 4.59534228e-01
-5.47905326e-01 4.25195456e-01 1.20162658e-01 -7.11025357e-01
-1.02104056e+00 3.36966753e-01 1.72111213e-01 -5.16431034e-01
-5.08233190e-01 1.01769948e+00 -4.58054870e-01 -3.14660333e-02
7.89993703e-01 -7.81768739e-01 -5.53386984e-03 -2.24835530e-01
6.10886991e-01 4.07785237e-01 2.86820859e-01 -4.76010740e-01
-1.49046257e-01 5.93876362e-01 -3.62936825e-01 9.80620757e-02
1.15328622e+00 -1.35443181e-01 1.94659695e-01 1.27147779e-01
1.53852534e+00 -1.41941920e-01 -1.46956265e+00 -2.37950012e-01
-7.02843070e-02 -5.82230270e-01 3.48096132e-01 -4.87071544e-01
-1.12319958e+00 9.46132362e-01 8.07119608e-01 1.95996776e-01
1.05679393e+00 -3.17871362e-01 9.58081305e-01 5.54932892e-01
1.66774437e-01 -1.26919425e+00 1.71531603e-01 7.62203932e-01
6.62720144e-01 -1.11234903e+00 3.06012571e-01 -3.38667899e-01
-4.47347641e-01 1.08965790e+00 3.90527934e-01 -2.19198063e-01
7.87163973e-01 4.47930753e-01 -1.86078042e-01 2.08226562e-01
-6.72427356e-01 2.94445276e-01 4.79638636e-01 5.40607572e-01
6.78335428e-01 -2.30875686e-01 -2.13599876e-01 5.20297527e-01
-3.93600136e-01 -6.67842552e-02 2.14710116e-01 7.97308862e-01
-7.47266829e-01 -1.20386839e+00 -1.90998256e-01 9.41900194e-01
-6.11596346e-01 -4.73626822e-01 4.69667584e-01 5.03208280e-01
7.77679384e-02 8.17153931e-01 3.74842465e-01 -6.61470175e-01
3.37382704e-01 1.20674133e-01 2.99587309e-01 -5.19730985e-01
-4.17404860e-01 1.43495247e-01 -9.90627706e-02 -6.91367567e-01
1.24576613e-02 -1.59780532e-01 -1.38003182e+00 -6.34609640e-01
-5.14115453e-01 8.67787302e-02 5.46946764e-01 6.54722095e-01
4.11317319e-01 5.88465631e-01 3.37569982e-01 -9.66245294e-01
-1.19971335e+00 -9.86877680e-01 -6.07019067e-01 4.10748512e-01
3.33633989e-01 -5.56919873e-01 -6.50604129e-01 1.15749128e-01] | [8.65261459350586, 3.264244794845581] |
19bff1ff-8598-4adf-8c32-031a3394de67 | attribute-value-generation-from-product-title | null | null | https://aclanthology.org/2021.ecnlp-1.2 | https://aclanthology.org/2021.ecnlp-1.2.pdf | Attribute Value Generation from Product Title using Language Models | Identifying the value of product attribute is essential for many e-commerce functions such as product search and product recommendations. Therefore, identifying attribute values from unstructured product descriptions is a critical undertaking for any e-commerce retailer. What makes this problem challenging is the diversity of product types and their attributes and values. Existing methods have typically employed multiple types of machine learning models, each of which handles specific product types or attribute classes. This has limited their scalability and generalization for large scale real world e-commerce applications. Previous approaches for this task have formulated the attribute value extraction as a Named Entity Recognition (NER) task or a Question Answering (QA) task. In this paper we have presented a generative approach to the attribute value extraction problem using language models. We leverage the large-scale pretraining of the GPT-2 and the T5 text-to-text transformer to create fine-tuned models that can effectively perform this task. We show that a single general model is very effective for this task over a broad set of product attribute values with the open world assumption. Our approach achieves state-of-the-art performance for different attribute classes, which has previously required a diverse set of models. | ['Manish Pandey', 'Pawan Goyal', 'Kalyani Roy'] | null | null | null | null | acl-ecnlp-2021-8 | ['attribute-value-extraction'] | ['natural-language-processing'] | [ 1.70733705e-01 1.36340320e-01 -5.54213226e-01 -8.51016998e-01
-1.05177510e+00 -9.32244480e-01 5.00673950e-01 2.74321586e-01
-2.97667474e-01 4.96900052e-01 3.40826623e-02 -4.60038185e-01
-1.25599101e-01 -1.14249861e+00 -4.43900645e-01 -4.86864477e-01
4.70103323e-02 1.10757124e+00 -9.55183804e-03 -5.70182383e-01
5.18794730e-02 2.44690254e-01 -1.53106177e+00 5.88500321e-01
7.12518930e-01 1.37634695e+00 -1.45250373e-02 5.36118269e-01
-8.59742999e-01 5.66064119e-01 -5.76654434e-01 -1.12481165e+00
4.30403948e-01 -1.71728194e-01 -9.30756629e-01 5.54770939e-02
1.57428056e-01 -3.33532915e-02 4.36141938e-01 1.11844850e+00
3.51453960e-01 1.90150738e-01 7.96477556e-01 -1.48202252e+00
-8.09258342e-01 8.93172264e-01 -2.38319471e-01 -2.54040271e-01
2.55464703e-01 -1.61191255e-01 1.66817629e+00 -5.32075942e-01
4.53213930e-01 1.09649372e+00 7.24740624e-01 3.20610017e-01
-1.30660796e+00 -4.63143438e-01 1.38199091e-01 -1.93345491e-02
-1.33352911e+00 -8.52933601e-02 6.86565399e-01 -2.58190662e-01
1.30079556e+00 9.68481749e-02 5.84348917e-01 8.32043111e-01
6.43489286e-02 8.11794400e-01 9.25649464e-01 -2.23620549e-01
2.11980268e-01 5.14484763e-01 9.83373001e-02 6.53087422e-02
1.06797367e-01 -3.32688898e-01 -7.57958889e-02 -2.73623019e-01
6.51958406e-01 -4.14246581e-02 3.43952894e-01 -3.94725710e-01
-1.16340590e+00 1.23944020e+00 1.59251198e-01 1.37300402e-01
-6.10474646e-01 -9.49724764e-02 4.04051036e-01 6.77243590e-01
5.25633335e-01 8.15567970e-01 -1.27156889e+00 -2.02204615e-01
-5.16285360e-01 3.39347243e-01 1.55017829e+00 1.42008209e+00
6.16473079e-01 -2.75555164e-01 4.15070914e-02 9.97964144e-01
3.00592780e-01 4.11703825e-01 2.63126552e-01 -5.04226267e-01
5.56099296e-01 7.63983309e-01 1.50906801e-01 -6.67078376e-01
-2.36355990e-01 -4.51374650e-01 -4.70406175e-01 -1.55623406e-01
5.67955494e-01 -1.27189711e-01 -1.12308216e+00 1.57800996e+00
4.79057521e-01 -4.27233517e-01 3.22075248e-01 6.02596641e-01
7.83931732e-01 7.17751622e-01 6.00780368e-01 2.44361565e-01
1.73030329e+00 -6.92567348e-01 -6.76156402e-01 -3.32271367e-01
9.22086895e-01 -1.00804031e+00 1.03942871e+00 2.74255931e-01
-8.23510349e-01 -3.41136605e-01 -8.08726192e-01 -2.60651439e-01
-1.02469766e+00 -1.70577109e-01 1.09695804e+00 8.66144300e-01
-4.45812285e-01 3.97048682e-01 -2.45712861e-01 -3.80671710e-01
3.78403574e-01 7.48479366e-01 -3.26452941e-01 -1.10026844e-01
-1.48719025e+00 8.84305716e-01 5.78870237e-01 -1.52164131e-01
-2.97260046e-01 -6.18599594e-01 -1.03956771e+00 3.74339849e-01
7.06476033e-01 -8.75207901e-01 1.53376782e+00 -9.95673120e-01
-1.30835426e+00 7.10231006e-01 1.15926087e-01 -3.86031657e-01
-7.90013075e-02 -1.35559961e-01 -7.50293434e-01 -4.17330742e-01
1.82596564e-01 5.54242611e-01 6.10878468e-01 -9.72340107e-01
-8.73983860e-01 -5.64817607e-01 1.60463303e-01 1.69926688e-01
-1.58297241e-01 2.51409739e-01 -2.86415875e-01 -6.13962829e-01
9.63865295e-02 -9.57790792e-01 -2.75796801e-01 -5.46007454e-01
-2.51347810e-01 -4.74964529e-01 2.33718619e-01 -4.07502681e-01
1.03772461e+00 -1.95962036e+00 -2.22810686e-01 4.04325843e-01
-1.20873556e-01 5.27749099e-02 -2.91388094e-01 5.42615116e-01
4.60256487e-02 2.84074932e-01 3.77067178e-02 6.37485459e-02
4.39898133e-01 4.58575875e-01 -2.66392499e-01 -1.51165038e-01
2.90807813e-01 1.23127854e+00 -6.89472377e-01 -4.35243785e-01
-1.21173397e-01 5.63950479e-01 -4.56831127e-01 2.66189784e-01
-7.11639941e-01 -4.84117419e-02 -8.50557625e-01 8.63886476e-01
5.52001536e-01 -2.96654642e-01 2.50207871e-01 -3.10147911e-01
3.55028093e-01 4.16009188e-01 -1.30734444e+00 1.53468454e+00
-6.54734313e-01 -1.27050027e-01 -1.50292171e-02 -6.99746192e-01
1.11714554e+00 2.21447036e-01 7.64021695e-01 -6.81961894e-01
4.24178571e-01 4.16270584e-01 2.35298514e-01 -3.82840037e-01
6.23440564e-01 -4.69484657e-01 -5.30880272e-01 2.44444191e-01
2.78454244e-01 -3.39460045e-01 2.60980338e-01 -7.53847435e-02
7.96238303e-01 2.14083970e-01 4.74067837e-01 -2.44061559e-01
4.55573887e-01 3.13654482e-01 5.96576452e-01 6.10974669e-01
2.26010829e-02 4.21450824e-01 4.90616918e-01 -5.99708200e-01
-1.20740700e+00 -7.99263775e-01 -2.51897007e-01 1.55157053e+00
2.71329726e-03 -4.75309551e-01 -5.87901115e-01 -9.31255400e-01
2.45185375e-01 9.83801067e-01 -4.08301741e-01 -5.71970455e-02
-3.60881269e-01 -8.45528841e-01 1.31819680e-01 5.31386793e-01
4.19229150e-01 -1.17028022e+00 -1.58884302e-01 5.29178143e-01
-1.38419107e-01 -1.35526168e+00 -3.78547758e-01 5.15563011e-01
-6.17630899e-01 -8.36681008e-01 -4.11338896e-01 -1.02571225e+00
3.22341740e-01 -2.70300955e-01 1.74516213e+00 -4.49015379e-01
2.01326072e-01 1.49947315e-01 -5.72788417e-01 -8.18959534e-01
-5.57850361e-01 7.05889046e-01 -3.84657353e-01 2.43939698e-01
1.21072149e+00 -2.41201907e-01 -3.56854141e-01 6.11050785e-01
-8.11191678e-01 -3.19044501e-01 1.04069912e+00 7.64658928e-01
8.70532274e-01 3.99478108e-01 8.46851766e-01 -1.52371407e+00
8.08324695e-01 -7.42736638e-01 -3.74075830e-01 2.81391740e-01
-8.57615829e-01 3.03038388e-01 6.95924938e-01 -5.96917033e-01
-1.07499397e+00 3.05427849e-01 -7.14977145e-01 3.16973388e-01
-3.97958815e-01 6.97925985e-01 -7.73821354e-01 2.48942018e-01
3.17623377e-01 -3.78905386e-02 -1.82625905e-01 -6.85205817e-01
5.63484848e-01 9.20131326e-01 3.29361588e-01 -4.93468612e-01
4.39832240e-01 -7.14803189e-02 -1.34896832e-02 -4.73675311e-01
-1.19647944e+00 -7.89053798e-01 -7.16780126e-01 4.21552926e-01
8.50205421e-01 -8.07034791e-01 -7.22784042e-01 2.63245195e-01
-9.10392404e-01 6.93042809e-03 -4.71390218e-01 3.49799365e-01
-5.32063425e-01 -1.61529675e-01 -5.77059686e-01 -5.22097051e-01
-4.28310454e-01 -1.03627527e+00 1.17236674e+00 2.64617484e-02
-3.02011460e-01 -1.09924686e+00 -1.24067582e-01 4.87673908e-01
6.84113562e-01 3.34307924e-02 1.31960130e+00 -1.37532914e+00
-3.38728398e-01 -5.13891220e-01 -2.17484031e-02 3.23479712e-01
1.84049711e-01 -6.23233318e-01 -7.86079168e-01 5.10735363e-02
-2.68798843e-02 -2.31751099e-01 6.04953825e-01 2.16525599e-01
1.02229702e+00 -4.48336929e-01 -2.13183984e-01 5.62644541e-01
1.29110336e+00 3.78689677e-01 6.94100797e-01 4.49122101e-01
6.35004282e-01 8.76348674e-01 9.59998250e-01 2.22275227e-01
6.02210402e-01 7.99424648e-01 2.86109120e-01 -2.94196755e-01
2.64631271e-01 -3.20323855e-01 9.25918389e-03 3.27682883e-01
1.58561617e-01 -3.20658982e-01 -6.82645798e-01 7.06946194e-01
-1.84896553e+00 -8.52137148e-01 -4.35157120e-02 2.13522291e+00
1.00878739e+00 1.80362076e-01 2.81879067e-01 -1.21625267e-01
5.09911060e-01 -2.00052977e-01 -6.56126916e-01 -6.84860051e-01
2.19473504e-02 1.23744704e-01 6.03888094e-01 3.41420434e-02
-1.34026706e+00 1.01313162e+00 6.28596878e+00 8.46171498e-01
-6.00280643e-01 1.39146652e-02 7.04204559e-01 1.78889543e-01
-4.86164570e-01 -1.83281466e-01 -1.14457262e+00 3.82081240e-01
1.19506061e+00 -9.27454606e-02 1.92424953e-01 1.13413250e+00
-4.16802466e-01 2.35626385e-01 -1.28223217e+00 1.04054654e+00
-4.88991998e-02 -7.84236372e-01 2.01845407e-01 3.13266248e-01
5.48638761e-01 -1.92653820e-01 1.32951021e-01 6.35286868e-01
7.14278281e-01 -1.11485791e+00 4.30656701e-01 -7.55900070e-02
6.77639067e-01 -8.21069896e-01 1.11448693e+00 1.35821253e-01
-1.30102694e+00 -2.44712338e-01 -3.75570238e-01 2.25169092e-01
3.84464771e-01 5.08773267e-01 -8.94164026e-01 4.30488735e-01
6.42900944e-01 4.93214637e-01 -3.45485866e-01 7.09754229e-01
-8.74065757e-02 4.12646443e-01 -4.04073328e-01 -1.35870472e-01
2.59678990e-01 -2.80478179e-01 1.22454144e-01 1.04816270e+00
3.61652255e-01 3.66785303e-02 2.11037010e-01 8.41154337e-01
-3.09774965e-01 3.78925532e-01 -4.95164037e-01 -4.47238892e-01
3.11666429e-01 1.46569443e+00 -6.46285951e-01 -1.68766260e-01
-8.68200779e-01 6.94693744e-01 3.04132476e-02 9.43276659e-02
-7.06815183e-01 -6.58907473e-01 9.11511898e-01 2.41543502e-01
7.39042699e-01 -5.06845824e-02 -3.39416116e-01 -1.04891825e+00
-1.94563083e-02 -1.01304138e+00 7.10849881e-01 -3.76374692e-01
-1.79056466e+00 8.07352483e-01 -3.39232339e-03 -1.11475003e+00
-7.80795276e-01 -7.98688054e-01 6.38082949e-03 8.16762388e-01
-1.60294044e+00 -1.63876307e+00 1.26697376e-01 5.00474095e-01
4.50337589e-01 -2.84139335e-01 1.00336802e+00 3.68006527e-01
-5.13759442e-02 6.72099292e-01 3.51375826e-02 4.23795819e-01
5.55964053e-01 -1.44091058e+00 7.60721862e-01 3.12010437e-01
3.71531487e-01 7.82474160e-01 6.74944580e-01 -6.29252136e-01
-1.54598355e+00 -1.01775646e+00 1.39435828e+00 -8.23218286e-01
8.10117126e-01 -7.46985674e-01 -8.39961112e-01 7.96171427e-01
1.13358617e-01 -1.54727161e-01 1.18648529e+00 6.19410813e-01
-5.36180615e-01 -1.45180956e-01 -1.33830714e+00 1.26477212e-01
8.38764250e-01 -4.34033871e-01 -6.32398784e-01 1.79630429e-01
6.87793791e-01 -9.36001241e-02 -1.53133905e+00 1.71418518e-01
6.42386317e-01 -5.57801485e-01 1.01945555e+00 -9.42538917e-01
2.56085902e-01 -4.55717184e-02 -2.91274905e-01 -1.37042522e+00
-2.59369016e-01 -4.44967538e-01 1.54749025e-03 1.63280404e+00
9.80414748e-01 -7.61960149e-01 7.09524214e-01 1.36138737e+00
7.50332996e-02 -8.19663525e-01 -4.25665766e-01 -8.15673470e-01
1.72154859e-01 -3.69908065e-01 1.33326471e+00 8.62094998e-01
-2.78688580e-01 8.57114077e-01 -4.92189452e-02 -5.71635701e-02
4.22783405e-01 5.72244287e-01 5.97953737e-01 -1.48869824e+00
-4.40711111e-01 -2.83066779e-01 -4.92383897e-01 -9.74680603e-01
1.15408316e-01 -9.03382242e-01 -1.38660610e-01 -1.59209919e+00
1.02400012e-01 -7.94969738e-01 -3.29238802e-01 7.47843206e-01
4.37717810e-02 1.10412002e-01 1.67023987e-01 1.63250864e-01
-4.54396784e-01 2.49284759e-01 1.09974241e+00 -2.88908631e-01
-2.49697089e-01 3.64171296e-01 -1.12244952e+00 4.40087676e-01
8.87849450e-01 -6.63687348e-01 -5.03860414e-01 -2.68400133e-01
7.76930749e-01 -2.21136972e-01 -5.82139306e-02 -3.59980702e-01
-3.97485718e-02 -2.73979753e-01 3.31381112e-01 -3.33649069e-01
2.73554265e-01 -1.12524903e+00 9.11412910e-02 -1.70087993e-01
-4.61491704e-01 1.34245351e-01 3.75555120e-02 3.03909034e-01
-5.20221174e-01 -3.49178284e-01 3.65949184e-01 -3.68247271e-01
-1.07505059e+00 4.10001695e-01 3.00218035e-02 1.55183256e-01
9.53305542e-01 -2.06342980e-01 -1.33871481e-01 -3.77114713e-01
-7.56906450e-01 2.57545859e-01 2.88735360e-01 6.80593312e-01
1.81811333e-01 -1.24425578e+00 -7.59895623e-01 3.71673912e-01
4.68454212e-01 -1.24983296e-01 -1.91015497e-01 3.28725219e-01
1.79441031e-02 4.41096991e-01 -2.23183095e-01 -3.00636828e-01
-1.00932252e+00 7.95100868e-01 1.69873223e-01 -5.96286654e-01
-1.86298996e-01 6.38305843e-01 7.41813406e-02 -7.07494736e-01
-1.57024458e-01 -1.54084966e-01 -5.22367656e-01 2.54206657e-01
2.95773029e-01 -1.18954912e-01 2.61794120e-01 -8.59548390e-01
-4.89276275e-02 5.42684376e-01 -3.91965926e-01 1.12460464e-01
1.44672596e+00 -1.98512852e-01 1.18886255e-01 8.83284956e-02
1.34877968e+00 -3.31238568e-01 -9.30280149e-01 -2.51172334e-01
2.56800383e-01 -5.20918846e-01 8.56963173e-02 -1.13482845e+00
-1.07333744e+00 6.57169342e-01 1.74724028e-01 5.72706521e-01
1.16779399e+00 3.86354208e-01 1.08060265e+00 3.86158973e-01
6.80772483e-01 -1.11972570e+00 -4.92873281e-01 4.52084094e-01
4.58936423e-01 -1.45865011e+00 -2.72570401e-01 -7.90721595e-01
-1.16321468e+00 7.88518786e-01 3.92159909e-01 4.88372028e-01
5.98092794e-01 3.18692327e-01 2.66391218e-01 -3.02511036e-01
-5.99371314e-01 -5.86927116e-01 4.35010761e-01 6.50316954e-01
6.43937051e-01 1.38307437e-01 -5.69105186e-02 1.01221550e+00
-4.24392551e-01 -2.23015398e-02 1.88628286e-01 9.08319533e-01
-1.04682297e-01 -1.63908720e+00 1.38505742e-01 7.81248391e-01
-7.89753377e-01 -3.14820707e-01 -3.95992577e-01 7.47358024e-01
8.34861547e-02 1.07172811e+00 1.34448623e-02 -3.43299389e-01
5.04622281e-01 3.17278504e-01 3.61554205e-01 -7.80685782e-01
-9.41927373e-01 1.00754410e-01 4.43753600e-01 -4.01423752e-01
-2.82310992e-01 -6.07148647e-01 -1.18686771e+00 -2.19702974e-01
-3.55461389e-01 3.06176096e-01 8.45846653e-01 6.44579113e-01
7.11231053e-01 9.19776931e-02 5.06490529e-01 -2.27797627e-01
-9.87062037e-01 -7.99719393e-01 -9.56615865e-01 9.52936232e-01
-6.14539720e-02 -6.50991797e-01 -5.93507811e-02 1.11867294e-01] | [9.975106239318848, 6.28148889541626] |
f4cfee60-f07b-4b29-a0ac-a273b68b40bb | framework-for-2d-ad-placements-in-lineartv | 2212.02450 | null | https://arxiv.org/abs/2212.02450v1 | https://arxiv.org/pdf/2212.02450v1.pdf | Framework for 2D Ad placements in LinearTV | Virtual Product placement(VPP) is the advertising technique of digitally placing a branded object into the scene of a movie or TV show. This type of advertising provides the ability for brands to reach consumers without interrupting the viewing experience with a commercial break, as the products are seen in the background or as props. Despite this being a billion-dollar industry, ad rendering technique is currently executed at post production stage, manually either with the help of VFx artists or through semi-automated solutions. In this paper, we demonstrate a fully automated framework to digitally place 2-D ads in linear TV cooking shows captured using single-view camera with small camera movements. Without access to full video or production camera configuration, this framework performs the following tasks (i) identifying empty space for 2-D ad placement (ii) kitchen scene understanding (iii) occlusion handling (iv) ambient lighting and (v) ad tracking. | ['Sia Gholami', 'Karan Sindwani', 'Divya Bhargavi'] | 2022-12-05 | null | null | null | null | ['occlusion-handling'] | ['computer-vision'] | [ 5.34234047e-01 6.47333115e-02 1.08873390e-01 -2.61855632e-01
-4.34916258e-01 -1.36410046e+00 5.46662450e-01 4.96847212e-01
1.68833837e-01 -8.99584889e-02 -1.20233171e-01 -6.50295794e-01
2.90793777e-01 -7.23468482e-01 -8.74572515e-01 -7.63701871e-02
2.20071405e-01 6.38431251e-01 6.16088629e-01 -1.39379725e-01
2.02560350e-01 5.66522717e-01 -1.77712739e+00 4.56590593e-01
2.19141871e-01 1.28585696e+00 4.05954957e-01 1.07935774e+00
-3.89021546e-01 7.76212692e-01 -3.61508876e-01 -5.08033752e-01
7.66267061e-01 -1.71490565e-01 -1.64925009e-01 8.49372804e-01
7.63075650e-01 -8.41094673e-01 -2.47777570e-02 8.60820234e-01
5.27941659e-02 3.24908048e-02 6.59040734e-02 -1.27290845e+00
-3.59636694e-01 4.36405629e-01 -5.72415590e-01 -2.17235684e-01
9.62359190e-01 7.47841150e-02 5.81466854e-01 -5.24213195e-01
7.66090631e-01 1.14598179e+00 4.30618286e-01 -1.32147372e-01
-1.29622126e+00 -5.25090039e-01 2.72456586e-01 -8.73316750e-02
-1.14630282e+00 -4.39268261e-01 1.07724476e+00 -4.57497925e-01
5.69100559e-01 6.97184861e-01 1.01056266e+00 6.47166252e-01
3.50612029e-02 5.69884181e-01 1.01223886e+00 -7.25899935e-01
4.55305159e-01 7.73309648e-01 1.39294893e-01 5.72329819e-01
-1.56095222e-01 -1.47310063e-01 -3.89710009e-01 -1.41711399e-01
1.08222628e+00 1.49197534e-01 1.81141034e-01 -4.96076792e-01
-8.23083758e-01 6.47971809e-01 1.24410279e-01 -1.79759800e-01
-5.97068250e-01 2.66064294e-02 3.92657787e-01 4.14456986e-02
4.13616717e-01 1.80345416e-01 -3.64279985e-01 -1.05117708e-01
-8.44871938e-01 3.66651118e-01 5.97877562e-01 1.34070373e+00
3.77914310e-01 -1.41671374e-01 3.41515422e-01 5.57844877e-01
5.84792018e-01 2.93278754e-01 -1.09315157e-01 -8.87927651e-01
3.47885370e-01 8.69081199e-01 4.92430329e-01 -1.26742375e+00
-8.45400468e-02 2.59916425e-01 7.93270543e-02 5.48304319e-01
4.72147763e-01 1.31192235e-02 -8.98256958e-01 8.71899247e-01
8.49732101e-01 3.92969418e-03 -7.22213805e-01 8.91032994e-01
6.65402532e-01 7.78822660e-01 4.32730168e-02 -2.14661762e-01
1.88445449e+00 -8.85705113e-01 -1.07086682e+00 -2.00961947e-01
-4.76358607e-02 -1.49806106e+00 1.09350717e+00 8.60874295e-01
-1.38939059e+00 -5.43402076e-01 -1.38260388e+00 -3.53698641e-01
-5.06709754e-01 -1.18210770e-01 5.32988846e-01 1.04907846e+00
-7.44801104e-01 3.62191051e-01 -7.45437145e-01 -3.18861336e-01
2.18460962e-01 3.31370294e-01 -4.82728370e-02 -2.44111508e-01
-5.52929521e-01 5.70499957e-01 -3.25404108e-01 2.81831082e-02
-1.17232132e+00 -9.08573508e-01 -8.35309803e-01 -1.65616423e-01
6.19513512e-01 -2.98442572e-01 1.40213275e+00 -1.37260473e+00
-1.68087828e+00 1.01680923e+00 2.27701142e-01 -1.00802332e-01
7.46115983e-01 -5.57299912e-01 -5.22795618e-01 2.64693290e-01
2.29783654e-02 5.72827995e-01 1.14054763e+00 -1.51481545e+00
-6.90452158e-01 -6.94493830e-01 5.14242828e-01 2.71012276e-01
2.84539819e-01 2.32821614e-01 -4.15081292e-01 -4.51929629e-01
3.75952214e-01 -1.01609564e+00 -1.12130545e-01 5.43687820e-01
-5.60004771e-01 3.11549157e-01 1.15586340e+00 -9.94342923e-01
1.06336570e+00 -2.36732292e+00 -4.31711733e-01 2.00536907e-01
5.42062558e-02 1.08489014e-01 4.90684301e-01 4.23413336e-01
-1.04099803e-01 -3.53019863e-01 4.50584650e-01 -3.77747476e-01
2.50818163e-01 -9.60144252e-02 -1.06797710e-01 4.73747700e-01
-4.56766218e-01 1.63160056e-01 -6.59082651e-01 -3.81296784e-01
7.31371403e-01 6.17161572e-01 -4.43710178e-01 2.36111686e-01
-5.02713323e-01 1.70191545e-02 -2.22629383e-01 8.54692101e-01
1.16446769e+00 2.01062188e-01 4.96107191e-01 -4.36523169e-01
-4.73559499e-01 2.66784579e-01 -1.88075984e+00 1.69857597e+00
-2.92857885e-01 3.97518039e-01 6.46740973e-01 -2.43199036e-01
5.91000915e-01 3.34761977e-01 4.25471425e-01 -6.72007263e-01
2.66330987e-01 -3.68680432e-02 -8.67209077e-01 -4.45969224e-01
9.65826273e-01 1.24748021e-01 5.16703054e-02 6.23079121e-01
-5.57723403e-01 6.64727688e-02 6.22058958e-02 2.40579784e-01
8.29859614e-01 4.66201752e-01 2.46384457e-01 -2.47864444e-02
8.56825039e-02 5.30805707e-01 2.78198998e-02 1.83879346e-01
-1.53315529e-01 5.96717536e-01 9.60824490e-02 -5.59185684e-01
-1.30501187e+00 -1.11405265e+00 6.14550011e-03 1.41468477e+00
5.59862077e-01 -3.68735701e-01 -8.95762384e-01 -3.86532903e-01
-1.64174929e-01 1.02309704e+00 -2.25974187e-01 3.82029921e-01
-6.41511381e-01 1.58623040e-01 -2.91390538e-01 2.89610177e-01
2.86678523e-01 -6.93356395e-01 -1.00676560e+00 3.07775527e-01
1.61388457e-01 -1.04616344e+00 -9.40065384e-01 2.82028556e-01
-5.38928747e-01 -9.74223912e-01 -4.19309825e-01 -8.10470581e-01
7.55445302e-01 8.18602562e-01 1.19538629e+00 -1.62351906e-01
-4.94234413e-01 6.69698119e-01 -4.85765725e-01 -5.21947980e-01
-3.88262302e-01 -6.16319001e-01 -2.50167996e-01 -5.56007773e-02
3.83461595e-01 -4.64323759e-01 -9.54169035e-01 5.00420928e-01
-8.23284686e-01 3.63608658e-01 -4.63532545e-02 -9.93568003e-02
7.53088832e-01 4.40472603e-01 -2.99593151e-01 -7.71449506e-01
3.47720087e-01 -8.13483968e-02 -1.03398359e+00 -5.68187854e-04
-9.26449746e-02 -6.82206571e-01 3.61398965e-01 -4.84748691e-01
-9.52253938e-01 5.14462054e-01 1.16435260e-01 -3.22008431e-01
-4.83431637e-01 -4.01656210e-01 -4.25060719e-01 2.87748069e-01
2.58465022e-01 -2.38557354e-01 -7.05123320e-02 -5.72316408e-01
5.64408123e-01 6.95748448e-01 4.37667876e-01 4.42957878e-02
4.66795146e-01 5.20000160e-01 -3.59880894e-01 -5.84166706e-01
-5.31046808e-01 -5.96207440e-01 -7.21959352e-01 -5.82521141e-01
1.15579426e+00 -7.41914928e-01 -9.46481943e-01 1.55639714e-02
-1.07167852e+00 -1.54083446e-01 -2.05682367e-01 -3.65482015e-03
-2.83351362e-01 -2.11072084e-03 -3.81937921e-01 -9.71338272e-01
-2.68584862e-03 -1.35654187e+00 1.22403538e+00 1.16568737e-01
-2.93868631e-01 -6.12186849e-01 -1.99334279e-01 7.68647552e-01
-1.53746502e-02 4.17954385e-01 6.50122046e-01 -1.66066378e-01
-8.65661800e-01 -3.77778918e-01 -7.43574947e-02 1.57518923e-01
1.65204629e-01 1.08608805e-01 -1.12778306e+00 8.06768388e-02
-1.31838247e-01 3.41757834e-01 -2.24629328e-01 6.93053007e-01
7.65841842e-01 -3.76852363e-01 -3.49112421e-01 1.96313635e-01
1.61067247e+00 7.70072699e-01 6.79104209e-01 5.33748984e-01
4.60094839e-01 5.59005916e-01 9.88402903e-01 5.15754461e-01
2.20218256e-01 1.02344251e+00 7.52599776e-01 -3.25625986e-01
-2.30099320e-01 -5.23281813e-01 2.68109918e-01 1.35196839e-02
1.97255179e-01 -3.06701571e-01 -2.97621459e-01 2.84422368e-01
-1.49876618e+00 -8.05055737e-01 -4.81053174e-01 2.39169526e+00
9.21621844e-02 2.72152543e-01 5.53407907e-01 2.70883203e-01
9.70129371e-01 -2.74692178e-01 -3.67900461e-01 -8.76105547e-01
4.08758372e-01 -1.31438464e-01 8.46811712e-01 4.37189758e-01
-1.14510274e+00 4.62444901e-01 6.11476851e+00 3.02389115e-01
-8.48160565e-01 2.88954228e-01 5.90791225e-01 -2.43647665e-01
-2.29329705e-01 -9.54577029e-02 -6.96709394e-01 5.40176868e-01
2.61331260e-01 5.34195900e-01 6.63194716e-01 1.15635598e+00
4.32356298e-01 -4.42262560e-01 -1.22105753e+00 1.00340629e+00
5.03069833e-02 -1.12351263e+00 -3.25762510e-01 2.59819180e-01
4.15394068e-01 -8.31281006e-01 1.65576085e-01 -2.33702332e-01
2.76237100e-01 -6.24764740e-01 1.46919107e+00 -1.71753258e-01
6.16677165e-01 -6.83134198e-01 1.16190761e-01 -7.01943040e-02
-1.23424423e+00 -4.92815785e-02 -1.33013055e-01 9.70370024e-02
8.73759627e-01 2.14010894e-01 -1.04502726e+00 -3.13892327e-02
6.09280288e-01 -1.06699746e-02 -3.71575803e-01 7.78991282e-01
2.09473103e-01 1.43756881e-01 -4.47527826e-01 2.47120753e-01
2.76830494e-01 -4.25017983e-01 5.18533528e-01 9.36952114e-01
2.89602250e-01 1.38497502e-01 1.49665669e-01 5.03308475e-01
2.54056215e-01 1.25296503e-01 -4.77470726e-01 -4.91957515e-02
2.02378809e-01 1.37241566e+00 -1.29875839e+00 -2.85975952e-02
-3.82006317e-01 1.36148703e+00 -5.63738704e-01 -7.71899670e-02
-1.17644250e+00 -2.80813221e-02 4.31456327e-01 1.12912309e+00
6.31115913e-01 -2.61771142e-01 -1.45071402e-01 -4.12213326e-01
-5.70526831e-02 -7.64658451e-01 1.82878271e-01 -1.22652376e+00
-6.95938766e-01 4.66992795e-01 -5.37714772e-02 -1.21820986e+00
1.52172253e-01 -5.51173091e-01 -2.63652831e-01 2.15941578e-01
-8.98348927e-01 -1.46251130e+00 -4.43413347e-01 4.72118169e-01
1.00639558e+00 1.53801695e-01 6.43545330e-01 8.27429771e-01
-3.13892394e-01 1.99732825e-01 -3.08368001e-02 -6.77681863e-01
5.67198396e-01 -1.27196896e+00 3.02175619e-02 5.62026978e-01
8.86699483e-02 2.10344419e-01 1.24547195e+00 -7.07696855e-01
-1.82442498e+00 -6.53391004e-01 5.37429631e-01 -4.58909273e-01
3.88267249e-01 -8.65492821e-01 -2.50953615e-01 8.46853316e-01
5.33675671e-01 -4.44555014e-01 7.10372627e-01 -8.45320448e-02
1.22945875e-01 -2.82956243e-01 -1.47094274e+00 3.95821571e-01
7.74007916e-01 -8.08217973e-02 -6.66003078e-02 7.36171544e-01
6.75213754e-01 -7.20468640e-01 -5.34053981e-01 -4.61649925e-01
8.85685146e-01 -9.75792110e-01 1.21537352e+00 -6.13993593e-02
3.54452580e-01 -5.29287815e-01 -3.96229118e-01 -6.76906765e-01
-3.62590775e-02 -8.46814454e-01 3.19912523e-01 1.45724344e+00
3.08503956e-01 -2.61419569e-03 6.59533083e-01 1.18936443e+00
-2.85052985e-01 -6.05399497e-02 -4.42204177e-01 -1.23591617e-01
-1.08527279e+00 -6.89030349e-01 5.84078848e-01 7.00376511e-01
-1.20341614e-01 2.45134130e-01 -3.42787892e-01 4.94869679e-01
4.84555215e-01 -3.22639057e-03 8.80406260e-01 -8.36946487e-01
-3.86293739e-01 5.73919713e-02 -3.47413719e-01 -1.00226045e+00
-7.57854044e-01 -2.73277968e-01 -1.39904290e-01 -1.50778627e+00
1.77116320e-02 -5.82059264e-01 3.95120591e-01 -1.26825035e-01
5.37652850e-01 3.61328959e-01 2.59344250e-01 -2.05281213e-01
-6.49749279e-01 -3.29269618e-01 1.17305064e+00 1.17277443e-01
-4.05945748e-01 7.13852495e-02 -5.84740520e-01 8.23603690e-01
3.64982128e-01 -4.01448011e-01 -7.26372302e-01 -3.94893378e-01
5.00732005e-01 1.36738062e-01 4.92646545e-01 -4.97726083e-01
1.45323724e-01 1.85251255e-02 5.43111324e-01 -1.09811461e+00
5.61261892e-01 -1.56083369e+00 8.84416342e-01 1.34665683e-01
-7.09739998e-02 4.66385603e-01 2.33063325e-01 6.07337773e-01
4.58208740e-01 -3.09821069e-01 5.50473750e-01 -5.29462576e-01
-5.40527701e-01 -2.42644265e-01 -4.97009754e-01 -7.78013229e-01
1.58345592e+00 -7.27855623e-01 -6.12571128e-02 -3.28811526e-01
-1.07366288e+00 -2.13847980e-01 6.43625617e-01 2.35317349e-01
3.63292724e-01 -1.14364552e+00 8.54201689e-02 3.75505537e-01
-3.06900829e-01 4.79060933e-02 5.29937029e-01 3.93391937e-01
-1.36222625e+00 2.91811228e-01 -8.27974528e-02 -6.78300619e-01
-1.67751181e+00 9.69332874e-01 -2.41013570e-03 8.91939178e-02
-8.85436118e-01 7.00689673e-01 3.20544273e-01 9.54808220e-02
3.43455672e-01 -3.21811289e-01 -7.01849163e-02 8.97405818e-02
9.28071678e-01 6.91877842e-01 3.30316722e-01 -5.12059748e-01
-2.83139795e-01 3.04617941e-01 -1.30782396e-01 -1.98447421e-01
1.29125667e+00 -7.20546365e-01 3.61205786e-01 2.85236955e-01
8.53271306e-01 1.56237826e-01 -1.53232908e+00 3.24852109e-01
-5.00599623e-01 -9.48263407e-01 4.87789512e-01 -8.77442181e-01
-8.95061553e-01 4.73776162e-01 1.07277060e+00 6.01733148e-01
1.06549037e+00 1.05133861e-01 1.07261813e+00 -4.97953802e-01
3.98724854e-01 -1.35848868e+00 9.63074714e-02 -6.15449786e-01
7.42096245e-01 -9.77660835e-01 2.55535036e-01 -7.24141598e-01
-9.33586180e-01 9.11340356e-01 3.13090801e-01 -1.05161369e-01
6.95061803e-01 6.48716986e-01 1.00167863e-01 -6.07239306e-01
-2.93869585e-01 2.81928062e-01 -1.10746324e-02 6.61051452e-01
2.94239074e-01 2.28339419e-01 7.75477439e-02 3.30847174e-01
-2.15171233e-01 -2.71696150e-01 4.81378257e-01 1.22983050e+00
-4.71174210e-01 -8.72601867e-01 -7.12998569e-01 4.60201651e-02
-5.20513892e-01 1.19266607e-01 -2.31017441e-01 7.00532854e-01
3.47543210e-01 1.16478205e+00 3.30448300e-01 6.01122491e-02
6.93351328e-01 -2.32612088e-01 5.52477419e-01 -4.66960996e-01
-9.49840009e-01 7.49369919e-01 4.21576560e-01 -5.39431453e-01
-9.87852588e-02 -8.38346601e-01 -6.86685026e-01 -5.06012201e-01
-3.30739379e-01 -4.80026245e-01 1.36320138e+00 4.87995386e-01
2.03804761e-01 5.70931077e-01 5.40881932e-01 -9.20720220e-01
9.00247768e-02 -4.62884396e-01 -7.95997679e-01 3.24122518e-01
1.64014757e-01 -7.01187670e-01 -9.04172286e-02 5.89431942e-01] | [9.407934188842773, -2.6435210704803467] |
1989d763-434b-4e2c-9469-67adabc94182 | amrnet-chips-augmentation-in-areial-images | 2009.07168 | null | https://arxiv.org/abs/2009.07168v2 | https://arxiv.org/pdf/2009.07168v2.pdf | AMRNet: Chips Augmentation in Aerial Images Object Detection | Object detection in aerial images is a challenging task due to the following reasons: (1) objects are small and dense relative to images; (2) the object scale varies in a wide range; (3) the number of object in different classes is imbalanced. Many current methods adopt cropping idea: splitting high resolution images into serials subregions (chips) and detecting on them. However, some problems such as scale variation, object sparsity, and class imbalance exist in the process of training network with chips. In this work, three augmentation methods are introduced to relieve these problems. Specifically, we propose a scale adaptive module, which dynamically adjusts chip size to balance object scale, narrowing scale variation in training. In addtion, we introduce mosaic to augment datasets, relieving object sparity problem. To balance catgory, we present mask resampling to paste object in chips with panoramic segmentation. Our model achieves state-of-the-art perfomance on two popular aerial image datasets of VisDrone and UAVDT. Remarkably, three methods can be independently applied to detectiors, increasing performance steady without the sacrifice of inference efficiency. | ['Hongpeng Wang', 'Ye Tian', 'Xinghao Song', 'Zhiwei Wei', 'Chenzhen Duan'] | 2020-09-15 | null | null | null | null | ['object-detection-in-aerial-images'] | ['computer-vision'] | [ 3.54897052e-01 -3.27562392e-01 -9.36678201e-02 -2.00158492e-01
-2.75194466e-01 -7.05453753e-01 -5.14709577e-02 -3.48065794e-01
-3.42137545e-01 4.85209823e-01 -4.53961700e-01 9.33795720e-02
-2.38449369e-02 -9.78500605e-01 -8.90443981e-01 -6.12759233e-01
1.60644636e-01 2.82960415e-01 9.21663463e-01 -3.95025350e-02
8.59050751e-02 6.33314490e-01 -1.69219112e+00 2.86952615e-01
9.97256517e-01 1.19750977e+00 4.06356037e-01 5.04145324e-01
-1.19492777e-01 5.05276382e-01 -6.37780547e-01 -1.33466125e-01
8.03043306e-01 -3.37149024e-01 -5.12581944e-01 4.91016388e-01
6.83670163e-01 -5.35702884e-01 -1.20179042e-01 1.37659287e+00
2.68278778e-01 -2.94031292e-01 6.50763035e-01 -1.28010702e+00
-3.01859438e-01 6.52607679e-01 -1.41947496e+00 3.06717366e-01
-5.13364971e-01 1.45542800e-01 7.11821616e-01 -6.68969035e-01
1.99025184e-01 1.21960938e+00 6.04274929e-01 2.45472029e-01
-1.11464667e+00 -1.07725120e+00 3.97267431e-01 -2.25064978e-01
-1.47145319e+00 -8.58193114e-02 5.95314264e-01 -4.53459024e-01
4.66737002e-01 5.09844065e-01 7.80221999e-01 5.18126130e-01
6.07657433e-03 7.75111318e-01 1.07458532e+00 -2.28185415e-01
2.68319994e-01 1.65622652e-01 -6.86404854e-02 7.55800128e-01
4.96146649e-01 -2.38766938e-01 -2.05206856e-01 9.19613242e-02
1.06163287e+00 2.70922929e-01 -1.92059278e-01 -3.75011861e-01
-1.06391335e+00 4.76984322e-01 6.95316613e-01 1.45063952e-01
-1.28067017e-01 1.27949372e-01 2.42800787e-01 6.84175966e-03
2.70091683e-01 5.54152548e-01 -7.25649416e-01 5.84629834e-01
-1.06792414e+00 3.06162685e-01 3.25985074e-01 1.04113400e+00
7.53475666e-01 1.22960500e-01 -2.18002424e-01 8.28881204e-01
6.39106929e-02 5.41496992e-01 5.74770927e-01 -8.25979948e-01
3.83007526e-01 9.77236331e-01 -1.33523839e-02 -1.07673657e+00
-3.91795397e-01 -6.69114113e-01 -1.16621721e+00 3.65729690e-01
2.14688301e-01 -1.21855572e-01 -1.34810722e+00 1.48402560e+00
4.52642351e-01 1.76175609e-02 -3.87180537e-01 1.03254533e+00
8.26813519e-01 6.22196138e-01 -2.29729936e-02 -9.53759998e-02
1.64086652e+00 -9.30778086e-01 -5.23113132e-01 -5.79057813e-01
1.65842056e-01 -7.33276486e-01 9.55198884e-01 4.55288589e-01
-1.00794411e+00 -7.36900091e-01 -1.36945486e+00 1.74307510e-01
-3.96300822e-01 6.65257454e-01 6.23319983e-01 6.24061823e-01
-4.55094039e-01 3.24641168e-01 -5.65477252e-01 -1.82024941e-01
7.98538744e-01 5.19636691e-01 1.06080137e-01 1.05438069e-01
-6.76801860e-01 1.84430271e-01 6.64814472e-01 6.87856972e-02
-6.08008564e-01 -7.62735903e-01 -4.73560721e-01 -6.92961970e-03
7.36944497e-01 -5.29711902e-01 8.81614447e-01 -1.21202767e+00
-1.21587968e+00 8.82690549e-01 4.31398302e-01 -3.72739583e-01
5.18410146e-01 -1.46992505e-01 -1.97956219e-01 -6.31743670e-02
1.74949706e-01 1.01970184e+00 1.09721565e+00 -1.04716897e+00
-9.70214248e-01 -6.58613026e-01 7.13822767e-02 2.07721964e-01
-4.38935995e-01 -1.02560982e-01 -6.61598206e-01 -7.80661821e-01
5.42196393e-01 -6.88070714e-01 -2.47751579e-01 2.79604048e-01
-4.35213894e-01 1.37240306e-01 1.07703221e+00 -3.09744418e-01
1.43629766e+00 -2.20127368e+00 3.34239863e-02 -1.69699322e-02
3.39194357e-01 2.85195619e-01 -4.14161645e-02 -3.12843740e-01
1.15389563e-01 1.68676496e-01 -5.24450839e-01 1.53368384e-01
-2.93083131e-01 2.20968556e-02 -3.94899279e-01 4.15794820e-01
4.41340297e-01 6.14071548e-01 -4.40766096e-01 -7.66443133e-01
6.31092340e-02 -1.36055410e-01 -5.94465494e-01 1.58996969e-01
-3.29811096e-01 -3.81328762e-02 -4.42316025e-01 9.89263415e-01
1.16536748e+00 -1.87351465e-01 2.54511531e-03 -6.26167119e-01
-3.57972622e-01 -2.74783283e-01 -1.57764065e+00 1.41635573e+00
9.46145654e-02 1.79886118e-01 4.45189238e-01 -8.39990377e-01
8.29638958e-01 -2.26462409e-01 2.23138466e-01 -3.55761766e-01
4.14709538e-01 6.54918253e-02 1.05131693e-01 -4.57158059e-01
5.41602969e-01 3.21763232e-02 5.93399600e-05 2.94126961e-02
-1.20561801e-01 -6.66641176e-01 4.73421901e-01 -6.21021725e-02
7.20526218e-01 1.06877517e-02 3.46938312e-01 -4.71964359e-01
9.15869325e-02 9.27419439e-02 8.48401070e-01 5.71129084e-01
-2.37590462e-01 7.49652028e-01 5.36028981e-01 -4.14924413e-01
-9.39967036e-01 -6.49786472e-01 -4.31628853e-01 9.62043762e-01
4.55797881e-01 -8.18107724e-02 -9.65143085e-01 -6.19323432e-01
-4.82354648e-02 -1.45371363e-01 -6.20767653e-01 2.34352127e-01
-5.97555041e-01 -1.37443519e+00 5.41551173e-01 6.72493219e-01
1.09792793e+00 -7.31077790e-01 -1.03048146e+00 -1.41551256e-01
1.92190483e-02 -1.17540693e+00 -4.30198044e-01 4.31114107e-01
-1.00683987e+00 -1.00999153e+00 -5.75488865e-01 -9.39186990e-01
7.88617849e-01 5.83492875e-01 1.02709568e+00 1.44631704e-02
-8.44537199e-01 -3.54556233e-01 -1.79912865e-01 -4.63949323e-01
2.06320107e-01 2.55935818e-01 1.66774467e-02 -2.18535021e-01
1.18621469e-01 -3.48153979e-01 -7.91442215e-01 5.14051318e-01
-1.25970972e+00 1.30864501e-01 8.94748867e-01 6.63915455e-01
7.71500409e-01 6.04031086e-01 1.71360627e-01 -7.75709510e-01
-6.03067204e-02 -1.08973354e-01 -9.92124259e-01 2.03671217e-01
-3.22602063e-01 -2.64043957e-01 5.12634933e-01 -5.07979274e-01
-8.05369437e-01 3.65578711e-01 2.79776454e-01 -4.35084373e-01
-1.99991360e-01 -2.34413281e-01 -3.79834533e-01 -2.31123552e-01
5.66128850e-01 -1.24411359e-01 -1.15703486e-01 -1.92660093e-01
1.97625577e-01 6.47534311e-01 4.40564483e-01 -2.21774176e-01
9.06002939e-01 5.72143495e-01 2.44874321e-02 -6.98699296e-01
-8.88282239e-01 -2.83760726e-01 -7.44484603e-01 -5.46616614e-02
8.29299271e-01 -1.09524977e+00 -4.54773843e-01 6.02718413e-01
-8.05583119e-01 -3.77837151e-01 -2.77795553e-01 1.78602114e-01
4.49304562e-03 1.55538365e-01 -5.63104510e-01 -6.41889155e-01
-3.05144042e-01 -1.02367282e+00 1.03269076e+00 6.38942599e-01
2.34864891e-01 -8.90086889e-02 -4.97384489e-01 1.02740787e-01
2.28755698e-01 2.52362192e-01 7.87671745e-01 1.89742967e-02
-8.80349219e-01 -7.58291185e-02 -6.86125755e-01 4.06636059e-01
9.03300717e-02 2.27580726e-01 -8.69627357e-01 -2.91875094e-01
-8.42579827e-02 -3.94578904e-01 9.42788482e-01 6.44563854e-01
1.60806167e+00 -1.17655240e-01 -5.36242008e-01 7.85315037e-01
1.46970725e+00 1.16155073e-01 5.12191653e-01 2.10004747e-01
5.83305895e-01 5.90715587e-01 8.40074241e-01 3.20984572e-01
-1.83067843e-01 5.78255296e-01 5.33665121e-01 -4.53337908e-01
-1.86704800e-01 7.01345876e-02 -1.14348512e-02 2.32297599e-01
1.38198033e-01 -2.78524697e-01 -6.18082106e-01 4.46972668e-01
-1.65543973e+00 -5.60697377e-01 -1.93137661e-01 1.83026016e+00
7.52442598e-01 3.05278420e-01 2.26342708e-01 3.35054249e-02
7.74800956e-01 2.35454336e-01 -7.73857653e-01 3.65303636e-01
-3.54485363e-01 -1.56604603e-01 1.00597703e+00 1.10476064e-02
-1.31447506e+00 1.00681198e+00 5.87207842e+00 1.02852845e+00
-1.07962894e+00 -6.87330216e-02 1.01194096e+00 -2.51109868e-01
2.89918572e-01 -3.18641007e-01 -1.12207079e+00 5.03507257e-01
-4.86482196e-02 4.54067886e-01 3.61154258e-01 1.02072895e+00
-2.90002584e-01 -3.14975679e-01 -5.67986190e-01 1.02922821e+00
1.78454220e-01 -9.01524305e-01 -2.90691443e-02 -2.18097135e-01
8.07338476e-01 -1.62946686e-01 2.10287869e-01 1.76440865e-01
1.86064929e-01 -7.64417231e-01 7.78537631e-01 -5.91418706e-02
9.26390290e-01 -6.94370329e-01 5.83411455e-01 2.78305084e-01
-1.36513948e+00 -3.71389270e-01 -7.38247097e-01 1.03155129e-01
-2.27727637e-01 7.68850803e-01 -5.25887370e-01 2.54268289e-01
1.11832321e+00 2.98851281e-01 -8.22304189e-01 9.16005552e-01
5.78283742e-02 3.94509196e-01 -5.38433790e-01 1.96351349e-01
1.49106875e-01 -2.98286229e-01 3.47971588e-01 1.05849469e+00
2.51347393e-01 3.50539744e-01 2.91824281e-01 9.92181063e-01
-2.55978912e-01 -1.32934481e-01 -2.65841275e-01 1.53203949e-01
5.64647734e-01 1.53488147e+00 -1.30891240e+00 -2.17359081e-01
-1.86198026e-01 9.69282389e-01 4.49767672e-02 3.12957093e-02
-9.13996577e-01 -4.61865842e-01 3.63056183e-01 4.18162286e-01
5.39818764e-01 -2.03931937e-03 -4.89729047e-01 -9.14934814e-01
6.06337227e-02 -9.24807787e-01 3.46499622e-01 -4.51068789e-01
-9.06407654e-01 4.62581962e-01 -7.31561482e-02 -1.34806800e+00
5.12129068e-01 -6.95572793e-01 -3.47021848e-01 3.46426845e-01
-1.42712188e+00 -1.13793612e+00 -7.24331975e-01 8.00569877e-02
9.53708410e-01 -1.72793508e-01 2.99647063e-01 5.06021023e-01
-9.02417421e-01 3.72784346e-01 -1.72457904e-01 1.83791503e-01
7.64538288e-01 -1.03705311e+00 2.75311351e-01 1.01620221e+00
-7.13075250e-02 2.92378187e-01 4.96211559e-01 -6.65460765e-01
-1.19284344e+00 -1.34459352e+00 -4.62481342e-02 -2.27663353e-01
2.93630928e-01 -4.88122135e-01 -8.07492256e-01 5.04401624e-01
-1.24466889e-01 3.35248172e-01 2.71106452e-01 -1.92222446e-01
-2.05530867e-01 -4.17365760e-01 -1.14378190e+00 3.73898029e-01
9.79505241e-01 2.02385589e-01 -2.88140718e-02 2.74528176e-01
7.47068286e-01 -6.32864535e-01 -4.23521578e-01 8.21572006e-01
6.25932276e-01 -1.16480005e+00 9.39086258e-01 -2.33599931e-01
5.07854044e-01 -8.06609631e-01 -7.26515055e-02 -8.22636127e-01
-5.67596555e-01 -1.91388860e-01 -1.21154015e-04 1.25709772e+00
2.39229396e-01 -3.40260744e-01 8.80396724e-01 2.23153189e-01
-3.81509624e-02 -7.86515951e-01 -5.38478792e-01 -6.12240493e-01
-2.01745391e-01 2.35333189e-01 8.45261514e-01 8.18631053e-01
-6.51552022e-01 3.02556306e-01 -2.89886087e-01 5.04258275e-01
5.44012606e-01 3.89752597e-01 8.01971197e-01 -1.24409568e+00
-4.60005999e-01 -3.80280852e-01 -1.30384177e-01 -1.21366727e+00
-4.72407460e-01 -3.14172268e-01 3.13546062e-01 -1.05269754e+00
4.54726517e-01 -6.15289390e-01 -3.79807130e-02 3.81196439e-01
-2.75749236e-01 7.56202102e-01 1.43305182e-01 2.33429730e-01
-5.81197679e-01 2.74722397e-01 1.21448767e+00 -1.03814125e-01
-2.81247079e-01 -1.17020741e-01 -5.59468031e-01 9.85304415e-01
7.76041865e-01 -4.53307658e-01 -3.17566335e-01 -5.88268220e-01
2.07097515e-01 -3.06843430e-01 3.49528104e-01 -1.38293660e+00
1.07238621e-01 -1.75056219e-01 9.06993151e-01 -8.96052599e-01
1.55254960e-01 -8.33110332e-01 -5.11416644e-02 5.49758792e-01
2.15450618e-02 -4.25440371e-02 4.10491437e-01 5.16832113e-01
-8.62008706e-02 -2.45279893e-01 1.27252531e+00 -3.21624458e-01
-6.91588759e-01 4.11268979e-01 -1.52339622e-01 -1.17059834e-01
1.09910762e+00 -2.81574130e-01 -2.75656372e-01 3.98797631e-01
-4.93223481e-02 4.18966770e-01 5.68013251e-01 2.43286967e-01
2.09990174e-01 -8.91102552e-01 -4.56462711e-01 4.64949042e-01
-1.25153765e-01 6.36752903e-01 4.11853343e-01 7.57771134e-01
-6.15873158e-01 1.74826682e-01 -3.59423161e-01 -6.39655054e-01
-1.38140202e+00 3.92275661e-01 3.53613049e-01 -1.24492817e-01
-4.54177320e-01 1.19174397e+00 7.58889437e-01 -3.84705454e-01
2.38418132e-01 -6.26696050e-01 -1.23290025e-01 2.23797381e-01
5.12372911e-01 3.40737134e-01 -4.75929752e-02 -5.93226328e-02
-2.33152971e-01 8.32118154e-01 -3.14835876e-01 3.39675963e-01
1.07480264e+00 -7.47016370e-02 -2.22567603e-01 1.55586749e-01
5.52998960e-01 -1.43551424e-01 -1.47345376e+00 5.65830655e-02
-5.51261485e-01 -5.67355514e-01 1.88935250e-01 -6.57965958e-01
-1.11504972e+00 5.32593012e-01 9.61934865e-01 3.45878392e-01
