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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
534afc1e-7dfa-4d01-96bf-caa131a84217 | adaptive-modeling-of-satellite-derived | 2306.08501 | null | https://arxiv.org/abs/2306.08501v1 | https://arxiv.org/pdf/2306.08501v1.pdf | Adaptive Modeling of Satellite-Derived Nighttime Lights Time-Series for Tracking Urban Change Processes Using Machine Learning | Remotely sensed nighttime lights (NTL) uniquely capture urban change processes that are important to human and ecological well-being, such as urbanization, socio-political conflicts and displacement, impacts from disasters, holidays, and changes in daily human patterns of movement. Though several NTL products are global in extent, intrinsic city-specific factors that affect lighting, such as development levels, and social, economic, and cultural characteristics, are unique to each city, making the urban processes embedded in NTL signatures difficult to characterize, and limiting the scalability of urban change analyses. In this study, we propose a data-driven approach to detect urban changes from daily satellite-derived NTL data records that is adaptive across cities and effective at learning city-specific temporal patterns. The proposed method learns to forecast NTL signatures from past data records using neural networks and allows the use of large volumes of unlabeled data, eliminating annotation effort. Urban changes are detected based on deviations of observed NTL from model forecasts using an anomaly detection approach. Comparing model forecasts with observed NTL also allows identifying the direction of change (positive or negative) and monitoring change severity for tracking recovery. In operationalizing the model, we consider ten urban areas from diverse geographic regions with dynamic NTL time-series and demonstrate the generalizability of the approach for detecting the change processes with different drivers and rates occurring within these urban areas based on NTL deviation. This scalable approach for monitoring changes from daily remote sensing observations efficiently utilizes large data volumes to support continuous monitoring and decision making. | ['Eleanor C. Stokes', 'Srija Chakraborty'] | 2023-06-14 | null | null | null | null | ['anomaly-detection'] | ['methodology'] | [ 2.38749478e-02 -5.64245820e-01 -2.73597926e-01 -4.17134970e-01
-2.94224888e-01 -5.86445451e-01 9.06134844e-01 4.45655733e-01
-1.98074669e-01 7.56106496e-01 5.92621148e-01 -6.64616466e-01
-1.59539133e-01 -1.23392320e+00 -3.19796711e-01 -7.18520105e-01
-1.77042469e-01 7.71863312e-02 -5.72040975e-02 -3.97689879e-01
-1.38219237e-01 8.67224276e-01 -1.71807981e+00 -1.30932495e-01
8.39416921e-01 5.31794548e-01 1.24185897e-01 5.33054769e-01
-3.36087197e-01 4.55556899e-01 -2.32550219e-01 6.53918386e-01
3.77522498e-01 -1.80733457e-01 -1.63026854e-01 3.12265426e-01
3.23781192e-01 -2.02145606e-01 -2.78126210e-01 6.59647644e-01
3.99427205e-01 1.66918248e-01 6.10168099e-01 -1.43003607e+00
-6.07652962e-01 2.94884324e-01 -4.28226918e-01 6.48316145e-01
1.49794176e-01 5.68966150e-01 8.55399668e-01 -7.86456704e-01
2.83304542e-01 1.08195972e+00 7.85843909e-01 5.11212554e-03
-1.33145499e+00 -7.87440121e-01 5.19952476e-01 1.73947245e-01
-1.59682930e+00 -5.80503404e-01 6.20189548e-01 -8.83855641e-01
9.69577730e-01 5.11687636e-01 9.11436379e-01 6.06088638e-01
-3.73856127e-02 5.91502130e-01 1.08598006e+00 -1.25559658e-01
3.15546930e-01 -2.66257703e-01 -5.04352786e-02 3.19544077e-01
1.76735833e-01 1.78228617e-01 -5.47219953e-03 -1.08089164e-01
6.20812885e-02 5.57396948e-01 -1.07023895e-01 2.70782471e-01
-1.13774669e+00 6.03664517e-01 6.20321989e-01 3.82316351e-01
-6.14244938e-01 1.54404163e-01 1.93686947e-01 2.74744660e-01
9.55658555e-01 -1.15273118e-01 -6.46474957e-01 -7.07314983e-02
-9.62009072e-01 3.37288752e-02 5.20623699e-02 5.23826182e-01
1.16743624e+00 3.09402585e-01 -3.52211259e-02 7.02621698e-01
4.22023565e-01 1.35713613e+00 3.60064864e-01 -5.88288665e-01
3.17555457e-01 9.22876954e-01 1.88857540e-01 -1.22043157e+00
-8.67380619e-01 -2.55946815e-01 -8.88536513e-01 -1.93814766e-02
2.13043317e-01 -2.84084529e-01 -9.55725133e-01 1.77092314e+00
5.75525880e-01 2.87120461e-01 -1.72084123e-01 4.05412257e-01
3.92675728e-01 8.96503985e-01 3.64758223e-01 -2.80161440e-01
8.09674084e-01 -2.21407413e-01 -6.35478735e-01 -3.45676333e-01
8.34773064e-01 -2.98385531e-01 9.48182821e-01 -4.73026425e-01
-2.72781372e-01 -3.48700792e-01 -3.76365364e-01 5.06428421e-01
-9.93104935e-01 -7.01452345e-02 4.65213925e-01 3.60622615e-01
-1.08313465e+00 1.83870181e-01 -9.17152524e-01 -9.40906584e-01
3.55071276e-01 6.82597309e-02 5.58439344e-02 2.14161858e-01
-1.10885787e+00 6.41425967e-01 2.88585812e-01 2.73832053e-01
-4.58432972e-01 -7.88261116e-01 -8.24554563e-01 -7.63907954e-02
-1.53627098e-01 -1.47902921e-01 8.83592904e-01 -9.22833145e-01
-8.51766527e-01 7.38360584e-01 -3.31067115e-01 -2.64132053e-01
2.29847401e-01 4.64267284e-01 -1.22940993e+00 -5.84745407e-01
5.65664172e-01 5.17158210e-01 4.49255198e-01 -1.00360250e+00
-1.29977059e+00 -5.18494129e-01 -5.32780290e-01 8.60223472e-02
-3.02351475e-01 -9.32943895e-02 1.39884874e-01 -5.16216815e-01
1.71149552e-01 -9.98490691e-01 -3.53555202e-01 -1.29400700e-01
4.43932302e-02 -3.10707033e-01 1.11968517e+00 -7.04270124e-01
1.61294484e+00 -2.27713895e+00 -9.30714965e-01 5.93720257e-01
-1.31410472e-02 1.66361183e-01 -2.12359965e-01 5.18391669e-01
8.54735821e-02 1.97227880e-01 -4.70494747e-01 1.15000702e-01
3.28337736e-02 6.12958133e-01 -4.84625489e-01 6.67517602e-01
3.85917835e-02 9.43602324e-01 -1.11236370e+00 5.24927303e-02
6.06487751e-01 1.79067865e-01 -8.40657111e-03 -4.31678325e-01
-1.46357626e-01 6.60472453e-01 -3.92118424e-01 7.86749065e-01
5.18512011e-01 1.56193480e-01 -1.41494393e-01 4.44014966e-01
-8.31336081e-01 3.23009551e-01 -1.05861974e+00 6.58764839e-01
-5.16585767e-01 1.28288865e+00 -1.97312266e-01 -7.31847107e-01
1.03179049e+00 3.20754766e-01 5.83992898e-01 -1.02893221e+00
-2.69779086e-01 8.60036686e-02 -3.31727147e-01 -5.58450401e-01
3.37387770e-01 8.29966441e-02 -1.19193360e-01 7.04043984e-01
-9.37841654e-01 1.27091065e-01 1.32362470e-01 -4.19971228e-01
1.07023144e+00 -4.79750007e-01 4.29710299e-01 -8.77070948e-02
4.28203642e-01 2.31165379e-01 7.47156203e-01 5.67218184e-01
-4.23904717e-01 -3.35312709e-02 -6.02379963e-02 -1.05079412e+00
-1.03157151e+00 -1.02819610e+00 -1.98635846e-01 1.27921677e+00
-3.26622516e-01 1.81463435e-01 2.43565347e-02 -3.01836371e-01
3.36075217e-01 1.15231586e+00 -4.46283460e-01 1.15521267e-01
-5.41312158e-01 -1.29002357e+00 5.67006290e-01 3.86279702e-01
5.55180669e-01 -9.95499074e-01 -5.10174513e-01 3.68364424e-01
-3.09430867e-01 -8.78944099e-01 -1.64014444e-01 -2.38889852e-03
-8.95561397e-01 -1.06249678e+00 -7.28702694e-02 -4.83297557e-01
7.68374026e-01 6.25298381e-01 9.19908285e-01 -1.86419398e-01
-2.44317532e-01 5.77927828e-01 5.21403290e-02 -6.30709291e-01
-2.62226015e-01 1.99360885e-02 2.64275223e-01 2.19649136e-01
6.16689086e-01 -8.77400100e-01 -5.24375200e-01 4.04855520e-01
-7.77676284e-01 -1.66928396e-01 1.84181631e-01 1.15000315e-01
4.33845013e-01 5.93846083e-01 6.91077888e-01 -3.86203200e-01
4.66449201e-01 -9.48146880e-01 -7.82199979e-01 6.65359050e-02
-9.74397957e-01 -3.42672616e-01 4.30744916e-01 -2.68573403e-01
-1.10252297e+00 1.90868884e-01 2.16200203e-01 1.64312929e-01
-8.27909470e-01 5.80684125e-01 1.70089230e-01 4.31578308e-01
7.34957576e-01 3.47708881e-01 -5.03858447e-01 -2.65044659e-01
3.26300383e-01 9.22857106e-01 4.98946339e-01 -1.04081787e-01
1.22089362e+00 9.63052511e-01 -2.66581655e-01 -1.06460655e+00
-6.25553012e-01 -9.91244614e-01 -7.71800280e-01 -4.87641960e-01
6.08018756e-01 -1.26926351e+00 -4.07407165e-01 6.71672761e-01
-7.48890162e-01 -6.05822802e-01 -5.11221826e-01 4.04232621e-01
-5.36712026e-03 -3.20098937e-01 -5.87484054e-02 -9.82029498e-01
-4.92146462e-02 -3.63049537e-01 7.23540366e-01 2.01212108e-01
-3.20148915e-01 -1.41103232e+00 6.48889840e-01 5.93059473e-02
7.28244543e-01 6.53349519e-01 9.27574039e-01 -2.07285732e-01
-4.82131600e-01 -1.63102388e-01 -2.26763025e-01 -2.39520129e-02
9.82642114e-01 2.65533954e-01 -9.40059721e-01 -1.93226308e-01
-5.44281602e-01 4.44451928e-01 6.90694153e-01 5.38296998e-01
7.00779676e-01 -6.57669544e-01 -5.57187855e-01 4.73351717e-01
1.31610179e+00 3.01890969e-01 3.85335535e-01 6.69175863e-01
7.34386563e-01 5.33945799e-01 1.69515640e-01 6.77126229e-01
8.25674057e-01 3.12342048e-01 4.85229939e-01 -4.52065408e-01
-2.10160520e-02 -8.34231824e-02 6.63973331e-01 5.40927112e-01
-1.03983395e-01 4.36521024e-02 -1.50573277e+00 1.15238595e+00
-1.88677883e+00 -1.52605867e+00 -6.79222286e-01 2.18914580e+00
4.46054965e-01 -1.63116127e-01 1.55344933e-01 -5.21444604e-02
8.09948802e-01 5.19947290e-01 -8.47684264e-01 -1.70024574e-01
-2.91483343e-01 -2.05008835e-01 1.08692908e+00 4.15877193e-01
-1.05531132e+00 8.93273711e-01 6.68753147e+00 2.50865668e-01
-1.45424366e+00 1.68034270e-01 5.21714151e-01 -1.18680485e-01
-5.58892131e-01 -2.01213792e-01 -7.17094064e-01 4.03687119e-01
1.11947620e+00 -2.23007992e-01 5.94234765e-01 5.76274991e-01
1.47311640e+00 -7.02415556e-02 -5.50164998e-01 6.72238648e-01
-3.86914551e-01 -1.20512176e+00 -2.18698651e-01 2.19141528e-01
1.23245978e+00 8.39829504e-01 -1.54560036e-03 2.49672130e-01
8.58662188e-01 -7.12307692e-01 4.63083446e-01 7.04932153e-01
6.81885064e-01 -5.59409499e-01 1.53539538e-01 3.68008614e-01
-1.71570516e+00 -5.70687473e-01 -1.55755887e-02 -5.50665736e-01
-2.67873239e-02 8.39100242e-01 -1.16227901e+00 -5.53123057e-02
7.28155434e-01 1.16349983e+00 -5.96008599e-01 7.20437706e-01
-3.68894309e-01 1.08269584e+00 -6.17029488e-01 3.71443212e-01
5.49828351e-01 -2.53825873e-01 6.04350030e-01 1.18215489e+00
5.86663485e-01 1.42356232e-01 3.66706401e-01 6.87567949e-01
1.11829512e-01 -2.79577188e-02 -1.04038143e+00 1.28992319e-01
5.53583324e-01 1.12971771e+00 -5.53603053e-01 -1.22694537e-01
-4.63873029e-01 3.83436561e-01 -2.03302443e-01 6.74424827e-01
-5.76260328e-01 2.34012306e-01 1.22776258e+00 4.30731446e-01
1.26941532e-01 -5.19405305e-01 -1.89756706e-01 -7.22237349e-01
-1.74065202e-01 -4.07686859e-01 3.55013609e-01 -6.48195744e-01
-1.19458461e+00 -2.61378046e-02 -1.51098952e-01 -1.34517169e+00
-3.34987849e-01 -6.48917630e-02 -1.12971509e+00 6.02995813e-01
-2.01926088e+00 -1.19750297e+00 -5.43449759e-01 7.93386161e-01
4.94921684e-01 1.72474399e-01 6.23259127e-01 2.23674178e-01
-7.41240382e-01 -1.64114758e-01 8.15365672e-01 1.74399391e-01
3.58613759e-01 -9.97271180e-01 6.90895259e-01 1.01471722e+00
-9.08320174e-02 -3.40384692e-02 3.39622319e-01 -7.62135506e-01
-6.72067225e-01 -2.01190782e+00 1.42675495e+00 -1.83929041e-01
1.05628586e+00 4.37977165e-02 -6.54967487e-01 8.99774849e-01
-3.23000491e-01 -1.16407990e-01 7.70443320e-01 2.48073906e-01
-1.47447616e-01 -5.53505361e-01 -1.20996749e+00 7.24529147e-01
9.52186286e-01 -7.70800829e-01 -2.48942211e-01 5.42605221e-01
2.44390607e-01 3.26234788e-01 -5.07231355e-01 1.71302095e-01
2.36814171e-01 -5.78573525e-01 6.16072834e-01 -3.43048990e-01
-2.28695318e-01 -7.04026937e-01 -2.96019852e-01 -1.53625405e+00
-8.31859827e-01 -1.95793524e-01 4.96780932e-01 1.32013667e+00
3.22076142e-01 -9.09490585e-01 3.75817299e-01 8.78514647e-01
-4.55118977e-02 5.10620028e-02 -1.02327013e+00 -6.37931883e-01
-1.24326311e-01 -9.06761050e-01 1.18362868e+00 1.31113267e+00
-4.61125970e-01 -4.64036241e-02 -2.81329285e-02 8.64732206e-01
3.05887878e-01 8.09335336e-02 8.38337362e-01 -1.55591607e+00
6.12496197e-01 -6.83809280e-01 -3.59592140e-01 -4.06222790e-01
2.16991916e-01 -8.97560120e-01 -3.13657336e-02 -1.67253137e+00
-1.45726964e-01 -7.74069846e-01 -2.92449057e-01 1.05227923e+00
1.02942444e-01 1.86849028e-01 -1.36875004e-01 4.80285048e-01
-1.36439830e-01 6.29810750e-01 5.50947726e-01 -6.15695894e-01
-7.64742970e-01 1.49733052e-01 -2.12506533e-01 8.71814668e-01
1.03111267e+00 -5.07497311e-01 -2.84576029e-01 -4.04870123e-01
4.91368979e-01 -4.44132805e-01 5.06381869e-01 -1.25586784e+00
8.31550583e-02 -8.65993440e-01 2.90689677e-01 -8.99809420e-01
-2.41248935e-01 -1.13810587e+00 4.47005868e-01 5.22774160e-01
-4.44922186e-02 2.52434909e-01 2.47144058e-01 5.49756050e-01
4.00698632e-02 6.62527800e-01 5.17978549e-01 6.50396124e-02
-1.09970295e+00 6.19295478e-01 -1.22522569e+00 -2.05498219e-01
1.04316437e+00 -4.36127514e-01 -2.23041236e-01 -5.45814514e-01
-6.34976268e-01 5.69261968e-01 5.71500719e-01 5.31740963e-01
1.42571434e-01 -1.52623403e+00 -9.06410933e-01 3.64667416e-01
4.01084244e-01 -2.97403216e-01 7.00702146e-02 7.57673383e-01
-2.36715272e-01 4.39640999e-01 9.95621737e-03 -5.44703424e-01
-1.10099709e+00 2.20226511e-01 5.59837282e-01 -1.25072107e-01
-5.39477706e-01 -1.35180950e-02 -5.78943342e-02 -9.63643849e-01
-2.96221197e-01 -6.85252726e-01 -2.65831441e-01 4.80109781e-01
4.09146339e-01 6.15189672e-01 5.55886403e-02 -9.73352432e-01
-3.73968601e-01 4.66513008e-01 7.03474343e-01 1.06850907e-01
1.61710227e+00 -8.37895155e-01 -4.02431376e-02 9.53708231e-01
8.77984166e-01 -1.04089551e-01 -1.26313508e+00 -5.40214837e-01
3.37230742e-01 -2.91136384e-01 3.58300060e-01 -6.65410042e-01
-1.09928143e+00 3.89905602e-01 1.12091482e+00 3.22766751e-01
1.26939392e+00 -1.71860168e-03 9.78112757e-01 5.70526898e-01
4.12538797e-02 -1.33047581e+00 -6.66159093e-01 5.47317922e-01
4.75587666e-01 -1.37562060e+00 -2.71350503e-01 3.13912541e-01
-1.68855011e-01 7.81929970e-01 1.50917530e-01 4.25294608e-01
9.23608840e-01 -1.75577149e-01 3.03194135e-01 -2.07755521e-01
-6.60973370e-01 -8.70170295e-01 1.06525138e-01 6.84696972e-01
8.02237168e-02 7.52777576e-01 1.50275007e-01 -4.26215619e-01
-6.54725134e-02 -1.80464879e-01 3.17268968e-01 7.13243484e-01
-7.25872278e-01 -6.22160852e-01 -7.76383340e-01 5.87503433e-01
2.03658953e-01 -2.06708938e-01 -1.72527626e-01 7.03341424e-01
5.25738537e-01 1.23073387e+00 7.30242789e-01 -9.61233079e-02
3.12457025e-01 2.26704776e-01 -4.40859318e-01 -2.20627427e-01
-3.69201750e-01 -3.36705297e-01 -8.53711441e-02 -4.34788287e-01
-5.95171332e-01 -1.21701133e+00 -1.31481731e+00 -7.24776268e-01
2.26591036e-01 -3.88708770e-01 8.47664058e-01 1.27214181e+00
4.60322589e-01 1.98073670e-01 1.32033837e+00 -6.11575365e-01
3.71482044e-01 -8.65381420e-01 -6.50955319e-01 4.42823082e-01
6.93647385e-01 -4.37361568e-01 -4.59505200e-01 1.83092594e-01] | [9.431089401245117, -1.3093488216400146] |
ea48c47b-7dff-41d8-abed-a8f3547f2aa6 | learning-to-predict-scene-level-implicit-3d-1 | 2306.08671 | null | https://arxiv.org/abs/2306.08671v1 | https://arxiv.org/pdf/2306.08671v1.pdf | Learning to Predict Scene-Level Implicit 3D from Posed RGBD Data | We introduce a method that can learn to predict scene-level implicit functions for 3D reconstruction from posed RGBD data. At test time, our system maps a previously unseen RGB image to a 3D reconstruction of a scene via implicit functions. While implicit functions for 3D reconstruction have often been tied to meshes, we show that we can train one using only a set of posed RGBD images. This setting may help 3D reconstruction unlock the sea of accelerometer+RGBD data that is coming with new phones. Our system, D2-DRDF, can match and sometimes outperform current methods that use mesh supervision and shows better robustness to sparse data. | ['David F. Fouhey', 'Justin Johnson', 'Linyi Jin', 'Nilesh Kulkarni'] | 2023-06-14 | learning-to-predict-scene-level-implicit-3d | http://openaccess.thecvf.com//content/CVPR2023/html/Kulkarni_Learning_To_Predict_Scene-Level_Implicit_3D_From_Posed_RGBD_Data_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Kulkarni_Learning_To_Predict_Scene-Level_Implicit_3D_From_Posed_RGBD_Data_CVPR_2023_paper.pdf | cvpr-2023-1 | ['3d-reconstruction'] | ['computer-vision'] | [ 2.97535509e-01 4.45164353e-01 -4.41268049e-02 -5.32085180e-01
-7.86627352e-01 -5.34118533e-01 3.70353371e-01 -3.12210768e-01
-3.35852019e-02 4.59431618e-01 2.67625093e-01 -3.38301569e-01
2.27476254e-01 -8.66997898e-01 -1.31051862e+00 -1.56414688e-01
1.04813501e-01 7.53281891e-01 2.05750108e-01 -1.78566664e-01
2.90421154e-02 7.70428360e-01 -1.40837276e+00 2.70719409e-01
1.06474549e-01 1.20222104e+00 3.40376310e-02 1.04746926e+00
-4.35767602e-03 5.70033073e-01 -4.13804054e-01 1.90744996e-02
6.04894936e-01 -1.26293739e-02 -7.11606622e-01 4.38959956e-01
7.73707390e-01 -6.75842941e-01 -3.55261981e-01 6.24329090e-01
2.68765628e-01 -1.65418357e-01 3.49888772e-01 -1.21718335e+00
-3.52005750e-01 -2.81495690e-01 -2.95831263e-01 -3.17009538e-01
1.00701582e+00 4.45365719e-03 5.30950725e-01 -1.02242470e+00
8.31551194e-01 1.21378589e+00 1.28788304e+00 4.76831347e-01
-1.47433686e+00 -2.87281781e-01 -2.68743970e-02 -5.15360653e-01
-1.37059665e+00 -4.11207795e-01 8.73281896e-01 -3.18311542e-01
1.05882442e+00 4.45664644e-01 1.07143962e+00 1.24690378e+00
-5.17260358e-02 5.12754738e-01 1.23255777e+00 -1.14721293e-02
3.35329264e-01 3.28709818e-02 -1.92179129e-01 8.35610688e-01
-9.14744735e-02 -3.19199786e-02 -1.00911725e+00 -2.14842632e-01
1.31548870e+00 1.28665492e-01 -3.01858753e-01 -7.54320920e-01
-1.22703993e+00 4.82197821e-01 6.78605199e-01 -3.57212096e-01
-1.10640980e-01 5.52257359e-01 -1.69545233e-01 3.59549522e-01
5.10194540e-01 1.88590884e-01 -7.66368747e-01 -4.02200997e-01
-7.57874608e-01 2.87944004e-02 1.00308454e+00 1.18876207e+00
1.23906147e+00 -6.03491664e-02 1.01078677e+00 6.02671281e-02
6.23951912e-01 9.32552338e-01 6.05713800e-02 -1.48487759e+00
3.21311444e-01 5.34901321e-01 2.43854180e-01 -9.60654557e-01
-4.14582968e-01 6.26772791e-02 -4.21177417e-01 5.55455506e-01
1.73666626e-01 2.43016630e-01 -9.16863859e-01 1.01672292e+00
4.87237513e-01 6.89222813e-01 -4.88624871e-02 1.06200075e+00
7.18024909e-01 2.92357326e-01 -8.56975138e-01 3.11868578e-01
5.44079840e-01 -3.21482867e-01 -3.52581143e-01 -3.85912180e-01
4.77355361e-01 -2.40379840e-01 1.38782310e+00 6.58617735e-01
-8.86388838e-01 -4.93048608e-01 -1.33886099e+00 -5.84132075e-01
-2.30033502e-01 -3.86902004e-01 4.75935012e-01 6.77740097e-01
-1.34263206e+00 7.59288073e-01 -1.31551230e+00 -3.59868556e-01
1.84801355e-01 6.51988387e-01 -7.24322855e-01 -1.25046313e-01
-6.04452908e-01 8.69553030e-01 -2.29778454e-01 7.04640672e-02
-7.21031427e-01 -4.83113974e-01 -1.05572581e+00 -8.95766854e-01
4.35198471e-02 -8.63259792e-01 1.15684891e+00 -5.98950982e-01
-1.57295299e+00 1.09836411e+00 -1.18185915e-01 -3.20189625e-01
3.68665785e-01 -6.29759133e-01 -1.55912675e-02 2.66467541e-01
6.39285594e-02 7.36443222e-01 9.06038404e-01 -1.78458369e+00
-5.90704121e-02 -5.40667415e-01 3.93766731e-01 2.93942869e-01
8.67750198e-02 -8.24144185e-01 -3.96243513e-01 -2.00036287e-01
8.16161454e-01 -9.25289631e-01 -2.11832136e-01 6.52764261e-01
-4.59416211e-01 5.67427576e-01 1.10381961e+00 -5.69508493e-01
2.35047981e-01 -2.03454852e+00 4.00143355e-01 4.28518444e-01
1.42748103e-01 -5.35213888e-01 2.16344908e-01 -3.26029435e-02
1.80576935e-01 2.09900647e-01 -4.13936287e-01 -1.09029996e+00
3.43222544e-02 9.18443620e-01 -5.18903136e-01 8.57846022e-01
2.42441431e-01 5.61453402e-01 -9.51652288e-01 -3.02627623e-01
5.03912985e-01 9.44018900e-01 -6.14716828e-01 3.56034130e-01
-2.63871938e-01 8.24305534e-01 -3.51315558e-01 9.11303639e-01
6.98786199e-01 -3.55145693e-01 -1.65375713e-02 -3.18798184e-01
5.47471270e-02 7.67309785e-01 -1.33228946e+00 2.62694168e+00
-4.73088801e-01 4.36436057e-01 3.01279545e-01 -7.89546669e-01
8.43569756e-01 1.48092210e-01 7.67371655e-01 -7.21345484e-01
-3.78325805e-02 2.50080019e-01 -9.50982928e-01 -3.41379523e-01
5.61509430e-01 -2.92161912e-01 -2.35726595e-01 3.95155698e-01
-8.02104324e-02 -9.86741483e-01 -9.60124552e-01 1.63645446e-01
1.45609677e+00 9.71518993e-01 -1.79233849e-01 1.10091984e-01
-9.82367396e-02 1.63981974e-01 2.59178966e-01 5.58653712e-01
2.79339403e-01 1.08909202e+00 7.92115852e-02 -5.72578311e-01
-1.05642045e+00 -1.45471215e+00 -3.43398809e-01 2.86596626e-01
5.02537012e-01 -4.19935495e-01 -3.25012803e-01 -6.58199847e-01
1.88754693e-01 4.91148263e-01 -7.36357450e-01 3.35524455e-02
-6.68359101e-01 -7.11579397e-02 4.08210665e-01 6.01395667e-01
3.20218801e-01 -2.14928746e-01 -7.59716749e-01 5.50777232e-03
5.67083769e-02 -1.26119125e+00 -1.25366583e-01 6.46921098e-01
-1.38361025e+00 -1.12562263e+00 -7.24909604e-02 -3.53271097e-01
7.68356919e-01 2.51656353e-01 1.42991090e+00 1.01764411e-01
2.65893936e-01 1.16495776e+00 -4.45364326e-01 -5.77374958e-02
-2.90607661e-01 -2.42783651e-01 4.45356041e-01 -2.32380718e-01
-2.91477758e-02 -9.84624922e-01 -4.48562801e-01 4.22361255e-01
-6.29854560e-01 2.97333211e-01 -4.94858995e-02 4.21326429e-01
1.30581367e+00 -3.26865137e-01 -3.98076355e-01 -6.76989555e-01
-2.44480640e-01 -4.47383165e-01 -4.02479380e-01 -3.84001583e-01
-8.95867422e-02 5.50754964e-02 2.38492712e-01 -3.11055481e-01
-3.64680558e-01 7.92889297e-01 -3.47531855e-01 -9.88771379e-01
-2.02005997e-01 1.49362057e-01 -2.53317416e-01 -3.28225136e-01
7.47973800e-01 -1.02161027e-01 -3.50782275e-02 -9.03609574e-01
4.71548378e-01 3.81741375e-01 7.33728051e-01 -6.40400231e-01
1.04909372e+00 9.53753471e-01 2.05233648e-01 -7.59615302e-01
-7.66484022e-01 -1.54875681e-01 -1.03794682e+00 -1.38103798e-01
8.28814149e-01 -1.25434947e+00 -5.44885755e-01 4.88128141e-02
-1.20257509e+00 -8.32885206e-01 -7.95545459e-01 2.79813707e-01
-8.43524873e-01 3.53813693e-02 -4.48355049e-01 -6.10882878e-01
3.89618874e-01 -1.02039325e+00 1.95567894e+00 -3.48905474e-01
-3.54149848e-01 -9.43126619e-01 2.65180379e-01 1.36897504e-01
-1.01343937e-01 8.56150925e-01 1.99015856e-01 1.53710857e-01
-8.70482087e-01 -1.30696774e-01 3.42400312e-01 3.08957368e-01
3.76591027e-01 -3.23475331e-01 -1.31705487e+00 -9.68988910e-02
4.06923503e-01 -5.08148432e-01 2.34294415e-01 1.04756849e-02
1.04368460e+00 -4.18464601e-01 -3.07578117e-01 1.14423704e+00
1.39015698e+00 -4.63548362e-01 4.88768518e-01 5.25269210e-01
1.14296329e+00 2.36261487e-01 5.15498161e-01 3.70640934e-01
8.39123368e-01 7.36998558e-01 1.00992727e+00 -4.47886854e-01
-1.93651259e-01 -6.01114273e-01 4.94803011e-01 9.98021007e-01
-4.16235775e-02 2.41786614e-01 -9.21477258e-01 1.09934181e-01
-1.64789033e+00 -2.09620252e-01 -3.61597240e-01 2.14139056e+00
6.41963840e-01 4.25630867e-01 1.40223861e-01 4.30820793e-01
6.26732856e-02 -2.11968925e-02 -9.47041214e-01 -2.41309643e-01
-2.44963974e-01 5.12219131e-01 8.19536567e-01 9.16518569e-01
-7.19446719e-01 4.12338018e-01 7.52433205e+00 -1.34555429e-01
-1.04410791e+00 1.23426393e-01 2.96921343e-01 -3.18350613e-01
-7.87015915e-01 2.57023931e-01 -4.70005065e-01 2.00965524e-01
9.20911610e-01 5.39623737e-01 4.88692790e-01 8.49518955e-01
4.03105505e-02 -2.14010462e-01 -1.59657836e+00 1.22815752e+00
2.33717784e-02 -1.39312494e+00 -1.93786219e-01 3.84973317e-01
5.42793810e-01 3.52438629e-01 -9.27027985e-02 -1.34888232e-01
4.70032662e-01 -1.08788681e+00 1.08228695e+00 7.59150743e-01
1.17917192e+00 -2.19510585e-01 2.32130677e-01 6.16570652e-01
-1.12549984e+00 4.67439383e-01 -1.49860740e-01 -3.81249547e-01
2.18855977e-01 5.97741306e-01 -1.01510441e+00 2.95717716e-01
9.83256757e-01 1.13191974e+00 -5.32347202e-01 4.89896119e-01
-2.45319143e-01 3.91046613e-01 -9.27697599e-01 4.85041142e-01
-1.22751720e-01 -1.10550523e-01 5.38332105e-01 6.11971438e-01
5.33193588e-01 2.30992526e-01 1.74417362e-01 6.45301342e-01
9.60542355e-03 -5.76138616e-01 -1.35401928e+00 4.87539500e-01
3.11400741e-01 6.78119361e-01 -5.82924008e-01 -1.94407552e-01
-4.17064965e-01 1.29909503e+00 9.76758674e-02 1.53405145e-01
-7.60380805e-01 3.72429639e-01 7.11010635e-01 5.90382040e-01
2.01566666e-01 -9.14139271e-01 -6.42385542e-01 -1.33211017e+00
-2.31018383e-02 -5.16752899e-01 -4.50518057e-02 -1.43310499e+00
-1.07700264e+00 2.63610840e-01 -8.76817107e-02 -1.62296855e+00
-1.44677743e-01 -7.71331489e-01 7.12495670e-02 5.44215083e-01
-1.03622258e+00 -9.04656231e-01 -4.45634186e-01 8.79643321e-01
2.55668402e-01 5.15499592e-01 1.07053661e+00 1.68218955e-01
3.84497419e-02 4.24111076e-02 -1.57550365e-01 -4.03735526e-02
3.14254493e-01 -1.56649256e+00 7.54298866e-01 5.24448216e-01
4.96394932e-01 3.76660466e-01 6.84734285e-01 -7.58507788e-01
-2.18805075e+00 -7.00419188e-01 2.47828811e-02 -1.20272636e+00
3.97114486e-01 -7.88072169e-01 -6.55103505e-01 1.16673958e+00
-3.59955162e-01 5.01803041e-01 5.01338422e-01 3.58435651e-03
-4.96861011e-01 -1.96894616e-01 -1.31791592e+00 1.68457717e-01
1.54262292e+00 -1.16235030e+00 -5.50237954e-01 3.64527702e-01
1.03880668e+00 -1.31227338e+00 -1.13625205e+00 4.10317779e-01
5.61960816e-01 -1.09663010e+00 1.31474400e+00 -2.38974616e-01
1.59183517e-01 -6.29668176e-01 -9.57382441e-01 -1.15551102e+00
2.20435053e-01 -6.77589774e-01 -6.85790658e-01 5.78241289e-01
1.30690215e-03 -3.89407814e-01 1.14772582e+00 8.35106015e-01
-3.35094541e-01 -5.03744900e-01 -1.12459934e+00 -6.55136585e-01
-1.91287339e-01 -1.01001871e+00 6.95313632e-01 1.00678587e+00
-4.85605836e-01 2.54916281e-01 -3.50618035e-01 5.51026106e-01
6.75085306e-01 9.67185274e-02 1.18701637e+00 -1.02716410e+00
-4.46543843e-01 1.47482410e-01 -7.69880176e-01 -1.73449981e+00
-1.04851969e-01 -8.49386513e-01 3.20802703e-02 -1.65201080e+00
-5.82994819e-01 -7.86179364e-01 5.88435940e-02 6.09911442e-01
5.37567139e-01 8.45271349e-01 -1.39847055e-01 1.51614606e-01
-6.00294530e-01 3.78641814e-01 1.01984882e+00 -9.03650150e-02
-1.94706425e-01 -2.50997454e-01 -4.11933720e-01 9.16430116e-01
4.49023843e-01 -6.04873121e-01 -2.77306527e-01 -1.01380920e+00
6.29533768e-01 2.30831504e-01 6.54767990e-01 -1.09931898e+00
-2.47245841e-03 1.63677998e-03 7.69021928e-01 -8.62177253e-01
9.87515569e-01 -1.23033822e+00 7.75833309e-01 2.45792270e-01
1.80249393e-01 2.49148577e-01 3.00526619e-01 5.78427911e-01
2.15317026e-01 1.90920413e-01 1.64628610e-01 -3.02165180e-01
-4.28185016e-01 3.76153260e-01 -5.51445074e-02 -1.09875470e-01
3.96818370e-01 -8.47573817e-01 1.11864030e-01 -3.44424874e-01
-9.33410168e-01 -1.97107434e-01 1.30265176e+00 2.64722854e-01
1.02362311e+00 -1.64383447e+00 -1.85463160e-01 3.45313162e-01
-1.76594913e-01 9.23307896e-01 -3.84580940e-01 6.02166355e-01
-7.79569149e-01 -2.57659853e-01 -1.16525581e-02 -1.25371957e+00
-8.06127667e-01 1.16203420e-01 5.59589267e-01 2.84548789e-01
-1.13414407e+00 8.25974166e-01 -1.42098665e-01 -9.38624859e-01
2.86949784e-01 -7.33159304e-01 6.79504693e-01 -3.16699535e-01
-2.95908246e-02 9.64549631e-02 3.49857211e-01 -8.71602774e-01
-6.53752208e-01 9.67435896e-01 8.41442406e-01 -4.01562154e-01
1.50529075e+00 -2.51056969e-01 1.28390476e-01 1.11266637e+00
1.49189961e+00 2.73208112e-01 -1.85454977e+00 -1.01819269e-01
-2.44406402e-01 -7.15836227e-01 6.47382736e-02 -4.64594364e-01
-8.87450874e-01 9.56270337e-01 6.24413550e-01 2.95065075e-01
9.25617754e-01 1.45382464e-01 8.15273106e-01 6.11433923e-01
9.34007823e-01 -7.21473575e-01 4.33190018e-01 3.98592323e-01
8.30010235e-01 -1.12649930e+00 4.53359783e-01 -2.29625985e-01
-1.96779579e-01 9.60507274e-01 3.03317398e-01 -4.69542265e-01
8.69427323e-01 7.08056450e-01 5.53039089e-02 -3.74195188e-01
-2.43231878e-01 3.69274281e-02 7.55336359e-02 9.99445915e-01
-6.67769909e-02 -1.90890692e-02 1.03460014e+00 9.77462754e-02
-7.41840899e-01 -4.74685840e-02 3.68610054e-01 1.21717370e+00
-2.74340510e-01 -1.09780204e+00 -7.33637691e-01 3.25529993e-01
1.92857772e-01 2.34142527e-01 -5.59555471e-01 9.32769179e-01
2.89643526e-01 6.28212452e-01 1.71573654e-01 -8.30408037e-01
7.21679807e-01 -1.82626233e-01 1.06307364e+00 -6.72385573e-01
-2.90086061e-01 7.18309358e-02 1.85066834e-01 -1.22105730e+00
-7.99141824e-01 -7.69795060e-01 -1.59514475e+00 -2.69108444e-01
1.40363611e-02 -2.78299928e-01 9.05100048e-01 8.36272180e-01
3.14654052e-01 1.14046596e-01 8.04293931e-01 -1.51031017e+00
1.54865980e-01 -4.16554809e-01 -5.57742596e-01 2.52110273e-01
9.03637230e-01 -7.61433303e-01 -4.64482486e-01 3.41763675e-01] | [8.486034393310547, -2.818328857421875] |
cab6c556-f0fe-4706-8890-4a33997e2af6 | fine-tuning-pre-trained-language-model-with | 2010.07835 | null | https://arxiv.org/abs/2010.07835v3 | https://arxiv.org/pdf/2010.07835v3.pdf | Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach | Fine-tuned pre-trained language models (LMs) have achieved enormous success in many natural language processing (NLP) tasks, but they still require excessive labeled data in the fine-tuning stage. We study the problem of fine-tuning pre-trained LMs using only weak supervision, without any labeled data. This problem is challenging because the high capacity of LMs makes them prone to overfitting the noisy labels generated by weak supervision. To address this problem, we develop a contrastive self-training framework, COSINE, to enable fine-tuning LMs with weak supervision. Underpinned by contrastive regularization and confidence-based reweighting, this contrastive self-training framework can gradually improve model fitting while effectively suppressing error propagation. Experiments on sequence, token, and sentence pair classification tasks show that our model outperforms the strongest baseline by large margins on 7 benchmarks in 6 tasks, and achieves competitive performance with fully-supervised fine-tuning methods. | ['Chao Zhang', 'Tuo Zhao', 'Wendi Ren', 'Haoming Jiang', 'Simiao Zuo', 'Yue Yu'] | 2020-10-15 | null | https://aclanthology.org/2021.naacl-main.84 | https://aclanthology.org/2021.naacl-main.84.pdf | naacl-2021-4 | ['sentence-pair-classification'] | ['natural-language-processing'] | [ 3.48224938e-01 -4.82234769e-02 -5.47187030e-01 -6.85189545e-01
-1.07883918e+00 -5.89314818e-01 6.38305724e-01 3.05607349e-01
-8.92018557e-01 7.91072309e-01 3.70298594e-01 -2.67165244e-01
3.10692728e-01 -3.45601916e-01 -6.88462973e-01 -4.09409940e-01
1.92359298e-01 4.80137825e-01 1.68521985e-01 -4.03334737e-01
1.87911078e-01 -8.69997740e-02 -1.04623747e+00 4.94987458e-01
9.79283392e-01 6.59700513e-01 1.52566105e-01 4.89686638e-01
-3.98863643e-01 7.04418778e-01 -4.78659242e-01 -4.14907783e-01
2.02363487e-02 -9.70773250e-02 -7.29965687e-01 -1.39623001e-01
5.24197876e-01 9.34260860e-02 3.36024091e-02 9.21855152e-01
4.59806770e-01 2.84388572e-01 5.15733600e-01 -8.97533774e-01
-7.16485381e-01 1.03507710e+00 -5.28453588e-01 2.34957129e-01
1.52858412e-02 8.46934468e-02 1.25639880e+00 -1.17004120e+00
3.16290498e-01 1.39345646e+00 1.01259959e+00 8.51865709e-01
-1.61504412e+00 -7.52134919e-01 4.01607454e-01 -2.45854691e-01
-1.35194755e+00 -7.20229566e-01 4.09722388e-01 -2.88664162e-01
1.16864872e+00 -5.16028889e-02 -1.90054998e-01 1.27581406e+00
-1.73672512e-02 8.27490091e-01 1.04885638e+00 -7.29019821e-01
9.90956947e-02 3.10284197e-01 1.21740423e-01 6.03333235e-01
-1.03379473e-01 -1.21039033e-01 -4.29197878e-01 -2.21448377e-01
3.91875982e-01 -1.71898514e-01 -3.00728474e-02 -2.44662613e-01
-1.29274476e+00 8.61290038e-01 2.22959444e-01 5.57997525e-01
2.35420689e-02 1.32951126e-01 6.65187716e-01 5.33628345e-01
8.05810273e-01 8.31933737e-01 -1.10994065e+00 -2.26336807e-01
-1.26011193e+00 -1.75449714e-01 8.12920928e-01 8.79218757e-01
7.70564675e-01 1.53211951e-01 -5.69077015e-01 1.32606745e+00
2.67348260e-01 4.03477818e-01 8.98938119e-01 -8.07072163e-01
7.83620179e-01 3.63024116e-01 -1.52243331e-01 -4.77804750e-01
-4.50408667e-01 -7.35467494e-01 -1.02894831e+00 -7.20273107e-02
3.39124501e-01 -3.03716063e-01 -9.06855345e-01 1.99450922e+00
1.89462851e-03 2.48505399e-01 -1.50379851e-01 4.55699831e-01
5.89092553e-01 6.15029097e-01 4.15339321e-01 -3.18419307e-01
9.46586192e-01 -1.47190619e+00 -5.28119683e-01 -6.14609659e-01
8.98842633e-01 -7.77220070e-01 1.62858236e+00 3.37393612e-01
-1.05835247e+00 -7.02622831e-01 -1.03828561e+00 -1.36600494e-01
-7.66995624e-02 2.08502024e-01 4.35226083e-01 4.85384256e-01
-9.39059436e-01 7.62306988e-01 -8.44465315e-01 -1.28969148e-01
3.15781087e-01 3.01065862e-01 -3.69781226e-01 -5.36603741e-02
-1.30833924e+00 9.05544639e-01 3.40695739e-01 -1.51414469e-01
-6.65865600e-01 -8.99503946e-01 -8.57213259e-01 2.20555320e-01
3.80656838e-01 -5.43654501e-01 1.50183010e+00 -8.76259625e-01
-1.65270329e+00 8.63222182e-01 -3.34356904e-01 -5.43350518e-01
5.04794955e-01 -3.61457586e-01 -3.65069836e-01 -3.06504458e-01
1.20749660e-01 6.77199900e-01 1.07601476e+00 -1.02153301e+00
-4.45050836e-01 8.55809078e-03 -2.84287483e-01 1.08441792e-01
-8.02885413e-01 1.13587990e-01 -5.18865705e-01 -8.60022724e-01
-2.69875735e-01 -7.61297405e-01 -6.31846488e-01 -4.18027133e-01
-3.08234900e-01 -4.42667872e-01 3.56333733e-01 -1.20558806e-01
1.49872184e+00 -2.07436085e+00 -2.05370501e-01 7.82973692e-02
1.86676547e-01 6.50651336e-01 -7.62680769e-01 3.02248925e-01
-4.63317297e-02 2.40703031e-01 -2.95725107e-01 -9.06703234e-01
1.39328673e-01 3.13515007e-01 -2.77859598e-01 1.08802475e-01
4.70658034e-01 8.93894255e-01 -1.15068316e+00 -4.31419969e-01
-8.01544860e-02 3.62416655e-01 -6.78964496e-01 2.14648515e-01
-4.27949846e-01 3.94028813e-01 -3.29730093e-01 1.46870151e-01
4.09335494e-01 -5.89279652e-01 1.87219828e-01 -4.66714017e-02
6.21903315e-02 6.15235627e-01 -1.12811077e+00 1.85162544e+00
-7.83286631e-01 4.60738003e-01 -1.60841905e-02 -1.06304169e+00
9.84006763e-01 2.00018257e-01 1.84076101e-01 -5.65110207e-01
-1.07301943e-01 1.87752053e-01 -1.94704041e-01 -4.50920880e-01
4.51296121e-01 -2.66419470e-01 -6.21814728e-02 5.97390771e-01
2.61823088e-01 -1.99302748e-01 3.29503655e-01 2.67813653e-01
1.25708449e+00 8.28832909e-02 2.97723711e-01 -1.29584491e-01
6.40883446e-01 -1.50902808e-01 7.94988215e-01 1.00520778e+00
-2.09293947e-01 5.80194473e-01 1.01146419e-02 -2.19968315e-02
-1.05198812e+00 -9.36544180e-01 -2.27926746e-01 1.88014412e+00
-4.77623761e-01 -8.05315673e-01 -7.07590401e-01 -9.46262896e-01
3.17189060e-02 5.84684074e-01 -4.00574952e-01 -2.93882817e-01
-7.27724910e-01 -8.94602716e-01 7.43682683e-01 6.42365158e-01
2.25972980e-01 -9.54904735e-01 2.36816049e-01 4.63010073e-01
-3.07722777e-01 -1.19472671e+00 -8.21302950e-01 7.01464593e-01
-9.59735036e-01 -5.77389240e-01 -5.56695044e-01 -1.02691126e+00
7.66021967e-01 2.37955615e-01 1.59102583e+00 5.74209429e-02
1.40966810e-02 1.07153222e-01 -2.73762345e-01 -2.38470688e-01
-6.61408484e-01 5.15967906e-01 3.59502077e-01 -2.93919265e-01
6.44536197e-01 -4.96758908e-01 -2.08822206e-01 2.86513120e-01
-7.87689686e-01 -3.23871642e-01 6.67767048e-01 1.30423939e+00
5.31436026e-01 -8.73985812e-02 9.02669847e-01 -1.21092188e+00
8.77708793e-01 -4.14541185e-01 -3.94062907e-01 4.55542296e-01
-9.56037760e-01 5.00984013e-01 9.39654827e-01 -6.02493465e-01
-1.12876821e+00 -6.45586327e-02 -2.90511340e-01 -2.00281858e-01
-1.07349366e-01 6.01679862e-01 4.25047651e-02 -1.94693953e-02
1.07599270e+00 -1.27768084e-01 -2.60505021e-01 -7.55577266e-01
6.26943827e-01 7.68186510e-01 6.10040843e-01 -6.80253923e-01
8.60124230e-01 1.52759388e-01 -4.19944435e-01 -5.82939863e-01
-1.46670401e+00 -6.36447787e-01 -7.22462952e-01 4.34545010e-01
3.49280387e-01 -1.14226985e+00 -3.24707359e-01 7.92245492e-02
-9.36928332e-01 -5.95579207e-01 -3.82102609e-01 4.44941401e-01
-2.81413525e-01 4.84247923e-01 -9.06746745e-01 -4.85722810e-01
-4.35384154e-01 -8.14256370e-01 9.37434196e-01 -1.89910993e-01
-3.89557153e-01 -1.28772020e+00 2.96594292e-01 4.15770143e-01
6.80318534e-01 -3.94797176e-01 7.01037228e-01 -7.98849821e-01
3.54234246e-03 -1.68718353e-01 -1.04639955e-01 8.30475152e-01
2.24389777e-01 -1.26836225e-01 -1.13743830e+00 -4.43206459e-01
6.47007721e-03 -8.05417478e-01 1.30838597e+00 1.39476821e-01
1.07289875e+00 -2.40372747e-01 -3.58152717e-01 6.75555408e-01
1.09337974e+00 -6.46519184e-01 -3.60039771e-02 3.06819320e-01
5.80969572e-01 3.67293775e-01 4.81502235e-01 4.57681477e-01
4.16761011e-01 4.87013787e-01 -7.04482645e-02 -4.67280224e-02
-1.88037366e-01 -5.30851603e-01 4.56922919e-01 1.27008963e+00
2.79281765e-01 -1.89750969e-01 -8.93943191e-01 3.78118932e-01
