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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
62e800db-0752-43ca-af59-50364d8841d7 | elevation-estimation-driven-building-3d | 2301.04581 | null | https://arxiv.org/abs/2301.04581v1 | https://arxiv.org/pdf/2301.04581v1.pdf | Elevation Estimation-Driven Building 3D Reconstruction from Single-View Remote Sensing Imagery | Building 3D reconstruction from remote sensing images has a wide range of applications in smart cities, photogrammetry and other fields. Methods for automatic 3D urban building modeling typically employ multi-view images as input to algorithms to recover point clouds and 3D models of buildings. However, such models rely heavily on multi-view images of buildings, which are time-intensive and limit the applicability and practicality of the models. To solve these issues, we focus on designing an efficient DSM estimation-driven reconstruction framework (Building3D), which aims to reconstruct 3D building models from the input single-view remote sensing image. First, we propose a Semantic Flow Field-guided DSM Estimation (SFFDE) network, which utilizes the proposed concept of elevation semantic flow to achieve the registration of local and global features. Specifically, in order to make the network semantics globally aware, we propose an Elevation Semantic Globalization (ESG) module to realize the semantic globalization of instances. Further, in order to alleviate the semantic span of global features and original local features, we propose a Local-to-Global Elevation Semantic Registration (L2G-ESR) module based on elevation semantic flow. Our Building3D is rooted in the SFFDE network for building elevation prediction, synchronized with a building extraction network for building masks, and then sequentially performs point cloud reconstruction, surface reconstruction (or CityGML model reconstruction). On this basis, our Building3D can optionally generate CityGML models or surface mesh models of the buildings. Extensive experiments on ISPRS Vaihingen and DFC2019 datasets on the DSM estimation task show that our SFFDE significantly improves upon state-of-the-arts. Furthermore, our Building3D achieves impressive results in the 3D point cloud and 3D model reconstruction process. | ['Kun fu', 'Xian Sun', 'Wenhui Diao', 'Zhirui Wang', 'Wenjie Liu', 'Deke Tang', 'Wei Chen', 'Liangjin Zhao', 'Kaiqiang Chen', 'Yongqiang Mao'] | 2023-01-11 | null | null | null | null | ['point-cloud-reconstruction'] | ['computer-vision'] | [ 6.53294921e-02 -2.08762795e-01 2.89899051e-01 -4.17574286e-01
-5.91074109e-01 -3.18561912e-01 7.14972675e-01 -2.08593801e-01
3.31854224e-01 2.90592998e-01 4.62555401e-02 -3.75530541e-01
-1.93226859e-01 -1.73799205e+00 -4.29417580e-01 -4.84129012e-01
2.36419976e-01 6.81613922e-01 3.78575176e-01 -3.20960969e-01
9.33540314e-02 8.92883241e-01 -1.73541081e+00 -1.15155065e-02
1.08309686e+00 1.01184511e+00 5.66266000e-01 2.96751112e-01
-6.38144314e-01 1.27495518e-02 -1.75181888e-02 3.64824645e-02
4.46770668e-01 5.60203940e-02 -6.51580989e-01 7.00796902e-01
3.09370935e-01 -4.38504905e-01 3.90701601e-03 9.28303123e-01
4.42771167e-01 -1.50935873e-01 4.92183715e-01 -1.27495384e+00
-1.42178014e-01 -3.34589295e-02 -6.06207788e-01 -4.84446138e-01
2.57419527e-01 9.91525874e-02 6.97920740e-01 -1.25721085e+00
4.44245309e-01 1.27332497e+00 7.49263346e-01 -8.62909667e-03
-1.05019188e+00 -6.27744555e-01 4.50355113e-01 -4.38746773e-02
-1.91930246e+00 -2.81076103e-01 1.08142114e+00 -4.32564616e-01
5.08201540e-01 6.02937639e-01 1.09002280e+00 2.16590628e-01
-3.71418774e-01 4.93901074e-01 1.01508892e+00 2.70586903e-03
3.08581382e-01 -2.71386534e-01 -2.36661762e-01 5.87642074e-01
1.31644636e-01 -5.08180745e-02 -2.42058784e-01 -2.39116773e-01
1.13277793e+00 4.58335429e-01 -2.24967867e-01 -3.32511693e-01
-1.03522444e+00 6.45483673e-01 7.19629943e-01 6.25198632e-02
-4.67970133e-01 1.64017498e-01 -2.29383871e-01 -1.22764386e-01
9.08006370e-01 -5.02375245e-01 -3.28390181e-01 5.02708375e-01
-1.22904563e+00 2.10048735e-01 3.55148315e-01 1.11635983e+00
1.38247693e+00 -1.48902796e-02 5.43875039e-01 7.57675111e-01
9.59983110e-01 9.63487983e-01 -1.99192524e-01 -1.05772221e+00
6.96026266e-01 1.22541118e+00 -3.78185022e-03 -1.19845223e+00
-3.64010543e-01 -2.47807026e-01 -1.07616675e+00 7.91800246e-02
-2.99159676e-01 4.87374097e-01 -9.17499423e-01 1.05870342e+00
9.06391144e-01 4.38726127e-01 -1.01705573e-01 7.35958099e-01
9.79596496e-01 6.65589392e-01 -9.60876606e-03 1.28037944e-01
1.21148551e+00 -3.95258009e-01 -2.35887721e-01 -4.25733566e-01
6.16691172e-01 -6.35117829e-01 8.90789747e-01 -1.02599867e-01
-8.35848391e-01 -4.29134130e-01 -7.71217704e-01 -7.48691335e-02
-3.06575805e-01 -8.13157707e-02 4.75778103e-01 4.54159588e-01
-1.12775075e+00 4.10709321e-01 -8.35333169e-01 -4.49734658e-01
5.69698155e-01 1.54751554e-01 -8.64105150e-02 -3.16856951e-01
-8.56394291e-01 6.51364803e-01 3.12314093e-01 4.24946427e-01
-5.04152060e-01 -8.80988598e-01 -8.29137564e-01 -7.95536116e-02
9.32595730e-02 -1.11816454e+00 6.63059950e-01 -3.45721632e-01
-1.07441676e+00 1.07147241e+00 -1.56519383e-01 6.18102066e-02
5.16654432e-01 2.97652811e-01 -4.99031812e-01 -2.04528011e-02
5.03396809e-01 5.07159114e-01 4.09325302e-01 -1.79987240e+00
-7.61630356e-01 -9.06265020e-01 -1.68004379e-01 3.48121196e-01
1.19871289e-01 -4.33771849e-01 -4.64842379e-01 -3.13661516e-01
1.22659898e+00 -6.54931068e-01 -5.39152801e-01 1.68502614e-01
-3.41187716e-01 8.52752924e-02 1.19488609e+00 -8.95637989e-01
1.10819256e+00 -2.09334588e+00 -2.78565526e-01 7.74423897e-01
1.51130900e-01 -1.43774167e-01 8.96086991e-02 2.59679556e-01
1.38689309e-01 3.42527807e-01 -7.61593819e-01 -5.37125826e-01
-8.24945867e-02 4.42772418e-01 -2.71202505e-01 4.39673781e-01
-2.30612159e-01 7.48340607e-01 -6.46570086e-01 -6.76246345e-01
8.02536488e-01 6.47718668e-01 -4.28483218e-01 -8.17440525e-02
-2.35597044e-01 6.48350000e-01 -1.04751050e+00 8.21878731e-01
1.40394056e+00 -3.76530260e-01 1.61037371e-01 -2.48751983e-01
-5.39371133e-01 3.09826672e-01 -1.59109986e+00 1.82521296e+00
-8.24414551e-01 -1.76039383e-01 1.88910514e-01 -6.89508557e-01
1.18366241e+00 1.49693638e-01 7.86833227e-01 -4.93130863e-01
-3.10504317e-01 4.41355735e-01 -8.68276954e-01 -4.38828915e-02
3.87721717e-01 -8.96239802e-02 3.92903201e-02 1.80113792e-01
-6.88808322e-01 -9.10584807e-01 -4.28030640e-01 1.67526394e-01
5.17978430e-01 4.21140462e-01 1.82366475e-01 -2.06762552e-01
7.97562718e-01 2.36915573e-01 4.48558122e-01 1.73931047e-02
4.15923238e-01 7.88534343e-01 -2.65791833e-01 -7.00179994e-01
-1.01461256e+00 -1.27698696e+00 -2.41530970e-01 2.82429069e-01
4.87186402e-01 -4.03941840e-01 -7.09082246e-01 -4.11883175e-01
1.00463897e-01 7.58189201e-01 6.47352636e-02 3.50856811e-01
-6.12902880e-01 -7.85489440e-01 1.84495538e-01 2.33483151e-01
1.11577880e+00 -5.47677934e-01 -3.88654441e-01 3.09902817e-01
-4.99383479e-01 -1.12351251e+00 -1.17175974e-01 -5.14363050e-01
-1.30070734e+00 -1.06685197e+00 -1.84733003e-01 -5.34456015e-01
8.34113300e-01 6.70855165e-01 1.11288619e+00 4.99657631e-01
1.68295264e-01 4.02260065e-01 -7.81756565e-02 -3.38934399e-02
-2.00186312e-01 8.05898458e-02 -2.33483762e-01 9.11423713e-02
2.02232450e-01 -1.34013259e+00 -7.66406298e-01 5.82238853e-01
-9.07519221e-01 6.10495627e-01 1.65894598e-01 -1.67775303e-02
9.92529213e-01 1.54929772e-01 1.55941471e-01 -5.19465446e-01
-1.68382302e-01 -4.74311382e-01 -9.93419945e-01 1.07655302e-01
-5.98140001e-01 -3.68463159e-01 2.51683772e-01 3.89116883e-01
-1.02999139e+00 4.43405807e-01 -4.30471063e-01 -2.59656519e-01
-3.38521808e-01 4.54766035e-01 -7.69746184e-01 6.11480027e-02
2.29693845e-01 5.67040145e-01 -3.63773704e-01 -8.09976101e-01
4.69321340e-01 9.07513082e-01 2.86521822e-01 -3.51247698e-01
1.34431088e+00 1.07483566e+00 2.95538604e-01 -1.00317562e+00
-7.74380386e-01 -5.78004539e-01 -8.43997717e-01 -4.48525518e-01
8.41833234e-01 -1.29566097e+00 -4.36391681e-01 4.23016518e-01
-1.33862507e+00 -9.61982831e-02 -2.04800442e-01 2.49912784e-01
-6.19644880e-01 3.03542286e-01 -3.18067260e-02 -8.00431132e-01
-2.95966715e-01 -1.08512270e+00 1.50506139e+00 -1.41252190e-01
2.54114240e-01 -8.75733614e-01 1.02233132e-02 5.28877556e-01
2.38139465e-01 5.27068257e-01 8.10299575e-01 3.08463663e-01
-1.48716164e+00 2.61100173e-01 -2.60885686e-01 1.06205665e-01
3.79657447e-01 -7.59007633e-02 -1.04868805e+00 1.01831652e-01
4.25590687e-02 5.77257335e-01 5.34938812e-01 4.96843934e-01
1.03154063e+00 -1.78097293e-01 -4.95777994e-01 1.05631232e+00
1.78072858e+00 -1.14094280e-01 9.02259231e-01 5.34609914e-01
1.00900316e+00 5.89067698e-01 4.76722091e-01 5.85157156e-01
1.10967720e+00 7.13553250e-01 9.78094280e-01 -2.44791746e-01
-1.55227929e-01 -6.01954818e-01 -7.16634095e-02 1.06989539e+00
-4.05899137e-01 1.16330378e-01 -1.15428746e+00 5.18441617e-01
-1.81840742e+00 -8.51623297e-01 -7.82154679e-01 2.26843953e+00
2.52043009e-01 -2.81741589e-01 -2.33043313e-01 1.03173763e-01
8.54091525e-01 4.56361651e-01 -4.05465454e-01 2.80920804e-01
-1.09968752e-01 1.20093539e-01 6.40246928e-01 6.47603750e-01
-8.46601307e-01 1.02156305e+00 4.58911324e+00 8.20602357e-01
-8.65323007e-01 3.74945760e-01 3.77251327e-01 2.94667572e-01
-8.71869862e-01 3.59414935e-01 -8.86281669e-01 3.56652468e-01
6.16167426e-01 -1.73602998e-02 3.67626280e-01 1.08337307e+00
7.26742923e-01 5.90571947e-02 -4.35835451e-01 1.08206677e+00
-4.45487350e-01 -1.67313838e+00 2.59060264e-01 5.76641619e-01
8.51011574e-01 3.36217791e-01 -4.33445007e-01 -1.84202582e-01
4.13382769e-01 -6.10123038e-01 6.99936748e-01 5.12117743e-01
1.05942857e+00 -7.78417468e-01 2.11990148e-01 7.33634412e-01
-2.07915998e+00 3.12233180e-01 -2.81158477e-01 1.69592932e-01
5.55235147e-01 1.06592083e+00 -6.21585786e-01 1.13593936e+00
6.33965313e-01 8.29161108e-01 -3.39574933e-01 8.74891937e-01
-3.47126663e-01 2.14553371e-01 -6.68018103e-01 6.20289385e-01
1.45900503e-01 -8.20209384e-01 3.45165282e-01 5.65771222e-01
7.68470585e-01 2.91151911e-01 4.78557825e-01 1.14211166e+00
-2.48619504e-02 1.47905394e-01 -7.68649936e-01 6.63197398e-01
6.26152337e-01 1.27136910e+00 -7.29413331e-01 -3.77133906e-01
-2.30051428e-01 7.77823448e-01 -6.93612099e-02 2.36082986e-01
-7.90989935e-01 1.72454789e-01 7.46694446e-01 8.37172806e-01
1.42379686e-01 -4.69409466e-01 -6.33202612e-01 -1.20811081e+00
1.18517503e-01 -2.54650742e-01 -8.97991508e-02 -9.95447874e-01
-9.91273999e-01 2.56729037e-01 4.52457815e-02 -1.57354689e+00
8.06480870e-02 2.24110838e-02 -5.46750367e-01 9.57541883e-01
-1.88049853e+00 -1.72666156e+00 -8.44935000e-01 7.36941218e-01
5.48904955e-01 5.23861945e-01 6.16380870e-01 4.65745449e-01
-1.44129977e-01 -6.16980731e-01 3.74720357e-02 -1.87957048e-01
1.54952817e-02 -8.23513985e-01 8.31466913e-01 8.25840294e-01
-7.56510720e-02 2.03294083e-01 3.39418426e-02 -9.10420179e-01
-1.06097829e+00 -1.81974220e+00 9.62759793e-01 -2.82704443e-01
1.84133872e-01 -1.64273188e-01 -7.31817424e-01 7.65242875e-01
-6.29182339e-01 1.05798692e-01 1.67104721e-01 -3.56457978e-01
-7.47453468e-03 -2.75926024e-01 -1.36170113e+00 5.12774527e-01
1.63300097e+00 -5.32579720e-01 -2.71289855e-01 5.12889504e-01
6.43061817e-01 -3.17016184e-01 -1.03903687e+00 6.05538130e-01
2.02714384e-01 -1.13766944e+00 1.38343441e+00 2.09478468e-01
2.55327433e-01 -9.37256634e-01 -7.00120091e-01 -1.08346272e+00
-2.62132496e-01 -7.74288252e-02 1.52015254e-01 1.21011174e+00
-4.24918756e-02 -7.21699357e-01 8.87709081e-01 5.71969926e-01
-4.96470898e-01 -3.73374224e-01 -1.11664498e+00 -6.02440000e-01
-2.38206297e-01 -8.95880342e-01 1.53806424e+00 9.99016762e-01
-9.41477120e-01 9.43601504e-02 1.75878197e-01 8.77874136e-01
8.87153506e-01 5.99891305e-01 1.02417147e+00 -1.77359855e+00
4.65872020e-01 -2.54730612e-01 -3.71109009e-01 -1.22895908e+00
4.95565347e-02 -9.67264056e-01 -2.71610409e-01 -2.20253158e+00
-1.35228500e-01 -9.74994123e-01 3.20980757e-01 2.71901190e-01
1.74943626e-01 2.22276673e-01 5.28139211e-02 5.24765491e-01
-1.81817234e-01 8.44135582e-01 1.13197351e+00 -1.31648093e-01
-2.89758116e-01 2.98872411e-01 -3.12121779e-01 1.02753782e+00
7.20802844e-01 -4.47905481e-01 -3.60159546e-01 -6.47412896e-01
2.33498544e-01 8.78824107e-03 8.37951899e-01 -9.38391507e-01
9.33813304e-02 -3.76232296e-01 2.21040323e-01 -1.54808009e+00
6.15766823e-01 -1.27603948e+00 9.11572397e-01 3.55499089e-01
6.80513203e-01 -8.18220377e-02 -2.12491274e-01 3.21294188e-01
-4.45152931e-02 1.16207778e-01 6.83445930e-01 -5.87919176e-01
-7.51380324e-01 8.29652488e-01 4.42183726e-02 -2.88454831e-01
8.28617632e-01 -6.82523072e-01 4.77942340e-02 -1.43144503e-01
-7.25943148e-01 3.63734424e-01 1.00175738e+00 -8.65144134e-02
9.13550198e-01 -1.61873901e+00 -6.86552763e-01 3.93390745e-01
1.63023755e-01 9.42499518e-01 4.80723321e-01 5.72126210e-01
-7.60377407e-01 1.51677251e-01 3.53064984e-01 -9.53901947e-01
-1.05773890e+00 9.06837266e-03 4.51625973e-01 -2.91435391e-01
-8.33309829e-01 2.20457509e-01 2.42020860e-01 -1.15335095e+00
-5.70683539e-01 -4.41166192e-01 8.39983448e-02 -1.05339050e-01
1.26268908e-01 5.15165210e-01 3.50816309e-01 -9.96670723e-01
-5.15524209e-01 1.16438937e+00 8.52717161e-01 -1.51272625e-01
1.70087600e+00 -5.74233830e-01 -3.40086758e-01 1.43706605e-01
9.08678055e-01 -2.10650079e-02 -1.04530561e+00 -2.80970186e-01
1.08031761e-02 -7.48043954e-01 4.88578618e-01 -2.85299450e-01
-1.25445712e+00 7.99411118e-01 5.32009184e-01 1.03250012e-01
1.08653855e+00 9.26804021e-02 8.93264949e-01 1.31426156e-01
8.78202677e-01 -8.47948432e-01 -5.53568840e-01 2.81202585e-01
8.32533121e-01 -9.10932720e-01 3.21224481e-01 -1.00068772e+00
-5.67270145e-02 9.73462045e-01 3.33784848e-01 6.45102859e-02
9.56356764e-01 -1.60935283e-01 -2.26499707e-01 -5.46235561e-01
-4.43704203e-02 -4.74478871e-01 7.45142922e-02 6.66305304e-01
-2.00853392e-01 2.58024871e-01 1.80295870e-01 8.07266980e-02
-3.44166309e-01 9.85959992e-02 9.36770290e-02 8.71334314e-01
-7.13857770e-01 -1.13448215e+00 -7.92773962e-01 3.11752766e-01
4.72181767e-01 -3.20527792e-01 5.68147860e-02 7.31158793e-01
3.90144229e-01 8.60559106e-01 2.23678216e-01 -3.25166881e-01
5.64383030e-01 -1.63738132e-01 1.25917628e-01 -6.83818638e-01
-2.31822029e-01 2.15818763e-01 -7.78200105e-02 -6.25523448e-01
-6.93708658e-01 -6.28344953e-01 -1.32023966e+00 -6.01247966e-01
-3.27931851e-01 -2.41641641e-01 1.03461063e+00 8.82987678e-01
5.20897925e-01 1.68782905e-01 1.03510511e+00 -1.10428345e+00
4.18242663e-02 -6.09817028e-01 -8.37430537e-01 1.66237205e-01
-7.66942650e-02 -5.63642383e-01 -3.50516051e-01 3.00370622e-02] | [8.168229103088379, -2.7890162467956543] |
56950e29-194a-4272-8dfb-a0e6d5c88a3d | flownet-20-evolution-of-optical-flow | 1612.01925 | null | http://arxiv.org/abs/1612.01925v1 | http://arxiv.org/pdf/1612.01925v1.pdf | FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks | The FlowNet demonstrated that optical flow estimation can be cast as a
learning problem. However, the state of the art with regard to the quality of
the flow has still been defined by traditional methods. Particularly on small
displacements and real-world data, FlowNet cannot compete with variational
methods. In this paper, we advance the concept of end-to-end learning of
optical flow and make it work really well. The large improvements in quality
and speed are caused by three major contributions: first, we focus on the
training data and show that the schedule of presenting data during training is
very important. Second, we develop a stacked architecture that includes warping
of the second image with intermediate optical flow. Third, we elaborate on
small displacements by introducing a sub-network specializing on small motions.
FlowNet 2.0 is only marginally slower than the original FlowNet but decreases
the estimation error by more than 50%. It performs on par with state-of-the-art
methods, while running at interactive frame rates. Moreover, we present faster
variants that allow optical flow computation at up to 140fps with accuracy
matching the original FlowNet. | ['Thomas Brox', 'Nikolaus Mayer', 'Alexey Dosovitskiy', 'Margret Keuper', 'Eddy Ilg', 'Tonmoy Saikia'] | 2016-12-06 | flownet-2-0-evolution-of-optical-flow | http://openaccess.thecvf.com/content_cvpr_2017/html/Ilg_FlowNet_2.0_Evolution_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Ilg_FlowNet_2.0_Evolution_CVPR_2017_paper.pdf | cvpr-2017-7 | ['dense-pixel-correspondence-estimation'] | ['computer-vision'] | [-2.72626877e-01 -1.94863543e-01 -3.45978700e-02 -2.27852017e-02
-2.32387856e-01 -5.35079420e-01 4.52862084e-01 -4.35369939e-01
-6.23117983e-01 1.04238319e+00 2.86431909e-01 -2.57459313e-01
1.40109479e-01 -5.56617737e-01 -6.61954045e-01 -4.74353105e-01
-3.40178490e-01 2.46801555e-01 7.17910886e-01 -2.35759333e-01
4.37416285e-01 5.45441628e-01 -1.49642169e+00 9.30553302e-02
6.80375874e-01 7.05517650e-01 1.83698490e-01 1.24209785e+00
-1.09830640e-01 1.30654299e+00 -6.03066862e-01 -2.52724916e-01
4.07176018e-01 -7.06042469e-01 -1.25342822e+00 -1.54953394e-02
1.16209662e+00 -9.63673174e-01 -6.27490640e-01 6.28068328e-01
5.27735293e-01 4.94162917e-01 1.97394103e-01 -1.34320664e+00
-4.11021501e-01 1.35274351e-01 -2.86402881e-01 6.48089051e-01
3.22114468e-01 4.08594072e-01 9.52131748e-01 -8.27515066e-01
1.12000966e+00 1.26541209e+00 5.49187720e-01 9.50352132e-01
-1.34014976e+00 -2.96039939e-01 1.56250060e-01 4.69641566e-01
-8.26533556e-01 -6.04156494e-01 4.03666377e-01 -5.27275920e-01
1.22873247e+00 1.21894039e-01 7.45259047e-01 7.79334486e-01
1.94438517e-01 9.17408884e-01 6.30329013e-01 -3.04355413e-01
-4.10977639e-02 -1.89376578e-01 -2.29949027e-01 8.28745782e-01
-8.28988254e-02 2.91732669e-01 -5.02633035e-01 4.45993990e-01
1.01784289e+00 -3.11119080e-01 -7.06286252e-01 -3.65569234e-01
-1.37783015e+00 6.58689082e-01 6.25094891e-01 3.80554914e-01
2.87583265e-02 6.00036860e-01 4.98499870e-01 3.32300752e-01
6.13607168e-01 3.69368911e-01 -3.81895810e-01 -5.90449870e-01
-1.35531592e+00 4.58293676e-01 9.20164406e-01 6.89032376e-01
9.98494267e-01 3.15208137e-01 -1.30039304e-01 4.82510030e-01
2.54388303e-01 3.43082666e-01 2.59123892e-01 -1.78562570e+00
4.54596341e-01 -8.69368166e-02 2.96759903e-01 -9.25575078e-01
-4.38497603e-01 -1.87546372e-01 -5.61055481e-01 6.69242084e-01
9.06558335e-01 -4.24878716e-01 -7.15665758e-01 1.67915690e+00
1.51889607e-01 8.32331300e-01 -2.33263105e-01 1.12594128e+00
5.70661604e-01 8.07457030e-01 -1.72814384e-01 -1.87603608e-01
7.72497892e-01 -1.51360977e+00 -7.73120522e-01 -1.90937132e-01
7.64278591e-01 -9.99020517e-01 8.18427861e-01 3.79920036e-01
-1.51359272e+00 -9.29637372e-01 -9.53731000e-01 -4.23885971e-01
-6.56570494e-02 -2.90901661e-01 6.37348533e-01 6.19231582e-01
-1.71282411e+00 1.18498027e+00 -1.09440327e+00 -2.94714451e-01
4.37630266e-01 4.40762669e-01 -4.40383613e-01 9.17841196e-02
-1.01871455e+00 9.20534730e-01 2.47563705e-01 1.25913784e-01
-6.31478906e-01 -1.27247548e+00 -9.93373036e-01 8.54926780e-02
2.07392588e-01 -1.10336316e+00 1.34783542e+00 -1.02888334e+00
-1.95516181e+00 5.24662375e-01 -4.31740493e-01 -5.25290430e-01
9.70043540e-01 -4.42122549e-01 -9.17514339e-02 5.24089277e-01
-5.08359484e-02 1.16716325e+00 8.20374966e-01 -1.00088370e+00
-8.79674852e-01 1.81500480e-01 3.07320595e-01 -8.20168108e-02
-4.70509613e-03 -1.94595024e-01 -4.04821455e-01 -4.10543829e-01
-5.25869727e-01 -9.51222122e-01 -1.73178986e-01 3.67375970e-01
9.86827984e-02 -1.10501498e-01 1.06294751e+00 -3.38323355e-01
1.17376125e+00 -1.92760408e+00 1.52842775e-01 -4.32126135e-01
4.37428623e-01 6.91398501e-01 -1.79839224e-01 2.04584390e-01
-1.40409887e-01 5.03375493e-02 -1.04109704e-01 -7.20932603e-01
-1.67134568e-01 3.27366501e-01 -3.04806918e-01 5.13594568e-01
3.96606684e-01 1.00465167e+00 -1.18988836e+00 -5.55318892e-01
7.26906538e-01 6.61225796e-01 -1.16699457e+00 2.44583473e-01
-2.39379592e-02 7.11822510e-01 1.46339666e-02 -1.07193328e-02
7.49038279e-01 -2.33762473e-01 -1.40826613e-01 -1.67178169e-01
-4.12485629e-01 2.21748993e-01 -1.35763288e+00 2.13836193e+00
-4.56906468e-01 1.23711085e+00 6.35951906e-02 -7.91187108e-01
3.75370383e-01 2.36983627e-01 8.12406719e-01 -4.23356026e-01
6.75378460e-03 2.35468030e-01 5.53757288e-02 -6.24521792e-01
6.80753708e-01 -1.39377162e-01 7.57736623e-01 4.92266148e-01
4.96096522e-01 -2.64317870e-01 9.18990076e-01 1.52307764e-01
9.09619927e-01 7.02840149e-01 -1.45907819e-01 -3.15465599e-01
7.85486281e-01 -8.00212994e-02 4.25026923e-01 5.58877647e-01
-6.64926291e-01 8.37302327e-01 6.11557007e-01 -8.55651498e-01
-1.12330091e+00 -8.99462759e-01 -2.02990435e-02 9.68180299e-01
2.47883409e-01 -4.93743390e-01 -8.26365113e-01 -7.59941161e-01
-7.61968493e-02 2.42895842e-01 -5.79665601e-01 2.19096646e-01
-1.06349087e+00 -4.42292392e-01 3.25370878e-01 5.63341737e-01
7.46015489e-01 -1.21330762e+00 -8.55316937e-01 4.56113666e-01
-3.19421828e-01 -1.50716364e+00 -6.54340148e-01 -4.84162718e-01
-1.04990530e+00 -1.15853512e+00 -8.67079377e-01 -5.63120961e-01
2.51228988e-01 3.11411858e-01 1.45212018e+00 3.19343925e-01
-1.88701764e-01 2.36951888e-01 -2.33715117e-01 9.52039659e-02
-3.65770370e-01 2.98475772e-01 -2.68567890e-01 -1.26754731e-01
-1.04848251e-01 -6.46077752e-01 -1.04797888e+00 1.68873861e-01
-9.08819377e-01 -8.43594745e-02 9.69438478e-02 7.19547212e-01
-2.04316918e-02 -4.43834215e-01 1.88216329e-01 -7.10089266e-01
2.49836892e-01 -1.57707483e-02 -5.59156060e-01 -2.51756519e-01
-4.23121303e-01 2.40664870e-01 7.72466898e-01 -1.55005872e-01
-1.12558472e+00 4.08984832e-02 -3.70289952e-01 -5.68358004e-01
3.06484513e-02 -1.35604993e-01 5.16765535e-01 -2.59791225e-01
6.43673837e-01 -2.57374167e-01 2.24249750e-01 -1.03912815e-01
4.50743228e-01 4.12145443e-02 5.16294479e-01 -3.91067117e-01
6.20524645e-01 8.17575693e-01 3.59882474e-01 -8.78440082e-01
-6.32251263e-01 -4.75060046e-01 -9.88033473e-01 -4.65240866e-01
9.15049136e-01 -6.59526885e-01 -1.18244135e+00 5.92554271e-01
-1.39750683e+00 -7.69159377e-01 -4.12765712e-01 7.11199582e-01
-8.31361473e-01 6.03521347e-01 -9.38908696e-01 -4.58248258e-01
-1.54096395e-01 -1.22879076e+00 8.88914704e-01 2.01865524e-01
-4.23391126e-02 -1.52686357e+00 3.68031234e-01 6.04446977e-02
6.84178412e-01 2.24599451e-01 1.01617374e-01 1.61471456e-01
-8.01273167e-01 3.49169850e-01 -4.17866856e-01 5.07915735e-01
-3.34383734e-02 5.07519901e-01 -1.11571288e+00 -3.83393615e-01
-1.76911250e-01 -2.60306001e-01 1.20723557e+00 6.90249562e-01
8.61909330e-01 1.64408475e-01 -1.53515758e-02 9.21835065e-01
1.46198142e+00 -6.96118921e-02 9.54495490e-01 2.08380669e-01
8.35121512e-01 6.54373825e-01 2.24644572e-01 2.34404802e-01
3.01755250e-01 5.41611016e-01 5.12163162e-01 -1.76249370e-01
-6.42254174e-01 7.68394256e-03 4.38321799e-01 6.36523545e-01
-5.16271710e-01 -2.61680365e-01 -6.04102492e-01 5.69356322e-01
-1.97585738e+00 -1.32783890e+00 -4.39786047e-01 1.94636953e+00
5.13864815e-01 2.35935658e-01 2.46419132e-01 1.76820412e-01
4.26700175e-01 5.27336597e-01 -2.43233889e-01 -4.59456742e-01
1.17570512e-01 4.29294050e-01 4.89051878e-01 1.27366841e+00
-1.08922434e+00 1.12825620e+00 7.03276443e+00 3.38310778e-01
-1.36605167e+00 6.17143884e-02 4.08514291e-01 -2.38389775e-01
9.69841853e-02 -1.50034791e-02 -6.00140750e-01 4.71884429e-01
9.20606554e-01 9.59170982e-02 3.87255102e-01 4.03606832e-01
5.36318421e-01 -2.37643704e-01 -1.20943522e+00 9.44794893e-01
1.30057959e-02 -1.66234767e+00 -6.42745718e-02 1.25512972e-01
8.09187770e-01 2.16189608e-01 -1.86385378e-01 1.53288141e-01
1.72074631e-01 -8.97958636e-01 6.00979626e-01 4.31233883e-01
6.46846473e-01 -5.68523347e-01 7.11433887e-01 3.49921826e-03
-1.26266527e+00 1.48425728e-01 -3.73565942e-01 -3.94301713e-01
5.93004167e-01 4.85698730e-01 -3.88286412e-01 5.48617184e-01
5.99422812e-01 1.20684731e+00 -4.01251405e-01 1.17372143e+00
-2.40957245e-01 5.18932462e-01 -2.21066609e-01 2.61652261e-01
6.01722538e-01 -1.07233696e-01 4.34340924e-01 1.46749151e+00
1.58778876e-01 -1.33891121e-01 1.17344469e-01 8.90523732e-01
1.98961925e-02 -1.67191476e-01 -3.74726862e-01 4.39352512e-01
-1.46277249e-01 1.21031654e+00 -6.07760549e-01 -4.69718993e-01
-5.05628586e-01 1.05051470e+00 3.72280031e-01 4.53873843e-01
-7.71257579e-01 -3.82812828e-01 9.41561818e-01 1.74411774e-01
4.32717234e-01 -4.56171155e-01 1.35242552e-01 -1.47040939e+00
-5.18271104e-02 -1.65242150e-01 6.10550009e-02 -7.95985222e-01
-8.93864989e-01 6.13669217e-01 -2.55916774e-01 -1.18397093e+00
-7.04643726e-01 -9.29096699e-01 -5.15822470e-01 8.88148189e-01
-2.00424695e+00 -5.70131540e-01 -5.19406378e-01 6.21921062e-01
7.17058480e-01 4.12046582e-01 6.64028406e-01 6.34408951e-01
-4.15299118e-01 2.25253299e-01 -9.78910998e-02 3.56008917e-01
9.93280828e-01 -1.48846531e+00 6.89923704e-01 1.14945972e+00
2.02557176e-01 2.91919947e-01 8.10007036e-01 -5.93487248e-02
-9.41684783e-01 -7.85832465e-01 1.10065246e+00 -6.38338327e-01
8.10641587e-01 -2.02888161e-01 -8.26642513e-01 7.35425472e-01
5.13530254e-01 8.66710484e-01 1.16862342e-01 -2.22548962e-01
-8.16234425e-02 -1.54980555e-01 -9.33933973e-01 5.07355750e-01
1.16081345e+00 -3.73592854e-01 -2.69554108e-01 1.30294070e-01
5.98395467e-01 -7.35018373e-01 -6.40717208e-01 3.82868052e-02
6.07093036e-01 -1.64090872e+00 1.00038767e+00 -6.18189156e-01
8.38870108e-01 -2.83080071e-01 3.90888304e-01 -1.43517637e+00
-3.72454971e-01 -1.08284283e+00 -3.38052958e-01 7.79830754e-01
1.28453612e-01 -5.40655136e-01 1.05265975e+00 3.15518796e-01
-1.87285662e-01 -3.70137304e-01 -8.54221761e-01 -7.44881153e-01
3.07036847e-01 -5.40042818e-01 2.13660654e-02 8.09817016e-01
-5.97529188e-02 3.81731451e-01 -4.26932454e-01 -3.08516532e-01
5.06605327e-01 -1.09331429e-01 8.42519343e-01 -9.90115762e-01
-2.60653168e-01 -6.69714212e-01 -6.44880474e-01 -1.58351791e+00
3.37704062e-01 -5.96295059e-01 -5.96502312e-02 -1.43234003e+00
-3.24697375e-01 -4.51901788e-03 -8.48181546e-02 1.62820637e-01
-3.69562358e-01 4.88227904e-01 7.29712844e-01 1.34673238e-01
-5.11754990e-01 2.41487265e-01 1.81998110e+00 1.89556092e-01
-2.25656971e-01 -1.29770324e-01 -4.24013548e-02 5.00785470e-01
7.21562684e-01 -8.77379179e-02 -5.62444866e-01 -7.78073132e-01
-8.15963522e-02 9.96506438e-02 4.08401757e-01 -1.23243153e+00
3.05081248e-01 1.19753383e-01 1.71458066e-01 -3.05951118e-01
1.72089294e-01 -6.15319431e-01 -2.92552322e-01 6.56070590e-01
-2.32505947e-01 1.65941492e-01 3.28933716e-01 2.04532593e-01
-2.71609306e-01 -1.84422404e-01 9.37083662e-01 -1.91211209e-01
-9.11839902e-01 4.07999277e-01 -2.68934339e-01 5.20364285e-01
7.96817601e-01 -2.83989072e-01 -2.91156352e-01 -4.06200707e-01
-7.65383005e-01 7.49521032e-02 1.85528666e-01 3.85023952e-01
3.33202004e-01 -1.04323614e+00 -8.72712553e-01 3.15273255e-01
-4.37109113e-01 2.26717703e-02 2.35266998e-01 8.87399554e-01
-1.26485431e+00 6.14231646e-01 -4.22104150e-01 -7.97068775e-01
-9.89445984e-01 5.11005402e-01 6.47964120e-01 -2.38453507e-01
-6.13452792e-01 8.31863940e-01 1.19290113e-01 -8.62157419e-02
1.69425964e-01 -5.14382482e-01 -5.86010329e-02 4.75305617e-02
8.03972304e-01 7.83963799e-01 3.54224700e-03 -5.71067035e-01
-3.65613848e-01 8.46886337e-01 1.71740353e-01 -1.85096473e-01
1.24429250e+00 -2.38127455e-01 -4.73876670e-02 2.19548196e-01
1.44648707e+00 -1.12475827e-01 -2.02864385e+00 1.49726376e-01
-3.72565836e-01 -8.25306177e-01 7.27118254e-02 -5.02937257e-01
-1.59849823e+00 1.18521070e+00 4.91623193e-01 2.51045197e-01
8.48876834e-01 -5.21377802e-01 9.54171896e-01 1.11811541e-01
8.99883658e-02 -9.50113356e-01 9.51243863e-02 7.60129452e-01
4.40320611e-01 -1.36333442e+00 -1.02447189e-01 -4.08296406e-01
-1.95140481e-01 1.56601334e+00 5.28833091e-01 -5.00390828e-01
7.03808129e-01 3.50186974e-01 3.48390639e-01 2.01585427e-01
-6.91904306e-01 -3.49097878e-01 2.10604787e-01 6.03045166e-01
7.18385398e-01 -5.09090066e-01 -2.10804418e-01 -6.71073079e-01
-1.85508296e-01 4.54085618e-01 7.98842728e-01 7.74501085e-01
-4.21096444e-01 -1.10861778e+00 -1.14041172e-01 -2.07051530e-01
-6.78623378e-01 -8.42601880e-02 1.51155487e-01 1.01087987e+00
4.77241501e-02 1.00200951e+00 3.73474926e-01 -8.44332129e-02
4.07942146e-01 -1.65238112e-01 8.26771915e-01 -3.01152796e-01
-6.34122491e-01 -1.58062711e-01 -2.26690508e-02 -1.17957735e+00
-9.79999900e-01 -4.09973800e-01 -1.28121436e+00 -8.71208370e-01
4.91779074e-02 1.52016971e-02 3.14208210e-01 9.66303647e-01
2.95177370e-01 6.82144642e-01 5.18314660e-01 -1.44846952e+00
-1.01117477e-01 -7.46000588e-01 -3.08837563e-01 4.43798691e-01
9.60186541e-01 -5.23780465e-01 -6.21653140e-01 4.88890201e-01] | [8.77621078491211, -1.8085678815841675] |
0e87bf7d-2221-4c98-aad2-950c432cab2d | generic-event-boundary-captioning-a-benchmark | 2204.00486 | null | https://arxiv.org/abs/2204.00486v4 | https://arxiv.org/pdf/2204.00486v4.pdf | GEB+: A Benchmark for Generic Event Boundary Captioning, Grounding and Retrieval | Cognitive science has shown that humans perceive videos in terms of events separated by the state changes of dominant subjects. State changes trigger new events and are one of the most useful among the large amount of redundant information perceived. However, previous research focuses on the overall understanding of segments without evaluating the fine-grained status changes inside. In this paper, we introduce a new dataset called Kinetic-GEB+. The dataset consists of over 170k boundaries associated with captions describing status changes in the generic events in 12K videos. Upon this new dataset, we propose three tasks supporting the development of a more fine-grained, robust, and human-like understanding of videos through status changes. We evaluate many representative baselines in our dataset, where we also design a new TPD (Temporal-based Pairwise Difference) Modeling method for visual difference and achieve significant performance improvements. Besides, the results show there are still formidable challenges for current methods in the utilization of different granularities, representation of visual difference, and the accurate localization of status changes. Further analysis shows that our dataset can drive developing more powerful methods to understand status changes and thus improve video level comprehension. The dataset is available at https://github.com/showlab/GEB-Plus | ['Mike Zheng Shou', 'Matt Feiszli', 'Stan Weixian Lei', 'Licheng Yu', 'Difei Gao', 'Yuxuan Wang'] | 2022-04-01 | null | null | null | null | ['boundary-grounding', 'boundary-captioning'] | ['computer-vision', 'computer-vision'] | [ 1.87552631e-01 -3.13320041e-01 -2.34638527e-01 -6.80628777e-01
-5.26385784e-01 -7.47082531e-01 4.18962151e-01 2.42371231e-01
-2.25583285e-01 3.31865251e-01 7.04553664e-01 -5.05110472e-02
9.46799368e-02 -4.27192971e-02 -9.11384463e-01 -2.43287891e-01
-2.38972783e-01 -1.95672885e-01 4.74961996e-01 -3.17478001e-01
4.90420938e-01 6.59763068e-02 -1.93469381e+00 9.29632604e-01
8.37711155e-01 1.07631898e+00 3.40808809e-01 7.61996865e-01
3.23842317e-01 7.67942071e-01 -7.60668874e-01 -1.53460711e-01
2.31180623e-01 -5.23184419e-01 -8.27942610e-01 -1.95728820e-02
1.05836475e+00 -7.26134539e-01 -4.95394617e-01 8.73206913e-01
6.02859855e-01 3.37103307e-01 2.17324883e-01 -1.58670616e+00
-5.79968452e-01 4.99151111e-01 -4.18822229e-01 9.13441360e-01
1.11450207e+00 4.11772311e-01 8.41246009e-01 -7.29357541e-01
8.22330117e-01 1.37906957e+00 4.02394235e-01 4.74589467e-01
-1.03723252e+00 -6.52226090e-01 6.17670774e-01 1.03719831e+00
-1.14077568e+00 -6.31441116e-01 6.38203442e-01 -6.05730116e-01
1.21910548e+00 6.32103026e-01 8.95436347e-01 1.64594710e+00
9.75860357e-02 9.63514864e-01 1.20675111e+00 -9.41852406e-02
1.06155261e-01 -1.92357838e-01 3.98186594e-01 4.15674865e-01
5.60309701e-02 2.96659861e-02 -1.17369521e+00 2.88285345e-01
4.50139940e-01 -8.18431601e-02 -7.77069330e-01 -3.07207197e-01
-1.74791312e+00 9.62919444e-02 2.89095402e-01 8.47880840e-02
-2.63388097e-01 1.63909554e-01 6.20022476e-01 2.95007706e-01
4.07925487e-01 3.07336003e-01 -5.18319070e-01 -8.11254084e-01
-8.59843850e-01 2.63689607e-01 4.85476434e-01 9.30040956e-01
4.83994663e-01 -3.94516289e-01 -6.07359529e-01 4.60871994e-01
1.36796117e-01 3.39752138e-01 3.60153019e-01 -1.38080478e+00
6.28295720e-01 3.60812217e-01 3.23410064e-01 -1.34744751e+00
-3.28370392e-01 -8.65059793e-02 -2.38718346e-01 -9.55006853e-03
3.95733416e-01 1.87350586e-01 -8.81891608e-01 1.94458616e+00
3.13666224e-01 5.17063379e-01 -3.69764864e-01 1.06511009e+00
8.54281962e-01 9.12146866e-01 1.63948834e-01 -1.47205308e-01
1.44334555e+00 -9.44539666e-01 -1.02383971e+00 -1.94469258e-01
4.03713405e-01 -5.12926996e-01 1.32005858e+00 5.97431242e-01
-1.18538213e+00 -9.63515341e-01 -1.09817934e+00 -2.23041028e-01
-3.39094698e-01 -1.10471472e-01 4.87347841e-01 3.55127305e-01
-1.33114684e+00 6.12173080e-01 -8.18069816e-01 -6.19101524e-01
3.09600383e-01 -2.76794042e-02 -3.08431745e-01 -1.83303386e-01
-1.32346678e+00 8.77662539e-01 5.41578770e-01 1.81648135e-01
-9.95456398e-01 -8.03428233e-01 -9.55135643e-01 2.10338105e-02
5.86377263e-01 -4.30783272e-01 1.29919517e+00 -1.08361793e+00
-1.06202161e+00 7.70654321e-01 -3.55578780e-01 -2.67344624e-01
4.82190728e-01 -3.81458491e-01 -5.23553133e-01 5.57508826e-01
-2.24543847e-02 1.10887241e+00 6.62685752e-01 -1.08499193e+00
-6.14899039e-01 -1.54778659e-01 3.37847143e-01 4.42235440e-01
-2.03063831e-01 1.77437708e-01 -8.07400167e-01 -7.48945475e-01
-1.72663212e-01 -7.43201017e-01 3.58089387e-01 1.38645753e-01
-6.31020069e-02 -2.15711296e-01 7.45415926e-01 -1.19890320e+00
1.60765100e+00 -2.42484283e+00 2.68637002e-01 -2.24695414e-01
3.98950189e-01 1.25553265e-01 -3.03242147e-01 4.28687453e-01
-3.09976190e-01 2.70310462e-01 3.18002939e-01 -2.10489303e-01
6.64602146e-02 -6.07591495e-02 -2.34993309e-01 2.94164240e-01
3.51978727e-02 9.40701246e-01 -1.05473053e+00 -3.19297791e-01
3.60561073e-01 3.10976774e-01 -6.20142043e-01 1.75453275e-01
-2.09021270e-01 4.90661651e-01 -1.21248048e-02 6.23711705e-01
7.14216590e-01 -1.73431098e-01 -2.24063881e-02 -8.14022779e-01
-5.53380102e-02 3.53292316e-01 -1.17697978e+00 1.98041534e+00
2.11043194e-01 1.14053833e+00 -2.44693846e-01 -7.81994581e-01
3.51591498e-01 1.68914706e-01 3.98911834e-01 -8.82267594e-01
2.42313482e-02 -3.87625456e-01 8.90800357e-02 -9.47230101e-01
7.76570201e-01 4.80253100e-01 -4.75257114e-02 1.96461067e-01
-6.84023499e-02 2.51366377e-01 5.95986545e-01 5.66659510e-01
9.80890512e-01 2.62261897e-01 3.10430914e-01 -1.77151710e-02
6.83251321e-02 -3.01738195e-02 6.15446866e-01 7.56743312e-01
-8.70209694e-01 6.71930552e-01 5.82842886e-01 -2.86422670e-01
-7.63943970e-01 -1.09920669e+00 2.14188665e-01 1.32578981e+00
7.64175653e-01 -8.19862902e-01 -6.70150399e-01 -4.10168409e-01
-1.63821951e-01 6.51330829e-01 -7.36503124e-01 -2.68710971e-01
-4.54128683e-01 -4.08499539e-01 4.86402333e-01 7.98628926e-01
7.55910516e-01 -9.20511007e-01 -6.92821622e-01 -9.37813222e-02
-9.94609714e-01 -1.28207052e+00 -7.60901451e-01 -4.37374651e-01
-6.79115891e-01 -1.07149291e+00 -3.83708835e-01 -5.00522494e-01
4.01585937e-01 6.39239669e-01 1.11376083e+00 -5.31529933e-02
-2.36442894e-01 7.19231725e-01 -6.09741092e-01 -1.69150203e-01
-1.54765517e-01 -6.19138002e-01 4.77631167e-02 -1.42363131e-01
2.95142889e-01 -1.85537562e-01 -8.90646458e-01 5.88448644e-01
-8.02510381e-01 4.54556525e-01 3.78084153e-01 3.61441553e-01
4.21315581e-01 -1.27742559e-01 2.77020037e-01 -4.03987586e-01
5.21980822e-01 -3.78176332e-01 -1.26936957e-01 3.57307494e-01
-2.08343655e-01 -2.83091486e-01 2.01464482e-02 -7.49158442e-01
-1.34639001e+00 -3.82329166e-01 2.10385412e-01 -3.31552327e-01
-3.47902894e-01 4.26750004e-01 -1.58485055e-01 1.32565290e-01
6.06267691e-01 1.24046147e-01 -3.09930801e-01 -3.00770700e-01
4.94825840e-01 4.86447632e-01 7.02905059e-01 -5.16089320e-01
1.86410069e-01 5.01649499e-01 -5.54801106e-01 -6.03521466e-01
-6.83435202e-01 -7.17876434e-01 -5.18197775e-01 -8.46628606e-01
9.54205096e-01 -1.23342228e+00 -6.93262100e-01 7.34321892e-01
-1.03567314e+00 -5.14725566e-01 -2.03436524e-01 4.73214328e-01
-6.11939371e-01 7.27537215e-01 -7.00436473e-01 -2.37746567e-01
1.64255589e-01 -1.13465691e+00 1.03163195e+00 2.94507027e-01
-4.81360525e-01 -5.97873449e-01 -3.28021124e-02 7.95176983e-01
2.62027621e-01 3.28075379e-01 6.61583543e-01 -3.53676468e-01
-7.77971148e-01 5.36174104e-02 -2.86101341e-01 2.92473733e-01
2.77594998e-02 1.12379111e-01 -7.83676863e-01 -3.97354603e-01
-1.29492909e-01 -3.59466136e-01 8.43734741e-01 4.51228023e-01
1.38594675e+00 -6.32636026e-02 -3.35587472e-01 4.50967878e-01
9.82691765e-01 4.01149660e-01 8.93642902e-01 1.73039630e-01
6.46652102e-01 5.48857212e-01 1.00645256e+00 4.33333755e-01
6.78077817e-01 7.86170840e-01 3.55870664e-01 1.02475122e-01
-4.51584727e-01 -1.78407013e-01 8.08997750e-01 8.49029243e-01
-2.66242921e-01 -6.77336335e-01 -7.55500853e-01 5.56575298e-01
-1.89658737e+00 -1.37515235e+00 -5.58238961e-02 2.02193856e+00
8.50270152e-01 1.07548922e-01 -5.75787127e-02 -3.11951637e-02
8.45310867e-01 3.58771771e-01 -7.51577795e-01 -2.66619921e-01
-4.20153961e-02 -2.85389036e-01 9.37293395e-02 3.39891165e-01
-9.89282370e-01 8.19864869e-01 6.73256588e+00 6.29445374e-01
-1.04296803e+00 -1.83133222e-02 5.30522048e-01 -3.97996455e-01
-6.86811805e-02 -1.00102410e-01 -5.85721731e-01 7.11063147e-01
9.13506269e-01 -1.17185563e-01 3.87178868e-01 3.80115330e-01
8.01960170e-01 -5.76971889e-01 -1.39066327e+00 1.24374592e+00
4.66372341e-01 -1.09392476e+00 1.13401040e-01 -2.81397849e-01
6.08833790e-01 -6.49386346e-02 5.54025499e-03 3.58710438e-01
-2.98983902e-01 -7.76856661e-01 1.17990375e+00 5.92326283e-01
7.83188939e-01 -1.07086726e-01 2.77968228e-01 2.87154224e-02
-1.25278986e+00 4.71945181e-02 -6.06360063e-02 -3.41821969e-01
3.23455036e-01 2.17597529e-01 -5.97971737e-01 3.71068090e-01
1.21958733e+00 1.24889147e+00 -9.74320710e-01 1.10334539e+00
-1.04358375e-01 7.14658082e-01 -9.51218456e-02 2.22294927e-01
-2.75247753e-01 2.35690132e-01 6.19262695e-01 1.27263439e+00
2.59744763e-01 2.33634621e-01 2.29362026e-01 5.32628834e-01
1.18752852e-01 -2.41649151e-01 -2.57584572e-01 -8.54827017e-02
5.45037985e-01 9.55831647e-01 -7.66409039e-01 -7.46562243e-01
-5.12668490e-01 1.24985015e+00 1.55722573e-01 5.96293092e-01
-1.31480479e+00 -1.37844816e-01 9.28342044e-01 -3.71103808e-02
3.44380438e-01 -3.58316481e-01 1.08402103e-01 -1.58464956e+00
2.83686161e-01 -1.22517252e+00 5.08208156e-01 -1.35409188e+00
-9.35890317e-01 2.06998408e-01 5.73902488e-01 -1.21528661e+00
-3.38759162e-02 -6.06900215e-01 -3.93110096e-01 4.11616623e-01
-1.35370088e+00 -8.08629155e-01 -9.82174993e-01 6.61691666e-01
9.44861531e-01 5.87210894e-01 3.27208370e-01 3.91141653e-01
-6.45496607e-01 3.74047995e-01 -2.52698213e-01 -1.98218361e-01
1.31776619e+00 -1.13978672e+00 4.28757846e-01 1.10141790e+00
-4.42828313e-02 5.74194610e-01 9.48949575e-01 -7.95713484e-01
-1.14300787e+00 -6.89712167e-01 4.79048222e-01 -7.33888030e-01
3.99466127e-01 -4.68614697e-01 -9.75972772e-01 8.84245157e-01
4.65067804e-01 -2.30174512e-01 6.23082519e-01 -8.63021463e-02
-4.47317272e-01 -6.26723617e-02 -8.25416327e-01 6.34143591e-01
1.43869865e+00 -5.19165516e-01 -7.91374564e-01 7.32581504e-03
8.52986872e-01 -5.99761248e-01 -7.07576394e-01 4.55192596e-01
6.83850288e-01 -1.22167754e+00 9.63981390e-01 -5.53093135e-01
3.90507281e-01 -5.82707047e-01 -5.77283725e-02 -1.39060771e+00
-4.31126684e-01 -4.53031957e-01 -3.03165913e-01 1.20515931e+00
9.69373994e-03 -2.22612113e-01 3.73384476e-01 8.24088693e-01
-2.71549046e-01 -4.49082017e-01 -5.31046629e-01 -7.37142801e-01
-6.15450263e-01 -5.81357300e-01 2.73247182e-01 8.87068748e-01
2.73577213e-01 -3.01721729e-02 -5.29416203e-01 1.71536028e-01
2.33484268e-01 -7.21440464e-03 5.43971002e-01 -8.15614462e-01
-2.38444000e-01 -3.10711384e-01 -6.47165775e-01 -1.38489532e+00
-2.10612416e-01 -4.58596766e-01 9.02239531e-02 -1.56486034e+00
5.59499502e-01 2.46683642e-01 -4.09716517e-01 5.15854657e-01
-4.49058354e-01 1.48045987e-01 4.30187076e-01 7.36280158e-02
-1.21982861e+00 4.00532126e-01 1.36147094e+00 -8.16871822e-02
-8.79008174e-02 -6.02503538e-01 -6.52444482e-01 6.39143407e-01
6.63531363e-01 5.53312562e-02 -5.57667851e-01 -5.54203749e-01
2.77291238e-01 2.14739498e-02 5.74427843e-01 -1.05829358e+00
1.41180903e-01 -2.50600040e-01 4.75318849e-01 -8.05737257e-01
5.10870337e-01 -4.88969922e-01 2.19652534e-01 3.12726945e-01
-4.67760235e-01 3.20126951e-01 5.27001858e-01 7.05709219e-01
-3.23622495e-01 2.30441570e-01 3.72558802e-01 3.00038084e-02
-1.71179724e+00 1.15620941e-02 -4.78261530e-01 1.68686390e-01
1.28342032e+00 -5.10116100e-01 -7.60883510e-01 -6.41167045e-01
-7.17851281e-01 3.69633287e-01 4.66721147e-01 8.08183789e-01
6.74109101e-01 -1.20156837e+00 -4.49831039e-01 -3.44409719e-02
3.57292354e-01 -3.96361142e-01 8.45080793e-01 8.86391342e-01
-5.00433624e-01 2.32606530e-01 -6.07147098e-01 -7.20798254e-01
-1.69156075e+00 6.58418894e-01 1.54024854e-01 2.81543314e-01
-4.30875301e-01 8.83517027e-01 4.43273485e-01 3.03194463e-01
3.95536572e-01 -8.90887201e-01 -2.24699095e-01 3.29550385e-01
7.86574960e-01 5.95927835e-01 -1.55251831e-01 -6.61554456e-01
-4.54019904e-01 4.10316765e-01 -1.76182538e-01 7.54460096e-02
9.71926689e-01 -6.91280603e-01 1.20770566e-01 5.05733252e-01
9.68346238e-01 -1.24136925e-01 -1.73799849e+00 1.19392723e-01
-2.73837686e-01 -7.56372869e-01 -1.39647722e-01 -1.25079095e+00
-7.02675819e-01 7.59623110e-01 8.28697801e-01 -7.13756233e-02
1.49106574e+00 1.61609426e-01 7.05932260e-01 1.84462801e-01
2.17857420e-01 -1.00763965e+00 5.26292741e-01 2.28429615e-01
9.80340362e-01 -1.36878693e+00 -1.37143329e-01 -5.96435666e-01
-7.87254691e-01 7.19032347e-01 9.53319550e-01 3.66555423e-01
1.65564090e-01 -1.39157236e-01 1.23556867e-01 -1.67725861e-01
-1.11568546e+00 -1.62202537e-01 5.68870544e-01 5.97246706e-01
3.14047277e-01 -8.48895907e-02 -1.89075291e-01 5.04151821e-01
-9.10975505e-03 4.12669182e-02 6.49397790e-01 9.61630762e-01
-3.92141581e-01 -7.62088001e-01 -2.89279968e-01 2.47546017e-01
-1.92602754e-01 -3.52529511e-02 -2.87390649e-01 6.08742476e-01
2.35908568e-01 1.14478183e+00 1.16677329e-01 -5.14431238e-01
4.23225194e-01 2.91665010e-02 6.96189642e-01 -4.08830106e-01
-2.46019036e-01 -1.77706555e-01 1.71184003e-01 -1.30862689e+00
-7.65524626e-01 -7.86252975e-01 -9.70865607e-01 -2.80540377e-01
-1.77888386e-02 -2.79059559e-01 4.53941673e-01 7.20506430e-01
7.10296988e-01 7.08890617e-01 1.72782898e-01 -9.74929869e-01
-8.75084549e-02 -9.84522820e-01 -2.43430987e-01 1.02586699e+00
2.15188652e-01 -7.96207249e-01 -3.48108053e-01 5.67780197e-01] | [9.599076271057129, 0.6333010792732239] |
f947f17c-54b5-4b82-9e66-01cf5eb2e5f7 | revisiting-click-based-interactive-video | 2203.01784 | null | https://arxiv.org/abs/2203.01784v2 | https://arxiv.org/pdf/2203.01784v2.pdf | Revisiting Click-based Interactive Video Object Segmentation | While current methods for interactive Video Object Segmentation (iVOS) rely on scribble-based interactions to generate precise object masks, we propose a Click-based interactive Video Object Segmentation (CiVOS) framework to simplify the required user workload as much as possible. CiVOS builds on de-coupled modules reflecting user interaction and mask propagation. The interaction module converts click-based interactions into an object mask, which is then inferred to the remaining frames by the propagation module. Additional user interactions allow for a refinement of the object mask. The approach is extensively evaluated on the popular interactive~DAVIS dataset, but with an inevitable adaptation of scribble-based interactions with click-based counterparts. We consider several strategies for generating clicks during our evaluation to reflect various user inputs and adjust the DAVIS performance metric to perform a hardware-independent comparison. The presented CiVOS pipeline achieves competitive results, although requiring a lower user workload. | ['Rainer Stiefelhagen', 'Michael Arens', 'Norbert Scherer-Negenborn', 'Stefan Becker', 'Sebastian Bullinger', 'Stephane Vujasinovic'] | 2022-03-03 | null | null | null | null | ['interactive-video-object-segmentation'] | ['computer-vision'] | [ 4.98275936e-01 -1.46852478e-01 2.83280760e-02 -5.60399652e-01
-4.96296495e-01 -7.89219737e-01 5.95442116e-01 2.61242986e-01
-7.76088893e-01 2.96132445e-01 -3.54365796e-01 -3.89268547e-01
2.54631668e-01 -6.04121625e-01 -7.14025080e-01 -2.41083637e-01
1.41042277e-01 6.62380457e-01 1.37178266e+00 5.62468916e-02
5.34606695e-01 7.34759152e-01 -1.65017629e+00 6.57124937e-01
7.68709123e-01 8.15274119e-01 5.10200441e-01 1.25281858e+00
-4.29354757e-01 4.96158928e-01 -8.24883640e-01 -2.59894460e-01
4.32197392e-01 -3.31978768e-01 -7.41956353e-01 2.89198965e-01
4.88445997e-01 -4.34992254e-01 3.91583964e-02 7.41098583e-01
3.30427319e-01 9.80205759e-02 4.63256329e-01 -1.23420346e+00
7.78717771e-02 7.95372903e-01 -4.69106555e-01 3.15063328e-01
5.58076084e-01 4.79710460e-01 8.51188660e-01 -9.26533401e-01
7.91255295e-01 1.01180112e+00 4.30405587e-01 3.89142424e-01
-1.35873067e+00 -5.28433442e-01 2.61227399e-01 1.81277767e-01
-1.49606919e+00 -3.57295364e-01 5.29184103e-01 -6.17709279e-01
8.12616944e-01 6.88127518e-01 7.80791402e-01 5.89028478e-01
-1.06777914e-01 1.00158107e+00 8.98185968e-01 -4.62476224e-01
4.34840024e-01 4.11519259e-01 4.01227713e-01 6.14534378e-01
1.81633085e-01 -4.36622947e-01 -5.19070506e-01 -4.45807278e-02
1.02045906e+00 -1.54090792e-01 -3.13191026e-01 -4.46366817e-01
-1.19310153e+00 3.53442430e-01 2.18066335e-01 5.00170104e-02
-2.10083425e-01 3.94477755e-01 3.09983641e-01 1.06781803e-01
3.25852782e-02 2.79019773e-01 -3.93153578e-01 -2.47041211e-01
-1.54538727e+00 5.39917767e-01 1.02431881e+00 1.28678560e+00
9.08329844e-01 -2.14856431e-01 -6.37195885e-01 5.63650191e-01
2.86525577e-01 -9.59590226e-02 3.85446660e-02 -9.67785299e-01
2.98893124e-01 6.19963527e-01 4.43201482e-01 -8.55174839e-01
-2.20748276e-01 -2.72771180e-01 9.18003693e-02 6.33321166e-01
5.54921091e-01 -8.93380567e-02 -1.15286565e+00 1.24846280e+00
4.38806921e-01 1.25730142e-01 -5.65519929e-01 9.30217922e-01
6.46088004e-01 4.73632395e-01 1.89546317e-01 3.81979235e-02
1.43651676e+00 -1.10872495e+00 -5.71563244e-01 -2.53508538e-02
5.16062200e-01 -7.64427841e-01 1.30568385e+00 6.13062859e-01
-1.27687538e+00 -6.48022592e-01 -1.17805278e+00 -1.68172926e-01
-1.78807944e-01 3.08661222e-01 3.85985345e-01 7.91761756e-01
-1.11755872e+00 6.19338989e-01 -1.09943056e+00 -3.83430302e-01
3.15902680e-01 5.71960092e-01 1.52123541e-01 5.58644652e-01
-4.97224987e-01 4.68977511e-01 4.59181488e-01 -2.14768127e-02
-7.65590847e-01 -7.15268195e-01 -4.87776726e-01 8.74565393e-02
7.32007623e-01 -5.88178158e-01 1.35055530e+00 -1.03524959e+00
-1.68217397e+00 7.06670702e-01 -1.41811013e-01 -5.46296895e-01
1.08384132e+00 -3.91216278e-01 1.04827158e-01 4.68353450e-01
-2.28523463e-01 1.07448316e+00 1.00011182e+00 -1.48276818e+00
-7.62668669e-01 1.15751289e-01 4.03459460e-01 8.51241723e-02
-4.36073765e-02 2.33340207e-02 -1.39577270e+00 -6.44457042e-01
-1.16501629e-01 -1.01397610e+00 -2.14990526e-01 1.13268279e-01
-5.31440318e-01 -3.33605371e-02 1.13789582e+00 -3.18470865e-01
1.61557555e+00 -2.09240127e+00 1.91583097e-01 1.75367549e-01
2.18468979e-01 3.14160287e-01 -6.98976442e-02 3.33545685e-01
1.40300184e-01 1.54456764e-01 -1.71220362e-01 -6.73829079e-01
-6.63870201e-02 2.01395666e-03 -7.73465037e-02 1.49935961e-01
1.44770160e-01 6.05038464e-01 -6.89440668e-01 -7.83749640e-01
4.87257570e-01 2.30658069e-01 -1.17885900e+00 3.34679633e-01
-7.14545310e-01 3.03552568e-01 -3.31217319e-01 5.01245558e-01
7.66848743e-01 -1.01354182e-01 1.33249655e-01 -4.31965798e-01
-4.22276199e-01 2.90646963e-02 -1.41551554e+00 1.80039728e+00
-2.81103641e-01 7.70773530e-01 1.91037059e-01 -4.78520691e-01
6.88835621e-01 4.61046472e-02 3.82605672e-01 -3.93821113e-02
3.51327479e-01 5.08813560e-02 -8.11586305e-02 -5.18945277e-01
7.77073324e-01 5.08497059e-01 1.72815278e-01 3.47701907e-01
-8.57224129e-03 -2.54328519e-01 5.99197030e-01 5.59014618e-01
1.14557850e+00 7.02473998e-01 -2.37410069e-02 -3.81836861e-01
5.74885011e-01 1.93609238e-01 2.72984684e-01 9.22468424e-01
-2.12053299e-01 9.02682126e-01 6.49013519e-01 -8.63894522e-02
-8.14116180e-01 -1.04919255e+00 -9.32194851e-03 1.23347127e+00
4.58607286e-01 -7.55233765e-01 -1.22728896e+00 -8.96133423e-01
-3.58478785e-01 8.27820957e-01 -3.68102849e-01 5.45437694e-01
-6.55792892e-01 -3.87870789e-01 2.69087464e-01 4.20440853e-01
3.10204268e-01 -1.01000154e+00 -1.07268238e+00 2.30611518e-01
2.49412179e-01 -1.13018310e+00 -6.80874646e-01 5.58466129e-02
-7.28354335e-01 -9.44188833e-01 -5.43424547e-01 -4.10329729e-01
8.04444194e-01 2.03078523e-01 1.05287993e+00 2.47310013e-01
-5.79311550e-01 4.72349674e-01 -5.23282588e-01 -2.88409650e-01
-3.92169386e-01 2.21604511e-01 -3.42286497e-01 1.63889453e-01
1.09076947e-01 -4.33137566e-01 -9.45803583e-01 5.31774879e-01
-1.22831309e+00 3.61638159e-01 2.71187037e-01 2.87595451e-01
5.76787114e-01 -3.10720742e-01 -1.26455734e-02 -1.19819045e+00
2.29509264e-01 -1.19529724e-01 -9.17568445e-01 -1.56002585e-03
-3.33418101e-01 2.72887535e-02 3.83024007e-01 -5.38287699e-01
-9.41633642e-01 4.45411265e-01 -2.48359799e-01 -4.33540761e-01
-1.83094189e-01 5.43638058e-02 -1.93137124e-01 -1.13630190e-01
6.85351849e-01 -1.14286393e-01 -5.38517356e-01 -6.19772971e-01
4.02926058e-01 6.13768876e-01 5.02840042e-01 -5.29981256e-01
6.23623788e-01 3.75158221e-01 -6.14470661e-01 -7.17187941e-01
-3.59851450e-01 -4.51464355e-01 -7.77427852e-01 -4.12846535e-01
1.11220300e+00 -5.18901408e-01 -6.97282374e-01 2.79746026e-01
-1.24571359e+00 -5.13191164e-01 -3.16966236e-01 2.65506446e-01
-5.08557916e-01 2.22737268e-01 -4.74074155e-01 -6.65585160e-01
-1.37472386e-02 -1.68321431e+00 1.19778776e+00 3.35449666e-01
-5.24291813e-01 -4.54922676e-01 -2.40619540e-01 2.51016051e-01
1.54815242e-01 1.66535955e-02 5.77936471e-01 -7.56001532e-01
-1.28245175e+00 -2.49778956e-01 -4.36464131e-01 -2.21427809e-02
-1.51639372e-01 5.07550299e-01 -9.97174323e-01 -9.91273895e-02
-4.67960030e-01 2.03220963e-01 6.27553999e-01 1.93775132e-01
1.40319264e+00 1.59507558e-01 -4.67749298e-01 5.72788656e-01
1.44844592e+00 4.52505499e-01 6.77676976e-01 2.61082023e-01
7.20917165e-01 2.13320822e-01 5.09626269e-01 4.50718015e-01
1.98473305e-01 9.50208664e-01 3.91956091e-01 -2.74294261e-02
-2.72620350e-01 -1.02456518e-01 2.50056088e-01 3.70273262e-01
-1.28962398e-01 -5.82687974e-01 -7.89468527e-01 3.68440002e-01
-1.62796021e+00 -6.79020166e-01 -4.02042389e-01 2.24080729e+00
7.53988683e-01 5.55951893e-01 4.54275697e-01 1.31858975e-01
5.76279223e-01 1.75384462e-01 -3.75551432e-01 -5.36745846e-01
5.01923501e-01 3.21340472e-01 5.55849850e-01 7.16755390e-01
-8.59968960e-01 1.09996986e+00 6.57021046e+00 7.67759740e-01
-1.14572072e+00 2.09213421e-01 4.48501915e-01 -1.91233575e-01
-2.75183022e-01 1.54523000e-01 -1.14728534e+00 2.46226773e-01
5.34404755e-01 4.73470539e-02 2.82035828e-01 7.49144018e-01
5.24403632e-01 -6.31203055e-01 -1.38863385e+00 9.15528655e-01
-8.10920745e-02 -1.40867960e+00 8.37134570e-02 -3.38055432e-01
3.63305420e-01 -3.97336632e-02 -3.43956172e-01 1.62945762e-01
1.15665808e-01 -3.58707994e-01 1.10248923e+00 4.30141777e-01
7.93565512e-01 -3.99897069e-01 1.41346276e-01 1.58702090e-01
-1.13957238e+00 2.99313754e-01 1.30749404e-01 9.76811051e-02
3.17659706e-01 2.87424952e-01 -1.07208979e+00 1.39095291e-01
6.38981342e-01 1.35386556e-01 -8.20647061e-01 1.28169847e+00
1.30153775e-01 6.77815199e-01 -4.21823591e-01 -7.39444941e-02
6.22830689e-02 -1.02880605e-01 7.20968306e-01 1.66501200e+00
-6.64923936e-02 -3.76091786e-02 3.24393809e-01 9.31403935e-01
7.69521967e-02 1.46686643e-01 4.29410525e-02 -7.39079937e-02
3.72061551e-01 1.34876156e+00 -1.49380791e+00 -5.25180340e-01
-3.35799962e-01 1.53466642e+00 8.29556361e-02 4.68875229e-01
-1.26913035e+00 -5.88857174e-01 3.51409733e-01 6.36551738e-01
6.07894123e-01 -4.70239639e-01 -3.35323334e-01 -9.94055569e-01
-1.77920628e-02 -6.58040166e-01 1.19695082e-01 -7.40135670e-01
-4.05591756e-01 7.12548971e-01 3.65835071e-01 -9.52483535e-01
-4.04347777e-02 -3.83005708e-01 -6.78705275e-01 6.51867390e-01
-8.49503994e-01 -8.46245706e-01 -6.10931635e-01 2.89783388e-01
9.78743017e-01 3.70213538e-01 3.55562955e-01 4.25473839e-01
-6.21570110e-01 5.53711116e-01 -5.31371057e-01 -1.40897296e-02
4.92671967e-01 -1.43481982e+00 5.52934706e-01 8.86812985e-01
2.29114786e-01 6.15341425e-01 9.65566218e-01 -7.32656419e-01
-1.25478184e+00 -8.66057873e-01 3.62142891e-01 -4.94747937e-01
3.20107460e-01 -6.38211489e-01 -7.83886790e-01 7.25641847e-01
2.54483730e-01 2.41494980e-02 3.60400319e-01 -3.62795591e-01
-7.74100274e-02 -6.63635805e-02 -9.54298079e-01 9.22400117e-01
1.13212895e+00 -3.07230860e-01 -3.17405134e-01 7.40575045e-02
7.35313773e-01 -5.34286141e-01 -5.17182350e-01 2.05271676e-01
5.73197007e-01 -1.25914454e+00 8.41117859e-01 -1.47875831e-01
1.64105557e-02 -9.16170776e-01 1.56213000e-01 -7.05447316e-01
1.91702060e-02 -9.33430731e-01 9.05722976e-02 1.27262187e+00
4.57059979e-01 -4.78288606e-02 1.03425479e+00 8.20255935e-01
-1.68138921e-01 -6.67806387e-01 -4.49639708e-01 -3.32601756e-01
-7.85927832e-01 -6.95489705e-01 3.12864810e-01 3.63843143e-01
-1.36456132e-01 -9.15060192e-03 3.12270373e-02 2.37037823e-01
4.11793739e-01 -3.54441106e-02 1.17976499e+00 -8.31878603e-01
-6.58269584e-01 -5.01348376e-01 -3.40453029e-01 -1.47714436e+00
-4.07801330e-01 -5.65380871e-01 1.40598282e-01 -1.30917227e+00
6.63536862e-02 -6.20391250e-01 8.39844272e-02 1.24716178e-01
-2.44039536e-01 7.33179510e-01 4.55534309e-01 7.23832324e-02
-8.42806220e-01 -8.22564811e-02 9.53244567e-01 1.34126201e-01
-9.09316003e-01 1.33800730e-01 -1.36379465e-01 9.26161230e-01
3.77879441e-01 -3.77339065e-01 -5.29171765e-01 -4.49955076e-01
8.21582302e-02 4.13374566e-02 3.55538666e-01 -1.31782162e+00
2.79238969e-01 -8.96586031e-02 2.16720253e-01 -9.34964895e-01
3.11748326e-01 -8.07210147e-01 3.50084126e-01 4.05322790e-01
-3.99945706e-01 -6.99305311e-02 3.19051087e-01 4.61463928e-01
1.24176271e-01 -5.33975780e-01 7.90879607e-01 -2.33527068e-02
-6.16073012e-01 2.63330221e-01 -6.75681353e-01 -2.68354833e-01
1.18365633e+00 -6.87753797e-01 1.92463711e-01 -1.78263202e-01
-1.15672910e+00 4.20566313e-02 6.57747746e-01 2.55781561e-01
4.51204062e-01 -7.38496542e-01 -3.06430995e-01 2.28712872e-01
-3.69666778e-02 5.93214743e-02 -4.80505712e-02 7.13628411e-01
-1.25956929e+00 1.21405073e-01 8.77970234e-02 -8.15286219e-01
-1.46454573e+00 5.06760001e-01 1.76645800e-01 -4.11785673e-03
-6.85219049e-01 1.05277526e+00 3.80002648e-01 1.96252704e-01
4.33190554e-01 -7.37479031e-01 2.80971210e-02 5.64835072e-02
5.47885954e-01 5.04680634e-01 7.25898221e-02 -8.38953182e-02
-2.85652965e-01 2.88507700e-01 -2.07945079e-01 -5.47852576e-01
9.69371736e-01 -2.54885912e-01 3.40893924e-01 4.33647275e-01
7.11399078e-01 3.45963657e-01 -1.56019437e+00 1.89371780e-01
1.69124827e-01 -6.28193319e-01 -9.35282186e-02 -7.51395047e-01
-8.63240302e-01 6.85922027e-01 5.04490912e-01 2.76049286e-01
1.13412642e+00 -1.62843451e-01 6.57024384e-01 6.88988417e-02
5.31893373e-01 -1.02053785e+00 5.78853264e-02 2.25430317e-02
7.25553274e-01 -7.23605454e-01 1.67717710e-01 -9.27137971e-01
-4.62122202e-01 1.02875412e+00 7.75837421e-01 -1.83979124e-01
4.85467196e-01 6.88443124e-01 -5.28112799e-02 2.30461750e-02
-4.74417478e-01 -1.57741129e-01 1.38310552e-01 2.45921090e-01
4.01334941e-01 -2.73107708e-01 -7.43783891e-01 4.05188173e-01
1.25759080e-01 2.73155391e-01 5.42629063e-01 9.44307327e-01
-5.09483099e-01 -1.25519991e+00 -4.07372028e-01 3.23474109e-01
-4.48293030e-01 -1.59365460e-01 -3.46488386e-01 7.49960005e-01
1.82866141e-01 6.91161096e-01 2.87222505e-01 -3.10674489e-01
2.14472964e-01 -2.99468916e-02 5.37500560e-01 -9.09094274e-01
-9.52982843e-01 4.71772164e-01 1.49348006e-01 -8.65507483e-01
-3.73152196e-01 -6.38688028e-01 -1.41832101e+00 9.29361209e-02
-4.74075705e-01 1.16543986e-01 7.96315372e-01 7.37305462e-01
4.39847738e-01 6.20545566e-01 1.89406484e-01 -1.29801488e+00
1.20485723e-01 -7.35985935e-01 -6.23310246e-02 3.42318475e-01
4.42719497e-02 -4.49296981e-01 -8.46581385e-02 5.09373665e-01] | [9.258901596069336, -0.17697866261005402] |
6d78856a-ee01-4ed1-ae6d-c2189a75a0b0 | guided-deep-generative-model-based-spatial | 2306.17197 | null | https://arxiv.org/abs/2306.17197v1 | https://arxiv.org/pdf/2306.17197v1.pdf | Guided Deep Generative Model-based Spatial Regularization for Multiband Imaging Inverse Problems | When adopting a model-based formulation, solving inverse problems encountered in multiband imaging requires to define spatial and spectral regularizations. In most of the works of the literature, spectral information is extracted from the observations directly to derive data-driven spectral priors. Conversely, the choice of the spatial regularization often boils down to the use of conventional penalizations (e.g., total variation) promoting expected features of the reconstructed image (e.g., piecewise constant). In this work, we propose a generic framework able to capitalize on an auxiliary acquisition of high spatial resolution to derive tailored data-driven spatial regularizations. This approach leverages on the ability of deep learning to extract high level features. More precisely, the regularization is conceived as a deep generative network able to encode spatial semantic features contained in this auxiliary image of high spatial resolution. To illustrate the versatility of this approach, it is instantiated to conduct two particular tasks, namely multiband image fusion and multiband image inpainting. Experimental results obtained on these two tasks demonstrate the benefit of this class of informed regularizations when compared to more conventional ones. | ['Jie Chen', 'Nicolas Dobigeon', 'Min Zhao'] | 2023-06-29 | null | null | null | null | ['image-inpainting'] | ['computer-vision'] | [ 8.10760140e-01 2.22434893e-01 5.31415381e-02 -2.25172341e-01
-1.05328047e+00 -2.80810386e-01 7.69533515e-01 -1.71211641e-02
-3.53880376e-01 8.94265652e-01 -4.52290475e-02 1.63551971e-01
-6.62943304e-01 -7.28864014e-01 -7.09437728e-01 -1.07260561e+00
2.27384850e-01 -8.81552473e-02 -1.50939345e-01 -2.46737301e-01
1.83278665e-01 7.35020697e-01 -1.59243965e+00 1.96814254e-01
1.02858770e+00 1.07912350e+00 5.67089140e-01 3.67560863e-01
1.35667965e-01 5.73111713e-01 -1.25245601e-01 -8.23521838e-02
2.63020575e-01 -5.94956100e-01 -6.74198627e-01 5.65862358e-01
1.41314015e-01 7.78708467e-03 2.78620087e-02 1.06991696e+00
1.72794491e-01 2.49008149e-01 7.13422298e-01 -6.88868582e-01
-5.57915151e-01 2.15053782e-01 -6.27570927e-01 5.09863384e-02
2.76556104e-01 -2.01959796e-02 9.10320282e-01 -9.98695970e-01
6.11655414e-01 8.48071456e-01 7.30504096e-01 1.10556148e-01
-1.77981412e+00 1.11206040e-01 2.96521131e-02 1.06996074e-02
-1.26237953e+00 -6.11826241e-01 1.25563097e+00 -5.63609183e-01
2.45828301e-01 2.92065442e-01 4.87076998e-01 1.10566294e+00
-1.20954543e-01 4.29094493e-01 1.48595405e+00 -7.52825379e-01
2.94957668e-01 2.23541826e-01 -1.50742298e-02 3.86910558e-01
9.70668793e-02 3.05346310e-01 -5.18265128e-01 -1.60420351e-02
8.13440621e-01 -1.37937471e-01 -6.81346416e-01 -6.96325779e-01
-9.07935500e-01 8.42407703e-01 4.33511615e-01 6.25034928e-01
-6.92137063e-01 -3.76316272e-02 2.51728464e-02 1.33802816e-01
8.40694904e-01 3.68969798e-01 -8.11146349e-02 2.71569729e-01
-1.17405188e+00 2.10576698e-01 3.69648278e-01 6.15137756e-01
1.05386603e+00 1.46774098e-01 -1.58274189e-01 8.72957528e-01
3.56968522e-01 1.23312026e-01 7.41144791e-02 -9.53280807e-01
8.99371803e-02 1.95281580e-01 2.90823549e-01 -9.03030992e-01
-4.26674724e-01 -7.56308198e-01 -8.65597367e-01 3.11222732e-01
3.66274238e-01 -2.97571700e-02 -8.28063905e-01 1.95115471e+00
3.23306441e-01 4.05658782e-01 8.56786445e-02 1.02914023e+00
2.92007178e-01 4.49188113e-01 -1.39036579e-02 -4.68393266e-01
1.09564853e+00 -4.93313581e-01 -7.40516007e-01 -1.63763002e-01
2.05050886e-01 -7.17553079e-01 1.08234882e+00 4.32625294e-01
-1.22167218e+00 -5.63268363e-01 -1.05616772e+00 3.68302967e-03
-3.37009043e-01 2.51921296e-01 3.74971151e-01 5.95251977e-01
-9.82990921e-01 7.36904263e-01 -7.32940555e-01 -2.58576453e-01
3.27444673e-01 1.06907403e-02 -2.46865466e-01 -5.73336519e-02
-9.78413165e-01 8.76374900e-01 4.15497452e-01 3.14736962e-01
-5.02803087e-01 -9.23821986e-01 -7.51378775e-01 2.95971688e-02
4.16353106e-01 -9.12650645e-01 7.69654632e-01 -1.31412911e+00
-1.48230946e+00 9.55725670e-01 7.61771947e-02 -5.31263351e-01
7.87504792e-01 -9.25219506e-02 -1.09218866e-01 5.18386722e-01
1.48719683e-01 4.39527668e-02 1.46736109e+00 -1.68693864e+00
-2.04065323e-01 -2.97629654e-01 1.68491140e-01 -2.74235066e-02
-2.31460869e-01 -2.93799490e-01 -1.56451911e-01 -8.89163375e-01
1.99844524e-01 -7.15214670e-01 -2.96900272e-01 -5.17884195e-02
-3.86314273e-01 5.16825616e-01 4.78166193e-01 -6.42936707e-01
8.01085353e-01 -2.14881110e+00 6.11746132e-01 2.65029192e-01
8.28822479e-02 3.34080420e-02 -4.73718531e-02 5.60208619e-01
-3.94906372e-01 -3.01433951e-01 -8.42566431e-01 -4.86070544e-01
-4.86810915e-02 1.26176506e-01 -3.50580186e-01 8.56786489e-01
5.00218987e-01 5.90140104e-01 -7.90484667e-01 -1.96097121e-01
6.75364733e-01 1.05950391e+00 -6.70966387e-01 1.62942320e-01
-1.72342435e-01 1.23396075e+00 -6.28506958e-01 1.50489569e-01
8.36972535e-01 -3.15322608e-01 1.35098338e-01 -5.67442358e-01
-4.14409667e-01 -8.23228657e-02 -1.13181984e+00 2.10067153e+00
-9.45160329e-01 1.83957353e-01 5.42208493e-01 -1.46463275e+00
6.30697310e-01 2.66521007e-01 8.01361084e-01 -4.85352218e-01
1.01260908e-01 4.50250983e-01 -5.07548094e-01 -4.26217198e-01
3.35323006e-01 -4.84311432e-01 2.38019377e-01 1.33132607e-01
1.58025056e-01 -2.27410123e-01 3.59620852e-03 -2.35490829e-01
5.66820800e-01 5.80340862e-01 3.99268657e-01 -4.60344374e-01
1.08792472e+00 -1.36400491e-01 2.20379680e-01 7.95780361e-01
3.14634502e-01 8.00076127e-01 2.19050720e-01 -1.61364064e-01
-1.12496042e+00 -9.96242940e-01 -4.66709197e-01 8.34526122e-01
-8.12766925e-02 -1.93073340e-02 -8.63471210e-01 -2.29827315e-01
-2.19323426e-01 5.82995474e-01 -6.54668152e-01 -1.51967540e-01
-6.51378810e-01 -9.72918034e-01 3.93525511e-02 -2.70467307e-02
5.01401365e-01 -6.95827007e-01 -8.92260373e-01 2.46719658e-01
-1.93517923e-01 -1.04352224e+00 9.11726207e-02 3.68751258e-01
-7.78464854e-01 -9.52614605e-01 -9.65667605e-01 -2.71996796e-01
3.93923312e-01 2.99908966e-01 1.01059842e+00 -1.94957569e-01
-4.66688395e-01 7.22688913e-01 -4.79693025e-01 1.27346352e-01
-2.81962454e-01 3.33341882e-02 -2.89812595e-01 7.13058889e-01
-3.15389335e-01 -1.19248509e+00 -6.00126326e-01 -2.72399575e-01
-1.36217916e+00 1.46871820e-01 8.14012170e-01 1.14150345e+00
6.15867555e-01 -1.72306895e-01 7.00646102e-01 -1.02559888e+00
4.55456823e-01 -4.98786479e-01 -6.61158025e-01 2.23781347e-01
-5.06344974e-01 2.10201785e-01 7.06307650e-01 -2.00182617e-01
-1.33424246e+00 2.50165582e-01 -3.39017153e-01 -3.91886562e-01
-2.96172351e-01 6.39030635e-01 -1.62356466e-01 -4.19443578e-01
4.31983739e-01 5.74153960e-01 -3.85986082e-03 -7.20032632e-01
6.51889801e-01 3.44246984e-01 5.53408325e-01 -7.28540659e-01
8.56599033e-01 8.09019744e-01 2.94151485e-01 -1.18229723e+00
-9.53100920e-01 -2.60373175e-01 -7.33242512e-01 -2.67801881e-01
9.88606691e-01 -7.48883367e-01 -4.61425602e-01 7.55166784e-02
-1.04085660e+00 -2.08021939e-01 -5.92166543e-01 4.41986740e-01
-1.10472119e+00 5.35701275e-01 -3.11056435e-01 -9.07436728e-01
1.24895155e-01 -1.16643679e+00 1.24815691e+00 -1.26916245e-01
1.53708398e-01 -1.24695694e+00 -7.87140280e-02 1.76947042e-01
4.52456892e-01 4.82618719e-01 8.64922523e-01 -1.63721129e-01
-6.72315419e-01 9.11638364e-02 -3.07089239e-01 3.99978369e-01
1.14390619e-01 -4.38666373e-01 -1.33522666e+00 -2.58716047e-01
6.08462811e-01 -2.20736146e-01 1.01987469e+00 6.57673776e-01
1.20451939e+00 -6.88571185e-02 -1.00565135e-01 9.48242903e-01
1.83920836e+00 -2.06131890e-01 4.54621911e-01 3.13724518e-01
5.00234723e-01 9.20281470e-01 5.51403701e-01 5.97216964e-01
-1.36970013e-01 1.05459344e+00 5.45184851e-01 -2.58093566e-01
-1.44359261e-01 3.12565304e-02 -1.45112559e-01 3.12073290e-01
-2.79451281e-01 1.31985350e-02 -5.53517401e-01 4.39504743e-01
-1.78089535e+00 -8.32852542e-01 -1.20289654e-01 2.16927075e+00
5.93410969e-01 -2.51328051e-01 -2.31355637e-01 2.67703265e-01
5.31649411e-01 5.52546442e-01 -3.34432691e-01 9.20983627e-02
-2.51977026e-01 3.28581214e-01 4.16901886e-01 8.37016046e-01
-1.21049476e+00 4.62262660e-01 5.25549936e+00 8.72004747e-01
-1.08473575e+00 2.74732113e-01 4.54712510e-01 1.01355568e-01
-4.46720839e-01 -1.51786376e-02 -2.40564883e-01 3.65953147e-01
8.04718196e-01 6.06433786e-02 6.10551655e-01 3.59358341e-01
5.22042274e-01 -1.18755527e-01 -9.59095478e-01 9.88784969e-01
-1.57128256e-02 -1.24981558e+00 -1.96358077e-02 -1.20819581e-03
5.99724472e-01 -4.29614276e-01 3.54646653e-01 -1.10906489e-01
-2.12756380e-01 -9.39750195e-01 8.81018400e-01 8.75123382e-01
7.10456848e-01 -6.58199787e-01 3.49253058e-01 3.55761141e-01
-8.85565817e-01 -7.20879212e-02 -5.75249493e-02 8.69059488e-02
4.46592957e-01 9.81388807e-01 -8.70998427e-02 1.13538229e+00
3.61379147e-01 8.31610858e-01 -1.57913193e-01 7.24228978e-01
-7.70347491e-02 1.89485744e-01 -1.70894057e-01 7.14252651e-01
3.53153497e-01 -7.57180452e-01 8.17943931e-01 1.19423878e+00
4.29636598e-01 -3.79070789e-02 9.59891751e-02 1.25388753e+00
2.01216251e-01 7.51903430e-02 -7.74751604e-01 2.72999525e-01
-7.32837543e-02 1.50753820e+00 -6.30861402e-01 -4.81025055e-02
-5.53384125e-01 9.85530436e-01 2.16842547e-01 5.62877595e-01
-8.05175364e-01 1.30500242e-01 4.28168237e-01 2.42181212e-01
4.78895247e-01 -2.26247922e-01 -3.32316786e-01 -1.11803365e+00
2.38396585e-01 -5.91032088e-01 3.79767716e-02 -8.04659247e-01
-1.34526372e+00 5.63976288e-01 1.07546657e-01 -1.34886324e+00
-2.65259176e-01 -5.20768881e-01 -3.55232537e-01 1.32623982e+00
-2.04097915e+00 -1.50161660e+00 -2.88654447e-01 6.48353577e-01
2.66642302e-01 3.31388384e-01 6.42000675e-01 3.87962937e-01
-3.29662859e-01 -7.01986253e-02 1.54566631e-01 -4.18377936e-01
4.59827542e-01 -1.17207193e+00 -3.70824724e-01 9.85615492e-01
-1.13007292e-01 6.11213565e-01 1.00526059e+00 -2.90738732e-01
-1.42815411e+00 -1.01581228e+00 4.24158007e-01 -9.58060939e-03
7.06597507e-01 -1.76788464e-01 -9.84007776e-01 5.63140631e-01
1.68549955e-01 3.38234544e-01 5.11598170e-01 -3.08661550e-01
-8.42083320e-02 -1.11992657e-01 -1.05874860e+00 3.35753322e-01
7.99760938e-01 -7.70796835e-01 -3.50017220e-01 3.02289605e-01
3.63173842e-01 -1.09885700e-01 -1.11713648e+00 5.09749353e-01
3.21398884e-01 -1.38790679e+00 1.40317035e+00 -3.35991383e-01
5.69058657e-01 -2.73222506e-01 -2.88376570e-01 -1.34266198e+00
-1.43204540e-01 -5.83663344e-01 5.45359142e-02 1.11931133e+00
9.98444036e-02 -6.74640179e-01 5.95836163e-01 3.08996856e-01
-4.43002254e-01 -6.30336761e-01 -9.88828540e-01 -5.82270384e-01
5.50570339e-02 -3.77803922e-01 2.18700111e-01 1.06467783e+00
-4.87184644e-01 3.92539129e-02 -6.04726553e-01 5.01321614e-01
8.52578759e-01 3.66837800e-01 3.34843963e-01 -1.25513172e+00
-5.95371962e-01 -3.66688043e-01 -4.41080071e-02 -9.10651386e-01
3.06795716e-01 -8.12945068e-01 1.28137805e-02 -1.25709236e+00
-1.10936150e-01 -3.93737733e-01 -2.56486773e-01 -4.49096859e-02
7.61505961e-02 4.76274133e-01 1.30951941e-01 1.34890467e-01
-1.00607030e-01 6.91962242e-01 1.22159755e+00 4.33065742e-02
-2.42492586e-01 -4.12228005e-03 -5.93268037e-01 8.55625808e-01
6.01448953e-01 -2.32755199e-01 -4.37932521e-01 -3.45090657e-01
3.00205499e-01 4.79484528e-01 9.48592186e-01 -8.58791292e-01
-9.67106409e-03 -1.65934101e-01 2.41318658e-01 -2.15066850e-01
6.99634731e-01 -1.04598749e+00 3.49821866e-01 1.73239350e-01
-3.46332937e-01 -7.12460697e-01 -1.47760818e-02 6.79341376e-01
-4.61493045e-01 -3.49664062e-01 1.12796795e+00 -2.03553960e-01
-5.78799486e-01 -7.74974888e-03 -2.70095944e-01 -1.85658589e-01
8.46353233e-01 -3.18020135e-01 2.60467499e-01 -2.12136418e-01
-1.10173404e+00 -4.53663409e-01 3.94472450e-01 -4.23312895e-02
5.23089767e-01 -1.12574780e+00 -6.03119314e-01 3.20931822e-01
1.87384516e-01 -2.86505014e-01 5.81879199e-01 1.48645139e+00
-1.68651015e-01 4.02893364e-01 -1.55689299e-01 -6.23666883e-01
-7.59060264e-01 5.68656564e-01 5.49967408e-01 -2.87300199e-01
-8.21514070e-01 5.64972937e-01 3.11625004e-01 -9.51631144e-02
-2.56127447e-01 -3.22746992e-01 -1.50110140e-01 2.18869388e-01
2.75544286e-01 2.24992558e-01 1.93567738e-01 -7.57147431e-01
-1.76113963e-01 7.61158049e-01 3.38498056e-01 -2.58135706e-01
1.66845822e+00 -3.92985970e-01 -2.69257367e-01 4.24755752e-01
1.19706726e+00 2.36908793e-01 -1.62763298e+00 -1.94466263e-01
2.01016381e-01 -4.60374534e-01 3.46653044e-01 -4.99697208e-01
-1.02535510e+00 8.47793818e-01 5.78804553e-01 4.81680423e-01
1.45831120e+00 -1.39283776e-01 2.74621069e-01 -3.74357142e-02
3.33134353e-01 -8.22081029e-01 -1.62479892e-01 5.44636212e-02
1.05374217e+00 -1.09703374e+00 -5.52587397e-02 -6.99535370e-01
-1.12032190e-01 1.27230990e+00 -2.12702334e-01 -2.35025659e-01
6.66664064e-01 8.04275721e-02 -2.84698099e-01 -1.19086467e-01
-1.02837145e-01 -5.60418427e-01 3.51213545e-01 5.43568313e-01
4.92845893e-01 -2.49633193e-01 -5.48680186e-01 4.43724036e-01
2.13728458e-01 1.10129543e-01 4.09909934e-01 7.73295283e-01
-2.57339090e-01 -1.13914251e+00 -4.88996536e-01 1.19927756e-01
-3.80836546e-01 -9.94676277e-02 -1.00340527e-02 7.57754683e-01
2.53792793e-01 8.84941518e-01 -3.70467871e-01 2.59057701e-01
2.67123193e-01 5.95827922e-02 7.16419220e-01 -4.95706946e-01
-3.17027777e-01 3.33548784e-01 -9.54012647e-02 -5.57509184e-01
-1.10285938e+00 -6.22417629e-01 -7.35082984e-01 2.18650416e-01
1.62024740e-02 -1.26258768e-02 6.72830284e-01 9.79657471e-01
2.13790666e-02 8.19374502e-01 4.92092252e-01 -1.32693255e+00
-4.61203903e-01 -8.31424952e-01 -8.43672633e-01 5.47983348e-01
7.26059079e-01 -9.14185643e-01 -3.41696709e-01 1.94206223e-01] | [11.542078971862793, -2.4391214847564697] |
365adf76-4eac-4140-888d-c8857c3a4d9a | learning-parallax-attention-for-stereo-image | 1903.05784 | null | http://arxiv.org/abs/1903.05784v3 | http://arxiv.org/pdf/1903.05784v3.pdf | Learning Parallax Attention for Stereo Image Super-Resolution | Stereo image pairs can be used to improve the performance of super-resolution
(SR) since additional information is provided from a second viewpoint. However,
it is challenging to incorporate this information for SR since disparities
between stereo images vary significantly. In this paper, we propose a
parallax-attention stereo superresolution network (PASSRnet) to integrate the
information from a stereo image pair for SR. Specifically, we introduce a
parallax-attention mechanism with a global receptive field along the epipolar
line to handle different stereo images with large disparity variations. We also
propose a new and the largest dataset for stereo image SR (namely, Flickr1024).
Extensive experiments demonstrate that the parallax-attention mechanism can
capture correspondence between stereo images to improve SR performance with a
small computational and memory cost. Comparative results show that our PASSRnet
achieves the state-of-the-art performance on the Middlebury, KITTI 2012 and
KITTI 2015 datasets. | ['Wei An', 'Longguang Wang', 'Jungang Yang', 'Zhengfa Liang', 'Zaiping Lin', 'Yulan Guo', 'Yingqian Wang'] | 2019-03-14 | learning-parallax-attention-for-stereo-image-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Wang_Learning_Parallax_Attention_for_Stereo_Image_Super-Resolution_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Wang_Learning_Parallax_Attention_for_Stereo_Image_Super-Resolution_CVPR_2019_paper.pdf | cvpr-2019-6 | ['stereo-image-super-resolution'] | ['computer-vision'] | [ 4.22549516e-01 -4.16504771e-01 1.18031152e-01 -4.25867587e-01
-1.01358426e+00 -3.32040489e-01 4.83352870e-01 -5.87451458e-01
-1.91677615e-01 6.63250208e-01 6.21544003e-01 2.45010868e-01
8.24034885e-02 -7.20341682e-01 -9.63012457e-01 -6.42241359e-01
3.81647289e-01 5.66111356e-02 6.78754926e-01 -6.04743898e-01
4.92473751e-01 6.33906066e-01 -1.89119899e+00 6.98849261e-01
1.02252102e+00 8.17763031e-01 7.39683330e-01 4.62987900e-01
2.02081695e-01 8.82990420e-01 -2.18830481e-01 1.14660328e-02
6.88322723e-01 -5.02629727e-02 -7.97065318e-01 6.02071453e-03
1.29383683e+00 -7.06717134e-01 -6.76785648e-01 1.10989892e+00
8.00412953e-01 3.05315644e-01 1.94607645e-01 -5.24298251e-01
-9.20910001e-01 3.91306788e-01 -1.11188591e+00 6.01027369e-01
5.04875898e-01 2.81962547e-02 7.83533514e-01 -1.08040178e+00
6.89275563e-01 1.51230204e+00 3.59291494e-01 3.17227989e-01
-1.33568037e+00 -9.33187008e-01 -1.02813095e-01 3.52208912e-01
-1.52455223e+00 -6.20026469e-01 7.05603123e-01 -2.18761265e-01
8.58182311e-01 2.06703976e-01 3.76992017e-01 7.92332113e-01
-5.79806566e-02 6.23494089e-01 1.38234234e+00 -1.02613792e-02
-1.42970964e-01 -2.91713595e-01 2.03614309e-02 5.72136231e-02
7.73917288e-02 4.87783492e-01 -8.16554785e-01 3.02479118e-01
1.61946630e+00 9.12149101e-02 -5.83536863e-01 -1.76222935e-01
-1.19745719e+00 6.23945832e-01 9.34786439e-01 1.78559020e-01
-1.76177621e-01 -6.60286099e-02 7.35273585e-02 3.21949542e-01
4.30980891e-01 3.50329578e-01 -3.26467365e-01 2.65925020e-01
-7.31082678e-01 1.12124391e-01 -1.25124276e-01 1.07750189e+00
8.27108502e-01 1.05932385e-01 -2.29997411e-01 1.04476964e+00
-1.61235392e-01 5.36051810e-01 2.83635736e-01 -1.54710221e+00
5.64415097e-01 3.25090170e-01 2.50397414e-01 -1.15033793e+00
-9.96656045e-02 -4.16466713e-01 -1.18746555e+00 1.56550676e-01
4.72322255e-02 4.80623692e-01 -7.64122307e-01 1.52895534e+00
2.83976167e-01 4.81204629e-01 3.19204815e-02 1.36214876e+00
1.27822387e+00 5.86575389e-01 -5.50905228e-01 -1.71858922e-01
9.88757730e-01 -9.84599829e-01 -4.43192810e-01 -3.64964634e-01
-1.24669187e-01 -8.14005792e-01 9.71939921e-01 2.67646760e-01
-1.33393049e+00 -1.09118605e+00 -8.95905852e-01 -8.14176261e-01
1.42850712e-01 -4.49368209e-02 3.98132712e-01 -2.18203431e-03
-1.24880505e+00 5.01697600e-01 -5.98780870e-01 -5.16686402e-02
4.48752135e-01 1.79694816e-01 -3.60999584e-01 -4.66534615e-01
-1.17421424e+00 5.79356790e-01 1.27468243e-01 7.17384517e-02
-6.69948876e-01 -8.17679703e-01 -1.07097995e+00 1.80780683e-02
1.16747841e-01 -7.61520386e-01 8.79969537e-01 -7.91919589e-01
-1.35494411e+00 9.17252123e-01 -4.59035873e-01 -2.73044884e-01
3.79755199e-01 -3.56862694e-01 -3.25351298e-01 2.98970938e-01
3.47586125e-01 8.50377500e-01 7.15028107e-01 -1.53298736e+00
-7.68623531e-01 -5.73054910e-01 2.24125981e-01 5.36479890e-01
2.15313643e-01 1.59901336e-01 -5.78536093e-01 -7.04314232e-01
4.23110962e-01 -5.01738310e-01 -1.86306208e-01 -1.71824709e-01
-1.65383533e-01 1.23646699e-01 6.19308650e-01 -7.71013856e-01
8.80635560e-01 -2.21487164e+00 3.74434173e-01 -3.95576805e-01
2.86421269e-01 1.84199527e-01 -3.42640996e-01 -1.52454739e-02
-1.98814064e-01 -1.68328971e-01 -1.76498204e-01 -2.32514754e-01
-5.40554583e-01 1.84430070e-02 -5.66685557e-01 3.95680577e-01
1.29318446e-01 7.27445602e-01 -8.72450292e-01 -3.25695187e-01
4.19483632e-01 9.14428890e-01 -6.47845328e-01 1.38521060e-01
2.86836505e-01 1.09459162e+00 -2.77413994e-01 4.37773019e-01
1.13150322e+00 -3.33283722e-01 -3.14221263e-01 -5.52608728e-01
-3.34439635e-01 3.28541517e-01 -1.29893708e+00 2.08519077e+00
-5.08996606e-01 8.25483203e-01 -1.55490816e-01 -3.33885729e-01
9.99019384e-01 -1.70361176e-01 3.35349023e-01 -1.13310778e+00
-1.15397207e-01 1.10705361e-01 -2.49891937e-01 -3.81439142e-02
8.44934642e-01 9.42410305e-02 3.04961890e-01 -3.32047455e-02
-2.80967861e-01 -2.09482238e-01 7.06560016e-02 3.69090140e-02
7.22776890e-01 5.97089417e-02 3.77334207e-01 -2.39897788e-01
8.34975660e-01 -3.96704257e-01 8.10502529e-01 6.44828081e-01
-1.37917409e-02 1.28150499e+00 -2.86787361e-01 -6.88132346e-01
-1.20894289e+00 -1.28531229e+00 -4.35982585e-01 7.58583248e-01
6.69046581e-01 -2.13044092e-01 -2.91455746e-01 -1.08712338e-01
-2.24489823e-01 3.00327569e-01 -4.28532422e-01 1.95944980e-01
-7.43262529e-01 -2.08945721e-01 -2.22407896e-02 7.53072083e-01
1.17812169e+00 -9.05865729e-01 -5.01164854e-01 -8.15339386e-02
-5.92038631e-01 -1.42299366e+00 -6.89742863e-01 -5.44196188e-01
-8.84138525e-01 -9.55433369e-01 -8.73695433e-01 -6.63540661e-01
4.74835515e-01 1.13952172e+00 1.24083710e+00 -2.50590742e-01
-2.00778544e-01 -1.18442282e-01 -1.60696760e-01 2.26596147e-01
9.97667834e-02 -3.58244628e-01 -3.12427320e-02 -9.27501768e-02
2.16510698e-01 -7.31327951e-01 -9.35436606e-01 6.67491257e-01
-8.28356326e-01 5.31352878e-01 3.96999508e-01 9.45307016e-01
9.92274225e-01 2.03668803e-01 1.24120586e-01 -5.61763108e-01
-2.63310998e-04 -7.09048137e-02 -7.48509467e-01 -8.72050747e-02
-5.72960936e-02 -1.94887027e-01 5.18306077e-01 -1.61183909e-01
-1.47613776e+00 2.28204411e-02 1.06356479e-01 -7.35109866e-01
-8.76849219e-02 1.60927400e-02 -2.23333776e-01 -3.36879313e-01
6.44841909e-01 5.50175607e-01 -3.08865279e-01 -6.55802071e-01
2.08045170e-01 6.59997821e-01 8.44918668e-01 -2.78650612e-01
8.23381841e-01 1.02117944e+00 5.07301539e-02 -8.04944456e-01
-1.18343401e+00 -6.17790937e-01 -8.11457694e-01 1.73939794e-01
7.93016970e-01 -1.68551195e+00 -5.18771231e-01 4.81182933e-01
-1.01628804e+00 -2.00791478e-01 -1.30461067e-01 3.46298635e-01
-7.90233016e-01 6.39147282e-01 -7.16046989e-01 -4.03471172e-01
-3.29539210e-01 -1.27653921e+00 1.54140067e+00 4.31014299e-01
3.77870739e-01 -3.99413854e-01 2.55279057e-03 8.63325834e-01
5.11733353e-01 -6.23895042e-02 1.79670721e-01 1.77515820e-01
-1.11701012e+00 3.55728775e-01 -9.45195436e-01 1.74522385e-01
2.32267797e-01 -4.47971642e-01 -1.13733327e+00 -5.54382205e-01
5.58343492e-02 -2.51472741e-01 1.16626132e+00 5.36052525e-01
1.43809628e+00 2.12575272e-02 1.08730190e-01 1.21909809e+00
1.57679641e+00 -9.11764279e-02 1.08468068e+00 4.40943152e-01
9.77314949e-01 4.73379046e-01 8.65842521e-01 2.22128645e-01
6.01854920e-01 1.10632598e+00 2.68267214e-01 -3.08363855e-01
-5.57585418e-01 -2.48565659e-01 2.50544697e-01 5.60978532e-01
-5.00295639e-01 2.79951066e-01 -5.32267034e-01 3.96555185e-01
-1.69599080e+00 -1.29244649e+00 -1.52064532e-01 2.39592671e+00
8.39823067e-01 -2.47033924e-01 -2.60434151e-01 -6.25813007e-02
7.75663555e-01 4.25452441e-01 -5.90504348e-01 9.78020504e-02
-6.60948277e-01 7.30424523e-02 5.37484646e-01 8.00128102e-01
-9.70450222e-01 1.16075051e+00 5.88120794e+00 1.01904058e+00
-1.16406488e+00 5.24769817e-03 6.00445271e-01 -1.03626363e-02
-2.86697507e-01 -2.33598217e-01 -9.40469027e-01 3.90712082e-01
4.06737328e-01 -8.04787502e-02 7.69658625e-01 5.20587802e-01
2.59553015e-01 -1.64892659e-01 -8.21288109e-01 1.43947423e+00
2.94554085e-01 -1.35268152e+00 1.69508994e-01 1.59502886e-02
1.21723700e+00 4.20654148e-01 2.08099097e-01 6.34707138e-02
3.09375167e-01 -1.04978859e+00 2.14199513e-01 4.49084669e-01
1.14622164e+00 -8.14054072e-01 6.34371579e-01 9.12570879e-02
-1.45628643e+00 -1.28579870e-01 -8.63817573e-01 6.23781644e-02
1.74054801e-01 4.40872848e-01 -7.78020322e-02 8.80505085e-01
1.27386081e+00 1.33255994e+00 -6.22644067e-01 9.84646857e-01
-2.15945244e-01 -1.25603735e-01 -1.36973545e-01 9.51763690e-01
-1.82492882e-01 -2.25479797e-01 7.70926595e-01 6.71495616e-01
4.00648236e-01 3.55708539e-01 5.05443998e-02 9.38220322e-01
9.99500751e-02 -2.23578081e-01 -7.10542023e-01 5.00000834e-01
4.99163866e-01 9.92195964e-01 -2.50073463e-01 -3.10425013e-01
-5.22972941e-01 1.16301596e+00 3.04268032e-01 5.24994314e-01
-5.98889410e-01 9.65424255e-02 9.01452184e-01 2.31978893e-01
4.91131514e-01 -1.76091939e-01 -3.67966175e-01 -1.68187439e+00
-5.00908531e-02 -1.03485751e+00 3.16856056e-01 -1.36097693e+00
-1.25314593e+00 8.37336063e-01 -1.50752157e-01 -1.54020751e+00
8.26397911e-02 -2.75546849e-01 -2.20986426e-01 1.10959756e+00
-2.09799170e+00 -1.17734933e+00 -8.34865630e-01 8.48828137e-01
9.22436535e-01 -4.16544899e-02 2.94459522e-01 2.83786416e-01
-2.56611586e-01 2.99561650e-01 2.14660764e-01 -3.75732221e-02
9.21171665e-01 -1.08918917e+00 4.76983070e-01 9.19641793e-01
-1.19022407e-01 6.61192834e-01 6.13837004e-01 -5.84944606e-01
-1.23264194e+00 -1.06107759e+00 6.34269357e-01 -3.81293833e-01
2.47131154e-01 -3.26400958e-02 -1.00942600e+00 6.93689883e-01
9.71011594e-02 2.72104621e-01 2.06902966e-01 -5.36253899e-02
-6.78103507e-01 -3.60439897e-01 -1.05366051e+00 6.11413658e-01
1.47483850e+00 -6.42348349e-01 -5.97568274e-01 -2.00197175e-02
9.76392746e-01 -7.94675767e-01 -9.35133100e-01 6.90023541e-01
4.26490903e-01 -1.47789216e+00 1.67760682e+00 -1.06223486e-01
8.29976737e-01 -7.22032845e-01 -6.12208307e-01 -1.22727346e+00
-6.79941595e-01 -4.19603467e-01 1.10659927e-01 9.99166012e-01
-1.52492970e-01 -6.09033465e-01 3.26049179e-01 2.84015417e-01
-7.58664608e-02 -3.02823395e-01 -9.42200482e-01 -7.15740979e-01
-1.02493808e-01 1.11527994e-01 7.45870113e-01 9.41729367e-01
-5.31293035e-01 4.95349288e-01 -5.86311936e-01 4.43846881e-01
9.78282332e-01 6.04309618e-01 9.07858014e-01 -1.21073949e+00
-3.29378426e-01 -2.57958800e-01 -4.98241007e-01 -1.55952024e+00
1.72163304e-02 -6.00490928e-01 -9.20629427e-02 -1.39601278e+00
7.19753802e-01 -2.61316057e-02 -1.80926234e-01 -1.45398895e-03
-3.32964897e-01 5.85917652e-01 1.97676376e-01 5.14495194e-01
-5.65619767e-01 7.96674073e-01 1.80400574e+00 6.39422014e-02
-1.81796029e-01 -3.05510014e-01 -7.82927394e-01 6.70551956e-01
5.27255416e-01 1.40765443e-01 -3.72778088e-01 -9.49294031e-01
1.00678436e-01 3.90854031e-01 4.33169097e-01 -9.43030119e-01
2.47595608e-01 -2.62568027e-01 6.35904133e-01 -1.02994883e+00
5.51947892e-01 -6.88326418e-01 2.09000617e-01 -9.23826918e-02
-3.15322220e-01 -8.29370171e-02 7.43694529e-02 6.92470551e-01
-3.98352414e-01 5.38349450e-01 1.15448928e+00 -1.60197288e-01
-8.63711298e-01 6.02795005e-01 3.70687127e-01 8.28084629e-03
6.45117283e-01 -3.68657470e-01 -6.78610146e-01 -3.85194153e-01
-4.48533237e-01 3.78296614e-01 9.89083290e-01 4.64252412e-01
9.38671947e-01 -1.43091261e+00 -9.05528247e-01 4.71330196e-01
1.17481232e-01 5.75177133e-01 8.57252061e-01 6.46525860e-01
-4.21108335e-01 3.41372222e-01 -6.63129747e-01 -9.02927041e-01
-1.42280066e+00 5.52369118e-01 3.80350828e-01 -5.04419878e-02
-9.55047786e-01 7.98475683e-01 9.53002095e-01 -3.64501625e-01
9.34782252e-03 -1.17058672e-01 -5.78917384e-01 -3.61240923e-01
1.00882423e+00 4.15700942e-01 -2.63930053e-01 -9.43662226e-01
9.57689807e-03 1.29727530e+00 -1.70848355e-01 5.23239188e-02
1.44674528e+00 -7.87824869e-01 -1.42957881e-01 2.83414155e-01
1.01193738e+00 1.35195047e-01 -1.59603679e+00 -7.25795805e-01
-5.78600943e-01 -1.39396191e+00 2.99410641e-01 -5.81788242e-01
-1.40557611e+00 9.50485289e-01 6.45247340e-01 -2.84572244e-01
1.50556862e+00 -1.36838734e-01 8.35468948e-01 7.54677728e-02
6.93605185e-01 -7.30480492e-01 9.64112431e-02 5.72918296e-01
1.34703374e+00 -1.66276515e+00 4.74635698e-02 -7.79729545e-01
-6.97557449e-01 1.04523039e+00 8.46914530e-01 -3.01727891e-01
3.52114797e-01 8.13301653e-03 -3.99143137e-02 -2.54194960e-02
-6.11239135e-01 -4.16563362e-01 4.43493485e-01 6.99977458e-01
3.90349686e-01 -1.71385825e-01 -7.28237033e-02 1.56269699e-01
-2.56753057e-01 -7.94647902e-04 6.56477630e-01 3.39644670e-01
-5.27537584e-01 -7.56797493e-01 -2.87752897e-01 3.54960747e-02
-3.13983858e-01 -4.02027935e-01 -1.87539637e-01 4.26413327e-01
-1.99272349e-01 1.00997353e+00 2.77137220e-01 -3.27157557e-01
1.40249699e-01 -9.06642199e-01 6.30322099e-01 -5.60493767e-01
-1.63212121e-01 3.37876707e-01 -1.20307140e-01 -1.03983176e+00
-6.29990757e-01 -4.32105601e-01 -1.06895542e+00 -3.85989845e-01
7.25757778e-02 -3.85113448e-01 1.82569176e-01 5.81445098e-01
6.22071266e-01 3.73518258e-01 9.37640667e-01 -1.16013110e+00
-2.07051083e-01 -9.16948080e-01 -5.85838556e-01 4.60134268e-01
5.88717937e-01 -6.35238171e-01 -4.00352120e-01 -8.19523111e-02] | [10.67194938659668, -2.1419496536254883] |
4dbab473-0ec1-45c8-88b0-72bc62d28450 | spatio-temporal-wind-speed-forecasting-using | 2208.13585 | null | https://arxiv.org/abs/2208.13585v2 | https://arxiv.org/pdf/2208.13585v2.pdf | Spatio-Temporal Wind Speed Forecasting using Graph Networks and Novel Transformer Architectures | This study focuses on multi-step spatio-temporal wind speed forecasting for the Norwegian continental shelf. The study aims to leverage spatial dependencies through the relative physical location of different measurement stations to improve local wind forecasts. Our multi-step forecasting models produce either 10-minute, 1- or 4-hour forecasts, with 10-minute resolution, meaning that the models produce more informative time series for predicted future trends. A graph neural network (GNN) architecture was used to extract spatial dependencies, with different update functions to learn temporal correlations. These update functions were implemented using different neural network architectures. One such architecture, the Transformer, has become increasingly popular for sequence modelling in recent years. Various alterations have been proposed to better facilitate time series forecasting, of which this study focused on the Informer, LogSparse Transformer and Autoformer. This is the first time the LogSparse Transformer and Autoformer have been applied to wind forecasting and the first time any of these or the Informer have been formulated in a spatio-temporal setting for wind forecasting. By comparing against spatio-temporal Long Short-Term Memory (LSTM) and Multi-Layer Perceptron (MLP) models, the study showed that the models using the altered Transformer architectures as update functions in GNNs were able to outperform these. Furthermore, we propose the Fast Fourier Transformer (FFTransformer), which is a novel Transformer architecture based on signal decomposition and consists of two separate streams that analyse the trend and periodic components separately. The FFTransformer and Autoformer were found to achieve superior results for the 10-minute and 1-hour ahead forecasts, with the FFTransformer significantly outperforming all other models for the 4-hour ahead forecasts. | ['Paal Engelstad', 'Roy Stenbro', 'Narada Dilp Warakagoda', 'Lars Ødegaard Bentsen'] | 2022-08-29 | null | null | null | null | ['weather-forecasting', 'spatio-temporal-forecasting'] | ['miscellaneous', 'time-series'] | [-1.99714169e-01 -3.73535603e-01 2.64737636e-01 -7.01135993e-02
6.20524883e-02 -5.97942889e-01 8.51868570e-01 -1.23210968e-02
-3.43017787e-01 8.73054385e-01 4.21148121e-01 -8.11542809e-01
-8.67410064e-01 -8.82722557e-01 -3.07933837e-01 -8.09422016e-01
-8.38294148e-01 -6.44608587e-02 -6.78167492e-02 -4.53395873e-01
-8.47577304e-02 5.04449606e-01 -1.53951621e+00 1.63669139e-01
7.33386397e-01 9.01784182e-01 3.07197779e-01 1.07603431e+00
-1.13051206e-01 8.49459112e-01 -6.12859607e-01 -1.39667988e-02
2.63247013e-01 -3.00257146e-01 -4.12308723e-01 -6.48873329e-01
1.27360418e-01 -1.72337040e-01 9.21083167e-02 4.70959991e-01
5.80726862e-01 6.45916879e-01 3.01568836e-01 -8.61010790e-01
-2.70213038e-01 8.43908489e-01 -2.39267245e-01 6.65077686e-01
5.80927990e-02 -1.70471102e-01 8.39804113e-01 -8.42223763e-01
2.03633472e-01 1.01561451e+00 1.17576015e+00 -1.11882865e-01
-1.15110552e+00 -5.28280497e-01 2.84222573e-01 2.67522544e-01
-9.54148412e-01 -2.22944051e-01 7.86273360e-01 -7.20339298e-01
1.55112600e+00 2.32641101e-01 8.31467807e-01 1.01937902e+00
6.70981228e-01 3.14542092e-02 1.13361347e+00 -3.75050783e-01
1.72367781e-01 -1.01580881e-01 1.53096855e-01 9.18073505e-02
5.35625145e-02 7.44322300e-01 -5.50252676e-01 -8.00128207e-02
5.46685278e-01 1.58304721e-01 -5.15127957e-01 4.90518183e-01
-9.57033217e-01 9.63203013e-01 5.87290168e-01 8.66494894e-01
-9.22347963e-01 1.37717694e-01 4.78536069e-01 4.92948532e-01
1.06303823e+00 4.23509836e-01 -8.79368842e-01 -3.24022293e-01
-1.38785732e+00 -3.75285069e-03 8.31829667e-01 -8.08587521e-02
4.24939960e-01 1.00065696e+00 1.85792178e-01 4.95577991e-01
3.75628382e-01 6.88099384e-01 9.33633745e-01 -3.20946306e-01
4.37476993e-01 2.83780787e-02 -1.21571086e-01 -1.25582886e+00
-9.15526986e-01 -1.10246527e+00 -1.20378530e+00 3.71095836e-01
2.07313895e-01 -7.84204304e-01 -8.71607304e-01 1.38689077e+00
1.16794765e-01 8.79830182e-01 2.39002377e-01 6.91728592e-01
4.85436231e-01 1.24788713e+00 2.28190608e-02 -3.11727613e-01
1.00656593e+00 -7.82348633e-01 -9.72862124e-01 -4.48799282e-02
7.78142393e-01 -8.09458315e-01 4.38754797e-01 3.40736330e-01
-6.03457570e-01 -1.02777565e+00 -1.04309893e+00 5.47281802e-01
-1.10561526e+00 2.74581075e-01 2.36878291e-01 6.41407907e-01
-1.52336061e+00 1.09587145e+00 -1.05523574e+00 -2.08074063e-01
-5.33009470e-01 -4.50479984e-02 -1.90201521e-01 6.48480177e-01
-1.70965350e+00 1.06819463e+00 5.30192137e-01 7.63979554e-01
-2.46096164e-01 -9.72166181e-01 -8.77432227e-01 2.33269989e-01
-1.94633752e-01 -7.09136784e-01 7.85492003e-01 -9.14349854e-01
-1.76288986e+00 -2.82724082e-01 -3.83287430e-01 -1.08917308e+00
1.44658938e-01 -2.46456161e-01 -9.60805714e-01 -2.10196003e-01
-2.80729622e-01 -5.60343675e-02 1.00004840e+00 -7.98099279e-01
-7.54497170e-01 -4.93324697e-02 -2.83856422e-01 -5.52784316e-02
-1.51744485e-01 -2.68649012e-01 8.33158791e-01 -1.06874382e+00
-3.00154716e-01 -6.12947524e-01 -2.60278523e-01 -8.77072155e-01
9.79650617e-02 -4.66908589e-02 1.02023005e+00 -1.13166833e+00
1.67633963e+00 -2.00982642e+00 2.61196401e-02 3.50614041e-01
1.11120977e-02 5.24227858e-01 -1.62958130e-01 1.14868212e+00
-4.78433520e-01 1.55902103e-01 -7.95580745e-02 -5.40688217e-01
-2.22871661e-01 5.39580703e-01 -9.74903584e-01 2.45456457e-01
1.07644156e-01 6.63492501e-01 -7.44584620e-01 3.76692265e-01
5.75868189e-01 9.46614385e-01 5.79830632e-02 -4.98221219e-02
1.67938620e-01 3.22754920e-01 1.24689050e-01 -1.96041435e-01
7.21556485e-01 -3.78539935e-02 -1.54553309e-01 8.06697458e-03
-6.29779637e-01 3.55596781e-01 -1.06322122e+00 1.04232812e+00
-8.57251465e-01 8.72626245e-01 -3.53840515e-02 -8.18192959e-01
1.11869752e+00 8.07211697e-01 2.67155170e-01 -5.25730968e-01
-3.15515727e-01 2.63547391e-01 1.49717927e-01 -4.04662818e-01
5.93759477e-01 -4.13202643e-01 6.20723009e-01 2.40895540e-01
-1.47358840e-02 2.45011494e-01 2.79862165e-01 -3.45212311e-01
7.31358945e-01 3.67136359e-01 1.32242993e-01 -3.74002397e-01
5.82805634e-01 -2.43037865e-01 4.21966910e-01 5.07952631e-01
4.00137842e-01 3.76607180e-01 1.74243316e-01 -9.08688545e-01
-5.84196210e-01 -6.01158559e-01 7.18616694e-02 1.01642966e+00
-6.88729167e-01 -5.44042706e-01 -5.06919399e-02 -1.97372034e-01
2.22867541e-02 1.09221935e+00 -7.49496400e-01 2.33923614e-01
-6.36617303e-01 -8.11082065e-01 5.66381335e-01 5.85620821e-01
2.15230078e-01 -8.41890574e-01 -8.34255457e-01 6.44266963e-01
1.69901997e-01 -7.69558191e-01 1.37092724e-01 7.15576530e-01
-1.18476319e+00 -5.81922352e-01 -7.45210648e-01 -1.96689934e-01
1.45474687e-01 -6.49654642e-02 8.29690158e-01 -2.01006442e-01
4.22605067e-01 1.98898271e-01 -7.01098263e-01 -5.37233233e-01
-6.45438656e-02 1.12822413e-01 1.76392391e-01 2.60442317e-01
-1.18890710e-01 -1.06095767e+00 -3.25963974e-01 -1.26218684e-02
-9.25709128e-01 -6.10612556e-02 3.93904448e-01 9.10183370e-01
8.86488035e-02 4.20014858e-01 6.87238932e-01 -6.08726263e-01
8.61429334e-01 -7.52224028e-01 -7.65769899e-01 -9.11262333e-02
-8.02915335e-01 1.66594639e-01 1.26009870e+00 -2.09066197e-01
-9.55530107e-01 -3.21035564e-01 -4.77064580e-01 -5.16136050e-01
-8.66287500e-02 1.40575922e+00 7.26205885e-01 4.37350534e-02
4.50932950e-01 4.76430714e-01 -2.02539772e-01 -8.02697957e-01
1.06443219e-01 1.94199637e-01 1.35996431e-01 2.18638871e-02
8.09376836e-01 3.46791416e-01 1.71319231e-01 -1.16974711e+00
-3.81501943e-01 -4.66196060e-01 -6.86079562e-01 -4.41910237e-01
7.76529610e-01 -8.18437278e-01 -3.09293151e-01 5.47762871e-01
-1.33070827e+00 -4.79555458e-01 -2.64333963e-01 9.69205737e-01
6.54000416e-02 3.68521437e-02 -5.01500249e-01 -1.19007671e+00
-4.66856211e-01 -6.96061075e-01 7.36589193e-01 1.67465329e-01
-1.64788246e-01 -1.84664273e+00 4.89135444e-01 -5.54454088e-01
1.11172223e+00 5.37632465e-01 6.43077016e-01 -6.39277220e-01
8.42409655e-02 -4.36301045e-02 2.54183739e-01 4.53399688e-01
2.36258432e-01 2.57958204e-01 -1.07964206e+00 -4.11482036e-01
2.03648433e-01 6.72103941e-01 9.30266798e-01 8.92917037e-01
2.49351770e-01 -6.02405906e-01 -1.15170509e-01 7.90707886e-01
1.48183894e+00 5.26352584e-01 3.79598618e-01 3.38677198e-01
4.53922391e-01 5.72956741e-01 1.23458661e-01 3.93485010e-01
2.81757236e-01 2.73964882e-01 3.71949792e-01 -2.73351163e-01
5.41672967e-02 -5.23492284e-02 5.47538579e-01 1.20334113e+00
-7.47382224e-01 -5.43313622e-01 -9.66000855e-01 6.19466066e-01
-1.85170186e+00 -1.20284176e+00 -4.87135231e-01 2.08554935e+00
1.04014792e-01 2.25061074e-01 -8.14797953e-02 4.50036734e-01
1.95633516e-01 8.02430093e-01 5.15930615e-02 -7.79730380e-01
-1.19588293e-01 5.89417219e-01 7.85810709e-01 7.57043242e-01
-1.07498193e+00 3.52198631e-01 5.48191118e+00 5.90529740e-01
-1.70316565e+00 2.00568244e-01 2.82091677e-01 1.27127573e-01
-1.53074279e-01 4.17217463e-02 -7.86092818e-01 5.75131357e-01
1.72446728e+00 2.54636873e-02 3.77343208e-01 5.71111023e-01
9.55319941e-01 -4.36642906e-03 -5.33436537e-01 5.09476900e-01
-2.21081063e-01 -1.30489850e+00 -7.10867792e-02 7.93193430e-02
6.43969059e-01 3.78795564e-01 -8.35645944e-02 1.81529313e-01
1.40839607e-01 -1.13740838e+00 5.59618294e-01 1.22939265e+00
2.21771255e-01 -7.34108746e-01 1.35040271e+00 3.57248664e-01
-1.75280261e+00 -9.71096531e-02 -9.89276618e-02 -7.39750266e-01
6.13884747e-01 9.93942022e-01 -7.54386842e-01 1.23826540e+00
8.41982305e-01 1.08081412e+00 -2.31502071e-01 9.88237262e-01
-6.11024164e-02 1.18655860e+00 -5.56345463e-01 -1.07569523e-01
7.53605187e-01 -3.41096520e-01 7.73815155e-01 1.24592888e+00
8.59668195e-01 -1.89255744e-01 -3.31790708e-02 1.87333494e-01
6.81292832e-01 -1.24420971e-01 -6.78595304e-01 1.01398915e-01
2.54383832e-01 1.08311546e+00 -5.10930896e-01 -2.51673609e-01
-3.87900382e-01 4.09585059e-01 -1.36412039e-01 9.52419877e-01
-6.46541059e-01 -6.06500924e-01 7.52621293e-01 1.76867321e-01
6.95064783e-01 -5.57896197e-01 -1.89908698e-01 -6.28031313e-01
1.38070546e-02 -3.11127990e-01 4.93171304e-01 -8.70115757e-01
-1.24451697e+00 1.10729969e+00 1.28049612e-01 -1.31171346e+00
-8.67405951e-01 -7.88915098e-01 -9.61059391e-01 1.38984525e+00
-1.80529106e+00 -1.07009125e+00 1.18007004e-01 2.65884697e-01
4.18689787e-01 -7.19776899e-02 9.28309381e-01 1.55828893e-01
-3.94186646e-01 -1.21988088e-01 5.84596038e-01 -1.19755216e-01
2.93136865e-01 -1.42779028e+00 4.52692032e-01 1.21921992e+00
2.08731920e-01 5.81816852e-01 8.46741080e-01 -6.78184569e-01
-9.96137500e-01 -1.20102429e+00 1.36441696e+00 -3.32434685e-03
9.05197501e-01 3.27541679e-02 -1.05683696e+00 7.77616978e-01
6.21252716e-01 3.05462666e-02 4.91793007e-01 1.71176732e-01
1.01089746e-01 -4.04757589e-01 -5.42166531e-01 1.91892922e-01
4.27133739e-01 -4.22814876e-01 -7.69149363e-01 -2.17505218e-03
5.50448477e-01 -2.43865639e-01 -1.03389096e+00 4.98505265e-01
4.99283731e-01 -1.19187236e+00 7.36837447e-01 -1.99617282e-01
2.77236015e-01 -6.04983449e-01 2.89295930e-02 -2.09528041e+00
-5.40866613e-01 -6.99036419e-01 -1.86939389e-01 9.96806920e-01
6.92752123e-01 -1.57096851e+00 1.62782952e-01 -7.89310411e-02
-3.09172094e-01 -6.99679911e-01 -1.20401919e+00 -8.25666010e-01
-1.03037305e-01 -6.83890462e-01 7.18064964e-01 9.71340239e-01
-2.82207578e-01 2.15490505e-01 -6.59218252e-01 4.43254530e-01
-9.90565345e-02 8.75682905e-02 4.00540978e-01 -1.38703513e+00
-3.80176574e-01 -6.94543958e-01 -2.20298901e-01 -8.22974145e-01
-1.32722825e-01 -6.15830064e-01 -2.50662208e-01 -1.70407617e+00
-1.01928377e+00 2.99191289e-02 -4.47191954e-01 4.57279950e-01
-5.08103184e-02 -1.73731670e-01 1.35259777e-01 -5.77043332e-02
5.52867830e-01 5.44417381e-01 6.53893352e-01 2.56763220e-01
-3.57985526e-01 3.83215159e-01 1.98483229e-01 6.14188135e-01
8.00783277e-01 -3.31669420e-01 -6.19016171e-01 -4.48156476e-01
5.86713731e-01 1.30180374e-01 3.74414891e-01 -1.28595138e+00
4.18647766e-01 2.48681858e-01 6.37626052e-01 -9.45238769e-01
3.74655426e-01 -8.84363890e-01 8.02842855e-01 6.76488757e-01
7.69014955e-02 8.59250009e-01 7.35159338e-01 4.97967362e-01
-6.07535720e-01 4.23081423e-04 1.21693507e-01 1.01729698e-01
-6.95340991e-01 -1.14430532e-01 -8.43023896e-01 -5.04736245e-01
6.37671649e-01 -3.84844154e-01 -2.83419609e-01 -5.25633633e-01
-8.68751764e-01 1.24140754e-01 -3.64296734e-01 2.60088295e-01
4.46336359e-01 -8.90156329e-01 -9.61781085e-01 3.51495087e-01
-6.07686341e-01 -4.80165929e-01 4.45850939e-01 1.14509177e+00
-4.63046193e-01 8.13008785e-01 1.28791705e-02 -4.15383011e-01
-1.01392472e+00 2.26763949e-01 6.80184066e-01 -8.24536443e-01
-5.88416576e-01 7.46348500e-01 -2.95005590e-01 -4.53786969e-01
-2.42980599e-01 -1.02958703e+00 -6.94426000e-01 6.26968324e-01
3.92999709e-01 5.75390577e-01 3.61634314e-01 -6.74924076e-01
-3.09530795e-01 8.42921555e-01 4.73731071e-01 -2.33796656e-01
1.65052879e+00 -2.17542663e-01 -1.44067243e-01 9.75175381e-01
9.32104647e-01 1.67286336e-01 -1.01062095e+00 9.51296017e-02
7.53522292e-02 -2.79983170e-02 4.26355362e-01 -8.82082462e-01
-1.02446330e+00 9.62716460e-01 4.17878181e-01 8.92555654e-01
1.35522854e+00 -9.12939250e-01 7.44185209e-01 -5.29958718e-02
9.88027975e-02 -5.98575234e-01 -8.93272877e-01 1.04533124e+00
9.41253424e-01 -5.88041008e-01 -2.05886438e-01 2.37565249e-01
-2.59294569e-01 1.67209184e+00 -1.61157459e-01 -1.06797948e-01
1.17246902e+00 3.94660294e-01 2.39722669e-01 -5.22416271e-02
-9.21066046e-01 -2.86980718e-01 5.84497273e-01 3.10710341e-01
4.84595776e-01 1.76668867e-01 -2.39824638e-01 4.17896658e-01
-7.08907068e-01 7.44029821e-05 3.34293664e-01 7.34529555e-01
-7.14205131e-02 -8.02497029e-01 -5.21365643e-01 5.31040907e-01
-3.40265542e-01 -3.85375082e-01 2.71023005e-01 8.58907700e-01
8.37761760e-02 1.13829803e+00 2.48062119e-01 -6.34862185e-01
2.88756043e-01 4.48808461e-01 -3.79065365e-01 -1.63507402e-01
-1.15392625e+00 2.02678721e-02 2.59535134e-01 -3.76768827e-01
-5.56875229e-01 -4.61374193e-01 -8.68917704e-01 -2.43971735e-01
-2.78700054e-01 5.14022827e-01 8.77482176e-01 1.12452281e+00
4.36467588e-01 1.00569987e+00 6.88488543e-01 -1.27314639e+00
-1.82907999e-01 -1.33976150e+00 -6.72126412e-01 -4.31911707e-01
1.00072753e+00 -4.32546347e-01 -7.63491988e-01 3.42020057e-02] | [6.466240406036377, 2.9011495113372803] |
0d2eb971-bc4d-42bb-a35d-32c2c3febd5e | learning-reliable-representations-for | 2303.17117 | null | https://arxiv.org/abs/2303.17117v1 | https://arxiv.org/pdf/2303.17117v1.pdf | Learning Reliable Representations for Incomplete Multi-View Partial Multi-Label Classification | As a cross-topic of multi-view learning and multi-label classification, multi-view multi-label classification has gradually gained traction in recent years. The application of multi-view contrastive learning has further facilitated this process, however, the existing multi-view contrastive learning methods crudely separate the so-called negative pair, which largely results in the separation of samples belonging to the same category or similar ones. Besides, plenty of multi-view multi-label learning methods ignore the possible absence of views and labels. To address these issues, in this paper, we propose an incomplete multi-view partial multi-label classification network named RANK. In this network, a label-driven multi-view contrastive learning strategy is proposed to leverage supervised information to preserve the structure within view and perform consistent alignment across views. Furthermore, we break through the view-level weights inherent in existing methods and propose a quality-aware sub-network to dynamically assign quality scores to each view of each sample. The label correlation information is fully utilized in the final multi-label cross-entropy classification loss, effectively improving the discriminative power. Last but not least, our model is not only able to handle complete multi-view multi-label datasets, but also works on datasets with missing instances and labels. Extensive experiments confirm that our RANK outperforms existing state-of-the-art methods. | ['Min Zhang', 'Liqiang Nie', 'Yong Xu', 'Jie Wen', 'Chengliang Liu'] | 2023-03-30 | null | null | null | null | ['multi-view-learning', 'multi-label-learning'] | ['computer-vision', 'methodology'] | [ 2.06331909e-01 -2.00097203e-01 -5.46720862e-01 -6.28666818e-01
-8.37077379e-01 -5.27110934e-01 4.11276609e-01 2.60902584e-01
-7.37250522e-02 3.82452756e-01 1.93974122e-01 3.45556498e-01
-2.51233757e-01 -4.84566450e-01 -1.86711207e-01 -9.83237863e-01
5.13602078e-01 2.22120926e-01 -2.61327103e-02 1.67945206e-01
-1.15361298e-02 2.10318223e-01 -1.51248586e+00 5.90062618e-01
7.97792792e-01 9.65999484e-01 7.57397413e-02 9.79265571e-02
3.56906652e-02 7.09759355e-01 -4.25713658e-02 -7.09019303e-01
1.13429777e-01 -4.11871493e-01 -6.80098474e-01 4.89127189e-01
7.88641870e-01 -1.76112220e-01 1.03238992e-01 1.25819170e+00
5.14971972e-01 -1.26435339e-01 6.79169476e-01 -1.39324760e+00
-5.16212285e-01 2.10845336e-01 -1.06116354e+00 -1.04537539e-01
1.89069077e-01 -1.95190176e-01 1.47084630e+00 -1.17861569e+00
6.08342767e-01 1.25384235e+00 6.74623787e-01 4.31468606e-01
-1.34727728e+00 -7.78881133e-01 5.06364584e-01 2.95266509e-01
-1.27842569e+00 -2.51176149e-01 1.16150117e+00 -4.72293824e-01
3.50626618e-01 2.34681115e-01 5.91324985e-01 1.12241316e+00
1.67270347e-01 9.44840908e-01 1.64324749e+00 -3.11509609e-01
-2.09917083e-01 2.23238826e-01 4.73545417e-02 8.88005495e-01
7.18012303e-02 -4.47687544e-02 -5.16162872e-01 -3.05317163e-01
2.53307104e-01 5.50669491e-01 -2.24281922e-01 -1.01262677e+00
-1.24361360e+00 8.56491745e-01 2.85762519e-01 1.65142223e-01
-1.62256286e-01 -5.10834992e-01 7.43569136e-01 9.92261767e-02
8.70901585e-01 -1.58909988e-02 -5.79289496e-01 2.71745473e-01
-9.23218668e-01 1.72957443e-02 4.46362674e-01 7.56282985e-01
7.83052385e-01 -4.06676412e-01 -1.81901470e-01 1.22080123e+00
4.63699937e-01 2.83095300e-01 3.60363215e-01 -8.14340889e-01
6.90402985e-01 1.06344390e+00 -3.95329058e-01 -1.13780284e+00
-6.67794287e-01 -7.43051767e-01 -1.17408061e+00 4.97735515e-02
1.82677969e-01 2.65980363e-01 -5.29087961e-01 1.76366293e+00
6.24946535e-01 -2.34965757e-02 -1.53786123e-01 6.72634423e-01
7.49583602e-01 1.96165130e-01 9.16660950e-02 -6.05205178e-01
1.55935717e+00 -1.27726138e+00 -8.01155567e-01 -8.48365873e-02
5.65903246e-01 -6.73999250e-01 8.23290825e-01 4.25775528e-01
-8.60899687e-01 -5.25830984e-01 -1.04056907e+00 8.54397938e-02
-2.42458761e-01 1.62189543e-01 3.72734785e-01 5.65799296e-01
-4.34249282e-01 4.51794356e-01 -5.36791563e-01 -5.61169013e-02
5.40437639e-01 7.42569417e-02 -5.81119537e-01 -5.89636981e-01
-9.22437251e-01 7.05407023e-01 3.45185697e-01 -1.14605598e-01
-4.89059716e-01 -6.25981092e-01 -7.56468058e-01 -8.20624232e-02
5.91740251e-01 -7.36919820e-01 7.65408337e-01 -9.97978806e-01
-1.09150922e+00 1.07116544e+00 -1.65613711e-01 3.23285878e-01
4.50571716e-01 2.43074093e-02 -6.91710532e-01 3.33455324e-01
2.92070627e-01 5.01945853e-01 7.47192025e-01 -1.80294156e+00
-8.36082935e-01 -7.40704834e-01 8.63994658e-02 4.72715586e-01
-5.32679141e-01 -1.69449344e-01 -4.14612234e-01 -7.33691514e-01
4.49361801e-01 -1.12907505e+00 1.41518876e-01 -1.46738077e-02
-3.13197464e-01 -2.94950247e-01 7.51534522e-01 -5.14832795e-01
1.32371402e+00 -2.07764316e+00 4.85152096e-01 3.49137932e-03
4.52954262e-01 2.56323665e-02 -7.33142942e-02 2.78619438e-01
-3.78408909e-01 3.35882716e-02 -1.33719459e-01 -6.49351954e-01
-1.45937771e-01 1.38347998e-01 1.99089468e-01 7.09930420e-01
8.11441243e-02 6.81264102e-01 -9.73488331e-01 -8.85052681e-01
1.15028933e-01 4.36911315e-01 -5.36751270e-01 6.98264837e-02
3.00041199e-01 5.77695310e-01 -3.52319717e-01 8.70386958e-01
9.22240674e-01 -6.94814980e-01 3.71227086e-01 -7.01980710e-01
1.13868512e-01 -1.50716171e-01 -1.20901978e+00 1.86615229e+00
-5.30281126e-01 -9.37695876e-02 1.47642925e-01 -1.02684891e+00
5.25381207e-01 4.67475682e-01 9.93105888e-01 -5.65679848e-01
-8.25616568e-02 1.44465551e-01 -3.72289419e-01 -4.08162862e-01
2.62332261e-01 -4.97310936e-01 2.24813763e-02 4.80797738e-01
1.59245282e-01 3.41511279e-01 6.59583211e-02 1.04979157e-01
5.33700228e-01 1.21988803e-01 6.31727815e-01 -7.39107898e-04
8.26456845e-01 -5.29057741e-01 1.06673288e+00 1.79540843e-01
-4.69874144e-01 7.58229256e-01 3.45677167e-01 -3.25373799e-01
-7.60679364e-01 -8.56709540e-01 -3.21124732e-01 1.06845021e+00
2.75577307e-01 -5.38008332e-01 -4.33085740e-01 -1.36622918e+00
-1.07401766e-01 2.64817834e-01 -5.40031791e-01 -1.47467643e-01
-3.35804880e-01 -9.35373008e-01 8.48683119e-02 3.36513042e-01
4.68655586e-01 -6.15252793e-01 -1.17280908e-01 1.24528058e-01
-6.18562162e-01 -1.04795456e+00 -6.23899162e-01 2.88469315e-01
-9.51184809e-01 -1.29379427e+00 -6.66704655e-01 -7.56022096e-01
7.29265034e-01 6.82006836e-01 1.22675467e+00 -8.02455842e-02
3.36949304e-02 3.83103997e-01 -4.84397382e-01 3.19388025e-02
-3.25178385e-01 1.10368602e-01 -4.91029210e-02 4.27507758e-01
4.01811957e-01 -6.86468840e-01 -6.69735312e-01 5.65071940e-01
-9.43291962e-01 2.34310448e-01 5.44588685e-01 1.20950413e+00
1.00710642e+00 -4.02099802e-04 7.16048777e-01 -1.07648766e+00
8.38934332e-02 -4.36612755e-01 -2.33096451e-01 6.38764918e-01
-1.11681890e+00 -1.75543085e-01 6.80658340e-01 -4.29785252e-01
-9.66215491e-01 1.26221776e-01 1.17937498e-01 -5.25717735e-01
-1.14802599e-01 3.98443252e-01 -5.38906515e-01 4.69972417e-02
8.88230577e-02 8.43475685e-02 -8.68157204e-03 -4.12653744e-01
4.27248776e-01 5.35155416e-01 -1.25555754e-01 -1.88697740e-01
6.11391604e-01 7.50350058e-01 2.70004332e-01 -1.58501714e-01
-1.47393227e+00 -7.90327132e-01 -9.01406407e-01 -3.58848572e-01
8.59714210e-01 -1.22961223e+00 -5.89708805e-01 4.67628092e-01
-7.97011018e-01 4.96921033e-01 -9.23679844e-02 4.46247458e-01
-4.06238645e-01 6.40727282e-01 -5.73048532e-01 -4.52760369e-01
-1.96845531e-01 -1.23957002e+00 1.18375027e+00 1.37847051e-01
1.98175654e-01 -1.09188664e+00 1.72806367e-01 9.25012410e-01
-1.82018895e-02 1.26065269e-01 9.90483105e-01 -6.72282457e-01
-2.74289370e-01 -3.67112607e-02 -3.63145620e-01 6.09552503e-01
3.77843678e-01 -4.56002980e-01 -1.15413809e+00 -6.31344020e-01
2.32972667e-01 -5.94319344e-01 9.28080797e-01 2.03292236e-01
1.14219272e+00 -9.36465114e-02 -3.73236924e-01 5.40693104e-01
1.62067533e+00 -8.50100890e-02 7.64604881e-02 1.37591690e-01
1.17742109e+00 9.08890724e-01 7.79293716e-01 3.95778149e-01
7.48978794e-01 8.45128477e-01 5.75473368e-01 -1.69509932e-01
3.56803946e-02 -1.32173255e-01 1.93959966e-01 1.35295713e+00
1.88100561e-01 -2.34316573e-01 -4.04876262e-01 2.72312254e-01
-1.71696353e+00 -1.16776550e+00 -4.52193655e-02 1.97890687e+00
6.25470817e-01 -2.02264741e-01 2.23247722e-01 8.63233060e-02
8.73131096e-01 6.30927742e-01 -6.57864571e-01 1.15333669e-01
-2.32859030e-01 -4.37679857e-01 3.04599136e-01 3.97897474e-02
-1.54711711e+00 3.36228013e-01 5.28998804e+00 9.59367394e-01
-9.18798447e-01 4.95084256e-01 7.49641240e-01 -2.11649582e-01
-3.92285019e-01 -6.28318042e-02 -8.13525498e-01 3.32433403e-01
4.29325461e-01 2.23237902e-01 1.60377592e-01 7.10569918e-01
-2.62374971e-02 6.87716678e-02 -1.08378041e+00 1.17063713e+00
5.25979102e-01 -7.97625542e-01 -1.14889778e-02 3.63055706e-01
8.69192421e-01 -1.92478314e-01 1.30452931e-01 2.48968497e-01
-8.16546902e-02 -5.65174282e-01 5.15229642e-01 2.93089032e-01
7.53922284e-01 -1.02368402e+00 6.00365818e-01 3.88184696e-01
-1.46695125e+00 -3.10590804e-01 -1.38161302e-01 4.34659779e-01
1.55482456e-01 9.17321086e-01 -2.39068180e-01 1.16023266e+00
4.70088214e-01 1.15367866e+00 -7.84285545e-01 5.68107963e-01
1.21384822e-01 2.63363570e-01 2.45116562e-01 5.28172135e-01
6.21382631e-02 -3.47537130e-01 2.93073922e-01 8.46165895e-01
1.53544575e-01 -9.87591892e-02 6.81373537e-01 3.09225410e-01
-1.14409924e-01 4.25278574e-01 -4.95989859e-01 4.61054742e-01
2.08602399e-01 1.62222779e+00 -6.54175401e-01 -1.95704505e-01
-1.02845013e+00 1.00376904e+00 4.92734134e-01 3.56750824e-02
-7.37252116e-01 2.84387380e-01 4.23204213e-01 -1.45513788e-01
3.62223297e-01 1.83601812e-01 -2.01376736e-01 -1.59650254e+00
1.45075545e-01 -1.06089187e+00 7.10746944e-01 -3.91835690e-01
-1.77423894e+00 3.03213120e-01 -8.53310898e-02 -1.79356229e+00
3.96294110e-02 -3.41460228e-01 4.51058596e-02 5.34160912e-01
-1.80345404e+00 -1.59832573e+00 -1.58428699e-01 4.12836432e-01
5.56482553e-01 -2.35246673e-01 8.79950762e-01 7.29476094e-01
-5.92309535e-01 7.29145765e-01 3.37249577e-01 -1.11703709e-01
1.21394181e+00 -1.28103983e+00 -4.03053313e-01 5.21823645e-01
2.03025207e-01 3.58803779e-01 1.94375485e-01 -5.07258654e-01
-9.34698105e-01 -1.18677831e+00 7.28691280e-01 -3.43101293e-01
3.44020039e-01 -2.18441963e-01 -9.68520761e-01 4.40770179e-01
1.82287782e-01 3.15294176e-01 1.27391815e+00 2.80764550e-01
-7.80210495e-01 -4.61904734e-01 -1.18038857e+00 3.41945410e-01
9.68729794e-01 -7.37020969e-01 -2.14451045e-01 3.60336363e-01
4.94211376e-01 -1.51562750e-01 -1.25809920e+00 6.73795700e-01
7.19124496e-01 -1.38807428e+00 1.02223587e+00 -3.24690849e-01
6.04911089e-01 -3.18214536e-01 -2.51608372e-01 -1.39383769e+00
-5.73037982e-01 1.97860882e-01 -8.85639042e-02 1.68401659e+00
1.76704198e-01 -5.03018081e-01 6.31075740e-01 1.83202237e-01
-1.09892618e-02 -1.03041387e+00 -9.85226512e-01 -5.68493724e-01
-9.00556799e-03 2.92616375e-02 3.53998154e-01 1.37422216e+00
-2.16509148e-01 5.55758715e-01 -6.74192369e-01 1.42794594e-01
8.81594539e-01 5.41760802e-01 2.20461532e-01 -1.47041821e+00
-3.71602327e-01 -2.74683386e-01 -2.15312660e-01 -7.32300103e-01
4.15411800e-01 -1.22310126e+00 -3.42027664e-01 -1.36868417e+00
1.01921380e+00 -4.52696055e-01 -6.77687049e-01 3.07758242e-01
-5.03827095e-01 2.17233464e-01 2.72275239e-01 4.25023794e-01
-8.95488024e-01 6.55594051e-01 1.51870048e+00 -2.15624422e-01
1.97998881e-01 5.05543090e-02 -8.63701880e-01 7.63270617e-01
5.18186152e-01 -6.86873734e-01 -5.06498575e-01 -6.72487617e-02
2.62993902e-01 1.43670514e-01 2.17461497e-01 -7.33419240e-01
4.79426160e-02 -1.75934196e-01 3.32118243e-01 -7.69855201e-01
3.25532585e-01 -1.15412092e+00 6.52266145e-02 2.64361382e-01
-3.17091852e-01 1.54280052e-01 -4.10285085e-01 9.63469565e-01
-4.43181694e-01 -1.41644180e-01 1.01151121e+00 -2.23024979e-01
-4.19600606e-01 4.54894692e-01 1.31902575e-01 1.84258103e-01
1.02883494e+00 4.81913425e-02 -2.64810026e-01 -8.36536568e-03
-6.78214610e-01 4.15782779e-01 6.49958313e-01 4.75211382e-01
3.67027283e-01 -1.71892762e+00 -5.47423542e-01 6.65435195e-02
6.51228189e-01 -6.47633374e-02 7.71252453e-01 1.04567337e+00
1.83790699e-01 1.93883851e-01 -1.30955473e-01 -7.41541564e-01
-1.58795357e+00 8.08022559e-01 1.76296115e-01 -9.08870697e-01
-3.47597599e-01 4.79976952e-01 4.91026074e-01 -6.76702321e-01
-1.99612491e-02 2.75411963e-01 -5.89874625e-01 7.70423830e-01
4.34787840e-01 4.32047904e-01 9.98946130e-02 -9.51180100e-01
-4.23898518e-01 9.79765654e-01 -3.48887324e-01 1.70900777e-01
1.40127563e+00 -5.45999885e-01 -2.24139214e-01 8.68470550e-01
1.66940105e+00 2.47398810e-03 -1.11498737e+00 -4.51757580e-01
-2.04672232e-01 -3.62890273e-01 -2.93416288e-02 -7.41954684e-01
-1.37698913e+00 8.72368217e-01 8.64420474e-01 3.11155561e-02
1.37413192e+00 -8.88053402e-02 6.43316507e-01 -7.06428066e-02
2.66103387e-01 -1.22373652e+00 4.70755607e-01 2.19076142e-01
4.12999928e-01 -1.73656785e+00 4.30057615e-01 -6.96317673e-01
-7.83139765e-01 9.00476456e-01 6.80370629e-01 3.45592469e-01
6.97747350e-01 -1.39976367e-01 3.22423540e-02 -3.40536445e-01
-6.82174802e-01 8.89462009e-02 4.30065751e-01 2.63752282e-01
4.52701390e-01 1.01989061e-01 -4.22191679e-01 3.79217952e-01
5.46343386e-01 -2.31599107e-01 1.43208936e-01 9.03246522e-01
-1.17322221e-01 -1.35089970e+00 -1.83102280e-01 4.87554818e-01
-6.99786842e-01 7.27641135e-02 -8.19359645e-02 5.61770678e-01
4.09946620e-01 9.21190798e-01 -3.41487139e-01 -5.53843081e-01
3.02332401e-01 3.72394413e-01 4.16726023e-01 -5.40085614e-01
-7.15300083e-01 3.33281606e-01 -7.79778212e-02 -4.65174675e-01
-1.09146154e+00 -9.14510787e-01 -7.38294840e-01 6.10256530e-02
-6.95610762e-01 -2.22882301e-01 3.84052128e-01 9.40550447e-01
3.73235106e-01 4.95484382e-01 1.18649864e+00 -6.11974120e-01
-6.94761097e-01 -7.31544375e-01 -8.26399684e-01 7.38380790e-01
2.43081450e-01 -8.70311439e-01 -4.90620673e-01 -5.17476797e-02] | [8.612852096557617, 4.448630332946777] |
1a0445cc-c5a4-4aa8-a0ac-1a817de18ddf | mutexmatch-semi-supervised-learning-with-1 | 2203.14316 | null | https://arxiv.org/abs/2203.14316v2 | https://arxiv.org/pdf/2203.14316v2.pdf | MutexMatch: Semi-Supervised Learning with Mutex-Based Consistency Regularization | The core issue in semi-supervised learning (SSL) lies in how to effectively leverage unlabeled data, whereas most existing methods tend to put a great emphasis on the utilization of high-confidence samples yet seldom fully explore the usage of low-confidence samples. In this paper, we aim to utilize low-confidence samples in a novel way with our proposed mutex-based consistency regularization, namely MutexMatch. Specifically, the high-confidence samples are required to exactly predict "what it is" by conventional True-Positive Classifier, while the low-confidence samples are employed to achieve a simpler goal -- to predict with ease "what it is not" by True-Negative Classifier. In this sense, we not only mitigate the pseudo-labeling errors but also make full use of the low-confidence unlabeled data by consistency of dissimilarity degree. MutexMatch achieves superior performance on multiple benchmark datasets, i.e., CIFAR-10, CIFAR-100, SVHN, STL-10, mini-ImageNet and Tiny-ImageNet. More importantly, our method further shows superiority when the amount of labeled data is scarce, e.g., 92.23% accuracy with only 20 labeled data on CIFAR-10. Our code and model weights have been released at https://github.com/NJUyued/MutexMatch4SSL. | ['Yang Gao', 'Yinghuan Shi', 'Luping Zhou', 'Lei Wang', 'Lei Qi', 'Zhen Zhao', 'Yue Duan'] | 2022-03-27 | mutexmatch-semi-supervised-learning-with | https://openreview.net/forum?id=r5hq-Ooh_Ba | https://openreview.net/pdf?id=r5hq-Ooh_Ba | null | ['semi-supervised-image-classification'] | ['computer-vision'] | [-2.59812146e-01 -6.52938858e-02 -4.76050019e-01 -6.35553896e-01
-9.01768863e-01 -3.46167296e-01 2.87194967e-01 -2.75477618e-02
-5.56236207e-01 1.12909973e+00 -1.93301424e-01 -2.63417900e-01
1.35539740e-01 -5.67275047e-01 -5.89170098e-01 -7.28272557e-01
3.34303886e-01 2.66703725e-01 9.28867906e-02 1.01449326e-01
1.21584516e-02 5.28398529e-02 -1.29560769e+00 4.37348001e-02
9.49298859e-01 1.11228216e+00 4.56667691e-02 4.27798703e-02
2.51877084e-02 7.74636686e-01 -2.22987175e-01 -2.75451064e-01
3.52840871e-01 -3.31668317e-01 -6.55054271e-01 6.95654526e-02
4.59274948e-01 -1.90068886e-01 -4.85852435e-02 1.32364845e+00
3.15455377e-01 1.91786550e-02 6.43468201e-01 -1.35716939e+00
-5.45342088e-01 7.31263280e-01 -9.33768630e-01 2.21221566e-01
-2.53917694e-01 1.71869129e-01 1.02497876e+00 -1.20800912e+00
4.92689520e-01 9.63278234e-01 6.20628297e-01 5.02007663e-01
-1.26602817e+00 -1.16240764e+00 2.82004386e-01 1.42963082e-01
-1.64553463e+00 -3.63962203e-01 6.91409171e-01 -4.56806958e-01
2.66068101e-01 1.29950032e-01 3.88512731e-01 9.56167459e-01
-1.37395382e-01 8.46645534e-01 1.48833501e+00 -3.35667133e-01
3.54289562e-01 4.23132509e-01 5.89991271e-01 6.30492866e-01
3.39152932e-01 1.00332014e-01 -2.23933145e-01 -9.95602366e-03
5.45124114e-01 2.75618523e-01 -2.74356902e-01 -1.42240331e-01
-1.08809519e+00 8.81421626e-01 3.18814933e-01 2.05523908e-01
-1.99329048e-01 -2.88334042e-01 4.02162254e-01 2.44591847e-01
7.28243649e-01 1.10868894e-01 -6.90873504e-01 2.20421165e-01
-1.08728290e+00 -1.05563506e-01 5.54082572e-01 9.85276878e-01
1.07016051e+00 1.26436979e-01 -1.00590639e-01 1.10352814e+00
3.87262881e-01 5.57278037e-01 6.46798074e-01 -7.69769847e-01
4.51260924e-01 5.07361770e-01 5.11023439e-02 -6.47014916e-01
-2.00403735e-01 -7.85003245e-01 -1.14139044e+00 2.02280864e-01
3.14014256e-01 -2.85434276e-01 -1.05040312e+00 1.82258141e+00
3.80508423e-01 4.19534206e-01 8.03113729e-02 1.03331351e+00
7.72364557e-01 4.16030109e-01 8.83413702e-02 -2.78254569e-01
1.25440800e+00 -1.15559995e+00 -4.71068263e-01 -2.96411365e-01
6.90669417e-01 -6.57977283e-01 1.28974843e+00 3.57217461e-01
-6.45641267e-01 -6.19010210e-01 -1.14153481e+00 2.85220176e-01
-2.14572996e-01 5.29531360e-01 5.64214826e-01 5.47342241e-01
-5.18011272e-01 4.56101865e-01 -5.52073240e-01 9.75066945e-02
5.37503183e-01 1.58451676e-01 -5.04542768e-01 -2.07851619e-01
-1.12701046e+00 4.15029317e-01 7.08439887e-01 5.45927063e-02
-6.91603124e-01 -7.29039609e-01 -6.27276182e-01 -7.03633875e-02
6.16888165e-01 -1.95979670e-01 1.15714872e+00 -1.05563676e+00
-1.08830905e+00 8.21838379e-01 -3.74300331e-02 -5.41557670e-01
6.11144304e-01 -1.09653428e-01 -6.06583297e-01 7.05868006e-02
3.24528158e-01 7.99777985e-01 5.64965248e-01 -1.32407463e+00
-7.00084090e-01 -2.18617052e-01 -1.84600234e-01 -1.25863021e-02
-4.29152310e-01 -3.54960978e-01 -3.78786832e-01 -7.70345032e-01
3.17091584e-01 -1.12356591e+00 -1.55296370e-01 -3.21125761e-02
-6.19439304e-01 -1.79963991e-01 7.08584189e-01 -2.37428591e-01
1.18677151e+00 -2.22245407e+00 -6.28814697e-01 2.69146174e-01
3.18399996e-01 6.89567029e-01 -1.24639243e-01 1.35528501e-02
-1.69180930e-01 2.25314990e-01 -3.03908914e-01 -2.65727580e-01
-2.07160950e-01 2.37721756e-01 -2.43467405e-01 3.01793188e-01
1.39595032e-01 6.46466315e-01 -9.19703603e-01 -6.95393860e-01
2.96663016e-01 2.97448784e-01 -3.32238257e-01 -1.34295942e-02
-7.76548982e-02 3.25736642e-01 -4.94692653e-01 6.57155275e-01
8.94573390e-01 -6.66465819e-01 3.64830717e-02 -3.98816764e-01
-3.41409184e-02 -9.74305123e-02 -1.34794796e+00 1.25648236e+00
-3.02002102e-01 2.34426022e-01 -1.09191075e-01 -9.95779276e-01
1.07349551e+00 2.24915683e-01 3.22097838e-01 -5.81740022e-01
7.49815023e-03 4.23017889e-01 -8.18978548e-02 -2.09115207e-01
1.93087697e-01 -1.50828779e-01 1.99917495e-01 2.94000387e-01
-4.90511321e-02 2.72149116e-01 3.26164246e-01 3.19611698e-01
6.59168720e-01 -6.51714299e-03 3.70095372e-01 -3.50522071e-01
5.92449784e-01 1.54641634e-02 9.96490240e-01 6.03316307e-01
-3.49136800e-01 6.70971513e-01 2.93003231e-01 -1.93542123e-01
-7.89469182e-01 -1.05938315e+00 -2.78987437e-01 7.63602734e-01
1.24196097e-01 -2.76263952e-01 -5.06464720e-01 -1.01020789e+00
1.81522444e-02 7.35182703e-01 -5.52378714e-01 -1.38874471e-01
-1.27439946e-01 -8.55026484e-01 4.43359971e-01 5.40552139e-01
9.85297024e-01 -8.24143529e-01 -1.63091328e-02 8.83841813e-02
-2.00968429e-01 -1.09546185e+00 -4.99153197e-01 2.21641883e-01
-8.46218586e-01 -1.19455111e+00 -5.44278622e-01 -7.98089862e-01
8.26406896e-01 4.50688273e-01 1.04026926e+00 9.15665105e-02
-8.08212981e-02 -1.46015108e-01 -4.95095283e-01 -2.65063018e-01
-2.60840990e-02 4.60012965e-02 5.85059561e-02 6.41771406e-02
4.47858512e-01 -4.36494529e-01 -6.49188578e-01 7.09618986e-01
-6.81431651e-01 1.06409989e-01 6.93057239e-01 1.16365671e+00
1.14749622e+00 3.44922021e-02 1.03717732e+00 -1.27997005e+00
1.76563337e-01 -7.39955306e-01 -4.76251096e-01 4.20921117e-01
-1.03013778e+00 -1.55293599e-01 9.62436855e-01 -7.16645241e-01
-9.63847935e-01 -2.57485248e-02 -2.59137481e-01 -7.14716077e-01
-1.16763771e-01 5.98441064e-01 -1.09289683e-01 7.41168782e-02
6.63393259e-01 2.03377634e-01 -1.49595574e-01 -5.12215555e-01
4.70247529e-02 7.64501989e-01 4.37659711e-01 -7.01068401e-01
6.79947793e-01 5.90235949e-01 -1.75202385e-01 -6.45644903e-01
-1.31338096e+00 -5.64456522e-01 -3.49513888e-01 5.29657863e-02
3.63593608e-01 -1.17020869e+00 -4.76465434e-01 4.33636278e-01
-4.30365711e-01 -4.25671369e-01 -3.29773456e-01 8.82898629e-01
-4.81632091e-02 4.57516015e-01 -6.63446724e-01 -8.16853344e-01
-5.82307220e-01 -1.03709114e+00 6.68586791e-01 3.22841644e-01
1.44134045e-01 -8.45658004e-01 -1.11233532e-01 4.28268522e-01
1.68470651e-01 2.80057965e-03 4.82016325e-01 -9.67205167e-01
-3.25141966e-01 -2.23201200e-01 -5.52686870e-01 8.43076468e-01
2.63532817e-01 6.75822198e-02 -1.05538762e+00 -4.38651294e-01
-1.57825172e-01 -7.90276170e-01 9.47694600e-01 3.22785378e-01
1.41246891e+00 -1.38013855e-01 -5.69070056e-02 5.84654450e-01
1.48111701e+00 -3.56568471e-02 3.92194867e-01 5.93704544e-02
5.92810392e-01 4.14854079e-01 1.16560531e+00 4.71605986e-01
4.46896493e-01 2.75509596e-01 2.73368359e-01 -1.23679616e-01
-3.53607349e-02 -3.78628165e-01 -9.14343670e-02 1.04248917e+00
1.64887667e-01 -7.55937472e-02 -8.01455557e-01 2.83834845e-01
-1.91852808e+00 -7.92258203e-01 -2.74395555e-01 2.18760276e+00
1.02257001e+00 3.69975656e-01 -1.00001410e-01 2.20398620e-01
8.35856199e-01 -3.76591906e-02 -6.88911915e-01 3.44654232e-01
-2.40506753e-01 4.78501134e-02 4.93046969e-01 2.84052879e-01
-1.26032782e+00 7.98530281e-01 5.03056335e+00 1.29770613e+00
-1.20233488e+00 1.84748143e-01 1.16612911e+00 -5.33302128e-03
-1.79096803e-01 1.06674656e-01 -1.01946414e+00 8.07842910e-01
6.07590139e-01 9.36588943e-02 7.86114763e-03 1.06837285e+00
2.92392433e-01 -2.86760837e-01 -9.26580310e-01 1.01553845e+00
-1.79044276e-01 -1.18270183e+00 -2.64003605e-01 -1.33718938e-01
8.41450930e-01 1.62910655e-01 3.87913100e-02 5.93280554e-01
2.50972152e-01 -7.91296065e-01 7.04887331e-01 3.61887395e-01
1.04504526e+00 -6.35387421e-01 8.87194395e-01 6.43026769e-01
-1.17004025e+00 1.96801230e-01 -5.01150906e-01 2.17547700e-01
-1.05289370e-01 1.17888474e+00 -6.16988063e-01 5.15990317e-01
6.71314836e-01 8.41644347e-01 -3.96436453e-01 1.01045573e+00
-3.73613656e-01 1.03327739e+00 -3.65286618e-01 4.00995538e-02
2.44407326e-01 -2.18787357e-01 2.03026328e-02 1.13857543e+00
1.97146654e-01 2.30352953e-01 5.33865809e-01 6.47217333e-01
-2.31199488e-01 2.27343589e-01 -1.53109342e-01 1.71702459e-01
7.08924770e-01 1.41678381e+00 -7.33610272e-01 -6.47345722e-01
-4.98384297e-01 4.94747728e-01 4.11708117e-01 3.45813096e-01
-8.92185986e-01 -2.33388260e-01 3.24451208e-01 -1.89683661e-02
2.67998755e-01 -8.41381215e-03 -2.10972518e-01 -1.44684613e+00
2.98564485e-03 -8.01745892e-01 4.51162308e-01 -7.42733657e-01
-1.70964825e+00 6.47187054e-01 -6.43788055e-02 -1.45834994e+00
3.84179801e-02 -4.57858622e-01 -5.52030861e-01 8.65713656e-01
-1.68962240e+00 -1.03324974e+00 -3.86975914e-01 6.83281243e-01
4.89626557e-01 -9.14000943e-02 5.35803258e-01 6.23213887e-01
-7.71228313e-01 8.45832646e-01 1.83789521e-01 3.96702915e-01
9.41747129e-01 -1.01595855e+00 -8.40533245e-03 7.15683818e-01
1.46181002e-01 6.32896841e-01 4.40743834e-01 -6.43147111e-01
-6.90406859e-01 -1.27553523e+00 5.90936363e-01 1.12603866e-01
4.91800010e-01 -1.43143132e-01 -1.09659398e+00 5.87826550e-01
-2.86401212e-01 7.10212409e-01 7.77700841e-01 6.52127266e-02
-7.26296902e-01 -3.76125544e-01 -1.39236009e+00 5.12544334e-01
8.94245684e-01 -4.51094389e-01 -3.27077568e-01 2.82808095e-01
5.86799920e-01 -1.40965685e-01 -8.41255128e-01 6.67483866e-01
4.59411502e-01 -9.69329417e-01 8.57356966e-01 -3.73094887e-01
3.49787056e-01 -5.20754337e-01 -3.08966160e-01 -1.20581341e+00
-1.92589134e-01 7.85206184e-02 -8.30616206e-02 1.32442236e+00
5.14370501e-01 -1.01528227e+00 1.03020906e+00 3.99740785e-01
-1.10062152e-01 -1.02890050e+00 -7.41988838e-01 -8.40391278e-01
1.86899155e-02 -4.61149991e-01 4.18261856e-01 1.33735716e+00
-3.59124750e-01 1.96541891e-01 -5.56930780e-01 1.90014094e-02
8.38391960e-01 1.93835676e-01 6.44327700e-01 -1.17927039e+00
-3.99697095e-01 -1.53262928e-01 -6.33159950e-02 -8.88463616e-01
1.85733259e-01 -8.51900816e-01 -6.82848245e-02 -1.02348101e+00
3.90971363e-01 -1.05970275e+00 -6.51151597e-01 8.15382719e-01
-2.81489909e-01 5.02178431e-01 4.41756621e-02 5.02841890e-01
-7.61783421e-01 5.60902059e-01 1.17076385e+00 -5.81133626e-02
5.10698669e-02 1.14581600e-01 -7.42252171e-01 7.40845919e-01
1.23423266e+00 -5.20080626e-01 -5.57080269e-01 -1.17068447e-01
-3.53035852e-02 -3.38700823e-02 3.92279446e-01 -1.02803981e+00
7.20119774e-02 -2.74362177e-01 3.65021348e-01 -7.02032030e-01
8.75744969e-02 -6.55402660e-01 3.45196538e-02 3.78978878e-01
-4.39495295e-01 -2.45219424e-01 -9.73678753e-02 5.37132740e-01
-3.80956054e-01 -4.44201916e-01 8.99572194e-01 -5.18829376e-02
-7.81381249e-01 5.72142065e-01 1.36618525e-01 4.54968274e-01
9.32344258e-01 -9.82321203e-02 -3.62082601e-01 -6.66214898e-02
-6.25125945e-01 4.03066158e-01 3.66266131e-01 1.67987227e-01
5.35066724e-01 -1.36855805e+00 -7.19722331e-01 2.16080338e-01
3.41362745e-01 2.88266260e-02 3.70890796e-01 8.22113395e-01
-2.36976728e-01 2.06809074e-01 -6.31145090e-02 -6.51632667e-01
-1.12356460e+00 4.46827680e-01 1.33497238e-01 -3.06319773e-01
-3.53921592e-01 7.77669787e-01 1.76017672e-01 -5.50096691e-01
8.78519341e-02 -8.05345774e-02 -2.14427263e-02 -4.36377823e-02
4.17613149e-01 2.60560095e-01 5.90609275e-02 -3.38543653e-01
-3.72586459e-01 3.43673617e-01 -3.22368860e-01 2.01368257e-02
1.07904577e+00 -9.25927684e-02 1.91543534e-01 5.05095363e-01
1.30921924e+00 3.66905294e-02 -1.37638688e+00 -6.29750073e-01
-1.10210082e-03 -5.53922951e-01 -4.35491987e-02 -8.09123337e-01
-1.32716644e+00 8.07449400e-01 6.35245323e-01 -1.40932262e-01
9.57822680e-01 -1.21203400e-01 6.03922188e-01 4.97760147e-01
4.75578696e-01 -1.24665105e+00 1.56013370e-01 4.01230723e-01
2.83493936e-01 -1.78943777e+00 1.26823053e-01 -3.87785316e-01
-8.44280243e-01 7.38552690e-01 8.69975448e-01 -7.77218118e-02
8.53344679e-01 8.17351714e-02 9.97807458e-02 6.79272339e-02
-8.19825828e-01 -1.23980962e-01 1.87424019e-01 2.48174652e-01
5.37576854e-01 2.77999729e-01 -3.79325867e-01 6.83358192e-01
5.71450852e-02 1.75027996e-01 1.97625533e-01 7.97202766e-01
-4.05875891e-01 -9.35810268e-01 -2.08568007e-01 8.97060752e-01
-4.94040698e-01 -2.43310809e-01 8.39177519e-02 8.82134855e-01
2.42586344e-01 9.74593997e-01 -1.74133793e-01 -4.59279716e-01
-2.73423363e-02 -8.66939723e-02 1.15166083e-02 -6.63755059e-01
-3.02270591e-01 6.41834363e-02 1.02712415e-01 -1.72888175e-01
-4.89584416e-01 -4.17414963e-01 -1.34399176e+00 -3.08192879e-01
-6.11234307e-01 3.68007153e-01 4.72749323e-01 8.81237328e-01
1.44675523e-01 2.59438068e-01 8.46702754e-01 -2.85963297e-01
-8.99710238e-01 -9.65252399e-01 -7.44794548e-01 4.43177253e-01
5.18353917e-02 -5.67134976e-01 -5.36152482e-01 -2.51364648e-01] | [9.419179916381836, 3.7201550006866455] |
47c89756-7500-463e-821d-bbaf1e48e552 | dimension-independent-mixup-for-hard-negative | 2306.15905 | null | https://arxiv.org/abs/2306.15905v1 | https://arxiv.org/pdf/2306.15905v1.pdf | Dimension Independent Mixup for Hard Negative Sample in Collaborative Filtering | Collaborative filtering (CF) is a widely employed technique that predicts user preferences based on past interactions. Negative sampling plays a vital role in training CF-based models with implicit feedback. In this paper, we propose a novel perspective based on the sampling area to revisit existing sampling methods. We point out that current sampling methods mainly focus on Point-wise or Line-wise sampling, lacking flexibility and leaving a significant portion of the hard sampling area un-explored. To address this limitation, we propose Dimension Independent Mixup for Hard Negative Sampling (DINS), which is the first Area-wise sampling method for training CF-based models. DINS comprises three modules: Hard Boundary Definition, Dimension Independent Mixup, and Multi-hop Pooling. Experiments with real-world datasets on both matrix factorization and graph-based models demonstrate that DINS outperforms other negative sampling methods, establishing its effectiveness and superiority. Our work contributes a new perspective, introduces Area-wise sampling, and presents DINS as a novel approach that achieves state-of-the-art performance for negative sampling. Our implementations are available in PyTorch. | ['Philip S. Yu', 'Xiaolong Liu', 'Tianyu Lin', 'Chao Zhou', 'Jibing Gong', 'Liangwei Yang', 'Xi Wu'] | 2023-06-28 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [ 1.49261847e-01 -2.74289817e-01 -7.17775285e-01 -5.17560482e-01
-4.84837860e-01 -4.43758875e-01 5.42898238e-01 4.77679260e-02
-3.68000716e-01 8.36986601e-01 2.30902269e-01 -6.88444614e-01
-2.46916369e-01 -9.94709790e-01 -3.56462538e-01 -2.18382269e-01
-2.31837705e-01 4.45792854e-01 3.83233041e-01 -3.84755313e-01
3.49381715e-01 1.85052693e-01 -1.42003834e+00 8.30115676e-01
1.12256742e+00 9.02436376e-01 1.89118385e-01 5.41078150e-01
-2.98836082e-01 6.38668954e-01 -4.64238405e-01 -4.84949082e-01
2.77674943e-01 -3.24164540e-01 -7.73797452e-01 -4.12026852e-01
3.23269933e-01 -2.50106424e-01 9.66737419e-02 7.71315157e-01
6.06107116e-01 5.36294878e-01 5.71763754e-01 -1.28067827e+00
-6.88407362e-01 8.40699613e-01 -5.85669339e-01 2.63917387e-01
5.54677784e-01 -1.48227468e-01 1.21270585e+00 -1.16739583e+00
4.72384512e-01 1.28461683e+00 8.67807209e-01 6.75640881e-01
-1.27268648e+00 -8.38436425e-01 9.57630873e-01 1.69251129e-01
-1.29829192e+00 -1.05206996e-01 7.45411336e-01 -2.46885747e-01
9.29448485e-01 7.59828687e-01 1.11135364e+00 9.86842334e-01
-2.80371130e-01 1.43563187e+00 1.01863027e+00 -5.79109371e-01
4.44286168e-01 4.78670090e-01 5.01187801e-01 3.72144639e-01
2.14406788e-01 3.50793488e-02 -6.56791449e-01 -7.42690563e-01
7.27554977e-01 3.96064043e-01 -5.21750927e-01 -5.49290180e-01
-7.40731895e-01 9.34839487e-01 3.99282098e-01 2.93845832e-01
-2.60965139e-01 -3.65287811e-01 2.10931018e-01 5.11070430e-01
7.98354506e-01 4.01387662e-01 -7.78757989e-01 -1.40028849e-01
-1.02968884e+00 3.97509634e-01 1.18420732e+00 1.01842749e+00
5.77793658e-01 -3.48598003e-01 -4.65764791e-01 1.01751542e+00
4.75578398e-01 2.95850456e-01 4.40766394e-01 -5.19734085e-01
4.35460210e-01 7.87447035e-01 3.14823598e-01 -9.09210742e-01
-2.36459345e-01 -6.79002523e-01 -9.30336773e-01 -1.64956018e-01
1.34125933e-01 -1.80452809e-01 -6.84726000e-01 1.46469998e+00
5.30517995e-01 2.59460658e-01 -5.39778471e-01 9.08662915e-01
7.18245864e-01 3.66345286e-01 1.48428991e-01 -4.10040826e-01
8.79042208e-01 -1.41559434e+00 -6.45297468e-01 -2.72119157e-02
7.90178657e-01 -7.06410706e-01 1.34929955e+00 9.01586115e-01
-9.75160301e-01 -4.82108831e-01 -8.79936814e-01 2.65967995e-01
-6.01705372e-01 1.52657315e-01 1.29707658e+00 1.22665632e+00
-1.14560902e+00 7.71879911e-01 -4.49084818e-01 -2.27235973e-01
3.63530487e-01 5.49549103e-01 3.65894698e-02 -1.73042223e-01
-1.43723714e+00 3.34041893e-01 -1.16377287e-01 -1.42153082e-02
-2.17192888e-01 -1.01291180e+00 -5.21154165e-01 4.55209240e-02
4.37348008e-01 -8.81833375e-01 1.31896842e+00 -1.06940186e+00
-1.55687881e+00 6.93860948e-02 -5.43910563e-01 -4.83846515e-01
6.67797863e-01 -5.10319650e-01 -5.66045046e-01 -2.28544250e-01
-1.58887535e-01 2.04890668e-01 6.97205961e-01 -1.16431773e+00
-3.74591678e-01 -3.48341465e-01 3.09015989e-01 2.73592085e-01
-9.94937897e-01 -2.69464612e-01 -3.54291052e-01 -7.04567611e-01
-6.76102340e-02 -6.89633787e-01 -7.77735054e-01 -1.55307800e-01
-2.04447329e-01 -5.81752181e-01 6.62797689e-01 -9.82374996e-02
2.21945953e+00 -1.79551244e+00 -3.17278117e-01 5.93186319e-01
3.53570431e-01 7.53458738e-01 -2.07374409e-01 7.99189866e-01
1.87742397e-01 4.61739570e-01 9.86839831e-02 -6.41947031e-01
-5.64022250e-02 -7.73018748e-02 -1.40668601e-01 1.44788384e-01
-9.78808776e-02 6.38958812e-01 -1.33247268e+00 -2.93275744e-01
2.38907501e-01 5.90434551e-01 -1.06475914e+00 1.06509469e-01
-2.26628467e-01 2.03274861e-01 -4.15003866e-01 6.34740472e-01
1.09126508e+00 -3.85149151e-01 5.88086128e-01 -1.97268352e-01
3.39719132e-02 4.43739414e-01 -1.56287432e+00 1.41574514e+00
-3.58556926e-01 -8.46947581e-02 4.55110408e-02 -7.03895688e-01
7.43005395e-01 1.17979497e-01 3.58304143e-01 -4.87514526e-01
-5.08029014e-02 2.76600808e-01 -3.69017661e-01 -1.64850593e-01
8.99322808e-01 1.77621663e-01 4.19725329e-01 6.00936711e-01
3.72528136e-02 2.71223366e-01 3.92545044e-01 5.71709931e-01
8.00952137e-01 -4.39201854e-02 3.67886841e-01 -4.12593186e-01
5.57255089e-01 -5.96472025e-01 4.60465252e-01 1.05399859e+00
-2.58105397e-01 5.85392714e-01 2.04033747e-01 -3.10826749e-01
-3.23208809e-01 -8.02637935e-01 1.29453123e-01 1.41650200e+00
2.01935284e-02 -1.21376371e+00 -5.05478501e-01 -1.22835422e+00
1.11540832e-01 6.17941380e-01 -9.28575456e-01 1.14483312e-01
-2.89467663e-01 -7.83826470e-01 -1.57538220e-01 5.46804667e-01
2.81831294e-01 -7.57821739e-01 2.01344028e-01 -1.77517685e-03
-1.49426907e-01 -3.07242572e-01 -6.50858283e-01 1.27238547e-02
-1.09697318e+00 -1.20785630e+00 -7.71471441e-01 -5.35517991e-01
7.55576491e-01 9.11955655e-01 1.49748814e+00 3.60211849e-01
2.21469030e-01 4.14929867e-01 -6.56383336e-01 -3.95989984e-01
3.16565722e-01 5.02245486e-01 2.35324223e-02 1.69829831e-01
8.92675400e-01 -6.93593681e-01 -1.09078157e+00 5.36148608e-01
-6.71641588e-01 -1.79234073e-01 4.51061666e-01 8.57214749e-01
4.07839566e-01 -4.04956967e-01 7.72290170e-01 -1.75571299e+00
1.21879911e+00 -7.20955253e-01 -1.23460829e-01 2.23117873e-01
-1.01024044e+00 -3.64695042e-01 7.06198335e-01 -7.34695137e-01
-9.46898341e-01 -1.22855499e-01 -3.30420256e-01 -1.98915273e-01
2.16729492e-01 5.88610649e-01 3.58375125e-02 -1.18715987e-01
1.00645113e+00 -8.64206329e-02 -3.04709196e-01 -6.93644822e-01
4.36584741e-01 9.31993425e-01 -4.63184983e-01 -6.51148617e-01
3.31524193e-01 3.86778504e-01 -4.65433896e-01 -7.98414290e-01
-8.82303894e-01 -1.00976920e+00 -4.61923003e-01 -2.06852928e-01
7.97547549e-02 -7.04050958e-01 -7.58319080e-01 4.68049012e-02
-5.91431797e-01 -3.61966521e-01 -3.79299760e-01 5.70009351e-01
-8.25609788e-02 3.78776640e-01 -4.15430516e-01 -1.50278044e+00
-7.21928120e-01 -6.01308465e-01 7.88384199e-01 4.84049506e-02
-4.44024354e-01 -1.11092818e+00 1.62742198e-01 2.47767687e-01
7.82780468e-01 -3.68034810e-01 4.63112056e-01 -6.19764924e-01
-2.72186011e-01 -2.60602504e-01 -1.42095461e-01 3.80741715e-01
-7.66065642e-02 -6.52266219e-02 -9.70566630e-01 -5.64844310e-01
-1.97599858e-01 -2.95916468e-01 1.11067772e+00 3.70307833e-01
1.32127571e+00 -2.45719984e-01 -5.77055395e-01 4.27112758e-01
1.33062720e+00 -5.94659746e-02 5.58435380e-01 2.04531118e-01
3.88718337e-01 2.39509717e-01 7.90741324e-01 6.95061326e-01
3.45254064e-01 2.97224671e-01 3.49679321e-01 -3.16939473e-01
1.22945085e-01 -3.78540069e-01 2.57785738e-01 7.71401703e-01
-1.71451584e-01 -3.30794513e-01 -2.27455065e-01 4.34089005e-01
-2.05720496e+00 -9.91948903e-01 -3.21978748e-01 2.31527209e+00
7.73397803e-01 1.17380388e-01 5.49921930e-01 4.09503460e-01
6.74944401e-01 1.16079748e-01 -3.30448210e-01 -4.20484543e-01
3.38210404e-01 3.55097562e-01 2.94582009e-01 6.07630372e-01
-1.05202222e+00 7.53340781e-01 6.72955608e+00 1.02136946e+00
-9.21181142e-01 1.09580196e-01 4.94458914e-01 -2.58485943e-01
-6.24778628e-01 -1.42116576e-01 -1.09308827e+00 4.10816640e-01
7.59548187e-01 7.18339309e-02 4.23401833e-01 1.02105236e+00
2.28412345e-01 -8.15025046e-02 -9.28732872e-01 7.55582094e-01
-4.04137224e-02 -1.26214921e+00 2.75783718e-01 -1.73828885e-01
8.80181909e-01 5.54783046e-02 -8.17081891e-03 7.10756242e-01
4.91187960e-01 -6.39791787e-01 4.54979241e-01 5.10663390e-01
6.36696935e-01 -7.19349802e-01 6.71921313e-01 4.76599693e-01
-1.42377555e+00 -1.79466203e-01 -4.41893727e-01 -4.74680990e-01
1.42545789e-01 1.01198888e+00 -7.36585200e-01 6.86173618e-01
7.18393505e-01 9.51749861e-01 -5.36522090e-01 1.27077830e+00
-6.46173507e-02 1.02816808e+00 -3.00395995e-01 -4.69140530e-01
-9.59270746e-02 -2.57436633e-01 2.22234949e-01 1.49092948e+00
1.72146842e-01 5.40227331e-02 4.01104540e-01 5.14285564e-01
-7.86528364e-02 4.21068251e-01 -4.09472495e-01 -1.30628161e-02
6.59803867e-01 1.15146494e+00 -5.72842181e-01 -5.45043051e-01
-5.10998130e-01 6.64052665e-01 4.32823122e-01 4.79110777e-01
-5.27767897e-01 -3.08155388e-01 5.90870559e-01 3.41678143e-01
3.55998278e-01 -1.02888472e-01 -3.44704121e-01 -1.36800289e+00
2.59479601e-02 -9.09884691e-01 5.66601694e-01 -1.61407232e-01
-1.69591475e+00 4.74931031e-01 -1.16790637e-01 -1.47817111e+00
-1.13945175e-02 -3.82260978e-01 -3.90873611e-01 9.89013612e-01
-1.55971527e+00 -9.13498700e-01 -2.91715413e-01 7.42045879e-01
4.98310477e-01 5.85417636e-02 1.09660757e+00 6.21720850e-01
-3.12689483e-01 9.33278441e-01 7.55803734e-02 -2.27284446e-01
8.15409720e-01 -1.23908913e+00 5.49692571e-01 4.66492444e-01
1.39092086e-02 1.51815343e+00 3.16718966e-01 -6.63066387e-01
-1.30358863e+00 -1.04929173e+00 1.03905821e+00 -3.43399525e-01
4.19745564e-01 -7.27821410e-01 -6.51161730e-01 3.73621792e-01
-3.26769762e-02 4.10970934e-02 1.30865979e+00 8.91509116e-01
-2.77844459e-01 -2.29166105e-01 -1.33295357e+00 5.68571627e-01
1.51670611e+00 -3.30258191e-01 -1.46215335e-01 4.42955583e-01
6.62458062e-01 -6.97557777e-02 -7.76986241e-01 4.44597155e-01
8.87412667e-01 -1.43680084e+00 1.00038564e+00 -4.80229974e-01
-1.15260214e-01 -7.17838183e-02 1.25471335e-02 -1.30264580e+00
-6.40235722e-01 -8.65429878e-01 -5.75506926e-01 1.13184869e+00
6.57821000e-01 -9.34654832e-01 1.22584319e+00 4.59673196e-01
1.68351918e-01 -1.24807727e+00 -2.16337487e-01 -5.86960614e-01
-2.12267980e-01 -5.44640064e-01 7.91363537e-01 1.21163309e+00
1.39393896e-01 4.59493548e-01 -4.66061592e-01 -3.82627130e-01
2.09498331e-01 4.05470543e-02 1.00332451e+00 -1.51399851e+00
-6.71555996e-01 -2.10674688e-01 2.70795763e-01 -1.79089344e+00
-4.70933825e-01 -7.34705567e-01 -5.11447847e-01 -1.76645303e+00
7.29831830e-02 -7.43398190e-01 -4.85562533e-01 1.18813500e-01
-3.12739462e-01 4.17492270e-01 8.85837823e-02 1.36451021e-01
-8.54487836e-01 4.85474318e-01 1.40953863e+00 -4.15122397e-02
-6.39932990e-01 5.92122912e-01 -1.00685585e+00 5.14339685e-01
8.40557992e-01 -1.21975645e-01 -1.00047803e+00 -1.12727284e-01
5.99406183e-01 -3.82184178e-01 -3.38115424e-01 -8.17241728e-01
1.32856444e-01 -3.88418794e-01 5.28836846e-01 -7.81776547e-01
3.32007825e-01 -7.82569468e-01 -2.70685285e-01 3.09228987e-01
-4.14255202e-01 -5.03574573e-02 -5.72663806e-02 9.03944969e-01
-5.79502434e-02 -1.74535271e-02 2.10581332e-01 -3.51407528e-01
-4.05341446e-01 4.99152154e-01 -2.50571311e-01 -1.64777320e-02
5.83581686e-01 -2.01431409e-01 -1.02015480e-01 -4.57635641e-01
-7.10585535e-01 2.94795036e-01 1.31216228e-01 3.84300619e-01
5.81270695e-01 -1.19819856e+00 -2.69778728e-01 6.71552002e-01
-4.83012088e-02 -2.88532436e-01 2.47740135e-01 9.74798977e-01
2.34841742e-02 3.57905805e-01 4.20852929e-01 -3.30439270e-01
-1.39569092e+00 5.29386282e-01 -3.57891619e-03 -5.76894462e-01
-1.49654493e-01 1.25341272e+00 -3.56532708e-02 -8.67156208e-01
4.47161227e-01 -3.70085627e-01 -6.02992058e-01 1.08902417e-01
8.07918012e-01 5.32591879e-01 3.13019007e-01 -8.10777843e-02
-2.09962621e-01 1.03639670e-01 -1.59740523e-01 1.05706438e-01
1.14297128e+00 -2.35743392e-02 -1.86600029e-01 5.68772972e-01
9.85187829e-01 3.11400235e-01 -7.53945112e-01 -3.13535929e-01
-2.18324244e-01 -1.02516532e+00 7.01370556e-03 -9.69660759e-01
-9.11033750e-01 7.31097281e-01 4.63074148e-01 6.92722976e-01
1.09830129e+00 -6.28978252e-01 9.04113650e-01 2.79526025e-01
6.87699556e-01 -1.24761641e+00 -1.79590896e-01 6.23601913e-01
6.65193439e-01 -1.20628774e+00 2.56236076e-01 -8.16529334e-01
-3.36389631e-01 8.54311049e-01 7.60282636e-01 -2.26910383e-01
1.23692846e+00 -6.49729967e-02 -1.45655677e-01 1.31640792e-01
-9.92433965e-01 -1.32692814e-01 4.89443630e-01 7.56314993e-01
1.11639619e+00 2.34914556e-01 -1.08118761e+00 9.84250784e-01
-8.61367807e-02 5.20653129e-01 1.29624689e-03 1.09428990e+00
-4.30843323e-01 -1.72349584e+00 4.28056233e-02 9.17887688e-01
-2.47212186e-01 -3.83885503e-01 -6.16184533e-01 5.62907398e-01
6.63563460e-02 1.23246598e+00 -3.83281589e-01 -9.12929535e-01
3.58910203e-01 -6.15398884e-02 3.98433894e-01 -7.86665559e-01
-1.12352788e+00 -1.89027816e-01 3.32751811e-01 -7.01125324e-01
-4.47087884e-01 -4.27826375e-01 -7.07764626e-01 -4.41902190e-01
-8.46796751e-01 7.11228549e-01 4.57788557e-01 5.14044821e-01
6.95483625e-01 3.67511272e-01 6.59957469e-01 -9.75399256e-01
-6.48774683e-01 -1.18521643e+00 -6.06641889e-01 2.40128756e-01
2.10645288e-01 -7.38368034e-01 -5.74781418e-01 -5.38019478e-01] | [10.048665046691895, 5.631604194641113] |
97306bcb-b924-40b6-a188-1a149a1f13fc | text-detection-and-recognition-in-the-wild-a | 2006.04305 | null | https://arxiv.org/abs/2006.04305v2 | https://arxiv.org/pdf/2006.04305v2.pdf | Text Detection and Recognition in the Wild: A Review | Detection and recognition of text in natural images are two main problems in the field of computer vision that have a wide variety of applications in analysis of sports videos, autonomous driving, industrial automation, to name a few. They face common challenging problems that are factors in how text is represented and affected by several environmental conditions. The current state-of-the-art scene text detection and/or recognition methods have exploited the witnessed advancement in deep learning architectures and reported a superior accuracy on benchmark datasets when tackling multi-resolution and multi-oriented text. However, there are still several remaining challenges affecting text in the wild images that cause existing methods to underperform due to there models are not able to generalize to unseen data and the insufficient labeled data. Thus, unlike previous surveys in this field, the objectives of this survey are as follows: first, offering the reader not only a review on the recent advancement in scene text detection and recognition, but also presenting the results of conducting extensive experiments using a unified evaluation framework that assesses pre-trained models of the selected methods on challenging cases, and applies the same evaluation criteria on these techniques. Second, identifying several existing challenges for detecting or recognizing text in the wild images, namely, in-plane-rotation, multi-oriented and multi-resolution text, perspective distortion, illumination reflection, partial occlusion, complex fonts, and special characters. Finally, the paper also presents insight into the potential research directions in this field to address some of the mentioned challenges that are still encountering scene text detection and recognition techniques. | ['Steven Wardell', 'Paul Fieguth', 'Mohamed A. Naiel', 'Zobeir Raisi', 'John Zelek'] | 2020-06-08 | null | null | null | null | ['scene-text-detection'] | ['computer-vision'] | [ 8.17171037e-01 -8.27048719e-01 3.15088220e-02 -1.50103554e-01
-2.62526691e-01 -4.68385547e-01 8.47048879e-01 -7.64660165e-02
-3.93953383e-01 2.26003334e-01 -2.93397047e-02 -3.37884426e-02
1.10185288e-01 -3.40934485e-01 -5.64905643e-01 -8.22968006e-01
5.40903687e-01 4.30385470e-01 3.40587974e-01 -1.33504272e-01
9.96290624e-01 5.76223433e-01 -2.17939091e+00 4.88734841e-01
4.19294685e-01 7.99131155e-01 2.89502442e-01 8.95081341e-01
-3.08492839e-01 8.04813087e-01 -8.93419206e-01 -3.22012633e-01
1.83332741e-01 -2.23188490e-01 -4.41700608e-01 8.52201283e-01
1.00103545e+00 -4.14402723e-01 -5.20976841e-01 7.94750869e-01
6.87546134e-01 1.29055336e-01 8.36030900e-01 -9.72371817e-01
-4.08749968e-01 1.28058717e-01 -9.45370018e-01 5.59264123e-01
5.51005900e-01 1.09987527e-01 3.96834403e-01 -1.21185780e+00
4.78176445e-01 1.17019737e+00 4.49927837e-01 3.66427779e-01
-5.68260193e-01 -3.74652535e-01 1.73931614e-01 3.56339604e-01
-1.06704259e+00 -2.69624501e-01 8.17406237e-01 -7.66871810e-01
1.00343573e+00 4.03690845e-01 2.89917588e-01 1.30757606e+00
6.07658923e-01 1.03232968e+00 1.12422025e+00 -7.69329071e-01
-1.48240238e-01 2.23857701e-01 2.94430822e-01 5.47427058e-01
3.67582738e-01 -1.18003897e-01 -9.24361169e-01 3.35220575e-01
4.53590035e-01 -5.86865880e-02 5.61325327e-02 -3.39303702e-01
-1.42242455e+00 6.21354640e-01 -3.42945814e-01 4.80059117e-01
9.16284472e-02 -3.30676973e-01 7.82908559e-01 1.52892709e-01
2.19448909e-01 1.90312982e-01 -1.17751315e-01 -1.66898102e-01
-9.78647530e-01 2.64201254e-01 5.53454816e-01 8.40558171e-01
2.74441540e-01 6.53397501e-01 -6.20243698e-02 9.14605021e-01
-5.38585410e-02 9.44262385e-01 9.46523845e-01 3.81692755e-03
8.54261816e-01 5.57040215e-01 -1.35440722e-01 -1.15292597e+00
-4.59603280e-01 -1.31012052e-01 -7.62024939e-01 3.26010585e-01
3.44094485e-01 -9.00675822e-03 -1.05428326e+00 4.73915160e-01
2.54677504e-01 -3.96374881e-01 -3.13016698e-02 8.30659747e-01
1.21398866e+00 5.49211144e-01 -4.79731798e-01 3.98585349e-02
1.46026492e+00 -1.20894623e+00 -9.58651125e-01 -7.10461617e-01
4.23838556e-01 -1.50285637e+00 9.73922014e-01 7.62654483e-01
-7.81694889e-01 -8.37513387e-01 -1.31121552e+00 -9.33041349e-02
-8.48710954e-01 7.79470026e-01 1.64603189e-01 9.24324095e-01
-4.54789579e-01 2.22068191e-01 -5.17321050e-01 -7.70383000e-01
1.08684734e-01 2.82934755e-01 -1.92950740e-01 -2.38580316e-01
-6.37628078e-01 9.75270033e-01 1.87000155e-01 3.76134515e-01
-5.09443223e-01 -1.75327547e-02 -6.60212636e-01 -3.71726841e-01
5.37603676e-01 -2.59155065e-01 8.06782544e-01 -1.11266100e+00
-1.43668270e+00 1.21613371e+00 -2.40992740e-01 -2.30765760e-01
8.53893161e-01 -5.14695287e-01 -6.96924388e-01 8.32615420e-02
8.73247832e-02 2.54143387e-01 1.36089790e+00 -9.10787880e-01
-6.89574301e-01 -4.73720521e-01 -4.57868427e-01 6.23838365e-01
-2.69364893e-01 3.61199021e-01 -3.85564387e-01 -8.11211884e-01
4.12583128e-02 -6.22019827e-01 2.95645446e-01 -1.61460996e-01
-4.49435323e-01 -2.96152562e-01 1.51918781e+00 -5.46849489e-01
8.60401869e-01 -2.10048985e+00 -1.33917071e-02 -3.60798955e-01
-9.14081261e-02 5.23150206e-01 4.29904684e-02 8.13361704e-01
4.24395241e-02 -5.41166924e-02 2.44762391e-01 -1.48214206e-01
-6.68677539e-02 -3.09462827e-02 -5.00161648e-01 7.95866668e-01
3.02612428e-02 4.45861965e-01 -1.64833680e-01 -6.54136002e-01
1.09174609e+00 3.42446387e-01 3.84468287e-01 -1.16582215e-01
1.94617677e-02 1.50553092e-01 -4.38751221e-01 9.09735322e-01
7.53745615e-01 2.52431542e-01 -2.64801323e-01 -2.95545548e-01
-3.62822890e-01 -2.06236422e-01 -1.57119441e+00 1.10497248e+00
9.56317708e-02 1.37999403e+00 9.72891226e-04 -1.24940419e+00
9.88629460e-01 1.89755689e-02 3.57162446e-01 -7.63193905e-01
5.87496638e-01 1.85408294e-01 -3.19498062e-01 -1.19888902e+00
9.98639524e-01 3.37968677e-01 2.95164257e-01 2.18438357e-01
-2.50898778e-01 -1.57342657e-01 5.81868351e-01 -1.12835474e-01
4.90480691e-01 1.59844905e-01 1.87731847e-01 -1.95906833e-02
9.52559173e-01 1.54297754e-01 -4.75515500e-02 8.86078954e-01
-2.54465222e-01 8.04971159e-01 1.95423812e-01 -9.08365548e-01
-1.05431700e+00 -4.80472773e-01 -3.72893572e-01 1.19245970e+00
4.35468853e-01 -9.58413556e-02 -4.97839063e-01 -4.45172846e-01
-1.75947934e-01 2.29346737e-01 -5.62155545e-01 2.39001274e-01
-7.01510131e-01 -9.03245687e-01 7.08468854e-01 5.37688911e-01
8.55225623e-01 -1.07707262e+00 -8.29474688e-01 -1.79668501e-01
-2.63793081e-01 -1.65372241e+00 -1.70422167e-01 1.17915198e-01
-9.31398273e-01 -1.15778923e+00 -8.97939861e-01 -1.02436173e+00
4.87306595e-01 6.89940155e-01 8.94774199e-01 6.68533668e-02
-8.13787222e-01 5.79333365e-01 -4.97088730e-01 -6.91152096e-01
-2.55935073e-01 -2.08721384e-01 -1.52039111e-01 2.14464054e-01
7.62838125e-01 3.71905655e-01 -2.59622991e-01 6.69801474e-01
-9.37896729e-01 6.94470108e-03 7.37132192e-01 9.13893461e-01
3.29582959e-01 3.41506422e-01 1.28700599e-01 -6.65473044e-01
3.70953947e-01 2.37465724e-01 -5.87061942e-01 1.07954852e-01
-3.17190796e-01 -4.47522849e-01 4.69312757e-01 -5.37615001e-01
-1.00433564e+00 2.89144129e-01 7.50629380e-02 -1.57283589e-01
-7.87685931e-01 1.84604526e-01 1.27885371e-01 -3.60629886e-01
7.45704770e-01 7.02767730e-01 -1.64090335e-01 -1.67708471e-01
-1.30801633e-01 1.08926189e+00 2.93043017e-01 -2.56435841e-01
8.03371131e-01 7.45288312e-01 1.91062227e-01 -1.64815581e+00
-6.73708200e-01 -1.01429701e+00 -8.71503294e-01 -4.46836025e-01
9.70842123e-01 -7.74950385e-01 -3.50204021e-01 1.22933245e+00
-7.72300184e-01 6.27871649e-03 1.05490528e-01 6.01033866e-01
-4.58045125e-01 1.11535561e+00 -2.75439173e-01 -9.40210283e-01
-2.88252562e-01 -1.28848708e+00 1.55709684e+00 2.32766524e-01
2.22669289e-01 -8.84904623e-01 -3.35501790e-01 1.01770961e+00
2.26848558e-01 2.00275317e-01 7.00198174e-01 -5.34143806e-01
-4.52014655e-01 -5.16841710e-01 -5.96833453e-02 3.64230931e-01
5.89890741e-02 3.61004680e-01 -1.18344295e+00 -3.18261266e-01
-3.53477849e-03 -3.59744459e-01 8.34702611e-01 4.06017512e-01
8.52057099e-01 3.26081306e-01 -2.86021769e-01 3.23238879e-01
1.32200706e+00 2.52889574e-01 4.93618518e-01 7.74125993e-01
7.82804012e-01 7.19222307e-01 9.26132321e-01 3.30525070e-01
-2.16116875e-01 6.67841315e-01 4.18895453e-01 -2.87501097e-01
-3.12858224e-01 3.44934464e-01 4.75793660e-01 4.07056361e-01
-9.65229142e-03 -5.64134061e-01 -8.28016996e-01 2.83887744e-01
-1.55649054e+00 -8.77355874e-01 -8.07129204e-01 2.06848145e+00
-1.38200864e-01 1.00143604e-01 2.64823824e-01 7.05149591e-01
1.13530087e+00 3.06245536e-01 -6.93034768e-01 -4.42752630e-01
-6.42327130e-01 -9.25633311e-02 6.99971437e-01 -8.18215683e-02
-1.55859685e+00 1.04355729e+00 6.61161327e+00 7.20870912e-01
-1.57840645e+00 -3.75661433e-01 2.69103348e-01 1.65979996e-01
7.78751612e-01 -4.93905693e-01 -9.98362184e-01 1.93964779e-01
3.07931662e-01 8.09632316e-02 1.96363792e-01 8.48153412e-01
1.42497391e-01 -3.89627874e-01 -8.23057711e-01 1.41125464e+00
1.06070936e+00 -1.02340972e+00 2.17567667e-01 -4.03666049e-02
8.64991784e-01 1.81289449e-01 3.24090809e-01 5.53782545e-02
-4.78185803e-01 -8.51320326e-01 6.24115705e-01 7.28974715e-02
5.86016476e-01 -4.30331469e-01 8.37396622e-01 2.82840669e-01
-1.15150869e+00 -1.56153440e-01 -6.68696940e-01 -3.13601606e-02
-2.63685197e-01 3.65545690e-01 -8.29115570e-01 6.66680098e-01
7.92399466e-01 9.69891548e-01 -9.86361802e-01 9.73578990e-01
1.20889723e-01 3.14853430e-01 -1.97776422e-01 -5.43116927e-01
1.86564192e-01 -8.71361122e-02 5.75471044e-01 1.48302913e+00
1.42463401e-01 -6.20651722e-01 3.58231753e-01 3.30123186e-01
4.00350332e-01 5.89887440e-01 -8.61022949e-01 -1.49885684e-01
-1.58458263e-01 1.22235250e+00 -1.16457295e+00 -3.25713277e-01
-6.64237797e-01 8.47783625e-01 -5.06859064e-01 4.84764695e-01
-6.01851463e-01 -4.42218721e-01 9.67543293e-03 1.50837258e-01
4.67150837e-01 -3.70889008e-01 -5.65350413e-01 -1.20484471e+00
3.68782640e-01 -1.22604680e+00 4.54459697e-01 -1.14209569e+00
-9.56014216e-01 3.81385297e-01 -1.10952638e-01 -1.37984979e+00
2.57032663e-02 -1.25972450e+00 -6.05688095e-01 5.53036094e-01
-1.47803760e+00 -1.51732039e+00 -6.34952843e-01 6.64569855e-01
1.59335303e+00 -7.20612586e-01 4.11063164e-01 1.86233327e-01
-7.28100479e-01 4.00964200e-01 4.63394672e-01 2.82869488e-01
9.63620663e-01 -9.84295845e-01 2.45608151e-01 1.11379528e+00
1.51107907e-01 9.88981966e-03 8.39061916e-01 -5.61632395e-01
-1.71999514e+00 -6.88290775e-01 5.51997006e-01 -6.04417920e-01
4.61263925e-01 -4.57331985e-01 -6.12659037e-01 5.88539183e-01
3.11388046e-01 -3.98531705e-01 2.57013083e-01 -1.83855370e-01
-6.52674586e-02 5.96433459e-03 -8.55336547e-01 5.19299388e-01
3.94555271e-01 -1.56237513e-01 -5.94817996e-01 6.71183884e-01
-3.04147720e-01 -6.51693761e-01 -4.15013283e-01 1.11094743e-01
7.04923749e-01 -1.13754666e+00 7.43509531e-01 -2.04269379e-01
3.38330925e-01 -3.19884419e-01 -1.09963916e-01 -5.94986737e-01
1.60056964e-01 -3.19708437e-01 3.17464769e-01 1.10119784e+00
-9.99160260e-02 -6.78434849e-01 9.24072981e-01 -2.81090979e-02
-1.85977757e-01 -2.61891097e-01 -8.50845039e-01 -6.26780689e-01
7.43270293e-02 -2.67025501e-01 2.78935898e-02 8.71419609e-01
-2.53768712e-01 3.86713892e-01 -7.41578817e-01 -1.02683809e-02
4.42495763e-01 1.33138478e-01 1.27783453e+00 -1.23420393e+00
1.55926540e-01 -4.45589960e-01 -9.45994914e-01 -1.05650699e+00
-1.12059392e-01 -2.65093297e-01 5.00807539e-02 -1.54729521e+00
6.51558414e-02 2.78812200e-01 3.18715394e-01 -1.43554017e-01
-1.81572326e-02 3.16208452e-01 2.20874488e-01 1.95163086e-01
-5.10460913e-01 1.48073211e-01 1.35779178e+00 -4.92624044e-01
2.67944396e-01 1.04673706e-01 -1.42646283e-01 8.33657324e-01
8.13304424e-01 -2.45969459e-01 -1.89733312e-01 -5.72696567e-01
2.65548557e-01 -3.76950055e-01 2.73738086e-01 -1.34434199e+00
3.57775390e-01 -4.98987101e-02 6.79258168e-01 -1.32751191e+00
4.58512038e-01 -8.31744611e-01 -4.33613479e-01 3.19132686e-01
-1.38876989e-01 2.84581184e-01 2.38158047e-01 5.76450169e-01
-3.17608595e-01 -6.73546791e-01 8.60825300e-01 -1.07948005e-01
-9.67456639e-01 -2.35557452e-01 -1.14471745e+00 -1.78444773e-01
1.04576528e+00 -9.70170319e-01 -4.20746446e-01 -2.49456689e-01
-2.93932378e-01 -1.50228605e-01 3.28025311e-01 9.22189236e-01
6.06771767e-01 -6.47861123e-01 -7.34832287e-01 4.67573702e-01
2.82339513e-01 -2.33039588e-01 4.59017098e-01 8.99314046e-01
-8.48905623e-01 6.90729022e-01 -4.22666192e-01 -1.10984373e+00
-1.96654522e+00 6.63191497e-01 3.69835109e-01 -1.57092378e-01
-7.14183509e-01 2.41679087e-01 1.38287127e-01 -1.36332244e-01
5.28248072e-01 -2.83967584e-01 -4.56041843e-01 -1.37823075e-01
5.44408858e-01 6.98232591e-01 5.34789622e-01 -9.95133162e-01
-1.30996659e-01 1.41606879e+00 -1.91667423e-01 3.01767647e-01
7.17548430e-01 -2.55501360e-01 2.82620698e-01 6.21105492e-01
5.93275309e-01 4.26058508e-02 -7.36162841e-01 -8.06013197e-02
-3.51811469e-01 -5.64005554e-01 6.74049705e-02 -7.76209772e-01
-9.06481624e-01 1.30098486e+00 1.01362216e+00 3.17305446e-01
9.32225943e-01 -5.51830232e-01 3.85374516e-01 8.53834808e-01
-2.11729825e-01 -1.75375307e+00 5.00934422e-01 6.23595715e-01
7.29039371e-01 -1.46325946e+00 4.39561009e-01 -6.33866429e-01
-7.95947731e-01 1.77712131e+00 6.32377386e-01 -8.51002336e-03
2.12688968e-01 3.91527206e-01 5.02804875e-01 -2.94454306e-01
-2.74237096e-01 -2.47839645e-01 1.73924938e-01 7.77137339e-01
6.29318714e-01 -3.84527475e-01 -3.02022725e-01 -4.32721496e-01
-2.10865289e-02 -1.50013641e-01 8.81023407e-01 1.12127280e+00
-5.13877571e-01 -8.67690742e-01 -1.00954151e+00 6.10199094e-01
-7.52083957e-01 2.29662642e-01 -8.33242714e-01 1.08678567e+00
-6.42555207e-02 1.21716213e+00 -4.12776694e-02 -3.08511347e-01
4.83395189e-01 3.34719494e-02 3.60498637e-01 -2.62461960e-01
-6.50496185e-01 3.76629353e-01 8.36973637e-02 1.14290081e-01
-6.54091299e-01 -8.50867510e-01 -7.18082547e-01 -2.14978546e-01
-7.82216668e-01 -3.92187297e-01 1.13959610e+00 1.00090826e+00
-4.89796214e-02 6.11956358e-01 5.03260911e-01 -1.02685344e+00
-4.13739592e-01 -1.13865292e+00 -7.64082611e-01 4.27367657e-01
2.51676381e-01 -7.14030623e-01 -5.04531264e-01 5.12079477e-01] | [11.963662147521973, 2.32021427154541] |
121ff187-682b-43d9-9b76-87964df2a928 | capturing-and-inferring-dense-full-body-human-1 | 2206.09553 | null | https://arxiv.org/abs/2206.09553v1 | https://arxiv.org/pdf/2206.09553v1.pdf | Capturing and Inferring Dense Full-Body Human-Scene Contact | Inferring human-scene contact (HSC) is the first step toward understanding how humans interact with their surroundings. While detecting 2D human-object interaction (HOI) and reconstructing 3D human pose and shape (HPS) have enjoyed significant progress, reasoning about 3D human-scene contact from a single image is still challenging. Existing HSC detection methods consider only a few types of predefined contact, often reduce body and scene to a small number of primitives, and even overlook image evidence. To predict human-scene contact from a single image, we address the limitations above from both data and algorithmic perspectives. We capture a new dataset called RICH for "Real scenes, Interaction, Contact and Humans." RICH contains multiview outdoor/indoor video sequences at 4K resolution, ground-truth 3D human bodies captured using markerless motion capture, 3D body scans, and high resolution 3D scene scans. A key feature of RICH is that it also contains accurate vertex-level contact labels on the body. Using RICH, we train a network that predicts dense body-scene contacts from a single RGB image. Our key insight is that regions in contact are always occluded so the network needs the ability to explore the whole image for evidence. We use a transformer to learn such non-local relationships and propose a new Body-Scene contact TRansfOrmer (BSTRO). Very few methods explore 3D contact; those that do focus on the feet only, detect foot contact as a post-processing step, or infer contact from body pose without looking at the scene. To our knowledge, BSTRO is the first method to directly estimate 3D body-scene contact from a single image. We demonstrate that BSTRO significantly outperforms the prior art. The code and dataset are available at https://rich.is.tue.mpg.de. | ['Michael J. Black', 'Daniel Scharstein', 'Senya Polikovsky', 'Tsvetelina Alexiadis', 'Matvey Safroshkin', 'Markus Höschle', 'Hongwei Yi', 'Chun-Hao P. Huang'] | 2022-06-20 | capturing-and-inferring-dense-full-body-human | http://openaccess.thecvf.com//content/CVPR2022/html/Huang_Capturing_and_Inferring_Dense_Full-Body_Human-Scene_Contact_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Huang_Capturing_and_Inferring_Dense_Full-Body_Human-Scene_Contact_CVPR_2022_paper.pdf | cvpr-2022-1 | ['markerless-motion-capture', 'monocular-3d-human-pose-estimation', 'human-scene-contact-detection'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 4.11977798e-01 1.15058720e-02 -1.23964660e-01 -2.56019145e-01
-6.19030774e-01 -3.51768911e-01 4.06213045e-01 -2.65973985e-01
-2.52614677e-01 2.60945559e-01 2.34166324e-01 1.86811343e-01
2.09178880e-01 -7.30504453e-01 -9.83622670e-01 -1.79581270e-01
3.51219587e-02 1.06375945e+00 5.18353820e-01 -3.13392788e-01
-3.39175403e-01 5.35558164e-01 -1.49278903e+00 7.51555562e-02
5.40315926e-01 9.67952847e-01 1.15559876e-01 8.41377020e-01
5.36076546e-01 3.43327522e-01 -1.53375894e-01 -2.90912449e-01
5.46659350e-01 -4.50612128e-01 -7.09998846e-01 4.29803133e-01
8.18824828e-01 -7.36460388e-01 -3.71382594e-01 5.64834774e-01
5.12974739e-01 4.90958095e-02 4.34371442e-01 -1.37602401e+00
1.51743874e-01 -9.25121829e-02 -7.06116378e-01 -1.95923433e-01
1.07705092e+00 3.04660350e-01 6.64578319e-01 -9.86262977e-01
8.81627560e-01 1.35039842e+00 9.77113962e-01 5.67428827e-01
-1.22031558e+00 -3.84014368e-01 1.72762237e-02 -4.52111475e-02
-1.35935998e+00 -4.60345596e-01 8.34001958e-01 -5.48029244e-01
8.92711163e-01 4.99244928e-01 1.37226880e+00 1.14176714e+00
1.25520095e-01 9.24012542e-01 1.02005541e+00 -3.81146580e-01
-8.44337195e-02 -3.37358057e-01 -2.75468621e-02 1.09629655e+00
2.70308495e-01 -6.23352453e-02 -8.16237569e-01 -1.41430601e-01
1.25909734e+00 1.64227694e-01 -3.67975742e-01 -8.10842872e-01
-1.41797733e+00 2.85240710e-01 4.43072617e-01 -2.45346650e-01
-2.35941410e-01 3.50678176e-01 -1.84686519e-02 -1.07731052e-01
4.18712974e-01 -6.06485270e-02 -4.21108872e-01 -1.32139623e-01
-7.49925673e-01 4.97682065e-01 7.54460812e-01 9.34648395e-01
7.54399121e-01 -4.07255918e-01 2.66586810e-01 4.86981720e-01
3.68755549e-01 8.03954899e-01 -2.50040203e-01 -1.21847272e+00
4.61865246e-01 6.36236072e-01 1.90066963e-01 -1.00749540e+00
-5.55459321e-01 -7.41229504e-02 -7.34988689e-01 2.68184960e-01
5.83816588e-01 -3.52465338e-03 -1.10629058e+00 1.30148542e+00
9.52017665e-01 5.29836118e-02 -4.22149152e-01 1.27509427e+00
9.05862927e-01 2.91864693e-01 -3.53789628e-01 3.46757919e-01
1.40950739e+00 -9.14353669e-01 -4.34037894e-01 -6.28965914e-01
3.14231575e-01 -4.01972622e-01 1.09746850e+00 3.82119298e-01
-1.22391880e+00 -4.53864694e-01 -8.99298608e-01 -4.60534215e-01
-1.17326841e-01 2.24405136e-02 7.15071917e-01 2.69294530e-01
-7.96931446e-01 4.72846210e-01 -1.09857345e+00 -7.10611820e-01
1.58478037e-01 5.53122103e-01 -7.41333246e-01 -3.45740050e-01
-1.00818086e+00 8.56911063e-01 -1.68195277e-01 3.71507138e-01
-7.76796460e-01 -5.05680621e-01 -1.24334288e+00 -4.72778231e-01
8.40127528e-01 -1.28316045e+00 1.04684162e+00 -6.67491198e-01
-1.23861551e+00 1.26207292e+00 -2.61510849e-01 -2.22518086e-01
8.71040821e-01 -8.23153436e-01 8.90827179e-02 3.63872498e-01
7.88183659e-02 7.37390161e-01 7.24402487e-01 -1.56041396e+00
-3.79875481e-01 -6.55954063e-01 8.31394866e-02 5.56690693e-01
5.13749003e-01 -9.77338478e-02 -9.50879335e-01 -3.46797496e-01
5.15920222e-01 -1.14417970e+00 -2.38814399e-01 5.07509947e-01
-7.04500437e-01 1.37706578e-01 6.78983510e-01 -7.89832413e-01
5.36562502e-01 -1.75052965e+00 2.71974325e-01 3.14278990e-01
2.39633620e-01 -1.54532179e-01 1.87888891e-01 2.24038601e-01
3.68803620e-01 -1.64963871e-01 -4.69872266e-01 -6.29117548e-01
3.68501395e-02 4.73520100e-01 1.12561055e-01 8.64599228e-01
-3.55848223e-02 1.09225452e+00 -8.50225687e-01 -7.65956104e-01
5.84199369e-01 7.06349194e-01 -6.42754853e-01 3.12544465e-01
-1.96235701e-01 7.01449156e-01 -2.64256269e-01 1.11007786e+00
5.32660186e-01 -2.15076983e-01 -3.58079150e-02 -5.00638068e-01
9.13391635e-02 -7.61843547e-02 -1.37071431e+00 2.01764679e+00
-2.84907501e-03 2.60058254e-01 4.77730870e-01 -7.03093410e-01
3.46158803e-01 2.62639433e-01 8.29825282e-01 -3.36779356e-01
1.32725045e-01 -2.93168742e-02 -3.34057540e-01 -3.46165568e-01
2.50511527e-01 -7.96977952e-02 -2.28800163e-01 6.31529987e-02
-1.11390159e-01 -4.11505580e-01 -2.46510550e-01 6.13286160e-02
1.05143070e+00 7.55099058e-01 2.34670103e-01 1.75179407e-01
-3.21561731e-02 2.89055049e-01 5.95382988e-01 6.90691531e-01
-1.22571506e-01 1.10010624e+00 6.59717917e-02 -4.39816475e-01
-8.47471476e-01 -1.53462481e+00 3.30035575e-02 7.92800188e-01
6.85986161e-01 -4.29578424e-01 -7.34448969e-01 -2.61334121e-01
8.88148919e-02 -1.22979917e-02 -6.46981478e-01 1.39786899e-01
-9.29441690e-01 -2.26104751e-01 3.53010148e-01 6.75307214e-01
6.73047364e-01 -7.45785117e-01 -1.10954845e+00 -5.48765548e-02
-6.37813687e-01 -1.43579328e+00 -5.22893488e-01 2.29209065e-02
-7.55166352e-01 -1.31424212e+00 -8.08937132e-01 -4.05864686e-01
6.97918475e-01 5.02311178e-02 1.33104944e+00 2.65189782e-02
-8.32886577e-01 9.85271990e-01 -1.44416556e-01 -1.53425694e-01
1.32244363e-01 -3.62563252e-01 3.03805649e-01 -2.17620581e-01
-1.27043463e-02 -6.46937907e-01 -7.56789804e-01 4.98728752e-01
-3.17133099e-01 5.10147452e-01 4.54729050e-01 6.50906444e-01
1.00385439e+00 -1.99380532e-01 -4.91469800e-01 -5.42845666e-01
-1.27673417e-01 -2.78698027e-01 -1.28654048e-01 7.82537553e-03
9.18600410e-02 -4.52789068e-01 -8.68707299e-02 -3.36730659e-01
-1.00358951e+00 7.05960572e-01 -1.84235945e-01 -6.67786479e-01
-6.25297487e-01 6.77433237e-02 -3.97093207e-01 1.93509996e-01
5.38175106e-01 -3.86943901e-03 -3.44473459e-02 -5.97686708e-01
2.16295063e-01 8.74554589e-02 1.05286682e+00 -7.88472354e-01
7.93679357e-01 9.64547753e-01 1.58770770e-01 -9.74007070e-01
-8.36265743e-01 -7.06158698e-01 -1.29081619e+00 -5.28366089e-01
1.13249958e+00 -1.02777326e+00 -7.65908659e-01 4.97642845e-01
-1.06240642e+00 -7.11771488e-01 -3.21132392e-01 4.89582270e-01
-7.87762046e-01 4.08259481e-01 -7.32723355e-01 -8.58649850e-01
-2.32656196e-01 -8.06095243e-01 1.80184722e+00 -2.59880990e-01
-6.08324289e-01 -6.23212993e-01 -1.07209034e-01 7.76717424e-01
-3.36915582e-01 9.95568991e-01 3.23313385e-01 1.41567931e-01
-6.48857713e-01 -2.04965144e-01 1.76866189e-01 -1.08379409e-01
1.08210161e-01 -2.67445385e-01 -8.40542376e-01 -6.24091551e-02
-1.85291454e-01 -3.56892824e-01 6.05496883e-01 6.12528682e-01
7.67182529e-01 -1.23824634e-01 -5.28093219e-01 8.06098104e-01
1.04966080e+00 -3.70492905e-01 4.90270644e-01 2.47240484e-01
1.29750562e+00 7.59162962e-01 8.39707553e-01 4.76411343e-01
5.92236638e-01 1.01260602e+00 5.23321807e-01 -3.87489080e-01
-3.02102953e-01 -5.51862895e-01 3.76973540e-01 5.95661461e-01
-6.07746422e-01 -8.27477407e-03 -1.26175594e+00 4.68967736e-01
-1.81887960e+00 -8.18574011e-01 -4.65112180e-01 2.14875031e+00
7.11711347e-01 1.83243543e-01 3.57671648e-01 2.17268169e-01
4.13596779e-01 -1.48349851e-01 -6.99460447e-01 2.65808642e-01
3.67955342e-02 7.25305676e-02 4.48745608e-01 6.80502534e-01
-1.08693159e+00 9.50486898e-01 5.59797716e+00 3.03907692e-01
-7.20493674e-01 7.05381706e-02 2.00948238e-01 -2.44143471e-01
8.37293337e-04 1.82315867e-04 -7.01588929e-01 1.24167232e-02
2.08345205e-01 5.18236160e-01 1.08152963e-01 8.10988069e-01
4.63837534e-02 -6.42329872e-01 -1.39203477e+00 1.37596011e+00
1.72971785e-01 -8.09703946e-01 -2.19338566e-01 1.61253616e-01
3.08154732e-01 1.09264590e-01 -2.94062674e-01 -7.69382119e-02
2.04013571e-01 -1.13404214e+00 9.91641819e-01 8.20827365e-01
9.86852646e-01 -3.47711146e-01 4.24265504e-01 7.80269980e-01
-1.59890175e+00 3.70993912e-01 -3.55120935e-02 -2.92329609e-01
5.17771006e-01 5.93352318e-01 -5.27337849e-01 4.42127913e-01
1.04185152e+00 8.11718345e-01 -5.09185135e-01 6.98515773e-01
-1.26149982e-01 3.56469631e-01 -9.00760293e-01 5.72652340e-01
-2.39262775e-01 -2.17454165e-01 6.39213681e-01 9.76648152e-01
1.42556965e-01 4.61418658e-01 6.70233250e-01 7.63435781e-01
2.70044833e-01 -2.68886626e-01 -7.63816237e-01 4.36254084e-01
3.81584093e-02 1.02248704e+00 -8.98778915e-01 -3.97048622e-01
-1.56919196e-01 1.44335330e+00 -3.63108739e-02 2.13626146e-01
-8.10629427e-01 1.49312422e-01 6.77978337e-01 7.94916153e-01
-1.05908252e-01 -5.71809888e-01 -1.89149022e-01 -1.38698256e+00
3.51069570e-01 -6.77794993e-01 1.76212534e-01 -9.31349456e-01
-1.01997161e+00 8.01399127e-02 4.77406889e-01 -1.17460561e+00
-2.41961852e-01 -4.88400042e-01 -1.78762376e-01 5.31052232e-01
-7.97326386e-01 -1.50966609e+00 -7.46257901e-01 1.03482783e+00
6.93791211e-01 7.94551373e-01 6.67588830e-01 1.57184228e-01
-2.36030236e-01 3.02020639e-01 -6.01983845e-01 3.12164545e-01
5.72901905e-01 -1.21494389e+00 5.10956526e-01 5.60590923e-01
1.08833782e-01 4.78246778e-01 6.23509526e-01 -1.08855414e+00
-1.92187190e+00 -7.30649829e-01 5.62859118e-01 -1.12521636e+00
-6.88364450e-03 -7.64937043e-01 -4.73129362e-01 8.97038281e-01
-3.33262622e-01 2.91771263e-01 3.71545821e-01 1.46533670e-02
-1.62824560e-02 9.15611833e-02 -1.17062020e+00 7.32882261e-01
1.78024852e+00 -5.53255737e-01 -6.02263570e-01 2.66358435e-01
4.69682306e-01 -1.15626097e+00 -9.46837723e-01 7.38626540e-01
9.57094669e-01 -1.05641508e+00 1.49356699e+00 -3.06925662e-02
2.18893826e-01 -3.52887481e-01 -3.50484997e-01 -8.61045420e-01
9.47729573e-02 -3.33804131e-01 -4.83629137e-01 6.36756241e-01
-2.15781793e-01 -2.76260406e-01 1.11262298e+00 9.18094456e-01
-1.45771965e-01 -8.82935703e-01 -8.11648965e-01 -5.59894383e-01
-4.13068593e-01 -6.99380875e-01 2.22353190e-01 8.11388016e-01
-2.91694075e-01 1.65554941e-01 -7.01251924e-01 3.68271738e-01
1.07447505e+00 9.60232019e-02 1.37941027e+00 -1.37852299e+00
-3.93960446e-01 1.42396420e-01 -5.17806411e-01 -1.36690509e+00
-1.20198026e-01 -4.12879407e-01 4.13743734e-01 -1.87851167e+00
2.16462001e-01 -3.91069353e-01 3.76288176e-01 6.06666327e-01
4.97838594e-02 6.01893783e-01 1.21766083e-01 2.97144383e-01
-4.89512205e-01 3.19857210e-01 1.36018467e+00 9.04953182e-02
-3.26330066e-02 -1.90172214e-02 5.80349006e-02 1.30183744e+00
5.29000998e-01 -2.23458856e-01 -2.66349018e-01 -3.65059882e-01
8.96246955e-02 2.69880712e-01 1.12282944e+00 -1.34464085e+00
1.75444290e-01 -2.00995043e-01 9.22184467e-01 -1.04281712e+00
1.09683013e+00 -8.13423574e-01 6.56919420e-01 5.10342002e-01
1.19704552e-01 -1.58211425e-01 -1.46260541e-02 5.80340505e-01
1.84829488e-01 4.02956635e-01 4.45417136e-01 -6.16338193e-01
-7.98617065e-01 5.54772675e-01 -1.19802184e-01 4.82191816e-02
8.14119339e-01 -8.32438469e-01 2.73031622e-01 -3.75256598e-01
-1.04115677e+00 2.81842172e-01 9.55153286e-01 3.48966002e-01
8.96007657e-01 -1.16904247e+00 -4.66218323e-01 2.48152778e-01
-9.98481810e-02 6.80908203e-01 2.21446514e-01 9.48286831e-01
-5.76281786e-01 9.63341668e-02 -1.39315441e-01 -1.16823542e+00
-1.61891258e+00 5.92838936e-02 5.28449059e-01 8.06807801e-02
-1.25236809e+00 8.03912997e-01 4.34275806e-01 -7.86416829e-01
2.89884657e-01 -5.08315802e-01 3.28216463e-01 -4.07286197e-01
9.52159539e-02 4.41366494e-01 -2.53240883e-01 -1.15870357e+00
-3.67867708e-01 1.26186657e+00 5.68917453e-01 -3.19173694e-01
1.03112698e+00 -2.94369608e-01 -9.09886695e-03 6.69379294e-01
1.03087008e+00 3.68047990e-02 -1.63522339e+00 -1.61704704e-01
-3.49300086e-01 -6.94018841e-01 -3.38529855e-01 -5.35978973e-01
-6.57051027e-01 8.99315119e-01 5.70499599e-01 -4.24812645e-01
9.84115601e-01 4.30255473e-01 9.42802846e-01 3.81015241e-01
8.91521513e-01 -1.11757231e+00 1.81963176e-01 4.63648111e-01
9.48170185e-01 -1.30201149e+00 3.63834798e-01 -9.33642149e-01
-4.77044642e-01 7.25252211e-01 8.78010809e-01 -5.71196787e-02
5.58389127e-01 4.02969539e-01 -1.90182015e-01 -5.74724257e-01
-2.19316825e-01 -3.68448675e-01 4.98227477e-01 7.46479571e-01
1.42288700e-01 1.15967236e-01 4.01640505e-01 2.69067377e-01
-4.16805893e-01 4.74137552e-02 1.34784847e-01 1.20600069e+00
-2.47794494e-01 -8.41567814e-01 -7.90749192e-01 4.53708231e-01
-3.05165768e-01 2.16190845e-01 -6.17352664e-01 8.89705539e-01
4.84055072e-01 7.14238882e-01 6.68849349e-02 -4.39853877e-01
6.12055361e-01 -1.06987901e-01 9.45846736e-01 -7.31340647e-01
-2.01424628e-01 9.94948521e-02 2.52713978e-01 -1.13511491e+00
-6.39468193e-01 -6.06145859e-01 -1.45962489e+00 -3.03810239e-01
-2.32183471e-01 -3.21670502e-01 4.92413193e-01 1.00188851e+00
1.30792126e-01 2.25124866e-01 -2.54599154e-02 -1.72209096e+00
1.04351833e-01 -6.19459391e-01 -5.93484521e-01 5.24048686e-01
5.17751992e-01 -9.55083847e-01 -9.48357359e-02 3.47931653e-01] | [7.032830715179443, -1.0670313835144043] |
3e0a2744-72f1-457a-afc2-40000137e6d4 | a-prospective-observational-study-to | 2106.02118 | null | https://arxiv.org/abs/2106.02118v2 | https://arxiv.org/pdf/2106.02118v2.pdf | A Prospective Observational Study to Investigate Performance of a Chest X-ray Artificial Intelligence Diagnostic Support Tool Across 12 U.S. Hospitals | Importance: An artificial intelligence (AI)-based model to predict COVID-19 likelihood from chest x-ray (CXR) findings can serve as an important adjunct to accelerate immediate clinical decision making and improve clinical decision making. Despite significant efforts, many limitations and biases exist in previously developed AI diagnostic models for COVID-19. Utilizing a large set of local and international CXR images, we developed an AI model with high performance on temporal and external validation. Conclusions and Relevance: AI-based diagnostic tools may serve as an adjunct, but not replacement, for clinical decision support of COVID-19 diagnosis, which largely hinges on exposure history, signs, and symptoms. While AI-based tools have not yet reached full diagnostic potential in COVID-19, they may still offer valuable information to clinicians taken into consideration along with clinical signs and symptoms. | ['Christopher Tignanelli', 'Erich Kummerfeld', 'Judy Wawira Gichoya', 'Scott D. Steenburg', 'Tadashi Allen', 'Kun Huang', 'John L. Burns', 'Sean Switzer', 'Daniel Boley', 'Eric Murray', 'Nicholas Ingraham', 'Genevieve B. Melton', 'Zach Zaiman', 'Dyah Adila', 'Taihui Li', 'Le Peng', 'Ju Sun'] | 2021-06-03 | null | null | null | null | ['covid-19-detection'] | ['medical'] | [ 1.85673982e-01 -2.54654288e-01 -4.60057497e-01 -3.94609213e-01
-8.10882390e-01 -5.99271715e-01 2.95226663e-01 7.95013428e-01
-4.32910532e-01 5.90339720e-01 3.75931352e-01 -1.18139672e+00
-8.55882406e-01 -4.75365162e-01 2.45533651e-03 -3.60984236e-01
-1.39359146e-01 1.34736967e+00 -1.70241281e-01 4.28645343e-01
-8.84308890e-02 9.39066768e-01 -6.54301941e-01 2.24316001e-01
8.85395646e-01 8.68354857e-01 2.82908112e-01 1.03933859e+00
2.58725137e-01 1.48014247e+00 -4.77561921e-01 -1.35355994e-01
1.01870112e-01 -7.59931087e-01 -8.34862351e-01 -1.75326765e-01
-1.31078765e-01 -8.13397229e-01 -2.61607498e-01 1.55952536e-02
5.76678336e-01 -3.03159356e-01 1.14295113e+00 -6.78932130e-01
-3.94567639e-01 2.97304183e-01 -2.64678895e-01 7.73926318e-01
3.27699870e-01 7.03598082e-01 6.26941800e-01 -5.92388093e-01
7.58413017e-01 6.36093438e-01 8.95655751e-01 4.93320644e-01
-8.73790622e-01 -4.63739425e-01 -1.96207523e-01 3.05364490e-01
-1.00478446e+00 4.56764489e-01 1.98574394e-01 -7.81383216e-01
9.01374400e-01 6.42656267e-01 1.27572751e+00 6.20465934e-01
4.24975514e-01 4.16935265e-01 1.08305740e+00 -2.94787258e-01
-3.38771120e-02 -1.03475474e-01 3.47107977e-01 5.40712476e-01
3.88717473e-01 -2.27447785e-02 -7.81963170e-02 -7.73892522e-01
7.75210559e-01 5.83526552e-01 -1.42041966e-01 3.04846764e-01
-1.25651956e+00 1.03807545e+00 4.62502539e-01 3.73701960e-01
-9.72739398e-01 -8.33540335e-02 4.69858021e-01 4.63056229e-02
1.27201468e-01 7.15439856e-01 -2.98336953e-01 -4.11261976e-01
-8.25124621e-01 1.31457925e-01 4.05812353e-01 -5.82979992e-04
-2.65451759e-01 -1.27273232e-01 -1.46497905e-01 7.87924290e-01
1.38831288e-01 9.91446912e-01 4.69026864e-01 -8.34942162e-01
2.12486327e-01 7.51844287e-01 1.43592190e-02 -8.06154966e-01
-7.25498736e-01 -2.47742057e-01 -6.70305192e-01 -3.01975429e-01
1.16662718e-01 -2.54352421e-01 -1.04865348e+00 1.14274490e+00
-2.02282537e-02 -1.65155120e-02 -1.03958905e-01 9.30883288e-01
3.20701152e-01 3.18949789e-01 3.39753658e-01 -3.17714721e-01
1.43660831e+00 -2.73210973e-01 -3.08402687e-01 -1.45517334e-01
8.35759938e-01 -5.91223955e-01 5.20900965e-01 5.85525692e-01
-1.12138653e+00 -5.77523038e-02 -7.97157884e-01 4.98329937e-01
-1.63711831e-01 1.01522468e-01 7.06890345e-01 6.58841074e-01
-7.91714549e-01 5.35883248e-01 -1.13465476e+00 -6.14495039e-01
3.40057760e-01 2.87950099e-01 -2.70242929e-01 -2.89631575e-01
-1.03533447e+00 1.37386596e+00 2.38540336e-01 6.85681850e-02
-6.59602106e-01 -6.22841954e-01 -2.96693504e-01 -2.55849123e-01
3.66571814e-01 -1.25376940e+00 1.24309337e+00 -7.15152696e-02
-6.71889365e-01 8.22483838e-01 -2.76950806e-01 -7.21700430e-01
3.89010906e-01 -6.98251426e-02 -4.09731239e-01 7.81098187e-01
-5.11098467e-02 2.00481370e-01 4.25624460e-01 -7.88735569e-01
-6.90003037e-01 -5.89918315e-01 -4.52407569e-01 3.23049158e-01
-8.54046419e-02 4.61181551e-01 3.22919041e-02 -8.09049249e-01
1.25257313e-01 -9.06096339e-01 -5.79940379e-01 -3.24383974e-02
1.77006334e-01 -1.29348651e-01 6.56994641e-01 -9.88166392e-01
1.37728333e+00 -1.61005461e+00 -3.97071749e-01 3.39854121e-01
6.70335412e-01 7.73596585e-01 6.27080262e-01 4.23052073e-01
-5.19589297e-02 4.45895612e-01 -2.83528149e-01 3.06502432e-01
-5.99176228e-01 2.68503934e-01 7.46723190e-02 4.62132037e-01
6.62207305e-01 1.12424898e+00 -8.44769537e-01 -7.81586111e-01
9.76329565e-01 3.82639259e-01 -3.87459069e-01 4.36121523e-01
9.88393207e-04 5.44194758e-01 -7.12741256e-01 6.54522836e-01
1.67668313e-01 -9.85090256e-01 3.23429942e-01 5.08707047e-01
1.85025975e-01 2.15375960e-01 -6.52154505e-01 8.31864655e-01
-3.40748578e-01 4.24510211e-01 -1.79609209e-01 -6.46162033e-01
6.43998742e-01 6.44120395e-01 8.23842108e-01 -3.95695925e-01
3.07363003e-01 2.67819077e-01 6.54292285e-01 -5.29609859e-01
-1.77249148e-01 -5.02103567e-01 3.52737993e-01 9.99988437e-01
-6.04611039e-01 -4.62667972e-01 -2.86034960e-02 1.76683098e-01
1.48137271e+00 -5.45334399e-01 6.50371671e-01 1.36013642e-01
4.36469108e-01 6.40764534e-01 4.52301413e-01 1.07047951e+00
-1.31110758e-01 8.21557164e-01 1.95233133e-02 -4.57830667e-01
-9.39006150e-01 -1.25451827e+00 -8.53136063e-01 7.96776935e-02
-4.22979683e-01 -1.66580006e-01 -5.86391985e-02 -3.94713223e-01
9.77176577e-02 6.73790932e-01 -5.64631283e-01 -8.37407634e-02
-6.53245449e-01 -9.33555186e-01 5.12766004e-01 9.69687819e-01
-1.52476788e-01 -8.37724507e-01 -1.09340954e+00 7.66351104e-01
-9.63118449e-02 -6.53065979e-01 2.09772095e-01 2.08159879e-01
-8.83804262e-01 -1.42124224e+00 -1.20840383e+00 -3.38279814e-01
4.30119395e-01 9.94417891e-02 9.10516858e-01 5.93585491e-01
-8.96607697e-01 8.28312397e-01 -3.25270206e-01 -8.19552600e-01
-6.83929861e-01 -1.78550243e-01 3.77355181e-02 -5.54580510e-01
3.27370703e-01 -2.88196445e-01 -1.12094724e+00 3.20697248e-01
-7.78953910e-01 6.11728104e-03 6.86455965e-01 8.03013325e-01
6.16514742e-01 -4.53528911e-01 5.41604161e-01 -8.91283035e-01
8.01352739e-01 -8.98158610e-01 8.73321146e-02 3.32051098e-01
-1.21729386e+00 -5.00702679e-01 6.53217494e-01 -1.45733863e-01
-1.03019464e+00 -5.52233398e-01 -8.36560950e-02 -4.77748752e-01
-9.70378965e-02 7.35988379e-01 1.02415085e+00 2.67258883e-01
6.55445218e-01 -2.38041696e-03 1.52641982e-01 -4.66939241e-01
-2.99629867e-01 9.87611294e-01 4.89925951e-01 -4.34073269e-01
4.84281331e-01 6.01811469e-01 2.39205196e-01 -5.51074862e-01
-5.81733763e-01 -8.85681570e-01 -5.31145096e-01 -5.74819520e-02
1.15429521e+00 -6.81472182e-01 -5.46542346e-01 1.13263823e-01
-6.99157059e-01 3.39318961e-02 -3.53238255e-01 9.39026296e-01
-1.65249228e-01 1.28164187e-01 -6.30223870e-01 -7.30412543e-01
-6.63286030e-01 -1.24761510e+00 5.76166093e-01 8.43492597e-02
-8.75530124e-01 -1.21196377e+00 2.88886011e-01 4.07194346e-01
5.43236911e-01 3.34044665e-01 1.38041139e+00 -9.66886699e-01
-4.61076975e-01 -7.32645631e-01 -3.60021114e-01 1.94848314e-01
4.69908625e-01 2.62957901e-01 -3.27480227e-01 7.91240707e-02
1.34111673e-01 -2.87620455e-01 4.96351361e-01 4.83268440e-01
9.28649724e-01 1.31000308e-02 -4.05234367e-01 4.56578821e-01
1.29440057e+00 1.31159675e+00 3.58673483e-02 2.01759070e-01
4.25705731e-01 6.13614544e-02 5.71590781e-01 5.60626924e-01
4.02734131e-01 2.31175438e-01 -4.74813059e-02 -3.58956367e-01
2.10138559e-02 -1.57848354e-02 -6.01172805e-01 6.18031442e-01
-3.61291289e-01 -1.89250052e-01 -1.84869206e+00 5.60974360e-01
-1.32245493e+00 -6.89509749e-01 -3.08546871e-01 1.86636841e+00
5.62727869e-01 1.14738226e-01 1.70594528e-01 5.98672079e-03
3.10713708e-01 -7.45137185e-02 -7.34301984e-01 -6.32651567e-01
-9.66047496e-02 4.22147900e-01 1.87429622e-01 9.22197700e-02
-4.08209056e-01 2.05501080e-01 7.99762392e+00 6.81885704e-02
-1.51373816e+00 1.58466443e-01 8.50327969e-01 1.07694920e-02
-3.37383121e-01 1.93230752e-02 -2.97721088e-01 3.25184524e-01
8.09179366e-01 -3.81966174e-01 1.65037796e-01 6.96971834e-01
2.86400408e-01 -3.77081603e-01 -8.65626514e-01 6.29589021e-01
3.43500264e-02 -1.67223907e+00 -1.17147826e-01 3.90673652e-02
5.09189129e-01 2.65337527e-01 1.02178082e-01 -2.08660111e-01
4.20691252e-01 -9.88013983e-01 7.89605826e-02 3.38500261e-01
1.23395157e+00 -3.92501205e-01 1.08129954e+00 -3.71605195e-02
-8.41847539e-01 -2.60883532e-02 -4.22333181e-02 -1.04268417e-01
3.94040823e-01 1.58349648e-01 -1.82726073e+00 4.19278026e-01
5.55201411e-01 3.35911006e-01 -4.46796805e-01 7.99029827e-01
-1.10331044e-01 1.08697450e+00 -4.89452749e-01 -2.64825281e-02
4.26930219e-01 4.55869734e-02 3.74096692e-01 1.01036108e+00
-2.12749783e-02 1.10091496e+00 2.25951336e-03 5.48779130e-01
3.44110221e-01 2.98143417e-01 -6.82204127e-01 -4.12769705e-01
3.67818296e-01 8.29928517e-01 -9.62173820e-01 -7.35591412e-01
-4.49631661e-01 3.79105657e-01 -8.00358132e-03 2.00379476e-01
-6.06165409e-01 2.68098451e-02 2.64075667e-01 5.27204216e-01
-1.82847545e-01 1.59116983e-01 -7.40734935e-01 -7.59316802e-01
-2.91311532e-01 -1.00825012e+00 8.42134953e-01 -8.56611252e-01
-8.98581684e-01 6.03364468e-01 1.11202665e-01 -9.14498329e-01
-1.08943248e+00 -4.27336931e-01 -8.16524208e-01 8.11435044e-01
-9.64900494e-01 -1.01869583e+00 -1.97229624e-01 1.48295477e-01
1.77308887e-01 -1.25041440e-01 1.04815054e+00 -2.14237258e-01
-3.10298622e-01 1.23827159e-01 3.54340762e-01 1.39560103e-01
4.74436969e-01 -1.17838407e+00 4.99754064e-02 5.60488999e-01
-4.47000533e-01 8.95855546e-01 3.59934598e-01 -9.17819142e-01
-1.02356660e+00 -6.64019525e-01 8.04275751e-01 -7.62005150e-01
5.58525145e-01 6.75258577e-01 -8.45149577e-01 7.50934839e-01
-2.23208487e-01 -3.13574135e-01 1.04639888e+00 -7.10853040e-02
1.40782982e-01 3.30589682e-01 -1.20249820e+00 3.96317393e-01
6.89284921e-01 -5.98706067e-01 -9.96393204e-01 3.94382119e-01
4.40627426e-01 -1.62746325e-01 -1.42070079e+00 7.34989107e-01
5.23681402e-01 -7.07046032e-01 1.10560155e+00 -8.51250470e-01
3.82241338e-01 -1.49368623e-03 2.26031154e-01 -9.37863588e-01
-4.22903091e-01 -3.34725648e-01 4.26584691e-01 3.75925601e-01
3.95050228e-01 -8.43352199e-01 6.70931220e-01 1.10885656e+00
2.33013615e-01 -1.00782776e+00 -6.89170241e-01 -3.09530914e-01
-2.13757530e-03 -7.37796307e-01 4.56500262e-01 9.28410172e-01
2.20595393e-02 -1.35156631e-01 -5.58190085e-02 -5.73775508e-02
2.39474267e-01 2.22793400e-01 2.60098130e-01 -1.15596879e+00
-3.40987891e-01 -4.25579369e-01 -4.02895391e-01 -1.26840815e-01
-5.06344438e-01 -7.97510624e-01 -7.21803606e-01 -1.97960305e+00
6.41652942e-01 -8.16683888e-01 -5.96844971e-01 4.37155902e-01
-5.65460980e-01 1.48149222e-01 3.13749522e-01 5.06471932e-01
1.97894946e-02 -1.01236559e-01 9.09398913e-01 1.93124577e-01
-2.15644330e-01 9.38591808e-02 -6.49717808e-01 8.88970017e-01
7.08523810e-01 -7.93985724e-01 -4.59282011e-01 -2.93885231e-01
-6.54065832e-02 8.91450405e-01 3.70292753e-01 -8.09643030e-01
5.98471090e-02 -6.56729698e-01 5.73849678e-01 -9.33423102e-01
2.02337831e-01 -7.72855818e-01 4.54187334e-01 1.24157488e+00
-1.26387298e-01 6.17425442e-01 2.32973620e-01 2.47834176e-01
-2.01813132e-01 -1.26726031e-01 6.03574991e-01 -2.96657115e-01
-4.17540789e-01 2.19030663e-01 -9.79492545e-01 3.52468193e-01
1.13658285e+00 -2.39911228e-01 -3.30383480e-01 -4.81279790e-01
-7.66122580e-01 7.47921243e-02 4.99585032e-01 1.79822519e-01
7.70392299e-01 -8.88784826e-01 -8.77654433e-01 -3.87646593e-02
1.63344875e-01 1.50525346e-01 3.06806654e-01 1.17027998e+00
-1.34920263e+00 9.62070346e-01 -1.22355320e-01 -8.43571603e-01
-1.38534367e+00 6.19910657e-01 2.30799839e-01 -4.31133479e-01
-6.86839759e-01 6.80025458e-01 6.59158155e-02 2.14457497e-01
1.17960393e-01 -2.13161409e-01 1.88320994e-01 -1.69481665e-01
6.20213926e-01 2.78237432e-01 -2.08039675e-02 -3.96238029e-01
-5.65492034e-01 5.44666052e-01 -3.78926516e-01 -1.91795081e-01
1.32963145e+00 -7.84324929e-02 1.61882967e-01 3.41191173e-01
8.08130801e-01 -3.68350953e-01 -3.98787498e-01 -2.20795736e-01
-2.08233401e-01 -5.21238923e-01 -2.15909397e-03 -1.17210793e+00
-5.73083639e-01 9.71954882e-01 5.70435226e-01 -1.74305856e-01
1.12054133e+00 2.37924337e-01 6.04819715e-01 3.17508250e-01
-1.26809543e-02 -7.64051259e-01 1.74571725e-03 -1.37292176e-01
5.60540736e-01 -1.26452124e+00 2.44392782e-01 2.30689123e-02
-9.55189586e-01 8.08435380e-01 3.16580266e-01 1.06654741e-01
6.40463710e-01 1.00619569e-01 4.24734175e-01 -3.80754799e-01
-1.00219202e+00 1.24094065e-03 3.74347717e-02 7.17978537e-01
4.56218690e-01 4.74571228e-01 -5.19731045e-01 4.53879535e-01
2.19617993e-01 2.42230475e-01 2.49667749e-01 1.07408786e+00
-2.34247357e-01 -1.07030094e+00 -6.12314045e-01 1.22684538e+00
-9.69911337e-01 -2.81373203e-01 -5.76345503e-01 9.80158329e-01
2.12773114e-01 6.74909472e-01 1.38560878e-02 -2.49648914e-01
1.45166785e-01 1.61275491e-01 2.70567894e-01 -4.50366348e-01
-8.08470249e-01 2.80402694e-02 1.54597864e-01 -2.28730552e-02
-2.08036646e-01 -8.62354338e-01 -1.32393301e+00 -1.67715356e-01
-2.52684176e-01 7.48239905e-02 4.94093657e-01 8.73355567e-01
2.79833078e-01 5.82354367e-01 4.99801278e-01 1.42497241e-01
-6.12659752e-01 -7.18995214e-01 -3.18982512e-01 3.03869635e-01
3.19129318e-01 -4.27034497e-01 -2.72414595e-01 -2.03222647e-01] | [15.308751106262207, -1.9004192352294922] |
6ba35688-fb7b-4c86-9e51-9b921ba21858 | drinet-efficient-voxel-as-point-point-cloud | 2111.08318 | null | https://arxiv.org/abs/2111.08318v1 | https://arxiv.org/pdf/2111.08318v1.pdf | DRINet++: Efficient Voxel-as-point Point Cloud Segmentation | Recently, many approaches have been proposed through single or multiple representations to improve the performance of point cloud semantic segmentation. However, these works do not maintain a good balance among performance, efficiency, and memory consumption. To address these issues, we propose DRINet++ that extends DRINet by enhancing the sparsity and geometric properties of a point cloud with a voxel-as-point principle. To improve efficiency and performance, DRINet++ mainly consists of two modules: Sparse Feature Encoder and Sparse Geometry Feature Enhancement. The Sparse Feature Encoder extracts the local context information for each point, and the Sparse Geometry Feature Enhancement enhances the geometric properties of a sparse point cloud via multi-scale sparse projection and attentive multi-scale fusion. In addition, we propose deep sparse supervision in the training phase to help convergence and alleviate the memory consumption problem. Our DRINet++ achieves state-of-the-art outdoor point cloud segmentation on both SemanticKITTI and Nuscenes datasets while running significantly faster and consuming less memory. | ['Qifeng Chen', 'Tongyi Cao', 'Shuangjie Xu', 'Rui Wan', 'Maosheng Ye'] | 2021-11-16 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [-3.65188569e-02 -2.85495073e-01 -1.96206048e-01 -5.40357649e-01
-6.26986325e-01 -1.36995539e-01 1.82368100e-01 1.16634369e-01
-8.62320289e-02 2.63482183e-01 9.46640149e-02 2.41140991e-01
2.24005952e-02 -1.17921495e+00 -9.16901112e-01 -4.54784185e-01
1.56723455e-01 3.87683183e-01 6.16754889e-01 -2.70656973e-01
2.28937715e-01 5.64820588e-01 -1.69688344e+00 1.86723486e-01
1.22850883e+00 1.01675773e+00 5.21898568e-01 3.02994937e-01
-4.81172264e-01 3.82210791e-01 -2.79841840e-01 -5.16975559e-02
4.10226971e-01 7.18076527e-02 -6.65416539e-01 1.74997225e-01
5.32191932e-01 -3.02876413e-01 -5.63128471e-01 1.07487345e+00
3.71152520e-01 1.72424197e-01 6.82726353e-02 -1.12078238e+00
-4.92125034e-01 2.67522901e-01 -1.01832032e+00 -5.16614728e-02
5.53403534e-02 3.41631635e-03 9.19431329e-01 -1.03465343e+00
4.10391748e-01 1.18452895e+00 7.31228411e-01 3.04898284e-02
-7.49940872e-01 -8.51822138e-01 5.39286733e-01 -4.32362361e-03
-1.69507372e+00 -2.36131642e-02 9.24603105e-01 -1.43424615e-01
1.01848221e+00 1.88763648e-01 8.70441437e-01 4.87900615e-01
-3.35035205e-01 8.35670352e-01 5.80918908e-01 8.16628039e-02
1.81571186e-01 -3.57669771e-01 2.98399143e-02 8.30089331e-01
2.31423244e-01 1.00601956e-01 -3.72102708e-01 -1.92625344e-01
1.15357697e+00 4.68752265e-01 -8.02257806e-02 -4.71187562e-01
-9.43225145e-01 8.24287593e-01 1.01899076e+00 2.39409268e-01
-3.00554335e-01 3.22156638e-01 1.74660668e-01 -1.96566507e-01
5.09745717e-01 1.53141156e-01 -4.23874348e-01 6.12654649e-02
-1.28020716e+00 3.51996213e-01 3.51155221e-01 1.16792142e+00
1.19527435e+00 1.77033409e-01 2.27804303e-01 9.02935028e-01
5.11267066e-01 8.30490708e-01 2.12325096e-01 -1.26291978e+00
5.35585642e-01 9.44149137e-01 -1.99579209e-01 -1.28466010e+00
-1.51516244e-01 -7.03678846e-01 -8.31992030e-01 5.46375699e-02
-3.17474931e-01 2.02857971e-01 -1.10587668e+00 1.26123965e+00
5.61015248e-01 7.53118932e-01 -1.12576716e-01 1.15897644e+00
1.17603421e+00 1.00134003e+00 1.90483436e-01 3.68147463e-01
9.86015856e-01 -1.32163668e+00 -3.68085772e-01 -3.88512760e-01
5.76146960e-01 -7.06500947e-01 9.76862967e-01 4.27392200e-02
-1.09080374e+00 -5.52845657e-01 -9.82441127e-01 -5.00026107e-01
-1.12347417e-01 1.64997175e-01 1.03713608e+00 2.20106542e-01
-6.81575179e-01 5.03593206e-01 -1.17074287e+00 -3.12329344e-02
7.30320513e-01 4.30822551e-01 -1.81708515e-01 -3.98550481e-01
-6.97220683e-01 1.73821807e-01 1.04291268e-01 -7.60810822e-02
-5.21482706e-01 -1.05432284e+00 -1.15933609e+00 3.57946008e-01
1.97928324e-01 -8.42791975e-01 8.58823180e-01 -7.33975410e-01
-1.26446533e+00 6.27471149e-01 -2.76320040e-01 -2.42559686e-01
9.34360176e-03 -4.61496949e-01 -1.75898060e-01 2.91799694e-01
4.76793855e-01 1.03198934e+00 6.22393966e-01 -1.46991253e+00
-6.89016581e-01 -5.50184369e-01 -1.30615577e-01 3.93321037e-01
-1.22628948e-02 -1.97058275e-01 -8.99397552e-01 -6.12935841e-01
6.63614511e-01 -7.48258948e-01 -6.21474743e-01 1.67013049e-01
-1.44937471e-01 -2.14986950e-02 1.13487399e+00 -5.03106177e-01
9.09089804e-01 -2.62464285e+00 2.22255066e-02 2.73801744e-01
2.78938562e-01 1.99945822e-01 -3.30043018e-01 7.89778382e-02
6.22233748e-02 -1.31192524e-03 -6.14508390e-01 -6.32825077e-01
-2.45879307e-01 4.90791351e-01 -2.85186231e-01 4.28846925e-01
3.37424189e-01 8.93075109e-01 -9.21799302e-01 -6.02778196e-01
5.98928452e-01 8.09710443e-01 -8.70843291e-01 3.76047865e-02
-3.53403151e-01 3.55922908e-01 -8.13332260e-01 9.42765832e-01
1.13213134e+00 -4.83450532e-01 -3.57924879e-01 -1.80447280e-01
-2.74120212e-01 8.24520290e-02 -1.32644939e+00 2.36331081e+00
-4.47260261e-01 1.91223264e-01 3.49081218e-01 -7.47257710e-01
9.62622106e-01 -1.29176244e-01 8.38287354e-01 -6.79859936e-01
1.37823939e-01 2.52186179e-01 -3.86097163e-01 -1.04080573e-01
8.90394390e-01 1.63210481e-01 2.97165006e-01 -1.76774319e-02
-8.82159993e-02 -6.64824784e-01 -2.22725511e-01 3.36530268e-01
8.23187292e-01 3.23637366e-01 -1.57614145e-02 -6.64391294e-02
4.24790263e-01 3.93217236e-01 7.29596078e-01 2.06909999e-01
1.85331315e-01 1.04002106e+00 -6.65355101e-02 -3.13480973e-01
-8.62079024e-01 -9.76264477e-01 -9.83954370e-02 8.37180614e-01
7.79573441e-01 -5.91480613e-01 -4.77505535e-01 -3.11046958e-01
1.66440651e-01 2.85842597e-01 -2.39957541e-01 7.86351636e-02
-7.38884509e-01 -4.83529747e-01 8.42817128e-02 7.94104218e-01
8.66411030e-01 -7.52080441e-01 -3.23421448e-01 9.60030779e-02
-1.87766761e-01 -1.37136316e+00 -3.22195947e-01 -9.48164910e-02
-1.23463631e+00 -9.28684831e-01 -4.08542365e-01 -8.41509879e-01
7.63980687e-01 9.31978524e-01 1.15058720e+00 3.77908945e-01
-1.48015872e-01 6.47988841e-02 -4.33095813e-01 -2.19565600e-01
4.27954435e-01 3.68434429e-01 -4.55561370e-01 -2.52662659e-01
1.34229213e-01 -8.11895072e-01 -7.45939672e-01 3.07850927e-01
-9.83334243e-01 4.37487625e-02 5.66107035e-01 4.39079762e-01
1.16238332e+00 -1.89216826e-02 -1.05719686e-01 -6.33194745e-01
-1.19783595e-01 -6.56953275e-01 -5.51760912e-01 -1.22924827e-01
-1.87898830e-01 -2.44021118e-01 4.12935495e-01 8.10824782e-02
-8.10787320e-01 4.35564101e-01 -6.37746274e-01 -8.20327282e-01
-5.37034571e-02 3.75055283e-01 -2.45892644e-01 -5.20812333e-01
2.06573099e-01 3.41326177e-01 -1.79220647e-01 -6.56701088e-01
4.66210306e-01 4.40541804e-01 6.25904441e-01 -6.87210381e-01
9.91436005e-01 9.24780726e-01 -6.50180653e-02 -9.42288578e-01
-1.02595246e+00 -8.89988720e-01 -5.82345724e-01 2.05380931e-01
7.58646548e-01 -1.52450204e+00 -3.06518525e-01 4.09976721e-01
-1.04411733e+00 -1.88417405e-01 -6.38104081e-01 3.32176328e-01
-3.70267659e-01 5.37150741e-01 -3.83329332e-01 -2.94179559e-01
-3.95510256e-01 -1.11215603e+00 1.77342844e+00 4.12886977e-01
1.94556117e-01 -6.85816288e-01 -7.95837343e-02 5.16726315e-01
3.68441194e-01 2.90428847e-01 5.03885567e-01 4.15618308e-02
-1.14839184e+00 -1.63188696e-01 -5.81244349e-01 2.19044015e-01
-7.57054165e-02 -1.75998248e-02 -7.59042084e-01 -2.79810041e-01
-5.62967025e-02 -1.35501623e-01 9.16190922e-01 4.51835722e-01
1.68223178e+00 5.53893372e-02 -4.83150095e-01 1.37376022e+00
1.70412505e+00 -2.10130945e-01 6.65333390e-01 2.63597190e-01
1.26828444e+00 9.50006470e-02 7.19597936e-01 4.78231996e-01
6.68977320e-01 4.70107704e-01 8.42305660e-01 -3.73930633e-01
-2.57181823e-01 -4.78840113e-01 -1.66111112e-01 8.46710026e-01
1.42255649e-02 2.62989365e-02 -8.00587475e-01 7.12179899e-01
-1.85154760e+00 -7.45645761e-01 -3.39870185e-01 1.77643919e+00
3.06896001e-01 -1.23467498e-01 -8.82180855e-02 6.84684440e-02
4.83503968e-01 4.96930718e-01 -5.55745006e-01 1.43420294e-01
-2.62873948e-01 5.45836031e-01 6.40357733e-01 4.83054012e-01
-9.94095385e-01 1.33717680e+00 5.83045292e+00 8.97185981e-01
-1.29017448e+00 1.78551584e-01 4.20193046e-01 -5.41383699e-02
-6.23759389e-01 1.20715201e-01 -7.25570023e-01 3.71103287e-01
3.15806538e-01 2.87074417e-01 1.55154318e-01 1.32107306e+00
-1.44820334e-02 -1.14178643e-01 -4.84973013e-01 1.39946222e+00
-2.87112012e-03 -1.60950851e+00 7.00295419e-02 3.02698296e-02
1.01704073e+00 6.28417075e-01 2.97219642e-02 1.53786078e-01
2.81096548e-01 -9.61628735e-01 6.29426420e-01 2.71850497e-01
6.31235898e-01 -7.81127155e-01 5.76427639e-01 2.20534101e-01
-1.54123151e+00 9.05552879e-02 -5.71989536e-01 -8.86648223e-02
1.70037761e-01 8.88230920e-01 -1.01599708e-01 7.31078625e-01
8.25123250e-01 1.17868328e+00 -4.64205295e-01 1.21150208e+00
-1.59193933e-01 4.72429574e-01 -6.96668625e-01 4.21944499e-01
6.16890788e-01 -4.32612062e-01 4.65968579e-01 9.36303675e-01
3.52990508e-01 3.99681240e-01 6.18987203e-01 8.80433619e-01
-3.31382565e-02 6.99328929e-02 -3.83545816e-01 1.81521550e-01
6.14851534e-01 1.16353786e+00 -7.79237330e-01 -3.14556956e-01
-5.79674840e-01 8.64822328e-01 1.86259374e-01 1.60909981e-01
-7.30458021e-01 -2.34810963e-01 1.07011271e+00 3.29764873e-01
5.12461185e-01 -7.13796616e-01 -8.36308062e-01 -1.23440027e+00
7.15701878e-02 -4.80790883e-01 -9.03420057e-03 -8.23687613e-01
-9.86221373e-01 4.97875005e-01 -2.13369206e-01 -1.24345553e+00
4.04260874e-01 -1.33310378e-01 -5.85653841e-01 7.31440783e-01
-1.99286568e+00 -1.33655512e+00 -8.23407948e-01 7.18158603e-01
6.43274188e-01 1.64524943e-01 4.93101001e-01 3.42376709e-01
-5.09289801e-01 5.96810430e-02 3.86557356e-02 2.59287208e-02
1.67957172e-01 -1.01144588e+00 7.17849135e-01 7.69261539e-01
3.52339796e-03 4.84086066e-01 1.72799721e-01 -7.53099203e-01
-1.28996861e+00 -1.42200291e+00 5.52538872e-01 -9.12786797e-02
2.67967552e-01 -1.92854777e-01 -1.03653610e+00 4.08148617e-01
-3.59130979e-01 5.10773778e-01 4.79063779e-01 2.32577249e-02
-2.55025655e-01 -3.22986431e-02 -1.20104444e+00 2.90699542e-01
1.48260832e+00 -4.65266496e-01 -4.72627699e-01 4.90914553e-01
1.09869969e+00 -8.53553236e-01 -8.89106214e-01 5.70284188e-01
-2.12012846e-02 -8.79996300e-01 1.39183939e+00 -9.16209072e-02
6.45882428e-01 -5.59771538e-01 -4.71195161e-01 -1.00527048e+00
-5.16675234e-01 -1.00617297e-01 -1.20966569e-01 9.95837331e-01
-1.25896677e-01 -3.25770140e-01 1.17991924e+00 2.00615153e-01
-6.40771627e-01 -9.45790529e-01 -6.89252675e-01 -5.82250476e-01
-4.70928289e-02 -5.94834805e-01 1.24319994e+00 8.76366317e-01
-5.81232369e-01 4.56504337e-02 -4.75850590e-02 6.64491713e-01
6.40090704e-01 5.93641400e-01 8.89497757e-01 -1.15744805e+00
2.66473275e-02 -1.83645800e-01 -6.09221280e-01 -1.43035603e+00
1.58247173e-01 -8.49046230e-01 -1.81276381e-01 -1.72887909e+00
3.34577635e-02 -8.71773660e-01 9.63716675e-03 4.42394137e-01
-2.90349901e-01 3.15365881e-01 3.04565996e-01 4.87224936e-01
-7.81717181e-01 8.99238348e-01 1.31258631e+00 -2.03572184e-01
-3.27825099e-01 -2.25876406e-01 -7.29821503e-01 7.85200059e-01
6.66584730e-01 -4.20237273e-01 -3.93131286e-01 -1.05047882e+00
7.21032470e-02 2.10978780e-02 5.11999905e-01 -1.45806253e+00
1.76976055e-01 -2.41593048e-01 3.59229296e-01 -9.72915888e-01
6.44360542e-01 -9.05362189e-01 1.93114076e-02 2.11000144e-01
2.45533973e-01 -1.03994599e-02 1.34911850e-01 5.97599208e-01
-5.48027992e-01 7.03556836e-02 9.31770861e-01 -2.94133067e-01
-1.05229485e+00 9.71356571e-01 3.19543421e-01 -1.12082213e-01
9.53408241e-01 -3.14136744e-01 -9.96727198e-02 9.70437676e-02
-5.58887422e-01 6.27712846e-01 8.19646835e-01 4.18138474e-01
9.41205382e-01 -1.14015353e+00 -4.44232583e-01 5.23152828e-01
-8.54643155e-03 9.34497833e-01 6.52546525e-01 4.12697256e-01
-9.47439492e-01 1.89217523e-01 -1.59264460e-01 -8.90658200e-01
-9.91306424e-01 3.36641520e-01 2.28380293e-01 4.03655097e-02
-9.44356322e-01 1.09442151e+00 2.86340624e-01 -5.77608109e-01
6.60820454e-02 -3.68323267e-01 2.78231427e-02 -4.97966141e-01
2.58943200e-01 2.88726926e-01 1.16231740e-01 -7.66466200e-01
-3.52590919e-01 1.06651568e+00 8.61553326e-02 2.51833469e-01
1.59661651e+00 -8.83945972e-02 -1.28127292e-01 -1.62353795e-02
1.10749686e+00 -3.08303013e-02 -1.45300198e+00 -3.23569298e-01
-5.06961823e-01 -1.02751970e+00 5.13815939e-01 -3.75790268e-01
-1.65587699e+00 7.87156284e-01 4.92432505e-01 -3.66009504e-01
1.03969991e+00 -5.10078408e-02 1.44305634e+00 1.34336486e-01
5.69528103e-01 -8.42435181e-01 -6.16645664e-02 5.86002588e-01
6.11962676e-01 -1.41054475e+00 2.95162827e-01 -1.19765580e+00
-4.72441614e-01 7.06301093e-01 7.25791037e-01 -7.18098283e-01
9.75668490e-01 2.57778287e-01 -1.79716289e-01 -5.38579047e-01
-2.25341439e-01 -3.79401058e-01 1.37962714e-01 6.55559659e-01
1.52285710e-01 7.49910399e-02 -7.58424848e-02 4.10280287e-01
-4.38535303e-01 1.04346603e-01 8.85941237e-02 9.62356627e-01
-5.39118290e-01 -8.76926064e-01 -3.79163384e-01 4.11939263e-01
6.29528090e-02 -9.01633948e-02 -1.39865011e-01 6.19160652e-01
3.75483811e-01 7.38903701e-01 3.17231327e-01 -3.72625262e-01
3.48906428e-01 -5.33335984e-01 1.90032840e-01 -9.00331974e-01
-5.04091442e-01 -3.03586759e-02 -5.00565231e-01 -1.01288450e+00
-4.79566008e-01 -6.17517292e-01 -1.58292270e+00 -4.37759697e-01
-2.35119447e-01 1.95725784e-01 9.79594409e-01 8.99336874e-01
7.04393446e-01 5.70826590e-01 6.85694814e-01 -1.05834651e+00
-5.12260944e-02 -5.65270126e-01 -6.28571272e-01 1.87746510e-01
9.65115428e-02 -6.33488476e-01 -2.48570889e-02 -2.78389961e-01] | [8.015181541442871, -3.4083328247070312] |
fa5a90de-868b-4a2a-81b8-5e26846ed188 | session-aware-information-embedding-for-e | 1707.05955 | null | http://arxiv.org/abs/1707.05955v2 | http://arxiv.org/pdf/1707.05955v2.pdf | Session-aware Information Embedding for E-commerce Product Recommendation | Most of the existing recommender systems assume that user's visiting history
can be constantly recorded. However, in recent online services, the user
identification may be usually unknown and only limited online user behaviors
can be used. It is of great importance to model the temporal online user
behaviors and conduct recommendation for the anonymous users. In this paper, we
propose a list-wise deep neural network based architecture to model the limited
user behaviors within each session. To train the model efficiently, we first
design a session embedding method to pre-train a session representation, which
incorporates different kinds of user search behaviors such as clicks and views.