1.32611704e+00 -1.61418617e-01 7.06411302e-01 8.02089944e-02
1.65939599e-01 -1.28004587e+00 2.29660720e-01 1.50879294e-01
5.96113324e-01 -1.28902698e+00 4.72676933e-01 -8.50026429e-01
-4.51585829e-01 1.03052175e+00 1.14741397e+00 -2.85967886e-01
3.51103336e-01 6.74745202e-01 -2.16762841e-01 -2.60622174e-01
-2.52146423e-01 -2.04006597e-01 1.06151447e-01 2.08595216e-01
1.49958253e-01 3.90346982e-02 -2.46963441e-01 6.18034363e-01
-3.44673991e-02 -1.30126163e-01 3.55351925e-01 9.32991624e-01
-7.39667177e-01 -5.69707811e-01 -6.20827377e-01 6.43066585e-01
-5.14842927e-01 7.16261789e-02 -3.07234734e-01 7.56556749e-01
6.66492045e-01 5.48263311e-01 3.53673398e-01 -2.35035464e-01
2.35915497e-01 -4.13419306e-01 4.32085186e-01 -5.56311250e-01
-3.09437573e-01 2.07494989e-01 -3.18324447e-01 -3.85347962e-01
-3.11909497e-01 -2.80273229e-01 -1.00082302e+00 1.49857968e-01
-7.51190603e-01 -2.36286521e-01 4.41027313e-01 6.18770599e-01
2.60796428e-01 9.73898292e-01 4.01119500e-01 -7.38192797e-01
-5.01100183e-01 -8.91695559e-01 -9.25052524e-01 -5.20763546e-02
1.02300465e-01 -5.90938747e-01 -3.91815394e-01 2.88440436e-01] | [8.741085052490234, -0.7671900987625122] |
9867e3b5-a5d3-4e28-892f-8e123a827ff7 | tbi-gan-an-adversarial-learning-approach-for | 2208.06099 | null | https://arxiv.org/abs/2208.06099v1 | https://arxiv.org/pdf/2208.06099v1.pdf | TBI-GAN: An Adversarial Learning Approach for Data Synthesis on Traumatic Brain Segmentation | Brain network analysis for traumatic brain injury (TBI) patients is critical for its consciousness level assessment and prognosis evaluation, which requires the segmentation of certain consciousness-related brain regions. However, it is difficult to construct a TBI segmentation model as manually annotated MR scans of TBI patients are hard to collect. Data augmentation techniques can be applied to alleviate the issue of data scarcity. However, conventional data augmentation strategies such as spatial and intensity transformation are unable to mimic the deformation and lesions in traumatic brains, which limits the performance of the subsequent segmentation task. To address these issues, we propose a novel medical image inpainting model named TBI-GAN to synthesize TBI MR scans with paired brain label maps. The main strength of our TBI-GAN method is that it can generate TBI images and corresponding label maps simultaneously, which has not been achieved in the previous inpainting methods for medical images. We first generate the inpainted image under the guidance of edge information following a coarse-to-fine manner, and then the synthesized intensity image is used as the prior for label inpainting. Furthermore, we introduce a registration-based template augmentation pipeline to increase the diversity of the synthesized image pairs and enhance the capacity of data augmentation. Experimental results show that the proposed TBI-GAN method can produce sufficient synthesized TBI images with high quality and valid label maps, which can greatly improve the 2D and 3D traumatic brain segmentation performance compared with the alternatives. | ['Lichi Zhang', 'Zengxin Qi', 'Qian Wang', 'Zheren Li', 'Xuehai Wu', 'Ruizhe Zheng', 'Zhe Wang', 'Zeyu Wei', 'Kai Xuan', 'Zhenrong Shen', 'Sheng Wang', 'Di Zang', 'Xiangyu Zhao'] | 2022-08-12 | null | null | null | null | ['brain-segmentation'] | ['medical'] | [ 4.63037461e-01 4.84038237e-03 2.03181863e-01 -3.32999796e-01
-5.84888220e-01 -1.83730230e-01 1.30396619e-01 -1.82798430e-01
-3.55909169e-01 7.11967289e-01 2.82985926e-01 5.73361106e-02
7.18827099e-02 -8.62512827e-01 -3.62585574e-01 -8.11267376e-01
4.02446568e-01 5.25420606e-01 2.41916433e-01 -7.37527609e-02
4.36264230e-03 4.76471007e-01 -1.15499425e+00 1.00060686e-01
1.34231687e+00 9.35868442e-01 4.48997378e-01 6.33625984e-02
-3.73831004e-01 5.75828969e-01 -5.22192061e-01 -1.10854067e-01
4.57146555e-01 -8.10767651e-01 -7.14427888e-01 3.86338592e-01
3.42184678e-02 -5.36557376e-01 -6.49459839e-01 1.02078283e+00
4.91003811e-01 -5.03822230e-03 5.56525528e-01 -1.23574948e+00
-3.25260401e-01 3.99463177e-01 -8.31867993e-01 1.51615366e-01
-4.45836410e-03 2.70235270e-01 1.91254303e-01 -5.98332644e-01
3.68769616e-01 8.51894796e-01 2.62686253e-01 8.21201861e-01
-9.88847792e-01 -9.06953692e-01 -1.43128410e-01 2.27824733e-01
-1.11016119e+00 -1.10579975e-01 1.01713955e+00 -5.03464282e-01
2.80118853e-01 2.84057371e-02 1.16026068e+00 6.58415020e-01
1.66823909e-01 7.07959175e-01 1.52592397e+00 -9.96244326e-02
9.40199941e-02 -6.26273334e-01 -3.71398777e-02 7.86408544e-01
7.40786865e-02 -1.06731758e-01 -3.71050686e-01 2.80727804e-01
1.15413582e+00 3.27201337e-01 -7.02495635e-01 -1.18365109e-01
-1.42513239e+00 5.51856995e-01 7.23795056e-01 2.32731804e-01
-6.84946179e-01 -5.39412796e-02 3.13467681e-01 5.20057371e-03
5.14157832e-01 1.35780349e-01 1.27974793e-01 3.60134989e-02
-1.18936801e+00 -1.86016142e-01 1.24826275e-01 7.55048692e-01
7.02700138e-01 2.63271213e-01 -2.88174659e-01 9.19099092e-01
-6.47218972e-02 2.91012973e-01 6.75001085e-01 -7.88388073e-01
3.65185529e-01 7.61347950e-01 -4.09461170e-01 -5.36052942e-01
-4.05438215e-01 -2.77001679e-01 -1.12922204e+00 2.88191587e-01
3.08632076e-01 -1.61287278e-01 -1.50916743e+00 1.60667431e+00
3.56665164e-01 3.25470030e-01 -2.14559495e-01 1.13896239e+00
6.95592403e-01 4.13957208e-01 -3.06616761e-02 -5.06883264e-01
1.20731032e+00 -1.10907841e+00 -9.06169057e-01 -3.02567095e-01
4.30591583e-01 -5.56522191e-01 1.15338755e+00 2.05646023e-01
-1.16056335e+00 -3.14042568e-02 -8.53447497e-01 2.67244130e-02
2.73251951e-01 -2.81904519e-01 5.05118310e-01 4.37882394e-01
-8.64733875e-01 1.12787694e-01 -9.92640615e-01 7.42718205e-02
9.82029378e-01 4.32526380e-01 -4.92565185e-01 -4.60268140e-01
-9.23490226e-01 9.05784130e-01 3.88548672e-01 1.36655509e-01
-9.36534941e-01 -8.48598838e-01 -7.63000309e-01 -2.08516106e-01
1.87833250e-01 -6.96975887e-01 9.07905817e-01 -7.00912118e-01
-1.31648493e+00 7.74830163e-01 -1.28564954e-01 -1.50710285e-01
7.45211899e-01 2.46874660e-01 -6.63874149e-02 3.98210078e-01
1.06791809e-01 1.03867543e+00 6.49498343e-01 -1.40490675e+00
-3.28300238e-01 -5.56505263e-01 -2.59694904e-01 3.06007147e-01
-1.21005692e-01 -1.20215013e-03 -5.44338465e-01 -8.86223733e-01
4.88326669e-01 -7.00451910e-01 -3.43141139e-01 -1.73636600e-02
-2.96157479e-01 3.62709731e-01 1.07512748e+00 -1.02912009e+00
7.35464990e-01 -1.99479687e+00 4.44575995e-02 3.17591637e-01
3.31480563e-01 3.22001241e-02 -1.23666741e-01 -2.99461395e-01
-7.92677850e-02 5.95112739e-04 -9.62054431e-01 -4.44116652e-01
-4.77753758e-01 3.69049668e-01 -6.39016228e-03 4.92535621e-01
-3.30677703e-02 1.00286734e+00 -6.78105295e-01 -8.37125242e-01
2.30976492e-01 6.16337180e-01 -5.17255485e-01 4.49424505e-01
-9.52632353e-02 1.20542026e+00 -5.00539660e-01 6.78700805e-01
7.36390054e-01 1.96882114e-01 -2.86714971e-01 -3.18505794e-01
9.73208770e-02 5.66930994e-02 -6.90286994e-01 1.93127394e+00
-4.08513844e-01 2.13643476e-01 2.12627649e-01 -1.01995385e+00
8.87056589e-01 4.78059500e-01 1.03528404e+00 -8.02513957e-01
4.45575356e-01 1.75465688e-01 1.97006047e-01 -4.74203825e-01
-5.94711564e-02 -6.09387279e-01 1.52899995e-01 6.73715115e-01
-2.00795710e-01 -5.90467095e-01 1.61754712e-01 -2.85193343e-02
9.29345131e-01 -1.25822455e-01 -4.38020349e-01 -2.55396981e-02
2.98994690e-01 -1.80275775e-02 9.11475658e-01 9.60327014e-02
-9.68858153e-02 1.18483555e+00 3.05394292e-01 -1.55164286e-01
-1.12892914e+00 -8.81171584e-01 -1.29835129e-01 1.96646661e-01
1.10716455e-01 2.21848920e-01 -1.36546040e+00 -6.64580226e-01
-5.71746290e-01 5.54964602e-01 -5.33918798e-01 -3.13941181e-01
-7.55131602e-01 -1.03232098e+00 4.05228913e-01 5.09754539e-01
1.00505161e+00 -1.34348619e+00 -5.95282972e-01 3.57458174e-01
-6.48907721e-01 -1.13042402e+00 -8.27383339e-01 -8.13547447e-02
-1.10699105e+00 -1.09706604e+00 -9.05856013e-01 -9.97995913e-01
1.12898433e+00 2.46345222e-01 5.35583377e-01 4.19539958e-01
-2.21799955e-01 -3.55427340e-02 -2.89545149e-01 -2.17947483e-01
-4.47407901e-01 -2.53716648e-01 6.81060925e-02 9.36451033e-02
-1.47164032e-01 -9.32427943e-01 -1.03116977e+00 2.82869458e-01
-1.46501029e+00 5.56061506e-01 6.26760483e-01 9.32865500e-01
6.99249208e-01 2.08063647e-01 6.50635779e-01 -6.42079115e-01
7.24340737e-01 -1.28559813e-01 -2.54221648e-01 2.28192374e-01
-5.23847401e-01 -1.96177617e-01 4.56283718e-01 -5.08033574e-01
-1.02588463e+00 1.35209560e-01 -3.49191874e-01 -5.94509780e-01
-2.05078930e-01 5.34404516e-01 -4.36636150e-01 -3.05650592e-01
1.20842829e-01 3.55233759e-01 4.12847906e-01 -2.92694747e-01
1.00305475e-01 4.92864251e-01 8.59022498e-01 -5.87063193e-01
5.09422898e-01 4.57107663e-01 -4.99493144e-02 -4.48447317e-01
-4.88130420e-01 -1.46950752e-01 -7.63138413e-01 -4.14690554e-01
1.07515025e+00 -5.60364485e-01 -3.29104751e-01 8.38028550e-01
-1.09106183e+00 -5.40585637e-01 -3.74365568e-01 6.75061285e-01
-5.43650031e-01 6.05369508e-01 -9.10385728e-01 -2.52292693e-01
-7.35896766e-01 -1.75306356e+00 6.81776643e-01 1.84618294e-01
2.09024385e-01 -6.94046080e-01 -2.82113463e-01 8.51190269e-01
3.96559298e-01 5.02386332e-01 1.26998699e+00 -1.42722830e-01
-4.61030811e-01 -8.14773962e-02 -2.78049260e-01 4.71793771e-01
4.03062761e-01 -5.63340962e-01 -6.30068362e-01 -2.17026561e-01
3.93118799e-01 -2.74275512e-01 7.11811006e-01 3.67471486e-01
1.30915415e+00 -1.39850497e-01 -6.42384887e-02 6.90457225e-01
1.15405297e+00 4.12650317e-01 9.16536629e-01 2.59536535e-01
1.03490674e+00 6.80651248e-01 3.25996041e-01 1.04167260e-01
3.26921642e-01 5.28175831e-01 5.07691085e-01 -4.92150873e-01
-5.54516554e-01 -5.82444444e-02 1.34866655e-01 1.23155737e+00
-2.78423931e-02 -5.58026396e-02 -1.06795216e+00 6.98919833e-01
-1.42240632e+00 -5.97604573e-01 -1.81641236e-01 2.09808469e+00
1.02598846e+00 -1.90956071e-01 -7.23426118e-02 2.49411657e-01
8.77926528e-01 -3.57152641e-01 -6.92353010e-01 1.75542936e-01
3.81063558e-02 5.15482605e-01 1.98051453e-01 3.47818524e-01
-6.05208695e-01 8.69884908e-01 5.60176897e+00 7.69508064e-01
-1.35865080e+00 5.23808837e-01 7.76542187e-01 -6.81749061e-02
-3.78709137e-01 -1.03654653e-01 -9.16946232e-02 6.83250189e-01
3.59300524e-01 9.95302647e-02 5.06684840e-01 2.63079911e-01
4.70912814e-01 -1.72666281e-01 -7.06066012e-01 9.77620065e-01
9.07475576e-02 -1.04123807e+00 2.13641658e-01 1.83540002e-01
8.16834867e-01 -8.39075074e-03 -1.09438509e-01 -3.20248194e-02
1.40514195e-01 -1.03864658e+00 5.91598213e-01 4.70729977e-01
1.11073351e+00 -9.18114126e-01 6.53059483e-01 3.69018108e-01
-8.06019425e-01 1.53058201e-01 -1.71510175e-01 4.54102367e-01
6.10693097e-01 7.35823810e-01 -8.46315801e-01 5.72946966e-01
5.57351291e-01 2.49859318e-01 -3.34029943e-01 1.40494490e+00
-3.83673638e-01 6.35498166e-01 -4.49529767e-01 6.60184860e-01
1.52009547e-01 -6.51201069e-01 3.40474635e-01 7.74496019e-01
5.14380157e-01 5.42791784e-01 2.93189973e-01 9.91387546e-01
-2.81615138e-01 9.77524817e-02 -3.20363641e-01 3.62927794e-01
2.59699047e-01 1.20191169e+00 -9.97247040e-01 -1.55740395e-01
-3.85580003e-01 1.01982069e+00 1.51246414e-01 3.16432714e-01
-7.67713010e-01 1.98921207e-02 2.40190357e-01 3.27635556e-01
-2.09713653e-01 -2.86251545e-01 -6.51923716e-01 -1.21291852e+00
7.16680754e-03 -9.52239871e-01 2.30324596e-01 -7.56969392e-01
-1.07451856e+00 8.49462092e-01 -4.04505171e-02 -1.16331482e+00
1.35604322e-01 1.21819749e-02 -9.02648628e-01 7.42315888e-01
-1.29579473e+00 -1.34667993e+00 -7.38988876e-01 8.76171410e-01
4.30842429e-01 1.36535481e-01 4.59952384e-01 4.44191724e-01
-6.84093893e-01 3.92140567e-01 -3.14003855e-01 2.11570814e-01
4.85784858e-01 -7.90196538e-01 -8.94100294e-02 1.08058703e+00
-2.82576352e-01 2.12201715e-01 2.83063501e-01 -9.01508629e-01
-9.58181143e-01 -1.23650360e+00 9.71829742e-02 1.45482823e-01
2.92306423e-01 -4.44603264e-02 -1.11647952e+00 4.79942709e-01
2.28109479e-01 2.22901292e-02 5.12587190e-01 -8.95705998e-01
2.41093755e-01 6.81051402e-04 -1.58709884e+00 5.71165919e-01
9.56391335e-01 -2.85419047e-01 -5.88303387e-01 3.45923126e-01
5.22213995e-01 -5.31087518e-01 -9.11009669e-01 8.48212898e-01
1.05258934e-01 -6.33971691e-01 7.82648206e-01 -1.17799953e-01
4.17073637e-01 -4.52533424e-01 2.79606938e-01 -1.49732435e+00
5.96377403e-02 -2.14987680e-01 3.03537130e-01 1.31252718e+00
3.53919156e-02 -6.73388779e-01 8.96560490e-01 1.00064349e+00
-5.62433660e-01 -8.42729628e-01 -1.05616915e+00 -5.15551507e-01
3.08781207e-01 -4.03433830e-01 7.86616206e-01 9.93958354e-01
-8.89152363e-02 -5.92152923e-02 -1.60109878e-01 -4.00883295e-02
6.84745848e-01 1.02912530e-01 3.45349491e-01 -9.71321106e-01
1.93942919e-01 -5.27157187e-01 -3.17650288e-01 -8.23359907e-01
2.70663738e-01 -1.07955289e+00 1.41357243e-01 -1.95524228e+00
3.06665570e-01 -8.12588394e-01 -2.49713331e-01 7.46527672e-01
-1.45552292e-01 7.40138173e-01 5.04333340e-02 5.10959148e-01
1.12925261e-01 9.79658186e-01 2.08270502e+00 -2.65329331e-01
-1.04831748e-01 -2.62994319e-01 -3.80730122e-01 6.27597690e-01
8.75210881e-01 -5.30563593e-01 -5.01498044e-01 -5.53289771e-01
-4.53674942e-01 3.55018288e-01 3.67057085e-01 -1.07707012e+00
3.22589338e-01 -1.45517647e-01 3.39462519e-01 -5.11740565e-01
2.02073812e-01 -8.19586694e-01 1.60481542e-01 3.52841824e-01
9.14567038e-02 1.64976522e-01 1.03708103e-01 2.29024425e-01
-4.84586865e-01 -1.30968288e-01 9.86187339e-01 -2.03962386e-01
-2.87330359e-01 9.28174198e-01 -4.20778930e-01 3.87600176e-02
1.08664274e+00 -3.88656437e-01 -9.60857123e-02 -3.89001548e-01
-5.51695406e-01 2.13704765e-01 8.27287316e-01 2.09570512e-01
1.01849747e+00 -1.30702448e+00 -7.14917362e-01 2.00847968e-01
-2.12729529e-01 6.19014382e-01 4.18940306e-01 1.34877741e+00
-7.43659377e-01 -9.08648819e-02 -6.03126526e-01 -4.70887184e-01
-9.41747606e-01 4.19821978e-01 4.62237537e-01 -2.59118050e-01
-9.17991281e-01 4.02159303e-01 5.06139159e-01 -2.71389425e-01
-9.24722403e-02 -2.75479823e-01 -1.42449439e-01 -2.37964988e-01
3.25763285e-01 6.79013878e-02 3.05350453e-01 -7.97635734e-01
-1.15488030e-01 5.57639062e-01 -1.17738374e-01 -1.73214510e-01
1.44133484e+00 -1.29972577e-01 -4.56392139e-01 -1.76524833e-01
9.04266894e-01 -1.91711098e-01 -1.35645700e+00 -6.59382492e-02
-5.02627432e-01 -4.71886128e-01 3.86607587e-01 -7.38667548e-01
-1.84823537e+00 1.05826485e+00 6.70848489e-01 -3.65132093e-01
1.47374380e+00 -2.28487015e-01 1.54397166e+00 -3.10531020e-01
6.44552588e-01 -6.95639074e-01 7.54450187e-02 1.88883319e-01
9.67339933e-01 -8.69833827e-01 -3.63889515e-01 -6.70659482e-01
-6.63447857e-01 9.99929905e-01 8.45176697e-01 2.27739021e-01
2.81617224e-01 4.32724863e-01 2.55238354e-01 -3.46909910e-01
-1.58302132e-02 -1.58629701e-01 2.05129549e-01 5.93320310e-01
1.59624815e-01 -1.48145825e-01 -4.76300716e-01 4.08706486e-01
-2.47968510e-01 2.54592579e-02 5.36456466e-01 9.17284727e-01
-1.48297474e-01 -1.26011074e+00 -4.28969383e-01 5.97490191e-01
-4.20304328e-01 -1.28302619e-01 -1.52126357e-01 4.64518577e-01
3.27037454e-01 8.44790936e-01 -3.72746699e-02 -3.19969803e-01
2.69223005e-01 5.48538715e-02 6.73078239e-01 -5.55261374e-01
-4.82705027e-01 1.05858199e-01 -3.45885754e-01 -2.44869277e-01
-3.42082262e-01 -5.12252033e-01 -1.75938511e+00 -2.07640007e-01
-3.01809490e-01 5.80211990e-02 5.93902826e-01 1.09274042e+00
1.56420752e-01 6.23234868e-01 5.77562809e-01 -7.84388125e-01
1.84586778e-01 -1.03298473e+00 -4.56695437e-01 7.14521646e-01
2.26847809e-02 -8.05841208e-01 -9.63820294e-02 1.84046909e-01] | [14.104812622070312, -2.200671911239624] |
ef8abe12-5f90-4c5d-88fb-ee32575293aa | automatic-stroke-classification-of-tabla | 2104.09064 | null | https://arxiv.org/abs/2104.09064v1 | https://arxiv.org/pdf/2104.09064v1.pdf | Automatic Stroke Classification of Tabla Accompaniment in Hindustani Vocal Concert Audio | The tabla is a unique percussion instrument due to the combined harmonic and percussive nature of its timbre, and the contrasting harmonic frequency ranges of its two drums. This allows a tabla player to uniquely emphasize parts of the rhythmic cycle (theka) in order to mark the salient positions. An analysis of the loudness dynamics and timing deviations at various cycle positions is an important part of musicological studies on the expressivity in tabla accompaniment. To achieve this at a corpus-level, and not restrict it to the few recordings that manual annotation can afford, it is helpful to have access to an automatic tabla transcription system. Although a few systems have been built by training models on labeled tabla strokes, the achieved accuracy does not necessarily carry over to unseen instruments. In this article, we report our work towards building an instrument-independent stroke classification system for accompaniment tabla based on the more easily available tabla solo audio tracks. We present acoustic features that capture the distinctive characteristics of tabla strokes and build an automatic system to predict the label as one of a reduced, but musicologically motivated, target set of four stroke categories. To address the lack of sufficient labeled training data, we turn to common data augmentation methods and find the use of pitch-shifting based augmentation to be most promising. We then analyse the important features and highlight the problem of their instrument-dependence while motivating the use of more task-specific data augmentation strategies to improve the diversity of training data. | ['Preeti Rao', 'Rohit M. A.'] | 2021-04-19 | null | null | null | null | ['stroke-classification'] | ['methodology'] | [ 4.80028391e-01 -1.83816671e-01 -1.71961024e-01 -2.17580348e-02
-7.09876001e-01 -9.17341828e-01 4.04415369e-01 -1.02137052e-01
-1.82204589e-01 4.68278140e-01 4.23340350e-01 -1.44610777e-01
-5.23111522e-01 -2.11149812e-01 -1.88138373e-02 -5.96195340e-01
1.37730213e-02 5.65140426e-01 2.65412241e-01 -4.68332350e-01
3.34301054e-01 7.32212126e-01 -1.85045004e+00 3.42880905e-01
5.02054691e-01 7.43339598e-01 2.76642084e-01 9.30471003e-01
-2.11806819e-02 5.63697815e-01 -8.36655676e-01 -7.26153627e-02
6.32051453e-02 -9.07336414e-01 -6.49259567e-01 6.44972622e-02
3.20643991e-01 1.17116526e-01 -5.75162238e-03 5.36500037e-01
5.89603305e-01 3.27838898e-01 7.93093383e-01 -9.57583964e-01
1.83043964e-02 9.24037814e-01 -2.59112656e-01 3.89298439e-01
3.51394713e-01 -2.68440604e-01 1.09770906e+00 -6.22034550e-01
6.57229900e-01 6.11662388e-01 9.65581000e-01 3.77269357e-01
-1.28916824e+00 -5.84457099e-01 -4.50158358e-01 1.66296765e-01
-1.21928537e+00 -5.79766512e-01 1.24817550e+00 -6.02721512e-01
5.67234933e-01 6.51371002e-01 9.29288387e-01 9.09268022e-01
-3.82889181e-01 7.11423695e-01 1.14583933e+00 -8.07559788e-01
-5.76325320e-03 1.18169719e-02 2.11640745e-02 1.66347966e-01
-2.69984692e-01 -5.23946136e-02 -6.64431095e-01 -5.41599467e-03
6.56659782e-01 -4.90308672e-01 -5.48637807e-01 -2.63196945e-01
-9.40366328e-01 5.82103550e-01 4.22866717e-02 8.10972929e-01
-2.10303336e-01 -1.16431512e-01 6.76711738e-01 2.34322697e-01
-8.86368752e-02 9.47447121e-01 -4.63918477e-01 -6.83802962e-01
-1.36950946e+00 2.06359103e-01 7.90521741e-01 4.56310838e-01
2.26258978e-01 3.22607189e-01 2.50787102e-02 1.11504745e+00
-2.05385134e-01 7.87955001e-02 6.51620090e-01 -9.11212742e-01
4.76392061e-02 3.73427778e-01 -1.12984404e-01 -8.07818830e-01
-5.87793171e-01 -5.25516987e-01 -3.31985474e-01 2.10685179e-01
6.50454938e-01 -3.13266702e-02 -4.77059007e-01 1.63471174e+00
-1.56422500e-02 -1.67857781e-01 -1.52049512e-01 8.96048307e-01
5.84919631e-01 5.26042640e-01 -2.28532866e-01 -3.83012205e-01
1.32152200e+00 -4.74533468e-01 -7.90514171e-01 2.04580352e-01
6.21493816e-01 -1.20060980e+00 1.57333398e+00 7.73228645e-01
-9.32666302e-01 -7.23992527e-01 -1.17762458e+00 1.12503441e-02
-1.30807683e-01 5.16110003e-01 4.88627195e-01 8.51491034e-01
-2.71446437e-01 7.89249301e-01 -6.10232115e-01 -2.38742754e-01
-6.18614815e-03 2.36240983e-01 -1.91683158e-01 6.34215355e-01
-8.62231314e-01 6.92719460e-01 4.05859143e-01 2.32918258e-03
-2.91629940e-01 -5.92546344e-01 -4.95332807e-01 -2.07144152e-02
3.30356449e-01 9.87263918e-02 1.35882008e+00 -1.06891251e+00
-1.61625385e+00 8.05914998e-01 1.86155096e-01 -1.74840301e-01
4.23096299e-01 -1.51922703e-01 -5.21491706e-01 2.60976553e-01
-1.95169866e-01 3.28051686e-01 8.24143946e-01 -1.15126884e+00
-6.74353540e-01 -6.16551451e-02 -3.53004605e-01 1.37050703e-01
-4.51391280e-01 1.47592664e-01 -2.58669466e-01 -1.06702518e+00
8.55608881e-02 -1.23443592e+00 3.82804245e-01 -6.00464582e-01
-3.68110418e-01 -1.41347215e-01 9.08114910e-01 -6.68318272e-01
1.50224185e+00 -2.44928813e+00 2.86934465e-01 2.66696632e-01
-2.17908815e-01 3.64922374e-01 1.03563771e-01 5.87880254e-01
-1.24615192e-01 -2.19998002e-01 -1.99460134e-01 -1.38297260e-01
-7.90691189e-03 2.74179518e-01 -4.71489668e-01 2.83494025e-01
1.29487249e-03 6.04581952e-01 -5.88166416e-01 -3.48150164e-01
1.98520944e-01 4.75362986e-01 -4.21683908e-01 7.15159252e-02
-4.19006348e-02 5.76912582e-01 -1.77195042e-01 3.15635026e-01
7.11100921e-02 5.03951430e-01 6.87854216e-02 -4.29678679e-01
-3.53990048e-01 6.01358712e-01 -1.30692792e+00 1.66063333e+00
-3.72002989e-01 9.33695555e-01 -1.07021309e-01 -7.81154454e-01
1.18756557e+00 4.81213808e-01 4.98662710e-01 -2.67646223e-01
2.55672961e-01 5.54597735e-01 5.91233611e-01 -6.14672542e-01
7.61218250e-01 -4.55930620e-01 -1.63407281e-01 4.46162790e-01
2.27259584e-02 -4.90328997e-01 3.23403895e-01 -2.45919347e-01
6.39582038e-01 2.97699153e-01 2.11779445e-01 -2.19621763e-01
4.87426996e-01 1.29770311e-02 2.97034860e-01 4.56566095e-01
1.05109274e-01 1.01853883e+00 3.09273273e-01 -2.59873003e-01
-9.58496809e-01 -7.22910106e-01 -2.53679037e-01 1.33641303e+00
-4.77202266e-01 -6.30297005e-01 -6.09180987e-01 -1.60206541e-01
-2.50031978e-01 6.08945668e-01 -5.55739343e-01 -4.66510057e-02
-7.97106922e-01 -5.24148464e-01 8.52159679e-01 5.58883786e-01
7.98640028e-02 -1.47388422e+00 -9.59882855e-01 3.67729843e-01
-2.53055811e-01 -7.73003519e-01 -3.96759450e-01 7.51563489e-01
-6.21004641e-01 -1.09934556e+00 -6.22430801e-01 -8.20859015e-01
-1.55183151e-01 -2.37734765e-01 6.94074273e-01 -1.37559444e-01
-4.01586473e-01 2.93535829e-01 -7.31408060e-01 -6.86174035e-01
-5.80575466e-01 3.98832887e-01 1.02354705e-01 -9.38836634e-02
1.84717983e-01 -7.87327230e-01 -2.51187414e-01 3.05566192e-01
-7.83440232e-01 -2.70141065e-01 1.87337488e-01 6.97023153e-01
5.15741825e-01 -2.44770385e-02 6.50858700e-01 -6.09135151e-01
8.56996715e-01 1.59186542e-01 -8.03394318e-02 -1.94652334e-01
-8.60260129e-02 -2.67758727e-01 6.04576588e-01 -8.55963528e-01
-6.81307554e-01 1.34738654e-01 -4.04273599e-01 -2.70132780e-01
-1.65354148e-01 5.66079080e-01 -1.37884244e-01 5.26541546e-02
8.89230907e-01 9.70675722e-02 1.20294295e-01 -7.58911073e-01
2.11679235e-01 9.16834354e-01 1.02682471e+00 -5.13483465e-01
6.43766701e-01 7.48057216e-02 -4.88764942e-02 -1.35055685e+00
-1.01212347e+00 -6.20689690e-01 -7.54157841e-01 -4.17125046e-01
6.73541129e-01 -2.91082740e-01 -4.83739763e-01 1.04375459e-01
-7.05437541e-01 -3.68525505e-01 -7.47088909e-01 5.42313218e-01
-9.07868922e-01 3.52236539e-01 -4.90674466e-01 -9.29233670e-01
-2.42052704e-01 -9.54165995e-01 7.56397247e-01 4.78534810e-02
-9.87027347e-01 -6.47489727e-01 4.04504180e-01 3.37677896e-01
2.15794206e-01 2.88896501e-01 9.65522766e-01 -7.11090088e-01
2.46563762e-01 -3.79253596e-01 4.82712895e-01 3.76360804e-01
4.19415474e-01 2.00444266e-01 -1.17076921e+00 -1.05882496e-01
9.84446630e-02 -4.65903133e-01 6.83573067e-01 3.89475584e-01
9.47887242e-01 1.35123450e-03 1.71763703e-01 5.28930962e-01
8.39843571e-01 3.61867934e-01 5.64438641e-01 5.60228527e-01
5.10197461e-01 8.09995532e-01 7.68923402e-01 2.75131553e-01
-3.20567846e-01 1.05702412e+00 3.80246304e-02 -4.20044661e-02
-5.23527145e-01 -2.99755156e-01 2.68905729e-01 1.07599497e+00
-4.56023514e-01 -3.47151980e-02 -6.35681510e-01 5.68359017e-01
-1.41547549e+00 -9.78338122e-01 -3.99408489e-01 2.24075508e+00
8.31797361e-01 2.87728518e-01 7.47178853e-01 1.16079533e+00
5.04152715e-01 2.06610523e-02 -1.05751239e-01 -7.76698232e-01
-2.40565896e-01 4.17005509e-01 6.26551360e-02 2.49392465e-01
-1.10003924e+00 7.17751443e-01 6.23995209e+00 9.25162673e-01
-1.26406074e+00 -2.44716093e-01 3.66663262e-02 -1.75655052e-01
-1.16573479e-02 -1.13992088e-01 -5.57215869e-01 3.88329268e-01
8.19578350e-01 1.61703989e-01 4.02162760e-01 6.15466774e-01
3.04700673e-01 1.03032477e-01 -1.03947091e+00 1.02381682e+00
1.20049626e-01 -1.12607455e+00 -2.00876012e-01 -6.06015995e-02
5.07534146e-01 -2.44032308e-01 4.88850772e-02 1.54901832e-01
-6.26682639e-01 -1.01448822e+00 9.36436832e-01 2.84335196e-01
7.83921778e-01 -8.90288889e-01 4.97795969e-01 1.93384737e-01
-1.18137240e+00 -1.39953941e-01 -6.56881928e-02 -2.80765384e-01
1.45188928e-01 -1.60215408e-01 -8.81360292e-01 2.55035490e-01
4.66421425e-01 3.46513152e-01 -4.86358881e-01 1.24545383e+00
-2.02716872e-01 1.06942225e+00 -5.49359560e-01 -1.40656039e-01
6.77689612e-02 -1.43193096e-01 7.72261083e-01 1.49092197e+00
3.72919321e-01 1.26328155e-01 -3.08040734e-02 5.05184054e-01
2.61558026e-01 4.63816196e-01 -3.83693099e-01 -1.62308261e-01
4.72202301e-01 1.33951664e+00 -9.42905426e-01 -2.39377767e-02
-2.79883463e-02 7.54836917e-01 -7.47408941e-02 -5.68419807e-02
-4.50242549e-01 -5.54990947e-01 2.55728424e-01 2.89668977e-01
3.67304921e-01 -2.59752601e-01 -4.19782668e-01 -5.82506180e-01
-1.58620268e-01 -1.10013819e+00 3.87036890e-01 -7.00506032e-01
-9.25017893e-01 7.48722494e-01 -1.57624289e-01 -1.42587018e+00
-5.23207545e-01 -3.95585477e-01 -7.42741108e-01 7.66085804e-01
-9.75071669e-01 -1.13961351e+00 -4.63547893e-02 3.58180016e-01
5.78595936e-01 -2.78490007e-01 1.11247075e+00 2.49339357e-01
-2.10727736e-01 6.09145641e-01 2.94379629e-02 1.19599514e-02
7.32949793e-01 -1.39234865e+00 -9.22361538e-02 4.11804497e-01
8.56240928e-01 3.45707506e-01 1.20210469e+00 -2.61308372e-01
-8.93563628e-01 -4.44330812e-01 8.25758755e-01 -4.56167608e-01
7.80914903e-01 -2.93292046e-01 -9.83068764e-01 3.17653775e-01
-6.28782287e-02 -5.88140011e-01 1.02562642e+00 3.24661553e-01
-6.82869181e-02 -2.52244659e-02 -5.24852455e-01 5.23001492e-01
7.02971876e-01 -6.45448685e-01 -1.13128698e+00 -6.59816489e-02
5.56232035e-03 -1.64736465e-01 -8.09329987e-01 4.38502759e-01
8.76721919e-01 -9.81668949e-01 6.86171055e-01 -3.73264492e-01
2.87301213e-01 -3.42474848e-01 1.01212613e-01 -1.22911465e+00
-1.75761998e-01 -8.40713918e-01 3.07525188e-01 1.59705293e+00
2.65084147e-01 -3.77952754e-02 9.43994045e-01 -1.97040793e-02
-3.76967251e-01 -3.17065299e-01 -8.75669777e-01 -7.37616241e-01
-5.00602834e-02 -6.22169495e-01 1.94640487e-01 9.90303040e-01
3.49922448e-01 4.71365869e-01 -6.15356922e-01 -3.09231043e-01
1.18515089e-01 3.91787171e-01 7.89586484e-01 -1.50897479e+00
-6.11564815e-01 -6.32186890e-01 -4.25987333e-01 -7.15330422e-01
-1.18007407e-01 -9.35676992e-01 3.97500098e-02 -9.32800114e-01
-7.04598501e-02 -5.89009523e-01 -4.24442679e-01 4.18639719e-01
1.27259210e-01 9.10649121e-01 4.79899466e-01 3.32713425e-01
-1.17108621e-01 4.55204755e-01 1.11072946e+00 1.00579508e-01
-8.86756122e-01 3.83863091e-01 -4.04219419e-01 9.12493467e-01
9.52848136e-01 -3.07016283e-01 -4.23336595e-01 1.05887830e-01
9.92670208e-02 5.02946787e-02 -1.81026086e-02 -1.25488865e+00
-3.11709493e-02 2.82469958e-01 2.27406979e-01 -6.63697898e-01
6.13557398e-01 -6.45729184e-01 2.82745868e-01 3.85097772e-01
-5.49316049e-01 -1.08852461e-01 3.30744296e-01 -2.09110025e-02
-2.27687076e-01 -6.17012739e-01 6.62222683e-01 1.88987136e-01
-4.15466100e-01 -3.14446092e-01 -7.72653282e-01 -2.59987712e-02
5.23106217e-01 -5.34953237e-01 3.43234599e-01 -3.77169013e-01
-9.70400989e-01 -4.32819605e-01 4.15448278e-01 5.02457380e-01
1.92091241e-01 -1.15783155e+00 -5.94649673e-01 2.09154218e-01
2.60901004e-01 -4.13477153e-01 9.33926627e-02 8.00522804e-01
-5.01027882e-01 2.38570109e-01 -4.44430351e-01 -4.12593126e-01
-1.64501917e+00 4.13481116e-01 2.18907908e-01 3.88191640e-02
-8.23705077e-01 7.62228370e-01 -2.16989160e-01 -1.29284993e-01
3.67667139e-01 -1.86115772e-01 -6.90144658e-01 5.04542589e-01
4.80941474e-01 5.00725925e-01 1.63819686e-01 -8.88677120e-01
-7.11769015e-02 5.87862730e-01 2.79598027e-01 -2.80841738e-01
1.25240064e+00 7.41924569e-02 1.30597472e-01 9.69525099e-01
8.76243770e-01 6.77223623e-01 -1.01826489e+00 2.99262077e-01
1.88174620e-01 -3.67560774e-01 1.93377174e-02 -8.24023724e-01
-5.80804408e-01 9.31284130e-01 4.93993104e-01 5.32842457e-01
1.21650028e+00 -4.34131809e-02 5.99341512e-01 2.02189371e-01
-4.85257106e-03 -1.12112308e+00 5.96575029e-02 4.35041517e-01
1.04674828e+00 -5.86286008e-01 -1.64495185e-01 -3.62413287e-01
-6.89419448e-01 1.33952332e+00 1.57825891e-02 -8.46961513e-02
3.58617783e-01 4.21861053e-01 3.00348759e-01 -1.22920051e-01
-2.37271756e-01 -3.30228060e-01 5.42574346e-01 6.78262532e-01
8.43586743e-01 -2.78858859e-02 -7.30870187e-01 6.93401396e-01
-1.11955059e+00 -8.71935859e-02 4.95144993e-01 6.33212328e-01
-4.69777346e-01 -1.44188654e+00 -5.28895378e-01 4.16168720e-01
-7.09613502e-01 3.42804343e-02 -7.56980896e-01 9.50492978e-01
3.06781858e-01 9.54346716e-01 5.44381700e-02 -3.61376613e-01
5.54965734e-01 3.75152647e-01 7.57535815e-01 -4.03690428e-01
-9.23777342e-01 6.41393363e-01 3.06706041e-01 3.04499865e-02
-4.79233146e-01 -8.63904655e-01 -1.10429668e+00 9.06502679e-02
-3.58238071e-01 4.12400961e-01 5.41966975e-01 8.38684559e-01
-3.58023196e-01 7.52053916e-01 4.62278068e-01 -1.16557348e+00
-3.35241318e-01 -1.23127103e+00 -8.31830800e-01 6.67938411e-01
1.86595410e-01 -5.75236201e-01 -3.42512369e-01 1.27329230e-01] | [15.854453086853027, 5.321338176727295] |
8bdc2fe2-b5f6-4cdd-a757-0068726d9bc9 | physics-informed-neural-networks-for-1 | null | null | http://phmpapers.org/index.php/phmconf/article/view/814 | http://phmpapers.org/index.php/phmconf/article/download/814/phmc_19_814 | Physics-informed neural networks for corrosion-fatigue prognosis | In this paper, we present a novel physics-informed neural network modeling approach for corrosion-fatigue. The hybrid approach is designed to merge physics- informed and data-driven layers within deep neural networks. The result is a cumulative damage model where the physics-informed layers are used to model the relatively well understood physics (crack growth through Paris law) and the data-driven layers account for the hard to model effects (bias in damage accumulation due to corrosion). A numerical experiment is used to present the main features of the proposed physics-informed recurrent neural network for damage accumulation. The test problem consists of predicting corrosion-fatigue of an Al 2024-T3 alloy used on panels of aircraft wing. Besides cyclic loading, the panels are also subjected to saline corrosion. The physics-informed neural network is trained using full observation of inputs (far-field loads, stress ratio and a corrosivity index – defined per airport) and very limited observation of outputs (crack length at inspection for only a small portion of the fleet). Results show that the physics-informed neural network is able to learn the correction in the original fatigue model due to corrosion and predictions are accurate enough for ranking damage in different airplanes in the fleet (which can be used to prioritizing inspection). | ['Arinan Dourado', 'Felipe A. C. Viana'] | 2019-09-22 | null | null | null | annual-conference-of-the-phm-society-2019-9 | ['physics-informed-machine-learning', 'graph-regression', 'graph-to-sequence'] | ['graphs', 'graphs', 'natural-language-processing'] | [-2.72163842e-02 -5.87972626e-02 6.19869947e-01 -2.10676283e-01
-3.50947589e-01 -1.05476499e-01 6.63291663e-02 1.94844127e-01
-1.20891720e-01 5.52482426e-01 -5.29415607e-02 -1.80872887e-01
-8.64234686e-01 -9.82919037e-01 -8.51643145e-01 -1.03428924e+00
-3.38729203e-01 8.06781948e-01 2.97166884e-01 -7.75641024e-01
2.17224866e-01 8.90150845e-01 -1.78607750e+00 5.65786362e-01
4.01859701e-01 9.71777976e-01 3.46500069e-01 8.69909585e-01
6.67270303e-01 7.83944428e-01 -6.39349639e-01 2.02456743e-01
-8.49031806e-02 -2.17034981e-01 -7.32173681e-01 -5.19263208e-01
-1.79082051e-01 -3.51492822e-01 -1.60089493e-01 5.65408528e-01
6.09050632e-01 2.47719347e-01 1.02112401e+00 -7.12459028e-01
-3.39297146e-01 7.50904441e-01 9.63942613e-03 2.25217506e-01
1.78176627e-01 -1.53989792e-02 6.70292616e-01 -9.12944794e-01
2.44766295e-01 9.92278337e-01 1.01604283e+00 3.53551626e-01
-5.79896927e-01 1.82142705e-01 -1.87618375e-01 3.83600980e-01
-1.04053330e+00 4.47454840e-01 1.09565759e+00 -8.29390526e-01
1.18183112e+00 8.52093920e-02 4.94819522e-01 1.00519037e+00
9.33376849e-01 -8.33347514e-02 7.63280690e-01 -2.94013679e-01
4.43589628e-01 -1.29608423e-01 3.41123134e-01 5.84984660e-01
4.39644635e-01 7.90120423e-01 -2.19837397e-01 1.24904104e-01
5.35821378e-01 1.11188859e-01 -1.96835175e-01 2.65509516e-01
-3.43444943e-01 6.41245544e-01 8.08713496e-01 5.38622737e-01
-9.02174950e-01 4.35379684e-01 4.80233580e-01 3.69587392e-01
4.34071392e-01 3.20544779e-01 -7.39287019e-01 1.26115903e-02
-1.08556128e+00 5.27815580e-01 7.08984017e-01 -3.68479379e-02
6.45623863e-01 7.30626047e-01 1.36063531e-01 6.30778015e-01
5.43999970e-01 5.44272065e-01 5.52325547e-01 -6.32141054e-01
-5.54557852e-02 4.58387434e-01 4.85697389e-02 -9.89338636e-01
-9.40031707e-01 -7.78551519e-01 -8.02951455e-01 7.80260324e-01
9.90125537e-02 -6.77544296e-01 -8.85234237e-01 1.19157803e+00
-3.13769504e-02 -8.10202062e-02 -1.19308278e-01 1.09336829e+00
6.69667482e-01 8.89930904e-01 -2.08029583e-01 1.10532612e-01
1.15313125e+00 -4.63162273e-01 -2.66816914e-01 -4.67034914e-02
7.96585560e-01 -2.42127582e-01 7.80245662e-01 8.73103499e-01
-1.31136942e+00 -9.28398550e-01 -1.40585816e+00 3.96016866e-01
-6.98006213e-01 2.51254011e-02 -1.91877410e-01 5.79537094e-01
-1.07158661e+00 1.21393263e+00 -7.28006899e-01 1.74261853e-02
-1.33136094e-01 4.41826344e-01 1.35304376e-01 -2.74263080e-02
-1.63149536e+00 1.19622564e+00 3.33740711e-01 9.99816656e-01
-1.46132731e+00 -4.64069933e-01 -5.53253412e-01 2.36758947e-01
1.16213955e-01 -6.29234016e-01 8.79522920e-01 -5.25593162e-01
-1.33522391e+00 2.06516296e-01 5.25302887e-01 -6.62993908e-01
6.09268397e-02 -4.97737914e-01 -3.43492806e-01 2.18713686e-01
-6.00072682e-01 -4.85827768e-04 6.37754917e-01 -1.75726342e+00
-1.59541994e-01 4.28595953e-02 1.15533456e-01 -1.84270218e-01
-2.74420142e-01 -2.40375862e-01 4.83152926e-01 -7.10720181e-01
-2.51920849e-01 -8.13887358e-01 -2.37281770e-01 -7.74422526e-01
-2.71798909e-01 -3.27327847e-02 7.46480644e-01 -9.73206401e-01
1.06383324e+00 -1.75319934e+00 3.49797100e-01 4.63308573e-01
-8.12825784e-02 1.28186986e-01 1.19254418e-01 1.04117787e+00
-6.05552018e-01 -1.32750645e-01 -8.42072010e-01 -1.84757262e-01
-8.43801424e-02 4.27740335e-01 -8.88075307e-02 4.07452881e-01
2.14671195e-01 4.25504237e-01 -4.25769627e-01 7.21483603e-02
5.31442463e-02 5.09010851e-01 -3.77329648e-01 1.94322243e-01
-2.28067592e-01 1.83545519e-02 -2.13155597e-01 5.43382347e-01
7.21160769e-01 6.11517370e-01 -4.73516077e-01 -4.39169109e-01
-1.56023934e-01 -2.32559979e-01 -9.35012698e-01 1.33509016e+00
-6.65479600e-01 3.51812035e-01 -6.41821548e-02 -1.08821332e+00
1.16411269e+00 4.77846473e-01 5.03091514e-01 -2.97863543e-01
5.46695352e-01 4.13740635e-01 1.69079334e-01 -9.09387708e-01
3.45805943e-01 -6.07892394e-01 1.59528982e-02 3.85803699e-01
3.37421782e-02 -3.12685490e-01 2.27935221e-02 -2.59589940e-01
1.26144826e+00 3.21311772e-01 -8.11235547e-01 -4.94632244e-01
7.22320735e-01 9.10437331e-02 4.77518767e-01 3.47726852e-01
3.88280064e-01 7.60500371e-01 5.46593249e-01 -6.01838708e-01
-1.12515521e+00 -8.01923633e-01 2.40808234e-01 8.23744655e-01
-2.31537566e-01 4.63021040e-01 -1.02830458e+00 -2.61326045e-01
-1.48284793e-01 9.99017894e-01 -1.00622463e+00 -7.15464652e-01
-7.35714912e-01 -1.00104284e+00 3.28390419e-01 7.88961709e-01
-6.38428107e-02 -1.23965597e+00 -8.79558980e-01 3.09232771e-01
3.56882542e-01 -4.33749825e-01 3.59371483e-01 5.94102144e-01
-1.07956624e+00 -1.07225108e+00 -5.47265708e-01 -5.94501019e-01
5.41054666e-01 -5.54601967e-01 9.66405213e-01 3.80196631e-01
-3.66061240e-01 4.13201839e-01 -4.27612811e-01 -5.61736405e-01
-7.60427654e-01 -5.78487925e-02 1.74906459e-02 -2.51074824e-02
-2.51305848e-01 -4.42704707e-01 -6.35007679e-01 2.55688876e-01
-1.11799109e+00 -6.97209120e-01 5.44812620e-01 8.91976416e-01
2.05157280e-01 8.21577787e-01 5.14069796e-01 -7.99827993e-01
7.61612594e-01 -5.41723728e-01 -3.10603201e-01 -7.02937692e-02
-7.06507862e-01 -1.10001273e-01 7.03188062e-01 -1.21497512e-01
-1.24124289e+00 -3.37930441e-01 -7.11829126e-01 -5.02817214e-01
-2.68862963e-01 1.28605378e+00 -8.37892592e-02 1.05429307e-01
7.97318578e-01 -3.21249008e-01 -1.36073887e-01 -6.86958611e-01
-3.03374946e-01 1.54729739e-01 6.09481812e-01 -7.70077705e-01
8.00782681e-01 -1.21270167e-02 3.81804585e-01 -6.46669269e-01
-6.21097326e-01 1.12145573e-01 -6.93213165e-01 -8.63560617e-01
6.11992896e-01 -4.86912578e-01 -7.34359503e-01 8.28995407e-01
-1.16739297e+00 -5.46360552e-01 -6.64552093e-01 5.68057716e-01
-3.89501303e-01 1.50049403e-01 -1.05736053e+00 -1.23921442e+00
-4.21038061e-01 -1.19937038e+00 6.85316443e-01 -2.21064519e-02
1.81766331e-01 -1.33375788e+00 3.38405430e-01 1.32126912e-01
3.23151082e-01 9.27863061e-01 1.22333801e+00 -5.17090142e-01
3.89860459e-02 -6.78797662e-01 2.92855561e-01 1.03796852e+00
-3.13961983e-01 4.58955765e-01 -1.08308923e+00 -2.86181122e-01
6.87599480e-01 -1.81304336e-01 8.89013410e-01 6.78548217e-01
9.02944744e-01 8.24232325e-02 1.03724629e-01 -8.65043607e-03
1.96150839e+00 3.44690353e-01 6.26080155e-01 1.50856018e-01
6.20848656e-01 9.95090008e-01 5.14558017e-01 4.08604532e-01
-1.99232772e-01 3.16020221e-01 1.17541885e+00 -1.97911233e-01
5.79363704e-02 3.97450179e-01 5.19616902e-01 9.27871585e-01
-5.94885528e-01 -4.93597567e-01 -1.12474501e+00 5.89815140e-01
-1.41915870e+00 -8.61632705e-01 -8.13055396e-01 2.16633844e+00
4.07070965e-01 7.59792864e-01 2.77514886e-02 7.46077776e-01
6.14102066e-01 -3.01789701e-01 -3.58677357e-01 -1.20158637e+00
-2.84761012e-01 3.03934783e-01 3.72141242e-01 5.99732637e-01
-4.49184984e-01 6.08955994e-02 6.21937323e+00 5.71062684e-01
-9.60030794e-01 2.24511147e-01 4.95416939e-01 -1.29897147e-02
-1.29405066e-01 -1.34467065e-01 -4.25682664e-01 5.05498707e-01
1.57827175e+00 6.20806396e-01 9.43573713e-02 5.10937989e-01
4.92074847e-01 -2.33846292e-01 -8.84501040e-01 2.02199623e-01
2.94554029e-02 -1.12229133e+00 -9.70200449e-02 6.49750754e-02
5.01703084e-01 1.18015543e-01 7.64725450e-03 -4.16924618e-03
8.25941283e-03 -8.59685779e-01 8.51558208e-01 1.42787707e+00
5.27961552e-01 -1.02105045e+00 1.66506577e+00 2.20472306e-01
-5.82627892e-01 -6.73067331e-01 -3.36899191e-01 -2.44407237e-01
4.56116587e-01 1.06337869e+00 -5.35311818e-01 1.04083395e+00
9.79457140e-01 5.00833690e-01 -3.34082842e-01 7.88606524e-01
1.10083528e-01 1.00192177e+00 -4.41811323e-01 1.00181162e-01
3.76733452e-01 -9.49629545e-02 4.12464291e-01 1.02419198e+00
5.59754431e-01 -2.72903621e-01 -3.14797103e-01 8.43781531e-01
4.91488904e-01 -5.41364133e-01 -5.36360621e-01 4.72421914e-01
-1.28665298e-01 1.00868571e+00 -5.21770835e-01 6.02982566e-02
1.23631284e-01 3.89702648e-01 -3.07428271e-01 4.26432103e-01
-9.00247276e-01 -4.51834977e-01 1.94416016e-01 5.38525343e-01
2.70895004e-01 -2.79710859e-01 -3.68579805e-01 -3.41209583e-02
-2.43628919e-01 -2.16107339e-01 1.46545559e-01 -1.45236158e+00
-1.31841052e+00 8.90576303e-01 4.37388688e-01 -1.03512168e+00
-2.66422957e-01 -1.21696460e+00 -1.09459245e+00 1.04158580e+00
-1.05004287e+00 -1.05867386e+00 -1.71428055e-01 9.96636599e-02
4.53078449e-01 -1.32948697e-01 4.00709450e-01 2.78846025e-01
-7.66147494e-01 3.60594615e-02 2.39020601e-01 -1.96649238e-01
3.35306704e-01 -1.17751181e+00 8.35146457e-02 7.55877674e-01
-5.43968499e-01 2.66042858e-01 1.32697713e+00 -8.74333143e-01
-1.08541584e+00 -8.82460177e-01 5.51356614e-01 -2.26605803e-01
6.59223318e-01 -8.79632533e-02 -1.21622086e+00 4.10962999e-02
4.36416805e-01 -4.51512069e-01 3.81607711e-01 -2.26676375e-01
3.56939793e-01 -3.70028436e-01 -1.13060749e+00 5.33141643e-02
3.15644652e-01 -1.92597777e-01 -7.51707911e-01 3.75562638e-01
5.03225863e-01 -3.33587706e-01 -1.23593163e+00 6.67089164e-01
3.35835576e-01 -1.24155009e+00 8.92203093e-01 -1.62766591e-01
7.07022488e-01 -7.23723173e-02 2.19823062e-01 -1.30809999e+00
-3.56314063e-01 -1.59641519e-01 -2.12590322e-01 1.01147652e+00
4.42188770e-01 -6.59375191e-01 6.13004148e-01 3.13766658e-01
-7.07601666e-01 -1.28041208e+00 -8.54882956e-01 -5.61407447e-01
2.10504293e-01 -4.53076452e-01 3.11465621e-01 2.39282176e-01
-8.02151561e-01 -3.73070799e-02 -7.06072152e-02 4.28255230e-01
2.12251708e-01 -4.47864175e-01 -1.29970104e-01 -1.68349755e+00
-1.79792240e-01 -2.57469028e-01 -2.78794169e-01 -9.16174799e-02
-1.63358122e-01 -1.50167778e-01 2.56967008e-01 -1.67017674e+00
-4.73888725e-01 -2.73348898e-01 -7.03681350e-01 2.83775151e-01
3.25774282e-01 4.19512391e-02 -2.70569891e-01 2.14357432e-02
4.69350338e-01 4.23077077e-01 9.18254673e-01 -2.33285241e-02
1.54750958e-01 1.68177143e-01 -2.55019963e-01 6.25453234e-01
7.25426316e-01 -7.09441662e-01 -3.85957986e-01 -4.44205791e-01
1.02632117e+00 1.41449511e-01 9.10688460e-01 -1.60470319e+00
1.08674981e-01 3.65983069e-01 2.49747694e-01 -1.11952484e+00
3.56958121e-01 -1.13793981e+00 3.67168695e-01 8.58536839e-01
-1.43427961e-02 5.05319953e-01 3.80206138e-01 5.46136081e-01
-4.03092623e-01 -9.03023660e-01 5.92395782e-01 -3.89452488e-03
-5.46004996e-02 -6.06904961e-02 -8.74889672e-01 -5.97089350e-01
5.72108567e-01 -5.49603105e-01 -1.58364847e-02 -8.74684975e-02
-9.82030690e-01 -1.53358638e-01 2.43282333e-01 -2.85893418e-02