-1.84530854e+00 -7.35550165e-01 2.93419324e-02 2.07487679e+00
1.51727235e+00 4.67557847e-01 -9.46292207e-02 4.29415628e-02
5.76035202e-01 7.39521384e-02 -5.42507648e-01 -3.96952033e-01
-2.91111022e-01 3.87126088e-01 4.18247372e-01 5.83802819e-01
-1.33196032e+00 1.25835741e+00 6.81577921e+00 1.15286994e+00
-1.13618088e+00 3.61603677e-01 5.99662960e-01 -8.95768926e-02
-4.15065944e-01 -9.34671611e-02 -1.23854756e+00 2.92663783e-01
1.21845233e+00 4.59054969e-02 3.42791617e-01 7.25551367e-01
4.02070761e-01 2.21878663e-01 -1.24791467e+00 7.50037432e-01
8.79288316e-02 -1.24995446e+00 -1.95368573e-01 -4.82431889e-01
1.17316353e+00 6.47981822e-01 8.28928873e-02 7.36018896e-01
7.72803187e-01 -1.04447150e+00 3.20871711e-01 2.49850124e-01
8.92928541e-01 -4.04147714e-01 6.45673692e-01 7.49215603e-01
-9.41336632e-01 -1.44798145e-01 -5.23859262e-01 -1.21526606e-01
2.19675675e-01 9.30723131e-01 -9.44935143e-01 5.52117378e-02
4.38653618e-01 7.47998416e-01 -5.35719991e-01 8.59777033e-01
-5.34450948e-01 1.12710130e+00 -2.68626571e-01 -5.92398504e-03
3.85377437e-01 1.69604629e-01 2.13954538e-01 1.80052125e+00
-1.83349937e-01 -2.81249166e-01 3.53512973e-01 6.85990095e-01
-4.53371972e-01 2.64509439e-01 -9.52535793e-02 -4.06400068e-03
5.46631813e-01 1.31608605e+00 -2.78218418e-01 -4.16309953e-01
-5.41600108e-01 8.17942321e-01 8.31726909e-01 3.47053438e-01
-4.92520452e-01 -2.93360829e-01 4.65609610e-01 -9.48957130e-02
2.48596832e-01 -3.10620278e-01 -4.50998604e-01 -1.47146964e+00
-5.71865402e-02 -9.63154972e-01 4.67209011e-01 -3.47070605e-01
-1.78650224e+00 6.57721221e-01 -3.99061829e-01 -1.02053273e+00
-4.06701803e-01 -5.10830820e-01 -4.20046926e-01 7.84967601e-01
-2.00266027e+00 -1.03216910e+00 7.51686320e-02 4.80814040e-01
5.96201539e-01 -9.05731320e-02 1.02603662e+00 4.69862998e-01
-5.48475862e-01 1.15543890e+00 6.14339769e-01 1.92054674e-01
1.50921369e+00 -1.35078204e+00 5.80707669e-01 5.66922724e-01
3.13332498e-01 8.11863184e-01 4.73634601e-01 -4.95045424e-01
-7.88530052e-01 -1.19519222e+00 1.38549340e+00 -6.73407376e-01
8.97991180e-01 -5.86237848e-01 -1.01574719e+00 5.14958382e-01
3.61836441e-02 3.57926220e-01 8.10593545e-01 6.29858255e-01
-8.33359063e-01 -9.82172564e-02 -9.16618049e-01 3.65459263e-01
1.05540133e+00 -6.91698730e-01 -8.32468510e-01 5.38322031e-01
9.89896238e-01 -1.92875981e-01 -7.82737792e-01 4.44894224e-01
2.47887567e-01 -4.24240261e-01 1.00614083e+00 -1.02083313e+00
4.02603686e-01 3.43900248e-02 7.75195733e-02 -1.65044725e+00
-5.61345160e-01 -7.92455435e-01 -5.11703156e-02 1.55130315e+00
7.11161792e-01 -3.15744847e-01 7.92126715e-01 4.87589747e-01
-2.23910704e-01 -5.90319216e-01 -5.54098487e-01 -1.00157475e+00
4.81937021e-01 -6.67675793e-01 2.81315058e-01 1.07436252e+00
3.34021747e-01 8.10590446e-01 -4.07487810e-01 -2.41424158e-01
5.72653174e-01 -2.75588900e-01 5.59051991e-01 -1.13400519e+00
-4.59950119e-01 -5.14287293e-01 5.15574291e-02 -1.45307755e+00
5.83570182e-01 -1.08138454e+00 4.41366732e-01 -1.16517794e+00
3.25328767e-01 -7.53905058e-01 -6.32193983e-01 7.60198057e-01
-5.59312105e-01 3.97324860e-01 8.46359432e-02 2.11992860e-01
-1.11616838e+00 6.22421026e-01 9.73884165e-01 -2.24418968e-01
-2.59941459e-01 1.29776552e-01 -7.96357512e-01 8.15791070e-01
7.84745872e-01 -6.55449271e-01 -2.35639617e-01 -7.85526633e-01
4.34181571e-01 -5.15181839e-01 -2.31466845e-01 -7.66072273e-01
4.04121667e-01 1.01787876e-02 1.66890100e-01 -1.83614656e-01
1.70350730e-01 -4.44438219e-01 -7.10013151e-01 1.15899585e-01
-9.80759978e-01 -1.05485864e-01 2.20489338e-01 5.60209870e-01
-3.56680572e-01 -3.34497750e-01 9.24660146e-01 -1.31622270e-01
-4.82757151e-01 1.89003557e-01 -3.29220593e-01 8.13931406e-01
3.05412889e-01 2.54267991e-01 -1.85340524e-01 -2.35184208e-01
-7.72735357e-01 5.36189675e-01 8.68161619e-02 5.58188617e-01
2.41847321e-01 -1.27297795e+00 -9.28919315e-01 3.11779231e-01
2.07489759e-01 4.54365648e-02 -5.40368780e-02 7.47116804e-01
2.73649931e-01 5.64444065e-01 3.14518750e-01 -5.48311472e-01
-1.06725371e+00 4.57565874e-01 4.33420055e-02 -7.87447810e-01
-2.11600989e-01 1.33051586e+00 3.82823683e-02 -9.89839733e-01
5.53245366e-01 -4.74564046e-01 -1.43895015e-01 -1.71960935e-01
6.43549204e-01 1.60957262e-01 3.06937337e-01 -3.01160097e-01
-1.89798042e-01 4.27846760e-01 -3.48929524e-01 3.41055542e-02
1.42582190e+00 -3.16449374e-01 1.00437002e-02 5.73478639e-01
1.22737551e+00 1.81339055e-01 -1.27169979e+00 -1.01877522e+00
5.00156939e-01 -9.90732163e-02 -1.25256730e-02 -9.15811360e-01
-5.93313336e-01 1.06487596e+00 7.34141991e-02 -9.18933307e-04
7.62419522e-01 -1.59806803e-01 8.88460457e-01 8.14985216e-01
1.01666458e-01 -1.40426052e+00 3.43906611e-01 1.13322830e+00
4.31229770e-01 -1.72497606e+00 -2.76024163e-01 -1.68877169e-01
-6.63950205e-01 1.05469990e+00 6.83204591e-01 -1.09169170e-01
6.67507172e-01 5.75812280e-01 2.75641799e-01 3.55567396e-01
-1.18559527e+00 -9.33233928e-03 4.03072536e-01 4.45444912e-01
8.81181121e-01 -1.05687961e-01 -1.12698346e-01 7.65232980e-01
-1.13785520e-01 1.35946289e-01 9.96520668e-02 6.25794470e-01
-6.13967478e-01 -1.48000824e+00 -2.29360953e-01 4.97464448e-01
-6.12698615e-01 -6.49460673e-01 -7.27793649e-02 2.59086698e-01
2.36946307e-02 1.19384873e+00 -9.47661102e-02 -1.70387223e-01
1.67841703e-01 1.80563301e-01 2.76170313e-01 -1.00899160e+00
-8.99578452e-01 1.47524804e-01 2.00610086e-01 -4.57825482e-01
-3.46769780e-01 -5.07355034e-01 -1.17701173e+00 -3.62063795e-02
-4.60779101e-01 4.85209823e-01 5.12086272e-01 1.13142538e+00
3.61671895e-01 4.52204376e-01 7.49076128e-01 -5.52092016e-01
-1.44567835e+00 -1.22331929e+00 -2.35570267e-01 4.54616815e-01
5.50337255e-01 -2.78033257e-01 -5.82739055e-01 1.52666532e-02] | [10.826324462890625, 8.385525703430176] |
53187776-e9e9-46fb-a534-cbce68de90bf | selective-token-generation-for-few-shot-1 | 2209.08206 | null | https://arxiv.org/abs/2209.08206v1 | https://arxiv.org/pdf/2209.08206v1.pdf | Selective Token Generation for Few-shot Natural Language Generation | Natural language modeling with limited training data is a challenging problem, and many algorithms make use of large-scale pretrained language models (PLMs) for this due to its great generalization ability. Among them, additive learning that incorporates a task-specific adapter on top of the fixed large-scale PLM has been popularly used in the few-shot setting. However, this added adapter is still easy to disregard the knowledge of the PLM especially for few-shot natural language generation (NLG) since an entire sequence is usually generated by only the newly trained adapter. Therefore, in this work, we develop a novel additive learning algorithm based on reinforcement learning (RL) that selectively outputs language tokens between the task-general PLM and the task-specific adapter during both training and inference. This output token selection over the two generators allows the adapter to take into account solely the task-relevant parts in sequence generation, and therefore makes it more robust to overfitting as well as more stable in RL training. In addition, to obtain the complementary adapter from the PLM for each few-shot task, we exploit a separate selecting module that is also simultaneously trained using RL. Experimental results on various few-shot NLG tasks including question answering, data-to-text generation and text summarization demonstrate that the proposed selective token generation significantly outperforms the previous additive learning algorithms based on the PLMs. | ['Sungwoong Kim', 'Eun-Sol Kim', 'Taehwan Kwon', 'DaeJin Jo'] | 2022-09-17 | null | https://aclanthology.org/2022.coling-1.510 | https://aclanthology.org/2022.coling-1.510.pdf | coling-2022-10 | ['data-to-text-generation'] | ['natural-language-processing'] | [ 5.31446457e-01 3.05206954e-01 -3.23611110e-01 -9.37053263e-02
-1.04150569e+00 -1.42318189e-01 7.42040098e-01 1.59006059e-01
-4.13490117e-01 1.01029706e+00 2.92520493e-01 -1.26150295e-01
1.62556112e-01 -1.04525661e+00 -7.23232806e-01 -8.18167865e-01
4.26820457e-01 4.78759646e-01 2.31537625e-01 -4.69877511e-01
1.71191692e-01 -1.57020301e-01 -1.46060586e+00 2.19523892e-01
1.73740268e+00 7.31456518e-01 9.35757756e-01 5.64735889e-01
-7.54221737e-01 9.71325755e-01 -6.77780986e-01 -2.46892586e-01
-1.87360615e-01 -8.43702197e-01 -3.84166598e-01 -5.88522758e-03
6.98250625e-03 -3.27531457e-01 -2.98894420e-02 7.86339819e-01
8.90052199e-01 4.54080701e-01 6.35942459e-01 -1.05230355e+00
-5.26688516e-01 1.06004703e+00 -4.00883794e-01 -1.69738308e-01
3.83457482e-01 3.79661858e-01 9.71456587e-01 -8.28869402e-01
5.10980666e-01 1.16574395e+00 2.45220765e-01 7.87953198e-01
-9.68429863e-01 -5.65335870e-01 2.11561441e-01 1.14206374e-01
-1.16983116e+00 -4.34638351e-01 8.89867008e-01 -1.44385770e-01
1.11003733e+00 -7.79991224e-02 3.69116962e-01 1.20402145e+00
2.89677590e-01 1.26820993e+00 6.71647906e-01 -6.52473092e-01
3.48307341e-01 2.75266260e-01 6.01647124e-02 6.55522764e-01
1.18595324e-01 -2.73857474e-01 -3.93433750e-01 -1.68937951e-01
4.46772546e-01 -7.82808661e-02 -2.84803897e-01 -6.41617104e-02
-1.16139603e+00 8.22298944e-01 1.70534909e-01 3.64409238e-01
-5.02339065e-01 3.37515436e-02 4.79960382e-01 1.77825063e-01
4.92193252e-01 4.55532193e-01 -3.24224710e-01 -2.17675865e-01
-1.00985932e+00 2.39874139e-01 6.68195188e-01 1.19710553e+00
9.74721551e-01 5.58677495e-01 -7.70268559e-01 1.17863166e+00
-5.04905246e-02 3.49310398e-01 1.25903583e+00 -3.29738855e-01
7.36123800e-01 8.27667713e-01 -2.07957439e-02 -5.58338404e-01
-1.73714072e-01 -4.89348114e-01 -1.18163514e+00 -2.32558548e-01
-7.76434457e-03 -5.23264945e-01 -7.91376889e-01 1.99549103e+00
2.71556586e-01 3.82196605e-01 3.64995033e-01 4.00362670e-01
7.81886101e-01 1.10884833e+00 1.37984857e-01 -6.60575509e-01
1.09989369e+00 -1.18026435e+00 -8.17009866e-01 -4.13275301e-01
9.22274649e-01 -5.59980810e-01 1.31215596e+00 6.80796728e-02
-9.13511038e-01 -9.05718803e-01 -1.00907683e+00 -1.49743289e-01
-2.25381792e-01 1.87179998e-01 3.42767239e-01 3.26031655e-01
-5.52446246e-01 5.59803069e-01 -3.57751042e-01 -2.27166966e-01
8.60804170e-02 3.31573337e-02 6.91225007e-02 -1.04265921e-01
-1.65734088e+00 7.82712996e-01 9.16616321e-01 -1.09927543e-01
-6.85920835e-01 -4.71904606e-01 -8.98345768e-01 1.45791620e-01
6.91086888e-01 -9.44816411e-01 1.32777858e+00 -9.26451683e-01
-1.87091064e+00 2.03037828e-01 -3.25026810e-01 -6.10463560e-01
4.94453996e-01 -2.69021451e-01 -1.42787859e-01 -1.22950794e-02
9.78029072e-02 5.23831725e-01 1.00407588e+00 -9.70314026e-01
-5.19279063e-01 -6.36078697e-03 3.55883427e-02 4.63691622e-01
-4.89179999e-01 -3.21253598e-01 -1.37434766e-01 -9.31952000e-01
-4.39006060e-01 -5.37677944e-01 -2.89999634e-01 -6.05862439e-01
-3.94395709e-01 -6.82273269e-01 4.13581669e-01 -6.20929778e-01
1.51172960e+00 -1.74759078e+00 1.82915822e-01 -3.49438548e-01
-3.49032909e-01 6.00004137e-01 -3.67381841e-01 6.45347834e-01
6.10985681e-02 -9.15201157e-02 -4.30040896e-01 -4.94507879e-01
-1.19290769e-01 1.32444382e-01 -4.48127806e-01 -2.95798093e-01
3.22344720e-01 1.04619467e+00 -1.28717470e+00 -7.49367476e-01
1.29840836e-01 -5.07857539e-02 -4.16064620e-01 5.80383182e-01
-6.67078733e-01 1.46886572e-01 -5.47310174e-01 3.25154483e-01
4.99379814e-01 -1.02497995e-01 -5.78364450e-03 1.12305477e-01
-1.37718901e-01 2.74066091e-01 -1.12278998e+00 1.87672639e+00
-8.81821394e-01 -3.87487262e-02 -3.47956151e-01 -1.05186820e+00
1.18073964e+00 5.66767633e-01 1.39820725e-01 -4.83229339e-01
-2.43481640e-02 3.30359727e-01 1.21851504e-01 -6.65595353e-01
6.35957301e-01 -4.16660786e-01 -1.50791556e-01 8.66130650e-01
4.40088809e-01 -1.75350532e-01 4.51010495e-01 3.48832250e-01
8.63459826e-01 3.19480211e-01 7.58379877e-01 8.33575279e-02
7.84425974e-01 -1.90634862e-01 6.36966586e-01 8.94955754e-01
1.81164175e-01 4.64419425e-01 2.43469924e-01 1.55470654e-01
-8.32354188e-01 -7.71952033e-01 4.10735548e-01 1.28999853e+00
8.86801165e-03 -3.70832026e-01 -7.63646781e-01 -6.47605300e-01
-3.49286675e-01 1.38076591e+00 -3.00690979e-01 -6.62498295e-01
-6.20704412e-01 -7.29050517e-01 5.41611075e-01 5.07792056e-01
5.14449775e-01 -1.59340835e+00 -2.87242889e-01 6.82080090e-01
-4.89079863e-01 -8.36313963e-01 -6.96137190e-01 6.82262406e-02
-9.85014677e-01 -6.50681436e-01 -8.89758825e-01 -8.28746498e-01
4.75734711e-01 4.30963695e-01 7.78421700e-01 -1.02053531e-01
1.03438394e-02 6.15358166e-02 -7.34320998e-01 -4.63666052e-01
-7.66239345e-01 4.67332602e-01 1.60589593e-03 2.72086084e-01
1.47960275e-01 -5.95277607e-01 -1.18780509e-01 -2.00488925e-01
-1.05722082e+00 3.81198049e-01 9.52141464e-01 1.05977297e+00
4.11953092e-01 -9.63807553e-02 1.36778331e+00 -9.75986242e-01
1.07499397e+00 -5.84236562e-01 -1.51490375e-01 6.19552493e-01
-3.74140173e-01 3.61903220e-01 1.27921546e+00 -7.14481592e-01
-1.57910120e+00 -2.13956624e-01 -2.47501954e-01 -2.63766080e-01
-4.04996192e-03 5.99053860e-01 -5.41917562e-01 5.82907677e-01
6.67140722e-01 9.56319153e-01 7.07456023e-02 -3.94229382e-01
5.79389691e-01 8.49271536e-01 2.20435277e-01 -4.83536422e-01
7.23465145e-01 -1.24520138e-01 -4.66398150e-01 -9.94248569e-01
-8.17625225e-01 -3.83464664e-01 -4.05118465e-01 4.85929810e-02
6.05094910e-01 -8.67251277e-01 -1.26992956e-01 4.98043507e-01
-1.47962141e+00 -3.52246076e-01 -5.71573615e-01 4.14683521e-01
-8.08633626e-01 5.54805219e-01 -4.15146530e-01 -1.15010452e+00
-9.26519334e-01 -1.05565226e+00 1.03546107e+00 3.51684123e-01
-2.33002603e-01 -9.73922372e-01 8.77780542e-02 6.37597814e-02
3.97570372e-01 -3.95822749e-02 1.14747739e+00 -8.91131341e-01
-4.56641406e-01 -1.76262304e-01 1.29010081e-01 5.91735184e-01
3.62885147e-01 -3.11402142e-01 -7.77488708e-01 -1.17316127e-01
2.87086040e-01 -6.04314506e-01 1.05118620e+00 1.83815539e-01
9.48855817e-01 -4.98177528e-01 -1.71300650e-01 1.54842272e-01
1.27368224e+00 9.92419496e-02 6.78477526e-01 -2.01595481e-02
6.42988622e-01 6.32065475e-01 8.55123580e-01 4.33652282e-01
1.87024608e-01 5.36049724e-01 -2.36686561e-02 2.48329133e-01
-3.73097137e-02 -6.73766971e-01 7.44432569e-01 1.19132876e+00
-1.02690700e-02 -4.16652650e-01 -4.20174807e-01 4.20009375e-01
-2.01065946e+00 -1.18750691e+00 1.35494381e-01 2.40934086e+00
1.24875677e+00 1.91074252e-01 -1.11925453e-01 4.76769023e-02
8.58772397e-01 4.21500027e-01 -6.49899542e-01 -2.85690278e-01
-1.26698151e-01 2.83512741e-01 -1.08245201e-01 4.11683500e-01
-7.18178868e-01 9.95089173e-01 4.52508402e+00 1.60253942e+00
-9.02296424e-01 1.24386825e-01 2.94938773e-01 -6.80515692e-02
-3.99766326e-01 2.08476149e-02 -9.96000051e-01 6.68014109e-01
8.73481512e-01 -7.99240828e-01 2.41991684e-01 9.48408544e-01
4.79846656e-01 -4.68306504e-02 -9.61860418e-01 6.67436719e-01
2.04607353e-01 -1.10405493e+00 6.76646292e-01 -3.05314511e-01
8.40989888e-01 -3.93180758e-01 -3.58716965e-01 8.39782953e-01
8.41686204e-02 -6.32905066e-01 4.79190648e-01 5.73532343e-01
7.24685431e-01 -7.42728114e-01 7.51436949e-01 1.18120658e+00
-1.13840473e+00 -1.19328469e-01 -7.25711286e-01 -6.03661081e-03
4.34770256e-01 7.34139025e-01 -1.07558668e+00 8.76631796e-01
-2.78453737e-01 5.60651243e-01 -2.60143220e-01 8.51400495e-01
-4.35370773e-01 6.37456298e-01 -1.46106333e-02 -2.83171296e-01
2.80726522e-01 -1.58191159e-01 6.84065878e-01 1.21805215e+00
6.13996208e-01 -8.09581503e-02 3.20639968e-01 8.34017158e-01
-2.78939426e-01 5.96522331e-01 -6.80662394e-01 -1.23387188e-01
4.88176882e-01 1.34588993e+00 -2.27460727e-01 -7.37685919e-01
-2.59228915e-01 1.04411221e+00 4.12223190e-01 2.61494219e-01
-6.65585637e-01 -6.45101786e-01 2.50845581e-01 -1.19839022e-02
8.73850957e-02 2.40634661e-02 -7.86163211e-02 -1.28809071e+00
7.40616379e-05 -8.09648633e-01 1.95270389e-01 -6.87216103e-01
-1.28373075e+00 6.27078891e-01 -7.49649778e-02 -1.59609377e+00
-7.82380521e-01 -1.55161219e-02 -1.03756332e+00 1.03988469e+00
-1.53025210e+00 -1.13102376e+00 -1.28087133e-01 4.15317684e-01
1.21393430e+00 -3.78132105e-01 8.45459640e-01 -7.80951753e-02
-8.21103454e-01 7.16091394e-01 1.22845307e-01 -2.60853946e-01
8.83969903e-01 -1.19389391e+00 2.03862727e-01 7.86409616e-01
-1.54726401e-01 7.75946379e-01 4.91761833e-01 -7.77948499e-01
-1.16815197e+00 -1.41113663e+00 1.08904040e+00 8.40327293e-02
5.36519349e-01 -3.44757795e-01 -1.02439582e+00 5.09810805e-01
4.19819832e-01 -6.60314679e-01 6.51480436e-01 -2.32472315e-01
-5.62425181e-02 -1.85066685e-01 -7.76979089e-01 7.77291536e-01
8.02888274e-01 -2.29686484e-01 -1.00643766e+00 4.62192565e-01
1.04565275e+00 -6.81851506e-02 -3.27278912e-01 2.75887102e-01
2.67834038e-01 -6.64040446e-01 6.36380613e-01 -4.79458362e-01
6.78176582e-01 -1.57370910e-01 2.08178505e-01 -1.66607797e+00
-1.04712054e-01 -6.76816106e-01 -3.94344330e-01 1.49242437e+00
3.78699571e-01 -6.24375522e-01 3.80868316e-01 2.20987067e-01
-3.78108650e-01 -7.64776647e-01 -6.38442636e-01 -9.56366599e-01
-1.85508225e-02 -2.71853477e-01 5.40578723e-01 6.23263657e-01
1.02278985e-01 9.38217461e-01 -7.89609551e-01 -4.34500098e-01
4.35955286e-01 2.44235486e-01 9.06955242e-01 -1.01261532e+00
-6.31643057e-01 -4.37334329e-01 3.69677246e-01 -1.38500321e+00
2.91265398e-01 -1.13933933e+00 4.39802289e-01 -1.68942487e+00
2.58852988e-01 -2.64361888e-01 -1.73397675e-01 2.99536169e-01
-8.54088664e-01 -3.70968342e-01 2.06873730e-01 8.82528201e-02
-5.87680817e-01 1.08519161e+00 1.43497181e+00 -1.12431407e-01
-4.84009653e-01 2.38478258e-01 -8.52063537e-01 5.33043683e-01
8.39100182e-01 -3.44793528e-01 -7.38868058e-01 -2.14176193e-01
-1.03005864e-01 2.48664156e-01 -1.62924737e-01 -7.74889648e-01
3.08124423e-01 -2.14298457e-01 8.09125900e-02 -5.52675307e-01
2.17888623e-01 -2.45483458e-01 -2.40139857e-01 4.52921778e-01
-5.80112576e-01 -4.16542828e-01 1.63596887e-02 7.02050030e-01
-2.05957547e-01 -8.01258147e-01 6.96603835e-01 -4.97431159e-01
-6.68759406e-01 2.97058254e-01 -3.30077529e-01 1.38589948e-01
8.80139887e-01 -9.94751975e-02 -1.83699742e-01 -4.98997271e-01
-2.74862856e-01 3.78937453e-01 8.61191675e-02 4.85368818e-01
7.06806362e-01 -1.33303797e+00 -9.36160266e-01 9.17727500e-02
2.56475419e-01 1.75072938e-01 6.00484490e-01 6.84446871e-01
9.65182185e-02 3.86990637e-01 3.91145647e-02 -9.94861871e-02
-8.30058038e-01 6.49622500e-01 -2.79558226e-02 -8.32511067e-01
-3.69348884e-01 6.09322488e-01 3.08208823e-01 -4.22892898e-01
-2.71685515e-02 -8.08737576e-02 -3.66491973e-01 3.29558015e-01
8.14193249e-01 4.13727671e-01 -1.87748298e-01 -2.95799106e-01
1.38591915e-01 2.73952544e-01 -2.44521394e-01 -9.05035064e-03
1.06470466e+00 -7.47667700e-02 -8.55405703e-02 8.18408608e-01
1.02172267e+00 -7.41006434e-02 -9.08371270e-01 -4.68185574e-01
-1.12665959e-01 3.57947499e-02 -3.74950707e-01 -5.16735017e-01
-4.22098845e-01 1.15525186e+00 -9.71127823e-02 1.41531482e-01
1.04593277e+00 -3.22737962e-01 1.09667385e+00 5.06770551e-01
4.66108114e-01 -1.27439880e+00 4.03921336e-01 6.43616140e-01
9.88380492e-01 -1.00039899e+00 -2.68538415e-01 -1.51214242e-01
-8.06255460e-01 1.14915705e+00 9.12738860e-01 -8.66636932e-02
1.46238908e-01 -1.20601557e-01 -3.56752396e-01 5.09912968e-01
-1.02801704e+00 -3.04723233e-01 5.98590402e-03 4.74901289e-01
4.79754806e-01 -4.04387526e-02 -6.98815644e-01 9.28545415e-01
-1.24606967e-01 2.79369682e-01 5.63652515e-01 7.44704008e-01
-8.47591877e-01 -1.37292242e+00 -7.87706897e-02 6.27162158e-01
4.18219566e-02 -3.95447254e-01 -1.61590189e-01 3.88302565e-01
1.76478833e-01 8.86217773e-01 -1.84841067e-01 -3.10022324e-01
2.44917080e-01 5.46301961e-01 2.83764273e-01 -1.09259212e+00
-5.42091727e-01 2.31777150e-02 -4.80483100e-02 -8.25991258e-02
-1.48359090e-01 -4.80680078e-01 -1.34030700e+00 1.86851040e-01
-6.12645149e-01 2.82991916e-01 1.96618319e-01 1.15099549e+00
3.53935868e-01 7.05723345e-01 7.56192386e-01 -7.00877368e-01
-9.29328680e-01 -1.58982980e+00 -5.85258067e-01 2.39828557e-01
-3.72464880e-02 -4.86582369e-01 -2.53055573e-01 -5.02485856e-02] | [11.818882942199707, 8.911764144897461] |
38cae7ac-ff66-4399-ad71-8d1bca484b34 | integrated-face-analytics-networks-through | 1711.06055 | null | http://arxiv.org/abs/1711.06055v1 | http://arxiv.org/pdf/1711.06055v1.pdf | Integrated Face Analytics Networks through Cross-Dataset Hybrid Training | Face analytics benefits many multimedia applications. It consists of a number
of tasks, such as facial emotion recognition and face parsing, and most
existing approaches generally treat these tasks independently, which limits
their deployment in real scenarios. In this paper we propose an integrated Face
Analytics Network (iFAN), which is able to perform multiple tasks jointly for
face analytics with a novel carefully designed network architecture to fully
facilitate the informative interaction among different tasks. The proposed
integrated network explicitly models the interactions between tasks so that the
correlations between tasks can be fully exploited for performance boost. In
addition, to solve the bottleneck of the absence of datasets with comprehensive
training data for various tasks, we propose a novel cross-dataset hybrid
training strategy. It allows "plug-in and play" of multiple datasets annotated
for different tasks without the requirement of a fully labeled common dataset
for all the tasks. We experimentally show that the proposed iFAN achieves
state-of-the-art performance on multiple face analytics tasks using a single
integrated model. Specifically, iFAN achieves an overall F-score of 91.15% on
the Helen dataset for face parsing, a normalized mean error of 5.81% on the
MTFL dataset for facial landmark localization and an accuracy of 45.73% on the
BNU dataset for emotion recognition with a single model. | ['Jian Zhao', 'Terence Sim', 'Jianshu Li', 'Fang Zhao', 'Shengtao Xiao', 'Jiashi Feng', 'Shuicheng Yan', 'Jianan Li'] | 2017-11-16 | null | null | null | null | ['facial-emotion-recognition', 'face-parsing'] | ['computer-vision', 'computer-vision'] | [-3.16181383e-03 1.62464440e-01 6.26178160e-02 -7.64236033e-01
-6.65443778e-01 -4.16327983e-01 4.02466595e-01 -2.32325837e-01
-4.10774708e-01 2.06479028e-01 -2.32997254e-01 9.07532871e-02
-3.68366390e-02 -4.10477430e-01 -8.09586048e-01 -6.19716525e-01
1.01901628e-02 5.02769828e-01 5.75830089e-03 1.03576720e-01
-3.94550860e-01 6.42165840e-01 -1.82547534e+00 3.58394504e-01
6.60305202e-01 1.49113905e+00 2.03961745e-01 5.16903460e-01
-2.75871307e-01 8.10207248e-01 -4.86369133e-01 -8.12731266e-01
1.02449544e-01 6.94537833e-02 -5.92483699e-01 8.62919167e-02
5.27820289e-01 -3.96585852e-01 2.52289884e-02 8.84066761e-01
6.44484460e-01 -6.63919076e-02 3.09162319e-01 -1.76845086e+00
-7.16736764e-02 4.19682950e-01 -6.83045268e-01 -2.59429067e-01
5.55671491e-02 -8.41381401e-02 8.09120476e-01 -1.10093808e+00
5.02677262e-01 1.35690665e+00 5.51066577e-01 8.12793195e-01
-9.58867729e-01 -1.07286465e+00 2.85779238e-01 1.04922056e-02
-1.31396914e+00 -9.71336782e-01 7.91631937e-01 -3.55911076e-01
7.92160690e-01 -5.55456504e-02 2.37546891e-01 1.01935363e+00
-2.30800778e-01 7.64568388e-01 7.63003647e-01 -2.22068548e-01
5.00993729e-02 2.98190713e-02 1.70041516e-01 1.01403415e+00
8.84100199e-02 -3.74340087e-01 -8.68794858e-01 1.64658844e-01
3.88313621e-01 1.02373146e-01 3.36418524e-02 -1.65786222e-01
-6.99288845e-01 5.21320879e-01 2.08704755e-01 9.48933810e-02
-3.33553821e-01 1.94117904e-01 6.23074472e-01 8.43793303e-02
6.70126736e-01 -2.06946135e-01 -6.54688954e-01 6.28200248e-02
-8.45138073e-01 -8.25244114e-02 7.13042855e-01 8.93010378e-01
8.71900797e-01 5.91654256e-02 -6.87832162e-02 8.88276398e-01
5.13304770e-01 5.42712629e-01 3.40989158e-02 -1.00917816e+00
4.01535243e-01 7.39859581e-01 -2.58099139e-01 -8.07620823e-01
-7.13400602e-01 -3.29373509e-01 -8.32916319e-01 9.29255337e-02
4.99966234e-01 -3.91160011e-01 -7.68640220e-01 2.06700873e+00
6.71000779e-01 5.11252820e-01 3.58650764e-03 6.72750413e-01
1.13555872e+00 4.03291464e-01 3.28830510e-01 -2.15451717e-01
1.57641482e+00 -1.11559987e+00 -6.94879472e-01 -2.95755088e-01
6.49343371e-01 -7.15944409e-01 1.06468868e+00 4.28314626e-01
-9.48180437e-01 -6.28980100e-01 -7.23520577e-01 -1.61209375e-01
-2.70566463e-01 5.76089263e-01 8.48367214e-01 6.85794294e-01
-1.20585763e+00 1.78423092e-01 -8.58424366e-01 -3.14945489e-01
7.76445687e-01 7.65623450e-01 -7.46282756e-01 -1.43033132e-01
-5.59291184e-01 3.50576997e-01 1.64772376e-01 3.19387257e-01
-1.02595150e+00 -8.83002043e-01 -8.30337822e-01 3.17296892e-01
7.05982745e-01 -5.19465685e-01 1.20311570e+00 -1.07766998e+00
-1.52939856e+00 8.69603217e-01 -4.13843066e-01 -1.18813030e-01
3.39629024e-01 -4.44897443e-01 -3.93937618e-01 2.81035900e-01
-5.40522970e-02 1.00115156e+00 8.24803710e-01 -1.06610990e+00
-4.72130835e-01 -8.10896516e-01 8.73517096e-02 -9.24691707e-02
-6.66341484e-01 3.09113085e-01 -9.95523214e-01 -2.26736233e-01
-2.34135792e-01 -9.10138607e-01 1.68427274e-01 1.69619635e-01
-2.08527774e-01 -3.37984473e-01 1.03124845e+00 -5.87245286e-01
9.97123241e-01 -2.25120544e+00 4.13479470e-02 5.37573285e-02
3.90704960e-01 3.42127353e-01 -4.48232621e-01 -3.47676091e-02
-8.11737403e-02 3.31071019e-02 3.85423750e-02 -1.09660006e+00
6.38870373e-02 1.87247947e-01 1.45838231e-01 1.76193699e-01
3.04589659e-01 6.91124618e-01 -5.44205248e-01 -5.81998527e-01
-2.16066875e-02 8.23208809e-01 -6.49534881e-01 4.32752669e-01
-2.33692661e-01 6.12568557e-01 -3.51402879e-01 8.98287356e-01
9.07350242e-01 -2.36695617e-01 3.26708943e-01 -1.95808008e-01
2.64555365e-01 -1.83095515e-01 -1.19349301e+00 1.81491899e+00
-5.99901438e-01 3.44061345e-01 7.15534389e-01 -7.48648643e-01
9.70533192e-01 4.66404617e-01 6.33022428e-01 -6.58394635e-01
1.32629052e-01 1.23854056e-01 -1.00725867e-01 -4.51317966e-01
6.88647404e-02 8.15846920e-02 2.62031369e-02 3.62565398e-01
5.27280986e-01 5.45783341e-01 2.10568771e-01 9.92484465e-02
1.18605781e+00 7.33634606e-02 2.83390880e-02 -1.13006987e-01
7.05568910e-01 -5.51054418e-01 9.05488670e-01 3.64153445e-01
-1.89685360e-01 2.61820048e-01 8.10326278e-01 -5.18797994e-01
-5.48214614e-01 -7.56325483e-01 4.33624387e-02 1.69366384e+00
-2.68317252e-01 -6.49663389e-01 -1.09450424e+00 -1.05052042e+00
-2.89934389e-02 2.33761132e-01 -5.75408578e-01 1.56485006e-01
-3.95603776e-01 -7.00978696e-01 8.25763822e-01 4.74728912e-01
7.32174635e-01 -1.16037202e+00 -3.92732352e-01 -2.45152980e-01
-1.72280297e-01 -1.64008844e+00 -1.32694826e-01 -1.98998186e-03
-6.85438752e-01 -1.31998456e+00 -2.13171855e-01 -6.56351626e-01
6.20835662e-01 1.44794554e-01 1.15066051e+00 1.64475009e-01
-2.21288130e-01 4.60042983e-01 -2.20745414e-01 -3.70518684e-01
-7.67734647e-02 1.53637037e-01 7.39574581e-02 5.11056185e-01
3.47710431e-01 -5.02199531e-01 -3.19657594e-01 4.17411625e-01
-7.77177036e-01 1.61363615e-03 5.75343370e-01 8.60602081e-01
4.86793250e-01 -1.86121568e-01 8.65565360e-01 -1.00870979e+00
8.38474482e-02 -5.27155638e-01 -7.09174931e-01 4.44013059e-01
-3.33072960e-01 -2.23571569e-01 5.59605360e-01 -2.46017769e-01
-1.32593465e+00 4.35981244e-01 -2.13233978e-01 -7.40629315e-01
-3.23004246e-01 2.03982443e-01 -7.82766223e-01 -1.21556483e-01
1.61857113e-01 -7.40266666e-02 1.38418421e-01 -5.75466275e-01
2.61481166e-01 6.20634258e-01 4.27059174e-01 -7.52930760e-01
3.78645748e-01 4.44772422e-01 2.60660201e-01 -8.86298597e-01
-8.63057733e-01 -3.02222639e-01 -7.77717471e-01 -3.76346797e-01
9.48524117e-01 -1.09978712e+00 -1.25972378e+00 6.66339338e-01
-1.33733594e+00 -3.44908148e-01 1.41247049e-01 2.40944982e-01
-3.07613134e-01 1.67223796e-01 -4.89299595e-01 -8.62324059e-01
-4.75173324e-01 -1.33527219e+00 1.41426432e+00 3.89533997e-01
9.78605151e-02 -8.44089150e-01 -4.71969992e-01 6.04032218e-01
2.23324522e-01 7.44852126e-02 6.24113142e-01 -8.43562841e-01
-5.15604258e-01 4.12374688e-03 -4.48791355e-01 4.04274553e-01
-2.77320258e-02 9.54694673e-03 -1.56757426e+00 -2.48565793e-01
-1.26529440e-01 -6.04908645e-01 7.93133736e-01 1.69066399e-01
1.46293402e+00 -2.06866837e-03 -2.20051527e-01 7.74066627e-01
1.06039786e+00 1.66480988e-01 4.96221334e-01 -1.80084467e-01
8.55480433e-01 9.83193278e-01 5.41143954e-01 5.31010509e-01
4.52378601e-01 6.83351338e-01 6.32928193e-01 -7.60405064e-02
-1.30493362e-02 -5.04301190e-02 4.29684371e-01 4.64689374e-01
1.23871282e-01 -2.59226114e-01 -9.19074118e-01 2.28774920e-01
-1.87258708e+00 -6.10045254e-01 5.40853739e-02 1.91940916e+00
3.20464224e-01 -1.52873740e-01 1.18798681e-01 -8.96741375e-02
6.71251237e-01 1.31234840e-01 -4.90001619e-01 -2.05520272e-01
7.18175024e-02 3.93784255e-01 -4.29000333e-03 2.90588915e-01
-1.26009583e+00 1.16645753e+00 5.48892689e+00 8.58566046e-01
-1.10383892e+00 3.62466782e-01 9.09680307e-01 -2.35373572e-01
2.50414312e-01 -3.56511682e-01 -1.09536099e+00 3.50478947e-01
9.90743101e-01 2.98709124e-01 3.98488432e-01 9.67950106e-01
9.14593041e-02 5.05306348e-02 -1.13835537e+00 1.20366859e+00
1.32092968e-01 -1.03530478e+00 -2.83941301e-03 -1.24518317e-03
3.50782156e-01 -1.46588907e-01 1.39607370e-01 3.57061595e-01
-6.83249459e-02 -1.08401227e+00 4.55841124e-01 4.25742447e-01
8.89813781e-01 -8.90888751e-01 7.51711249e-01 3.25848758e-01
-1.37993264e+00 -1.87561333e-01 -1.69522300e-01 1.52517512e-01
-4.24215496e-02 7.50537813e-01 -6.60096586e-01 5.31652987e-01
8.99741650e-01 4.73129183e-01 -5.59552789e-01 5.27235508e-01
-5.55743165e-02 4.17853802e-01 -5.18359661e-01 3.66255075e-01
-6.05422519e-02 -1.05457464e-02 1.90236598e-01 1.11110032e+00
3.68708849e-01 -1.33198783e-01 3.07730079e-01 4.61488754e-01
-5.86072981e-01 3.00867349e-01 -6.17837965e-01 -4.34149355e-02
5.31993330e-01 1.85180926e+00 -6.27408922e-01 -1.57001704e-01
-5.97356498e-01 8.16840827e-01 5.38054705e-01 2.28052258e-01
-1.03156567e+00 -4.11493145e-02 9.22680438e-01 -3.74509953e-02
2.55830556e-01 2.61305887e-02 -1.20644517e-01 -9.22484636e-01
3.00773531e-01 -7.39722788e-01 4.81318027e-01 -6.35954857e-01
-1.12814045e+00 8.30679059e-01 -1.28339842e-01 -6.37417436e-01
-2.28698984e-01 -8.59047055e-01 -6.41656280e-01 6.28989041e-01
-1.55658388e+00 -1.63179433e+00 -6.76263630e-01 8.43169630e-01
4.21938568e-01 -4.68519479e-01 7.31347978e-01 8.03055286e-01
-1.18354702e+00 9.25572157e-01 -3.92881542e-01 2.89877206e-01
1.00148201e+00 -9.69784081e-01 -3.69909108e-02 8.28337789e-01
1.29233882e-01 3.66400033e-01 8.07114840e-02 -5.02506793e-01
-1.45358074e+00 -1.32354736e+00 8.52109492e-01 -1.17565155e-01
2.82742053e-01 -6.70851529e-01 -7.98451245e-01 8.44830334e-01
-1.39655117e-02 4.51411277e-01 8.15518677e-01 4.38702255e-01
-5.66957355e-01 -6.59125865e-01 -1.28892684e+00 2.24772766e-01
1.20210552e+00 -5.08784771e-01 1.99997663e-01 4.16224211e-01
6.91123843e-01 -2.75986820e-01 -7.62114465e-01 5.30545056e-01
5.21461546e-01 -1.17414820e+00 9.32891011e-01 -5.43769419e-01
4.22721148e-01 -2.77241226e-02 -2.96840966e-01 -8.54901910e-01
1.08093202e-01 -6.49918377e-01 -2.68870205e-01 1.77115226e+00
1.67509452e-01 -6.40487015e-01 8.99747908e-01 6.01666510e-01
-1.73574790e-01 -8.17875981e-01 -9.67048109e-01 -4.96897876e-01
-5.51147401e-01 -6.20663166e-01 7.68454313e-01 7.93401897e-01
-4.84479040e-01 5.24686575e-01 -3.93998474e-01 3.41260791e-01
7.20478833e-01 -1.90739587e-01 1.00401318e+00 -1.48153341e+00
-1.59522325e-01 -2.47519493e-01 -2.96380341e-01 -7.63235092e-01
6.83722973e-01 -8.61068428e-01 -2.09572479e-01 -9.97344851e-01
2.45921835e-01 -4.56642985e-01 -1.60550013e-01 8.17220926e-01
1.73018500e-02 2.37777799e-01 3.48878622e-01 -1.48269042e-01
-9.23296332e-01 5.98727703e-01 8.11088562e-01 2.27398306e-01
-9.59571972e-02 -1.75402850e-01 -7.26495802e-01 1.06588507e+00
6.33559883e-01 -3.78458858e-01 -6.11074686e-01 -6.45657122e-01
-1.42295718e-01 9.07216147e-02 3.23571920e-01 -1.14606714e+00
3.33515227e-01 1.66200042e-01 4.95958388e-01 -4.01086211e-01
6.09781384e-01 -9.79345739e-01 1.20084323e-01 -3.79221477e-02
1.93864536e-02 1.03174105e-01 4.41281348e-01 4.57158983e-01
-2.74430066e-01 8.25929455e-03 7.85235405e-01 3.08801711e-01
-8.16033542e-01 5.11422396e-01 2.76282549e-01 -2.71862298e-02
1.23406076e+00 1.61394268e-01 -3.85716170e-01 -2.27725506e-01
-8.66884530e-01 4.30020988e-01 1.57314345e-01 3.58270049e-01
4.12629843e-01 -1.14930689e+00 -5.47417700e-01 3.83232445e-01
-8.07424728e-03 2.10496709e-01 6.64747238e-01 8.47086251e-01
-3.38602036e-01 2.19900995e-01 -3.28473449e-01 -7.00352013e-01
-1.77313626e+00 3.38289112e-01 3.38643223e-01 -5.06836116e-01
-2.67168552e-01 9.04610276e-01 5.59557080e-01 -6.25727057e-01
3.92910093e-01 1.72387406e-01 -3.21398407e-01 4.10073906e-01
6.39846027e-01 4.12329733e-01 2.81395614e-01 -7.18116105e-01
-5.21471143e-01 5.51129162e-01 -5.31013981e-02 1.39713824e-01
1.45780432e+00 -4.18112287e-03 -3.15376610e-01 9.39748064e-02
1.17762077e+00 -1.35170847e-01 -1.25694239e+00 -2.39755869e-01
-9.01641548e-02 -2.23669738e-01 3.83961648e-02 -8.99429500e-01
-1.63801205e+00 1.23794854e+00 5.97238004e-01 -3.06394041e-01
1.67746627e+00 6.59813955e-02 6.16430938e-01 3.90060663e-01
1.67191684e-01 -8.94159198e-01 2.51321524e-01 4.53929573e-01
7.12613761e-01 -1.29702008e+00 -2.74411678e-01 -8.71583760e-01
-6.43099666e-01 1.12358177e+00 9.79578316e-01 3.29510838e-01
7.26279974e-01 4.92239833e-01 -1.25378482e-02 -2.84680665e-01
-9.17912006e-01 -2.29767442e-01 1.99620724e-01 3.64472359e-01
3.90837342e-01 -6.74255118e-02 2.28605062e-01 1.10311234e+00
2.31311873e-01 1.18340060e-01 -1.17440604e-01 6.41655445e-01
-1.07323609e-01 -1.13580739e+00 -2.02488720e-01 3.00374001e-01
-6.36779189e-01 1.10071994e-01 -2.79214650e-01 8.77636671e-01
3.13143134e-01 9.16558564e-01 2.85988927e-01 -5.00323117e-01
1.98288828e-01 5.93923151e-01 3.44095707e-01 -5.32227516e-01
-7.26738393e-01 2.73141980e-01 1.63864702e-01 -9.08882499e-01
-4.35516685e-01 -4.47941542e-01 -1.26038218e+00 -4.47264373e-01
1.12148732e-01 -6.69657513e-02 7.02569664e-01 9.72693026e-01
8.60916555e-01 5.91262698e-01 5.63240051e-01 -8.34848583e-01
-1.61093190e-01 -8.81935716e-01 -5.80056846e-01 2.09648266e-01
-3.44325863e-02 -9.23773706e-01 -2.41468716e-02 -3.62924114e-02] | [13.455460548400879, 0.7850288152694702] |
47790e2d-a071-42a5-97b7-58d7263d7d92 | a-hierarchical-game-theoretic-decision-making | 2303.16641 | null | https://arxiv.org/abs/2303.16641v1 | https://arxiv.org/pdf/2303.16641v1.pdf | A Hierarchical Game-Theoretic Decision-Making for Cooperative Multi-Agent Systems Under the Presence of Adversarial Agents | Underlying relationships among Multi-Agent Systems (MAS) in hazardous scenarios can be represented as Game-theoretic models. This paper proposes a new hierarchical network-based model called Game-theoretic Utility Tree (GUT), which decomposes high-level strategies into executable low-level actions for cooperative MAS decisions. It combines with a new payoff measure based on agent needs for real-time strategy games. We present an Explore game domain, where we measure the performance of MAS achieving tasks from the perspective of balancing the success probability and system costs. We evaluate the GUT approach against state-of-the-art methods that greedily rely on rewards of the composite actions. Conclusive results on extensive numerical simulations indicate that GUT can organize more complex relationships among MAS cooperation, helping the group achieve challenging tasks with lower costs and higher winning rates. Furthermore, we demonstrated the applicability of the GUT using the simulator-hardware testbed - Robotarium. The performances verified the effectiveness of the GUT in the real robot application and validated that the GUT could effectively organize MAS cooperation strategies, helping the group with fewer advantages achieve higher performance. | ['Ramviyas Parasuraman', 'Qin Yang'] | 2023-03-28 | null | null | null | null | ['real-time-strategy-games'] | ['playing-games'] | [-4.44846630e-01 6.25523806e-01 3.79155606e-01 3.20609897e-01
-1.48597792e-01 -3.38602781e-01 5.97839653e-01 6.65108413e-02