Based on the learnt session representation, we further propose a list-wise
ranking model to generate the recommendation result for each anonymous user
session. We conduct quantitative experiments on a recently published dataset
from an e-commerce company. The evaluation results validate the effectiveness
of the proposed method, which can outperform the state-of-the-art
significantly. | ['Si Luo', 'Yan Ming', 'Wu Chen'] | 2017-11-20 | null | null | null | null | ['product-recommendation'] | ['miscellaneous'] | [-1.71198159e-01 -4.17043418e-01 -3.52706760e-01 -9.82336462e-01
-2.49150857e-01 -4.75911081e-01 4.95326102e-01 -2.20578253e-01
-3.34602803e-01 3.66061032e-01 4.29571003e-01 -5.11161208e-01
-4.07945454e-01 -1.05122626e+00 -5.30353487e-01 -3.56629372e-01
-1.17945373e-01 3.66410792e-01 4.97460775e-02 -3.41186076e-01
2.86091775e-01 3.68709341e-02 -1.40454698e+00 4.03975964e-01
9.37562764e-01 1.13663399e+00 2.09613666e-01 3.75996679e-01
-3.53542268e-01 3.94015223e-01 -2.80584127e-01 -5.51109731e-01
2.38101721e-01 -3.49069953e-01 -5.34306347e-01 7.74782151e-02
1.22292235e-01 -9.60350633e-01 -9.39295948e-01 8.11295688e-01
4.38315749e-01 7.52933264e-01 4.28862453e-01 -9.76388037e-01
-1.26944590e+00 8.53636742e-01 -2.49888897e-01 1.58306703e-01
3.99394214e-01 -2.40174845e-01 1.36903059e+00 -6.21381938e-01
4.05113459e-01 9.33648169e-01 3.48911375e-01 6.89515889e-01
-9.66650069e-01 -7.44861007e-01 6.90501153e-01 5.46142161e-01
-9.75591838e-01 -1.04502901e-01 9.80071783e-01 -1.17719412e-01
4.92329180e-01 3.13606352e-01 5.28577685e-01 1.32693958e+00
-2.30634108e-01 1.19256675e+00 6.24985039e-01 2.23195955e-01
-2.70432830e-02 5.80619276e-01 4.98866647e-01 3.78066927e-01
8.96125138e-02 8.04797634e-02 -2.11077377e-01 -2.83677280e-01
7.86715031e-01 8.19426239e-01 -2.59754330e-01 -4.31740433e-01
-9.12970364e-01 9.68251109e-01 7.90376663e-01 2.39083722e-01
-3.85369927e-01 -2.54680097e-01 5.15109658e-01 3.99750084e-01
2.85848826e-01 1.60562083e-01 -4.04234499e-01 5.76893650e-02
-7.02948987e-01 7.09892958e-02 1.09074008e+00 8.39688301e-01
5.67953169e-01 -2.25371532e-02 -1.49753347e-01 1.19452441e+00
5.94170272e-01 -2.35828082e-03 7.91405261e-01 -5.11236668e-01
1.65836960e-01 7.20630884e-01 3.23004603e-01 -1.07074487e+00
-1.73084512e-01 -5.96500337e-01 -8.06216061e-01 -6.87363565e-01
1.18371814e-01 -1.06581159e-01 -6.68838084e-01 1.59621263e+00
1.10027909e-01 5.00881314e-01 -1.32982939e-01 1.03493261e+00
1.07782090e+00 8.88816714e-01 -7.01075047e-02 -3.72847587e-01
8.60384583e-01 -1.20809603e+00 -6.81362152e-01 7.15667307e-02
4.66530710e-01 -2.84463108e-01 1.17114758e+00 2.67359495e-01
-7.20878780e-01 -6.58657491e-01 -9.23256099e-01 2.77400780e-02
-4.79742646e-01 4.10008550e-01 1.13427472e+00 6.52951539e-01
-7.02071130e-01 7.28030026e-01 -8.40798914e-01 -4.96830672e-01
2.07085669e-01 6.64215386e-01 -8.62110704e-02 -1.22897625e-01
-1.43773389e+00 1.40116543e-01 1.71164244e-01 3.53803486e-01
-7.65418351e-01 -5.41795433e-01 -4.87182230e-01 4.63125646e-01
4.29086149e-01 -3.62886488e-01 1.28534448e+00 -6.75733268e-01
-1.51363516e+00 4.25382942e-01 -2.89169043e-01 -3.27849269e-01
3.33788157e-01 -8.12890381e-02 -1.02740467e+00 -5.33612013e-01
-3.05870146e-01 -1.03306100e-01 2.58662313e-01 -1.20859718e+00
-7.06759810e-01 -5.04590392e-01 6.20359480e-01 2.38775417e-01
-1.05870044e+00 -1.84300154e-01 -8.80503118e-01 -3.58583272e-01
2.72989143e-02 -9.24545288e-01 -3.64066094e-01 -3.84088159e-01
-3.24065089e-01 -4.97263670e-01 7.68001437e-01 -7.08288074e-01
1.65126193e+00 -2.07109356e+00 -1.42726466e-01 3.29332262e-01
-1.92068219e-02 2.08493531e-01 -1.53437912e-01 5.98635733e-01
2.49480158e-01 1.04433000e-02 1.85537770e-01 -3.69360656e-01
2.74660379e-01 1.76836267e-01 -5.40099084e-01 1.60810128e-01
-7.27790654e-01 8.12995255e-01 -9.01500642e-01 1.62973069e-02
1.71312600e-01 4.54561710e-01 -8.26392233e-01 6.94000900e-01
-9.46288034e-02 1.92923278e-01 -8.28559577e-01 3.30021739e-01
7.65725911e-01 -5.05603492e-01 5.17471611e-01 -2.77843028e-01
1.38131425e-01 5.86845219e-01 -1.12012506e+00 1.78892648e+00
-8.04826319e-01 8.71088952e-02 -2.46387973e-01 -8.09507132e-01
9.44070935e-01 7.50204176e-02 5.90694427e-01 -7.92947650e-01
2.03548610e-01 6.88662976e-02 1.47987306e-01 -3.22366863e-01
7.00889528e-01 6.17377400e-01 -2.99719325e-03 1.01201487e+00
8.65005106e-02 9.87053335e-01 2.71541297e-01 3.91112566e-01
9.74957168e-01 6.05832338e-02 -4.02088426e-02 4.59876135e-02
9.19300914e-01 -6.39861643e-01 7.13993907e-01 8.14963400e-01
-2.97282767e-02 3.95628214e-01 1.48171082e-01 -4.78331417e-01
-7.04858005e-01 -9.84946907e-01 -1.80795547e-02 1.54088235e+00
3.91282827e-01 -5.00392139e-01 -4.59873706e-01 -1.03496730e+00
7.27949664e-02 6.13843918e-01 -7.67118454e-01 -5.04736841e-01
-4.69296813e-01 -8.04702818e-01 -1.86060265e-01 6.02690279e-01
6.33777916e-01 -1.23779583e+00 3.45635444e-01 3.64216357e-01
-4.41571847e-02 -7.75179029e-01 -9.10869062e-01 -4.52878952e-01
-7.89177537e-01 -8.01007986e-01 -6.16978288e-01 -8.51367354e-01
7.18402624e-01 5.73037088e-01 8.10070455e-01 3.00958514e-01
2.25890487e-01 2.15626895e-01 -5.29300809e-01 6.32066876e-02
1.70158625e-01 4.19653773e-01 2.67697036e-01 5.87400973e-01
8.60812247e-01 -8.08079541e-01 -1.24179649e+00 6.70220077e-01
-9.23668087e-01 -3.14523518e-01 4.35279340e-01 8.49469304e-01
3.23908210e-01 1.04727253e-01 7.07160771e-01 -1.68340814e+00
1.01388979e+00 -8.07034850e-01 -2.73980439e-01 3.49281937e-01
-9.86323476e-01 -3.58462900e-01 9.71994519e-01 -6.78031504e-01
-1.26979947e+00 -1.12212285e-01 -1.96567103e-01 -3.05746317e-01
1.39131630e-03 7.52308726e-01 -3.49235982e-01 1.23989157e-01
3.60304683e-01 5.58552086e-01 -3.32224548e-01 -9.80576277e-01
4.16833699e-01 1.08624649e+00 1.95617974e-01 -2.28219882e-01
6.51199937e-01 3.67215097e-01 -7.99688756e-01 -3.78562599e-01
-9.37362254e-01 -7.83584535e-01 -3.29662204e-01 1.72856711e-02
3.20629656e-01 -6.94062769e-01 -1.08160830e+00 2.39278033e-01
-7.60471344e-01 -6.92835897e-02 2.18102366e-01 5.47300577e-01
-2.44937003e-01 4.53418523e-01 -8.38030457e-01 -7.52359986e-01
-4.62428659e-01 -8.91353607e-01 4.81934071e-01 5.88862062e-01
1.24473989e-01 -1.16200161e+00 9.74287242e-02 3.74485046e-01
5.66206038e-01 -5.59463680e-01 7.40187705e-01 -1.20542753e+00
-7.38581359e-01 -5.10321975e-01 -3.89746398e-01 2.62269109e-01
4.06260908e-01 -3.15502971e-01 -7.22742200e-01 -4.03187692e-01
-2.33063161e-01 7.90060759e-02 8.28756213e-01 6.02404550e-02
1.83390844e+00 -5.31026006e-01 -3.75880241e-01 7.13271558e-01
1.22249341e+00 4.82282162e-01 5.65541983e-01 2.47544959e-01
8.54795694e-01 4.20428306e-01 5.42821944e-01 5.02105653e-01
6.79913938e-01 5.66370547e-01 2.52726257e-01 2.57162750e-01
4.33577687e-01 -7.32917905e-01 2.05275089e-01 1.02462435e+00
-8.91733915e-02 -3.70230645e-01 -1.26364246e-01 4.16077346e-01
-2.15668917e+00 -1.02775443e+00 2.24786445e-01 2.36916065e+00
5.07942736e-01 -3.08936872e-02 3.25155973e-01 -1.98094279e-01
6.55767441e-01 2.93421924e-01 -1.00084901e+00 -3.39009374e-01
4.94696110e-01 -2.38009200e-01 3.15812886e-01 1.76696673e-01
-9.74578857e-01 8.02582681e-01 5.72643232e+00 4.62703645e-01
-9.51176763e-01 5.18856309e-02 3.39370579e-01 -1.70878232e-01
-6.79776073e-01 -2.37767920e-01 -8.63627851e-01 8.55032802e-01
9.67074037e-01 -5.75522304e-01 7.36997604e-01 1.12596250e+00
3.15490991e-01 6.35948479e-01 -1.16381299e+00 1.01795709e+00
2.68936425e-01 -1.06290066e+00 2.63610870e-01 3.84159476e-01
8.37071717e-01 -1.92889124e-01 2.53132820e-01 9.09441769e-01
4.13813531e-01 -6.08567953e-01 -2.13764504e-01 5.87782800e-01
2.22428635e-01 -8.22182298e-01 6.71744287e-01 3.57985735e-01
-1.08460033e+00 -5.61394095e-01 -6.64627552e-01 2.37858117e-01
2.84658790e-01 3.26778024e-01 -5.04877388e-01 5.60362160e-01
7.51345515e-01 1.24420369e+00 -4.65619266e-01 1.38731265e+00
3.78059708e-02 8.45061123e-01 -1.52791381e-01 -3.91946256e-01
1.37384370e-01 -7.29984999e-01 7.36331567e-02 8.86423349e-01
4.86223936e-01 1.98223114e-01 2.64073789e-01 7.22934127e-01
-5.10592878e-01 4.59142804e-01 -4.32420164e-01 -3.11256647e-01
3.75167102e-01 1.54060960e+00 -2.58374155e-01 -9.90692675e-02
-7.61880219e-01 1.24575436e+00 3.75325024e-01 5.43591857e-01
-7.26562977e-01 -3.94339263e-01 6.73391998e-01 1.32845312e-01
3.45414221e-01 -1.10821523e-01 1.36849701e-01 -1.52401936e+00
1.05323590e-01 -6.13276005e-01 6.52700305e-01 -5.15296638e-01
-1.64856303e+00 5.08584082e-01 -5.07845998e-01 -1.26931024e+00
-3.27413827e-01 -4.51078534e-01 -1.01880407e+00 7.70455182e-01
-1.17409515e+00 -1.22134948e+00 -3.52210075e-01 6.42956853e-01
5.61272979e-01 -4.83595103e-01 8.44389439e-01 8.33212256e-01
-7.83692062e-01 1.11123955e+00 5.67205667e-01 2.65716881e-01
6.14127398e-01 -1.11430919e+00 4.90255594e-01 3.95888686e-01
3.07800621e-01 1.53241789e+00 3.40842366e-01 -4.47932750e-01
-1.57784259e+00 -1.01000249e+00 7.10544646e-01 -4.18126464e-01
6.99159205e-01 -4.19357270e-01 -1.12002110e+00 9.90090489e-01
-5.35438508e-02 -3.16567063e-01 1.15693271e+00 8.27961326e-01
-3.00862223e-01 -5.05395353e-01 -9.66002047e-01 7.05654979e-01
1.29611313e+00 -7.38257349e-01 -4.82196629e-01 3.82918686e-01
7.74535596e-01 -1.76111758e-01 -7.23891318e-01 1.87595472e-01
8.09169471e-01 -7.68606484e-01 1.03182256e+00 -9.24954116e-01
2.70799428e-01 -1.26114398e-01 -2.31183078e-02 -1.33649957e+00
-5.60647905e-01 -3.87918741e-01 -5.52201211e-01 1.42632651e+00
5.15923858e-01 -6.11652732e-01 1.25745356e+00 9.85673368e-01
9.22576115e-02 -8.06529760e-01 -3.19938183e-01 -4.16590095e-01
-4.94420826e-01 4.70414050e-02 1.00552750e+00 9.52825546e-01
1.53821688e-02 4.19367254e-01 -8.25714767e-01 3.05747632e-02
4.63623434e-01 5.95674813e-01 7.58841932e-01 -1.40243089e+00
-5.89243054e-01 -1.06012434e-01 -7.31454864e-02 -1.60849488e+00
1.00669645e-01 -8.70563030e-01 -3.24033946e-01 -1.72995269e+00
3.78812879e-01 -3.34889591e-01 -1.28719461e+00 4.25658152e-02
-8.12272802e-02 -1.00347884e-01 -2.57544100e-01 1.98950380e-01
-8.79585564e-01 6.26589298e-01 1.18140614e+00 -7.23937824e-02
-5.14808774e-01 6.44501567e-01 -1.22777009e+00 4.51214880e-01
6.37842596e-01 -1.81336831e-02 -9.21209991e-01 -4.57845390e-01
3.07022929e-01 5.90087147e-03 -1.44808665e-01 -5.73591113e-01
3.38679820e-01 -8.51888210e-02 2.93742627e-01 -6.49465501e-01
5.56294799e-01 -9.71142054e-01 -2.45620549e-01 -7.09534362e-02
-7.81837404e-01 -3.30948830e-01 -3.62339556e-01 1.09148192e+00
-6.96715862e-02 -9.38226879e-02 1.77002385e-01 -1.26725763e-01
-8.71344149e-01 1.15666747e+00 -1.15212360e-02 -6.35153770e-01
7.40860403e-01 6.57396689e-02 -1.02364756e-02 -6.85558975e-01
-7.93364048e-01 5.51579356e-01 1.41298383e-01 9.91037369e-01
6.27442658e-01 -1.67090964e+00 -2.09555104e-01 1.55231640e-01
3.41497779e-01 -7.03193963e-01 7.37088084e-01 1.75076798e-01
6.41359091e-02 4.15820271e-01 -9.68421176e-02 2.63153724e-02
-1.03466940e+00 6.98073566e-01 1.41144574e-01 -7.71884173e-02
-2.55359113e-01 6.90119445e-01 1.96241185e-01 -9.93442535e-01
5.12385845e-01 2.95530617e-01 -9.83153343e-01 -5.97225726e-02
7.65720963e-01 3.13978970e-01 -2.24695951e-01 -3.45032126e-01
2.07872335e-02 1.57379285e-01 -8.99458110e-01 1.53207004e-01
1.43353176e+00 -3.58087718e-01 8.41744542e-02 5.08339703e-01
1.52532029e+00 -1.72968775e-01 -8.58666241e-01 -6.69578493e-01
-3.97045225e-01 -8.67761552e-01 1.75040975e-01 -7.23973274e-01
-1.33609045e+00 8.51052642e-01 6.34690046e-01 5.30932784e-01
9.24740970e-01 -4.02745903e-01 1.39461017e+00 7.56304801e-01
4.49573338e-01 -1.22422516e+00 -1.84885293e-01 2.92205036e-01
5.08141220e-01 -1.41168296e+00 -2.11891413e-01 -1.95594534e-01
-6.24706089e-01 7.45248258e-01 1.02617180e+00 -2.32455179e-01
1.07297516e+00 -6.77900553e-01 4.52779084e-02 1.28762990e-01
-6.74273312e-01 -6.50640801e-02 4.02895451e-01 2.18441024e-01
7.36680984e-01 4.20285761e-02 -7.25596428e-01 1.19856107e+00
-1.13604464e-01 1.40881687e-01 4.81820256e-01 5.62729597e-01
-2.47706473e-01 -1.53346229e+00 5.42556405e-01 1.09986103e+00
-4.33911562e-01 -1.88155193e-02 -2.76146859e-01 1.77194521e-01
-3.84738177e-01 1.00171924e+00 -3.28760408e-02 -9.11588192e-01
4.72700357e-01 -7.26312622e-02 -1.08780265e-01 -7.25236714e-01
-4.81471837e-01 -2.33172715e-01 2.84502711e-02 -7.59671688e-01
-2.91062891e-02 -5.52877069e-01 -9.25689220e-01 -5.57006121e-01
-3.61409366e-01 5.12349606e-01 7.07489789e-01 5.37863731e-01
5.43312848e-01 4.60456997e-01 1.24915266e+00 -5.02259672e-01
-7.98136592e-01 -9.33903813e-01 -9.61386740e-01 6.97268963e-01
2.70789824e-02 -5.52733243e-01 -4.16970223e-01 -2.95464575e-01] | [10.159995079040527, 5.623106002807617] |
481a78d9-6813-4bd6-8e64-9994f0327ba2 | scannet-a-fast-and-dense-scanning-framework | 1707.09597 | null | http://arxiv.org/abs/1707.09597v1 | http://arxiv.org/pdf/1707.09597v1.pdf | ScanNet: A Fast and Dense Scanning Framework for Metastatic Breast Cancer Detection from Whole-Slide Images | Lymph node metastasis is one of the most significant diagnostic indicators in
breast cancer, which is traditionally observed under the microscope by
pathologists. In recent years, computerized histology diagnosis has become one
of the most rapidly expanding fields in medical image computing, which
alleviates pathologists' workload and reduces misdiagnosis rate. However,
automatic detection of lymph node metastases from whole slide images remains a
challenging problem, due to the large-scale data with enormous resolutions and
existence of hard mimics. In this paper, we propose a novel framework by
leveraging fully convolutional networks for efficient inference to meet the
speed requirement for clinical practice, while reconstructing dense predictions
under different offsets for ensuring accurate detection on both micro- and
macro-metastases. Incorporating with the strategies of asynchronous sample
prefetching and hard negative mining, the network can be effectively trained.
Extensive experiments on the benchmark dataset of 2016 Camelyon Grand Challenge
corroborated the efficacy of our method. Compared with the state-of-the-art
methods, our method achieved superior performance with a faster speed on the
tumor localization task and surpassed human performance on the WSI
classification task. | ['Pheng-Ann Heng', 'Qi Dou', 'Hao Chen', 'Huangjing Lin', 'Liansheng Wang', 'Jing Qin'] | 2017-07-30 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 2.74210602e-01 -1.03788964e-01 -5.48889995e-01 -7.63282850e-02
-1.05711293e+00 -1.81845650e-01 5.84443621e-02 4.77856725e-01
-7.38912880e-01 6.30813777e-01 -8.90897512e-02 -7.23100722e-01
-1.65027156e-01 -7.41248310e-01 -2.98604250e-01 -1.28460610e+00
2.36752018e-01 6.31868064e-01 3.01392555e-01 1.66945845e-01
9.88247693e-02 6.52624905e-01 -1.15153623e+00 4.08459365e-01
6.25014663e-01 1.20747209e+00 3.46015215e-01 7.78337419e-01
-5.38630523e-02 6.95355415e-01 -2.11631238e-01 -3.03797215e-01
7.33321235e-02 1.31758198e-01 -5.64366937e-01 -2.27330625e-01
1.54269278e-01 -3.78156275e-01 -3.95702332e-01 1.09702408e+00
6.84306562e-01 -6.34573638e-01 3.54831845e-01 -8.20391059e-01
-4.78432328e-02 3.83268565e-01 -8.64051044e-01 3.27462971e-01
-3.25704813e-01 2.23986372e-01 7.97096193e-01 -7.19702303e-01
6.86182857e-01 6.71704650e-01 8.01809371e-01 5.55513799e-01
-1.11638510e+00 -9.43338096e-01 -3.20701689e-01 4.52367783e-01
-1.49423611e+00 -5.03726244e-01 1.38412938e-01 -9.53885242e-02
7.56039619e-01 3.63959819e-01 7.45667875e-01 9.14570332e-01
6.72067463e-01 7.90734470e-01 6.64911032e-01 -2.59906054e-01
9.66580659e-02 3.42763811e-02 -8.89396667e-02 9.76382792e-01
4.84988809e-01 -1.64296910e-01 -3.13706487e-01 -2.67720073e-01
7.01509893e-01 6.90057576e-01 -4.43315059e-01 -1.17495373e-01
-1.45246339e+00 6.53287828e-01 4.98258561e-01 3.38839442e-01
-6.97619691e-02 -1.87857095e-02 7.64298856e-01 -1.12507544e-01
5.08309722e-01 7.22571975e-03 -3.86539906e-01 5.50658256e-02
-8.09433043e-01 -5.77458814e-02 7.42068410e-01 4.77449149e-01
2.25432485e-01 -7.08478212e-01 -3.14568609e-01 5.19558609e-01
8.97068083e-02 5.65225780e-01 8.56234610e-01 -4.18030620e-01
7.68497363e-02 9.24954891e-01 -1.60979807e-01 -7.75653183e-01
-7.98998892e-01 -7.13128746e-01 -1.72870851e+00 -1.36972457e-01
6.58175170e-01 2.61512697e-01 -9.45024669e-01 1.28049552e+00
5.47265768e-01 3.53895366e-01 -1.79569945e-01 7.41737187e-01
7.50464797e-01 2.60291100e-01 2.68246550e-02 -2.35964239e-01
1.76386023e+00 -7.74843812e-01 -6.53058290e-01 -6.82986826e-02
1.16568899e+00 -5.41915119e-01 7.09139764e-01 2.81332403e-01
-6.91479206e-01 -1.15737997e-01 -7.90378213e-01 -1.30246222e-01
-1.68683797e-01 4.13822025e-01 9.47061539e-01 2.32490644e-01
-1.07129812e+00 3.26852202e-01 -1.28589714e+00 -4.16109413e-01
9.53021109e-01 5.33315539e-01 -5.08965790e-01 -1.97789818e-01
-6.92692459e-01 5.19776940e-01 2.57082433e-01 2.50010312e-01
-7.02845871e-01 -1.10160410e+00 -5.05116701e-01 1.27982244e-01
4.27806795e-01 -6.72073543e-01 1.30691051e+00 -4.37395155e-01
-1.24758804e+00 1.03552568e+00 -3.16177785e-01 -4.82917249e-01
5.48066497e-01 2.66894192e-01 -2.04895154e-01 2.16550663e-01
8.40181559e-02 6.93608880e-01 3.40254277e-01 -3.11358631e-01
-9.90779638e-01 -6.31522894e-01 -5.60546279e-01 -5.96603490e-02
-5.89604080e-01 -4.63124782e-01 -7.71030009e-01 -2.18537763e-01
1.51197329e-01 -8.89466286e-01 -6.27875328e-01 5.91338277e-01
-5.15366077e-01 -3.67430747e-01 6.08514428e-01 -4.42244649e-01
1.21439862e+00 -2.40207791e+00 7.23381415e-02 2.65339106e-01
6.91411674e-01 3.92488390e-01 2.22193778e-01 -3.34503762e-02
1.81908101e-01 2.29919655e-03 9.12626386e-02 -3.11304390e-01
-3.19127798e-01 6.94683567e-02 6.29542470e-02 8.70273054e-01
-1.15148261e-01 1.03975964e+00 -8.93616676e-01 -9.26947594e-01
-1.13670610e-01 3.78266662e-01 -4.70683753e-01 8.97117481e-02
-1.63409114e-02 2.60884583e-01 -3.40345889e-01 9.19690847e-01
4.76601213e-01 -1.06442392e+00 4.45565104e-01 -2.25567266e-01
3.13695073e-01 -1.14602678e-01 -7.38409221e-01 1.65381801e+00
-2.02101886e-01 5.51858008e-01 3.36371809e-01 -9.28536594e-01
3.67697269e-01 3.30057025e-01 4.79479969e-01 -7.66494572e-01
2.38998696e-01 4.83322650e-01 1.73337653e-01 -6.12767458e-01
1.69846043e-02 -1.06093585e-01 1.70962319e-01 2.78202504e-01
-4.56307441e-01 3.19699883e-01 2.38755286e-01 1.24637119e-01
1.48293996e+00 -5.33378780e-01 6.54125631e-01 -2.86081553e-01
5.97428918e-01 6.27997518e-02 6.97500706e-01 5.94989538e-01
-3.72728825e-01 2.58546114e-01 5.38341403e-01 -7.66863346e-01
-6.51083171e-01 -7.99439847e-01 -4.40981299e-01 7.22273350e-01
9.87918079e-02 -2.05234602e-01 -3.91943425e-01 -8.24114382e-01
1.88199729e-01 -1.51889518e-01 -8.18095267e-01 2.20609121e-02
-4.55533504e-01 -1.10471523e+00 6.11777961e-01 5.56562662e-01
2.86547035e-01 -8.20036054e-01 -4.65475917e-01 2.29702711e-01
-2.73020148e-01 -1.05730355e+00 -2.29354531e-01 2.70449013e-01
-9.78446841e-01 -1.32454419e+00 -6.50461137e-01 -9.27214980e-01
1.00674462e+00 5.50597072e-01 1.01582146e+00 5.79003215e-01
-1.14660287e+00 -3.37479889e-01 -1.94559395e-02 -6.67714596e-01
-4.35658634e-01 3.99526626e-01 -2.57803559e-01 9.05652810e-03
6.85114384e-01 -1.47390947e-01 -1.10671997e+00 9.85404477e-02
-1.10932696e+00 4.68243212e-01 1.19356883e+00 1.36560464e+00
1.03452361e+00 5.57849631e-02 4.29243743e-01 -1.18426335e+00
1.79426596e-02 -4.91608113e-01 -5.24005234e-01 1.17266916e-01
-4.20936555e-01 -1.54829174e-01 7.39656448e-01 -2.92335331e-01
-7.47210801e-01 2.33882934e-01 -3.99487495e-01 5.82456589e-04
-1.96070410e-02 5.17667055e-01 1.43889368e-01 -8.66156518e-02
4.83516335e-01 2.53269762e-01 4.87810254e-01 -9.77719724e-02
-3.87446016e-01 7.09954739e-01 4.31042552e-01 6.05411641e-02
4.18070406e-01 9.98273492e-01 4.32238489e-01 -6.49715722e-01
-1.11125159e+00 -7.73775756e-01 -1.73421159e-01 2.37520989e-02
6.62013054e-01 -1.01415431e+00 -1.19589865e+00 5.15734971e-01
-8.52275372e-01 -3.18039745e-01 -1.82560049e-02 2.36180618e-01
4.28844318e-02 2.89198309e-01 -1.18255591e+00 -1.68639973e-01
-7.62895048e-01 -1.22106719e+00 1.36364508e+00 2.10232660e-01
-5.84985912e-02 -9.40579891e-01 -9.11106020e-02 2.18463436e-01
6.08213723e-01 1.18045270e-01 1.02569020e+00 -6.17555499e-01
-7.17467189e-01 -5.93920648e-01 -5.53360879e-01 -1.65358812e-01
9.52299908e-02 4.44847457e-02 -8.57112646e-01 -5.85265815e-01
-3.68011147e-01 -3.21224004e-01 9.58883405e-01 3.12505692e-01
1.77431166e+00 -9.92749780e-02 -1.13509476e+00 9.25042033e-01
1.58452177e+00 -3.08348596e-01 4.05386955e-01 3.16498607e-01
4.30594236e-01 1.08513884e-01 5.29639602e-01 7.06880987e-01
2.66900152e-01 2.98881769e-01 6.67401791e-01 -4.75039095e-01
1.02982692e-01 1.04869846e-02 -2.85938621e-01 6.77533388e-01
1.61763310e-01 -3.20976824e-01 -9.99682903e-01 5.37459850e-01
-1.87309539e+00 -7.25581348e-01 -2.75624432e-02 1.94497943e+00
1.18120170e+00 1.78999975e-01 -4.29752976e-01 2.91256726e-01
4.05002117e-01 -2.07543164e-01 -7.31664062e-01 4.32705134e-01
2.81027228e-01 2.20954075e-01 6.62460804e-01 4.62422036e-02
-8.89315903e-01 4.00082260e-01 5.24988174e+00 1.29451501e+00
-1.32636476e+00 9.52529907e-02 1.18730962e+00 -2.37500876e-01
1.94799066e-01 -5.05592048e-01 -1.19155598e+00 5.39246857e-01
7.59420097e-01 4.02838141e-02 -2.12134033e-01 7.93383539e-01
4.94517162e-02 -2.31716380e-01 -1.29614985e+00 1.04338384e+00
-1.17101371e-01 -1.93299830e+00 -6.48233742e-02 4.82234031e-01
5.44062853e-01 1.34600475e-01 3.28489169e-02 2.14339495e-01
1.94195285e-01 -1.10464013e+00 -8.40218067e-02 4.01710123e-01
1.23519146e+00 -6.09881103e-01 1.44460428e+00 7.13129520e-01
-9.06100690e-01 -4.97279279e-02 -4.49192196e-01 1.58394918e-01
-2.48128667e-01 9.57718909e-01 -1.30660188e+00 2.56349206e-01
6.96906507e-01 8.66889536e-01 -6.41896546e-01 9.08390582e-01
2.45465785e-01 6.65640116e-01 -4.87207651e-01 -3.29240352e-01
-4.37752530e-02 4.15947676e-01 -3.38019244e-02 1.27454185e+00
5.44284105e-01 4.72132489e-02 1.32347599e-01 1.22505918e-01
-3.04765433e-01 7.74417818e-02 -1.57825395e-01 1.60087079e-01
6.31716669e-01 1.72564042e+00 -9.79987383e-01 -3.35773319e-01
-2.63963580e-01 6.15222156e-01 5.15578330e-01 -7.65501037e-02
-7.87197948e-01 -2.38540769e-01 3.25702280e-01 3.84230614e-01
9.48064402e-02 2.79795051e-01 -3.51052761e-01 -9.90304470e-01
-2.26567015e-01 -7.61514783e-01 5.56429148e-01 -1.65553503e-02
-1.27178192e+00 3.64936739e-01 -6.88124716e-01 -1.30178499e+00
1.51791498e-01 -8.56103659e-01 -5.08256137e-01 5.42721450e-01
-1.87825489e+00 -9.26218271e-01 -7.27238417e-01 5.18320322e-01
3.01327527e-01 1.46186640e-02 1.12672532e+00 4.01014417e-01
-8.06295991e-01 9.08119321e-01 3.66755784e-01 2.49068365e-01
9.65809584e-01 -1.05865943e+00 -1.44245639e-01 4.36399996e-01
-5.16083360e-01 3.47917914e-01 2.01896057e-01 -2.54920244e-01
-1.73032582e+00 -1.21280468e+00 8.46491754e-01 -1.58606633e-03
8.10172141e-01 -3.20080161e-01 -8.23702335e-01 4.50526625e-01
-2.24813402e-01 5.15269578e-01 1.19771230e+00 -2.24407479e-01
-2.07188372e-02 -5.50056577e-01 -9.99504447e-01 7.01273263e-01
7.18036413e-01 -1.19404890e-01 3.25856060e-01 7.47253060e-01
4.89722908e-01 -7.80300975e-01 -8.29949319e-01 5.65261364e-01
5.07512391e-01 -7.93291271e-01 8.32268298e-01 -1.41312063e-01
5.06650329e-01 -1.13166869e-01 3.08580995e-01 -8.34576964e-01
-5.38231969e-01 -2.65949905e-01 -1.08225502e-01 7.24919558e-01
4.09749687e-01 -6.85270429e-01 1.26322913e+00 2.09960282e-01
-2.04083286e-02 -1.26937950e+00 -1.16764724e+00 -1.67352811e-01
-2.63304085e-01 -6.67328620e-03 3.46031398e-01 4.67360198e-01
5.71995787e-02 4.25633565e-02 1.07948594e-01 8.80586449e-03
6.34145141e-01 3.43489125e-02 6.94597602e-01 -1.20342469e+00
-4.81025398e-01 -6.60418630e-01 -6.24379873e-01 -9.55027223e-01
-7.28890821e-02 -1.11277711e+00 7.07887411e-02 -1.28753281e+00
7.49832332e-01 -4.80998933e-01 -4.65396732e-01 6.27924681e-01
-5.50986767e-01 6.36475682e-01 -4.18004096e-01 3.60374510e-01
-9.12830651e-01 -7.90770799e-02 1.45649743e+00 -2.66474515e-01
3.22926700e-01 -8.83177891e-02 -8.20911825e-01 7.33705938e-01
7.07074940e-01 -4.31620479e-01 -2.57262308e-02 -2.86512464e-01
5.29221594e-01 4.80443060e-01 2.24822417e-01 -9.59022641e-01
6.89124227e-01 -7.97950476e-02 7.55758941e-01 -7.95856655e-01
1.17423743e-01 -8.99877667e-01 3.46295722e-02 9.87628520e-01
-2.23714188e-01 -2.33065054e-01 2.77642310e-01 8.36144745e-01
-2.34098986e-01 2.13745371e-01 8.39848995e-01 1.07818786e-02
-3.02901149e-01 7.86167145e-01 -3.21506560e-01 -3.37321967e-01
1.35936201e+00 -3.06652691e-02 -5.32175839e-01 3.16649437e-01
-1.96329921e-01 4.19598937e-01 3.60447526e-01 -1.25274122e-01
6.30087376e-01 -9.52480197e-01 -9.13376629e-01 3.49756628e-01
2.91938603e-01 6.07783794e-01 7.15136051e-01 1.35105574e+00
-8.46326113e-01 4.29269016e-01 2.09801465e-01 -9.92899716e-01
-1.40327978e+00 3.63007873e-01 2.79455453e-01 -1.15793133e+00
-7.94134676e-01 1.01357615e+00 3.57030362e-01 -1.67551339e-01
5.26284397e-01 -2.54657567e-01 -6.65488988e-02 -3.47450346e-01
8.88342142e-01 1.26507640e-01 3.26725900e-01 1.26066536e-01
-3.00272435e-01 7.66296908e-02 -8.97499800e-01 5.80988586e-01
1.35006487e+00 1.45781279e-01 -3.40092093e-01 1.66505828e-01
1.09820247e+00 -2.72410780e-01 -1.07464600e+00 -5.34752250e-01
-1.56020150e-01 -2.98756450e-01 1.23115778e-01 -8.07874501e-01
-1.15627766e+00 8.66017997e-01 5.77629507e-01 -1.57048747e-01
1.23645127e+00 -8.40377286e-02 1.09176326e+00 4.52037543e-01
3.16109687e-01 -7.08756924e-01 -9.14507359e-02 1.46567583e-01
2.22576931e-01 -1.66006446e+00 2.28822619e-01 -4.92358327e-01
-2.25753840e-02 1.29691720e+00 5.70218503e-01 7.79210404e-02
7.20195591e-01 7.40928710e-01 1.06008910e-01 -1.05571732e-01
-1.18948793e+00 5.07224798e-01 -1.93904668e-01 7.93763027e-02
4.64477599e-01 2.66213059e-01 -9.38386694e-02 5.90122163e-01
1.61593452e-01 3.69635075e-01 2.89517879e-01 1.02702594e+00
-4.60800499e-01 -8.54618609e-01 -7.74854496e-02 1.00962472e+00
-8.45546603e-01 -2.75372237e-01 1.24512650e-01 8.23886037e-01
-1.30816504e-01 4.37421590e-01 1.03020534e-01 -8.79105851e-02
-1.06294990e-01 -3.45696479e-01 1.09387763e-01 -5.17015815e-01
-4.49477673e-01 5.25924601e-02 -3.55782717e-01 -5.54670095e-01
-1.07624665e-01 -4.14698541e-01 -1.40115559e+00 -5.00118077e-01
-3.55386615e-01 3.72226685e-02 6.28382921e-01 7.99040616e-01
6.14222407e-01 7.74858117e-01 4.91629899e-01 -6.11498773e-01
-8.91396344e-01 -9.75917041e-01 -6.43926442e-01 2.31968150e-01
3.87872517e-01 -9.58616957e-02 -4.01755869e-01 4.59920689e-02] | [15.18452262878418, -2.9556844234466553] |
8f419b78-696a-4c82-9a06-daa6bfba78da | joint-supervised-and-self-supervised-learning | 2004.07392 | null | https://arxiv.org/abs/2004.07392v1 | https://arxiv.org/pdf/2004.07392v1.pdf | Joint Supervised and Self-Supervised Learning for 3D Real-World Challenges | Point cloud processing and 3D shape understanding are very challenging tasks for which deep learning techniques have demonstrated great potentials. Still further progresses are essential to allow artificial intelligent agents to interact with the real world, where the amount of annotated data may be limited and integrating new sources of knowledge becomes crucial to support autonomous learning. Here we consider several possible scenarios involving synthetic and real-world point clouds where supervised learning fails due to data scarcity and large domain gaps. We propose to enrich standard feature representations by leveraging self-supervision through a multi-task model that can solve a 3D puzzle while learning the main task of shape classification or part segmentation. An extensive analysis investigating few-shot, transfer learning and cross-domain settings shows the effectiveness of our approach with state-of-the-art results for 3D shape classification and part segmentation. | ['Tatiana Tommasi', 'Antonio Alliegro', 'Davide Boscaini'] | 2020-04-15 | null | null | null | null | ['3d-shape-retrieval'] | ['computer-vision'] | [ 6.39035180e-02 -1.05563521e-01 -7.32547268e-02 -3.03668082e-01
-7.44719386e-01 -7.07018733e-01 7.09002554e-01 2.77821571e-01
-2.73413092e-01 4.38476980e-01 -3.01207304e-01 1.35737509e-02
-2.43028879e-01 -8.20970178e-01 -8.41948450e-01 -4.62159574e-01
-9.19359922e-02 1.29667139e+00 6.04561627e-01 -3.89991492e-01
3.51693898e-01 8.33821177e-01 -1.69320035e+00 1.99520499e-01
9.70444381e-01 9.40135777e-01 4.19283926e-01 4.82957214e-01
-5.52438915e-01 2.42641062e-01 -5.03158867e-01 -1.84898123e-01
7.70453095e-01 2.60345787e-01 -8.21616888e-01 5.35578310e-01
2.81077176e-01 -7.15633109e-02 -2.86884932e-03 8.54187131e-01
4.00565684e-01 1.64578214e-01 7.64458001e-01 -1.33010900e+00
-4.87820387e-01 -2.41236061e-01 -4.69696134e-01 3.06723751e-02
3.55906367e-01 3.17919254e-01 6.70101464e-01 -6.24866307e-01
8.54899943e-01 1.15513313e+00 8.26210618e-01 3.72735977e-01
-1.32505453e+00 -3.25068980e-01 3.54844555e-02 2.73141474e-01
-9.62915659e-01 -1.45559698e-01 9.59961653e-01 -5.53931117e-01
1.27681470e+00 -3.48516554e-01 7.86408126e-01 1.01811719e+00
-3.03096443e-01 9.08329308e-01 7.98925042e-01 -2.14049697e-01
4.67957169e-01 -3.54132243e-02 -1.42590031e-01 7.58056521e-01
1.55045882e-01 2.99628992e-02 -4.01155114e-01 -3.54450047e-02
9.17779863e-01 -3.78676457e-03 1.27133161e-01 -1.08433390e+00
-1.16451502e+00 8.63839269e-01 5.01904786e-01 3.97765338e-01
-4.08170789e-01 -1.21607892e-01 3.57075989e-01 5.57696998e-01
7.95861363e-01 5.80067098e-01 -7.60914445e-01 -1.91237360e-01
-8.58460963e-01 6.13872170e-01 8.22944701e-01 1.15348434e+00
8.88744056e-01 5.66531606e-02 4.11182255e-01 8.24769795e-01
9.97283608e-02 5.86596072e-01 1.44908190e-01 -1.10984802e+00
4.54504311e-01 1.05933905e+00 2.33598381e-01 -4.62228924e-01
-4.94393021e-01 -1.60480812e-01 -4.88245100e-01 7.59882212e-01
6.83369517e-01 1.45732030e-01 -1.16535652e+00 1.20739293e+00
5.45372665e-01 3.85142803e-01 1.02263719e-01 9.34358537e-01
8.51680636e-01 3.75540465e-01 -2.14211904e-02 3.32931876e-01
1.15554607e+00 -8.92124176e-01 -1.32891312e-01 -4.03383136e-01
4.75269020e-01 -6.16116166e-01 1.05272353e+00 1.55207291e-01
-9.18338954e-01 -4.86445159e-01 -1.07453740e+00 -1.27889261e-01
-5.25521278e-01 -6.80067763e-02 8.28862727e-01 3.56079370e-01
-5.90783954e-01 7.20861673e-01 -1.00019777e+00 -5.07891893e-01
1.13163292e+00 4.54456985e-01 -6.07438326e-01 -2.68017709e-01
-5.06993651e-01 1.01113856e+00 3.23960215e-01 -4.29784894e-01
-8.30179155e-01 -9.78058517e-01 -8.96412194e-01 -1.73103169e-01
6.22329593e-01 -1.02801180e+00 1.19290447e+00 -7.05252171e-01
-1.56925261e+00 1.25447214e+00 2.70525843e-01 -4.92615014e-01
5.58375120e-01 -1.45325363e-01 2.32449949e-01 2.22031236e-01
1.19161479e-01 9.50555861e-01 9.09596503e-01 -1.49479973e+00
-5.03512084e-01 -8.13296378e-01 1.95925295e-01 3.12381536e-01
2.33700916e-01 -3.35736275e-01 -2.17982009e-01 -2.90563107e-01
3.42033118e-01 -9.76797044e-01 -3.91321242e-01 1.95814967e-01
-9.14156809e-02 -5.06115377e-01 1.02982461e+00 -1.67100489e-01
-8.63141492e-02 -1.83120048e+00 3.84275556e-01 -1.17553964e-01
-5.27169295e-02 3.89148742e-01 -3.27687472e-01 3.26396018e-01
2.18719572e-01 -1.17615566e-01 -5.38592339e-01 -5.11627793e-01
9.28979516e-02 4.90569711e-01 -1.76351696e-01 3.93199772e-01
4.46540058e-01 1.10824239e+00 -7.89626956e-01 -1.16825253e-01
4.56604093e-01 2.18548000e-01 -4.78707880e-01 2.75861412e-01
-8.50142360e-01 7.84986198e-01 -6.55773461e-01 9.35384035e-01
6.78430498e-01 -4.26299214e-01 -4.41456974e-01 1.49843097e-01
-1.31822294e-02 7.99767580e-03 -1.17728257e+00 2.43655252e+00
-3.69061768e-01 2.73167819e-01 1.59447253e-01 -1.30705595e+00
1.03050530e+00 1.66900516e-01 6.81840003e-01 -5.11314988e-01
1.03078932e-01 1.78769693e-01 -1.54129788e-01 -7.23754108e-01
1.80570573e-01 -4.53572929e-01 -1.13803688e-02 5.17903149e-01
3.29739064e-01 -1.02093339e+00 -8.33683908e-02 -1.48360446e-01
1.15922868e+00 5.32774031e-01 1.71886861e-01 -4.18055132e-02
3.04859102e-01 6.42627537e-01 3.98805529e-01 5.07861555e-01
-2.60184228e-01 4.83057529e-01 2.16066301e-01 -7.29932904e-01
-1.34702635e+00 -1.07658112e+00 1.27313927e-01 9.43505824e-01
2.29928866e-01 1.71699226e-01 -4.42050010e-01 -7.32487142e-01
4.24410969e-01 5.68047523e-01 -5.18803239e-01 8.29725116e-02
-4.15470600e-01 -3.63900155e-01 2.42284939e-01 4.17107850e-01
4.99759614e-01 -1.25796449e+00 -9.60255802e-01 7.56678879e-02
2.29929894e-01 -1.48769653e+00 3.11602116e-01 1.22499041e-01
-1.17398036e+00 -1.22193074e+00 -7.16654360e-01 -7.99860418e-01
5.35079598e-01 5.44051588e-01 1.18097413e+00 -5.63913658e-02
-4.83051151e-01 6.65214837e-01 -4.50882912e-01 -7.79730320e-01
-2.77259707e-01 1.66154787e-01 -1.24665268e-01 -3.29863757e-01
4.55408096e-01 -9.96966124e-01 -2.31729239e-01 1.61841437e-01
-8.51889968e-01 -8.04483443e-02 6.29996359e-01 6.99783742e-01
6.54184759e-01 -2.64598817e-01 7.67506659e-01 -6.39987230e-01
3.44059914e-01 -4.72548246e-01 -7.71473944e-01 1.19576231e-01
-2.25489452e-01 -4.87771407e-02 2.93489248e-01 -3.69672775e-01
-1.12995851e+00 3.19997668e-01 -8.36079270e-02 -8.54720712e-01
-8.39664876e-01 1.70956090e-01 -1.86719671e-01 -2.45160908e-01
5.51421285e-01 1.10652447e-01 2.91152537e-01 -7.14355469e-01
4.27643061e-01 4.05632973e-01 3.67282003e-01 -7.27805853e-01
9.33296800e-01 8.35094690e-01 2.67637759e-01 -9.07871366e-01
-9.12429988e-01 -5.73635638e-01 -1.28502297e+00 -3.79326865e-02
9.67437208e-01 -9.14483011e-01 -6.98158085e-01 4.58661467e-01
-1.22572935e+00 -4.92315233e-01 -5.87824583e-01 2.04888210e-01
-1.14015567e+00 3.89313877e-01 -2.00015247e-01 -6.65612638e-01
-1.49054483e-01 -1.07383358e+00 1.47501922e+00 3.14594328e-01
1.32791013e-01 -1.00338888e+00 2.29749233e-01 6.46451831e-01
2.18246251e-01 3.90770555e-01 1.21116185e+00 -1.02972102e+00
-8.82868409e-01 -1.93124071e-01 -2.37043098e-01 -4.00804579e-02
1.46579728e-01 -4.74739760e-01 -8.73786449e-01 -9.05578882e-02
5.92649542e-02 -8.36423755e-01 6.84830725e-01 2.42283300e-01
1.02316225e+00 4.33133423e-01 -2.56364882e-01 4.39763278e-01
1.29312694e+00 -3.90680097e-02 2.59071618e-01 3.61864150e-01
4.94789809e-01 8.16453934e-01 6.47386372e-01 5.66960335e-01
4.79146451e-01 7.00438380e-01 6.05081439e-01 1.67117611e-01
-1.07322194e-01 -1.49880096e-01 -2.55943567e-01 3.90738010e-01
-2.05956876e-01 4.59959693e-02 -1.32849824e+00 7.15022564e-01
-1.92523837e+00 -9.14664924e-01 9.22349002e-03 2.00326157e+00
3.89071882e-01 8.36966187e-02 1.62922904e-01 8.12558010e-02
3.71996224e-01 -1.13771353e-02 -8.70223880e-01 8.98267776e-02
-7.55067244e-02 2.63656139e-01 1.43016815e-01 1.17755048e-01
-1.37151849e+00 1.16500580e+00 5.73383331e+00 7.22163022e-01
-8.63016248e-01 1.53255403e-01 2.06126750e-01 6.34774342e-02
-3.54056582e-02 -1.59392849e-01 -3.62605661e-01 6.05951529e-03
4.53394532e-01 5.53186722e-02 4.51750100e-01 9.03725445e-01
-2.19851043e-02 -2.37624168e-01 -1.20613003e+00 1.20616138e+00
1.48804694e-01 -1.53057635e+00 -2.71503419e-01 7.78302103e-02
8.17141235e-01 4.97741699e-01 -9.50810760e-02 3.53410035e-01
3.23474824e-01 -8.02474856e-01 4.02587622e-01 5.73083639e-01
4.60867107e-01 -5.47468424e-01 5.03816605e-01 8.46471071e-01
-1.04945707e+00 7.35271350e-03 -5.10478199e-01 -1.70513153e-01
2.26265043e-01 2.93406039e-01 -9.04847443e-01 7.11152017e-01
6.12074852e-01 1.01884449e+00 -4.08106536e-01 1.38642037e+00
-1.38512179e-01 5.87673634e-02 -5.24129152e-01 1.11871593e-01
3.19970578e-01 -4.55774546e-01 9.07442808e-01 6.24149919e-01
2.74184495e-01 2.89911866e-01 5.31243324e-01 1.13345075e+00
5.17051481e-02 -1.23600304e-01 -1.06831133e+00 -6.68938830e-02
2.24979565e-01 1.11727130e+00 -1.06840003e+00 -3.03555936e-01
-5.71572423e-01 9.69860137e-01 4.61434782e-01 8.79195333e-02
-4.30278838e-01 -7.29598403e-02 8.64680409e-01 6.47557294e-03
6.47250712e-01 -7.21194327e-01 -7.21836448e-01 -1.09610689e+00
-1.41588315e-01 -5.91670215e-01 2.14314654e-01 -6.34790778e-01
-1.63407815e+00 2.41078228e-01 8.83997306e-02 -1.39619327e+00
-1.18682131e-01 -7.33567536e-01 -7.38192916e-01 3.92093629e-01
-1.75654924e+00 -1.48278880e+00 -3.29338223e-01 5.08384287e-01
1.01190865e+00 -5.29985189e-01 8.81104350e-01 4.28782217e-02
-2.85141598e-02 -2.63475180e-02 -7.95904174e-02 -4.98149619e-02
3.79218280e-01 -1.17308044e+00 5.49214005e-01 3.24966192e-01
5.72569251e-01 4.82375594e-03 5.16068459e-01 -7.88567007e-01
-1.69171619e+00 -8.65378559e-01 3.44094515e-01 -5.89558065e-01
5.39059877e-01 -5.07937193e-01 -1.13828707e+00 4.47994977e-01
-3.38831022e-02 3.58965605e-01 6.38805807e-01 1.79800689e-01
-4.26699787e-01 2.97057599e-01 -1.43845010e+00 2.48688295e-01
1.39765096e+00 -3.57988834e-01 -1.05512238e+00 4.78565872e-01
6.57556355e-01 -3.94323021e-01 -7.77014613e-01 5.61142504e-01
2.38209620e-01 -8.84872735e-01 1.10592723e+00 -9.86696005e-01
2.91260302e-01 -1.93821684e-01 -2.67096192e-01 -1.47279537e+00
1.89346001e-02 -2.55197316e-01 -3.51309590e-02 7.44821250e-01
1.98890656e-01 -1.92059040e-01 1.35073841e+00 5.75529754e-01
-3.68368715e-01 -5.42813301e-01 -1.18186593e+00 -1.00927460e+00
1.75089344e-01 -6.20668888e-01 6.71220362e-01 8.68869066e-01
-2.17300043e-01 3.25557411e-01 3.35975103e-02 9.11192000e-02
6.78821564e-01 7.39786685e-01 1.13548672e+00 -1.79209709e+00
-1.19444616e-01 -4.88362104e-01 -6.27812386e-01 -1.02248204e+00
5.64945996e-01 -1.07157326e+00 -4.42239493e-02 -1.72373164e+00
1.08076716e-02 -7.24711716e-01 2.13192105e-01 5.99525511e-01
2.68958449e-01 2.13129967e-01 5.16026497e-01 8.35056156e-02
-7.66062915e-01 9.84929979e-01 1.34514141e+00 -4.05327618e-01
-3.31844151e-01 3.18830341e-01 -5.15353493e-02 9.30021346e-01
6.00197017e-01 -3.23894650e-01 -3.56835425e-01 -7.30594635e-01
-2.23529376e-02 8.21827725e-02 4.08870757e-01 -1.03229713e+00
3.91357869e-01 -1.67870715e-01 3.64925832e-01 -6.15681648e-01
8.87227416e-01 -1.16797316e+00 -2.15547785e-01 2.39644766e-01
1.16393059e-01 -1.66866332e-01 4.02164131e-01 7.73788154e-01
-1.46840662e-01 -2.60322809e-01 6.85631454e-01 -5.52034199e-01
-9.44886982e-01 5.57926476e-01 -6.40392900e-02 1.62815660e-01
1.33942878e+00 -4.72288579e-01 -1.24129757e-01 -1.14208266e-01
-9.49417949e-01 2.81993151e-01 7.19623029e-01 6.11549675e-01
8.42514932e-01 -1.23104370e+00 -7.06876040e-01 3.62574518e-01
2.79390842e-01 3.62115145e-01 3.19273829e-01 5.34036756e-01
-3.06436867e-01 3.05177391e-01 -5.74561834e-01 -9.81743097e-01
-1.17276371e+00 5.19931436e-01 1.75229579e-01 -5.35841845e-02
-9.22778308e-01 8.23173821e-01 -1.22240990e-01 -9.14811730e-01
1.58453017e-01 -2.95444310e-01 -6.04426563e-02 9.45129842e-02
8.20400640e-02 3.49778414e-01 1.80920422e-01 -6.28589988e-01
-2.18972921e-01 8.58249843e-01 2.87533432e-01 1.78961784e-01
1.97090411e+00 1.61407411e-01 1.19011886e-01 4.66797709e-01
8.38221192e-01 -4.67851043e-01 -1.48750210e+00 -4.14403707e-01
4.04874384e-01 -6.83391094e-01 -2.01333195e-01 -7.85275877e-01
-7.61888683e-01 1.16890013e+00 4.73995775e-01 2.10584458e-02
7.54893064e-01 4.29596871e-01 7.46490955e-01 7.27837801e-01
8.58264685e-01 -9.83246207e-01 3.99085313e-01 8.15557420e-01
8.88248622e-01 -1.73946214e+00 -3.69306952e-02 -4.11322922e-01
-7.37440526e-01 1.15280020e+00 5.93556225e-01 -5.39310932e-01
8.14665496e-01 2.32633755e-01 -4.17002998e-02 -4.47360665e-01
-6.83297634e-01 -6.97231591e-01 2.71278799e-01 9.40183342e-01
-2.38060981e-01 -1.64529830e-01 3.53006810e-01 5.25347471e-01
-8.95146132e-02 -1.11071475e-01 3.38891685e-01 9.85867083e-01
-6.19445682e-01 -1.35834265e+00 -2.40188912e-01 4.07047719e-01
-4.28967774e-02 5.56854427e-01 -4.36385661e-01 7.83746421e-01
2.59539992e-01 6.37314022e-01 2.30650142e-01 -4.08482179e-02
4.94962215e-01 4.25991178e-01 9.39422429e-01 -9.59569156e-01
-3.88124704e-01 -1.33991361e-01 -1.44915938e-01 -4.92294937e-01
-6.45193517e-01 -8.31924796e-01 -1.43504751e+00 1.07668735e-01
-6.99860975e-02 -2.23885924e-01 8.86002302e-01 1.26250970e+00
6.15510821e-01 2.73118436e-01 2.92326003e-01 -1.56972444e+00
-4.94672120e-01 -7.85868943e-01 -3.65807801e-01 5.31543553e-01
1.92220762e-01 -1.17111623e+00 6.24807365e-02 1.48343469e-03] | [8.016220092773438, -3.042501449584961] |
9adab9ef-da8d-4685-92f2-fbd29143051b | adversarial-feature-augmentation-for-cross | 2208.11021 | null | https://arxiv.org/abs/2208.11021v1 | https://arxiv.org/pdf/2208.11021v1.pdf | Adversarial Feature Augmentation for Cross-domain Few-shot Classification | Existing methods based on meta-learning predict novel-class labels for (target domain) testing tasks via meta knowledge learned from (source domain) training tasks of base classes. However, most existing works may fail to generalize to novel classes due to the probably large domain discrepancy across domains. To address this issue, we propose a novel adversarial feature augmentation (AFA) method to bridge the domain gap in few-shot learning. The feature augmentation is designed to simulate distribution variations by maximizing the domain discrepancy. During adversarial training, the domain discriminator is learned by distinguishing the augmented features (unseen domain) from the original ones (seen domain), while the domain discrepancy is minimized to obtain the optimal feature encoder. The proposed method is a plug-and-play module that can be easily integrated into existing few-shot learning methods based on meta-learning. Extensive experiments on nine datasets demonstrate the superiority of our method for cross-domain few-shot classification compared with the state of the art. Code is available at https://github.com/youthhoo/AFA_For_Few_shot_learning | ['Andy J. Ma', 'Yanxu Hu'] | 2022-08-23 | null | null | null | null | ['cross-domain-few-shot'] | ['computer-vision'] | [ 2.56488413e-01 5.22490358e-03 -4.83370215e-01 -4.16068524e-01
-1.01931727e+00 -4.38637733e-01 6.03864729e-01 -1.74200997e-01
-6.14981577e-02 1.08193672e+00 -5.31642288e-02 1.20910361e-01
1.53401764e-02 -1.00575089e+00 -6.86042905e-01 -7.10356593e-01
3.89744282e-01 4.32770461e-01 3.65863889e-01 -3.72966409e-01
6.75916150e-02 4.46131602e-02 -1.64530098e+00 3.09193641e-01
1.03912556e+00 8.80428791e-01 1.47415400e-01 2.42261648e-01
-3.04490209e-01 6.13767385e-01 -7.11596906e-01 -3.88549924e-01
3.24372977e-01 -8.17597926e-01 -5.31507015e-01 1.53388372e-02
1.28776282e-01 -3.13926786e-01 -4.29959953e-01 1.20305037e+00
5.65438926e-01 4.93939787e-01 8.53887677e-01 -1.70547795e+00
-9.42989171e-01 2.51125991e-01 -3.73342514e-01 2.90743500e-01
1.43444479e-01 8.80218744e-02 6.77708387e-01 -9.90554214e-01
7.84834921e-01 9.45536256e-01 4.36235309e-01 1.11369669e+00
-1.06228948e+00 -1.04180753e+00 -2.26100138e-03 4.36862141e-01
-1.18278003e+00 -4.51350689e-01 1.04893970e+00 -3.77981186e-01
6.38227999e-01 -1.71973854e-01 2.75545716e-01 1.64304924e+00
-3.39480899e-02 8.11512649e-01 9.38566804e-01 -4.30675000e-01
6.46937191e-01 5.44190109e-01 1.21876411e-02 4.83716339e-01
7.97368586e-02 2.44225785e-01 -4.80237007e-01 -3.03061396e-01
4.27286446e-01 4.27384496e-01 -1.82295054e-01 -7.95042336e-01
-8.19807470e-01 1.08010590e+00 1.84543863e-01 1.72654286e-01
-2.38206703e-02 -2.07172051e-01 7.03203201e-01 6.77648127e-01
8.28494251e-01 3.57092798e-01 -4.56808746e-01 -1.20834164e-01
-5.73637724e-01 3.83271456e-01 5.88013113e-01 1.15384448e+00
9.17325079e-01 2.86165953e-01 -3.27976286e-01 1.20268881e+00
-7.58263022e-02 3.00480664e-01 1.08881474e+00 -7.32221067e-01
4.39554930e-01 5.12777150e-01 3.78650315e-02 -3.38867992e-01
1.74385175e-01 -3.58769506e-01 -5.42452395e-01 3.45114440e-01
2.20341116e-01 -3.74790639e-01 -9.76207733e-01 1.75567722e+00
5.98859787e-01 5.98057091e-01 2.58138537e-01 6.30256891e-01
9.29825842e-01 6.45565271e-01 1.68534145e-01 -1.15527265e-01
1.04050815e+00 -1.02411306e+00 -5.96863568e-01 -3.68611038e-01
6.66712046e-01 -4.05965984e-01 1.04566479e+00 5.18492088e-02
-7.19900846e-01 -7.84081578e-01 -1.29710531e+00 2.73400277e-01
-6.91710889e-01 -3.69170934e-01 3.08005154e-01 6.25865459e-01
-2.81881005e-01 8.41853797e-01 -4.33340490e-01 -2.30863124e-01
7.76406229e-01 -4.63467976e-03 -3.91568452e-01 -2.49489456e-01
-1.51957262e+00 7.18987942e-01 5.76593935e-01 -7.96054244e-01
-1.12414408e+00 -9.91200686e-01 -1.05775559e+00 -1.09553151e-02
5.12742341e-01 -6.87513173e-01 1.36434853e+00 -1.09915066e+00
-1.48604119e+00 7.84694910e-01 1.18838884e-01 -3.88386279e-01
4.78535473e-01 -4.97734882e-02 -5.63946903e-01 5.56678139e-03
2.38851249e-01 4.29585159e-01 1.18502021e+00 -1.10428524e+00
-5.84979653e-01 -4.39265698e-01 -6.77087754e-02 1.12799272e-01
-5.86003125e-01 -3.47961545e-01 6.43509552e-02 -8.37441504e-01
-3.14402550e-01 -6.01869822e-01 -4.02531549e-02 6.00146763e-02
3.51864360e-02 -2.52859473e-01 9.69600022e-01 -2.49162033e-01
8.73366177e-01 -2.40047836e+00 -8.80579650e-02 -3.32855284e-01
-5.09568751e-02 6.04544759e-01 -3.04782659e-01 4.50992376e-01
-2.03990653e-01 -2.65695274e-01 -2.99326897e-01 -1.32151380e-01
-3.08484696e-02 1.78707644e-01 -4.63576257e-01 4.16055232e-01
3.09664994e-01 8.27253222e-01 -1.15802717e+00 -3.92778814e-01
4.05244678e-01 1.80941552e-01 -3.33201587e-01 5.06296337e-01
-2.57561594e-01 3.71983171e-01 -5.26337981e-01 8.47626507e-01
8.02201033e-01 -8.08310136e-02 -3.50602232e-02 2.66937017e-01
3.20469201e-01 -8.87739100e-03 -9.82827783e-01 1.85592473e+00
-5.13998747e-01 2.48303413e-01 -4.49488938e-01 -1.40855873e+00
1.32507372e+00 4.54720229e-01 3.33274752e-01 -7.23660529e-01
2.70262152e-01 2.79387653e-01 -9.55944657e-02 -4.01719660e-01
1.07689075e-01 -6.96871281e-01 -1.96569353e-01 3.21215987e-01
6.76277936e-01 -4.57898863e-02 -1.80160273e-02 -1.51843220e-01
1.04910541e+00 2.79607356e-01 7.40658164e-01 1.21050514e-01
4.21185911e-01 1.57348394e-01 8.59605968e-01 6.68034971e-01
-7.13446558e-01 7.02772737e-01 3.28760475e-01 -2.45538279e-01
-1.21100783e+00 -1.18033183e+00 -1.69957727e-01 1.43170619e+00
1.81311622e-01 -4.07971963e-02 -5.40899575e-01 -1.15354872e+00
2.14843214e-01 9.14123297e-01 -8.66176128e-01 -7.03020334e-01
-1.61213741e-01 -3.74610156e-01 3.43880057e-01 7.87706792e-01
4.81197327e-01 -1.13419700e+00 -3.94094557e-01 2.68530071e-01
1.67067721e-01 -8.47358823e-01 -1.82062954e-01 3.29623938e-01
-8.58923435e-01 -1.00951791e+00 -1.10931265e+00 -9.44294691e-01
3.82916719e-01 3.22245628e-01 7.57235527e-01 -5.38958430e-01
-3.30027431e-01 2.23945946e-01 -7.96061814e-01 -5.70601106e-01
-4.69227076e-01 -6.74692243e-02 1.10401668e-01 -7.06158951e-02
8.05988133e-01 -8.38550687e-01 -4.29227203e-01 2.90171385e-01
-9.24935877e-01 -2.86984563e-01 3.41253787e-01 1.47292304e+00
5.10319293e-01 -1.18072601e-02 1.18983078e+00 -1.29493105e+00
5.01696169e-01 -1.10191941e+00 -2.99334735e-01 2.43386716e-01
-4.91242707e-01 -5.36027923e-02 9.72532690e-01 -8.26252639e-01
-1.22622192e+00 -1.50195703e-01 2.61957254e-02 -8.65229368e-01
-5.54663897e-01 7.08173290e-02 -3.88761729e-01 7.35638738e-02
8.73084188e-01 4.43720460e-01 5.95627651e-02 -5.37330270e-01
3.73652905e-01 8.20642889e-01 3.15726399e-01 -5.42844892e-01
8.38601112e-01 4.95966315e-01 -3.35450023e-01 -5.24587750e-01
-9.79692221e-01 -5.47366440e-01 -6.75907314e-01 -1.22706033e-02
4.29619491e-01 -1.02995455e+00 5.26566282e-02 3.84757757e-01
-7.54379630e-01 -2.73288369e-01 -8.26124191e-01 4.30145800e-01
-8.54890704e-01 7.96549916e-02 -1.60928041e-01 -6.11761093e-01
-2.31117815e-01 -7.62411535e-01 7.27399290e-01 3.31056029e-01
-1.37089595e-01 -1.10061336e+00 4.34738040e-01 2.39424229e-01
2.71628380e-01 3.47137570e-01 9.65590298e-01 -1.26531422e+00
8.99037346e-02 -2.99476326e-01 3.30560543e-02 6.00319266e-01
3.00777376e-01 -5.54269612e-01 -1.26713419e+00 -3.24889302e-01
1.94944903e-01 -8.53275001e-01 9.50928926e-01 1.14219368e-01
1.11914968e+00 -9.64085571e-03 -2.98863977e-01 5.92517555e-01
1.50269914e+00 3.94758135e-01 4.65787411e-01 4.34696168e-01
3.86889368e-01 3.53690803e-01 1.16821146e+00 7.88964033e-01
-1.58539444e-01 5.36352336e-01 2.59466916e-01 3.70865136e-01
-2.46084020e-01 -4.21408087e-01 3.25747699e-01 3.27590466e-01
2.68248647e-01 1.89944413e-02 -8.36094379e-01 8.37959528e-01
-1.82393849e+00 -1.14525139e+00 5.96813202e-01 2.07858205e+00
8.72996211e-01 2.27051511e-01 2.10448653e-01 7.71613643e-02
9.51606989e-01 3.35034102e-01 -1.00717700e+00 -3.92032415e-01
2.22268105e-01 4.90398377e-01 7.69165903e-02 5.86206615e-02
-1.19675124e+00 8.33315134e-01 4.83717155e+00 1.20511293e+00
-8.68666291e-01 5.71405888e-01 2.32670575e-01 -1.64857686e-01
-2.51926601e-01 -1.34377316e-01 -9.23921168e-01 5.55008471e-01
7.91343093e-01 -6.39741480e-01 7.86689073e-02 1.46757710e+00
-3.74389857e-01 3.46682221e-01 -1.10224628e+00 7.29641438e-01
2.34140128e-01 -1.16291285e+00 5.40149920e-02 -1.95198864e-01
1.05378377e+00 -1.32537603e-01 2.56897897e-01 9.79196668e-01
1.41297087e-01 -5.90454102e-01 2.95185626e-01 2.61981934e-01
1.04150176e+00 -9.34477448e-01 6.81318641e-01 5.31811893e-01
-9.11548197e-01 -4.13066477e-01 -7.70934403e-01 -3.00968681e-02
-1.26971200e-01 4.13661867e-01 -9.33475018e-01 4.69512343e-01
3.68588537e-01 8.64658117e-01 -3.75526875e-01 8.86907876e-01
-5.31903990e-02 3.99969339e-01 3.44862819e-01 -4.12069224e-02
1.22194268e-01 9.83259007e-02 6.65697396e-01 8.30229163e-01
4.35021281e-01 -3.12263169e-03 1.41157120e-01 9.04779375e-01
-2.70227075e-01 8.29523280e-02 -1.14882290e+00 -1.33647263e-01
7.04306722e-01 9.04025435e-01 -2.31771007e-01 -5.29150367e-01
-7.36827493e-01 1.15171206e+00 4.25340176e-01 2.78396338e-01
-9.47265685e-01 -8.28001142e-01 8.36924314e-01 1.55257270e-01
5.23936629e-01 3.46822292e-01 -1.68473929e-01 -1.31207693e+00
-3.43407057e-02 -8.69904220e-01 6.15943313e-01 -5.14268279e-01
-1.80252147e+00 3.90187055e-01 2.59143300e-02 -1.91488576e+00
-4.14420247e-01 -4.67480749e-01 -8.69708717e-01 6.85841918e-01
-1.62773192e+00 -1.07488430e+00 -3.10447663e-01 8.37187648e-01
1.05588877e+00 -7.96287060e-01 1.10351872e+00 2.56262362e-01
-2.90214300e-01 9.73570049e-01 5.65082014e-01 1.04956634e-01
1.12114811e+00 -1.03346372e+00 3.59231621e-01 4.32893366e-01
-1.16696090e-01 3.14554095e-01 5.09820819e-01 -6.91348195e-01
-9.22463894e-01 -1.37710905e+00 5.18345416e-01 -2.50496984e-01
6.64249480e-01 -3.55458409e-01 -1.19203019e+00 6.39500141e-01
3.03867292e-02 3.42658222e-01 1.15956748e+00 1.07315695e-03
-7.02454269e-01 -1.27472803e-01 -1.44480598e+00 3.48162115e-01
8.83680582e-01 -4.63204175e-01 -1.01969159e+00 2.48086482e-01
8.17173839e-01 -1.16833426e-01 -7.72320271e-01 4.82691258e-01
4.75132674e-01 -8.90087724e-01 8.44034612e-01 -1.02518308e+00
7.51361668e-01 7.22706169e-02 -3.04396719e-01 -1.63096273e+00
-3.85988653e-01 -1.26085818e-01 -5.09523630e-01 1.35288990e+00
1.13226913e-01 -7.06079364e-01 8.38417828e-01 1.60385728e-01
-6.24977797e-02 -6.74585462e-01 -9.36154604e-01 -1.30897129e+00
3.84492874e-01 -1.60954833e-01 6.76931441e-01 1.21259224e+00
1.64179146e-01 2.73918927e-01 -4.71214592e-01 -1.36670321e-01
7.74051607e-01 3.18747640e-01 7.84547210e-01 -1.29063284e+00
-5.27823985e-01 1.93473455e-02 -5.44045746e-01 -5.24623275e-01
5.19598305e-01 -9.80072021e-01 -8.19731355e-02 -1.03115463e+00
4.34162349e-01 -1.83785781e-01 -6.36746466e-01 5.35807312e-01
-2.11026028e-01 2.12572053e-01 1.07873760e-01 1.36686517e-02
-6.62693143e-01 9.50995386e-01 1.31399870e+00 -2.83525258e-01
-1.97386757e-01 1.85105518e-01 -7.84881651e-01 6.47469461e-01
1.08050895e+00 -7.50589192e-01 -6.55684650e-01 6.07537851e-02
-4.75876600e-01 6.90902993e-02 3.03202927e-01 -1.26092935e+00
8.05290118e-02 -4.12450999e-01 6.06714249e-01 -2.70072341e-01
5.50822914e-01 -7.13442087e-01 -3.14577997e-01 4.93320674e-01
-2.76490569e-01 -5.79047859e-01 1.45326823e-01 7.88814783e-01
-2.86943138e-01 -5.68839729e-01 1.15642643e+00 -3.39839548e-01
-1.10918510e+00 4.58068460e-01 3.92696932e-02 4.92062002e-01
1.49096346e+00 -1.52997240e-01 -4.83681768e-01 -2.03362107e-01
-8.55747283e-01 -5.18010221e-02 4.87653613e-01 6.17669702e-01
8.48246694e-01 -1.70726657e+00 -6.43095016e-01 4.61077094e-01
7.30926096e-01 -2.93223023e-01 5.86596251e-01 2.21405402e-01
1.29472971e-01 4.96775582e-02 -7.24620581e-01 -2.44630933e-01
-9.61487830e-01 8.16119611e-01 1.35714054e-01 -7.14564621e-02
-5.24780333e-01 8.41372013e-01 1.52498901e-01 -5.86767912e-01
1.32582664e-01 4.22005594e-01 6.16775302e-04 2.03125194e-01
7.49458134e-01 5.60477197e-01 -2.10382625e-01 -1.90305546e-01
-2.26505563e-01 2.61926442e-01 -3.35858881e-01 6.73284382e-03
1.44563890e+00 7.85494000e-02 6.08663619e-01 7.84490466e-01
1.43858314e+00 -4.59116459e-01 -1.27896333e+00 -5.06701589e-01
-3.93014491e-01 -7.62990773e-01 -3.96906555e-01 -6.64013445e-01
-6.21675670e-01 1.12021220e+00 7.95889735e-01 -6.17322251e-02
9.79633868e-01 5.36132380e-02 9.38610971e-01 2.63043076e-01
3.95657808e-01 -1.28276277e+00 4.43581372e-01 4.14553046e-01
5.73935151e-01 -1.56025970e+00 -1.46851704e-01 -1.59605712e-01
-9.76794660e-01 8.88867795e-01 1.02183127e+00 -4.22187924e-01
8.14007461e-01 -6.74246550e-02 -1.05697550e-01 8.89552906e-02
-8.22930336e-01 -1.75067052e-01 1.24179393e-01 1.07365322e+00
1.57439321e-01 -6.93739252e-03 -1.98667333e-01 1.01344788e+00
2.27544382e-01 1.87121332e-01 3.43934029e-01 1.04073179e+00
-6.77082717e-01 -1.28745162e+00 -9.80558917e-02 4.80009407e-01
-2.65882671e-01 7.03659356e-02 -2.03716561e-01 7.68757939e-01
3.80762070e-01 7.59559751e-01 -3.46429236e-02 -5.38768172e-01
3.68106812e-01 6.62217557e-01 4.39277679e-01 -1.17051554e+00
-1.70949504e-01 -3.93696606e-01 -2.36955702e-01 -3.35009575e-01
-1.24323696e-01 -4.29430574e-01 -9.80479717e-01 -1.15239099e-02
-2.20107794e-01 2.24017188e-01 2.64607251e-01 9.09272313e-01
2.98389912e-01 6.23098552e-01 9.94702756e-01 -5.95229149e-01
-1.10315919e+00 -9.97634828e-01 -9.66704249e-01 7.51782656e-01
3.12897474e-01 -1.12473297e+00 -5.03658652e-01 -5.95292374e-02] | [10.077657699584961, 3.0714919567108154] |
572f6c1d-73bd-4135-a31d-f56aab96a670 | towards-large-vocabulary-kazakh-russian-sign | null | null | https://aclanthology.org/2022.signlang-1.24 | https://aclanthology.org/2022.signlang-1.24.pdf | Towards Large Vocabulary Kazakh-Russian Sign Language Dataset: KRSL-OnlineSchool | This paper presents a new dataset for Kazakh-Russian Sign Language (KRSL) created for the purposes of Sign Language Processing. In 2020, Kazakhstan’s schools were quickly switched to online mode due to the COVID-19 pandemic. Every working day, the El-arna TV channel was broadcasting video lessons for grades from 1 to 11 with sign language translation. This opportunity allowed us to record a corpus with a large vocabulary and spontaneous SL interpretation. To this end, this corpus contains video recordings of Kazakhstan’s online school translated to Kazakh-Russian sign language by 7 interpreters. At the moment we collected and cleaned 890 hours of video material. A custom annotation tool was created to make the process of data annotation simple and easy-to-use by the Deaf community. To date, around 325 hours of videos have been annotated with glosses and 4,009 lessons out of 4,547 were transcribed with automatic speech-to-text software. The KRSL-OnlineSchool dataset will be made publicly available at https://krslproject.github.io/online-school/ | ['Anara Sandygulova', 'Vadim Kimmelman', 'Aigerim Kydyrbekova', 'Medet Mukushev'] | null | null | null | null | signlang-lrec-2022-6 | ['sign-language-translation'] | ['computer-vision'] | [-6.35153428e-02 1.51074141e-01 -2.53031373e-01 -3.08501303e-01
-1.05853450e+00 -7.92077184e-01 4.49776053e-01 -2.85405904e-01
-5.18622816e-01 6.59274757e-01 8.53664041e-01 -4.35681880e-01
2.08694823e-02 -4.80801821e-01 -5.06453037e-01 -4.54189599e-01
3.94553065e-01 4.22421575e-01 3.53784114e-01 -1.45456046e-01
2.43797600e-01 2.90107399e-01 -1.71650410e+00 1.14293836e-01
7.81645536e-01 1.11760035e-01 2.70850867e-01 8.87054145e-01
-1.53629512e-01 1.06615639e+00 -5.94781101e-01 -5.36522448e-01
2.46323273e-01 -6.22608364e-01 -1.13593042e+00 -2.08210036e-01
8.30497503e-01 -9.11658168e-01 -6.79115772e-01 8.81055593e-01
7.16848850e-01 -7.18090013e-02 1.56462848e-01 -9.47037637e-01
-3.98433447e-01 9.93390799e-01 1.88576251e-01 -2.17361942e-01
7.43205309e-01 2.47065201e-01 9.30706084e-01 -6.46525681e-01
1.42689979e+00 8.05718064e-01 2.26094931e-01 6.57175720e-01
-5.15881479e-01 -1.04089391e+00 -1.58943206e-01 5.11493564e-01
-1.25348520e+00 -5.93162715e-01 6.07706904e-01 -6.04947746e-01
5.56481540e-01 2.66627103e-01 1.34235501e+00 1.10122490e+00
-5.78471065e-01 1.12008274e+00 1.22177577e+00 -8.10850084e-01
-8.22846442e-02 -2.15249032e-01 1.26956165e-01 6.59242332e-01
1.74347535e-01 -4.49963093e-01 -9.52178299e-01 3.01565051e-01
7.56231070e-01 -7.35815167e-01 -3.16875160e-01 5.62308580e-02
-1.14930570e+00 4.75580633e-01 -4.55746055e-01 4.02558565e-01
3.39654237e-02 -9.26634446e-02 4.96998191e-01 7.74033427e-01
1.51856571e-01 -2.53760740e-02 -1.51954457e-01 -1.09510791e+00
-7.02227414e-01 2.89827317e-01 6.04435086e-01 1.06238997e+00
1.03003755e-01 1.17503218e-02 2.92939991e-01 9.52301919e-01
4.86451864e-01 6.81212962e-01 3.72193366e-01 -1.04651642e+00
7.03016937e-01 4.40178990e-01 -2.55917490e-01 -3.65536451e-01
-1.16050631e-01 3.64089847e-01 -1.40765443e-01 1.41301751e-01
1.14410281e+00 -1.12233520e-01 -1.15489137e+00 1.10431170e+00
3.72962534e-01 -4.51724418e-02 9.33460891e-03 1.09432101e+00
1.31954110e+00 3.23735476e-01 1.49566424e-03 -1.27452105e-01
1.17077005e+00 -7.48095989e-01 -9.58019495e-01 3.79640132e-01
9.71967161e-01 -1.30595160e+00 1.16521060e+00 6.08891129e-01
-1.17277491e+00 5.99535368e-02 -5.46052039e-01 -3.26051563e-01
-2.33285248e-01 -9.13100764e-02 5.26925206e-01 7.85441458e-01
-1.01932657e+00 1.10787094e-01 -8.36323738e-01 -6.71086788e-01
4.41050529e-01 -2.12136097e-02 -6.37715340e-01 -2.16582760e-01
-8.31875801e-01 5.67446649e-01 6.24088570e-02 2.96280850e-02
-4.65857983e-01 -3.33363533e-01 -7.69615412e-01 -6.29042983e-01
2.48310626e-01 2.22795568e-02 1.60370767e+00 -9.60628748e-01
-2.02207088e+00 1.38523316e+00 -4.92583439e-02 3.93110849e-02
9.22047615e-01 -3.09089631e-01 -4.37340915e-01 3.98281783e-01
1.49157748e-01 1.77974522e-01 3.53917301e-01 -7.02386558e-01
-8.55160356e-01 -1.42564625e-01 1.29454513e-03 1.38780415e-01
2.74312887e-02 8.12342942e-01 -7.73215890e-01 -7.56763637e-01
3.95438939e-01 -8.46572876e-01 3.96475673e-01 -1.52198315e-01
-5.63184954e-02 -4.01495919e-02 5.58036745e-01 -1.39637196e+00
1.26454222e+00 -2.03680944e+00 -1.53961986e-01 2.33488545e-01
3.86749059e-02 5.98409116e-01 -8.83429199e-02 8.11058342e-01
1.78342998e-01 -1.38607696e-01 -1.00312652e-02 -8.23085755e-02
7.06095994e-02 2.97594577e-01 -6.96449950e-02 5.97712934e-01
-3.88189226e-01 6.31608665e-01 -1.21248102e+00 -7.57634580e-01
3.61063629e-01 4.71979886e-01 -3.90061677e-01 3.21474150e-02
1.43293172e-01 8.84754598e-01 -3.86853725e-01 1.14375687e+00
4.73852396e-01 4.15036440e-01 3.85348856e-01 5.92653513e-01
-6.75596774e-01 5.53447843e-01 -1.07241297e+00 1.73639035e+00
-2.40892723e-01 1.31932306e+00 2.80647278e-01 -4.93597388e-01
5.45575976e-01 6.21811867e-01 5.45231581e-01 -7.05683291e-01
2.83277333e-01 6.32890403e-01 -9.83307138e-02 -9.32358742e-01
6.20132267e-01 3.48143846e-01 8.99988320e-03 3.98541957e-01
8.25954899e-02 -5.19796491e-01 5.48428774e-01 1.97756197e-02
1.09443021e+00 4.47625637e-01 -7.69424811e-03 2.70509928e-01
4.58967358e-01 2.87382334e-01 5.44397354e-01 5.03131926e-01
-3.32756191e-01 5.53402662e-01 2.18114957e-01 -4.29464579e-01
-8.50951910e-01 -9.80012774e-01 -1.58291757e-01 9.46202338e-01
-5.44080675e-01 -5.01321912e-01 -7.52440751e-01 -2.96280116e-01
-5.48172534e-01 5.98279595e-01 1.21932626e-01 6.84193850e-01
-8.94698381e-01 4.02884811e-01 9.07222331e-01 4.49576452e-02
6.08776450e-01 -1.27004647e+00 -6.94635510e-01 9.63232368e-02
-3.63001019e-01 -1.19732404e+00 -5.08278131e-01 -5.52629173e-01
-4.87689525e-01 -1.20346498e+00 -9.68568683e-01 -1.30306077e+00
6.55822098e-01 -1.11376211e-01 5.46468019e-01 -2.81935520e-02
-2.57249981e-01 6.88314140e-01 -8.16328704e-01 -4.54530060e-01
-6.57440782e-01 -1.89199522e-01 -1.81810185e-01 -4.91349548e-01
4.71374512e-01 -4.89128798e-01 -8.47184509e-02 -5.73099777e-02
-8.08293819e-01 2.88103610e-01 2.60705292e-01 5.72405338e-01
3.55392665e-01 -4.92517889e-01 5.90833314e-02 -5.15695572e-01
2.19972834e-01 2.39136890e-01 -9.85974371e-01 8.80422816e-02
-7.72370398e-02 -4.82562602e-01 2.22426921e-01 -2.07495674e-01
-9.53787208e-01 1.77529976e-01 -4.26558912e-01 6.26571989e-03
-3.90681803e-01 3.81825656e-01 -7.43742436e-02 -1.49832040e-01
-1.69643909e-02 3.81194919e-01 -1.88095748e-01 -5.10197163e-01
4.08976614e-01 1.43979049e+00 7.69327700e-01 -3.88036847e-01
6.70357347e-01 2.31459886e-01 -2.25224718e-01 -1.40503252e+00
-3.56772274e-01 -5.98767996e-01 -8.84651780e-01 -1.05819297e+00
8.22190404e-01 -1.04246068e+00 -7.46507943e-01 1.38523757e+00
-8.13552141e-01 -8.29111397e-01 -3.24000746e-01 8.83267999e-01
-6.38354003e-01 4.65046138e-01 -6.53620481e-01 -7.79708564e-01
-5.60541302e-02 -9.36747909e-01 6.89036906e-01 3.16510379e-01
-2.05366507e-01 -5.69776177e-01 1.49533376e-01 1.04403830e+00
-7.53511488e-02 1.03689969e-01 3.57501566e-01 -3.23550045e-01
-5.63096225e-01 -2.13682562e-01 -6.39717430e-02 4.66255337e-01
3.82682793e-02 3.59537870e-01 -7.69049823e-01 -7.99722821e-02
-6.88574612e-01 -3.48775566e-01 2.95051724e-01 3.28880161e-01
5.73817849e-01 -3.02760839e-01 4.73367333e-01 3.28680307e-01
1.10260403e+00 3.55575621e-01 6.78672194e-01 5.40848374e-01
6.39952898e-01 4.37221378e-01 7.05031097e-01 3.43370229e-01
9.15384173e-01 6.74240410e-01 -2.24865273e-01 4.06818837e-01
-6.62822366e-01 -6.11218572e-01 5.40088117e-01 1.67739165e+00
-6.03718400e-01 1.81271493e-01 -1.33861935e+00 1.13076377e+00
-1.61223567e+00 -1.08321238e+00 -4.52508569e-01 2.34129453e+00
9.84081328e-01 -1.35817230e-01 2.06392661e-01 2.04424918e-01
3.76538128e-01 -4.70791869e-02 1.24771036e-01 -4.35026973e-01
-1.55581400e-01 3.13584596e-01 6.87888503e-01 8.50745201e-01
-7.75103509e-01 1.39779305e+00 5.54128933e+00 5.32710016e-01
-1.15160418e+00 5.61786555e-02 -1.85719997e-01 -1.52247712e-01
-3.00828129e-01 1.03043877e-01 -6.90511346e-01 4.24992502e-01
7.94960082e-01 -7.36143887e-02 5.12962818e-01 4.52610850e-01
7.27769196e-01 -3.50415856e-01 -4.81615722e-01 1.00708568e+00
6.03433661e-02 -1.30269861e+00 -2.58110583e-01 1.58171654e-01
8.25156271e-01 4.44808692e-01 -6.09497249e-01 1.62744313e-01
4.67161447e-01 -6.23027205e-01 1.12960625e+00 6.54888749e-01
1.25370061e+00 -4.43213880e-01 5.01612723e-01 1.49725556e-01
-1.16715825e+00 7.44134188e-02 2.66398996e-01 -3.55649292e-01
4.10566658e-01 4.71725203e-02 -8.85686278e-01 3.32201183e-01
6.82202756e-01 6.43788457e-01 -2.43842959e-01 1.13963521e+00
-7.89890587e-01 1.12172973e+00 -4.76716816e-01 -1.36512876e-01
2.59707987e-01 -4.79487658e-01 9.13722277e-01 1.00012910e+00
3.11158866e-01 2.47495219e-01 -1.18573755e-01 -7.28857815e-02
4.55251485e-02 5.09404540e-01 -7.74828494e-01 -2.94807404e-01
4.80495661e-01 5.98542392e-01 -7.25876808e-01 -2.73077905e-01
-6.12127006e-01 8.09885323e-01 -1.09030418e-01 3.29411924e-01
-4.30299073e-01 -5.03611743e-01 7.44041085e-01 3.90366375e-01
8.85631889e-02 -5.92049122e-01 -9.66893658e-02 -1.19102597e+00
1.68887421e-01 -9.21622992e-01 3.08419317e-01 -8.39830518e-01
-5.79746604e-01 6.60662502e-02 1.37344465e-01 -1.29211581e+00
-4.26850766e-01 -5.84242463e-01 -2.94654846e-01 4.55023259e-01
-1.22560322e+00 -1.41008031e+00 -1.85633153e-01 5.03629386e-01
7.52826929e-01 -1.53628081e-01 7.93311000e-01 6.43631577e-01
-2.90370047e-01 4.40700352e-01 1.58640563e-01 6.86302900e-01
8.01608384e-01 -9.77720022e-01 1.48880765e-01 1.02441120e+00
2.15362623e-01 2.31595963e-01 6.25301123e-01 -7.42645800e-01
-1.39378905e+00 -5.21919668e-01 1.80502987e+00 -3.33034664e-01
9.89592910e-01 -8.79982039e-02 -3.90229464e-01 8.05590868e-01
3.35971355e-01 -4.54818100e-01 6.48330510e-01 -5.63946545e-01
-5.53359911e-02 2.72105128e-01 -9.04362679e-01 7.81667113e-01
1.40583980e+00 -8.29391718e-01 -6.63313925e-01 4.41068709e-01
1.01470150e-01 -9.36155677e-01 -9.10235047e-01 -1.04790092e-01
1.03964055e+00 -6.14527762e-01 2.34101295e-01 -3.37221295e-01
3.35173965e-01 -3.30977440e-01 -7.23837838e-02 -7.58646429e-01
4.75508869e-01 -1.09911215e+00 3.84257078e-01 1.27565122e+00
1.25556096e-01 -7.63974190e-01 6.95025265e-01 5.18128693e-01
-3.71510237e-01 -1.15112461e-01 -1.35292649e+00 -6.24684870e-01
2.69881450e-02 -9.68124866e-01 4.34252828e-01 1.03425324e+00
3.78226340e-01 -3.53524059e-01 -3.57754618e-01 8.08863807e-03
4.31046456e-01 -9.65145826e-02 1.20074153e+00 -9.94917452e-01
8.93686339e-02 -6.31504714e-01 -7.03504622e-01 -8.18778038e-01
1.34617195e-01 -8.85359764e-01 -8.19120109e-02 -1.85776508e+00
-2.99357563e-01 5.31184897e-02 4.59101439e-01 8.53140771e-01
5.16328812e-01 4.21667993e-01 4.14630473e-01 2.37898722e-01
-2.10196614e-01 8.36763531e-03 1.45549333e+00 1.02585688e-01
-4.61558670e-01 -3.05714994e-03 6.34572236e-03 6.69893980e-01
7.93490469e-01 -1.98820502e-01 1.53707881e-02 -6.15813017e-01
1.65115416e-01 2.84731835e-02 9.40015540e-02 -8.31501961e-01
1.06573783e-01 -2.14065373e-01 -4.31501001e-01 -7.48741567e-01
7.85196349e-02 -7.05448806e-01 3.05569887e-01 3.38539451e-01
-9.61889848e-02 -3.93649697e-01 1.48253115e-02 -3.57100248e-01
-4.71439898e-01 -2.84446687e-01 4.90522444e-01 -3.32731642e-02
-9.81312513e-01 -1.82336599e-01 -1.05081141e+00 1.81347892e-01
1.05019927e+00 -6.73396111e-01 -3.40189666e-01 -7.80465066e-01
-6.08956158e-01 2.79516399e-01 5.74018896e-01 3.26857299e-01
5.00353634e-01 -1.21580780e+00 -7.65129328e-01 4.23781425e-01
1.82884932e-01 4.20738570e-02 3.05058360e-01 8.30960095e-01
-1.31565237e+00 4.04019982e-01 -4.69805211e-01 -2.26333946e-01
-1.89331019e+00 -5.43329120e-01 -2.45761722e-01 1.44968435e-01
-1.19886982e+00 6.24660134e-01 -7.66276777e-01 -4.59613055e-01
4.10227984e-01 -4.79656786e-01 -1.91334233e-01 2.26346523e-01
5.50543010e-01 4.68688369e-01 -2.17022941e-01 -1.22762644e+00
-1.64448723e-01 7.67234206e-01 4.03515995e-01 -5.72333992e-01
1.39421189e+00 -1.91351995e-01 8.42007715e-03 5.80045104e-01
8.08654666e-01 6.12399995e-01 -6.97923243e-01 -2.15970702e-03
-1.11083053e-01 -6.38697028e-01 -1.30970301e-02 -6.63142800e-01
-8.21940541e-01 3.86854529e-01 3.76487762e-01 -2.36091226e-01
1.21014547e+00 9.13183093e-02 8.70264173e-01 2.68052518e-01
4.73614424e-01 -1.62320721e+00 -7.07340240e-01 8.84458005e-01
7.46791005e-01 -9.72444594e-01 -1.98150307e-01 -2.64311492e-01
-5.82308590e-01 1.26446295e+00 4.02418599e-02 3.23790848e-01
5.92518926e-01 9.02338326e-02 8.53033304e-01 -9.03452635e-02
-1.86226398e-01 -4.31586146e-01 1.25906885e-01 6.75077617e-01
4.64723200e-01 3.35783631e-01 -9.49329436e-01 1.67430818e-01
-9.34101164e-01 5.13213098e-01 1.07636392e+00 1.27193069e+00
-6.20366484e-02 -1.26058972e+00 -4.34644699e-01 1.60152599e-01
-4.78936106e-01 1.23042516e-01 -4.73200142e-01 9.70808804e-01
-8.92634392e-02 8.60675752e-01 -7.24571198e-02 -9.09066480e-03
4.51240897e-01 3.62286150e-01 5.40001690e-01 -2.91984439e-01
-3.65441233e-01 8.47826153e-02 6.53587520e-01 -4.36254025e-01
-5.91834426e-01 -1.15648234e+00 -1.41447556e+00 -3.97431016e-01
1.07419223e-01 -1.07969165e-01 8.78661215e-01 9.18595135e-01
-2.94840544e-01 -1.22857364e-02 1.38897285e-01 -5.68871379e-01
-2.12807003e-02 -8.90188992e-01 -4.93058860e-01 1.35399967e-01
1.81959361e-01 -1.10897005e-01 -1.66165456e-01 3.52250010e-01] | [9.143714904785156, -6.458750247955322] |
f74d35da-d4ba-4c91-a89d-6196c0f9be52 | phrasetransformer-an-incorporation-of-local | null | null | https://link.springer.com/article/10.1007/s10489-022-04246-0 | https://link.springer.com/content/pdf/10.1007/s10489-022-04246-0.pdf | PhraseTransformer: An Incorporation of Local Context Information into Sequence-to-sequence Semantic Parsing | Semantic parsing is a challenging task mapping a natural language utterance to machine-understandable information representation. Recently, approaches using neural machine translation (NMT) have achieved many promising results, especially the Transformer. However, the typical drawback of adapting the vanilla Transformer to semantic parsing is that it does not consider the phrase in expressing the information of sentences while phrases play an important role in constructing the sentence meaning. Therefore, we propose an architecture, PhraseTransformer, that is capable of a more detailed meaning representation by learning the phrase dependencies in the sentence. The main idea is to incorporate Long Short-Term Memory into the Self-Attention mechanism of the original Transformer to capture the local context of a word. Experimental results show that our proposed model performs better than the original Transformer in terms of understanding sentences structure as well as logical representation and raises the model local context-awareness without any support from external tree information. Besides, although the recurrent architecture is integrated, the number of sequential operations of the PhraseTransformer is still (1) similar to the original Transformer. Our proposed model achieves strong competitive performance on Geo and MSParS datasets, and leads to SOTA performance on the Atis dataset for methods using neural networks. In addition, to prove the generalization of our proposed model, we also conduct extensive experiments on three translation datasets IWLST14 German-English, IWSLT15 Vietnamese-English, WMT14 English-German, and show significant improvement. Our code is available at https://github.com/phuongnm94/PhraseTransformer.git. | ['Minh Le Nguyen', 'Vu Tran', 'Huy Tien Nguyen', 'Tung Le', 'Phuong Minh Nguyen'] | 2022-11-29 | null | null | null | applied-intelligence-2022-11 | ['nmt', 'semantic-parsing'] | ['computer-code', 'natural-language-processing'] | [ 3.20406139e-01 3.64453495e-01 -3.23516011e-01 -6.20187163e-01
-7.59218395e-01 -3.55661482e-01 2.83990473e-01 -1.23205483e-01
-4.97748405e-01 6.93621278e-01 4.51586127e-01 -5.77727675e-01
3.18125993e-01 -9.89213765e-01 -9.22316253e-01 -5.47889233e-01
5.56339741e-01 4.46042269e-01 1.26530752e-01 -4.44592118e-01
9.12370384e-02 -1.83554396e-01 -9.52551603e-01 6.07388675e-01
1.03588223e+00 8.12308311e-01 7.63868392e-01 2.52221227e-01
-7.18972683e-01 7.56009161e-01 -3.96582156e-01 -4.47706193e-01
-9.37756822e-02 -5.70719957e-01 -1.16293502e+00 -1.99030727e-01
4.47428226e-02 1.16021223e-02 -2.31200960e-02 1.05775666e+00
3.35255057e-01 -1.81862097e-02 1.84994623e-01 -7.70833015e-01
-9.70294595e-01 1.09406626e+00 -4.42203075e-01 2.05279186e-01
3.35407495e-01 -2.30941013e-01 1.17734361e+00 -1.03252184e+00
5.00280797e-01 1.63776183e+00 5.34418881e-01 5.87611914e-01
-7.77661324e-01 -6.86052799e-01 6.18408144e-01 4.62314993e-01
-1.09802377e+00 -2.79030591e-01 7.24310815e-01 2.59828895e-01
1.31342626e+00 1.89454630e-01 2.89813876e-01 1.24929821e+00
3.33725601e-01 9.55488920e-01 1.17586005e+00 -6.27935529e-01
4.78781536e-02 9.38675478e-02 4.96052176e-01 7.47451663e-01
-1.75477013e-01 -3.27361882e-01 -4.67003614e-01 6.59251884e-02
6.20044589e-01 -2.89403573e-02 -2.53228962e-01 1.96912706e-01
-1.07466352e+00 9.03028011e-01 6.56001866e-01 5.88038325e-01
-2.54401386e-01 2.31273517e-01 4.76432532e-01 4.61250812e-01
5.37267625e-01 8.98050144e-02 -7.76715517e-01 -8.74194652e-02
-2.23627940e-01 -3.78001258e-02 6.02688611e-01 9.62172925e-01
6.87331319e-01 -8.68125781e-02 -1.56184182e-01 9.71303642e-01
2.26898730e-01 5.95856786e-01 8.91918838e-01 -7.10152030e-01
9.85347986e-01 6.89733028e-01 -2.60668576e-01 -8.58023942e-01
-3.39400351e-01 -5.48708618e-01 -8.86459589e-01 -6.45164013e-01
3.02035920e-02 -2.04816699e-01 -8.41692567e-01 2.09629846e+00
3.87171209e-02 1.07021913e-01 4.30877388e-01 7.57642269e-01
9.55167532e-01 9.87219453e-01 1.84717566e-01 -2.11786956e-01
1.71086001e+00 -1.22661436e+00 -9.34208810e-01 -7.10800111e-01
7.58306742e-01 -6.53402269e-01 1.37611496e+00 -1.43868506e-01
-1.01414692e+00 -6.40264869e-01 -8.17249775e-01 -3.72727871e-01
-3.53091747e-01 1.07619740e-01 6.29492104e-01 2.74819911e-01
-9.96367395e-01 3.52086544e-01 -7.63394475e-01 -6.19586289e-01
1.66471422e-01 2.46202871e-01 -2.46758923e-01 -1.23695068e-01
-1.64800322e+00 8.72720420e-01 7.33920455e-01 3.70150506e-01
-3.28378439e-01 -3.76463652e-01 -1.00701523e+00 2.71419078e-01
4.63440031e-01 -9.51637506e-01 1.35687685e+00 -9.57295299e-01
-1.48652017e+00 5.78428030e-01 -8.73861849e-01 -5.66893697e-01
1.93069037e-02 -3.12952518e-01 -2.39458829e-01 8.30792189e-02
2.92236328e-01 6.98825836e-01 5.26514113e-01 -8.84695590e-01
-5.24486125e-01 -4.88370538e-01 2.26612434e-01 3.27068031e-01
-4.44126278e-01 1.15413278e-01 -4.40269172e-01 -7.22004890e-01
3.89099509e-01 -9.66867268e-01 -1.82186678e-01 -5.99804580e-01
-4.52392101e-01 -5.54008842e-01 7.23870933e-01 -7.37086177e-01
1.21160352e+00 -2.08878183e+00 1.87339246e-01 -3.46418679e-01
-2.94731677e-01 2.20026568e-01 -3.79407763e-01 5.92092574e-01
-2.63578385e-01 2.19057634e-01 -4.55460399e-01 -3.78469437e-01
-7.68546760e-02 6.34548128e-01 -6.12036228e-01 -2.50315785e-01
4.06861961e-01 1.33119631e+00 -6.65162027e-01 -3.29744071e-01
-3.74211892e-02 2.39995673e-01 -4.62583095e-01 1.31150251e-02
-3.10778767e-01 4.24129426e-01 -7.72407830e-01 3.59118044e-01
5.97940743e-01 -2.18851760e-01 3.56583714e-01 -1.55465022e-01
6.72169477e-02 9.00245905e-01 -7.32115626e-01 1.95674324e+00
-8.33333373e-01 1.21558055e-01 -7.85110220e-02 -1.28179550e+00
9.53197479e-01 4.78421569e-01 -1.12314597e-01 -9.76887524e-01
1.74902752e-01 2.59465486e-01 -3.60351615e-02 -4.32551652e-01
2.83674419e-01 -4.82671738e-01 -3.24432939e-01 4.27851051e-01
5.12110395e-03 9.29628760e-02 2.62327511e-02 9.98702496e-02
9.21563566e-01 3.11730385e-01 2.99538165e-01 -2.72919834e-01
7.04996705e-01 -9.62802917e-02 8.42447400e-01 4.11113352e-01
1.39693409e-01 4.42843646e-01 4.76440519e-01 -4.14626032e-01
-6.13883674e-01 -7.52835393e-01 4.00795303e-02 1.07583082e+00
1.46748379e-01 -4.54504490e-01 -9.16142583e-01 -7.83434331e-01
-4.44811285e-01 1.08028090e+00 -4.86378759e-01 -2.36198589e-01
-9.70708251e-01 -7.98460305e-01 4.62530702e-01 8.01387727e-01
1.01715171e+00 -1.33382106e+00 -2.62908638e-01 4.30663675e-01
-9.33196723e-01 -1.49166977e+00 -5.82985938e-01 2.09105104e-01
-1.01030421e+00 -5.77494204e-01 -2.94277787e-01 -1.05568552e+00
5.52069783e-01 2.76049197e-01 1.00957870e+00 3.37591320e-02
2.69164950e-01 -1.56114101e-01 -4.25442725e-01 -3.27593744e-01
-3.91338915e-01 3.35313529e-01 -1.56267285e-01 -7.78439045e-02
5.78007102e-01 -6.18118882e-01 -1.48563534e-01 2.62396336e-01
-6.76284194e-01 4.38753098e-01 7.25893736e-01 1.01278734e+00
5.24465263e-01 1.27315400e-02 8.15779448e-01 -1.02995205e+00
5.50322831e-01 -3.90091836e-01 -2.02537298e-01 3.37585390e-01
-4.17467535e-01 3.37942392e-01 9.08575952e-01 -8.15809593e-02
-1.58133435e+00 -2.56321222e-01 -5.16345501e-01 -5.84774725e-02
-9.72039327e-02 7.05111742e-01 -3.99515867e-01 4.92235452e-01
1.74918070e-01 4.38184053e-01 -1.71309456e-01 -7.09208369e-01
2.07277462e-01 6.10696971e-01 2.85500646e-01 -7.38304317e-01
6.20413601e-01 2.05785275e-01 -3.07840973e-01 -4.85591918e-01
-1.22078288e+00 -1.23574674e-01 -5.89238524e-01 4.70599860e-01
9.92537916e-01 -8.77132535e-01 -4.29055274e-01 3.00730377e-01
-1.58394492e+00 -9.06663314e-02 -4.03254963e-02 3.81280065e-01
-4.71454173e-01 3.74040157e-01 -9.79840934e-01 -4.67479020e-01
-6.80033028e-01 -1.05639219e+00 1.00466728e+00 1.61624357e-01
-1.07165620e-01 -1.13040686e+00 -2.61427283e-01 6.95245683e-01
4.25714165e-01 -2.17500806e-01 1.37067807e+00 -6.71512365e-01
-5.37692785e-01 1.12683721e-01 -3.58602762e-01 4.33230370e-01
1.98490769e-01 -6.80041254e-01 -9.72807884e-01 -1.53683767e-01
4.53787327e-01 -2.68610328e-01 1.01259398e+00 2.21651480e-01
9.74450767e-01 -4.28756088e-01 -2.28545085e-01 4.07899410e-01
1.41550350e+00 2.98480719e-01 6.36574745e-01 3.24160695e-01
6.79988205e-01 6.46479726e-01 7.25139976e-01 -2.41037995e-01
6.79367602e-01 5.97363055e-01 2.42980763e-01 -1.70673758e-01
-2.11867929e-01 -5.78630149e-01 7.61604786e-01 1.34603691e+00
1.69397648e-02 -1.93135500e-01 -7.27772415e-01 3.88428807e-01
-2.06117463e+00 -7.29790032e-01 -1.22611955e-01 1.83944547e+00
8.57098162e-01 3.69688898e-01 -4.91040677e-01 -1.78265959e-01
6.84877455e-01 2.02093408e-01 -3.85313451e-01 -7.81330705e-01
-1.10170037e-01 2.77874708e-01 1.46927014e-01 5.99178851e-01
-7.79680550e-01 1.43082678e+00 5.14321852e+00 9.09458578e-01
-1.03306699e+00 4.00748819e-01 6.33905113e-01 3.13059926e-01
-3.44841629e-01 1.40740201e-01 -9.04858828e-01 3.96167517e-01
1.04882228e+00 -1.85367435e-01 3.27485561e-01 5.64154208e-01
3.84910941e-01 1.82275459e-01 -9.66957688e-01 6.81214631e-01
6.53173551e-02 -1.18473494e+00 3.24731916e-01 -2.18610212e-01
2.36325011e-01 1.59047186e-01 -1.50827885e-01 5.86631000e-01
-4.37050359e-03 -8.35630596e-01 6.02463484e-01 8.21696296e-02
4.57033813e-01 -7.07482100e-01 1.11778879e+00 6.08237326e-01
-1.37750971e+00 -9.67010260e-02 -6.00419343e-01 -2.98585415e-01
1.38947591e-01 3.62612039e-01 -7.84933209e-01 9.58409190e-01
6.32854998e-01 7.88905442e-01 -5.11582613e-01 1.19899429e-01
-7.43170381e-01 7.97079265e-01 -1.32957548e-01 -3.65758911e-02
4.24527675e-01 -2.24790096e-01 4.15091515e-01 1.12447464e+00
2.67915517e-01 2.36574337e-01 2.30694801e-01 9.27108526e-01
-1.70121819e-01 3.27458113e-01 -5.18369138e-01 -3.72217931e-02
3.76426339e-01 9.98773456e-01 -6.32998049e-01 -4.77217674e-01
-4.93969232e-01 1.02209044e+00 5.65904140e-01 3.80679637e-01
-8.54667068e-01 -2.62792528e-01 3.96140754e-01 -1.09096684e-01
3.33005458e-01 -1.16503902e-01 -2.73203611e-01 -1.36171079e+00
5.29959142e-01 -8.48268986e-01 4.15310323e-01 -8.71172428e-01
-1.10420275e+00 9.11001086e-01 1.00533720e-02 -8.90266240e-01
-2.55903512e-01 -5.92747152e-01 -7.51323521e-01 9.94572520e-01
-1.66706443e+00 -1.33976173e+00 1.19286567e-01 4.61730212e-01
1.13080835e+00 3.59325446e-02 1.07092285e+00 1.67868719e-01
-7.88875937e-01 5.11535764e-01 -3.61029625e-01 1.97785184e-01
4.88305241e-01 -1.04078913e+00 7.44444013e-01 9.59652007e-01
1.57382190e-01 8.59496713e-01 5.51957190e-01 -5.15669882e-01
-1.33180881e+00 -1.11544871e+00 1.44662642e+00 -3.80444944e-01
6.45822823e-01 -5.01712024e-01 -1.12843943e+00 9.50790942e-01
5.55137813e-01 -4.38738376e-01 3.98881793e-01 2.60097831e-01
-3.20939153e-01 -2.85373721e-02 -8.40225756e-01 5.29314697e-01
1.28100717e+00 -4.78812635e-01 -1.17039299e+00 2.98002034e-01
1.33342302e+00 -2.90427297e-01 -5.22844076e-01 4.51279402e-01
1.42930865e-01 -7.55843878e-01 6.79693341e-01 -4.51968819e-01
5.27951777e-01 -1.95667073e-01 -2.47380212e-01 -1.23135126e+00
-2.72724956e-01 -2.96760380e-01 2.56482422e-01 1.35149622e+00
6.78224027e-01 -1.08224738e+00 5.79390228e-01 3.86545211e-01
-3.65121931e-01 -9.10628676e-01 -1.07583439e+00 -7.20487118e-01
3.03930402e-01 -5.10684311e-01 7.46708095e-01 9.35318530e-01
6.92610666e-02 1.09068930e+00 -1.79086059e-01 8.05529207e-02
3.27956080e-01 4.41067427e-01 3.65238756e-01 -8.42502654e-01
-4.38237906e-01 -1.32028207e-01 9.79900211e-02 -1.39810443e+00
5.83558261e-01 -1.17758298e+00 -1.18249081e-01 -1.86253333e+00
3.79181713e-01 -2.71390378e-01 -3.90864313e-01 9.91516292e-01
-3.65284145e-01 -1.25331119e-01 2.33864993e-01 3.57763730e-02
-2.44342372e-01 8.24475527e-01 1.42194951e+00 -1.78220168e-01
1.89671051e-02 1.40163186e-03 -9.07010555e-01 8.64623845e-01
9.48974609e-01 -6.58569515e-01 -5.09476602e-01 -9.57772672e-01
-3.21427472e-02 1.18477330e-01 1.69232130e-01 -5.00323415e-01
9.61269438e-02 -5.41451164e-02 7.70624587e-03 -6.03854239e-01
3.04936320e-01 -6.69388354e-01 -1.25401363e-01 5.46531439e-01
-2.70097375e-01 4.03284609e-01 2.67826319e-01 3.44938427e-01
-4.23945010e-01 -3.37786257e-01 4.99820143e-01 -4.63771284e-01
-7.85288632e-01 3.93574052e-02 -2.36528337e-01 3.05767469e-02
4.60038602e-01 -4.79423627e-02 -4.09930348e-01 -1.33348137e-01
-5.22591472e-01 3.12299758e-01 1.46511286e-01 9.21985328e-01
5.68190157e-01 -1.23650992e+00 -6.71661079e-01 2.34219581e-01
-1.65234022e-02 4.58036065e-02 2.27455303e-01 7.80615628e-01
-1.37368292e-01 7.98367262e-01 4.74772863e-02 -4.37037796e-01
-1.03329766e+00 4.84524935e-01 2.26675034e-01 -4.24727112e-01
-8.01514328e-01 7.46386945e-01 6.20875716e-01 -7.75892317e-01
-9.54957157e-02 -6.63160086e-01 -1.98984206e-01 -2.49757960e-01
4.65074897e-01 -1.90504670e-01 1.05043545e-01 -5.92121005e-01
-4.21304345e-01 7.28510916e-01 -2.45822087e-01 7.43857911e-03
1.15741634e+00 -2.81655312e-01 -5.18875957e-01 4.26955551e-01
1.19799507e+00 -2.09017470e-01 -5.05167663e-01 -5.60722232e-01
2.56977320e-01 -5.98361678e-02 -1.76771417e-01 -7.59451210e-01
-8.76289308e-01 1.02486289e+00 2.50805587e-01 -1.47256389e-01
1.19338846e+00 1.14546120e-01 1.29782903e+00 7.05748558e-01
6.09831035e-01 -8.87967825e-01 -6.42555729e-02 9.11292315e-01
9.83066142e-01 -1.09998131e+00 -5.47344387e-01 -7.72371292e-01
-5.78827262e-01 9.80357707e-01 7.39832342e-01 8.31236318e-02
3.65511268e-01 1.30455196e-01 2.32222110e-01 -7.19549209e-02
-1.07113218e+00 -3.21229585e-02 -1.23705447e-01 1.44803718e-01
5.45922637e-01 6.48649409e-02 -5.84947169e-01 8.67191017e-01
-3.79318863e-01 -6.43690005e-02 3.57193649e-01 8.61042440e-01
-4.40414369e-01 -1.46266866e+00 -1.45244554e-01 3.49310860e-02
-4.62030500e-01 -4.74593699e-01 -3.62182975e-01 5.70534050e-01
-3.69817652e-02 1.06209016e+00 -1.51750863e-01 -1.91014513e-01
4.72216904e-01 4.23963040e-01 1.79726928e-01 -9.31679368e-01
-5.81771314e-01 1.31712139e-01 3.59779745e-01 -5.46861887e-01
-3.43893915e-01 -4.11301851e-01 -1.79119289e+00 -2.56674946e-04
-2.82559991e-01 5.17071962e-01 5.48391223e-01 1.24424529e+00
4.97059166e-01 6.75542951e-01 6.03258073e-01 -2.59295613e-01
-5.35423279e-01 -1.17525494e+00 -1.76720232e-01 2.79135853e-01
-9.33052227e-03 -4.31330770e-01 -1.89181626e-01 -1.41273424e-01] | [10.678293228149414, 9.2100191116333] |
e0bf30a1-60a4-45bb-8a09-d9b7b4dfaa51 | self-supervised-relative-depth-learning-for | 1712.04850 | null | http://arxiv.org/abs/1712.04850v2 | http://arxiv.org/pdf/1712.04850v2.pdf | Self-Supervised Relative Depth Learning for Urban Scene Understanding | As an agent moves through the world, the apparent motion of scene elements is
(usually) inversely proportional to their depth. It is natural for a learning
agent to associate image patterns with the magnitude of their displacement over
time: as the agent moves, faraway mountains don't move much; nearby trees move
a lot. This natural relationship between the appearance of objects and their
motion is a rich source of information about the world. In this work, we start
by training a deep network, using fully automatic supervision, to predict
relative scene depth from single images. The relative depth training images are
automatically derived from simple videos of cars moving through a scene, using
recent motion segmentation techniques, and no human-provided labels. This proxy
task of predicting relative depth from a single image induces features in the
network that result in large improvements in a set of downstream tasks
including semantic segmentation, joint road segmentation and car detection, and
monocular (absolute) depth estimation, over a network trained from scratch. The
improvement on the semantic segmentation task is greater than those produced by
any other automatically supervised methods. Moreover, for monocular depth
estimation, our unsupervised pre-training method even outperforms supervised
pre-training with ImageNet. In addition, we demonstrate benefits from learning
to predict (unsupervised) relative depth in the specific videos associated with
various downstream tasks. We adapt to the specific scenes in those tasks in an
unsupervised manner to improve performance. In summary, for semantic
segmentation, we present state-of-the-art results among methods that do not use
supervised pre-training, and we even exceed the performance of supervised
ImageNet pre-trained models for monocular depth estimation, achieving results
that are comparable with state-of-the-art methods. | ['Erik Learned-Miller', 'Huaizu Jiang', 'Greg Shakhnarovich', 'Michael Maire', 'Gustav Larsson'] | 2017-12-13 | self-supervised-relative-depth-learning-for-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Huaizu_Jiang_Self-Supervised_Relative_Depth_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Huaizu_Jiang_Self-Supervised_Relative_Depth_ECCV_2018_paper.pdf | eccv-2018-9 | ['road-segementation'] | ['computer-vision'] | [ 4.18050557e-01 1.70510247e-01 -3.58684599e-01 -5.27464926e-01
-5.68991780e-01 -6.43038452e-01 9.45550084e-01 -2.53232956e-01
-7.12159812e-01 5.23570716e-01 -5.72358556e-02 -2.93368958e-02
3.79899293e-01 -9.82755005e-01 -8.29409838e-01 -6.31655455e-01
-4.64258641e-02 7.87272215e-01 7.45867431e-01 -5.11343079e-03
3.00168902e-01 6.13718271e-01 -1.59010231e+00 1.32121220e-01
5.32254875e-01 8.00052881e-01 5.39934397e-01 1.02845895e+00
-2.41000000e-02 1.20661247e+00 -3.04662615e-01 -3.44487466e-02
3.44556838e-01 -2.90624619e-01 -1.21164870e+00 5.42190194e-01
8.02437305e-01 -8.35667968e-01 -5.74961066e-01 8.26113045e-01
-1.57016978e-01 1.98835343e-01 8.04563820e-01 -1.08348370e+00
-9.59565267e-02 2.45425433e-01 -5.98986387e-01 2.77936816e-01
2.48554692e-01 4.02024627e-01 9.12482977e-01 -6.57428205e-01
9.77703035e-01 1.30649638e+00 5.14731467e-01 4.66188997e-01
-1.13500524e+00 -1.02731705e-01 4.98794913e-01 1.66493565e-01
-1.04793620e+00 -3.58582050e-01 7.92367935e-01 -5.67276597e-01
1.13079274e+00 -1.71187252e-01 6.52402341e-01 6.02986336e-01
-1.06191315e-01 1.09473908e+00 7.11862624e-01 -3.16187024e-01
2.71153092e-01 -1.93312671e-02 -2.56612271e-01 7.53982484e-01
-2.07932182e-02 3.18834126e-01 -2.06929207e-01 4.81199354e-01
9.98988569e-01 -3.92012186e-02 -4.20754515e-02 -6.75246596e-01
-1.27014923e+00 7.78600097e-01 6.91399157e-01 2.96067506e-01
-4.94149774e-02 6.34881020e-01 1.58937663e-01 2.85249110e-02
6.63402438e-01 2.99721032e-01 -6.25722528e-01 -1.12924419e-01
-1.07428277e+00 2.33905807e-01 6.26136184e-01 8.98283839e-01
1.34402514e+00 4.56102155e-02 4.91088390e-01 5.35937786e-01
1.80371627e-01 4.65223193e-01 3.33105326e-01 -1.66924536e+00
4.24176157e-01 4.87345427e-01 8.72475803e-02 -8.88603091e-01
-5.86186647e-01 -1.65777370e-01 -5.51636755e-01 5.31415105e-01
1.00609958e+00 -1.72709793e-01 -1.18292904e+00 1.80615389e+00
4.23686385e-01 1.89428911e-01 1.61604077e-01 8.54310393e-01
5.78688085e-01 8.17497909e-01 -6.60106689e-02 5.79327606e-02
9.94453728e-01 -1.32751870e+00 -2.07721725e-01 -9.09324169e-01
1.09452069e+00 -4.34029728e-01 9.59439695e-01 4.09494996e-01
-1.04802442e+00 -7.03782916e-01 -9.53857243e-01 -2.21183002e-01
-4.45013493e-01 -6.18880056e-03 8.47407341e-01 3.75973165e-01
-1.24616098e+00 7.16311753e-01 -9.34202075e-01 -4.16476667e-01
5.18207192e-01 2.33220235e-01 -3.83959562e-01 -2.00814396e-01
-8.03146958e-01 8.27369750e-01 3.72024059e-01 -6.74430504e-02
-1.12641037e+00 -5.66112518e-01 -1.19574404e+00 -3.95493209e-01
4.14370894e-01 -7.79300392e-01 1.53777158e+00 -1.35856318e+00
-1.39898491e+00 1.13802552e+00 -2.48163059e-01 -6.06892645e-01
6.77176952e-01 -1.57917857e-01 2.50267148e-01 4.99555498e-01
3.82561475e-01 1.61047649e+00 5.62431574e-01 -1.50285828e+00
-1.11837959e+00 -3.58948201e-01 2.85810918e-01 5.28364837e-01
2.58841127e-01 -4.22054142e-01 -7.07513690e-01 -1.66766763e-01
5.50916716e-02 -9.54958856e-01 -5.45807302e-01 1.34693161e-01
-3.00612867e-01 -8.69537070e-02 8.84852767e-01 -3.43129098e-01
5.32719910e-01 -1.96415901e+00 1.85285304e-02 -2.21332356e-01
1.07663788e-01 -1.85692951e-01 -1.37248456e-01 -5.64599559e-02
1.58232316e-01 -2.47678041e-01 -5.79763472e-01 -5.26285529e-01
-2.38911673e-01 4.36511189e-01 7.38782436e-03 6.92295671e-01
9.58664939e-02 1.01617992e+00 -1.18147206e+00 -5.04226804e-01
5.60153902e-01 1.50273398e-01 -5.63178778e-01 7.05543011e-02
-6.17909849e-01 5.56298256e-01 -2.90531784e-01 4.28044796e-01
6.03354454e-01 -3.48035127e-01 1.05394144e-02 7.09950477e-02
-2.24464417e-01 3.09295982e-01 -1.02211761e+00 1.88007295e+00
-4.87821639e-01 1.07641149e+00 -1.86927989e-02 -1.19830489e+00
4.94382739e-01 1.60432123e-02 6.49901986e-01 -6.75957918e-01
3.64995301e-02 1.01430178e-01 -5.26906922e-02 -3.11853528e-01
5.07875979e-01 1.50533944e-01 -3.52590345e-02 3.73367429e-01
-5.34502603e-02 -8.84809375e-01 6.07324302e-01 3.10171396e-01
9.38181818e-01 3.23423296e-01 8.80081356e-02 -2.65354142e-02
4.25960213e-01 4.84564066e-01 3.28053653e-01 6.68430686e-01
-1.81789383e-01 6.99444950e-01 5.12105048e-01 -5.75186133e-01
-1.20828271e+00 -9.87264276e-01 -7.92784467e-02 9.78220344e-01
5.51350951e-01 3.87056395e-02 -9.85913992e-01 -6.23457491e-01
-1.07016303e-01 4.30850029e-01 -6.31549716e-01 1.71513617e-01
-8.50715756e-01 -5.70345461e-01 3.10503989e-01 8.46125245e-01
8.34017992e-01 -1.10327959e+00 -8.14484417e-01 2.11444348e-01
-1.59602031e-01 -1.82095563e+00 -3.81864697e-01 2.43897721e-01
-1.06701219e+00 -1.03294456e+00 -7.13428319e-01 -9.46442723e-01
6.73652232e-01 4.49272186e-01 1.25478327e+00 -1.32738009e-01
-3.28281552e-01 4.93657798e-01 -2.69998200e-02 -1.03806533e-01
-3.15796524e-01 1.81827798e-01 -3.74517798e-01 -2.35656083e-01
1.90501764e-01 -5.10099173e-01 -8.63830626e-01 2.94324458e-01
-7.57516921e-01 2.98588097e-01 5.31531215e-01 4.03131217e-01
5.45732975e-01 2.51784027e-02 1.01916559e-01 -9.90426123e-01
-3.18016738e-01 -2.19837472e-01 -8.26582372e-01 -4.33735490e-01
-3.05158585e-01 4.81887162e-02 4.50165868e-01 -2.03432873e-01
-1.16024327e+00 6.12024844e-01 -1.19457766e-01 -1.37918442e-01
-7.29366004e-01 1.20538753e-02 4.53783572e-02 9.12435651e-02
6.10747397e-01 1.07376695e-01 -1.39935791e-01 -1.19858481e-01
6.95389628e-01 2.25744918e-01 7.86283135e-01 -1.39069140e-01
8.14562321e-01 1.05008733e+00 4.89293374e-02 -9.75289345e-01
-1.00164163e+00 -6.83835626e-01 -1.06753623e+00 -2.12890670e-01
1.36674654e+00 -1.17599297e+00 -4.32858855e-01 8.03573370e-01
-1.27947605e+00 -1.21194279e+00 -2.95205981e-01 4.27862346e-01
-9.35363829e-01 3.75362456e-01 -8.24318767e-01 -5.43970466e-01
2.76594430e-01 -1.02509761e+00 1.22925544e+00 1.82774197e-02
-1.28349811e-01 -1.38114393e+00 4.62257192e-02 5.73048413e-01
1.30488332e-02 1.65119663e-01 6.58303738e-01 -1.06094085e-01
-1.22029710e+00 1.95621848e-01 -3.89940262e-01 3.62376213e-01
1.51445970e-01 -3.21158469e-02 -1.12116611e+00 1.03022560e-01
-2.30644971e-01 -1.69934392e-01 1.15899324e+00 8.72373104e-01
9.54375684e-01 3.30624841e-02 -5.14816403e-01 7.18769848e-01
1.38436151e+00 2.04979166e-01 6.42534316e-01 5.94105244e-01
1.15966809e+00 8.54975164e-01 7.73299754e-01 4.85107563e-02
5.76683044e-01 6.94895029e-01 6.66035235e-01 -4.39155877e-01
-2.63073683e-01 -1.75968066e-01 4.16599602e-01 1.21017806e-01
7.40751578e-03 -2.00918600e-01 -9.52783048e-01 7.98592448e-01
-1.79903615e+00 -9.95263398e-01 -3.22368741e-01 1.86319971e+00
7.23021030e-01 3.83672625e-01 3.59866321e-01 -2.14184634e-03
3.41651380e-01 3.15837979e-01 -7.46613443e-01 -2.14943722e-01
-1.64935306e-01 -2.50390410e-01 1.05517805e+00 9.33598340e-01
-1.38948381e+00 1.47783792e+00 6.67506838e+00 4.71303612e-01
-1.15803921e+00 2.94459295e-02 1.09012377e+00 2.58847047e-02
-1.93119377e-01 1.04866095e-01 -7.85708725e-01 2.08859131e-01
6.95990562e-01 3.15523654e-01 3.58460456e-01 1.01641285e+00
3.67735684e-01 -6.78148150e-01 -1.41711152e+00 8.68862450e-01
1.61827169e-02 -1.39018381e+00 -1.95759192e-01 1.33942857e-01
1.26937091e+00 4.96671557e-01 4.78868559e-03 -3.05921603e-02
5.43081045e-01 -1.06551826e+00 7.76653111e-01 2.22203761e-01
5.42419195e-01 -6.37092173e-01 5.53079247e-01 5.61262608e-01
-1.30515146e+00 -1.78068951e-02 -1.85037762e-01 -3.34105432e-01
4.40961182e-01 6.91171348e-01 -8.35732520e-01 8.86322111e-02
5.83512783e-01 1.17089558e+00 -5.58396041e-01 6.17998838e-01
-3.35551500e-01 1.57222643e-01 -4.15198892e-01 2.77171969e-01
7.37070739e-01 -2.74534285e-01 1.97790891e-01 1.21478498e+00
-6.59074336e-02 -2.93470523e-03 2.85646737e-01 8.91249001e-01
-5.67806372e-03 -2.70036280e-01 -9.54255760e-01 3.39414924e-01
1.47697344e-01 1.10937762e+00 -1.12703359e+00 -5.77485561e-01
-5.38563967e-01 1.03895366e+00 1.50923714e-01 5.03000975e-01
-6.78952336e-01 -3.62414867e-02 6.91596568e-01 3.70152861e-01
4.70053881e-01 -4.39180940e-01 -3.76398027e-01 -8.92359853e-01
-2.23414332e-01 -3.87188673e-01 6.04396015e-02 -1.00856960e+00
-8.05618525e-01 3.48078460e-01 2.34901920e-01 -1.09016728e+00
-7.40043700e-01 -7.51690209e-01 -5.35030901e-01 4.11916167e-01
-1.73834717e+00 -9.19711888e-01 -4.26018745e-01 4.34797525e-01
9.54819083e-01 2.55944908e-01 1.68004408e-01 1.66935712e-01
-1.11876242e-01 -2.40027867e-02 -1.47413418e-01 3.17194492e-01
3.97715122e-01 -1.48225570e+00 6.81890786e-01 9.94310677e-01
2.84448773e-01 -8.32606107e-02 5.33204675e-01 -5.02427459e-01
-9.02353823e-01 -1.13569081e+00 6.45691872e-01 -8.02442789e-01
5.13524294e-01 -2.56079227e-01 -5.18643796e-01 8.76679063e-01
6.57587945e-02 7.67384320e-02 -1.36891976e-01 -3.03665519e-01
-7.27697602e-03 -6.35597259e-02 -8.57795477e-01 6.05517745e-01
1.27440560e+00 -5.63732147e-01 -2.92616993e-01 4.16068286e-01
6.68683887e-01 -4.62508261e-01 -3.28519523e-01 3.21955413e-01
2.16474682e-01 -1.22637618e+00 1.21665144e+00 -3.47313583e-02
7.10974574e-01 -3.27198625e-01 -6.41813427e-02 -1.17169094e+00
3.34794745e-02 -3.22098851e-01 1.34607852e-01 6.97480679e-01
4.59814966e-01 -3.30289721e-01 1.44475210e+00 4.12608862e-01
-2.05119401e-01 -4.93361771e-01 -6.58550084e-01 -6.11904442e-01
2.52458483e-01 -5.86416900e-01 6.11144379e-02 7.67494977e-01
-4.12499785e-01 4.61607158e-01 1.89849716e-02 1.83748186e-01
7.75159538e-01 5.66778630e-02 1.08556437e+00 -1.20547783e+00
-2.82837361e-01 -4.80565846e-01 -6.57133877e-01 -1.95364892e+00
3.00766170e-01 -6.92398071e-01 3.28253806e-01 -1.74601293e+00
1.71465397e-01 -3.60090613e-01 4.09365445e-01 2.43120849e-01
7.97440559e-02 4.11093354e-01 5.18457368e-02 1.78394467e-01
-6.72974229e-01 2.20328674e-01 1.60141861e+00 -1.82272300e-01
-3.29227090e-01 -6.51364550e-02 -2.36464038e-01 1.27040422e+00
6.85715437e-01 -3.30505908e-01 -7.04218984e-01 -6.24872804e-01
5.80203533e-02 6.29874840e-02 6.02593422e-01 -9.74660277e-01
2.33505622e-01 -2.08387122e-01 4.04576749e-01 -7.57154942e-01
6.11804307e-01 -6.62348330e-01 -2.42666304e-01 4.45839137e-01
-1.66607827e-01 -1.82087258e-01 3.12626809e-02 5.33196449e-01
-2.23533228e-01 -2.88410991e-01 9.30578709e-01 -5.55370629e-01
-1.27187455e+00 3.59122783e-01 -7.77282655e-01 2.62222767e-01
9.92873609e-01 -7.23452330e-01 -2.27569118e-01 -6.53697133e-01
-7.68293858e-01 1.35488585e-01 7.26086318e-01 1.25589982e-01
5.44098258e-01 -9.70581591e-01 -3.99687409e-01 5.13730422e-02
-2.70363353e-02 6.49767280e-01 9.68520865e-02 7.38698304e-01
-1.02986419e+00 5.77953279e-01 -1.10690914e-01 -1.11580896e+00
-1.02737856e+00 3.89836490e-01 6.03330731e-01 -1.67720407e-01
-7.63681948e-01 8.25407863e-01 8.44356120e-01 -5.10286868e-01
1.56653255e-01 -7.12253451e-01 -5.85150085e-02 -1.21262036e-01
2.68125772e-01 2.04858944e-01 -1.80175349e-01 -6.05903149e-01
-2.12540030e-01 9.44890022e-01 -2.53629107e-02 -4.55293953e-01
1.27176559e+00 -4.27960426e-01 2.09422141e-01 4.78953362e-01
1.32669103e+00 -2.63827562e-01 -2.04340577e+00 -1.64597735e-01
-1.56479314e-01 -5.04453957e-01 2.49162063e-01 -4.47069019e-01
-1.32491255e+00 9.45118487e-01 3.24366719e-01 8.38814452e-02
8.46307039e-01 3.72333050e-01 7.57411301e-01 4.55192387e-01
3.71844918e-01 -1.17028761e+00 4.74041373e-01 7.43148565e-01
1.91714451e-01 -1.51033080e+00 -1.03110850e-01 -6.36087775e-01
-6.88856721e-01 1.02010465e+00 6.93494618e-01 -2.32147485e-01
5.26707053e-01 3.16949517e-01 2.36562163e-01 -1.89654306e-01
-6.23567700e-01 -5.37847996e-01 4.92801629e-02 7.80858636e-01
2.60110766e-01 -2.60265350e-01 4.61739033e-01 -3.42079997e-01
-1.68414474e-01 -2.01406419e-01 6.90243602e-01 8.07596087e-01
-7.51973689e-01 -8.84390593e-01 -1.45052806e-01 6.45761192e-02
-6.35238662e-02 -3.58479284e-02 -3.90697241e-01 1.06905842e+00
2.55039632e-01 9.63671982e-01 5.68840563e-01 1.00609772e-01
1.34513840e-01 -4.09264237e-01 6.91188216e-01 -7.19830513e-01
1.08799353e-01 1.10892924e-02 2.15760008e-01 -6.86642468e-01
-7.94604897e-01 -8.15375507e-01 -1.59313226e+00 -3.18021894e-01
-1.30597390e-02 -3.64922881e-01 4.85792667e-01 1.27264690e+00
-1.33130923e-01 3.95142704e-01 5.83604515e-01 -1.49558115e+00
1.49694756e-01 -5.93715668e-01 -4.50971574e-01 5.94927669e-01
4.12171900e-01 -5.69239080e-01 -5.63029289e-01 5.08522272e-01] | [8.574723243713379, -1.7208459377288818] |
a473d871-b07f-48f1-9dbf-63333754ef75 | rats-nas-redirection-of-adjacent-trails-on | 2305.04206 | null | https://arxiv.org/abs/2305.04206v2 | https://arxiv.org/pdf/2305.04206v2.pdf | RATs-NAS: Redirection of Adjacent Trails on GCN for Neural Architecture Search | Various hand-designed CNN architectures have been developed, such as VGG, ResNet, DenseNet, etc., and achieve State-of-the-Art (SoTA) levels on different tasks. Neural Architecture Search (NAS) now focuses on automatically finding the best CNN architecture to handle the above tasks. However, the verification of a searched architecture is very time-consuming and makes predictor-based methods become an essential and important branch of NAS. Two commonly used techniques to build predictors are graph-convolution networks (GCN) and multilayer perceptron (MLP). In this paper, we consider the difference between GCN and MLP on adjacent operation trails and then propose the Redirected Adjacent Trails NAS (RATs-NAS) to quickly search for the desired neural network architecture. The RATs-NAS consists of two components: the Redirected Adjacent Trails GCN (RATs-GCN) and the Predictor-based Search Space Sampling (P3S) module. RATs-GCN can change trails and their strengths to search for a better neural network architecture. P3S can rapidly focus on tighter intervals of FLOPs in the search space. Based on our observations on cell-based NAS, we believe that architectures with similar FLOPs will perform similarly. Finally, the RATs-NAS consisting of RATs-GCN and P3S beats WeakNAS, Arch-Graph, and others by a significant margin on three sub-datasets of NASBench-201. | ['Kuo-Chin Fan', 'Chun-Chieh Lee', 'Jun-Wei Hsieh', 'Yu-Ming Zhang'] | 2023-05-07 | null | null | null | null | ['architecture-search'] | ['methodology'] | [-2.62844086e-01 -2.21172825e-01 -1.63782146e-02 -1.65304810e-01
-1.20561734e-01 -2.93932080e-01 3.40006173e-01 -1.22322209e-01
-5.21636248e-01 5.94621658e-01 -1.69821545e-01 -6.21574163e-01
-1.36556819e-01 -8.59156787e-01 -7.88444519e-01 -5.73359787e-01
-2.13010967e-01 2.60860384e-01 6.85819864e-01 -3.36298496e-01
4.30958629e-01 6.84120834e-01 -1.58156419e+00 4.35206234e-01
5.08057058e-01 1.30088460e+00 4.23928022e-01 6.57904148e-01
-1.26123950e-01 7.73297250e-01 -2.46101916e-01 -1.10115498e-01
5.12172461e-01 -2.13750660e-01 -8.37119281e-01 -8.32405567e-01
2.17797726e-01 1.56516403e-01 -3.70494008e-01 1.01204967e+00
5.99777281e-01 2.92016089e-01 3.77101988e-01 -1.29679883e+00
-2.50100702e-01 8.60489666e-01 -2.83120006e-01 6.67106569e-01
-2.23572701e-01 3.62650126e-01 9.05035615e-01 -1.16069031e+00
5.50053716e-01 1.21879637e+00 8.55009079e-01 5.62488377e-01
-9.96024132e-01 -9.84352946e-01 2.21517935e-01 3.89645636e-01
-1.61960077e+00 -4.48838562e-01 6.55256927e-01 -3.01710159e-01
1.43036830e+00 2.80064404e-01 9.04645920e-01 9.63263750e-01
4.07957256e-01 6.06853902e-01 8.64273965e-01 -3.65258217e-01
5.87117493e-01 -1.06052063e-01 2.07444936e-01 8.77660036e-01
1.83121219e-01 1.82373360e-01 -6.19107068e-01 -1.50739979e-02
1.05379772e+00 -9.51457545e-02 -3.51161063e-01 2.17773259e-01
-1.15027857e+00 5.41087568e-01 1.15813577e+00 3.78061086e-01
-6.04943991e-01 4.43796217e-01 3.62400979e-01 4.08641368e-01
-1.75798535e-02 9.24705088e-01 -7.40689456e-01 3.70131023e-02
-9.93958056e-01 1.66238070e-01 8.06874990e-01 9.88868237e-01
8.04723024e-01 1.57994211e-01 -1.86563492e-01 8.17761898e-01
3.11085880e-01 -1.43518820e-01 6.81161046e-01 -4.88736451e-01
3.42882842e-01 9.82318401e-01 -6.43858492e-01 -9.70570624e-01
-5.95387161e-01 -8.58015895e-01 -1.20381010e+00 1.37451127e-01
1.74875870e-01 -1.43737689e-01 -1.07791150e+00 1.47232449e+00
8.37423876e-02 3.34725767e-01 -2.91851074e-01 8.16860020e-01
1.05072773e+00 6.97542250e-01 2.83051804e-02 2.39798278e-01
1.25316858e+00 -1.24533200e+00 1.35468498e-01 -5.51614881e-01
6.35195374e-01 -4.39606428e-01 1.00439143e+00 2.64851183e-01
-1.02195835e+00 -6.62277460e-01 -1.12570655e+00 9.84603390e-02
-4.47396666e-01 2.17980444e-01 4.43111062e-01 6.48914725e-02
-1.77227890e+00 7.94468522e-01 -7.68239021e-01 -4.30353731e-01
4.97600019e-01 7.30542719e-01 -1.23887777e-01 1.86968461e-01
-1.10478473e+00 8.04802239e-01 5.70428431e-01 3.97743583e-01
-9.45364535e-01 -7.20171094e-01 -4.45725650e-01 4.07021105e-01
2.19565675e-01 -7.25006938e-01 1.15864336e+00 -9.84056234e-01
-1.46892893e+00 5.63733101e-01 -8.65715593e-02 -7.21345961e-01
1.19684942e-01 3.53160441e-01 -5.37247956e-01 -2.70264089e-01
-2.19130829e-01 9.40744460e-01 5.81132352e-01 -7.56416321e-01
-6.65465772e-01 -1.81573927e-01 -1.52964637e-01 1.29892314e-02
-3.47016007e-01 1.60734095e-02 -6.06586337e-01 -5.49421251e-01
3.83870870e-01 -1.05252111e+00 -4.42892730e-01 -1.34181008e-01
-7.11252689e-01 -3.07746142e-01 5.23550391e-01 -1.82501748e-01
1.53277242e+00 -2.02105045e+00 1.01684742e-02 4.37787116e-01
3.50372523e-01 5.40475309e-01 -3.46649319e-01 1.17722735e-01
-1.71604633e-01 1.89822078e-01 2.28390440e-01 -3.13647687e-01
-4.21267420e-01 2.90212445e-02 -6.65054545e-02 1.31628022e-01
1.86811775e-01 1.05340517e+00 -6.47933722e-01 -3.62666041e-01
-2.83669859e-01 2.25898460e-01 -6.00932181e-01 -8.28863904e-02
-1.98367372e-01 6.55883253e-02 -4.93756771e-01 9.55004454e-01
2.92655855e-01 -5.96517324e-01 2.25916412e-02 -2.62732983e-01
-3.99736732e-01 4.36376601e-01 -9.37556863e-01 1.27695310e+00
-3.21479678e-01 8.23312938e-01 -5.20963222e-02 -8.16899776e-01
1.38287520e+00 2.98468620e-02 -8.40598494e-02 -7.70057857e-01
3.88328731e-01 6.57460690e-01 3.86004299e-01 -7.91086480e-02
3.86319995e-01 4.35538292e-01 3.91289234e-01 9.00179669e-02
2.04818502e-01 5.37922323e-01 1.26746327e-01 -1.56632662e-01
1.54044628e+00 -3.93689513e-01 1.98391899e-01 -6.14041150e-01
6.14351511e-01 5.87402284e-03 6.92224383e-01 6.89035773e-01
-2.52803564e-01 5.22641480e-01 5.42039931e-01 -8.11917007e-01
-9.50345337e-01 -5.43960810e-01 9.67024192e-02 1.08100414e+00
-1.14191085e-01 -3.21456134e-01 -5.89574993e-01 -5.70166111e-01
-2.47736603e-01 6.49780273e-01 -6.82773650e-01 -1.25445262e-01
-8.13355923e-01 -5.21400928e-01 8.02944303e-01 7.49881327e-01
9.12449837e-01 -1.60309958e+00 -7.28228211e-01 3.10510278e-01
4.29371595e-01 -7.06624508e-01 -3.40814203e-01 6.60833061e-01
-1.11060798e+00 -9.30420101e-01 -6.95873260e-01 -1.19390440e+00
9.00002778e-01 -5.70725016e-02 1.21071649e+00 4.88781780e-01
1.65666882e-02 -4.91335362e-01 -2.13693678e-01 -3.63604844e-01
-2.57384419e-01 5.96142292e-01 -1.41437486e-01 -1.50215015e-01
1.25131205e-01 -7.58502901e-01 -7.04809964e-01 4.51310903e-01
-4.35298800e-01 3.15061778e-01 8.73245478e-01 8.19452405e-01
8.62777472e-01 -8.44920650e-02 5.36868095e-01 -8.12384009e-01
6.89256132e-01 -4.17224735e-01 -8.76705706e-01 2.12492183e-01
-1.04198015e+00 2.72665679e-01 8.82645011e-01 -4.17512625e-01
-3.17190409e-01 6.88912794e-02 -2.40253970e-01 -9.25895333e-01
2.64840554e-02 8.54878187e-01 7.92526752e-02 -3.42137307e-01
8.78628433e-01 3.84396136e-01 -7.53498524e-02 -5.25310278e-01
-2.15782657e-01 1.87327147e-01 4.49511886e-01 -1.02314860e-01
4.37056780e-01 -4.30716462e-02 1.58283547e-01 -3.74846071e-01
-2.82809854e-01 -3.26651186e-01 -3.22556227e-01 -2.31026769e-01
5.66684008e-01 -7.39663005e-01 -6.77099347e-01 5.05845487e-01
-1.25678289e+00 -7.07975030e-01 -5.13963923e-02 2.51247227e-01
9.18567739e-03 -3.85468423e-01 -6.69222534e-01 -4.04255748e-01
-8.80506635e-01 -1.45735729e+00 4.80818838e-01 5.18920958e-01
-1.24062963e-01 -9.69345868e-01 -2.84560379e-02 -3.77732754e-01
8.14253986e-01 -2.80715562e-02 1.11401200e+00 -9.26256597e-01
-7.90866852e-01 -9.66553092e-02 -3.94874662e-01 2.61877298e-01
-3.07543904e-01 1.10681079e-01 -8.54553819e-01 -3.75792891e-01
-3.59127760e-01 -4.48992029e-02 1.11999261e+00 5.57000399e-01
1.42727101e+00 -4.45870906e-01 -6.19841576e-01 9.85134184e-01
1.55566728e+00 5.91753662e-01 8.82750154e-01 7.72310078e-01
5.66610396e-01 -6.70022592e-02 1.60762474e-01 7.42656738e-02
2.63613582e-01 4.39356148e-01 7.07127929e-01 -6.87476993e-02
-2.89225161e-01 -1.76542357e-01 2.43768185e-01 1.00021112e+00
-2.51344204e-01 -4.26512867e-01 -1.28314233e+00 4.82766032e-01
-1.79614091e+00 -5.75470030e-01 2.14376319e-02 1.86028504e+00
6.78940117e-01 3.68516773e-01 -1.66191354e-01 9.12940204e-02
6.42701626e-01 6.00569732e-02 -6.48058951e-01 -5.75875401e-01
-2.29378361e-02 2.73806512e-01 7.45038092e-01 7.21535087e-02
-7.41956294e-01 8.62825751e-01 5.95336342e+00 9.21203434e-01
-1.49648345e+00 -1.72469497e-01 9.81777251e-01 -3.39810438e-02
-1.31058171e-01 6.37844428e-02 -1.32454693e+00 4.89746183e-01
8.41125071e-01 1.34126440e-01 6.60835922e-01 1.13041270e+00
-1.37740597e-01 5.51315695e-02 -1.23377275e+00 1.05975354e+00
-3.27513069e-01 -2.09003901e+00 1.33604527e-01 4.77706566e-02
7.43266940e-01 6.24541521e-01 -1.20732211e-01 5.54328561e-01
2.41368800e-01 -1.16891229e+00 9.08927560e-01 5.19869745e-01
7.79349387e-01 -5.01434386e-01 8.32079768e-01 3.63680065e-01
-1.42082620e+00 -2.88595438e-01 -5.20273805e-01 7.68711120e-02
-3.54795963e-01 5.63653171e-01 -1.18299294e+00 2.38869429e-01
9.06352043e-01 6.22867286e-01 -7.80938506e-01 1.51536560e+00
-1.50775522e-01 7.60685623e-01 -3.26476127e-01 -7.32874155e-01
4.82305110e-01 1.78246036e-01 5.02759337e-01 1.15051734e+00
4.97998714e-01 -9.11519527e-02 -1.43022344e-01 1.04913795e+00
-3.73328626e-01 -1.35190606e-01 -5.55826798e-02 1.44758746e-01
9.01299953e-01 1.27725101e+00 -1.00229979e+00 -2.32253090e-01
-9.20410976e-02 4.67710257e-01 5.17507315e-01 2.98755556e-01
-6.74956203e-01 -3.20601851e-01 5.09788513e-01 3.09557587e-01
3.75379860e-01 1.12567477e-01 -4.82910514e-01 -5.94012976e-01
-2.60469228e-01 -1.03152966e+00 4.67110574e-01 -7.35831499e-01
-9.83956099e-01 1.26892471e+00 -4.22481179e-01 -1.15317798e+00
3.74821499e-02 -6.48419619e-01 -9.84307647e-01 9.55641508e-01
-1.47406328e+00 -6.84512556e-01 -5.27829170e-01 7.71069348e-01
7.20109522e-01 -5.31275630e-01 7.27604806e-01 3.44239622e-01
-9.34268296e-01 7.76193380e-01 -1.14386134e-01 1.87668338e-01
2.05781773e-01 -8.29733133e-01 7.54216135e-01 7.14058280e-01
1.33700833e-01 6.23265207e-01 3.27969134e-01 -5.73724866e-01
-1.36707580e+00 -1.34575069e+00 8.41862321e-01 2.98421949e-01
5.28052270e-01 -2.04307288e-01 -1.01080692e+00 5.98886013e-01
-8.88606608e-02 3.94675940e-01 -3.66417095e-02 1.35321930e-01
-2.03733802e-01 -3.98039579e-01 -7.56588697e-01 7.83054769e-01
1.10136390e+00 -1.50381416e-01 4.16350402e-02 9.22827944e-02
6.80266082e-01 -5.67672193e-01 -4.17306542e-01 4.27914381e-01
3.84018987e-01 -9.08001363e-01 7.64808893e-01 -4.08508360e-01
3.63718331e-01 -3.34559321e-01 -4.44988348e-02 -1.51188540e+00
-6.82487130e-01 -5.74702024e-01 -8.63442197e-02 7.89652526e-01
9.38400626e-01 -8.33411396e-01 1.02975988e+00 3.29394311e-01
-5.53886473e-01 -1.41764760e+00 -9.93713200e-01 -8.69824231e-01
-2.04939127e-01 -3.80940676e-01 8.36076736e-01 8.21727812e-01
-3.08191210e-01 3.74532223e-01 1.05000764e-01 1.07145607e-01
1.45103633e-01 -1.41050637e-01 5.09860098e-01 -1.28455281e+00
-3.33609730e-01 -9.23322082e-01 -4.89976108e-01 -9.78161633e-01
-2.89316744e-01 -9.80322361e-01 1.42877385e-01 -1.38309133e+00
-2.06244409e-01 -9.58537102e-01 -6.99251890e-01 8.08897197e-01
1.28527790e-01 -9.34075937e-03 1.24701060e-01 4.81693387e-01
-4.75354671e-01 2.79909164e-01 1.20695019e+00 -9.91278067e-02
-4.91517961e-01 3.84919569e-02 -5.33616304e-01 8.35570455e-01
9.02453899e-01 -5.11225641e-01 -4.34522331e-01 -5.13510406e-01
5.12628138e-01 -1.93816289e-01 5.11982858e-01 -1.41010666e+00
9.53270018e-01 5.62422015e-02 4.24114406e-01 -7.14963675e-01
7.76998177e-02 -3.89170110e-01 3.84503573e-01 7.68225849e-01
-4.46510136e-01 6.00912452e-01 3.84746552e-01 2.47826055e-01
-2.88123727e-01 -1.76871732e-01 7.26557553e-01 -2.51817465e-01
-9.89413977e-01 5.30014098e-01 -2.83413082e-01 -1.63425699e-01
7.04963624e-01 -4.51915741e-01 -5.35882056e-01 1.99999418e-02
-4.89067405e-01 3.84402066e-01 -9.64528136e-03 2.25402787e-01
9.40383255e-01 -1.07793248e+00 -4.74739313e-01 4.20118153e-01
-7.04577416e-02 3.01040918e-01 -1.00222729e-01 8.86471748e-01
-8.24900746e-01 6.12941146e-01 -2.90240884e-01 -5.40152609e-01
-1.10266101e+00 3.36656511e-01 6.07483149e-01 -5.15345156e-01
-4.21371907e-01 1.53393245e+00 -2.64796615e-02 -2.72272944e-01
3.68454158e-01 -6.39176309e-01 -2.71921068e-01 -1.95303395e-01
3.53453517e-01 3.15221369e-01 4.25086856e-01 -1.81748256e-01
-6.29122615e-01 8.81326571e-02 -2.06574172e-01 3.19772154e-01
1.45754433e+00 4.07907695e-01 -2.26244062e-01 5.58524057e-02
1.11195505e+00 -5.99356532e-01 -1.08219457e+00 -2.77051538e-01
2.00384129e-02 3.84758860e-02 4.29090470e-01 -8.52582872e-01
-1.50600529e+00 6.43585622e-01 4.69182402e-01 1.10315561e-01
1.40900218e+00 -5.68783879e-02 7.06745982e-01 5.13436735e-01
6.04745746e-01 -8.69833052e-01 3.66659388e-02 9.13402855e-01
1.09865379e+00 -6.84697628e-01 -1.31469235e-01 -2.77072459e-01
-2.64564097e-01 1.31154644e+00 1.12524819e+00 -2.71572381e-01
9.36274588e-01 3.56920570e-01 -2.73229122e-01 -4.72882628e-01
-1.22120702e+00 1.74334899e-01 4.04994160e-01 1.32953122e-01
1.58211142e-01 -6.20503128e-02 1.07391536e-01 7.21506774e-01
-3.75908524e-01 2.81390827e-02 7.52541050e-02 7.90293038e-01
-4.13686424e-01 -7.36212254e-01 -1.26801610e-01 7.88807929e-01
-1.39517356e-02 -6.05382860e-01 -2.64076918e-01 7.59427965e-01
1.23100705e-01 4.46244717e-01 2.37642378e-01 -1.00454247e+00
3.42373937e-01 -7.33227655e-02 8.58330652e-02 -4.58225071e-01
-1.12178183e+00 -7.33271167e-02 6.18600734e-02 -7.43768215e-01
1.29727781e-01 -3.27707171e-01 -1.31739414e+00 -5.28637886e-01
-5.16548276e-01 -1.93364739e-01 6.07732594e-01 6.97751760e-01
7.04490602e-01 7.70880461e-01 3.50718260e-01 -8.48738253e-01
-3.16266268e-01 -8.56249154e-01 -4.38112438e-01 -4.28552121e-01
1.08424544e-01 -4.33568597e-01 -2.21744239e-01 -3.76301557e-01] | [8.578907012939453, 3.242985725402832] |
7d1ad0ff-ed2d-422f-b890-bfec8ba1b232 | cross-domain-transfer-via-semantic-skill | 2212.07407 | null | https://arxiv.org/abs/2212.07407v1 | https://arxiv.org/pdf/2212.07407v1.pdf | Cross-Domain Transfer via Semantic Skill Imitation | We propose an approach for semantic imitation, which uses demonstrations from a source domain, e.g. human videos, to accelerate reinforcement learning (RL) in a different target domain, e.g. a robotic manipulator in a simulated kitchen. Instead of imitating low-level actions like joint velocities, our approach imitates the sequence of demonstrated semantic skills like "opening the microwave" or "turning on the stove". This allows us to transfer demonstrations across environments (e.g. real-world to simulated kitchen) and agent embodiments (e.g. bimanual human demonstration to robotic arm). We evaluate on three challenging cross-domain learning problems and match the performance of demonstration-accelerated RL approaches that require in-domain demonstrations. In a simulated kitchen environment, our approach learns long-horizon robot manipulation tasks, using less than 3 minutes of human video demonstrations from a real-world kitchen. This enables scaling robot learning via the reuse of demonstrations, e.g. collected as human videos, for learning in any number of target domains. | ['Akshara Rai', 'Dhruv Batra', 'Joseph J. Lim', 'Franziska Meier', 'Vikash Kumar', 'Ruta Desai', 'Karl Pertsch'] | 2022-12-14 | null | null | null | null | ['robot-manipulation'] | ['robots'] | [ 6.32160828e-02 2.09989905e-01 9.01468918e-02 -8.81265402e-02
-4.06962305e-01 -1.00402105e+00 7.47528732e-01 -3.24007034e-01
-5.82336605e-01 9.57515299e-01 -1.80303305e-01 -3.11080635e-01
-8.82001221e-02 -5.75463831e-01 -1.23905587e+00 -4.26715255e-01
-5.09872913e-01 7.03062415e-01 2.22560391e-01 -2.44904056e-01
2.50055045e-01 6.93353951e-01 -1.56417537e+00 -7.16918381e-03
5.77972412e-01 3.54406118e-01 8.95388782e-01 1.02200472e+00
1.70343220e-01 1.01613557e+00 -5.95033526e-01 5.05200505e-01
4.59394276e-01 -5.39958239e-01 -7.67084479e-01 2.99372315e-01
2.96520852e-02 -8.48172188e-01 -6.65032506e-01 8.41134906e-01
9.96150598e-02 6.04592681e-01 6.69554651e-01 -1.71355188e+00
-2.15702266e-01 5.26642919e-01 -1.98206887e-01 -2.76436716e-01
9.13541496e-01 9.28196728e-01 5.95514953e-01 -2.14156166e-01
9.86096084e-01 1.56696963e+00 1.82146490e-01 5.34518838e-01
-1.16569960e+00 -6.35588825e-01 1.03490941e-01 2.39435136e-01
-8.57307017e-01 -4.24483307e-02 3.15020800e-01 -4.91333842e-01
1.14090264e+00 -4.18223232e-01 8.17137420e-01 1.55330181e+00
1.93605199e-01 9.56346035e-01 1.16717255e+00 -3.11810136e-01
5.31585395e-01 -1.57472134e-01 -4.94060665e-01 7.89354920e-01
-2.37474069e-02 6.31791353e-01 -2.79786736e-01 5.47147840e-02
1.32716453e+00 1.43436983e-01 -2.88165450e-01 -1.04310715e+00
-1.71635568e+00 6.36778414e-01 4.42662716e-01 1.06994763e-01
-5.32145262e-01 6.32317662e-01 5.77871680e-01 9.29630160e-01
-5.57493150e-01 8.45836580e-01 -7.08869398e-01 -4.74354684e-01
-2.42310837e-01 7.33335495e-01 1.10475862e+00 1.46202159e+00
8.27479362e-01 -1.80878453e-02 2.14384139e-01 3.18735301e-01
5.33883050e-02 6.47389233e-01 2.70798653e-01 -1.72144461e+00
6.82411909e-01 2.07243443e-01 7.57010162e-01 -1.19536601e-01
-3.80221307e-01 3.47118974e-01 -1.54546216e-01 1.09061193e+00
7.15921164e-01 -3.32227230e-01 -8.58973205e-01 1.76637959e+00
4.67113674e-01 1.97593689e-01 4.10102248e-01 1.17236304e+00
2.31998950e-01 8.21864367e-01 4.26266491e-02 1.87830925e-01
1.29134762e+00 -1.37896621e+00 -1.49600059e-01 -1.48749948e-01
8.93718183e-01 -4.02514130e-01 1.33496344e+00 5.71763635e-01
-8.24906707e-01 -6.95176661e-01 -9.28833842e-01 8.08637962e-02
-3.23515207e-01 -1.17493756e-01 4.31626529e-01 -6.56048879e-02
-8.13874125e-01 1.16046214e+00 -1.11492169e+00 -7.26143301e-01
-5.55515476e-02 4.96439308e-01 -7.54793584e-01 -3.31091285e-01
-8.67225826e-01 1.25905716e+00 6.76916599e-01 -3.55237007e-01
-1.89976013e+00 -5.83139062e-01 -1.24290514e+00 8.74444395e-02
7.44459093e-01 -7.65299857e-01 1.84143937e+00 -8.69542301e-01
-2.15208173e+00 4.43859696e-01 4.83924627e-01 -2.58539528e-01
7.22580612e-01 -3.35156918e-01 2.36189738e-01 4.17365581e-01
9.14292708e-02 1.04153109e+00 8.60900581e-01 -1.26499057e+00
-6.25512183e-01 -1.06498692e-03 8.39845955e-01 4.67114985e-01
3.92911345e-01 -1.37517512e-01 1.98136747e-01 -2.37990096e-01
-3.51169527e-01 -1.43883324e+00 -2.06377566e-01 3.08391154e-01
1.29769266e-01 -2.57752001e-01 9.84038055e-01 -4.48061079e-01
-1.41203314e-01 -2.06788707e+00 8.61646175e-01 -1.87355392e-02
-1.75901517e-01 -1.34460196e-01 -4.86066818e-01 9.29103553e-01
-7.40011688e-03 -5.67642391e-01 -2.30166718e-01 -1.53087646e-01
3.82466197e-01 6.85893357e-01 -8.84779822e-03 4.11070257e-01
-4.78190854e-02 8.80731583e-01 -1.58249760e+00 -2.36569569e-01
4.12615716e-01 2.11415842e-01 -5.21986842e-01 5.79090357e-01
-6.36484623e-01 1.14829493e+00 -5.40150046e-01 1.53783441e-01
1.33676112e-01 -1.00127846e-01 3.17780495e-01 3.37038696e-01
-1.37667522e-01 2.14606464e-01 -1.07918227e+00 2.58811927e+00
-1.03945613e+00 5.54315925e-01 3.55373353e-01 -9.90610480e-01
6.89696193e-01 3.96873236e-01 2.30079994e-01 -4.82156783e-01
-1.71734557e-01 3.22879523e-01 1.68211877e-01 -8.99950743e-01
2.55095065e-01 -9.30882841e-02 -3.72540087e-01 6.37482643e-01
4.41523403e-01 -1.03318346e+00 3.25740874e-01 6.59833550e-02
1.50054741e+00 1.14876032e+00 4.72167373e-01 5.73501512e-02
1.17881246e-01 6.63835585e-01 7.90943429e-02 7.59107769e-01
-2.11912528e-01 1.07028365e-01 3.54225904e-01 -1.62737176e-01
-1.50683677e+00 -1.26766241e+00 6.83618367e-01 1.05207026e+00
2.09596157e-01 -1.20354757e-01 -5.56821764e-01 -6.01970434e-01
1.57941267e-01 8.53844464e-01 -4.30415750e-01 3.06285471e-02
-1.05914140e+00 4.80410516e-01 2.17395321e-01 5.09025633e-01
4.32073563e-01 -1.64264441e+00 -1.57468700e+00 2.03696191e-01
1.62895620e-01 -1.11811662e+00 -2.42817268e-01 3.09441388e-01
-9.07574832e-01 -1.19436920e+00 -8.69625747e-01 -9.60228682e-01
5.95618904e-01 3.59404445e-01 9.03062820e-01 -3.38562906e-01
-4.67823684e-01 1.15745330e+00 -6.08262002e-01 -3.17767859e-02
-7.35269547e-01 -3.37443322e-01 1.76027954e-01 -9.52175200e-01
-1.75672621e-01 -1.00016272e+00 -6.38470888e-01 1.54454350e-01
-6.45127177e-01 2.41033107e-01 9.19827878e-01 1.14532197e+00
-3.78111824e-02 -5.49463592e-02 2.85328060e-01 -2.23864451e-01
4.60003912e-01 -4.80285376e-01 -8.41613233e-01 5.23995198e-02
-8.08596890e-03 9.76609662e-02 8.44942570e-01 -9.58128810e-01
-9.04841721e-01 2.51122594e-01 4.44961786e-01 -6.11019671e-01
-6.48848236e-01 2.51581609e-01 2.27172464e-01 1.79802939e-01
6.13421261e-01 2.72514969e-01 2.89877772e-01 -1.72623038e-01
6.66674495e-01 2.37285972e-01 7.65155733e-01 -1.08715558e+00
9.45836782e-01 8.23575780e-02 3.51239964e-02 -6.19512856e-01
2.32496127e-01 -3.96809220e-01 -7.20766604e-01 -1.39833495e-01
8.11822116e-01 -9.46221471e-01 -1.26916218e+00 2.70611912e-01
-1.25135112e+00 -1.36921835e+00 -3.49217087e-01 9.01015818e-01
-1.47039354e+00 2.49894202e-01 -8.30357909e-01 -5.75961053e-01
3.01974654e-01 -1.21847498e+00 1.20972800e+00 1.90893024e-01
-5.08483708e-01 -6.62634194e-01 2.03278348e-01 -1.28833586e-02
1.17723420e-01 3.52840006e-01 7.69085407e-01 -2.70679921e-01
-8.03436756e-01 2.21455052e-01 -1.28225490e-01 3.75125706e-02
2.10628986e-01 -5.41167319e-01 -4.41274226e-01 -6.97686136e-01
-2.11938187e-01 -9.92434859e-01 1.32089972e-01 -1.05441928e-01
6.93981767e-01 -1.31095693e-01 -3.50433499e-01 1.75701633e-01
1.13563168e+00 6.24159217e-01 5.85332334e-01 5.51379025e-01
2.87008494e-01 7.45004296e-01 1.15063071e+00 3.92603457e-01
2.22090840e-01 7.93905079e-01 4.71946746e-01 2.07968056e-01
-6.18268140e-02 -5.31321764e-01 7.31683135e-01 1.14232369e-01
-2.34092660e-02 2.31471881e-01 -6.66121364e-01 5.30988455e-01
-2.03080106e+00 -9.03275430e-01 5.60213625e-01 1.70323694e+00
5.45723021e-01 -1.75803334e-01 3.26681107e-01 -2.06070870e-01
3.57518524e-01 -2.82022148e-01 -9.57502723e-01 -2.04208732e-01
7.04435050e-01 1.28344670e-01 2.73681104e-01 4.63337988e-01
-6.75107956e-01 9.10732746e-01 5.08431721e+00 3.34579885e-01
-8.89168441e-01 2.84899119e-02 -3.31079572e-01 -1.60005182e-01
2.66220838e-01 8.28111917e-02 1.26437498e-02 3.67218226e-01
7.82061815e-01 4.06053402e-02 9.50865746e-01 1.18670595e+00
3.70675296e-01 -5.34089088e-01 -1.79476690e+00 1.03365839e+00
-4.06915933e-01 -8.87726665e-01 -4.49931949e-01 -1.78168342e-01
4.63402033e-01 1.10375442e-01 -2.62642413e-01 1.00722933e+00
1.03465056e+00 -9.64346826e-01 7.33404875e-01 -4.92113791e-02
7.76920140e-01 -3.51880491e-01 2.12823957e-01 9.18215275e-01
-9.61189330e-01 -2.80382425e-01 -4.07817736e-02 -4.17787582e-01
2.90989220e-01 -5.76139688e-01 -1.23463488e+00 4.34736341e-01
6.67892933e-01 6.45732224e-01 4.10548419e-01 6.64254308e-01
-5.14888704e-01 -8.85901451e-02 -4.25487548e-01 -3.20313156e-01
6.10076487e-01 -5.17664313e-01 4.79630440e-01 9.74124789e-01
5.69156587e-01 3.19501787e-01 4.35681134e-01 8.33890676e-01
2.19143033e-01 -3.61863345e-01 -1.18595684e+00 9.38692391e-02
3.07734609e-01 8.44182968e-01 -5.09810448e-01 -4.64456022e-01
-2.30099797e-01 1.39557540e+00 4.33308870e-01 6.44872308e-01
-1.03382003e+00 -5.42158246e-01 6.62050366e-01 -2.34934077e-01
5.10395825e-01 -8.11687648e-01 5.98690331e-01 -1.01935244e+00
5.35299955e-03 -1.17634797e+00 -1.33535996e-01 -1.19549930e+00
-7.23461211e-01 3.80902998e-02 4.70171392e-01 -1.47518706e+00
-7.13015378e-01 -7.28764117e-01 -6.29618466e-01 6.04621232e-01
-1.48098660e+00 -8.10269058e-01 -4.75335926e-01 7.00142801e-01
1.07267284e+00 -1.27275452e-01 8.90457451e-01 -6.79096654e-02
1.33835077e-01 -1.15769982e-01 -1.08725712e-01 -2.29725718e-01
5.87605715e-01 -1.25887406e+00 4.19842422e-01 5.55937216e-02
-1.48958012e-01 4.87171680e-01 9.44674015e-01 -3.10764045e-01
-1.73949289e+00 -6.06464624e-01 3.59688029e-02 -2.99751520e-01
9.41635787e-01 -4.74287242e-01 -6.39018595e-01 9.96206045e-01
2.12597311e-01 -1.78200930e-01 -1.17797010e-01 -2.74956793e-01
-4.22817677e-01 2.60896534e-01 -1.27003527e+00 9.85395312e-01
1.42838395e+00 -4.81331646e-01 -1.00889778e+00 5.69126785e-01
8.48119736e-01 -7.10688889e-01 -7.89257646e-01 6.63755238e-02
6.82060838e-01 -5.27410805e-01 9.62662339e-01 -7.83828855e-01
6.47122860e-01 -5.71985126e-01 -5.61589450e-02 -1.94460142e+00
-6.54331297e-02 -8.07202518e-01 -1.85893476e-01 2.89481670e-01
1.29059121e-01 -6.49788737e-01 5.21137238e-01 9.17465836e-02
-1.61909595e-01 -2.40205258e-01 -7.58678973e-01 -1.39685392e+00
2.91176233e-03 -7.36158788e-02 4.32846308e-01 6.24295592e-01
5.75317740e-01 2.50773191e-01 -2.94560075e-01 2.78803259e-01
4.81035590e-01 2.88683146e-01 1.41055560e+00 -7.10957468e-01
-8.27360034e-01 -1.69821113e-01 -2.42402971e-01 -1.55427969e+00
6.34382010e-01 -7.73838401e-01 4.30441886e-01 -1.39611518e+00
5.45968451e-02 -5.46829462e-01 1.41531333e-01 3.85894716e-01
4.05953914e-01 -5.30512094e-01 5.94390571e-01 2.32267678e-01
-5.65314233e-01 6.17312372e-01 1.68629813e+00 -1.83148563e-01
-2.86252201e-01 -3.31576347e-01 2.27966309e-01 4.70971793e-01
8.92175376e-01 -4.37259406e-01 -7.42859244e-01 -4.36957449e-01
-3.22418451e-01 8.14894736e-01 6.78040385e-01 -1.14328682e+00
2.52512485e-01 -4.74550664e-01 9.24394056e-02 1.16854891e-01
6.18633687e-01 -1.04436779e+00 1.98501050e-01 1.05714273e+00
-3.72315079e-01 1.92988887e-01 2.42833421e-01 7.94304132e-01
1.93072259e-02 -2.96647966e-01 4.21461582e-01 -6.33270383e-01
-1.12042761e+00 -2.13661984e-01 -4.64738399e-01 -2.32729211e-01
1.49068761e+00 -2.24690974e-01 -2.91546732e-01 -5.73612273e-01
-9.12291229e-01 5.20502806e-01 7.75977790e-01 4.25126463e-01
6.29981875e-01 -9.45128500e-01 -4.79797632e-01 -2.43673265e-01
1.22333750e-01 1.45102605e-01 6.59087077e-02 4.57785338e-01
-7.58261919e-01 2.56501019e-01 -7.07577527e-01 -7.26861775e-01
-1.10241723e+00 8.20751488e-01 1.71532527e-01 -2.57005036e-01
-1.28970790e+00 4.04947668e-01 3.76531869e-01 -8.65086377e-01
2.70074487e-01 -6.31523013e-01 3.25003952e-01 -7.36464083e-01
1.84430107e-01 2.93716848e-01 -6.21165991e-01 2.15031058e-01
-3.22654583e-02 5.02989888e-01 2.15267271e-01 -4.47729468e-01
1.34129953e+00 4.88706343e-02 2.04663038e-01 5.40772319e-01
9.89583790e-01 -6.06194079e-01 -1.98920441e+00 -1.01629961e-02
-6.16823584e-02 -3.53195339e-01 -7.11625576e-01 -8.36488307e-01
-2.37799674e-01 6.21080458e-01 2.80447632e-01 -3.96620184e-01
6.67002857e-01 1.43749133e-01 5.73178649e-01 1.33793163e+00
1.34582460e+00 -1.09593606e+00 7.06732631e-01 6.19589090e-01
1.11114466e+00 -1.30423558e+00 -1.38858661e-01 -3.67090553e-02
-7.02972233e-01 1.15950906e+00 5.65466821e-01 -4.86548483e-01
3.19660380e-02 2.63066560e-01 -8.02334771e-03 -1.34805560e-01
-7.30314255e-01 1.48712676e-02 -5.36923409e-01 8.96416724e-01
-1.96861878e-01 2.18222052e-01 2.97987700e-01 -2.63797224e-01
-8.72931555e-02 2.71431416e-01 7.46082246e-01 1.36112833e+00
-4.30016756e-01 -9.53901350e-01 -4.22792315e-01 -1.20116107e-01
4.23950195e-01 5.14921606e-01 1.00201681e-01 1.30701315e+00
-1.10760137e-01 7.66518295e-01 -4.67699990e-02 9.60440710e-02
5.51933408e-01 8.25009868e-02 1.22861254e+00 -9.14389551e-01
-3.57566863e-01 -4.15337265e-01 8.58181491e-02 -8.42355013e-01
-3.65502566e-01 -7.01676369e-01 -1.58337665e+00 -1.88663676e-01
3.43520284e-01 1.23099193e-01 7.57125854e-01 7.91356683e-01
6.77306056e-02 5.19259274e-01 3.32053363e-01 -1.66801560e+00
-9.79273140e-01 -1.02259934e+00 -5.72166383e-01 7.16055572e-01
5.40606797e-01 -9.61239278e-01 -2.92942345e-01 2.28979643e-02] | [4.549839973449707, 0.84914231300354] |
040aa910-8901-4f4c-a9ec-51fd14ec29ca | unsupervised-chinese-word-segmentation-with | null | null | https://openreview.net/forum?id=DXbxULpbXZ4 | https://openreview.net/pdf?id=DXbxULpbXZ4 | Unsupervised Chinese Word Segmentation with BERT Oriented Probing and Transformation | Word Segmentation is a fundamental step for understanding Chinese language. Previous neural approaches for unsupervised Chinese Word Segmentation (CWS) only exploits shallow semantic information, which can miss important context. Large scale Pre-trained language models (PLM) have achieved great success in many areas because of its ability to capture the deep contextual semantic relation. In this paper, we propose to take advantage of the deep semantic information embedded in PLM (e.g., BERT) with a self-training manner, which iteratively probes and transforms the semantic information in PLM into explicit word segmentation ability. Extensive experiment results show that our proposed approach achieves state-of-the-art F1 score on two CWS benchmark datasets. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['chinese-word-segmentation'] | ['natural-language-processing'] | [ 2.76158482e-01 1.08286403e-01 -4.47788745e-01 -5.90058923e-01
-5.79651654e-01 -4.15840000e-01 2.12765068e-01 1.36285797e-01
-7.11342335e-01 4.19493884e-01 3.65970820e-01 -5.25539160e-01
4.98308718e-01 -7.34002948e-01 -4.37798440e-01 -3.63974184e-01
3.69876385e-01 3.91670495e-01 6.54670954e-01 -1.52477860e-01
4.01188940e-01 -1.40292376e-01 -7.35478997e-01 2.74011046e-01
1.45817149e+00 6.04148984e-01 8.16001713e-01 2.93605447e-01
-8.58072400e-01 5.06763697e-01 -6.11264288e-01 -1.84480205e-01
-2.77346373e-01 -6.55020714e-01 -1.21281183e+00 -2.67983489e-02
-2.98267186e-01 -1.48749694e-01 1.11860096e-01 1.34652984e+00
5.58651350e-02 2.25680217e-01 2.46212751e-01 -5.41917264e-01
-7.29050398e-01 1.09674525e+00 -5.85982740e-01 2.32863232e-01
-5.25574796e-02 2.27766000e-02 1.32332194e+00 -7.96481729e-01
5.34455001e-01 1.32377470e+00 4.40979958e-01 7.54174829e-01
-9.98383939e-01 -6.20484352e-01 8.36356878e-01 5.78866042e-02
-1.15921438e+00 1.27529085e-01 8.09857428e-01 2.26777475e-02
1.00754952e+00 -1.65032558e-02 7.01776862e-01 8.48158062e-01
-1.04947351e-01 1.64909112e+00 1.01929915e+00 -6.98234975e-01
2.82888025e-01 -1.77006274e-01 8.65241647e-01 4.38518912e-01
4.93435740e-01 -5.27421236e-01 -2.21813634e-01 2.76666075e-01
7.06272900e-01 -1.11134298e-01 -2.51183044e-02 3.08463573e-01
-7.60792613e-01 9.51947629e-01 2.59479135e-01 7.20254481e-01
-1.17754012e-01 1.46220520e-01 2.92783886e-01 -7.41824508e-02
5.59808493e-01 5.85643411e-01 -8.81966054e-01 -1.81648448e-01
-7.90580690e-01 -2.87977904e-01 5.56948900e-01 8.89897466e-01
7.57357419e-01 3.57873105e-02 -2.15304449e-01 8.11710179e-01
3.16352904e-01 1.31789908e-01 9.58557129e-01 -5.83113968e-01
4.31554466e-01 6.93022668e-01 -2.25927025e-01 -5.74418008e-01
-4.07399148e-01 -4.40850109e-01 -5.50149739e-01 -6.30344391e-01
2.35888362e-01 -4.44976538e-01 -1.43313920e+00 1.82646036e+00
3.97995207e-03 3.82981569e-01 2.46642366e-01 8.23958576e-01
7.84626663e-01 7.59232759e-01 6.91487372e-01 1.46379635e-01
1.39500189e+00 -1.22589374e+00 -7.83055663e-01 -9.12433982e-01
7.46724069e-01 -5.49632251e-01 1.41105700e+00 2.13733837e-01
-8.30940425e-01 -6.29249096e-01 -7.56992877e-01 -1.37812719e-01
-4.95203286e-01 1.01589359e-01 7.54130363e-01 7.58871436e-01
-8.55328202e-01 3.07854563e-01 -1.17368948e+00 -3.50080341e-01
6.92305684e-01 2.53997386e-01 1.58816814e-01 -3.00796121e-01
-1.51958919e+00 4.84420359e-01 9.55557644e-01 2.74466932e-01
-6.88711643e-01 -3.32285196e-01 -1.01877081e+00 1.91202983e-01
8.28540981e-01 -1.69439629e-01 1.25269902e+00 -1.15252352e+00
-1.48324752e+00 7.05967605e-01 -4.57575113e-01 -4.22783017e-01
6.08409196e-02 -7.51718163e-01 -2.02880934e-01 2.56458014e-01
1.75347835e-01 9.47934687e-01 4.71442163e-01 -1.26877069e+00
-6.32912159e-01 -2.56214291e-01 -4.67315204e-02 7.25612268e-02
-4.47614759e-01 2.81177551e-01 -9.87906516e-01 -8.72247577e-01
2.72947043e-01 -5.30294597e-01 -6.89450145e-01 -8.95669162e-01
-6.21152341e-01 -4.57512915e-01 7.24216163e-01 -8.24680328e-01
1.37199807e+00 -1.74214244e+00 -8.97703990e-02 6.04114868e-02
-7.33198151e-02 7.89876044e-01 -4.35552508e-01 1.97191641e-01
2.79686242e-01 5.47697604e-01 -3.82053047e-01 -4.90079969e-01
-9.35572907e-02 5.51407397e-01 -2.63386339e-01 -1.08122036e-01
4.52061623e-01 1.44177091e+00 -1.00763583e+00 -5.70546389e-01
-6.87965080e-02 1.65910736e-01 -5.82337797e-01 2.45423004e-01
-7.50541747e-01 5.32762468e-01 -9.03431833e-01 4.59237009e-01
6.62934959e-01 -3.51266921e-01 4.38876122e-01 3.14387590e-01
6.60954267e-02 5.90208709e-01 -7.82333672e-01 1.90577543e+00
-3.61145705e-01 4.26038235e-01 -7.30624571e-02 -1.27219093e+00
7.90211916e-01 1.02193020e-01 -2.62449831e-02 -8.00641477e-01
3.39178860e-01 1.62097737e-01 6.72110319e-02 -1.79179013e-01
5.97540259e-01 -1.24477521e-01 -2.82480419e-01 2.76342392e-01
2.26470202e-01 5.04756682e-02 1.91400379e-01 3.16334069e-01
5.97333491e-01 2.34982356e-01 2.45089069e-01 -5.54477155e-01
5.94185352e-01 1.22255675e-01 8.56987953e-01 7.19733417e-01
-1.64368555e-01 6.26749516e-01 6.93789065e-01 -8.96598175e-02
-5.40912628e-01 -8.49753976e-01 2.40429536e-01 1.11135817e+00
3.87051463e-01 -4.42095488e-01 -1.20633531e+00 -7.80071318e-01
-4.91374046e-01 9.73084450e-01 -4.91211891e-01 -2.63740003e-01
-9.52182591e-01 -8.78112972e-01 5.03500819e-01 9.67826784e-01
7.22475588e-01 -1.39858413e+00 -4.94248509e-01 4.23621446e-01
-2.99857199e-01 -1.54614604e+00 -4.24598068e-01 3.10615778e-01
-1.07727134e+00 -8.09634984e-01 -5.97372711e-01 -1.30011344e+00
9.51898336e-01 3.99988413e-01 1.08910370e+00 4.25723106e-01
-6.25960007e-02 9.80638992e-03 -6.99204147e-01 -3.16255689e-01
-1.35572746e-01 6.47215843e-01 -5.42081296e-01 -1.31518066e-01
8.59816849e-01 -2.55837012e-02 -5.78530431e-01 1.84076905e-01
-1.05789399e+00 3.59393090e-01 6.61914647e-01 7.67284453e-01
6.49651349e-01 1.81860059e-01 6.79403603e-01 -1.54202116e+00
6.20341063e-01 -3.34220588e-01 -5.61166823e-01 2.91150331e-01
-5.21411836e-01 7.83226416e-02 6.51796699e-01 -3.83528531e-01
-1.45295763e+00 -1.89470693e-01 -4.32365686e-01 3.01024914e-01
-4.63754505e-01 8.56131911e-01 -5.75189114e-01 4.83498096e-01
-3.13721225e-02 3.78204823e-01 -5.61801136e-01 -9.56039310e-01
4.61035609e-01 6.07197702e-01 5.91355860e-01 -9.93456662e-01
3.87875855e-01 3.18530411e-01 -6.26623809e-01 -7.83245146e-01
-1.33739769e+00 -8.02169919e-01 -9.23269391e-01 3.88014376e-01
1.36526811e+00 -9.39849019e-01 -1.57058448e-01 8.26283216e-01
-1.16325366e+00 -6.99834168e-01 -7.29906708e-02 2.56114990e-01
1.14528961e-01 5.18748343e-01 -9.08546925e-01 -6.91765964e-01
-4.06927377e-01 -9.67179298e-01 9.33609307e-01 7.74035037e-01
-1.87526941e-01 -1.19240117e+00 -3.18504393e-01 3.24368238e-01
3.27612817e-01 -1.30478308e-01 8.83246899e-01 -1.00107384e+00
-5.23477733e-01 -1.86313599e-01 -5.20952761e-01 4.66101170e-01
1.55003682e-01 -4.59875405e-01 -7.23590195e-01 -7.81569481e-02
-1.79188624e-01 -2.23372474e-01 1.42043149e+00 3.62691700e-01
1.36393619e+00 -1.58639073e-01 -4.12921607e-01 3.42856973e-01
1.58024168e+00 2.71906674e-01 5.41361153e-01 1.89451411e-01
1.00795841e+00 5.55137336e-01 6.93416655e-01 7.93780312e-02
4.30821329e-01 -4.07449454e-02 2.36865386e-01 -2.62642771e-01
5.38803190e-02 -4.69449461e-01 2.20506519e-01 1.09776115e+00
3.25753063e-01 -3.15396726e-01 -1.14126766e+00 9.97777998e-01
-1.84141493e+00 -1.33998975e-01 -2.44301874e-02 1.62403286e+00
1.20095146e+00 6.52216017e-01 -2.84806758e-01 -2.66211390e-01
6.92225337e-01 3.35354507e-01 -4.91945773e-01 -3.42336476e-01
-1.35160625e-01 6.28139377e-01 4.83336508e-01 6.81985199e-01
-1.05284858e+00 2.05639768e+00 5.44463825e+00 1.08974028e+00
-9.13307250e-01 1.40346259e-01 8.46771121e-01 6.53999090e-01
-6.07933402e-01 8.99221599e-02 -1.06306088e+00 3.27091634e-01
8.33103716e-01 2.64425665e-01 6.84672669e-02 7.83584237e-01
1.56726718e-01 -4.40636337e-01 -4.81136829e-01 2.85486490e-01
-1.00888796e-01 -1.15816462e+00 2.58882284e-01 -3.07311058e-01
1.03741348e+00 1.77460060e-01 -2.22924456e-01 3.65852624e-01
7.68270850e-01 -1.04962993e+00 5.86546063e-01 -3.70319076e-02
7.14968503e-01 -7.88399398e-01 9.36761498e-01 5.88473678e-01
-1.26857388e+00 2.10486963e-01 -4.57124412e-01 -1.53638020e-01
2.27991864e-01 4.81127024e-01 -4.99031872e-01 3.38768601e-01
5.14527261e-01 6.37651026e-01 -6.36015058e-01 6.76539183e-01
-1.00213754e+00 1.35559976e+00 -1.54087886e-01 -3.74249905e-01
8.89503241e-01 -3.08777064e-01 1.14156112e-01 1.55387735e+00
-2.21678570e-01 4.08768862e-01 4.14594501e-01 9.16918576e-01
-1.01086430e-01 1.80720031e-01 1.89718097e-01 -4.39580917e-01
3.26782525e-01 9.77704108e-01 -1.45386064e+00 -5.42785287e-01
-5.49142241e-01 1.11236918e+00 3.60145241e-01 4.90368009e-01
-5.04405439e-01 -4.89883840e-01 7.87224531e-01 -4.02237117e-01
6.68087542e-01 -5.91104388e-01 -6.82864428e-01 -1.23651695e+00
-2.40059912e-01 -5.01718581e-01 5.09957373e-01 -4.23887879e-01
-9.44745243e-01 4.47686136e-01 -1.72379777e-01 -3.93807083e-01
1.18938528e-01 -7.48469055e-01 -9.37479079e-01 9.73212838e-01
-1.83214676e+00 -1.33393204e+00 -1.29327610e-01 2.28897169e-01
1.03663993e+00 1.86388552e-01 6.95323467e-01 -8.25747252e-02
-8.11808288e-01 4.53054160e-01 -2.55768429e-02 7.70563304e-01
1.94750533e-01 -1.41665184e+00 1.03642988e+00 1.34466314e+00
3.28831047e-01 9.05276239e-01 4.45538551e-01 -9.05652761e-01
-9.26793039e-01 -1.18017793e+00 9.10606742e-01 -2.27405712e-01
4.81650621e-01 -6.16064966e-01 -1.11404991e+00 7.35869586e-01
3.39247257e-01 -2.91982681e-01 7.01332271e-01 1.67963043e-01
-2.53522664e-01 3.57306838e-01 -7.41782546e-01 6.92237139e-01
9.50176001e-01 -3.77007306e-01 -1.03056693e+00 -1.29132196e-01
1.40655088e+00 -2.78875917e-01 -1.03970371e-01 1.84675127e-01
1.58390820e-01 -4.69011277e-01 7.79388487e-01 -9.47100639e-01
3.95153046e-01 -1.78576574e-01 -9.71525069e-03 -1.24748707e+00
-5.73931001e-02 -5.98009765e-01 2.37454876e-01 1.28806651e+00
4.37564969e-01 -5.71759343e-01 9.37546968e-01 4.45234656e-01
-2.88289368e-01 -7.37640619e-01 -4.34838653e-01 -6.74632072e-01
4.01234537e-01 -7.98575044e-01 6.07919872e-01 9.16815281e-01
-6.44715279e-02 5.10529637e-01 4.08066995e-02 1.19577982e-01
2.73994118e-01 2.94994652e-01 2.30144292e-01 -1.10052204e+00
-1.65769994e-01 -5.70227206e-01 1.67542860e-01 -1.67438900e+00
5.33489227e-01 -6.94750011e-01 4.28945363e-01 -1.81659830e+00
2.95598298e-01 -3.57277185e-01 -6.30130887e-01 5.62590897e-01
-8.66698861e-01 2.14023009e-01 1.46881536e-01 -1.10171631e-01
-9.80038226e-01 3.90459865e-01 1.21262896e+00 4.40658033e-02
-3.65455300e-01 -1.56581670e-01 -1.03135026e+00 9.80157197e-01
1.00004053e+00 -3.65989834e-01 -2.69242734e-01 -9.01037037e-01
4.03207056e-02 -4.03482199e-01 -2.37232819e-01 -6.16556525e-01
1.47846062e-02 -4.43710089e-01 2.33715132e-01 -7.07581639e-01
-1.57677770e-01 -4.89159375e-01 -8.58012259e-01 5.71895897e-01
-3.29830527e-01 -3.81940067e-01 2.32456505e-01 4.89345729e-01
-4.57830191e-01 -4.75917727e-01 7.77621567e-01 -5.26438355e-01
-1.33495319e+00 1.58221081e-01 -4.34944153e-01 5.05584478e-01
5.75549543e-01 1.93719529e-02 -1.32562280e-01 -1.03474796e-01
-4.71607566e-01 5.92545748e-01 1.70165658e-01 6.13249362e-01
4.99144226e-01 -8.62997174e-01 -4.26400870e-01 1.67746052e-01
-1.22250751e-01 6.62664413e-01 -1.07232472e-02 3.93947303e-01
-5.46816111e-01 6.49622679e-01 1.84098363e-01 -3.43184114e-01
-8.18543077e-01 4.32518810e-01 -2.58864891e-02 -4.75383401e-01
-6.38334811e-01 1.19389296e+00 4.75098342e-01 -2.92361021e-01
1.02361426e-01 -8.02617669e-01 -4.15052205e-01 -2.12900154e-02
5.38933158e-01 -1.88833937e-01 -2.44247168e-01 -5.15078664e-01
-3.27208817e-01 6.53306782e-01 -2.33870074e-01 -8.38674009e-02
1.20532835e+00 -3.30228895e-01 -1.96521252e-01 2.06398457e-01
1.14445484e+00 2.88897101e-02 -1.17707396e+00 -5.97544551e-01
6.92316771e-01 -1.77148372e-01 1.12612262e-01 -8.30106318e-01
-1.09289443e+00 1.31751227e+00 -1.27766222e-01 -1.29574224e-01
1.20255971e+00 1.96704879e-01 1.54419804e+00 2.28386134e-01
1.72878847e-01 -1.46247661e+00 1.43905833e-01 9.75282013e-01
1.02853522e-01 -1.37181425e+00 -3.88888180e-01 -5.52258670e-01
-7.68859327e-01 1.00215507e+00 9.16688740e-01 -2.29050756e-01
6.17851853e-01 1.31062597e-01 1.98155150e-01 -2.18872055e-02
-3.61380905e-01 -6.30169928e-01 2.14128479e-01 4.23936903e-01
4.04573083e-01 2.36577332e-01 -6.38650537e-01 1.44763231e+00
9.34729576e-02 -3.95243138e-01 1.98924839e-01 9.67698097e-01
-9.40084934e-01 -1.33599138e+00 -5.89642972e-02 2.79381603e-01
-7.17487693e-01 -5.23434103e-01 -5.57851493e-01 5.93474090e-01
7.13913888e-02 8.65619481e-01 5.82366623e-02 -7.77680501e-02
-6.58486458e-03 2.16939569e-01 3.69377658e-02 -9.78598237e-01
-3.93082768e-01 4.28413987e-01 -1.35210082e-01 -5.12034535e-01
-2.45256945e-01 -6.26388848e-01 -2.07253933e+00 3.85823429e-01
-2.92145044e-01 3.70624840e-01 3.84836257e-01 1.39075530e+00
-2.05032937e-02 6.09824419e-01 2.84932643e-01 -2.33389735e-01
-8.35482255e-02 -9.86234009e-01 -3.27700168e-01 3.50507408e-01
1.26684889e-01 -1.70797497e-01 9.34038237e-02 -2.93661328e-03] | [9.952720642089844, 10.109447479248047] |
78fe8d6d-6953-444f-8c95-a4aea6f82e7e | keep-learning-self-supervised-meta-learning | null | null | https://aclanthology.org/2021.eacl-main.6 | https://aclanthology.org/2021.eacl-main.6.pdf | Keep Learning: Self-supervised Meta-learning for Learning from Inference | A common approach in many machine learning algorithms involves self-supervised learning on large unlabeled data before fine-tuning on downstream tasks to further improve performance. A new approach for language modelling, called dynamic evaluation, further fine-tunes a trained model during inference using trivially-present ground-truth labels, giving a large improvement in performance. However, this approach does not easily extend to classification tasks, where ground-truth labels are absent during inference. We propose to solve this issue by utilizing self-training and back-propagating the loss from the model{'}s own class-balanced predictions (pseudo-labels), adapting the Reptile algorithm from meta-learning, combined with an inductive bias towards pre-trained weights to improve generalization. Our method improves the performance of standard backbones such as BERT, Electra, and ResNet-50 on a wide variety of tasks, such as question answering on SQuAD and NewsQA, benchmark task SuperGLUE, conversation response selection on Ubuntu Dialog corpus v2.0, as well as image classification on MNIST and ImageNet without any changes to the underlying models. Our proposed method outperforms previous approaches, enables self-supervised fine-tuning during inference of any classifier model to better adapt to target domains, can be easily adapted to any model, and is also effective in online and transfer-learning settings. | ['Sai Chetan Chinthakindi', 'Akhil Kedia'] | 2021-04-01 | null | null | null | eacl-2021-2 | ['answer-selection'] | ['natural-language-processing'] | [ 1.83138564e-01 3.15648764e-01 -2.76456356e-01 -7.96877801e-01
-9.47741210e-01 -6.03102386e-01 5.69566607e-01 1.90749571e-01
-8.57578278e-01 1.00992465e+00 5.26456460e-02 -4.00666088e-01
3.12105156e-02 -8.10260952e-01 -8.25128734e-01 -4.15229470e-01
1.11246340e-01 1.01037943e+00 5.58624387e-01 -6.32332265e-01
1.20965764e-01 -9.72457230e-03 -1.28003252e+00 7.41373122e-01
7.67345786e-01 1.13091242e+00 1.96004346e-01 7.62935340e-01
-2.59914309e-01 1.22030723e+00 -6.26396358e-01 -9.01599526e-01
-4.28156182e-02 -2.46555716e-01 -1.39142156e+00 -8.35091919e-02
3.58157367e-01 -1.75474703e-01 1.50025534e-02 8.33045423e-01
5.91525555e-01 1.61531791e-01 5.22396863e-01 -8.99249792e-01
-5.55629969e-01 1.01116717e+00 -1.75775960e-01 2.24442586e-01
1.73144531e-03 2.28187144e-01 1.09867740e+00 -6.80867374e-01
4.54699874e-01 1.58346081e+00 6.21733069e-01 8.76520753e-01
-1.35329151e+00 -6.80636168e-01 2.39228249e-01 4.26340759e-01
-1.00846624e+00 -4.87716466e-01 3.90705258e-01 -2.26132572e-01
1.08902490e+00 2.34540150e-01 2.79468726e-02 1.23447263e+00
-5.60296513e-02 9.31711674e-01 1.22778857e+00 -4.35627013e-01
2.93747544e-01 5.41756868e-01 1.66836560e-01 6.85910165e-01
-6.24746561e-01 -1.04195207e-01 -4.37650681e-01 -2.64393896e-01
2.44933411e-01 -3.61552715e-01 -1.01679012e-01 -2.92552471e-01
-1.33108187e+00 1.03509116e+00 5.98406851e-01 2.00854197e-01
-2.19339713e-01 -8.91264342e-03 6.12492859e-01 7.16481924e-01
7.65884936e-01 7.49523818e-01 -9.08859253e-01 6.74425140e-02
-6.98297203e-01 2.28264123e-01 1.01878548e+00 6.18168354e-01
1.01996446e+00 -2.43099853e-01 -5.98545492e-01 1.16519523e+00
1.41906261e-01 1.15678482e-01 1.03313303e+00 -1.11848211e+00
5.46767294e-01 5.75517297e-01 -1.24655208e-02 -3.40577960e-01
-4.80699182e-01 -6.87357008e-01 -8.73559296e-01 4.50057276e-02
5.05773485e-01 -2.68310934e-01 -8.76117885e-01 1.94013429e+00
3.32972676e-01 4.12143916e-02 2.59662658e-01 8.23057473e-01
9.50707972e-01 5.88405490e-01 2.47641966e-01 3.02014947e-02
1.13681543e+00 -1.48545933e+00 -2.61260360e-01 -3.84218365e-01
8.41535330e-01 -6.00353718e-01 1.20402730e+00 4.07358438e-01
-1.02893221e+00 -7.07607150e-01 -8.11334789e-01 -8.25943053e-02
-3.48939002e-01 -8.19399580e-02 5.62702835e-01 5.46570361e-01
-1.26319242e+00 5.60581863e-01 -4.57258523e-01 -2.25945428e-01
2.60553777e-01 6.41275048e-01 -8.58037546e-02 4.97995988e-02
-1.59061384e+00 1.25645363e+00 7.20147014e-01 -4.91400361e-02
-1.03526771e+00 -6.81893528e-01 -4.84817564e-01 3.65868062e-02
4.84778851e-01 -8.40804696e-01 1.69371247e+00 -1.47432268e+00
-2.05420709e+00 1.15199172e+00 6.51585460e-02 -9.41392124e-01
6.70105100e-01 -9.61618870e-03 -2.40202442e-01 -2.32558653e-01
-1.33727059e-01 1.16060281e+00 8.09722483e-01 -8.73097539e-01
-5.63291788e-01 -1.70994967e-01 5.41606128e-01 4.17534977e-01
-3.89769524e-01 -2.08440676e-01 -3.71472687e-01 -2.54140079e-01
-3.41970533e-01 -8.63367558e-01 -5.10285974e-01 -4.31321502e-01
-4.26722437e-01 -5.59547424e-01 3.08822572e-01 -2.02817395e-01
9.40863073e-01 -1.85428762e+00 2.55493790e-01 -5.05742058e-02
1.25160709e-01 5.73559642e-01 -4.28235084e-01 2.09300488e-01
4.01650146e-02 -1.91616282e-01 -2.12893501e-01 -3.60043764e-01
-5.35909608e-02 2.21120596e-01 -3.16302061e-01 2.95910966e-02
2.14639530e-01 9.67241943e-01 -9.60322917e-01 -4.14429098e-01
1.52008429e-01 8.12037587e-02 -7.58886814e-01 4.23893392e-01
-7.80187011e-01 5.38447559e-01 -3.88238341e-01 5.20122647e-02
3.08733284e-01 -6.45947874e-01 4.31193523e-02 8.68994147e-02
2.26261839e-01 6.44156039e-01 -9.33148444e-01 1.74325132e+00
-9.36061442e-01 2.75405288e-01 -7.05605894e-02 -1.37749898e+00
9.07455325e-01 1.04792617e-01 4.56535518e-02 -8.60594571e-01
-4.15401021e-03 7.94176534e-02 5.90355508e-02 -4.46227938e-01
3.00342202e-01 7.90412724e-03 -1.42375054e-03 4.80776936e-01
5.01133442e-01 6.85575977e-02 2.53891021e-01 2.81489968e-01
1.03884435e+00 1.02708608e-01 1.78237453e-01 -2.71243840e-01
8.93577099e-01 1.32198811e-01 2.36079752e-01 9.31653738e-01
9.17051658e-02 2.73472846e-01 2.29726046e-01 -4.05545741e-01
-7.30051875e-01 -9.66735482e-01 -1.92758054e-01 2.14138603e+00
-4.37349491e-02 -2.86587298e-01 -9.35444534e-01 -1.00587142e+00
-3.35537903e-02 9.37976897e-01 -6.02803588e-01 -2.82585531e-01
-4.29696172e-01 -1.09898973e+00 5.50132453e-01 2.31944710e-01
6.54514432e-01 -1.29147232e+00 -1.81604877e-01 4.10747617e-01
-4.57835764e-01 -9.69139159e-01 -1.13045022e-01 3.73527080e-01
-8.64590228e-01 -7.23808050e-01 -6.60477221e-01 -7.19383717e-01
4.82026041e-01 -2.95243770e-01 1.62175560e+00 1.14459939e-01
-1.65160354e-02 1.61533117e-01 -2.49462485e-01 -1.84699655e-01
-8.09236526e-01 5.98481417e-01 -2.07838431e-01 1.95538148e-01
3.14020932e-01 -3.01499754e-01 -5.25901377e-01 6.44154370e-01
-6.23251021e-01 2.01460481e-01 5.90126514e-01 1.14603996e+00
4.48794007e-01 -3.23330998e-01 1.04256713e+00 -1.34301865e+00
6.90002024e-01 -5.28399050e-01 -4.48368013e-01 5.83583891e-01
-7.03728199e-01 3.80346805e-01 7.40187168e-01 -4.73290205e-01
-1.23964906e+00 -2.26723775e-01 -4.54541296e-01 -1.18263997e-01
-1.35786250e-01 3.96947652e-01 3.87100428e-02 4.47129346e-02
1.19152248e+00 -4.63593341e-02 -8.90796855e-02 -4.60383594e-01
6.49498940e-01 8.12845409e-01 4.63753253e-01 -5.51069558e-01
3.67388129e-01 2.17455864e-01 -5.02289891e-01 -4.05022383e-01
-1.44598448e+00 -4.81400162e-01 -4.32197303e-01 1.18115388e-01
6.68613255e-01 -9.59566653e-01 -9.01032865e-01 4.33967710e-01
-9.67169106e-01 -8.49634409e-01 -1.58138528e-01 3.48563373e-01
-6.33189321e-01 -3.15450206e-02 -7.54057050e-01 -4.20472413e-01
-5.57093918e-01 -1.22025037e+00 8.22067797e-01 1.56454131e-01
-2.90901363e-01 -1.28951097e+00 8.89433473e-02 7.74117768e-01
7.35500574e-01 -4.93064046e-01 9.09814000e-01 -1.05257559e+00
-4.58999574e-01 5.92295714e-02 -1.61794245e-01 7.15573132e-01
-2.42763937e-01 -4.39272225e-01 -1.25059283e+00 -2.89219797e-01
-1.65862843e-01 -1.18646324e+00 1.18494630e+00 9.14339200e-02
1.41328454e+00 -3.27529937e-01 -1.96487233e-01 3.68861794e-01
9.72358882e-01 -3.68683398e-01 3.05494219e-01 4.60935056e-01
3.67875397e-01 7.66736925e-01 5.39981961e-01 1.52759165e-01
6.86520755e-01 7.08202720e-01 4.64007854e-01 1.35563016e-02
-8.50064829e-02 -1.28101304e-01 2.95336604e-01 7.31695831e-01
1.98595524e-01 -4.22243983e-01 -8.75503659e-01 3.52524132e-01
-1.94480729e+00 -7.16929436e-01 2.29897931e-01 2.13550568e+00
1.38620138e+00 4.28525507e-01 6.37311190e-02 -4.12209123e-01
6.71699107e-01 -3.24343033e-02 -8.05953979e-01 -3.88972938e-01
-6.87071681e-02 3.30457568e-01 4.57740486e-01 7.30673909e-01
-1.14526165e+00 1.32232440e+00 6.05806112e+00 1.10465145e+00
-1.21637630e+00 5.50909460e-01 1.07100904e+00 2.89661177e-02
-2.46632293e-01 -8.56070518e-02 -1.03741717e+00 2.63896793e-01
1.22270107e+00 1.45636573e-02 5.36538005e-01 8.63865554e-01
-1.81102782e-01 -6.23441041e-02 -1.23793530e+00 5.76392055e-01
1.29133061e-01 -1.55819511e+00 7.71558434e-02 -5.00148833e-01
7.46450722e-01 5.39604843e-01 6.21519014e-02 9.98389721e-01
8.32471848e-01 -9.56283927e-01 4.50864583e-01 3.20956141e-01
6.30115032e-01 -5.94378531e-01 7.40384161e-01 8.98576081e-01
-3.16478074e-01 -8.80070552e-02 -4.70916033e-01 -1.00113936e-01
-3.01516671e-02 3.60783070e-01 -1.25318480e+00 2.06482619e-01
5.67859948e-01 2.29207173e-01 -7.39196420e-01 7.19183862e-01
-2.76261777e-01 9.14208710e-01 -1.98330402e-01 -3.10171545e-01
4.62809652e-01 2.99244434e-01 1.21472307e-01 1.30004883e+00
-3.95630270e-01 -3.51682723e-01 3.24851155e-01 5.47335982e-01
-4.22452807e-01 2.25967154e-01 6.97577670e-02 3.54537249e-01
2.16967478e-01 1.27706611e+00 -4.25580770e-01 -6.24310374e-01
-2.01904863e-01 8.72714341e-01 7.75249481e-01 3.38600636e-01
-7.83864141e-01 -1.25768185e-01 1.48809478e-01 -1.50335832e-02
7.09463954e-02 3.42765361e-01 -9.90514010e-02 -1.20259380e+00
-3.58477324e-01 -1.06291461e+00 6.23559535e-01 -6.39783978e-01
-1.43900645e+00 1.01520348e+00 -8.69757235e-02 -8.08025956e-01
-7.72436500e-01 -5.60647786e-01 -3.70004296e-01 8.64393950e-01
-1.72548628e+00 -1.09385431e+00 -1.05768897e-01 7.84459829e-01
6.66171670e-01 -4.10866320e-01 1.06051958e+00 3.10124189e-01
-2.94388026e-01 8.45928788e-01 1.59907356e-01 9.43806246e-02
1.07642722e+00 -1.38609421e+00 4.31557864e-01 2.30832547e-01
2.40173504e-01 2.80940861e-01 6.15059376e-01 -2.17422053e-01
-7.57667482e-01 -1.20644760e+00 8.12133551e-01 -5.08115530e-01
9.41190481e-01 -4.72328365e-01 -9.18951333e-01 7.26515889e-01
4.22162235e-01 -9.56513286e-02 5.95244288e-01 5.50243080e-01
-4.46341127e-01 -2.34967560e-01 -1.03411257e+00 4.87073481e-01
8.87765408e-01 -3.21138710e-01 -6.21583998e-01 9.58728790e-01
7.68734992e-01 -6.62744462e-01 -7.00031936e-01 4.94724363e-01
2.04408005e-01 -9.38165128e-01 9.69887733e-01 -1.07105017e+00
4.72252607e-01 2.22828552e-01 3.76745895e-03 -1.55531526e+00
-2.36993447e-01 -5.89876413e-01 1.63778901e-01 1.07706439e+00
9.13999379e-01 -6.93634808e-01 8.66600752e-01 3.63017887e-01
1.18540286e-03 -8.05588841e-01 -8.42394531e-01 -4.02968198e-01
1.81508482e-01 -3.70332569e-01 4.22627866e-01 9.31135416e-01
-2.37994671e-01 8.26020300e-01 -2.94057459e-01 -1.77615508e-01
5.47349870e-01 1.65159568e-01 8.78750503e-01 -1.08131015e+00
-7.57824183e-01 -5.13223767e-01 -1.15918308e-01 -1.28026986e+00
6.20450497e-01 -1.19741273e+00 4.94865067e-02 -1.16124439e+00
9.70188081e-02 -7.51210034e-01 -4.14841771e-01 5.69055855e-01
-2.92241156e-01 4.45138246e-01 3.69653222e-04 8.13967884e-02
-1.11926198e+00 4.98967588e-01 1.21667647e+00 -2.67941624e-01
-7.28170425e-02 4.57901359e-01 -6.31781280e-01 8.14888954e-01
7.46699154e-01 -4.42777812e-01 -4.81108516e-01 -5.79749525e-01
3.64431173e-01 1.00693598e-01 3.17358375e-01 -6.65421188e-01
2.98200786e-01 1.53633747e-02 2.09759980e-01 -2.70470023e-01
3.64119053e-01 -4.19439942e-01 -3.28937262e-01 2.81438768e-01
-1.04210877e+00 -1.56874552e-01 1.07044198e-01 4.02997881e-01
-2.50396162e-01 -5.56959689e-01 9.36107337e-01 -4.12191749e-01
-8.75124753e-01 6.55534938e-02 -1.00827992e-01 4.65391964e-01
7.59081900e-01 3.13429683e-01 -4.46939677e-01 -4.65916216e-01
-1.10194731e+00 6.92263424e-01 -1.52331283e-02 6.19657636e-01
1.44601226e-01 -9.80258703e-01 -9.93779778e-01 6.92741647e-02
1.60178468e-01 4.89140339e-02 3.70056331e-01 7.31929362e-01
-3.62689167e-01 4.59619105e-01 5.07844379e-03 -8.78318608e-01
-9.19348121e-01 3.44233841e-01 5.93258739e-01 -7.92654991e-01
-1.75970376e-01 1.35665512e+00 3.32050443e-01 -1.06067646e+00
4.50500220e-01 -4.78583165e-02 -2.80526310e-01 -7.60228792e-03
5.60980260e-01 6.54598475e-02 3.55264217e-01 -3.15304428e-01
-1.50419652e-01 -1.39517644e-02 -4.45628434e-01 -1.49395913e-01
1.00866711e+00 -2.62898535e-01 -1.04147874e-01 6.33005440e-01
1.17352033e+00 -3.99788350e-01 -1.15738046e+00 -7.42420554e-01
1.06477000e-01 1.22998647e-01 -1.33390108e-03 -1.34731698e+00
-9.01667595e-01 9.25451219e-01 3.94473642e-01 6.30352348e-02
9.31491494e-01 1.00646652e-01 4.84436214e-01 9.99228954e-01
3.46329987e-01 -1.03868699e+00 2.54810452e-01 7.37312615e-01
7.31579423e-01 -1.55430079e+00 -4.16090548e-01 -3.11150663e-02
-8.70785594e-01 9.52505529e-01 9.10116374e-01 -6.87477440e-02
5.08978546e-01 1.21118957e-02 2.84024268e-01 1.99767649e-01
-1.36969256e+00 -1.70886293e-01 3.24666440e-01 1.95051700e-01
5.19586444e-01 4.74714451e-02 -4.75222804e-03 3.61706913e-01
-3.03580999e-01 2.19416525e-02 1.35080129e-01 4.76109415e-01
-4.29859549e-01 -1.14852834e+00 -9.38927755e-02 4.98878926e-01
-5.72140634e-01 -3.63248318e-01 -2.51655787e-01 4.81818438e-01
6.61784783e-02 9.02289808e-01 8.15250501e-02 -3.06133568e-01
2.13953272e-01 2.77321011e-01 3.76736522e-01 -8.40073526e-01
-9.78810668e-01 -1.63339004e-01 4.60112453e-01 -5.24364889e-01
-5.02517700e-01 -2.06949741e-01 -1.03293610e+00 -1.28489077e-01
-4.65918869e-01 4.58696127e-01 6.53519928e-01 1.18930876e+00
3.57363075e-01 5.68806052e-01 6.03357673e-01 -5.41085184e-01
-1.11647022e+00 -1.19810653e+00 6.43052831e-02 4.17549342e-01
1.62620634e-01 -5.22475660e-01 -3.63798946e-01 5.07628210e-02] | [11.120630264282227, 8.314643859863281] |
b4ad1a4e-4da6-4077-bfa3-5040ef0d922f | tuning-ir-cut-filter-for-illumination-aware | 2103.14708 | null | https://arxiv.org/abs/2103.14708v1 | https://arxiv.org/pdf/2103.14708v1.pdf | Tuning IR-cut Filter for Illumination-aware Spectral Reconstruction from RGB | To reconstruct spectral signals from multi-channel observations, in particular trichromatic RGBs, has recently emerged as a promising alternative to traditional scanning-based spectral imager. It has been proven that the reconstruction accuracy relies heavily on the spectral response of the RGB camera in use. To improve accuracy, data-driven algorithms have been proposed to retrieve the best response curves of existing RGB cameras, or even to design brand new three-channel response curves. Instead, this paper explores the filter-array based color imaging mechanism of existing RGB cameras, and proposes to design the IR-cut filter properly for improved spectral recovery, which stands out as an in-between solution with better trade-off between reconstruction accuracy and implementation complexity. We further propose a deep learning based spectral reconstruction method, which allows to recover the illumination spectrum as well. Experiment results with both synthetic and real images under daylight illumination have shown the benefits of our IR-cut filter tuning method and our illumination-aware spectral reconstruction method. | ['Yinqiang Zheng', 'Xiao Zhou', 'Junchi Yan', 'Bo Sun'] | 2021-03-26 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Sun_Tuning_IR-Cut_Filter_for_Illumination-Aware_Spectral_Reconstruction_From_RGB_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Sun_Tuning_IR-Cut_Filter_for_Illumination-Aware_Spectral_Reconstruction_From_RGB_CVPR_2021_paper.pdf | cvpr-2021-1 | ['spectral-reconstruction'] | ['computer-vision'] | [ 6.44743860e-01 -7.02078938e-01 1.40326083e-01 -1.82744250e-01
-7.30711877e-01 -6.39908671e-01 1.38190970e-01 -6.68395936e-01
-2.80533791e-01 4.78866398e-01 6.86846823e-02 -1.67192474e-01
-3.23489249e-01 -5.84936500e-01 -5.49075723e-01 -1.06339943e+00
4.40179229e-01 -1.30958572e-01 1.06525943e-01 -2.72132128e-01
2.88304359e-01 4.14965957e-01 -1.69850981e+00 1.70432582e-01
1.00292861e+00 1.26764226e+00 4.44629252e-01 6.87626064e-01
-1.51262984e-01 4.33920801e-01 -3.29733402e-01 -6.81020021e-02
7.75556147e-01 -5.67225575e-01 -2.35196412e-01 3.82707685e-01
3.32416236e-01 -5.30558765e-01 -4.12494183e-01 1.08967423e+00
5.81783831e-01 -1.13214821e-01 4.36511487e-01 -7.29077876e-01
-7.10417151e-01 2.11091921e-01 -9.39140856e-01 -2.86477357e-01
3.72693449e-01 4.09586817e-01 7.58402169e-01 -6.03453636e-01
-3.23794186e-02 7.63552725e-01 6.83906913e-01 1.94257185e-01
-1.29079306e+00 -7.73955822e-01 -4.32613462e-01 4.30270523e-01
-1.40588546e+00 -4.71268058e-01 1.14754558e+00 -6.06486052e-02
4.33668286e-01 2.33100906e-01 1.04677904e+00 8.91033828e-01
1.47563946e-02 3.66386414e-01 1.88680518e+00 -5.77759862e-01
-2.41445284e-02 9.19113532e-02 -4.81085956e-01 5.10994673e-01
2.31385067e-01 3.78950596e-01 -5.15861273e-01 8.79984945e-02
1.03257740e+00 -8.54818523e-02 -8.35498095e-01 -2.44402930e-01
-9.77190137e-01 4.04787630e-01 4.26401943e-01 -6.84395730e-02
-1.51802778e-01 1.20040685e-01 -1.08902911e-02 3.54973525e-01
1.70351565e-01 5.25602996e-01 -3.92971933e-01 2.26573065e-01
-7.63627410e-01 -2.42648929e-01 4.02321100e-01 6.38361037e-01
9.29529488e-01 1.98484659e-01 2.45611891e-01 1.11304498e+00
4.80246425e-01 9.82029438e-01 3.17586303e-01 -1.01228273e+00
4.60019931e-02 5.87345004e-01 4.22850162e-01 -6.21675789e-01
-5.44499815e-01 -4.71350998e-01 -7.58031189e-01 1.20486945e-01
3.94117534e-01 -4.80318666e-02 -7.75340915e-01 1.33795691e+00
2.33973935e-01 1.78322449e-01 6.10001625e-05 1.51489580e+00
4.04937625e-01 6.28300428e-01 -4.90538865e-01 -4.90111470e-01
1.13505256e+00 -5.01312613e-01 -6.26591563e-01 8.32764804e-03
2.16010828e-02 -1.09993029e+00 1.44869316e+00 7.84472525e-01
-8.98336232e-01 -5.59123039e-01 -1.02659464e+00 7.71165639e-03
-2.32771318e-02 4.27659631e-01 7.62253881e-01 1.03854907e+00
-9.34063911e-01 3.33858311e-01 -4.09781516e-01 -2.83717155e-01
2.65383325e-03 9.01294574e-02 -2.38402206e-02 -1.31669790e-01
-8.61501992e-01 7.46314168e-01 3.42120945e-01 3.23677540e-01
-4.29526359e-01 -6.12700105e-01 -1.35405585e-01 -2.43285805e-01
5.37511945e-01 -5.35584211e-01 9.14218187e-01 -1.37644863e+00
-2.32315159e+00 8.21034431e-01 9.90637690e-02 1.57031137e-02
1.56862020e-01 -8.91922191e-02 -7.31168985e-01 3.14240187e-01
-5.28415561e-01 -8.70072544e-02 1.07252014e+00 -1.46617198e+00
-3.11012536e-01 -2.94440806e-01 -3.77118289e-02 1.99130371e-01
-5.12356639e-01 -1.15629256e-01 -3.70561540e-01 -2.34079808e-01
5.20196617e-01 -8.47817600e-01 -1.21476864e-02 2.54877508e-02
-3.60631436e-01 4.08730567e-01 6.63202286e-01 -5.29898584e-01
9.52311754e-01 -1.88953209e+00 -1.92720398e-01 1.64987907e-01
-3.00714314e-01 4.41272438e-01 -1.65002957e-01 4.86530155e-01
-1.71027228e-01 -3.39272320e-01 -3.44006479e-01 2.39311427e-01
-4.90429193e-01 -1.25647962e-01 -2.14490876e-01 8.52427304e-01
2.98141669e-02 4.55364078e-01 -4.91863996e-01 -1.28349960e-01
4.86260921e-01 5.61096489e-01 -3.42147350e-01 2.25902930e-01
-1.73108384e-01 5.82600534e-01 -2.40555152e-01 8.69155347e-01
1.12458062e+00 -2.03642339e-01 3.47700566e-01 -8.98452938e-01
-4.70988274e-01 -1.58285871e-01 -1.37432969e+00 1.98864961e+00
-7.59331226e-01 3.59676391e-01 3.87188852e-01 -7.24740088e-01
9.50797319e-01 9.78745371e-02 6.16384149e-01 -1.12360704e+00
1.39570057e-01 4.26926970e-01 -2.00706556e-01 -7.55397916e-01
4.52360749e-01 -3.71550977e-01 3.71633083e-01 2.72449374e-01
-3.07348907e-01 -2.40918174e-01 -3.16010624e-01 -3.67228895e-01
6.09407783e-01 7.15541780e-01 1.99747324e-01 -1.94251627e-01
7.72577524e-01 1.27782242e-03 4.04877931e-01 4.84237194e-01
1.50413796e-01 7.90382743e-01 -1.00565396e-01 -3.17942291e-01
-1.01759303e+00 -9.32326734e-01 -1.79105774e-01 7.69756436e-01
5.41267693e-01 4.94983904e-02 -4.84720945e-01 2.43333966e-01
-1.50261164e-01 4.15562630e-01 -6.58003166e-02 -7.22613558e-02
-4.84257936e-01 -1.00397038e+00 2.14798450e-01 2.43498143e-02
1.00975966e+00 -4.16139275e-01 -8.61578047e-01 6.19118251e-02
-1.85383171e-01 -8.92027140e-01 -6.86880946e-02 1.68156385e-01
-7.74688661e-01 -1.38760340e+00 -4.66054201e-01 -1.42230317e-01
2.98444957e-01 9.26432371e-01 7.64496446e-01 -7.01928511e-02
-5.05290151e-01 6.02108300e-01 -6.60649180e-01 -1.82523072e-01
-7.40982071e-02 -2.58816212e-01 -1.60210222e-01 5.30607224e-01
-2.28161574e-03 -6.84563398e-01 -1.11351228e+00 4.16624755e-01
-1.01424789e+00 3.48667115e-01 8.82886112e-01 6.66661143e-01
5.92052162e-01 6.59041554e-02 2.93616503e-01 -8.08308005e-01
3.00518930e-01 -2.61871237e-02 -1.11253953e+00 4.98391360e-01
-7.59426832e-01 -4.31249328e-02 8.20429921e-01 -2.06366614e-01
-1.37772131e+00 3.52035791e-01 -6.27245679e-02 -3.00601006e-01
5.80745889e-03 2.96365887e-01 -2.92134613e-01 -5.60083926e-01
7.30068922e-01 4.37880456e-01 4.63273786e-02 -5.16389072e-01
4.34883684e-01 9.54633474e-01 6.40279353e-01 -5.89151084e-01
7.87708461e-01 8.38238180e-01 1.26831710e-01 -1.22385216e+00
-6.44810855e-01 -7.75103033e-01 -3.89539927e-01 -5.34699738e-01
6.22394025e-01 -1.14228570e+00 -1.12677038e+00 6.55582130e-01
-9.76483762e-01 -2.70911694e-01 1.54223368e-01 4.94743735e-01
-4.88996714e-01 4.55559701e-01 -4.15893018e-01 -9.53653336e-01
-2.54317760e-01 -1.15820718e+00 1.16444588e+00 4.15204406e-01
5.66955507e-01 -5.75827479e-01 1.04146020e-03 5.20775557e-01
6.89818382e-01 1.97644159e-02 7.29030669e-01 5.87967098e-01
-1.03854430e+00 -1.17470445e-02 -7.19234407e-01 2.42666975e-01
1.75908238e-01 8.04493055e-02 -1.44887078e+00 -3.01336646e-01
1.76471129e-01 -2.55794495e-01 9.83148694e-01 4.70691502e-01
1.35986865e+00 2.08024886e-02 -1.75709673e-03 1.23929405e+00
2.25070643e+00 5.62211052e-02 9.49012816e-01 4.52894807e-01
7.41715312e-01 4.96614963e-01 4.74326909e-01 5.94772935e-01
1.12204559e-01 6.69673443e-01 7.40382969e-01 -3.38847220e-01
-2.79126465e-01 3.88672799e-02 2.96908915e-01 6.58651471e-01
-3.55411708e-01 -3.20892304e-01 -5.16790271e-01 -2.22720541e-02
-1.55987322e+00 -7.11169004e-01 -5.24758458e-01 2.42874813e+00
7.89149880e-01 -4.86441642e-01 1.38086721e-01 8.36350769e-02
4.83775198e-01 7.30661526e-02 -6.46560967e-01 1.28390804e-01
-6.03983819e-01 4.83937919e-01 1.12609208e+00 2.81401098e-01
-7.53123224e-01 6.96335435e-01 6.58260012e+00 7.39270449e-01
-1.59340942e+00 5.97889361e-04 1.90921381e-01 -8.02125707e-02
-5.54445028e-01 1.46514416e-01 -3.09029847e-01 3.00720036e-01
8.15567553e-01 1.86085328e-01 1.01712501e+00 2.57560372e-01
5.75776398e-01 -5.06414115e-01 -4.84972030e-01 1.47031760e+00
1.26418825e-02 -1.06210327e+00 -4.66061272e-02 9.78891775e-02
5.63905418e-01 -2.52397895e-01 2.54416883e-01 -5.30254841e-01
-4.56530340e-02 -7.22033739e-01 7.12697327e-01 7.47865677e-01
1.01890612e+00 -6.09964609e-01 3.79890919e-01 1.10501893e-01
-1.15555406e+00 -3.36496562e-01 -5.49433291e-01 7.19092190e-02
-1.87405437e-01 6.84159279e-01 -5.66659510e-01 7.51465559e-01
8.20998847e-01 5.77947438e-01 -5.22888660e-01 1.03753304e+00
-3.00669491e-01 6.59617782e-01 -2.48370618e-01 1.03374362e-01
-2.01382384e-01 -9.05100644e-01 2.70017654e-01 9.45723712e-01
5.31466722e-01 3.48256886e-01 1.53878033e-01 1.15279007e+00
1.71460733e-01 8.01177919e-02 -3.31561476e-01 -1.56024024e-01
2.07259163e-01 1.61999607e+00 -5.92173576e-01 3.99640530e-01
-6.87989771e-01 9.31540728e-01 -3.30186903e-01 6.13459110e-01
-9.20343816e-01 -1.67703122e-01 5.64267695e-01 3.40009451e-01
5.38010150e-02 -4.73899066e-01 -2.43568212e-01 -1.08740532e+00
-8.63483101e-02 -7.78725266e-01 5.18058240e-02 -1.28851283e+00
-1.11693978e+00 9.22080949e-02 -3.69228929e-01 -1.52003002e+00
5.21039963e-01 -9.67372775e-01 -1.97545946e-01 1.00842977e+00
-1.96161151e+00 -1.51015222e+00 -5.04934013e-01 8.49343657e-01
1.92396805e-01 9.60770249e-02 6.62957370e-01 3.27367336e-01
-4.42389101e-01 9.45590883e-02 3.88063997e-01 -3.62907916e-01
9.10491943e-01 -8.43664944e-01 -3.88653755e-01 8.86173606e-01
-9.09592956e-02 4.81457651e-01 7.65608490e-01 -2.84091234e-01
-2.37427974e+00 -7.08274007e-01 -4.49855387e-01 1.83941886e-01
4.96954918e-01 -8.12676772e-02 -7.30445564e-01 -4.04219935e-03
3.77099514e-01 -2.19536856e-01 7.47362077e-01 -2.23444179e-01
-5.34061968e-01 -8.35136652e-01 -8.82358909e-01 3.47336471e-01
1.00248301e+00 -7.42434502e-01 1.36905536e-01 3.63931209e-01
4.89195347e-01 -2.05428198e-01 -6.01983249e-01 3.93673509e-01
9.17731702e-01 -1.54489934e+00 1.16059673e+00 2.06888214e-01
2.95145720e-01 -7.83663571e-01 -3.39622319e-01 -1.28858483e+00
-2.81964302e-01 -5.40296793e-01 4.51643497e-01 1.06363618e+00
1.89501837e-01 -8.41101766e-01 6.68479860e-01 1.67799205e-01
-2.12349087e-01 -2.47681320e-01 -4.73646671e-01 -6.95512950e-01
-4.37822372e-01 -4.16543216e-01 5.81350386e-01 7.52182543e-01
-5.94783008e-01 9.21417773e-02 -6.33781493e-01 5.60941696e-01
8.53946805e-01 6.87698126e-01 8.64962041e-01 -1.20555162e+00
-5.19844711e-01 -3.05827975e-01 1.52366772e-01 -9.66788530e-01
-7.73912966e-02 -5.82657218e-01 6.73404634e-02 -1.32998514e+00
1.71097592e-01 -4.21318740e-01 -2.50284374e-01 5.08145429e-02
2.78798155e-02 4.28142816e-01 -5.33533953e-02 4.01736110e-01
-3.79247516e-01 4.05513674e-01 1.42566657e+00 -4.28740866e-02
-3.97183210e-01 -8.93909782e-02 -7.21518338e-01 5.28384030e-01
6.19951785e-01 5.32612950e-02 -4.93908614e-01 -5.31557977e-01
7.47818530e-01 2.75518060e-01 5.12487829e-01 -1.16570842e+00
2.84562707e-01 -4.26995218e-01 4.47570145e-01 -3.71018142e-01
5.24553835e-01 -1.11417091e+00 5.97059548e-01 2.50821233e-01
-3.04015186e-02 -5.01953602e-01 -6.49690703e-02 7.08152771e-01
-6.01495393e-02 5.70793040e-02 1.06388950e+00 -4.03516293e-01
-1.02859771e+00 5.83360605e-02 -4.97273467e-02 -3.14340234e-01
6.45389080e-01 -6.17896974e-01 -4.34046477e-01 -2.67453492e-01
1.00441940e-01 -2.46151015e-01 7.03452885e-01 -1.67887524e-01
7.00777471e-01 -1.07894921e+00 -3.96704465e-01 2.36785546e-01
1.84107363e-01 -3.97269815e-01 4.34450477e-01 8.79607439e-01
-9.17600214e-01 1.93921790e-01 -2.79029578e-01 -6.91710055e-01
-1.02994871e+00 3.77107024e-01 5.61462879e-01 4.53125954e-01
-4.63419527e-01 7.49332130e-01 -2.34813571e-01 -2.33583137e-01
-1.36819273e-01 -2.16002032e-01 6.95816502e-02 -1.45151064e-01
3.03842962e-01 4.55544084e-01 2.40015760e-01 -4.24375832e-01
-6.60707429e-02 1.04563284e+00 6.17961407e-01 3.16133872e-02
1.37312305e+00 -4.79787737e-01 -4.09996033e-01 4.74591106e-01
9.79479313e-01 1.36942402e-01 -1.36712801e+00 -1.68616042e-01
-3.45484108e-01 -9.83000636e-01 3.91515642e-01 -9.28495407e-01
-1.23530936e+00 7.86010444e-01 8.75078321e-01 2.17936963e-01
1.80831051e+00 -4.95132327e-01 7.37909853e-01 1.59698501e-01
6.13351464e-01 -1.29166532e+00 2.10540116e-01 2.25123931e-02
5.97183049e-01 -1.26101542e+00 2.77432442e-01 -5.67223310e-01
-2.20264629e-01 1.55642259e+00 2.91931957e-01 8.34500864e-02
3.12143624e-01 5.22492789e-02 1.91411659e-01 3.32106426e-02
-2.05973387e-01 -5.03291428e-01 1.32567108e-01 6.01620913e-01
4.41725016e-01 1.56926364e-01 -7.16353906e-03 8.71333182e-02
1.00715809e-01 1.07379816e-01 5.58521569e-01 4.57211703e-01
-6.93919480e-01 -1.07166231e+00 -7.88954020e-01 2.78523862e-01