7.13065326e-01 -8.33556414e-01 -9.31335211e-01 6.29419880e-03
-5.60642302e-01 1.17737629e-01 9.74965394e-01 6.76176667e-01
-1.11047173e+00 7.61840716e-02 -3.17311913e-01 -4.64360744e-01
-6.60875916e-01 5.37284076e-01 8.42101634e-01 -3.42193931e-01
-7.12757409e-02 1.01718068e+00 -2.59981632e-01 -1.89729646e-01
-3.11477482e-01 -7.30431020e-01 -4.47159022e-01 1.86720267e-01
-2.33052507e-01 8.58834445e-01 7.67354667e-01 -5.12698352e-01
-1.76457122e-01 8.68519604e-01 3.78264517e-01 9.37789753e-02
1.78709638e+00 1.38469905e-01 -2.69981444e-01 9.58788753e-01
9.20611501e-01 -3.47734839e-01 -1.28574979e+00 4.83564258e-01
-1.71683475e-01 6.84464335e-01 3.96678954e-01 -1.27737546e+00
-1.35470688e+00 1.16537261e+00 8.63044858e-01 6.12237036e-01
1.51096058e+00 -5.41051745e-01 8.50358903e-01 3.41253549e-01
-6.64483756e-02 -1.43226850e+00 -1.16101995e-01 5.82937002e-01
1.25946200e+00 -4.16552871e-01 2.21963599e-01 3.24513800e-02
-4.53139067e-01 1.47422528e+00 4.20608640e-01 -6.82307899e-01
1.12343788e+00 4.87873346e-01 -1.33834600e-01 -8.35074544e-01
-6.86638176e-01 3.06962073e-01 2.39309773e-01 2.95378327e-01
2.04002455e-01 -3.17121089e-01 -1.08010761e-01 8.58947515e-01
7.01548755e-02 -2.23459274e-01 6.48858845e-01 1.12150586e+00
-6.53793693e-01 -8.47591639e-01 -7.69975662e-01 3.91702265e-01
-2.33027667e-01 7.72590786e-02 -1.91067517e-01 7.57482171e-01
5.20436466e-01 8.89440298e-01 -1.48309663e-01 -6.02393925e-01
6.14281118e-01 7.19384626e-02 1.02810599e-01 -4.20090020e-01
-1.22336853e+00 -5.77184046e-03 9.95037630e-02 -2.28034928e-01
-2.78960407e-01 -5.26083231e-01 -1.26230168e+00 5.90339005e-02
-4.55411911e-01 3.18640351e-01 1.02397311e+00 9.11948144e-01
1.44875586e-01 1.30216587e+00 8.49773228e-01 -1.34443605e+00
-2.96966553e-01 -1.22903979e+00 -1.06428671e+00 -4.80675511e-03
3.39166522e-01 -1.08912599e+00 -6.54025197e-01 2.69025475e-01] | [6.725043773651123, 2.4964754581451416] |
3dc55d07-ccc9-418a-af64-f9d95e980a18 | total-energy-shaping-with-neural | 2112.12999 | null | https://arxiv.org/abs/2112.12999v2 | https://arxiv.org/pdf/2112.12999v2.pdf | Total Energy Shaping with Neural Interconnection and Damping Assignment -- Passivity Based Control | In this work we exploit the universal approximation property of Neural Networks (NNs) to design interconnection and damping assignment (IDA) passivity-based control (PBC) schemes for fully-actuated mechanical systems in the port-Hamiltonian (pH) framework. To that end, we transform the IDA-PBC method into a supervised learning problem that solves the partial differential matching equations, and fulfills equilibrium assignment and Lyapunov stability conditions. A main consequence of this, is that the output of the learning algorithm has a clear control-theoretic interpretation in terms of passivity and Lyapunov stability. The proposed control design methodology is validated for mechanical systems of one and two degrees-of-freedom via numerical simulations. | ['Bayu Jayawardhana', 'Rodolfo Reyes-Baez', 'Santiago Sanchez-Escalonilla'] | 2021-12-24 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [-7.52690760e-03 8.76249552e-01 -4.39124584e-01 2.95447737e-01
2.74281621e-01 -4.81109560e-01 4.25878674e-01 -1.47834435e-01
-3.12663913e-01 1.07631767e+00 -3.33669245e-01 -4.22035545e-01
-1.03023124e+00 -5.66544056e-01 -5.14501452e-01 -1.03990710e+00
-2.57570416e-01 1.18380018e-01 -1.49184644e-01 -7.06686974e-01
-2.76948184e-01 5.88682830e-01 -9.21889603e-01 -8.49778771e-01
5.96978426e-01 8.42851222e-01 -1.97595209e-01 5.79156041e-01
9.73030090e-01 6.68740988e-01 1.99721187e-01 3.49412888e-01
3.62348974e-01 -2.96380371e-01 -1.00428152e+00 -1.75147876e-01
-2.40913451e-01 -6.51945397e-02 -5.80117881e-01 9.43461001e-01
4.99187678e-01 6.42785192e-01 1.05807829e+00 -1.26009488e+00
-3.61350298e-01 6.43292248e-01 -2.55824788e-03 -1.65600061e-01
-1.89031437e-01 8.98322538e-02 1.02805185e+00 -8.66227090e-01
5.37236094e-01 9.34451640e-01 7.83802271e-01 7.31268644e-01
-1.51674604e+00 -2.34009922e-01 -3.84633392e-01 -7.02687353e-02
-1.40647936e+00 1.47414207e-03 9.78540719e-01 -5.81804097e-01
6.26318514e-01 2.92307496e-01 7.67933190e-01 6.82561934e-01
7.02279687e-01 3.06709856e-01 6.97207630e-01 -5.07254660e-01
4.09504086e-01 1.44343913e-01 2.95505166e-01 7.79888868e-01
4.32012081e-01 5.21440148e-01 1.88151076e-01 -5.26764579e-02
1.23081052e+00 -3.09453815e-01 -3.44372332e-01 -7.53362775e-01
-9.63264048e-01 9.53762054e-01 5.81969678e-01 4.15963739e-01
-5.31464577e-01 1.73620805e-01 2.56560147e-01 5.73261678e-01
-8.72831792e-02 9.28675532e-01 -2.39259452e-01 5.18728256e-01
-2.35427797e-01 3.55337024e-01 1.01414824e+00 7.10555315e-01
5.18201530e-01 6.84967279e-01 2.73401737e-01 5.31956375e-01
4.08051342e-01 6.33144259e-01 1.13006473e-01 -1.02360702e+00
-2.35078067e-01 4.03218120e-01 1.95769653e-01 -1.24118114e+00
-8.17496538e-01 -6.11697316e-01 -1.51253068e+00 4.60264087e-01
1.62006557e-01 -8.32522511e-01 -1.72410890e-01 1.98170495e+00
3.47890615e-01 -3.35881919e-01 1.77079812e-01 1.05844879e+00
1.11010924e-01 7.94521093e-01 -2.63551354e-01 -5.33355117e-01
7.67331123e-01 -3.12420011e-01 -9.42924380e-01 3.22785258e-01
3.39456499e-01 -6.38951138e-02 6.20914459e-01 2.34457806e-01
-1.25614119e+00 -5.38410485e-01 -1.38499081e+00 4.59767222e-01
-2.14606017e-01 2.39373207e-01 -6.23616427e-02 -6.04453683e-03
-9.75502968e-01 1.17509627e+00 -8.35205734e-01 -1.74760029e-01
-5.54439962e-01 7.90318847e-01 -1.09744802e-01 1.27305901e+00
-1.62766135e+00 1.11010873e+00 7.28482544e-01 4.80606288e-01
-5.04658759e-01 -8.14675450e-01 -4.70747799e-01 -4.11140844e-02
3.72623324e-01 -8.33361626e-01 1.01544702e+00 -6.24723196e-01
-1.94300246e+00 2.27990896e-01 4.47381854e-01 -4.81707722e-01
4.27115589e-01 1.81756854e-01 -8.82897601e-02 4.42399830e-01
-5.43215871e-01 1.33614570e-01 9.96579885e-01 -1.20634794e+00
1.78055204e-02 9.24862921e-02 1.19401455e-01 -1.26142532e-01
-6.64825201e-01 -7.67770350e-01 6.16792381e-01 -5.81272662e-01
2.21488953e-01 -8.85363102e-01 -3.38439673e-01 1.38433933e-01
-6.45084381e-01 -1.81182459e-01 8.86728287e-01 -3.94980282e-01
1.29858804e+00 -1.60384035e+00 1.00186396e+00 7.62453198e-01
2.08522946e-01 3.12081575e-01 2.80470371e-01 9.58051682e-01
-2.25260869e-01 -8.08661506e-02 -5.04411310e-02 1.54072076e-01
1.23321772e-01 1.80325747e-01 -3.30982625e-01 5.74424505e-01
4.12026972e-01 4.70119059e-01 -5.54571629e-01 -4.79122221e-01
3.43925118e-01 4.49165225e-01 -4.88129944e-01 9.99778509e-02
5.54151535e-02 6.97030008e-01 -8.46053362e-01 1.12714872e-01
3.22817862e-01 -1.29264295e-01 4.58174706e-01 -4.66476113e-01
-6.75573885e-01 -2.01415405e-01 -1.34091508e+00 8.49819839e-01
-5.76781511e-01 3.03472638e-01 8.37388694e-01 -1.56850469e+00
9.16049659e-01 7.19528377e-01 7.40582585e-01 -2.94331551e-01
7.12002993e-01 3.66624385e-01 2.78967712e-02 -4.42781359e-01
6.73261359e-02 -3.34761143e-01 -1.16729975e-01 2.19612017e-01
2.46272877e-01 -8.16740096e-02 1.09767281e-01 -1.46931231e-01
7.78365433e-01 -3.42208058e-01 4.30922300e-01 -1.10283327e+00
1.32457006e+00 1.12795942e-02 5.22304833e-01 3.63895535e-01
-1.27658471e-01 -1.59022778e-01 5.31867266e-01 -1.79188624e-01
-1.52644789e+00 -8.98004949e-01 -2.98341274e-01 4.04479861e-01
-7.92425126e-02 4.29291248e-01 -7.60376930e-01 4.59559768e-01
1.04462676e-01 2.75599360e-01 -5.07741451e-01 -6.99790716e-01
-8.40821862e-01 -1.26842149e-02 3.53059232e-01 3.41355234e-01
2.29552850e-01 -5.69231987e-01 -6.76485598e-01 4.34762120e-01
2.09813297e-01 -7.40436554e-01 -3.17994086e-03 3.64888638e-01
-7.29312658e-01 -1.17268753e+00 -6.52588308e-01 -1.24362576e+00
5.96019566e-01 -5.86355329e-01 4.43735063e-01 -1.73780859e-01
-1.05781049e-01 5.86334109e-01 1.74200192e-01 5.55459410e-02
-7.47502208e-01 2.77858645e-01 7.50434220e-01 -7.17434958e-02
-6.94344163e-01 -8.04239154e-01 -3.73764068e-01 5.14243782e-01
-6.97516203e-01 3.96137051e-02 4.87347811e-01 1.13483346e+00
6.41514659e-01 3.51323247e-01 1.00945306e+00 -1.63793802e-01
8.04848552e-01 -7.49020576e-02 -1.35177469e+00 -1.69073455e-02
-6.81228101e-01 1.22012496e-01 1.30048156e+00 -4.72061485e-01
-8.73148501e-01 3.44321996e-01 -4.55584861e-02 -4.12291259e-01
2.52600729e-01 6.40683532e-01 2.70816032e-03 -4.08420563e-01
4.23100233e-01 2.43959904e-01 6.68297231e-01 -2.20526218e-01
2.95350492e-01 4.40499842e-01 7.56047964e-01 -6.59838498e-01
1.28847122e+00 3.81010436e-02 1.03424001e+00 -1.08015800e+00
-1.34832352e-01 -1.52076542e-01 -7.58561552e-01 -5.88151753e-01
7.49049425e-01 -3.43754143e-01 -1.66659188e+00 4.70659465e-01
-9.44842637e-01 -4.06306267e-01 -4.55206543e-01 6.54491544e-01
-9.95565176e-01 8.92938152e-02 -9.98798132e-01 -1.49610472e+00
-4.04384524e-01 -7.91156590e-01 2.09106117e-01 2.23496541e-01
-1.32660806e-01 -1.46230447e+00 2.54065633e-01 -5.20889878e-01
3.74848306e-01 7.06987798e-01 1.03906512e+00 3.87791470e-02
-2.13097021e-01 -4.60863531e-01 2.90660292e-01 4.85996991e-01
-4.24944341e-01 3.72169763e-02 -4.99079913e-01 -6.37282550e-01
2.34872580e-01 -1.92994654e-01 1.59685433e-01 4.98240829e-01
2.76457042e-01 -7.42154837e-01 -2.36889467e-01 -4.17537205e-02
2.03722668e+00 1.91409588e-01 -2.37322580e-02 -3.82590406e-02
5.46061516e-01 7.08305597e-01 7.86809400e-02 4.01070237e-01
-7.53882974e-02 5.66748857e-01 5.51141143e-01 6.92925304e-02
3.83194327e-01 -5.28904535e-02 2.61667937e-01 1.06285036e+00
-2.07843751e-01 2.74044126e-01 -7.72321165e-01 4.64408070e-01
-1.79418731e+00 -7.77378082e-01 -2.88336873e-01 2.17440581e+00
8.48902345e-01 -1.29999533e-01 3.66106719e-01 6.28977776e-01
9.85905766e-01 -3.87164861e-01 -4.90401685e-01 -5.83719313e-01
-1.46368831e-01 3.91264558e-02 7.10862458e-01 7.61944771e-01
-1.10049188e+00 -5.82215711e-02 6.05142450e+00 5.10422111e-01
-1.23764229e+00 -2.32800514e-01 2.35135481e-01 5.03283501e-01
3.91964316e-01 -1.84406787e-01 -5.66450059e-01 1.48215130e-01
9.85831976e-01 -2.87446588e-01 5.47451675e-01 8.42034459e-01
7.28358984e-01 2.47945577e-01 -9.35799837e-01 3.13273251e-01
-5.65661550e-01 -1.21830845e+00 -4.41629082e-01 4.00640722e-03
7.91743159e-01 -5.32279611e-01 2.50412583e-01 2.85525024e-02
-3.49055856e-01 -5.22173941e-01 7.42629468e-01 7.44797468e-01
4.67890114e-01 -6.68593347e-01 5.18496752e-01 4.69891071e-01
-1.21188521e+00 -5.52000284e-01 -1.79834232e-01 -4.13867325e-01
2.90950835e-01 3.94646794e-01 -3.32924515e-01 6.88633025e-01
-1.20657682e-01 7.51746953e-01 1.27810866e-01 7.75259078e-01
1.10734031e-01 3.90525341e-01 -4.42108095e-01 -8.62765431e-01
4.08008456e-01 -4.75050092e-01 9.49899912e-01 6.26389205e-01
-1.35396063e-01 9.86819938e-02 2.89388895e-02 1.29405832e+00
3.63966286e-01 -9.06076580e-02 -5.35476804e-01 1.63399801e-01
2.62116700e-01 1.24769342e+00 -4.07277226e-01 -2.23872233e-02
2.95013338e-01 3.56165431e-02 -1.71009272e-01 4.17331904e-01
-6.62376642e-01 -8.96183670e-01 5.87956548e-01 5.73585890e-02
1.53807610e-01 -3.28172952e-01 -2.38139093e-01 -8.52665782e-01
1.29881371e-02 -2.05270335e-01 -8.59189406e-03 -3.95514011e-01
-1.28129041e+00 1.28424242e-01 2.21624687e-01 -1.61846292e+00
-6.29043341e-01 -8.36939275e-01 -6.47712946e-01 7.40566373e-01
-1.27817738e+00 -7.62986124e-01 3.78338516e-01 5.68297565e-01
-2.38466710e-01 1.42237246e-01 8.23687792e-01 3.98397565e-01
-8.40519309e-01 1.04684472e-01 4.52665001e-01 2.57669628e-01
-2.32846662e-02 -1.38408136e+00 -5.31506300e-01 7.42565393e-01
-1.13866436e+00 6.83541715e-01 1.14109850e+00 -2.04480484e-01
-1.83545935e+00 -1.04439676e+00 7.74986327e-01 4.40721840e-01
1.16394579e+00 -3.62400562e-02 -8.16079199e-01 2.50756890e-01
3.18942875e-01 -1.51252925e-01 -1.63044274e-01 -3.36427569e-01
3.11814368e-01 -2.05935404e-01 -1.08522272e+00 6.78939760e-01
4.84788269e-01 -3.60300541e-01 -4.85671669e-01 2.58241057e-01
5.17987132e-01 -1.70321390e-01 -1.40078497e+00 3.99772525e-01
7.39271998e-01 -3.23678181e-02 1.10311794e+00 -6.21811211e-01
4.86225188e-01 -3.60653073e-01 1.50577605e-01 -1.22439063e+00
-3.90397340e-01 -1.06606448e+00 -2.88288534e-01 7.41824150e-01
4.89737242e-01 -8.55084419e-01 4.22040850e-01 3.53721321e-01
-1.95390597e-01 -1.14192688e+00 -1.14790857e+00 -1.17266047e+00
6.31674051e-01 2.43703634e-01 -2.09619492e-01 9.03756857e-01
8.72818708e-01 5.42874634e-01 -5.84448159e-01 4.89433557e-01
6.39392316e-01 -1.90729007e-01 1.80270195e-01 -1.33334684e+00
-5.99429548e-01 -5.75575948e-01 -2.10732892e-01 -5.60523570e-01
3.82308543e-01 -5.59145689e-01 1.18129939e-01 -1.29292059e+00
-5.99152565e-01 -2.26588994e-01 -1.39559060e-01 3.87922823e-01
4.03103501e-01 8.31488296e-02 -1.29932940e-01 2.91567683e-01
-1.05439648e-01 6.60349607e-01 1.19519758e+00 1.56185990e-02
-5.53983569e-01 1.92133322e-01 2.34313652e-01 5.03760576e-01
1.05574691e+00 -3.34537700e-02 -6.53911829e-01 1.27084211e-01
2.40534306e-01 5.18323958e-01 7.19308496e-01 -1.37533009e+00
3.76501888e-01 -1.13135390e-01 -2.12601647e-01 -5.36080480e-01
2.03623652e-01 -1.01296067e+00 3.20663571e-01 1.41650665e+00
-6.67522728e-01 -1.59130365e-01 -2.07962140e-01 4.61199909e-01
-1.60264179e-01 -2.01498210e-01 1.14744651e+00 4.47062373e-01
-1.86962977e-01 -1.59287348e-01 -9.23473358e-01 -2.23961696e-01
6.91936195e-01 -2.69447360e-03 -8.23140517e-02 -3.02385151e-01
-1.00728762e+00 3.99926215e-01 -7.58621544e-02 -2.79100418e-01
1.92345202e-01 -1.36487818e+00 -2.07134247e-01 3.26372802e-01
-3.13978493e-01 -4.61231440e-01 2.44768515e-01 1.25696135e+00
-3.55681747e-01 8.57164085e-01 -4.62668687e-01 -4.99269605e-01
-6.17565453e-01 2.27363691e-01 7.28404224e-01 -6.20079301e-02
-4.14522678e-01 9.33823287e-02 -5.60839176e-01 -6.34263575e-01
1.03101231e-01 -3.77484679e-01 -3.36685210e-01 -2.98695207e-01
-1.88852653e-01 5.95430136e-01 -5.91417216e-02 -5.38099587e-01
7.65079781e-02 6.98179007e-01 6.50390625e-01 -3.01329285e-01
1.34502268e+00 -1.55936182e-01 -1.84209704e-01 4.03402686e-01
1.24879420e+00 -6.88887835e-01 -1.08893669e+00 -1.67731732e-01
-2.27298468e-01 5.69551289e-01 1.27799064e-01 -4.35275972e-01
-9.76470947e-01 5.42972803e-01 4.30592626e-01 7.65533686e-01
9.89131689e-01 -4.37633365e-01 5.47020853e-01 8.97742569e-01
1.95995644e-01 -1.43822753e+00 -1.91014260e-01 6.28586292e-01
8.68523836e-01 -4.20865208e-01 -1.45681053e-01 -5.33557653e-01
1.64067283e-01 1.52407217e+00 4.81617719e-01 -8.63979518e-01
1.18998790e+00 1.03250995e-01 -4.67534065e-01 2.24208459e-01
-5.35774946e-01 1.38597116e-01 2.90414155e-01 4.89324629e-01
2.57558346e-01 -2.50326823e-02 -1.00838113e+00 4.97497350e-01
1.53374434e-01 3.64840254e-02 6.53992891e-01 9.88775253e-01
-4.48140264e-01 -8.13056827e-01 -1.15125895e-01 -1.91001832e-01
-3.41631770e-02 7.26828575e-01 -1.94792107e-01 1.30874503e+00
-2.67012596e-01 7.63362825e-01 -2.14129135e-01 -3.96275878e-01
5.82498014e-01 3.25136050e-03 1.62541091e-01 -2.10763067e-01
-4.74530548e-01 2.89742202e-01 -4.86607403e-02 -1.81844175e-01
-5.21454990e-01 -1.95735827e-01 -1.44735432e+00 -2.09717467e-01
-6.82224810e-01 3.47289503e-01 5.47680557e-01 6.74654961e-01
1.56605646e-01 6.54987454e-01 9.21043634e-01 -8.39203298e-01
-1.29173422e+00 -8.50632489e-01 -7.55936325e-01 -1.39124051e-01
6.36362314e-01 -9.15644765e-01 -6.26832902e-01 -1.69656456e-01] | [5.477433204650879, 2.654442071914673] |
74f35c44-bb84-4be3-81c2-7ccdcfab0b9a | bundlerecon-ray-bundle-based-3d-neural | 2305.07342 | null | https://arxiv.org/abs/2305.07342v1 | https://arxiv.org/pdf/2305.07342v1.pdf | BundleRecon: Ray Bundle-Based 3D Neural Reconstruction | With the growing popularity of neural rendering, there has been an increasing number of neural implicit multi-view reconstruction methods. While many models have been enhanced in terms of positional encoding, sampling, rendering, and other aspects to improve the reconstruction quality, current methods do not fully leverage the information among neighboring pixels during the reconstruction process. To address this issue, we propose an enhanced model called BundleRecon. In the existing approaches, sampling is performed by a single ray that corresponds to a single pixel. In contrast, our model samples a patch of pixels using a bundle of rays, which incorporates information from neighboring pixels. Furthermore, we design bundle-based constraints to further improve the reconstruction quality. Experimental results demonstrate that BundleRecon is compatible with the existing neural implicit multi-view reconstruction methods and can improve their reconstruction quality. | ['Jianke Zhu', 'Weikun Zhang'] | 2023-05-12 | null | null | null | null | ['neural-rendering'] | ['computer-vision'] | [ 2.08649486e-02 -1.55599192e-01 -3.98096442e-02 -3.30919802e-01
-3.71604770e-01 -1.14188731e-01 5.46215177e-01 -1.89852089e-01
-2.04515785e-01 8.24114382e-01 3.78431559e-01 1.48898035e-01
2.12251201e-01 -1.40955520e+00 -7.63104796e-01 -6.54403389e-01
5.00929356e-01 3.34641822e-02 5.44285893e-01 -7.94788003e-02
1.26464874e-01 4.73130018e-01 -1.40664351e+00 4.56872523e-01
1.10175049e+00 8.02862763e-01 3.04999262e-01 2.79976904e-01
-2.35791832e-01 5.25651574e-01 -1.10851079e-01 -2.18580708e-01
1.87076986e-01 -5.41852236e-01 -2.80722141e-01 1.37131037e-02
5.32521009e-01 -8.94706249e-01 -3.52975607e-01 1.05523086e+00
6.60557985e-01 1.30365193e-01 2.29611769e-01 -7.09698975e-01
-6.30712271e-01 3.77367944e-01 -8.09943199e-01 -3.85954976e-01
1.35291070e-01 -9.47917253e-02 8.45237195e-01 -1.16609466e+00
7.87752867e-01 1.21870315e+00 7.67145872e-01 5.47004223e-01
-1.41698205e+00 -7.08029211e-01 4.05566812e-01 -1.24811940e-01
-1.12803245e+00 -2.50039637e-01 1.10888374e+00 -3.07024211e-01
4.53188837e-01 2.78282523e-01 1.26466227e+00 7.44714677e-01
1.53535292e-01 7.49011815e-01 1.34993565e+00 -7.96547458e-02
1.55679911e-01 -2.15454012e-01 -2.21195891e-02 6.85901523e-01
-2.97464691e-02 2.84284025e-01 -7.82094240e-01 -2.00986952e-01
1.59432089e+00 1.95872650e-01 -4.91293937e-01 -7.81438589e-01
-1.21899509e+00 8.58351350e-01 5.33264816e-01 3.86316068e-02
-4.33224201e-01 2.93611854e-01 2.56114483e-01 -1.68669835e-01
9.14501011e-01 1.32626444e-01 -1.39342442e-01 1.06144227e-01
-1.16146719e+00 3.61543179e-01 4.61268544e-01 7.62434363e-01
1.11455810e+00 1.47192463e-01 1.40345544e-01 1.03878677e+00
5.22636473e-01 1.60474911e-01 8.39827955e-03 -1.51130199e+00
2.53751278e-01 5.84550142e-01 -3.60451490e-02 -1.16448486e+00
-1.11129507e-01 -5.93170762e-01 -1.09376419e+00 6.09506547e-01
-1.92288328e-02 1.90101385e-01 -6.39655113e-01 1.59974611e+00
6.14389002e-01 3.30188036e-01 -3.54927152e-01 9.86086845e-01
7.57726610e-01 7.17452288e-01 -2.82862365e-01 -2.58849084e-01
8.23386490e-01 -1.12793958e+00 -8.06251526e-01 2.11850449e-01
2.18764931e-01 -7.69349337e-01 9.41674769e-01 6.75284684e-01
-1.40203834e+00 -6.25523567e-01 -1.01848567e+00 -3.85811955e-01
1.71156853e-01 -1.61147073e-01 8.07556152e-01 3.90534282e-01
-9.84126210e-01 8.18497658e-01 -9.27620292e-01 1.02031231e-01
5.85695207e-01 7.75569603e-02 -2.03462422e-01 -2.35873803e-01
-7.92875230e-01 3.02595139e-01 1.92983244e-02 1.34165287e-01
-7.39771426e-01 -9.02189791e-01 -5.71059883e-01 -2.60853022e-01
1.26236722e-01 -1.15029287e+00 8.47646177e-01 -9.05185580e-01
-1.59248388e+00 5.13198316e-01 -2.59932160e-01 3.01733799e-02
7.78246701e-01 -4.88030076e-01 1.60056606e-01 2.00108573e-01
-2.20275614e-02 6.98522210e-01 5.36206603e-01 -1.80564284e+00
-3.98999661e-01 -4.14058864e-01 3.98837894e-01 3.49374056e-01
-2.05051705e-01 -3.99041116e-01 -6.39607430e-01 -7.66919672e-01
6.12326622e-01 -6.59284472e-01 -4.73454446e-01 6.32432818e-01
-3.34968925e-01 1.89772278e-01 9.00354505e-01 -6.24489784e-01
1.16967452e+00 -2.15013409e+00 4.45953012e-01 -9.15964022e-02
4.76177454e-01 -4.50052321e-02 -9.42836457e-04 3.76357675e-01
2.01226026e-01 8.17068517e-02 -4.44469184e-01 -7.35139489e-01
-3.36566061e-01 5.20552874e-01 -2.41493270e-01 4.62208062e-01
-2.91768014e-01 6.35897994e-01 -8.58155012e-01 -5.30688941e-01
3.98950994e-01 1.04496992e+00 -1.06696486e+00 2.42101803e-01
-3.15753877e-01 9.46135998e-01 -7.10330784e-01 4.15441096e-01
1.08252299e+00 -2.34759867e-01 -1.58177391e-02 -4.56995696e-01
-4.95548517e-01 3.74051034e-01 -1.27545142e+00 2.23127604e+00
-3.99249882e-01 2.35939115e-01 4.13252473e-01 -6.31982088e-01
5.94012082e-01 3.97347987e-01 7.79034615e-01 -5.11181474e-01
-1.97704136e-01 1.01094864e-01 -3.42861444e-01 -1.41736612e-01
6.75141752e-01 -1.74267530e-01 5.76967955e-01 4.60896194e-01
-5.23082197e-01 -2.14635387e-01 -1.98999867e-01 1.53518379e-01
6.83064461e-01 8.12532365e-01 9.67423245e-02 -9.24988836e-02
1.89420298e-01 -1.80389002e-01 9.49329197e-01 2.90079206e-01
1.14136845e-01 1.17680573e+00 1.18121058e-01 -4.55867499e-01
-1.21935332e+00 -1.11751413e+00 -2.77734071e-01 5.88549197e-01
3.56081128e-01 -6.25584185e-01 -8.07875276e-01 -2.40498379e-01
-1.84192196e-01 4.39802021e-01 -5.17791927e-01 2.57238060e-01
-8.44881594e-01 -6.60621822e-01 8.81734937e-02 5.46797037e-01
7.84681618e-01 -8.06606472e-01 -7.40204513e-01 5.53989232e-01
-3.01356167e-01 -8.29319537e-01 -4.17063773e-01 -2.65024275e-01
-1.46970713e+00 -8.67810965e-01 -7.70966291e-01 -3.68638456e-01
7.34948993e-01 5.80952585e-01 1.21600831e+00 2.37743080e-01
7.22018182e-02 2.52802789e-01 -1.39104724e-01 7.96780959e-02
-1.32585227e-01 -4.46397401e-02 -3.34431440e-01 1.40002416e-02
-4.22585547e-01 -1.13588047e+00 -9.17826653e-01 3.11229795e-01
-9.59059954e-01 8.49743426e-01 2.84273893e-01 8.24082434e-01
1.21116459e+00 -2.78419167e-01 3.02919388e-01 -9.99031603e-01
2.47209966e-01 -2.52927303e-01 -4.46116239e-01 -7.15179741e-02
-4.28750843e-01 -1.68147221e-01 5.37973702e-01 -4.15741473e-01
-1.13463497e+00 6.21499270e-02 -4.80833232e-01 -5.87943792e-01
-1.65650179e-03 5.48133969e-01 -3.44167531e-01 -2.57070601e-01
1.43335044e-01 2.99503982e-01 9.28633939e-03 -7.83988595e-01
3.18051070e-01 4.99068350e-02 2.04529446e-02 -6.04861498e-01
4.39325184e-01 8.63746822e-01 8.97308253e-03 -8.85038972e-01
-7.43708432e-01 -8.46080109e-02 -7.62405932e-01 -4.25399840e-01
8.94114971e-01 -1.01039112e+00 -5.30876160e-01 6.41910017e-01
-1.41402233e+00 -4.17075843e-01 -3.47017199e-01 4.70719486e-01
-6.27374530e-01 5.80215514e-01 -8.49970460e-01 -7.06197143e-01
-2.15156019e-01 -1.52790868e+00 1.00868416e+00 -3.18589509e-02
7.95661937e-03 -9.86056447e-01 3.37091051e-02 2.95844615e-01
4.96781707e-01 4.22436684e-01 8.99012864e-01 4.92793202e-01
-1.15842879e+00 2.86042362e-01 -2.13618606e-01 3.05644721e-01
1.11975923e-01 4.21852991e-02 -9.01346147e-01 -1.94113657e-01
2.56796479e-01 -1.15156919e-01 8.30541492e-01 5.17275274e-01
1.39015388e+00 -1.58509016e-01 -4.03708607e-01 1.07457530e+00
1.65528572e+00 1.87092870e-01 8.79627705e-01 2.87979245e-01
1.14961207e+00 4.25125211e-01 2.28919700e-01 3.97174001e-01
5.12810707e-01 7.03140676e-01 7.89458692e-01 -3.01262796e-01
-2.44272321e-01 -4.41986650e-01 -5.19055612e-02 1.35644674e+00
-4.83365774e-01 2.78038215e-02 -3.33108217e-01 1.62099779e-01
-1.87839484e+00 -8.91566515e-01 -2.83239722e-01 2.06924033e+00
8.17298591e-01 -2.16223136e-01 -3.73540893e-02 8.87356326e-02
3.73705089e-01 5.85585833e-01 -6.73929811e-01 8.99081528e-02
-1.97237790e-01 -1.45071745e-01 2.00798735e-01 5.85338354e-01
-6.68736041e-01 7.17304945e-01 6.88266373e+00 6.26133502e-01
-1.21127677e+00 2.64849693e-01 6.71067894e-01 -1.21361718e-01
-1.04798400e+00 1.40474081e-01 -7.13121355e-01 2.78561383e-01
1.70998529e-01 2.39633888e-01 5.38819075e-01 6.98469698e-01
4.35690820e-01 -3.03042144e-01 -9.11605358e-01 8.29050958e-01
-1.14703309e-02 -1.63601494e+00 3.56285274e-01 3.50334793e-01
8.71675789e-01 -1.16080701e-01 -6.48697987e-02 -2.40742981e-01
2.01426908e-01 -8.49898756e-01 8.65667641e-01 7.96051979e-01
6.26842201e-01 -7.79953659e-01 9.06467065e-02 5.15838921e-01
-1.30243790e+00 5.02078533e-01 -4.36693043e-01 4.68145572e-02
5.50517917e-01 9.19925392e-01 6.59502447e-02 7.13014007e-01
6.31585181e-01 1.08772779e+00 -8.87802839e-02 9.97388721e-01
-1.09746434e-01 5.08509576e-01 -1.89918190e-01 4.67421263e-01
6.22555204e-02 -9.17221069e-01 6.42193973e-01 5.95709145e-01
3.40033621e-01 1.43704399e-01 3.30837250e-01 1.25525343e+00
6.44593611e-02 1.12241492e-01 -6.13379717e-01 4.81513739e-01
2.75832027e-01 1.01834476e+00 -4.38584864e-01 -4.07083362e-01
-6.38551235e-01 1.10902309e+00 4.13472980e-01 5.98516226e-01
-7.86827803e-01 2.97988087e-01 6.85156703e-01 5.44110775e-01
1.72736704e-01 -6.11425340e-01 -6.27147496e-01 -1.18367267e+00
7.39814043e-02 -7.79878736e-01 -2.60769904e-01 -7.59185135e-01
-1.21231866e+00 5.56043804e-01 -8.47758427e-02 -1.28848708e+00
1.53580740e-01 -1.65898889e-01 -2.91225791e-01 9.60296154e-01
-1.70402968e+00 -1.23480177e+00 -2.71780521e-01 3.70863855e-01
5.96914411e-01 2.92448133e-01 6.54376745e-01 4.77256447e-01
-3.23525280e-01 8.53887647e-02 1.78069323e-01 -1.38629600e-01
5.08090794e-01 -9.31861341e-01 1.28870189e-01 6.70677483e-01
-1.53824627e-01 8.73482466e-01 3.61554086e-01 -7.26314664e-01
-1.35979867e+00 -9.29216504e-01 2.20375434e-01 1.44238278e-01
7.80961439e-02 -2.16289058e-01 -1.18571556e+00 5.48265934e-01
2.34929577e-01 2.07176864e-01 7.32437849e-01 -1.28267810e-01
-2.39709988e-01 -1.30973160e-01 -1.04298174e+00 6.17889047e-01
1.35913384e+00 -5.44789732e-01 -2.09277440e-02 -9.41547975e-02
6.66297615e-01 -5.81085801e-01 -1.09753311e+00 5.60437262e-01
7.30633080e-01 -1.64806938e+00 1.28882241e+00 -6.84623048e-02
7.41072416e-01 -4.85449165e-01 -3.29116970e-01 -1.44238806e+00
-5.20047426e-01 -2.33224630e-01 -1.65148869e-01 1.14060569e+00
-1.08405761e-01 -6.92334533e-01 9.47497368e-01 4.87108648e-01
-4.03961092e-01 -1.24369740e+00 -1.00046134e+00 -3.31458420e-01
1.43243656e-01 -3.68365020e-01 7.06904531e-01 8.82863879e-01
-6.40351295e-01 7.11958110e-02 -5.65020680e-01 1.48797482e-01
9.78647768e-01 4.38443065e-01 7.93494523e-01 -1.34093106e+00
-5.22990644e-01 -3.52474064e-01 1.91640511e-01 -1.74076807e+00
-2.37492189e-01 -6.69973373e-01 -1.18589856e-01 -2.03900814e+00
2.44816735e-01 -8.68457139e-01 2.08888873e-02 7.53745139e-02
-1.40397549e-01 3.29778910e-01 2.12738082e-01 4.46728021e-01
-2.37893805e-01 8.79233599e-01 2.00724077e+00 2.69543350e-01
-3.29596549e-02 -2.86822110e-01 -4.47595745e-01 9.79705274e-01
4.87402111e-01 -2.16279700e-01 -3.24149996e-01 -9.57817256e-01
4.05936360e-01 2.68955052e-01 4.24049854e-01 -8.90029550e-01
1.47146359e-01 -2.69950688e-01 4.65454280e-01 -1.10892367e+00
8.32689404e-01 -8.47368002e-01 6.27544880e-01 3.43345910e-01
-8.46544430e-02 1.29287630e-01 -1.05616614e-01 5.76706946e-01
-1.81577101e-01 -7.00951815e-02 9.28481102e-01 -4.80816156e-01
-3.67857158e-01 6.13967776e-01 -2.00468436e-01 -3.50646138e-01
7.85610318e-01 -4.59012419e-01 7.09920079e-02 -1.42226249e-01
-4.14811850e-01 -1.35642380e-01 1.04509079e+00 4.50769067e-02
9.12044048e-01 -1.49239016e+00 -4.69432086e-01 3.63430470e-01
-3.47928196e-01 6.60252333e-01 2.32122079e-01 7.07792819e-01
-6.00320458e-01 -1.49889395e-01 -2.09389523e-01 -6.57522440e-01
-1.10436976e+00 2.04970062e-01 4.05087352e-01 -1.51440546e-01
-9.09982979e-01 5.99455297e-01 6.90052450e-01 -5.05300105e-01
-3.50041501e-02 -2.01217443e-01 -1.36528403e-01 -3.86735685e-02
3.47285897e-01 5.25618494e-01 -3.84112626e-01 -5.93856335e-01
2.86497056e-01 9.84153271e-01 1.67082116e-01 -2.01113775e-01
1.63587129e+00 -1.66636392e-01 -5.16008556e-01 7.84526944e-01
1.02993906e+00 4.06275600e-01 -1.77078104e+00 -1.18092127e-01
-7.80678928e-01 -8.34260702e-01 3.82466882e-01 -4.08032477e-01
-1.55072868e+00 1.05863047e+00 3.31030577e-01 -6.25227997e-03
1.05725121e+00 -4.66290861e-01 1.15557122e+00 -3.07397664e-01
8.93803477e-01 -7.90241897e-01 1.20459706e-01 4.49429542e-01
8.86389196e-01 -8.31259966e-01 2.61159450e-01 -9.60809767e-01
-1.83003366e-01 1.05144477e+00 6.41665280e-01 -2.91525781e-01
7.36294568e-01 4.55927819e-01 -1.97659746e-01 -1.52349442e-01
-4.29527014e-01 3.32115829e-01 7.95644075e-02 4.27493572e-01
8.02454591e-01 -6.03826568e-02 -5.65493226e-01 1.28071606e-01
1.61953256e-01 2.03126356e-01 2.97104686e-01 8.82676005e-01
-3.03858191e-01 -1.41868961e+00 -2.90915430e-01 3.75224382e-01
-2.31762633e-01 -1.50194675e-01 9.24299434e-02 4.61996228e-01
2.83752263e-01 4.54326719e-01 1.26250595e-01 -2.01004058e-01
2.74260581e-01 -3.30031991e-01 5.44182539e-01 -6.50658011e-01
-5.60519874e-01 4.63554949e-01 -9.80994552e-02 -8.30214322e-01
-7.41438448e-01 -6.34999454e-01 -1.20961785e+00 -3.20555598e-01
-3.49031597e-01 -1.66585311e-01 6.74468577e-01 5.94389558e-01
5.15670598e-01 7.00145364e-01 4.86647725e-01 -1.17415094e+00
-4.15156186e-02 -5.69746673e-01 -5.80579042e-01 1.54715165e-01
1.53185338e-01 -7.29662478e-01 -1.36618912e-01 -1.94865875e-02] | [9.249850273132324, -3.227365016937256] |
141bb994-eda6-47d4-8da6-cb6e5600176c | realized-recurrent-conditional | 2302.08002 | null | https://arxiv.org/abs/2302.08002v1 | https://arxiv.org/pdf/2302.08002v1.pdf | Realized recurrent conditional heteroskedasticity model for volatility modelling | We propose a new approach to volatility modelling by combining deep learning (LSTM) and realized volatility measures. This LSTM-enhanced realized GARCH framework incorporates and distills modeling advances from financial econometrics, high frequency trading data and deep learning. Bayesian inference via the Sequential Monte Carlo method is employed for statistical inference and forecasting. The new framework can jointly model the returns and realized volatility measures, has an excellent in-sample fit and superior predictive performance compared to several benchmark models, while being able to adapt well to the stylized facts in volatility. The performance of the new framework is tested using a wide range of metrics, from marginal likelihood, volatility forecasting, to tail risk forecasting and option pricing. We report on a comprehensive empirical study using 31 widely traded stock indices over a time period that includes COVID-19 pandemic. | ['Robert Kohn', 'Minh-Ngoc Tran', 'Chao Wang', 'Chen Liu'] | 2023-02-16 | null | null | null | null | ['econometrics'] | ['miscellaneous'] | [-8.60560894e-01 -2.80978769e-01 -4.35030609e-02 -4.33064848e-01
-7.77576745e-01 -3.92140657e-01 1.27161944e+00 -8.00581351e-02
-4.37191635e-01 8.67732584e-01 2.41470858e-01 -6.48146510e-01
-2.66985595e-01 -1.25795841e+00 -2.31011927e-01 -6.22391760e-01
-5.46401620e-01 6.50620162e-01 -3.41848850e-01 -2.80858856e-02
4.17810440e-01 4.68520522e-01 -9.33745980e-01 -1.59967050e-01
3.12511891e-01 1.70315182e+00 -2.61056036e-01 2.49619514e-01
-5.69038689e-01 9.11778271e-01 -6.12708807e-01 -8.18010867e-01
4.51719880e-01 2.20874131e-01 -1.03832304e-01 -4.34928954e-01
-1.21372141e-01 -6.43352866e-01 -1.83453158e-01 9.22164440e-01
2.35689014e-01 2.48747960e-01 9.67882097e-01 -7.86790669e-01
-7.95489192e-01 1.13878310e+00 -5.79070151e-01 6.29888475e-01
-4.99517292e-01 4.98958789e-02 1.15680718e+00 -8.18453312e-01
1.16568260e-01 1.51381791e+00 8.21536660e-01 3.38052139e-02
-1.17936468e+00 -7.43680835e-01 2.11428832e-02 -3.06583792e-01
-7.96815872e-01 7.01106191e-02 6.76486969e-01 -7.00726092e-01
1.29549146e+00 -1.82119712e-01 7.47413993e-01 1.32883239e+00
8.97529662e-01 4.39281166e-01 1.34991109e+00 1.15313888e-01
6.16782963e-01 2.03337763e-02 9.88287404e-02 1.33714259e-01
5.60647845e-01 9.09564257e-01 -1.69068873e-01 -3.45864713e-01
1.12634885e+00 3.47528458e-01 6.48893714e-01 1.71073645e-01
-8.23619902e-01 1.30088377e+00 5.25610000e-02 2.40870088e-01
-8.77116144e-01 5.46353161e-01 5.06004691e-01 8.00174832e-01
8.90913069e-01 1.96993545e-01 -8.15249860e-01 -2.80517191e-01
-1.22111309e+00 4.86491233e-01 1.22393882e+00 4.34269160e-01
3.18059713e-01 1.29263985e+00 -3.74379367e-01 4.69645232e-01
5.49912572e-01 9.41556215e-01 8.36698413e-01 -1.00191307e+00
2.99946100e-01 -4.53296863e-02 8.88435319e-02 -8.07321906e-01
-2.87212789e-01 -3.92789066e-01 -1.09951591e+00 4.39159542e-01
1.77530661e-01 -5.33352315e-01 -8.32976282e-01 1.52136433e+00
-2.31226772e-01 4.00010884e-01 1.47557497e-01 1.60635889e-01
4.73601550e-01 7.06628263e-01 9.82330665e-02 -3.51958632e-01
1.03336859e+00 -4.21477735e-01 -7.59426236e-01 1.58446968e-01
1.15437536e-02 -4.87074673e-01 2.55494446e-01 5.14134645e-01
-1.18702030e+00 -3.90645385e-01 -7.89339542e-01 4.43743408e-01
-6.40673637e-01 -5.93877673e-01 7.82925546e-01 7.59810388e-01
-1.14693356e+00 9.92559314e-01 -8.54401529e-01 6.21980786e-01
1.94994718e-01 1.66050673e-01 1.90508395e-01 9.52262998e-01
-1.39731693e+00 8.39868903e-01 5.25229454e-01 8.60279873e-02
-9.13554490e-01 -4.86233413e-01 -7.93783784e-01 4.80349571e-01
4.55254056e-02 -7.59531736e-01 1.64928138e+00 -4.20647115e-01
-2.03214931e+00 2.87654757e-01 4.57168519e-01 -1.40576673e+00
8.92781138e-01 -2.18324035e-01 -5.23407638e-01 -4.37548876e-01
-3.45037460e-01 2.43002728e-01 1.05630493e+00 -3.42286468e-01
-5.13164937e-01 -2.43170410e-01 -5.29337049e-01 -3.83245677e-01
3.15236114e-02 -1.96142167e-01 4.85480577e-01 -1.21279240e+00
-1.44273788e-01 -5.13431370e-01 -2.91904241e-01 -8.97958100e-01
-1.93643197e-01 -9.91087332e-02 4.48825121e-01 -7.44253933e-01
1.22167003e+00 -1.67287898e+00 -5.95604360e-01 5.83405018e-01
-1.44183682e-02 -4.34505828e-02 1.77220061e-01 7.14425802e-01
-4.41714525e-02 2.06083536e-01 -1.24655694e-01 -5.63740194e-01
6.42907679e-01 1.64197654e-01 -1.29473853e+00 8.78881589e-02
2.57762093e-02 1.42110646e+00 -3.31351310e-01 1.61440060e-01
4.38023001e-01 3.65833759e-01 -1.61460578e-01 1.31970078e-01
-4.66084450e-01 1.45654947e-01 -1.90274358e-01 5.12470067e-01
6.44797623e-01 -1.62979513e-01 2.89450660e-02 6.71619594e-01
-2.33582348e-01 3.64971966e-01 -1.24773395e+00 6.94713593e-01
-3.73885453e-01 4.53605056e-01 -5.04106522e-01 -7.34325230e-01
1.46204448e+00 3.67723763e-01 1.58688992e-01 -9.16770518e-01
2.86231160e-01 6.24103248e-01 -1.27347603e-01 2.15188071e-01
4.90065426e-01 -5.41010737e-01 -3.94666433e-01 8.06816161e-01
5.10744117e-02 -1.82124581e-02 -9.66818333e-02 -3.70874524e-01
5.30848324e-01 -2.26924807e-01 4.93132085e-01 -4.17517394e-01
1.16523586e-01 -7.59556413e-01 5.16118348e-01 1.01116109e+00
8.21969807e-02 -2.55734585e-02 8.13061535e-01 -7.70422220e-01
-1.00533366e+00 -1.46406186e+00 -2.96577334e-01 7.34444320e-01
-8.57021034e-01 2.20833644e-01 -3.71627241e-01 -9.39956233e-02
6.11448288e-01 9.95584249e-01 -9.38981831e-01 1.58281907e-01
-3.16743642e-01 -9.02514815e-01 4.20369506e-01 8.56024086e-01
5.80502450e-01 -1.51066422e+00 -6.52984798e-01 6.38573766e-01
7.20317304e-01 -6.93887055e-01 2.02277140e-03 4.06658411e-01
-1.40967751e+00 -5.79847574e-01 -8.31037760e-01 -1.22541644e-01
-5.26584387e-01 -5.22233427e-01 1.30097079e+00 -6.12124085e-01
1.70863211e-01 1.99082240e-01 3.30104768e-01 -8.75288427e-01
-3.91273737e-01 -3.14530879e-01 1.55711308e-01 -7.23348483e-02
6.15105152e-01 -6.73050821e-01 -4.08645719e-01 -3.78840923e-01
-8.03814828e-01 -7.23014355e-01 4.42484975e-01 9.14189219e-01
5.85009575e-01 -5.73084243e-02 7.97104597e-01 -6.20520473e-01
1.09313202e+00 -7.06975758e-01 -1.34195089e+00 4.90364991e-02
-1.12940824e+00 2.53581285e-01 2.55572885e-01 -2.33964935e-01
-1.31943393e+00 -8.12302530e-01 1.69518236e-02 -5.10287821e-01
1.20792836e-01 9.51453507e-01 4.52628762e-01 4.99338746e-01
-3.20243239e-01 2.45532960e-01 -1.11957803e-01 -8.66081715e-01
2.37371162e-01 1.58198178e-02 6.62606061e-01 -2.68611252e-01
5.91730058e-01 3.75526845e-01 1.73342019e-01 -6.18695557e-01
-5.31400919e-01 2.54639953e-01 -5.16182303e-01 2.05189317e-01
5.92358172e-01 -9.77666497e-01 -8.50073993e-01 1.03526306e+00
-9.21221077e-01 -3.40081185e-01 -6.07171535e-01 7.82307863e-01
-7.13348269e-01 7.90236145e-02 -1.27712858e+00 -1.48130739e+00
-6.79529309e-01 -7.88498819e-01 8.39192092e-01 1.04157522e-01
-9.60753411e-02 -1.85217500e+00 5.76144576e-01 -5.91761589e-01
8.42448354e-01 5.44207990e-01 1.19802797e+00 -1.29801238e+00
-5.15587449e-01 -3.60920012e-01 -8.67096521e-03 5.06264031e-01
-5.84465824e-02 2.50039011e-01 -9.57983315e-01 -1.02783695e-01
5.27189195e-01 2.59879399e-02 1.37021971e+00 1.04253745e+00
7.87727058e-01 -2.77964294e-01 3.80774528e-01 7.54836321e-01
1.28667712e+00 5.35600841e-01 5.34096241e-01 6.77359879e-01
5.67854010e-02 4.42817867e-01 2.18587756e-01 1.00080895e+00
2.86862552e-01 -1.00048795e-01 2.79005349e-01 4.63207096e-01
7.87612259e-01 -2.77622819e-01 4.69628066e-01 1.12082279e+00
-4.23950627e-02 -4.77168821e-02 -8.47225070e-01 5.22094667e-02
-1.66847992e+00 -1.42385757e+00 7.55403563e-02 2.22239065e+00
3.72590810e-01 6.31434619e-01 3.40941429e-01 -9.76741612e-02
4.62089479e-01 4.48608667e-01 -6.30255938e-01 -8.44048619e-01
-3.53469133e-01 4.93581980e-01 5.01381874e-01 5.21136642e-01
-1.20053875e+00 9.10671651e-01 7.72060061e+00 6.40859008e-01
-1.10326576e+00 -1.16492696e-01 1.00711036e+00 -1.13013059e-01
-6.56751513e-01 -2.91711330e-01 -9.55464065e-01 6.81024611e-01
1.54671788e+00 -6.68838203e-01 1.11163132e-01 7.71855950e-01
1.39907762e-01 1.88121080e-01 -8.10179651e-01 8.68696451e-01
-5.98478734e-01 -1.54614806e+00 3.01149487e-01 6.35707796e-01
7.97154307e-01 3.81503493e-01 9.32657003e-01 7.93773055e-01
7.83370495e-01 -1.14255011e+00 7.34650910e-01 1.29071152e+00
2.20511913e-01 -1.23564756e+00 1.02706146e+00 9.99685526e-02
-1.05269909e+00 -4.68537569e-01 -5.13639092e-01 -1.48138255e-01
3.51724982e-01 8.23372126e-01 -4.86967474e-01 1.82432294e-01
7.12511897e-01 5.15133440e-01 -2.89940715e-01 9.64703023e-01
7.24981353e-02 4.87209648e-01 -3.39275897e-01 -3.53226252e-02
7.74682164e-01 -7.20184445e-01 3.88309866e-01 1.10454202e+00
7.56702006e-01 -3.38502616e-01 -2.80887872e-01 1.15952635e+00
1.16705753e-01 -9.43854526e-02 -8.68061006e-01 -3.70202392e-01
5.04443586e-01 7.62697160e-01 -5.90256214e-01 -6.90097213e-01
-5.01425624e-01 2.62681961e-01 -7.20967129e-02 4.99339908e-01
-6.74700379e-01 -1.05355494e-01 5.37301362e-01 -3.77464771e-01
7.75467455e-01 -3.34325820e-01 -5.34634888e-01 -1.11461580e+00
-2.15225682e-01 -6.38339400e-01 6.44478738e-01 -3.01385909e-01
-1.79224396e+00 6.35642111e-01 2.16893315e-01 -9.68171000e-01
-1.24011397e+00 -8.72880936e-01 -1.18895936e+00 1.06870699e+00
-1.66684449e+00 -5.06632984e-01 4.75873947e-01 3.57833445e-01
2.54194289e-01 -1.07885325e+00 6.76559031e-01 -1.69862717e-01
-4.45286781e-01 1.92925110e-01 9.76525605e-01 7.12780328e-03
6.84102029e-02 -1.68950355e+00 1.51868236e+00 4.31079209e-01
3.81694227e-01 6.84759498e-01 4.78733182e-01 -1.02107775e+00
-8.42983246e-01 -7.99820423e-01 7.55006373e-01 -2.92768657e-01
1.34500110e+00 -1.95130669e-02 -1.05619895e+00 8.83774817e-01
3.13054383e-01 -4.72818881e-01 6.58509851e-01 -1.82781562e-01
-4.30121839e-01 -3.43347564e-02 -9.87674534e-01 3.71756047e-01
2.36446615e-02 -5.84342659e-01 -9.80471373e-01 -2.56753355e-01
5.91967642e-01 1.36180729e-01 -1.10211563e+00 4.15702552e-01
8.48254085e-01 -1.20645511e+00 7.50208437e-01 -7.81975448e-01
8.37416872e-02 6.49334729e-01 -4.60446835e-01 -1.29626644e+00
-4.18989629e-01 -9.65662062e-01 -1.02565396e+00 1.11873460e+00
3.57822120e-01 -1.17816710e+00 5.24101913e-01 7.84674704e-01
3.04190010e-01 -4.83549207e-01 -1.22418034e+00 -1.15306497e+00
7.66747952e-01 -7.09971964e-01 1.21062183e+00 6.98335826e-01
-5.41942477e-01 -2.61731446e-01 -4.42840844e-01 -2.89405435e-01
8.67249429e-01 3.42963278e-01 3.62767249e-01 -1.82207453e+00
-4.41048443e-01 -1.07013631e+00 -9.19822231e-02 -5.35493851e-01
4.25120085e-01 -5.46211541e-01 -6.91133499e-01 -9.37595367e-01
-4.97493520e-02 3.60772550e-01 -9.42168474e-01 -1.05984718e-01
4.49923903e-01 -1.44269437e-01 1.01717979e-01 9.66380686e-02
-5.44268042e-02 8.48374665e-01 6.58351660e-01 -2.44443864e-02
-5.65147288e-02 4.23532277e-01 -2.63073713e-01 9.59343851e-01
9.88694251e-01 -2.98181564e-01 -1.10222332e-01 -3.01672854e-02
2.64481187e-01 4.96116102e-01 4.46460664e-01 -5.93848228e-01
-1.10538036e-01 -9.38860700e-02 8.14006031e-01 -1.04040682e+00
4.54262286e-01 -2.60704100e-01 2.04718053e-01 6.75513625e-01
-3.10926735e-01 9.24841225e-01 3.34102780e-01 8.45966637e-01
-5.19205868e-01 -1.29205197e-01 7.38599241e-01 -4.15491998e-01
-6.45106137e-01 6.14737988e-01 -5.95924437e-01 1.65111765e-01
7.56288886e-01 5.84891327e-02 -1.48958206e-01 -7.00127304e-01
-9.59481537e-01 8.68418366e-02 -1.74170136e-02 5.03478706e-01
6.68351829e-01 -1.40745807e+00 -6.26621425e-01 4.51451749e-01
-6.39445782e-01 -7.90784001e-01 1.60832569e-01 5.98733842e-01
-3.92993808e-01 1.00693095e+00 -1.43600389e-01 -7.27216303e-02
-3.30890030e-01 4.79372114e-01 4.91443247e-01 -6.37367606e-01
-6.56904578e-01 6.18543327e-01 9.22129601e-02 -2.23579034e-01
4.76313293e-01 -6.80901766e-01 -3.37923110e-01 6.78792238e-01
7.01586485e-01 7.67668188e-01 -1.37278453e-01 -3.48882228e-01
1.55354664e-01 2.15189144e-01 -4.31600399e-02 -6.82300568e-01
1.51700318e+00 -2.40478795e-02 -8.62407014e-02 1.18802392e+00
1.01041448e+00 -5.20373881e-01 -1.42279732e+00 -3.52630883e-01
9.25001502e-01 -1.83543086e-01 8.81704018e-02 -6.31624937e-01
-1.22202909e+00 1.29364657e+00 5.85932493e-01 6.90015078e-01
3.00994366e-01 -5.57841778e-01 9.26275909e-01 7.89148033e-01
3.16175461e-01 -1.40959907e+00 2.29885969e-02 1.05543232e+00
7.12800682e-01 -1.13176417e+00 -2.69717366e-01 7.87481427e-01
-8.09169173e-01 1.29908621e+00 3.24558429e-02 -6.59576595e-01
1.49551022e+00 4.26964998e-01 2.56028414e-01 -2.91183054e-01
-1.33690655e+00 1.85349643e-01 2.77665228e-01 1.91176400e-01
3.28110158e-01 3.34611177e-01 1.13176331e-01 7.71346390e-01
-6.82410538e-01 -2.60557622e-01 4.69002128e-01 5.00251889e-01
-5.22980928e-01 -7.96501279e-01 -1.57086268e-01 7.73066342e-01
-1.01637244e+00 -3.32595319e-01 -1.55741617e-01 1.05546153e+00
-8.88467193e-01 5.12475312e-01 6.46933079e-01 -1.61306467e-02
-1.08668312e-01 4.21486408e-01 -3.33592035e-02 -2.28184983e-01