-5.26005030e-01 1.02265394e+00 -2.79683560e-01 -6.25729486e-02
-9.71458435e-01 -9.27951932e-01 3.00384723e-02 -6.55453920e-01
-8.42592955e-01 7.02908754e-01 5.76290190e-01 -1.09706628e+00
3.20215851e-01 8.83948430e-02 -1.79944324e+00 -2.18665615e-01
9.37808096e-01 7.63003826e-01 7.57281721e-01 7.15854466e-01
2.44343668e-01 1.25579667e+00 -9.61507201e-01 -2.20968723e-02
4.44734067e-01 -3.74228239e-01 -7.55544305e-01 -7.77301788e-02
-1.22927213e+00 -2.79344350e-01 6.09654374e-02 7.56341875e-01
5.41995823e-01 1.42302558e-01 6.89880550e-01 -2.15118551e+00
1.06470957e-01 1.06916058e+00 -4.57727253e-01 -1.59885794e-01
3.98014009e-01 1.35777965e-01 8.45918357e-01 -6.09871186e-02
6.37602210e-01 1.31197584e+00 2.26363525e-01 3.24119866e-01
-8.55118573e-01 -3.48698229e-01 -3.66625823e-02 3.12260509e-01
-1.26160991e+00 2.91357636e-01 3.93035620e-01 -2.75540859e-01
1.24301159e+00 3.85191262e-01 8.44665945e-01 2.49116883e-01
6.97542906e-01 4.86146837e-01 1.05827105e+00 -4.21053022e-01
6.02394402e-01 -1.44474670e-01 -1.87668160e-01 3.39679778e-01
4.02896196e-01 1.90313473e-01 -1.44308344e-01 -3.58046174e-01
8.49057794e-01 -4.61854488e-01 2.01184914e-01 -7.62457013e-01
-1.18312085e+00 7.43823051e-01 3.73944640e-01 5.24843514e-01
-9.39189732e-01 2.74682313e-01 4.53081131e-01 7.51778066e-01
3.64160314e-02 9.46653724e-01 -2.56688207e-01 -3.20338160e-01
-1.55993611e-01 5.08847833e-01 1.15659153e+00 8.92515540e-01
6.11850381e-01 1.54858455e-01 6.04647622e-02 3.91806155e-01
3.72816563e-01 -4.03274186e-02 1.32564366e-01 -1.37719417e+00
-6.47422373e-02 7.79963851e-01 3.68994683e-01 -8.34125459e-01
-8.51370096e-01 -3.02859634e-01 -6.61676228e-01 9.33378458e-01
-1.86365657e-02 -5.39488256e-01 1.43863307e-02 1.64171171e+00
1.73155785e-01 -2.12165579e-01 4.87971216e-01 7.34822512e-01
4.64260340e-01 5.69559574e-01 2.92709027e-03 -5.88608205e-01
1.13115072e+00 -8.74548733e-01 -8.59637797e-01 1.61521494e-01
7.10295677e-01 -1.94981456e-01 6.72404826e-01 5.97267628e-01
-1.37652111e+00 -2.76891768e-01 -1.09178567e+00 9.63776886e-01
-2.43654475e-01 -3.30974996e-01 6.94517255e-01 8.53684843e-01
-1.81997931e+00 4.25532997e-01 -5.10032594e-01 -5.77124238e-01
-4.06121284e-01 7.77243853e-01 -5.85053302e-03 5.39181173e-01
-1.13078272e+00 1.18650484e+00 5.20665109e-01 -3.28151047e-01
-1.24739385e+00 -1.68375015e-01 -4.45513427e-01 3.73480201e-01
9.33158040e-01 -7.46055186e-01 1.22931349e+00 -5.87737560e-01
-1.83718014e+00 4.34058726e-01 9.96187508e-01 -6.11321747e-01
4.76729870e-01 2.86469787e-01 1.56759337e-01 1.75282493e-01
2.75370985e-01 5.90850294e-01 1.78032458e-01 -1.51846600e+00
-1.05146515e+00 1.68445870e-01 8.29642832e-01 7.36239135e-01
-1.08061559e-01 9.01558846e-02 3.79994780e-01 4.20727804e-02
-5.75144768e-01 -8.66783023e-01 -9.07537162e-01 -8.48335743e-01
6.63727820e-02 -3.56395096e-01 3.92549753e-01 1.30763963e-01
9.64411795e-01 -1.82079387e+00 7.36774921e-01 2.07907557e-01
6.59263730e-01 -1.37750119e-01 -1.73780754e-01 1.04482186e+00
4.39716071e-01 8.12765863e-03 2.90525824e-01 1.66567340e-01
3.47090155e-01 2.32968003e-01 5.40314496e-01 3.94473448e-02
-1.98654771e-01 4.58814591e-01 -1.07984066e+00 -2.80420244e-01
2.42700055e-01 -3.47074717e-01 -4.10198629e-01 5.29246867e-01
-2.82778740e-02 -5.11286035e-03 -7.36329079e-01 2.48258039e-01
3.99447709e-01 2.44753927e-01 7.37543225e-01 4.43596691e-01
-4.76253301e-01 -1.95318326e-01 -1.34645188e+00 1.09684825e+00
-3.50883782e-01 7.77472779e-02 8.58714461e-01 -1.05200589e+00
1.06313205e+00 3.47388119e-01 9.42855895e-01 -4.68759507e-01
6.22917652e-01 2.00232849e-01 5.59160173e-01 -2.14163437e-01
7.93903649e-01 2.92225569e-01 -5.97952247e-01 5.65156758e-01
2.43510440e-01 -6.88152134e-01 4.41041052e-01 4.50067550e-01
1.43870187e+00 1.63431058e-03 8.81529629e-01 -8.27735305e-01
5.62913597e-01 4.04130459e-01 1.82482198e-01 1.09715259e+00
-5.23870468e-01 -2.99510986e-01 1.04165852e+00 -7.99996778e-03
-6.17167830e-01 -8.16980958e-01 7.82267630e-01 1.24700928e+00
7.43276656e-01 -4.94364142e-01 -9.62864101e-01 -3.06075990e-01
-2.22150177e-01 6.83869779e-01 -4.86321270e-01 -6.82530850e-02
-1.44307464e-01 -8.10007691e-01 2.96723187e-01 -1.50613472e-01
4.99152303e-01 -1.32000065e+00 -1.40534663e+00 4.76963311e-01
-4.60047536e-02 -7.93948114e-01 1.55515626e-01 4.02320206e-01
-4.35800582e-01 -1.14112699e+00 -3.15715939e-01 -7.37537324e-01
2.72043377e-01 5.83672523e-01 1.04502010e+00 3.84905577e-01
-1.33925989e-01 8.99032593e-01 -8.85022402e-01 -6.48154378e-01
-6.30969405e-01 1.17952496e-01 3.79364073e-01 -5.30927420e-01
-9.86941159e-02 -7.40431309e-01 -3.97230357e-01 6.51001692e-01
-6.34115160e-01 1.30296946e-01 6.33291423e-01 6.86805010e-01
-1.39527604e-01 6.41457617e-01 1.03405786e+00 -2.25470394e-01
1.72502375e+00 -6.21362627e-01 -6.00418925e-01 2.29610726e-01
-4.54308391e-01 -1.67427659e-01 4.30022150e-01 -1.75728157e-01
-7.91643023e-01 -2.45674163e-01 3.34929734e-01 9.31538120e-02
1.15149982e-01 6.35849535e-01 -1.16523951e-01 -2.37769216e-01
6.02042317e-01 1.43201568e-03 4.58477855e-01 3.90246630e-01
3.02875727e-01 5.92976332e-01 2.59387810e-02 -8.35602939e-01
3.77530664e-01 -1.65145591e-01 2.53741562e-01 -5.63375771e-01
5.64765871e-01 -2.16250494e-01 -2.46425346e-01 -1.04243791e+00
6.69578016e-01 -6.90567553e-01 -1.50390124e+00 6.48875356e-01
-8.80517304e-01 -6.14174604e-01 -2.11693406e-01 2.61217773e-01
-1.10456812e+00 2.24556878e-01 -6.03895783e-01 -1.39793813e+00
-1.34057343e-01 -1.39862335e+00 4.19696778e-01 3.51261407e-01
4.13080566e-02 -6.30274594e-01 -2.46082265e-02 -1.58288136e-01
7.75109291e-01 2.16057733e-01 5.50897360e-01 -5.23075223e-01
-6.98742509e-01 1.03713244e-01 1.94550440e-01 -1.18174799e-01
5.38672172e-02 -8.71465877e-02 -1.70664564e-01 -2.65572280e-01
1.58549294e-01 -4.89359915e-01 -2.68477082e-01 4.43033576e-01
1.25175327e-01 -1.36775300e-02 -2.98013061e-01 -3.12206656e-01
1.45415783e+00 7.59931684e-01 8.38798106e-01 8.57151628e-01
-1.21408358e-01 1.03115809e+00 1.44250834e+00 8.39917064e-01
5.42124927e-01 9.21503544e-01 1.35489130e+00 1.16522960e-01
4.38564926e-01 3.64982545e-01 6.35592222e-01 7.71759033e-01
-8.14449847e-01 -2.85426736e-01 -7.73777425e-01 4.20068830e-01
-2.45225000e+00 -8.69024396e-01 -2.10976273e-01 1.84870803e+00
3.02047461e-01 4.56244379e-01 7.89771199e-01 1.06358275e-01
6.61921561e-01 -2.64169335e-01 -1.34920850e-01 -5.04108727e-01
1.66228250e-01 -2.49039799e-01 4.87240404e-01 5.68077624e-01
-5.54693758e-01 9.72649932e-01 6.68485880e+00 1.12987912e+00
-3.02457839e-01 1.70191735e-01 3.39385092e-01 -5.04526049e-02
1.51048884e-01 -1.28089160e-01 -2.23196477e-01 6.54139668e-02
8.86293888e-01 -1.21127939e+00 6.76831305e-01 1.01955724e+00
5.86827159e-01 -6.10288739e-01 -4.74618882e-01 5.69879830e-01
-4.58317697e-01 -1.17467129e+00 -3.59482229e-01 3.97883743e-01
3.16492528e-01 -9.09933150e-02 -4.32112187e-01 6.25687599e-01
1.17303860e+00 -6.67356133e-01 8.50377440e-01 1.87742233e-01
3.15971434e-01 -9.93721128e-01 1.23200655e+00 7.02591419e-01
-1.53249073e+00 -3.61603558e-01 -3.78961474e-01 -9.83668268e-01
3.27138454e-01 -1.97225362e-01 -1.05794275e+00 1.25013578e+00
4.86966789e-01 6.53598085e-02 -2.90895939e-01 9.93538857e-01
1.21444650e-01 -9.35760289e-02 -1.73164338e-01 -8.31145048e-01
5.19257963e-01 -7.12432921e-01 6.62209392e-01 6.07341468e-01
4.35034990e-01 4.48736399e-01 4.63117093e-01 6.59709990e-01
7.73110807e-01 2.02582851e-01 -8.36626410e-01 2.48942465e-01
6.62119508e-01 1.58802629e+00 -1.00137746e+00 7.48067871e-02
-3.86383496e-02 3.72289956e-01 1.12946250e-01 -1.12418123e-01
-9.47639763e-01 -3.64313513e-01 6.38625145e-01 -2.10595325e-01
-3.03981841e-01 -2.13851005e-01 -1.47471458e-01 -4.12187546e-01
-4.91882741e-01 -8.69284213e-01 1.32406175e-01 -1.02170753e+00
-9.78477478e-01 1.05820465e+00 3.74335676e-01 -1.56839013e+00
-7.61798859e-01 -5.03656089e-01 -6.17130101e-01 5.01769841e-01
-7.39830494e-01 -8.84150267e-01 -3.06006730e-01 4.93745536e-01
3.76074016e-01 -7.18828261e-01 7.41826892e-01 -1.97683260e-01
-3.54386270e-01 -2.54239375e-03 -2.73002777e-02 -6.93045199e-01
1.05195008e-01 -1.37271392e+00 -1.71960726e-01 4.39326823e-01
-8.46370101e-01 -4.60190587e-02 9.43243325e-01 -4.83073562e-01
-1.17040503e+00 -3.48763764e-01 5.65145090e-02 8.21947530e-02
9.13792968e-01 -2.40455940e-01 -1.25747532e-01 9.34020653e-02
7.68624306e-01 -8.88662219e-01 2.36436933e-01 -9.73582417e-02
7.46078014e-01 -1.57799292e-03 -1.32304418e+00 8.39225471e-01
1.09998429e+00 2.42790326e-01 -4.22456294e-01 7.76597485e-02
9.48411584e-01 -2.36711532e-01 -6.53563976e-01 1.66350797e-01
1.95403114e-01 -1.31280684e+00 7.55320132e-01 -2.47887939e-01
3.58272702e-01 -3.54642421e-01 -1.12694189e-01 -1.87576878e+00
-6.61194205e-01 -7.68421650e-01 4.21525866e-01 8.51187050e-01
1.23204030e-01 -7.36765087e-01 5.18663466e-01 3.51644039e-01
-1.39942572e-01 -5.01660824e-01 -9.81973469e-01 -1.02546430e+00
-1.17544383e-01 9.42183807e-02 6.31184995e-01 6.96192503e-01
9.02720928e-01 3.99264932e-01 -4.09846663e-01 4.32839617e-02
6.66138649e-01 -5.62551260e-01 9.07300234e-01 -1.23921645e+00
-4.92375404e-01 -7.49307990e-01 -7.16649473e-01 -2.28524610e-01
2.38001458e-02 -2.46789753e-01 2.05673248e-01 -1.83706772e+00
-4.39208932e-02 -5.41580379e-01 -2.05961779e-01 2.83058345e-01
-4.25992720e-02 -4.38155293e-01 6.49305880e-01 1.32858232e-01
-1.04459786e+00 5.41619301e-01 1.09353864e+00 1.20195754e-01
-4.48054671e-01 2.09726151e-02 -6.70343578e-01 5.29091775e-01
1.00503135e+00 -2.32206210e-01 -8.03713202e-01 2.69017547e-01
2.24597126e-01 9.53083932e-01 -7.04445243e-02 -1.22707462e+00
3.76791656e-01 -6.71804667e-01 -1.00615239e+00 -2.34099939e-01
1.22365996e-01 -1.00855660e+00 5.51333964e-01 1.17280829e+00
-4.25554104e-02 3.07466209e-01 2.64443737e-02 5.19145906e-01
-2.21712455e-01 -4.83516484e-01 2.61011422e-01 -1.72019988e-01
-8.23206723e-01 -4.29222584e-01 -1.11001551e+00 -4.80725765e-01
1.90859032e+00 -2.96400994e-01 -4.25934613e-01 -8.98575425e-01
-6.97433531e-01 7.91057110e-01 3.62265646e-01 1.69690534e-01
3.12870920e-01 -9.13110077e-01 -8.33412707e-01 -5.09797037e-01
-1.61793455e-01 -5.64355969e-01 3.26764494e-01 7.20172286e-01
-8.29385936e-01 1.61223263e-01 -1.18173218e+00 -2.60922134e-01
-1.39773917e+00 5.32378495e-01 6.67247832e-01 -8.71386111e-01
-1.49135590e-01 4.05255884e-01 4.09907937e-01 -5.25690019e-01
1.03815809e-01 5.12746982e-02 -6.52704835e-01 -2.88862400e-02
2.27663741e-01 6.04640484e-01 -3.07065278e-01 -1.15776636e-01
-3.90506297e-01 -3.14080343e-02 4.04589146e-01 -5.99317491e-01
1.34194338e+00 -4.73367512e-01 -4.16981339e-01 1.07256837e-01
-1.87906310e-01 -3.39044869e-01 -9.90922034e-01 3.45651716e-01
3.07628326e-02 -1.85405731e-01 -2.24973321e-01 -7.49765694e-01
-8.26398969e-01 3.90341878e-02 2.94644088e-01 1.22020686e+00
1.30118835e+00 -3.14540088e-01 -1.35598585e-01 4.24925685e-01
1.51293302e+00 -1.16018403e+00 3.67234677e-01 5.53365886e-01
9.06088114e-01 -6.34913862e-01 -2.98268378e-01 -8.21429014e-01
-1.05465508e+00 1.01895988e+00 9.58109975e-01 -3.88942897e-01
4.70529318e-01 6.57182992e-01 -1.73779190e-01 -4.61577117e-01
-1.13891566e+00 -5.94644725e-01 -7.44213283e-01 1.25732255e+00
-1.47049397e-01 5.19591510e-01 -9.14732933e-01 1.02992857e+00
-2.48955488e-01 6.56910092e-02 1.46782839e+00 1.05371010e+00
-9.21595097e-01 -1.26016378e+00 -4.42781597e-01 4.57329303e-01
4.65316586e-02 3.17665935e-01 -4.66711521e-01 1.12609124e+00
-2.71549188e-02 1.58682251e+00 -2.35106587e-01 -7.51551807e-01
5.96883476e-01 -5.43947399e-01 3.27972680e-01 -5.49907565e-01
-1.05254757e+00 9.89924520e-02 6.05098426e-01 -5.55903494e-01
-6.44622803e-01 -5.34985423e-01 -1.14779234e+00 -5.86044908e-01
-1.76699743e-01 6.18632972e-01 2.65639961e-01 7.10158050e-01
3.14239502e-01 1.09709227e+00 8.05837810e-01 -1.17382824e+00
-6.73172832e-01 -9.58066821e-01 -1.21728694e+00 -1.37865871e-01
-4.07525986e-01 -1.10947776e+00 -3.09287012e-01 -5.73443830e-01] | [3.7276811599731445, 1.8770866394042969] |
b3edeb0a-ce24-4926-85a9-664d36951bac | learning-geometry-disentangled-representation | 2012.10921 | null | https://arxiv.org/abs/2012.10921v3 | https://arxiv.org/pdf/2012.10921v3.pdf | Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloud | In 2D image processing, some attempts decompose images into high and low frequency components for describing edge and smooth parts respectively. Similarly, the contour and flat area of 3D objects, such as the boundary and seat area of a chair, describe different but also complementary geometries. However, such investigation is lost in previous deep networks that understand point clouds by directly treating all points or local patches equally. To solve this problem, we propose Geometry-Disentangled Attention Network (GDANet). GDANet introduces Geometry-Disentangle Module to dynamically disentangle point clouds into the contour and flat part of 3D objects, respectively denoted by sharp and gentle variation components. Then GDANet exploits Sharp-Gentle Complementary Attention Module that regards the features from sharp and gentle variation components as two holistic representations, and pays different attentions to them while fusing them respectively with original point cloud features. In this way, our method captures and refines the holistic and complementary 3D geometric semantics from two distinct disentangled components to supplement the local information. Extensive experiments on 3D object classification and segmentation benchmarks demonstrate that GDANet achieves the state-of-the-arts with fewer parameters. Code is released on https://github.com/mutianxu/GDANet. | ['Yu Qiao', 'Xiaojuan Qi', 'Mingye Xu', 'Zhipeng Zhou', 'Junhao Zhang', 'Mutian Xu'] | 2020-12-20 | null | null | null | null | ['3d-object-classification', '3d-part-segmentation'] | ['computer-vision', 'computer-vision'] | [-1.77212745e-01 -6.32360280e-02 -1.78194568e-01 -4.01345998e-01
-6.09230638e-01 -7.99125493e-01 5.54605722e-01 7.90402368e-02
1.25368059e-01 1.90802976e-01 8.00588951e-02 -3.74429189e-02
-2.45020226e-01 -7.75362611e-01 -7.87094891e-01 -8.36804032e-01
1.74062833e-01 5.06714046e-01 8.13017860e-02 -3.10171545e-01
1.67666748e-01 6.80325687e-01 -1.44414651e+00 2.47896269e-01
8.45024049e-01 1.08642960e+00 -1.40635222e-02 3.48925978e-01
-2.33201116e-01 -3.32932472e-02 -3.14737171e-01 -2.02189147e-01
5.55159688e-01 6.98963180e-02 -4.86821651e-01 1.89677119e-01
8.46456170e-01 -3.81144255e-01 -2.07053289e-01 1.23383319e+00
2.07175508e-01 -9.06170458e-02 7.35800803e-01 -1.32035327e+00
-1.29794610e+00 2.19216883e-01 -1.15420139e+00 1.43355310e-01
2.86769271e-01 2.18135595e-01 1.11363637e+00 -1.02553749e+00
1.55598968e-01 1.41426849e+00 5.94819665e-01 1.67405665e-01
-1.25133455e+00 -8.02885473e-01 6.16057992e-01 -8.17328170e-02
-1.47451246e+00 -2.77744737e-02 1.28112519e+00 -4.22510386e-01
6.48432255e-01 5.10709465e-01 7.60337353e-01 9.09268200e-01
-5.28168231e-02 9.34210181e-01 9.67119217e-01 9.18477103e-02
-2.24335581e-01 -1.88072100e-02 4.53222841e-01 6.46855831e-01
5.98243058e-01 -1.65403225e-02 -1.37018070e-01 9.68440995e-02
1.16766846e+00 4.32405829e-01 -4.20602769e-01 -6.92312121e-01
-1.24446464e+00 5.84484994e-01 7.44957387e-01 1.42193124e-01
-2.90847957e-01 -6.04421683e-02 5.82533255e-02 -3.92957553e-02
6.68487787e-01 2.38747865e-01 -5.30695498e-01 2.43215829e-01
-5.30386329e-01 4.72317874e-01 3.77036184e-01 1.34715414e+00
1.04259825e+00 -1.59255490e-01 -5.42695308e-03 5.05048394e-01
4.86255258e-01 5.66880226e-01 2.42678523e-01 -7.75470436e-01
4.46996778e-01 1.18697882e+00 -1.28539475e-02 -1.33634543e+00
-4.50906068e-01 -6.86821342e-01 -1.00079155e+00 3.02210927e-01
2.26106688e-01 2.09696501e-01 -1.22170663e+00 1.59566021e+00
5.18477380e-01 4.31208253e-01 -3.62568170e-01 1.32738113e+00
1.18917596e+00 4.44282353e-01 -3.06995362e-01 2.85964489e-01
1.74370217e+00 -9.16547120e-01 -5.12989402e-01 -3.00026536e-01
8.70685056e-02 -5.41018605e-01 1.09665215e+00 2.15782240e-01
-1.23718774e+00 -6.89844251e-01 -1.13959372e+00 -6.34638071e-01
-4.42933947e-01 1.31722704e-01 5.78348160e-01 3.82329971e-01
-7.18610466e-01 3.96629393e-01 -8.89256954e-01 1.76694337e-02
8.68503511e-01 3.20971996e-01 -4.32302982e-01 -1.26287225e-04
-8.84481251e-01 5.45623183e-01 -1.97425764e-02 3.36454123e-01
-4.01500136e-01 -1.06472623e+00 -9.51035440e-01 2.12762848e-01
3.31553400e-01 -9.81154561e-01 1.07336068e+00 -6.43490374e-01
-1.19677150e+00 1.12047589e+00 -2.60677952e-02 1.41825035e-01
3.61528248e-01 -3.33290994e-01 -2.15220340e-02 1.52498826e-01
1.46534145e-01 4.61379766e-01 8.27067494e-01 -1.73555660e+00
-2.45486453e-01 -1.01396823e+00 2.31349349e-01 4.71737117e-01
1.92839041e-01 -5.52620828e-01 -5.52962124e-01 -6.63643479e-01
6.88513577e-01 -5.97593188e-01 7.30677098e-02 3.28942597e-01
-6.31876349e-01 -1.80585161e-01 8.86691272e-01 -3.50243121e-01
7.76437104e-01 -2.23523831e+00 3.58545214e-01 4.91434485e-02
8.64786386e-01 -3.12362779e-02 -1.77133754e-01 -1.63199268e-02
-3.14815015e-01 3.57367992e-01 -2.21380726e-01 -6.23745263e-01
2.54095167e-01 2.35635623e-01 -3.00153166e-01 6.30294323e-01
5.99571109e-01 1.05677986e+00 -9.79826689e-01 -2.42058009e-01
3.02200675e-01 6.38197124e-01 -5.42553842e-01 -1.14813082e-01
-4.12175804e-02 4.86740351e-01 -8.07245314e-01 8.35399032e-01
1.37405515e+00 -2.25290939e-01 -3.88038218e-01 -5.95026553e-01
-1.57049194e-01 1.20094255e-01 -1.07998824e+00 2.04735065e+00
-2.53902614e-01 3.12035233e-01 2.49755710e-01 -1.03846490e+00
1.15298641e+00 4.22028787e-02 4.68328059e-01 -5.26707649e-01
3.47837150e-01 4.52460423e-02 -2.87421852e-01 -4.92070496e-01
3.96962404e-01 -2.21621558e-01 1.06306896e-02 1.59024656e-01
2.04419438e-02 -4.22727853e-01 -5.41858613e-01 5.06102815e-02
4.95583206e-01 2.76351333e-01 1.44722670e-01 -3.55013549e-01
4.38623786e-01 -2.78657973e-01 5.34773052e-01 4.50917989e-01
-3.59564573e-01 1.14011216e+00 5.35379708e-01 -5.35687506e-01
-7.57394433e-01 -1.47236049e+00 -1.66831091e-01 5.89568555e-01
9.17672515e-01 -1.19789116e-01 -5.58952689e-01 -5.73303223e-01
3.48879725e-01 4.93171632e-01 -7.98668623e-01 -2.11288258e-01
-5.31329095e-01 -4.22245234e-01 1.95787296e-01 5.70880890e-01
5.53259552e-01 -4.86455232e-01 -3.60946923e-01 -3.28254640e-01
5.66185378e-02 -1.07422471e+00 -7.49056101e-01 -8.67540389e-02
-7.96642601e-01 -1.10643554e+00 -7.32257605e-01 -7.33458757e-01
6.40124500e-01 8.68701756e-01 1.23992276e+00 2.69141775e-02
-2.39095792e-01 1.50573283e-01 -3.01582366e-01 -4.65531319e-01
4.59072858e-01 -4.93010767e-02 -1.04332015e-01 1.40182734e-01
4.99657661e-01 -1.02031839e+00 -7.72943854e-01 2.68554538e-01
-7.84029663e-01 3.31507146e-01 4.46803421e-01 5.52377284e-01
8.39982450e-01 -2.09483668e-01 1.42587155e-01 -6.86617255e-01
3.99343073e-01 -7.01595902e-01 -3.59167188e-01 4.56341095e-02
-2.11437121e-01 -9.86421295e-03 3.65012050e-01 -3.59913707e-01
-8.99283171e-01 -9.11849067e-02 2.32201722e-02 -1.06669056e+00
-4.83638972e-01 2.40001097e-01 -6.29389286e-01 9.22175273e-02
2.89048165e-01 2.26517826e-01 -3.11587937e-02 -6.98732436e-01
5.72696507e-01 3.74596566e-01 5.23505032e-01 -7.04261363e-01
1.03706145e+00 8.57864738e-01 3.88750038e-03 -6.39219940e-01
-9.05297458e-01 -5.02736807e-01 -7.16879785e-01 8.47293809e-02
1.01558483e+00 -9.24995005e-01 -8.27873170e-01 5.45734227e-01
-1.25055361e+00 1.40225664e-01 -4.30824220e-01 2.24126711e-01
-4.77636933e-01 2.52768308e-01 -2.29696631e-01 -5.43013573e-01
-1.17369965e-01 -1.24519122e+00 1.81677294e+00 4.18147713e-01
1.38326168e-01 -8.09904814e-01 8.59400723e-04 3.89283150e-01
-3.84805910e-02 6.83133364e-01 9.40258861e-01 -4.82428312e-01
-9.71524417e-01 -2.02698216e-01 -5.31133413e-01 1.23009674e-01
3.33649665e-01 -1.82605051e-02 -1.03685617e+00 -4.18104231e-02
2.28854820e-01 5.25395051e-02 9.14203227e-01 4.31983829e-01
1.35661376e+00 -2.94856340e-01 -1.87757745e-01 1.16235328e+00
1.38719070e+00 1.87921785e-02 3.90416682e-01 -3.22645850e-04
1.15467906e+00 4.51246172e-01 2.80163050e-01 2.59341449e-01
6.30862832e-01 6.13479733e-01 7.47938514e-01 -2.74876952e-01
-8.89984518e-02 -3.06424320e-01 -1.07505575e-01 7.06981361e-01
-2.35322177e-01 -1.66147143e-01 -7.43481457e-01 4.09386098e-01
-1.68433893e+00 -7.61635840e-01 -3.31847191e-01 1.80723321e+00
4.29593295e-01 7.88500607e-02 1.32016148e-02 2.40438506e-02
7.06855893e-01 4.65204298e-01 -7.95096517e-01 -1.73610002e-01
-2.50463933e-01 2.08323106e-01 3.18697870e-01 4.13717121e-01
-1.08948231e+00 5.95069945e-01 4.20541000e+00 8.74213159e-01
-9.86322999e-01 -8.16239193e-02 5.81708491e-01 -1.80184454e-01
-7.60407329e-01 -5.17531447e-02 -4.69083875e-01 3.10738504e-01
-4.53239791e-02 -9.44856778e-02 2.41345465e-01 6.29512608e-01
1.01659605e-02 3.56103748e-01 -1.13810921e+00 1.17301071e+00
-1.20861400e-02 -1.24903202e+00 4.28691894e-01 1.11176856e-02
6.37297869e-01 -9.96001586e-02 4.84865636e-01 1.72871917e-01
-4.28564325e-02 -1.08505547e+00 1.08629262e+00 6.79075897e-01
7.56940663e-01 -6.27871692e-01 5.02682984e-01 3.44356596e-01
-1.55504107e+00 1.79848060e-01 -1.21418774e-01 -1.04358889e-01
-7.01553822e-02 4.91575867e-01 -2.33250946e-01 1.07035673e+00
7.52785027e-01 1.09531081e+00 -3.25557798e-01 8.18102539e-01
-6.17060326e-02 3.63602191e-02 -3.60823959e-01 2.18891382e-01
3.16230655e-01 -6.93657935e-01 9.78091419e-01 8.33046198e-01
1.52606651e-01 5.00086904e-01 5.29152639e-02 1.58539975e+00
-6.16477728e-02 -3.05659443e-01 -5.38568556e-01 2.21072733e-01
3.99062991e-01 1.13124347e+00 -7.16768920e-01 -1.51723161e-01
-4.18883532e-01 8.30103755e-01 2.15206042e-01 6.27469420e-01
-8.57400775e-01 -3.35124195e-01 1.14857173e+00 2.16927767e-01
2.67471850e-01 -3.25529724e-01 -7.90889263e-01 -1.41321826e+00
3.38581920e-01 -5.70608616e-01 1.72602311e-02 -1.02931237e+00
-1.58852053e+00 6.61505878e-01 1.56686217e-01 -1.35769606e+00
5.71821928e-01 -6.85871124e-01 -8.08042109e-01 1.14870679e+00
-1.56643534e+00 -1.52434325e+00 -7.79373944e-01 5.47199607e-01
5.96155405e-01 2.96032012e-01 4.92507994e-01 1.69529885e-01
-6.92327678e-01 5.39223194e-01 -1.43148258e-01 1.04873173e-01
2.92733610e-01 -1.43765140e+00 5.88577271e-01 4.82776970e-01
3.61723676e-02 7.38218129e-01 3.46274763e-01 -4.02009726e-01
-1.41068208e+00 -9.64869857e-01 3.53093386e-01 -6.83472931e-01
3.89734685e-01 -5.58953106e-01 -1.12448156e+00 5.97768009e-01
7.27979094e-02 1.96163639e-01 2.41600677e-01 -4.49152999e-02
-6.55065536e-01 -8.69348869e-02 -1.08509278e+00 5.11793375e-01
1.20526421e+00 -5.78424394e-01 -7.43867576e-01 2.04740360e-01
1.26669943e+00 -7.43684113e-01 -6.83178127e-01 5.32218754e-01
4.79671896e-01 -1.14255393e+00 1.34489274e+00 -5.16095579e-01
6.34941161e-01 -3.91433775e-01 -1.56314582e-01 -1.13251805e+00
-4.92078781e-01 -4.47813362e-01 -1.98224872e-01 8.79180014e-01
7.14746416e-02 -8.43963504e-01 7.99467385e-01 3.62993032e-01
-5.30818999e-01 -1.23468637e+00 -9.19439733e-01 -5.55996835e-01
2.80744284e-01 -3.62486571e-01 1.24022782e+00 1.19182229e+00
-4.13983405e-01 3.11834157e-01 1.15376987e-01 5.42012870e-01
6.35100067e-01 6.01306617e-01 5.62910676e-01 -1.37222147e+00
3.84866260e-02 -8.48574460e-01 -6.20071769e-01 -1.53553998e+00
-1.00509562e-01 -1.07776618e+00 -2.38704041e-01 -1.39184928e+00
2.79085428e-01 -5.10888159e-01 -1.58623204e-01 2.89173216e-01
-2.64746070e-01 1.61117569e-01 2.90738881e-01 9.08462703e-02
-1.74166679e-01 8.56560230e-01 1.72441554e+00 -3.23591262e-01
-2.46819183e-01 -4.51613292e-02 -1.12297118e+00 8.70822906e-01
7.11386085e-01 -1.50214240e-01 -4.53382462e-01 -8.81462872e-01
9.25492719e-02 -1.54313847e-01 8.56892705e-01 -6.34791493e-01
6.99430555e-02 -2.12188214e-01 4.23112631e-01 -9.04673278e-01
4.87343013e-01 -1.00880945e+00 6.14675581e-02 -6.24442706e-03
5.94984647e-03 7.61501789e-02 3.67288411e-01 8.04854155e-01
-2.57587343e-01 1.98165357e-01 6.64647281e-01 -1.87488869e-01
-3.91079843e-01 8.45549285e-01 3.23511302e-01 -1.38298061e-03
1.00271046e+00 -6.79189086e-01 -4.21944141e-01 -6.74519539e-02
-7.43649483e-01 3.78395379e-01 5.48980653e-01 5.46130598e-01
8.71749163e-01 -1.41245699e+00 -6.59957528e-01 6.75827146e-01
6.32024407e-02 9.60814834e-01 6.60517395e-01 7.84899533e-01
-6.19790137e-01 1.83154926e-01 -1.65132791e-01 -9.58496451e-01
-9.27482784e-01 5.27556002e-01 5.50983906e-01 5.09281456e-02
-6.59166634e-01 1.01603496e+00 8.97105813e-01 -5.00007331e-01
7.69160464e-02 -9.93849993e-01 -1.72663227e-01 2.60463320e-02
1.50398776e-01 2.08642274e-01 -8.10057148e-02 -7.61998534e-01
-4.83239621e-01 1.23108029e+00 -1.85919309e-03 2.29151085e-01
1.41033828e+00 -1.60344899e-01 -1.15268536e-01 5.07429540e-01
1.54792237e+00 -1.32601401e-02 -1.37163305e+00 -1.94616035e-01
-5.59050620e-01 -7.11176693e-01 8.24075267e-02 -4.70491946e-01
-1.38420224e+00 1.15849292e+00 5.19916117e-01 2.57375866e-01
1.30747461e+00 1.46754071e-01 7.27049708e-01 -1.46209255e-01
1.40202761e-01 -2.39641622e-01 5.66409677e-02 4.35515791e-01
1.21382451e+00 -1.26162934e+00 2.39950558e-03 -6.98699117e-01
-3.94135088e-01 9.82027709e-01 7.97922254e-01 -6.34926260e-01
9.83712673e-01 -4.29684520e-02 -2.04069093e-01 -6.47927403e-01
-4.06558663e-01 -1.03599168e-01 8.45845163e-01 4.54868406e-01
4.15675640e-02 2.80827671e-01 6.64109215e-02 1.18286288e+00
-2.00453386e-01 -4.67470407e-01 1.68643340e-01 6.30389988e-01
-1.78285286e-01 -5.69948316e-01 -3.18868995e-01 1.87215000e-01
4.99918242e-04 -5.14466465e-02 -4.46875274e-01 1.03065896e+00
6.32877052e-01 5.22617817e-01 5.37939668e-01 -4.66230541e-01
6.73985839e-01 -3.05657536e-01 4.55149889e-01 -5.99193335e-01
-3.92448366e-01 1.65678039e-01 -4.94464844e-01 -7.24688947e-01
-2.25790396e-01 -5.40905774e-01 -1.33289325e+00 -1.78743437e-01
-2.68135995e-01 -1.31305188e-01 4.14640635e-01 6.15108848e-01
4.75080818e-01 6.13128543e-01 6.65033102e-01 -1.18171287e+00
-4.34562236e-01 -7.69646525e-01 -6.86418831e-01 5.57844698e-01
8.93215239e-01 -8.88311207e-01 -4.15670991e-01 -2.05004305e-01] | [8.015222549438477, -3.475912094116211] |
b191549f-ed45-4d15-8f90-fb9dd7bc2395 | learning-attribute-structure-co-evolutions-in | 2007.13004 | null | https://arxiv.org/abs/2007.13004v1 | https://arxiv.org/pdf/2007.13004v1.pdf | Learning Attribute-Structure Co-Evolutions in Dynamic Graphs | Most graph neural network models learn embeddings of nodes in static attributed graphs for predictive analysis. Recent attempts have been made to learn temporal proximity of the nodes. We find that real dynamic attributed graphs exhibit complex co-evolution of node attributes and graph structure. Learning node embeddings for forecasting change of node attributes and birth and death of links over time remains an open problem. In this work, we present a novel framework called CoEvoGNN for modeling dynamic attributed graph sequence. It preserves the impact of earlier graphs on the current graph by embedding generation through the sequence. It has a temporal self-attention mechanism to model long-range dependencies in the evolution. Moreover, CoEvoGNN optimizes model parameters jointly on two dynamic tasks, attribute inference and link prediction over time. So the model can capture the co-evolutionary patterns of attribute change and link formation. This framework can adapt to any graph neural algorithms so we implemented and investigated three methods based on it: CoEvoGCN, CoEvoGAT, and CoEvoSAGE. Experiments demonstrate the framework (and its methods) outperform strong baselines on predicting an entire unseen graph snapshot of personal attributes and interpersonal links in dynamic social graphs and financial graphs. | ['Meng Jiang', 'Yihong Ma', 'Zhihan Zhang', 'Tianwen Jiang', 'Daheng Wang', 'Tong Zhao', 'Nitesh V. Chawla'] | 2020-07-25 | null | null | null | null | ['graph-structure-learning'] | ['graphs'] | [-3.13580066e-01 4.54229027e-01 -2.90984094e-01 -3.64140332e-01
6.39551520e-01 -4.78224874e-01 9.77435529e-01 6.48891866e-01
-6.82147518e-02 6.28060400e-01 2.65644401e-01 -1.60336271e-01
-4.49858785e-01 -1.47824204e+00 -6.77609682e-01 -3.98461998e-01
-9.80896592e-01 1.00682461e+00 3.49711806e-01 -5.22550285e-01
-1.59420252e-01 4.96422768e-01 -1.02837348e+00 -2.24957392e-01
5.59751749e-01 4.74323332e-01 -2.94314951e-01 1.04748940e+00
-3.24394107e-01 9.82811332e-01 -3.29721332e-01 -1.20012701e+00
4.70099077e-02 -1.58296719e-01 -6.78273261e-01 -2.47545525e-01
1.46293893e-01 2.21489385e-01 -1.20357108e+00 7.57716417e-01
4.49823052e-01 -4.37641218e-02 9.95323420e-01 -1.88120246e+00
-1.44769013e+00 1.01290643e+00 -4.49749678e-01 6.40234113e-01
2.57269859e-01 -4.28167582e-02 1.43383896e+00 -3.73170763e-01
9.83728647e-01 1.36608231e+00 1.12777150e+00 3.63538533e-01
-1.24885762e+00 -2.89690524e-01 5.34824312e-01 6.08904302e-01
-1.32208633e+00 1.36976331e-01 1.13093770e+00 -6.68634295e-01
1.03302026e+00 1.37575284e-01 1.23063278e+00 1.57379866e+00
6.93372369e-01 3.90063465e-01 3.04884315e-01 -9.21355411e-02
-2.80942559e-01 -8.09936374e-02 4.58879620e-01 1.07675242e+00
3.01158309e-01 1.64405063e-01 -4.74475741e-01 -4.53553230e-01
5.11576891e-01 1.89128295e-01 4.99890232e-03 -2.36152560e-01
-9.04739022e-01 8.06753337e-01 8.53471577e-01 3.62924159e-01
-3.74227256e-01 5.87256730e-01 4.27659124e-01 7.99042881e-01
8.71091783e-01 1.06779367e-01 -4.76652771e-01 1.66844234e-01
-2.14660227e-01 -6.15849346e-02 1.15061879e+00 8.79194796e-01
7.12854743e-01 7.86699802e-02 -1.05364896e-01 7.45978236e-01
2.63714939e-01 1.50830060e-01 4.07851160e-01 -3.10548246e-01
5.28768934e-02 9.21029508e-01 -5.90731263e-01 -1.70458376e+00
-5.35674810e-01 -6.97738051e-01 -1.09732163e+00 -4.39538509e-01
4.14669104e-02 2.07259855e-03 -7.95738876e-01 1.74413884e+00
4.88551915e-01 6.62944794e-01 -4.38272297e-01 2.21840277e-01
8.81564856e-01 6.15915418e-01 1.97479084e-01 -5.02167866e-02
7.71117926e-01 -7.26311147e-01 -9.18898284e-01 -3.68208662e-02
5.53099334e-01 -7.26359189e-02 5.83905160e-01 -2.45736793e-01
-7.33574867e-01 -4.54875082e-01 -8.93226445e-01 1.26585737e-01
-8.00632060e-01 -7.75397480e-01 1.11122358e+00 2.48963758e-01
-1.52169788e+00 1.23838532e+00 -7.09419787e-01 -6.90497339e-01
3.54071885e-01 5.24831235e-01 -3.36570412e-01 1.94244996e-01
-1.58370149e+00 5.96103132e-01 4.31029588e-01 1.12990446e-01
-6.69367731e-01 -6.50112927e-01 -8.01607370e-01 2.36823648e-01
2.11129576e-01 -1.01466298e+00 5.73697448e-01 -9.56484616e-01
-1.02428973e+00 9.10911560e-01 1.48679502e-02 -5.48581481e-01
3.23939770e-01 1.03069291e-01 -1.13714826e+00 -3.78525108e-01
-2.12085158e-01 1.75807446e-01 8.19331408e-01 -8.47607195e-01
-3.45057368e-01 -5.81253171e-01 -2.35480815e-01 -5.96349724e-02
-9.42221701e-01 -6.15179598e-01 -4.59473878e-01 -7.41776288e-01
-1.77705854e-01 -1.06660473e+00 -7.58722499e-02 -1.93096206e-01
-5.00856102e-01 -8.40812683e-01 7.60902166e-01 -4.10003930e-01
1.68719709e+00 -1.82154703e+00 4.88051444e-01 3.43835950e-01
8.35088491e-01 2.34554410e-02 -3.57695907e-01 9.48076606e-01
-6.05205074e-02 4.05392975e-01 1.04155086e-01 -4.35132720e-02
-6.72853440e-02 5.17417252e-01 -1.94994137e-02 5.16169369e-01
-3.49144600e-02 1.49962687e+00 -1.10506487e+00 -5.87573290e-01
-4.18267936e-01 6.09126151e-01 -3.31782967e-01 2.64589041e-01
-2.60687947e-01 -3.32315192e-02 -3.75882924e-01 8.25309873e-01
2.73176193e-01 -6.50010943e-01 5.53701520e-01 -6.26570657e-02
5.96020162e-01 -2.23148018e-01 -6.75374091e-01 1.15233636e+00
-5.57221323e-02 7.59931445e-01 -2.48998657e-01 -8.46757174e-01
1.23475420e+00 -6.14822609e-03 6.12541556e-01 -4.08062369e-01
2.67356671e-02 -1.05423480e-01 2.67377943e-01 -6.33674860e-01
2.89962500e-01 3.33238989e-01 7.15079084e-02 4.56867665e-01
3.44281614e-01 5.84343195e-01 1.56585395e-01 8.80356789e-01
1.52207351e+00 -2.16873139e-01 3.24572176e-01 -2.25634411e-01
5.56404650e-01 -3.66254866e-01 6.36076212e-01 6.09225214e-01
-3.14270884e-01 -1.22930244e-01 9.83430505e-01 -1.06582403e+00
-1.16533816e+00 -1.21811998e+00 1.93196058e-01 1.48726594e+00
-7.44366571e-02 -7.51643717e-01 -2.45098978e-01 -8.94396722e-01
6.47073507e-01 4.47590172e-01 -1.10561001e+00 -5.40232718e-01
-6.80432260e-01 -9.41910505e-01 3.19020957e-01 4.86343741e-01
-1.19691171e-01 -1.26855516e+00 6.77813470e-01 5.15931010e-01
2.40906239e-01 -9.04134512e-01 -6.77525461e-01 -1.14959650e-01
-7.96555698e-01 -1.44138920e+00 -8.47063437e-02 -9.91431415e-01
5.18673718e-01 -3.83831859e-01 1.63022149e+00 6.47846043e-01
-3.70327115e-01 7.66279638e-01 -2.80543000e-01 -3.05518925e-01
-5.20460904e-01 5.11405766e-01 1.79401800e-01 3.90722066e-01
3.60599130e-01 -1.32738388e+00 -4.52376157e-01 1.39302015e-01
-5.33496797e-01 -4.37426597e-01 3.35851520e-01 8.01729500e-01
3.46038133e-01 -1.07336566e-01 6.94219947e-01 -1.63937151e+00
1.13952041e+00 -9.57942426e-01 -2.15003595e-01 4.74082083e-01
-1.31519008e+00 2.65982270e-01 5.94231486e-01 -6.21132553e-01
-4.50209051e-01 -3.78076524e-01 9.04670432e-02 -4.06094939e-01
4.84505385e-01 6.85411096e-01 1.62677079e-01 1.05166204e-01
4.16944563e-01 2.47742370e-01 8.60830471e-02 -2.96556592e-01
5.25021791e-01 1.77858993e-01 4.48608845e-01 -2.56396532e-01
1.01598728e+00 3.03973049e-01 3.18497211e-01 -6.98537946e-01
-5.24823785e-01 -2.08987549e-01 -9.53736246e-01 -4.77172345e-01
5.62059164e-01 -5.10975897e-01 -9.57985699e-01 5.43787062e-01
-1.09738803e+00 -3.19183975e-01 -9.72020477e-02 -1.22446204e-02
-2.92403344e-02 2.92641550e-01 -1.03365195e+00 -3.89993787e-01
-5.10909021e-01 -1.72882691e-01 3.59314352e-01 -3.42861339e-02
-1.43436059e-01 -2.04887176e+00 5.88167071e-01 -4.20645446e-01
4.90923017e-01 8.02583039e-01 1.30182111e+00 -9.45472479e-01
-4.84487981e-01 -2.72098869e-01 4.28646579e-02 -2.68448144e-01
2.62380689e-01 4.37173098e-01 -4.83114153e-01 -5.38497806e-01
-6.91730261e-01 1.23776481e-01 1.04879475e+00 2.16263965e-01
1.08424687e+00 -7.16466248e-01 -9.58748579e-01 9.08124030e-01
1.28485799e+00 -3.68507542e-02 3.57835829e-01 1.96291670e-01
1.11808777e+00 5.79311013e-01 -9.54470634e-02 2.72690207e-01
8.61860156e-01 4.74067748e-01 8.13451767e-01 1.81562364e-01
-2.29498729e-01 -6.40690386e-01 2.66522259e-01 1.46113181e+00
-2.99139708e-01 -7.24254787e-01 -8.86720598e-01 6.51271164e-01
-1.93896282e+00 -9.83118057e-01 -5.84878385e-01 1.68721735e+00
3.52575421e-01 2.51638472e-01 2.21214592e-01 -3.00879031e-01
1.11027634e+00 6.05435908e-01 -8.96893680e-01 -5.47008097e-01
-2.04905480e-01 -1.93214864e-01 4.04141158e-01 6.21349216e-01
-8.95977259e-01 8.18860888e-01 6.09962130e+00 2.75761247e-01
-5.79108775e-01 1.86292350e-01 6.36154532e-01 2.46968716e-01
-6.47158682e-01 -5.06022535e-02 -6.30108058e-01 5.72941899e-01
1.36340439e+00 -7.30402112e-01 5.38840652e-01 6.14019990e-01
-5.68171620e-01 9.58948493e-01 -1.31926131e+00 6.60860062e-01
8.81983414e-02 -1.42075229e+00 2.52738148e-01 2.06275433e-01
6.34033382e-01 2.79091537e-01 4.69301417e-02 5.53331017e-01
9.75416780e-01 -1.12623250e+00 -2.58325897e-02 1.00711429e+00
5.89005649e-01 -7.41917193e-01 7.52905786e-01 7.61784241e-02
-1.55301857e+00 -3.64284456e-01 -3.96589577e-01 1.03038378e-01
1.18881561e-01 6.23376489e-01 -1.12564921e+00 6.40192807e-01
5.97528577e-01 1.42231011e+00 -1.00822711e+00 5.55683076e-01
-2.40256280e-01 7.34468222e-01 1.40192807e-01 -2.55597055e-01
5.34137078e-02 -3.93599898e-01 8.39177191e-01 1.06671572e+00
2.02811480e-01 -1.40048144e-02 8.75822082e-02 6.07853889e-01
-4.75065827e-01 -2.97972970e-02 -1.03888345e+00 -3.82806331e-01
7.81979740e-01 1.36724746e+00 -5.79136312e-01 -7.61361197e-02
-2.37082005e-01 1.06218898e+00 1.01698124e+00 3.77051532e-01
-7.74259746e-01 -2.19395071e-01 7.84791708e-01 6.68495715e-01
1.94337815e-01 -1.74919635e-01 4.68423516e-01 -8.89793575e-01
-2.72679746e-01 -3.05413455e-01 1.15687811e+00 -3.92351270e-01
-2.03295350e+00 1.03294814e+00 -4.33592737e-01 -6.61646664e-01
-2.41859972e-01 -5.07003665e-01 -1.00068974e+00 3.91511530e-01
-1.38710225e+00 -1.20383096e+00 -3.86704803e-01 8.89925778e-01