-9.75020826e-02 1.70237431e-03 -2.01293394e-01 5.22059500e-01
2.42403552e-01 1.15912151e+00 -4.25313205e-01 -5.03250718e-01
4.09194469e-01 2.23864783e-02 7.99127281e-01 -1.68526903e-01
-3.09084862e-01 5.27051687e-01 -2.48918593e-01 -7.53175497e-01
-8.85291994e-01 -3.58002871e-01 -9.68075097e-01 -2.01352865e-01
-6.76433086e-01 -2.75605917e-01 8.44699442e-01 6.45696402e-01
7.51519799e-02 3.49695235e-01 8.80746722e-01 -7.10148513e-01
-3.47034305e-01 -6.14926875e-01 -1.04961789e+00 3.06398839e-01
3.85586590e-01 -5.66107333e-01 -4.22535956e-01 -2.94887535e-02] | [10.299229621887207, -2.5225107669830322] |
74fea1c3-dfb7-4e3b-966d-ca5727aecf28 | batchgfn-generative-flow-networks-for-batch | 2306.15058 | null | https://arxiv.org/abs/2306.15058v1 | https://arxiv.org/pdf/2306.15058v1.pdf | BatchGFN: Generative Flow Networks for Batch Active Learning | We introduce BatchGFN -- a novel approach for pool-based active learning that uses generative flow networks to sample sets of data points proportional to a batch reward. With an appropriate reward function to quantify the utility of acquiring a batch, such as the joint mutual information between the batch and the model parameters, BatchGFN is able to construct highly informative batches for active learning in a principled way. We show our approach enables sampling near-optimal utility batches at inference time with a single forward pass per point in the batch in toy regression problems. This alleviates the computational complexity of batch-aware algorithms and removes the need for greedy approximations to find maximizers for the batch reward. We also present early results for amortizing training across acquisition steps, which will enable scaling to real-world tasks. | ['Yarin Gal', 'Yoshua Bengio', 'Tristan Deleu', 'Nikolay Malkin', 'Moksh Jain', 'Andrew Jesson', 'Salem Lahlou', 'Shreshth A. Malik'] | 2023-06-26 | null | null | null | null | ['active-learning', 'active-learning'] | ['methodology', 'natural-language-processing'] | [-4.89631034e-02 6.20738864e-01 -3.30037564e-01 -6.68101490e-01
-1.31287718e+00 -3.13059241e-01 5.05450666e-01 6.79153129e-02
-7.52873123e-01 1.11123538e+00 2.29891956e-01 -3.10702678e-02
-3.31724614e-01 -6.34448290e-01 -9.22050953e-01 -8.19453597e-01
-4.01968300e-01 9.16513324e-01 5.89144900e-02 3.20313632e-01
1.83398634e-01 3.05246860e-01 -1.15237105e+00 1.77819312e-01
6.59180701e-01 8.34656715e-01 2.85695851e-01 8.58709455e-01
-1.71426535e-01 1.01386440e+00 -7.41651058e-01 -4.38561559e-01
4.35850441e-01 -6.86696112e-01 -7.78872430e-01 2.18113646e-01
2.48397160e-02 -4.79150414e-01 2.05210656e-01 6.33573651e-01
4.39440787e-01 5.64063728e-01 6.42805099e-01 -1.11470139e+00
-1.96231961e-01 1.10603082e+00 -3.44427228e-01 1.73475668e-01
-3.86055447e-02 3.69014978e-01 1.29770267e+00 -8.14144075e-01
6.42845273e-01 1.05794466e+00 4.37027633e-01 5.55790365e-01
-1.45961475e+00 -4.70065445e-01 3.73247832e-01 -3.95091295e-01
-9.46131229e-01 -9.67324615e-01 4.31703627e-01 -1.74282402e-01
9.17914391e-01 1.16590172e-01 1.02723837e+00 8.59502912e-01
-6.21842220e-02 1.36713922e+00 6.39519215e-01 -3.69379103e-01
7.62170792e-01 2.11100668e-01 -6.36509955e-02 7.41469562e-01
1.21091440e-01 -9.44661628e-03 -9.47966516e-01 -4.16966558e-01
9.61327434e-01 1.97393671e-01 -2.96640955e-02 -6.69072330e-01
-9.03402627e-01 1.12193406e+00 2.35992014e-01 -1.59012929e-01
-3.33450615e-01 5.28874338e-01 2.23827362e-01 4.27792788e-01
8.87909889e-01 8.04381311e-01 -5.23240805e-01 -3.23860347e-01
-1.29511511e+00 4.13751960e-01 1.03214729e+00 9.97865558e-01
8.23657751e-01 -3.88732506e-03 -3.37035328e-01 7.55170345e-01
7.21490145e-01 1.05643839e-01 1.03853941e-01 -1.19199169e+00
5.08737445e-01 4.39227641e-01 3.30233157e-01 -1.60841197e-02
-6.70527741e-02 -4.20351714e-01 -7.77495503e-02 3.04055333e-01
6.43143833e-01 -4.93101299e-01 -7.30557024e-01 1.52551854e+00
4.73842114e-01 -7.83799514e-02 -2.34207109e-01 4.76799965e-01
2.14499906e-01 5.91185689e-01 1.18437344e-02 -6.84732318e-01
4.33354318e-01 -9.09850717e-01 -3.17928702e-01 -3.53935868e-01
7.43895113e-01 -3.58502090e-01 1.03205550e+00 4.20398533e-01
-1.70549560e+00 -5.06838448e-02 -8.21347296e-01 2.06854641e-01
1.99217647e-01 -2.53120512e-01 1.08468151e+00 4.58352655e-01
-1.10517454e+00 8.60952616e-01 -1.18090963e+00 2.19600677e-01
1.02337551e+00 6.35453045e-01 -1.10020094e-01 -1.61201898e-02
-4.55609500e-01 4.76116806e-01 4.69850153e-01 1.22461163e-01
-1.39759314e+00 -1.06090069e+00 -7.20427513e-01 1.46276236e-01
5.66957712e-01 -5.71200252e-01 1.66213703e+00 -1.13418555e+00
-1.70173478e+00 4.19688135e-01 -2.34982386e-01 -7.04023778e-01
5.69145441e-01 -4.46184993e-01 4.28071946e-01 -1.86283793e-02
-7.20768496e-02 7.95550108e-01 9.46583331e-01 -8.70919883e-01
-3.29682142e-01 -9.82667878e-02 1.01966202e-01 5.36572933e-01
-1.23718806e-01 3.80520038e-02 -3.34613979e-01 -1.98796943e-01
2.81975828e-02 -6.96935654e-01 -6.62110448e-01 2.47985139e-01
-1.43695518e-01 -4.20715988e-01 2.30992675e-01 -1.16791852e-01
8.04609537e-01 -2.01648831e+00 3.61719690e-02 2.21504539e-01
4.36781555e-01 -1.86014786e-01 -8.90523382e-03 3.38829190e-01
3.42909813e-01 -4.43710499e-02 -3.63939792e-01 -8.41307163e-01
-1.94329843e-02 3.54124576e-01 -2.52416342e-01 4.98795211e-01
4.57152665e-01 1.02502954e+00 -1.21222973e+00 -6.82205200e-01
-9.28815231e-02 -5.03531732e-02 -7.35379517e-01 6.32629037e-01
-6.88815475e-01 2.21697986e-01 -3.67801011e-01 2.50878841e-01
2.69136816e-01 -3.60915512e-01 1.49760097e-01 6.04467452e-01
2.55905509e-01 5.82558692e-01 -1.12837160e+00 1.66244674e+00
-5.83413124e-01 5.51826835e-01 1.47073627e-01 -9.60938990e-01
8.14638674e-01 2.42010087e-01 5.85731506e-01 -2.02697709e-01
-7.42697492e-02 1.34934232e-01 -1.61588952e-01 -8.62529427e-02
3.98344725e-01 -2.56885260e-01 5.03075682e-02 9.37670648e-01
6.47988975e-01 -1.72603250e-01 3.64965230e-01 4.27442670e-01
1.12214184e+00 4.93232459e-01 3.87659431e-01 -2.94715822e-01
-2.03887820e-01 -1.70567125e-01 4.12781507e-01 1.19923961e+00
3.50391939e-02 6.41080379e-01 6.77718580e-01 -2.42007226e-01
-1.02810562e+00 -1.26964426e+00 1.54973730e-01 1.46555972e+00
-4.83526260e-01 -2.57061899e-01 -3.57611239e-01 -1.03102481e+00
-9.92118791e-02 8.50400865e-01 -6.72232926e-01 2.17545219e-02
-4.58624572e-01 -8.44181955e-01 1.35697111e-01 5.50785661e-01
-5.00009321e-02 -1.29164600e+00 -5.31807303e-01 2.63731271e-01
3.23545933e-01 -4.57627803e-01 -5.74369609e-01 8.77708256e-01
-1.41738605e+00 -7.52536714e-01 -6.41787291e-01 -4.26089048e-01
9.71372724e-01 -2.85892397e-01 1.45327318e+00 -8.66654515e-02
-2.35016242e-01 3.14761370e-01 -9.41776782e-02 -5.34396410e-01
-3.58451277e-01 2.16874018e-01 -4.23289269e-01 -2.00818703e-01
5.92578612e-02 -4.35588598e-01 -5.84001303e-01 2.74576880e-02
-5.92532337e-01 3.31992342e-04 5.87169766e-01 8.88435066e-01
6.38934076e-01 -4.83205914e-01 6.60737872e-01 -1.28497350e+00
6.76823020e-01 -7.19907761e-01 -9.16768551e-01 1.47540346e-01
-6.69502556e-01 5.13272047e-01 4.89379734e-01 -5.86200237e-01
-1.27030253e+00 3.10980111e-01 1.07972577e-01 -2.90815622e-01
1.98971927e-01 1.58571899e-01 1.54063374e-01 3.16186637e-01
9.12935972e-01 3.01520731e-02 5.39511107e-02 -2.55801082e-01
4.64947551e-01 2.58895874e-01 -6.06963187e-02 -7.18163013e-01
4.36230779e-01 2.43546620e-01 -1.79089308e-01 -5.82393289e-01
-1.16230094e+00 -6.63774014e-01 -4.15638864e-01 -1.88404292e-01
2.81152785e-01 -8.44902694e-01 -7.58983672e-01 5.52499034e-02
-7.57744491e-01 -8.97021234e-01 -1.27517092e+00 6.63539767e-01
-9.64563608e-01 -1.73826426e-01 -5.03153086e-01 -1.29422581e+00
-4.27660435e-01 -7.67520547e-01 8.87858450e-01 3.21287543e-01
-3.14594209e-01 -1.24456751e+00 4.25984502e-01 1.88458264e-01
3.40711772e-01 -4.06018436e-01 4.57954794e-01 -1.07734072e+00
-9.31609452e-01 -2.40372956e-01 3.29532683e-01 2.25110978e-01
-2.04644218e-01 -9.51147079e-02 -9.43727076e-01 -2.33969942e-01
2.55073495e-02 -8.37662041e-01 9.64870870e-01 6.59404337e-01
1.02995265e+00 -5.26958108e-01 -1.60360426e-01 4.15379941e-01
1.20581830e+00 1.76455304e-02 5.93328834e-01 -1.80604354e-01
3.06207597e-01 1.99468866e-01 5.92752218e-01 6.61658406e-01
-2.07066387e-01 2.41619036e-01 1.50822163e-01 1.00746319e-01
2.79637694e-01 -3.03811163e-01 4.13146079e-01 5.22798777e-01
-3.81759368e-03 -3.52783680e-01 -4.25379634e-01 6.30104363e-01
-1.87881410e+00 -1.06037223e+00 4.63588297e-01 2.52674818e+00
1.38013339e+00 4.89452124e-01 3.70144635e-01 -2.46155158e-01
2.06866354e-01 1.13341711e-01 -7.55130529e-01 -2.82822967e-01
3.52591485e-01 5.52862167e-01 4.59064662e-01 6.84730351e-01
-7.63540924e-01 7.49832332e-01 7.76580906e+00 7.64831722e-01
-5.63528538e-01 1.64592862e-01 8.44741046e-01 -8.21754575e-01
-5.37625313e-01 2.65614778e-01 -9.96259928e-01 2.36012682e-01
1.08140349e+00 3.74584533e-02 5.40603340e-01 1.11901069e+00
1.24722153e-01 -4.68838155e-01 -1.44513917e+00 4.69282746e-01
-2.13062897e-01 -1.48336411e+00 -4.08009291e-01 1.95063189e-01
7.81633735e-01 3.90021324e-01 -3.12164664e-01 3.60019356e-01
1.03954434e+00 -9.02880430e-01 7.54803002e-01 6.88588738e-01
3.14553261e-01 -7.00869679e-01 3.46912891e-01 7.26849794e-01
-6.65758312e-01 -1.38275146e-01 -5.81429541e-01 8.03405046e-03
1.47744402e-01 8.73352051e-01 -1.59830499e+00 -6.36213347e-02
2.41270989e-01 3.02552968e-01 -2.41536170e-01 1.28159225e+00
-1.58973560e-01 1.05274498e+00 -7.56815076e-01 -4.67521936e-01
1.01628140e-01 -3.74721438e-01 3.70929897e-01 1.09576380e+00
1.34634320e-02 -3.04160327e-01 1.76037282e-01 1.12737477e+00
-1.93342403e-01 3.14468108e-02 -5.22209167e-01 -3.62853736e-01
5.26986599e-01 1.26993370e+00 -8.68135631e-01 -3.03003877e-01
-1.13842525e-02 7.41314888e-01 7.89752424e-01 1.74823478e-01
-4.91795868e-01 -1.72073051e-01 1.88960180e-01 1.53941184e-01
4.55606192e-01 -3.11447799e-01 -1.82569385e-01 -8.87733996e-01
-7.94020519e-02 -5.79077899e-01 3.61354351e-01 -4.96450067e-01
-1.20469594e+00 8.63750950e-02 3.23629767e-01 -7.47438550e-01
-8.69080722e-01 -1.62606791e-01 -7.64331937e-01 8.89360785e-01
-8.30288172e-01 -5.80801010e-01 2.50669986e-01 3.34942758e-01
8.53478312e-01 -1.51871875e-01 6.81238174e-01 -3.11371267e-01
-4.29401219e-01 6.12778068e-01 1.22507498e-01 2.38013156e-02
3.09994489e-01 -1.52200353e+00 6.11059308e-01 6.78687155e-01
5.62440515e-01 9.02609169e-01 6.29096031e-01 -5.31367540e-01
-1.09760523e+00 -6.15716696e-01 5.83842993e-01 -5.15894651e-01
4.63504225e-01 -5.91610909e-01 -5.55086970e-01 7.96296656e-01
8.45778137e-02 1.22165836e-01 8.23752344e-01 5.75675666e-01
-3.29897553e-02 -1.94507539e-01 -1.11075032e+00 4.14201379e-01
1.05814886e+00 -2.67259091e-01 -2.27890491e-01 5.05474806e-01
6.82770133e-01 -3.45546663e-01 -7.46335328e-01 -1.86705574e-01
4.85784143e-01 -9.65624392e-01 6.66153431e-01 -7.04748690e-01
2.93023735e-01 1.75101265e-01 2.31295362e-01 -1.25754189e+00
-1.02966920e-01 -1.01544201e+00 -5.77726781e-01 1.02334690e+00
8.92783642e-01 -5.09214163e-01 1.11509907e+00 7.46718824e-01
-1.01463236e-01 -1.21558762e+00 -7.19390631e-01 -4.91455078e-01
-1.52247787e-01 -4.23394322e-01 2.66363651e-01 3.18549633e-01
-1.12012766e-01 4.17310536e-01 -2.49069750e-01 -4.74513680e-01
8.23569953e-01 -9.39405411e-02 5.92298985e-01 -1.17807567e+00
-9.90343273e-01 -1.00463539e-01 8.42217356e-02 -1.24927235e+00
-5.51111326e-02 -7.13456511e-01 4.85263556e-01 -1.50828040e+00
3.30523163e-01 -7.85584807e-01 -2.82730699e-01 5.10814786e-01
-1.57524943e-01 -1.08732931e-01 1.62222609e-01 1.83860525e-01
-1.02195632e+00 4.06789720e-01 1.14495826e+00 1.74569994e-01
-5.38285136e-01 5.35517812e-01 -8.30787003e-01 5.71883619e-01
6.98098481e-01 -1.01382077e+00 -9.61082757e-01 -1.05427496e-01
6.31924212e-01 2.40264043e-01 4.03569229e-02 -5.72310567e-01
1.99463874e-01 -4.30271477e-01 7.21806645e-01 -4.94673073e-01
4.02422249e-01 -3.77738684e-01 8.96451697e-02 1.73529938e-01
-1.03443098e+00 -3.34009796e-01 -3.03289771e-01 6.56134903e-01
3.13397735e-01 -8.75658751e-01 7.35812724e-01 -5.33027291e-01
1.36465719e-03 3.68445665e-01 -2.62600899e-01 5.79647422e-01
7.93633521e-01 -1.18913226e-01 1.05825804e-01 -4.71482158e-01
-6.79503381e-01 2.14022785e-01 2.16944098e-01 -9.91348922e-02
5.46917081e-01 -8.43630195e-01 -6.43453360e-01 2.05957830e-01
-2.38490805e-01 6.21593833e-01 -2.19035909e-01 6.62064612e-01
-2.94732779e-01 -1.15456255e-02 1.25739306e-01 -5.40918410e-01
-8.40354741e-01 1.49832785e-01 4.62143451e-01 -6.49284720e-01
-4.52927947e-01 1.40478122e+00 7.84528553e-02 -2.31887400e-01
3.03626508e-01 -2.06434000e-02 1.85800239e-01 7.07294717e-02
4.84153509e-01 3.54440242e-01 1.74927171e-02 3.64538223e-01
1.90191586e-02 -4.09284294e-01 -4.49325085e-01 -6.32915199e-01
1.52896810e+00 1.41075060e-01 1.55221403e-01 8.92858028e-01
1.15443933e+00 -4.49940264e-02 -2.08521223e+00 -1.60891548e-01
1.95979560e-03 -4.08768535e-01 1.13450572e-01 -6.04108751e-01
-1.00600839e+00 4.95286822e-01 2.67031163e-01 1.63376704e-01
6.50012910e-01 2.98644692e-01 3.61970961e-01 7.45146573e-01
3.58218014e-01 -1.25053859e+00 2.64607191e-01 1.04209244e-01
5.11641204e-01 -1.01753438e+00 4.67595547e-01 5.88045344e-02
-7.17426181e-01 9.29351807e-01 3.52979153e-01 -2.82887876e-01
5.22451162e-01 5.46729088e-01 -2.66832262e-01 -1.73264563e-01
-1.18032467e+00 5.23915105e-02 1.77996829e-01 4.70092833e-01
5.55838764e-01 3.70749086e-02 8.50970149e-02 1.46855772e-01
3.68292704e-02 7.75080398e-02 2.69752771e-01 1.07426870e+00
-6.82809174e-01 -1.09659433e+00 1.09446853e-01 9.50505316e-01
-4.17145044e-01 -1.58187509e-01 -2.47884721e-01 5.31189501e-01
-3.34931284e-01 6.99618518e-01 3.24259907e-01 2.13478744e-01
-2.20904440e-01 2.11612329e-01 9.27772224e-01 -1.02374232e+00
-7.22359061e-01 2.06977978e-01 2.73809046e-01 -5.41154265e-01
-2.22028375e-01 -8.62755954e-01 -1.28299510e+00 -5.16840210e-03
-6.89870179e-01 4.01612401e-01 4.67613310e-01 1.20774138e+00
1.53944060e-01 1.37345567e-01 7.89209902e-01 -8.99104953e-01
-9.80484366e-01 -9.90922689e-01 -6.08374774e-01 2.30476004e-03
2.29169309e-01 -3.11572224e-01 -4.71776336e-01 -5.04522547e-02] | [8.226856231689453, 3.9991605281829834] |
c0ddf3b5-6d82-4940-9767-1e205d2b5f77 | continual-learning-for-automated-audio | 2107.08028 | null | https://arxiv.org/abs/2107.08028v1 | https://arxiv.org/pdf/2107.08028v1.pdf | Continual Learning for Automated Audio Captioning Using The Learning Without Forgetting Approach | Automated audio captioning (AAC) is the task of automatically creating textual descriptions (i.e. captions) for the contents of a general audio signal. Most AAC methods are using existing datasets to optimize and/or evaluate upon. Given the limited information held by the AAC datasets, it is very likely that AAC methods learn only the information contained in the utilized datasets. In this paper we present a first approach for continuously adapting an AAC method to new information, using a continual learning method. In our scenario, a pre-optimized AAC method is used for some unseen general audio signals and can update its parameters in order to adapt to the new information, given a new reference caption. We evaluate our method using a freely available, pre-optimized AAC method and two freely available AAC datasets. We compare our proposed method with three scenarios, two of training on one of the datasets and evaluating on the other and a third of training on one dataset and fine-tuning on the other. Obtained results show that our method achieves a good balance between distilling new knowledge and not forgetting the previous one. | ['Konstantinos Drossos', 'Jan Berg'] | 2021-07-16 | null | null | null | null | ['audio-captioning'] | ['audio'] | [ 6.42814517e-01 2.11701736e-01 2.18154699e-01 -3.39600325e-01
-1.04643118e+00 -8.47800732e-01 4.18619573e-01 2.85381377e-01
-4.39228356e-01 1.13624692e+00 6.61395371e-01 1.49013802e-01
-5.78359440e-02 -3.03091049e-01 -7.16210365e-01 -4.36693728e-01
1.85917527e-03 8.30132365e-01 4.17502314e-01 -5.52202873e-02
2.61204302e-01 1.48677096e-01 -1.84521544e+00 7.44305015e-01
7.47237742e-01 7.87592232e-01 6.01074517e-01 1.16285312e+00
-3.41105431e-01 6.06893063e-01 -7.58312106e-01 -3.45176309e-01
-1.05511948e-01 -4.40116853e-01 -1.14517701e+00 -1.22077763e-02
1.90142900e-01 5.15485033e-02 -9.40034818e-03 6.00604534e-01
6.73620224e-01 3.52008879e-01 4.52654541e-01 -1.10101044e+00
-1.31246224e-01 1.02602994e+00 3.23089331e-01 1.34134680e-01
7.38999665e-01 7.62181357e-02 8.38077843e-01 -6.09623969e-01
6.31818354e-01 9.51072216e-01 6.47221446e-01 7.73138523e-01
-9.79225814e-01 -4.19720441e-01 3.67043763e-01 3.65636677e-01
-1.32874310e+00 -7.72529483e-01 1.00062597e+00 -4.48687732e-01
5.53582311e-01 4.74438906e-01 7.96087384e-01 1.26636255e+00
-2.89377838e-01 7.73009300e-01 8.04665446e-01 -6.43344760e-01
4.64070857e-01 6.99861526e-01 -1.55603811e-01 5.91506734e-02
-2.15153173e-01 -3.41381013e-01 -6.75180078e-01 -3.93766582e-01
2.87029911e-02 -6.94156826e-01 -5.63064694e-01 -3.01099837e-01
-1.25051248e+00 3.19101185e-01 -9.04882997e-02 5.30039191e-01
-5.07266819e-01 2.00972818e-02 5.25640428e-01 2.27998361e-01
1.13967724e-01 7.78667748e-01 -8.03984106e-01 -6.90686882e-01
-9.44091260e-01 2.42196247e-01 1.01113236e+00 8.21227014e-01
7.50675797e-01 -1.94766685e-01 -3.77418280e-01 9.26864743e-01
1.53686911e-01 1.73076242e-01 9.86047328e-01 -6.19309723e-01
5.76660693e-01 1.37697086e-01 5.04596055e-01 -5.25451005e-01
-2.44817942e-01 -5.18800676e-01 -2.10141897e-01 -2.88506299e-01
2.30280787e-01 -3.20267677e-01 -9.13589776e-01 1.99919808e+00
2.50457019e-01 4.07532752e-01 1.52839065e-01 5.67713261e-01
5.79768062e-01 9.41885650e-01 1.24470942e-01 -5.98737657e-01
8.90832722e-01 -1.07267606e+00 -9.75342453e-01 -2.98404068e-01
3.70473683e-01 -1.07604897e+00 1.14080703e+00 6.15650594e-01
-8.58666718e-01 -7.14695513e-01 -1.07820249e+00 3.13796997e-01
-3.82835388e-01 1.33591041e-01 1.74457073e-01 5.49724698e-01
-1.11121297e+00 4.44355488e-01 -4.10991579e-01 -4.75894392e-01
-9.74332094e-02 4.27081019e-01 -1.17624216e-01 6.29941151e-02
-1.30436587e+00 5.99096119e-01 7.68222511e-01 -8.18641037e-02
-1.01530063e+00 -5.85640073e-01 -5.19382060e-01 4.64362130e-02
2.22530156e-01 -6.34122550e-01 1.67412925e+00 -1.33606505e+00
-1.99077010e+00 2.49106809e-01 -1.08391538e-01 -5.72634816e-01
5.61702609e-01 -4.49511856e-01 -7.31035769e-01 6.10781610e-02
-9.41624492e-02 8.44675899e-01 1.05787718e+00 -1.58766735e+00
-6.64338589e-01 2.96376012e-02 3.25631574e-02 4.48958129e-01
-5.74705303e-01 -4.98335361e-01 -5.16243160e-01 -6.76925063e-01
-2.55788833e-01 -1.02581847e+00 2.17511922e-01 -2.95697927e-01
-3.83176267e-01 1.19810350e-01 9.86583173e-01 -8.14271331e-01
1.59310341e+00 -2.02176356e+00 2.45381862e-01 1.65435687e-01
-4.22123671e-01 4.72183436e-01 -3.58946770e-01 5.82927823e-01
-1.64416477e-01 -1.99770667e-02 -3.62775743e-01 -4.72229928e-01
-2.52032727e-01 1.34337023e-01 -3.49342614e-01 -4.80049253e-02
-8.85192528e-02 2.32652202e-01 -1.11604965e+00 -5.10789037e-01
7.42732435e-02 6.49409890e-01 -6.42389297e-01 5.49443901e-01
-5.48963368e-01 8.07317615e-01 -3.19287151e-01 1.91304043e-01
3.68609220e-01 1.25235960e-01 2.35279202e-01 -3.39927971e-01
-1.13043942e-01 2.23598003e-01 -1.54659665e+00 1.96993148e+00
-8.67668152e-01 6.86237037e-01 -2.15327039e-01 -7.00404584e-01
8.72823417e-01 8.63319755e-01 4.20343429e-01 -2.27858648e-01
2.08623745e-02 4.00552899e-01 -1.58991367e-01 -8.61554563e-01
4.68149930e-01 2.04764586e-02 1.25587761e-01 4.73329693e-01
2.97132760e-01 -3.96359921e-01 2.98903137e-01 -7.39822909e-02
9.93426681e-01 2.99395770e-01 3.96083117e-01 1.10663906e-01
1.03494668e+00 -1.35912657e-01 1.49653062e-01 8.74026000e-01
-1.20575882e-01 7.91425705e-01 1.06540799e-01 -4.96901304e-01
-1.14427590e+00 -8.28247905e-01 5.74182682e-02 1.08364463e+00
-3.01310062e-01 -5.30543089e-01 -8.16833198e-01 -6.25442982e-01
-3.34365159e-01 1.04690182e+00 -4.79267895e-01 -2.01629370e-01
-6.03857934e-01 -3.04153025e-01 4.81612056e-01 6.52696118e-02
5.61925411e-01 -1.09924483e+00 -4.86533642e-01 6.10551417e-01
-5.59900224e-01 -1.01267898e+00 -6.11500323e-01 2.53080688e-02
-7.14123905e-01 -6.49069309e-01 -7.27663517e-01 -8.78695369e-01
4.14988488e-01 -2.52893627e-01 9.39596593e-01 -1.22937739e-01
1.98340133e-01 7.12583065e-01 -5.36499143e-01 -4.68475491e-01
-9.68344986e-01 5.34729362e-01 7.76970536e-02 4.46030289e-01
-2.21668378e-01 -7.45132565e-01 -3.64419073e-01 1.38246521e-01
-9.76265013e-01 2.05993414e-01 6.27349555e-01 5.48420548e-01
4.82779831e-01 -9.67348814e-02 7.79012620e-01 -8.20894361e-01
8.66334081e-01 -3.66794616e-01 -1.93366051e-01 4.69761044e-01
-5.34266114e-01 3.95628512e-01 7.63500750e-01 -8.71073723e-01
-1.16725123e+00 6.91549778e-01 -3.15097332e-01 -2.90890276e-01
-3.27581465e-01 4.88405645e-01 -1.73784658e-01 1.63407758e-01
7.26154745e-01 2.47468457e-01 -3.57765734e-01 -6.35546565e-01
3.96776050e-01 1.04329610e+00 7.14240074e-01 -7.00536311e-01
7.10373223e-01 1.12894336e-02 -5.06043315e-01 -6.00896001e-01
-7.92470932e-01 -4.07312900e-01 -7.80689657e-01 -5.47010839e-01
6.14211261e-01 -6.98480666e-01 -2.88338423e-01 3.20769280e-01
-1.30859077e+00 -2.43981391e-01 -4.72856104e-01 6.41727030e-01
-7.56996155e-01 1.90387174e-01 4.06365171e-02 -8.44889164e-01
-2.43066072e-01 -1.17628860e+00 1.09693539e+00 8.72922689e-02
-5.28666973e-01 -9.49166238e-01 6.11732781e-01 2.45396540e-01
4.86550182e-01 -3.70388217e-02 8.46522331e-01 -9.85767245e-01
-1.90372556e-01 -2.88421094e-01 3.14435542e-01 3.90822560e-01
4.60992783e-01 -2.40272075e-01 -1.47758698e+00 -3.21017206e-01
-1.54812336e-01 -7.95723721e-02 4.19876844e-01 -3.82375196e-02
1.22518790e+00 -7.28335679e-01 -2.51264513e-01 4.78963591e-02
1.25678122e+00 5.70949972e-01 6.21292531e-01 4.27050680e-01
2.92696804e-01 2.52352059e-01 5.00791430e-01 5.69648862e-01
1.45621300e-01 9.77275968e-01 3.10373574e-01 3.46997023e-01
-3.13985616e-01 -5.25736094e-01 4.03331548e-01 1.04035890e+00
1.47624359e-01 -6.33377254e-01 -9.46239352e-01 8.28203261e-01
-1.75121212e+00 -7.15341389e-01 5.00771284e-01 2.49340391e+00
1.24913216e+00 1.82256728e-01 1.16232157e-01 2.78243542e-01
1.00135708e+00 -1.14158310e-01 -5.54914117e-01 -4.64978278e-01
2.10282728e-01 1.85599536e-01 -1.43399566e-01 8.85627329e-01
-1.07456076e+00 7.58066952e-01 6.27740908e+00 4.06705588e-01
-1.18662560e+00 5.94505761e-03 2.70349920e-01 -4.52799276e-02
-3.16550732e-01 5.73920971e-03 -4.33967531e-01 5.70053399e-01
1.29081225e+00 -5.42339623e-01 6.15341067e-01 7.18459010e-01
2.55568773e-01 1.19888440e-01 -1.34998262e+00 1.02073205e+00
2.64891207e-01 -1.20009530e+00 6.66844904e-01 -5.35358131e-01
6.55457795e-01 -3.83007675e-01 -5.03232144e-02 3.88059109e-01
-1.48539141e-01 -6.18308187e-01 8.45341504e-01 7.31531262e-01
6.09314919e-01 -5.81110954e-01 9.14391816e-01 2.02282503e-01
-1.27321637e+00 -1.52006418e-01 2.41883352e-01 8.03508013e-02
2.44370729e-01 2.24630907e-01 -1.47647357e+00 4.77407664e-01
6.31330431e-01 3.68991882e-01 -7.79681504e-01 1.59530997e+00
-2.93021202e-01 7.47744918e-01 -2.48506397e-01 -7.51113817e-02
1.92831922e-02 3.89262170e-01 9.15323913e-01 1.35878015e+00
5.11177421e-01 -2.54748911e-01 4.27558180e-03 3.68476093e-01
-3.72746252e-02 4.76695120e-01 -3.33051682e-01 -1.12721227e-01
5.83432496e-01 9.56005573e-01 -2.65024871e-01 -5.24074674e-01
1.04946066e-02 9.52446043e-01 2.38963552e-02 3.83539617e-01
-7.22972333e-01 -3.82254511e-01 2.73241550e-01 3.81992161e-02
3.00545901e-01 1.25707269e-01 2.15899095e-01 -1.03282833e+00
-2.21437830e-02 -9.22393382e-01 6.27509594e-01 -1.36442184e+00
-9.10992026e-01 1.03775465e+00 2.75000900e-01 -1.46295714e+00
-6.51002884e-01 -1.13332823e-01 -3.63030434e-01 5.60901701e-01
-1.28151059e+00 -8.58692586e-01 -4.40133065e-01 5.48810542e-01
1.00939429e+00 -2.12938339e-01 9.50110257e-01 4.63960648e-01
-4.28485498e-02 4.29461956e-01 -3.47220153e-02 -3.76553506e-01
1.03899562e+00 -1.12144208e+00 -3.78683694e-02 6.49282873e-01
3.87641698e-01 2.58220941e-01 1.30589700e+00 -5.87606490e-01
-1.00928152e+00 -9.76068914e-01 8.12473953e-01 -1.78095385e-01
5.81932604e-01 -3.55286688e-01 -1.15214598e+00 6.67687953e-01
3.46062601e-01 -4.99648631e-01 7.02237964e-01 -1.83038026e-01
2.87272725e-02 -3.57900470e-01 -1.13019669e+00 4.48105127e-01
7.66162872e-01 -4.84733731e-01 -8.79997432e-01 4.69582170e-01
9.00969625e-01 -5.70789397e-01 -6.32273257e-01 2.78355807e-01
5.95187366e-01 -4.33618248e-01 6.02040946e-01 -5.37092507e-01
-1.67400286e-01 -5.55409968e-01 -1.50167018e-01 -1.70890105e+00
8.76280200e-03 -1.01167524e+00 -3.84306610e-02 1.37137640e+00
6.97670579e-01 -3.74739677e-01 5.70685267e-01 3.48059863e-01
-3.62648338e-01 -3.04907143e-01 -1.04723120e+00 -5.67462504e-01
-4.16776866e-01 -6.52551711e-01 9.80736673e-01 6.85630918e-01
-5.95833920e-02 3.98238659e-01 -5.30851901e-01 4.39085543e-01
1.77850887e-01 -2.48912543e-01 8.50858629e-01 -1.23838699e+00
-2.75891513e-01 1.08311906e-01 -4.97776479e-01 -7.49271691e-01
-2.94505507e-02 -6.90758705e-01 2.87599295e-01 -1.45086503e+00
-5.74273523e-03 -2.66916990e-01 -3.96480590e-01 4.58594263e-01
2.67194095e-03 5.60606038e-03 2.26681232e-01 2.46409059e-01
-5.70872605e-01 5.08341253e-01 8.98599625e-01 -3.81140769e-01
-7.92190015e-01 3.47920001e-01 -3.71022016e-01 5.16205549e-01
9.23450530e-01 -6.71490669e-01 -6.49103940e-01 -3.78878921e-01
1.95498079e-01 -1.18608430e-01 1.39038458e-01 -1.60543025e+00
3.38480204e-01 1.20192088e-01 1.78035989e-01 -3.88816684e-01
4.41931248e-01 -9.44873512e-01 5.63169539e-01 2.06421062e-01
-7.25913882e-01 1.53010832e-02 6.34647250e-01 6.27419472e-01
-2.83811808e-01 -6.42020166e-01 5.81362188e-01 -1.64707258e-01
-8.20580184e-01 5.11215860e-03 -5.60460448e-01 -3.94511111e-02
6.91879630e-01 -9.36831459e-02 -2.59963181e-02 -9.07802284e-01
-1.32030237e+00 -6.11754879e-02 1.25109896e-01 7.58316875e-01
4.60868597e-01 -1.24878967e+00 -7.13869274e-01 -3.90819609e-02
2.77002484e-01 -3.05180132e-01 1.56408548e-01 2.69020587e-01
-3.05664539e-01 3.98096085e-01 -2.67468899e-01 -5.12855649e-01
-1.28668153e+00 5.68688333e-01 4.21282381e-01 -3.05884689e-01
-2.74966955e-01 4.87788677e-01 -1.84566766e-01 -1.76865265e-01
4.91432190e-01 -2.54580110e-01 -5.49364924e-01 1.66767955e-01
5.68922579e-01 7.98182860e-02 2.74761945e-01 -5.69448411e-01
-1.36063978e-01 5.42365849e-01 -1.51279524e-01 -7.05689013e-01
1.33134592e+00 -2.58096516e-01 1.73306331e-01 9.06846583e-01
1.09722364e+00 2.50660211e-01 -1.01784134e+00 -2.49726847e-01
6.48425743e-02 -3.39175969e-01 -1.59632042e-01 -1.05367053e+00
-6.02522016e-01 6.17871940e-01 9.53700602e-01 8.24309587e-02
1.27453768e+00 -8.23603794e-02 6.08895838e-01 7.56587684e-01
3.11165005e-01 -1.26400101e+00 3.26363772e-01 3.40063900e-01
1.27699983e+00 -6.96132958e-01 1.03073893e-02 -1.27719194e-01
-7.82590568e-01 1.44907844e+00 5.20613551e-01 4.94536430e-01
4.25200731e-01 -1.02822341e-01 4.38975617e-02 2.74356067e-01
-9.25702453e-01 -6.27205595e-02 4.43587095e-01 7.83834577e-01
1.92952543e-01 -2.37044021e-01 -8.83210525e-02 4.41946566e-01
-4.93117511e-01 3.15936625e-01 7.73299038e-01 8.84883106e-01
-3.82141173e-01 -1.32100964e+00 -3.66715848e-01 2.73204774e-01
-1.51553795e-01 2.14433968e-02 -4.74624813e-01 6.98082387e-01
2.18689546e-01 1.02688766e+00 -1.03347607e-01 -3.89909059e-01
4.71464068e-01 6.21841848e-01 1.35054901e-01 -7.67776906e-01
-5.49626648e-01 -1.33749500e-01 3.56430441e-01 -7.53397495e-02
-5.27705014e-01 -8.42754900e-01 -8.87507081e-01 5.37496507e-01
-2.22055838e-01 6.66034043e-01 9.46134627e-01 9.35727775e-01
4.13892537e-01 5.46352148e-01 6.57394826e-01 -7.68266380e-01
-2.38223314e-01 -9.34693635e-01 2.41192020e-02 3.47450286e-01
5.61000526e-01 -5.93866706e-01 -5.28999388e-01 6.06572807e-01] | [15.276043891906738, 4.846409320831299] |
dacbf7a0-ba69-466f-9c6c-bd3e9f16f715 | corpus-augmentation-by-sentence-segmentation | 1905.08945 | null | https://arxiv.org/abs/1905.08945v1 | https://arxiv.org/pdf/1905.08945v1.pdf | Corpus Augmentation by Sentence Segmentation for Low-Resource Neural Machine Translation | Neural Machine Translation (NMT) has been proven to achieve impressive results. The NMT system translation results depend strongly on the size and quality of parallel corpora. Nevertheless, for many language pairs, no rich-resource parallel corpora exist. As described in this paper, we propose a corpus augmentation method by segmenting long sentences in a corpus using back-translation and generating pseudo-parallel sentence pairs. The experiment results of the Japanese-Chinese and Chinese-Japanese translation with Japanese-Chinese scientific paper excerpt corpus (ASPEC-JC) show that the method improves translation performance. | ['Jinyi Zhang', 'Tadahiro Matsumoto'] | 2019-05-22 | null | null | null | null | ['low-resource-neural-machine-translation'] | ['natural-language-processing'] | [ 3.12415868e-01 -3.23867708e-01 -2.22393587e-01 -1.78467989e-01
-1.16543627e+00 -6.95452332e-01 6.79850519e-01 -4.29631293e-01
-4.26195025e-01 1.27487445e+00 3.48754108e-01 -7.70649910e-01
6.45944238e-01 -3.86507660e-01 -6.67230487e-01 -3.11908454e-01
4.34231907e-01 7.97478497e-01 -3.00216287e-01 -4.99479920e-01
3.03847671e-01 1.53563514e-01 -5.41193604e-01 5.73176682e-01
1.35430646e+00 1.90286208e-02 6.13262057e-01 5.96980274e-01
-4.24527138e-01 4.82468866e-02 -7.03336835e-01 -8.53655279e-01
6.38471425e-01 -1.06419766e+00 -9.67330515e-01 -1.46907672e-01
1.73576787e-01 6.31510913e-02 -9.35186520e-02 1.17436695e+00
6.18791878e-01 -3.21249932e-01 4.18879479e-01 -7.82650173e-01
-1.24619997e+00 1.08308864e+00 -6.28047764e-01 4.01093990e-01
1.17310010e-01 -2.10298955e-01 8.62283111e-01 -1.18109739e+00
9.14598525e-01 1.09255648e+00 4.25474197e-01 7.30156898e-01
-9.72923517e-01 -5.19418120e-01 -3.14103842e-01 2.13628709e-01
-1.28907728e+00 -1.19879872e-01 5.44754624e-01 9.40383524e-02
1.35795045e+00 4.12055224e-01 5.60345173e-01 1.33178508e+00
7.42529452e-01 8.37539017e-01 1.24190056e+00 -8.94998074e-01
-2.27035776e-01 1.72704831e-01 -1.38608562e-02 1.65981531e-01
3.63648236e-02 -4.86585312e-02 -3.81679416e-01 -1.34787112e-01
7.47847021e-01 -4.10267591e-01 -2.19126314e-01 5.73840916e-01
-1.72580993e+00 6.54946446e-01 -1.34063572e-01 7.46599853e-01
-3.52866024e-01 -4.36367840e-01 8.68972957e-01 8.71387780e-01
6.81090355e-01 6.82769120e-01 -6.00811958e-01 -3.60326797e-01
-8.14939857e-01 1.96646675e-02 7.00633109e-01 1.38532352e+00
4.26110595e-01 -1.14667937e-02 -2.35685989e-01 1.16210437e+00
-3.07192564e-01 8.90530348e-01 9.83527541e-01 -4.23837423e-01
1.16163647e+00 2.70171225e-01 -4.56968173e-02 -6.34556651e-01
3.59668396e-02 -3.15283895e-01 -9.09262896e-01 -5.34283996e-01
-2.07581725e-02 -5.27579546e-01 -5.09580433e-01 1.50803196e+00
3.06377467e-02 -4.10366207e-01 6.42131865e-01 8.91896844e-01
4.14190769e-01 1.17109811e+00 -2.94524938e-01 -6.91597402e-01
1.01181984e+00 -1.55236840e+00 -1.14118314e+00 -1.57504857e-01
7.42329240e-01 -1.59693170e+00 1.24401295e+00 1.49803516e-02
-1.33833468e+00 -7.50227690e-01 -8.93385053e-01 -1.05966151e-01
-1.65491536e-01 4.11978781e-01 2.16013446e-01 3.61513436e-01
-8.14161718e-01 6.93305194e-01 -6.98734462e-01 -4.88436610e-01
-1.51282966e-01 3.22171479e-01 -4.72868711e-01 -1.61461711e-01
-1.30077922e+00 1.25642991e+00 5.54252863e-01 1.88342720e-01
-3.46634209e-01 -8.36334378e-02 -2.24530488e-01 -1.70621425e-01
-2.31665760e-01 -7.90393531e-01 1.32965064e+00 -1.31814170e+00
-1.92599440e+00 5.79188347e-01 -3.71540725e-01 -4.30984080e-01
4.59320128e-01 -4.08521116e-01 -6.42696977e-01 -1.05619030e-02
2.32411906e-01 5.94118595e-01 3.00112575e-01 -6.93076253e-01
-5.02878547e-01 -3.32290649e-01 -5.45903563e-01 5.60387373e-01
-4.22131181e-01 7.68060207e-01 -4.38488126e-01 -8.80766213e-01
-1.10218599e-01 -1.19492018e+00 -5.13193190e-01 -8.05914879e-01
-3.59088242e-01 -2.38258854e-01 6.85854197e-01 -1.14255834e+00
1.17005599e+00 -1.68823516e+00 3.35457742e-01 -2.45692149e-01
-5.91659546e-01 3.85571629e-01 -6.16566479e-01 8.82120967e-01
3.70046962e-03 2.61640459e-01 -3.21906656e-01 -1.85151279e-01
-2.77995378e-01 5.17857969e-01 -3.71361405e-01 6.66196719e-02
3.39069963e-01 1.32066488e+00 -7.06657469e-01 -6.31354094e-01
-3.27381074e-01 -4.32714000e-02 -7.16953799e-02 3.61801982e-01
-1.13986932e-01 6.15779579e-01 -4.54408228e-01 4.17203218e-01
5.72667181e-01 6.89757094e-02 2.32763991e-01 3.53760660e-01
-2.87958562e-01 5.96216321e-01 -1.77210823e-01 2.03654146e+00
-4.90529478e-01 7.82304168e-01 -3.68714660e-01 -8.08944464e-01
1.16671312e+00 6.69771373e-01 -1.46767572e-02 -7.54784107e-01
1.25420555e-01 7.45638192e-01 4.27206516e-01 -7.71318734e-01
7.85218716e-01 -2.41971970e-01 2.65997946e-02 5.27514338e-01
-1.25111109e-02 -2.48094305e-01 6.10295355e-01 -1.22526489e-01
7.55255342e-01 4.56670731e-01 1.49582848e-01 -3.56325179e-01
6.21904552e-01 5.97154081e-01 8.39082837e-01 1.54303148e-01
5.62078916e-02 8.87615800e-01 2.52530556e-02 -4.13637042e-01
-1.81933475e+00 -7.25944936e-01 1.54786512e-01 5.96604705e-01
-2.38550916e-01 -3.67779106e-01 -1.03920341e+00 -5.63945174e-01
-6.84406221e-01 7.37471044e-01 -1.62674546e-01 9.61221904e-02
-1.35146356e+00 -1.01936364e+00 6.64176762e-01 4.23709542e-01
5.22876322e-01 -1.17227948e+00 3.46782982e-01 4.38484967e-01
-8.75309706e-01 -1.38922095e+00 -1.13061547e+00 -1.89134538e-01
-1.20895493e+00 -4.93335217e-01 -1.10966623e+00 -1.34077561e+00
7.03353584e-01 3.02416950e-01 1.16854680e+00 -2.96153575e-01
2.67263740e-01 -4.96985555e-01 -5.16471326e-01 -4.28117275e-01
-1.13333988e+00 3.30572456e-01 2.37603694e-01 -4.04768825e-01
7.46443093e-01 -5.74615002e-01 -7.56988069e-04 2.95443475e-01
-5.15704453e-01 5.56805670e-01 1.13251519e+00 1.02323890e+00
5.00025511e-01 -6.71607077e-01 6.43894672e-01 -5.74314952e-01
1.15746653e+00 -1.31547213e-01 -4.44324702e-01 6.53873146e-01
-7.29929328e-01 5.71547300e-02 1.02356744e+00 -7.68757463e-01
-1.11904049e+00 -3.33380282e-01 -1.40318155e-01 -2.52948970e-01
1.04340427e-01 6.31852746e-01 -7.76981637e-02 2.90445089e-01
7.44367838e-01 8.13001752e-01 -1.20618850e-01 -6.38472140e-01
9.78937745e-02 1.12380207e+00 5.80477774e-01 -6.40769541e-01
7.52519011e-01 -2.43188933e-01 -2.40783632e-01 -5.40172398e-01
-2.17755422e-01 -1.71765551e-01 -9.46412802e-01 2.32721522e-01
7.79789686e-01 -7.89284408e-01 -4.05856706e-02 1.35994762e-01
-1.91386771e+00 -1.41617179e-01 6.63359910e-02 1.02254665e+00
-3.25151175e-01 5.88268995e-01 -1.13946533e+00 -2.77807206e-01
-1.11076331e+00 -1.23167419e+00 7.10276783e-01 3.02043166e-02
-2.92110652e-01 -6.92307413e-01 4.09516275e-01 3.81564945e-01
3.64077538e-01 -3.25327605e-01 9.60182726e-01 -7.74703085e-01
-3.07053268e-01 -1.67689204e-01 -9.45139080e-02 6.58389449e-01
6.51525036e-02 -9.16544050e-02 -1.32046178e-01 -1.19483076e-01
1.38996527e-01 -2.17451043e-02 1.95708543e-01 4.53248769e-02
4.29707974e-01 -4.61942762e-01 -1.39994979e-01 2.22649202e-01
1.21354616e+00 4.13429230e-01 6.52844846e-01 3.07558358e-01
5.82423508e-01 5.01740336e-01 6.16193235e-01 -3.00494492e-01
1.61551744e-01 6.33911610e-01 -4.50934529e-01 -6.64897710e-02
-2.33623281e-01 -2.51280308e-01 8.12952518e-01 2.07630467e+00
-3.35795879e-01 -2.75777012e-01 -8.67672920e-01 6.05960906e-01
-1.77402925e+00 -7.85874248e-01 -6.70478165e-01 1.87177086e+00
1.27602351e+00 -8.67985561e-02 -2.24748954e-01 -4.06196147e-01
9.72926855e-01 -3.88254970e-01 -8.29602182e-02 -1.03116167e+00
-5.75993299e-01 1.72317147e-01 4.86631334e-01 5.01856387e-01
-3.49454790e-01 1.29673851e+00 6.64074039e+00 8.67366850e-01
-1.03247440e+00 4.66132492e-01 5.01545668e-01 3.67545150e-02
-2.86665440e-01 9.73382145e-02 -7.35903859e-01 4.64899540e-01
1.37107289e+00 -6.49131954e-01 5.43183327e-01 5.70202410e-01
3.38641107e-01 4.34043288e-01 -1.11054945e+00 8.50132287e-01
9.20068994e-02 -1.33242774e+00 3.70613158e-01 5.86737730e-02
1.16072059e+00 4.42242116e-01 -3.49618822e-01 3.90643835e-01
1.91520918e-02 -5.88156700e-01 4.65967149e-01 -2.35044770e-02
6.57189250e-01 -6.93386793e-01 1.03742802e+00 6.11748278e-01
-6.98663890e-01 3.48405778e-01 -8.61061752e-01 -1.73379049e-01
2.53517628e-01 4.73459989e-01 -8.77748251e-01 9.15969074e-01
3.27609360e-01 4.96728092e-01 -1.88601136e-01 5.77120483e-01
-4.89959627e-01 9.12913978e-01 -3.10693104e-02 -3.48107666e-01
3.17509115e-01 -8.99338663e-01 5.96035600e-01 1.38690448e+00
5.63320756e-01 7.41517469e-02 3.04497406e-02 8.04726124e-01
-3.08500141e-01 8.31233561e-01 -6.20366096e-01 -4.20349568e-01
2.39546627e-01 1.01316428e+00 -3.68717521e-01 -6.05687380e-01
-5.47606051e-01 1.48259866e+00 3.55259031e-01 4.22145456e-01
-5.50872147e-01 -3.24264467e-01 3.50625962e-01 -4.56212759e-01
-1.33873969e-01 -4.84076619e-01 -6.50443912e-01 -1.36017621e+00
4.17290837e-01 -1.10322011e+00 -2.36097530e-01 -7.59789288e-01
-1.37242413e+00 1.32679284e+00 -2.48524740e-01 -1.45178497e+00
-3.67176831e-01 -3.50497037e-01 -6.74839020e-01 1.41635633e+00
-1.04111481e+00 -1.26040149e+00 4.88251179e-01 1.34056300e-01
1.11614120e+00 -6.29372537e-01 1.04209983e+00 4.48315173e-01
-7.61102855e-01 5.73497891e-01 5.85907996e-01 2.39142388e-01
9.29604292e-01 -9.39111769e-01 1.16525662e+00 1.07435715e+00
2.43258402e-01 7.96733022e-01 6.64013922e-01 -8.61784577e-01
-1.40078163e+00 -8.98127675e-01 1.83475208e+00 -3.46507907e-01
5.36752582e-01 -4.05431896e-01 -8.53001475e-01 6.06407344e-01
1.13141739e+00 -4.93810803e-01 6.64525151e-01 -7.98715055e-02
-1.53706577e-02 1.61187068e-01 -6.19927466e-01 1.04550791e+00
1.03898263e+00 -4.70376372e-01 -9.67832148e-01 6.72890425e-01
1.04239976e+00 -3.63933235e-01 -9.16652203e-01 3.90012443e-01
4.07510489e-01 -2.58506477e-01 4.82541323e-01 -8.97454500e-01
9.60900426e-01 -1.27711847e-01 -5.50781116e-02 -1.58369505e+00
-2.26050720e-01 -1.04033494e+00 2.82802612e-01 1.17050767e+00
9.05584872e-01 -4.63232458e-01 5.67177474e-01 2.56607980e-01
-5.31869173e-01 -6.02854848e-01 -8.69618416e-01 -1.15773284e+00
6.52542651e-01 1.92210808e-01 6.90042377e-01 1.06813550e+00
4.69337076e-01 1.10667312e+00 -4.99515533e-01 -1.85586736e-01
8.71109441e-02 3.89382422e-01 8.24077606e-01 -6.59942687e-01
-4.29148793e-01 -4.93119001e-01 3.60076785e-01 -1.23751104e+00
2.58784711e-01 -1.10632348e+00 -1.19769229e-02 -1.52354038e+00
5.22769690e-01 -6.80473680e-03 1.40018120e-01 1.44117653e-01
-6.26844823e-01 2.51925617e-01 1.24485202e-01 6.81560099e-01
-1.83794498e-02 6.82030022e-01 1.75186670e+00 -1.69731632e-01
-1.70757309e-01 -1.55608281e-01 -1.68996066e-01 9.76481959e-02
1.17531097e+00 -6.39134645e-01 -4.65670675e-02 -9.36936975e-01
-6.51443154e-02 2.58487165e-01 -6.47080481e-01 -5.61117351e-01
-1.07571304e-01 -3.81929934e-01 1.58822179e-01 -6.86677814e-01
-3.16977762e-02 -4.73999530e-01 7.81778544e-02 5.60066521e-01
-4.28134322e-01 1.13170898e+00 2.86283016e-01 7.54227266e-02
-4.47228312e-01 -3.20667058e-01 4.67054278e-01 -3.20753962e-01
9.03240696e-04 -1.21043278e-02 -4.78419602e-01 -1.00484267e-01
6.25091136e-01 -8.16291571e-02 -2.27361843e-01 2.28950325e-02
-6.40484989e-02 6.81856275e-02 4.21199203e-01 8.47922504e-01
3.17539990e-01 -1.51253664e+00 -1.50815558e+00 1.12029009e-01
-9.87820476e-02 -3.95154655e-01 -1.53854042e-01 8.73762846e-01
-6.92398608e-01 6.86172664e-01 -5.63878417e-01 -3.79890710e-01
-1.38884795e+00 6.95930004e-01 -5.99786937e-02 -4.64102030e-01
-4.57083493e-01 5.25127530e-01 -1.39333233e-01 -6.97663367e-01
-2.99699426e-01 -7.67921135e-02 7.60131925e-02 -7.94231832e-01
4.16820586e-01 2.30248258e-01 1.44056424e-01 -6.31669104e-01
2.44396463e-01 2.87550807e-01 -3.86170149e-01 -4.76693511e-01
1.07201600e+00 -2.02053085e-01 -7.35529065e-01 2.39722908e-01
1.32413900e+00 5.73393777e-02 -2.31971741e-01 -3.24463785e-01
2.06116974e-01 -3.47610354e-01 -4.74048942e-01 -8.72731686e-01
-5.36033452e-01 9.26236629e-01 2.41771311e-01 -3.00359040e-01
9.69083905e-01 -4.96553451e-01 1.44937551e+00 6.55712187e-01
4.13178116e-01 -1.34192300e+00 -3.43755633e-01 8.45212877e-01
1.05071020e+00 -1.13616121e+00 -3.52244973e-01 -3.69717628e-01
-5.77842176e-01 1.32515419e+00 6.37086391e-01 8.98289680e-02
-1.26085818e-01 1.50692299e-01 5.62709332e-01 4.57357079e-01
-8.09570849e-01 3.44768792e-01 1.63706645e-01 2.94096351e-01
8.34076643e-01 1.76773623e-01 -1.62603188e+00 2.24940419e-01
-4.87769008e-01 -5.07641137e-02 5.02483666e-01 7.58733094e-01
-7.11402968e-02 -2.01595140e+00 -5.95577359e-01 7.40978047e-02
-7.84188092e-01 -6.41855359e-01 -8.29424083e-01 5.08055687e-01
-2.49044105e-01 1.00219524e+00 -8.25173035e-02 -4.37739223e-01
1.56181008e-01 2.86831588e-01 6.41776204e-01 -5.64659119e-01
-9.27783489e-01 4.05056387e-01 3.16040576e-01 7.01024979e-02
-3.14555466e-01 -5.86422563e-01 -1.06925750e+00 -2.85757184e-01
-4.72217441e-01 7.92563856e-01 9.20233011e-01 9.34524596e-01
4.48022455e-01 1.68833002e-01 8.11403155e-01 -5.23974538e-01
-7.04601943e-01 -1.63961017e+00 -6.24424815e-02 2.59352773e-01
-2.84310222e-01 3.39194208e-01 2.26607904e-01 1.31948471e-01] | [11.579689979553223, 10.385865211486816] |
7a01c8d1-b07b-4822-bc39-bf1112a326c1 | targeted-vae-structured-inference-and | 2009.13472 | null | https://arxiv.org/abs/2009.13472v5 | https://arxiv.org/pdf/2009.13472v5.pdf | Targeted VAE: Variational and Targeted Learning for Causal Inference | Undertaking causal inference with observational data is incredibly useful across a wide range of tasks including the development of medical treatments, advertisements and marketing, and policy making. There are two significant challenges associated with undertaking causal inference using observational data: treatment assignment heterogeneity (\textit{i.e.}, differences between the treated and untreated groups), and an absence of counterfactual data (\textit{i.e.}, not knowing what would have happened if an individual who did get treatment, were instead to have not been treated). We address these two challenges by combining structured inference and targeted learning. In terms of structure, we factorize the joint distribution into risk, confounding, instrumental, and miscellaneous factors, and in terms of targeted learning, we apply a regularizer derived from the influence curve in order to reduce residual bias. An ablation study is undertaken, and an evaluation on benchmark datasets demonstrates that TVAE has competitive and state of the art performance. | ['Richard Bowden', 'Necati Cihan Camgoz', 'Matthew James Vowels'] | 2020-09-28 | null | null | null | null | ['miscellaneous'] | ['miscellaneous'] | [ 9.13762152e-01 3.90216917e-01 -9.40619290e-01 -3.78936082e-01
-6.35835350e-01 -4.28463340e-01 7.43019581e-01 1.94503918e-01
-4.65609252e-01 1.16882527e+00 7.77544260e-01 -9.25689697e-01
-6.11398518e-01 -7.36583889e-01 -9.94910121e-01 -6.49805129e-01
-3.47787708e-01 4.43778992e-01 -3.12178701e-01 4.74198908e-01
2.39332080e-01 -9.46145784e-03 -1.05354166e+00 1.54579893e-01
8.91578972e-01 4.18909162e-01 -3.02270442e-01 6.29022717e-02
2.10763887e-01 7.62661576e-01 -2.52699196e-01 -4.54185605e-01
-1.28660768e-01 -4.16614771e-01 -5.90403140e-01 -2.28863627e-01
4.76423711e-01 -4.47156668e-01 -3.28014083e-02 7.99753904e-01
5.09064615e-01 -2.04316482e-01 1.13024759e+00 -1.29266918e+00
-4.83464926e-01 9.59720016e-01 -8.67271423e-01 7.35446811e-02
2.56598920e-01 1.93504348e-01 1.06782949e+00 -2.63680071e-01
5.17749548e-01 1.55638599e+00 4.74558115e-01 3.43833774e-01
-1.70795488e+00 -1.00538802e+00 3.74051511e-01 -1.67118609e-01
-8.20990086e-01 -4.23963368e-01 4.79407191e-01 -7.58985877e-01
2.34568477e-01 2.56419390e-01 5.51174693e-02 1.51115572e+00
4.55885708e-01 6.34702325e-01 1.34641480e+00 -3.86637777e-01
5.59921622e-01 -9.40077975e-02 -1.23725012e-01 2.02317059e-01
3.16903383e-01 7.90419519e-01 -2.02266991e-01 -6.90576017e-01
6.21439338e-01 1.05534576e-01 -1.92790553e-01 -5.39067745e-01
-1.22050428e+00 1.23512828e+00 2.34505355e-01 -2.60731101e-01
-5.96544266e-01 1.79402232e-01 4.51375186e-01 -8.88287947e-02
3.79439384e-01 4.10347909e-01 -7.00690091e-01 2.31052592e-01
-7.59694457e-01 5.25963008e-01 6.05297506e-01 4.98424381e-01
1.67406440e-01 -2.43742794e-01 -4.33066458e-01 6.66893303e-01
2.12079734e-01 7.45431662e-01 2.35986412e-01 -9.90418255e-01
5.43674648e-01 3.87262106e-01 4.00422245e-01 -6.51475072e-01
-4.94391173e-01 -4.21254821e-02 -1.09446013e+00 9.13237967e-03
6.32451653e-01 -6.70984030e-01 -1.07111883e+00 2.08872533e+00
4.54808116e-01 5.23949802e-01 -1.91890031e-01 7.62543917e-01
6.14543974e-01 2.84762025e-01 7.32508302e-01 -6.34788156e-01
1.44884121e+00 -1.63329944e-01 -8.28498602e-01 -4.69294459e-01
7.08342135e-01 -4.37308818e-01 7.11943448e-01 2.20372289e-01
-1.01886821e+00 5.01930751e-02 -6.48303986e-01 2.90649980e-01
-4.60798778e-02 -2.17880383e-01 9.58447874e-01 6.03149593e-01
-3.48278344e-01 5.14383852e-01 -5.49019933e-01 3.57661210e-02
6.44539654e-01 5.89021623e-01 -2.85803139e-01 -7.31634200e-02
-1.46375835e+00 5.44443071e-01 1.13535367e-01 -1.55703664e-01
-9.54652488e-01 -1.42095554e+00 -7.98813581e-01 2.74582863e-01
1.01587176e+00 -1.01196992e+00 1.13334382e+00 -6.25513017e-01
-9.37900245e-01 4.64250475e-01 -3.14065129e-01 -4.52555627e-01
6.52029753e-01 4.08985168e-02 -2.53936499e-01 -4.49633867e-01
3.93207371e-01 2.42865041e-01 7.92369127e-01 -9.02356863e-01
-9.27001655e-01 -7.12149203e-01 -2.22746521e-01 6.74747378e-02
-8.61073378e-03 1.53913379e-01 1.62563510e-02 -8.19227457e-01
-1.85610518e-01 -1.00843191e+00 -4.20387506e-01 -3.35994631e-01
-5.14491439e-01 -2.80292034e-01 4.52613115e-01 -6.20125413e-01
1.33603120e+00 -1.91870487e+00 9.99738723e-02 3.30273241e-01
2.12217242e-01 -4.42982279e-02 1.52665079e-01 1.49035975e-01
-5.64022958e-01 4.27599877e-01 -3.27963859e-01 2.39894539e-01
-8.99378583e-02 -1.97308417e-02 -3.28655154e-01 4.27298635e-01
1.57503739e-01 5.93718588e-01 -9.29519355e-01 -3.54792237e-01
1.12552837e-01 1.85198113e-01 -7.25722969e-01 -2.63621882e-02
-2.45836914e-01 6.40864551e-01 -6.20792210e-01 1.59595028e-01
5.31333685e-01 -1.59043372e-01 5.76885760e-01 7.58457780e-02
-1.22299217e-01 5.49805641e-01 -1.38262665e+00 9.37731385e-01
-5.81653774e-01 8.75516608e-02 -7.55337551e-02 -1.11368287e+00
2.91583538e-01 5.37295282e-01 5.57799399e-01 -6.80187702e-01
1.32827103e-01 8.13541338e-02 3.46390724e-01 -5.72086990e-01
-1.89292863e-01 -6.02127016e-01 -2.20565736e-01 5.69181442e-01
-2.91237533e-01 3.75953972e-01 -2.01062009e-01 2.42283419e-02
1.17192090e+00 -1.52305305e-01 3.70610654e-01 -2.28904650e-01
9.23677832e-02 -2.52304375e-01 8.37359786e-01 1.07864237e+00
1.30673870e-01 3.42626661e-01 1.11680889e+00 -2.45254368e-01
-8.70080054e-01 -1.12798703e+00 -3.30232888e-01 8.96852851e-01
-3.93556029e-01 1.64053649e-01 -2.99702376e-01 -7.51353443e-01
3.59376222e-01 1.07788825e+00 -1.09121084e+00 -4.23261940e-01
-3.84031296e-01 -1.24591410e+00 2.37334386e-01 5.33839464e-01
-3.50502543e-02 -9.24142599e-01 -2.32264563e-01 2.74353325e-01
-1.04692765e-01 -6.80723906e-01 -4.64888692e-01 8.77631679e-02
-8.40929747e-01 -1.13161969e+00 -4.18054223e-01 -2.04941139e-01
3.17087889e-01 -1.27329022e-01 1.10136378e+00 -4.26548421e-01
-1.40392393e-01 -1.61733553e-01 1.13947757e-01 -8.44539106e-01
-3.65073562e-01 -2.46403128e-01 -7.83812031e-02 -4.03639637e-02
3.63252401e-01 -4.79373425e-01 -8.71149957e-01 1.52948126e-01
-8.61944437e-01 -1.67030990e-01 5.87536395e-01 8.98014665e-01
3.63414049e-01 1.21470168e-01 7.60045946e-01 -1.56640410e+00
4.15536284e-01 -8.05721760e-01 -7.06522703e-01 2.55980730e-01
-7.25515008e-01 1.52280360e-01 2.10024565e-01 -6.25199258e-01
-1.33370554e+00 -1.47903323e-01 2.30257049e-01 1.02928095e-01
-1.93206266e-01 6.79497957e-01 -3.22703212e-01 6.84261501e-01
5.23735404e-01 -5.54659128e-01 1.21880420e-01 -3.28157187e-01
2.93345660e-01 7.56901979e-01 1.31624609e-01 -5.09549618e-01
2.84268171e-01 6.14084005e-01 2.00945765e-01 -3.35051358e-01
-9.47128594e-01 -2.42514879e-01 -8.18815921e-03 4.26335692e-01
7.37668395e-01 -9.69583452e-01 -1.14385176e+00 1.56578831e-02
-7.59355187e-01 -4.40964282e-01 -7.69689530e-02 8.72628570e-01
-4.32770967e-01 -1.02617033e-01 -2.94490010e-02 -8.42276633e-01
-1.78748682e-01 -1.08835804e+00 8.54605317e-01 6.60656765e-02
-5.48296273e-01 -1.27103579e+00 1.64066013e-02 4.68822598e-01
-2.78640464e-02 4.78386879e-01 1.38954592e+00 -5.84869564e-01
-2.22055048e-01 -1.83415547e-01 -2.21711367e-01 -1.13449901e-01
4.25142258e-01 -1.96101040e-01 -6.47164583e-01 -1.32605702e-01
-1.87403291e-01 -2.80242115e-02 7.37475216e-01 1.22212279e+00
1.37490642e+00 -6.07333362e-01 -7.11586773e-01 1.84989408e-01
1.06916380e+00 3.95381212e-01 6.29233420e-01 -1.22693002e-01
6.98487222e-01 8.55232835e-01 4.08955842e-01 4.90950853e-01
5.25523305e-01 8.46912742e-01 2.95238495e-01 -2.60689974e-01
1.20834857e-01 -4.17049646e-01 9.92846396e-03 -5.49390316e-01
6.73671216e-02 -1.21300258e-01 -7.16383696e-01 6.79165959e-01
-1.85133243e+00 -1.12621915e+00 -4.28391337e-01 2.66503620e+00
9.87471759e-01 1.72702715e-01 2.91085273e-01 -2.15445578e-01
6.86482370e-01 -2.03521803e-01 -7.07767785e-01 -4.28362191e-01
6.74739927e-02 2.27184027e-01 8.70429218e-01 5.13122857e-01
-1.07410789e+00 5.20165265e-01 6.20621395e+00 7.12375641e-01
-9.83482063e-01 6.36911690e-02 8.56493533e-01 -4.18530740e-02
-5.22348225e-01 3.30430567e-01 -7.21596479e-01 6.28052890e-01
9.49624777e-01 -1.79645613e-01 8.76723081e-02 1.83454365e-01
8.39213550e-01 -2.41584688e-01 -1.28682113e+00 1.67514682e-01
-3.84641737e-01 -1.23542523e+00 -8.94959942e-02 3.83155435e-01
8.61629486e-01 -3.57778698e-01 1.39921263e-01 2.08064884e-01
9.22082067e-01 -1.08590686e+00 4.38297242e-01 3.77356082e-01
8.31112683e-01 -5.66350698e-01 9.11091924e-01 2.44138598e-01
-2.71116883e-01 -3.77551585e-01 7.28693902e-02 -1.79392457e-01
1.18549451e-01 9.22145009e-01 -9.44223464e-01 6.21752858e-01
5.60063779e-01 4.68155563e-01 2.74904608e-03 8.97300661e-01
-3.78295213e-01 1.01124346e+00 -7.61545002e-02 2.90968627e-01
-6.94313422e-02 -9.41245481e-02 3.10591727e-01 7.21315563e-01
1.18289813e-01 9.45567116e-02 -2.95674782e-02 7.52615154e-01
-1.85952723e-01 -8.72981250e-02 -6.62035823e-01 1.44080082e-02
3.55326056e-01 7.11517513e-01 -3.35965186e-01 -2.05465198e-01
-4.71577525e-01 2.14533538e-01 -8.57073218e-02 5.09473622e-01
-6.81232929e-01 2.20884755e-01 4.75753009e-01 3.61384988e-01
1.01437815e-01 6.34669244e-01 -6.11754000e-01 -7.83637285e-01
-2.40408167e-01 -1.00526822e+00 9.38012719e-01 -3.76601189e-01
-1.33819568e+00 -5.10912836e-01 4.98523563e-01 -6.15259886e-01
-3.85811836e-01 -2.97562569e-01 -4.71476704e-01 1.13478649e+00
-1.10278130e+00 -9.28207874e-01 4.04571444e-01 1.69134676e-01
1.58131808e-01 2.66246974e-01 5.87261796e-01 2.64445126e-01
-5.10811746e-01 4.49949116e-01 1.07977390e-01 -1.15406267e-01
9.38007534e-01 -1.22964704e+00 7.81542342e-03 2.87132263e-01
-5.15599072e-01 6.39330149e-01 7.83399820e-01 -1.03402925e+00
-9.77033377e-01 -1.09169555e+00 1.11989009e+00 -5.33535480e-01
8.12891066e-01 -4.63939071e-01 -7.36517191e-01 8.48316550e-01
-2.51167834e-01 -3.73581618e-01 6.40989542e-01 8.14240575e-01
-3.34915757e-01 -4.84550707e-02 -1.07955468e+00 9.85503078e-01
9.22868490e-01 -6.64045736e-02 -4.84197766e-01 5.33043265e-01
7.12499678e-01 -1.13144435e-01 -7.52416253e-01 7.96622753e-01
7.26317763e-01 -6.29763126e-01 1.04411733e+00 -1.16361797e+00
7.40912020e-01 4.45531644e-02 2.42645532e-01 -1.35724592e+00
-2.80660927e-01 -4.85190958e-01 1.42358497e-01 1.15558624e+00
6.33887768e-01 -5.98655105e-01 6.61010981e-01 9.19728220e-01
3.54674786e-01 -5.44104755e-01 -9.42391038e-01 -4.62513059e-01
4.79043543e-01 -1.57042161e-01 6.72308505e-01 1.08043444e+00
-1.42875627e-01 6.28991485e-01 -6.35708928e-01 2.14308545e-01
8.05875063e-01 2.18496978e-01 5.53985596e-01 -1.30366755e+00
-2.84255743e-01 -2.25050166e-01 1.45476356e-01 -3.80970776e-01
1.84162870e-01 -4.44178611e-01 -7.04829693e-02 -1.47014248e+00
6.50120854e-01 -6.07237339e-01 -2.23269582e-01 5.56048036e-01
-7.80986965e-01 -4.30220574e-01 -1.99794680e-01 -1.60433590e-01
4.19946760e-02 3.43768209e-01 1.20986831e+00 -1.41508564e-01
-3.40221405e-01 4.69226420e-01 -1.08040130e+00 7.51233995e-01
6.17560327e-01 -8.98706079e-01 -5.56132376e-01 -4.05317508e-02
2.39493698e-01 5.26217818e-01 6.99568331e-01 -3.37121263e-02
-9.39204097e-02 -6.04998887e-01 2.40632400e-01 -2.87434440e-02
-2.65561283e-01 -8.36012125e-01 4.58612561e-01 7.02943027e-01
-7.24876404e-01 -2.59107769e-01 7.72266090e-02 7.23989248e-01
1.88430235e-01 -8.97466689e-02 4.21234667e-01 4.37777713e-02
-4.69933636e-02 4.32917625e-02 -1.56286791e-01 3.17183107e-01
7.15379655e-01 2.48869881e-01 -4.36074108e-01 -4.08426672e-01
-6.87802255e-01 6.08165205e-01 -8.55877101e-02 4.34752613e-01
1.90153241e-01 -1.10141659e+00 -9.57826793e-01 -8.11684206e-02
-2.11318489e-03 -1.58820555e-01 5.56309223e-01 1.01496446e+00
4.13766563e-01 3.47095072e-01 3.14873219e-01 -2.40257606e-01
-1.15034962e+00 7.25126624e-01 -2.48657465e-02 -6.10939741e-01
-2.14498460e-01 3.51047426e-01 7.96240389e-01 -5.16552150e-01
2.19918579e-01 -1.97804421e-02 -1.68879926e-01 2.60429591e-01
3.85848731e-01 4.92202371e-01 -1.27224736e-02 -1.05657196e-02
-3.24603945e-01 7.82364309e-02 -2.66375631e-01 -6.06545843e-02
1.36614048e+00 4.10659425e-02 -1.49271235e-01 4.25038457e-01
9.05760527e-01 -9.31911021e-02 -1.30491304e+00 -1.74518645e-01
7.95642659e-02 -4.15265352e-01 2.00109094e-01 -1.20004475e+00
-7.38428295e-01 5.68537831e-01 5.57013452e-01 1.75878510e-01
9.10084069e-01 5.48101030e-03 6.92939386e-02 -3.06358516e-01
6.65800124e-02 -8.46844018e-01 -4.22459781e-01 -8.88465494e-02
7.53601551e-01 -1.49911976e+00 4.10098255e-01 -5.12066662e-01
-4.58666831e-01 2.44234607e-01 1.15883239e-01 1.35925815e-01
6.98912144e-01 1.07419722e-01 -1.35211274e-01 -2.06266999e-01
-8.21300328e-01 7.04284012e-02 3.87137771e-01 3.82903963e-01
6.39262855e-01 3.73864323e-01 -7.84012139e-01 4.96230930e-01
4.21214961e-02 3.40767503e-01 3.61427993e-01 6.33465767e-01
3.20635289e-01 -1.08418727e+00 -4.86733556e-01 9.90429819e-01
-8.43000352e-01 -1.12067699e-01 -1.38902485e-01 9.64084148e-01
1.84286118e-01 1.06454289e+00 9.50982273e-02 2.21814200e-01
6.29960716e-01 -1.07383184e-01 2.36651987e-01 -6.14514887e-01
-2.62829334e-01 2.97514409e-01 2.88765430e-01 -3.94881934e-01
-5.45105696e-01 -1.07317936e+00 -8.23670983e-01 -1.93919942e-01
-3.54771495e-01 1.46796852e-01 4.30816740e-01 1.18381357e+00
2.16957361e-01 7.43404627e-01 6.95678771e-01 -2.18159452e-01
-9.15619075e-01 -7.69152522e-01 -3.18287462e-01 3.73970956e-01
4.67109531e-01 -8.27409685e-01 -5.64568758e-01 -3.94468084e-02] | [8.04552173614502, 5.381961345672607] |
1e62bfa9-1027-4042-8299-4315ec3a882c | subgraph-networks-based-contrastive-learning | 2306.03506 | null | https://arxiv.org/abs/2306.03506v1 | https://arxiv.org/pdf/2306.03506v1.pdf | Subgraph Networks Based Contrastive Learning | Graph contrastive learning (GCL), as a self-supervised learning method, can solve the problem of annotated data scarcity. It mines explicit features in unannotated graphs to generate favorable graph representations for downstream tasks. Most existing GCL methods focus on the design of graph augmentation strategies and mutual information estimation operations. Graph augmentation produces augmented views by graph perturbations. These views preserve a locally similar structure and exploit explicit features. However, these methods have not considered the interaction existing in subgraphs. To explore the impact of substructure interactions on graph representations, we propose a novel framework called subgraph network-based contrastive learning (SGNCL). SGNCL applies a subgraph network generation strategy to produce augmented views. This strategy converts the original graph into an Edge-to-Node mapping network with both topological and attribute features. The single-shot augmented view is a first-order subgraph network that mines the interaction between nodes, node-edge, and edges. In addition, we also investigate the impact of the second-order subgraph augmentation on mining graph structure interactions, and further, propose a contrastive objective that fuses the first-order and second-order subgraph information. We compare SGNCL with classical and state-of-the-art graph contrastive learning methods on multiple benchmark datasets of different domains. Extensive experiments show that SGNCL achieves competitive or better performance (top three) on all datasets in unsupervised learning settings. Furthermore, SGNCL achieves the best average gain of 6.9\% in transfer learning compared to the best method. Finally, experiments also demonstrate that mining substructure interactions have positive implications for graph contrastive learning. | ['Xiaoniu Yang', 'Qi Xuan', 'Shanqing Yu', 'Zeyu Wang', 'Jiafei Shao', 'Jinhuan Wang'] | 2023-06-06 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 4.66299087e-01 6.42309964e-01 -5.83754480e-01 -1.27128139e-01
-4.32406336e-01 -4.90601957e-01 7.04062283e-01 2.17292562e-01
6.79293796e-02 6.78616703e-01 1.49461418e-01 -2.69110143e-01
-3.07562530e-01 -1.05191374e+00 -7.97538280e-01 -7.18290985e-01
-6.05825484e-01 3.26328665e-01 1.35529712e-01 -2.49962360e-01
1.29807284e-02 4.64073181e-01 -1.16971231e+00 2.51745969e-01
9.28326368e-01 4.37733650e-01 1.25662179e-03 2.30914176e-01
-2.19700694e-01 7.35855162e-01 -2.20180258e-01 -3.05469483e-01
4.37460482e-01 -6.36073351e-01 -8.04074049e-01 3.27513307e-01
4.18773711e-01 3.71157050e-01 -5.63549519e-01 1.13342214e+00
3.28087062e-01 9.22819749e-02 7.36883581e-01 -1.82780051e+00
-5.95662773e-01 9.73004460e-01 -9.85371411e-01 4.37814564e-01
3.68576705e-01 1.55692160e-01 1.45829630e+00 -9.09535050e-01
8.43866825e-01 1.21034098e+00 5.88733971e-01 2.88685501e-01
-1.46047831e+00 -7.27144599e-01 6.34976327e-01 8.35596472e-02
-1.14490259e+00 4.62654568e-02 1.21677506e+00 -3.16199124e-01
1.01826167e+00 4.06846464e-01 6.16991103e-01 1.06867576e+00
1.04278643e-02 6.55207455e-01 1.16030824e+00 -5.38968742e-01
2.86639463e-02 -6.67463019e-02 2.47650102e-01 1.11504269e+00
5.20331681e-01 1.05415300e-01 -6.55178905e-01 -1.59208521e-01
5.02819598e-01 2.86871544e-03 -3.97100747e-01 -7.23731279e-01
-1.16502357e+00 8.07611883e-01 8.37523878e-01 3.04404616e-01
-6.02189787e-02 -1.10272624e-01 2.96059996e-01 5.53335726e-01
7.15350747e-01 8.33436608e-01 -5.95839620e-01 5.16016603e-01
-2.69759357e-01 -1.45306155e-01 7.04034925e-01 1.06599283e+00
1.16601181e+00 -3.13439332e-02 -1.56518593e-01 5.85515857e-01
2.05308840e-01 -5.98342158e-02 3.68044466e-01 -2.71276653e-01
7.10266113e-01 1.32288980e+00 -7.87395775e-01 -1.27360880e+00
-5.01275182e-01 -7.04113066e-01 -1.10494065e+00 -1.99665159e-01
6.65617138e-02 -5.06305136e-02 -1.03559792e+00 1.89414954e+00
4.46400940e-01 5.76276302e-01 -4.44595516e-03 4.55221325e-01
1.26874936e+00 2.55709618e-01 5.83062284e-02 -2.64760345e-01
1.01004207e+00 -1.14524031e+00 -5.41759193e-01 -3.76060605e-01
1.13190114e+00 -2.37522587e-01 1.18227482e+00 -7.37232864e-02
-6.31790340e-01 -3.55288267e-01 -1.03889525e+00 4.21366811e-01
-5.75870693e-01 -2.40957901e-01 8.47072005e-01 3.68718356e-01
-9.97115374e-01 9.29784298e-01 -5.04703522e-01 -2.77252108e-01
4.65167433e-01 2.68695682e-01 -7.73308158e-01 -1.41987577e-01
-1.21081543e+00 4.77925718e-01 6.36317432e-01 -2.63249576e-01
-5.55937231e-01 -9.61975455e-01 -1.32195842e+00 3.00270140e-01
9.42634940e-01 -5.55916309e-01 4.85630631e-01 -8.96435678e-01
-8.09121370e-01 9.42695618e-01 6.80976361e-02 -2.97508538e-01
2.70869970e-01 2.46010333e-01 -5.49861431e-01 7.01559559e-02
2.67585188e-01 4.05056059e-01 6.45769775e-01 -1.36500883e+00
-3.31671715e-01 -5.90075314e-01 3.22877765e-02 4.42045867e-01
-5.11839271e-01 -6.12028718e-01 -4.79257286e-01 -9.68238115e-01
3.88794899e-01 -9.90675509e-01 -3.85938257e-01 -3.48453879e-01
-8.57527494e-01 -2.35843703e-01 9.94395971e-01 -2.53199339e-01
1.25785398e+00 -1.91333890e+00 2.88828433e-01 5.57915747e-01
8.53651047e-01 1.66552097e-01 -5.07657349e-01 7.57203698e-01
-6.67423129e-01 4.11484182e-01 -3.05595577e-01 -3.39054018e-02
-3.52502972e-01 1.03877239e-01 3.19367163e-02 3.83306533e-01
3.17937583e-01 1.20562482e+00 -1.20928407e+00 -4.79973853e-01
2.78285351e-02 9.93437245e-02 -5.33499539e-01 2.55614460e-01
-1.07328072e-01 4.01140064e-01 -4.61087286e-01 6.65615380e-01
5.46727240e-01 -7.69580483e-01 8.22258711e-01 -3.95538419e-01
4.40717399e-01 8.88840333e-02 -1.02976239e+00 1.33516574e+00
-6.78059086e-02 2.91428119e-01 -4.92207915e-01 -1.32893169e+00
1.06627071e+00 -1.12686217e-01 6.07804835e-01 -5.67800045e-01
-8.07158872e-02 -7.00623393e-02 1.17297173e-01 -2.91159898e-01
-6.66318014e-02 1.66862607e-01 -6.57881051e-02 4.77390140e-01
3.18838865e-01 2.56247312e-01 3.51918578e-01 8.33509862e-01
1.53210652e+00 -1.38192503e-02 6.95051014e-01 -4.26124007e-01
4.86642778e-01 -1.68870479e-01 6.07020736e-01 4.81733680e-01
7.40278065e-02 2.84698248e-01 1.00237012e+00 -3.43984485e-01
-5.70363820e-01 -9.95756924e-01 5.23148954e-01 9.99361813e-01
3.32216501e-01 -8.35107386e-01 -4.92380708e-01 -1.35518742e+00
1.65227249e-01 4.38944161e-01 -8.64806235e-01 -8.06287944e-01
-4.07489508e-01 -8.90200317e-01 1.71728104e-01 4.96695191e-01
5.11346281e-01 -9.76476550e-01 2.86677301e-01 -1.92625239e-01
-9.11748186e-02 -1.16086102e+00 -6.68950200e-01 2.59239703e-01
-9.95381355e-01 -1.63825750e+00 -1.42324373e-01 -1.09506166e+00
1.08394313e+00 5.75457036e-01 1.30943775e+00 2.79438555e-01
-2.39231944e-01 3.12942833e-01 -5.71618736e-01 -3.62589434e-02
-3.12853485e-01 3.33272725e-01 -7.69696608e-02 6.62788004e-02
1.89052254e-01 -1.01157153e+00 -1.77048713e-01 2.06252366e-01
-6.92575276e-01 2.31813565e-01 7.51432240e-01 1.01182735e+00
7.57882178e-01 2.18799151e-02 8.04447412e-01 -1.73666751e+00
6.84862435e-01 -6.21224642e-01 -3.85870039e-01 4.31504399e-01
-1.14538395e+00 3.78005117e-01 7.19735026e-01 -2.86431402e-01
-8.20125341e-01 8.78846347e-02 4.97973531e-01 -5.53812921e-01
1.81743562e-01 9.82012808e-01 -5.43188035e-01 -1.77144960e-01
7.42190063e-01 2.32091427e-01 1.38390973e-01 -3.65985453e-01
5.45814157e-01 1.64394509e-02 3.03513259e-01 -2.97375679e-01
1.00429535e+00 3.57700139e-01 4.07698572e-01 -6.75547659e-01
-9.37548995e-01 -6.29759550e-01 -8.60782504e-01 -2.11369872e-01
2.76149988e-01 -8.17233562e-01 -5.29439330e-01 2.24482208e-01
-5.01896203e-01 -2.87188500e-01 -2.54589945e-01 1.43489420e-01
-4.22783941e-01 7.21988976e-01 -3.41401219e-01 -3.54750752e-01
-3.44759196e-01 -7.14534700e-01 8.51731181e-01 -6.80035055e-02
1.02799553e-02 -1.10815811e+00 1.56272471e-01 1.99852511e-01
-2.08148822e-01 7.26728797e-01 1.20041955e+00 -1.18302262e+00
-6.23359025e-01 -3.75200585e-02 -2.78295338e-01 1.12883293e-03
6.01575255e-01 -3.29750925e-01 -7.33209550e-01 -6.27666593e-01
-6.44182861e-01 -2.20740050e-01 1.10238671e+00 8.90942439e-02
1.21090400e+00 -4.71705705e-01 -8.29388440e-01 6.76446021e-01
1.35118544e+00 -9.34030786e-02 4.35450345e-01 1.31876528e-01
1.27023411e+00 6.71074212e-01 6.39082134e-01 2.47381046e-01
2.87860543e-01 4.12149042e-01 5.53445816e-01 -4.63472396e-01
-3.19568783e-01 -6.45691037e-01 9.77214351e-02 7.95239568e-01
-7.04315081e-02 -3.74218404e-01 -8.70049179e-01 3.93051416e-01
-1.95385599e+00 -7.61310339e-01 -2.94294745e-01 2.08350444e+00
6.02890670e-01 3.48329186e-01 1.23049475e-01 1.14817135e-01
8.80380929e-01 3.47348601e-01 -6.84280097e-01 6.28366843e-02
-1.84150353e-01 2.07260862e-01 4.62844402e-01 2.80030221e-01
-1.16949701e+00 1.12218463e+00 5.42622995e+00 6.86911345e-01
-7.72798657e-01 -1.90358862e-01 7.99066722e-01 3.18279594e-01
-5.94186187e-01 1.01592258e-01 -4.05286819e-01 2.67679960e-01
4.62244332e-01 -4.56290185e-01 3.90138388e-01 8.32934856e-01
-1.86105773e-01 3.94873381e-01 -1.21754229e+00 7.69628704e-01
9.33061168e-02 -1.42579436e+00 3.42423141e-01 3.53218347e-01
9.61397171e-01 2.77620051e-02 -1.63433969e-01 4.97424304e-01
4.85152364e-01 -9.77042854e-01 -3.76209505e-02 3.51167172e-01
9.37232852e-01 -8.60181689e-01 6.22402072e-01 3.34167272e-01
-1.69303477e+00 2.51037069e-02 6.18363917e-02 2.97039840e-02
-2.58368701e-01 6.48914635e-01 -1.05209458e+00 1.04490566e+00
6.40068173e-01 1.15380418e+00 -9.11207616e-01 7.08349466e-01
-3.60114783e-01 5.55191517e-01 1.33222356e-01 8.67531002e-02
2.41210312e-01 -4.17420000e-01 7.09311724e-01 9.85499024e-01
-1.16905294e-01 -2.03687456e-02 5.73617518e-01 7.88987100e-01
-4.98140693e-01 1.92741409e-01 -1.20273745e+00 -2.72448033e-01
5.75121939e-01 1.44806135e+00 -9.08635318e-01 -2.51034737e-01
-4.74240303e-01 8.22504878e-01 9.02860522e-01 3.23222518e-01
-5.97262442e-01 -3.27642083e-01 4.54799235e-01 2.10852146e-01
1.40869603e-01 1.47425964e-01 6.50700480e-02 -1.19860661e+00
5.80228604e-02 -1.07591665e+00 9.60575879e-01 -3.07733297e-01
-1.59338379e+00 7.06161976e-01 1.42309442e-01 -1.25462437e+00
-1.29689664e-01 -3.99004281e-01 -7.76352763e-01 4.71252620e-01
-1.46748996e+00 -1.40567219e+00 -6.22267842e-01 5.14569640e-01
2.93670297e-01 -4.54151332e-01 6.88778222e-01 1.17884586e-02
-6.87935174e-01 8.15129280e-01 -2.67101347e-01 2.18419507e-01
4.77802217e-01 -1.49077630e+00 8.12488258e-01 9.11726773e-01
4.06083256e-01 4.16914344e-01 3.10075462e-01 -1.06770051e+00
-1.31553924e+00 -1.65245152e+00 6.32752776e-01 -1.95350602e-01
6.24200761e-01 -4.89881456e-01 -1.08629537e+00 1.02434361e+00
-3.40182967e-02 4.78490889e-01 6.54308081e-01 3.65194738e-01
-6.20916784e-01 -7.83352938e-04 -1.09573567e+00 8.00833285e-01
1.77996242e+00 -3.64231855e-01 -2.98403263e-01 4.58315462e-01
9.90413964e-01 -1.60024047e-01 -7.90186703e-01 8.76164436e-01
1.33292973e-01 -7.94587493e-01 8.75243485e-01 -1.15504181e+00
3.35354865e-01 -4.72372957e-02 1.58126161e-01 -1.52537596e+00
-5.46007812e-01 -7.97313392e-01 -3.96810889e-01 1.23284471e+00
6.46665990e-01 -9.38513339e-01 1.11698127e+00 6.33344054e-02
-1.81668147e-01 -1.05068421e+00 -4.55411345e-01 -1.05410039e+00
-3.14933449e-01 1.34880289e-01 5.53597152e-01 1.67001438e+00
2.63043553e-01 8.23058188e-01 -2.65030861e-01 1.64598897e-01
7.84541845e-01 3.36019486e-01 9.29735780e-01 -1.49495173e+00
-2.89101213e-01 -4.27332401e-01 -5.52054048e-01 -4.82601315e-01
5.45848072e-01 -1.58459640e+00 -4.39757615e-01 -1.55173063e+00
5.08526087e-01 -2.72703230e-01 -4.22312737e-01 6.50173604e-01
-6.01026118e-01 -5.39807938e-02 -2.51272619e-02 -4.49378369e-03
-4.77505565e-01 6.90671325e-01 1.40000808e+00 -3.31207156e-01
-3.75222981e-01 -1.08158350e-01 -1.00892437e+00 6.18232846e-01
8.69961679e-01 -2.77871132e-01 -8.71488273e-01 2.36965016e-01
5.54502681e-02 -2.39945352e-01 1.99748814e-01 -6.18844151e-01
4.12693843e-02 -1.34582564e-01 7.67992064e-02 -4.21284676e-01
-7.64821172e-02 -5.84296703e-01 1.32134005e-01 5.65283716e-01
-3.81925076e-01 1.25202119e-01 2.83079240e-02 1.13474500e+00
-1.28698573e-01 2.09753796e-01 6.42471194e-01 -3.59072447e-01
-8.83838773e-01 6.55324161e-01 1.20353766e-01 4.15331036e-01
1.14111853e+00 -2.46068314e-01 -4.42150444e-01 -2.63761848e-01
-8.78420353e-01 3.75269741e-01 1.95718840e-01 5.39063871e-01
7.28246570e-01 -1.50220060e+00 -6.71997845e-01 3.21363598e-01
5.64549387e-01 -5.03576733e-02 4.21245471e-02 6.70904398e-01
-1.03334956e-01 5.43139167e-02 -1.38239637e-01 -3.74242902e-01
-1.53019178e+00 8.31233382e-01 2.09646001e-01 -8.12704802e-01
-7.22050667e-01 9.26164091e-01 5.68383574e-01 -7.55350471e-01
-2.00061244e-03 7.19245803e-03 -4.14465755e-01 -3.25151421e-02
4.76909764e-02 4.00658906e-01 9.15219039e-02 -4.93448049e-01
-4.04347390e-01 2.72381246e-01 -3.55024576e-01 5.66523850e-01
1.39788723e+00 2.05620736e-01 -1.76704720e-01 2.38455549e-01
1.33736503e+00 -1.06271602e-01 -9.07752514e-01 -5.85915804e-01
3.67447406e-01 -5.32940328e-01 -5.58133945e-02 -5.92697442e-01
-1.43967390e+00 3.28757912e-01 1.45314440e-01 2.36380056e-01
1.09603548e+00 3.30537438e-01 3.43031496e-01 4.52993751e-01
2.54969299e-01 -6.05321884e-01 6.12513721e-01 1.80502161e-01
9.35342431e-01 -1.43684804e+00 2.62220323e-01 -1.36427462e+00
-5.68816423e-01 7.38578081e-01 1.18035209e+00 -1.31279424e-01
7.54003108e-01 7.13064149e-02 -3.75910997e-01 -7.86184132e-01
-8.60299647e-01 -4.60221618e-01 6.65340304e-01 7.37687051e-01
3.21395814e-01 1.30901366e-01 -2.00450808e-01 4.06679094e-01
4.41524340e-03 -6.11273229e-01 3.24028760e-01 8.14253271e-01
-9.04458612e-02 -1.12140834e+00 3.46027732e-01 9.03315187e-01
-3.89023162e-02 -2.79482454e-01 -1.01565254e+00 1.13423860e+00
-2.17182323e-01 7.11289167e-01 -1.36274159e-01 -7.59896636e-01
3.82409662e-01 -3.08425039e-01 3.44946861e-01 -1.02800405e+00
-3.91182512e-01 -5.75844795e-02 3.01772147e-01 -5.93392611e-01
-3.28442186e-01 -3.68465096e-01 -1.23154676e+00 -6.09695949e-02
-6.09778345e-01 9.74002406e-02 -1.32452503e-01 8.56625676e-01
7.01846182e-01 8.12865853e-01 9.38792467e-01 -4.87061560e-01
-3.29545319e-01 -9.33369637e-01 -8.44589591e-01 6.70087636e-01
5.60862087e-02 -8.23385239e-01 -5.35856962e-01 -1.85631931e-01] | [7.302801132202148, 6.234770774841309] |
d2d4fcf5-a764-4fb1-a579-7ae65821b0d5 | selective-feature-compression-for-efficient | 2104.00179 | null | https://arxiv.org/abs/2104.00179v2 | https://arxiv.org/pdf/2104.00179v2.pdf | Selective Feature Compression for Efficient Activity Recognition Inference | Most action recognition solutions rely on dense sampling to precisely cover the informative temporal clip. Extensively searching temporal region is expensive for a real-world application. In this work, we focus on improving the inference efficiency of current action recognition backbones on trimmed videos, and illustrate that one action model can also cover then informative region by dropping non-informative features. We present Selective Feature Compression (SFC), an action recognition inference strategy that greatly increase model inference efficiency without any accuracy compromise. Differently from previous works that compress kernel sizes and decrease the channel dimension, we propose to compress feature flow at spatio-temporal dimension without changing any backbone parameters. Our experiments on Kinetics-400, UCF101 and ActivityNet show that SFC is able to reduce inference speed by 6-7x and memory usage by 5-6x compared with the commonly used 30 crops dense sampling procedure, while also slightly improving Top1 Accuracy. We thoroughly quantitatively and qualitatively evaluate SFC and all its components and show how does SFC learn to attend to important video regions and to drop temporal features that are uninformative for the task of action recognition. | ['Joseph Tighe', 'Davide Modolo', 'Hao Chen', 'Xinyu Li', 'Chunhui Liu'] | 2021-04-01 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Liu_Selective_Feature_Compression_for_Efficient_Activity_Recognition_Inference_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Liu_Selective_Feature_Compression_for_Efficient_Activity_Recognition_Inference_ICCV_2021_paper.pdf | iccv-2021-1 | ['feature-compression'] | ['computer-vision'] | [ 7.32548952e-01 -1.46808296e-01 -5.62740743e-01 -2.74923027e-01
-8.28179598e-01 -3.02672654e-01 6.00506067e-01 -2.60373324e-01
-6.66131914e-01 7.62313902e-01 4.51306045e-01 -4.79666926e-02
-2.16564074e-01 -3.97553086e-01 -8.28530073e-01 -6.08558297e-01
-2.40253046e-01 3.89651805e-01 6.18467689e-01 3.66018504e-01
3.93349171e-01 5.84085405e-01 -1.82810342e+00 7.46881127e-01
3.49206090e-01 1.08447468e+00 -5.01997732e-02 8.55480969e-01
1.48462757e-01 1.13172781e+00 -6.42639458e-01 -2.00344518e-01
2.17634037e-01 -5.07150829e-01 -1.16656113e+00 2.84680009e-01
6.98398530e-01 -7.44348347e-01 -5.22853792e-01 5.96980453e-01
3.17007005e-01 2.35106066e-01 5.22715628e-01 -1.24611914e+00
1.04044199e-01 5.69411695e-01 -4.88175184e-01 6.86454177e-01
4.06155676e-01 2.81462848e-01 8.50294530e-01 -5.52905858e-01
7.40457058e-01 1.31971979e+00 5.30347586e-01 5.93209982e-01
-1.07420695e+00 -4.76663560e-01 4.06156838e-01 8.11222911e-01
-1.25796092e+00 -6.87142909e-01 3.29577565e-01 -1.46272287e-01
1.34064639e+00 4.35282558e-01 8.05997610e-01 1.27335930e+00
1.21941231e-02 1.20803106e+00 7.26651371e-01 -2.01020822e-01
4.08095777e-01 -6.42765462e-01 3.22464257e-02 6.55213296e-01
-1.98896844e-02 -1.69413909e-03 -8.69440079e-01 -1.12986036e-01
7.14703441e-01 1.58459634e-01 -3.60081643e-01 -2.96840847e-01
-1.24930108e+00 6.59515440e-01 1.44137172e-02 8.56178775e-02
-4.17993158e-01 6.74167871e-01 6.87509358e-01 1.70400321e-01
1.66697234e-01 2.39335284e-01 -4.87261772e-01 -1.00056326e+00
-1.12820601e+00 3.21249634e-01 7.30203867e-01 9.73542392e-01
4.98781443e-01 2.01912392e-02 -5.27888834e-01 6.06435180e-01
-1.50631115e-01 4.29339051e-01 4.23146546e-01 -1.68676841e+00
3.70963991e-01 4.28706855e-01 7.29134353e-03 -7.17054367e-01
-9.71613750e-02 -1.14801545e-02 -6.66399777e-01 -2.00183526e-01
5.88543653e-01 1.27892435e-01 -8.00940752e-01 1.39110625e+00
2.89942026e-01 4.88903791e-01 -3.27179372e-01 7.97486305e-01
2.53505170e-01 4.47130948e-01 7.57182911e-02 -1.60041541e-01
1.31963897e+00 -1.00068021e+00 -6.18089139e-01 -2.04439923e-01
8.81706774e-01 -4.61084157e-01 1.01182318e+00 5.30217648e-01
-1.02919745e+00 -3.93770754e-01 -8.45337629e-01 -8.82153288e-02
-1.41123533e-01 2.12101892e-01 8.71605635e-01 3.36872578e-01
-7.10205495e-01 8.60171437e-01 -1.32206821e+00 -3.87610406e-01
9.12817478e-01 3.17838341e-01 -5.49266517e-01 -3.65933001e-01
-9.76403296e-01 7.63484180e-01 3.92991900e-01 -1.86419159e-01
-1.00933075e+00 -8.53678465e-01 -7.88311720e-01 1.38247192e-01
7.51834989e-01 -3.52305174e-01 1.29548621e+00 -8.24842393e-01
-1.44544911e+00 3.21061492e-01 -3.66495192e-01 -1.03527009e+00
5.01728892e-01 -4.89708245e-01 -1.99962214e-01 7.32364714e-01
-1.76787719e-01 1.07530224e+00 8.50417554e-01 -3.71089399e-01
-7.00278759e-01 -2.06946477e-01 2.99827516e-01 1.36706263e-01
-3.66564840e-01 6.38650507e-02 -7.91286111e-01 -6.33376002e-01
-1.18472479e-01 -9.26162660e-01 -1.48662895e-01 3.40425730e-01
-1.43752933e-01 -1.39071405e-01 1.04869020e+00 -5.19597292e-01
1.30057383e+00 -2.01310205e+00 6.00759219e-03 2.65674610e-02
-1.09423205e-01 4.11518574e-01 -1.95135146e-01 2.88499087e-01
1.15009695e-01 -1.35788411e-01 -3.24965686e-01 -1.57586142e-01
-7.57242963e-02 6.80685282e-01 -2.26142257e-01 6.45271778e-01
2.73573399e-01 8.52022707e-01 -7.82129049e-01 -7.62518167e-01
4.01648730e-01 6.82049632e-01 -9.72247660e-01 -1.81381971e-01
-2.68035471e-01 1.68764547e-01 -4.64840889e-01 6.98712111e-01
3.49013209e-01 -3.78723294e-01 1.22810259e-01 -1.73083276e-01
1.40591443e-01 3.24613214e-01 -1.15769291e+00 1.83293700e+00
-2.27382898e-01 7.02963710e-01 -1.75154462e-01 -1.18373370e+00
3.87654811e-01 -6.99473172e-02 7.92872667e-01 -6.40430272e-01
-1.72007874e-01 -1.44460797e-01 -1.90576479e-01 -4.35885310e-01
2.60622710e-01 3.05531412e-01 1.55180365e-01 3.33421290e-01
-2.53956835e-03 4.99636143e-01 4.17374671e-01 1.32506043e-01
1.61724126e+00 3.43591928e-01 2.26520911e-01 -7.27638826e-02
5.45850813e-01 -1.62794963e-01 6.26625001e-01 7.71290064e-01
-3.91520351e-01 5.49980819e-01 7.52482474e-01 -5.47946215e-01
-5.98104835e-01 -7.47165322e-01 -7.81464800e-02 1.22616470e+00
-1.67314172e-01 -8.12233150e-01 -7.96909690e-01 -9.00072634e-01
-6.57613529e-03 6.29805863e-01 -6.08926237e-01 -2.64920205e-01
-8.30726445e-01 -4.34780300e-01 7.49470651e-01 8.77328932e-01
9.04276550e-01 -1.01505828e+00 -1.24625623e+00 7.73777440e-02
-4.15584326e-01 -1.33845782e+00 -4.95078474e-01 1.54156387e-01
-9.57388222e-01 -1.23589802e+00 -6.51050985e-01 -1.11332275e-01
4.75466967e-01 7.28178024e-02 1.07555425e+00 -1.48516977e-02
-6.22134745e-01 4.81884271e-01 -6.23021781e-01 -2.79550590e-02
-1.54088484e-02 2.24413306e-01 -2.45381206e-01 -1.71417594e-02
6.52298093e-01 -5.20147502e-01 -7.03557134e-01 4.11526918e-01
-8.91963542e-01 1.05884604e-01 5.94635606e-01 6.52692974e-01
6.82546020e-01 -4.24038470e-02 1.00483820e-01 -7.47588277e-01
-1.48204848e-01 -2.67807245e-01 -3.87875974e-01 5.50538041e-02
-3.89176458e-01 3.10115725e-01 3.96112800e-01 -4.27035004e-01
-8.33104432e-01 4.97185767e-01 -1.23648852e-01 -8.34518373e-01
-2.23170862e-01 3.09507027e-02 -4.23066095e-02 1.32389396e-01
5.52479386e-01 3.90422970e-01 1.13811389e-01 -5.61022878e-01
9.03830901e-02 3.97159338e-01 3.53159755e-01 -4.52540159e-01
1.26851231e-01 7.76113093e-01 1.04552358e-01 -8.28629613e-01
-7.75407910e-01 -5.15197873e-01 -6.63847983e-01 -1.89112231e-01
8.36836696e-01 -7.77009726e-01 -9.80760396e-01 2.48790979e-01
-8.48272622e-01 -5.52267551e-01 -5.03693223e-01 6.12674952e-01
-9.08479273e-01 6.01491749e-01 -5.91075301e-01 -6.63871169e-01
-7.14035183e-02 -1.08828247e+00 1.41002989e+00 -2.10522354e-01
-2.12070808e-01 -5.12482882e-01 -1.09556079e-01 4.13869917e-01
3.83668542e-01 1.01312533e-01 4.70070153e-01 -4.18926209e-01
-7.33333290e-01 -1.33170143e-01 -2.23143473e-01 2.42817223e-01
-1.24285430e-01 -1.56099677e-01 -8.92339230e-01 -1.99396148e-01
-3.79197389e-01 -3.84018689e-01 1.36666346e+00 4.62257385e-01
1.69160092e+00 -3.77434969e-01 -2.94602185e-01 6.39729738e-01
1.05684090e+00 4.29527909e-02 1.06297302e+00 1.29626155e-01
4.73076940e-01 2.64186323e-01 7.77236104e-01 7.83415794e-01
-1.51467277e-02 8.94586861e-01 4.35038179e-01 2.38643005e-01
-2.31881812e-01 -1.34448111e-01 6.95671797e-01 1.19083695e-01
-3.31263512e-01 -5.62817380e-02 -5.71408987e-01 5.31850100e-01
-1.95410776e+00 -1.29693949e+00 2.99285084e-01 2.02211571e+00
9.06170487e-01 2.01706618e-01 4.74075854e-01 4.07970369e-01
4.25198346e-01 2.71903276e-01 -6.33309484e-01 -2.36175537e-01
2.34069340e-02 3.64771128e-01 6.88710570e-01 3.66336346e-01
-1.23872614e+00 9.68603671e-01 6.84476042e+00 1.09455204e+00
-7.46758044e-01 -3.66074480e-02 4.97037709e-01 -7.28838384e-01
2.26508930e-01 -2.01052856e-02 -9.54356790e-01 4.44467455e-01
1.22453189e+00 2.98522890e-01 4.03641224e-01 1.00943017e+00
1.60550177e-01 -3.51140857e-01 -1.31551659e+00 1.18980634e+00
1.48941623e-02 -1.51496339e+00 1.95423797e-01 2.05748633e-01
3.51869911e-01 -4.35362272e-02 -2.90908337e-01 4.11951661e-01
5.86051680e-02 -1.03216660e+00 4.07374054e-01 6.54231548e-01
7.49787211e-01 -6.33289754e-01 5.43532491e-01 1.75979868e-01
-1.14904034e+00 -1.61474690e-01 -4.88253564e-01 -1.19533516e-01
1.68314666e-01 4.06170607e-01 -5.54696798e-01 8.95364508e-02
8.03640306e-01 1.04452598e+00 -7.06447423e-01 9.48952734e-01
1.07110657e-01 7.63724148e-01 -5.14897823e-01 5.04456311e-02
4.08966392e-01 2.63020068e-01 2.69756585e-01 1.46544766e+00
8.12577605e-02 1.20047480e-01 1.15190171e-01 4.51220930e-01
2.22107857e-01 -3.41289848e-01 -4.87951368e-01 -1.65179744e-01
2.58998126e-01 7.00208366e-01 -6.88188970e-01 -5.33635378e-01
-3.90175730e-01 1.36826813e+00 5.90731427e-02 2.05699965e-01
-1.24154878e+00 -2.22662255e-01 7.77652621e-01 3.03014219e-01
9.21236694e-01 -8.37481841e-02 1.63951844e-01 -1.12457716e+00
7.91300237e-02 -9.17358398e-01 6.13937199e-01 -4.43625987e-01
-6.23952389e-01 8.37662444e-02 3.87172729e-01 -1.23667157e+00
-4.61212575e-01 -5.58599651e-01 -7.09792133e-03 1.89994380e-01
-1.39710617e+00 -7.75870442e-01 -1.82258636e-01 8.34897459e-01
9.44374561e-01 3.13505083e-02 8.68552864e-01 4.39197510e-01
-5.82789004e-01 6.61577880e-01 -1.56226531e-01 1.65326998e-01
4.51452196e-01 -9.18298960e-01 3.05231959e-01 7.12404668e-01
2.67230719e-01 3.24158490e-01 5.40761888e-01 -4.57828283e-01
-1.53799009e+00 -1.16659081e+00 7.72689700e-01 -3.15639734e-01
3.33667576e-01 -2.23757550e-01 -6.54836357e-01 7.43460178e-01
-3.00655495e-02 4.64297950e-01 4.56124514e-01 6.13807468e-03
-5.81805468e-01 -2.95393944e-01 -9.76597190e-01 6.53110683e-01
1.56424785e+00 -5.02172291e-01 -4.07824010e-01 4.71567094e-01
3.15484881e-01 -2.11780146e-01 -9.31257606e-01 4.09870476e-01
6.90876126e-01 -1.04238677e+00 1.15145087e+00 -8.87284935e-01
4.34265941e-01 -1.99987680e-01 -3.90528589e-01 -6.01930439e-01
-2.67228693e-01 -6.77282155e-01 -7.78576314e-01 7.26410747e-01
4.53798175e-02 -1.83854029e-01 1.14762414e+00 4.57846522e-01
2.53705066e-02 -7.58364916e-01 -1.12152922e+00 -9.58600581e-01
-2.84565151e-01 -7.14873731e-01 3.10539007e-01 4.37446356e-01
-5.34104463e-03 3.07292677e-02 -4.15278941e-01 -2.45437965e-01
4.22328025e-01 7.07248524e-02 6.78146064e-01 -7.63016582e-01
-5.38679481e-01 -4.39386070e-01 -6.63239479e-01 -1.48174679e+00
1.25904366e-01 -4.54558164e-01 -3.95780914e-02 -1.11888337e+00
2.95746863e-01 4.28897478e-02 -2.73594439e-01 8.21198881e-01
1.70177698e-01 3.47593129e-01 2.08559573e-01 7.91043639e-02
-1.07964134e+00 4.86211866e-01 1.15789807e+00 -1.20541446e-01
1.47169545e-01 -8.65357742e-02 -2.17779994e-01 6.81838155e-01
5.22579074e-01 -3.51053536e-01 -7.19083786e-01 -3.21051866e-01
-1.51332021e-01 1.41998781e-02 5.61998963e-01 -1.38721859e+00
1.00849740e-01 -2.20482647e-01 5.43430686e-01 -7.45833457e-01
7.29568064e-01 -8.27959836e-01 1.21221775e-02 6.37292922e-01
-6.64067686e-01 -1.99628189e-01 2.06839025e-01 7.16039181e-01
-1.18916661e-01 -9.91569459e-02 7.42043138e-01 -2.11241275e-01
-9.57357407e-01 2.09205925e-01 -6.32017851e-01 1.34095311e-01
1.10458088e+00 -4.53763127e-01 -1.23411246e-01 -3.60234410e-01
-7.64600337e-01 2.16067005e-02 2.80320585e-01 1.85646385e-01
6.93409801e-01 -1.24338198e+00 -4.90024567e-01 3.25716168e-01
1.56764109e-02 -3.39284927e-01 2.93996364e-01 1.15360618e+00
-5.62265933e-01 6.68380618e-01 -2.35502154e-01 -7.77347624e-01
-1.43945074e+00 5.17744303e-01 2.45536402e-01 -3.24737132e-01
-1.01709855e+00 8.11143875e-01 -4.86980639e-02 3.16419631e-01
7.53816068e-01 -7.35170066e-01 -6.93860427e-02 1.13028688e-02
9.58908617e-01 6.27050579e-01 -1.20554287e-02 -3.29590380e-01
-6.52457058e-01 4.22539622e-01 -2.55785048e-01 2.66943648e-02
1.28343940e+00 1.44688904e-01 4.15308446e-01 1.75266564e-01
1.31139195e+00 -5.53983688e-01 -1.75617158e+00 -2.34336220e-02
-1.52136432e-02 -7.14365840e-01 1.55371398e-01 -6.31068826e-01
-1.26904023e+00 8.07297170e-01 5.01626611e-01 -7.36206695e-02
1.31535780e+00 -8.88536870e-02 8.80466938e-01 5.41788936e-01
3.79383236e-01 -1.29847383e+00 1.08830832e-01 4.61738855e-01
7.91222751e-01 -8.93273294e-01 2.58300811e-01 -4.61318374e-01
-7.05289721e-01 1.01633763e+00 4.93367970e-01 -1.40562192e-01
6.04074657e-01 4.22675371e-01 -7.25075066e-01 -1.96621478e-01
-1.06381047e+00 -2.16460571e-01 6.12758957e-02 5.03592372e-01
1.31356463e-01 -3.05977464e-01 -1.96740717e-01 1.55559495e-01
1.63068995e-01 3.66988927e-01 1.41603932e-01 1.18297529e+00
-4.09333050e-01 -8.93836915e-01 -5.40870428e-02 6.17949247e-01
-5.42330801e-01 8.57338756e-02 -2.80403167e-01 8.87333274e-01
-6.41616136e-02 7.08953559e-01 4.31107908e-01 -3.47146899e-01
1.38943091e-01 2.48141259e-01 6.73293650e-01 -2.13828444e-01
-3.11416417e-01 1.10229075e-01 3.07820439e-01 -1.40757084e+00
-7.42552519e-01 -1.01617348e+00 -1.28292549e+00 -3.73345107e-01
-5.32695018e-02 -2.31769934e-01 2.38194764e-01 1.00679100e+00
6.47531986e-01 4.70221668e-01 1.43140465e-01 -7.24558711e-01
-6.42780244e-01 -8.97894442e-01 -3.66493821e-01 3.40217382e-01
2.88228393e-01 -8.52678299e-01 -3.00880671e-01 3.57440233e-01] | [8.456531524658203, 0.5132519602775574] |
bbdb64f2-3835-47e7-964e-6956cdaf7760 | a-quantum-inspired-probabilistic-model-for | 2011.05511 | null | https://arxiv.org/abs/2011.05511v1 | https://arxiv.org/pdf/2011.05511v1.pdf | A Quantum-Inspired Probabilistic Model for the Inverse Design of Meta-Structures | In quantum mechanics, a norm squared wave function can be interpreted as the probability density that describes the likelihood of a particle to be measured in a given position or momentum. This statistical property is at the core of the microcosmos. Meanwhile, machine learning inverse design of materials raised intensive attention, resulting in various intelligent systems for matter engineering. Here, inspired by quantum theory, we propose a probabilistic deep learning paradigm for the inverse design of functional meta-structures. Our probability-density-based neural network (PDN) can accurately capture all plausible meta-structures to meet the desired performances. Local maxima in probability density distribution correspond to the most likely candidates. We verify this approach by designing multiple meta-structures for each targeted transmission spectrum to enrich design choices. | ['XueFeng Zhu', 'Yingtao Luo'] | 2020-11-11 | null | null | null | null | ['probabilistic-deep-learning'] | ['computer-vision'] | [ 1.64727911e-01 3.16999033e-02 -1.95831403e-01 -3.15485656e-01
-3.77940744e-01 -5.42544499e-02 4.86357540e-01 -1.07505724e-01
-1.10698916e-01 9.13650095e-01 3.87155294e-01 -1.23323165e-01
-6.33445144e-01 -1.27643073e+00 -7.91480422e-01 -1.25483680e+00
8.70743021e-02 4.54778582e-01 -1.02446407e-01 -1.40428111e-01
4.61799681e-01 3.12367946e-01 -1.11152565e+00 3.57313156e-01
6.48831427e-01 1.19021916e+00 4.24465537e-01 2.39726424e-01
7.14778006e-02 3.76386553e-01 -3.79850537e-01 -1.31441876e-01
-9.46165547e-02 -5.38260698e-01 -4.76892531e-01 -4.96859759e-01
-3.31631154e-01 -2.50478446e-01 -8.45592022e-01 1.28073180e+00
5.81793964e-01 5.65991327e-02 1.32761800e+00 -6.82201564e-01
-1.01388228e+00 1.06335795e+00 -1.09873950e-01 -8.78588855e-02
1.15005873e-01 4.31464761e-01 1.32535613e+00 -4.17496383e-01
1.03923254e-01 8.24084938e-01 2.06213281e-01 6.27357960e-01
-1.16830039e+00 -3.81482989e-01 -7.10059345e-01 4.92506742e-01
-1.33763993e+00 -2.67107219e-01 1.00892603e+00 -3.40978563e-01
7.23608613e-01 -2.47951314e-01 7.06679583e-01 1.12775183e+00
1.05518460e+00 5.29456854e-01 8.65435779e-01 -4.31188822e-01
5.26628315e-01 -2.20945682e-02 -2.81698376e-01 5.94587386e-01
3.18513155e-01 5.43932676e-01 -6.69866681e-01 -5.06027639e-02
5.47993124e-01 -3.99136096e-02 -1.91694260e-01 -3.53114337e-01
-1.12927663e+00 6.50725722e-01 6.63890421e-01 2.53441542e-01
-6.08107090e-01 5.73524177e-01 -8.96494091e-02 -1.99613646e-01
-4.02224958e-01 8.81269753e-01 -3.01057547e-01 -8.38565081e-02
-5.98255932e-01 3.02614123e-01 6.63021266e-01 5.59635818e-01
7.24079609e-01 8.43250155e-02 -4.12731916e-01 2.31582195e-01
7.06203520e-01 7.32446194e-01 4.17337231e-02 -9.01341259e-01
-4.54462953e-02 3.95072639e-01 2.91233331e-01 -6.81458294e-01
-5.77069461e-01 -8.44816327e-01 -1.15437233e+00 -1.37448609e-01
1.59739226e-01 -2.54881352e-01 -6.89852655e-01 1.72981620e+00
9.39650685e-02 -2.76360899e-01 5.93251176e-02 8.38532746e-01
5.23603916e-01 1.06882572e+00 -2.88557351e-01 -1.66261122e-01
1.07751882e+00 -1.63208067e-01 -4.68160689e-01 1.45294800e-01
2.21605822e-01 -4.13080066e-01 7.13594615e-01 3.89976591e-01
-6.71758473e-01 -1.22279592e-01 -1.33515453e+00 4.65230644e-01
1.00885101e-01 -9.13570076e-02 7.94742644e-01 8.07809532e-01
-5.07393777e-01 1.07364953e+00 -6.09467983e-01 4.56007719e-01
4.05662835e-01 5.39196193e-01 -9.69444460e-04 1.21232770e-01
-1.21336246e+00 7.62952626e-01 4.83689964e-01 3.19242328e-01
-1.08673942e+00 -6.24242902e-01 -1.58417687e-01 1.50229931e-01
-5.58590703e-02 -9.35275257e-01 1.02301884e+00 -1.74553499e-01
-1.88790596e+00 1.70087352e-01 1.35183454e-01 -3.24715555e-01
-1.81234643e-01 6.55412301e-02 -4.26704913e-01 -2.41771135e-02
-1.59308612e-01 3.55980158e-01 7.72687852e-01 -9.94430423e-01
-1.29873231e-01 -3.17601971e-02 -7.67865330e-02 -2.09939778e-01
-1.67187557e-01 -5.14429808e-01 4.33260471e-01 -4.34671938e-02
3.02279770e-01 -7.81032145e-01 -1.63311362e-01 -4.44887906e-01
-7.48869061e-01 -3.26046377e-01 1.89127907e-01 -7.14916885e-02
1.11547565e+00 -1.88252103e+00 3.43450069e-01 2.55711675e-01
2.63895124e-01 -6.42190725e-02 1.38091221e-01 5.67156911e-01
1.93329230e-01 -3.19371931e-02 -4.53159437e-02 4.89935815e-01
3.71671677e-01 -2.03449532e-01 -2.13007525e-01 7.52009809e-01
1.46841928e-01 7.42802978e-01 -7.94575036e-01 -7.03545809e-02
3.17799509e-01 2.89069086e-01 -9.01681721e-01 2.64954001e-01
-7.60425925e-01 5.70519626e-01 -8.26110363e-01 3.42353553e-01
7.07296610e-01 -3.38571131e-01 8.56456682e-02 -6.98452532e-01
-1.61198393e-01 4.87018734e-01 -7.16774702e-01 1.29613495e+00
-3.38394552e-01 3.61639172e-01 -1.97558716e-01 -1.01450825e+00
9.03522313e-01 1.68808743e-01 7.42984354e-01 -9.89482880e-01
3.94984037e-01 2.64104545e-01 5.92711210e-01 -5.42377949e-01
5.18037200e-01 -6.25987649e-01 -3.59644413e-01 5.33485353e-01
1.29142359e-01 -4.01873380e-01 -1.98046431e-01 -2.11631104e-01
9.97144818e-01 3.20988595e-02 8.87869745e-02 -4.16638643e-01
2.92820215e-01 -4.26096857e-01 4.72203255e-01 9.08110142e-01
-1.04066074e-01 3.48582000e-01 4.48574960e-01 -5.34513354e-01
-1.32245672e+00 -1.49086094e+00 -4.97004867e-01 5.50802827e-01
4.56183344e-01 -1.75358579e-01 -4.87600327e-01 1.32086024e-01
-2.25759134e-01 9.84052777e-01 -2.14618802e-01 -7.69450366e-01
-2.78505325e-01 -1.22266352e+00 3.17361921e-01 -3.28739993e-02
3.68923277e-01 -1.17936075e+00 -2.82261252e-01 4.57337528e-01
4.16153762e-03 -6.81356370e-01 3.61544155e-02 4.22913760e-01
-3.23807716e-01 -6.37933254e-01 -2.68122911e-01 -3.15200716e-01
3.02732617e-01 -3.58573169e-01 8.79195035e-01 -5.03228664e-01
-4.42645937e-01 5.03270701e-02 -1.00702345e-01 -1.87746555e-01
-7.25382864e-01 1.11847259e-01 4.70056742e-01 7.72366747e-02
3.35460573e-01 -9.79013324e-01 -9.27915454e-01 1.06169442e-02
-4.82131362e-01 1.12239264e-01 8.02834392e-01 6.40579641e-01
6.15740359e-01 5.84694684e-01 6.20842218e-01 -1.50981784e-01
5.51519096e-01 -5.91148555e-01 -6.89129770e-01 3.48241240e-01
-1.04101472e-01 7.89198041e-01 8.21726263e-01 -1.06429689e-01
-8.19952011e-01 -3.35502923e-01 -3.13850075e-01 1.62741542e-01
-1.16596438e-01 4.60264802e-01 -4.96948332e-01 -2.11317353e-02
5.95833540e-01 4.91118133e-01 -4.01417255e-01 -1.65636271e-01
4.24818844e-01 7.09156752e-01 2.05884933e-01 -9.57313240e-01
5.07300198e-01 1.49525866e-01 6.19066715e-01 -1.00466442e+00
-1.13059437e+00 1.29429042e-01 -2.60554701e-01 -2.30482146e-01
8.81707013e-01 -5.51891327e-01 -1.41611195e+00 4.11243826e-01
-1.19189692e+00 1.51323145e-02 5.92617691e-02 8.35493386e-01
-8.39133620e-01 -1.57583714e-01 -3.04374546e-01 -1.07196498e+00
-3.02104086e-01 -1.24978411e+00 8.95588398e-01 4.91509378e-01
-9.53224376e-02 -8.04188192e-01 1.65760711e-01 -1.22328341e-01
5.93001127e-01 1.72289819e-01 1.44485998e+00 -4.78096381e-02
-9.98990536e-01 -2.83108652e-01 -1.28337130e-01 1.40200898e-01
1.09376132e-01 -1.51503772e-01 -7.40748227e-01 3.51560079e-02
2.82253712e-01 -1.42814219e-01 8.62438619e-01 8.93242300e-01
1.23192370e+00 -2.40195394e-01 -3.02604258e-01 6.29425287e-01
1.44703293e+00 3.95993978e-01 5.30089438e-01 -4.71923910e-02
6.28127098e-01 2.48910129e-01 -1.66347608e-01 9.03164089e-01
1.50007844e-01 5.27740359e-01 5.28113842e-01 7.57065535e-01
3.11484575e-01 -3.25172931e-01 3.50892037e-01 6.83564782e-01
1.20477349e-01 -6.70835078e-01 -6.89445078e-01 1.80805504e-01
-1.43870258e+00 -1.17762935e+00 -7.72864383e-04 2.10805464e+00
7.27874339e-01 3.64729166e-01 -2.06301495e-01 -1.76966995e-01
8.11845720e-01 8.05821270e-02 -6.99946642e-01 -1.90110989e-02
-1.90509096e-01 2.37536475e-01 5.16901195e-01 2.41200626e-01
-8.39574695e-01 5.21583736e-01 6.62689447e+00 1.00290942e+00
-1.00277579e+00 -8.09370875e-02 4.86191511e-01 -1.43674836e-01
-7.49751508e-01 1.48268893e-01 -8.19314122e-01 7.59037673e-01
9.95724499e-01 -2.20565245e-01 7.45323300e-01 4.40109164e-01
2.26694703e-01 -1.35676309e-01 -9.87600803e-01 9.40045297e-01
-6.51752114e-01 -1.84996617e+00 1.49871647e-01 3.71278673e-01
8.27811062e-01 1.08457603e-01 4.07011092e-01 3.23000699e-02
1.03269085e-01 -1.10020053e+00 8.72201204e-01 8.67310584e-01
6.05989754e-01 -5.94894111e-01 4.65488076e-01 3.03046137e-01
-8.28892827e-01 -1.40425280e-01 -5.76122522e-01 -3.94274481e-02
3.95115972e-01 1.01509631e+00 -7.44488239e-01 3.30210626e-01
2.74513155e-01 3.72939765e-01 2.19306752e-01 1.03302372e+00
-1.69584677e-01 5.54519832e-01 -5.32729506e-01 -9.55232024e-01
1.10275842e-01 -4.16844159e-01 6.85326099e-01 5.32263637e-01
5.83114803e-01 -6.15414567e-02 -2.42714226e-01 1.69604838e+00
-2.11683795e-01 -4.13159490e-01 -4.21164066e-01 -5.27725756e-01
5.62880039e-01 1.06079900e+00 -5.52658677e-01 2.40427062e-01
4.66411561e-02 5.01634955e-01 1.10857710e-01 1.89116359e-01
-8.72470260e-01 -3.41806680e-01 5.69792747e-01 1.17812194e-01
7.45655075e-02 -4.87030387e-01 -2.44695768e-01 -8.24810922e-01
-1.93628103e-01 -1.91565260e-01 -2.50921398e-01 -5.41307807e-01
-1.59077644e+00 2.41732374e-01 -1.44740850e-01 -1.02137220e+00
2.71425486e-01 -1.07719481e+00 -6.47975743e-01 7.13802993e-01
-1.09676373e+00 -5.74227750e-01 2.71290541e-01 6.03220537e-02
-3.81585062e-01 -2.59375781e-01 7.83882380e-01 3.79464924e-02
-5.07179737e-01 2.54237950e-01 6.31896853e-01 -6.20825775e-02
1.67226419e-01 -1.29968548e+00 1.68313801e-01 4.28554147e-01
1.30716478e-02 7.75000393e-01 1.09802771e+00 -5.22960067e-01
-1.96561611e+00 -7.60825634e-01 2.01664031e-01 -2.14297965e-01
7.98808277e-01 -3.71739537e-01 -4.40300226e-01 -1.12527974e-01
1.38230445e-02 -1.16601244e-01 7.50988185e-01 -4.76426221e-02
1.00516908e-01 -1.51395798e-01 -1.08449876e+00 5.24104238e-01
8.89615536e-01 -5.83155632e-01 -3.94750595e-01 5.94489634e-01
8.06840301e-01 -1.09457418e-01 -9.65532601e-01 4.47957188e-01
7.66270459e-01 -1.10683370e+00 9.11058307e-01 -4.01999623e-01
6.66959226e-01 -4.05956835e-01 -6.38490200e-01 -1.27383506e+00
-6.39438629e-01 -7.88655460e-01 -2.12771013e-01 7.19700754e-01
5.48510373e-01 -2.67472178e-01 9.55076516e-01 4.94159937e-01
-2.42344901e-01 -6.60516322e-01 -1.27985609e+00 -5.80199420e-01
1.61108240e-01 -5.82693279e-01 8.32967937e-01 1.52955815e-01
1.19087081e-02 3.96009982e-01 -4.43911493e-01 4.07476068e-01
9.06982541e-01 2.51890779e-01 6.06985204e-02 -9.67770636e-01
-5.61531842e-01 -7.71145761e-01 -3.91512960e-01 -1.10481739e+00
6.11062422e-02 -9.74706352e-01 2.97348261e-01 -1.65381420e+00
2.88747340e-01 -5.26515484e-01 -5.18860340e-01 -2.45983154e-01
2.18803912e-01 -1.99789405e-01 -3.22266936e-01 -1.01411775e-01
-6.22681379e-01 1.34453106e+00 1.22679293e+00 -1.85062230e-01
-1.51080966e-01 4.05394614e-01 -6.75852537e-01 3.43610525e-01
8.86601925e-01 -5.32598555e-01 4.66603413e-03 -2.24332809e-01
9.56943035e-01 2.28759587e-01 2.80415803e-01 -1.35495985e+00
1.62880793e-01 -4.12578195e-01 3.66466671e-01 -5.38571894e-01
3.72958869e-01 -5.96662998e-01 1.46253556e-01 5.23923218e-01
-2.10698649e-01 -6.26310349e-01 -3.88693750e-01 6.79429889e-01
1.75385445e-01 -5.66812515e-01 9.12590742e-01 -1.35462955e-01
-3.08074564e-01 5.37553370e-01 -6.13310635e-01 -1.50085330e-01
5.68916619e-01 1.03017256e-01 -2.93992937e-01 -2.04706997e-01
-4.70300287e-01 -1.49553120e-01 2.26446554e-01 1.60549786e-02
6.46170139e-01 -1.37009668e+00 -4.93537903e-01 1.11744106e-01
-2.21211184e-02 -1.91480458e-01 4.97797310e-01 4.20553207e-01
-5.94173253e-01 4.87231404e-01 -1.60128593e-01 -6.73228204e-01
1.46541759e-01 3.05021197e-01 7.23064601e-01 2.92827841e-02
-2.83849090e-01 8.05173397e-01 1.16545551e-01 -3.44178945e-01
-2.69074351e-01 -2.35824555e-01 -7.25611439e-03 -5.52193820e-01
3.12685907e-01 7.26035163e-02 -2.67182082e-01 -3.65949988e-01
-4.35990393e-01 4.94585633e-01 2.52222717e-01 -1.74152195e-01
1.56349754e+00 7.43700340e-02 -2.93251961e-01 2.74817824e-01
1.16419721e+00 -1.37914166e-01 -1.17920721e+00 -1.11262649e-01
-1.44542828e-01 1.08661629e-01 4.57162946e-01 -6.36965811e-01
-6.08639538e-01 8.98607671e-01 5.61066151e-01 3.27433228e-01
6.03078246e-01 1.46219537e-01 7.71817029e-01 7.88089037e-01
7.05077410e-01 -9.50472057e-01 1.37827218e-01 4.95370656e-01
3.98135990e-01 -9.91432905e-01 4.48114946e-02 2.93227911e-01
-2.43333712e-01 1.35571063e+00 3.93626183e-01 -2.59894967e-01
1.03705370e+00 1.43043891e-01 -8.74535322e-01 -4.22462761e-01
-5.70253491e-01 2.48121738e-01 3.41605514e-01 5.45328021e-01
4.62593675e-01 5.08517802e-01 3.17751586e-01 8.89236450e-01
-4.13952559e-01 -3.64831716e-01 5.10301769e-01 5.44422030e-01
-9.48339283e-01 -9.80919361e-01 -1.01722874e-01 5.99457920e-01
-2.18368903e-01 -1.74596235e-01 3.44970785e-02 -1.02529466e-01
2.37238407e-01 7.87019730e-01 2.44989321e-02 -6.51012242e-01
1.45969419e-02 -2.90118605e-01 9.73390579e-01 -5.73983014e-01
3.18977982e-01 -2.56503969e-01 -3.54997694e-01 -2.02714369e-01
-2.99250990e-01 -3.41372311e-01 -1.30869281e+00 -3.38016570e-01
-4.84687299e-01 2.79382646e-01 5.34204662e-01 1.24513423e+00
3.94101948e-01 5.95261216e-01 7.88021922e-01 -8.32449257e-01
-5.73822796e-01 -8.14988315e-01 -7.99536526e-01 -2.43276343e-01
4.09683615e-01 -8.11369836e-01 -2.69590974e-01 -5.04610598e-01] | [5.418069362640381, 5.088384628295898] |
431dd1fb-150b-41c4-a999-33fd50313cab | affinitynet-semi-supervised-few-shot-learning | 1805.08905 | null | http://arxiv.org/abs/1805.08905v2 | http://arxiv.org/pdf/1805.08905v2.pdf | AffinityNet: semi-supervised few-shot learning for disease type prediction | While deep learning has achieved great success in computer vision and many
other fields, currently it does not work very well on patient genomic data with
the "big p, small N" problem (i.e., a relatively small number of samples with
high-dimensional features). In order to make deep learning work with a small
amount of training data, we have to design new models that facilitate few-shot
learning. Here we present the Affinity Network Model (AffinityNet), a data
efficient deep learning model that can learn from a limited number of training
examples and generalize well. The backbone of the AffinityNet model consists of
stacked k-Nearest-Neighbor (kNN) attention pooling layers. The kNN attention
pooling layer is a generalization of the Graph Attention Model (GAM), and can
be applied to not only graphs but also any set of objects regardless of whether
a graph is given or not. As a new deep learning module, kNN attention pooling
layers can be plugged into any neural network model just like convolutional
layers. As a simple special case of kNN attention pooling layer, feature
attention layer can directly select important features that are useful for
classification tasks. Experiments on both synthetic data and cancer genomic
data from TCGA projects show that our AffinityNet model has better
generalization power than conventional neural network models with little
training data. The code is freely available at
https://github.com/BeautyOfWeb/AffinityNet . | ['Aidong Zhang', 'Tianle Ma'] | 2018-05-22 | null | null | null | null | ['type-prediction'] | ['computer-code'] | [-1.29987270e-01 3.17134172e-01 -4.10712719e-01 -3.89982730e-01
-2.73148328e-01 -5.78196421e-02 1.87311441e-01 2.73786455e-01
-2.36878872e-01 5.44615209e-01 1.98658735e-01 -4.13097739e-01
-1.68539360e-01 -1.23324883e+00 -6.17219925e-01 -6.34909689e-01
-2.03490168e-01 4.57391560e-01 3.38223547e-01 -2.79671013e-01
-1.61830395e-01 4.69997048e-01 -1.07068908e+00 3.69859397e-01
6.22791648e-01 7.23022342e-01 5.00422157e-02 8.55316281e-01
-1.20630577e-01 7.60794699e-01 -3.76080662e-01 -1.46153092e-01
-7.94988777e-03 -2.26270616e-01 -1.01624215e+00 -3.18334162e-01
1.11694977e-01 -1.90011084e-01 -5.87910473e-01 1.07153785e+00
7.87619889e-01 1.47096619e-01 4.50436831e-01 -1.24510455e+00
-1.13569021e+00 4.72819924e-01 -2.34369412e-01 4.44533110e-01
5.18613635e-03 2.79761493e-01 1.17074227e+00 -8.58816564e-01
6.89551711e-01 1.04916418e+00 1.02259159e+00 6.64329112e-01
-1.15124559e+00 -4.11422163e-01 -2.50438391e-03 3.35283726e-01
-1.34476650e+00 9.24605131e-02 4.48322088e-01 -3.41651469e-01
1.22709692e+00 2.38325879e-01 8.32144320e-01 1.12507236e+00
5.33497036e-01 9.36697841e-01 3.42950284e-01 -1.42787799e-01
2.68133253e-01 -2.23238334e-01 7.69062340e-01 8.16592634e-01
2.33087063e-01 -1.37580439e-01 -7.91565329e-02 -3.44614267e-01
7.91838229e-01 6.47251070e-01 -3.78570646e-01 -1.45237699e-01
-9.85327303e-01 1.08107710e+00 1.13276315e+00 4.70812351e-01
-1.64556235e-01 3.58471066e-01 4.88035917e-01 4.18988734e-01
4.75606203e-01 6.20082438e-01 -6.89831018e-01 2.21581519e-01
-2.55305439e-01 -5.58386557e-02 6.42925084e-01 8.84344757e-01
7.72006691e-01 -1.61268532e-01 -4.65402275e-01 8.21414292e-01
-1.36521325e-01 -1.28652871e-01 7.95620322e-01 -2.86130995e-01
-1.56411588e-01 9.48142290e-01 -3.29812586e-01 -7.41744757e-01
-9.05952990e-01 -4.74817365e-01 -1.20660698e+00 1.16372975e-02
2.79516727e-01 -2.87366062e-01 -1.35806513e+00 1.59014165e+00
1.87679991e-01 4.07211840e-01 4.58280034e-02 7.94943213e-01
1.43072534e+00 6.76892817e-01 7.67902285e-02 2.46907428e-01
1.38979936e+00 -1.05842209e+00 -5.78651667e-01 -2.64201581e-01
9.42650974e-01 -1.90306753e-01 1.19915056e+00 6.16762750e-02
-6.46995246e-01 -3.83396566e-01 -8.93064737e-01 -5.36286533e-01
-9.75589931e-01 -3.14607739e-01 1.07986581e+00 4.62021351e-01
-1.26585090e+00 6.79634988e-01 -7.39826798e-01 -8.80472481e-01
7.39532888e-01 6.35088265e-01 -5.03702402e-01 -3.86372834e-01
-1.38384438e+00 5.87569892e-01 5.62852919e-01 -6.71626031e-02
-8.35562468e-01 -7.22547531e-01 -1.00643635e+00 5.31195402e-01
3.57875586e-01 -9.27816927e-01 9.97430384e-01 -9.27045941e-01
-1.18025017e+00 6.54110909e-01 2.93154716e-01 -3.67910951e-01
-7.45039135e-02 6.66521490e-03 -2.26786971e-01 -1.53245956e-01
-2.54936069e-01 7.42156565e-01 3.05687845e-01 -3.43218356e-01
-2.16674566e-01 -2.57184386e-01 6.00159727e-02 -9.93091166e-02
-5.10336876e-01 -8.64949450e-02 -4.65887398e-01 -5.38400650e-01
-1.47339866e-01 -9.13922429e-01 -6.69911742e-01 -5.36103100e-02
-6.86332822e-01 -5.23318172e-01 7.75775135e-01 -4.39665139e-01
8.86561871e-01 -2.14161015e+00 5.80690801e-02 1.50714919e-01
4.37627733e-01 4.50573117e-01 -4.16827738e-01 4.06178087e-01
-4.17953461e-01 2.76315033e-01 -5.26122898e-02 3.39159966e-01
-1.95245162e-01 2.92174518e-01 3.24958354e-01 3.86213005e-01
4.79054689e-01 1.43412948e+00 -1.22897446e+00 -4.67845276e-02
-5.03475219e-02 4.89044815e-01 -7.29075193e-01 1.20907299e-01
-3.21978629e-01 -2.11387780e-03 -5.63407004e-01 8.15315247e-01
4.20395255e-01 -8.75993371e-01 1.75756782e-01 6.20383434e-02
5.26311576e-01 -1.72429666e-01 -5.93708754e-01 1.68362486e+00
-6.25606924e-02 5.89823961e-01 -4.01259422e-01 -1.12145209e+00
8.16821277e-01 1.53549463e-01 2.83137858e-01 -2.89478183e-01
4.27625775e-01 -1.34371892e-02 3.62829804e-01 -6.47082448e-01
1.28208533e-01 -1.55809512e-02 -1.87634110e-01 2.29926795e-01
5.08195519e-01 3.63897830e-01 5.89492619e-02 1.90098614e-01
1.65603590e+00 -5.80427587e-01 7.56219864e-01 -4.01825666e-01
1.52633741e-01 -8.22681040e-02 8.59712780e-01 8.38163733e-01
-1.66504815e-01 6.82402670e-01 8.64271700e-01 -1.05653358e+00
-8.63248527e-01 -6.43221200e-01 -3.86864468e-02 1.27051127e+00
-2.60986269e-01 -6.08711362e-01 -5.53855658e-01 -9.03745055e-01
1.00530230e-01 3.94603044e-01 -1.14822257e+00 -5.33322453e-01
-8.89508128e-02 -1.18164730e+00 3.23138952e-01 7.55147338e-01
1.75661787e-01 -1.24137533e+00 -1.94194868e-01 7.60861859e-02
3.11926007e-01 -5.84425211e-01 -5.31641841e-01 4.03359026e-01
-6.46083891e-01 -1.37539208e+00 -7.19955564e-01 -9.88189459e-01
8.60896289e-01 1.12433210e-01 1.16419709e+00 4.55208033e-01
-7.07187116e-01 1.31352887e-01 -4.68554258e-01 -7.10273147e-01
-1.56076491e-01 4.34216797e-01 -1.90083146e-01 -2.02619676e-02
7.90256321e-01 -4.30051625e-01 -6.34067357e-01 1.54516613e-02
-8.73004258e-01 -5.61222956e-02 6.31138325e-01 1.34355342e+00
5.75143397e-01 -1.40644595e-01 7.40079284e-01 -1.22178340e+00
7.45084882e-01 -8.22310150e-01 -2.79915839e-01 2.12911636e-01
-2.72787243e-01 9.24368296e-03 8.04381192e-01 -3.28046650e-01
-3.61953408e-01 9.27667320e-02 -4.55567777e-01 -4.51942593e-01
-3.80121656e-02 7.73478031e-01 -1.67725161e-01 -5.36081791e-02
6.94149733e-01 5.16679697e-03 3.03124428e-01 -4.78655607e-01
2.10845500e-01 6.23898208e-01 4.60944101e-02 4.94145527e-02
1.30548522e-01 2.37155527e-01 6.72909841e-02 -7.96726406e-01
-7.63480067e-01 -3.83566409e-01 -5.80119431e-01 1.90733060e-01
9.70736563e-01 -5.95647871e-01 -7.46987879e-01 4.37158972e-01
-7.26762295e-01 -7.10864723e-01 -3.57200384e-01 4.17516530e-01
-2.82112807e-01 -1.65578462e-02 -9.87772882e-01 -1.16220661e-01
-6.24783635e-01 -1.01928365e+00 7.82488644e-01 4.04989958e-01
-1.60260066e-01 -1.24884391e+00 6.65019229e-02 -3.16129595e-01
4.08648252e-01 2.23479703e-01 1.23030984e+00 -1.21274698e+00
-4.15410429e-01 -6.25112474e-01 -2.55360305e-01 2.15079114e-01
2.40184203e-01 5.00567304e-03 -9.68572319e-01 -4.98068303e-01
-5.29443085e-01 -5.24247050e-01 1.27972317e+00 6.52328253e-01
1.67999709e+00 -1.94342554e-01 -7.54407287e-01 9.18174028e-01
1.46888924e+00 1.06070630e-01 8.32824826e-01 -4.90621943e-03
7.91044056e-01 1.30840927e-01 2.09541842e-01 1.24971755e-01
2.22642630e-01 2.51080871e-01 5.22111118e-01 -4.78301376e-01
-9.91771221e-02 -4.57959436e-02 -8.71833935e-02 6.36852026e-01
1.19368490e-02 -3.79408270e-01 -1.14260662e+00 4.96426105e-01
-1.96486259e+00 -8.03113461e-01 9.69244912e-02 1.95063317e+00
5.37020385e-01 -9.40726325e-02 4.23532352e-02 -2.36731440e-01
6.73564315e-01 -5.66199049e-02 -7.27339625e-01 -6.23125911e-01
6.86614960e-02 3.17133397e-01 3.26866031e-01 2.83212990e-01
-1.15375710e+00 8.70653868e-01 6.35839558e+00 8.54294479e-01
-1.13894618e+00 9.82277915e-02 1.07031596e+00 -8.69572014e-02
-2.48441979e-01 -3.06854159e-01 -6.80381060e-01 4.10437077e-01
1.03663683e+00 -2.70741075e-01 3.63047533e-02 1.05383205e+00
-2.16874495e-01 3.84909213e-01 -1.20675790e+00 8.14207613e-01
-6.94621205e-02 -1.69659865e+00 2.71256436e-02 1.11092240e-01
5.40342987e-01 3.94869566e-01 -9.05289222e-03 6.88172340e-01
6.45226359e-01 -1.39357889e+00 -3.32516462e-01 5.41131318e-01
7.37100601e-01 -7.87694216e-01 1.15362501e+00 2.77753532e-01
-7.92378128e-01 -2.73870826e-01 -9.17100966e-01 -1.07938483e-01
-2.67399609e-01 7.15974569e-01 -1.21253657e+00 4.41352725e-01
8.74099314e-01 1.00333071e+00 -6.32042646e-01 1.32937038e+00
-1.61924303e-01 4.69530195e-01 -2.02730939e-01 -4.09384072e-01
4.83616024e-01 1.28969654e-01 1.91891462e-01 1.25331616e+00
4.59743649e-01 2.07893029e-01 2.75418311e-01 6.17218137e-01
-3.83993447e-01 2.55414635e-01 -7.75754213e-01 -8.11479092e-02
1.27436623e-01 1.48793972e+00 -9.12897289e-01 -3.78643334e-01
-5.42476892e-01 8.57055187e-01 8.22034895e-01 3.51072669e-01
-5.31120420e-01 -7.86694527e-01 6.75383866e-01 3.17082740e-02
4.17265028e-01 3.69092613e-01 2.49187350e-02 -9.28608000e-01
-5.70449948e-01 -6.32375300e-01 9.58061099e-01 -6.81537151e-01
-1.70238304e+00 6.33552670e-01 -6.54891074e-01 -9.75002527e-01
2.13390347e-02 -9.60922480e-01 -1.03673708e+00 7.70739496e-01
-1.25122535e+00 -1.05962551e+00 -4.49868202e-01 5.36373377e-01
3.15817535e-01 -2.03882173e-01 1.18189454e+00 3.49655822e-02
-8.13394547e-01 6.80643082e-01 2.45346069e-01 5.45601726e-01
6.06253803e-01 -1.36387348e+00 6.67348146e-01 4.49639469e-01
-1.91682741e-01 6.62572324e-01 2.79287457e-01 -6.46991372e-01
-1.29858458e+00 -1.61649013e+00 7.11099386e-01 -2.25151062e-01
5.51023602e-01 -5.35660088e-01 -1.10648835e+00 1.06625342e+00
2.71518350e-01 7.35386908e-01 1.14372671e+00 5.34148276e-01
-1.75200671e-01 -2.14526728e-02 -1.12276673e+00 5.32547295e-01
9.75995958e-01 -2.57804245e-01 -2.02787817e-01 8.08775783e-01
9.90674675e-01 -3.07485193e-01 -1.07402933e+00 3.94747138e-01
2.87212700e-01 -7.50460565e-01 9.12315249e-01 -1.24907815e+00
2.62871116e-01 -3.11032850e-02 1.34654224e-01 -1.60507250e+00
-1.07550490e+00 -4.08354402e-01 -5.84352575e-02 6.69837415e-01
5.99514186e-01 -8.26124489e-01 9.73947167e-01 4.83394265e-01
-4.49346006e-01 -1.25891435e+00 -6.67034626e-01 -8.04382026e-01
2.06038743e-01 -7.24843442e-02 8.68624926e-01 1.22723770e+00
4.17930275e-01 5.66565216e-01 -3.16635549e-01 4.22548726e-02
1.59606650e-01 4.46589850e-03 5.58170617e-01 -1.19362104e+00
-3.94217014e-01 -4.84266818e-01 -9.60385680e-01 -4.78700817e-01
-2.22215299e-02 -1.43147540e+00 -1.47075891e-01 -1.79018021e+00
5.65717816e-01 -2.51742363e-01 -7.68325031e-01 1.02383316e+00
-5.09063721e-01 1.95919037e-01 -1.67695940e-01 -2.40470707e-01
-6.10141397e-01 4.21211183e-01 1.12379932e+00 -4.04485404e-01
-3.01739961e-01 4.88520227e-03 -8.51402998e-01 6.07721448e-01
1.00289130e+00 -2.55030274e-01 -3.06347996e-01 -2.62028277e-01
2.25914270e-01 -1.99345961e-01 2.35693872e-01 -8.83890033e-01
3.37228686e-01 3.22245285e-02 7.29747593e-01 -3.44571084e-01
7.23592564e-02 -4.78104591e-01 1.52866945e-01 7.31147170e-01
-2.78109670e-01 -7.93883577e-02 1.72440439e-01 5.31012118e-01
-2.77554914e-02 -1.65685788e-01 6.33958459e-01 -4.28683788e-01
-9.24296677e-01 9.67275977e-01 -3.64273280e-01 -2.01100990e-01
1.07049978e+00 -4.26894203e-02 -5.73286891e-01 -2.25984126e-01
-1.09433031e+00 5.33810258e-01 3.88674587e-01 3.90386462e-01
6.91291928e-01 -1.49717855e+00 -6.80862606e-01 3.73173863e-01
4.32534814e-01 7.33208656e-02 5.19301713e-01 7.41601586e-01
-5.33617914e-01 3.59032512e-01 -3.16296995e-01 -4.21724558e-01
-1.06481183e+00 1.13233197e+00 4.82795656e-01 -2.97997266e-01
-8.11400294e-01 1.25476217e+00 4.02491540e-01 -6.32696986e-01
7.92750344e-02 -5.02693415e-01 -3.21783096e-01 -1.72479242e-01
8.39382648e-01 1.52904436e-01 3.80057171e-02 -1.42589239e-02
-4.14460957e-01 1.58894703e-01 -4.90833133e-01 8.65077436e-01
1.64427245e+00 5.44826150e-01 -2.71331072e-01 3.47909212e-01
1.35767758e+00 -6.45840347e-01 -9.95865703e-01 -1.99916303e-01
-5.40147014e-02 -1.94873273e-01 2.21114457e-02 -6.35741532e-01
-1.31953740e+00 9.57865059e-01 5.13609767e-01 4.12884593e-01
1.01986754e+00 2.68213153e-01 7.35571742e-01 5.01278400e-01
1.43684968e-02 -9.14239526e-01 1.90834090e-01 5.92030227e-01
8.07113111e-01 -1.28675902e+00 -1.95716068e-01 -3.59099597e-01
-4.03194040e-01 1.20196378e+00 8.08276653e-01 -2.69287527e-01
9.45927680e-01 2.37921834e-01 -2.57474840e-01 -6.15555882e-01
-1.07745230e+00 -3.15656841e-01 1.99340433e-01 6.68531001e-01
5.27321458e-01 1.45337686e-01 -6.34207129e-02 1.08700848e+00
5.86088374e-02 3.30575705e-01 6.59030318e-01 7.56258249e-01
-5.89640200e-01 -8.54612172e-01 1.05893925e-01 1.13693833e+00
-4.24789757e-01 -2.93501645e-01 -4.37812835e-01 7.88596988e-01
3.40948440e-02 3.09074968e-01 2.49111801e-01 -6.65172935e-01
2.34197021e-01 1.10287629e-01 1.58482954e-01 -9.81626928e-01
-5.16161382e-01 -2.37427980e-01 -4.09661904e-02 -6.54587328e-01
1.74736753e-01 -1.92897424e-01 -1.17136002e+00 -4.22450185e-01
-1.88656837e-01 7.81524647e-03 7.22329840e-02 5.89430749e-01
6.64311230e-01 8.43158484e-01 2.27633521e-01 -6.28073096e-01
-2.65280724e-01 -1.04289615e+00 -8.36699963e-01 3.14862907e-01
3.02212089e-01 -3.49225581e-01 5.17643988e-02 -3.99007261e-01] | [6.922696590423584, 6.250348091125488] |
8c5f1821-1b9e-495f-af5a-c886b43e8a29 | long-document-re-ranking-with-modular-re | 2205.04275 | null | https://arxiv.org/abs/2205.04275v2 | https://arxiv.org/pdf/2205.04275v2.pdf | Long Document Re-ranking with Modular Re-ranker | Long document re-ranking has been a challenging problem for neural re-rankers based on deep language models like BERT. Early work breaks the documents into short passage-like chunks. These chunks are independently mapped to scalar scores or latent vectors, which are then pooled into a final relevance score. These encode-and-pool methods however inevitably introduce an information bottleneck: the low dimension representations. In this paper, we propose instead to model full query-to-document interaction, leveraging the attention operation and modular Transformer re-ranker framework. First, document chunks are encoded independently with an encoder module. An interaction module then encodes the query and performs joint attention from the query to all document chunk representations. We demonstrate that the model can use this new degree of freedom to aggregate important information from the entire document. Our experiments show that this design produces effective re-ranking on two classical IR collections Robust04 and ClueWeb09, and a large-scale supervised collection MS-MARCO document ranking. | ['Jamie Callan', 'Luyu Gao'] | 2022-05-09 | null | null | null | null | ['document-ranking'] | ['natural-language-processing'] | [ 3.15658636e-02 9.66575965e-02 -3.69617432e-01 -3.37336153e-01
-1.44959402e+00 -6.75018787e-01 9.63381112e-01 5.44986367e-01
-5.43735921e-01 5.25195837e-01 8.94235790e-01 1.64309561e-01
-3.12937230e-01 -4.43037450e-01 -7.26355076e-01 -2.62607276e-01
-3.00835520e-01 8.70497227e-01 1.08468622e-01 -2.94737935e-01
6.49745703e-01 6.14722446e-02 -1.68689728e+00 8.82311285e-01
9.47306633e-01 6.65828884e-01 4.73728776e-01 1.01196718e+00
-3.27493876e-01 1.08771431e+00 -7.35814214e-01 -3.27118903e-01
2.02277422e-01 -1.60237402e-01 -1.13169849e+00 -5.24321079e-01
5.72260857e-01 -8.87809694e-01 -3.61580908e-01 7.17436075e-01
7.00982213e-01 4.65845734e-01 7.04788744e-01 -6.37989402e-01
-1.15681696e+00 1.25639820e+00 -2.75622398e-01 6.93481192e-02
4.36028183e-01 -5.68558395e-01 1.76283836e+00 -1.16012788e+00
6.70963585e-01 1.36107516e+00 1.96965531e-01 5.52047372e-01
-1.14035165e+00 -4.15055990e-01 6.87614679e-02 1.84405953e-01
-9.10279155e-01 -4.04831529e-01 5.06406784e-01 -3.48939776e-01
1.12622476e+00 3.07167619e-01 4.29820895e-01 7.56430089e-01
2.37911195e-02 1.21669102e+00 4.17951256e-01 -4.33256090e-01
1.69136375e-02 -5.86458258e-02 8.62853229e-01 4.04584408e-01
1.01712130e-01 -1.76927254e-01 -7.54154503e-01 -3.27528507e-01
2.75827527e-01 2.67099649e-01 -9.31553990e-02 -3.53917956e-01
-1.23180997e+00 9.05505896e-01 7.89818883e-01 2.67421782e-01
-2.29355812e-01 3.85608286e-01 4.10529971e-01 5.64291418e-01
5.51931798e-01 8.70156288e-01 -3.30606043e-01 5.66850528e-02
-1.12613404e+00 3.70807201e-01 7.10005581e-01 6.93020821e-01
5.95217824e-01 -6.65755093e-01 -9.07257020e-01 1.18093646e+00
5.17681658e-01 3.12046170e-01 8.13278258e-01 -8.00495803e-01
6.02622569e-01 4.24049646e-01 2.22157821e-01 -6.64556623e-01
-2.54311532e-01 -5.83091736e-01 -5.77889383e-01 -1.86432153e-01
-1.31392866e-01 2.82539010e-01 -9.83757079e-01 1.67693937e+00
-2.17092976e-01 -3.72270465e-01 2.60595322e-01 8.47035885e-01
9.12206113e-01 8.63531590e-01 7.02511221e-02 5.10944054e-02
1.19101822e+00 -1.40370882e+00 -4.05543149e-01 -1.29103214e-01
8.52618456e-01 -7.71620214e-01 1.10995328e+00 1.66510612e-01
-1.50338948e+00 -6.31558299e-01 -1.17239583e+00 -7.91388690e-01
-2.77966291e-01 2.87379682e-01 5.50667822e-01 -8.60678479e-02
-1.39303601e+00 6.56401813e-01 -5.04599631e-01 -2.36394972e-01
1.62370250e-01 6.28132284e-01 -2.58322954e-01 -2.03769147e-01
-1.28669524e+00 8.48679245e-01 3.65161687e-01 7.24068433e-02
-8.28728080e-01 -5.96318841e-01 -5.71269751e-01 6.55584991e-01
-9.35338885e-02 -9.20236528e-01 1.65346515e+00 -6.59305394e-01
-1.63100207e+00 6.53889656e-01 -2.86462098e-01 -3.67547840e-01
1.55579988e-02 -7.31158674e-01 -4.00785916e-02 1.25553131e-01
1.50910214e-01 8.00928116e-01 4.96775538e-01 -1.14732981e+00
-5.39177120e-01 -3.27282846e-01 3.36963415e-01 6.57947242e-01
-5.57431340e-01 2.75142759e-01 -5.97113132e-01 -6.15781248e-01
8.42497349e-02 -8.50652337e-01 -4.67959158e-02 -4.83502090e-01
-3.21352005e-01 -5.47226429e-01 1.32641092e-01 -6.56370640e-01
1.33568192e+00 -1.93243682e+00 4.89778996e-01 1.71414763e-01
4.79915410e-01 2.04488486e-02 -7.47871816e-01 7.47082770e-01
6.67132735e-02 2.22030148e-01 1.42841175e-01 -6.24088347e-01
1.92447230e-01 -2.04721704e-01 -6.52412236e-01 2.22545974e-02
-7.40283355e-02 1.29869509e+00 -1.15897048e+00 -3.90177310e-01
-4.17148054e-01 3.69488776e-01 -7.73524165e-01 3.62509012e-01
-3.19186628e-01 -1.65344015e-01 -5.24761140e-01 3.26908022e-01
3.35507512e-01 -4.28297430e-01 8.18566382e-02 -1.20829873e-01
1.23969384e-01 7.96728969e-01 -6.39889181e-01 2.01401877e+00
-5.38678765e-01 6.88445270e-01 -1.50816768e-01 -7.23524272e-01
7.22672164e-01 1.04428381e-01 4.18129086e-01 -8.24135602e-01
-2.29264393e-01 3.01495939e-01 -2.44162291e-01 -4.29021977e-02
1.31586707e+00 3.09087276e-01 -2.92996615e-01 9.44866061e-01
1.10597268e-01 -9.44177341e-03 4.98763472e-01 9.81141031e-01
1.15972698e+00 -2.42564246e-01 -6.46659508e-02 -2.41579950e-01
2.36432090e-01 -2.05162585e-01 -1.18990466e-01 1.02384007e+00
4.84476030e-01 7.48827338e-01 4.63871121e-01 -1.16588704e-01
-1.04088402e+00 -1.10023034e+00 3.63542698e-02 1.89162207e+00
1.28460065e-01 -7.27462709e-01 -2.84333557e-01 -7.00365663e-01
1.70269519e-01 4.74739313e-01 -7.25703835e-01 -3.26515466e-01
-5.46546161e-01 -5.16987145e-01 2.54412383e-01 5.43160975e-01
2.72802711e-02 -1.11105454e+00 -1.81854188e-01 4.48660731e-01
-2.47221723e-01 -2.74521440e-01 -6.86784267e-01 4.57377672e-01
-8.95183027e-01 -7.00314283e-01 -1.16662765e+00 -7.91819990e-01
6.62491858e-01 5.49612582e-01 1.44908071e+00 2.81024873e-01
1.46334767e-02 3.59819412e-01 -5.57090878e-01 -3.72294635e-02
-3.17022741e-01 4.67269361e-01 -8.62884447e-02 -3.17836285e-01
2.24630654e-01 -2.26159960e-01 -7.20171869e-01 5.24330661e-02
-8.98312032e-01 4.96240258e-02 9.25400376e-01 1.14314556e+00
4.68244910e-01 -5.09592414e-01 6.41746759e-01 -7.01478183e-01
1.39649343e+00 -5.14455020e-01 -3.90384287e-01 6.25587106e-01
-4.84716982e-01 7.01023221e-01 2.82496274e-01 -2.87156105e-01
-9.11615193e-01 -1.53631195e-01 2.53383756e-01 9.63776708e-02
6.41430676e-01 9.21030700e-01 1.68950096e-01 5.19872546e-01
6.16286039e-01 1.41671047e-01 -3.87419134e-01 -8.24583054e-01
7.97850072e-01 9.61121082e-01 4.66077179e-01 -6.60084009e-01
5.21037936e-01 4.09406647e-02 -5.07250488e-01 -2.92467237e-01
-1.03565693e+00 -8.60115767e-01 -7.18830645e-01 1.67279765e-01
6.24501526e-01 -1.17219055e+00 -4.10690635e-01 1.22637831e-01
-1.49006307e+00 -2.88939923e-01 -5.23233950e-01 5.09907603e-01
-2.74330616e-01 3.41542333e-01 -9.05516386e-01 -3.79697621e-01
-7.13348329e-01 -9.34736431e-01 1.47490966e+00 2.39923418e-01
-2.20276415e-01 -5.17478883e-01 6.23583198e-01 3.50372404e-01
5.04105389e-01 -7.02250302e-01 1.02256298e+00 -8.17500830e-01
-6.79657817e-01 -4.10944551e-01 -5.50413132e-01 1.63246840e-01
-2.28607401e-01 -1.97494045e-01 -9.85747159e-01 -5.29191375e-01
-5.90029538e-01 -7.61298776e-01 1.53738725e+00 1.38463542e-01
1.15282464e+00 -3.96052659e-01 -3.01961064e-01 2.68228650e-01
1.02072656e+00 -1.73501953e-01 5.54053426e-01 2.07023576e-01
7.44608700e-01 4.94186729e-01 4.14445907e-01 2.56568104e-01
5.40016890e-01 7.83786476e-01 1.17749706e-01 1.97678506e-01
-1.27925411e-01 -2.20252767e-01 6.75220728e-01 1.24176860e+00
1.54165132e-03 -5.68900168e-01 -7.21422553e-01 3.98653179e-01
-1.94425941e+00 -1.20416260e+00 2.88774997e-01 2.31912279e+00
1.09457660e+00 -1.58026487e-01 -2.30526194e-01 -2.56246239e-01
5.02780914e-01 2.94694990e-01 -4.34906334e-01 -4.30228174e-01
-9.91248619e-03 4.97571751e-02 2.40593523e-01 7.75801957e-01
-9.72509563e-01 9.80673552e-01 6.43907166e+00 6.04506135e-01
-7.95031667e-01 -3.50057259e-02 3.51495475e-01 -6.15375817e-01
-7.57682383e-01 -1.26470208e-01 -8.45080614e-01 2.87927449e-01
1.08278716e+00 -4.27948803e-01 4.65083897e-01 7.34508634e-01
-4.14451838e-01 1.01450279e-01 -1.50694919e+00 8.39234531e-01
3.03067803e-01 -1.32062268e+00 3.86856467e-01 1.19711347e-01
8.07712793e-01 4.58852738e-01 1.34087533e-01 8.03033113e-01
8.04539680e-01 -1.01428390e+00 5.86497843e-01 7.85025656e-01
7.00257123e-01 -6.25421047e-01 7.00519800e-01 8.99830610e-02
-1.04547763e+00 -3.64100456e-01 -8.86981189e-01 1.84021890e-01
3.10802832e-02 6.06037140e-01 -8.03436756e-01 3.99874359e-01
5.19189417e-01 8.09676886e-01 -8.89290452e-01 1.14117754e+00
-8.93464610e-02 1.05481125e-01 -1.54734179e-01 -2.46280298e-01
1.50171012e-01 3.12381208e-01 3.80364746e-01 1.20060968e+00
4.35974419e-01 -1.42329589e-01 1.18593052e-01 4.26674217e-01
-6.25726223e-01 2.00475246e-01 -2.20794350e-01 -3.59107047e-01
3.32873285e-01 1.16475153e+00 -2.37683862e-01 -7.68329382e-01
-1.85153827e-01 1.32826257e+00 8.58589292e-01 4.94001627e-01
-2.97633678e-01 -6.15174651e-01 4.64400649e-01 -2.01516107e-01
2.09413260e-01 -1.27637491e-01 3.12329531e-02 -1.46644878e+00
1.18168503e-01 -7.03412950e-01 4.02165085e-01 -8.47816706e-01
-1.29207480e+00 7.30572164e-01 -6.86367303e-02 -1.00087953e+00
-6.15607202e-01 -2.71848202e-01 -5.10761857e-01 9.03539538e-01
-1.53544235e+00 -1.16336513e+00 -1.15211658e-01 3.33144158e-01
6.12927020e-01 -2.71597475e-01 9.29020464e-01 2.46104226e-01
-2.10611701e-01 7.17322886e-01 8.40537429e-01 1.91855605e-03
9.42890704e-01 -1.50804543e+00 3.76141816e-01 3.91380221e-01
5.17607749e-01 1.09435833e+00 1.85753971e-01 -5.47211349e-01
-1.19480765e+00 -7.15106785e-01 1.39836240e+00 -7.58030474e-01
4.97037888e-01 -5.86101949e-01 -8.91041040e-01 4.27908927e-01
3.37475896e-01 -2.04493791e-01 7.83709705e-01 5.42853653e-01
-6.76824152e-01 -8.85909945e-02 -4.57509726e-01 5.97246289e-01
8.65018010e-01 -9.70481217e-01 -1.06275475e+00 5.10207236e-01
1.07031393e+00 2.95784650e-03 -6.96180642e-01 8.31436068e-02
9.79797721e-01 -5.73836803e-01 1.29512680e+00 -8.19199741e-01
6.57081485e-01 -1.15207024e-01 -2.26424679e-01 -1.46107173e+00
-6.63708270e-01 -4.47527528e-01 -4.16975170e-01 1.05347311e+00
5.00569522e-01 -6.85433075e-02 6.15056872e-01 6.23250544e-01
-2.23060414e-01 -5.95568061e-01 -6.14563048e-01 -4.81220871e-01
1.83288708e-01 3.39540206e-02 7.00035989e-01 5.14633715e-01
4.63656664e-01 7.96335101e-01 -2.60617733e-01 -2.36763716e-01
5.03614128e-01 3.22712839e-01 6.49065673e-01 -1.59098673e+00
-5.27862489e-01 -7.30002165e-01 1.12032376e-01 -1.64707184e+00
1.48949221e-01 -1.40807128e+00 2.88549960e-01 -1.94420207e+00
8.42671335e-01 -3.55152875e-01 -8.02600026e-01 3.25669199e-01
-3.93826067e-01 3.19513455e-02 2.68751830e-01 7.61513650e-01
-1.19971359e+00 6.88449979e-01 1.02289355e+00 -5.98637998e-01
-3.93753290e-01 -2.23166928e-01 -8.83752644e-01 1.14904530e-01
6.27042532e-01 -5.83239913e-01 -5.50041556e-01 -1.00354016e+00
6.67249560e-01 9.89838466e-02 2.67916799e-01 -7.43416190e-01
4.95253384e-01 2.37537116e-01 4.68431234e-01 -8.37704659e-01
2.41383299e-01 -4.35904294e-01 -5.33077657e-01 2.43211776e-01
-1.34718621e+00 6.81822971e-02 -1.42805830e-01 5.84092617e-01
-4.15569901e-01 -3.21788311e-01 3.22738200e-01 -4.07424942e-02
-4.94476736e-01 2.46466279e-01 -2.15493038e-01 -7.91231915e-02
2.82708079e-01 2.11286366e-01 -6.13522530e-01 -5.07751644e-01
-5.53873837e-01 3.98598880e-01 2.83044547e-01 6.34092093e-01
5.81950486e-01 -1.43720543e+00 -1.01806426e+00 1.00502893e-01
3.62382352e-01 -5.94919771e-02 2.36688644e-01 2.61100233e-01
-1.91437379e-01 9.28556621e-01 -4.28791679e-02 -2.57574439e-01
-1.15145278e+00 4.34489399e-01 -1.16237454e-01 -8.72871578e-01
-3.81860882e-01 1.06053233e+00 3.43653053e-01 -3.46466452e-01
4.65518981e-01 -4.48074877e-01 -4.62872267e-01 5.41358232e-01
8.67002130e-01 7.65073821e-02 3.84316593e-02 -3.72657955e-01
-2.10325748e-01 4.95029092e-01 -7.99657106e-01 -4.51780856e-01
1.40033937e+00 -2.39383295e-01 -2.73501486e-01 1.83322370e-01
1.65418875e+00 1.12055384e-01 -8.04936528e-01 -4.26722288e-01
3.24010968e-01 -2.70267874e-01 1.46157831e-01 -9.57866311e-01
-6.00424826e-01 9.81098533e-01 4.17864174e-01 1.03746511e-01
7.35955834e-01 3.19561332e-01 7.62039661e-01 1.10497856e+00
1.57959402e-01 -1.09298861e+00 5.47262490e-01 9.12885129e-01
1.27565634e+00 -1.20097339e+00 6.09804615e-02 4.99196827e-01
-5.61838090e-01 1.02085745e+00 3.81109744e-01 -5.48247807e-02
4.39746201e-01 -2.36971498e-01 -1.95950344e-01 -2.51635224e-01
-1.08255184e+00 -1.55059651e-01 8.28787267e-01 2.34725758e-01
8.28284800e-01 -7.52281472e-02 -2.81710625e-01 8.39614928e-01
-1.32833049e-01 1.48771415e-02 3.09800863e-01 7.58756638e-01
-7.79601276e-01 -1.34298158e+00 -1.33980289e-02 6.25472963e-01
-4.16893929e-01 -6.95968866e-01 -4.87648576e-01 1.54822975e-01
-5.30578077e-01 6.45876586e-01 2.31106207e-01 -4.22086358e-01
-3.42947729e-02 1.18614718e-01 2.68327653e-01 -1.05403864e+00
-7.50464320e-01 -4.33069207e-02 5.09436131e-02 -6.38989985e-01
-6.42034635e-02 -3.50100428e-01 -1.14726722e+00 6.38135821e-02
-3.61627460e-01 6.64387107e-01 7.10457563e-01 4.92253840e-01
7.47759938e-01 4.53815967e-01 5.88097930e-01 -1.06710362e+00
-8.83571506e-01 -1.15825510e+00 -4.91598099e-01 4.12581980e-01
5.33828080e-01 -3.53763878e-01 -4.08455491e-01 -1.37611449e-01] | [11.441839218139648, 7.632616996765137] |
000dcad9-33ba-4c82-a394-590ea8d09b35 | systematic-literature-review-on-application | 2305.12695 | null | https://arxiv.org/abs/2305.12695v1 | https://arxiv.org/pdf/2305.12695v1.pdf | Systematic Literature Review on Application of Machine Learning in Continuous Integration | This research conducted a systematic review of the literature on machine learning (ML)-based methods in the context of Continuous Integration (CI) over the past 22 years. The study aimed to identify and describe the techniques used in ML-based solutions for CI and analyzed various aspects such as data engineering, feature engineering, hyper-parameter tuning, ML models, evaluation methods, and metrics. In this paper, we have depicted the phases of CI testing, the connection between them, and the employed techniques in training the ML method phases. We presented nine types of data sources and four taken steps in the selected studies for preparing the data. Also, we identified four feature types and nine subsets of data features through thematic analysis of the selected studies. Besides, five methods for selecting and tuning the hyper-parameters are shown. In addition, we summarised the evaluation methods used in the literature and identified fifteen different metrics. The most commonly used evaluation methods were found to be precision, recall, and F1-score, and we have also identified five methods for evaluating the performance of trained ML models. Finally, we have presented the relationship between ML model types, performance measurements, and CI phases. The study provides valuable insights for researchers and practitioners interested in ML-based methods in CI and emphasizes the need for further research in this area. | ['Muhammad Ali Babar', 'Mansooreh Zahedi', 'Triet Huynh Minh Le', 'Ali Kazemi Arani'] | 2023-05-22 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [-9.78298672e-03 5.77465072e-02 -9.57952201e-01 -2.27702975e-01
-7.81222463e-01 -3.27256352e-01 4.80062902e-01 5.90701103e-01
-4.19274807e-01 6.56686842e-01 -7.29539618e-02 -2.93415546e-01
-7.13514388e-01 -6.74816728e-01 -4.84421015e-01 -4.57775712e-01
-9.05436575e-02 3.49721700e-01 5.63392136e-03 1.82390183e-01
9.61405694e-01 5.05919933e-01 -2.04040241e+00 5.19154727e-01
9.86404181e-01 1.02856159e+00 1.19391873e-01 4.59833622e-01
-3.30022514e-01 7.74922669e-01 -1.12031066e+00 -4.37463224e-01
-2.08038196e-01 -2.42648154e-01 -4.71575290e-01 -1.56342089e-01
1.51887581e-01 2.15427220e-01 3.15240204e-01 3.06315243e-01
5.02886951e-01 -1.15805812e-01 1.03838575e+00 -1.54419732e+00
-3.68889660e-01 4.19349313e-01 5.04587032e-02 -1.02206413e-02
6.16626859e-01 -2.49027926e-02 7.03731298e-01 -1.38375092e+00
6.09257877e-01 1.00189471e+00 8.02372515e-01 3.14528406e-01
-8.53913367e-01 -7.29871273e-01 -1.87945440e-01 1.90402299e-01
-1.30208015e+00 -2.62492716e-01 4.60448146e-01 -1.04155338e+00
1.43863618e+00 3.50480676e-01 9.92778301e-01 5.30635238e-01
4.97442365e-01 7.50053406e-01 1.10458302e+00 -9.89630699e-01
3.72168571e-01 1.01099253e+00 3.64781469e-01 2.77446419e-01
6.61287248e-01 1.12106308e-01 -4.00705785e-01 -4.55925196e-01
5.25333345e-01 -3.31627846e-01 2.86993653e-01 -1.10956408e-01
-7.03209162e-01 1.15072560e+00 -4.97480817e-02 6.12357855e-01
-4.15964276e-01 -5.86133778e-01 6.33132935e-01 1.57617196e-01
3.44078928e-01 7.70448506e-01 -8.42620432e-01 -1.82681978e-01
-7.79195070e-01 2.49204636e-01 9.92259860e-01 1.11627543e+00
4.07758862e-01 5.74969389e-02 -1.79513231e-01 8.99952114e-01
4.16795880e-01 1.80476516e-01 6.80371344e-01 -7.89277077e-01
5.33645749e-01 1.03786731e+00 -6.45802096e-02 -1.07343781e+00
-4.82194364e-01 -2.42046610e-01 -3.36175412e-01 8.99578184e-02
-1.35738805e-01 -2.57736176e-01 -4.84024614e-01 1.04636514e+00
9.17687546e-04 -5.78741789e-01 2.01875895e-01 4.79398780e-02
1.36390042e+00 3.93140405e-01 2.63118684e-01 -6.54410839e-01
1.02296770e+00 -7.45128155e-01 -1.02784514e+00 3.94329131e-02
9.19960737e-01 -7.85101593e-01 1.00035214e+00 5.94361365e-01
-1.23173714e+00 -7.47581005e-01 -1.12224782e+00 4.42730010e-01
-8.28874052e-01 3.54965478e-01 5.22868097e-01 9.96684432e-01
-5.05660355e-01 4.19602245e-01 -5.01721919e-01 -2.41474405e-01
1.74537167e-01 3.98651630e-01 -3.79782319e-02 3.73067170e-01
-1.20168543e+00 1.23623633e+00 5.08113623e-01 -4.86523509e-02
-3.44692230e-01 -8.14052701e-01 -8.08425903e-01 -3.34374279e-01
1.47032738e-01 -3.94643694e-01 9.90703464e-01 -7.75528729e-01
-1.11520112e+00 5.74165106e-01 -1.06185652e-01 -7.14705363e-02
1.06795117e-01 -2.20299065e-01 -6.65453076e-01 -1.25148311e-01
-5.77129237e-02 3.99176002e-01 3.27563435e-01 -1.24056900e+00
-1.05298996e+00 -1.88394576e-01 -3.43773425e-01 1.87468573e-01
-1.87209651e-01 3.26224655e-01 -4.70308185e-01 -4.86896694e-01
-1.07971661e-01 -5.24250507e-01 2.25699507e-02 -4.87799257e-01
2.46205553e-02 -5.38559973e-01 5.13388157e-01 -7.34920144e-01
2.05588293e+00 -2.04616714e+00 -2.86298275e-01 4.05769229e-01
-2.15166677e-02 4.61439162e-01 2.88192540e-01 9.52673018e-01
1.48864925e-01 5.47695994e-01 -1.73809171e-01 -5.35449721e-02
-1.92663949e-02 -5.75230345e-02 2.32890233e-01 5.13924211e-02
1.23707369e-01 6.67319298e-01 -5.12598574e-01 -5.92622101e-01
5.06102562e-01 6.29158974e-01 -1.67968690e-01 2.68469334e-01
2.11792048e-02 4.16288758e-03 -1.01821065e-01 7.67096102e-01
4.06464607e-01 2.54869044e-01 1.91334844e-01 -4.58965689e-01
-4.74899381e-01 3.20299864e-01 -1.22845006e+00 8.74675870e-01
-5.84139287e-01 6.09260976e-01 -5.27909458e-01 -7.79937923e-01
1.32716548e+00 6.33516014e-01 6.41720653e-01 -7.76170015e-01
7.02068657e-02 7.36572146e-01 9.93989408e-02 -9.56180990e-01
7.10807368e-02 3.73263091e-01 4.18335684e-02 1.91832989e-01
2.13391513e-01 -2.74373263e-01 3.98122430e-01 -3.44723433e-01
3.37762296e-01 1.46879107e-01 7.03732491e-01 -5.28127700e-02
6.64108753e-01 1.58361211e-01 3.84039670e-01 6.15981042e-01
7.49782473e-02 2.81886697e-01 4.15260315e-01 -1.95997894e-01
-7.05427825e-01 -4.02104735e-01 -5.46069622e-01 8.75596941e-01
-6.02770567e-01 -7.94059455e-01 -6.79060757e-01 -4.67178762e-01
1.01388134e-01 9.40554798e-01 -6.96252406e-01 -1.66281581e-01
-2.30945975e-01 -7.92326510e-01 5.04489601e-01 5.54786325e-01
2.81323910e-01 -1.22384834e+00 -8.34799886e-01 3.04528505e-01
-1.17812036e-02 -5.99718153e-01 3.54407787e-01 1.39882997e-01
-1.25744510e+00 -1.31649184e+00 -1.69055209e-01 -8.44946682e-01
2.53176659e-01 -1.93958655e-01 1.06445622e+00 8.15047175e-02
-4.22144860e-01 4.24009740e-01 -3.56513083e-01 -8.70824218e-01
-4.33333606e-01 2.10221142e-01 -1.97638139e-01 -6.76682234e-01
8.64041388e-01 9.41428617e-02 -2.94936031e-01 3.13409179e-01
-8.10586393e-01 -4.12373453e-01 7.66169012e-01 6.25488818e-01
5.88392913e-01 1.03223093e-01 1.01374161e+00 -9.07863796e-01
1.08430421e+00 -6.66676700e-01 -3.04028898e-01 7.05657005e-01
-1.44905233e+00 -3.62365723e-01 -8.19472820e-02 -3.11099082e-01
-8.95311415e-01 -4.29864556e-01 -2.29334190e-01 1.66984618e-01
-1.99777409e-01 9.01541114e-01 -2.12178379e-01 -1.30448654e-01
8.00695240e-01 -2.09074050e-01 3.38569820e-01 -4.56947953e-01
-1.18269119e-02 1.05364037e+00 -2.36556202e-01 -5.48909962e-01
-1.07208423e-01 -3.38202029e-01 -3.32448572e-01 -9.06238914e-01
-3.86552393e-01 -4.94938821e-01 -7.51344442e-01 -3.18943888e-01
4.48689133e-01 -5.92557609e-01 -4.94344413e-01 3.18275899e-01
-5.90587139e-01 -1.68457344e-01 -3.25873971e-01 7.87059724e-01
-4.68580723e-01 -1.92150369e-01 -2.18971655e-01 -9.88525271e-01
-6.87518716e-01 -1.28861654e+00 4.38082725e-01 3.39986026e-01
-5.30015171e-01 -1.26673079e+00 1.69585258e-01 2.57614881e-01
6.90727293e-01 3.98844630e-01 1.17753005e+00 -8.74887288e-01
2.86767393e-01 -4.39838529e-01 7.01682940e-02 3.17472458e-01
3.25978279e-01 6.65800095e-01 -8.99449229e-01 -2.56062504e-02
1.42189085e-01 -9.98420417e-02 1.26399666e-01 6.03613913e-01
1.00986671e+00 -6.22877896e-01 -4.58874077e-01 1.89904392e-01
1.50626016e+00 9.38367903e-01 4.79791850e-01 6.86372697e-01
3.13221961e-01 1.05694580e+00 1.33201134e+00 3.91979218e-01
2.14794084e-01 6.88702047e-01 -8.30322132e-02 4.55191650e-04
-3.10256612e-04 -2.09777672e-02 1.35830164e-01 8.96677673e-01
-2.49534696e-01 1.04856826e-01 -1.01818156e+00 2.71140933e-01
-1.70711958e+00 -6.55596733e-01 -1.70890197e-01 2.44028950e+00
7.63230443e-01 2.86652416e-01 5.68662226e-01 5.26411355e-01
2.66734242e-01 -5.75106621e-01 -9.44236144e-02 -8.45631719e-01
-7.65989674e-03 3.01175416e-01 2.33023152e-01 4.58752275e-01
-1.06395674e+00 5.09869814e-01 7.39337111e+00 6.93710923e-01
-1.05873704e+00 -1.49171352e-02 3.48132104e-01 1.65578142e-01
-2.93349743e-01 -1.90335110e-01 -1.07953799e+00 2.95003891e-01
1.26655149e+00 -3.36730242e-01 -1.30015612e-01 7.95645773e-01
1.47199094e-01 -3.98466796e-01 -9.95515049e-01 6.12459958e-01
1.60361633e-01 -1.39310193e+00 6.53396472e-02 -9.34234113e-02
6.85322106e-01 -4.24890161e-01 3.32934447e-02 5.03453314e-01
-4.36753869e-01 -9.94708419e-01 5.38036823e-01 6.86084449e-01
8.00330341e-01 -8.59062850e-01 1.16387057e+00 1.60347328e-01
-1.04101515e+00 -4.62981254e-01 -2.29296327e-01 -4.14584950e-02
-4.10939008e-01 4.36133832e-01 -8.68487716e-01 9.22874928e-01
7.01518297e-01 6.31198764e-01 -8.77766192e-01 1.33072364e+00
4.38837074e-02 7.69517839e-01 -1.83123827e-01 -2.29382232e-01
-1.13280080e-01 -1.47467285e-01 7.60058453e-03 1.68104279e+00
1.41485095e-01 -4.01650816e-01 -2.18696848e-01 5.09860933e-01
7.19310343e-01 6.53063357e-01 -7.34517992e-01 -6.08678311e-02
1.01999831e+00 8.69174898e-01 -3.96145731e-01 -2.49073312e-01
-7.33507037e-01 -1.20932221e-01 -2.03286722e-01 3.46269637e-01
-4.77168351e-01 -6.26590967e-01 1.38456404e-01 1.94877684e-01
-1.75910681e-01 1.61370993e-01 -7.59675741e-01 -3.52384269e-01
-2.40357034e-02 -1.18403924e+00 6.84599102e-01 -2.95965642e-01
-1.04041862e+00 4.08262014e-01 7.92553961e-01 -1.37712622e+00
-3.71982872e-01 -4.15720612e-01 -3.56656253e-01 9.38979983e-01
-7.28644609e-01 -9.64305282e-01 -2.13542968e-01 1.26129851e-01
5.90787947e-01 -4.97371554e-01 1.08349597e+00 3.29389066e-01
-6.68546140e-01 6.95052981e-01 2.53597170e-01 -2.85720259e-01
5.60227096e-01 -8.05126071e-01 2.26018275e-03 1.41239047e-01
-4.25942093e-01 8.90099287e-01 3.01548928e-01 -8.43526185e-01
-1.13399482e+00 -6.86844528e-01 1.25942397e+00 -8.28091130e-02
2.17798024e-01 -1.37194410e-01 -8.78514349e-01 5.02631247e-01
4.82135899e-02 -7.84557939e-01 1.21943295e+00 2.80602509e-03
1.77081659e-01 -2.06126630e-01 -1.45603442e+00 1.65726930e-01
2.51718044e-01 -2.22357199e-01 -4.85527277e-01 7.60665834e-02
4.28063482e-01 -1.35227740e-01 -1.41560483e+00 8.09510648e-01
9.43357706e-01 -7.45477378e-01 1.02263880e+00 -4.84727621e-01
1.44730240e-01 1.76911235e-01 -1.66248202e-01 -8.92342508e-01
-1.82768837e-01 1.96891706e-02 -3.00754964e-01 1.34918404e+00
9.09424663e-01 -7.61384308e-01 3.18956941e-01 1.04134285e+00
-8.86510536e-02 -1.29116511e+00 -5.00917256e-01 -5.67086399e-01
2.01136127e-01 -5.23263872e-01 4.44148153e-01 7.48060644e-01
1.70504943e-01 8.73566642e-02 1.18327267e-01 -4.56310749e-01
3.23695481e-01 -3.51166219e-01 6.43608689e-01 -1.32243073e+00
1.67214945e-01 -7.76631534e-01 -2.96705544e-01 -5.71171716e-02
-3.77821743e-01 -6.37021959e-01 -4.14585382e-01 -1.81215787e+00
5.61237969e-02 -8.37281942e-01 -1.26382127e-01 2.62097836e-01
-2.02454582e-01 -3.56946409e-01 5.98679520e-02 4.46815759e-01
2.70331860e-01 -4.41287719e-02 7.64755785e-01 2.32323498e-01
-7.23299980e-01 4.83422101e-01 -6.19436920e-01 4.77995247e-01
1.26974964e+00 -5.81044197e-01 -7.11631000e-01 2.14226753e-01
3.73366505e-01 -1.82952166e-01 -1.53695449e-01 -8.03228736e-01
1.85095876e-01 -4.31103826e-01 4.84540105e-01 -8.72890890e-01
8.37583840e-02 -1.02835441e+00 5.09726167e-01 5.17114460e-01
-3.99560243e-01 2.67344743e-01 5.42141318e-01 -2.38524452e-01
-2.82921314e-01 -8.92815590e-01 4.23161030e-01 6.28967136e-02
-4.70555097e-01 -2.74201423e-01 -5.82597136e-01 -1.66571051e-01
1.20659626e+00 -5.48780739e-01 3.82279628e-03 2.36782227e-02
-7.74541676e-01 1.78383154e-04 1.65014416e-01 4.36599344e-01
7.46548235e-01 -1.17670846e+00 -4.38840747e-01 2.29637280e-01
1.82425246e-01 -2.75703549e-01 -1.59625039e-01 1.05534315e+00
-5.06026328e-01 8.00433636e-01 -3.13106924e-01 -4.13562179e-01
-1.31895661e+00 5.51827610e-01 3.13910276e-01 -6.28304332e-02
-2.28348270e-01 4.42621320e-01 -2.46486723e-01 -6.60104930e-01
8.82271349e-01 -2.66870201e-01 -9.62435782e-01 5.51416337e-01
3.31215680e-01 8.94512892e-01 4.69880044e-01 -3.07712853e-01
-4.17013168e-01 7.14444697e-01 1.89148560e-01 -2.32283682e-01
1.06696129e+00 5.18597141e-02 -3.80190164e-02 1.16593683e+00
1.07559288e+00 -3.03739667e-01 -4.83212799e-01 1.53731585e-01
3.80175620e-01 -2.50009954e-01 -3.49034995e-01 -1.06754160e+00
-7.46454060e-01 7.08474159e-01 7.56274700e-01 2.12173119e-01
1.15659821e+00 -2.17933491e-01 -1.31588995e-01 1.43639401e-01
2.56725289e-02 -1.47226107e+00 -1.00625716e-01 1.33017525e-01
1.21300781e+00 -1.16487265e+00 2.12047368e-01 -4.67039078e-01
-9.07524824e-01 1.38001108e+00 6.58368886e-01 2.67750919e-01
1.17327487e+00 3.25435311e-01 -4.99230511e-02 -3.70842218e-01
-7.45869040e-01 2.38163739e-01 6.80603385e-01 8.20658624e-01
9.43158209e-01 1.71268255e-01 -1.10572588e+00 9.13105249e-01
-1.09385140e-01 2.69868672e-01 1.73537254e-01 1.45959222e+00
-4.26731914e-01 -1.18175983e+00 -4.36676264e-01 8.16902518e-01
-4.82871294e-01 2.33990669e-01 -6.04274631e-01 1.43596053e+00
2.49584943e-01 1.24823713e+00 -1.84498549e-01 -6.71876132e-01
6.28669262e-01 1.98448330e-01 4.38046187e-01 -6.10218287e-01
-1.14043009e+00 3.56918536e-02 3.01733524e-01 -1.95358098e-01
-4.79552597e-01 -9.10768509e-01 -1.01340544e+00 -7.73369148e-02
-5.99459648e-01 4.55781341e-01 1.16656339e+00 8.00023079e-01
2.62466997e-01 5.32258809e-01 5.95290303e-01 -3.23186874e-01
-9.80029926e-02 -1.39820254e+00 3.67481224e-02 -7.06054494e-02
-6.65600896e-02 -7.97618449e-01 -5.17588198e-01 -1.62873983e-01] | [8.43787670135498, 4.626473426818848] |
3c3e8583-e33c-48b0-8d2f-c1c82223aa23 | counter-strike-deathmatch-with-large-scale | 2104.04258 | null | https://arxiv.org/abs/2104.04258v2 | https://arxiv.org/pdf/2104.04258v2.pdf | Counter-Strike Deathmatch with Large-Scale Behavioural Cloning | This paper describes an AI agent that plays the popular first-person-shooter (FPS) video game `Counter-Strike; Global Offensive' (CSGO) from pixel input. The agent, a deep neural network, matches the performance of the medium difficulty built-in AI on the deathmatch game mode, whilst adopting a humanlike play style. Unlike much prior work in games, no API is available for CSGO, so algorithms must train and run in real-time. This limits the quantity of on-policy data that can be generated, precluding many reinforcement learning algorithms. Our solution uses behavioural cloning - training on a large noisy dataset scraped from human play on online servers (4 million frames, comparable in size to ImageNet), and a smaller dataset of high-quality expert demonstrations. This scale is an order of magnitude larger than prior work on imitation learning in FPS games. | ['Jun Zhu', 'Tim Pearce'] | 2021-04-09 | null | null | null | null | ['fps-games'] | ['playing-games'] | [ 1.60008464e-02 2.54123569e-01 -1.58744797e-01 -5.97829930e-03
-7.55230844e-01 -7.50249267e-01 8.14398885e-01 -6.30049109e-01
-9.86978292e-01 9.38044012e-01 7.20115975e-02 -3.79168004e-01
-4.91120759e-03 -5.60308278e-01 -7.80263484e-01 -4.54316884e-01
-3.46575916e-01 7.19385922e-01 7.15104163e-01 -8.12833428e-01
3.73988926e-01 1.36950701e-01 -1.65288913e+00 2.89653659e-01
2.00440452e-01 7.53115833e-01 3.04329962e-01 1.45668566e+00
5.32723129e-01 1.69761121e+00 -1.03704035e+00 -3.02668720e-01
8.20040643e-01 -6.73976183e-01 -9.54089344e-01 -2.05375627e-01
2.15079188e-01 -1.01403773e+00 -7.87186325e-01 7.68517137e-01
7.73616135e-01 2.48538658e-01 2.24763200e-01 -1.65241027e+00
-8.49996358e-02 5.80898166e-01 -8.77694488e-02 5.69502652e-01
6.11342788e-01 1.13860464e+00 7.10394979e-01 7.99644366e-02
9.91554022e-01 1.23345947e+00 7.19368994e-01 7.56778896e-01
-7.68110335e-01 -7.28813410e-01 -2.62566447e-01 1.28369451e-01
-8.00266445e-01 -3.04358572e-01 2.60807246e-01 -3.82433593e-01
1.34312916e+00 -1.33144390e-02 1.32307208e+00 1.86689830e+00
2.04824239e-01 9.31446254e-01 1.16471040e+00 -1.09816901e-01
3.98256958e-01 -4.55343246e-01 -8.86723042e-01 6.07957661e-01
-3.99850458e-01 7.71882415e-01 -5.73034942e-01 -2.67115355e-01
1.52357149e+00 -3.62219155e-01 5.51921912e-02 -2.06103608e-01
-1.31485498e+00 8.18638861e-01 1.44588619e-01 -4.08344790e-02
-4.90974754e-01 9.11686003e-01 7.31508493e-01 9.72147763e-01
-9.24042910e-02 8.05777729e-01 -3.96124274e-01 -1.32250738e+00
-7.17276394e-01 1.13880789e+00 1.05167472e+00 7.87906647e-01
1.82753310e-01 4.96566236e-01 2.82618910e-01 2.66208857e-01
-6.90887198e-02 2.39085242e-01 5.12770295e-01 -2.03062892e+00
3.19373041e-01 -9.98106822e-02 2.68970311e-01 -6.74503803e-01
-3.35130423e-01 2.94487998e-02 -3.08118071e-02 1.25714397e+00
7.77265370e-01 -5.52210152e-01 -6.42817795e-01 1.58252978e+00
1.70589343e-01 4.01489586e-01 9.42843407e-02 1.14028573e+00
5.61828852e-01 6.32479966e-01 2.10336186e-02 2.83391386e-01
9.58570242e-01 -1.12068260e+00 -6.73007742e-02 -1.72952548e-01
5.29073715e-01 -4.34920520e-01 1.11821973e+00 8.04996490e-01
-1.32390141e+00 -3.49412262e-01 -9.13868308e-01 1.25547796e-01
-1.04024149e-01 -6.21954560e-01 8.39190423e-01 4.57900345e-01
-1.12985599e+00 1.01401210e+00 -9.72394168e-01 -4.32747722e-01
2.70919621e-01 3.85023743e-01 -4.08723533e-01 2.99767673e-01
-1.13154554e+00 1.11689651e+00 6.10977411e-01 -4.12619382e-01
-1.74332464e+00 -5.09178817e-01 -5.32499492e-01 -3.31394792e-01
7.74652898e-01 -5.39695680e-01 2.03289866e+00 -1.42949080e+00
-2.19520354e+00 9.50508952e-01 1.00311470e+00 -7.91907430e-01
1.08457065e+00 -2.45238796e-01 -8.14576671e-02 5.18855035e-01
2.02750593e-01 1.02582753e+00 9.50998306e-01 -7.98905671e-01
-8.91142070e-01 1.96355030e-01 8.76732528e-01 7.38162756e-01
4.54968423e-01 4.98911083e-01 -1.69654384e-01 -5.85643589e-01
-8.06079507e-01 -8.42298865e-01 -3.95987719e-01 2.99877860e-02
1.42691776e-01 -1.67288676e-01 7.56834269e-01 -6.38750851e-01
7.19136953e-01 -1.95674038e+00 3.06276560e-01 -1.30500570e-01
3.19723725e-01 1.93636894e-01 -2.85101235e-01 7.51648962e-01
1.50760934e-01 -2.72270799e-01 6.81637973e-02 2.47972980e-01
2.82302618e-01 6.21046007e-01 -1.76315501e-01 4.34407234e-01
-1.08615190e-01 9.35087800e-01 -1.33868337e+00 -5.30681789e-01
6.81515709e-02 -2.47380927e-01 -8.62467408e-01 2.70097136e-01
-4.27862942e-01 4.67453450e-01 -4.75176841e-01 5.47734201e-01
-1.31455749e-01 3.11498493e-01 -8.17630626e-03 5.67605078e-01
-2.78452784e-01 3.96725535e-01 -9.35321152e-01 2.08285618e+00
-1.85782984e-02 6.96116388e-01 4.58116651e-01 -9.63896871e-01
4.51375782e-01 5.17926395e-01 6.36329472e-01 -7.63185740e-01
3.85308713e-01 2.27257937e-01 4.25069153e-01 -6.30311549e-01
6.09190762e-01 -5.09069003e-02 -4.51935261e-01 6.53117299e-01
4.14922595e-01 -7.64044464e-01 5.61152816e-01 2.62957811e-01
1.76464117e+00 8.70570123e-01 4.61776257e-02 -5.18542901e-02
-3.47326159e-01 6.76375329e-01 4.53661054e-01 1.28447950e+00
-7.17435896e-01 4.07230943e-01 8.53300571e-01 -6.04729176e-01
-1.59873116e+00 -9.19491291e-01 5.65530300e-01 1.43415201e+00
2.55566780e-02 -3.82907361e-01 -1.11926270e+00 -4.61134374e-01
-1.01500988e-01 4.44307417e-01 -4.71449405e-01 -3.79941389e-02
-7.95453131e-01 -3.09989657e-02 1.28981006e+00 4.52148110e-01
6.29674852e-01 -1.82620358e+00 -1.37850046e+00 6.38607800e-01
1.47058949e-01 -7.89957285e-01 -1.69959977e-01 5.46255410e-02
-3.39197487e-01 -1.33548808e+00 -4.90276963e-01 -5.81310451e-01
-1.99387461e-01 -1.73678443e-01 1.36781001e+00 9.07927155e-02
-4.14048016e-01 6.00888729e-01 -4.90356535e-01 -5.21831036e-01
-7.85212815e-01 -1.23874247e-01 2.76710629e-01 -9.43976343e-01
1.88585699e-01 -8.21816921e-01 -5.04386067e-01 1.14936784e-01
-7.92440176e-01 3.39881092e-01 4.31824356e-01 1.00086772e+00
-1.68648764e-01 -9.46620293e-03 3.33304793e-01 -2.94347882e-01
9.60935235e-01 -4.28631037e-01 -8.21291745e-01 -4.79697108e-01
-2.31430586e-02 -4.89442587e-01 8.67597222e-01 -7.17850864e-01
-6.66527450e-01 -1.42645091e-01 -2.74678051e-01 -5.46720982e-01
-2.69045413e-01 2.29400042e-02 5.65709054e-01 -1.83038086e-01
1.00021183e+00 1.38768941e-01 4.25145239e-01 -3.66930924e-02
4.84956980e-01 4.97691125e-01 1.00526047e+00 -8.97467732e-01
6.38403058e-01 1.84958622e-01 -3.65219533e-01 -6.33576274e-01
-4.23504189e-02 -1.09370679e-01 -3.22121561e-01 -5.69937229e-01
8.16555023e-01 -9.54053283e-01 -1.41028857e+00 6.95654809e-01
-7.25873113e-01 -1.21461487e+00 -6.74426675e-01 4.92968172e-01
-1.50849915e+00 -2.02042684e-02 -1.08759737e+00 -6.18733108e-01
1.16474628e-02 -1.07131958e+00 7.84797192e-01 2.92805046e-01
-3.11057478e-01 -5.77097774e-01 4.64571178e-01 2.66278356e-01
3.90064895e-01 3.42264712e-01 1.65308386e-01 -4.40478355e-01
-4.71136838e-01 -1.01916559e-01 1.34939209e-01 2.60271490e-01
-5.37601948e-01 -9.64466762e-03 -3.62892717e-01 -3.40462595e-01
-1.98655665e-01 -1.19301760e+00 2.53010213e-01 2.83198565e-01
6.70154154e-01 -3.21531862e-01 9.82948989e-02 4.52055722e-01
1.03014445e+00 3.75635684e-01 8.55683446e-01 9.41573262e-01
3.91723096e-01 2.07371503e-01 7.40943968e-01 6.29120648e-01
2.98183769e-01 5.24062753e-01 6.59344077e-01 -1.03382729e-01
1.49676815e-01 -4.36824888e-01 6.74471557e-01 1.31547198e-01
-7.54056871e-01 2.64085247e-03 -8.22669208e-01 3.98832649e-01
-1.95656335e+00 -1.87072158e+00 4.10404742e-01 1.60612500e+00
8.71167898e-01 4.36438769e-01 9.03237164e-01 -1.77006558e-01
3.19858372e-01 2.02209219e-01 -5.39323866e-01 -7.54277587e-01
1.14678070e-01 5.21474481e-01 7.16142237e-01 3.35134894e-01
-1.06550479e+00 1.33750331e+00 7.47530317e+00 1.00527024e+00
-8.16323340e-01 2.52509844e-02 3.79033536e-01 -7.44309247e-01
2.97415018e-01 7.38821104e-02 -2.39265978e-01 5.71577847e-01
1.00612772e+00 -1.87702134e-01 1.06113732e+00 1.11111736e+00
4.28500712e-01 -4.01594669e-01 -8.50630045e-01 8.57970178e-01
-8.71332511e-02 -1.39369071e+00 -7.45043695e-01 1.85642987e-01
4.97191459e-01 4.62414414e-01 -1.16041020e-01 8.27287078e-01
1.46593678e+00 -1.21416509e+00 1.20405328e+00 2.11315155e-01
7.20605671e-01 -8.71528983e-01 3.42149317e-01 8.03448200e-01
-6.81820810e-01 -3.65989238e-01 -2.64903396e-01 -6.14238143e-01
2.73392379e-01 -7.29810297e-01 -6.19886816e-01 -1.05509676e-01
1.07300794e+00 3.98603737e-01 -1.56556338e-01 8.24626565e-01
-2.56562799e-01 6.40162945e-01 -6.24635935e-01 -2.82487363e-01
1.02056289e+00 -2.19465747e-01 7.32772827e-01 8.88969064e-01
-4.73932317e-03 4.80317146e-01 4.77583259e-01 6.04392767e-01
2.57472068e-01 -4.35396373e-01 -9.41556036e-01 -2.74986565e-01
1.79428205e-01 1.05224609e+00 -6.11636281e-01 -5.87326646e-01
-1.99417204e-01 1.22772300e+00 3.12345535e-01 2.14223154e-02
-1.07086813e+00 -1.86642274e-01 8.17912638e-01 2.77571734e-02
2.87121236e-01 -4.66736078e-01 4.98475850e-01 -9.25991833e-01
-2.74927676e-01 -1.74194169e+00 1.75037637e-01 -9.09340084e-01
-1.04029703e+00 3.63175184e-01 1.63522467e-01 -1.18860424e+00
-1.10202670e+00 -5.16660213e-01 -7.28318930e-01 2.76956320e-01
-5.64672470e-01 -8.64427924e-01 -1.72825868e-03 7.05113053e-01
6.41782045e-01 -4.74888235e-01 7.77029097e-01 1.14297690e-02
-1.60652012e-01 2.68707454e-01 4.73084301e-02 2.21727252e-01
3.88254285e-01 -1.25842249e+00 7.82525003e-01 4.68498558e-01
-2.02024713e-01 -2.64560404e-05 9.19676423e-01 -5.89592040e-01
-1.36690259e+00 -3.24701130e-01 -2.93381035e-01 -5.52513003e-01
1.18185687e+00 -2.43172899e-01 -3.91777873e-01 7.21447349e-01
5.79828799e-01 -3.63767296e-01 1.51690930e-01 -3.67353052e-01
-9.42650661e-02 5.42112291e-01 -1.03171492e+00 1.04743528e+00
1.39305484e+00 -4.35714066e-01 -7.73374498e-01 3.29761475e-01
4.20579582e-01 -8.99478912e-01 -6.76723301e-01 -9.91613567e-02
7.86994755e-01 -1.24919033e+00 9.21940207e-01 -9.93506074e-01
8.27228367e-01 -6.72400594e-02 1.36421353e-01 -1.37839711e+00
-2.39501283e-01 -1.18871140e+00 8.06962848e-02 3.99322271e-01
-1.19220592e-01 -2.91962326e-01 1.08290684e+00 4.16441828e-01
-1.91067651e-01 -6.43009365e-01 -1.01854217e+00 -9.31137979e-01
2.45440096e-01 -4.72262591e-01 3.31816137e-01 8.37797523e-01
2.74279982e-01 5.49290888e-02 -6.93168581e-01 -4.93179023e-01
4.17722285e-01 -5.20969868e-01 1.35547864e+00 -7.22234964e-01
-1.07872045e+00 -6.25818729e-01 -6.78111970e-01 -1.09506857e+00
1.37180015e-01 -4.97420102e-01 1.96729571e-01 -1.09963560e+00
-1.21684767e-01 -2.11431086e-01 2.71795034e-01 6.38398707e-01
2.88420409e-01 3.80573452e-01 4.90634263e-01 2.49569222e-01
-8.23112488e-01 4.57670659e-01 1.27005458e+00 1.28891364e-01
-1.06503777e-01 -2.56113827e-01 -3.24661046e-01 1.05884087e+00
9.38568950e-01 -4.64457482e-01 -3.73386472e-01 -3.05895776e-01
4.70436960e-02 3.99482429e-01 6.05017185e-01 -1.30459750e+00
1.85318202e-01 -5.32285929e-01 9.04126614e-02 -3.28247324e-02
5.22744298e-01 -4.32976604e-01 1.29776195e-01 6.39806747e-01
-2.58542120e-01 3.36880684e-01 2.00565770e-01 3.93602878e-01
8.94017816e-02 -3.22799474e-01 6.02397978e-01 -8.73986185e-01
-1.23867512e+00 6.42291754e-02 -1.10005093e+00 3.29976261e-01
1.21427524e+00 -5.56459665e-01 -3.92663002e-01 -9.40701365e-01
-4.27276850e-01 3.33245724e-01 7.98537552e-01 3.74041170e-01
3.13326389e-01 -1.12275410e+00 -7.27363110e-01 -1.18774459e-01
-2.03771919e-01 -2.23289266e-01 1.08067997e-01 4.23368454e-01
-1.37992442e+00 -2.61193871e-01 -8.08922172e-01 -2.65876532e-01
-1.01858985e+00 3.85719597e-01 5.99604666e-01 -1.74658731e-01
-1.22490811e+00 8.75917017e-01 -1.99797109e-01 -5.48639297e-01
1.33287922e-01 1.84001923e-01 1.65799484e-01 -6.38401210e-01
5.00195682e-01 3.76805693e-01 -5.55486381e-01 -3.76653224e-01
1.09699249e-01 -2.02045724e-01 1.35214822e-02 -8.71949255e-01
1.36641812e+00 4.16493654e-01 2.07878575e-01 2.82357663e-01
6.94462240e-01 -4.61613029e-01 -2.00277400e+00 2.59863555e-01
-3.43716800e-01 -4.64382172e-01 -1.96746141e-01 -5.34166813e-01
-4.93967474e-01 3.84500831e-01 3.13115031e-01 2.06757426e-01
6.59654438e-01 -1.74029663e-01 9.14621949e-01 7.59023190e-01
8.16492140e-01 -1.71970367e+00 4.79243815e-01 7.35788286e-01
9.00553346e-01 -9.64440405e-01 -2.52896488e-01 4.13480937e-01
-9.45676386e-01 9.11655486e-01 8.26136708e-01 -9.01062489e-01
2.95952875e-02 7.20292211e-01 2.93792963e-01 -3.99657965e-01
-1.04912663e+00 -4.12864387e-02 -6.67104959e-01 8.84062409e-01
-1.92841619e-01 -1.11107729e-01 -8.77821818e-03 2.48612210e-01
-8.59618962e-01 2.33269915e-01 7.36614227e-01 1.45131123e+00
-6.15976155e-01 -9.14973378e-01 -3.96588445e-01 2.64069527e-01
-4.95260745e-01 1.22388050e-01 -4.84177381e-01 1.12481272e+00
8.55984762e-02 7.91970611e-01 1.46125734e-01 -3.61631960e-01
2.53648579e-01 -3.10428202e-01 9.14261937e-01 -3.57432276e-01
-1.02650750e+00 -2.29276538e-01 4.30051148e-01 -1.19079864e+00
-1.74294829e-01 -4.14626122e-01 -1.16724420e+00 -1.01402056e+00
2.08595276e-01 -1.30337283e-01 4.59053755e-01 8.32886219e-01
-6.98387027e-02 3.41743648e-01 2.60304034e-01 -1.60085225e+00
-6.83191359e-01 -7.12069571e-01 -6.80141628e-01 5.75741410e-01
-4.49427851e-02 -6.84242904e-01 -1.74636036e-01 -1.08290032e-01] | [3.7074337005615234, 1.5004788637161255] |
bbd06f1a-f3bf-4728-a20c-eec34354a52f | homados-at-semeval-2021-task-6-multi-task | null | null | https://aclanthology.org/2021.semeval-1.141 | https://aclanthology.org/2021.semeval-1.141.pdf | HOMADOS at SemEval-2021 Task 6: Multi-Task Learning for Propaganda Detection | Among the tasks motivated by the proliferation of misinformation, propaganda detection is particularly challenging due to the deficit of fine-grained manual annotations required to train machine learning models. Here we show how data from other related tasks, including credibility assessment, can be leveraged in multi-task learning (MTL) framework to accelerate the training process. To that end, we design a BERT-based model with multiple output layers, train it in several MTL scenarios and perform evaluation against the SemEval gold standard. | ['Piotr Przyby{\\l}a', "Konrad Kaczy{\\'n}ski"] | 2021-08-01 | null | null | null | semeval-2021 | ['propaganda-detection'] | ['natural-language-processing'] | [-2.87089944e-02 2.79720664e-01 -3.19424510e-01 -4.37812477e-01
-1.34118664e+00 -6.78841531e-01 1.16284490e+00 5.54683924e-01
-5.88477969e-01 8.02406013e-01 3.70223701e-01 -7.21646547e-01
1.72589615e-01 -4.67219979e-01 -6.30414665e-01 -3.66452903e-01
1.13763817e-01 4.67774272e-01 1.91784799e-01 -2.46172890e-01
5.49868762e-01 -6.89624846e-02 -6.08733296e-01 5.27681351e-01
1.03149116e+00 8.62166762e-01 -4.60507393e-01 3.70270580e-01
2.09107190e-01 1.24012733e+00 -9.22670543e-01 -9.13525999e-01
-2.61151135e-01 4.89801094e-02 -9.27196562e-01 -6.90087259e-01
2.98516065e-01 -3.25100720e-01 1.49632320e-01 1.08133852e+00
4.74223733e-01 -1.09678335e-01 1.07174671e+00 -9.69910204e-01
-7.97339857e-01 1.23951674e+00 -6.16497815e-01 3.72520745e-01
2.96008021e-01 -9.25946757e-02 8.62707436e-01 -1.07072651e+00
5.82745790e-01 1.52900076e+00 7.07623482e-01 4.67633277e-01
-8.08059275e-01 -6.83958352e-01 -8.52500647e-02 3.95954132e-01
-7.93094516e-01 -3.67272079e-01 6.83603644e-01 -7.16101348e-01
4.82313007e-01 -2.73347318e-01 8.40523243e-02 2.03921294e+00
2.38987148e-01 9.05327737e-01 1.58078134e+00 -3.26368272e-01
1.24846376e-01 1.02174342e-01 3.53916854e-01 7.38533795e-01
2.60595620e-01 1.31210580e-01 -4.20611382e-01 -2.82998860e-01
1.29121214e-01 -5.14855504e-01 1.39867365e-01 5.69719851e-01
-8.56170356e-01 1.27029943e+00 4.08494920e-01 4.85224873e-01
-1.85312286e-01 3.35977256e-01 5.92089713e-01 3.21606874e-01
9.00471985e-01 5.73946297e-01 -2.80926019e-01 -1.51461303e-01
-1.03830230e+00 2.94624805e-01 7.64398396e-01 2.54075199e-01
2.47559756e-01 6.35623261e-02 -5.54531515e-01 5.69996774e-01
3.94869119e-01 4.39332575e-01 1.41137913e-01 -6.92715764e-01
7.72970438e-01 2.92663902e-01 1.29212856e-01 -9.44225430e-01
-5.50095141e-01 -6.25643969e-01 -7.19754338e-01 2.82998502e-01
5.41825354e-01 -5.49035311e-01 -8.04825425e-01 1.53546214e+00
2.07188204e-01 5.03900945e-02 -7.83897936e-02 7.22545922e-01
7.70737588e-01 6.97448015e-01 4.16301847e-01 6.94057252e-03
1.29699373e+00 -9.87954140e-01 -7.35641897e-01 -4.27434534e-01
8.44925523e-01 -5.24556577e-01 9.06036258e-01 5.27316511e-01
-8.00623715e-01 -7.96105042e-02 -1.14353132e+00 -1.75704062e-02
-4.19662178e-01 -3.45872715e-02 3.58023643e-01 4.64418113e-01
-4.79421347e-01 4.42234188e-01 -5.70901692e-01 3.33077341e-01
8.51245403e-01 -5.70841491e-01 -1.79091632e-01 1.04358017e-01
-1.82476366e+00 1.72203457e+00 6.14749134e-01 2.33632907e-01
-1.69852030e+00 -4.51487213e-01 -7.85481870e-01 2.01665796e-02
4.64722842e-01 -2.49252290e-01 1.39370131e+00 -7.12820590e-01
-1.09893811e+00 8.25344503e-01 5.96498609e-01 -6.10192716e-01
7.56861687e-01 -3.57822597e-01 -3.57144445e-01 -3.52931619e-02
2.80304253e-01 4.30505536e-02 1.22716773e+00 -1.24313569e+00
-4.68205988e-01 -7.57233426e-02 2.70277169e-02 -2.55235314e-01
-3.80725563e-01 7.49792814e-01 4.85946506e-01 -6.90850854e-01
-6.89370990e-01 -4.83796686e-01 -2.53846198e-01 -5.15460312e-01
-4.99755144e-01 -6.09959245e-01 9.27285314e-01 -8.95150304e-01
1.36221159e+00 -1.86394262e+00 2.80690920e-02 1.75705850e-01
5.40778399e-01 6.18705809e-01 -5.61402515e-02 2.55464613e-01
3.96004587e-01 6.25131011e-01 1.07133230e-02 -2.68678665e-01
-3.29362601e-02 -1.27850592e-01 -3.15179437e-01 3.85946870e-01
4.17735189e-01 9.83033538e-01 -1.25224721e+00 -6.78773701e-01
-2.94666499e-01 2.95414835e-01 -5.49831130e-02 5.65085560e-02
-4.33042496e-01 6.04982913e-01 -8.40697825e-01 5.80924988e-01
3.40148717e-01 -5.12011588e-01 -6.66625500e-02 -1.21853994e-02
-4.60755043e-02 6.34756744e-01 -4.80323851e-01 1.34392297e+00
-6.46909297e-01 6.12145841e-01 1.83280513e-01 -9.84900773e-01
5.71460485e-01 6.83453202e-01 -1.45888671e-01 -5.38998604e-01
6.48948371e-01 3.47355157e-01 -7.64987618e-02 -5.76077759e-01
1.57349244e-01 -2.76885837e-01 -5.34706950e-01 1.05073905e+00
1.81278944e-01 -8.76775086e-02 -6.31828010e-02 4.81094629e-01
9.86171186e-01 -1.72359496e-01 3.73911113e-01 -1.23130217e-01
3.62907857e-01 7.92828426e-02 3.82881463e-01 7.63654292e-01
-2.45043993e-01 -3.23522240e-02 3.96206588e-01 -5.27530193e-01
-1.06980455e+00 -8.76598656e-01 -2.89564669e-01 1.42163277e+00
-3.11211854e-01 -3.48470271e-01 -6.19114339e-01 -1.14651883e+00
-2.07841113e-01 9.23959494e-01 -8.47964346e-01 -7.06127957e-02
-5.43883979e-01 -8.98392081e-01 1.11400032e+00 3.69463176e-01
4.71094906e-01 -9.27739024e-01 -6.73858047e-01 3.81052136e-01
-3.32912266e-01 -1.13544095e+00 -1.95560548e-02 9.21065137e-02
-4.54840809e-01 -1.16030431e+00 -3.83212358e-01 -4.92771298e-01
2.08599269e-01 -1.49037540e-01 1.11604154e+00 3.02800596e-01
1.93747729e-01 -1.97929084e-01 -6.03141069e-01 -6.99601293e-01
-1.12562418e+00 1.96022257e-01 -9.11575481e-02 -2.95007735e-01
-4.79626991e-02 -3.58042508e-01 -1.32275194e-01 -3.05997618e-02
-9.74659383e-01 1.13204129e-01 5.39885223e-01 9.85249996e-01
-1.65276557e-01 -2.46109113e-01 1.14936876e+00 -1.11803246e+00
1.19023407e+00 -7.53574669e-01 -5.42620420e-01 5.57700038e-01
-7.28870213e-01 1.68873444e-01 5.50147295e-01 -3.58738214e-01
-1.15608287e+00 -6.74150169e-01 -7.88396597e-02 -4.43711244e-02
3.15595329e-01 1.27241457e+00 3.99111480e-01 1.91768240e-02
1.20992899e+00 -3.61592382e-01 -2.73861915e-01 -5.30159593e-01
7.99320757e-01 8.14637959e-01 5.58292747e-01 -5.25973380e-01
1.09131718e+00 5.56984395e-02 -2.81351507e-01 4.05598693e-02
-1.84979331e+00 8.05462822e-02 -5.55972815e-01 -3.09547782e-01
8.27661693e-01 -8.87521565e-01 -1.87419593e-01 3.42905909e-01
-1.60816288e+00 -2.90288210e-01 3.79597247e-01 3.26642871e-01
-1.72115564e-01 1.60191268e-01 -8.55050325e-01 -1.06015933e+00
-8.27316999e-01 -7.80124366e-01 7.80874908e-01 -1.79530978e-02
2.78412849e-02 -1.33856642e+00 2.49225140e-01 5.46946585e-01
5.35734296e-01 4.45514649e-01 1.17866981e+00 -1.19838595e+00
1.03907362e-01 -3.61978590e-01 -3.56509060e-01 3.80679816e-01
-1.35337487e-01 6.92317337e-02 -1.08411491e+00 -2.25373387e-01
1.32299483e-01 -1.09758842e+00 1.09199309e+00 2.79730290e-01
8.44658136e-01 -7.84750938e-01 -3.99135649e-01 -5.99520467e-02
9.52993333e-01 -4.10242110e-01 1.46258533e-01 6.47420526e-01
4.84556437e-01 5.63546836e-01 5.93438506e-01 3.00445378e-01
6.30007625e-01 4.61917400e-01 4.81575310e-01 -6.52690008e-02
1.48886487e-01 -3.45783621e-01 4.94357675e-01 5.52737057e-01
-1.12521723e-01 -2.59046167e-01 -1.32445705e+00 3.18719566e-01
-1.83660972e+00 -1.26066375e+00 -1.65710509e-01 1.61596072e+00
1.31184721e+00 3.25696319e-01 -3.28931846e-02 1.50240716e-02
7.10656583e-01 3.58361959e-01 -3.81830670e-02 -6.51555002e-01
-2.34433234e-01 -5.05883247e-02 1.52502701e-01 6.78402126e-01
-1.56881046e+00 1.18034482e+00 7.43660593e+00 1.21703255e+00
-9.38594878e-01 8.14076841e-01 9.12175357e-01 2.38110900e-01
-3.31042349e-01 1.24928476e-02 -6.43510342e-01 3.76923919e-01
1.22224677e+00 -3.51743817e-01 -1.03259999e-02 6.77868366e-01
3.17566603e-01 7.63147995e-02 -8.18906009e-01 4.04872119e-01
4.49037626e-02 -1.47383308e+00 7.13637425e-03 -2.81605452e-01
7.73712933e-01 4.22090888e-01 6.94989488e-02 5.70834875e-01
8.54597330e-01 -1.34935856e+00 8.96777153e-01 1.48875983e-02
7.56157339e-01 -1.38180330e-01 8.95673037e-01 7.60510087e-01
-5.41625738e-01 -3.45825404e-01 -5.48805855e-02 -7.24965706e-02
6.31959915e-01 8.44718575e-01 -1.13862062e+00 5.07850945e-01
2.53232509e-01 5.14999270e-01 -7.05036283e-01 8.42182577e-01
-8.80164802e-01 1.14393592e+00 -1.07114585e-02 1.69264823e-02
7.06671298e-01 3.41595918e-01 4.27334964e-01 1.46415067e+00
1.07812084e-01 -1.72269493e-01 3.03807706e-01 9.06168044e-01
-4.80829388e-01 -8.35875571e-02 -4.70559031e-01 -2.21303672e-01
6.23477459e-01 1.57490873e+00 4.04953631e-03 -4.29449648e-01
-1.94341525e-01 3.62467676e-01 6.38862550e-01 2.02096975e-03
-1.13433909e+00 1.82433933e-01 -3.45012635e-01 -1.33452639e-01
-3.37187707e-01 -2.26449445e-01 -2.49750584e-01 -1.09116185e+00
-3.18781525e-01 -8.95181000e-01 5.42430997e-01 -4.86085296e-01
-1.65250647e+00 7.79692173e-01 2.15787560e-01 -7.62084365e-01
-6.31431997e-01 -4.92409408e-01 -1.01122570e+00 7.19023764e-01
-1.71494246e+00 -1.67192757e+00 3.76528949e-02 1.86110452e-01
4.46541637e-01 -3.09221119e-01 6.19629443e-01 2.91114658e-01
-4.70049173e-01 5.09399474e-01 -1.24155253e-01 4.49202031e-01
8.89534354e-01 -1.14467847e+00 4.49380487e-01 8.22461665e-01
1.99871920e-02 1.40192762e-01 6.99115872e-01 -7.79743791e-01
-6.82048082e-01 -1.08455229e+00 1.06041229e+00 -9.46416974e-01
1.36718011e+00 -3.71950597e-01 -1.03784525e+00 6.65203691e-01
1.69479147e-01 -2.49493867e-01 6.29909277e-01 3.78887802e-01
-1.14577329e+00 5.20448327e-01 -1.08302283e+00 1.02168676e-02
5.22793889e-01 -6.43780470e-01 -8.13969374e-01 5.17747521e-01
5.78150749e-01 -2.67808199e-01 -8.89653087e-01 2.67629236e-01
5.34264982e-01 -2.60535955e-01 6.47854626e-01 -1.07635045e+00
9.54644799e-01 1.55975953e-01 1.47418395e-01 -1.62845564e+00
-2.92610288e-01 -4.79995281e-01 -1.98768109e-01 1.26202977e+00
1.08305478e+00 -2.80264795e-01 1.40570775e-01 2.64708936e-01
-2.72245128e-02 -5.79431236e-01 -1.32892287e+00 -6.20295644e-01
6.93575680e-01 -2.88296402e-01 7.72986785e-02 1.14117956e+00
3.06560308e-01 8.23841512e-01 -8.36596489e-01 3.48889157e-02
6.28430307e-01 -8.66937712e-02 1.67313024e-01 -1.49139321e+00
-6.78344071e-02 -4.69957769e-01 1.81067407e-01 -4.94042397e-01
3.90424848e-01 -1.00524592e+00 1.14791855e-01 -1.46588290e+00
4.42115963e-01 -6.04356349e-01 -2.36665860e-01 7.58293450e-01
-4.65969950e-01 9.31184962e-02 3.26564848e-01 2.56892294e-01
-8.66979301e-01 4.98694599e-01 1.09389317e+00 -3.93190354e-01
3.15724045e-01 -1.32884145e-01 -7.36634672e-01 8.35643530e-01
7.56853223e-01 -8.44251335e-01 1.49840578e-01 -9.10908580e-01
7.63963521e-01 4.83371913e-02 4.63878453e-01 -6.79698110e-01
2.11208567e-01 -2.97300607e-01 1.50357142e-01 -3.59891325e-01
1.77925587e-01 -3.14157397e-01 -3.30309302e-01 4.21774864e-01
-6.14215016e-01 1.31541910e-02 -3.30800153e-02 6.13135695e-01
-1.10330924e-01 -4.58448946e-01 6.44848049e-01 1.67121203e-03
-5.49131706e-02 2.04065219e-01 -5.57172120e-01 2.46662810e-01
6.52915299e-01 4.80013013e-01 -1.13845992e+00 -3.65907282e-01
-5.12090862e-01 4.15715456e-01 6.25403272e-03 3.34615439e-01
4.39546734e-01 -1.30085409e+00 -1.46751583e+00 -5.90815723e-01
1.16223492e-01 -2.18096703e-01 -1.47624984e-01 1.00067043e+00
-5.15559874e-02 2.43039593e-01 -6.24124967e-02 -9.65042263e-02
-8.88903081e-01 4.48664069e-01 2.23665148e-01 -7.74787009e-01
-3.24492246e-01 6.41472399e-01 -2.62167305e-01 -1.44641712e-01
5.16813472e-02 3.28281745e-02 -6.30905271e-01 2.16428280e-01
7.49840140e-01 3.82112265e-01 -8.92474800e-02 -3.91118526e-01
-2.59867847e-01 2.81782672e-02 -2.43553773e-01 -3.13897341e-01
1.50976706e+00 2.64277518e-01 -1.03149757e-01 3.28861296e-01
7.92666137e-01 9.46481228e-02 -9.90188479e-01 -3.99787933e-01
6.30103469e-01 -2.41353750e-01 4.14589584e-01 -1.50605905e+00
-6.95488691e-01 1.10401273e+00 2.27652341e-01 2.55859315e-01
5.20683587e-01 2.29199871e-01 6.20679021e-01 4.20842826e-01
3.87197018e-01 -1.35075688e+00 4.51191723e-01 7.20098078e-01
1.10062003e+00 -1.46301389e+00 -7.14904442e-02 -9.04212669e-02
-7.43440390e-01 1.01482749e+00 3.50247860e-01 2.17114594e-02
5.76661229e-01 2.88208127e-01 1.12893328e-01 -5.49262404e-01
-9.78142679e-01 2.26583138e-01 3.43652129e-01 4.41279739e-01
5.96756518e-01 -1.01744831e-02 -6.04738951e-01 8.10582638e-01
1.58533111e-01 -1.30449846e-01 5.98386168e-01 4.70338345e-01
-7.39575148e-01 -9.86683607e-01 -2.52226174e-01 4.10683900e-01
-9.45377767e-01 -4.18301314e-01 -5.00967920e-01 2.72467524e-01
-1.96418434e-01 1.25423932e+00 -6.96443856e-01 -3.64986449e-01
-2.24210873e-01 -9.58238542e-02 2.32265145e-01 -5.88287592e-01
-1.08954573e+00 -1.63981169e-01 7.19870448e-01 -2.04878047e-01
-5.21337509e-01 -3.11928630e-01 -8.75317216e-01 -2.76274860e-01
-5.57779849e-01 4.16994363e-01 8.05421591e-01 1.24817801e+00
-3.43497260e-03 1.81051120e-01 6.31564319e-01 -5.71308374e-01
-1.16097987e+00 -1.50684679e+00 3.13479006e-02 4.04610693e-01
2.55143195e-01 -8.12821448e-01 -5.87983429e-01 -2.24654853e-01] | [8.464176177978516, 10.657181739807129] |