-6.60781860e-01 6.01670563e-01 -2.14253336e-01 -2.60155737e-01
1.33514076e-01 -1.00016510e+00 -7.07705855e-01 -6.11599147e-01
8.85769650e-02 5.42940199e-02 5.19027829e-01 8.41336071e-01
-7.59361719e-04 3.13294709e-01 9.22504604e-01 -8.76906872e-01
-1.59487367e+00 -1.06820476e+00 -1.44555080e+00 -1.59692451e-01
4.19582337e-01 -6.39477968e-01 -4.64814603e-01 -6.75920308e-01] | [4.636246204376221, 4.135196208953857] |
a7318c4f-6189-4e9e-8320-0b1cd9f0eada | sports-camera-calibration-via-synthetic-data | 1810.10658 | null | http://arxiv.org/abs/1810.10658v1 | http://arxiv.org/pdf/1810.10658v1.pdf | Sports Camera Calibration via Synthetic Data | Calibrating sports cameras is important for autonomous broadcasting and
sports analysis. Here we propose a highly automatic method for calibrating
sports cameras from a single image using synthetic data. First, we develop a
novel camera pose engine. The camera pose engine has only three significant
free parameters so that it can effectively generate a lot of camera poses and
corresponding edge (i.e, field marking) images. Then, we learn compact deep
features via a siamese network from paired edge image and camera pose and build
a feature-pose database. After that, we use a novel two-GAN (generative
adversarial network) model to detect field markings in real images. Finally, we
query an initial camera pose from the feature-pose database and refine camera
poses using truncated distance images. We evaluate our method on both synthetic
and real data. Our method not only demonstrates the robustness on the synthetic
data but also achieves the state-of-the-art accuracy on a standard soccer
dataset and very high performance on a volleyball dataset. | ['James J. Little', 'Jianhui Chen'] | 2018-10-25 | null | null | null | null | ['sports-analytics'] | ['computer-vision'] | [ 1.93591058e-01 -3.06981951e-01 4.65038419e-02 -3.85478348e-01
-1.41526425e+00 -9.18200910e-01 3.12596709e-01 -4.07578558e-01
-7.54174411e-01 4.43466961e-01 -1.19773418e-01 4.36000377e-01
5.25227129e-01 -8.08569729e-01 -1.44423759e+00 -6.43486738e-01
4.02018249e-01 4.65304106e-01 4.73939985e-01 -3.46279263e-01
2.08046019e-01 2.56396204e-01 -1.37912464e+00 -9.72986892e-02
5.54535151e-01 9.08566892e-01 -9.40563157e-02 8.89081419e-01
6.01687074e-01 5.23216605e-01 -7.47168243e-01 -8.61632347e-01
5.39864838e-01 -4.54870194e-01 -2.58504480e-01 4.30421084e-01
7.31503606e-01 -4.25297827e-01 -6.55520976e-01 1.26882100e+00
5.50398171e-01 2.14635488e-02 2.42186889e-01 -1.34335268e+00
-3.11547130e-01 3.53372484e-01 -7.38793075e-01 2.68325433e-02
7.04169929e-01 4.75819647e-01 5.39376676e-01 -6.69177592e-01
7.19624043e-01 7.72869051e-01 9.08837974e-01 4.08530295e-01
-1.16577840e+00 -9.25761938e-01 -2.03426734e-01 3.11010182e-02
-1.61663401e+00 -2.03236178e-01 9.45831239e-01 -2.31910273e-01
8.52556303e-02 4.59872559e-02 1.15648520e+00 1.29656923e+00
1.11988790e-01 6.71025991e-01 9.52772141e-01 -2.86299676e-01
1.49109110e-01 -3.45007300e-01 -4.10406172e-01 8.09889555e-01
3.44860554e-02 2.24083766e-01 -7.99984992e-01 3.00860792e-01
1.15235639e+00 -1.54292703e-01 -2.79391080e-01 -4.94626105e-01
-1.42977786e+00 8.60296726e-01 3.95680755e-01 -4.05877888e-01
-8.77941847e-02 5.17495990e-01 2.02261910e-01 6.30643964e-02
-9.36591029e-02 2.84588575e-01 1.72079712e-01 -3.52724642e-01
-7.74298787e-01 4.67372119e-01 5.45250058e-01 1.13894606e+00
6.76725984e-01 1.31100759e-01 1.17172457e-01 6.70116961e-01
-7.92142004e-02 7.58096933e-01 2.95403689e-01 -1.23730516e+00
6.21531367e-01 2.71030456e-01 -5.40953055e-02 -1.11500025e+00
-1.27870157e-01 -3.83029073e-01 -6.53714895e-01 8.62395465e-02
5.57556748e-01 -2.03206196e-01 -7.49654949e-01 1.64595664e+00
3.69972676e-01 5.95840454e-01 -8.93555023e-03 1.36367381e+00
6.30813837e-01 5.20995677e-01 -4.42867488e-01 2.79983014e-01
1.27823007e+00 -1.00872421e+00 -2.70788997e-01 -5.92209518e-01
1.19785510e-01 -7.51664639e-01 1.06792891e+00 6.89015210e-01
-1.14362490e+00 -5.59966385e-01 -1.36928535e+00 -2.11358577e-01
-2.72714887e-02 3.61626714e-01 4.10427570e-01 6.20674610e-01
-6.90156519e-01 1.63798988e-01 -9.07567739e-01 3.57838087e-02
-1.07758809e-02 2.98524618e-01 -6.11658752e-01 -3.11173350e-01
-1.20323491e+00 4.12574261e-01 3.35228741e-01 2.12891176e-01
-1.16546333e+00 -4.85129148e-01 -1.29521167e+00 -2.73106426e-01
5.16485095e-01 -6.32750511e-01 1.09232807e+00 -1.02224541e+00
-1.95560789e+00 1.10566092e+00 4.32779074e-01 -3.61935228e-01
8.28837633e-01 -3.67692441e-01 -4.07317311e-01 3.10766578e-01
2.24757403e-01 8.12737644e-01 8.16395283e-01 -1.22635448e+00
-7.02075541e-01 -1.81365803e-01 1.03793487e-01 3.82579744e-01
3.20968963e-02 -1.46340385e-01 -1.28354049e+00 -8.43831897e-01
2.02384502e-01 -1.45131528e+00 -1.21292964e-01 4.74699140e-02
-5.43763220e-01 5.55084944e-01 4.14412796e-01 -5.29145598e-01
7.69060612e-01 -2.10936427e+00 3.24089199e-01 2.55455047e-01
9.85947028e-02 -8.62841010e-02 -3.60489339e-02 9.26658325e-03
-1.48684354e-02 -4.27928060e-01 -2.09634036e-01 -1.29919901e-01
-1.89875916e-01 2.05290481e-01 -2.02277256e-03 7.55523384e-01
1.74995884e-02 8.84466052e-01 -8.48210216e-01 -6.12325728e-01
3.23287070e-01 4.75551546e-01 -8.11258137e-01 2.18538672e-01
1.00014620e-01 4.87671256e-01 -8.70165974e-02 6.05327368e-01
7.77091324e-01 2.69956261e-01 -3.39191221e-02 -2.76051670e-01
4.23037186e-02 -4.80736405e-01 -1.56735766e+00 2.21909070e+00
-2.01804891e-01 6.41825497e-01 -4.96296361e-02 -7.49523818e-01
9.74543095e-01 -1.68240994e-01 4.31259811e-01 -4.01556253e-01
2.87728906e-01 6.72477530e-03 -2.40320966e-01 -3.74585718e-01
6.57399893e-01 -8.76900926e-02 -8.48440528e-01 2.08348632e-01
2.51415670e-01 -6.27821684e-01 1.68031707e-01 1.88528925e-01
8.14664364e-01 3.05274099e-01 -2.08885044e-01 8.11424628e-02
4.09692585e-01 6.44435585e-02 8.86328399e-01 4.36740309e-01
5.10037020e-02 1.36125875e+00 5.60378313e-01 -5.24703264e-01
-1.34045029e+00 -1.23083079e+00 2.89425611e-01 6.38442874e-01
7.48439908e-01 -2.77671069e-01 -1.14860666e+00 -5.32486320e-01
-1.54009417e-01 3.37529898e-01 -5.75137019e-01 -5.31159699e-01
-7.42696643e-01 -4.30546910e-01 8.17159116e-01 7.86232233e-01
7.67752171e-01 -5.65546036e-01 -3.45238537e-01 5.98420389e-02
-3.89402002e-01 -1.57849205e+00 -1.17852676e+00 -1.18501373e-01
-4.21190143e-01 -1.19273520e+00 -7.64537394e-01 -9.75714207e-01
5.91392875e-01 8.13618228e-02 1.08083940e+00 -2.32052028e-01
-1.56015307e-01 1.13328129e-01 -2.80320436e-01 -1.33615941e-01
-2.95499742e-01 -4.67536505e-04 1.20119594e-01 2.72748858e-01
1.93340451e-01 -2.28351966e-01 -7.40499258e-01 7.13724375e-01
-9.88276184e-01 4.67804521e-02 4.71804470e-01 8.51650655e-01
9.89807844e-01 -1.36336833e-01 8.11341107e-02 -6.21270955e-01
3.18119764e-01 2.79897209e-02 -1.07416606e+00 1.27144188e-01
-1.04022011e-01 -7.16761127e-02 5.53183913e-01 -5.64250946e-01
-7.76269555e-01 7.91416585e-01 -2.87919790e-01 -6.45120919e-01
1.08127624e-01 1.03476681e-01 -3.86989713e-01 -3.95262033e-01
7.78381050e-01 3.64758134e-01 -1.51075786e-02 -1.04113050e-01
4.43725169e-01 4.51605588e-01 1.53829360e+00 -6.68423176e-01
1.11839700e+00 5.46498001e-01 -6.62553459e-02 -5.42807698e-01
-8.27091277e-01 -2.00228095e-01 -7.89376736e-01 -5.36846101e-01
1.07710075e+00 -1.23687994e+00 -9.10654604e-01 8.34540904e-01
-8.83632541e-01 -9.39404815e-02 -2.40010306e-01 8.12884271e-01
-8.13456118e-01 4.48324263e-01 -8.30359697e-01 -1.03728548e-01
1.25492681e-02 -1.46465147e+00 1.48558116e+00 4.44721639e-01
2.65216172e-01 -5.56118488e-01 1.62181556e-01 5.47008455e-01
-3.24596882e-01 7.88121045e-01 -7.63504058e-02 8.73961002e-02
-8.40509355e-01 -6.34007037e-01 2.24103421e-01 4.21320438e-01
-5.00368595e-01 -6.38325885e-03 -6.53736293e-01 -2.80344129e-01
-2.15315700e-01 -4.45428103e-01 6.05536342e-01 2.94540435e-01
1.21840453e+00 2.33227968e-01 7.33600631e-02 1.38724053e+00
1.39591682e+00 -1.05771609e-01 8.62961173e-01 4.97725040e-01
8.43815804e-01 1.28110051e-01 7.89230049e-01 2.97972262e-01
4.20176804e-01 9.06365693e-01 4.23622340e-01 -2.68245965e-01
-2.50001363e-02 -8.18523228e-01 5.40650249e-01 7.44085789e-01
-1.44374311e-01 -6.55934066e-02 -6.38541579e-01 2.85705149e-01
-1.57028079e+00 -7.55712032e-01 -1.04874119e-01 2.29587173e+00
7.80789435e-01 4.68844473e-01 3.80807787e-01 1.17078006e-01
1.15388620e+00 -8.80005658e-02 -6.35334373e-01 9.52735078e-03
-3.28531951e-01 1.64728567e-01 1.03963220e+00 3.42444718e-01
-1.16932976e+00 1.02354348e+00 5.90722322e+00 6.55953646e-01
-1.11334145e+00 -8.09620544e-02 5.01509309e-01 -2.49004185e-01
1.10249240e-02 -6.86434582e-02 -8.21503878e-01 6.93982184e-01
4.89102274e-01 2.28915438e-02 3.92317384e-01 9.61432576e-01
-1.97383180e-01 -1.41678572e-01 -9.35399890e-01 1.46777785e+00
5.79591036e-01 -1.47763872e+00 -1.89760432e-01 4.53975461e-02
7.70030022e-01 -2.50461996e-01 5.94381578e-02 1.90081477e-01
3.78492147e-01 -7.82340944e-01 1.20725656e+00 5.26815295e-01
1.25520992e+00 -1.13125610e+00 7.54209340e-01 2.70365536e-01
-1.08748829e+00 6.17742464e-02 -3.44411969e-01 1.45573199e-01
4.22538012e-01 7.33589604e-02 -2.39384860e-01 3.14927310e-01
6.94251597e-01 6.43970132e-01 -6.69088185e-01 1.20869493e+00
-5.63392937e-01 3.47629696e-01 -4.46782827e-01 3.21141779e-01
1.02619983e-01 -3.86602819e-01 4.08612877e-01 9.12657082e-01
3.99217248e-01 9.53145549e-02 1.75660551e-01 5.13076901e-01
-2.91046858e-01 -2.48305276e-01 -4.47695643e-01 3.96893799e-01
3.84929657e-01 1.27677405e+00 -8.21415603e-01 -1.15239106e-01
-2.58938998e-01 1.32535231e+00 3.48597132e-02 6.01505637e-02
-1.40864813e+00 -5.39623916e-01 4.31748033e-01 1.41046762e-01
2.27217808e-01 -1.87014908e-01 2.47967690e-02 -1.63848388e+00
1.27656803e-01 -1.10431111e+00 3.32589000e-01 -1.05131269e+00
-9.07995582e-01 5.51562250e-01 -3.60282287e-02 -1.63704300e+00
-3.36665362e-01 -3.34054530e-01 -5.33110380e-01 3.21673125e-01
-9.13833916e-01 -1.28843558e+00 -7.34631300e-01 7.54719317e-01
5.00523984e-01 -1.26805767e-01 3.97046655e-01 5.39467096e-01
-7.05466330e-01 9.70149159e-01 7.09724948e-02 6.69164002e-01
8.55207622e-01 -1.19870591e+00 6.29953206e-01 1.02392864e+00
3.97504687e-01 -1.67891663e-02 5.86456954e-01 -3.91390741e-01
-1.58825302e+00 -8.95605564e-01 3.32482718e-02 -6.09730184e-01
4.44822282e-01 -7.51240730e-01 -3.41981232e-01 8.50642383e-01
-6.43075444e-03 2.20117927e-01 2.66397387e-01 -5.23283362e-01
1.95372589e-02 -4.74282205e-01 -9.37282681e-01 4.91685510e-01
9.55915451e-01 -3.53090912e-01 -5.45830190e-01 2.50187129e-01
5.97935200e-01 -1.11645663e+00 -8.56966317e-01 8.23338181e-02
6.65507853e-01 -9.08358574e-01 1.07889080e+00 -1.70741707e-01
4.75564152e-01 -5.93879819e-01 5.45666702e-02 -1.31425965e+00
-2.26364862e-02 -8.76027822e-01 4.51086253e-01 1.25465763e+00
2.71295935e-01 -2.40309730e-01 1.30801129e+00 4.07303721e-01
-1.97157487e-01 -3.56136888e-01 -8.12601388e-01 -6.83995724e-01
-1.27253935e-01 -4.87517059e-01 5.52540362e-01 6.79663718e-01
-3.36186498e-01 2.24063262e-01 -6.36780202e-01 2.54988074e-01
9.90671396e-01 8.03033188e-02 1.43276072e+00 -7.45028317e-01
-3.39352280e-01 1.28610179e-01 -9.76119101e-01 -1.21243250e+00
4.82133478e-02 -4.29846078e-01 2.22204447e-01 -8.76966894e-01
1.35296762e-01 -7.51691908e-02 1.92522839e-01 -3.95347811e-02
-9.74897519e-02 9.78215218e-01 2.05866873e-01 1.82573143e-02
-6.19185269e-01 3.39880496e-01 1.25719714e+00 -1.17866658e-02
4.89298813e-02 4.27341647e-02 -4.46611524e-01 9.19438481e-01
4.42697406e-01 -6.11798048e-01 -8.65072664e-03 -5.23077488e-01
2.67232358e-01 3.93974662e-01 6.96069419e-01 -1.49283218e+00
3.97897243e-01 -4.49119061e-02 6.50284708e-01 -4.80014771e-01
6.25529170e-01 -6.02959692e-01 2.77344763e-01 4.31287438e-01
-2.00448275e-01 3.78855556e-01 -3.21337916e-02 7.16937780e-01
-4.93044585e-01 -1.19671069e-01 9.93942142e-01 -8.95173773e-02
-8.77923071e-01 2.70313442e-01 5.50180376e-02 4.60437596e-01
1.39572525e+00 -3.96570295e-01 -1.29162297e-01 -6.87577009e-01
-7.26856470e-01 3.54098856e-01 9.41454828e-01 3.95442396e-01
7.47514427e-01 -1.56013656e+00 -6.03176236e-01 5.24292529e-01
3.62674862e-01 2.57608265e-01 3.05072725e-01 5.75263798e-01
-1.39131141e+00 -1.68962836e-01 -4.38457519e-01 -9.95014966e-01
-1.10167527e+00 4.94453251e-01 4.76608425e-01 1.91247612e-01
-6.68191910e-01 8.21940005e-01 1.79462023e-02 -3.50703567e-01
1.22400485e-01 -3.05268735e-01 2.04807267e-01 -2.74251372e-01
2.97536522e-01 1.41287923e-01 -1.40603080e-01 -9.16670263e-01
-2.53935754e-01 1.08036053e+00 2.18580544e-01 -3.60512942e-01
1.27884090e+00 5.33724278e-02 5.55444598e-01 4.58989181e-02
1.27527523e+00 2.41249993e-01 -1.89434755e+00 7.00171143e-02
-5.67890406e-01 -8.20270061e-01 2.27757227e-02 -3.25719476e-01
-1.56532526e+00 7.85902202e-01 6.70616984e-01 -3.72939318e-01
1.10690176e+00 -1.65337622e-01 1.13490379e+00 1.26193210e-01
3.37545753e-01 -1.44341171e+00 2.73703665e-01 5.74879050e-02
6.07946813e-01 -1.17466712e+00 4.13856506e-02 -3.67342085e-01
-1.16238713e+00 9.99165833e-01 6.96658075e-01 -6.80319428e-01
2.16340557e-01 6.57493055e-01 3.23126256e-01 -2.50884425e-02
-5.55285141e-02 -1.79514475e-02 1.01472765e-01 6.85822725e-01
-1.79439485e-01 -1.14065655e-01 -5.63481473e-04 4.28020805e-01
-5.79209626e-01 1.40520588e-01 7.92987168e-01 6.79255068e-01
-5.69018684e-02 -1.06924784e+00 -6.17474079e-01 -1.48551241e-01
-4.28349704e-01 1.66183621e-01 -3.66340965e-01 9.88726556e-01
1.45755082e-01 5.93716741e-01 1.09066062e-01 -7.55251050e-01
6.21033967e-01 -3.81515861e-01 5.74130893e-01 -3.54320705e-01
-6.19594157e-01 2.03157634e-01 -9.72823352e-02 -6.48838878e-01
-2.08731160e-01 -6.49121463e-01 -9.20399666e-01 -3.37210506e-01
-3.76117736e-01 1.03405617e-01 7.40061045e-01 5.44021010e-01
3.56580056e-02 3.93012196e-01 7.27167428e-01 -9.96328354e-01
-3.58307987e-01 -5.52197993e-01 -7.48031378e-01 6.92370534e-01
-9.45601985e-02 -7.34565973e-01 -1.16313852e-01 4.65066493e-01] | [7.509117603302002, -1.4040178060531616] |
3065f3e9-7996-48dd-9fda-22593145b8d6 | system-identification-with-copula-entropy | 2304.12922 | null | https://arxiv.org/abs/2304.12922v1 | https://arxiv.org/pdf/2304.12922v1.pdf | System Identification with Copula Entropy | Identifying differential equation governing dynamical system is an important problem with wide applications. Copula Entropy (CE) is a mathematical concept for measuring statistical independence in information theory. In this paper we propose a method for identifying differential equation of dynamical systems with CE. The problem is considered as a variable selection problem and solved with the previously proposed CE-based method for variable selection. The proposed method composed of two components: the difference operator and the CE estimator. Since both components can be done non-parametrically, the proposed method is therefore model-free and hyperparameter-free. The simulation experiment with the 3D Lorenz system verified the effectiveness of the proposed method. | ['Jian Ma'] | 2023-04-23 | null | null | null | null | ['variable-selection'] | ['methodology'] | [-1.25446573e-01 -4.23491925e-01 2.28488401e-01 3.24576534e-02
-2.21101403e-01 -5.70578218e-01 5.26306033e-01 -8.37736428e-02
-5.57246745e-01 1.16036379e+00 -6.12287462e-01 -3.05018097e-01
-6.84400201e-01 -4.17741984e-01 5.82653992e-02 -1.07329106e+00
-2.58031338e-01 4.55760807e-01 4.94415127e-02 -1.30663425e-01
3.73920798e-01 6.23684287e-01 -1.37295198e+00 -9.07963097e-01
1.03807247e+00 9.72417712e-01 1.23851791e-01 8.86516273e-01
1.27642587e-01 3.23003411e-01 -7.11898029e-01 1.76612139e-01
4.68246579e-01 -8.46109867e-01 -2.58628875e-01 -1.46883652e-01
-6.78293526e-01 1.15705192e-01 6.23261370e-02 1.26093078e+00
5.96624970e-01 3.95411164e-01 1.33507025e+00 -1.42499208e+00
-1.92515835e-01 3.12097490e-01 -6.24132156e-01 6.62065864e-01
-1.95292346e-02 -3.18146497e-01 5.61913848e-01 -6.67450905e-01
1.86022207e-01 1.04246259e+00 4.95559573e-01 2.84144133e-01
-1.01614523e+00 -6.36968732e-01 -5.53755939e-01 1.31035194e-01
-1.78908002e+00 -1.72535852e-01 8.54448497e-01 -6.51676714e-01
5.52243233e-01 2.08896503e-01 6.33783340e-01 3.12407255e-01
9.69898880e-01 2.48960510e-01 1.41035569e+00 -4.70199645e-01
4.43063140e-01 7.12545156e-01 4.57979143e-01 5.35663128e-01
8.55073690e-01 2.02523068e-01 3.11687678e-01 -5.72634459e-01
8.58217239e-01 -2.58726537e-01 -1.42350525e-01 -4.92138386e-01
-5.93016267e-01 9.41355586e-01 -3.17478567e-01 4.84280854e-01
-3.28578740e-01 -1.36145800e-01 1.83986694e-01 5.01781940e-01
4.61411029e-01 5.22028983e-01 -3.49439472e-01 -9.36260149e-02
-6.32606804e-01 2.76136041e-01 1.18067598e+00 1.08251333e+00
3.79191309e-01 2.54829198e-01 4.45626536e-03 4.48438913e-01
4.98677760e-01 1.14729369e+00 4.51267838e-01 -4.04717416e-01
-3.73975858e-02 3.07578802e-01 2.69556463e-01 -8.36372018e-01
-5.31088591e-01 -4.43608731e-01 -8.42566013e-01 1.98362961e-01
2.47635514e-01 -7.58894682e-01 -4.93783236e-01 1.60901546e+00
3.53192598e-01 2.49234587e-01 2.05192417e-01 5.97999990e-01
4.28601086e-01 7.92627871e-01 -9.42827538e-02 -8.38387311e-01
9.95732367e-01 -7.11224228e-02 -1.16867340e+00 5.84391654e-01
1.32489324e-01 -6.61040068e-01 1.62347078e-01 3.30895454e-01
-7.59092629e-01 -1.94822043e-01 -1.22180653e+00 7.07604885e-01
-2.10870907e-01 -1.48480251e-01 3.20265740e-01 9.05784786e-01
-9.18904364e-01 2.92503357e-01 -6.74000919e-01 -2.51525760e-01
-2.07816184e-01 5.97295105e-01 -5.02418131e-02 7.07047582e-01
-1.28360677e+00 1.17178023e+00 4.87817168e-01 -2.22639330e-02
-4.17325526e-01 -2.78275788e-01 -7.05069065e-01 2.01379418e-01
9.60959122e-02 -4.92320120e-01 9.87725496e-01 -8.18613946e-01
-1.85308754e+00 3.76328379e-01 -2.62747020e-01 -1.99864864e-01
5.38640738e-01 2.34367520e-01 -4.68019426e-01 1.44657314e-01
-1.46371648e-01 -5.19209206e-01 1.02040410e+00 -1.21172428e+00
-3.27874899e-01 -2.13156968e-01 -5.15608966e-01 6.86103702e-02
2.94026077e-01 2.42554639e-02 1.49917975e-01 -3.31664473e-01
2.37077847e-01 -1.01269376e+00 -5.89012206e-02 -5.77320874e-01
-1.77016839e-01 -1.11399829e-01 7.01175034e-01 -4.83683407e-01
1.32720697e+00 -1.97827542e+00 2.10844520e-02 8.40131760e-01
1.53296560e-01 3.11686397e-01 4.70773280e-01 5.49229085e-01
-1.43763870e-01 1.55341402e-01 -5.38667142e-01 -7.73376375e-02
-6.45381510e-02 -2.43217289e-01 -3.02226376e-03 9.01183724e-01
1.89425439e-01 8.44647363e-02 -6.96429014e-01 -7.56317914e-01
5.01347125e-01 5.79945266e-01 -1.10675260e-01 2.59598464e-01
3.68440717e-01 6.00816071e-01 -8.74507606e-01 3.44412297e-01
1.07841194e+00 1.37111515e-01 -6.34993240e-02 3.07531208e-01
-3.39573026e-01 -5.67149222e-01 -1.59247351e+00 5.77695668e-01
-3.29474270e-01 7.02402413e-01 6.03571087e-02 -1.07813275e+00
1.29646444e+00 4.40164745e-01 6.32963777e-01 -1.54832050e-01
9.53042984e-01 1.63212076e-01 4.20186609e-01 -7.21690476e-01
3.67547095e-01 -3.80017549e-01 -1.26005292e-01 3.17458153e-01
1.88118577e-01 -3.14413548e-01 1.92288846e-01 -9.73261669e-02
8.37069392e-01 -1.93794280e-01 1.16103601e+00 -9.57785726e-01
1.10428250e+00 -3.24356079e-01 6.39275968e-01 7.03245580e-01
-5.52309453e-01 -3.23184021e-02 6.21757209e-01 3.26408923e-01
-1.04818881e+00 -9.66159225e-01 -8.25057626e-01 1.93256997e-02
5.07722676e-01 6.70788363e-02 -7.22835779e-01 -4.16342281e-02
2.11444497e-01 6.48109078e-01 -5.18660963e-01 -9.30431187e-02
-2.36899734e-01 -7.67216265e-01 3.67821753e-01 -1.23698615e-01
6.72809303e-01 -6.28853381e-01 -6.06141508e-01 2.50976861e-01
2.75980353e-01 -6.24915421e-01 6.25857860e-02 2.72350580e-01
-7.44919181e-01 -1.02817440e+00 -7.33417690e-01 -5.40562332e-01
6.34043455e-01 5.05533852e-02 6.78396463e-01 -9.09837112e-02
-1.84863657e-01 4.88379002e-01 -3.00695926e-01 -6.81887865e-01
-5.86120844e-01 -4.05516028e-01 3.09554100e-01 1.46406814e-01
4.53059018e-01 -3.55884224e-01 -3.27975184e-01 2.81012923e-01
-7.19993293e-01 -8.44297409e-01 5.21666765e-01 8.64326298e-01
4.25881684e-01 4.97476786e-01 7.12814093e-01 -9.02424097e-01
1.18326080e+00 -6.97651923e-01 -1.22402525e+00 2.08463475e-01
-7.23555386e-01 4.38682973e-01 7.08147287e-01 -3.18680733e-01
-1.23677027e+00 6.99141622e-02 2.66824275e-01 -2.21914902e-01
4.96255569e-02 6.79846287e-01 -2.39148885e-01 -3.77691716e-01
2.83172101e-01 3.48442346e-01 1.62854731e-01 -2.69249648e-01
-2.86750466e-01 9.08808112e-01 2.42671221e-01 -2.40155146e-01
7.54822016e-01 2.33616792e-02 8.06236625e-01 -1.22717214e+00
9.30672437e-02 -5.88292360e-01 -5.17624259e-01 -2.77433783e-01
6.68312907e-01 -4.48290914e-01 -1.07873213e+00 5.54307640e-01
-1.01619542e+00 2.49887049e-01 1.53615758e-01 8.43651474e-01
-5.14785647e-01 3.97714764e-01 -1.80802435e-01 -1.55518818e+00
-2.79114604e-01 -9.48083043e-01 2.86728024e-01 5.24672806e-01
8.56228173e-02 -1.40637553e+00 5.09658813e-01 -5.54763198e-01
5.01223087e-01 4.51895773e-01 6.21353447e-01 -9.35266197e-01
-1.17813885e-01 -5.78759789e-01 1.03738971e-01 3.61832798e-01
1.96101859e-01 3.61147076e-01 -6.90839052e-01 -3.62141430e-01
7.83740401e-01 5.78091085e-01 4.31090653e-01 7.23021984e-01
5.55238843e-01 4.31355052e-02 -3.04758400e-01 4.98320043e-01
1.73241222e+00 7.30829597e-01 2.95700461e-01 1.78338021e-01
1.95591703e-01 2.13829413e-01 6.92848861e-01 8.83716106e-01
-2.20431015e-01 3.88022184e-01 -1.17785372e-01 2.10673437e-01
6.67097270e-01 2.77192652e-01 1.77798703e-01 9.62861955e-01
2.22473107e-02 -5.81410944e-01 -9.43931639e-01 2.05863580e-01
-1.82917106e+00 -1.16857755e+00 -4.23577517e-01 2.46524906e+00
5.02168834e-01 -1.06112920e-01 -4.20556217e-02 2.69263715e-01
9.58534718e-01 -4.05185938e-01 -5.95997095e-01 -7.10906267e-01
-4.18610536e-02 2.84102470e-01 9.47993219e-01 7.93795764e-01
-1.15609419e+00 3.30739647e-01 6.82888126e+00 7.30915785e-01
-1.14018309e+00 9.69720185e-02 1.82384364e-02 5.45139790e-01
7.11031109e-02 6.95402920e-02 -8.60380352e-01 7.01209307e-01
9.36614394e-01 -8.58419001e-01 6.61102831e-02 6.76612079e-01
3.57164651e-01 -7.40747571e-01 -6.36321902e-01 1.20049751e+00
4.88227652e-03 -4.11060750e-01 -5.04519701e-01 1.08984984e-01
7.38406122e-01 -4.87654954e-01 8.86389613e-02 2.02936698e-02
-1.33671118e-02 -7.23685384e-01 1.54265702e-01 9.53227699e-01
5.76023042e-01 -1.02981973e+00 1.37631726e+00 2.92631716e-01
-1.18058991e+00 -3.86515558e-02 -4.66740042e-01 -2.54969180e-01
8.08981955e-02 6.70550108e-01 -7.10248709e-01 4.83483672e-01
1.62481606e-01 4.82940674e-01 -3.06544602e-01 1.62460029e+00
-3.60578857e-02 5.51975787e-01 -5.75010657e-01 -5.87079406e-01
-1.21200979e-01 -7.65748262e-01 1.19699323e+00 1.08947098e+00
5.52610397e-01 2.71201521e-01 -3.00831050e-01 8.61689270e-01
6.50420010e-01 2.46873274e-01 -8.52899730e-01 1.29400417e-01
9.13298965e-01 1.09358346e+00 -8.67630780e-01 -2.58032471e-01
1.60264056e-02 5.20865977e-01 -3.61514539e-01 2.85514891e-01
-8.07526171e-01 -8.19838762e-01 2.77265519e-01 -3.24325383e-01
3.04243863e-01 -3.13541621e-01 -3.51919919e-01 -1.03350091e+00
-3.17314893e-01 -3.08362484e-01 2.66216189e-01 -2.04349875e-01
-1.13363922e+00 6.44511938e-01 5.92744887e-01 -1.63483512e+00
-5.32060325e-01 -6.43476903e-01 -6.08111024e-01 1.23465383e+00
-1.03342581e+00 -2.76383728e-01 -1.22431576e-01 7.07262337e-01
-8.54149275e-03 -4.62297231e-01 7.50732660e-01 6.24109991e-02
-6.17173016e-01 5.13930857e-01 1.05078208e+00 -2.80844241e-01
5.51369369e-01 -1.41821969e+00 -4.13247645e-01 9.74890709e-01
-6.39725626e-01 7.88834095e-01 1.22133052e+00 -7.35536397e-01
-1.19998646e+00 -3.84955674e-01 7.00538814e-01 -1.96719781e-01
6.21223450e-01 -3.35536122e-01 -6.95535779e-01 2.39198461e-01
1.27707914e-01 -3.62417132e-01 6.60010219e-01 -2.11937457e-01
3.89012158e-01 -1.21203743e-01 -1.61865568e+00 1.24277614e-01
1.62450597e-01 -9.69600677e-02 -7.48455048e-01 4.61182930e-02
1.34114355e-01 -2.04045713e-01 -9.94039416e-01 1.78811461e-01
4.45820719e-01 -7.56278038e-01 4.24587727e-01 2.96496274e-03
-2.97385246e-01 -3.99496883e-01 1.28072262e-01 -1.41649246e+00
-3.02011311e-01 -9.40683305e-01 1.83365822e-01 1.26920724e+00
2.89055943e-01 -1.28954482e+00 8.73909704e-03 7.12514222e-01
3.87727827e-01 -3.44540805e-01 -1.12660468e+00 -1.18327248e+00
3.68687138e-02 1.60075441e-01 4.06095386e-01 7.83632576e-01
2.85740316e-01 3.41308355e-01 -2.98800081e-01 1.58191934e-01
8.62456143e-01 -4.43018556e-01 5.12878716e-01 -1.54868031e+00
-2.85871238e-01 -4.50377226e-01 -6.52246892e-01 -4.43767875e-01
1.63583219e-01 -4.95080829e-01 4.01411682e-01 -9.27936435e-01
2.31218129e-01 -4.76110965e-01 -4.99837995e-01 -4.31737751e-01
-2.18271300e-01 -5.72157145e-01 -2.59740967e-02 2.61413723e-01
-2.70694774e-02 6.18082285e-01 8.94209445e-01 3.53644311e-01
-4.95975345e-01 4.87978995e-01 -1.14511006e-01 4.41688180e-01
1.05502808e+00 -7.90995538e-01 -7.45612264e-01 5.06920516e-01
-1.14775792e-01 4.35323417e-01 6.22769967e-02 -1.07569814e+00
1.19384773e-01 -9.27290544e-02 9.63172019e-02 -6.51835024e-01
1.64406985e-01 -1.05763459e+00 4.93204385e-01 5.79071939e-01
-2.58596707e-02 2.46101156e-01 3.14980328e-01 5.37248731e-01
-1.95394158e-01 -7.35799968e-01 1.03552914e+00 2.59613484e-01
-4.22558308e-01 6.83524162e-02 -7.19031751e-01 -7.12737590e-02
1.27196670e+00 -2.19588637e-01 -2.51082271e-01 -5.54917097e-01
-4.25169468e-01 1.26588210e-01 1.88970491e-01 -1.92155659e-01
4.20781970e-01 -9.03751791e-01 -6.85243964e-01 3.49392176e-01
-3.04494858e-01 -5.62057018e-01 7.15069920e-02 1.03369868e+00
-6.59108222e-01 5.46712101e-01 -3.46482575e-01 -4.98420328e-01
-1.20840526e+00 5.33485889e-01 3.81479651e-01 -1.74007401e-01
-1.46109954e-01 4.71718520e-01 -1.04591414e-01 2.35264171e-02
-2.36611590e-01 1.46891894e-02 -5.74927688e-01 -5.09078726e-02
3.42769206e-01 4.91108358e-01 -3.14492166e-01 -7.66794443e-01
-5.31957626e-01 7.62250602e-01 3.03954065e-01 -5.59029281e-01
1.06355572e+00 -4.88335162e-01 -3.82684290e-01 7.95222878e-01
1.40924573e+00 9.16562974e-02 -8.75303686e-01 -4.10720035e-02
4.33324985e-02 -3.88373882e-01 2.40693063e-01 -5.71459293e-01
-6.13131464e-01 4.98596817e-01 9.25975144e-01 6.55667603e-01
1.11563826e+00 -5.24713516e-01 1.28947690e-01 4.03267771e-01
2.98302263e-01 -1.24076343e+00 -5.76994836e-01 6.44223392e-01
6.65996313e-01 -1.25463068e+00 2.14620352e-01 -3.71860862e-01
-5.40136933e-01 1.19317710e+00 3.48278999e-01 -5.17400980e-01
1.38210368e+00 4.23936486e-01 -1.47331759e-01 5.60777523e-02
-6.13688290e-01 -3.40435863e-01 3.24928224e-01 5.60325801e-01
3.42633307e-01 9.23696533e-02 -1.28673995e+00 6.74967706e-01
-8.74931142e-02 -1.15717553e-01 7.57918596e-01 9.90940630e-01
-4.85375196e-01 -8.82510662e-01 -6.25555098e-01 4.33540493e-01
-5.07574499e-01 5.81468083e-02 -2.56597787e-01 1.21064854e+00
-2.98999727e-01 1.20490193e+00 7.74788111e-02 -4.21776861e-01
4.02359776e-02 -5.05599529e-02 2.73414910e-01 1.71425045e-02
-2.82865524e-01 1.10450648e-01 -2.19964072e-01 1.30350497e-02
-3.70853275e-01 -9.82741594e-01 -1.17723072e+00 -3.34748179e-01
-8.21059465e-01 6.72804296e-01 9.08708870e-01 9.20610607e-01
1.64974168e-01 3.37950349e-01 1.18121886e+00 -4.40380007e-01
-7.70324588e-01 -1.07735384e+00 -1.24254107e+00 -2.09223740e-02
3.64924073e-01 -1.13561428e+00 -1.05457711e+00 -3.22344095e-01] | [7.113379955291748, 4.015048027038574] |
fd4321d8-f504-4c48-85ef-e8a11d216e07 | an-approach-to-intelligent-pneumonia | 2012.03487 | null | https://arxiv.org/abs/2012.03487v1 | https://arxiv.org/pdf/2012.03487v1.pdf | An Approach to Intelligent Pneumonia Detection and Integration | Each year, over 2.5 million people, most of them in developed countries, die from pneumonia [1]. Since many studies have proved pneumonia is successfully treatable when timely and correctly diagnosed, many of diagnosis aids have been developed, with AI-based methods achieving high accuracies [2]. However, currently, the usage of AI in pneumonia detection is limited, in particular, due to challenges in generalizing a locally achieved result. In this report, we propose a roadmap for creating and integrating a system that attempts to solve this challenge. We also address various technical, legal, ethical, and logistical issues, with a blueprint of possible solutions. | ['Vamsi S. Pidikiti', 'Sayali R. Rajhans', 'Alena Iureva', 'Bonaventure F. P. Dossou'] | 2020-12-07 | null | null | null | null | ['pneumonia-detection'] | ['medical'] | [ 1.72526136e-01 -1.63033605e-01 -2.87475586e-01 4.49070595e-02
-5.33724725e-01 -4.55131263e-01 2.65096933e-01 2.64186323e-01
-4.37874675e-01 1.05027342e+00 2.35446319e-01 -3.46402079e-01
-1.60636708e-01 -6.15592122e-01 -2.89187543e-02 -6.15367293e-01
1.38520911e-01 9.78078425e-01 2.16187879e-01 3.89012843e-01
2.94882089e-01 7.50161827e-01 -1.25428963e+00 1.14031292e-01
1.12113881e+00 4.57169712e-01 2.20537648e-01 7.74711311e-01
-2.88600504e-01 9.88815188e-01 -5.84701836e-01 -2.72734553e-01
3.07848509e-02 -3.47041249e-01 -7.25620091e-01 -3.37007731e-01
1.60762966e-01 -1.18415368e+00 1.04864947e-01 4.12749350e-01
4.93507892e-01 -6.26408994e-01 1.22550416e+00 -1.33480585e+00
-3.40081036e-01 -8.39291811e-02 -1.41837159e-02 2.62724012e-01
4.57586288e-01 1.98911369e-01 5.08219123e-01 -6.11875951e-01
5.19050956e-01 9.81126249e-01 7.89508462e-01 9.03424025e-01
-6.84415102e-01 -5.82001984e-01 -3.53118986e-01 3.12022865e-01
-1.31433940e+00 -1.95498466e-01 2.65266567e-01 -7.02005804e-01
1.29969084e+00 3.86804938e-01 9.40974295e-01 1.00027692e+00
2.50522107e-01 7.46671140e-01 9.22978520e-01 -2.99259305e-01
1.41058089e-02 4.68732119e-01 -7.69891366e-02 3.29666734e-01
9.30440962e-01 -4.28356767e-01 3.90102267e-02 -6.53684139e-01
8.11354458e-01 5.29017806e-01 -1.77787334e-01 -3.69193584e-01
-9.72632766e-01 7.05494821e-01 -3.98178399e-02 6.60538793e-01
-8.31847072e-01 -2.45269120e-01 1.67525366e-01 -4.09244567e-01
2.51270890e-01 2.97570795e-01 -2.76507467e-01 -2.75752217e-01
-8.50870609e-01 1.65577903e-01 7.63773739e-01 3.03982317e-01
2.27766305e-01 -1.55837819e-01 -1.18109308e-01 8.01962316e-01
3.44389617e-01 9.89932120e-01 2.76799619e-01 -9.78766084e-01
2.08649337e-01 6.13691986e-01 4.41321403e-01 -1.05879498e+00
-5.01976967e-01 1.06484778e-01 -8.31718981e-01 -1.89354762e-01
1.92848012e-01 -1.80188194e-01 -6.78411245e-01 1.23049021e+00
1.83777988e-01 -1.53792456e-01 1.24923646e-01 3.93665850e-01
2.68394411e-01 5.47348380e-01 5.05993307e-01 -5.01108587e-01
1.18026209e+00 -7.02259719e-01 -8.56775463e-01 -3.43388766e-01
5.33037663e-01 -3.97126406e-01 5.64238846e-01 5.04642725e-01
-1.01191843e+00 1.58593550e-01 -5.00183582e-01 4.35088754e-01
-3.64036858e-01 -1.67882085e-01 2.82433182e-01 9.33688104e-01
-1.14927673e+00 7.25502297e-02 -9.20541465e-01 -1.43606663e+00
4.26934063e-01 3.34616423e-01 -7.13051856e-02 -2.00751290e-01
-8.38007808e-01 1.26485729e+00 2.96294898e-01 -2.54541546e-01
-4.85088825e-01 -5.55154860e-01 -1.14653967e-01 -9.74843502e-02
7.81055763e-02 -1.15460324e+00 1.13446796e+00 -1.06235445e-01
-9.67871547e-01 8.76593113e-01 -2.50668377e-01 -3.77299547e-01
4.07614291e-01 -4.45469618e-01 -3.07790160e-01 6.11151457e-01
2.93824375e-01 5.77729881e-01 2.96318799e-01 -1.00274873e+00
-1.04451525e+00 -3.32100242e-01 -4.33354586e-01 2.44176060e-01
-7.16120541e-01 6.28146708e-01 -9.83307883e-02 -3.30537230e-01
-3.12971085e-01 -6.50341809e-01 -3.62233400e-01 2.82091826e-01
1.67636916e-01 -4.44695413e-01 7.56970286e-01 -8.57798815e-01
1.45977235e+00 -1.64804423e+00 -2.88598627e-01 -5.51657490e-02
2.83193558e-01 7.42272496e-01 3.36422235e-01 6.62194192e-01
4.42745894e-01 5.16693115e-01 -4.36353117e-01 1.91722333e-01
-3.55081767e-01 2.58977681e-01 -1.72164798e-01 1.63931727e-01
3.27340484e-01 7.81520128e-01 -9.97987330e-01 -7.74070561e-01
6.55843019e-01 8.31776381e-01 -4.89212662e-01 5.68406582e-01
8.27736631e-02 3.66312087e-01 -8.31665695e-01 7.20762491e-01
4.20061827e-01 -4.08514857e-01 3.77929091e-01 4.82136369e-01
-1.50722697e-01 2.56973088e-01 -5.73010564e-01 5.97296953e-01
5.41511830e-03 1.65804535e-01 3.07929460e-02 -7.47943819e-01
5.81514359e-01 8.38952184e-01 9.75983799e-01 -1.96586907e-01
2.72572488e-02 5.12114525e-01 -2.29313791e-01 -9.38440025e-01
-2.33263895e-02 -8.34286287e-02 5.25112212e-01 5.69549799e-01
-7.48568118e-01 -3.30801517e-01 9.65318903e-02 9.90682468e-02
1.39349210e+00 -1.49834588e-01 8.11013162e-01 -1.89430802e-03
5.44008017e-01 4.31031108e-01 4.41078067e-01 7.06191540e-01
-7.12568283e-01 6.91043615e-01 1.49457544e-01 -4.28356528e-01
-9.13347244e-01 -1.16868520e+00 -1.78787202e-01 2.89744854e-01
-4.14883226e-01 1.16992285e-02 -1.03646231e+00 -6.37974501e-01
-1.74961295e-02 8.47674072e-01 -1.86859578e-01 2.65287399e-01
-7.64334857e-01 -8.61230254e-01 6.42695308e-01 5.29459178e-01
5.32325566e-01 -1.19514072e+00 -1.19817555e+00 4.87278253e-01
-4.52448994e-01 -8.79267156e-01 2.18641862e-01 -2.91998804e-01
-1.00156772e+00 -1.23090088e+00 -1.30835664e+00 -6.74658298e-01
6.31467342e-01 5.13549566e-01 8.08459222e-01 4.19535249e-01
-5.30315816e-01 8.43518555e-01 -1.72383308e-01 -5.08540511e-01
-5.37820041e-01 -2.48398080e-01 1.80218950e-01 -7.38408864e-01
8.42587709e-01 -1.74859151e-01 -7.03061342e-01 1.10327393e-01
-8.05811465e-01 -1.91079140e-01 8.05002034e-01 1.40975416e-01
4.78965640e-01 -2.59618610e-02 8.83179903e-01 -7.42794633e-01
6.19383752e-01 -8.29529941e-01 3.03124357e-02 4.29077715e-01
-8.90340984e-01 -5.96358180e-01 2.92622954e-01 -1.23857781e-01
-8.57229590e-01 -1.13299802e-01 -7.69139230e-02 -1.97622195e-01
-5.66207767e-01 1.97587535e-01 2.76853025e-01 3.04344475e-01
5.10566711e-01 1.14820294e-01 3.16796541e-01 -4.05081749e-01
-3.48905414e-01 1.36669266e+00 1.48688897e-01 -2.20534444e-01
6.01333499e-01 4.41684365e-01 -1.64160207e-01 -9.03538883e-01
-6.11289024e-01 -7.00434923e-01 -3.32030594e-01 -5.32244325e-01
1.14087284e+00 -5.03924966e-01 -2.99369991e-01 8.14410508e-01
-9.80345309e-01 -2.40267769e-01 1.16533503e-01 6.55279338e-01
-4.25605834e-01 5.29220879e-01 -5.22897184e-01 -1.01049399e+00
-8.22988808e-01 -9.13800061e-01 5.20226717e-01 2.22962722e-01
-7.20887363e-01 -9.77551937e-01 5.72929680e-01 7.21996069e-01
9.20186818e-01 1.77396700e-01 9.52577829e-01 -6.32871628e-01
-3.81679446e-01 -1.48963481e-01 -5.11743009e-01 2.78847128e-01
7.36487150e-01 2.90544093e-01 -6.85561538e-01 -2.84308523e-01
3.21447611e-01 -1.65527880e-01 1.72202840e-01 4.91468757e-01
7.02335596e-01 -5.15213311e-01 -9.63782191e-01 -2.73484081e-01
1.21356070e+00 7.36511111e-01 5.13827741e-01 3.81723583e-01
4.07555819e-01 6.76107228e-01 5.88661909e-01 2.79908359e-01
4.91411746e-01 3.51130158e-01 1.82307139e-01 1.33319974e-01
-8.48997012e-02 7.36537650e-02 1.35620594e-01 6.76954567e-01
-2.81041116e-01 -4.38712269e-01 -1.49244356e+00 7.78742015e-01
-1.50391579e+00 -9.09648359e-01 -7.67441168e-02 2.07551455e+00
6.04623616e-01 4.83700149e-02 3.28220427e-01 3.65224220e-02
6.28079653e-01 -3.02998185e-01 -3.80299360e-01 -2.94253647e-01
7.60019943e-02 5.06299771e-02 2.23387450e-01 2.03853726e-01
-9.64955211e-01 4.66125816e-01 8.26163960e+00 1.88015163e-01
-9.90026653e-01 -9.87364724e-02 5.11896193e-01 2.54998118e-01
-2.98442185e-01 -4.74040508e-01 -5.27881444e-01 3.36478949e-01
8.52329016e-01 -2.90838361e-01 2.81416655e-01 7.99010694e-01
1.07797451e-01 -1.33602232e-01 -8.58292460e-01 4.61021602e-01
2.45491028e-01 -9.74133730e-01 1.43034279e-01 6.00967407e-02
4.25546110e-01 2.02576518e-02 -1.27983108e-01 -6.54110909e-02
-5.46370558e-02 -8.72666717e-01 1.02132864e-01 4.45948005e-01
7.16178060e-01 -3.88233393e-01 8.28958929e-01 3.13694775e-01
-9.43861663e-01 -1.25722438e-01 2.92658457e-03 -2.92051017e-01
1.34054586e-01 4.01671171e-01 -1.67732191e+00 1.49780408e-01
9.22140241e-01 3.31654757e-01 -3.37793082e-01 1.20690310e+00
-2.21169934e-01 6.78657949e-01 -4.98832315e-01 -3.74686986e-01
-1.62416007e-02 -1.44469351e-01 3.64773661e-01 1.15124655e+00
4.26041156e-01 4.42640871e-01 -1.02623835e-01 6.45675898e-01
6.20801151e-01 3.83735955e-01 -1.04716122e+00 -3.51993769e-01
6.01826787e-01 8.37244451e-01 -6.14292085e-01 -4.27812308e-01
-6.36886775e-01 7.40438163e-01 1.71869993e-01 3.59523982e-01
-5.33073008e-01 -3.64931524e-02 5.84536433e-01 1.49840832e-01
-1.27996549e-01 1.19542196e-01 -3.03522110e-01 -6.66311443e-01
-1.51159927e-01 -7.92326272e-01 5.70473313e-01 -7.52916396e-01
-9.10134494e-01 2.16691837e-01 3.18705410e-01 -9.24129784e-01
-7.43777037e-01 -5.48083007e-01 -4.99162227e-01 5.79225957e-01
-1.05296540e+00 -7.78500080e-01 -5.47473170e-02 1.53344646e-02
3.25191408e-01 -1.00287776e-02 1.14770877e+00 3.12918574e-01
-7.20580161e-01 1.47169322e-01 1.93226546e-01 -1.86898008e-01
6.40361905e-01 -7.90846586e-01 8.31671432e-03 5.51590502e-01
-7.20761180e-01 6.54774845e-01 4.65425253e-01 -1.01201379e+00
-1.10636222e+00 -8.55759144e-01 1.50219321e+00 -5.43137252e-01
4.54737097e-01 2.76698172e-01 -9.68844950e-01 6.52494848e-01
4.64310832e-02 -8.44088495e-01 6.70385540e-01 -3.67592305e-01
2.64903288e-02 1.30173847e-01 -1.78196561e+00 6.30100846e-01
6.03857934e-01 -2.28694007e-01 -9.35857236e-01 5.52401543e-01
2.22030312e-01 1.94142342e-01 -7.99036443e-01 6.34048104e-01
7.05781877e-01 -9.74099815e-01 1.08819759e+00 -2.79601991e-01
1.80101663e-01 -6.85908720e-02 1.19769767e-01 -6.29387081e-01
-4.87765223e-01 4.87746820e-02 5.07402197e-02 1.03507066e+00
7.77387097e-02 -1.01569784e+00 8.87795031e-01 9.33564901e-01
2.05799639e-01 -7.50082552e-01 -8.31484556e-01 -5.77857256e-01
9.35764611e-02 -4.17400673e-02 3.75851780e-01 9.07872558e-01
6.62868097e-02 -1.57531798e-01 -3.22568625e-01 7.83046037e-02
5.53102434e-01 -4.70333666e-01 4.31171954e-01 -1.30136561e+00
3.03695589e-01 -6.21084929e-01 -2.04001501e-01 -1.97230026e-01
-3.42281193e-01 -2.72993922e-01 -2.35747188e-01 -2.36825442e+00
8.70746613e-01 -5.51011503e-01 -3.39348197e-01 5.28153002e-01
-3.16137701e-01 1.10594660e-01 2.98367459e-02 5.78760922e-01
-1.70922264e-01 -2.33793691e-01 8.31002831e-01 -3.84238129e-03
-1.08736657e-01 -1.14736490e-01 -5.66085279e-01 8.00922394e-01
1.47411895e+00 -5.24834275e-01 -3.25258315e-01 -2.54134327e-01
-4.84951288e-02 -4.08259705e-02 7.97436014e-02 -1.17553663e+00
7.13363243e-03 -8.48167479e-01 1.74810305e-01 -1.02533424e+00
2.27886483e-01 -1.08223057e+00 6.33867860e-01 1.27208388e+00
2.78263748e-01 -2.48202965e-01 8.46258625e-02 1.30421028e-01
2.93914735e-01 -4.45872962e-01 5.50862074e-01 -5.62363453e-02
-2.73780704e-01 -3.49621437e-02 -1.03751266e+00 -6.51219711e-02
1.26986063e+00 -2.62921393e-01 -5.42978823e-01 -2.88996965e-01
2.83477493e-02 3.93078506e-01 7.46063471e-01 1.12475276e-01
5.01854599e-01 -9.05810416e-01 -6.41822696e-01 -4.73012030e-02
-6.14241995e-02 1.00111719e-02 1.99713871e-01 1.03722739e+00
-1.10741782e+00 1.07242227e+00 -4.85985190e-01 -4.91248250e-01
-1.23962784e+00 9.18799043e-01 2.20593005e-01 -2.50904322e-01
-5.85008204e-01 3.75462294e-01 2.57115781e-01 -1.10340625e-01
4.35042143e-01 5.98291978e-02 -4.34411466e-01 -4.56161331e-03
6.61659062e-01 8.35101902e-01 -3.47110003e-01 -2.92212844e-01
-8.33030224e-01 6.24541998e-01 -1.00930557e-01 7.61651546e-02
9.53861237e-01 -7.49013796e-02 -4.02706236e-01 2.95004636e-01
5.51484942e-01 -1.08451679e-01 -4.22143996e-01 2.61836588e-01
-6.76221251e-02 -2.47131124e-01 -3.57458532e-01 -8.70835960e-01
-6.70626402e-01 7.28886724e-01 6.09822690e-01 3.48055333e-01
1.16963363e+00 6.69408543e-03 8.50510657e-01 5.46487272e-01
4.08230126e-01 -8.10736537e-01 -1.84573203e-01 1.48789719e-01
6.81913018e-01 -9.50569451e-01 2.16408093e-02 -2.68689275e-01
-4.67938304e-01 8.79941523e-01 3.81314427e-01 3.20467532e-01
5.02988875e-01 1.58632606e-01 4.52051729e-01 -1.50633603e-02
-5.87523520e-01 2.80011278e-02 -2.88075656e-01 9.45104897e-01
4.62850034e-01 2.31167510e-01 -8.45198452e-01 2.22898260e-01
3.26984793e-01 5.45515120e-01 2.61030734e-01 1.19788051e+00
-9.66963232e-01 -1.17944479e+00 -7.65853405e-01 8.99926662e-01
-6.80169880e-01 1.70699611e-01 -7.50120401e-01 7.76247859e-01
-8.37057605e-02 1.09868658e+00 2.05380499e-01 -8.10630396e-02
3.85826230e-02 2.76759058e-01 1.49402201e-01 -3.92511010e-01
-3.86064500e-01 1.82395224e-02 3.14729661e-01 -2.91624337e-01
-6.92029119e-01 -9.26713467e-01 -1.49578965e+00 -2.79193491e-01
1.95116088e-01 9.65228155e-02 4.37058896e-01 7.85775363e-01
2.65707284e-01 1.55088212e-02 4.48699623e-01 -2.41335660e-01
-3.23879361e-01 -7.44412005e-01 -3.74971569e-01 -3.72598693e-02
5.97177930e-02 -4.62608933e-01 -6.03917718e-01 -2.35422507e-01] | [15.578423500061035, -1.6638038158416748] |
91f4c1b6-fba0-4534-a793-1de8cca06d6e | dichromatic-model-based-temporal-color | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Yoo_Dichromatic_Model_Based_Temporal_Color_Constancy_for_AC_Light_Sources_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Yoo_Dichromatic_Model_Based_Temporal_Color_Constancy_for_AC_Light_Sources_CVPR_2019_paper.pdf | Dichromatic Model Based Temporal Color Constancy for AC Light Sources | Existing dichromatic color constancy approach commonly requires a number of spatial pixels which have high specularity. In this paper, we propose a novel approach to estimate the illuminant chromaticity of AC light source using high-speed camera. We found that the temporal observations of an image pixel at a fixed location distribute on an identical dichromatic plane. Instead of spatial pixels with high specularity, multiple temporal samples of a pixel are exploited to determine AC pixels for dichromatic plane estimation, whose pixel intensity is sinusoidally varying well. A dichromatic plane is calculated per each AC pixel, and illuminant chromaticity is determined by the intersection of dichromatic planes. From multiple dichromatic planes, an optimal illuminant is estimated with a novel MAP framework. It is shown that the proposed method outperforms both existing dichromatic based methods and temporal color constancy methods, irrespective of the amount of specularity.