3.56652886e-02 -5.56712508e-01 7.28858769e-01 1.30489305e-01
-5.77149689e-01 6.44830644e-01 2.18797043e-01 4.92735267e-01
3.60333115e-01 -1.52135193e+00 8.47431183e-01 4.08671767e-01
5.49110413e-01 3.36106896e-01 5.39009213e-01 -1.03847754e+00
-1.48147583e+00 -1.47575247e+00 1.03297496e+00 -6.28916860e-01
1.23689163e+00 -5.44216156e-01 -1.14780354e+00 1.18462455e+00
3.29377502e-01 5.34409583e-01 5.42918921e-01 6.26736701e-01
-3.55893552e-01 -3.94011378e-01 -7.08228886e-01 3.95118624e-01
1.73343289e+00 -5.39923728e-01 -5.31831533e-02 6.43735647e-01
1.22204006e+00 1.67371839e-01 -1.22700191e+00 3.52751464e-01
3.88280600e-01 -6.80150986e-01 1.09750044e+00 -1.14929533e+00
2.23237649e-01 1.95795476e-01 2.68167466e-01 -1.73624551e+00
-1.06558907e+00 -8.67993772e-01 -1.01705337e+00 1.44396639e+00
4.49008465e-01 -9.91992414e-01 8.18823636e-01 1.92315519e-01
1.41072407e-01 -8.43430817e-01 -7.09960580e-01 -6.98480368e-01
-1.09800987e-01 2.09143326e-01 8.77121866e-01 1.40665615e+00
-2.91760087e-01 7.70494759e-01 -4.87997115e-01 2.25308001e-01
7.80272722e-01 4.63846214e-02 5.57822764e-01 -2.02042460e+00
-2.41669342e-01 -5.19102097e-01 -9.29145157e-01 -3.13218802e-01
5.56751311e-01 -1.58764803e+00 -9.47496355e-01 -1.64866531e+00
1.48411274e-01 -3.21649939e-01 -5.68368673e-01 2.78965890e-01
-4.04091269e-01 -2.48567358e-01 -1.32755220e-01 2.70325840e-01
-5.97421587e-01 8.46842170e-01 1.11248600e+00 -4.61389631e-01
-1.67109787e-01 -7.92896897e-02 -3.86277586e-01 4.36423868e-01
7.32187033e-01 -7.29775131e-01 -6.31275833e-01 -3.09133857e-01
7.00233996e-01 9.62269977e-02 2.52046704e-01 -5.59615731e-01
4.60285574e-01 1.13472618e-01 3.72921824e-01 -2.95387715e-01
1.17598899e-01 -7.72744775e-01 6.39015853e-01 6.67660654e-01
-4.17833030e-01 7.75497139e-01 -3.92271161e-01 1.47434390e+00
2.18303092e-02 1.89162701e-01 2.12176487e-01 -9.76111740e-02
-7.85632849e-01 1.13145053e+00 5.24380282e-02 1.79119617e-01
1.15173519e+00 -1.09828882e-01 -4.64755177e-01 -6.35951102e-01
-1.26565385e+00 5.75313568e-01 1.42299816e-01 8.29602301e-01
6.90080881e-01 -1.71757579e+00 -1.00816560e+00 4.67287190e-03
-3.24757174e-02 -5.41639149e-01 3.18186730e-02 5.65938592e-01
-3.35674614e-01 -1.40573785e-01 -8.49247798e-02 -4.04144764e-01
-1.34510338e+00 9.61198151e-01 4.13437188e-01 -6.09541357e-01
-7.49069870e-01 1.00658000e+00 -2.98993677e-01 -7.08011508e-01
1.15874395e-01 4.36350852e-01 -6.72836006e-01 3.19059193e-01
7.46657467e-03 4.96676832e-01 -2.74770081e-01 -6.29010499e-01
-2.52978623e-01 2.72962838e-01 -3.68173599e-01 6.62130117e-01
1.86239922e+00 -1.60657346e-01 -6.10291660e-01 7.95346081e-01
1.37662113e+00 -1.90315887e-01 -9.34572458e-01 -5.28889239e-01
3.56302142e-01 -3.35826725e-01 -2.81321824e-01 -4.64049786e-01
-1.42610931e+00 5.08310139e-01 2.71095157e-01 7.95655966e-01
6.28191113e-01 2.79681206e-01 7.46078372e-01 3.20703000e-01
1.33675188e-01 -8.12009573e-01 3.26520145e-01 4.15166885e-01
7.84164310e-01 -9.53203499e-01 1.53095186e-01 -5.33673525e-01
-4.72335756e-01 1.23653686e+00 8.85100007e-01 -2.13021189e-01
1.31592786e+00 -3.71258892e-02 -4.78103369e-01 -7.99229443e-01
-1.22755921e+00 1.11899756e-01 4.39107269e-01 7.62205482e-01
3.07769388e-01 1.99406937e-01 -1.48583442e-01 2.50519425e-01
-1.88897923e-01 -5.65673769e-01 5.80921054e-01 3.05691212e-01
-1.23342350e-01 -1.20645380e+00 3.35571766e-01 7.80899704e-01
-2.52669334e-01 -1.02884889e-01 -7.81807065e-01 9.51150477e-01
-1.52675480e-01 4.64391500e-01 3.28720957e-01 -8.43530655e-01
2.02437058e-01 1.54391006e-01 2.05939606e-01 -3.03830832e-01
-4.96757835e-01 -6.35786474e-01 2.42429093e-01 -4.36029553e-01
-1.80551529e-01 -5.91190517e-01 -7.76494265e-01 -9.12900031e-01
-7.17592686e-02 1.75286084e-01 9.26604792e-02 2.60885417e-01
6.33746207e-01 9.31516528e-01 7.97583640e-01 -2.85077631e-01
-2.19219074e-01 -1.06651831e+00 -9.17211413e-01 8.60840499e-01
2.19346046e-01 -5.77956200e-01 -4.92621273e-01 -1.96177676e-01] | [7.172128677368164, 6.10564661026001] |
9874106c-e210-47df-82ce-b07c60dbde9a | open-vocabulary-object-detection-via-scene | 2307.03339 | null | https://arxiv.org/abs/2307.03339v1 | https://arxiv.org/pdf/2307.03339v1.pdf | Open-Vocabulary Object Detection via Scene Graph Discovery | In recent years, open-vocabulary (OV) object detection has attracted increasing research attention. Unlike traditional detection, which only recognizes fixed-category objects, OV detection aims to detect objects in an open category set. Previous works often leverage vision-language (VL) training data (e.g., referring grounding data) to recognize OV objects. However, they only use pairs of nouns and individual objects in VL data, while these data usually contain much more information, such as scene graphs, which are also crucial for OV detection. In this paper, we propose a novel Scene-Graph-Based Discovery Network (SGDN) that exploits scene graph cues for OV detection. Firstly, a scene-graph-based decoder (SGDecoder) including sparse scene-graph-guided attention (SSGA) is presented. It captures scene graphs and leverages them to discover OV objects. Secondly, we propose scene-graph-based prediction (SGPred), where we build a scene-graph-based offset regression (SGOR) mechanism to enable mutual enhancement between scene graph extraction and object localization. Thirdly, we design a cross-modal learning mechanism in SGPred. It takes scene graphs as bridges to improve the consistency between cross-modal embeddings for OV object classification. Experiments on COCO and LVIS demonstrate the effectiveness of our approach. Moreover, we show the ability of our model for OV scene graph detection, while previous OV scene graph generation methods cannot tackle this task. | ['Jianfei Cai', 'Munawar Hayat', 'Hengcan Shi'] | 2023-07-07 | null | null | null | null | ['open-vocabulary-object-detection', 'scene-graph-generation', 'object-detection', 'object-localization', 'graph-generation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'graphs'] | [ 9.97729599e-02 1.07305437e-01 -2.31172815e-01 -3.27713937e-01
-4.53500301e-01 -4.52712178e-01 6.55053973e-01 2.76564747e-01
-1.64484203e-01 3.41833732e-03 2.51319587e-01 -2.42799848e-01
2.02594727e-01 -9.95218933e-01 -9.52515483e-01 -4.30327028e-01
2.99397260e-01 5.67295253e-02 3.95939440e-01 -9.83588845e-02
9.17377770e-02 2.34050527e-01 -1.55759597e+00 2.24602640e-01
7.52826571e-01 1.14351416e+00 7.16387570e-01 4.12418783e-01
-5.42879760e-01 1.02561164e+00 -3.77391785e-01 -4.28513587e-01
1.76420167e-01 -1.50792003e-01 -4.91008103e-01 3.20675462e-01
7.50590205e-01 -1.46046624e-01 -6.96750343e-01 1.16978765e+00
2.72195429e-01 3.55040543e-02 7.44807720e-01 -1.43543220e+00
-1.27795196e+00 5.88533342e-01 -5.32582164e-01 7.39206895e-02
2.24907666e-01 3.49085957e-01 1.50184870e+00 -1.30336893e+00
6.71987593e-01 1.41939151e+00 3.88903648e-01 3.13580543e-01
-1.10695279e+00 -5.88579774e-01 4.37155634e-01 3.90118182e-01
-1.69018042e+00 -2.48888209e-01 1.01413274e+00 -4.63277400e-01
8.24543417e-01 1.03221945e-01 9.06286657e-01 1.03060281e+00
-2.82515194e-02 1.29915607e+00 6.25255942e-01 -4.89541322e-01
-4.19979617e-02 1.56471461e-01 2.47685984e-01 7.61729240e-01
5.56843877e-01 -1.49376616e-01 -3.99850547e-01 2.84858823e-01
7.86993265e-01 3.04663807e-01 -3.69425774e-01 -7.31853783e-01
-1.25872004e+00 9.88421023e-01 1.07979488e+00 1.64400458e-01
-2.73409933e-01 1.83918372e-01 4.28714216e-01 1.64923090e-02
3.39642197e-01 3.91728193e-01 -1.72497064e-01 5.14691770e-01
-4.83048767e-01 1.11479841e-01 4.56049770e-01 1.48088062e+00
8.36566985e-01 2.13974208e-01 -5.21408975e-01 1.06105304e+00
7.86625743e-01 6.40924871e-01 2.99221873e-01 -2.32488349e-01
5.54870903e-01 9.73346770e-01 -5.18732131e-01 -1.19767010e+00
-2.78934777e-01 -4.46079761e-01 -5.88850498e-01 -4.39796537e-01
-5.49662625e-03 2.80201346e-01 -1.18268323e+00 1.67350686e+00
4.80703801e-01 3.60477030e-01 1.73640579e-01 1.06118417e+00
1.64395404e+00 6.13645792e-01 2.19561592e-01 1.47141635e-01
1.62011659e+00 -1.21131849e+00 -5.85473895e-01 -5.81166267e-01
8.28922272e-01 -4.50158626e-01 1.29951143e+00 -3.54773290e-02
-5.31023800e-01 -5.42921901e-01 -9.77888525e-01 -3.80382955e-01
-6.19270504e-01 3.70585956e-02 7.66139507e-01 3.03962559e-01
-7.78501391e-01 -2.36245513e-01 -4.17261422e-01 -5.06884754e-01
6.71747088e-01 -4.43028398e-02 -3.16622317e-01 -3.61283213e-01
-1.01436305e+00 4.76061165e-01 6.40371084e-01 7.73035586e-02
-1.12910593e+00 -6.12423778e-01 -1.72287250e+00 1.03235967e-01
6.70555115e-01 -7.58120537e-01 9.10870969e-01 -7.23450840e-01
-8.28182399e-01 9.55440223e-01 -1.56822905e-01 -3.52806062e-01
-6.80062398e-02 6.28461391e-02 -3.71927679e-01 2.18385130e-01
3.37971121e-01 9.39730644e-01 1.01030183e+00 -1.44586265e+00
-6.05054736e-01 -2.12229148e-01 3.96727204e-01 2.95122713e-01
-4.01414543e-01 -8.78309179e-03 -9.19983625e-01 -7.34084070e-01
4.14008319e-01 -6.20848298e-01 -1.47574961e-01 1.48850400e-02
-8.97022367e-01 -5.20275176e-01 7.68835187e-01 -3.78527194e-01
1.17672813e+00 -2.32534122e+00 -4.35215198e-02 -1.01406937e-02
6.24412894e-01 7.52609000e-02 -4.23601538e-01 1.48500904e-01
-1.01401834e-02 7.60909216e-03 -8.33104625e-02 -4.68474209e-01
1.33095026e-01 3.40047866e-01 -4.32073534e-01 3.78785580e-01
5.47860384e-01 1.45867240e+00 -8.99247110e-01 -8.10114861e-01
3.59987438e-01 3.04565996e-01 -7.19464481e-01 9.43765119e-02
-5.37876546e-01 7.79869556e-02 -6.01774573e-01 9.56121683e-01
6.25129282e-01 -5.43755472e-01 -2.20258340e-01 -4.81731266e-01
1.04965232e-01 1.38243288e-01 -1.03193212e+00 1.57221389e+00
-4.68989670e-01 7.82880962e-01 -3.38577837e-01 -9.02793169e-01
9.02216673e-01 -1.45744354e-01 1.72398806e-01 -7.97398090e-01
2.54467547e-01 -2.80967169e-02 -1.47188157e-01 -4.64091599e-01
6.68355227e-01 1.88310906e-01 -1.90866694e-01 -1.52311087e-01
3.49180996e-01 -2.19778910e-01 1.32632241e-01 6.42437220e-01
1.00846267e+00 -7.18337297e-02 3.23702246e-01 -2.47806907e-02
5.80127895e-01 1.25202790e-01 4.88056839e-01 6.75048053e-01
-1.65729702e-01 7.77306795e-01 4.52588826e-01 -6.89953268e-02
-7.68787742e-01 -1.02700114e+00 -2.58101344e-01 9.74673092e-01
8.35924029e-01 -7.55189657e-01 -3.88457626e-01 -8.44612718e-01
8.89864564e-02 7.44638145e-01 -5.60937464e-01 -3.42797011e-01
-1.88018188e-01 -3.45222801e-01 4.03257877e-01 5.88220656e-01
3.16847652e-01 -1.12235034e+00 -2.33459979e-01 -6.86615184e-02
-9.37204584e-02 -1.54587519e+00 -6.67887807e-01 1.24042779e-01
-4.51682508e-01 -1.07618570e+00 -3.32244545e-01 -1.19338012e+00
7.94695854e-01 8.52850080e-01 1.03873944e+00 9.25836414e-02
-2.88696498e-01 6.46489382e-01 -5.92550755e-01 -6.29170299e-01
-7.14090466e-02 -1.89811245e-01 -6.42130822e-02 3.17794025e-01
5.70928216e-01 -1.54423982e-01 -4.58956420e-01 1.03375778e-01
-8.67597878e-01 2.43596241e-01 6.37249231e-01 8.29588413e-01
8.80104363e-01 -1.66455433e-01 2.20711887e-01 -7.36160219e-01
2.90200174e-01 -6.36572003e-01 -6.63226187e-01 2.12632611e-01
-3.60652208e-01 -1.18702799e-01 4.89391506e-01 -5.34848928e-01
-7.30330765e-01 -1.23219108e-02 1.12156004e-01 -1.11026621e+00
-1.75307944e-01 6.24209225e-01 -6.00906789e-01 -1.03751324e-01
4.10531849e-01 4.67188150e-01 -3.80806059e-01 -3.23551625e-01
7.66292751e-01 6.04564548e-01 4.67313975e-01 -1.35016769e-01
1.05863726e+00 4.47444826e-01 -2.36737952e-01 -1.02498126e+00
-1.14206302e+00 -9.42890406e-01 -4.35235232e-01 -1.65555298e-01
1.13125134e+00 -1.40182734e+00 -3.33760411e-01 2.55289227e-01
-1.22875011e+00 -2.66481042e-01 -3.65390897e-01 4.32126850e-01
-2.30777785e-01 3.31017286e-01 -4.16748524e-01 -6.57883763e-01
-1.63267419e-01 -9.89817619e-01 1.65724289e+00 2.65667975e-01
2.89230317e-01 -9.72340703e-01 -2.85450250e-01 2.90609062e-01
-1.54763326e-01 1.33097813e-01 6.24526441e-01 -8.65791917e-01
-9.35716629e-01 -1.41387597e-01 -8.72495532e-01 2.43834272e-01
3.27608315e-03 -3.71871173e-01 -1.02321410e+00 -4.79868501e-02
-3.58025223e-01 -3.66493255e-01 9.81944263e-01 2.73976117e-01
1.20845103e+00 -1.63807899e-01 -5.71767151e-01 9.32288408e-01
1.52885878e+00 -2.22956076e-01 4.78246450e-01 3.13591152e-01
1.71206129e+00 4.66319472e-01 7.67846525e-01 1.96659967e-01
8.75508130e-01 7.14158058e-01 6.10088229e-01 -2.35519454e-01
-5.98986924e-01 -5.73193848e-01 4.20951366e-01 7.86741495e-01
5.52020669e-01 -3.84463847e-01 -8.87542903e-01 8.72231245e-01
-1.64827168e+00 -6.59849882e-01 -6.31035268e-01 1.75831020e+00
4.71709549e-01 1.81416750e-01 -2.31329709e-01 -2.02501208e-01
6.42665744e-01 3.54263216e-01 -4.98351038e-01 -5.74493669e-02
-3.73139352e-01 -1.71409503e-01 6.08099341e-01 1.65864393e-01
-1.35492134e+00 1.38517439e+00 4.33844709e+00 1.00000107e+00
-9.39365387e-01 3.15638602e-01 1.78181604e-01 1.86903700e-01
-5.35930932e-01 1.32518351e-01 -9.78498578e-01 3.25529605e-01
2.94575632e-01 -1.72923543e-02 3.21185775e-02 1.38196766e+00
-1.44740298e-01 2.15674192e-01 -1.10282445e+00 1.32310140e+00
5.94206870e-01 -1.28178930e+00 3.36093545e-01 6.12594709e-02
5.63793480e-01 1.97795704e-01 -2.42664635e-01 8.04872394e-01
2.64094859e-01 -7.69335687e-01 1.12165606e+00 2.12554246e-01
7.76490510e-01 -4.52730775e-01 5.75509012e-01 1.98539734e-01
-1.76332295e+00 -1.22965291e-01 -5.72925508e-01 2.70446151e-01
2.02591196e-01 5.37668109e-01 -6.88101709e-01 5.93416870e-01
5.39249599e-01 1.07590985e+00 -9.50353861e-01 1.22335100e+00
-6.12504005e-01 6.58492208e-01 -2.05412939e-01 -5.46760485e-02
3.83033395e-01 -1.31885350e-01 7.21187294e-01 1.05442786e+00
-4.74993512e-02 -6.59791529e-02 5.09978116e-01 1.23114634e+00
-2.09092945e-01 3.26565772e-01 -8.95757735e-01 -1.12780422e-01
4.58859801e-01 1.33594120e+00 -7.55850255e-01 -2.45708793e-01
-7.56259620e-01 9.24539268e-01 4.61749852e-01 2.37898797e-01
-9.40185010e-01 -3.80826086e-01 5.91343641e-01 -1.32236823e-01
6.66284919e-01 -2.00958967e-01 -1.38954833e-01 -1.41621101e+00
6.66337386e-02 -5.42562008e-01 4.53906685e-01 -9.72118676e-01
-1.49544287e+00 3.59529793e-01 3.83445024e-02 -1.38120699e+00
3.21798295e-01 -6.85468435e-01 -5.04633009e-01 4.59391624e-01
-1.73635936e+00 -1.70646739e+00 -5.16827524e-01 7.01116264e-01
8.72527659e-01 -3.39611508e-02 1.89439565e-01 2.88106233e-01
-6.86302483e-01 6.39708400e-01 -2.83851147e-01 4.80088979e-01
3.39868277e-01 -1.20538843e+00 4.17523563e-01 1.15819144e+00
8.20411026e-01 6.23302996e-01 2.74612665e-01 -7.92856455e-01
-1.78338420e+00 -1.78045845e+00 7.29845762e-01 -6.47099733e-01
8.00122082e-01 -8.24037075e-01 -9.41811621e-01 8.21004689e-01
-1.94303781e-01 5.21915555e-01 3.54707479e-01 7.13306144e-02
-7.22305834e-01 9.95441452e-02 -6.09985590e-01 8.79313350e-01
1.41807485e+00 -7.59698868e-01 -6.56340897e-01 5.76889515e-01
1.41501236e+00 -5.50189674e-01 -5.24434865e-01 5.36678851e-01
1.27594993e-01 -4.47350502e-01 1.12890577e+00 -4.40744728e-01
3.26213509e-01 -5.90334654e-01 -4.40731585e-01 -1.02873051e+00
-3.11482400e-01 -2.57776249e-02 -3.31907690e-01 1.42048800e+00
3.06895614e-01 -4.89919424e-01 4.19494420e-01 4.65316214e-02
-5.52512765e-01 -6.42637193e-01 -4.82423306e-01 -8.69593143e-01
-4.20518309e-01 -1.05614245e+00 5.70783138e-01 9.45447087e-01
-3.12308520e-01 7.95098603e-01 -1.28602251e-01 5.96389413e-01
5.37956476e-01 4.51988250e-01 9.77138042e-01 -1.14510441e+00
-1.72198668e-01 -4.57111448e-01 -9.14218485e-01 -1.27143812e+00
2.42180824e-01 -1.32484555e+00 2.60696113e-01 -1.84924579e+00
3.17768812e-01 -4.94073570e-01 -2.37494990e-01 5.22419572e-01
-3.95867407e-01 3.77233237e-01 4.53878999e-01 1.12807021e-01
-1.16836786e+00 9.05881643e-01 1.27654183e+00 -4.37466145e-01
-2.37196714e-01 -5.27097642e-01 -8.36847782e-01 7.54862666e-01
3.93861443e-01 -3.77875060e-01 -4.92175639e-01 -4.50365990e-01
1.40460506e-01 -4.70555872e-01 8.12787950e-01 -7.82707155e-01
2.32213691e-01 2.49037854e-02 1.32450551e-01 -7.73073673e-01
2.10071772e-01 -7.43110597e-01 -3.02080721e-01 1.05875626e-01
-8.38769823e-02 -5.11213481e-01 9.89489704e-02 1.00857866e+00
-2.48895511e-01 -4.01206724e-02 5.14466047e-01 1.44578248e-01
-1.29170084e+00 6.22741997e-01 4.34247814e-02 2.20532298e-01
1.13039017e+00 -3.19148481e-01 -2.88220376e-01 -3.71895619e-02
-3.89053673e-01 5.96317410e-01 4.01066929e-01 7.81706989e-01
9.50656593e-01 -1.49331391e+00 -5.01871645e-01 2.31886685e-01
1.06030011e+00 5.46612382e-01 9.03022811e-02 8.76278639e-01
-3.10452074e-01 3.58863235e-01 4.97365147e-01 -1.07570076e+00
-1.28598523e+00 8.96562696e-01 1.03156738e-01 1.45122573e-01
-9.93636549e-01 1.25395358e+00 1.00185120e+00 -5.51665843e-01
1.96500182e-01 -4.97711271e-01 -5.03491402e-01 8.18249509e-02
4.26244080e-01 -3.10576260e-01 -1.44865885e-01 -9.77754414e-01
-2.99722135e-01 5.71679533e-01 4.39907089e-02 3.95564437e-01
1.07183564e+00 -3.08357626e-01 -1.26524329e-01 4.98212725e-01
1.14990985e+00 3.59982960e-02 -1.07039678e+00 -6.29724801e-01
-3.22761796e-02 -4.16810960e-01 3.26849788e-01 -1.79790661e-01
-1.09459221e+00 8.94782186e-01 3.27220738e-01 7.14363232e-02
9.70300317e-01 5.51641643e-01 7.86309123e-01 3.14186066e-01
4.32069927e-01 -6.00032210e-01 2.99657971e-01 5.74857473e-01
7.62459815e-01 -1.50964320e+00 -2.51689255e-01 -8.91650856e-01
-7.29245126e-01 8.16182613e-01 8.64788532e-01 -6.11615181e-02
6.12773776e-01 -1.41018406e-01 -2.59274065e-01 -7.12611973e-01
-5.85958958e-01 -1.01539409e+00 6.18169606e-01 7.13786542e-01
-9.98656824e-02 1.18052550e-01 1.31066695e-01 8.70603442e-01
-7.70259053e-02 -5.05946875e-01 3.83472145e-01 6.99800730e-01
-4.28038299e-01 -6.24327481e-01 -3.07065040e-01 6.35333300e-01
2.55596936e-02 -5.88083804e-01 -3.28697115e-01 7.87448108e-01
4.22515005e-01 9.75284696e-01 1.29611775e-01 -3.76887590e-01
3.63327265e-01 -3.05844605e-01 1.83854386e-01 -1.18634033e+00
-4.11003008e-02 2.47594826e-02 4.19575255e-03 -6.45659983e-01
-1.64342135e-01 -5.31136155e-01 -1.27618897e+00 2.91481107e-01
-7.36291111e-01 -2.09581152e-01 5.41408539e-01 9.93780375e-01
3.52937222e-01 7.60411739e-01 5.49329460e-01 -6.32794440e-01
8.51261243e-02 -7.55953372e-01 -6.94415748e-01 5.06740808e-01
3.43851268e-01 -9.61913049e-01 -3.79929662e-01 3.39975134e-02] | [10.345049858093262, 1.5066074132919312] |
f687f3f0-fed8-4610-b193-9f9847613d89 | sequence-learning-with-rnns-for-medical | 1811.11523 | null | http://arxiv.org/abs/1811.11523v2 | http://arxiv.org/pdf/1811.11523v2.pdf | Sequence Learning with RNNs for Medical Concept Normalization in User-Generated Texts | In this work, we consider the medical concept normalization problem, i.e.,
the problem of mapping a disease mention in free-form text to a concept in a
controlled vocabulary, usually to the standard thesaurus in the Unified Medical
Language System (UMLS). This task is challenging since medical terminology is
very different when coming from health care professionals or from the general
public in the form of social media texts. We approach it as a sequence learning
problem, with recurrent neural networks trained to obtain semantic
representations of one- and multi-word expressions. We develop end-to-end
neural architectures tailored specifically to medical concept normalization,
including bidirectional LSTM and GRU with an attention mechanism and additional
semantic similarity features based on UMLS. Our evaluation over a standard
benchmark shows that our model improves over a state of the art baseline for
classification based on CNNs. | ['Elena Tutubalina', 'Zulfat Miftahutdinov', 'Sergey Nikolenko', 'Valentin Malykh'] | 2018-11-28 | null | null | null | null | ['medical-concept-normalization'] | ['medical'] | [ 9.04866874e-01 4.49071467e-01 -4.30802077e-01 -4.47662711e-01
-9.59858239e-01 -1.90700769e-01 4.07913864e-01 8.05650830e-01
-1.08856916e+00 6.67476773e-01 6.48363054e-01 -4.67170149e-01
3.40105779e-02 -7.95646608e-01 -4.65916842e-01 -4.62517887e-01
2.30951816e-01 6.38064206e-01 -1.78041235e-01 -6.21541739e-01
-2.18225583e-01 1.21504188e-01 -1.00542939e+00 4.32928890e-01
3.64719361e-01 7.73891151e-01 -1.92253992e-01 5.96761405e-01
-3.76078069e-01 1.08265960e+00 -6.68638468e-01 -5.27088821e-01
-1.90247610e-01 -4.70869213e-01 -1.30509412e+00 -4.27855223e-01
2.24911973e-01 1.59086630e-01 -1.88119672e-02 1.34058774e+00
5.98113120e-01 3.34665090e-01 6.89894676e-01 -6.91875637e-01
-9.56829548e-01 7.06119716e-01 -2.83046722e-01 1.64087757e-01
2.89524674e-01 -6.05232954e-01 1.10937130e+00 -4.04811352e-01
1.11108315e+00 1.37130201e+00 8.42160106e-01 1.01596653e+00
-1.21177077e+00 -4.70011413e-01 1.98144928e-01 -7.19655529e-02
-1.31610477e+00 -3.16424310e-01 2.63504852e-02 -4.17543948e-01
1.42247736e+00 2.34357908e-01 7.47995526e-02 1.53790820e+00
4.89044785e-01 4.91186887e-01 3.42183739e-01 -6.25287890e-01
6.91430643e-02 -1.01456106e-01 4.63220567e-01 5.21661043e-01
4.15732741e-01 -3.91911179e-01 -6.75782114e-02 -3.18574488e-01
3.78566444e-01 3.34665060e-01 -1.81778014e-01 -1.09816045e-01
-1.20214844e+00 1.04714549e+00 4.97082204e-01 8.29568505e-01
-3.01739037e-01 2.17267737e-01 8.24464560e-01 3.66757333e-01
8.31828773e-01 5.82119346e-01 -8.44092309e-01 3.54567558e-01
-6.33079052e-01 1.88895270e-01 9.05169308e-01 8.81687164e-01
2.42189154e-01 -4.02594656e-01 -2.91011512e-01 9.93275523e-01
1.76785707e-01 7.80493915e-02 1.06833065e+00 -2.44755730e-01
3.18262577e-01 6.87356591e-01 -2.02209488e-01 -6.41468346e-01
-8.11194122e-01 -6.16689086e-01 -1.14133799e+00 -3.03049415e-01
-4.52363417e-02 -2.17827976e-01 -1.28824317e+00 2.01505733e+00
1.18945800e-01 3.52650762e-01 3.87283057e-01 5.22587299e-01
1.34480727e+00 3.18042427e-01 7.16610968e-01 -1.56529754e-01
1.75331891e+00 -9.53493476e-01 -1.13712955e+00 -3.14179063e-01
1.26621187e+00 -5.77216744e-01 5.87842107e-01 -1.44200757e-01
-7.72924483e-01 -2.09439605e-01 -9.59658444e-01 -5.36961555e-01
-1.04274762e+00 -2.90836364e-01 3.90877217e-01 5.48302174e-01
-9.93567050e-01 7.48757958e-01 -1.00772023e+00 -9.13658440e-01
4.71373975e-01 3.43577921e-01 -5.31384706e-01 -9.23314095e-02
-1.61150110e+00 1.15357494e+00 5.90164423e-01 -2.80711859e-01
-2.04390392e-01 -9.89352167e-01 -1.27000415e+00 6.96592033e-02
3.62801939e-01 -1.25145853e+00 1.44751120e+00 -1.26698315e+00
-1.07702947e+00 1.42868841e+00 2.70959060e-03 -7.89324582e-01
4.17166902e-03 -1.84287235e-01 -6.83684349e-01 -3.18477273e-01
4.23494697e-01 4.80955780e-01 4.08854812e-01 -5.07626414e-01
-5.86871445e-01 -3.68580461e-01 8.87481272e-02 2.01208219e-02
-3.31669986e-01 5.62867284e-01 -3.02811921e-01 -1.04979420e+00
-1.92257300e-01 -8.32474113e-01 -8.29722226e-01 -2.94204772e-01
-4.91673768e-01 -2.56669462e-01 6.47385642e-02 -8.05930018e-01
1.26859760e+00 -1.87497663e+00 2.42671207e-01 1.92954019e-01
2.91901648e-01 2.31299445e-01 -2.69527465e-01 1.52599767e-01
-7.63088524e-01 3.41016859e-01 -6.15620136e-01 -3.21666658e-01
-1.70494273e-01 3.60792577e-01 -7.82120526e-02 3.50159168e-01
2.82533735e-01 1.04507434e+00 -1.11446762e+00 -2.60847241e-01
-2.77452916e-01 6.03661716e-01 -5.74451983e-01 1.38451174e-01
-2.27677852e-01 2.56442130e-01 -4.55667019e-01 4.93315816e-01
2.88459718e-01 -6.42950416e-01 3.96811843e-01 -5.49673848e-02
4.17310625e-01 6.75252140e-01 -6.10585928e-01 2.32448983e+00
-6.91079140e-01 9.48814154e-02 -1.01559512e-01 -1.25750351e+00
5.40461361e-01 6.79243445e-01 6.48534358e-01 -5.31647682e-01
5.96519947e-01 7.31098205e-02 1.28392167e-02 -4.66827214e-01
3.68846774e-01 -4.52095866e-01 -2.97487527e-01 2.75309801e-01
6.55600846e-01 3.78114283e-01 3.11699361e-01 9.78243351e-02
1.30186856e+00 3.78829017e-02 1.01022387e+00 -3.28318030e-01
6.95905507e-01 -2.13598892e-01 5.84907293e-01 7.93347597e-01
3.04059744e-01 6.24768138e-01 3.68230343e-01 -5.11914909e-01
-7.13623822e-01 -7.01956153e-01 -2.40045890e-01 1.48955989e+00
-3.97332937e-01 -5.85594475e-01 -8.96389186e-01 -7.91277528e-01
-1.04023524e-01 5.34112513e-01 -1.12804937e+00 -5.07746339e-01
-4.78259981e-01 -1.31287408e+00 5.31090498e-01 6.27504945e-01
-2.03487113e-01 -1.07132387e+00 -3.70301247e-01 6.12229824e-01
-1.16813190e-01 -1.21963286e+00 -6.23984098e-01 4.53775406e-01
-5.57747543e-01 -1.14199722e+00 -8.51620913e-01 -9.57729518e-01
6.65247262e-01 -3.99880111e-01 1.53315258e+00 1.29078582e-01
-6.24398172e-01 2.77379841e-01 -3.92503649e-01 -7.66637683e-01
-7.12218404e-01 7.05840170e-01 -1.93265855e-01 1.48435321e-03
8.72934222e-01 -2.49812469e-01 -3.39335203e-01 -2.27518827e-01
-1.27357268e+00 6.92134500e-02 3.77804041e-01 1.02901852e+00
5.57007074e-01 -6.79155231e-01 6.23050034e-01 -1.47667503e+00
7.47772574e-01 -7.12590873e-01 -1.72450364e-01 4.62902576e-01
-4.93994087e-01 2.67388701e-01 3.12882990e-01 -2.93986619e-01
-9.14987266e-01 1.01826563e-01 -6.18972778e-01 2.82256324e-02
-3.17301363e-01 8.18521202e-01 -5.27313128e-02 1.81418329e-01
9.81071770e-01 -2.01245442e-01 -2.33782485e-01 -5.87258101e-01
6.19869709e-01 7.71578193e-01 6.86736882e-01 -2.21678615e-01
1.61862612e-01 2.93954462e-01 -7.58988336e-02 -5.79377294e-01
-1.45659876e+00 -8.60571384e-01 -7.00255215e-01 5.21795571e-01
1.41073072e+00 -9.72219408e-01 -6.06914997e-01 -5.75502329e-02
-1.33059347e+00 1.22045405e-01 -3.29907805e-01 4.47496861e-01
-2.99209654e-01 -1.32861249e-02 -7.86572278e-01 -2.96665609e-01
-9.28642392e-01 -7.91431010e-01 1.13030648e+00 -1.90769378e-02
-6.91430151e-01 -1.28041124e+00 4.32744831e-01 2.18755499e-01
4.63733613e-01 3.91190261e-01 1.14576805e+00 -1.27628553e+00
2.65421391e-01 -4.22835171e-01 -2.08340764e-01 3.21627915e-01
4.88012463e-01 -5.42196572e-01 -8.34529877e-01 -2.97502786e-01
2.08826121e-02 -2.24273875e-01 1.26140618e+00 3.09221268e-01
1.06768608e+00 -3.65631968e-01 -5.73395967e-01 7.61153281e-01
1.29439998e+00 9.13033485e-02 4.34479356e-01 4.09599423e-01
8.77762377e-01 5.41333199e-01 1.03450544e-01 6.04832694e-02
1.83389276e-01 4.45172727e-01 8.14557299e-02 -5.90494454e-01
4.23773366e-04 1.19375870e-01 -1.17655136e-01 8.39940190e-01
1.30292743e-01 -3.19944292e-01 -1.16069925e+00 6.40288591e-01
-1.88296735e+00 -5.25505543e-01 1.89710259e-01 1.93764126e+00
1.21863246e+00 -8.28222781e-02 -3.61334741e-01 -2.82085538e-01
7.13864148e-01 -1.17594875e-01 -3.96479189e-01 -5.27457178e-01
-6.01280332e-02 7.57892013e-01 7.00067461e-01 3.46390814e-01
-1.52510571e+00 8.75414312e-01 5.79755354e+00 8.61470580e-01
-1.08632112e+00 5.27065575e-01 7.83192754e-01 -1.12812206e-01
1.43716782e-01 -5.94959199e-01 -7.58115768e-01 1.86583146e-01
1.34311926e+00 -1.79879174e-01 -1.59897804e-01 6.70039952e-01
-3.62103820e-01 4.82498258e-01 -1.32605326e+00 9.00651336e-01
4.12507355e-01 -1.40472567e+00 3.95356089e-01 -3.40364784e-01
6.59584820e-01 4.38938439e-01 -2.34307602e-01 2.77473658e-01
6.71558857e-01 -1.39280915e+00 1.12030409e-01 5.47062278e-01
9.74418223e-01 -6.14705741e-01 1.31233263e+00 -1.77109241e-01
-9.46959257e-01 2.76289850e-01 -1.56218886e-01 3.21369320e-01
8.77397060e-02 2.15359494e-01 -9.07654583e-01 8.55455160e-01
5.53781450e-01 1.16362154e+00 -4.13294256e-01 6.62667871e-01
-1.24366380e-01 4.01902556e-01 -3.39642353e-02 1.51191697e-01
4.25803214e-01 1.03534766e-01 1.58847675e-01 1.85944760e+00
1.89076275e-01 4.36931057e-03 3.88219953e-02 6.53896511e-01
-7.21644998e-01 7.75618792e-01 -4.31241632e-01 -5.57089113e-02
-3.44213814e-01 1.27412343e+00 -8.19624543e-01 -5.94682693e-01
-6.64647222e-01 7.85751104e-01 2.62017310e-01 3.40850860e-01
-3.86934847e-01 -5.30907393e-01 9.49016452e-01 -2.40366254e-02
-4.52704169e-03 7.57214904e-01 -4.43314798e-02 -1.36227310e+00
-3.31791848e-01 -7.80753076e-01 9.53833163e-01 -5.20180702e-01
-1.72242081e+00 8.43743205e-01 -3.51687223e-01 -1.00122070e+00
-4.07617182e-01 -9.64838624e-01 -1.43413410e-01 8.41543317e-01
-1.48710346e+00 -1.19552314e+00 1.96141481e-01 5.77834129e-01
3.18481237e-01 -2.91051954e-01 1.50423133e+00 6.90135002e-01
-5.21726489e-01 5.33603966e-01 1.12292647e-01 5.47958672e-01
1.03175306e+00 -1.22720051e+00 1.00476444e+00 5.59797168e-01
-8.38736165e-03 9.16264534e-01 5.19072890e-01 -6.28403604e-01
-5.32507002e-01 -1.60802388e+00 1.51584160e+00 -5.30212224e-01
8.13889325e-01 -5.64716995e-01 -1.18750858e+00 9.35670137e-01
1.58415183e-01 1.03854120e-01 1.25525391e+00 2.59107172e-01
-5.34560025e-01 3.02749038e-01 -9.59855437e-01 5.21681070e-01
1.06196666e+00 -6.14408910e-01 -9.98128831e-01 6.78222179e-01
1.16976392e+00 -5.29507875e-01 -1.00917137e+00 4.44461137e-01
1.87138513e-01 1.59289539e-01 1.24375641e+00 -1.66219079e+00
5.66190839e-01 1.18778422e-01 -9.72523391e-02 -1.27138698e+00
-2.75439441e-01 -5.22731304e-01 3.26191157e-01 7.42691457e-01
7.18470037e-01 -5.06221473e-01 4.97301459e-01 4.19780880e-01
-3.91159207e-02 -6.16521299e-01 -1.05450058e+00 -3.46046746e-01
4.27237988e-01 -3.37281644e-01 4.64255452e-01 1.49474502e+00
2.07633868e-01 8.45580816e-01 -3.81683081e-01 -2.12802827e-01
1.22819267e-01 -2.38834143e-01 2.52300706e-02 -1.47978830e+00
-6.57404736e-02 -5.21703959e-01 -6.08903766e-01 -4.18812484e-01
5.34951448e-01 -1.45225918e+00 -7.91571885e-02 -1.63781023e+00
2.70800769e-01 1.82687238e-01 -8.57355237e-01 7.49534905e-01
-3.04430008e-01 2.52718925e-01 -2.27642611e-01 -4.53544140e-01
-7.37457812e-01 1.53837785e-01 6.43518388e-01 -4.76937294e-01
-4.94457446e-02 -1.53548688e-01 -1.04494584e+00 9.66597199e-01
6.45934939e-01 -1.10774004e+00 -7.61258900e-02 -4.66673940e-01
6.48743391e-01 -5.46189733e-02 -1.42669678e-05 -4.54610258e-01
2.95785606e-01 1.85953602e-01 7.44614378e-02 -7.55205974e-02
3.75517309e-02 -6.49629533e-01 -1.81697309e-01 6.56567514e-01
-8.53668749e-01 -2.64763609e-02 2.63407111e-01 5.48930407e-01
-2.74933398e-01 -3.01650792e-01 6.62241340e-01 -3.63686711e-01
-2.71492243e-01 2.31570244e-01 -5.46811521e-01 3.97942543e-01
4.23385531e-01 4.64348495e-01 -1.09262392e-01 -5.86719289e-02
-1.18660843e+00 1.63528115e-01 -5.09316735e-02 8.12917650e-01
2.57825464e-01 -1.26456106e+00 -1.02029729e+00 -4.62556332e-02
5.29453933e-01 -1.84529901e-01 8.00376087e-02 6.27445698e-01
-2.81158924e-01 6.79849148e-01 -4.86872643e-02 -2.13331610e-01
-1.37811327e+00 7.06759274e-01 6.96279287e-01 -7.46320903e-01
-5.14003098e-01 1.01343346e+00 5.21945000e-01 -9.31756496e-01
3.35404605e-01 -8.39765429e-01 -6.02897167e-01 2.94026852e-01
7.24290013e-01 -1.66057646e-01 7.31421947e-01 -8.14569771e-01
-4.94265735e-01 4.64390248e-01 -3.65733027e-01 2.33414978e-01
1.42955756e+00 2.34265938e-01 -5.74108005e-01 2.92222977e-01
1.67397392e+00 -5.88554621e-01 4.53661382e-02 -7.17043459e-01
5.14756382e-01 3.38960171e-01 2.13129120e-03 -8.18565845e-01
-1.04445708e+00 5.57130933e-01 6.80661321e-01 -2.98907608e-01
8.93025756e-01 8.62208754e-02 7.22260892e-01 9.18267488e-01
-1.75425619e-01 -1.02629304e+00 -4.12097335e-01 7.22762287e-01
7.80094981e-01 -1.19853842e+00 -1.64217591e-01 -3.58180851e-01
-3.60263109e-01 1.04321194e+00 1.12814613e-01 1.51393553e-02
9.13415015e-01 4.66786444e-01 3.90096009e-01 -2.60988712e-01
-7.13473856e-01 -3.08764160e-01 5.10814726e-01 3.16623330e-01
1.03787506e+00 -1.01658061e-01 -4.15507108e-01 1.03362668e+00
-1.20306700e-01 2.17694044e-01 4.51094478e-01 8.44317436e-01
1.54589429e-01 -1.33479035e+00 -4.11960222e-02 7.31803000e-01
-1.37578213e+00 -6.02278948e-01 -2.44630426e-01 3.93255174e-01
3.09315085e-01 6.95127130e-01 2.01909512e-01 9.75574180e-02
5.57881534e-01 3.11150998e-01 7.50187337e-02 -1.28103983e+00
-1.18733227e+00 2.55717300e-02 1.60033837e-01 -6.50304377e-01
-7.12815166e-01 -3.82455885e-01 -1.38808072e+00 4.12427843e-01
-3.84197272e-02 7.73207769e-02 5.69446385e-01 1.06332409e+00
2.36365229e-01 1.14274991e+00 -2.44109839e-01 -9.62809771e-02
-9.55166593e-02 -1.06838679e+00 -2.42215887e-01 8.42531562e-01
5.98315537e-01 -2.98617512e-01 8.43565837e-02 3.00288707e-01] | [8.517056465148926, 8.68575668334961] |
ec153fe1-5ba7-486c-8fb1-a2e34fd71550 | comparision-of-adversarial-and-non | 2211.00731 | null | https://arxiv.org/abs/2211.00731v1 | https://arxiv.org/pdf/2211.00731v1.pdf | Comparision Of Adversarial And Non-Adversarial LSTM Music Generative Models | Algorithmic music composition is a way of composing musical pieces with minimal to no human intervention. While recurrent neural networks are traditionally applied to many sequence-to-sequence prediction tasks, including successful implementations of music composition, their standard supervised learning approach based on input-to-output mapping leads to a lack of note variety. These models can therefore be seen as potentially unsuitable for tasks such as music generation. Generative adversarial networks learn the generative distribution of data and lead to varied samples. This work implements and compares adversarial and non-adversarial training of recurrent neural network music composers on MIDI data. The resulting music samples are evaluated by human listeners, their preferences recorded. The evaluation indicates that adversarial training produces more aesthetically pleasing music. | ['Johan Pieter de Villiers', 'Anna Sergeevna Bosman', "Moseli Mots'oehli"] | 2022-11-01 | null | null | null | null | ['music-generation', 'music-generation'] | ['audio', 'music'] | [ 5.22248805e-01 2.30553299e-01 2.87953824e-01 -8.18413496e-02
-6.49452150e-01 -9.06887710e-01 5.30195355e-01 -6.16451144e-01
-6.81299493e-02 8.05663884e-01 3.95393759e-01 2.09677871e-02
-9.17593092e-02 -7.80472934e-01 -5.87540150e-01 -7.37453043e-01
1.35374188e-01 6.28622472e-01 -2.79379785e-01 -6.36531651e-01
3.86479944e-02 4.13231671e-01 -1.39033067e+00 6.71013713e-01
3.92253995e-01 7.21002519e-01 -2.40166619e-01 1.02686942e+00
-8.44602287e-02 9.28771555e-01 -1.21743381e+00 -7.28035748e-01
5.28277695e-01 -1.05417073e+00 -3.74460101e-01 -4.02731568e-01
3.50721329e-01 1.23243723e-02 -1.57138750e-01 9.99781609e-01
8.90091658e-01 3.03523690e-01 9.04527605e-01 -9.22837436e-01
-8.57071400e-01 1.24593079e+00 1.25805020e-01 -3.06427866e-01
5.38913727e-01 1.72868073e-01 1.10774958e+00 -6.43523872e-01
7.41719127e-01 1.07106519e+00 9.48693395e-01 8.41860056e-01
-1.72130120e+00 -8.55960190e-01 -5.36031246e-01 -2.13763993e-02
-1.14458716e+00 -5.10632932e-01 1.27798724e+00 -2.66481161e-01
7.02742398e-01 6.81650937e-01 9.41400290e-01 1.56796002e+00
7.45297372e-02 7.71091998e-01 1.05955315e+00 -5.27132928e-01
2.65814483e-01 -7.58777708e-02 -8.55051816e-01 -4.14044224e-02
-4.70341802e-01 5.81328690e-01 -6.90603077e-01 -1.13153912e-01
9.18888509e-01 -2.96298087e-01 -4.85199615e-02 -8.69072825e-02
-1.09997594e+00 8.32411945e-01 5.16140401e-01 4.04043943e-01
-5.89502811e-01 3.37222159e-01 6.07928932e-01 8.90270293e-01
9.30299461e-02 1.17646396e+00 -1.75172418e-01 -3.84854734e-01
-1.13526464e+00 7.52243221e-01 8.68765950e-01 6.67375743e-01
1.08161926e-01 9.81358409e-01 -2.11769417e-01 1.11494207e+00
-4.37571928e-02 4.59684461e-01 8.73427927e-01 -9.49702084e-01
-1.03026673e-01 1.91188276e-01 -1.98437229e-01 -8.53257418e-01
-5.02164476e-02 -5.65708280e-01 -9.96554017e-01 1.05753016e+00
4.43998665e-01 -2.30404228e-01 -7.45240927e-01 1.56731260e+00
-2.86088049e-01 4.30426784e-02 1.76808566e-01 8.27935815e-01
7.92297304e-01 5.85936308e-01 -1.47703081e-01 -4.49487977e-02
6.35585248e-01 -6.40531361e-01 -6.74613655e-01 8.54503587e-02
-3.92048866e-01 -1.33724916e+00 1.32971561e+00 8.09651792e-01
-1.41923356e+00 -7.28822291e-01 -9.99908268e-01 2.05643162e-01
-5.39384596e-02 -2.68727899e-01 4.19345558e-01 8.94754469e-01
-8.36145461e-01 1.16714621e+00 -3.28684032e-01 2.36597121e-01
4.38552558e-01 3.48681659e-01 -2.78756972e-02 6.19380295e-01
-1.04158151e+00 7.59910762e-01 5.05758584e-01 4.59026992e-02
-8.68317544e-01 -7.66075611e-01 -2.30560735e-01 -9.46259871e-02
-4.92128879e-02 -7.50094175e-01 1.60706532e+00 -1.71589184e+00
-2.17982364e+00 6.70790255e-01 5.14231980e-01 -7.35782683e-01
7.20796645e-01 -2.27857172e-01 -5.26349127e-01 -4.19104367e-01
-3.96661252e-01 4.80725527e-01 1.23848844e+00 -1.35844231e+00
-1.35497645e-01 1.86465681e-01 -2.40434304e-01 6.57389536e-02
-8.06120783e-03 -8.26250240e-02 2.62665451e-01 -1.51211059e+00