30c6f112-01eb-41d6-98c3-533e59e74aa3 | do-all-roads-lead-to-rome-understanding-the | 2002.12867 | null | https://arxiv.org/abs/2002.12867v1 | https://arxiv.org/pdf/2002.12867v1.pdf | Do all Roads Lead to Rome? Understanding the Role of Initialization in Iterative Back-Translation | Back-translation provides a simple yet effective approach to exploit monolingual corpora in Neural Machine Translation (NMT). Its iterative variant, where two opposite NMT models are jointly trained by alternately using a synthetic parallel corpus generated by the reverse model, plays a central role in unsupervised machine translation. In order to start producing sound translations and provide a meaningful training signal to each other, existing approaches rely on either a separate machine translation system to warm up the iterative procedure, or some form of pre-training to initialize the weights of the model. In this paper, we analyze the role that such initialization plays in iterative back-translation. Is the behavior of the final system heavily dependent on it? Or does iterative back-translation converge to a similar solution given any reasonable initialization? Through a series of empirical experiments over a diverse set of warmup systems, we show that, although the quality of the initial system does affect final performance, its effect is relatively small, as iterative back-translation has a strong tendency to convergence to a similar solution. As such, the margin of improvement left for the initialization method is narrow, suggesting that future research should focus more on improving the iterative mechanism itself. | ['Mikel Artetxe', 'Noe Casas', 'Gorka Labaka', 'Eneko Agirre'] | 2020-02-28 | null | null | null | null | ['unsupervised-machine-translation'] | ['natural-language-processing'] | [ 3.79446089e-01 1.30047485e-01 -2.85407186e-01 -5.37750959e-01
-1.05679321e+00 -7.64132857e-01 1.03621233e+00 -2.29351804e-01
-4.13358986e-01 8.18251550e-01 3.36593866e-01 -8.42824578e-01
4.46968645e-01 -3.59624803e-01 -9.77288246e-01 -6.10498250e-01
5.35809278e-01 8.39809000e-01 -1.75342441e-01 -4.33108807e-01
1.01430573e-01 1.17424041e-01 -8.25724602e-01 2.00712979e-01
8.18608940e-01 1.65998638e-01 1.50343493e-01 5.40892959e-01
-1.56731114e-01 5.22958815e-01 -7.44818926e-01 -7.15922236e-01
5.47604263e-01 -1.18184447e+00 -1.02205670e+00 2.56556630e-01
2.05015868e-01 -1.41156539e-01 1.05352163e-01 1.00698686e+00
5.23837984e-01 -1.07866719e-01 6.28139138e-01 -7.97721863e-01
-6.05162442e-01 1.08048332e+00 -4.82275814e-01 1.41020834e-01
7.77856037e-02 3.08113366e-01 9.59154487e-01 -1.04372859e+00
6.44045651e-01 1.11632299e+00 7.66046107e-01 4.75539833e-01
-1.72438252e+00 -5.92512488e-01 -1.12065725e-01 -5.93601406e-01
-1.21830654e+00 -7.67288923e-01 5.78652501e-01 -2.10424706e-01
8.31970155e-01 1.94737375e-01 5.36449015e-01 1.19685245e+00
2.44089216e-01 4.09361094e-01 1.43206406e+00 -8.79992664e-01
1.91149791e-03 3.85112703e-01 -2.24799048e-02 2.98090309e-01
2.71229684e-01 1.26015067e-01 -4.23866540e-01 -3.43635976e-01
8.17438781e-01 -6.82504058e-01 -1.97553426e-01 -1.50434315e-01
-1.33641469e+00 8.27878177e-01 2.70046651e-01 5.86851954e-01
-3.28358889e-01 1.12043969e-01 3.19184124e-01 8.20840299e-01
5.09039998e-01 9.75633502e-01 -4.03137922e-01 -2.16198578e-01
-1.22173953e+00 1.20117478e-01 8.78503978e-01 6.63352489e-01
7.53026128e-01 6.69102594e-02 3.96697596e-02 8.17333698e-01
4.92243581e-02 3.84608507e-01 6.94532692e-01 -7.90359497e-01
8.18322062e-01 1.81033909e-01 2.47595266e-01 -5.01478016e-01
-3.88850532e-02 -5.01834869e-01 -1.81563035e-01 -5.92041425e-02
9.74453330e-01 -5.96480966e-01 -9.78504598e-01 2.00904775e+00
3.20573039e-02 -2.05412343e-01 2.11822122e-01 9.02509391e-01
1.29221544e-01 8.09023738e-01 -1.07783295e-01 -3.58364910e-01
8.79562974e-01 -1.02493596e+00 -5.44732213e-01 -4.65278596e-01
7.89331377e-01 -1.45098400e+00 1.32854843e+00 6.77296370e-02
-1.43570077e+00 -3.81872803e-01 -1.09737384e+00 3.71708050e-02
2.11364571e-02 1.75412253e-01 4.83907163e-01 5.59768558e-01
-1.03290915e+00 8.10578465e-01 -9.67588425e-01 -6.11901343e-01
-5.74733503e-02 5.19670784e-01 -2.45596498e-01 3.44219059e-02
-1.14232492e+00 1.41620588e+00 2.56525397e-01 1.15849666e-01
-4.30951416e-01 -1.78506210e-01 -4.17748153e-01 -2.76059061e-01
-3.41209918e-02 -8.55798483e-01 1.54064715e+00 -1.85308802e+00
-1.84598136e+00 7.99999356e-01 -4.31449771e-01 -2.38713235e-01
6.39228702e-01 -8.98772329e-02 -3.81901674e-02 -2.72797912e-01
5.33958077e-02 6.68612719e-01 8.52454364e-01 -1.35796535e+00
-2.67315686e-01 -1.75891072e-01 -1.81341022e-01 5.34335971e-01
-2.40666926e-01 5.08529305e-01 -4.81762290e-01 -6.45129144e-01
2.27869421e-01 -1.41870272e+00 -3.05897236e-01 -6.53689861e-01
-1.87042758e-01 4.18718494e-02 2.83241093e-01 -5.92408776e-01
9.65482533e-01 -1.83071578e+00 3.89948934e-01 3.04131448e-01
-1.33011624e-01 1.53939083e-01 -3.13702494e-01 6.59037232e-01
-1.30983949e-01 2.33945966e-01 -3.36438835e-01 -3.71109843e-01
-2.35791087e-01 3.25426489e-01 -3.51224750e-01 3.46922159e-01
3.30250263e-01 1.11089814e+00 -8.33680868e-01 -3.23964685e-01
-3.46642971e-01 2.65446991e-01 -3.88995528e-01 1.92962244e-01
-9.58758742e-02 5.64597964e-01 -1.90550759e-01 2.19539732e-01
3.20848078e-01 -2.36485258e-01 5.51721215e-01 3.11252594e-01
-1.23925090e-01 7.38786459e-01 -8.42499018e-01 1.57651174e+00
-3.74366343e-01 7.35245585e-01 -5.84301613e-02 -9.12201941e-01
8.35517585e-01 5.80710530e-01 9.21362564e-02 -6.31780267e-01
2.53125370e-01 6.71260178e-01 6.01227582e-01 -1.93355694e-01
6.44361556e-01 -5.19983470e-01 2.34827116e-01 1.03464007e+00
-5.91473691e-02 -3.29012066e-01 1.86152771e-01 7.71681219e-03
7.78981149e-01 3.74017149e-01 -4.10006307e-02 -3.77901584e-01
1.42624840e-01 4.52450097e-01 5.32630980e-01 6.66460037e-01
1.66000158e-01 8.73333752e-01 2.66262978e-01 -3.38200927e-01
-1.54564703e+00 -8.26062500e-01 -9.15404558e-02 9.50087070e-01
-1.68344766e-01 -2.88471073e-01 -1.04604352e+00 -6.13154292e-01
-4.31554139e-01 8.81598234e-01 -3.42577487e-01 -1.00677721e-01
-1.04210937e+00 -1.15261626e+00 7.50370324e-01 3.85980844e-01
1.23843968e-01 -9.33304787e-01 -3.52808923e-01 4.14527953e-01
-5.11142492e-01 -8.56810749e-01 -7.16876864e-01 4.48270619e-01
-1.33074367e+00 -4.53458965e-01 -9.48956251e-01 -9.07684684e-01
1.01154852e+00 3.44050407e-01 1.38812745e+00 1.64555743e-01
6.45334840e-01 -1.86140016e-01 -1.58666953e-01 -3.61379743e-01
-1.11748147e+00 4.30375755e-01 1.73441932e-01 -1.59172386e-01
3.58563989e-01 -7.03166902e-01 -1.30869418e-01 4.56941038e-01
-7.13587523e-01 3.15002114e-01 8.96248937e-01 1.01164615e+00
1.93233237e-01 -3.45910966e-01 4.96245861e-01 -8.98649812e-01
1.18840420e+00 -2.88454771e-01 -3.38486075e-01 2.25221306e-01
-7.92607009e-01 3.45816791e-01 7.11126149e-01 -8.38133454e-01
-8.73964071e-01 3.63535546e-02 -4.87149581e-02 -3.69319171e-01
1.93522140e-01 6.23188078e-01 8.07241052e-02 9.06624049e-02
1.04639447e+00 3.56052309e-01 3.00642133e-01 -4.18946266e-01
4.22199696e-01 5.63932836e-01 3.77870947e-01 -8.35706234e-01
1.21379840e+00 -6.58485070e-02 -5.04421115e-01 -4.67197865e-01
-4.93646055e-01 1.03072159e-01 -6.12908900e-01 4.25372161e-02
5.94878495e-01 -9.57055271e-01 3.53268981e-01 2.89393932e-01
-1.16835499e+00 -8.03988159e-01 -1.64935783e-01 6.88381791e-01
-4.84125346e-01 1.21477619e-01 -7.86645174e-01 -4.11498845e-01
-3.10651571e-01 -1.45823789e+00 6.97443068e-01 -2.33429316e-02
-7.68885016e-01 -9.52336967e-01 2.83662289e-01 4.46439624e-01
5.55983603e-01 -2.18044952e-01 9.68191087e-01 -7.68456697e-01
-2.93663621e-01 -1.82146113e-02 1.08964756e-01 3.60419393e-01
2.65295148e-01 5.23342155e-02 -6.39356017e-01 -3.20134342e-01
2.24080592e-01 -4.19315726e-01 4.64812666e-01 1.86493784e-01
1.00184873e-01 -3.68857563e-01 -9.18535218e-02 3.28864068e-01
1.11838949e+00 5.41434810e-02 4.82795656e-01 4.29220527e-01
4.82260823e-01 5.26051462e-01 3.19533736e-01 -3.48797768e-01
2.11033717e-01 7.16537714e-01 -2.71340251e-01 -3.93527985e-01
-9.36851501e-02 -3.72765273e-01 6.75986648e-01 1.26004374e+00
-1.83086976e-01 2.05093529e-03 -9.38170075e-01 4.37296569e-01
-1.74809229e+00 -8.64225686e-01 -2.89575756e-02 2.32107472e+00
1.32390547e+00 3.79502892e-01 2.74380058e-01 -6.85731843e-02
6.27294958e-01 -5.17841689e-02 -1.60542086e-01 -8.94971192e-01
-1.07595988e-01 2.79756248e-01 4.86403108e-01 5.26911497e-01
-4.61691350e-01 1.29484165e+00 7.32288694e+00 4.97309387e-01
-1.58108389e+00 1.84559420e-01 6.35401368e-01 -2.43700277e-02
-4.33846623e-01 3.95274371e-01 -4.39549297e-01 4.23184812e-01
1.25064552e+00 -1.41980916e-01 7.11518228e-01 3.61435711e-01
5.05467117e-01 2.90422607e-02 -1.15480101e+00 3.26361269e-01
-1.52955309e-01 -1.15843236e+00 2.96013318e-02 3.12241822e-01
9.70391035e-01 2.98495233e-01 -3.52386385e-02 2.11504444e-01
6.21524692e-01 -8.08140457e-01 9.26332474e-01 -1.96780469e-02
6.09364331e-01 -6.70117795e-01 5.18036664e-01 5.71321011e-01
-4.85898376e-01 2.19699889e-01 -2.90900379e-01 -3.16676855e-01
1.33423924e-01 3.33420157e-01 -1.01756215e+00 4.27049011e-01
6.14740886e-02 3.60691175e-02 -4.01156515e-01 5.48688114e-01
-4.37009752e-01 1.19732106e+00 -3.24339807e-01 1.98395967e-01
4.66652840e-01 -4.02383119e-01 4.74947035e-01 1.18049276e+00
1.58179849e-01 -1.24259464e-01 1.01788566e-01 7.79457808e-01
-2.66185910e-01 3.53235841e-01 -6.47715986e-01 -4.10543919e-01
2.88787067e-01 9.68312621e-01 -8.32043946e-01 -4.27675337e-01
-5.35496414e-01 1.03388524e+00 4.26813245e-01 5.78400493e-01
-8.32783937e-01 -4.92904149e-02 4.63356346e-01 1.27167806e-01
9.11024511e-02 -4.72483128e-01 -7.17530370e-01 -1.25201440e+00
9.66004729e-02 -1.49589741e+00 -2.14307606e-01 -6.90965533e-01
-8.58657479e-01 8.43592942e-01 -2.52187967e-01 -1.05137956e+00
-7.44097769e-01 2.37380713e-03 -7.32918441e-01 1.39195931e+00
-1.06575930e+00 -9.92076755e-01 4.31512743e-01 2.29036540e-01
6.64401650e-01 -6.13015369e-02 7.26499677e-01 1.32207751e-01
-6.44533157e-01 8.44119251e-01 1.80202574e-01 1.34294286e-01
1.10740602e+00 -1.03905964e+00 8.64241660e-01 1.24093151e+00
6.34939373e-01 1.21734905e+00 1.03116286e+00 -6.96495712e-01
-1.35668814e+00 -8.31260800e-01 1.24951577e+00 -8.28711569e-01
6.09748483e-01 -3.06435496e-01 -8.48808110e-01 9.24328446e-01
5.59710979e-01 -5.84120274e-01 4.41543311e-01 2.60793537e-01
-1.92447945e-01 2.98664242e-01 -7.84082234e-01 8.64192545e-01
7.02383101e-01 -4.47095782e-01 -8.95735800e-01 2.33046591e-01
7.16778040e-01 -4.88952696e-01 -6.86139405e-01 9.78982300e-02
5.32056451e-01 -5.53538263e-01 5.66768825e-01 -7.03866422e-01
7.25329161e-01 -2.23743677e-01 -3.10673974e-02 -1.53427160e+00
-3.91976118e-01 -9.34773445e-01 2.62407780e-01 1.19532979e+00
1.08988678e+00 -7.30466723e-01 7.19940186e-01 6.10546350e-01
-3.93641032e-02 -7.81529009e-01 -6.08123541e-01 -7.34648705e-01
6.29871130e-01 -3.33049625e-01 5.01482546e-01 1.14609945e+00
-8.45272001e-03 8.84435058e-01 -3.56220573e-01 -6.87449649e-02
2.54687101e-01 7.12183351e-03 1.13987434e+00 -7.55610645e-01
-5.14204443e-01 -4.35569048e-01 3.46251935e-01 -1.02392554e+00
2.91745402e-02 -1.07619774e+00 7.62262419e-02 -1.23617435e+00
2.59655297e-01 -4.76132810e-01 -4.30489406e-02 4.18255091e-01
-3.80010277e-01 4.89097714e-01 2.90473193e-01 7.77051032e-01
1.62371650e-01 1.11219294e-01 1.27010643e+00 1.63504407e-01
-6.46458745e-01 1.55178905e-01 -1.01843405e+00 4.79493469e-01
8.64023268e-01 -6.66480839e-01 -4.75404382e-01 -8.70784163e-01
1.31193817e-01 4.26356383e-02 -2.00102344e-01 -5.79531372e-01
1.27664646e-02 -2.67310798e-01 2.97780693e-01 4.71429601e-02
1.38984352e-01 -4.38135624e-01 2.94913083e-01 3.67366284e-01
-5.00890374e-01 5.80654442e-01 2.05795512e-01 -1.64267663e-02
-6.15413859e-02 -3.55316222e-01 8.09683383e-01 -2.29254156e-01
6.04397990e-02 -2.39616871e-01 -7.30099380e-01 7.86299482e-02
3.60305369e-01 -3.92979473e-01 1.81723848e-01 -3.74102205e-01
-3.91012788e-01 -3.67031768e-02 9.12547290e-01 4.43897396e-01
-1.23283088e-01 -1.30368912e+00 -8.87701035e-01 6.82147592e-02
-1.51483327e-01 -3.80918235e-01 -7.72926152e-01 1.05396402e+00
-2.74373353e-01 3.66851538e-01 -2.86764503e-02 -4.80619103e-01
-9.27699327e-01 1.42716005e-01 3.90742600e-01 -2.98190415e-01
-6.11326694e-01 7.02084541e-01 -6.11556433e-02 -5.52854002e-01
-9.88265425e-02 -2.06355006e-01 3.75500321e-01 -1.36236235e-01
1.94170147e-01 -1.89296529e-01 8.96750391e-02 -8.02692950e-01
3.41785257e-03 2.50732064e-01 -2.55059779e-01 -8.62829566e-01
1.17874634e+00 -7.75249377e-02 -1.28910363e-01 5.45503438e-01
8.81332457e-01 2.51115263e-01 -1.03214836e+00 -3.45940858e-01
1.55902300e-02 -2.20431298e-01 -1.77939996e-01 -8.62735391e-01
-7.70865381e-01 5.81650913e-01 1.69018462e-01 -1.71513155e-01
8.00365388e-01 -3.11325729e-01 7.96203315e-01 3.97252232e-01
4.20812786e-01 -1.03274608e+00 -2.13252217e-01 4.95170653e-01
7.24886477e-01 -1.16276956e+00 -3.15548815e-02 6.16923198e-02
-5.77615798e-01 1.05593073e+00 3.33857328e-01 1.12009852e-03
2.80704862e-03 3.33416522e-01 6.75692916e-01 1.68160409e-01
-7.74980426e-01 9.62570235e-02 7.75941610e-02 1.55178651e-01
8.67040694e-01 1.70690231e-02 -7.67151654e-01 8.91565904e-02
-6.54124677e-01 -1.64203197e-02 5.84896207e-01 1.05257988e+00
-2.42419466e-01 -1.67723989e+00 -6.45956874e-01 1.81592718e-01
-6.58086479e-01 -3.20688337e-01 -9.09044445e-01 8.34848583e-01
-1.69312969e-01 1.00334740e+00 -1.01643860e-01 -4.16375756e-01
6.10255077e-02 5.35724878e-01 6.39431655e-01 -6.66771114e-01
-1.06233633e+00 3.97303551e-01 3.15707058e-01 -1.54490516e-01
-1.79874390e-01 -8.39416265e-01 -1.11110425e+00 -3.61413211e-01
-6.18235171e-01 4.12628263e-01 8.71148109e-01 1.16115856e+00
1.44282967e-01 3.29524316e-02 6.37603939e-01 -8.46087754e-01
-9.52577055e-01 -1.11349893e+00 1.80115163e-01 3.38666052e-01
-3.35377641e-02 -1.47065043e-01 -2.69046634e-01 6.37558335e-03] | [11.59279727935791, 10.184125900268555] |
edae532c-3bfb-405d-aad9-4414dafaa8f0 | retinexformer-one-stage-retinex-based | 2303.06705 | null | https://arxiv.org/abs/2303.06705v1 | https://arxiv.org/pdf/2303.06705v1.pdf | Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement | When enhancing low-light images, many deep learning algorithms are based on the Retinex theory. However, the Retinex model does not consider the corruptions hidden in the dark or introduced by the light-up process. Besides, these methods usually require a tedious multi-stage training pipeline and rely on convolutional neural networks, showing limitations in capturing long-range dependencies. In this paper, we formulate a simple yet principled One-stage Retinex-based Framework (ORF). ORF first estimates the illumination information to light up the low-light image and then restores the corruption to produce the enhanced image. We design an Illumination-Guided Transformer (IGT) that utilizes illumination representations to direct the modeling of non-local interactions of regions with different lighting conditions. By plugging IGT into ORF, we obtain our algorithm, Retinexformer. Comprehensive quantitative and qualitative experiments demonstrate that our Retinexformer significantly outperforms state-of-the-art methods on seven benchmarks. The user study and application on low-light object detection also reveal the latent practical values of our method. Codes and pre-trained models will be released. | ['Yulun Zhang', 'Radu Timofte', 'Haoqian Wang', 'Jing Lin', 'Hao Bian', 'Yuanhao Cai'] | 2023-03-12 | null | null | null | null | ['image-enhancement', 'low-light-image-enhancement'] | ['computer-vision', 'computer-vision'] | [ 3.43547881e-01 -5.41865528e-01 2.20536977e-01 -6.01604342e-01
-4.23444569e-01 -2.55126506e-01 5.97630024e-01 -5.51980674e-01
-1.55058131e-01 5.67471981e-01 2.33762816e-01 -1.53316066e-01
2.08586350e-01 -7.25362837e-01 -8.48993719e-01 -1.04642320e+00
5.44420421e-01 -4.07004267e-01 9.75938663e-02 -2.37972543e-01
1.72122777e-01 4.83310997e-01 -1.77269030e+00 5.52685142e-01
7.59595156e-01 8.91814530e-01 3.93565983e-01 6.17964804e-01
5.98569550e-02 8.87144148e-01 -3.73248100e-01 -8.38784352e-02
3.39792520e-01 -4.25571531e-01 -2.19776377e-01 1.95119023e-01
5.79122901e-01 -6.51411951e-01 -3.37294132e-01 1.14380765e+00
7.18225598e-01 7.49974176e-02 3.36691201e-01 -9.12635326e-01
-9.89368141e-01 2.07524952e-02 -8.63272190e-01 2.47086927e-01
2.56950378e-01 5.33173203e-01 6.79890931e-01 -1.30708098e+00
2.36470371e-01 1.10013461e+00 6.07997954e-01 3.09869438e-01
-1.18999434e+00 -6.31466329e-01 1.24514394e-01 2.57104069e-01
-1.04735076e+00 -6.48990691e-01 8.65906417e-01 -1.31186530e-01
7.80568063e-01 2.13247344e-01 5.91390550e-01 1.02867317e+00
4.18323159e-01 6.78041458e-01 1.84755683e+00 -5.82737803e-01
6.53157905e-02 7.98966829e-03 1.18841797e-01 6.27880633e-01
7.88578540e-02 6.19167209e-01 -7.87419021e-01 3.04981142e-01
6.74744666e-01 1.67279243e-01 -5.50166249e-01 1.99594013e-02
-9.20487225e-01 2.21198142e-01 7.35653937e-01 5.40855043e-02
-3.93623590e-01 2.83074439e-01 -2.93050744e-02 -2.10154839e-02
8.24017584e-01 4.83574346e-02 -2.63533890e-01 4.66924667e-01
-7.18804002e-01 -1.04190141e-01 1.84811726e-01 6.75102293e-01
1.16086066e+00 -7.31930956e-02 -3.92557263e-01 9.70795214e-01
5.44375837e-01 5.14551640e-01 -1.36726618e-01 -7.88762629e-01
1.14789285e-01 3.48601043e-01 2.13282749e-01 -6.27377689e-01
-5.95686495e-01 -6.95618033e-01 -8.91792715e-01 6.26853168e-01
2.05326527e-01 3.38798501e-02 -1.05904412e+00 1.53111768e+00
3.37698847e-01 3.48166615e-01 -2.15945560e-02 1.31337452e+00
9.82744396e-01 6.54052198e-01 -5.91848120e-02 -4.85028923e-01
1.33138955e+00 -1.10096419e+00 -9.85210478e-01 -2.95583546e-01
1.41019240e-01 -1.12730825e+00 1.22767067e+00 6.18701279e-01
-1.41408491e+00 -7.91215003e-01 -9.54121113e-01 -5.01667500e-01
-2.64096379e-01 3.55655909e-01 7.32106090e-01 6.49168670e-01
-1.15459073e+00 2.86196768e-01 -6.07573390e-01 -1.57221496e-01
5.16694844e-01 -2.26039765e-03 9.06466171e-02 -4.99133974e-01
-1.04674470e+00 9.89784360e-01 -2.16449782e-01 6.06667757e-01
-1.12847948e+00 -7.19761133e-01 -6.78451359e-01 -9.99553427e-02
2.50565559e-01 -6.98965907e-01 9.08030689e-01 -8.90916288e-01
-1.75006342e+00 7.97903240e-01 -5.83274066e-01 9.45436675e-03
2.67233610e-01 -4.41013783e-01 -3.35750520e-01 -4.77024801e-02
-2.05434874e-01 3.26756895e-01 1.07659042e+00 -1.68443143e+00
-5.11326611e-01 -2.62292475e-01 3.82993013e-01 4.00859863e-01
-1.39930755e-01 3.10523659e-01 -7.39517689e-01 -4.23943728e-01
7.69608319e-02 -7.04237998e-01 2.12050676e-02 6.15716055e-02
-3.12591881e-01 7.15375170e-02 7.79501021e-01 -4.20827299e-01
8.78356397e-01 -2.12139153e+00 -3.41571927e-01 -1.12207569e-01
4.51081395e-01 4.41886455e-01 -2.33669147e-01 2.06775919e-01
-2.70424575e-01 -3.96885008e-01 5.06357402e-02 -5.78732133e-01
-3.81820686e-02 1.61098942e-01 -5.33197641e-01 8.43428969e-01
1.56757757e-01 9.84291196e-01 -1.00590086e+00 -2.10737735e-01
6.78333938e-01 9.19181645e-01 -5.27034760e-01 4.04183537e-01
-2.43172366e-02 5.90661108e-01 1.18490215e-02 7.46788204e-01
1.12865615e+00 -1.40113235e-01 -1.86610237e-01 -9.16298270e-01
-4.56355810e-01 1.74413860e-01 -8.32200229e-01 1.57174003e+00
-7.94667482e-01 8.35277855e-01 3.47708724e-02 -4.10954297e-01
8.66505444e-01 -6.83115125e-02 3.42086047e-01 -1.05195093e+00
3.04455221e-01 2.49298979e-02 -3.95144880e-01 -7.09117830e-01
3.78081739e-01 -2.37674460e-01 6.83554113e-01 5.13518333e-01
-1.62147775e-01 -1.50696069e-01 -6.56547248e-02 -1.07980467e-01
7.81378150e-01 6.14834785e-01 1.54334649e-01 -1.30129941e-02
5.23835003e-01 -7.15751886e-01 3.67299229e-01 6.99283123e-01
-3.21579516e-01 8.27523589e-01 4.70286943e-02 -5.14893889e-01
-5.83054602e-01 -9.79106963e-01 -2.86508322e-01 1.24963307e+00
4.12779361e-01 -2.24662647e-01 -6.02016747e-01 -3.99975061e-01
-4.61287707e-01 8.73725057e-01 -5.93308985e-01 -1.29420966e-01
-4.12503898e-01 -1.12274325e+00 1.08816683e-01 3.45538408e-01
7.53465414e-01 -8.45304370e-01 -6.17962301e-01 -6.66542277e-02
-4.98903424e-01 -1.17699611e+00 -2.62921691e-01 2.38120243e-01
-3.67756277e-01 -1.05890751e+00 -5.23413181e-01 -3.94899726e-01
7.80978322e-01 9.58692789e-01 1.08121467e+00 1.25740454e-01
-4.39761221e-01 4.20225859e-01 -1.96540847e-01 -6.10741735e-01
-3.18877771e-02 -6.41237855e-01 -1.96029127e-01 4.23009962e-01
5.29324353e-01 -5.12256563e-01 -1.07676399e+00 3.93876910e-01
-8.90061736e-01 4.04027700e-01 7.51308620e-01 9.20196533e-01
6.45080388e-01 2.63654411e-01 1.09586202e-01 -6.18063390e-01
3.71976674e-01 -1.24793254e-01 -7.27312267e-01 2.41822988e-01
-6.37325227e-01 -2.57970363e-01 4.27932560e-01 -2.58285552e-01
-1.71233380e+00 2.77599245e-02 -1.02754898e-01 -4.23287392e-01
-3.07042360e-01 4.43125516e-03 -2.12299258e-01 -3.42019886e-01
7.39358366e-01 2.66120404e-01 -4.62667674e-01 -4.79555130e-01
4.88044292e-01 4.49526817e-01 5.67979515e-01 -3.73192072e-01
7.57188559e-01 9.96652186e-01 2.50269920e-01 -7.38047481e-01
-1.36658943e+00 -3.00069422e-01 -5.35587490e-01 -7.02466190e-01
8.31775904e-01 -1.08685684e+00 -9.62105989e-01 6.54391885e-01
-1.31515777e+00 -5.73226988e-01 -2.49893665e-01 2.69786209e-01
-4.22513604e-01 1.22348249e-01 -3.82955641e-01 -9.90661204e-01
-1.52734607e-01 -1.20037091e+00 1.31662798e+00 3.51653099e-01
4.62420881e-01 -8.72095287e-01 5.19753247e-02 4.68469083e-01
5.43788373e-01 -7.72255659e-02 5.66199541e-01 4.17953819e-01
-9.73794222e-01 1.29288688e-01 -7.80649900e-01 5.25549293e-01
2.55430192e-01 -1.60304591e-01 -1.73729587e+00 -2.37129539e-01
2.42056578e-01 -3.24849546e-01 1.02946520e+00 6.54419899e-01
1.31548083e+00 3.57315838e-02 4.50620688e-02 1.20131266e+00
1.62623167e+00 -2.54189909e-01 9.83630836e-01 3.02964300e-01
8.33497226e-01 6.58181190e-01 5.81529498e-01 4.24205989e-01
4.21442568e-01 6.17086232e-01 8.50163758e-01 -8.96423876e-01
-7.78076172e-01 -6.28905892e-02 4.03527647e-01 3.59826952e-01
-4.95521635e-01 -2.93118626e-01 -6.01828575e-01 1.71343803e-01
-1.56790006e+00 -9.24694896e-01 -5.25891244e-01 1.98668110e+00
9.53041971e-01 -1.32888466e-01 -5.23966849e-01 2.48322003e-02
3.98848236e-01 3.36362809e-01 -4.53705579e-01 -5.20111807e-02
-3.89181048e-01 2.55382866e-01 4.44770575e-01 6.83769584e-01
-9.42732215e-01 9.54181492e-01 6.73292446e+00 4.23025280e-01
-1.21646869e+00 3.39644223e-01 6.55899227e-01 -1.87474713e-01
-2.05947578e-01 2.96124034e-02 -8.14666748e-01 3.06494147e-01
4.26866084e-01 3.63400489e-01 7.93967426e-01 2.77936757e-01
8.10752928e-01 -4.25390512e-01 -9.62787449e-01 1.07821274e+00
2.94144422e-01 -8.61902654e-01 -2.36923456e-01 1.65544320e-02
9.54688847e-01 2.03479484e-01 3.31657708e-01 -1.00049183e-01
2.19832972e-01 -7.73311675e-01 6.26619697e-01 9.03451920e-01
9.07201827e-01 -3.86137098e-01 4.74746764e-01 8.63521546e-02
-9.82436359e-01 -1.31489381e-01 -5.10219753e-01 -7.90255442e-02
1.99040860e-01 9.63447928e-01 -4.15346801e-01 4.69670624e-01
9.32676196e-01 9.86715138e-01 -6.49465978e-01 8.87226403e-01
-7.63204396e-01 5.87578714e-01 -9.21262354e-02 4.09098953e-01
-4.87828180e-02 -3.29462618e-01 4.09308344e-01 1.32548285e+00
-5.89554943e-03 2.57551014e-01 -1.08523950e-01 1.20833826e+00
-8.82220827e-03 -3.57172668e-01 -3.67259055e-01 4.82367575e-01
-1.74388602e-01 1.58507240e+00 -4.73355502e-01 -1.63981244e-01
-6.66019261e-01 9.91777360e-01 5.99267718e-04 9.92302775e-01
-1.04643142e+00 -2.07780585e-01 6.19108856e-01 2.27322415e-01
2.42639729e-03 -1.09217592e-01 -5.57093680e-01 -1.14827240e+00
-2.37531960e-02 -7.23419428e-01 -1.42184421e-01 -1.61402190e+00
-1.26104999e+00 4.37735349e-01 -1.26462609e-01 -1.19925237e+00
4.53788370e-01 -6.62306368e-01 -5.70611358e-01 1.13084030e+00
-2.43887591e+00 -1.39085209e+00 -9.46801305e-01 9.70631540e-01
5.06365359e-01 2.76319355e-01 4.12800223e-01 5.14445662e-01
-6.95537448e-01 2.33098432e-01 -5.09802550e-02 -2.61338145e-01
1.05584061e+00 -1.05938697e+00 -1.77456613e-03 1.37066782e+00
-1.08460523e-01 6.98770463e-01 7.40031004e-01 -3.47519726e-01
-1.42117190e+00 -1.18142772e+00 7.01586723e-01 -3.33802581e-01
4.32312399e-01 -3.71898860e-01 -8.37327421e-01 6.06492162e-01
4.17942017e-01 4.89735633e-01 4.59350586e-01 5.62002696e-02
-4.09043491e-01 -5.07200956e-01 -8.00939560e-01 6.86691225e-01
1.16863823e+00 -6.18850529e-01 -3.52622390e-01 4.76515412e-01
3.60986292e-01 -4.56351668e-01 -3.33355874e-01 3.28024685e-01
5.49522519e-01 -1.60697460e+00 1.22754014e+00 -9.48765203e-02
5.95654190e-01 -5.36193371e-01 -2.73244798e-01 -1.27012753e+00
-4.69029218e-01 -7.24077404e-01 -8.35506842e-02 1.11161721e+00
-2.12837920e-01 -8.22932720e-01 2.58176684e-01 5.22460282e-01
-2.85213828e-01 -6.13678217e-01 -3.34529161e-01 -3.73733252e-01
-4.23588037e-01 -5.88958740e-01 4.75385517e-01 7.52750874e-01
-5.98769546e-01 2.54046470e-01 -5.65712392e-01 4.56782699e-01
1.12145758e+00 4.98373853e-03 8.67165685e-01 -8.09797645e-01
-2.50773996e-01 -1.32834509e-01 1.74430579e-01 -1.14089930e+00
-2.13138275e-02 -4.96729136e-01 3.77615988e-01 -1.69133067e+00
3.63491237e-01 -2.98925489e-01 -5.40611327e-01 6.94184601e-01
-4.30948973e-01 6.93541110e-01 3.72079015e-02 2.62463659e-01
-5.81780970e-01 6.23740554e-01 1.47952127e+00 -2.19616364e-03
-1.25297830e-01 -1.48035347e-01 -7.43545949e-01 1.01444840e+00
4.89268452e-01 -2.13929832e-01 -4.03087407e-01 -8.73774230e-01
4.52689171e-01 -6.67078137e-01 8.03240061e-01 -7.86367178e-01
2.50020087e-01 -2.65245587e-01 6.95475996e-01 -8.07579935e-01
5.47213316e-01 -8.87205660e-01 -5.45817949e-02 3.09745297e-02
-1.66284263e-01 -5.51165104e-01 1.04482979e-01 4.43167686e-01
-1.13441225e-03 1.44087553e-01 1.20479751e+00 4.04535197e-02
-4.34980869e-01 1.43486857e-01 -1.52574524e-01 -3.20403397e-01
6.95951819e-01 -6.64127022e-02 -7.14842975e-01 -2.83840656e-01
-2.87746191e-01 -1.69636071e-01 4.02113736e-01 9.66753215e-02
7.22457409e-01 -1.17401266e+00 -6.32421255e-01 5.08849740e-01
5.78499585e-02 -1.85478315e-01 4.73143101e-01 1.09889376e+00
-2.97395945e-01 2.02495858e-01 -7.02201203e-02 -7.13922083e-01
-1.28914070e+00 5.38346648e-01 5.51420093e-01 -5.99181838e-02
-6.30382895e-01 8.07414114e-01 9.13846374e-01 6.89008385e-02
5.07865585e-02 -3.63387018e-01 -1.26823425e-01 -2.13133201e-01
8.49098682e-01 4.55254376e-01 1.75453916e-01 -5.07366478e-01
-1.30003318e-01 7.38361537e-01 1.06546514e-01 1.13554128e-01
1.50837564e+00 -6.13920510e-01 -3.40570986e-01 3.19718212e-01
1.12293196e+00 7.83181116e-02 -1.61312270e+00 -5.25278509e-01
-7.94155598e-01 -8.99317324e-01 6.86623275e-01 -1.10000336e+00
-1.26824975e+00 1.08490121e+00 8.40424001e-01 -1.54555306e-01
1.89732647e+00 -1.85667530e-01 4.92115527e-01 2.08614632e-01
1.94520727e-01 -9.89286602e-01 2.48916686e-01 4.74173963e-01
1.10183752e+00 -1.37005079e+00 3.24248195e-01 -4.46251810e-01
-2.88060606e-01 1.17167902e+00 5.34399927e-01 1.91279888e-01
7.43183672e-01 4.59834397e-01 3.98809642e-01 -2.52412200e-01
-7.16567814e-01 -5.31157911e-01 3.79488409e-01 7.48117268e-01
5.03283143e-01 -3.25073689e-01 1.36378324e-02 2.49472409e-01
1.87893659e-01 1.71526745e-01 3.60837787e-01 6.46969497e-01
-3.33628088e-01 -7.48015642e-01 -7.16718912e-01 -1.17155768e-01
-2.76241124e-01 -3.95418078e-01 -1.46404773e-01 3.38467300e-01
6.10138595e-01 1.22517133e+00 -2.21374139e-01 -1.96245015e-01
3.15235734e-01 -3.62390101e-01 6.31856322e-01 -4.38002765e-01
-4.49528664e-01 6.54884577e-01 -2.54826576e-01 -1.05177426e+00
-8.69185925e-01 -3.63644749e-01 -7.62497485e-01 -8.48534778e-02
-4.83333886e-01 -7.52037942e-01 8.74164104e-01 8.54240358e-01
1.68016389e-01 9.34335947e-01 7.81347573e-01 -1.42253911e+00
4.47336212e-02 -9.06062782e-01 -4.60075051e-01 4.42544788e-01
6.00290895e-01 -7.37622738e-01 -4.99869347e-01 2.84682423e-01] | [10.699371337890625, -2.5032215118408203] |
5722c615-a98e-441a-b708-6139f5d1be2f | probabilistic-knowledge-distillation-of-face | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Xu_Probabilistic_Knowledge_Distillation_of_Face_Ensembles_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Xu_Probabilistic_Knowledge_Distillation_of_Face_Ensembles_CVPR_2023_paper.pdf | Probabilistic Knowledge Distillation of Face Ensembles | Mean ensemble (i.e. averaging predictions from multiple models) is a commonly-used technique in machine learning that improves the performance of each individual model. We formalize it as feature alignment for ensemble in open-set face recognition and generalize it into Bayesian Ensemble Averaging (BEA) through the lens of probabilistic modeling. This generalization brings up two practical benefits that existing methods could not provide: (1) the uncertainty of a face image can be evaluated and further decomposed into aleatoric uncertainty and epistemic uncertainty, the latter of which can be used as a measure for out-of-distribution detection of faceness; (2) a BEA statistic provably reflects the aleatoric uncertainty of a face image, acting as a measure for face image quality to improve recognition performance. To inherit the uncertainty estimation capability from BEA without the loss of inference efficiency, we propose BEA-KD, a student model to distill knowledge from BEA. BEA-KD mimics the overall behavior of ensemble members and consistently outperforms SOTA knowledge distillation methods on various challenging benchmarks. | ['Bryan Hooi', 'Shouhong Ding', 'Jiaxiang Wu', 'Jiaying Wu', 'Miao Xiong', 'Ailin Deng', 'Shen Li', 'Jianqing Xu'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['face-image-quality', 'face-recognition'] | ['computer-vision', 'computer-vision'] | [-9.90106445e-03 1.75728440e-01 1.47147611e-01 -7.04302669e-01
-1.10609949e+00 -5.34869075e-01 8.01503658e-01 -2.61637986e-01
3.05329692e-02 8.07066202e-01 1.93649188e-01 -1.73122168e-01
-5.62771320e-01 -7.15046346e-01 -9.27572966e-01 -1.07502878e+00
1.28474414e-01 4.63941246e-01 -3.14980626e-01 9.92693156e-02
3.70151475e-02 5.42071223e-01 -1.71070647e+00 2.15461224e-01
9.25547302e-01 1.26234913e+00 -6.34350240e-01 3.38782728e-01
3.39151435e-02 5.13472736e-01 -4.87293392e-01 -8.70527267e-01
1.48103029e-01 -3.53701383e-01 -6.83264792e-01 -4.88828532e-02
6.01989031e-01 -3.32754374e-01 -9.32873040e-02 1.14842498e+00
5.35265148e-01 -1.25606507e-01 1.26561165e+00 -1.34774542e+00
-5.14694750e-01 9.29813802e-01 -6.56590879e-01 9.89587680e-02
2.30085880e-01 -2.81285942e-02 9.42741573e-01 -1.08166289e+00
1.11942939e-01 1.72107589e+00 7.01946378e-01 5.32534242e-01
-1.30364978e+00 -8.87175024e-01 -3.11824195e-02 1.32798374e-01
-1.55489779e+00 -7.52798855e-01 4.32688445e-01 -5.99428654e-01
5.68021953e-01 2.16172397e-01 1.03351794e-01 1.22362673e+00
2.01725066e-01 7.92808414e-01 1.31943488e+00 -3.90690148e-01
4.49687511e-01 1.66366979e-01 3.34103107e-01 6.83620691e-01
2.76845366e-01 4.14655715e-01 -7.94841468e-01 -4.82555240e-01
3.03507060e-01 -3.60193044e-01 -2.68275917e-01 -1.44595027e-01
-7.67195106e-01 6.55021906e-01 8.88222829e-02 -2.19481606e-02
-2.80421257e-01 2.31377169e-01 1.79193452e-01 2.72922784e-01
6.45481467e-01 9.78277922e-02 -4.08017099e-01 -5.31869829e-02
-1.00326157e+00 2.69675970e-01 9.35574889e-01 3.30946088e-01
6.70871675e-01 1.62960082e-01 -5.99157810e-01 7.01135159e-01
6.16435289e-01 9.52885747e-01 1.65615797e-01 -1.00780249e+00
-2.14126855e-01 1.77335232e-01 1.02613144e-01 -7.71202326e-01
2.33272731e-01 -3.01691264e-01 -7.20459759e-01 4.14294213e-01
4.79162961e-01 -1.58934742e-01 -8.13576400e-01 1.82880235e+00
2.39794776e-01 9.00884092e-01 -2.40713805e-02 3.27109069e-01
6.42450452e-01 3.56095642e-01 -4.44987565e-02 -3.92360806e-01
1.34474325e+00 -4.20820951e-01 -5.31140149e-01 2.61061847e-01
1.91034555e-01 -3.92124981e-01 2.24109456e-01 6.02085352e-01
-7.40801930e-01 -3.49826008e-01 -1.16449940e+00 6.37372017e-01
-2.28125170e-01 -1.14770390e-01 5.02848208e-01 1.33925128e+00
-1.03937256e+00 7.99870253e-01 -6.86501980e-01 1.68403015e-01
8.07870984e-01 3.40543509e-01 -5.02720952e-01 -1.78132504e-01
-1.03424394e+00 9.33639467e-01 4.94163513e-01 1.21439166e-01
-1.16910768e+00 -1.05764639e+00 -8.24128687e-01 1.06146984e-01
4.89770770e-01 -6.71975970e-01 1.10218501e+00 -7.65027881e-01
-1.59397578e+00 6.32932425e-01 -3.74180287e-01 -3.59085262e-01
5.17776191e-01 -3.22885454e-01 -3.43248516e-01 -2.70608634e-01
-3.14106137e-01 3.87856692e-01 1.37395811e+00 -1.34988499e+00
-4.02679205e-01 -8.48142743e-01 -2.50586450e-01 -2.81397581e-01
-1.87777013e-01 -9.63158067e-03 -1.38222389e-02 -5.69138646e-01
-8.04817528e-02 -7.87626624e-01 1.85778305e-01 -1.80487514e-01
-3.22642565e-01 -5.90229869e-01 8.13192129e-01 -5.62399685e-01
1.14208066e+00 -2.09127092e+00 2.00157925e-01 4.89470810e-01
1.96455553e-01 2.83728093e-01 -1.86173290e-01 -4.88061868e-02
-1.45011276e-01 1.65217653e-01 -3.54800940e-01 -3.09242129e-01
1.21404678e-01 1.51348054e-01 -5.76259315e-01 3.38599712e-01
5.29606640e-01 8.22744310e-01 -8.38914275e-01 -3.50272059e-01
5.70831336e-02 7.53745914e-01 -4.61904526e-01 3.35868821e-02
-3.04767102e-01 5.04929900e-01 -3.21467370e-01 5.83365679e-01
9.40768957e-01 -8.57286677e-02 8.47568363e-02 -1.79220468e-01
5.17284393e-01 -9.71338972e-02 -1.38540518e+00 1.15498734e+00
-2.69311309e-01 2.01993659e-01 -2.54319221e-01 -7.29710817e-01
7.93344855e-01 2.57287443e-01 2.95831800e-01 5.25366552e-02
3.33272636e-01 1.40142456e-01 -5.21849841e-03 -1.02317005e-01
1.34312836e-02 -7.16497330e-03 7.90413097e-02 6.86969101e-01
5.82357705e-01 2.72773411e-02 -1.18942015e-01 8.14438239e-02
8.38593602e-01 1.11127861e-01 2.97161788e-01 -5.27329683e-01
5.96954405e-01 -8.45397413e-01 5.65806031e-01 9.58331943e-01
-1.86715350e-01 3.49941790e-01 5.12768805e-01 -1.56386212e-01
-6.06438339e-01 -1.37823188e+00 -6.40671611e-01 1.08980262e+00
-4.10941273e-01 -3.33045065e-01 -9.62159038e-01 -1.03735900e+00
4.17092979e-01 7.19812214e-01 -9.48072612e-01 -2.17771307e-01
9.66030434e-02 -9.72456753e-01 7.12678134e-01 6.35756552e-01
3.62476289e-01 -5.92155397e-01 -3.94058079e-02 -2.65817225e-01
-6.91111088e-02 -9.41191256e-01 -2.19150603e-01 -2.08097100e-02
-4.70161676e-01 -1.11089265e+00 -5.95384181e-01 1.78955093e-01
4.41624701e-01 -1.37649521e-01 1.24621153e+00 -5.41968010e-02
-7.83506706e-02 8.09375346e-01 -1.41805694e-01 -8.90011609e-01
-5.09558558e-01 -6.22059226e-01 6.05432570e-01 4.82233077e-01
7.68332839e-01 -7.15088308e-01 -2.71878392e-01 2.53395945e-01
-6.68955922e-01 -4.76191968e-01 3.61641586e-01 1.00261211e+00
4.36341494e-01 9.25503820e-02 8.72179449e-01 -7.54702508e-01
5.55063665e-01 -5.03356874e-01 -6.56532586e-01 6.84175134e-01
-8.44986677e-01 5.36888659e-01 -1.66844204e-01 -1.15438081e-01
-1.31466496e+00 -1.78339481e-01 -4.58389968e-02 -4.77912068e-01
-1.13658354e-01 4.51506376e-01 -3.94703060e-01 -2.27129772e-01
5.88594079e-01 7.86184892e-02 1.70949683e-01 -2.70893127e-01
5.14554143e-01 8.72955441e-01 6.02446318e-01 -1.03947663e+00
5.80477715e-01 4.42439377e-01 2.31453389e-01 -6.39247417e-01
-1.12813580e+00 -1.16038442e-01 -6.01386070e-01 -4.15053964e-01
6.00099683e-01 -9.59905088e-01 -9.81332600e-01 7.39464462e-01
-1.00449467e+00 2.29086608e-01 -1.48135707e-01 4.11673635e-01
-4.72228229e-01 4.54912543e-01 -2.00814411e-01 -1.28186810e+00
-3.16342980e-01 -1.03634810e+00 1.03951919e+00 2.97917604e-01
-8.22170377e-02 -7.96976447e-01 -4.17215340e-02 3.41764957e-01
1.83562472e-01 1.89370841e-01 7.94516444e-01 -8.71243119e-01
-2.53818035e-01 -2.09670186e-01 -2.07710549e-01 8.20488036e-01
-1.38975471e-01 1.84669420e-01 -1.68300343e+00 -2.27395847e-01
3.58625725e-02 -4.52464998e-01 1.16207027e+00 3.96783412e-01
1.33317673e+00 -1.29977092e-01 -1.53691575e-01 1.83919251e-01
1.07557189e+00 -1.85443819e-01 6.90438747e-01 -3.52055937e-01
2.65475750e-01 5.38338900e-01 1.89523563e-01 7.66401947e-01
1.38919696e-01 5.03611147e-01 3.59604985e-01 8.63514960e-01
-2.46647745e-02 -2.41655424e-01 7.35636175e-01 5.22502303e-01
-2.89695114e-01 5.02758957e-02 -7.87868500e-01 1.93855837e-01
-1.81170988e+00 -1.14631295e+00 2.05154032e-01 2.34301496e+00
8.57880414e-01 -2.59091169e-01 1.43796995e-01 6.61171451e-02
6.33499503e-01 -8.55217502e-02 -6.03083730e-01 -5.69492131e-02
-1.55514494e-01 2.66489625e-01 5.86200058e-02 4.65850145e-01
-9.00085032e-01 4.14241314e-01 6.60266733e+00 1.29028463e+00
-6.52212501e-01 1.68640688e-01 9.21918154e-01 1.94869444e-01
-4.33648407e-01 -1.55741721e-01 -1.01360190e+00 4.81537551e-01
1.14764893e+00 -2.74902225e-01 4.10680205e-01 8.44049156e-01
-4.13138479e-01 -2.34040454e-01 -1.59924424e+00 1.07432568e+00
3.46072644e-01 -1.00412345e+00 1.77372292e-01 2.41667941e-01
9.93223965e-01 -1.61304697e-01 4.47316408e-01 5.25781751e-01
6.50339127e-01 -1.51026678e+00 6.08970582e-01 9.62632418e-01
6.09164715e-01 -8.37504566e-01 1.02214062e+00 3.70813996e-01
-6.33189738e-01 -1.14389069e-01 -1.83379263e-01 2.56329089e-01
-2.08601028e-01 1.01209998e+00 -5.96520841e-01 8.50215673e-01
6.58509672e-01 3.44299316e-01 -6.14873230e-01 7.77478456e-01
-1.37861326e-01 6.79136634e-01 -5.84631681e-01 2.98793286e-01
-1.88274294e-01 -1.78952903e-01 6.67637765e-01 9.96288240e-01
4.68492329e-01 3.75005305e-02 -2.14579433e-01 9.70810294e-01
-2.86150575e-01 -3.36820990e-01 -4.28203642e-01 -1.20713048e-01
6.96767151e-01 1.05521810e+00 -1.94263130e-01 -2.62070924e-01
-1.04521103e-01 6.65995300e-01 3.54053438e-01 1.75764382e-01
-7.56658554e-01 1.08054459e-01 7.10822821e-01 -4.46867287e-01
4.42933530e-01 3.37696880e-01 -1.61028072e-01 -1.21470046e+00
-4.75511141e-02 -9.36020136e-01 5.58954120e-01 -6.75277054e-01
-1.76021171e+00 4.83416200e-01 1.70112997e-01 -8.02230775e-01
-4.88377571e-01 -9.63794291e-01 -5.67465007e-01 8.90667319e-01
-1.29717541e+00 -1.22797751e+00 -1.69031501e-01 4.66957837e-01
1.18697122e-01 -2.77223259e-01 1.11545181e+00 -1.31663010e-01
-6.81887150e-01 9.86925960e-01 2.90296733e-01 7.40364790e-02
7.51249313e-01 -1.17056000e+00 -5.16315401e-02 7.18742490e-01
3.64534050e-01 5.27338445e-01 7.31934130e-01 -4.74153221e-01
-1.11201251e+00 -9.83593643e-01 5.72645545e-01 -1.22342014e+00
5.80986381e-01 8.64956975e-02 -9.02000070e-01 6.26902103e-01
-8.39112699e-02 2.29589328e-01 8.77204359e-01 6.39307022e-01
-1.03426909e+00 -1.90275043e-01 -1.35250461e+00 2.46692851e-01
6.79327369e-01 -6.61448359e-01 -6.71090961e-01 5.20985108e-03
1.84692487e-01 -9.73426402e-02 -1.14538467e+00 7.88594007e-01
8.83085132e-01 -1.26966763e+00 8.90685976e-01 -7.76362956e-01
2.22099304e-01 -2.41843343e-01 -4.79514897e-01 -1.41143787e+00
-1.16000615e-01 -4.70658034e-01 -6.42744362e-01 1.53749347e+00
2.84197897e-01 -7.08567858e-01 3.08389604e-01 7.31039226e-01
3.72197360e-01 -7.57815003e-01 -1.18942273e+00 -1.00212908e+00
4.10940260e-01 -8.17128360e-01 1.06761730e+00 6.28286362e-01
-1.81027085e-01 5.41945770e-02 -2.74955094e-01 3.78997386e-01
1.10960233e+00 -2.51740664e-01 5.77435374e-01 -1.66815031e+00
-2.78513134e-01 -4.39074367e-01 -6.13750577e-01 -6.13780141e-01
6.18300915e-01 -9.04099166e-01 1.82085596e-02 -6.57768965e-01
6.43638968e-01 -9.62703153e-02 -6.26880646e-01 2.66342461e-01
-2.74504960e-01 1.96905226e-01 1.58484466e-02 -1.44126294e-02
-5.55598497e-01 6.54254019e-01 7.52208471e-01 -8.00481662e-02
3.79127622e-01 1.06214151e-01 -7.97090411e-01 8.79313290e-01
3.53990853e-01 -3.32210362e-01 -2.59864330e-01 2.55591553e-02
1.53396919e-01 -6.53502485e-03 6.52188122e-01 -9.37618732e-01
4.83154021e-02 1.31934937e-02 5.57601810e-01 -3.83535862e-01
2.38196746e-01 -6.88097417e-01 1.82943493e-01 5.73358871e-02
-1.04043186e-01 -6.34729087e-01 1.54913783e-01 1.04635108e+00
-8.18241537e-02 -3.07779849e-01 7.88540483e-01 1.64677143e-01
-5.44343829e-01 3.04109156e-01 -3.50665078e-02 -1.56570330e-01
9.94770050e-01 -3.15739512e-02 -2.52574652e-01 -3.58549654e-01
-7.85751402e-01 1.69506818e-01 -1.74112264e-02 3.91904294e-01
4.52706009e-01 -1.39331138e+00 -1.12592113e+00 3.28109801e-01
1.54118299e-01 -3.46621484e-01 6.10796154e-01 8.90475154e-01
2.41232574e-01 3.44785899e-01 6.57230541e-02 -8.48397613e-01
-1.22809780e+00 2.46042833e-01 4.53932405e-01 -3.11441958e-01
-5.15727811e-02 1.06157672e+00 1.19198196e-01 -4.79990304e-01
4.04531568e-01 2.81323135e-01 -2.38794070e-02 1.31403375e-02
1.04405034e+00 4.32933480e-01 2.06679046e-01 -5.82919121e-01
-4.29497153e-01 3.64241987e-01 -1.03462011e-01 -1.96657911e-01
1.18613851e+00 -3.13743837e-02 -3.97426456e-01 5.14485776e-01
6.96576118e-01 -7.60847852e-02 -1.23558235e+00 -3.63942564e-01
4.87850755e-02 -4.46580082e-01 3.14700007e-01 -1.20642066e+00
-7.15047300e-01 9.30367649e-01 8.36768329e-01 -1.28875807e-01
1.01334357e+00 -2.10125209e-03 7.26464167e-02 5.92087448e-01
6.45714879e-01 -7.53521800e-01 1.20260324e-02 5.54506898e-01
1.00074816e+00 -1.49638498e+00 1.97786652e-02 -2.49686047e-01
-4.41458255e-01 1.19677234e+00 4.40249324e-01 1.68914065e-01
1.17867529e+00 3.09295654e-01 -2.95583546e-01 -6.67170957e-02
-6.89741492e-01 -1.35721847e-01 6.74694479e-01 7.25726843e-01
1.46007463e-01 1.71481147e-01 2.10462034e-01 1.09387231e+00
-5.64717688e-02 1.14537269e-01 5.66009097e-02 3.86161774e-01
-3.17069888e-01 -8.45806777e-01 -4.63688344e-01 6.23915911e-01
-5.35660028e-01 7.73885697e-02 -2.17519328e-01 2.91925073e-01
5.57157956e-02 1.04940629e+00 -4.08470295e-02 -5.08996964e-01
-1.93349347e-01 6.37603045e-01 8.15246999e-01 -4.01019812e-01
-7.14708269e-02 -3.76511514e-01 -6.16508275e-02 -4.03578877e-01
-4.95607734e-01 -6.71392560e-01 -5.68500042e-01 -5.33740401e-01
-6.64495826e-01 2.51563966e-01 5.43418646e-01 1.32793605e+00
5.27549446e-01 2.61860967e-01 6.78777575e-01 -7.02861905e-01
-1.24608898e+00 -1.15869975e+00 -8.14215422e-01 1.59253001e-01
1.44674838e-01 -9.63909268e-01 -6.31307900e-01 -4.03010845e-02] | [7.7538676261901855, 4.021604537963867] |
71851d07-0f7e-4c9e-968f-106db91485c9 | enhancing-aspect-extraction-for-hindi | null | null | https://aclanthology.org/2021.ecnlp-1.17 | https://aclanthology.org/2021.ecnlp-1.17.pdf | Enhancing Aspect Extraction for Hindi | Aspect extraction is not a well-explored topic in Hindi, with only one corpus having been developed for the task. In this paper, we discuss the merits of the existing corpus in terms of quality, size, sparsity, and performance in aspect extraction tasks using established models. To provide a better baseline corpus for aspect extraction, we translate the SemEval 2014 aspect-based sentiment analysis dataset and annotate the aspects in that data. We provide rigorous guidelines and a replicable methodology for this task. We quantitatively evaluate the translations and annotations using inter-annotator agreement scores. We also evaluate our dataset using state-of-the-art neural aspect extraction models in both monolingual and multilingual settings and show that the models perform far better on our corpus than on the existing Hindi dataset. With this, we establish our corpus as the gold-standard aspect extraction dataset in Hindi. | ['Manish Shrivastava', 'Alok Debnath', 'Arghya Bhattacharya'] | null | null | null | null | acl-ecnlp-2021-8 | ['aspect-extraction'] | ['natural-language-processing'] | [-1.00963879e-02 3.11380804e-01 -4.23684448e-01 -5.09836495e-01
-1.34726322e+00 -8.20077181e-01 8.92430425e-01 2.83633411e-01
-6.99137449e-01 8.68817627e-01 6.29894793e-01 -1.81945711e-01
2.77931631e-01 -5.61710894e-01 -5.06341100e-01 -3.23309332e-01
3.02389592e-01 9.16699469e-01 -1.14980854e-01 -3.72718036e-01
1.52192200e-02 -2.73884356e-01 -9.80217636e-01 5.07197678e-01
5.15323043e-01 6.27509832e-01 -2.39630252e-01 1.64359361e-01
3.05998847e-02 6.98537409e-01 -9.01591599e-01 -1.01240671e+00
1.22281797e-01 -2.97927618e-01 -9.77625012e-01 -1.55600207e-02
3.35320830e-01 9.78717878e-02 8.66027176e-03 7.09316492e-01
6.42524123e-01 -3.22485566e-01 7.59159088e-01 -9.78706181e-01
-7.87534177e-01 1.29415882e+00 -4.97049719e-01 -8.27923790e-02
2.52910644e-01 -1.24853186e-01 1.57306743e+00 -8.43952954e-01
1.15984130e+00 6.25406623e-01 8.05691123e-01 2.36580461e-01
-8.51468921e-01 -4.80601192e-01 -3.92589681e-02 2.05668107e-01
-1.05599928e+00 -3.93592000e-01 6.85029924e-01 -2.06693858e-01
1.56896973e+00 2.90212426e-02 9.81933951e-01 1.10419130e+00
3.07699889e-01 1.37206304e+00 1.32464707e+00 -6.52880251e-01
8.95647705e-02 5.17376781e-01 3.12499881e-01 1.98020577e-01
4.83702779e-01 -4.35857296e-01 -7.62376428e-01 2.68563647e-02
9.26591232e-02 -7.15920448e-01 6.17820211e-02 -5.52991405e-02
-1.30829942e+00 1.08849096e+00 -1.02871642e-01 4.71832871e-01
-4.22356039e-01 -1.75929084e-01 7.15985417e-01 4.19595450e-01
1.00666833e+00 7.02764094e-01 -1.06236017e+00 -3.66735786e-01
-1.11281824e+00 4.42527264e-01 1.17505813e+00 1.36680007e+00
6.11456752e-01 6.15707934e-02 -1.35908186e-01 1.05960631e+00
2.48585597e-01 8.09401393e-01 2.65057147e-01 -6.84143245e-01
8.71347666e-01 7.43697286e-01 -4.93194610e-01 -5.33923209e-01
-3.02829951e-01 -5.00229120e-01 -5.01078129e-01 5.53514622e-02
1.69693127e-01 -3.34843367e-01 -8.49880755e-01 1.47027743e+00
3.32879305e-01 -9.37274158e-01 8.56412873e-02 3.55849385e-01
1.18993270e+00 4.12372321e-01 7.79342130e-02 -5.98888798e-03
1.75405717e+00 -1.22267878e+00 -8.52232039e-01 -6.23293638e-01
7.80524194e-01 -1.29193366e+00 9.34968591e-01 3.93496037e-01
-9.92083967e-01 9.46905985e-02 -1.29072058e+00 -8.80664513e-02
-7.00001001e-01 3.33075464e-01 9.29441333e-01 5.18843234e-01
-9.32707965e-01 6.10784367e-02 -6.49174750e-01 -5.49236059e-01
4.82100606e-01 9.25081223e-02 -6.32442057e-01 1.05268523e-01
-1.15261829e+00 1.03591108e+00 4.00008023e-01 -2.18364000e-01
-6.42146647e-01 -8.35438371e-01 -1.20144951e+00 -4.46046531e-01
3.20636570e-01 -4.70732957e-01 1.45839655e+00 -6.76308632e-01
-9.92200732e-01 1.30412745e+00 -9.08562019e-02 -4.22066957e-01
4.85067954e-03 -2.93228149e-01 -3.86521161e-01 -1.08905569e-01
6.12605095e-01 5.64242125e-01 6.11322939e-01 -9.93747473e-01
-8.74911010e-01 -5.50110519e-01 3.82417977e-01 2.89939642e-01
-3.55204254e-01 4.26894456e-01 -7.83020020e-01 -7.78482914e-01
-2.72819012e-01 -1.30234039e+00 -2.70717621e-01 -7.48336256e-01
-4.77319390e-01 -1.47208641e-03 5.54992855e-01 -8.55911255e-01
1.39153922e+00 -1.73632979e+00 -1.63122967e-01 -2.45046411e-02
6.43155873e-02 4.22064625e-02 -3.66001219e-01 4.62016076e-01
2.26248339e-01 7.25494176e-02 -4.07261193e-01 -4.96023804e-01
3.26319076e-02 2.05260083e-01 8.50720629e-02 3.40176314e-01
1.73676938e-01 1.25507927e+00 -4.84584898e-01 -6.73845530e-01
-2.65445173e-01 6.18402421e-01 -7.40727007e-01 9.09085050e-02
-2.11319059e-01 2.18428671e-01 -2.43658856e-01 9.24430311e-01
3.44538122e-01 1.22125246e-01 4.28358406e-01 -3.49595606e-01
-1.96154136e-02 8.95233810e-01 -7.47526646e-01 1.88376498e+00
-8.38218629e-01 8.40410590e-01 -8.21383446e-02 -6.42110467e-01
7.52443373e-01 5.49408019e-01 4.81924862e-01 -7.60222137e-01
1.05076388e-01 3.55741173e-01 -5.22266813e-02 -2.30045930e-01
9.56007719e-01 -2.42212906e-01 -6.86316490e-01 9.01247263e-01
7.40546167e-01 -6.71635687e-01 7.83490479e-01 1.82343990e-01
8.84434581e-01 3.94578457e-01 9.45419848e-01 -4.37578619e-01
3.54125500e-01 4.67968762e-01 3.81864667e-01 3.45721573e-01
-5.30269369e-02 9.00756001e-01 4.32978302e-01 -4.31983918e-01
-1.34097922e+00 -5.97170711e-01 -2.72719830e-01 9.45866406e-01
-6.79613411e-01 -9.57648814e-01 -7.34662056e-01 -1.03142846e+00
-6.35443389e-01 8.78813744e-01 -8.40612948e-01 5.25956452e-01
-8.04569006e-01 -1.29820418e+00 4.72545445e-01 4.35456514e-01
3.73305351e-01 -1.47351730e+00 -4.73797947e-01 3.44041049e-01
-5.07093191e-01 -1.44053853e+00 -2.18796194e-01 5.59578598e-01
-3.81632626e-01 -8.95691097e-01 -4.04089987e-01 -8.81573081e-01
2.07475796e-01 -1.03735276e-01 1.93273282e+00 -3.79613787e-01
9.38413292e-02 2.20874652e-01 -5.31727612e-01 -8.33536506e-01
-4.29414123e-01 9.22022700e-01 -5.58329463e-01 -8.30707133e-01
9.26990509e-01 -4.58193898e-01 -3.67768943e-01 2.64396332e-02
-1.02350080e+00 -2.50398546e-01 6.34429872e-01 5.18498540e-01
6.27381623e-01 -1.97958335e-01 1.55497566e-01 -1.48558724e+00
6.27296209e-01 -3.46503407e-01 -2.45659009e-01 3.99772339e-02
-9.54973519e-01 -1.15060255e-01 2.61090010e-01 1.25803992e-01
-1.17477930e+00 -2.83133179e-01 -5.09697795e-01 7.82181323e-01
-1.49757653e-01 1.07905459e+00 -2.85893947e-01 3.59460354e-01
4.44227189e-01 1.04973450e-01 -4.81712520e-01 -4.00440842e-01
7.39862561e-01 7.69427657e-01 1.65497690e-01 -4.83895987e-01
6.01383805e-01 4.43247765e-01 -3.13936949e-01 -9.70027387e-01
-1.46885002e+00 -4.56516862e-01 -1.09076560e+00 1.55422568e-01
7.78221548e-01 -1.35613263e+00 2.65317500e-01 3.49114478e-01
-1.15726459e+00 -1.14518695e-01 -5.22480011e-01 1.78048879e-01
-5.61939597e-01 1.40218034e-01 -8.61019194e-01 -5.15075445e-01
-9.50888574e-01 -1.14442861e+00 1.32481372e+00 -3.23233485e-01
-9.66752350e-01 -1.14415622e+00 6.64411485e-01 7.65464604e-01
5.07107615e-01 2.24494889e-01 1.02196193e+00 -8.45108986e-01
-4.56435323e-01 -2.86981106e-01 5.57487905e-02 3.01786691e-01
3.13067902e-03 -6.82145879e-02 -9.79331732e-01 -2.49172933e-02
5.26450351e-02 -4.08662081e-01 5.11705995e-01 2.71223694e-01
-1.50206126e-02 -6.59723639e-01 8.95438716e-02 4.23437148e-01
1.46452439e+00 6.44128257e-03 6.18405759e-01 1.36873114e+00
6.61626577e-01 6.43191814e-01 7.61126280e-01 2.48110473e-01
9.94743526e-01 6.27309144e-01 -2.28549168e-01 -1.62611045e-02
-2.84254491e-01 -9.64122638e-02 4.49009955e-01 1.66870081e+00
-1.48465768e-01 -2.85496213e-03 -7.98033714e-01 1.16374326e+00
-1.60761583e+00 -6.09039366e-01 -6.39970899e-02 1.77051997e+00
1.26169002e+00 3.10866594e-01 2.48163328e-01 9.22745168e-02
-1.50952429e-01 4.86498982e-01 2.17827961e-01 -6.31360948e-01
-5.22636354e-01 5.29085696e-01 1.72392592e-01 3.37828666e-01
-1.34024334e+00 1.16942275e+00 6.73879957e+00 8.99578333e-01
-6.65684938e-01 6.45853281e-01 5.21867514e-01 -2.54513174e-01
-4.70844269e-01 1.08974725e-01 -1.10073936e+00 -2.60956120e-02
1.07126606e+00 1.02402819e-02 1.33576527e-01 1.05096924e+00
-2.02212065e-01 -6.13442017e-03 -9.39846098e-01 5.52821577e-01
5.42737603e-01 -1.17207468e+00 5.97706251e-02 1.80414155e-01
1.01598608e+00 6.29141748e-01 -2.12592334e-01 7.07000256e-01
3.14294636e-01 -8.56034935e-01 9.26538229e-01 -3.63404602e-01
6.65115178e-01 -7.42029250e-01 1.17157316e+00 -2.39722058e-01
-1.13763809e+00 6.41545713e-01 -2.00751543e-01 9.12822112e-02
3.98601949e-01 6.40667140e-01 -6.34744287e-01 5.76550126e-01
7.46804118e-01 9.85178649e-01 -8.54324281e-01 6.38574541e-01
-9.23173368e-01 9.61241663e-01 -3.76757771e-01 -9.85286087e-02
5.01599014e-01 -5.46995163e-01 5.88892341e-01 1.43460357e+00
5.64179420e-02 -6.70072556e-01 -1.16652492e-02 2.87335932e-01
-4.05509435e-02 7.01458752e-01 -7.84751236e-01 -1.73780203e-01
-5.97067103e-02 1.57776999e+00 -6.48192763e-01 -4.72107619e-01
-9.48688686e-01 7.47097790e-01 5.85491180e-01 3.30473930e-01
-6.08189583e-01 -2.76167721e-01 6.39760375e-01 -5.42365909e-02
5.80793738e-01 -2.59986311e-01 -8.44896972e-01 -1.50714552e+00
2.72383958e-01 -1.42073166e+00 5.78303576e-01 -5.12620091e-01
-1.01152730e+00 1.34186673e+00 9.31842178e-02 -1.25456691e+00
-4.28633541e-01 -6.62296176e-01 -3.34150732e-01 4.18496668e-01
-1.73686063e+00 -1.96207535e+00 2.47803986e-01 1.42964020e-01
6.69054568e-01 -4.85211939e-01 9.49642360e-01 5.30526698e-01
-2.75850266e-01 6.52106225e-01 -2.40732402e-01 2.34861210e-01
6.67428613e-01 -1.30602336e+00 8.15830648e-01 7.66833186e-01
6.64005697e-01 7.40855813e-01 7.18101680e-01 -7.54093349e-01
-1.09998453e+00 -1.01791346e+00 1.62007177e+00 -1.00365472e+00
1.19265187e+00 -4.50053066e-01 -5.24213254e-01 1.15408468e+00
8.74140739e-01 -8.18361759e-01 1.22993040e+00 7.63479471e-01
-5.50070524e-01 1.71651930e-01 -7.94768453e-01 5.82069457e-01
7.89292037e-01 -5.50581455e-01 -8.67128134e-01 3.11400682e-01
6.62942588e-01 -1.17040478e-01 -1.18043506e+00 4.90935594e-01
8.05247188e-01 -6.66744232e-01 6.77254975e-01 -4.64683264e-01
9.01621163e-01 -9.41292048e-02 -4.76182044e-01 -1.50301850e+00
-1.66971564e-01 -4.26522285e-01 2.38713413e-01 1.77055490e+00
1.57261550e+00 -3.75721306e-01 8.09744179e-01 4.11540717e-01
1.65004894e-01 -9.88340676e-01 -7.96386719e-01 -4.55150992e-01
3.82513970e-01 -9.61410522e-01 5.82460701e-01 7.61039197e-01
2.33092159e-01 1.29298007e+00 -3.42338741e-01 -5.46463430e-01
5.78042090e-01 5.36978066e-01 8.87588501e-01 -6.88998640e-01
-4.70157743e-01 -2.34047040e-01 -1.80732593e-01 -3.32553655e-01
1.67331666e-01 -1.13972521e+00 -7.57793561e-02 -1.85510254e+00
7.68779814e-01 -1.64319754e-01 5.34105264e-02 7.80221641e-01
-3.48140836e-01 8.32825899e-01 1.81248516e-01 3.29196393e-01
-6.99973106e-01 6.60194337e-01 8.86224508e-01 -4.80698943e-01
6.15194328e-02 -1.06786482e-01 -1.15881753e+00 8.29324424e-01
8.90380919e-01 -7.76634574e-01 -1.44805193e-01 -5.60612082e-01
7.55196273e-01 -7.14513719e-01 -6.31411254e-01 -8.34168851e-01
-4.51214433e-01 2.89762497e-01 2.11719170e-01 -7.95168400e-01
4.22640830e-01 -7.32699096e-01 -2.59427607e-01 -1.18665799e-01
-2.00818509e-01 3.47779930e-01 2.11017773e-01 -1.01443186e-01
-5.06586552e-01 -3.69294792e-01 4.66313779e-01 -2.96976864e-01
-5.15690923e-01 2.46675238e-01 -8.03413928e-01 7.20543683e-01
5.09911120e-01 2.71471500e-01 -2.93228954e-01 -5.87193131e-01
-2.55307972e-01 -1.63029760e-01 4.76551563e-01 3.77963513e-01
2.21303895e-01 -1.29611456e+00 -9.54668343e-01 -4.70285043e-02
6.90032244e-01 -4.22238320e-01 -3.28644335e-01 8.16111326e-01
-4.40954506e-01 7.01766014e-01 -2.13821575e-01 -2.11841375e-01
-1.17962933e+00 2.44022354e-01 -4.68267053e-01 -9.29050863e-01
-4.18581992e-01 2.39640176e-01 -5.64845279e-02 -1.30314028e+00
-3.25696945e-01 -1.37953147e-01 -5.79507351e-01 3.33081871e-01
4.15697575e-01 -6.04092441e-02 2.57060081e-01 -8.60073328e-01
-3.69837344e-01 5.51848412e-01 -9.38166827e-02 -6.51915491e-01
1.82670021e+00 -1.83623761e-01 -4.07822758e-01 5.91577470e-01
1.25839996e+00 2.47460976e-01 -1.84796482e-01 -1.96086168e-01
2.31594175e-01 7.98418149e-02 -2.03596160e-01 -9.94184375e-01
-1.08659101e+00 4.39172536e-01 -7.85764307e-02 3.71074043e-02
8.30567420e-01 2.25039423e-01 9.83562887e-01 5.49018681e-01
3.26570719e-01 -1.44568586e+00 -4.63895798e-01 1.04583800e+00
8.16664457e-01 -1.39357412e+00 4.24777299e-01 -4.53560978e-01
-1.10068989e+00 5.71426451e-01 3.03601235e-01 7.48521984e-02
9.29191411e-01 7.49328434e-01 6.87536716e-01 -5.78309059e-01
-6.92642272e-01 -3.83296251e-01 5.06968677e-01 6.68047071e-01
1.12143517e+00 1.93602026e-01 -7.76332796e-01 7.50006974e-01
-9.95556712e-01 -1.62740141e-01 5.83176732e-01 1.10555196e+00
3.21704969e-02 -1.22857034e+00 1.77495286e-01 4.77212310e-01
-1.08400989e+00 -7.05286384e-01 -4.50210482e-01 1.10793006e+00
-3.20619047e-01 9.94267762e-01 -6.08926535e-01 -6.14528880e-02
3.88278782e-01 2.61165407e-02 4.51572835e-01 -7.85787284e-01
-8.76009822e-01 3.42359513e-01 9.52160418e-01 -3.34059656e-01
-8.32693219e-01 -7.47754216e-01 -5.20779014e-01 1.44357413e-01
-9.15384740e-02 3.00906807e-01 1.06277096e+00 1.22574925e+00
1.79511681e-01 2.89667666e-01 8.24135989e-02 -4.36835408e-01
1.66787580e-02 -1.41080916e+00 -4.55052823e-01 4.84369397e-01
-4.20805693e-01 -8.92471075e-02 -2.17856169e-01 1.75980300e-01] | [11.40244197845459, 6.792319297790527] |
226f5c6a-22e5-4715-a73f-4256867e7c4b | sentibase-sentiment-analysis-in-twitter-on-a | null | null | https://aclanthology.org/S15-2098 | https://aclanthology.org/S15-2098.pdf | Sentibase: Sentiment Analysis in Twitter on a Budget | null | ['Aditya Joshi', 'Vasudeva Varma', 'Satarupa Guha'] | 2015-06-01 | null | null | null | semeval-2015-6 | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.181724548339844, 3.8228163719177246] |
9cc0031c-f0ad-4bb2-97b3-8ab82a5b23b4 | patient-risk-assessment-and-warning-symptom | 1809.10804 | null | http://arxiv.org/abs/1809.10804v1 | http://arxiv.org/pdf/1809.10804v1.pdf | Patient Risk Assessment and Warning Symptom Detection Using Deep Attention-Based Neural Networks | We present an operational component of a real-world patient triage system.
Given a specific patient presentation, the system is able to assess the level
of medical urgency and issue the most appropriate recommendation in terms of
best point of care and time to treat. We use an attention-based convolutional
neural network architecture trained on 600,000 doctor notes in German. We
compare two approaches, one that uses the full text of the medical notes and
one that uses only a selected list of medical entities extracted from the text.
These approaches achieve 79% and 66% precision, respectively, but on a
confidence threshold of 0.6, precision increases to 85% and 75%, respectively.
In addition, a method to detect warning symptoms is implemented to render the
classification task transparent from a medical perspective. The method is based
on the learning of attention scores and a method of automatic validation using
the same data. | ['Ce Zhang', 'Lorenz Kuhn', 'An-phi Nguyen', 'Ivan Girardi', 'Nora Hollenstein', 'Adam Ivankay', 'Pengfei Ji', 'Chiara Marchiori'] | 2018-09-28 | patient-risk-assessment-and-warning-symptom-1 | https://aclanthology.org/W18-5616 | https://aclanthology.org/W18-5616.pdf | ws-2018-10 | ['deep-attention', 'deep-attention'] | ['computer-vision', 'natural-language-processing'] | [ 2.30386391e-01 5.60015559e-01 -2.24779293e-01 -3.73774409e-01
-7.02803314e-01 -4.82596636e-01 3.80242229e-01 1.16446877e+00
-7.83341110e-01 3.67193758e-01 4.51740265e-01 -5.87897420e-01
-5.02663493e-01 -8.63602877e-01 -1.53624505e-01 -3.16480160e-01
-2.55968515e-02 7.85863042e-01 -5.49398325e-02 -3.62844504e-02
1.62443042e-01 3.99040043e-01 -1.28529489e+00 6.45141840e-01
9.20434415e-01 1.28096652e+00 -4.28446755e-02 9.48625207e-01
1.08459957e-01 8.80079150e-01 -6.80503607e-01 -4.31719363e-01
1.32229924e-01 -2.98568398e-01 -8.35759461e-01 -1.39998525e-01
3.62490773e-01 -3.57292950e-01 -1.91302106e-01 6.73690557e-01
7.13057697e-01 7.29808658e-02 6.80223525e-01 -5.62177300e-01
-4.76385236e-01 4.65263397e-01 3.57692987e-01 4.08013612e-01
5.79071939e-01 8.39465931e-02 8.89964461e-01 -5.21460235e-01
6.60911560e-01 5.78844428e-01 5.99915802e-01 6.38548017e-01
-8.67228508e-01 -2.67700523e-01 -3.69333625e-02 -9.33411568e-02
-9.77717817e-01 -2.57959455e-01 5.48047498e-02 -6.52136803e-01
1.18762672e+00 5.45367241e-01 7.28441298e-01 7.78789580e-01
3.78802150e-01 1.82338655e-01 8.04521322e-01 -4.75229383e-01
2.67967552e-01 4.65300381e-01 2.95200199e-01 6.06727540e-01
2.81200528e-01 -1.37719616e-01 -6.00684434e-02 -6.47204399e-01
3.71911526e-01 3.63443702e-01 -3.46651107e-01 2.63296813e-02
-1.17725456e+00 8.11336756e-01 5.61685622e-01 6.61222219e-01
-7.76349187e-01 -5.33488452e-01 5.90474844e-01 -1.18629988e-02
3.03784728e-01 9.11406934e-01 -5.51593721e-01 1.14565320e-01
-1.09068382e+00 3.10046464e-01 1.12410903e+00 2.88582593e-01
-7.18839467e-02 -4.65830863e-01 -8.05341244e-01 5.05275071e-01
2.70603657e-01 3.95060718e-01 7.02778339e-01 -6.40498757e-01
3.85069013e-01 9.05482650e-01 3.34071189e-01 -9.21047628e-01
-7.17701375e-01 -5.88511050e-01 -5.55328667e-01 1.36032477e-02
2.63715774e-01 -5.38003623e-01 -1.19971502e+00 1.40010059e+00
1.08002685e-01 -1.10693321e-01 3.24372232e-01 7.83268750e-01
1.05057299e+00 2.93851733e-01 4.14542824e-01 -1.57291964e-01
1.72362971e+00 -6.43074572e-01 -8.60436916e-01 -6.91860765e-02
8.50406468e-01 -5.00537336e-01 5.74461102e-01 4.21119362e-01
-1.23630154e+00 -3.08973700e-01 -1.04878104e+00 1.41295260e-02
-5.78269362e-01 4.79353458e-01 1.65027678e-01 5.02616346e-01
-9.96187150e-01 9.68471587e-01 -7.60007918e-01 -6.19736075e-01
2.73341209e-01 6.11173749e-01 -2.57384658e-01 1.57554299e-01
-1.28277397e+00 1.23732686e+00 3.23310226e-01 -1.20789126e-01
-2.78899312e-01 -6.68662310e-01 -8.04077804e-01 5.31204045e-01
9.97854210e-03 -9.35659766e-01 1.36017120e+00 -8.21306765e-01
-1.21668518e+00 1.15709972e+00 2.46774465e-01 -8.24739993e-01
6.26213193e-01 -2.87022680e-01 -6.14573896e-01 4.52211499e-01
5.06804772e-02 2.52279103e-01 4.00976896e-01 -5.96319258e-01
-1.04375303e+00 -2.24337876e-01 1.57081842e-01 1.36922851e-01
-4.41019684e-01 -1.31383501e-02 -4.59095389e-01 -4.97919172e-01
-3.68215650e-01 -9.19667721e-01 -5.39168477e-01 -1.24471411e-01
-3.27095479e-01 -6.70267195e-02 2.17782721e-01 -1.00813341e+00
1.55949533e+00 -2.06553864e+00 -9.63620171e-02 4.33013827e-01
5.96149683e-01 7.29988158e-01 2.26888552e-01 3.98803800e-01
-2.71974027e-01 1.45978287e-01 -1.44481167e-01 -2.07298326e-05
-3.54266614e-01 -1.68940872e-01 -4.84010689e-02 1.03070490e-01
3.19363862e-01 6.19837761e-01 -9.95845020e-01 -3.47289413e-01
4.82934475e-01 6.66097403e-01 -7.16359556e-01 4.13466483e-01
5.17026484e-02 3.91047835e-01 -6.04194701e-01 3.93820584e-01
8.25529620e-02 -5.59516609e-01 3.08016717e-01 -2.03895956e-01
-3.68322134e-02 7.49083936e-01 -8.88893545e-01 1.21308923e+00
-5.69250524e-01 2.90009528e-01 -1.98741153e-01 -7.75952637e-01
6.50055110e-01 7.59680271e-01 7.04223633e-01 -4.43567991e-01
3.75298381e-01 2.45666638e-01 2.94840753e-01 -8.28819931e-01
3.50681245e-01 1.56267270e-01 -1.67026687e-02 1.94090143e-01
-1.45548508e-01 3.55115086e-01 3.73882473e-01 3.43753994e-01
1.48389220e+00 -3.31664473e-01 7.36246765e-01 -3.10878307e-01
6.40685439e-01 1.74429640e-02 3.12220603e-01 8.10282290e-01
-1.00302100e-01 5.41459560e-01 5.06451786e-01 -8.06200325e-01
-7.53785729e-01 -4.87099111e-01 -2.61147112e-01 5.60107112e-01
-2.82783240e-01 -6.75639570e-01 -8.31896663e-01 -9.28921103e-01
5.22001050e-02 9.11243021e-01 -1.06936169e+00 -2.79307336e-01
-4.14206982e-01 -7.09191620e-01 1.40690237e-01 5.82202375e-01
-3.44013572e-02 -1.36444163e+00 -1.44983828e+00 2.77544230e-01
-2.34172791e-02 -8.32588017e-01 -2.45690510e-01 2.30198622e-01
-6.81909204e-01 -1.23892283e+00 -9.54353988e-01 -5.67358673e-01
6.63426280e-01 -4.95008945e-01 1.24115682e+00 4.31931049e-01
-3.72374386e-01 3.05830568e-01 -2.80952811e-01 -4.14275348e-01
-6.22396410e-01 2.19977617e-01 -2.08644405e-01 -1.90428291e-02
4.57954824e-01 -6.93291351e-02 -9.77557719e-01 -1.57010972e-01
-7.44341254e-01 -1.42216131e-01 6.57939076e-01 7.94101894e-01
2.31800333e-01 -6.27233267e-01 1.63619161e-01 -1.01420319e+00
8.71119499e-01 -5.82241654e-01 -3.20353031e-01 2.29259849e-01
-1.04750824e+00 1.09671593e-01 5.22993326e-01 -2.64464706e-01
-5.98503351e-01 1.70925006e-01 -3.54653239e-01 -2.59000510e-01
-4.51695293e-01 6.25091791e-01 5.06481409e-01 4.28045243e-01
7.56531417e-01 -2.93132961e-01 -2.91881353e-01 -3.76408309e-01
7.34912977e-02 9.04571891e-01 4.52290505e-01 1.54239789e-01
9.54538137e-02 1.43047124e-01 -1.15674905e-01 -2.22945049e-01
-1.00276279e+00 -5.99841654e-01 -3.92939210e-01 1.07285328e-01
1.13537908e+00 -4.69454020e-01 -7.77745724e-01 -3.27478111e-01
-9.96337891e-01 -1.41475514e-01 -3.73824567e-01 7.19615757e-01
-1.95013359e-01 3.04447971e-02 -6.39724374e-01 -6.57697141e-01
-9.12538171e-01 -1.06364918e+00 8.88527095e-01 8.89574364e-02
-6.91204190e-01 -8.65501046e-01 1.68174252e-01 4.31400612e-02
5.38683653e-01 3.81624341e-01 9.62615967e-01 -1.49607003e+00
1.07848674e-01 -6.93584621e-01 -1.85314313e-01 1.28607079e-01
3.57158661e-01 3.70295905e-02 -8.19505751e-01 -9.63785350e-02
-3.04512549e-02 2.38673046e-01 8.11676204e-01 5.61285079e-01
9.99145865e-01 -4.09951955e-01 -5.79672754e-01 4.21335012e-01
1.22647703e+00 5.79625607e-01 4.13793594e-01 2.48097375e-01
3.97898495e-01 5.53075552e-01 4.96059716e-01 4.48508888e-01
2.98124343e-01 5.28653443e-01 5.53555191e-01 -3.10925961e-01
1.08691111e-01 1.85493484e-01 -2.09124625e-01 3.63036633e-01
-2.09519625e-01 -4.28221881e-01 -1.23860121e+00 6.00655138e-01
-1.81105566e+00 -7.42231429e-01 2.85754818e-02 2.47380090e+00
5.98868728e-01 2.94893473e-01 9.08431336e-02 1.20396964e-01
6.26437545e-01 -4.03749496e-01 -3.27395260e-01 -7.36440420e-01
4.43075657e-01 2.51657963e-01 6.86483502e-01 4.22907591e-01
-1.19034183e+00 3.52351636e-01 6.72425556e+00 7.43495822e-02
-1.31471264e+00 -1.30544528e-01 8.44151795e-01 -7.87982494e-02
7.10592642e-02 -4.65264440e-01 -6.43314123e-01 6.29190683e-01
1.39045000e+00 1.80045646e-02 -9.03982446e-02 7.42671013e-01
1.25438944e-01 4.13469039e-02 -1.16219306e+00 7.61562169e-01
2.48466194e-01 -1.42968202e+00 -6.95594698e-02 -1.21876532e-02
2.90330768e-01 6.06998764e-02 2.45773466e-03 1.65315166e-01
1.62617728e-01 -9.48103309e-01 3.14842612e-01 7.49985099e-01
9.10294116e-01 -4.54534203e-01 1.13427365e+00 1.21287152e-01
-8.30663800e-01 -2.45146066e-01 1.90443277e-01 1.66874990e-01
1.40091861e-02 3.27093095e-01 -1.26544178e+00 5.39655387e-01
8.26748490e-01 6.02680981e-01 -4.93570030e-01 1.12395144e+00
-1.33777648e-01 4.73958552e-01 -2.50489205e-01 -3.46448552e-03
3.17753762e-01 1.80841848e-01 4.39589024e-01 1.43118525e+00
5.07515550e-01 1.95785582e-01 1.49839312e-01 4.49964494e-01
-6.26678718e-03 5.93864679e-01 -5.07532775e-01 -2.69742124e-02
9.21273306e-02 1.56483090e+00 -8.65017354e-01 -1.00643361e+00
-2.06016585e-01 8.75392973e-01 1.12239540e-01 7.12285563e-02
-5.46369076e-01 -5.71312487e-01 2.31086746e-01 4.11089242e-01
3.08944017e-01 5.38380623e-01 -1.63214445e-01 -9.62742865e-01
-2.36850321e-01 -8.08890045e-01 8.34096611e-01 -6.36349678e-01
-9.07242060e-01 1.05126941e+00 -4.22861397e-01 -1.29611659e+00
-6.90702200e-01 -7.38816381e-01 -5.17928123e-01 9.82226729e-01
-1.39884293e+00 -6.25666618e-01 -3.59024048e-01 2.66298413e-01
2.93196738e-01 -8.01789165e-02 1.39216506e+00 6.16481960e-01
-3.55042279e-01 4.11252558e-01 -1.58466578e-01 3.87597024e-01
8.76295924e-01 -1.45031703e+00 2.49597505e-01 2.09283292e-01
-2.38569394e-01 7.93294966e-01 7.69327104e-01 -5.79504132e-01
-6.58371568e-01 -1.02424836e+00 1.58321774e+00 -6.98592484e-01
3.49089801e-01 1.88473850e-01 -9.79490519e-01 4.59099710e-01
2.13820696e-01 -8.79282653e-02 1.02013409e+00 -4.88989390e-02
-5.02153821e-02 1.22247264e-01 -1.34716630e+00 5.15221894e-01
4.27564144e-01 -2.32312441e-01 -8.42417419e-01 6.23945534e-01
5.90255439e-01 -7.20421672e-01 -1.20681953e+00 4.91880119e-01
6.51910603e-01 -7.57923543e-01 7.48583496e-01 -9.42466557e-01
5.14633179e-01 6.86264858e-02 2.29380965e-01 -1.25772262e+00
-4.06865388e-01 -4.92585033e-01 -1.14784949e-01 5.99729598e-01
7.57685542e-01 -6.75293803e-01 3.36624056e-01 9.05920744e-01
1.66450404e-02 -1.07083225e+00 -5.85608840e-01 1.93589218e-02
-2.26777643e-01 8.37829411e-02 3.62059951e-01 9.13613975e-01
2.03537926e-01 3.91435087e-01 -9.66619700e-02 1.40884817e-01
-1.98315457e-01 1.33308724e-01 3.34272161e-02 -1.50789630e+00
-1.30993798e-01 -4.66091752e-01 -3.49589288e-01 -1.66564524e-01
-4.31116074e-01 -8.53286088e-01 -2.65012473e-01 -2.03966856e+00
7.69802779e-02 -2.23207861e-01 -9.42281246e-01 6.47016227e-01
-1.33264244e-01 1.57114744e-01 7.54780546e-02 1.49035662e-01
-4.69329059e-01 -2.51489103e-01 6.22393727e-01 5.82176596e-02
-5.12264669e-01 9.54119116e-02 -6.60819530e-01 8.04651737e-01
9.21048403e-01 -6.61792815e-01 5.75440191e-02 -1.86842099e-01
3.35760653e-01 2.01366678e-01 1.77563161e-01 -9.76020217e-01
1.02214582e-01 1.41338110e-01 4.48235154e-01 -3.74986172e-01
1.53865159e-01 -8.43273759e-01 -1.86345443e-01 1.08992434e+00
-8.16553056e-01 2.55499572e-01 3.21903050e-01 3.34102750e-01
-6.83439448e-02 -1.81804061e-01 5.16959488e-01 -2.20875755e-01
-7.54296482e-02 1.90929249e-01 -4.50041503e-01 -2.04435602e-01
8.88294160e-01 1.84902042e-01 6.29930350e-04 -2.34700799e-01
-1.13564587e+00 1.77181482e-01 5.03651425e-02 2.65484422e-01
3.26695770e-01 -9.82998669e-01 -7.49655008e-01 6.94812760e-02
3.25883478e-01 -4.66423869e-01 1.71413943e-01 1.02852583e+00
-7.01264262e-01 6.73572958e-01 -3.52292024e-02 -3.68409425e-01
-1.46348333e+00 9.58065629e-01 5.87800860e-01 -6.28852606e-01
-8.26312006e-01 4.25513595e-01 2.89018359e-02 -1.05706267e-01
2.85764992e-01 -7.43630111e-01 -8.77056479e-01 1.08329289e-01
8.80244434e-01 -5.44878431e-02 5.89851737e-01 -6.15290165e-01
-5.46463430e-01 3.69139820e-01 -1.80431709e-01 -5.29196784e-02
1.19689333e+00 1.62570089e-01 1.11384667e-01 1.70755774e-01
7.42520571e-01 1.27881765e-01 -5.93378425e-01 -7.55518675e-02
8.98102522e-02 -3.38571258e-02 2.20462650e-01 -1.35745621e+00
-8.77825081e-01 8.06955814e-01 9.11672115e-01 7.12406397e-01
1.10818994e+00 -2.18227252e-01 6.00534320e-01 4.32406068e-01
-5.45755565e-01 -9.27058399e-01 -5.00766873e-01 2.51185447e-01
7.08094299e-01 -1.25814855e+00 -1.04233809e-01 -1.24282859e-01
-6.94625258e-01 1.17842758e+00 1.37348667e-01 -1.97792694e-01
7.15235233e-01 1.76400855e-01 3.90726715e-01 -4.29994375e-01
-8.34103584e-01 -1.64941087e-01 6.93847954e-01 2.09476426e-01
6.67582870e-01 1.85145400e-02 -6.40295267e-01 7.64805913e-01
7.75011480e-02 2.43551716e-01 2.72892088e-01 9.69933629e-01
-2.65773892e-01 -8.59345138e-01 -2.09771007e-01 8.27276468e-01
-1.20745301e+00 -3.65902066e-01 -3.34807932e-01 5.02324820e-01
2.45378911e-01 9.01913762e-01 9.39687490e-02 -1.78257525e-01
6.12673759e-01 3.75797302e-01 3.69600430e-02 -9.17852044e-01
-1.27570391e+00 8.30937400e-02 3.33214968e-01 -5.29548407e-01
-2.48689637e-01 -3.87109041e-01 -1.24531507e+00 2.75814056e-01
-7.83310011e-02 4.17981297e-01 6.81446254e-01 9.44944620e-01
6.58183813e-01 1.01088870e+00 1.62669435e-01 -4.29342479e-01
-4.28397357e-01 -8.90989363e-01 -8.98547843e-02 5.55883169e-01
6.92561150e-01 -3.41532141e-01 -1.57129630e-01 2.50343271e-02] | [8.309685707092285, 8.006129264831543] |
9dc117a1-edf1-469a-b495-3f9042f70300 | population-diversity-leads-to-short-running | 2204.06461 | null | https://arxiv.org/abs/2204.06461v1 | https://arxiv.org/pdf/2204.06461v1.pdf | Population Diversity Leads to Short Running Times of Lexicase Selection | In this paper we investigate why the running time of lexicase parent selection is empirically much lower than its worst-case bound of O(N*C). We define a measure of population diversity and prove that high diversity leads to low running times O(N + C) of lexicase selection. We then show empirically that genetic programming populations evolved under lexicase selection are diverse for several program synthesis problems, and explore the resulting differences in running time bounds. | ['William La Cava', 'Johannes Lengler', 'Thomas Helmuth'] | 2022-04-13 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 4.68605548e-01 1.62932817e-02 -3.86514395e-01 -2.35844657e-01
-6.56849384e-01 -8.18710685e-01 2.21736133e-01 2.45992839e-01
-6.00348532e-01 9.36627269e-01 -2.18171895e-01 -5.10418952e-01
-2.00130582e-01 -9.70788062e-01 -7.26193011e-01 -8.33363533e-01
-3.05134445e-01 6.11428022e-01 4.60769683e-01 -1.84219688e-01
7.69167840e-01 6.37192428e-01 -2.02961230e+00 8.98499861e-02
1.10978234e+00 3.57263505e-01 -3.19724232e-01 1.36371791e+00
-1.88817173e-01 -5.98518318e-03 -1.10473669e+00 -2.35505030e-01
2.62748301e-01 -1.26088428e+00 -7.44894564e-01 -8.78098831e-02
3.13974351e-01 4.90471870e-01 2.14350834e-01 1.27350473e+00
5.83559275e-01 8.78554955e-02 4.89723295e-01 -1.40288973e+00
-5.46565294e-01 1.20630765e+00 -7.67458677e-01 5.24388790e-01
4.51795757e-01 4.60978001e-02 7.03673959e-01 -1.07411720e-01
6.42872989e-01 1.14196515e+00 5.58645904e-01 4.43249911e-01
-1.66794086e+00 -5.20939112e-01 -4.23897021e-02 -4.78929102e-01
-1.57155979e+00 -3.12535942e-01 9.04661119e-02 -1.14728130e-01
1.23317754e+00 7.44768023e-01 9.13731158e-01 5.17187357e-01
4.86937374e-01 5.07786989e-01 9.86371100e-01 -7.04792380e-01
4.72931415e-01 -1.59205034e-01 1.91653207e-01 8.76007557e-01
8.71922910e-01 3.86423022e-01 -3.72813642e-01 -6.90144598e-01
6.08569086e-01 -6.96182251e-01 -4.77721095e-01 -2.48310834e-01
-1.04167473e+00 9.72982943e-01 -2.59951562e-01 2.96623200e-01
-3.74155156e-02 5.34138978e-01 1.95808724e-01 8.37005794e-01
-1.33320928e-01 9.11259055e-01 -5.01942873e-01 -6.35409296e-01
-7.18598127e-01 3.90881777e-01 1.28315401e+00 1.34899199e+00
3.38506281e-01 1.50883138e-01 1.80322006e-01 6.81438923e-01
-2.93901086e-01 6.88024700e-01 4.47980493e-01 -1.08933163e+00
1.60016805e-01 1.30272880e-01 -7.97243863e-02 -6.66402698e-01
-2.60475367e-01 -4.14557457e-01 -3.94592702e-01 2.83590347e-01
4.64211822e-01 -4.13462579e-01 -3.53426784e-01 1.80399871e+00
1.41052261e-01 -5.14137924e-01 3.29585262e-02 1.95145741e-01
-1.22272279e-02 7.52103031e-01 -2.53860980e-01 -7.62415230e-01
8.14184546e-01 -7.88546026e-01 -2.35506967e-01 9.49799642e-02
8.63218188e-01 -6.76485002e-01 9.59515810e-01 5.15097916e-01
-1.51779389e+00 6.26738966e-02 -9.66042221e-01 6.25735462e-01
6.83223456e-02 -4.06493932e-01 7.48870015e-01 1.34324574e+00
-1.16891682e+00 5.77609539e-01 -7.81972587e-01 -5.68301201e-01
-5.10190893e-03 6.11515641e-01 8.74780193e-02 3.47798944e-01
-4.43891764e-01 4.58161712e-01 6.58791721e-01 -2.09124655e-01
-4.29669172e-01 -4.60180789e-01 -3.75321925e-01 1.58955127e-01
3.91464680e-01 -6.69856787e-01 9.25028086e-01 -7.00007558e-01
-1.40397596e+00 8.13129544e-01 -1.18268147e-01 -3.24843556e-01
4.53247637e-01 3.01599294e-01 -1.07306309e-01 -1.04441270e-01
-3.52169067e-01 3.46131861e-01 4.03709441e-01 -1.22101521e+00
-1.14277577e+00 -2.16192469e-01 1.25391725e-02 -7.96173699e-03
-2.64374971e-01 3.45590532e-01 -1.58821777e-01 -3.86981338e-01
6.69563338e-02 -1.26899791e+00 -5.63100338e-01 -4.11156863e-01
-1.25439927e-01 -1.50313288e-01 -2.94871852e-02 7.18198419e-02
1.43830693e+00 -2.05689049e+00 3.61849874e-01 5.34706593e-01
-4.28522341e-02 -6.04896024e-02 -3.37365538e-01 5.04982471e-01
8.30222517e-02 5.15004635e-01 -2.92035460e-01 7.10649729e-01
7.03723431e-02 1.65446222e-01 5.20272937e-04 5.42634308e-01
-1.17000721e-01 5.94837725e-01 -9.32651341e-01 -5.12848616e-01
-6.85515761e-01 1.07272156e-01 -1.01379859e+00 -2.14979872e-01
-1.88510031e-01 -3.61789823e-01 -3.06326002e-01 5.77430189e-01
5.58010221e-01 -9.69737247e-02 4.88467425e-01 3.63294244e-01
-2.94949740e-01 4.83185239e-03 -1.03742671e+00 1.04154897e+00
-3.22687566e-01 6.32581770e-01 -1.86086923e-01 -4.68820959e-01
9.16212440e-01 -1.23357520e-01 2.34071985e-01 -4.64877099e-01
3.58811915e-01 5.33339143e-01 4.44700748e-01 -4.02279913e-01
4.30189520e-01 -3.72552685e-02 -2.09608957e-01 8.10523927e-01
-5.40269494e-01 -4.48154479e-01 6.05058134e-01 -2.47614741e-01
1.43679452e+00 -2.54969478e-01 3.03986907e-01 -6.61625504e-01
4.68650401e-01 4.54314411e-01 6.17410064e-01 1.17637074e+00
-7.22583011e-02 2.39019662e-01 1.14594567e+00 -1.72911450e-01
-1.21473873e+00 -6.36375010e-01 -1.73362181e-01 1.28134489e+00
1.85371816e-01 -4.22885984e-01 -9.30881977e-01 -4.42069978e-01
1.27201341e-02 8.38473976e-01 -7.05699980e-01 -3.73728156e-01
-9.86294091e-01 -1.28530788e+00 8.81534040e-01 4.76909727e-01
9.22417790e-02 -9.32801127e-01 -1.21082699e+00 1.89595088e-01
4.03675437e-01 -3.38379055e-01 -4.84202325e-01 7.09047794e-01
-1.14395952e+00 -8.82706165e-01 -5.41989744e-01 -9.27782774e-01
8.05288076e-01 2.90304516e-02 1.34441078e+00 5.66697240e-01
-3.14297944e-01 6.01265356e-02 -4.22278047e-01 -2.29990765e-01
-7.52925277e-01 4.86285865e-01 -2.61582434e-01 -1.12677646e+00
4.35314864e-01 -4.66763228e-01 -4.59260307e-02 1.57260031e-01
-7.53449142e-01 -4.57041264e-01 4.77937222e-01 1.13159907e+00
5.69464505e-01 6.90216005e-01 8.10317695e-02 -9.79021966e-01
1.01412737e+00 -2.71123081e-01 -1.19443393e+00 5.38301349e-01
-8.80138934e-01 5.45056283e-01 5.31780720e-01 -7.83240855e-01
-6.69364512e-01 4.93098125e-02 9.48536769e-02 3.74377578e-01
2.16341779e-01 4.75582987e-01 8.13078657e-02 -4.94846642e-01
8.11921954e-01 1.39781311e-01 -1.11336350e-01 -7.40516335e-02
-1.96279511e-01 2.69629240e-01 3.51871192e-01 -1.08466852e+00
6.41810238e-01 4.86178417e-03 2.14570984e-01 -6.39646113e-01
-3.43131989e-01 3.15723985e-01 -1.98296696e-01 4.09353435e-01
4.65820193e-01 -1.52688637e-01 -6.65418506e-01 2.82804761e-02
-8.74311686e-01 -4.58071649e-01 -3.82376522e-01 2.43302003e-01
-7.67125964e-01 -6.14879057e-02 -3.92966062e-01 -8.55958283e-01
-1.81362219e-02 -1.03704941e+00 7.09098876e-01 3.49124134e-01
-2.49261513e-01 -6.36843264e-01 3.77872825e-01 -2.50347704e-01
2.24285081e-01 4.07776088e-01 1.44168019e+00 -4.90331709e-01
-5.17558277e-01 7.35485181e-02 7.42414966e-02 -4.75322276e-01
-3.63636464e-01 4.69666034e-01 -1.16964713e-01 -4.57834810e-01
-1.42994344e-01 9.78476703e-02 5.41655779e-01 2.57566065e-01
9.89470780e-01 -2.36064196e-01 -5.77079177e-01 9.92989182e-01
1.83549643e+00 9.09266651e-01 6.35797918e-01 7.19110787e-01
2.06245169e-01 4.69716758e-01 7.84501255e-01 5.69027066e-01
-2.39732698e-01 3.64312768e-01 -8.34366605e-02 2.52682686e-01
2.96565115e-01 1.83053583e-01 3.65108192e-01 7.63756275e-01
-1.28154695e-01 -8.53340089e-01 -1.31346977e+00 7.08034039e-01
-1.68371594e+00 -9.31735456e-01 -1.78733677e-01 2.19651008e+00
1.12987983e+00 3.22947770e-01 5.92820942e-01 1.81995720e-01
9.11983013e-01 -2.96186209e-01 -4.32928503e-01 -1.15271640e+00
-2.87242532e-01 4.46280092e-01 9.88322735e-01 6.60142899e-01
-4.44392741e-01 8.51183832e-01 8.59788513e+00 5.66922486e-01
-8.33293617e-01 -2.14495569e-01 3.80301267e-01 -4.53060389e-01
-5.71684062e-01 -5.26230969e-02 -8.37516308e-01 5.34816802e-01
1.32674527e+00 -9.36665297e-01 5.16912878e-01 7.61119246e-01
-4.63321239e-01 -1.85759231e-01 -1.25681281e+00 7.16618538e-01
1.26946703e-01 -1.15493429e+00 -6.61847293e-02 4.14399266e-01
1.22518122e+00 -3.49612743e-01 -1.46021415e-03 1.02387983e-02
5.89405358e-01 -1.11260521e+00 5.26734173e-01 -1.36852816e-01
5.39636374e-01 -1.52446818e+00 6.40914559e-01 2.03272834e-01
-8.64411950e-01 -2.94255316e-01 -5.90039551e-01 3.36521655e-01
1.96282372e-01 3.08620185e-01 -4.49281335e-01 -8.74049515e-02
5.24328113e-01 -2.53357351e-01 -6.16195321e-01 1.49018216e+00
1.18118851e-03 3.60912234e-01 -5.43003917e-01 -7.67283201e-01
2.49714881e-01 -2.21064940e-01 4.90334183e-01 1.23203242e+00
6.48087680e-01 4.00118679e-01 -1.17091887e-01 7.02836990e-01
2.39088759e-01 -1.64570585e-02 -3.67029935e-01 -5.26193917e-01
7.75622964e-01 3.53229076e-01 -1.13607311e+00 -2.19954744e-01
1.07505664e-01 8.97766113e-01 1.18029155e-01 9.74265784e-02
-9.96934116e-01 -1.18051457e+00 4.96884912e-01 -1.65556669e-01
5.95069468e-01 -2.50968039e-01 -3.38261604e-01 -8.01799655e-01
-1.41872942e-01 -1.29009581e+00 2.44919851e-01 -1.17722429e-01
-8.77513826e-01 8.98200333e-01 2.18629241e-01 -5.80091655e-01
-3.51358324e-01 -5.13111711e-01 -3.59009773e-01 6.48159325e-01
-7.33951449e-01 -9.16835219e-02 -7.09075853e-02 -1.18944332e-01
1.60808742e-01 -2.65860558e-01 7.04795599e-01 3.80730629e-02
-6.57612503e-01 1.02736175e+00 3.86955380e-01 -5.12442648e-01
1.88823983e-01 -1.35616994e+00 4.79929775e-01 8.36393476e-01
-2.72989094e-01 8.49939823e-01 1.04578769e+00 -6.37158811e-01
-1.70512688e+00 -4.37267095e-01 9.23743844e-01 -2.14456990e-01
6.31911516e-01 -2.28710100e-02 -6.25904560e-01 3.70367795e-01
1.78861037e-01 -3.84494811e-01 7.32613206e-01 2.36049816e-01
-2.89236486e-01 2.64234900e-01 -1.23495090e+00 9.68860328e-01
1.60309887e+00 -2.88505316e-01 -2.06965417e-01 2.18955219e-01
6.57604039e-01 -3.45422894e-01 -9.42513824e-01 1.69293448e-01
6.98036611e-01 -1.17716515e+00 6.92236304e-01 -5.27521789e-01
3.22286934e-01 -2.43615508e-01 -5.73004335e-02 -1.13601351e+00
-2.14310899e-01 -9.81836438e-01 8.12180042e-02 9.59737301e-01
7.31843829e-01 -9.08426166e-01 1.10281909e+00 2.74382919e-01
2.45849460e-01 -7.85874963e-01 -5.20827234e-01 -1.19722712e+00
6.00598514e-01 3.69416028e-01 1.06733370e+00 8.62520874e-01
4.33804505e-02 9.39282402e-03 1.61982715e-01 -8.33733380e-03
5.95136166e-01 6.23765707e-01 5.17896235e-01 -9.99178350e-01
-7.45494545e-01 -9.49211180e-01 -4.77373868e-01 -7.89774597e-01
1.66465268e-01 -6.39817238e-01 2.72184640e-01 -8.72067988e-01
3.65223914e-01 -7.53036737e-01 -1.64664090e-01 9.32854265e-02
-1.89929321e-01 8.18234757e-02 5.17919175e-02 -2.35072613e-01
-3.29576522e-01 2.51653716e-02 8.64639759e-01 3.25550921e-02
-5.82437098e-01 -3.38590920e-01 -7.21417964e-01 4.43497211e-01
8.65507782e-01 -9.40886974e-01 -4.92639065e-01 -6.81015193e-01
7.92128205e-01 2.34946266e-01 -4.86282915e-01 -1.08053172e+00
1.40688971e-01 -6.94235384e-01 -6.98568597e-02 -5.04823446e-01
-4.04237896e-01 -5.74801624e-01 5.09967089e-01 1.00218737e+00
-8.83966863e-01 7.47112572e-01 1.94217354e-01 3.10153335e-01
6.77102879e-02 -8.89064014e-01 9.04328048e-01 -7.41631538e-02
-1.98086992e-01 -4.99574617e-02 -6.33701146e-01 2.35527813e-01
1.29567969e+00 -5.91495693e-01 -4.77450550e-01 4.22055721e-02
-3.90922844e-01 7.89641216e-02 1.19658244e+00 -2.21420333e-01
1.25071421e-01 -9.62085128e-01 -5.01907706e-01 1.17663413e-01
1.49165355e-02 -5.10285974e-01 -5.48829556e-01 6.81366980e-01
-1.26847494e+00 2.01439276e-01 -4.01826829e-01 -5.62317789e-01
-1.69203436e+00 6.22959435e-01 2.14497760e-01 1.16178542e-01
-4.16054606e-01 1.40433991e+00 -2.26689905e-01 -2.61638403e-01
4.11866128e-01 -1.61261912e-02 3.68912309e-01 -1.33870304e-01
4.40908372e-01 5.77081919e-01 -3.63665640e-01 -2.37375930e-01
-4.51613516e-01 7.15580463e-01 5.31932265e-02 -4.92020488e-01
1.34497750e+00 1.90563619e-01 -4.83869046e-01 2.32280552e-01
1.19811738e+00 4.20750529e-01 -5.56082487e-01 5.31146526e-01
3.79389733e-01 -9.34492052e-01 -2.42896125e-01 -4.57608163e-01
-9.88957882e-01 3.36782157e-01 6.66409135e-01 5.35482168e-01
1.42231703e+00 -1.94364816e-01 6.92667961e-01 8.90276492e-01
7.58986771e-01 -9.96244669e-01 -3.65921050e-01 4.70199168e-01
5.58197737e-01 -6.31751895e-01 3.30160528e-01 -7.82873988e-01
-3.78651947e-01 1.09208620e+00 6.73143268e-01 -2.90694356e-01
2.38110244e-01 9.06399965e-01 -2.74510980e-01 -2.44227201e-02
-1.15341878e+00 -5.09931296e-02 -3.31635356e-01 7.27833390e-01
6.04984224e-01 -1.02034621e-01 -9.61560488e-01 -2.28491910e-02
-8.12661469e-01 -1.07422657e-01 8.29405069e-01 1.45635772e+00
-7.70395935e-01 -1.37445080e+00 -4.99029100e-01 2.22178221e-01
-3.97147775e-01 -8.47805664e-02 -7.74168909e-01 8.45817983e-01
9.41121578e-02 7.42495298e-01 2.03236565e-01 -1.61720753e-01
9.80683491e-02 -1.97009012e-01 1.10060620e+00 -4.76710379e-01
-1.11200309e+00 1.48006484e-01 4.21323061e-01 -3.86025697e-01
-2.42064163e-01 -7.05059171e-01 -1.29472113e+00 -9.01969492e-01
-5.57427824e-01 5.90933084e-01 4.81415659e-01 2.52449721e-01
2.65132815e-01 3.46390635e-01 4.43765491e-01 -1.47373572e-01
-6.14036202e-01 1.20146967e-01 -5.27275980e-01 -1.64015293e-01
2.38953857e-03 -1.63266629e-01 -6.39928102e-01 -1.31287590e-01] | [7.989263534545898, 7.147768974304199] |
874d19ba-e9f8-49ca-9acc-5462faad523d | robust-video-object-tracking-via-bayesian | 1604.00475 | null | http://arxiv.org/abs/1604.00475v3 | http://arxiv.org/pdf/1604.00475v3.pdf | Robust video object tracking via Bayesian model averaging based feature fusion | In this article, we are concerned with tracking an object of interest in
video stream. We propose an algorithm that is robust against occlusion, the
presence of confusing colors, abrupt changes in the object feature space and
changes in object size. We develop the algorithm within a Bayesian modeling
framework. The state space model is used for capturing the temporal correlation
in the sequence of frame images by modeling the underlying dynamics of the
tracking system. The Bayesian model averaging (BMA) strategy is proposed for
fusing multi-clue information in the observations. Any number of object
features are allowed to be involved in the proposed framework. Every feature
represents one source of information to be fused and is associated with an
observation model. The state inference is performed by employing the particle
filter methods. In comparison with related approaches, the BMA based tracker is
shown to have robustness, expressivity, and comprehensibility. | ['Bin Liu', 'Yi Dai'] | 2016-04-02 | null | null | null | null | ['video-object-tracking'] | ['computer-vision'] | [ 1.48720190e-01 -6.48947954e-01 1.14785545e-01 -9.13720950e-02
-1.77510321e-01 -3.90419871e-01 7.91954279e-01 1.13004580e-01
-4.11191016e-01 6.68360293e-01 -1.96526051e-01 2.45321751e-01
-4.13778335e-01 -3.32787365e-01 -5.69933712e-01 -9.40401971e-01
-2.06385151e-01 7.77007714e-02 6.01136029e-01 3.55225265e-01
2.01933697e-01 8.41793537e-01 -1.74218559e+00 -1.87991768e-01
4.86342311e-01 1.01143968e+00 5.28847098e-01 1.11637104e+00
-3.33901728e-03 9.28700387e-01 -7.96149015e-01 3.56699601e-02
2.60659605e-01 -3.08208197e-01 -1.72374427e-01 5.80112398e-01
3.89494270e-01 -2.09911361e-01 -1.18930429e-01 1.23102868e+00
2.17248082e-01 2.59076029e-01 6.52319074e-01 -1.36157238e+00
2.40144413e-02 -9.33171809e-02 -6.10158801e-01 5.08685052e-01
3.75110030e-01 -7.44918734e-02 3.38427126e-01 -7.16284275e-01
4.80260521e-01 1.57875574e+00 3.58562350e-01 4.04065162e-01
-1.01468050e+00 -3.58300894e-01 3.93040091e-01 3.63574266e-01
-1.21610141e+00 -2.57608622e-01 7.79336214e-01 -8.18185508e-01
2.78414577e-01 3.74426901e-01 8.41970205e-01 6.70063078e-01
7.64770567e-01 7.25109398e-01 1.06863487e+00 -5.44926584e-01
3.39244485e-01 3.09111569e-02 4.12429392e-01 6.28971815e-01
7.19638467e-01 4.72449183e-01 -7.37776339e-01 -3.60045314e-01
7.80540049e-01 2.89304942e-01 -1.18878804e-01 -4.79360014e-01
-1.03105485e+00 4.56852317e-01 -2.93429732e-01 9.12390500e-02
-5.67275405e-01 3.95638317e-01 8.64080861e-02 1.34487092e-01
4.19179738e-01 -1.17541440e-01 -1.32601514e-01 -1.06151506e-01
-7.84927726e-01 4.05821770e-01 7.38804996e-01 1.06125069e+00
3.55423957e-01 1.35013685e-01 -3.18104386e-01 3.11440546e-02
8.42411101e-01 9.36993778e-01 1.39002457e-01 -1.16326821e+00
-1.74330339e-01 1.45770699e-01 7.30054557e-01 -8.66811335e-01
-9.47766453e-02 -4.37466741e-01 -3.72110099e-01 7.24354684e-01
3.56306940e-01 -8.53026956e-02 -6.05971754e-01 1.57546687e+00
7.98594296e-01 6.03814542e-01 -9.62767191e-03 4.72240925e-01
4.81671095e-01 6.48461998e-01 -3.40191312e-02 -9.33028162e-01
1.45423567e+00 -5.60671270e-01 -1.38106990e+00 3.38733196e-01
-3.18264693e-01 -1.16477418e+00 -2.67555332e-03 6.61159217e-01
-1.06163692e+00 -9.56715941e-01 -9.42254484e-01 6.30957067e-01
-2.34319597e-01 9.06949863e-02 -1.67363100e-02 7.22256601e-01
-7.68562376e-01 6.06990397e-01 -1.20326233e+00 -4.91500050e-01
-7.40109533e-02 4.63410109e-01 7.33184069e-02 8.82741660e-02
-5.10142565e-01 1.10347009e+00 4.56838727e-01 3.14032853e-01
-9.96772945e-01 -3.52745235e-01 -4.85349238e-01 -1.95771545e-01
3.03982943e-01 -6.07843339e-01 1.20321679e+00 -8.07785213e-01
-1.52584207e+00 1.22881792e-01 -5.16805530e-01 -3.03143650e-01
6.39134347e-01 -5.71212769e-01 -6.08621180e-01 2.09203735e-01
-2.86262244e-01 1.08296327e-01 1.54477203e+00 -1.33816814e+00
-1.01382005e+00 -1.93862185e-01 -2.04876378e-01 1.55930921e-01
2.20894486e-01 3.26850682e-01 -2.87788898e-01 -6.30841732e-01
1.42592594e-01 -8.43012094e-01 -2.00585146e-02 1.14664838e-01
2.94840746e-02 -8.19897205e-02 1.42526269e+00 -3.65989715e-01
1.10359287e+00 -2.22081947e+00 -4.12710421e-02 6.07856996e-02
6.40122294e-02 2.42429152e-01 2.19295248e-01 4.27244812e-01
2.95973748e-01 -5.34178555e-01 2.32668504e-01 -3.94131362e-01
-1.46816507e-01 3.71572763e-01 -2.31032893e-01 8.41229141e-01
-1.96600929e-02 2.63549894e-01 -8.55450392e-01 -7.38839388e-01
6.15690708e-01 6.74012363e-01 -5.24050407e-02 4.27384526e-01
-2.67928056e-02 4.94293630e-01 -6.30738497e-01 4.35145289e-01
9.26244676e-01 -3.50346714e-02 -4.71673831e-02 -3.82459879e-01
-4.78287786e-01 -3.70644659e-01 -1.74973559e+00 1.25959039e+00
2.77531557e-02 5.47526479e-01 2.23947853e-01 -5.38373232e-01
7.54089892e-01 5.89784861e-01 6.86875999e-01 5.72460890e-02
2.63865232e-01 -1.36885867e-01 7.96336457e-02 -6.06124282e-01
5.53547978e-01 -1.33020207e-02 3.92090440e-01 3.28108966e-01
1.58783525e-01 1.59815237e-01 4.94433641e-01 1.08285509e-01
8.77729416e-01 3.60767931e-01 4.46051180e-01 -3.60191435e-01
6.22095704e-01 -2.37035319e-01 8.31565261e-01 9.81756568e-01
-3.83916110e-01 -1.83800105e-02 -1.36339933e-01 -3.69218081e-01
-7.81050444e-01 -1.03051603e+00 -2.45087832e-01 5.08307636e-01
4.47803587e-01 -8.82812142e-02 -4.80888665e-01 -1.75786644e-01
-2.64307838e-02 3.51963013e-01 -5.77458382e-01 -1.23764999e-01
-2.96684384e-01 -7.25665033e-01 -1.92820430e-01 2.86391795e-01
2.58830905e-01 -7.00215220e-01 -1.26848757e+00 6.67383671e-01
1.00443617e-01 -1.16061103e+00 -3.36626798e-01 -8.06265920e-02
-1.04099941e+00 -1.15533674e+00 -3.66563648e-01 -3.89885679e-02
5.28244853e-01 3.12342465e-01 8.55135977e-01 -2.02022642e-01
-4.28547978e-01 1.06304216e+00 -3.70196730e-01 -8.12201142e-01
-5.66089332e-01 -9.27589357e-01 3.34034860e-01 5.49591601e-01
2.65419513e-01 -1.37234805e-03 -4.59354997e-01 1.49655700e-01
-8.15751135e-01 -1.23266585e-01 2.82614022e-01 6.54677391e-01
3.77583712e-01 3.47544193e-01 -1.24489134e-02 -3.53051245e-01
3.13299000e-01 -1.08065978e-01 -1.12641001e+00 3.78163457e-01
-3.48876983e-01 -1.05286226e-01 4.33700457e-02 -7.54134417e-01
-1.40305674e+00 2.28725180e-01 6.92769527e-01 -5.19377649e-01
-3.06656927e-01 8.62639919e-02 -2.07154050e-01 -3.18144798e-01
1.10035695e-01 1.51547834e-01 2.44046152e-01 -5.15994966e-01
2.01476470e-01 2.37952664e-01 4.09723252e-01 -6.18373513e-01
8.09803665e-01 8.03386927e-01 5.89547932e-01 -8.74161780e-01
-6.05206609e-01 -7.15834081e-01 -7.06211507e-01 -8.26106369e-01
8.50886226e-01 -7.61347592e-01 -1.04814398e+00 7.43342698e-01
-1.33239031e+00 3.01575422e-01 -3.42769802e-01 1.09568894e+00
-5.65326095e-01 4.92779583e-01 -3.90573531e-01 -1.54523480e+00
1.61856636e-01 -1.07731318e+00 7.24271297e-01 4.78194594e-01
3.00796442e-02 -1.07033765e+00 2.89445490e-01 -1.43246993e-01
2.09844097e-01 3.22456330e-01 1.57109663e-01 -5.36276102e-01
-9.71616924e-01 -4.25454140e-01 2.73764521e-01 2.92978644e-01
3.11311543e-01 5.42427599e-01 -9.27207947e-01 -5.35292566e-01
5.56515574e-01 4.53283846e-01 4.15540725e-01 8.22783589e-01
5.62930703e-01 -1.17842168e-01 -3.32930177e-01 -9.01001226e-03
1.48116589e+00 6.86206281e-01 -1.04474220e-02 1.17143951e-01
4.01546776e-01 4.34780419e-01 1.03628755e+00 8.54492784e-01
1.45479129e-03 8.09705496e-01 6.23319387e-01 1.06106982e-01
1.67543702e-02 2.79835403e-01 5.26385427e-01 6.19844496e-01
-1.62212312e-01 -4.26946610e-01 -3.76973599e-01 3.14506024e-01
-2.08582306e+00 -1.30064940e+00 -4.75864381e-01 2.34862494e+00
1.42859727e-01 -4.88035614e-03 1.25479832e-01 -5.65764867e-03
1.07305467e+00 -1.91360731e-02 -4.16830242e-01 4.67053093e-02
3.51488069e-02 -2.53051579e-01 5.11249185e-01 6.73136234e-01
-1.12050378e+00 2.91300386e-01 6.86552715e+00 5.46822131e-01
-6.27016246e-01 1.73961863e-01 -2.36021027e-01 5.66177443e-03
3.67505103e-01 1.97592318e-01 -1.12249351e+00 6.92605853e-01
7.95730352e-01 -1.75737321e-01 4.45779637e-02 3.72205406e-01
5.94563961e-01 -7.66579866e-01 -9.19566572e-01 8.29010010e-01
2.15871528e-01 -8.09206009e-01 -2.01283139e-03 1.69957921e-01
5.64999759e-01 -3.75796109e-01 1.44154742e-01 -3.82393479e-01
4.28799957e-01 -3.38884622e-01 8.76303852e-01 1.22318530e+00
-6.50165528e-02 -5.11118710e-01 5.75299442e-01 4.01900470e-01
-1.37699544e+00 -1.88917547e-01 -2.25652874e-01 -1.58741102e-01
4.76208091e-01 5.37365377e-01 -4.26406652e-01 7.63067544e-01
6.08869910e-01 7.60401845e-01 -3.08735758e-01 1.56264639e+00
1.52622133e-01 3.81716430e-01 -5.33029437e-01 9.30376537e-03
-1.17082424e-01 -2.84616560e-01 9.89687204e-01 9.45173264e-01
4.98919785e-01 9.64279473e-02 4.96327847e-01 5.51089048e-01
6.92866027e-01 -1.71329349e-01 -3.49964023e-01 3.06532919e-01
4.77340788e-01 1.11121058e+00 -7.52476096e-01 -7.76310682e-01
-6.65472567e-01 6.43740475e-01 -4.27146494e-01 4.41286653e-01
-7.70980299e-01 1.20217010e-01 3.73991966e-01 -2.19383150e-01
6.08070135e-01 -2.73269862e-01 3.13979238e-01 -1.12104332e+00
-1.66301936e-01 -5.61900973e-01 3.89359236e-01 -8.16114545e-01
-1.08418703e+00 3.45710546e-01 4.95164752e-01 -1.63625586e+00
-1.44334778e-01 -5.76398432e-01 -6.39331698e-01 9.35825229e-01
-1.42750180e+00 -1.00320637e+00 -1.91677436e-01 7.94519126e-01
6.67300344e-01 -2.64355928e-01 3.72737885e-01 2.58094333e-02
-4.35325593e-01 -3.39442760e-01 3.88743699e-01 -2.66842306e-01
5.40187120e-01 -1.08876359e+00 -2.86451280e-01 1.27574015e+00
1.49171241e-02 6.65435553e-01 1.43752491e+00 -8.83615017e-01
-1.44042337e+00 -8.37901175e-01 5.75428724e-01 -4.41495568e-01
7.23869085e-01 -1.98728708e-03 -7.99381912e-01 5.52925169e-01
1.87160596e-01 2.74985552e-01 5.34745753e-01 -3.24761510e-01
1.69436708e-01 -1.58136964e-01 -1.03154492e+00 -1.18036065e-02
3.07709783e-01 -2.19044089e-01 -8.04141343e-01 -2.83100586e-02
9.09705013e-02 -3.42142969e-01 -9.03182447e-01 3.71212274e-01
7.63894260e-01 -8.29300761e-01 8.52425575e-01 -4.42635357e-01
-6.55093908e-01 -9.76641476e-01 -1.49277136e-01 -8.54693294e-01
-5.74361920e-01 -7.66388774e-01 -6.45537972e-01 1.16860294e+00
-2.54804552e-01 -3.92039955e-01 3.46765310e-01 4.59410220e-01
2.89474249e-01 3.42369117e-02 -1.09290814e+00 -1.19259107e+00
-5.52316904e-01 -2.23945618e-01 2.73142129e-01 2.82168865e-01
-1.62071526e-01 -1.01694502e-01 -7.08515346e-01 7.18561769e-01
1.42408681e+00 2.01463297e-01 8.69194269e-01 -1.56728685e+00
-4.03427631e-01 -4.22343612e-02 -6.19991004e-01 -7.90460825e-01
-9.68011990e-02 3.45247658e-03 1.47213846e-01 -1.16734970e+00
4.77228910e-01 1.58727974e-01 -4.72022623e-01 -3.88033539e-01
-2.13156387e-01 -2.24346071e-01 6.03904784e-01 2.23165259e-01
-7.96152711e-01 3.37923616e-01 1.03506172e+00 1.33504003e-01
1.56219006e-01 4.45973605e-01 3.18210393e-01 9.15410995e-01
5.27761281e-01 -7.62774527e-01 -2.58212626e-01 -6.61656708e-02
-2.22834811e-01 3.02735806e-01 6.32047176e-01 -1.14144170e+00
4.73448217e-01 -2.20080107e-01 4.97632176e-01 -1.15408480e+00
6.63851202e-01 -1.43203855e+00 6.59335971e-01 9.04014647e-01
-5.38016576e-03 2.95528442e-01 3.77261490e-02 1.08560014e+00
-1.92661643e-01 -5.49444497e-01 8.09689581e-01 -8.94699916e-02
-7.12074816e-01 3.67357492e-01 -7.76969075e-01 -5.72534561e-01
1.16547918e+00 -4.05661166e-01 -2.66156215e-02 -2.41202533e-01
-1.11545813e+00 6.09937706e-04 3.31767082e-01 2.38237977e-01
3.34126979e-01 -1.28993082e+00 -4.12791222e-01 2.17282660e-02
-1.06202595e-01 -5.25930166e-01 4.35000658e-01 8.74770403e-01
-9.63873491e-02 2.14243218e-01 -2.24810958e-01 -8.83521676e-01
-1.74602282e+00 8.23871136e-01 2.03336626e-01 -1.42771497e-01
-3.30361992e-01 3.11508447e-01 6.76724911e-02 6.24773979e-01
3.32837552e-01 -2.89484352e-01 -4.39932227e-01 4.90903854e-02
7.47612357e-01 7.92143583e-01 -4.83236492e-01 -8.91043067e-01
-2.43003115e-01 8.72917771e-01 4.89505269e-02 -3.14092964e-01
9.24700081e-01 -6.11084044e-01 -1.40517242e-02 1.02884626e+00
6.40498281e-01 -1.02632761e-01 -1.77546036e+00 -3.81400883e-01
-7.90917650e-02 -6.64592862e-01 9.78637561e-02 -5.63993931e-01
-6.44932926e-01 6.45667970e-01 1.20323193e+00 3.39087188e-01
9.27712262e-01 -2.49914154e-01 -3.55179934e-03 6.95547508e-03
3.83881718e-01 -1.01968861e+00 -3.39261666e-02 3.25073451e-01
4.61998582e-01 -1.11762536e+00 4.10827339e-01 -3.01790386e-01
-2.02278674e-01 1.20033956e+00 3.30474049e-01 -2.36400366e-01
9.78976548e-01 6.06798768e-01 -2.77733486e-02 -1.12606712e-01
-1.01197112e+00 -4.76547241e-01 3.55901748e-01 5.99282920e-01
4.74215522e-02 -7.29534626e-02 -4.64684725e-01 -2.84320805e-02
5.39816082e-01 1.69398040e-01 5.52618444e-01 1.29433429e+00
-8.31765950e-01 -1.02280176e+00 -1.13039708e+00 -4.64194119e-02
-6.37831032e-01 4.29002821e-01 3.03095788e-01 6.69693589e-01
2.03070655e-01 1.17040312e+00 2.44130984e-01 2.53809690e-01
1.92564055e-01 -6.69420436e-02 7.73507118e-01 -2.49602452e-01
-1.63576186e-01 7.79661298e-01 -3.21102232e-01 -5.49670279e-01
-1.15629518e+00 -1.24263072e+00 -7.99283981e-01 2.89030988e-02
-6.41723692e-01 2.69198269e-01 7.66413510e-01 9.41564381e-01
1.22032622e-02 6.02300823e-01 5.30464947e-01 -7.74108291e-01
-4.17534739e-01 -9.43312466e-01 -8.84392917e-01 1.49306223e-01
6.46129489e-01 -1.13099682e+00 -3.05222958e-01 5.29341936e-01] | [6.644790172576904, -1.989187479019165] |
1439c0e6-c31a-4681-83d6-a1b50471ccc5 | 3d-regnet-deep-learning-model-for-covid-19 | 2107.04055 | null | https://arxiv.org/abs/2107.04055v1 | https://arxiv.org/pdf/2107.04055v1.pdf | 3D RegNet: Deep Learning Model for COVID-19 Diagnosis on Chest CT Image | In this paper, a 3D-RegNet-based neural network is proposed for diagnosing the physical condition of patients with coronavirus (Covid-19) infection. In the application of clinical medicine, lung CT images are utilized by practitioners to determine whether a patient is infected with coronavirus. However, there are some laybacks can be considered regarding to this diagnostic method, such as time consuming and low accuracy. As a relatively large organ of human body, important spatial features would be lost if the lungs were diagnosed utilizing two dimensional slice image. Therefore, in this paper, a deep learning model with 3D image was designed. The 3D image as input data was comprised of two-dimensional pulmonary image sequence and from which relevant coronavirus infection 3D features were extracted and classified. The results show that the test set of the 3D model, the result: f1 score of 0.8379 and AUC value of 0.8807 have been achieved. | ['Xinyu Liu', 'YuHan Wang', 'Haibo Qi'] | 2021-07-08 | null | null | null | null | ['covid-19-detection'] | ['medical'] | [-2.00308740e-01 -1.65816963e-01 2.00061388e-02 -3.30398947e-01
1.41785100e-01 -2.98413306e-01 9.66698397e-03 -1.78921402e-01
-2.58578002e-01 5.56496501e-01 5.66633977e-02 -4.30309564e-01
-2.75047898e-01 -6.59126103e-01 -2.68351138e-01 -6.95804656e-01
-3.94440562e-01 6.51292562e-01 6.70563430e-02 4.57961142e-01
-2.05563173e-01 1.05869079e+00 -9.25170898e-01 2.85670072e-01
5.82461059e-01 1.13476920e+00 4.36577946e-01 6.97346449e-01
-9.95104909e-02 5.36706924e-01 -5.26466191e-01 3.69369626e-01
3.19059491e-01 -4.43200976e-01 -5.72115898e-01 4.17249054e-02
6.13023303e-02 -9.67123866e-01 -1.14219025e-01 6.62311792e-01
6.76952899e-01 -6.50033429e-02 9.68190432e-01 -8.01069438e-01
-4.18298304e-01 -2.85430104e-01 -2.81848818e-01 5.10344148e-01
-1.38660580e-01 6.80520535e-02 2.24556714e-01 -7.91830122e-01
5.94503999e-01 1.07173979e+00 9.34884429e-01 3.95122141e-01
-4.00946796e-01 -5.34498751e-01 -4.34508353e-01 2.22936813e-02
-1.13126612e+00 4.88183260e-01 4.87973779e-01 -8.82467866e-01
9.64524567e-01 9.83711332e-02 8.99912000e-01 5.92226148e-01
8.41130078e-01 4.78520483e-01 1.05534530e+00 6.24893829e-02
-7.77298352e-03 1.70316771e-01 1.17037542e-01 8.25227678e-01
5.32397509e-01 4.12630200e-01 3.86997521e-01 -2.08672628e-01
1.00000668e+00 5.82028031e-01 -1.46786675e-01 -4.95222360e-01
-8.54754448e-01 8.60264122e-01 6.94516063e-01 5.44781446e-01
-7.77872443e-01 -3.70611966e-01 6.90480769e-01 -2.92583811e-03
2.62776971e-01 9.39625651e-02 -6.02958441e-01 3.46602857e-01
-6.26092315e-01 1.80158615e-01 4.22655404e-01 5.00752330e-01
1.22450173e-01 2.21717715e-01 -2.83914179e-01 5.38728833e-01
5.82671702e-01 8.24498713e-01 6.45815432e-01 -4.77959424e-01
5.80010787e-02 7.86896586e-01 -1.64885774e-01 -1.03426147e+00
-9.17366207e-01 -5.54529667e-01 -1.38087153e+00 2.11021811e-01
-6.06597885e-02 -3.75894129e-01 -1.25738883e+00 1.32113647e+00
4.54546779e-01 1.99633732e-01 1.92788139e-01 1.26520252e+00
1.23658419e+00 4.81287986e-01 1.90418556e-01 -1.92505851e-01
1.49353063e+00 -6.67460978e-01 -6.32879734e-01 1.85305104e-01
6.32705271e-01 -4.24650967e-01 5.49687743e-01 1.11094922e-01
-6.62568688e-01 -4.89299029e-01 -1.03761041e+00 4.98413622e-01
-4.38832462e-01 2.10055098e-01 5.75614452e-01 5.20242274e-01
-8.83978367e-01 5.84520936e-01 -6.86836243e-01 -7.20456243e-01
6.28754735e-01 2.99163669e-01 -3.75893235e-01 -1.59572750e-01
-1.22677207e+00 1.09515643e+00 5.07029176e-01 2.51948327e-01
-9.98233736e-01 -4.34148133e-01 -3.46944928e-01 -9.14272480e-03
-1.61762819e-01 -1.04852486e+00 1.13558543e+00 -3.55773926e-01
-8.90667677e-01 1.04042828e+00 2.39690766e-01 -3.72102529e-01
4.11605120e-01 -9.15202722e-02 -3.90962750e-01 4.60021257e-01
-6.31578732e-04 3.32860053e-01 3.65343213e-01 -1.10789144e+00
-3.99558783e-01 -7.71819770e-01 -3.19542915e-01 1.53686374e-01
2.75630087e-01 2.26389512e-01 -1.50578246e-01 -4.71708894e-01
8.44348967e-02 -9.40614223e-01 -2.75215298e-01 2.23606080e-02
-1.97538450e-01 -1.90249741e-01 1.15546978e+00 -8.86889219e-01
7.89056897e-01 -1.95060539e+00 -5.58888912e-01 1.94915697e-01
4.57613379e-01 8.61786664e-01 1.67363375e-01 -1.05594069e-01
-6.07815795e-02 3.89534861e-01 -1.27042890e-01 4.15166855e-01
-4.15514350e-01 6.79891706e-02 2.33385786e-01 5.83846331e-01
2.25802567e-02 9.80126023e-01 -4.24672455e-01 -8.29188347e-01
4.12070304e-01 7.92996466e-01 -4.17861074e-01 6.59755766e-01
4.22282368e-02 4.54832107e-01 -9.77436900e-01 6.52982116e-01
9.60433483e-01 -5.92625380e-01 -3.94755378e-02 -1.64163217e-01
1.65694445e-01 -2.30182797e-01 -4.96895462e-01 1.15849519e+00
-1.29778758e-01 3.43989521e-01 1.26522064e-01 -8.52941096e-01
7.96345353e-01 8.76401484e-01 7.49789953e-01 -6.25718772e-01
7.40506053e-01 -2.92968210e-02 5.98168336e-02 -1.32147133e+00
-2.90252149e-01 -4.07898724e-01 4.25521851e-01 3.32725108e-01
-3.56334299e-01 -7.25249723e-02 -2.88856030e-01 -3.88153315e-01
9.03458893e-01 -5.48978858e-02 4.56937253e-01 -2.57689267e-01
4.99508381e-01 2.84974903e-01 5.87471724e-01 5.45779586e-01
-5.36292613e-01 7.56218910e-01 2.47189566e-01 -8.25971186e-01
-1.09758043e+00 -1.18459678e+00 -2.34161735e-01 8.18830878e-02
-5.02518155e-02 5.48651457e-01 -7.20983863e-01 -7.77675211e-01
-1.51029155e-02 2.35996991e-01 -6.36497378e-01 2.98115574e-02
-7.72524595e-01 -5.24120271e-01 5.77715993e-01 7.24035978e-01
6.61813021e-01 -1.31918049e+00 -1.11847341e+00 1.85484104e-02
2.22107828e-01 -9.01900947e-01 -2.56183267e-01 -2.82825399e-02
-1.47651327e+00 -1.40652406e+00 -9.47843432e-01 -1.13934135e+00
6.25594676e-01 3.75606269e-01 8.62539768e-01 3.78267497e-01
-6.12047255e-01 -5.23630343e-02 -1.34498268e-01 -6.39417350e-01
-3.63773972e-01 -2.59355456e-01 1.48811415e-01 -6.34965003e-01
5.13250947e-01 -4.22060370e-01 -8.79208446e-01 9.27211344e-02
-7.17811763e-01 -1.46382317e-01 1.05429327e+00 6.96607292e-01
6.34082079e-01 -3.50066684e-02 6.78957820e-01 -7.86793351e-01
7.51995027e-01 -5.35945296e-01 -4.23690170e-01 2.70206984e-02
-5.80791652e-01 -5.97906172e-01 7.44379342e-01 -1.19575977e-01
-9.22129750e-01 -1.22488886e-01 -2.29616072e-02 -1.03877783e+00
-5.75900197e-01 4.57413971e-01 1.08320355e-01 4.40187752e-01
6.31838560e-01 1.62062600e-01 5.83918750e-01 -4.44262296e-01
-2.36970022e-01 9.89752471e-01 2.47204006e-01 -1.32917138e-02
4.14227158e-01 3.24536681e-01 1.98869780e-01 -8.39498639e-01
-7.22452760e-01 -6.47969007e-01 -6.90369725e-01 -3.09938163e-01
1.50356340e+00 -8.41647863e-01 -7.47840047e-01 4.45212275e-01
-1.33852959e+00 7.45876059e-02 1.05935000e-01 1.18530822e+00
-3.94415557e-01 3.26694608e-01 -6.05258226e-01 -5.23447931e-01
-1.01156330e+00 -1.12769604e+00 5.44086277e-01 2.97673911e-01
-1.31346077e-01 -1.15985370e+00 9.83915702e-02 4.81619775e-01
6.09284878e-01 4.10621017e-01 1.25121212e+00 -1.05470860e+00
-4.80984598e-01 -3.93632799e-01 -4.48725551e-01 6.27419233e-01
2.84454256e-01 -3.38722318e-01 -8.17536175e-01 -2.72669822e-01
5.30100703e-01 -1.60129502e-01 2.46714413e-01 8.76212895e-01
1.23105538e+00 -2.86535144e-01 -4.50888693e-01 5.97263038e-01
1.43573999e+00 1.00890994e+00 3.06627333e-01 -6.50608763e-02
5.17079711e-01 3.28451037e-01 6.75471246e-01 2.46695414e-01
-1.15928225e-01 7.62373209e-02 5.76028526e-01 -4.50545996e-01
-4.62611392e-02 2.99237780e-02 -2.99030036e-01 8.71956289e-01
-9.35842618e-02 -3.99863273e-01 -1.01860666e+00 5.80041289e-01
-1.27189922e+00 -8.47449839e-01 -2.04898208e-01 1.77568090e+00
1.04480170e-01 -1.58371031e-01 -4.38752919e-02 -1.86351076e-01
8.34275007e-01 -1.13992922e-01 -6.74481988e-01 -4.36794907e-01
1.46616474e-01 6.14934862e-02 2.32912526e-01 2.15985682e-02
-1.15289152e+00 3.27072859e-01 6.64201880e+00 7.63306022e-02
-1.45390201e+00 -2.32554730e-02 5.45866072e-01 8.08960274e-02
1.40708521e-01 -6.19976401e-01 -3.36578101e-01 4.47457999e-01
5.30349910e-01 1.06567293e-01 -8.82806182e-02 9.32302177e-01
7.09634185e-01 1.76250458e-01 -6.30627275e-01 9.07801986e-01
2.20107719e-01 -1.09539092e+00 4.86114845e-02 1.85335591e-01
4.45147574e-01 4.47633266e-01 -2.24740744e-01 1.66527197e-01
-2.16563374e-01 -1.32732129e+00 -2.75158137e-01 4.87873554e-01
8.28244567e-01 -5.40735126e-01 1.34727323e+00 4.61843580e-01
-8.23181748e-01 2.29074940e-01 -3.13617915e-01 2.46706635e-01
1.57703876e-01 5.65739632e-01 -1.64223778e+00 3.36090744e-01
8.01061511e-01 1.90477222e-01 -6.45733997e-02 1.32038033e+00
1.19308531e-01 6.17200196e-01 -3.60423684e-01 -3.71846944e-01
3.19045186e-01 -3.48125368e-01 4.79317635e-01 1.15357506e+00
6.21834099e-01 3.79184425e-01 5.83784729e-02 8.96158338e-01
1.67696655e-01 3.67355973e-01 -1.27450013e+00 5.61444946e-02
1.12284318e-01 1.15840471e+00 -7.10234165e-01 -3.87023687e-01
-3.03151190e-01 5.46906412e-01 -3.90655875e-01 2.35436618e-01
-9.03498769e-01 -1.16744682e-01 3.35593969e-01 3.00856352e-01
4.54963386e-01 2.10140109e-01 -3.51530820e-01 -6.28537595e-01
-1.76901564e-01 -6.05223536e-01 4.19523478e-01 -1.07059312e+00
-1.14964354e+00 7.43784010e-01 -3.19622234e-02 -1.24665320e+00
-1.59719229e-01 -9.32334304e-01 -8.68555129e-01 8.83470953e-01
-1.10601997e+00 -9.94962275e-01 -4.51075077e-01 2.60276556e-01
2.81095475e-01 -4.71941143e-01 9.53268230e-01 3.69454622e-01
-2.72120684e-01 5.56206703e-02 1.46928683e-01 2.85174906e-01
1.53647885e-01 -8.87719750e-01 -8.02598894e-02 4.54655796e-01
-7.73080468e-01 6.75306976e-01 2.79454023e-01 -9.78199542e-01
-9.47927535e-01 -1.32455897e+00 8.05584133e-01 -7.53685087e-02
-5.33034466e-02 3.08851808e-01 -7.82552779e-01 5.53301275e-01
1.22627355e-01 3.07249933e-01 8.20485592e-01 -4.85775977e-01
1.95329309e-01 1.95888445e-01 -1.71801949e+00 2.24063993e-01
6.89665318e-01 -2.03907907e-01 -8.79859865e-01 3.86878431e-01
6.50005102e-01 -4.04610872e-01 -1.20423222e+00 9.44192946e-01
6.78869903e-01 -8.55647445e-01 9.68684196e-01 -9.20506716e-01
2.37918898e-01 -4.19342935e-01 -1.02043757e-02 -1.01269090e+00
-3.11730802e-01 3.58816892e-01 9.80374664e-02 3.48589361e-01
2.27972999e-01 -6.13988876e-01 1.03003395e+00 2.94302404e-02
-2.40988478e-01 -1.19429266e+00 -7.15672135e-01 -6.84660554e-01
1.80423692e-01 -7.39772841e-02 5.62996686e-01 9.32868481e-01
-5.43218315e-01 3.95380944e-01 -7.08693489e-02 2.43502215e-01
3.77911925e-01 9.10761207e-02 3.60524356e-01 -1.51598501e+00
1.74164966e-01 -1.18272327e-01 -3.71523023e-01 -8.11529517e-01
-2.33079597e-01 -9.13649499e-01 -1.79107741e-01 -2.01133204e+00
4.17131543e-01 -4.29944098e-01 -4.54106092e-01 1.23327993e-01
1.83691129e-01 -9.65431780e-02 6.39385208e-02 2.71625847e-01
-1.23910673e-01 1.69060305e-01 1.83094072e+00 -9.30760354e-02
-1.07849248e-01 3.51480097e-01 -1.55282706e-01 7.72158086e-01
1.28855634e+00 -4.11165208e-01 -5.82824647e-01 -1.70729533e-01
-3.32282215e-01 4.48065728e-01 3.52681905e-01 -8.96305025e-01
-3.72864716e-02 -4.82693687e-02 8.20828259e-01 -1.39950609e+00
2.68578321e-01 -1.06170011e+00 2.95262009e-01 9.05128181e-01
1.92020148e-01 2.10098386e-01 1.34742275e-01 2.18404636e-01
-2.10940182e-01 -1.59670636e-01 9.10965502e-01 -4.64685470e-01
-2.38709003e-01 7.00269282e-01 -5.60254335e-01 4.91881557e-02
1.16838455e+00 -5.38339138e-01 -9.41344649e-02 -8.03287923e-02
-6.07161045e-01 -1.84089504e-02 2.05070540e-01 2.08073348e-01
8.64660978e-01 -1.29128242e+00 -6.07915580e-01 2.35725671e-01
-1.65530697e-01 4.65206414e-01 6.83428109e-01 9.58340883e-01
-1.16104841e+00 9.78131950e-01 -3.40471923e-01 -8.62692773e-01
-1.44433630e+00 6.81664288e-01 8.15029204e-01 -2.38268927e-01
-7.34629095e-01 3.75263751e-01 4.90123272e-01 -7.14709640e-01
1.79697096e-01 -5.05219758e-01 -6.12933397e-01 -2.33029619e-01
2.01861084e-01 1.73238501e-01 1.52360871e-01 -6.30757868e-01
-5.37073612e-01 7.54289150e-01 -2.05250401e-02 5.53767383e-01
1.44473469e+00 2.09624976e-01 -1.08466551e-01 8.48750025e-02
1.44035244e+00 -4.66471612e-01 -7.04818249e-01 -3.36746797e-02
-4.88452673e-01 -1.92666233e-01 1.26836495e-03 -1.01742637e+00
-1.20743978e+00 9.53415990e-01 1.38056350e+00 8.70676637e-02
9.33250964e-01 -8.08541477e-02 8.21423411e-01 3.74106973e-01
-8.65749270e-02 -7.58948088e-01 -1.08367980e-01 5.56829929e-01
7.54502892e-01 -1.28579545e+00 5.67671983e-03 -9.21043307e-02
-5.38392067e-01 9.31971312e-01 7.02917278e-01 -1.39247164e-01
1.00763547e+00 1.42085761e-01 5.99392056e-01 -7.46925354e-01
-4.51387107e-01 2.39956915e-01 8.03206339e-02 7.53165185e-01
5.06660759e-01 1.83351159e-01 -3.78100097e-01 5.43560088e-01
6.33488595e-02 3.90074074e-01 1.13082066e-01 9.42537606e-01
-5.56381464e-01 -2.17278495e-01 -3.82271081e-01 8.00327241e-01
-8.46025288e-01 9.91215780e-02 -2.16563106e-01 1.06374717e+00
3.03178877e-01 5.18386900e-01 9.62320045e-02 -2.61945993e-01
2.65273303e-01 1.23359337e-01 1.92940384e-01 -3.61046106e-01
-3.76583815e-01 -2.37881020e-01 -2.11142778e-01 -1.22095749e-01
-3.66194308e-01 -1.55095458e-01 -1.61431372e+00 -2.48923432e-02
-2.07145765e-01 1.75346583e-01 8.46074164e-01 8.15809667e-01
4.29180562e-01 5.94375014e-01 5.39144695e-01 -4.07994032e-01
-5.19238710e-01 -1.12069178e+00 -6.43508196e-01 2.81674623e-01
4.37446505e-01 -4.53224957e-01 -3.80823106e-01 -1.95817173e-01] | [15.613738059997559, -1.662722110748291] |
443e11bb-b6e9-415d-b01a-dee137e79aca | realistic-face-animation-generation-from | 2103.14984 | null | https://arxiv.org/abs/2103.14984v1 | https://arxiv.org/pdf/2103.14984v1.pdf | Realistic face animation generation from videos | 3D face reconstruction and face alignment are two fundamental and highly related topics in computer vision. Recently, some works start to use deep learning models to estimate the 3DMM coefficients to reconstruct 3D face geometry. However, the performance is restricted due to the limitation of the pre-defined face templates. To address this problem, some end-to-end methods, which can completely bypass the calculation of 3DMM coefficients, are proposed and attract much attention. In this report, we introduce and analyse three state-of-the-art methods in 3D face reconstruction and face alignment. Some potential improvement on PRN are proposed to further enhance its accuracy and speed. | ['Minshan Xie', 'Zihao Jian'] | 2021-03-27 | null | null | null | null | ['face-alignment', 'face-reconstruction'] | ['computer-vision', 'computer-vision'] | [-1.38936877e-01 -7.95294940e-02 3.52679193e-02 -6.74121141e-01
-2.65535027e-01 -3.86095531e-02 4.68993276e-01 -6.20515764e-01
-5.72955459e-02 1.86180502e-01 -3.30372527e-02 -4.38609421e-02
5.93221970e-02 -6.53092980e-01 -4.29857016e-01 -4.96479094e-01
2.14877188e-01 5.97293198e-01 -2.07092285e-01 -3.03182527e-02
2.81086504e-01 1.08296871e+00 -1.65223944e+00 2.76900735e-02
3.78652066e-01 9.75206256e-01 -1.22842088e-01 2.08165854e-01
-4.28842813e-01 6.99193776e-02 -3.38558614e-01 -6.62544727e-01
4.88467842e-01 -3.85888368e-01 -4.32352126e-01 2.94493198e-01
8.88448060e-01 -4.51952279e-01 -3.16998810e-01 1.01264608e+00
9.97044861e-01 -2.03088745e-01 6.50793552e-01 -1.19473481e+00
-4.97562736e-01 -6.49124756e-02 -1.00656259e+00 4.51856740e-02
4.32867974e-01 -1.71279475e-01 3.64952266e-01 -1.30889916e+00
5.14080822e-01 1.49613726e+00 8.60390127e-01 8.40991974e-01
-7.65263975e-01 -8.82978916e-01 -7.11004436e-02 4.56125379e-01
-1.58945942e+00 -8.82812500e-01 1.18655527e+00 -3.07781309e-01
9.05503690e-01 9.51123387e-02 7.34043002e-01 7.85519242e-01
-5.26408665e-02 5.81586361e-01 1.02934074e+00 -5.74078381e-01
-1.75469488e-01 -2.86574304e-01 -2.40977600e-01 8.98011506e-01
-6.49978518e-02 2.06935033e-01 -5.43340981e-01 -2.63844430e-03
1.10866797e+00 -1.15723647e-01 -8.46749619e-02 -3.10040593e-01
-6.16670847e-01 6.48258388e-01 1.17270850e-01 4.05948758e-01
-3.74819785e-01 -1.97412208e-01 2.10108832e-01 3.48660350e-02
7.43091583e-01 -1.86707363e-01 -4.00813609e-01 -1.98694617e-02
-1.14761579e+00 2.03986302e-01 6.31969929e-01 9.75561619e-01
7.01800823e-01 2.06074014e-01 -2.04654224e-02 1.04727888e+00
6.29581273e-01 5.62333047e-01 5.91613464e-02 -1.02612674e+00
2.82636851e-01 4.97227281e-01 -2.63746947e-01 -1.20257974e+00
-4.48593825e-01 -3.16874087e-01 -1.00547898e+00 2.68998891e-01
2.09462330e-01 1.55846953e-01 -9.05742645e-01 1.35862041e+00
7.01334000e-01 5.68473458e-01 -3.88873100e-01 9.79541898e-01
1.22175395e+00 4.74069208e-01 -3.02214861e-01 -3.91412973e-01
1.24723053e+00 -8.12219381e-01 -8.40450644e-01 -1.18181020e-01
2.84699723e-02 -1.31202030e+00 5.95746756e-01 2.27201775e-01
-1.36800563e+00 -8.53450656e-01 -8.27432752e-01 -1.33266225e-01
1.09951906e-01 2.30029643e-01 4.44946229e-01 8.15536022e-01
-1.19498265e+00 4.31083173e-01 -6.69453025e-01 -3.08052212e-01
8.89820218e-01 6.57556057e-01 -6.16790593e-01 -1.28583118e-01
-8.98117363e-01 1.04723430e+00 2.96330103e-03 6.38864040e-01
-6.43451810e-01 -5.48755765e-01 -7.26693213e-01 -6.22090884e-02
7.61964992e-02 -6.91106796e-01 1.17616117e+00 -6.72135770e-01
-1.88198531e+00 1.43404639e+00 -3.71676743e-01 2.08784103e-01
5.03720880e-01 -8.96032900e-02 -3.18998992e-01 -6.24454916e-02
-1.62815213e-01 6.76234663e-01 1.15592229e+00 -1.07388663e+00
-2.22753152e-01 -7.61727095e-01 -2.48566747e-01 8.23371038e-02
-3.78853172e-01 3.81755561e-01 -1.03844333e+00 -5.13833642e-01
3.22282612e-01 -7.72426546e-01 -2.03237459e-01 4.00795132e-01
-6.59536198e-02 -2.93279469e-01 1.03732145e+00 -9.34595466e-01
7.37576902e-01 -1.95137632e+00 2.48099446e-01 1.54166088e-01
-9.76797380e-03 7.58162618e-01 -3.23918283e-01 1.87546946e-02
-6.23972192e-02 -1.60697699e-01 -1.60767391e-01 -8.67506146e-01
-1.99203923e-01 5.70553914e-02 -6.22977726e-02 8.20277393e-01
1.93859190e-01 6.97165012e-01 -3.20283443e-01 -7.17261910e-01
4.79450911e-01 9.90253448e-01 -6.34492755e-01 3.63246858e-01
-1.27885095e-03 5.85580826e-01 -3.18330318e-01 8.14296424e-01
1.46153104e+00 1.97831348e-01 2.10648283e-01 -4.54717904e-01
-4.36245203e-02 1.49137955e-02 -1.27099335e+00 1.68470490e+00
-4.81482118e-01 3.82154614e-01 4.01919872e-01 -1.14785075e+00
1.20580792e+00 3.44668448e-01 7.05392420e-01 -6.01482809e-01
4.59425181e-01 2.30098054e-01 -1.31492421e-01 -4.93501663e-01
8.59892815e-02 -2.09552318e-01 5.40041089e-01 3.97186607e-01
1.34636685e-01 -2.75447160e-01 -2.75260627e-01 -4.89256680e-01
2.66565919e-01 4.21348304e-01 4.16432887e-01 -2.65473545e-01
1.03500473e+00 -5.21333337e-01 7.74254382e-01 -5.64915650e-02
-3.46496642e-01 7.60825455e-01 3.23575199e-01 -7.43761480e-01
-1.03671408e+00 -7.77669251e-01 -3.47365707e-01 5.60477853e-01
-4.84169275e-02 -3.08018923e-01 -1.00520575e+00 -7.56424785e-01
-6.19923919e-02 1.09368801e-01 -3.55902791e-01 2.05177844e-01
-1.05266225e+00 -7.37229168e-01 5.20895183e-01 4.35466588e-01
6.77787721e-01 -1.02488470e+00 -2.10088924e-01 3.05190515e-02
-6.90813065e-02 -1.31029868e+00 -5.60391963e-01 -6.33171439e-01
-1.14089632e+00 -1.02893817e+00 -8.78086507e-01 -1.15116322e+00
1.03867352e+00 3.80525768e-01 1.00940025e+00 4.62419868e-01
-3.08585405e-01 1.31126791e-01 -1.59128785e-01 -4.27607983e-01
-3.27462196e-01 -5.81434220e-02 3.36668253e-01 2.28858571e-02
5.29223144e-01 -7.77868032e-01 -4.78763133e-01 5.43820441e-01
-4.02363390e-01 -4.03204970e-02 4.63072449e-01 5.12178183e-01
7.62737334e-01 -1.34551644e-01 2.82526553e-01 -5.62592924e-01
1.96494460e-01 1.32853448e-01 -7.92882025e-01 1.01761408e-01
-3.71028006e-01 -5.75388819e-02 5.18557966e-01 -2.89851576e-01
-1.02724206e+00 3.38797510e-01 -8.33616376e-01 -6.60886526e-01
-1.69898868e-01 2.96405200e-02 -5.78238368e-01 -5.48296511e-01
3.41674536e-01 2.48508617e-01 4.56403404e-01 -7.17324376e-01
2.29595527e-01 6.91971719e-01 2.79516429e-01 -4.14696366e-01
1.00266290e+00 5.27163029e-01 3.73177469e-01 -9.75666583e-01
-7.84883261e-01 -1.75719768e-01 -8.44522834e-01 -3.64578635e-01
5.66491604e-01 -7.83534765e-01 -9.04359043e-01 7.10527658e-01
-1.57366896e+00 1.39906570e-01 2.15671629e-01 4.60489452e-01
-6.26414001e-01 6.85069919e-01 -4.98679668e-01 -7.05034673e-01
-7.85351455e-01 -1.29214573e+00 1.17784810e+00 2.96172827e-01
2.50382349e-02 -6.96140289e-01 -7.31264725e-02 4.40317959e-01
4.36983645e-01 -2.30374970e-02 7.16814518e-01 -4.60367724e-02
-4.38785672e-01 -1.30516112e-01 -2.86814749e-01 4.53109920e-01
2.75825143e-01 1.81352988e-01 -1.03144848e+00 -4.19565648e-01
3.80166054e-01 7.91160464e-02 3.86824489e-01 5.08466840e-01
1.31660080e+00 -1.24461584e-01 -3.82378697e-01 9.64401603e-01
1.14723659e+00 1.81971431e-01 7.19449461e-01 -5.89279830e-02
6.85153425e-01 7.87069857e-01 4.91348267e-01 4.42636430e-01
1.83699980e-01 1.09392226e+00 4.85913098e-01 -6.84140995e-02
-3.98365825e-01 -2.50712156e-01 1.46778584e-01 9.92847204e-01
-4.67810810e-01 2.61417836e-01 -7.12495327e-01 2.39812527e-02
-1.42900360e+00 -9.57046747e-01 -1.32160068e-01 2.01798153e+00
6.27628982e-01 -1.41997218e-01 -5.43616107e-03 2.11832747e-01
8.32020819e-01 2.50632375e-01 -4.47300404e-01 -1.37784839e-01
4.15576575e-03 4.94470596e-01 -4.43283617e-02 6.54780328e-01
-1.02189541e+00 1.19332993e+00 6.84461403e+00 8.12806189e-01
-1.20020163e+00 8.68893638e-02 6.15400910e-01 1.90707147e-01
-2.78557986e-02 -1.99203625e-01 -1.28014982e+00 2.41493344e-01
3.52629215e-01 2.96539009e-01 4.32867736e-01 8.48521411e-01
1.41082108e-01 3.22394431e-01 -1.01832938e+00 1.49886215e+00
4.78694201e-01 -1.21730649e+00 1.31691843e-01 2.29348794e-01
6.16634905e-01 -4.44066167e-01 7.13229850e-02 3.81770693e-02
-4.74501103e-01 -1.15680671e+00 3.79565507e-01 5.22511065e-01
1.02575791e+00 -8.54874134e-01 7.03270078e-01 2.86251307e-01
-1.17101955e+00 3.16695482e-01 -7.80011833e-01 -1.06587872e-01
1.33510873e-01 6.69597268e-01 -8.75650644e-01 4.88074481e-01
7.81322300e-01 6.84604108e-01 -4.22917426e-01 9.66810882e-01
-2.21938133e-01 2.25352064e-01 -3.75096053e-01 1.62956432e-01
-2.18438908e-01 -3.32328111e-01 3.80992591e-01 8.57429564e-01
6.54771149e-01 3.08652252e-01 -1.16600886e-01 8.15412700e-01
-1.51086748e-01 1.62428796e-01 -5.19447088e-01 2.20673069e-01
2.73638457e-01 1.40360510e+00 -5.97192645e-01 7.83903301e-02
-5.30197799e-01 8.45624685e-01 2.82478184e-01 -4.68304493e-02
-7.92333603e-01 5.56683280e-02 8.68299603e-01 2.76843280e-01
2.75929451e-01 -5.64353645e-01 -2.84872442e-01 -9.28359568e-01
1.19714439e-01 -9.31418657e-01 5.06328046e-02 -3.87576789e-01
-1.27449512e+00 5.95246553e-01 -5.70638999e-02 -1.18780720e+00
-2.29273409e-01 -6.43727779e-01 -5.54726839e-01 9.48035181e-01
-1.60640538e+00 -1.21892810e+00 -2.88195103e-01 6.33143902e-01
6.31693363e-01 -5.28535426e-01 7.89758563e-01 6.58002079e-01
-4.78860676e-01 9.66188967e-01 -2.08910614e-01 1.85264051e-01
5.28024733e-01 -3.47761214e-01 7.17594028e-01 6.76181495e-01
3.19629937e-01 5.41404307e-01 5.22621453e-01 -4.51632053e-01
-1.63439310e+00 -7.97703683e-01 1.07879543e+00 -2.87079871e-01
-2.50611216e-01 -3.68783176e-01 -8.52977097e-01 3.92011136e-01
-8.31859559e-03 2.16070265e-01 4.90869254e-01 3.68210785e-02
-1.84655309e-01 -4.20945376e-01 -1.50649929e+00 6.30443692e-01
1.46021950e+00 -3.62869948e-01 -4.93276834e-01 1.34509444e-01
1.82048157e-01 -6.00965381e-01 -6.68533862e-01 7.19496250e-01
7.70352542e-01 -1.18053913e+00 1.18235302e+00 -2.50298262e-01
1.05134733e-01 -3.83618683e-01 -8.03072751e-02 -9.81223702e-01
-2.60266215e-01 -6.00217342e-01 -2.90363818e-01 1.20466185e+00
-2.02833652e-01 -2.91407317e-01 1.15300953e+00 1.96220130e-01
3.04435808e-02 -8.14061463e-01 -1.34497178e+00 -4.88433838e-01
-9.03197899e-02 -2.84584045e-01 8.84076893e-01 8.30086708e-01
-5.47279477e-01 4.02092874e-01 -5.41506290e-01 1.88563660e-01
7.02210605e-01 2.86684275e-01 8.62676859e-01 -1.46748245e+00
2.06787780e-01 -5.15931249e-01 -5.01168311e-01 -1.29110241e+00
5.51914454e-01 -8.60805631e-01 -1.33309841e-01 -1.34979248e+00
1.65711250e-02 -3.63573134e-01 2.29042307e-01 3.47086668e-01
7.78620020e-02 5.31950533e-01 1.39867082e-01 9.31915175e-03
-2.80506127e-02 8.56677890e-01 1.38050175e+00 -3.13593298e-02
-6.72892928e-02 2.54470646e-01 -4.61115658e-01 9.03384924e-01
8.28951895e-01 -2.56032050e-01 -1.72992796e-01 -6.66316926e-01
-2.91942388e-01 4.94195633e-02 3.58294696e-02 -8.60972047e-01
8.28692243e-02 -1.29791528e-01 8.60445023e-01 -9.01353061e-01
6.93854213e-01 -9.26457286e-01 1.93539351e-01 3.48313719e-01
2.61729956e-01 1.82806626e-01 2.47747362e-01 3.39127192e-03
-1.99551582e-01 -3.08508813e-01 1.22360945e+00 -1.70636922e-01
-4.46164399e-01 1.00751424e+00 -4.32095155e-02 -3.64035249e-01
9.75865960e-01 -3.93473566e-01 2.62951910e-01 -2.23538727e-01
-3.79803717e-01 -1.62967548e-01 3.11503619e-01 4.84236449e-01
1.04765046e+00 -1.59744132e+00 -9.28742230e-01 7.48621762e-01
-3.99664551e-01 8.61261860e-02 4.73783374e-01 5.37396252e-01
-6.14446640e-01 5.01531124e-01 -5.57172835e-01 -5.94803870e-01
-1.56065631e+00 5.48988402e-01 4.91473079e-01 9.68850702e-02
-4.98975635e-01 8.64403307e-01 -1.39244393e-01 -6.95019364e-01
2.94130385e-01 2.79878497e-01 -3.69328499e-01 -2.92095691e-02
6.16666257e-01 4.04325962e-01 1.91877201e-01 -1.15185702e+00
-5.54050565e-01 1.49168050e+00 1.31644383e-01 1.49693102e-01
1.30314457e+00 7.67512619e-02 -4.61096495e-01 -4.66890097e-01
1.29470897e+00 -6.53096363e-02 -1.05848050e+00 -1.44321844e-01
-2.41332054e-01 -9.18651521e-01 4.07399833e-02 -2.69233435e-01
-1.62040937e+00 1.26860714e+00 7.47957766e-01 -3.23703587e-01
1.27747047e+00 -1.34956032e-01 9.19738710e-01 1.21323064e-01
4.39661086e-01 -8.93911064e-01 -3.02176271e-02 6.24868870e-01
1.17259955e+00 -1.15849090e+00 3.10343921e-01 -8.16272378e-01
-6.36229813e-02 1.41785204e+00 9.34450150e-01 3.67693007e-02
9.35383677e-01 2.48256236e-01 1.33269116e-01 -5.19817285e-02
-1.20802335e-01 -1.37970731e-01 2.66979039e-01 7.48705566e-01
6.42302096e-01 -4.02238041e-01 -5.23079753e-01 2.76173830e-01
-2.06931993e-01 1.28150806e-01 -2.91452091e-02 5.77046216e-01
-1.06873527e-01 -1.53518224e+00 -6.26770735e-01 2.35946655e-01
-4.93658185e-01 1.12120412e-01 -2.02746168e-01 5.74554205e-01
1.39035985e-01 6.77466750e-01 1.50605058e-02 -3.89281005e-01
4.98572975e-01 -2.88544819e-02 1.04290581e+00 -2.97812670e-01
-2.04495072e-01 8.98303166e-02 -2.81270713e-01 -3.96887273e-01
-7.02580154e-01 -5.12509227e-01 -9.69204187e-01 -5.42804182e-01
-2.85543323e-01 -1.52759239e-01 6.80558562e-01 8.03391933e-01
6.19537473e-01 -8.27371180e-02 9.16073203e-01 -1.28743923e+00
-4.91417497e-01 -9.96625543e-01 -3.85865927e-01 3.26849781e-02
5.22249602e-02 -8.16945553e-01 -1.56927601e-01 -9.92078856e-02] | [13.219151496887207, 0.3117883503437042] |
5b8a2fcb-ad9b-4b1a-bf38-9d339a1a3bc8 | zeroforge-feedforward-text-to-shape-without | 2306.08183 | null | https://arxiv.org/abs/2306.08183v2 | https://arxiv.org/pdf/2306.08183v2.pdf | ZeroForge: Feedforward Text-to-Shape Without 3D Supervision | Current state-of-the-art methods for text-to-shape generation either require supervised training using a labeled dataset of pre-defined 3D shapes, or perform expensive inference-time optimization of implicit neural representations. In this work, we present ZeroForge, an approach for zero-shot text-to-shape generation that avoids both pitfalls. To achieve open-vocabulary shape generation, we require careful architectural adaptation of existing feed-forward approaches, as well as a combination of data-free CLIP-loss and contrastive losses to avoid mode collapse. Using these techniques, we are able to considerably expand the generative ability of existing feed-forward text-to-shape models such as CLIP-Forge. We support our method via extensive qualitative and quantitative evaluations | ['Chinmay Hegde', 'Adarsh Krishnamurthy', 'Aditya Balu', 'Anushrut Jignasu', 'Ameya Joshi', 'Minh Pham', 'Kelly O. Marshall'] | 2023-06-14 | null | null | null | null | ['text-to-shape-generation'] | ['computer-vision'] | [ 4.68398064e-01 3.59673172e-01 1.76560476e-01 -3.38276237e-01
-1.09859359e+00 -7.07099140e-01 9.76694763e-01 -1.63019598e-02
1.49754420e-01 6.53152168e-01 4.12934631e-01 -3.47247988e-01
5.35716340e-02 -1.11032057e+00 -7.96804309e-01 -3.51062775e-01
1.68290213e-01 8.21406841e-01 1.06173418e-01 -4.02458578e-01
2.42540061e-01 6.44648731e-01 -1.51614249e+00 4.28447038e-01
6.12978876e-01 9.45807636e-01 -1.28951073e-01 6.03540301e-01
-3.81767213e-01 5.46807289e-01 -2.54059464e-01 -7.54839361e-01
3.74792725e-01 -3.45492959e-01 -6.07012749e-01 9.73288491e-02
4.99380112e-01 -4.36039567e-01 -1.57722160e-01 4.77909088e-01
7.45587945e-01 2.00885326e-01 1.07110584e+00 -9.45833087e-01
-9.97536480e-01 5.29763937e-01 -2.85076529e-01 -3.87960494e-01
1.24681219e-01 2.75234014e-01 1.03380609e+00 -1.55295646e+00
9.75602984e-01 1.17806768e+00 9.19579923e-01 6.99976385e-01
-1.62632120e+00 -5.47415137e-01 -2.73610111e-02 -4.21218246e-01
-1.42844582e+00 -9.61772025e-01 1.11968541e+00 -4.97334361e-01
1.24521399e+00 3.35102230e-01 8.28555882e-01 9.60773170e-01
-1.58901289e-01 7.02392876e-01 5.32675207e-01 -7.07951486e-01
3.58308107e-01 1.49444891e-02 -6.55776739e-01 6.16970181e-01
6.48603961e-02 3.93602520e-01 -3.65089506e-01 -3.18210959e-01
1.06850183e+00 -2.97504812e-01 1.76411867e-03 -7.49130905e-01
-7.12400019e-01 8.79431903e-01 2.87068427e-01 2.31962666e-01
-2.21938580e-01 4.90181059e-01 3.20783287e-01 -5.59031032e-03
9.49304700e-01 4.27721947e-01 -3.18264931e-01 3.57466787e-02
-1.48615444e+00 7.25086629e-01 8.08036625e-01 1.27748132e+00
6.76305175e-01 5.77607095e-01 -4.43334430e-01 8.41302752e-01
5.76577246e-01 4.91597742e-01 1.22940563e-01 -9.97368932e-01
3.24360639e-01 2.40786925e-01 1.15686469e-01 -8.08273435e-01
3.13347615e-02 -3.66315246e-01 -4.91036236e-01 4.64237362e-01
2.67371118e-01 -1.00374840e-01 -1.26177180e+00 1.45007813e+00
1.68576732e-01 1.37992268e-02 -1.86511993e-01 5.70773065e-01
7.82834828e-01 4.92815942e-01 1.93543732e-02 2.35883184e-02
8.67257357e-01 -6.80064082e-01 -4.54509497e-01 -1.10293580e-02
4.60764408e-01 -9.66221929e-01 1.05045784e+00 6.12615943e-02
-1.60756636e+00 -1.95268959e-01 -9.95333135e-01 -2.04179600e-01
-2.94213325e-01 1.23065762e-01 6.87254012e-01 8.41075480e-01
-1.09015787e+00 6.97367787e-01 -6.57680929e-01 -8.90344381e-02
8.56228471e-01 2.73427367e-01 -5.11099026e-02 6.00720234e-02
-7.32006669e-01 6.37909532e-01 7.15015382e-02 -2.04715148e-01
-8.69549096e-01 -1.06922185e+00 -1.00961685e+00 -1.98684987e-02
2.36493334e-01 -1.06113565e+00 1.47895479e+00 -8.63982975e-01
-1.66545737e+00 9.96891201e-01 -1.31088629e-01 -3.52906287e-01
4.91737276e-01 -1.41514316e-01 1.83170542e-01 1.14431717e-02
-1.75582573e-01 1.00708270e+00 9.17089581e-01 -1.76309502e+00
4.98602027e-03 6.29511401e-02 -5.73632605e-02 1.39301077e-01
-2.85276175e-01 -1.38514265e-01 -4.73615199e-01 -1.22808909e+00
-6.79597408e-02 -6.17843866e-01 -4.01694357e-01 5.89729607e-01
-4.38652039e-01 1.52790353e-01 8.95166814e-01 -3.83381575e-01
9.66114342e-01 -1.84119868e+00 -5.61269186e-02 2.34369934e-01
6.38720989e-02 1.75505564e-01 -3.03170413e-01 5.20726621e-01
-4.13115583e-02 3.69313270e-01 -3.89447331e-01 -1.00555801e+00
1.57300800e-01 1.39717951e-01 -6.77555680e-01 1.25330776e-01
5.49699247e-01 1.15448427e+00 -6.62284493e-01 -7.24859715e-01
3.97409201e-01 6.20461047e-01 -1.03801751e+00 2.09157363e-01
-7.70600200e-01 2.60971040e-01 -3.12000513e-01 6.43820465e-01
4.42013472e-01 -6.89941123e-02 -1.85514346e-01 -2.39329562e-01
-2.34028809e-02 2.26537362e-01 -8.03464174e-01 2.03931999e+00
-4.87139285e-01 4.01594698e-01 -2.47815270e-02 -6.07438326e-01
1.09635592e+00 3.86913657e-01 5.36153495e-01 -4.78340238e-01
3.45343858e-01 2.74553150e-01 -5.04943132e-01 7.03148022e-02
7.11400270e-01 -6.19757652e-01 1.44756213e-01 5.03861904e-01
3.10463011e-01 -8.49516034e-01 1.73708342e-03 2.67510593e-01
7.71945000e-01 7.66647935e-01 -8.86153951e-02 -2.49775469e-01
1.53209031e-01 -3.24389823e-02 1.94074944e-01 4.73533630e-01
4.09318477e-01 1.30127704e+00 7.46910367e-03 -2.90076315e-01
-1.68683875e+00 -1.19144928e+00 -2.29087044e-02 7.93683350e-01
-4.27058190e-01 -4.25081849e-01 -7.99131632e-01 -4.76473063e-01
-4.08011749e-02 1.16978180e+00 -5.48765957e-01 2.35130973e-02
-7.07800031e-01 -4.47169334e-01 8.32347214e-01 6.50681734e-01
-2.43357047e-01 -1.05703449e+00 -2.12666273e-01 3.43392432e-01
2.12567896e-01 -7.97284842e-01 -6.27407014e-01 1.02549709e-01
-1.05872047e+00 -6.08260691e-01 -1.05971253e+00 -8.04087937e-01
9.60767090e-01 -5.98012134e-02 1.40401113e+00 1.29794315e-01
-3.65564108e-01 2.93162793e-01 -2.97960430e-01 -4.34563011e-01
-6.69572949e-01 5.18990960e-03 -4.34201598e-01 -1.15709625e-01
-1.68685198e-01 -9.33838010e-01 -4.42205966e-01 -4.55328450e-02
-9.90903795e-01 4.21492696e-01 4.31660384e-01 7.32299685e-01
8.75136852e-01 -3.95580143e-01 5.95174074e-01 -7.40770638e-01
5.91895521e-01 -1.90950692e-01 -5.67967236e-01 1.01809315e-01
-6.08630538e-01 3.63030940e-01 6.48023188e-01 -4.01088446e-01
-1.34052336e+00 1.93989903e-01 -5.53800166e-01 -1.04636323e+00
-5.04351482e-02 3.60661179e-01 -1.04672775e-01 -1.39087617e-01
9.12099302e-01 3.45713407e-01 2.42219195e-02 -5.45414150e-01
1.16772151e+00 2.79766023e-01 4.70423371e-01 -8.28760803e-01
1.11795783e+00 5.70169091e-01 -1.19424097e-01 -7.27921784e-01
-6.19287074e-01 8.29256997e-02 -6.88069165e-01 -1.28507778e-01
6.26581609e-01 -7.39510715e-01 -1.21874347e-01 3.88307363e-01
-1.23913395e+00 -6.58614099e-01 -9.45777595e-01 -2.18130007e-01
-1.07321775e+00 1.39779970e-01 -5.45117199e-01 -9.30631399e-01
-8.32740784e-01 -6.53095901e-01 1.38713586e+00 -3.98916677e-02
-3.79155278e-01 -1.10003865e+00 1.76071286e-01 1.15875147e-01
7.52339363e-01 2.76461333e-01 1.04992795e+00 -5.13911545e-01
-7.03283429e-01 -1.23941198e-01 -1.29842192e-01 1.91594288e-01
9.19352919e-02 3.48663926e-02 -9.66444135e-01 -2.46098891e-01
-4.27233666e-01 -5.39438665e-01 7.45210290e-01 4.82387871e-01
1.04021227e+00 -5.62684298e-01 -3.77361625e-01 7.62413681e-01
1.43558538e+00 -5.26283532e-02 6.99386775e-01 -2.34969676e-01
6.27034605e-01 4.08606321e-01 3.21483463e-01 5.96577108e-01
3.04375291e-01 6.84215426e-01 2.45553851e-01 -1.96505815e-01
-6.73843205e-01 -7.75266111e-01 -4.07953486e-02 7.05435455e-01
-8.69757682e-02 -3.99896562e-01 -6.68840289e-01 8.02166879e-01
-1.57737505e+00 -1.00872171e+00 7.40652457e-02 2.11621308e+00
1.10434449e+00 1.61298990e-01 -2.48275790e-02 1.37041241e-01
3.49398494e-01 2.21275896e-01 -3.18724841e-01 -3.35238189e-01
3.05243544e-02 5.76873004e-01 2.94443190e-01 5.81180155e-01
-9.39488709e-01 1.12023413e+00 7.49149179e+00 1.25080717e+00
-1.03912866e+00 -9.65751242e-03 7.69158900e-01 -4.28386837e-01
-1.08702445e+00 1.68740913e-01 -6.18626058e-01 6.14522174e-02
7.29372680e-01 -2.73738921e-01 5.23089170e-01 7.59652793e-01
2.15485543e-01 2.16349676e-01 -1.11029541e+00 8.03729892e-01
3.03316385e-01 -1.79911768e+00 3.04545432e-01 -7.72081837e-02
7.93632686e-01 -2.07336277e-01 -1.17226332e-01 2.68281341e-01
4.82971281e-01 -1.16051054e+00 1.04091930e+00 6.97264373e-01
1.38025653e+00 -6.19228959e-01 9.42038149e-02 2.59688437e-01
-1.27506793e+00 3.53540212e-01 -1.92974925e-01 2.93345660e-01
5.57423949e-01 7.25670159e-01 -9.30926204e-01 5.41468918e-01
8.66072029e-02 4.82690275e-01 -2.28237256e-01 9.18231368e-01
6.53532371e-02 5.97799778e-01 -4.84905571e-01 6.28160536e-02
-1.60035193e-01 3.89850028e-02 6.26368225e-01 1.16048419e+00
5.46646178e-01 6.63009807e-02 1.79502830e-01 1.58699930e+00
-7.70471990e-02 1.38361007e-01 -9.84756291e-01 -1.82530284e-01
5.44332862e-01 1.02709198e+00 -6.37471795e-01 -3.22866201e-01
-1.64959714e-01 8.44881475e-01 1.60445824e-01 3.21997821e-01
-5.62843740e-01 -3.97664636e-01 3.06684166e-01 6.23118341e-01
5.19534647e-01 -4.14963931e-01 -7.19968379e-01 -8.70762765e-01
-1.83148816e-01 -5.34956515e-01 -1.33505374e-01 -1.05355752e+00
-1.36362207e+00 4.75782990e-01 1.75868347e-01 -1.23317039e+00
-5.51901579e-01 -2.22780198e-01 -8.01920056e-01 8.52700651e-01
-1.20741260e+00 -1.96363866e+00 6.32178560e-02 4.67592001e-01
6.87926829e-01 -2.14578867e-01 1.03565741e+00 3.71569842e-01
6.50156885e-02 6.84535325e-01 -2.41656914e-01 3.22429016e-02
3.80701900e-01 -1.01609349e+00 9.57410693e-01 5.79100013e-01
3.24292988e-01 5.02525926e-01 7.74215877e-01 -8.20540965e-01
-1.29991090e+00 -1.20829415e+00 6.85242236e-01 -5.24365366e-01
2.72038519e-01 -5.61016858e-01 -6.19918823e-01 6.43923342e-01
1.85410485e-01 7.86914118e-03 4.97878343e-01 -2.20086649e-02
-2.35556260e-01 2.08901778e-01 -1.15597510e+00 8.23088050e-01
1.27184963e+00 -3.77636284e-01 -5.87799728e-01 7.77320191e-02
8.57867718e-01 -5.48490167e-01 -8.27682316e-01 5.12256205e-01
5.59967637e-01 -7.18918443e-01 1.23863304e+00 -4.88989621e-01
5.87093771e-01 -4.49418202e-02 -1.37718648e-01 -1.24691069e+00
-3.42031538e-01 -8.04782987e-01 -2.88987696e-01 1.46130145e+00
5.62195420e-01 -6.63689449e-02 1.12727392e+00 6.09109163e-01
-4.73780692e-01 -1.07071877e+00 -8.77706409e-01 -6.46864891e-01
5.79258561e-01 -5.81855953e-01 6.12722158e-01 6.03354335e-01
-2.90740907e-01 2.53680944e-01 -4.20354545e-01 -4.62207109e-01
6.52832031e-01 1.37891725e-01 7.43738592e-01 -1.01909578e+00
-3.17544609e-01 -5.79595864e-01 3.98605457e-03 -1.10801780e+00
-1.22961827e-01 -1.07505345e+00 2.18693078e-01 -1.69101930e+00
8.32415074e-02 -7.76340306e-01 2.72320092e-01 6.26077414e-01
3.22039425e-01 5.08645058e-01 2.23376393e-01 1.13775238e-01
-1.64578870e-01 1.08950543e+00 1.44543731e+00 -1.48421898e-01
-1.94410086e-01 -2.21154049e-01 -6.64446890e-01 5.81710637e-01
6.86376810e-01 -5.29308140e-01 -6.66789412e-01 -4.27260071e-01
2.80366212e-01 4.46468452e-03 3.87939066e-01 -7.64128864e-01
-3.61280702e-02 -2.53842175e-01 4.45405036e-01 -7.73454785e-01
7.88846016e-01 -5.81358135e-01 1.39966562e-01 8.83826613e-02
-3.56345654e-01 -2.69310862e-01 4.13342416e-01 3.82928371e-01
1.31220609e-01 -3.31181645e-01 7.69719064e-01 -3.09433013e-01
-8.28384236e-02 3.73793602e-01 -2.61230767e-01 1.89306587e-01
8.72427940e-01 -4.06971186e-01 -1.45973384e-01 -5.85499585e-01
-6.94678962e-01 -2.23429799e-01 6.19052947e-01 5.52668810e-01
8.87356460e-01 -1.64792848e+00 -8.66170049e-01 4.41342026e-01
-4.78397831e-02 1.03996433e-01 1.50421401e-02 3.57653856e-01
-3.89902830e-01 3.59466076e-01 -5.75130107e-03 -2.70575851e-01
-1.13442326e+00 4.53182071e-01 3.67126852e-01 -1.51489809e-01
-8.01263630e-01 8.28192711e-01 2.70303190e-02 -7.03264892e-01
6.87148124e-02 -1.29505873e-01 1.66722879e-01 -1.60664260e-01
2.27996796e-01 -2.74640601e-02 1.26729846e-01 -4.05231833e-01
4.59678890e-03 5.45177758e-01 1.48826957e-01 -6.98996842e-01
1.36833715e+00 2.28263631e-01 2.92262971e-01 1.81778252e-01
8.82476628e-01 2.25875363e-01 -1.34010124e+00 -7.80678317e-02
-4.10391480e-01 -4.70809937e-01 9.82360616e-02 -8.75351548e-01
-1.03134179e+00 8.41884136e-01 3.18125099e-01 2.92543303e-02
8.93209100e-01 -5.72751984e-02 1.03727210e+00 8.24506879e-02
6.14988506e-01 -9.70596015e-01 2.46378481e-01 4.08093035e-01
1.37990689e+00 -6.60135388e-01 1.35901511e-01 -5.61675370e-01
-3.83454949e-01 9.26855505e-01 3.48217070e-01 -3.10933858e-01
8.60089600e-01 7.21319735e-01 -1.23865828e-01 -2.01520070e-01
-1.07084632e+00 -8.45048502e-02 5.43410957e-01 7.68191695e-01
6.19118810e-01 -2.15578765e-01 -2.11560965e-01 5.23778081e-01
-2.55122811e-01 2.88239419e-01 2.27780998e-01 9.91649508e-01
-3.23791593e-01 -1.54427743e+00 -2.64360338e-01 5.48504293e-01
-2.92294592e-01 -4.44940865e-01 -5.27997434e-01 4.80029434e-01
-2.58366484e-02 7.08955109e-01 1.15220010e-01 -1.51903927e-01
2.47967333e-01 2.76228547e-01 7.78815627e-01 -6.93879664e-01
-4.81027007e-01 3.89225364e-01 3.22689354e-01 -2.99490184e-01
-1.24671645e-01 -6.13179088e-01 -1.26915157e+00 -3.04931164e-01
-5.04732251e-01 -4.64771062e-01 6.28130615e-01 7.73803949e-01
6.48670852e-01 4.42616791e-01 3.02339315e-01 -1.46500325e+00
-5.00668943e-01 -8.47223520e-01 -1.94949791e-01 5.36222160e-01
-5.89622334e-02 -6.45389140e-01 -1.63998567e-02 4.02695447e-01] | [9.02125358581543, -3.583132028579712] |
e94bb59a-7365-4450-a529-744124eec1fe | act-aware-slot-value-predicting-in-multi | 2208.02462 | null | https://arxiv.org/abs/2208.02462v1 | https://arxiv.org/pdf/2208.02462v1.pdf | Act-Aware Slot-Value Predicting in Multi-Domain Dialogue State Tracking | As an essential component in task-oriented dialogue systems, dialogue state tracking (DST) aims to track human-machine interactions and generate state representations for managing the dialogue. Representations of dialogue states are dependent on the domain ontology and the user's goals. In several task-oriented dialogues with a limited scope of objectives, dialogue states can be represented as a set of slot-value pairs. As the capabilities of dialogue systems expand to support increasing naturalness in communication, incorporating dialogue act processing into dialogue model design becomes essential. The lack of such consideration limits the scalability of dialogue state tracking models for dialogues having specific objectives and ontology. To address this issue, we formulate and incorporate dialogue acts, and leverage recent advances in machine reading comprehension to predict both categorical and non-categorical types of slots for multi-domain dialogue state tracking. Experimental results show that our models can improve the overall accuracy of dialogue state tracking on the MultiWOZ 2.1 dataset, and demonstrate that incorporating dialogue acts can guide dialogue state design for future task-oriented dialogue systems. | ['Biing-Hwang Juang', 'Ting-Wei Wu', 'Ruolin Su'] | 2022-08-04 | null | null | null | null | ['dialogue-state-tracking', 'machine-reading-comprehension', 'task-oriented-dialogue-systems'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 2.49933854e-01 6.96261287e-01 -3.56160074e-01 -6.22232378e-01
-4.35262501e-01 -8.19532990e-01 1.19766450e+00 3.12157542e-01
-2.21021101e-01 8.03992450e-01 9.41772342e-01 -6.90657735e-01
4.43281885e-03 -5.70088625e-01 4.63969886e-01 2.65345335e-01
7.16891587e-02 9.38432097e-01 3.62460911e-01 -1.11030185e+00
3.95259649e-01 2.31112167e-02 -1.24197304e+00 5.21516860e-01
8.89564395e-01 5.14513969e-01 4.18729335e-02 1.10022604e+00
-7.15409935e-01 1.30003095e+00 -9.92868245e-01 4.08635549e-02
-1.91969752e-01 -7.95858920e-01 -1.74178708e+00 2.61061460e-01
-2.21091062e-01 -5.74825585e-01 -2.54161358e-01 5.83594024e-01
4.14746761e-01 5.01797497e-01 7.43813396e-01 -1.24928296e+00
9.65255946e-02 6.70553446e-01 1.86238527e-01 1.02194183e-01
1.16337502e+00 1.51931018e-01 1.08687401e+00 -1.29777744e-01
7.94035017e-01 1.59946597e+00 3.45289528e-01 1.19607067e+00
-1.17587304e+00 -7.57360607e-02 1.26919553e-01 -1.13423996e-01
-6.20541096e-01 -1.05157721e+00 5.33649564e-01 -4.55877513e-01
1.32799971e+00 4.56119329e-01 5.82452118e-01 8.05893958e-01
-1.15551716e-02 9.90105093e-01 1.04369438e+00 -8.35101426e-01
1.62785769e-01 2.22832248e-01 3.66735280e-01 6.05366409e-01
-7.55933523e-01 -3.42721969e-01 -7.29243457e-01 -4.68005270e-01
5.68629742e-01 -6.61063552e-01 -2.56698318e-02 -1.63949907e-01
-1.09618378e+00 1.10665560e+00 -3.34617406e-01 2.71018744e-01
-7.90292099e-02 -5.04135251e-01 8.56472373e-01 5.73846042e-01
4.79987234e-01 1.02544880e+00 -5.13013721e-01 -9.89592135e-01
-3.50422293e-01 6.00267172e-01 1.66525757e+00 8.44195902e-01
4.30993557e-01 -1.58753380e-01 -6.91549778e-01 1.40130210e+00
3.60580087e-01 -8.49339813e-02 4.67530727e-01 -1.24506605e+00
4.71744895e-01 1.12645531e+00 5.16544580e-01 -4.32346821e-01
-8.20853055e-01 5.22440553e-01 -1.23515688e-01 -1.35314837e-01
8.20702851e-01 -4.20699716e-01 -4.40306544e-01 1.68776453e+00
4.97782052e-01 -6.51754439e-01 5.31145692e-01 6.38286889e-01
9.87140179e-01 7.40447164e-01 4.66053754e-01 -1.99214861e-01
1.71235442e+00 -5.82595885e-01 -1.15713584e+00 -4.50071812e-01
1.31963193e+00 -7.69520938e-01 1.00497687e+00 -1.74751901e-03
-1.10066438e+00 -7.31711462e-02 -6.28497124e-01 -1.34519428e-01
-1.58578724e-01 -2.57717431e-01 7.14859128e-01 6.87174082e-01
-8.45516622e-01 1.58487067e-01 -6.72278225e-01 -6.68023169e-01
-2.07314372e-01 6.56648725e-02 -6.06824942e-02 3.77095342e-01
-1.78443241e+00 1.45048714e+00 5.14459729e-01 -2.07465410e-01
-3.46510172e-01 -1.96189269e-01 -1.20102096e+00 -8.46073218e-03
3.27694595e-01 -3.61261487e-01 2.08956361e+00 -3.55804741e-01
-2.04036832e+00 7.25701094e-01 -2.83985943e-01 -3.95027727e-01
3.62534046e-01 3.97000946e-02 -4.20295149e-01 -6.94673788e-03
3.57076749e-02 6.59677744e-01 1.09570459e-01 -8.47075820e-01
-9.93261516e-01 -2.29602918e-01 6.72233820e-01 1.00512755e+00
-3.00469905e-01 2.95233846e-01 -2.38194302e-01 2.06209645e-01
-1.57625861e-02 -8.24371636e-01 -2.37541884e-01 -2.95249552e-01
-4.79501247e-01 -7.59321809e-01 7.21690416e-01 -5.96813083e-01
1.54595613e+00 -1.65302467e+00 -2.91072652e-02 -1.44798264e-01
2.31777295e-01 3.01376492e-01 -1.19073447e-02 8.97957563e-01
4.87621605e-01 -1.19170606e-01 1.94808021e-01 -2.52502739e-01
4.50843126e-01 3.43899429e-01 -9.01301056e-02 -1.14998087e-01
6.36194870e-02 6.95397019e-01 -1.03472435e+00 -7.48002410e-01
5.10534525e-01 -1.94004849e-01 -6.36058748e-01 7.09859014e-01
-7.80841172e-01 5.31975687e-01 -7.31401145e-01 1.10958494e-01
-2.12456919e-02 -3.29627633e-01 6.50255501e-01 1.10329524e-01
-3.37496012e-01 1.35028791e+00 -8.77814591e-01 1.75120604e+00
-6.54421985e-01 5.38234651e-01 2.90426910e-01 -5.22741616e-01
1.05477703e+00 6.14856839e-01 4.24539089e-01 -6.07764065e-01
5.26189581e-02 -1.18232384e-01 3.93991083e-01 -5.83632529e-01
1.06037056e+00 -1.32534012e-01 -6.11467481e-01 1.01796126e+00
-9.84373037e-03 -4.10855204e-01 4.88861144e-01 4.85761285e-01
1.10396945e+00 -8.80905539e-02 6.23108268e-01 -1.68507040e-01
5.03448606e-01 6.42375529e-01 2.46217191e-01 7.11424828e-01
-5.12515843e-01 -1.27097622e-01 7.33032286e-01 -2.41121143e-01
-7.84007788e-01 -4.48567301e-01 -2.05643043e-01 1.74591684e+00
-7.01832622e-02 -5.91829360e-01 -7.62958884e-01 -6.36329949e-01
-1.68337986e-01 1.03336370e+00 -1.44791663e-01 -2.27734938e-01
-4.77729499e-01 -3.52915138e-01 8.20337534e-01 1.08993709e-01
6.82189643e-01 -1.19556868e+00 -7.65448570e-01 6.47761583e-01
-6.34041429e-01 -1.04594159e+00 -3.32640529e-01 1.31168678e-01
-5.14459729e-01 -1.13361764e+00 -2.71528482e-01 -7.37875938e-01
7.06212670e-02 -6.77596265e-03 1.27334332e+00 -2.53322516e-02
1.10078886e-01 7.76493311e-01 -4.40428287e-01 -3.22307289e-01
-1.47741902e+00 4.16826040e-01 -1.05358206e-01 -5.31347156e-01
6.02928996e-01 7.75068030e-02 -3.78954440e-01 5.41663349e-01
-5.82919955e-01 4.79399532e-01 -3.69603276e-01 1.01357174e+00
-6.94327176e-01 -2.30546027e-01 7.56772041e-01 -1.11563015e+00
1.65905058e+00 -3.67932647e-01 -2.13823408e-01 4.50786233e-01
-5.50357521e-01 2.13399917e-01 3.23132187e-01 -1.23304442e-01
-1.71239221e+00 -1.20128199e-01 -2.31324434e-01 5.74556768e-01
-3.89383703e-01 6.63890600e-01 8.97716954e-02 3.75498563e-01
9.01153684e-01 2.10388258e-01 5.41579902e-01 -2.87365645e-01
5.29660940e-01 1.14874184e+00 2.29458287e-01 -9.98393178e-01
1.04624644e-01 -1.89081594e-01 -5.21796823e-01 -1.03499436e+00
-7.29831636e-01 -7.83866048e-01 -5.93280673e-01 -3.53768319e-01
7.49541461e-01 -7.88891554e-01 -1.07985675e+00 3.14144939e-01
-1.26943648e+00 -7.87394822e-01 -1.75450385e-01 7.71071985e-02
-7.71382093e-01 4.85225618e-01 -8.89333487e-01 -1.33091855e+00
-3.45321715e-01 -9.87397015e-01 8.22034061e-01 2.92437732e-01
-1.27882755e+00 -1.42678082e+00 2.35376567e-01 6.30140603e-01
3.74125719e-01 -1.71444342e-01 1.12396932e+00 -1.20939004e+00
6.74585477e-02 -1.77547801e-02 1.32702425e-01 -1.79920197e-01
4.86408651e-01 -2.81140804e-01 -8.30939233e-01 -9.64157283e-03
-1.32829174e-01 -8.42948914e-01 1.07066538e-02 3.64234708e-02
2.23114997e-01 -6.04226828e-01 -8.99752676e-02 -3.59906644e-01
4.86308634e-01 5.94285488e-01 3.27225447e-01 3.55776221e-01
1.81841582e-01 1.35075021e+00 8.70415747e-01 8.55901659e-01
1.12485731e+00 8.90155137e-01 -8.05633366e-02 1.30361840e-01
1.15553871e-01 -2.91612327e-01 2.99654722e-01 7.24624336e-01
3.97679597e-01 -3.06970209e-01 -1.31273055e+00 6.00264847e-01
-1.93034768e+00 -9.92106795e-01 2.12314595e-02 1.76562130e+00
1.51163566e+00 1.32920071e-01 2.97192991e-01 -1.26772821e-01
6.94152951e-01 5.54217935e-01 -5.16537130e-01 -8.86492908e-01
3.59217763e-01 -1.11261293e-01 -1.30253986e-01 8.82722259e-01
-8.61319482e-01 1.33578634e+00 6.54947758e+00 3.66297454e-01
-6.57126069e-01 -1.93012446e-01 3.50837886e-01 2.62621343e-01
-1.73129857e-01 6.41922578e-02 -8.59346211e-01 1.48200646e-01
1.03851569e+00 -5.59554100e-01 3.88083726e-01 4.92803395e-01
3.83843035e-01 -3.79450828e-01 -1.19249940e+00 3.71840566e-01
-3.42387229e-01 -1.16129923e+00 -2.59105355e-01 -1.15992660e-02
2.33865380e-01 -4.74978477e-01 -3.41777056e-01 7.32897699e-01
1.00031960e+00 -8.21022332e-01 2.49314666e-01 6.80298433e-02
6.04127705e-01 -4.73701894e-01 2.60646999e-01 7.31079757e-01
-9.74731565e-01 -8.33109207e-03 6.88039437e-02 -4.32772070e-01
4.20944154e-01 -3.99328202e-01 -1.67274308e+00 3.77108566e-02
3.23809646e-02 3.75011951e-01 -1.89724401e-01 3.91822606e-01
-2.16711145e-02 5.00558078e-01 -2.31287673e-01 -4.36763465e-01
1.77662015e-01 -1.18514784e-01 5.87811172e-01 1.12574470e+00
-2.96429098e-01 3.89745414e-01 6.17037952e-01 6.19904995e-01
3.67561221e-01 4.91048507e-02 -6.15532935e-01 -3.90015632e-01
9.47349310e-01 1.05690384e+00 -4.39885855e-01 -3.58127266e-01
-3.02509665e-01 6.75393164e-01 3.05944160e-02 8.85900110e-02
-1.18936092e-01 -2.92144924e-01 1.05272734e+00 -7.96695948e-02
-5.22915363e-01 -2.52755225e-01 -3.27746332e-01 -1.11843753e+00
-5.28262794e-01 -1.30959487e+00 5.43081522e-01 -4.88105774e-01
-9.17823195e-01 4.46077585e-01 2.42701679e-01 -8.86578381e-01
-1.22048962e+00 -4.28377360e-01 -6.87432349e-01 1.04418159e+00
-1.12071085e+00 -9.41927910e-01 6.15977775e-03 3.90249968e-01
1.04551566e+00 -2.55229145e-01 1.46218014e+00 -3.56107742e-01
-4.28822905e-01 3.47854406e-01 -3.18609089e-01 3.54907006e-01
8.26333344e-01 -1.46487451e+00 6.64649129e-01 2.89632857e-01
-3.70878190e-01 8.19696665e-01 9.97693062e-01 -6.96371257e-01
-1.27181113e+00 -4.19160843e-01 1.10660326e+00 -5.91026783e-01
6.60769641e-01 -3.68257880e-01 -9.86849666e-01 5.20262003e-01
4.73940641e-01 -8.94751072e-01 8.83393764e-01 7.45341599e-01
3.53296869e-03 7.51532376e-01 -1.07821548e+00 9.09640491e-01
8.45455229e-01 -9.55869079e-01 -1.12968028e+00 2.73388565e-01
4.50229436e-01 -9.58836079e-01 -1.24966419e+00 3.94297950e-02
4.14381921e-01 -7.80033886e-01 7.04036713e-01 -9.95036125e-01
3.44671071e-01 9.14917588e-02 1.83531344e-01 -1.53207743e+00
-2.56323460e-02 -1.01095319e+00 -2.55894363e-01 1.25490785e+00
4.37750041e-01 -5.61068714e-01 8.79557490e-01 1.62526727e+00
-1.75015911e-01 -2.96433270e-01 -7.87767589e-01 -2.33165592e-01
9.17004272e-02 -1.32491261e-01 6.35040402e-01 1.02102149e+00
1.17018437e+00 1.05876017e+00 -2.91349828e-01 -3.11440796e-01
3.79529893e-02 8.51479638e-03 1.04411268e+00 -1.50857806e+00
-4.04554754e-02 -3.66076618e-01 1.31330729e-01 -1.66539979e+00
3.35125268e-01 -5.16613543e-01 3.62336904e-01 -1.75435543e+00
-3.50555718e-01 -6.16672993e-01 4.35329199e-01 6.77996814e-01
-2.09604010e-01 -7.32575774e-01 1.50731459e-01 2.92756200e-01
-8.21629882e-01 7.20691681e-01 1.32064152e+00 -1.33535892e-01
-6.23917758e-01 2.35336825e-01 -6.33788764e-01 5.47246337e-01
8.55034709e-01 8.13573450e-02 -8.01573873e-01 -4.93614078e-02
-1.29613087e-01 8.47098827e-01 -3.63148600e-01 -6.15559101e-01
4.73853141e-01 -5.43441534e-01 -1.00149833e-01 -2.86783248e-01
7.26825833e-01 -3.46159130e-01 -4.53806818e-01 3.62785995e-01
-1.11897826e+00 -1.32653341e-01 2.15372056e-01 2.17322409e-01
-1.01347663e-01 -3.96781445e-01 5.66497922e-01 -2.48219132e-01
-9.05920982e-01 -2.78338671e-01 -1.18988812e+00 3.81507486e-01
7.87146211e-01 -3.68457645e-01 -6.73829556e-01 -8.84705424e-01
-7.01596618e-01 7.37048149e-01 3.97729754e-01 6.87737286e-01
3.17384601e-01 -1.02193642e+00 -4.46819305e-01 1.76919270e-02
4.44241434e-01 -1.84901264e-02 1.28737658e-01 1.23061411e-01
-2.19098002e-01 7.69151211e-01 -3.62013489e-01 -5.40125370e-01
-1.41933692e+00 -4.61137474e-01 4.94600415e-01 -6.14048839e-01
-3.16141188e-01 5.32034218e-01 -6.43612817e-02 -1.00006449e+00
2.53201902e-01 3.01516261e-02 -8.94385695e-01 3.09545875e-01
5.15029669e-01 3.00994217e-01 -2.01178670e-01 -7.22670913e-01
3.75989154e-02 -4.60686952e-01 -4.60297346e-01 -6.33921921e-01
9.48628604e-01 -4.88228768e-01 2.57211085e-02 7.37898648e-01
6.41045868e-01 -4.58102167e-01 -1.27873170e+00 -5.42920649e-01
4.22951818e-01 -3.00462931e-01 -1.34117737e-01 -9.89322543e-01
3.85254696e-02 6.04825139e-01 2.46609449e-01 9.24293518e-01
4.28535104e-01 -1.30209759e-01 7.91419387e-01 5.73227882e-01
5.71096420e-01 -1.41358745e+00 2.09565178e-01 1.32110047e+00
6.93818271e-01 -1.34158039e+00 -2.19252869e-01 -2.11191937e-01
-1.16731894e+00 1.15418327e+00 1.19952953e+00 8.62706602e-01
1.23771220e-01 6.93508610e-02 6.71278238e-01 -3.77229691e-01
-1.30295467e+00 -1.72265470e-01 7.94767812e-02 6.11419022e-01
9.21026707e-01 2.99761742e-02 -3.34852040e-01 1.85840487e-01
-3.22711796e-01 -2.23611832e-01 6.56758010e-01 1.38700700e+00
-7.04733074e-01 -1.55222511e+00 -2.94018447e-01 7.26999879e-01
-1.70587286e-01 -5.26324520e-03 -7.57855892e-01 5.58373570e-01
-9.36299503e-01 1.60404229e+00 9.46231708e-02 -2.39327595e-01
5.92947662e-01 6.71875834e-01 2.62185127e-01 -1.23048878e+00
-1.10927355e+00 -2.28405684e-01 1.15571153e+00 -2.17466682e-01
-2.84256309e-01 -5.51424026e-01 -1.32589805e+00 -4.91308928e-01
-4.87611115e-01 5.73217511e-01 4.68851566e-01 9.98673618e-01
2.66823173e-01 3.20029140e-01 5.02272129e-01 -3.86984348e-01
-1.03443360e+00 -1.30335736e+00 -3.33010614e-01 5.15661836e-01
1.92496747e-01 -5.63055158e-01 1.05858751e-01 -1.18375115e-01] | [12.927939414978027, 7.954102516174316] |
0cd25e60-e6fb-4610-b000-e8038f1cf6a4 | a-simple-and-powerful-global-optimization-for | 2209.09341 | null | https://arxiv.org/abs/2209.09341v2 | https://arxiv.org/pdf/2209.09341v2.pdf | A Simple and Powerful Global Optimization for Unsupervised Video Object Segmentation | We propose a simple, yet powerful approach for unsupervised object segmentation in videos. We introduce an objective function whose minimum represents the mask of the main salient object over the input sequence. It only relies on independent image features and optical flows, which can be obtained using off-the-shelf self-supervised methods. It scales with the length of the sequence with no need for superpixels or sparsification, and it generalizes to different datasets without any specific training. This objective function can actually be derived from a form of spectral clustering applied to the entire video. Our method achieves on-par performance with the state of the art on standard benchmarks (DAVIS2016, SegTrack-v2, FBMS59), while being conceptually and practically much simpler. Code is available at https://ponimatkin.github.io/ssl-vos. | ['Vincent Lepetit', 'Renaud Marlet', 'Yuming Du', 'Yang Xiao', 'Nermin Samet', 'Georgy Ponimatkin'] | 2022-09-19 | null | null | null | null | ['unsupervised-object-segmentation', 'unsupervised-video-object-segmentation'] | ['computer-vision', 'computer-vision'] | [ 1.31545082e-01 7.42661357e-02 -3.31883371e-01 -2.37902269e-01
-5.29884517e-01 -7.77437985e-01 4.71414804e-01 -1.77786186e-01
-6.44981146e-01 5.50302505e-01 -1.03705414e-01 -1.62711635e-01
6.90090135e-02 -3.49550575e-01 -9.10378456e-01 -7.51037419e-01
-4.29829694e-02 3.86595815e-01 7.64136136e-01 -2.58709714e-02
1.49660900e-01 2.37148434e-01 -1.47139323e+00 1.06376477e-01
7.25647151e-01 1.22056568e+00 5.63585103e-01 7.84970224e-01
4.68071513e-02 8.69757295e-01 -1.96366563e-01 -2.19243854e-01
4.89090234e-01 -5.58924317e-01 -1.26329911e+00 5.68488240e-01
6.47001147e-01 -1.39714167e-01 -5.31542301e-01 1.05885839e+00
1.14382757e-02 2.40880683e-01 5.88676810e-01 -9.96622264e-01
-2.31588319e-01 3.85403842e-01 -4.99871343e-01 4.95871514e-01
1.47246188e-02 2.31018767e-01 1.05626416e+00 -8.79075825e-01
9.14513826e-01 8.87583435e-01 5.13607264e-01 2.96662658e-01
-1.35325515e+00 -1.55783638e-01 1.53309286e-01 4.34696794e-01
-1.22416341e+00 -4.26915109e-01 4.74369556e-01 -6.85555935e-01
4.78786975e-01 2.52653241e-01 7.14737356e-01 7.02808976e-01
-3.75859976e-01 1.27144289e+00 1.02531314e+00 -2.78316796e-01
2.51274556e-01 -1.04710065e-01 1.58463597e-01 9.64371502e-01
2.62077823e-02 -2.54038215e-01 -3.47628295e-01 1.84604838e-01
8.55565429e-01 -5.04225120e-02 -5.46442389e-01 -6.49817228e-01
-1.26463401e+00 7.22136855e-01 3.46717417e-01 2.97861546e-01
-1.04113378e-01 2.68762410e-01 2.67836720e-01 1.30219281e-01
4.53602523e-01 2.60539144e-01 -5.85365534e-01 -2.38778099e-01
-1.33832347e+00 1.67045575e-02 7.28868663e-01 1.03445065e+00
1.08603323e+00 -6.05285391e-02 1.39770834e-02 4.80073005e-01
1.56938642e-01 3.65917236e-01 3.03640813e-01 -1.64618385e+00
1.32997334e-01 3.21109384e-01 1.90046459e-01 -7.25590050e-01
-2.90419936e-01 -2.56740987e-01 -5.00854909e-01 1.98488280e-01
7.75384605e-01 -1.42987952e-01 -1.11034179e+00 1.35047424e+00
3.58374149e-01 5.21092296e-01 -3.01577955e-01 1.18175030e+00
7.14214683e-01 6.10143960e-01 -3.32480669e-01 -2.41959900e-01
9.74443316e-01 -1.55751944e+00 -5.12862921e-01 -2.90301919e-01
4.37301904e-01 -6.98163986e-01 9.40208673e-01 4.63575542e-01
-1.29257691e+00 -5.21105647e-01 -8.72455835e-01 -1.84009120e-01
-1.80963039e-01 1.87232763e-01 6.98602319e-01 4.78722483e-01
-1.28662729e+00 8.24364901e-01 -1.01141644e+00 -4.20711398e-01
5.83969951e-01 2.80788273e-01 -2.40941361e-01 9.02942270e-02
-6.71156883e-01 4.14651215e-01 4.91673082e-01 1.48556717e-02
-1.07951343e+00 -6.38970256e-01 -8.85366201e-01 -1.41557768e-01
7.47356772e-01 -5.31367540e-01 1.30662549e+00 -1.39386761e+00
-1.67943203e+00 1.05789208e+00 -2.80043304e-01 -5.91204762e-01
7.49735057e-01 -3.78252774e-01 1.52358279e-01 7.16117620e-01
1.06738940e-01 8.95765305e-01 1.12870634e+00 -1.21122205e+00
-5.69954753e-01 -7.87789971e-02 7.85719231e-02 8.89394060e-02
-1.34147018e-01 1.45730242e-01 -9.78261530e-01 -7.04817593e-01
-1.84817448e-01 -1.03521323e+00 -3.82959098e-01 1.17418222e-01
-4.83223230e-01 -5.10480329e-02 8.22817087e-01 -6.58916235e-01
1.01902938e+00 -2.12354207e+00 4.55240399e-01 -4.62297201e-02
1.34029403e-01 4.06711370e-01 -1.71674922e-01 1.65395349e-01
6.92558512e-02 3.24684493e-02 -7.63755739e-01 -4.84129101e-01
-2.28801042e-01 2.05791861e-01 -6.54022470e-02 8.68364990e-01
1.68445647e-01 1.07833326e+00 -1.04281974e+00 -7.13273048e-01
4.67268020e-01 7.67654926e-02 -5.52289188e-01 5.71580902e-02
-3.63251239e-01 6.44092977e-01 -2.59752005e-01 5.08079112e-01
5.73204875e-01 -7.19704866e-01 9.46218595e-02 5.30540012e-02
-2.05683962e-01 1.59070835e-01 -1.19743359e+00 2.18738675e+00
1.96417794e-02 8.33946645e-01 2.27485716e-01 -1.49385357e+00
4.66322690e-01 7.82842841e-03 7.58180439e-01 -1.94315091e-01
2.56279796e-01 1.03747703e-01 -1.91999629e-01 -5.66650927e-01
9.78335664e-02 3.20384979e-01 1.96781248e-01 1.65168479e-01
5.65210164e-01 -5.09746522e-02 8.66466880e-01 3.27713877e-01
1.02641106e+00 4.13299173e-01 1.87555209e-01 -6.34282589e-01
5.28529763e-01 1.21284097e-01 6.16800904e-01 7.50404894e-01
-1.99685857e-01 8.94950449e-01 5.00495970e-01 -2.12969571e-01
-9.10798311e-01 -1.20918894e+00 -1.07179791e-01 8.82103026e-01
5.39453745e-01 -4.43777710e-01 -9.39259827e-01 -8.67818773e-01
-7.02288672e-02 1.87385708e-01 -6.06624067e-01 4.47691470e-01
-4.66009855e-01 -3.88013303e-01 1.88609138e-01 4.18146700e-01
4.05237705e-01 -9.43902016e-01 -7.19644606e-01 5.24100401e-02
-2.86265552e-01 -1.57190323e+00 -6.69695616e-01 1.24250904e-01
-1.06099319e+00 -1.23728347e+00 -8.09580743e-01 -8.13503146e-01
6.38252258e-01 3.66324127e-01 1.28099501e+00 1.62409604e-01
-5.13229132e-01 5.21083474e-01 -3.42820197e-01 -7.05160201e-02
-1.64415818e-02 8.91348198e-02 -2.15998664e-01 2.62979776e-01
3.34149227e-02 -4.82461900e-01 -8.37681055e-01 4.44461942e-01
-8.84796917e-01 8.36778358e-02 3.68063509e-01 6.38714075e-01
7.93051481e-01 -2.44751662e-01 1.89188823e-01 -7.65934527e-01
-2.60931253e-01 -3.22992682e-01 -7.22302139e-01 -2.18039770e-02
-1.80665344e-01 -1.14134096e-01 7.27503419e-01 -1.97735861e-01
-7.95996189e-01 4.34004486e-01 1.45714059e-01 -6.99556351e-01
-3.41158926e-01 1.28929317e-01 7.49155656e-02 -9.31386873e-02
2.55912334e-01 2.51539409e-01 7.15553537e-02 -5.99165261e-01
5.43694973e-01 3.30466300e-01 7.44383991e-01 -1.96419582e-01
8.25033367e-01 1.04727602e+00 -8.07234347e-02 -9.65796232e-01
-1.00625598e+00 -8.86428297e-01 -9.62587178e-01 -2.99050361e-01
1.00013292e+00 -7.95606315e-01 -5.47610104e-01 4.14159417e-01
-1.01826477e+00 -8.55959415e-01 -4.33089703e-01 4.73048866e-01
-8.36769164e-01 6.54554307e-01 -7.39531100e-01 -4.40137893e-01
-8.91631097e-02 -8.77587676e-01 9.04357910e-01 2.62206852e-01
4.09397185e-02 -9.83025193e-01 -6.70806393e-02 4.75517541e-01
-6.45703822e-02 2.00705156e-01 1.00766256e-01 -3.60172302e-01
-9.27207708e-01 2.53387034e-01 -2.21453831e-01 7.27119982e-01
8.10016990e-02 2.00344279e-01 -8.13552856e-01 -1.88123837e-01
-6.33712262e-02 -2.96144307e-01 1.37357664e+00 6.92796767e-01
1.35990369e+00 -3.11674953e-01 -2.12313235e-01 9.42514062e-01
1.52597427e+00 -1.17557354e-01 5.61295331e-01 2.89416581e-01
7.96930730e-01 5.69853544e-01 5.88517368e-01 2.25439563e-01
3.33566159e-01 6.16177619e-01 5.18349528e-01 -3.23111355e-01
-3.61606628e-01 3.59180309e-02 4.03152287e-01 7.12588727e-01
-2.91576058e-01 -1.09304383e-01 -7.36162364e-01 8.34618211e-01
-2.19587111e+00 -9.69111383e-01 -3.70389909e-01 1.97112286e+00
6.97544515e-01 -4.50700223e-02 4.33515906e-01 5.94045594e-02
6.83847904e-01 3.26788992e-01 -5.34700096e-01 -2.78611444e-02
-2.19557345e-01 1.85422122e-01 7.71204591e-01 5.86530805e-01
-1.46014047e+00 1.10006547e+00 6.22072315e+00 8.02544832e-01
-9.71060693e-01 3.10529947e-01 7.76434064e-01 -2.40151003e-01
7.12041631e-02 9.35307443e-02 -5.45808077e-01 6.34270847e-01
5.66169262e-01 1.97922997e-02 5.66474736e-01 6.56246006e-01
2.90198982e-01 -3.73866528e-01 -9.97096777e-01 8.94217432e-01
1.09446838e-01 -1.56939650e+00 -4.28214669e-01 -1.09910987e-01
1.07531381e+00 5.03700435e-01 -1.23579666e-01 -2.47367129e-01
8.78641382e-02 -7.55133212e-01 8.91242921e-01 3.34436834e-01
4.47125822e-01 -4.46437627e-01 4.62099642e-01 1.74285635e-01
-1.16607165e+00 -3.01741771e-02 -2.15069458e-01 -1.15924394e-02
2.92029053e-01 6.30927861e-01 -3.49106342e-01 4.73720074e-01
1.03155839e+00 1.03526902e+00 -6.24605358e-01 1.24674177e+00
-3.59240770e-01 8.27386260e-01 -4.52831745e-01 2.53464639e-01
4.82437938e-01 -5.43608010e-01 6.45229220e-01 1.53861260e+00
1.03086688e-01 -5.43764643e-02 4.39347357e-01 6.47264302e-01
-1.05136991e-01 1.98919594e-01 -3.45530301e-01 3.45868878e-02
-3.50751635e-03 1.42774892e+00 -1.14852726e+00 -4.96844649e-01
-5.61657965e-01 1.22108817e+00 2.88013637e-01 5.20873070e-01
-8.83307099e-01 -1.37253255e-01 4.73805934e-01 2.42662728e-01
1.08361340e+00 -4.07357544e-01 -2.00009808e-01 -1.42158043e+00
7.44957626e-02 -6.28557384e-01 4.34308261e-01 -5.12670040e-01
-9.80211139e-01 4.24196750e-01 -1.08430952e-01 -1.19333506e+00
-9.57950130e-02 -7.13546932e-01 -5.29825985e-01 2.90595472e-01
-1.66435230e+00 -6.80074453e-01 -3.53424907e-01 7.60363102e-01
6.63756907e-01 1.37330011e-01 3.27958107e-01 2.04553619e-01
-5.25275648e-01 2.20595300e-01 2.37164363e-01 2.98471391e-01
5.62245905e-01 -1.38931859e+00 3.71073335e-01 1.15998805e+00
3.77315819e-01 1.17994741e-01 6.60281479e-01 -3.04025322e-01
-1.16259336e+00 -1.07200527e+00 5.43447077e-01 -2.89376348e-01
1.02341700e+00 -4.27593619e-01 -7.51708269e-01 6.51596248e-01
4.21697676e-01 4.21556830e-01 4.36359495e-01 -2.99853474e-01
-1.10151730e-01 -1.85635500e-02 -7.27410972e-01 3.66894931e-01
1.35183442e+00 -2.57863969e-01 -3.45208585e-01 4.90073383e-01
6.12111449e-01 -3.95835608e-01 -7.71192312e-01 3.55962902e-01
1.35036260e-01 -1.14769602e+00 1.03368890e+00 -5.44915259e-01
5.30690372e-01 -5.49593568e-01 6.75509498e-02 -9.10999238e-01
-2.49344349e-01 -1.07901907e+00 -4.41989809e-01 8.52927029e-01
4.11856055e-01 -3.43950689e-01 8.73593688e-01 1.13203116e-01
-2.49816462e-01 -8.73752832e-01 -8.29544067e-01 -1.04017222e+00
-2.36992657e-01 -2.96989053e-01 -7.74956355e-03 8.48043144e-01
-1.35331810e-01 4.82310355e-02 -2.90641278e-01 1.19568042e-01
1.02385426e+00 3.55045140e-01 6.94980085e-01 -1.08611155e+00
-5.41814983e-01 -6.00731254e-01 -5.57403922e-01 -1.48002589e+00
2.52553165e-01 -9.43336070e-01 1.47684082e-01 -1.55225503e+00
1.91758409e-01 -1.76486999e-01 -3.06467414e-01 4.80095088e-01
-1.41598582e-01 6.61079347e-01 4.24992323e-01 2.93722659e-01
-1.12256229e+00 5.10246634e-01 1.33258176e+00 -1.19632229e-01
-2.07478434e-01 2.34434009e-02 -3.45021188e-01 1.05309367e+00
1.05327427e+00 -3.77171457e-01 -3.13593179e-01 -4.82747912e-01
-2.08239526e-01 -1.57583907e-01 5.33933640e-01 -9.42519188e-01
1.25528470e-01 -1.48703367e-01 8.40815380e-02 -3.70362788e-01
2.37408206e-01 -5.76375186e-01 -1.36267126e-01 3.78106803e-01
-9.39793736e-02 -1.89018145e-01 1.06519550e-01 6.39966488e-01
-2.70224333e-01 -4.47817028e-01 9.21104848e-01 -2.37496883e-01
-9.62570965e-01 4.81070280e-01 -3.18394482e-01 3.61306310e-01
1.16474438e+00 -2.52233475e-01 -1.25498220e-01 -2.51163393e-01
-7.78072059e-01 3.29402655e-01 8.38071346e-01 3.90197277e-01
3.74936134e-01 -1.01195073e+00 -6.52615130e-01 -2.22867683e-01
-1.91916883e-01 1.98071510e-01 1.84473678e-01 1.13897049e+00
-7.63120174e-01 4.12267625e-01 -3.90634201e-02 -9.71763372e-01
-1.11192596e+00 6.83318198e-01 3.06795716e-01 -8.17379877e-02
-7.84897864e-01 8.77573490e-01 3.65842789e-01 3.56005691e-03
1.87852725e-01 -2.90937066e-01 1.03728689e-01 -6.64172992e-02
4.52629864e-01 4.45404053e-01 -2.82350272e-01 -6.32191420e-01
-4.64433134e-01 6.45723224e-01 1.73069894e-01 -6.74743578e-02
1.31154943e+00 -3.38975549e-01 -1.96521908e-01 5.23548484e-01
1.36089373e+00 -2.24323183e-01 -1.77880347e+00 -2.68384278e-01
1.17513679e-01 -6.14489853e-01 1.00389265e-01 -3.71660978e-01
-1.40631330e+00 7.08551586e-01 3.12284440e-01 2.90730357e-01
1.14992595e+00 4.43926007e-01 7.62908220e-01 2.68416852e-01
2.29374275e-01 -1.19011068e+00 2.03446656e-01 4.48691219e-01
5.73382437e-01 -1.35013676e+00 3.97604443e-02 -6.07540965e-01
-6.81178153e-01 1.07351422e+00 2.67317533e-01 -3.70620549e-01
6.82839155e-01 3.20840091e-01 7.54788239e-03 4.36821990e-02
-4.56738800e-01 -6.16456687e-01 4.00368452e-01 3.95783424e-01
2.35063151e-01 -1.68401793e-01 -3.38832170e-01 8.14078972e-02
5.73447831e-02 8.48845840e-02 6.13460839e-01 7.42472172e-01
-4.89472538e-01 -8.61048341e-01 -5.31509100e-03 4.09275621e-01
-6.45490229e-01 -1.25588939e-01 -1.10789992e-01 6.01489186e-01
1.41608521e-01 8.67105722e-01 2.12934002e-01 1.70359947e-02
-1.16685055e-01 -1.65473849e-01 6.30147696e-01 -6.43402696e-01
-1.68544859e-01 3.09400797e-01 -9.54012498e-02 -9.89333749e-01
-1.07529461e+00 -9.21378672e-01 -1.32662272e+00 -1.88180916e-02
-2.32471257e-01 -1.95645471e-03 3.47591281e-01 9.08273220e-01
3.20068181e-01 2.19849601e-01 6.73624396e-01 -1.18596029e+00
6.65082037e-02 -7.38929152e-01 -5.46003163e-01 4.19504732e-01
4.28267330e-01 -5.06442606e-01 -4.29564536e-01 5.59040606e-01] | [9.100717544555664, -0.19488321244716644] |
674daf15-a943-438e-9043-81b07a3f76c6 | learning-polysemantic-spoof-trace-a-multi | 2212.03943 | null | https://arxiv.org/abs/2212.03943v1 | https://arxiv.org/pdf/2212.03943v1.pdf | Learning Polysemantic Spoof Trace: A Multi-Modal Disentanglement Network for Face Anti-spoofing | Along with the widespread use of face recognition systems, their vulnerability has become highlighted. While existing face anti-spoofing methods can be generalized between attack types, generic solutions are still challenging due to the diversity of spoof characteristics. Recently, the spoof trace disentanglement framework has shown great potential for coping with both seen and unseen spoof scenarios, but the performance is largely restricted by the single-modal input. This paper focuses on this issue and presents a multi-modal disentanglement model which targetedly learns polysemantic spoof traces for more accurate and robust generic attack detection. In particular, based on the adversarial learning mechanism, a two-stream disentangling network is designed to estimate spoof patterns from the RGB and depth inputs, respectively. In this case, it captures complementary spoofing clues inhering in different attacks. Furthermore, a fusion module is exploited, which recalibrates both representations at multiple stages to promote the disentanglement in each individual modality. It then performs cross-modality aggregation to deliver a more comprehensive spoof trace representation for prediction. Extensive evaluations are conducted on multiple benchmarks, demonstrating that learning polysemantic spoof traces favorably contributes to anti-spoofing with more perceptible and interpretable results. | ['Di Huang', 'Biao Wang', 'Pengyu Li', 'Binghui Chen', 'Hongyu Yang', 'Kaicheng Li'] | 2022-12-07 | null | null | null | null | ['face-anti-spoofing'] | ['computer-vision'] | [ 7.51334250e-01 -4.20166969e-01 -4.44886118e-01 -1.27108008e-01