| [' Jong-Ok Kim', 'Jun-Sang Yoo'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['color-constancy'] | ['computer-vision'] | [ 4.39567149e-01 -8.69299769e-01 2.61919707e-01 -1.78164646e-01
-2.40058899e-01 -8.62705588e-01 3.58072639e-01 -7.55949438e-01
-3.34446967e-01 8.15898657e-01 -1.26841232e-01 3.30173492e-01
2.82909065e-01 -5.73939383e-01 -5.62974930e-01 -1.08118260e+00
3.67451102e-01 -6.41237423e-02 3.01610351e-01 2.34325215e-01
5.14205456e-01 3.64187628e-01 -1.50705254e+00 7.80529678e-02
8.20729256e-01 9.47828770e-01 4.14350271e-01 1.11824024e+00
1.11191362e-01 4.34410989e-01 -7.73111343e-01 -9.29358900e-02
5.25489688e-01 -5.51101923e-01 -2.31452048e-01 1.65249243e-01
6.44079387e-01 -4.81492817e-01 -7.55424947e-02 1.44735348e+00
1.65190324e-01 2.11009011e-01 6.60240114e-01 -9.51293170e-01
-4.46101815e-01 -3.20808798e-01 -1.11310923e+00 2.68379152e-02
2.89185792e-01 2.01380000e-01 7.99354315e-01 -7.46899545e-01
5.87686896e-01 8.16367865e-01 3.96201253e-01 3.72116268e-02
-1.14054549e+00 -5.83408892e-01 -1.96376294e-01 2.38724098e-01
-1.48120284e+00 -2.91347504e-01 1.25404966e+00 -7.96692744e-02
3.92305106e-01 4.26015377e-01 1.04248023e+00 3.13907325e-01
6.68942213e-01 2.77233839e-01 1.87304759e+00 -7.09232390e-01
3.48405212e-01 2.28401855e-01 -1.73847780e-01 3.71716350e-01
2.87176609e-01 3.19564432e-01 -6.08730257e-01 -5.08349277e-02
1.02077019e+00 -1.22439772e-01 -5.72296500e-01 -3.70817751e-01
-1.16218626e+00 -2.88606994e-03 3.13420802e-01 1.77168339e-01
-2.56597131e-01 2.75697261e-01 -3.78237575e-01 1.60393745e-01
4.15005267e-01 7.16516495e-01 -2.36702442e-01 -3.71242538e-02
-6.96361959e-01 -9.32418481e-02 6.14175737e-01 8.63997936e-01
1.18264127e+00 -2.06746943e-02 2.12235287e-01 7.77664602e-01
2.12845076e-02 1.00818121e+00 1.12479828e-01 -1.24842334e+00
-4.26923186e-02 2.19400495e-01 6.23115480e-01 -8.28534544e-01
-1.94199055e-01 -2.37222254e-01 -5.85312426e-01 5.89498878e-01
6.46302104e-01 -3.43329579e-01 -6.97394609e-01 1.40631175e+00
3.62001091e-01 5.67284763e-01 1.01521775e-01 1.33216417e+00
-1.57028548e-02 8.78960252e-01 -6.35418415e-01 -6.33511662e-01
1.03924406e+00 -7.77881324e-01 -7.52188742e-01 1.66570544e-01
-6.17926940e-02 -1.30465305e+00 9.57622468e-01 7.70183504e-01
-1.22948444e+00 -3.43813598e-01 -1.10819650e+00 4.41579968e-02
5.89650944e-02 2.07117632e-01 5.11506021e-01 8.49405408e-01
-1.13527751e+00 -4.15045470e-02 -3.69864523e-01 7.17603788e-02
-4.34408754e-01 -3.13495755e-01 1.91098824e-01 -2.55345196e-01
-7.30599225e-01 6.61866367e-01 3.44095565e-02 1.21444955e-01
-4.87324148e-01 -9.64211941e-01 -4.92938489e-01 -3.44079167e-01
1.11197799e-01 -3.20196986e-01 1.11610019e+00 -1.54435861e+00
-1.82409716e+00 5.83254695e-01 -7.99797952e-01 3.17459628e-02
4.03615415e-01 -1.06902337e-02 -8.65914047e-01 5.14597833e-01
-6.11464530e-02 2.92558700e-01 1.05628479e+00 -1.72496033e+00
-5.60363948e-01 -1.96357712e-01 -1.84322238e-01 6.86168194e-01
1.45385012e-01 -2.38515045e-02 -6.39029503e-01 3.71513069e-02
2.70662934e-01 -7.60213017e-01 5.92414252e-02 2.50176579e-01
-6.38300478e-01 4.38055933e-01 8.38437915e-01 -1.33119285e-01
8.72162104e-01 -2.25660992e+00 -5.60655177e-01 2.33429119e-01
1.42271206e-01 -1.49512023e-01 4.18771543e-02 1.70819774e-01
1.04210280e-01 -5.85570097e-01 -2.46060714e-01 6.14985898e-02
-4.14406449e-01 -3.26186538e-01 -1.16739273e-01 8.91020238e-01
-1.03635922e-01 3.61894071e-01 -9.26314831e-01 -2.97072738e-01
5.41285515e-01 5.87143779e-01 -4.06605363e-01 1.08439222e-01
-2.99476027e-01 3.09665382e-01 -1.39678940e-01 7.18177617e-01
1.63749731e+00 4.79378700e-02 3.02421480e-01 -4.69061017e-01
-8.81329358e-01 -9.37967896e-02 -1.43340766e+00 1.30224490e+00
-5.64813375e-01 1.02818346e+00 -2.01940741e-02 -1.18452244e-01
9.93192375e-01 6.29858300e-02 4.94027525e-01 -9.79247630e-01
-1.41036034e-01 3.75855505e-01 -2.72653967e-01 -3.17884058e-01
9.17190075e-01 -1.30978733e-01 2.83857584e-01 5.55098832e-01
-5.91012716e-01 -7.30354071e-01 -4.59698178e-02 -1.62291542e-01
5.15132785e-01 1.88277066e-01 4.33331639e-01 -4.50333387e-01
3.66646558e-01 1.07378133e-01 4.78849083e-01 6.21337950e-01
-2.36200705e-01 9.09355223e-01 2.65491456e-01 -2.91289777e-01
-1.23215270e+00 -1.36152494e+00 -5.07545352e-01 5.54303467e-01
1.09463370e+00 6.54081479e-02 -6.51787043e-01 3.87098759e-01
-2.20404491e-01 7.56531119e-01 -2.90981621e-01 3.30313683e-01
-3.64386767e-01 -7.35049248e-01 -2.02851892e-02 -1.63561836e-01
1.09084487e+00 -6.23538852e-01 -9.60324228e-01 -8.25122669e-02
-2.04675660e-01 -9.92153943e-01 -6.04764819e-01 -2.75470130e-02
-4.64452207e-01 -1.18564963e+00 -8.47436070e-01 -3.09133172e-01
6.78136826e-01 1.14131343e+00 1.14290094e+00 -4.30098981e-01
-5.82001925e-01 5.85084498e-01 -8.43720138e-02 -5.51747322e-01
3.42627093e-02 -6.13764524e-01 -2.36176461e-01 3.61081183e-01
6.21720195e-01 -2.85252035e-01 -1.14429569e+00 4.82067525e-01
-7.04155564e-01 3.45329314e-01 3.86048615e-01 3.87258470e-01
8.06808114e-01 5.05712152e-01 -1.83634788e-01 -7.01505959e-01
2.05297507e-02 -3.07455420e-01 -1.17405868e+00 4.53269080e-04
-4.20096010e-01 -2.42186993e-01 6.48642242e-01 -2.73844540e-01
-1.73618174e+00 -6.18619695e-02 8.01409066e-01 -1.54251263e-01
-1.83482185e-01 -6.91760108e-02 1.82804942e-01 -2.51708150e-01
8.23362470e-01 3.36779118e-01 -2.23947778e-01 -1.76400766e-01
1.88732907e-01 4.24559802e-01 6.40932441e-01 -3.13600659e-01
6.48530304e-01 1.19991112e+00 3.17316800e-01 -1.18242323e+00
-6.48194015e-01 -8.84857595e-01 -5.05711019e-01 -6.32115364e-01
7.63758481e-01 -1.04545653e+00 -8.68173599e-01 9.26887155e-01
-1.09454811e+00 -4.23898011e-01 -3.96070965e-02 8.49587083e-01
-4.10836577e-01 3.20932209e-01 -2.58764505e-01 -1.10205066e+00
2.83808708e-01 -7.34377682e-01 1.02824473e+00 5.56599557e-01
2.19449788e-01 -9.00483966e-01 4.30183917e-01 8.98175538e-02
2.75076479e-01 -1.17618982e-02 5.12252331e-01 1.01171803e+00
-1.22382212e+00 7.12346584e-02 -6.58004701e-01 4.95620705e-02
4.01201844e-01 3.78077418e-01 -1.30719042e+00 -4.80452217e-02
3.07766646e-01 9.23907086e-02 6.41501367e-01 9.22790289e-01
9.50657368e-01 -7.76430964e-02 -1.60132185e-01 9.37589526e-01
2.14682245e+00 4.47094977e-01 7.20494628e-01 3.79937589e-01
5.59756756e-01 4.25169557e-01 7.73764312e-01 3.86433691e-01
7.58826360e-02 6.34121358e-01 6.23345852e-01 -4.45523262e-01
-2.50565529e-01 2.70844400e-01 2.41656497e-01 2.05780894e-01
-2.44745150e-01 -5.86281657e-01 -3.72479618e-01 5.83639264e-01
-1.18553364e+00 -1.06410682e+00 -5.70466459e-01 2.42253399e+00
9.34207141e-01 -5.24086773e-01 -4.99824062e-02 -8.79592970e-02
8.66713643e-01 3.02530646e-01 -7.14555204e-01 -3.98369849e-01
-6.36733949e-01 -6.06991388e-02 7.90327787e-01 9.75951493e-01
-6.66752398e-01 7.49094367e-01 6.96054935e+00 2.49538004e-01
-1.39685357e+00 -2.71083385e-01 4.06490862e-01 -1.30150035e-01
-8.18043947e-01 -4.26795520e-02 -7.19627619e-01 6.70219839e-01
4.45444554e-01 -8.84177014e-02 6.72196269e-01 4.42432940e-01
5.40079772e-01 -1.19799805e+00 -5.51796019e-01 1.11398244e+00
2.51174748e-01 -8.76813293e-01 -2.40715340e-01 -2.10063621e-01
1.39891434e+00 -2.12101400e-01 5.45284629e-01 -9.18056369e-01
2.75416434e-01 -6.22315347e-01 4.46805090e-01 8.54094207e-01
1.03897214e+00 -5.75849831e-01 2.21582964e-01 -3.76897305e-02
-1.19769728e+00 3.41206968e-01 -7.26876676e-01 -1.50868550e-01
5.64979203e-02 9.11112666e-01 -7.29007483e-01 1.95942596e-01
7.03070819e-01 6.84934497e-01 -1.52152613e-01 1.29773307e+00
-1.01909712e-01 3.13849628e-01 -3.22553754e-01 -2.43801698e-01
1.66513488e-01 -8.72366071e-01 6.52392864e-01 9.43816900e-01
5.77271700e-01 1.80712536e-01 -4.60535794e-01 1.18433821e+00
1.97277531e-01 -1.08725414e-01 -5.11762977e-01 4.97060001e-01
4.53761011e-01 1.26201975e+00 -4.23585653e-01 -5.89842796e-02
-6.18585169e-01 1.37268734e+00 -3.81036192e-01 1.14447665e+00
-6.55362487e-01 -3.15277815e-01 7.15254664e-01 1.15725756e-01
-2.34393865e-01 -4.22107309e-01 -6.10669494e-01 -1.01537192e+00
1.44334912e-01 -2.34672099e-01 -4.62634973e-02 -1.47576630e+00
-7.59834290e-01 1.02135219e-01 -1.10026248e-01 -1.67814064e+00
2.18699470e-01 -5.89438617e-01 -7.20195293e-01 1.30433667e+00
-2.09760642e+00 -7.41195858e-01 -7.14108825e-01 9.51574206e-01
2.15652764e-01 2.31407166e-01 4.31354135e-01 -2.16861621e-01
7.04538524e-02 -1.11092785e-02 7.84135282e-01 -5.41434348e-01
1.07884455e+00 -1.45419145e+00 -1.87333688e-01 1.19906938e+00
-2.24075988e-01 4.58771139e-01 7.87508428e-01 -4.23390329e-01
-1.41101432e+00 -1.01872325e+00 5.99280953e-01 -1.49703056e-01
4.10581231e-01 -1.94754586e-01 -4.31143612e-01 3.43834519e-01
3.07206184e-01 -1.12481579e-01 4.52381819e-01 -4.00412083e-01
-2.79913545e-01 -4.77386862e-01 -1.10085130e+00 8.52751672e-01
5.72758675e-01 -6.20583355e-01 1.10099591e-01 3.34469080e-01
3.37791532e-01 -2.80218005e-01 -3.53147537e-01 2.95191985e-02
8.51981342e-01 -1.49190784e+00 8.03489923e-01 4.31566775e-01
3.06512475e-01 -7.97155082e-01 -1.35749757e-01 -1.36149776e+00
-1.97775230e-01 -7.60011971e-01 5.37916005e-01 6.73817456e-01
2.69096166e-01 -9.73637819e-01 7.14921832e-01 2.85575867e-01
-1.67014614e-01 5.43288179e-02 -6.13015592e-01 -5.73399663e-01
-4.25503999e-01 -2.60627240e-01 3.72064203e-01 8.86381447e-01
-3.33782017e-01 -1.89979181e-01 -4.15791094e-01 6.38755500e-01
1.17380536e+00 7.07616389e-01 6.73774362e-01 -1.06272137e+00
-1.58062264e-01 -7.76859224e-02 -3.19236964e-02 -1.18130648e+00
-1.93487898e-01 -1.84530839e-01 2.21742108e-01 -9.62973356e-01
3.98535222e-01 -5.30094087e-01 -2.57893980e-01 -4.73253906e-01
-8.27289000e-02 7.74663210e-01 -2.37405926e-01 3.52512985e-01
-2.97206551e-01 6.15015812e-02 1.72354257e+00 8.58693291e-03
-6.76575541e-01 -7.65770972e-02 -2.89834768e-01 7.28324115e-01
7.38566399e-01 1.54736534e-01 -6.52150154e-01 -4.83517230e-01
4.80598629e-01 7.24412603e-05 4.76823330e-01 -1.01992893e+00
2.23148301e-01 -5.90228558e-01 7.39675939e-01 -7.19443858e-01
7.39455163e-01 -7.86670625e-01 3.17666888e-01 1.05549067e-01
1.25824705e-01 -3.46163273e-01 -1.34983897e-01 5.07765353e-01
-9.98722240e-02 -1.49701327e-01 1.24288976e+00 -2.32960194e-01
-9.07248259e-01 9.90592539e-02 -5.06297290e-01 -3.11347067e-01
1.09263682e+00 -8.38573992e-01 -6.87450290e-01 -5.14016390e-01
1.51738644e-01 -4.48938489e-01 1.01184356e+00 -2.95756191e-01
4.55640316e-01 -1.11477649e+00 -2.50587881e-01 3.99374336e-01
5.46780638e-02 -3.45265329e-01 5.85561037e-01 7.47367859e-01
-1.06238699e+00 3.50875497e-01 -2.43076831e-01 -7.78056026e-01
-1.26455235e+00 1.75889775e-01 7.20697999e-01 5.14653325e-01
-4.43004221e-01 7.96035767e-01 6.93836570e-01 2.76983440e-01
-3.22395146e-01 -2.82568306e-01 6.02068678e-02 -3.30980986e-01
4.90843683e-01 6.72376752e-01 -2.60062784e-01 -5.45343161e-01
-3.44981663e-02 1.29043376e+00 3.04145634e-01 -4.73542511e-01
8.08237076e-01 -6.52268171e-01 -2.96575308e-01 8.28038216e-01
1.13810122e+00 6.01795316e-01 -1.68086493e+00 -8.84302929e-02
-9.03095126e-01 -1.05058861e+00 1.14387803e-01 -6.62047923e-01
-7.41810381e-01 6.26100123e-01 5.55498958e-01 2.22309887e-01
1.57477546e+00 -3.47730875e-01 3.89638960e-01 1.00127295e-01
3.80053252e-01 -1.37323213e+00 2.22768068e-01 2.16455609e-01
4.01334226e-01 -9.98777092e-01 -4.70724516e-02 -7.34758675e-01
-5.84050238e-01 1.29161477e+00 5.58222532e-01 -1.52782381e-01
4.95035738e-01 3.01964909e-01 3.70144129e-01 -1.49991974e-01
-3.72130334e-01 -1.65004302e-02 2.32586473e-01 7.07707524e-01
3.63930583e-01 1.40475407e-01 -1.27578750e-01 -5.32108605e-01
-1.82500824e-01 -3.45556647e-01 1.08800125e+00 3.57952505e-01
-5.85869431e-01 -4.29137439e-01 -6.59233630e-01 1.97939165e-02
-2.05070391e-01 -2.28724360e-01 -1.67076647e-01 6.41506732e-01
4.54588793e-03 9.23789144e-01 7.32440829e-01 1.60601035e-01
-2.27909237e-01 -3.88345569e-01 6.05604053e-01 -3.16611081e-01
1.29150301e-01 4.10696059e-01 -1.53274640e-01 -6.70637608e-01
-7.44697988e-01 -6.28025889e-01 -8.70270967e-01 -3.44982773e-01
-2.22561046e-01 1.13369671e-04 8.19126785e-01 2.97986805e-01
-1.79855004e-01 -9.42105725e-02 1.20348001e+00 -5.52924037e-01
4.64100182e-01 -5.26273906e-01 -1.36377811e+00 1.52059898e-01
6.65535152e-01 -2.93630838e-01 -9.53581393e-01 2.37240136e-01] | [10.367412567138672, -2.667145252227783] |
03550177-c6d9-4edc-9f59-8ab0f2d3b31e | use-of-extended-kalman-filtering-in-detecting | null | null | https://www.sciencedirect.com/science/article/abs/pii/S0017931006006533#preview-section-abstract | https://www.sciencedirect.com/science/article/abs/pii/S0017931006006533 | Use of extended Kalman filtering in detecting fouling in heat exchangers | This paper is concerned with how non-linear physical state space models can be applied to on-line detection of fouling in heat exchangers. The model parameters are estimated by using an extended Kalman filter and measurements of inlet and outlet temperatures and mass flow rates. In contrast to most conventional methods, fouling can be detected when the heat exchanger operates in transient states. Measurements from a clean counterflow heat exchanger are first used to optimize the Kalman filter. Then fouling is considered.The results show that the proposed method is very sensitive, hence well suited for fouling detection. | ['Bernard Desmet', 'Olafur P. Palsson', 'Sylvain Lalot', 'Gudmundur R. Jonsson'] | 2007-01-17 | null | null | null | journal-2007-1 | ['line-detection'] | ['computer-vision'] | [-1.73271447e-02 -4.89178121e-01 9.95018110e-02 1.98558077e-01
2.66498297e-01 -6.20017171e-01 2.67902404e-01 4.05230910e-01
-4.67415825e-02 1.08525884e+00 -4.82325673e-01 -5.34301937e-01
-2.28237018e-01 -6.36011899e-01 -3.24258417e-01 -7.07037747e-01
-5.39729238e-01 1.65774018e-01 -8.82436931e-02 9.66762453e-02
2.35987306e-01 8.62178981e-01 -1.28108776e+00 4.57952060e-02
7.87987411e-01 8.63113940e-01 5.41015744e-01 1.20423520e+00
1.86814561e-01 3.31976175e-01 -6.45868480e-01 7.76195705e-01
2.34455302e-01 -5.88801384e-01 -8.61937702e-01 2.70671755e-01
-2.98927754e-01 -3.58497679e-01 -1.17946953e-01 1.16879559e+00
3.21556866e-01 3.68078500e-01 4.87396628e-01 -4.47968572e-01
3.81804377e-01 6.00081719e-02 3.37341040e-01 1.61325693e-01
2.60684043e-01 2.37485319e-01 3.11779827e-01 -6.17084086e-01
1.63343549e-02 1.20706356e+00 5.59976697e-01 2.04668701e-01
-1.52525055e+00 -2.97019422e-01 -1.64977796e-02 -4.30254430e-01
-1.05228817e+00 -4.28468436e-01 2.13046297e-01 -6.01751208e-01
1.24649477e+00 6.49551928e-01 7.79506743e-01 4.69978973e-02
7.61799455e-01 6.46576822e-01 1.50462067e+00 -3.76137763e-01
4.42732036e-01 3.46089602e-01 4.98899408e-02 2.45110393e-01
6.30230606e-01 6.97496176e-01 1.89192936e-01 -1.46071747e-01
8.35016310e-01 2.83075310e-02 -4.37074929e-01 -3.14539671e-01
-7.49534011e-01 4.70944524e-01 -1.01260029e-01 3.87396783e-01
-6.65265620e-01 -2.75474995e-01 6.69170618e-01 7.41178334e-01
5.45723855e-01 1.02661812e+00 -8.18932712e-01 -3.54624093e-02
-8.00445676e-01 2.15559378e-01 1.46553481e+00 8.25161159e-01
4.07014102e-01 2.44047239e-01 1.32798761e-01 4.40898329e-01
6.20601296e-01 1.10049331e+00 1.27735779e-01 -7.90610790e-01
-1.24447718e-01 3.18476319e-01 1.09722173e+00 5.25041223e-02
-3.95469218e-01 3.70746804e-03 -3.50653112e-01 5.41848600e-01
1.61205947e-01 -9.12390292e-01 -8.33064556e-01 5.95369637e-01
1.93038344e-01 -1.30016208e-01 3.80487949e-01 7.90190995e-01
9.02765542e-02 8.36888492e-01 8.57578963e-02 -9.66691554e-01
1.03350043e+00 -4.81671274e-01 -1.61553824e+00 1.80628896e-02
4.48878288e-01 -9.54063475e-01 -4.98459153e-02 4.93249744e-01
-1.01610172e+00 -3.73213321e-01 -1.00303626e+00 8.27378213e-01
-4.58691925e-01 1.37028649e-01 2.01549113e-01 9.56166387e-01
-4.83477592e-01 1.47673261e+00 -1.36044753e+00 -4.46236312e-01
-3.60585779e-01 5.57974935e-01 5.28838225e-02 5.33833444e-01
-1.24638331e+00 1.44447303e+00 7.11833894e-01 8.26625049e-01
-8.23562205e-01 -3.97458732e-01 -8.11661005e-01 -8.28037634e-02
-4.17693220e-02 -1.66827157e-01 1.77144992e+00 -2.36435950e-01
-2.16741037e+00 2.20839247e-01 -2.53873408e-01 -3.36229891e-01
6.55609548e-01 -8.21154594e-01 -8.03844571e-01 1.34228975e-01
-2.79998034e-01 -9.80195224e-01 6.28251672e-01 -1.04012656e+00
-9.44929600e-01 -6.98294071e-03 -3.75745207e-01 1.02432154e-01
2.55792260e-01 -1.83609426e-02 5.97231984e-01 2.26113215e-01
3.31730098e-01 -8.33955407e-01 -3.62095445e-01 -5.76169372e-01
-2.30510235e-01 2.37286463e-01 1.13307023e+00 -6.57522619e-01
1.37298501e+00 -1.61944497e+00 -3.38837057e-01 4.81597096e-01
-8.43424737e-01 7.52128065e-01 6.65577948e-01 1.01530957e+00
-1.69585839e-01 -1.45139053e-01 1.10664338e-01 3.90410274e-01
9.17115360e-02 3.34768414e-01 -3.72353494e-02 1.15584183e+00
3.39145571e-01 3.30271572e-01 -1.21746826e+00 2.31108353e-01
1.29398108e+00 1.20441027e-01 1.89283878e-01 4.80661899e-01
3.14239115e-01 2.45670542e-01 -5.32893300e-01 5.11943579e-01
8.63574564e-01 3.17970753e-01 4.50309873e-01 -1.91928521e-01
-9.25114751e-01 3.71847421e-01 -1.73669732e+00 9.68557477e-01
-5.23463845e-01 3.46698225e-01 8.41045022e-01 -9.34759676e-01
1.12144911e+00 9.58159804e-01 6.88528597e-01 -4.85008627e-01
2.63216555e-01 3.76154363e-01 -4.45440769e-01 -9.50156033e-01
5.11412263e-01 -7.98555315e-01 4.58923548e-01 -1.53872818e-01
-5.74893467e-02 -3.93259227e-01 3.58997494e-01 -5.51944375e-01
8.15348029e-01 2.11572051e-01 4.21741277e-01 -1.27367985e+00
1.07413375e+00 -8.08706209e-02 5.68960428e-01 4.45965558e-01
7.65912980e-02 -2.42677763e-01 8.93873051e-02 -3.53891760e-01
-9.56698418e-01 -6.38305962e-01 -8.36340487e-01 3.72625500e-01
7.37978399e-01 1.95822101e-02 -3.50881279e-01 1.28335565e-01
4.76817697e-01 6.41251802e-01 -2.14031413e-01 9.86012351e-03
-4.72375989e-01 -9.49246109e-01 3.85230556e-02 5.06853878e-01
2.30579048e-01 -6.63019776e-01 -7.42075264e-01 1.04173934e+00
5.17239034e-01 -5.14868081e-01 3.43700856e-01 6.65707707e-01
-1.50828707e+00 -1.41268289e+00 -5.46296060e-01 -3.82255644e-01
7.42947876e-01 3.03793214e-02 8.43504608e-01 1.41779333e-01
-2.94570595e-01 1.97731137e-01 -2.39963010e-01 -7.43103027e-01
-9.45339799e-01 -3.77910584e-01 1.05319701e-01 -2.48465985e-01
3.23275059e-01 3.26918930e-01 -6.55246496e-01 8.07907164e-01
-6.39247537e-01 -4.34610486e-01 2.30832517e-01 1.05126035e+00
1.91251382e-01 3.94122839e-01 5.63169956e-01 -1.25923431e+00
6.59071445e-01 -2.08082706e-01 -1.10816443e+00 2.76170284e-01
-9.12832558e-01 5.18607423e-02 7.28776455e-01 -2.03936860e-01
-1.51013887e+00 2.68666208e-01 1.91563278e-01 -6.35572374e-02
-5.93236446e-01 5.24833858e-01 -3.57509255e-02 -1.00646056e-01
5.37281334e-01 -3.99641454e-01 4.13049668e-01 -8.32135201e-01
-4.24224287e-01 5.66251278e-01 3.71636868e-01 -3.03169668e-01
6.51115417e-01 1.81120783e-01 3.53825629e-01 -1.23072970e+00
6.10934235e-02 -1.01602912e+00 -5.44269502e-01 -3.53395194e-01
3.67789656e-01 -9.49760735e-01 -1.01317561e+00 8.00463796e-01
-8.01396310e-01 -5.35152316e-01 -2.92443961e-01 1.16201615e+00
-4.19665992e-01 2.90830135e-01 -1.08046186e+00 -1.74897087e+00
-3.47694695e-01 -7.32573152e-01 6.42176986e-01 3.78028750e-01
-1.32133469e-01 -1.62995446e+00 1.38127446e-01 -5.10325611e-01
5.26623011e-01 2.61700362e-01 1.58209741e-01 -3.03918898e-01
-9.23783407e-02 -6.52805448e-01 3.92099649e-01 6.93177700e-01
7.30042160e-01 1.99311748e-01 -1.17038465e+00 -1.02108359e+00
3.71581197e-01 3.65750641e-01 5.76745450e-01 5.72521031e-01
1.88406467e-01 -1.19466007e-01 -6.60236955e-01 3.41019452e-01
1.66950095e+00 6.82959139e-01 3.77040446e-01 5.34696341e-01
7.04558939e-03 2.35076442e-01 1.17054284e+00 6.46336257e-01
-7.55699873e-01 -6.52718619e-02 2.65577048e-01 -4.12757277e-01
7.02779293e-01 2.15554550e-01 4.07429338e-01 3.74782473e-01
-1.39865130e-01 -3.36866975e-01 -5.99048316e-01 5.35623789e-01
-1.65559077e+00 -9.46274340e-01 -7.06880748e-01 2.40956259e+00
4.60654557e-01 -6.35260046e-02 -2.52633244e-01 3.37411463e-01
1.15690494e+00 -3.64433438e-01 -3.47582430e-01 -7.23925471e-01
2.31802970e-01 2.98722893e-01 1.33360851e+00 8.00400853e-01
-1.45552206e+00 1.35924697e-01 7.96251917e+00 -1.66407675e-01
-1.03614771e+00 -5.85814834e-01 -2.89110273e-01 2.26024762e-01
4.85001922e-01 4.31408703e-01 -7.93088257e-01 3.03299278e-01
1.21896398e+00 -3.65730554e-01 3.05765390e-01 5.04866242e-01
6.68922722e-01 -6.51723444e-01 -1.06888998e+00 3.54154021e-01
-6.30835116e-01 -6.44601405e-01 -8.28421772e-01 -8.21607038e-02
8.26130867e-01 -2.85613090e-01 -6.10710084e-01 1.63841352e-01
3.33033665e-03 -3.82282168e-01 3.08170885e-01 7.30201423e-01
1.88269779e-01 -6.21570826e-01 1.18739545e+00 3.18305373e-01
-1.48588061e+00 -3.66284043e-01 -3.75293255e-01 -6.08263731e-01
7.78961003e-01 6.29197478e-01 -1.07607424e+00 9.07728732e-01
3.82781178e-01 6.39837027e-01 9.65458229e-02 1.57994902e+00
-1.81156443e-03 3.61563623e-01 -6.19785130e-01 -7.98116550e-02
-2.22313441e-02 -3.12754959e-01 3.98771137e-01 1.40618467e+00
4.04456943e-01 1.52645512e-02 9.14885625e-02 7.47622669e-01
7.81793058e-01 -4.70248312e-02 -5.10352194e-01 -1.28730282e-01
3.85466278e-01 7.31859505e-01 -6.46816432e-01 -7.39350080e-01
-3.09075832e-01 6.75616562e-01 -9.09162641e-01 4.69787985e-01
-2.69496709e-01 -8.58772814e-01 7.91969001e-01 3.43285315e-02
2.80976325e-01 -1.55254826e-01 1.86887175e-01 -8.57177973e-01
-5.07202268e-01 -7.95396939e-02 3.56693566e-01 -2.51338065e-01
-1.12846231e+00 -1.76872283e-01 3.19949627e-01 -1.28813219e+00
-3.53243828e-01 -1.28058934e+00 -7.62188137e-01 1.53966033e+00
-1.57195580e+00 -1.54391855e-01 3.95431258e-02 -2.68123602e-03
3.95322651e-01 2.12812662e-01 9.35803533e-01 1.69723570e-01
-7.47741699e-01 -3.64476919e-01 1.15689838e+00 -2.39477530e-01
5.33718884e-01 -1.63093555e+00 4.38729107e-01 9.74630356e-01
-1.14323390e+00 7.20674157e-01 1.42520154e+00 -1.09469569e+00
-1.75303555e+00 -8.81798506e-01 6.83962703e-01 1.34493813e-01
5.80909014e-01 1.00954168e-01 -1.25875151e+00 2.84212142e-01
2.70023465e-01 -3.33319157e-02 1.77487969e-01 -9.43395644e-02
1.08253467e+00 -7.05819065e-03 -1.05069613e+00 -3.80057603e-01
-3.43933254e-02 -6.31807387e-01 -7.00628698e-01 4.69197333e-01
-2.92367071e-01 -9.26080585e-01 -1.45425701e+00 6.46578670e-01
5.95750749e-01 -4.55505311e-01 4.54159468e-01 -6.51120186e-01
-7.00395763e-01 -4.29553092e-01 7.09410548e-01 -1.41184127e+00
-3.04560453e-01 -1.10996139e+00 -1.26954857e-02 8.14356565e-01
1.48317873e-01 -9.14003968e-01 4.10355210e-01 8.52607846e-01
-2.84396708e-01 -7.87524059e-02 -5.64970136e-01 -1.17446709e+00
-2.98443437e-01 2.88908333e-01 3.54723752e-01 7.85003364e-01
5.59378028e-01 -5.29612042e-02 -4.41002131e-01 8.40608656e-01
3.57089579e-01 2.16863617e-01 4.38399822e-01 -1.48332751e+00
1.21518925e-01 1.35245219e-01 -9.81042236e-02 -9.51783657e-01
-1.96056768e-01 -1.46911964e-01 7.13373303e-01 -1.60210323e+00
-7.61035800e-01 4.35339026e-02 -6.20299220e-01 -1.27449512e-01
-1.85075030e-01 -5.86064041e-01 -8.20516199e-02 -1.91384017e-01
-2.78151091e-02 2.80655950e-01 1.07323360e+00 1.25710845e-01
-6.36811972e-01 4.90243971e-01 4.70530629e-01 3.43619883e-01
8.21991444e-01 -3.18391532e-01 3.11143901e-02 2.28259400e-01
-5.85912243e-02 5.03669679e-01 1.65111601e-01 -1.11596143e+00
8.65854025e-02 -2.39451677e-01 4.98300701e-01 -6.86603248e-01
-9.66986790e-02 -1.34115100e+00 8.23007464e-01 1.16526020e+00
6.72980398e-02 -4.63630743e-02 5.77541173e-01 6.30492389e-01
-3.15462291e-01 -7.43149400e-01 7.55998552e-01 -3.41355354e-01
-7.16008902e-01 -2.12861180e-01 -1.00975442e+00 -7.00939059e-01
9.31121647e-01 -2.96762854e-01 -3.14182907e-01 3.27854007e-01
-1.15922260e+00 6.50571883e-01 4.36815649e-01 -4.02042195e-02
9.44659412e-02 -9.61658895e-01 -3.69518995e-01 7.44475543e-01
-4.51558411e-01 -1.43143937e-01 3.80757004e-01 9.29831862e-01
-9.75039244e-01 6.27631009e-01 -2.51113564e-01 -6.23024642e-01
-1.16328299e+00 6.92580402e-01 8.35645974e-01 2.69473158e-02
-5.00081599e-01 2.14239419e-01 -2.33816952e-01 1.15525469e-01
-3.02550524e-01 -6.64016247e-01 -3.03429533e-02 4.61043045e-02
5.12327611e-01 8.60060096e-01 4.54868585e-01 -2.67732948e-01
-2.19520539e-01 1.67598918e-01 3.88542712e-01 2.75935113e-01
1.06634879e+00 -3.96558344e-01 -8.39803666e-02 9.10273492e-01
1.01290047e+00 -5.71622908e-01 -1.53336179e+00 4.72737588e-02
7.55392686e-02 -5.83391249e-01 3.00166517e-01 -5.17692327e-01
-4.83551949e-01 5.47389805e-01 1.03478742e+00 8.65591586e-01
9.55844760e-01 -8.12161624e-01 1.93823829e-01 6.21425152e-01
1.04654655e-01 -1.58603704e+00 -8.57869446e-01 5.32957554e-01
4.88224000e-01 -7.57114708e-01 5.32500565e-01 -3.11444402e-01
-2.38677591e-01 1.60124636e+00 3.72095734e-01 -3.66595536e-01
6.85611546e-01 5.24353325e-01 2.56832361e-01 -1.40562445e-01
-7.51914799e-01 -5.73676974e-02 -2.60799587e-01 1.30333334e-01
7.10658014e-01 1.83593288e-01 -4.96847481e-01 -5.89144170e-01
6.00534141e-01 1.11620352e-01 3.75547171e-01 1.54802394e+00
-8.38731468e-01 -9.90073681e-01 -7.89880991e-01 7.36928701e-01
-6.87395096e-01 6.93078458e-01 1.53557524e-01 8.63793671e-01
-3.45070094e-01 1.12842309e+00 -1.17881402e-01 8.14807937e-02
1.17568755e+00 3.98976505e-01 5.30369245e-02 -5.83797216e-01
-7.92471707e-01 5.20776629e-01 3.67806792e-01 -5.46709538e-01
-6.19997084e-01 -8.21307480e-01 -1.14493465e+00 7.20230788e-02
-1.37348187e+00 8.69712889e-01 7.32895017e-01 7.82089055e-01
-3.26027542e-01 6.50458455e-01 9.93483961e-01 -9.11630630e-01
-7.81837165e-01 -1.13927209e+00 -1.59693313e+00 2.19838038e-01
8.89316082e-01 -9.15131807e-01 -8.66375506e-01 -4.00666296e-02] | [6.041321754455566, 2.6544368267059326] |
2b45a392-9fbb-4876-814b-b9c771d158e3 | simple-and-effective-unsupervised-speech-1 | 2210.10191 | null | https://arxiv.org/abs/2210.10191v1 | https://arxiv.org/pdf/2210.10191v1.pdf | Simple and Effective Unsupervised Speech Translation | The amount of labeled data to train models for speech tasks is limited for most languages, however, the data scarcity is exacerbated for speech translation which requires labeled data covering two different languages. To address this issue, we study a simple and effective approach to build speech translation systems without labeled data by leveraging recent advances in unsupervised speech recognition, machine translation and speech synthesis, either in a pipeline approach, or to generate pseudo-labels for training end-to-end speech translation models. Furthermore, we present an unsupervised domain adaptation technique for pre-trained speech models which improves the performance of downstream unsupervised speech recognition, especially for low-resource settings. Experiments show that unsupervised speech-to-text translation outperforms the previous unsupervised state of the art by 3.2 BLEU on the Libri-Trans benchmark, on CoVoST 2, our best systems outperform the best supervised end-to-end models (without pre-training) from only two years ago by an average of 5.0 BLEU over five X-En directions. We also report competitive results on MuST-C and CVSS benchmarks. | ['Juan Pino', 'Michael Auli', 'Wei-Ning Hsu', 'Yun Tang', 'Ilia Kulikov', 'Peng-Jen Chen', 'Hirofumi Inaguma', 'Changhan Wang'] | 2022-10-18 | null | null | null | null | ['speech-to-text-translation', 'unsupervised-speech-recognition'] | ['natural-language-processing', 'speech'] | [ 4.72138166e-01 3.41094077e-01 -4.86476481e-01 -6.66075826e-01
-1.68686581e+00 -7.27704048e-01 7.73109496e-01 -4.55884159e-01
-3.47685069e-01 9.11347151e-01 5.22469103e-01 -9.24392581e-01
6.12387836e-01 -1.42633349e-01 -7.63915479e-01 -4.14497793e-01
6.68789864e-01 1.09006143e+00 -1.08721264e-01 -3.38310093e-01
-2.86003172e-01 -1.10585995e-01 -6.91836417e-01 5.73546052e-01
1.05796337e+00 4.25388694e-01 3.12705845e-01 6.38293207e-01
-2.66374946e-01 4.95708883e-01 -6.67884827e-01 -5.88272810e-01
1.48853153e-01 -8.97944272e-01 -1.03865290e+00 3.20437193e-01
2.18270004e-01 -1.50286794e-01 -3.00475150e-01 8.77833903e-01
8.12412083e-01 -1.86015889e-02 6.21647179e-01 -9.03205752e-01
-8.84613037e-01 1.20446622e+00 6.98187277e-02 1.43996654e-02
1.19054049e-01 2.36382380e-01 8.49770010e-01 -1.36756563e+00
7.69011915e-01 1.19437575e+00 2.11113319e-01 9.98499274e-01
-1.20117795e+00 -4.36036229e-01 -2.19470382e-01 -1.20124489e-01
-1.19396412e+00 -1.35535800e+00 3.23668838e-01 -2.20897913e-01
1.43824077e+00 2.27963641e-01 -4.71902974e-02 1.67039502e+00
-4.99893844e-01 9.25382137e-01 1.22270548e+00 -8.79749119e-01
2.31463686e-01 2.18178019e-01 -3.74033660e-01 3.86881948e-01
-3.65032405e-01 2.65802056e-01 -8.62092257e-01 1.51909009e-01
1.74026534e-01 -6.54635489e-01 1.65429264e-02 3.37297469e-01
-1.79942513e+00 6.69794679e-01 -1.60160497e-01 2.96231836e-01
-2.05861926e-01 -2.37232327e-01 5.98902285e-01 7.11562574e-01
8.98121655e-01 4.66318101e-01 -9.28394794e-01 -4.71154302e-01
-1.12196994e+00 -3.65079880e-01 9.94422913e-01 1.50946105e+00
4.43078369e-01 5.93394339e-01 -4.57757711e-01 1.27539968e+00
2.27952942e-01 1.13934028e+00 6.50702178e-01 -7.36871779e-01
1.13244903e+00 2.92737503e-02 -1.57613754e-01 1.36564910e-01
1.26158521e-01 -4.71166819e-01 -6.06730402e-01 -6.06294692e-01
3.23838681e-01 -5.36560833e-01 -1.41323161e+00 1.51173234e+00
6.74998984e-02 -8.77563879e-02 6.75703585e-01 7.96461344e-01
8.87760043e-01 1.01535201e+00 5.29161282e-02 -3.96941900e-01
1.08250535e+00 -1.71801937e+00 -8.52442086e-01 -6.88992441e-01
1.00504887e+00 -1.14470661e+00 1.31882238e+00 1.23801693e-01
-1.15096176e+00 -4.81681526e-01 -7.03309059e-01 -1.88489519e-02
-2.53404826e-01 6.53588474e-01 1.77887604e-01 8.31559479e-01
-1.34025633e+00 1.16014466e-01 -8.97076249e-01 -7.52556980e-01
5.77761792e-02 1.56483501e-01 -2.44041160e-01 6.56586960e-02
-1.23393190e+00 1.05030680e+00 2.37345502e-01 -1.67605102e-01
-1.16757810e+00 -3.91054183e-01 -6.74735904e-01 -5.43010309e-02
2.59404987e-01 -4.82382268e-01 1.86320555e+00 -1.09918189e+00
-2.20451283e+00 8.79326284e-01 -5.10147035e-01 -5.92027545e-01
4.20701116e-01 8.83380771e-02 -7.58629262e-01 -1.11354403e-01
1.98749885e-01 7.73951709e-01 5.91114342e-01 -8.80192220e-01
-4.34596390e-01 -9.56966653e-02 -5.95417500e-01 3.98138642e-01
-3.17856431e-01 6.81205750e-01 -4.33553100e-01 -6.61513031e-01
-3.91894616e-02 -1.17045236e+00 -2.33168796e-01 -6.43993855e-01
-5.54107368e-01 -2.27073848e-01 6.51646554e-01 -9.58298922e-01
1.00602543e+00 -1.94162691e+00 1.97432518e-01 -2.32956842e-01
-4.96023893e-01 5.78383803e-01 -3.78259629e-01 6.67509854e-01
2.07826663e-02 2.47045428e-01 -4.90243107e-01 -8.55480134e-01
2.46236641e-02 2.95174897e-01 -5.69856226e-01 2.66166553e-02
3.78987700e-01 1.07840502e+00 -9.53568816e-01 -2.77146399e-01
2.85226330e-02 1.97605580e-01 -1.75786301e-01 4.85569447e-01
-3.89000624e-01 8.33909333e-01 -8.78355429e-02 7.24012196e-01
1.71323344e-01 -7.49384165e-02 2.88106948e-01 5.95205247e-01
-7.23427981e-02 1.36420596e+00 -3.97510916e-01 2.10044622e+00
-7.51119435e-01 6.41386807e-01 -4.78620604e-02 -9.60857332e-01
1.08314574e+00 1.03521109e+00 9.44751576e-02 -6.87269807e-01
-6.34416938e-02 8.06100905e-01 -2.78200693e-02 -2.71320820e-01
2.50044882e-01 -1.97311521e-01 -1.94998711e-01 5.08979917e-01
4.78129089e-01 -4.36948866e-01 -2.11507529e-02 4.08661067e-02
9.34259474e-01 1.25876203e-01 -4.65463996e-02 -1.16699338e-01
3.05521876e-01 3.66229385e-01 3.90923589e-01 4.63312656e-01
-1.47183105e-01 7.40827501e-01 5.79939596e-02 3.71998474e-02
-1.25468135e+00 -8.29031229e-01 2.24434540e-01 1.43530476e+00
-5.74019670e-01 -3.36512119e-01 -1.07697177e+00 -9.90492523e-01
-4.96098608e-01 1.07884359e+00 8.70737508e-02 -2.37746537e-02
-6.38455808e-01 -4.82004434e-01 1.04137719e+00 3.68848085e-01
2.93171138e-01 -1.17523515e+00 6.10074341e-01 4.24017876e-01
-6.85248137e-01 -1.66807783e+00 -8.54235291e-01 3.34167749e-01
-8.53157878e-01 1.93355121e-02 -9.79313612e-01 -1.14650989e+00
4.51047450e-01 2.45020896e-01 1.19927335e+00 -3.01935285e-01
6.52066827e-01 -1.69418052e-01 -6.05858564e-01 -4.20661867e-01
-1.29254234e+00 8.27287138e-01 3.30889612e-01 -7.43778050e-02
5.45321405e-01 -3.05363894e-01 -6.80769309e-02 5.79901278e-01
-4.41001117e-01 2.74044394e-01 6.85504854e-01 8.89679551e-01
3.77538204e-01 -8.83830011e-01 1.02855146e+00 -8.13622773e-01
5.43735027e-01 -4.42567319e-01 -3.45398843e-01 4.86458182e-01
-7.00631142e-01 1.53810114e-01 8.11424553e-01 -3.81316662e-01
-1.23003924e+00 2.34977201e-01 -3.55680436e-01 -2.87708521e-01
-2.10914090e-01 5.81509769e-01 -2.22376049e-01 4.03813034e-01
9.52849388e-01 5.76760709e-01 -8.90807733e-02 -6.48258984e-01
6.36372089e-01 1.60439444e+00 4.39221948e-01 -4.87562090e-01
6.70569181e-01 -1.09795034e-01 -6.87468112e-01 -8.05553615e-01
-7.17959106e-01 -4.65213716e-01 -7.18661904e-01 1.72892377e-01
8.31037760e-01 -1.22055829e+00 1.94619641e-01 2.43792221e-01
-1.41452885e+00 -6.61125183e-01 -1.02154724e-01 6.32416546e-01
-7.27831066e-01 1.40267372e-01 -7.07742333e-01 -5.91296315e-01
-7.59406447e-01 -1.41188169e+00 1.33754671e+00 -4.19206321e-01
-3.88604671e-01 -8.24524760e-01 2.62785316e-01 9.02310431e-01
5.87816715e-01 -6.68797016e-01 6.57105446e-01 -1.09632552e+00
-3.76637071e-01 1.87668681e-01 1.32620946e-01 6.88641369e-01
2.44603291e-01 -2.00608656e-01 -9.54027891e-01 -2.95324296e-01
-3.15653503e-01 -6.45152032e-01 5.33470869e-01 5.42040989e-02
5.51434338e-01 -4.48837876e-01 -1.52577862e-01 4.34888870e-01
6.32293105e-01 2.21654713e-01 3.80838424e-01 -2.27497704e-02
6.03542507e-01 7.01933444e-01 5.63790679e-01 -1.46957710e-01
3.51786047e-01 8.70020211e-01 -2.23829970e-01 -2.36515492e-01
-6.84940636e-01 -5.96019149e-01 9.67762053e-01 1.82190740e+00
3.43966156e-01 -5.47699988e-01 -1.12442160e+00 8.07975233e-01
-1.64498389e+00 -5.50143838e-01 -1.48882776e-01 2.17291975e+00
1.32648814e+00 -6.56895712e-02 1.62630156e-01 -3.96299392e-01
8.05338383e-01 2.71634143e-02 -2.96213806e-01 -6.12519562e-01
-2.28286669e-01 3.65645170e-01 5.57106137e-01 5.95691502e-01
-8.40944946e-01 1.88310683e+00 6.81580257e+00 1.00048435e+00
-1.26720655e+00 6.36804760e-01 6.67222738e-01 7.73794577e-02
-3.37219864e-01 2.84531146e-01 -9.93146837e-01 4.96685088e-01
1.85583675e+00 1.02028279e-02 6.12236857e-01 9.25749719e-01
4.66132700e-01 6.24461532e-01 -1.03786588e+00 7.17353404e-01
4.36040461e-02 -1.29595423e+00 -2.58089192e-02 -8.36833939e-02
1.10619998e+00 8.07514012e-01 -4.57095653e-02 5.80196977e-01
6.29882395e-01 -8.99218321e-01 7.58152366e-01 -2.14569822e-01
1.41095757e+00 -5.01380920e-01 4.32318807e-01 5.69483399e-01
-6.18862987e-01 5.50600708e-01 -3.10207278e-01 2.27162793e-01
4.19652641e-01 2.92224735e-01 -1.65671694e+00 4.38129723e-01
1.37520671e-01 4.62841213e-01 -2.77233094e-01 5.75766921e-01
-6.32764935e-01 1.44429791e+00 -2.86808163e-01 -1.94697186e-01
4.81506616e-01 -5.47880083e-02 4.90231246e-01 1.54520166e+00
5.99138856e-01 -2.26867229e-01 2.20188901e-01 4.88326818e-01
-4.82125789e-01 3.18168163e-01 -6.42515302e-01 -4.39377546e-01
6.37844145e-01 9.17766571e-01 -4.07892585e-01 -6.41382277e-01
-5.10194421e-01 1.33105743e+00 2.82701224e-01 4.98087943e-01
-5.43840528e-01 -1.63560942e-01 3.26517731e-01 -3.34690213e-02
3.51419020e-03 -5.31508923e-01 -3.93312216e-01 -1.37986255e+00
4.08204198e-02 -1.51900411e+00 -7.07592741e-02 -6.89319432e-01
-1.16835022e+00 1.13394642e+00 -4.22612965e-01 -1.32625854e+00
-8.08072090e-01 -3.52294147e-01 -3.38835239e-01 1.08687627e+00
-1.34835541e+00 -1.43887269e+00 3.57731849e-01 3.58431041e-01
1.29130220e+00 -7.61091053e-01 1.11436808e+00 3.86299431e-01
-5.44330835e-01 8.19868207e-01 4.82673258e-01 3.25247407e-01
1.21245646e+00 -1.03026903e+00 1.34512806e+00 1.13600147e+00
6.41211927e-01 3.44917566e-01 4.43695009e-01 -7.76416898e-01
-1.36460876e+00 -1.17626178e+00 1.69329774e+00 -7.45598376e-01
7.48915255e-01 -7.84897685e-01 -6.87626660e-01 8.56425226e-01
6.11352861e-01 -1.55120656e-01 6.79607868e-01 1.51153669e-01
-2.72003561e-01 4.34194319e-02 -7.34390080e-01 7.44328558e-01
1.24077106e+00 -8.16256881e-01 -5.46525836e-01 8.03556979e-01
1.27820849e+00 -3.76805127e-01 -6.89234912e-01 2.26260170e-01
9.43306461e-02 -1.24322154e-01 5.80872953e-01 -9.51605916e-01
2.94367880e-01 -1.59228072e-01 -3.74068946e-01 -1.86804700e+00
6.14872463e-02 -1.30403197e+00 1.52773857e-01 1.39571357e+00
1.32304788e+00 -4.51741308e-01 6.71936154e-01 2.26880461e-01
-7.89670646e-01 -3.08045626e-01 -1.06286550e+00 -1.13675237e+00
2.98533708e-01 -3.31271499e-01 5.46928942e-01 1.08892202e+00
2.14182138e-01 8.73660266e-01 -4.55413401e-01 -4.59842347e-02
2.56963909e-01 -2.94460654e-01 8.23903382e-01 -4.97815639e-01
-5.18834591e-01 -2.20211372e-01 1.77781820e-01 -1.51193845e+00
4.52270269e-01 -1.36497307e+00 4.90531296e-01 -1.44034314e+00
-9.02496427e-02 -3.96833479e-01 1.71531573e-01 6.85007691e-01
-1.42106250e-01 4.56287824e-02 8.14576373e-02 4.02895570e-01
-3.07249725e-01 7.19295323e-01 1.09162474e+00 -2.14818090e-01
-2.91427553e-01 1.44348845e-01 -5.25981665e-01 2.01369300e-01
8.27037752e-01 -7.49588013e-01 -4.55950171e-01 -1.12086010e+00
-2.54319042e-01 3.59142661e-01 -4.79236543e-01 -7.14648187e-01
2.29291663e-01 -2.14838371e-01 -2.58184582e-01 -4.48345810e-01
2.28649810e-01 -4.07582849e-01 -1.67922959e-01 9.58207771e-02
-5.98133802e-01 -3.45984511e-02 -1.46754161e-02 5.59388176e-02
-3.82509947e-01 -1.68145284e-01 6.67677641e-01 -6.69876039e-02
-3.45073879e-01 2.16847643e-01 -6.31809831e-01 2.30318040e-01
5.31576931e-01 1.65636301e-01 -4.09526587e-01 -6.84652150e-01
-7.07646728e-01 1.16830245e-01 1.63046420e-01 7.34718740e-01
3.47125232e-01 -1.36598313e+00 -1.34474850e+00 3.17686528e-01
4.11811262e-01 -1.56823725e-01 -3.03848892e-01 8.37700784e-01
-3.08748543e-01 7.70273745e-01 3.46132934e-01 -7.18213558e-01
-1.03770757e+00 2.37815440e-01 1.47372156e-01 -1.54196963e-01
-1.46244720e-01 9.13818240e-01 -3.36638600e-01 -1.15169549e+00
2.52658546e-01 -1.67753533e-01 4.40623939e-01 -3.69018197e-01
2.48013377e-01 1.69717640e-01 6.10036790e-01 -8.61291111e-01
-3.40968519e-01 -3.65879424e-02 -1.24266699e-01 -8.62299263e-01
1.10152996e+00 -2.87571758e-01 1.82967514e-01 3.38557690e-01
1.13486683e+00 2.57229861e-02 -1.00543439e+00 -6.21981680e-01
2.82350093e-01 -7.59502724e-02 3.40636484e-02 -1.25741947e+00
-6.68107390e-01 1.09162641e+00 8.33752453e-02 -1.34122416e-01
7.91114926e-01 1.82894096e-01 1.11337209e+00 6.88832343e-01
4.28554058e-01 -1.27685857e+00 -8.07508454e-02 1.01101184e+00
8.10409427e-01 -1.42251098e+00 -6.95110261e-01 -4.00141835e-01
-1.11044538e+00 7.76122332e-01 3.47346842e-01 5.09696603e-01
2.41558358e-01 2.36695066e-01 6.25664055e-01 3.75787944e-01
-8.68499219e-01 -2.96514213e-01 3.11167419e-01 5.87415934e-01
7.60868132e-01 3.96144718e-01 -2.16238827e-01 3.69303018e-01
-4.96495515e-01 -2.33660191e-01 3.04233551e-01 6.34384751e-01
-3.97631735e-01 -1.63092315e+00 -1.05652444e-01 -7.12159183e-03
-6.36538029e-01 -7.52009451e-01 -8.44594896e-01 2.72809386e-01
-5.09356081e-01 1.50091720e+00 -1.55596986e-01 -4.23683524e-01
3.61280084e-01 5.85594058e-01 8.98217186e-02 -1.12831771e+00
-6.18514299e-01 4.78058875e-01 7.78735638e-01 -1.49218574e-01
-2.46516198e-01 -5.41782320e-01 -1.07828867e+00 -1.29135340e-01
-4.66479838e-01 3.66638541e-01 1.10941291e+00 9.70469117e-01
5.76423168e-01 2.02473223e-01 9.40675080e-01 -4.80086625e-01
-8.82664561e-01 -1.55363405e+00 1.04807176e-01 4.11485732e-02
1.40378490e-01 5.00204451e-02 -3.64320695e-01 4.29353803e-01] | [14.489651679992676, 7.166929721832275] |
f33c714f-7a27-44a3-8bcc-945445a581ec | counterfactual-explanations-in-sequential | 2107.02776 | null | https://arxiv.org/abs/2107.02776v2 | https://arxiv.org/pdf/2107.02776v2.pdf | Counterfactual Explanations in Sequential Decision Making Under Uncertainty | Methods to find counterfactual explanations have predominantly focused on one step decision making processes. In this work, we initiate the development of methods to find counterfactual explanations for decision making processes in which multiple, dependent actions are taken sequentially over time. We start by formally characterizing a sequence of actions and states using finite horizon Markov decision processes and the Gumbel-Max structural causal model. Building upon this characterization, we formally state the problem of finding counterfactual explanations for sequential decision making processes. In our problem formulation, the counterfactual explanation specifies an alternative sequence of actions differing in at most k actions from the observed sequence that could have led the observed process realization to a better outcome. Then, we introduce a polynomial time algorithm based on dynamic programming to build a counterfactual policy that is guaranteed to always provide the optimal counterfactual explanation on every possible realization of the counterfactual environment dynamics. We validate our algorithm using both synthetic and real data from cognitive behavioral therapy and show that the counterfactual explanations our algorithm finds can provide valuable insights to enhance sequential decision making under uncertainty. | ['Manuel Gomez-Rodriguez', 'Abir De', 'Stratis Tsirtsis'] | 2021-07-06 | null | http://proceedings.neurips.cc/paper/2021/hash/fd0a5a5e367a0955d81278062ef37429-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/fd0a5a5e367a0955d81278062ef37429-Paper.pdf | neurips-2021-12 | ['decision-making-under-uncertainty', 'counterfactual-explanation', 'decision-making-under-uncertainty'] | ['medical', 'miscellaneous', 'reasoning'] | [ 7.44876146e-01 9.20327544e-01 -3.60769242e-01 -1.96264178e-01
-3.54939610e-01 -4.51872647e-01 9.12057936e-01 2.16968566e-01
-3.56578141e-01 1.07430458e+00 6.95048749e-01 -1.03935897e+00
-8.18633199e-01 -6.44401848e-01 -4.06256706e-01 -5.27511954e-01
-5.30326843e-01 8.01512003e-01 -3.76376420e-01 1.45026073e-01
3.59235942e-01 2.75238544e-01 -1.32692838e+00 3.93051505e-01
7.35700250e-01 4.76938963e-01 2.26823047e-01 8.44999969e-01
-9.49158594e-02 7.39441037e-01 -2.54034191e-01 -3.30479354e-01
3.65732491e-01 -7.46188104e-01 -1.06308913e+00 3.91234875e-01
-5.78402340e-01 -4.08678859e-01 -1.98945105e-02 9.40291405e-01
3.31893235e-01 2.87773103e-01 7.67576098e-01 -1.46359146e+00
-3.56929272e-01 1.14036357e+00 -4.26665068e-01 2.13882968e-01
5.37696183e-01 4.63867813e-01 8.81194949e-01 6.10268489e-02
4.81793702e-01 1.95632601e+00 6.54342175e-02 9.67996657e-01
-1.78961933e+00 -3.46344888e-01 6.48176312e-01 1.58960372e-01
-5.37811100e-01 -1.23296969e-01 4.54125047e-01 -2.79545397e-01
7.14115024e-01 4.20887172e-01 8.51084650e-01 1.04762685e+00
5.83620191e-01 9.74500537e-01 1.60095048e+00 -6.04437649e-01
7.82925904e-01 -4.24933761e-01 -8.03347081e-02 1.20470218e-01
4.99521285e-01 7.43066847e-01 -1.03781655e-01 -6.13352299e-01
5.25525510e-01 3.11160833e-01 -2.18566686e-01 -3.41954589e-01
-1.10156739e+00 8.62150311e-01 -1.43571511e-01 -1.86614469e-02
-1.11367691e+00 2.83612341e-01 7.85048157e-02 4.81095374e-01
-4.25255531e-03 6.20474815e-01 -5.32441497e-01 3.76014970e-02
-3.15428227e-01 6.19918406e-01 9.21678424e-01 2.97473013e-01
1.68463483e-01 -2.76716799e-01 -6.98520482e-01 -2.61342246e-02
2.92611688e-01 5.92138648e-01 3.07795614e-01 -1.40825641e+00
4.55704570e-01 6.89538419e-02 9.02159393e-01 -3.15946370e-01
-3.22173923e-01 -1.04660727e-02 -2.67595440e-01 4.09300506e-01
4.99736577e-01 -5.98406613e-01 -7.80060530e-01 1.84390688e+00
1.93674996e-01 2.10882694e-01 2.86165595e-01 5.86645246e-01
-7.48762190e-01 4.38322932e-01 3.79342437e-01 -1.14354634e+00
9.78437960e-01 -2.18689963e-01 -1.02328098e+00 3.53858098e-02
5.81680715e-01 -5.11849642e-01 6.42643154e-01 3.39139909e-01
-1.12032223e+00 1.44725949e-01 -5.03327370e-01 9.21338677e-01
5.67990899e-01 -7.01133728e-01 7.59950042e-01 7.20887363e-01
-8.91921699e-01 9.49123442e-01 -8.43325973e-01 -1.46343827e-01
3.04275721e-01 1.68893889e-01 2.99540460e-01 -9.40330885e-03
-1.10253131e+00 8.35514784e-01 7.56420195e-01 -1.01662584e-01
-1.43208313e+00 -5.39777815e-01 -2.75222361e-01 2.21632078e-01
1.18233669e+00 -1.04699540e+00 1.90747869e+00 -1.12573731e+00
-1.64924312e+00 1.83858976e-01 -2.22640082e-01 -1.04122591e+00
9.93593395e-01 6.66456148e-02 -2.93575555e-01 -7.23519474e-02
2.49353081e-01 1.99835822e-01 7.20882595e-01 -1.12525749e+00
-8.52895916e-01 -5.51076353e-01 3.01982075e-01 4.22085494e-01
5.52323699e-01 -2.85742860e-02 6.19750142e-01 -3.11632067e-01
2.89510656e-02 -1.20091367e+00 -1.01099050e+00 -5.36905944e-01
-7.46055126e-01 -1.48174673e-01 7.70176426e-02 -9.31011885e-02
1.02011502e+00 -1.55707204e+00 -1.71784043e-01 1.06811903e-01
1.44346859e-02 -4.09840524e-01 4.35812585e-02 4.79291111e-01
-4.92226362e-01 4.14583892e-01 -3.35861593e-01 5.25561981e-02
1.99419454e-01 3.15865338e-01 -7.39145994e-01 4.68365639e-01
-2.40187973e-01 7.52140224e-01 -8.83470297e-01 2.84725204e-02
2.13329658e-01 -5.39040804e-01 -3.74703556e-01 1.02076650e-01
-5.90206206e-01 2.69729584e-01 -6.89543903e-01 9.83744189e-02
5.01324296e-01 5.06816171e-02 8.49046648e-01 7.68731534e-01
-2.89244145e-01 3.67885351e-01 -1.34489763e+00 9.99279797e-01
-3.08226019e-01 1.80389494e-01 -4.11651760e-01 -8.94786894e-01
1.50923222e-01 6.53871715e-01 5.12337208e-01 -2.40666807e-01
3.42205226e-01 4.16101180e-02 3.53814006e-01 -3.16570789e-01
5.34761250e-02 -1.17755735e+00 -1.95151150e-01 1.11458027e+00
-5.16334534e-01 7.32720494e-02 -1.41719475e-01 1.19202144e-01
1.11206222e+00 -1.87256530e-01 9.65103149e-01 -3.32443386e-01
2.00060278e-01 1.09462459e-02 6.92416310e-01 1.43380511e+00
-3.81902009e-01 1.00821899e-02 1.05518723e+00 -4.40783352e-01
-8.92380714e-01 -1.10936618e+00 1.87900439e-01 4.66166347e-01
1.73970871e-02 1.58409923e-01 -4.69697654e-01 -6.62480712e-01
1.18171692e-01 1.65939176e+00 -9.15439785e-01 -1.13286719e-01
-1.89846456e-01 -9.12973523e-01 -1.42816827e-01 3.00544649e-01