-7.53799826e-03 -1.05876935e+00 -1.52028009e-01 -3.81978840e-01
-6.73246145e-01 2.51579434e-01 4.50760394e-01 -5.82863033e-01
1.00721967e+00 -1.92756224e+00 2.92526811e-01 4.16352719e-01
-2.96753049e-01 3.24004173e-01 -1.52704746e-01 5.97393334e-01
-5.10075629e-01 2.73411851e-02 -1.96934827e-02 -5.33104874e-02
2.75082201e-01 -9.37217753e-03 -7.36394882e-01 -1.88911166e-02
-8.90015066e-02 1.25408125e+00 -7.32235610e-01 1.97981652e-02
1.29327416e-01 6.02287948e-01 -6.16619706e-01 4.50200558e-01
-5.62246323e-01 6.25735641e-01 9.65060014e-03 5.29558003e-01
-5.34015670e-02 4.91294295e-01 2.75180768e-02 -1.43150270e-01
3.07662725e-01 3.02806467e-01 -9.77969229e-01 1.90928149e+00
-5.11434793e-01 8.20610225e-01 -4.65385348e-01 -6.26124144e-01
1.36572766e+00 6.40005469e-01 2.65163779e-01 -5.47388375e-01
2.35601664e-01 3.21184576e-01 4.56418425e-01 -1.48123965e-01
5.64033866e-01 -1.00040889e+00 -2.02583641e-01 7.28426576e-01
-6.63011670e-02 -6.35129213e-01 -2.73774713e-01 -2.91928619e-01
9.70570266e-01 5.80828965e-01 1.61385313e-01 1.05038725e-01
9.58824828e-02 -5.88785149e-02 3.25805098e-01 5.81279576e-01
3.30059588e-01 1.02978468e+00 2.28666395e-01 -6.33050084e-01
-1.65622568e+00 -1.35544884e+00 4.06462103e-01 1.13600850e+00
-5.08880436e-01 -1.33392870e-01 -8.41074169e-01 -9.30744782e-02
-2.43206307e-01 1.05832529e+00 -7.14246213e-01 -3.28093797e-01
-7.31671393e-01 -1.98593244e-01 1.06479454e+00 3.55236083e-01
-9.24772322e-02 -2.10135484e+00 -3.80117089e-01 5.66170275e-01
7.55973831e-02 -2.01401547e-01 -5.64741433e-01 1.88774511e-01
-8.50997448e-01 -6.76633596e-01 -9.17777359e-01 -7.20571578e-01
1.26754180e-01 -5.75062931e-01 1.53639066e+00 -2.84830809e-01
-4.22745973e-01 6.42530695e-02 -4.55039412e-01 -9.04333889e-01
-1.24702954e+00 1.44662410e-01 1.18996881e-01 -1.22866690e-01
8.55602995e-02 -1.30495119e+00 -5.89794099e-01 1.60251573e-01
-1.01092970e+00 8.40515941e-02 6.30400896e-01 1.02286601e+00
6.57119930e-01 -2.16319308e-01 8.58485341e-01 -1.09341729e+00
1.22197938e+00 1.18369371e-01 -3.72438552e-03 -2.51531582e-02
-4.58866447e-01 1.03749953e-01 1.00905037e+00 -1.08650827e+00
-7.70307600e-01 5.94511554e-02 -4.21684563e-01 -8.56990159e-01
-1.11862637e-01 2.33825937e-01 -1.01821490e-01 1.10383444e-01
1.32601297e+00 2.12085292e-01 7.35920796e-04 -3.54623705e-01
5.60738623e-01 3.40180844e-01 9.84535038e-01 -5.50461829e-01
1.26317632e+00 -8.92829448e-02 -1.56470850e-01 -4.71561044e-01
-3.94763350e-01 3.09468269e-01 -4.92774397e-01 -4.41051871e-01
5.43295979e-01 -3.53783727e-01 -7.80723333e-01 2.61674792e-01
-8.71086657e-01 -5.74193299e-01 -1.21798635e+00 1.72471657e-01
-1.15092576e+00 -2.46300623e-01 -5.03430605e-01 -8.90455127e-01
-7.33982682e-01 -7.20568419e-01 5.03852129e-01 9.96729583e-02
-9.61246610e-01 -7.10399270e-01 4.50408578e-01 -4.55771312e-02
6.12366498e-01 7.66931415e-01 9.66267347e-01 -6.29030347e-01
-2.84145266e-01 -3.88550609e-01 6.20185256e-01 5.12872636e-01
2.35637039e-01 1.36318058e-01 -1.23536575e+00 1.50664169e-02
-2.17912287e-01 -4.73276407e-01 4.55886841e-01 2.63839483e-01
8.25833380e-01 -5.33229411e-01 4.87385273e-01 7.26443768e-01
1.19760585e+00 4.77583319e-01 1.10434413e+00 3.11865181e-01
7.02572644e-01 3.76372039e-01 1.66627988e-01 2.69030750e-01
-7.35892594e-01 7.66307294e-01 3.16437006e-01 2.83532459e-02
-4.73930508e-01 -8.18367243e-01 5.79442739e-01 8.43431056e-01
-6.13903046e-01 -9.36398953e-02 -5.62182546e-01 3.10344934e-01
-1.43346310e+00 -1.58356786e+00 3.14959228e-01 2.06898952e+00
9.66070533e-01 1.58898681e-01 5.25287092e-01 8.44026625e-01
5.09640813e-01 1.99353858e-03 -5.92994153e-01 -8.88268292e-01
-3.12135637e-01 9.98706758e-01 1.68371648e-01 1.19040169e-01
-7.79997766e-01 9.07260299e-01 6.88520050e+00 9.58380580e-01
-1.16835356e+00 -1.35668695e-01 3.37078303e-01 -5.68126559e-01
-6.29278421e-01 -2.65319884e-01 1.03824988e-01 3.67375165e-01
8.92183304e-01 -2.73781925e-01 1.14554596e+00 6.98371530e-01
4.14636247e-02 7.55304933e-01 -1.21694601e+00 1.16780913e+00
-5.61734177e-02 -1.53741360e+00 4.59632814e-01 -1.45613760e-01
9.63962972e-01 -3.74594063e-01 4.34829354e-01 3.47665191e-01
4.50940162e-01 -1.85253441e+00 1.22650385e+00 8.56384277e-01
1.05344152e+00 -1.11862016e+00 4.75635707e-01 8.67463946e-02
-7.51228511e-01 4.41075265e-02 -6.32735118e-02 -1.13030009e-01
1.48661107e-01 8.51950888e-03 -9.76231933e-01 2.09894106e-01
3.50017548e-01 3.70776981e-01 -2.29224205e-01 9.15482044e-01
-3.32426906e-01 9.61808145e-01 8.06582496e-02 -2.84540653e-01
-6.41238764e-02 -2.56820679e-01 9.01859999e-01 1.06482029e+00
5.02463520e-01 -1.79353848e-01 -1.81291416e-01 1.15687609e+00
-1.62072733e-01 3.07029426e-01 -7.97387064e-01 -3.52571487e-01
3.11451018e-01 7.46993661e-01 -5.17530918e-01 6.36227056e-02
3.59615862e-01 1.28265095e+00 -1.28162429e-01 3.58259499e-01
-5.63077450e-01 -3.97675008e-01 6.45785928e-01 3.05236757e-01
1.78036347e-01 1.12799183e-01 -7.08162606e-01 -5.78060508e-01
-3.25547248e-01 -1.50351095e+00 -8.73209815e-03 -1.04866755e+00
-1.56057072e+00 9.18034434e-01 -5.83153129e-01 -1.48033452e+00
-9.51882243e-01 -3.75192195e-01 -1.00494099e+00 1.10968101e+00
-4.44387764e-01 -1.28759539e+00 7.31989890e-02 3.93513709e-01
5.56788862e-01 -1.02923524e+00 1.44409513e+00 1.20310485e-01
1.85445473e-01 8.80670607e-01 1.95107788e-01 1.99408725e-01
6.54537857e-01 -1.37180626e+00 6.74649954e-01 3.19102317e-01
9.38011110e-01 4.41854417e-01 1.28325522e+00 -2.58494765e-01
-1.10892844e+00 -9.05947268e-01 4.65079337e-01 -3.78105164e-01
4.95177120e-01 -2.12630391e-01 -8.26681197e-01 3.48470390e-01
4.28463668e-01 -6.10228956e-01 9.97333229e-01 1.22684665e-01
-3.99501920e-01 -1.79348662e-02 -9.83798623e-01 8.77497673e-01
9.26978707e-01 -7.32794583e-01 -7.70529151e-01 -3.63604389e-02
4.60194767e-01 -4.17563796e-01 -1.02079940e+00 1.09280199e-01
1.07582104e+00 -1.10483074e+00 1.23815501e+00 -8.22107971e-01
9.16975021e-01 -2.78849095e-01 -8.02561343e-02 -1.64635801e+00
-3.56242567e-01 -1.23354602e+00 1.94394186e-01 1.26265264e+00
6.16919875e-01 -4.05116118e-02 9.41909313e-01 1.93657488e-01
-5.69002703e-02 -4.21604842e-01 -8.13411772e-01 -7.52451897e-01
2.40209118e-01 -5.63940108e-01 7.90330112e-01 8.69180024e-01
-3.61423016e-01 3.68643135e-01 -7.61688828e-01 -5.30118823e-01
2.66262203e-01 3.01123559e-01 1.12802768e+00 -1.19233787e+00
-8.76502752e-01 -9.34110940e-01 -5.09684980e-01 -1.70452163e-01
1.39893815e-01 -1.13932955e+00 1.48578703e-01 -1.00899279e+00
-2.50518292e-01 -4.36416119e-01 -3.59191090e-01 3.15490842e-01
1.87287241e-01 9.13249135e-01 4.43263769e-01 1.48874030e-01
7.27520138e-02 5.30395746e-01 1.34721637e+00 -3.26376498e-01
-5.58381855e-01 4.33152050e-01 -5.70074201e-01 6.43061161e-01
9.60859179e-01 -5.86066663e-01 -4.03873771e-01 5.51550277e-02
5.68324387e-01 -3.08543202e-02 3.44106704e-01 -1.11187899e+00
-2.10047841e-01 -2.41217390e-02 6.30077183e-01 -1.47260860e-01
4.56933588e-01 -5.97814798e-01 1.06552029e+00 4.51612204e-01
-7.79423714e-01 -1.32509306e-01 2.80149460e-01 2.50687689e-01
-3.43578607e-01 -4.94242579e-01 7.19700098e-01 -3.28989506e-01
-2.65829742e-01 3.67739834e-02 -2.03423172e-01 6.44215867e-02
5.45277178e-01 -5.26961803e-01 3.53805721e-01 -7.66850948e-01
-1.23104453e+00 -9.14210618e-01 5.24158418e-01 5.18695474e-01
5.03980041e-01 -1.76142228e+00 -1.06688261e+00 1.64673209e-01
-1.63468778e-01 -3.07533413e-01 1.21618927e-01 -6.18091747e-02
-6.03287220e-01 -1.06538229e-01 -6.83304965e-01 -1.46352291e-01
-1.45368671e+00 5.10625660e-01 5.24542809e-01 -8.94676894e-03
-8.14159393e-01 8.88080955e-01 -2.24840179e-01 -5.30544579e-01
2.13431850e-01 1.06227271e-01 -1.50816422e-02 3.04797087e-02
3.39989722e-01 2.95163244e-01 -1.94145352e-01 -5.18910408e-01
2.60426551e-01 3.03466171e-01 3.58073264e-01 -6.03443503e-01
1.52506483e+00 6.52517855e-01 1.20629042e-01 8.82493198e-01
7.41768062e-01 2.98199326e-01 -1.03744805e+00 7.04016909e-02
-9.46795195e-02 -3.84488344e-01 -4.15225267e-01 -1.15529227e+00
-9.93434966e-01 7.52764642e-01 6.83934927e-01 2.76648611e-01
1.31919670e+00 -4.93350297e-01 7.05846190e-01 3.86601686e-01
3.35775912e-01 -8.99473190e-01 1.98691383e-01 3.46492678e-01
1.48529863e+00 -6.42886579e-01 -2.21729755e-01 2.71532565e-01
-7.34086752e-01 1.13663435e+00 -1.11197317e-02 -6.64909363e-01
3.30412954e-01 4.97702241e-01 4.65533823e-01 3.10317576e-01
-5.58216214e-01 1.91407114e-01 6.07810080e-01 9.11358178e-01
7.40171134e-01 2.78345108e-01 2.64589358e-02 7.77598977e-01
-1.22367907e+00 1.26786068e-01 2.34533519e-01 3.74196142e-01
-4.60563600e-03 -1.59932184e+00 -5.74137390e-01 3.39180797e-01
-6.02174997e-01 -1.36344254e-01 -8.23799014e-01 5.43799460e-01
3.28475595e-01 5.43499351e-01 -5.94033897e-02 -7.94815242e-01
4.91527438e-01 3.45447630e-01 7.28502214e-01 -4.29597855e-01
-1.43768692e+00 2.04450846e-01 1.36567980e-01 -2.01217726e-01
-2.89004743e-01 -6.40163302e-01 -9.85875189e-01 -4.37975943e-01
8.38383213e-02 7.92356730e-02 6.62057519e-01 3.55126947e-01
-8.61195996e-02 1.07146132e+00 5.53651512e-01 -1.14099813e+00
-6.22175217e-01 -1.27032983e+00 -7.58212268e-01 7.92024195e-01
4.14880291e-02 -4.01460938e-02 7.27809966e-02 4.91394907e-01] | [16.01044464111328, 5.5565009117126465] |
0a34c4fb-8015-4379-80c9-f342db3f70d8 | temporally-coherent-video-anonymization | 2106.02328 | null | https://arxiv.org/abs/2106.02328v1 | https://arxiv.org/pdf/2106.02328v1.pdf | Temporally coherent video anonymization through GAN inpainting | This work tackles the problem of temporally coherent face anonymization in natural video streams.We propose JaGAN, a two-stage system starting with detecting and masking out faces with black image patches in all individual frames of the video. The second stage leverages a privacy-preserving Video Generative Adversarial Network designed to inpaint the missing image patches with artificially generated faces. Our initial experiments reveal that image based generative models are not capable of inpainting patches showing temporal coherent appearance across neighboring video frames. To address this issue we introduce a newly curated video collection, which is made publicly available for the research community along with this paper. We also introduce the Identity Invariance Score IdI as a means to quantify temporal coherency between neighboring frames. | ['Torsten Schön', 'Marcel Wasserer', 'Raphael Mitsch', 'Georg Göri', 'Patrick Blies', 'Thangapavithraa Balaji'] | 2021-06-04 | null | null | null | null | ['face-anonymization'] | ['computer-vision'] | [ 5.82132339e-01 1.00044094e-01 4.27669212e-02 -2.17766941e-01
-6.07510507e-01 -8.89331162e-01 5.21500051e-01 -5.97576559e-01
-1.59283295e-01 7.56512523e-01 2.37333074e-01 -4.65754420e-02
2.66736627e-01 -4.81814593e-01 -9.40562904e-01 -6.68568432e-01
-3.26353490e-01 -2.31146991e-01 -2.42130250e-01 1.67865261e-01
-5.43456860e-02 5.30666173e-01 -1.26212609e+00 5.81490457e-01
2.90137529e-01 8.69811893e-01 -5.53237140e-01 7.46683598e-01
5.50841391e-01 9.69256282e-01 -6.86283112e-01 -7.71777511e-01
9.67670739e-01 -9.16283131e-01 -5.79161346e-01 4.48215306e-01
8.83955061e-01 -7.87946522e-01 -7.31737316e-01 1.09383881e+00
2.23732144e-01 1.64850280e-02 2.21548557e-01 -1.82649446e+00
-7.15972722e-01 3.58061016e-01 -6.14929318e-01 3.05307090e-01
5.03890455e-01 4.34125453e-01 6.39646888e-01 -7.45499611e-01
9.95451272e-01 1.26114416e+00 6.84113801e-01 8.22661400e-01
-1.34187162e+00 -7.99284816e-01 -1.97068211e-02 -3.93027738e-02
-1.66566432e+00 -9.47126031e-01 9.35021937e-01 -3.38753164e-01
4.06835973e-01 5.45785487e-01 5.83167791e-01 1.33928406e+00
1.13910839e-01 5.15752494e-01 1.15013885e+00 -7.05930367e-02
1.22700274e-01 -1.43617079e-01 -5.24047971e-01 5.79447806e-01
1.84363753e-01 3.79060060e-01 -7.61548221e-01 -5.30952513e-01
1.07527685e+00 2.08267078e-01 -4.24824029e-01 -3.76346827e-01
-8.82964075e-01 8.46463203e-01 5.20920418e-02 6.96420670e-02
-4.28049982e-01 2.96898782e-01 1.32432953e-01 5.84781647e-01
5.91287017e-01 1.77265957e-01 -3.90670449e-02 2.40534142e-01
-1.26035321e+00 4.69340175e-01 6.47298396e-01 8.97110581e-01
7.34329283e-01 2.53001839e-01 -2.31422931e-01 1.80175453e-01
-4.87629138e-02 3.55251551e-01 1.14643186e-01 -1.51695478e+00
8.80127996e-02 -6.59388024e-03 5.31251356e-02 -1.33497560e+00
3.93964589e-01 1.68299884e-01 -6.99903250e-01 2.14601755e-01
2.69898713e-01 -3.41675520e-01 -8.12174141e-01 2.05891752e+00
3.52138251e-01 6.12280607e-01 1.19665772e-01 7.75967598e-01
5.52570522e-01 4.09615755e-01 -1.77296754e-02 -5.06833613e-01
1.23678792e+00 -8.09495151e-01 -7.63593316e-01 -2.12589372e-03
-1.73207819e-01 -7.41538644e-01 5.15042007e-01 8.48567113e-02
-1.24535418e+00 -2.76547879e-01 -8.82515728e-01 2.04012804e-02
-9.90663320e-02 -2.69887388e-01 5.03253400e-01 9.18071032e-01
-1.45915675e+00 4.03502494e-01 -8.65685403e-01 -2.39093706e-01
9.36885655e-01 3.76592666e-01 -9.06787872e-01 -9.34257805e-02
-9.70113277e-01 2.30882689e-01 4.76419518e-04 -1.98774755e-01
-1.12547791e+00 -8.50759149e-01 -9.64007258e-01 -1.46691576e-01
2.79367477e-01 -7.33579755e-01 9.58241522e-01 -1.77425742e+00
-1.23174906e+00 1.15814662e+00 -3.70807707e-01 -6.95820570e-01
8.88030469e-01 -4.31768000e-02 -5.42574525e-01 5.84805667e-01
2.47738376e-01 7.72929549e-01 1.64173341e+00 -1.44366276e+00
-3.52862477e-01 -3.81875217e-01 -7.29885250e-02 -1.22380555e-01
-8.22282583e-02 3.00734758e-01 -5.78827500e-01 -1.31787336e+00
-3.01906407e-01 -9.79954839e-01 -4.57291268e-02 6.78419247e-02
-4.66554165e-01 6.17838800e-01 1.37892222e+00 -8.65211666e-01
9.56061780e-01 -2.23304367e+00 -7.66652077e-02 2.53863662e-01
4.37436461e-01 1.21614531e-01 -2.68667221e-01 2.94293404e-01
-2.95445323e-01 3.14069390e-01 -3.71569633e-01 -6.22269928e-01
-3.15140396e-01 1.06338665e-01 -6.57153606e-01 7.37090707e-01
4.22716290e-01 8.76384735e-01 -7.53552198e-01 -2.59347200e-01
-6.92538992e-02 7.55138278e-01 -7.85273433e-01 2.61203945e-01
-1.34509012e-01 8.63127828e-01 -1.64246857e-01 8.68754566e-01
1.12544751e+00 8.63695592e-02 2.39847407e-01 5.84711805e-02
1.45958990e-01 -2.14235708e-01 -7.88193464e-01 1.53932738e+00
2.64185578e-01 8.10323238e-01 4.76650715e-01 -7.00409770e-01
3.64905328e-01 4.87889946e-01 7.94734538e-01 -1.68973088e-01
2.52788574e-01 -2.17467770e-01 -3.66762698e-01 -4.19309735e-01
5.30935287e-01 -5.25760241e-02 -5.47860041e-02 6.97434425e-01
-4.07703705e-02 1.85254470e-01 -1.01348855e-01 2.64398694e-01
1.36005843e+00 -1.26468958e-02 6.66626096e-02 4.95127290e-02
2.20572859e-01 -2.98542291e-01 6.81551337e-01 6.86359942e-01
-4.94047940e-01 9.74435151e-01 5.76148331e-01 -5.73988795e-01
-1.19916689e+00 -1.08252859e+00 1.84756279e-01 6.92900419e-01
6.83106259e-02 -5.01016438e-01 -1.15099549e+00 -8.88512135e-01
1.68254554e-01 -1.28226271e-02 -9.61619973e-01 -1.22912683e-01
-5.85128248e-01 -4.37582940e-01 7.98576951e-01 7.06559718e-02
5.76499164e-01 -8.92718434e-01 -4.63167131e-01 -1.92450657e-01
-3.48150313e-01 -1.25547802e+00 -8.88752759e-01 -4.39101368e-01
-4.91617918e-01 -1.17901289e+00 -6.72552705e-01 -6.53274715e-01
9.23977137e-01 5.03503799e-01 1.06909800e+00 1.43490911e-01
-4.41040039e-01 6.91289604e-01 -3.26462001e-01 -9.01796967e-02
-5.82071900e-01 -3.43411505e-01 8.93047974e-02 6.51623845e-01
2.89267063e-01 -7.10105777e-01 -6.65817201e-01 9.66131762e-02
-1.33651519e+00 -1.72221631e-01 1.15147658e-01 7.47818410e-01
5.86784422e-01 1.88904077e-01 1.11989744e-01 -9.60588694e-01
4.20303315e-01 -6.18027627e-01 -4.23776865e-01 7.06949681e-02
-1.12623371e-01 -3.36912930e-01 4.41765100e-01 -4.97426599e-01
-7.00629890e-01 2.81437516e-01 1.99968815e-01 -1.13853455e+00
-1.90951034e-01 -1.01884328e-01 -3.81080747e-01 -4.73699778e-01
2.47487321e-01 3.39349061e-01 2.61802197e-01 -8.11704099e-02
4.17467922e-01 2.32943028e-01 1.00350106e+00 -2.70122677e-01
1.15597022e+00 9.47960556e-01 -1.77913159e-01 -7.20822453e-01
-2.33269513e-01 -7.17112049e-02 -4.19770390e-01 -2.79673606e-01
8.53778481e-01 -1.14282596e+00 -6.23976290e-01 6.36799335e-01
-1.08406913e+00 -1.35500193e-01 -3.83605808e-01 5.23083806e-02
-7.26028323e-01 5.52411854e-01 -5.75398386e-01 -6.19423151e-01
-1.88177615e-01 -8.48983228e-01 1.14544022e+00 1.32048592e-01
-2.56232709e-01 -7.09292650e-01 -2.41807383e-02 3.47461522e-01
2.91215897e-01 8.09215546e-01 2.72982389e-01 -3.59501332e-01
-9.18294787e-01 -2.11873144e-01 1.22941382e-01 3.49609733e-01
3.18116665e-01 2.00841948e-01 -1.08986235e+00 -6.40904248e-01
4.53618646e-01 -8.38877708e-02 8.63592029e-01 4.16040212e-01
1.05371988e+00 -8.58757198e-01 -1.96411401e-01 1.03293073e+00
1.31983626e+00 2.71435171e-01 9.71906543e-01 1.48026764e-01
7.61863053e-01 4.94050711e-01 1.23975188e-01 6.41846001e-01
4.64399979e-02 4.89664078e-01 2.79605240e-01 -2.27004513e-01
-5.14002666e-02 -5.83547473e-01 5.50373256e-01 9.44297388e-02
6.31548613e-02 -4.97153580e-01 -3.93641323e-01 5.70980728e-01
-1.70149803e+00 -1.50174177e+00 5.11218309e-01 1.93894088e+00
6.30047739e-01 -4.36501324e-01 2.85938084e-01 -1.16816245e-01
8.92672837e-01 3.94981295e-01 -3.82828742e-01 -2.45406721e-02
-4.69622314e-01 3.90359014e-01 6.55595303e-01 4.64858592e-01
-1.38085461e+00 1.05431950e+00 7.03812599e+00 5.01858294e-01
-1.03825569e+00 2.77603954e-01 9.97286618e-01 -3.67559880e-01
-3.52110893e-01 1.43387794e-01 -2.48105124e-01 7.88657069e-01
8.26273918e-01 -2.71964163e-01 6.97760761e-01 5.90285659e-01
2.45244473e-01 1.33582130e-01 -1.11830461e+00 9.99859571e-01
3.60872209e-01 -1.50956714e+00 2.64921546e-01 2.57550955e-01
1.01367962e+00 -4.08126831e-01 5.85993290e-01 -4.26388025e-01
4.11206216e-01 -1.25842881e+00 8.02842081e-01 5.15451789e-01
1.01306784e+00 -8.66221666e-01 2.62989193e-01 -1.85991988e-01
-9.51407194e-01 1.38737276e-01 -1.48051217e-01 3.54521312e-02
1.26311764e-01 1.90200090e-01 -4.73689437e-01 4.12705541e-01
8.31075370e-01 8.08375001e-01 -4.10792261e-01 6.14791036e-01
-5.74491099e-02 4.66358364e-01 -2.90971935e-01 9.52618718e-01
-1.26145199e-01 -2.47281581e-01 8.11478138e-01 8.46843779e-01
3.71445149e-01 2.86878496e-01 -8.47259983e-02 9.61104214e-01
-5.30669332e-01 -2.97598064e-01 -1.14132285e+00 -1.25632957e-01
5.18912315e-01 9.49960828e-01 -6.56494319e-01 -1.15989342e-01
-3.83952647e-01 1.51670003e+00 -1.60316184e-01 5.26671946e-01
-8.60941768e-01 -2.58722361e-02 1.15567780e+00 2.04436764e-01
5.39142489e-01 -5.34662642e-02 -1.51003769e-04 -1.41388047e+00
2.30086148e-01 -1.18707132e+00 3.65001351e-01 -5.48543751e-01
-1.26048625e+00 5.91734290e-01 -2.45572448e-01 -1.18948126e+00
-4.01089251e-01 5.45898043e-02 -6.30642295e-01 6.54486656e-01
-1.12685907e+00 -1.26146138e+00 -1.71520516e-01 1.07357883e+00
2.27599487e-01 -2.90749669e-01 7.15043366e-01 1.61614463e-01
-6.44134939e-01 9.11250770e-01 8.69894475e-02 4.24166054e-01
7.62226105e-01 -6.49110854e-01 6.65148795e-01 1.42235434e+00
2.33582512e-01 7.04938829e-01 7.09075451e-01 -6.26996517e-01
-1.53884339e+00 -1.35784805e+00 6.72761619e-01 -5.56795776e-01
2.57565856e-01 -5.12186170e-01 -7.88227439e-01 1.09999609e+00
4.45037067e-01 3.24630171e-01 6.96048379e-01 -8.19060862e-01
-5.60381472e-01 3.89416260e-03 -1.60744393e+00 5.45621216e-01
9.81781304e-01 -8.98973525e-01 -2.31482357e-01 1.40165627e-01
6.55515194e-01 -2.23122001e-01 -6.60514951e-01 2.91741431e-01
4.84824121e-01 -1.21846902e+00 9.89633858e-01 -5.60938716e-01
4.10962582e-01 -4.14435536e-01 -1.50144622e-01 -7.84930527e-01
-8.59538466e-03 -1.48430777e+00 -4.75977845e-02 1.41182899e+00
-9.71332639e-02 -4.98163670e-01 1.06035542e+00 7.28434086e-01
5.18909931e-01 -2.27890220e-02 -9.61184382e-01 -6.70442283e-01
-2.40648776e-01 -1.64816588e-01 6.05242550e-01 1.09362996e+00
-3.90156239e-01 -2.75358588e-01 -8.83979738e-01 2.30247751e-01
7.83685446e-01 -5.61682284e-02 9.81365442e-01 -6.49424613e-01
-1.90008223e-01 -6.28165528e-02 -7.40916133e-01 -5.80497205e-01
3.87684673e-01 -5.02413929e-01 -2.15646297e-01 -4.29071397e-01
3.03857118e-01 1.71291754e-01 3.96084636e-02 4.08467352e-01
-9.15478263e-03 9.59721684e-01 1.04735591e-01 3.67031127e-01
-3.75688225e-01 3.58670473e-01 8.74513686e-01 -8.72437134e-02
1.19528115e-01 -4.74907719e-02 -8.02814126e-01 4.50872630e-01
6.75760269e-01 -6.97776735e-01 -5.55794358e-01 -3.42999071e-01
-3.94290984e-01 1.93079963e-01 8.20318460e-01 -9.31657732e-01
-3.10155377e-03 -2.00991049e-01 5.60789585e-01 -1.98199660e-01
3.84634912e-01 -1.00719273e+00 8.14803958e-01 3.47602218e-01
-2.82002866e-01 3.37660670e-01 2.71838248e-01 7.21139252e-01
-3.80231649e-01 3.79184276e-01 9.55750942e-01 -1.64715454e-01
-5.13343275e-01 7.01911747e-01 -2.88004935e-01 4.65640388e-02
1.34416795e+00 -3.71703029e-01 -1.37765840e-01 -8.34637403e-01
-7.13107526e-01 -1.97099492e-01 1.17603195e+00 4.55937475e-01
6.38718426e-01 -1.34655941e+00 -8.05954337e-01 8.51224720e-01
-2.49071673e-01 -5.09945333e-01 4.24924850e-01 4.77824479e-01
-6.23578370e-01 -6.50845766e-02 -4.75234181e-01 -2.74333864e-01
-1.57141674e+00 7.79840291e-01 2.71683931e-01 1.36076167e-01
-5.98349333e-01 9.71077144e-01 4.67717230e-01 4.02871370e-01
1.73853375e-02 1.03975691e-01 3.04735988e-01 -2.00322479e-01
7.39539623e-01 -1.67071763e-02 -1.88691571e-01 -1.18539858e+00
-3.79336149e-01 2.45203495e-01 -9.94250178e-02 -3.44343364e-01
1.19933319e+00 -4.63110179e-01 -3.45409483e-01 -1.73080012e-01
1.54998040e+00 3.00241113e-01 -1.67220044e+00 3.86790149e-02
-4.08270091e-01 -1.00956976e+00 -2.89018571e-01 -3.08585137e-01
-1.49871087e+00 2.09106624e-01 4.44681168e-01 2.45018676e-01
1.47487342e+00 -1.80946916e-01 9.88435864e-01 -3.39011371e-01
2.22388297e-01 -5.70090413e-01 -9.79066044e-02 1.42263621e-01
6.64839983e-01 -1.01437199e+00 -8.97760466e-02 -5.29882669e-01
-5.78207731e-01 7.48876452e-01 1.67656839e-01 -2.27120116e-01
6.08988523e-01 5.55303097e-01 1.64903313e-01 -3.22423913e-02
-7.39101350e-01 1.39620230e-01 4.19099033e-02 9.34331179e-01
1.01748435e-02 -9.84668657e-02 1.12740332e-02 4.38059241e-01
-3.58010143e-01 2.69631371e-02 4.85118389e-01 1.10608160e+00
2.95701802e-01 -1.19835222e+00 -4.95285094e-01 -9.63340104e-02
-9.32985842e-01 -9.06366408e-02 -8.54671836e-01 6.30652487e-01
2.18673661e-01 9.86434937e-01 3.16156387e-01 -5.81266284e-01
-1.19626127e-01 -7.05416733e-03 4.00563419e-01 -1.45178884e-01
-5.56549966e-01 7.22386539e-02 -2.98008978e-01 -8.61872613e-01
-6.80987239e-01 -9.20031071e-01 -6.24106407e-01 -8.48748684e-01
3.87107730e-01 -1.90345496e-02 2.58260041e-01 4.79093403e-01
6.85589552e-01 1.47419004e-03 1.04167426e+00 -8.98463130e-01
2.00793911e-02 -3.28910828e-01 -6.22816622e-01 6.70645058e-01
8.40749979e-01 -2.43106022e-01 -4.95684117e-01 8.11203599e-01] | [12.726192474365234, 0.20487245917320251] |
3612baeb-19c6-4ee3-b739-4d6af377e3b6 | damp-doubly-aligned-multilingual-parser-for | 2212.08054 | null | https://arxiv.org/abs/2212.08054v2 | https://arxiv.org/pdf/2212.08054v2.pdf | DAMP: Doubly Aligned Multilingual Parser for Task-Oriented Dialogue | Modern virtual assistants use internal semantic parsing engines to convert user utterances to actionable commands. However, prior work has demonstrated that semantic parsing is a difficult multilingual transfer task with low transfer efficiency compared to other tasks. In global markets such as India and Latin America, this is a critical issue as switching between languages is prevalent for bilingual users. In this work we dramatically improve the zero-shot performance of a multilingual and codeswitched semantic parsing system using two stages of multilingual alignment. First, we show that constrastive alignment pretraining improves both English performance and transfer efficiency. We then introduce a constrained optimization approach for hyperparameter-free adversarial alignment during finetuning. Our Doubly Aligned Multilingual Parser (DAMP) improves mBERT transfer performance by 3x, 6x, and 81x on the Spanglish, Hinglish and Multilingual Task Oriented Parsing benchmarks respectively and outperforms XLM-R and mT5-Large using 3.2x fewer parameters. | ['Rushin Shah', 'Diyi Yang', 'Rahul Goel', 'Eric Zhu', 'Fei Liu', 'Christopher Hidey', 'William Held'] | 2022-12-15 | null | null | null | null | ['semantic-parsing', 'xlm-r'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.16873789e-02 3.31462830e-01 -1.38675943e-01 -6.10789061e-01
-1.36831665e+00 -1.19112086e+00 3.98465604e-01 -2.59344041e-01
-8.98790836e-01 1.02289581e+00 2.89896607e-01 -9.23266590e-01
4.91052926e-01 -5.09069085e-01 -1.01521277e+00 -1.45596802e-01
4.00576055e-01 9.51844990e-01 -9.92324576e-02 -6.47863090e-01
-1.19599782e-01 -5.28329387e-02 -7.56882966e-01 2.16899037e-01
1.22766483e+00 3.10664773e-01 4.18699086e-01 9.20575440e-01
-4.57661211e-01 5.83672822e-01 -4.97265816e-01 -8.42884064e-01
3.85033548e-01 -2.89458334e-01 -1.38238931e+00 -5.94264746e-01
6.55801475e-01 -3.80571425e-01 -1.43468738e-01 1.15014148e+00
6.59416854e-01 3.28293554e-02 2.01677471e-01 -1.10920382e+00
-6.64451957e-01 1.14385200e+00 -3.48308653e-01 -6.68377802e-03
3.66945326e-01 -2.78930366e-02 1.18080473e+00 -5.40700376e-01
8.17848563e-01 1.67940676e+00 6.59067154e-01 8.65324378e-01
-9.61792171e-01 -1.04093349e+00 3.32671881e-01 -1.16118804e-01
-8.40987623e-01 -4.00837690e-01 3.07556897e-01 -1.64989978e-01
1.25425851e+00 2.35120893e-01 2.28465279e-03 1.54015088e+00
1.13146594e-02 8.46914709e-01 1.24724555e+00 -5.37665367e-01
-5.93755208e-02 2.45146483e-01 2.02568322e-01 8.33596647e-01
-1.78952768e-01 -1.14545427e-01 -3.95848304e-01 -2.99700424e-02
4.88263577e-01 -5.85275948e-01 1.97734103e-01 1.62243061e-02
-1.16556180e+00 1.07108176e+00 2.94377387e-01 1.17953695e-01
5.38131408e-02 2.06516251e-01 7.93009877e-01 6.51578784e-01
3.75245571e-01 4.40841556e-01 -8.96116138e-01 -5.72285056e-01
-3.10117304e-01 1.65912688e-01 9.02652085e-01 1.26845312e+00
5.37646294e-01 -7.33692572e-02 1.54376984e-01 1.16259944e+00
-2.52283253e-02 1.00853801e+00 6.18468106e-01 -1.05267835e+00
1.21383464e+00 7.34787062e-02 -2.13254448e-02 -3.03904176e-01
-4.43557650e-01 -3.90332304e-02 -3.67137939e-01 -2.29895696e-01
4.92447764e-01 -5.57696342e-01 -7.66907334e-01 2.19442201e+00
2.31809378e-01 -1.13490514e-01 4.91205126e-01 6.44877493e-01
6.76182032e-01 5.76909542e-01 7.65470564e-01 3.07406068e-01
1.45425606e+00 -1.12261617e+00 -6.68587804e-01 -6.87458634e-01
8.15782249e-01 -9.90047812e-01 1.61804342e+00 -7.92974830e-02
-1.07614326e+00 -3.94469529e-01 -9.32669461e-01 -4.03685600e-01
-3.58622134e-01 -3.41188699e-01 9.60860252e-01 8.16621423e-01
-1.06165850e+00 2.15044618e-01 -9.55017090e-01 -7.45095968e-01
4.65152144e-01 4.20595944e-01 -5.78849077e-01 -2.42611486e-02
-1.35974443e+00 1.00725639e+00 3.04229379e-01 -5.25700033e-01
-6.42139494e-01 -7.77517259e-01 -8.56921673e-01 -1.60704285e-01
8.87228996e-02 -6.47024155e-01 1.73092759e+00 -9.58615780e-01
-1.79059422e+00 1.16750407e+00 -3.85928378e-02 -4.64600503e-01
7.34724998e-01 -7.62019336e-01 -1.23267539e-01 -6.95724413e-02
4.29466605e-01 1.07351375e+00 3.89665782e-01 -9.63172436e-01
-8.71118009e-01 -4.66663569e-01 2.84805238e-01 8.50125670e-01
-2.23009482e-01 4.19322371e-01 -4.63109583e-01 -6.70831382e-01
-1.38027534e-01 -1.22204852e+00 -2.49050319e-01 -6.41099632e-01
-4.32381451e-01 5.07138334e-02 5.44106781e-01 -1.17268407e+00
5.81517935e-01 -2.01605368e+00 3.94570291e-01 -1.95581645e-01
-2.29063153e-01 1.58412442e-01 -4.61288452e-01 8.40528235e-02
2.43450090e-01 2.44548082e-01 -2.61073709e-01 -4.00740027e-01
8.91851261e-02 4.28923666e-01 -3.98324192e-01 -5.43143861e-02
-1.18675299e-01 1.13446820e+00 -9.04990971e-01 -4.43190068e-01
2.67030280e-02 2.93861240e-01 -9.97475564e-01 4.58185583e-01
-8.22992250e-02 6.59223199e-01 -4.38020617e-01 6.05121911e-01
4.90879893e-01 1.43003374e-01 5.89121878e-01 2.45726809e-01
1.68796539e-01 5.03432870e-01 -6.20998383e-01 2.33287930e+00
-1.18938828e+00 4.91862953e-01 2.84364879e-01 -7.25865245e-01
6.30926311e-01 2.91018128e-01 1.64792523e-01 -9.86308157e-01
1.03429653e-01 2.97833502e-01 -3.53692055e-01 -1.66297674e-01
8.13391685e-01 -1.40437126e-01 -7.58125961e-01 4.50288951e-01
2.46424660e-01 -1.18715666e-01 -2.98234046e-01 4.10686195e-01
9.84848261e-01 3.57606769e-01 2.16946557e-01 -5.81354022e-01
2.50469893e-01 4.43325154e-02 5.16441941e-01 8.66748571e-01
-3.40212643e-01 2.48363167e-01 3.43115240e-01 -1.96565777e-01
-1.06824255e+00 -1.42802882e+00 7.52469376e-02 1.93828583e+00
1.29519254e-01 -1.22404359e-01 -1.38345265e+00 -9.90425587e-01
-5.19943237e-02 7.91419685e-01 -9.63350609e-02 3.71184908e-02
-1.02231896e+00 -7.39021719e-01 8.75087738e-01 6.18900180e-01
7.37205386e-01 -1.22221196e+00 -3.11760366e-01 2.82336652e-01
-7.39640415e-01 -1.65984702e+00 -6.70893192e-01 3.28709215e-01
-7.98520625e-01 -7.57133067e-01 -2.34605506e-01 -1.19547629e+00
4.58561629e-01 -3.96870077e-02 1.40529883e+00 -2.65117615e-01
-9.56496224e-02 3.06167513e-01 -3.64902526e-01 -3.59354556e-01
-7.87472308e-01 7.51504958e-01 1.19943060e-01 -6.18174374e-01
5.80417812e-01 -5.97184062e-01 -1.95121065e-01 1.51773378e-01
-4.60062176e-01 2.17035607e-01 4.63084191e-01 8.68177772e-01
2.43126482e-01 -7.30381310e-01 6.86649978e-01 -1.42673492e+00
6.86336935e-01 -3.59508455e-01 -6.14477098e-01 3.73946607e-01
-3.78548235e-01 3.21893632e-01 7.75090575e-01 -2.25254446e-01
-1.35810983e+00 1.19514853e-01 -3.47036690e-01 1.71551973e-01
-4.20123748e-02 -2.03250200e-02 -4.55433130e-01 1.24515168e-01
7.83154190e-01 -4.71722558e-02 -1.56938270e-01 -4.27025020e-01
8.15158486e-01 1.04036915e+00 8.54053617e-01 -1.20295918e+00
7.74374187e-01 1.50804773e-01 -7.05697834e-01 -6.29643619e-01
-7.57895708e-01 -2.53733724e-01 -5.13219476e-01 2.42981106e-01
1.29574585e+00 -1.26907015e+00 -6.00665033e-01 3.49337667e-01
-1.25240314e+00 -9.04520333e-01 9.45500731e-02 2.72531033e-01
-7.12824941e-01 1.26687527e-01 -1.05572367e+00 -1.87233403e-01
-7.67426908e-01 -1.49194598e+00 1.00281036e+00 4.93027130e-03
-4.43222821e-01 -1.08458793e+00 3.95209976e-02 8.64802122e-01
6.14133060e-01 -2.10309491e-01 1.03147185e+00 -8.09748471e-01
-3.54771614e-01 3.13660890e-01 -1.21762618e-01 2.81216145e-01
2.48766821e-02 -6.28204167e-01 -9.49790359e-01 -5.14846623e-01
-4.08271521e-01 -6.32534981e-01 3.02281648e-01 1.42050207e-01
7.37966001e-01 -4.01033610e-01 -2.32518464e-01 9.01270807e-01
1.19284999e+00 2.49834910e-01 3.52541536e-01 6.22343063e-01
8.73878896e-01 4.93891209e-01 5.67488492e-01 -5.46763837e-02
7.00307906e-01 7.05805600e-01 8.17969143e-02 -8.26716349e-02
-6.72607571e-02 -5.01387715e-01 6.26157999e-01 1.21351731e+00
1.91993326e-01 -7.64742568e-02 -9.16599929e-01 3.44401181e-01
-1.78786767e+00 -3.84752989e-01 4.24225539e-01 2.08789039e+00
1.18110168e+00 1.69711336e-02 -1.68151528e-01 -6.67028189e-01
7.87701368e-01 -5.23842983e-02 -4.93304640e-01 -8.16534221e-01
-9.82123539e-02 6.81854725e-01 8.55899930e-01 1.16151178e+00
-1.20889878e+00 2.03662705e+00 6.35867214e+00 5.71227729e-01
-8.20792973e-01 6.25287056e-01 6.70249045e-01 3.86224836e-02
-3.08656812e-01 -4.15652134e-02 -9.91188169e-01 3.91604573e-01
1.19460225e+00 -4.84362757e-03 9.50457752e-01 1.00462210e+00
-3.05629969e-01 3.89589578e-01 -9.24566269e-01 8.51895392e-01
-2.06725031e-01 -8.55506063e-01 4.00475673e-02 -2.66321778e-01
6.56221211e-01 4.95522588e-01 9.75708142e-02 8.10420930e-01
1.19043243e+00 -1.14001667e+00 6.02326572e-01 -3.91587168e-01
9.12296176e-01 -9.45122659e-01 6.00810647e-01 2.28227326e-03
-9.97807205e-01 8.26101750e-02 -3.29076886e-01 2.03896612e-01
2.34303966e-01 -2.81443536e-01 -6.69495344e-01 3.11922699e-01
8.65452230e-01 2.63117522e-01 -3.68702918e-01 -1.48837373e-01
-4.86728698e-01 6.05069876e-01 -1.49283275e-01 1.97566912e-01
3.94989550e-01 -3.45278591e-01 4.34641927e-01 1.41550684e+00
1.10405654e-01 -5.53443357e-02 2.55760998e-01 3.58902216e-01
-4.85629946e-01 3.48679274e-01 -5.81506193e-01 4.60457876e-02
7.45085120e-01 1.16167367e+00 -4.48198527e-01 -2.79219776e-01
-5.53778350e-01 1.51735842e+00 7.18334675e-01 3.70985031e-01
-9.31013227e-01 -2.72584170e-01 1.08612955e+00 -4.11041349e-01
-7.73738995e-02 -2.04407275e-01 -3.10404390e-01 -1.26268458e+00
-1.04859196e-01 -1.38973701e+00 5.23611009e-01 -4.81537342e-01
-9.84108031e-01 8.09955418e-01 -2.51299322e-01 -5.42128146e-01
-4.39014852e-01 -6.75120950e-01 -1.96741611e-01 9.58379984e-01
-1.50430334e+00 -1.47076201e+00 -1.32856011e-01 4.94510293e-01
8.84088337e-01 -6.50628507e-01 1.24939966e+00 4.57892805e-01
-5.17363012e-01 1.16184473e+00 7.63889179e-02 4.79580790e-01
9.67250943e-01 -1.47313428e+00 1.08610511e+00 8.47262144e-01
4.96270321e-02 5.92234373e-01 6.09410524e-01 -5.27108610e-01
-1.27266252e+00 -1.19206691e+00 9.75826204e-01 -7.99871087e-01
8.08593571e-01 -6.63673997e-01 -6.38395667e-01 1.31267130e+00
5.29862046e-01 -3.08556676e-01 5.72129607e-01 4.06541377e-01
-6.86334848e-01 1.50369108e-01 -1.10909665e+00 8.60867858e-01
1.31913006e+00 -9.12284911e-01 -5.82209527e-01 5.04227996e-01
1.18885064e+00 -7.00688064e-01 -8.57865095e-01 1.52963489e-01
4.61709499e-01 -4.78124797e-01 1.02107573e+00 -1.04145467e+00
3.98182511e-01 4.48810518e-01 -5.61468363e-01 -1.39791679e+00
4.60499302e-02 -9.96251225e-01 4.82797772e-01 1.22881210e+00
5.26590645e-01 -8.68736327e-01 7.53031492e-01 5.44817090e-01
-9.62532535e-02 -2.86891032e-02 -8.50688517e-01 -8.38697076e-01
8.35528076e-01 -2.87418693e-01 5.45458555e-01 1.22900200e+00
6.74169809e-02 7.04044700e-01 -2.04849824e-01 1.52028695e-01
6.15123034e-01 -8.61633644e-02 8.60018015e-01 -8.33241642e-01
-4.74158049e-01 -3.62827659e-01 -1.72562543e-02 -9.62222040e-01
9.05069709e-01 -1.38535118e+00 2.44451463e-02 -1.13659894e+00
2.59257734e-01 -8.04623425e-01 -2.28204906e-01 6.54237688e-01
-5.23069203e-01 3.36447954e-02 2.71244258e-01 -4.02739495e-02
-5.62991202e-01 2.01830849e-01 8.03475022e-01 -8.98030773e-02
-1.92717195e-01 -4.15719986e-01 -8.48924220e-01 4.37259436e-01
1.04940522e+00 -5.06756961e-01 -4.79863971e-01 -9.42315519e-01
5.46421111e-02 9.80928987e-02 -1.80667832e-01 -5.50173461e-01
-2.60174423e-01 -7.51530528e-02 -2.29086474e-01 1.92993477e-01
6.97896332e-02 -5.24199367e-01 -1.87025964e-01 4.23274249e-01
-5.13923824e-01 3.81709129e-01 5.26963353e-01 1.85367793e-01
-4.24443111e-02 3.99663858e-02 7.64536023e-01 -3.41335922e-01
-7.80821383e-01 9.39547941e-02 -2.79975623e-01 6.42325640e-01
7.50736475e-01 3.39195758e-01 -5.52650213e-01 -3.20936203e-01
-5.54874182e-01 3.94627452e-01 2.74074733e-01 8.81187916e-01
2.31906236e-03 -1.22651541e+00 -8.00675988e-01 7.63126686e-02
-3.25675197e-02 -8.73352513e-02 7.91920945e-02 2.85817713e-01
-9.38528478e-01 6.24508440e-01 -5.12682438e-01 -4.00907725e-01
-1.34903634e+00 1.84252515e-01 3.53501946e-01 -3.29997808e-01
-5.73193789e-01 1.08785915e+00 3.85362566e-01 -1.40742397e+00
3.13249409e-01 -3.32329050e-02 8.29829425e-02 -3.47758263e-01
2.20652282e-01 2.13792101e-01 -1.42383222e-02 -5.93371451e-01
-3.13924074e-01 4.36967731e-01 -3.38706315e-01 -4.33401287e-01
1.18192375e+00 -8.47018957e-02 -4.91856709e-02 3.25649008e-02
1.37881041e+00 9.33162794e-02 -1.10258400e+00 -3.89999479e-01
1.33168578e-01 2.36451775e-02 -2.74548262e-01 -9.87640977e-01
-8.70631337e-01 7.14240313e-01 5.75370848e-01 -4.36180353e-01
7.43250668e-01 -4.69315611e-03 1.19255865e+00 6.60069585e-01
5.68881810e-01 -1.34754467e+00 -1.36809289e-01 7.21179843e-01
6.00998282e-01 -1.49845755e+00 -5.44322252e-01 -4.60477710e-01
-8.40438604e-01 5.64931571e-01 7.83945620e-01 2.80425787e-01
1.86674267e-01 5.03271639e-01 6.10376537e-01 2.95024216e-01
-3.98103446e-01 1.46235645e-01 -2.55134761e-01 5.37982523e-01
5.96784592e-01 6.18141353e-01 -1.83619440e-01 6.10774457e-01
-8.37304771e-01 -5.83748043e-01 2.06114367e-01 9.95664954e-01
-2.45748252e-01 -1.62104964e+00 -1.54063746e-01 -3.82452793e-02
-8.26684415e-01 -5.54763556e-01 -2.45713010e-01 8.85416389e-01
-5.29801726e-01 7.56959260e-01 1.23681441e-01 -2.57091224e-01