-4.47789341e-01 -6.46179676e-01 8.50675285e-01 -3.44920456e-02
4.03186768e-01 3.27456981e-01 1.57827228e-01 -3.37154865e-01
-3.03254515e-01 -8.15999389e-01 -4.00476098e-01 -9.90698695e-01
-1.64727673e-01 2.48900667e-01 -5.61855510e-02 -4.32813108e-01
5.88864870e-02 8.63253057e-01 -1.65645206e+00 6.43340588e-01
6.59613073e-01 9.25047040e-01 -6.07452393e-01 6.08291447e-01
1.74009010e-01 6.08590901e-01 -9.24986601e-01 -8.90057921e-01
1.61952570e-01 -3.41758758e-01 -3.31129402e-01 -1.46842584e-01
7.18219876e-01 -2.36048430e-01 -6.46614492e-01 1.29549527e+00
5.88082671e-01 -3.32562327e-01 5.84032893e-01 -1.69938719e+00
-4.50231224e-01 3.94933462e-01 -6.80740356e-01 3.32048804e-01
8.59029353e-01 3.43826741e-01 9.42847431e-01 -7.19752491e-01
2.83716440e-01 1.75880539e+00 2.89330363e-01 6.31988406e-01
-1.28344214e+00 -1.32249570e+00 -1.91846997e-01 4.90830451e-01
-1.12871897e+00 -7.00519621e-01 1.39443314e+00 -3.20640922e-01
5.83450854e-01 3.33551973e-01 4.70958441e-01 1.84478450e+00
2.90825844e-01 7.73775280e-01 1.38354266e+00 -2.76334941e-01
-3.86461735e-01 2.01554984e-01 -1.08847767e-03 7.80201912e-01
5.07197082e-01 6.11131251e-01 -8.12540472e-01 -4.04906303e-01
2.91681498e-01 2.20221534e-01 -2.87270755e-01 -4.55677599e-01
-9.07606781e-01 8.29918027e-01 3.73739243e-01 2.83619255e-01
-9.14739296e-02 -3.23296487e-01 6.96805239e-01 4.52153385e-01
1.40353650e-01 3.29362750e-01 3.22869010e-02 4.60753411e-01
-9.06355202e-01 7.11292103e-02 8.71106565e-01 1.89095944e-01
6.92232430e-01 1.79717526e-01 -1.09473169e-02 8.12666059e-01
4.30629194e-01 9.15970147e-01 1.14766881e-01 -2.79323012e-01
5.66972077e-01 5.04849017e-01 -4.73779738e-01 -1.93988347e+00
-2.33965605e-01 -4.73711908e-01 -1.09241796e+00 1.70352280e-01
1.93500593e-01 2.78257340e-01 -7.77060449e-01 1.72208917e+00
3.17316115e-01 6.64942801e-01 9.90169123e-02 4.69133794e-01
6.17455125e-01 4.50302422e-01 3.77119593e-02 -2.86427081e-01
1.54678679e+00 -4.70381111e-01 -7.50934064e-01 -2.33359471e-01
8.25152025e-02 -8.27268898e-01 5.85609734e-01 6.46441102e-01
-4.60721493e-01 -3.80492896e-01 -1.33972311e+00 4.66609567e-01
-4.84554648e-01 -2.96250075e-01 7.62908459e-01 1.47299004e+00
-6.23170972e-01 2.07257882e-01 -5.36503792e-01 -4.12039049e-02
8.13014925e-01 4.05370682e-01 -6.12998426e-01 -4.00008798e-01
-1.49292982e+00 6.11466825e-01 5.53440869e-01 7.20906779e-02
-1.43017590e+00 -6.35504723e-01 -8.17144096e-01 -7.36989677e-02
2.58365899e-01 -4.80328470e-01 2.62227148e-01 -6.92808509e-01
-1.45130980e+00 9.43517506e-01 1.02970339e-01 -1.59248859e-01
3.23981732e-01 1.70119941e-01 -1.31188941e+00 3.71303290e-01
-1.68189347e-01 3.63761522e-02 1.58387399e+00 -1.53715050e+00
-2.02413216e-01 -6.96880460e-01 4.04456630e-02 -3.48995626e-01
-6.92023993e-01 5.04766881e-01 -1.02014974e-01 -8.66345763e-01
8.38279910e-03 -7.49443054e-01 3.63981307e-01 -2.38826737e-01
-6.13246739e-01 -1.69790313e-02 1.20498645e+00 -6.46088600e-01
1.23239899e+00 -2.06591797e+00 1.58185765e-01 4.11145091e-01
2.76774794e-01 7.91340470e-01 -2.34313443e-01 4.98164862e-01
-5.26471198e-01 9.78796631e-02 -1.61157012e-01 -2.86798716e-01
-1.94060221e-01 3.09059620e-02 -6.10808134e-01 8.07886124e-01
4.43860680e-01 5.48497677e-01 -1.02890003e+00 -5.18374622e-01
3.98053735e-01 8.24966013e-01 -5.58175266e-01 1.80503026e-01
9.72185656e-02 6.26770020e-01 -2.50070304e-01 1.09352601e+00
1.17686939e+00 -2.58678962e-02 3.09889108e-01 -4.35024858e-01
7.60540247e-01 1.39371112e-01 -1.06393349e+00 1.31215465e+00
-4.83399510e-01 5.05054355e-01 1.02560334e-01 -8.95935655e-01
1.01486349e+00 3.61936659e-01 5.81841469e-01 -5.36422074e-01
3.30695271e-01 2.91180849e-01 3.06502338e-02 -7.04402626e-01
-1.40447542e-03 -1.39035076e-01 -1.93572640e-02 3.74743104e-01
3.32571447e-01 2.01593861e-01 -1.15759484e-01 1.76593199e-01
9.78487551e-01 -2.89392233e-01 1.79556549e-01 8.97271186e-02
1.00864661e+00 -5.13728440e-01 4.48913842e-01 4.43794996e-01
-5.88520944e-01 1.16666116e-01 5.17089844e-01 -3.18729460e-01
-4.62303847e-01 -1.30093348e+00 -2.87322611e-01 1.01862061e+00
4.25842762e-01 -5.66780448e-01 -4.04964805e-01 -1.18203259e+00
1.72464415e-01 3.17716509e-01 -5.63184559e-01 -6.00222528e-01
-6.57912314e-01 -9.83180821e-01 1.20890343e+00 -1.43331289e-02
3.99962097e-01 -8.62337530e-01 -1.05935201e-01 -3.08855474e-01
-1.90945357e-01 -1.29028106e+00 1.07322939e-01 -3.10700446e-01
-4.71829444e-01 -1.49246228e+00 -2.04710186e-01 -4.48703885e-01
3.51664901e-01 5.48641324e-01 6.85791731e-01 2.10583225e-01
-3.73277634e-01 1.45445779e-01 -1.82290420e-01 -2.69817322e-01
-7.52929449e-01 -3.48168433e-01 3.34022343e-01 7.04390585e-01
5.84399164e-01 -6.42216086e-01 -4.04348552e-01 4.50935125e-01
-1.06771708e+00 -3.32209945e-01 4.42041337e-01 9.38744009e-01
-9.42274854e-02 3.07532698e-01 3.87541860e-01 -8.04122865e-01
4.62174773e-01 -9.13408637e-01 -1.11342110e-01 4.22897041e-01
-6.78915203e-01 -7.22803324e-02 4.66603011e-01 -6.73496127e-01
-1.01399755e+00 -4.22639549e-01 -8.88945982e-02 -6.38112366e-01
-3.84924412e-01 6.42543808e-02 -7.51638234e-01 -4.05462623e-01
7.31656134e-01 2.68028885e-01 2.08876938e-01 -2.02505678e-01
2.12949172e-01 6.22978330e-01 4.69242632e-01 -4.30065721e-01
1.37281466e+00 7.22949505e-01 2.27754042e-01 -9.56843555e-01
-3.93344373e-01 -3.70192200e-01 -3.49633217e-01 -3.58747929e-01
4.10007179e-01 -7.24149168e-01 -1.04147494e+00 9.56794620e-01
-1.20412934e+00 4.65930164e-01 4.21950102e-01 1.82493940e-01
-2.04199776e-01 9.02051985e-01 -6.53747797e-01 -7.81583428e-01
-6.12673983e-02 -1.41241336e+00 1.00941885e+00 5.05143851e-02
9.37069207e-02 -9.92442369e-01 1.79372445e-01 6.53292477e-01
2.02861056e-01 6.34275496e-01 8.19573760e-01 -1.04103458e+00
-3.14011991e-01 -6.38570845e-01 -4.37770337e-01 3.88044924e-01
3.59707028e-01 -6.93476573e-02 -1.42645669e+00 -5.89345813e-01
6.02527298e-02 -4.11398947e-01 9.79345202e-01 -3.53258908e-01
9.90131378e-01 -4.74893689e-01 -4.81770039e-01 8.73134792e-01
1.17370725e+00 -1.71470642e-03 5.98029792e-01 1.07813939e-01
1.09800446e+00 1.13101625e+00 1.56271815e-01 3.87753189e-01
8.67854431e-03 8.09412003e-01 9.08472061e-01 3.60026926e-01
-2.71340370e-01 -4.23326105e-01 6.92354441e-01 4.14890885e-01
1.56825066e-01 -5.10999262e-01 -8.27111840e-01 -1.53782377e-02
-1.21655297e+00 -1.20712173e+00 1.19503006e-01 1.99838412e+00
4.28474247e-01 4.30969484e-02 8.33836868e-02 7.19781160e-01
8.31263959e-01 9.49121952e-01 -3.33413184e-01 -2.95848519e-01
-4.47264731e-01 1.40334472e-01 1.90612167e-01 4.98987764e-01
-1.06402409e+00 9.84955549e-01 5.47061062e+00 1.13571334e+00
-1.18408298e+00 2.01065972e-01 2.40548611e-01 1.49593577e-01
-4.43566173e-01 -2.31593966e-01 -7.39328563e-01 6.83932245e-01
7.17483640e-01 3.90872985e-01 5.79931319e-01 3.73299420e-01
-3.61386091e-01 5.39841712e-01 -7.81590343e-01 1.31769156e+00
6.37323499e-01 -1.08623803e+00 4.15249676e-01 2.62636602e-01
4.82801288e-01 -3.42092216e-01 3.92770231e-01 -8.76733661e-02
2.88988441e-01 -1.15685916e+00 3.61162364e-01 1.45041719e-01
7.92645991e-01 -8.77401888e-01 5.17112136e-01 1.23919316e-01
-1.10081208e+00 -5.01057386e-01 -4.88310233e-02 4.49314237e-01
1.47393852e-01 5.50645411e-01 -8.46050739e-01 9.24226761e-01
4.73779678e-01 8.19837451e-01 -5.66879869e-01 5.01261890e-01
-5.60445070e-01 5.79736471e-01 -2.27504596e-01 3.68394941e-01
-2.99783349e-01 2.23555639e-01 9.73052740e-01 1.27584434e+00
9.54307765e-02 -3.51063490e-01 4.89740111e-02 5.54423988e-01
9.60628912e-02 -1.15351021e-01 -9.44066763e-01 -2.64910340e-01
6.47331476e-01 1.15526378e+00 -3.99118870e-01 -4.47010361e-02
-5.40596470e-02 9.41932678e-01 9.41521674e-02 3.31690490e-01
-7.96292007e-01 -2.96068907e-01 1.07416606e+00 -1.21923856e-01
-2.89260335e-02 -1.86605932e-04 -1.09663904e-01 -1.36603057e+00
-2.61563927e-01 -1.37634134e+00 8.45430791e-01 -1.83965147e-01
-1.44319570e+00 7.41262853e-01 -7.66223297e-02 -1.00731301e+00
-2.76449144e-01 -9.65183675e-01 -5.06430149e-01 7.27178335e-01
-1.73216915e+00 -1.62373066e+00 -2.77063042e-01 1.02659309e+00
2.31811956e-01 -6.13427460e-01 8.77631068e-01 4.72375393e-01
-1.04101837e+00 1.11426485e+00 -4.89046931e-01 2.17716038e-01
7.53585160e-01 -5.47647774e-01 5.99083267e-02 1.02486849e+00
2.63008267e-01 8.02984238e-01 7.21425712e-01 -7.18918443e-01
-1.69211876e+00 -6.96745217e-01 6.23117447e-01 -5.46699524e-01
8.85585010e-01 -3.86169612e-01 -9.06125486e-01 3.42829973e-01
1.57875083e-02 3.33587140e-01 9.04961705e-01 -1.54659405e-01
-1.52951276e+00 -3.45590740e-01 -1.31717563e+00 4.29048806e-01
9.20826077e-01 -1.17374897e+00 -2.04381675e-01 1.11015588e-01
5.16736507e-01 -8.11565109e-03 -6.31978631e-01 6.80880189e-01
8.25513482e-01 -1.37998116e+00 1.55771661e+00 -8.03887725e-01
3.51946443e-01 -7.86384642e-02 -3.76566321e-01 -9.38914359e-01
-1.48192421e-01 -9.29129899e-01 -7.72423804e-01 1.44951832e+00
-1.64464995e-01 -8.31626594e-01 7.33658850e-01 -4.09437686e-01
4.66198921e-01 -5.72269082e-01 -9.14730012e-01 -7.06827044e-01
-2.34905615e-01 -6.38026297e-01 9.93362367e-01 1.21039450e+00
-8.13306794e-02 8.52366164e-02 -8.30992162e-01 7.67049372e-01
1.14185441e+00 -1.00276627e-01 7.52540648e-01 -1.38372004e+00
-4.48352039e-01 -6.23675764e-01 -7.07908511e-01 -7.12157488e-01
5.99234998e-01 -1.06611788e+00 -3.47378016e-01 -4.88840669e-01
6.61978573e-02 -3.94395232e-01 -6.77429974e-01 1.99915990e-01
-2.32251734e-01 7.47760177e-01 2.59813964e-01 3.18715423e-01
-9.17119756e-02 2.24547490e-01 1.06327784e+00 -6.03559017e-01
1.70071021e-01 1.13963522e-01 -6.33952975e-01 5.65145314e-01
6.69513881e-01 -4.78536844e-01 -4.19479430e-01 -2.35016242e-01
2.52956510e-01 2.19644919e-01 7.42437482e-01 -9.80911195e-01
5.59920035e-02 -1.28839225e-01 2.53045619e-01 -2.63451010e-01
5.65117121e-01 -9.79709327e-01 1.74730159e-02 7.87145913e-01
-1.27853036e-01 -2.46194124e-01 1.09912694e-01 9.33363974e-01
-1.25444055e-01 2.17711136e-01 9.53319430e-01 3.50269973e-01
-3.95531744e-01 5.55467188e-01 2.52405256e-02 -2.39885882e-01
9.65324521e-01 -4.00216907e-01 -6.49667561e-01 -7.32886419e-02
-4.37250584e-01 -3.27554405e-01 3.15870315e-01 6.90679312e-01
9.34149384e-01 -1.39491057e+00 -7.54549921e-01 9.32845414e-01
3.90328288e-01 -8.09253275e-01 4.96578693e-01 6.74275696e-01
-1.99435189e-01 2.27105230e-01 -5.66455007e-01 -6.86181962e-01
-1.45546567e+00 1.02715218e+00 1.82006955e-01 -4.76269335e-01
-1.40975803e-01 8.00066710e-01 2.16003843e-02 -2.05694214e-01
2.07107395e-01 6.36341691e-01 -5.51041424e-01 4.17585582e-01
9.30824876e-01 6.90607667e-01 -6.64076880e-02 -1.28520858e+00
-5.80327332e-01 6.06279373e-01 -8.61588195e-02 1.47659793e-01
9.61143076e-01 -1.36972860e-01 -4.10525471e-01 -1.08473197e-01
1.46890438e+00 3.94696921e-01 -9.88769293e-01 -3.74266267e-01
-1.09381929e-01 -8.71208549e-01 -3.38636726e-01 -6.34839177e-01
-1.26876354e+00 1.23868215e+00 6.16932690e-01 4.31761056e-01
1.34652269e+00 -6.75648451e-02 8.06903362e-01 -9.46483016e-02
4.67100620e-01 -1.41516462e-01 5.69865048e-01 2.03699276e-01
7.01804221e-01 -1.31794035e+00 -1.49845362e-01 -8.30978453e-01
-3.06118280e-01 1.06945050e+00 3.83358508e-01 -6.08413257e-02
7.27121770e-01 1.70114890e-01 -1.89085066e-01 -4.25817430e-01
-1.92253724e-01 2.85877049e-01 3.66953909e-01 1.08904099e+00
-1.00189343e-01 1.97211698e-01 3.41405869e-01 2.44565919e-01
-1.18931852e-01 -4.85084325e-01 5.15181920e-04 5.36130667e-01
3.43093053e-02 -1.34280622e+00 -7.09431469e-01 2.38162978e-03
-5.87425351e-01 1.52751803e-01 -4.38096911e-01 2.57938981e-01
3.53302717e-01 1.32869995e+00 -4.00283962e-01 -1.09844530e+00
-5.86467907e-02 -1.17442131e-01 5.69396615e-01 -3.56267124e-01
-4.29327399e-01 -3.48822922e-01 -8.76387805e-02 -7.82007575e-01
-5.00678897e-01 -3.67081851e-01 -4.74992663e-01 -6.53898835e-01
-3.41939420e-01 -9.86378938e-02 6.02043331e-01 8.90609086e-01
3.44442964e-01 4.15245146e-01 1.17934704e+00 -9.13244009e-01
-6.52658165e-01 -5.77649295e-01 -4.01783794e-01 6.25868082e-01
9.20115530e-01 -1.06322122e+00 -5.63899815e-01 -2.88993567e-01] | [13.064651489257812, 1.2070266008377075] |
bb8bcac3-90eb-4e1c-b63c-664e98741c46 | joint-incremental-disfluency-detection-and-1 | null | null | https://aclanthology.org/Q14-1011 | https://aclanthology.org/Q14-1011.pdf | Joint Incremental Disfluency Detection and Dependency Parsing | We present an incremental dependency parsing model that jointly performs disfluency detection. The model handles speech repairs using a novel non-monotonic transition system, and includes several novel classes of features. For comparison, we evaluated two pipeline systems, using state-of-the-art disfluency detectors. The joint model performed better on both tasks, with a parse accuracy of 90.5{\%} and 84.0{\%} accuracy at disfluency detection. The model runs in expected linear time, and processes over 550 tokens a second. | ['Mark Johnson', 'Matthew Honnibal'] | 2014-01-01 | null | null | null | tacl-2014-1 | ['transition-based-dependency-parsing'] | ['natural-language-processing'] | [-2.50806153e-01 3.78166705e-01 -5.16541563e-02 -3.17464083e-01
-1.37737703e+00 -9.01924551e-01 2.31590867e-01 4.14364070e-01
-5.49666882e-01 5.59065759e-01 4.82962936e-01 -7.10047960e-01
7.00979292e-01 -2.93655664e-01 -4.20777708e-01 -2.52510816e-01
-2.79453337e-01 4.87030387e-01 7.70775914e-01 -1.48105815e-01
-7.33702481e-02 2.31693700e-01 -1.11282778e+00 6.07481956e-01
7.09046066e-01 5.26470065e-01 3.73981416e-01 1.20664763e+00
-7.83560536e-05 1.25389397e+00 -1.28322971e+00 -2.70782828e-01
-4.48943049e-01 -7.96222687e-02 -1.33269167e+00 -4.79251385e-01
2.27304876e-01 -3.27051133e-01 -5.21202266e-01 7.82707632e-01
4.83482182e-01 2.88208183e-02 1.22789614e-01 -5.46808064e-01
-6.86836720e-01 8.59756470e-01 1.27069101e-01 1.23057580e+00
8.93817782e-01 4.77844104e-02 1.06861472e+00 -8.33489776e-01
5.96230030e-01 1.32278752e+00 3.92262280e-01 8.10638905e-01
-1.14006734e+00 -5.87284446e-01 3.22857410e-01 4.20938450e-04
-9.17715251e-01 -7.32945144e-01 2.23427415e-01 -2.80460089e-01
2.44027853e+00 1.02127627e-01 6.18906952e-02 9.41646516e-01
2.51084983e-01 7.25901425e-01 9.25045609e-01 -6.49484038e-01
-1.46587566e-02 -3.49485308e-01 8.77137721e-01 7.31066227e-01
-1.13657922e-01 1.23460755e-01 -9.38185036e-01 -1.69789866e-01
3.13618243e-01 -8.14399362e-01 -4.12387043e-01 9.90099847e-01
-7.31777012e-01 5.68048060e-01 -2.60059714e-01 5.93461394e-01
7.28616863e-02 -1.33925289e-01 6.61545575e-01 4.29815859e-01
7.85722315e-01 -2.09485972e-03 -7.74120808e-01 -8.28287363e-01
-7.75094748e-01 8.99899974e-02 9.21583474e-01 1.05369449e+00
3.72024804e-01 2.21882150e-01 -2.74378538e-01 8.70298326e-01
6.52691573e-02 3.21144700e-01 3.47998410e-01 -8.83753240e-01
9.71713960e-01 1.68154433e-01 1.00107104e-01 3.91580770e-03
-7.63953924e-01 -1.04958050e-01 -2.09770441e-01 2.06843000e-02
4.40297991e-01 -2.46631235e-01 -1.22875214e+00 1.92343926e+00
-2.98524406e-02 1.52446534e-02 1.65497899e-01 4.62591529e-01
7.45727122e-01 7.58449733e-01 6.08457923e-01 -7.56312132e-01
1.42044735e+00 -8.08010042e-01 -1.12486243e+00 -4.82457936e-01
9.88348186e-01 -1.03233802e+00 1.12854946e+00 4.43775922e-01
-1.71513522e+00 -4.75033551e-01 -9.13693190e-01 -3.68708551e-01
1.74543276e-01 -2.28933766e-01 4.46680963e-01 8.52464914e-01
-1.38214087e+00 6.42144740e-01 -1.15650105e+00 -1.47007450e-01
-1.75868019e-01 2.24316046e-01 -4.01141733e-01 1.70097247e-01
-1.34275103e+00 1.12417555e+00 2.61610627e-01 -2.71272361e-01
-6.05277598e-01 -4.32911962e-01 -9.36041474e-01 1.71823606e-01
-1.34616330e-01 -3.80728133e-02 1.95231593e+00 -3.04003775e-01
-1.57393873e+00 1.00405157e+00 -6.17630124e-01 -3.84965777e-01
-9.86404121e-02 -6.12594128e-01 -1.04112077e+00 1.73120737e-01
-3.19115329e-03 -3.53395976e-02 1.92663744e-01 -6.52296484e-01
-8.11062753e-01 -1.48417056e-01 -1.87503949e-01 -4.02486436e-02
-9.56375245e-03 1.03031683e+00 -2.87267387e-01 -7.77183294e-01
1.50711626e-01 -6.48692071e-01 2.93866485e-01 -1.13774431e+00
-4.20590281e-01 -6.60706162e-01 5.81429660e-01 -1.30117643e+00
1.84609365e+00 -2.06039834e+00 -2.58684950e-03 -3.07392240e-01
-2.47761700e-02 7.24936903e-01 3.35071124e-02 4.00215954e-01
-2.21169382e-01 5.68787038e-01 -3.69680792e-01 -8.56789768e-01
-3.23490113e-01 3.26771379e-01 -1.40926242e-01 1.85353592e-01
5.32424688e-01 5.02797306e-01 -9.77924705e-01 -2.50951082e-01
2.13727579e-02 1.36180699e-01 -5.20268559e-01 7.45916545e-01
-4.97446507e-02 2.59729743e-01 1.02775328e-01 8.75514388e-01
6.90186739e-01 4.16083813e-01 6.00026309e-01 4.51485842e-01
-4.97328579e-01 1.40526783e+00 -6.54976964e-01 1.67704022e+00
-4.54089493e-01 4.01006669e-01 3.30319554e-01 -4.67694998e-01
7.48062074e-01 9.03497279e-01 -2.20812708e-01 -5.65009713e-01
-4.01368588e-02 2.65426308e-01 1.14289317e-02 -6.27200842e-01
4.08594996e-01 -1.98765069e-01 -4.17889744e-01 3.68603677e-01
4.63802278e-01 7.40668401e-02 3.09517771e-01 3.28821301e-01
1.99704874e+00 -2.61112720e-01 2.30427086e-01 -3.44975531e-01
2.21587405e-01 -1.42984316e-02 9.72481787e-01 6.65204644e-01
-5.47232568e-01 5.89890957e-01 7.16333508e-01 -3.86628360e-01
-5.17824113e-01 -1.55056334e+00 1.66186392e-02 1.23008919e+00
-4.85409647e-01 -8.45556557e-01 -1.01837540e+00 -9.04671371e-01
-4.41149056e-01 9.75831807e-01 -1.92856267e-01 -1.38103247e-01
-1.31020689e+00 -7.28626609e-01 1.15148056e+00 9.55909729e-01
1.24246709e-01 -1.68077898e+00 -3.40516537e-01 5.87144375e-01
-7.10213125e-01 -1.25734782e+00 -5.55773795e-01 6.94869339e-01
-6.96542084e-01 -8.22520614e-01 1.53324544e-01 -1.27049303e+00
-6.78960383e-02 -1.55894458e-01 1.71525002e+00 4.19545084e-01
-8.52214321e-02 -1.50098026e-01 -7.82103002e-01 2.44182646e-02
-1.09017515e+00 5.84427007e-02 1.06149256e-01 -9.49454308e-01
5.66461086e-01 -5.15433192e-01 -1.40440449e-01 -8.73906985e-02
-4.26884085e-01 -6.17063582e-01 2.59085864e-01 7.26396799e-01
4.57774043e-01 -4.23239291e-01 6.55218542e-01 -9.94012415e-01
7.25865006e-01 -5.12796342e-01 -8.51102695e-02 1.35206610e-01
-4.29900825e-01 3.01368125e-02 6.68746829e-01 -1.38238788e-01
-1.38619649e+00 2.84968931e-02 -9.63275015e-01 -7.61086568e-02
-5.09616435e-01 1.36568531e-01 -3.92956853e-01 6.84895575e-01
5.39455175e-01 -2.31741726e-01 -4.85048324e-01 -1.04755914e+00
3.23837280e-01 7.22290456e-01 9.72198665e-01 -3.65875781e-01
1.46349907e-01 -1.50305927e-01 -8.06475461e-01 -5.76963186e-01
-6.43836975e-01 -6.36320055e-01 -6.53040707e-01 2.63841033e-01
1.02033734e+00 -8.12793672e-01 -5.56915820e-01 7.99279988e-01
-1.71632922e+00 -4.97462481e-01 2.50354147e-04 4.12878066e-01
-4.49380606e-01 5.70254505e-01 -1.55537784e+00 -8.11774671e-01
-6.51407659e-01 -8.72421741e-01 7.58841276e-01 -1.57925770e-01
-3.24126124e-01 -6.87987864e-01 4.20138568e-01 1.68624714e-01
3.00107058e-02 -3.07507753e-01 9.66279566e-01 -7.20531523e-01
1.12059377e-01 8.40363577e-02 1.28574036e-02 6.25944853e-01
1.15862042e-01 -1.95639506e-02 -1.28441501e+00 -1.33374572e-01
-2.40433458e-02 -3.20570171e-01 1.15509474e+00 2.49462157e-01
7.99743414e-01 -1.33496717e-01 -2.37092376e-01 2.89112836e-01
7.95434535e-01 4.33339775e-01 6.24112844e-01 -4.77964245e-02
4.18177754e-01 3.94201577e-01 5.48844993e-01 2.80003458e-01
6.21939957e-01 5.57839692e-01 3.39685202e-01 2.41874024e-01
-5.06009281e-01 -1.15867950e-01 8.04297149e-01 1.17113149e+00
-2.56007202e-02 -5.85545123e-01 -1.15870714e+00 7.59013593e-01
-1.55656660e+00 -1.02672184e+00 -4.96290445e-01 1.91425180e+00
1.15693772e+00 6.30451679e-01 1.72706991e-01 3.89731497e-01
9.26112056e-01 1.32949725e-01 6.76755011e-02 -1.19756997e+00
-1.34500608e-01 7.98888922e-01 2.11441755e-01 9.98719990e-01
-1.18171132e+00 1.67237210e+00 7.80871964e+00 3.66167426e-01
-7.17030287e-01 6.15228534e-01 3.31860095e-01 -4.78418283e-02
-6.05409406e-02 7.21669346e-02 -1.04101884e+00 3.96975398e-01
1.75134599e+00 2.58185059e-01 2.82773256e-01 5.44294536e-01
8.89568627e-02 -2.22867399e-01 -7.74715781e-01 4.41739321e-01
-6.06033318e-02 -1.06626976e+00 -4.59494174e-01 -5.46065271e-01
6.19909205e-02 3.00048023e-01 -3.68101478e-01 4.57735449e-01
6.51450038e-01 -7.51963019e-01 1.01069999e+00 9.39475968e-02
1.06797373e+00 -1.00139463e+00 5.48114181e-01 2.75349647e-01
-1.31280529e+00 -3.14519592e-02 -1.49063081e-01 -3.88129681e-01
8.26788127e-01 8.91584218e-01 -8.38355362e-01 2.50402898e-01
1.11203694e+00 2.91939706e-01 -3.47253472e-01 5.00971258e-01
-8.02480221e-01 1.31564069e+00 -2.76886582e-01 3.61714453e-01
-2.05946073e-01 5.76075375e-01 5.96249342e-01 2.07398129e+00
-1.17565349e-01 3.61277789e-01 2.56307721e-01 9.95816216e-02
-1.17615685e-01 -1.07550353e-01 -2.69978344e-02 3.79797816e-01
9.55222309e-01 1.03160298e+00 -5.71233511e-01 -3.43977451e-01
-3.98790985e-01 1.26038051e+00 1.06338251e+00 -4.77376096e-02
-7.95538902e-01 -3.72357488e-01 9.34901595e-01 -2.20488802e-01
1.74122721e-01 -5.48586011e-01 -2.44311348e-01 -9.34657991e-01
1.89954668e-01 -5.45851469e-01 8.17543566e-01 -2.33107314e-01
-1.03292906e+00 1.20527506e+00 -1.59602523e-01 -5.54291904e-01
-5.91502547e-01 -5.99588454e-01 -8.65866661e-01 9.67396259e-01
-1.61677039e+00 -8.35312426e-01 2.09192470e-01 5.74617743e-01
6.72216296e-01 2.22617343e-01 1.25563693e+00 4.12967473e-01
-9.54390287e-01 8.62625539e-01 -4.74238157e-01 3.08434933e-01
8.51525664e-01 -1.62975669e+00 1.45724642e+00 1.33838403e+00
-5.76007999e-02 5.77999532e-01 5.60624182e-01 -8.33799243e-01
-6.62809432e-01 -1.08750308e+00 1.91208744e+00 -6.27032697e-01
5.63650072e-01 -5.38904667e-01 -1.18766737e+00 8.94617319e-01
3.17116886e-01 1.90327898e-01 5.32797992e-01 5.41188419e-01
-6.42739356e-01 4.91341114e-01 -1.23736572e+00 -2.98473816e-02
1.50782537e+00 -6.56727016e-01 -8.22538435e-01 3.31840128e-01
1.08484387e+00 -7.03551590e-01 -7.83682346e-01 2.55102068e-01
1.49654627e-01 -1.12888420e+00 5.73261857e-01 -6.37988925e-01
-9.64373201e-02 1.96433038e-01 -1.93608671e-01 -1.38040507e+00
-7.30258346e-01 -8.34499776e-01 -5.86425126e-01 1.47147977e+00
6.88888669e-01 -5.72700918e-01 1.02769151e-01 4.12707508e-01
-1.02450311e+00 -3.66838813e-01 -1.43875551e+00 -9.23250020e-01
4.06278640e-01 -7.12324560e-01 2.47894049e-01 4.73692298e-01
7.18658566e-01 2.98898518e-01 -7.34286532e-02 3.90201688e-01
-1.18193299e-01 -3.13466787e-01 -2.64850944e-01 -9.04909730e-01
-2.95683473e-01 -1.03349686e-01 -8.93503502e-02 -8.67706418e-01
6.07769370e-01 -7.59236991e-01 3.24318141e-01 -1.32265544e+00
-2.37588525e-01 -5.41728139e-01 -4.06096071e-01 8.34016204e-01
-5.02296150e-01 1.13490716e-01 6.78168610e-02 1.07349999e-01
-3.28819782e-01 -1.18128508e-01 3.54561478e-01 4.51202780e-01
-3.66773009e-01 -8.33482045e-05 -3.96113515e-01 8.06705654e-01
1.19254959e+00 -6.54844165e-01 4.22195392e-03 -9.04942870e-01
-1.56583026e-01 3.25737059e-01 -1.70809567e-01 -8.68866563e-01
-7.09632710e-02 9.70451683e-02 -8.55737329e-02 -4.31099921e-01
1.60659119e-01 3.03677678e-01 -5.05667448e-01 3.61291528e-01
-9.79033485e-03 6.58206284e-01 4.26417738e-01 1.90483063e-01
-1.78829834e-01 -1.46966740e-01 8.66399527e-01 -2.45428681e-01
-5.91021597e-01 -1.64274544e-01 -1.02135789e+00 4.84336644e-01
5.69574535e-01 3.26794297e-01 -7.56757915e-01 -2.31811050e-02
-1.01290441e+00 -2.03855395e-01 4.14194427e-02 6.55219495e-01
3.33833933e-01 -5.69911599e-01 -6.86806619e-01 2.81061143e-01
-7.30353594e-02 -3.27683091e-01 1.34033084e-01 4.14331734e-01
-4.55527276e-01 4.49805111e-01 2.07852378e-01 -4.34496477e-02
-1.33959937e+00 4.64008629e-01 2.48217165e-01 -6.00308478e-01
-6.62410140e-01 1.30660748e+00 -3.42339993e-01 -1.37806922e-01
1.41660467e-01 -6.87128305e-01 -1.90606207e-01 -9.34017152e-02
8.49790454e-01 5.32332778e-01 8.45675290e-01 -7.42923141e-01
-8.53771567e-01 -1.16315253e-01 -3.56687784e-01 -3.57474566e-01
8.53671491e-01 -4.67792340e-02 -6.72774985e-02 2.82505304e-01
7.35411406e-01 3.94055605e-01 -1.01782036e+00 4.98128273e-02
3.59929591e-01 -2.05718316e-02 5.17581105e-02 -1.31872046e+00
-6.00257993e-01 7.84851730e-01 3.68037045e-01 4.11705554e-01
1.07735205e+00 4.68762785e-01 1.06212258e+00 -1.44072650e-02
2.51634270e-01 -1.20156133e+00 -1.66808411e-01 1.10555744e+00
6.47380948e-01 -7.88450480e-01 -6.86802804e-01 -9.54387426e-01
-4.09623802e-01 9.23327208e-01 7.19035387e-01 -1.10370636e-01
7.02357650e-01 1.00801051e+00 3.08153063e-01 2.09122777e-01
-1.03304565e+00 -4.00943816e-01 -3.71164292e-01 6.74988508e-01
9.83199060e-01 4.46171552e-01 -7.03406751e-01 1.07958353e+00
-1.93029135e-01 -5.89992285e-01 4.67855483e-01 1.28450012e+00
-7.65293658e-01 -1.55802572e+00 -1.75953880e-01 1.18278310e-01
-1.00442874e+00 -5.50295293e-01 -2.99058795e-01 4.72422421e-01
9.96440277e-02 1.69827569e+00 2.88246870e-01 -4.84023571e-01
6.47248447e-01 5.70698500e-01 3.93654048e-01 -1.10580933e+00
-1.27433431e+00 3.63660082e-02 8.40167761e-01 -7.82668531e-01
3.74498479e-02 -9.36840117e-01 -1.96286809e+00 -2.02822670e-01
-4.13499653e-01 1.02927715e-01 2.14281589e-01 1.06970930e+00
4.23459172e-01 4.11759615e-01 5.14788508e-01 -4.65954363e-01
-3.26425046e-01 -1.38210666e+00 -4.71797943e-01 2.16750190e-01
3.85842711e-01 -4.22334164e-01 -5.43462813e-01 6.91555440e-02] | [10.476283073425293, 9.832993507385254] |
610a0c58-fee5-4a89-b025-1d0cfc916128 | cover-tree-compressed-sensing-for-fast-mr | 1706.07834 | null | http://arxiv.org/abs/1706.07834v2 | http://arxiv.org/pdf/1706.07834v2.pdf | Cover Tree Compressed Sensing for Fast MR Fingerprint Recovery | We adopt data structure in the form of cover trees and iteratively apply
approximate nearest neighbour (ANN) searches for fast compressed sensing
reconstruction of signals living on discrete smooth manifolds. Levering on the
recent stability results for the inexact Iterative Projected Gradient (IPG)
algorithm and by using the cover tree's ANN searches, we decrease the
projection cost of the IPG algorithm to be logarithmically growing with data
population for low dimensional smooth manifolds. We apply our results to
quantitative MRI compressed sensing and in particular within the Magnetic
Resonance Fingerprinting (MRF) framework. For a similar (or sometimes better)
reconstruction accuracy, we report 2-3 orders of magnitude reduction in
computations compared to the standard iterative method which uses brute-force
searches. | ['Zhouye Chen', 'Yves Wiaux', 'Mohammad Golbabaee', 'Mike E. Davies'] | 2017-06-23 | null | null | null | null | ['magnetic-resonance-fingerprinting'] | ['medical'] | [ 9.84691858e-01 5.57592690e-01 -5.08859940e-02 -1.93836957e-01
-9.96458113e-01 -2.31905043e-01 1.07724860e-01 1.19802721e-01
-2.04227775e-01 6.22249842e-01 2.92441875e-01 -5.01122773e-01
-6.06437862e-01 -4.51990873e-01 -8.23388338e-01 -5.53358436e-01
-7.36813009e-01 3.16322237e-01 1.51491106e-01 -1.07081428e-01
2.97730118e-01 6.95171833e-01 -9.91849124e-01 -1.10839233e-01
7.51352310e-01 1.05633855e+00 4.50266860e-02 6.94641829e-01
1.25556260e-01 5.10940671e-01 4.04177248e-01 -1.63531929e-01
4.30172980e-01 -5.11848807e-01 -1.08341396e+00 1.42889708e-01
3.33461851e-01 -4.32863891e-01 -5.67674160e-01 1.20125484e+00
5.15446186e-01 7.06381872e-02 4.96249676e-01 -6.76839650e-01
-2.23384321e-01 6.83701396e-01 -6.02801025e-01 3.07851166e-01
4.99565214e-01 -6.21785104e-01 4.06887949e-01 -1.19446480e+00
9.43340540e-01 8.89641464e-01 1.27004981e+00 3.59194547e-01
-1.39322674e+00 -1.24474183e-01 -3.78535837e-01 9.50509682e-02
-1.47406888e+00 -8.16468596e-01 6.82274520e-01 -2.99088866e-01
6.73635483e-01 6.45856380e-01 7.02059448e-01 2.78886110e-01
3.77368808e-01 4.35948431e-01 1.07649386e+00 -5.93112588e-01
4.53333259e-01 -1.97724164e-01 -1.82938669e-02 8.91161501e-01
3.27569544e-01 3.23419958e-01 -3.58643740e-01 -7.67188430e-01
8.95446599e-01 2.18031302e-01 -4.60671544e-01 -6.77184522e-01
-1.49341297e+00 9.98321950e-01 3.68918806e-01 2.62107223e-01
-5.42297482e-01 3.99666935e-01 3.20961654e-01 4.92311418e-01
5.00878870e-01 1.32175222e-01 -5.77319935e-02 1.22197904e-02
-1.30268764e+00 5.04609756e-02 9.62497771e-01 7.62619615e-01
6.46121383e-01 -1.95983917e-01 2.61664152e-01 4.35833305e-01
3.95684153e-01 4.85682070e-01 2.90277660e-01 -1.34087372e+00
3.49934429e-01 -7.28019029e-02 -1.88082654e-03 -1.50201285e+00
-4.54975307e-01 -3.00869018e-01 -1.01152802e+00 3.85515764e-02
1.69857606e-01 -4.21117023e-02 -5.95689356e-01 1.42612016e+00
7.23933697e-01 3.12228113e-01 -6.76605925e-02 8.24568450e-01
4.65057045e-02 3.39471549e-01 -5.14333963e-01 -7.64246225e-01
1.09657764e+00 -2.91054606e-01 -6.49552286e-01 8.13839063e-02
8.69513392e-01 -6.08279228e-01 3.50374877e-01 3.81755561e-01
-1.12845957e+00 1.40955850e-01 -1.26618373e+00 1.17929317e-01
1.77917778e-01 -3.77579361e-01 5.06773472e-01 7.92488873e-01
-1.32248759e+00 1.34733689e+00 -1.16004241e+00 -9.30127427e-02
5.82940280e-01 5.73427022e-01 -4.97831613e-01 -3.21220070e-01
-8.43401074e-01 6.55263424e-01 -1.04394995e-01 1.88350547e-02
-5.20162702e-01 -7.03447580e-01 -8.22066188e-01 -4.46684569e-01
2.19348576e-02 -4.81273562e-01 6.44555032e-01 -2.75077194e-01
-1.38892365e+00 7.99384892e-01 -1.84059203e-01 -8.41361344e-01
6.16313577e-01 7.80830234e-02 -1.95045516e-01 7.98656404e-01
-1.39833674e-01 2.65942067e-01 1.04024565e+00 -5.31983435e-01
2.23099068e-01 -8.20532799e-01 -3.63805085e-01 1.24568760e-01
-1.41158313e-01 -1.49933219e-01 3.13215822e-01 -6.48008227e-01
1.00327313e+00 -1.09552515e+00 -8.04295421e-01 6.23414516e-01
-1.40636861e-01 4.30296928e-01 6.80792511e-01 -9.76420999e-01
1.17096257e+00 -2.22904372e+00 4.13781226e-01 5.96526980e-01
4.93904382e-01 -2.63655156e-01 1.95596963e-01 3.57303113e-01
-1.42882153e-01 -3.09042484e-01 -7.89214492e-01 -1.41713783e-01
-3.14488739e-01 5.95310405e-02 -2.17177838e-01 1.25116622e+00
-2.98775107e-01 6.88621700e-01 -8.30545545e-01 -7.01451123e-01
-4.00604215e-03 4.46861982e-01 -6.86546862e-01 -4.16554838e-01
4.43470478e-01 4.14289802e-01 -6.99243784e-01 4.19647008e-01
8.28165770e-01 -4.40218419e-01 1.22824930e-01 -4.22645658e-02
1.88222378e-01 8.79750997e-02 -1.43317008e+00 2.20910001e+00
-1.36177912e-01 2.26802438e-01 6.06035650e-01 -1.40710056e+00
7.82580614e-01 5.59279859e-01 1.01811802e+00 -2.07233980e-01
-1.60314009e-01 8.55416417e-01 -6.13437533e-01 -2.13311180e-01
6.81085438e-02 -5.45869827e-01 2.46423736e-01 4.89352077e-01
-2.51151949e-01 6.68870583e-02 -4.92113113e-01 2.61772364e-01
1.42307115e+00 -1.68762982e-01 4.18866873e-01 -8.31840694e-01
5.00287235e-01 -5.89844363e-04 1.49601951e-01 5.93992054e-01
-3.07758301e-01 5.46131432e-01 -8.08306225e-03 -5.90842128e-01
-1.03439677e+00 -9.26543176e-01 -9.10197198e-01 4.23476875e-01
3.11202388e-02 -1.73998013e-01 -7.60803759e-01 -2.96621203e-01
4.89718355e-02 3.29926312e-01 -4.44362462e-01 -5.13808876e-02
-7.13488102e-01 -6.60127282e-01 3.76873702e-01 1.27857000e-01
2.23665327e-01 -5.61772823e-01 -8.05304468e-01 5.84301829e-01
1.21650606e-01 -9.00885642e-01 -6.70784295e-01 1.08567514e-01
-1.58645654e+00 -8.26592028e-01 -1.06315374e+00 -6.24362350e-01
6.67620957e-01 1.36118203e-01 6.91842556e-01 -1.75829336e-01
-3.98460686e-01 6.65658653e-01 -2.01456204e-01 7.90452361e-02
-4.09883291e-01 -1.82967067e-01 3.66924375e-01 -4.44782488e-02
-5.61030991e-02 -1.00754762e+00 -7.62507439e-01 1.01235420e-01
-7.59982705e-01 -1.14140719e-01 3.79557312e-01 8.26780319e-01
9.64895070e-01 -1.58613399e-01 4.89878505e-01 -8.22439492e-01
3.13852608e-01 -6.58325970e-01 -5.51420271e-01 -6.77606091e-02
-1.06372786e+00 2.94028819e-01 1.63442478e-01 -3.18717003e-01
-2.38219857e-01 7.27493882e-01 -6.84005325e-04 -6.98763788e-01
5.23370206e-01 4.60901141e-01 6.29665732e-01 -8.91000450e-01
9.23119366e-01 4.57402468e-01 5.91710031e-01 -5.19190431e-01
3.25444221e-01 6.28632724e-01 5.61658382e-01 -3.01375657e-01
6.43647134e-01 8.47471237e-01 5.51573277e-01 -7.52637208e-01
-3.02882791e-01 -3.33604932e-01 -4.99616086e-01 -1.04894042e-01
7.77104914e-01 -6.47763431e-01 -4.57406163e-01 -1.49594381e-01
-8.34904492e-01 8.60737935e-02 -6.66116297e-01 7.43489206e-01
-1.10228908e+00 7.33172476e-01 -7.29589880e-01 -9.15717006e-01
-5.62652767e-01 -1.00835681e+00 1.03637052e+00 -5.25294662e-01
-3.47965099e-02 -9.48041677e-01 1.37196794e-01 -3.85960750e-02
5.93261003e-01 6.60034180e-01 5.08138061e-01 -2.27426231e-01
-4.51290667e-01 -3.87877464e-01 1.77612215e-01 -7.70120844e-02
-1.09537520e-01 -9.73655760e-01 -5.09041011e-01 -5.72027981e-01
8.48438501e-01 9.97554585e-02 5.69223881e-01 6.21784329e-01
9.38598633e-01 -6.50560439e-01 -6.03066266e-01 8.40080142e-01
1.53973472e+00 -4.19234663e-01 6.54310107e-01 3.59567329e-02
3.47281814e-01 5.40288448e-01 4.01083738e-01 5.11652172e-01
-5.18547855e-02 5.56757092e-01 1.12486027e-01 1.43795714e-01
-1.18225671e-01 -1.90017521e-01 2.64446974e-01 1.02636170e+00
7.17714131e-02 9.10127163e-01 -7.76350558e-01 6.52546704e-01
-1.60001671e+00 -8.25065255e-01 -1.81294307e-01 2.40620422e+00
8.69213045e-01 -2.74049342e-01 2.05492243e-01 5.10915160e-01
8.39342058e-01 -5.64504601e-02 -6.95999026e-01 -2.57757723e-01
1.98843911e-01 3.29316914e-01 1.04286098e+00 8.04661572e-01
-8.26894343e-01 8.29872563e-02 7.42582226e+00 7.80738473e-01
-8.39075565e-01 4.37609166e-01 6.06818914e-01 -1.46212578e-02
-4.62346524e-01 1.50827363e-01 -1.63262412e-01 1.79838166e-01
1.10393715e+00 -3.81935269e-01 1.14813292e+00 7.73897409e-01
-3.78244705e-02 1.74561203e-01 -9.39714134e-01 1.35581124e+00
-2.52112627e-01 -1.58974946e+00 -3.62080365e-01 5.64408600e-01
6.06309474e-01 3.50846529e-01 -1.14144191e-01 -5.44883549e-01
-3.09111416e-01 -9.56095219e-01 4.90771294e-01 3.73957455e-01
1.21708810e+00 -4.81965601e-01 1.21758558e-01 4.43957746e-01
-1.18136656e+00 7.33634681e-02 -4.24791932e-01 1.04261406e-01
3.39617193e-01 9.57876742e-01 -5.87412059e-01 4.02100891e-01
4.80676115e-01 6.14757776e-01 1.65720120e-01 8.80202651e-01
7.44097114e-01 4.27717328e-01 -8.59630823e-01 2.64277458e-01
1.62319034e-01 -3.56779218e-01 1.10197771e+00 8.74963760e-01
6.21190369e-01 4.79131192e-01 -1.08092904e-01 7.04358518e-01
2.95379490e-01 2.46362090e-01 -8.83919001e-01 9.34178531e-02
4.62614417e-01 8.10296535e-01 -6.74912095e-01 -2.34397933e-01
-1.53320506e-01 1.21527529e+00 -8.11695904e-02 8.81700814e-02
-1.46195352e-01 -5.38691401e-01 1.61568686e-01 6.96148753e-01
6.37826264e-01 -3.08482379e-01 -3.27869356e-01 -9.31213558e-01
3.78907204e-01 -6.47198498e-01 3.99264812e-01 -2.26296961e-01
-1.07531238e+00 4.96695042e-01 1.72281891e-01 -1.26590788e+00
-4.97039437e-01 -3.30986619e-01 -5.91052473e-02 7.31503427e-01
-1.17879760e+00 -4.68004733e-01 4.11582172e-01 8.22045267e-01
-5.25753610e-02 1.77841142e-01 1.06629324e+00 2.77453452e-01
2.70019054e-01 3.98640990e-01 4.24666196e-01 -2.78194249e-01
-2.12282419e-01 -9.24131036e-01 2.68142909e-01 6.41521454e-01
3.79928984e-02 5.34339249e-01 7.71408200e-01 -6.68366432e-01
-2.08934832e+00 -7.58122921e-01 7.08506525e-01 -1.01566710e-01
7.18441725e-01 -1.61888488e-02 -8.44247282e-01 5.92007995e-01
-4.49706823e-01 6.32397294e-01 3.48262310e-01 -3.09739232e-01
-3.64943705e-02 6.15559518e-02 -1.99817729e+00 -1.28204999e-02
1.43323445e+00 -6.97543085e-01 -3.99045944e-01 8.49437714e-01
6.05808735e-01 -6.40522480e-01 -1.30007362e+00 3.44468534e-01
4.28226739e-01 -6.10288441e-01 1.17987931e+00 -2.16336161e-01
5.23616932e-02 -8.24909136e-02 -6.41393661e-01 -7.06325769e-01
-3.16230625e-01 -1.34778714e+00 -5.42915165e-01 2.49559775e-01
1.06681414e-01 -8.35986257e-01 9.03355718e-01 5.12538195e-01
-1.19945826e-03 -1.15425324e+00 -1.73632848e+00 -7.73603082e-01
1.36382908e-01 -3.89437526e-01 3.02274734e-01 9.57868516e-01
6.54074728e-01 -1.12901315e-01 -3.36271942e-01 6.85329288e-02
1.29262614e+00 -2.68116537e-02 -1.29082128e-01 -1.16976154e+00
-6.81402504e-01 9.09862891e-02 -9.25932229e-01 -1.05109262e+00
-4.15613919e-01 -1.39590394e+00 -2.18623087e-01 -1.07521164e+00
4.90897149e-02 -6.83288634e-01 -2.33057648e-01 3.44013870e-02
3.37266147e-01 3.86040300e-01 5.33072501e-02 5.82259595e-01
-5.03227711e-01 3.57302368e-01 1.09785616e+00 2.01426208e-01
-5.24049215e-02 7.35262930e-02 -5.54699421e-01 5.09106338e-01
2.66125381e-01 -8.33689988e-01 -2.21195623e-01 -2.54521579e-01
3.63609850e-01 8.24963748e-01 1.86260998e-01 -1.16469562e+00
2.63322890e-01 2.53508955e-01 1.01471439e-01 -3.22895139e-01
2.96444565e-01 -7.98890769e-01 4.65715557e-01 1.02992976e+00
-3.32993627e-01 -6.43032864e-02 -1.51696563e-01 9.56511438e-01
2.54689097e-01 -4.29688126e-01 7.79950976e-01 -2.20706969e-01
2.14914302e-03 4.76477921e-01 -9.92062017e-02 -1.13035269e-01
8.00350547e-01 -4.76883799e-01 5.33972919e-01 -4.31099921e-01
-1.28844857e+00 -2.35038832e-01 4.80978936e-01 -4.60803658e-01
9.30502355e-01 -1.63144040e+00 -9.33068931e-01 2.73477972e-01
-3.34075361e-01 -2.68546969e-01 1.24955222e-01 1.40357280e+00
-6.54730022e-01 2.99112320e-01 2.01768011e-01 -7.67826021e-01
-7.85090208e-01 7.11237431e-01 5.34453809e-01 -2.20818892e-01
-1.29862404e+00 6.45031691e-01 -5.36550403e-01 -3.36205393e-01
-1.58971343e-02 2.20097918e-02 3.28938484e-01 -5.07075191e-01
7.38195240e-01 8.12962770e-01 1.89327732e-01 -5.09561300e-01
-6.20038927e-01 8.21610510e-01 2.08340093e-01 -4.55934972e-01
1.65032530e+00 -3.11262161e-01 -1.71285719e-01 2.52394736e-01
1.68959951e+00 9.03045908e-02 -1.13901043e+00 -4.05705959e-01
1.42177530e-02 -3.95157665e-01 4.62589145e-01 -1.39862776e-01
-8.57331216e-01 4.44608539e-01 9.53384697e-01 3.10412169e-01
1.09935904e+00 2.40594447e-01 9.25801754e-01 3.99955481e-01
9.33446050e-01 -7.43740320e-01 -6.21398866e-01 -2.00447768e-01
1.05607843e+00 -7.50468016e-01 4.42355365e-01 -6.10389888e-01
1.61294434e-02 1.12847722e+00 -8.51969779e-01 -5.78775227e-01
1.24136972e+00 2.77676255e-01 -5.42553484e-01 -4.46992874e-01
-1.77365780e-01 6.01427197e-01 1.88864753e-01 4.25229222e-01
6.56049103e-02 2.53002912e-01 -6.77783966e-01 6.26404025e-03
-1.68375015e-01 3.49802822e-01 3.63312244e-01 9.16188002e-01
-5.22046208e-01 -8.25033128e-01 -5.17901599e-01 7.52859175e-01
-5.49233735e-01 -7.28386566e-02 1.73405290e-01 2.40325928e-01
-5.51835835e-01 7.10617959e-01 -2.06359848e-01 -1.73815176e-01
-7.48170540e-02 -9.16561950e-03 8.51909876e-01 -1.61235899e-01
2.53455844e-02 -6.33093715e-02 -8.81160274e-02 -1.07896972e+00
-4.85956997e-01 -1.22104859e+00 -1.50158465e+00 -3.95200759e-01
-3.21634114e-01 1.74046308e-01 8.76894772e-01 8.71163666e-01
6.37833893e-01 -2.45954603e-01 8.35374475e-01 -9.43239570e-01
-1.31733155e+00 -6.51482463e-01 -8.00237238e-01 2.65359700e-01
5.38681388e-01 -3.84673059e-01 -5.12139320e-01 -2.50865459e-01] | [13.41998291015625, -2.4340717792510986] |
94b3a62e-0136-4b3c-9837-10226d6992b4 | inheriting-the-wisdom-of-predecessors-a | null | null | https://www.ijcai.org/proceedings/2022/572 | https://www.ijcai.org/proceedings/2022/0572.pdf | Inheriting the Wisdom of Predecessors: A Multiplex Cascade Framework for Unified Aspect-based Sentiment Analysis | So far, aspect-based sentiment analysis (ABSA) has involved with total seven subtasks, in which, however the interactions among them have been left unexplored sufficiently. This work presents a novel multiplex cascade framework for unified ABSA and maintaining such interactions. First, we model total seven subtasks as a hierarchical dependency in the easy-to-hard order, based on which we then propose a multiplex decoding mechanism, transferring the sentiment layouts and clues in lower tasks to upper ones. The multiplex strategy enables highly-efficient subtask interflows and avoids repetitive training; meanwhile it sufficiently utilizes the existing data without requiring any further annotation. Further, based on the characteristics of aspect-opinion term extraction and pairing, we enhance our multiplex framework by integrating POS tag and syntactic dependency information for term boundary and pairing identification. The proposed Syntax-aware Multiplex (SyMux) framework enhances the ABSA performances on 28 subtasks (7×4 datasets) with big margins. | ['Donghong Ji', 'Jingye Li', 'Shengqiong Wu', 'Chenliang Li', 'Fei Li', 'Hao Fei'] | 2022-07-30 | null | null | null | conference-2022-7 | ['term-extraction', 'aspect-based-sentiment-analysis', 'aspect-sentiment-triplet-extraction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 2.35792071e-01 -1.33326724e-01 1.13617823e-01 -6.42454624e-01
-8.19412947e-01 -7.70437062e-01 4.73660290e-01 2.13708386e-01
-4.47164923e-01 2.97851771e-01 1.69362620e-01 -5.94067454e-01
-1.80835888e-01 -6.08299792e-01 -6.97709918e-01 -6.18223310e-01
2.99855649e-01 2.83557847e-02 3.42888296e-01 -5.96155345e-01
1.70172080e-01 -2.46033028e-01 -1.38844991e+00 9.16046619e-01
1.05539083e+00 1.30064201e+00 1.76226810e-01 4.29287553e-01
-5.10534346e-01 5.87331355e-01 -5.21139383e-01 -9.31566417e-01
-5.97570762e-02 -2.03332752e-01 -9.26323891e-01 -1.01803504e-01
-3.23658660e-02 5.21990210e-02 3.15565407e-01 7.61114419e-01
4.40314829e-01 -7.62853250e-02 4.64712769e-01 -1.07636356e+00
-5.34369290e-01 9.72372174e-01 -9.18920279e-01 4.36151354e-03
1.95765823e-01 -1.87180609e-01 1.60099316e+00 -9.79765594e-01
1.95839822e-01 1.10832572e+00 6.19297206e-01 7.13122264e-02
-6.93621695e-01 -7.25012839e-01 1.01552653e+00 2.22112253e-01
-1.01705575e+00 -3.58875960e-01 8.33670437e-01 -2.68154055e-01
1.43385375e+00 2.78770894e-01 7.98919857e-01 8.39185715e-01
8.15980416e-03 1.18397617e+00 9.89614785e-01 -4.63430196e-01
-9.27166119e-02 -2.24837177e-02 6.15770698e-01 7.31013358e-01
2.06001759e-01 -6.04795933e-01 -7.42187321e-01 1.29662767e-01
4.09763772e-03 -4.86899167e-03 -4.94683953e-03 3.62136215e-02
-1.07359004e+00 4.36488628e-01 2.49165103e-01 3.56960982e-01
-2.14063004e-01 -1.85327455e-01 6.70537710e-01 4.84588623e-01
4.47308868e-01 1.54133350e-01 -1.05170488e+00 3.83947156e-02
-6.26556337e-01 -7.92886764e-02 7.01669991e-01 1.25211918e+00
9.55907583e-01 -2.28445187e-01 -3.29254061e-01 8.03866506e-01
4.10921663e-01 2.91806102e-01 5.26512265e-01 2.20919307e-02
8.60736251e-01 1.26206553e+00 -1.76786810e-01 -9.48801935e-01
-6.63222432e-01 -6.59774840e-01 -6.11906528e-01 -3.76367331e-01
1.35716155e-01 -2.55134284e-01 -8.35658789e-01 1.57191503e+00
3.91712397e-01 -1.74720377e-01 1.31892664e-02 6.82405472e-01
8.23198676e-01 4.73840654e-01 2.11364746e-01 -1.71280891e-01
1.98167658e+00 -1.30442691e+00 -7.61390746e-01 -4.42180455e-01
9.71072912e-01 -1.20268142e+00 1.29573166e+00 4.95029658e-01
-8.12467396e-01 -4.23864037e-01 -1.24817455e+00 -4.72632229e-01
-6.30848348e-01 5.74837685e-01 1.08745646e+00 6.43067837e-01
-9.52603400e-01 -4.45075892e-03 -7.49884248e-01 -2.79217493e-03
2.58720726e-01 6.18078053e-01 -7.56710917e-02 1.16027363e-01
-1.42536569e+00 4.08346713e-01 3.46261352e-01 8.05699766e-01
-4.20795858e-01 -4.78908598e-01 -9.52203989e-01 1.27218395e-01
6.30436063e-01 -7.81771600e-01 1.26288021e+00 -8.97493660e-01
-1.61890459e+00 6.36302233e-01 -4.45635676e-01 -1.03665497e-02
8.66095945e-02 -4.50782001e-01 -4.01167691e-01 -3.69055182e-01
1.07605815e-01 3.40492874e-01 7.56262541e-01 -1.05605769e+00
-9.72099721e-01 -7.22080171e-01 6.36311471e-01 4.95394111e-01
-8.43636394e-01 3.16647738e-01 -7.27264225e-01 -7.03630328e-01
5.64928316e-02 -7.29421079e-01 -1.29904658e-01 -5.87780833e-01
-5.74443579e-01 -3.53099614e-01 3.34136069e-01 -4.97712284e-01
1.66555822e+00 -2.00432825e+00 2.23743886e-01 2.46190980e-01
2.62207687e-01 2.55150385e-02 -1.11395858e-01 3.05024713e-01
4.15155739e-02 5.72729409e-02 -1.08964518e-01 -6.35430753e-01
2.02585399e-01 -3.57613303e-02 -2.55789816e-01 -1.39724031e-01
2.54489839e-01 9.72602487e-01 -5.88073075e-01 -6.52994573e-01
-1.61535531e-01 1.58991814e-01 -6.27061188e-01 1.44120604e-01
-3.06986779e-01 1.05191283e-01 -8.51604462e-01 8.95344734e-01
8.96259785e-01 -2.55970299e-01 2.69532919e-01 -4.40673500e-01
-2.55365133e-01 4.33281302e-01 -9.75614250e-01 1.77207768e+00
-9.26231802e-01 1.50500564e-02 2.79356509e-01 -8.48202527e-01
9.15058851e-01 5.99071197e-02 2.54848063e-01 -5.74635148e-01
3.21047127e-01 2.69873977e-01 -2.49224864e-02 -5.03304720e-01
5.27475655e-01 4.14690450e-02 -4.95535672e-01 4.34003860e-01
1.74909115e-01 1.62170202e-01 2.99848855e-01 8.56313556e-02
9.46439922e-01 2.78401643e-01 5.20729274e-02 -2.56111056e-01
9.08726454e-01 -2.28607585e-03 5.96111953e-01 4.83882248e-01
-2.61365231e-02 3.17865759e-01 7.32073665e-01 -2.81834513e-01
-4.43579644e-01 -4.68520492e-01 7.87158161e-02 1.71514142e+00
2.07127616e-01 -7.37913489e-01 -7.53835142e-01 -9.25012887e-01
-3.12744617e-01 2.86658883e-01 -7.19320774e-01 5.37902750e-02
-5.51825643e-01 -1.19126022e+00 4.65917587e-01 6.30869687e-01
5.79747260e-01 -9.66747105e-01 4.25930619e-02 1.99056417e-01
-2.18106747e-01 -1.30418193e+00 -6.06482625e-01 5.82160056e-01
-5.78845859e-01 -9.29460883e-01 -4.28155869e-01 -1.11862397e+00
5.24732471e-01 3.06253940e-01 1.10371923e+00 -3.50991040e-02
7.80394450e-02 -2.24342540e-01 -7.03165233e-01 -4.89821404e-01
2.60404974e-01 6.71279073e-01 -3.28835189e-01 3.18218291e-01
8.06492865e-01 -6.12538815e-01 -7.97272444e-01 2.98206300e-01
-8.46422970e-01 2.60389507e-01 9.73899126e-01 6.22731388e-01
4.85694140e-01 -1.58850357e-01 6.91073835e-01 -1.10456729e+00
7.06891477e-01 -3.38743657e-01 -5.61076403e-01 4.99253184e-01
-5.58895767e-01 -1.10496812e-01 8.94775629e-01 -7.37330467e-02
-1.42216635e+00 5.11736190e-03 -2.95898646e-01 1.54749870e-01
4.53199111e-02 7.15853512e-01 -5.68269610e-01 8.34538564e-02
2.01326221e-01 2.36682951e-01 -3.48790437e-01 -5.19437194e-01
4.09357011e-01 8.51251841e-01 7.29124993e-04 -6.48382187e-01
4.31625843e-01 3.04989934e-01 -2.02368677e-01 -4.07996714e-01
-1.29212129e+00 -7.00707912e-01 -6.48794472e-01 5.76863326e-02
9.38363254e-01 -1.07038128e+00 -1.03747761e+00 8.57666492e-01
-1.25357747e+00 1.39860108e-01 3.50216836e-01 3.13000113e-01
9.94268525e-03 3.87191623e-01 -8.11110735e-01 -6.64222538e-01
-6.24563515e-01 -1.49557912e+00 1.30160034e+00 1.66106075e-01
-9.65784043e-02 -6.41709864e-01 -2.53299385e-01 7.55331159e-01
2.62696058e-01 -3.75731587e-01 1.23255622e+00 -8.08625996e-01
-4.24822479e-01 1.76086519e-02 -3.05644840e-01 2.78503656e-01
1.35215387e-01 -1.91591054e-01 -1.12873113e+00 -1.32290022e-02
-9.34728608e-03 -3.64041835e-01 1.08089960e+00 1.75912648e-01
1.09503460e+00 4.57174443e-02 -1.82834253e-01 5.09768724e-01
9.89437461e-01 1.94109321e-01 1.42925873e-01 6.26052856e-01
1.00827277e+00 7.47808039e-01 6.22528017e-01 3.04866433e-01
8.69040906e-01 4.40778375e-01 3.23467523e-01 -1.94899887e-01
1.61495373e-01 -5.36186174e-02 5.29376984e-01 1.61753166e+00
7.29782926e-03 -3.47168505e-01 -6.29239440e-01 6.01663828e-01
-1.88047099e+00 -1.59279302e-01 -2.57879376e-01 1.67399168e+00
9.85549510e-01 2.94106930e-01 -9.03451741e-02 1.58492371e-01
5.00047624e-01 4.14997280e-01 -4.35506195e-01 -5.67219436e-01
-1.48475736e-01 5.54245040e-02 9.27024633e-02 4.29106146e-01
-1.31457114e+00 1.04999769e+00 4.77624893e+00 1.18608809e+00
-7.77000844e-01 1.50849804e-01 6.50660455e-01 1.88194904e-02
-7.05311179e-01 1.05297901e-01 -1.05449438e+00 2.80332446e-01
3.97902846e-01 2.13138640e-01 2.32022643e-01 7.07465768e-01
-7.11144581e-02 3.82348001e-02 -8.54753375e-01 6.21260822e-01
3.76625210e-02 -8.21164012e-01 3.90905440e-01 -4.22878325e-01
6.03256226e-01 -3.14487934e-01 1.52914628e-01 7.20593333e-01
9.84555632e-02 -6.27331793e-01 5.92505276e-01 1.30855665e-01
4.19222325e-01 -9.06324506e-01 1.14137089e+00 6.43004179e-02
-1.57635188e+00 -1.42820582e-01 -1.42279372e-01 2.68149078e-02
3.92534137e-02 7.19131887e-01 -3.12944621e-01 1.14293981e+00
6.83556497e-01 6.15163267e-01 -7.06169248e-01 6.29526794e-01
-3.77267212e-01 6.32038414e-01 -2.07243070e-01 -3.82721782e-01
5.14287710e-01 -3.52039278e-01 2.07708180e-01 1.46776986e+00
1.15092978e-01 -7.81542882e-02 1.35531947e-01 2.76340812e-01
-9.81398001e-02 7.48672783e-01 -7.81922713e-02 5.80371066e-04
1.53377593e-01 1.58901882e+00 -8.38696301e-01 -9.27882865e-02
-8.17670822e-01 8.27784896e-01 3.53438824e-01 3.64523947e-01
-7.39825070e-01 -7.43896246e-01 4.82994974e-01 -3.45431745e-01
5.58969498e-01 9.82196331e-02 -5.90403736e-01 -1.46893466e+00
5.79836905e-01 -9.45134103e-01 5.69591999e-01 -6.04841113e-01
-1.11907756e+00 1.01839197e+00 -2.16254368e-01 -1.07837868e+00
1.78111404e-01 -8.87366295e-01 -5.47620893e-01 8.03167105e-01
-1.74809730e+00 -1.91862822e+00 -9.56068411e-02 5.09656727e-01
6.53587759e-01 -1.30520239e-01 6.63902998e-01 5.78641891e-01
-9.71615136e-01 9.61089790e-01 -2.41256595e-01 4.32905257e-02
7.51969039e-01 -1.31116796e+00 4.45604414e-01 8.40808213e-01
-1.23012282e-01 9.31997001e-01 1.64449766e-01 -5.16651988e-01
-1.49796879e+00 -9.33924437e-01 1.20820618e+00 -2.67761707e-01
1.02597904e+00 -7.24889696e-01 -7.27224946e-01 6.26390040e-01
3.17710012e-01 -4.21842009e-01 7.09519267e-01 7.88945258e-01
-4.89486605e-01 -4.53634709e-01 -3.57147425e-01 6.05798006e-01
1.07146525e+00 -6.51221573e-01 -5.28491318e-01 2.03007892e-01
1.06321812e+00 -2.80551881e-01 -4.82877880e-01 5.67790687e-01
5.84811032e-01 -7.83078492e-01 7.02511489e-01 -4.52522546e-01
5.38667321e-01 -4.86249447e-01 -1.28513742e-02 -1.05640817e+00
-1.64260060e-01 -6.13838673e-01 7.75293708e-02 1.58414340e+00
9.55636859e-01 -4.96474206e-01 5.36348820e-01 3.09358478e-01
-4.85728949e-01 -1.25508857e+00 -6.35038793e-01 -1.78464845e-01
-1.83707327e-01 -7.16825068e-01 9.52514887e-01 7.44681418e-01
9.26676691e-02 8.59856546e-01 -1.32555306e-01 1.21892579e-01
7.10078925e-02 6.08134270e-01 3.69900525e-01 -8.64613771e-01
-4.30421412e-01 -5.59447706e-01 -7.89063722e-02 -1.37608206e+00
8.83794948e-02 -1.02475786e+00 -8.54387954e-02 -1.21930695e+00
3.26771230e-01 -5.42695582e-01 -7.72709131e-01 6.93485379e-01
-6.03164375e-01 1.58932835e-01 -8.07484314e-02 -4.50509880e-03
-1.15557706e+00 7.27150917e-01 1.21782529e+00 -1.79952830e-01
-1.20884135e-01 9.07652900e-02 -1.10021913e+00 8.29336464e-01
5.17667174e-01 -2.47021079e-01 -5.44091105e-01 -8.77805591e-01
8.76893818e-01 -3.09085280e-01 -2.27313563e-01 -4.12126750e-01
3.03939551e-01 2.20925167e-01 2.52171934e-01 -7.68463433e-01
1.14683807e-01 -6.97191894e-01 -5.46514511e-01 -9.88997817e-02
-2.45997384e-01 2.68247366e-01 3.91014993e-01 6.27278388e-01
-5.17720401e-01 -2.28486165e-01 6.62564253e-03 8.10947046e-02
-6.05584264e-01 1.22832134e-01 -2.74071842e-01 9.60454866e-02
8.24628949e-01 6.56229183e-02 -2.36028194e-01 4.32277061e-02
-6.94315255e-01 6.06982946e-01 -2.49166731e-02 4.40815687e-01
2.69817501e-01 -1.07232940e+00 -4.27708477e-01 2.20387772e-01
2.99092054e-01 2.91512936e-01 6.55469716e-01 9.98016596e-01
-2.29182780e-01 4.01200593e-01 1.84156716e-01 -3.90198231e-01
-1.29331660e+00 4.86780554e-01 8.84067714e-02 -6.93639219e-01
-9.09793302e-02 1.22291875e+00 5.30727208e-01 -7.00124025e-01
7.01715946e-02 -5.22208273e-01 -7.47379959e-01 5.83856165e-01
4.04612392e-01 3.46601121e-02 6.27143204e-01 -4.83398199e-01
-4.99312729e-01 8.38038504e-01 -6.08616769e-01 8.82259160e-02
1.24326599e+00 -3.60239595e-01 -5.23523510e-01 2.72791326e-01
1.07165337e+00 1.52346894e-01 -6.63828135e-01 -1.95335001e-01
1.15106240e-01 -1.52907474e-02 5.65665923e-02 -8.89655173e-01
-1.07378566e+00 1.09420896e+00 -3.39369848e-02 2.73946732e-01
1.40483201e+00 -1.79675102e-01 1.06381392e+00 5.14138937e-01
2.16408521e-01 -1.03132617e+00 -1.95370600e-01 7.43059695e-01
5.82712948e-01 -1.11921453e+00 -1.41982287e-01 -7.88743854e-01
-7.82506883e-01 1.00526786e+00 7.88701296e-01 2.83669263e-01
7.24561334e-01 4.68532294e-01 2.19589174e-01 -2.03807101e-01
-9.32488084e-01 -3.51532340e-01 3.32426488e-01 1.07524149e-01
5.71408987e-01 -6.04927316e-02 -6.72769547e-01 1.61798000e+00
-2.63921171e-01 -1.91105068e-01 -5.82664795e-02 9.12289739e-01
-8.08099508e-02 -1.20293057e+00 3.98072153e-02 4.12629545e-01
-7.78829873e-01 -6.17368817e-01 -2.27293178e-01 4.71082330e-01
3.42722505e-01 1.15159142e+00 -3.45869392e-01 -6.56363666e-01
6.07040048e-01 1.19804285e-01 6.25318512e-02 -4.19786960e-01
-1.18260336e+00 4.35081124e-01 4.41465616e-01 -4.01194096e-01
-3.95595074e-01 -4.58892703e-01 -1.08943152e+00 1.13648534e-01
-5.74741602e-01 2.60509461e-01 4.93551940e-01 1.29180515e+00
6.43295586e-01 8.11626434e-01 7.16939688e-01 -4.64969814e-01
-3.53447467e-01 -1.10368979e+00 -3.68790209e-01 1.42412320e-01
2.58224249e-01 -4.58835632e-01 -2.02471539e-01 -7.54905641e-02] | [11.523127555847168, 6.594694137573242] |
112ecbd2-ea85-4aa2-b6e6-29dca6186824 | a-survey-of-machine-learning-techniques-in | 2010.09680 | null | https://arxiv.org/abs/2010.09680v1 | https://arxiv.org/pdf/2010.09680v1.pdf | A Survey of Machine Learning Techniques in Adversarial Image Forensics | Image forensic plays a crucial role in both criminal investigations (e.g., dissemination of fake images to spread racial hate or false narratives about specific ethnicity groups) and civil litigation (e.g., defamation). Increasingly, machine learning approaches are also utilized in image forensics. However, there are also a number of limitations and vulnerabilities associated with machine learning-based approaches, for example how to detect adversarial (image) examples, with real-world consequences (e.g., inadmissible evidence, or wrongful conviction). Therefore, with a focus on image forensics, this paper surveys techniques that can be used to enhance the robustness of machine learning-based binary manipulation detectors in various adversarial scenarios. | ['Kim-Kwang Raymond Choo', 'Reza M. Parizi', 'Ali Dehghantanha', 'Ehsan Nowroozi'] | 2020-10-19 | null | null | null | null | ['image-forensics'] | ['computer-vision'] | [ 5.85951924e-01 -2.02972978e-01 -1.84648469e-01 -1.67938117e-02
-5.14554203e-01 -8.40133548e-01 7.51360953e-01 4.40931678e-01
-3.96945357e-01 1.01231110e+00 -2.74906188e-01 -6.89884722e-01
1.91833615e-01 -8.95507038e-01 -6.11753941e-01 -7.81376481e-01
7.48230964e-02 -8.30795057e-03 -1.25450626e-01 1.09921746e-01
9.63256240e-01 8.21384907e-01 -1.00716007e+00 4.04810578e-01
5.69916785e-01 5.96018195e-01 -5.05540967e-01 7.19341874e-01
9.69464798e-03 9.86774921e-01 -1.15472281e+00 -1.18885005e+00
2.52944261e-01 -2.95869708e-01 -5.86840510e-01 1.76041454e-01
5.12829125e-01 -7.67749250e-01 -6.32394731e-01 1.66775978e+00
3.81513864e-01 -3.51867437e-01 6.50561988e-01 -1.41085052e+00
-8.44237566e-01 2.69881368e-01 -9.10209298e-01 6.74396574e-01
4.18462604e-01 5.46115041e-01 -2.79090852e-02 -7.36074388e-01
5.91072738e-01 1.52873778e+00 4.61079359e-01 3.84436518e-01
-9.08230364e-01 -1.19430220e+00 -5.88286817e-01 4.90037858e-01
-1.16649270e+00 -5.80089331e-01 7.45450675e-01 -6.28960788e-01
3.37309808e-01 2.75694966e-01 4.06487942e-01 1.32840562e+00
4.70203012e-01 6.04694426e-01 1.61706102e+00 -5.37645698e-01
7.68876150e-02 3.14767867e-01 -1.50750250e-01 5.79527438e-01
9.20147479e-01 3.29870522e-01 -2.50366598e-01 -7.88699865e-01
7.26636708e-01 -1.56096280e-01 -3.22049633e-02 2.26944715e-01
-6.69277489e-01 1.34909523e+00 1.55843392e-01 5.25761135e-02
-3.98275793e-01 -1.85877737e-02 6.02678001e-01 6.42275512e-02
3.43490809e-01 5.47123551e-01 3.48299176e-01 -1.30085483e-01
-9.65047777e-01 2.73951560e-01 3.15180361e-01 3.14649880e-01
5.02926469e-01 2.93647081e-01 1.65805608e-01 4.03202891e-01
7.64813125e-02 6.57862723e-01 1.94717824e-01 -6.30873621e-01
6.82257473e-01 1.53641373e-01 4.57571112e-02 -1.57466090e+00
3.29860181e-01 1.56574756e-01 -8.13814938e-01 5.42692244e-01
4.93129045e-01 -2.17256278e-01 -9.20870125e-01 1.17136359e+00
4.13000226e-01 2.27833673e-01 -4.96622659e-02 8.09744477e-01
5.22465885e-01 4.07743007e-01 4.46431875e-01 -1.05735108e-01
1.41442561e+00 -3.95589709e-01 -7.90227532e-01 -4.66037840e-01
2.08148122e-01 -8.60617518e-01 6.03865504e-01 5.08110821e-01
-9.82953668e-01 -3.50458622e-02 -7.65941322e-01 4.11223710e-01
-5.97407103e-01 -4.78304058e-01 5.81381977e-01 1.49921596e+00
-3.22033733e-01 4.75603491e-01 -3.50243062e-01 -2.54017562e-01
1.11729527e+00 8.09631050e-02 -5.06964982e-01 -3.79319847e-01
-1.22180212e+00 9.50133502e-01 5.48682094e-01 -5.63827716e-02
-8.29616368e-01 -3.94214213e-01 -6.22810304e-01 -2.72097915e-01
3.57735157e-01 -3.48106027e-02 5.89258373e-01 -9.11758125e-01
-5.97016573e-01 1.38485599e+00 1.60156891e-01 -7.77316093e-01
8.96023810e-01 -3.81965935e-02 -7.09189534e-01 5.64951301e-01
1.68894693e-01 3.96009266e-01 1.40470958e+00 -1.25072789e+00
-2.48581007e-01 -5.14263153e-01 1.58110499e-01 -3.05652291e-01
-6.18562698e-01 7.38162398e-01 3.05715173e-01 -1.02679586e+00
-3.38168859e-01 -7.51564980e-01 -1.21909238e-01 3.10499407e-02
-8.12671661e-01 2.76426584e-01 1.25833464e+00 -1.10247445e+00
1.20827794e+00 -2.16733122e+00 -5.89474142e-01 1.84679717e-01
1.42353982e-01 9.61484313e-01 1.54912174e-01 3.16002399e-01
-3.69422175e-02 7.48655140e-01 -3.96071702e-01 2.47093305e-01
-4.95308727e-01 4.68247123e-02 -5.40700555e-01 8.65679264e-01
4.23619837e-01 7.78880119e-01 -9.10081744e-01 -6.80821538e-01
3.09940755e-01 2.11391032e-01 -5.36214784e-02 -2.01954246e-01
3.83109510e-01 6.88123286e-01 -5.61157107e-01 8.89681935e-01
9.55267489e-01 2.25169733e-01 7.06614107e-02 -6.62191957e-02
1.28496170e-01 -7.09930435e-02 -7.58201122e-01 4.97250229e-01
-2.51288060e-02 8.73006463e-01 8.04196373e-02 -1.05162311e+00
7.12226152e-01 2.18220785e-01 -1.06205896e-01 -4.86846983e-01
3.52579504e-01 1.97546944e-01 7.07717463e-02 -9.59509552e-01
7.61684060e-01 -3.22742581e-01 7.38506094e-02 8.74287307e-01
-4.31018502e-01 2.20967442e-01 -1.45495106e-02 6.51133284e-02
1.22071683e+00 -4.12601739e-01 5.57796597e-01 2.39003852e-01
5.27379096e-01 3.11018646e-01 2.72225142e-01 8.72182429e-01
-6.18488312e-01 3.88751388e-01 3.89429569e-01 -5.64274848e-01
-1.45101893e+00 -7.90830195e-01 -1.50981853e-02 4.26553369e-01
2.71550417e-01 3.36794883e-01 -7.69747972e-01 -9.76843536e-01
2.26085261e-01 7.54591703e-01 -3.53323787e-01 -5.25670826e-01
-7.42682278e-01 -7.64846623e-01 1.25097013e+00 1.80828065e-01
6.74755037e-01 -1.21840656e+00 -7.87146747e-01 7.81186447e-02
-1.09878041e-01 -1.30157399e+00 -9.34461206e-02 -5.73667049e-01
-8.23307514e-01 -1.42165637e+00 -5.13177037e-01 -1.19916901e-01
8.54173660e-01 5.01055658e-01 5.27961671e-01 4.54966635e-01
-7.97738552e-01 2.05995694e-01 -3.80352706e-01 -4.80645180e-01
-1.02856982e+00 -5.94442010e-01 -1.42607272e-01 1.34087905e-01
3.76648158e-01 -9.57028121e-02 -4.93866771e-01 6.55908361e-02
-1.46439195e+00 -5.05285442e-01 5.13016224e-01 6.08458400e-01
-6.95603788e-02 5.29795408e-01 4.90838438e-01 -1.17173696e+00
1.03417432e+00 -8.18871498e-01 -4.27755207e-01 3.58995020e-01
-3.03702712e-01 -5.01919329e-01 5.23750246e-01 -1.09063697e+00
-1.14285195e+00 -5.64954758e-01 1.07239433e-01 -6.07636988e-01
-5.04900396e-01 4.18525040e-01 1.08353406e-01 -5.14035404e-01
7.34529376e-01 1.99878007e-01 8.69515464e-02 -1.77816436e-01
1.72831997e-01 9.05678153e-01 5.76652884e-01 -3.03137809e-01
1.22722626e+00 7.07446754e-01 1.43146589e-01 -8.46026182e-01
-6.12293839e-01 -1.36976138e-01 -2.68310875e-01 -6.04633033e-01
6.27561092e-01 -5.81325114e-01 -4.70241070e-01 8.44049573e-01
-1.31935787e+00 3.22151273e-01 3.70406568e-01 3.20756048e-01
-9.87224802e-02 9.39241767e-01 -6.04351282e-01 -1.23106229e+00
-1.52077392e-01 -9.73091960e-01 5.27147293e-01 2.78962374e-01
-3.07139158e-01 -8.62110019e-01 -2.77460009e-01 1.11593688e+00
2.74899393e-01 9.20965672e-01 8.80586505e-01 -5.83622038e-01
-3.69689286e-01 -7.11311519e-01 -4.45111245e-01 4.64309543e-01
1.32353544e-01 1.34725586e-01 -9.17919278e-01 -3.60032111e-01
6.37509525e-02 -5.89553773e-01 4.41403061e-01 -1.07619623e-02
9.05188382e-01 -8.73188317e-01 -3.30818176e-01 1.82133421e-01
1.36831224e+00 6.05332553e-01 1.26932764e+00 5.92127562e-01
4.45888847e-01 8.26298952e-01 8.91159654e-01 5.27794182e-01
-3.25353295e-01 3.47906411e-01 5.38073123e-01 8.22791830e-02
-1.52337719e-02 -1.73523262e-01 3.15028936e-01 -2.54650474e-01
-1.74739391e-01 -3.11492085e-01 -8.81184518e-01 3.44506651e-01
-1.41303670e+00 -1.47608066e+00 -3.23779583e-01 2.16626525e+00
4.24058944e-01 2.20366478e-01 1.29458889e-01 3.03227782e-01
1.50517380e+00 3.02820206e-01 -6.89889669e-01 -7.06845522e-01
-2.59212017e-01 3.21979612e-01 7.25550711e-01 9.52256620e-02
-1.34933901e+00 1.09607697e+00 6.19110727e+00 1.09903276e+00
-1.14887059e+00 4.19474334e-01 7.65323937e-01 2.50519156e-01
5.65709695e-02 -1.70519892e-02 -2.78201967e-01 7.11143851e-01
4.98153478e-01 -1.07173271e-01 3.85205626e-01 6.60258472e-01
2.51512408e-01 -2.97488183e-01 -1.73300728e-01 9.86231208e-01
5.63876212e-01 -1.39902651e+00 2.15599574e-02 3.93557936e-01
5.46393394e-01 -8.15225363e-01 2.67046213e-01 -1.04483820e-01
1.12111621e-01 -1.09438229e+00 9.54874992e-01 -1.30229384e-01
9.86110926e-01 -8.52018893e-01 6.66564465e-01 2.99555480e-01
-4.57936019e-01 -1.15884915e-01 -4.29410756e-01 -1.74687535e-01
3.78987640e-01 6.87850058e-01 -7.43113995e-01 1.77586168e-01
5.45547783e-01 2.26386830e-01 -5.28808355e-01 9.13736403e-01
-6.73404038e-01 7.38898456e-01 7.67602101e-02 8.29559043e-02
3.84735227e-01 -2.43500948e-01 8.68770421e-01 1.04079068e+00
2.19570428e-01 1.55044585e-01 -3.18580180e-01 8.40016603e-01
3.32682277e-03 -2.06700489e-01 -1.13402641e+00 -6.26181483e-01
6.42511427e-01 1.20366311e+00 -1.00062430e+00 -2.89126039e-01
-1.41974628e-01 9.71789002e-01 -7.79762864e-02 2.63899148e-01
-1.00194156e+00 -2.61685729e-01 5.14701605e-01 4.95820493e-01
1.08481303e-01 -1.69436932e-01 -1.38029009e-01 -9.97953117e-01
-1.98076636e-01 -1.20989537e+00 3.27940434e-01 -6.10823810e-01
-1.29236042e+00 2.74960548e-01 -4.33383621e-02 -1.09841204e+00
-6.19402193e-02 -4.93356854e-01 -9.07208383e-01 5.13224959e-01
-1.17688215e+00 -1.23336375e+00 2.20957682e-01 4.33055788e-01
2.79645085e-01 -5.18695772e-01 3.01721007e-01 3.49677056e-01
-4.09989059e-01 4.36630398e-01 -3.18605870e-01 6.44095004e-01
5.40933073e-01 -4.87731546e-01 4.17094022e-01 1.17142105e+00
-1.40270680e-01 5.21308362e-01 5.86694062e-01 -1.13627017e+00
-1.27703440e+00 -1.01649833e+00 5.58492422e-01 -1.14636935e-01
8.58240664e-01 -8.90395343e-02 -7.71472156e-01 5.07289350e-01
-1.23559892e-01 -1.04477368e-01 6.18239462e-01 -4.62392658e-01
-6.03183031e-01 6.09910131e-01 -1.85127091e+00 7.49949992e-01
6.12574756e-01 -6.89579964e-01 -3.72111499e-01 4.03804541e-01
6.35599941e-02 -1.10378213e-01 -3.92683208e-01 1.66528393e-02
6.48933709e-01 -9.39460576e-01 1.29015696e+00 -1.01825750e+00
6.79169774e-01 -7.44455829e-02 1.92760289e-01 -6.26549244e-01
-5.99899478e-02 -4.97050852e-01 -1.05811253e-01 1.34853292e+00
-6.85775802e-02 -8.21702898e-01 6.56945884e-01 4.39555943e-01
4.70299751e-01 -3.01515877e-01 -9.92764056e-01 -8.48629594e-01
3.49758752e-02 -2.81593591e-01 4.36179757e-01 1.53514767e+00
-3.53021473e-01 -2.45799139e-01 -8.38491023e-01 3.29957902e-01
9.69981790e-01 -3.44124675e-01 7.91329563e-01 -6.65003717e-01
6.97443858e-02 -2.82976389e-01 -8.43119502e-01 -7.68744126e-02
1.33074656e-01 -3.62996012e-01 -4.87447411e-01 -6.52334452e-01
4.60855931e-01 -3.73800129e-01 1.35276303e-01 1.98519170e-01
-4.64870125e-01 6.58421516e-01 5.01498699e-01 4.15016264e-01
-1.20513864e-01 -1.37100771e-01 1.17644823e+00 -3.73029798e-01
4.36006427e-01 1.46221332e-02 -7.42278576e-01 9.02147770e-01
1.13262606e+00 -1.02119911e+00 6.58637434e-02 -9.74180549e-02
2.43616492e-01 9.97797847e-02 8.24183464e-01 -6.81937993e-01
1.25943054e-03 -5.80100477e-01 3.24949145e-01 -2.42022097e-01
1.17365122e-01 -5.14257073e-01 1.24746285e-01 9.35349584e-01
7.81941637e-02 1.55844539e-01 1.09452188e-01 7.12606013e-01
-2.88239866e-01 -7.49368608e-01 1.11687779e+00 -3.70684236e-01
-4.26667660e-01 6.42350242e-02 -6.56640053e-01 4.20037247e-02
1.35091865e+00 -6.53182745e-01 -8.50128651e-01 -5.16080856e-01
-3.26715887e-01 -2.89897084e-01 6.05369091e-01 1.66047320e-01
8.76982272e-01 -1.12015593e+00 -8.30087781e-01 -1.86531156e-01
-1.10998169e-01 -7.24236906e-01 2.00502604e-01 6.15171790e-01
-6.96742833e-01 -5.45462370e-02 -4.45689291e-01 2.13809684e-03
-1.74597764e+00 7.96178162e-01 -1.42613158e-01 -2.30806917e-01
-3.72914940e-01 4.97569919e-01 -8.27736408e-02 8.95842239e-02
-1.76195532e-01 7.90878713e-01 -1.06890097e-01 -1.11276992e-01
7.71478534e-01 8.77651036e-01 -2.86641598e-01 -1.00365376e+00
-3.83144200e-01 2.84826428e-01 -1.15769707e-01 -5.10433363e-03
9.39615905e-01 2.59451717e-02 -3.26196462e-01 -1.13036685e-01
8.77215207e-01 1.92057267e-01 -8.54661345e-01 5.11964336e-02
1.57459244e-01 -1.20409417e+00 -1.37098655e-01 -7.69685566e-01
-9.88872468e-01 9.54697132e-01 4.83600080e-01 5.09698510e-01
8.31642807e-01 -1.16018035e-01 8.63116503e-01 4.40264866e-02
3.64839971e-01 -9.82369542e-01 3.04120570e-01 -1.87207833e-01
7.15720952e-01 -1.47558296e+00 4.60625529e-01 -5.08973181e-01
-6.93667948e-01 1.18316150e+00 2.84710765e-01 2.17965934e-02
2.41811514e-01 2.70784378e-01 -9.15729161e-03 -1.57287776e-01
1.19425334e-01 3.20135981e-01 -6.00454919e-02 1.00075316e+00
2.28546672e-02 2.77818054e-01 -5.04716337e-01 -6.10727295e-02
1.61542855e-02 -2.04279751e-01 8.39577794e-01 1.26201427e+00
-3.79912227e-01 -1.12832201e+00 -1.51690578e+00 4.82423723e-01
-9.06117558e-01 -1.17841475e-01 -8.54816735e-01 9.56526816e-01
4.75097209e-01 1.14380658e+00 -9.74449143e-02 -2.09958375e-01
-2.35679969e-01 -2.45133013e-01 5.67306340e-01 -2.97347099e-01
-6.05349422e-01 -2.14085668e-01 1.44784078e-01 -2.36580640e-01
-4.41972971e-01 -7.51942456e-01 -6.99675441e-01 -9.21330094e-01
-4.30512518e-01 -4.12531227e-01 7.27971196e-01 9.03651416e-01
4.65288237e-02 -1.47018656e-01 4.74021554e-01 -4.82019007e-01
-4.21109378e-01 -7.83279955e-01 -4.75033939e-01 7.97426283e-01
1.58942834e-01 -6.39226615e-01 -4.48835790e-01 1.36908293e-01] | [12.492029190063477, 1.0879186391830444] |
d253ecd1-f2f0-482f-8cda-12b179c47e8f | unsupervised-pre-training-on-patient | 2203.12616 | null | https://arxiv.org/abs/2203.12616v2 | https://arxiv.org/pdf/2203.12616v2.pdf | Unsupervised Pre-Training on Patient Population Graphs for Patient-Level Predictions | Pre-training has shown success in different areas of machine learning, such as Computer Vision (CV), Natural Language Processing (NLP) and medical imaging. However, it has not been fully explored for clinical data analysis. Even though an immense amount of Electronic Health Record (EHR) data is recorded, data and labels can be scarce if the data is collected in small hospitals or deals with rare diseases. In such scenarios, pre-training on a larger set of EHR data could improve the model performance. In this paper, we apply unsupervised pre-training to heterogeneous, multi-modal EHR data for patient outcome prediction. To model this data, we leverage graph deep learning over population graphs. We first design a network architecture based on graph transformer designed to handle various input feature types occurring in EHR data, like continuous, discrete, and time-series features, allowing better multi-modal data fusion. Further, we design pre-training methods based on masked imputation to pre-train our network before fine-tuning on different end tasks. Pre-training is done in a fully unsupervised fashion, which lays the groundwork for pre-training on large public datasets with different tasks and similar modalities in the future. We test our method on two medical datasets of patient records, TADPOLE and MIMIC-III, including imaging and non-imaging features and different prediction tasks. We find that our proposed graph based pre-training method helps in modeling the data at a population level and further improves performance on the fine tuning tasks in terms of AUC on average by 4.15% for MIMIC and 7.64% for TADPOLE. | ['Nassir Navab', 'Anees Kazi', 'Chantal Pellegrini'] | 2022-03-23 | null | null | null | null | ['length-of-stay-prediction', 'unsupervised-pre-training'] | ['medical', 'methodology'] | [ 3.41948390e-01 3.36607754e-01 -2.56407142e-01 -6.22650564e-01
-8.96269262e-01 -2.90827483e-01 2.62101948e-01 7.52506077e-01
-3.08804601e-01 5.56220949e-01 5.35104752e-01 -5.05432665e-01
-2.97124803e-01 -9.27158654e-01 -7.12631047e-01 -5.27048767e-01
-2.96768099e-01 6.50140047e-01 -2.23058254e-01 1.31550118e-01
-4.40669686e-01 2.88171440e-01 -9.32360947e-01 6.33564711e-01
8.35054517e-01 9.68597710e-01 -6.46511987e-02 4.71349776e-01
2.90880166e-02 9.31362629e-01 -1.72421917e-01 -4.40021157e-01
2.15603426e-01 -2.45818049e-01 -6.94772005e-01 -6.54998869e-02
1.12769164e-01 6.85871691e-02 -4.06352520e-01 7.07710445e-01
8.75736713e-01 -3.11100513e-01 6.78473711e-01 -1.02861726e+00
-5.67365408e-01 8.64673913e-01 -5.59060693e-01 -2.80385315e-01
1.62138283e-01 3.12009305e-01 7.69218624e-01 -3.70509088e-01
6.76980257e-01 1.00430822e+00 1.15526617e+00 4.50838506e-01
-1.44463980e+00 -4.72164899e-01 -2.83665568e-01 6.89217867e-03
-1.16222572e+00 -1.52348980e-01 6.77508831e-01 -5.99391222e-01
6.73895419e-01 2.29574248e-01 5.47487617e-01 1.23523605e+00
5.14794052e-01 6.80332541e-01 1.06419230e+00 -9.06831473e-02
-1.42770978e-02 -7.29478374e-02 2.90488958e-01 6.36619687e-01
1.54602274e-01 1.06812306e-02 -1.50589257e-01 -4.59610730e-01
4.85744148e-01 6.12705112e-01 -2.09557533e-01 -1.02642946e-01
-1.48180652e+00 7.58373678e-01 5.47346175e-01 2.26071894e-01
-6.69114769e-01 4.51690257e-02 5.99012733e-01 3.61532986e-01
4.12682831e-01 4.44106698e-01 -6.64862216e-01 2.53073454e-01
-8.84203076e-01 -1.56957172e-02 6.80746198e-01 5.28356433e-01
4.78985399e-01 -3.08634192e-01 -6.25476003e-01 8.86879086e-01
2.25068629e-01 4.10239190e-01 4.86906797e-01 -5.42885125e-01
6.28160655e-01 8.41327548e-01 -4.43168730e-01 -8.66867781e-01
-1.06480730e+00 -6.61419392e-01 -1.52030444e+00 -3.60384256e-01
4.71359402e-01 -4.39024210e-01 -1.37937701e+00 1.88763165e+00
2.31621459e-01 3.30666184e-01 1.12102613e-01 7.86624789e-01
1.17942703e+00 4.01791096e-01 3.53147864e-01 -1.68887511e-01
1.63554347e+00 -5.54986417e-01 -6.11844599e-01 1.33740872e-01
1.16735601e+00 -3.87213856e-01 9.19008553e-01 2.50036895e-01
-6.41851485e-01 -2.64984399e-01 -4.52447295e-01 1.39505729e-01
-3.35684329e-01 1.34136885e-01 7.99060166e-01 3.32594573e-01
-8.41284394e-01 6.15395308e-01 -9.10844743e-01 -4.38365012e-01
7.32526660e-01 5.15561640e-01 -6.16650999e-01 -3.94380391e-01
-1.25141680e+00 6.15469873e-01 2.57155836e-01 -2.34224424e-02
-7.05947220e-01 -1.13638794e+00 -8.41814280e-01 1.82458460e-01
2.90410995e-01 -1.38332176e+00 6.94048226e-01 -5.77600062e-01
-1.07561636e+00 8.36491108e-01 2.20344916e-01 -7.02021182e-01
4.91465628e-01 1.93919137e-01 -4.94478583e-01 -1.86218768e-01
-2.93005388e-02 5.81454039e-01 5.64019918e-01 -9.24840987e-01
-1.88528389e-01 -7.39148498e-01 -3.48011076e-01 -1.81798965e-01
-2.56848186e-01 -2.27150917e-01 -1.97285771e-01 -7.40422368e-01
-1.59619376e-01 -1.04355097e+00 -6.44369721e-01 -4.96940255e-01
-6.81765974e-01 -6.42966852e-02 4.35982913e-01 -7.56052136e-01
1.12351489e+00 -2.05284119e+00 -2.49372777e-02 3.59687507e-01
6.92982078e-01 1.21489085e-01 -1.82082757e-01 4.65696275e-01
-3.48254055e-01 1.74567133e-01 -3.98583472e-01 -4.18899059e-01
-3.28747064e-01 1.08291335e-01 1.32798329e-01 3.41739297e-01
3.55704039e-01 1.26115274e+00 -6.59991384e-01 -5.68002820e-01
6.65521249e-02 4.44508433e-01 -7.59537518e-01 2.50000268e-01
-5.93998730e-02 9.46081221e-01 -5.40108919e-01 6.93577170e-01
5.82850218e-01 -8.31426024e-01 1.53804153e-01 -4.16555077e-01
3.67786467e-01 1.24573708e-01 -8.20390642e-01 1.96870220e+00
-4.67772126e-01 1.93184599e-01 -1.81770951e-01 -1.31610715e+00
6.86645031e-01 3.89407575e-01 1.22330832e+00 -6.75736129e-01
2.37139195e-01 -8.37293714e-02 3.45874518e-01 -6.37182415e-01
-5.55318110e-02 -1.47878766e-01 -2.04476655e-01 2.79270440e-01
4.71422710e-02 2.96935260e-01 -1.14517482e-02 1.83298945e-01
1.67767692e+00 -4.45389152e-01 5.36158383e-02 -3.87512632e-02
4.07502770e-01 1.72019884e-01 6.53935790e-01 8.23400259e-01
7.18230009e-02 8.23458672e-01 5.13626814e-01 -4.82227325e-01
-7.84488022e-01 -8.99174750e-01 -5.14619946e-01 6.91632032e-01
-2.82147050e-01 -5.94773054e-01 -4.10495192e-01 -8.43598425e-01
2.50851601e-01 3.26859504e-01 -6.71771407e-01 -3.14003974e-01
-3.30764025e-01 -1.32374513e+00 6.36692345e-01 4.67869818e-01
2.57986426e-01 -1.06926537e+00 -2.04545557e-01 3.18046629e-01
-1.14887804e-01 -1.21106815e+00 -2.01001987e-01 2.90280610e-01
-1.00173867e+00 -1.06424987e+00 -6.69160426e-01 -5.16164362e-01
5.73909938e-01 -2.65334129e-01 1.22819197e+00 7.00377449e-02
-4.04310733e-01 3.63621682e-01 -3.23133528e-01 -5.81918180e-01
-4.93954808e-01 3.62117559e-01 -1.63846761e-01 2.31681749e-01
2.34463185e-01 -5.21031797e-01 -7.38856435e-01 -9.71973129e-03
-9.43803668e-01 2.98106819e-01 8.60250235e-01 1.16634130e+00
8.02951813e-01 -2.49206915e-01 7.68102705e-01 -1.40615273e+00
5.12124717e-01 -8.05098355e-01 -1.53098479e-01 3.28754902e-01
-7.86319196e-01 1.65048048e-01 7.36145973e-01 -3.84333819e-01
-6.27736270e-01 2.57210702e-01 -2.77181149e-01 -5.39150178e-01
-2.60290295e-01 1.10623276e+00 1.32280841e-01 2.47563154e-01
6.71988487e-01 -1.11668237e-01 2.64826179e-01 -4.84857619e-01
6.94852546e-02 6.11977816e-01 2.71657556e-01 -3.36126894e-01
5.97626746e-01 3.32782060e-01 3.87931347e-01 -4.56530839e-01
-7.83563673e-01 -4.13593858e-01 -3.14579606e-01 1.60646617e-01
1.05414617e+00 -1.00766337e+00 -7.91331232e-01 3.05868119e-01
-7.96616137e-01 -3.20991933e-01 -4.13662791e-01 6.64344668e-01
-2.69551039e-01 9.09971073e-02 -6.29706919e-01 -3.45578581e-01
-7.43154764e-01 -1.19875908e+00 1.24184227e+00 -3.76140773e-02
-5.46886735e-02 -1.13575578e+00 2.61950016e-01 4.25578713e-01
4.57840800e-01 6.58192992e-01 1.13570201e+00 -9.95790362e-01
-3.53270143e-01 -2.86776930e-01 -3.46914649e-01 1.62152395e-01
2.62878925e-01 -3.88630807e-01 -8.33383977e-01 -2.24677458e-01
-2.71430910e-01 -3.21564645e-01 1.08778417e+00 7.55764008e-01
1.64969957e+00 -1.41624942e-01 -5.05678296e-01 8.86442065e-01
1.34075403e+00 -6.80668205e-02 5.97721040e-01 -5.59099354e-02
1.09166789e+00 5.41273534e-01 1.85382172e-01 4.70705211e-01
8.39194715e-01 4.91346776e-01 4.24625993e-01 -6.91680193e-01
1.72718000e-02 -9.06569213e-02 2.90483385e-02 6.94023788e-01
-1.14817157e-01 -1.41846582e-01 -1.36180818e+00 6.12509847e-01
-2.00535464e+00 -7.19913840e-01 -2.81442404e-01 2.13005853e+00
8.23954701e-01 -1.59020662e-01 1.73758432e-01 -1.63097039e-01
4.87382561e-01 -2.10165575e-01 -5.98463118e-01 -1.65304076e-02
-1.75912082e-02 3.28654975e-01 6.69150174e-01 -8.71422663e-02
-1.14182365e+00 5.66622496e-01 5.18858433e+00 5.64249754e-01
-1.36600924e+00 2.17740923e-01 1.07669830e+00 9.37139317e-02
-2.89327353e-01 -8.66133645e-02 -4.32076663e-01 5.36574781e-01
1.14486790e+00 1.23763993e-01 1.64039284e-01 5.32435358e-01
3.05424899e-01 2.79453993e-01 -1.20594633e+00 1.14095592e+00
-2.60658979e-01 -1.56654000e+00 -6.03332818e-02 2.69148886e-01
6.04039371e-01 4.79827970e-01 -9.71225426e-02 5.26294887e-01
2.86235332e-01 -1.33128262e+00 -5.85215501e-02 6.53221786e-01
8.79360199e-01 -3.90021890e-01 1.13923717e+00 4.20641929e-01
-7.68904626e-01 -4.58471067e-02 -1.01111062e-01 3.18341464e-01
1.89692631e-01 1.05646491e+00 -1.10080695e+00 9.71848488e-01
6.31310463e-01 1.05518675e+00 -6.90816641e-01 1.02240920e+00
3.68570089e-01 9.15533781e-01 -3.74144405e-01 3.99262577e-01
-1.06599763e-01 -1.40349790e-01 1.69030234e-01 1.19496393e+00
3.68260592e-01 2.18516633e-01 4.55585241e-01 7.23996997e-01
-3.71853143e-01 2.85922378e-01 -6.96454704e-01 -6.38866499e-02
1.81241818e-02 1.55576110e+00 -4.61863369e-01 -3.89811158e-01
-5.43065906e-01 3.07712078e-01 1.90316364e-01 3.03138465e-01
-1.00372422e+00 -5.89859188e-02 2.08431870e-01 4.08198714e-01
-1.59818590e-01 2.00451180e-01 -4.56243455e-01 -1.48245132e+00
-3.77398938e-01 -8.84003997e-01 7.97709227e-01 -2.85801530e-01
-1.77928185e+00 5.96020877e-01 -2.10913241e-01 -1.21932065e+00
-4.34832215e-01 -5.23323536e-01 -6.09450698e-01 6.69686556e-01
-1.43127573e+00 -1.41266978e+00 -5.02887726e-01 8.46198320e-01
-1.12931542e-01 -2.30811611e-01 8.64166737e-01 6.68635845e-01
-7.26561129e-01 7.53646910e-01 8.82142596e-03 4.72682148e-01
9.35559511e-01 -1.13874710e+00 2.21898593e-02 4.54261839e-01
6.75529167e-02 5.57278872e-01 2.08461806e-01 -6.70419395e-01
-1.61064541e+00 -1.47805297e+00 5.82820892e-01 -3.77055436e-01
5.85797846e-01 -2.67011732e-01 -1.03511047e+00 6.34806216e-01
-4.82333340e-02 3.76169294e-01 9.35333252e-01 4.15730536e-01
-5.73749319e-02 -3.37654352e-01 -1.25013173e+00 2.54775345e-01
9.04473484e-01 -3.15732837e-01 -2.33945310e-01 5.45455277e-01
7.59718359e-01 -4.47858900e-01 -1.63082957e+00 1.02451169e+00
3.81059706e-01 -5.05737126e-01 9.67049181e-01 -9.17815030e-01
5.48821032e-01 -3.70670892e-02 -5.56598008e-02 -1.46850944e+00
-2.73224622e-01 -3.98613036e-01 -6.15834258e-02 1.12436497e+00
6.86628997e-01 -8.22528183e-01 7.50617027e-01 6.37384653e-01
-1.97204843e-01 -1.06876647e+00 -6.69993937e-01 -3.15114826e-01
-1.73832271e-02 -5.90963900e-01 6.54644728e-01 1.27885163e+00
-1.28280565e-01 5.51326156e-01 -3.62729937e-01 -1.87369231e-02
5.65271974e-01 2.12866873e-01 8.68086100e-01 -1.47988451e+00
-5.07498145e-01 -1.15605772e-01 -4.70101237e-01 -3.11040282e-01
-2.39615765e-04 -1.39397395e+00 -4.59767044e-01 -1.76823127e+00
4.64546621e-01 -8.46019566e-01 -6.60360813e-01 9.29103971e-01
-2.90753752e-01 4.82391939e-02 3.05833984e-02 1.99061483e-01
-4.22740459e-01 4.14802462e-01 1.29359853e+00 -3.92116189e-01
-3.60461682e-01 7.57585317e-02 -6.84362113e-01 5.15923858e-01
7.06721961e-01 -6.70144141e-01 -2.38879114e-01 -3.24940413e-01
7.08323196e-02 5.62363267e-01 4.84306574e-01 -8.89483035e-01
-1.84724256e-02 8.41655135e-02 5.42398155e-01 -4.96002108e-01
-3.64470370e-02 -1.00133502e+00 4.37440276e-01 5.42264998e-01
-3.93975377e-01 1.21462911e-01 1.20404653e-01 6.09574854e-01
-2.80484885e-01 3.55689913e-01 5.42253792e-01 2.20707636e-02
-2.08960831e-01 6.94529533e-01 -1.56897737e-03 1.58645615e-01
8.82682800e-01 1.32219285e-01 -3.86135161e-01 -3.17519873e-01
-1.09063077e+00 5.24541795e-01 1.64737448e-01 2.75947869e-01
4.38810498e-01 -1.20730531e+00 -1.05876589e+00 2.45218188e-01
3.57503176e-01 1.77209675e-01 5.38988709e-01 1.38341129e+00
-2.63038248e-01 2.29978561e-01 -1.61149930e-02 -1.07671642e+00
-1.03091931e+00 6.17312491e-01 1.77672997e-01 -9.06242728e-01
-6.24303102e-01 1.88831955e-01 3.27462465e-01 -6.83727443e-01
2.19385587e-02 -6.06012523e-01 -3.34499031e-01 -1.59991104e-02
1.04021162e-01 -1.24874815e-01 3.85276526e-01 -4.51079100e-01
-4.41014826e-01 4.64694142e-01 -1.54671907e-01 3.96654367e-01
1.72390079e+00 2.43159264e-01 -1.31779388e-01 2.23244965e-01
1.39677227e+00 -2.19771788e-01 -6.35014296e-01 -3.11690032e-01
1.43955480e-02 4.68611233e-02 2.50164997e-02 -8.85240018e-01
-1.38643515e+00 7.08650112e-01 7.09843874e-01 1.71641856e-01
1.36726713e+00 9.08916667e-02 7.39934206e-01 2.84589767e-01
2.60660619e-01 -5.88525414e-01 -2.11757898e-01 3.53995591e-01
5.45559824e-01 -1.53638828e+00 -9.43834484e-02 -2.45986432e-01
-8.12397182e-01 8.03485215e-01 3.39469552e-01 -8.04923251e-02
9.36608732e-01 2.97222495e-01 -6.10656627e-02 -4.45787013e-01
-8.30161691e-01 -1.88782081e-01 6.05840087e-01 3.77833962e-01
6.34209454e-01 3.27735662e-01 -2.09589183e-01 8.78551364e-01
-4.96259541e-04 3.22253704e-01 3.71732622e-01 5.63552976e-01
2.88377196e-01 -1.20125329e+00 -1.98253915e-01 1.12990129e+00
-8.68671715e-01 -2.92638451e-01 -6.65933117e-02 6.57148123e-01
2.13073179e-01 6.76286578e-01 -1.49715289e-01 -5.64992547e-01
4.35440570e-01 1.13246523e-01 2.03084245e-01 -7.00902700e-01
-8.88087034e-01 5.99460192e-02 1.02801628e-01 -6.35083735e-01
-4.12208140e-01 -4.98011887e-01 -1.30712855e+00 -1.50429279e-01
-1.55868858e-01 1.76554918e-03 5.52098870e-01 8.94630313e-01
8.80046725e-01 9.84093666e-01 5.43400645e-01 -3.50206792e-01
-4.34118599e-01 -1.00284278e+00 -2.94326723e-01 7.86585271e-01
2.94291764e-01 -4.24739838e-01 4.33347523e-02 -3.03853280e-03] | [7.917031288146973, 6.38846492767334] |
4a5f9645-8bd4-4bda-8a98-9f0e11ba350d | continual-facial-expression-recognition-a | 2305.06448 | null | https://arxiv.org/abs/2305.06448v1 | https://arxiv.org/pdf/2305.06448v1.pdf | Continual Facial Expression Recognition: A Benchmark | Understanding human affective behaviour, especially in the dynamics of real-world settings, requires Facial Expression Recognition (FER) models to continuously adapt to individual differences in user expression, contextual attributions, and the environment. Current (deep) Machine Learning (ML)-based FER approaches pre-trained in isolation on benchmark datasets fail to capture the nuances of real-world interactions where data is available only incrementally, acquired by the agent or robot during interactions. New learning comes at the cost of previous knowledge, resulting in catastrophic forgetting. Lifelong or Continual Learning (CL), on the other hand, enables adaptability in agents by being sensitive to changing data distributions, integrating new information without interfering with previously learnt knowledge. Positing CL as an effective learning paradigm for FER, this work presents the Continual Facial Expression Recognition (ConFER) benchmark that evaluates popular CL techniques on FER tasks. It presents a comparative analysis of several CL-based approaches on popular FER datasets such as CK+, RAF-DB, and AffectNet and present strategies for a successful implementation of ConFER for Affective Computing (AC) research. CL techniques, under different learning settings, are shown to achieve state-of-the-art (SOTA) performance across several datasets, thus motivating a discussion on the benefits of applying CL principles towards human behaviour understanding, particularly from facial expressions, as well the challenges entailed. | ['Hatice Gunes', 'German I. Parisi', 'Tolga Dimlioglu', 'Nikhil Churamani'] | 2023-05-10 | null | null | null | null | ['facial-expression-recognition'] | ['computer-vision'] | [ 2.30657876e-01 1.02550704e-02 5.04321791e-02 -5.70244908e-01
5.40187247e-02 -2.15370268e-01 6.04456961e-01 9.86010954e-02
-5.93219280e-01 8.27156305e-01 -1.03377432e-01 4.87644911e-01
-1.16056584e-01 -3.40660572e-01 -3.79571497e-01 -6.61320925e-01
-4.95129704e-01 5.06798506e-01 -3.40980828e-01 -7.40948319e-01
-3.18529516e-01 7.29062796e-01 -1.96326256e+00 4.24482763e-01
1.48026779e-01 1.08830893e+00 -2.96981454e-01 5.17800450e-01
1.24984317e-01 1.17614543e+00 -5.99935532e-01 -5.78622818e-01
-2.21808076e-01 -3.14075738e-01 -8.43782961e-01 1.50548026e-01
1.09266832e-01 -1.48928374e-01 -3.16073373e-02 5.05252421e-01
5.20953774e-01 2.78502226e-01 5.46735227e-01 -1.63063717e+00
-3.52081478e-01 2.59095997e-01 -2.47228295e-01 -2.45066155e-02
6.03172302e-01 2.67680794e-01 4.43482637e-01 -7.70712852e-01
7.41654932e-01 1.22980940e+00 1.00031304e+00 9.89394963e-01
-1.12517416e+00 -4.93767142e-01 2.97480583e-01 5.19795954e-01
-1.00575745e+00 -7.88248420e-01 8.37841153e-01 -2.52993792e-01
1.23750162e+00 2.77485460e-01 8.43465388e-01 1.65866292e+00
1.60639450e-01 9.68937278e-01 1.18209434e+00 -3.76416355e-01
4.40737277e-01 3.79758269e-01 -2.61644889e-02 5.03508389e-01
-4.05560941e-01 1.49456412e-01 -8.42929244e-01 -1.15630828e-01
1.39801905e-01 -1.26185566e-01 3.81188467e-02 -2.30786890e-01
-5.85312366e-01 4.16331291e-01 3.24007213e-01 6.41377747e-01
-8.78552735e-01 2.95637604e-02 9.69381988e-01 7.28661656e-01
5.32984674e-01 3.46958667e-01 -7.47727334e-01 -7.65570939e-01
-4.62717086e-01 6.23705685e-02 7.57295012e-01 3.27273637e-01
9.83250797e-01 3.67710382e-01 -4.02761959e-02 9.63758707e-01
-1.53329253e-01 2.82088071e-01 7.91839719e-01 -1.03741360e+00
-5.63559175e-01 5.97921371e-01 -3.49734314e-02 -1.13344634e+00
-7.27135003e-01 -2.01662391e-01 -7.80052245e-01 2.54512817e-01
-8.98728967e-02 -3.71587515e-01 -6.17992461e-01 2.23606920e+00
3.45918566e-01 1.89242482e-01 1.75520942e-01 4.61912572e-01
4.88765597e-01 2.53348500e-01 5.03430963e-01 -6.46843076e-01
8.24614704e-01 -5.93854606e-01 -9.03174400e-01 -2.82529980e-01
6.50618672e-01 -2.71696091e-01 8.95586491e-01 6.57791734e-01
-7.91800201e-01 -4.80225593e-01 -6.69671714e-01 4.78288233e-01
-7.13007092e-01 -1.97682336e-01 1.08201265e+00 6.78414643e-01
-1.24613142e+00 3.68743718e-01 -6.11192524e-01 -6.07519031e-01
3.42843115e-01 6.78643763e-01 -5.71658492e-01 2.05920190e-01
-1.33995640e+00 1.11213255e+00 -3.90181541e-02 2.71457076e-01
-7.72488415e-01 -5.22716284e-01 -6.50491893e-01 -1.53600439e-01
2.95669436e-01 -4.35718328e-01 1.38457119e+00 -2.24900985e+00
-2.06738853e+00 9.00813162e-01 9.35106259e-03 -5.31365812e-01
4.56138849e-01 -2.11220756e-01 -6.21820867e-01 9.07677710e-02
-4.72765863e-01 7.64820516e-01 1.00417125e+00 -1.18380463e+00
-2.76624799e-01 -5.52947342e-01 1.26931787e-01 2.84685999e-01
-4.64927763e-01 -6.87668324e-02 1.15455218e-01 -4.25324775e-02
-7.49778807e-01 -1.06313205e+00 -5.59484363e-02 -6.57447577e-02
4.69777375e-01 -4.13154364e-01 1.07639146e+00 -2.81944841e-01
1.15404153e+00 -2.34070563e+00 3.10915262e-01 1.38053268e-01
-1.70109287e-01 5.13106704e-01 -1.90166593e-01 4.47793782e-01
-3.33337873e-01 -2.99176782e-01 1.07250102e-01 -6.61727250e-01
1.49840876e-01 5.31782568e-01 -2.11220570e-02 2.51060337e-01
1.80953547e-01 8.91198933e-01 -8.50597262e-01 -9.92718637e-02
3.78552347e-01 6.12069845e-01 -2.81643152e-01 2.89136976e-01
-1.68351859e-01 3.21035177e-01 1.59578361e-02 5.32163918e-01
3.41997921e-01 2.89864928e-01 3.35496515e-01 -8.38493034e-02
2.33592242e-01 -5.30684531e-01 -7.71443069e-01 1.40882266e+00
-8.18659961e-01 6.12817287e-01 2.33847663e-01 -9.44991052e-01
9.96751904e-01 4.60179716e-01 7.81469285e-01 -1.06243241e+00
3.23919326e-01 -5.94929140e-03 -1.41546041e-01 -6.42266750e-01
1.87743887e-01 -3.42820048e-01 -1.56188771e-01 3.55033159e-01
4.74672705e-01 2.43303254e-01 1.78278238e-02 -6.55845255e-02
1.19284129e+00 1.13889612e-01 3.77077192e-01 1.85063958e-01
6.16221011e-01 -4.49220419e-01 4.80372399e-01 3.71520668e-01
-7.64115751e-01 -1.27447486e-01 4.10561651e-01 -9.83706772e-01
-4.67354000e-01 -6.73895180e-01 1.57177240e-01 1.71169829e+00
-2.58714676e-01 -2.18184680e-01 -7.15587139e-01 -5.69908440e-01
-2.22095966e-01 7.91639924e-01 -1.18560398e+00 -7.66343653e-01
-2.01191172e-01 -8.35010946e-01 5.40653110e-01 3.18973660e-01
4.06892568e-01 -1.67467928e+00 -1.18927848e+00 4.36176628e-01
2.13942081e-02 -1.03282619e+00 3.23237658e-01 3.47725987e-01
-2.87924498e-01 -8.83770823e-01 -1.53347790e-01 -1.97442606e-01
2.92667776e-01 -2.45933637e-01 1.19291592e+00 5.26679307e-02
-4.30212080e-01 1.17045999e+00 -4.62070525e-01 -5.68592370e-01
-3.57178450e-01 -1.26847982e-01 4.04063553e-01 4.94498849e-01
7.16421723e-01 -7.97443688e-01 -1.92777470e-01 1.17918305e-01
-9.09058630e-01 -2.12920308e-01 4.92427319e-01 8.64429772e-01
1.19895689e-01 -1.54842883e-01 9.48849380e-01 -7.51984954e-01
8.39524269e-01 -7.09718347e-01 1.91017300e-01 3.51475239e-01
-5.13890445e-01 -3.93954545e-01 3.93918872e-01 -6.56447053e-01
-1.39024460e+00 1.03725225e-01 -1.99890554e-01 -5.63935280e-01
-5.44695139e-01 4.12596703e-01 3.41632143e-02 -4.08883579e-02
7.08801270e-01 3.30532119e-02 3.82547081e-01 1.74533743e-02
1.85385257e-01 7.05995381e-01 3.97655249e-01 -6.81320310e-01
-1.16540618e-01 3.78634006e-01 -6.79366961e-02 -8.25199246e-01
-7.62301087e-01 -1.53207675e-01 -8.76577973e-01 -8.61921489e-01
5.13392925e-01 -7.85018981e-01 -1.05104351e+00 7.92651951e-01
-8.17325294e-01 -6.84210122e-01 -4.76586998e-01 5.52006476e-02
-8.10736299e-01 -2.25688770e-01 -5.15705705e-01 -1.22176623e+00
-3.67348403e-01 -7.43243992e-01 9.25135791e-01 3.07967871e-01
-7.70067573e-01 -1.23868442e+00 3.23014408e-01 8.10685977e-02
6.26126289e-01 6.11563981e-01 7.05191255e-01 -5.46736479e-01
4.07207638e-01 -1.97041631e-01 3.38088781e-01 4.53241020e-01
2.02822194e-01 1.31391585e-01 -1.35112178e+00 -1.78806618e-01
-1.36611164e-01 -1.02691138e+00 3.54597807e-01 -1.78037509e-01
9.66146290e-01 -3.15289378e-01 6.16647266e-02 1.89882979e-01
1.05326593e+00 2.25204378e-01 6.62552953e-01 3.68823260e-01
1.95301697e-01 7.31489837e-01 6.33741856e-01 9.08402801e-01
4.90576088e-01 6.38273478e-01 5.51057458e-01 -8.62901732e-02
1.54272020e-01 6.98633194e-02 7.57090390e-01 4.01636124e-01
-2.09221572e-01 1.45695314e-01 -7.65281677e-01 3.40619236e-01
-2.02907252e+00 -1.19833648e+00 4.59773242e-01 1.76413560e+00
8.58662486e-01 -1.64770827e-01 3.05144906e-01 -5.82793262e-03
2.41614550e-01 2.38200910e-02 -8.62982512e-01 -1.16930819e+00
-2.49846816e-01 3.92339468e-01 -2.92935848e-01 3.45074415e-01
-9.55620229e-01 1.05414617e+00 5.97502089e+00 5.43287575e-01
-1.50995457e+00 9.59295034e-02 8.23166907e-01 -1.54594094e-01
5.80574423e-02 -6.43352270e-01 -1.62789091e-01 1.91900060e-01
1.36265171e+00 -1.11956783e-01 6.91457450e-01 1.00789213e+00
4.22938243e-02 -1.82464764e-01 -1.10846782e+00 1.14620233e+00
3.47408652e-01 -7.12551653e-01 -1.88605443e-01 -3.92875850e-01
3.92535836e-01 4.11036350e-02 3.95176500e-01 9.14783597e-01
1.41061783e-01 -1.03289294e+00 4.73748088e-01 8.65870953e-01
6.64231837e-01 -9.24661398e-01 9.74673927e-01 2.27600738e-01
-5.88859737e-01 -3.49997014e-01 4.84842286e-02 -4.65714276e-01
-2.26165861e-01 2.00726032e-01 -7.53246784e-01 3.53154600e-01
9.58074868e-01 8.64947259e-01 -6.66816533e-01 1.94979489e-01
1.75438806e-01 3.96157652e-01 -3.66403729e-01 -1.10317521e-01
2.19317496e-01 1.19865470e-01 3.18138689e-01 1.40692139e+00
-6.18372411e-02 3.97267267e-02 -8.77237990e-02 2.37159312e-01
6.91270903e-02 1.51190400e-01 -6.78510308e-01 4.17679474e-02
1.04830012e-01 1.43542349e+00 -2.05837205e-01 -1.44930065e-01
-1.14129737e-01 1.23604631e+00 6.61279321e-01 2.73625761e-01
-6.17153823e-01 3.04666817e-01 9.24291551e-01 -1.94837183e-01
-4.97259498e-02 2.62857825e-02 3.35874319e-01 -8.48624527e-01
-2.76630491e-01 -1.17163408e+00 5.26246190e-01 -8.50887477e-01
-1.35973513e+00 8.56661737e-01 4.22829809e-03 -6.63249791e-01
-6.38348460e-01 -4.89385456e-01 -4.22577262e-01 1.15810178e-01
-1.34746289e+00 -1.36013556e+00 -4.23106939e-01 8.64396930e-01
4.69214827e-01 -2.37085432e-01 1.39616227e+00 2.59012669e-01
-5.85901976e-01 5.99998236e-01 -1.87922031e-01 -3.85116577e-01
8.28653276e-01 -9.93303537e-01 -4.56561238e-01 -2.59004621e-04
3.84997167e-02 2.41016835e-01 7.63944209e-01 -1.31958336e-01
-1.33131135e+00 -8.65898252e-01 5.89017570e-01 -4.20920342e-01
6.37906313e-01 -3.14367712e-01 -9.17142153e-01 7.31635988e-01
3.75227183e-01 2.11484637e-03 1.00358820e+00 3.61261010e-01
-3.13471556e-01 -3.95789415e-01 -1.30370760e+00 5.16535699e-01
9.73651826e-01 -5.44470608e-01 -2.48672754e-01 -2.48080632e-03
1.22295171e-01 -7.20584244e-02 -7.05680013e-01 6.95380986e-01
8.58549595e-01 -1.32946491e+00 6.34933889e-01 -7.33789921e-01
4.06437218e-02 2.45161459e-01 -1.26929849e-01 -1.55780876e+00
-1.83652207e-01 -8.89139771e-01 -1.68493271e-01 1.17684364e+00
-2.52761059e-02 -5.17640829e-01 4.44477022e-01 8.88068080e-01
2.32757434e-01 -9.44275320e-01 -1.12042189e+00 -2.88166016e-01
-2.15317890e-01 -5.90486825e-01 5.47508717e-01 1.17024815e+00
2.44758964e-01 3.87565702e-01 -6.10929787e-01 -3.80170047e-01
4.46025021e-02 -3.09626877e-01 9.50627148e-01 -1.21354687e+00
-1.07499763e-01 -4.20096189e-01 -7.09169865e-01 1.79257989e-02
7.17051983e-01 -3.65260452e-01 -1.55424625e-01 -6.91113830e-01
3.20740379e-02 -1.57881349e-01 -6.51906788e-01 8.18063259e-01
-3.74909304e-02 2.48997465e-01 2.45389581e-01 -1.13588624e-01
-1.07889199e+00 9.44920063e-01 5.87039590e-01 4.15425599e-02
-2.69351810e-01 -9.16670933e-02 -3.69191766e-01 1.05980861e+00
8.64890933e-01 -8.79986882e-02 -4.86018568e-01 1.07095614e-02
4.34395701e-01 -2.73777544e-01 2.31132731e-01 -1.17859757e+00
1.90762524e-02 -2.21295401e-01 4.56377953e-01 2.01048985e-01
7.54827023e-01 -1.09932089e+00 2.91332513e-01 1.68881819e-01
-2.92259276e-01 1.31604791e-01 5.74805737e-01 3.36037308e-01
-1.82768688e-01 4.64606695e-02 7.13188350e-01 -2.05021411e-01
-1.17974782e+00 2.92987376e-01 -6.61455274e-01 -2.99235404e-01
1.26280499e+00 -1.49399757e-01 8.83205011e-02 -7.53095269e-01
-1.25558400e+00 5.69115654e-02 2.41625771e-01 6.94093168e-01
4.56545264e-01 -1.10632563e+00 -4.57879990e-01 1.98236898e-01
2.89501935e-01 -6.63315415e-01 5.70380986e-01 8.45430672e-01
4.85854410e-02 -1.39637023e-01 -7.36391902e-01 -4.06539768e-01
-1.56264532e+00 4.08773124e-01 7.59963870e-01 -1.73744947e-01
-3.87614667e-02 8.30443919e-01 -8.38781595e-02 -4.38058227e-01
2.91847050e-01 3.50043744e-01 -4.09848481e-01 4.58093882e-01
5.27196109e-01 1.83975026e-01 1.37868389e-01 -8.40971887e-01
-3.23549837e-01 1.65418789e-01 -9.45145488e-02 2.40546856e-02
1.61406827e+00 -3.85296196e-01 -2.19663337e-01 1.03589582e+00
1.11007488e+00 -5.59314787e-01 -1.16828537e+00 -2.58765936e-01
6.97131529e-02 -1.49153486e-01 -3.19994949e-02 -1.20869577e+00
-8.23675036e-01 8.01415861e-01 9.45500076e-01 -1.08118467e-01
1.54537976e+00 -1.62360385e-01 3.84163946e-01 6.43567026e-01
6.10346496e-01 -1.54251575e+00 6.03928626e-01 5.50605655e-01
1.05410838e+00 -1.12713182e+00 -2.13802323e-01 2.21109286e-01
-1.10904002e+00 1.31397295e+00 8.89254451e-01 2.85394251e-01
7.02440619e-01 3.08109283e-01 2.87160397e-01 -2.47009605e-01
-1.33804023e+00 -2.04584524e-01 -2.62521684e-01 7.03789532e-01
2.64521390e-01 1.75725102e-01 2.09965080e-01 5.81826031e-01
-4.23970073e-02 2.48454049e-01 2.97269821e-01 1.01044428e+00
-8.95905271e-02 -9.38254297e-01 1.13207541e-01 1.05385758e-01
-3.30432862e-01 4.37255532e-01 -8.52675915e-01 7.21420109e-01
5.00286937e-01 7.00331032e-01 1.59179643e-01 -3.97529364e-01
4.59701985e-01 5.06869078e-01 5.68005145e-01 -2.68031269e-01
-8.30851793e-01 -3.79978359e-01 2.49232218e-01 -7.46492326e-01
-9.12139833e-01 -9.79665339e-01 -1.07080805e+00 -2.65657216e-01
5.63811623e-02 -1.17019281e-01 5.04020393e-01 1.00217700e+00
5.76980948e-01 4.25485462e-01 7.86252141e-01 -9.14501727e-01
-1.62882075e-01 -9.71289873e-01 -4.41572219e-01 8.48721802e-01
2.48489544e-01 -9.28476095e-01 -3.00976872e-01 1.53694838e-01] | [13.549811363220215, 1.999802589416504] |
e6b6cd33-d726-402d-a2f0-d2ff3efb6fa3 | multi-task-consistency-for-active-learning | 2306.12398 | null | https://arxiv.org/abs/2306.12398v1 | https://arxiv.org/pdf/2306.12398v1.pdf | Multi-Task Consistency for Active Learning | Learning-based solutions for vision tasks require a large amount of labeled training data to ensure their performance and reliability. In single-task vision-based settings, inconsistency-based active learning has proven to be effective in selecting informative samples for annotation. However, there is a lack of research exploiting the inconsistency between multiple tasks in multi-task networks. To address this gap, we propose a novel multi-task active learning strategy for two coupled vision tasks: object detection and semantic segmentation. Our approach leverages the inconsistency between them to identify informative samples across both tasks. We propose three constraints that specify how the tasks are coupled and introduce a method for determining the pixels belonging to the object detected by a bounding box, to later quantify the constraints as inconsistency scores. To evaluate the effectiveness of our approach, we establish multiple baselines for multi-task active learning and introduce a new metric, mean Detection Segmentation Quality (mDSQ), tailored for the multi-task active learning comparison that addresses the performance of both tasks. We conduct extensive experiments on the nuImages and A9 datasets, demonstrating that our approach outperforms existing state-of-the-art methods by up to 3.4% mDSQ on nuImages. Our approach achieves 95% of the fully-trained performance using only 67% of the available data, corresponding to 20% fewer labels compared to random selection and 5% fewer labels compared to state-of-the-art selection strategy. Our code will be made publicly available after the review process. | ['Alois C. Knoll', 'Alvaro Marcos-Ramiro', 'Michael Schmidt', 'Walter Zimmer', 'Philipp Friedrich', 'Aral Hekimoglu'] | 2023-06-21 | null | null | null | null | ['active-learning', 'active-learning'] | ['methodology', 'natural-language-processing'] | [ 6.05850637e-01 5.32254055e-02 -3.42594028e-01 -4.34594333e-01
-1.54354751e+00 -5.23780644e-01 5.44744313e-01 9.89127681e-02
-7.86342561e-01 4.72500861e-01 -2.10829392e-01 1.05345249e-01
-6.96526840e-02 -1.84713185e-01 -7.14075446e-01 -7.63620675e-01
2.46480823e-01 4.84363437e-01 7.12027311e-01 5.76375127e-01
3.01811963e-01 1.57923073e-01 -1.59091091e+00 2.84778178e-01
9.35131490e-01 1.17666388e+00 5.13236344e-01 6.08807981e-01
-1.17224723e-01 6.57888114e-01 -6.63462162e-01 -3.17658991e-01
3.32368225e-01 -1.15539476e-01 -9.23442423e-01 4.24747288e-01
9.25087452e-01 -2.78515846e-01 4.61698323e-01 9.47029591e-01
5.91604829e-01 1.38168797e-01 6.06982768e-01 -1.35151446e+00
-2.82608539e-01 3.53697628e-01 -7.08818197e-01 2.09425718e-01
-5.95235918e-03 3.50094646e-01 1.22841930e+00 -1.20050073e+00
4.27820235e-01 1.04349148e+00 5.92950940e-01 4.50369358e-01
-1.32527781e+00 -5.95791817e-01 4.45481002e-01 1.31874129e-01
-1.13515031e+00 -8.24204326e-01 5.93071878e-01 -6.58296406e-01
8.19536090e-01 1.93931386e-02 4.87349510e-01 1.10662401e+00
-1.85020149e-01 1.11981928e+00 1.05730724e+00 -5.73758662e-01
3.93728018e-01 1.55748025e-01 3.76350522e-01 8.76305878e-01
4.07700717e-01 -7.04218820e-02 -7.13130832e-01 -1.72487810e-01
6.30565584e-01 -2.76763052e-01 -5.96486516e-02 -5.86551547e-01
-1.16213155e+00 8.39062035e-01 8.57687220e-02 -7.03650564e-02
-1.86753199e-01 1.43982649e-01 2.93155283e-01 -1.26099616e-01
8.32721114e-01 4.76954162e-01 -5.50016046e-01 -4.76235226e-02
-1.07854903e+00 4.90596704e-02 5.66078961e-01 7.50950038e-01
7.63517559e-01 -1.76016361e-01 -4.59984154e-01 1.11720204e+00
6.75262034e-01 2.98492014e-01 -2.81741824e-02 -1.28449762e+00
3.24824095e-01 6.41184986e-01 5.70464656e-02 -5.43345034e-01
-2.60317266e-01 -4.34501678e-01 -2.72138923e-01 4.79787529e-01
6.25362456e-01 -1.50212020e-01 -1.10647929e+00 1.56442678e+00
3.15512955e-01 1.89270154e-01 -3.16518635e-01 7.51615942e-01
6.74816370e-01 2.95199245e-01 2.62043029e-01 -3.29728186e-01
1.10121167e+00 -1.33259010e+00 -4.36302811e-01 -6.72263861e-01
3.40457976e-01 -7.61723816e-01 1.22912681e+00 4.18642730e-01
-1.08332253e+00 -5.92319131e-01 -1.01450932e+00 3.61090302e-02
3.71314283e-03 4.32275295e-01 4.94871020e-01 5.77221870e-01
-9.34049785e-01 1.36986539e-01 -9.94490206e-01 -2.11499691e-01
7.67806053e-01 1.95243955e-01 -2.94538517e-03 -1.66342072e-02
-5.56925535e-01 7.09848106e-01 3.58473927e-01 -7.51154870e-02
-1.13425803e+00 -7.75925040e-01 -7.01563299e-01 -1.86714187e-01
8.13059032e-01 -6.01819336e-01 1.37692368e+00 -1.05042779e+00
-1.12175107e+00 1.08087766e+00 -7.79418796e-02 -2.70880163e-01
6.58396959e-01 -3.45586896e-01 8.55435953e-02 2.08358362e-01
3.08804065e-01 1.00812483e+00 9.04904008e-01 -1.51970661e+00
-9.53169227e-01 -3.35345954e-01 1.46582827e-01 2.94954479e-01
-1.91770867e-01 1.37264162e-01 -1.01503432e+00 -5.51762402e-01
6.36986867e-02 -1.03128505e+00 -1.92954555e-01 5.60328364e-01
-3.73500764e-01 -5.01976967e-01 8.57282043e-01 -3.62647265e-01
8.34586620e-01 -2.08213782e+00 -7.61228576e-02 -8.62000808e-02
3.79230797e-01 4.12998676e-01 -2.11398929e-01 -2.05672950e-01
4.20092791e-01 5.35129011e-02 -5.16242981e-01 -1.03011346e+00
-1.11769147e-01 1.33119479e-01 8.53697956e-02 3.24325591e-01
4.52480495e-01 8.20103526e-01 -9.65469122e-01 -8.67618620e-01
2.73649096e-01 3.02473515e-01 -1.93055347e-01 2.82850504e-01
-3.45824003e-01 4.34199274e-01 -3.15529406e-01 9.26851809e-01
4.66654480e-01 -4.43515807e-01 6.38695732e-02 -2.95122117e-01
-2.67554093e-02 -9.60181095e-03 -1.17821491e+00 1.72791278e+00
-2.44398758e-01 6.17730021e-01 1.22204117e-01 -9.68433678e-01
7.13217497e-01 1.39820978e-01 5.89047968e-01 -6.54714167e-01
-1.54204413e-01 2.02638626e-01 -6.49723783e-02 -4.99788821e-01
2.87242770e-01 4.57709223e-01 1.60359442e-01 5.07764041e-01
2.54725218e-01 2.18170453e-02 4.15120125e-01 1.13281809e-01
1.05132711e+00 3.18533927e-01 3.45564425e-01 -7.37438351e-02
2.22935528e-01 -3.25634144e-02 6.53435767e-01 1.20162499e+00
-6.86122358e-01 6.54228568e-01 1.68184683e-01 -9.14159790e-02
-8.76799107e-01 -9.85362709e-01 -1.51673600e-01 1.46871233e+00
8.40772241e-02 -1.39866263e-01 -6.66132748e-01 -1.02589190e+00
-1.39281154e-01 5.29137492e-01 -5.20899653e-01 1.53983429e-01
-4.11177278e-01 -9.65027034e-01 3.96712482e-01 5.19279122e-01
5.87247193e-01 -1.06127608e+00 -8.28470051e-01 -1.73361063e-01
-2.19584495e-01 -1.29868615e+00 -3.91965419e-01 3.25317383e-01
-7.57659972e-01 -1.35099435e+00 -6.31137133e-01 -5.68121314e-01
6.40038192e-01 4.13240910e-01 1.30340493e+00 -1.11451615e-02
-3.41310024e-01 6.49120569e-01 -4.14381057e-01 -5.14221251e-01
-3.32684398e-01 1.62238434e-01 -2.76279747e-01 -7.72043243e-02
1.75576165e-01 1.20265754e-02 -5.63137650e-01 4.64574873e-01
-7.76493609e-01 1.24526143e-01 7.70368934e-01 5.70349097e-01
8.97331357e-01 -4.04541671e-01 4.83717203e-01 -8.95448625e-01
2.87241399e-01 -2.09628895e-01 -6.50578976e-01 5.80137610e-01
-8.19707930e-01 -9.33110788e-02 -1.61614940e-02 -4.53497797e-01
-1.04213333e+00 5.44973791e-01 8.10342878e-02 -3.30426455e-01
-1.46805882e-01 2.40163803e-01 -1.48639992e-01 -3.53364468e-01
5.97381771e-01 -2.03251690e-01 -1.69183910e-01 -3.16135049e-01
3.67882699e-01 6.61062658e-01 4.94548500e-01 -5.31539738e-01
5.06936133e-01 4.61070508e-01 -2.85293847e-01 -5.51634610e-01
-1.37250733e+00 -7.58785248e-01 -7.22499013e-01 -4.57826942e-01
8.18349719e-01 -9.33923841e-01 -3.86578619e-01 6.19670451e-01
-1.07995450e+00 -4.84727561e-01 -2.64975101e-01 3.97055298e-01
-6.09661341e-01 4.05400425e-01 -2.75347322e-01 -1.11327744e+00
-5.02221882e-01 -1.47982562e+00 1.31047428e+00 6.82206675e-02
-1.96115211e-01 -9.51137185e-01 -1.58775803e-02 8.49660993e-01
2.15616316e-01 -3.60254534e-02 5.60536861e-01 -8.39896142e-01
-7.82824218e-01 1.16730727e-01 -3.62172544e-01 4.25609946e-01
2.07381248e-02 1.82979122e-01 -1.28555477e+00 -3.37800860e-01
-1.16320655e-01 -7.34808147e-01 1.28822637e+00 6.71302259e-01
1.24022186e+00 7.67446756e-02 -3.70308131e-01 3.79255444e-01
1.26137829e+00 2.14243308e-01 3.76114190e-01 3.32326174e-01
7.27970362e-01 6.02713883e-01 9.18358505e-01 3.06480825e-01
3.75837952e-01 8.53792191e-01 5.02719462e-01 -3.17404360e-01
-3.99307936e-01 2.40261316e-01 3.52887899e-01 3.96080852e-01
7.78664872e-02 -2.68077672e-01 -1.01270366e+00 5.14539719e-01
-2.08073306e+00 -7.08796144e-01 -8.36819857e-02 2.42621636e+00
9.80594158e-01 5.13960123e-01 2.40761057e-01 7.75571093e-02
6.90371692e-01 1.42640620e-01 -8.18630278e-01 1.84524283e-01
7.59691596e-02 -3.72024230e-03 4.65218872e-01 4.96192604e-01
-1.57726717e+00 8.03357840e-01 6.76989746e+00 9.46309745e-01
-9.19898570e-01 3.57829452e-01 7.85603940e-01 -1.52419731e-01
1.43370181e-01 -1.71267223e-02 -1.00102460e+00 4.04594898e-01
5.11119545e-01 3.10140580e-01 5.10308333e-03 8.37467253e-01
3.23225647e-01 -5.67049623e-01 -1.22926331e+00 1.01361513e+00
3.42432588e-01 -1.09927821e+00 -2.06613004e-01 -1.02544479e-01
8.36884439e-01 1.94632098e-01 -8.12364928e-03 5.11982739e-02
3.49379659e-01 -6.63401783e-01 9.20889318e-01 4.89249915e-01
7.13615298e-01 -2.95682967e-01 4.69046295e-01 3.39763373e-01
-1.09418595e+00 -1.43996075e-01 -1.11036703e-01 2.16165408e-01
1.58914328e-01 8.16812098e-01 -8.59323800e-01 3.63622338e-01
7.40322411e-01 6.05705678e-01 -9.80376482e-01 1.32257593e+00
-9.22938511e-02 7.38864005e-01 -1.63870171e-01 2.72476047e-01
4.93409857e-02 -7.47368718e-03 5.43522954e-01 1.13092816e+00
-1.06525280e-01 -5.06594419e-01 7.03725636e-01 9.96714056e-01
-3.85828726e-02 -2.17042685e-01 -6.20261207e-02 1.42170444e-01
6.42307818e-01 1.40152049e+00 -1.01382899e+00 -3.21308881e-01
-4.27587658e-01 8.96418393e-01 3.67245048e-01 3.74033213e-01
-8.31540227e-01 1.70365963e-02 3.10024649e-01 -2.83370644e-01
2.36878961e-01 -2.86984265e-01 -4.51223791e-01 -8.99299085e-01
7.35864639e-02 -7.52911150e-01 4.00321871e-01 -6.56721711e-01
-1.49187493e+00 3.32162768e-01 2.06070036e-01 -9.85765636e-01
-2.49138013e-01 -5.53285539e-01 -3.80469114e-01 6.43138289e-01
-1.49354494e+00 -1.32383859e+00 -5.63579917e-01 2.49578029e-01
9.50003088e-01 -2.32800409e-01 6.41525507e-01 3.26924860e-01
-7.82681525e-01 5.24441421e-01 -1.51010200e-01 1.66517287e-01
9.31633830e-01 -1.27808738e+00 4.49925840e-01 1.03602612e+00
3.24892104e-01 3.41435015e-01 3.83311421e-01 -6.05241656e-01
-8.80373716e-01 -1.11299920e+00 4.46531981e-01 -6.91047788e-01
3.46276909e-01 -3.67480248e-01 -7.42409348e-01 6.05426550e-01
-8.73151701e-03 2.50061035e-01 6.97096527e-01 1.26475736e-01
-3.98297340e-01 -1.04970530e-01 -9.61523294e-01 2.18924746e-01
1.04776359e+00 -4.51530069e-01 -3.69844198e-01 3.80124956e-01
7.20162511e-01 -3.01566273e-01 -4.91170794e-01 6.00471437e-01
5.08447945e-01 -9.67308044e-01 1.02485609e+00 -1.22280568e-01
2.06706002e-01 -3.10588896e-01 -5.61295450e-02 -1.00061953e+00
-2.94791609e-01 -1.43699020e-01 -2.43474022e-01 1.30103183e+00
6.76675498e-01 -3.54568094e-01 7.30021417e-01 5.16831934e-01
-3.39138448e-01 -9.80019987e-01 -8.67528677e-01 -5.62916219e-01
-2.94527084e-01 -5.55044055e-01 -2.51100361e-01 7.33927965e-01
-7.45979071e-01 5.21234870e-01 -1.36078209e-01 -2.05677897e-02
9.48582649e-01 -6.76007792e-02 5.67067802e-01 -1.42502177e+00
-3.48822773e-01 -4.62047398e-01 -3.30356397e-02 -8.92112970e-01
1.71260878e-01 -5.98893166e-01 3.56620789e-01 -1.80449951e+00
6.48900628e-01 -6.83543563e-01 -2.45809659e-01 7.90384769e-01
-4.43021595e-01 5.58688939e-01 2.42487863e-01 5.33152521e-01
-1.12895739e+00 3.12740445e-01 8.88523936e-01 -3.67298484e-01
-2.41510540e-01 1.10508129e-01 -5.75709522e-01 8.41793776e-01
5.13114691e-01 -4.26252097e-01 -5.55214524e-01 -5.18036723e-01
3.47801410e-02 -4.06072736e-01 5.57864845e-01 -1.09011805e+00
3.59448701e-01 -1.13047004e-01 3.82850677e-01 -6.09299958e-01
5.52100122e-01 -6.44486129e-01 -2.15103701e-01 1.71089470e-01
-6.28289700e-01 -2.74310738e-01 2.46826541e-02 6.57801569e-01
4.07466628e-02 -4.16363597e-01 9.86460030e-01 -1.74483970e-01
-1.04272437e+00 3.36427778e-01 -2.36207135e-02 2.76255250e-01
1.08566093e+00 -3.76033157e-01 -4.44225222e-01 -9.35895462e-03
-5.86204052e-01 3.02461594e-01 3.75863671e-01 3.58057737e-01
6.15734696e-01 -8.45766783e-01 -6.89928234e-01 -1.04542181e-01
4.23839122e-01 3.00084203e-01 9.18324199e-03 9.97327447e-01
-1.46336243e-01 1.79815575e-01 7.10111707e-02 -1.02614069e+00
-1.53680587e+00 -2.91297375e-03 4.01707947e-01 -2.19003171e-01
-3.72215569e-01 9.55839157e-01 7.96499774e-02 -3.38274032e-01
5.61862886e-01 -6.79177567e-02 -1.69083863e-01 3.72552305e-01
3.41369748e-01 4.93330121e-01 2.85524964e-01 -3.47061992e-01
-4.44752425e-01 5.44827163e-01 -1.35111973e-01 -2.20526591e-01
1.06514311e+00 -2.46093646e-01 8.77221599e-02 7.00933337e-01
9.52754557e-01 -2.02287480e-01 -1.70398581e+00 -3.66057843e-01
2.33241513e-01 -4.79555100e-01 2.04472229e-01 -1.06621540e+00
-1.04471421e+00 6.44269347e-01 9.04128194e-01 9.34660956e-02
1.04788136e+00 2.23595858e-01 3.64524305e-01 3.87340218e-01
2.06707805e-01 -1.27331984e+00 5.86413741e-01 3.62303078e-01
5.83044827e-01 -1.74574208e+00 1.73349321e-01 -5.95498502e-01
-8.03680003e-01 6.90596163e-01 8.47136140e-01 2.18176648e-01
3.06145847e-01 3.41526002e-01 2.51135439e-01 -1.29184335e-01
-7.22891867e-01 -4.83248740e-01 5.02752542e-01 6.37036741e-01
3.84221315e-01 -2.25614786e-01 -3.11654899e-02 1.80046797e-01
4.85768378e-01 -3.92749943e-02 2.36643344e-01 9.25194025e-01
-5.94799280e-01 -9.47274566e-01 -3.45794618e-01 5.39875269e-01
-3.73252779e-01 -7.12828769e-04 -5.53405344e-01 4.95506376e-01
3.25547457e-01 1.18858898e+00 2.23848924e-01 -8.08486938e-02
4.15791683e-02 1.30905077e-01 5.88536322e-01 -8.41779172e-01
-3.46614450e-01 2.41199255e-01 1.95822701e-01 -6.18929625e-01
-1.09706497e+00 -6.27209723e-01 -1.01025629e+00 2.91540205e-01
-5.41296899e-01 -4.04900223e-01 5.67595720e-01 1.13697302e+00
3.46634001e-01 4.74008203e-01 4.01832283e-01 -9.49757934e-01
-6.27913356e-01 -8.82804096e-01 -2.91197062e-01 3.69298279e-01
3.16053122e-01 -9.10012960e-01 -3.52526695e-01 1.91711575e-01] | [9.257956504821777, 1.2883515357971191] |
b1f26c81-1932-4549-8fe8-02c37751907b | bert-erc-fine-tuning-bert-is-enough-for | 2301.06745 | null | https://arxiv.org/abs/2301.06745v1 | https://arxiv.org/pdf/2301.06745v1.pdf | BERT-ERC: Fine-tuning BERT is Enough for Emotion Recognition in Conversation | Previous works on emotion recognition in conversation (ERC) follow a two-step paradigm, which can be summarized as first producing context-independent features via fine-tuning pretrained language models (PLMs) and then analyzing contextual information and dialogue structure information among the extracted features. However, we discover that this paradigm has several limitations. Accordingly, we propose a novel paradigm, i.e., exploring contextual information and dialogue structure information in the fine-tuning step, and adapting the PLM to the ERC task in terms of input text, classification structure, and training strategy. Furthermore, we develop our model BERT-ERC according to the proposed paradigm, which improves ERC performance in three aspects, namely suggestive text, fine-grained classification module, and two-stage training. Compared to existing methods, BERT-ERC achieves substantial improvement on four datasets, indicating its effectiveness and generalization capability. Besides, we also set up the limited resources scenario and the online prediction scenario to approximate real-world scenarios. Extensive experiments demonstrate that the proposed paradigm significantly outperforms the previous one and can be adapted to various scenes. | ['Li Wang', 'Bin Wang', 'Jian Luan', 'Yanran Li', 'Tingting Zhang', 'Jinshi Cui', 'Zhiyu Wu', 'Xiangyu Qin'] | 2023-01-17 | null | null | null | null | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [ 2.05745444e-01 -3.30047339e-01 8.64086077e-02 -5.79596877e-01
-5.29741585e-01 -4.38915551e-01 7.17269242e-01 -5.12717701e-02
-6.26501739e-01 4.94544089e-01 4.75005865e-01 -1.18004337e-01
6.17608465e-02 -4.50909823e-01 -9.63397920e-02 -5.21183193e-01
2.46508002e-01 3.36442441e-01 1.05151333e-01 -4.59683150e-01
4.95420128e-01 3.30039054e-01 -1.38196504e+00 6.79019511e-01
1.11197793e+00 1.17642474e+00 3.33053738e-01 6.41642749e-01
-6.12307549e-01 6.82876468e-01 -6.51928306e-01 -4.79074299e-01
-3.22152197e-01 -4.83062565e-01 -8.40894639e-01 2.63158798e-01
-5.91594517e-01 -1.61445826e-01 -8.55891407e-03 5.61073422e-01
4.77821529e-01 4.48474467e-01 5.74584126e-01 -1.06592989e+00
-3.69668365e-01 4.62276667e-01 -4.82832968e-01 -5.18716350e-02
5.22864938e-01 -2.22465172e-01 9.25489068e-01 -1.19622600e+00
3.18638235e-01 1.41728354e+00 5.59320569e-01 6.98043466e-01
-5.69332302e-01 -5.60902655e-01 7.39605188e-01 3.04037303e-01
-1.08465028e+00 -5.15305758e-01 1.01560116e+00 -3.14060867e-01
1.03841853e+00 2.59855062e-01 4.15183187e-01 1.24507272e+00
-1.50451332e-01 1.22097754e+00 1.47943509e+00 -6.04201376e-01
1.99883282e-01 6.78974688e-01 1.97803169e-01 5.19293547e-01
-5.26792765e-01 -3.10817748e-01 -6.40316546e-01 -2.25457683e-01
2.77788073e-01 -1.39778018e-01 -2.31196329e-01 7.35842809e-03
-1.01598418e+00 8.21997702e-01 -2.21921224e-02 4.32469964e-01
-3.18233311e-01 -5.41157782e-01 8.50025237e-01 2.73435831e-01
5.75993657e-01 1.86690912e-01 -7.01580346e-01 -5.78097939e-01
-5.03463447e-01 -9.20005366e-02 1.10334933e+00 1.09195697e+00
5.19529104e-01 -1.55998677e-01 -2.38109291e-01 1.37253380e+00
3.52764696e-01 2.52661228e-01 8.53153288e-01 -4.95218545e-01
7.16399193e-01 7.45946944e-01 9.15667042e-03 -1.04846358e+00
-5.43722212e-01 -2.08632797e-01 -9.68300462e-01 -5.43817341e-01
-1.11553751e-01 -4.88518000e-01 -3.57047945e-01 1.51237428e+00
3.34167272e-01 1.53892729e-02 4.49091196e-01 7.65364468e-01
1.07502031e+00 8.61624599e-01 2.48182133e-01 -5.79427898e-01
1.42372179e+00 -1.42298353e+00 -1.04954195e+00 -2.75041670e-01
6.12868905e-01 -8.53224397e-01 1.35200965e+00 6.21322453e-01
-6.35398924e-01 -4.77861315e-01 -7.51878738e-01 -4.83621545e-02
-5.91655672e-01 5.86980402e-01 7.63959289e-01 6.10436618e-01
-6.70880854e-01 2.28570014e-01 -1.94244280e-01 -4.45419222e-01
-1.41091764e-01 2.34097540e-02 -1.45335361e-01 1.77851737e-01
-1.42317879e+00 6.47336721e-01 6.48636341e-01 5.07265627e-01
-3.97381186e-01 -5.18777706e-02 -7.42198706e-01 2.04079643e-01
6.35949433e-01 -1.67684034e-01 1.17992485e+00 -1.09336078e+00
-2.26473141e+00 5.29247046e-01 -4.15125728e-01 -3.88950557e-02
4.19041604e-01 -3.13589334e-01 -7.24556804e-01 5.21732680e-02
-3.77661973e-01 2.80098289e-01 6.69193566e-01 -1.22700131e+00
-9.48251665e-01 -2.30058149e-01 1.58715814e-01 6.35181189e-01
-6.54936731e-01 3.84534299e-01 -8.69367480e-01 -5.96390903e-01
-8.00363496e-02 -7.91784883e-01 -1.70583874e-01 -6.09989107e-01
-2.52566010e-01 -5.48565388e-01 7.48359084e-01 -5.24752796e-01
1.65479183e+00 -2.19984913e+00 8.13377351e-02 2.00305328e-01
-3.18569802e-02 3.39405864e-01 -1.98801503e-01 6.29024506e-01
8.09708238e-02 1.84102774e-01 7.67048448e-02 -5.53238273e-01
2.86336511e-01 1.01973519e-01 -2.05763683e-01 -1.69066608e-01
1.54918015e-01 7.96848834e-01 -6.78757727e-01 -7.56144047e-01
2.85129696e-01 1.76165566e-01 -4.56544518e-01 4.51518118e-01
-5.44742793e-02 4.01870072e-01 -7.01950550e-01 4.98167723e-01
5.81786990e-01 -1.65997490e-01 4.71827179e-01 -6.22653663e-02
-1.23240627e-01 1.56181380e-01 -1.26228702e+00 1.38957381e+00
-8.76395822e-01 3.15345705e-01 2.48187885e-01 -1.18590796e+00
1.27836621e+00 3.82793993e-01 1.92257822e-01 -6.40880287e-01
2.36671478e-01 2.36735225e-01 -2.21870720e-01 -8.18351686e-01
7.51489997e-01 -7.18583316e-02 -4.46995229e-01 4.57854480e-01
2.22082380e-02 2.98874497e-01 6.75244490e-03 1.20170392e-01
5.85522354e-01 3.44247408e-02 4.93853748e-01 6.25015795e-02
1.06128931e+00 -2.69729137e-01 7.52395272e-01 6.21883273e-01
-4.61544633e-01 4.08833325e-02 5.42581975e-01 -1.40402853e-01
-3.79763842e-01 -2.96177953e-01 8.73033330e-02 1.53600585e+00
2.31226698e-01 -6.60363793e-01 -7.20946372e-01 -1.01116478e+00
-4.83033031e-01 6.72817230e-01 -4.94138926e-01 -6.32852167e-02
-5.11201143e-01 -7.88987756e-01 3.86876017e-01 5.17592251e-01
8.55628371e-01 -1.31696939e+00 -3.36927861e-01 3.24071735e-01
-6.64714754e-01 -1.31517065e+00 -5.12033284e-01 6.96482211e-02
-4.88069862e-01 -7.70152569e-01 -3.39365482e-01 -8.00729454e-01
3.05338532e-01 3.80048901e-01 9.94481087e-01 1.39526397e-01
5.47819957e-02 4.31960315e-01 -9.48932290e-01 -3.39737833e-01
-1.75090432e-01 2.06336826e-01 -1.67909920e-01 4.61637706e-01
5.82705200e-01 -2.66208917e-01 -4.98616129e-01 4.95902240e-01
-6.01297200e-01 3.22621733e-01 7.83518314e-01 1.04272008e+00
2.65790939e-01 2.29643032e-01 8.05542588e-01 -1.08823740e+00
1.16511154e+00 -4.26371962e-01 9.41654854e-03 6.73904359e-01
-6.53312147e-01 -1.74312219e-01 9.24467385e-01 -5.25008619e-01
-1.75907421e+00 2.95956875e-03 -4.04202878e-01 -4.23402973e-02
-3.94899666e-01 6.67781115e-01 -4.96152818e-01 1.42588124e-01
1.05674706e-01 6.00224078e-01 -3.86070311e-01 -5.94225287e-01
2.05515683e-01 1.24831247e+00 2.82449752e-01 -9.54669476e-01
4.00731593e-01 6.90860162e-03 -5.47592103e-01 -7.37521827e-01
-8.26039016e-01 -6.52938008e-01 -6.43236220e-01 -3.86231482e-01
6.34503841e-01 -8.44700873e-01 -1.08561230e+00 4.83700573e-01
-1.10821629e+00 -1.19601816e-01 2.54757494e-01 5.06223023e-01
-4.29229885e-01 4.67877537e-01 -7.70957053e-01 -1.24283683e+00
-4.82711643e-01 -1.03865612e+00 1.18618834e+00 4.44146186e-01
3.39055993e-02 -1.13989985e+00 -2.32486874e-01 3.97922486e-01
4.00662571e-01 -2.53732409e-02 9.76107717e-01 -1.06598759e+00
5.34584112e-02 -6.86583370e-02 -1.99274853e-01 3.39562029e-01
-8.20995718e-02 -1.22137420e-01 -1.06273508e+00 -8.15119594e-02
1.04013130e-01 -6.84968412e-01 5.92163086e-01 -2.65568823e-01
1.41055715e+00 -2.61319757e-01 -1.54540718e-01 3.60338718e-01
1.04475355e+00 5.49258530e-01 4.35278654e-01 4.48122174e-01
2.87471384e-01 9.98278379e-01 1.01082432e+00 7.20284104e-01
6.25632584e-01 7.18869805e-01 -9.38051343e-02 -1.48865521e-01
4.76415902e-01 -2.63187557e-01 3.63805473e-01 1.26842535e+00
5.70400916e-02 -2.88657337e-01 -5.71224153e-01 1.50551423e-01
-2.16026950e+00 -8.13207209e-01 2.74203718e-01 1.57034028e+00
8.09156060e-01 7.74463639e-02 4.04552221e-02 -8.78190473e-02
6.31560326e-01 2.06598967e-01 -5.80313981e-01 -6.90603912e-01
-8.71572569e-02 -1.86129898e-01 -3.06217581e-01 3.37445825e-01
-1.04159093e+00 1.09251916e+00 5.89563513e+00 1.04222250e+00
-1.16539752e+00 -4.62904014e-02 7.68597007e-01 1.77234501e-01
-2.04307079e-01 -1.81711704e-01 -9.70950127e-01 5.45113742e-01
7.82701254e-01 -1.90431923e-01 5.59458137e-01 9.98647988e-01
4.25545990e-01 1.02786422e-01 -8.27383339e-01 1.21477985e+00
3.88305813e-01 -9.06229854e-01 4.40915339e-02 -3.49464387e-01
2.22843140e-01 -5.77167690e-01 -3.13141197e-01 1.09103227e+00
1.52446285e-01 -7.80409873e-01 3.90130669e-01 5.94108045e-01
4.87463176e-01 -8.23541641e-01 9.71664369e-01 6.49919152e-01
-1.22325444e+00 -1.32364959e-01 -2.43565574e-01 9.95529890e-02
-5.09933978e-02 8.94833207e-02 -5.95936954e-01 7.81712532e-01
8.34692538e-01 6.34949803e-01 -5.46227753e-01 5.12333393e-01
-3.64946835e-02 5.74156761e-01 -1.38791233e-01 -5.40022135e-01
3.81057411e-01 -4.10673678e-01 9.45013762e-02 1.62742293e+00
4.48729806e-02 4.45167392e-01 5.81823111e-01 3.69206131e-01
6.52994774e-03 7.81968236e-01 -8.89585465e-02 3.71363349e-02
6.57128751e-01 1.63494277e+00 -4.89062488e-01 -3.80775660e-01
-5.65057933e-01 1.12611198e+00 5.66806614e-01 4.60634708e-01
-8.13537061e-01 -7.82783270e-01 1.40507564e-01 -7.37631142e-01
2.25478828e-01 1.21473782e-01 -6.02444559e-02 -1.39243114e+00
1.54007569e-01 -1.20518363e+00 4.18441981e-01 -5.80369413e-01
-1.17348182e+00 9.60140944e-01 -1.69523150e-01 -1.16355586e+00
-2.50349551e-01 -7.40571618e-01 -7.88194358e-01 7.38109410e-01
-1.65592992e+00 -1.23814106e+00 -4.78161216e-01 8.02489102e-01
1.03431964e+00 -2.80567437e-01 9.92472470e-01 5.93268499e-02
-9.78848040e-01 7.65910625e-01 3.29689026e-01 4.58505265e-02
9.35309470e-01 -1.20377362e+00 -2.21702576e-01 5.68520725e-01
-2.50685334e-01 6.75496280e-01 2.85903484e-01 -2.80855775e-01
-1.29367125e+00 -8.97717476e-01 9.20253336e-01 5.67321107e-02
7.10647941e-01 -7.21886039e-01 -8.84395659e-01 4.58962768e-01
2.40207225e-01 -4.42223638e-01 1.01341724e+00 5.99733233e-01
-2.04880878e-01 -1.54335469e-01 -1.01802647e+00 6.75232232e-01
8.66966307e-01 -4.91020113e-01 -6.56007230e-01 1.80631831e-01
6.60974741e-01 -4.54253644e-01 -9.49007988e-01 4.04851168e-01
6.62719667e-01 -8.02404761e-01 4.49612707e-01 -5.94965100e-01
4.09038126e-01 9.73795727e-02 -1.84047699e-01 -1.40907931e+00
-1.84898958e-01 -8.37387025e-01 -4.11696024e-02 1.73735619e+00
3.72868598e-01 -6.78855717e-01 4.73741531e-01 4.82598394e-01
-2.33113810e-01 -1.07865107e+00 -4.71171677e-01 -3.57794166e-01
-7.41478950e-02 -3.80208492e-01 8.19424748e-01 8.77820611e-01
3.22101980e-01 7.26553559e-01 -6.75893664e-01 -1.49634182e-01
-6.74826503e-02 4.41464096e-01 9.25295830e-01 -1.05586910e+00
-3.13013524e-01 -4.29218143e-01 3.80479455e-01 -1.53846622e+00
4.21048373e-01 -4.22918320e-01 3.64178181e-01 -1.20008147e+00
2.32677132e-01 -4.89670843e-01 -4.21101809e-01 1.93633690e-01
-5.61077118e-01 -2.77976632e-01 2.05083519e-01 2.34175891e-01
-1.04769468e+00 8.70685279e-01 1.17966306e+00 5.81258908e-02
-3.35094362e-01 1.13559581e-01 -9.21176493e-01 8.69015992e-01
8.34236681e-01 -1.41220831e-03 -4.36594337e-01 -6.70738593e-02
1.69976766e-03 2.24671155e-01 -6.77138790e-02 -5.20013273e-01
3.15055549e-01 -4.56083894e-01 4.17569458e-01 -7.08740771e-01
4.23589557e-01 -8.86625290e-01 -4.56390530e-01 -3.73830721e-02
-6.26371741e-01 -1.42834902e-01 2.21014470e-01 6.25911415e-01
-4.28029716e-01 -3.69925886e-01 5.23693383e-01 -7.68615156e-02
-1.08029437e+00 1.26566440e-02 -5.29959798e-01 3.32299098e-02
9.27567840e-01 -5.10269254e-02 -3.48151773e-01 -4.61812615e-01
-5.51968455e-01 5.71605921e-01 -7.27128536e-02 6.03652477e-01
5.88686645e-01 -1.23432767e+00 -5.63354433e-01 2.33220041e-01
4.02585030e-01 -3.80355150e-01 5.56499243e-01 8.31063211e-01
-3.37770805e-02 5.92639565e-01 1.33437961e-01 -3.21951509e-01
-1.47578979e+00 4.45913196e-01 2.05211863e-01 -7.14444697e-01
-3.94133449e-01 4.79443043e-01 2.45521620e-01 -8.12325418e-01
5.40848434e-01 -3.83077413e-02 -7.22207546e-01 1.03900783e-01
5.16172469e-01 1.96449235e-02 -1.00124478e-01 -5.94796121e-01
-3.03764731e-01 5.84924042e-01 -4.36962903e-01 6.20426610e-02
1.10435462e+00 -6.00630164e-01 -6.11497834e-02 6.41818643e-01
1.02750659e+00 1.39818683e-01 -9.79460597e-01 -6.45920932e-01
2.12164372e-01 -2.03724921e-01 -1.46594837e-01 -1.11303329e+00
-6.99546993e-01 9.11305070e-01 3.67538273e-01 2.59782195e-01
1.51482797e+00 -2.97369987e-01 7.19903111e-01 8.02411437e-01
1.68093428e-01 -1.59860814e+00 2.53551483e-01 9.41865563e-01
9.30533171e-01 -1.32988012e+00 -2.62886763e-01 -5.13838887e-01
-1.25255537e+00 1.26405108e+00 9.62060630e-01 2.64428377e-01
6.89097047e-01 1.60982966e-01 1.48839995e-01 1.99648854e-03
-1.19440377e+00 -7.60478526e-02 2.52830178e-01 2.64347672e-01
6.14784479e-01 8.00849572e-02 -5.97386360e-01 1.18799722e+00
-1.25246838e-01 -8.94778967e-02 1.31776139e-01 7.04522908e-01
-4.96064931e-01 -1.01805186e+00 -9.67702046e-02 2.82615155e-01
-3.28894049e-01 -8.42490122e-02 -6.94509327e-01 8.18172991e-01
-8.66879001e-02 1.38284397e+00 -2.37776831e-01 -6.93713605e-01
5.32205522e-01 3.09379607e-01 -2.10560232e-01 -3.94993007e-01
-7.17089176e-01 3.14042449e-01 4.96850997e-01 -4.12401915e-01
-5.64591944e-01 -3.91430616e-01 -1.03561580e+00 -8.45583528e-02
-6.62869990e-01 7.15420842e-01 5.98844349e-01 1.20163536e+00
4.64665979e-01 5.16480148e-01 1.25682807e+00 -6.03746593e-01
-5.42748153e-01 -1.34776247e+00 -5.19485354e-01 4.45363969e-01
-1.05507031e-01 -6.28596008e-01 -3.33665639e-01 -4.29646224e-02] | [13.024367332458496, 6.098376750946045] |
e73173b8-28f2-4400-9742-8e37955b5767 | is-synthetic-dataset-reliable-for | 2209.05047 | null | https://arxiv.org/abs/2209.05047v1 | https://arxiv.org/pdf/2209.05047v1.pdf | Is Synthetic Dataset Reliable for Benchmarking Generalizable Person Re-Identification? | Recent studies show that models trained on synthetic datasets are able to achieve better generalizable person re-identification (GPReID) performance than that trained on public real-world datasets. On the other hand, due to the limitations of real-world person ReID datasets, it would also be important and interesting to use large-scale synthetic datasets as test sets to benchmark person ReID algorithms. Yet this raises a critical question: is synthetic dataset reliable for benchmarking generalizable person re-identification? In the literature there is no evidence showing this. To address this, we design a method called Pairwise Ranking Analysis (PRA) to quantitatively measure the ranking similarity and perform the statistical test of identical distributions. Specifically, we employ Kendall rank correlation coefficients to evaluate pairwise similarity values between algorithm rankings on different datasets. Then, a non-parametric two-sample Kolmogorov-Smirnov (KS) test is performed for the judgement of whether algorithm ranking correlations between synthetic and real-world datasets and those only between real-world datasets lie in identical distributions. We conduct comprehensive experiments, with ten representative algorithms, three popular real-world person ReID datasets, and three recently released large-scale synthetic datasets. Through the designed pairwise ranking analysis and comprehensive evaluations, we conclude that a recent large-scale synthetic dataset ClonedPerson can be reliably used to benchmark GPReID, statistically the same as real-world datasets. Therefore, this study guarantees the usage of synthetic datasets for both source training set and target testing set, with completely no privacy concerns from real-world surveillance data. Besides, the study in this paper might also inspire future designs of synthetic datasets. | ['Cuicui Kang'] | 2022-09-12 | null | null | null | null | ['generalizable-person-re-identification'] | ['computer-vision'] | [-1.37157604e-01 -4.63614702e-01 1.61786601e-02 -6.33911133e-01
-4.10113275e-01 -8.49152565e-01 7.65049398e-01 2.88005620e-01
-7.42466092e-01 1.09518230e+00 -6.71652481e-02 -7.63212815e-02
-2.71637440e-01 -9.76061881e-01 -6.96202397e-01 -4.26486045e-01
-1.57914251e-01 5.76766968e-01 7.13285850e-03 -1.11374818e-01
2.56880164e-01 5.26347876e-01 -1.82595265e+00 -2.14692336e-02
1.02191222e+00 4.46140498e-01 -4.33097333e-01 3.46642584e-01
2.97872484e-01 1.51040778e-01 -1.00934207e+00 -1.05356288e+00
7.66172051e-01 -2.43974179e-01 -4.67529774e-01 -5.40783465e-01
6.16379559e-01 -4.85744178e-01 -3.21455449e-01 1.12289274e+00
8.27457368e-01 4.60321084e-02 5.97376704e-01 -1.81133282e+00
-7.68065095e-01 4.36673313e-01 -4.43814784e-01 -3.13242519e-04
7.77992189e-01 2.67681748e-01 5.74305058e-01 -4.01909679e-01
5.85940957e-01 1.11312342e+00 9.08499479e-01 5.95068216e-01
-1.20315754e+00 -1.32501292e+00 -2.99335331e-01 1.42745942e-01
-1.70128548e+00 -2.16191113e-01 4.73356754e-01 -4.17729765e-01
3.61120164e-01 7.70550013e-01 4.09217119e-01 1.53668916e+00
-2.75935363e-02 5.49809933e-01 1.40605223e+00 -9.47051048e-02
1.43692017e-01 5.14081180e-01 4.55317527e-01 2.38468006e-01
1.05422235e+00 4.75504220e-01 -3.99902731e-01 -5.08414745e-01
2.54935473e-01 5.44528365e-02 -3.77612293e-01 -4.12094772e-01
-1.41240084e+00 5.08730829e-01 1.83640748e-01 5.51725514e-02
-9.85583663e-02 -3.53901863e-01 7.20289528e-01 5.08546591e-01
1.68766156e-01 5.00301600e-01 -3.13303202e-01 -1.01893283e-02
-8.10659111e-01 7.58313298e-01 8.07155788e-01 1.10514355e+00
5.66111267e-01 -4.95517284e-01 -2.25781962e-01 6.67714894e-01
-8.15761974e-04 7.96114326e-01 7.01930702e-01 -4.58027601e-01
4.58026916e-01 6.77410781e-01 3.83607209e-01 -1.23659956e+00
-2.63174057e-01 -2.48982817e-01 -9.86488938e-01 -1.25808224e-01
7.79660821e-01 8.91100150e-03 -3.02403718e-01 1.93692279e+00
2.58302331e-01 2.51126051e-01 1.37313008e-02 7.94316411e-01
9.51501369e-01 1.34411767e-01 1.03336193e-01 4.95588779e-02
1.41692030e+00 -3.45227212e-01 -2.84148157e-01 3.86692345e-01
4.97945219e-01 -4.96699035e-01 1.02372420e+00 3.81646752e-02
-6.35924399e-01 -6.08462989e-01 -9.85308111e-01 2.31889874e-01
-8.78446877e-01 1.05576567e-01 5.51901221e-01 1.14448702e+00
-8.94355476e-01 3.38062316e-01 -3.86005670e-01 -8.91884446e-01
1.39207765e-01 5.21679282e-01 -8.68841648e-01 3.34358700e-02
-1.37622130e+00 7.24303484e-01 5.25864005e-01 -5.08343168e-02
-5.93380094e-01 -8.53338778e-01 -6.97034001e-01 -2.73597121e-01
-1.41130826e-02 -8.65672469e-01 8.18128526e-01 -6.62750185e-01
-9.40699637e-01 1.31027818e+00 -1.06785353e-02 -4.88705069e-01
1.03561449e+00 -4.91910577e-02 -6.63597882e-01 -4.23138469e-01
3.09792489e-01 4.15252924e-01 3.08963388e-01 -1.30722106e+00
-3.93934458e-01 -5.63564062e-01 -8.30546170e-02 -1.96931764e-01
-3.78347576e-01 1.81955740e-01 -2.68014669e-01 -7.49605775e-01
-3.52126718e-01 -1.03202462e+00 2.10695654e-01 -3.83687139e-01
-6.46602631e-01 -2.50089049e-01 3.92776757e-01 -5.71971774e-01
1.00518978e+00 -1.99179137e+00 -4.73419368e-01 4.86645013e-01
4.60013859e-02 4.41425532e-01 -3.54454428e-01 4.50724959e-01
-1.97472438e-01 2.08016470e-01 -7.37351850e-02 -2.16057837e-01
3.79420333e-02 -2.65437663e-01 -3.59839350e-01 7.53429830e-01
-2.15165958e-01 9.48308110e-01 -1.00419033e+00 -4.40284133e-01
1.66633740e-01 7.32138827e-02 -1.93526044e-01 3.53895426e-01
3.97354871e-01 4.00495738e-01 -4.59018350e-01 5.30921638e-01
1.01811063e+00 2.91000694e-01 7.16627166e-02 -1.90490246e-01
7.67908767e-02 -2.06883207e-01 -1.16792953e+00 1.20437038e+00
6.63626790e-02 5.41332066e-01 -4.66203988e-01 -5.64076602e-01
1.03324747e+00 -1.32474065e-01 3.58925790e-01 -8.57022643e-01
1.07910976e-01 1.35989174e-01 -1.60624415e-01 -2.17874125e-01
6.94215178e-01 2.70667523e-01 -3.02840412e-01 5.45585334e-01
-3.56183916e-01 3.95368248e-01 3.40013146e-01 2.06790239e-01
1.04034579e+00 -2.72863567e-01 1.39988348e-01 -4.38955396e-01
6.67217910e-01 -7.77950734e-02 5.19178867e-01 1.00287426e+00
-4.10043269e-01 6.14758849e-01 3.20947528e-01 -4.77188587e-01
-1.03827894e+00 -1.44570601e+00 -3.44130754e-01 6.96744204e-01
5.08659124e-01 -3.25278133e-01 -9.54974234e-01 -8.25400114e-01
4.90161628e-01 5.66740453e-01 -8.99246395e-01 -3.29584442e-02
-3.91954094e-01 -1.05197108e+00 1.12707436e+00 2.99292594e-01
8.07784200e-01 -7.97353148e-01 -2.58083373e-01 -2.28850305e-01
-1.17797010e-01 -9.66090500e-01 -5.96771419e-01 -7.29238689e-01
-4.06115860e-01 -1.57792342e+00 -9.09144223e-01 -5.42440474e-01
7.42479384e-01 4.64157581e-01 1.05556202e+00 2.32917756e-01
-3.05740207e-01 2.73138911e-01 -3.36975574e-01 -3.30775291e-01
-3.15620184e-01 5.17859794e-02 5.98520637e-01 -5.23940027e-02
8.04115534e-01 -3.57285678e-01 -6.87813580e-01 1.06940114e+00
-7.40068376e-01 -1.00314811e-01 3.93949717e-01 6.49957299e-01
2.00520471e-01 8.66946578e-02 7.98527002e-01 -8.54633152e-01
9.06915605e-01 -4.61056381e-01 -7.30940759e-01 7.03071058e-01
-7.28862047e-01 -1.43397868e-01 6.99040115e-01 -4.29640323e-01
-8.28282833e-01 -5.23237884e-01 3.83383512e-01 4.14037183e-02
-1.72061831e-01 1.00938506e-01 -2.42184460e-01 7.18317926e-02
8.51419389e-01 3.39872867e-01 -5.27553335e-02 -2.38665134e-01
1.26482114e-01 9.77759123e-01 8.34159613e-01 -7.68426239e-01
1.26157475e+00 4.09714252e-01 -1.89360231e-01 -6.40520692e-01
-2.94248015e-01 -5.40730119e-01 -4.75477904e-01 -5.75530566e-02
7.26141751e-01 -1.01534152e+00 -1.20354283e+00 8.49714756e-01
-1.03306496e+00 -1.02401944e-02 4.92402054e-02 4.96853411e-01
-1.88250259e-01 6.37707233e-01 -4.37146693e-01 -7.34177291e-01
-5.19029558e-01 -8.48722100e-01 8.08561802e-01 3.37270200e-01
-5.00017524e-01 -6.61564708e-01 2.43378371e-01 5.60446799e-01
3.81768107e-01 5.24409354e-01 4.88999784e-01 -1.01087034e+00
-3.72931004e-01 -7.10848451e-01 -6.02141321e-01 2.61802766e-02
2.56609768e-01 5.40721640e-02 -9.82508838e-01 -6.95030987e-01
-4.54389244e-01 -1.15681432e-01 4.60570782e-01 4.46242141e-03
1.24159253e+00 -1.94716558e-01 -5.02300918e-01 5.05476058e-01
1.20365989e+00 -1.72873214e-02 7.04146683e-01 6.19328320e-01
4.85797822e-01 8.41110528e-01 7.59713113e-01 3.61754656e-01
5.94149351e-01 7.64463127e-01 -1.70088544e-01 9.81108993e-02
3.27694237e-01 -7.25920439e-01 2.17357382e-01 2.10124806e-01
-1.25954524e-01 -4.76514935e-01 -9.68754113e-01 3.17078263e-01
-1.78547084e+00 -1.17568374e+00 -1.47717521e-01 2.91153359e+00
5.51297307e-01 -9.11235660e-02 5.04320800e-01 2.14938950e-02
1.02669322e+00 -2.15278104e-01 -4.67653394e-01 -4.81282584e-02
-3.05060863e-01 -3.61007929e-01 7.55030632e-01 -7.43672550e-02
-9.84333813e-01 4.96697068e-01 6.09190416e+00 6.13469124e-01
-8.22394252e-01 -1.29099190e-01 5.66466510e-01 5.64618371e-02
-2.19788954e-01 -2.57291585e-01 -6.87107146e-01 9.09478664e-01
1.08403993e+00 -7.39694297e-01 2.87044138e-01 6.71261370e-01
9.87286940e-02 5.03248908e-02 -1.34585500e+00 1.31302464e+00
8.74316469e-02 -9.27595854e-01 2.61115283e-01 1.75355941e-01
6.73961639e-01 -1.39551491e-01 1.07058965e-01 3.88777494e-01
3.41031164e-01 -1.11862552e+00 4.30761606e-01 5.13360262e-01
8.92752886e-01 -7.86500871e-01 1.19535470e+00 3.55071425e-01
-1.10436380e+00 1.88067421e-01 -4.68699127e-01 2.56499439e-01
-2.45343432e-01 3.57196927e-01 -8.73771429e-01 8.37216079e-01
8.34301531e-01 4.50715423e-01 -1.02556884e+00 1.30745971e+00
8.98488835e-02 1.65953219e-01 -2.58611143e-01 -5.23289405e-02
-4.71591979e-01 -1.48237780e-01 3.02860677e-01 1.12316310e+00
2.43409961e-01 1.53108025e-02 -1.89029291e-01 7.90836096e-01
-1.65377572e-01 2.34448299e-01 -7.33496010e-01 4.07784060e-02
8.11347663e-01 9.16527510e-01 -4.12185073e-01 -1.84848592e-01
-2.56848961e-01 1.03665781e+00 1.14752509e-01 4.49523687e-01
-9.95035350e-01 -2.33228713e-01 8.13053787e-01 1.92903265e-01
-5.81915438e-01 -4.94119674e-02 -2.23129556e-01 -1.31237459e+00
1.26496747e-01 -1.00860989e+00 6.48423076e-01 -3.97044539e-01
-1.86481404e+00 4.63483751e-01 3.26503158e-01 -1.49968326e+00
-1.54772237e-01 -3.90124232e-01 -6.48157775e-01 1.01992369e+00
-1.18743467e+00 -1.05476332e+00 -6.87812269e-01 6.75065398e-01
-9.56899896e-02 -5.48650980e-01 8.62579346e-01 3.90828431e-01
-7.63446748e-01 1.43953729e+00 3.81493986e-01 5.96308410e-01
1.13599968e+00 -1.07755363e+00 7.26874471e-01 7.58764148e-01
-2.14713320e-01 1.27494276e+00 8.26947629e-01 -7.89028227e-01
-1.36912322e+00 -1.07192993e+00 6.25830173e-01 -9.60911393e-01
2.58856833e-01 -4.49137717e-01 -7.54015923e-01 5.98323166e-01
-2.93967932e-01 -1.18063159e-01 9.81299281e-01 1.77055895e-01
-6.37177765e-01 -2.96378016e-01 -1.74403989e+00 4.84411240e-01
1.20604682e+00 -5.39117157e-01 -5.73214233e-01 1.06693529e-01
3.61869693e-01 -7.86853731e-02 -9.56359208e-01 4.79923189e-01
9.03676689e-01 -1.12552810e+00 1.19628060e+00 -6.41297519e-01
6.08509928e-02 -5.19609451e-01 -6.64634183e-02 -1.15998602e+00
1.78205311e-01 -3.34188282e-01 5.03603160e-01 1.59501588e+00
2.67647803e-01 -1.13172793e+00 9.04159844e-01 1.14024925e+00
5.23286164e-01 -2.56515563e-01 -7.29801953e-01 -1.17555797e+00
8.41150284e-02 -1.55024752e-01 1.13686514e+00 1.13282549e+00
-2.13173375e-01 -3.33455712e-01 -3.69496435e-01 3.64662677e-01
1.21588027e+00 9.50462278e-03 1.62917924e+00 -1.36119187e+00
1.79323442e-02 -2.80658334e-01 -9.28487599e-01 -3.75348181e-01
4.05581623e-01 -8.79519224e-01 -3.34980190e-01 -1.10124588e+00
7.09835052e-01 -6.72708809e-01 -3.91057372e-01 2.02176899e-01
-2.38068044e-01 3.08545440e-01 2.17488967e-02 3.35671276e-01
-4.19474602e-01 5.17045856e-01 8.28972936e-01 -3.06346953e-01
-1.17822401e-01 9.17590931e-02 -6.54882789e-01 1.65151998e-01
5.81837058e-01 -3.42609227e-01 -4.16103154e-01 4.45857346e-02
-3.67702767e-02 -3.72825116e-01 6.45112276e-01 -1.04313314e+00
3.76437902e-01 -1.16573431e-01 5.35667181e-01 -3.98676038e-01
-1.32673785e-01 -6.19551480e-01 5.60009539e-01 3.72951597e-01
-1.80034488e-01 4.01533723e-01 2.08767563e-01 5.40251076e-01
1.55207217e-02 2.75366813e-01 6.01570368e-01 1.19256333e-01
-6.25125349e-01 4.66339290e-01 2.76015520e-01 1.38287898e-02
1.32290018e+00 -5.96306384e-01 -7.14776993e-01 -1.78102359e-01
1.32684454e-01 4.93456990e-01 1.10231864e+00 6.46428466e-01
3.86457473e-01 -1.22817183e+00 -1.00940621e+00 3.77024174e-01
6.34889185e-01 -5.44068217e-01 2.85969794e-01 3.04946333e-01
-4.89215553e-01 3.67973059e-01 -4.60902691e-01 -4.26388800e-01
-1.46095645e+00 7.57831156e-01 3.48219693e-01 3.01019959e-02
-2.48123243e-01 5.61807454e-01 1.22804463e-01 -9.38610077e-01
6.12979531e-02 1.25858173e-01 1.05476342e-01 -1.89785082e-02
8.31389248e-01 6.27785921e-01 -1.16247423e-01 -8.48976016e-01
-6.65775836e-01 4.04352486e-01 -2.46318370e-01 4.32609580e-02
9.80082035e-01 6.86804280e-02 -2.20363781e-01 -3.51451375e-02
1.22080362e+00 1.90447032e-01 -6.76596820e-01 4.64608409e-02
2.79824194e-02 -8.60391259e-01 -8.55994642e-01 -7.96874344e-01
-6.17098629e-01 1.92837000e-01 9.70413387e-01 2.55887538e-01
8.66913378e-01 -3.99078220e-01 6.54370129e-01 3.54736775e-01
9.91475880e-01 -8.27730894e-01 -3.79620224e-01 2.88810581e-02
7.48667657e-01 -1.47376215e+00 4.83419634e-02 -3.30619723e-01
-4.47122246e-01 6.36767030e-01 7.31661618e-01 1.28553331e-01
4.68668133e-01 -1.68130368e-01 1.23257734e-01 9.59402099e-02
-1.05664797e-01 1.74983412e-01 3.15818876e-01 7.37117648e-01
4.01662618e-01 4.27012533e-01 -5.81010520e-01 5.84946930e-01
-8.15084457e-01 6.72646761e-02 3.63865495e-01 4.77350712e-01
2.23911732e-01 -1.46047425e+00 -4.95281577e-01 6.09027982e-01
-1.56839564e-01 3.05593044e-01 -5.88441432e-01 1.01426375e+00
4.99910712e-02 1.08641946e+00 -2.39936095e-02 -7.41404355e-01
6.51600122e-01 -1.92102566e-01 2.36769274e-01 -2.11485907e-01
-7.93231249e-01 -1.22103822e+00 2.26324704e-02 -5.12832284e-01
-3.04149777e-01 -7.17578232e-01 -8.04830551e-01 -1.08532894e+00
-9.19518545e-02 3.64909619e-01 4.45924759e-01 6.14575088e-01
3.45351845e-01 -2.80913860e-01 6.45219505e-01 -3.92349154e-01
-6.83424413e-01 -8.44484210e-01 -5.58930576e-01 1.07254577e+00
-1.34603875e-02 -6.03498876e-01 -3.64494264e-01 -4.51102138e-01] | [14.705273628234863, 1.0530718564987183] |
4cf7269a-dd67-4a9e-b6a0-4064e3a25b80 | a-context-aware-loss-function-for-action | 1912.01326 | null | https://arxiv.org/abs/1912.01326v3 | https://arxiv.org/pdf/1912.01326v3.pdf | A Context-Aware Loss Function for Action Spotting in Soccer Videos | In video understanding, action spotting consists in temporally localizing human-induced events annotated with single timestamps. In this paper, we propose a novel loss function that specifically considers the temporal context naturally present around each action, rather than focusing on the single annotated frame to spot. We benchmark our loss on a large dataset of soccer videos, SoccerNet, and achieve an improvement of 12.8% over the baseline. We show the generalization capability of our loss for generic activity proposals and detection on ActivityNet, by spotting the beginning and the end of each activity. Furthermore, we provide an extended ablation study and display challenging cases for action spotting in soccer videos. Finally, we qualitatively illustrate how our loss induces a precise temporal understanding of actions and show how such semantic knowledge can be used for automatic highlights generation. | ['Thomas B. Moeslund', 'Adrien Deliège', 'Marc Van Droogenbroeck', 'Anthony Cioppa', 'Rikke Gade', 'Silvio Giancola', 'Bernard Ghanem'] | 2019-12-03 | a-context-aware-loss-function-for-action-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Cioppa_A_Context-Aware_Loss_Function_for_Action_Spotting_in_Soccer_Videos_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Cioppa_A_Context-Aware_Loss_Function_for_Action_Spotting_in_Soccer_Videos_CVPR_2020_paper.pdf | cvpr-2020-6 | ['action-spotting'] | ['computer-vision'] | [ 5.61741352e-01 7.33100697e-02 -3.23599249e-01 -2.35569000e-01
-8.09914947e-01 -6.39542639e-01 6.56809211e-01 -1.39312129e-02
-5.02252460e-01 7.02614963e-01 6.48066461e-01 2.42972031e-01
1.05707578e-01 -2.68081754e-01 -7.75900483e-01 -6.01461768e-01
-4.97323692e-01 5.88847511e-02 6.93859637e-01 -3.03700520e-03
1.57164201e-01 4.64338779e-01 -1.44108689e+00 7.77511299e-01
2.55433261e-01 8.81484866e-01 1.82496846e-01 7.37506628e-01
4.42458838e-01 1.19674444e+00 -9.76279795e-01 -1.41246721e-01
3.39533240e-01 -6.63895726e-01 -6.68909192e-01 5.11592031e-01
7.17053473e-01 -3.41894895e-01 -6.70167148e-01 5.18111646e-01
2.37913296e-01 5.27944028e-01 2.28566766e-01 -1.45761907e+00
1.39143229e-01 5.60704350e-01 -6.59829736e-01 5.69481730e-01
7.52916634e-01 3.08599651e-01 1.10435605e+00 -7.66705751e-01
1.07023525e+00 1.09943140e+00 6.22815251e-01 4.72238749e-01
-1.01256704e+00 -5.87495744e-01 5.50986826e-01 5.94915450e-01
-1.16338289e+00 -4.90446687e-01 6.96549237e-01 -4.28637445e-01
8.38846743e-01 5.60644008e-02 9.54245567e-01 1.36066437e+00
2.36055721e-02 1.37867856e+00 5.69348931e-01 -1.50805205e-01
1.06993988e-01 -4.95549381e-01 -1.48955330e-01 5.43105185e-01
-2.20297858e-01 9.09390524e-02 -9.67877865e-01 1.10969394e-01
9.20741439e-01 -4.04660404e-02 -3.95143390e-01 -4.90931809e-01
-1.64154077e+00 5.71484208e-01 1.77233905e-01 3.93399596e-02
-4.22949880e-01 6.83305860e-01 7.65842140e-01 -1.98493525e-02
3.78074020e-01 3.91633451e-01 -4.00882661e-01 -6.43114209e-01
-1.14238930e+00 4.36145037e-01 5.13909161e-01 9.19485509e-01
5.18788755e-01 -1.17459066e-01 -7.01318681e-01 4.13560599e-01
-3.13777506e-01 8.72557089e-02 -9.12158471e-03 -1.58320844e+00
4.43151861e-01 3.58908325e-01 4.87668782e-01 -9.14969265e-01
-3.23530316e-01 -2.32789963e-01 -1.16896600e-01 -3.49969119e-02
6.68647707e-01 -2.06426337e-01 -8.01872909e-01 1.78392458e+00
1.90161467e-01 8.97981107e-01 -1.24160893e-01 1.06739557e+00
3.45486462e-01 5.99614501e-01 2.35812411e-01 -1.81353569e-01
1.36666870e+00 -1.33026183e+00 -7.43630707e-01 -3.55246574e-01
5.83758593e-01 -5.00376284e-01 1.05948567e+00 4.84317213e-01
-1.00535619e+00 -5.36814153e-01 -9.19697821e-01 -6.43054396e-02
-8.84186476e-03 3.86240512e-01 7.40029454e-01 1.34982437e-01
-8.54381800e-01 6.11674011e-01 -1.10461688e+00 -4.65063125e-01
2.83915371e-01 -2.54009906e-02 -3.83904010e-01 5.90114407e-02
-9.99103725e-01 6.28408194e-01 6.16245270e-01 -1.98911086e-01
-1.35953915e+00 -8.18986177e-01 -9.07415330e-01 -7.47643188e-02
8.74617875e-01 -5.08715630e-01 1.33971560e+00 -1.20531034e+00
-1.16505933e+00 8.66539121e-01 -3.67251426e-01 -9.77013290e-01
7.72361517e-01 -6.39126003e-01 -3.39298278e-01 6.48923993e-01
2.93736666e-01 9.85659301e-01 5.44652283e-01 -9.60269928e-01
-1.11553657e+00 6.87399507e-02 5.13526797e-01 2.44730696e-01
-4.87525612e-02 2.70047247e-01 -9.01079178e-01 -9.57033277e-01
-1.19792059e-01 -8.82718205e-01 -1.95791811e-01 1.68000698e-01
-2.70328850e-01 -1.75550193e-01 1.04931116e+00 -6.72598302e-01
1.32054234e+00 -2.13252974e+00 7.08451271e-02 -1.04768306e-01
-1.32797388e-02 -9.98756010e-03 -1.49638668e-01 4.63820994e-01
-8.32154304e-02 -2.21464008e-01 -2.76476920e-01 -2.03168035e-01
-8.92221853e-02 2.11337969e-01 -4.23347473e-01 3.80050361e-01
2.11661145e-01 9.24085438e-01 -1.31577301e+00 -4.69972908e-01
3.28958243e-01 4.29530263e-01 -6.18287683e-01 -5.69758974e-02
-3.96195233e-01 6.19783998e-01 -3.29522818e-01 6.05240583e-01
6.95632398e-02 -1.76363602e-01 2.06957400e-01 -3.03720981e-01
-1.91398207e-02 2.89004296e-01 -1.16954887e+00 2.20197582e+00
-3.55976820e-01 1.01909506e+00 -1.85798913e-01 -8.27164352e-01
4.04391110e-01 3.01868290e-01 9.14969742e-01 -5.10859847e-01
-2.65537798e-01 -2.54484326e-01 -2.98130304e-01 -4.87851441e-01
5.29502094e-01 1.58433512e-01 -2.71969497e-01 5.44018686e-01
7.86968395e-02 2.24980474e-01 7.29205966e-01 3.86645049e-01
1.41436315e+00 7.40435302e-01 3.55970055e-01 8.55928361e-02
3.23966533e-01 4.93068807e-02 6.73583031e-01 8.37461650e-01
-4.81546193e-01 7.57414937e-01 7.77724683e-01 -6.54077053e-01
-9.80507314e-01 -9.94225681e-01 2.69977719e-01 1.38152361e+00
3.77187490e-01 -6.95892990e-01 -6.12219274e-01 -1.07730174e+00
-1.87253147e-01 7.70883799e-01 -8.37728143e-01 -2.61586588e-02
-1.11738646e+00 -4.86114323e-01 6.32358313e-01 9.03493643e-01
6.40250266e-01 -1.22953439e+00 -1.01018715e+00 2.88738132e-01
-5.80859125e-01 -1.57412970e+00 -9.05072629e-01 -1.24916606e-01
-7.24765778e-01 -1.31994522e+00 -6.50699854e-01 -4.50406373e-01
2.67615259e-01 2.33376741e-01 1.37337589e+00 -9.76840407e-02
-3.71145993e-01 6.81931019e-01 -5.19188702e-01 -1.01342484e-01
-1.24826990e-01 -1.16360463e-01 -1.57346100e-01 2.01752275e-01
3.18563432e-01 -3.74444366e-01 -9.19906735e-01 5.17505825e-01
-6.75321221e-01 2.61598229e-01 2.57820874e-01 5.23288310e-01
4.39997196e-01 -2.77751446e-01 3.18592101e-01 -5.13452411e-01
9.81356800e-02 -2.14248449e-01 -2.77194858e-01 2.94882447e-01
1.66575953e-01 -1.67304382e-01 1.51664689e-01 -3.55789363e-01
-1.07685935e+00 4.79492188e-01 2.20684767e-01 -5.28920412e-01
-2.56554604e-01 -1.53440246e-02 -1.28358370e-02 2.37534389e-01
6.08150065e-01 2.23789185e-01 -4.41398680e-01 -2.93083370e-01
5.05844355e-01 -8.57975259e-02 7.87325561e-01 -6.68529689e-01
4.21603531e-01 8.27792406e-01 -1.62898675e-01 -6.49923205e-01
-9.87112820e-01 -6.30953848e-01 -6.55108988e-01 -6.05255246e-01
9.94485378e-01 -1.07868469e+00 -8.24071765e-01 2.50658870e-01
-1.25663412e+00 -5.82399189e-01 -4.94055152e-01 4.86714214e-01
-9.14108813e-01 5.17811835e-01 -7.49933898e-01 -6.99686587e-01
7.19668940e-02 -8.51034760e-01 1.41809690e+00 -3.07436958e-02
-4.99445289e-01 -9.42381084e-01 5.63290119e-02 3.26818854e-01
-2.42050827e-01 5.79983652e-01 1.21406712e-01 -4.98459041e-01
-7.17191875e-01 -1.97677821e-01 -2.10482821e-01 1.13335706e-01
1.11307159e-01 -1.64944734e-02 -7.78377235e-01 -2.33398184e-01
-3.88786823e-01 -2.92613953e-01 1.19543481e+00 5.26903093e-01
1.27479982e+00 -1.42221600e-01 -4.50921476e-01 5.27919114e-01
1.10800672e+00 3.28443199e-01 7.41669297e-01 2.54796654e-01
6.08902574e-01 5.67123234e-01 1.20893574e+00 6.53461218e-01
6.45017996e-02 1.07800508e+00 2.96345800e-01 -1.57790825e-01
-2.42150605e-01 -4.98936027e-01 6.01943731e-01 -3.63388248e-02
-5.39359570e-01 -3.75140041e-01 -6.95050895e-01 7.83902645e-01
-2.16473746e+00 -1.32422459e+00 2.95436889e-01 2.18399906e+00
5.35339713e-01 2.16281623e-01 5.25920391e-01 -4.26986702e-02
8.91891122e-01 4.70321029e-01 -4.75735605e-01 1.88481808e-01
-2.18447015e-01 2.05535248e-01 5.84052801e-01 4.47807550e-01
-1.52552199e+00 1.23496068e+00 7.08113384e+00 9.20821846e-01
-7.48745143e-01 1.44673273e-01 4.42984849e-01 -5.98813534e-01
1.99500814e-01 7.90463835e-02 -6.33581042e-01 3.56047750e-01
5.21156073e-01 -1.06734157e-01 1.24207787e-01 6.65563047e-01
8.40674102e-01 -3.78214568e-01 -1.32104766e+00 8.96078706e-01
1.90210491e-01 -1.43214631e+00 1.41227305e-01 -2.43019655e-01
8.45007598e-01 -3.12442809e-01 -4.43550527e-01 9.13624242e-02
2.94409573e-01 -8.43987584e-01 9.43420410e-01 4.21865553e-01
5.22711098e-01 -7.50887036e-01 3.42437595e-01 -5.96121363e-02
-1.49175274e+00 -2.51477122e-01 9.19331536e-02 -2.15512857e-01
7.70069122e-01 2.44579658e-01 -8.20352316e-01 3.16967934e-01
5.71119130e-01 1.31254375e+00 -4.53751057e-01 1.19830287e+00
-5.50066531e-01 5.36602020e-01 -2.38527536e-01 2.88792074e-01
4.69568580e-01 2.82856375e-02 7.18231380e-01 1.48488367e+00
3.08557272e-01 1.98495418e-01 6.21511102e-01 6.29330575e-01
1.66789129e-01 -2.67062962e-01 -5.24006963e-01 -5.76579198e-02
2.33693749e-01 1.09094751e+00 -9.92450297e-01 -6.71910882e-01
-3.06577682e-01 1.27193832e+00 3.08645666e-02 4.89497483e-01
-1.31087935e+00 -2.61134803e-01 8.13584447e-01 2.75470376e-01
5.12133598e-01 -2.67319143e-01 -8.47533531e-03 -1.19727409e+00
1.30142152e-01 -7.06470430e-01 5.93315780e-01 -9.15919065e-01
-7.23998308e-01 1.82738125e-01 3.06079417e-01 -1.34450853e+00
-3.45534921e-01 -4.55609798e-01 -7.68501878e-01 1.00083448e-01
-8.41299534e-01 -9.52349246e-01 -4.35094059e-01 5.74376166e-01
1.06426561e+00 2.81799555e-01 2.70231962e-01 3.77417862e-01
-4.32589233e-01 2.67633229e-01 -5.53738058e-01 3.66647333e-01
7.86739171e-01 -1.19539797e+00 5.42058945e-01 1.03204799e+00
4.82277781e-01 2.74722815e-01 1.01183128e+00 -5.73063314e-01
-6.56442702e-01 -9.90126729e-01 6.91057503e-01 -5.81948936e-01
7.68152833e-01 -3.95539075e-01 -6.39048159e-01 1.00189698e+00
1.71711072e-01 -1.82172850e-01 3.06999385e-01 -1.70914158e-01
-9.75745395e-02 -5.03099859e-02 -7.35153675e-01 7.92579889e-01
1.65857112e+00 -3.69037390e-01 -5.09471774e-01 4.95229065e-01
6.76290751e-01 -4.28221613e-01 -5.34534097e-01 5.24754047e-01
6.29010320e-01 -1.07396555e+00 1.16349161e+00 -7.64174223e-01
4.99633044e-01 -4.88087147e-01 1.33449554e-01 -9.21397328e-01
4.82465662e-02 -7.18709826e-01 -2.02819258e-01 6.91690505e-01
1.66639432e-01 1.25292152e-01 1.05433536e+00 3.62780392e-01
-2.76988834e-01 -5.00340641e-01 -7.73884475e-01 -7.09704876e-01
-5.72923601e-01 -7.15218067e-01 1.29354537e-01 7.68263578e-01
1.68469891e-01 -1.33214727e-01 -7.23379731e-01 -1.12497419e-01
5.55552244e-01 -8.34840816e-03 9.15097654e-01 -6.21812820e-01
-4.73920524e-01 -2.40499526e-01 -6.31892204e-01 -1.50951934e+00
1.21359900e-01 -4.19568419e-01 1.82523608e-01 -1.33531892e+00
3.62738043e-01 1.83016330e-01 -4.58926708e-01 5.77656746e-01
-6.62056729e-02 5.93097448e-01 3.94453168e-01 -5.11112772e-02
-1.22451949e+00 3.16692799e-01 1.33387363e+00 -9.69227105e-02
-2.34698400e-01 -9.27150175e-02 -1.03273012e-01 8.84738147e-01
5.42978644e-01 -3.44911903e-01 -5.06934762e-01 -1.42332613e-01
-1.48069514e-02 2.85273165e-01 8.85582089e-01 -1.31460226e+00
1.99828118e-01 -2.83547461e-01 2.18087986e-01 -6.37968481e-01
5.43217003e-01 -4.07020986e-01 5.56067787e-02 3.57959419e-01
-6.08178139e-01 -1.35523900e-01 2.00899631e-01 8.03825319e-01
-3.52826923e-01 9.57715511e-02 4.75176096e-01 -2.33643830e-01
-1.38279784e+00 2.93336958e-01 -4.21015114e-01 2.98879892e-01
1.39266849e+00 -4.09309447e-01 -7.73044769e-03 -5.72678566e-01
-1.24594414e+00 2.48782843e-01 4.70877290e-01 5.65495670e-01
5.21286130e-01 -1.29909754e+00 -6.20707989e-01 -2.70496547e-01
2.35336080e-01 -3.95580918e-01 3.32227618e-01 1.00347579e+00
-6.25105083e-01 3.30069035e-01 -2.30462372e-01 -7.07232535e-01
-1.32701242e+00 4.61980045e-01 3.40367526e-01 -2.08512679e-01
-9.27657366e-01 7.81920195e-01 5.92403233e-01 3.49399745e-01
4.67773706e-01 -4.63709474e-01 -1.05978027e-01 9.69060138e-02
7.13293850e-01 5.50957918e-01 -3.91109526e-01 -5.30441165e-01
-5.26851237e-01 4.15420175e-01 1.96384907e-01 -3.82460207e-01
9.26209390e-01 -9.52266157e-02 3.01850796e-01 4.49773222e-01
9.23286498e-01 -8.34200680e-02 -2.19293880e+00 -4.92258370e-02
2.31995180e-01 -5.20941257e-01 -3.81956160e-01 -7.40915656e-01
-1.22814941e+00 6.04892790e-01 3.16783696e-01 -2.85767615e-01
1.13278508e+00 1.16093457e-01 8.71820509e-01 3.88618022e-01
5.06489277e-01 -1.22006023e+00 4.86647815e-01 3.25917602e-01
7.98328161e-01 -1.00357497e+00 -4.48993582e-04 -4.71801937e-01
-9.24395561e-01 9.77779925e-01 6.29553854e-01 -2.50012577e-01
2.44284302e-01 2.98961610e-01 -2.07398087e-02 -2.02291802e-01
-8.15742731e-01 -3.19913208e-01 2.56746203e-01 4.05212641e-01
3.40361089e-01 -1.27360612e-01 -3.94830406e-01 3.08031559e-01
1.57540664e-01 2.95204580e-01 5.28597474e-01 1.00737286e+00
-4.65890288e-01 -8.10768425e-01 -1.62519440e-01 2.16117185e-02
-5.41591287e-01 6.81899861e-02 -4.90164489e-01 9.06264067e-01
3.30790430e-01 7.63395727e-01 2.70020455e-01 -2.21392050e-01
3.93378079e-01 1.07114047e-01 6.07819498e-01 -6.05952442e-01
-3.28865290e-01 1.91131517e-01 4.01542068e-01 -1.13889372e+00
-8.04084718e-01 -7.80721724e-01 -1.24205160e+00 -1.23186382e-02
1.91716924e-01 1.15044534e-01 5.96363954e-02 7.84976244e-01
2.61609823e-01 4.94559377e-01 2.35031471e-01 -1.08597803e+00
1.25839919e-01 -7.26746500e-01 -5.27429819e-01 8.10559869e-01
2.15697378e-01 -8.32892299e-01 -4.30124342e-01 5.64192295e-01] | [8.305116653442383, 0.48820042610168457] |
a6ccc840-4c45-40f7-8cff-98de073ca9fe | end-to-end-speaker-attributed-asr-with | 2104.02128 | null | https://arxiv.org/abs/2104.02128v1 | https://arxiv.org/pdf/2104.02128v1.pdf | End-to-End Speaker-Attributed ASR with Transformer | This paper presents our recent effort on end-to-end speaker-attributed automatic speech recognition, which jointly performs speaker counting, speech recognition and speaker identification for monaural multi-talker audio. Firstly, we thoroughly update the model architecture that was previously designed based on a long short-term memory (LSTM)-based attention encoder decoder by applying transformer architectures. Secondly, we propose a speaker deduplication mechanism to reduce speaker identification errors in highly overlapped regions. Experimental results on the LibriSpeechMix dataset shows that the transformer-based architecture is especially good at counting the speakers and that the proposed model reduces the speaker-attributed word error rate by 47% over the LSTM-based baseline. Furthermore, for the LibriCSS dataset, which consists of real recordings of overlapped speech, the proposed model achieves concatenated minimum-permutation word error rates of 11.9% and 16.3% with and without target speaker profiles, respectively, both of which are the state-of-the-art results for LibriCSS with the monaural setting. | ['Takuya Yoshioka', 'Zhuo Chen', 'Zhong Meng', 'Xiaofei Wang', 'Yashesh Gaur', 'Guoli Ye', 'Naoyuki Kanda'] | 2021-04-05 | null | null | null | null | ['speaker-identification'] | ['speech'] | [ 1.11794956e-01 -1.03640422e-01 3.67718250e-01 -5.12198389e-01
-1.81271493e+00 -1.59483567e-01 1.44536555e-01 -2.07723960e-01
-4.49150622e-01 2.85389304e-01 3.35379928e-01 -2.20593557e-01
3.32861245e-01 5.94787821e-02 -7.38042474e-01 -8.04828703e-01
2.56510749e-02 3.72901082e-01 1.41783897e-02 3.74754071e-02
7.74244145e-02 -1.75443667e-04 -1.81053960e+00 3.36285710e-01
6.55759037e-01 1.14699316e+00 3.32492322e-01 1.02655566e+00
9.76591408e-02 8.44266951e-01 -1.07582235e+00 -3.86390716e-01
-2.72528410e-01 -3.66906524e-01 -6.43121660e-01 -1.28445461e-01
9.16830361e-01 -2.56941259e-01 -3.63907844e-01 8.02974164e-01
1.18731213e+00 8.16637725e-02 3.59991401e-01 -8.63510489e-01
-3.30827206e-01 1.15391552e+00 -3.11164409e-01 4.15089011e-01
9.27994028e-02 2.66955197e-02 9.46279764e-01 -1.19006407e+00
-3.71621132e-01 1.31001151e+00 6.93533599e-01 6.26953661e-01
-7.49814510e-01 -9.70573604e-01 4.18764204e-02 4.80360895e-01
-1.87353909e+00 -1.16658294e+00 4.67575699e-01 -2.32989013e-01
1.32201910e+00 5.21632135e-01 2.22993210e-01 1.09327114e+00
-3.00797999e-01 8.07017267e-01 5.52642047e-01 -6.42370641e-01
7.60938227e-02 7.94206709e-02 2.79306412e-01 3.92706513e-01
-5.10662913e-01 5.16557992e-02 -1.07048953e+00 9.25756618e-02
1.61666200e-01 -5.77358603e-01 -3.47886235e-01 4.51513886e-01
-1.09639704e+00 6.30323887e-01 7.62260929e-02 3.22599977e-01
-2.08306029e-01 1.82192057e-01 6.02280676e-01 1.95565000e-02
7.30936348e-01 2.49911938e-02 -4.18195933e-01 -4.01731133e-01
-1.11976278e+00 -1.67361885e-01 6.85914636e-01 1.13627088e+00
3.53073031e-02 7.53067076e-01 -3.69451404e-01 1.45933008e+00
3.70049506e-01 7.91080475e-01 6.44399405e-01 -5.29139042e-01
5.71530998e-01 -6.71855211e-02 -1.19399182e-01 -1.93725571e-01
3.23502049e-02 -9.14646685e-01 -5.47891021e-01 -2.09689334e-01
1.71608478e-01 -1.50355384e-01 -9.47488904e-01 1.92738235e+00
1.80001184e-02 4.80247259e-01 1.37830064e-01 9.75146711e-01
1.06211925e+00 8.39213371e-01 -2.03384329e-02 -2.14301437e-01
1.53325033e+00 -1.21092248e+00 -1.08222663e+00 -2.69897014e-01
2.64519125e-01 -9.69408035e-01 1.23529553e+00 3.62806857e-01
-1.27453160e+00 -5.35674930e-01 -1.09121287e+00 -5.14989812e-03
-9.78221372e-02 3.78283799e-01 -1.87057391e-01 1.20919001e+00
-1.14569819e+00 -2.44810563e-02 -7.12345123e-01 -1.73466489e-01
1.33836284e-01 2.72684485e-01 1.77234143e-01 3.69379729e-01
-1.29336083e+00 7.42191434e-01 -3.79489735e-02 1.25542045e-01
-1.45816422e+00 -8.18409026e-01 -7.66068518e-01 4.45722848e-01
1.83253661e-02 -3.39738518e-01 1.84311211e+00 -6.13686085e-01
-1.85377526e+00 8.02027464e-01 -8.09450626e-01 -8.04356754e-01
1.74426213e-01 -2.45243400e-01 -7.97724724e-01 -9.98500139e-02
-2.21292019e-01 4.13885921e-01 8.61994803e-01 -9.10561919e-01
-6.88748538e-01 -3.75930637e-01 -6.38794422e-01 2.66164064e-01
-6.50325656e-01 6.08188093e-01 -3.29410851e-01 -6.69395924e-01
3.17187048e-02 -5.05292177e-01 3.52255076e-01 -7.24748909e-01
-5.85382581e-01 -4.15141106e-01 5.62171817e-01 -1.06858003e+00
1.31447923e+00 -2.34503126e+00 -1.46475255e-01 -1.85202077e-01
-9.94238257e-02 3.71885508e-01 -3.19072194e-02 2.07054138e-01
-1.52002379e-01 -2.00597599e-01 -1.89285159e-01 -1.08988678e+00
1.48658976e-01 -3.10671628e-01 -4.24230129e-01 3.44159037e-01
-1.00276619e-01 2.55962402e-01 -4.17125285e-01 -2.38426238e-01
3.30794096e-01 7.59107649e-01 -2.29338527e-01 3.71536613e-01
2.73240685e-01 2.28322819e-01 4.92489308e-01 7.27239788e-01
6.31422877e-01 3.87133330e-01 -2.58402646e-01 1.24045402e-01
-3.56209636e-01 9.99746084e-01 -9.81589317e-01 1.64229858e+00
-6.78672135e-01 8.74800086e-01 3.59003454e-01 -4.90973979e-01
9.60681379e-01 7.45137870e-01 -1.23107947e-01 -5.34078479e-01
1.78352222e-01 4.79336292e-01 1.33286091e-02 -2.98172146e-01
4.66394693e-01 -1.64525926e-01 1.19509965e-01 1.72755018e-01
3.44086021e-01 1.23447761e-01 -2.46533647e-01 -7.24429190e-02
7.52594650e-01 -7.01234341e-01 -1.92423463e-01 -2.04064965e-01
8.06288958e-01 -6.76385283e-01 3.64311486e-01 7.63496697e-01
-3.02699208e-01 8.22311878e-01 -6.54188991e-02 -1.38974674e-02
-7.93269575e-01 -1.26504576e+00 -3.22071195e-01 1.59357607e+00
-5.13121426e-01 -2.37640798e-01 -1.16354144e+00 -3.07957176e-02
-3.82976204e-01 1.08328450e+00 -1.70695394e-01 -3.85766812e-02
-6.20833516e-01 -5.41431785e-01 1.20614481e+00 5.37042618e-01
3.14812124e-01 -8.97206306e-01 -1.50842816e-01 2.90623873e-01
-6.81777477e-01 -1.28230083e+00 -1.05382311e+00 2.59563893e-01
-3.04699033e-01 -5.34978569e-01 -1.05594039e+00 -1.17110205e+00
7.95386359e-03 9.21893045e-02 9.30218399e-01 -5.75282693e-01
-4.08696197e-02 1.03921711e-01 -8.46206918e-02 -5.43946147e-01
-5.96147537e-01 2.71577150e-01 4.35769528e-01 2.57735342e-01
5.70238709e-01 -4.63647455e-01 -3.45904231e-01 3.88950616e-01
-1.92209795e-01 -3.12825739e-01 3.42365801e-01 7.78090954e-01
2.84514278e-01 -3.80295217e-01 9.92610216e-01 -3.92994322e-02
4.17956352e-01 -3.97582427e-02 -5.50975382e-01 1.91475198e-01
-2.49327213e-01 -2.09518507e-01 3.36742967e-01 -5.99433661e-01
-9.87123013e-01 -4.29558121e-02 -6.10900342e-01 -5.80775619e-01
-2.45948389e-01 6.97562173e-02 -4.28798944e-01 8.92055780e-02
4.28514123e-01 7.38446295e-01 -2.18936861e-01 -7.84420192e-01
-6.31215870e-02 1.59455049e+00 7.89711833e-01 -8.55931789e-02
1.26409177e-02 -1.28590629e-01 -7.76387572e-01 -1.18168569e+00
-5.53802013e-01 -6.63738132e-01 -1.70888528e-01 -2.15399146e-01
7.26999581e-01 -1.45124125e+00 -1.11853945e+00 9.33862209e-01
-1.32642722e+00 -2.15514749e-01 -9.59450193e-03 7.14265168e-01
-5.57149351e-01 1.52897820e-01 -7.97793031e-01 -1.53407025e+00
-7.96707094e-01 -1.29921699e+00 1.17608654e+00 -2.21957505e-01
-1.10928833e-01 -4.38253522e-01 -3.62662449e-02 5.76785505e-01
6.35974646e-01 -5.59947789e-01 6.69903219e-01 -8.81637514e-01
-2.85913080e-01 -1.25334129e-01 1.72264680e-01 7.05685854e-01
-7.60966167e-02 -4.61960554e-01 -1.77000177e+00 -3.88053507e-01
1.72560871e-01 -8.65425244e-02 8.99768472e-01 6.34204686e-01
1.17053521e+00 -4.59417820e-01 1.02748394e-01 5.24264753e-01
7.84813106e-01 3.21870863e-01 5.79706490e-01 -1.33765697e-01
5.52466094e-01 4.69427913e-01 1.23482682e-01 3.95514637e-01
5.18153012e-01 8.98099363e-01 3.75184923e-01 3.86746824e-02
-6.35744989e-01 -3.05622727e-01 6.97654963e-01 1.64894533e+00
4.67283458e-01 -3.75227064e-01 -9.07744944e-01 1.00220132e+00
-1.26341224e+00 -8.62854898e-01 -2.04658821e-01 2.60914040e+00
8.63956451e-01 -1.46790490e-01 3.23600203e-01 4.76961255e-01
1.26113677e+00 1.26263753e-01 -3.15970004e-01 -6.15195274e-01
-1.19949326e-01 1.95210323e-01 4.95013654e-01 7.12812722e-01
-1.04318058e+00 8.99909794e-01 6.66423798e+00 1.03282809e+00
-1.16401505e+00 6.51828349e-01 4.45142120e-01 -5.44755280e-01
2.58429855e-01 -6.39217794e-01 -1.13559258e+00 5.91537774e-01
1.86713719e+00 2.89504267e-02 4.97875094e-01 7.84484565e-01
2.45505050e-01 3.34241450e-01 -1.07109976e+00 1.37760258e+00
5.56956351e-01 -7.23096550e-01 -6.04660064e-02 -1.74444273e-01
2.41293177e-01 3.10600549e-01 3.66432428e-01 4.54799563e-01
-1.25146031e-01 -9.54460621e-01 1.13290501e+00 1.06258243e-01
9.75396633e-01 -1.08631372e+00 7.90403426e-01 3.37475091e-01
-1.17938137e+00 -9.71072819e-03 -8.47879574e-02 1.86551183e-01
2.68604726e-01 4.18880790e-01 -1.27447307e+00 1.12109996e-01
8.97487402e-01 7.47339204e-02 -2.46091604e-01 1.26911712e+00
-6.31571412e-02 1.20894587e+00 -3.92510653e-01 -9.94481817e-02
-3.93601619e-02 5.44384420e-01 8.90285730e-01 1.52672780e+00
5.33815920e-01 -3.26050222e-01 -3.27556640e-01 6.38702810e-01
-2.50828922e-01 4.45587300e-02 -4.57637012e-02 1.47653997e-01
8.34688365e-01 8.67958546e-01 1.77633509e-01 -4.26496416e-01
-1.06500737e-01 8.03798437e-01 9.39868242e-02 4.00154442e-01
-1.05552983e+00 -7.14785993e-01 8.19766164e-01 -3.55827548e-02
5.76548815e-01 5.38646504e-02 -2.53981948e-01 -8.71792674e-01
7.85927027e-02 -8.78873885e-01 1.57586083e-01 -7.42367625e-01
-1.03361702e+00 1.00151694e+00 -3.30977291e-01 -1.10865688e+00
-4.00498033e-01 -3.41720164e-01 -6.49919212e-01 1.20901442e+00
-1.40988684e+00 -1.15850437e+00 -4.26467434e-02 5.32300174e-01
1.07837856e+00 -4.47670370e-01 1.10771048e+00 8.85811806e-01
-7.20707476e-01 1.43910813e+00 1.23095959e-01 5.25558852e-02
8.83821011e-01 -1.07739031e+00 7.13749349e-01 9.03419971e-01
1.93121091e-01 4.53718603e-01 7.35940933e-01 -3.44705656e-02
-1.03391898e+00 -1.10279000e+00 1.35217655e+00 -2.62650669e-01
4.09752399e-01 -8.90119255e-01 -9.12466109e-01 6.39502227e-01
4.36574042e-01 -3.69227022e-01 7.79744804e-01 3.02202851e-01
-4.32320654e-01 -3.68436307e-01 -9.76454079e-01 2.74783283e-01
6.79089010e-01 -8.55050623e-01 -6.08710349e-01 3.81135255e-01
9.68190253e-01 -4.90964204e-01 -6.30730391e-01 2.06429690e-01
5.14488101e-01 -6.88634634e-01 8.52410078e-01 -1.10565469e-01
-2.32641011e-01 -1.77619681e-01 -4.44977313e-01 -1.38771641e+00
9.99441836e-04 -7.17620015e-01 -1.15539499e-01 1.67616880e+00
6.02531731e-01 -7.21551299e-01 3.62001270e-01 -1.58588476e-02
-8.74529600e-01 -2.35002264e-01 -1.55436850e+00 -9.99583721e-01
-2.19039135e-02 -7.18656480e-01 6.83344424e-01 5.70551693e-01
-8.85482058e-02 4.37831759e-01 -5.22852302e-01 5.01264513e-01
6.89836323e-01 -4.05761868e-01 4.90738243e-01 -7.40773797e-01
-1.89371422e-01 -3.90577018e-01 -2.57681787e-01 -1.16364944e+00
1.79331347e-01 -6.55217469e-01 6.68371677e-01 -1.10632598e+00
5.26296534e-03 1.59807459e-01 -5.08433461e-01 3.49808455e-01
2.91778333e-02 1.48960218e-01 -3.69062759e-02 -8.21096599e-02
-4.75158960e-01 8.64343345e-01 4.41256970e-01 -3.04941982e-01
-1.74214125e-01 2.75446147e-01 -3.84120256e-01 6.41026020e-01
7.63975322e-01 -5.79501271e-01 1.07680112e-01 -7.10298121e-01
-6.55698240e-01 2.14304745e-01 2.20823690e-01 -1.31240356e+00
4.63053465e-01 5.64343154e-01 -9.48811397e-02 -9.83316958e-01
8.25492740e-01 -3.27287674e-01 -2.45440334e-01 4.61681306e-01
-6.36071324e-01 -2.51567543e-01 4.29652989e-01 3.00912768e-01
-4.48051006e-01 -1.56901218e-02 9.02507305e-01 2.20455781e-01
-2.41545752e-01 -8.79949927e-02 -7.87797153e-01 4.68937755e-02
4.70027506e-01 1.26041353e-01 -2.35404938e-01 -4.47069019e-01
-7.40180492e-01 -3.22892405e-02 -4.09174234e-01 5.58899462e-01
6.48629189e-01 -1.20703912e+00 -1.21713233e+00 3.84328246e-01
2.62319744e-01 -3.92241329e-01 4.96470988e-01 8.24422240e-01
-7.82132670e-02 8.15025210e-01 3.53297889e-01 -8.86814654e-01
-1.69831717e+00 1.13389783e-01 5.49974442e-01 3.82996619e-01
-2.71711707e-01 1.43939590e+00 2.16556534e-01 -4.41597641e-01
9.07228947e-01 -4.68679518e-01 -1.83101068e-03 -1.01898298e-01
8.39426935e-01 5.94105244e-01 6.58841968e-01 -8.74397933e-01
-6.90198362e-01 2.18629286e-01 -1.63136899e-01 -5.39997876e-01
9.58875835e-01 -3.41245055e-01 2.25400582e-01 8.39078426e-01
1.12126493e+00 1.52589038e-01 -8.82650495e-01 -2.58188903e-01
-2.56947905e-01 -2.81448245e-01 4.90267277e-01 -1.00997066e+00
-8.76456976e-01 1.43700337e+00 1.01961875e+00 2.23563686e-02
1.00973570e+00 5.17039858e-02 1.05248260e+00 1.34748906e-01
1.17572837e-01 -9.70165074e-01 -1.58496261e-01 7.94144988e-01
1.02427936e+00 -1.01157463e+00 -9.02884781e-01 -1.07234553e-01
-6.02992296e-01 7.86165953e-01 5.30737877e-01 4.23445672e-01
3.16733271e-01 3.18980902e-01 1.85586840e-01 2.39353895e-01
-7.82400191e-01 -2.31189504e-01 3.05729002e-01 4.57194716e-01
5.48487246e-01 2.26035193e-01 3.03780705e-01 7.94297516e-01
-6.43454194e-01 -5.04592001e-01 2.86285222e-01 1.71242148e-01
-6.47294164e-01 -5.08243740e-01 -8.02887022e-01 -4.53431569e-02
-6.87615514e-01 -6.71244681e-01 -3.49599719e-01 6.19091429e-02
-2.37256005e-01 1.53624940e+00 3.03602874e-01 -3.97251606e-01
4.41930354e-01 5.76519251e-01 1.80686742e-01 -4.17884171e-01
-8.32122266e-01 4.58652973e-01 1.66359648e-01 -2.79948544e-02
4.43892069e-02 -6.01500750e-01 -1.02721846e+00 -2.71777064e-01
-6.63491249e-01 2.52383828e-01 1.12169456e+00 7.90207267e-01
3.95937622e-01 1.07997060e+00 6.47959530e-01 -6.29839659e-01
-7.97307730e-01 -1.74700117e+00 -6.50408745e-01 -2.06668854e-01
8.01188052e-01 -3.61231923e-01 -7.24435747e-01 -6.57981858e-02] | [14.631935119628906, 6.151705265045166] |