3.90686512e-01 -1.06563663e+00 -1.09179723e+00 5.28758407e-01
-1.14524901e-01 -4.81940150e-01 -2.18474999e-01 -1.08404182e-01
-1.22116756e+00 -1.36440921e+00 -4.84406829e-01 5.49985409e-01
3.92287433e-01 1.83061242e-01 6.52117550e-01 -3.86443317e-01
2.02406317e-01 4.16030586e-01 1.53064594e-01 -1.03947115e+00
-7.31124341e-01 -8.32629025e-01 4.21223909e-01 5.56905530e-02
1.87424958e-01 -3.50266904e-01 -6.86419308e-01 -2.92521268e-02
-8.39694262e-01 4.12295192e-01 3.49626958e-01 5.57231903e-01
5.43658435e-01 3.51809591e-01 4.60775465e-01 -8.97067428e-01
9.53794241e-01 -6.58653140e-01 -7.08131671e-01 4.55120921e-01
-8.99021685e-01 6.09533012e-01 4.87299442e-01 -3.92953426e-01
-1.80845034e+00 3.14749889e-02 3.74672413e-01 1.97599474e-02
-2.99453527e-01 3.54624867e-01 -3.87350261e-01 9.61086452e-01
4.36217427e-01 9.40013677e-02 -3.00773662e-02 -3.27851057e-01
6.13288343e-01 2.86335856e-01 2.36212432e-01 -7.18412757e-01
2.45568469e-01 8.74656677e-01 2.43872389e-01 -2.18564913e-01
-7.64200985e-01 4.46591601e-02 -2.38847733e-01 -2.38765419e-01
8.63289416e-01 -4.41320896e-01 -1.34624457e+00 8.30612555e-02
-1.16793048e+00 -3.84872586e-01 -6.48326278e-01 6.87665224e-01
-1.47982562e+00 1.33931786e-01 7.91737810e-03 -1.69206607e+00
2.58287638e-01 -8.59482944e-01 4.40055758e-01 9.23915654e-02
-4.83158410e-01 -9.74218190e-01 3.33313048e-01 2.01040298e-01
-1.25350565e-01 5.86494565e-01 8.84734511e-01 -5.48455417e-01
-6.82702363e-01 2.17860132e-01 3.99596602e-01 -3.49618763e-01
2.25315928e-01 -1.88174322e-01 -6.00904703e-01 -1.48766980e-01
3.84370089e-01 3.73177767e-01 5.08446991e-01 9.79246736e-01
1.00343728e+00 -9.48551595e-01 -7.42187083e-01 -1.77213147e-01
1.40787327e+00 1.00013328e+00 4.89893228e-01 5.04464805e-01
-1.05791405e-01 9.67484653e-01 8.49569261e-01 6.77552938e-01
1.79743811e-01 3.78088087e-01 4.90950823e-01 5.25158048e-01
7.89693773e-01 -4.84928042e-01 2.98274219e-01 -6.52702630e-01
-2.91273713e-01 -1.88876390e-01 -7.88660407e-01 8.37929070e-01
-2.18011618e+00 -1.49691021e+00 -1.47501305e-01 2.55548573e+00
5.41748703e-01 2.21745819e-01 3.31677049e-01 6.98778033e-02
8.39660108e-01 -3.44762146e-01 -7.90169358e-01 -7.90504098e-01
1.79688975e-01 -1.18255094e-01 7.25299895e-01 7.72962451e-01
-5.57263792e-01 5.14497280e-01 6.60424232e+00 3.30158651e-01
-5.04505634e-01 1.33286431e-01 7.98602641e-01 -5.13991535e-01
-7.77980685e-01 3.72908026e-01 -4.35491353e-01 4.98735040e-01
1.43170106e+00 -1.07690895e+00 2.91740268e-01 5.06844819e-01
1.11258519e+00 -4.16806519e-01 -1.40061224e+00 3.45496535e-01
-9.49179053e-01 -1.47362185e+00 7.84332305e-02 3.18085581e-01
1.09264791e+00 -6.10271811e-01 -2.24120423e-01 -4.38726060e-02
1.50447690e+00 -8.60649109e-01 1.02925444e+00 6.25204325e-01
2.64125466e-01 -1.07828963e+00 5.41277170e-01 7.55884051e-01
-3.82928252e-01 -7.48655438e-01 -4.86949570e-02 -7.26751566e-01
5.87616920e-01 5.32770455e-01 -1.26280689e+00 6.56097353e-01
1.37133896e-01 1.05339596e-02 4.47422117e-01 8.23002458e-01
-7.08056629e-01 6.12616241e-01 -1.20163701e-01 1.90211590e-02
3.27537060e-01 -1.12674455e-03 6.78525746e-01 6.07613087e-01
3.40370387e-01 7.08858788e-01 8.30964074e-02 7.92211831e-01
4.07360852e-01 -3.32087010e-01 -9.29494143e-01 -7.92508665e-03
4.85459268e-01 2.46210337e-01 -1.02489841e+00 -4.19565111e-01
4.68870625e-02 8.44287276e-01 -8.07298645e-02 3.84710759e-01
-5.05203068e-01 4.46690381e-01 1.00172877e+00 -1.29925292e-02
-7.29964375e-02 2.50769585e-01 -6.89050078e-01 -9.70906317e-01
-4.22114521e-01 -7.93590128e-01 9.68694627e-01 -6.37560129e-01
-1.01813066e+00 -2.03019883e-02 4.52794433e-01 -6.69511437e-01
-6.43890321e-01 -1.07671060e-01 -9.09306407e-01 1.06019306e+00
-9.00602877e-01 -1.31624609e-01 7.21080840e-01 1.62268177e-01
7.26443529e-01 3.24892431e-01 5.67951083e-01 -9.89502728e-01
-4.21781510e-01 -4.27621603e-01 2.40164638e-01 -7.40898609e-01
-4.93792519e-02 -1.45441067e+00 3.41641784e-01 1.05255544e+00
-2.89270103e-01 6.26394928e-01 1.34026742e+00 -9.32511628e-01
-8.99094999e-01 -7.48922884e-01 7.93426514e-01 -3.81970108e-01
8.41945589e-01 3.21607113e-01 -6.07095242e-01 1.00035322e+00
2.42182404e-01 -9.58927870e-01 3.89121115e-01 2.36643702e-02
4.93739396e-01 3.40132326e-01 -1.24333596e+00 1.20196867e+00
1.11036253e+00 -8.23051631e-02 -1.09676862e+00 2.64223784e-01
8.91350210e-01 4.57068607e-02 -1.45252347e-01 1.36384843e-02
4.16875005e-01 -7.88646221e-01 8.42525482e-01 -1.27674139e+00
5.75112998e-01 -1.51404276e-01 -8.23559836e-02 -1.60059452e+00
-2.87808567e-01 -1.32962561e+00 7.03167543e-02 6.24302685e-01
3.69824052e-01 -8.26976299e-01 5.54695189e-01 1.23491549e+00
3.09475601e-01 -5.72994590e-01 -1.19418991e+00 -9.71273601e-01
5.40240966e-02 -9.24733460e-01 1.08768129e+00 6.68866396e-01
2.25760579e-01 6.85024932e-02 -3.30729097e-01 1.63337365e-01
1.05593228e+00 5.21063149e-01 3.32972974e-01 -9.43741858e-01
-5.76489091e-01 -3.00804943e-01 2.55996615e-01 -6.23339236e-01
2.90113747e-01 -4.41161960e-01 1.37211069e-01 -1.57997561e+00
5.45205772e-01 -1.69449583e-01 -1.67125523e-01 3.17087770e-01
-3.95577282e-01 -9.39177334e-01 5.92415988e-01 1.15543507e-01
-2.91858524e-01 6.31532311e-01 1.22649300e+00 4.69646268e-02
-4.93727118e-01 6.03478014e-01 -1.16823161e+00 9.72197235e-01
8.20358217e-01 -7.21658051e-01 -7.13732958e-01 1.73460290e-01
6.15612827e-02 1.03381610e+00 3.10172617e-01 -1.52644992e-01
-2.32458353e-01 -1.28635049e+00 -9.33144316e-02 -3.74626845e-01
-2.42525131e-01 -7.00961471e-01 7.08219767e-01 1.26430237e+00
-9.50469911e-01 -7.96948671e-02 3.75149548e-02 1.27641225e+00
5.01063704e-01 -4.99706954e-01 5.10645330e-01 -3.49145442e-01
-5.98499715e-01 6.55472800e-02 -1.08234429e+00 -3.27522486e-01
1.46863651e+00 1.42082628e-02 -8.28894135e-03 -8.25999737e-01
-1.21766520e+00 4.10283655e-01 1.84727281e-01 1.83084205e-01
4.43435192e-01 -1.05899906e+00 -7.48687863e-01 -5.85365355e-01
-3.09092343e-01 -5.82947135e-01 4.00136650e-01 5.38897991e-01
1.10292934e-01 8.32267761e-01 -2.26782501e-01 -3.18532251e-02
-1.19783711e+00 8.99658740e-01 5.21608710e-01 -5.98008394e-01
-4.75529283e-01 1.96595341e-01 4.42123920e-01 4.09847088e-02
-5.57759464e-01 -3.03939044e-01 -1.26366258e-01 -8.44160542e-02
8.19876671e-01 8.53321254e-01 -5.03573179e-01 1.80785269e-01
-1.35198429e-01 -5.00774086e-01 2.95597743e-02 -9.90323305e-01
1.19711483e+00 -6.82292998e-01 1.26029521e-01 4.67890561e-01
2.63674706e-01 -4.09944147e-01 -1.66804564e+00 -7.80290067e-02
4.89492804e-01 -7.33372092e-01 -2.30253428e-01 -9.28404331e-01
-2.27915719e-01 3.97770554e-01 3.66673708e-01 6.90558970e-01
1.05405319e+00 6.68225884e-02 -4.21073660e-02 2.27153644e-01
7.29499698e-01 -8.68812859e-01 -2.95632690e-01 5.63347228e-02
9.34157550e-01 -8.53403389e-01 -1.37026221e-01 -2.14873090e-01
-8.64597142e-01 8.31671298e-01 1.23497164e-02 1.50499359e-01
4.16395575e-01 3.57991941e-02 -2.79999048e-01 -1.09457590e-01
-1.36850417e+00 -4.34208751e-01 -2.77827829e-01 5.84522843e-01
9.58779529e-02 9.30808723e-01 -8.06581438e-01 4.84621525e-01
-2.50191540e-01 3.95381749e-01 1.08643055e+00 8.83347690e-01
-4.17479575e-01 -9.73619282e-01 -8.45940590e-01 5.36671937e-01
-5.62214077e-01 1.40070304e-01 -3.20905685e-01 6.42125905e-01
-2.69886613e-01 1.14151752e+00 6.76157475e-02 3.40238959e-01
4.64484155e-01 2.15973705e-01 6.23602152e-01 -5.91110826e-01
1.42841965e-01 8.52754787e-02 1.71915352e-01 -6.91549480e-01
-3.32994759e-01 -1.26136458e+00 -1.28977883e+00 -5.71397424e-01
1.20916627e-01 1.01771265e-01 2.43343979e-01 1.52074456e+00
3.54123563e-02 5.95579088e-01 6.93580627e-01 -3.60150903e-01
-1.23476529e+00 -6.80922747e-01 -5.55043638e-01 1.18448153e-01
2.78540373e-01 -5.94411135e-01 -2.37774551e-01 -4.89419736e-02] | [8.109895706176758, 5.55803918838501] |
00b93cdb-b8db-4ab9-a5d3-47f79c49aa76 | situatedgen-incorporating-geographical-and | 2306.12552 | null | https://arxiv.org/abs/2306.12552v1 | https://arxiv.org/pdf/2306.12552v1.pdf | SituatedGen: Incorporating Geographical and Temporal Contexts into Generative Commonsense Reasoning | Recently, commonsense reasoning in text generation has attracted much attention. Generative commonsense reasoning is the task that requires machines, given a group of keywords, to compose a single coherent sentence with commonsense plausibility. While existing datasets targeting generative commonsense reasoning focus on everyday scenarios, it is unclear how well machines reason under specific geographical and temporal contexts. We formalize this challenging task as SituatedGen, where machines with commonsense should generate a pair of contrastive sentences given a group of keywords including geographical or temporal entities. We introduce a corresponding English dataset consisting of 8,268 contrastive sentence pairs, which are built upon several existing commonsense reasoning benchmarks with minimal manual labor. Experiments show that state-of-the-art generative language models struggle to generate sentences with commonsense plausibility and still lag far behind human performance. Our dataset is publicly available at https://github.com/yunx-z/situated_gen. | ['Xiaojun Wan', 'Yunxiang Zhang'] | 2023-06-21 | null | null | null | null | ['text-generation'] | ['natural-language-processing'] | [ 5.36261737e-01 5.44698596e-01 2.51598358e-02 -4.31215644e-01
-9.72825050e-01 -8.19374919e-01 1.40860260e+00 9.29875225e-02
1.02528244e-01 1.09015942e+00 7.25190341e-01 -5.89993954e-01
1.24413118e-01 -1.11095643e+00 -6.80080652e-01 -6.42555803e-02
5.70786655e-01 7.87140191e-01 -5.97744621e-02 -7.50986457e-01
2.53551066e-01 -1.85978398e-01 -1.22955728e+00 5.48385382e-01
1.33050060e+00 5.16371548e-01 1.77525446e-01 7.84569085e-01
-2.46599346e-01 1.38718975e+00 -1.01722884e+00 -1.15961611e+00
-2.68165231e-01 -1.04937744e+00 -1.30160546e+00 -3.68750811e-01
4.23191369e-01 3.25230649e-03 -3.24694991e-01 1.18658173e+00
4.39988464e-01 2.99702436e-01 7.62255847e-01 -1.46348214e+00
-1.58820760e+00 1.60901868e+00 1.07826732e-01 3.48674834e-01
8.71715188e-01 2.74154127e-01 1.40485835e+00 -6.55670941e-01
8.61554980e-01 1.35766876e+00 6.14413798e-01 7.24467099e-01
-1.37570536e+00 -3.88190269e-01 -1.60131618e-01 3.48883867e-01
-1.10088837e+00 -4.09183711e-01 1.01083207e+00 -1.47927895e-01
1.27154088e+00 6.34608567e-01 7.23604023e-01 1.76026690e+00
2.30465472e-01 7.08798707e-01 1.26154685e+00 -3.33470643e-01
3.34610015e-01 -2.54032075e-01 9.88222286e-02 4.59891915e-01
4.09011990e-01 -2.70449698e-01 -3.99413198e-01 -1.78328529e-01
3.32797498e-01 -1.85198277e-01 -1.79316536e-01 4.25989926e-01
-1.81939125e+00 1.06332088e+00 5.47042131e-01 4.98288929e-01
-4.16211516e-01 5.91361821e-01 2.90430069e-01 -8.42389241e-02
4.72184092e-01 8.45435977e-01 -1.04466744e-01 -4.57787514e-01
-9.98840034e-01 8.52879941e-01 1.03188133e+00 1.19296789e+00
2.46881142e-01 -6.47720620e-02 -4.51963246e-01 7.66130686e-01
-1.18125295e-02 8.12867403e-01 4.94553655e-01 -9.27713633e-01
8.34782839e-01 5.29470384e-01 -8.59595761e-02 -1.00333214e+00
-3.93340178e-02 4.38689115e-03 -8.80921841e-01 -5.04192173e-01
1.25595897e-01 -2.36461043e-01 -6.66505337e-01 1.90822554e+00
-2.58651003e-02 1.23450048e-01 4.95799124e-01 5.79737186e-01
1.21692026e+00 8.17445934e-01 1.16576508e-01 -4.86774743e-02
1.49454451e+00 -7.41707385e-01 -7.84598291e-01 -7.01150358e-01
4.18676555e-01 -6.73488677e-01 1.50514424e+00 -2.54461020e-02
-1.47388530e+00 -1.32352009e-01 -7.40548432e-01 -6.81096017e-01
-7.65627801e-01 -2.09790140e-01 8.28570008e-01 2.76104152e-01
-9.08368170e-01 4.13223296e-01 -3.87704581e-01 -2.29681402e-01
3.91668856e-01 -5.83349884e-01 3.40931974e-02 1.18605629e-01
-1.86227822e+00 1.28218198e+00 7.40221441e-01 -8.43641758e-02
-7.49300957e-01 -7.75771677e-01 -1.09823763e+00 -1.65827990e-01
3.68771046e-01 -1.25907409e+00 1.65382886e+00 -2.99096048e-01
-1.18576467e+00 1.16088629e+00 -1.73949316e-01 -7.97601640e-01
7.78423786e-01 -2.37735644e-01 -5.47617555e-01 -1.50381058e-01
6.72320127e-01 5.73925972e-01 3.23984861e-01 -1.16465282e+00
-3.41046244e-01 1.82714686e-02 4.38239932e-01 1.63688242e-01
4.04743373e-01 8.63890648e-02 2.82479078e-01 -9.10153687e-01
-6.01125434e-02 -7.59099901e-01 -3.97132011e-04 -6.17642999e-01
-1.18897760e+00 -6.48371041e-01 3.84303153e-01 -6.76009893e-01
1.22279537e+00 -1.40442598e+00 1.73644736e-01 -3.14084649e-01
1.54985145e-01 -3.81240100e-01 2.26033002e-01 6.71934605e-01
-1.61828861e-01 5.66935360e-01 -5.79194546e-01 -2.22107619e-01
6.01937234e-01 2.28626877e-01 -1.14162481e+00 -1.63867727e-01
3.20064038e-01 1.48778892e+00 -1.51327980e+00 -7.15394318e-01
4.23035398e-02 1.54593378e-01 -4.96332526e-01 -8.41448531e-02
-6.77229583e-01 -2.75791824e-01 -1.96320340e-01 5.24788439e-01
2.10806385e-01 -3.77547175e-01 4.06803600e-02 3.80568765e-02
2.90325671e-01 1.06849945e+00 -6.01529181e-01 1.81837869e+00
-4.96799290e-01 7.61707067e-01 -8.85598242e-01 -4.22253966e-01
7.04212308e-01 2.12875798e-01 -3.02114874e-01 -4.18262035e-01
1.08183056e-01 1.75477952e-01 -1.10957928e-01 -3.17506552e-01
9.61708426e-01 -7.11926997e-01 -7.68000424e-01 6.28801346e-01
-9.32054594e-02 -1.23637795e+00 6.12520039e-01 5.85349619e-01
1.08570874e+00 5.00596734e-03 6.18269503e-01 -1.57680258e-01
9.05371234e-02 4.10079569e-01 2.34862164e-01 8.25736105e-01
1.01499848e-01 6.00375891e-01 7.70311773e-01 -2.41445526e-01
-9.68046486e-01 -1.61416495e+00 1.19018711e-01 7.72081554e-01
1.35802433e-01 -6.40634000e-01 -7.22267032e-01 -4.96925026e-01
-1.97699338e-01 2.05391407e+00 -8.53459835e-01 -2.18861088e-01
-5.06716967e-01 -5.68895578e-01 1.09780455e+00 5.08109450e-01
6.07496679e-01 -1.63922465e+00 -7.61533618e-01 2.52366483e-01
-9.38113689e-01 -1.10421979e+00 -2.45332152e-01 -6.00538738e-02
-4.89416301e-01 -9.45956707e-01 -3.18105996e-01 -4.40409988e-01
2.96872735e-01 -3.96307856e-02 1.89857459e+00 -1.53789759e-01
-2.15526059e-01 -4.51406352e-02 -2.78334975e-01 -8.23763847e-01
-9.61652756e-01 -1.64334830e-02 -3.06827962e-01 -8.23640525e-01
2.12275341e-01 -5.96028268e-01 -5.00028282e-02 -4.72293288e-01
-9.85563576e-01 6.66982830e-01 5.70883937e-02 8.84206235e-01
4.13924783e-01 -5.95786497e-02 5.45175791e-01 -9.80557442e-01
1.32681417e+00 -8.03561687e-01 7.01027513e-02 5.29623985e-01
-2.47677535e-01 1.36582464e-01 6.34352684e-01 -3.59926701e-01
-1.26859415e+00 -8.32301974e-01 5.44854142e-02 7.63973147e-02
-7.45021179e-02 5.82496107e-01 1.32144243e-03 1.10129201e+00
9.79657531e-01 5.61513424e-01 -4.55724746e-01 3.61630917e-01
1.00707293e+00 3.18712354e-01 9.59433913e-01 -9.94236171e-01
7.83393443e-01 3.07083368e-01 -1.34749696e-01 -5.10683060e-01
-1.33157778e+00 2.53019959e-01 -2.05976143e-01 -5.23477746e-03
9.68632221e-01 -7.84625053e-01 -2.64868230e-01 6.98599145e-02
-1.69677389e+00 -5.06889999e-01 -6.40121520e-01 -5.84677793e-02
-1.06133223e+00 1.38750285e-01 -5.98303854e-01 -5.93574345e-01
-7.20619559e-01 -6.90691411e-01 1.12802768e+00 6.05925657e-02
-1.05941916e+00 -1.22933590e+00 1.85309723e-01 5.33669293e-01
2.64594197e-01 7.18220174e-01 1.07970452e+00 -6.29372776e-01
-2.37704098e-01 3.16335522e-02 -2.94380542e-03 4.60888296e-02
3.18292111e-01 1.76145479e-01 -7.29150832e-01 6.33293808e-01
-2.07513779e-01 -4.89197910e-01 8.33585143e-01 8.53784904e-02
1.03014028e+00 -7.62653470e-01 -1.57237619e-01 2.35003963e-01
1.08419788e+00 1.49388397e-02 8.19397449e-01 2.13350952e-01
4.01897728e-01 2.72778064e-01 3.60210240e-01 3.83677900e-01
9.71711338e-01 2.26285532e-01 2.83781230e-01 2.61094481e-01
-2.72828698e-01 -7.68391132e-01 2.10188046e-01 4.22918469e-01
-3.17857385e-01 -4.89810497e-01 -1.17307138e+00 9.68926966e-01
-1.74544442e+00 -1.92190886e+00 -1.52968839e-01 1.29351962e+00
1.54610229e+00 2.05086276e-01 -1.80579081e-01 1.52287602e-01
6.19208872e-01 4.76711571e-01 -3.08431178e-01 -5.79630136e-01
-5.60412109e-01 1.87025726e-01 -2.13483274e-01 6.39326572e-01
-8.11849773e-01 1.20938873e+00 5.75447273e+00 8.10377300e-01
-6.73935711e-01 1.00737289e-01 4.89069909e-01 -2.26993099e-01
-1.11405933e+00 1.31583810e-01 -4.33855951e-01 7.27981329e-01
9.11193192e-01 -7.33528674e-01 5.99134266e-01 8.24815750e-01
2.12737974e-02 -8.40728506e-02 -1.16453850e+00 9.60888028e-01
3.32557678e-01 -1.79754674e+00 4.00060445e-01 -3.46523672e-01
1.02645028e+00 -1.12171285e-01 -8.34539011e-02 4.72140372e-01
9.31955159e-01 -1.16385031e+00 1.38650692e+00 7.45335698e-01
6.04255021e-01 -6.33086503e-01 4.50065017e-01 4.75105166e-01
-7.43701398e-01 2.51448333e-01 -2.88931549e-01 -2.57915348e-01
6.24132276e-01 8.42087090e-01 -1.06315601e+00 4.30667669e-01
3.77796441e-01 5.37023604e-01 -7.13797927e-01 1.40097320e-01
-1.03430307e+00 5.87092519e-01 -1.24864951e-01 -6.44412220e-01
2.81303883e-01 3.93423028e-02 6.55872643e-01 1.46703064e+00
4.41255540e-01 3.47269505e-01 -4.24519658e-01 1.75889957e+00
-3.10975909e-01 -3.85862648e-01 -9.68921602e-01 -4.53052551e-01
6.62801623e-01 9.89498913e-01 -6.23013496e-01 -8.32386911e-01
2.20716164e-01 1.15451801e+00 4.27926123e-01 1.25137493e-01
-1.35363579e+00 -4.12942529e-01 6.28049016e-01 -5.31588942e-02
-1.63551867e-01 -2.14466415e-02 -7.82064557e-01 -1.25302184e+00
1.67516395e-01 -7.41192818e-01 3.20569664e-01 -1.67859769e+00
-1.48414719e+00 6.58320725e-01 4.12830085e-01 -6.47059321e-01
-8.45942974e-01 -2.08526298e-01 -1.05562389e+00 8.45797062e-01
-9.64488745e-01 -1.39924836e+00 -3.71147841e-01 3.64257187e-01
8.65905464e-01 1.17938735e-01 8.82951736e-01 -5.45767188e-01
-1.65114045e-01 6.78307563e-02 -5.99641442e-01 1.22130617e-01
3.57722610e-01 -1.66987360e+00 1.10436952e+00 1.08407378e+00
3.55951428e-01 1.08346987e+00 1.19002748e+00 -9.81883526e-01
-9.39020991e-01 -1.15267515e+00 1.79644930e+00 -1.09766674e+00
1.03723276e+00 -3.72053653e-01 -5.03623307e-01 1.02029526e+00
8.45430076e-01 -2.95792013e-01 6.45499170e-01 5.69999553e-02
-5.32781184e-01 5.38566649e-01 -1.14005530e+00 1.31110990e+00
1.55862832e+00 -7.73028493e-01 -1.53406310e+00 7.42284834e-01
1.02845073e+00 -6.63809061e-01 -5.46842873e-01 -5.42626018e-03
2.31365804e-02 -7.80924559e-01 9.46098268e-01 -9.87476468e-01
1.43240607e+00 -3.89559388e-01 -2.95246571e-01 -1.79280198e+00
-2.17358693e-01 -8.17718804e-01 -2.88627863e-01 1.27124619e+00
6.80344343e-01 -5.39889157e-01 2.25217398e-02 6.87421203e-01
-1.92656532e-01 -6.10900939e-01 -6.85927629e-01 -7.49017537e-01
3.41032416e-01 -6.67598307e-01 1.00140202e+00 1.14375460e+00
3.46575826e-01 8.50466728e-01 8.06619674e-02 -2.66642869e-01
4.98055905e-01 4.56881732e-01 4.00233328e-01 -9.29095507e-01
-1.15475677e-01 -7.43322968e-01 -1.46356046e-01 -4.66743201e-01
5.72482884e-01 -1.22724271e+00 2.19097838e-01 -2.25514913e+00
2.67230034e-01 -7.89834931e-02 4.42241877e-01 6.87999785e-01
-5.69209874e-01 1.26238912e-01 2.47691736e-01 -1.55932739e-01
-4.60730761e-01 6.11734152e-01 1.23880494e+00 -3.39564830e-01
1.01771392e-01 -3.69954467e-01 -1.20877194e+00 7.23872006e-01
1.05775487e+00 -2.63572901e-01 -6.78224266e-01 -4.45691645e-01
8.93296480e-01 -7.96919167e-02 1.05978441e+00 -6.71916842e-01
7.07227662e-02 -7.85773575e-01 2.43474156e-01 -6.36727154e-01
3.13810736e-01 -1.35525525e-01 3.83824050e-01 2.69692272e-01
-7.74030268e-01 4.90853518e-01 2.51940005e-02 1.72766820e-01
-9.92413089e-02 -1.33661643e-01 3.80849391e-01 -3.82892221e-01
-6.09643996e-01 -3.56559187e-01 -4.71504580e-04 9.40554976e-01
1.00189590e+00 -1.84120890e-02 -8.82015049e-01 -4.67435628e-01
-4.09480542e-01 -2.15240195e-01 2.80404598e-01 5.36282063e-01
7.16571808e-01 -1.46556437e+00 -1.11669230e+00 -5.11417985e-01
2.30579272e-01 6.88009039e-02 1.98795304e-01 4.15618658e-01
-6.04219258e-01 4.08608884e-01 1.12969436e-01 -3.03834118e-02
-7.75429130e-01 5.37064910e-01 2.15607733e-01 -1.71999753e-01
-6.18215740e-01 9.76389229e-01 -2.22459078e-01 -4.55824524e-01
-6.48159027e-01 -9.63149786e-01 2.69837350e-01 -7.66980648e-02
5.42967737e-01 3.34558249e-01 -2.47369498e-01 -5.06401718e-01
-4.16463882e-01 -1.51451349e-01 3.07074100e-01 -4.57145184e-01
1.14296389e+00 2.60736704e-01 -4.25102025e-01 6.12995565e-01
7.40608811e-01 1.29352555e-01 -5.00649869e-01 4.33478914e-02
-1.05644085e-01 -3.53023969e-02 -3.49590749e-01 -1.19568682e+00
-1.87427282e-01 6.02463782e-01 -6.49001718e-01 6.23855293e-01
7.06584930e-01 5.06463230e-01 9.43405211e-01 4.87492949e-01
3.49075437e-01 -9.62771773e-01 -7.76143605e-03 9.87887621e-01
1.58778560e+00 -1.07325828e+00 -1.91445470e-01 -3.13365400e-01
-1.03543520e+00 8.05102468e-01 3.84446949e-01 -1.63901225e-01
1.80365562e-01 1.83699995e-01 -2.06497908e-01 -3.68381768e-01
-1.04380631e+00 -2.03227580e-01 2.80021429e-01 5.93415916e-01
6.79468989e-01 6.78683937e-01 -3.13916117e-01 8.87428463e-01
-1.42536771e+00 -1.72047600e-01 6.97480738e-01 7.18593657e-01
-1.43995211e-01 -4.69792604e-01 -2.05027699e-01 4.48948711e-01
-1.18212193e-01 -7.15827644e-01 -9.62887108e-01 7.74556518e-01
1.42851412e-01 1.03294122e+00 -3.08262389e-02 -8.99427608e-02
1.76591426e-01 2.66491890e-01 6.56618118e-01 -7.54737735e-01
-4.81678814e-01 -8.71079206e-01 5.78668952e-01 -3.10915411e-01
-2.25633711e-01 -7.89086461e-01 -1.55098653e+00 -6.02040052e-01
1.28960639e-01 8.51703063e-02 3.47783267e-01 1.07803357e+00
1.61334828e-01 6.87535465e-01 -1.03812322e-01 -3.69033188e-01
-6.98615015e-01 -1.13293707e+00 -1.90431982e-01 5.66295147e-01
7.86420330e-02 -3.12255740e-01 -3.37494045e-01 4.68699545e-01] | [11.199752807617188, 8.743819236755371] |
444ff8af-fdb4-4356-97cb-701e30ed78b2 | generative-one-class-models-for-text-based | 1611.05915 | null | http://arxiv.org/abs/1611.05915v1 | http://arxiv.org/pdf/1611.05915v1.pdf | Generative One-Class Models for Text-based Person Retrieval in Forensic Applications | Automatic forensic image analysis assists criminal investigation experts in
the search for suspicious persons, abnormal behaviors detection and identity
matching in images. In this paper we propose a person retrieval system that
uses textual queries (e.g., "black trousers and green shirt") as descriptions
and a one-class generative color model with outlier filtering to represent the
images both to train the models and to perform the search. The method is
evaluated in terms of its efficiency in fulfilling the needs of a forensic
retrieval system: limited annotation, robustness, extensibility, adaptability
and computational cost. The proposed generative method is compared to a
corresponding discriminative approach. Experiments are carried out using a
range of queries in three different databases. The experiments show that the
two evaluated algorithms provide average retrieval performance and adaptable to
new datasets. The proposed generative algorithm has some advantages over the
discriminative one, specifically its capability to work with very few training
samples and its much lower computational requirements when the number of
training examples increases. | ['Hedvig Kjellström', 'David Gerónimo'] | 2016-11-17 | null | null | null | null | ['person-retrieval', 'nlp-based-person-retrival'] | ['computer-vision', 'computer-vision'] | [ 1.94718644e-01 -5.99736571e-01 2.94876456e-01 -4.50386226e-01
-6.68236494e-01 -6.32323086e-01 7.58973420e-01 1.51395068e-01
-6.57956839e-01 5.87907672e-01 -3.63629490e-01 -9.75464508e-02
-5.22122025e-01 -5.86033285e-01 -6.44270480e-02 -6.12331033e-01
1.32959455e-01 1.05242908e+00 4.32313472e-01 4.35332535e-03
6.36513889e-01 1.05052173e+00 -1.54275131e+00 4.40950431e-02
7.23576725e-01 7.15597510e-01 5.76445833e-02 5.84468544e-01
3.13770548e-02 5.85720599e-01 -8.52864504e-01 -7.58751750e-01
3.55930179e-01 -1.40161410e-01 -5.74002266e-01 3.24437529e-01
3.01870912e-01 -3.31747144e-01 -3.93712580e-01 1.06860197e+00
7.13231087e-01 2.48149797e-01 8.88669789e-01 -1.10277843e+00
-7.45575726e-01 -1.91749707e-01 -5.87414265e-01 7.24898934e-01
6.23167992e-01 9.58055854e-02 3.95365089e-01 -6.76401794e-01
7.60168850e-01 1.29720926e+00 6.42548621e-01 4.98166084e-01
-1.12221038e+00 -4.84858274e-01 -3.21948767e-01 5.61847925e-01
-1.66635466e+00 -4.06485766e-01 7.82865047e-01 -5.29595017e-01
7.85668910e-01 3.68917674e-01 3.18656981e-01 8.27195764e-01
-3.23265605e-02 5.06203294e-01 1.17045391e+00 -4.83299106e-01
1.94842294e-01 6.06011152e-01 4.53128815e-01 7.30156958e-01
6.35137856e-01 -2.87872165e-01 -3.47977906e-01 -7.78024793e-01
4.18899149e-01 2.12816641e-01 2.59173848e-02 -1.00282850e-02
-3.97742391e-01 7.01852024e-01 -1.67976007e-01 6.13871634e-01
-4.31378514e-01 -7.65975118e-02 5.26590347e-01 2.32674092e-01
3.28114212e-01 1.50710478e-01 3.80367100e-01 -1.71687737e-01
-1.25467300e+00 2.66814947e-01 6.07836843e-01 7.67603219e-01
5.00582039e-01 1.44939467e-01 3.96063849e-02 8.71852517e-01
2.47021377e-01 7.12772548e-01 6.04295313e-01 -2.43262172e-01
4.14682955e-01 7.17796504e-01 1.08836673e-01 -1.27896988e+00
-1.48898229e-01 -2.03360189e-02 -2.14573175e-01 1.09626651e-01
2.74669468e-01 3.18605006e-01 -9.09503043e-01 1.17718291e+00
1.38639882e-01 -8.34306628e-02 8.12425162e-04 5.71318269e-01
7.26540148e-01 4.41191614e-01 1.85735688e-01 -1.85220048e-01
1.34222662e+00 -1.71376318e-01 -6.38878465e-01 -8.31028167e-03
9.00991112e-02 -1.13728666e+00 6.92763686e-01 4.20876563e-01
-8.76933873e-01 -5.91316760e-01 -7.60192275e-01 2.69996643e-01
-5.65103829e-01 5.20926952e-01 3.90350074e-01 1.17831218e+00
-8.14302087e-01 3.99298072e-01 -5.87116241e-01 -7.53714025e-01
2.27746561e-01 4.91398007e-01 -3.72649074e-01 -7.92663544e-02
-6.93006575e-01 7.39549279e-01 6.89420283e-01 7.49090686e-02
-6.91210866e-01 5.90885207e-02 -5.65681100e-01 -7.96703026e-02
6.36609495e-02 -2.16915369e-01 5.45345604e-01 -8.10602486e-01
-8.43373775e-01 1.30351138e+00 3.38093229e-02 -2.73504496e-01
7.86986232e-01 -1.93442330e-01 -7.44330347e-01 6.74191236e-01
3.57680500e-01 9.49353874e-02 8.96602571e-01 -1.11715615e+00
-2.37923056e-01 -8.38242173e-01 -4.09915924e-01 -1.45873323e-01
-3.06591660e-01 5.49851477e-01 -5.96452177e-01 -6.97090149e-01
1.04255222e-01 -9.32402492e-01 3.07698935e-01 -3.47323090e-01
-3.19725066e-01 -2.14586198e-01 1.29472733e+00 -9.06867623e-01
1.07898521e+00 -2.17374849e+00 -3.75516802e-01 8.94881725e-01
-2.31841460e-01 5.56059062e-01 2.61448473e-01 7.04370975e-01
6.92790328e-03 -1.99746370e-01 -1.57105461e-01 -3.29907119e-01
-6.32969514e-02 2.26469681e-01 5.51025793e-02 7.07554221e-01
7.41970018e-02 4.55443799e-01 -6.81235909e-01 -1.05116761e+00
1.89212456e-01 4.92040843e-01 -9.10969749e-02 1.91796854e-01
3.70636940e-01 1.69638067e-01 -5.89812934e-01 1.03160870e+00
6.96492970e-01 2.17969060e-01 2.16702059e-01 -7.71014094e-02
2.58198351e-01 -4.87802148e-01 -1.47645104e+00 9.64002371e-01
1.10231921e-01 6.25106335e-01 -1.94245771e-01 -9.59835172e-01
1.32723045e+00 3.21915299e-01 3.10663283e-01 -6.90863729e-01
1.71924636e-01 4.65533018e-01 -1.76846206e-01 -1.07575464e+00
5.99616230e-01 -8.75326768e-02 1.19885303e-01 7.14796424e-01
5.55758961e-02 4.40638930e-01 7.46037006e-01 1.99001461e-01
8.07848990e-01 1.57433212e-01 1.20871603e-01 -4.18762118e-02
8.09018075e-01 -1.28754377e-01 3.56055528e-01 7.21999526e-01
-7.28394883e-03 3.20495039e-01 2.22100735e-01 -4.99621063e-01
-1.07764268e+00 -9.66919005e-01 -1.16118573e-01 6.94086552e-01
1.06025361e-01 6.04133978e-02 -7.44720459e-01 -5.47162056e-01
-4.11209874e-02 5.24289012e-01 -4.29397345e-01 -1.46028260e-02
-7.22142398e-01 -9.27246034e-01 8.16664934e-01 2.10599616e-01
6.47240400e-01 -1.01017964e+00 -9.54697132e-01 -4.07966860e-02
4.69686538e-02 -8.83777320e-01 -8.60222429e-02 -3.91001731e-01
-7.81932771e-01 -1.36685455e+00 -6.73212826e-01 -5.77433705e-01
9.54837859e-01 2.43172929e-01 7.42928088e-01 3.21257144e-01
-9.12370801e-01 8.83856416e-01 -4.92345512e-01 -2.28601262e-01
-3.71806681e-01 -2.72535861e-01 -4.45464551e-02 3.59789610e-01
7.95752943e-01 -4.58668858e-01 -5.00026703e-01 1.71718404e-01
-1.34935379e+00 -9.17159319e-01 5.39271057e-01 8.25390816e-01
3.03083509e-01 9.21436250e-02 4.26212668e-01 -1.10919833e+00
1.01041448e+00 -4.04900074e-01 -5.06655037e-01 6.34098768e-01
-6.98836744e-01 -8.77193213e-02 3.60697776e-01 -3.82805794e-01
-1.33498895e+00 -4.63569053e-02 1.31634042e-01 -2.97951639e-01
-3.17101002e-01 -4.25762460e-02 -1.30947292e-01 -3.53945255e-01
7.54406154e-01 5.72657585e-01 -8.52036104e-02 -7.50231326e-01
-1.11250326e-01 8.06603253e-01 7.04480350e-01 -7.19485521e-01
9.02090847e-01 4.17706698e-01 1.09611012e-01 -9.00652587e-01
7.32008815e-02 -9.48583901e-01 -7.39353776e-01 -4.74913478e-01
6.40265465e-01 -2.83034205e-01 -6.53409243e-01 3.45218569e-01
-1.02673948e+00 7.31978774e-01 1.29182041e-01 4.01305765e-01
-3.58141571e-01 1.11175966e+00 -4.51470971e-01 -1.38323784e+00
-6.35002315e-01 -9.56387758e-01 9.02330101e-01 2.74416447e-01
-4.29256372e-02 -8.53396654e-01 6.55953586e-03 5.26262999e-01
1.32306322e-01 3.40700209e-01 8.53005767e-01 -1.15889335e+00
-3.27326685e-01 -9.35624957e-01 -2.56063879e-01 1.99844509e-01
-7.12567195e-02 1.73960663e-02 -1.00277197e+00 -3.88019741e-01
-2.54939273e-02 -2.14901626e-01 7.14165449e-01 -1.88205600e-01
7.79420316e-01 -1.70055583e-01 -3.20528507e-01 3.80137205e-01
1.68967235e+00 8.14236164e-01 8.16637039e-01 5.57279170e-01
3.27529132e-01 6.25576079e-01 6.99557006e-01 5.74159086e-01
-3.85377258e-01 6.37379289e-01 2.16653273e-01 1.10162906e-01
1.63550943e-01 8.85080695e-02 7.63714910e-02 7.04105198e-02
-3.06372911e-01 -2.67827511e-01 -9.27753806e-01 5.33830762e-01
-1.81228805e+00 -1.54908180e+00 -7.90520459e-02 2.40180659e+00
1.47930413e-01 1.54373888e-02 4.91913110e-01 4.35134500e-01
1.04868960e+00 -2.58821547e-01 -2.03245282e-01 -4.25925732e-01
-8.92212018e-02 3.85539263e-01 3.18646461e-01 2.12372169e-01
-1.03852558e+00 5.73188186e-01 5.86640453e+00 9.31305289e-01
-9.99297619e-01 1.92636207e-01 5.68707645e-01 1.43011771e-02
9.73015726e-02 -1.22669011e-01 -6.43761814e-01 7.50753522e-01
6.87344432e-01 8.26238319e-02 9.48096067e-02 7.19060123e-01
7.34374374e-02 -3.61203521e-01 -7.31635988e-01 1.28512383e+00
6.71896338e-01 -8.43217731e-01 2.75415391e-01 6.82103783e-02
3.02278340e-01 -5.78707278e-01 1.92401946e-01 -2.49500081e-01
-4.64703619e-01 -8.16206217e-01 6.42433643e-01 9.68890727e-01
4.47859257e-01 -1.00511384e+00 1.01878011e+00 1.42443582e-01
-7.75060773e-01 -8.97021890e-02 -5.32245278e-01 5.57347536e-01
5.36898635e-02 2.31897179e-03 -1.05703807e+00 6.29295886e-01
5.14205873e-01 -5.25082387e-02 -9.26328778e-01 1.33552122e+00
2.53801107e-01 3.79492760e-01 -3.72427136e-01 -1.03315704e-01
2.98752844e-01 -4.78089958e-01 5.96623182e-01 1.75922275e+00
5.20876944e-01 1.12130018e-02 -3.48255015e-03 7.06748128e-01
5.06139636e-01 4.70982701e-01 -7.07451999e-01 1.95176825e-02
4.54125017e-01 1.01004744e+00 -1.18172836e+00 -5.15639782e-01
-4.98488061e-02 9.99923885e-01 -2.26565257e-01 2.79585451e-01
-5.72768033e-01 -7.22439647e-01 -1.60018340e-01 3.63615215e-01
1.55578449e-01 -4.91791740e-02 1.59844965e-01 -8.75920594e-01
2.28348777e-01 -7.64165640e-01 7.87499487e-01 -6.36192501e-01
-1.17132139e+00 8.38205576e-01 5.01340628e-01 -1.25420606e+00
-5.38562834e-01 -8.04781497e-01 -5.40288627e-01 8.45045686e-01
-8.78385007e-01 -1.37387836e+00 -2.46590465e-01 9.42636967e-01
3.08210611e-01 -8.64725649e-01 6.23698592e-01 6.22798741e-01
-5.18720686e-01 5.12639105e-01 2.39946961e-01 2.52032101e-01
7.21569121e-01 -1.00603318e+00 -3.11050117e-01 1.19386911e+00
1.54536098e-01 1.05092597e+00 7.84983218e-01 -8.69507015e-01
-1.14078987e+00 -5.74933529e-01 8.94084334e-01 -1.44600600e-01
2.87653565e-01 5.76323308e-02 -8.62226069e-01 2.20174238e-01
-4.49184664e-02 -1.74138114e-01 9.90846276e-01 -2.41757557e-01
-3.26934844e-01 -2.33188093e-01 -1.61294413e+00 1.10062473e-01
4.26243633e-01 -5.95233858e-01 -6.75088644e-01 3.46408993e-01
-5.26298165e-01 6.95760548e-02 -5.32219827e-01 -1.75583273e-01
6.03787184e-01 -1.12202549e+00 9.77807224e-01 -4.96705562e-01
-3.17390323e-01 -3.83215547e-01 3.94661017e-02 -3.32759112e-01
-1.37205437e-01 -3.50702137e-01 4.05482292e-01 1.33936715e+00
1.06578767e-01 -6.12029552e-01 7.61524677e-01 7.67463982e-01
2.94531256e-01 -2.61335701e-01 -9.89871860e-01 -8.68507326e-01
-7.87412882e-01 -2.07534865e-01 3.75237197e-01 6.71963334e-01
-2.57507622e-01 -4.72570285e-02 -5.89973092e-01 -1.03127714e-02
9.10593987e-01 1.19520701e-01 6.71780825e-01 -1.33844793e+00
-5.57499826e-01 -1.80595666e-01 -1.14245832e+00 -5.63583970e-02
-1.27242312e-01 -6.18875206e-01 -6.08309865e-01 -1.06097734e+00
6.29103363e-01 -3.82149428e-01 -1.21694505e-01 2.48829857e-01
-1.29879072e-01 6.22484088e-01 2.73686945e-01 5.68351805e-01
-5.69632530e-01 -1.48639709e-01 4.64487702e-01 -1.27169192e-01
-6.44011423e-02 2.26271540e-01 -1.91038668e-01 6.87671065e-01
6.96757615e-01 -4.96301442e-01 -4.81157154e-01 2.07972527e-03
4.68419306e-03 -1.49698958e-01 6.41057551e-01 -1.00645041e+00
2.08092317e-01 1.36807799e-01 7.15270042e-01 -6.77476346e-01
5.92913389e-01 -8.12436759e-01 5.41006148e-01 5.58062792e-01
1.35715371e-02 5.92945397e-01 8.55343533e-04 6.50209069e-01
-2.33428419e-01 -9.87756073e-01 7.56497622e-01 -3.11153084e-01
-9.12185907e-01 -3.68788984e-04 -2.37932950e-01 -3.19558859e-01
1.22742701e+00 -8.49330068e-01 -2.55523592e-01 -3.56827259e-01
-1.00111032e+00 -3.81439298e-01 4.71009076e-01 -5.50637059e-02
8.23257685e-01 -1.14326215e+00 -5.38960218e-01 4.82625663e-02
2.90640742e-01 -9.74417269e-01 2.68729210e-01 6.93243563e-01
-1.06756055e+00 3.52706373e-01 -5.89387238e-01 -3.88296574e-01
-1.87879682e+00 6.45247459e-01 3.00418455e-02 -3.02898765e-01
-3.51863980e-01 4.09869045e-01 -4.43238497e-01 2.28061482e-01
1.41874388e-01 5.10338545e-01 -3.99162829e-01 1.09260269e-01
4.78849888e-01 7.46880889e-01 4.55327407e-02 -1.00770581e+00
-3.58972251e-01 7.04329073e-01 -1.58122316e-01 -1.14907704e-01
1.30558622e+00 2.32150793e-01 -2.38292739e-01 1.32828876e-01
9.28587258e-01 1.19882047e-01 -5.24510562e-01 -1.32881626e-01
2.23052144e-01 -9.58117247e-01 -4.07439142e-01 -5.12574375e-01
-8.83400321e-01 8.59645486e-01 1.13168693e+00 3.89626354e-01
1.17989516e+00 -1.96859807e-01 4.15116668e-01 6.39617205e-01
3.63564849e-01 -1.38587046e+00 5.91936335e-02 -1.52150556e-01
5.75062156e-01 -9.63343441e-01 2.78820127e-01 -1.83365017e-01
-5.96671820e-01 1.49720979e+00 3.48419487e-01 -2.23260164e-01
1.96763158e-01 -2.24718690e-01 -9.04961228e-02 -4.74193871e-01
-1.78727239e-01 -3.50284696e-01 3.67013097e-01 8.68547320e-01
4.43406761e-01 -1.57113492e-01 -8.58383954e-01 -4.96189035e-02
1.91034362e-01 -2.69201636e-01 1.22639321e-01 1.05141366e+00
-4.82938826e-01 -1.39749265e+00 -9.21563148e-01 4.62111354e-01
-9.58983421e-01 2.36094221e-01 -6.67101800e-01 1.17127585e+00
4.94295061e-01 8.89772713e-01 -2.57204890e-01 -2.71096136e-02
2.99066633e-01 2.02139765e-01 5.72331846e-01 -3.27865303e-01
-6.99796259e-01 2.60887444e-01 1.22855783e-01 -2.56080598e-01
-5.63039243e-01 -1.07195508e+00 -6.20561302e-01 -1.53720662e-01
-2.25627810e-01 5.09356439e-01 6.78727448e-01 8.01909089e-01
1.35004101e-02 -1.45911202e-01 4.04396355e-01 -2.76060730e-01
-5.94384611e-01 -9.99564528e-01 -7.02301979e-01 8.53467226e-01
-2.57135808e-01 -7.17784047e-01 9.66059193e-02 3.09742361e-01] | [12.395585060119629, 0.9116864204406738] |
fcc4c7ae-a11a-42cf-a9ef-103bc74a9d5a | mac-a-novel-stochastic-optimization-method | 2304.12248 | null | https://arxiv.org/abs/2304.12248v1 | https://arxiv.org/pdf/2304.12248v1.pdf | MAC, a novel stochastic optimization method | A novel stochastic optimization method called MAC was suggested. The method is based on the calculation of the objective function at several random points and then an empirical expected value and an empirical covariance matrix are calculated. The empirical expected value is proven to converge to the optimum value of the problem. The MAC algorithm was encoded in Matlab and the code was tested on 20 test problems. Its performance was compared with those of the interior point method (Matlab name: fmincon), simplex, pattern search (PS), simulated annealing (SA), particle swarm optimization (PSO), and genetic algorithm (GA) methods. The MAC method failed two test functions and provided inaccurate results on four other test functions. However, it provided accurate results and required much less CPU time than the widely used optimization methods on the other 14 test functions. | ['János Tóth', 'Tamás Turányi', 'Goitom Simret Kidane', 'Attila László Nagy'] | 2023-04-14 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [ 3.59189250e-02 -3.76900196e-01 9.05063227e-02 8.63584206e-02
-3.05205673e-01 -4.75280106e-01 2.06914589e-01 1.50362924e-01
-6.32508039e-01 1.54214776e+00 -6.00755930e-01 -4.99802202e-01
-7.60354400e-01 -7.54509389e-01 -1.18432939e-01 -1.21719813e+00
-3.37119997e-01 5.60395837e-01 1.31926134e-01 -2.04895392e-01
9.04117465e-01 4.39856052e-01 -1.62907600e+00 -5.40195704e-01
1.30537808e+00 1.07685971e+00 2.54695743e-01 6.94303274e-01
-4.87801358e-02 2.30281621e-01 -1.09548259e+00 -7.33178481e-02
2.71791458e-01 -6.58496737e-01 -4.18101728e-01 5.02840690e-02
-8.83041382e-01 3.71889830e-01 3.97782117e-01 1.11912203e+00
5.70656836e-01 6.28581822e-01 5.28768897e-01 -1.45789051e+00
-3.36989075e-01 1.92372382e-01 -7.48552382e-01 1.84562102e-01
5.57394981e-01 2.18153726e-02 4.20891464e-01 -8.25489044e-01
2.50844240e-01 9.07699108e-01 8.00035775e-01 -5.87359108e-02
-9.96327162e-01 -5.65592587e-01 -5.06304026e-01 1.85509220e-01
-1.66784322e+00 2.07504228e-01 3.43828827e-01 2.46769190e-01
8.67715240e-01 5.79901099e-01 1.01886463e+00 1.44570038e-01
6.52445674e-01 2.86893815e-01 1.32952416e+00 -6.15460038e-01
8.04821014e-01 1.04055673e-01 -1.40928388e-01 6.41555667e-01
8.16027105e-01 3.79577696e-01 7.34039173e-02 -5.30563176e-01
3.77071828e-01 -3.39131862e-01 -4.05946106e-01 -1.15207590e-01
-1.17146373e+00 9.38475847e-01 5.77839017e-02 3.97716969e-01
-7.27678180e-01 -1.82550192e-01 -5.73157631e-02 7.97936022e-02
4.76474315e-02 7.94317067e-01 -4.48863268e-01 -4.71771717e-01
-7.61854589e-01 3.62090886e-01 1.29273438e+00 2.93629140e-01
3.60292137e-01 4.42165732e-01 3.62672657e-02 6.78993464e-01
5.63146889e-01 9.67427731e-01 4.92394894e-01 -9.27708924e-01
4.36097533e-01 6.04111254e-01 6.89126909e-01 -1.25883234e+00
-6.10094965e-01 -6.41160786e-01 -5.22058487e-01 6.29913867e-01
6.59135997e-01 -6.89335883e-01 -5.46194911e-01 9.83725429e-01
3.75738949e-01 -2.57002860e-02 2.71142304e-01 7.24369287e-01
2.53489494e-01 1.09856343e+00 -5.51368475e-01 -7.40960419e-01
7.18007147e-01 -9.04005945e-01 -9.92330074e-01 2.14728847e-01
1.94938570e-01 -1.07152843e+00 5.94531536e-01 8.24498951e-01
-1.17424154e+00 -2.47214928e-01 -1.49005127e+00 1.13034999e+00
-3.91334534e-01 1.26155568e-02 5.10255039e-01 1.38297737e+00
-9.85524237e-01 6.64973736e-01 -6.14465594e-01 -1.14324816e-01
-9.55063030e-02 7.66324461e-01 2.11574346e-01 2.14163467e-01
-6.63577378e-01 9.37907755e-01 5.96666813e-01 4.33361262e-01
-1.59041926e-01 -1.88240156e-01 -7.75899529e-01 -1.78371056e-03
2.48487160e-01 -2.85993457e-01 5.41522324e-01 -7.88789332e-01
-2.14581203e+00 9.22188684e-02 -2.14169964e-01 -1.77017808e-01
2.76558608e-01 4.80141252e-01 -5.01747608e-01 -3.66283990e-02
-6.83240369e-02 2.39336081e-02 6.70369208e-01 -1.48256958e+00
-7.30199695e-01 1.42554805e-01 -4.30359721e-01 2.21706733e-01
8.25907141e-02 1.69630945e-01 4.25747745e-02 -4.87830102e-01
2.12634876e-01 -8.36148739e-01 -4.66332465e-01 -6.39792383e-01
-3.46005261e-01 1.89766530e-02 4.09618229e-01 -6.07604742e-01
1.59884846e+00 -1.81833422e+00 7.23800808e-02 1.12281787e+00
-4.48321342e-01 1.92929983e-01 1.21273316e-01 6.43241584e-01
-1.67432800e-01 1.32963866e-01 -4.00138706e-01 2.20010191e-01
3.07408944e-02 2.21726760e-01 5.06781280e-01 7.82074094e-01
-4.48911399e-01 3.80680740e-01 -9.41236556e-01 -3.95566255e-01
2.82295793e-01 1.67394485e-02 -4.64866489e-01 -7.30177909e-02
2.49905482e-01 3.63877118e-02 -3.40219259e-01 9.73837376e-01
8.87731910e-01 -1.77655414e-01 2.72978228e-02 3.52764651e-02
-4.27567035e-01 -4.36259329e-01 -1.90367723e+00 6.93886876e-01
-2.02192664e-01 4.08313304e-01 2.10056975e-01 -9.17096138e-01
1.12088144e+00 3.94866109e-01 6.86320662e-01 -4.53294605e-01
6.08605981e-01 5.09990454e-01 3.69954526e-01 -4.56596613e-01
4.45574909e-01 2.46537596e-01 1.90311760e-01 7.55743146e-01
-5.27464807e-01 -3.51010442e-01 5.75947225e-01 -4.00153428e-01
8.13770831e-01 -3.05826604e-01 5.29493570e-01 -4.57986116e-01
9.46595252e-01 3.57746869e-01 4.82841223e-01 6.51023746e-01
-2.50569701e-01 2.29070440e-01 2.44018421e-01 -3.63751978e-01
-8.36532891e-01 -9.10560012e-01 -2.82428592e-01 2.31288582e-01
6.45905554e-01 -1.54656470e-01 -6.12224758e-01 -2.15552196e-01
2.48511136e-02 1.12767911e+00 -3.85781080e-01 -7.98751786e-03
-3.77911329e-01 -1.42222166e+00 1.25041995e-02 -2.46241510e-01
7.50208735e-01 -1.00737584e+00 -5.72789907e-01 5.14574051e-01
2.65517663e-02 -6.86462402e-01 -9.02520642e-02 1.25401039e-02
-6.67677701e-01 -1.14602065e+00 -5.46578526e-01 -5.16170621e-01
9.23327267e-01 2.59759724e-02 6.60651386e-01 5.26842415e-01
-2.93931395e-01 1.73260331e-01 -4.40860778e-01 -5.34344971e-01
-3.64066333e-01 -4.22226608e-01 2.10944027e-01 3.11381798e-02
-2.19110996e-02 -2.81471550e-01 -2.64302343e-01 7.50919044e-01
-5.48549235e-01 -7.24396110e-01 2.46417001e-01 1.23409975e+00
6.71800196e-01 1.22316957e+00 3.54646653e-01 -4.98346612e-02
1.16205168e+00 -4.62986261e-01 -1.12090755e+00 2.71267742e-01
-8.13303530e-01 -5.60899219e-03 6.60345435e-01 -3.00933570e-01
-8.81507277e-01 -1.69927299e-01 9.09224004e-02 5.40280007e-02
3.72786224e-01 5.16838491e-01 1.24804713e-01 -7.75870562e-01
5.46529293e-01 3.42983454e-01 4.56894606e-01 -2.54179418e-01
-5.10087192e-01 6.48886085e-01 2.31025457e-01 -4.11178410e-01
8.94435704e-01 2.01321736e-01 3.05437654e-01 -6.45314693e-01
-2.38756955e-01 -2.92115510e-02 3.40809091e-03 -3.77897322e-01
6.22328341e-01 7.14699477e-02 -1.17062974e+00 5.86717844e-01
-8.16977918e-01 1.01338983e-01 9.86765511e-03 7.51547873e-01
-3.45401108e-01 3.11093748e-01 1.07338272e-01 -1.23962271e+00
-1.59557849e-01 -1.29689252e+00 1.59111947e-01 7.11107194e-01
-8.20297673e-02 -1.04668701e+00 -2.22953290e-01 1.22994162e-01
4.92041767e-01 6.81624174e-01 5.86745799e-01 -4.20186698e-01
-1.04263037e-01 -5.67700684e-01 1.33558661e-01 1.49211675e-01
2.66558617e-01 6.08973980e-01 -3.00592128e-02 -4.10627991e-01
4.60678875e-01 3.28281581e-01 -3.28628898e-01 7.34615564e-01
9.15063262e-01 -3.17991167e-01 -4.42725390e-01 6.68549061e-01
1.92835844e+00 1.14506578e+00 6.33610845e-01 9.40701962e-01
-2.07257733e-01 2.89889395e-01 9.74777102e-01 8.11557651e-01
1.86920688e-02 5.26440799e-01 5.56095302e-01 1.23899966e-01
6.81327581e-01 4.47695285e-01 1.30094960e-01 4.77879643e-01
-3.77900988e-01 -5.42556345e-01 -1.02916968e+00 1.71430439e-01
-1.64083552e+00 -9.14217472e-01 -2.39358634e-01 2.23716593e+00
3.99586707e-01 1.56997159e-01 1.32405624e-01 6.93334520e-01
9.92909551e-01 -4.54019725e-01 -3.76560569e-01 -1.09468377e+00
-1.03790238e-01 2.73355186e-01 9.90914285e-01 4.34577078e-01
-6.76766396e-01 1.91934824e-01 7.21336985e+00 9.25635159e-01
-8.56751502e-01 -2.28469759e-01 3.76779258e-01 3.41548361e-02
1.54738976e-02 -1.23474054e-01 -6.38764262e-01 9.72387433e-01
7.79215932e-01 -6.84631765e-01 9.73813236e-01 4.58368748e-01
4.10119265e-01 -8.85062277e-01 -2.59778470e-01 1.05072057e+00
-1.60123512e-01 -1.28684568e+00 -6.02185309e-01 1.55586541e-01
1.13119638e+00 -4.06969756e-01 -7.17802942e-02 -9.36328918e-02
2.31159136e-01 -1.24389124e+00 2.12913901e-01 4.98590738e-01
1.46639654e-02 -1.36193645e+00 1.37665200e+00 3.01848590e-01
-9.50484693e-01 -5.68482816e-01 -2.35223070e-01 -1.14427626e-01
3.01766276e-01 3.83028656e-01 -6.25860870e-01 7.21766651e-01
7.50730574e-01 -7.94430524e-02 -2.43317783e-01 1.91357160e+00
6.89925775e-02 5.40147185e-01 -8.64873528e-01 -1.01452041e+00
5.14114678e-01 -8.51090133e-01 1.13059402e+00 3.83594483e-01
5.69285333e-01 1.90904215e-02 -1.67324245e-01 5.97478747e-01
9.34579909e-01 3.14975530e-01 1.27355471e-01 -3.44055071e-02
7.74449646e-01 8.59747708e-01 -1.09449911e+00 1.13903739e-01
-6.92889392e-02 4.70602036e-01 -5.09574831e-01 5.00295579e-01
-1.10631478e+00 -1.05624163e+00 3.12499762e-01 -4.26683545e-01
4.24702048e-01 -3.24868023e-01 -7.24542320e-01 -3.45786870e-01
-1.04096323e-01 -8.63911748e-01 3.72180074e-01 -7.25186646e-01
-9.27726805e-01 5.91081202e-01 2.11001441e-01 -1.32150543e+00
-3.79320711e-01 -6.93766654e-01 -7.99299777e-01 9.99684751e-01
-8.58541131e-01 9.93535742e-02 -1.62733227e-01 4.40407485e-01
1.49063021e-01 -7.10490584e-01 6.73033893e-01 1.15308009e-01
-8.26348484e-01 6.08120084e-01 8.78090978e-01 -6.19087934e-01
-8.96341875e-02 -1.08214462e+00 -4.42538857e-01 7.71063030e-01
-6.82329476e-01 3.78981262e-01 1.15712404e+00 -4.55871105e-01
-1.57404864e+00 -1.51144609e-01 6.85735226e-01 1.68262780e-01
7.00065315e-01 3.59080374e-01 -3.08513254e-01 -1.59362465e-01
3.26976418e-01 -3.37004751e-01 5.31766236e-01 -4.58044708e-01
1.02182889e+00 2.12861560e-02 -1.80426753e+00 4.64377791e-01
2.70964593e-01 6.35836422e-01 -3.06257755e-01 3.83884847e-01
1.55144811e-01 -6.48911178e-01 -9.40861344e-01 3.09087187e-01
3.94662440e-01 -9.11164880e-01 1.05986273e+00 2.00081661e-01
-1.55373499e-01 -6.83355272e-01 -7.30489418e-02 -1.40931404e+00
-1.28108159e-01 -8.59529436e-01 1.40387103e-01 8.87688458e-01
7.12370872e-01 -1.47914279e+00 7.22932756e-01 5.12355924e-01
1.81399643e-01 -1.21860254e+00 -1.09644485e+00 -1.09007847e+00
-4.51931894e-01 -1.56418942e-02 9.38004017e-01 6.35639787e-01
-7.29679614e-02 -2.97684073e-01 -7.11615607e-02 2.48971075e-01
9.19175029e-01 1.72623068e-01 4.13889617e-01 -9.02038038e-01
-2.90166855e-01 -4.40656751e-01 -4.04583097e-01 -1.90441161e-01
-1.89089999e-01 -3.20583910e-01 -2.46168569e-01 -1.70749378e+00
-4.88048583e-01 -6.18066192e-01 -1.54902562e-01 1.25106975e-01