1.84924260e-01 4.80432183e-01 4.89799768e-01 -5.30080497e-01
-8.06128621e-01 -3.42469960e-01 3.14276218e-01 -7.63137460e-01
8.75682160e-02 -5.75749755e-01 -1.61196172e+00 -4.00030792e-01
2.15493530e-01 9.22939479e-02 9.99474823e-01 8.92489374e-01
3.49821210e-01 4.15732056e-01 3.31159860e-01 -3.17982495e-01
-7.43787766e-01 -8.72426748e-01 7.90426414e-03 5.47197044e-01
-1.12823233e-01 -1.38334349e-01 -6.17978647e-02 1.21364854e-01] | [10.875717163085938, 9.41396427154541] |
9fc3f6ba-6a8e-4401-af47-2ab6b339bf33 | adversarial-capsule-networks-for-romanian | 2306.07845 | null | https://arxiv.org/abs/2306.07845v1 | https://arxiv.org/pdf/2306.07845v1.pdf | Adversarial Capsule Networks for Romanian Satire Detection and Sentiment Analysis | Satire detection and sentiment analysis are intensively explored natural language processing (NLP) tasks that study the identification of the satirical tone from texts and extracting sentiments in relationship with their targets. In languages with fewer research resources, an alternative is to produce artificial examples based on character-level adversarial processes to overcome dataset size limitations. Such samples are proven to act as a regularization method, thus improving the robustness of models. In this work, we improve the well-known NLP models (i.e., Convolutional Neural Networks, Long Short-Term Memory (LSTM), Bidirectional LSTM, Gated Recurrent Units (GRUs), and Bidirectional GRUs) with adversarial training and capsule networks. The fine-tuned models are used for satire detection and sentiment analysis tasks in the Romanian language. The proposed framework outperforms the existing methods for the two tasks, achieving up to 99.08% accuracy, thus confirming the improvements added by the capsule layers and the adversarial training in NLP approaches. | ['Florin Pop', 'Dumitru-Clementin Cercel', 'Andrei-Marius Avram', 'Răzvan-Alexandru Smădu', 'Sebastian-Vasile Echim'] | 2023-06-13 | null | null | null | null | ['sentiment-analysis', 'satire-detection'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.21917057e-01 -1.37027279e-01 -1.22813463e-01 -1.26378655e-01
-5.41954875e-01 -7.80690849e-01 7.13167489e-01 -1.44889683e-01
-5.52324712e-01 7.73366630e-01 5.35801232e-01 -1.87024072e-01
4.68175143e-01 -7.06265628e-01 -5.19562542e-01 -7.24974573e-01
1.30934089e-01 5.33847734e-02 -3.94756526e-01 -8.48957181e-01
5.37279665e-01 4.44022655e-01 -1.09630144e+00 7.61310101e-01
7.61426926e-01 1.05248296e+00 -2.50966251e-01 5.50039470e-01
-3.49332571e-01 1.40879476e+00 -6.58351123e-01 -7.04744816e-01
-1.11514116e-02 -4.80633259e-01 -5.37975788e-01 -5.75786591e-01
7.48521686e-02 -7.28101954e-02 -1.98408589e-01 1.27277160e+00
7.16003835e-01 -7.75955841e-02 4.71352637e-01 -7.44504988e-01
-1.10212243e+00 9.95026290e-01 -4.76489782e-01 -7.70800635e-02
3.45544130e-01 -2.13925183e-01 8.14703763e-01 -8.56416047e-01
6.51642203e-01 1.29804063e+00 9.32947218e-01 9.10444677e-01
-7.56351948e-01 -9.56284106e-01 -2.14489087e-01 3.03180832e-02
-9.72010195e-01 -2.87006497e-01 1.25409734e+00 -2.44650528e-01
6.81200206e-01 2.27877289e-01 3.27930868e-01 1.77531147e+00
5.20258844e-01 1.06601489e+00 1.52188575e+00 -2.99099237e-01
5.09896539e-02 4.27186161e-01 -3.32496054e-02 3.84777755e-01
-2.92840868e-01 -7.06471177e-03 -5.46828389e-01 -2.78264105e-01
4.50351685e-01 -1.85348950e-02 -8.89514610e-02 2.20231071e-01
-9.19766366e-01 1.31493723e+00 7.39160001e-01 7.95937300e-01
-4.61086780e-01 6.17235415e-02 1.21724248e+00 4.35700506e-01
7.74404824e-01 9.24723208e-01 -7.04531014e-01 2.84581095e-01
-9.17603552e-01 1.17085114e-01 8.71690869e-01 5.69736362e-01
-5.50034866e-02 6.98950052e-01 -2.48098627e-01 9.04574275e-01
-1.89329721e-02 5.28377235e-01 1.22820437e+00 -5.90040535e-02
3.70482057e-01 6.58066034e-01 -6.12338260e-02 -1.31785202e+00
-4.67272639e-01 -4.96813893e-01 -1.09048939e+00 2.94055641e-01
1.30885497e-01 -4.35275257e-01 -6.02865577e-01 1.67260945e+00
-2.18030170e-01 2.00323015e-02 7.05408812e-01 8.93327296e-01
1.28936601e+00 8.44083369e-01 4.66662675e-01 -4.09472622e-02
1.36637235e+00 -9.13591206e-01 -8.53799462e-01 -3.19345504e-01
6.06680870e-01 -1.13629615e+00 1.21092415e+00 5.29175460e-01
-8.16868782e-01 -5.09738505e-01 -1.14655876e+00 -1.31644517e-01
-9.00325775e-01 2.10016832e-01 8.46673489e-01 5.93394637e-01
-5.81110179e-01 8.99531960e-01 -2.47704253e-01 7.57253170e-02
5.39835632e-01 2.75527239e-01 -1.71785012e-01 4.68650520e-01
-1.71192849e+00 9.67277288e-01 1.97959065e-01 4.85242486e-01
-7.49533296e-01 -5.28110564e-01 -9.92405891e-01 -1.52205691e-01
-1.93263531e-01 -6.37717918e-02 9.15973961e-01 -2.03121257e+00
-1.85660970e+00 1.14847064e+00 4.51933920e-01 -8.06339979e-01
3.68488878e-01 -3.87748539e-01 -4.91820097e-01 -7.84584880e-02
-2.86731035e-01 4.32094961e-01 1.02126372e+00 -1.10087001e+00
1.06141813e-01 -3.75160843e-01 2.62724180e-02 1.27634749e-01
-6.52793944e-01 4.67549950e-01 4.88391608e-01 -9.56174850e-01
-1.03655513e-02 -8.11038256e-01 -2.45849237e-01 -5.04819095e-01
-3.56336594e-01 -6.83823824e-02 8.25974643e-01 -1.08174241e+00
8.90677989e-01 -2.11909389e+00 3.59259956e-02 -1.43490613e-01
-1.52097583e-01 4.42955792e-01 -1.35202348e-01 2.65942574e-01
-2.33728066e-01 2.28892818e-01 9.81605379e-04 -2.02395767e-01
7.54157156e-02 -4.97842617e-02 -6.85977221e-01 5.72466373e-01
2.04065681e-01 1.12195110e+00 -8.54281843e-01 -2.02580303e-01
2.31364891e-01 6.27383769e-01 -3.19085240e-01 1.04262773e-02
-2.01809838e-01 6.20737255e-01 -5.38563073e-01 7.17856109e-01
4.80889231e-01 4.03460056e-01 4.83057238e-02 -1.90838933e-01
-7.15619768e-04 3.68016303e-01 -5.87725222e-01 1.59815693e+00
-9.62009430e-01 6.13808215e-01 5.99105917e-02 -9.16415393e-01
1.36219335e+00 5.37747562e-01 2.12031841e-01 -8.33123684e-01
7.81857610e-01 4.43762988e-01 -1.03804305e-01 -4.58560675e-01
4.77262467e-01 -7.28695154e-01 -5.87951243e-01 7.83705711e-02
1.28753223e-02 -1.94611073e-01 -3.07106793e-01 -1.69832811e-01
6.54796183e-01 2.56486833e-01 2.99245149e-01 -4.60184753e-01
1.03241432e+00 -1.06786452e-01 4.13952827e-01 4.90142316e-01
-1.14105798e-01 3.44007999e-01 5.10469437e-01 -5.57264805e-01
-1.14015114e+00 -4.63383824e-01 -9.17577595e-02 1.17732430e+00
-1.53452948e-01 1.18888870e-01 -7.03421593e-01 -5.69333851e-01
-3.63478124e-01 9.07888293e-01 -8.46525013e-01 -3.12379926e-01
-5.19454122e-01 -8.72116804e-01 1.03816426e+00 4.40942466e-01
5.70083082e-01 -2.01836944e+00 -5.98434865e-01 1.91296652e-01
1.09129623e-02 -9.53007698e-01 2.74652205e-02 4.63435918e-01
-7.95100212e-01 -8.32638502e-01 -7.99204886e-01 -9.67532992e-01
2.27858230e-01 -5.67016423e-01 1.02466333e+00 -1.31355986e-01
-3.66440453e-02 -2.94526160e-01 -6.72597766e-01 -4.54631180e-01
-6.55298889e-01 -3.12009696e-02 -6.34759814e-02 -2.94795576e-02
4.80750680e-01 -5.43874204e-01 -4.01444882e-01 -2.66498327e-01
-1.03351998e+00 -1.25052661e-01 5.90751410e-01 1.17238498e+00
2.45926976e-01 -2.22097561e-01 9.05700564e-01 -1.25946164e+00
9.98735309e-01 -4.59548503e-01 -3.19649994e-01 -1.18991017e-01
-2.70484120e-01 1.39914313e-02 1.33274472e+00 -8.31183612e-01
-1.10586679e+00 2.08293777e-02 -5.86608827e-01 -4.13053960e-01
-1.11189343e-01 4.29027319e-01 -2.22449880e-02 -2.51907229e-01
8.83968592e-01 1.36336297e-01 -3.58080715e-01 -3.25905859e-01
5.21764040e-01 8.36641610e-01 4.98589426e-01 -2.52781004e-01
4.32937622e-01 3.32500696e-01 -3.11156482e-01 -5.57166934e-01
-1.21575201e+00 -2.26339057e-01 -5.54001391e-01 7.88284987e-02
1.01266372e+00 -8.44599426e-01 -7.98453271e-01 5.15848458e-01
-1.11890900e+00 -3.05656880e-01 -9.00773779e-02 4.09464180e-01
-4.85704035e-01 1.97566688e-01 -1.07509375e+00 -9.22819734e-01
-1.24660099e+00 -9.57967639e-01 6.85433269e-01 2.02934116e-01
-3.24776560e-01 -1.22380197e+00 2.26338916e-02 3.17385465e-01
5.19904137e-01 7.37010360e-01 8.90511870e-01 -9.42386329e-01
6.72275543e-01 -4.45101559e-01 -5.90359755e-02 8.17782879e-01
-2.33596504e-01 -3.77658367e-01 -1.35154819e+00 4.59721833e-02
7.28134990e-01 -7.95305371e-01 7.97247171e-01 3.04495424e-01
9.34922218e-01 -5.74819624e-01 3.55012447e-01 5.71707129e-01
1.56215644e+00 1.60956264e-01 1.02243686e+00 5.95006108e-01
6.23453617e-01 5.32056093e-01 3.55312824e-01 2.49995798e-01
-3.54508102e-01 3.02587207e-02 5.61679423e-01 -3.99935007e-01
7.95147046e-02 -1.64490581e-01 7.42549419e-01 8.24241817e-01
3.13303769e-01 -5.54295406e-02 -4.64732051e-01 3.32091391e-01
-1.63446367e+00 -1.11248410e+00 -3.43038678e-01 1.83063161e+00
1.01513779e+00 3.16866010e-01 -4.38917309e-01 3.77280116e-01
5.07149100e-01 4.11618054e-01 -3.34976822e-01 -1.23887300e+00
-6.30782545e-01 8.27522993e-01 5.80258429e-01 5.26651323e-01
-1.37124074e+00 1.37004542e+00 5.54766941e+00 8.74789953e-01
-1.57005775e+00 2.71776199e-01 7.94975400e-01 1.41014636e-01
-1.79154396e-01 -4.44961756e-01 -4.12544578e-01 3.18656743e-01
9.25837219e-01 2.61156738e-01 5.56452215e-01 1.22872186e+00
3.14195335e-01 3.34218293e-01 -5.94390810e-01 7.45008945e-01
6.09320164e-01 -9.92278814e-01 1.91562593e-01 -6.03154838e-01
9.77576852e-01 3.83057743e-02 3.24583769e-01 6.69407666e-01
1.15246266e-01 -1.26852226e+00 6.81399286e-01 3.97683978e-01
5.90146482e-01 -1.03529751e+00 1.31065738e+00 1.31480664e-01
-7.42211938e-01 -2.21924052e-01 -5.05189359e-01 -3.94937843e-01
-1.74191803e-01 2.50400037e-01 -6.07189298e-01 3.25257123e-01
5.06124020e-01 7.42951035e-01 -5.06939352e-01 2.03827560e-01
-4.85479742e-01 6.36169672e-01 3.65059897e-02 -4.27709877e-01
7.02070594e-01 -1.62503496e-01 2.65607595e-01 1.36277092e+00
-1.29856214e-01 -5.69676198e-02 -5.87405153e-02 9.51560140e-01
-2.99051374e-01 7.17611909e-01 -7.14845359e-01 -4.07847106e-01
-2.83833265e-01 1.84001303e+00 -6.44779980e-01 -3.59190702e-01
-1.45976946e-01 1.24482262e+00 7.05146044e-02 1.08169906e-01
-8.48732173e-01 -3.53892565e-01 2.57262856e-01 -1.97755694e-01
-2.11002436e-02 1.42272964e-01 -8.54425251e-01 -1.09908760e+00
-3.58806640e-01 -1.19977188e+00 2.72329599e-01 -8.34377408e-01
-1.63612688e+00 9.00925517e-01 -7.53182769e-01 -1.06372166e+00
-1.94562674e-02 -9.25204515e-01 -7.96601653e-01 8.25533688e-01
-1.56731915e+00 -1.53380334e+00 1.46350622e-01 7.25934684e-01
8.24688077e-01 -2.60391831e-01 1.16011310e+00 3.01669925e-01
-2.30548993e-01 4.60778534e-01 8.62797499e-02 3.52365553e-01
6.10231280e-01 -1.08872354e+00 -1.10286988e-01 5.50426602e-01
-5.12175895e-02 6.04680896e-01 7.95317113e-01 -5.27317047e-01
-1.29075181e+00 -1.01171839e+00 1.05776918e+00 -1.07035458e-01
1.18044162e+00 -6.21279538e-01 -6.36144400e-01 3.82462293e-01
4.65660274e-01 -3.71687651e-01 7.45994210e-01 -3.64906266e-02
-4.39303577e-01 1.59596518e-01 -1.25982249e+00 6.70752108e-01
4.26786810e-01 -5.39000452e-01 -7.77989388e-01 4.55463588e-01
4.34627175e-01 -2.02777579e-01 -6.66394651e-01 4.98244435e-01
4.69810158e-01 -6.45564377e-01 9.45884824e-01 -6.33891582e-01
1.11982441e+00 1.15085863e-01 -1.92204073e-01 -1.01225197e+00
-1.42369699e-02 -4.44863975e-01 3.97004634e-01 1.13165581e+00
4.57593977e-01 -4.05785769e-01 5.38248003e-01 1.44178763e-01
-6.31296784e-02 -6.86754227e-01 -5.78106880e-01 -1.15141854e-01
4.26779658e-01 -3.18245649e-01 -5.52289328e-03 1.30225706e+00
1.12337165e-01 7.32788444e-01 -8.28929603e-01 -2.41835579e-01
1.32607609e-01 2.77621984e-01 3.58394742e-01 -1.01640153e+00
-9.67341065e-02 -5.39621055e-01 -1.45381391e-01 -4.10450071e-01
8.10813427e-01 -1.05469668e+00 3.25459731e-03 -1.11028862e+00
2.12928187e-02 -2.31991261e-01 -4.80822295e-01 5.24968565e-01
1.10699199e-01 7.63193250e-01 1.44660911e-02 -5.10382950e-02
-1.47815734e-01 6.48677766e-01 1.26947379e+00 -1.79985628e-01
-3.88090193e-01 -1.09092660e-01 -8.69476736e-01 1.24486196e+00
1.32080138e+00 -5.18481374e-01 2.80105583e-02 3.83881629e-02
5.76085329e-01 -2.78491914e-01 3.80479217e-01 -6.89893186e-01
-9.43590105e-02 2.80912425e-02 7.13395059e-01 -5.13540149e-01
3.86416316e-01 -6.33097231e-01 -3.24697584e-01 8.60183358e-01
-5.72228491e-01 -1.95282865e-02 4.20187980e-01 -8.00715014e-02
-2.81986356e-01 -4.05633122e-01 9.84296322e-01 -4.97162670e-01
-6.10813916e-01 -2.15771556e-01 -3.67641121e-01 -1.71034038e-01
6.37108028e-01 2.08929479e-01 -1.31391212e-01 -3.38588089e-01
-6.74054086e-01 -3.14412117e-01 3.56926881e-02 6.23508453e-01
4.84965473e-01 -1.22997224e+00 -8.05908501e-01 -1.71748865e-02
-1.65040553e-01 -6.03641927e-01 1.51298732e-01 7.95994520e-01
-7.43333995e-01 2.07547262e-01 -4.80244249e-01 7.14604184e-02
-1.17101336e+00 8.46509397e-01 3.36105704e-01 -4.78815675e-01
-4.88737553e-01 7.95235395e-01 9.08034891e-02 -7.49140143e-01
-5.35850972e-02 -6.64266720e-02 -8.73281658e-01 5.83998524e-02
4.00110424e-01 -2.70136967e-02 -1.70628428e-02 -8.09283316e-01
-1.41204134e-01 2.96276510e-01 9.18066055e-02 1.38228759e-01
1.42072964e+00 3.52687180e-01 -3.47720265e-01 7.11692572e-01
1.14214957e+00 3.53974789e-01 -5.06753087e-01 4.56534252e-02
3.63346306e-03 1.47492647e-01 1.01534486e-01 -1.08581805e+00
-1.04143846e+00 9.29988503e-01 5.51360369e-01 6.44232556e-02
1.10440421e+00 -3.33582133e-01 9.61427867e-01 3.23696047e-01
-1.81648768e-02 -1.11773205e+00 6.83037341e-02 6.60655200e-01
1.16206682e+00 -9.55333292e-01 -2.97157317e-01 3.97009216e-02
-9.13462400e-01 1.39008665e+00 3.56881529e-01 -9.37608540e-01
3.02273124e-01 4.82922673e-01 4.69419688e-01 -3.11443985e-01
-2.86205411e-01 6.43891767e-02 1.77122205e-01 1.31517455e-01
7.84887254e-01 5.49517199e-02 -9.40152943e-01 9.30281341e-01
-5.90499341e-01 -3.11632156e-01 3.23781312e-01 4.16387558e-01
-1.24897577e-01 -9.00956452e-01 -4.26193386e-01 1.23086974e-01
-1.13609636e+00 -7.24264324e-01 -8.82460237e-01 5.60243905e-01
2.43770093e-01 9.01062667e-01 -3.78534019e-01 -3.05886865e-01
1.97439209e-01 1.70389384e-01 2.12818086e-01 -3.85569364e-01
-1.58704710e+00 -2.34476048e-02 1.25245824e-01 -2.88036436e-01
-4.10177022e-01 -1.53837785e-01 -1.20342588e+00 -2.23210335e-01
-4.44654226e-01 8.18528086e-02 8.94872069e-01 1.04128373e+00
-2.07092077e-01 6.46618426e-01 8.72763336e-01 -7.43143380e-01
-6.92248106e-01 -1.41234291e+00 -5.36103189e-01 8.32654476e-01
-2.74593942e-02 -2.09176227e-01 -5.11096954e-01 1.14704691e-01] | [11.200509071350098, 7.033242225646973] |
39659789-5ef7-46e2-a95c-eec7c603f701 | image-moment-invariants-to-rotational-motion | 2303.14566 | null | https://arxiv.org/abs/2303.14566v1 | https://arxiv.org/pdf/2303.14566v1.pdf | Image Moment Invariants to Rotational Motion Blur | Rotational motion blur caused by the circular motion of the camera or/and object is common in life. Identifying objects from images affected by rotational motion blur is challenging because this image degradation severely impacts image quality. Therefore, it is meaningful to develop image invariant features under rotational motion blur and then use them in practical tasks, such as object classification and template matching. This paper proposes a novel method to generate image moment invariants under general rotational motion blur and provides some instances. Further, we achieve their invariance to similarity transform. To the best of our knowledge, this is the first time that moment invariants for rotational motion blur have been proposed in the literature. We conduct extensive experiments on various image datasets disturbed by similarity transform and rotational motion blur to test these invariants' numerical stability and robustness to image noise. We also demonstrate their performance in image classification and handwritten digit recognition. Current state-of-the-art blur moment invariants and deep neural networks are chosen for comparison. Our results show that the moment invariants proposed in this paper significantly outperform other features in various tasks. | ['Guoying Zhao', 'Hongxiang Hao', 'Hanlin Mo'] | 2023-03-25 | null | null | null | null | ['template-matching', 'handwritten-digit-recognition'] | ['computer-vision', 'computer-vision'] | [ 1.01267457e-01 -8.36638272e-01 -8.31386482e-04 -2.69509673e-01
-1.44274965e-01 -5.45186818e-01 4.72005695e-01 -3.83809686e-01
-3.45734894e-01 6.22943342e-01 1.89807639e-02 -7.80446380e-02
-4.32864040e-01 -1.03761517e-01 -4.41345006e-01 -8.36562335e-01
-2.47160513e-02 -2.40170389e-01 2.50243992e-01 1.99098006e-01
7.40513384e-01 1.00720501e+00 -1.57122207e+00 -9.76098552e-02
8.53500247e-01 1.11048174e+00 3.30211580e-01 8.30118001e-01
4.65042770e-01 9.03780162e-01 -8.61042917e-01 -4.36165869e-01
2.80720651e-01 -1.28496960e-01 -8.24515104e-01 6.27550006e-01
5.84243774e-01 -7.41269231e-01 -9.79649305e-01 1.38217926e+00
5.77241778e-01 2.35496044e-01 7.44267106e-01 -1.04862952e+00
-1.70313108e+00 2.47590274e-01 -8.16448987e-01 6.29005015e-01
3.40037405e-01 1.83973089e-01 3.57271373e-01 -9.44155276e-01
2.80091286e-01 1.06530106e+00 6.04947329e-01 4.15711105e-01
-9.47062135e-01 -1.43325433e-01 -3.77668709e-01 7.40510762e-01
-1.23473561e+00 -5.20909846e-01 1.09747267e+00 -4.22504455e-01
7.43809402e-01 5.42456865e-01 1.22843251e-01 8.19205642e-01
4.06243771e-01 7.45888948e-01 1.15947926e+00 -4.77714419e-01
-2.23647337e-02 -3.04065108e-01 3.72604668e-01 6.72383904e-01
5.88806927e-01 -4.52764072e-02 -1.25635430e-01 1.11001037e-01
1.03573072e+00 7.36755580e-02 -7.60554075e-01 -2.96142727e-01
-1.42420173e+00 2.68864334e-01 4.55593020e-01 4.35475856e-01
-3.07455778e-01 1.72585219e-01 1.46442205e-01 3.36510032e-01
3.09219897e-01 5.20776749e-01 -2.44227618e-01 -3.29237372e-01
-7.97784984e-01 -2.55479850e-02 8.01298559e-01 8.79033804e-01
4.17708129e-01 1.56855837e-01 -2.89363056e-01 1.17437136e+00
1.55064926e-01 6.51802361e-01 7.60614812e-01 -1.06386793e+00
5.04872426e-02 2.36449286e-01 2.80595332e-01 -1.05461597e+00
-1.19718850e-01 -1.72611222e-01 -1.11818326e+00 1.16560459e-01
6.65281892e-01 1.44617468e-01 -1.00306571e+00 1.03084624e+00
-1.65428668e-01 2.53875583e-01 7.66285881e-02 1.49152243e+00
6.14179254e-01 1.66320518e-01 -6.78134143e-01 -2.09562793e-01
1.51682389e+00 -7.64121532e-01 -1.00196671e+00 6.90353587e-02
-7.75636211e-02 -1.46869779e+00 7.60244966e-01 3.65510434e-01
-1.04750621e+00 -7.94008732e-01 -1.19429791e+00 -1.71009660e-01
-1.61535442e-01 4.73637015e-01 8.48227680e-01 7.61344075e-01
-1.02406418e+00 7.53793657e-01 -7.82639682e-01 -2.63877511e-01
2.97408819e-01 3.21513504e-01 -2.84909844e-01 -1.77125037e-01
-1.05988193e+00 1.13232243e+00 -1.20740391e-01 4.62010384e-01
-4.96928543e-01 -2.52817392e-01 -8.68397415e-01 -9.78551954e-02
-1.94536150e-01 -5.73460042e-01 1.65019643e+00 -7.22297132e-01
-1.43130100e+00 7.59453952e-01 -3.26176733e-01 -3.93957227e-01
6.68748736e-01 -4.08067673e-01 -4.41558033e-01 3.00392956e-01
-4.34814245e-02 1.77567542e-01 1.28264880e+00 -9.34491873e-01
-1.65706098e-01 -2.13235900e-01 -1.33056715e-01 -6.26069382e-02
-1.74091861e-01 3.07518035e-01 -3.03962022e-01 -8.31747174e-01
3.39548700e-02 -8.65111709e-01 3.76920439e-02 -6.03448600e-02
-1.01725757e-01 -1.88541010e-01 1.18753409e+00 -5.95515072e-01
9.14437413e-01 -2.04079771e+00 -1.41870260e-01 -2.60695785e-01
-3.29910330e-02 4.52967197e-01 -7.35120252e-02 1.88325211e-01
-2.31773451e-01 -3.26831758e-01 4.66988906e-02 9.27947834e-02
-1.44966900e-01 1.20119639e-01 -5.29371142e-01 9.06526983e-01
2.94435084e-01 9.66461480e-01 -6.44870758e-01 -2.53162652e-01
5.74611783e-01 5.39514840e-01 -7.30176568e-02 8.04593712e-02
4.97355133e-01 1.32325724e-01 -2.76195347e-01 8.45728457e-01
1.03750074e+00 -2.95179874e-01 -5.49356937e-01 -8.41564476e-01
-1.72742635e-01 -3.87347281e-01 -8.53385746e-01 1.14955389e+00
-8.46216828e-02 1.15125740e+00 -2.54249543e-01 -8.92038524e-01
8.81167352e-01 8.78972411e-02 4.74094540e-01 -3.45654935e-01
1.85285181e-01 2.89207816e-01 3.38226199e-01 -9.20139968e-01
7.69808233e-01 2.92584509e-01 3.95320475e-01 2.76091397e-01
-1.24010377e-01 -2.85743296e-01 2.96487957e-01 -2.02335238e-01
9.83999252e-01 -1.80279985e-01 7.47946724e-02 -2.69019425e-01
8.22207928e-01 -4.15699542e-01 4.67155427e-02 9.37708974e-01
-6.08851671e-01 9.73640382e-01 9.09283236e-02 -6.52645767e-01
-1.26669931e+00 -1.01244235e+00 -3.94349158e-01 4.49284256e-01
5.01424551e-01 5.12750685e-01 -7.53187120e-01 -1.43235549e-01
1.56343788e-01 1.90542504e-01 -3.77190948e-01 -1.34750560e-01
-4.76753086e-01 -8.01675797e-01 6.64118826e-01 5.73163509e-01
1.10505390e+00 -7.54525423e-01 -4.89014924e-01 -7.83685967e-02
5.59348427e-02 -1.41587675e+00 -9.25693333e-01 -3.81962746e-01
-7.85368621e-01 -1.33420002e+00 -1.47683263e+00 -1.05280149e+00
9.58936512e-01 8.64708424e-01 4.66899097e-01 -1.82282075e-01
-7.20016599e-01 6.36344612e-01 -2.39539787e-01 1.87937822e-02
-2.57139146e-01 -3.59595001e-01 2.92197883e-01 2.28480339e-01
3.34727436e-01 -2.25017130e-01 -9.04318273e-01 6.63807392e-01
-1.09110296e+00 -2.01126039e-01 8.42701554e-01 9.74615574e-01
-4.02039476e-02 2.65211821e-01 1.84741527e-01 1.76098086e-02
1.15409732e+00 3.47591460e-01 -7.92830646e-01 3.29795659e-01
-4.30437922e-01 2.91470826e-01 2.77325630e-01 -7.48550057e-01
-1.26680660e+00 -8.35708529e-03 2.20067456e-01 -6.47023857e-01
-3.82014781e-01 5.26498318e-01 2.80631870e-01 -5.54348409e-01
7.27422714e-01 4.38273996e-01 1.25085115e-01 -5.34985125e-01
1.74550816e-01 1.35307336e+00 9.92426991e-01 -3.57902735e-01
9.41220820e-01 4.84744459e-01 2.33304948e-01 -1.28774607e+00
-6.53262436e-01 -6.34679437e-01 -5.19580781e-01 -3.30759406e-01
6.91485882e-01 -6.83569610e-01 -9.99760926e-01 1.40795517e+00
-1.53185201e+00 1.56457111e-01 2.52853453e-01 7.78741717e-01
-4.95182246e-01 1.07889795e+00 -1.17195785e+00 -7.71059752e-01
-3.59023899e-01 -1.31237817e+00 9.23495054e-01 4.55554336e-01
2.26658314e-01 -8.52719128e-01 -3.17414731e-01 3.08674395e-01
6.83857858e-01 -1.76964790e-01 1.30316809e-01 2.23498903e-02
-8.70797098e-01 -2.86626488e-01 -7.10268080e-01 5.75616896e-01
9.40219700e-01 1.03284575e-01 -1.11791706e+00 -3.85034233e-01
3.94406408e-01 1.44028172e-01 1.03364253e+00 6.18400276e-01
1.28155780e+00 -3.71008247e-01 -2.37195548e-02 6.21789873e-01
1.07014775e+00 4.11244839e-01 9.12079751e-01 4.70581770e-01
6.94572508e-01 1.35270640e-01 5.66500425e-01 3.08066636e-01
-2.54252613e-01 6.30394101e-01 1.30971342e-01 -3.72523889e-02
-2.28952050e-01 6.13909006e-01 3.03718925e-01 6.88387930e-01
-4.55345690e-01 2.24886298e-01 -6.89128518e-01 4.71143246e-01
-1.67141771e+00 -1.11620975e+00 -2.18023866e-01 1.91826689e+00
8.89022350e-01 -9.72235277e-02 -4.07982230e-01 1.07217669e-01
1.12240148e+00 9.44399163e-02 -5.30519187e-01 -2.43500456e-01
-2.84231037e-01 5.24972007e-02 7.60284424e-01 3.64104688e-01
-1.53040767e+00 6.15192652e-01 5.98238707e+00 4.01268542e-01
-1.43157327e+00 -3.27188104e-01 5.60563147e-01 4.81919318e-01
5.10186493e-01 -1.96997672e-01 -4.68956888e-01 4.53175455e-01
3.37821484e-01 -3.37403864e-01 6.54512107e-01 9.16059434e-01
2.06091672e-01 -1.36821091e-01 -9.52582181e-01 1.38630736e+00
3.83795708e-01 -9.53960717e-01 -6.38409257e-02 -6.07983135e-02
8.59594285e-01 -1.62725598e-01 4.82129127e-01 -4.33512777e-01
-1.96829081e-01 -8.98759663e-01 4.52411085e-01 8.43245447e-01
7.01780200e-01 -5.14785647e-01 8.80968928e-01 -1.21473067e-01
-7.83040047e-01 6.69333190e-02 -6.36173725e-01 -2.09456816e-01
-3.98063920e-02 7.68396199e-01 -6.68343365e-01 5.08910716e-01
4.54312235e-01 8.80329132e-01 -8.34131539e-01 1.57123256e+00
1.61244310e-02 6.36457726e-02 3.30418736e-01 -8.09733421e-02
3.14489529e-02 -1.12795606e-01 5.93192577e-01 1.11084759e+00
6.92623913e-01 -2.56668925e-01 -3.92446935e-01 6.29666448e-01
7.19658062e-02 -5.06482005e-01 -6.74440920e-01 -8.89662281e-02
8.77348706e-02 1.19632912e+00 -6.52580202e-01 -3.33282679e-01
-5.51177561e-01 1.45394695e+00 -2.43197143e-01 5.99629521e-01
-8.06527078e-01 -9.69677806e-01 9.72736835e-01 -3.40139598e-01
4.25853252e-01 -6.20311022e-01 -3.50954533e-02 -1.66857302e+00
1.68256387e-01 -5.96729040e-01 -3.11376024e-02 -1.11993086e+00
-1.62588036e+00 4.04250950e-01 9.20897424e-02 -1.38601780e+00
-5.70296161e-02 -1.15956128e+00 -3.38463008e-01 9.00717497e-01
-1.37683105e+00 -8.57925177e-01 -6.65716588e-01 3.80112201e-01
7.26225495e-01 -2.33857796e-01 3.13096941e-01 2.48310000e-01
-3.66850793e-01 3.88677508e-01 2.77498543e-01 4.46656853e-01
8.70550811e-01 -1.26524234e+00 5.30367792e-01 1.08264971e+00
-1.44191146e-01 8.99899423e-01 9.47487354e-01 -3.98067355e-01
-1.66298544e+00 -8.46272409e-01 4.87297535e-01 -1.73553616e-01
8.74435663e-01 1.22419961e-01 -9.21202660e-01 3.12201947e-01
4.52977479e-01 1.40394703e-01 6.95022866e-02 -5.95379174e-01
-2.13134766e-01 -3.58614385e-01 -9.66284931e-01 7.29606092e-01
8.01049471e-01 -6.09649658e-01 -8.43585193e-01 2.11069658e-01
3.20632637e-01 -7.12822199e-01 -9.19090807e-01 5.42307377e-01
7.94496715e-01 -1.02084494e+00 1.12842834e+00 -2.08062381e-01
2.35574767e-01 -5.83664596e-01 6.15415350e-02 -1.34328878e+00
-4.82246935e-01 -8.58174622e-01 -2.59704143e-01 1.13827562e+00
-2.74867326e-01 -8.14408004e-01 4.91797328e-01 5.15481949e-01
-1.96752653e-01 -1.80429608e-01 -8.26555431e-01 -1.10841131e+00
-2.76608825e-01 -1.31756114e-02 4.45701122e-01 1.00598550e+00
-8.08768868e-02 4.17307168e-02 -4.67461467e-01 2.86179960e-01
8.82006168e-01 2.70322859e-01 6.06689572e-01 -9.57621336e-01
-1.86893661e-02 -5.97202897e-01 -8.91640007e-01 -1.33059502e+00
5.05671538e-02 -3.80801439e-01 -2.50395183e-02 -1.25871766e+00
3.45211446e-01 1.62995428e-01 -2.69926578e-01 -1.50502503e-01
-2.42084846e-01 2.98717141e-01 6.19572103e-02 4.76240069e-01
-2.82192171e-01 3.43151897e-01 1.56020892e+00 -4.77977723e-01
3.34043242e-02 1.14604965e-01 -4.02461469e-01 6.64567709e-01
1.01347280e+00 1.26384586e-01 -1.48826346e-01 -5.39686739e-01
-4.97814298e-01 -2.50612855e-01 6.04986191e-01 -1.26595044e+00
3.55922103e-01 -1.71980217e-01 8.51992667e-01 -4.33778226e-01
3.14994276e-01 -7.68167913e-01 9.88025144e-02 6.73025310e-01
-1.39367446e-01 1.93469658e-01 8.78868625e-02 2.63159901e-01
-3.68084997e-01 -5.00887692e-01 9.98748064e-01 -5.23216091e-02
-8.57032657e-01 -5.74953035e-02 -4.88732219e-01 -4.78040755e-01
8.64594936e-01 -3.78987908e-01 -9.43989933e-01 -5.15191317e-01
-2.67435253e-01 -4.38610494e-01 5.54004371e-01 5.57063460e-01
1.05093920e+00 -1.34517026e+00 -3.46450090e-01 2.95913637e-01
-1.00629970e-01 -4.96301621e-01 2.51946896e-01 1.09983873e+00
-1.00651836e+00 7.77665913e-01 -3.31476867e-01 -7.41019964e-01
-1.53412414e+00 8.46271396e-01 4.77659464e-01 3.29155356e-01
-1.27844557e-01 4.35145736e-01 6.31531514e-03 3.36001068e-01
1.69009969e-01 -7.79834330e-01 1.58147402e-02 -3.49270105e-01
6.02304637e-01 5.69028437e-01 1.21103913e-01 -7.54887640e-01
-3.48036855e-01 8.54182661e-01 -2.42778718e-01 8.74278471e-02
1.09288085e+00 -1.60762906e-01 -3.40579271e-01 8.28111470e-02
1.21869171e+00 -2.07722381e-01 -1.26578057e+00 -5.72932102e-02
1.48946747e-01 -9.60058868e-01 -8.52911770e-02 -4.40340638e-01
-1.08354700e+00 7.15731382e-01 8.56667638e-01 4.14246231e-01
1.34310007e+00 -1.15744308e-01 5.79965234e-01 7.15765297e-01
1.24151908e-01 -9.68759179e-01 4.44748789e-01 6.53481424e-01
1.09927797e+00 -1.04377651e+00 4.31788675e-02 -3.44578356e-01
-4.05999959e-01 1.49941540e+00 6.18770361e-01 -1.81638613e-01
4.75776464e-01 1.27197608e-01 2.36628711e-01 2.45058835e-01
-3.57847661e-01 -3.09407234e-01 5.84379673e-01 6.40220881e-01
5.43300092e-01 -3.48228246e-01 -3.66765559e-01 -1.49290442e-01
4.57153395e-02 2.21671626e-01 8.06853354e-01 9.99867380e-01
-4.63447779e-01 -8.70562673e-01 -7.89268017e-01 3.30379039e-01
-9.31919932e-01 1.26031846e-01 -3.84121060e-01 4.63622242e-01
-4.25324291e-01 9.23495114e-01 5.59286177e-02 -7.02282041e-02
2.36649543e-01 -2.14881569e-01 1.02073109e+00 1.19419187e-01
1.21870138e-01 1.53699681e-01 -2.54459441e-01 -7.76458904e-02
-7.71865785e-01 -6.39158428e-01 -6.27290845e-01 -2.27931485e-01
-4.04233187e-01 -7.82085061e-02 6.64600968e-01 6.09347582e-01
2.04272181e-01 2.87680089e-01 6.50199771e-01 -1.08906734e+00
-1.00798988e+00 -1.26287520e+00 -6.48584008e-01 6.33511961e-01
7.57222116e-01 -7.65752017e-01 -4.53602195e-01 4.78048593e-01] | [11.608992576599121, -2.752147674560547] |
b9148382-5bae-49be-b710-6d4818d21408 | acdmsr-accelerated-conditional-diffusion | 2307.00781 | null | https://arxiv.org/abs/2307.00781v1 | https://arxiv.org/pdf/2307.00781v1.pdf | ACDMSR: Accelerated Conditional Diffusion Models for Single Image Super-Resolution | Diffusion models have gained significant popularity in the field of image-to-image translation. Previous efforts applying diffusion models to image super-resolution (SR) have demonstrated that iteratively refining pure Gaussian noise using a U-Net architecture trained on denoising at various noise levels can yield satisfactory high-resolution images from low-resolution inputs. However, this iterative refinement process comes with the drawback of low inference speed, which strongly limits its applications. To speed up inference and further enhance the performance, our research revisits diffusion models in image super-resolution and proposes a straightforward yet significant diffusion model-based super-resolution method called ACDMSR (accelerated conditional diffusion model for image super-resolution). Specifically, our method adapts the standard diffusion model to perform super-resolution through a deterministic iterative denoising process. Our study also highlights the effectiveness of using a pre-trained SR model to provide the conditional image of the given low-resolution (LR) image to achieve superior high-resolution results. We demonstrate that our method surpasses previous attempts in qualitative and quantitative results through extensive experiments conducted on benchmark datasets such as Set5, Set14, Urban100, BSD100, and Manga109. Moreover, our approach generates more visually realistic counterparts for low-resolution images, emphasizing its effectiveness in practical scenarios. | ['Yanning Zhang', 'In So Kweon', 'Yu Zhu', 'Jinqiu Sun', 'Kang Zhang', 'Pham Xuan Trung', 'Axi Niu'] | 2023-07-03 | null | null | null | null | ['image-super-resolution', 'image-to-image-translation', 'super-resolution', 'image-to-image-translation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'miscellaneous'] | [ 6.18452132e-01 -1.25861457e-02 -3.05397548e-02 -6.88816234e-02
-1.02492619e+00 -6.84307888e-02 7.46376395e-01 -5.71060836e-01
-2.86321938e-01 8.28598082e-01 5.40987730e-01 5.22100441e-02
-1.11794747e-01 -9.91050720e-01 -6.22753322e-01 -6.93518162e-01
2.75808871e-01 -3.63913216e-02 3.83084297e-01 -3.99513990e-01
2.44421050e-01 5.16383529e-01 -1.39325821e+00 4.36594367e-01
9.74834442e-01 5.66605747e-01 6.25259161e-01 7.59461045e-01
2.25488073e-03 1.00889945e+00 -4.72563475e-01 -2.21611202e-01
2.53460437e-01 -7.61037827e-01 -8.52353632e-01 7.54505470e-02
5.97408056e-01 -7.16905057e-01 -3.31936300e-01 1.31593430e+00
6.35067999e-01 1.74842596e-01 5.61623693e-01 -4.42162991e-01
-1.55359268e+00 6.98205113e-01 -1.09646499e+00 6.78012550e-01
3.55271846e-01 -2.91398186e-02 6.99304342e-01 -1.02388263e+00
1.06995821e+00 1.47390842e+00 7.11824358e-01 6.95751905e-01
-1.68064117e+00 -5.73618412e-01 -1.39252558e-01 4.78134155e-02
-1.43774581e+00 -4.10700232e-01 7.63122261e-01 -7.37208501e-02
7.32687056e-01 4.34596725e-02 8.43067095e-02 1.24195254e+00
1.88974872e-01 5.12979031e-01 1.59821820e+00 -3.31562579e-01
1.44354254e-01 -9.53286812e-02 -3.44883770e-01 3.50633144e-01
7.93781355e-02 3.15861374e-01 -5.59103847e-01 1.95867702e-01
1.84928858e+00 -2.52458543e-01 -3.41974646e-01 7.71981850e-02
-1.21360922e+00 6.87481821e-01 6.72075868e-01 6.74064159e-01
-7.13929296e-01 -5.12493867e-03 -4.33994755e-02 2.42913544e-01
9.59054649e-01 2.25982577e-01 1.91051047e-02 1.67241737e-01
-1.30677068e+00 6.84528723e-02 8.35124180e-02 6.58085823e-01
5.67107201e-01 3.95170778e-01 -2.61195004e-01 1.01955771e+00
-1.04213789e-01 3.87065589e-01 2.83013850e-01 -1.58417487e+00
1.19844869e-01 -7.93401375e-02 3.07484299e-01 -1.06871247e+00
-6.12345673e-02 -6.66364551e-01 -1.50900257e+00 6.02018416e-01
1.28190666e-01 8.11686590e-02 -1.09527385e+00 1.57309759e+00
1.94593772e-01 3.87037635e-01 2.95053184e-01 1.17462158e+00
9.13947463e-01 6.76353037e-01 1.15754560e-01 -3.58696491e-01
1.01607418e+00 -8.87855649e-01 -9.91857529e-01 9.44459587e-02
-8.23929161e-02 -8.54857206e-01 9.93451953e-01 4.80923504e-01
-1.57398307e+00 -8.51476789e-01 -9.82097983e-01 -4.01179135e-01
8.40757340e-02 -7.64974952e-02 2.74304509e-01 3.68496776e-01
-1.60359144e+00 8.87501955e-01 -5.72707951e-01 -2.57103503e-01
6.74837112e-01 -1.21990405e-02 -3.13554496e-01 -4.17079836e-01
-1.27419543e+00 1.01936960e+00 1.98139042e-01 -9.39015970e-02
-9.56680596e-01 -8.37678313e-01 -7.44955003e-01 -1.70478791e-01
1.98468179e-01 -9.34382021e-01 8.99577975e-01 -9.43167448e-01
-1.70369089e+00 6.30649745e-01 -2.71524340e-01 -6.17583930e-01
6.76741540e-01 -1.08720504e-01 -4.65464950e-01 4.18282926e-01
1.48327738e-01 7.77695119e-01 1.20202768e+00 -1.71890712e+00
-6.67293310e-01 -1.16887316e-01 1.46246120e-01 2.79004902e-01
-7.49728456e-02 2.12716937e-01 -4.38848555e-01 -8.61795902e-01
1.66180253e-01 -3.81062239e-01 -5.62937379e-01 -6.19541779e-02
-4.51565348e-02 2.05410004e-01 7.77163327e-01 -9.73030984e-01
1.10542977e+00 -2.04502201e+00 3.62203091e-01 -1.03624724e-01
5.77242315e-01 1.46095932e-01 -3.38802695e-01 -1.16459884e-01
1.03381770e-02 2.15140074e-01 -4.70651507e-01 -4.11366612e-01
-6.31885171e-01 2.13489637e-01 -1.84814230e-01 2.92689353e-01
3.04547399e-01 9.78134394e-01 -1.00918984e+00 -7.32859969e-01
4.27309602e-01 1.35899234e+00 -3.78259897e-01 4.38266098e-02
5.08752689e-02 8.45123529e-01 -1.38908282e-01 3.73996377e-01
9.75313663e-01 -4.74978626e-01 -1.77846447e-01 -4.96896386e-01
-2.38589793e-01 -3.09052855e-01 -1.33753216e+00 1.68241513e+00
-7.09041297e-01 5.77483058e-01 3.01751107e-01 -4.33455050e-01
9.10034955e-01 7.32137188e-02 4.47334915e-01 -1.11187506e+00
-2.40785971e-01 -4.09941413e-02 -4.37076181e-01 -2.65961826e-01
8.76691282e-01 -3.09254855e-01 5.57943881e-01 3.61534387e-01
-6.87931553e-02 -1.22010678e-01 1.62098587e-01 4.22809064e-01
7.38994896e-01 4.08153266e-01 7.89825916e-02 -1.22072078e-01
5.67588747e-01 -1.90726534e-01 3.11010629e-01 8.18358481e-01
-4.41523343e-02 1.25558126e+00 -2.59396899e-02 -1.17419049e-01
-1.46788919e+00 -1.30831659e+00 -1.54052347e-01 8.09694529e-01
3.19573075e-01 -8.95241462e-03 -9.23098683e-01 -2.16812223e-01
-5.52590191e-01 7.02236414e-01 -7.87135661e-01 2.95813978e-01
-5.67201018e-01 -8.92891705e-01 3.77191603e-01 5.57949603e-01
9.88078177e-01 -1.05459499e+00 -2.76456594e-01 1.79881006e-01
-6.48005545e-01 -1.28915942e+00 -3.73053402e-01 -3.43929648e-01
-9.65365112e-01 -5.65692544e-01 -1.30753088e+00 -7.77421474e-01
6.54015958e-01 6.34436846e-01 1.09276009e+00 -1.27627179e-01
-1.60503596e-01 1.94524333e-01 -1.74075454e-01 3.41347158e-01
-7.32237637e-01 -2.31087536e-01 -5.47710583e-02 2.81501804e-02
-1.78483352e-01 -6.13317907e-01 -8.04724038e-01 2.59877354e-01
-1.28758836e+00 3.36450577e-01 7.95296729e-01 8.29454243e-01
1.01226807e+00 5.46660006e-01 5.93679845e-01 -8.09143126e-01
9.75270569e-01 -2.92809129e-01 -3.43911529e-01 5.60146272e-02
-8.30201745e-01 6.29905015e-02 4.92399484e-01 -5.94200611e-01
-1.72183919e+00 -2.91668624e-01 -2.23196000e-01 -5.12130678e-01
-4.25855629e-02 3.39722633e-01 2.31052145e-01 -2.54311234e-01
1.00055397e+00 5.45630395e-01 6.37924075e-02 -5.33262014e-01
7.73464620e-01 3.41901392e-01 8.77844214e-01 -2.52302170e-01
9.97916877e-01 9.28419530e-01 -4.86106351e-02 -8.80039334e-01
-8.86518061e-01 -6.48074001e-02 -7.47665823e-01 -1.60081193e-01
1.00294018e+00 -1.18411803e+00 -2.82213628e-01 2.57037967e-01
-1.02890575e+00 -4.05943245e-01 -3.98045778e-01 3.05559963e-01
-5.96858025e-01 5.57214618e-01 -1.04688692e+00 -6.86095715e-01
-3.96863788e-01 -1.06004298e+00 1.01456213e+00 3.63197654e-01
-9.26223174e-02 -1.00749171e+00 -1.68811396e-01 4.75098997e-01