579cdff2-ea1b-40a3-becd-c7995bb818a6 | scope-and-arbitration-in-machine-learning | 2303.06386 | null | https://arxiv.org/abs/2303.06386v1 | https://arxiv.org/pdf/2303.06386v1.pdf | Scope and Arbitration in Machine Learning Clinical EEG Classification | A key task in clinical EEG interpretation is to classify a recording or session as normal or abnormal. In machine learning approaches to this task, recordings are typically divided into shorter windows for practical reasons, and these windows inherit the label of their parent recording. We hypothesised that window labels derived in this manner can be misleading for example, windows without evident abnormalities can be labelled `abnormal' disrupting the learning process and degrading performance. We explored two separable approaches to mitigate this problem: increasing the window length and introducing a second-stage model to arbitrate between the window-specific predictions within a recording. Evaluating these methods on the Temple University Hospital Abnormal EEG Corpus, we significantly improved state-of-the-art average accuracy from 89.8 percent to 93.3 percent. This result defies previous estimates of the upper limit for performance on this dataset and represents a major step towards clinical translation of machine learning approaches to this problem. | ['David Western', 'Luke J. W. Canham', 'Yixuan Zhu'] | 2023-03-11 | null | null | null | null | ['eeg', 'eeg'] | ['methodology', 'time-series'] | [ 7.51059353e-01 3.92657936e-01 -7.11797178e-02 -7.74828315e-01
-1.30403471e+00 -6.36290014e-01 2.29258478e-01 6.27129018e-01
-6.40453160e-01 9.77589846e-01 1.68893382e-01 -7.03985810e-01
-9.05023366e-02 -1.83812767e-01 -4.63978976e-01 -6.27621830e-01
-4.34548080e-01 3.72469395e-01 3.56547743e-01 4.83192027e-01
4.75109011e-01 3.46827477e-01 -1.00592434e+00 8.20860863e-01
6.02211475e-01 8.88678312e-01 -6.83592707e-02 6.77985668e-01
2.54819065e-01 6.32413328e-01 -1.10721958e+00 -1.28462613e-01
-1.64057128e-02 -7.30501533e-01 -9.21087563e-01 8.05672556e-02
-3.74633186e-02 -2.13862211e-01 1.19669862e-01 7.65814424e-01
6.01093233e-01 -1.36164367e-01 7.53567159e-01 -1.02514744e+00
2.37597059e-03 7.60896623e-01 -5.41411102e-01 8.43968153e-01
3.78459573e-01 -6.83351830e-02 9.34789360e-01 -4.26829129e-01
3.26442301e-01 2.84841985e-01 8.52046728e-01 4.73926991e-01
-1.66083360e+00 -8.01600993e-01 1.81378156e-01 1.83641434e-01
-1.25172603e+00 -6.42273366e-01 4.84009475e-01 -8.81284535e-01
1.22831774e+00 6.09467268e-01 8.69406700e-01 9.08998907e-01
5.75155795e-01 4.61531609e-01 1.17853200e+00 -4.50743288e-01
3.72275889e-01 1.09442882e-01 1.92592829e-01 -4.68882062e-02
2.12571487e-01 -1.47091344e-01 -5.01069844e-01 -5.39756179e-01
5.14439583e-01 -3.51506174e-01 -4.42922115e-01 1.55348092e-01
-1.23048663e+00 5.61934710e-01 -1.45769998e-01 3.48906666e-01
-3.74688804e-01 -2.44119644e-01 6.55761719e-01 4.18847114e-01
5.16518354e-01 8.59789729e-01 -6.29630804e-01 -5.28718412e-01
-1.35971105e+00 1.18063398e-01 6.44469380e-01 3.58683556e-01
-2.28850543e-02 -3.63254100e-01 -4.97525698e-03 8.31200778e-01
1.97680574e-02 -3.49595606e-01 6.28842711e-01 -7.44614124e-01
5.37260354e-01 2.55158216e-01 9.87615585e-02 -4.32566285e-01
-8.00393105e-01 -5.07807076e-01 -2.67905205e-01 2.01923206e-01
5.07042110e-01 -3.48247319e-01 -1.04320610e+00 1.53897583e+00
-2.22883344e-01 -1.65140510e-01 -7.87545368e-02 5.69417417e-01
2.82083005e-01 2.25834846e-01 2.58090407e-01 -5.82128167e-01
1.24909401e+00 -3.33143830e-01 -6.16823256e-01 -4.39865142e-01
9.41704512e-01 -7.41313696e-01 7.29679525e-01 9.33593094e-01
-1.06633353e+00 -2.99831294e-02 -1.18519461e+00 4.90692049e-01
6.61201254e-02 -5.33733256e-02 4.25401151e-01 7.29492426e-01
-9.86040473e-01 6.31176293e-01 -1.15364325e+00 -1.96025386e-01
6.32627666e-01 6.14736021e-01 -3.54502589e-01 5.01489699e-01
-9.39915597e-01 1.07928765e+00 5.52875578e-01 1.10860541e-01
-3.83422047e-01 -6.58520460e-01 -4.24401104e-01 -8.04730803e-02
-2.77329814e-02 -5.82342409e-02 1.37754595e+00 -9.51050997e-01
-6.00510240e-01 1.13723159e+00 -1.72809035e-01 -6.49899125e-01
5.17837405e-01 -7.86710903e-02 -6.78973138e-01 2.72051655e-02
1.11100256e-01 2.30754271e-01 7.79473960e-01 -9.36863601e-01
-8.22639227e-01 -5.23448229e-01 -2.29672447e-01 4.81415652e-02
2.10939676e-01 1.62411675e-01 1.84600838e-02 -8.24347615e-01
3.54595989e-01 -8.48233819e-01 -8.71106684e-02 -4.79946375e-01
-3.49497527e-01 -6.78262413e-02 2.66345888e-01 -8.98456156e-01
1.56966722e+00 -2.12275767e+00 -3.73191178e-01 4.08647329e-01
3.98469001e-01 6.00601062e-02 4.11179721e-01 2.49773711e-01
-6.11818373e-01 1.61258191e-01 -3.58078659e-01 2.24548206e-02
-4.13435429e-01 -3.27969603e-02 -3.78614455e-01 6.95593476e-01
1.01729088e-01 4.52309072e-01 -6.51806355e-01 -4.40970212e-01
-7.39133805e-02 1.81579754e-01 -4.99693573e-01 1.89641625e-01
4.05819356e-01 5.83750844e-01 -3.27055529e-02 3.30944747e-01
1.25989541e-01 -7.03067258e-02 2.64538407e-01 1.00163708e-03
1.11437097e-01 8.62601399e-01 -8.61340463e-01 1.29824150e+00
1.56852212e-02 8.43082726e-01 -3.66022140e-01 -1.16318059e+00
7.16431797e-01 6.07473016e-01 5.73116779e-01 -2.58829027e-01
-2.18168907e-02 3.11889112e-01 7.05726802e-01 -5.41592538e-01
8.41390938e-02 -4.16816354e-01 1.31193906e-01 5.84721684e-01
-2.08880812e-01 -3.46800201e-02 -9.51837301e-02 -1.98330387e-01
1.47801960e+00 -1.03403009e-01 4.68653083e-01 -3.39418858e-01
5.72196357e-02 2.05145013e-02 6.58799827e-01 8.62145126e-01
-3.18846762e-01 9.60672975e-01 8.51165473e-01 -2.97556251e-01
-8.13897669e-01 -1.05567789e+00 -6.97070122e-01 8.06190729e-01
-5.92666805e-01 -5.29828191e-01 -8.55192780e-01 -7.66095936e-01
-2.94707447e-01 1.05546379e+00 -6.49566710e-01 -4.21056062e-01
-5.22333682e-01 -1.17604434e+00 7.22236335e-01 8.59541357e-01
-2.63528347e-01 -1.08224976e+00 -1.23519886e+00 5.22554815e-01
-2.03062505e-01 -8.80073726e-01 -2.30107382e-01 8.27628434e-01
-9.73250151e-01 -1.11889684e+00 -5.30728459e-01 -5.57750821e-01
7.69011497e-01 -3.54313225e-01 1.08466160e+00 4.66651320e-02
-2.60154396e-01 6.33585528e-02 -4.63803321e-01 -6.18656754e-01
-3.85677338e-01 7.68117681e-02 1.19880354e-03 -2.49461025e-01
6.96631968e-01 -6.67504907e-01 -7.09843099e-01 9.12137106e-02
-7.93068111e-01 -1.22594982e-01 6.23624623e-01 9.31701660e-01
3.62034827e-01 -8.08179826e-02 8.93850923e-01 -1.01295769e+00
8.39099288e-01 -4.28048223e-01 -1.57917351e-01 1.17628455e-01
-8.41116726e-01 -2.09164545e-01 3.55169445e-01 -5.72247684e-01
-5.22332549e-01 5.90102412e-02 -3.00089896e-01 7.59311244e-02
-3.46988827e-01 3.63853604e-01 1.16326898e-01 1.49811208e-01
6.72556221e-01 1.28753722e-01 -2.59660184e-01 -2.42703840e-01
-4.89615649e-01 9.91763175e-01 5.44711351e-01 -3.45179290e-01
1.46434620e-01 2.01257586e-01 -3.33692193e-01 -5.56835234e-01
-7.06057906e-01 -4.89732444e-01 -6.15666509e-01 -1.42745733e-01
8.10219646e-01 -6.30716622e-01 -3.73814374e-01 1.08372688e-01
-9.95068014e-01 -4.22677249e-01 -2.49598935e-01 7.80008197e-01
-6.77761316e-01 8.73829052e-02 -2.94275761e-01 -7.44825542e-01
-3.04449677e-01 -1.11609364e+00 7.88215339e-01 -2.09169820e-01
-1.15571928e+00 -7.83965230e-01 1.55205771e-01 9.73450914e-02
-2.12599020e-02 2.62326181e-01 1.20141339e+00 -1.41036046e+00
2.97087282e-01 -4.75975901e-01 8.69410709e-02 3.19308221e-01
2.18876451e-01 -3.27846855e-01 -1.19490600e+00 -2.74380535e-01
3.44429344e-01 -1.19570419e-01 5.22954881e-01 4.28400934e-01
1.29441988e+00 1.68813225e-02 -4.38195258e-01 2.98286736e-01
1.03092313e+00 8.48847747e-01 5.92480898e-01 5.33594191e-01
1.39163479e-01 4.43794012e-01 3.97891343e-01 4.59402114e-01
1.23851253e-02 5.32113910e-01 -1.00669079e-01 -1.65438391e-02
4.34248835e-01 1.11162528e-01 1.21351488e-01 4.78162229e-01
1.34564359e-02 2.44931597e-03 -1.24004734e+00 6.99626327e-01
-1.63330376e+00 -7.33713686e-01 1.22511506e-01 2.47589231e+00
7.62152970e-01 8.31577718e-01 1.48799315e-01 5.74998081e-01
5.91162205e-01 -7.12333620e-02 -6.24083996e-01 -6.13571465e-01
2.67132819e-01 2.50252843e-01 3.86248380e-01 2.04029083e-01
-8.17922235e-01 3.99029016e-01 7.45683146e+00 4.17955309e-01
-1.06191897e+00 1.58756658e-01 9.18820560e-01 -3.89185995e-01
1.25688374e-01 -1.47888809e-01 -6.05077147e-01 5.90095282e-01
1.30773377e+00 -7.57825598e-02 1.28076792e-01 4.00855124e-01
3.85466874e-01 -5.16374409e-01 -1.57101524e+00 9.97218192e-01
8.47532079e-02 -1.01854706e+00 -4.11473870e-01 6.46269247e-02
2.32598990e-01 1.13385819e-01 -1.86217606e-01 1.08559042e-01
-3.40469450e-01 -1.18053782e+00 6.81627333e-01 3.10870647e-01
8.55494857e-01 -7.25704610e-01 7.99162328e-01 3.13937157e-01
-6.29077852e-01 -3.41372080e-02 8.51120101e-04 -1.17481902e-01
1.97527170e-01 4.06600028e-01 -1.19082737e+00 -5.28576933e-02
7.94796586e-01 3.96293640e-01 -6.46709442e-01 1.29182637e+00
6.30642753e-03 1.05915368e+00 -1.63361371e-01 4.90361035e-01
1.08750239e-01 1.83420628e-01 5.07357717e-01 1.25403333e+00
1.03853300e-01 2.60865599e-01 -1.41136274e-01 5.87368906e-01
2.80363172e-01 -1.33023694e-01 -4.32088822e-01 -3.63648944e-02
4.02328640e-01 8.09808791e-01 -1.14085472e+00 -4.01395023e-01
-4.97329175e-01 8.34119022e-01 1.85052603e-01 2.34291688e-01
-8.31186354e-01 -6.28644645e-01 2.84831643e-01 2.24788174e-01
-2.13683583e-02 2.73607850e-01 -7.83152938e-01 -9.11515594e-01
3.79491925e-01 -8.89996469e-01 5.23843169e-01 -5.36717653e-01
-9.63401616e-01 8.51369679e-01 5.38994595e-02 -1.21170986e+00
-4.75773960e-01 -4.43419755e-01 -6.12121403e-01 8.55084538e-01
-8.19297850e-01 -4.45150018e-01 2.85116613e-01 2.72271514e-01
6.14646316e-01 1.42157033e-01 1.06525970e+00 1.61018342e-01
-3.01218003e-01 7.31779933e-01 -2.27319002e-02 2.07326844e-01
8.92797410e-01 -1.30453193e+00 3.22238892e-01 6.91509306e-01
1.95394963e-01 5.46967387e-01 9.33193624e-01 -6.30657136e-01
-4.27955955e-01 -6.61758721e-01 9.87810194e-01 -6.00830793e-01
6.67363703e-01 -2.44833440e-01 -1.23270810e+00 1.01087439e+00
-1.23193674e-02 -3.39286864e-01 1.16390383e+00 2.82120794e-01
-6.60080016e-02 1.14256419e-01 -1.17200947e+00 3.32874775e-01
7.55763173e-01 -5.38494170e-01 -1.20424974e+00 3.31645548e-01
1.39500037e-01 -5.00170827e-01 -1.13426042e+00 4.81935620e-01
8.60132039e-01 -1.03374100e+00 5.21503031e-01 -8.09645474e-01
1.39083415e-01 1.38715491e-01 1.34346828e-01 -1.51405501e+00
-1.33949593e-01 -6.48384988e-01 1.05117552e-01 8.76747847e-01
8.99161220e-01 -7.89663732e-01 7.02989459e-01 1.02294242e+00
-3.13407034e-01 -1.27455127e+00 -9.31875050e-01 -6.09244823e-01
1.18373618e-01 -8.52842152e-01 5.02169788e-01 7.67257512e-01
7.18545496e-01 9.13549885e-02 -9.99375340e-03 8.63478631e-02
2.39399567e-01 -3.77465218e-01 -4.39556800e-02 -1.13488913e+00
-2.77209014e-01 -5.36456048e-01 -7.69951522e-01 -5.52911937e-01
-2.94835977e-02 -7.54660726e-01 2.35064849e-01 -1.31789458e+00
2.38310695e-01 -4.70511168e-01 -6.58059239e-01 5.57235956e-01
-1.95518985e-01 4.38785017e-01 -1.20308794e-01 2.33522698e-01
-1.83551937e-01 -3.35066408e-01 4.60202962e-01 1.09916471e-01
-4.54297602e-01 3.65341485e-01 -7.39914417e-01 9.36080515e-01
1.09203100e+00 -7.53340960e-01 -4.94238645e-01 -2.61003345e-01
1.13812916e-01 1.11876234e-01 5.51380254e-02 -9.46422637e-01
7.54136667e-02 1.68624863e-01 8.08474779e-01 -5.92451870e-01
2.05401137e-01 -7.66391993e-01 1.46643206e-01 4.32926148e-01
-6.79039180e-01 3.56738627e-01 4.31178302e-01 3.11943501e-01
1.08075598e-02 -2.05997795e-01 7.77500749e-01 -1.08566098e-01
-1.92330375e-01 -2.31025487e-01 -8.52909923e-01 3.14044714e-01
1.10252309e+00 -4.60272044e-01 -2.04240866e-02 -1.52201504e-01
-1.23266041e+00 -3.49450856e-02 1.97877720e-01 2.60169387e-01
5.88032484e-01 -8.83609951e-01 -5.89215994e-01 3.88827354e-01
2.17301518e-01 -2.80707389e-01 -7.63680413e-02 1.34085524e+00
-3.78233641e-01 5.65314293e-01 -1.86080605e-01 -6.02113307e-01
-1.47314763e+00 5.30810468e-02 3.43778640e-01 -1.47046492e-01
-1.04169631e+00 8.88927400e-01 8.50647539e-02 2.00761080e-01
3.96619767e-01 -4.97792751e-01 -1.62164539e-01 2.06753835e-01
7.68173873e-01 1.99182048e-01 5.66671193e-01 -4.82323706e-01
-5.21819234e-01 -5.61830960e-02 -5.08460402e-01 -4.19786602e-01
1.14903426e+00 6.27621263e-02 3.62626575e-02 9.38473463e-01
1.00122333e+00 -2.15925291e-01 -9.76824582e-01 3.03951770e-01
3.79499465e-01 -2.73441851e-01 9.66118053e-02 -1.08991015e+00
-6.11088097e-01 6.83171690e-01 6.95158958e-01 4.17211324e-01
1.17325830e+00 8.40543136e-02 5.16319692e-01 3.93279046e-02
3.63498002e-01 -1.01577950e+00 -4.94784504e-01 -8.55466053e-02
7.09177971e-01 -9.69435334e-01 1.75340205e-01 3.32588926e-02
-6.94655240e-01 1.10590446e+00 3.85698915e-01 -8.67739767e-02
5.20107448e-01 3.32000524e-01 1.73848554e-01 -2.18470052e-01
-9.26261902e-01 3.54715019e-01 2.99989641e-01 5.23857534e-01
7.38736153e-01 2.98544526e-01 -6.86882436e-01 7.67519951e-01
-3.28482717e-01 1.04023226e-01 7.55599499e-01 1.14777160e+00
-3.59147370e-01 -8.78262758e-01 -4.32571977e-01 1.31782925e+00
-1.12168837e+00 -2.32440040e-01 -2.31849954e-01 7.04956234e-01
1.04705691e-01 1.02970040e+00 3.52765381e-01 -4.27487314e-01
2.91148871e-01 5.59591234e-01 5.45910537e-01 -7.96488881e-01
-6.59592807e-01 3.68970543e-01 2.69006431e-01 -4.00780261e-01
-1.81007147e-01 -9.86466527e-01 -1.27494550e+00 2.86517799e-01
-4.29455936e-01 5.13034940e-01 4.55685794e-01 1.21119034e+00
8.69141370e-02 8.32276046e-01 1.50747359e-01 -5.49619555e-01
-5.29253185e-01 -9.51564729e-01 -6.97096586e-01 1.05449207e-01
6.60285592e-01 -4.79252160e-01 -5.42863429e-01 2.97023922e-01] | [13.335137367248535, 3.4931297302246094] |
85b95c98-63fe-40ab-9ec1-3a4bc9563c33 | graformer-graph-oriented-transformer-for-3d | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Zhao_GraFormer_Graph-Oriented_Transformer_for_3D_Pose_Estimation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Zhao_GraFormer_Graph-Oriented_Transformer_for_3D_Pose_Estimation_CVPR_2022_paper.pdf | GraFormer: Graph-Oriented Transformer for 3D Pose Estimation | In 2D-to-3D pose estimation, it is important to exploit the spatial constraints of 2D joints, but it is not yet well modeled. To better model the relation of joints for 3D pose estimation, we propose an effective but simple network, called GraFormer, where a novel transformer architecture is designed via embedding graph convolution layers after multi-head attention block. The proposed GraFormer is built by repeatedly stacking the GraAttention block and the ChebGConv block. The proposed GraAttention block is a new transformer block designed for processing graph-structured data, which is able to learn better features through capturing global information from all the nodes as well as the explicit adjacency structure of nodes. To model the implicit high-order connection relations among non-neighboring nodes, the ChebGConv block is introduced to exchange information between non-neighboring nodes and attain a larger receptive field. We have empirically shown the superiority of GraFormer through extensive experiments on popular public datasets. Specifically, GraFormer outperforms the state-of-the-art GraghSH on the Human3.6M dataset yet only contains 18% parameters of it | ['Yunjie Tian', 'Weiqiang Wang', 'Weixi Zhao'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['3d-pose-estimation'] | ['computer-vision'] | [-4.73113954e-01 5.81998527e-01 4.27119248e-02 -3.77530485e-01
7.32911751e-02 -6.13860749e-02 4.91222829e-01 -4.53924946e-02
-5.18248379e-01 2.96971649e-01 3.94317001e-01 1.42741874e-01
-2.61048764e-01 -7.65769899e-01 -1.02868247e+00 -5.89058757e-01
-5.29951155e-01 6.61590815e-01 3.61894131e-01 -3.76996279e-01
-1.11546226e-01 4.74903226e-01 -1.00153232e+00 -6.89790584e-03
7.07562506e-01 1.01224816e+00 2.21595734e-01 1.90434292e-01
9.50524881e-02 5.26454747e-01 -3.18177998e-01 -3.58662903e-01
5.26418276e-02 6.13855831e-02 -7.58079767e-01 -1.35233654e-02
5.22092640e-01 -3.60212177e-01 -6.16237044e-01 6.99339807e-01
6.54134274e-01 1.20248735e-01 7.31592059e-01 -1.04806912e+00
-7.59127378e-01 7.10296810e-01 -6.87207222e-01 -9.78456065e-03
1.26149356e-01 1.27417728e-01 1.24449837e+00 -9.36655223e-01
7.14508832e-01 1.59524870e+00 5.83104193e-01 3.67870718e-01
-1.02906692e+00 -5.65086782e-01 5.30042827e-01 4.28653836e-01
-1.43260252e+00 1.61543950e-01 1.17080593e+00 -1.30767867e-01
9.13998485e-01 -1.96152702e-02 8.82130742e-01 1.23150349e+00
6.27899170e-01 7.94378400e-01 6.69538558e-01 1.02670155e-01
-1.73276693e-01 -3.22180122e-01 2.89460838e-01 9.85661387e-01
7.69365206e-02 6.01461120e-02 -4.54381913e-01 1.66242659e-01
1.14626801e+00 -2.88186944e-03 -1.72011808e-01 -8.09025943e-01
-1.12640476e+00 7.32007623e-01 1.53134978e+00 2.08355427e-01
-4.27127719e-01 5.69459677e-01 1.33898914e-01 3.59320492e-02
3.72612417e-01 2.89498389e-01 -3.53126436e-01 4.95976031e-01
-4.12776887e-01 1.67465195e-01 5.52244544e-01 9.19015825e-01
7.86542118e-01 -1.04535513e-01 -2.80513823e-01 6.87473178e-01
7.24647939e-01 8.94208997e-02 2.67577797e-01 -5.52675068e-01
6.50345683e-01 9.23657358e-01 -3.57272536e-01 -1.33033395e+00
-9.23212826e-01 -9.86084580e-01 -1.20092881e+00 -6.52327240e-02
1.70822397e-01 1.55499745e-02 -1.05106473e+00 1.78933442e+00
4.89710599e-01 2.07517192e-01 -4.59059983e-01 1.28835940e+00
1.10091376e+00 5.50005436e-01 -4.69396152e-02 4.50887918e-01
1.42610288e+00 -1.14688301e+00 -3.31930667e-01 -4.16590154e-01
5.26270866e-01 -3.41227144e-01 8.09827447e-01 4.55578938e-02
-1.02421379e+00 -7.47343957e-01 -1.01833534e+00 -3.99003595e-01
-3.05197597e-01 2.30432481e-01 8.43397141e-01 -5.72568513e-02
-1.16021824e+00 5.46131790e-01 -9.16800499e-01 -3.43558431e-01
3.78218055e-01 6.82153404e-01 -7.58151591e-01 -1.23425409e-01
-1.26496923e+00 8.72677684e-01 3.89486641e-01 7.42858231e-01
-8.47194850e-01 -4.60845172e-01 -1.19247520e+00 2.73395866e-01
6.11910939e-01 -9.35346127e-01 5.53645432e-01 -4.83850181e-01
-1.42438579e+00 5.13333917e-01 2.38718659e-01 -2.54412442e-01
3.60493541e-01 -4.35153991e-01 4.74661356e-03 1.85483083e-01
-3.03118259e-01 8.88558745e-01 8.40617657e-01 -1.14353156e+00
2.17574939e-01 -7.44070053e-01 1.71433702e-01 4.81108844e-01
-3.53967965e-01 -5.52997887e-01 -8.87094498e-01 -6.65598631e-01
5.24679899e-01 -1.00118840e+00 -3.22710425e-01 7.43172318e-02
-5.86939931e-01 -5.03668904e-01 7.25385964e-01 -6.24453783e-01
1.05346227e+00 -2.05529308e+00 5.60912669e-01 4.66784209e-01
7.34158635e-01 1.62598267e-01 -5.58233202e-01 5.13660491e-01
-4.53847684e-02 -1.89261824e-01 1.12389468e-01 -4.40909207e-01
7.61383921e-02 2.70281792e-01 3.99836659e-01 6.26352489e-01
4.62768793e-01 1.30485070e+00 -7.94786870e-01 -3.92072082e-01
4.47256267e-01 8.46380532e-01 -7.83271670e-01 2.22968400e-01
-7.62769207e-02 3.36184680e-01 -7.68475890e-01 3.80317241e-01
7.78497458e-01 -4.95289117e-01 1.54777691e-01 -5.82728326e-01
2.82709837e-01 3.97332579e-01 -1.09661555e+00 1.99443865e+00
-3.07014912e-01 1.70701370e-01 2.40262657e-01 -9.81591523e-01
1.03230977e+00 -1.75318662e-02 4.28922892e-01 -6.79743290e-01
4.19383794e-01 -2.21270651e-01 1.40025601e-01 -1.89657792e-01
2.51156598e-01 3.87000471e-01 -9.68796015e-02 -1.51176555e-02
3.39275926e-01 9.48133767e-02 -1.87978446e-01 3.30922306e-01
1.06369388e+00 2.44356588e-01 4.12825271e-02 -5.25260687e-01
6.13929629e-01 -5.22596061e-01 3.17510456e-01 4.09645885e-01
6.94998726e-02 4.61247027e-01 5.51582754e-01 -5.92214108e-01
-6.62475884e-01 -1.13330686e+00 2.57533431e-01 9.40076888e-01
4.71691251e-01 -5.96295416e-01 -6.81034923e-01 -6.98712587e-01
1.95777342e-01 1.55699417e-01 -9.73414183e-01 -4.50814247e-01
-7.72484958e-01 -4.76629376e-01 2.43786424e-01 7.36094594e-01
7.13927448e-01 -8.95501852e-01 -4.39621985e-01 1.39911368e-01
1.45516172e-01 -1.02665544e+00 -6.66922629e-01 2.07940280e-01
-7.15464354e-01 -9.99661326e-01 -6.51020050e-01 -7.83967376e-01
7.58659184e-01 -9.34621617e-02 1.12612796e+00 2.76791960e-01
-1.22879103e-01 4.73152027e-02 -2.93208987e-01 -2.77074613e-02
3.21129978e-01 6.44900203e-01 -2.13580191e-01 1.31284297e-01
9.79246274e-02 -7.16986358e-01 -7.89543092e-01 4.77108896e-01
-5.44101775e-01 2.55773038e-01 9.61436689e-01 1.00242138e+00
4.73544002e-01 -2.79416889e-01 2.32203841e-01 -7.90784061e-01
2.98485756e-01 -2.23651707e-01 -4.11354780e-01 4.15251888e-02
-3.70004654e-01 4.17589784e-01 5.70621848e-01 -2.90237039e-01
-6.72944129e-01 1.04170941e-01 -3.63755792e-01 -5.21704137e-01
9.78062768e-03 5.74723661e-01 -4.09174263e-01 -1.20983370e-01
1.97214782e-01 -2.98625343e-02 1.12591080e-01 -8.13861847e-01
3.00099045e-01 2.09611773e-01 4.24355716e-01 -4.47772503e-01
8.45177472e-01 2.61680037e-01 3.65781486e-01 -8.72853756e-01
-4.92309988e-01 -1.70164704e-01 -8.03671837e-01 -1.18548468e-01
9.85195220e-01 -8.99922550e-01 -9.65933681e-01 5.81258833e-01
-1.21362793e+00 -2.76883334e-01 1.52667433e-01 6.28133833e-01
-2.58201450e-01 3.03903401e-01 -8.68405223e-01 -4.25813645e-01
-3.69696975e-01 -1.15984094e+00 1.36503983e+00 -5.74122667e-02
-1.33194745e-01 -9.66503024e-01 -2.15984747e-01 2.27438942e-01
2.88770258e-01 1.95839107e-01 1.07930946e+00 -4.44806099e-01
-7.53584087e-01 -2.09094882e-01 -3.89871657e-01 1.42193273e-01
-2.33789816e-01 -3.10664475e-01 -5.91529369e-01 -4.81935471e-01
-3.79202962e-01 -2.37424195e-01 1.17395723e+00 3.10639143e-01
1.18879580e+00 -2.08592072e-01 -3.35438043e-01 7.81786859e-01
1.05339336e+00 -4.12840754e-01 7.87419736e-01 1.16388360e-02
1.28071833e+00 6.51538074e-01 2.54985631e-01 1.55548528e-01
7.28439867e-01 7.09327579e-01 1.10884619e+00 -3.06262404e-01
-1.27093166e-01 -4.00959164e-01 2.69577771e-01 9.95254219e-01
-4.41096842e-01 -4.07038271e-01 -6.85493290e-01 2.08151385e-01
-1.80722713e+00 -4.16475981e-01 -1.01464711e-01 1.96466660e+00
4.10390973e-01 3.08745295e-01 -5.29621467e-02 -1.96515784e-01
5.01694858e-01 6.59986258e-01 -4.17399198e-01 -1.17442615e-01
1.23571701e-01 2.03006208e-01 3.66963089e-01 5.34146369e-01
-1.10904098e+00 1.08869374e+00 5.20164251e+00 7.60720909e-01
-9.70826626e-01 -2.19877720e-01 3.69487584e-01 1.90876782e-01
-3.23797166e-01 -1.39841929e-01 -7.27288187e-01 1.15355603e-01
3.39941829e-01 5.50382435e-01 2.94356555e-01 6.40690207e-01
-9.45795178e-02 1.41424745e-01 -1.18550658e+00 7.87498176e-01
2.22599879e-02 -1.04806530e+00 2.25227728e-01 2.09824547e-01
2.61250138e-01 1.34334266e-01 -7.57086724e-02 3.23173374e-01
8.90967026e-02 -1.15510297e+00 5.89522302e-01 5.39930522e-01
4.68397677e-01 -7.95909464e-01 9.41933632e-01 3.85783941e-01
-1.54384196e+00 1.20579839e-01 -5.07889152e-01 -7.86083937e-02
6.97440293e-04 5.95057845e-01 -6.58847213e-01 8.63480568e-01
6.32044494e-01 1.01457977e+00 -7.80656219e-01 9.23596859e-01
-5.13123691e-01 1.98655739e-01 -4.45483476e-01 -1.65363267e-01
4.77027446e-01 -1.97040603e-01 5.04455507e-01 7.98991740e-01
1.47775725e-01 -1.85912535e-01 2.61690170e-01 8.29927087e-01
-1.66185766e-01 -5.87644540e-02 -3.98769468e-01 8.82954970e-02
1.28171042e-01 1.39657927e+00 -5.13814390e-01 -2.92395614e-02
-3.20271462e-01 1.08131540e+00 8.06429625e-01 5.01224637e-01
-9.51358080e-01 -2.97682613e-01 6.20535612e-01 3.46365690e-01
4.46618974e-01 -5.79107642e-01 1.10333823e-01 -9.91246283e-01
9.57425386e-02 -5.44440567e-01 3.93428653e-01 -7.60657847e-01
-1.29005671e+00 8.29413056e-01 -5.42400964e-02 -8.51255953e-01
-3.85521315e-02 -8.06293786e-01 -5.63722193e-01 8.88830543e-01
-1.27317023e+00 -1.56013131e+00 -5.60216248e-01 7.57827818e-01
1.09459899e-01 2.60450542e-01 5.85124195e-01 1.78910941e-01
-5.92264354e-01 6.34596050e-01 -4.16061610e-01 4.00872141e-01
4.20038044e-01 -1.11103690e+00 4.83836323e-01 3.91939402e-01
1.90024897e-01 7.20042586e-01 3.09618294e-01 -6.90436304e-01
-1.58062124e+00 -1.08635056e+00 8.47983778e-01 -1.56798467e-01
5.13184488e-01 -8.51805031e-01 -9.47656870e-01 7.54475057e-01
1.47814453e-01 2.58265853e-01 4.66218591e-02 2.82589525e-01
-4.52584028e-01 -2.79096931e-01 -5.82068622e-01 5.39001584e-01
1.40595698e+00 -3.76709998e-01 -6.84955478e-01 4.66716066e-02
8.23686361e-01 -5.78369141e-01 -8.92269194e-01 5.85964262e-01
5.91837525e-01 -7.88433194e-01 1.15704477e+00 -3.99407238e-01
4.49568123e-01 -1.72296867e-01 1.41197383e-01 -1.39010477e+00
-6.25635862e-01 -3.96775693e-01 -2.29619116e-01 8.13868403e-01
3.06855559e-01 -5.92559040e-01 1.04506755e+00 2.78306812e-01
-3.64018381e-01 -1.17659366e+00 -9.44792569e-01 -6.66555524e-01
-3.23134684e-03 -1.21872388e-01 5.52946687e-01 7.05012679e-01
-2.35679105e-01 8.32557797e-01 -5.73686600e-01 2.08185792e-01
5.80252409e-01 -6.39237538e-02 8.07257771e-01 -1.27311814e+00
-4.02034849e-01 -2.21109435e-01 -7.36789286e-01 -1.60941398e+00
1.04901806e-01 -9.48056638e-01 -1.28276706e-01 -1.70120466e+00
-1.14472059e-03 -2.72758067e-01 -4.43239778e-01 4.44610447e-01
-1.00530855e-01 2.11520806e-01 1.40894935e-01 -1.89478919e-01
-5.59794843e-01 1.05860448e+00 1.94326544e+00 -2.72028655e-01
-1.81085244e-01 -1.41943514e-01 -3.68619174e-01 7.83734560e-01
3.67269248e-01 -1.67368576e-01 -4.98283416e-01 -8.00618649e-01
1.75193533e-01 -4.24827263e-02 6.96370125e-01 -1.03059733e+00
2.40447670e-01 1.86232239e-01 6.91578746e-01 -9.83289540e-01
6.30530775e-01 -1.04019821e+00 1.37652561e-01 5.08913815e-01
-2.11163774e-01 1.87146559e-01 -1.27788693e-01 6.46350980e-01
-1.28222287e-01 3.45521837e-01 4.55248147e-01 -1.38220116e-01
-7.17076540e-01 8.56619835e-01 8.72404594e-03 -1.98931113e-01
8.40674341e-01 -3.44570652e-02 -1.71113431e-01 -3.42690647e-01
-8.88501167e-01 5.69057882e-01 2.24312812e-01 5.01325011e-01
7.61800885e-01 -1.47621834e+00 -5.94262540e-01 4.95952904e-01
1.29140109e-01 3.75467896e-01 4.42780167e-01 8.99669528e-01
-4.26064610e-01 3.45773697e-01 -2.75636584e-01 -6.86391890e-01
-1.02271581e+00 5.50158441e-01 2.65602678e-01 -5.19275069e-01
-7.75961816e-01 1.15427804e+00 6.32954717e-01 -5.41411877e-01
3.25102538e-01 -6.14824474e-01 -4.23737943e-01 -1.04475141e-01
2.58306377e-02 2.32743055e-01 -5.23113050e-02 -8.28565240e-01
-6.44948363e-01 7.27249622e-01 -1.24828354e-01 3.15944731e-01
1.41952503e+00 -3.70738991e-02 -2.26810694e-01 9.72437486e-02
1.40541768e+00 -3.05053443e-01 -1.32789850e+00 -1.97912917e-01
-2.49833629e-01 -2.22068742e-01 6.64951578e-02 -4.28643346e-01
-1.50977135e+00 1.12308061e+00 2.67816007e-01 -2.00105384e-01
9.61082160e-01 2.03037053e-01 7.94926107e-01 4.77514565e-01
4.59949702e-01 -9.30910826e-01 4.09003198e-01 7.19622552e-01
1.21707392e+00 -9.85647380e-01 7.85015449e-02 -8.14213932e-01
-3.67099732e-01 9.40615654e-01 9.59048927e-01 -5.94945014e-01
8.53595614e-01 -1.68109223e-01 -2.61514992e-01 -5.23270607e-01
-5.74305236e-01 -1.23550333e-01 8.16720665e-01 4.48235750e-01
3.92438680e-01 -1.01297684e-02 -3.09540510e-01 7.10172653e-01
-1.32902458e-01 -4.56022829e-01 -2.08781987e-01 5.29192626e-01
-1.78705588e-01 -1.05446863e+00 -4.99814712e-02 3.99019539e-01
-4.98097017e-02 -6.22434821e-03 -5.82577705e-01 1.15193987e+00
1.01527318e-01 4.20783252e-01 1.80975899e-01 -7.46320784e-01
4.10416245e-01 -2.75926054e-01 5.31149387e-01 -5.57503939e-01
-7.11661100e-01 3.84131938e-01 6.02276176e-02 -1.03256679e+00
-2.63970912e-01 -2.34279588e-01 -1.32167435e+00 -2.31915548e-01
-3.05264860e-01 -1.08948629e-02 3.93191993e-01 8.81525040e-01
5.39537072e-01 8.94451499e-01 4.19226140e-01 -1.13951361e+00
-4.86397266e-01 -1.11474693e+00 -8.05187345e-01 4.44174975e-01
1.81174666e-01 -1.07298112e+00 -6.31070286e-02 -7.11209476e-01] | [7.058216571807861, -0.6285943984985352] |
4ec01265-5ff3-4164-a3c2-0b839c29a77f | filtering-aggression-from-the-multilingual | null | null | https://aclanthology.org/W18-4423 | https://aclanthology.org/W18-4423.pdf | Filtering Aggression from the Multilingual Social Media Feed | This paper describes the participation of team DA-LD-Hildesheim from the Information Retrieval Lab(IRLAB) at DA-IICT Gandhinagar, India in collaboration with the University of Hildesheim, Germany and LDRP-ITR, Gandhinagar, India in a shared task on Aggression Identification workshop in COLING 2018. The objective of the shared task is to identify the level of aggression from the User-Generated contents within Social media written in English, Devnagiri Hindi and Romanized Hindi. Aggression levels are categorized into three predefined classes namely: {`}Overtly Aggressive{`}, {`}Covertly Aggressive{`} and {`}Non-aggressive{`}. The participating teams are required to develop a multi-class classifier which classifies User-generated content into these pre-defined classes. Instead of relying on a bag-of-words model, we have used pre-trained vectors for word embedding. We have performed experiments with standard machine learning classifiers. In addition, we have developed various deep learning models for the multi-class classification problem. Using the validation data, we found that validation accuracy of our deep learning models outperform all standard machine learning classifiers and voting based ensemble techniques and results on test data support these findings. We have also found that hyper-parameters of the deep neural network are the keys to improve the results. | ['M', 'Prasenjit Majumder', 'ip', 'Thomas l', 'S Modha'] | 2018-08-01 | null | null | null | coling-2018-8 | ['aggression-identification'] | ['natural-language-processing'] | [-4.46392089e-01 1.93239432e-02 2.50040084e-01 -3.00418168e-01
-4.89176989e-01 -4.20839220e-01 4.61342216e-01 5.62970221e-01
-9.24128711e-01 5.09199321e-01 4.72372562e-01 -1.60686001e-01
-5.34939110e-01 -8.07307243e-01 -1.83348116e-02 -4.69209373e-01
-9.67720672e-02 7.10177541e-01 -5.56992218e-02 -7.19519496e-01
4.99775797e-01 4.34646964e-01 -1.27496660e+00 5.30778110e-01
3.23957384e-01 6.31552041e-01 -3.35966170e-01 9.57250535e-01
3.16564530e-01 9.97738421e-01 -6.94080293e-01 -7.95659065e-01
3.16817254e-01 -1.80522904e-01 -1.14429247e+00 -4.77541298e-01
2.95640528e-01 -4.35087472e-01 -6.41575038e-01 8.21208358e-01
6.89069808e-01 3.15281779e-01 1.04063118e+00 -1.25970852e+00
-3.76656234e-01 7.38963068e-01 -5.67559779e-01 6.10876739e-01
1.91577271e-01 -1.10939294e-01 1.22197390e+00 -7.60637999e-01
4.94755059e-01 1.18237138e+00 7.43341029e-01 6.98042989e-01
-6.38163924e-01 -9.52263892e-01 -3.75540167e-01 6.43181324e-01
-1.28241861e+00 -1.29467860e-01 7.50414729e-01 -6.05247915e-01
1.08172262e+00 2.86264360e-01 3.98370951e-01 1.32823110e+00
-8.36215019e-02 6.68009877e-01 8.91152322e-01 -4.26793218e-01
6.20219074e-02 3.49941254e-01 6.52558804e-01 5.52857816e-01
2.73343503e-01 -1.22505054e-01 -3.38263988e-01 -3.45776021e-01
2.85868913e-01 -3.03680211e-01 3.50608110e-01 3.23941320e-01
-4.69767064e-01 1.61019742e+00 1.21935573e-03 5.89615226e-01
-3.80921513e-01 3.48332934e-02 1.03118944e+00 2.67840385e-01
5.18014252e-01 4.48750049e-01 -4.52793926e-01 -5.30334413e-01
-6.85649157e-01 4.29595798e-01 5.97991705e-01 3.14276785e-01
1.04642197e-01 5.69887124e-02 -4.29981723e-02 1.51348937e+00
3.53653520e-01 -4.01063859e-02 7.53991961e-01 -7.69802809e-01
5.42252600e-01 6.30302429e-01 -2.56742626e-01 -1.11082864e+00
-6.95247650e-01 -2.93337345e-01 -5.48440635e-01 2.02462554e-01
3.53529125e-01 -3.70511591e-01 -5.72959781e-01 1.60967994e+00
2.31577456e-01 -2.46129125e-01 1.09419331e-01 6.37345612e-01
8.10593367e-01 5.57031572e-01 2.79863745e-01 -2.03215592e-02
9.91717935e-01 -8.46465051e-01 -5.60012281e-01 6.87670559e-02
1.15442419e+00 -5.80994844e-01 6.45027220e-01 8.02974522e-01
-1.15088642e+00 -3.12485874e-01 -1.01767743e+00 1.19422995e-01
-6.96288645e-01 -2.29935437e-01 5.37738323e-01 1.07414448e+00
-8.28524113e-01 4.76537198e-01 -3.42071652e-01 -5.46095192e-01
3.89655441e-01 6.49321258e-01 -7.16261268e-01 3.37084353e-01
-1.38625407e+00 1.10601652e+00 4.60366756e-01 -1.98746070e-01
-6.11779809e-01 -5.08496284e-01 -6.44331515e-01 -2.42512137e-01
-3.82463574e-01 1.23805881e-01 8.22840154e-01 -9.03649330e-01
-1.15887392e+00 1.20785785e+00 6.09673858e-01 -2.87458092e-01
4.91169721e-01 -1.16460316e-01 -5.19829690e-01 -2.30366550e-02
-5.97374998e-02 1.24552175e-01 3.34519655e-01 -8.57211053e-01
-8.53788137e-01 -7.47017920e-01 1.72690704e-01 2.39164025e-01
-8.50469232e-01 6.46536291e-01 2.10355178e-01 -5.89632571e-01
-3.73252064e-01 -9.39297020e-01 1.66732684e-01 -6.95365489e-01
-1.48975298e-01 -5.60614765e-01 8.55330825e-01 -9.93244648e-01
1.60596573e+00 -2.03941870e+00 2.47116968e-01 2.81614602e-01
2.33979926e-01 6.42527163e-01 5.54615771e-03 7.86088943e-01
-3.21236342e-01 -8.50316975e-03 1.45115227e-01 -1.86089486e-01
2.66855583e-02 3.21249396e-01 8.83551966e-03 6.64713502e-01
-2.85588503e-01 2.35214129e-01 -7.27245152e-01 -3.42311174e-01
2.48519227e-01 2.65445560e-01 -6.79918468e-01 2.47034535e-01
5.83812118e-01 1.48261249e-01 -1.16186984e-01 2.54263848e-01
5.75340867e-01 5.84986985e-01 -1.90442074e-02 1.21971317e-01
-8.88718516e-02 1.29635379e-01 -9.89523232e-01 9.71356571e-01
-4.90664452e-01 6.54323697e-01 -6.48385882e-02 -1.25643206e+00
8.54399383e-01 2.51695633e-01 3.93737316e-01 -5.07208705e-01
4.91976529e-01 2.71031827e-01 5.92961967e-01 -6.17101490e-01
5.50588250e-01 -1.01854622e-01 -3.13192010e-01 5.18637180e-01
2.68280387e-01 2.76121974e-01 2.58985311e-01 2.06698298e-01
1.35080755e+00 -4.86900508e-01 3.07766110e-01 -3.01072448e-01
7.26312280e-01 -3.96784633e-01 2.77858675e-01 7.19976664e-01
-5.72674572e-01 4.93043065e-01 5.60224831e-01 -4.56754029e-01
-1.23682189e+00 -8.85617197e-01 -4.10142004e-01 1.75647247e+00
-4.54041630e-01 -6.23426735e-01 -6.96295559e-01 -7.55375266e-01
-3.23135644e-01 1.15826368e+00 -1.01539075e+00 -3.49383742e-01
-6.43863618e-01 -8.49217772e-01 1.18133783e+00 3.57012987e-01
2.83069998e-01 -1.18333519e+00 -5.22487164e-01 1.66253835e-01
-1.94237139e-02 -7.99051762e-01 1.14048578e-01 3.74183089e-01
-1.62100881e-01 -9.88886058e-01 -4.82736617e-01 -5.24058402e-01
1.14046074e-01 -3.35993618e-01 6.03533566e-01 2.38662720e-01
-6.39812291e-01 2.18324959e-01 -7.43987203e-01 -7.03718722e-01
-5.38919210e-01 1.02660963e-02 2.87981987e-01 -7.34007284e-02
6.46688759e-01 -4.29979503e-01 -5.66346824e-01 1.43515125e-01
-7.76638269e-01 -3.56271565e-01 1.29289135e-01 7.58573651e-01
-3.76416981e-01 1.12647824e-01 5.28657794e-01 -9.38889444e-01
9.08535540e-01 -8.96292448e-01 2.04271615e-01 -2.27851734e-01
-2.06599668e-01 -5.32454967e-01 4.89940584e-01 -5.25750995e-01
-7.54315853e-01 -6.63524747e-01 -8.00373733e-01 -5.80854565e-02
-2.98742890e-01 3.83291215e-01 2.10462749e-01 1.03492029e-01
6.40041471e-01 -9.83951017e-02 -3.68541181e-01 -2.29657754e-01
1.73034277e-02 1.35652745e+00 -2.94662099e-02 -3.55352789e-01
4.13737595e-01 4.98782247e-02 -1.68234229e-01 -1.16916347e+00
-7.20565081e-01 -5.55498719e-01 -6.65797830e-01 -4.35363024e-01
1.15059602e+00 -6.56397402e-01 -6.62221134e-01 8.05287004e-01
-1.00947702e+00 -2.41976961e-01 1.43432274e-01 3.56363714e-01
-3.35580051e-01 3.59689742e-01 -8.30172896e-01 -1.15664542e+00
-5.22411168e-01 -7.90359795e-01 5.09413540e-01 1.36870399e-01
-9.42669690e-01 -1.07554901e+00 5.69603622e-01 9.13812935e-01
3.40439975e-01 1.80311888e-01 1.24903178e+00 -1.66514385e+00
8.88568282e-01 -6.35474503e-01 2.10227389e-02 6.93456411e-01
-1.91173807e-01 2.86920637e-01 -8.71635914e-01 -1.74162894e-01
-7.21989423e-02 -7.01373935e-01 3.90251994e-01 -1.12414770e-01
1.22549975e+00 -3.27615768e-01 1.13319091e-01 1.43324748e-01
1.04436445e+00 4.59499359e-01 6.18580341e-01 7.87120581e-01
5.79948902e-01 7.20137000e-01 3.94373626e-01 8.01360190e-01
4.39847738e-01 7.25942135e-01 3.33750933e-01 3.41678739e-01
1.75404981e-01 3.09588797e-02 3.94011557e-01 8.29228401e-01
-2.99650609e-01 -2.97326416e-01 -1.12159991e+00 5.59748530e-01
-1.79581571e+00 -1.31087017e+00 -5.38534164e-01 1.99359214e+00
7.29088902e-01 -3.60852331e-02 6.84130847e-01 6.74398124e-01
5.23510575e-01 1.19801762e-03 4.34466079e-02 -1.44662619e+00
1.48288295e-01 6.22808397e-01 3.06604475e-01 4.96141613e-01
-1.03716946e+00 9.30335939e-01 5.18925095e+00 8.09389234e-01
-9.43979204e-01 3.54969293e-01 4.38639462e-01 -4.49175358e-01
3.05490673e-01 -5.84050059e-01 -7.85448790e-01 5.62914371e-01
1.54555225e+00 -2.64121704e-02 3.43575895e-01 6.88633025e-01
-3.02988738e-02 3.92997637e-02 -7.94397950e-01 8.69806468e-01
3.73483419e-01 -8.11578214e-01 -2.03673229e-01 3.30617130e-01
4.34916377e-01 2.54033059e-01 1.45712882e-01 8.47261250e-01
4.69134033e-01 -1.21273458e+00 5.72077513e-01 5.47383213e-03
2.58003294e-01 -1.10439134e+00 9.93708074e-01 5.00678480e-01
-3.38531494e-01 -6.97003901e-01 -3.96926403e-01 -3.04235041e-01
-2.74384916e-01 -8.67020860e-02 -6.77087843e-01 3.30543727e-01
9.64804113e-01 1.48565784e-01 -6.17124438e-01 6.98783338e-01
6.16197623e-02 9.51681197e-01 -2.41659135e-01 -2.74602443e-01
4.86420691e-01 -1.97317407e-01 2.65309185e-01 1.39352000e+00
1.39978364e-01 2.34994873e-01 -3.06172222e-01 7.91107416e-02
2.15939641e-01 4.35946524e-01 -6.04012072e-01 -2.12822497e-01
1.99195538e-02 1.38552940e+00 -4.42882240e-01 5.81205562e-02
-4.50379580e-01 8.34365845e-01 7.58515358e-01 -1.46193698e-01
-9.14268374e-01 -5.28911531e-01 6.95661724e-01 1.96704626e-01
6.15883656e-02 -1.08992733e-01 -3.27548414e-01 -7.24497855e-01
-6.14339769e-01 -9.39682484e-01 8.45573425e-01 -3.60753298e-01
-1.37263536e+00 6.23264134e-01 2.83698767e-01 -6.55176222e-01
-1.92113444e-01 -8.18437397e-01 -7.60243058e-01 7.38496661e-01
-5.19541860e-01 -1.23832262e+00 -1.11404927e-02 6.07302368e-01
5.43909907e-01 -6.01848722e-01 9.65293944e-01 5.59784234e-01
-9.74630654e-01 8.70487988e-01 2.29960367e-01 5.49409330e-01
5.09899080e-01 -1.09665072e+00 -2.41110608e-01 3.52926373e-01
1.40837263e-02 5.38295627e-01 9.70066428e-01 -4.67870265e-01
-7.04427063e-01 -9.14561629e-01 9.15287435e-01 -5.43903291e-01
1.09481096e+00 -5.91253817e-01 -6.78561389e-01 6.97157264e-01
4.66074884e-01 -5.37008762e-01 1.46751904e+00 2.64276266e-01
-3.63332033e-01 2.64515907e-01 -1.27652836e+00 4.89282101e-01
7.34485805e-01 -2.37662449e-01 -7.88852572e-01 5.83174765e-01
-9.74228457e-02 -1.51733056e-01 -7.78717041e-01 2.01288342e-01
6.55677378e-01 -1.12215364e+00 8.38145316e-01 -1.21954024e+00
9.47792053e-01 4.56490189e-01 -4.49839354e-01 -1.31312263e+00
-3.96859407e-01 -6.87488168e-02 2.19020531e-01 1.22732115e+00
3.64651173e-01 -4.55037355e-01 7.08430588e-01 6.81717813e-01
-1.97650939e-02 -6.49157047e-01 -1.26351345e+00 -4.95252550e-01
7.54971981e-01 -5.62221646e-01 -2.15394959e-01 1.11527610e+00
2.11716384e-01 5.06691456e-01 -5.17514110e-01 -1.19581617e-01
5.20003021e-01 -7.53585875e-01 8.55961621e-01 -1.23082328e+00
-3.32827598e-01 -5.28044343e-01 -9.39282477e-01 2.44220167e-01
4.39939171e-01 -8.78874481e-01 -5.26908815e-01 -1.46829534e+00
2.76058227e-01 -2.80139834e-01 -3.79681140e-01 4.69866186e-01
5.17113656e-02 5.92889965e-01 6.24156371e-02 -8.87621865e-02
-2.85207301e-01 2.18702257e-01 5.03738463e-01 4.60242890e-02
1.09873608e-01 -6.56899661e-02 -7.20216990e-01 8.37380648e-01
9.29671347e-01 -5.46702981e-01 -3.97607684e-01 -1.87011898e-01
3.01513404e-01 -2.74109781e-01 1.10254996e-01 -9.28489327e-01
-1.10506617e-01 -1.80052176e-01 3.28254879e-01 -3.39099318e-01
7.00056016e-01 -5.31361222e-01 -9.54244509e-02 3.80573124e-01
-5.97569525e-01 8.50994363e-02 2.32543036e-01 3.86535116e-02
7.95316994e-02 -8.42250943e-01 9.42443609e-01 -8.91994014e-02
-3.11495870e-01 9.18650255e-02 -7.55761683e-01 1.47848159e-01
1.47650254e+00 -2.16229141e-01 -2.84123182e-01 -4.76384699e-01
-7.29987621e-01 -5.95007576e-02 -2.84806304e-02 5.24518669e-01
4.01419014e-01 -1.13703096e+00 -1.06552541e+00 4.89752591e-02
2.30437964e-01 -8.54721785e-01 4.63368386e-01 7.46886075e-01
-9.31452751e-01 2.52874851e-01 -6.57585382e-01 4.63653319e-02
-1.69580019e+00 1.75621584e-01 -4.47995327e-02 -2.74060398e-01
-2.17977390e-01 1.03900397e+00 -1.04460381e-01 -6.75448239e-01
1.23773582e-01 7.68722773e-01 -7.17006624e-01 3.69487852e-01
4.59169567e-01 6.95782840e-01 1.80483192e-01 -1.16719508e+00
-2.52330720e-01 9.57488120e-02 -5.43599069e-01 -2.74606884e-01
1.64343131e+00 2.42601082e-01 -1.48115143e-01 5.33602059e-01
1.64373922e+00 7.74796680e-02 -1.10202663e-01 3.89660716e-01
-2.52463180e-03 -3.91089976e-01 2.81805724e-01 -7.63993621e-01
-7.69459724e-01 1.02983367e+00 6.58631980e-01 3.49600405e-01
7.04573274e-01 -1.59726322e-01 7.11759865e-01 3.34555745e-01
1.56022474e-01 -1.57393968e+00 1.62900314e-01 8.43897700e-01
9.84605968e-01 -9.46546197e-01 -7.98737928e-02 1.07207462e-01
-9.81918097e-01 1.24516666e+00 8.41829956e-01 -3.50725740e-01
6.68610513e-01 2.11080220e-02 1.86217457e-01 -2.89300412e-01
-6.98618054e-01 2.75661442e-02 3.42824124e-02 6.32782519e-01
6.73290372e-01 -7.68422484e-02 -9.22807992e-01 8.39470804e-01
-5.69549024e-01 -3.83085191e-01 7.16084421e-01 7.45194197e-01
-2.80795574e-01 -1.10520887e+00 -1.46027192e-01 7.49521732e-01
-8.94372582e-01 5.37964590e-02 -7.21558154e-01 1.07325375e+00
1.51391000e-01 1.03444350e+00 1.21403351e-01 -7.94198394e-01
3.30641240e-01 1.16166435e-01 4.64674294e-01 -6.32862151e-01
-1.23704588e+00 -3.09936196e-01 5.37787139e-01 -1.04295634e-01
1.66128233e-01 -7.60033965e-01 -1.03035975e+00 -5.74193776e-01
-1.29538760e-01 2.87988633e-01 7.61327446e-01 9.71458733e-01
-2.78812557e-01 2.19550595e-01 5.78544736e-01 -5.69283664e-01
-6.16743743e-01 -1.30423570e+00 -6.90342605e-01 7.92447746e-01
-2.57886767e-01 -7.73602962e-01 -5.96468925e-01 -6.18445635e-01] | [8.822958946228027, 10.774117469787598] |
d2f2865c-48c4-4034-9e66-4e3b64b6dfc3 | cascades-of-regression-tree-fields-for-image | 1404.2086 | null | http://arxiv.org/abs/1404.2086v2 | http://arxiv.org/pdf/1404.2086v2.pdf | Cascades of Regression Tree Fields for Image Restoration | Conditional random fields (CRFs) are popular discriminative models for
computer vision and have been successfully applied in the domain of image
restoration, especially to image denoising. For image deblurring, however,
discriminative approaches have been mostly lacking. We posit two reasons for
this: First, the blur kernel is often only known at test time, requiring any
discriminative approach to cope with considerable variability. Second, given
this variability it is quite difficult to construct suitable features for
discriminative prediction. To address these challenges we first show a
connection between common half-quadratic inference for generative image priors
and Gaussian CRFs. Based on this analysis, we then propose a cascade model for
image restoration that consists of a Gaussian CRF at each stage. Each stage of
our cascade is semi-parametric, i.e. it depends on the instance-specific
parameters of the restoration problem, such as the blur kernel. We train our
model by loss minimization with synthetically generated training data. Our
experiments show that when applied to non-blind image deblurring, the proposed
approach is efficient and yields state-of-the-art restoration quality on images
corrupted with synthetic and real blur. Moreover, we demonstrate its
suitability for image denoising, where we achieve competitive results for
grayscale and color images. | ['Jeremy Jancsary', 'Uwe Schmidt', 'Stefan Roth', 'Carsten Rother', 'Sebastian Nowozin'] | 2014-04-08 | null | null | null | null | ['blind-image-deblurring'] | ['computer-vision'] | [ 3.71713281e-01 -4.07714635e-01 2.65197784e-01 -3.22634250e-01
-8.99430394e-01 -4.00615662e-01 6.88669980e-01 -2.72252619e-01
-4.06415910e-01 6.23045802e-01 1.19332217e-01 -1.09887868e-01
-2.25080386e-01 -3.88191819e-01 -7.77082622e-01 -9.88604248e-01
2.42521971e-01 2.99763650e-01 5.83587997e-02 5.34207746e-02
3.24163109e-01 3.86479020e-01 -1.42564833e+00 -4.98201400e-02
1.20218992e+00 8.00465345e-01 6.36768401e-01 6.85471475e-01
3.94961864e-01 7.78259039e-01 -3.33027810e-01 -3.17172676e-01
5.28164320e-02 -4.12217915e-01 -7.32111752e-01 5.86133242e-01
3.70143801e-01 -3.20827007e-01 -3.92270178e-01 1.36792707e+00
3.92209858e-01 2.11367637e-01 9.04671371e-01 -7.44303226e-01
-7.34854877e-01 1.95764095e-01 -4.69270110e-01 6.31403700e-02
3.07350084e-02 1.16764188e-01 7.34969795e-01 -1.02472866e+00
4.28922445e-01 1.18539691e+00 6.30090594e-01 3.86267215e-01
-1.69247723e+00 -6.32046536e-02 -1.69490635e-01 4.41034406e-01
-1.30358112e+00 -7.40499198e-01 6.65110469e-01 -6.50018811e-01
4.21903163e-01 7.71024004e-02 1.68269306e-01 1.04238188e+00
2.56907761e-01 7.01299369e-01 1.52799010e+00 -5.25220752e-01
3.49031389e-01 7.77637884e-02 3.06818839e-02 3.79474699e-01
-4.06029448e-03 8.00930411e-02 -3.17993790e-01 -6.47465587e-02
8.34028602e-01 -1.34117350e-01 -7.01090515e-01 -4.10665333e-01
-9.79752183e-01 6.92860484e-01 4.52240229e-01 4.48792338e-01
-5.42795599e-01 1.59028485e-01 3.22154127e-02 1.77780509e-01
7.22832263e-01 5.93713671e-02 -1.96461141e-01 5.19185700e-02
-1.22751451e+00 1.40377507e-01 6.18567288e-01 5.68565488e-01
9.60544109e-01 -1.34441763e-01 -2.33414441e-01 1.20110416e+00
4.57482338e-01 5.10957062e-01 2.75531113e-01 -9.36668694e-01
1.01977766e-01 -9.35257226e-02 4.66386884e-01 -9.27296758e-01
2.47101416e-03 -3.95145118e-01 -1.26092505e+00 3.81331414e-01
5.72586417e-01 2.11276531e-01 -1.01659596e+00 1.59752512e+00
-6.24929788e-03 4.66958374e-01 3.88323795e-04 1.12114882e+00
1.64159015e-01 5.13293862e-01 -7.56890653e-03 -3.51987451e-01
1.23962808e+00 -8.47394824e-01 -7.76736259e-01 -4.93991435e-01
8.84118006e-02 -1.02902603e+00 8.40221703e-01 5.88202119e-01
-9.71343815e-01 -6.59209728e-01 -6.30761623e-01 -8.36495981e-02
1.41580448e-01 3.39326024e-01 2.68739969e-01 7.19354570e-01
-1.21159589e+00 7.69668043e-01 -1.00353408e+00 -2.32559368e-01
3.32368344e-01 1.80037215e-01 -3.05433184e-01 -5.33693552e-01
-8.76420736e-01 1.09104812e+00 8.68189335e-02 5.38541257e-01
-1.03293800e+00 -2.39159092e-01 -8.08583140e-01 8.55974257e-02
1.53399378e-01 -8.91133845e-01 1.08031797e+00 -9.30156946e-01
-1.52806973e+00 6.92627728e-01 -4.75623786e-01 -4.80738997e-01
6.99048042e-01 -4.55869555e-01 -7.38885924e-02 1.23537511e-01
-2.76614167e-03 3.27669054e-01 1.49615645e+00 -1.61271882e+00
-1.81113660e-01 -2.67284542e-01 -2.24513084e-01 -4.10482241e-03
-1.12775095e-01 9.28732976e-02 -5.73807597e-01 -8.52742732e-01
1.61699995e-01 -8.90954673e-01 -5.05007267e-01 -1.61060423e-01
-4.28047806e-01 1.17595717e-01 5.80628633e-01 -1.02564824e+00
9.23689008e-01 -2.21961737e+00 4.91488665e-01 8.07053819e-02
2.42617633e-02 2.08400160e-01 -3.41713838e-02 1.28394157e-01
-3.66207324e-02 -1.84418440e-01 -7.22247005e-01 -8.98272157e-01
6.04446558e-03 3.60252082e-01 -3.34828138e-01 8.03778112e-01
2.89292634e-01 7.24083960e-01 -7.60511041e-01 -3.39844435e-01
5.62395096e-01 9.52405870e-01 -4.97815698e-01 5.33769965e-01
-1.38113454e-01 8.89897346e-01 -2.52150983e-01 3.01783711e-01
9.82195616e-01 -2.86015898e-01 4.85971160e-02 -3.80772620e-01
-5.03416620e-02 -1.47690058e-01 -1.21562910e+00 1.66301668e+00
-5.83877206e-01 5.12689292e-01 3.36446583e-01 -1.14038873e+00
7.63488889e-01 2.84311473e-01 1.86493784e-01 -2.37360328e-01
1.72126796e-02 4.61441040e-01 -3.85516196e-01 -5.10235071e-01
4.06680137e-01 -2.68506825e-01 3.96708518e-01 1.80619448e-01
1.65573403e-01 -4.38093692e-01 1.55537486e-01 2.62276968e-03
1.01125050e+00 2.72012085e-01 -1.34724965e-02 -2.90806442e-01
7.75793254e-01 -3.93516541e-01 3.50273371e-01 8.84136438e-01
4.37517688e-02 1.13025367e+00 1.74911290e-01 -1.48541912e-01
-9.63675320e-01 -8.77152205e-01 -2.66387224e-01 5.86597323e-01
2.80522376e-01 -2.00766400e-02 -9.98480618e-01 -3.68429095e-01
-3.52050245e-01 6.63700879e-01 -3.79748255e-01 1.04597777e-01
-5.90975404e-01 -1.05083799e+00 5.64091392e-02 1.90690402e-02
4.33591217e-01 -8.85983586e-01 -2.76328117e-01 3.12512487e-01
-4.22648877e-01 -1.32976937e+00 -4.77514237e-01 1.46169588e-01
-8.92074108e-01 -1.08642220e+00 -1.06566870e+00 -7.35177815e-01
7.97759950e-01 4.53182608e-01 1.18525612e+00 1.60533041e-02
-1.17840357e-01 4.77903396e-01 -2.41026625e-01 3.65560621e-01
-7.72466600e-01 -4.26189929e-01 -9.67760980e-02 6.00058258e-01
-2.42667094e-01 -6.99224770e-01 -7.81160831e-01 4.18137252e-01
-1.30725288e+00 -5.60599891e-03 7.86339223e-01 1.10687900e+00
5.97552955e-01 2.15296358e-01 1.83768108e-01 -7.19126284e-01
5.29112697e-01 -2.96690106e-01 -6.92681491e-01 3.19274783e-01
-5.83096027e-01 2.29267746e-01 6.41316831e-01 -4.00495321e-01
-1.26991546e+00 1.73962638e-01 -1.58339992e-01 -5.91218829e-01
-4.29683179e-01 4.98579413e-01 -1.03532910e-01 -2.96853065e-01
6.45082355e-01 4.02403563e-01 5.41783497e-02 -8.61086845e-01
4.63095963e-01 6.62050903e-01 8.65704536e-01 -5.02973080e-01
1.00456691e+00 4.77860570e-01 4.11430560e-02 -1.02488983e+00
-6.55239046e-01 -4.77714658e-01 -6.15713477e-01 -2.48818547e-01
8.88627350e-01 -9.83755648e-01 -4.45482999e-01 1.02663124e+00
-1.35685861e+00 -4.67579812e-01 5.50893061e-02 5.06204367e-01
-7.64688730e-01 8.58286917e-01 -7.84315825e-01 -8.29667330e-01
-4.38857675e-02 -1.49864638e+00 1.00772393e+00 1.55802056e-01
3.82763594e-01 -1.10596645e+00 -1.76634807e-02 4.72675651e-01
6.35657549e-01 -1.12059802e-01 7.68583775e-01 -2.95650531e-02
-7.14047551e-01 -2.52666585e-02 -4.52373475e-01 9.29282665e-01
2.33344257e-01 -3.73266935e-01 -1.08669627e+00 -4.66780633e-01
3.98844928e-01 -2.64490902e-01 1.19647956e+00 6.43642545e-01
1.23447108e+00 -1.88019708e-01 -1.29718959e-01 5.74832976e-01
1.57946730e+00 -4.83735144e-01 9.59601521e-01 1.89609200e-01
7.13384509e-01 4.91369665e-01 3.91623020e-01 2.82319427e-01
2.87916034e-01 9.20877874e-01 4.93619055e-01 -1.21715397e-01
-3.76940936e-01 2.06440896e-01 4.96768892e-01 6.29608154e-01
-1.69825643e-01 -2.62832671e-01 -6.41373873e-01 6.15410507e-01
-1.73772764e+00 -7.77542114e-01 -3.49246532e-01 2.35929060e+00
1.03629911e+00 -1.40048325e-01 -4.00574625e-01 8.36600177e-03
7.89700031e-01 1.62731200e-01 -2.11844280e-01 1.38776436e-01
-3.05697590e-01 2.11749911e-01 3.95892560e-01 8.66542876e-01
-1.27257133e+00 8.53142023e-01 5.52385378e+00 8.92502427e-01
-9.89228010e-01 2.10845098e-01 8.06324780e-01 5.63270748e-01
-2.20518574e-01 1.73012048e-01 -4.13186908e-01 6.40531301e-01
6.53310120e-01 2.90311128e-01 7.46432483e-01 3.89735788e-01
4.65641528e-01 -3.67028862e-01 -9.33905959e-01 1.03203988e+00
2.48816479e-02 -9.03887987e-01 -1.95412070e-01 4.89885844e-02
8.26857924e-01 5.27184941e-02 7.50604495e-02 -1.51692912e-01
1.91116005e-01 -1.06235254e+00 7.57859826e-01 8.40917528e-01
5.24117231e-01 -3.45364809e-01 7.72568107e-01 4.74570632e-01
-5.60375571e-01 1.54545277e-01 -4.15844679e-01 2.92294353e-01
4.51581150e-01 1.21219206e+00 -4.90886390e-01 6.85559094e-01
5.52479267e-01 7.02332377e-01 -5.12940824e-01 1.41338170e+00
-5.34731865e-01 7.27967262e-01 -1.22499973e-01 6.44390345e-01
-9.68178138e-02 -4.95161831e-01 6.00995183e-01 1.31221187e+00
5.37464857e-01 -2.07832366e-01 -4.12179669e-03 9.43297148e-01
9.45694968e-02 -1.84046701e-01 -1.15575254e-01 3.35492760e-01
-1.22070294e-02 1.18104613e+00 -5.67732334e-01 -9.81656685e-02
-2.87319988e-01 1.36335647e+00 1.07384190e-01 5.96052349e-01
-7.03565776e-01 1.00003049e-01 5.95272779e-01 -1.39637977e-01
6.31578088e-01 -4.50445652e-01 -1.54543906e-01 -1.48130989e+00
2.05377210e-02 -8.95708621e-01 -1.10789984e-01 -8.58828366e-01
-1.61914778e+00 7.01095700e-01 -1.42637864e-01 -1.07716310e+00
-3.21676195e-01 -5.85392058e-01 -4.32222962e-01 1.22302365e+00
-1.92650783e+00 -1.25714576e+00 -4.02255833e-01 6.55465066e-01
5.29806197e-01 2.59357780e-01 7.72249043e-01 3.82731497e-01
-4.72567141e-01 7.37909749e-02 4.54039454e-01 -1.50002971e-01
9.59590793e-01 -1.30455852e+00 3.60235155e-01 1.26569128e+00
2.23718584e-01 5.25807917e-01 1.08390629e+00 -5.11635900e-01
-1.34950769e+00 -9.37199712e-01 7.78639436e-01 -1.95264444e-01
6.33504331e-01 -1.76474437e-01 -1.18042123e+00 3.63651574e-01
2.58550763e-01 3.38171422e-01 2.00945154e-01 -1.57368943e-01
-3.09497744e-01 -2.53127236e-02 -9.67412889e-01 2.74812907e-01
5.67256451e-01 -5.96977532e-01 -2.69590110e-01 4.11300093e-01
1.37374759e-01 -3.53905678e-01 -6.44024014e-01 3.22875619e-01
1.94363788e-01 -1.26815021e+00 1.07643116e+00 -2.75174454e-02
4.11754698e-01 -6.32579386e-01 -1.35513216e-01 -1.49765289e+00
-3.59904379e-01 -6.35709047e-01 -1.06954351e-01 1.14446425e+00
3.15689459e-03 -4.87622708e-01 3.93698782e-01 3.82771224e-01
-9.71976444e-02 -2.78726608e-01 -9.02029812e-01 -6.53690875e-01
-1.02951139e-01 -4.82093692e-01 -6.91643208e-02 6.71768785e-01
-7.34992206e-01 2.01007277e-01 -8.50356340e-01 4.78504628e-01
1.00763559e+00 1.13310054e-01 6.56643152e-01 -1.07666814e+00
-6.28866911e-01 -2.84776330e-01 -2.42534623e-01 -1.58849013e+00
1.93469569e-01 -5.30949414e-01 6.31548405e-01 -1.43570864e+00
3.32303792e-01 -4.80865598e-01 -9.37097594e-02 1.86696678e-01
-4.13441509e-01 4.37080175e-01 1.70857683e-02 5.48471212e-01
-3.85087907e-01 6.42058194e-01 1.12926579e+00 -1.69213846e-01
3.72833014e-02 2.38739908e-01 -3.80375832e-01 5.73510706e-01
4.44617867e-01 -3.69546056e-01 -5.25531843e-02 -6.48769140e-01
-6.21450618e-02 8.67059752e-02 9.23243642e-01 -8.55734646e-01
3.22199494e-01 -4.01217416e-02 2.28669837e-01 -1.54673249e-01
3.31812322e-01 -8.17632973e-01 1.93218648e-01 1.57048240e-01
-4.75626439e-02 -5.21851420e-01 -1.30743280e-01 8.25757742e-01
-4.41597968e-01 -5.19521475e-01 1.05795562e+00 6.68240488e-02
-5.79137743e-01 9.90516320e-02 -3.69121253e-01 -2.29075834e-01
3.61186713e-01 7.05962703e-02 -8.54625702e-02 -5.41409433e-01
-8.05480957e-01 -1.59258828e-01 6.14309788e-01 1.46612212e-01
5.43180823e-01 -9.06497538e-01 -9.72438931e-01 1.19041257e-01
-2.06808329e-01 -1.84524488e-02 5.14367998e-01 1.15962780e+00
-3.57038617e-01 2.99430303e-02 1.08266272e-01 -7.53854215e-01
-1.11966527e+00 5.14445126e-01 3.29377919e-01 -2.31245860e-01
-4.69567299e-01 6.89402938e-01 3.08347881e-01 8.53056461e-02
8.14762861e-02 -2.56493747e-01 -1.82598203e-01 -2.92500943e-01
4.91606116e-01 2.50544697e-01 2.25610435e-01 -8.61075461e-01
-1.18500538e-01 6.34267211e-01 1.70489654e-01 -1.93013772e-01
1.52313614e+00 -5.45615971e-01 -5.58026075e-01 3.88642005e-03
1.06254959e+00 1.12416424e-01 -1.69417667e+00 -3.32726359e-01
7.05742911e-02 -7.00293005e-01 3.76824498e-01 -6.07624948e-01
-1.06211603e+00 8.61384094e-01 7.33285367e-01 3.33036602e-01
1.41530168e+00 -7.28169270e-03 4.66372252e-01 9.84168705e-03
3.49214882e-01 -5.88580251e-01 -2.95501333e-02 4.37078208e-01
9.39870477e-01 -1.34537292e+00 -8.14516470e-02 -5.13334453e-01
-3.50666314e-01 1.15364683e+00 -9.87941250e-02 -1.94634810e-01
6.36502981e-01 6.38185162e-03 7.03793317e-02 1.32589072e-01
-2.93346614e-01 -2.46538579e-01 4.62612182e-01 5.10172486e-01
2.79447675e-01 -1.30505502e-01 -2.32250005e-01 2.09564924e-01
2.65786588e-01 1.41175985e-01 4.98089463e-01 4.45434600e-01
-3.05375099e-01 -1.42787826e+00 -6.52393460e-01 -3.10017355e-02
-5.88806808e-01 -2.56927133e-01 1.68280974e-01 1.64085597e-01
8.03648159e-02 1.19565070e+00 -3.65034252e-01 -7.08727399e-03
9.91158560e-02 -1.64369941e-01 6.15441501e-01 -3.81473213e-01
-1.84699982e-01 4.48480636e-01 -2.70055264e-01 -3.87910306e-01
-7.28399277e-01 -7.33681977e-01 -5.54140985e-01 -9.19160172e-02
-4.72169638e-01 1.60288528e-01 7.91873872e-01 1.09630454e+00
1.72838554e-01 2.51483291e-01 7.09875405e-01 -1.19154644e+00
-7.25453436e-01 -1.11893415e+00 -6.93665683e-01 4.64342654e-01
5.62472939e-01 -5.97239852e-01 -5.71997166e-01 6.15762174e-01] | [11.577362060546875, -2.4904654026031494] |
2ec1f28b-8ec7-42d4-b8d7-627b26201f3e | lagr-label-aligned-graphs-for-better | null | null | https://openreview.net/forum?id=WtK3-ws0P3_ | https://openreview.net/pdf?id=WtK3-ws0P3_ | LAGr: Label Aligned Graphs for Better Systematic Generalization in Semantic Parsing | Semantic parsing is the task of producing structured meaning representations for natural language sentences.
Recent research has pointed out that the commonly-used sequence-to-sequence (seq2seq) semantic parsers struggle to generalize systematically, i.e. to handle examples that require recombining known knowledge in novel settings. In this work, we show that better systematic generalization can be achieved by producing the meaning representation directly as a graph and not as a sequence. To this end we propose LAGr (Label Aligned Graphs), a general framework to produce semantic parses by independently predicting node and edge labels for a complete multi-layer input-aligned graph. The strongly-supervised LAGr algorithm requires aligned graphs as inputs, whereas weakly-supervised LAGr infers alignments for originally unaligned target graphs using approximate maximum-a-posteriori inference. Experiments demonstrate that LAGr achieves significant improvements in systematic generalization upon the baseline seq2seq parsers in both strongly- and weakly-supervised settings. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['systematic-generalization'] | ['reasoning'] | [ 9.43816483e-01 1.09336436e+00 -8.60843062e-02 -9.66461003e-01
-1.04155016e+00 -1.08468330e+00 3.49531561e-01 2.68823117e-01
-1.68681085e-01 8.27600300e-01 2.64236093e-01 -5.38133919e-01
1.88031659e-01 -9.33831155e-01 -1.14021456e+00 -5.26370347e-01
-6.12570941e-02 6.30149484e-01 2.84304079e-02 3.54303457e-02
-1.32037243e-02 1.40979588e-01 -1.32456338e+00 5.81444740e-01
6.47799313e-01 2.51819640e-01 4.86978859e-01 9.48423088e-01
-6.98670089e-01 5.94707429e-01 -4.95640606e-01 -6.90381229e-01
2.74786577e-02 -1.10190344e+00 -1.32127964e+00 -6.77960739e-02
7.07660735e-01 2.71455169e-01 9.38281417e-02 1.21645367e+00
2.42426638e-02 1.56119689e-01 4.76833493e-01 -1.15947819e+00
-7.40586758e-01 1.12724149e+00 -1.61605135e-01 6.13963790e-02
4.85871792e-01 -5.05341142e-02 1.50750220e+00 -5.04579544e-01
8.70746613e-01 1.76798105e+00 5.53184807e-01 1.01732445e+00
-1.40393066e+00 -4.49378103e-01 5.05588293e-01 -1.30083486e-01
-8.33771050e-01 -1.07216246e-01 5.70971668e-01 -4.44425531e-02
1.46258605e+00 1.79674327e-01 1.80871770e-01 1.29351342e+00
-1.59312692e-02 7.39747345e-01 9.47717667e-01 -6.24871314e-01
3.66080910e-01 -2.81440347e-01 3.70200962e-01 1.06035745e+00
3.66156787e-01 -5.20208962e-02 -4.98072624e-01 -2.44387495e-03
4.95597690e-01 -4.51151252e-01 8.82903934e-02 -3.14643353e-01
-1.00454748e+00 8.84393036e-01 4.76371169e-01 2.60145694e-01
-1.16124332e-01 5.69216013e-01 6.60420060e-01 4.16022569e-01
4.66899961e-01 6.53170228e-01 -8.65412295e-01 -1.03809141e-01
-4.38432664e-01 2.02158093e-01 1.05046952e+00 1.36556756e+00
1.09773529e+00 -3.74128222e-02 3.69684435e-02 5.32828867e-01
4.68303300e-02 4.16825533e-01 4.37063128e-01 -9.69526112e-01
5.85426509e-01 4.54732090e-01 -4.64526266e-01 -3.98417592e-01
-5.26231289e-01 -3.68483216e-01 -4.79641050e-01 -3.14951837e-01
4.32099670e-01 -3.65448236e-01 -9.67343509e-01 2.36096883e+00
2.20631614e-01 3.82670969e-01 6.95889175e-01 5.71743488e-01
8.54310513e-01 6.40965998e-01 7.01599598e-01 -1.00368299e-01
1.21655548e+00 -1.10657609e+00 -4.58093047e-01 -9.17056799e-01
1.18061578e+00 -2.05512360e-01 1.30138695e+00 4.05371450e-02
-9.02135670e-01 -4.88010496e-01 -7.26527512e-01 -2.16997460e-01
-4.00499016e-01 -3.54353815e-01 7.95343161e-01 5.89379787e-01
-1.23528993e+00 7.12640464e-01 -8.16591501e-01 -5.45384109e-01
4.57656950e-01 2.12979972e-01 -4.77538794e-01 -2.86742955e-01
-9.96164024e-01 7.31102943e-01 9.57626820e-01 -1.41889766e-01
-7.11144149e-01 -5.98661184e-01 -1.41796505e+00 2.61184931e-01
5.83505213e-01 -1.07950675e+00 1.61011600e+00 -1.27171624e+00
-1.31760633e+00 1.13571429e+00 -4.63573128e-01 -6.14917397e-01
-9.08971652e-02 -2.64096648e-01 3.22247520e-02 1.74368367e-01
3.06244701e-01 1.02245867e+00 4.49124098e-01 -1.21385920e+00
-5.34559608e-01 -6.28788412e-01 1.84992194e-01 1.94197938e-01
2.02146694e-01 -1.53817892e-01 -4.25450020e-02 -6.30933285e-01
8.81391987e-02 -9.56535339e-01 -5.78601062e-01 -3.53606462e-01
-4.64856833e-01 -4.18603092e-01 5.36571920e-01 -6.03909671e-01
7.04031765e-01 -1.80215991e+00 3.70395660e-01 -1.52418196e-01
-5.74748032e-02 1.86828792e-01 -5.02475560e-01 4.70238447e-01
-2.14431241e-01 2.52184540e-01 -7.83952951e-01 -3.33296269e-01
3.38678733e-02 9.56937969e-01 -3.49149197e-01 -9.80287790e-02
5.66263258e-01 1.33111227e+00 -1.60647833e+00 -3.16543907e-01
-1.30686566e-01 4.73805554e-02 -5.53680420e-01 2.71660209e-01
-7.44929016e-01 4.32437658e-01 -5.50883055e-01 3.30187529e-01
4.31552649e-01 -4.22579497e-01 7.52834857e-01 2.35088751e-01
5.20521343e-01 5.40398180e-01 -5.84113657e-01 2.42008686e+00
-8.56727123e-01 4.07337904e-01 -1.26860037e-01 -1.44827318e+00
1.06381381e+00 3.00309658e-02 -3.38141054e-01 -2.71866977e-01
-2.70068496e-01 3.07124317e-01 -2.88596064e-01 -4.58621144e-01
2.56408840e-01 -7.12109804e-01 -5.81362724e-01 2.82973886e-01
4.94404197e-01 -3.06730300e-01 9.74231437e-02 3.59899521e-01
1.31175840e+00 6.56807661e-01 4.58441168e-01 -3.22318912e-01
4.31511849e-01 2.35831663e-01 4.77322966e-01 9.14488018e-01
2.35694125e-01 3.28239411e-01 7.23616898e-01 -3.97240251e-01
-9.46431220e-01 -1.43869936e+00 4.48966265e-01 1.33316100e+00
1.02583617e-01 -4.83457714e-01 -9.96416628e-01 -1.47688293e+00
-1.25680789e-01 1.26169264e+00 -2.93648213e-01 -2.96087891e-01
-8.93732548e-01 -3.57703924e-01 6.37722969e-01 9.00259674e-01
1.13267481e-01 -1.44352090e+00 -5.08015811e-01 3.55778545e-01
7.69146904e-02 -1.36085618e+00 -1.53986350e-01 6.21209800e-01
-1.09735250e+00 -1.03076088e+00 -2.54523247e-01 -1.28384292e+00
9.01020288e-01 -1.01801135e-01 1.37493706e+00 -1.10086732e-01
-1.31383702e-01 2.54160762e-01 -4.40004259e-01 -1.11822307e-01
-9.67761517e-01 2.89325237e-01 -3.32631558e-01 -4.05316621e-01
4.57073778e-01 -6.84757769e-01 9.84309018e-02 -3.35754782e-01
-9.99552250e-01 4.16165918e-01 7.12801933e-01 9.45818603e-01
5.37692428e-01 -3.12960595e-01 1.08957934e+00 -1.56425607e+00
5.61900556e-01 -4.84687686e-01 -4.43400830e-01 5.35266042e-01
-4.24221486e-01 8.15680563e-01 1.09971249e+00 2.71390211e-02
-1.35744154e+00 3.96021962e-01 -2.89091617e-01 1.17333628e-01
-6.94593430e-01 5.19998729e-01 -4.04970139e-01 4.16687250e-01
6.80058002e-01 2.50452727e-01 -1.62982255e-01 -5.12757242e-01
9.61472094e-01 5.91577947e-01 9.40217197e-01 -1.05731797e+00
5.30103147e-01 1.77945316e-01 3.24183583e-01 -7.42183685e-01
-1.36493456e+00 -3.27986240e-01 -9.38702106e-01 3.38124841e-01
1.06704080e+00 -7.09877670e-01 -2.13158011e-01 8.18006024e-02
-1.45079458e+00 -6.31135881e-01 -4.06091720e-01 1.08134158e-01
-8.28442037e-01 4.82317835e-01 -3.79085809e-01 -6.46271467e-01
-2.98364192e-01 -7.35652268e-01 1.53935397e+00 1.76801488e-01
-4.97699112e-01 -1.36419773e+00 -3.57843414e-02 2.37163827e-01
-3.88887972e-02 3.88322383e-01 1.26359940e+00 -1.12077963e+00
-4.29256678e-01 2.81510442e-01 -2.59024769e-01 3.87327224e-01
2.97526032e-01 -7.18264401e-01 -9.91764605e-01 -1.51746988e-01
-3.46453995e-01 -4.94915485e-01 8.40556145e-01 1.18863843e-02
1.16801202e+00 -4.90102202e-01 -4.66925830e-01 4.76206779e-01
1.29338109e+00 -4.17188890e-02 2.41797671e-01 -1.95619580e-03
7.97754467e-01 8.84140015e-01 4.72391278e-01 -1.29049435e-01
3.18753272e-01 2.73069888e-01 2.16212869e-01 1.51609540e-01
-3.06481749e-01 -9.07785118e-01 3.82321835e-01 6.09127343e-01
5.30838132e-01 -4.39377934e-01 -8.90577137e-01 6.68372214e-01
-1.84773636e+00 -5.69156051e-01 -7.45053813e-02 1.74542367e+00
9.71103966e-01 1.35822639e-01 -2.51402259e-01 -2.94008970e-01
8.60792577e-01 8.32366720e-02 -6.37031138e-01 -7.23950863e-01
-1.24388278e-01 7.67603636e-01 5.09829581e-01 7.87173748e-01
-8.83549929e-01 1.63027990e+00 5.96454620e+00 5.81035078e-01
-5.42624712e-01 1.08977787e-01 4.79249150e-01 3.02376539e-01
-7.48897016e-01 4.27107245e-01 -8.30426872e-01 4.90250252e-02
1.19487548e+00 -1.68048620e-01 3.62005860e-01 9.47014809e-01
-2.87466764e-01 1.94507435e-01 -1.48203480e+00 7.04547703e-01
6.74560890e-02 -1.25803375e+00 4.06734079e-01 -4.20452744e-01
5.67838252e-01 -1.17137589e-01 -4.61882025e-01 4.96672630e-01
7.69360363e-01 -1.11538708e+00 3.62939745e-01 1.35720387e-01
7.62157738e-01 -4.97056723e-01 6.19176090e-01 4.12383199e-01
-1.15056980e+00 5.22069931e-02 -3.27597618e-01 -3.24609846e-01
4.62391824e-01 3.22654098e-01 -1.07572174e+00 6.94309056e-01
3.01501304e-01 9.61703122e-01 -3.81277084e-01 3.06212634e-01
-1.13146782e+00 7.27716148e-01 -1.42244756e-01 -2.59650975e-01
4.63883877e-01 -6.08914830e-02 4.70184654e-01 1.63753998e+00
1.63209274e-01 1.73057154e-01 3.08395952e-01 9.82188523e-01
-2.49928787e-01 -1.00222290e-01 -7.81991541e-01 -3.22012424e-01
3.28354836e-01 1.05484855e+00 -7.80916750e-01 -6.20994031e-01
-3.89663965e-01 1.27083147e+00 7.20211685e-01 3.17516714e-01
-3.15021276e-01 -3.67139190e-01 5.62928915e-01 -3.63140911e-01
3.88755798e-01 -1.96758255e-01 -2.72279292e-01 -1.08400679e+00
-6.68817982e-02 -6.01280868e-01 7.72915363e-01 -9.69365776e-01
-1.41262245e+00 4.97992992e-01 2.73675650e-01 -2.38937065e-01
-5.67374408e-01 -8.67494643e-01 -4.81124252e-01 8.60346794e-01
-1.50567067e+00 -1.30180395e+00 -5.44390418e-02 5.92183284e-02
1.00393343e+00 3.39558683e-02 1.14534545e+00 -5.14595747e-01
-1.30149975e-01 4.92428869e-01 -4.19800431e-01 -9.83279664e-03
2.61505574e-01 -1.61339486e+00 1.25423777e+00 9.52776134e-01
6.12439334e-01 5.27347147e-01 7.53753841e-01 -7.83875287e-01
-1.27201259e+00 -1.31754744e+00 1.28972995e+00 -4.42611307e-01
5.92541933e-01 -6.96377695e-01 -1.11974406e+00 1.16241777e+00
9.05250907e-02 3.21760029e-02 6.05199099e-01 1.98468208e-01
-6.47458434e-01 3.91261667e-01 -1.09793556e+00 4.85255390e-01
1.86734533e+00 -7.67405510e-01 -1.12713504e+00 4.72387820e-01
1.20948684e+00 -3.85866135e-01 -5.84208310e-01 2.72873014e-01
9.13100839e-02 -5.89675725e-01 7.12109268e-01 -1.26389229e+00
4.32963371e-01 2.67807650e-03 -2.92849332e-01 -1.47978508e+00
-3.25044952e-02 -7.89411008e-01 3.57425809e-02 1.02515233e+00
6.81516647e-01 -7.07303226e-01 1.01840627e+00 2.17355847e-01
-5.44018924e-01 -6.13692582e-01 -7.59772480e-01 -1.07713592e+00
3.16677362e-01 -4.36244160e-01 6.41291857e-01 7.69506156e-01
2.67465115e-01 9.89356995e-01 1.91078365e-01 4.04044390e-01
6.77263975e-01 4.01328325e-01 5.57837069e-01 -1.12040472e+00
-5.63231587e-01 -2.11136520e-01 -4.93457705e-01 -1.34980392e+00
1.08902586e+00 -1.45622981e+00 3.10155600e-01 -1.65175152e+00
2.54533798e-01 -2.45734021e-01 -2.03896165e-02 8.38828623e-01
-4.18212265e-01 -2.04199627e-01 -5.10041714e-02 -3.79121065e-01
-7.24953711e-01 3.26846510e-01 1.12680078e+00 1.09937834e-02
5.17445952e-02 -2.69626051e-01 -9.65663254e-01 6.84802353e-01
9.06980038e-01 -7.95198500e-01 -7.46197879e-01 -5.70413589e-01
4.19160247e-01 1.57549858e-01 3.49379778e-01 -4.18932348e-01
-8.69191885e-02 -1.17973387e-01 -2.25354031e-01 -1.52683407e-01
-1.60384979e-02 -3.98397803e-01 -1.47847518e-01 3.89801174e-01
-8.16448987e-01 2.94433814e-02 2.69157678e-01 8.14926922e-01
-1.39929995e-01 -5.90538323e-01 4.62781638e-01 -5.68997860e-01
-9.69420433e-01 -5.20434529e-02 -4.72829379e-02 5.93709290e-01
8.08328807e-01 2.84270812e-02 -3.85209322e-01 -3.30802649e-01
-1.00034714e+00 2.19931960e-01 3.64123583e-01 5.36447227e-01
4.79780465e-01 -8.72596562e-01 -6.52965784e-01 4.87192273e-02
1.75368503e-01 4.60718632e-01 -5.53205051e-02 4.79059331e-02
-3.19882482e-01 4.28125262e-01 7.49886036e-02 -5.04194796e-01
-1.34521246e+00 6.06778562e-01 2.17326283e-01 -2.60656565e-01
-6.92360401e-01 1.10707891e+00 6.63456917e-01 -1.03118443e+00
2.20156591e-02 -6.19311929e-01 2.74390906e-01 -6.50992811e-01
1.96807623e-01 3.43743265e-02 -1.21801533e-01 -5.14247596e-01
-3.79635423e-01 4.95254278e-01 3.66998352e-02 1.89294610e-02
1.42714119e+00 -7.44139627e-02 -1.42616823e-01 3.47863942e-01
1.42585599e+00 -3.36149454e-01 -1.17707121e+00 -2.72595167e-01
6.04304254e-01 -5.42161725e-02 -4.14457947e-01 -8.06379259e-01
-6.14507794e-01 8.82452965e-01 -2.06885830e-01 5.67108542e-02
8.58987987e-01 7.08568811e-01 1.02432775e+00 7.72541940e-01
3.86655569e-01 -5.55318356e-01 -5.59679084e-02 6.15933597e-01
5.45938909e-01 -9.56006348e-01 -4.15949821e-01 -9.02195454e-01
-6.21336877e-01 1.16383529e+00 5.10066390e-01 -4.85661142e-02
5.65459169e-02 2.39535049e-01 -4.48750108e-02 -2.63614655e-01
-7.14718521e-01 -2.37376735e-01 2.01268774e-02 8.70151997e-01
3.98941517e-01 2.22066030e-01 -8.42907056e-02 8.27169418e-01
-4.54495907e-01 -2.85110831e-01 4.46119577e-01 1.04776752e+00
-5.48492134e-01 -1.47730637e+00 1.15680754e-01 1.96085885e-01
-3.38856488e-01 -3.12563896e-01 -7.81092823e-01 6.32620037e-01
-1.05319224e-01 8.90148163e-01 3.58050615e-02 -1.16502494e-03
1.98713705e-01 5.23534000e-01 8.34418476e-01 -1.28371966e+00
-4.44726229e-01 -4.30765539e-01 6.56366229e-01 -5.42238176e-01
-4.38559383e-01 -6.33879960e-01 -1.95220375e+00 5.13389170e-01
7.80809438e-03 3.57612967e-01 5.95080912e-01 1.16981351e+00
3.68667036e-01 5.71674526e-01 2.01818153e-01 -4.22484964e-01
-7.70203292e-01 -7.59673595e-01 -2.81572312e-01 6.45965993e-01
2.07193986e-01 -1.60704032e-01 -3.59385490e-01 2.98985213e-01] | [10.512675285339355, 9.096501350402832] |
8631d0b6-b315-4cc8-a9be-d89973c43b1d | toward-a-unified-framework-for-unsupervised | 2212.10097 | null | https://arxiv.org/abs/2212.10097v1 | https://arxiv.org/pdf/2212.10097v1.pdf | Toward a Unified Framework for Unsupervised Complex Tabular Reasoning | Structured tabular data exist across nearly all fields. Reasoning task over these data aims to answer questions or determine the truthiness of hypothesis sentences by understanding the semantic meaning of a table. While previous works have devoted significant efforts to the tabular reasoning task, they always assume there are sufficient labeled data. However, constructing reasoning samples over tables (and related text) is labor-intensive, especially when the reasoning process is complex. When labeled data is insufficient, the performance of models will suffer an unendurable decline. In this paper, we propose a unified framework for unsupervised complex tabular reasoning (UCTR), which generates sufficient and diverse synthetic data with complex logic for tabular reasoning tasks, assuming no human-annotated data at all. We first utilize a random sampling strategy to collect diverse programs of different types and execute them on tables based on a "Program-Executor" module. To bridge the gap between the programs and natural language sentences, we design a powerful "NL-Generator" module to generate natural language sentences with complex logic from these programs. Since a table often occurs with its surrounding texts, we further propose novel "Table-to-Text" and "Text-to-Table" operators to handle joint table-text reasoning scenarios. This way, we can adequately exploit the unlabeled table resources to obtain a well-performed reasoning model under an unsupervised setting. Our experiments cover different tasks (question answering and fact verification) and different domains (general and specific), showing that our unsupervised methods can achieve at most 93% performance compared to supervised models. We also find that it can substantially boost the supervised performance in low-resourced domains as a data augmentation technique. Our code is available at https://github.com/leezythu/UCTR. | ['Jianyong Wang', 'Ning Liu', 'Bowen Dong', 'Zhichao Duan', 'Xiuxing Li', 'Zhenyu Li'] | 2022-12-20 | null | null | null | null | ['fact-verification'] | ['natural-language-processing'] | [ 2.06451282e-01 4.93779957e-01 -4.66037571e-01 -5.57765663e-01
-1.07795060e+00 -7.87919641e-01 5.34574747e-01 1.24375597e-01
2.68067658e-01 7.60658920e-01 6.23402596e-02 -8.13677132e-01
2.90693104e-01 -1.27404034e+00 -1.09421742e+00 -2.74817087e-02
4.41956282e-01 7.23484814e-01 2.60937810e-01 -3.46890271e-01
-9.14627407e-03 -9.57110822e-02 -1.56597126e+00 8.86217773e-01
1.17279804e+00 6.54708683e-01 -1.20384090e-01 3.79706204e-01
-6.49188280e-01 1.53628671e+00 -5.96533418e-01 -9.83151555e-01
2.51336753e-01 -5.47812164e-01 -9.93315995e-01 2.22036064e-01
3.39017451e-01 -1.35654464e-01 -1.60411105e-01 1.25955319e+00
-1.15477793e-01 -2.60089278e-01 3.98718268e-01 -1.42997026e+00
-4.92797971e-01 1.35748434e+00 -3.23424995e-01 -2.94703335e-01
7.79286206e-01 1.66434243e-01 1.39508915e+00 -7.12626338e-01
7.46896982e-01 1.39435422e+00 4.54194814e-01 5.72100878e-01
-1.25233364e+00 -5.01907527e-01 8.35989341e-02 1.04579583e-01
-1.07046664e+00 -3.23688596e-01 8.24481606e-01 -3.96049500e-01
7.08664894e-01 5.31518281e-01 2.79276580e-01 9.99159157e-01
-3.62014063e-02 1.14447069e+00 8.59894216e-01 -5.62745810e-01
1.43102437e-01 5.36018550e-01 3.84509861e-01 9.40341711e-01
5.84294140e-01 -6.29675269e-01 -3.16737115e-01 -2.45457053e-01
2.34449521e-01 -3.82448137e-02 -1.63857996e-01 -4.28923845e-01
-1.27219391e+00 8.84999573e-01 -2.37620510e-02 1.85453430e-01
3.72871123e-02 -8.11198056e-02 6.80839062e-01 5.22437215e-01
9.20843557e-02 6.08885467e-01 -7.82190263e-01 -3.84250544e-02
-6.37433946e-01 5.93092442e-01 1.35405374e+00 1.45002341e+00
7.71909773e-01 -2.69077510e-01 -2.14742362e-01 7.70155430e-01
7.77110755e-02 7.01315343e-01 2.57021159e-01 -8.76022935e-01
1.13215971e+00 1.18360972e+00 5.99359088e-02 -9.90676224e-01
-4.42278311e-02 9.76632684e-02 -6.82526708e-01 -2.04648361e-01
6.49991632e-01 -1.21215917e-01 -5.17751753e-01 1.72901487e+00
3.43186021e-01 -6.62210941e-01 3.96192223e-01 5.31340182e-01
8.79507959e-01 5.78302324e-01 -1.78210109e-01 -2.35975236e-02
1.77216971e+00 -1.03119230e+00 -8.84220064e-01 -4.75998551e-01
9.90130305e-01 -3.44096214e-01 1.77228940e+00 4.10227418e-01
-9.49940503e-01 -3.16386253e-01 -9.71397996e-01 -2.80483425e-01
-5.78321993e-01 2.89907962e-01 8.62040699e-01 7.10847318e-01
-4.67742831e-01 7.01925084e-02 -6.60315812e-01 -5.83045632e-02
4.37811613e-01 3.99808921e-02 -2.86889672e-01 -4.33691502e-01
-1.37027431e+00 5.60027003e-01 5.17639697e-01 -7.97650442e-02
-6.01959586e-01 -5.77647209e-01 -1.27248287e+00 1.59018803e-02
1.20970070e+00 -6.41801476e-01 1.47909582e+00 -5.64250171e-01
-1.08760500e+00 8.02849174e-01 -3.73936027e-01 -5.83056748e-01
6.29447877e-01 -2.25385558e-02 -2.75183052e-01 -8.30110535e-02
4.24284250e-01 2.97472477e-01 2.61769176e-01 -1.21186399e+00
-3.31118226e-01 -2.78334111e-01 5.92392325e-01 -1.38747662e-01
-1.83454007e-01 1.17251398e-02 -5.85193217e-01 -4.75161582e-01
2.64078587e-01 -7.81570017e-01 -1.55057296e-01 -2.01055840e-01
-9.42574084e-01 -2.99559742e-01 6.07915401e-01 -5.11150599e-01
1.31994593e+00 -1.90149474e+00 -2.05769092e-01 3.08424048e-02
2.89288044e-01 1.18376483e-04 2.61048943e-01 4.79061544e-01
9.94157419e-03 4.26711947e-01 -4.75892127e-01 2.19627358e-02
3.82408828e-01 3.47904980e-01 -6.81307793e-01 -2.18146682e-01
4.82137799e-01 1.07611954e+00 -7.22987890e-01 -8.32841992e-01
-1.55891389e-01 -3.68195087e-01 -8.80641162e-01 1.66080728e-01
-1.10290825e+00 1.21494960e-02 -6.90579534e-01 9.96029139e-01
6.01617694e-01 -6.37530863e-01 5.47720611e-01 -2.33987230e-03
3.76896799e-01 6.92200959e-01 -1.24106288e+00 1.48773324e+00
-4.88828987e-01 3.91128570e-01 -2.86826968e-01 -9.09085155e-01
9.35188711e-01 1.14115015e-01 7.80700799e-03 -3.96365345e-01
5.17219231e-02 1.53037310e-01 -2.47797407e-02 -8.92506719e-01
2.70267576e-01 -2.16781318e-01 -5.36242008e-01 4.95316058e-01
-1.91446364e-01 -5.76994896e-01 8.21445882e-01 5.39634168e-01
1.42700958e+00 1.23472333e-01 2.45851472e-01 -1.42735422e-01
8.90323818e-01 6.58866346e-01 6.04672074e-01 7.22253442e-01
1.76360667e-01 3.38886768e-01 1.36795127e+00 -4.42440867e-01
-1.03852642e+00 -7.41773486e-01 -5.80535531e-02 8.34577441e-01
-4.77407873e-02 -8.07240784e-01 -7.59771705e-01 -1.10781991e+00
5.78795634e-02 1.04624021e+00 -5.34965336e-01 -5.21509117e-03
-4.94534940e-01 -5.58180630e-01 7.26034045e-01 4.60820347e-01
6.78327262e-01 -1.22366369e+00 -1.87125161e-01 -8.82614702e-02
-6.35206223e-01 -1.43527532e+00 -9.44834054e-02 2.46788412e-01
-5.98919988e-01 -1.35131729e+00 2.36450046e-01 -6.54824138e-01
8.34172964e-01 -3.51552330e-02 1.48365712e+00 -7.60085322e-03
1.97224505e-02 -2.67793965e-02 -4.00489897e-01 -3.95307750e-01
-8.62105787e-01 1.37705375e-02 -3.06604713e-01 -2.89743006e-01
6.80413663e-01 -2.78008789e-01 1.30847663e-01 2.84655124e-01
-1.18122804e+00 3.49442810e-01 4.16430652e-01 8.65111470e-01
4.19011354e-01 5.26113570e-01 3.77523899e-01 -1.78819346e+00
6.05696917e-01 -4.36442137e-01 -7.35963523e-01 5.25953770e-01
-4.39568669e-01 5.13831019e-01 1.24527609e+00 -3.44704211e-01
-1.02031457e+00 -1.21858552e-01 1.05238892e-01 -2.90740222e-01
-1.12596154e-01 7.23759949e-01 -7.25432336e-01 5.88848174e-01
8.67773294e-01 2.48493612e-01 -1.06066018e-01 -7.25435093e-02
3.70046943e-01 4.67787772e-01 3.49533707e-01 -1.25922942e+00
1.24462247e+00 3.67247760e-01 -2.01231852e-01 -1.18329391e-01
-1.20098448e+00 -5.64372577e-02 -4.68432158e-01 1.96140334e-01
5.74490249e-01 -8.78020644e-01 -8.63941431e-01 -3.89256589e-02
-1.09035742e+00 -5.30823886e-01 -2.49962658e-01 7.02680871e-02
-4.69161570e-01 3.27219903e-01 -5.26011467e-01 -8.83683503e-01
-1.69788197e-01 -1.24774158e+00 9.85076904e-01 -1.91449672e-01
-3.27368230e-01 -8.68860602e-01 -1.13649622e-01 8.31358016e-01
-1.23935930e-01 2.17288032e-01 1.34669745e+00 -9.70687091e-01
-8.09247851e-01 -4.18005466e-01 -2.55498558e-01 3.61362368e-01
1.84451848e-01 9.44040418e-02 -8.40776682e-01 3.41361076e-01
1.86601952e-01 -7.83557832e-01 4.47846383e-01 -3.54881257e-01
1.49182534e+00 -6.97773516e-01 -5.25820926e-02 1.37110516e-01
1.24455905e+00 9.02149826e-02 6.73735142e-01 2.26895481e-01
6.52762711e-01 8.13331604e-01 8.10026050e-01 2.50947863e-01
7.18042016e-01 2.23847002e-01 -1.60772272e-03 1.56854331e-01
2.06195146e-01 -7.01311469e-01 3.34418327e-01 5.69117844e-01
5.09805322e-01 -2.31004700e-01 -1.30538249e+00 5.36979914e-01
-1.83566725e+00 -1.08352005e+00 -2.90738285e-01 1.90465224e+00
1.35486674e+00 5.85705221e-01 -1.18422776e-01 4.33824807e-01
5.21732390e-01 -4.14800942e-02 -5.23228645e-01 -2.06484392e-01
-1.00065462e-01 -6.16828026e-03 1.35811806e-01 4.33461308e-01
-9.57456470e-01 9.23883557e-01 5.20083332e+00 7.97267973e-01
-6.69268370e-01 -2.05263823e-01 6.92978978e-01 2.93536216e-01
-7.97093689e-01 2.47303650e-01 -8.28032613e-01 2.69728303e-01
8.68951678e-01 -3.20935279e-01 4.27762210e-01 1.26449680e+00
-3.07823415e-03 -9.82458293e-02 -1.40675426e+00 6.97504282e-01
5.25390320e-02 -1.29357123e+00 3.75518709e-01 -2.99348831e-01
5.56352615e-01 -4.96861279e-01 -2.20520318e-01 8.82953703e-01
5.17453253e-01 -8.62501085e-01 7.54509747e-01 1.58604220e-01
7.12644517e-01 -4.38979924e-01 7.61460721e-01 7.14874089e-01
-1.04465652e+00 -1.03417851e-01 -3.20181817e-01 -8.73259753e-02
-3.15392345e-01 8.93863440e-01 -1.01709318e+00 6.70338929e-01
4.58372831e-01 5.70739925e-01 -8.02363753e-01 2.12941155e-01
-5.33025980e-01 4.84229118e-01 -1.64322183e-01 -3.47748309e-01
-2.94723380e-02 -1.28224030e-01 3.79906297e-02 1.08677769e+00
-1.14956878e-01 1.13013394e-01 2.74195522e-01 1.42604828e+00
-4.10759568e-01 1.12173714e-01 -9.01594877e-01 -3.31304371e-01
3.27380806e-01 1.19192338e+00 -6.95619226e-01 -7.98199713e-01
-6.74858093e-01 3.42530727e-01 3.46035779e-01 2.37906262e-01
-1.03803360e+00 -4.06895012e-01 2.16718435e-01 2.20386192e-01
7.67306983e-02 8.50550309e-02 -7.81044543e-01 -1.67738307e+00
7.03957558e-01 -1.52137995e+00 4.98686254e-01 -9.16578174e-01
-1.21030998e+00 5.82851827e-01 1.61829412e-01 -1.23155165e+00
-3.40632170e-01 -6.74302757e-01 -3.06433052e-01 5.93656898e-01
-1.18152165e+00 -9.37137663e-01 -3.36283982e-01 6.03678644e-01
6.09999657e-01 -1.12074703e-01 7.20519185e-01 1.76949680e-01
-6.25805020e-01 6.49012387e-01 -4.48709965e-01 6.09650910e-01
5.69106817e-01 -1.43878412e+00 4.59165871e-01 9.75738466e-01
9.96533409e-02 1.10551178e+00 7.20958948e-01 -7.29312479e-01
-1.98246634e+00 -1.18874049e+00 1.01650357e+00 -9.51288521e-01
1.02796459e+00 -9.61478114e-01 -1.03087425e+00 1.01762056e+00
-6.60896227e-02 -1.23289891e-01 5.39059460e-01 2.56318867e-01
-8.01787913e-01 -1.24879941e-01 -1.18269515e+00 9.88378882e-01
8.62221420e-01 -8.10760438e-01 -7.83311605e-01 6.84265971e-01
9.92266655e-01 -5.39083242e-01 -7.23812044e-01 5.13824642e-01
2.01246426e-01 -8.49824011e-01 5.38532674e-01 -7.32879817e-01
1.15484512e+00 -5.57826161e-01 -3.57923597e-01 -6.34360254e-01
3.53383303e-01 -5.26445568e-01 -2.02096775e-01 1.29679060e+00
9.51782763e-01 -5.77727318e-01 9.42905188e-01 1.05862820e+00
5.08100651e-02 -7.64869034e-01 -3.52216840e-01 -5.39286375e-01
2.81094182e-02 -8.11600089e-01 7.37978518e-01 1.01437044e+00
5.91031671e-01 6.07913017e-01 4.91427742e-02 1.54186949e-01
4.18856651e-01 5.99821150e-01 1.20014620e+00 -9.31135893e-01
-4.68430102e-01 -1.34884715e-01 1.98643412e-02 -9.17163610e-01
4.91692424e-01 -1.04267931e+00 1.03466794e-01 -1.48001027e+00
4.16859090e-01 -4.12573665e-01 4.22589183e-01 8.39059472e-01
-3.05430025e-01 -3.38998169e-01 -4.62832600e-02 -2.11949367e-02
-6.71075165e-01 3.59093398e-01 1.40413427e+00 -4.12965059e-01
5.35496064e-02 8.25573578e-02 -1.03755641e+00 6.60610616e-01
6.91663682e-01 -3.98863673e-01 -7.48888254e-01 -1.88015252e-01
5.81089258e-01 4.15718496e-01 2.66598135e-01 -9.61441994e-01
2.04641476e-01 -2.90268749e-01 -1.39825657e-01 -4.79119450e-01
-6.25490397e-02 -8.13793600e-01 -5.94213381e-02 3.31492543e-01
-6.58124149e-01 4.53033485e-02 2.75463760e-01 2.70056576e-01
-4.68882889e-01 -3.57433587e-01 4.46260124e-01 -3.53561640e-01
-2.81274408e-01 3.98011729e-02 -2.41406634e-01 5.40726423e-01
7.25628555e-01 2.89027363e-01 -6.06612086e-01 -3.63934487e-01
-4.52523768e-01 3.94628614e-01 4.80317771e-01 2.24842787e-01
3.91763180e-01 -1.11995673e+00 -5.81594944e-01 3.30905944e-01
5.31958103e-01 4.07805473e-01 -1.06385782e-01 6.81472838e-01
-6.26063883e-01 5.28762639e-01 2.46369377e-01 -3.86652619e-01
-8.74771774e-01 9.59259212e-01 1.46940246e-01 -5.91083944e-01
-3.97708297e-01 4.77540344e-01 1.83234617e-01 -1.01696181e+00
1.59497723e-01 -8.69907439e-01 1.23586826e-01 -2.72352904e-01
4.26347196e-01 -2.74271011e-01 1.69942766e-01 1.85687855e-01
-1.50994167e-01 2.56422441e-02 -1.94694430e-01 7.66454190e-02
1.05003560e+00 3.16230029e-01 -4.43877995e-01 6.78457618e-01
7.28155255e-01 4.41670179e-01 -5.60053468e-01 -3.19493055e-01
2.75037885e-01 -3.86388958e-01 -6.14501536e-01 -8.14385235e-01
-7.09435046e-01 7.29238808e-01 -5.57655573e-01 5.67783117e-01
9.27144170e-01 1.39938623e-01 5.87401748e-01 9.86163497e-01
5.74578822e-01 -6.28964841e-01 4.20827977e-02 6.30548716e-01
7.78399825e-01 -1.48680639e+00 -1.08720645e-01 -9.48876262e-01
-8.31092894e-01 9.45048690e-01 9.51600790e-01 2.55869001e-01
2.30398133e-01 6.99785054e-01 1.36212241e-02 -2.56609887e-01
-1.22698021e+00 -5.38685098e-02 -8.71604681e-02 2.41761819e-01
5.71542084e-01 -2.95986626e-02 -1.00594327e-01 8.41780782e-01
-5.67648888e-01 -8.73399340e-03 7.10022032e-01 9.30080652e-01
-2.35178962e-01 -1.32896185e+00 -4.78402466e-01 6.15475655e-01
-3.42145026e-01 -3.09069216e-01 -4.86131459e-01 9.92001116e-01
-1.25706509e-01 9.50252295e-01 -3.87478888e-01 -4.03742373e-01
4.06279922e-01 3.84823352e-01 2.57583231e-01 -9.11108494e-01
-4.05486882e-01 -3.34064931e-01 5.59711397e-01 -3.30184847e-01
-1.56644017e-01 -6.75855577e-01 -1.40035558e+00 -5.32803655e-01
-1.21394835e-01 4.09581870e-01 1.31912082e-01 9.23548400e-01
2.97356653e-03 5.99091947e-01 4.35434878e-01 2.07189202e-01
-6.45495832e-01 -9.05134082e-01 -3.51707429e-01 5.78282356e-01
1.11217070e-02 -3.82301629e-01 -3.04057360e-01 3.15202743e-01] | [9.578991889953613, 7.742423057556152] |
3b4016b9-92a7-4029-b449-de99aa6da478 | handling-motion-blur-in-multi-frame-super | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Ma_Handling_Motion_Blur_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Ma_Handling_Motion_Blur_2015_CVPR_paper.pdf | Handling Motion Blur in Multi-Frame Super-Resolution | Ubiquitous motion blur easily fails multi-frame super-resolution (MFSR). Our method proposed in this paper tackles this issue by optimally searching least blurred pixels in MFSR. An EM framework is proposed to guide residual blur estimation and high-resolution image reconstruction. To suppress noise, we employ a family of sparse penalties as natural image priors, along with an effective solver. Theoretical analysis is performed on how and when our method works. The relationship between estimation errors of motion blur and the quality of input images is discussed. Our method produces sharp and higher-resolution results given input of challenging low-resolution noisy and blurred sequences. | ['Ziyang Ma', 'Xin Tao', 'Li Xu', 'Jiaya Jia', 'Renjie Liao', 'Enhua Wu'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['multi-frame-super-resolution'] | ['computer-vision'] | [ 4.50918734e-01 -4.14820462e-01 2.25764796e-01 -8.13772678e-02
-6.92455888e-01 -1.47790387e-01 2.57105708e-01 -9.51757967e-01
-3.13467026e-01 1.06127787e+00 5.09141207e-01 2.28637233e-01
-2.58521169e-01 -3.02988619e-01 -5.30420661e-01 -6.66713059e-01
2.91369915e-01 -3.97301644e-01 2.68635839e-01 9.37302932e-02
4.44419086e-01 1.85650766e-01 -1.22150707e+00 2.70489961e-01
1.25900626e+00 5.47276199e-01 1.10811770e+00 9.57177460e-01
4.22018349e-01 1.19555175e+00 -3.91599834e-01 -4.64170612e-02
3.13871682e-01 -5.87126970e-01 -1.01934350e+00 5.09029508e-01
4.56255943e-01 -7.18670726e-01 -5.30481994e-01 1.35631073e+00
3.96828055e-01 3.58978808e-01 3.58555913e-01 -3.09978038e-01
-1.10741150e+00 2.91905642e-01 -1.02751577e+00 9.15719330e-01
5.42872250e-01 2.25317806e-01 3.21710706e-01 -1.31134546e+00
7.19943881e-01 1.18125129e+00 6.56416178e-01 3.87140602e-01
-1.28149509e+00 -2.42823772e-02 -3.46585959e-01 3.98932964e-01
-1.50438643e+00 -6.35678947e-01 6.43185019e-01 -2.32363880e-01
5.35176933e-01 3.98057073e-01 2.00743601e-01 7.50444770e-01
5.38836777e-01 4.61030155e-01 1.44791853e+00 -3.66288036e-01
5.73891997e-02 -2.37002254e-01 3.59442122e-02 2.74718463e-01
3.17437708e-01 3.57244849e-01 -5.44918835e-01 -1.37545064e-01
1.56374764e+00 -7.25614578e-02 -9.97882724e-01 1.97838113e-01
-1.18478751e+00 3.12536657e-01 2.09910437e-01 5.00181496e-01
-6.38248742e-01 1.66750744e-01 -1.55236974e-01 2.74229050e-02
5.44682682e-01 4.65582967e-01 -1.95572346e-01 -1.17291197e-01
-1.22214901e+00 1.34164497e-01 1.66517287e-01 7.82328427e-01
6.53861344e-01 2.70619422e-01 -2.72126019e-01 1.11906779e+00
1.42303497e-01 4.36917126e-01 3.95846397e-01 -1.75905788e+00
1.34607106e-01 -2.70488679e-01 7.26279974e-01 -8.98008168e-01
8.15310106e-02 -3.13125312e-01 -1.11387551e+00 2.00004339e-01
3.68656397e-01 -3.98234604e-03 -9.26817358e-01 1.24703240e+00
2.99827486e-01 7.85048246e-01 -3.72654311e-02 1.61230707e+00
4.53563303e-01 7.19237328e-01 -3.03424269e-01 -6.66866243e-01
1.25737607e+00 -8.86057198e-01 -1.37986946e+00 -3.54479700e-01
-2.78543979e-01 -1.24027765e+00 6.98599696e-01 2.92447001e-01
-1.80410409e+00 -9.18685973e-01 -7.55636394e-01 -3.52592647e-01
5.16346633e-01 -1.86526608e-02 1.85974941e-01 3.06004643e-01
-1.31406724e+00 7.25658000e-01 -6.71262741e-01 1.48479179e-01
3.95948321e-01 5.17793596e-02 -1.26994625e-01 -4.99457300e-01
-1.14306855e+00 1.38080752e+00 1.83812261e-01 4.25398618e-01
-7.06879854e-01 -7.57654011e-01 -9.11523461e-01 -6.91811815e-02
1.19336903e-01 -1.20306838e+00 1.19272208e+00 -8.07651281e-01
-1.45104825e+00 5.33106029e-01 -5.46855867e-01 -4.76935387e-01
5.81124246e-01 -6.62321806e-01 -2.56652266e-01 5.46419382e-01
4.53434102e-02 2.55644679e-01 1.44163620e+00 -1.44944406e+00
-6.04224384e-01 7.48604909e-02 -2.48546690e-01 4.07980740e-01
3.82215291e-01 2.35138595e-01 -2.80443698e-01 -9.80849206e-01
1.21750064e-01 -3.77721816e-01 -5.43155849e-01 -3.80465418e-01
1.01978041e-01 5.61645269e-01 8.37706685e-01 -1.16059041e+00
1.39105201e+00 -2.10000324e+00 3.07937443e-01 -4.00618404e-01
3.91862869e-01 2.14750782e-01 -8.17319204e-04 -1.25407994e-01
1.61686796e-03 -3.82709414e-01 -5.57640254e-01 -1.86997145e-01
-5.26452124e-01 2.53136218e-01 -3.98121446e-01 7.92801619e-01
1.84284240e-01 7.95132041e-01 -9.68216956e-01 -4.20888901e-01
6.73045933e-01 9.26623464e-01 -3.43867034e-01 4.44081128e-01
3.32485765e-01 8.23531389e-01 -1.45838335e-01 3.69454443e-01
1.34964788e+00 -4.84015882e-01 -2.57133394e-01 -5.05861640e-01
-4.33758646e-01 -1.55733749e-01 -1.22138941e+00 1.70523596e+00
-5.17274499e-01 7.48615384e-01 6.43938541e-01 -3.32042545e-01
5.28796136e-01 1.52332962e-01 3.58068287e-01 -3.71861815e-01
-4.35813107e-02 1.30606309e-01 -3.25419605e-01 -7.20403731e-01
8.89202714e-01 -1.96924806e-01 6.43642366e-01 8.17291662e-02
-2.96901792e-01 -2.67828345e-01 1.15567774e-01 7.30758309e-02
1.06006241e+00 2.31760934e-01 3.99890929e-01 -4.45261806e-01
7.82740116e-01 -2.05438167e-01 4.15955424e-01 7.66593099e-01
-2.78021455e-01 1.23585272e+00 -3.96013260e-01 -3.91804993e-01
-1.35334539e+00 -9.93141234e-01 -3.59531194e-01 5.30460000e-01
6.54479265e-01 9.34991315e-02 -8.40849519e-01 1.97246224e-01
-3.90375078e-01 5.61171949e-01 -3.90184104e-01 8.18936676e-02
-9.03406680e-01 -8.52952480e-01 -6.98107034e-02 2.19699949e-01
7.99165964e-01 -6.70158923e-01 -7.39750922e-01 1.78456321e-01
-6.97426736e-01 -1.47496831e+00 -8.81647706e-01 -2.81448722e-01
-1.04989374e+00 -9.58822846e-01 -1.20721233e+00 -7.25909173e-01
8.02778363e-01 8.36944222e-01 1.01238155e+00 -1.27255142e-01
-2.94740081e-01 2.57769018e-01 -3.86740007e-02 6.48047626e-01
-4.66205120e-01 -7.89835930e-01 3.55936997e-02 2.34441906e-01
-2.65144676e-01 -4.13575470e-01 -9.72463131e-01 4.90389317e-01
-1.05177665e+00 3.42519253e-01 7.53959954e-01 1.12174833e+00
4.64338332e-01 2.77350545e-01 3.53501886e-02 -4.85583961e-01
7.18502104e-01 -3.12590599e-01 -6.82679951e-01 -1.81350322e-03
-6.63083673e-01 -2.00301632e-02 5.20177305e-01 -5.23889959e-01
-1.78274512e+00 -1.16389923e-01 2.91648597e-01 -9.47459996e-01
-7.11444467e-02 7.12009966e-02 2.43049487e-01 -2.13616312e-01
6.61847115e-01 4.49973494e-01 -2.73964018e-01 -6.40739083e-01
4.61868465e-01 6.12995148e-01 1.07947886e+00 -2.97934115e-01
7.70471334e-01 6.89818203e-01 -1.03187621e-01 -9.49835479e-01
-7.74001360e-01 -6.19380534e-01 -6.16802573e-01 -2.63079494e-01
8.47400308e-01 -1.11117566e+00 -4.26867574e-01 5.36864281e-01
-1.34816861e+00 -3.63338709e-01 -1.21089406e-01 5.99775791e-01
-6.72392190e-01 8.83542657e-01 -1.09073615e+00 -9.72223282e-01
-2.50559837e-01 -1.10759163e+00 8.75768065e-01 3.00459862e-01
-6.86163008e-02 -9.94604170e-01 -7.00525641e-02 5.54954469e-01
7.84824550e-01 -5.27311640e-04 -9.99690220e-03 8.60634267e-01
-1.05166686e+00 5.41350603e-01 -6.89273596e-01 4.78090614e-01
3.41912776e-01 -3.81991684e-01 -9.25954938e-01 -3.50675315e-01
8.96587610e-01 3.40179205e-01 8.73253882e-01 1.22811735e+00
9.59168494e-01 -4.82643336e-01 -8.55077654e-02 1.00571764e+00
1.78759480e+00 -1.93514023e-02 1.00000453e+00 4.03813779e-01
7.48507440e-01 2.71117926e-01 8.19435537e-01 4.19851184e-01
4.93297651e-02 6.76533043e-01 1.16846360e-01 -7.45263919e-02
-5.50932050e-01 3.10819179e-01 2.98787266e-01 5.57495415e-01
-3.57356548e-01 1.10867664e-01 -4.79963183e-01 7.19818175e-01
-1.84618127e+00 -1.22341919e+00 -5.59454381e-01 1.85852301e+00
1.08553886e+00 -3.37269872e-01 -3.90746772e-01 -9.72813889e-02
1.12597287e+00 2.90289432e-01 -3.27197790e-01 1.68606751e-02
-4.48916525e-01 8.43641907e-02 5.91942132e-01 1.04897201e+00
-9.51030195e-01 7.15469241e-01 7.24380398e+00 8.49865675e-01
-6.59525752e-01 3.94041091e-01 6.83341205e-01 2.79510766e-02
-1.53536066e-01 -1.47342738e-02 -4.06146079e-01 7.92038798e-01
7.57635891e-01 -2.66073495e-01 9.33305383e-01 4.85798210e-01
8.32523465e-01 -5.35260797e-01 -4.73073363e-01 1.51879954e+00
-2.73497477e-02 -1.40091789e+00 -3.21158499e-01 -2.00847521e-01
1.05051577e+00 -1.92032188e-01 2.63772368e-01 -6.05900466e-01
3.08375835e-01 -1.06397450e+00 6.06584072e-01 8.24365675e-01
9.03353572e-01 -3.95342618e-01 6.04497731e-01 1.10229269e-01
-9.86946166e-01 -6.87294453e-02 -5.82735062e-01 -2.35637009e-01
6.15219355e-01 9.32407260e-01 -3.05673093e-01 6.83789372e-01
7.76987553e-01 9.16101754e-01 -1.74490392e-01 1.17848003e+00
-1.79670557e-01 2.61814028e-01 1.73379853e-01 8.45248163e-01
-1.60585091e-01 -5.99094689e-01 7.82270432e-01 1.33199120e+00
4.50409085e-01 6.51405990e-01 -3.12180132e-01 1.04680026e+00
3.20877016e-01 -4.98485953e-01 -2.32292980e-01 5.95743835e-01
2.71929979e-01 1.20330179e+00 -2.56602079e-01 -3.63077551e-01
-3.56025696e-01 1.45422888e+00 -4.72483449e-02 8.12732339e-01
-7.44327605e-01 1.40169933e-01 7.67319918e-01 1.12373762e-01
4.20518517e-01 -3.20476025e-01 -6.00075543e-01 -1.55716610e+00
-2.51876533e-01 -6.77203655e-01 2.16972247e-01 -1.39584649e+00
-1.14282608e+00 5.65993071e-01 -7.14159831e-02 -1.32856786e+00
-7.84509778e-02 -1.91484213e-01 -3.08756113e-01 1.31769645e+00
-1.76902258e+00 -6.88895881e-01 -4.85183477e-01 5.32196879e-01
9.93869543e-01 3.79593492e-01 7.16199353e-02 4.31827724e-01
-2.93345630e-01 -2.25269258e-01 2.41284907e-01 -3.14265221e-01
7.12198794e-01 -1.10249817e+00 2.31117159e-01 1.54099691e+00
-4.09916788e-01 7.11724401e-01 1.18475127e+00 -7.42996275e-01
-1.23206174e+00 -9.50465560e-01 5.41431308e-01 -3.08305889e-01
5.64828336e-01 4.20233965e-01 -1.29390371e+00 3.93082082e-01
5.44159532e-01 2.21617475e-01 -2.01885346e-02 -7.41002023e-01
1.86864331e-01 2.44371906e-01 -1.21196425e+00 4.16192502e-01
9.13348198e-01 -5.17288923e-01 -7.49077737e-01 -5.33668324e-02
6.06713951e-01 -8.87758791e-01 -8.38348687e-01 3.64963233e-01
2.22382411e-01 -1.15994203e+00 1.36320639e+00 1.99349195e-01
7.41730094e-01 -8.17690134e-01 2.69688852e-02 -1.31063461e+00
-9.71284509e-01 -1.03153789e+00 -7.04238594e-01 7.44767308e-01
-2.35253155e-01 -2.14817882e-01 2.87141144e-01 4.86817241e-01
-1.57061905e-01 -2.57880479e-01 -6.81098759e-01 -5.11272132e-01
-5.53349316e-01 -5.57822883e-02 -4.73733097e-02 9.56878543e-01
-3.52706432e-01 1.32122397e-01 -9.80924129e-01 5.26014864e-01
1.29327786e+00 4.66120504e-02 3.55489820e-01 -5.70567846e-01
-2.59278089e-01 -2.16966406e-01 2.94090454e-02 -1.38182390e+00
-2.39111677e-01 -9.03371945e-02 1.35318100e-01 -1.47878516e+00
4.39043373e-01 1.75905123e-01 -1.87719420e-01 -3.56045455e-01
-6.50406539e-01 3.75118524e-01 -1.57137625e-02 6.13504827e-01
-4.44589287e-01 3.50946665e-01 1.72042084e+00 2.60027975e-01
-5.19937761e-02 -1.42070815e-01 -4.08446819e-01 6.80378675e-01
5.22806466e-01 -1.24044076e-01 -2.72465199e-01 -7.13236690e-01
-2.65462488e-01 3.78392667e-01 6.62667394e-01 -7.88948238e-01
2.52467901e-01 -4.09863979e-01 6.96938813e-01 -5.43868244e-01
4.87201005e-01 -7.33408034e-01 6.17433071e-01 3.88804853e-01
-3.00677240e-01 -1.01518437e-01 -2.89363693e-02 5.93982875e-01
-2.85837471e-01 -2.97081321e-01 1.47705555e+00 -3.37166429e-01
-7.57982373e-01 -1.05158672e-01 -3.43875974e-01 -1.55806109e-01
6.11436784e-01 -3.67365062e-01 -3.87727112e-01 -4.08758461e-01
-7.46214926e-01 -1.06439389e-01 7.41702318e-01 7.47088194e-02
1.08819723e+00 -1.11859334e+00 -8.77924681e-01 5.02248928e-02
-5.81598043e-01 1.48228602e-02 7.52268136e-01 1.04514050e+00
-8.12568963e-01 1.81458205e-01 -2.87672013e-01 -5.66373646e-01
-1.23035884e+00 6.56108558e-01 4.59282875e-01 -1.70901883e-02
-7.76247740e-01 7.90181100e-01 3.82218957e-01 6.02857471e-01
-2.26232782e-01 -2.58150637e-01 -9.06329826e-02 -6.27690971e-01
1.09577572e+00 8.43357027e-01 -3.87849957e-01 -8.56472254e-01
-7.83436447e-02 7.38207817e-01 2.62275077e-02 -2.53265202e-01
1.30485153e+00 -1.22219920e+00 -3.33460420e-01 9.76043344e-02
7.72620678e-01 -7.30458787e-03 -1.83315945e+00 -1.91198140e-01
-1.95381157e-02 -1.32892644e+00 4.19972032e-01 -5.35796285e-01
-9.22118366e-01 3.70777190e-01 5.68531334e-01 2.15141382e-02
1.56534374e+00 -2.37031445e-01 9.46410418e-01 -4.38459903e-01
2.47830585e-01 -9.32089865e-01 7.37731019e-03 3.83442670e-01
9.23278749e-01 -1.26010776e+00 2.39858210e-01 -5.96033633e-01
-4.94804502e-01 9.84846830e-01 4.56789732e-01 -3.19239408e-01
8.89308304e-02 3.70344281e-01 2.75933836e-02 1.31158724e-01
-6.58272326e-01 -2.61744589e-01 3.74974996e-01 5.47531605e-01
3.95060837e-01 -5.84871531e-01 -5.55870473e-01 -7.66838260e-04
3.17421496e-01 4.05307978e-01 9.44518626e-01 5.95486283e-01
-6.01282299e-01 -5.67949772e-01 -1.03881145e+00 -4.89255823e-02
-7.19953954e-01 -2.93476373e-01 4.07537997e-01 1.16413444e-01
-3.03556379e-02 1.13644922e+00 -1.09025635e-01 6.38769865e-02
3.65446173e-02 -4.92565840e-01 7.05069125e-01 -1.93524003e-01
-1.16782993e-01 6.09000981e-01 -2.78515220e-01 -6.56808913e-01
-7.80269623e-01 -4.68661517e-01 -9.00580466e-01 -4.09790337e-01
-3.36605906e-01 4.87341695e-02 1.91907868e-01 6.19164586e-01
2.44304374e-01 5.53811848e-01 6.63180709e-01 -1.08516419e+00
-3.29877675e-01 -9.93753612e-01 -7.54736364e-01 4.68676925e-01
8.44539583e-01 -3.34643215e-01 -6.93271935e-01 5.27386427e-01] | [11.457064628601074, -2.645368814468384] |
2dc072d5-6275-4ff8-a97c-8cc6e86be181 | using-rhetorical-structure-theory-for | null | null | https://aclanthology.org/W17-3608 | https://aclanthology.org/W17-3608.pdf | Using Rhetorical Structure Theory for Detection of Fake Online Reviews | null | ['Olu Popoola'] | 2017-09-01 | null | null | null | ws-2017-9 | ['deception-detection'] | ['miscellaneous'] | [-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] | [-7.248286724090576, 3.568211317062378] |
b9305e95-2352-4415-a01a-24d806b1240a | annotation-and-classification-of-evidence-and-1 | 2107.06990 | null | https://arxiv.org/abs/2107.06990v1 | https://arxiv.org/pdf/2107.06990v1.pdf | Annotation and Classification of Evidence and Reasoning Revisions in Argumentative Writing | Automated writing evaluation systems can improve students' writing insofar as students attend to the feedback provided and revise their essay drafts in ways aligned with such feedback. Existing research on revision of argumentative writing in such systems, however, has focused on the types of revisions students make (e.g., surface vs. content) rather than the extent to which revisions actually respond to the feedback provided and improve the essay. We introduce an annotation scheme to capture the nature of sentence-level revisions of evidence use and reasoning (the `RER' scheme) and apply it to 5th- and 6th-grade students' argumentative essays. We show that reliable manual annotation can be achieved and that revision annotations correlate with a holistic assessment of essay improvement in line with the feedback provided. Furthermore, we explore the feasibility of automatically classifying revisions according to our scheme. | ['Richard Correnti', 'Lindsay C. Matsumura', 'Diane Litman', 'Elaine Wang', 'Tazin Afrin'] | 2021-07-14 | annotation-and-classification-of-evidence-and | https://aclanthology.org/2020.bea-1.7 | https://aclanthology.org/2020.bea-1.7.pdf | ws-2020-7 | ['automated-writing-evaluation'] | ['natural-language-processing'] | [ 1.29248992e-01 5.49058974e-01 -5.38121164e-01 -3.38399738e-01
-8.59300792e-01 -1.05407333e+00 6.01652741e-01 1.03219008e+00
-4.86313730e-01 7.49518752e-01 6.85949624e-01 -8.96078169e-01
-3.76721114e-01 -8.52771938e-01 -4.34236139e-01 1.48296922e-01
1.13427472e+00 2.21194446e-01 2.27086455e-01 -4.96366322e-01
1.07338858e+00 4.19751443e-02 -1.26053882e+00 4.24876720e-01
1.21517837e+00 3.12556289e-02 -3.23217250e-02 1.03127933e+00
-4.09388989e-01 1.04567063e+00 -1.08860540e+00 -7.67051697e-01
-5.07364333e-01 -6.70434177e-01 -1.31331742e+00 1.96304005e-02
7.61907160e-01 -3.21966738e-01 1.96379945e-01 9.53896582e-01
1.15704946e-01 5.80368079e-02 6.17671669e-01 -6.19356751e-01
-1.07185090e+00 1.00318110e+00 -1.89838409e-01 5.45205534e-01
7.78339446e-01 -2.86513627e-01 1.12843168e+00 -8.64222765e-01
9.14805949e-01 8.88944387e-01 7.51703501e-01 6.17995799e-01
-9.99244452e-01 -2.95636654e-01 2.40184695e-01 1.81014568e-01
-2.76457429e-01 -4.35562164e-01 7.23260164e-01 -9.12817538e-01
4.82610315e-01 3.60068351e-01 8.40786874e-01 7.58140206e-01
2.06839785e-01 7.11653471e-01 1.22370708e+00 -8.00220251e-01
3.57979760e-02 5.35517275e-01 1.00436103e+00 6.05939806e-01
6.47080481e-01 -6.68248951e-01 -7.98335671e-01 9.20428932e-02
-7.35517144e-02 -3.76542747e-01 -3.95582169e-01 2.54933715e-01
-1.02865589e+00 6.13925219e-01 -2.49008372e-01 3.67878884e-01
-4.64697443e-02 -8.55976890e-04 4.06687826e-01 6.26845360e-01
4.24017698e-01 8.90124559e-01 -5.55454195e-01 -5.88919222e-01
-1.15520656e+00 3.97808820e-01 8.78036320e-01 5.16342223e-01
3.34902734e-01 -1.90970555e-01 -3.76444548e-01 8.78980279e-01
6.65720046e-01 2.78072745e-01 7.45516181e-01 -1.03312349e+00
6.09201193e-01 9.77176666e-01 6.63722306e-02 -7.29384661e-01
4.87746559e-02 -5.77703297e-01 1.02833897e-01 1.19391635e-01
4.36367601e-01 -2.54906528e-02 -2.28136584e-01 1.51485026e+00
2.29317188e-01 -7.84785092e-01 2.07277924e-01 4.41448867e-01
1.23455966e+00 7.22984672e-02 7.62272477e-02 -3.23426843e-01
1.30753946e+00 -8.97721887e-01 -9.55381811e-01 1.83088601e-01
1.12555814e+00 -1.28426194e+00 1.18408787e+00 4.49645489e-01
-1.66335535e+00 -4.28844899e-01 -1.09887588e+00 -3.40896636e-01
6.24196380e-02 2.53978521e-01 -1.53925200e-03 8.70070815e-01
-9.95850325e-01 7.64722109e-01 -4.97518450e-01 -1.64241374e-01
3.26474696e-01 -2.57328004e-01 -7.78497010e-02 3.52078080e-01
-7.66145170e-01 1.03691339e+00 -3.02251697e-01 -3.03318232e-01
-1.00606225e-01 -1.05307007e+00 -6.74751878e-01 8.86339471e-02
-3.56977917e-02 -4.26339746e-01 2.08624721e+00 -7.40031838e-01
-1.62451863e+00 1.01848519e+00 -1.38916001e-01 1.41709939e-01
6.46240294e-01 -3.22384685e-01 1.81150183e-01 1.29278079e-01
4.22649205e-01 1.19552165e-01 1.44204050e-01 -7.68935800e-01
-6.07645035e-01 -2.51322150e-01 2.89623469e-01 3.34288418e-01
-7.08460808e-01 1.66373670e-01 3.63626212e-01 -7.69566715e-01
3.14283162e-01 -6.35041296e-01 2.62760043e-01 4.05754484e-02
-2.21522197e-01 -9.06128645e-01 7.91579664e-01 -6.68617845e-01
1.65316379e+00 -1.73870051e+00 -1.06100135e-01 9.85117927e-02
5.36736190e-01 2.27930412e-01 1.05855398e-01 5.99818289e-01
5.95117360e-02 6.57103419e-01 1.95604023e-02 -1.91067994e-01
3.64348203e-01 -8.19174945e-02 -4.07487780e-01 1.83465868e-01
-2.53130764e-01 7.81368971e-01 -1.18934667e+00 -4.64150906e-01
-3.07408154e-01 -1.90213621e-01 -4.31426406e-01 9.29456428e-02
-1.37950620e-02 3.09844175e-03 -4.91927147e-01 4.32874471e-01
1.04171880e-01 -1.77792162e-01 1.47795483e-01 5.83312511e-01
-5.23805559e-01 1.41947365e+00 -7.81078339e-01 1.07444239e+00
-3.28396112e-01 1.10538435e+00 -1.63215742e-01 -6.58084512e-01
9.58304167e-01 4.74549085e-01 -1.10277750e-01 -7.36787975e-01
-3.18838388e-01 3.46698076e-01 2.54090905e-01 -3.79715919e-01
8.44284594e-01 8.26541036e-02 8.63002986e-02 1.59605646e+00
-2.97732532e-01 -4.25796896e-01 7.35242784e-01 7.31753290e-01
1.04497826e+00 2.01678932e-01 5.26879847e-01 -4.88973111e-01
5.38274348e-01 4.81783524e-02 1.37978494e-01 8.79742026e-01
-1.42862573e-01 1.10500008e-01 5.90687811e-01 -1.71492055e-01
-9.76914167e-01 -6.49652898e-01 -2.02621207e-01 1.28313160e+00
-3.49336743e-01 -6.69248700e-01 -7.99406767e-01 -9.61257398e-01
-7.98136964e-02 9.84331846e-01 -5.16096175e-01 -5.62788881e-02
-5.61970711e-01 -2.58365333e-01 5.25480688e-01 6.35964215e-01
1.66699931e-01 -1.09545195e+00 -9.62014735e-01 3.90944213e-01
-2.03434423e-01 -4.78458405e-01 -3.76433283e-01 1.00563793e-02
-8.35658133e-01 -1.33981693e+00 -1.89222619e-01 -5.66947937e-01
8.94866765e-01 3.69333774e-01 1.15753174e+00 8.83095026e-01
3.43021274e-01 9.04550791e-01 -5.34744024e-01 -7.26902068e-01
-8.31411004e-01 1.11471921e-01 -2.67210871e-01 -9.39212799e-01
2.84751832e-01 -3.37383747e-01 -2.32962221e-01 -1.52795807e-01
-6.85275972e-01 3.23676795e-01 2.13107735e-01 6.73276067e-01
6.18059300e-02 -5.89022398e-01 8.07008386e-01 -1.45397639e+00
1.43850720e+00 -2.41441295e-01 -2.50214159e-01 5.35854518e-01
-1.26111472e+00 1.01600654e-01 2.19355240e-01 -1.61530808e-01
-1.14708066e+00 -8.37989450e-01 -1.72179163e-01 7.50650585e-01
9.19537768e-02 8.42709720e-01 5.59574842e-01 -2.13198483e-01
9.43378389e-01 -1.83092374e-02 -2.91977704e-01 -1.18048698e-01
7.49036595e-02 5.98560154e-01 4.65656579e-01 -1.20414007e+00
7.67058790e-01 -3.01311880e-01 -1.72939241e-01 -3.77979577e-01
-1.54151118e+00 -3.48864287e-01 -7.50171840e-01 -6.02067173e-01
3.57268304e-01 -5.73026240e-01 -7.53701270e-01 5.48139699e-02
-1.27862751e+00 -5.03632009e-01 -6.10016227e-01 5.41942239e-01
-2.69061446e-01 5.75528800e-01 -9.13988888e-01 -8.13814521e-01
-4.90506172e-01 -9.60453212e-01 4.87095833e-01 5.02845407e-01
-1.09786427e+00 -1.36030364e+00 5.35362840e-01 8.80106509e-01
2.06328571e-01 -1.06466621e-01 1.35437000e+00 -7.57251143e-01
-1.39595360e-01 4.24317457e-02 -3.90089699e-04 2.47693881e-01
-1.59524187e-01 4.76250917e-01 -6.74205065e-01 2.07364514e-01
7.02511817e-02 -5.24535239e-01 6.17129683e-01 -8.31525698e-02
8.74106646e-01 -7.33671010e-01 1.97290078e-01 -1.81645289e-01
8.79293621e-01 -2.24086955e-01 2.30001971e-01 7.46972680e-01
3.73379201e-01 8.12105417e-01 5.21011531e-01 1.73416123e-01
7.49920785e-01 3.44041228e-01 -1.62291631e-01 5.67998171e-01
-5.44111073e-01 -3.05037141e-01 5.68972528e-01 1.45289505e+00
1.34761825e-01 -1.93701684e-01 -1.06150949e+00 7.65030086e-01
-1.74881029e+00 -1.00826299e+00 -1.07414067e+00 1.88788712e+00
1.47354043e+00 3.96729648e-01 -2.92782456e-01 6.94007158e-01
3.52358967e-01 -1.42992353e-02 4.91581820e-02 -1.11236107e+00
1.41925201e-01 6.49642169e-01 -7.39509165e-02 1.05632699e+00
-1.84506983e-01 5.87626576e-01 6.67433023e+00 1.65605590e-01
-8.25531542e-01 1.61766425e-01 4.33157325e-01 2.22945511e-01
-1.20423543e+00 6.07170202e-02 -9.41089690e-01 3.53995204e-01
9.06972826e-01 -4.32316601e-01 -1.07258081e-01 4.94387448e-01
2.69313067e-01 -2.30576456e-01 -1.05880773e+00 4.23324741e-02
2.00853348e-01 -1.50570405e+00 -1.00526333e-01 -9.86440182e-02
1.15002429e+00 -6.77500784e-01 -6.91080093e-03 3.03343564e-01
6.20211840e-01 -5.96439123e-01 1.33968747e+00 4.55861300e-01
4.50469702e-01 -2.79611140e-01 7.46157110e-01 5.26664734e-01
-6.38896048e-01 1.74326003e-01 8.85775760e-02 -7.18026102e-01
-4.20254976e-01 4.14701819e-01 -1.34834075e+00 -2.32229033e-03
2.44684637e-01 6.97970212e-01 -9.53243077e-01 7.28260577e-01
-1.17776561e+00 1.16320753e+00 3.33407789e-01 -5.70577145e-01
2.69471407e-02 5.28655611e-02 4.52240288e-01 1.16438580e+00
3.68697196e-01 2.70675868e-01 2.25189514e-02 6.73144400e-01
-4.22043890e-01 2.62809634e-01 7.90505931e-02 -2.38621771e-01
7.27636516e-01 1.16525257e+00 -5.66051364e-01 -5.42353809e-01
-6.98454604e-02 3.67803216e-01 6.40935540e-01 1.28998622e-01
-2.11062822e-02 -1.61275342e-01 1.37806430e-01 2.62253553e-01
-1.23512506e-01 -9.31337550e-02 -1.12367594e+00 -8.48341942e-01
3.02949220e-01 -7.82725394e-01 2.98416585e-01 -9.01523232e-01
-1.02383780e+00 -2.54478380e-02 -4.73234266e-01 -6.02991462e-01
1.24125918e-02 -4.77721900e-01 -1.14122093e+00 8.64148080e-01
-1.62126684e+00 -6.04860723e-01 -4.40636367e-01 -1.09756753e-01
6.89426661e-01 7.20366314e-02 7.75388539e-01 -8.55959728e-02
-4.31029588e-01 7.90505648e-01 -9.60576087e-02 -1.57818198e-02
1.12735128e+00 -1.75363684e+00 8.64954963e-02 8.48356962e-01
2.16255814e-01 8.87949705e-01 7.11186469e-01 -7.22850740e-01
-7.80658484e-01 -6.27927184e-01 1.70553279e+00 -7.76428044e-01
8.73026609e-01 4.18523014e-01 -1.16316330e+00 5.26821256e-01
4.53361958e-01 -6.51492476e-01 1.21425521e+00 5.30421853e-01
-5.00597715e-01 2.90368915e-01 -8.18136632e-01 5.38409293e-01
7.48234332e-01 -7.52687097e-01 -1.46944046e+00 3.95058453e-01
3.54919523e-01 -6.65698051e-01 -9.51164782e-01 3.85232791e-02
5.37088752e-01 -6.74717486e-01 3.45394313e-01 -8.95980835e-01
1.23094702e+00 -1.40505418e-01 4.30914104e-01 -1.26279747e+00
-3.24832112e-01 -3.57270926e-01 -1.06851451e-01 1.28590584e+00
4.84440058e-01 -2.83625931e-01 6.73755646e-01 6.84937894e-01
-7.52984822e-01 -9.86190438e-01 -4.06711787e-01 3.41097675e-02
2.76607782e-01 -2.35424563e-01 3.08487177e-01 1.26607120e+00
7.04606950e-01 3.62362385e-01 5.36983550e-01 -4.11959171e-01
2.40463018e-01 1.76392853e-01 7.79297233e-01 -1.69777095e+00
2.33578682e-02 -1.09019935e+00 5.69195524e-02 -6.95954621e-01
3.26523513e-01 -1.00611746e+00 -7.67049864e-02 -2.04242468e+00
4.29188520e-01 -5.22059679e-01 4.24806438e-02 5.79650223e-01
-7.13093400e-01 1.77594528e-01 1.48576319e-01 3.76383692e-01
-7.67820954e-01 1.86872005e-03 1.23015797e+00 8.09661597e-02
-3.64186354e-02 -9.50906053e-02 -1.27593088e+00 8.82511497e-01
6.85993969e-01 -4.24328297e-01 -2.60692030e-01 -4.62538719e-01
9.63130713e-01 4.03601900e-02 2.39447981e-01 -5.10835290e-01
6.18140280e-01 -5.64302981e-01 3.60227227e-01 -4.38725114e-01
-4.74486619e-01 1.05337642e-01 -5.02748847e-01 3.72519940e-01
-1.18708599e+00 4.56236869e-01 9.89398509e-02 8.48179962e-03
-1.84164509e-01 -9.80803549e-01 5.82268775e-01 -6.72018081e-02
2.13180054e-02 -4.62976545e-01 -6.70122862e-01 2.92701274e-01
6.81160986e-01 -3.31204742e-01 -1.01662064e+00 -3.29352170e-01
-3.56429607e-01 1.15674235e-01 5.08002877e-01 1.31364524e-01
4.17818367e-01 -1.22468650e+00 -8.79922509e-01 -3.25459063e-01
6.27090633e-02 -1.94230691e-01 -1.68604314e-01 9.08800364e-01
-6.13418996e-01 6.03748858e-01 -6.34435490e-02 -1.32338941e-01
-1.56887162e+00 -3.30334872e-01 -5.27380593e-02 -5.98816454e-01
-4.24015999e-01 9.35948610e-01 -3.19335490e-01 -5.27336180e-01
6.40758798e-02 -2.81362981e-01 -6.65243089e-01 3.67896765e-01
6.37507677e-01 5.57913065e-01 3.78192812e-01 -2.11560100e-01
1.18509561e-01 2.33464241e-01 -2.71498680e-01 -2.82908529e-01
1.15569329e+00 -1.43303066e-01 -1.90711364e-01 7.72618651e-01
3.81639987e-01 9.65054870e-01 -6.53234601e-01 -2.08280474e-01
2.74437219e-01 -3.17187279e-01 -1.13332577e-01 -1.18755162e+00
-1.67837337e-01 7.83824861e-01 -3.23014557e-01 3.57542545e-01
3.49878371e-01 -2.80085206e-01 5.57570696e-01 6.73293233e-01
-2.62155771e-01 -1.46302521e+00 4.24964398e-01 7.16628730e-01
8.08650970e-01 -9.63797629e-01 5.03378451e-01 -3.15095067e-01
-1.70532882e-01 1.61074722e+00 7.09774375e-01 3.11222017e-01
1.09818026e-01 -1.03187568e-01 1.15371354e-01 -2.56733209e-01
-1.03354728e+00 3.76244187e-01 6.33393347e-01 8.61799791e-02
1.34450352e+00 1.63860619e-01 -1.14337933e+00 7.00622559e-01
-8.60604882e-01 -2.90937334e-01 1.46264195e+00 9.61496592e-01
-9.17439818e-01 -1.36535561e+00 -5.39670587e-01 4.98710245e-01
-5.70588052e-01 -1.37724966e-01 -9.86172855e-01 4.30315375e-01
-1.00236870e-01 9.73686397e-01 -2.12131202e-01 2.73353755e-01
2.72262007e-01 4.23405826e-01 6.38351858e-01 -1.25569832e+00
-1.25124753e+00 -3.95355314e-01 4.20552522e-01 2.08494410e-01
-4.74443734e-01 -9.85674620e-01 -1.41231143e+00 -5.38225114e-01
-3.91893417e-01 7.42852271e-01 6.93401515e-01 1.40184224e+00
-6.84129596e-02 8.01336348e-01 1.31438151e-01 1.40899435e-01
-9.35440421e-01 -1.06445968e+00 3.79834063e-02 2.45347455e-01
2.84315467e-01 -3.11674327e-01 -2.95632631e-01 1.86434045e-01] | [11.274574279785156, 9.24040699005127] |
b81b6199-e112-4462-ad18-a413418cd804 | interactive-models-for-post-editing | null | null | https://aclanthology.org/2021.triton-1.19 | https://aclanthology.org/2021.triton-1.19.pdf | Interactive Models for Post-Editing | Despite the increasingly good quality of Machine Translation (MT) systems, MT outputs require corrections. Automatic Post-Editing (APE) models have been introduced to perform these corrections without human intervention. However, no system has been able to fully automate the Post-Editing (PE) process. Moreover, while numerous translation tools, such as Translation Memories (TMs), largely benefit from translators’ input, Human-Computer Interaction (HCI) remains limited when it comes to PE. This research-in-progress paper discusses APE models and suggests that they could be improved in more interactive scenarios, as previously done in MT with the creation of Interactive MT (IMT) systems. Based on the hypothesis that PE would benefit from HCI, two methodologies are proposed. Both suggest that traditional batch learning settings are not optimal for PE. Instead, online techniques are recommended to train and update PE models on the fly, via either real or simulated interactions with the translator. | ['Ruslan Mitkov', 'Marie Escribe'] | null | null | null | null | triton-2021-7 | ['automatic-post-editing', 'automatic-post-editing'] | ['computer-vision', 'natural-language-processing'] | [ 5.81638336e-01 4.99335796e-01 -5.14156297e-02 -4.06969845e-01
-8.54922295e-01 -7.55917549e-01 8.84918511e-01 7.58364648e-02
-4.77556139e-01 5.89753747e-01 1.09592550e-01 -9.67593193e-01
2.59415209e-01 -3.20572436e-01 -7.65731990e-01 1.98086262e-01
5.48549891e-01 8.65819216e-01 2.69552283e-02 -3.45507205e-01
4.62737083e-01 2.66762674e-01 -1.27536619e+00 5.60527682e-01
1.27403831e+00 6.22953176e-02 5.37635744e-01 7.79869795e-01
-4.01382208e-01 5.67781508e-01 -6.61035776e-01 -6.60074294e-01
2.27448329e-01 -8.05458546e-01 -9.19577837e-01 -2.32507020e-01
2.17032954e-01 -4.04527932e-01 1.34253368e-01 6.48531318e-01
6.00767970e-01 -4.42101061e-02 4.50216711e-01 -1.09565318e+00
-8.86835337e-01 8.30621064e-01 4.23088409e-02 -1.62527919e-01
7.84607351e-01 1.91719651e-01 4.49748009e-01 -1.37686443e+00
9.10785973e-01 1.11744165e+00 6.50135577e-01 5.38738549e-01
-1.23200190e+00 -4.69796121e-01 -3.63415390e-01 1.26256064e-01
-8.67922902e-01 -6.93778932e-01 4.32956576e-01 -4.44751203e-01
1.54794550e+00 4.11652237e-01 7.02393293e-01 1.20672464e+00
4.13685679e-01 7.88553596e-01 1.27436018e+00 -1.26296043e+00
-1.49442941e-01 8.77246976e-01 -2.49071226e-01 3.87493104e-01
8.43661502e-02 -7.08453804e-02 -7.34071076e-01 1.60539731e-01
6.90753102e-01 -4.80083495e-01 3.13648432e-02 -1.54969350e-01
-1.34207630e+00 3.01026195e-01 -2.60610044e-01 6.11976922e-01
-4.39362049e-01 -2.16189787e-01 3.55481267e-01 8.76472235e-01
5.57681978e-01 7.55215585e-01 -5.30521572e-01 -7.84251750e-01
-1.21250522e+00 6.55544028e-02 1.05406713e+00 1.32363319e+00
7.92815387e-01 -2.93559730e-01 -3.43476593e-01 8.01300943e-01
1.17233902e-01 4.44939375e-01 4.13880348e-01 -7.68233478e-01
8.35166872e-01 7.43020535e-01 4.90925729e-01 -8.35897923e-01
5.06457360e-03 -1.70884788e-01 -2.58209348e-01 1.01322047e-01
2.92949528e-01 -1.97164550e-01 -8.04665625e-01 1.15335810e+00
1.60867408e-01 -3.39202523e-01 -1.55054659e-01 5.75138271e-01
3.31267208e-01 6.62301421e-01 6.77087232e-02 -4.17255521e-01
7.54453659e-01 -1.11716092e+00 -9.78918254e-01 -3.74901891e-01
1.11838794e+00 -1.39820731e+00 1.52026546e+00 4.29429442e-01
-1.37386453e+00 -7.22932875e-01 -9.27702606e-01 -2.64927238e-01
-4.03230935e-01 3.74195725e-01 3.79120111e-01 9.23582494e-01
-1.12419784e+00 8.93402100e-01 -1.01661777e+00 -8.36477041e-01
-9.19818059e-02 5.14086246e-01 -1.45819515e-01 3.26718688e-02
-1.07914329e+00 1.60917306e+00 -2.55996250e-02 3.40585023e-01
-4.17061478e-01 -5.53232312e-01 -4.98996854e-01 -2.41916180e-01
1.30344164e-02 -7.34189868e-01 1.77966809e+00 -1.61961317e+00
-2.08880281e+00 5.63368559e-01 -3.49438041e-01 -9.18896496e-02
9.25795734e-01 -6.33260131e-01 -2.08094746e-01 -1.44935906e-01
-8.98732841e-02 7.90749550e-01 6.16156042e-01 -1.17734981e+00
-5.45309544e-01 1.21770389e-01 -9.66136083e-02 4.35805887e-01
-5.19339204e-01 5.07727146e-01 -4.42018658e-01 -3.16935033e-01
-2.14357257e-01 -1.15574443e+00 -5.68973646e-03 -2.52943933e-01
7.42095560e-02 -2.37984955e-02 7.51761973e-01 -1.16462803e+00
1.56405437e+00 -1.64384794e+00 1.51846305e-01 8.45053941e-02
-3.18466395e-01 6.44424140e-01 -2.98497021e-01 1.02857637e+00
2.37585947e-01 3.61747712e-01 -3.59085873e-02 -5.72116256e-01
4.19428423e-02 1.46351889e-01 -7.39592612e-02 7.87812192e-03
2.60190696e-01 1.09968913e+00 -8.20758879e-01 -4.56699014e-01
1.94077045e-01 2.96748966e-01 -4.25657928e-01 3.12562793e-01
-3.40821028e-01 5.33335030e-01 -1.34119079e-01 3.77325714e-01
5.20231187e-01 1.26745760e-01 4.46038634e-01 3.44296813e-01
-4.61368829e-01 5.47629774e-01 -8.04015815e-01 1.72230101e+00
-7.26119339e-01 8.69247973e-01 5.92681877e-02 -4.57582295e-01
9.37718630e-01 5.56725860e-01 1.09382562e-01 -8.71405542e-01
-2.55728755e-02 5.76084673e-01 2.92289287e-01 -7.27642596e-01
9.94364619e-01 1.19005397e-01 4.18838412e-01 1.02732325e+00
-1.49559632e-01 -1.79464877e-01 1.32641137e-01 1.41957447e-01
8.88781846e-01 7.66204536e-01 4.78698686e-02 6.61924481e-02
1.84989735e-01 5.53069592e-01 2.00873271e-01 7.26511717e-01
-2.30252910e-02 3.73221695e-01 -4.64387089e-02 -1.48810536e-01
-1.37034893e+00 -4.94914919e-01 3.61060619e-01 1.10166144e+00
-1.79937616e-01 -5.82016587e-01 -1.18531370e+00 -4.22308654e-01
-5.69923818e-01 1.14186275e+00 -6.92826435e-02 -5.40208407e-02
-8.14864576e-01 -1.78067669e-01 3.85981888e-01 4.09178674e-01
1.54234052e-01 -1.29643333e+00 -8.60176802e-01 6.77544653e-01
-4.53027874e-01 -8.28553498e-01 -4.30859804e-01 -7.09430268e-03
-1.18928528e+00 -3.41322333e-01 -5.72508872e-01 -6.95980847e-01
6.75481558e-01 3.34074050e-01 9.66294408e-01 2.07882658e-01
2.47324094e-01 4.64689791e-01 -6.93778336e-01 -4.91620481e-01
-1.13741755e+00 4.81180131e-01 7.18199089e-02 -5.30962706e-01
4.46787030e-01 -5.02869189e-01 -3.45146060e-01 4.42024112e-01
-7.12042511e-01 6.69970632e-01 9.94588792e-01 6.82289660e-01
1.95974767e-01 -6.08346581e-01 4.66481000e-01 -1.05541515e+00
1.06936669e+00 -9.76315439e-02 -1.58873245e-01 6.20029926e-01
-1.14258003e+00 -2.11533472e-01 4.19760674e-01 -8.44644070e-01
-1.30063689e+00 -7.74958581e-02 -1.62586942e-02 -1.57364950e-01
-2.63232421e-02 8.21907759e-01 1.95171479e-02 -8.06080028e-02
7.24223495e-01 9.63709131e-02 -3.47723179e-02 -3.76051128e-01
2.57558405e-01 1.11129701e+00 1.32533297e-01 -5.82970321e-01
6.87416613e-01 -3.99811655e-01 -5.47572553e-01 -5.58164835e-01
-1.37108028e-01 -1.20921008e-01 -9.23118353e-01 -4.27702606e-01
4.17032838e-01 -6.22482359e-01 3.14710313e-03 1.89443484e-01
-1.19832325e+00 -7.91865468e-01 1.32912591e-01 5.18731833e-01
-7.10385978e-01 3.41122746e-01 -6.34563982e-01 -9.25083280e-01
-5.43539464e-01 -1.08742118e+00 7.90293336e-01 2.28071541e-01
-1.09120190e+00 -8.29591572e-01 7.13151842e-02 6.10854566e-01
6.29477262e-01 -3.24848086e-01 1.11672759e+00 -5.25481343e-01
-5.63820958e-01 -3.75756681e-01 3.69403623e-02 2.58399367e-01
1.91730142e-01 3.24510127e-01 -6.96610570e-01 -2.21393272e-01
-1.25294179e-01 -1.80057794e-01 -1.87671229e-01 -3.61342542e-02
3.35583448e-01 -4.41230625e-01 -3.33104700e-01 1.56239076e-02
8.88258576e-01 3.96719694e-01 7.31707096e-01 5.98124802e-01
5.16828775e-01 7.86867619e-01 1.01112330e+00 3.71907242e-02
3.00477266e-01 7.42691696e-01 -3.01225722e-01 -8.79647359e-02
-1.43564180e-01 -7.09714890e-01 9.17834401e-01 1.32577598e+00
-1.25459716e-01 -3.86406481e-01 -1.19446242e+00 4.30125564e-01
-1.91958404e+00 -6.19150877e-01 -4.36944038e-01 2.12022090e+00
8.28526080e-01 2.31349707e-01 -1.34138063e-01 -1.65399104e-01
5.98130345e-01 -5.64881325e-01 -2.77347505e-01 -1.12375641e+00
2.69477576e-01 4.19771403e-01 2.87580341e-01 6.00354850e-01
-3.64972651e-01 1.28791749e+00 6.76800203e+00 3.30843806e-01
-1.17020082e+00 3.19646984e-01 3.14598590e-01 9.22327712e-02
-3.97116035e-01 4.43956196e-01 -4.80330527e-01 1.96169958e-01
1.48208582e+00 -1.53363258e-01 6.06356025e-01 7.41447508e-01
6.97626829e-01 -1.58227324e-01 -1.29081583e+00 5.98741770e-01
-1.29335627e-01 -1.02666008e+00 1.17677554e-01 6.20306097e-02
6.59759164e-01 -2.09642038e-01 -7.31427073e-02 6.21901810e-01
1.51190208e-02 -7.07395792e-01 8.76192689e-01 5.19695580e-01
8.52504194e-01 -5.48310995e-01 6.52880907e-01 7.53476620e-01
-6.14396334e-01 1.72736183e-01 -2.55412340e-01 -3.59624535e-01
2.37455636e-01 1.49712741e-01 -1.35492027e+00 5.80511928e-01
3.61857235e-01 3.75652403e-01 -7.78268397e-01 6.36496425e-01
-3.10409695e-01 8.46394777e-01 -1.16123915e-01 -1.91874355e-01
-2.37835739e-02 -4.79331702e-01 3.21398586e-01 1.56451380e+00
7.10324526e-01 -7.44451061e-02 2.12175380e-02 5.87880075e-01
7.04314336e-02 4.97043252e-01 -4.96237189e-01 -5.54825485e-01
6.03650749e-01 1.18033409e+00 -5.97502172e-01 -4.62295622e-01
-3.44479591e-01 1.50583470e+00 1.56388372e-01 3.68549675e-01
-5.60841918e-01 -2.90896058e-01 2.90089101e-01 3.81910801e-01
-3.03488318e-02 -4.24461126e-01 -5.83574593e-01 -9.38647151e-01
2.63695866e-01 -1.49319100e+00 -2.81919748e-01 -1.03717470e+00
-8.57778490e-01 5.83775520e-01 -3.38971242e-02 -1.21509862e+00
-5.30943394e-01 -3.87264252e-01 -4.55567360e-01 1.18807721e+00
-8.15430105e-01 -1.32210433e+00 1.38276488e-01 9.09013450e-02
8.13604057e-01 2.24811569e-01 8.92814577e-01 3.26814055e-01
-3.06279540e-01 7.07562268e-01 6.22756593e-02 -3.23197693e-01
1.17591774e+00 -9.72089946e-01 7.36613810e-01 9.43548441e-01
1.17158808e-01 1.05808663e+00 9.17240739e-01 -1.06325293e+00
-1.78765392e+00 -7.57341862e-01 1.70477891e+00 -8.34673882e-01
4.47517931e-01 -5.94002068e-01 -9.10459518e-01 7.36982524e-01
6.85584784e-01 -8.55632663e-01 5.59714019e-01 2.19570160e-01
6.51689321e-02 2.18514234e-01 -9.06742513e-01 9.10451770e-01
1.10031438e+00 -8.09260190e-01 -6.48643851e-01 2.91441143e-01
5.80708683e-01 -5.88753402e-01 -7.39850044e-01 1.47452340e-01
6.85642421e-01 -4.47543383e-01 4.27117825e-01 -3.75207633e-01
5.45869052e-01 -1.96773157e-01 2.41390154e-01 -1.47260404e+00
-1.82524770e-01 -1.10992587e+00 -3.99064198e-02 1.42864490e+00
8.71271670e-01 -5.52278996e-01 4.16226327e-01 1.02780688e+00
-3.97532433e-01 -4.58369970e-01 -5.95493972e-01 -7.40022719e-01
5.06350137e-02 -5.63367009e-01 4.64093447e-01 9.75444496e-01
4.11722153e-01 6.38621449e-01 -5.51390827e-01 -1.18955210e-01
-1.38546586e-01 -2.87512660e-01 1.16667140e+00 -9.63570356e-01
-3.11650068e-01 -3.57212096e-01 2.52528340e-01 -9.16884303e-01
-1.60751984e-01 -8.58218610e-01 2.90707827e-01 -1.84289646e+00
9.64716673e-02 -2.32563078e-01 2.34211475e-01 3.91472191e-01
-2.31376603e-01 -1.78177506e-01 2.63251603e-01 4.97034788e-01
-1.77148879e-01 3.02687734e-01 1.30893409e+00 1.72054946e-01
-6.16056919e-01 -8.83337259e-02 -4.58307803e-01 2.78859764e-01
9.23722804e-01 -6.59049451e-01 -5.18436849e-01 -8.89889598e-01
4.63394523e-01 8.36410522e-02 -1.45091727e-01 -8.36777627e-01
4.52898443e-01 -1.72718719e-01 2.66148984e-01 -5.81373632e-01
-7.07179755e-02 -7.32297659e-01 6.55518651e-01 2.37273976e-01
-3.72814029e-01 7.26467788e-01 2.82579392e-01 -1.93677954e-02
-2.44940650e-02 -3.98265958e-01 1.87340170e-01 -1.08507887e-01
-1.53962404e-01 -2.88818091e-01 -1.02399290e+00 -6.45679653e-01
5.93068779e-01 -7.20543861e-01 -6.83618933e-02 -4.40140247e-01
-4.86451209e-01 -5.85431838e-03 7.05401540e-01 5.35629749e-01
3.94228578e-01 -9.43384886e-01 -5.38298666e-01 -5.90872578e-02
6.01949207e-02 -3.90447199e-01 -8.90437812e-02 9.20275509e-01
-6.34247482e-01 5.14245868e-01 -3.39352906e-01 -3.48838180e-01
-1.26573467e+00 4.25291151e-01 -1.42748384e-02 -4.03654665e-01
-5.01521647e-01 3.25169921e-01 -6.04142964e-01 -7.69567907e-01
9.50525254e-02 -2.01068714e-01 1.50967851e-01 -4.18158546e-02
4.04031813e-01 4.43235874e-01 4.54807729e-01 -2.44706288e-01
4.57488373e-02 7.07550943e-02 -3.69353294e-01 -8.69671822e-01
1.20965219e+00 -4.74173248e-01 -1.63942337e-01 6.29172981e-01
5.59955001e-01 -8.65141675e-02 -8.33113968e-01 -3.21318395e-02
5.11789680e-01 -5.07130206e-01 -2.43277565e-01 -1.32692254e+00
-8.31546709e-02 1.10908604e+00 6.19556606e-01 -2.83730835e-01
9.32836413e-01 -4.80404556e-01 7.91879535e-01 7.74308324e-01
5.10323405e-01 -1.67090344e+00 -2.10453570e-01 6.62464321e-01
8.71590376e-01 -1.01018023e+00 -1.64093092e-01 -3.47632468e-01
-6.77324772e-01 1.19554543e+00 6.33876741e-01 4.94076669e-01
1.08015470e-01 2.03031316e-01 3.63682866e-01 2.97052950e-01
-1.07035041e+00 3.22797924e-01 4.40661646e-02 6.06230617e-01
1.10732007e+00 9.84491259e-02 -8.60965550e-01 2.93674737e-01
-1.87690303e-01 4.91110414e-01 5.33876538e-01 1.31549942e+00
-1.51184529e-01 -1.74910843e+00 -3.80386472e-01 3.31484854e-01
-8.34758207e-02 -3.01516175e-01 -9.88554358e-01 7.91090488e-01
5.49656376e-02 1.30310440e+00 -2.98239976e-01 -5.74391842e-01
4.83575553e-01 5.03492832e-01 6.45977557e-01 -8.69632423e-01
-1.40634882e+00 8.94501656e-02 4.98529971e-01 -2.93801963e-01
-1.64226204e-01 -8.01650047e-01 -8.18817675e-01 -4.05309707e-01
-4.25225258e-01 1.30309075e-01 9.75019813e-01 9.96849537e-01
5.71176291e-01 9.01344121e-02 2.22963154e-01 -7.84406841e-01
-3.86969507e-01 -1.52127779e+00 3.51904631e-01 1.53474733e-01
-3.63539934e-01 -1.14412032e-01 1.04591213e-01 3.65796477e-01] | [11.63212776184082, 10.27624225616455] |
fa355070-ce4c-4cb5-acfa-7e0c574d5f3e | random-graph-matching-at-otter-s-threshold | 2209.12313 | null | https://arxiv.org/abs/2209.12313v2 | https://arxiv.org/pdf/2209.12313v2.pdf | Random graph matching at Otter's threshold via counting chandeliers | We propose an efficient algorithm for graph matching based on similarity scores constructed from counting a certain family of weighted trees rooted at each vertex. For two Erd\H{o}s-R\'enyi graphs $\mathcal{G}(n,q)$ whose edges are correlated through a latent vertex correspondence, we show that this algorithm correctly matches all but a vanishing fraction of the vertices with high probability, provided that $nq\to\infty$ and the edge correlation coefficient $\rho$ satisfies $\rho^2>\alpha \approx 0.338$, where $\alpha$ is Otter's tree-counting constant. Moreover, this almost exact matching can be made exact under an extra condition that is information-theoretically necessary. This is the first polynomial-time graph matching algorithm that succeeds at an explicit constant correlation and applies to both sparse and dense graphs. In comparison, previous methods either require $\rho=1-o(1)$ or are restricted to sparse graphs. The crux of the algorithm is a carefully curated family of rooted trees called chandeliers, which allows effective extraction of the graph correlation from the counts of the same tree while suppressing the undesirable correlation between those of different trees. | ['Sophie H. Yu', 'Jiaming Xu', 'Yihong Wu', 'Cheng Mao'] | 2022-09-25 | null | null | null | null | ['graph-matching'] | ['graphs'] | [ 3.59951913e-01 2.85532236e-01 -1.76880032e-01 -5.96635044e-03
-6.61554217e-01 -4.06918555e-01 4.56211269e-02 5.01140833e-01
-2.58999228e-01 3.92728448e-01 -5.48256755e-01 -5.35178900e-01
-6.01677775e-01 -1.40847552e+00 -4.76665705e-01 -7.30007052e-01
-7.74720669e-01 8.31406176e-01 4.49230492e-01 -3.32619905e-01
1.83903888e-01 3.50820959e-01 -1.53216767e+00 -3.81396830e-01
7.84229040e-01 9.59124923e-01 -5.15233390e-02 1.01040268e+00
-1.61373168e-01 7.76045322e-01 -1.86599284e-01 -8.39907527e-01
6.09104931e-01 -6.73500836e-01 -8.76951814e-01 -1.02729216e-01
4.84842122e-01 2.84967721e-01 -6.26377642e-01 1.61787915e+00
1.47332609e-01 -1.47500664e-01 6.38457060e-01 -1.32558310e+00
-2.39500761e-01 7.75672972e-01 -9.41541791e-01 3.50798428e-01
4.72756267e-01 -3.64142239e-01 1.47801995e+00 -3.79540682e-01
6.93622947e-01 8.92746627e-01 8.70288551e-01 -3.02051343e-02
-1.30018330e+00 -1.07486320e+00 -3.42365503e-01 1.49649903e-01
-1.76321101e+00 4.81928475e-02 7.86626577e-01 -2.90999651e-01
6.41370654e-01 5.58575153e-01 6.05663538e-01 1.37651771e-01
-4.84350771e-02 -4.10715863e-02 1.06129932e+00 -7.58717239e-01
-3.60225961e-02 -1.67873561e-01 4.29288685e-01 1.28297389e+00
7.96673357e-01 2.90395021e-01 -3.29092622e-01 -3.49154174e-01
4.65804875e-01 -2.13000804e-01 -5.52392974e-02 -5.19723058e-01
-8.07377219e-01 9.43345368e-01 3.94261837e-01 6.51793063e-01
1.11124516e-01 4.25870240e-01 1.24587610e-01 5.41329741e-01
2.81448007e-01 3.87720764e-02 -1.30099118e-01 3.36941957e-01
-8.00874114e-01 3.14049907e-02 8.75595987e-01 1.26645017e+00
1.19955337e+00 -2.06845090e-01 3.04437220e-01 3.69183630e-01
6.92039281e-02 6.88669086e-01 -4.86419499e-01 -8.85985911e-01
4.76876646e-01 8.10762048e-01 -2.38102645e-01 -1.52556002e+00
-2.80594647e-01 -4.83184785e-01 -1.12036300e+00 2.43674442e-01
7.93415308e-01 3.74028087e-01 -6.63211882e-01 2.05453992e+00
2.49856398e-01 1.01747938e-01 -4.12619889e-01 3.52866977e-01
4.21829611e-01 3.32561731e-01 -1.47297949e-01 -5.23313999e-01
1.42125690e+00 -3.72164905e-01 -4.39682484e-01 -8.22972357e-02
9.31228757e-01 -7.32653141e-01 6.87547982e-01 1.65584683e-01
-1.21676922e+00 -1.94566607e-01 -1.02041817e+00 2.40960583e-01
-9.78612304e-02 -3.99870753e-01 7.80345738e-01 1.17394710e+00
-1.20692182e+00 7.56131172e-01 -2.94964910e-01 -2.19548360e-01
9.18301046e-02 6.34553432e-01 -4.32547212e-01 -3.01301062e-01
-9.54811335e-01 5.91735959e-01 9.01604146e-02 -2.91945010e-01
-3.26146752e-01 -2.67513663e-01 -7.37030387e-01 8.39941502e-02
5.03475308e-01 -4.17485476e-01 7.14432001e-01 -8.02957773e-01
-6.76124454e-01 1.12782741e+00 -2.57969439e-01 -3.40427995e-01
2.57317215e-01 6.07983112e-01 -4.65753317e-01 4.68917727e-01
3.91054094e-01 -1.58550173e-01 5.15351176e-01 -1.02569747e+00
-5.79126000e-01 -6.85421407e-01 -1.24185771e-01 -2.34769806e-01
-8.61058291e-03 2.97165215e-01 -3.61568540e-01 -4.09102559e-01
4.66084152e-01 -9.08617318e-01 -5.27818799e-01 -2.31619745e-01
-2.25137681e-01 -5.53119145e-02 3.55512708e-01 -3.94103646e-01
1.40343225e+00 -2.09535551e+00 7.86664803e-03 1.22107077e+00
9.72831964e-01 -7.26010054e-02 -1.35143012e-01 6.69237196e-01
-3.54040802e-01 1.70588374e-01 -3.66140485e-01 9.59578902e-02
-1.50837526e-01 9.21642184e-02 2.12963849e-01 9.26139593e-01
-3.63395452e-01 3.48625541e-01 -9.36635494e-01 -6.58930898e-01
-9.89872962e-02 4.75431383e-02 -4.31512237e-01 -4.28838469e-02
2.59892810e-02 -1.48572102e-01 -1.66146174e-01 4.76897687e-01
9.19453859e-01 -5.63219368e-01 7.69771934e-01 2.20955834e-01
2.55223997e-02 1.72793642e-01 -1.59375095e+00 1.02031016e+00
3.26494612e-02 1.96058661e-01 3.81241322e-01 -1.15066504e+00
1.03947735e+00 1.47750258e-01 6.65841222e-01 -6.89691305e-01
4.55809891e-01 5.34808218e-01 1.46085799e-01 1.29275069e-01
3.59177232e-01 -2.51220942e-01 -2.08757013e-01 5.02777219e-01
-1.33650601e-01 2.28988379e-03 5.96385717e-01 6.06006026e-01
1.84746993e+00 -8.06133628e-01 3.98846388e-01 -6.43943667e-01
5.13612986e-01 -6.73986226e-02 5.29589891e-01 8.85401189e-01
-1.12619668e-01 2.42084742e-01 7.78386474e-01 -4.88886870e-02
-1.14852214e+00 -1.04216003e+00 -1.01598546e-01 1.01307082e+00
4.43451643e-01 -8.04814577e-01 -7.67543435e-01 -3.81124824e-01
3.92073393e-03 2.89686143e-01 -6.67754412e-01 -1.54670700e-01
-4.73997295e-01 -7.33247340e-01 3.53718877e-01 4.00465161e-01
2.53546834e-01 -5.87942958e-01 -2.16672212e-01 1.42610490e-01
-4.61086631e-02 -9.87095416e-01 -4.96359885e-01 4.90841717e-01
-8.88487577e-01 -1.47457755e+00 -2.10068852e-01 -9.67523575e-01
8.86430860e-01 5.54515362e-01 1.39326441e+00 6.43877387e-01
-3.82977337e-01 5.49881682e-02 -2.38503039e-01 1.77055225e-01
-4.62055624e-01 -2.07147710e-02 5.03510982e-02 -4.13291603e-01
5.16422570e-01 -9.60500002e-01 -3.58862460e-01 3.97535801e-01
-4.93173242e-01 -3.21390271e-01 3.07326019e-01 7.88478374e-01
5.40717125e-01 4.72106904e-01 -1.37204736e-01 -1.03667819e+00
1.26898646e-01 -3.54592502e-01 -1.05227256e+00 2.15547591e-01
-8.69452477e-01 2.96610892e-01 6.22822583e-01 -2.05475330e-01
-2.24619612e-01 1.08579710e-01 5.48281595e-02 -4.38325927e-02
3.96608800e-01 2.44170040e-01 -4.43257205e-02 -5.17925858e-01
6.97730422e-01 -1.50345445e-01 -1.86791718e-01 -2.14581251e-01
2.78600812e-01 2.12657124e-01 5.58211207e-01 -4.86561477e-01
1.16804552e+00 5.17667890e-01 7.90638685e-01 -6.73768759e-01
-5.33048868e-01 -6.20819807e-01 -2.33552992e-01 -2.10986263e-03
4.93099868e-01 -5.04630685e-01 -1.27229905e+00 1.16927184e-01
-9.33784604e-01 2.55258586e-02 -3.84326458e-01 4.83346790e-01
-4.45283085e-01 7.93822229e-01 -6.23337865e-01 -1.18375087e+00
-8.50096047e-02 -7.55946279e-01 6.16480410e-01 -2.56633222e-01
-1.70452565e-01 -6.99820995e-01 3.10301244e-01 1.36766225e-01
7.86147639e-02 2.52483457e-01 1.35197079e+00 -5.70101380e-01
-7.05193996e-01 -6.16502762e-01 -6.57332182e-01 -3.40423584e-02
-1.67542458e-01 1.62959844e-03 -3.99526298e-01 -2.75409669e-01
-1.40606120e-01 1.34408116e-01 7.10779905e-01 1.62741244e-01
9.11584496e-01 -3.64423454e-01 -5.48888385e-01 4.00439709e-01
1.59002554e+00 7.16276839e-02 5.81244707e-01 -2.01418102e-01
5.09646833e-01 4.20681953e-01 1.70153439e-01 2.56258160e-01
1.68700129e-01 4.66557205e-01 3.99280190e-01 -4.16356549e-02
1.06428020e-01 -2.48179525e-01 1.01464227e-01 1.29024875e+00
-3.41018707e-01 -1.24894239e-01 -8.63120914e-01 5.03947198e-01
-1.53122377e+00 -1.10761881e+00 -8.85518789e-01 2.55847430e+00
6.12165570e-01 4.17706341e-01 3.28315556e-01 5.85802913e-01
1.09757388e+00 9.08406451e-02 -9.79834981e-03 -4.89250034e-01
-1.72453821e-01 1.15092969e+00 1.11486387e+00 8.62285733e-01
-4.53326195e-01 5.40379405e-01 6.69528294e+00 1.06891835e+00
-1.82558417e-01 1.07752949e-01 3.41352940e-01 3.27103674e-01
-6.38628602e-01 6.16551459e-01 -5.22254050e-01 3.66135716e-01
1.03980637e+00 -3.32251072e-01 4.66432244e-01 7.94802189e-01
-5.57621956e-01 -3.39286268e-01 -8.27706099e-01 8.92744422e-01
-1.47669852e-01 -1.07793522e+00 -4.47148412e-01 4.12090361e-01
6.11269414e-01 -2.70996511e-01 -1.70925692e-01 -8.43116827e-03
9.23250198e-01 -1.06636477e+00 2.67200410e-01 9.78751183e-02
1.07896066e+00 -9.22477245e-01 5.37431717e-01 3.14283043e-01
-1.97283840e+00 3.90683077e-02 -2.54230797e-01 -3.66201028e-02
-1.44615144e-01 9.16326404e-01 -4.10343587e-01 7.48810291e-01
6.16352201e-01 -6.69818446e-02 -2.04281032e-01 8.27182174e-01
1.12526350e-01 3.54703218e-01 -7.23107696e-01 -1.02640912e-02
-3.20815556e-02 -6.88186646e-01 4.81889695e-01 9.40165401e-01
5.25236845e-01 4.88568038e-01 5.82897738e-02 4.62960124e-01
-2.25267738e-01 4.20805931e-01 -8.04044425e-01 3.86641324e-02
5.14176369e-01 9.57928061e-01 -1.06705284e+00 -2.06555188e-01
-3.72744620e-01 6.18763566e-01 5.09430647e-01 -2.05130845e-01
-6.64141893e-01 -6.77376807e-01 4.11396444e-01 5.13124883e-01
3.25478196e-01 -4.33897614e-01 -2.58481264e-01 -8.50729823e-01
-3.33392955e-02 -8.07372332e-01 6.67219579e-01 -2.66782105e-01
-1.28551388e+00 6.44119680e-01 -1.27777666e-01 -7.42289603e-01
-4.39326316e-02 -6.27923489e-01 -3.38079751e-01 8.14622462e-01
-9.79975641e-01 -5.18601239e-01 -1.73209131e-01 7.21745610e-01
-5.71807742e-01 4.82519455e-02 9.34958816e-01 4.30127740e-01
-1.68302253e-01 7.97514200e-01 3.28018293e-02 4.89429664e-03
1.01776928e-01 -1.23945951e+00 2.19089478e-01 9.50423837e-01
4.22694653e-01 2.59638608e-01 9.35455799e-01 -8.06044161e-01
-1.30591083e+00 -5.46594024e-01 1.39132881e+00 -2.74932802e-01
8.37560654e-01 -5.41197360e-01 -7.52548933e-01 4.09273505e-01
-1.19654603e-01 1.62670761e-01 7.16928124e-01 3.29420209e-01
-7.38493443e-01 -2.25752711e-01 -1.29400289e+00 5.08493066e-01
1.63272524e+00 -5.99015415e-01 -1.23755217e-01 4.45164293e-01
4.64251697e-01 -1.20807558e-01 -9.83912468e-01 4.70789135e-01
3.21927875e-01 -1.11775184e+00 8.75469208e-01 -4.05038208e-01
-1.17264949e-01 -2.28566960e-01 -4.47916150e-01 -3.60427707e-01
-5.41293979e-01 -8.86419594e-01 1.49201602e-01 7.84741700e-01
3.63314152e-01 -5.30952096e-01 1.04900908e+00 1.44119099e-01
3.90492260e-01 -3.02970290e-01 -1.28048563e+00 -1.01857233e+00
-1.06321201e-01 -6.22651219e-01 5.67443490e-01 9.88406777e-01
2.50524819e-01 2.40620032e-01 -2.26210713e-01 3.17770720e-01
1.10026062e+00 1.32611379e-01 7.06779361e-01 -1.67798805e+00
-5.36565125e-01 -4.98051941e-01 -7.39282966e-01 -7.00120687e-01
-2.67933793e-02 -1.12770343e+00 -3.81179452e-01 -1.12713015e+00
5.31229079e-01 -8.63010228e-01 -4.79997247e-02 1.88259706e-01
4.16871831e-02 3.52254331e-01 -1.25261426e-01 -3.57876346e-02
-4.95405912e-01 1.10396467e-01 8.90281677e-01 1.74993142e-01
1.83165997e-01 1.27599165e-01 -5.87120116e-01 6.91652060e-01
5.32882869e-01 -8.22229743e-01 -3.16662639e-02 2.76743323e-01
7.97626972e-01 6.21724963e-01 2.97090858e-01 -8.83107781e-01
2.82732248e-01 -1.10250777e-02 -2.83283949e-01 -4.32655036e-01
1.04619242e-01 -9.47734594e-01 6.33403778e-01 6.97893858e-01
-4.21462171e-02 4.39318866e-01 -3.84903342e-01 7.70128369e-01
3.67853642e-02 -4.26451713e-01 9.57670212e-01 -3.52188721e-02
-1.07951090e-01 4.87486809e-01 -7.40654022e-02 1.92557454e-01
9.89522755e-01 -4.29430604e-01 -2.57218719e-01 -4.97873604e-01
-4.72867489e-01 -3.90946269e-02 5.79336762e-01 -4.44346279e-01
2.20841974e-01 -1.24980187e+00 -5.97802877e-01 1.61722213e-01
1.61023095e-01 -2.70136058e-01 5.85955083e-02 8.82252574e-01
-5.58697164e-01 -4.74090651e-02 7.59684592e-02 -4.71380353e-01
-1.46932757e+00 7.14803934e-01 2.34548330e-01 -6.37385309e-01
-5.20413041e-01 8.69101465e-01 -1.20266207e-01 -5.97969890e-02
2.67397147e-03 -7.16915578e-02 4.64023888e-01 -2.49110028e-01
1.54067263e-01 6.20243371e-01 6.82437718e-02 -7.79550076e-01
-4.74651158e-01 7.54981041e-01 1.41809043e-02 -8.23913366e-02
9.90730524e-01 2.85831653e-03 -4.34954762e-01 5.17043509e-02
1.24758279e+00 5.66727042e-01 -5.74560702e-01 -3.56258810e-01
9.80525911e-02 -7.35526741e-01 -4.26347166e-01 -1.20646872e-01
-1.18879104e+00 5.90493500e-01 4.76636857e-01 8.09139729e-01
1.03927660e+00 2.34509706e-01 7.97368824e-01 1.39408857e-01
7.15806842e-01 -9.02164042e-01 -1.29786193e-01 3.80428672e-01
4.20348160e-02 -8.11343968e-01 1.48936421e-01 -9.67404485e-01
-3.81981470e-02 8.69645834e-01 1.66170269e-01 -5.04236341e-01
7.04512715e-01 4.78777826e-01 -6.84459686e-01 -4.76382822e-01
-4.37724650e-01 -6.11875653e-01 7.49809146e-02 5.32069325e-01
2.59555757e-01 1.70952827e-01 -7.06568897e-01 2.96995133e-01
-4.97401416e-01 -3.03336084e-01 2.14213088e-01 7.03870773e-01
-7.81951427e-01 -1.43567753e+00 -2.48337388e-01 5.79934061e-01
-4.31410670e-01 -2.38561332e-01 -3.61572593e-01 1.01993358e+00
1.27695695e-01 9.11291778e-01 -5.20550087e-02 -4.80172038e-01
1.55314073e-01 -1.19610518e-01 7.42253900e-01 -3.68330538e-01
-5.81198335e-01 -1.53223306e-01 1.41692340e-01 -4.73588943e-01
-3.34905744e-01 -5.34900904e-01 -1.24625969e+00 -1.20903146e+00
-6.84602320e-01 4.80013132e-01 2.84146011e-01 7.02607572e-01
-4.11923975e-02 4.53458354e-03 9.26146746e-01 -9.67685878e-02
-4.31025147e-01 -4.76421475e-01 -1.19136429e+00 3.19536477e-01
-1.74146071e-01 -5.91749966e-01 -8.76013339e-01 -4.91032273e-01] | [6.814311504364014, 5.16959810256958] |
6acefab2-1bd4-4ccd-b755-afa97ca9a2a4 | weighted-sparse-subspace-representation-a | 2106.04330 | null | https://arxiv.org/abs/2106.04330v1 | https://arxiv.org/pdf/2106.04330v1.pdf | Weighted Sparse Subspace Representation: A Unified Framework for Subspace Clustering, Constrained Clustering, and Active Learning | Spectral-based subspace clustering methods have proved successful in many challenging applications such as gene sequencing, image recognition, and motion segmentation. In this work, we first propose a novel spectral-based subspace clustering algorithm that seeks to represent each point as a sparse convex combination of a few nearby points. We then extend the algorithm to constrained clustering and active learning settings. Our motivation for developing such a framework stems from the fact that typically either a small amount of labelled data is available in advance; or it is possible to label some points at a cost. The latter scenario is typically encountered in the process of validating a cluster assignment. Extensive experiments on simulated and real data sets show that the proposed approach is effective and competitive with state-of-the-art methods. | ['Nicos G. Pavlidis', 'Hankui Peng'] | 2021-06-08 | null | null | null | null | ['motion-segmentation'] | ['computer-vision'] | [ 3.48160595e-01 -2.20793232e-01 -3.05956781e-01 -2.38613307e-01
-8.89011800e-01 -7.15674996e-01 4.39100713e-01 2.76715606e-01
-5.11185050e-01 4.88023371e-01 -1.99731410e-01 -7.14120567e-02
-3.22121739e-01 -2.52022445e-01 -4.69143301e-01 -1.02971184e+00
-3.00425917e-01 8.73805642e-01 1.60886750e-01 4.46413368e-01
2.57221758e-01 5.34094572e-01 -1.19861650e+00 -1.12051278e-01
1.05954170e+00 5.82333863e-01 2.84456253e-01 4.48112041e-01
-3.26899350e-01 1.77635640e-01 -2.01310143e-01 6.07901886e-02
3.89531225e-01 -4.17727351e-01 -6.80257976e-01 7.52487242e-01
-9.35892910e-02 3.68593544e-01 2.49240473e-01 1.07498908e+00
2.59483099e-01 4.66720074e-01 7.20658779e-01 -1.23459363e+00
2.26542592e-01 2.02718824e-01 -8.33693147e-01 -1.09702080e-01
2.13891268e-01 -1.49694696e-01 6.30146563e-01 -1.07745278e+00
6.80848479e-01 7.93306649e-01 4.49381232e-01 2.80568659e-01
-1.59058666e+00 -2.97173917e-01 1.27669454e-01 1.89274386e-01
-1.78949142e+00 -6.07588232e-01 9.83014643e-01 -6.93982005e-01
4.12736714e-01 3.49351853e-01 5.95336318e-01 6.46788359e-01
-4.24919188e-01 9.70073342e-01 9.72746193e-01 -5.67165732e-01
9.50766444e-01 5.01626991e-02 1.54977307e-01 4.68612462e-01
1.11569069e-01 -2.69271225e-01 -4.43476796e-01 -5.77382565e-01
5.42210281e-01 1.65359065e-01 -3.48982960e-01 -1.29111958e+00
-1.22840786e+00 7.80254364e-01 7.96180964e-02 2.47959107e-01
-4.93477970e-01 -2.81202346e-01 2.01712772e-01 -1.29775509e-01
4.07197207e-01 2.74421275e-01 -2.67849546e-02 -8.16203728e-02
-1.48699987e+00 -2.56146658e-02 7.64430523e-01 8.99664700e-01
7.91237175e-01 -1.45070791e-01 3.90498221e-01 8.07633221e-01
3.27612758e-01 1.78561676e-02 2.18298957e-01 -1.15393162e+00
9.76764187e-02 6.31074309e-01 3.87993664e-01 -7.30102360e-01
-1.89131245e-01 -1.67576209e-01 -7.95749187e-01 6.41656220e-02
5.33186734e-01 6.01863340e-02 -7.85510540e-01 1.43763804e+00
6.00720644e-01 9.03116167e-01 -6.71793297e-02 1.02247000e+00
2.27858841e-01 5.42334378e-01 -8.13221857e-02 -1.07362306e+00
5.25883079e-01 -6.45638585e-01 -6.44534886e-01 3.82820107e-02
3.32828432e-01 -7.81063795e-01 5.86684585e-01 5.34842908e-01
-7.85526216e-01 -3.16682160e-01 -7.20266640e-01 4.15353209e-01
-1.52464986e-01 1.71094701e-01 6.31374657e-01 7.58259356e-01
-8.13987672e-01 2.96847403e-01 -1.13747752e+00 -6.98987365e-01
2.66349941e-01 5.66934049e-01 -2.60359704e-01 -1.40615180e-01
-4.15167421e-01 2.33928427e-01 4.31227326e-01 2.60601163e-01
-5.57764649e-01 -2.48885095e-01 -5.81708968e-01 -9.33490917e-02
6.52243972e-01 -4.42765892e-01 8.36814523e-01 -8.04992497e-01
-1.29914427e+00 7.38249838e-01 -5.93377650e-01 -2.55741984e-01
5.47829032e-01 -9.21527669e-02 -1.48610681e-01 2.57780075e-01
5.90586700e-02 4.06209201e-01 7.90065289e-01 -1.32347083e+00
-3.27402532e-01 -5.07138610e-01 -4.21017230e-01 3.77545327e-01
-2.73098916e-01 1.14525639e-01 -8.74860168e-01 -4.22652453e-01
5.72868586e-01 -1.27874613e+00 -7.42639065e-01 -1.04772843e-01
-4.70254898e-01 -2.46038884e-01 9.90206659e-01 -2.06804261e-01
8.32307339e-01 -2.05139160e+00 6.35900080e-01 7.25797653e-01
3.98824774e-02 1.75502494e-01 1.85153365e-01 5.28923512e-01
-2.07369745e-01 -2.49909863e-01 -6.99104965e-01 -3.34697902e-01
-2.80152023e-01 9.45421383e-02 -1.99581444e-01 9.63294208e-01
1.41441934e-02 4.20714617e-01 -1.03390324e+00 -7.80524373e-01
3.83185089e-01 2.53714889e-01 -3.89602691e-01 1.64325330e-02
-4.36805263e-02 6.73949003e-01 -4.05496806e-01 6.57680392e-01
5.81959546e-01 -2.92655021e-01 4.53336507e-01 9.09932703e-02
-1.31629035e-01 -2.86602169e-01 -1.78588581e+00 1.95205188e+00
3.20858181e-01 3.57807547e-01 3.09161752e-01 -1.52593446e+00
6.23865724e-01 3.92436862e-01 1.04479432e+00 2.26523489e-01
-3.82142663e-02 1.87978432e-01 -1.08773503e-02 -1.65116891e-01
3.15355211e-01 4.53639552e-02 8.40961486e-02 4.33428138e-01
-1.27145126e-01 -5.35429269e-02 4.74774539e-01 2.72961795e-01
8.64310861e-01 2.33863071e-01 3.70422602e-01 -2.86557138e-01
6.78011179e-01 3.73814374e-01 9.30972576e-01 4.82933104e-01
-3.43491912e-01 6.56786203e-01 2.58355945e-01 1.83240339e-01
-9.39936578e-01 -1.06727433e+00 -1.13810383e-01 7.44350612e-01
1.62916422e-01 -3.27312589e-01 -9.27030623e-01 -5.56701481e-01
-3.13212872e-01 3.94770652e-01 -2.08888024e-01 3.00920099e-01
-4.40570265e-01 -8.26920390e-01 -8.71392116e-02 3.48968089e-01
1.55622363e-01 -5.16845167e-01 -3.98931235e-01 2.45815650e-01
-4.63225245e-02 -8.49979043e-01 -4.40640122e-01 2.19916090e-01
-1.05422127e+00 -1.08635283e+00 -7.83087313e-01 -8.45043957e-01
1.04970634e+00 5.50214589e-01 5.51141322e-01 -2.23179191e-01
-2.44262815e-01 4.15683150e-01 -3.54391277e-01 -1.22453071e-01
-7.63880983e-02 -3.20671648e-02 4.67793584e-01 3.26736003e-01
5.20855010e-01 -3.82417232e-01 -3.40510994e-01 3.11785221e-01
-7.99341500e-01 -1.76649660e-01 3.05862874e-01 7.74950624e-01
1.03576720e+00 4.62797552e-01 3.46129715e-01 -1.02562380e+00
4.01191503e-01 -4.44127202e-01 -8.23569417e-01 3.01856846e-01
-4.25163269e-01 -3.15473467e-01 5.43095410e-01 -4.79904115e-01
-7.97249615e-01 1.17107069e+00 3.93489510e-01 -8.48932028e-01
-4.58996803e-01 5.92284739e-01 -3.24934661e-01 -5.21590635e-02
4.83060569e-01 3.37935656e-01 1.31678045e-01 -4.17590946e-01
3.88526350e-01 9.05623794e-01 5.42068064e-01 -4.34094399e-01
9.07543182e-01 6.71554387e-01 2.34860420e-01 -1.14651477e+00
-2.74637043e-01 -1.31111085e+00 -1.13592088e+00 -3.06542069e-01
4.75489885e-01 -7.89005518e-01 -6.08177662e-01 2.30945051e-01
-6.48575246e-01 -1.57999862e-02 -4.24634814e-02 5.18537819e-01
-7.93914914e-01 7.42143869e-01 -1.84663385e-01 -1.08239973e+00
1.03019640e-01 -1.31103778e+00 7.97168672e-01 1.33133560e-01
-4.07772332e-01 -9.08538401e-01 1.90503135e-01 4.66524035e-01
-1.75253376e-01 2.61396050e-01 5.97547889e-01 -7.33834267e-01
-5.26510060e-01 -1.72658086e-01 3.37695599e-01 -1.01424985e-01
2.34151065e-01 1.03791982e-01 -7.20579922e-01 -6.17716610e-01
2.32588053e-02 -1.17267534e-01 6.70432329e-01 4.00886893e-01
1.03924859e+00 1.71802133e-01 -6.76253259e-01 4.13855404e-01
1.31047595e+00 3.15358520e-01 2.78155833e-01 -1.71215057e-01
7.25595951e-01 7.48007476e-01 8.67167473e-01 1.76469415e-01
-1.41369119e-01 7.59717166e-01 -6.88849092e-02 -2.23128408e-01
4.58676517e-01 1.93662301e-01 -1.86803006e-02 9.64977443e-01
1.53263196e-01 2.49386691e-02 -1.21250224e+00 6.78562880e-01
-2.34204435e+00 -9.56128240e-01 -2.79733211e-01 2.53762889e+00
5.63988626e-01 -2.20575780e-01 5.16915858e-01 4.16770965e-01
9.67151582e-01 -2.76685953e-01 -8.01147580e-01 -3.39233316e-04
-1.01628229e-01 -1.68289185e-01 4.18727458e-01 3.18294257e-01
-1.32316136e+00 7.56814420e-01 6.31644106e+00 7.81251848e-01
-9.95736659e-01 -1.36473149e-01 5.18024564e-01 -6.87065870e-02
1.38379276e-01 1.96216986e-01 -3.57066065e-01 4.07382905e-01
5.67221940e-01 -1.98280871e-01 4.13090408e-01 7.89908886e-01
4.13415611e-01 -3.50257397e-01 -1.19638681e+00 1.19955850e+00
-8.71938914e-02 -1.16399515e+00 -3.10505927e-01 2.05857113e-01
8.89703989e-01 -1.42288744e-01 -7.81412497e-02 -4.12319213e-01
2.72374898e-01 -7.85808742e-01 3.20521742e-01 4.67993617e-01
5.73248327e-01 -9.76605773e-01 3.79096746e-01 7.64089584e-01
-1.08266234e+00 -2.75107510e-02 -3.51247370e-01 1.87967584e-01
6.50943369e-02 6.23894393e-01 -1.04565287e+00 5.99507093e-01
4.94432956e-01 6.26570821e-01 -3.23609799e-01 1.55481601e+00
5.98764792e-02 8.78201544e-01 -5.52952349e-01 8.93037692e-02
1.67104974e-01 -8.14290404e-01 7.54866242e-01 1.14089108e+00
1.79312393e-01 2.93983102e-01 5.22336423e-01 7.30548024e-01
2.58678913e-01 3.98677737e-01 -6.56998992e-01 -2.47240022e-01
6.72471941e-01 1.38874376e+00 -1.38189995e+00 -1.59477741e-01
-5.17088950e-01 1.16032314e+00 3.23230863e-01 6.23429894e-01
-2.55210668e-01 -1.41439050e-01 3.54349285e-01 -7.96081126e-02
3.25802028e-01 -6.46087706e-01 -2.94417322e-01 -1.13076973e+00
-2.73329943e-01 -6.54437482e-01 4.72022355e-01 -5.12414396e-01
-9.99665916e-01 -4.01528105e-02 5.52988946e-02 -1.36599803e+00
-5.04095793e-01 -3.16262513e-01 -4.96220678e-01 7.61576951e-01
-7.69098043e-01 -6.73832774e-01 -6.57423073e-03 6.03118956e-01
5.06021678e-01 -3.32738876e-01 7.87148595e-01 1.77076161e-02
-8.55787694e-01 1.04105681e-01 7.77592838e-01 6.03925958e-02
5.99381626e-01 -1.30959857e+00 -5.20614572e-02 1.21566010e+00
4.95204687e-01 1.01702416e+00 7.90815115e-01 -6.14619374e-01
-1.55302441e+00 -8.61974239e-01 5.37318528e-01 -1.86527133e-01
3.81950945e-01 -4.71041858e-01 -9.16953206e-01 3.92870605e-01
-4.48062718e-02 -6.32095560e-02 1.05295408e+00 1.81731597e-01
1.79245979e-01 -6.35641888e-02 -1.15922379e+00 5.15037119e-01
7.13826895e-01 -4.07929868e-01 -3.73558104e-01 4.66374069e-01
1.47827035e-02 -1.40006512e-01 -6.47055089e-01 3.40010494e-01
7.59084672e-02 -7.22815275e-01 9.56449389e-01 -4.17015910e-01
-3.32489491e-01 -8.34084451e-01 -1.17004722e-01 -1.26548958e+00
-3.69482130e-01 -8.65833998e-01 -6.31846115e-02 1.11646116e+00
1.38764158e-01 -2.16313750e-01 1.33450484e+00 6.18345559e-01
5.30433133e-02 -5.29358327e-01 -1.11769879e+00 -7.07955360e-01
-2.25449666e-01 -3.08175743e-01 1.27003461e-01 1.32713282e+00
2.43742406e-01 2.39636347e-01 -1.38654992e-01 2.59882152e-01
1.02771294e+00 4.72998857e-01 8.08302522e-01 -1.39064479e+00
-2.33642116e-01 -1.40069842e-01 -4.94759470e-01 -8.63242447e-01
2.87808597e-01 -6.12643778e-01 1.82746336e-01 -1.40418780e+00
4.89977002e-01 -6.26624227e-01 -2.36355513e-01 1.83970392e-01
-3.33329648e-01 2.03077942e-01 -4.40834239e-02 4.90939051e-01
-8.38830888e-01 1.31938934e-01 5.26495039e-01 -2.94387694e-02
-5.26649654e-01 1.85582086e-01 -3.01268727e-01 6.92950130e-01
6.28932536e-01 -2.00178206e-01 -6.43030107e-01 1.15357883e-01
-1.71823129e-01 2.22433954e-01 4.19994593e-02 -8.22934568e-01
6.82953000e-01 -3.97344142e-01 3.31936508e-01 -8.44745457e-01
3.80288154e-01 -1.03786480e+00 5.22455454e-01 2.16325060e-01
-1.02506831e-01 -4.74715173e-01 -1.88212648e-01 8.94292593e-01
-1.67983249e-01 -2.80425787e-01 8.59549344e-01 -1.18418671e-01
-9.45040762e-01 9.78289545e-02 -3.36282581e-01 -2.77782887e-01
1.25670469e+00 -6.19211793e-01 3.27390254e-01 -2.66677350e-01
-8.24908555e-01 3.42587858e-01 8.21812689e-01 1.56444654e-01
5.39515257e-01 -1.15237761e+00 -3.83462578e-01 1.06685869e-01
1.54968724e-01 1.22546203e-01 -9.90038514e-02 1.04965961e+00
-2.87461251e-01 5.29327750e-01 1.45214006e-01 -1.07172132e+00
-1.54076779e+00 8.78064990e-01 -1.19494759e-02 3.67064089e-01
-2.68436432e-01 6.05060816e-01 1.16952628e-01 -3.43221486e-01
2.49759898e-01 2.88430125e-01 -1.35140225e-01 7.72041604e-02
1.48605496e-01 6.70609951e-01 -6.52376860e-02 -1.06377101e+00
-4.85381275e-01 6.72205925e-01 6.38170093e-02 -1.95890144e-01
1.26156592e+00 -2.43786067e-01 -1.23342641e-01 7.12655485e-01
1.06878757e+00 -3.60185355e-02 -1.14594197e+00 -4.53949630e-01
3.56282055e-01 -3.90027136e-01 9.81760770e-02 -2.59946138e-01
-7.19043732e-01 5.57726800e-01 6.53596103e-01 1.54394612e-01
1.23613501e+00 7.67367706e-03 3.70609105e-01 5.25170326e-01
5.95957458e-01 -1.38541472e+00 -1.94967791e-01 2.64748987e-02
2.63632357e-01 -1.19635916e+00 1.70349836e-01 -8.74932468e-01
-5.09008169e-01 7.90010452e-01 2.07539722e-01 -3.89107279e-02
4.60853130e-01 9.32382345e-02 6.83968365e-02 1.09800110e-02
-4.97568130e-01 -3.39318097e-01 3.30295146e-01 5.88576436e-01
4.47367489e-01 1.01046093e-01 -3.61190796e-01 3.86478975e-02
1.69733867e-01 -1.85813587e-02 3.72838646e-01 1.08396518e+00
-4.74483967e-01 -1.19376445e+00 -8.00789237e-01 3.35043460e-01
-3.37900519e-01 2.76062548e-01 -7.08274543e-01 4.10783201e-01
-3.25068273e-02 1.17126286e+00 8.11396837e-02 6.76506907e-02
-1.48471713e-01 2.39152893e-01 3.96092892e-01 -6.84337676e-01
-2.97302920e-02 9.66994405e-01 -1.58259973e-01 -4.98570234e-01
-8.02530527e-01 -1.25068545e+00 -1.44000661e+00 2.00382203e-01
-3.90869260e-01 5.85489750e-01 6.31545007e-01 7.29473114e-01
2.49878034e-01 -4.58656214e-02 8.31438184e-01 -6.13234639e-01
-1.42205238e-01 -5.85859418e-01 -7.96070397e-01 3.91466528e-01
8.32034871e-02 -6.18213594e-01 -2.06165165e-01 4.54814792e-01] | [7.764566898345947, 4.424886703491211] |
4a7e690a-f0ab-42b6-8144-d125259ec315 | recurrent-attention-models-for-depth-based | 1611.07212 | null | http://arxiv.org/abs/1611.07212v1 | http://arxiv.org/pdf/1611.07212v1.pdf | Recurrent Attention Models for Depth-Based Person Identification | We present an attention-based model that reasons on human body shape and
motion dynamics to identify individuals in the absence of RGB information,
hence in the dark. Our approach leverages unique 4D spatio-temporal signatures
to address the identification problem across days. Formulated as a
reinforcement learning task, our model is based on a combination of
convolutional and recurrent neural networks with the goal of identifying small,
discriminative regions indicative of human identity. We demonstrate that our
model produces state-of-the-art results on several published datasets given
only depth images. We further study the robustness of our model towards
viewpoint, appearance, and volumetric changes. Finally, we share insights
gleaned from interpretable 2D, 3D, and 4D visualizations of our model's
spatio-temporal attention. | ['Li Fei-Fei', 'Albert Haque', 'Alexandre Alahi'] | 2016-11-22 | recurrent-attention-models-for-depth-based-1 | http://openaccess.thecvf.com/content_cvpr_2016/html/Haque_Recurrent_Attention_Models_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Haque_Recurrent_Attention_Models_CVPR_2016_paper.pdf | cvpr-2016-6 | ['person-identification'] | ['computer-vision'] | [-1.65676102e-01 -1.49124563e-01 1.17643684e-01 -1.95563257e-01
-2.23303437e-01 -5.55662036e-01 6.30360126e-01 1.01553552e-01
-2.59263128e-01 2.01682776e-01 6.51635945e-01 1.89872041e-01
-6.00813963e-02 -5.88444412e-01 -3.49543542e-01 -2.27552816e-01
-6.22866511e-01 2.75891334e-01 -2.31629729e-01 -1.76837891e-01
1.26741379e-01 7.16419995e-01 -1.51652026e+00 4.95122820e-02
2.39490867e-01 9.50548708e-01 -6.02806032e-01 7.14551091e-01
4.50154483e-01 6.60002410e-01 -3.98780018e-01 -4.11555737e-01
6.48605287e-01 -4.63899642e-01 -6.38264239e-01 2.70156652e-01
9.57578063e-01 -5.90140462e-01 -7.88279474e-01 4.60565180e-01
6.34639561e-01 1.61028698e-01 4.76672918e-01 -1.12202096e+00
-1.06048179e+00 -2.65363038e-01 -5.81286848e-01 5.50864935e-01
6.79903924e-01 8.04032147e-01 7.66977668e-01 -5.53625107e-01
6.29297674e-01 1.25577867e+00 1.10979414e+00 5.42840540e-01
-1.16933072e+00 -3.26464027e-01 3.10789227e-01 4.87567894e-02
-1.25059569e+00 -3.05803686e-01 8.90098989e-01 -7.69995809e-01
9.90258098e-01 4.88870293e-02 1.18846214e+00 1.23750484e+00
-1.13027461e-01 7.01049507e-01 9.99730229e-01 -9.47628394e-02
-7.85357207e-02 -4.28209007e-01 -1.15585342e-01 9.63265359e-01
1.60452574e-01 4.92583901e-01 -7.85207689e-01 1.43630862e-01
1.20722413e+00 7.91870058e-02 -3.87323238e-02 -6.03706777e-01
-1.15697479e+00 6.22624755e-01 6.96068108e-01 -1.21221192e-01
-4.07332540e-01 5.44637024e-01 1.98371813e-01 -1.43513814e-01
6.66060030e-01 4.45252687e-01 -2.44962052e-01 -2.24963874e-01
-7.97221243e-01 4.31428254e-01 3.10376018e-01 5.20793915e-01
4.77003008e-01 1.98210910e-01 -3.13908994e-01 4.93567973e-01
3.05542707e-01 7.02635884e-01 2.51129299e-01 -1.13373458e+00
6.86968938e-02 7.78042912e-01 5.44560440e-02 -1.21180522e+00
-7.90960908e-01 -4.36118245e-01 -6.61716402e-01 4.00771320e-01
4.67683673e-01 -1.26347035e-01 -8.63704622e-01 1.70645857e+00
3.98380011e-01 2.67718673e-01 -4.37125474e-01 1.26848686e+00
7.21677840e-01 1.74894314e-02 2.04128608e-01 4.60396111e-01
9.87989306e-01 -7.90778458e-01 -1.34625256e-01 -2.59854108e-01
1.67753085e-01 2.91815829e-02 7.02103078e-01 9.76749603e-03
-1.09692049e+00 -7.31164455e-01 -7.01863527e-01 -1.91174150e-01
-5.29318631e-01 2.61192381e-01 6.43613279e-01 6.30073667e-01
-1.20291865e+00 9.23282087e-01 -8.20913017e-01 -7.88437486e-01
4.20088947e-01 2.93023348e-01 -5.12936831e-01 4.18512732e-01
-8.73627603e-01 8.09912145e-01 -3.35363060e-01 2.44334251e-01
-7.63060927e-01 -6.65130913e-01 -1.04429173e+00 -2.67952114e-01
-1.59803569e-01 -1.18472326e+00 9.43780601e-01 -7.97084570e-01
-1.36436284e+00 1.32076263e+00 -1.49246797e-01 -5.05106509e-01
8.39994490e-01 -4.63445783e-01 -1.97833359e-01 5.44211149e-01
1.72607452e-01 6.89598143e-01 7.27763891e-01 -1.03288186e+00
-3.50034833e-01 -7.57738352e-01 -3.03332936e-02 1.51668325e-01
-1.52752608e-01 -2.40057826e-01 -5.27869761e-01 -8.21606100e-01
-1.98579021e-02 -8.91041279e-01 -1.29243255e-01 5.59382081e-01
-2.29390457e-01 4.24088500e-02 5.42201221e-01 -1.01086247e+00
8.99808824e-01 -2.03133798e+00 4.32675600e-01 2.33945668e-01
3.13531190e-01 -1.54164523e-01 3.26089598e-02 1.49363115e-01
4.93255863e-03 2.34223410e-01 -1.46359056e-01 -6.97496355e-01
1.50078282e-01 -1.68671548e-01 -6.34825677e-02 9.50782895e-01
4.05981839e-01 1.22483575e+00 -1.06315112e+00 -3.39365125e-01
6.52985513e-01 5.41114748e-01 -4.26871717e-01 2.73312479e-01
1.67529047e-01 8.47241938e-01 -7.51098394e-02 9.26777065e-01
2.15718895e-01 -3.93218160e-01 2.83236871e-03 -1.91015407e-01
-1.24178335e-01 1.71260498e-02 -6.70122862e-01 1.92920697e+00
-1.89154297e-01 8.63477468e-01 -1.76956102e-01 -6.66881263e-01
7.08618999e-01 -4.28059846e-02 9.03209686e-01 -9.85251188e-01
9.87657830e-02 -3.60058546e-01 -3.92535210e-01 -6.05780840e-01
3.72613519e-01 1.39252409e-01 -2.49920338e-02 4.56052393e-01
-2.86329165e-03 3.99691164e-01 -2.92393178e-01 -5.26550263e-02
9.41756129e-01 5.89715183e-01 -5.04931509e-02 -2.73184199e-02
1.16050839e-02 -3.06659430e-01 3.16048265e-01 8.08901489e-01
-4.99206245e-01 9.04283047e-01 1.64668575e-01 -9.55551326e-01
-1.20965517e+00 -1.23874366e+00 1.47543237e-01 9.23973501e-01
2.13924900e-01 -2.60680705e-01 -2.46384054e-01 -4.68845099e-01
5.91014087e-01 9.77520123e-02 -1.45359623e+00 -1.64410681e-01
-5.34450293e-01 -5.50720036e-01 7.21458554e-01 8.00074816e-01
6.22061670e-01 -8.26011121e-01 -1.30898333e+00 -3.92427184e-02
9.24593434e-02 -9.97047186e-01 -4.03842747e-01 -4.01288211e-01
-7.02155352e-01 -1.22180915e+00 -8.75279903e-01 -2.22251281e-01
4.65952396e-01 1.20767049e-01 1.29426396e+00 8.52513835e-02
-7.03420818e-01 1.11853123e+00 -3.01638961e-01 -6.95026666e-02
3.51431251e-01 -1.07226789e-01 1.61861956e-01 2.05817759e-01
1.84652820e-01 -6.75633788e-01 -1.05319178e+00 1.17129646e-01
-2.33548373e-01 -4.50483970e-02 2.67444342e-01 5.40496171e-01
2.87362337e-01 -5.61725616e-01 1.63446143e-01 -2.26547748e-01
3.43880028e-01 -5.11597157e-01 -2.33033001e-01 1.75590858e-01
-3.16056043e-01 -6.74959272e-02 1.37258902e-01 -2.17966706e-01
-6.12196922e-01 1.91799596e-01 2.92642981e-01 -8.63619268e-01
-2.67801434e-01 -3.28383856e-02 3.01730245e-01 -2.77826577e-01
5.96704304e-01 2.94829637e-01 1.34015188e-01 -6.12687111e-01
4.03677315e-01 1.71140969e-01 9.75690305e-01 -5.31106770e-01
7.07866251e-01 1.03060138e+00 1.45910293e-01 -6.85657680e-01
-5.28341770e-01 -4.95065480e-01 -1.19437850e+00 -6.11936390e-01
1.00058377e+00 -1.13744843e+00 -1.17664123e+00 5.83656728e-01
-9.25179064e-01 -6.75346434e-01 -3.31556201e-01 1.57856375e-01
-5.89079857e-01 4.01821464e-01 -6.27384365e-01 -1.12414706e+00
-3.95668715e-01 -4.88336265e-01 1.52758765e+00 1.76874727e-01
-6.26659989e-01 -1.05570328e+00 2.34966695e-01 1.82335764e-01
2.71894872e-01 9.57602739e-01 4.52625722e-01 -1.09806232e-01
-5.54799080e-01 -1.25657812e-01 -2.05370590e-01 -1.63015276e-01
1.15721188e-01 -7.64540047e-04 -1.03528118e+00 -3.17375600e-01
-6.38061523e-01 -2.74472773e-01 9.46920812e-01 5.12464464e-01
9.94696677e-01 -2.34725460e-01 -1.40958771e-01 1.09800065e+00
9.45838213e-01 -4.11610305e-01 3.75081986e-01 5.01849890e-01
9.93299544e-01 8.31864178e-01 7.77951702e-02 6.79977536e-01
6.34319246e-01 7.95633972e-01 4.04701442e-01 -3.99789959e-01
-4.34845358e-01 -4.38127846e-01 1.63488373e-01 8.81093070e-02
-5.07774293e-01 3.18875164e-01 -1.02171683e+00 7.96508908e-01
-1.77374828e+00 -1.14482570e+00 1.66914999e-01 2.19316888e+00
2.50057489e-01 -9.11390036e-02 8.16707551e-01 -2.72634625e-01
5.02126992e-01 2.03394160e-01 -8.45614493e-01 -1.32898912e-01
-2.21906036e-01 3.51595469e-02 5.46507239e-01 1.46128654e-01
-1.39535654e+00 6.71442091e-01 7.42039919e+00 -2.88248390e-01
-1.14930356e+00 -2.58400351e-01 5.90185881e-01 -5.97660005e-01
-1.59232989e-01 -3.66066068e-01 -3.10207248e-01 4.92368817e-01
6.61720395e-01 9.19384360e-02 3.94742757e-01 6.20813251e-01
2.84514517e-01 5.79755940e-02 -1.15709996e+00 1.26174319e+00
3.79960477e-01 -1.28388691e+00 -5.31889386e-02 1.99540094e-01
5.73206842e-01 5.61109297e-02 3.82104993e-01 -3.09873205e-02
5.15983224e-01 -1.32628441e+00 9.09714162e-01 7.88371801e-01
9.20813322e-01 -5.62659681e-01 3.00894886e-01 -2.26843625e-01
-1.34907353e+00 -2.57354140e-01 1.69655189e-01 -3.58728051e-01
6.02223165e-02 9.00870115e-02 -6.10426188e-01 3.99030030e-01
1.08952999e+00 1.38088691e+00 -9.86590326e-01 1.02101505e+00
-1.28116712e-01 1.41586602e-01 -3.02807599e-01 2.53987581e-01
7.02146068e-02 4.43722168e-03 4.45628315e-01 1.32594681e+00
2.32642323e-01 1.56485319e-01 1.30216464e-01 1.10427487e+00
7.37028662e-03 -2.56269485e-01 -9.04342949e-01 2.20852986e-01
1.72756121e-01 9.14789677e-01 -5.76113582e-01 -1.08840264e-01
-3.04852098e-01 1.30906975e+00 4.54752833e-01 4.54029590e-01
-6.66401267e-01 1.74847528e-01 9.82582927e-01 2.56523669e-01
3.92805874e-01 -5.41962862e-01 -4.53013271e-01 -1.15550816e+00
-6.20677248e-02 -4.10177231e-01 6.64107323e-01 -8.53419483e-01
-1.62668979e+00 2.73609877e-01 -1.11113921e-01 -1.09603894e+00
-3.29253942e-01 -6.25893712e-01 -5.77873468e-01 9.21282947e-01
-1.22613072e+00 -1.43452954e+00 -6.17721796e-01 5.47963321e-01
2.05739796e-01 -1.62953474e-02 7.32830763e-01 1.32241055e-01
-7.50804126e-01 5.67571282e-01 -1.52447358e-01 6.51383936e-01
5.40166259e-01 -1.34426880e+00 1.01043892e+00 8.58186066e-01
1.17007457e-01 5.82643092e-01 4.86563623e-01 -7.51177549e-01
-1.34987581e+00 -1.08208978e+00 3.80234241e-01 -1.06649411e+00
3.56610686e-01 -2.05885649e-01 -4.60637391e-01 7.83812761e-01
-9.76205170e-02 2.24979818e-01 7.89585292e-01 4.43824828e-01
-5.24246216e-01 4.81074415e-02 -1.05234444e+00 4.44601595e-01
1.39842045e+00 -7.88857996e-01 -5.36071301e-01 -1.55550689e-02
2.38864198e-01 -5.28442800e-01 -8.59858394e-01 2.65680045e-01
1.06550276e+00 -1.12768066e+00 1.41414964e+00 -7.85799086e-01
4.55125988e-01 -1.76596776e-01 6.43707067e-02 -1.08858049e+00
-6.41394138e-01 -4.37413126e-01 -3.39619040e-01 6.56871915e-01
-1.22704998e-01 -1.71529710e-01 9.29603696e-01 7.78421700e-01
1.55696273e-01 -6.35390341e-01 -1.08385980e+00 -6.38427734e-01
-5.69042712e-02 -1.90346569e-01 4.16102648e-01 8.54430735e-01
-1.31516427e-01 1.85081456e-03 -6.80517673e-01 1.64192930e-01
8.15889001e-01 4.39366519e-01 9.39061761e-01 -1.21868122e+00
-3.93980175e-01 -4.49697316e-01 -7.51608431e-01 -9.61502194e-01
7.23000839e-02 -6.20978951e-01 -3.22430074e-01 -1.39525592e+00
9.13973749e-02 -4.07531261e-02 -2.38228574e-01 6.94274306e-01
-1.08503915e-01 7.01336801e-01 4.47509080e-01 3.11124235e-01
-7.38971591e-01 6.12946630e-01 8.16540301e-01 -2.94194788e-01
-1.54126361e-01 -3.03690434e-01 -6.29255652e-01 5.54847956e-01
6.35460436e-01 9.48825106e-02 -6.00652508e-02 -6.12241328e-01
1.54909268e-02 -2.24191487e-01 1.22522485e+00 -1.17922366e+00
1.51925609e-01 1.60090886e-02 1.16686666e+00 -4.66723800e-01
6.31828070e-01 -4.25681859e-01 -3.37095186e-02 4.95038867e-01
-3.31572652e-01 3.69129598e-01 3.66648704e-01 6.98924422e-01
2.92390466e-01 7.72477806e-01 5.56833744e-01 -4.53638703e-01
-9.63560462e-01 5.42641878e-01 3.47306430e-02 1.32823661e-01
8.10736418e-01 -4.25891608e-01 -1.48345456e-01 -5.98202646e-01
-8.18799913e-01 3.23954433e-01 8.35980475e-01 4.85575110e-01
6.58105135e-01 -1.35907459e+00 -8.02621365e-01 2.50053793e-01
1.45297766e-01 -3.30996066e-01 3.95751864e-01 6.17936075e-01
-5.31391919e-01 3.21367502e-01 -6.73339903e-01 -7.92326152e-01
-1.02609229e+00 4.42336738e-01 7.11774349e-01 8.64476040e-02
-8.78103375e-01 7.52376318e-01 -1.90556590e-02 -3.04921180e-01
1.39285147e-01 -8.89495760e-02 -5.39109781e-02 -4.89444993e-02
3.11536372e-01 5.03089488e-01 -2.80529737e-01 -9.18379486e-01
-6.02776587e-01 9.34496462e-01 4.65073407e-01 -1.60980731e-01
1.31597435e+00 -4.65369463e-01 2.33795479e-01 5.60270190e-01
1.14078557e+00 -1.85971692e-01 -1.79200196e+00 1.27583772e-01
-3.60145122e-01 -5.86009920e-01 -3.36881101e-01 -9.20682490e-01
-1.09206069e+00 1.06627774e+00 8.97685766e-01 -8.86240825e-02
8.47829819e-01 -5.13953669e-03 7.04392195e-01 -3.65474038e-02
2.37785876e-01 -9.89619136e-01 3.17819983e-01 2.43647411e-01
8.27244461e-01 -1.30658913e+00 3.05142194e-01 7.48443380e-02
-4.73184437e-01 1.15067184e+00 7.56114960e-01 -2.04370141e-01
3.85392845e-01 -2.42873371e-01 5.10221533e-03 -4.22514111e-01
-3.41194242e-01 -5.61703146e-01 4.91954058e-01 8.89293790e-01
2.43409708e-01 -3.92525690e-03 4.31292683e-01 3.90231282e-01
-2.94355929e-01 -1.69835672e-01 5.73196784e-02 7.21650958e-01
1.18458336e-02 -5.19110799e-01 -2.60029554e-01 1.57504380e-01
-8.74453038e-02 2.18828768e-01 -7.83345222e-01 8.27758372e-01
3.35545033e-01 6.17852151e-01 3.74177009e-01 -5.89603126e-01
3.93943071e-01 7.16409273e-03 6.66098356e-01 -4.20461535e-01
-6.61843538e-01 -3.61208141e-01 -9.40687805e-02 -8.97239447e-01
-5.52835286e-01 -7.96618998e-01 -8.04415464e-01 -4.28635657e-01
5.88304341e-01 -4.55925047e-01 3.31374556e-01 6.27792954e-01
6.94162369e-01 5.01152813e-01 4.65257645e-01 -1.16588426e+00
-1.48568943e-01 -6.52795792e-01 -4.50723588e-01 9.90036964e-01
6.95847571e-01 -8.37928057e-01 -1.40991434e-01 1.23821050e-01] | [7.09236478805542, -0.9200131893157959] |