-1.15534097e-01 7.76530281e-02 -8.08486715e-02 -6.70015067e-02
-6.65645748e-02 4.59623486e-01 1.32363343e+00 2.83797354e-01
-5.51196277e-01 5.92301369e-01 -4.01361078e-01 5.38169265e-01
1.13708615e+00 -6.77049458e-01 -2.70776719e-01 2.14915425e-01
3.00395578e-01 1.65989459e-01 -1.76179662e-01 -1.34990478e+00
3.69450182e-01 -4.30904329e-01 5.13881266e-01 -9.19355333e-01
2.87945002e-01 -1.26926386e+00 5.18655479e-01 9.12138939e-01
4.81511354e-01 5.39641500e-01 2.29201570e-01 2.52026260e-01
-2.70773739e-01 -8.23141754e-01 7.19717920e-01 2.60061681e-01
-4.42027569e-01 -3.24545383e-01 -6.65434837e-01 -2.86712736e-01
1.56018949e+00 -9.82356191e-01 -3.92123349e-02 -4.56679195e-01
-6.80764794e-01 5.17402709e-01 4.90635455e-01 -2.28674918e-01
6.24932230e-01 -1.19093895e+00 -5.30716062e-01 1.74930274e-01
-6.55250847e-01 -5.27399659e-01 -2.73649871e-01 1.02349854e+00
-1.30779350e+00 2.26202294e-01 -1.84739813e-01 -4.90261197e-01
-1.23454106e+00 3.12736332e-01 5.27175605e-01 -1.91633701e-01
5.92017546e-02 7.41333246e-01 -1.10573208e+00 -4.40894753e-01
2.26299003e-01 -9.46386624e-03 -3.21501762e-01 -1.89422145e-01
3.85926515e-01 1.15438676e+00 -1.69462934e-01 -4.17200297e-01
-5.33411145e-01 8.37393820e-01 7.06943333e-01 -3.01434278e-01
1.32401991e+00 -2.14455351e-01 -5.63470244e-01 -1.69202670e-01
1.22171211e+00 1.84171394e-01 -6.36735678e-01 4.81786430e-01
-1.22309044e-01 -7.93703318e-01 8.35893899e-02 -1.00711107e+00
-9.00906205e-01 -2.05655694e-01 7.76161492e-01 6.71122611e-01
1.44234180e+00 -9.11954284e-01 5.97977817e-01 5.11015356e-01
5.90946913e-01 -1.43483281e+00 -4.21768248e-01 5.84524155e-01
6.61695659e-01 -1.00841022e+00 4.50896472e-01 -3.52291971e-01
-5.92056394e-01 1.43494427e+00 4.05378640e-01 -2.32521057e-01
1.03018057e+00 6.06905043e-01 -2.53466666e-01 4.10687551e-02
-5.00513315e-01 1.79208294e-01 6.56700283e-02 4.66585577e-01
-1.01491757e-01 -1.38483316e-01 -1.08705819e+00 3.15126508e-01
-3.42788637e-01 -1.09598888e-02 4.79834080e-01 1.22012937e+00
-4.78599966e-01 -9.64606941e-01 -1.19063771e+00 3.25044066e-01
-4.89067823e-01 4.64090854e-02 3.83277945e-02 1.04173172e+00
2.87835523e-02 1.38402712e+00 -4.91347164e-02 -2.62934953e-01
6.83923960e-02 -4.48217124e-01 4.17947471e-01 9.31661054e-02
-7.92595863e-01 -2.71799386e-01 2.54651755e-01 -5.91948926e-01
-2.29618996e-01 -7.37328708e-01 -1.38492048e+00 -5.54433465e-01
-5.02960026e-01 9.32472944e-01 1.07941377e+00 9.30292130e-01
-1.03449367e-01 4.01024640e-01 1.06123841e+00 -8.75432074e-01
-4.58117664e-01 -4.00708199e-01 -5.92295408e-01 -4.54359800e-01
-1.57378137e-01 -8.81633043e-01 -9.02812064e-01 -5.69050014e-01] | [5.708166599273682, 3.4743728637695312] |
31148a15-3b4f-45d9-b8cd-56adf48d5b9f | learning-disentangled-label-representations | 2212.01461 | null | https://arxiv.org/abs/2212.01461v1 | https://arxiv.org/pdf/2212.01461v1.pdf | Learning Disentangled Label Representations for Multi-label Classification | Although various methods have been proposed for multi-label classification, most approaches still follow the feature learning mechanism of the single-label (multi-class) classification, namely, learning a shared image feature to classify multiple labels. However, we find this One-shared-Feature-for-Multiple-Labels (OFML) mechanism is not conducive to learning discriminative label features and makes the model non-robustness. For the first time, we mathematically prove that the inferiority of the OFML mechanism is that the optimal learned image feature cannot maintain high similarities with multiple classifiers simultaneously in the context of minimizing cross-entropy loss. To address the limitations of the OFML mechanism, we introduce the One-specific-Feature-for-One-Label (OFOL) mechanism and propose a novel disentangled label feature learning (DLFL) framework to learn a disentangled representation for each label. The specificity of the framework lies in a feature disentangle module, which contains learnable semantic queries and a Semantic Spatial Cross-Attention (SSCA) module. Specifically, learnable semantic queries maintain semantic consistency between different images of the same label. The SSCA module localizes the label-related spatial regions and aggregates located region features into the corresponding label feature to achieve feature disentanglement. We achieve state-of-the-art performance on eight datasets of three tasks, \ie, multi-label classification, pedestrian attribute recognition, and continual multi-label learning. | ['Kaiqi Huang', 'Xiaotang Chen', 'Naiyu Gao', 'Fei He', 'Jian Jia'] | 2022-12-02 | null | null | null | null | ['pedestrian-attribute-recognition', 'multi-label-learning'] | ['computer-vision', 'methodology'] | [ 3.02555174e-01 -1.15722001e-01 -4.66853857e-01 -7.58571267e-01
-9.95967150e-01 -5.72763324e-01 5.13787270e-01 -4.22527045e-02
-2.05441877e-01 6.51050866e-01 -1.65668711e-01 1.27343655e-01
-3.86702865e-01 -6.42623663e-01 -6.22135639e-01 -1.16962445e+00
4.43615645e-01 1.84506088e-01 2.71189376e-03 2.39674762e-01
1.90264985e-01 5.32775104e-01 -1.88486254e+00 6.56855524e-01
6.17022336e-01 1.22831738e+00 8.62080529e-02 2.83946276e-01
6.61312044e-02 5.99341691e-01 -2.80173540e-01 -3.24605852e-01
1.53999075e-01 -4.20486242e-01 -9.31092203e-01 1.39850855e-01
7.26041973e-01 8.26566070e-02 -1.57699153e-01 1.05402815e+00
2.95651495e-01 -6.17378354e-02 1.02053761e+00 -1.80967927e+00
-8.99887204e-01 -8.89480487e-02 -7.05693722e-01 -3.43217850e-02
3.00152451e-01 -1.46571144e-01 1.41428888e+00 -1.04516685e+00
4.21896905e-01 1.32684088e+00 5.94239533e-01 3.50714505e-01
-1.37617099e+00 -8.87374997e-01 3.42833281e-01 2.85614461e-01
-1.66416824e+00 -3.21547329e-01 7.86314487e-01 -5.10816097e-01
5.38382590e-01 4.75007653e-01 4.58548635e-01 1.00867009e+00
3.40522200e-01 8.39414299e-01 1.53696585e+00 -4.97282714e-01
-8.63179937e-02 3.61507297e-01 2.91420966e-01 1.10986078e+00
2.01404601e-01 2.82965958e-01 -7.17941403e-01 -1.70472965e-01
5.22931993e-01 1.73914999e-01 -4.31797989e-02 -8.38939905e-01
-1.19481003e+00 1.02658188e+00 4.81533796e-01 1.69094682e-01
1.87921911e-01 7.00975806e-02 3.74858320e-01 3.28524947e-01
4.05006498e-01 3.25503826e-01 -5.43800175e-01 5.04945099e-01
-6.53479636e-01 1.87428385e-01 3.95570546e-01 9.46194828e-01
1.21978319e+00 -4.23492342e-01 -1.71702132e-01 6.33421540e-01
5.69424331e-01 5.15652716e-01 3.39561105e-01 -8.80537212e-01
3.41167212e-01 7.78983414e-01 -2.73872644e-01 -1.14584267e+00
-5.86966932e-01 -3.68055910e-01 -7.22759902e-01 3.08784038e-01
3.72824252e-01 2.34151214e-01 -5.57810068e-01 1.99788034e+00
4.23332602e-01 2.21471086e-01 -1.97340548e-02 8.33676815e-01
7.74826348e-01 1.61549836e-01 3.50180238e-01 -1.47459075e-01
1.55766702e+00 -1.08643782e+00 -4.72254604e-01 -1.44129440e-01
9.26849782e-01 -4.79098231e-01 9.99821782e-01 6.43300414e-02
-5.25388181e-01 -5.35103500e-01 -1.16729236e+00 -5.34037203e-02
-6.25284076e-01 1.78017363e-01 5.72315693e-01 5.58209538e-01
-6.25570714e-01 2.77424693e-01 -2.73859471e-01 -3.86037886e-01
5.42443931e-01 3.91213119e-01 -9.24263775e-01 -1.50115609e-01
-1.03032327e+00 9.82109308e-01 3.03113371e-01 -1.14353910e-01
-7.18142211e-01 -6.28499508e-01 -9.70322549e-01 -6.13885261e-02
2.85498798e-01 -7.41667449e-01 6.66364610e-01 -1.16549599e+00
-1.01726556e+00 1.31445289e+00 -1.94947571e-01 2.51025259e-01
6.35013655e-02 9.43855494e-02 -6.10998869e-01 2.33573511e-01
5.62319458e-01 6.55563653e-01 8.03242266e-01 -1.68354034e+00
-6.18395925e-01 -6.70559287e-01 -8.81626308e-02 2.25390822e-01
-3.37981246e-02 2.02602595e-02 3.26490760e-01 -5.48988163e-01
4.86356914e-01 -1.10996139e+00 2.69899547e-01 2.38766998e-01
-4.61974591e-01 -3.30050230e-01 8.74956608e-01 -3.31157357e-01
8.35333109e-01 -2.38994908e+00 1.41520172e-01 1.02984898e-01
2.58575946e-01 -2.16523498e-01 -3.34205776e-01 1.50449887e-01
-4.53306884e-01 1.78301886e-01 7.98360035e-02 -3.76285434e-01
8.93491805e-02 2.47904122e-01 -3.24489921e-02 8.35458517e-01
3.41586739e-01 1.04876220e+00 -8.46625626e-01 -8.68004501e-01
9.97228846e-02 2.06064805e-01 -3.49238724e-01 1.11565650e-01
1.61855012e-01 6.77072763e-01 -6.51510239e-01 8.40837419e-01
6.45199537e-01 -4.81474847e-01 2.48101667e-01 -6.05058968e-01
-4.84337248e-02 -2.76796132e-01 -1.18506134e+00 1.79180336e+00
-3.86908233e-01 1.82166398e-01 -1.33780763e-01 -1.05267501e+00
7.64738441e-01 3.78129929e-01 7.45794296e-01 -7.03062832e-01
9.20130461e-02 2.61105627e-01 -4.16325718e-01 -6.86819732e-01
-1.05807625e-01 -2.30150908e-01 -3.82397383e-01 5.05421698e-01
3.32008630e-01 3.68565440e-01 -3.04168344e-01 -1.50811881e-01
7.32114851e-01 1.17491290e-01 7.58631945e-01 -5.31005442e-01
5.30120552e-01 -3.36808085e-01 6.88258410e-01 5.36705077e-01
-6.60861135e-01 6.25221670e-01 3.73978198e-01 -4.81656730e-01
-7.42858052e-01 -1.28520715e+00 -4.68192488e-01 1.42729008e+00
5.01204669e-01 -1.08471215e-01 -3.73045534e-01 -1.23231208e+00
2.66392350e-01 5.30029893e-01 -8.29689324e-01 -2.68808156e-01
-2.99900800e-01 -6.85747802e-01 5.81938505e-01 4.12254125e-01
6.37524188e-01 -7.36752629e-01 -5.27229905e-01 -2.65290290e-01
-1.56377017e-01 -9.11237895e-01 -5.59303403e-01 4.24484521e-01
-2.92569011e-01 -1.40206218e+00 -1.50331736e-01 -9.23530102e-01
6.98065817e-01 4.76740986e-01 9.50150073e-01 -1.08430363e-01
-2.90349066e-01 4.47219700e-01 -3.69663984e-01 1.04060724e-01
1.37279499e-02 -6.19872995e-02 -2.51519922e-02 4.12758231e-01
4.94125098e-01 -5.23684800e-01 -5.69987178e-01 6.51244283e-01
-6.81276262e-01 6.71857074e-02 6.57723665e-01 1.24210513e+00
7.69749165e-01 -2.71609992e-01 8.74781251e-01 -6.39297903e-01
1.99923784e-01 -6.22946024e-01 -2.68066347e-01 8.22891116e-01
-7.75473058e-01 2.62584925e-01 3.12146962e-01 -3.88500303e-01
-8.97095978e-01 3.69920790e-01 1.60522416e-01 -3.15094501e-01
-6.00373745e-01 5.41732125e-02 -7.17156291e-01 -4.62460458e-01
3.59882385e-01 1.84109837e-01 1.89427033e-01 -3.03423434e-01
6.38760865e-01 5.02536952e-01 8.83726850e-02 -7.18382895e-01
4.87053990e-01 4.69086200e-01 4.68141913e-01 -2.84546196e-01
-1.46929085e+00 -5.99278808e-01 -1.07173777e+00 -3.01336974e-01
1.26191604e+00 -1.00531352e+00 -8.95908237e-01 2.78098583e-01
-9.91764128e-01 3.15659225e-01 -2.83493519e-01 3.99291277e-01
-8.46203029e-01 2.79462487e-01 -3.29687536e-01 -3.65259677e-01
1.56477287e-01 -1.41163361e+00 1.35133195e+00 1.67393953e-01
-4.29222696e-02 -9.66331720e-01 -6.47148001e-04 4.21081692e-01
7.42862448e-02 4.03836250e-01 1.22498834e+00 -8.95003676e-01
-5.64003766e-01 -2.85579767e-02 -6.57990396e-01 2.43398681e-01
5.40928245e-02 -5.02561390e-01 -1.22870100e+00 -4.33428347e-01
-2.57458864e-03 -6.92875266e-01 8.49727213e-01 8.37303102e-02
1.08541358e+00 -2.40096822e-01 -5.45133114e-01 6.40999019e-01
1.58668041e+00 -7.21106976e-02 2.35824987e-01 3.07421744e-01
8.43010426e-01 6.52271867e-01 5.84965050e-01 2.47841746e-01
6.61832690e-01 1.02661455e+00 5.22110522e-01 -1.01446256e-01
-2.25995436e-01 -7.26353079e-02 1.09404311e-01 5.79230726e-01
2.78034300e-01 -4.48199995e-02 -5.23191929e-01 1.20398194e-01
-1.99994719e+00 -9.45353150e-01 -9.55283083e-03 1.94946086e+00
5.27390659e-01 -4.58806276e-01 6.48219436e-02 -1.53903738e-02
7.29239166e-01 3.20065141e-01 -5.87871969e-01 -2.61318803e-01
-3.61839235e-01 -2.22337350e-01 2.97689468e-01 3.53977442e-01
-1.54025400e+00 6.78337634e-01 5.81807184e+00 8.03424895e-01
-7.77900338e-01 4.89648849e-01 4.56658602e-01 1.29398480e-01
-1.98480338e-01 3.37024443e-02 -9.69227433e-01 2.89215595e-01
5.01396835e-01 2.88364768e-01 1.47205681e-01 8.73178601e-01
-2.99157888e-01 -1.20590515e-01 -1.32789814e+00 1.12811124e+00
4.14376944e-01 -9.02734518e-01 2.58971810e-01 1.99966863e-01
5.02723455e-01 -4.37127173e-01 2.53125966e-01 1.62335560e-01
-1.52907893e-01 -1.06993949e+00 8.00135851e-01 6.45177841e-01
9.88317072e-01 -6.66708291e-01 5.05379498e-01 2.78488904e-01
-1.41355658e+00 -2.52462387e-01 -1.65444091e-01 2.29556039e-01
-1.05053626e-01 3.79537344e-01 -2.74429828e-01 8.66691649e-01
3.39205682e-01 7.44540215e-01 -9.49700654e-01 7.40781724e-01
2.34472416e-02 1.41744260e-02 1.48489684e-01 1.64085597e-01
1.19780570e-01 -9.72366780e-02 3.70747656e-01 1.01629043e+00
9.37876478e-02 -1.07715733e-01 5.12887716e-01 8.12911391e-01
1.84493616e-01 2.01248839e-01 -7.14127839e-01 4.69221383e-01
3.43321174e-01 1.27061820e+00 -6.41225815e-01 6.06821552e-02
-7.74297833e-01 1.11464965e+00 6.08986795e-01 4.24837291e-01
-8.10276389e-01 -1.04413241e-01 7.52509892e-01 -1.51601464e-01
8.02823231e-02 1.07524858e-03 -4.53294486e-01 -1.23990130e+00
-1.99725069e-02 -5.70664167e-01 5.69565296e-01 -5.13928533e-01
-1.67220759e+00 3.54213834e-01 4.82644662e-02 -1.16792977e+00
6.03285804e-02 -6.43784285e-01 -1.45339370e-01 8.31126273e-01
-1.57566261e+00 -1.93573368e+00 -2.28431091e-01 7.67379165e-01
4.52572703e-01 -2.55804121e-01 1.26270533e+00 4.00344193e-01
-5.51514506e-01 9.46247458e-01 1.74330026e-01 -2.09875852e-02
8.54073286e-01 -1.13898039e+00 -5.01965821e-01 4.00446117e-01
3.62064719e-01 3.01711679e-01 2.78449744e-01 -3.66872996e-01
-9.81010914e-01 -1.16857684e+00 9.79978859e-01 -5.98623931e-01
4.08221036e-01 -4.78388369e-01 -6.33479059e-01 7.04812825e-01
2.08542459e-02 4.58121687e-01 1.17291892e+00 1.64203852e-01
-9.84245360e-01 -1.68806940e-01 -1.22104454e+00 2.33852252e-01
1.06782854e+00 -8.25912654e-01 -3.92705470e-01 4.03754532e-01
7.28022695e-01 1.53293893e-01 -9.94436085e-01 6.01212978e-01
8.37975085e-01 -1.03638232e+00 1.12747955e+00 -6.51484251e-01
9.02170837e-02 -3.75345379e-01 -7.63113499e-01 -1.02481771e+00
-6.02278233e-01 2.61555225e-01 2.03013029e-02 1.17604256e+00
3.34281266e-01 -7.92734146e-01 5.04942715e-01 3.92404258e-01
-5.99968471e-02 -1.03140914e+00 -1.21660936e+00 -8.06517959e-01
1.07509688e-01 -8.16690996e-02 5.71380794e-01 1.23513758e+00
-8.48403573e-02 3.78186882e-01 -3.28300178e-01 3.25343281e-01
7.62927055e-01 4.69157904e-01 2.03755558e-01 -1.47695410e+00
-1.61102265e-01 -3.20524573e-01 -6.16124630e-01 -7.14771867e-01
7.90072560e-01 -1.28808010e+00 -1.80191010e-01 -1.17365766e+00
6.75351262e-01 -5.76800346e-01 -6.88305020e-01 6.23680651e-01
-7.33378008e-02 2.75797427e-01 3.22161943e-01 3.52765650e-01
-1.05225289e+00 6.13948822e-01 1.38918924e+00 -1.55130371e-01
3.47239494e-01 -8.35518911e-02 -7.49784708e-01 5.38376033e-01
4.50495869e-01 -6.90586329e-01 -2.86913455e-01 -6.16078600e-02
8.64093006e-02 -1.26818687e-01 7.98723400e-01 -8.87626410e-01
1.31947190e-01 -2.47342154e-01 4.85048324e-01 -4.08049792e-01
4.08316195e-01 -1.00210750e+00 1.20137669e-01 1.46766469e-01
-4.45183456e-01 2.85087619e-02 -2.44224146e-01 7.96907485e-01
-2.38908723e-01 -1.30185395e-01 9.24084365e-01 -6.65211007e-02
-9.37159061e-01 1.80885911e-01 -2.30982415e-02 2.71984786e-02
1.42065454e+00 -2.50502348e-01 -5.27165473e-01 2.06534058e-01
-8.77261162e-01 5.34756668e-02 3.66861641e-01 4.81221050e-01
6.40310287e-01 -1.64479017e+00 -4.76769984e-01 5.77123404e-01
5.54521680e-01 -2.95705080e-01 3.40213150e-01 7.82930195e-01
9.68494639e-02 4.94949996e-01 -2.89158404e-01 -7.24884093e-01
-1.26871717e+00 7.63697684e-01 4.54368502e-01 -2.41079092e-01
-3.52807105e-01 7.90964723e-01 5.74003816e-01 -6.44280672e-01
-1.09893031e-01 2.54481316e-01 -1.72348127e-01 3.29867125e-01
2.37201720e-01 3.28814894e-01 -1.36025071e-01 -1.20431256e+00
-5.87519109e-01 1.12092936e+00 -5.20619825e-02 2.07910597e-01
9.64261949e-01 -3.68013024e-01 -1.91578880e-01 7.08917201e-01
1.74472761e+00 -3.11992079e-01 -1.26100516e+00 -2.67920107e-01
1.02040373e-01 -5.08322895e-01 -1.80420712e-01 -8.36800754e-01
-9.39220846e-01 7.11196542e-01 1.02705431e+00 -5.11149950e-02
1.01160264e+00 2.97026724e-01 3.04159969e-01 1.53995305e-01
4.93729979e-01 -8.56068015e-01 4.89856333e-01 3.12403411e-01
7.02112317e-01 -1.57547593e+00 -1.62885025e-01 -6.78586841e-01
-7.58809924e-01 1.01886642e+00 7.21989930e-01 6.27461374e-02
1.01239669e+00 -4.30293940e-02 1.83727238e-02 -5.01311243e-01
-4.00543660e-01 -2.36684382e-01 5.79166472e-01 4.95229006e-01
1.77186012e-01 2.18049169e-01 -1.89810202e-01 5.93205214e-01
3.32608610e-01 -4.87408847e-01 -2.64261127e-01 7.25595534e-01
-3.97604555e-01 -1.02448928e+00 -9.22028050e-02 3.48779321e-01
-6.28978163e-02 2.43250802e-01 -3.11594196e-02 9.06821012e-01
6.35770857e-01 1.04484451e+00 -6.20795377e-02 -4.93059158e-01
2.24739209e-01 5.01441061e-01 6.26280427e-01 -5.45357645e-01
-3.99093539e-01 -2.71421313e-01 -1.41213819e-01 -7.40327537e-01
-6.73930824e-01 -7.74549007e-01 -9.64237213e-01 7.14572445e-02
-5.25902450e-01 -1.44794673e-01 5.43501556e-01 1.17042649e+00
4.71464336e-01 3.06947082e-01 8.14921737e-01 -6.08531356e-01
-5.36181986e-01 -6.29115462e-01 -9.78824973e-01 7.41266131e-01
3.43075007e-01 -1.24017501e+00 -4.01112020e-01 -1.71948597e-01] | [9.440239906311035, 4.061663627624512] |
f0a2477a-d836-44d8-a1c5-ce1d017b7d6a | shell-theory-a-statistical-model-of-reality | null | null | https://ieeexplore.ieee.org/document/9444188 | http://www.kind-of-works.com/papers/shell_theory_preprint.pdf | Shell Theory: A Statistical Model of Reality | The foundational assumption of machine learning is that the data under consideration is separable into classes; while intuitively reasonable, separability constraints have proven remarkably difficult to formulate mathematically. We believe this problem is
rooted in the mismatch between existing statistical techniques and commonly encountered data; object representations are typically high dimensional but statistical techniques tend to treat high dimensions a degenerate case. To address this problem, we develop a dedicated statistical framework for machine learning in high dimensions. The framework derives from the observation that object relations form a natural hierarchy; this leads us to model objects as instances of a high dimensional, hierarchal generative processes. Using a distance based statistical technique, also developed in this paper, we show that in such generative processes, instances of each process in the hierarchy, are almost-always encapsulated by a distinctive-shell that excludes almost-all other instances. The result is shell theory, a statistical machine learning framework in which separability constraints (distinctive-shells) are formally derived from the assumed generative process. | ['Yasuyuki Matsushita', 'Hongdong Li', 'Ngai-Man Cheung', 'Changhao Ren', 'Siying Liu', 'Wen-Yan Lin'] | 2021-05-28 | null | null | null | ieee-transactions-on-pattern-analysis-and-15 | ['unsupervised-anomaly-detection-with-specified-5', 'unsupervised-anomaly-detection-with-specified-4', 'unsupervised-anomaly-detection-with-specified-7', 'unsupervised-anomaly-detection-with-specified-6', 'unsupervised-anomaly-detection-with-specified', 'one-class-classifier'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'methodology'] | [ 1.66949302e-01 5.32651007e-01 9.36288163e-02 -3.59593332e-01
-1.56721890e-01 -6.21377051e-01 1.09429049e+00 9.23023596e-02
-1.81267243e-02 4.00747061e-01 1.63510248e-01 -1.30793869e-01
-7.31662571e-01 -9.78154957e-01 -3.91373515e-01 -1.04949439e+00
-1.28540441e-01 1.18129790e+00 1.53287232e-01 -4.72860970e-02
1.63706735e-01 6.71916485e-01 -1.71114385e+00 6.16023429e-02
6.50094032e-01 7.30942845e-01 -1.64762035e-01 5.63866615e-01
-2.41216019e-01 6.69861376e-01 -1.58638716e-01 -5.71288764e-01
3.47292542e-01 -6.36310816e-01 -9.05633807e-01 4.02596921e-01
2.42216825e-01 2.89652765e-01 -2.79204369e-01 1.09726322e+00
-7.32849836e-02 -3.53520140e-02 1.49644053e+00 -1.45344770e+00
-7.28722751e-01 4.75106180e-01 -3.51624221e-01 -3.18655185e-02
3.72630209e-02 -5.90904176e-01 1.35634959e+00 -8.23474586e-01
4.87523288e-01 1.20823181e+00 6.89433992e-01 5.34622371e-01
-1.70081425e+00 4.57828790e-02 -2.27809072e-01 -2.82279193e-01
-1.56065166e+00 -1.94475114e-01 8.33782315e-01 -1.00969887e+00
4.95968968e-01 4.80327725e-01 7.52220392e-01 5.94160736e-01
3.69564682e-01 6.73608720e-01 1.16877735e+00 -5.16147912e-01
5.43853402e-01 2.94305652e-01 5.53197503e-01 5.64788759e-01
7.69171178e-01 -3.06971725e-02 -3.79559338e-01 -2.80682951e-01
8.13781261e-01 1.24481618e-01 7.45280758e-02 -1.08689153e+00
-7.66623437e-01 1.13671112e+00 9.16180089e-02 5.86836755e-01
-2.39636019e-01 -1.32676706e-01 8.42050239e-02 1.18022494e-01
2.99229026e-01 2.87879676e-01 -1.88679516e-01 3.76176476e-01
-9.79838848e-01 4.63519484e-01 1.26355016e+00 9.09963965e-01
8.20593119e-01 -3.30103070e-01 4.51505959e-01 4.92063880e-01
5.76445103e-01 9.07134339e-02 3.13288748e-01 -6.58513069e-01
-1.53700896e-02 6.39038920e-01 -1.66653275e-01 -9.41357851e-01
-4.03235227e-01 -6.54951453e-01 -1.02971816e+00 1.46720380e-01
3.34987670e-01 3.85991335e-01 -4.79546547e-01 1.75895655e+00
2.26782814e-01 -3.09742004e-01 2.12233886e-01 4.91809368e-01
2.76931912e-01 2.87017286e-01 -1.58435777e-01 -4.94023710e-01
1.14362323e+00 -3.95472050e-01 -5.77736974e-01 3.37994039e-01
5.77988446e-01 -4.05603558e-01 6.89379811e-01 5.24612129e-01
-1.17324460e+00 -2.30033398e-01 -8.85641456e-01 6.53045550e-02
-2.69606143e-01 -2.38233358e-01 9.72136199e-01 7.52943158e-01
-1.07829297e+00 6.72357440e-01 -8.84140611e-01 -5.50980091e-01
1.97073698e-01 4.40612793e-01 -3.06128442e-01 3.17626536e-01
-7.08821058e-01 7.67344773e-01 3.64800572e-01 1.24509543e-01
-6.90469742e-01 -3.02637905e-01 -5.95652342e-01 -2.22055376e-01
8.15129727e-02 -7.54643619e-01 1.08125138e+00 -8.73809516e-01
-9.26954925e-01 1.15270436e+00 -1.32353872e-01 -3.99802864e-01
5.81095636e-01 -1.21519350e-01 -2.14605287e-01 9.97315794e-02
5.02633341e-02 8.89533088e-02 9.23062325e-01 -1.69290125e+00
-3.29242587e-01 -7.98607409e-01 -6.13548979e-02 4.21673879e-02
-1.97175339e-01 -6.33238852e-02 3.85394394e-02 -4.40822333e-01
8.89613271e-01 -1.07597435e+00 -3.06481421e-01 -3.16141695e-01
-5.90062320e-01 -5.70421100e-01 6.28101528e-01 -4.43576500e-02
1.01395929e+00 -2.08612251e+00 5.99821508e-01 4.75069076e-01
6.17848158e-01 -3.51653159e-01 4.60272044e-01 5.92293322e-01
-2.40559340e-01 2.53846943e-01 -6.44263864e-01 -3.73973429e-01
3.58577251e-01 5.23491621e-01 -4.12579477e-01 7.64470577e-01
9.07844603e-02 4.97086376e-01 -6.07412875e-01 -5.79231560e-01
1.86581314e-01 4.08935636e-01 -6.05792820e-01 5.04565760e-02
-1.08747736e-01 3.46138924e-02 -5.19313753e-01 2.64853626e-01
7.82439351e-01 -2.69927561e-01 2.52640396e-01 -1.00961380e-01
-2.01737761e-01 1.74556524e-01 -1.41438770e+00 1.34683442e+00
1.75740942e-01 2.52106965e-01 2.11263262e-02 -1.22324240e+00
9.18012738e-01 1.72928110e-01 6.51495278e-01 -4.14895713e-02
1.98495999e-01 2.33001396e-01 2.70547509e-01 -1.93849578e-01
4.23985481e-01 -7.44541824e-01 -2.66617179e-01 5.08577347e-01
8.59976187e-02 -5.40010810e-01 2.24434897e-01 3.87515247e-01
8.81267488e-01 -1.07062347e-02 3.13197941e-01 -8.17856848e-01
3.56026471e-01 1.56298000e-02 4.85234976e-01 7.88167655e-01
9.13090557e-02 5.63367307e-01 7.39600122e-01 -4.45655912e-01
-1.39802670e+00 -1.59094644e+00 -7.81136811e-01 6.20508313e-01
9.26483050e-03 -5.25292993e-01 -9.23629165e-01 -3.50898921e-01
6.38547689e-02 5.52893221e-01 -1.00584769e+00 1.04610715e-02
-9.98063162e-02 -1.06060481e+00 2.73635030e-01 2.71836668e-01
-2.07163766e-02 -6.20790064e-01 -5.34231722e-01 1.71449810e-01
4.67276752e-01 -7.39507794e-01 3.03743273e-01 4.88680303e-01
-9.78771150e-01 -1.13360894e+00 -3.87872487e-01 -5.34664571e-01
7.05468535e-01 7.61739314e-02 1.07249475e+00 4.20668907e-02
-5.71492553e-01 5.28782487e-01 -2.02904448e-01 -2.71737486e-01
-6.69626713e-01 -7.75540546e-02 4.16394502e-01 4.03411090e-02
6.42683089e-01 -7.78761506e-01 -1.76634237e-01 5.07243760e-02
-1.18378484e+00 2.61324998e-02 4.84223038e-01 4.35464054e-01
4.45817530e-01 4.94694024e-01 2.52119303e-01 -7.86873043e-01
3.22179854e-01 -5.72318017e-01 -4.73212630e-01 1.28640041e-01
-4.76505011e-01 3.58227938e-01 4.03297573e-01 -1.06374815e-01
-8.04356098e-01 -1.30533800e-01 1.84351996e-01 -4.55513149e-02
-3.98693353e-01 4.60135072e-01 -4.71042365e-01 2.88407326e-01
4.49085623e-01 2.95101583e-01 2.16976628e-01 -5.90702713e-01
2.88583517e-01 5.12561142e-01 2.20699698e-01 -8.00036371e-01
1.07501054e+00 7.31786430e-01 6.42019272e-01 -1.14310110e+00
-1.03487206e+00 -3.72695923e-01 -1.30353725e+00 -1.43509274e-02
1.01139987e+00 -6.42217100e-01 -5.67010581e-01 1.86919063e-01
-8.68653417e-01 1.02410968e-02 -4.73104060e-01 5.34045458e-01
-8.94845963e-01 3.33416224e-01 -4.35710073e-01 -1.19692731e+00
2.66925722e-01 -7.04027891e-01 9.18874860e-01 -1.09199189e-01
-4.60759908e-01 -1.23853278e+00 4.57705110e-01 2.07797617e-01
-3.14826854e-02 1.59518942e-01 1.29225409e+00 -1.09868217e+00
-3.62662643e-01 -1.75458029e-01 4.98335734e-02 3.27066451e-01
1.06644258e-01 2.65449256e-01 -6.81442320e-01 -3.42212208e-02
4.20691609e-01 -1.05472825e-01 8.22130740e-01 1.86696827e-01
9.17897701e-01 -1.65968344e-01 -3.23768735e-01 3.12859416e-01
1.64536512e+00 -6.66177273e-03 4.46022332e-01 6.96259141e-02
7.05388546e-01 1.12479591e+00 4.07941043e-02 4.07020181e-01
2.98701167e-01 5.42260230e-01 2.14983925e-01 1.08324356e-01
2.70489573e-01 -2.33490057e-02 3.74835767e-02 1.00762343e+00
-4.31294441e-01 1.36961758e-01 -1.00825298e+00 3.51184607e-01
-1.75680923e+00 -1.31360269e+00 -8.41314614e-01 2.38504577e+00
8.22577655e-01 2.23891288e-01 3.86182487e-01 4.99308646e-01
5.80290675e-01 -2.16310784e-01 -1.07039943e-01 -2.40214065e-01
-4.46699500e-01 1.64931089e-01 1.51442617e-01 5.29948771e-01
-1.08617866e+00 5.54997087e-01 7.01061392e+00 5.14859915e-01
-4.42726433e-01 7.13717118e-02 1.23515002e-01 5.26531152e-02
-4.91675913e-01 1.81852013e-01 -8.71187925e-01 2.45367333e-01
7.70980656e-01 -2.78817803e-01 -4.13194001e-02 8.15080583e-01
-7.59072229e-02 -4.72897775e-02 -1.60650694e+00 5.68913877e-01
2.26770788e-01 -1.04418385e+00 3.54478866e-01 7.15068758e-01
7.84706056e-01 -3.19593668e-01 9.54640880e-02 6.12794235e-02
6.40958011e-01 -1.18898845e+00 8.95833671e-01 5.05878210e-01
2.02076882e-01 -7.35200047e-01 3.48675400e-01 4.87357557e-01
-1.04861629e+00 -2.38453839e-02 -5.31079471e-01 -2.64142424e-01
1.64398085e-02 7.72583485e-01 -3.50996107e-01 6.07724726e-01
3.67073059e-01 5.34373820e-01 -4.65821266e-01 9.39881921e-01
7.14922324e-02 4.40876245e-01 -3.02960813e-01 1.62249789e-01
2.25645453e-01 -7.14724004e-01 5.38972616e-01 1.04728305e+00
2.96701103e-01 1.24525145e-01 3.26432623e-02 9.57475662e-01
3.81763279e-01 2.96531413e-02 -8.71193469e-01 -8.27172622e-02
3.30481343e-02 1.24137855e+00 -1.06497526e+00 -2.04942688e-01
-4.70518559e-01 6.65965855e-01 1.68573245e-01 1.40382722e-01
-5.85470438e-01 1.35237724e-01 6.00971520e-01 4.10411805e-01
2.58976877e-01 -5.00926256e-01 -4.22079295e-01 -1.26827121e+00
-1.23314373e-01 -5.01357436e-01 3.14146280e-01 -4.38515544e-01
-1.70150423e+00 4.46804851e-01 1.11794114e-01 -9.69515264e-01
-2.00249493e-01 -8.58545482e-01 -3.43062788e-01 7.58719087e-01
-7.83116996e-01 -1.19137108e+00 6.79391399e-02 6.70569599e-01
2.60145456e-01 8.56384635e-03 8.67595494e-01 -1.63233206e-01
-4.23325688e-01 -3.81156728e-02 4.49384093e-01 1.73574910e-02
3.80583405e-02 -1.69792557e+00 -1.30225986e-01 4.83119994e-01
4.73233163e-01 9.14380789e-01 1.00611055e+00 -7.99491644e-01
-1.45356739e+00 -8.46486568e-01 1.06843078e+00 -8.79001439e-01
1.06587338e+00 -7.25578547e-01 -1.05956435e+00 6.50721550e-01
1.13011226e-02 -9.57489386e-02 1.21053278e+00 4.01178688e-01
-4.39088851e-01 2.42848054e-01 -8.44121516e-01 3.16234976e-01
1.10225677e+00 -6.61437154e-01 -1.18592000e+00 3.53527308e-01
1.91376582e-01 4.37500149e-01 -9.51080501e-01 3.72390270e-01
4.51418281e-01 -1.16316581e+00 7.57167876e-01 -8.86019289e-01
5.06027281e-01 -3.03581625e-01 -5.50829232e-01 -7.76990175e-01
-2.63886392e-01 -5.57364285e-01 -1.38468444e-01 1.15254557e+00
2.60063082e-01 -4.11244541e-01 9.52589691e-01 8.75922143e-01
-3.47542353e-02 -6.06411397e-01 -8.67477119e-01 -1.13529825e+00
6.02002442e-01 -3.66609812e-01 2.38501981e-01 9.97281373e-01
9.33529660e-02 6.21415675e-01 -3.90190557e-02 7.79365674e-02
1.33753753e+00 1.48047209e-01 4.83028382e-01 -2.20898795e+00
-5.07291496e-01 -5.83793700e-01 -7.93179393e-01 -6.74395084e-01
3.23170215e-01 -9.52865422e-01 4.76199435e-03 -1.61746490e+00
6.17854893e-01 -7.27339625e-01 -2.99563054e-02 -6.43396154e-02
4.22152728e-01 8.82355645e-02 -4.99818102e-02 7.71969259e-01
-4.04831171e-01 5.36342561e-01 7.80335069e-01 3.76221925e-01
-6.95713088e-02 1.42813653e-01 -7.15339482e-01 1.19124806e+00
4.86412525e-01 -4.91797715e-01 -4.56822664e-01 4.01378609e-02
5.94754219e-01 -1.92184895e-01 6.04200423e-01 -9.12143290e-01
2.59660423e-01 2.48147920e-02 3.40160906e-01 -7.20805466e-01
3.81633490e-01 -8.75497520e-01 5.11799335e-01 4.32326794e-01
-3.59229892e-01 -3.18006277e-01 -3.94379854e-01 5.35115898e-01
-6.40583932e-02 -5.70460618e-01 6.67280436e-01 -1.27030134e-01
-2.01815948e-01 2.35839233e-01 -4.26689893e-01 4.17558253e-02
1.19560182e+00 -4.06439275e-01 1.01164073e-01 -4.99410443e-02
-1.17956185e+00 -3.85678142e-01 7.03672230e-01 -9.75309685e-02
2.90901333e-01 -1.26170039e+00 -5.96902251e-01 7.27685615e-02
-3.10124885e-02 4.17419672e-02 1.40319848e-02 8.96268427e-01
-2.27427721e-01 3.85136217e-01 4.81412522e-02 -7.12366402e-01
-1.07681096e+00 6.53178811e-01 2.53140509e-01 4.53621335e-02
-6.23790503e-01 7.55119145e-01 6.83855176e-01 -3.64950329e-01
-9.13544446e-02 2.19746046e-02 -5.35041243e-02 2.56213754e-01
2.18704507e-01 1.73433945e-01 -5.55595495e-02 -9.46090698e-01
-2.38728121e-01 5.93678534e-01 -1.17135094e-02 -3.37691784e-01
1.46454453e+00 -6.53424487e-02 -5.56563377e-01 1.01017630e+00
1.07129991e+00 1.17051855e-01 -9.27775860e-01 -3.11954647e-01
3.36161137e-01 -2.29864493e-01 -3.48290116e-01 -1.97549075e-01
-3.74123693e-01 9.16594982e-01 5.46331629e-02 1.17292368e+00
7.34926522e-01 7.59301364e-01 -4.18901592e-02 2.77993739e-01
3.20084900e-01 -9.55519855e-01 -3.19935642e-02 4.41045523e-01
9.10199583e-01 -7.30540872e-01 1.46408854e-02 -6.96519196e-01
-2.48790413e-01 1.19586599e+00 2.42115498e-01 -4.42710847e-01
8.17176104e-01 3.96110058e-01 -6.45271599e-01 -5.18508077e-01
-6.54996753e-01 -2.47750983e-01 3.30646008e-01 5.46465874e-01
5.65963268e-01 1.61494076e-01 -2.19882771e-01 7.25580394e-01
-6.16861939e-01 -4.52749938e-01 6.03983700e-01 8.39662492e-01
-7.53217161e-01 -1.13180351e+00 -3.28375548e-01 5.37278712e-01
-4.18005884e-01 -8.17866847e-02 -6.37318909e-01 1.03867149e+00
2.78886616e-01 6.37909412e-01 4.91501391e-01 -6.23966642e-02
1.41650327e-02 1.24517761e-01 8.85717809e-01 -8.69903326e-01
-1.54490754e-01 3.16247284e-01 -1.63015574e-01 -5.65081164e-02
-7.22031713e-01 -1.12179863e+00 -1.13528633e+00 -3.01153541e-01
-3.18289816e-01 6.33896470e-01 7.59211302e-01 1.10853255e+00
-1.90691829e-01 3.10022533e-01 4.94656771e-01 -6.10312521e-01
-8.67090642e-01 -6.80742383e-01 -1.31690753e+00 2.79808491e-01
2.24996284e-01 -9.14960623e-01 -8.98600817e-01 2.75283575e-01] | [7.314537048339844, 4.594597816467285] |
e6da29a0-02dc-45fa-82c8-d0734ec68372 | combined-machine-learning-and-physics-based | 2303.09073 | null | https://arxiv.org/abs/2303.09073v1 | https://arxiv.org/pdf/2303.09073v1.pdf | Combined Machine Learning and Physics-Based Forecaster for Intra-day and 1-Week Ahead Solar Irradiance Forecasting Under Variable Weather Conditions | Power systems engineers are actively developing larger power plants out of photovoltaics imposing some major challenges which include its intermittent power generation and its poor dispatchability. The issue is that PV is a variable generation source unless additional planning and system additions for mitigation of generation intermittencies. One underlying factor that can enhance the applications around mitigating distributed energy resource intermittency challenges is forecasting the generation output. This is challenging especially with renewable energy sources which are weather dependent as due to the random nature of weather variance. This work puts forth a forecasting model which uses the solar variables to produce a PV generation forecast and evaluates a set of machine learning models for this task. In this paper, a forecaster for irradiance prediction for intra-day is proposed. This forecaster is capable of forecasting 15 minutes and hourly irradiance up to one week ahead. The paper performed a correlation and sensitivity analysis of the strength of the relationship between local weather parameters and system generation. In this study performance of SVM, CART, ANN, and Ensemble learning were analyzed for the prediction of 15-minute intraday and day-ahead irradiance. The results show that SVM and Ensemble learning yielded the lowest MAE for 15-minute intraday and day-ahead irradiance, respectively. | ['Arif Sarwat', 'Mohd Tariq', 'Shahid Tufail', 'Hugo Riggs'] | 2023-03-16 | null | null | null | null | ['solar-irradiance-forecasting'] | ['time-series'] | [-9.21560526e-02 -3.60071212e-01 1.31299078e-01 -5.04087880e-02
1.10973030e-01 -1.03954756e+00 8.58041823e-01 1.68654352e-01
5.31442821e-01 1.45658112e+00 9.61918756e-02 -5.35716534e-01
-4.71228123e-01 -1.04434943e+00 1.00628853e-01 -1.10265183e+00
2.59584673e-02 -7.98595622e-02 -4.10139740e-01 -4.16128516e-01
3.18348378e-01 6.93856061e-01 -1.69193363e+00 -8.87349546e-02
1.42469561e+00 7.64856160e-01 5.56766391e-01 1.01317680e+00
2.27067828e-01 3.01837862e-01 -1.06776571e+00 5.98938346e-01
3.23950887e-01 -5.80948949e-01 5.39914817e-02 -3.64143610e-01
-1.22079566e-01 -2.54619330e-01 3.50518286e-01 6.24891579e-01
8.64752769e-01 1.27817333e-01 9.76553202e-01 -1.67810738e+00
-3.78078312e-01 2.90174574e-01 -1.60259157e-01 5.78005373e-01
1.57110065e-01 3.21750820e-01 3.25793684e-01 -5.23728669e-01
-2.29285136e-01 5.12538612e-01 4.93022919e-01 -2.35760108e-01
-8.52060258e-01 -2.58046329e-01 -3.50580394e-01 2.56141067e-01
-7.99070895e-01 7.08821863e-02 6.89250231e-01 -7.86156118e-01
1.66427433e+00 5.83611548e-01 8.99656475e-01 4.27954972e-01
9.19205487e-01 -1.84699699e-01 1.82548416e+00 -6.32160246e-01
5.67001939e-01 2.90850312e-01 1.06625007e-02 -1.44488633e-01
3.42898756e-01 5.12075365e-01 -5.63301593e-02 1.16753779e-01
1.29938439e-01 -2.44875327e-01 -4.12216753e-01 4.96262282e-01
-4.57788765e-01 8.75018775e-01 3.97688508e-01 6.58773601e-01
-4.56811547e-01 -5.49007058e-01 1.17731303e-01 4.02214259e-01
5.49293041e-01 3.46831262e-01 -8.91540170e-01 -9.13255587e-02
-1.02152753e+00 -1.18965618e-01 1.05764675e+00 5.18441617e-01
2.73316920e-01 9.39369321e-01 -3.48887816e-02 4.09850895e-01
1.60209462e-01 1.04636395e+00 7.31786132e-01 -4.35031772e-01
6.17138781e-02 5.05150855e-01 3.51347327e-01 -6.24060631e-01
-5.26800215e-01 -7.42607117e-01 -1.12523568e+00 9.15400207e-01
-2.89341956e-02 -9.58530247e-01 -9.13852394e-01 1.02463758e+00
2.09555313e-01 -2.88163394e-01 3.09637338e-01 4.82681364e-01
3.64468992e-01 1.26252759e+00 -5.50640374e-02 -8.80513906e-01
8.32784176e-01 -1.02145326e+00 -9.37822223e-01 3.01483840e-01
2.75293410e-01 -8.41383874e-01 4.23317105e-02 3.83041680e-01
-8.57652485e-01 -7.60919452e-01 -1.12416911e+00 4.89322096e-01
-1.20512462e+00 3.89982522e-01 1.68366432e-01 9.76489484e-01
-9.55832481e-01 8.71191800e-01 -4.05322582e-01 -5.07714689e-01
-1.50278047e-01 1.20835833e-01 2.18370557e-01 7.83449292e-01
-1.09533989e+00 1.85005450e+00 8.73018742e-01 4.41977143e-01
-2.52089560e-01 -7.68845439e-01 -5.28645575e-01 4.02044117e-01
-2.88042843e-01 -6.77286267e-01 8.78420472e-01 -7.78864145e-01
-1.74158728e+00 -4.34790969e-01 -2.60906398e-01 -6.48352802e-01
3.26839060e-01 1.97271660e-01 -9.64786351e-01 -4.17987347e-01
-1.06311359e-01 -2.51950473e-01 5.93810201e-01 -8.90305340e-01
-1.01370692e+00 -2.84334451e-01 -6.04167402e-01 3.74476552e-01
-4.30022515e-02 -5.10831714e-01 1.20977974e+00 -4.25503939e-01
-3.53669852e-01 -6.07862353e-01 -9.14370567e-02 -9.88753378e-01
-1.58954993e-01 -5.02784908e-01 1.33164775e+00 -1.13415575e+00
1.08496439e+00 -1.48185158e+00 -2.08719268e-01 4.03898507e-01
-7.95626879e-01 4.06499326e-01 3.15276265e-01 1.12703013e+00
-4.59191114e-01 7.40988925e-03 -4.80855629e-02 3.65238011e-01
-1.48313334e-02 5.95272243e-01 -6.35419130e-01 5.21928445e-02
1.31101355e-01 6.30314946e-01 -6.72937512e-01 4.03292358e-01
1.00388896e+00 4.63354021e-01 5.94707310e-01 6.77243397e-02
-2.53038526e-01 3.55534256e-01 8.40541273e-02 5.61567724e-01
7.72392094e-01 3.11713871e-02 7.06940219e-02 -2.04446211e-01
-8.89276087e-01 -1.30668342e-01 -1.13517857e+00 8.58358622e-01
-6.99838161e-01 7.42443919e-01 -4.04976338e-01 -1.15226674e+00
1.06922293e+00 4.87804413e-01 2.89354295e-01 -6.39196813e-01
-1.60602763e-01 3.70182216e-01 8.27531591e-02 -4.98543471e-01
3.64081711e-01 9.59518328e-02 7.32157826e-01 2.07123369e-01
-7.66497329e-02 -4.02608722e-01 4.96522009e-01 -4.24212396e-01
3.21837664e-01 9.57388505e-02 8.28826427e-01 -6.39248133e-01
7.12657690e-01 -2.13146433e-01 4.81001496e-01 9.63086560e-02
1.89300224e-01 1.01424232e-01 9.00276527e-02 -2.38535702e-01
-1.00097871e+00 -7.97660351e-01 -4.25658315e-01 5.23632586e-01
-3.08918744e-01 1.92543715e-01 -3.53021741e-01 -3.11444134e-01
2.70380795e-01 1.60305357e+00 -3.81873697e-01 4.40312296e-01
-1.47443429e-01 -1.06243503e+00 -3.76839131e-01 3.83400261e-01
5.46704769e-01 -7.86636353e-01 -7.23951459e-01 3.13653976e-01
4.09885526e-01 -4.47373807e-01 2.73935199e-01 6.40149951e-01
-8.74916971e-01 -8.10206294e-01 -8.56982648e-01 -3.90768737e-01
6.11290872e-01 6.76210085e-03 8.51881802e-01 -1.55745059e-01
-8.88768733e-02 3.44618820e-02 -2.60691196e-01 -1.11322939e+00
-4.01772231e-01 5.61605357e-02 -1.64546687e-02 -7.73764074e-01
-1.46724150e-01 -8.40705633e-01 -6.48367703e-01 1.46318316e-01
-4.10499394e-01 7.86425695e-02 4.36539322e-01 8.01634729e-01
-2.88959360e-03 8.54247987e-01 1.26047325e+00 -2.76279688e-01
7.62323320e-01 -9.39098597e-01 -1.17153752e+00 4.95795399e-01
-1.41908002e+00 -2.99557924e-01 8.67880166e-01 1.61195487e-01
-1.32927346e+00 -1.87357232e-01 2.71780670e-01 3.15049469e-01
-4.31881189e-01 6.83778584e-01 1.32218108e-01 -1.18627973e-01
3.49369228e-01 4.02667433e-01 -2.55808562e-01 -4.86846179e-01
3.02152280e-02 6.47873163e-01 7.77155757e-02 1.20290272e-01
1.08223820e+00 -1.58152476e-01 6.32201195e-01 -1.06835604e+00
-2.17681482e-01 -3.20958734e-01 -3.57288063e-01 -5.08609176e-01
5.85339487e-01 -8.15231740e-01 -5.84310293e-01 8.53100240e-01
-8.74342620e-01 -3.23765308e-01 -2.79401839e-01 3.08732510e-01
-8.71287510e-02 9.01730955e-02 1.48738012e-01 -1.39536321e+00
-7.59619296e-01 -5.86290896e-01 2.64319062e-01 1.06233406e+00
8.70645419e-02 -1.53780448e+00 2.75151402e-01 3.30164507e-02
9.66937959e-01 8.01232994e-01 8.28767180e-01 -1.12265572e-01
-2.47016355e-01 5.29932752e-02 7.85608962e-02 7.90538073e-01
7.18326390e-01 6.17674530e-01 -9.83137012e-01 -4.25933361e-01
3.45086567e-02 2.84662396e-01 1.39370844e-01 6.15137935e-01
6.48154020e-01 -5.24646223e-01 -1.26672596e-01 3.54926229e-01
2.17938042e+00 8.23593616e-01 4.45093006e-01 3.73937577e-01
1.32312700e-02 4.11375225e-01 4.47141916e-01 5.90888977e-01
2.23189682e-01 2.34291360e-01 5.51043391e-01 -5.92709295e-02
2.42806956e-01 -2.70552118e-03 2.62576431e-01 7.08368242e-01
-4.43043798e-01 -5.49648046e-01 -5.76084971e-01 7.14853704e-01
-1.62356496e+00 -1.33446896e+00 -5.20414770e-01 1.93392873e+00
4.55080301e-01 -2.85390764e-01 2.84370836e-02 5.25661767e-01
3.42790693e-01 1.16076186e-01 -4.61106479e-01 -9.99776006e-01
-5.33617854e-01 4.95775431e-01 7.12203622e-01 5.93395054e-01
-7.14771509e-01 -1.14791684e-01 5.95232821e+00 3.54235530e-01
-1.42413926e+00 -6.87426478e-02 5.84205329e-01 1.34084702e-01
-2.06630707e-01 -2.78140083e-02 -7.86179841e-01 1.28533483e+00
1.22182798e+00 -7.70317554e-01 4.89730328e-01 7.38220990e-01
9.20094609e-01 -8.90773237e-01 -5.89907348e-01 2.45640844e-01
-1.86906889e-01 -1.18401051e+00 -3.19211215e-01 8.59050751e-02
1.68446255e+00 2.83435285e-01 -2.92241901e-01 9.93083492e-02
6.17198460e-02 -1.04627895e+00 1.70710925e-02 1.01467121e+00
1.76138002e-02 -7.43261874e-01 1.14087760e+00 7.24144578e-01
-1.03694892e+00 -5.04163682e-01 -1.77971512e-01 -7.07435429e-01
3.99001449e-01 1.01249695e+00 -9.50689018e-01 1.21854162e+00
7.24981725e-01 4.57004786e-01 -4.11202282e-01 1.08616018e+00
-3.38450640e-01 6.01728737e-01 -7.58294821e-01 -2.45354638e-01
-2.05348521e-01 -5.73396623e-01 2.95027584e-01 1.03210270e+00
1.12602890e+00 2.51670539e-01 -3.06397736e-01 3.34098577e-01
7.39226401e-01 -1.22588083e-01 -7.67478049e-01 3.18022996e-01
7.15109110e-01 1.46692574e+00 -2.97355950e-01 -3.39429379e-01
-1.80891320e-01 5.29008985e-01 -5.54677248e-01 6.77209198e-01
-4.49876845e-01 -2.77120382e-01 3.95126194e-01 -3.76174450e-01
3.60501677e-01 -2.11643785e-01 -1.00433147e+00 -6.90899193e-01
2.19381362e-01 -3.51637661e-01 3.19528937e-01 -1.33022451e+00
-1.45375919e+00 1.69033051e-01 5.09877354e-02 -1.07481432e+00
-9.63752627e-01 -6.18006349e-01 -1.62117660e+00 1.88348293e+00
-1.96084571e+00 -9.26923156e-01 -5.11991084e-01 2.81016231e-01
6.93296075e-01 -4.46118325e-01 1.09981096e+00 -3.03735316e-01
-6.51194155e-01 -1.74659371e-01 9.57943797e-01 -6.72847927e-01
1.45232761e-02 -1.70056117e+00 -2.31947660e-01 1.07452512e+00
-3.08307648e-01 -4.79211546e-02 8.85145068e-01 -6.69401586e-01
-1.01744485e+00 -8.72984529e-01 1.10758781e+00 -2.32519209e-02
4.26029444e-01 4.19278890e-01 -6.59335971e-01 3.11425567e-01
1.31744206e+00 -5.79160929e-01 6.00686133e-01 -2.70915836e-01
5.06740689e-01 -3.69476169e-01 -1.52423155e+00 1.33786112e-01
-2.07651138e-01 -1.60196081e-01 -6.09520078e-01 5.08183956e-01
-1.76230133e-01 -2.30239749e-01 -1.37613654e+00 5.06040931e-01
4.72690016e-01 -1.00871980e+00 4.13829446e-01 -1.85395554e-02
2.17522025e-01 -6.38168573e-01 5.13842218e-02 -2.20324564e+00
-2.67654479e-01 -7.57870436e-01 -4.93881196e-01 1.39063823e+00
7.06351638e-01 -1.24073255e+00 2.92398244e-01 5.03218114e-01
1.44026745e-02 -6.88020170e-01 -8.40522945e-01 -8.11527789e-01
8.44575688e-02 4.02267575e-01 7.02796817e-01 1.19140267e+00
4.36483044e-03 1.97486520e-01 -4.00889032e-02 7.26338744e-01
5.77132821e-01 3.37353915e-01 6.60707831e-01 -1.14629328e+00
1.07636727e-01 -3.97716165e-01 3.37767303e-02 1.35896683e-01
-4.11907077e-01 -3.45387369e-01 -3.88533235e-01 -2.12498832e+00
-5.74061096e-01 1.38177916e-01 -3.90644252e-01 4.88758326e-01
-1.76265150e-01 -3.31320226e-01 3.52832586e-01 -1.39401138e-01
1.13601780e+00 6.63429439e-01 7.67672420e-01 1.24656931e-01
-3.34106266e-01 4.30139542e-01 -6.05725087e-02 4.93494958e-01
1.61882007e+00 1.09164581e-01 -7.37964869e-01 -1.45371050e-01
2.95844469e-02 2.34551415e-01 2.25592837e-01 -1.24726748e+00
1.61035687e-01 -5.25693774e-01 8.49741518e-01 -1.15515029e+00
-8.25210661e-02 -1.19489837e+00 8.18881631e-01 5.85399091e-01
3.95330459e-01 3.91777486e-01 3.20445210e-01 2.80676723e-01
-7.70771652e-02 -3.20122212e-01 4.19213831e-01 3.29836160e-01
-7.27658212e-01 -3.20662409e-01 -3.80723596e-01 -7.10647225e-01
1.49693644e+00 -4.75358397e-01 -1.06668663e+00 -2.85475612e-01
-6.37006640e-01 5.13256967e-01 2.68892318e-01 3.99097204e-01
-6.17809854e-02 -9.35479462e-01 -7.77596116e-01 1.65957779e-01
-6.37448132e-01 -5.39266407e-01 2.76010782e-01 5.82130432e-01
-4.65488434e-01 6.27367616e-01 -6.73718750e-01 -2.82594293e-01
-1.24092805e+00 2.71152377e-01 5.84914565e-01 -1.43047318e-01
3.33367847e-02 1.01464525e-01 -6.18518054e-01 9.55787748e-02
-3.88042957e-01 -3.91647965e-01 -5.14138460e-01 4.55358356e-01
7.28521645e-02 9.46914077e-01 3.73175502e-01 -1.61028609e-01
5.53514808e-02 5.05379915e-01 7.68411756e-01 2.50359148e-01
1.36634457e+00 -2.04273403e-01 -2.57739156e-01 5.85459709e-01
8.03317606e-01 -1.51567087e-01 -1.09348309e+00 6.19187117e-01
-1.47292629e-01 -1.66800112e-01 2.23027304e-01 -1.91032946e+00
-7.79761672e-01 3.99306685e-01 9.59190488e-01 1.11276340e+00
1.62214482e+00 -1.03526604e+00 3.01917642e-01 1.20702356e-01
1.41655192e-01 -1.47756505e+00 -9.56413627e-01 4.95167941e-01
1.07961965e+00 -1.23000097e+00 3.22131068e-01 -1.07571676e-01
-2.32973322e-01 1.41309083e+00 3.04474235e-01 9.02781412e-02
8.32697630e-01 4.15427953e-01 1.14000784e-02 4.17226434e-01
-8.37596178e-01 -9.90552679e-02 1.50947779e-01 8.23932350e-01
5.25013745e-01 4.77682918e-01 -7.96870291e-01 -8.65675230e-03
-5.20660460e-01 1.87141657e-01 5.17600417e-01 7.59202361e-01
-4.47290152e-01 -9.90115047e-01 -5.00630379e-01 6.07493639e-01
-2.84474880e-01 1.56024739e-03 8.08317885e-02 6.83560133e-01
5.86213410e-01 1.30053592e+00 1.17407991e-02 2.57602870e-01
1.86118722e-01 1.71304300e-01 4.92826775e-02 -9.04976055e-02
-7.15446711e-01 -2.86297888e-01 1.45534113e-01 2.71041572e-01
-3.36028010e-01 -6.93201303e-01 -8.71084273e-01 -3.14240783e-01
-5.57881892e-01 5.00394464e-01 1.57656157e+00 1.10775292e+00
1.09368384e-01 5.84296405e-01 1.42160249e+00 -8.51751387e-01
-7.78971672e-01 -1.27035844e+00 -6.48070812e-01 -2.70946503e-01
1.94208205e-01 -2.17623785e-01 -7.98831761e-01 6.20686682e-03] | [6.218637943267822, 2.8170254230499268] |
7cf0b05d-ce41-4095-a95e-cc3c6fc73adf | matching-cnn-meets-knn-quasi-parametric-human | 1504.01220 | null | http://arxiv.org/abs/1504.01220v1 | http://arxiv.org/pdf/1504.01220v1.pdf | Matching-CNN Meets KNN: Quasi-Parametric Human Parsing | Both parametric and non-parametric approaches have demonstrated encouraging
performances in the human parsing task, namely segmenting a human image into
several semantic regions (e.g., hat, bag, left arm, face). In this work, we aim
to develop a new solution with the advantages of both methodologies, namely
supervision from annotated data and the flexibility to use newly annotated
(possibly uncommon) images, and present a quasi-parametric human parsing model.