1.10302222e+00 2.57582039e-01 6.14138365e-01 3.45404506e-01
-8.20429802e-01 4.35144603e-01 -8.08026135e-01 5.91205120e-01
1.50034055e-01 -1.66796416e-01 -7.66728163e-01 -1.78918198e-01
1.10559434e-01 -9.39653590e-02 9.78514791e-01 7.90460587e-01
9.48100448e-01 -3.42321508e-02 5.77896461e-02 6.98666036e-01
1.74250376e+00 -2.52175570e-01 1.10283470e+00 4.58398730e-01
6.62271023e-01 2.58309275e-01 5.35643578e-01 5.98984547e-02
3.05406541e-01 4.58796114e-01 2.52342999e-01 -6.91152036e-01
-1.01592386e+00 -1.80152878e-01 2.18446016e-01 4.98594433e-01
-7.13273585e-01 8.54732171e-02 -3.11420381e-01 6.64067268e-01
-1.51953208e+00 -1.28085363e+00 -1.14558786e-01 2.00227070e+00
1.06377256e+00 3.37805897e-02 -3.19858603e-02 -4.11611497e-02
7.86043167e-01 3.03011745e-01 -4.82817262e-01 7.08052190e-03
-6.64052069e-01 1.96638063e-01 3.83641899e-01 8.80078912e-01
-8.07689607e-01 1.18732166e+00 6.90412760e+00 1.11515546e+00
-8.93308461e-01 4.80687231e-01 9.94404197e-01 6.36885539e-02
-4.46365416e-01 -4.55338061e-01 -6.05874777e-01 8.74021277e-02
8.00890625e-01 -1.33062631e-01 7.35262692e-01 4.45482403e-01
5.71116626e-01 -1.40608341e-01 -4.92310375e-01 9.65340853e-01
1.02259614e-01 -1.61358285e+00 3.49361658e-01 -5.33439070e-02
1.30992353e+00 3.79850455e-02 5.47831118e-01 -8.41467679e-02
5.79136193e-01 -1.23363495e+00 4.00784850e-01 6.79597557e-01
1.16966927e+00 -7.91391313e-01 5.69975674e-01 1.06057443e-01
-1.00373971e+00 1.15316153e-01 -6.26732767e-01 2.55464405e-01
3.71807933e-01 6.76587939e-01 -4.51712787e-01 6.82020724e-01
9.64778662e-01 7.90140450e-01 -3.58004391e-01 5.55412114e-01
-3.13683808e-01 4.51745749e-01 6.20866828e-02 7.45898008e-01
2.62650084e-02 -3.91319305e-01 5.73396742e-01 1.21142566e+00
3.71350348e-01 3.37441117e-01 -2.87435293e-01 1.35054815e+00
-7.16296881e-02 -3.78484353e-02 -4.45513517e-01 2.92026728e-01
1.29048660e-01 1.22864556e+00 -8.40652049e-01 -4.90383327e-01
-3.89826089e-01 1.35462677e+00 1.61557332e-01 8.04584026e-01
-8.09563696e-01 1.40151128e-01 4.43765104e-01 2.76444435e-01
4.87387478e-01 -3.33607763e-01 -5.91848135e-01 -1.04051781e+00
-3.13805014e-01 -9.28317606e-01 1.09401792e-01 -1.09222651e+00
-1.18348992e+00 1.03804445e+00 -6.94166869e-02 -1.11057377e+00
-2.03048214e-01 8.13728869e-02 -8.68201703e-02 1.26921022e+00
-1.85628831e+00 -1.36467671e+00 -4.34816748e-01 7.22183645e-01
8.99891675e-01 2.23927468e-01 4.63342756e-01 3.23327273e-01
-2.01754570e-01 2.02438965e-01 2.17686176e-01 -1.39878750e-01
6.72795057e-01 -1.10414708e+00 5.40373981e-01 1.20487106e+00
-9.92925614e-02 5.67655861e-01 9.88546789e-01 -9.29377198e-01
-9.20394003e-01 -1.12600887e+00 6.20028853e-01 -3.79983515e-01
5.27109206e-01 1.56146467e-01 -1.20419168e+00 3.79674733e-01
4.00601208e-01 -1.23824785e-02 -3.08173196e-03 -2.88989216e-01
-1.88526139e-01 1.48212984e-01 -1.46117723e+00 8.81013274e-01
1.21933234e+00 -5.63658297e-01 -3.13853890e-01 -1.20129190e-01
8.97261024e-01 -5.24605930e-01 -1.33175015e+00 3.47190112e-01
2.15927646e-01 -1.10826576e+00 1.53610122e+00 -3.56759690e-02
1.02788997e+00 -4.42523599e-01 -1.43580243e-01 -1.18555045e+00
-8.86981487e-01 -5.72613239e-01 -2.67483205e-01 1.21326530e+00
1.97081909e-01 -3.07490140e-01 4.93009835e-01 4.59755391e-01
3.92836660e-01 -4.83268857e-01 -7.06102490e-01 -5.85274220e-01
1.14155054e-01 -1.94360375e-01 3.76335651e-01 9.94749129e-01
-8.15124869e-01 2.69340068e-01 -7.39438117e-01 3.28427583e-01
1.23645627e+00 -3.96196432e-02 4.93527830e-01 -9.20433462e-01
-9.51560289e-02 -5.10800540e-01 8.03815052e-02 -1.16027057e+00
-1.78767085e-01 -5.35447836e-01 -1.48144871e-01 -1.89097393e+00
4.01152849e-01 -1.16472207e-01 -2.95772910e-01 -3.54092335e-03
-2.65159398e-01 9.10235941e-01 -7.70112798e-02 4.04659450e-01
-4.90110248e-01 4.27514434e-01 1.90773165e+00 -5.51522709e-02
-1.97694704e-01 -2.34335348e-01 -9.92059350e-01 5.87992072e-01
5.98266244e-01 -2.10151881e-01 -4.60623324e-01 -4.86885279e-01
8.77968818e-02 2.31700480e-01 3.48513067e-01 -6.87937737e-01
1.67329475e-01 -1.41778588e-01 6.83694899e-01 -4.87703860e-01
3.22114050e-01 -6.75499916e-01 1.99750349e-01 2.41195261e-01
-5.83757341e-01 -2.33902380e-01 4.08374965e-02 6.96595669e-01
-2.24091172e-01 2.84912866e-02 1.18593204e+00 -2.40478516e-01
-9.47967231e-01 2.42084548e-01 -1.77434579e-01 -1.41613662e-01
6.63413167e-01 -4.53434318e-01 -3.13118130e-01 -6.80376649e-01
-8.69081259e-01 -2.78343976e-01 6.33497179e-01 4.69375610e-01
9.43734586e-01 -1.23296440e+00 -1.23292744e+00 3.24291922e-02
-3.85615557e-01 8.72191191e-02 6.28667414e-01 7.48115361e-01
-3.56503516e-01 -1.16323447e-02 -2.91847616e-01 -5.93005061e-01
-1.28880715e+00 7.03256726e-01 2.84712225e-01 -2.84823477e-01
-1.11435246e+00 7.33107269e-01 3.84889334e-01 1.10791758e-01
-7.38098547e-02 -1.71732187e-01 -3.30346107e-01 -3.92821193e-01
9.25166965e-01 5.03468812e-01 -3.96741122e-01 -8.83557677e-01
1.04975961e-01 7.60218680e-01 -2.77441442e-01 -2.83469081e-01
1.53575790e+00 -8.05460215e-01 -1.12019472e-01 4.32783663e-02
8.75765324e-01 -3.96265741e-03 -1.69031394e+00 -7.30494261e-01
-3.16260666e-01 -5.96748233e-01 5.71083128e-01 -8.96313906e-01
-1.29629087e+00 4.99849737e-01 7.33403444e-01 -2.40333769e-02
1.48466349e+00 -5.61738387e-02 7.82762289e-01 -2.43909106e-01
3.69904965e-01 -1.02838302e+00 2.46262968e-01 2.71387964e-01
1.09514010e+00 -1.43872130e+00 1.80149466e-01 -4.15440291e-01
-6.77160025e-01 8.80477548e-01 4.24252689e-01 -1.31697521e-01
3.18541825e-01 4.95071471e-01 1.82298675e-01 -1.76162682e-02
-4.86190915e-01 -3.01979154e-01 2.71169275e-01 8.53951871e-01
3.95381838e-01 -2.53691763e-01 -2.40133375e-01 1.91318586e-01
1.12943262e-01 3.80420506e-01 8.06355417e-01 4.32410389e-01
-4.47679460e-01 -6.65433586e-01 -6.84777737e-01 3.52893993e-02
-6.30969584e-01 -4.39499140e-01 2.09103730e-02 5.93011498e-01
-9.64645892e-02 9.89213407e-01 -1.33070901e-01 -1.88289925e-01
4.32799272e-02 -5.66234708e-01 6.12379849e-01 -2.05918685e-01
-1.87508926e-01 3.56398076e-01 -2.61388361e-01 -7.17835903e-01
-8.72617900e-01 -3.51965487e-01 -1.06347954e+00 -4.12614256e-01
-7.87120536e-02 -2.41312921e-01 6.09935343e-01 6.86718643e-01
4.23167497e-01 9.00764883e-01 3.63048106e-01 -1.07384646e+00
-3.08885068e-01 -1.06206381e+00 -5.04383802e-01 4.95373428e-01
4.75813627e-01 -3.89695257e-01 -2.59618640e-01 3.77640128e-01] | [11.203714370727539, -1.9890614748001099] |
6dbf7e65-606a-4047-826a-f8171ea28777 | csi-based-outdoor-localization-for-massive | 1806.07447 | null | http://arxiv.org/abs/1806.07447v1 | http://arxiv.org/pdf/1806.07447v1.pdf | CSI-based Outdoor Localization for Massive MIMO: Experiments with a Learning Approach | We report on experimental results on the use of a learning-based approach to
infer the location of a mobile user of a cellular network within a cell, for a
5G-type Massive multiple input, multiple output (MIMO) system. We describe how
the sample spatial covariance matrix computed from the CSI can be used as the
input to a learning algorithm which attempts to relate it to user location. We
discuss several learning approaches, and analyze in depth the application of
extreme learning machines, for which theoretical approximate performance
benchmarks are available, to the localization problem. We validate the proposed
approach using experimental data collected on a Huawei 5G testbed, provide some
performance and robustness benchmarks, and discuss practical issues related to
the deployment of such a technique in 5G networks. | ['Luis García Ordóñez', 'Li Bojie', 'Paul Ferrand', 'Alexis Decurninge', 'Zhang Wei', 'Maxime Guillaud', 'He Gaoning'] | 2018-06-19 | null | null | null | null | ['outdoor-localization'] | ['robots'] | [-4.78757918e-01 2.22628668e-01 -1.40690550e-01 -1.40931100e-01
-8.50985885e-01 -3.33346844e-01 4.57400046e-02 -1.02026656e-01
1.00965217e-01 1.22235703e+00 -2.31436133e-01 -1.03777635e+00
-5.41489005e-01 -5.10589302e-01 -5.28075397e-01 -9.79968309e-01
-1.06131816e+00 4.25910711e-01 -4.49277997e-01 3.13448668e-01
1.24470755e-01 6.74002230e-01 -3.68429661e-01 -2.79338006e-02
3.09235871e-01 1.51381254e+00 1.74331367e-01 1.06522357e+00
3.12443316e-01 9.93717372e-01 -8.69550109e-01 -4.92110141e-02
4.26801682e-01 -3.00322741e-01 -6.23772264e-01 3.37574661e-01
-7.08656460e-02 9.71967503e-02 -4.96177226e-01 4.83452588e-01
9.63200390e-01 -2.74217159e-01 8.42362940e-01 -1.31974280e+00
-1.47773013e-01 6.71869040e-01 -1.13858618e-01 5.03742874e-01
-7.04004616e-02 -3.32339853e-01 8.10551584e-01 -1.03982723e+00
6.88900356e-04 8.45890403e-01 9.28956449e-01 -1.93797752e-01
-6.92638099e-01 -3.67780358e-01 -2.18137026e-01 9.64740068e-02
-1.71941435e+00 -8.38731766e-01 3.36980164e-01 -3.39763254e-01
5.81212759e-01 1.33225471e-01 6.69497967e-01 3.74743372e-01
1.77926138e-01 5.85863113e-01 5.63304245e-01 -9.01614904e-01
7.97310948e-01 3.51246864e-01 -5.10921776e-01 9.32236195e-01
3.36260438e-01 5.35584241e-02 1.54013634e-01 -7.40614831e-01
9.50094759e-01 -2.27464616e-01 -2.46501699e-01 -4.98933971e-01
-7.94841409e-01 6.01318181e-01 4.78978127e-01 3.86004299e-01
-5.93551338e-01 6.06306911e-01 -4.05551076e-01 5.82881331e-01
3.35562050e-01 2.67258227e-01 -5.79140007e-01 8.72259587e-02
-1.12137330e+00 -3.53569210e-01 1.21697056e+00 1.39613605e+00
6.58344924e-01 4.99201000e-01 -2.38907814e-01 4.66921836e-01
7.58387566e-01 5.76596141e-01 -9.38224271e-02 -1.11364043e+00
4.82006997e-01 -2.08986223e-01 3.79397720e-01 -6.88004613e-01
-6.86983466e-01 -1.46311760e+00 -8.64318252e-01 -7.15986416e-02
4.17779028e-01 -1.65480316e+00 -2.92246789e-01 1.20541501e+00
-1.62742883e-01 1.12505162e+00 4.58851248e-01 4.20155555e-01
2.85230398e-01 4.17856812e-01 -5.13803363e-01 -5.25916278e-01
4.12740797e-01 -7.40930736e-01 -3.37635368e-01 -1.76192507e-01
8.80917430e-01 -5.17368913e-01 -1.36308922e-02 8.05963799e-02
-1.04441130e+00 -1.66873395e-01 -1.15303826e+00 1.00243843e+00
-9.67963487e-02 6.10347748e-01 5.29707193e-01 1.25071788e+00
-1.33322418e+00 3.62600237e-01 -4.58975911e-01 -6.77762687e-01
7.31234670e-01 8.57999802e-01 3.76683086e-01 2.78987348e-01
-9.71732914e-01 4.84223813e-01 2.89495528e-01 6.36747256e-02
-5.11291325e-01 -5.29354751e-01 -2.83595949e-01 4.86680865e-01
-5.42116212e-03 -6.58250868e-01 1.21476948e+00 -5.06161749e-01
-1.40244985e+00 2.11493954e-01 -2.09514201e-01 -7.34797537e-01
2.55750477e-01 4.61581409e-01 -9.10747647e-01 1.13656759e-01
-1.82990775e-01 -2.19582260e-01 5.37583709e-01 -1.27105975e+00
-1.24921679e+00 -5.79905733e-02 2.81728115e-02 -3.37766767e-01
-2.00035751e-01 -4.46502775e-01 -1.54725477e-01 -4.62993711e-01
2.50426710e-01 -1.18368173e+00 -7.01229632e-01 -4.91069406e-01
-4.23769146e-01 3.12998682e-01 9.18123007e-01 -3.10131878e-01
1.31143093e+00 -1.86779189e+00 -2.12369233e-01 1.15364945e+00
5.95500171e-02 5.48636727e-02 2.50362784e-01 6.27146721e-01
1.31530955e-01 1.07634924e-01 5.77233195e-01 -2.51533538e-01
-4.51266319e-01 -8.70278105e-02 -3.20105970e-01 5.03510654e-01
-6.04539633e-01 8.25151920e-01 -8.15768600e-01 -7.64148533e-02
-1.85293593e-02 4.64284793e-03 -6.24216139e-01 6.18627816e-02
2.64789343e-01 3.79022688e-01 -6.68937027e-01 8.12799871e-01
3.98115218e-01 -7.19631910e-01 4.47176427e-01 -1.86662555e-01
2.23024428e-01 -5.49314737e-01 -1.00504518e+00 8.45085382e-01
-8.94432306e-01 8.65141451e-01 -3.24499980e-02 -1.24110639e+00
7.42928922e-01 8.48136663e-01 7.60225356e-01 1.26067311e-01
4.14336622e-01 4.18437123e-01 7.85972625e-02 -3.25423360e-01
-3.76280755e-01 8.19092914e-02 -5.94670000e-03 5.66412807e-01
4.69408810e-01 4.90214437e-01 -2.38240853e-01 1.84769243e-01
1.20316124e+00 -7.47044265e-01 5.74019313e-01 -3.58807683e-01
6.54046655e-01 -7.98355579e-01 4.77320015e-01 1.38296127e+00
-3.08849126e-01 8.29666704e-02 5.41775584e-01 -3.84890229e-01
-8.95405412e-01 -1.19671595e+00 1.20724209e-01 5.64508200e-01
-4.64920029e-02 -1.85521752e-01 -6.76561236e-01 -3.59346807e-01
1.47101805e-01 7.13413835e-01 -2.04647169e-01 -3.01498268e-02
-7.91153871e-03 -1.21688080e+00 5.83551764e-01 3.05285871e-01
6.45011425e-01 -1.85854718e-01 5.07849269e-02 4.41634834e-01
1.47798015e-02 -1.31941676e+00 1.08912148e-01 2.80038685e-01
-6.98822796e-01 -6.39454842e-01 -6.47837043e-01 -1.04374349e+00
4.81491029e-01 4.28772457e-02 7.23728836e-01 7.53893480e-02
2.76679814e-01 6.36827588e-01 7.27147460e-02 -4.44800019e-01
-2.69308537e-01 3.43195289e-01 5.77173889e-01 4.56421882e-01
2.17780441e-01 -9.93229687e-01 -3.96285415e-01 3.62056196e-01
3.72446954e-01 -6.92111671e-01 8.33536565e-01 5.99137962e-01
6.05726168e-02 2.84111083e-01 1.18296361e+00 -8.74253631e-01
5.55397749e-01 -9.30092394e-01 -7.19665647e-01 5.00961125e-01
-4.42391694e-01 -2.15334177e-01 6.41562045e-01 1.53375696e-02
-5.65966010e-01 1.44077420e-01 -6.31036833e-02 -2.76915312e-01
1.99264303e-01 6.73325181e-01 -2.82268375e-01 -1.08378375e+00
7.67442167e-01 -3.39240097e-02 -2.84254521e-01 -1.05561525e-01
1.84398696e-01 1.42227757e+00 1.30764350e-01 -4.86894250e-01
1.07554591e+00 4.22872096e-01 3.66833955e-01 -1.28753102e+00
-7.77609825e-01 -4.55588460e-01 -6.33489192e-01 -2.45922133e-01
-7.96241537e-02 -1.16305804e+00 -7.01709509e-01 4.63872068e-02
-6.41812742e-01 -4.12253439e-01 7.57358270e-03 7.32424498e-01
-1.14476454e+00 -7.78524727e-02 -6.66489184e-01 -9.65301335e-01
1.23948213e-02 -4.54278171e-01 2.94940978e-01 2.75109649e-01
2.47377738e-01 -1.51703978e+00 -2.98509896e-01 1.54909328e-01
7.74475574e-01 -1.25323847e-01 9.89021957e-01 -5.94431758e-01
-9.06349957e-01 -3.55808079e-01 -4.97077731e-03 -2.20677048e-01
1.54383779e-01 -3.75716865e-01 -1.05634093e+00 -7.31901109e-01
-9.28324759e-02 3.58291537e-01 8.16591978e-02 9.83159959e-01
1.05280960e+00 -4.49177474e-01 -1.04898226e+00 7.56797969e-01
1.62888765e+00 9.31298435e-02 5.76463461e-01 -1.12608122e-02
2.25739673e-01 -2.92923123e-01 4.40993607e-01 8.75662327e-01
1.64319053e-01 5.29340863e-01 3.81023943e-01 -3.36691678e-01
2.59354621e-01 2.19153017e-01 -1.64159283e-01 4.46211576e-01
-9.81299952e-02 -7.38218009e-01 -6.84608996e-01 1.54461756e-01
-1.96803629e+00 -9.79364157e-01 1.92725241e-01 2.17486429e+00
-1.42568424e-01 1.71580195e-01 1.76865280e-01 5.87913841e-02
6.91919684e-01 -3.81086737e-01 -3.31395328e-01 3.75235640e-02
-1.05673119e-01 -3.35682303e-01 1.17492640e+00 7.22169042e-01
-1.22330606e+00 7.96231747e-01 7.25011969e+00 7.38840520e-01
-1.11435342e+00 1.82384402e-01 1.07287920e+00 -6.07999526e-02
3.48590612e-01 -4.87519726e-02 -5.37541628e-01 3.49745452e-01
1.41343832e+00 -2.37835377e-01 3.95372987e-01 7.18671560e-01
7.67869711e-01 1.10960841e-01 -1.20889115e+00 1.36984742e+00
-3.26965541e-01 -1.87993348e+00 -2.41931036e-01 5.63702941e-01
9.86312747e-01 3.79439622e-01 2.22173229e-01 3.75545263e-01
6.43809065e-02 -7.83225238e-01 2.75953889e-01 7.55457461e-01
4.89638597e-01 -1.15369534e+00 1.06660581e+00 3.79246145e-01
-1.22762048e+00 -1.07227325e+00 -2.99948394e-01 -2.14167297e-01
3.80621612e-01 9.47032869e-01 -1.37417877e+00 4.00778919e-01
1.26898751e-01 3.24908435e-01 -4.04165566e-01 2.03715849e+00
2.69183636e-01 1.05735075e+00 -3.56743097e-01 -1.66685119e-01
3.66705894e-01 -1.72042232e-02 6.02819026e-01 1.15718114e+00
1.13984108e+00 1.55533910e-01 4.46030468e-01 3.57278705e-01
-7.63783082e-02 1.18969634e-01 -5.45897424e-01 3.65489483e-01
9.46443141e-01 1.22638452e+00 -5.17704666e-01 -3.07025671e-01
-6.66465580e-01 6.68568552e-01 8.96207243e-02 1.18294585e+00
-5.54027796e-01 -6.92949653e-01 5.89576364e-01 2.78156251e-01
5.85546315e-01 -4.71270680e-01 -3.46385658e-01 -8.30266833e-01
-3.83342505e-01 -2.70720333e-01 2.83018947e-01 -5.40626228e-01
-1.01910663e+00 2.31208906e-01 -5.69940031e-01 -1.62773502e+00
-7.76764214e-01 -4.75244969e-01 -6.73665941e-01 8.50040019e-01
-1.00997782e+00 -8.45281780e-01 1.40433609e-01 3.53124291e-01
-3.35886143e-02 -1.26523292e+00 9.86266673e-01 5.07806122e-01
-4.01933342e-01 8.81799638e-01 8.46289515e-01 3.92327100e-01
9.10843983e-02 -1.26875949e+00 5.84833845e-02 5.70572972e-01
3.70717227e-01 3.24158162e-01 5.32918513e-01 -2.13514522e-01
-1.13972998e+00 -1.28029156e+00 6.28018975e-01 -1.01382501e-01
5.26322424e-01 -1.74351498e-01 1.41568348e-01 1.16774237e+00
1.30490080e-01 6.38127923e-01 1.11311245e+00 3.18683863e-01
8.50527704e-01 -3.21070522e-01 -1.31987071e+00 7.81023741e-01
7.61027813e-01 -3.02954525e-01 1.89500839e-01 6.21489406e-01
-9.89343449e-02 -2.26773828e-01 -9.24452424e-01 6.19800799e-02
3.43440086e-01 -6.98192894e-01 1.17130065e+00 -6.46025121e-01
-9.03873205e-01 -3.46530199e-01 -4.66183543e-01 -1.68972743e+00
-8.06073785e-01 -1.02649844e+00 -7.87151873e-01 7.44983971e-01
6.91933215e-01 -5.62081754e-01 1.30388117e+00 -1.69236645e-01
2.52852410e-01 -8.53578746e-01 -1.03982329e+00 -8.40253651e-01
-6.19209278e-03 -1.24334596e-01 3.92626405e-01 5.50897121e-01
3.34212452e-01 4.79090780e-01 -4.05371159e-01 1.06612062e+00
8.15261602e-01 -1.48824126e-01 7.01368749e-01 -1.32926667e+00
-6.53344631e-01 -1.44469157e-01 -9.45645809e-01 -1.32149708e+00
-2.62245219e-02 -7.77532458e-01 -4.40565616e-01 -1.16090131e+00
-4.96627212e-01 -9.44192708e-01 -5.51444650e-01 -1.07446820e-01
4.77266461e-01 1.53624386e-01 -3.04993801e-02 -4.50960845e-02
-7.58189082e-01 1.50932506e-01 2.60256231e-01 -7.66572505e-02
-3.00798565e-01 1.36860371e+00 -7.39667177e-01 8.55338275e-01
1.03428328e+00 -1.34261819e-02 -4.62341845e-01 5.81244566e-02
-1.68242127e-01 5.43473959e-01 1.04695342e-01 -1.91168475e+00
6.22086942e-01 1.24020010e-01 1.01092029e+00 -4.12347913e-01
3.15121144e-01 -1.19590056e+00 -7.32109845e-02 5.28400421e-01
6.38026968e-02 -6.23643041e-01 -1.60319969e-01 8.06456327e-01
2.08954513e-01 -2.09764913e-01 7.20426977e-01 1.46189243e-01
-6.29859447e-01 5.83980441e-01 -9.46123421e-01 -4.86365855e-02
1.16982329e+00 -1.74225345e-01 3.98235738e-01 -1.26723945e+00
-1.35350394e+00 4.19218987e-01 -2.65149593e-01 -1.53728604e-01
4.43916082e-01 -1.45942044e+00 -6.77867174e-01 7.70654902e-02
-1.44833803e-01 -1.27012634e+00 -3.41819435e-01 7.80731022e-01
-3.16640556e-01 8.42602670e-01 1.40553772e-01 -6.57650054e-01
-9.87491190e-01 3.42949748e-01 1.32105315e+00 1.59149557e-01
-1.17385937e-02 5.12008190e-01 -5.11449456e-01 -1.51958853e-01
4.20193553e-01 7.25217313e-02 -1.23799689e-01 -4.45714504e-01
4.38509077e-01 6.07985556e-01 2.34420478e-01 -3.54023844e-01
-2.91704625e-01 3.48346025e-01 3.87363046e-01 -2.41250947e-01
8.99043083e-01 -9.17689681e-01 3.14925551e-01 4.75980669e-01
1.13368762e+00 2.57829517e-01 -9.70220149e-01 -5.93511879e-01
4.24374253e-01 -3.17153633e-01 2.25123614e-01 -6.42508686e-01
-1.32078648e+00 5.06811380e-01 1.21068776e+00 4.04548824e-01
7.60836899e-01 1.12401403e-01 3.34873915e-01 1.17183053e+00
1.05821776e+00 -1.03948689e+00 -2.72988498e-01 5.04025102e-01
1.31012261e-01 -9.16582406e-01 -2.11688146e-01 -4.19783413e-01
2.58964002e-01 1.32170808e+00 1.09462351e-01 8.23658183e-02
1.54702282e+00 4.57052797e-01 2.33472332e-01 4.75428887e-02
-6.58430994e-01 -2.27879748e-01 -2.33986154e-01 8.87562454e-01
3.51751566e-01 1.15244649e-01 3.53918761e-01 4.17405009e-01
-9.18296799e-02 1.38587356e-01 7.53919840e-01 7.70405710e-01
-1.06133413e+00 -8.95679772e-01 -4.56854641e-01 8.59411001e-01
-4.61968213e-01 -5.71322581e-03 7.93844983e-02 2.79826283e-01
4.86312211e-02 1.14407587e+00 -8.94388705e-02 -3.26947272e-01
-2.46469397e-02 -3.73952240e-01 4.43676561e-01 -3.75469625e-01
3.73846322e-01 7.35168234e-02 2.75027633e-01 -1.25003085e-01
1.80647567e-01 -6.32806242e-01 -1.27116024e+00 -1.19701937e-01
-8.07419717e-02 6.80478990e-01 2.32354924e-01 1.41143453e+00
6.19623959e-01 3.86077374e-01 1.27594447e+00 -7.70364285e-01
-4.63055223e-01 -6.77602410e-01 -1.16810155e+00 -7.15911031e-01
5.35318196e-01 -2.68657386e-01 -2.64178514e-01 -2.14725822e-01] | [6.114151954650879, 1.5997627973556519] |
ccb7835b-589b-47c1-8dbc-0b06e168de46 | lmscnet-lightweight-multiscale-3d-semantic | 2008.10559 | null | https://arxiv.org/abs/2008.10559v2 | https://arxiv.org/pdf/2008.10559v2.pdf | LMSCNet: Lightweight Multiscale 3D Semantic Completion | We introduce a new approach for multiscale 3Dsemantic scene completion from voxelized sparse 3D LiDAR scans. As opposed to the literature, we use a 2D UNet backbone with comprehensive multiscale skip connections to enhance feature flow, along with 3D segmentation heads. On the SemanticKITTI benchmark, our method performs on par on semantic completion and better on occupancy completion than all other published methods -- while being significantly lighter and faster. As such it provides a great performance/speed trade-off for mobile-robotics applications. The ablation studies demonstrate our method is robust to lower density inputs, and that it enables very high speed semantic completion at the coarsest level. Our code is available at https://github.com/cv-rits/LMSCNet. | ['Anne Verroust-Blondet', 'Luis Roldão', 'Raoul de Charette'] | 2020-08-24 | null | null | null | null | ['3d-semantic-scene-completion', '3d-semantic-scene-completion-from-a-single'] | ['computer-vision', 'computer-vision'] | [ 1.47500157e-01 1.50257617e-01 -2.42289871e-01 -3.38729799e-01
-8.48023117e-01 -4.45968390e-01 4.45689023e-01 -7.89295062e-02
-5.14419615e-01 4.13501441e-01 2.03787386e-01 -3.82441968e-01
6.22842135e-03 -7.66964674e-01 -8.16572249e-01 -6.26411522e-03
-3.06560714e-02 8.48103464e-01 5.46870768e-01 -8.62633958e-02
1.87748358e-01 6.71576619e-01 -1.66530454e+00 -1.28787935e-01
8.16627741e-01 8.78049433e-01 3.65688920e-01 5.65615892e-01
-2.79512465e-01 3.31103712e-01 1.80271804e-01 -1.27346307e-01
5.72609961e-01 2.28316486e-01 -8.87923121e-01 1.09972898e-03
1.00109398e+00 -3.82279545e-01 -2.98878342e-01 7.94451177e-01
3.70440215e-01 2.72286624e-01 5.66700459e-01 -1.07670820e+00
-3.27189595e-01 4.11043555e-01 -9.29088831e-01 7.71128833e-02
3.57811302e-01 2.87564874e-01 1.09415925e+00 -1.57694745e+00
7.68934071e-01 1.46561515e+00 9.97505128e-01 3.10117006e-01
-1.49827695e+00 -7.23905385e-01 2.24849433e-01 -5.13138808e-02
-1.53727424e+00 -5.15049279e-01 4.84519422e-01 -1.56900674e-01
1.21267319e+00 1.80547331e-02 8.89993906e-01 7.98027217e-01
4.19912860e-02 8.96002233e-01 1.04533684e+00 4.48664688e-02
3.51599544e-01 -3.81820083e-01 2.32754543e-01 1.02854395e+00
2.31452122e-01 8.99738073e-02 -6.19698405e-01 -9.37756523e-03
9.93091524e-01 -3.52594443e-02 1.15933523e-01 -8.45244706e-01
-1.22789121e+00 8.98078501e-01 8.47640038e-01 -9.10971165e-02
-2.00590119e-01 6.91962779e-01 2.35576794e-01 -8.31713751e-02
6.13874733e-01 9.36230496e-02 -3.25127810e-01 -7.42087960e-02
-1.21967685e+00 4.02405530e-01 6.87853038e-01 1.26080298e+00
1.06964040e+00 1.61957368e-01 1.54697135e-01 6.96664512e-01
1.34568423e-01 8.19446862e-01 -1.53422564e-01 -1.40734017e+00
5.31230927e-01 3.16282362e-01 -4.05240618e-02 -6.11708045e-01
-6.77059412e-01 -4.28883404e-01 -6.09250546e-01 5.13435900e-01
2.40434557e-01 1.99937806e-01 -1.45897806e+00 1.46832693e+00
6.10837460e-01 4.54094172e-01 -2.13509038e-01 9.98205423e-01
8.73492360e-01 3.27221394e-01 1.45598471e-01 3.46991092e-01
1.14235985e+00 -1.10174000e+00 -3.00836951e-01 -5.02886176e-01
5.18904686e-01 -7.24058449e-01 1.37373435e+00 3.62504482e-01
-1.22676003e+00 -4.99956518e-01 -9.69088554e-01 -6.76558018e-01
-2.20959067e-01 -2.79509157e-01 1.11146665e+00 4.89855587e-01
-1.33745944e+00 6.22142136e-01 -1.17849123e+00 -5.56568861e-01
8.78785789e-01 3.60002726e-01 -4.26556766e-01 -5.16119421e-01
-5.68229496e-01 8.43828499e-01 2.02257216e-01 -1.68827325e-01
-1.03052187e+00 -1.00829232e+00 -1.18053925e+00 -2.71199316e-01
3.60076457e-01 -1.12128985e+00 1.55497313e+00 -7.65978992e-02
-1.18427920e+00 8.59343886e-01 -3.55042160e-01 -6.63492858e-01
6.69895947e-01 -4.70132351e-01 4.55808729e-01 9.67110842e-02
4.98698771e-01 1.50031996e+00 5.59233248e-01 -1.17958343e+00
-6.69263422e-01 -5.00636637e-01 9.32747778e-03 6.07705474e-01
9.89128724e-02 -5.67526579e-01 -7.25904286e-01 -3.63495469e-01
5.48574448e-01 -9.32614386e-01 -6.89903080e-01 3.99646401e-01
-3.02014679e-01 -1.13503680e-01 8.59147727e-01 -3.42336506e-01
4.73227829e-01 -2.02897787e+00 8.30835551e-02 1.67270467e-01
3.30542892e-01 -1.04232065e-01 -2.42422253e-01 2.32286319e-01
2.63334125e-01 1.00187659e-01 -8.39482903e-01 -9.17342186e-01
1.21359527e-02 2.56583482e-01 -2.12686598e-01 5.51123857e-01
1.93506017e-01 1.06251323e+00 -8.86719108e-01 -5.50666928e-01
7.07793355e-01 6.55520678e-01 -8.97394538e-01 -3.43964189e-01
-2.58008838e-01 4.58257049e-01 -5.17464578e-01 9.70489621e-01
9.24735963e-01 -1.96887195e-01 -2.61246681e-01 -1.22432254e-01
-8.98151398e-02 2.56699860e-01 -1.26317203e+00 2.68305683e+00
-6.96777463e-01 3.29218984e-01 4.05953079e-01 -6.51471794e-01
5.97977936e-01 -2.78641701e-01 4.88224685e-01 -6.28774464e-01
-7.20706210e-02 3.09477955e-01 -4.42075044e-01 7.63989836e-02
7.98880279e-01 -1.29065618e-01 -1.49141371e-01 2.42104366e-01
1.09991245e-01 -1.05613303e+00 7.22738157e-04 5.71981728e-01
1.25218797e+00 5.19098639e-01 1.34466350e-01 -4.30596799e-01
-6.11281320e-02 3.93167645e-01 4.43317622e-01 7.24039912e-01
-2.59286957e-03 8.31247985e-01 -2.62069646e-02 -1.82213649e-01
-1.07511735e+00 -1.58901703e+00 -3.53055596e-01 7.37803042e-01
4.93034542e-01 -4.59661901e-01 -4.99815166e-01 -3.83884668e-01
4.49196011e-01 6.75430894e-01 -2.90814042e-01 3.23986083e-01
-4.33048755e-01 -2.95776516e-01 5.83775997e-01 7.36400187e-01
5.90599477e-01 -8.05882573e-01 -6.21461034e-01 2.12633029e-01
1.17680272e-02 -1.46420062e+00 -4.02835190e-01 4.01636124e-01
-1.35343373e+00 -9.04782653e-01 -4.47908401e-01 -7.80824363e-01
5.07763088e-01 6.75261021e-01 1.22463453e+00 8.50175787e-03
-6.58452332e-01 4.37314123e-01 -1.46025822e-01 -2.70052582e-01
1.94957942e-01 4.59918320e-01 9.57279354e-02 -7.20774472e-01
5.63191958e-02 -9.89635170e-01 -7.09758162e-01 3.04585951e-03
-8.66698086e-01 3.26358438e-01 5.12563765e-01 3.64315540e-01
1.00293934e+00 -2.99045533e-01 2.32306480e-01 -9.39751208e-01
3.02432597e-01 -3.74044001e-01 -5.60894012e-01 -5.33989549e-01
-5.36551297e-01 6.28502518e-02 3.41762632e-01 1.13142192e-01
-8.09525788e-01 4.93222982e-01 -4.78683382e-01 -6.94571912e-01
-3.44195962e-01 2.35035062e-01 7.94885978e-02 -1.37658969e-01
7.43648350e-01 -7.44188279e-02 -6.46465982e-04 -7.05709040e-01
9.31498766e-01 1.78857967e-01 6.20027065e-01 -7.25621879e-01
1.05238390e+00 1.02749968e+00 1.08491912e-01 -8.75853062e-01
-6.81482196e-01 -5.75971484e-01 -8.01880658e-01 7.39872381e-02
7.10938454e-01 -1.28567696e+00 -2.55639553e-01 2.59549469e-01
-9.97236907e-01 -7.60797739e-01 -6.63405001e-01 4.14913744e-01
-7.73708582e-01 3.49055916e-01 -6.99354708e-01 -6.29335523e-01
-4.24982935e-01 -1.11175907e+00 1.35834718e+00 4.56724018e-02
-2.09777161e-01 -7.55239785e-01 -5.95639497e-02 4.15824354e-01
2.76611686e-01 2.09324345e-01 5.62179327e-01 -5.22370674e-02
-9.03715789e-01 1.06593184e-01 -4.44249123e-01 1.05920732e-01
-1.65857643e-01 -2.00716540e-01 -1.04617488e+00 -2.59847105e-01
-3.30668390e-01 -3.42693895e-01 1.30485463e+00 5.88424087e-01
1.12865961e+00 1.30587399e-01 -4.62306529e-01 1.01305759e+00
1.47408938e+00 -5.96844256e-01 4.50158566e-01 4.27659638e-02
8.46543312e-01 4.05688673e-01 7.75288045e-01 3.19328398e-01
7.38610983e-01 4.55406040e-01 7.49001920e-01 -3.25348377e-01
-6.11485302e-01 -4.13159817e-01 2.32312709e-01 7.29681730e-01
1.37920722e-01 1.05012327e-01 -1.10580516e+00 6.88447773e-01
-1.91823900e+00 -5.66831589e-01 -3.08771044e-01 1.98291361e+00
3.65922749e-01 2.86661744e-01 8.84535834e-02 -1.15630671e-01
3.87424976e-01 3.06635171e-01 -8.39221776e-01 -1.28893942e-01
-3.63562480e-02 8.09506774e-01 8.09646726e-01 9.80022311e-01
-1.08392167e+00 1.59895527e+00 6.40027618e+00 1.01055765e+00
-5.73463440e-01 4.40605104e-01 3.44512880e-01 -5.16330481e-01
-6.01369858e-01 1.95029035e-01 -8.60509932e-01 -6.05727434e-02
5.13962805e-01 2.08849102e-01 4.26899761e-01 9.51635659e-01
3.07763427e-01 -5.22897124e-01 -9.69916105e-01 9.04850483e-01
-1.39003724e-01 -1.50617278e+00 -6.40650373e-03 1.89937681e-01
6.26452923e-01 8.37448955e-01 -3.66274291e-03 2.24381790e-01
6.96396053e-01 -1.07482624e+00 9.50610638e-01 1.03873715e-01
1.06315994e+00 -8.29982162e-01 2.95914441e-01 3.68331373e-01
-1.72681856e+00 1.43310189e-01 -4.12724614e-01 -2.28133962e-01
3.41003090e-01 8.76427710e-01 -7.73425519e-01 4.78833914e-01
8.91527593e-01 8.25591087e-01 -6.17177308e-01 1.07331169e+00
-2.66347349e-01 2.95392126e-01 -8.64603102e-01 2.49681875e-01
3.06909472e-01 -1.31126404e-01 5.53322434e-01 1.31201839e+00
2.70580590e-01 -3.22876386e-02 4.69737113e-01 9.26543236e-01
-1.09112889e-01 -2.07016487e-02 -1.02033532e+00 4.47405249e-01
6.65374696e-01 1.42920709e+00 -1.20273590e+00 -2.36790881e-01
-2.92599708e-01 9.45479512e-01 5.67576587e-01 2.28404954e-01
-7.97914684e-01 -1.60541922e-01 7.50409067e-01 2.26245761e-01
3.90986383e-01 -8.48439753e-01 -9.04168069e-01 -9.97318685e-01
-3.79084819e-03 -2.57169336e-01 2.76426405e-01 -8.23599160e-01
-1.14455450e+00 2.11965784e-01 1.09254323e-01 -8.26236844e-01
1.69002473e-01 -4.72246230e-01 -2.81012565e-01 6.35149956e-01
-1.64066184e+00 -1.20811915e+00 -4.48403329e-01 6.43321753e-01
7.93546140e-01 5.59659541e-01 6.12209439e-01 2.19841823e-01
-1.84281692e-01 1.81407779e-01 -3.25988799e-01 -2.34736651e-01
2.33939722e-01 -1.27339399e+00 1.07906985e+00 9.93793011e-01
2.96481133e-01 3.61953229e-01 3.89003545e-01 -8.69058847e-01
-1.60966921e+00 -1.07411849e+00 3.40603232e-01 -6.73480034e-01
5.36977589e-01 -6.16418481e-01 -5.81513941e-01 6.80563092e-01
-4.20657694e-02 6.09718375e-02 2.75399119e-01 2.31249973e-01
-3.91299039e-01 1.16583265e-01 -1.21452415e+00 6.49705529e-01
1.85115349e+00 -4.59545016e-01 -5.07879853e-01 3.45025837e-01
1.05562508e+00 -7.11747348e-01 -8.67447019e-01 7.27443695e-01
4.18797702e-01 -9.23017502e-01 1.34195995e+00 2.69362610e-02
3.13398570e-01 -4.71497357e-01 -4.89577860e-01 -9.91663873e-01
-3.66486400e-01 -3.05394292e-01 -1.92793369e-01 6.48089111e-01
3.88025641e-01 -4.06349123e-01 1.12589312e+00 3.53403062e-01
-5.60186923e-01 -7.36746550e-01 -1.24131978e+00 -8.15707147e-01
1.17042996e-01 -9.68402147e-01 4.11955446e-01 5.92072666e-01
-4.54576075e-01 5.18201888e-01 3.34857143e-02 2.28633970e-01
9.71833348e-01 1.95679232e-01 9.70397770e-01 -1.09945822e+00
1.53340414e-01 -5.65529227e-01 -2.19017461e-01 -1.28688323e+00
7.04875290e-02 -1.45387101e+00 1.19756244e-01 -2.04270267e+00
-1.09495400e-02 -6.68266833e-01 7.07743987e-02 6.83011770e-01
1.39006510e-01 7.65080512e-01 2.25146532e-01 1.39267698e-01
-5.54023921e-01 8.22575867e-01 1.33747065e+00 -8.19657519e-02
-2.46544793e-01 -2.42367506e-01 -5.83444417e-01 7.86659122e-01
8.63487780e-01 -5.08806825e-01 -4.98683900e-01 -6.35398030e-01
-4.28524464e-02 -2.08212048e-01 6.43334329e-01 -1.32788706e+00
1.63142025e-01 -1.20471800e-02 3.23476732e-01 -1.06150961e+00
7.72080600e-01 -6.16344213e-01 5.20899370e-02 4.40625906e-01
8.68726075e-02 9.61209610e-02 3.61364394e-01 6.95916176e-01
2.32552245e-01 1.28488287e-01 7.25927472e-01 -4.36748207e-01
-1.03551877e+00 8.06707919e-01 -2.69100636e-01 2.18653128e-01
7.24059463e-01 -4.75925714e-01 -8.95298049e-02 -1.85680419e-01
-6.90503895e-01 4.44750339e-01 8.63096893e-01 4.81589645e-01
8.89893949e-01 -1.18930030e+00 -6.55148566e-01 9.81627032e-02
3.04665305e-02 7.11592197e-01 2.49012545e-01 8.04015636e-01
-8.14567208e-01 2.77709156e-01 2.27149837e-02 -1.05236220e+00
-8.45767498e-01 7.43481591e-02 1.28782749e-01 -1.33925269e-03
-1.20279336e+00 1.03417027e+00 2.40501016e-01 -9.60693240e-01
3.14058483e-01 -6.03705287e-01 4.37649369e-01 -1.86687499e-01
-4.57436172e-03 3.18039984e-01 1.44539461e-01 -4.64496553e-01
-5.69282949e-01 8.35044265e-01 1.22750193e-01 -4.01907712e-01
1.38894224e+00 -6.77842125e-02 1.26944050e-01 3.56940746e-01
1.00085330e+00 -1.95008725e-01 -1.45464027e+00 -3.04733753e-01
-5.53430803e-02 -5.88471234e-01 2.74944156e-01 -4.57361758e-01
-1.04006732e+00 9.49248672e-01 5.57234526e-01 -2.34373942e-01
6.44865096e-01 2.04147696e-01 9.04002428e-01 4.55847055e-01
8.49288821e-01 -1.03302383e+00 3.88832279e-02 7.77262568e-01
7.82105148e-01 -1.15497839e+00 3.04102927e-01 -6.55139565e-01
-4.54327762e-01 6.53062403e-01 6.59653306e-01 -5.40474057e-01
7.58230209e-01 6.06239438e-01 -1.72720537e-01 -4.26620185e-01
-3.86385888e-01 -6.44124627e-01 1.82063924e-03 7.97406614e-01
1.09891109e-01 1.37071893e-01 1.56497862e-02 4.72754985e-02
-6.18558943e-01 -3.44864391e-02 3.38875353e-01 9.23361540e-01
-5.56861162e-01 -8.56652975e-01 -2.40692377e-01 6.07228041e-01
1.19454376e-01 -3.64139736e-01 -1.95417687e-01 9.22660410e-01
-9.70954299e-02 7.87327170e-01 1.70477346e-01 -4.90924567e-01
4.80170399e-01 -2.18821950e-02 5.97854435e-01 -9.85960960e-01
-3.13073397e-01 7.46843666e-02 1.32888392e-01 -1.18782461e+00
-3.22467476e-01 -7.40783811e-01 -1.66510081e+00 -5.23802936e-01
-1.65731296e-01 -2.36953259e-01 7.76372731e-01 6.72681570e-01
5.77347040e-01 3.22455198e-01 2.44197637e-01 -1.48256290e+00
-5.36140092e-02 -6.87998712e-01 -4.55299944e-01 2.62699462e-02
9.72257406e-02 -8.23991716e-01 -2.06111565e-01 -4.00608897e-01] | [8.276215553283691, -2.83524489402771] |
3019bfc4-4188-4faa-ac54-b7e25ca652d1 | detection-of-abnormal-behavior-with-self | 2107.06530 | null | https://arxiv.org/abs/2107.06530v1 | https://arxiv.org/pdf/2107.06530v1.pdf | Detection of Abnormal Behavior with Self-Supervised Gaze Estimation | Due to the recent outbreak of COVID-19, many classes, exams, and meetings have been conducted non-face-to-face. However, the foundation for video conferencing solutions is still insufficient. So this technology has become an important issue. In particular, these technologies are essential for non-face-to-face testing, and technology dissemination is urgent. In this paper, we present a single video conferencing solution using gaze estimation in preparation for these problems. Gaze is an important cue for the tasks such as analysis of human behavior. Hence, numerous studies have been proposed to solve gaze estimation using deep learning, which is one of the most prominent methods up to date. We use these gaze estimation methods to detect abnormal behavior of video conferencing participants. Our contribution is as follows. i) We find and apply the optimal network for the gaze estimation method and apply a self-supervised method to improve accuracy. ii) For anomaly detection, we present a new dataset that aggregates the values of a new gaze, head pose, etc. iii) We train newly created data on Multi Layer Perceptron (MLP) models to detect anomaly behavior based on deep learning. We demonstrate the robustness of our method through experiments. | ['Seong-Whan Lee', 'Suneung-Kim'] | 2021-07-14 | null | null | null | null | ['gaze-estimation'] | ['computer-vision'] | [ 7.69542605e-02 4.33923490e-03 2.32778579e-01 -6.05147243e-01
-2.15439647e-01 -3.43022309e-02 6.03454262e-02 -1.41885668e-01
-3.05656910e-01 7.08095312e-01 -2.37687930e-01 -5.12840748e-02
-3.75641882e-01 -3.02625716e-01 -5.79174936e-01 -7.94411719e-01
-4.36055548e-02 -3.22409086e-02 2.02611789e-01 -1.52347624e-01
6.40429556e-01 4.45863664e-01 -2.30960488e+00 1.51902333e-01
9.68490303e-01 1.05273128e+00 -1.02800258e-01 6.50930822e-01
2.68599279e-02 5.24807930e-01 -7.51132727e-01 -3.10930640e-01
-2.10215852e-01 -3.34669501e-01 -5.46425045e-01 -1.02185734e-01
5.70975959e-01 -5.94551206e-01 -4.59159631e-03 1.05528128e+00
6.14911973e-01 1.13820747e-01 4.73322779e-01 -1.96876776e+00
-1.49534941e-01 1.67175144e-01 -1.00995088e+00 4.11878973e-01
7.11966455e-01 -8.82777646e-02 4.07032281e-01 -5.37432909e-01
-1.80462804e-02 1.10458744e+00 5.17519534e-01 8.38198245e-01
-5.27708948e-01 -9.97997463e-01 1.55011103e-01 6.84818625e-01
-1.33441842e+00 -7.48727143e-01 8.85158420e-01 -3.48203391e-01