321d96af-33f6-474a-a985-f31495b04d1b | weighted-clustering-ensemble-a-review | 1910.02433 | null | https://arxiv.org/abs/1910.02433v2 | https://arxiv.org/pdf/1910.02433v2.pdf | Weighted Clustering Ensemble: A Review | Clustering ensemble, or consensus clustering, has emerged as a powerful tool for improving both the robustness and the stability of results from individual clustering methods. Weighted clustering ensemble arises naturally from clustering ensemble. One of the arguments for weighted clustering ensemble is that elements (clusterings or clusters) in a clustering ensemble are of different quality, or that objects or features are of varying significance. However, it is not possible to directly apply the weighting mechanisms from classification (supervised) domain to clustering (unsupervised) domain, also because clustering is inherently an ill-posed problem. This paper provides an overview of weighted clustering ensemble by discussing different types of weights, major approaches to determining weight values, and applications of weighted clustering ensemble to complex data. The unifying framework presented in this paper will help clustering practitioners select the most appropriate weighting mechanisms for their own problems. | ['Mimi Zhang'] | 2019-10-06 | null | null | null | null | ['clustering-ensemble'] | ['graphs'] | [ 2.41889119e-01 -4.64994282e-01 2.25499883e-01 -6.00199759e-01
-6.88257039e-01 -7.60459661e-01 4.38780546e-01 2.20302135e-01
-2.98447281e-01 5.07573783e-01 3.09689313e-01 -2.03195170e-01
-9.02223706e-01 -6.53211415e-01 2.04999104e-01 -1.51584494e+00
-3.06789935e-01 4.09803569e-01 -1.18386880e-01 6.20809160e-02
7.34349191e-01 5.25871694e-01 -1.91013741e+00 2.82532036e-01
1.24695730e+00 6.56168342e-01 1.16732053e-01 8.12812090e-01
-2.82014608e-01 3.03018808e-01 -8.47813189e-01 2.05761045e-02
8.50965381e-02 -6.01691306e-01 -5.75813949e-01 2.25544021e-01
3.28918286e-05 6.26976013e-01 3.51338446e-01 9.38959837e-01
6.81785822e-01 5.72971165e-01 1.28291321e+00 -1.56151986e+00
-5.03906071e-01 7.09846735e-01 -7.11814642e-01 -1.32742152e-01
1.12407856e-01 -1.83081433e-01 9.14843440e-01 -9.20932114e-01
2.53533781e-01 1.18466127e+00 7.24683285e-01 3.52028549e-01
-1.45744419e+00 -5.13056636e-01 2.28951409e-01 1.48620382e-01
-1.60824871e+00 -3.78822088e-01 6.01980269e-01 -5.82345784e-01
5.85287392e-01 5.92585862e-01 3.18448871e-01 7.29580998e-01
2.92769931e-02 2.50116855e-01 1.28702998e+00 -5.14561415e-01
6.37710869e-01 1.30933210e-01 5.31450391e-01 1.27963424e-01
4.81207669e-01 -1.30501971e-01 -1.57352090e-01 -4.66273576e-01
-3.07869855e-02 2.75264005e-03 -2.22742587e-01 -4.84177351e-01
-1.27494180e+00 8.51091385e-01 1.83415070e-01 7.54286706e-01
-2.51013041e-01 1.14103064e-01 2.06524894e-01 3.13083559e-01
5.21201551e-01 6.94463372e-01 -4.40830946e-01 1.01029545e-01
-1.13937795e+00 2.87619054e-01 5.09484529e-01 6.29689991e-01
6.18726373e-01 -6.46063238e-02 1.59605369e-01 9.39664185e-01
4.35707808e-01 6.02927804e-01 4.61886078e-01 -1.26110566e+00
-7.51408935e-02 7.90256679e-01 5.79455793e-02 -1.22585690e+00
-4.65271622e-01 -1.03330158e-01 -1.12759733e+00 6.24456704e-01
4.85644400e-01 -4.83353108e-01 -6.76869631e-01 1.61374414e+00
2.29405135e-01 -1.07703879e-01 -5.44133224e-03 5.50379813e-01
6.91124439e-01 6.14246666e-01 5.90174422e-02 -6.19715273e-01
1.13551652e+00 -4.75044698e-01 -7.05087185e-01 2.18575031e-01
4.88208085e-01 -8.62622201e-01 5.36610901e-01 6.28771365e-01
-8.15655291e-01 -7.25826144e-01 -9.18157458e-01 5.52204609e-01
-6.80027902e-01 -2.94909239e-01 2.49161497e-01 8.72760415e-01
-1.13135624e+00 7.30830014e-01 -5.36782861e-01 -6.09108150e-01
1.00701591e-02 6.20454133e-01 -2.54104346e-01 8.37458074e-02
-8.09873223e-01 1.08639038e+00 6.55890405e-01 -1.04654551e-01
-7.53116682e-02 -5.46531618e-01 -5.75258136e-01 -1.54435232e-01
1.24424867e-01 -4.41819042e-01 5.55692136e-01 -9.67528045e-01
-8.78068447e-01 5.28332412e-01 -5.44582546e-01 2.18512759e-01
1.67537197e-01 2.39359066e-01 -7.93402910e-01 -2.32677013e-01
-2.42223404e-02 2.68133700e-01 7.85971224e-01 -1.90193689e+00
-9.12934422e-01 -5.71011662e-01 -6.48222268e-01 1.77728891e-01
-2.33651027e-01 7.43199438e-02 1.14819460e-01 -6.49585724e-01
4.08463836e-01 -7.50285268e-01 -7.19108224e-01 -5.56664348e-01
-1.84364274e-01 -5.37405491e-01 1.11202347e+00 -2.18033701e-01
1.69352543e+00 -2.21121264e+00 4.37337875e-01 9.19894516e-01
3.73929828e-01 -1.12270877e-01 -1.96050592e-02 7.36885369e-01
-3.96629095e-01 3.68769258e-01 -6.73597634e-01 -7.91536365e-03
2.84935832e-02 2.79803127e-01 -4.88839224e-02 4.43256646e-01
-1.32471891e-02 3.08451802e-01 -1.18021595e+00 -7.86500752e-01
5.87923229e-01 4.52060699e-01 -7.70290717e-02 3.76515128e-02
4.96615589e-01 4.98266846e-01 -1.77541539e-01 5.17189145e-01
7.64608681e-01 -5.30618504e-02 4.77767050e-01 -1.47535980e-01
-3.19287717e-01 -3.59687030e-01 -1.87410223e+00 1.19106793e+00
-7.13290051e-02 4.64029253e-01 1.31727261e-02 -1.11175025e+00
1.01226604e+00 4.46996331e-01 9.53177392e-01 2.35273361e-01
7.96876177e-02 2.41546899e-01 3.99030447e-01 -4.90069717e-01
5.29244900e-01 -1.70695752e-01 -1.28056973e-01 9.54047978e-01
1.61669418e-01 -3.16688776e-01 5.44933677e-01 2.15859249e-01
9.11265671e-01 -3.92900914e-01 5.98991573e-01 -8.39182734e-01
6.53442621e-01 4.36871499e-02 4.92529005e-01 6.63790166e-01
-1.30911291e-01 7.61478364e-01 1.48218706e-01 -7.97079951e-02
-9.39727068e-01 -1.22227371e+00 -3.34381580e-01 1.13658643e+00
-3.47459614e-02 -3.51108462e-01 -6.25959456e-01 -5.91829002e-01
-1.28000351e-02 6.32400632e-01 -6.94444776e-01 8.12038872e-03
-2.87239134e-01 -1.10285318e+00 2.00847492e-01 4.76310551e-01
1.50087625e-01 -8.12066555e-01 -3.72578800e-01 3.73654306e-01
-2.13459805e-01 -3.14618915e-01 -2.49033004e-01 5.29052258e-01
-1.00853896e+00 -1.28496075e+00 -4.58753139e-01 -5.74495435e-01
9.41841483e-01 7.66878128e-01 1.18926549e+00 4.29419339e-01
1.08349323e-01 5.95220447e-01 -7.21841633e-01 -4.97461140e-01
-3.10223669e-01 1.51049215e-02 5.16462624e-01 3.48566612e-03
7.06526279e-01 -6.03259206e-01 -4.93425548e-01 5.85203052e-01
-8.72584105e-01 -7.06875861e-01 1.24313489e-01 8.34776044e-01
1.92641616e-01 1.04071617e+00 7.03657150e-01 -8.75776172e-01
1.00734413e+00 -6.23537958e-01 -3.95803005e-02 5.19563973e-01
-8.90238345e-01 -7.85099566e-02 2.84739912e-01 -3.64385247e-01
-1.03624821e+00 1.50499031e-01 1.47782505e-01 3.70655246e-02
-4.89952505e-01 4.00919825e-01 -2.62152404e-01 1.16332486e-01
9.43252563e-01 -8.26851204e-02 9.75846872e-02 -1.81943968e-01
5.72251081e-01 8.25361907e-01 2.98872292e-01 -9.14265335e-01
8.27562630e-01 5.49340546e-01 -1.13515243e-01 -7.45029926e-01
-2.78989255e-01 -8.57369661e-01 -1.20305312e+00 -6.24934077e-01
8.94488096e-01 -3.74725223e-01 -5.73600769e-01 2.31841683e-01
-8.34244311e-01 1.68520972e-01 -1.79377601e-01 2.98526824e-01
-2.69641936e-01 4.16165411e-01 4.42291275e-02 -1.13808894e+00
-9.44932178e-03 -1.21088922e+00 5.64777076e-01 9.64481682e-02
-9.39396441e-01 -1.33326995e+00 4.85520884e-02 2.24797606e-01
2.89811403e-01 6.10322237e-01 7.97417700e-01 -6.87246442e-01
2.96599358e-01 -3.59722286e-01 7.25022554e-02 2.04345867e-01
5.23407638e-01 9.30557251e-01 -1.09990096e+00 -3.24396998e-01
9.46664512e-02 7.36876801e-02 8.50311637e-01 5.93780994e-01
1.15761101e+00 1.69962585e-01 -6.01493716e-01 -2.49981671e-03
1.33203757e+00 4.86592472e-01 4.48277622e-01 -1.72782764e-01
4.24427837e-01 1.16924572e+00 4.52560484e-01 3.53354156e-01
9.34201702e-02 3.08770597e-01 6.65661972e-03 6.02064766e-02
4.98711348e-01 4.78140414e-01 1.76452860e-01 1.21971166e+00
-7.35497773e-01 -2.26243883e-01 -1.06898212e+00 3.90614003e-01
-2.12442040e+00 -1.43735027e+00 -8.01121950e-01 2.14531469e+00
4.53033388e-01 -2.31923148e-01 2.27343768e-01 7.05081701e-01
1.11712205e+00 6.37170800e-04 -1.70849845e-01 -3.70033026e-01
-2.03893051e-01 1.37447745e-01 2.46819809e-01 5.34574986e-01
-1.08035219e+00 2.97744334e-01 7.67836094e+00 9.98083234e-01
-4.09517378e-01 -1.60600632e-01 4.09138530e-01 1.13085270e-01
-1.74308524e-01 -4.65668589e-02 -1.69822812e-01 6.80849612e-01
7.51787901e-01 -3.34214211e-01 1.36513367e-01 5.21597564e-01
3.01848352e-01 -3.17946434e-01 -1.12932050e+00 7.26399302e-01
-1.48618191e-01 -8.60005140e-01 -2.43912518e-01 2.64352024e-01
1.12514663e+00 -5.25472820e-01 1.50110871e-01 2.42255013e-02
9.79667306e-01 -1.20294988e+00 8.13612044e-02 5.62452435e-01
3.90964955e-01 -1.05603397e+00 9.59003806e-01 1.42889410e-01
-1.30257821e+00 -3.32576632e-02 -3.42923760e-01 -1.05048552e-01
1.28872275e-01 1.02372921e+00 -1.81009933e-01 7.65361607e-01
1.02281046e+00 6.05791032e-01 -6.12852931e-01 9.18837786e-01
1.89992040e-01 4.51959103e-01 -1.92887694e-01 8.59762654e-02
1.66393325e-01 -6.03761256e-01 3.68463933e-01 1.41595602e+00
5.15448868e-01 4.37514812e-01 2.17253752e-02 4.90001142e-01
4.69163269e-01 -8.03973228e-02 -7.49035120e-01 1.58662841e-01
9.91948783e-01 1.47106194e+00 -1.19800735e+00 -3.89775693e-01
-2.68539190e-01 5.07942855e-01 -1.07312381e-01 4.07937557e-01
-4.01088804e-01 -3.71430606e-01 8.52243543e-01 -3.36857647e-01
-3.18047330e-02 -2.88869023e-01 -7.28061974e-01 -5.89194775e-01
-5.48860133e-01 -1.00983584e+00 7.80886531e-01 -5.37337363e-01
-1.70263958e+00 2.69196391e-01 1.98901087e-01 -1.45594966e+00
-1.15782581e-01 -5.91036677e-01 -9.40331757e-01 6.34209931e-01
-4.19909418e-01 -5.23107290e-01 -3.67801130e-01 5.37563622e-01
3.68172020e-01 -2.96509326e-01 9.73218918e-01 -1.51252240e-01
-4.01040763e-01 1.19119577e-01 7.16318846e-01 -8.75349250e-03
1.04191136e+00 -1.73451817e+00 -3.38385075e-01 5.72448850e-01
-6.26585558e-02 9.62884068e-01 1.00725031e+00 -3.38625014e-01
-8.64450932e-01 -8.85504603e-01 7.97314048e-01 -9.58600640e-01
5.20410955e-01 1.24919331e-02 -1.04083431e+00 6.92894533e-02
5.49919605e-01 -6.79871857e-01 1.38817620e+00 4.69456524e-01
-1.88793406e-01 -1.00772180e-01 -1.45688486e+00 7.24163651e-01
8.16112518e-01 -5.40570449e-03 -6.53012395e-01 6.28848672e-02
1.21947438e-01 3.17850798e-01 -1.24058056e+00 4.19153839e-01
4.73919094e-01 -1.27577758e+00 1.05452895e+00 -3.85600090e-01
1.91364601e-01 -7.60677993e-01 -4.21046585e-01 -1.88816047e+00
-7.58585095e-01 -2.73535140e-02 2.17284322e-01 1.45492136e+00
4.46167082e-01 -5.47689974e-01 5.18250227e-01 7.01224208e-01
5.30253425e-02 -3.52901459e-01 -6.86941803e-01 -6.07273221e-01
3.14630330e-01 -7.21738338e-01 7.34454036e-01 1.36081970e+00
3.71437043e-01 2.32158303e-01 -2.64056586e-02 1.05923995e-01
1.14598191e+00 -4.01085429e-02 6.08359218e-01 -1.73598123e+00
2.15313867e-01 -1.05330598e+00 -4.03176367e-01 -3.44905071e-02
6.43896461e-02 -6.10294342e-01 8.28326717e-02 -1.54004645e+00
1.05579942e-01 -4.64370906e-01 -6.38211310e-01 -4.15490307e-02
-5.25609016e-01 4.71210301e-01 8.84100273e-02 3.52961898e-01
-6.30249500e-01 -3.86282727e-02 8.04866314e-01 -2.11902916e-01
-1.23441808e-01 3.04467138e-02 -9.40101743e-01 9.39258456e-01
1.10444772e+00 -4.49800044e-01 -5.60151935e-01 3.76603678e-02
1.21115379e-01 -4.58721310e-01 -1.44180849e-01 -1.02990413e+00
4.75559592e-01 -4.50574785e-01 5.71380794e-01 -7.82641649e-01
2.21673623e-02 -1.26493275e+00 5.44209301e-01 3.17123264e-01
-2.51564741e-01 2.74228811e-01 -3.63910317e-01 5.24651885e-01
-2.28304520e-01 -4.12801564e-01 9.71125484e-01 -2.68891662e-01
-7.41417766e-01 -2.05136746e-01 -7.55827248e-01 -1.33664042e-01
1.24522638e+00 -7.49183357e-01 1.70462579e-01 -2.31557161e-01
-1.01577044e+00 6.61244094e-01 5.70397735e-01 2.36045659e-01
4.98819768e-01 -1.43096447e+00 -9.41828787e-01 -2.77385488e-02
2.56966352e-01 -3.63833122e-02 2.31338546e-01 8.50747943e-01
-2.12806866e-01 -2.62819100e-02 -1.20703857e-02 -7.57374883e-01
-1.57596374e+00 6.74559653e-01 1.65109843e-01 1.01069018e-01
-4.77811620e-02 7.65243709e-01 -2.24531218e-01 -6.73607528e-01
2.13201120e-01 2.69218385e-01 -4.10444349e-01 5.72672784e-01
5.61640859e-01 1.12018847e+00 -7.98494183e-03 -7.31556118e-01
-4.77576762e-01 9.08153236e-01 1.21034920e-01 -1.42743945e-01
1.23491609e+00 -2.50862569e-01 -6.60485685e-01 8.17748308e-01
7.16631949e-01 -2.27593824e-01 -7.11154997e-01 -4.15525325e-02
4.79138315e-01 -3.94304544e-01 -1.72807410e-01 -6.65897965e-01
-8.62426102e-01 6.24942005e-01 5.31044304e-01 7.20581293e-01
1.22804618e+00 -4.42955010e-02 -1.49299309e-01 3.02347004e-01
2.74884492e-01 -1.64933741e+00 -9.82547626e-02 2.20345363e-01
7.33421803e-01 -1.50475526e+00 2.68947273e-01 -1.51644379e-01
-7.71895826e-01 1.26593578e+00 5.14000535e-01 -2.17089534e-01
1.04282558e+00 3.41585696e-01 3.59230578e-01 -2.63921976e-01
-6.03745937e-01 -2.77018040e-01 4.49323803e-01 1.11706519e+00
9.22506630e-01 2.56432354e-01 -6.87317789e-01 2.22251251e-01
-1.25808522e-01 -6.28633797e-01 2.49977618e-01 8.16447496e-01
-7.83218384e-01 -1.21596575e+00 -1.07413697e+00 7.11946011e-01
-1.60107061e-01 2.29791775e-01 -8.33960474e-01 6.80563092e-01
6.51931167e-01 1.81559753e+00 1.87654331e-01 -6.86051130e-01
1.67189941e-01 3.45150888e-01 8.04102868e-02 -4.64018285e-01
-7.92503834e-01 1.22318469e-01 -1.84921637e-01 -1.18233003e-01
-1.17073452e+00 -7.71301985e-01 -1.17213047e+00 -1.00935292e+00
-5.43096423e-01 7.53101051e-01 4.23023582e-01 7.30159938e-01
-2.07173936e-02 5.58883190e-01 9.48444188e-01 -9.91835654e-01
-1.51965410e-01 -1.16503692e+00 -9.37132359e-01 4.75491673e-01
-4.61340696e-02 -7.27087975e-01 -6.85471296e-01 2.14406192e-01] | [7.605652809143066, 4.555628776550293] |
11322dcf-e42b-4480-9c0e-1739a0ed246c | proceedings-of-the-2nd-workshop-on-logic-and | 2211.09923 | null | https://arxiv.org/abs/2211.09923v1 | https://arxiv.org/pdf/2211.09923v1.pdf | Proceedings of the 2nd Workshop on Logic and Practice of Programming (LPOP) | This proceedings contains abstracts and position papers for the work presented at the second Logic and Practice of Programming (LPOP) Workshop. The workshop was held online, virtually in place of Chicago, USA, on November 15, 2010, in conjunction with the ACM SIGPLAN Conference on Systems, Programming, Languages, and Applications: Software for Humanity (SPLASH) 2020. The purpose of this workshop is to be a bridge between different areas of computer science that use logic as a practical tool. We take advantage of the common language of formal logic to exchange ideas between these different areas. | ['Yanhong A. Liu', 'Peter Van Roy', 'David S. Warren'] | 2022-11-17 | null | null | null | null | ['formal-logic'] | ['reasoning'] | [-1.21095195e-01 4.12407130e-01 -1.50659174e-01 -3.19462270e-01
-1.31416962e-01 -8.16472948e-01 5.60726643e-01 6.33365333e-01
-1.18341751e-01 5.32475948e-01 2.12624725e-02 -7.62542188e-01
9.06218216e-02 -8.09210539e-01 -5.40145397e-01 1.92480907e-01
-5.09522200e-01 6.00283146e-02 6.51493371e-01 -3.04879993e-01
2.82878548e-01 4.76149589e-01 -1.44659603e+00 5.94957829e-01
4.72226471e-01 6.52704954e-01 -4.00962174e-01 4.81928349e-01
-2.62075514e-01 8.94327223e-01 -1.63601920e-01 -5.05852878e-01
2.02616036e-01 -2.95059960e-02 -1.03047121e+00 -4.03079957e-01
1.62301987e-01 6.68787584e-02 -1.48668170e-01 1.26255059e+00
-1.86977834e-01 -3.76575351e-01 -1.76241249e-01 -1.61032963e+00
1.04742855e-01 1.20915532e+00 -6.25808060e-01 -1.90110892e-01
6.86178446e-01 7.38473982e-02 1.10150409e+00 -5.05617559e-01
6.51889741e-01 1.42250562e+00 7.08147585e-01 1.97121039e-01
-1.37213242e+00 -4.20621544e-01 1.86944246e-01 -3.04037593e-02
-1.62298417e+00 -5.94404817e-01 6.40057087e-01 -5.61266005e-01
1.12852228e+00 6.55156016e-01 8.33962142e-01 6.22015633e-02
4.13465470e-01 8.58590007e-01 9.70225990e-01 -1.07727039e+00
2.79131860e-01 6.51351750e-01 6.56421721e-01 5.83154023e-01
6.50253594e-01 -1.53121337e-01 -1.27551839e-01 -8.18762541e-01
6.59874797e-01 -2.63957590e-01 -4.98007834e-02 -5.80574334e-01
-1.04886603e+00 4.91193026e-01 -2.69286454e-01 7.52469838e-01
-3.65738943e-02 4.82211262e-01 5.23671091e-01 3.24892968e-01
-1.49816543e-01 4.22345102e-01 -3.39673281e-01 -2.76438057e-01
-4.85610664e-01 5.16256690e-01 1.49289024e+00 7.17085004e-01
2.11995542e-01 -3.05193275e-01 3.32583249e-01 2.24662513e-01
9.71572280e-01 5.18449619e-02 -3.29177260e-01 -1.34862244e+00
1.60526171e-01 6.91861272e-01 3.62707794e-01 -1.11787844e+00
6.40277192e-02 4.47717637e-01 -4.38262075e-02 4.64276075e-01
2.08972216e-01 -3.51768762e-01 -2.72045225e-01 1.38700211e+00
3.77258360e-02 -1.93505242e-01 4.21965234e-02 3.00665677e-01
7.27317274e-01 8.70700359e-01 1.44010693e-01 -3.71097416e-01
8.66282582e-01 -4.12713408e-01 -6.48212194e-01 1.27403602e-01
8.01279008e-01 -6.59982800e-01 5.42381883e-01 8.02136540e-01
-1.69240403e+00 2.40407847e-02 -1.23809040e+00 4.03751582e-01
-2.69849479e-01 -4.53358978e-01 9.74977672e-01 7.63219297e-01
-1.49567914e+00 2.01349601e-01 -1.13531661e+00 -4.88423884e-01
-2.89024059e-02 3.37063223e-01 1.12060696e-01 -6.98125064e-02
-1.20652652e+00 1.02784741e+00 5.34911573e-01 -8.57570097e-02
-7.87718371e-02 -5.09380877e-01 -8.28122258e-01 -1.37394324e-01
4.13403422e-01 -5.30293547e-02 1.24008131e+00 -8.47126126e-01
-1.27033758e+00 1.23370624e+00 -6.79121178e-04 -7.08521962e-01
2.19336301e-01 1.52585253e-01 -5.86779714e-01 -1.65073946e-01
-1.20264605e-01 2.79411256e-01 -1.88599959e-01 -1.19353151e+00
-8.74286115e-01 -1.61645561e-01 9.33435082e-01 -4.97129887e-01
1.15769416e-01 7.06267297e-01 -3.37631673e-01 1.40957788e-01
9.33982432e-02 -6.57734513e-01 -2.63639778e-01 -1.71244249e-01
2.54556954e-01 -4.66092825e-01 5.10181189e-01 -2.04529300e-01
1.55558944e+00 -2.43711972e+00 -7.32224509e-02 5.43845296e-01
1.63279846e-01 2.64890134e-01 4.43955898e-01 1.05408871e+00
-4.47388440e-02 3.74843180e-01 4.42783721e-02 4.32432592e-01
4.58006501e-01 -1.06151469e-01 -4.77442086e-01 3.76981109e-01
-1.77234054e-01 5.19517541e-01 -8.88897598e-01 -6.84369206e-01
1.29393741e-01 9.92194191e-02 -4.51032221e-01 -4.70687538e-01
-5.82724094e-01 -3.35505515e-01 -5.01955509e-01 4.68978524e-01
7.75446951e-01 2.55894978e-02 9.97108281e-01 7.82312974e-02
-8.76560152e-01 4.38904315e-01 -1.60954225e+00 1.45748186e+00
-2.72614777e-01 5.67009449e-01 3.88326257e-01 -6.07724130e-01
6.94495857e-01 8.83015215e-01 2.59196132e-01 -3.20328176e-02
4.11683023e-02 3.04453522e-01 2.14974076e-01 -3.50005209e-01
2.35153139e-01 -2.20511869e-01 -2.02049866e-01 8.73489439e-01
-5.54310739e-01 -6.61803186e-01 8.23183775e-01 3.83918107e-01
1.16551578e+00 4.00491655e-01 5.02628744e-01 -6.36271834e-01
9.04409111e-01 1.49694443e-01 6.67373359e-01 2.80949801e-01
-3.49088043e-01 -1.02520697e-01 9.62304473e-01 -8.50425720e-01
-7.41368353e-01 -8.73122633e-01 -1.65774986e-01 5.46436131e-01
2.28485614e-02 -1.25995851e+00 -4.22442257e-01 -4.60741431e-01
-9.36592221e-02 8.56140018e-01 -3.79020274e-02 1.14371702e-01
-6.45937383e-01 4.65294085e-02 6.31658375e-01 3.97853583e-01
3.08513552e-01 -1.03717101e+00 -9.09417093e-01 -2.05747224e-02
1.15145363e-01 -9.90301609e-01 -9.42182839e-02 5.04708700e-02
-6.44130647e-01 -1.00774980e+00 2.25655183e-01 -9.12744403e-01
4.06453639e-01 -2.04064354e-01 1.10970187e+00 6.43688679e-01
-3.08070779e-01 6.83813691e-01 -2.61313200e-01 -5.95231771e-01
-5.13822973e-01 -5.95710576e-01 -2.90149093e-01 -4.18086261e-01
4.86762851e-01 -4.21151489e-01 1.66118070e-01 -6.70734569e-02
-1.10151482e+00 2.77943932e-03 1.13161035e-01 4.16817337e-01
1.16189182e-01 5.75352669e-01 -9.13905259e-03 -7.20739186e-01
5.30525625e-01 -8.72504935e-02 -1.27405965e+00 6.63172841e-01
-5.61231256e-01 -3.03652108e-01 2.15740904e-01 2.71522731e-01
-6.38206065e-01 -7.52901807e-02 2.54414499e-01 2.21168712e-01
1.93194255e-01 1.00963271e+00 -4.45231408e-01 -2.27397114e-01
2.98913866e-01 3.92301753e-02 -1.04493789e-01 1.97276577e-01
-1.48233771e-01 7.27148414e-01 1.35628939e-01 -1.16702259e+00
6.11360073e-01 3.92566562e-01 2.24237636e-01 -6.40957475e-01
-1.21808805e-01 8.17534234e-03 -2.65736312e-01 -1.39484286e-01
2.82288104e-01 -6.73016369e-01 -8.46880376e-01 1.71922684e-01
-1.31783307e+00 -3.15607965e-01 -3.44293565e-01 2.33876884e-01
-1.17925450e-01 2.07200170e-01 -2.72926599e-01 -1.15761566e+00
1.03850953e-01 -7.46798873e-01 4.42791253e-01 1.41253145e-02
-7.04762101e-01 -1.42748380e+00 2.80077249e-01 7.85340741e-02
8.72823782e-03 5.27233005e-01 9.15542305e-01 -4.36213642e-01
-6.24753833e-01 -7.11050093e-01 -1.77327201e-01 5.14474690e-01
-1.28530964e-01 9.58742678e-01 -4.60772693e-01 -9.11146179e-02
-2.27501690e-01 -3.48239452e-01 -1.14111967e-01 1.71381190e-01
7.08414733e-01 -1.52791291e-01 -5.64987183e-01 -3.73367697e-01
1.76596498e+00 3.45484793e-01 1.02000165e+00 3.76137167e-01
-1.41047329e-01 6.12790167e-01 5.09272158e-01 5.12510180e-01
6.20361388e-01 6.17860913e-01 3.54251228e-02 5.55713415e-01
2.93340206e-01 -7.06326589e-02 3.86976868e-01 4.03461248e-01
-2.61111468e-01 2.97534674e-01 -1.73818827e+00 7.35971749e-01
-2.06726408e+00 -9.97256517e-01 -4.73759681e-01 2.41341639e+00
9.64924753e-01 6.58003390e-01 1.04845241e-01 2.22280115e-01
5.84480941e-01 4.33134176e-02 5.24161398e-01 -8.52440774e-01
4.40987796e-01 4.28116381e-01 2.07725495e-01 8.39257300e-01
-8.23654175e-01 5.39456308e-01 7.01633644e+00 3.29909623e-01
-1.21829033e+00 -1.53780123e-02 -1.76768243e-01 2.34783873e-01
-8.28804672e-01 8.22951436e-01 -5.98522484e-01 1.57715201e-01
1.07132387e+00 -9.32350814e-01 6.51178300e-01 6.44011140e-01
1.00155920e-01 -3.57379526e-01 -1.17600608e+00 3.66284132e-01
4.04432192e-02 -1.46178555e+00 -4.97165352e-01 -1.44658491e-01
8.00392926e-01 -2.22137943e-01 -3.21140349e-01 1.43609598e-01
4.94659871e-01 -1.14617550e+00 1.04396653e+00 3.91235977e-01
2.69736856e-01 -7.60907769e-01 8.16649735e-01 1.11657821e-01
-8.96294177e-01 2.15987153e-02 3.34135801e-01 -7.75370777e-01
-5.61283231e-02 4.46313500e-01 -4.57976520e-01 5.09665012e-01
6.94336712e-01 3.29916418e-01 -3.94385792e-02 1.25003469e+00
-1.68162778e-01 3.51575732e-01 -4.89396214e-01 -2.42983699e-01
-3.74694690e-02 2.23601297e-01 5.71995854e-01 1.08930123e+00
-4.00033087e-01 3.64791930e-01 3.77055794e-01 9.22588527e-01
4.40119535e-01 -4.51206207e-01 -4.68576014e-01 -4.20440346e-01
5.53449214e-01 9.71961319e-01 -8.51014912e-01 -3.42534989e-01
-6.01904035e-01 -1.08038329e-01 -5.64717412e-01 1.36980325e-01
-8.47821474e-01 -7.95963466e-01 9.32364836e-02 7.26393163e-01
4.05348130e-02 -5.10913670e-01 -7.26001441e-01 -7.82596290e-01
1.85277164e-01 -9.83871758e-01 2.01901272e-01 -5.96329153e-01
-8.80279064e-01 2.60240674e-01 5.99689364e-01 -6.96980357e-01
-2.12979749e-01 -4.26671594e-01 -7.88450122e-01 8.41598332e-01
-7.46819317e-01 -1.17493796e+00 1.21580601e-01 -5.70216924e-02
-3.24152440e-01 1.97717607e-01 1.02672911e+00 1.75108507e-01
2.24760491e-02 2.85799265e-01 -1.89066038e-01 -6.88930079e-02
4.45092350e-01 -1.09115183e+00 1.28436640e-01 7.04468787e-01
-1.40728742e-01 1.25976050e+00 1.03024483e+00 -3.12613994e-01
-2.01942921e+00 -6.05129421e-01 1.68571627e+00 -4.74231333e-01
1.01255906e+00 -2.15702608e-01 -4.02666181e-01 1.38235188e+00
5.17479956e-01 -1.05300419e-01 5.11020064e-01 8.90539400e-03
-2.06126913e-01 -4.85358208e-01 -1.27829754e+00 6.44119501e-01
4.52139586e-01 -4.90307480e-01 -7.17915237e-01 1.09720260e-01
6.83972299e-01 -4.86053973e-01 -8.70751858e-01 3.85124922e-01
7.10983813e-01 -7.66836047e-01 4.08973038e-01 -8.09024870e-02
3.77530336e-01 -7.97850728e-01 -5.75138986e-01 -3.52829993e-01
1.47618800e-01 -9.82356191e-01 3.08559150e-01 1.29224563e+00
4.62852091e-01 -9.30382252e-01 3.60762060e-01 1.07433200e+00
-9.57997739e-02 -3.53199452e-01 -7.05177784e-01 -5.76173604e-01
1.42898113e-01 -9.52297390e-01 5.78570366e-01 6.77818418e-01
1.29776096e+00 -1.30922412e-02 3.92675430e-01 -2.17307627e-01
5.44742167e-01 1.94068581e-01 7.96286881e-01 -1.30795753e+00
-3.74312073e-01 -5.78385949e-01 -9.21896458e-01 -2.64890432e-01
-9.92673710e-02 -9.43179309e-01 -3.24531496e-02 -1.60827351e+00
-5.34210391e-02 -2.73579687e-01 -6.00411296e-02 7.63094902e-01
6.72886968e-01 -8.61692056e-02 3.61263603e-01 -2.67071482e-02
-7.72704244e-01 -1.98270693e-01 6.67643547e-01 -4.16869782e-02
-2.59599388e-01 3.21175531e-02 -1.05177307e+00 7.77812481e-01
3.20081860e-01 -5.78787290e-02 -3.53517830e-01 -3.59229982e-01
6.83048368e-01 2.27749288e-01 4.20260131e-01 -8.80562544e-01
2.95263499e-01 -7.05301881e-01 -4.17317659e-01 -2.85973430e-01
3.86174209e-02 -1.22128034e+00 6.95714593e-01 8.12634587e-01
-1.88419282e-01 -1.31212354e-01 7.77844667e-01 -2.48386875e-01
-4.12475914e-01 -4.76341844e-01 4.14118677e-01 -2.02972427e-01
-5.92904806e-01 -3.49534690e-01 -5.15404701e-01 -3.79426293e-02
1.79490936e+00 -8.15387964e-02 -2.62911588e-01 -6.45422116e-02
-5.89130640e-01 6.40664458e-01 8.64460170e-01 -2.46242091e-01
4.02921379e-01 -9.05787885e-01 -4.57453907e-01 -5.49819618e-02
-9.30966735e-02 -2.78621465e-01 -1.01297200e-01 1.02329016e+00
-1.00341117e+00 9.43113506e-01 -2.37236977e-01 -3.29833359e-01
-1.32821548e+00 2.22790867e-01 8.40669200e-02 -3.96148533e-01
-3.72680634e-01 4.75536793e-01 -3.48310143e-01 -2.55417168e-01
3.94271582e-01 -2.60657907e-01 2.14761183e-01 -5.49668252e-01
8.06870818e-01 2.20624089e-01 -3.08403671e-01 7.51446038e-02
-8.45975041e-01 2.24168643e-01 1.92043241e-02 -6.06563687e-01
1.58373225e+00 -1.96976904e-02 -1.22262704e+00 1.11895096e+00
5.54215133e-01 2.39807159e-01 -4.86278415e-01 4.15542163e-02
3.39354634e-01 -2.70060420e-01 -7.08060637e-02 -8.22675824e-01
-2.97560751e-01 5.63776314e-01 1.96835205e-01 6.25723362e-01
7.94359267e-01 1.36311483e-02 3.71816844e-01 5.23079000e-02
8.82473290e-01 -9.25838947e-01 -6.28898501e-01 5.71605623e-01
1.05944359e+00 -3.78839850e-01 3.15008521e-01 -6.74965858e-01
-3.48356545e-01 1.48644996e+00 5.40708899e-01 -1.73265293e-01
1.11615884e+00 6.79852188e-01 -2.24083930e-01 -2.17100725e-01
-1.01882219e+00 3.12468261e-01 -4.53478873e-01 6.21509850e-01
8.21085155e-01 1.56320766e-01 -1.00646353e+00 7.33495176e-01
7.39412233e-02 9.24840748e-01 8.28282893e-01 1.82903576e+00
-3.79702091e-01 -1.66633368e+00 -6.36350513e-01 -1.06992722e-01
-5.35726130e-01 -1.76969953e-02 -5.86583912e-01 1.00637925e+00
2.77809918e-01 9.56736565e-01 -5.24524339e-02 -3.38372827e-01
3.53545755e-01 1.30002424e-01 8.65814745e-01 -8.02212715e-01
-7.01292396e-01 -3.05730969e-01 6.60218060e-01 -2.72186995e-01
-3.19631100e-01 -7.96656489e-01 -1.70893216e+00 -1.03112090e+00
-7.47634172e-02 7.70686030e-01 6.79156840e-01 8.38334024e-01
-2.39412133e-02 3.19303155e-01 3.17045510e-01 -3.70484233e-01
-3.48291397e-01 -4.69645321e-01 -9.10688519e-01 -5.43588042e-01
-7.18007013e-02 -2.77046859e-01 -7.43831396e-02 2.50466228e-01] | [8.685700416564941, 6.726314067840576] |
a780dced-2ba5-4457-8eca-1bfa1db4ed9b | most-a-multi-oriented-scene-text-detector | 2104.01070 | null | https://arxiv.org/abs/2104.01070v2 | https://arxiv.org/pdf/2104.01070v2.pdf | MOST: A Multi-Oriented Scene Text Detector with Localization Refinement | Over the past few years, the field of scene text detection has progressed rapidly that modern text detectors are able to hunt text in various challenging scenarios. However, they might still fall short when handling text instances of extreme aspect ratios and varying scales. To tackle such difficulties, we propose in this paper a new algorithm for scene text detection, which puts forward a set of strategies to significantly improve the quality of text localization. Specifically, a Text Feature Alignment Module (TFAM) is proposed to dynamically adjust the receptive fields of features based on initial raw detections; a Position-Aware Non-Maximum Suppression (PA-NMS) module is devised to selectively concentrate on reliable raw detections and exclude unreliable ones; besides, we propose an Instance-wise IoU loss for balanced training to deal with text instances of different scales. An extensive ablation study demonstrates the effectiveness and superiority of the proposed strategies. The resulting text detection system, which integrates the proposed strategies with a leading scene text detector EAST, achieves state-of-the-art or competitive performance on various standard benchmarks for text detection while keeping a fast running speed. | ['Xiang Bai', 'Yongpan Wang', 'Cong Yao', 'Wenqing Cheng', 'Jun Tang', 'Humen Zhong', 'Zhibo Yang', 'Minghui Liao', 'Minghang He'] | 2021-04-02 | null | http://openaccess.thecvf.com//content/CVPR2021/html/He_MOST_A_Multi-Oriented_Scene_Text_Detector_With_Localization_Refinement_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/He_MOST_A_Multi-Oriented_Scene_Text_Detector_With_Localization_Refinement_CVPR_2021_paper.pdf | cvpr-2021-1 | ['scene-text-detection'] | ['computer-vision'] | [ 5.69761992e-01 -5.07509887e-01 1.60325050e-01 -2.27769330e-01
-6.88470364e-01 -2.87069291e-01 8.61817360e-01 4.20719266e-01
-6.39935136e-01 2.17606962e-01 -9.74907503e-02 -2.17573628e-01
-1.37105985e-02 -6.39974952e-01 -2.86538512e-01 -8.61410618e-01
4.16771084e-01 3.16063076e-01 9.04742599e-01 -1.99445650e-01
7.01710105e-01 4.83164847e-01 -1.71721601e+00 1.37654051e-01
9.22447860e-01 9.37705338e-01 5.51774263e-01 7.71282494e-01
-4.51013356e-01 4.05569643e-01 -5.87628961e-01 -3.02938342e-01
3.48935336e-01 -2.03734219e-01 -3.28652292e-01 4.20339048e-01
4.41322595e-01 -1.91262603e-01 -2.61730373e-01 1.02335668e+00
7.04949081e-01 1.27469659e-01 7.01506376e-01 -7.64097393e-01
1.28813805e-02 3.48796099e-01 -9.52861369e-01 4.67583001e-01
2.80590862e-01 1.11622354e-02 1.01282930e+00 -1.13310969e+00
2.95791328e-01 8.58253181e-01 6.01439655e-01 1.66119263e-01
-8.33489478e-01 -4.12960619e-01 5.21525025e-01 -2.46699993e-02
-1.43716598e+00 -4.95998293e-01 5.98983049e-01 -1.62327379e-01
1.02892351e+00 3.86353225e-01 3.66345018e-01 8.15600932e-01
2.34218925e-01 9.09904122e-01 6.80029631e-01 -7.07169473e-01
2.40533888e-01 1.85748443e-01 9.22887102e-02 6.08372092e-01
4.26584721e-01 -5.24460435e-01 -7.36967623e-01 -4.27164994e-02
4.14411217e-01 6.66972399e-02 -1.42599091e-01 -4.36950326e-01
-1.31704247e+00 6.70584023e-01 1.89637229e-01 5.18960357e-01
-3.89090478e-02 -2.17693999e-01 5.65046430e-01 1.24647342e-01
5.63883007e-01 2.96946406e-01 -1.98894083e-01 1.20214187e-01
-1.22259927e+00 1.50950983e-01 5.39158404e-01 8.62862647e-01
4.44865495e-01 3.78384301e-03 -3.99750203e-01 9.52354193e-01
1.24066003e-01 6.41228199e-01 4.87627089e-01 1.63247809e-01
8.54500473e-01 8.65319133e-01 8.07970166e-02 -1.11069822e+00
-6.19166970e-01 -4.36783254e-01 -7.40043581e-01 -9.13498327e-02
1.74482688e-01 1.91808417e-01 -1.03355694e+00 9.88779902e-01
5.45509279e-01 -7.34194592e-02 -2.34947577e-01 8.63315821e-01
5.57747841e-01 6.57134831e-01 -6.84534684e-02 -1.57765836e-01
1.29222417e+00 -7.92291403e-01 -6.31537795e-01 -5.24452567e-01
7.60439336e-01 -1.06540608e+00 9.21448767e-01 3.92886758e-01
-8.79393995e-01 -3.86591971e-01 -1.21127033e+00 1.26246020e-01
-5.50825417e-01 5.93536854e-01 2.79683352e-01 8.25670540e-01
-8.62452507e-01 2.55009294e-01 -7.85973072e-01 -8.64394963e-01
1.71082988e-01 5.35888731e-01 -9.11095887e-02 1.94793388e-01
-7.01611340e-01 5.96855342e-01 4.44410712e-01 2.81082779e-01
-3.42006624e-01 -6.80631128e-05 -7.01709449e-01 6.41825721e-02
6.15957677e-01 -4.80314672e-01 8.55278373e-01 -9.29357469e-01
-1.31548107e+00 7.22559631e-01 -1.85658574e-01 -3.93460572e-01
8.75582397e-01 -2.66808361e-01 -2.82874644e-01 2.30312690e-01
2.75005609e-01 3.72689158e-01 1.40300620e+00 -8.65651846e-01
-1.01446962e+00 -4.63281423e-01 -3.56160492e-01 6.79285824e-01
-8.42876017e-01 2.38211378e-01 -7.24062622e-01 -8.35559547e-01
4.70365644e-01 -5.67122996e-01 -1.03527412e-01 -6.03003129e-02
-5.67443848e-01 -1.27526492e-01 9.55086827e-01 -2.88195193e-01
1.33086967e+00 -2.16355824e+00 6.53903261e-02 1.86300322e-01
1.99384227e-01 3.62980932e-01 4.34462652e-02 4.60250765e-01
3.58056486e-01 -2.33276725e-01 -1.72097862e-01 -6.79659724e-01
-1.18883848e-01 -3.20161879e-01 -4.58695799e-01 8.17689896e-01
1.74831346e-01 3.54464084e-01 -6.37789786e-01 -7.41927087e-01
7.36720145e-01 1.84827343e-01 -2.53336310e-01 4.96212021e-02
-1.64439753e-01 6.85843751e-02 -7.52601743e-01 7.06624329e-01
9.08141792e-01 -1.93752870e-01 7.91089758e-02 8.61065909e-02
-4.69232261e-01 8.89351293e-02 -1.59373605e+00 1.33536005e+00
-1.23094894e-01 6.57803893e-01 1.83259234e-01 -9.01844203e-01
9.61075366e-01 -6.22607879e-02 2.44484231e-01 -6.88077927e-01
4.64638203e-01 2.29469121e-01 -3.70274484e-01 -3.73566091e-01
1.00780928e+00 3.75419021e-01 -7.53645748e-02 2.21271113e-01
-2.97384769e-01 -5.91273941e-02 2.56158978e-01 1.99973747e-01
1.23956299e+00 -9.79296397e-03 3.16203654e-01 -2.26849690e-01
8.62348616e-01 -2.67654173e-02 8.50284379e-03 1.10275555e+00
-1.82062656e-01 6.46552920e-01 2.40306735e-01 -3.95064563e-01
-8.47024500e-01 -7.59646237e-01 -2.16578975e-01 1.46522510e+00
4.84375834e-01 -4.93110865e-01 -6.24170423e-01 -7.97030330e-01
-1.57168388e-01 4.00384963e-01 -5.35391748e-01 9.58951861e-02
-5.63059628e-01 -9.96093214e-01 4.24958289e-01 4.63464171e-01
6.72298372e-01 -9.59736526e-01 -8.58957946e-01 8.74003395e-02
4.01172638e-02 -1.27533424e+00 -3.79694611e-01 5.42410553e-01
-6.90657794e-01 -7.30716467e-01 -7.63305664e-01 -9.74709570e-01
7.38005638e-01 8.66108358e-01 6.00263476e-01 1.57032952e-01
-4.30748373e-01 1.12072520e-01 -7.36540616e-01 -2.13798761e-01
-1.84454456e-01 3.67279917e-01 -4.95817252e-02 2.56649822e-01
2.40714028e-01 -6.52659237e-02 -5.62865615e-01 5.06116629e-01
-1.04383707e+00 9.68105942e-02 7.70559311e-01 9.12130117e-01
4.12787050e-01 2.91169733e-01 3.59287739e-01 -7.61422276e-01
3.72346759e-01 -8.55580792e-02 -7.88355052e-01 1.76792443e-01
-5.33800304e-01 4.14215699e-02 9.08749938e-01 -3.70722502e-01
-1.02478755e+00 3.68149400e-01 -9.10874456e-02 -4.67391871e-02
-8.77065212e-02 1.68529108e-01 -4.76876348e-02 -2.15830967e-01
6.29575193e-01 7.08287656e-01 -6.78942382e-01 -3.31545323e-01
1.15597978e-01 1.03661609e+00 2.90528715e-01 -2.54299194e-01
8.56622458e-01 7.73150325e-01 -8.09744373e-02 -1.24316776e+00
-6.87447369e-01 -1.11063981e+00 -7.36109257e-01 -5.16099147e-02
5.28109908e-01 -8.92364502e-01 -2.33184487e-01 6.67569101e-01
-7.98835576e-01 -6.57602251e-02 1.96959242e-01 1.36381835e-01
-2.26922348e-01 7.65149713e-01 -2.78794169e-01 -9.91337061e-01
-6.15135014e-01 -9.98536587e-01 1.59517574e+00 1.76530749e-01
2.78591007e-01 -6.11998558e-01 -5.13367765e-02 3.93506661e-02
2.45867908e-01 -2.47780740e-01 6.07338548e-01 -8.14328372e-01
-4.74097073e-01 -5.38667202e-01 -4.80321258e-01 -5.62884398e-02
3.07991933e-02 -5.07642254e-02 -9.59901690e-01 -4.49273080e-01
-1.46558464e-01 2.33369227e-02 1.18960559e+00 3.54163885e-01
9.23745573e-01 2.62060523e-01 -6.66648328e-01 5.13104975e-01
1.38375962e+00 -3.35909799e-02 3.42028886e-01 6.38325691e-01
6.91548347e-01 3.60462785e-01 8.73256743e-01 7.94957757e-01
7.79870944e-03 7.69390881e-01 3.91257018e-01 -1.25187770e-01
1.72536790e-01 -2.26918682e-02 2.46612266e-01 4.98133242e-01
3.89586508e-01 -6.46781147e-01 -7.78572202e-01 5.91285169e-01
-1.96701193e+00 -4.82188493e-01 -2.31120005e-01 2.38150001e+00
1.53233349e-01 6.58456028e-01 2.21076488e-01 4.64207917e-01
1.01765859e+00 4.35131550e-01 -4.80060488e-01 -1.26575887e-01
-2.44584754e-01 -6.76224567e-03 6.27068818e-01 1.18501261e-01
-1.57669353e+00 1.23247039e+00 5.64262009e+00 1.06105602e+00
-1.17313385e+00 -4.28748764e-02 2.00513020e-01 1.26272021e-02
3.09470743e-01 -1.38715491e-01 -1.22488523e+00 2.60525554e-01
4.05019760e-01 5.62965088e-02 1.05601780e-01 8.55851054e-01
2.54671931e-01 -3.87470305e-01 -6.41874850e-01 9.85976577e-01
3.61884326e-01 -9.42841709e-01 1.64182872e-01 -2.45570138e-01
5.65981150e-01 -5.20485220e-03 1.12968341e-01 9.12785456e-02
-3.48169446e-01 -5.42908728e-01 8.29412460e-01 -2.23897910e-03
6.94427133e-01 -6.92395926e-01 7.98713386e-01 4.59531337e-01
-1.51970661e+00 -1.85422644e-01 -6.53122723e-01 9.07436758e-02
-1.74231201e-01 6.90121531e-01 -1.02502859e+00 5.64463556e-01
6.41733706e-01 5.71369350e-01 -1.01594841e+00 1.22780693e+00
-4.66312096e-02 4.25444722e-01 -5.71661532e-01 -6.63822830e-01
2.59112984e-01 1.44074485e-01 7.35263586e-01 1.57819033e+00
3.26848119e-01 -3.60815048e-01 2.16675967e-01 2.96555132e-01
2.14402691e-01 7.21457481e-01 -4.43461448e-01 2.04639032e-01
3.79632264e-01 1.53847373e+00 -1.51815236e+00 -2.48950541e-01
-4.08922911e-01 1.13030684e+00 2.30899140e-01 2.41703466e-02
-7.91363895e-01 -6.42482102e-01 -3.55427153e-02 1.52042851e-01
6.14263356e-01 -8.19378346e-02 -4.43376362e-01 -1.30603254e+00
3.21996063e-01 -5.97386956e-01 4.76850688e-01 -5.76963007e-01
-8.96622121e-01 6.80468857e-01 -2.19084516e-01 -1.42635238e+00
7.14631900e-02 -7.07647979e-01 -6.34838939e-01 4.89893794e-01
-1.39745796e+00 -1.17901599e+00 -4.68890101e-01 5.39092600e-01
1.09251273e+00 -3.86562161e-02 2.91377664e-01 2.35590905e-01
-8.18636358e-01 7.60103106e-01 4.23307747e-01 2.48276703e-02
8.56781960e-01 -1.14966190e+00 6.56364202e-01 1.23223841e+00
1.05752088e-01 2.87407190e-01 8.70463967e-01 -7.53563523e-01
-1.44721413e+00 -1.11526263e+00 7.22617447e-01 -2.24279001e-01
6.65381670e-01 -8.40337753e-01 -8.12741220e-01 3.58725786e-01
-2.14986950e-01 -1.79553032e-01 -6.33154511e-02 1.89094599e-02
-1.54260486e-01 -4.25861366e-02 -9.24097657e-01 7.72497654e-01
9.34236467e-01 -1.33200169e-01 -4.37717319e-01 5.89753032e-01
2.08022088e-01 -4.73189682e-01 3.69636863e-02 3.72818291e-01
4.35422599e-01 -1.10311794e+00 6.99296772e-01 1.34669051e-01
-9.89803821e-02 -5.49509287e-01 4.12545400e-03 -6.89714789e-01
-1.89886212e-01 -6.33752108e-01 1.38512388e-01 1.14024162e+00
1.38178393e-01 -6.31148756e-01 8.72619331e-01 -1.69411451e-01
-1.95602074e-01 -4.55016583e-01 -1.10317886e+00 -5.83507836e-01
-2.99238414e-01 -3.84121269e-01 2.47282945e-02 5.07359803e-01
1.48759447e-02 3.66268188e-01 -4.30154711e-01 3.16983879e-01
3.65702897e-01 1.77186131e-01 8.13157618e-01 -1.16544819e+00
-1.49647459e-01 -5.77295423e-01 -6.01114273e-01 -1.38300073e+00
-1.34760037e-01 -5.07065117e-01 5.65220714e-01 -1.36713398e+00
3.48411173e-01 -3.14840943e-01 -2.19106600e-01 1.79223105e-01
-5.29446006e-01 2.41858959e-01 6.52748644e-02 1.66269362e-01
-1.11587250e+00 4.51624483e-01 6.65420592e-01 -8.98446888e-02
-4.32429999e-01 2.08774924e-01 -3.33499551e-01 6.33325934e-01
5.97316802e-01 -4.37452495e-01 -1.78351253e-01 -2.58975178e-01
4.78040636e-01 -3.41962159e-01 1.38920143e-01 -1.29936290e+00
5.11077940e-01 1.29511490e-01 5.27237892e-01 -1.11067295e+00
2.05252379e-01 -6.67607069e-01 -4.74378139e-01 2.63545245e-01
-1.99587032e-01 -6.45655394e-02 3.04145515e-01 7.70198882e-01
9.42322761e-02 -4.26890314e-01 8.28816354e-01 3.15371096e-01
-6.94031179e-01 -7.61562362e-02 -6.13338411e-01 -2.85017759e-01
9.29205835e-01 -3.99009675e-01 -3.24418843e-01 -5.17747775e-02
-1.54520437e-01 2.17847362e-01 6.20030582e-01 3.32098335e-01
5.93824565e-01 -6.20380878e-01 -6.11903250e-01 3.51940781e-01
4.38574016e-01 -6.41681924e-02 3.00842643e-01 9.87056792e-01
-5.24153829e-01 5.16763687e-01 2.71271229e-01 -8.19322407e-01
-1.42034292e+00 5.83603799e-01 1.94026485e-01 -4.74039763e-01
-9.93439853e-01 8.03416312e-01 3.31708074e-01 -9.92626324e-02
3.67662311e-01 -3.44649702e-01 -2.83234715e-01 -4.08058949e-02
6.24406099e-01 2.80730784e-01 5.52610219e-01 -6.32028461e-01
-3.76070678e-01 7.82420695e-01 -4.26152885e-01 -1.44221149e-02
9.39117134e-01 -3.91721278e-01 3.26173365e-01 3.48871082e-01
5.36316752e-01 2.81794578e-01 -1.11881971e+00 -1.77306235e-01
-1.44060412e-02 -5.61836183e-01 2.72619069e-01 -5.70802033e-01
-5.99714279e-01 7.18518436e-01 6.88731790e-01 5.35738409e-01
1.39184463e+00 -1.63222298e-01 6.12475872e-01 6.71235561e-01
1.49134681e-01 -1.21215272e+00 1.05795331e-01 6.33989990e-01
3.50955576e-01 -1.20794129e+00 2.67712027e-01 -5.64678073e-01
-4.28309649e-01 1.26780856e+00 4.25652951e-01 -1.77987993e-01
2.94013381e-01 4.15068001e-01 -2.07532078e-01 -7.69994780e-02
-5.34839272e-01 -5.07328570e-01 1.26731291e-01 1.66968420e-01
1.44054860e-01 -2.28982672e-01 -4.90947753e-01 7.19406307e-02
2.49985605e-01 -4.42001164e-01 3.25190693e-01 1.10708129e+00
-9.73947465e-01 -7.50580251e-01 -6.54839158e-01 7.48733521e-01
-5.36389947e-01 -2.10196674e-01 -5.26902437e-01 7.87609816e-01
-9.88741666e-02 1.04251432e+00 -2.31092740e-02 -4.62452650e-01
3.23718250e-01 -9.61358920e-02 2.82265306e-01 -5.72742283e-01
-7.57978022e-01 5.80865800e-01 -1.45968333e-01 -2.38672182e-01
-3.56357336e-01 -7.13811994e-01 -1.12918091e+00 -3.79242934e-02
-9.74666536e-01 -1.50073364e-01 7.59118378e-01 1.05230618e+00
1.07907191e-01 3.99709910e-01 8.07423592e-01 -9.73973691e-01
-5.28665900e-01 -1.04099190e+00 -6.52738690e-01 1.35855719e-01
2.62848079e-01 -6.43374801e-01 -6.47012472e-01 -8.85861441e-02] | [12.011011123657227, 2.3204824924468994] |
a18576be-61d9-4810-81cb-645501d6b688 | infinite-action-contextual-bandits-with | 2302.08551 | null | https://arxiv.org/abs/2302.08551v2 | https://arxiv.org/pdf/2302.08551v2.pdf | Infinite Action Contextual Bandits with Reusable Data Exhaust | For infinite action contextual bandits, smoothed regret and reduction to regression results in state-of-the-art online performance with computational cost independent of the action set: unfortunately, the resulting data exhaust does not have well-defined importance-weights. This frustrates the execution of downstream data science processes such as offline model selection. In this paper we describe an online algorithm with an equivalent smoothed regret guarantee, but which generates well-defined importance weights: in exchange, the online computational cost increases, but only to order smoothness (i.e., still independent of the action set). This removes a key obstacle to adoption of smoothed regret in production scenarios. | ['Paul Mineiro', 'Yinglun Zhu', 'Mark Rucker'] | 2023-02-16 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [ 2.04120561e-01 3.07659477e-01 -6.13001108e-01 -3.43409777e-01
-1.20686221e+00 -8.70809317e-01 1.90393806e-01 1.37031332e-01
-4.95985150e-01 1.28083014e+00 2.52632856e-01 -6.37860775e-01
-9.03306961e-01 -6.02703631e-01 -9.42712188e-01 -7.03129590e-01
-1.52633682e-01 5.62255144e-01 -1.55421898e-01 -5.62247150e-02
2.88777858e-01 4.23999757e-01 -1.57835960e+00 8.40997770e-02
6.66353405e-01 1.02757382e+00 -1.84650362e-01 7.86661506e-01
-7.15574920e-02 8.72184694e-01 -2.85410196e-01 -6.23983026e-01
8.45543444e-01 -5.54313540e-01 -8.26139092e-01 -2.54616123e-02
2.53828347e-01 -4.23203886e-01 -2.24840373e-01 8.71393561e-01
1.77572861e-01 3.47466409e-01 2.70858526e-01 -1.06089449e+00
-4.35031801e-01 7.11084723e-01 -4.83646005e-01 8.92469361e-02
-6.83447868e-02 2.07072586e-01 1.48575330e+00 -3.02984454e-02
7.33845294e-01 1.09825218e+00 6.35036588e-01 5.40652573e-01
-1.56665754e+00 -2.46254593e-01 5.97881317e-01 -1.69517204e-01
-7.88620532e-01 -6.24039114e-01 3.20233107e-01 -2.26516783e-01
7.14132607e-01 9.48723495e-01 9.72288907e-01 8.30886066e-01
-3.20600033e-01 8.04791629e-01 1.05793500e+00 -2.08384991e-01
6.33139670e-01 -1.64484575e-01 2.78963089e-01 2.45433554e-01
5.29895186e-01 5.11555374e-01 -5.81189573e-01 -2.32489362e-01
8.00878525e-01 1.22194961e-02 1.41030073e-01 -5.64739168e-01
-6.84950113e-01 8.74548316e-01 9.27086771e-02 -8.01086128e-02
-3.86825949e-01 5.89863241e-01 3.83811146e-01 7.02423632e-01
7.18004227e-01 5.75863719e-01 -8.41622472e-01 -5.66372514e-01
-9.38769460e-01 5.32349527e-01 8.37119877e-01 9.62815821e-01
5.86666465e-01 -1.54818714e-01 -3.98616195e-01 5.36855459e-01
-1.17488772e-01 3.19718331e-01 -1.06622793e-01 -1.25355566e+00
6.10200107e-01 3.63503009e-01 7.25856245e-01 -3.76366943e-01
-5.18701911e-01 -7.37089097e-01 -4.76385981e-01 2.93247312e-01
9.54912663e-01 -3.30796450e-01 -7.64163435e-01 1.57108104e+00
5.40122926e-01 -1.77136391e-01 -1.13355733e-01 1.04383588e+00
-1.05534121e-01 2.43180752e-01 9.47066694e-02 -4.38637823e-01
8.10263515e-01 -8.34162891e-01 -5.24161637e-01 -3.34237009e-01
6.24000013e-01 -3.64684403e-01 9.90808249e-01 5.26279747e-01
-1.33558667e+00 7.56671354e-02 -7.10099995e-01 1.60072446e-02
-8.36116970e-02 -4.55709547e-01 1.43639350e+00 8.54381382e-01
-7.68050551e-01 9.96660531e-01 -7.83763826e-01 -3.97704169e-02
5.84367990e-01 4.08349633e-01 -1.40935034e-01 4.53302786e-02
-8.40620518e-01 9.01378751e-01 2.50951618e-01 -2.05487739e-02
-7.01759160e-01 -1.30948186e+00 -4.07571554e-01 2.61496186e-01
6.32009864e-01 -6.81944549e-01 1.57604349e+00 -1.35129774e+00
-1.58052230e+00 5.25227249e-01 -3.94585319e-02 -8.13023150e-01
1.21520698e+00 -3.71014535e-01 -7.65567720e-02 -2.83707559e-01
-2.16949955e-01 8.49184319e-02 6.08573735e-01 -8.41664493e-01
-1.04622471e+00 -4.27440286e-01 4.44667071e-01 3.44245285e-01
-1.63122769e-02 1.03891917e-01 -3.04000407e-01 -2.76141971e-01
-3.21063250e-02 -7.70624399e-01 -7.02584624e-01 -5.23739792e-02
-7.51226321e-02 2.02470452e-01 -1.96409613e-01 -5.90588272e-01
1.46681046e+00 -1.98672462e+00 -1.47136539e-01 3.80654842e-01
-2.68473668e-04 -2.96888649e-01 -9.48823690e-02 4.05200511e-01
1.62586235e-02 5.06644905e-01 2.04328634e-02 -9.77144092e-02
2.92314172e-01 5.37382774e-02 -2.34427050e-01 7.83148885e-01
4.98980619e-02 7.30313599e-01 -1.05285823e+00 -1.16834000e-01
2.38225579e-01 -2.99742848e-01 -9.81436014e-01 -8.12404603e-02
-4.52500194e-01 3.09694916e-01 -5.95621645e-01 4.87373859e-01
4.94296253e-01 -1.14576466e-01 2.05305070e-01 3.56043756e-01
-4.94795084e-01 2.79167712e-01 -1.22601557e+00 1.43088114e+00
-4.70877767e-01 3.49178702e-01 3.59872907e-01 -1.05761969e+00
4.11248118e-01 -1.26049131e-01 8.99538875e-01 -6.53416216e-01
-2.52109915e-01 1.94360122e-01 9.13096517e-02 -1.63019136e-01
3.48633051e-01 -3.01390111e-01 7.13585168e-02 4.00219560e-01
-1.80861771e-01 1.74220175e-01 3.77151191e-01 -2.53720760e-01
1.26177502e+00 3.15564096e-01 1.11611597e-01 -5.53002298e-01
-7.74087459e-02 2.93044716e-01 6.56777442e-01 1.22850311e+00
-9.48064253e-02 4.43382174e-01 8.62570405e-01 -6.34947956e-01
-1.39838934e+00 -7.97457397e-01 -1.95393413e-01 1.52214611e+00
2.34275796e-02 -9.58641842e-02 -6.10517681e-01 -5.36819935e-01
6.88856721e-01 8.43483329e-01 -8.24895918e-01 -4.60267030e-02
-3.24781269e-01 -7.09984660e-01 2.42578492e-01 1.69939905e-01
-2.52878785e-01 -7.19804943e-01 -5.80612838e-01 5.47711730e-01
3.10156584e-01 -5.73824763e-01 -2.27783754e-01 3.43439579e-01
-1.00366211e+00 -9.80182827e-01 -4.52727228e-01 2.22706750e-01
5.09714127e-01 1.50104836e-01 1.26774669e+00 -9.95434076e-02
-1.40897393e-01 2.21118644e-01 -2.27636978e-01 -7.01959848e-01
-1.13316372e-01 -9.32185799e-02 -2.14324832e-01 -7.39389658e-02
3.52112055e-02 -4.24094856e-01 -9.06085014e-01 2.03477353e-01
-7.21473873e-01 3.95496283e-03 6.08213603e-01 9.73950624e-01
7.03505576e-01 -2.58822262e-01 7.36046135e-01 -1.10540330e+00
5.33097684e-01 -4.59909588e-01 -1.21690083e+00 4.52320367e-01
-9.00317430e-01 3.15280080e-01 5.72821558e-01 -2.15924323e-01
-1.14347768e+00 9.44257677e-02 1.15926869e-01 -1.89573795e-01
2.69183993e-01 5.67236125e-01 3.59109640e-01 1.52320832e-01
5.89225471e-01 -6.63889050e-02 -1.26863057e-02 -6.92956686e-01
5.75813949e-01 2.15002164e-01 -9.42226779e-03 -5.88911593e-01
3.50236982e-01 4.76827472e-01 2.60466129e-01 -4.70548987e-01
-1.17131686e+00 -4.00804460e-01 -1.36579245e-01 -9.09743086e-02
3.81878167e-01 -6.49986684e-01 -1.05467355e+00 -3.43694203e-02
-6.06832981e-01 -5.75750530e-01 -7.74520338e-01 6.52336180e-01
-1.04141426e+00 -1.23881221e-01 -3.77485454e-01 -1.48065007e+00
-2.12563738e-01 -6.39269888e-01 7.07991123e-01 -1.99272297e-02
-1.25110850e-01 -7.28460729e-01 -8.23841915e-02 5.61682820e-01
2.74415821e-01 3.33544165e-01 4.99246359e-01 -6.23884261e-01
-6.47338629e-01 -3.48209888e-01 -2.31256291e-01 8.09548125e-02
-1.99845135e-02 2.64807165e-01 -7.84785807e-01 -2.62966096e-01
-2.23459363e-01 -2.60742307e-01 8.20620358e-01 7.51639366e-01
1.36257863e+00 -6.00851774e-01 1.60958946e-01 6.04104280e-01
1.36850202e+00 -3.70924268e-03 3.21890861e-01 3.84862125e-01
1.82833374e-01 5.52159727e-01 9.02480066e-01 9.42116082e-01
-1.86235443e-01 2.48142332e-01 4.70850557e-01 -1.04243763e-01
2.84483641e-01 -3.84212404e-01 1.64521396e-01 -3.75213474e-02
-2.67821282e-01 6.12896793e-02 -6.78375006e-01 5.14430881e-01
-2.28515172e+00 -1.17731190e+00 -1.36830240e-01 2.62227821e+00
1.11718655e+00 4.32387382e-01 4.38485265e-01 -2.27236465e-01
6.05740130e-01 -2.19427459e-02 -1.07463753e+00 -6.78283572e-01
2.24517304e-02 2.02514321e-01 1.50832331e+00 7.25695729e-01
-9.05843079e-01 7.26836801e-01 6.76212120e+00 1.01350808e+00
-7.58947313e-01 1.90943986e-01 8.86742413e-01 -8.74018252e-01
-4.69050825e-01 2.75195539e-01 -5.65778017e-01 4.71834183e-01
1.13842034e+00 -5.64041734e-01 1.04770350e+00 1.16056442e+00
6.18778706e-01 -2.01285362e-01 -1.16265202e+00 7.02039123e-01
-8.80024850e-01 -1.62255907e+00 -4.76410776e-01 2.79473782e-01
8.96106601e-01 1.22561604e-01 -1.81787625e-01 3.38195473e-01
8.05236816e-01 -9.44284797e-01 9.71376121e-01 6.40565872e-01
7.99756408e-01 -1.15753651e+00 5.01812041e-01 4.22244072e-01
-4.99290168e-01 -4.15031046e-01 -4.92395908e-01 -4.30345684e-01
-5.16281649e-02 8.55898619e-01 -4.42431241e-01 5.60140491e-01
5.62942803e-01 2.44682163e-01 7.13740615e-03 1.08076632e+00
1.01115674e-01 6.56412840e-01 -7.58746684e-01 -3.10165793e-01
4.03680950e-01 -4.79393244e-01 2.81896949e-01 1.06526124e+00
1.04714856e-01 1.22654088e-01 9.08008665e-02 5.70905805e-01
-9.85514149e-02 8.14388543e-02 -3.72487664e-01 -2.76120692e-01
4.74115133e-01 8.64253938e-01 -3.34054232e-01 -1.26911670e-01
-4.16178703e-01 4.41887289e-01 5.00819862e-01 3.37863088e-01
-7.26145923e-01 1.07625887e-01 1.11932719e+00 3.32310125e-02
1.84195355e-01 4.85787131e-02 -8.07034373e-01 -8.82419348e-01
1.36277661e-01 -6.90063894e-01 6.51586413e-01 -2.71528345e-02
-1.45407748e+00 -3.13456774e-01 -1.07560880e-01 -9.15363848e-01
-1.43346012e-01 -5.61857164e-01 -2.32011065e-01 8.70008826e-01
-1.37326574e+00 -7.81957865e-01 3.38508010e-01 2.95315772e-01
3.72282594e-01 4.15172487e-01 5.66244721e-01 1.60327092e-01
-5.45686066e-01 6.55779779e-01 8.37959945e-01 -3.84943038e-01
4.84553516e-01 -1.34440076e+00 3.93174767e-01 6.28218591e-01
-1.76851034e-01 4.79506403e-01 9.14379358e-01 -5.96080720e-01
-1.80296731e+00 -9.27593350e-01 5.06855667e-01 -4.23257351e-01
1.18622720e+00 -2.88776398e-01 -2.24597842e-01 6.20741785e-01
-3.33449721e-01 4.73512383e-03 5.23966253e-01 8.08967650e-01
2.86752041e-02 -5.41502178e-01 -1.24251199e+00 7.71444201e-01
1.46975708e+00 -1.88026354e-01 -3.33961472e-02 6.56362712e-01
6.73106372e-01 -4.31844324e-01 -7.86004424e-01 2.72984505e-01
9.42666292e-01 -1.10674381e+00 8.09815824e-01 -1.41179502e+00
3.21866482e-01 2.91741371e-01 7.57114077e-03 -9.93746161e-01
-4.56290811e-01 -1.07450366e+00 -3.35456818e-01 9.98751521e-01
8.16966534e-01 -7.82553375e-01 1.03438163e+00 1.43604040e+00
-5.13756387e-02 -9.75515723e-01 -1.28867972e+00 -1.06407571e+00
6.82446584e-02 -6.25592113e-01 5.22997975e-01 7.02924728e-01
3.47620174e-02 -2.75265556e-02 -5.77857673e-01 -1.90537021e-01
7.76582003e-01 4.19935822e-01 7.62520254e-01 -1.14464271e+00
-5.68843067e-01 -7.46955991e-01 5.99060431e-02 -8.87854636e-01
-3.29837024e-01 -3.64932239e-01 -4.56143878e-02 -1.18086278e+00
1.15555227e-01 -6.97110951e-01 -6.93361878e-01 3.18644077e-01
-1.50913283e-01 -3.01590264e-01 2.11326376e-01 -5.44986948e-02
-8.78129125e-01 3.82631242e-01 1.19650960e+00 1.31354943e-01
-3.67808133e-01 2.67331302e-01 -9.65681612e-01 6.08847618e-01
1.12815726e+00 -6.40994191e-01 -4.48583841e-01 -2.40965426e-01
4.81028885e-01 3.14542472e-01 1.56144023e-01 -5.46556473e-01
-2.55308121e-01 -1.06537199e+00 2.10215643e-01 -2.90166020e-01
4.98432778e-02 -8.44223559e-01 4.02556926e-01 5.72730958e-01
-7.52195537e-01 -2.25673497e-01 2.10022386e-02 7.23761857e-01
2.92895615e-01 -4.24237430e-01 6.28681839e-01 -3.74836296e-01
-1.81949928e-01 2.28936195e-01 1.61015578e-02 1.87449872e-01
9.59619522e-01 -1.65856779e-01 -3.04069489e-01 -4.53237712e-01
-6.19933665e-01 3.77368808e-01 4.16704923e-01 6.94746450e-02
-3.70919071e-02 -1.13376474e+00 -9.55346823e-01 -2.31917188e-01
-1.26010492e-01 -9.63014588e-02 4.15773809e-01 9.47343886e-01
-4.21708971e-01 4.30083603e-01 8.75570476e-02 -3.30950990e-02
-8.99439692e-01 7.39683867e-01 2.03585610e-01 -6.06823623e-01
-4.94857788e-01 9.23946083e-01 5.62840365e-02 -1.18005611e-01
3.21138322e-01 -3.59846681e-01 6.08751357e-01 2.42873281e-01
5.04926026e-01 6.14866376e-01 1.65091872e-01 3.84155363e-01
-1.42916709e-01 -1.85224116e-01 1.54735789e-01 -1.86408848e-01
1.61870348e+00 -7.25899115e-02 9.21469331e-02 3.48391443e-01
8.81049693e-01 8.67856741e-02 -1.82837832e+00 1.21750019e-01
3.51195157e-01 -7.52046466e-01 2.08603039e-01 -1.05106044e+00
-8.12877715e-01 2.61899501e-01 3.50714713e-01 5.65418422e-01
8.87692451e-01 -2.35218242e-01 4.17198360e-01 5.85654199e-01
5.05804300e-01 -1.66496658e+00 -6.97554648e-01 2.90993601e-01
9.78969693e-01 -1.16614342e+00 3.23068976e-01 -9.96660069e-02
-6.00351930e-01 7.92169571e-01 2.42197603e-01 -1.76102072e-01
3.71811658e-01 1.56664819e-01 -2.08685771e-01 1.38625726e-01
-1.07760155e+00 -4.67910677e-01 3.57908458e-02 2.05885097e-01
2.18080878e-01 2.88102120e-01 -7.89388239e-01 6.90910161e-01
-1.71653017e-01 4.48134281e-02 1.36056691e-01 8.62444282e-01
-4.27416623e-01 -9.24758792e-01 -2.89456904e-01 9.81799841e-01
-7.29913712e-01 -3.01010489e-01 -3.35103124e-01 7.63346255e-01
-3.08632225e-01 9.80583727e-01 6.16767220e-02 6.02180362e-02
4.13810313e-01 -3.81200165e-02 5.23582757e-01 -2.66173065e-01
-7.85996437e-01 8.27555284e-02 7.06125617e-01 -8.45162928e-01
-9.36254486e-02 -7.11131513e-01 -1.03180659e+00 -6.13362074e-01
-4.00854051e-01 3.30435097e-01 6.75376594e-01 6.46689594e-01
4.38571304e-01 2.87573874e-01 7.68536329e-01 -5.80946922e-01
-1.30537260e+00 -7.41984010e-01 -6.09788001e-01 5.22728503e-01
6.04484558e-01 -4.43224251e-01 -3.03064406e-01 -2.83268809e-01] | [4.535826683044434, 3.293484926223755] |
326769b9-6de0-4c51-9309-be19fdd55ad0 | neural-face-identification-in-a-2d-wireframe-1 | 2203.04229 | null | https://arxiv.org/abs/2203.04229v1 | https://arxiv.org/pdf/2203.04229v1.pdf | Neural Face Identification in a 2D Wireframe Projection of a Manifold Object | In computer-aided design (CAD) systems, 2D line drawings are commonly used to illustrate 3D object designs. To reconstruct the 3D models depicted by a single 2D line drawing, an important key is finding the edge loops in the line drawing which correspond to the actual faces of the 3D object. In this paper, we approach the classical problem of face identification from a novel data-driven point of view. We cast it as a sequence generation problem: starting from an arbitrary edge, we adopt a variant of the popular Transformer model to predict the edges associated with the same face in a natural order. This allows us to avoid searching the space of all possible edge loops with various hand-crafted rules and heuristics as most existing methods do, deal with challenging cases such as curved surfaces and nested edge loops, and leverage additional cues such as face types. We further discuss how possibly imperfect predictions can be used for 3D object reconstruction. | ['Zihan Zhou', 'Jia Zheng', 'Kehan Wang'] | 2022-03-08 | neural-face-identification-in-a-2d-wireframe | http://openaccess.thecvf.com//content/CVPR2022/html/Wang_Neural_Face_Identification_in_a_2D_Wireframe_Projection_of_a_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_Neural_Face_Identification_in_a_2D_Wireframe_Projection_of_a_CVPR_2022_paper.pdf | cvpr-2022-1 | ['3d-object-reconstruction', 'object-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 3.97284418e-01 2.10835859e-01 -2.39827279e-02 -2.45872065e-02
-2.88406700e-01 -7.01721489e-01 4.57070112e-01 -1.62388325e-01
5.13460577e-01 1.61113754e-01 -1.25056699e-01 -4.65024084e-01
-1.93485525e-02 -7.50479877e-01 -5.03070951e-01 -1.81206539e-01
2.25400776e-02 7.86881745e-01 1.74619272e-01 -1.68882668e-01
5.25805533e-01 1.16092253e+00 -1.68073881e+00 2.78614730e-01
4.71592039e-01 8.45293999e-01 1.04583621e-01 5.52343428e-01
-4.39680338e-01 4.61295806e-02 -3.00782263e-01 -6.84374511e-01
4.60992187e-01 -3.41081947e-01 -4.67987984e-01 8.03203166e-01
2.82937527e-01 -4.30465311e-01 1.40211433e-02 8.27496350e-01
2.63021260e-01 -3.85942310e-01 9.30269241e-01 -1.21300840e+00
-1.53340146e-01 1.07641518e-01 -1.01116586e+00 -5.72658658e-01
9.42190886e-01 -1.63442805e-01 8.43274415e-01 -1.32954144e+00
9.97478783e-01 1.48556054e+00 8.18351984e-01 3.94356132e-01
-1.33608496e+00 -5.33793747e-01 4.44916375e-02 1.39308712e-02
-1.49408579e+00 -4.96538550e-01 1.44158971e+00 -6.91859484e-01
4.44706768e-01 3.48221183e-01 6.80410147e-01 7.90154159e-01
-1.54573068e-01 7.87546635e-01 8.64874363e-01 -6.69923544e-01
2.82146454e-01 -3.15355137e-03 -1.80157498e-01 9.88834083e-01
4.92380047e-03 2.29455787e-03 -3.84579420e-01 -4.88141537e-01
1.22496784e+00 -3.89592350e-03 -2.56212682e-01 -1.06441176e+00
-9.79738176e-01 6.22256815e-01 1.58037603e-01 -1.11496001e-01
-2.52783090e-01 -1.84961379e-01 9.10067484e-02 1.58573806e-01
2.78517604e-01 4.28660750e-01 -4.24628347e-01 1.75378516e-01
-7.63144016e-01 4.15121078e-01 9.24875915e-01 1.39472187e+00
9.13281620e-01 -3.39780480e-01 2.55584896e-01 8.11581910e-01
4.09596086e-01 -6.70516817e-03 -3.23291749e-01 -9.61632729e-01
3.44675303e-01 8.34556818e-01 2.51453251e-01 -1.14475858e+00
-2.49449506e-01 -1.27567247e-01 -5.93496025e-01 4.46283668e-01
4.16292876e-01 1.56247154e-01 -8.54906738e-01 1.03477776e+00
6.62145913e-01 1.28309116e-01 -4.79120642e-01 6.70820534e-01
4.92174357e-01 1.88608572e-01 -6.11976683e-01 -2.38777339e-01
1.42058468e+00 -6.72294617e-01 -2.59463400e-01 -1.88854143e-01
5.82992792e-01 -1.10973346e+00 7.63006985e-01 3.06709498e-01
-1.03777409e+00 -3.04650873e-01 -1.05402124e+00 -6.46218285e-02
2.66078562e-02 2.55241096e-01 3.42284441e-01 5.36153078e-01
-6.76410496e-01 7.05208778e-01 -4.92513806e-01 -1.33135155e-01
5.51787317e-01 2.44023398e-01 -4.40793872e-01 -3.57526451e-01
-3.48553389e-01 6.95196331e-01 5.20765968e-02 3.12126905e-01
-4.53189522e-01 -7.06962585e-01 -5.97940743e-01 -2.38732502e-01
8.37326825e-01 -7.43997395e-01 1.29347408e+00 -6.75038993e-01
-1.50479519e+00 1.28486931e+00 -4.30366606e-01 2.32007340e-01
8.87128651e-01 -2.00397253e-01 -1.09772541e-01 4.70871590e-02
-1.54892564e-01 3.83858800e-01 1.10301626e+00 -1.77478921e+00
-2.72377014e-01 -5.65501750e-01 1.29365819e-02 -2.18417585e-01
4.24261421e-01 -2.46167287e-01 -7.74539948e-01 -5.89813709e-01
6.58811510e-01 -9.35203135e-01 -2.62893885e-01 7.62610018e-01
-9.80850875e-01 -1.63531244e-01 8.96510363e-01 -5.50759852e-01
1.04468596e+00 -2.05071878e+00 6.93886355e-02 6.38564527e-01
2.32870698e-01 1.01181448e-01 -2.87048593e-02 7.21699893e-01
-4.98896837e-02 2.13246897e-01 -3.91708970e-01 -3.82161528e-01
-7.54712150e-02 5.85653298e-02 -1.39655441e-01 3.47571939e-01
3.15621883e-01 5.88287354e-01 -9.03327703e-01 -4.17769879e-01
2.60923356e-01 2.35345915e-01 -7.57331073e-01 1.56900704e-01
-4.81769890e-01 4.59174871e-01 -8.70177925e-01 6.76346779e-01
9.58172977e-01 -1.92802206e-01 3.94588381e-01 -3.55001241e-01
-1.51566967e-01 -2.57914960e-02 -1.49686992e+00 1.66943610e+00
-4.44680482e-01 3.02385658e-01 9.27057937e-02 -7.02808678e-01
1.36104238e+00 2.66812712e-01 4.41893905e-01 -1.22224949e-01
4.81292158e-02 4.51849312e-01 -2.58272886e-01 -3.77598971e-01
3.00692379e-01 -1.42118111e-01 4.73145097e-02 5.24430096e-01
-4.17682916e-01 -4.89771515e-01 -2.83255786e-01 -1.61071673e-01
8.77110302e-01 3.53152901e-01 5.61716497e-01 -8.24373141e-02
3.90323311e-01 -7.19488859e-02 3.50451797e-01 5.21300077e-01
3.15884233e-01 9.88744080e-01 8.31493676e-01 -5.34088552e-01
-1.47675145e+00 -8.00276756e-01 -9.08496305e-02 1.56961903e-01
1.74631417e-01 -5.78152955e-01 -5.55488765e-01 -5.42428017e-01
2.65087247e-01 4.76573139e-01 -5.35672188e-01 1.27816156e-01
-6.35535419e-01 4.17185016e-02 -1.74567506e-01 2.81390548e-01
8.17877427e-02 -6.23328745e-01 -7.08748817e-01 1.80107161e-01
3.80558640e-01 -1.03192437e+00 -4.83753830e-01 2.74222549e-02
-9.18181598e-01 -1.25987816e+00 -7.84969985e-01 -8.93664122e-01
1.14612448e+00 4.41978365e-01 1.19054139e+00 6.02047443e-01
-4.98089641e-01 1.91389993e-01 -2.63761729e-01 -1.45114124e-01
-4.89441484e-01 -2.45167807e-01 -2.62344450e-01 3.19070071e-02
1.42733335e-01 -7.24313080e-01 -5.16375721e-01 5.94076931e-01
-5.27014971e-01 5.50745189e-01 3.88990134e-01 7.82392681e-01
8.32448959e-01 -1.04106648e-03 2.21122622e-01 -9.67328370e-01
2.93760508e-01 -2.91166306e-01 -6.62754834e-01 3.59941870e-01
-9.89366770e-02 1.71055034e-01 5.83888233e-01 -3.68513018e-01
-7.06024349e-01 5.57248294e-01 -3.47624011e-02 -8.68054152e-01
-3.05117995e-01 2.83736736e-01 -4.18888241e-01 6.47028908e-02
2.97155976e-01 5.88236898e-02 3.96918431e-02 -8.48514378e-01
4.49334532e-01 5.17108917e-01 3.82365227e-01 -6.72259390e-01
1.02733862e+00 3.90417486e-01 3.85748953e-01 -9.41516101e-01
-5.08025825e-01 -1.96479619e-01 -1.11262226e+00 -2.86255985e-01
3.78407806e-01 -4.03636545e-01 -7.73712873e-01 3.24104846e-01
-1.53126585e+00 -8.73191133e-02 -1.56650379e-01 -1.13739900e-01
-8.57815266e-01 5.26577175e-01 -1.35206506e-01 -1.01530111e+00
1.79185137e-01 -1.15158284e+00 1.50898898e+00 -6.30949140e-02
-4.93359298e-01 -5.76198995e-01 -2.23531321e-01 -9.87033620e-02
-3.18690866e-01 4.23270106e-01 1.45184600e+00 -3.16307396e-01
-7.02312231e-01 -3.84388536e-01 -1.14230536e-01 -2.26420283e-01
4.10528928e-01 3.49193931e-01 -7.70252407e-01 1.05611145e-01
-2.26800695e-01 1.50919527e-01 2.37931311e-01 1.13211989e-01
1.23799706e+00 -2.66760707e-01 -7.56667316e-01 5.20969570e-01
1.23663664e+00 3.27295899e-01 4.85710829e-01 -1.08995460e-01
7.14603186e-01 1.02362025e+00 3.84663314e-01 6.86389267e-01
8.25845897e-02 8.79421294e-01 3.07398647e-01 2.12928578e-01
4.11640964e-02 -7.96912968e-01 -2.43575454e-01 3.76854926e-01
1.58561498e-01 -1.61596775e-01 -9.89923894e-01 2.16036782e-01
-1.61424649e+00 -5.88205397e-01 -2.47228980e-01 2.44259715e+00
4.65757370e-01 2.53291786e-01 1.85065150e-01 4.42342639e-01
8.67669880e-01 -1.56726792e-01 -6.16040409e-01 -1.86786577e-01
2.19639167e-01 -5.46854399e-02 9.35288966e-02 3.81834984e-01
-6.00535572e-01 6.58861160e-01 6.53434229e+00 7.18782723e-01
-7.88047493e-01 -6.35948181e-01 5.54864824e-01 1.82264984e-01
-7.39945412e-01 1.81199029e-01 -6.79653108e-01 2.32460126e-01
3.39666642e-02 -6.83631673e-02 2.98068762e-01 7.68157303e-01
1.60422042e-01 -9.07667279e-02 -1.59352505e+00 1.33042479e+00
-9.03036669e-02 -1.27799046e+00 1.31652907e-01 2.34729186e-01
5.84967315e-01 -8.87173653e-01 -1.31592646e-01 -3.13991666e-01
1.23004876e-01 -1.00961816e+00 1.00864458e+00 4.37491924e-01
9.93635833e-01 -5.89629233e-01 9.12144184e-02 4.74421531e-01
-1.29376137e+00 9.82847586e-02 -2.94380691e-02 1.59819171e-01
5.26447713e-01 6.78227246e-01 -1.17506588e+00 6.13602579e-01
1.39510900e-01 6.21183395e-01 -2.21152857e-01 1.21258044e+00
-1.67408600e-01 7.60931373e-02 -4.97793287e-01 5.33411689e-02
-1.28050119e-01 -4.28374171e-01 7.09491670e-01 6.05181456e-01
6.65709436e-01 3.15079808e-01 1.57101408e-01 8.64217460e-01
-3.77933122e-02 -6.97853714e-02 -9.88258362e-01 1.35823190e-01
8.31343353e-01 8.21603060e-01 -1.08234584e+00 -3.46073285e-02
-4.59484249e-01 1.05005109e+00 1.65493518e-01 1.81025445e-01
-3.57549429e-01 -1.77575648e-02 4.42572862e-01 5.07153749e-01
6.95823193e-01 -4.22945976e-01 -4.49619919e-01 -7.39683568e-01
2.85297871e-01 -7.51393616e-01 -4.97912206e-02 -9.92879570e-01
-1.29360807e+00 4.71937031e-01 -7.13891983e-02 -1.63464546e+00
-2.30081677e-01 -7.64796734e-01 -6.32162809e-01 6.26825631e-01
-1.27434385e+00 -1.13344443e+00 -3.47442664e-02 3.45521450e-01
6.35088146e-01 2.28894591e-01 7.95013607e-01 3.93050537e-03
-3.47595841e-01 3.56634349e-01 8.43220670e-03 -7.53495470e-02
2.15877369e-01 -8.71896267e-01 6.71332121e-01 4.22032923e-01
1.84198156e-01 4.46456939e-01 7.54705846e-01 -8.25094223e-01
-1.85786498e+00 -6.43123090e-01 7.09890783e-01 -2.65580744e-01
4.35939461e-01 -4.86190230e-01 -8.35366726e-01 5.41484594e-01
-5.01821041e-01 -8.18487108e-02 2.77721733e-01 -2.41151806e-02
-3.37527961e-01 3.58658791e-01 -1.07033062e+00 1.00159168e+00
1.60130346e+00 -5.07209003e-01 -2.96768636e-01 2.12809324e-01
2.54980445e-01 -4.46093917e-01 -6.63261831e-01 1.90547064e-01
8.52246463e-01 -9.59206879e-01 1.08665359e+00 -5.05591691e-01
6.13582551e-01 -4.57713246e-01 -3.43295522e-02 -1.25884819e+00
-1.48015916e-01 -9.52161312e-01 -1.69110253e-01 1.01497483e+00
2.91531235e-01 -3.05772394e-01 1.09935987e+00 6.20408893e-01
-2.21037999e-01 -9.58290100e-01 -9.17689800e-01 -6.34064078e-01
-4.05470133e-01 -3.96747977e-01 9.85322297e-01 5.77827275e-01
-2.15203986e-01 2.19166994e-01 -3.83760720e-01 1.31654456e-01
7.19868839e-01 5.47305286e-01 1.06933284e+00 -1.62792146e+00
-4.67494130e-02 -5.15356839e-01 -4.64251041e-01 -1.57646120e+00
1.46771297e-02 -7.94414937e-01 1.05647065e-01 -1.17169964e+00
-1.80983737e-01 -8.61886442e-01 5.64370155e-01 9.64734554e-02
7.66569972e-02 -1.18755125e-01 9.87907574e-02 7.45655075e-02
5.35026491e-02 4.11549300e-01 1.45927000e+00 6.73995987e-02
-2.06468418e-01 1.97947040e-01 -6.29239380e-01 9.83917892e-01
3.14928621e-01 -3.27913910e-01 -2.79391468e-01 -5.90884745e-01
2.03976110e-01 5.81818342e-01 3.45527023e-01 -4.33379650e-01
1.80132866e-01 -1.49561897e-01 4.47121590e-01 -1.00600350e+00
4.38967258e-01 -1.32768548e+00 7.29456842e-01 2.70705193e-01
-1.53738722e-01 1.43606275e-01 -4.92639653e-02 5.64004958e-01
1.86764717e-01 -6.10911071e-01 5.46015441e-01 -2.72218108e-01
-4.17827964e-01 5.49055994e-01 -3.94893736e-02 -4.89506155e-01
1.22786510e+00 -8.94873679e-01 2.23509029e-01 -1.87417969e-01
-7.03662574e-01 2.15958238e-01 9.40065384e-01 4.26234752e-01
8.08166504e-01 -1.42272317e+00 -5.32288373e-01 5.53403676e-01
2.33634055e-01 3.66188437e-01 1.55354559e-01 3.75160575e-01
-5.44022322e-01 1.79380596e-01 7.30126584e-03 -7.09362686e-01
-1.22477496e+00 6.91626132e-01 2.75363952e-01 6.25485480e-02
-9.70683575e-01 6.30011380e-01 2.12747157e-01 -2.44480625e-01
-6.52559027e-02 -1.28657177e-01 9.58136376e-03 8.63217190e-02
4.04165387e-01 4.50518370e-01 2.06577897e-01 -6.12578690e-01
-2.37767413e-01 1.09359717e+00 1.90614462e-02 8.29198882e-02
1.39773846e+00 -3.89283560e-02 1.86795428e-01 3.49496484e-01
1.31236792e+00 1.81327105e-01 -1.29949999e+00 -2.52682537e-01
2.50343710e-01 -8.35140049e-01 -2.86929041e-01 -3.37560326e-01
-1.05494511e+00 8.90710056e-01 1.34116337e-01 2.46551871e-01
7.57228076e-01 1.95990920e-01 6.70950949e-01 2.53713746e-02
6.98631763e-01 -6.04370475e-01 1.15171587e-02 1.57315612e-01
1.12847483e+00 -8.01140070e-01 1.10312505e-02 -1.08136904e+00
-2.59726375e-01 1.66526568e+00 2.58838922e-01 -2.28139117e-01
7.66806364e-01 3.35029274e-01 -2.92643130e-01 -5.15012801e-01
-6.04998887e-01 1.01465471e-01 2.77777642e-01 5.61998069e-01
1.95968196e-01 -2.82327030e-02 4.60744500e-02 2.80570954e-01
-3.26502055e-01 1.61662593e-01 4.93990451e-01 1.11408615e+00
-2.66163111e-01 -1.33939159e+00 -4.86306041e-01 5.80595076e-01
-4.10569496e-02 3.15659136e-01 -4.25018698e-01 5.85759342e-01
-1.94539249e-01 6.01641655e-01 1.76114962e-01 -3.66717726e-01
7.02007949e-01 8.37063789e-02 6.68540001e-01 -7.36727238e-01
4.63948026e-03 1.61800399e-01 1.17962264e-01 -2.29391858e-01
-1.98510244e-01 -7.78012037e-01 -8.14856470e-01 -2.64262646e-01
-5.04620373e-01 -1.43623948e-01 4.53335434e-01 7.93843985e-01
6.13564730e-01 3.34252305e-02 9.94779885e-01 -1.22204340e+00
-2.99666405e-01 -3.66411209e-01 -6.44437373e-01 3.06865126e-01
4.61273283e-01 -1.02582836e+00 -2.18683764e-01 1.74993470e-01] | [8.686009407043457, -3.500701904296875] |
4285e33d-fb7d-4f3a-a2b8-bcee0894098d | frankmocap-a-monocular-3d-whole-body-pose | 2108.06428 | null | https://arxiv.org/abs/2108.06428v1 | https://arxiv.org/pdf/2108.06428v1.pdf | FrankMocap: A Monocular 3D Whole-Body Pose Estimation System via Regression and Integration | Most existing monocular 3D pose estimation approaches only focus on a single body part, neglecting the fact that the essential nuance of human motion is conveyed through a concert of subtle movements of face, hands, and body. In this paper, we present FrankMocap, a fast and accurate whole-body 3D pose estimation system that can produce 3D face, hands, and body simultaneously from in-the-wild monocular images. The core idea of FrankMocap is its modular design: We first run 3D pose regression methods for face, hands, and body independently, followed by composing the regression outputs via an integration module. The separate regression modules allow us to take full advantage of their state-of-the-art performances without compromising the original accuracy and reliability in practice. We develop three different integration modules that trade off between latency and accuracy. All of them are capable of providing simple yet effective solutions to unify the separate outputs into seamless whole-body pose estimation results. We quantitatively and qualitatively demonstrate that our modularized system outperforms both the optimization-based and end-to-end methods of estimating whole-body pose. | ['Hanbyul Joo', 'Takaaki Shiratori', 'Yu Rong'] | 2021-08-13 | null | null | null | null | ['3d-pose-estimation', '3d-human-reconstruction'] | ['computer-vision', 'computer-vision'] | [-4.24078226e-01 -6.06211834e-02 -3.72021049e-02 -1.88114196e-01
-5.67846179e-01 -5.52230895e-01 3.39546531e-01 -4.86825019e-01
-1.71989739e-01 2.78435767e-01 3.15810770e-01 2.91338801e-01
3.43626797e-01 -3.07468772e-01 -6.04020476e-01 -3.14557403e-01
-6.07225895e-02 6.03050590e-01 2.45803714e-01 -1.79041281e-01
-3.29284012e-01 4.69186991e-01 -1.54801965e+00 -2.26782441e-01
3.51627856e-01 1.00621045e+00 -4.19072121e-01 8.32241476e-01
4.46811557e-01 3.90102148e-01 -4.34657633e-01 -5.32069147e-01
4.20340359e-01 -3.80804718e-01 -4.48905051e-01 2.91413963e-01
9.17045474e-01 -5.88780999e-01 -4.04113114e-01 3.87155741e-01
1.02063727e+00 -2.43147045e-01 3.81710261e-01 -1.23606157e+00
2.14944705e-01 -5.29391393e-02 -8.48398030e-01 -3.49470079e-01
9.39540088e-01 2.12077707e-01 6.55345559e-01 -1.05761623e+00
7.21665084e-01 1.40598941e+00 9.59141195e-01 6.65221274e-01
-1.08045697e+00 -5.43927789e-01 1.76400155e-01 -4.14180517e-01
-1.65294278e+00 -7.93922901e-01 6.22062564e-01 -5.32658935e-01
7.74150848e-01 2.85576403e-01 9.83368099e-01 8.87981236e-01
3.52750421e-01 9.07568634e-01 7.24488139e-01 -2.16981977e-01
-2.26416364e-01 -2.41359189e-01 -2.31230006e-01 8.86581302e-01
2.27462739e-01 7.27727190e-02 -1.00736952e+00 -1.54561609e-01
8.61363888e-01 -2.08558962e-02 -2.20016286e-01 -1.00039923e+00
-1.28326285e+00 2.57809788e-01 3.01423848e-01 -1.78299174e-01
-1.19037680e-01 6.26398265e-01 2.56156325e-01 -1.77992671e-03
3.52747291e-01 -2.02937290e-01 -5.62971473e-01 -1.92498729e-01
-1.10919154e+00 5.87078691e-01 7.80402541e-01 1.06894279e+00
3.60752940e-01 -2.01806426e-01 -4.97718304e-02 4.64190573e-01
5.01276255e-01 7.32379913e-01 1.89681008e-01 -9.43682075e-01
3.53933454e-01 5.16841292e-01 2.45470956e-01 -9.05227184e-01
-7.29741573e-01 -4.68072832e-01 -3.82512420e-01 1.09067470e-01
3.92999053e-01 -4.20559466e-01 -6.72064006e-01 1.77984881e+00
1.11419261e+00 -3.77807826e-01 -4.72895116e-01 1.32232797e+00
8.41492474e-01 -6.98296120e-03 -1.92722917e-01 1.31256595e-01
1.41272211e+00 -1.13232970e+00 -5.52334607e-01 -5.01619577e-01
3.61113787e-01 -8.48091900e-01 7.94457734e-01 2.42391393e-01
-1.38687634e+00 -6.54505908e-01 -1.16185677e+00 -3.51666689e-01
9.34533551e-02 3.77286643e-01 5.67841470e-01 8.39414597e-01
-8.08290601e-01 5.34477353e-01 -1.14930093e+00 -4.85729605e-01
-1.10252365e-01 6.61609173e-01 -7.29295135e-01 3.22156519e-01
-6.58325374e-01 9.17083502e-01 -1.34683475e-01 2.46904522e-01
-4.86469328e-01 -4.98672426e-01 -1.09891498e+00 -3.13057005e-01
5.97255826e-01 -1.27879512e+00 1.53820443e+00 -5.61955154e-01
-1.80504084e+00 1.16198099e+00 -1.93510190e-01 6.30618781e-02
1.03657746e+00 -9.67675805e-01 9.03950781e-02 1.11303747e-01
-1.57178670e-01 9.11146760e-01 8.63187909e-01 -8.33834529e-01
-2.53278673e-01 -8.14474165e-01 -3.31014842e-01 4.79263574e-01
-2.85416394e-02 -1.06160492e-01 -1.22186911e+00 -5.60342371e-01
2.52300173e-01 -1.33811498e+00 2.93768179e-02 6.38765693e-01
-5.01789153e-01 2.83609420e-01 6.12900198e-01 -7.34935880e-01
1.12392795e+00 -2.02668786e+00 5.22684753e-01 1.11541189e-02
2.95870334e-01 -4.07920405e-03 1.90749019e-01 4.90263730e-01
1.60015538e-01 -4.48875546e-01 1.67111292e-01 -8.06679547e-01
9.14665759e-02 -1.51479155e-01 2.26699680e-01 9.04149592e-01
2.56847180e-02 1.04444170e+00 -6.93887532e-01 -6.09859109e-01
4.27734286e-01 6.85780287e-01 -5.60736954e-01 2.37608328e-01
8.45050439e-03 6.69652283e-01 -2.38296449e-01 1.02240825e+00
6.43885672e-01 -1.32067055e-01 4.29720849e-01 -2.42225081e-01
1.69231042e-01 1.46311194e-01 -1.44109821e+00 2.07545304e+00
-1.13744602e-01 2.37464473e-01 4.37769532e-01 -1.08565271e-01
6.51894748e-01 2.83096015e-01 6.39794528e-01 -1.70857936e-01
3.83185983e-01 1.99408695e-01 -2.64780372e-01 -1.05837852e-01
3.48483860e-01 -1.23273343e-01 -3.08664411e-01 3.29165816e-01
1.09590478e-01 -2.06353128e-01 -1.31925017e-01 -6.91522881e-02
6.35984063e-01 9.13511515e-01 6.27914667e-01 4.19598189e-04
4.04986292e-01 -2.37883687e-01 3.02227676e-01 1.56656370e-01
-2.65491784e-01 9.59032059e-01 2.93969780e-01 -2.38737926e-01
-7.78429568e-01 -1.24003196e+00 2.06384167e-01 1.05080342e+00
1.58819839e-01 -7.44268775e-01 -7.00141430e-01 -5.97649217e-01
4.98836160e-01 -2.61802733e-01 -6.57012165e-01 2.72628199e-02
-5.56226552e-01 -3.49426210e-01 4.63623315e-01 7.43628383e-01
3.27623904e-01 -3.13799113e-01 -1.03617513e+00 -2.96898168e-02
-2.94917405e-01 -1.30694306e+00 -8.38108480e-01 -2.06372485e-01
-9.41644192e-01 -1.11611509e+00 -9.45050061e-01 -3.54774058e-01
5.17061710e-01 3.18066686e-01 1.07952952e+00 2.11064741e-02
-4.66932297e-01 5.52332401e-01 -3.52446400e-02 -1.93270475e-01
1.11252882e-01 2.88135447e-02 3.77613127e-01 -1.01732656e-01
-5.05096419e-03 -6.10554934e-01 -9.33676481e-01 6.43518925e-01
-2.21155986e-01 1.63038760e-01 4.21682477e-01 4.60639745e-01
4.44038481e-01 -7.86790252e-01 -1.29849970e-01 -3.05777818e-01
1.48945916e-02 8.17810744e-02 -4.71674144e-01 -7.15758875e-02
-1.44733414e-01 -1.37081578e-01 8.43976885e-02 -3.79739046e-01
-6.41807079e-01 6.94949925e-01 -2.28189379e-01 -4.60568190e-01
1.79258704e-01 6.41077980e-02 -1.83061600e-01 -2.78103143e-01
5.28294623e-01 -3.07428418e-04 4.55797166e-01 -6.86821043e-01
6.19270980e-01 5.21045804e-01 8.14801157e-01 -4.79847670e-01
7.62533784e-01 6.76675737e-01 1.05593920e-01 -6.39737368e-01
-6.76041663e-01 -6.47183001e-01 -1.05468953e+00 -5.77097654e-01
6.27486289e-01 -1.25872421e+00 -1.11399543e+00 6.74220204e-01
-8.76044452e-01 -2.10154191e-01 -1.87488981e-02 4.40399349e-01
-7.18474567e-01 5.42033017e-01 -6.30268633e-01 -7.45339215e-01
-5.31143248e-01 -8.97811770e-01 1.74694943e+00 -1.09688677e-02
-6.77273035e-01 -4.58751082e-01 2.50645846e-01 3.81167442e-01
-1.32073276e-02 6.25312924e-01 6.80926964e-02 9.78208557e-02
-2.80641496e-01 -6.59371138e-01 6.06530868e-02 -2.71735117e-02
-1.41629446e-02 6.41970113e-02 -9.40100551e-01 -5.86514056e-01
-4.00076896e-01 -4.50866818e-01 3.34701210e-01 3.78017843e-01
5.43558478e-01 1.02809034e-02 -4.74985719e-01 7.36890793e-01
1.10259950e+00 -5.55582881e-01 3.42133939e-01 -6.90336972e-02
7.42249966e-01 7.31479645e-01 6.46094084e-01 5.32165110e-01
5.05953193e-01 1.25861812e+00 3.26029539e-01 -1.56121954e-01
-3.37722242e-01 -4.15831864e-01 6.75121307e-01 7.07158446e-01
-2.15926811e-01 2.28124022e-01 -7.02981234e-01 3.02588284e-01
-1.87368047e+00 -5.18609226e-01 6.67754263e-02 2.39233160e+00
7.15406954e-01 -7.37957954e-02 7.28292108e-01 1.07811233e-02
4.22772467e-01 6.37607202e-02 -4.56130594e-01 3.93598061e-03
3.50867927e-01 7.35842157e-03 4.47103888e-01 4.33222473e-01
-1.17025781e+00 8.60280633e-01 7.17734051e+00 3.05437803e-01
-1.12731564e+00 -9.55782980e-02 1.37896270e-01 -7.67278314e-01
3.60688627e-01 -3.28012109e-01 -9.75801110e-01 1.71891242e-01
3.90307546e-01 2.18137085e-01 1.01262145e-01 9.47481513e-01
7.22437054e-02 -2.16447696e-01 -1.35184431e+00 1.13394129e+00
1.53730363e-01 -7.25584090e-01 -3.05900276e-01 1.14111394e-01
4.41944838e-01 -1.70513749e-01 -1.86711818e-01 1.33592576e-01
-8.95354301e-02 -8.53250980e-01 1.17634106e+00 3.12073171e-01
9.53682661e-01 -4.77836818e-01 4.95016396e-01 5.43004394e-01
-1.35263515e+00 2.00546011e-01 8.46543014e-02 -3.16090375e-01
3.55793953e-01 4.33694243e-01 -4.74526763e-01 5.98776937e-01
6.29626155e-01 4.22056764e-01 -4.72941339e-01 8.58505130e-01
-2.72024661e-01 -4.90817055e-02 -5.57907820e-01 1.75566509e-01
-2.86335528e-01 1.11460268e-01 5.46677291e-01 1.11717224e+00
1.26387239e-01 -1.09230228e-01 2.01168090e-01 4.95422721e-01
4.04858440e-02 5.77240139e-02 -3.97558898e-01 2.39895687e-01
1.73024848e-01 1.28702414e+00 -5.61289072e-01 -3.86539921e-02
-4.02093828e-01 1.19172692e+00 2.56487012e-01 -1.60449132e-01
-1.03019762e+00 -3.87134440e-02 9.25616980e-01 4.75279897e-01
5.36484361e-01 -6.16500318e-01 -2.63910294e-01 -1.24892819e+00
4.54601824e-01 -9.48802829e-01 2.45351166e-01 -6.97355092e-01
-7.97016621e-01 3.01927865e-01 9.82845426e-02 -1.20809960e+00
-5.53149283e-01 -6.77602053e-01 -4.01389860e-02 6.28897369e-01
-9.27790344e-01 -1.45177543e+00 -6.81250334e-01 5.73842227e-01
2.57517904e-01 5.61936736e-01 7.91250944e-01 3.44552249e-01
-5.00645339e-01 7.94856608e-01 -4.53073531e-01 2.00291798e-01
1.02053809e+00 -9.95586276e-01 7.78900266e-01 6.77896321e-01
1.51231401e-02 6.79570735e-01 7.06214428e-01 -6.62980318e-01
-1.86602736e+00 -5.75358748e-01 8.03241193e-01 -9.83463466e-01
9.90162492e-02 -7.54783690e-01 -6.14098348e-02 7.57387757e-01
-3.19567114e-01 1.56170398e-01 5.42185009e-01 3.07150632e-01
-4.36678857e-01 -1.58235028e-01 -9.76127207e-01 7.98419893e-01
1.31490302e+00 -3.23289424e-01 -5.78318536e-01 5.27922325e-02
4.39847171e-01 -1.11726010e+00 -7.99718022e-01 5.87123156e-01
1.35165071e+00 -1.06060708e+00 1.34794378e+00 -3.31378609e-01
2.55992621e-01 -4.98802185e-01 -7.26075098e-02 -7.36424267e-01
-1.20562963e-01 -8.05339932e-01 -6.06650174e-01 7.77642965e-01
7.76975751e-02 -2.29721248e-01 1.03402495e+00 4.52207953e-01
3.00167412e-01 -8.89756680e-01 -9.02312994e-01 -6.57815337e-01
-3.49278271e-01 -3.02390039e-01 3.94871473e-01 4.31379706e-01
1.97458044e-01 4.10582840e-01 -8.33475471e-01 -2.11688620e-03
5.91732562e-01 3.15457046e-01 1.68844485e+00 -9.94981229e-01
-5.23079872e-01 -1.09344482e-01 -5.86952388e-01 -1.49536026e+00
-5.80304921e-01 -3.26835543e-01 9.31758881e-02 -1.14659381e+00
3.52639079e-01 1.51284814e-01 2.45947629e-01 4.90016222e-01
-1.73963100e-01 7.42255628e-01 4.35283214e-01 2.03457877e-01
-5.24717450e-01 4.04844075e-01 1.31758606e+00 3.43013883e-01
-2.49820650e-01 -4.27535910e-04 -5.03689706e-01 8.03157032e-01
1.96875095e-01 -3.08365077e-01 -1.93364710e-01 -3.96041751e-01
1.15890846e-01 2.76774108e-01 6.31344914e-01 -1.21493423e+00
1.35085717e-01 1.59081131e-01 8.05718660e-01 -7.37766981e-01
7.75472224e-01 -7.34518766e-01 4.80306298e-01 7.01627135e-01
3.18735093e-01 -1.28127476e-02 2.67387360e-01 3.73206645e-01
-1.86867323e-02 6.14674032e-01 7.82191455e-01 -1.32106915e-01
-3.17444086e-01 3.04398984e-01 -2.26823948e-02 -1.91909093e-02
9.97611761e-01 -5.24757683e-01 1.14535227e-01 -6.04179263e-01
-8.84199142e-01 2.07606822e-01 8.02996337e-01 6.44272506e-01
3.20536703e-01 -1.29628968e+00 -6.88253641e-01 3.04127783e-01
1.06808268e-01 -4.85193357e-02 9.86621976e-02 1.18699825e+00
-7.32918680e-01 4.88370925e-01 -2.34858379e-01 -9.37190235e-01
-1.70724452e+00 2.96671361e-01 5.62064409e-01 -1.62291273e-01
-5.94167948e-01 7.43998587e-01 1.07386492e-01 -7.43717253e-01
3.16384941e-01 -1.76796496e-01 4.30828691e-01 -6.83776885e-02
2.94796199e-01 5.05588889e-01 6.46540523e-02 -7.67775059e-01
-7.56858706e-01 1.18639219e+00 3.44449639e-01 -2.54139602e-01
9.61205721e-01 -3.00641596e-01 1.66259274e-01 2.62038499e-01
9.71144557e-01 5.10660291e-01 -1.31992888e+00 -5.24663106e-02
-5.76254070e-01 -6.47809505e-01 -2.32130200e-01 -7.78842807e-01
-9.61162210e-01 7.98256457e-01 4.57311630e-01 -5.70253253e-01
9.79797781e-01 4.91537750e-02 9.23358917e-01 3.68755907e-02
7.18361974e-01 -1.00519109e+00 9.51792747e-02 2.25016057e-01
8.99866164e-01 -8.68836105e-01 5.97827673e-01 -9.41381514e-01
-4.29738104e-01 9.46563482e-01 6.38022006e-01 -7.52892494e-02
4.95160788e-01 3.86929452e-01 2.01101273e-01 -2.27902308e-01
-3.75115573e-01 -2.35423058e-01 6.89585567e-01 3.49126279e-01
7.28018045e-01 1.30692963e-02 -3.16637576e-01 5.55031478e-01
-2.84288824e-01 6.23374060e-02 -2.56980896e-01 1.23382545e+00
-1.47120968e-01 -1.14711428e+00 -7.32575893e-01 -1.62623286e-01
-4.29524779e-01 4.31308627e-01 -8.21525812e-01 1.14235806e+00
3.67673524e-02 6.14031553e-01 -1.82862714e-01 -5.34297049e-01
5.83655179e-01 1.32613271e-01 1.02405274e+00 -4.28330004e-01
-6.57646418e-01 4.04592514e-01 1.47975013e-01 -1.15348053e+00
-2.23106489e-01 -6.01611435e-01 -9.78799224e-01 -6.26586258e-01
-4.09520447e-01 -4.38058794e-01 6.37303710e-01 7.35038936e-01
5.66507220e-01 1.46136716e-01 3.01116228e-01 -1.37729132e+00
-6.57360673e-01 -8.48470628e-01 -4.54461426e-01 2.12223187e-01
3.80155802e-01 -8.63652587e-01 -5.21940365e-02 -4.79904078e-02] | [7.050110816955566, -0.9125545024871826] |
6305a0c8-13cc-4b23-b4c5-19f2f61633e7 | deep-stock-predictions | 2006.04992 | null | https://arxiv.org/abs/2006.04992v1 | https://arxiv.org/pdf/2006.04992v1.pdf | Deep Stock Predictions | Forecasting stock prices can be interpreted as a time series prediction problem, for which Long Short Term Memory (LSTM) neural networks are often used due to their architecture specifically built to solve such problems. In this paper, we consider the design of a trading strategy that performs portfolio optimization using the LSTM stock price prediction for four different companies. We then customize the loss function used to train the LSTM to increase the profit earned. Moreover, we propose a data driven approach for optimal selection of window length and multi-step prediction length, and consider the addition of analyst calls as technical indicators to a multi-stack Bidirectional LSTM strengthened by the addition of Attention units. We find the LSTM model with the customized loss function to have an improved performance in the training bot over a regressive baseline such as ARIMA, while the addition of analyst call does improve the performance for certain datasets. | ['Akash Doshi', 'Somnath Rakshit', 'Sina Rafati', 'Puneet Sachdeva', 'Alexander Issa'] | 2020-06-08 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-1.94968387e-01 -8.80196914e-02 -9.68471915e-03 -4.94007826e-01
-4.05502707e-01 -4.09804970e-01 6.24345541e-01 -2.56576657e-01
-5.19113243e-01 5.04551351e-01 8.21665749e-02 -7.77173102e-01
-1.96483806e-01 -9.74778891e-01 -7.12562799e-01 -5.54758430e-01
-1.15707844e-01 4.12011266e-01 1.02752179e-01 -2.69046158e-01
5.10417163e-01 4.05257583e-01 -1.12778080e+00 3.79967242e-01
6.51810050e-01 1.66333091e+00 2.32643768e-01 4.15239245e-01
-5.50192237e-01 1.14319134e+00 -5.62403858e-01 -7.43600309e-01
6.35206521e-01 -5.98829910e-02 -4.63126332e-01 -3.51914376e-01
1.42562687e-01 -4.91026908e-01 7.97996446e-02 7.15388477e-01
3.50905865e-01 1.38702333e-01 1.74703583e-01 -8.67468596e-01
-5.10224223e-01 1.22612834e+00 -4.39605206e-01 4.93199497e-01
-4.07315910e-01 9.46844518e-02 1.33114195e+00 -6.54129446e-01
1.03081979e-01 9.25743043e-01 7.25256801e-01 2.16128036e-01
-1.23778212e+00 -7.15594590e-01 3.75903487e-01 1.61613464e-01
-5.46754479e-01 -1.25974163e-01 7.73066103e-01 -5.27977347e-01
1.33164001e+00 1.58155873e-01 5.99187374e-01 9.37650383e-01
5.77324867e-01 4.46078479e-01 9.60081339e-01 -4.21827346e-01
-1.88287254e-02 3.33238810e-01 2.96363086e-01 2.28026628e-01
1.43587500e-01 2.54695266e-01 -1.50115222e-01 -1.67811438e-01
7.82106340e-01 2.62067676e-01 2.28254020e-01 1.97522432e-01
-9.09144759e-01 1.07106948e+00 4.06383187e-01 6.11262441e-01
-7.24252462e-01 3.02168369e-01 4.35682178e-01 7.60711312e-01
8.64116430e-01 8.76669168e-01 -8.18569958e-01 -9.58137736e-02
-1.27219582e+00 3.39774072e-01 9.08705056e-01 2.25076124e-01
2.94063836e-01 6.48095131e-01 -4.17674422e-01 6.67580307e-01
1.44461483e-01 1.46634489e-01 8.87706757e-01 -7.05339968e-01
6.11598730e-01 5.50791502e-01 1.79756656e-01 -7.28885233e-01
-4.01254356e-01 -8.83880079e-01 -3.75461578e-01 4.67061609e-01
3.00832719e-01 -4.31799442e-01 -8.14922094e-01 1.38160753e+00
-3.90690356e-01 3.49510461e-01 6.43978789e-02 4.52718049e-01
1.52229279e-01 8.20830643e-01 2.65613738e-02 -1.44341856e-01
1.16317821e+00 -1.14015400e+00 -7.57681429e-01 -2.82756090e-01
7.44498670e-01 -5.97129285e-01 8.83249521e-01 3.58714968e-01
-1.20264912e+00 -5.23343742e-01 -9.73254204e-01 2.56255895e-01
-6.12040639e-01 2.09921934e-02 3.33170533e-01 4.47322756e-01
-8.16824794e-01 1.08351004e+00 -8.61159623e-01 3.16186458e-01
1.41377961e-02 4.67933655e-01 4.92159873e-01 8.63246679e-01
-1.36190200e+00 1.29392195e+00 6.78911567e-01 3.86350334e-01
-2.30848297e-01 -7.18799174e-01 -3.54023248e-01 4.85123158e-01
2.66056303e-02 -5.56982517e-01 1.30131090e+00 -1.33421826e+00
-1.67698026e+00 4.61643428e-01 3.51182789e-01 -1.44015098e+00
6.39737248e-01 -2.70977557e-01 -4.72409308e-01 -4.78426158e-01
-4.97645199e-01 3.36450487e-01 8.99738729e-01 -5.88950753e-01
-8.91180694e-01 -1.75925314e-01 7.18327984e-02 -2.15238184e-01
-5.17918169e-01 2.43292153e-01 3.86545062e-01 -1.18666732e+00
-2.76210755e-01 -8.69489610e-01 -3.64101410e-01 -7.94854760e-01
-1.09883681e-01 -1.13416739e-01 6.31140769e-01 -1.16424954e+00
1.54653382e+00 -1.92944777e+00 -2.30694368e-01 3.67920220e-01
-2.96579361e-01 2.30739355e-01 9.12627503e-02 1.13319166e-01
-4.13063645e-01 1.22991286e-01 -1.33838668e-01 -4.26420689e-01
2.10885346e-01 1.12176664e-01 -8.80550802e-01 -6.27029017e-02
2.52845973e-01 1.13070667e+00 -1.38928503e-01 9.74047333e-02
-4.93790209e-02 2.49045014e-01 -3.14659089e-01 2.31867224e-01
-5.21111012e-01 1.80851240e-02 -2.96005577e-01 1.94090098e-01
4.65130150e-01 -2.34138742e-01 -1.58769846e-01 1.44979030e-01
-5.10899842e-01 5.31821489e-01 -8.49167526e-01 9.30389345e-01
-5.26968956e-01 4.72225606e-01 -3.96127731e-01 -9.05598283e-01
1.27661812e+00 4.10227805e-01 3.15962911e-01 -8.88736188e-01
1.15280420e-01 5.58488011e-01 3.22650611e-01 -2.46097654e-01
4.71251339e-01 -4.55425292e-01 1.62195802e-01 5.54657936e-01
-3.42042446e-01 5.51535010e-01 -2.37414893e-02 -6.11202598e-01
7.99043179e-01 1.13403006e-03 -2.99395740e-01 -6.57272264e-02
5.75706601e-01 -2.54226089e-01 5.47821224e-01 5.16008437e-01
2.84915864e-01 2.63691574e-01 5.65821826e-01 -8.04144979e-01
-1.04032123e+00 -4.21497673e-01 -4.12083678e-02 1.21570837e+00
-8.89937162e-01 3.17069381e-01 -4.39739436e-01 -4.60082322e-01
1.79043084e-01 1.25413632e+00 -5.16748667e-01 4.52520046e-03
-9.07550871e-01 -8.67095411e-01 3.70247781e-01 6.13729179e-01
4.85353649e-01 -1.54895651e+00 -1.00042033e+00 6.05243027e-01
3.27840835e-01 -8.82310569e-01 -2.08567768e-01 5.17962396e-01
-1.09878349e+00 -4.98426646e-01 -9.80274796e-01 -5.48463285e-01
5.38426712e-02 -6.06857955e-01 9.70708907e-01 -2.82488286e-01
4.51695323e-01 -3.34133133e-02 -1.61654487e-01 -7.97974646e-01
-1.35587037e-01 4.89365816e-01 -3.41933399e-01 2.95259356e-01
3.07480752e-01 -4.38676715e-01 -4.32400256e-01 -7.25575760e-02
-8.46007109e-01 -1.59807548e-01 4.28430676e-01 7.77047873e-01
3.00474495e-01 -3.83426994e-02 6.78184688e-01 -8.24100256e-01
9.05615568e-01 -3.67771059e-01 -1.11927021e+00 4.05063659e-01
-1.25804925e+00 3.63768995e-01 5.26454568e-01 -4.12776113e-01
-1.17344582e+00 -4.03619766e-01 -1.80257082e-01 -5.44658720e-01
5.61647475e-01 9.94143844e-01 3.80708694e-01 -7.81978592e-02
-1.31581515e-01 2.09954023e-01 1.07983844e-02 -7.10812032e-01
-1.78080410e-01 1.57940701e-01 -4.13461998e-02 -8.68969783e-02
3.38818491e-01 2.29649879e-02 -1.32940650e-01 -1.18915237e-01
-7.98314691e-01 2.98184603e-01 -3.80069315e-01 -4.15960699e-03
7.98390508e-01 -5.89374065e-01 -6.24576926e-01 6.82336628e-01
-1.08225083e+00 -5.12587726e-01 -3.61482918e-01 5.10466278e-01
-5.63396096e-01 -3.45621735e-01 -8.46955776e-01 -9.60157275e-01
-7.63334990e-01 -9.98832881e-01 3.79287630e-01 -2.01688390e-02
-1.25284031e-01 -1.41195786e+00 6.29884377e-02 8.94337371e-02
9.38587248e-01 3.63282681e-01 1.01881158e+00 -1.29475069e+00
-5.06087601e-01 -4.18088377e-01 -2.31083273e-03 7.45400131e-01
-1.01430193e-01 -1.11057952e-01 -9.47600305e-01 -2.12145016e-01
4.52861607e-01 2.73962408e-01 1.17851806e+00 7.44379401e-01
1.17497408e+00 -6.56027555e-01 1.09951302e-01 7.40206420e-01
1.39883471e+00 7.26917088e-01 5.37487924e-01 1.03909516e+00
5.12073457e-01 7.59757459e-01 4.06500787e-01 4.26877171e-01
2.74496406e-01 4.98336315e-01 3.01987678e-01 1.79384416e-03
6.63319886e-01 -2.74886191e-02 7.49047399e-01 7.46350944e-01
-2.10939214e-01 5.56728244e-02 -8.21014404e-01 1.69173896e-01
-1.91498458e+00 -1.14267206e+00 1.19389363e-01 2.27831578e+00
5.04496872e-01 7.60011613e-01 3.93692136e-01 1.56137362e-01
5.29964387e-01 4.53743458e-01 -5.12295425e-01 -8.94953728e-01
-1.87402874e-01 2.24675521e-01 9.28872824e-01 3.79464000e-01
-9.47895288e-01 6.92298293e-01 6.40919304e+00 5.83986759e-01
-1.62074125e+00 -8.97934719e-04 1.06017184e+00 -4.21425790e-01
-4.29389983e-01 -1.31227687e-01 -1.01107633e+00 8.28407645e-01
1.70933020e+00 -2.97465473e-01 1.81052193e-01 6.38611913e-01
3.38838607e-01 3.59260947e-01 -9.93964136e-01 5.23342133e-01
-4.24347103e-01 -1.46296120e+00 1.34880751e-01 2.90930271e-01
4.77359772e-01 4.59323451e-02 4.99350131e-01 4.51767772e-01
3.36393714e-02 -8.94274235e-01 9.35261011e-01 9.83211517e-01
-1.37709871e-01 -9.56985474e-01 1.02187681e+00 2.50780761e-01
-8.83074462e-01 -6.42498672e-01 -1.77173734e-01 -1.68822214e-01
2.88281858e-01 4.99077171e-01 -6.32971287e-01 1.66816249e-01
7.19294429e-01 4.75800574e-01 -4.72114027e-01 9.53753650e-01
2.15087548e-01 7.72254348e-01 -3.01065296e-01 3.33242454e-02
7.19322860e-01 -5.06834745e-01 4.28612769e-01 9.23213899e-01
7.89509058e-01 -2.43014887e-01 -1.24663621e-01 1.09427106e+00
1.56738028e-01 6.39827130e-03 -4.48966712e-01 -3.08682084e-01
1.03527442e-01 8.56187463e-01 -4.18222815e-01 -3.66398275e-01
-5.53133249e-01 5.67425251e-01 7.14887157e-02 4.19678748e-01
-8.98851693e-01 -3.09373021e-01 4.22392100e-01 3.59058917e-01
7.43233085e-01 6.11962937e-02 -6.83971584e-01 -9.16086853e-01
2.79822201e-01 -4.93386894e-01 5.32569885e-01 -5.35027981e-01
-1.21615720e+00 8.06537867e-01 -2.74764150e-01 -1.14516604e+00
-6.39137805e-01 -7.43627429e-01 -1.13585424e+00 1.13164723e+00
-1.95334268e+00 -8.94565642e-01 4.56983656e-01 2.32708111e-01
4.45927441e-01 -5.05371809e-01 4.81666982e-01 3.62553030e-01
-7.26364315e-01 5.04873157e-01 2.25220695e-01 7.59240910e-02
3.77039373e-01 -1.31436312e+00 6.73749208e-01 6.96004331e-01
-1.72798276e-01 4.13200706e-01 7.25963771e-01 -5.78925133e-01
-7.61173308e-01 -1.10622776e+00 1.07330370e+00 -1.01206452e-01
1.08135867e+00 -1.04632862e-01 -1.18533814e+00 1.16750288e+00
3.17950666e-01 -4.04918551e-01 4.46306229e-01 -1.40030667e-01
-7.33021740e-03 -4.67165709e-01 -9.54563737e-01 2.89749086e-01
1.32239044e-01 -1.82648793e-01 -7.12817550e-01 -1.12918258e-01
9.06251788e-01 -1.45845503e-01 -1.07886696e+00 2.63251036e-01
6.39258265e-01 -1.05352473e+00 7.93005288e-01 -5.50655663e-01
3.85530442e-01 3.14369231e-01 5.32625876e-02 -1.27786040e+00
-4.62822944e-01 -4.96470898e-01 -2.11943820e-01 1.29845166e+00
1.00517488e+00 -1.34507382e+00 8.38869035e-01 1.09419191e+00
3.21650617e-02 -9.49397445e-01 -8.83578002e-01 -8.28619897e-01
2.86940336e-01 -1.76968157e-01 8.45730007e-01 7.52596021e-01
-3.33837599e-01 -4.08813022e-02 -5.41584134e-01 -1.79147959e-01
3.29022676e-01 4.69022959e-01 1.67760909e-01 -1.40770090e+00
-3.37380260e-01 -8.27008069e-01 9.87283289e-02 -6.16147339e-01
3.18919867e-01 -5.64956009e-01 -3.87364030e-01 -1.15227938e+00
-4.04505968e-01 -3.04423809e-01 -9.20676231e-01 4.27973002e-01
2.07055844e-02 -4.23837334e-01 4.72605944e-01 1.30379125e-01
5.90643510e-02 4.79596883e-01 7.75564909e-01 -5.38287573e-02
-4.92081791e-01 6.53563738e-01 -4.41354275e-01 6.22007251e-01
9.00804162e-01 -3.93183649e-01 2.74317581e-02 -5.54124236e-01
5.42243481e-01 2.95894802e-01 1.92048535e-01 -7.56287217e-01
1.66188225e-01 -3.96953411e-02 4.37887818e-01 -6.54632986e-01
3.38898718e-01 -9.25590873e-01 2.76777983e-01 8.77000093e-01
-6.50181770e-01 8.15081537e-01 3.20832580e-01 2.33816192e-01
-5.36724150e-01 -4.90120769e-01 5.82676888e-01 -2.51229942e-01
-4.90907788e-01 2.09222108e-01 -3.13714653e-01 -4.03163314e-01
8.69388819e-01 -2.81898409e-01 9.80870798e-02 -2.20903531e-01
-7.42594302e-01 3.68014395e-01 -6.68415725e-02 4.97782052e-01
3.10976356e-01 -1.27770782e+00 -7.27037191e-01 1.65260375e-01
-4.66780782e-01 -5.07278740e-01 -1.33787468e-01 8.33954215e-01
-4.41381067e-01 8.00236821e-01 -3.40003669e-01 6.52832538e-02
-9.77919996e-01 3.76175106e-01 6.52919114e-01 -8.06175590e-01
-5.20435572e-01 8.47054362e-01 -1.16661377e-01 -1.36683807e-01
3.79963517e-01 -7.71208525e-01 -4.77516770e-01 5.86936235e-01
4.54857349e-01 6.34804010e-01 3.20862651e-01 -2.03125566e-01
5.28137945e-02 3.88350517e-01 -1.46215782e-01 -1.70737207e-01
1.90182185e+00 -2.25963071e-03 -1.60000414e-01 8.80868435e-01
8.96997213e-01 -4.60421324e-01 -1.17611647e+00 -2.08196297e-01
9.45268989e-01 8.18801578e-04 3.47479522e-01 -7.62130737e-01
-1.48541462e+00 7.50965357e-01 6.33338571e-01 5.98173857e-01
1.03585005e+00 -6.35767519e-01 1.18747878e+00 3.25344414e-01
-8.58828351e-02 -1.08486784e+00 -3.57639134e-01 8.09281588e-01
8.42178702e-01 -1.05470765e+00 -3.63089591e-01 4.74173516e-01
-6.86695933e-01 1.42489290e+00 3.37779611e-01 -4.53586310e-01
9.84875917e-01 3.25701296e-01 6.25349134e-02 -1.27809867e-01
-1.18248427e+00 7.63806626e-02 3.49891186e-01 -2.70505577e-01
3.70998293e-01 2.78759245e-02 -2.72680908e-01 8.91478837e-01
-4.36357886e-01 4.73816618e-02 3.67471069e-01 6.38504028e-01
-3.67464274e-01 -1.09490442e+00 -2.65241712e-01 8.36047709e-01
-1.03513253e+00 -2.66605586e-01 -1.77642986e-01 5.01777411e-01
-3.90998349e-02 4.41649258e-01 5.56784451e-01 -2.00214431e-01
5.49169779e-01 6.33414686e-01 4.91846614e-02 -3.59401882e-01
-1.42324424e+00 1.80440009e-01 4.94482368e-02 -1.61476374e-01
-1.41624466e-01 -6.73116803e-01 -1.07142353e+00 -3.21123093e-01
-3.28745395e-01 1.07320718e-01 5.96095204e-01 1.00991154e+00
2.78556108e-01 7.81369209e-01 7.72303581e-01 -7.36653745e-01
-1.01671839e+00 -1.05626154e+00 -7.70385385e-01 1.35542024e-02
5.92830718e-01 -4.07393485e-01 -4.50561732e-01 -2.00362369e-01] | [4.468695640563965, 4.194979190826416] |
c7dcd470-c9a0-4c5b-b892-ac1b462cb0ba | leveraging-generative-ai-models-for-synthetic | 2305.05247 | null | https://arxiv.org/abs/2305.05247v1 | https://arxiv.org/pdf/2305.05247v1.pdf | Leveraging Generative AI Models for Synthetic Data Generation in Healthcare: Balancing Research and Privacy | The widespread adoption of electronic health records and digital healthcare data has created a demand for data-driven insights to enhance patient outcomes, diagnostics, and treatments. However, using real patient data presents privacy and regulatory challenges, including compliance with HIPAA and GDPR. Synthetic data generation, using generative AI models like GANs and VAEs offers a promising solution to balance valuable data access and patient privacy protection. In this paper, we examine generative AI models for creating realistic, anonymized patient data for research and training, explore synthetic data applications in healthcare, and discuss its benefits, challenges, and future research directions. Synthetic data has the potential to revolutionize healthcare by providing anonymized patient data while preserving privacy and enabling versatile applications. | ['Shashank Kumar', 'Aryan Jadon'] | 2023-05-09 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation'] | ['medical', 'miscellaneous'] | [ 2.78176814e-01 7.82362759e-01 -1.19477473e-01 -7.05038428e-01
-7.29704320e-01 -4.39133734e-01 1.73263550e-01 3.50393951e-01
-2.45830137e-02 1.05814362e+00 7.56863117e-01 -3.05797011e-01
-1.53106242e-01 -8.76850486e-01 -2.32000098e-01 -6.25656068e-01
1.24573961e-01 8.05690706e-01 -1.05827379e+00 1.62185326e-01
-5.26593566e-01 5.29148340e-01 -7.86374032e-01 4.88339126e-01
9.96660769e-01 7.43380904e-01 -8.18234503e-01 3.28778058e-01
5.26824668e-02 6.11925244e-01 -9.60802794e-01 -8.75184596e-01
6.15667164e-01 -6.73819840e-01 -4.63281453e-01 -1.31368473e-01
-5.47557846e-02 -4.72520083e-01 -3.97382289e-01 8.31095278e-01
9.82695460e-01 -5.62837064e-01 2.04316869e-01 -1.45759976e+00
-1.14594853e+00 7.26526737e-01 -1.82256788e-01 -4.74666506e-01
3.35460544e-01 5.29647648e-01 3.93570095e-01 -5.02467006e-02
9.59179699e-01 8.32883060e-01 8.86490285e-01 1.20855546e+00
-1.27500224e+00 -6.02529168e-01 -5.80146194e-01 -5.03850162e-01
-1.09872532e+00 -3.35236400e-01 4.87439275e-01 -4.17490691e-01
4.15237904e-01 7.87966251e-01 1.06971061e+00 1.50578701e+00
4.25035834e-01 8.37626159e-01 9.06876862e-01 1.54888988e-01
4.64891315e-01 1.63479462e-01 -7.55215958e-02 1.43937215e-01
8.00128818e-01 2.86631733e-01 -4.01754975e-01 -8.41441035e-01
5.30496597e-01 5.16728997e-01 -2.13629201e-01 -3.52674663e-01
-1.13254559e+00 7.89144754e-01 1.73321486e-01 -9.21171680e-02
-7.98651636e-01 -1.21944979e-01 5.69162905e-01 1.04821302e-01
1.80364817e-01 6.99782610e-01 -2.92379260e-01 -1.10587865e-01
-8.05504322e-01 5.45754135e-01 7.27440476e-01 1.36050940e+00
-1.83403399e-02 2.01643243e-01 -4.28266555e-01 3.48786891e-01
1.50299460e-01 4.72212076e-01 5.71874440e-01 -8.96517813e-01
2.95367211e-01 8.50606322e-01 4.71622199e-02 -7.65841067e-01
-4.14304495e-01 -3.11923683e-01 -1.37194848e+00 -1.97756410e-01
6.08742163e-02 -5.03637731e-01 -1.07933605e+00 1.52881134e+00
3.44862550e-01 -4.76875305e-01 5.72575748e-01 5.97123682e-01
7.55619764e-01 3.27189624e-01 3.44871998e-01 -3.46363872e-01
1.29465795e+00 -1.42701060e-01 -1.22224653e+00 5.68267377e-03
6.82630956e-01 -5.22034615e-02 5.78575373e-01 2.68757463e-01
-1.20609283e+00 1.34992883e-01 -5.34484327e-01 -1.78923488e-01
-2.75842518e-01 -4.28640574e-01 7.21681833e-01 1.17571807e+00
-8.13016653e-01 1.53823480e-01 -9.59413946e-01 -2.74109066e-01
1.37557316e+00 4.41245198e-01 -5.68618536e-01 -2.60341465e-01
-1.26110256e+00 2.90891021e-01 3.11479032e-01 -3.65680503e-03
-6.58977449e-01 -1.23691154e+00 -9.80097294e-01 7.34397769e-02
-6.72787651e-02 -1.47196758e+00 7.66952753e-01 -3.41293693e-01
-1.14299107e+00 8.99802148e-01 2.43177801e-01 -1.02966797e+00
1.07925391e+00 -1.24365233e-01 -6.16554439e-01 -4.21475559e-01
-1.48832947e-01 6.09420359e-01 1.90505907e-02 -9.39429879e-01
-1.52392522e-01 -8.54689002e-01 -9.11503673e-01 -1.80975229e-01
-6.05500974e-02 -1.76970884e-01 2.00279713e-01 -7.30068326e-01
-2.68156439e-01 -7.85284519e-01 -7.45285153e-01 -9.65291783e-02
-7.91753590e-01 5.63119829e-01 8.05509150e-01 -9.72952485e-01
1.10971558e+00 -2.09726906e+00 -2.40027666e-01 3.67681175e-01
5.29757559e-01 4.80168790e-01 2.80571103e-01 5.00415921e-01
1.68174598e-02 5.95748305e-01 -5.05957782e-01 -1.81114748e-01
-2.26340622e-01 2.02119887e-01 -3.20480555e-01 2.66627699e-01
1.02193303e-01 1.58382487e+00 -6.22395396e-01 -2.07473516e-01
1.92227483e-01 7.09116995e-01 -7.64060915e-01 4.45458502e-01
-7.94615075e-02 9.81838226e-01 -6.51498199e-01 9.62998152e-01
6.85487866e-01 -1.66346729e-01 4.40162808e-01 7.62288347e-02
3.31433594e-01 -1.10742293e-01 -4.81538773e-01 1.39409935e+00
1.68587327e-01 2.20538288e-01 -2.29857210e-02 -3.48027080e-01
1.04543459e+00 4.08372700e-01 1.01983011e+00 -7.71259189e-01
3.90677303e-01 -1.48526639e-01 3.97733180e-03 -5.38826883e-01
3.52912396e-01 -2.78737664e-01 -4.50324535e-01 3.79631370e-01
-4.90516722e-01 1.67568587e-02 -4.37551230e-01 1.06893688e-01
1.08638489e+00 -2.26956964e-01 6.30671442e-01 -3.72834727e-02
-1.98583782e-01 5.56536913e-02 8.58836949e-01 4.81532037e-01
-3.62908155e-01 7.11858749e-01 3.97368073e-01 -8.01022768e-01
-1.30600333e+00 -1.06606364e+00 -1.88574478e-01 -9.48352739e-03
-4.95299518e-01 -4.39165264e-01 -8.68896723e-01 -6.00740671e-01
4.35267955e-01 9.01393056e-01 -7.46425807e-01 -5.56324065e-01
-3.56576055e-01 -1.02956402e+00 8.62441361e-01 5.28521299e-01
2.44884148e-01 -1.06617844e+00 -8.49840283e-01 3.66780251e-01
-7.47387856e-02 -8.53383720e-01 -3.79255056e-01 -2.00057641e-01
-8.31836224e-01 -8.98438871e-01 -6.23665869e-01 -7.69519508e-02
6.78443611e-01 -6.14462733e-01 1.04532528e+00 -4.92027968e-01
-9.36386943e-01 1.80106372e-01 -4.12878692e-02 -9.81175840e-01
-8.48107278e-01 -8.94660726e-02 -6.68579489e-02 -1.34792298e-01
3.99203956e-01 -2.72333592e-01 -8.95302176e-01 1.91669110e-02
-1.16443169e+00 5.09308577e-02 4.08644080e-01 9.61643219e-01
9.05057728e-01 -3.64207864e-01 8.92978013e-01 -1.70454860e+00
9.70011294e-01 -6.80357397e-01 -3.39178979e-01 3.54517430e-01
-1.09072435e+00 -4.75904122e-02 6.91228986e-01 9.48996469e-02
-8.61430228e-01 1.27611309e-01 -2.23524839e-01 -4.03014809e-01
-2.71626949e-01 3.19541931e-01 -6.84992194e-01 3.79472166e-01
6.80251420e-01 4.82892990e-02 5.59627533e-01 -3.88896048e-01
5.72065055e-01 9.34885502e-01 6.29268885e-01 -1.48455754e-01
1.83052897e-01 5.10225713e-01 -2.13809356e-01 -3.19877714e-01
-3.04620743e-01 9.93236676e-02 -1.11646295e-01 3.38113457e-01
9.05460119e-01 -8.93441439e-01 -9.51154411e-01 3.92545670e-01
-6.32593334e-01 1.21619575e-01 -1.11530066e+00 2.09183916e-01
-6.31813228e-01 5.13498224e-02 -4.76221442e-01 -7.35548437e-01
-1.03249323e+00 -1.06763160e+00 8.61811340e-01 1.80740535e-01
-7.51006007e-01 -6.32936180e-01 1.34244561e-01 6.27159595e-01
5.93658924e-01 1.12027609e+00 8.37048829e-01 -8.89688194e-01
-6.19899571e-01 -6.59238696e-01 1.06547378e-01 1.26660734e-01
6.51910305e-01 -2.67621368e-01 -7.92411327e-01 -3.86363029e-01
1.32032067e-01 -1.26370341e-01 1.76478431e-01 5.17115355e-01
1.46511972e+00 -1.02594543e+00 -4.92345035e-01 1.04957318e+00
1.05205560e+00 5.75736344e-01 1.00227118e+00 -2.44068831e-01
7.21515536e-01 5.95393717e-01 2.62574136e-01 8.86225760e-01
4.78862047e-01 1.34867474e-01 2.47662887e-01 -3.65525514e-01
2.88759202e-01 -5.16692579e-01 -4.05067593e-01 3.98299545e-01
3.79399419e-01 -5.61174035e-01 -1.04694057e+00 5.67757547e-01
-1.74130082e+00 -8.37256253e-01 8.15887079e-02 2.21970367e+00
9.43830192e-01 -5.10605395e-01 2.99035162e-01 -2.03962594e-01
3.29331219e-01 -3.42356712e-01 -1.07599473e+00 -3.87406617e-01
-3.75662357e-01 2.01178148e-01 7.71267533e-01 -1.34382635e-01
-5.53235173e-01 5.06655335e-01 7.56621408e+00 -1.25887662e-01
-9.02196646e-01 -4.85570356e-02 1.33249104e+00 -4.06392813e-01
-8.75669837e-01 -3.96360874e-01 -3.82402509e-01 5.44812739e-01
1.09424078e+00 -7.69060314e-01 1.83246300e-01 8.53634953e-01
2.48772621e-01 5.83761394e-01 -1.16629899e+00 9.42211628e-01
-1.25585303e-01 -1.57330787e+00 4.39071268e-01 6.72612250e-01
8.18212032e-01 -2.39011362e-01 3.99609506e-01 -3.84538889e-01
7.92901278e-01 -1.40808320e+00 1.97657093e-01 8.15740049e-01
1.09788096e+00 -8.29309702e-01 9.66677547e-01 -5.06047606e-02
-1.20926261e-01 -1.11966863e-01 3.48780002e-03 5.28285265e-01
4.45469290e-01 5.97734988e-01 -1.26214826e+00 7.46044397e-01
5.66733062e-01 5.16905367e-01 -2.25504488e-01 8.60536575e-01
4.76927668e-01 3.63853306e-01 -1.30020693e-01 3.87372255e-01
-3.37239057e-01 -3.26146781e-01 2.86588669e-01 8.57247293e-01
4.02789205e-01 2.55599201e-01 -1.66715771e-01 1.10847330e+00
-2.71264285e-01 1.34454924e-03 -9.83002126e-01 -5.87995052e-01
4.97379124e-01 7.33313680e-01 -1.04361493e-02 -1.66677490e-01
1.11439750e-02 7.79134393e-01 -9.23647881e-02 1.69971362e-01
-5.47309101e-01 -2.84400154e-02 1.22403586e+00 4.12696391e-01
-4.45919812e-01 2.50367641e-01 -8.51973534e-01 -9.42774057e-01
-1.52616724e-01 -1.41652822e+00 8.68482530e-01 -4.54367548e-01
-1.31341112e+00 5.99733233e-01 -3.55040073e-01 -1.12230980e+00
-7.32584059e-01 -6.39375895e-02 -4.53137644e-02 9.82534051e-01
-7.47323453e-01 -1.37217081e+00 -1.93517014e-01 4.58959758e-01
-2.36872494e-01 -3.74449432e-01 1.26768470e+00 3.05503190e-01
-6.22921228e-01 1.08688986e+00 3.10707569e-01 2.39927933e-01
4.07618016e-01 -9.09988165e-01 9.33258891e-01 3.95931751e-01
-2.03470007e-01 9.27983999e-01 5.59113920e-01 -9.89155233e-01
-1.53993201e+00 -1.46150994e+00 4.90955532e-01 -7.10917413e-01
-1.51175722e-01 -3.65420520e-01 -9.15214777e-01 1.08853996e+00
8.49487558e-02 -4.62669842e-02 1.53737557e+00 -3.13602686e-01
-1.15538396e-01 -2.66645759e-01 -2.05737925e+00 5.15380979e-01
7.62912869e-01 -3.75318527e-01 -6.20058142e-02 1.72657251e-01
9.18006659e-01 -5.39111555e-01 -1.44347954e+00 5.24032056e-01
6.32564664e-01 -7.00304151e-01 9.15455282e-01 -1.23105800e+00
1.76014841e-01 -1.78655293e-02 1.85995847e-01 -1.29708457e+00
-1.29381344e-01 -1.16957140e+00 -6.33408055e-02 1.23834085e+00
5.09794593e-01 -8.93668532e-01 1.23180294e+00 1.76570082e+00
1.77605093e-01 -7.53183901e-01 -6.83991849e-01 -2.87642747e-01
-1.94167346e-03 -9.48549733e-02 1.51199031e+00 1.17584610e+00
7.43399262e-02 -3.63607585e-01 -8.01822662e-01 -3.70322019e-02
6.72812641e-01 -9.86546874e-02 9.10528541e-01 -9.96547520e-01
3.97537015e-02 -7.65184984e-02 -6.78210258e-01 -8.32450986e-02
-3.64249885e-01 -8.12484860e-01 -5.39739072e-01 -1.61208260e+00
4.64367978e-02 -5.44745505e-01 -1.23041242e-01 6.23912811e-01
-5.40505573e-02 2.75422558e-02 4.42284644e-02 -6.40639886e-02
3.91771793e-02 4.74622488e-01 1.11679518e+00 -1.59637794e-01
-3.37853909e-01 9.63945389e-02 -1.38575709e+00 3.17032114e-02
9.38814878e-01 -4.12747532e-01 -4.22631145e-01 -3.22487414e-01
1.62504107e-01 3.93399745e-01 1.91279218e-01 -6.41902685e-01
7.78502747e-02 -2.68855810e-01 5.90425491e-01 -3.66150618e-01
1.67289842e-02 -1.10556746e+00 1.31439757e+00 7.35359550e-01
-5.02223372e-01 1.16469800e-01 2.06994787e-01 5.45401931e-01
-1.48664325e-01 5.91312468e-01 5.58378100e-01 -1.43936917e-01
1.64165065e-01 7.22392321e-01 -9.02601406e-02 1.82627335e-01
1.39201820e+00 -1.97033927e-01 -4.12419826e-01 -3.23218256e-01
-1.01702654e+00 5.17497659e-01 6.18426800e-01 4.58114833e-01
7.92106211e-01 -1.27290916e+00 -9.45533395e-01 7.76172400e-01
3.95885974e-01 4.13333476e-01 5.48141241e-01 1.25097245e-01
-5.38215041e-01 2.04724088e-01 -4.35791343e-01 -2.49901190e-01
-1.14964342e+00 8.62813354e-01 2.70804405e-01 -8.92869607e-02
-8.00114274e-01 2.51168728e-01 1.58952415e-01 -4.58215803e-01
-3.20454612e-02 -1.83592692e-01 3.79559547e-01 -3.23339939e-01
6.60051465e-01 2.83213794e-01 -1.76124610e-02 -2.32679754e-01
-2.45335832e-01 -2.29221895e-01 -2.69028813e-01 2.66958207e-01
1.38448060e+00 -5.38474768e-02 -4.54067551e-02 -1.06605321e-01
9.59354579e-01 -4.24025506e-02 -9.08229828e-01 2.75749743e-01
-1.49540037e-01 -6.57455206e-01 -4.81683195e-01 -1.31148815e+00
-1.26082492e+00 4.98542905e-01 6.00144148e-01 1.86205626e-01
1.21604526e+00 -1.80423036e-02 1.00648999e+00 -2.21433282e-01
3.33067477e-01 -3.18581343e-01 -7.77573884e-01 -4.04876977e-01
7.69387126e-01 -1.00827503e+00 -2.37113368e-02 -3.41431558e-01
-1.12173045e+00 3.59906375e-01 3.94581020e-01 3.74625236e-01
5.18658757e-01 6.75407469e-01 4.38134879e-01 -1.57967851e-01
-6.45203769e-01 5.90324879e-01 -7.42125288e-02 1.07801712e+00
2.97929794e-01 6.25609398e-01 -3.08644205e-01 9.40754652e-01
-5.99446356e-01 5.61413169e-01 3.14839005e-01 8.92194629e-01
5.42587578e-01 -1.47151828e+00 -7.18000382e-02 9.85496759e-01
-7.57363856e-01 -8.53738189e-02 -8.16834450e-01 5.17573595e-01
-2.13330723e-02 3.79342645e-01 -1.30711263e-03 -2.62327731e-01
5.46537101e-01 1.46647647e-01 -1.21795386e-03 -2.73130596e-01
-1.00497746e+00 -2.99927294e-01 -8.64176452e-02 -5.90643287e-01
1.83691069e-01 -8.04302990e-01 -1.08141184e+00 -4.55961913e-01
2.22204223e-01 1.84089839e-01 7.17577159e-01 2.09398046e-01
1.44507492e+00 5.98734915e-01 4.06722873e-01 3.91299397e-01
-5.61556995e-01 -4.28454667e-01 -4.06861812e-01 6.95056438e-01
4.00043815e-01 3.95184845e-01 2.81082481e-01 6.09279498e-02] | [6.24810266494751, 6.754917144775391] |
e8d9c8ad-efa8-4561-8570-f1c551f91ef9 | data-efficient-goal-oriented-conversation | 1910.01302 | null | https://arxiv.org/abs/1910.01302v1 | https://arxiv.org/pdf/1910.01302v1.pdf | Data-Efficient Goal-Oriented Conversation with Dialogue Knowledge Transfer Networks | Goal-oriented dialogue systems are now being widely adopted in industry where it is of key importance to maintain a rapid prototyping cycle for new products and domains. Data-driven dialogue system development has to be adapted to meet this requirement --- therefore, reducing the amount of data and annotations necessary for training such systems is a central research problem. In this paper, we present the Dialogue Knowledge Transfer Network (DiKTNet), a state-of-the-art approach to goal-oriented dialogue generation which only uses a few example dialogues (i.e. few-shot learning), none of which has to be annotated. We achieve this by performing a 2-stage training. Firstly, we perform unsupervised dialogue representation pre-training on a large source of goal-oriented dialogues in multiple domains, the MetaLWOz corpus. Secondly, at the transfer stage, we train DiKTNet using this representation together with 2 other textual knowledge sources with different levels of generality: ELMo encoder and the main dataset's source domains. Our main dataset is the Stanford Multi-Domain dialogue corpus. We evaluate our model on it in terms of BLEU and Entity F1 scores, and show that our approach significantly and consistently improves upon a series of baseline models as well as over the previous state-of-the-art dialogue generation model, ZSDG. The improvement upon the latter --- up to 10% in Entity F1 and the average of 3% in BLEU score --- is achieved using only the equivalent of 10% of ZSDG's in-domain training data. | ['Arash Eshghi', 'Igor Shalyminov', 'Sungjin Lee', 'Oliver Lemon'] | 2019-10-03 | data-efficient-goal-oriented-conversation-1 | https://aclanthology.org/D19-1183 | https://aclanthology.org/D19-1183.pdf | ijcnlp-2019-11 | ['goal-oriented-dialogue-systems'] | ['natural-language-processing'] | [ 1.33768424e-01 1.00518644e+00 9.29601565e-02 -3.84770453e-01
-9.19191062e-01 -5.01174748e-01 9.35291529e-01 7.42373690e-02
-3.66328359e-01 1.26296580e+00 4.09797579e-01 -2.59300381e-01
1.06707819e-01 -9.07336593e-01 -1.71000779e-01 -3.05039167e-01
1.26417980e-01 1.13047349e+00 4.18750077e-01 -1.18064606e+00
2.47526661e-01 -1.97714105e-01 -1.07739377e+00 3.24281693e-01
9.81753349e-01 8.22455764e-01 7.57323876e-02 9.04733777e-01
-4.35152322e-01 1.01790631e+00 -9.91323650e-01 -7.57831097e-01
-5.15089557e-02 -9.90612507e-01 -1.51501918e+00 1.10571444e-01
-1.91832364e-01 -5.54143488e-01 -8.56543928e-02 7.13115633e-01
7.84474790e-01 3.17396224e-01 6.88480854e-01 -1.02256978e+00
-5.00806272e-01 8.11691880e-01 -1.85708813e-02 -2.40723670e-01
4.90843862e-01 2.46531233e-01 1.09470797e+00 -5.64014852e-01
9.34126675e-01 1.19703186e+00 3.81621271e-01 1.16637731e+00
-1.23758972e+00 -1.86166137e-01 -3.71272147e-01 -1.68520272e-01
-7.65333295e-01 -5.70946634e-01 7.47565210e-01 -3.83885264e-01
1.29856730e+00 -1.32858396e-01 3.53390336e-01 1.19621396e+00
-1.01656057e-01 8.16807628e-01 8.75211179e-01 -7.54815876e-01
2.33479232e-01 4.09137189e-01 6.17292114e-02 5.83004594e-01
-3.44109714e-01 -2.40845695e-01 -4.52452660e-01 -5.97384088e-02
7.43374765e-01 -9.14305866e-01 -5.00758104e-02 -2.41414681e-01
-1.07702291e+00 1.26691675e+00 -8.56003761e-02 6.11435890e-01
-1.37469292e-01 -1.85666323e-01 8.34330678e-01 7.92287767e-01
7.68536747e-01 8.76427770e-01 -5.40114105e-01 -6.62849545e-01
-4.28078443e-01 5.40922582e-01 1.56716335e+00 1.15543354e+00
4.66778904e-01 3.26993838e-02 -3.45060080e-01 1.36837387e+00
3.92055511e-02 -9.19699017e-03 7.34178364e-01 -8.18027258e-01
7.12736607e-01 6.80289388e-01 2.47166276e-01 -4.07808304e-01
-2.88014501e-01 2.37516671e-01 -7.28930175e-01 -2.25949902e-02
5.72089314e-01 -7.83612728e-01 -5.54268301e-01 1.73520470e+00
3.24222237e-01 -5.19097328e-01 6.28119528e-01 5.44847369e-01
1.00980389e+00 8.88139427e-01 7.05825463e-02 -3.16167325e-01
1.23155510e+00 -1.12808108e+00 -8.73805821e-01 -1.19046442e-01
8.61776054e-01 -6.27904773e-01 1.03446233e+00 2.38322571e-01
-1.23837280e+00 -5.45991063e-01 -9.08141732e-01 -5.11055104e-02
-5.14917314e-01 -8.16343427e-02 6.13804400e-01 7.62221515e-01
-1.14651513e+00 6.99855268e-01 -2.84110993e-01 -5.73797882e-01
-7.46718794e-02 4.33634400e-01 -3.18325728e-01 2.03478351e-01
-1.63962150e+00 1.22503865e+00 6.95280552e-01 -2.69835204e-01
-7.96229482e-01 -5.24288058e-01 -8.88356864e-01 -6.62499741e-02
4.89097416e-01 -5.43262541e-01 2.07094073e+00 -6.31750762e-01
-2.31773472e+00 7.72823274e-01 3.82898360e-01 -5.59344292e-01
5.78491211e-01 -1.58880934e-01 -3.76400113e-01 9.20484122e-03
9.03942361e-02 7.04135656e-01 3.97576571e-01 -1.01167548e+00
-8.66343439e-01 4.60689068e-02 4.68045145e-01 4.05643523e-01
-4.06175494e-01 9.58380103e-02 -2.40483105e-01 -2.56212652e-01
-6.76857591e-01 -8.85017931e-01 -3.91915232e-01 -6.12928092e-01
-5.11570990e-01 -6.29257560e-01 6.01460814e-01 -7.61009634e-01
1.20173621e+00 -1.75421786e+00 2.48417348e-01 -3.21466863e-01
8.34297538e-02 4.08271372e-01 -1.22140802e-01 9.12281871e-01
5.50116673e-02 -1.18095782e-02 -2.32344091e-01 -2.89455831e-01
2.90670753e-01 1.72911629e-01 -1.00630177e-02 -2.26419017e-01
5.57280481e-01 7.48924017e-01 -1.20451164e+00 -3.99738103e-01
3.03809494e-01 1.73723102e-02 -4.70088214e-01 7.55234897e-01
-7.14774728e-01 5.89221895e-01 -3.83238465e-01 -7.86010083e-03
4.40181345e-02 -1.28971562e-01 3.68150890e-01 9.41519439e-02
-1.76758468e-02 7.01804757e-01 -8.61605823e-01 2.06199002e+00
-7.00213432e-01 4.77600485e-01 -1.98480457e-01 -9.08282757e-01
1.19705582e+00 8.68221521e-01 4.18906927e-01 -6.00601196e-01
2.47184232e-01 3.60536098e-01 2.67232567e-01 -4.24549371e-01
8.64751756e-01 -2.96027750e-01 -5.96178055e-01 4.91017312e-01
7.29655206e-01 -3.76046419e-01 6.23750985e-01 1.21231236e-01
9.86138403e-01 8.16986710e-02 5.09429097e-01 -1.24686293e-01
6.85205400e-01 3.89437765e-01 1.97428659e-01 4.40467179e-01
3.01156156e-02 3.22681636e-01 8.28164160e-01 -2.11427391e-01
-1.19682562e+00 -5.21000922e-01 1.83709219e-01 1.31467330e+00
-1.17786810e-01 -3.32946390e-01 -1.09020126e+00 -9.63902354e-01
-3.99343014e-01 1.16066444e+00 -4.62147623e-01 -1.66227549e-01
-6.44840121e-01 -6.42719984e-01 7.23215818e-01 1.48540661e-01
8.38037550e-01 -1.37317753e+00 -4.38833296e-01 6.79004967e-01
-2.86869884e-01 -1.01234996e+00 -1.48240462e-01 1.82541177e-01
-6.30694091e-01 -9.18653786e-01 -8.81898105e-01 -7.03846514e-01
1.93293750e-01 -3.50872010e-01 1.42689872e+00 -3.84462535e-01
-8.96792114e-03 2.40932152e-01 -7.20526636e-01 -4.32156295e-01
-1.21638608e+00 5.77214122e-01 -4.02485251e-01 -3.81593734e-01
3.98335040e-01 -3.63672793e-01 -2.53251433e-01 1.96064889e-01
-7.14226127e-01 1.93829834e-01 3.86515945e-01 1.23996592e+00
-2.10217074e-01 -1.19946487e-01 1.11558032e+00 -1.30596018e+00
1.37881446e+00 -3.83502334e-01 -3.08186114e-01 1.54647872e-01
-4.77057457e-01 4.91223782e-02 6.49110198e-01 -1.69284418e-01
-1.49965882e+00 -1.73068345e-01 -5.76970398e-01 2.41788656e-01
-4.11303639e-01 5.13327956e-01 -2.83518761e-01 2.38157019e-01
8.29898894e-01 -9.37169418e-03 1.25437528e-01 -3.47392917e-01
6.86894059e-01 1.09629035e+00 3.11207414e-01 -6.40252650e-01
3.84534836e-01 -4.06041741e-01 -4.56323475e-01 -9.62450325e-01
-6.23192906e-01 -4.26617891e-01 -7.95842052e-01 -1.46821424e-01
9.08143580e-01 -6.30517304e-01 -4.30030346e-01 4.64340299e-01
-1.42374182e+00 -7.83590198e-01 -4.12166327e-01 8.64559412e-02
-8.98793221e-01 2.49707922e-01 -8.49407911e-01 -8.75502765e-01
-6.13353372e-01 -8.64283800e-01 8.55126977e-01 2.53105730e-01
-4.57903266e-01 -1.36561894e+00 4.02751178e-01 2.74291009e-01
4.10867006e-01 2.40210816e-01 1.00602841e+00 -1.31014299e+00
4.89729457e-02 -8.58498812e-02 2.60249944e-03 5.16046941e-01
2.72101969e-01 -4.51104313e-01 -8.44677091e-01 -1.36236012e-01
-5.90514652e-02 -9.65524495e-01 1.77884504e-01 -1.31035522e-01
4.07584250e-01 -3.16891491e-01 5.26801962e-03 -4.03739512e-01
1.14884806e+00 5.87156475e-01 6.33380890e-01 1.77082270e-01
3.64097029e-01 1.03528392e+00 1.06573427e+00 5.27615547e-01
5.01299441e-01 8.02199185e-01 6.60693571e-02 -5.26242815e-02
-5.19885719e-02 -1.67882159e-01 2.91883916e-01 1.05989683e+00
-1.41692355e-01 -3.63121301e-01 -8.86762738e-01 8.82717848e-01
-2.11887264e+00 -8.30403328e-01 1.13098770e-01 2.01885200e+00
1.50276625e+00 2.17124701e-01 5.35110593e-01 -9.73182917e-02
6.42782748e-01 1.39635056e-01 -3.47829819e-01 -8.41322243e-01
4.29599345e-01 3.72214586e-01 -1.31871551e-02 5.21548927e-01
-1.02153087e+00 1.23989820e+00 5.81794024e+00 7.42322505e-01
-8.79305422e-01 1.28534868e-01 6.14703834e-01 3.97174090e-01
8.70459303e-02 -2.04330996e-01 -9.16199505e-01 2.97599405e-01
1.45721090e+00 -5.35527647e-01 2.09197551e-01 1.08205688e+00
-1.11447023e-02 3.32491249e-02 -1.31844556e+00 5.22612989e-01
3.74963731e-02 -1.14917755e+00 -1.23427004e-01 6.61624819e-02
5.71014881e-01 -3.04127485e-01 -5.32955647e-01 9.07796800e-01
9.99682665e-01 -9.10842121e-01 2.49154225e-01 7.80732110e-02
8.15617323e-01 -8.09131861e-01 1.03678894e+00 5.28823137e-01
-7.98609495e-01 2.36835226e-01 -3.54437351e-01 4.62155379e-02
4.37729269e-01 2.79357880e-01 -1.49820650e+00 8.34520340e-01
1.89594015e-01 3.00532788e-01 -5.23399971e-02 6.70577228e-01
-2.45411560e-01 3.79641742e-01 9.09074992e-02 -3.34054172e-01
4.06178743e-01 -1.90044090e-01 2.51893252e-01 1.37455857e+00
1.22599356e-01 1.83274686e-01 2.45644063e-01 7.08954096e-01
-1.75083444e-01 2.60238737e-01 -8.02265346e-01 -2.02147439e-01
3.58698130e-01 1.24310708e+00 -2.86558777e-01 -3.43332708e-01
-4.04110312e-01 1.25419033e+00 2.50380903e-01 -1.43512934e-01
-5.84669292e-01 -9.33321059e-01 4.34570789e-01 -2.98409879e-01
1.33908674e-01 -7.58709237e-02 8.46520737e-02 -8.34215224e-01
-3.76133561e-01 -1.12009442e+00 1.69436678e-01 -3.62247020e-01
-1.43662107e+00 9.25188065e-01 6.13219216e-02 -1.07830453e+00
-1.20785177e+00 -5.16230166e-01 -4.50974226e-01 1.11384702e+00
-1.43408573e+00 -1.30977213e+00 -4.77741919e-02 4.88332510e-01
1.16712832e+00 -3.86117458e-01 1.36430633e+00 1.90835014e-01
-3.10979843e-01 5.07100642e-01 5.22827879e-02 4.36255991e-01
9.73853946e-01 -1.66549492e+00 6.79348409e-01 1.60810128e-01
-2.15638861e-01 3.40945214e-01 7.28094578e-01 -5.70104539e-01
-9.98973489e-01 -9.67721701e-01 1.17706788e+00 -4.20062423e-01
8.59476209e-01 -3.46158803e-01 -7.20267534e-01 4.92956638e-01
7.71426141e-01 -6.73555255e-01 8.06512296e-01 2.44307339e-01
1.27343953e-01 2.78753340e-01 -1.18977749e+00 5.86751223e-01
7.80859411e-01 -3.16565096e-01 -1.01951718e+00 3.80986005e-01
8.68584931e-01 -6.43698156e-01 -1.39404774e+00 1.58377632e-01
1.58897549e-01 -8.87422562e-01 5.92071354e-01 -7.24369645e-01
8.89346600e-01 2.00190529e-01 1.38022736e-01 -1.79954481e+00
3.33355069e-02 -9.76253450e-01 3.06687709e-02 1.67490232e+00
7.29255617e-01 -3.21634650e-01 7.25024223e-01 7.19492197e-01
-3.30664068e-01 -5.57210743e-01 -7.63259590e-01 -6.51654541e-01
3.16231400e-01 3.21755707e-02 3.62091780e-01 1.01304889e+00
6.74782813e-01 1.21360946e+00 -5.70135534e-01 -5.38159192e-01
1.48748428e-01 -9.99607071e-02 1.14306104e+00 -1.24374938e+00
-3.27825576e-01 -2.53218174e-01 -1.18040517e-02 -1.13060820e+00
1.52802289e-01 -4.94355470e-01 3.55244368e-01 -1.72259164e+00
-2.67721236e-01 -3.05375963e-01 3.52857351e-01 4.98322874e-01
-1.22787803e-02 -1.41861320e-01 9.22644511e-02 -5.99359535e-02
-5.86586237e-01 7.42401481e-01 1.31902945e+00 -3.35701890e-02
-5.52101612e-01 2.84262821e-02 -6.55632913e-01 4.10442770e-01
8.92767131e-01 -6.12186529e-02 -7.40137339e-01 -7.31946379e-02
-1.71444133e-01 6.03148699e-01 -2.36070886e-01 -7.37201333e-01
1.02248207e-01 -7.40333870e-02 -2.49804452e-01 -2.43844569e-01
4.33751076e-01 -4.12000567e-01 -3.02954614e-01 2.62370348e-01
-5.92773557e-01 -2.31614634e-01 2.22581431e-01 3.26544374e-01
-4.62326050e-01 -6.82138920e-01 6.60170257e-01 -4.56668437e-01
-9.02188241e-01 -1.88403234e-01 -5.33698320e-01 3.76985133e-01
9.50204968e-01 -1.39705678e-02 -4.69892234e-01 -7.31226981e-01
-5.88970244e-01 2.87292808e-01 1.40814602e-01 4.85460848e-01
2.90050924e-01 -1.11242187e+00 -7.58839965e-01 -2.48628318e-01
2.75900573e-01 9.44742784e-02 -3.82978506e-02 4.03960615e-01
-3.91024411e-01 6.33988857e-01 -4.47650045e-01 -1.97836056e-01
-1.09176445e+00 1.61180180e-02 2.53190994e-01 -9.71729517e-01
-4.99547362e-01 6.73758626e-01 -1.14971943e-01 -9.03111041e-01
5.06709963e-02 9.61189717e-02 -5.84764421e-01 1.70192972e-01
3.47764432e-01 2.48613909e-01 9.26778540e-02 -3.69728506e-01
2.19853401e-01 -7.60110095e-02 -3.82532507e-01 -4.87967700e-01
1.35426223e+00 -7.05230981e-02 -2.56877653e-02 6.42742634e-01
8.72146428e-01 -3.19477499e-01 -9.35953736e-01 -1.29155412e-01
3.33793342e-01 -1.35926589e-01 -4.01867896e-01 -1.09857571e+00
-4.41341043e-01 8.13239038e-01 2.27949083e-01 7.26027906e-01
6.66542947e-01 -4.46576886e-02 9.04979467e-01 6.42394841e-01
5.54421365e-01 -1.37889814e+00 3.89729142e-01 9.38826084e-01
8.74006927e-01 -1.26679826e+00 -4.34191316e-01 -4.96384650e-01
-1.15082788e+00 1.12077987e+00 8.55634689e-01 4.15285565e-02
2.16248646e-01 -4.33500409e-02 1.10710762e-01 -1.82150602e-01
-8.72406304e-01 -3.10061991e-01 -8.54472667e-02 8.27189028e-01
9.12819803e-01 -1.40096515e-01 -5.32903731e-01 6.68748140e-01
-3.58155668e-01 9.24917758e-02 6.34450495e-01 9.13470984e-01
-4.32998270e-01 -1.68744469e+00 1.09096549e-01 3.34956497e-01
-4.30093467e-01 -5.46512716e-02 -8.20770085e-01 1.11965549e+00
-3.97974193e-01 1.28545594e+00 -1.91206455e-01 -5.04333854e-01
6.60253108e-01 4.44008023e-01 3.15048367e-01 -1.14329338e+00
-9.18473125e-01 -1.54463872e-02 1.15825868e+00 -1.78422213e-01
-5.26536167e-01 -2.64118463e-01 -1.06435192e+00 -3.40982705e-01
-5.09148002e-01 6.05227053e-01 6.12194538e-01 8.75916839e-01
1.97808146e-01 6.98506117e-01 6.53349936e-01 -7.46273339e-01
-8.91604424e-01 -1.44380021e+00 -6.06553555e-01 3.83056730e-01
-1.30044326e-01 -4.61836547e-01 3.87651324e-02 7.50187486e-02] | [12.723152160644531, 7.965518951416016] |
c7cd3382-ff9b-40f8-bc22-d924c00f8917 | nvidia-unibz-submission-for-epic-kitchens-100 | 2206.10869 | null | https://arxiv.org/abs/2206.10869v1 | https://arxiv.org/pdf/2206.10869v1.pdf | NVIDIA-UNIBZ Submission for EPIC-KITCHENS-100 Action Anticipation Challenge 2022 | In this report, we describe the technical details of our submission for the EPIC-Kitchen-100 action anticipation challenge. Our modelings, the higher-order recurrent space-time transformer and the message-passing neural network with edge learning, are both recurrent-based architectures which observe only 2.5 seconds inference context to form the action anticipation prediction. By averaging the prediction scores from a set of models compiled with our proposed training pipeline, we achieved strong performance on the test set, which is 19.61% overall mean top-5 recall, recorded as second place on the public leaderboard. | ['Simon See', 'Ka-Chun Cheung', 'Cheng-Kuang Lee', 'Sze-Sen Poon', 'Yi-Kwan Wong', 'Giuseppe Fiameni', 'Oswald Lanz', 'Tsung-Ming Tai'] | 2022-06-22 | null | null | null | null | ['action-anticipation'] | ['computer-vision'] | [ 4.64442730e-01 4.09567326e-01 -3.95837665e-01 -8.84081796e-02
-1.14111710e+00 -2.16216192e-01 8.93486857e-01 2.48386189e-02
-3.11748534e-01 5.95592618e-01 8.10695112e-01 -5.62613487e-01
6.50970042e-02 -3.75550240e-01 -8.09928954e-01 -2.91724265e-01
-5.21494150e-01 4.50080872e-01 2.43893296e-01 -3.03864986e-01
2.80539453e-01 -2.35220283e-01 -1.44117141e+00 8.80272090e-01
2.42480576e-01 1.35120296e+00 -2.33205363e-01 1.30002236e+00
3.48721415e-01 1.80917454e+00 -3.32561702e-01 -1.57268971e-01
1.25000164e-01 -2.71338910e-01 -9.04531181e-01 -6.30810201e-01
9.03045163e-02 -3.31660569e-01 -8.75166535e-01 1.25761569e-01
5.92923760e-01 1.49332672e-01 1.68565996e-02 -1.34294188e+00
-1.90532088e-01 1.17107987e+00 -6.31768927e-02 5.18911362e-01
6.45367205e-01 4.67641264e-01 1.27085841e+00 -9.31053162e-01
8.21554601e-01 1.06992662e+00 1.02046609e+00 3.28415543e-01
-9.23422813e-01 -4.48141694e-01 3.66628170e-01 8.38613331e-01
-1.14383972e+00 -6.27998769e-01 4.30138022e-01 -1.75572485e-01
2.09295964e+00 2.99198538e-01 9.00241673e-01 1.63570964e+00
8.50938678e-01 1.38859534e+00 9.82241333e-01 -2.24933639e-01
2.21578076e-01 -5.79698324e-01 2.18928978e-01 3.92978758e-01
-5.65051138e-01 4.98554856e-01 -1.20286930e+00 -1.12066478e-01
-2.80085411e-02 -1.88342091e-02 2.50913203e-01 3.89460355e-01
-1.31826055e+00 3.28666240e-01 1.95161298e-01 -3.31575386e-02
-6.97090924e-01 1.04205656e+00 9.48360860e-01 1.99107260e-01
7.46662319e-01 7.36529455e-02 -6.62236273e-01 -1.03369987e+00
-5.08510351e-01 2.99448699e-01 7.79428780e-01 9.05244827e-01
1.38407901e-01 -9.49706286e-02 -8.17704618e-01 1.14514925e-01
2.24864021e-01 1.85242742e-01 5.64244270e-01 -1.11365271e+00
5.01521349e-01 4.18594807e-01 2.23186567e-01 -5.79826057e-01
-6.24327123e-01 -6.07583404e-01 -5.32195091e-01 -1.84109747e-01
7.26505071e-02 -4.96730596e-01 -7.73303926e-01 1.58142316e+00
1.29626423e-01 1.01434422e+00 2.97645181e-01 3.97537947e-01
6.50595367e-01 6.27879500e-01 2.63258308e-01 -1.20203800e-01
1.11922383e+00 -1.56655109e+00 -6.96371555e-01 -5.52351534e-01
1.14463460e+00 -6.99002683e-01 8.43519211e-01 5.32325268e-01
-1.14614284e+00 -4.78783727e-01 -1.08725989e+00 -1.58631951e-01
-7.73212165e-02 6.10382520e-02 8.70526195e-01 -5.94639871e-03
-1.24254632e+00 8.76812994e-01 -1.15417147e+00 -2.38076732e-01
3.00740227e-02 6.29074723e-02 1.17116719e-01 2.16623291e-01
-1.49743998e+00 1.03530025e+00 1.31617025e-01 2.81991929e-01
-1.41657424e+00 -8.75739396e-01 -3.07098716e-01 4.98747677e-02
4.14441347e-01 -6.35876417e-01 1.97488177e+00 -3.81511003e-01
-1.86179984e+00 4.71704900e-01 -2.76650548e-01 -1.32247949e+00
7.73483276e-01 -9.25547600e-01 -7.11579621e-01 -3.38325053e-01
-1.63455486e-01 4.27783310e-01 3.83033991e-01 -8.23927000e-02
-1.03834140e+00 4.51911576e-02 -6.63824007e-02 1.57983348e-01
3.30004126e-01 1.86671898e-01 -1.25488982e-01 -2.20271945e-01
-2.48290569e-01 -1.13011658e+00 -5.80079317e-01 -8.21586549e-01
-3.84363890e-01 -5.15715957e-01 6.86176598e-01 -7.09522009e-01
1.44811952e+00 -2.11580634e+00 -1.44845963e-01 -1.78401455e-01
7.25740790e-02 -1.38156302e-02 -3.41451883e-01 1.03445637e+00
-3.80372733e-01 -2.02722430e-01 2.87587523e-01 -6.02956951e-01
2.18582034e-01 -7.03772604e-02 -8.41104269e-01 1.45134985e-01
9.22980085e-02 1.16817570e+00 -1.13584268e+00 -2.01946601e-01
1.45995542e-01 3.18311274e-01 -3.47114861e-01 1.66148916e-01
-6.05595112e-01 1.79176629e-01 -3.19826871e-01 4.55403894e-01
-2.28565529e-01 -2.94308007e-01 3.62868071e-01 2.71112412e-01
-2.55972743e-01 1.03169668e+00 -5.68182170e-01 1.99068284e+00
-5.40714204e-01 9.52234089e-01 -7.80993879e-01 -3.97108585e-01
5.15626013e-01 5.87037265e-01 6.08761132e-01 -9.97581065e-01
-9.17850658e-02 -5.19319288e-02 -1.11345820e-01 -3.38334650e-01
7.40505099e-01 4.27464128e-01 -4.54773158e-01 7.00886011e-01
-2.31208712e-01 4.71403033e-01 7.04000667e-02 4.34726059e-01
1.98865008e+00 3.99228543e-01 3.92270349e-02 2.43705079e-01
4.59848903e-02 -2.03452017e-02 7.43105948e-01 8.85443866e-01
-2.66864359e-01 2.45961681e-01 7.05293775e-01 -8.41414392e-01
-8.57352316e-01 -8.34410608e-01 5.84392428e-01 1.43003988e+00
-3.70580286e-01 -1.26583552e+00 -3.08495045e-01 -5.09944022e-01
-3.28965306e-01 1.19961476e+00 -5.49791098e-01 -3.95157695e-01
-6.79205358e-01 -4.29871738e-01 8.16466630e-01 9.46291447e-01
2.31128529e-01 -1.24401605e+00 -1.01220965e+00 6.26568317e-01
-4.77124691e-01 -1.10037982e+00 -3.21844488e-01 2.47088686e-01
-7.21719682e-01 -8.94593954e-01 1.02625176e-01 -1.41563416e-01
-1.96899027e-01 -2.70215690e-01 1.49077916e+00 -9.37720537e-02
-5.83511628e-02 3.25146198e-01 -3.38930398e-01 -4.66476291e-01
-3.17028761e-01 1.85134724e-01 -2.86090747e-02 -3.29028666e-01
3.83311898e-01 -8.54310334e-01 -7.31336117e-01 3.83411236e-02
-3.58717114e-01 3.54574561e-01 5.61749637e-01 7.44641840e-01
4.74190086e-01 -3.76625299e-01 3.17561507e-01 -6.68790698e-01
4.63228285e-01 -4.70303565e-01 -4.14535701e-01 2.15352923e-01
-6.15367949e-01 -1.75673317e-03 2.92636126e-01 -1.69002384e-01
-9.71318007e-01 3.87262911e-01 -2.33800113e-01 -1.61151215e-01
1.62217468e-01 6.31063223e-01 5.36209762e-01 5.39399862e-01
6.23400807e-01 3.43426734e-01 -4.14956033e-01 -2.74677396e-01
5.97949207e-01 3.38713169e-01 7.24593103e-01 -2.67217726e-01
4.78169397e-02 2.18443185e-01 -6.99350312e-02 -1.46254301e-01
-1.07373834e+00 -2.09680930e-01 -1.55413643e-01 -3.65896821e-01
6.99279428e-01 -1.28356838e+00 -1.18148351e+00 6.44763827e-01
-1.31634068e+00 -9.46753263e-01 -3.87388796e-01 4.15504873e-01
-9.58281219e-01 -1.02860317e-01 -9.10404027e-01 -1.10339606e+00
-7.88190901e-01 -7.36489058e-01 1.16087317e+00 -7.84528404e-02
-7.12666333e-01 -7.82209933e-01 5.37191331e-01 9.69223902e-02
4.40148592e-01 2.40233719e-01 3.72896761e-01 -7.01361895e-01
-8.83068860e-01 -2.65101552e-01 2.37890333e-01 -3.50468382e-02
-5.59488237e-01 6.77229539e-02 -1.12762523e+00 -1.19682133e-01
-3.21973920e-01 -5.17719507e-01 1.08832729e+00 3.47515672e-01
1.11684537e+00 -2.38475606e-01 -5.70338845e-01 1.51559487e-01
9.92364645e-01 1.67238951e-01 8.50970984e-01 2.04617992e-01
3.00702125e-01 -4.85657081e-02 9.75084603e-01 9.39987838e-01
5.99961817e-01 6.25301719e-01 5.18233895e-01 2.88644165e-01
-1.68348700e-02 -8.03165376e-01 9.98050988e-01 8.38522434e-01
-7.86465183e-02 -3.05819064e-01 -1.01936936e+00 5.51896513e-01
-2.38544607e+00 -1.13205385e+00 -1.73988760e-01 1.86316979e+00
6.34236217e-01 5.47785461e-01 1.07574858e-01 1.14875004e-01
2.92482346e-01 4.46882993e-01 -5.53719640e-01 -6.75594389e-01
1.80407003e-01 1.78673685e-01 5.81877351e-01 7.14874148e-01
-1.02864408e+00 1.06657612e+00 8.32558537e+00 8.54186773e-01
-8.68438482e-01 1.71842381e-01 5.67044199e-01 -5.06390452e-01
-4.76163402e-02 5.09628654e-01 -9.34962749e-01 5.04236698e-01
2.00857639e+00 -2.78162241e-01 2.84003794e-01 7.66667008e-01
2.92874157e-01 -2.03812510e-01 -1.41161156e+00 6.00363612e-01
-3.78108829e-01 -1.59423959e+00 -6.02010250e-01 -3.29644442e-01
6.27275050e-01 7.49169111e-01 1.61076754e-01 8.09383452e-01
9.82204318e-01 -1.01843691e+00 8.14708114e-01 1.04954779e+00
4.43991899e-01 -5.56837440e-01 5.54460049e-01 3.12805235e-01
-1.07980931e+00 -2.21123606e-01 1.21939652e-01 -8.10679018e-01
5.76807976e-01 7.44992495e-01 -1.07060051e+00 4.43647712e-01
5.30979753e-01 1.15195858e+00 -3.68346542e-01 8.15347791e-01
-4.76399958e-01 9.79790151e-01 -4.04295743e-01 -1.07309885e-01
3.27956080e-01 4.88028854e-01 5.95767677e-01 1.03466010e+00
1.57636017e-01 6.37669233e-04 1.52323827e-01 1.52315453e-01
-1.18029587e-01 -5.53713083e-01 -5.10204732e-01 -2.64542818e-01
4.64662284e-01 1.12513566e+00 -1.48137614e-01 -6.56588316e-01
-6.95545226e-02 8.76065195e-01 3.40879828e-01 2.14928880e-01
-1.40552711e+00 -1.13771074e-01 7.17179418e-01 -3.67986262e-01
2.12208420e-01 -1.97193906e-01 -1.90005928e-01 -7.14003861e-01
-1.28878094e-02 -7.26591647e-01 6.72216892e-01 -9.49765921e-01
-7.87213922e-01 5.08040488e-01 -2.51633495e-01 -1.07786334e+00
-8.66501629e-01 -1.26778230e-01 -9.89627659e-01 4.57393438e-01
-1.10774517e+00 -9.70523655e-01 6.57469407e-02 1.40836537e-01
6.22772872e-01 -2.09369790e-02 9.81137693e-01 3.86978537e-01
-6.01360500e-01 5.15595436e-01 9.12091881e-03 -3.45757723e-01
4.88742679e-01 -9.75187480e-01 1.20235872e+00 7.45692968e-01
1.16467439e-01 1.67440176e-01 1.05196583e+00 -5.57812631e-01
-1.59534621e+00 -1.20871496e+00 1.55462551e+00 -8.25437009e-01
1.03212237e+00 -3.82107705e-01 -3.15503448e-01 1.55516624e+00
5.30482650e-01 -1.82020992e-01 4.57774132e-01 5.13230920e-01
-1.19887747e-01 -1.98857430e-02 -4.36603755e-01 7.02753007e-01
1.39251256e+00 -6.45519733e-01 -5.33476949e-01 7.59065092e-01
1.10271919e+00 -8.11871588e-01 -9.29530501e-01 3.54958773e-01
7.11666524e-01 -1.00429988e+00 7.58783996e-01 -1.04446638e+00
8.65522385e-01 -5.52580915e-02 -6.95153251e-02 -1.08938181e+00
-3.61656576e-01 -1.14470375e+00 -5.99692881e-01 7.70863414e-01
6.79286003e-01 -5.77945054e-01 8.45943868e-01 3.98417503e-01
-4.83268976e-01 -1.00603795e+00 -1.00106812e+00 -6.39595330e-01
-5.79790115e-01 -1.02986526e+00 3.13804656e-01 3.75249058e-01
3.72449785e-01 6.41009152e-01 -7.89769113e-01 -2.51821578e-01
1.48236260e-01 4.62050885e-02 8.21850061e-01 -6.01383686e-01
-6.05967045e-01 -1.46496266e-01 -2.14068204e-01 -1.21510971e+00
4.75038812e-02 -7.20780671e-01 8.95586014e-02 -1.37411523e+00
5.78245185e-02 8.31408950e-04 -7.16063142e-01 7.63847530e-01
1.78046659e-01 -2.68661492e-02 1.78369313e-01 -1.65908575e-01
-1.34339404e+00 8.54705513e-01 5.24496138e-01 -3.82534601e-02
-7.22788051e-02 6.88553462e-03 -4.52417850e-01 6.50286973e-01
7.98166394e-01 -4.92932051e-01 -3.58464152e-01 -5.37086964e-01
6.65419281e-01 4.33883697e-01 3.37486744e-01 -1.19391346e+00
4.99619454e-01 -2.84846369e-02 2.76656076e-02 -1.13414598e+00
5.86543679e-01 -2.50994653e-01 5.29691458e-01 8.25723708e-01
-8.17031622e-01 4.93867785e-01 2.71050125e-01 7.16594994e-01
1.30311683e-01 5.70013106e-01 -5.14268950e-02 2.22938091e-01
-1.05372930e+00 4.91658300e-02 -7.26184666e-01 1.00679711e-01
9.95250523e-01 2.14768276e-01 -5.16904533e-01 -6.30093038e-01
-9.32320476e-01 5.39520323e-01 -2.14725181e-01 6.72143340e-01
4.94515240e-01 -1.31052649e+00 -8.29973757e-01 -3.60367209e-01
1.12945288e-01 -4.63234395e-01 4.69306380e-01 1.29700768e+00
-2.14121472e-02 7.41042018e-01 -5.65917939e-02 -5.04677951e-01
-1.15648580e+00 3.62879753e-01 2.30949968e-01 -8.79149318e-01
-9.76599514e-01 8.75066578e-01 -5.35442591e-01 -9.11296383e-02
3.44977021e-01 -5.03848553e-01 1.96294233e-01 -2.67348170e-01
7.09739625e-01 7.71231890e-01 5.11946902e-02 -2.40561087e-02
-6.46295011e-01 -1.42443806e-01 -1.11041442e-01 -2.42647395e-01
1.24101865e+00 1.57524899e-01 2.54218187e-03 7.44903386e-01
9.81336176e-01 -3.90649498e-01 -1.20272326e+00 -2.63929456e-01
4.50167120e-01 -3.75895910e-02 8.74290913e-02 -1.05455339e+00
-5.50352454e-01 6.15539908e-01 6.61896050e-01 4.49648649e-02
9.33725893e-01 -1.69280857e-01 1.12561870e+00 5.15349090e-01
2.96018988e-01 -1.48611188e+00 1.06673792e-03 1.07024789e+00
9.01258647e-01 -7.60590255e-01 -1.29952550e-01 2.48258300e-02
-6.43987477e-01 5.20975232e-01 5.18372297e-01 -3.05764765e-01
3.79630536e-01 3.12987626e-01 -1.01481542e-01 -1.20914727e-01
-2.21003437e+00 1.01819128e-01 -1.30281657e-01 1.35480076e-01
6.41416550e-01 2.67821819e-01 -3.79562438e-01 6.02155566e-01
-2.36272201e-01 4.43013877e-01 3.94161075e-01 1.01941514e+00
-1.94282427e-01 -9.42501128e-01 2.75810093e-01 3.33463609e-01
-4.23168719e-01 -3.39497238e-01 -4.67791915e-01 3.59392762e-01
-4.14620191e-01 1.01626706e+00 2.01410085e-01 -9.61878955e-01
4.28615540e-01 3.39463830e-01 2.84613550e-01 -8.00798684e-02
-9.22759056e-01 -4.61515188e-01 7.92551041e-01 -1.43677056e+00
-2.30804291e-02 -5.62602282e-01 -1.36564553e+00 -6.03704035e-01
1.83584824e-01 2.01319959e-02 4.27432656e-01 9.18909729e-01
9.76323664e-01 9.91286457e-01 5.51982164e-01 -6.62043035e-01
-7.09320009e-01 -1.16887033e+00 -8.13175142e-02 -2.02003852e-01
7.70309716e-02 -3.19261968e-01 -1.87228799e-01 -3.06182027e-01] | [8.075424194335938, 0.5675957798957825] |
15f4b808-930f-45a6-a76a-9341268541e1 | illinois-cross-lingual-wikifier-grounding | null | null | https://aclanthology.org/C16-2031 | https://aclanthology.org/C16-2031.pdf | Illinois Cross-Lingual Wikifier: Grounding Entities in Many Languages to the English Wikipedia | We release a cross-lingual wikification system for all languages in Wikipedia. Given a piece of text in any supported language, the system identifies names of people, locations, organizations, and grounds these names to the corresponding English Wikipedia entries. The system is based on two components: a cross-lingual named entity recognition (NER) model and a cross-lingual mention grounding model. The cross-lingual NER model is a language-independent model which can extract named entity mentions in the text of any language in Wikipedia. The extracted mentions are then grounded to the English Wikipedia using the cross-lingual mention grounding model. The only resources required to train the proposed system are the multilingual Wikipedia dump and existing training data for English NER. The system is online at \url{http://cogcomp.cs.illinois.edu/page/demo_view/xl_wikifier} | ['Chen-Tse Tsai', 'Dan Roth'] | 2016-12-01 | illinois-cross-lingual-wikifier-grounding-1 | https://aclanthology.org/C16-2031 | https://aclanthology.org/C16-2031.pdf | coling-2016-12 | ['cross-lingual-ner'] | ['natural-language-processing'] | [-7.88317382e-01 2.79441297e-01 -4.49871153e-01 -2.86474556e-01
-1.02324855e+00 -8.03928137e-01 6.30965531e-01 4.15632367e-01
-8.41028452e-01 1.12617254e+00 5.64033210e-01 -2.57555932e-01
2.35353500e-01 -1.09678185e+00 -6.74821019e-01 -1.13834878e-02
1.16541728e-01 5.29767394e-01 2.68119156e-01 -4.09485459e-01
2.21622244e-01 1.87503144e-01 -1.04407799e+00 2.42882133e-01
1.22415984e+00 4.03868824e-01 4.37360018e-01 3.87068003e-01
-9.94331956e-01 7.98293173e-01 -4.36369181e-01 -7.88854182e-01
4.99949567e-02 -9.18327719e-02 -1.14175189e+00 -4.04140323e-01
2.32552826e-01 3.27962548e-01 -3.48664552e-01 1.32845140e+00
3.18419188e-01 3.72708477e-02 5.16580880e-01 -9.58921909e-01
-8.17790687e-01 1.11504900e+00 -1.70966357e-01 -5.55526391e-02
4.33114201e-01 -4.89804983e-01 8.75106096e-01 -1.13478291e+00
1.10268152e+00 7.52504230e-01 7.30432272e-01 3.27469289e-01
-4.11244512e-01 -6.34238124e-01 -1.89582214e-01 -1.20219357e-01
-1.78347671e+00 -3.53541434e-01 8.37757438e-02 -6.48680210e-01
1.21117842e+00 -2.64555782e-01 1.10055417e-01 5.44814110e-01
7.32231140e-02 2.50040442e-01 9.57050800e-01 -8.32216918e-01
-5.35387397e-02 6.28355384e-01 5.78701258e-01 1.03742230e+00
8.37127388e-01 -4.44836497e-01 -4.45043802e-01 -1.50572851e-01
5.76295376e-01 -3.28673899e-01 -1.27563149e-01 -1.00967199e-01
-1.30259275e+00 4.68887448e-01 4.02622044e-01 7.29329586e-01
-5.52631140e-01 -2.81887919e-01 6.54493868e-01 -6.91531971e-03
5.32272398e-01 7.65612662e-01 -8.20789933e-01 2.24885345e-01
-8.54150653e-01 -2.12671369e-01 1.25720108e+00 1.46379125e+00
1.05440724e+00 -2.95803130e-01 3.43468249e-01 9.41404283e-01
4.72611636e-01 5.64389884e-01 6.65161550e-01 -4.80041742e-01
9.12237823e-01 1.06966102e+00 4.85453784e-01 -9.08578217e-01
-4.81816649e-01 -1.29574791e-01 -2.03550711e-01 -4.13883746e-01
5.10203660e-01 -6.07168436e-01 -6.85047686e-01 1.56243670e+00
4.43473458e-01 -2.09541470e-02 5.60090780e-01 4.36843544e-01
1.41094553e+00 6.01224601e-01 4.97409433e-01 3.51207666e-02
1.70540524e+00 -9.47076380e-01 -9.64047134e-01 -2.29580671e-01
1.19150460e+00 -6.89840019e-01 5.70011675e-01 -4.17023182e-01
-7.03345180e-01 -3.57810527e-01 -8.45658064e-01 -4.47921902e-01
-1.49267066e+00 7.86423266e-01 5.39847553e-01 3.74034494e-01
-8.48984957e-01 2.12631360e-01 -7.08606362e-01 -1.05559719e+00
-3.36754739e-01 4.08925302e-02 -8.09320331e-01 -1.48270071e-01
-1.73115289e+00 1.17199624e+00 1.12438250e+00 5.17743006e-02
-3.02159697e-01 -2.89169520e-01 -1.35412157e+00 -3.20689172e-01
3.73194933e-01 -3.29055518e-01 9.18917716e-01 -5.13334095e-01
-7.58634508e-01 1.32552516e+00 -1.26883492e-01 -3.00906599e-01
-1.02218790e-02 -3.28020185e-01 -1.07490277e+00 -2.08032712e-01
9.47267830e-01 3.09998780e-01 -1.27687126e-01 -1.08234823e+00
-1.05617595e+00 -3.48786622e-01 1.57837898e-01 3.01521212e-01
-2.66381025e-01 3.60496074e-01 -8.31248045e-01 -3.86747241e-01
-2.35523447e-01 -7.53504753e-01 -1.53016180e-01 -8.64412606e-01
-6.74873769e-01 -2.88930595e-01 7.37463459e-02 -1.31985974e+00
1.56961238e+00 -1.86779392e+00 -3.46857935e-01 5.47520816e-02
1.45270033e-02 1.25562057e-01 4.75723520e-02 8.58239710e-01
-3.87132801e-02 3.50375354e-01 1.28051549e-01 4.70989570e-02
1.95553124e-01 -2.42992379e-02 -1.66249573e-02 2.26366863e-01
-2.14149430e-01 6.75021231e-01 -1.23193049e+00 -8.08366716e-01
-1.45279497e-01 3.55383337e-01 -7.47663677e-02 8.10142308e-02
1.00170054e-01 -3.77967884e-03 -5.38262963e-01 4.40209597e-01
3.25591564e-01 6.66896254e-02 6.44347489e-01 -3.86467099e-01
-7.88772047e-01 8.51587772e-01 -1.47651422e+00 1.74710965e+00
-5.32166660e-01 3.25703889e-01 -1.94138587e-01 -4.35817659e-01
1.01671040e+00 5.46068847e-01 1.03173099e-01 -2.28317961e-01
2.00033486e-02 7.91588664e-01 -4.95988458e-01 -6.15487456e-01
1.06845033e+00 1.94446281e-01 -7.84247100e-01 2.05079317e-01
6.23153687e-01 4.65667963e-01 9.40714359e-01 4.13939208e-01
7.46095181e-01 5.24707556e-01 8.97103369e-01 -5.52441239e-01
4.98670220e-01 5.54312468e-01 7.80818045e-01 3.25917810e-01
1.22841991e-01 -2.68284827e-01 1.04769655e-01 -2.46075347e-01
-1.16040313e+00 -7.68771112e-01 -2.98079491e-01 1.08060598e+00
3.08172591e-02 -8.39744151e-01 -8.74411941e-01 -7.69964933e-01
-1.66847423e-01 1.02286160e+00 -3.64592135e-01 3.09284002e-01
-3.76690656e-01 -3.87020051e-01 8.45487535e-01 4.26053554e-01
6.40501797e-01 -1.12754524e+00 1.00984955e-02 1.99939892e-01
-6.02531135e-01 -1.31832671e+00 -4.94308352e-01 2.22523764e-01
-2.69573480e-01 -1.02422833e+00 -5.23220956e-01 -1.47240174e+00
8.98028672e-01 -2.71141738e-01 1.05607545e+00 -1.53765023e-01
1.11673065e-01 4.19561625e-01 -4.25133735e-01 -3.60328883e-01
-3.81064385e-01 5.65135837e-01 3.49127799e-01 -3.84984046e-01
9.13733006e-01 -1.54958828e-03 8.97156373e-02 1.22079194e-01
-5.34397364e-01 -1.45227179e-01 2.87287325e-01 4.50308412e-01
6.31937325e-01 1.26769409e-01 5.59028924e-01 -1.34910119e+00
3.79740059e-01 -8.01639915e-01 -5.37178159e-01 6.67678416e-01
-3.75225961e-01 1.77810505e-01 4.53605294e-01 1.59021154e-01
-1.29963207e+00 1.69950902e-01 -1.87751070e-01 5.61353266e-01
-5.20267963e-01 1.29134989e+00 -4.69947487e-01 2.86952436e-01
5.94680071e-01 1.70963809e-01 -8.22217584e-01 -6.69348121e-01
5.93070269e-01 1.05499482e+00 7.52438843e-01 -6.52715027e-01
6.39143705e-01 -1.48939207e-01 -6.20007455e-01 -9.76684809e-01
-1.03687513e+00 -8.54752421e-01 -1.48095250e+00 -1.56246066e-01
1.10215235e+00 -1.38460541e+00 4.19544131e-02 1.42310277e-01
-1.12737179e+00 -1.60185218e-01 -2.26778924e-01 7.98420846e-01
-2.14420140e-01 2.44218960e-01 -8.50196004e-01 -5.27463913e-01
-2.76837319e-01 -4.31995451e-01 5.17045557e-01 5.18751383e-01
-1.05384305e-01 -1.37418783e+00 4.35975581e-01 3.34121704e-01
-2.00598925e-01 2.43097514e-01 7.52797663e-01 -1.40298545e+00
-3.04316711e-02 -4.78433073e-01 -1.59858286e-01 1.73032433e-01
2.02522144e-01 6.67778999e-02 -3.82372379e-01 -7.94039443e-02
-6.92684770e-01 -8.41119662e-02 3.05257499e-01 -9.81508270e-02
-5.60610667e-02 -4.63697881e-01 -6.21941030e-01 4.13117021e-01
1.87659287e+00 2.20640469e-02 4.80832785e-01 8.87767553e-01
9.05294597e-01 5.83520055e-01 6.24937177e-01 1.14997946e-01
9.53201354e-01 4.77619380e-01 -3.24642748e-01 -1.60856709e-01
-8.36468637e-02 -5.38499951e-01 4.09773290e-01 1.39947772e+00
-2.05793649e-01 -1.43737480e-01 -1.52065051e+00 1.10243917e+00
-1.67891860e+00 -1.07917857e+00 -4.25106078e-01 2.20178342e+00
1.09453678e+00 -1.05472729e-01 -9.88365784e-02 -7.78698802e-01
1.08602166e+00 -3.33457023e-01 -9.63417366e-02 -2.59914696e-01
-1.06978938e-01 -1.63531736e-01 1.04774189e+00 5.50093412e-01
-1.22543621e+00 1.41319025e+00 5.02720022e+00 5.10124087e-01
-6.96609259e-01 3.44150484e-01 -2.50120103e-01 5.78049600e-01
2.70088874e-02 1.52410924e-01 -1.54212070e+00 3.95203292e-01
1.26619923e+00 -8.29858184e-01 2.21750978e-02 9.66483057e-01
-5.39253391e-02 1.40409144e-02 -5.37178218e-01 5.02776563e-01
1.82734057e-01 -1.30184436e+00 -1.14315294e-01 1.23987995e-01
8.36205482e-01 5.53600550e-01 -6.85887754e-01 6.25326216e-01
9.74575639e-01 -4.92222041e-01 7.38993466e-01 5.79338491e-01
1.11491370e+00 -8.31574798e-01 1.16533709e+00 3.69393170e-01
-1.64173973e+00 3.17508221e-01 -5.45576274e-01 3.59078616e-01
3.12346131e-01 4.37667757e-01 -6.89185798e-01 1.02433038e+00
6.90930545e-01 7.01223373e-01 -5.97092509e-01 1.02697921e+00
-7.76483774e-01 3.63711476e-01 -1.67309448e-01 7.38108307e-02
2.96424866e-01 -2.87860006e-01 2.77411699e-01 1.83551705e+00
5.30933261e-01 9.28871185e-02 3.94902885e-01 2.09200293e-01
-4.97349441e-01 8.47342134e-01 -8.24891269e-01 -3.75302762e-01
8.87938082e-01 1.42917144e+00 -6.84747756e-01 -6.41179860e-01
-6.31942272e-01 8.78426135e-01 7.74985850e-01 2.59384722e-01
-5.71843863e-01 -1.19207346e+00 3.98534775e-01 1.33286983e-01
2.86899984e-01 -4.15524691e-01 1.37717113e-01 -1.47838259e+00
-1.88112244e-01 -4.97188270e-01 6.90492809e-01 -8.88437450e-01
-1.24981761e+00 8.82192433e-01 -1.35676891e-01 -1.01692212e+00
-2.95491993e-01 -7.39204228e-01 -1.73465401e-01 1.02067041e+00
-1.36190009e+00 -1.34754467e+00 -4.42612022e-02 5.48914731e-01
2.32408166e-01 -2.79426992e-01 1.15601587e+00 7.42038608e-01
-6.86101377e-01 4.60627794e-01 3.51390749e-01 1.16142392e+00
1.08984268e+00 -1.33581305e+00 4.20274436e-01 1.10124362e+00
1.98617354e-01 1.10628021e+00 4.49086308e-01 -1.19840562e+00
-8.49538982e-01 -1.45895064e+00 2.08294964e+00 -6.69040084e-01
1.16327345e+00 -3.25086653e-01 -7.81381607e-01 1.28825319e+00
5.77017426e-01 -2.01803073e-01 1.16306460e+00 3.02792788e-01
-4.52570111e-01 3.00704062e-01 -1.01105225e+00 5.11267304e-01
8.70863020e-01 -8.10410023e-01 -1.06124449e+00 6.32288992e-01
5.89415073e-01 -4.65316772e-01 -1.44182706e+00 -2.74908841e-01
2.34127671e-01 -4.24483716e-02 7.19543159e-01 -6.74669147e-01
1.02286428e-01 -7.30285406e-01 -2.77014524e-01 -1.36076641e+00
-3.79919946e-01 -8.83949846e-02 4.06187564e-01 1.98503637e+00
1.04483402e+00 -7.66757727e-01 4.39969227e-02 5.81866026e-01
-3.19368720e-01 5.43182604e-02 -6.48090184e-01 -9.37176824e-01
4.56396379e-02 -1.73265249e-01 4.10366803e-01 1.57168496e+00
6.97574019e-01 6.41050816e-01 -2.05391884e-01 6.49014890e-01
6.86518610e-01 -1.10279888e-01 5.29439032e-01 -1.15803063e+00
3.25903744e-01 1.53073967e-01 -2.50180274e-01 -6.09955966e-01
3.93622428e-01 -1.41240966e+00 3.01985532e-01 -2.21323037e+00
6.96406588e-02 -6.08774602e-01 -2.04176903e-01 1.01332307e+00
3.81551310e-02 -3.59396962e-03 -4.40265909e-02 3.25962782e-01
-7.83428192e-01 -3.89741324e-02 4.04426694e-01 2.10036531e-01
-1.65374056e-01 -5.63384712e-01 -5.74563503e-01 8.28190744e-01
9.00873244e-01 -8.86404991e-01 2.85014331e-01 -5.02852559e-01
4.34284747e-01 -2.01166674e-01 -2.03619868e-01 -9.54051912e-01
6.35997295e-01 -1.62744224e-01 9.34088230e-03 -3.84103268e-01
-2.88586944e-01 -6.01638734e-01 2.44178280e-01 2.38298848e-01
-2.98721969e-01 2.59119838e-01 2.03528032e-01 1.01682857e-01
-3.44505966e-01 -6.32461667e-01 5.13321817e-01 -4.68153208e-01
-1.33106077e+00 7.66482875e-02 -3.93718302e-01 3.94243181e-01
9.08740222e-01 6.73582926e-02 -4.60777521e-01 2.70371526e-01
-8.00721407e-01 2.35761210e-01 6.28842771e-01 4.47465450e-01
-1.95141464e-01 -1.47693086e+00 -6.98044777e-01 -2.44474307e-01
4.84940410e-01 -5.82442522e-01 1.90313160e-02 4.77914244e-01
-7.16008902e-01 6.64890051e-01 -3.78137887e-01 2.51523793e-01
-8.75115752e-01 6.78080082e-01 2.03907743e-01 -3.91027182e-01
-3.27203333e-01 4.41248983e-01 -2.36637250e-01 -1.19880509e+00
-1.33952171e-01 2.13876963e-01 -7.56303251e-01 1.77862540e-01
5.68213224e-01 1.88225001e-01 5.16246818e-02 -1.23083508e+00
-6.20568037e-01 4.54311520e-01 1.00239970e-01 -1.18341818e-01
1.46682608e+00 -4.11595494e-01 -4.03335333e-01 8.76173615e-01
1.00731003e+00 6.05533659e-01 -1.90544978e-01 -4.46666002e-01
7.35822380e-01 1.54528871e-01 -1.41797915e-01 -8.43302548e-01
-7.19926775e-01 2.99385190e-01 -1.27481278e-02 5.48186898e-03
5.39102435e-01 1.82678521e-01 5.80695331e-01 5.75742960e-01
8.75934899e-01 -1.42790580e+00 -8.98131192e-01 1.03286541e+00
4.62413490e-01 -9.82964575e-01 -1.21157058e-01 -5.43486178e-01
-7.03734457e-01 9.60331976e-01 7.50329733e-01 -2.14656629e-02
9.10877228e-01 3.12941909e-01 3.95968109e-01 -3.81887496e-01
-3.35522234e-01 -5.86296260e-01 2.32275873e-01 6.82120085e-01
8.94950092e-01 1.96299076e-01 -8.43764603e-01 8.69892895e-01
-2.13812575e-01 4.44426052e-02 6.88112199e-01 9.12695110e-01
-5.42137682e-01 -1.10756946e+00 -8.79812315e-02 1.79896221e-01
-6.30703866e-01 -5.89136362e-01 -2.37148046e-01 9.75615144e-01
4.56745565e-01 8.09831977e-01 1.86093040e-02 -1.73902929e-01
5.03795624e-01 5.28798342e-01 -8.97292420e-02 -1.12627459e+00
-6.59258306e-01 -3.57584685e-01 6.83077991e-01 -6.34524822e-02
-4.86561924e-01 -5.67379355e-01 -1.70401132e+00 -2.64367491e-01
-3.12167138e-01 8.49370897e-01 8.49583685e-01 8.15671086e-01
4.21204895e-01 1.27118573e-01 1.49414226e-01 -3.65030676e-01
2.69111902e-01 -8.17149758e-01 -8.24248433e-01 2.35033840e-01
-5.25579929e-01 -4.83068526e-01 -1.91920534e-01 4.32017565e-01] | [9.712125778198242, 9.50732421875] |
c68c43ba-5306-4944-9d5e-e18d7b9d126b | a-new-comprehensive-benchmark-for-semi-1 | 2305.13611 | null | https://arxiv.org/abs/2305.13611v1 | https://arxiv.org/pdf/2305.13611v1.pdf | A New Comprehensive Benchmark for Semi-supervised Video Anomaly Detection and Anticipation | Semi-supervised video anomaly detection (VAD) is a critical task in the intelligent surveillance system. However, an essential type of anomaly in VAD named scene-dependent anomaly has not received the attention of researchers. Moreover, there is no research investigating anomaly anticipation, a more significant task for preventing the occurrence of anomalous events. To this end, we propose a new comprehensive dataset, NWPU Campus, containing 43 scenes, 28 classes of abnormal events, and 16 hours of videos. At present, it is the largest semi-supervised VAD dataset with the largest number of scenes and classes of anomalies, the longest duration, and the only one considering the scene-dependent anomaly. Meanwhile, it is also the first dataset proposed for video anomaly anticipation. We further propose a novel model capable of detecting and anticipating anomalous events simultaneously. Compared with 7 outstanding VAD algorithms in recent years, our method can cope with scene-dependent anomaly detection and anomaly anticipation both well, achieving state-of-the-art performance on ShanghaiTech, CUHK Avenue, IITB Corridor and the newly proposed NWPU Campus datasets consistently. Our dataset and code is available at: https://campusvad.github.io. | ['Yanning Zhang', 'Peng Wang', 'Yue Lu', 'Congqi Cao'] | 2023-05-23 | a-new-comprehensive-benchmark-for-semi | http://openaccess.thecvf.com//content/CVPR2023/html/Cao_A_New_Comprehensive_Benchmark_for_Semi-Supervised_Video_Anomaly_Detection_and_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Cao_A_New_Comprehensive_Benchmark_for_Semi-Supervised_Video_Anomaly_Detection_and_CVPR_2023_paper.pdf | cvpr-2023-1 | ['video-anomaly-detection'] | ['computer-vision'] | [-8.39001909e-02 -5.53364933e-01 1.07082695e-01 -1.87898025e-01
-1.03729896e-01 -2.98538119e-01 4.66110080e-01 1.58075213e-01
-9.34797823e-02 3.27778190e-01 4.44557471e-03 -3.95016223e-01
6.19911142e-02 -5.26444912e-01 -6.29059732e-01 -8.04242671e-01
-3.37858617e-01 1.01475380e-01 4.87376422e-01 -2.98068643e-01
1.39013067e-01 5.67186117e-01 -1.87565362e+00 1.42194912e-01
9.95274603e-01 1.43952894e+00 -2.21835494e-01 4.91544634e-01
5.75885316e-03 9.20725167e-01 -4.39318955e-01 -6.57518879e-02
5.62270045e-01 -5.55901110e-01 -4.93222564e-01 4.79396313e-01
5.71964383e-01 -7.12843657e-01 -6.94195390e-01 9.59963620e-01
2.37980455e-01 3.14433128e-02 3.89295131e-01 -1.80081403e+00
-3.78219932e-01 4.20943275e-02 -6.97435379e-01 8.56267869e-01
3.64902824e-01 3.69399637e-01 6.82550251e-01 -7.41458237e-01
5.43576658e-01 7.97955692e-01 3.90379906e-01 6.05127513e-01
-3.26870829e-01 -6.26330256e-01 5.79763889e-01 8.67482364e-01
-1.15128040e+00 -2.02484086e-01 1.03188133e+00 -5.02541900e-01
6.54985666e-01 3.28406811e-01 8.73388112e-01 1.40952885e+00
1.68519244e-01 1.00610316e+00 5.83868444e-01 3.78706083e-02
1.78972468e-01 -2.38127872e-01 1.84252486e-01 6.44214451e-01
2.99185634e-01 -2.37984106e-01 -1.30861044e-01 -1.91487759e-01
2.64852256e-01 4.77571934e-01 -3.49570394e-01 -3.72859001e-01
-1.12478495e+00 4.75093782e-01 -6.87901303e-02 1.53931960e-01
-4.70047504e-01 -5.23218691e-01 8.67088258e-01 5.92481971e-01
5.41239977e-01 7.73022743e-03 -4.84813213e-01 -3.68403822e-01
-4.91087645e-01 2.04663321e-01 4.29356813e-01 1.01651335e+00
1.39216900e-01 5.89941978e-01 -1.69643074e-01 4.45951402e-01
1.97213978e-01 6.23779953e-01 4.03917015e-01 -5.96320987e-01
3.19307715e-01 7.77576983e-01 9.33978036e-02 -1.25970685e+00
-5.08359253e-01 -3.16058815e-01 -1.03358114e+00 3.08998674e-02
5.80600858e-01 -2.34261945e-01 -7.28236616e-01 1.43395257e+00
4.93090838e-01 7.21960008e-01 6.23881333e-02 8.96211565e-01
8.15021455e-01 8.57369542e-01 -8.75741020e-02 -7.21364498e-01
1.06237721e+00 -9.44365442e-01 -1.02301407e+00 -6.17315508e-02
6.81406796e-01 -8.16571593e-01 8.21650505e-01 8.87342513e-01
-6.58919573e-01 -4.18632537e-01 -8.35641444e-01 2.93847084e-01
-2.54576087e-01 4.67165373e-02 6.91567123e-01 3.69933903e-01
-6.23183012e-01 1.60252117e-02 -8.09454441e-01 -7.00823069e-01
6.43944442e-01 -3.17621320e-01 -3.58807057e-01 -1.92797855e-01
-1.04179013e+00 6.34518266e-01 3.48806858e-01 3.53774250e-01
-1.12208533e+00 -4.74297434e-01 -8.39162946e-01 -3.39203626e-01
7.23183572e-01 -1.00017153e-01 9.93427277e-01 -1.07624376e+00
-9.06161964e-01 6.78537726e-01 -1.55032724e-01 -6.13713920e-01
7.13335335e-01 -4.90944624e-01 -1.12610102e+00 2.20390763e-02
-1.19634792e-01 -1.15481623e-01 9.99622643e-01 -9.41415250e-01
-6.74148440e-01 -3.96036923e-01 -1.28613964e-01 3.94047871e-02
-4.81943250e-01 7.59217143e-02 -3.19421500e-01 -6.79797411e-01
1.59024298e-01 -7.48304367e-01 -1.80071920e-01 1.60211697e-01
-4.72399950e-01 -3.64489257e-01 1.55878341e+00 -8.47515225e-01
1.49813974e+00 -2.52565837e+00 -1.57004535e-01 1.46106437e-01
8.13026875e-02 6.36584342e-01 -8.62306431e-02 1.59747288e-01
-1.99725807e-01 -3.59282106e-01 -4.36564654e-01 3.21492786e-03
-3.71481210e-01 3.69614303e-01 -4.42949176e-01 6.62039638e-01
2.78210789e-01 2.88742304e-01 -9.66755450e-01 -4.53729302e-01
3.48693877e-01 -6.09562844e-02 -5.15879810e-01 4.22875941e-01
-2.31803969e-01 7.90540397e-01 -6.41292334e-01 1.16483533e+00
1.02227521e+00 2.54317790e-01 -4.34337407e-01 2.00473685e-02
-3.85197878e-01 -4.44091022e-01 -1.18396890e+00 1.54997885e+00
3.59572619e-01 7.80733287e-01 -1.45242333e-01 -1.21666110e+00
1.04891837e+00 4.25737321e-01 8.67334664e-01 -5.34051359e-01
1.65513858e-01 1.99627027e-01 -4.27635461e-02 -1.04981518e+00
3.89427781e-01 7.24876940e-01 1.85407132e-01 1.55286223e-01
-9.14048702e-02 4.17164803e-01 4.39681292e-01 2.94339567e-01
1.28985369e+00 -2.82906126e-02 3.99088919e-01 -1.45588890e-01
8.37539434e-01 2.39704207e-01 9.80369687e-01 6.71241045e-01
-8.70059013e-01 4.95072782e-01 7.90284812e-01 -9.58414853e-01
-1.09862661e+00 -1.11613381e+00 -2.22070515e-01 6.86001301e-01
3.62333030e-01 -2.55994081e-01 -5.10077953e-01 -9.27983701e-01
-1.97530985e-01 6.58082068e-01 -4.08173531e-01 -1.86369032e-01
-5.35653591e-01 -8.74822021e-01 5.12833238e-01 1.90298751e-01
9.56866562e-01 -1.19701540e+00 -3.63819271e-01 -1.29358798e-01
-4.30018514e-01 -1.34048891e+00 -2.79306829e-01 -5.07848918e-01
-8.19763303e-01 -1.56317735e+00 -4.53129143e-01 -6.70701504e-01
7.78306007e-01 4.23896313e-01 9.76509690e-01 2.11080648e-02
-5.06610274e-01 6.20261371e-01 -6.50441647e-01 -6.70674622e-01
-2.27721021e-01 -4.73359197e-01 3.80636185e-01 4.28612590e-01
6.80959821e-01 -3.18584502e-01 -6.32836223e-01 4.85212147e-01
-9.06777799e-01 -3.18523824e-01 4.87202197e-01 5.62995553e-01
6.80896342e-01 7.96107724e-02 4.67112124e-01 -6.36910975e-01
1.76848114e-01 -9.86047149e-01 -6.80630088e-01 -9.50893313e-02
-2.47118473e-01 -6.44069314e-01 9.63944018e-01 -2.16637895e-01
-1.17066884e+00 -1.52883669e-02 -2.88221836e-01 -6.24221027e-01
-8.46408188e-01 -9.07811988e-03 -2.25734696e-01 2.15963379e-01
5.72599828e-01 5.94504058e-01 2.11236365e-02 -3.46724898e-01
-1.45396277e-01 5.66150784e-01 6.59690082e-01 8.57477076e-03
9.74018276e-01 6.95291102e-01 5.61608300e-02 -1.22518218e+00
-9.00571227e-01 -7.14844823e-01 -6.34498239e-01 -4.27429140e-01
9.54205215e-01 -9.21310604e-01 -2.79949427e-01 1.13557088e+00
-1.09846842e+00 1.04781471e-01 -2.76527315e-01 4.71975654e-01
-2.27013141e-01 8.18279326e-01 -3.08187336e-01 -8.80258262e-01
-2.18556359e-01 -8.68002594e-01 6.03048742e-01 9.35041606e-02
2.39156380e-01 -5.69216967e-01 6.59394264e-02 1.78094476e-01
4.06294674e-01 8.17178428e-01 3.92646641e-01 -9.34934258e-01
-6.11786485e-01 -3.38097692e-01 -1.04838558e-01 5.29201865e-01
2.48643488e-01 3.20789278e-01 -8.88497293e-01 -2.79156715e-01
1.69613644e-01 6.37914538e-02 8.03137004e-01 3.80661756e-01
1.66046238e+00 -2.81197071e-01 -4.90124337e-02 6.43649936e-01
1.02194369e+00 6.15480661e-01 7.78144717e-01 5.55513442e-01
8.91373217e-01 1.73646227e-01 1.16722941e+00 9.90237951e-01
1.80542231e-01 4.20994699e-01 1.00781691e+00 6.86406270e-02
3.90983254e-01 2.24958539e-01 5.78083992e-01 7.76315570e-01
-2.74171770e-01 -4.33899850e-01 -9.40284133e-01 6.31538153e-01
-1.92244923e+00 -1.34970295e+00 -6.57226682e-01 2.07795644e+00
-6.83910176e-02 1.53470621e-01 2.24334702e-01 4.05497611e-01
6.35885894e-01 5.23565590e-01 -7.54480779e-01 -3.19871545e-01
-3.69734675e-01 -4.41607714e-01 7.25257397e-02 -1.44380555e-01
-1.67909026e+00 6.27962053e-01 5.34861088e+00 5.21323621e-01
-8.31503928e-01 -7.11203963e-02 5.42460322e-01 -8.65107328e-02
2.86998332e-01 -3.55455995e-01 -3.87146920e-01 8.30431581e-01
6.55053735e-01 2.59149931e-02 1.11829571e-01 1.22548711e+00
4.53701138e-01 5.96370772e-02 -8.57336104e-01 9.87296522e-01
4.29967284e-01 -7.88325250e-01 3.07561159e-01 -3.35828543e-01
6.65728927e-01 -8.73001069e-02 -6.46226332e-02 1.59769624e-01
-4.66902226e-01 -4.95438069e-01 2.28044957e-01 6.07188106e-01
2.98609346e-01 -8.96097541e-01 1.09223592e+00 2.03152716e-01
-1.23988509e+00 -3.40676546e-01 -3.25125813e-01 -1.54604554e-01
8.96187946e-02 6.15838647e-01 -1.98561147e-01 6.13006115e-01
1.08517611e+00 1.31448591e+00 -7.76502728e-01 1.36047888e+00
9.52125818e-04 8.34145963e-01 -1.46711141e-01 3.16178769e-01
3.66351694e-01 -2.96244323e-01 1.17243469e+00 1.06692517e+00
5.91711938e-01 1.98998496e-01 4.21339869e-01 2.54669916e-02
2.98389763e-01 1.28434643e-01 -1.09782791e+00 2.91759700e-01
2.72876263e-01 9.93093491e-01 -2.31645748e-01 -3.55474800e-01
-7.07851052e-01 1.04325163e+00 -2.45663226e-01 2.81407833e-01
-1.12745380e+00 -2.82998204e-01 1.00397372e+00 -7.57176206e-02
1.01878971e-01 -2.30422452e-01 5.25724515e-02 -1.39784777e+00
4.37129647e-01 -9.88725483e-01 8.99865091e-01 -4.27462220e-01
-1.34536159e+00 5.92600882e-01 -6.22862391e-02 -1.98013580e+00
-8.46015140e-02 -7.11168587e-01 -1.14996290e+00 1.06491663e-01
-1.43748522e+00 -8.09126854e-01 -1.00969875e+00 1.09740913e+00
9.68221962e-01 -6.12325609e-01 6.50038600e-01 5.49479067e-01
-1.19562781e+00 4.58794504e-01 2.05559060e-01 2.89572090e-01
6.88920379e-01 -1.03687751e+00 2.76059061e-01 1.58211219e+00
-3.06281328e-01 -2.08529517e-01 8.68038118e-01 -5.66861391e-01
-1.41657615e+00 -1.36339939e+00 3.27534169e-01 -5.32669008e-01
7.14694440e-01 -1.69616386e-01 -1.27300537e+00 9.06496525e-01
1.72680840e-01 3.37469786e-01 4.04596359e-01 -4.36754018e-01
-1.04115434e-01 -2.35139549e-01 -1.28000569e+00 4.92977291e-01
1.34071088e+00 -4.09809239e-02 -4.19057250e-01 6.68220818e-01
5.93399167e-01 -6.51454985e-01 -4.11118478e-01 4.96024877e-01
1.62543535e-01 -1.23908937e+00 7.62662590e-01 -5.67526042e-01
4.63005662e-01 -5.44979572e-01 -3.93185252e-03 -1.16902292e+00
-2.36384779e-01 -4.51144934e-01 -7.84080744e-01 1.19539881e+00
7.29907863e-03 -8.66287053e-01 6.71989024e-01 1.74002081e-01
-6.57907963e-01 -6.49122596e-01 -9.79749978e-01 -9.27270055e-01
-6.86757982e-01 -7.02029228e-01 4.91003454e-01 1.04508281e+00
-3.70664448e-01 -5.65540850e-01 -9.75536048e-01 5.59049666e-01
8.24995637e-01 -9.24244300e-02 9.44422722e-01 -1.13765788e+00
1.75950810e-01 -3.01051766e-01 -9.62463439e-01 -7.60704637e-01
-2.17455804e-01 -3.68253261e-01 -1.55692562e-01 -1.13718927e+00
-7.80925080e-02 -4.92054895e-02 -5.13107657e-01 3.70451897e-01
-1.73550472e-01 5.97958900e-02 -2.17395544e-01 1.67229801e-01
-8.78664851e-01 6.26827121e-01 1.07340562e+00 -4.91873026e-02
-2.41163447e-01 9.50191021e-02 -2.63450801e-01 1.01608205e+00
1.17347813e+00 -1.37114361e-01 -3.55809689e-01 -3.72497439e-01
-4.86793965e-01 -3.14357251e-01 4.56447482e-01 -1.36819839e+00
1.21957093e-01 -4.27398741e-01 4.14778233e-01 -9.85682309e-01
3.01630422e-02 -1.05553865e+00 -1.79882899e-01 6.16644859e-01
1.51920855e-01 4.89822954e-01 3.93118471e-01 8.89372766e-01
-3.64400625e-01 2.00149313e-01 5.54624677e-01 1.17425650e-01
-1.57270694e+00 9.04821217e-01 -6.74446583e-01 2.53688693e-01
1.67662859e+00 -9.19975415e-02 -4.94898200e-01 -2.02140257e-01
-4.63035911e-01 6.07494295e-01 1.14190936e-01 8.32645655e-01
9.04855967e-01 -1.34871817e+00 -9.07660484e-01 4.81865197e-01
4.48172927e-01 1.61300749e-01 8.32609713e-01 9.86676812e-01
-7.33665109e-01 9.78692174e-02 -6.19510293e-01 -8.43307197e-01
-1.40903389e+00 7.17195570e-01 7.28594288e-02 2.89079957e-02
-8.40037048e-01 4.49740082e-01 -4.54373052e-03 3.62375304e-02
4.19804126e-01 1.65692344e-02 -6.01961970e-01 -1.81530166e-04
7.87242651e-01 6.15457535e-01 -1.66924611e-01 -7.06369102e-01
-4.24912184e-01 3.88986677e-01 -1.32282645e-01 8.31534982e-01
1.08767927e+00 -1.48078948e-01 -1.22538999e-01 4.13617402e-01
7.87473619e-01 -2.06868649e-01 -1.32746768e+00 -1.29208520e-01
-1.12990119e-01 -7.95388281e-01 -3.22965056e-01 -2.02221245e-01
-1.25491810e+00 6.40378654e-01 9.96643782e-01 3.46394777e-01
1.56677592e+00 -2.74361342e-01 1.08743024e+00 5.47636688e-01
1.81133971e-01 -1.05671299e+00 1.85841814e-01 5.92457354e-01
9.04327095e-01 -1.64945316e+00 -9.53538939e-02 -3.60519499e-01
-8.71583164e-01 9.11126375e-01 1.26946771e+00 -2.48336703e-01
6.92570090e-01 -1.54471368e-01 1.71132565e-01 -1.67354673e-01
-5.18784881e-01 -1.05311930e-01 2.66551286e-01 6.70244098e-01
-5.88156190e-03 -1.11577176e-01 -1.67236954e-01 2.16957018e-01
1.89415038e-01 -2.02112377e-01 6.49216056e-01 9.65318263e-01
-4.20647800e-01 -4.89632368e-01 -3.52315247e-01 8.49221468e-01
-4.45292979e-01 2.27906063e-01 -1.10196382e-01 8.49139094e-01
3.01528811e-01 9.12753463e-01 3.78704011e-01 -3.77758086e-01
5.78594029e-01 9.40384343e-02 -3.12729329e-01 1.24486163e-02
-3.34811732e-02 -4.83092695e-01 -2.46365648e-02 -9.79152441e-01
-2.94372797e-01 -7.97655225e-01 -8.33200037e-01 -4.22357976e-01
9.66185778e-02 -1.21611531e-03 2.02334285e-01 9.14801240e-01
3.22555095e-01 5.14527917e-01 1.01761210e+00 -5.28728604e-01
-8.02828744e-02 -7.62033343e-01 -5.40342569e-01 8.01695824e-01
5.90068758e-01 -6.30180955e-01 -6.21102095e-01 -1.52192801e-01] | [7.839961528778076, 1.5766037702560425] |
35f12f90-03a4-44d9-a495-950528649394 | syllable-based-sequence-to-sequence-speech | 1804.10752 | null | http://arxiv.org/abs/1804.10752v2 | http://arxiv.org/pdf/1804.10752v2.pdf | Syllable-Based Sequence-to-Sequence Speech Recognition with the Transformer in Mandarin Chinese | Sequence-to-sequence attention-based models have recently shown very
promising results on automatic speech recognition (ASR) tasks, which integrate
an acoustic, pronunciation and language model into a single neural network. In
these models, the Transformer, a new sequence-to-sequence attention-based model
relying entirely on self-attention without using RNNs or convolutions, achieves
a new single-model state-of-the-art BLEU on neural machine translation (NMT)
tasks. Since the outstanding performance of the Transformer, we extend it to
speech and concentrate on it as the basic architecture of sequence-to-sequence
attention-based model on Mandarin Chinese ASR tasks. Furthermore, we
investigate a comparison between syllable based model and context-independent
phoneme (CI-phoneme) based model with the Transformer in Mandarin Chinese.