Under the classic K Nearest Neighbor (KNN)-based nonparametric framework, the
parametric Matching Convolutional Neural Network (M-CNN) is proposed to predict
the matching confidence and displacements of the best matched region in the
testing image for a particular semantic region in one KNN image. Given a
testing image, we first retrieve its KNN images from the
annotated/manually-parsed human image corpus. Then each semantic region in each
KNN image is matched with confidence to the testing image using M-CNN, and the
matched regions from all KNN images are further fused, followed by a superpixel
smoothing procedure to obtain the ultimate human parsing result. The M-CNN
differs from the classic CNN in that the tailored cross image matching filters
are introduced to characterize the matching between the testing image and the
semantic region of a KNN image. The cross image matching filters are defined at
different convolutional layers, each aiming to capture a particular range of
displacements. Comprehensive evaluations over a large dataset with 7,700
annotated human images well demonstrate the significant performance gain from
the quasi-parametric model over the state-of-the-arts, for the human parsing
task. | ['Xiaochun Cao', 'Liang Lin', 'Xiaohui Shen', 'Luoqi Liu', 'Xiaodan Liang', 'Jianchao Yang', 'Si Liu', 'Changsheng Xu', 'Shuicheng Yan'] | 2015-04-06 | matching-cnn-meets-knn-quasi-parametric-human-1 | http://openaccess.thecvf.com/content_cvpr_2015/html/Liu_Matching-CNN_Meets_KNN_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Liu_Matching-CNN_Meets_KNN_2015_CVPR_paper.pdf | cvpr-2015-6 | ['human-parsing'] | ['computer-vision'] | [ 3.91050041e-01 5.65844178e-01 -1.50130719e-01 -7.18146443e-01
-1.03047967e+00 -3.83199632e-01 2.69207865e-01 -1.84357762e-02
-5.38562477e-01 3.55490595e-01 -1.45216882e-01 2.61173397e-01
-2.44451568e-01 -6.98460698e-01 -9.81143773e-01 -5.60266852e-01
2.26553872e-01 6.02544725e-01 5.54802001e-01 1.19056627e-01
-1.63859203e-02 6.11437678e-01 -1.52504909e+00 3.57341111e-01
8.32212150e-01 1.41993117e+00 4.14360702e-01 6.14816368e-01
-1.88497216e-01 2.54235804e-01 -6.31569982e-01 -5.82686901e-01
1.01536371e-01 -3.28588426e-01 -9.03522968e-01 2.09139869e-01
8.08827281e-01 -1.22998759e-01 -4.78103347e-02 1.15086603e+00
3.50515783e-01 2.03491017e-01 5.84114552e-01 -1.02315724e+00
-7.16191888e-01 4.47752863e-01 -4.14031208e-01 -1.58843353e-01
2.28305191e-01 -1.93691939e-01 8.70992541e-01 -7.71844804e-01
7.78947651e-01 1.23092425e+00 8.66366744e-01 6.48948789e-01
-1.10855079e+00 -5.35975695e-01 -6.81383759e-02 2.47750990e-02
-1.20542884e+00 5.75204473e-03 8.33121121e-01 -4.76320505e-01
6.19194686e-01 -7.68996477e-02 4.39167649e-01 8.79292846e-01
9.66623984e-03 9.74312425e-01 9.26370025e-01 -4.72970665e-01
2.55524099e-01 7.68497810e-02 1.73529580e-01 8.55942607e-01
-6.28655702e-02 3.60373929e-02 -4.52701241e-01 -1.48907769e-02
6.98465228e-01 -2.91330487e-01 2.30954178e-02 -6.42057419e-01
-8.95677447e-01 7.79487967e-01 7.40606964e-01 3.63916337e-01
-3.84218991e-01 -9.85894799e-02 4.16572481e-01 -3.48647296e-01
4.17890251e-01 1.10468663e-01 -8.22928727e-01 3.76571268e-01
-1.27090299e+00 2.64804840e-01 6.10205770e-01 9.24714863e-01
1.01243079e+00 -3.79228950e-01 -2.68752754e-01 9.61672008e-01
4.71155569e-02 1.47213191e-01 4.51955408e-01 -1.27837837e+00
5.73947787e-01 8.36984158e-01 -4.02847975e-02 -9.28193867e-01
-6.18820369e-01 -2.11150020e-01 -7.58532643e-01 1.97498649e-01
9.44145679e-01 6.11243909e-03 -1.14936709e+00 1.61048818e+00
5.35902500e-01 1.10719815e-01 1.93561465e-01 9.12666917e-01
1.01180840e+00 3.98340374e-01 3.98952305e-01 1.66310281e-01
1.56343031e+00 -1.12701321e+00 -2.73061603e-01 -4.94703144e-01
3.48482877e-01 -5.96082687e-01 1.04105604e+00 1.63553178e-01
-1.04869294e+00 -1.05348706e+00 -8.96151662e-01 -2.16242105e-01
-5.10344028e-01 5.58139443e-01 3.34765285e-01 5.09500444e-01
-9.63717282e-01 6.64059162e-01 -6.29554927e-01 -3.67689192e-01
5.16575038e-01 2.52001673e-01 -7.40662336e-01 -2.10600585e-01
-1.22941804e+00 5.72017729e-01 9.13973212e-01 4.56903130e-01
-4.89994913e-01 -6.54895425e-01 -1.19312167e+00 2.30376963e-02
3.79759580e-01 -6.01552784e-01 1.02907634e+00 -1.37180793e+00
-1.25015306e+00 1.32742620e+00 5.35920933e-02 -4.16893542e-01
6.22097313e-01 -9.54792574e-02 -1.46983936e-01 2.88353622e-01
4.30626214e-01 1.20284688e+00 8.68588150e-01 -1.34209180e+00
-8.33311558e-01 -5.39084613e-01 -7.20135272e-02 -7.74460286e-02
1.41845629e-01 4.78241816e-02 -9.66719985e-01 -6.91951454e-01
4.22623217e-01 -9.00741220e-01 -1.95890337e-01 -3.00614536e-01
-6.27038479e-01 -3.35351259e-01 7.21197009e-01 -9.50089574e-01
7.26738095e-01 -2.23212957e+00 2.00495958e-01 1.31960779e-01
-5.44740669e-02 2.19158813e-01 -1.16290942e-01 -2.92570025e-01
-1.62980109e-01 -2.09104463e-01 -5.58423042e-01 -5.20065188e-01
-5.64008113e-03 3.50614429e-01 2.96065986e-01 3.31705570e-01
3.47987205e-01 1.12258029e+00 -6.42830133e-01 -8.48860741e-01
1.32994026e-01 4.73826349e-01 -3.36781681e-01 3.29862058e-01
-2.39074990e-01 5.41015029e-01 -4.48319376e-01 7.89581180e-01
7.07861722e-01 -2.05261968e-02 7.32446983e-02 -3.78063142e-01
1.54391333e-01 -5.74042618e-01 -1.12497425e+00 1.68468976e+00
-2.16942385e-01 3.24567735e-01 1.98537558e-01 -1.13291323e+00
1.13074791e+00 8.40928257e-02 2.64235407e-01 -6.94716096e-01
1.06672190e-01 2.97666192e-01 -3.94099891e-01 -5.51445663e-01
4.10916895e-01 6.17564209e-02 -3.96689802e-01 -2.13785723e-01
4.41142380e-01 -3.41842659e-02 2.49010518e-01 -2.52791584e-01
6.19741023e-01 2.86878049e-01 2.35259011e-01 -2.67195284e-01
7.17593968e-01 2.70723581e-01 8.03169608e-01 8.39436173e-01
-4.85331535e-01 8.69776130e-01 7.80225635e-01 -6.84034944e-01
-1.09678352e+00 -1.07706666e+00 -1.81676194e-01 1.27970231e+00
2.22289234e-01 2.17662618e-01 -1.47546768e+00 -9.48365033e-01
2.26570275e-02 4.75604355e-01 -8.63778710e-01 -1.31361065e-02
-9.02351856e-01 -3.94488782e-01 6.10328078e-01 9.44367409e-01
8.41288447e-01 -1.54747224e+00 -6.76729321e-01 1.05956517e-01
-2.74248362e-01 -1.40390134e+00 -4.66649532e-01 2.26452708e-01
-7.57058263e-01 -1.19527090e+00 -7.29208827e-01 -1.26861989e+00
7.73381591e-01 -4.98129666e-01 1.18687713e+00 -3.25127155e-01
-4.98490810e-01 2.75564879e-01 -3.72164577e-01 2.22764304e-03
-4.23425764e-01 -9.18985456e-02 -3.95745814e-01 7.79888406e-02
2.63527244e-01 -1.21366223e-02 -5.51932454e-01 4.27143157e-01
-8.05303752e-01 -1.37211204e-01 8.14802170e-01 9.58070755e-01
1.03334522e+00 1.96058497e-01 3.53464425e-01 -1.08419621e+00
1.35498524e-01 -2.03706548e-01 -6.52300298e-01 4.10606205e-01
-2.71637261e-01 -4.47425619e-02 5.97527862e-01 -4.69167411e-01
-1.11801517e+00 7.13389635e-01 -5.58979698e-02 -3.82272422e-01
-6.13597929e-01 2.67612457e-01 -5.98492324e-01 1.69863194e-01
4.66076851e-01 -9.33449157e-03 -1.53116778e-01 -3.73273700e-01
5.54211736e-01 3.81133616e-01 1.27117074e+00 -7.05792964e-01
4.73222643e-01 3.53489459e-01 -1.27128720e-01 -3.66552174e-01
-1.03000283e+00 -4.78896290e-01 -1.13288713e+00 -1.95431560e-01
1.56689620e+00 -6.74927354e-01 -3.79921615e-01 5.42101979e-01
-1.27841544e+00 -3.96052182e-01 -2.65571237e-01 2.99130023e-01
-7.63206840e-01 2.96078593e-01 -5.62550902e-01 -4.51632589e-01
-2.63167530e-01 -1.34616113e+00 1.41583037e+00 6.27640128e-01
-1.11197971e-01 -8.47699821e-01 -2.30456069e-01 8.16139400e-01
-1.78894982e-01 3.38619262e-01 1.08590102e+00 -9.25317466e-01
-2.58900732e-01 -5.02699912e-01 -4.29455042e-01 7.12621510e-01
-2.13155821e-01 -1.72924757e-01 -9.66330171e-01 -9.14481096e-03
-2.87754446e-01 -2.05225676e-01 8.36178422e-01 7.90044904e-01
1.29314423e+00 2.99993698e-02 -2.67604202e-01 6.05334640e-01
1.30226004e+00 1.06644236e-01 5.49101651e-01 4.20070410e-01
6.45402610e-01 9.91024196e-01 9.60065722e-01 9.54165608e-02
1.31484300e-01 6.03298843e-01 4.92080301e-01 -3.03607583e-01
-1.25009567e-01 -1.58738539e-01 -3.33774462e-02 1.03248417e-01
1.02939032e-01 -4.47737835e-02 -8.09914470e-01 5.74648261e-01
-1.85487330e+00 -5.36263168e-01 7.85737038e-02 2.14142275e+00
5.61566412e-01 2.52017498e-01 3.98124047e-02 -2.80854374e-01
1.27844334e+00 -2.02423446e-02 -6.66357994e-01 -3.26219410e-01
-2.29644924e-01 4.03264672e-01 6.14189386e-01 2.80484766e-01
-1.65015256e+00 1.11311269e+00 5.64233112e+00 9.94171143e-01
-6.63408160e-01 -4.21106592e-02 9.40765202e-01 4.26191896e-01
3.77452999e-01 -2.22556472e-01 -1.06611228e+00 3.70720804e-01
7.85314023e-01 4.80508655e-01 7.98872858e-02 1.21308851e+00
8.52274075e-02 -2.66179025e-01 -1.07617354e+00 8.69027019e-01
1.38965815e-01 -1.12402153e+00 -1.14132546e-01 -2.11992577e-01
7.55242169e-01 -3.59400094e-01 8.38308409e-02 3.99699867e-01
-2.67739734e-03 -1.12538886e+00 8.31980526e-01 6.73776269e-01
6.95867658e-01 -8.88913989e-01 1.08421588e+00 3.91869605e-01
-1.16582990e+00 -1.34271234e-01 -4.10278976e-01 5.46053112e-01
-1.63153708e-02 5.54262996e-01 -6.45413756e-01 5.85995495e-01
1.24086058e+00 1.23061143e-01 -7.04796076e-01 5.95260441e-01
-2.33161688e-01 3.48506242e-01 -5.58260791e-02 3.37901473e-01
4.32665646e-01 -2.86383003e-01 1.52353764e-01 1.27169919e+00
2.09373668e-01 -1.95656810e-02 2.02375621e-01 1.10397029e+00
-7.10656643e-02 1.23437308e-01 -1.34077027e-01 3.28218758e-01
3.06137085e-01 1.44118667e+00 -9.92027879e-01 -4.28778082e-01
-3.83069843e-01 1.14195728e+00 5.28869271e-01 2.43133396e-01
-8.51678550e-01 -3.67404133e-01 2.77580082e-01 -2.03612614e-02
7.45720565e-01 6.91695064e-02 -2.56440222e-01 -4.42439169e-01
1.55131280e-01 -5.86157322e-01 6.56683087e-01 -9.61757064e-01
-1.24784088e+00 6.93856716e-01 1.14459634e-01 -8.72543335e-01
-3.76457095e-01 -8.49562049e-01 -2.95466155e-01 9.32636857e-01
-1.19611537e+00 -1.43436933e+00 -5.22589326e-01 4.86530840e-01
5.27018070e-01 5.10973521e-02 6.05186284e-01 1.33820832e-01
-5.04227459e-01 6.69221699e-01 -8.56847242e-02 6.01869941e-01
7.18451262e-01 -1.52442122e+00 3.17384392e-01 7.70362496e-01
-2.10591868e-01 3.14812571e-01 4.42374706e-01 -7.57786155e-01
-5.73362947e-01 -1.34273374e+00 8.47247779e-01 -3.32087815e-01
3.87832880e-01 -2.03231603e-01 -9.46828663e-01 5.87730467e-01
-2.86276728e-01 4.47724253e-01 2.89831102e-01 -2.59508073e-01
-4.00592804e-01 5.43335080e-03 -1.37370515e+00 2.97610104e-01
7.91224718e-01 -3.99705112e-01 -7.41889596e-01 9.99907479e-02
8.61459255e-01 -6.81885839e-01 -1.01897240e+00 7.74070263e-01
5.08173823e-01 -1.10985732e+00 1.13756490e+00 -4.66943443e-01
5.82876086e-01 -3.30312937e-01 -2.43041351e-01 -9.97803032e-01
-2.25584745e-01 -4.84274141e-02 3.20311785e-01 1.44500756e+00
3.70546073e-01 -2.82120466e-01 1.27587736e+00 7.30118275e-01
-3.05918276e-01 -7.28580594e-01 -1.16592312e+00 -6.44662499e-01
2.54337311e-01 -4.17014301e-01 4.06320751e-01 5.43984532e-01
-6.14115059e-01 -1.21985778e-01 -1.11795723e-01 5.55571198e-01
5.59098899e-01 9.73124057e-02 6.62025928e-01 -1.10805845e+00
-2.66414821e-01 -3.96923929e-01 -7.15357661e-01 -8.43828857e-01
6.17665887e-01 -7.64740169e-01 4.08577591e-01 -1.26619077e+00
3.17545474e-01 -2.65998125e-01 -1.09083757e-01 4.09660637e-01
-4.36396122e-01 3.72647762e-01 -7.94265233e-03 5.28633408e-02
-5.94401360e-01 7.40478262e-02 1.11829078e+00 -1.84456393e-01
-1.81078285e-01 1.47529513e-01 -2.92070985e-01 9.08898354e-01
6.30134523e-01 -3.60867649e-01 -5.50475232e-02 -4.40221876e-02
-2.97441900e-01 2.45191619e-01 6.74825847e-01 -1.05030572e+00
3.85742992e-01 2.66565979e-01 7.60345340e-01 -7.60361075e-01
1.92898884e-01 -8.22786331e-01 8.39748383e-02 3.04681867e-01
-4.02940840e-01 -9.06965956e-02 5.78944907e-02 6.28435075e-01
-4.35669243e-01 -5.78693271e-01 1.13633633e+00 -2.46546954e-01
-1.23904228e+00 2.34887362e-01 1.93363264e-01 9.80415940e-02
1.07997298e+00 -5.50528467e-01 -5.39894029e-02 1.06981829e-01
-1.21752453e+00 8.07848126e-02 2.74354994e-01 2.49578267e-01
5.13540387e-01 -1.02806604e+00 -4.81335878e-01 1.98712781e-01
2.63529986e-01 4.41187680e-01 6.41906142e-01 6.40920758e-01
-5.32085717e-01 2.61556149e-01 -1.13445237e-01 -8.33507538e-01
-1.26023793e+00 6.18483841e-01 5.48927009e-01 -3.45284998e-01
-5.77640414e-01 9.50993538e-01 4.27058399e-01 -8.11716914e-01
1.78852871e-01 -5.29041469e-01 -2.87102699e-01 -9.40200314e-02
2.57714093e-01 2.47646898e-01 1.50447726e-01 -1.01612127e+00
-2.72990853e-01 8.84572387e-01 1.40833184e-01 2.38038301e-01
1.07101059e+00 3.51616852e-02 -1.48408711e-01 8.83944184e-02
1.40795517e+00 -2.33834580e-01 -1.59231997e+00 -4.33551222e-01
1.64305866e-01 -7.90247098e-02 -2.91827142e-01 -8.39654148e-01
-1.18555486e+00 4.28504884e-01 6.24354601e-01 -9.87574011e-02
1.13260937e+00 6.26266778e-01 6.79895699e-01 -1.01356998e-01
1.59165233e-01 -1.24385357e+00 1.42932832e-02 3.73892277e-01
5.88734925e-01 -1.34622407e+00 -3.63393277e-01 -6.43870234e-01
-7.84706235e-01 1.36063492e+00 7.52182066e-01 -1.54490948e-01
4.43386078e-01 1.02645889e-01 9.61386934e-02 -2.03316972e-01
6.65713400e-02 -1.92844272e-01 6.28329396e-01 7.80037820e-01
7.42798522e-02 2.97593921e-01 2.16844492e-02 1.07580149e+00
-3.57150167e-01 -2.70873427e-01 -3.71038401e-03 5.75014234e-01
-4.44056422e-01 -7.64189303e-01 -5.61326146e-01 2.37000436e-01
-3.43748391e-01 1.94903672e-01 -3.06358159e-01 1.18910050e+00
6.34817719e-01 8.57217610e-01 2.96653688e-01 -1.19148239e-01
6.82531416e-01 2.32636929e-01 3.47578824e-01 -4.10872757e-01
-5.46020806e-01 7.16080517e-02 -2.01783389e-01 -7.37072647e-01
-4.99370009e-01 -7.09433198e-01 -1.50051975e+00 1.39861345e-01
-2.02180952e-01 5.46001121e-02 5.95717192e-01 9.85814691e-01
3.80590931e-02 4.83710110e-01 1.90127090e-01 -1.00956857e+00
-2.47516319e-01 -7.71792114e-01 -7.31274724e-01 6.71814084e-01
-1.25203803e-01 -7.03007817e-01 -2.57809043e-01 2.57406592e-01] | [8.671414375305176, 0.018070267513394356] |
5388d5c6-1f40-415d-871f-9149bb9943a0 | difer-differentiable-automated-feature | 2010.08784 | null | https://arxiv.org/abs/2010.08784v3 | https://arxiv.org/pdf/2010.08784v3.pdf | DIFER: Differentiable Automated Feature Engineering | Feature engineering, a crucial step of machine learning, aims to extract useful features from raw data to improve data quality. In recent years, great efforts have been devoted to Automated Feature Engineering (AutoFE) to replace expensive human labor. However, existing methods are computationally demanding due to treating AutoFE as a coarse-grained black-box optimization problem over a discrete space. In this work, we propose an efficient gradient-based method called DIFER to perform differentiable automated feature engineering in a continuous vector space. DIFER selects potential features based on evolutionary algorithm and leverages an encoder-predictor-decoder controller to optimize existing features. We map features into the continuous vector space via the encoder, optimize the embedding along the gradient direction induced by the predicted score, and recover better features from the optimized embedding by the decoder. Extensive experiments on classification and regression datasets demonstrate that DIFER can significantly improve the performance of various machine learning algorithms and outperform current state-of-the-art AutoFE methods in terms of both efficiency and performance. | ['Yihua Huang', 'Chunfeng Yuan', 'Xu Guo', 'Zhuoer Xu', 'Guanghui Zhu'] | 2020-10-17 | null | null | null | null | ['automated-feature-engineering'] | ['methodology'] | [ 2.02437833e-01 -3.20452213e-01 -8.50062668e-02 -6.48386478e-01
-8.84828210e-01 -3.11376989e-01 3.34472358e-01 2.32648291e-02
-3.14053595e-01 5.29462337e-01 1.36107862e-01 1.08067520e-01
-2.10890874e-01 -7.81751752e-01 -6.87557399e-01 -4.61745173e-01
5.25791720e-02 1.31856143e-01 -2.34065294e-01 -1.35269001e-01
4.53624010e-01 4.31063414e-01 -1.57464123e+00 1.58399660e-02
9.89015460e-01 1.13804734e+00 -8.29619821e-03 4.01862144e-01
1.93767177e-04 3.45050931e-01 -3.00116986e-01 -5.09008765e-01
3.30941141e-01 -5.21686137e-01 -5.19665062e-01 2.21427828e-02
-5.87735958e-02 -1.62591338e-01 -3.13695014e-01 9.89883125e-01
4.23398584e-01 5.19615300e-02 5.85949421e-01 -1.23085570e+00
-7.44746029e-01 2.17341974e-01 -2.71850944e-01 -1.95071578e-01
1.82725072e-01 1.30097315e-01 1.42902720e+00 -1.40570641e+00
4.65177923e-01 9.74018037e-01 5.87295115e-01 3.50239366e-01
-1.42341447e+00 -5.65501094e-01 2.01498419e-02 3.09437394e-01
-1.56481314e+00 -4.26246434e-01 9.05017674e-01 -5.56922019e-01
1.04308510e+00 1.59851089e-01 8.85296106e-01 5.32386720e-01
2.48714373e-01 8.87967169e-01 5.19649327e-01 -3.39380145e-01
3.34894985e-01 2.07074150e-01 3.64477560e-02 1.28704178e+00
7.80086368e-02 4.02160943e-01 -8.71305287e-01 -2.13760078e-01
5.24033546e-01 2.38158628e-01 -2.01847002e-01 -5.46821415e-01
-1.10794842e+00 1.22100961e+00 5.20347714e-01 -9.30523407e-03
-6.28192544e-01 1.10979527e-01 1.98484346e-01 4.61603701e-01
3.92543048e-01 8.91312003e-01 -6.48426831e-01 -4.77665573e-01
-8.25730085e-01 1.49907932e-01 7.22985148e-01 5.98497093e-01
1.13669991e+00 3.52879204e-02 -1.70040205e-01 8.50591838e-01
4.04666126e-01 3.05343956e-01 5.94575167e-01 -5.88898063e-01
2.72306561e-01 1.02127564e+00 -1.25229195e-01 -1.04138815e+00
-2.26731792e-01 -5.90640604e-01 -6.02984190e-01 5.71302958e-02
-1.51819065e-01 -3.01465750e-01 -6.07232630e-01 1.38992000e+00
4.53393996e-01 2.84344703e-03 -7.94341937e-02 9.23026502e-01
3.03105146e-01 7.26851881e-01 -3.53718877e-01 6.67056069e-02
8.92860472e-01 -1.01765740e+00 -4.33191538e-01 -1.61240131e-01
6.81960166e-01 -4.70208317e-01 1.18260026e+00 3.45160723e-01
-7.37864316e-01 -3.41285855e-01 -1.31839395e+00 -8.22997764e-02
-2.50963926e-01 5.50051868e-01 7.22580194e-01 4.22897339e-01
-5.42258501e-01 9.90981638e-01 -1.01751328e+00 3.04083705e-01
7.12743640e-01 6.89567924e-01 -4.67803001e-01 2.28374661e-03
-9.69717979e-01 6.72401130e-01 3.13479185e-01 1.61937743e-01
-7.38636374e-01 -9.17676032e-01 -1.03696918e+00 3.13981116e-01
3.95556629e-01 -6.98721945e-01 1.09313822e+00 -9.02040005e-01
-1.87909734e+00 2.40794763e-01 -2.05353230e-01 -3.11030477e-01
2.40681320e-01 -5.57864785e-01 -2.91605979e-01 -2.94098407e-01
-7.58931190e-02 4.57600802e-01 1.07545936e+00 -5.86994827e-01
-7.35297024e-01 -1.97010785e-01 -4.76437986e-01 -7.78572168e-03
-6.14923954e-01 -1.87877685e-01 -3.05909723e-01 -5.92916429e-01
-1.17177621e-01 -8.90243113e-01 -2.25461438e-01 2.19915777e-01
-4.58756983e-01 -2.26302132e-01 6.66940928e-01 -5.68586171e-01
1.46896315e+00 -2.40238643e+00 4.03629482e-01 5.06464422e-01
2.80142635e-01 2.92377144e-01 -1.07008465e-01 4.29866135e-01
2.01088101e-01 -8.72367099e-02 -3.23477119e-01 -1.53286457e-01
2.46557012e-01 -4.16986793e-02 -1.86368704e-01 5.48630416e-01
7.12027907e-01 1.07487118e+00 -8.65988076e-01 -2.30049387e-01
1.23251013e-01 6.49683893e-01 -1.02774680e+00 4.65005189e-01
-1.48278654e-01 1.60214186e-01 -7.78863251e-01 6.19221687e-01
1.39110729e-01 -4.42264736e-01 -1.90090716e-01 -1.71776265e-01
-2.32266009e-01 2.51074106e-01 -9.91263211e-01 1.70073545e+00
-5.36958933e-01 6.93750560e-01 -4.40242738e-01 -9.44312632e-01
1.13578773e+00 -1.18560292e-01 5.46128809e-01 -4.63100314e-01
1.52781889e-01 3.84010315e-01 -1.23355687e-02 -4.06078279e-01
3.60508442e-01 3.73764075e-02 -5.99540770e-02 2.74900883e-01
9.36816856e-02 -1.53234392e-01 3.45461965e-02 -2.99631923e-01
1.19342303e+00 1.42887145e-01 4.36719477e-01 1.24818332e-01
5.49630821e-01 3.02766114e-02 7.07536280e-01 3.66886914e-01
8.80861953e-02 4.66646165e-01 3.47281218e-01 -5.03798723e-01
-8.85649621e-01 -7.94403970e-01 -5.59390001e-02 1.07893682e+00
-1.87695742e-01 -6.82628155e-01 -4.49425250e-01 -9.25126076e-01
3.46745312e-01 7.00583279e-01 -5.45829058e-01 -7.54454255e-01
-3.74433130e-01 -6.56746268e-01 1.04131587e-01 4.70832229e-01
4.62624937e-01 -8.27159703e-01 -7.26116359e-01 5.65869093e-01
2.44717464e-01 -6.39641643e-01 -7.03720450e-01 2.94463933e-01
-6.37618303e-01 -8.40922534e-01 -3.62946302e-01 -6.80142462e-01
7.21509039e-01 -1.12572992e-02 8.07782829e-01 7.24052861e-02
-8.13512683e-01 1.20690009e-02 -3.95254254e-01 -2.15703830e-01
-3.43377665e-02 2.13099614e-01 -1.78611666e-01 4.21748936e-01
5.15319884e-01 -2.70882219e-01 -6.26026511e-01 2.24409088e-01
-6.90301299e-01 9.11498889e-02 8.80412102e-01 1.43873048e+00
9.45520878e-01 -5.09799309e-02 6.37802839e-01 -6.82350278e-01
6.31494999e-01 -3.53324831e-01 -9.07618165e-01 2.60751963e-01
-9.81474340e-01 7.92151451e-01 8.42431247e-01 -1.57465413e-01
-6.46516085e-01 5.28441727e-01 -2.18216613e-01 -3.50371420e-01
3.66696417e-01 6.69052243e-01 -1.71128467e-01 -2.24852130e-01
4.50934112e-01 4.93853301e-01 6.16909042e-02 -4.51207131e-01
3.14664513e-01 7.79374957e-01 2.05864787e-01 -2.73889065e-01
8.92466247e-01 2.03682017e-02 5.81059679e-02 -5.68909943e-01
-8.89502168e-01 -1.29929021e-01 -5.49295843e-01 7.95597211e-02
4.80805695e-01 -7.00521231e-01 -5.77130198e-01 1.61241293e-01
-6.54281199e-01 8.33039638e-03 -4.56055969e-01 5.88267326e-01
-5.66912532e-01 -1.29319057e-01 -1.47351965e-01 -3.31738681e-01
-6.65977538e-01 -1.29346418e+00 1.14380205e+00 2.89641231e-01
-2.04572082e-01 -7.25775778e-01 1.07301757e-01 -1.68355763e-01
4.94305879e-01 1.10497005e-01 1.02578723e+00 -5.81747413e-01
-4.98095095e-01 -5.91282725e-01 -7.43382871e-02 4.95589763e-01
2.69617975e-01 1.42800093e-01 -6.32240832e-01 -3.29014540e-01
-3.24863911e-01 -3.13924760e-01 7.61880517e-01 1.26885369e-01
1.36696351e+00 -3.59310359e-01 -3.43906999e-01 9.89405513e-01
1.40587389e+00 8.79974291e-02 1.92729563e-01 4.02618289e-01
4.55661893e-01 1.30915478e-01 8.83428991e-01 8.69603693e-01
2.25027442e-01 6.03418171e-01 2.66857415e-01 3.15449610e-02
1.20395780e-01 -5.11482120e-01 4.04648274e-01 6.43241346e-01
4.07064170e-01 3.27640951e-01 -5.90343177e-01 4.54102576e-01
-1.76561522e+00 -5.69632947e-01 5.39399981e-01 2.10823798e+00
1.05341101e+00 3.06088757e-02 -8.79334658e-02 1.50629923e-01
4.59803194e-01 -2.10327595e-01 -1.03245199e+00 -2.82640189e-01
2.47122124e-01 2.48793483e-01 1.99349150e-01 2.85328865e-01
-1.08089232e+00 9.34732318e-01 6.09922123e+00 5.13796687e-01
-1.37682438e+00 -2.90238708e-01 3.99593800e-01 -3.75222594e-01
-3.16502273e-01 -1.59985021e-01 -7.87721634e-01 4.81828272e-01
7.94172704e-01 -4.62221175e-01 7.73132205e-01 1.10984576e+00
-6.08764589e-02 3.30713958e-01 -1.37157691e+00 1.11506879e+00
-8.46581310e-02 -1.39115238e+00 -2.21132115e-01 1.58428773e-02
6.26111090e-01 -2.87808906e-02 3.17134112e-01 4.09855664e-01
1.98970661e-01 -1.10248601e+00 3.69871527e-01 6.07618392e-01
8.25095773e-01 -1.04061389e+00 5.70834994e-01 1.49719894e-01
-1.01385701e+00 -2.01844230e-01 -4.55621034e-01 2.28766859e-01
-2.87356246e-02 7.49637187e-01 -1.03768444e+00 2.76788890e-01
3.80704761e-01 9.60850239e-01 -6.22314095e-01 1.10777199e+00
-1.88844174e-01 5.99186003e-01 -2.56070971e-01 -3.80938977e-01
1.93483338e-01 -2.27752373e-01 3.50196064e-01 9.71272945e-01
4.72140342e-01 -1.88233659e-01 8.40953067e-02 8.86404455e-01
-2.63153583e-01 1.85021356e-01 -3.48044693e-01 -5.10805726e-01
6.12570465e-01 1.19586194e+00 -5.99949099e-02 -2.90094223e-02
-4.71334964e-01 1.10405409e+00 5.37605643e-01 1.70624373e-03
-6.52600944e-01 -1.07526708e+00 9.31919456e-01 -2.92165846e-01
6.56972885e-01 -1.32209882e-01 -1.73815683e-01 -1.24522483e+00
1.11782484e-01 -8.27062309e-01 1.07887082e-01 -2.61382431e-01
-1.12203789e+00 6.64644420e-01 -5.75384796e-01 -1.36788857e+00
-5.84454417e-01 -5.31889617e-01 -2.78560758e-01 7.30005085e-01
-1.59471762e+00 -6.65585041e-01 -1.59972176e-01 4.75065649e-01
5.41833520e-01 -4.99691784e-01 8.66672754e-01 1.54428288e-01
-7.97207832e-01 9.20245767e-01 5.36767185e-01 1.68545060e-02
4.47292775e-01 -1.08384228e+00 5.14355302e-01 6.06955945e-01
2.60835826e-01 5.04391491e-01 5.14611900e-01 -3.44533533e-01
-2.05000019e+00 -1.24107218e+00 8.82739961e-01 7.54854456e-02
6.20095134e-01 -3.87029707e-01 -7.59284317e-01 3.73024732e-01
-2.79375702e-01 3.58564794e-01 9.21891153e-01 3.74129936e-02
-3.00402194e-01 -2.49389246e-01 -1.15189457e+00 4.98529404e-01
8.23356390e-01 -4.92845327e-01 -4.35642362e-01 1.11974686e-01
5.91372192e-01 -1.59668267e-01 -1.07442701e+00 2.42821962e-01
7.65845001e-01 -5.77242970e-01 7.40202546e-01 -7.96132863e-01
5.11416733e-01 -2.84337580e-01 -2.22007006e-01 -1.86294234e+00
-4.15056318e-01 -7.16752946e-01 -3.43693703e-01 9.45007443e-01
5.45834959e-01 -7.60617018e-01 7.32151449e-01 6.21596336e-01
-7.53956288e-02 -1.25240159e+00 -8.04584742e-01 -6.41094923e-01
-3.29303384e-01 -2.41369784e-01 1.02795327e+00 5.24325490e-01
6.51152432e-02 3.89652729e-01 -3.38710159e-01 -5.91141880e-02
4.24813420e-01 4.45155919e-01 7.67887056e-01 -1.25188410e+00
-5.41176140e-01 -5.20557106e-01 -6.79159343e-01 -1.06129014e+00
2.82925934e-01 -1.03799427e+00 1.54551864e-01 -1.25417292e+00
4.47065681e-02 -2.54883468e-01 -3.35510761e-01 6.78875983e-01
-3.99662435e-01 2.93914825e-02 8.82413909e-02 -2.88148504e-02
-4.78084058e-01 1.16454637e+00 9.67821598e-01 -1.51069179e-01
-2.54449338e-01 -1.19021125e-02 -8.89433622e-01 4.19309735e-01
6.52757049e-01 -6.84575260e-01 -1.11761525e-01 -3.97853732e-01
3.03674787e-01 -1.04047388e-01 9.35287401e-03 -8.44405234e-01
6.01023696e-02 -1.80563405e-01 6.63183570e-01 -2.57339597e-01
4.39994156e-01 -9.26689208e-01 -8.05508345e-02 5.41960061e-01
-5.31258583e-01 1.77084059e-01 -1.32742435e-01 5.15147686e-01
-4.86361384e-01 -1.73202440e-01 5.51331162e-01 4.06353563e-01
-5.83436549e-01 4.03962135e-01 6.85030594e-02 -1.20161593e-01
1.00321472e+00 8.84521902e-02 -4.65947501e-02 -9.96093228e-02
-4.14530009e-01 2.25862443e-01 3.15189004e-01 4.13811326e-01
1.00979841e+00 -1.42027938e+00 -7.81437278e-01 7.28801906e-01
1.88380554e-01 -2.30097502e-01 -2.30956182e-01 5.80196261e-01
-2.66123533e-01 4.11182255e-01 -3.89888063e-02 -4.01116967e-01
-1.15013242e+00 2.66041011e-01 2.78549820e-01 -2.00131461e-01
-6.84395492e-01 9.47888851e-01 -3.07254940e-01 -3.32414955e-01
6.50983453e-02 -1.11690182e-02 7.41398707e-02 -1.43085539e-01
5.46374023e-01 4.07166749e-01 2.64549434e-01 -3.44201505e-01
-3.72136950e-01 5.26653707e-01 -3.03811282e-01 6.04838133e-02
1.77404130e+00 1.72854856e-01 1.66193873e-01 1.21782877e-01
1.74280417e+00 -2.98862666e-01 -1.54570580e+00 -4.08450305e-01
1.31917045e-01 -8.20004046e-01 4.06217784e-01 -7.76012540e-01
-1.17376649e+00 7.43200064e-01 6.04880095e-01 -1.51160508e-01
1.30825543e+00 -2.55490690e-01 7.65654027e-01 7.08894074e-01
2.07660675e-01 -1.08637917e+00 1.37805253e-01 4.87163901e-01
8.83420289e-01 -1.25913870e+00 -3.93689889e-03 -1.38025075e-01
-8.18216801e-01 1.30401695e+00 4.52303410e-01 -3.56858820e-01
8.11828196e-01 1.18876196e-01 -2.94778168e-01 -1.09245390e-01
-9.94639575e-01 -4.05191928e-02 6.27488971e-01 1.74126804e-01
3.82149935e-01 -6.91031888e-02 -1.59732148e-01 8.26235592e-01
-4.89636987e-01 1.86888963e-01 9.34865251e-02 1.10950983e+00
-5.48523784e-01 -1.18398011e+00 -7.75553882e-02 9.33176637e-01
-3.15482050e-01 -5.56180738e-02 -2.70674735e-01 4.71788734e-01
-2.00402349e-01 8.05759311e-01 1.79821104e-02 -7.08843470e-01
3.87825608e-01 1.36212975e-01 3.01465094e-01 -5.82128346e-01
-6.67532623e-01 -1.17110372e-01 -1.50319040e-01 -6.93302035e-01
3.26820314e-01 -8.33321154e-01 -1.30529094e+00 3.46175171e-02
-6.06259823e-01 3.16598237e-01 1.02199960e+00 9.99195158e-01
9.33427930e-01 6.12472177e-01 1.27065146e+00 -6.22081280e-01
-1.14667833e+00 -5.43519497e-01 -4.30306107e-01 2.22250193e-01
5.88434100e-01 -7.11596310e-01 -1.89828187e-01 -1.70343101e-01] | [9.364721298217773, 3.2788546085357666] |
645414a5-0c0b-44b6-918f-5773815300e8 | on-bottleneck-features-for-text-dependent | 2005.07383 | null | https://arxiv.org/abs/2005.07383v2 | https://arxiv.org/pdf/2005.07383v2.pdf | On Bottleneck Features for Text-Dependent Speaker Verification Using X-vectors | Applying x-vectors for speaker verification has recently attracted great interest, with the focus being on text-independent speaker verification. In this paper, we study x-vectors for text-dependent speaker verification (TD-SV), which remains unexplored. We further investigate the impact of the different bottleneck (BN) features on the performance of x-vectors, including the recently-introduced time-contrastive-learning (TCL) BN features and phone-discriminant BN features. TCL is a weakly supervised learning approach that constructs training data by uniformly partitioning each utterance into a predefined number of segments and then assigning each segment a class label depending on their position in the utterance. We also compare TD-SV performance for different modeling techniques, including the Gaussian mixture models-universal background model (GMM-UBM), i-vector, and x-vector. Experiments are conducted on the RedDots 2016 challenge database. It is found that the type of features has a marginal impact on the performance of x-vectors with the TCL BN feature achieving the lowest equal error rate, while the impact of features is significant for i-vector and GMM-UBM. The fusion of x-vector and i-vector systems gives a large gain in performance. The GMM-UBM technique shows its advantage for TD-SV using short utterances. | ['Zheng-Hua Tan', 'Achintya Kumar Sarkar'] | 2020-05-15 | null | null | null | null | ['text-independent-speaker-verification', 'text-dependent-speaker-verification'] | ['speech', 'speech'] | [-1.10655345e-01 -4.96097147e-01 -1.85755387e-01 -6.77914619e-01
-1.25794518e+00 -5.01646221e-01 9.05305624e-01 1.00459307e-01
-4.23050433e-01 1.86189935e-01 5.40732145e-01 -8.33739340e-01
-5.03421761e-02 9.29668397e-02 -1.95300132e-01 -1.13164330e+00
-7.33862072e-02 2.19215155e-01 6.20605014e-02 -2.54721135e-01
2.46122509e-01 5.24763465e-01 -1.52606916e+00 2.01198068e-02
3.69403720e-01 9.11543310e-01 -7.20765814e-02 9.51574802e-01
-8.95831883e-02 5.61282575e-01 -9.42452788e-01 -2.21508920e-01
1.59775198e-01 -4.86018687e-01 -6.68760598e-01 -1.06488213e-01
4.00739431e-01 -1.28385186e-01 -1.14964262e-01 6.04570448e-01
9.32656467e-01 2.15419233e-01 7.57652462e-01 -1.40670764e+00
2.07290221e-02 7.23444045e-01 -4.86910224e-01 3.16706181e-01
3.69089127e-01 -1.25744753e-02 8.22277904e-01 -8.55358720e-01
3.62469196e-01 1.50007057e+00 7.78871417e-01 6.73843563e-01
-1.35228288e+00 -7.21467912e-01 6.57185838e-02 4.46717530e-01
-1.57038021e+00 -9.14472818e-01 8.02049220e-01 -5.10893226e-01
1.01833916e+00 6.35582745e-01 -2.12166309e-02 1.08587503e+00
7.75572658e-02 8.31790268e-01 1.15534878e+00 -8.80051613e-01
4.08198088e-01 3.78076643e-01 7.37932086e-01 1.62034914e-01
-4.33091760e-01 3.95927221e-01 -7.77399540e-01 -4.24000800e-01
-8.59780014e-02 -4.15507376e-01 -4.24384594e-01 -2.63214540e-02
-8.53196621e-01 8.96638095e-01 -2.05475166e-01 5.95966876e-01
-9.96626988e-02 -1.56792536e-01 7.88716733e-01 4.23642069e-01
4.81216937e-01 -2.65006691e-01 -6.20852232e-01 -3.83274794e-01
-1.33770919e+00 1.52388394e-01 6.95805669e-01 6.86138868e-01
4.57132548e-01 4.42646742e-01 -4.81890023e-01 1.03078806e+00
5.43118954e-01 4.66350526e-01 7.92411685e-01 -3.63980591e-01
4.38677192e-01 4.03243629e-03 -7.18828365e-02 -3.41000855e-01
-3.74047875e-01 -3.22162390e-01 -4.60284054e-01 1.47788182e-01
5.53612590e-01 -2.25231677e-01 -9.53346550e-01 1.92245519e+00
3.72994006e-01 9.58546773e-02 1.75711229e-01 6.65149391e-01
7.84543395e-01 7.97510862e-01 8.87269825e-02 -3.66408497e-01
1.23798859e+00 -7.54318833e-01 -8.07994545e-01 1.32983059e-01
7.87117481e-01 -9.23324645e-01 7.84792721e-01 2.63790876e-01
-6.62851334e-01 -6.88805819e-01 -9.83222604e-01 3.30695212e-01
-3.62737894e-01 1.47625759e-01 -1.17020443e-01 1.44175720e+00
-1.22756648e+00 1.69239432e-01 -5.96206069e-01 -3.33765566e-01
-1.84314679e-02 4.53380495e-01 -3.75908434e-01 -9.29066679e-04
-1.06727076e+00 9.73791182e-01 9.83041972e-02 -3.73048447e-02
-8.82422924e-01 -7.04211712e-01 -9.35488224e-01 6.43750951e-02
-9.72782001e-02 7.70328864e-02 1.34936166e+00 -6.59965813e-01
-1.88137424e+00 5.51496327e-01 -7.32680976e-01 -5.14583230e-01
3.51018608e-01 1.96231544e-01 -7.34931469e-01 -1.24598376e-01
-1.98071793e-01 4.89546865e-01 1.12950051e+00 -1.36616135e+00
-4.61737275e-01 -5.88415027e-01 -5.34902513e-01 -1.08598046e-01
-2.51819372e-01 5.70230484e-01 -1.58082560e-01 -4.36492175e-01
9.79038402e-02 -1.05633259e+00 1.41378060e-01 -7.88679361e-01
-6.04286790e-01 -5.77343404e-01 1.27859783e+00 -1.03407264e+00
1.25610018e+00 -2.58498883e+00 -2.77563259e-02 2.61045337e-01
-2.79989630e-01 3.51270050e-01 -6.45502955e-02 4.54282016e-01
-3.12842846e-01 -2.22158842e-02 3.10303569e-02 -8.92548501e-01
2.43297935e-01 9.68744606e-02 -2.64024407e-01 6.32100165e-01
-1.42791301e-01 5.57757080e-01 -2.43827760e-01 -4.89901394e-01
4.14602965e-01 7.83884585e-01 -3.64640594e-01 1.07131295e-01
1.63213968e-01 3.57433319e-01 2.13973939e-01 4.46699947e-01
8.74310672e-01 3.26116174e-01 8.96816775e-02 -6.58574849e-02
-3.44635338e-01 5.29772818e-01 -1.09621072e+00 1.17321539e+00
-3.77994925e-01 9.00035977e-01 3.35491776e-01 -8.11471343e-01
9.48733628e-01 8.61638248e-01 1.97305992e-01 -4.97249573e-01
2.44486302e-01 9.08595994e-02 1.92589432e-01 -1.99942857e-01
4.17349219e-01 -4.53384727e-01 2.07154155e-02 3.80466342e-01
1.69106603e-01 1.62015498e-01 3.26350741e-02 2.26700366e-01
6.79481685e-01 -2.73747623e-01 3.37932981e-03 -3.89454156e-01
7.61261463e-01 -4.69966501e-01 4.29363221e-01 4.80377883e-01
-5.23540139e-01 5.82151532e-01 3.57774258e-01 3.15865159e-01
-5.61297238e-01 -9.35420752e-01 -4.64735508e-01 1.30704105e+00
-1.15804575e-01 -5.11777103e-01 -7.05073595e-01 -6.50754571e-01
-3.04804705e-02 1.22083354e+00 -5.20442665e-01 -1.60033986e-01
-4.58926290e-01 -7.62363970e-01 8.09913635e-01 3.83315921e-01
-6.68444857e-02 -7.57815421e-01 -2.21819237e-01 1.32777363e-01
-2.62532383e-01 -9.73972023e-01 -8.10661793e-01 4.90721226e-01
-6.83037102e-01 -3.76574427e-01 -7.93300569e-01 -6.95081174e-01
3.52566749e-01 9.22829285e-02 5.98766088e-01 -4.41512704e-01
4.45457958e-02 5.95148206e-01 -4.52388436e-01 -4.64465350e-01
-7.99423039e-01 -1.20102696e-01 3.29912841e-01 4.61467385e-01
5.94132960e-01 8.37431755e-03 -8.19740966e-02 5.30797243e-01
-6.25774384e-01 -5.78752220e-01 1.11401491e-01 9.31182325e-01
6.14237748e-02 -3.05101126e-02 5.82073271e-01 -3.16755503e-01
4.37140703e-01 -1.61658198e-01 -4.10447955e-01 6.00970350e-03
-6.14797235e-01 5.64209297e-02 1.48043841e-01 -5.56365311e-01
-1.01771033e+00 -1.61707565e-01 -5.50841808e-01 -4.96502221e-01
-2.49448866e-01 5.20361662e-01 -5.19546032e-01 -6.48070350e-02
4.66024399e-01 4.99665201e-01 2.42294312e-01 -6.23035550e-01
1.39339238e-01 1.08228815e+00 2.61326551e-01 -2.56995350e-01
4.92950886e-01 9.57858264e-02 -4.77219492e-01 -1.46048045e+00
-1.97703555e-01 -9.59037364e-01 -4.14184123e-01 -2.84925550e-01
8.85794818e-01 -7.61387765e-01 -6.20682597e-01 6.11453831e-01
-9.19225216e-01 -2.81839579e-01 -1.38558403e-01 7.37165630e-01
-9.13286880e-02 5.66028237e-01 -6.19873464e-01 -1.11869872e+00
-2.84490258e-01 -1.54651785e+00 1.14822304e+00 -5.98444529e-02
-4.41532165e-01 -9.22123253e-01 1.59713626e-01 5.69027364e-01
6.55778110e-01 -4.27738398e-01 8.47362280e-01 -1.09587669e+00
-5.50707132e-02 -3.14634383e-01 3.90768945e-02 6.67011857e-01
3.42616178e-02 -3.19530368e-02 -1.41557682e+00 -6.60525501e-01
2.66447663e-01 6.87885731e-02 8.29445899e-01 8.11497867e-01
6.54611051e-01 -1.37530982e-01 -3.26120079e-01 2.21470907e-01
1.09205103e+00 3.04115385e-01 4.20555800e-01 6.49491772e-02
4.16382462e-01 6.06653154e-01 4.65230763e-01 3.84720474e-01
3.67059737e-01 1.14615762e+00 -9.77244787e-03 -1.28078498e-02
-1.72426239e-01 5.75165823e-02 8.88152778e-01 1.00634241e+00
2.26936549e-01 -3.02320272e-01 -9.91870046e-01 5.61205924e-01
-1.23234010e+00 -1.09553409e+00 -9.41258743e-02 2.47655940e+00
5.66564798e-01 -1.51280258e-02 4.95936513e-01 9.48008955e-01
8.14785361e-01 1.36046425e-01 -4.95188385e-02 -4.64421928e-01
-2.00832129e-01 1.70467254e-02 3.03371519e-01 9.58653092e-01
-8.97130728e-01 4.51602191e-01 5.94810629e+00 1.00164044e+00
-1.70851099e+00 5.00659645e-01 4.86613780e-01 -1.64709836e-02
-2.43195165e-02 -4.28769067e-02 -1.36043119e+00 6.05521142e-01
1.50050461e+00 4.25298773e-02 -1.23765089e-01 6.68036938e-01
3.91997486e-01 -2.94007473e-02 -1.08992851e+00 1.04655361e+00
4.61538553e-01 -8.88749301e-01 -3.43833923e-01 3.29291821e-01
3.66082817e-01 2.30765626e-01 1.81397483e-01 5.35554349e-01
-1.66978434e-01 -6.08706892e-01 1.05502760e+00 -1.25513300e-01
7.60805249e-01 -7.11931407e-01 7.83267915e-01 1.75851479e-01
-1.05793202e+00 -1.46424204e-01 5.44271469e-02 3.99049789e-01
2.64078200e-01 3.86598676e-01 -1.16655135e+00 4.91837412e-01
4.88898516e-01 1.65689126e-01 -4.32754397e-01 9.05971467e-01
3.66004050e-01 1.37155068e+00 -3.01184148e-01 4.26360127e-03
1.83661774e-01 5.45607656e-02 7.77667105e-01 1.68987143e+00
2.08141670e-01 -3.53572696e-01 -3.53913456e-02 3.24891329e-01
4.31339473e-01 2.67800570e-01 -1.15660228e-01 2.44131923e-01
4.05303657e-01 8.47112298e-01 -5.09929419e-01 -4.46723759e-01
-2.95275778e-01 6.72108948e-01 -2.01961800e-01 4.64765370e-01
-7.24539816e-01 -2.42386028e-01 6.64796472e-01 8.62082839e-02
7.17552423e-01 -1.82103902e-01 -1.13515310e-01 -7.91382432e-01
-1.71815440e-01 -9.77100372e-01 3.44397634e-01 -2.83567548e-01
-9.71281230e-01 7.80796587e-01 7.82748684e-02 -1.22467017e+00
-4.98853952e-01 -4.38804060e-01 -6.39967084e-01 1.19732034e+00
-1.45507634e+00 -1.05826175e+00 1.52643397e-01 6.08177245e-01
5.82175374e-01 -4.37174380e-01 8.20341229e-01 4.53486353e-01
-7.10996032e-01 1.13697195e+00 4.06756818e-01 1.56845599e-01
7.58376122e-01 -1.15486610e+00 2.62418002e-01 7.92287886e-01
2.63373435e-01 6.23330116e-01 9.06493545e-01 -2.54224002e-01
-1.23644793e+00 -9.31354046e-01 1.24867380e+00 -4.45206702e-01
-1.40422874e-03 -5.54755449e-01 -8.39862466e-01 5.92439175e-01
2.11609796e-01 -2.46455371e-01 1.01837170e+00 3.10774863e-01
-6.46872997e-01 -1.87780932e-01 -1.22400057e+00 1.59193203e-01
1.21240623e-01 -8.60784352e-01 -3.56167197e-01 1.12651981e-01
5.89264154e-01 -3.47156897e-02 -7.05567896e-01 2.25397676e-01
4.79800701e-01 -1.07609296e+00 8.15555871e-01 -1.87720940e-01
-4.12994027e-01 -2.96855539e-01 -5.09370625e-01 -1.21503735e+00
-1.71963096e-01 -5.96175909e-01 1.33960992e-01 1.65682352e+00
7.20515847e-01 -7.32525647e-01 7.70143747e-01 4.75716889e-01
-5.33949770e-02 -3.37359995e-01 -1.49105895e+00 -1.04031408e+00
1.80154573e-03 -8.12277555e-01 6.57518029e-01 8.23180497e-01
1.66464284e-01 6.03873432e-01 -3.55459958e-01 2.42710754e-01
4.84787434e-01 -5.23751564e-02 8.17968249e-01 -9.26366270e-01
-2.93492377e-01 -5.37508786e-01 -6.61863089e-01 -1.05231571e+00
2.56611437e-01 -8.87020230e-01 1.36226192e-01 -8.87702882e-01
-5.23861907e-02 -4.21185553e-01 -2.83391446e-01 2.66060621e-01
-2.35111371e-01 -4.08459781e-03 3.54069531e-01 -5.34389094e-02
-2.15059087e-01 6.38697743e-01 3.73656899e-01 -2.66878545e-01
-4.20314848e-01 4.52107519e-01 -1.34493008e-01 1.70030132e-01
7.13165462e-01 -3.93183649e-01 -1.92933068e-01 -3.58532532e-03
-9.28750098e-01 3.58572304e-01 7.57088698e-03 -8.45624030e-01
1.81850284e-01 4.25881207e-01 7.06714392e-02 -9.06809509e-01
7.39629388e-01 -5.96240640e-01 -3.77331935e-02 5.37447572e-01
-2.87018627e-01 -6.50247419e-03 5.29867887e-01 2.98518360e-01
-4.09223199e-01 -1.51724681e-01 9.76950943e-01 5.92109144e-01
-3.48887175e-01 -6.52549043e-02 -6.54704750e-01 -3.39455128e-01
7.10755587e-01 -3.23272824e-01 2.44110480e-01 -6.61650956e-01
-6.05243206e-01 -2.50952572e-01 -1.65379383e-02 6.16060019e-01
4.03657615e-01 -1.13618481e+00 -1.08598375e+00 5.46761453e-01
1.33952811e-01 -7.48669624e-01 3.74441504e-01 9.02377784e-01
7.00210929e-02 7.77365685e-01 1.39785752e-01 -1.14365637e+00
-2.06360006e+00 4.36887175e-01 2.84725994e-01 -4.98766303e-02
-1.98856056e-01 9.43201363e-01 6.58925856e-03 -7.11343586e-01
5.82845688e-01 -2.56444782e-01 -1.18334249e-01 2.10248336e-01
6.06007278e-01 5.04873514e-01 4.47054863e-01 -1.40778375e+00
-5.96661448e-01 3.66769314e-01 -3.17216903e-01 -6.57467961e-01
9.26681817e-01 -1.87758178e-01 5.87375350e-02 6.55086219e-01
1.52738464e+00 2.11644202e-01 -6.87046766e-01 -2.31136799e-01
1.75543800e-01 -2.10626870e-01 4.02778417e-01 -6.68095350e-01
-7.83754349e-01 9.31215525e-01 9.29871023e-01 2.34076068e-01
8.00705552e-01 -8.04970274e-04 5.24604499e-01 -3.32589924e-01
1.50832981e-01 -6.58847034e-01 -4.21802610e-01 4.50919628e-01
7.84396052e-01 -1.12112868e+00 -2.95629799e-01 -1.16619296e-01
-6.37357831e-01 6.82576716e-01 2.19802886e-01 7.28048861e-01
9.86276865e-01 2.98401117e-01 6.00132525e-01 1.81285501e-01
-5.75993776e-01 3.36398147e-02 3.30823153e-01 5.24413049e-01
5.19680917e-01 3.12767684e-01 -1.52490754e-02 3.95472258e-01
-3.63218606e-01 -4.24934864e-01 4.56115380e-02 8.23738337e-01
-7.14324638e-02 -1.36185038e+00 -9.12734747e-01 3.42510968e-01
-3.62049133e-01 -2.95860376e-02 -1.98053613e-01 5.40048778e-01
-9.79590416e-02 1.39221621e+00 -1.07229680e-01 -5.60864985e-01
1.34816214e-01 6.23554528e-01 3.08167338e-01 -2.77789921e-01
-7.72119939e-01 3.60247076e-01 1.02513567e-01 -4.68574688e-02
-1.93302885e-01 -1.17778528e+00 -8.83994758e-01 -3.51286381e-01
-9.32426810e-01 5.02983212e-01 1.31358242e+00 8.92377675e-01
3.17131966e-01 2.29348823e-01 6.94800198e-01 -9.83228624e-01
-9.29420710e-01 -1.40341485e+00 -5.89367628e-01 2.93547511e-01
8.09141695e-01 -4.50541973e-01 -7.58720994e-01 -1.30180619e-03] | [14.352425575256348, 6.109110355377197] |
fa449b54-9ec0-43f2-a126-7d9e9e2e653b | toward-a-deep-neural-approach-for-knowledge | 1606.07211 | null | http://arxiv.org/abs/1606.07211v1 | http://arxiv.org/pdf/1606.07211v1.pdf | Toward a Deep Neural Approach for Knowledge-Based IR | This paper tackles the problem of the semantic gap between a document and a
query within an ad-hoc information retrieval task. In this context, knowledge
bases (KBs) have already been acknowledged as valuable means since they allow
the representation of explicit relations between entities. However, they do not
necessarily represent implicit relations that could be hidden in a corpora.
This latter issue is tackled by recent works dealing with deep representation
learn ing of texts. With this in mind, we argue that embedding KBs within deep
neural architectures supporting documentquery matching would give rise to
fine-grained latent representations of both words and their semantic relations.
In this paper, we review the main approaches of neural-based document ranking
as well as those approaches for latent representation of entities and relations
via KBs. We then propose some avenues to incorporate KBs in deep neural
approaches for document ranking. More particularly, this paper advocates that
KBs can be used either to support enhanced latent representations of queries
and documents based on both distributional and relational semantics or to serve
as a semantic translator between their latent distributional representations. | ['Nathalie Bricon-Souf', 'Gia-Hung Nguyen', 'Laure Soulier', 'Lynda Tamine'] | 2016-06-23 | null | null | null | null | ['ad-hoc-information-retrieval', 'implicit-relations'] | ['natural-language-processing', 'natural-language-processing'] | [-5.30551113e-02 5.63173890e-01 -5.01912415e-01 -6.20121419e-01
-8.46329629e-01 -5.77597082e-01 1.12614655e+00 7.31218040e-01
-5.38898110e-01 3.40609282e-01 8.01823854e-01 -2.39300489e-01
-6.28240168e-01 -1.24017310e+00 -4.98631507e-01 -2.92337656e-01
2.07396001e-01 1.00182760e+00 -2.74565686e-02 -5.69873750e-01
2.42922410e-01 5.37123382e-01 -1.46534872e+00 9.17764604e-01
6.57458186e-01 9.58288312e-01 -1.02231070e-01 2.06866458e-01
-7.11697042e-01 1.03455162e+00 -6.93806350e-01 -8.27885270e-01
-2.54374653e-01 4.89248298e-02 -1.44018281e+00 -7.19089866e-01
6.18665993e-01 -4.14230198e-01 -5.95143020e-01 8.28032076e-01
4.38899547e-01 3.15849781e-01 1.16810226e+00 -9.94903624e-01
-1.29226184e+00 9.72058356e-01 1.00466490e-01 2.29835123e-01
4.11940485e-01 -7.50383675e-01 1.76595497e+00 -8.41264486e-01
7.18505442e-01 1.46382368e+00 4.56284970e-01 4.14058506e-01
-1.06501544e+00 -1.88473985e-01 2.55093537e-02 5.36590993e-01
-1.18387556e+00 -3.80986542e-01 7.24127829e-01 -4.36250597e-01
1.65217733e+00 2.94772625e-01 4.53411877e-01 1.04669273e+00
-1.99316859e-01 9.51733530e-01 4.29020852e-01 -8.69226456e-01
-1.55988201e-01 3.53062809e-01 5.73513508e-01 4.31085318e-01
3.44428241e-01 -8.53398442e-02 -5.06241620e-01 -3.64956379e-01
4.83730376e-01 -1.13019936e-01 -2.24681422e-01 -3.92705828e-01
-9.34343040e-01 1.19110751e+00 7.98742473e-01 7.95734286e-01
-3.86541575e-01 4.04510915e-01 8.11640203e-01 2.71904111e-01
6.58789217e-01 9.17526901e-01 -4.14933980e-01 3.30557704e-01
-7.53533363e-01 3.09237480e-01 7.69157052e-01 6.88807070e-01
6.19475186e-01 -3.38063240e-01 -6.30000353e-01 1.10583079e+00
6.36750460e-01 1.12729810e-01 5.25519669e-01 -8.94953072e-01
5.71073472e-01 8.24625254e-01 3.07075605e-02 -1.53930330e+00
-1.19136930e-01 -2.94158041e-01 -4.88137782e-01 -4.04716223e-01
5.12720421e-02 4.94470388e-01 -4.79476929e-01 1.66198707e+00
-1.84648372e-02 -3.00935149e-01 6.46014154e-01 6.68834031e-01
1.23352647e+00 7.39024103e-01 2.13647261e-01 1.41453683e-01
1.50389576e+00 -9.79829431e-01 -8.84634197e-01 -1.51884869e-01
1.19559872e+00 -5.53921223e-01 8.96469057e-01 -3.19678932e-01
-9.85212743e-01 -3.54570508e-01 -7.30503380e-01 -6.38471305e-01
-1.15007031e+00 1.91854432e-01 9.06102419e-01 2.62632191e-01
-1.31038582e+00 5.49406171e-01 -5.15395939e-01 -6.80750668e-01
2.74734199e-01 3.27400446e-01 -2.35659212e-01 1.57118961e-01
-1.81771624e+00 1.24624324e+00 7.18622088e-01 2.00787798e-01
-3.77364695e-01 -4.98120427e-01 -9.29567873e-01 4.92407501e-01
8.90410244e-02 -1.02038562e+00 1.22973859e+00 -7.22011626e-01
-9.35976982e-01 1.29021907e+00 -2.28614986e-01 -5.90216994e-01
-2.01986089e-01 -5.32061875e-01 -4.14934784e-01 3.19474131e-01
2.20843822e-01 6.29975140e-01 2.74114877e-01 -1.06581128e+00
-2.69601643e-01 -4.08268362e-01 6.66310906e-01 4.29539502e-01
-7.22250879e-01 2.12849140e-01 -1.42745316e-01 -6.19855642e-01
-6.10030331e-02 -5.62065661e-01 2.72542775e-01 -2.60112226e-01
-4.25364733e-01 -9.17735994e-01 5.21964669e-01 -6.32307351e-01
1.13972700e+00 -1.67108548e+00 9.31924954e-02 1.16534874e-01
2.07739994e-01 5.06375074e-01 -2.18326047e-01 8.74766052e-01
-7.79068619e-02 1.20377474e-01 3.10496151e-01 -1.97077710e-02
3.24003726e-01 3.06360006e-01 -9.46905971e-01 -7.16006458e-02
9.87619236e-02 1.50741577e+00 -1.10294783e+00 -3.68077546e-01
-2.33981788e-01 7.41690457e-01 -3.21872652e-01 2.42798224e-01
-1.93200901e-01 -1.71409160e-01 -5.39302349e-01 2.16135964e-01
7.29413256e-02 -3.23943615e-01 4.37352836e-01 -5.61799467e-01
2.16324478e-01 9.35248375e-01 -6.21596575e-01 1.62595201e+00
-5.74885190e-01 9.02509451e-01 -4.44314420e-01 -1.28011823e+00
1.05886042e+00 6.13165200e-01 2.86727488e-01 -8.67762625e-01
-2.30538230e-02 4.08542722e-01 -4.02811617e-01 -3.97301257e-01
1.05717289e+00 -2.36461133e-01 -4.97056395e-02 6.76099896e-01
1.01122171e-01 -3.66931967e-02 2.07826391e-01 5.28612435e-01
7.67926335e-01 4.55661900e-02 2.03469142e-01 -2.48960659e-01
5.84440053e-01 3.90506387e-02 -3.17767620e-01 7.14779913e-01
4.50407237e-01 8.80521443e-03 3.90178174e-01 -5.58639586e-01
-9.77016091e-01 -8.73655558e-01 -1.40279397e-01 1.32028019e+00
1.89091952e-03 -4.30805802e-01 -3.66600275e-01 -6.05804980e-01
1.18389793e-01 8.36581409e-01 -7.28059649e-01 -5.78001201e-01
-3.99152577e-01 -3.48065376e-01 8.22906077e-01 8.39328289e-01
1.60637945e-01 -1.24806058e+00 -3.91285509e-01 1.22414015e-01
-3.55540633e-01 -9.41660643e-01 -7.85190016e-02 3.96055818e-01
-9.09709752e-01 -9.01364148e-01 -8.88737202e-01 -1.11227357e+00
5.57088733e-01 3.21129471e-01 1.51253200e+00 9.85582247e-02
-6.22225739e-02 8.04935753e-01 -5.45504272e-01 -1.71032563e-01
-4.02916312e-01 3.08733404e-01 6.07415549e-02 -4.35040087e-01
8.96587789e-01 -3.38634044e-01 -6.67045414e-01 -1.37136668e-01
-1.13365185e+00 -2.80227840e-01 2.84593254e-01 8.44573259e-01
2.04238817e-01 -6.81470484e-02 4.92941439e-01 -1.06123793e+00
1.22662377e+00 -6.19145632e-01 -3.17563236e-01 7.42739439e-01
-6.78209424e-01 5.06274343e-01 2.70053446e-01 -1.33032814e-01
-1.23186648e+00 -6.88493907e-01 -3.82407606e-02 -4.44254391e-02
-8.10969993e-03 1.02520454e+00 1.29018739e-01 3.20580900e-01
8.57588768e-01 -2.53552608e-02 -5.10832191e-01 -4.87699479e-01
9.52523172e-01 7.81168759e-01 1.83586016e-01 -9.67329860e-01
4.81068611e-01 4.33498710e-01 -2.34429821e-01 -3.82064044e-01
-1.33219814e+00 -8.03548574e-01 -7.93770313e-01 3.14853102e-01
8.85525942e-01 -7.69237638e-01 -4.34248924e-01 -2.18145356e-01
-1.76869142e+00 6.05717674e-02 -6.20799184e-01 3.59686524e-01
-5.32271922e-01 1.43718809e-01 -8.42309356e-01 -5.61962068e-01
-5.48153222e-01 -7.55900741e-01 1.23322606e+00 -5.68611547e-02
-5.54537952e-01 -1.53510976e+00 3.97172123e-01 4.46831077e-01
5.01492321e-01 -1.34941354e-01 1.58419049e+00 -1.12928617e+00
-1.99616030e-01 -4.60351884e-01 -5.74501157e-01 2.00645000e-01
1.65314034e-01 -2.86173582e-01 -1.17657661e+00 -9.47191045e-02
-3.78291935e-01 -6.28699899e-01 1.09388578e+00 1.25810415e-01
7.39018679e-01 -6.12284184e-01 -4.96683747e-01 2.78803289e-01
1.32042205e+00 1.01752937e-01 5.61249495e-01 6.51324034e-01
6.33183002e-01 1.25064385e+00 4.32447493e-01 3.70929874e-02
5.01566470e-01 8.35528135e-01 2.70235777e-01 -6.21044114e-02
-1.16006508e-01 -3.89047235e-01 1.46243706e-01 1.03150451e+00
-2.78837066e-02 -4.42732811e-01 -1.11463463e+00 5.46723902e-01
-1.90011823e+00 -1.06166172e+00 -2.84042377e-02 1.84471154e+00
1.03261364e+00 -3.19526076e-01 -4.13571358e-01 -1.22940302e-01
6.75638139e-01 2.86422580e-01 -2.69617617e-01 -6.96195900e-01
-2.68626124e-01 1.32136613e-01 2.89235469e-02 6.43886507e-01
-9.50178325e-01 1.24733603e+00 6.23902369e+00 6.56456888e-01
-8.80184948e-01 -4.72354107e-02 4.41484362e-01 3.15418422e-01
-7.56832302e-01 5.49093559e-02 -1.07109869e+00 -2.80862123e-01
9.88808811e-01 -2.78244644e-01 3.18586864e-02 7.82287180e-01
-3.54217172e-01 3.31323296e-01 -1.50386679e+00 9.20024157e-01
1.07921436e-01 -1.65159976e+00 7.24042475e-01 -1.97665580e-02
6.00731850e-01 -1.63026676e-01 1.00007914e-01 6.44418180e-01
5.30069172e-01 -1.16977811e+00 4.01541233e-01 6.28508389e-01
5.59384823e-01 -4.50566918e-01 1.06329870e+00 7.98702240e-02
-9.73493040e-01 2.03803629e-01 -7.23587275e-01 7.99686462e-02
-4.18161005e-02 2.93747514e-01 -9.61250484e-01 6.76376820e-01
5.44553936e-01 6.17648542e-01 -4.52477872e-01 5.69868207e-01
-3.91667187e-01 5.32727130e-02 7.96691477e-02 -1.00080967e-01
4.17192996e-01 1.04149342e-01 2.55937159e-01 1.42500257e+00
2.88927495e-01 -1.77986026e-01 -1.87165245e-01 1.01949751e+00
-2.23974973e-01 5.46249509e-01 -1.02864599e+00 -5.05469024e-01
5.13442516e-01 9.61139321e-01 -4.96052504e-01 -5.58649540e-01
-3.29333037e-01 1.01516652e+00 8.83366108e-01 6.33919477e-01
-3.66813004e-01 -5.11152923e-01 5.56163132e-01 8.19995776e-02
3.31591256e-02 -3.82677279e-02 2.78402001e-01 -1.27071404e+00
-9.18392166e-02 -4.78856951e-01 6.98086143e-01 -9.18206751e-01
-1.46034861e+00 8.27196121e-01 2.74769902e-01 -6.83392644e-01
-7.35004067e-01 -7.79059827e-01 -1.79238051e-01 9.13180828e-01
-1.94333708e+00 -1.26313317e+00 9.38179642e-02 3.53213161e-01
2.36171782e-01 -2.36766592e-01 1.40375996e+00 2.75931805e-01
6.07176572e-02 4.11247075e-01 3.51338118e-01 4.25046682e-01
6.65206850e-01 -1.14950204e+00 1.53660148e-01 1.66684777e-01
7.85391688e-01 1.33530211e+00 4.65262771e-01 -5.02956390e-01
-8.88268828e-01 -8.14780891e-01 1.53755069e+00 -6.53778911e-01
7.46058822e-01 -9.59130973e-02 -1.04485440e+00 7.65494287e-01
5.03223836e-01 -1.32563815e-01 1.13477957e+00 6.73651159e-01
-8.74810755e-01 2.64662374e-02 -6.13917649e-01 5.65099001e-01
7.02688277e-01 -1.26983786e+00 -1.07862127e+00 5.86175323e-01
9.11367893e-01 3.42331082e-03 -9.21541929e-01 3.64964902e-01
5.43465316e-01 -6.60317898e-01 1.30685914e+00 -1.20229936e+00
4.72294539e-01 7.84708187e-02 -3.92749518e-01 -1.17497778e+00
-3.45625997e-01 9.91491228e-02 -4.75670725e-01 1.44878757e+00
4.70999390e-01 -3.64604592e-01 6.40764236e-01 7.34286368e-01
1.83428898e-02 -7.15738177e-01 -6.22431815e-01 -8.07117343e-01
4.57447618e-01 -4.99858767e-01 5.06093442e-01 1.25441909e+00
2.97330409e-01 6.98392034e-01 5.34960330e-02 -8.53033215e-02
4.12433818e-02 2.05360711e-01 3.64768237e-01 -1.44663393e+00
-1.16553739e-01 -6.49851203e-01 -4.39897895e-01 -1.11552858e+00
8.00659657e-01 -1.57415187e+00 -2.59074062e-01 -2.20878410e+00
2.42792666e-01 -3.97420883e-01 -6.56398773e-01 4.98340011e-01
4.07608710e-02 1.15389645e-01 -1.07133418e-01 5.28180599e-01
-8.36242259e-01 6.62638783e-01 7.73672819e-01 -4.94663537e-01
1.27973318e-01 -3.37652057e-01 -8.81053269e-01 6.51518822e-01
7.67450571e-01 -5.64953625e-01 -5.70001185e-01 -9.66505289e-01
1.00426221e+00 -1.18445769e-01 3.76707584e-01 -2.24423513e-01
3.40401232e-01 1.03073478e-01 7.74995759e-02 -3.37832391e-01
4.72356856e-01 -9.09851372e-01 -4.34648633e-01 2.47614339e-01
-1.16391194e+00 2.10513175e-02 5.17101809e-02 5.13208389e-01
-7.26093531e-01 -6.41714990e-01 1.73503578e-01 -3.06104690e-01
-8.29244196e-01 -2.36378470e-03 -3.49672675e-01 2.37288311e-01
4.98657912e-01 -6.27221167e-02 -5.88804483e-01 -5.05038321e-01
-7.04419971e-01 1.06437251e-01 -4.50544022e-02 6.61962509e-01
6.80128932e-01 -1.63791382e+00 -5.19737780e-01 -2.95115113e-01
5.70552647e-01 -1.03517644e-01 -5.55725284e-02 3.60902190e-01
-4.53825086e-01 1.44177890e+00 6.66713342e-02 -1.63188145e-01
-1.14059520e+00 5.89324474e-01 2.73947120e-01 -7.40206838e-01
-3.89419138e-01 1.11917961e+00 3.18811417e-01 -6.42142773e-01
3.81005436e-01 -2.09856227e-01 -8.38110089e-01 6.31553173e-01
5.40767372e-01 5.65871336e-02 1.49032146e-01 -7.91706085e-01
-3.58724236e-01 4.40097868e-01 -3.00550997e-01 -8.78408700e-02
1.35805225e+00 -8.02319720e-02 -5.54697573e-01 6.57371402e-01
1.53254819e+00 -8.94597173e-02 -2.21787110e-01 -7.11078882e-01
7.30163813e-01 -2.07087129e-01 5.52634373e-02 -7.08995640e-01
-7.64731705e-01 1.17865586e+00 2.34883487e-01 3.40430766e-01
6.18308425e-01 5.88181376e-01 8.00565660e-01 9.53610539e-01
1.03090100e-01 -9.76129472e-01 3.09574485e-01 7.71366477e-01
9.32716548e-01 -1.11924052e+00 -1.87725276e-01 -8.91258270e-02
-4.53016579e-01 1.47586977e+00 4.21917498e-01 -5.95882200e-02
3.89876544e-01 -3.28015953e-01 1.49786025e-01 -6.93374217e-01
-7.64365852e-01 -3.83654684e-01 9.11707222e-01 6.52606964e-01
1.12965083e+00 -2.77295083e-01 -2.84470201e-01 3.91429543e-01
-1.97137848e-01 -3.32754850e-01 -1.00087777e-01 7.38943040e-01
-4.70227659e-01 -1.36935520e+00 -1.07454630e-02 3.19467634e-01
-4.90027100e-01 -6.19365275e-01 -6.04169130e-01 3.90257210e-01
-1.62655801e-01 6.63686693e-01 1.44177184e-01 -9.86137539e-02
2.09687382e-01 2.91053802e-01 3.12263668e-01 -9.99418795e-01
-4.69047308e-01 -4.13033903e-01 3.28300506e-01 -2.76089251e-01
-6.21157289e-01 -8.78849104e-02 -1.11922824e+00 1.37696892e-01
-4.63572383e-01 7.08224893e-01 5.68101645e-01 1.02046311e+00
3.42383742e-01 6.05166137e-01 -4.91570868e-02 -3.65439534e-01
-5.21559298e-01 -8.56412709e-01 -4.97321278e-01 4.80898649e-01
-2.39767488e-02 -5.33394575e-01 3.28824930e-02 -1.10950001e-01] | [10.117773056030273, 8.49109172821045] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.