Additionally, a greedy cascading decoder with the Transformer is proposed for
mapping CI-phoneme sequences and syllable sequences into word sequences.
Experiments on HKUST datasets demonstrate that syllable based model with the
Transformer performs better than CI-phoneme based counterpart, and achieves a
character error rate (CER) of \emph{$28.77\%$}, which is competitive to the
state-of-the-art CER of $28.0\%$ by the joint CTC-attention based
encoder-decoder network. | ['Bo Xu', 'Shuang Xu', 'Shiyu Zhou', 'Linhao Dong'] | 2018-04-28 | null | null | null | null | ['sequence-to-sequence-speech-recognition'] | ['speech'] | [ 4.68201727e-01 -1.06011666e-01 -3.05313990e-02 -2.45785251e-01
-1.24247169e+00 -2.07737967e-01 4.83491331e-01 -5.04569232e-01
-5.55627823e-01 6.31115913e-01 2.19633833e-01 -1.05007148e+00
7.06851840e-01 -2.05922008e-01 -8.58175159e-01 -7.48400509e-01
4.38703626e-01 4.96619523e-01 -1.18535221e-01 -3.59883398e-01
-8.77409503e-02 -9.82080866e-03 -1.06611431e+00 3.31338644e-01
9.55791652e-01 7.49588132e-01 6.73993349e-01 9.00066614e-01
-1.00580610e-01 5.39728165e-01 -5.91970146e-01 -5.59340835e-01
-1.03724509e-01 -7.42202640e-01 -5.70069730e-01 -1.74789056e-01
1.96120828e-01 -1.03434816e-01 -4.49393332e-01 1.08047748e+00
7.56673276e-01 5.54259755e-02 6.43739581e-01 -4.98531848e-01
-1.50639069e+00 9.34996426e-01 -4.06282365e-01 3.60110998e-01
-8.15302134e-02 8.34678710e-02 1.02957320e+00 -1.43097389e+00
1.19106844e-01 1.07571650e+00 4.05851215e-01 8.87371182e-01
-8.60132277e-01 -5.12956858e-01 1.80988193e-01 5.36427140e-01
-1.60343778e+00 -6.61525071e-01 4.90004510e-01 -1.67960078e-01
1.85743535e+00 4.42974925e-01 2.14236736e-01 1.24547243e+00
1.95820153e-01 8.17235112e-01 8.60528469e-01 -7.99175799e-01
3.78343761e-02 5.19291125e-02 -1.50248138e-02 4.24988508e-01
-3.09073776e-01 1.71392307e-01 -4.92017776e-01 4.83223796e-01
6.86195493e-01 -3.20966870e-01 -3.81987721e-01 5.60561717e-01
-1.33547568e+00 6.65472865e-01 2.19586074e-01 4.64711607e-01
-4.70560610e-01 3.19285780e-01 3.40688467e-01 3.65188897e-01
5.20753384e-01 -4.57382984e-02 -6.66916370e-01 -3.49008441e-01
-8.99772763e-01 -4.80285585e-01 3.81585926e-01 1.34330428e+00
4.01615083e-01 8.42320204e-01 -5.17101765e-01 1.23157620e+00
2.30824843e-01 8.50450754e-01 9.67347741e-01 -2.89647549e-01
6.60978198e-01 -1.18449710e-01 -2.84350902e-01 -8.41711983e-02
2.69645810e-01 -7.05242515e-01 -1.10321951e+00 -3.46755087e-01
-2.05268577e-01 -1.76192433e-01 -1.25508630e+00 1.78779221e+00
-1.60361320e-01 2.25493386e-01 1.72218978e-01 8.30831468e-01
5.66970468e-01 1.13380790e+00 -5.64553514e-02 -3.76066387e-01
1.31109154e+00 -1.46967793e+00 -1.00840068e+00 -3.12042117e-01
5.37389755e-01 -9.09332395e-01 1.38931048e+00 8.00409093e-02
-1.35493028e+00 -8.44994187e-01 -1.08989894e+00 -2.11329594e-01
-4.50828969e-01 4.77607846e-01 -8.60978588e-02 8.90878260e-01
-1.39143968e+00 1.87909946e-01 -6.93734944e-01 -2.53523678e-01
-7.29265213e-02 2.45235607e-01 -2.14113425e-02 1.03935987e-01
-1.36514640e+00 1.06147456e+00 3.95144969e-01 4.51693594e-01
-1.17513156e+00 -4.32711393e-01 -8.90205920e-01 4.06509101e-01
5.71075343e-02 -5.24600029e-01 1.45607233e+00 -8.62861872e-01
-2.09108686e+00 5.68515956e-01 -6.04954362e-01 -5.71461916e-01
-5.18610366e-02 -2.91730613e-01 -7.04738915e-01 -2.13997424e-01
-3.82383108e-01 6.83355331e-01 8.57177615e-01 -7.75413632e-01
-5.99191904e-01 9.17813554e-03 -5.57709992e-01 2.44779229e-01
-3.90273482e-01 5.15227139e-01 -4.87053692e-01 -9.49337065e-01
-2.33111456e-01 -9.52313542e-01 2.31959671e-02 -7.99967885e-01
-2.84377813e-01 -3.04225534e-01 6.95251107e-01 -1.24003518e+00
1.59988868e+00 -1.94781959e+00 3.23966026e-01 -3.78247917e-01
-3.80615294e-01 7.97316968e-01 -3.33787054e-01 3.91048253e-01
-2.39059150e-01 2.76972651e-01 -5.28005838e-01 -6.45433962e-01
4.09235470e-02 2.07129940e-01 -3.31865788e-01 1.47185147e-01
6.32294297e-01 1.39399338e+00 -5.87832332e-01 -9.78744030e-02
2.66017705e-01 7.30852544e-01 -2.53039747e-01 3.21886301e-01
-1.55976504e-01 3.91449094e-01 9.25201029e-02 7.01801777e-01
3.67679805e-01 1.04648270e-01 1.25544265e-01 4.18066591e-01
-2.52073467e-01 9.21985269e-01 -2.12758765e-01 1.65987718e+00
-6.72250748e-01 6.59965158e-01 2.57877149e-02 -9.79868114e-01
1.02742553e+00 8.71399283e-01 -2.92952955e-01 -8.39811921e-01
1.37453631e-01 5.97211301e-01 4.85986352e-01 -1.33487538e-01
5.11319339e-01 -2.13326052e-01 -5.62175810e-02 2.44322702e-01
2.85821944e-01 9.66262817e-02 -1.99554026e-01 -1.75562397e-01
9.72393513e-01 2.29863286e-01 2.03233570e-01 -3.50789815e-01
8.06898057e-01 -4.32838947e-01 4.08696204e-01 4.88067627e-01
-9.15955529e-02 9.29358304e-01 -8.27047825e-02 -5.45504643e-03
-1.42079794e+00 -9.58984017e-01 8.52333009e-02 1.16265893e+00
-5.49413681e-01 -7.80078992e-02 -1.14689636e+00 -3.88334513e-01
-6.67068899e-01 1.04361975e+00 -3.92964184e-01 -1.66023582e-01
-1.08553696e+00 -7.61796236e-01 7.60492980e-01 7.12069631e-01
4.69529092e-01 -1.34088624e+00 1.34783775e-01 5.48039854e-01
-5.33930898e-01 -1.21880460e+00 -1.11427736e+00 3.37531388e-01
-4.71712649e-01 -9.49960127e-02 -1.26267767e+00 -1.20786738e+00
3.44333977e-01 1.76187113e-01 8.83526921e-01 -1.83960274e-01
2.31623963e-01 -7.24831894e-02 -7.63393998e-01 -3.19514573e-01
-7.29593456e-01 3.69163752e-01 1.71363294e-01 2.25581437e-01
4.99878764e-01 -4.23957884e-01 -2.29867965e-01 2.52007037e-01
-5.74026585e-01 1.56122711e-04 8.74932647e-01 1.06290472e+00
6.04071617e-01 -6.01023376e-01 9.14784610e-01 -3.64811599e-01
4.85100001e-01 -2.97833830e-01 -4.40834910e-01 4.71290976e-01
-6.11275613e-01 -5.23849390e-02 9.31158543e-01 -3.66486073e-01
-1.06576240e+00 -5.63851669e-02 -6.27053559e-01 -7.29225934e-01
-1.16670720e-01 6.31590664e-01 -4.48414415e-01 4.06254441e-01
1.95093408e-01 1.13769269e+00 -2.84646809e-01 -7.20990717e-01
4.07766283e-01 1.33542669e+00 8.18115771e-01 -2.73974806e-01
5.35833478e-01 -3.26210856e-01 -6.67422652e-01 -9.78436768e-01
-3.01581025e-01 -3.66776079e-01 -6.75515115e-01 1.51425824e-01
1.15536320e+00 -1.08269179e+00 -3.94757897e-01 6.47671640e-01
-1.58084083e+00 -5.41996598e-01 1.19660527e-01 6.20893359e-01
-6.58052504e-01 4.23870414e-01 -1.01648831e+00 -1.01633310e+00
-7.42193937e-01 -1.26584029e+00 9.94319320e-01 -2.06833214e-01
1.06337227e-01 -8.29519749e-01 -1.83638886e-01 3.15554231e-01
7.94200599e-01 -5.82801521e-01 1.07211411e+00 -7.36181617e-01
-6.55048072e-01 -1.80629734e-02 -7.36175627e-02 8.12328815e-01
1.07033119e-01 -3.21130991e-01 -1.12598693e+00 -3.66641194e-01
1.01024397e-01 1.02116756e-01 9.01027739e-01 4.40640867e-01
1.07638323e+00 -4.24503893e-01 7.10258409e-02 4.96574730e-01
1.23714852e+00 7.75746226e-01 1.09115314e+00 -1.10693388e-01
7.76302218e-01 1.85201153e-01 2.24113181e-01 -1.05931379e-01
3.49363655e-01 8.66574228e-01 4.89115603e-02 -1.28263056e-01
-5.57174861e-01 -3.55447143e-01 9.85020399e-01 2.00768209e+00
-3.76443379e-02 -7.16887116e-01 -7.62350798e-01 7.72673845e-01
-1.66651428e+00 -5.98326564e-01 -9.07974094e-02 2.41608548e+00
8.64582539e-01 7.41288662e-02 -1.30416647e-01 3.07396450e-03
1.11472797e+00 2.23196037e-02 -1.31097376e-01 -1.07126081e+00
-2.90842503e-01 6.26904666e-01 4.12944406e-01 7.72886276e-01
-8.03485572e-01 1.50438011e+00 5.74047852e+00 1.30750620e+00
-1.23712492e+00 5.96486270e-01 6.23570740e-01 3.00555617e-01
-3.04978698e-01 -1.79553896e-01 -9.46881294e-01 5.96828759e-01
1.83428168e+00 5.82887828e-02 7.69133389e-01 6.25196159e-01
3.55012089e-01 6.11784458e-01 -1.09981179e+00 8.61260295e-01
2.19065666e-01 -1.14086652e+00 2.61629015e-01 9.32950005e-02
6.15744472e-01 3.86193156e-01 2.69355446e-01 5.91697931e-01
2.33672053e-01 -1.26295674e+00 8.55370104e-01 6.93982169e-02
1.45357502e+00 -7.42691219e-01 9.12165344e-01 3.21184695e-01
-1.37245870e+00 1.93294048e-01 -4.71004099e-01 -2.06024405e-02
4.62393761e-01 1.66526988e-01 -9.15253282e-01 6.81086361e-01
3.88152272e-01 4.89476293e-01 -1.75752863e-01 6.73762977e-01
-3.67909819e-01 1.20150530e+00 7.73344114e-02 -2.94363260e-01
4.76649284e-01 -1.63505003e-01 5.13894677e-01 1.73210430e+00
7.29998708e-01 1.90382358e-02 -3.98278534e-01 8.30449641e-01
-1.52990207e-01 1.42606646e-01 -3.00474077e-01 -3.40013474e-01
5.76331079e-01 8.81827950e-01 -3.54136378e-01 -5.56611478e-01
-4.66237485e-01 1.44263649e+00 4.02596056e-01 4.89212215e-01
-1.03452134e+00 -6.01575196e-01 6.83501959e-01 -2.79328674e-01
9.98656571e-01 -5.08430064e-01 -2.63600856e-01 -1.12436533e+00
6.93567097e-02 -9.67474878e-01 -2.88890660e-01 -8.39593112e-01
-1.06684947e+00 1.15875840e+00 -4.92671013e-01 -1.21729517e+00
-3.41627598e-01 -5.66317797e-01 -6.12916827e-01 1.37777710e+00
-1.74435663e+00 -1.24997246e+00 4.30887222e-01 3.77375901e-01
1.14345980e+00 -3.69061679e-01 1.04905057e+00 5.63479900e-01
-8.14352930e-01 1.16944134e+00 3.66119683e-01 2.86259800e-01
3.78509015e-01 -1.18738055e+00 1.23316717e+00 1.22717726e+00
3.41201752e-01 6.24675572e-01 1.37403101e-01 -5.07753909e-01
-1.34039295e+00 -1.30667579e+00 1.53817570e+00 -5.11219442e-01
4.84472305e-01 -7.50206411e-01 -1.00922298e+00 8.36843491e-01
7.83090413e-01 -3.68694305e-01 4.78487641e-01 -2.12796018e-01
-2.38619372e-01 8.54522958e-02 -5.77900231e-01 7.45079696e-01
9.82550383e-01 -8.05159092e-01 -6.36336148e-01 -5.89075983e-02
1.38742733e+00 -2.25640029e-01 -6.57423377e-01 3.01300883e-01
2.95053452e-01 -4.85397667e-01 6.61520779e-01 -5.02907813e-01
3.69405806e-01 -3.03888470e-01 -5.69942534e-01 -1.58311248e+00
-6.44334793e-01 -9.51931715e-01 -8.10685083e-02 1.27196765e+00
9.30320382e-01 -5.98635912e-01 2.60453939e-01 -9.36863571e-02
-1.13689363e+00 -5.81845045e-01 -1.29041946e+00 -1.09065247e+00
5.41429400e-01 -4.64482695e-01 6.47798717e-01 7.39521384e-01
-2.52067540e-02 7.57616818e-01 -6.49556696e-01 1.28302082e-01
1.32814720e-01 -5.12765110e-01 1.59319833e-01 -3.96017492e-01
-4.58854228e-01 -5.36180556e-01 1.33262267e-02 -1.47506833e+00
1.24438025e-01 -1.04712152e+00 4.90362972e-01 -1.30701828e+00
6.13169670e-02 -5.76158706e-03 -5.85264683e-01 4.86148328e-01
-3.57252657e-01 1.59484893e-01 2.76871175e-01 1.33385137e-02
-1.96299627e-01 9.92515445e-01 9.56961572e-01 -7.28856847e-02
-8.58384836e-03 -1.55094430e-01 -4.09713387e-01 1.20354004e-01
7.73839712e-01 -3.64415854e-01 -1.38164563e-02 -8.37134182e-01
-4.19173479e-01 2.41972640e-01 -1.38745621e-01 -8.34944367e-01
1.56405926e-01 1.59320191e-01 -9.37143490e-02 -7.34886289e-01
5.48558772e-01 -4.44409788e-01 -1.35795504e-01 5.69793880e-01
-4.58372176e-01 3.53324145e-01 1.43443450e-01 4.22846079e-01
-3.54407519e-01 -5.17509207e-02 7.32366383e-01 -1.59649149e-01
-5.49456179e-01 2.17324883e-01 -8.48992050e-01 -1.51997298e-01
5.00380576e-01 -2.86765754e-01 -2.34715253e-01 -3.98551732e-01
-6.40538692e-01 -1.31478786e-01 -1.36966228e-01 7.14616120e-01
7.05844581e-01 -1.39706504e+00 -1.26751661e+00 3.74612004e-01
5.66739291e-02 -4.10665274e-01 1.63866296e-01 8.39198470e-01
-2.07271874e-01 1.11898899e+00 8.11145455e-02 -4.87634301e-01
-1.08170915e+00 6.25015438e-01 2.37265348e-01 -5.42586744e-02
-1.87914222e-01 1.06107211e+00 3.70327353e-01 -5.77416718e-01
1.72670394e-01 -3.69676769e-01 1.12987585e-01 -5.91712296e-01
5.26106894e-01 1.95835710e-01 4.47079629e-01 -8.27152133e-01
-4.49903220e-01 3.53344947e-01 -2.53808439e-01 -4.68701005e-01
1.03715825e+00 -4.67588395e-01 -5.02344817e-02 4.14499521e-01
1.15253663e+00 -1.69294193e-01 -9.97986257e-01 -2.81783283e-01
-1.43892571e-01 1.66867673e-02 5.29687218e-02 -1.07691038e+00
-8.47358048e-01 1.51846766e+00 4.64391798e-01 -1.54177383e-01
1.04623854e+00 -2.05769002e-01 1.30010664e+00 1.43863812e-01
1.69066772e-01 -1.05024743e+00 -1.01305738e-01 1.12073600e+00
1.02472961e+00 -1.08985245e+00 -9.44256425e-01 -1.99596629e-01
-7.60011971e-01 9.25670803e-01 5.25069952e-01 1.24565512e-01
5.50947785e-01 2.37283126e-01 3.25862579e-02 5.77327073e-01
-1.02900290e+00 -3.84368062e-01 3.36907357e-01 7.69186258e-01
7.12904871e-01 4.56312984e-01 -3.76860082e-01 6.59511745e-01
-3.74154806e-01 -1.25728250e-01 5.74918211e-01 5.06179869e-01
-5.50023198e-01 -1.22515178e+00 -1.61114603e-01 -5.83347082e-02
-6.95655942e-01 -1.01863420e+00 -3.32318902e-01 3.59564841e-01
-2.38252446e-01 1.10698152e+00 2.00393736e-01 -7.19652355e-01
2.22982448e-02 4.77709383e-01 2.61628360e-01 -7.64198661e-01
-7.45823860e-01 6.25478446e-01 1.47970080e-01 -9.98351071e-03
1.11800238e-01 -4.37676698e-01 -9.55322504e-01 -1.39272019e-01
-4.61640835e-01 4.66134474e-02 8.41445565e-01 9.28131342e-01
5.41770279e-01 7.01529443e-01 7.72182286e-01 -6.72217369e-01
-6.79560602e-01 -1.52372229e+00 -3.15958112e-01 -2.24364370e-01
4.20357704e-01 1.49710141e-02 -1.46078631e-01 8.85257199e-02] | [14.443045616149902, 6.966318607330322] |
3f98b582-4a46-4c5b-b1bb-2dfa48273c75 | informing-unsupervised-pretraining-with | 1909.02339 | null | https://arxiv.org/abs/1909.02339v2 | https://arxiv.org/pdf/1909.02339v2.pdf | Specializing Unsupervised Pretraining Models for Word-Level Semantic Similarity | Unsupervised pretraining models have been shown to facilitate a wide range of downstream NLP applications. These models, however, retain some of the limitations of traditional static word embeddings. In particular, they encode only the distributional knowledge available in raw text corpora, incorporated through language modeling objectives. In this work, we complement such distributional knowledge with external lexical knowledge, that is, we integrate the discrete knowledge on word-level semantic similarity into pretraining. To this end, we generalize the standard BERT model to a multi-task learning setting where we couple BERT's masked language modeling and next sentence prediction objectives with an auxiliary task of binary word relation classification. Our experiments suggest that our "Lexically Informed" BERT (LIBERT), specialized for the word-level semantic similarity, yields better performance than the lexically blind "vanilla" BERT on several language understanding tasks. Concretely, LIBERT outperforms BERT in 9 out of 10 tasks of the GLUE benchmark and is on a par with BERT in the remaining one. Moreover, we show consistent gains on 3 benchmarks for lexical simplification, a task where knowledge about word-level semantic similarity is paramount. | ['Goran Glavaš', 'Edoardo Maria Ponti', 'Ivan Vulić', 'Anne Lauscher', 'Anna Korhonen'] | 2019-09-05 | null | https://aclanthology.org/2020.coling-main.118 | https://aclanthology.org/2020.coling-main.118.pdf | coling-2020-8 | ['lexical-simplification'] | ['natural-language-processing'] | [ 2.44687200e-01 3.51635695e-01 -5.19055486e-01 -4.93811250e-01
-7.60487378e-01 -4.96060431e-01 5.23634315e-01 7.60500133e-01
-9.91332769e-01 6.26642764e-01 5.77662587e-01 -5.51972926e-01
-1.68498710e-01 -7.78142333e-01 -6.18713796e-01 -5.03021955e-01
1.85038224e-01 7.33208418e-01 -7.07708150e-02 -6.68415129e-01
-3.93282995e-02 3.01972657e-01 -1.23215663e+00 3.29397529e-01
9.38610911e-01 6.38294458e-01 2.24923715e-01 3.55744541e-01
-3.84981573e-01 7.87255228e-01 -2.66268343e-01 -8.43496919e-01
6.96089584e-03 5.80529571e-02 -9.63379383e-01 -6.26427114e-01
3.97178859e-01 -2.82273814e-02 -3.04456413e-01 1.07168281e+00
4.53097880e-01 5.29473245e-01 7.04125047e-01 -8.29689503e-01
-1.05528128e+00 1.15134776e+00 -2.31117085e-01 3.93464595e-01
1.09888785e-01 -2.17364505e-01 1.64642119e+00 -1.12330091e+00
5.70804715e-01 1.53407454e+00 7.93082297e-01 6.06282711e-01
-1.56024730e+00 -4.09510702e-01 5.91049850e-01 4.59637910e-01
-1.36145282e+00 -5.50231457e-01 5.34487188e-01 -3.50056708e-01
1.69416964e+00 1.12686723e-01 4.87129167e-02 1.05753434e+00
8.00582692e-02 8.88031542e-01 7.63004065e-01 -7.63076186e-01
2.51647700e-02 8.08252469e-02 7.05638349e-01 5.15851736e-01
3.93297613e-01 5.04603535e-02 -6.11881554e-01 4.80973646e-02
5.31236082e-02 -1.99858218e-01 -2.15561718e-01 -4.09889579e-01
-1.04003680e+00 1.09403074e+00 2.63198376e-01 4.04672027e-01
-1.39575541e-01 1.21836588e-01 6.21475816e-01 5.74782073e-01
8.24373841e-01 6.42566025e-01 -7.85058558e-01 7.17932265e-03
-6.84431434e-01 2.16719490e-02 8.63324046e-01 7.10031748e-01
9.16842103e-01 6.42050877e-02 -2.35798791e-01 1.05495453e+00
4.95000705e-02 3.13560486e-01 7.97391653e-01 -6.22200251e-01
5.98967552e-01 2.98003554e-01 -3.14834535e-01 -6.87048793e-01
-4.29773033e-01 -3.24150711e-01 -6.73102617e-01 -8.74291360e-02
1.90046221e-01 -7.41266012e-02 -9.19367433e-01 2.21780586e+00
4.23530862e-02 1.14744142e-01 3.74319673e-01 4.95612442e-01
9.83006895e-01 4.48851615e-01 5.43484807e-01 -1.32969067e-01
1.43139350e+00 -1.14936531e+00 -8.95896435e-01 -5.52587271e-01
1.04485130e+00 -6.52904332e-01 1.40296388e+00 2.95748740e-01
-9.49516594e-01 -4.93901014e-01 -1.09612393e+00 -7.03685045e-01
-9.41055834e-01 -1.10243298e-01 1.00314677e+00 6.00095928e-01
-9.33643103e-01 6.08572483e-01 -6.47799432e-01 -5.16388059e-01
3.37540150e-01 3.41181517e-01 -3.73253495e-01 -2.42705122e-01
-1.67299199e+00 1.41010749e+00 4.76775914e-01 -3.04443181e-01
-5.84398210e-01 -1.05641711e+00 -1.21556699e+00 2.82566756e-01
3.63299817e-01 -9.59616959e-01 1.30708814e+00 -7.71366715e-01
-1.40496588e+00 1.14360166e+00 -3.55037749e-01 -8.73789787e-01
1.43913239e-01 -6.02161407e-01 -2.36503437e-01 -2.55867034e-01
9.12390649e-02 7.38744736e-01 4.96834278e-01 -1.06556404e+00
-4.21035290e-01 -1.97924152e-01 3.74558032e-01 3.70252609e-01
-6.27284348e-01 7.07157105e-02 -1.35276064e-01 -9.23279762e-01
-4.72477019e-01 -4.43788439e-01 -2.15692401e-01 -3.02530020e-01
-3.38849127e-01 -7.24043787e-01 3.71286839e-01 -6.19757533e-01
1.27503777e+00 -2.00654221e+00 5.15702426e-01 -1.49887219e-01
2.22803444e-01 6.08669162e-01 -4.67772096e-01 5.45049965e-01
-4.03853714e-01 2.67172098e-01 -2.77887702e-01 -8.36303830e-01
3.12086582e-01 5.45904756e-01 -6.47139907e-01 1.67571262e-01
2.73093462e-01 1.17021298e+00 -8.18651199e-01 -1.63414791e-01
3.18161309e-01 2.41721928e-01 -8.25093091e-01 3.04382741e-02
-2.97554463e-01 -9.02784429e-03 -1.07282132e-01 1.96060896e-01
4.99820083e-01 1.13844857e-01 1.00905590e-01 -1.37214154e-01
1.35291860e-01 7.94814587e-01 -6.88867569e-01 1.97740149e+00
-8.45950186e-01 4.49670076e-01 -8.11056271e-02 -1.35160077e+00
5.32638788e-01 2.07266092e-01 2.78641105e-01 -6.21865273e-01
1.34797290e-01 -1.18409516e-02 2.98997581e-01 -3.25706571e-01
5.49311697e-01 -5.58891118e-01 -1.79786801e-01 4.43429559e-01
4.04551655e-01 -2.55360395e-01 2.63123780e-01 2.67413199e-01
1.12528396e+00 -2.38164291e-02 6.83281243e-01 -4.19348896e-01
5.52276671e-01 -1.17255382e-01 5.80098391e-01 7.66202092e-01
-2.65711248e-02 3.75468224e-01 5.03756940e-01 -2.54437476e-01
-7.46412516e-01 -1.31256890e+00 -9.49708521e-02 1.73009229e+00
5.48239201e-02 -7.10389674e-01 -4.86486971e-01 -6.75957918e-01
2.98073590e-01 1.31238806e+00 -6.16281688e-01 -5.46739042e-01
-7.89397001e-01 -8.39148462e-01 7.53973961e-01 6.27367318e-01
-1.37499496e-01 -1.05560434e+00 -1.99989732e-02 1.76132724e-01
1.54364053e-02 -1.22550464e+00 -3.31829101e-01 5.81580937e-01
-5.79342961e-01 -8.49898040e-01 -2.31513873e-01 -9.40549731e-01
2.41072342e-01 6.93939440e-03 1.43330097e+00 -8.40248093e-02
-2.77847856e-01 2.68775016e-01 -4.60452080e-01 -6.42646611e-01
-2.30391890e-01 3.93896282e-01 2.90129602e-01 -2.33455241e-01
8.75039995e-01 -6.43916845e-01 -7.95525461e-02 -2.96387076e-01
-9.50970054e-01 -5.90877198e-02 5.00544071e-01 1.08521032e+00
5.22379756e-01 -1.88993067e-01 7.87780046e-01 -1.28012586e+00
9.38117027e-01 -4.90724742e-01 -6.38817996e-02 2.80036896e-01
-6.36002481e-01 2.55083829e-01 8.12443316e-01 -4.54521537e-01
-1.06902957e+00 -2.98347384e-01 -5.70276916e-01 -1.27437085e-01
-2.88139749e-02 7.21382320e-01 -4.67652559e-01 1.74978897e-01
5.82478642e-01 3.19911689e-02 -2.51229823e-01 -8.42605948e-01
7.94444919e-01 4.32900786e-01 3.84658068e-01 -8.13997507e-01
7.73527443e-01 2.81147182e-01 -2.67814457e-01 -7.50125885e-01
-1.39871740e+00 -5.24923384e-01 -8.86092126e-01 7.31435478e-01
9.60122347e-01 -9.30314779e-01 -4.12930220e-01 9.85961184e-02
-1.51232779e+00 -3.96224976e-01 -6.63746238e-01 5.40410578e-01
-4.12807882e-01 4.49968815e-01 -6.95232928e-01 -3.77562940e-01
-3.46297890e-01 -8.65579903e-01 9.94175851e-01 -2.04020098e-01
-5.34289479e-01 -1.48907566e+00 1.80695876e-01 1.85231507e-01
4.05877262e-01 -3.65323365e-01 1.67096770e+00 -1.22199714e+00
-2.57242825e-02 6.52316362e-02 -2.26677909e-01 7.48642981e-01
1.15763195e-01 -5.79974651e-01 -1.12517560e+00 -1.20762572e-01
7.70844370e-02 -4.39691901e-01 1.46650422e+00 2.69200027e-01
1.14520848e+00 -9.74603668e-02 -2.84925461e-01 7.65029490e-01
1.32774651e+00 -2.17535228e-01 3.95712942e-01 1.83004931e-01
9.54809725e-01 6.99146986e-01 4.57298458e-01 1.88910458e-02
4.55759764e-01 5.77676713e-01 8.92603695e-02 -1.63505599e-01
-4.51741517e-01 -3.00886869e-01 4.40692365e-01 1.12815106e+00
1.86097428e-01 -3.80919039e-01 -9.23857689e-01 7.16897368e-01
-1.80610144e+00 -6.53907537e-01 1.16999984e-01 1.98826313e+00
1.26885760e+00 1.26962036e-01 -3.31958741e-01 -5.60072847e-02
5.40657103e-01 4.03361738e-01 -4.88199145e-01 -8.13355446e-01
-4.05882329e-01 7.83949673e-01 5.89664996e-01 9.55722332e-01
-1.22117090e+00 1.54309893e+00 5.70707083e+00 1.09035122e+00
-7.06131160e-01 5.43710351e-01 3.20033163e-01 -1.31612971e-01
-6.12587631e-01 2.57267170e-02 -9.73983109e-01 6.47518188e-02
9.17726099e-01 -4.13752049e-01 3.89874876e-01 5.51979184e-01
7.59902298e-02 8.32723230e-02 -1.65084743e+00 8.49335909e-01
3.78973335e-01 -1.16049469e+00 4.31017011e-01 -2.06348315e-01
4.58412796e-01 2.02397048e-01 5.84593080e-02 7.91523814e-01
6.52889252e-01 -1.46545279e+00 4.13099766e-01 2.74045318e-01
6.85122311e-01 -8.26451659e-01 8.18727851e-01 4.37073499e-01
-9.62260962e-01 1.83397546e-01 -5.51085234e-01 -4.16277170e-01
4.03542340e-01 6.79895520e-01 -4.65240926e-01 5.38540781e-01
2.71708757e-01 8.02054286e-01 -6.02427661e-01 5.68661094e-01
-6.31546438e-01 6.45242214e-01 -1.31899983e-01 6.95044398e-02
2.69885540e-01 5.57992570e-02 6.23868823e-01 1.51797390e+00
-1.06032148e-01 2.72550285e-02 1.38756260e-01 5.46156228e-01
-3.90810370e-01 4.51818317e-01 -7.42455900e-01 -8.42415243e-02
4.92731929e-01 9.20538127e-01 -2.32809141e-01 -4.69984382e-01
-5.96743941e-01 8.47807467e-01 9.26569641e-01 2.26310596e-01
-6.17635131e-01 -6.06029391e-01 1.13925540e+00 -2.52173990e-01
5.39001822e-01 -2.69876510e-01 -5.35192132e-01 -1.35032856e+00
-1.23174749e-01 -7.15186656e-01 4.99954343e-01 -4.01911467e-01
-1.63215923e+00 3.38542193e-01 -1.58055983e-02 -4.96869147e-01
-1.16199829e-01 -1.09383798e+00 -5.04619718e-01 9.66772676e-01
-2.04595900e+00 -1.26920390e+00 2.98233896e-01 4.43872511e-01
8.45946670e-01 -2.11245418e-01 1.24177980e+00 3.51650387e-01
-5.96306682e-01 8.24233353e-01 1.21683642e-01 1.89927518e-01
8.96653175e-01 -1.47774374e+00 6.20780587e-01 7.76412368e-01
5.84649444e-01 8.96173418e-01 6.59470916e-01 -4.94491249e-01
-1.22226477e+00 -1.15908515e+00 1.51411593e+00 -5.87261081e-01
1.06178141e+00 -6.60427034e-01 -1.09812534e+00 1.07804513e+00
3.48358542e-01 -1.19722426e-01 1.02826309e+00 7.19197512e-01
-6.87327087e-01 3.85588855e-02 -7.89484620e-01 6.82800174e-01
1.31608033e+00 -7.72175074e-01 -1.31989217e+00 4.80060339e-01
1.15253139e+00 -1.56359807e-01 -8.10418546e-01 5.04398286e-01
2.32031688e-01 -4.77616012e-01 1.23509192e+00 -1.21810150e+00
3.30979258e-01 1.99757129e-01 -4.34446752e-01 -1.65674186e+00
-3.49832237e-01 -5.05787730e-01 -9.29909572e-02 1.47303712e+00
5.15878081e-01 -7.23520279e-01 5.19985616e-01 3.05160731e-01
-3.33523601e-01 -7.18892932e-01 -9.53515887e-01 -8.78187418e-01
8.75141263e-01 -7.83896744e-01 4.24944937e-01 1.05500555e+00
7.21420869e-02 9.34616685e-01 -2.26476207e-01 -1.79213528e-02
3.27024162e-01 7.73005607e-03 4.13952023e-01 -1.34961319e+00
-2.65611857e-01 -6.20234251e-01 -3.14541042e-01 -1.28553557e+00
1.01618218e+00 -1.46954000e+00 -5.10650128e-02 -1.50421560e+00
1.49769500e-01 -3.16038430e-01 -5.37011564e-01 5.38453341e-01
-3.68662745e-01 5.56855164e-02 1.00669220e-01 -1.89923853e-01
-5.79190433e-01 8.11161757e-01 8.68827581e-01 -2.20924780e-01
2.32868567e-02 -3.55596364e-01 -1.05069625e+00 1.02286136e+00
7.05703676e-01 -5.87487638e-01 -4.92772609e-01 -1.04964948e+00
2.55950868e-01 -6.71554744e-01 1.41283199e-01 -4.64596748e-01
3.21040660e-01 -7.89468363e-02 -1.52953535e-01 -2.70838350e-01
4.86669272e-01 -5.82674801e-01 -6.68814242e-01 3.12259197e-01
-5.82126915e-01 -4.35650200e-02 2.91977227e-01 5.14176071e-01
-3.17603379e-01 -4.00867552e-01 6.37168229e-01 3.17804851e-02
-1.02391052e+00 1.31573737e-01 -2.84937143e-01 6.29197657e-01
7.77014434e-01 4.64813365e-03 -3.23137850e-01 -6.85292110e-02
-8.26100230e-01 1.23970680e-01 1.15037039e-01 6.11497223e-01
4.54701722e-01 -1.20785022e+00 -7.20659673e-01 2.99496740e-01
2.77437210e-01 -4.51075519e-03 -5.39180897e-02 8.17634523e-01
-7.51343295e-02 6.82515681e-01 2.12485105e-01 -9.02924463e-02
-1.06995177e+00 8.60785186e-01 1.92524210e-01 -5.01793981e-01
-4.94078219e-01 1.20204079e+00 5.34329355e-01 -7.27238178e-01
3.37034792e-01 -7.31636524e-01 -2.20952064e-01 3.71957868e-01
4.11532402e-01 1.42078876e-01 4.53626484e-01 -5.58515668e-01
-4.51458126e-01 5.12430012e-01 -4.14834648e-01 1.25265554e-01
1.53451169e+00 -1.92069665e-01 -2.40668416e-01 6.37960792e-01
1.31476879e+00 2.79618770e-01 -6.14831865e-01 -7.05888569e-01
5.49036026e-01 -7.54970759e-02 6.25087544e-02 -7.20649898e-01
-6.74710751e-01 1.20232058e+00 -2.44138632e-02 -1.12049587e-01
8.11273456e-01 1.01929277e-01 9.80199993e-01 8.11516404e-01
1.72343343e-01 -1.03885770e+00 -3.45644653e-01 1.05227709e+00
6.47772491e-01 -1.27871001e+00 -2.37457547e-02 -5.45159876e-01
-6.30332470e-01 8.20792794e-01 4.46327329e-01 -2.38791555e-01
8.15442502e-01 4.44937408e-01 2.96064448e-02 -2.09699981e-02
-1.02896333e+00 -5.81292987e-01 2.71883607e-01 8.09420705e-01
6.03136539e-01 6.94741458e-02 -4.72944707e-01 1.01257133e+00
-4.43789423e-01 -4.31289881e-01 1.51381880e-01 5.27925491e-01
-6.06293261e-01 -1.37358785e+00 1.85696587e-01 3.60024393e-01
-5.00651002e-01 -9.02857900e-01 -4.14872020e-01 8.57433200e-01
2.38893405e-01 8.30239654e-01 7.79424831e-02 -1.08095936e-01
4.88546699e-01 4.00898844e-01 4.14284199e-01 -1.37787759e+00
-5.66331863e-01 -4.48661447e-01 3.93998921e-01 -4.60537046e-01
-1.89436153e-01 -5.39555073e-01 -1.11928964e+00 -3.42489958e-01
3.33264358e-02 1.46890610e-01 3.20325941e-01 1.30907393e+00
1.80207938e-01 6.73369408e-01 2.42152903e-03 -5.30755579e-01
-8.19436669e-01 -1.05302405e+00 -3.86142969e-01 5.66219687e-01
3.98744047e-01 -8.07457030e-01 -3.34050596e-01 -4.42041755e-02] | [10.721789360046387, 8.916720390319824] |
a16a1ec1-c4d6-4b59-938e-ba104bd8b855 | wuyun-exploring-hierarchical-skeleton-guided | 2301.04488 | null | https://arxiv.org/abs/2301.04488v2 | https://arxiv.org/pdf/2301.04488v2.pdf | WuYun: Exploring hierarchical skeleton-guided melody generation using knowledge-enhanced deep learning | Although deep learning has revolutionized music generation, existing methods for structured melody generation follow an end-to-end left-to-right note-by-note generative paradigm and treat each note equally. Here, we present WuYun, a knowledge-enhanced deep learning architecture for improving the structure of generated melodies, which first generates the most structurally important notes to construct a melodic skeleton and subsequently infills it with dynamically decorative notes into a full-fledged melody. Specifically, we use music domain knowledge to extract melodic skeletons and employ sequence learning to reconstruct them, which serve as additional knowledge to provide auxiliary guidance for the melody generation process. We demonstrate that WuYun can generate melodies with better long-term structure and musicality and outperforms other state-of-the-art methods by 0.51 on average on all subjective evaluation metrics. Our study provides a multidisciplinary lens to design melodic hierarchical structures and bridge the gap between data-driven and knowledge-based approaches for numerous music generation tasks. | ['Lingyun Sun', 'Songruoyao Wu', 'Qihao Liang', 'Xu Tan', 'Zhijie Huang', 'Tieyao Zhang', 'Xinda Wu', 'Kejun Zhang'] | 2023-01-11 | null | null | null | null | ['music-generation', 'music-generation'] | ['audio', 'music'] | [ 2.49239177e-01 -1.32477239e-01 2.87716240e-02 1.93576366e-01
-1.03825223e+00 -9.74276781e-01 2.03772172e-01 -3.18300724e-01
1.32511988e-01 8.54030669e-01 7.17143059e-01 1.07465848e-01
-3.15403640e-01 -8.29569519e-01 -4.77558702e-01 -6.58684254e-01
2.03424394e-01 3.90768111e-01 -2.66734332e-01 -6.50457203e-01
2.15443879e-01 2.31403023e-01 -1.28764534e+00 7.59780049e-01
6.54657304e-01 8.09822321e-01 1.22458234e-01 9.53664541e-01
1.86517425e-02 8.71395051e-01 -9.54914153e-01 -5.06918728e-01
3.85696709e-01 -1.07155657e+00 -8.11994433e-01 6.42366242e-03
1.83775231e-01 -1.70694426e-01 -1.53466582e-01 5.95482826e-01
9.95803893e-01 2.74258137e-01 5.18772185e-01 -7.47992992e-01
-8.00036013e-01 1.55194438e+00 -3.11367601e-01 -3.01702827e-01
3.11519414e-01 2.11433619e-01 1.54506242e+00 -9.83791351e-01
5.74878454e-01 7.31843174e-01 1.01559675e+00 6.51710391e-01
-1.41206133e+00 -7.65611172e-01 -3.77955824e-01 8.06073248e-02
-1.16922677e+00 -3.94806981e-01 1.27637076e+00 -3.42632771e-01
6.06129706e-01 3.29911858e-01 1.16594696e+00 1.14045787e+00
-1.21847279e-01 1.00361645e+00 7.23314404e-01 -5.38140833e-01
-2.59645488e-02 -7.95557678e-01 -5.69064021e-01 3.04919124e-01
-2.52760619e-01 2.94639260e-01 -1.03461444e+00 1.08631730e-01
1.17677402e+00 -4.59397167e-01 -1.36117026e-01 -1.50075117e-02
-1.40838253e+00 6.71811283e-01 3.59174401e-01 3.64741206e-01
-6.87788129e-01 1.79675207e-01 5.60634553e-01 1.56803846e-01
-9.04235020e-02 1.35150003e+00 -3.12122196e-01 -5.06244898e-01
-1.45083296e+00 7.44237244e-01 7.12485671e-01 5.45588315e-01
3.49179119e-01 7.96467185e-01 -7.17732430e-01 1.14279521e+00
-1.40439793e-01 2.30564922e-01 6.61005795e-01 -1.26671934e+00
2.85546690e-01 3.39847982e-01 3.77161913e-02 -5.45705736e-01
-1.57165617e-01 -9.11992073e-01 -9.86519814e-01 3.48223537e-01
3.42695624e-01 -3.39426219e-01 -7.68051147e-01 1.74872148e+00
-1.57221481e-01 2.68512249e-01 2.92400643e-02 8.89835298e-01
8.64524782e-01 6.97370708e-01 -1.78568929e-01 -2.27877632e-01
1.13838458e+00 -1.19541097e+00 -6.94917917e-01 2.48000130e-01
-6.29891530e-02 -1.13135827e+00 1.39068747e+00 7.20750511e-01
-1.74390900e+00 -1.12757385e+00 -1.23507261e+00 -1.37002289e-01
4.66581702e-01 3.11376721e-01 4.37783301e-01 3.86466324e-01
-7.49663651e-01 1.09030366e+00 -4.71582055e-01 5.38608253e-01
4.34444934e-01 1.45600475e-02 1.35043561e-01 6.54971302e-01
-1.23470092e+00 4.00799155e-01 7.19146609e-01 5.02003729e-02
-1.13310730e+00 -1.13983703e+00 -5.53006053e-01 1.79713160e-01
7.00707957e-02 -1.08937669e+00 1.68182766e+00 -8.25369835e-01
-1.98647785e+00 6.43066645e-01 2.29067981e-01 -6.08538210e-01
3.68684083e-01 -3.48798603e-01 -4.07183349e-01 -1.00293562e-01
-1.19514540e-02 7.34544814e-01 9.69280481e-01 -1.33423138e+00
-5.83047390e-01 2.47954860e-01 -1.41294748e-01 3.99458766e-01
-2.69267321e-01 -5.28629661e-01 -3.60958099e-01 -1.61841786e+00
-9.72566530e-02 -9.71016228e-01 -7.43964761e-02 -5.87195933e-01
-6.73011005e-01 9.33153778e-02 4.19228703e-01 -8.31299126e-01
1.90989769e+00 -1.96254635e+00 5.05077600e-01 4.29537557e-02
1.47105232e-01 2.70342529e-01 -3.51471156e-01 6.64376259e-01
-3.09659839e-01 -2.81497419e-01 -2.61640310e-01 -3.86691570e-01
1.60831466e-01 -8.92600492e-02 -7.89361537e-01 -4.73550975e-01
-6.76647797e-02 1.27045488e+00 -9.50750411e-01 -1.75477728e-01
2.01948844e-02 4.00585026e-01 -8.14535558e-01 2.60320783e-01
-4.34362650e-01 7.98827052e-01 1.12459501e-02 5.33874571e-01
1.44854993e-01 -3.50760892e-02 2.18718693e-01 -5.47321796e-01
-8.70005414e-02 4.64978188e-01 -9.78383780e-01 2.25687742e+00
-5.45100510e-01 6.41113162e-01 -3.19692373e-01 -4.84666586e-01
1.22251558e+00 5.17476916e-01 5.38299918e-01 -5.32061398e-01
-2.06678137e-02 3.48831356e-01 2.13077798e-01 -9.35849771e-02
8.22559595e-01 -6.51775301e-01 -3.08947623e-01 5.79874873e-01
3.81438658e-02 -5.11659086e-01 2.67124653e-01 -1.90075591e-01
8.99027169e-01 4.93204594e-01 3.87098074e-01 3.34029607e-02
3.70236099e-01 -5.26397601e-02 5.17351627e-01 6.42538667e-01
3.82185668e-01 1.01402891e+00 1.70942485e-01 -5.02132356e-01
-1.46861899e+00 -1.62432706e+00 4.51118141e-01 1.34139550e+00
-3.39522392e-01 -6.53214335e-01 -9.35921013e-01 -4.97897454e-02
-3.22959602e-01 7.69052446e-01 -4.96123910e-01 -3.69388431e-01
-8.93628955e-01 -3.66985112e-01 1.12539995e+00 8.40165794e-01
3.40465218e-01 -1.93071616e+00 -3.47579598e-01 7.00196922e-01
-6.17666543e-01 -2.93551385e-01 -7.91019201e-01 -1.31660979e-02
-8.62564087e-01 -7.85202742e-01 -8.86991262e-01 -1.01273835e+00
-4.35735583e-02 -4.41089571e-01 1.60416245e+00 -2.16133401e-01
-3.19387406e-01 -1.80021316e-01 -3.68840337e-01 -4.54860747e-01
-7.46560276e-01 3.19680542e-01 -3.28505412e-02 -1.13140680e-01
-2.31971532e-01 -1.09024596e+00 -8.47546279e-01 5.45506515e-02
-9.45773304e-01 3.67236376e-01 6.59040332e-01 9.37479496e-01
9.44051862e-01 1.37690231e-01 1.02589428e+00 -5.84234297e-01
1.21005666e+00 1.33113265e-01 -1.30420044e-01 -3.20964754e-02
-2.47617304e-01 1.01217590e-01 8.88751388e-01 -5.55739045e-01
-1.01356101e+00 3.51270698e-02 -4.72240150e-01 -4.94647473e-01
1.38783932e-01 6.56008959e-01 -1.90622043e-02 5.34355521e-01
9.44892049e-01 5.02719462e-01 -3.35547477e-01 -5.97011566e-01
7.22490132e-01 2.68727452e-01 1.23464191e+00 -8.66335511e-01
8.44715536e-01 1.63660303e-01 -1.06190048e-01 -4.80933487e-01
-1.25605714e+00 3.98826506e-03 -7.48380899e-01 -1.86873168e-01
7.47608900e-01 -8.86516809e-01 -7.24695146e-01 6.10765636e-01
-1.00106645e+00 -6.27664030e-01 -1.15560436e+00 1.16150297e-01
-1.09907377e+00 8.56677517e-02 -9.44570124e-01 -5.41080058e-01
-9.62905645e-01 -6.08298004e-01 9.93242085e-01 6.92084357e-02
-7.70785332e-01 -6.69057429e-01 5.24439991e-01 3.27367157e-01
2.05667302e-01 5.27399123e-01 1.02351189e+00 -1.04053028e-01
-3.47658694e-01 9.08119082e-02 3.00308436e-01 4.64912653e-01
3.50562394e-01 -1.40013576e-01 -8.14458907e-01 -1.38347909e-01
-2.94174582e-01 -4.80805516e-01 8.92324924e-01 4.90586132e-01
1.14183617e+00 -3.35301250e-01 3.32713932e-01 9.30221796e-01
9.72959757e-01 2.92385906e-01 8.43770027e-01 1.95688590e-01
8.56177568e-01 2.64996499e-01 3.74097168e-01 8.57286632e-01
8.22418854e-02 7.15773702e-01 3.13222073e-02 -3.92091125e-02
-6.65293694e-01 -7.42263913e-01 3.08886617e-01 1.37410259e+00
-5.97237289e-01 -8.83431509e-02 -5.01066685e-01 7.24937081e-01
-1.78161681e+00 -1.39731681e+00 1.44290477e-01 1.97774589e+00
1.48402059e+00 8.84736627e-02 5.72583377e-01 7.05411434e-01
5.70224822e-01 2.09870070e-01 -6.79484546e-01 -1.48258612e-01
-1.90359920e-01 9.55073893e-01 -2.96887010e-01 9.09565240e-02
-8.87316406e-01 1.16703343e+00 6.71327305e+00 1.17924297e+00
-1.09593832e+00 -2.79404938e-01 2.40673929e-01 -3.62787306e-01
-4.38554287e-01 -2.08600029e-01 -4.82579559e-01 2.26772562e-01
6.43179536e-01 -3.42451900e-01 8.84097397e-01 6.92660511e-01
2.68708795e-01 6.98933005e-01 -1.06818938e+00 1.15722215e+00
-1.71311930e-01 -1.97574067e+00 4.46120352e-01 -1.73716173e-01
1.28555512e+00 -5.14230490e-01 2.68275887e-01 4.55002874e-01
3.20332736e-01 -1.28318083e+00 1.07297957e+00 7.82992721e-01
1.07869840e+00 -1.14799678e+00 2.15017319e-01 -3.42420116e-03
-1.64993131e+00 -5.49034141e-02 6.04824312e-02 -1.84166208e-01
4.86107945e-01 3.15637022e-01 -9.28629220e-01 6.46326005e-01
3.28036636e-01 6.83970332e-01 -2.12081552e-01 1.18719387e+00
-4.52732652e-01 9.71520483e-01 2.06753403e-01 3.83863926e-01
1.86997131e-01 -1.02151014e-01 6.55487359e-01 1.09233141e+00
6.52108848e-01 -5.67635782e-02 1.90525264e-01 1.11555421e+00
-3.42335790e-01 -3.16707604e-03 -8.24613124e-02 -3.01245540e-01
7.11497128e-01 1.18153512e+00 -5.64854264e-01 -2.32636243e-01
3.30496728e-01 1.13367963e+00 3.45762931e-02 2.61521578e-01
-8.64730239e-01 -6.34159327e-01 5.07398665e-01 1.05462208e-01
5.02017558e-01 -1.48324981e-01 -5.95734596e-01 -8.41432035e-01
-2.41425499e-01 -1.24897933e+00 2.58332729e-01 -1.00105023e+00
-1.36496663e+00 8.03661287e-01 -4.57084596e-01 -1.57194793e+00
-8.76987219e-01 -1.22028746e-01 -8.39656353e-01 8.54295492e-01
-8.28916728e-01 -1.40477419e+00 -2.00166285e-01 5.75094283e-01
8.57383013e-01 -4.66762185e-01 1.11115515e+00 1.87132850e-01
1.14415564e-01 7.18290508e-01 -5.41313402e-02 4.83813763e-01
6.51595473e-01 -1.36542118e+00 6.99036241e-01 5.78908741e-01
8.51115942e-01 5.26579738e-01 4.46940988e-01 -6.21017575e-01
-8.61656308e-01 -9.91076946e-01 5.82350731e-01 -4.24211621e-01
5.77040553e-01 -1.58317536e-01 -6.68931484e-01 2.78080970e-01
3.70621026e-01 -5.69236159e-01 9.75281060e-01 2.05946252e-01
-3.26749384e-01 -1.43323436e-01 -4.53022569e-01 9.74251390e-01
1.27087891e+00 -6.17104828e-01 -9.62337017e-01 -2.56168872e-01
8.55800867e-01 -5.24926603e-01 -1.00922620e+00 4.68311727e-01
1.01896238e+00 -1.10659289e+00 1.16574025e+00 -5.36727667e-01
8.44272971e-01 -5.89114845e-01 4.63137180e-02 -1.65542519e+00
-7.05071568e-01 -1.24954152e+00 -4.11435694e-01 1.20097780e+00
5.66406250e-01 4.13791507e-01 9.42971885e-01 -4.78149056e-01
-5.49148917e-01 -6.73937976e-01 -5.10875404e-01 -6.56061709e-01
1.68923438e-02 -5.78972280e-01 8.57784927e-01 8.40772808e-01
-2.96482027e-01 6.74647331e-01 -8.44984353e-01 -3.97171050e-01
5.44466436e-01 6.86533570e-01 8.29577029e-01 -1.18564641e+00
-7.58303225e-01 -8.19475412e-01 1.27959579e-01 -9.86201584e-01
-1.01538762e-01 -9.35249686e-01 3.98515612e-02 -1.47010183e+00
-2.90591433e-03 -2.03617364e-01 -5.71628869e-01 4.39385295e-01
-1.53167725e-01 8.24513495e-01 6.24172509e-01 3.34250599e-01
-2.94394046e-01 7.96703041e-01 1.73460472e+00 -1.01313666e-01
-7.73527324e-01 2.30373174e-01 -9.99225736e-01 7.23923028e-01
6.75001442e-01 -1.10768393e-01 -5.42237997e-01 -2.74484575e-01
3.69505227e-01 2.97330856e-01 1.66761830e-01 -1.39855468e+00
4.94127795e-02 -1.22654317e-02 6.62464440e-01 -9.17820334e-01
4.85644996e-01 -1.91696793e-01 4.37882215e-01 3.89336795e-01
-6.79055095e-01 1.53562927e-04 2.45518491e-01 3.04946393e-01
-4.65789795e-01 -1.55206859e-01 7.12804973e-01 -2.28138298e-01
-5.07057011e-01 2.13254213e-01 -8.66241306e-02 3.10517579e-01
4.80794907e-01 -2.66066909e-01 1.28347903e-01 -5.30109406e-01
-1.14765978e+00 -4.30314124e-01 3.30435857e-02 4.64757442e-01
5.82606137e-01 -1.91236162e+00 -1.04172587e+00 1.85889229e-01
-1.25179719e-02 -6.23609088e-02 4.66325790e-01 2.83283204e-01
-6.14018917e-01 9.29601863e-02 -5.46944797e-01 -1.84103444e-01
-1.10178924e+00 2.57008463e-01 9.61500406e-02 -5.47607303e-01
-6.86575711e-01 1.02213919e+00 1.40373677e-01 -3.60339224e-01
1.57489330e-01 -2.65884638e-01 -1.63166225e-01 2.72043198e-01
5.01529813e-01 4.12377983e-01 -1.57198146e-01 -2.79002935e-01
1.20438173e-01 5.81371546e-01 2.70875514e-01 -5.66236079e-01
1.48772013e+00 2.33144954e-01 1.63253620e-02 5.87015688e-01
5.16689718e-01 5.70964098e-01 -1.53760684e+00 1.39555372e-02
-9.45795700e-02 -1.43785939e-01 -2.86380559e-01 -1.27011168e+00
-9.07615185e-01 6.45968139e-01 5.26285358e-02 1.15423545e-01
1.45030332e+00 -2.65034199e-01 1.40220928e+00 3.16546112e-01
1.89908385e-01 -1.14875078e+00 8.96790981e-01 7.04421937e-01
1.37600791e+00 -3.53042185e-01 -2.49884322e-01 1.06474847e-01
-1.04197526e+00 1.15187407e+00 4.38643306e-01 -4.05585438e-01
2.93956697e-01 3.37269127e-01 1.69947743e-01 1.30196467e-01
-6.14002168e-01 -2.45647311e-01 7.29098260e-01 5.92554927e-01
7.61038601e-01 1.31326318e-01 -4.46809158e-02 1.11395454e+00
-1.21106863e+00 9.38030258e-02 1.31052777e-01 3.60134006e-01
-3.57979059e-01 -1.56136453e+00 -3.31627220e-01 8.09760466e-02
-3.68094683e-01 -6.27914310e-01 -5.18544257e-01 5.05009413e-01
3.87793273e-01 6.47757351e-01 -1.05349988e-01 -6.23405457e-01
4.71186072e-01 2.48457357e-01 7.74355888e-01 -5.60606956e-01
-1.02867138e+00 6.54330552e-01 1.26519114e-01 -2.07022786e-01
-2.42345572e-01 -3.16865206e-01 -1.26011980e+00 -1.46049693e-01
8.96484256e-02 1.55509830e-01 2.11013988e-01 5.14641821e-01
4.21451151e-01 1.27823913e+00 6.09132588e-01 -1.26834643e+00
-4.66051519e-01 -1.05872405e+00 -5.41873276e-01 5.97607970e-01
1.52458295e-01 -1.45521134e-01 1.87661678e-01 5.18015206e-01] | [16.047391891479492, 5.5526227951049805] |
6e44391d-1b81-45de-a00b-4df9a1bff6f0 | han-ecg-an-interpretable-atrial-fibrillation | 2002.05262 | null | https://arxiv.org/abs/2002.05262v1 | https://arxiv.org/pdf/2002.05262v1.pdf | HAN-ECG: An Interpretable Atrial Fibrillation Detection Model Using Hierarchical Attention Networks | Atrial fibrillation (AF) is one of the most prevalent cardiac arrhythmias that affects the lives of more than 3 million people in the U.S. and over 33 million people around the world and is associated with a five-fold increased risk of stroke and mortality. like other problems in healthcare domain, artificial intelligence (AI)-based algorithms have been used to reliably detect AF from patients' physiological signals. The cardiologist level performance in detecting this arrhythmia is often achieved by deep learning-based methods, however, they suffer from the lack of interpretability. In other words, these approaches are unable to explain the reasons behind their decisions. The lack of interpretability is a common challenge toward a wide application of machine learning-based approaches in the healthcare which limits the trust of clinicians in such methods. To address this challenge, we propose HAN-ECG, an interpretable bidirectional-recurrent-neural-network-based approach for the AF detection task. The HAN-ECG employs three attention mechanism levels to provide a multi-resolution analysis of the patterns in ECG leading to AF. The first level, wave level, computes the wave weights, the second level, heartbeat level, calculates the heartbeat weights, and third level, window (i.e., multiple heartbeats) level, produces the window weights in triggering a class of interest. The detected patterns by this hierarchical attention model facilitate the interpretation of the neural network decision process in identifying the patterns in the signal which contributed the most to the final prediction. Experimental results on two AF databases demonstrate that our proposed model performs significantly better than the existing algorithms. Visualization of these attention layers illustrates that our model decides upon the important waves and heartbeats which are clinically meaningful in the detection task. | ['U. Rajendra Acharya', 'Fatemeh Afghah', 'Sajad Mousavi'] | 2020-02-12 | null | null | null | null | ['atrial-fibrillation-detection'] | ['medical'] | [ 2.95031995e-01 -5.90221770e-02 4.00551967e-02 -9.95863155e-02
-4.91076142e-01 -2.34011665e-01 -9.41813886e-02 1.79047719e-01
8.81398842e-02 7.91437984e-01 3.80523086e-01 -5.06512821e-01
-3.46766084e-01 -5.97764790e-01 -4.66083316e-03 -7.14279234e-01
-3.40250939e-01 1.45551980e-01 -2.26523787e-01 -2.10944843e-02
2.37759903e-01 4.43018705e-01 -1.18132091e+00 6.25583470e-01
1.05874741e+00 1.27504730e+00 -1.76699504e-01 6.31185651e-01
3.56599726e-02 7.46432781e-01 -7.60043085e-01 2.30245605e-01
-1.38733089e-01 -9.11882162e-01 -5.25489509e-01 -3.41999769e-01
-2.71683097e-01 -7.66776409e-03 3.42429802e-02 7.72442877e-01
8.92817378e-01 -5.00699937e-01 5.21632135e-01 -9.37395275e-01
-3.54538232e-01 6.25311792e-01 -4.65237349e-01 8.87845695e-01
2.09807962e-01 6.62255809e-02 5.93697488e-01 -1.09808075e+00
1.65730536e-01 9.86919641e-01 9.50056493e-01 3.71766359e-01
-8.02079380e-01 -6.23768568e-01 -7.18750730e-02 2.09940970e-01
-1.36793637e+00 -2.75088310e-01 9.88689899e-01 -5.74295700e-01
7.04538345e-01 5.01049221e-01 9.81687367e-01 8.88628602e-01
5.09094954e-01 3.71107966e-01 9.53919530e-01 -2.94127375e-01
-3.40223908e-02 -1.49059802e-01 4.74833578e-01 5.91061056e-01
2.54337609e-01 7.62258619e-02 -4.76409316e-01 -4.96083111e-01
8.09025764e-01 4.17573422e-01 -7.02624500e-01 3.66518825e-01
-1.46493876e+00 6.71409905e-01 4.65218693e-01 5.65160513e-01
-7.48519659e-01 1.15291728e-02 4.38502371e-01 1.71492413e-01
2.76483536e-01 5.27789235e-01 -4.77708727e-01 -1.63636476e-01
-7.99057901e-01 -3.85314934e-02 4.69672680e-01 2.42529258e-01
2.36025393e-01 3.88401181e-01 -4.62865025e-01 4.43067014e-01
3.36068571e-01 3.40260893e-01 7.96473801e-01 -5.89694142e-01
2.58020252e-01 9.85343456e-01 1.37209654e-01 -1.18308580e+00
-6.81437910e-01 -1.12583792e+00 -1.52658832e+00 1.64408118e-01
1.22572675e-01 -3.95932257e-01 -8.40053737e-01 1.48744440e+00
-1.02340197e-02 3.58230531e-01 1.34815395e-01 1.02557635e+00
9.19916928e-01 5.47992527e-01 1.31026953e-01 -6.85694337e-01
1.60361767e+00 -4.39647973e-01 -9.29806411e-01 -1.37645781e-01
1.91929698e-01 -5.23690641e-01 8.58433664e-01 4.51674104e-01
-8.87506425e-01 -7.68166840e-01 -1.10794961e+00 3.81840914e-01
8.32383428e-03 3.82434964e-01 4.11269575e-01 3.77242774e-01
-8.03033948e-01 5.49352825e-01 -7.29898274e-01 -2.55103320e-01
6.06946528e-01 2.50374407e-01 8.49192888e-02 5.26903987e-01
-1.44229245e+00 8.02964747e-01 2.10533902e-01 7.17778563e-01
-6.93711877e-01 -6.43250465e-01 -4.66948718e-01 1.88226342e-01
-2.32956216e-01 -9.07906175e-01 6.72357917e-01 -1.13591993e+00
-8.96558464e-01 6.69676661e-01 -4.03128237e-01 -5.44964433e-01
3.78612489e-01 -3.67810100e-01 -6.62885368e-01 -1.82244033e-02
5.24173863e-03 1.05834022e-01 8.44661593e-01 -9.60816324e-01
-6.17795527e-01 -4.31876928e-01 -4.02501911e-01 2.51229823e-01
-2.35047176e-01 -2.60738153e-02 1.46704931e-02 -9.75612462e-01
3.06866199e-01 -6.45662308e-01 -3.28742146e-01 -1.69982463e-01
-4.09643471e-01 -2.04385743e-01 8.41140330e-01 -8.16940367e-01
1.92103231e+00 -2.39638996e+00 1.90153390e-01 3.18383574e-01
6.71953022e-01 3.61776412e-01 4.90945876e-01 1.51954249e-01
-2.58184344e-01 3.75908554e-01 -4.11871076e-01 2.70176142e-01
-7.43667006e-01 1.61306486e-02 -4.40722227e-01 2.29784131e-01
2.91451126e-01 7.46475399e-01 -6.71342254e-01 -3.14088225e-01
5.93194403e-02 6.72640741e-01 -6.98951259e-02 4.40144509e-01
2.77716786e-01 8.03028047e-01 -7.09936798e-01 4.24462020e-01
1.84323162e-01 -4.53805625e-01 8.30986351e-02 -4.73225832e-01
-1.49275903e-02 7.89349303e-02 -1.02448118e+00 1.45875347e+00
-5.37593774e-02 4.02086347e-01 -3.84844363e-01 -1.08766258e+00
1.00826764e+00 8.62628102e-01 5.03010929e-01 -2.32661724e-01
1.39621019e-01 3.02876055e-01 5.56101680e-01 -8.45783710e-01
-4.08795506e-01 -1.20636903e-01 1.69388995e-01 4.63629842e-01
-5.21098912e-01 5.32504320e-01 -2.63895154e-01 -1.53664187e-01
1.06821966e+00 -1.96906433e-01 6.28152192e-01 -3.55667979e-01
7.57783353e-01 -1.92963436e-01 1.08034945e+00 7.00612307e-01
-2.73895830e-01 7.75047243e-01 3.67891282e-01 -1.28189695e+00
-5.59962749e-01 -8.33779752e-01 -2.04894930e-01 3.23314369e-01
-1.85890310e-02 -2.86491513e-01 -6.51515245e-01 -4.46655810e-01
-2.17022792e-01 2.55888999e-01 -7.65982985e-01 -3.44496250e-01
-6.42839730e-01 -1.06507158e+00 6.72461092e-01 6.93369269e-01
7.04736888e-01 -1.66942155e+00 -1.20004439e+00 4.76963729e-01
-5.24119675e-01 -6.49915934e-01 -2.23971948e-01 6.60994574e-02
-1.09520793e+00 -1.11881101e+00 -7.07619607e-01 -7.12835193e-01
5.51719069e-01 -3.45869541e-01 1.14613152e+00 4.94850159e-01
-7.52623916e-01 -2.05557317e-01 -1.46323219e-01 -8.38227093e-01
-1.99335456e-01 -6.50461614e-02 1.31345034e-01 4.77831990e-01
3.93187702e-01 -7.29649842e-01 -1.13211846e+00 3.35263647e-02
-3.28621268e-01 1.20414495e-01 8.24750781e-01 7.66674280e-01
5.97728372e-01 -6.29466325e-02 1.19552624e+00 -9.09326613e-01
1.08058512e+00 -5.43251395e-01 5.98487034e-02 1.60436600e-01
-8.12720001e-01 -1.35264680e-01 7.34191358e-01 -2.83231437e-01
-5.52296996e-01 -6.57864138e-02 5.18216528e-02 -3.40885848e-01
-1.88541174e-01 7.38336027e-01 -3.63783538e-02 3.87021393e-01
8.03442895e-01 3.97939801e-01 -9.34356675e-02 -1.39236227e-01
-2.39739403e-01 8.43402565e-01 5.53069353e-01 -2.47605160e-01
5.17906964e-01 1.69630602e-01 -9.90744457e-02 -5.59178293e-01
-7.60262549e-01 -3.01778227e-01 -4.16953743e-01 -3.57802123e-01
1.16088116e+00 -8.37007284e-01 -6.37436092e-01 3.06375384e-01
-1.33024359e+00 3.51330727e-01 -7.60608837e-02 4.54674155e-01
-1.94538131e-01 1.29761904e-01 -4.27853346e-01 -1.19497216e+00
-1.24392974e+00 -8.71284187e-01 7.23214328e-01 3.37993503e-01
-6.52325451e-01 -6.46456838e-01 -8.37212950e-02 7.48273879e-02
5.04734874e-01 7.41777062e-01 1.24639213e+00 -5.03660500e-01
-2.46649057e-01 5.03787659e-02 -1.11086555e-01 5.35775051e-02
5.07634699e-01 -2.38948241e-01 -1.09614933e+00 -1.04510218e-01
2.69119382e-01 2.55992025e-01 6.20314360e-01 6.56033754e-01
1.27197909e+00 -2.27436900e-01 -4.26060170e-01 5.25281489e-01
1.07033360e+00 8.04830730e-01 6.43639028e-01 -3.19501907e-02
6.72772050e-01 8.47089216e-02 3.47983152e-01 4.27956462e-01
1.29203692e-01 2.77605385e-01 2.92080551e-01 -7.39005804e-01
-6.43273294e-02 1.17310725e-01 1.29349977e-01 8.42297912e-01
-7.06433356e-01 5.52591607e-02 -1.08227754e+00 5.11366129e-01
-2.08276010e+00 -1.03865790e+00 -2.98426747e-01 2.18006635e+00
8.40865493e-01 2.34145135e-01 -3.90603673e-03 7.59699821e-01
7.68036067e-01 -1.97948024e-01 -7.98699796e-01 -4.48011994e-01
-9.15853083e-02 2.68348128e-01 -3.46805274e-01 1.72610953e-01
-9.95994151e-01 3.26375782e-01 5.81328392e+00 1.04812630e-01
-1.29150152e+00 4.55185911e-03 1.08694911e+00 2.66989797e-01
-9.38489661e-03 -3.26768190e-01 -3.18409175e-01 6.00854993e-01
7.67116189e-01 -1.48943216e-01 1.93320110e-01 4.16690022e-01
8.04333925e-01 3.80882442e-01 -1.14898992e+00 1.28619719e+00
-3.53910327e-02 -1.29984033e+00 1.76351115e-01 -2.10107446e-01
2.11767271e-01 -4.03625935e-01 -1.99508712e-01 -6.81305155e-02
-4.06639934e-01 -1.39418662e+00 4.05527860e-01 9.04958129e-01
9.05794859e-01 -7.77744114e-01 9.57990527e-01 3.22341055e-01
-1.33488405e+00 -3.97499561e-01 -4.50660437e-02 -1.85156018e-01
1.09886624e-01 9.45595622e-01 -6.88209891e-01 4.80600923e-01
9.90333080e-01 8.17052186e-01 -3.11550945e-01 9.18498874e-01
-1.73644558e-01 1.05438626e+00 -1.09331384e-01 1.94798842e-01
-2.65618891e-01 -7.70931244e-02 6.19099557e-01 1.06672072e+00
5.48420906e-01 4.54158336e-01 1.96017608e-01 8.85493577e-01
1.83568597e-01 1.55325253e-02 -6.85924351e-01 2.05858111e-01
3.47467899e-01 1.20427108e+00 -6.72165811e-01 -4.04124677e-01
-6.12750687e-02 7.71383584e-01 -2.22516060e-02 4.13728386e-01
-8.00805688e-01 -5.18556535e-01 3.52781266e-01 3.05646569e-01
-3.10993910e-01 2.71636635e-01 -7.25618899e-01 -9.14936185e-01
2.30698630e-01 -1.05892611e+00 6.06560826e-01 -6.57466054e-01
-1.07540095e+00 1.10012472e+00 -5.86820781e-01 -1.38794672e+00
-2.49772027e-01 -1.39059603e-01 -1.06879520e+00 1.30558634e+00
-1.31123078e+00 -8.50396931e-01 -5.67348123e-01 6.64585650e-01
6.83761835e-01 -1.67423993e-01 1.29183733e+00 4.11908954e-01
-6.06428027e-01 2.69752622e-01 -7.33700454e-01 3.31738293e-01
3.84199649e-01 -1.23793304e+00 1.83636516e-01 8.66654038e-01
8.72149095e-02 7.22111225e-01 5.02242804e-01 -5.66363811e-01
-9.23726976e-01 -9.87844944e-01 1.12537217e+00 -3.12503397e-01
4.15032879e-02 1.06832907e-01 -9.51364398e-01 2.73557484e-01
1.08887546e-01 1.50877729e-01 7.84650981e-01 -4.33420874e-02
2.14256108e-01 -4.11292434e-01 -8.67529452e-01 3.33477527e-01
7.77378082e-01 -2.42839485e-01 -7.77200937e-01 2.99954545e-02
3.03698003e-01 -1.77548155e-01 -8.70717525e-01 7.73509502e-01
8.03466260e-01 -9.73832369e-01 9.28518057e-01 -7.28950441e-01
3.08662623e-01 -6.27981305e-01 4.55901414e-01 -1.23489714e+00
-5.87322056e-01 -6.64068043e-01 -3.39096755e-01 1.01913154e+00
6.79467022e-01 -7.41133690e-01 4.46985811e-01 2.41224229e-01
-1.51700139e-01 -1.21474278e+00 -7.10025489e-01 5.56527823e-02
-3.21338356e-01 -1.06368855e-01 4.65405524e-01 9.94919062e-01
-5.98170049e-02 4.41600651e-01 -3.71596515e-01 3.43053550e-01
4.87095237e-01 3.04947227e-01 4.55081463e-02 -1.69424248e+00
-1.29249945e-01 -3.79628986e-01 -4.15903449e-01 -2.73720175e-01
-4.72238958e-01 -7.92042613e-01 -2.73562342e-01 -1.71841049e+00
1.15474634e-01 -5.29309750e-01 -8.39037061e-01 6.13718152e-01
-5.54922163e-01 2.26843148e-01 -1.48673505e-01 3.95149261e-01
1.01999743e-02 -4.90689557e-03 9.80367780e-01 -1.77893713e-01
-4.69281167e-01 2.14954957e-01 -9.24142361e-01 8.75502825e-01
9.27590251e-01 -3.35085481e-01 -4.90485221e-01 -2.75957674e-01
2.35633552e-01 3.33017826e-01 2.00414777e-01 -1.22462094e+00
9.75646302e-02 1.80318266e-01 1.02054346e+00 -3.57510775e-01
-8.31994712e-02 -7.39880621e-01 3.74317974e-01 9.93063331e-01
-2.54860699e-01 4.38580155e-01 1.08940322e-02 3.75640273e-01
-3.75895113e-01 3.91732395e-01 6.19054556e-01 -2.25830555e-01
-1.79864854e-01 4.20675486e-01 -5.58661938e-01 2.29126975e-01
8.34502399e-01 -2.73789048e-01 7.39173684e-03 -3.27574462e-01
-1.04936302e+00 6.09274171e-02 -5.15982926e-01 5.28927326e-01
9.22920525e-01 -1.30802178e+00 -1.04612613e+00 4.67302948e-01
5.21007851e-02 1.41777560e-01 9.36718211e-02 1.13428664e+00
-5.34464955e-01 1.20379373e-01 -1.39766216e-01 -6.97586417e-01
-1.27790582e+00 2.01363921e-01 7.03463852e-01 -4.04284418e-01
-1.01356947e+00 5.38778186e-01 1.23955153e-01 4.49424863e-01
3.12065661e-01 -5.04731894e-01 -8.45719337e-01 -4.39051725e-02
6.61302209e-01 2.03959897e-01 3.70762534e-02 -4.40928876e-01
-6.08739078e-01 7.91207790e-01 1.97318286e-01 3.10430378e-01
1.13768268e+00 1.97301894e-01 -2.66855329e-01 6.41327918e-01
4.42683280e-01 -1.83964625e-01 -5.87730646e-01 2.62716889e-01
-4.30439413e-02 1.00629777e-01 -4.34925593e-02 -1.19170249e+00
-1.20572424e+00 1.23478901e+00 1.06519759e+00 3.02781224e-01
1.41617346e+00 -4.93536413e-01 8.15439820e-01 3.35472226e-02
1.98671088e-01 -5.22400796e-01 -2.74785250e-01 -1.19085506e-01
1.07696259e+00 -9.10616457e-01 -1.60688952e-01 -1.90844253e-01
-4.98497576e-01 1.42921615e+00 4.34761614e-01 -4.63499346e-05
9.17947829e-01 2.92683601e-01 5.96969008e-01 -3.36624265e-01
-5.20120680e-01 2.16542512e-01 3.57518882e-01 3.91136914e-01
7.28012145e-01 9.18680429e-02 -5.50287604e-01 1.14001858e+00
1.50649836e-02 3.25206459e-01 1.45276576e-01 6.92713261e-01
-4.04352427e-01 -8.24205339e-01 -4.24810082e-01 7.86142766e-01
-8.51189852e-01 -2.47580916e-01 -3.87703300e-01 1.52279899e-01
5.43486595e-01 9.98295546e-01 -7.12656900e-02 -2.88902581e-01
1.86578065e-01 3.13693881e-01 -1.28283665e-01 -5.21371484e-01
-9.52042997e-01 1.25732332e-01 -6.14573024e-02 -3.55067819e-01
-2.98822820e-01 -2.89692551e-01 -1.61181450e+00 3.39822680e-01
-2.90183648e-02 3.87763500e-01 1.44732907e-01 9.11669910e-01
7.63189495e-01 9.46387053e-01 5.78701854e-01 -3.59270096e-01
-1.16143785e-01 -1.12053776e+00 -2.73337454e-01 5.04840314e-01
6.48779452e-01 -3.88383359e-01 -4.38786805e-01 3.13743293e-01] | [14.316999435424805, 3.2947115898132324] |
dade9dee-3afd-4494-91ac-d4c6c392bd2c | iranian-modal-music-dastgah-detection-using | 2203.15335 | null | https://arxiv.org/abs/2203.15335v3 | https://arxiv.org/pdf/2203.15335v3.pdf | Iranian Modal Music (Dastgah) detection using deep neural networks | Music classification and genre detection are topics in music information retrieval (MIR) that many articles have been published regarding their utilities in the modern world. However, this contribution is insufficient in non-western music, such as Iranian modal music. In this work, we have implemented several deep neural networks to recognize Iranian modal music in seven highly correlated categories. The best model, BiLGNet, which achieved 92 percent overall accuracy, uses an architecture inspired by autoencoders, including bidirectional LSTM and GRU layers. We trained the models using the Nava dataset, which includes 1786 records and up to 55 hours of music played solo by Kamanche, Tar, Setar, Reed, and Santoor (Dulcimer). We considered Multiple features such as MFCC, Chroma CENS, and Mel spectrogram as input. The results indicate that MFCC carries more valuable information for detecting Iranian modal music (Dastgah) than other sound representations. Moreover, the architecture inspired by autoencoders is robust in distinguishing highly correlated data like Dastgahs. It also shows that because of the precise order in Iranian Dastgah Music, Bidirectional Recurrent networks are more efficient than any other networks that have been implemented in this study. | ['Milad Dadgar', 'Farzad Didehvar', 'Danial Ebrat'] | 2022-03-29 | null | null | null | null | ['music-classification', 'music-information-retrieval'] | ['music', 'music'] | [-2.98764348e-01 -5.69044709e-01 7.85325542e-02 3.12873274e-01
-5.42350054e-01 -5.53937495e-01 4.10108954e-01 -3.40391189e-01
-3.10512483e-01 3.77446651e-01 5.45868635e-01 -2.50717886e-02
-6.27463579e-01 -7.21119642e-01 -2.73535609e-01 -6.56452417e-01
-3.13151032e-01 3.53867888e-01 -3.98006022e-01 -5.63657939e-01
5.81833661e-01 3.07488859e-01 -1.79106820e+00 5.76757729e-01
2.43150309e-01 9.43956912e-01 5.61931171e-02 7.46580005e-01
1.12250440e-01 9.67981756e-01 -1.13171768e+00 -8.52469206e-02
1.35742381e-01 -8.89785588e-01 -8.70970547e-01 -5.13652682e-01
1.20956481e-01 -1.48115873e-01 -3.33616823e-01 7.37421572e-01
1.10004067e+00 2.73787409e-01 8.10094535e-01 -6.18370950e-01
-7.03126848e-01 1.70827055e+00 -4.87117141e-01 2.85987258e-01
3.23901325e-01 -3.96710485e-01 1.10329330e+00 -6.93168581e-01
3.46155107e-01 1.10533822e+00 8.87658119e-01 2.58340746e-01
-7.77603686e-01 -1.02247906e+00 -5.74134171e-01 7.96581447e-01
-1.49755859e+00 -2.37973571e-01 1.05382085e+00 -4.63858694e-01
9.91595805e-01 4.24600124e-01 8.98296952e-01 1.12485516e+00
2.26975143e-01 7.73880601e-01 8.84758294e-01 -6.66388154e-01
-1.11881793e-01 -1.10149488e-01 6.11862987e-02 -1.31602347e-01
-2.37869576e-01 6.41675591e-02 -5.96461475e-01 -6.15558028e-02
7.09804475e-01 -2.80106664e-01 -2.04203799e-01 6.27642870e-01
-1.36415315e+00 8.32728565e-01 2.83246696e-01 1.11759114e+00
-6.76805437e-01 1.24830619e-01 8.32792521e-01 5.93003452e-01
1.82557672e-01 6.59384251e-01 -3.14475417e-01 -4.50260222e-01
-1.02018619e+00 5.78479134e-02 5.45219958e-01 2.10773975e-01
5.44591108e-03 8.35318923e-01 1.45534590e-01 1.32923353e+00
8.94186832e-03 3.79297018e-01 1.13670671e+00 -7.53198624e-01
8.51880014e-02 3.58995050e-01 -4.50977981e-01 -1.19288433e+00
-4.83886033e-01 -8.59248459e-01 -1.06798446e+00 5.59501983e-02
1.18329793e-01 -5.54306433e-02 -4.58468586e-01 1.32234681e+00
-3.41068238e-01 -1.80907607e-01 2.16657892e-01 1.15204835e+00
1.23871076e+00 9.22682285e-01 -3.90319228e-01 -6.86076209e-02
1.26000786e+00 -6.77172005e-01 -9.03453350e-01 3.95415813e-01
1.94135085e-01 -1.39813495e+00 8.80438030e-01 1.02424824e+00
-1.12816238e+00 -9.11692262e-01 -1.13763011e+00 2.17190996e-01
-3.90061557e-01 6.18827343e-01 5.65615892e-01 5.25226831e-01
-7.44050801e-01 9.64964211e-01 -2.40467325e-01 -2.35741407e-01
-1.41894102e-01 3.56581599e-01 2.77476776e-02 7.46611059e-01
-1.45714700e+00 6.01254642e-01 7.98349619e-01 2.22909734e-01
-6.88877463e-01 -1.32771343e-01 -2.12720633e-01 9.09939259e-02
-2.82883525e-01 -2.58191943e-01 1.14735019e+00 -1.28309333e+00
-1.80827987e+00 5.20212471e-01 4.50307697e-01 -6.11746669e-01
1.15279913e-01 -3.09241205e-01 -9.20662701e-01 6.78905919e-02
-1.79790705e-01 3.73484552e-01 6.69270277e-01 -7.04674363e-01
-4.64611620e-01 -2.64169395e-01 -2.67950982e-01 1.38549268e-01
-4.43759352e-01 4.71039027e-01 1.61457896e-01 -1.15370142e+00
2.63542622e-01 -9.69526470e-01 4.87670779e-01 -1.25969732e+00
-5.64303339e-01 -3.53048056e-01 6.34919405e-01 -9.44382727e-01
1.69555032e+00 -2.35301328e+00 1.07973635e-01 2.68187344e-01
-4.01516527e-01 1.64868742e-01 -1.45037800e-01 6.41744673e-01
-3.46356899e-01 -1.10275224e-02 1.83368966e-01 3.63742203e-01
9.58452147e-05 -4.69235890e-02 -3.66526872e-01 2.56755531e-01
-4.14972782e-01 6.34322405e-01 -2.65407592e-01 -2.08893299e-01
1.11580893e-01 7.73801982e-01 -1.24983169e-01 -1.75549954e-01
2.11282209e-01 2.72993177e-01 2.19643489e-02 7.75169909e-01
3.16494137e-01 2.51718462e-01 -4.43027243e-02 -3.31813186e-01
-4.28860217e-01 6.66461289e-01 -1.30609059e+00 1.70737553e+00
-2.62554497e-01 9.70956028e-01 -3.89114916e-01 -8.55183601e-01
1.15755880e+00 8.14575195e-01 5.79059005e-01 -7.10521758e-01
6.55017853e-01 4.53554124e-01 6.34817481e-01 -3.20667058e-01
7.13963807e-01 -1.32425383e-01 8.22898746e-02 4.51876163e-01
1.53433859e-01 1.76270142e-01 2.56226093e-01 -3.41571122e-01
6.87505901e-01 -6.70731813e-02 9.91077423e-02 -3.28994505e-02
5.36019206e-01 -1.94064826e-01 4.30680782e-01 6.28188908e-01
3.08456004e-01 6.81596696e-01 -2.27995086e-02 -3.53158712e-01
-9.97849762e-01 -7.53639400e-01 -2.78764844e-01 1.34690666e+00
-4.66013998e-01 -4.64101553e-01 -3.35499227e-01 1.20013028e-01
-3.33713561e-01 4.62398350e-01 -4.36765760e-01 -1.43645937e-02
-6.59166813e-01 -5.49262822e-01 1.08521152e+00 3.51321727e-01
6.69374108e-01 -1.71585763e+00 -4.98046488e-01 4.93283063e-01
-3.60820442e-01 -4.20492530e-01 -2.90819705e-01 2.68430561e-01
-9.81717944e-01 -9.19659734e-01 -1.04776382e+00 -1.06337726e+00
-6.02505326e-01 -6.51712343e-02 1.05553699e+00 -4.92638946e-01
-2.20098913e-01 5.63749229e-04 -6.58127129e-01 -6.57189488e-01
-4.00907308e-01 3.15467030e-01 4.58981059e-02 -9.12603140e-02
4.24377441e-01 -7.32937634e-01 -4.08926129e-01 1.80821687e-01
-6.17548645e-01 -3.19815099e-01 6.66125834e-01 7.62998700e-01
3.63076389e-01 4.32249248e-01 8.78323495e-01 -4.91654217e-01
8.65649581e-01 -2.94295907e-01 4.73280773e-02 -3.02204520e-01
-4.13703293e-01 -3.61030251e-01 4.99707848e-01 -7.26939976e-01
-7.53675580e-01 -3.83488834e-01 -4.92600650e-01 -3.81508410e-01
-1.84271723e-01 7.06177473e-01 3.30439150e-01 2.54182577e-01
9.17324841e-01 2.81650454e-01 -3.33141357e-01 -8.76157403e-01
-9.44078416e-02 1.21565139e+00 6.30705774e-01 -3.25921267e-01
3.36306274e-01 3.26340348e-02 -3.00166845e-01 -1.07771742e+00
-4.03143048e-01 -4.88032818e-01 -2.62744278e-01 -4.62859869e-01
8.06238234e-01 -8.79957020e-01 -1.07523572e+00 5.10660052e-01
-1.02136087e+00 2.15676367e-01 -3.60021889e-01 1.15615451e+00
-4.50539827e-01 9.03632864e-02 -1.26500070e+00 -1.17703450e+00
-9.37518895e-01 -7.27758944e-01 5.48130214e-01 9.12136212e-02
-6.36454999e-01 -4.56724375e-01 2.88427800e-01 2.13843837e-01
5.92476726e-01 5.12654670e-02 9.94635999e-01 -7.41192102e-01
1.59605339e-01 -1.21284284e-01 3.15570235e-01 5.55121541e-01
1.05721220e-01 1.01561889e-01 -1.28336465e+00 3.79309431e-02
7.37851411e-02 -2.95949519e-01 8.73827815e-01 6.63891673e-01
1.04088616e+00 -2.95865387e-01 5.67895412e-01 3.69498670e-01
1.06803656e+00 9.33413446e-01 8.12608898e-01 7.76921630e-01
4.44909155e-01 2.28174105e-01 3.01291198e-01 5.03966451e-01
-9.89940763e-02 5.15494525e-01 2.80240983e-01 -1.80262383e-02
-2.01945201e-01 -1.26800612e-01 7.72328615e-01 1.64576066e+00
-6.83705270e-01 2.14279648e-02 -5.92245698e-01 4.34051931e-01
-1.64120042e+00 -1.38772357e+00 -2.55262822e-01 2.06339478e+00
6.59346402e-01 -2.19407544e-01 6.30211890e-01 1.22973001e+00
7.17698395e-01 1.79792792e-02 -1.85342640e-01 -7.66591430e-01
-5.88184953e-01 5.17349541e-01 6.91065216e-04 -2.08807513e-01
-1.25595391e+00 7.45982766e-01 6.35985756e+00 1.00293005e+00
-1.25353599e+00 7.46569857e-02 1.02660313e-01 -2.36674547e-01
6.29014224e-02 -3.80825311e-01 -1.16200715e-01 4.48842973e-01
1.29336703e+00 2.18333155e-01 6.41932547e-01 7.24921048e-01
2.19531626e-01 2.65959114e-01 -6.60872281e-01 1.32407558e+00
2.23429546e-01 -1.14109540e+00 2.35451564e-01 2.47887783e-02
6.50587142e-01 1.36987135e-01 2.37844348e-01 5.21352887e-01
-1.61513045e-01 -9.55454946e-01 7.24105775e-01 5.53644359e-01
4.22607154e-01 -1.24287188e+00 1.13051701e+00 1.13610864e-01
-1.09121168e+00 -2.60528892e-01 -4.37534362e-01 -3.44708771e-01
-2.73271054e-01 3.23096037e-01 -6.66633427e-01 7.19042003e-01
9.39450204e-01 9.63605344e-01 -3.71364295e-01 1.06247020e+00
1.74626008e-01 1.09547949e+00 -2.45767668e-01 -5.57403192e-02
1.34542480e-01 -4.36763018e-01 6.76751196e-01 1.31438506e+00
6.92514420e-01 -2.90874064e-01 -3.00802231e-01 5.82001746e-01
2.58681625e-01 6.65714443e-01 -4.67096239e-01 -4.21539098e-01
2.43458003e-01 9.45814848e-01 -5.28296053e-01 -2.36210406e-01
8.89354572e-02 6.83704615e-01 -5.61889172e-01 2.89298445e-01
-7.06567943e-01 -6.17789567e-01 4.01986241e-01 -1.16766587e-01
4.85076271e-02 7.97956809e-02 -2.62005568e-01 -8.68381441e-01
-3.30768019e-01 -1.26445460e+00 7.72523880e-01 -1.05020714e+00
-1.29848480e+00 9.35778737e-01 -4.25826788e-01 -1.68608868e+00
-6.37268186e-01 -5.18717408e-01 -5.09427607e-01 7.88989544e-01
-7.93397307e-01 -9.78694201e-01 1.68588340e-01 8.39764059e-01
6.74288988e-01 -8.16502273e-01 1.26848590e+00 8.49585354e-01
-3.82502615e-01 4.64851409e-01 1.98231250e-01 4.95429546e-01
7.34231889e-01 -1.07446849e+00 -2.81419098e-01 2.25250065e-01
9.12186325e-01 4.80854362e-01 4.89939928e-01 -2.34828576e-01
-1.16434109e+00 -5.16500235e-01 8.06360543e-01 1.62422270e-01
6.35324419e-01 3.17824662e-01 -6.51921093e-01 5.44666529e-01
6.75774038e-01 -1.08878899e+00 9.49827015e-01 4.79096711e-01
-2.53380924e-01 -2.08738297e-01 -4.63541210e-01 4.32073891e-01
5.54385781e-01 -7.19557583e-01 -8.20308268e-01 3.73339839e-02
1.55927867e-01 -6.28811121e-02 -1.04068446e+00 4.62020665e-01
9.09448922e-01 -1.12714648e+00 8.45071316e-01 -3.23104799e-01
3.93482059e-01 -4.42138255e-01 -2.30685174e-01 -1.25398123e+00
-5.11072457e-01 -3.44218671e-01 6.59580603e-02 1.35907674e+00
3.88231307e-01 -4.45780009e-01 4.07590777e-01 -6.19013846e-01
-1.83781862e-01 -1.30126983e-01 -7.19299734e-01 -7.13905156e-01
-4.34869081e-02 -7.74930477e-01 3.74702543e-01 1.15856230e+00
1.40047669e-01 6.19103789e-01 -7.42921352e-01 -5.07131457e-01
1.55185699e-01 3.77200484e-01 6.34678602e-01 -1.47895181e+00
-5.80059767e-01 -9.21674788e-01 -8.05425704e-01 -3.88863206e-01
-7.42480680e-02 -1.01961052e+00 -4.38812613e-01 -1.43824685e+00
1.14726052e-01 -3.33312824e-02 -8.39445412e-01 5.62377334e-01
6.73652709e-01 7.79406130e-01 3.31255943e-01 4.14017856e-01
3.34621072e-02 2.72405565e-01 9.62938428e-01 -4.60589260e-01
-4.81985569e-01 2.55852342e-01 -4.11106765e-01 9.82373774e-01
1.10954607e+00 -3.36077958e-01 -9.07907262e-02 -2.07007080e-01
4.66470301e-01 6.70345426e-02 -3.15742916e-03 -1.45680439e+00
3.99708338e-02 4.50967491e-01 7.32094765e-01 -1.07995200e+00
6.12476408e-01 -5.51127315e-01 6.54441595e-01 4.36325908e-01
-4.99221355e-01 1.83984935e-01 1.80521190e-01 -5.35736606e-02
-7.35655665e-01 -1.65800244e-01 4.86189365e-01 -1.85143538e-02
-6.61708057e-01 -3.61450404e-01 -7.60724306e-01 -3.29034656e-01
2.47739226e-01 -2.48187959e-01 -2.32701488e-02 -7.27624238e-01
-9.89865243e-01 -6.49446487e-01 -3.44945073e-01 6.26497328e-01
5.32189906e-01 -1.63214278e+00 -9.89774585e-01 4.18568924e-02
-1.31825432e-01 -9.21799481e-01 4.53546792e-01 8.20369065e-01
-5.55693865e-01 4.21746790e-01 -5.22959352e-01 -4.77381110e-01
-1.36648023e+00 3.30737948e-01 1.71894535e-01 2.05622956e-01
-7.00530112e-01 5.35555005e-01 -3.20518106e-01 -2.29070723e-01
5.10137439e-01 -1.14613548e-01 -9.22093511e-01 7.01634526e-01
4.12303835e-01 7.03743219e-01 1.39600977e-01 -8.55161428e-01
-2.21777424e-01 8.04441929e-01 2.97368556e-01 -2.77570695e-01
1.30532384e+00 1.77186623e-01 -3.59460294e-01 1.18211520e+00
8.79976392e-01 2.67913342e-01 9.85889658e-02 2.53340751e-01
4.97515239e-02 -1.02768376e-01 1.63283944e-01 -1.15542424e+00
-1.22538590e+00 9.26024199e-01 9.87565398e-01 5.07367790e-01
1.36947167e+00 -3.32408071e-01 8.10541332e-01 4.74218518e-01
1.82219386e-01 -1.41348422e+00 -2.25699291e-01 8.09636652e-01
1.08923364e+00 -5.42312920e-01 -4.36032623e-01 5.54956615e-01
-6.69681966e-01 1.42847574e+00 6.61292747e-02 -3.48256230e-01
6.57268465e-01 -1.58220413e-03 4.13162887e-01 1.72012765e-02
-4.83414263e-01 -3.55558783e-01 5.23748815e-01 4.60866094e-02
1.08615994e+00 2.97493011e-01 -7.39519119e-01 8.27841520e-01
-1.03737974e+00 -1.34512573e-01 4.96366084e-01 3.44326913e-01
-3.20880741e-01 -8.43101323e-01 -8.16377521e-01 3.37559402e-01
-1.00524175e+00 -1.31563023e-01 -7.23543584e-01 9.30552959e-01
2.38723621e-01 1.18807566e+00 9.35326219e-02 -9.38568354e-01
2.89983928e-01 2.90567636e-01 3.26569438e-01 -8.19999352e-02
-1.24896502e+00 9.31538165e-01 9.58502665e-02 -9.13735330e-02
-6.26238406e-01 -2.48710275e-01 -1.09710240e+00 -4.24420744e-01
-3.29787850e-01 5.65810561e-01 8.97869468e-01 6.40388906e-01
9.01664346e-02 9.97645020e-01 5.92653871e-01 -1.00114369e+00
-2.06012622e-01 -1.68275809e+00 -1.19033408e+00 3.06474954e-01
9.74554475e-03 -3.54659349e-01 -3.72476041e-01 -1.70441404e-01] | [15.896685600280762, 5.182823181152344] |
b0281f7d-3c0b-4111-b4eb-bed931847434 | cpmlho-hyperparameter-tuning-via-cutting | 2212.06150 | null | https://arxiv.org/abs/2212.06150v1 | https://arxiv.org/pdf/2212.06150v1.pdf | CPMLHO:Hyperparameter Tuning via Cutting Plane and Mixed-Level Optimization | The hyperparameter optimization of neural network can be expressed as a bilevel optimization problem. The bilevel optimization is used to automatically update the hyperparameter, and the gradient of the hyperparameter is the approximate gradient based on the best response function. Finding the best response function is very time consuming. In this paper we propose CPMLHO, a new hyperparameter optimization method using cutting plane method and mixed-level objective function.The cutting plane is added to the inner layer to constrain the space of the response function. To obtain more accurate hypergradient,the mixed-level can flexibly adjust the loss function by using the loss of the training set and the verification set. Compared to existing methods, the experimental results show that our method can automatically update the hyperparameters in the training process, and can find more superior hyperparameters with higher accuracy and faster convergence. | ['Chen Zhu', 'Mana Zheng', 'Shaoyu Dou', 'Yang Jiao', 'Shuo Yang'] | 2022-12-11 | null | null | null | null | ['bilevel-optimization'] | ['methodology'] | [-5.05992651e-01 -1.75571725e-01 -3.44357789e-01 -4.72153455e-01
-3.12226802e-01 -2.70497829e-01 -3.27153862e-01 -4.30404574e-01
-6.98994398e-01 8.25996578e-01 -1.75318584e-01 -9.15327147e-02
-5.36835134e-01 -8.42710018e-01 -5.35537302e-01 -1.01791668e+00
3.34767193e-01 4.28953052e-01 3.34331989e-01 -1.67583480e-01
3.98728222e-01 3.82455915e-01 -9.74850118e-01 2.90815253e-02
9.86029267e-01 1.19662356e+00 2.07222879e-01 1.80360600e-01
-9.47237015e-02 -6.38359040e-02 -6.04411304e-01 -1.60262976e-02
3.40812951e-01 -2.14111641e-01 -5.91667771e-01 -2.14100862e-03
4.14525233e-02 -2.12177947e-01 -2.34734453e-02 1.13627863e+00
5.70551872e-01 1.19867392e-01 3.87620032e-01 -1.25247371e+00
-2.67442346e-01 5.08932889e-01 -5.92621863e-01 4.24334481e-02
-2.39488065e-01 8.12579971e-03 7.19584465e-01 -6.96158946e-01
2.95344472e-01 1.39489925e+00 7.26760626e-01 4.70980257e-01
-9.59383249e-01 -1.18201995e+00 3.83672595e-01 3.61165464e-01
-1.54517663e+00 9.49914530e-02 7.71389425e-01 -1.45452082e-01
5.30512094e-01 9.39641446e-02 8.24254394e-01 4.08528626e-01
5.30318171e-02 7.83855379e-01 5.21742761e-01 -3.62509221e-01
6.08108602e-02 4.35558677e-01 2.93261647e-01 1.12747574e+00
3.76652032e-01 3.77839282e-02 1.26657337e-02 -1.88239723e-01
1.00934303e+00 -2.02332258e-01 -6.76659048e-01 -2.51830697e-01
-8.47277462e-01 1.02215242e+00 6.05955422e-01 5.10403775e-02
-1.60777241e-01 1.52679354e-01 3.86929452e-01 2.30051443e-01
-5.06256882e-04 6.48740530e-01 -6.96787059e-01 4.86333370e-02
-5.61837614e-01 1.74294055e-01 8.38055611e-01 6.30933821e-01
8.65339935e-01 1.19761266e-01 -2.49413475e-01 1.36247468e+00
8.80846262e-01 3.32215518e-01 7.49920785e-01 -9.37802255e-01
7.84612954e-01 8.60581517e-01 -5.92938028e-02 -1.00566292e+00
-5.35834730e-01 -4.80234116e-01 -6.52773142e-01 2.49065936e-01
3.99318784e-01 -7.42183328e-01 -9.00272369e-01 1.52008307e+00
7.34690905e-01 1.03622392e-01 -3.48708034e-02 1.30186629e+00
6.95201337e-01 1.01595747e+00 -2.73943335e-01 -3.72802466e-01
1.06289172e+00 -1.04588878e+00 -7.62243211e-01 -7.99244419e-02
5.16527832e-01 -5.43557882e-01 1.20427668e+00 4.90641445e-01
-8.41960490e-01 -1.50970563e-01 -1.46615291e+00 2.27872536e-01
-4.14248258e-02 4.49664712e-01 4.65801686e-01 4.01296824e-01
-4.51868981e-01 5.66136658e-01 -6.49275839e-01 2.99194902e-01
3.28637511e-01 6.85957253e-01 -6.38910383e-02 2.43402287e-01
-1.51307106e+00 7.22182095e-01 1.04594338e+00 4.49898124e-01
-4.02880728e-01 -9.45867658e-01 -5.75163484e-01 1.90339684e-01
4.36784714e-01 -4.55406398e-01 9.09500122e-01 -7.74397850e-01
-1.95307875e+00 2.91477114e-01 2.70113170e-01 -6.92477673e-02
3.50778013e-01 -1.21447984e-02 -3.97085473e-02 -1.74399123e-01
-4.82354730e-01 4.94898319e-01 6.94156826e-01 -1.03564405e+00
-5.51951051e-01 -3.31384778e-01 -3.62805516e-01 5.95299363e-01
-6.20740652e-01 -2.84293056e-01 -8.56495619e-01 -3.34844381e-01
3.31759572e-01 -8.67994606e-01 -2.36537576e-01 -8.80994052e-02
-3.16701859e-01 -1.91937551e-01 9.14783299e-01 -5.05260050e-01
1.69483542e+00 -2.22487044e+00 2.00095862e-01 7.47848094e-01
-7.66228661e-02 5.44961870e-01 2.05392484e-02 -1.46536231e-01
1.78836510e-02 2.03126445e-01 -2.19640911e-01 4.16451305e-01
-9.87980664e-02 1.53662771e-01 -2.32372154e-02 2.97179878e-01
-2.06539333e-01 4.79448289e-01 -5.06536067e-01 -6.08583212e-01
-7.25745261e-02 5.11375964e-01 -6.39828146e-01 2.04237714e-01
-1.08412541e-01 -1.67240053e-01 -9.15723741e-01 2.02874362e-01
7.27460027e-01 -8.83070827e-02 -8.89426470e-02 -5.42979240e-01
-2.41901316e-02 -1.97145596e-01 -1.77482271e+00 1.08583832e+00
-4.55810726e-01 4.79219317e-01 2.39724442e-01 -8.20716500e-01
1.38923383e+00 1.69234678e-01 2.99665153e-01 -3.38436633e-01
4.86070514e-01 1.05924122e-01 -6.65573552e-02 -6.92303300e-01
-6.26626089e-02 5.09969145e-02 3.40946585e-01 1.71740681e-01
-3.08128059e-01 -2.75623575e-02 1.58402383e-01 -4.90265638e-01
2.90128529e-01 -6.60160407e-02 5.03305271e-02 -2.43672609e-01
9.30680096e-01 1.53657282e-02 1.00143111e+00 1.69965029e-01
1.67929351e-01 4.24696803e-01 7.56862104e-01 -5.45026720e-01
-9.19353426e-01 -4.25372541e-01 -5.68254530e-01 7.31393754e-01
1.70529187e-01 -5.99333178e-03 -1.05386019e+00 -7.21638441e-01
1.52209178e-01 2.98741549e-01 -3.01214486e-01 -4.27962184e-01
-8.04909647e-01 -1.26896667e+00 3.81154895e-01 2.31602445e-01
1.02015257e+00 -7.84191728e-01 -1.74420372e-01 1.54198274e-01
-1.28099591e-01 -5.06301224e-01 -7.96892583e-01 -8.02987814e-02
-1.00242245e+00 -1.04808760e+00 -7.05115736e-01 -1.07871974e+00
9.88504887e-01 -3.76135975e-01 5.22384048e-01 4.94701564e-01
-2.88184434e-01 -5.00378609e-01 1.56609956e-02 -2.69124746e-01
-4.58221324e-02 2.40188688e-01 -1.16235487e-01 -4.72631007e-02
-7.01496974e-02 -2.27955312e-01 -4.90825295e-01 7.61714160e-01
-6.81294560e-01 -1.77452877e-01 4.48772401e-01 1.07970488e+00
8.61373723e-01 1.00471467e-01 6.70718551e-01 -7.97200203e-01
1.05236411e+00 -2.33407781e-01 -1.22588563e+00 4.16802794e-01
-1.18817115e+00 5.52316427e-01 9.05573130e-01 -7.66917825e-01
-1.01479197e+00 -6.47285804e-02 -7.72506446e-02 -5.71616709e-01
4.73250002e-01 4.51825976e-01 -3.69646996e-01 -5.43607116e-01
5.63053846e-01 -1.69918478e-01 1.24192238e-01 -4.61545765e-01
4.82758656e-02 6.85932100e-01 2.18210667e-01 -4.99079049e-01
5.38079500e-01 -1.21711090e-01 2.07100451e-01 -3.35801870e-01
-7.52308786e-01 -2.84886658e-01 -3.15165132e-01 -1.89319611e-01
5.53392470e-01 -3.42460364e-01 -1.13409877e+00 4.15274948e-01
-8.24684501e-01 -2.80493170e-01 -2.28039566e-02 7.45574415e-01
-2.93948591e-01 2.01668069e-01 -3.85156989e-01 -3.74111533e-01
-7.37260222e-01 -1.35702670e+00 4.26079512e-01 7.01990128e-01
3.47331852e-01 -1.06901085e+00 4.76403572e-02 1.24616630e-01
6.36868775e-02 1.00198820e-01 9.86656547e-01 -3.51551175e-01
-3.30054313e-01 -3.53776366e-01 -1.33581623e-01 3.86265695e-01
-2.72546679e-01 2.83841163e-01 -3.99008900e-01 -3.27861816e-01
9.95270908e-02 -1.60602704e-01 6.56924963e-01 5.33994973e-01
1.49690533e+00 -6.87536299e-01 -3.71045530e-01 1.51666772e+00
1.67588305e+00 5.72926462e-01 5.74590623e-01 8.43163967e-01
6.05635464e-01 2.70445853e-01 7.50906765e-01 3.41279894e-01
8.26993212e-02 5.54889739e-01 1.62941068e-01 6.80655688e-02
1.45236373e-01 -2.23647177e-01 9.50913876e-03 5.97986698e-01
2.12665573e-01 2.38446351e-02 -8.58015001e-01 9.60560702e-03
-2.00074506e+00 -7.73920298e-01 -3.94248916e-03 2.37312913e+00
1.22106135e+00 1.50044203e-01 -9.46385339e-02 1.22547060e-01
7.98881531e-01 -1.43500671e-01 -9.12618220e-01 -5.31125367e-01
3.27384889e-01 -5.52223980e-01 6.90474391e-01 8.22850049e-01
-7.78749883e-01 7.30224073e-01 6.60518789e+00 9.29026484e-01
-1.42598987e+00 -3.44700426e-01 5.06234229e-01 -3.05152416e-01
-1.25048310e-01 -1.39515519e-01 -1.41712511e+00 7.39083767e-01
1.54233992e-01 -2.05722824e-02 7.79356301e-01 9.38380063e-01
3.29063654e-01 3.08168121e-02 -6.25806212e-01 1.14227760e+00
-1.41183913e-01 -1.14950752e+00 -1.33972108e-01 -1.93833202e-01
8.10907960e-01 -2.92867780e-01 7.03473762e-02 2.52828032e-01
-1.33772036e-02 -8.51920903e-01 1.06617041e-01 5.17009974e-01
4.37789917e-01 -1.08550560e+00 7.65258074e-01 4.61998343e-01
-1.03554714e+00 -4.42189813e-01 -6.27824187e-01 5.17922759e-01
-9.54281539e-02 5.18708169e-01 -9.28953946e-01 4.72091027e-02
6.05100751e-01 1.35773018e-01 -2.30284885e-01 1.55728722e+00
-3.64016473e-01 4.62575167e-01 -6.33885920e-01 -5.59381723e-01
3.18762302e-01 -7.28027523e-01 6.29543722e-01 9.69074965e-01
1.07519813e-01 2.06806101e-02 3.34761679e-01 6.96269453e-01
2.06871494e-03 4.44379836e-01 1.83838174e-01 3.21371198e-01
7.30415344e-01 1.29806292e+00 -2.79438198e-01 1.51022980e-02
2.10223436e-01 2.76488066e-01 5.89453042e-01 4.69984412e-01
-9.29639399e-01 -1.04538727e+00 2.77290821e-01 -1.35053620e-01
1.42922387e-01 9.19015408e-02 -4.77662206e-01 -6.75085664e-01
1.73586950e-01 -7.34762132e-01 5.27994990e-01 -6.69891417e-01
-7.75300682e-01 5.70427358e-01 1.77667513e-01 -8.78188550e-01
-1.35976970e-01 -6.22399986e-01 -7.84209847e-01 1.09295392e+00
-1.34485435e+00 -4.40293610e-01 -4.38678712e-01 3.65651429e-01
1.92806512e-01 -4.25703138e-01 3.43373686e-01 2.72602171e-01
-1.25629044e+00 1.04579747e+00 3.34376395e-01 -6.75778985e-02
4.82441932e-01 -8.26385081e-01 -4.84192163e-01 2.88827181e-01
-5.91137946e-01 4.69517291e-01 6.31861567e-01 -3.58910412e-01
-9.41396832e-01 -7.54958391e-01 2.90299356e-01 4.93875980e-01
4.00340587e-01 9.09640193e-02 -1.08827782e+00 2.53853172e-01
-3.19587916e-01 -2.11926475e-01 3.80617678e-01 9.19042826e-02
9.47364792e-02 -6.91943586e-01 -1.24974060e+00 6.03027940e-01
4.58576173e-01 3.08475792e-01 -2.43017420e-01 4.48819220e-01
7.82627642e-01 -6.45720005e-01 -1.08124256e+00 6.26282156e-01
5.69760323e-01 -2.10783243e-01 9.61097360e-01 -5.11792719e-01
-5.79584464e-02 -3.95120502e-01 3.91806006e-01 -1.58858168e+00
-4.48732346e-01 -5.52968383e-01 -3.65948938e-02 1.14675677e+00
7.77660966e-01 -8.84411812e-01 1.04139245e+00 6.41524136e-01
-1.06540984e-02 -1.74245751e+00 -7.45992422e-01 -5.56732714e-01
-5.86616397e-02 1.74776390e-01 1.00074768e+00 7.17160046e-01
5.82345994e-03 3.03002745e-01 -8.53267536e-02 1.25608623e-01
5.35408914e-01 -1.23731412e-01 3.65686297e-01 -1.17017114e+00
-1.62539542e-01 -5.96272349e-01 -1.09167330e-01 -9.36027527e-01
1.82666585e-01 -7.83329964e-01 8.92670527e-02 -1.49383903e+00
-4.02038507e-02 -9.16786134e-01 -4.08155099e-02 6.42660975e-01
-3.78860265e-01 -3.11773092e-01 -7.35530555e-02 2.55643815e-01
-6.11776412e-02 5.69257259e-01 1.78187239e+00 4.34741415e-02
-8.13916743e-01 2.30347350e-01 -4.06327903e-01 7.73686945e-01
1.17330754e+00 -4.87492144e-01 -5.75003088e-01 -5.81534863e-01
4.19629574e-01 3.60310241e-03 -1.87385976e-01 -6.41910732e-01
4.02642310e-01 -4.86519933e-01 5.22672415e-01 -3.75508517e-01
2.07028583e-01 -8.47957551e-01 -1.05791546e-01 5.26726723e-01
-4.42567289e-01 -3.25411521e-02 1.69665232e-01 3.01343620e-01
-3.34340185e-02 -9.86101449e-01 1.14624012e+00 3.51401828e-02
-3.02205026e-01 5.70667207e-01 1.37868688e-01 1.62566379e-01
1.09649014e+00 -3.38223428e-01 -2.54592538e-01 9.92306992e-02
-4.40770179e-01 9.92031515e-01 1.19504973e-01 2.31016099e-01
7.91482627e-01 -1.50129378e+00 -4.35261428e-01 3.54608834e-01
-5.00686646e-01 3.39122444e-01 -1.45348370e-01 4.65120673e-01
-8.08296382e-01 1.50217116e-01 -7.26276487e-02 -2.96653718e-01
-1.33743989e+00 1.88375577e-01 1.16689909e+00 -2.46454820e-01
-3.82253617e-01 9.38710153e-01 -2.16556042e-01 -6.45914137e-01
7.22589374e-01 -4.06434424e-02 -6.00265741e-01 -9.07346904e-02
5.03439188e-01 7.31555343e-01 -1.40322626e-01 -2.50701401e-02
-2.23639891e-01 9.51434195e-01 -1.05761282e-01 -1.30174488e-01
1.09143984e+00 1.02577344e-01 -3.08697820e-01 1.08956859e-01
1.66579032e+00 -4.67501163e-01 -1.32921779e+00 1.20606702e-02
-5.08224070e-01 -4.29898918e-01 5.65796435e-01 -8.92008245e-01
-1.45195472e+00 4.62266296e-01 7.80352056e-01 -8.93362537e-02
1.12216198e+00 -5.10520697e-01 9.64604378e-01 7.34132826e-01
-5.54538779e-02 -1.63835716e+00 -7.39218220e-02 6.17248714e-01
9.09004986e-01 -8.85028601e-01 3.24151814e-01 -4.47176814e-01
-5.64840734e-01 1.41158843e+00 1.14531517e+00 -2.28000924e-01
7.96070278e-01 2.89559931e-01 8.54943171e-02 -5.66998981e-02
-6.09779894e-01 5.46791017e-01 4.33918476e-01 1.69611245e-01
1.15574792e-01 -2.75334746e-01 -7.81701386e-01 5.75536311e-01
-2.48374894e-01 -1.29652694e-01 1.03786133e-01 3.74805450e-01
-9.01003301e-01 -1.05676532e+00 -7.01616347e-01 4.50962365e-01
-2.84620613e-01 1.83769777e-01 5.01661189e-02 5.43990433e-01
1.16142504e-01 6.39498651e-01 4.15035821e-02 -3.84909600e-01
6.43638968e-01 -1.45583212e-01 3.61913204e-01 -2.74630845e-01
-5.67826986e-01 -2.49158546e-01 -1.49451494e-01 -3.81115526e-01
5.00374019e-01 -3.01031500e-01 -1.86691868e+00 -1.16795376e-01
-9.57854152e-01 4.86136585e-01 9.01215017e-01 8.68971705e-01
2.77747773e-02 4.28232670e-01 1.14412320e+00 -2.26193443e-01
-9.95456755e-01 -6.90001488e-01 -2.21397385e-01 2.81550325e-02
1.53575078e-01 -6.42593861e-01 -4.48829085e-01 -3.85266542e-01] | [6.872664928436279, 4.009604454040527] |
abebd6b6-ba8c-4a31-9be8-ba610e058a24 | on-estimation-and-inference-of-large | 2211.01921 | null | https://arxiv.org/abs/2211.01921v2 | https://arxiv.org/pdf/2211.01921v2.pdf | On Estimation and Inference of Large Approximate Dynamic Factor Models via the Principal Component Analysis and its equivalence with Quasi Maximum Likelihood estimation | We provide an alternative derivation of the asymptotic results for the Principal Components estimator of a large approximate factor model and we prove that the derived estimator of the loadings is asymptotically equivalent to their Quasi Maximum Likelihood estimator. This result holds regardless of the specific second order structure of the idiosyncratic components, which can be both serially and cross-sectionally weakly correlated as well as heteroskedastic. Results are derived under a minimal set of assumptions and, in particular, we require only the existence of 4th order moments. A special focus is given to the time series setting, a case considered in almost all recent econometric applications of factor models. Hence, estimation is based on the classical $n\times n$ sample covariance matrix and not on a $T\times T$ covariance matrix often considered in the literature. Indeed, despite the two approaches being asymptotically equivalent, the former is more coherent with a time series setting and it immediately allows us to write more intuitive asymptotic expansions for the Principal Component estimators showing that they are equivalent to OLS as long as $\sqrt n/T\to 0$ and $\sqrt T/n\to 0$, that is the loadings are estimated in a time series regression as if the factors were known, while the factors are estimated in a cross-sectional regression as if the loadings were known. Finally, we give some alternative sets of primitive sufficient conditions for mean-squared consistency of the sample covariance matrix of the factors, of the idiosyncratic components, and of the observed time series, which is the starting point for Principal Component Analysis. | ['Matteo Barigozzi'] | 2022-11-03 | null | null | null | null | ['time-series-regression'] | ['time-series'] | [-1.88005358e-01 -1.24481127e-01 -6.38441667e-02 5.98571226e-02
-3.34485352e-01 -8.26441467e-01 3.88409227e-01 -2.47403830e-01
-4.78167892e-01 5.77963650e-01 8.19228441e-02 -6.05762064e-01
-8.06252301e-01 -6.95735872e-01 -6.39998198e-01 -8.02506387e-01
-2.44867533e-01 1.52527899e-01 -1.91504940e-01 -5.36934957e-02
1.30958650e-02 4.57904369e-01 -1.30197275e+00 -9.43217397e-01
5.94715714e-01 1.09526908e+00 -1.11926898e-01 6.59524143e-01
6.62391782e-02 3.13458830e-01 -3.00984085e-01 -5.59518158e-01
5.73035479e-01 -5.96899450e-01 -3.87038052e-01 3.85717809e-01
-2.30053157e-01 7.32613951e-02 -8.08681548e-02 1.08670902e+00
6.06913418e-02 1.84511125e-01 8.32125247e-01 -1.18076849e+00
-5.00162899e-01 5.44445992e-01 -6.12458289e-01 1.42834455e-01
2.64093131e-01 -1.59775972e-01 9.18621421e-01 -6.89709425e-01
4.07196999e-01 8.89686823e-01 8.94793153e-01 -1.01275355e-01
-1.45422995e+00 -3.20084274e-01 -8.00078064e-02 -4.14222538e-01
-1.29714096e+00 -2.99508244e-01 6.05834961e-01 -7.82813907e-01
5.99352777e-01 2.69437134e-01 5.94802558e-01 9.52862799e-01
3.66949141e-01 1.62279412e-01 1.18712175e+00 -5.21303296e-01
2.28007674e-01 2.58993506e-01 5.75635552e-01 2.84225792e-01
7.06473112e-01 2.71502912e-01 -7.08465651e-02 -2.72351980e-01
9.97070491e-01 1.21646017e-01 -2.03932136e-01 -5.17334759e-01
-1.11231601e+00 9.33502197e-01 -2.91125923e-01 6.74962282e-01
-6.29687190e-01 8.00019503e-02 2.70647347e-01 3.98118883e-01
3.48069787e-01 1.13436483e-01 -4.73628670e-01 -1.10469587e-01
-6.69198632e-01 1.58378631e-01 9.69673514e-01 9.35635686e-01
8.00992250e-01 4.03539985e-01 3.51908982e-01 5.31857073e-01
1.30528659e-01 1.09100926e+00 5.61287880e-01 -9.47020113e-01
1.15522414e-01 2.52461642e-01 2.51284599e-01 -9.73995328e-01
-3.84203255e-01 -8.07768583e-01 -9.82381582e-01 -1.96211189e-01
6.28528357e-01 -3.66174191e-01 -1.83943108e-01 2.16901779e+00
4.20258082e-02 -2.44163256e-02 -3.18796262e-02 4.22045618e-01
-4.25927103e-01 3.51474375e-01 -1.37003049e-01 -9.63302076e-01
1.21746516e+00 -2.77505934e-01 -7.83037186e-01 -6.36457279e-02
4.03474391e-01 -6.83024645e-01 7.55592823e-01 2.96012700e-01
-1.11033916e+00 -4.36467677e-01 -6.76862538e-01 5.85235476e-01
-1.49920553e-01 1.44148216e-01 6.51100516e-01 8.19929540e-01
-1.07042563e+00 6.55417860e-01 -8.20146143e-01 -3.74237597e-01
-4.20984447e-01 1.27608687e-01 -3.97113353e-01 3.04736644e-01
-9.24158037e-01 6.57971442e-01 2.30233930e-02 2.08569974e-01
-3.80234838e-01 -5.97931027e-01 -6.05770886e-01 2.92622566e-01
1.73586130e-01 -4.52276587e-01 1.22864294e+00 -1.25863528e+00
-1.18671417e+00 4.75707173e-01 -2.99949557e-01 -1.70540705e-01
6.14379287e-01 1.93384647e-01 -5.28732836e-01 3.98199335e-02
3.81893694e-01 -6.15338624e-01 8.83613586e-01 -6.47215903e-01
-3.35253924e-01 -6.05843902e-01 -2.25357279e-01 -3.33771437e-01
-2.75022745e-01 7.80798420e-02 -4.76377979e-02 -6.19511902e-01
2.37953678e-01 -9.44195628e-01 -3.43358159e-01 -5.06512523e-01
-8.80534574e-02 5.97184412e-02 3.45288627e-02 -5.45784235e-01
1.36415482e+00 -2.33900595e+00 1.67812437e-01 6.02072716e-01
-2.47744918e-01 -4.67745692e-01 3.95807512e-02 7.93172538e-01
-8.02732408e-01 -1.25357369e-02 -3.20434481e-01 -1.08908787e-01
3.07059228e-01 -1.00202456e-01 -5.49438477e-01 9.46427047e-01
-5.17655462e-02 5.11471808e-01 -4.51977611e-01 9.16923285e-02
5.23669943e-02 1.12731583e-01 -2.56454825e-01 -1.36977568e-01
3.10009271e-01 1.10770643e-01 -3.64872605e-01 5.99657595e-02
5.21239936e-01 -2.52441406e-01 2.76489019e-01 1.96851537e-01
-5.21350682e-01 -1.48369864e-01 -1.63065386e+00 1.14476335e+00
-3.46051008e-01 2.90930063e-01 4.65909988e-02 -1.40955722e+00
7.25520790e-01 5.57424307e-01 8.08413744e-01 -4.58462358e-01
1.63305670e-01 4.24591064e-01 -7.30249006e-03 -3.16753894e-01
2.18370974e-01 -7.72015452e-01 -3.31510305e-01 5.86982012e-01
8.07934999e-02 1.96540371e-01 5.29201686e-01 -2.14136131e-02
9.83735621e-01 -7.47436360e-02 7.14550495e-01 -7.59723008e-01
4.59295392e-01 -2.37364501e-01 5.93734086e-01 5.73395431e-01
1.86572775e-01 1.65835455e-01 1.04868329e+00 4.28847224e-02
-1.16485834e+00 -1.14706385e+00 -4.65552956e-01 5.25624096e-01
-4.27470922e-01 -2.04718634e-01 -6.40804768e-01 -1.24918163e-01
6.78155795e-02 3.87093723e-01 -7.30916858e-01 4.98056151e-02
-1.68252960e-01 -8.73094440e-01 2.02539742e-01 4.87960815e-01
-3.20785940e-02 -4.81834382e-01 -5.91603339e-01 2.33522803e-01
2.79361248e-01 -7.84276426e-01 -3.68762195e-01 3.72451097e-01
-9.57135379e-01 -1.25172591e+00 -8.77928734e-01 -2.57260442e-01
5.64558804e-01 1.72787175e-01 8.70064855e-01 -4.11116630e-01
1.03667594e-01 7.61816859e-01 -9.25114974e-02 -3.65456909e-01
-1.01838760e-01 -3.25242251e-01 3.74394655e-01 3.60152245e-01
3.78597081e-01 -9.52034771e-01 -4.60590899e-01 3.17152202e-01
-1.15841866e+00 -5.45479536e-01 5.66277862e-01 6.36300623e-01
3.15067023e-01 2.81869769e-01 3.78457338e-01 -6.25295103e-01
5.92413008e-01 -5.50609171e-01 -1.04114711e+00 7.36506134e-02
-6.94225252e-01 1.28570616e-01 8.87006879e-01 -6.56237304e-01
-8.26624870e-01 -1.76283181e-01 2.25218758e-01 -2.80558199e-01
-1.07293710e-01 9.60892379e-01 -5.83601035e-02 4.11317587e-01
3.85619998e-01 2.27135539e-01 -1.21126696e-01 -6.49793684e-01
9.21024680e-02 5.04511356e-01 3.78560185e-01 -5.38384855e-01
9.23151553e-01 5.70144236e-01 4.85970348e-01 -8.21404278e-01
-2.67604023e-01 -5.67935169e-01 -5.50175548e-01 6.55468926e-02
6.56798661e-01 -8.29600751e-01 -9.76199389e-01 4.71822143e-01
-7.21377969e-01 -1.44548848e-01 -5.46989858e-01 9.26787853e-01
-9.43883955e-01 3.44661862e-01 -5.24514556e-01 -1.37714076e+00
2.64654368e-01 -9.28072095e-01 6.04097188e-01 -5.58915175e-02
-2.69206882e-01 -1.22345710e+00 3.68362337e-01 -2.80462801e-01
2.71159708e-01 1.26978323e-01 1.08661318e+00 -6.40719891e-01
-6.43152446e-02 -5.13702393e-01 2.27190424e-02 4.63871658e-01
8.31241757e-02 1.85331747e-01 -4.36424583e-01 -2.37386152e-01
5.40222824e-01 5.96557617e-01 4.24560159e-01 7.67750144e-01
3.44430327e-01 -3.25567663e-01 -1.44919038e-01 4.26146388e-01
1.68531466e+00 3.09427172e-01 2.24431559e-01 9.92567092e-02
2.54027456e-01 8.08273792e-01 7.15700090e-02 6.89783990e-01
8.78320858e-02 5.02117753e-01 -1.88769519e-01 2.99485922e-01
6.69765949e-01 -1.53257430e-01 5.21105289e-01 1.07947016e+00
-9.75067317e-02 1.29792750e-01 -7.74872482e-01 7.01316357e-01
-1.83747327e+00 -1.10238779e+00 -5.72635353e-01 2.70416045e+00
4.10385966e-01 -1.20229289e-01 5.89729607e-01 1.41020745e-01
6.17366731e-01 -9.43032652e-02 -3.70107204e-01 -3.12026858e-01
-2.56609678e-01 5.23906708e-01 8.96749556e-01 5.12194157e-01
-8.02081108e-01 1.62534863e-01 7.21862984e+00 4.76652205e-01
-8.73876214e-01 9.30185616e-03 2.18281046e-01 1.01282494e-02
-4.14823592e-01 4.00606245e-01 -5.78149974e-01 6.64397776e-01
1.36098087e+00 -5.67950249e-01 2.87516296e-01 1.01618898e+00
5.17962575e-01 -3.68837655e-01 -9.28281665e-01 8.39476764e-01
-3.13628405e-01 -5.30928075e-01 -4.89713043e-01 6.27967119e-01
5.86938322e-01 -4.68153596e-01 2.32250422e-01 6.25057518e-02
3.08128834e-01 -6.79177165e-01 8.62471342e-01 6.71935916e-01
5.41174352e-01 -8.00565302e-01 5.98628819e-01 5.48356116e-01
-1.24854410e+00 -2.18752682e-01 -4.05800343e-01 -4.82658416e-01
3.86142612e-01 8.64348471e-01 1.62157059e-01 7.29759157e-01
4.22518075e-01 5.56032598e-01 -2.15693802e-01 7.85391212e-01
1.18514828e-01 5.44330239e-01 -4.16537255e-01 2.43336469e-01
2.07943797e-01 -1.00251603e+00 5.27500153e-01 9.17994380e-01
8.94317150e-01 5.71694113e-02 -3.13895166e-01 6.31175637e-01
4.87053931e-01 2.83502847e-01 -5.80394387e-01 -2.71040589e-01
4.49417271e-02 9.11479354e-01 -7.28575706e-01 -1.52946129e-01
-8.79664242e-01 5.98603070e-01 -5.86038157e-02 6.35047197e-01
-5.56369960e-01 -2.96341151e-01 7.17084408e-01 2.17380658e-01
5.63658655e-01 -5.61783254e-01 -3.77436280e-02 -1.39606404e+00
4.26737189e-01 -8.18550289e-01 3.71881843e-01 -4.23078626e-01
-1.41284430e+00 2.91298389e-01 3.88902843e-01 -1.09553444e+00
-8.47497404e-01 -8.75105202e-01 -4.02457714e-01 1.18622530e+00
-9.13057864e-01 -4.61844832e-01 4.49096292e-01 5.86741328e-01
-1.70218218e-02 2.05428023e-02 7.75348485e-01 -2.33940855e-02
-7.69999683e-01 9.30408165e-02 7.75250018e-01 7.87928700e-03
2.88323045e-01 -1.23102319e+00 1.10618941e-01 1.20730948e+00
7.50526562e-02 1.14120495e+00 9.92897689e-01 -6.19338870e-01
-1.48738098e+00 -4.34639394e-01 9.66559291e-01 -1.95105270e-01
1.34978795e+00 -2.83121854e-01 -6.83256686e-01 1.04856944e+00
9.13538598e-03 -2.53501385e-01 7.58594930e-01 2.51298308e-01
-2.03895733e-01 -1.55809715e-01 -5.74808300e-01 4.40905541e-01
6.89737558e-01 -4.44908082e-01 -3.89888078e-01 1.00510687e-01
3.02666903e-01 2.40901574e-01 -1.08199108e+00 2.17123643e-01
8.46896410e-01 -1.17094529e+00 7.26770878e-01 -6.82275832e-01
-7.75352344e-02 -1.15796536e-01 -4.62564975e-01 -9.61938620e-01
-4.68198240e-01 -8.21576715e-01 3.73489350e-01 1.20589185e+00
3.15173566e-01 -1.04548800e+00 2.86744773e-01 7.94611156e-01
2.33719170e-01 -2.03845188e-01 -1.10853541e+00 -1.18053341e+00
2.30034724e-01 -6.39127374e-01 3.52043897e-01 9.40532446e-01
1.92171738e-01 2.01607242e-01 -3.67484510e-01 3.25352214e-02
6.39240682e-01 7.35930577e-02 5.01734853e-01 -1.33339763e+00
-8.98323417e-01 -5.80955386e-01 -3.22752178e-01 -9.85641539e-01
3.95942032e-01 -3.62611145e-01 -1.70932531e-01 -8.23422074e-01
1.89077482e-01 -1.28397763e-01 -1.81701422e-01 -3.11845273e-01
4.92192842e-02 -1.82345837e-01 3.74074616e-02 2.49575064e-01
1.78326011e-01 2.43954837e-01 6.37555599e-01 3.82097155e-01
-9.51937661e-02 4.72240955e-01 -7.15587616e-01 8.53420377e-01
3.26288670e-01 -3.04162890e-01 -4.79712844e-01 -8.17692205e-02
6.87479675e-01 5.97088873e-01 3.56625438e-01 -6.12144470e-01
5.63686676e-02 -2.34035894e-01 1.52792081e-01 -2.38337725e-01
8.27192143e-03 -9.60936904e-01 9.28220809e-01 2.85099387e-01
1.84944086e-02 5.54223776e-01 6.35566488e-02 6.45549953e-01
-1.00542851e-01 -6.05959594e-01 4.00602877e-01 -1.75857484e-01
-1.70977503e-01 1.76840290e-01 -6.35252953e-01 -2.94591546e-01
8.06598246e-01 -1.90619722e-01 8.67690220e-02 -7.15538502e-01
-7.94072628e-01 -3.15580308e-01 5.47672689e-01 -1.56626821e-01
-2.52455790e-02 -1.24632776e+00 -3.97670627e-01 3.64406556e-01
-2.77030885e-01 -6.44644082e-01 4.51574206e-01 1.38486314e+00
-5.47735058e-02 7.57787526e-01 -1.94012374e-02 -3.41495007e-01
-6.41062915e-01 1.02989042e+00 1.98218331e-01 -1.18086115e-01
-1.63629740e-01 5.88990226e-02 4.14044023e-01 9.99569520e-03
-1.65415436e-01 -3.00761908e-01 2.23640516e-01 2.78736681e-01
3.98691893e-01 5.04502475e-01 -1.15184039e-01 -8.06332648e-01
-2.50943005e-01 8.48473907e-01 5.25089145e-01 -6.23721719e-01
1.34275627e+00 -4.80050772e-01 -4.28406239e-01 1.03700304e+00
1.28408456e+00 3.55939239e-01 -9.96205568e-01 -1.67185456e-01
1.69705033e-01 -1.14676677e-01 -4.98599499e-01 -1.49173379e-01
-8.02823782e-01 6.26745701e-01 1.64208502e-01 7.21330762e-01
1.17457795e+00 -1.10354938e-01 1.37922630e-01 6.44867495e-03
3.24073732e-01 -6.69682980e-01 -4.76424575e-01 3.57837617e-01
5.27069569e-01 -6.26254439e-01 -2.01187462e-01 -1.98566005e-01
-2.37299174e-01 1.03209722e+00 -2.56772995e-01 -5.47653258e-01
9.11818504e-01 2.27432683e-01 -1.49971575e-01 1.00880824e-01
-5.59905350e-01 -3.64908397e-01 2.00801626e-01 4.60414022e-01
6.12694860e-01 1.13180399e-01 -8.15483510e-01 8.76208425e-01
-3.81746113e-01 -2.82242835e-01 5.47862172e-01 5.31613588e-01
-1.08060986e-01 -1.15120280e+00 -4.84985560e-01 2.49247521e-01
-6.99561298e-01 -2.51652990e-02 -1.24933776e-02 1.18998086e+00
-3.48694921e-01 8.00081789e-01 1.91927209e-01 1.52594343e-01
4.90342885e-01 4.16134626e-01 2.98627675e-01 -2.53872186e-01
-2.15683594e-01 6.84949100e-01 -3.88552934e-01 -3.40526015e-01
-4.72096741e-01 -1.00405049e+00 -6.16161287e-01 -4.99154866e-01
-3.96911830e-01 5.04799426e-01 6.39249504e-01 1.04497159e+00
8.09165556e-03 1.12153485e-01 8.00890863e-01 -3.75981241e-01
-9.07485008e-01 -9.77709115e-01 -1.35913575e+00 3.60692561e-01
2.06168279e-01 -3.97903234e-01 -7.70694733e-01 9.47164372e-02] | [6.4563889503479, 4.159118175506592] |
a4c14d53-da09-419f-948b-8b800ca634a7 | explaining-the-performance-of-collaborative | 2303.11172 | null | https://arxiv.org/abs/2303.11172v1 | https://arxiv.org/pdf/2303.11172v1.pdf | Explaining the Performance of Collaborative Filtering Methods With Optimal Data Characteristics | The performance of a Collaborative Filtering (CF) method is based on the properties of a User-Item Rating Matrix (URM). And the properties or Rating Data Characteristics (RDC) of a URM are constantly changing. Recent studies significantly explained the variation in the performances of CF methods resulted due to the change in URM using six or more RDC. Here, we found that the significant proportion of variation in the performances of different CF techniques can be accounted to two RDC only. The two RDC are the number of ratings per user or Information per User (IpU) and the number of ratings per item or Information per Item (IpI). And the performances of CF algorithms are quadratic to IpU (or IpI) for a square URM. The findings of this study are based on seven well-established CF methods and three popular public recommender datasets: 1M MovieLens, 25M MovieLens, and Yahoo! Music Rating datasets | ['Marwan Bikdash', 'Samin Poudel'] | 2023-03-17 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [-2.62686461e-01 -5.90467632e-01 -2.80977219e-01 -4.44798559e-01
-1.99711561e-01 -7.29789674e-01 6.21756673e-01 2.67912507e-01
-4.93456095e-01 2.12883383e-01 6.71361566e-01 -1.53196409e-01
-7.73603499e-01 -9.47113037e-01 -4.06778336e-01 -3.62964123e-01
-3.49084198e-01 2.42435634e-01 2.55401045e-01 -4.14828688e-01
8.37413251e-01 1.49314255e-01 -1.96502364e+00 5.96519351e-01
6.75758243e-01 1.21265757e+00 2.28441402e-01 3.88717383e-01
-3.69177223e-03 6.26600325e-01 -5.94019473e-01 -4.32160079e-01
5.50942361e-01 -2.08422631e-01 -4.91231054e-01 -2.67459810e-01
1.42043516e-01 -3.86163709e-03 6.05567098e-02 7.20711768e-01
3.47095400e-01 7.20530450e-01 1.05804920e+00 -9.73020315e-01
-1.11017704e+00 8.53614867e-01 -3.10806364e-01 1.96837768e-01
5.71649075e-01 -7.06824362e-01 1.52099431e+00 -1.10775363e+00
4.96363878e-01 1.18412280e+00 6.55986011e-01 2.11765971e-02
-9.13437724e-01 -4.40061569e-01 2.67972291e-01 4.35348861e-02
-1.43081248e+00 -1.35576859e-01 4.43186574e-02 -6.19547069e-01
5.15231252e-01 3.72287452e-01 6.59350276e-01 4.89378124e-01
2.74881601e-01 2.82150775e-01 9.54842567e-01 -2.74804473e-01
4.65906560e-01 5.49632311e-01 3.48731041e-01 -6.10321586e-04
2.94337183e-01 1.12879694e-01 -4.08722848e-01 -4.07185167e-01
9.33134556e-01 3.80109400e-01 -1.31094709e-01 -1.52086526e-01
-8.08334708e-01 1.15599787e+00 2.75485039e-01 3.60723913e-01
-1.56111211e-01 -5.80579758e-01 1.75430372e-01 8.64311814e-01
5.00286579e-01 9.70249176e-01 -6.88020051e-01 -2.99992770e-01
-5.18148422e-01 4.86441597e-04 8.72995973e-01 8.04411471e-01
3.64505857e-01 -3.50632161e-01 -8.71483535e-02 1.43736720e+00
3.48428011e-01 2.85933048e-01 7.83512771e-01 -1.21700764e+00
1.48829997e-01 6.92430019e-01 4.50111002e-01 -1.11550534e+00
-4.80690271e-01 -5.43466091e-01 -5.76264322e-01 -1.10243388e-01
4.72989887e-01 -2.57999122e-01 -1.51474968e-01 1.25520945e+00
1.66969955e-01 -5.45312285e-01 -3.21288973e-01 8.81003082e-01
8.38870585e-01 4.81705487e-01 -4.13680583e-01 -4.46160674e-01
1.06794918e+00 -8.62912178e-01 -5.07712126e-01 1.01002738e-01
5.67917764e-01 -9.72526670e-01 1.08886230e+00 8.26052129e-01
-9.87614453e-01 -9.83750999e-01 -8.90777111e-01 3.67008984e-01
-3.51423025e-01 1.03025034e-01 7.53620267e-01 7.15584457e-01
-7.49949634e-01 9.72197473e-01 -2.93557048e-02 -1.50857493e-01
-1.05300330e-01 1.78818628e-01 -2.47055978e-01 -2.44460136e-01
-1.15510356e+00 8.15385342e-01 -2.27782935e-01 -3.33729953e-01
-2.60188907e-01 -8.98304820e-01 -2.78622776e-01 2.87126880e-02
8.77002999e-02 -4.72441167e-01 8.38163733e-01 -9.51616883e-01
-1.50710022e+00 2.55581364e-02 3.94197702e-01 3.42841633e-02
1.18618108e-01 -4.21275854e-01 -8.78674209e-01 -3.17834705e-01
-2.66539697e-02 3.86423320e-02 5.86520076e-01 -9.25361514e-01
-7.91056275e-01 -4.70087856e-01 1.71078891e-01 3.68961900e-01
-3.09110224e-01 3.32426071e-01 -9.85298380e-02 -6.14833057e-01
8.74831900e-02 -8.21916997e-01 -1.17681667e-01 -5.52696228e-01
5.39341927e-01 -4.20403868e-01 7.07282275e-02 -2.23471031e-01
1.77005970e+00 -2.45176721e+00 1.74958557e-01 3.66784781e-01
-1.12466954e-01 6.02736399e-02 -3.29610407e-01 8.99682462e-01
2.12492824e-01 6.43661991e-02 5.84948659e-01 9.38061625e-02
-8.17339122e-02 3.76677327e-02 -4.57016826e-02 4.00631011e-01
-6.33585691e-01 3.95372093e-01 -8.41632068e-01 3.94555852e-02
-6.75540790e-02 3.12072307e-01 -7.73730874e-01 2.98888505e-01
2.98911333e-01 2.83648729e-01 -6.19321391e-02 1.64543420e-01
5.56930900e-01 -4.02095467e-01 2.63327748e-01 -1.62912682e-01
-3.97852689e-01 6.05517685e-01 -1.51474166e+00 1.37440681e+00
-4.75274295e-01 7.97377005e-02 -2.34770238e-01 -3.05668712e-01
1.05247951e+00 2.18490109e-01 7.22744763e-01 -5.54430664e-01
1.49759874e-01 1.18353032e-01 2.58183360e-01 1.29371524e-01
7.45398462e-01 3.49951625e-01 2.83128202e-01 8.99883389e-01
-7.43354708e-02 2.93877959e-01 4.46805716e-01 4.54095930e-01
1.10747600e+00 -4.51198786e-01 2.77108282e-01 -4.55869794e-01
2.72664607e-01 -5.22053778e-01 5.71318746e-01 9.18636382e-01
1.82319298e-01 5.98469138e-01 2.33782396e-01 -1.37805060e-01
-9.22677755e-01 -1.00279117e+00 -6.07289433e-01 1.57444525e+00
6.61841482e-02 -9.20899570e-01 -7.24471956e-02 -4.39616412e-01
3.45696151e-01 5.91966093e-01 -8.43972445e-01 -1.78462237e-01
1.42718613e-01 -8.44699860e-01 -7.70922676e-02 5.03287911e-01
-9.71671045e-02 -8.23120832e-01 7.04435632e-02 2.35649601e-01
1.17785126e-01 -7.81033456e-01 -8.31102252e-01 -1.40034467e-01
-1.02099872e+00 -8.74271274e-01 -4.79710639e-01 -2.05745682e-01
5.43284059e-01 7.38955677e-01 1.42995834e+00 -3.52984779e-02
2.88010269e-01 5.14086246e-01 -1.15614617e+00 -1.76904023e-01
-6.70986399e-02 9.13071781e-02 4.50250447e-01 1.50238380e-01
2.23359495e-01 -6.51916206e-01 -9.69635367e-01 1.07284641e+00
-6.75552070e-01 -4.53121334e-01 4.03087407e-01 3.15683037e-01
3.52413923e-01 2.98209995e-01 8.68377090e-01 -1.21692002e+00
1.02510142e+00 -8.98069382e-01 -1.56780586e-01 8.28163475e-02
-1.46765816e+00 -2.58187950e-01 5.12879014e-01 -7.17308342e-01
-9.02988076e-01 -5.74118078e-01 7.84415156e-02 3.58106829e-02
3.30491424e-01 6.10594273e-01 2.67285079e-01 2.72705704e-01
8.34365666e-01 -3.94411415e-01 -1.46536231e-01 -9.42835152e-01
6.13011897e-01 1.00865161e+00 1.04783207e-01 -2.99939662e-01
3.69974941e-01 1.50362864e-01 -3.81296128e-01 -3.57814312e-01
-9.86309886e-01 -9.05835032e-01 -6.06710017e-01 -2.38118500e-01
3.97012115e-01 -8.60393584e-01 -6.32153153e-01 1.62031502e-02
-2.80528039e-01 1.57682866e-01 -3.70828211e-01 8.92104387e-01
2.59890147e-02 2.96185970e-01 -6.71698511e-01 -8.89950573e-01
-4.48392242e-01 -7.39569128e-01 3.24094385e-01 1.29777387e-01
-3.33796054e-01 -8.47097754e-01 3.62384588e-01 4.84362662e-01
6.89894736e-01 -4.59742188e-01 9.78940427e-01 -7.54210055e-01
1.93717107e-01 -4.39178079e-01 -1.52563021e-01 6.24319077e-01
2.90579855e-01 2.09036052e-01 -2.36576259e-01 -5.32662451e-01
-7.63781592e-02 1.77682303e-02 3.75868142e-01 4.73853618e-01
6.42122328e-01 -2.26263538e-01 2.15916082e-01 1.17292434e-01
1.35862017e+00 1.90469652e-01 5.85081518e-01 8.85233954e-02
3.24051052e-01 3.37210357e-01 8.90192330e-01 8.81651163e-01
2.58226573e-01 7.63906479e-01 2.74323404e-01 4.51674700e-01
1.68407962e-01 -1.47193208e-01 7.16967344e-01 1.29660416e+00
-5.48643112e-01 1.61273163e-02 -2.15828359e-01 7.84060583e-02
-1.83178616e+00 -7.36936688e-01 -4.74719167e-01 2.75611663e+00
4.49610025e-01 -1.59671485e-01 5.20518780e-01 2.26789236e-01
3.22768539e-01 -2.88998157e-01 -3.33182186e-01 -5.64997315e-01
-1.34528405e-03 1.78429447e-02 4.42525923e-01 1.38467282e-01
-3.33979994e-01 2.64706999e-01 7.09294462e+00 7.39259422e-01
-4.68991548e-01 1.05202012e-01 2.69261524e-02 -4.36453670e-01
-2.59666324e-01 -1.33428976e-01 -7.74755299e-01 6.86105072e-01
1.24065149e+00 1.27679352e-02 8.60242009e-01 7.93748319e-01
1.78808630e-01 -3.98478061e-01 -1.11344874e+00 7.62876034e-01
-3.37869935e-02 -9.69151020e-01 2.22247094e-02 3.49332154e-01
1.14258957e+00 1.98641106e-01 2.11796388e-01 4.27571326e-01
5.72010338e-01 -5.41957140e-01 2.87813842e-01 6.58641398e-01
6.30205214e-01 -6.16664767e-01 7.00137675e-01 3.51600379e-01
-1.07192361e+00 -6.38013422e-01 -8.55435193e-01 -4.40116525e-01
-2.28497803e-01 9.15659904e-01 -2.95872778e-01 4.30119514e-01
1.14651990e+00 9.61851478e-01 -5.81695378e-01 9.94868398e-01
1.71300143e-01 5.42664170e-01 7.29027018e-02 1.05781645e-01
-2.26234823e-01 -8.74521673e-01 1.49281174e-01 7.47889102e-01
2.79426754e-01 2.17816740e-01 -2.61585802e-01 3.16541553e-01
-1.66997865e-01 7.36709654e-01 -1.23600185e-01 6.02199994e-02
8.58718574e-01 1.50466478e+00 -3.79862219e-01 -7.41823763e-02
-7.90676534e-01 5.16154528e-01 2.02384725e-01 -1.89542770e-03
-5.96351445e-01 5.65261813e-03 9.71560240e-01 5.25073290e-01
4.28147584e-01 -4.44100685e-02 -4.83831763e-02 -1.26586950e+00
-4.18805867e-01 -1.08172226e+00 5.20506501e-01 -6.55521810e-01
-1.94257450e+00 3.47710341e-01 -2.87957191e-01 -1.42935348e+00
8.98530632e-02 -3.68448257e-01 -3.57467085e-01 6.66478157e-01
-7.51013994e-01 -3.34257990e-01 1.13559766e-02 6.00449979e-01
3.91664505e-01 -4.39793408e-01 9.54373598e-01 7.00429022e-01
-4.13264096e-01 6.86582327e-01 9.22217131e-01 -1.22742392e-01
1.06778932e+00 -1.15262187e+00 8.06516111e-02 2.23737746e-01
3.57810616e-01 1.11336815e+00 2.75032878e-01 -4.25207019e-01
-1.15425682e+00 -5.93128204e-01 6.19475484e-01 -8.36963296e-01
6.30464435e-01 -1.84869811e-01 -6.03323102e-01 3.44734699e-01
-1.30369261e-01 -3.20545048e-01 1.38704538e+00 9.51047361e-01
-7.37373829e-01 -2.74563730e-01 -1.15865600e+00 2.32749760e-01
1.25423408e+00 -3.44620973e-01 -2.64372408e-01 2.66802788e-01
5.04618287e-01 1.98995277e-01 -1.82505131e+00 3.93344104e-01
1.11888051e+00 -1.24331880e+00 1.01594079e+00 -6.70057535e-01
3.96983385e-01 -2.58514225e-01 -5.60723186e-01 -1.53176892e+00
-1.22804749e+00 -9.70961154e-02 -2.22577333e-01 1.05471575e+00
7.10031033e-01 -5.82068384e-01 3.00173491e-01 9.28310394e-01
9.80790481e-02 -8.42926025e-01 -4.33838189e-01 -6.77346468e-01
2.90270969e-02 -1.54819012e-01 6.78043842e-01 8.35058451e-01
1.69377938e-01 7.34696150e-01 -4.53166634e-01 -2.07855955e-01
6.91623390e-02 1.98167458e-01 6.69382989e-01 -1.82090318e+00
-7.92321980e-01 -2.92254299e-01 1.68055296e-02 -1.03084373e+00
-7.26704895e-01 -7.40895331e-01 -7.84915030e-01 -1.30913270e+00
2.86512852e-01 -7.37901211e-01 -6.38193667e-01 -2.25220993e-01
-3.05702649e-02 1.24359995e-01 4.12515074e-01 8.89809906e-01
-6.60941958e-01 9.94672552e-02 1.18129730e+00 4.30604339e-01
-4.44324911e-01 5.67381442e-01 -1.00057685e+00 5.58943033e-01
4.28733736e-01 -4.74285811e-01 -7.22754240e-01 -3.00821781e-01
1.30255926e+00 1.16016053e-01 -6.69961572e-01 -7.42673278e-01
1.88278958e-01 -6.78081959e-02 3.97312403e-01 -5.58356225e-01
7.27146640e-02 -8.01861584e-01 5.52354395e-01 -8.19432437e-02
-6.16829097e-01 3.12908858e-01 -5.60514688e-01 5.92863321e-01
4.60049994e-02 -4.84017372e-01 6.17493808e-01 -7.75492489e-02
-9.18343440e-02 8.60630721e-02 -3.87208253e-01 -1.16030075e-01
4.44056451e-01 -8.89398754e-02 -2.93948442e-01 -5.82678497e-01
-7.13523507e-01 -2.44244352e-01 2.17943981e-01 9.63701546e-01
2.38382235e-01 -1.40588045e+00 -7.01777399e-01 7.08392113e-02
9.65039134e-02 -6.79533362e-01 1.75875023e-01 6.87887311e-01
2.61328816e-01 2.66769290e-01 -2.96615362e-01 1.23191833e-01
-1.05483699e+00 6.89377189e-01 3.05641219e-02 -3.33501160e-01
-7.01487437e-02 6.30695820e-01 -1.25416726e-01 -4.16579038e-01
6.97257668e-02 -6.03894368e-02 -8.47392023e-01 5.90779543e-01
7.46139467e-01 9.77391243e-01 1.87805593e-01 -4.59829986e-01
-4.32528146e-02 4.50360984e-01 -3.21247727e-01 3.66353579e-02
1.34956336e+00 -4.57347453e-01 -9.36499387e-02 8.28694403e-01
1.00300622e+00 4.00774628e-01 -7.48555779e-01 -3.72440100e-01
-4.47648317e-01 -6.81410849e-01 3.67548466e-02 -1.04452634e+00
-1.14988613e+00 2.88619045e-02 5.54289639e-01 6.07861459e-01
9.12826300e-01 -2.42643327e-01 3.93140793e-01 -5.89423738e-02
7.57064342e-01 -1.26623929e+00 1.31808883e-02 4.70457405e-01
6.34868920e-01 -1.03082418e+00 3.96604151e-01 -1.71380356e-01
-8.49973500e-01 6.51308477e-01 2.77671307e-01 -1.06995322e-01
1.28773665e+00 -2.35690400e-01 -1.11622922e-01 3.06188632e-02
-9.76388693e-01 8.06113183e-02 7.55338430e-01 2.99844682e-01
9.47096705e-01 3.07291716e-01 -1.06005716e+00 1.12479436e+00
-3.21282476e-01 -2.58494914e-01 5.08993804e-01 4.18472677e-01
-5.99330902e-01 -1.38343573e+00 -1.20117381e-01 9.48607266e-01
-5.32863677e-01 -2.29108155e-01 -3.90134811e-01 3.50156724e-01
1.92550316e-01 1.59708118e+00 3.02921794e-02 -9.13636923e-01
3.14876288e-01 -2.65997440e-01 2.79934883e-01 -8.85913789e-01
-1.08999002e+00 -2.39380077e-01 5.05862795e-02 -5.82200468e-01
-3.33320528e-01 -8.85120511e-01 -9.42016780e-01 -5.58695436e-01
-8.27761233e-01 7.12132156e-01 8.16837013e-01 6.39539123e-01
6.99675918e-01 4.60021645e-02 1.28178012e+00 -1.95624232e-01
-6.12120330e-01 -1.28847742e+00 -1.33298516e+00 7.35135734e-01
-5.17059684e-01 -8.42936754e-01 -7.26575792e-01 -3.78098518e-01] | [10.026398658752441, 5.731016635894775] |
5b145131-e3b6-41eb-ab7f-da4ab0cc9a52 | weakly-supervised-contrastive-learning-1 | 2110.04770 | null | https://arxiv.org/abs/2110.04770v1 | https://arxiv.org/pdf/2110.04770v1.pdf | Weakly Supervised Contrastive Learning | Unsupervised visual representation learning has gained much attention from the computer vision community because of the recent achievement of contrastive learning. Most of the existing contrastive learning frameworks adopt the instance discrimination as the pretext task, which treating every single instance as a different class. However, such method will inevitably cause class collision problems, which hurts the quality of the learned representation. Motivated by this observation, we introduced a weakly supervised contrastive learning framework (WCL) to tackle this issue. Specifically, our proposed framework is based on two projection heads, one of which will perform the regular instance discrimination task. The other head will use a graph-based method to explore similar samples and generate a weak label, then perform a supervised contrastive learning task based on the weak label to pull the similar images closer. We further introduced a K-Nearest Neighbor based multi-crop strategy to expand the number of positive samples. Extensive experimental results demonstrate WCL improves the quality of self-supervised representations across different datasets. Notably, we get a new state-of-the-art result for semi-supervised learning. With only 1\% and 10\% labeled examples, WCL achieves 65\% and 72\% ImageNet Top-1 Accuracy using ResNet50, which is even higher than SimCLRv2 with ResNet101. | ['Chang Xu', 'Xiaogang Wang', 'ChangShui Zhang', 'Chen Qian', 'Shan You', 'Fei Wang', 'Mingkai Zheng'] | 2021-10-10 | weakly-supervised-contrastive-learning | http://openaccess.thecvf.com//content/ICCV2021/html/Zheng_Weakly_Supervised_Contrastive_Learning_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Zheng_Weakly_Supervised_Contrastive_Learning_ICCV_2021_paper.pdf | iccv-2021-1 | ['self-supervised-image-classification', 'semi-supervised-image-classification'] | ['computer-vision', 'computer-vision'] | [ 5.33198535e-01 9.80469659e-02 -4.88206565e-01 -2.08943784e-01
-4.90003884e-01 -2.80490220e-01 7.48947859e-01 2.73338616e-01
-4.18897092e-01 6.37425005e-01 -2.35979501e-02 -6.75157532e-02
-1.37430608e-01 -8.90467465e-01 -6.76291943e-01 -9.85069215e-01
1.38026699e-01 2.26865381e-01 4.10122067e-01 -5.91166504e-02
2.22998455e-01 1.64794281e-01 -1.74273825e+00 3.36306006e-01
1.05846107e+00 1.04854894e+00 3.62795949e-01 3.86092924e-02
-3.68267953e-01 1.00766838e+00 -2.50022769e-01 -3.72197539e-01
3.50741923e-01 -5.23397505e-01 -8.13773513e-01 3.87548923e-01
4.78191108e-01 1.76469535e-01 -6.14324361e-02 1.31027079e+00
3.59406531e-01 6.81709871e-03 7.70334601e-01 -1.42328644e+00
-6.60231709e-01 6.03327334e-01 -1.12694669e+00 2.19545782e-01
-9.75921843e-03 1.29710928e-01 1.05043757e+00 -1.03532565e+00
7.01444149e-01 1.11175394e+00 4.37865645e-01 3.92582774e-01
-1.23059416e+00 -9.78967905e-01 3.86388570e-01 3.43626767e-01
-1.37746990e+00 -1.16125032e-01 1.16034055e+00 -3.05322349e-01
4.14814711e-01 1.79515928e-02 6.90652728e-01 8.25082660e-01
-1.75989717e-01 9.24155951e-01 1.52842903e+00 -5.86639047e-01
1.66993961e-01 3.07121426e-01 3.50880325e-01 7.72764087e-01
1.89080253e-01 7.99823999e-02 -3.58069956e-01 1.60143062e-01
4.64204788e-01 4.11908478e-01 -3.26708198e-01 -6.18723750e-01
-1.05194855e+00 8.71866584e-01 1.06987202e+00 3.33944887e-01
-2.41930947e-01 -4.40753609e-01 4.28737611e-01 2.45757774e-01
5.23188889e-01 2.27586806e-01 -9.60350782e-02 5.72000861e-01
-9.05199051e-01 -1.40163591e-02 2.36488551e-01 7.27688670e-01
1.03200579e+00 -3.94870080e-02 -1.68822378e-01 1.04965949e+00
2.36868784e-01 3.08478355e-01 5.45132041e-01 -3.51561368e-01
5.21817148e-01 1.08309996e+00 -4.15188879e-01 -1.23059273e+00
-2.38297835e-01 -8.77339900e-01 -1.28609753e+00 4.28927958e-01
3.11921299e-01 1.79356739e-01 -1.24656808e+00 1.63083696e+00
2.93566436e-01 3.96226436e-01 1.81506708e-01 8.55275035e-01
9.31303322e-01 5.86264074e-01 3.32018733e-01 -5.20701826e-01
1.08369291e+00 -1.16839266e+00 -4.23345149e-01 -2.59163827e-01
6.12010181e-01 -6.68000042e-01 9.96195078e-01 3.21839124e-01
-7.99040973e-01 -8.03454280e-01 -1.28039885e+00 3.36351544e-01
-4.22765225e-01 -3.51034030e-02 5.31289697e-01 3.49088401e-01
-6.71170235e-01 4.12114948e-01 -3.15951854e-01 -1.17460921e-01
8.80725026e-01 1.24576949e-02 -4.66340661e-01 -5.52425385e-01
-9.57313716e-01 5.34683108e-01 7.38506258e-01 7.06806034e-02
-7.16984928e-01 -5.80225348e-01 -8.49097490e-01 -5.39778173e-02
4.77360815e-01 -2.02015296e-01 6.09576166e-01 -1.41881406e+00
-9.14591551e-01 1.12566423e+00 2.74585765e-02 -4.07155037e-01
4.89692390e-01 -9.57462378e-03 -4.81974483e-01 1.93580031e-01
3.26401979e-01 7.57942140e-01 9.88491297e-01 -1.62244904e+00
-6.68236196e-01 -4.66543913e-01 -1.05571732e-01 4.73094672e-01
-4.30647165e-01 -3.30881208e-01 -4.75482732e-01 -7.32617974e-01
3.04740667e-01 -9.74798799e-01 -2.37842277e-01 1.89225525e-02
-3.87382507e-01 -4.05180395e-01 9.84322667e-01 -2.46124759e-01
1.09860480e+00 -2.29657030e+00 1.11147955e-01 1.81672439e-01
3.87504905e-01 5.68348587e-01 -2.90261686e-01 1.19394861e-01
-4.18060720e-01 6.45144563e-03 -5.92418611e-01 -1.43577009e-01
-4.61008132e-01 5.31972013e-02 -3.93883646e-01 3.38602334e-01
4.22249287e-01 8.42119515e-01 -1.11050713e+00 -7.18953967e-01
2.51412749e-01 3.11869562e-01 -3.64098400e-01 2.99095213e-01
-1.15040541e-02 3.45463455e-01 -2.26557627e-01 5.06991744e-01
9.89214540e-01 -4.75143045e-01 2.36064225e-01 -3.00062954e-01
2.07267534e-02 -2.51504421e-01 -1.27764487e+00 1.59543824e+00
-1.37361929e-01 3.10900480e-01 -2.39496022e-01 -1.40429938e+00
1.03731120e+00 -1.79914683e-01 3.12361240e-01 -9.09021735e-01
-1.29477277e-01 1.75756291e-01 -2.04443708e-02 -2.87918389e-01
2.40390658e-01 -1.93186477e-01 2.33684689e-01 2.13790238e-01
-1.40328119e-02 -8.01272541e-02 1.42833307e-01 2.89032310e-01
6.74405277e-01 1.55180797e-01 4.61639792e-01 -3.72147977e-01
6.13332331e-01 6.12963624e-02 7.80833066e-01 5.57777464e-01
-2.18186975e-01 5.60578227e-01 3.19984645e-01 -4.47409004e-01
-7.06253529e-01 -1.02740335e+00 -2.37213448e-01 1.05775213e+00
5.57092071e-01 -2.55566806e-01 -6.11363769e-01 -1.12094831e+00
-1.26045406e-01 4.85869914e-01 -7.39988863e-01 -3.26759726e-01
-4.96444345e-01 -1.01771331e+00 4.12538797e-02 4.80521977e-01
9.98019457e-01 -1.26043630e+00 -2.76348650e-01 -3.16444524e-02
-3.52982171e-02 -9.51325536e-01 -1.39612228e-01 3.63329083e-01
-8.26822400e-01 -1.35428870e+00 -8.97802293e-01 -1.24950802e+00
1.00529182e+00 6.65848672e-01 1.00633836e+00 2.92397678e-01
-2.21105009e-01 -1.14153028e-02 -4.65239614e-01 -2.67636865e-01
-1.36903837e-01 -2.56873239e-02 -5.72598800e-02 3.13863128e-01
3.39439213e-01 -5.29279530e-01 -7.42300868e-01 3.40326071e-01
-9.32166755e-01 3.25300455e-01 7.74484873e-01 9.64349151e-01
9.02387321e-01 1.42451525e-01 6.87523127e-01 -1.06953168e+00
2.87772983e-01 -5.49272597e-01 -2.96543568e-01 2.86041737e-01
-8.43460262e-01 -2.50744633e-04 8.21691811e-01 -6.36549711e-01
-1.00626671e+00 1.76738232e-01 2.53412277e-01 -4.83886272e-01
-7.83385038e-02 3.83051336e-01 -1.28656432e-01 -2.07080157e-03
6.46691561e-01 6.47444367e-01 -9.69625041e-02 -1.95318207e-01
3.56924117e-01 6.40928686e-01 3.96522880e-01 -2.56276995e-01
9.37797725e-01 4.00004804e-01 -1.90769937e-02 -7.99497783e-01
-1.04372978e+00 -4.55844283e-01 -5.47514737e-01 -2.51677483e-01
8.40301991e-01 -1.03922570e+00 -1.24011122e-01 5.78864694e-01
-6.66268587e-01 -2.85484016e-01 -2.56314665e-01 3.40342462e-01
-2.77064085e-01 4.97366369e-01 -2.38818631e-01 -6.90504670e-01
-3.83851409e-01 -1.14091587e+00 8.02665174e-01 5.16936541e-01
3.07769954e-01 -6.86103582e-01 -7.59012531e-03 3.08396012e-01
2.21455961e-01 3.06489587e-01 9.80436862e-01 -6.65245473e-01
-4.46728796e-01 9.38565878e-04 -6.49230480e-01 4.89422172e-01
1.76137030e-01 -3.00201595e-01 -1.05983377e+00 -4.63499486e-01
-2.14405417e-01 -6.40233099e-01 1.21072459e+00 1.24147259e-01
1.32982504e+00 -8.30779970e-02 -4.67530400e-01 5.19966066e-01
1.57038164e+00 3.13340612e-02 6.81581795e-01 2.98215747e-01
9.71218050e-01 5.38794637e-01 7.23767400e-01 7.61265978e-02
2.41210714e-01 5.87503672e-01 5.72457016e-01 -3.50610882e-01
-4.24941570e-01 -3.30480158e-01 4.40149531e-02 8.30408990e-01
-1.49912417e-01 6.78946078e-02 -8.19310546e-01 5.21055639e-01
-1.78290033e+00 -8.56585145e-01 4.55428585e-02 2.24628615e+00
8.34623873e-01 4.13645118e-01 1.45822167e-01 4.46090251e-01
9.86003816e-01 3.48062605e-01 -5.72165549e-01 2.92204142e-01
-2.06839979e-01 1.45848095e-01 2.96338379e-01 2.08498463e-02
-1.22192669e+00 9.46438253e-01 5.08996916e+00 1.06899250e+00
-1.31624770e+00 -9.39930752e-02 9.99046147e-01 3.23401004e-01
-1.21617399e-01 -3.71671505e-02 -6.49704158e-01 5.90940654e-01
1.75971150e-01 1.70119554e-02 1.36109442e-01 8.86195123e-01
-2.07598329e-01 -4.45337407e-02 -7.53542125e-01 1.18935740e+00
1.60753861e-01 -1.22834539e+00 4.11746264e-01 -1.25525236e-01
9.20939684e-01 -1.11727990e-01 7.50191733e-02 5.45045078e-01
1.25520676e-01 -9.51584101e-01 3.57528239e-01 3.93884033e-01
7.81483710e-01 -8.96781623e-01 6.00703835e-01 4.17522043e-01
-1.56052327e+00 -2.26034969e-01 -5.44929922e-01 7.45193213e-02
-1.61897749e-01 6.60064220e-01 -5.25845289e-01 6.77578986e-01
8.02737474e-01 9.98841047e-01 -8.69437158e-01 1.13239944e+00
-2.37168655e-01 6.47865236e-01 5.57785295e-02 9.37466174e-02
3.16791773e-01 -1.48234889e-01 3.21164191e-01 1.09896278e+00
-1.00379370e-01 2.22398713e-02 6.22566402e-01 5.42906046e-01
-2.96109796e-01 3.58532786e-01 -6.67888999e-01 2.92359829e-01
3.79032791e-01 1.37583542e+00 -9.64890778e-01 -5.14577985e-01
-3.51417303e-01 1.00188136e+00 7.07129538e-01 3.13424021e-01
-6.35174155e-01 -3.92758101e-01 7.17678145e-02 1.36127427e-01
3.32824975e-01 1.42944694e-01 -9.73201990e-02 -1.09406650e+00
1.04186885e-01 -8.50326419e-01 6.01213276e-01 -6.03607178e-01
-1.39719975e+00 6.96619391e-01 -9.30721611e-02 -1.66580296e+00
-1.67791434e-02 -4.54594046e-01 -7.48033702e-01 4.91576105e-01
-2.03707910e+00 -1.20608246e+00 -6.12292051e-01 6.48266017e-01
7.50162780e-01 -3.00090075e-01 6.27871633e-01 1.23618864e-01
-6.09661222e-01 6.93592072e-01 -1.23889521e-01 2.68284857e-01
8.01957190e-01 -1.27008271e+00 -1.45140678e-01 7.67466009e-01
3.21475536e-01 3.64804566e-01 3.07407975e-01 -5.52139282e-01
-9.77831483e-01 -1.39797187e+00 5.38453281e-01 4.14790288e-02
3.65391850e-01 -1.32866472e-01 -1.15322340e+00 2.67656684e-01
2.17040092e-01 5.31065762e-01 5.62805951e-01 -1.63344666e-01
-6.06697023e-01 -3.89579684e-01 -1.12918901e+00 5.09423196e-01
1.20332336e+00 -4.22270983e-01 -6.67424500e-01 3.38870466e-01
6.80384338e-01 6.13677595e-03 -5.56458235e-01 6.77959442e-01
2.93799371e-01 -1.00645876e+00 9.47225332e-01 -4.41989243e-01
6.40949726e-01 -4.15307343e-01 -1.29112333e-01 -1.30232942e+00
-4.84635323e-01 -1.12161823e-01 -4.94499924e-03 1.29999077e+00
3.07169944e-01 -5.39006412e-01 6.81250095e-01 9.33419447e-03
6.37855679e-02 -8.48441601e-01 -5.63248038e-01 -8.53397250e-01
1.21294394e-01 -1.69372156e-01 2.81650007e-01 1.36401808e+00
-1.78805947e-01 5.26489019e-01 -3.03387195e-01 8.19003582e-02
9.02256012e-01 4.29343313e-01 6.48733616e-01 -1.38642025e+00
-1.47411317e-01 -3.75066727e-01 -5.03914833e-01 -7.76464105e-01
1.55537650e-01 -1.24216413e+00 -5.62996753e-02 -1.31470919e+00
6.25487506e-01 -6.21706963e-01 -7.19065070e-01 5.82170844e-01
-5.47953427e-01 6.71054363e-01 3.54409337e-01 4.49290574e-01
-8.11673760e-01 6.22534633e-01 1.26244378e+00 -4.53907728e-01
-2.30820298e-01 -2.27998644e-01 -7.86772430e-01 7.22883046e-01
7.20026791e-01 -4.59965527e-01 -5.78528285e-01 -1.50210947e-01
-1.35244638e-01 -3.82457286e-01 3.63404661e-01 -1.20647752e+00
1.10320240e-01 -1.90348580e-01 7.02278674e-01 -3.90269756e-01
-2.96291690e-02 -7.37170935e-01 -2.44592860e-01 5.72988808e-01
-4.02702928e-01 -1.72443137e-01 1.09031960e-01 7.92194068e-01
-3.73725265e-01 -1.25297353e-01 1.07865202e+00 -1.66010916e-01
-1.08541238e+00 4.72811699e-01 2.88527459e-01 2.05268547e-01
1.17544675e+00 -2.93181509e-01 -3.61215085e-01 7.79252499e-02
-5.27946413e-01 1.93617776e-01 3.23209405e-01 4.41380978e-01
7.44673908e-01 -1.44099760e+00 -8.17806602e-01 2.78236002e-01
5.94832480e-01 1.19995259e-01 2.40104452e-01 6.15950167e-01
-8.59240741e-02 -1.28011137e-01 -2.79132694e-01 -9.08817291e-01
-1.31688690e+00 7.93542981e-01 6.93332870e-04 -4.48500842e-01
-7.29153156e-01 6.22469306e-01 5.18919945e-01 -3.43611121e-01
1.32182553e-01 2.71465838e-01 -6.63730621e-01 3.01917434e-01
7.05175102e-01 1.24500044e-01 -1.16872817e-01 -7.04454422e-01
-4.94366348e-01 8.05135071e-01 -4.82133687e-01 3.26610833e-01
1.38502634e+00 3.82391065e-02 5.64500503e-02 3.66924196e-01
1.25783777e+00 -1.37174398e-01 -1.24738681e+00 -5.61657190e-01
-1.09966516e-01 -3.28109771e-01 1.19554803e-01 -7.53609896e-01
-1.24937522e+00 7.66283512e-01 9.14806306e-01 2.33045772e-01
1.34001696e+00 -2.31479108e-02 4.64056104e-01 3.18648219e-01
2.62177438e-01 -9.85595226e-01 2.99217939e-01 2.84098178e-01
7.41595209e-01 -1.65061867e+00 1.05397083e-01 -5.64567626e-01
-7.34353185e-01 8.71985734e-01 7.94144630e-01 -3.48922461e-01
5.44692278e-01 -1.57316417e-01 7.17555434e-02 -2.10784242e-01
-3.17584097e-01 -4.46414769e-01 3.87388140e-01 6.27891362e-01
2.63249993e-01 2.82042827e-02 -4.32821274e-01 3.54253709e-01
2.14391187e-01 -2.52366569e-02 8.79835114e-02 9.56225634e-01
-4.73616749e-01 -8.72988224e-01 -6.76410943e-02 5.21683514e-01
-2.21286610e-01 -1.62502840e-01 -3.31193656e-01 7.85680711e-01
1.68814242e-01 8.34306180e-01 -1.53355636e-02 -5.18018186e-01
1.98290199e-01 -1.94561202e-02 3.91717255e-01 -4.98981327e-01
-4.90684748e-01 -8.03942829e-02 -3.58791739e-01 -4.30393636e-01
-7.69388199e-01 -3.02425057e-01 -1.16612160e+00 1.71424672e-01
-3.95296663e-01 7.56738484e-02 1.22985467e-01 1.00171447e+00
2.13036895e-01 3.92956525e-01 1.08956778e+00 -7.54482090e-01
-3.25816512e-01 -9.02495980e-01 -5.05194068e-01 7.67042577e-01
1.49079695e-01 -8.12830508e-01 -4.91855949e-01 -3.44051346e-02] | [9.516084671020508, 2.761770486831665] |
5db2cbbb-31bc-44aa-93c7-1b534e893a26 | approximating-dtw-with-a-convolutional-neural | 2301.12873 | null | https://arxiv.org/abs/2301.12873v1 | https://arxiv.org/pdf/2301.12873v1.pdf | Approximating DTW with a convolutional neural network on EEG data | Dynamic Time Wrapping (DTW) is a widely used algorithm for measuring similarities between two time series. It is especially valuable in a wide variety of applications, such as clustering, anomaly detection, classification, or video segmentation, where the time-series have different timescales, are irregularly sampled, or are shifted. However, it is not prone to be considered as a loss function in an end-to-end learning framework because of its non-differentiability and its quadratic temporal complexity. While differentiable variants of DTW have been introduced by the community, they still present some drawbacks: computing the distance is still expensive and this similarity tends to blur some differences in the time-series. In this paper, we propose a fast and differentiable approximation of DTW by comparing two architectures: the first one for learning an embedding in which the Euclidean distance mimics the DTW, and the second one for directly predicting the DTW output using regression. We build the former by training a siamese neural network to regress the DTW value between two time-series. Depending on the nature of the activation function, this approximation naturally supports differentiation, and it is efficient to compute. We show, in a time-series retrieval context on EEG datasets, that our methods achieve at least the same level of accuracy as other DTW main approximations with higher computational efficiency. We also show that it can be used to learn in an end-to-end setting on long time series by proposing generative models of EEGs. | ['Laurent Heutte', 'Alain Rakotomamonjy', 'Romain Picot-Clemente', 'Hugo Lerogeron'] | 2023-01-30 | null | null | null | null | ['video-semantic-segmentation'] | ['computer-vision'] | [ 7.56343231e-02 -2.12346137e-01 1.00066036e-01 -3.32969069e-01
-6.07073665e-01 -6.27143383e-01 5.76875210e-01 4.30092484e-01
-6.15821600e-01 4.61660802e-01 -5.07446192e-02 -2.09987283e-01
-6.25942111e-01 -5.23203135e-01 -6.05467558e-01 -9.42650378e-01
-5.77685058e-01 3.42732459e-01 2.31164679e-01 -1.47475615e-01
1.20944656e-01 6.81514084e-01 -1.44896686e+00 -1.08222708e-01
8.74406874e-01 1.28243649e+00 6.09224848e-02 3.57390851e-01
1.39961960e-02 4.44033086e-01 -5.99874020e-01 -2.73678243e-01
3.38435650e-01 -4.97448087e-01 -6.40868962e-01 -1.47876352e-01
1.52912229e-01 1.33088052e-01 -4.47681069e-01 9.42821801e-01
4.87804681e-01 4.23469186e-01 7.27907956e-01 -1.44126987e+00
-5.08414865e-01 3.73785079e-01 -4.19221342e-01 4.19369757e-01
2.43663996e-01 -1.86274812e-01 9.38633382e-01 -6.68687761e-01
5.70023119e-01 8.53932321e-01 1.04777050e+00 4.17473674e-01
-1.57509255e+00 -3.59933287e-01 4.63912115e-02 3.61224353e-01
-1.28202152e+00 -6.06224574e-02 9.71816361e-01 -4.56126839e-01
7.15632319e-01 3.31060708e-01 8.27592909e-01 1.08578777e+00
3.16880465e-01 7.86328077e-01 1.14802766e+00 -1.32276192e-01
4.67570513e-01 -1.84331492e-01 1.24677941e-01 3.02124172e-01
-3.68110478e-01 -7.05634207e-02 -2.60569841e-01 -1.76886469e-01
5.27932823e-01 2.83284068e-01 -5.49474239e-01 -5.74656427e-01
-1.24691081e+00 7.79063940e-01 4.86908764e-01 8.14445376e-01
-4.22782421e-01 1.36748835e-01 5.62880635e-01 7.78034031e-01
8.01580369e-01 5.19911587e-01 -3.20574820e-01 -4.31390613e-01
-1.25265074e+00 3.35266113e-01 6.85438275e-01 4.32103872e-01
4.35313165e-01 -6.19582757e-02 -7.56110102e-02 6.77479506e-01
-1.79267198e-01 6.33236691e-02 1.08878469e+00 -7.51476586e-01
2.00014681e-01 2.58388221e-01 -6.26397654e-02 -9.80128109e-01
-6.43816888e-01 -3.42945129e-01 -9.65452075e-01 2.09042892e-01
7.52780974e-01 4.71586809e-02 -6.35710537e-01 1.93649054e+00
8.27391818e-02 4.78332311e-01 -2.52222896e-01 8.41972172e-01
1.79364741e-01 6.72716439e-01 -2.98228413e-01 -4.44529593e-01
9.73942876e-01 -7.08067656e-01 -7.63243496e-01 1.57576472e-01
4.86836493e-01 -5.69946945e-01 1.07951009e+00 5.13067245e-01
-1.17314017e+00 -4.34182107e-01 -1.06713450e+00 -7.84522295e-02
-6.78908646e-01 -9.60929245e-02 3.73127729e-01 3.28740001e-01
-1.14342618e+00 1.36974657e+00 -1.17665231e+00 -4.34983581e-01
2.49990776e-01 2.34383062e-01 -1.68284088e-01 4.56588089e-01
-1.08091688e+00 9.31015670e-01 1.99747473e-01 -8.50688666e-02
-4.12624568e-01 -8.88759851e-01 -5.64863563e-01 1.28316000e-01
5.84625499e-03 -3.53764534e-01 9.44050610e-01 -1.01364696e+00
-1.55166006e+00 7.53972471e-01 1.54961143e-02 -8.47220063e-01
9.00952995e-01 1.78444125e-02 -6.02501035e-01 1.00814968e-01
-7.32906628e-04 2.21823484e-01 1.16879511e+00 -4.87954944e-01
-2.83391565e-01 -5.38668096e-01 1.52333477e-03 -1.80518493e-01
-4.18253571e-01 -2.23956093e-01 -9.41843465e-02 -1.06859469e+00
1.91384017e-01 -9.49408531e-01 -1.36200279e-01 2.95080394e-01
2.29510758e-02 -4.62443918e-01 8.39770973e-01 -4.65201735e-01
1.20902920e+00 -2.41271091e+00 2.79033393e-01 2.33008936e-01
2.34429374e-01 5.97715341e-02 -2.07096487e-01 5.95117509e-01
-4.85282123e-01 -2.08226651e-01 -5.26999772e-01 -2.52693444e-01
8.25806409e-02 1.24470718e-01 -5.71716607e-01 7.80049980e-01
3.73398423e-01 7.61106968e-01 -1.08953989e+00 4.90952283e-02
6.20880127e-02 7.07366049e-01 -3.68889660e-01 1.14224508e-01
6.89413771e-02 4.26536590e-01 -1.58244763e-02 -3.35222147e-02
3.70392591e-01 -1.36366552e-02 -1.90448776e-01 -1.70424819e-01
-1.21540464e-01 2.23371089e-01 -1.06397581e+00 1.94149709e+00
-4.91183400e-01 1.01202905e+00 -3.64387751e-01 -1.50048566e+00
8.19646418e-01 3.81967753e-01 9.41640139e-01 -6.92858279e-01
2.42944770e-02 3.69013995e-01 2.88363937e-02 -4.45086867e-01
2.49092638e-01 -1.33562610e-01 1.70290247e-01 5.78553081e-01
1.77951083e-01 -1.15786634e-01 3.71572345e-01 -2.30378211e-01
1.11409795e+00 2.07359061e-01 7.19512478e-02 -3.79637569e-01
4.47774500e-01 -5.31398833e-01 2.16305301e-01 2.86253303e-01
-1.11634322e-02 8.02361131e-01 6.52513802e-01 -4.85292345e-01
-9.99311149e-01 -1.19786787e+00 -3.71457279e-01 7.60853648e-01
-7.93615356e-02 -3.09533000e-01 -7.79608548e-01 -4.89956975e-01
2.19572848e-03 6.07696354e-01 -6.91436946e-01 -4.53406751e-01
-5.72144687e-01 -6.24378979e-01 5.47590494e-01 5.68007886e-01
1.66852579e-01 -7.78790116e-01 -9.22285616e-01 4.45832729e-01
-6.86839670e-02 -8.93291712e-01 -8.00697327e-01 4.92543429e-01
-1.23510861e+00 -8.56879830e-01 -1.09974861e+00 -6.29634798e-01
5.45152962e-01 3.25950906e-02 8.81184518e-01 -3.65130156e-01
-2.15129972e-01 5.33975482e-01 -1.90713212e-01 -2.76900947e-01
-1.06767334e-01 -1.02593653e-01 1.49197176e-01 3.87558788e-01
2.18505993e-01 -1.05506182e+00 -6.29513860e-01 2.81638265e-01
-1.14107311e+00 -3.90099823e-01 1.54809728e-01 9.41418231e-01
6.64649427e-01 1.50507912e-01 5.75245917e-01 -5.04318058e-01
9.64445114e-01 -4.55957860e-01 -5.99227190e-01 2.21401781e-01
-7.50116348e-01 3.66162121e-01 9.65970159e-01 -8.24371636e-01
-3.24295640e-01 -2.11315751e-01 -2.51141816e-01 -6.71808183e-01
7.53988177e-02 5.44467747e-01 3.03756267e-01 -7.30717182e-02
5.65842509e-01 3.22205514e-01 1.04592644e-01 -5.64775705e-01
3.18159580e-01 3.13086987e-01 6.20867133e-01 -4.60254997e-01
7.18300521e-01 4.53603864e-01 2.27199961e-02 -8.21143687e-01
-5.22481322e-01 -4.76080596e-01 -6.57917142e-01 -1.84543990e-02
6.41386926e-01 -4.05724049e-01 -6.13823295e-01 3.10294092e-01
-1.04105985e+00 -3.22406918e-01 -8.03307295e-01 6.05470598e-01
-8.54528308e-01 4.26679909e-01 -4.56954002e-01 -6.15359306e-01
-2.25620002e-01 -9.21710730e-01 8.54572296e-01 -2.34690197e-02
-3.37736160e-01 -1.33843470e+00 2.21089467e-01 -4.91379380e-01
5.22294521e-01 5.58105290e-01 1.08728921e+00 -7.92791426e-01
-5.83460147e-04 -4.04229105e-01 1.65259227e-01 3.47153991e-01
1.05981782e-01 -3.96267772e-02 -9.51592028e-01 -4.37439263e-01
5.17573535e-01 -1.40076587e-02 7.31100202e-01 4.21979100e-01
1.53343213e+00 -2.69861549e-01 1.21241165e-02 8.01982284e-01
1.27510011e+00 3.06161523e-01 4.94022518e-01 2.11257979e-01
2.22853214e-01 6.10934436e-01 2.73399681e-01 3.52954507e-01
4.61827703e-02 7.48166621e-01 2.49211058e-01 1.42240077e-01
1.16549775e-01 -3.17208692e-02 5.33139944e-01 1.01986444e+00
-3.76498252e-01 2.64605135e-01 -7.36222982e-01 5.24390280e-01
-1.93445873e+00 -1.17954600e+00 1.38883308e-01 2.59015322e+00
8.29626858e-01 1.46490350e-01 4.77853507e-01 5.80942690e-01
3.66052836e-01 2.52474844e-01 -6.50353789e-01 -3.77179086e-01
-1.34486347e-01 5.79027176e-01 1.41243100e-01 1.92535922e-01
-9.61499214e-01 2.62222022e-01 5.82519293e+00 8.88014257e-01
-1.56654513e+00 2.68773019e-01 4.61437404e-01 -1.98222682e-01
-1.73226699e-01 -3.04056764e-01 6.26418144e-02 6.96900606e-01
1.10818815e+00 -4.76549357e-01 6.21072531e-01 6.56684935e-01
2.14922026e-01 2.24328741e-01 -1.48099411e+00 1.33401167e+00
-1.41110167e-01 -1.09377944e+00 -3.27180922e-01 -4.54004109e-02
4.09650773e-01 1.23842843e-01 1.98958546e-01 1.69253990e-01
-3.97678196e-01 -9.23407793e-01 8.53715837e-01 4.83980000e-01
5.20327330e-01 -7.73559511e-01 3.78377169e-01 2.91670382e-01
-1.17448151e+00 -7.20917434e-02 -3.31852496e-01 3.47057544e-02
8.11670441e-03 7.33964562e-01 -4.11952972e-01 5.14792085e-01
6.25074208e-01 8.29088509e-01 -4.52616751e-01 1.29363036e+00
8.38465914e-02 3.65449280e-01 -3.53072256e-01 -9.48996246e-02
4.12052989e-01 -5.83484411e-01 6.86739385e-01 1.18380404e+00
6.71312213e-01 -2.92527556e-01 -1.27802670e-01 9.29558337e-01
5.96037954e-02 4.57344800e-02 -7.27077365e-01 -1.72882363e-01
1.65790811e-01 1.10802782e+00 -8.38009834e-01 -5.85595146e-02
-2.72037596e-01 1.26094866e+00 2.60235339e-01 5.41100800e-01
-1.01329243e+00 -7.62286127e-01 6.37082398e-01 1.08191758e-01
2.68698812e-01 -4.71335769e-01 4.43294179e-03 -1.19146609e+00
3.16264868e-01 -5.33453643e-01 3.86721700e-01 -6.40782952e-01
-1.50722909e+00 8.14707518e-01 -4.47509624e-03 -1.55664551e+00
-5.13398945e-01 -6.41560256e-01 -5.66293299e-01 6.86788261e-01
-1.36874402e+00 -4.79167551e-01 -2.14864053e-02 9.17624831e-01
4.65585142e-01 9.81449857e-02 7.42111146e-01 4.45607424e-01
-3.49532932e-01 5.44885278e-01 4.82965827e-01 3.59683260e-02
5.35471022e-01 -1.53404379e+00 3.41670483e-01 7.64531255e-01
5.57647347e-01 6.13681316e-01 8.54753256e-01 3.56745906e-02
-1.35681987e+00 -9.94154036e-01 8.32701743e-01 -1.34583667e-01
1.02861357e+00 -3.53619695e-01 -1.18271554e+00 5.18640339e-01
1.48109615e-01 2.47564778e-01 5.45179486e-01 -3.35987359e-02
-3.69495720e-01 -4.05041397e-01 -1.01906157e+00 6.30630672e-01
9.24212813e-01 -7.79404283e-01 -6.62681937e-01 4.27741766e-01
4.19050872e-01 -2.75354475e-01 -1.11264586e+00 4.48123403e-02
6.23403430e-01 -1.06020355e+00 9.18059587e-01 -5.21703780e-01
3.42701524e-01 -8.45628753e-02 1.23600148e-01 -1.65526569e+00
-2.83994138e-01 -1.02536583e+00 -3.00609171e-01 1.04165483e+00
1.05801947e-01 -1.00559378e+00 3.83995146e-01 4.44543719e-01
-1.74231958e-02 -1.08080947e+00 -1.29713094e+00 -1.30339217e+00
1.45544857e-01 -5.83860695e-01 5.93706787e-01 9.86400306e-01
1.53198093e-01 1.20828748e-01 -6.13860078e-02 -1.17566563e-01
3.15796882e-01 4.07648116e-01 2.65410453e-01 -1.37141252e+00
-2.51901060e-01 -9.92073536e-01 -7.95382500e-01 -1.03579760e+00
9.61784273e-02 -1.02348316e+00 -1.24820799e-01 -9.89810169e-01
-5.07312775e-01 -2.29165897e-01 -6.44681275e-01 2.79919207e-01
2.21333116e-01 1.01828195e-01 7.72451237e-02 2.41641372e-01
-1.99282870e-01 6.62369907e-01 8.77478719e-01 -1.87906936e-01
-3.41199189e-01 2.52896637e-01 -1.07329383e-01 7.95997977e-01
6.44447684e-01 -4.53030705e-01 -5.64323425e-01 -1.76054284e-01
1.18654311e-01 2.29448788e-02 3.84045452e-01 -1.24725688e+00
2.44550183e-01 2.57169664e-01 1.60740495e-01 -3.17281842e-01
3.81282300e-01 -9.63482022e-01 1.05352648e-01 5.08653581e-01
-3.94966185e-01 4.56911862e-01 -2.58671939e-02 5.06850004e-01
-4.23622668e-01 -2.63809353e-01 7.31996715e-01 1.81813717e-01
-4.21171755e-01 4.85104978e-01 -2.66279548e-01 1.19283266e-01
9.23609018e-01 -2.78903782e-01 1.21449366e-01 -4.32994246e-01
-7.17862010e-01 -2.28564404e-02 2.72344828e-01 3.77344877e-01
5.63050091e-01 -1.59185922e+00 -4.21144217e-01 2.68164992e-01
2.83349794e-03 -3.75496298e-01 -1.92240123e-02 1.49474382e+00
-2.67873913e-01 1.12278722e-01 -2.59448051e-01 -7.30200768e-01
-8.13990474e-01 7.44824469e-01 3.87047708e-01 -2.99532473e-01
-9.15892243e-01 5.10778964e-01 -3.04943733e-02 -2.30722371e-02
4.25935775e-01 -8.09824646e-01 -4.01031822e-02 4.79641467e-01
5.24435401e-01 4.23802942e-01 3.31362426e-01 -3.49428415e-01
-2.73004919e-01 7.37609982e-01 2.47192740e-01 -1.78962320e-01
1.60448647e+00 1.75622199e-02 -1.22986399e-01 1.04802692e+00
1.58465350e+00 -2.05479786e-01 -1.28215003e+00 -2.66466200e-01
4.13551658e-01 -2.82791555e-01 -1.65730879e-01 -3.16290945e-01
-1.28448224e+00 1.12388396e+00 7.57623911e-01 8.14512789e-01
1.46558046e+00 -2.33698145e-01 9.42174435e-01 3.90184343e-01
3.83680224e-01 -1.17493308e+00 8.79792050e-02 3.73471320e-01
1.05965507e+00 -8.03859949e-01 -2.64129221e-01 1.74710810e-01
-2.98942029e-01 1.58058608e+00 -1.63611501e-01 -5.22312939e-01
8.22584093e-01 2.19259068e-01 -1.32773668e-01 1.64636634e-02
-5.73166728e-01 -2.00192854e-01 4.26486492e-01 6.58399999e-01
4.75749820e-01 -1.84605166e-01 -5.51132381e-01 4.05771852e-01
-2.81881392e-01 -1.52697504e-01 3.81326675e-01 5.52973270e-01
3.36269103e-02 -1.00055349e+00 -1.53785609e-02 3.41624975e-01
-4.84981805e-01 9.02276933e-02 -2.37509198e-02 7.14578807e-01
-2.34899092e-02 6.09094381e-01 1.85318783e-01 -2.42363811e-01
4.77285534e-01 3.80668253e-01 6.50173426e-01 -1.06120192e-01
-5.81002295e-01 5.95841147e-02 -4.86648411e-01 -8.14597189e-01
-5.93357146e-01 -6.89883471e-01 -1.14678562e+00 -4.19021994e-02
-1.84586495e-02 1.86161682e-01 7.11369216e-01 8.91072094e-01
3.42293173e-01 4.79022264e-01 8.82688582e-01 -9.39461827e-01
-7.73947477e-01 -7.15794027e-01 -7.99893677e-01 7.80341089e-01
5.64926982e-01 -5.65212905e-01 -3.95292282e-01 -3.05379387e-02] | [7.401034832000732, 3.3485796451568604] |
ca6e9db6-36bb-4a0b-bd30-3f3581d3ce44 | edbb-demo-biometrics-and-behavior-analysis | 2211.09210 | null | https://arxiv.org/abs/2211.09210v2 | https://arxiv.org/pdf/2211.09210v2.pdf | edBB-Demo: Biometrics and Behavior Analysis for Online Educational Platforms | We present edBB-Demo, a demonstrator of an AI-powered research platform for student monitoring in remote education. The edBB platform aims to study the challenges associated to user recognition and behavior understanding in digital platforms. This platform has been developed for data collection, acquiring signals from a variety of sensors including keyboard, mouse, webcam, microphone, smartwatch, and an Electroencephalography band. The information captured from the sensors during the student sessions is modelled in a multimodal learning framework. The demonstrator includes: i) Biometric user authentication in an unsupervised environment; ii) Human action recognition based on remote video analysis; iii) Heart rate estimation from webcam video; and iv) Attention level estimation from facial expression analysis. | ['Javier Ortega-Garcia', 'Julian Fierrez', 'Luis F. Gomez', 'Ruben Tolosana', 'Aythami Morales', 'Roberto Daza'] | 2022-11-16 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [ 5.50167225e-02 -1.35455355e-01 -4.79945773e-03 -2.48617426e-01
-1.70924485e-01 -3.92472595e-01 1.87560678e-01 4.67644364e-01
-3.72861505e-01 6.04620397e-01 -1.58052072e-01 -2.50313431e-01
-3.57880712e-01 -3.60941559e-01 -3.59974891e-01 -6.46424294e-01
1.56281367e-01 -1.64078012e-01 -2.42156461e-01 -2.03836337e-01
4.52292532e-01 8.34057570e-01 -2.20532513e+00 1.49926409e-01
7.06338465e-01 1.06019950e+00 -1.33672848e-01 1.27087677e+00
2.90768594e-01 9.79934514e-01 -9.51015592e-01 2.94620007e-01
-3.80967170e-01 -2.75458634e-01 -5.98422825e-01 -1.61710516e-01
3.53248954e-01 -3.93164724e-01 -2.26121470e-01 9.56434846e-01
1.02444005e+00 4.25182551e-01 1.90273836e-01 -1.45720792e+00
-1.72830343e-01 3.65521282e-01 2.92176176e-02 5.09373724e-01
1.28395271e+00 -5.67442644e-03 1.02475837e-01 -4.72104400e-01
1.05847791e-02 9.75052476e-01 3.10777307e-01 9.60280120e-01
-1.14011538e+00 -9.90570605e-01 -2.65154868e-01 7.00845122e-01
-1.47702718e+00 -3.31872791e-01 6.91721559e-01 -4.55057204e-01
8.89470875e-01 5.69102287e-01 1.21334815e+00 1.65013242e+00
5.54829016e-02 6.95861161e-01 1.43984354e+00 -7.39840329e-01
5.40255606e-01 8.15990925e-01 9.65991020e-01 5.15232682e-01
1.00311881e-03 -4.82725613e-02 -1.45952606e+00 9.10145268e-02
6.10826015e-01 -5.51827140e-02 -3.55323285e-01 9.28823836e-03
-9.47051466e-01 5.51037528e-02 -4.51818526e-01 6.22691154e-01
-6.09693170e-01 -4.80378754e-02 -9.68882293e-02 6.91647291e-01
-6.70688301e-02 1.84259549e-01 -3.29443127e-01 -9.68577921e-01
-7.11286783e-01 -2.78429657e-01 1.05947244e+00 8.90185177e-01
2.07564890e-01 2.04870373e-01 1.28386497e-01 3.88914257e-01
8.35227013e-01 6.95269942e-01 1.13981628e+00 -7.39788175e-01
-8.73903185e-02 8.11485112e-01 -4.45471145e-02 -5.33845842e-01
-5.66200316e-01 3.03713232e-01 -1.74015149e-01 3.40200216e-01
2.82771498e-01 -5.04302382e-01 -3.82774770e-01 1.31349909e+00
5.19680798e-01 7.52498388e-01 1.39174402e-01 6.27502322e-01
1.39823937e+00 4.65808868e-01 2.33705100e-02 -2.93565720e-01
1.51003003e+00 -1.50781006e-01 -1.21321428e+00 3.40795428e-01
2.20015392e-01 -2.16921762e-01 8.80082309e-01 1.23960757e+00
-1.06618166e+00 -6.94163859e-01 -1.23494184e+00 3.19598377e-01
-8.41163635e-01 -1.40478104e-01 3.97095352e-01 1.70406711e+00
-1.06150794e+00 3.53792042e-01 -1.02474856e+00 -5.75832605e-01
-9.97994188e-03 7.95758605e-01 -2.76207119e-01 5.02856076e-01
-8.90732527e-01 7.77572215e-01 7.93504864e-02 -2.75632352e-01
-9.93137062e-01 -6.46569729e-01 -8.03350568e-01 3.38959724e-01
-4.06233132e-01 -1.08747415e-01 9.87100124e-01 -1.00652587e+00
-2.52113342e+00 8.24292183e-01 -8.06997158e-03 -7.60212019e-02
-3.00935991e-02 -3.17433596e-01 -7.55384207e-01 2.45038420e-01
-1.03798532e+00 7.12800473e-02 6.67756140e-01 -6.90657258e-01
-1.40041068e-01 -7.46691763e-01 -5.58629394e-01 3.03377420e-01
-9.50384378e-01 2.34569699e-01 4.54269141e-01 2.78102141e-03
-2.69061327e-01 -5.31148076e-01 5.44148743e-01 -7.19853640e-01
3.22590262e-01 -3.47977459e-01 9.45171654e-01 -6.63120866e-01
1.30880797e+00 -2.03223562e+00 -5.41694239e-02 5.91982782e-01
-1.31322533e-01 5.37482560e-01 3.81815314e-01 1.85703799e-01
-4.56900209e-01 -2.31607154e-01 6.77124619e-01 5.11042625e-02
2.46229351e-01 -2.21743196e-01 1.48592204e-01 4.22748029e-01
-6.63582563e-01 3.19841295e-01 -6.14369631e-01 -4.00718987e-01
9.01302218e-01 7.09220707e-01 -1.45741031e-01 6.44299746e-01
7.62724936e-01 7.04580307e-01 -3.28331649e-01 9.24481869e-01
3.10016066e-01 4.55687612e-01 -1.90027103e-01 4.68435854e-01
-3.17502916e-01 2.62472868e-01 -1.49184465e+00 1.78116047e+00
-4.70267415e-01 9.70848501e-01 3.60930324e-01 -1.18468797e+00
1.12621796e+00 1.19231653e+00 6.31120384e-01 -5.86909294e-01
6.01395309e-01 -2.02808052e-01 -3.22953701e-01 -1.32819009e+00
-1.46950483e-01 3.52357000e-01 2.96315670e-01 4.61882502e-01
7.78024316e-01 -1.00881783e-02 -3.35339129e-01 -2.85888135e-01
1.10576093e+00 1.00464478e-01 3.81373584e-01 -2.84251094e-01
9.29495096e-01 -7.21191287e-01 4.97333370e-02 3.50779146e-01
-6.60969973e-01 -1.78753436e-01 -2.47592852e-03 -2.58230984e-01
1.55205086e-01 -7.60712922e-01 -2.03519106e-01 1.24494529e+00
-1.01901107e-01 -1.65007427e-01 -1.28668916e+00 -1.64652973e-01
-3.18791270e-01 8.32804918e-01 -2.35684812e-01 -4.12784576e-01
1.51683137e-01 -4.77204859e-01 4.66470659e-01 2.76101559e-01
3.36886704e-01 -1.30260861e+00 -1.12644267e+00 1.11219920e-01
1.73198029e-01 -1.05206859e+00 1.94663763e-01 4.20180291e-01
-9.43556488e-01 -8.17153335e-01 -2.10742518e-01 -7.73305714e-01
4.09934729e-01 -2.70457655e-01 6.71403050e-01 -4.40477952e-02
-6.16024256e-01 1.71358156e+00 -1.40549377e-01 -1.20938480e+00
-1.28953643e-02 -2.48941109e-01 4.21705872e-01 1.87424257e-01
1.14453745e+00 -9.88509715e-01 -4.87694621e-01 -3.72213759e-02
-5.69354892e-01 -4.61446404e-01 1.64045304e-01 9.04229879e-02
-9.42016095e-02 -2.88332969e-01 4.70906734e-01 -1.39991105e-01
7.88651347e-01 -3.68277252e-01 -5.44856310e-01 3.66301984e-01
-5.81346631e-01 -6.20691061e-01 6.44064248e-02 -8.46055925e-01
-1.00539756e+00 1.87592402e-01 -2.50438392e-01 2.41620187e-02
-1.33531189e+00 1.18559040e-01 -5.64167976e-01 -3.44862282e-01
7.05128610e-01 3.87663960e-01 -1.07294209e-01 -2.78492004e-01
-5.32625020e-01 1.62524760e+00 7.71415830e-01 -6.62538767e-01
7.43366927e-02 -2.91680038e-01 -3.82027417e-01 -1.46072543e+00
-1.70066372e-01 -5.05531967e-01 -8.80895853e-01 -1.11998868e+00
1.03751743e+00 -1.11468101e+00 -1.82933271e+00 7.76870370e-01
-7.64688671e-01 -3.74765903e-01 -1.35511354e-01 1.21141398e+00
-4.08593148e-01 -2.45079860e-01 -6.07875943e-01 -1.49254489e+00
-5.51039696e-01 -7.35804379e-01 6.05349660e-01 1.04549265e+00
-6.00159049e-01 -1.02689326e+00 2.58907259e-01 1.02063417e+00
2.96914965e-01 3.00103515e-01 3.01697344e-01 -9.84109640e-01
-4.80954312e-02 -6.12793803e-01 9.00766730e-01 3.85193706e-01
-1.09135628e-01 -1.59516837e-02 -1.58707094e+00 -8.01345333e-02
3.20434719e-01 -5.40035605e-01 -2.84277052e-01 1.54690892e-01
1.27477050e+00 1.40774801e-01 6.72195666e-03 2.81839192e-01
9.75592136e-01 8.46614122e-01 6.38140321e-01 8.89588892e-02
3.73867124e-01 4.85848874e-01 -2.42370907e-02 3.45861733e-01
9.00980607e-02 3.14597577e-01 3.15744907e-01 4.43745285e-01
3.86151642e-01 1.13995090e-01 1.01095772e+00 1.04324627e+00
-2.40329534e-01 -8.44739079e-02 -1.09937024e+00 2.09551558e-01
-1.56637621e+00 -9.84704912e-01 -1.80655912e-01 2.37301111e+00
6.66538239e-01 -2.56913632e-01 3.77961546e-01 7.69629180e-01
5.86755276e-01 -6.81920826e-01 -2.57809877e-01 -9.22510803e-01
6.10769451e-01 9.80625987e-01 -8.42233226e-02 3.13374728e-01
-6.57498121e-01 2.65481114e-01 6.42379284e+00 1.83300003e-01
-1.19047177e+00 1.84428811e-01 9.97512788e-02 -2.21423075e-01
1.81623787e-01 -6.87263906e-01 -7.27959812e-01 5.74665904e-01
1.81542850e+00 2.51614265e-02 6.88017964e-01 7.57365167e-01
2.05752522e-01 -2.74744958e-01 -1.19265866e+00 1.35721397e+00
5.15791297e-01 -6.03809237e-01 -4.47845310e-01 -6.14502886e-03
2.75804400e-01 -4.57856834e-01 -9.27789509e-02 4.03503925e-01
2.75269207e-02 -1.12063038e+00 1.94846585e-01 9.45969224e-01
3.77952427e-01 -8.94691050e-01 3.89835715e-01 6.91510797e-01
-5.76485097e-01 -1.70310318e-01 2.91886032e-01 -3.12778175e-01
-8.04869473e-01 2.90766079e-02 -4.05836403e-01 -1.27011165e-01
8.31740022e-01 4.76675749e-01 -4.01029319e-01 9.32222843e-01
-3.09515089e-01 1.04606318e+00 -2.78688222e-01 -5.26413381e-01
-4.69908416e-01 -1.94155052e-01 3.31816792e-01 1.12920010e+00
6.04336619e-01 5.57812870e-01 -4.11395133e-01 5.81358135e-01
2.49236181e-01 8.56155157e-02 -6.48661196e-01 1.98356673e-01
4.48294967e-01 1.65038574e+00 -6.77831352e-01 -1.93395376e-01
-4.30448085e-01 6.50246382e-01 -4.87170130e-01 4.14706498e-01
-6.63680732e-01 -7.27217317e-01 6.82252407e-01 -9.44743603e-02
-7.57540986e-02 -8.41124356e-02 -3.40904444e-01 -1.04491222e+00
-5.94143569e-01 -9.47030246e-01 2.97454774e-01 -1.13002968e+00
-4.62669432e-01 -1.24062680e-01 -1.03849964e-02 -7.63276994e-01
-6.23890013e-02 -8.32240760e-01 -1.04060817e+00 6.23062611e-01
-7.21229851e-01 -3.75428110e-01 -8.13863695e-01 1.45866191e+00
2.46889323e-01 -4.88983512e-01 1.41488147e+00 5.02003312e-01
-1.14035094e+00 6.80459678e-01 -3.29376310e-01 -6.55505946e-03
5.54446161e-01 -1.54723120e+00 -8.15444350e-01 5.65605938e-01
1.07332304e-01 5.50198972e-01 4.31611747e-01 -9.76056531e-02
-1.96310055e+00 -1.89729780e-01 6.81073368e-01 -6.58897400e-01
1.04944386e-01 -4.78981435e-01 -6.30316317e-01 7.81892002e-01
6.59803987e-01 -1.85546383e-01 1.49645519e+00 8.10784847e-02
4.54904705e-01 -5.07591367e-01 -1.44744658e+00 2.25025639e-01
2.51142442e-01 -5.98442614e-01 -8.40544879e-01 -3.08610257e-02
-1.67652801e-01 -4.53492820e-01 -1.40996814e+00 -2.70173728e-01
8.46115708e-01 -6.93165302e-01 8.20289075e-01 -4.23963636e-01
-1.81915045e-01 1.15134045e-01 2.70472795e-01 -1.13997591e+00
3.89675349e-01 -1.28647447e+00 -5.90882540e-01 1.51211965e+00
-9.32272077e-02 -5.44780314e-01 6.88183963e-01 1.30720055e+00
3.67336005e-01 2.69322190e-02 -1.04118240e+00 -5.08684218e-02
-5.19421220e-01 -6.56511486e-01 4.21833873e-01 9.59212720e-01
1.27035952e+00 4.19289380e-01 3.76832858e-02 2.68404007e-01
2.64784664e-01 -5.73399961e-01 6.70964897e-01 -1.69415033e+00
1.79823905e-01 -2.53022432e-01 -1.05544317e+00 -4.25373018e-01
-4.75883018e-03 -4.21573132e-01 -3.27288747e-01 -9.52564359e-01
-2.15301976e-01 6.89141095e-01 -6.82249844e-01 3.67712677e-01
3.08356971e-01 -8.68717209e-02 -3.02029520e-01 -7.58822024e-01
-4.33952838e-01 8.39275569e-02 6.06727004e-01 3.04022133e-01
-5.11696041e-01 1.51832625e-01 -2.22311258e-01 8.12900722e-01
7.56339908e-01 -3.44366610e-01 -4.04973924e-01 2.91266680e-01
1.24983288e-01 4.59881455e-01 3.70091558e-01 -1.67448950e+00
7.23814487e-01 6.85622618e-02 7.82355547e-01 -1.20428599e-01
2.62097836e-01 -1.50587988e+00 -1.84584484e-01 6.17581964e-01
-4.56784546e-01 -4.48020808e-02 4.00409490e-01 -8.29589888e-02
-2.71996558e-02 -2.31829718e-01 5.00877082e-01 1.27119631e-01
-3.98416221e-01 -4.11293864e-01 -1.38884771e+00 -5.92651069e-01
1.41272175e+00 -5.18042266e-01 -1.89463064e-01 -5.31037986e-01
-1.19590497e+00 1.16882659e-01 -3.70766342e-01 2.91141152e-01
6.51358247e-01 -1.13556707e+00 2.57545356e-02 1.00481486e+00
-3.26567352e-01 -4.33241814e-01 1.11438453e-01 1.04824471e+00
-1.46120504e-01 2.97499359e-01 -7.78770864e-01 -5.23375154e-01
-2.14772153e+00 1.99332952e-01 5.79903245e-01 1.64978057e-01
-6.08791336e-02 6.01486444e-01 -9.90825117e-01 -9.10319388e-02
1.03772926e+00 -2.95889169e-01 -9.53601420e-01 4.80309159e-01
1.02408767e+00 8.59159231e-01 3.99412304e-01 -1.81343302e-01
-3.19528937e-01 2.63120085e-01 5.24976790e-01 -3.08908552e-01
1.44403386e+00 -1.24355294e-01 2.98090070e-01 9.22623932e-01
7.38121212e-01 -1.54340893e-01 -4.74073321e-01 3.10323000e-01
5.49954847e-02 7.62109086e-02 4.65043664e-01 -9.98892725e-01
-7.64565349e-01 1.14038789e+00 1.43379402e+00 3.24204981e-01
1.38453114e+00 -9.50569332e-01 3.51171017e-01 6.85454667e-01
8.52044821e-02 -1.65409291e+00 2.44619116e-01 2.39547208e-01
3.69870216e-01 -9.54271913e-01 -1.76020578e-01 2.70416260e-01
-4.44731236e-01 1.48352897e+00 8.45325470e-01 1.66594461e-01
1.05933082e+00 6.72289133e-01 1.56375572e-01 -3.85192484e-01
-7.69214392e-01 2.26740111e-02 4.50522602e-01 6.90770328e-01
7.45434344e-01 -1.53794855e-01 3.58773172e-02 1.17067182e+00
-2.81717956e-01 5.15609205e-01 7.14705229e-01 1.12652719e+00
-5.69263756e-01 -5.70757627e-01 -1.05710208e+00 -1.05453700e-01
-6.57142580e-01 3.17282200e-01 -7.71228671e-01 4.73824888e-01
2.06288144e-01 1.28869295e+00 -2.05610320e-02 -5.34425557e-01
6.54221892e-01 1.07341886e+00 5.66510558e-01 -3.84059876e-01
-1.20711839e+00 -1.54501736e-01 -3.02035868e-01 -7.34801650e-01
-6.09537840e-01 -7.96293080e-01 -1.12937629e+00 -1.74153775e-01
5.49946986e-02 3.20736825e-01 1.33390331e+00 8.40338945e-01
1.23365946e-01 7.85985231e-01 4.50877488e-01 -8.52361083e-01
7.61186853e-02 -1.26539135e+00 -8.18861783e-01 -4.49954718e-02
3.44215810e-01 -2.69592375e-01 -3.56797695e-01 1.84817806e-01] | [13.49421215057373, 2.694260835647583] |
dd85b7cc-c9b3-4df6-96f4-a026a5821817 | boxvis-video-instance-segmentation-with-box | 2303.14618 | null | https://arxiv.org/abs/2303.14618v1 | https://arxiv.org/pdf/2303.14618v1.pdf | BoxVIS: Video Instance Segmentation with Box Annotations | It is expensive and labour-extensive to label the pixel-wise object masks in a video. As a results, the amount of pixel-wise annotations in existing video instance segmentation (VIS) datasets is small, limiting the generalization capability of trained VIS models. An alternative but much cheaper solution is to use bounding boxes to label instances in videos. Inspired by the recent success of box-supervised image instance segmentation, we first adapt the state-of-the-art pixel-supervised VIS models to a box-supervised VIS (BoxVIS) baseline, and observe only slight performance degradation. We consequently propose to improve BoxVIS performance from two aspects. First, we propose a box-center guided spatial-temporal pairwise affinity (STPA) loss to predict instance masks for better spatial and temporal consistency. Second, we collect a larger scale box-annotated VIS dataset (BVISD) by consolidating the videos from current VIS benchmarks and converting images from the COCO dataset to short pseudo video clips. With the proposed BVISD and the STPA loss, our trained BoxVIS model demonstrates promising instance mask prediction performance. Specifically, it achieves 43.2\% and 29.0\% mask AP on the YouTube-VIS 2021 and OVIS valid sets, respectively, exhibiting comparable or even better generalization performance than state-of-the-art pixel-supervised VIS models by using only 16\% annotation time and cost. Codes and data of BoxVIS can be found at \url{https://github.com/MinghanLi/BoxVIS}. | ['Lei Zhang', 'Minghan Li'] | 2023-03-26 | null | null | null | null | ['video-instance-segmentation'] | ['computer-vision'] | [ 3.38545293e-01 1.29118413e-01 -5.81713915e-01 -5.22469640e-01
-1.00618124e+00 -5.08369207e-01 3.82047832e-01 -8.72599781e-02
-5.48717618e-01 6.15746796e-01 -3.65480602e-01 -1.13628760e-01
2.35923082e-01 -5.12555718e-01 -1.08545530e+00 -5.56594074e-01
4.57265321e-03 2.55634785e-01 7.94817209e-01 1.84405148e-01
2.56325062e-02 1.77234441e-01 -1.44002974e+00 4.81423080e-01
9.69282746e-01 1.39517403e+00 2.81777799e-01 4.46583331e-01
-9.94968638e-02 6.75872743e-01 -3.77560407e-01 -5.13999343e-01
5.95593095e-01 -4.78972614e-01 -8.65373015e-01 3.06034178e-01
8.98344040e-01 -3.90582472e-01 -5.44530272e-01 9.92526054e-01
1.52550370e-01 2.31656596e-01 5.24469376e-01 -1.28712916e+00
-4.20426607e-01 6.05429173e-01 -9.20561433e-01 3.04030240e-01
-3.29627022e-02 2.56262720e-01 1.01227307e+00 -9.77848411e-01
8.57140243e-01 7.45601475e-01 6.27955258e-01 5.83413959e-01
-1.22831583e+00 -7.19746470e-01 5.52561462e-01 2.51063824e-01
-1.48379827e+00 -1.81658775e-01 6.39878750e-01 -5.47595680e-01
6.70712113e-01 4.80034649e-01 7.03588367e-01 1.03706563e+00
-3.23797613e-01 1.21571898e+00 1.18744111e+00 -3.57206054e-02
1.38015002e-01 1.63730174e-01 2.01253265e-01 8.03834915e-01
5.76630011e-02 -9.96614918e-02 -4.71417427e-01 3.59004974e-01
7.12240338e-01 8.48331824e-02 -3.71229619e-01 -3.74595344e-01
-1.21112418e+00 3.48811805e-01 6.50157154e-01 1.25499606e-01
-2.11873606e-01 3.05456340e-01 5.19521475e-01 -9.81223732e-02
6.93342090e-01 1.96318313e-01 -5.49566150e-01 -2.39364356e-01
-1.42360330e+00 2.03629732e-01 3.74586910e-01 1.13724744e+00
8.04607272e-01 -4.46275882e-02 -5.08977771e-01 8.75360370e-01
1.11099775e-03 3.48772287e-01 1.82035789e-01 -1.04393768e+00
6.44828439e-01 5.72708726e-01 2.86298189e-02 -6.20450199e-01
-1.53742120e-01 -5.54261565e-01 -4.82738435e-01 4.74225804e-02
5.99650562e-01 1.33179367e-01 -1.33561349e+00 1.34732223e+00
4.06751782e-01 4.98522580e-01 -4.01969761e-01 1.15812171e+00
1.08166826e+00 7.55535483e-01 1.90727949e-01 -1.55750722e-01
1.17573345e+00 -1.41171610e+00 -5.41142642e-01 -3.22316468e-01
7.22708344e-01 -3.96935016e-01 1.29700601e+00 4.97429311e-01
-1.24012554e+00 -6.70924783e-01 -9.88460898e-01 3.22039016e-02
-2.55631357e-01 3.17801297e-01 4.27486330e-01 5.05459130e-01
-7.21861184e-01 5.19637764e-01 -1.00404429e+00 -5.20512760e-02
1.05215240e+00 3.46228421e-01 -3.19303632e-01 -2.81815112e-01
-7.75078237e-01 2.34409571e-01 3.78197044e-01 3.66976485e-02
-1.16239142e+00 -1.03784502e+00 -8.36080909e-01 -1.04210593e-01
8.93832684e-01 -9.92875248e-02 1.04787397e+00 -1.20649123e+00
-1.04741454e+00 9.52714145e-01 -2.29768708e-01 -6.73187613e-01
8.74004185e-01 -3.21052462e-01 -1.60753459e-01 4.16260630e-01
1.68081671e-01 1.10911584e+00 7.57867575e-01 -1.52728915e+00
-6.97552919e-01 -7.22182840e-02 4.87613492e-02 2.90493704e-02
-3.36226135e-01 -1.09264456e-01 -1.25600052e+00 -8.62795174e-01
-3.78559716e-02 -9.72776413e-01 -6.29451424e-02 2.06304029e-01
-5.16582966e-01 -1.75394088e-01 8.35506797e-01 -8.54427993e-01
1.39807534e+00 -2.22245216e+00 5.73424697e-02 8.53291750e-02
1.58114776e-01 5.68738818e-01 -2.27979809e-01 4.47826684e-02
1.26785906e-02 2.01064780e-01 -6.73156857e-01 -6.09581828e-01
-1.69222996e-01 2.45992035e-01 -1.71319634e-01 3.88652831e-01
2.97607601e-01 1.10729241e+00 -7.92694569e-01 -7.00430274e-01
3.44926298e-01 3.30175161e-01 -8.01935375e-01 5.32418191e-02
-4.27168936e-01 5.39208531e-01 -2.64820158e-01 9.89367485e-01
7.18855083e-01 -2.32540131e-01 -3.23856473e-02 -3.76710594e-01
9.49045569e-02 5.84927425e-02 -9.53805327e-01 1.82059932e+00
-8.01635161e-02 7.41687775e-01 -1.46058291e-01 -1.23647058e+00
5.96404731e-01 8.23184401e-02 8.21497917e-01 -8.06903481e-01
8.21374729e-02 3.01830441e-01 -1.91071004e-01 -4.89507765e-01
3.64162356e-01 3.17073435e-01 1.60177857e-01 -8.16080254e-03
1.83623940e-01 -2.31432077e-02 6.42712772e-01 1.92804128e-01
9.26830769e-01 6.75107718e-01 -1.72018230e-01 -1.82132274e-01
4.74524647e-01 2.70034492e-01 7.94235706e-01 6.35737538e-01
-4.15099710e-01 1.03653574e+00 4.40731585e-01 -2.91919082e-01
-8.54954123e-01 -1.00047278e+00 -3.90061140e-01 9.06032205e-01
4.99891043e-01 -4.44628894e-01 -1.08688426e+00 -1.20274842e+00
-5.38006052e-02 3.94285768e-01 -6.25543535e-01 1.75017923e-01
-7.46374249e-01 -4.50845957e-01 4.23490524e-01 8.30412507e-01
7.42103934e-01 -9.17653203e-01 -2.55696476e-01 1.95881613e-02
-1.06797516e-01 -1.42168236e+00 -6.87759042e-01 3.25136781e-02
-7.13997364e-01 -9.75424170e-01 -1.01902497e+00 -7.60383606e-01
7.89713144e-01 1.79849178e-01 1.07665634e+00 7.71413073e-02
-3.65255207e-01 2.26312011e-01 -5.46178639e-01 -2.10239023e-01
1.39676839e-01 -3.19521055e-02 -1.65635422e-01 7.18135536e-02
2.34663934e-01 -1.80431604e-01 -1.00699866e+00 7.22354889e-01
-9.71842289e-01 2.75655031e-01 3.08595985e-01 7.81838953e-01
1.13239884e+00 -1.73115209e-01 2.71539867e-01 -1.09429741e+00
-2.73677915e-01 -2.74941027e-01 -6.13479972e-01 9.02456194e-02
-5.66983223e-01 -5.71958125e-01 3.77968848e-01 -4.69907939e-01
-8.85536194e-01 1.55004516e-01 -1.35271132e-01 -6.65754616e-01
-1.66953936e-01 3.35253865e-01 -4.47005220e-02 -5.96459918e-02
4.92138088e-01 1.98631555e-01 -2.21097425e-01 -3.94529521e-01
3.14385742e-01 6.07379735e-01 5.16577125e-01 -4.83000100e-01
6.34996414e-01 6.64336085e-01 -3.40277314e-01 -6.32525086e-01
-1.07541966e+00 -6.76227570e-01 -5.40953457e-01 -4.08606142e-01
1.11354268e+00 -1.08318174e+00 -3.24992895e-01 3.57022047e-01
-7.49040127e-01 -9.60100770e-01 -4.17219967e-01 3.77526939e-01
-6.12887025e-01 2.98918337e-01 -5.50102532e-01 -6.01371586e-01
-1.16044879e-01 -1.25058842e+00 1.25759578e+00 8.36445987e-02
-2.12625593e-01 -5.89702129e-01 -3.78885210e-01 8.32528472e-01
-3.21643390e-02 2.06841558e-01 4.35261548e-01 -4.44725573e-01
-8.19766879e-01 -2.98470318e-01 -4.74852949e-01 7.21088588e-01
-1.56312779e-01 -4.47399318e-02 -9.40953374e-01 -2.50755489e-01
-4.58145142e-01 -2.58822173e-01 1.17545068e+00 5.51053524e-01
1.81361008e+00 -1.36036888e-01 -3.78069580e-01 9.84522045e-01
1.40862155e+00 2.91849613e-01 5.55739999e-01 3.11037153e-01
1.07128155e+00 5.30244708e-01 1.11521411e+00 1.88482314e-01
1.95046589e-01 8.34247410e-01 4.24618393e-01 -3.54805171e-01
-4.42284882e-01 -3.46436262e-01 2.27495551e-01 4.50971425e-01
-3.14355433e-01 -3.28478664e-01 -8.47655475e-01 7.40292072e-01
-1.97137392e+00 -7.27677643e-01 -1.95808917e-01 1.98612368e+00
7.49076486e-01 2.46069193e-01 4.58553225e-01 1.35600209e-01
5.32502770e-01 2.97600985e-01 -5.38595200e-01 4.19799425e-02
-7.45994449e-02 1.74399734e-01 8.34187329e-01 1.88259467e-01
-1.40111637e+00 1.06645560e+00 4.94752598e+00 1.11905777e+00
-1.10798717e+00 3.76800716e-01 1.17713225e+00 -5.86035252e-01
-4.53370772e-02 -1.86800078e-01 -7.77735293e-01 6.86293960e-01
5.14196217e-01 4.65855777e-01 3.28352630e-01 8.95070910e-01
3.04245293e-01 -2.06042677e-01 -1.05108464e+00 1.09826410e+00
3.75732891e-02 -1.56110609e+00 -5.37496284e-02 -9.92222875e-02
1.05640030e+00 1.52882189e-01 1.17663503e-01 2.96284407e-01
-3.06687683e-01 -8.79181087e-01 1.06360531e+00 2.18892381e-01
1.07227516e+00 -5.16677916e-01 6.81140006e-01 9.55074728e-02
-1.25968671e+00 -9.50307995e-02 -1.37309238e-01 2.38804981e-01
2.55226135e-01 2.97973782e-01 -2.88219810e-01 4.28874344e-01
1.13148236e+00 9.83250916e-01 -5.49380064e-01 1.27312529e+00
-1.49575457e-01 1.01387405e+00 -2.47062832e-01 2.78981417e-01
5.48955679e-01 -4.23503876e-01 3.62216473e-01 1.14316499e+00
1.79623932e-01 2.55900137e-02 3.23545903e-01 9.01314735e-01
-2.98168421e-01 7.32460171e-02 -1.01240560e-01 -1.21810533e-01
2.13539645e-01 1.22573340e+00 -9.69430923e-01 -5.57035923e-01
-5.62075198e-01 1.12773442e+00 2.02357247e-01 6.56870008e-01
-1.23356545e+00 -6.52712584e-02 5.49477339e-01 4.56406593e-01
6.43557668e-01 -1.16296984e-01 -5.44823468e-01 -9.45496559e-01
3.20364088e-01 -7.88751245e-01 3.49888384e-01 -6.47864223e-01
-1.03562868e+00 6.22092366e-01 1.38281420e-01 -1.44375169e+00
3.16880733e-01 -6.73327684e-01 -4.07465369e-01 5.10280192e-01
-1.56510997e+00 -1.21375847e+00 -5.49923122e-01 3.78142446e-01
9.29716766e-01 -1.24204103e-02 1.95167899e-01 5.60560524e-01
-8.58054459e-01 7.96539783e-01 -4.56168465e-02 3.44681978e-01
6.23849750e-01 -1.12246072e+00 3.04740131e-01 9.26509857e-01
4.70779330e-01 2.89264649e-01 4.55514908e-01 -6.10421836e-01
-1.20818019e+00 -1.49476981e+00 2.93571353e-01 -5.08147061e-01
5.37079155e-01 -5.65532684e-01 -9.73991394e-01 6.69500232e-01
1.10242173e-01 6.06345296e-01 2.85830081e-01 -4.00291353e-01
-2.61344045e-01 -1.86333522e-01 -1.04501605e+00 7.19353795e-01
1.52786005e+00 -3.00674140e-01 -4.37931456e-02 4.69820023e-01
8.15793216e-01 -5.96650124e-01 -8.43709767e-01 7.42755592e-01
3.71666431e-01 -8.82938385e-01 1.01216197e+00 -4.19659317e-01
5.88639200e-01 -4.47043747e-01 -6.86581712e-03 -7.00753331e-01
-1.63686834e-02 -5.78358531e-01 -2.33658463e-01 1.34496737e+00
5.90996504e-01 -4.53119993e-01 1.16692686e+00 6.08990073e-01
-3.64959419e-01 -1.31635284e+00 -9.06641603e-01 -8.49708259e-01
-7.69282728e-02 -7.38763094e-01 3.40611875e-01 8.40613961e-01
-3.09275806e-01 -2.71963924e-01 -3.00174803e-01 4.67570834e-02
5.08958697e-01 9.78824310e-03 8.14732909e-01 -6.82776868e-01
-3.96933168e-01 -4.21430916e-01 -3.71250808e-01 -1.39061785e+00
9.74384099e-02 -9.40050304e-01 8.12628046e-02 -1.53901863e+00
2.37999246e-01 -7.01897323e-01 -4.28481340e-01 5.61639607e-01
-3.48615080e-01 1.02447259e+00 2.94955790e-01 3.38767231e-01
-7.81885743e-01 4.69779432e-01 1.32456088e+00 -3.17634881e-01
-2.64278293e-01 -1.92971796e-01 -2.61306971e-01 6.64694607e-01
6.01740956e-01 -3.31636071e-01 -4.64376539e-01 -3.93290371e-01
-1.57061577e-01 -1.94626555e-01 4.53324407e-01 -1.02396834e+00
-1.27902284e-01 -1.33464485e-01 3.21606874e-01 -7.77970314e-01
4.69406694e-01 -7.56944776e-01 -2.56180856e-02 2.92096376e-01
-2.22778589e-01 -4.00232613e-01 3.38044941e-01 6.32888377e-01
-3.84688675e-01 -1.29248172e-01 7.05848992e-01 1.92225520e-02
-1.12122250e+00 7.92574763e-01 8.95466954e-02 2.43250683e-01
1.34141314e+00 -5.89820564e-01 -3.15791667e-01 -1.54580576e-02
-5.20591676e-01 4.06339556e-01 5.25565624e-01 4.34292078e-01
5.30060768e-01 -1.03949046e+00 -4.22725827e-01 1.05564333e-01
3.84033740e-01 3.35340559e-01 4.97224718e-01 1.27343464e+00
-7.64282227e-01 3.07087958e-01 9.02197659e-02 -9.18969929e-01
-1.32856631e+00 4.62160498e-01 2.57304639e-01 1.67972162e-01
-8.05509090e-01 1.21888447e+00 5.79386592e-01 -2.26723388e-01
3.30770552e-01 -4.37638760e-01 4.88006473e-02 -1.10544361e-01
3.25156510e-01 2.97761112e-01 -5.61726093e-02 -6.24064326e-01
-2.89748847e-01 6.29289448e-01 -4.67221737e-02 8.82071480e-02
1.23599815e+00 7.96292052e-02 3.25148523e-01 2.61407912e-01
1.17863846e+00 -1.91797942e-01 -1.82569385e+00 -1.89273596e-01
-1.56900942e-01 -7.45133579e-01 9.80277546e-03 -7.14956641e-01
-1.38039649e+00 7.89570987e-01 5.44728100e-01 3.02823335e-02
1.09699559e+00 2.06626505e-01 1.02746749e+00 -1.70303032e-01
2.56206244e-01 -1.23348165e+00 2.15031691e-02 8.97368789e-02
6.95737720e-01 -1.46184087e+00 -2.83373008e-03 -9.43444550e-01
-9.33674634e-01 5.09591103e-01 8.38589847e-01 -1.79748610e-01
4.61423784e-01 1.29987925e-01 7.36010298e-02 -1.02004841e-01
-3.62947166e-01 -2.89627820e-01 6.21752799e-01 4.46373552e-01
3.38090986e-01 -6.20025806e-02 -4.49421108e-01 5.81357121e-01
1.96649209e-01 -2.40850933e-02 9.03414413e-02 6.23747468e-01
-6.01114631e-02 -8.94253552e-01 -7.86277875e-02 6.74398839e-01
-4.88133520e-01 -1.24824449e-01 -1.73031479e-01 8.28152299e-01
2.97546685e-01 6.78160548e-01 1.71061948e-01 -2.53160238e-01
2.62796879e-01 -1.03360370e-01 3.77613783e-01 -6.07649207e-01
-4.19224828e-01 1.59687102e-01 1.44781858e-01 -9.22588348e-01
-5.38915217e-01 -6.46058679e-01 -1.40959430e+00 1.09966911e-01
-2.93826461e-01 -1.47327334e-01 5.57083309e-01 8.77179325e-01
2.95508802e-01 6.73010945e-01 2.36121744e-01 -1.00147367e+00
-1.44688219e-01 -8.55708301e-01 -5.02094567e-01 7.01456964e-01
-5.20988088e-03 -7.09723771e-01 -1.88119397e-01 2.42345735e-01] | [9.273626327514648, 0.06983296573162079] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.