5.79018772e-01 2.50269562e-01 5.21212578e-01 1.22285092e+00
1.00238241e-01 6.49681807e-01 9.73691523e-01 -6.71107590e-01
-1.11131690e-01 2.08793059e-01 2.69899547e-01 8.02231133e-01
2.55504191e-01 -2.72020161e-01 -7.96931446e-01 -3.06936149e-02
3.79693598e-01 1.14292718e-01 -3.43610436e-01 4.89408039e-02
-6.18750393e-01 4.64673340e-01 1.14196144e-01 3.18955779e-01
-2.93024480e-01 -5.11485226e-02 1.85837016e-01 3.27677995e-01
5.94727993e-01 1.53997779e-01 -1.45829067e-01 -4.98233199e-01
-7.56846845e-01 1.63937956e-01 6.03795111e-01 3.83623302e-01
4.87018734e-01 -2.48134106e-01 2.88748648e-02 7.63255477e-01
7.52460063e-01 4.42088068e-01 5.07224441e-01 -8.35160494e-01
3.38353217e-01 6.62271500e-01 -1.13805257e-01 -1.21982300e+00
-5.61404824e-01 2.17457443e-01 -5.62142730e-01 3.33081096e-01
4.38461900e-01 -4.53663379e-01 -5.27616322e-01 1.69815493e+00
4.76218969e-01 6.37985647e-01 -4.23495442e-01 7.55595446e-01
8.62946212e-01 4.01789039e-01 -1.37395874e-01 -5.82701981e-01
1.16759932e+00 -5.52896500e-01 -1.10451543e+00 6.77552447e-02
5.70044518e-01 -6.05803251e-01 8.80750418e-01 6.40556276e-01
-9.86059546e-01 -2.39069328e-01 -8.68043005e-01 3.17082465e-01
-3.06264043e-01 -5.17430566e-02 2.35402614e-01 1.19120061e+00
-1.25003159e+00 3.44802886e-01 -7.16747344e-01 -6.59649134e-01
3.95698160e-01 7.37065017e-01 -3.36976320e-01 2.25940511e-01
-8.52965057e-01 8.16021204e-01 -4.07886580e-02 4.32782739e-01
-3.25313628e-01 -2.93861121e-01 -7.80870676e-01 5.65232895e-02
3.97284567e-01 -1.11030057e-01 1.15736985e+00 -1.02208495e+00
-1.59485412e+00 8.83885562e-01 -6.60383821e-01 1.58674363e-02
2.44155616e-01 -3.35921794e-01 -7.34286606e-01 1.32318407e-01
-4.36509103e-02 4.59724754e-01 9.71937895e-01 -1.08498108e+00
-7.71758318e-01 -6.91591680e-01 -1.08013578e-01 1.42505085e-02
-8.06338668e-01 6.14681184e-01 -1.86177805e-01 -1.18269511e-02
6.12700032e-03 -7.91756392e-01 6.45984650e-01 -3.63057815e-02
-4.05044198e-01 -7.05848038e-01 1.28488350e+00 -6.93392873e-01
1.53405452e+00 -2.27162433e+00 -2.22772673e-01 4.30398136e-01
6.31300688e-01 3.69462132e-01 2.33883798e-01 7.93545414e-03
-2.87168890e-01 -1.41426073e-02 -3.67326103e-02 -4.23632383e-01
-2.47149974e-01 -2.28637666e-01 1.26891032e-01 4.22913522e-01
9.53429863e-02 4.53616679e-01 -5.23405135e-01 -7.43854702e-01
8.68357792e-02 2.79017955e-01 -5.73696733e-01 4.44775164e-01
2.54066974e-01 5.91673672e-01 -3.94336462e-01 8.12339604e-01
7.81491995e-01 -1.12860970e-01 -1.71802074e-01 1.54446557e-01
-2.43930012e-01 -2.43685976e-01 -9.65848386e-01 1.14537513e+00
2.43523046e-02 9.23344433e-01 1.06796898e-01 -9.12958264e-01
9.02635753e-01 4.28449064e-01 5.49704075e-01 -4.65753227e-01
5.29094100e-01 1.15654230e-01 2.41605788e-01 -1.38245654e+00
2.90855616e-01 3.89566034e-01 3.74200314e-01 6.24457479e-01
-2.34514132e-01 5.12425125e-01 1.86610624e-01 -5.48877157e-02
9.92792487e-01 6.85769767e-02 4.16850746e-02 1.33906543e-01
8.49547744e-01 -6.32877886e-01 4.62832749e-01 3.36434871e-01
-7.25540280e-01 6.16624296e-01 8.30703974e-01 -3.32056314e-01
-4.92962748e-01 -5.82026958e-01 -2.15706676e-02 1.30318558e+00
8.59864876e-02 4.11425643e-02 -1.12637436e+00 -6.17815375e-01
-1.52800471e-01 5.50039232e-01 -5.25659025e-01 -2.70607799e-01
-5.73140144e-01 -6.88909590e-01 5.62520385e-01 2.47464433e-01
5.07242799e-01 -1.44378138e+00 -4.23543662e-01 -3.13346893e-01
-1.97976261e-01 -9.06924367e-01 -2.34339669e-01 -4.25937355e-01
-5.36173463e-01 -1.34432709e+00 -7.38184273e-01 -7.92416692e-01
7.91394770e-01 2.37662867e-01 4.76751536e-01 4.81434762e-01
-2.26109847e-02 4.37743276e-01 -2.86321193e-01 -7.42573798e-01
-1.04733437e-01 5.72664104e-03 6.29077852e-01 4.76628035e-01
8.91996324e-01 -5.97764969e-01 -4.31474864e-01 3.68745744e-01
-4.81960595e-01 -4.04504120e-01 4.33412820e-01 2.54530996e-01
-1.84968635e-01 -2.97901899e-01 6.96213543e-01 -6.26262963e-01
9.97014463e-01 -5.61723113e-01 -6.13119662e-01 3.71585757e-01
-4.33344007e-01 -3.80099356e-01 5.30344732e-02 -2.33341306e-01
-1.18654430e+00 -3.26891720e-01 -1.07966743e-01 -5.88441133e-01
-7.16495275e-01 2.11304665e-01 -1.50210544e-01 -3.06699067e-01
5.39218843e-01 -2.48962879e-01 4.57786292e-01 -3.20516467e-01
-4.62777704e-01 1.29164112e+00 2.68306106e-01 -2.51199931e-01
5.37135065e-01 2.11006001e-01 -1.42190503e-02 -1.14976645e+00
-6.85295939e-01 -4.01035160e-01 -6.51208222e-01 -7.50202060e-01
1.00970733e+00 -4.13522810e-01 -1.63582575e+00 1.07977688e+00
-1.25859725e+00 1.37046248e-01 5.99279106e-01 6.59323573e-01
-1.45428702e-01 3.51049691e-01 -2.84449309e-01 -1.29337406e+00
-2.46700093e-01 -1.03739238e+00 8.36877406e-01 7.37444460e-01
-2.19339326e-01 -7.39035666e-01 2.10105002e-01 5.32111526e-01
3.37721944e-01 9.95776206e-02 4.10191268e-01 -6.31851256e-01
-3.69938433e-01 -1.53940573e-01 -4.28112507e-01 1.98362663e-01
1.76527604e-01 4.86015171e-01 -1.42560339e+00 -1.87181830e-01
-7.79304504e-02 -2.02577963e-01 4.43357229e-01 5.97670615e-01
1.61831117e+00 1.07773855e-01 -5.50002754e-01 5.80973744e-01
6.83667362e-01 4.25408185e-01 7.08791316e-01 4.23210114e-01
8.34161937e-01 9.06287074e-01 4.70604479e-01 3.92261982e-01
4.60315406e-01 4.30161446e-01 6.67608917e-01 2.69162148e-01
4.49527204e-01 2.48248056e-01 3.80602807e-01 6.54347181e-01
-4.83066589e-01 -4.28019971e-01 -9.99180973e-01 2.95981288e-01
-1.67634869e+00 -1.07057464e+00 -3.39747161e-01 2.09017968e+00
3.21978331e-01 8.25849995e-02 4.08615410e-01 3.93487215e-01
1.10984075e+00 -2.16292560e-01 -5.21482408e-01 -4.65979934e-01
4.77579266e-01 4.11404297e-02 -1.87915429e-01 1.04699312e-02
-1.02079070e+00 4.34457392e-01 5.96233463e+00 2.10341051e-01
-1.36771905e+00 -1.31503090e-01 4.23312604e-01 -2.37324283e-01
2.35858142e-01 -6.54870152e-01 -8.82813692e-01 8.27783406e-01
9.66914475e-01 6.38186187e-02 3.34435642e-01 6.10705316e-01
4.76669818e-01 -3.04568827e-01 -1.05387139e+00 1.37790656e+00
5.33327281e-01 -7.16569185e-01 -5.04128516e-01 7.68428296e-02
2.28253081e-01 -1.89426079e-01 1.47187084e-01 1.63626447e-01
-4.83881533e-01 -8.97866845e-01 7.04677179e-02 6.82242095e-01
6.48663223e-01 -6.97755754e-01 9.73554492e-01 3.05179298e-01
-7.24293709e-01 -3.47771823e-01 8.01024139e-02 -1.58763319e-01
1.74195878e-02 3.29860538e-01 -7.10126340e-01 3.01996708e-01
1.27448952e+00 5.56129336e-01 -5.70006192e-01 1.42070818e+00
-6.31881654e-02 4.95353073e-01 -2.69333392e-01 -3.30453962e-01
-2.27796406e-01 -8.26826468e-02 5.14137983e-01 4.84864920e-01
5.44367492e-01 -1.40202614e-02 -3.51915747e-01 5.28638124e-01
-5.87158352e-02 -6.00207001e-02 -7.19996750e-01 2.24605739e-01
4.36891139e-01 1.18313050e+00 -4.82795000e-01 2.30296209e-01
-5.61589777e-01 6.11887455e-01 1.81175187e-01 2.71575123e-01
-8.65831494e-01 -6.69986069e-01 7.46701658e-01 2.75700241e-01
-1.58165365e-01 2.38229364e-01 -6.98901340e-02 -8.86541307e-01
2.78487325e-01 -9.12458241e-01 2.88167119e-01 -9.34247851e-01
-1.12162280e+00 3.98266315e-01 1.06427930e-01 -1.18944681e+00
-5.32805502e-01 -6.62163854e-01 -1.01045763e+00 7.38017440e-01
-1.50468051e+00 -6.71187341e-01 -7.86447525e-01 6.74735904e-01
2.61764318e-01 -4.60537612e-01 6.35800779e-01 5.47450185e-01
-1.14796531e+00 9.15572524e-01 -2.20395729e-01 1.88573062e-01
1.03166819e+00 -7.88618445e-01 -1.09099723e-01 7.82606006e-01
-3.39114428e-01 5.60014009e-01 5.13528705e-01 -3.92004669e-01
-9.39716995e-01 -5.61622798e-01 8.29296529e-01 -5.93465030e-01
3.83187801e-01 -1.88619822e-01 -1.15597737e+00 7.55119145e-01
3.28705996e-01 -2.06643194e-01 8.17626357e-01 3.47524792e-01
3.97243887e-01 -3.65175515e-01 -1.25039196e+00 4.03172165e-01
8.33968461e-01 -3.55115831e-01 -6.37200236e-01 3.49298939e-02
3.93590778e-01 -5.21660209e-01 -3.05266649e-01 2.89190441e-01
7.02197731e-01 -1.30868804e+00 3.99299711e-01 -4.68537927e-01
3.59278530e-01 -1.76762342e-01 3.79801899e-01 -1.05397320e+00
9.01142880e-02 -7.63377905e-01 -2.39236191e-01 1.36429787e+00
2.72901416e-01 -9.11529899e-01 9.71015692e-01 7.64508069e-01
-1.20627075e-01 -6.41097605e-01 -8.61494422e-01 -2.17518643e-01
-6.57070637e-01 -2.69787788e-01 6.41021252e-01 9.96654570e-01
1.43316448e-01 7.69713148e-02 -3.89609635e-01 3.16091239e-01
5.41018546e-01 -4.96521145e-01 8.92572343e-01 -1.65483689e+00
2.45037243e-01 -4.31012928e-01 -6.10922992e-01 -6.11114383e-01
2.62947410e-01 -7.13625178e-02 7.48735480e-03 -1.01731813e+00
9.16844159e-02 -1.51307091e-01 -3.78474176e-01 4.09422398e-01
-1.95888028e-01 1.55060187e-01 -1.60363942e-01 -3.24006192e-02
-5.37759602e-01 3.51265430e-01 1.02392280e+00 2.69224048e-01
-3.35743278e-01 4.58905667e-01 -6.75993860e-01 9.69715118e-01
8.87636960e-01 -4.26691473e-01 -3.02134693e-01 -3.34095031e-01
2.69042879e-01 -1.03372410e-01 2.64286399e-01 -1.11969614e+00
5.35729229e-01 -3.21419910e-02 3.86626631e-01 -7.09544420e-01
4.61365104e-01 -7.32872367e-01 -6.48594141e-01 2.41485555e-02
6.80211559e-02 2.27531180e-01 2.44109794e-01 3.52467716e-01
-1.57652885e-01 -3.30212504e-01 4.00740802e-01 3.09957832e-01
-6.03149891e-01 2.28107095e-01 -2.97083825e-01 -1.21596903e-01
1.40813220e+00 -4.64486957e-01 -4.46763009e-01 -5.47426760e-01
-5.67503095e-01 5.29324889e-01 5.63172139e-02 5.50733328e-01
6.91915393e-01 -8.76746416e-01 -4.68706936e-01 7.04725146e-01
5.06911650e-02 -8.90272185e-02 3.11900079e-01 1.16358232e+00
-4.77253824e-01 1.37505785e-01 -6.03639722e-01 -9.06340957e-01
-1.77641952e+00 1.05378218e-01 3.07202667e-01 4.52324390e-01
-3.14599387e-02 9.77709591e-01 1.24821169e-02 -2.70300806e-01
4.58841413e-01 -3.18006836e-02 -9.77992117e-01 1.83254436e-01
8.18811834e-01 6.31604612e-01 1.26228765e-01 -6.62642479e-01
-5.74394166e-01 6.46188200e-01 -2.76973635e-01 1.73271567e-01
1.27152789e+00 -2.70667702e-01 -3.46874684e-01 4.29392725e-01
1.01653302e+00 -6.92076832e-02 -8.67903590e-01 2.46936336e-01
-1.19552843e-01 -6.25425637e-01 6.75192103e-02 -3.57596278e-01
-1.07497287e+00 1.13596582e+00 8.90337229e-01 5.82573533e-01
1.28401065e+00 -1.84991643e-01 5.07306635e-01 4.84033197e-01
4.40477803e-02 -1.10963225e+00 1.20419674e-01 3.27083021e-01
5.31483591e-01 -1.60408485e+00 -1.70447648e-01 -3.05321306e-01
-3.86097759e-01 1.11900353e+00 1.10720789e+00 3.29025209e-01
8.41194034e-01 7.78384507e-02 2.99355745e-01 -3.57476741e-01
-5.10678113e-01 1.73014514e-02 2.55655587e-01 5.77018321e-01
5.24783611e-01 -3.51362765e-01 4.55800258e-02 3.21751207e-01
-7.57642090e-02 2.57114291e-01 4.88335043e-01 9.38411474e-01
-5.52516401e-01 -7.85909474e-01 -8.47401857e-01 8.12244594e-01
-6.25688016e-01 3.94558817e-01 -2.88897187e-01 6.75839663e-01
-2.69734897e-02 1.20179033e+00 3.57403010e-01 -4.99345601e-01
2.98878163e-01 3.09263140e-01 3.62316310e-01 -3.98715585e-01
-2.43815899e-01 -3.08421016e-01 -1.46650001e-01 -5.87813258e-01
-5.97978175e-01 -7.84442246e-01 -9.24080431e-01 -7.54896283e-01
-5.40763378e-01 -6.69571087e-02 7.23465204e-01 1.23764765e+00
3.55800331e-01 5.75043619e-01 7.47905135e-01 -8.94775748e-01
-1.16830572e-01 -1.15257347e+00 -5.84124625e-01 2.45193705e-01
5.50384641e-01 -1.04140484e+00 -6.04138970e-01 -6.75156042e-02] | [14.024275779724121, 0.2446908950805664] |
ac190f95-90f6-4865-a7f3-e7a620c7997e | robust-environment-perception-for-automated | 2206.03943 | null | https://arxiv.org/abs/2206.03943v1 | https://arxiv.org/pdf/2206.03943v1.pdf | Robust Environment Perception for Automated Driving: A Unified Learning Pipeline for Visual-Infrared Object Detection | The RGB complementary metal-oxidesemiconductor (CMOS) sensor works within the visible light spectrum. Therefore it is very sensitive to environmental light conditions. On the contrary, a long-wave infrared (LWIR) sensor operating in 8-14 micro meter spectral band, functions independent of visible light. In this paper, we exploit both visual and thermal perception units for robust object detection purposes. After delicate synchronization and (cross-) labeling of the FLIR [1] dataset, this multi-modal perception data passes through a convolutional neural network (CNN) to detect three critical objects on the road, namely pedestrians, bicycles, and cars. After evaluation of RGB and infrared (thermal and infrared are often used interchangeably) sensors separately, various network structures are compared to fuse the data at the feature level effectively. Our RGB-thermal (RGBT) fusion network, which takes advantage of a novel entropy-block attention module (EBAM), outperforms the state-of-the-art network [2] by 10% with 82.9% mAP. | ['Lutz Eckstein', 'Laurent Kloeker', 'Christian Mayr', 'Ali Kariminezhad', 'Mohsen Vadidar'] | 2022-06-08 | null | null | null | null | ['robust-object-detection'] | ['computer-vision'] | [ 7.27828920e-01 -2.60877550e-01 2.47231568e-03 -3.02919716e-01
-5.84674001e-01 -4.30649251e-01 4.42976713e-01 -1.66561127e-01
-7.67599523e-01 4.48550433e-01 -1.19229808e-01 -4.22068238e-01
4.29288238e-01 -9.85521019e-01 -6.39234960e-01 -9.75768209e-01
6.15217626e-01 -4.72786814e-01 3.76397610e-01 -2.55035609e-01
1.13108538e-01 4.56844449e-01 -1.98838437e+00 4.50711399e-01
6.58618808e-01 1.71792173e+00 3.43158096e-02 8.04683208e-01
1.20030835e-01 9.18457150e-01 -3.34198892e-01 -3.48192126e-01
2.11712226e-01 -1.33708656e-01 -2.86133498e-01 -5.37851393e-01
6.06197894e-01 -4.21672344e-01 -5.71857691e-01 1.14405453e+00
6.29837453e-01 2.01710358e-01 5.51269352e-01 -1.19427955e+00
-9.62224424e-01 4.31718439e-01 -5.84255695e-01 3.96603167e-01
2.54990876e-01 5.58391750e-01 7.47135401e-01 -6.26956403e-01
2.41441038e-02 1.04325283e+00 4.53767568e-01 4.53948617e-01
-6.23300195e-01 -7.78982222e-01 -2.01663122e-01 5.27950406e-01
-1.32514763e+00 -2.74741381e-01 7.54975736e-01 -1.52877718e-01
1.49697471e+00 3.16660494e-01 6.74296081e-01 1.00557590e+00
5.80777168e-01 3.63955885e-01 1.20246303e+00 -1.07113823e-01
2.64865398e-01 1.08943276e-01 2.64466345e-01 3.80810380e-01
3.84038776e-01 4.04216260e-01 -5.80093026e-01 5.40528476e-01
6.78442866e-02 2.30639413e-01 -1.74824342e-01 4.14251447e-01
-9.47252274e-01 2.13095009e-01 1.23105907e+00 3.01021278e-01
-9.87794027e-02 6.72785819e-01 1.60193041e-01 1.83427140e-01
1.10869855e-01 -9.71256569e-02 -2.51657248e-01 1.42139256e-01
-4.51052934e-01 -5.28053343e-01 2.92333394e-01 6.33948386e-01
1.07732308e+00 3.20779234e-02 3.37611884e-02 3.42702091e-01
6.47158563e-01 7.94772208e-01 2.15252817e-01 -6.35429323e-01
5.20933449e-01 8.18583012e-01 -1.07733399e-01 -4.96957719e-01
-5.88853598e-01 -2.80681968e-01 -1.09691811e+00 5.54069996e-01
1.61705092e-02 -9.75053012e-02 -1.22224319e+00 1.17435324e+00
1.17620386e-01 7.41693527e-02 3.48590463e-01 1.01528633e+00
1.29262650e+00 6.51830733e-01 1.44195348e-01 -7.09756017e-02
1.51862919e+00 -6.90088153e-01 -5.82694590e-01 -5.81295490e-01
1.70370474e-01 -3.52636397e-01 7.69414961e-01 1.65219799e-01
-6.97160959e-01 -1.05523872e+00 -1.44216406e+00 -6.21878147e-01
-1.07257569e+00 8.60425606e-02 3.14408958e-01 1.07354963e+00
-1.17219841e+00 2.29068667e-01 -7.32747972e-01 -1.38097212e-01
4.60508972e-01 4.06700969e-01 -1.42109603e-01 -6.06867224e-02
-1.31625271e+00 9.41635191e-01 3.83765936e-01 5.69933414e-01
-3.99835050e-01 -3.30917150e-01 -7.50314951e-01 -5.22205047e-03
2.43565142e-01 -4.57547128e-01 7.98854589e-01 -9.46523011e-01
-1.71059549e+00 7.48509109e-01 -6.42455295e-02 -4.35578972e-01
2.11629912e-01 -2.21843123e-01 -6.36364222e-01 2.72241592e-01
-3.82555515e-01 5.91798842e-01 7.61533260e-01 -1.03875899e+00
-6.37122154e-01 -6.08456016e-01 2.27525290e-02 4.09935927e-03
-2.84572452e-01 4.02647033e-02 -3.04825604e-01 9.85082015e-02
1.97186410e-01 -8.59250426e-01 1.01645224e-01 -4.82530007e-03
-4.73245561e-01 -1.80556282e-01 9.74617660e-01 -3.52517158e-01
8.97456348e-01 -2.27442312e+00 -4.51765299e-01 9.65294689e-02
2.29684636e-01 3.57626915e-01 1.52443036e-01 5.29812574e-02
-2.46600047e-01 1.58289135e-01 -2.87707090e-01 -2.15409994e-01
-9.92473960e-02 -6.45688102e-02 -8.95677879e-02 7.49841571e-01
1.19799912e-01 1.20460129e+00 -6.97145641e-01 -2.75841713e-01
7.08790839e-01 7.97547221e-01 1.38870075e-01 -2.72779584e-01
2.97556706e-02 3.30922186e-01 -2.97034860e-01 9.62953329e-01
9.42254603e-01 -2.39841249e-02 -6.99608773e-02 -6.39382422e-01
-4.80312884e-01 1.51997417e-01 -7.43294895e-01 1.46816421e+00
-3.85575980e-01 9.25921798e-01 -1.02382295e-01 -6.53700531e-01
1.04341972e+00 2.65578866e-01 4.13651347e-01 -1.44341075e+00
5.21917522e-01 3.10455084e-01 -1.56167641e-01 -6.20372474e-01
5.32123804e-01 -3.92661020e-02 -8.98502767e-02 2.68360577e-03
-3.33695531e-01 1.56829774e-01 -2.73580849e-01 -2.73506373e-01
1.06798387e+00 3.25756706e-02 -9.73046385e-03 3.37088376e-01
4.99083519e-01 -5.47975898e-01 3.86651158e-01 4.89100635e-01
-4.98185664e-01 5.02685905e-01 -8.33817646e-02 -4.66079235e-01
-6.55671120e-01 -1.19998157e+00 -1.10244080e-02 1.00866568e+00
6.38788223e-01 1.96079840e-03 -3.72334868e-01 -1.11439012e-01
-3.06295287e-02 6.75258636e-01 -4.68925506e-01 -3.99003178e-01
-5.06046474e-01 -9.02743459e-01 5.61250865e-01 8.34717512e-01
1.29442871e+00 -9.67380583e-01 -1.56660330e+00 6.57154247e-02
-1.87487260e-01 -1.45330727e+00 2.01253593e-01 5.23145676e-01
-4.06860679e-01 -1.02607167e+00 -8.50154236e-02 -2.66347140e-01
-4.28054668e-02 4.51709151e-01 9.31509197e-01 -1.91186950e-01
-5.97527921e-01 3.60216767e-01 -2.50054866e-01 -5.07940292e-01
-1.74052715e-01 -2.64638305e-01 -2.41514757e-01 2.18918234e-01
8.22102368e-01 -4.39738721e-01 -9.56829607e-01 -1.32154226e-01
-8.58891010e-01 -7.03062490e-02 8.09364915e-01 1.60136476e-01
3.88089746e-01 1.38801605e-01 4.87264767e-02 -1.69383779e-01
-5.83602339e-02 -8.00225884e-02 -6.92013025e-01 1.82861045e-01
-3.50301117e-01 -7.26590976e-02 5.35723865e-01 -1.72147583e-02
-1.02487087e+00 3.43619555e-01 -1.50633961e-01 -2.77767152e-01
-4.46749419e-01 8.85556731e-03 -3.91364247e-01 -2.42091075e-01
7.54887819e-01 1.11988522e-01 -4.96261597e-01 -2.14108318e-01
3.37639689e-01 1.11151898e+00 8.55735719e-01 3.01984549e-01
6.93597436e-01 8.41919303e-01 2.77823448e-01 -7.89077520e-01
-5.15788257e-01 -5.14776289e-01 -5.76819420e-01 -6.01264417e-01
1.43233240e+00 -1.18464220e+00 -1.30884445e+00 6.35422111e-01
-1.03463614e+00 8.17617476e-02 1.11383468e-01 4.73371267e-01
-1.85879245e-01 1.16318025e-01 -4.33572054e-01 -1.29025924e+00
-5.89166880e-01 -9.16310608e-01 1.06551743e+00 4.89536017e-01
4.48395252e-01 -3.45845342e-01 -3.20736319e-01 5.34290493e-01
4.25063103e-01 5.83878160e-01 4.34371293e-01 2.29958579e-01
-8.91190290e-01 -2.81028420e-01 -6.77060485e-01 5.86895645e-01
-2.73841806e-02 2.40700915e-01 -1.69096935e+00 4.95345034e-02
-7.88423046e-02 -2.51183301e-01 1.74686348e+00 3.34050536e-01
1.05113709e+00 1.56497419e-01 -2.65913904e-01 6.90651894e-01
1.81194508e+00 3.13155085e-01 1.08395684e+00 2.01105863e-01
9.54203963e-01 3.34143728e-01 2.04395309e-01 2.00832844e-01
5.89562833e-01 3.89936209e-01 1.03308094e+00 -2.53877282e-01
-4.31408547e-02 2.77400196e-01 7.03770339e-01 2.11802065e-01
-2.42448777e-01 -4.68298584e-01 -8.75879288e-01 1.56358495e-01
-1.51600444e+00 -1.05258393e+00 -4.60768193e-01 2.07294321e+00
4.12593484e-01 2.05710053e-01 -1.87127665e-01 4.38746959e-01
6.49183691e-01 1.82523161e-01 -8.77328396e-01 -5.12863636e-01
-2.81198204e-01 2.11946413e-01 1.16628397e+00 1.48673087e-01
-1.23968089e+00 5.06510317e-01 5.51692915e+00 3.20426702e-01
-1.41875911e+00 1.65254429e-01 6.59231544e-01 1.59408078e-02
-1.67768836e-01 -2.49262258e-01 -5.47523677e-01 4.96635199e-01
1.26750767e+00 5.68135500e-01 4.61263657e-01 4.15308088e-01
8.12210049e-03 -4.95796382e-01 -9.11427677e-01 1.34671843e+00
-3.70222959e-04 -8.50138545e-01 -5.13239145e-01 -8.21044296e-02
4.73685056e-01 5.08596122e-01 4.04762477e-01 3.32225338e-02
-1.13911539e-01 -1.07399285e+00 7.49966502e-01 6.47356033e-01
1.07164156e+00 -7.48961687e-01 7.46916473e-01 8.63156617e-02
-1.67365670e+00 -6.26404166e-01 -3.86006624e-01 -4.00303267e-02
-2.29071617e-01 8.09362590e-01 -1.21071395e-02 6.46353245e-01
1.11090171e+00 8.35281610e-01 -5.46436787e-01 5.92593133e-01
-1.55142412e-01 3.59120697e-01 -4.64782506e-01 -4.72285040e-02
2.89294094e-01 -1.71543419e-01 1.56977147e-01 1.10393381e+00
2.60304242e-01 1.80505142e-01 -2.41491124e-01 1.05166149e+00
-1.33470312e-01 -5.32310128e-01 -5.64258873e-01 9.90607589e-02
1.27409711e-01 1.44685662e+00 -6.71894968e-01 -4.51909602e-01
-6.44033194e-01 9.18423176e-01 -2.61078745e-01 5.97481906e-01
-1.02970481e+00 -4.57046360e-01 7.33079612e-01 -2.65012950e-01
3.41446519e-01 -1.88183621e-01 -5.97097993e-01 -9.21060920e-01
2.11788386e-01 -1.26049174e-02 1.43758193e-01 -1.11914003e+00
-8.49255562e-01 4.61696655e-01 -4.89372075e-01 -1.34267831e+00
2.53542691e-01 -9.81200457e-01 -2.84054995e-01 9.00781751e-01
-2.19042897e+00 -1.06002712e+00 -8.22204888e-01 5.59492707e-01
-1.14250913e-01 4.02685642e-01 3.69720668e-01 2.55831361e-01
-7.47043252e-01 4.41958994e-01 2.01774567e-01 1.84904113e-01
3.89072955e-01 -9.69897687e-01 4.40554023e-01 9.13058281e-01
-4.49454069e-01 2.41077706e-01 2.73722023e-01 -3.36167663e-01
-1.85472286e+00 -1.45327973e+00 5.47588050e-01 -2.39693448e-01
3.13523620e-01 -2.68656999e-01 -6.61720753e-01 3.10016692e-01
3.85509253e-01 3.01977992e-01 4.38537568e-01 -6.38786435e-01
-7.79837191e-01 -6.75761580e-01 -1.25476933e+00 2.76131004e-01
9.57803726e-01 -7.51408875e-01 -1.35765731e-01 9.12058428e-02
7.16152430e-01 -2.39958420e-01 -6.03503227e-01 5.55080175e-01
7.18458891e-01 -1.55949831e+00 1.11959457e+00 2.47723415e-01
3.18723857e-01 -5.18776774e-01 -3.27326655e-01 -6.91127598e-01
-1.67650923e-01 -4.10266608e-01 2.52778595e-03 6.38072312e-01
1.20746553e-01 -9.56158102e-01 4.20521289e-01 4.17522103e-01
-2.45534107e-01 -2.57658482e-01 -1.39699280e+00 -4.75195527e-01
-3.76631826e-01 -8.36900771e-01 4.83307987e-01 2.34762549e-01
-1.87009141e-01 3.52825284e-01 -1.22804374e-01 5.03031850e-01
5.85973620e-01 9.49471593e-02 1.61520943e-01 -1.09766626e+00
1.32961154e-01 -5.03845870e-01 -7.50023007e-01 -8.20516467e-01
-2.06933081e-01 -6.49986207e-01 2.70540148e-01 -1.72277057e+00
-1.33202597e-01 1.32497281e-01 -7.82688618e-01 5.46725750e-01
-1.79920644e-01 8.91997457e-01 2.23835215e-01 -2.61155397e-01
-6.76216185e-01 4.85368967e-01 9.87957954e-01 -3.73418093e-01
-4.44735885e-02 -2.31677711e-01 -6.69500232e-01 4.91389781e-01
8.16967428e-01 -5.43356687e-02 -8.91466960e-02 -4.14217591e-01
6.25237465e-01 -3.13745916e-01 7.61799157e-01 -1.60352600e+00
3.24475080e-01 1.24923691e-01 6.15647435e-01 -8.52330327e-01
5.72763205e-01 -1.09008455e+00 2.49288380e-02 6.37435555e-01
-5.49794324e-02 -1.81979656e-01 1.54311195e-01 6.24325156e-01
5.03116101e-02 2.95450747e-01 1.04487467e+00 -1.11497141e-01
-9.89483953e-01 1.16232917e-01 -5.76482236e-01 -4.29060876e-01
9.68342960e-01 -7.07450092e-01 -8.21694613e-01 -1.93636850e-01
-9.92272571e-02 -3.00173070e-02 2.37999991e-01 2.90298313e-01
8.50274146e-01 -9.49614346e-01 -3.93833131e-01 2.32792929e-01
2.37963110e-01 -4.62409146e-02 4.18498009e-01 7.88307250e-01
-3.14989984e-01 4.87465322e-01 -1.34641454e-01 -7.09811509e-01
-1.07896948e+00 6.46356523e-01 6.09149396e-01 3.24250370e-01
-4.53617752e-01 9.05828953e-01 -1.23209603e-01 -5.91851622e-02
1.79125533e-01 -9.16344464e-01 -2.62602359e-01 9.95850042e-02
6.08434796e-01 6.38818324e-01 2.22188100e-01 -8.15634489e-01
-7.57995903e-01 9.55869377e-01 6.39582217e-01 -7.10820733e-03
1.06500030e+00 -4.07024920e-01 -1.64049268e-01 6.63670361e-01
1.46263802e+00 -6.58781171e-01 -1.19628894e+00 -1.99635506e-01
-4.77481544e-01 -1.90419797e-02 4.11450565e-01 -8.08108926e-01
-1.18081915e+00 1.28258348e+00 1.23656392e+00 3.04148704e-01
1.73472524e+00 -3.19355547e-01 9.06600714e-01 4.85076338e-01
8.13238248e-02 -1.33936071e+00 1.79326739e-02 4.05737400e-01
3.55881214e-01 -1.31584656e+00 -7.07060993e-02 -3.28601807e-01
-3.06366414e-01 1.42423308e+00 5.85011780e-01 2.75297046e-01
6.17606878e-01 4.03643042e-01 1.35349393e-01 1.45517290e-02
-6.31851256e-01 -7.49682426e-01 1.82665467e-01 5.83000839e-01
2.59041160e-01 1.18922472e-01 3.37857962e-01 5.20242155e-02
4.32120748e-02 2.48632319e-02 3.43611687e-01 9.32295501e-01
-7.85996675e-01 -1.79877609e-01 -4.63832289e-01 3.39674383e-01
-2.28627518e-01 -2.03158811e-01 -3.76981795e-01 2.24358633e-01
5.98767996e-01 1.45644641e+00 2.23380417e-01 -8.63067985e-01
1.78569302e-01 -2.22016312e-02 2.45857939e-01 -6.04128800e-02
-7.47740209e-01 -5.25334626e-02 -2.63976038e-01 -7.85530031e-01
-9.41574097e-01 -2.07807258e-01 -1.35300946e+00 -3.49810392e-01
-3.60672057e-01 -7.19821393e-01 1.15212178e+00 9.97474968e-01
2.61048943e-01 9.32152152e-01 7.59265006e-01 -7.09936559e-01
-3.78039107e-03 -9.12000656e-01 -4.43955749e-01 -5.75558795e-03
7.58863628e-01 -3.83477122e-01 -3.31942707e-01 2.69128419e-02] | [9.466604232788086, -1.3364198207855225] |
61941fd6-569e-42b5-b98f-c57b0d0ec140 | wikicoder-learning-to-write-knowledge-powered | 2303.08574 | null | https://arxiv.org/abs/2303.08574v1 | https://arxiv.org/pdf/2303.08574v1.pdf | WikiCoder: Learning to Write Knowledge-Powered Code | We tackle the problem of automatic generation of computer programs from a few pairs of input-output examples. The starting point of this work is the observation that in many applications a solution program must use external knowledge not present in the examples: we call such programs knowledge-powered since they can refer to information collected from a knowledge graph such as Wikipedia. This paper makes a first step towards knowledge-powered program synthesis. We present WikiCoder, a system building upon state of the art machine-learned program synthesizers and integrating knowledge graphs. We evaluate it to show its wide applicability over different domains and discuss its limitations. WikiCoder solves tasks that no program synthesizers were able to solve before thanks to the use of knowledge graphs, while integrating with recent developments in the field to operate at scale. | ['Gaëtan Margueritte', 'Nathanaël Fijalkow', 'Théo Matricon'] | 2023-03-15 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 5.78234568e-02 6.04340672e-01 -3.38428259e-01 -1.22765735e-01
-4.01296705e-01 -7.40455806e-01 6.34893775e-01 4.07351494e-01
-2.72818625e-01 7.16070294e-01 -9.83153805e-02 -6.01804614e-01
-2.00784475e-01 -1.28363204e+00 -9.27270532e-01 1.75377697e-01
1.85024753e-01 4.93669331e-01 7.39088416e-01 -5.59704900e-01
3.41833442e-01 1.39829904e-01 -1.92916989e+00 3.23569596e-01
1.01785910e+00 1.27741545e-01 3.18855584e-01 7.90757477e-01
-6.85236990e-01 9.54809189e-01 -6.80654407e-01 -6.83834434e-01
2.08084937e-02 -3.65162343e-01 -1.12801433e+00 -4.66455609e-01
3.75308424e-01 1.00527093e-01 4.88413796e-02 1.06296706e+00
1.19954422e-01 -7.42766783e-02 3.71693283e-01 -1.57766473e+00
-7.31462836e-01 9.92601097e-01 2.36579448e-01 -7.82160759e-02
7.40131259e-01 1.16953447e-01 9.53571796e-01 -6.22973561e-01
1.06241167e+00 9.07551587e-01 8.23082745e-01 7.47036040e-01
-1.43391991e+00 -5.02288863e-02 -2.23335341e-01 2.42705494e-01
-1.10972071e+00 -2.35681728e-01 5.54669380e-01 -8.73144746e-01
1.57948101e+00 2.63651311e-01 7.31342912e-01 6.09705448e-01
-3.32875252e-01 6.07370436e-01 1.04120374e+00 -9.39117253e-01
5.03225252e-02 7.39909172e-01 4.55354214e-01 1.11929774e+00
4.35486227e-01 1.37882203e-01 -3.69204521e-01 -4.48796093e-01
7.91599393e-01 -5.79790056e-01 -3.97574067e-01 -7.70452321e-01
-1.29189646e+00 6.59828603e-01 2.04686642e-01 7.85961568e-01
7.75139332e-02 3.19815516e-01 5.69161296e-01 6.63174272e-01
-2.13750556e-01 9.62234318e-01 -6.23907089e-01 -3.24807227e-01
-9.15195584e-01 4.88915354e-01 1.59922600e+00 1.44227183e+00
1.08311689e+00 -6.28305366e-03 4.12172601e-02 2.95801133e-01
9.86024290e-02 3.13219130e-01 4.85116541e-01 -7.56930649e-01
5.81847787e-01 1.19040656e+00 5.12858108e-02 -8.68418694e-01
-9.62108076e-02 4.04382311e-02 2.54745781e-02 3.08221251e-01
5.04503489e-01 -5.56265935e-02 -4.48966980e-01 1.39819658e+00
5.04293218e-02 -1.54913902e-01 3.08657855e-01 3.17200899e-01
1.10965538e+00 3.70662957e-01 -7.47907311e-02 -1.21256493e-01
1.13172865e+00 -1.10386968e+00 -4.74067271e-01 -1.01381466e-01
1.09110570e+00 -5.27686715e-01 1.16641366e+00 4.50471312e-01
-8.77776504e-01 -5.35685539e-01 -1.09914780e+00 4.31098603e-02
-9.04372215e-01 1.38513297e-01 9.07822788e-01 9.38551068e-01
-1.24844766e+00 7.68753052e-01 -5.36705256e-01 -6.68240368e-01
7.57937059e-02 3.77941638e-01 -2.85550296e-01 1.65893644e-01
-9.38840091e-01 1.25328732e+00 9.47168827e-01 -4.92337704e-01
-5.93432724e-01 -8.86027694e-01 -9.19450641e-01 2.74996497e-02
7.94416428e-01 -9.81322765e-01 1.50861204e+00 -1.33743918e+00
-1.45121121e+00 1.05041087e+00 9.99599025e-02 -4.18849140e-01
5.14018416e-01 6.99937195e-02 -4.39570606e-01 -1.31792828e-01
-1.19125105e-01 2.14288026e-01 3.59096378e-01 -1.12014890e+00
-5.32362580e-01 4.61052470e-02 5.68003058e-01 -3.42288196e-01
-3.82607400e-01 1.04438521e-01 -3.09676588e-01 -3.99151593e-01
-6.10818088e-01 -8.51610959e-01 -2.83815235e-01 -3.48527521e-01
-2.64295697e-01 -4.00974214e-01 5.98374546e-01 -4.89018738e-01
1.50629354e+00 -1.71339643e+00 5.05204201e-01 3.27764809e-01
2.53719330e-01 5.80467105e-01 -2.83369422e-03 8.72190177e-01
-8.56338143e-02 3.20201397e-01 -3.18462610e-01 4.46231753e-01
5.36650538e-01 3.02888721e-01 -2.88615584e-01 -1.69011831e-01
1.68277577e-01 1.05307746e+00 -1.21024144e+00 -5.57092726e-01
1.19525686e-01 -3.38315144e-02 -6.93148971e-01 1.66242212e-01
-1.02496874e+00 -1.74476460e-01 -5.73972106e-01 3.18040133e-01
2.31884003e-01 -1.93346143e-01 4.00970221e-01 -5.33552989e-02
-2.45396495e-01 3.40883911e-01 -1.34616661e+00 1.93761635e+00
-7.42031693e-01 6.12134993e-01 -1.76602006e-01 -1.05321920e+00
1.09389901e+00 4.28966105e-01 -1.50012612e-01 -1.07160702e-01
-1.97765768e-01 5.18478692e-01 -6.49910420e-02 -8.68909836e-01
6.17482543e-01 1.11158274e-01 -1.16572335e-01 5.61838031e-01
5.02926528e-01 -7.69643664e-01 7.53294468e-01 3.69089365e-01
1.33592999e+00 6.21436477e-01 7.74495900e-01 -3.71914089e-01
8.11987877e-01 7.64641225e-01 2.35096067e-01 8.56866479e-01
3.76067460e-01 8.43944103e-02 8.36572826e-01 -4.37433898e-01
-1.12586761e+00 -6.59594357e-01 1.86221927e-01 9.59132016e-01
-4.56796110e-01 -1.08904254e+00 -9.03237700e-01 -9.84409511e-01
-1.31900981e-01 7.50266612e-01 -3.97586435e-01 2.41037816e-01
-5.87505698e-01 -2.67688721e-01 7.66393363e-01 5.18859506e-01
2.21881986e-01 -1.21770930e+00 -8.34297538e-01 1.47653997e-01
1.74519897e-01 -7.33037233e-01 1.08989693e-01 7.47789964e-02
-8.20729852e-01 -1.59879768e+00 -3.40573013e-01 -1.00374639e+00
6.33626342e-01 -3.11096132e-01 1.87421095e+00 4.99537796e-01
-2.67980486e-01 8.10387015e-01 -3.95852685e-01 -4.26055342e-01
-1.00371706e+00 4.45506096e-01 -3.94230664e-01 -8.38470638e-01
6.06746316e-01 -7.58275390e-01 2.31999069e-01 -6.08991608e-02
-1.02063537e+00 1.63122445e-01 4.79555547e-01 7.03404248e-01
2.25977808e-01 1.14145949e-01 5.15724361e-01 -1.51519847e+00
7.92004824e-01 -3.59612405e-01 -1.13063121e+00 7.33669460e-01
-5.30264199e-01 4.15865809e-01 7.32106209e-01 -2.14261144e-01
-9.69254315e-01 2.91244656e-01 8.30449089e-02 2.95694657e-02
-2.32551292e-01 9.01038170e-01 -2.55919933e-01 -2.22903371e-01
1.18041348e+00 2.08870381e-01 -2.27839500e-01 -2.58896410e-01
6.87452972e-01 3.52629155e-01 4.16654408e-01 -1.18017995e+00
1.00222981e+00 -2.13090777e-01 -9.25587676e-03 -6.20618641e-01
-2.03409031e-01 -2.99489498e-01 -7.26456165e-01 1.65220693e-01
4.29578453e-01 -5.99198222e-01 -5.49014986e-01 9.61096138e-02
-1.41461825e+00 -7.51039922e-01 -5.57170868e-01 7.95301795e-02
-8.61466229e-01 2.73978740e-01 -1.35698587e-01 -4.21624094e-01
-2.20996901e-01 -1.06759655e+00 4.95516866e-01 1.70519918e-01
-4.86327469e-01 -1.20705605e+00 5.00828445e-01 -6.13210499e-02
4.81918156e-01 1.75801128e-01 1.31347847e+00 -8.86881709e-01
-8.05114925e-01 -5.57241924e-02 2.48248931e-02 3.65014851e-01
-3.69569436e-02 4.40640837e-01 -7.83962131e-01 3.50245804e-01
-4.06411797e-01 -1.59655035e-01 3.84362757e-01 -4.13074881e-01
6.46968663e-01 -3.50775987e-01 -4.72870648e-01 1.59271330e-01
1.82484877e+00 -1.03789948e-01 6.91280246e-01 4.69828993e-01
6.03819489e-01 6.93563938e-01 2.90513426e-01 -2.79517882e-02
3.53248179e-01 8.56090188e-01 7.58182555e-02 3.50867927e-01
-2.06101224e-01 -4.66684431e-01 2.43317366e-01 7.60434985e-01
-9.73810926e-02 2.78651774e-01 -1.35532498e+00 1.06589437e+00
-2.07382727e+00 -9.00323391e-01 -4.81073290e-01 2.13546991e+00
1.12326741e+00 1.39184475e-01 2.47683406e-01 2.81051882e-02
4.42903280e-01 -3.39996189e-01 -7.71954954e-02 -6.27972722e-01
1.55939788e-01 6.59097075e-01 2.34887034e-01 5.08528590e-01
-7.55716920e-01 9.81953979e-01 6.39353704e+00 7.58669198e-01
-6.68323696e-01 4.74175736e-02 -4.02742624e-01 3.63491595e-01
-5.82365632e-01 4.85536039e-01 -6.74689293e-01 1.06544588e-02
1.06327307e+00 -9.20067191e-01 6.11725211e-01 1.42396200e+00
-4.17086840e-01 -2.13547230e-01 -1.59839785e+00 5.34350455e-01
3.84031907e-02 -1.62351608e+00 -8.82001370e-02 -4.38279450e-01
8.73196900e-01 -1.60520986e-01 -4.58624303e-01 6.04668796e-01
8.49504113e-01 -1.00252497e+00 5.21335483e-01 6.37032986e-01
4.07438457e-01 -5.29966414e-01 5.60046613e-01 4.43450987e-01
-1.17147291e+00 9.94124413e-02 -8.55730176e-02 -1.69932365e-01
-2.58910358e-01 6.69589162e-01 -1.03363311e+00 8.62404704e-01
3.11591804e-01 6.15361273e-01 -8.35997105e-01 1.20063531e+00
-6.67843580e-01 2.37828583e-01 -3.31342846e-01 -3.31398427e-01
-1.59342617e-01 -4.89822328e-02 4.33642209e-01 1.48457265e+00
5.91211915e-01 3.08844708e-02 1.86671048e-01 1.41324282e+00
8.59475955e-02 2.66019762e-01 -1.14310575e+00 -3.06796312e-01
3.02951992e-01 1.15675581e+00 -4.29313332e-01 -7.76358724e-01
-7.26700008e-01 5.23978829e-01 4.46019709e-01 1.71136171e-01
-5.75504005e-01 -8.28171551e-01 1.66635782e-01 6.86023105e-03
2.39538357e-01 -2.73581207e-01 -1.10657260e-01 -1.29145885e+00
2.53868103e-01 -1.10658288e+00 2.51613140e-01 -9.42295432e-01
-8.89537454e-01 6.89585388e-01 3.02563548e-01 -9.01708901e-01
-5.11112869e-01 -7.08054423e-01 -6.14414334e-01 9.83166516e-01
-1.43202281e+00 -1.09716797e+00 -3.32953930e-01 4.25389767e-01
1.87424436e-01 -2.24033445e-01 1.18521631e+00 1.09166220e-01
-1.41165644e-01 2.92839766e-01 -3.66629064e-01 2.40408942e-01
5.11887014e-01 -1.57205462e+00 4.39795882e-01 8.39061916e-01
2.33829498e-01 9.84133840e-01 7.66541362e-01 -6.85324192e-01
-1.79114568e+00 -9.11398411e-01 1.15077233e+00 -8.43024433e-01
1.09014392e+00 -1.08709157e-01 -9.74144459e-01 8.63995373e-01
3.26515585e-01 4.65920269e-02 5.80098629e-01 1.40172929e-01
-6.63995564e-01 1.65763900e-01 -8.57977688e-01 4.63004977e-01
1.16697764e+00 -7.22516894e-01 -1.08398783e+00 4.28519100e-01
6.09130561e-01 -5.75743675e-01 -1.03006709e+00 1.77712664e-01
2.19805002e-01 -1.04683089e+00 6.54720426e-01 -6.40929401e-01
6.96563184e-01 -6.33135617e-01 -2.60272180e-03 -1.46768022e+00
3.39401245e-01 -7.57616758e-01 -2.99876899e-01 1.47377038e+00
7.39559174e-01 -5.44853508e-01 7.21223176e-01 8.01384091e-01
-1.81208417e-01 -1.13600969e-01 -4.74048048e-01 -9.39697623e-01
1.10095575e-01 -4.90029007e-01 7.05827534e-01 1.03189254e+00
8.46775413e-01 3.03986847e-01 1.00476526e-01 -1.26398817e-01
3.19573551e-01 4.27561849e-01 1.27171111e+00 -1.27074862e+00
-8.14333737e-01 -5.93567669e-01 -4.06726927e-01 -5.51674485e-01
2.84210742e-01 -1.27970243e+00 -2.70395935e-01 -1.65100896e+00
-2.17818227e-02 -4.64754194e-01 3.47681373e-01 7.52007484e-01
4.39230986e-02 -2.35993385e-01 2.52525151e-01 -1.05939195e-01
-5.61856449e-01 1.16928164e-02 8.84736657e-01 -1.78011298e-01
-3.83445024e-01 -6.64351135e-02 -5.58942795e-01 6.86400533e-01
6.71519518e-01 -6.42377496e-01 -3.74668539e-01 -2.58285403e-01
9.65375125e-01 5.38635664e-02 4.48940903e-01 -1.16833067e+00
4.54303920e-01 -2.72433192e-01 -4.51858431e-01 1.23890573e-02
-2.10053012e-01 -1.06308603e+00 5.31918228e-01 5.09435773e-01
-2.21815422e-01 4.03259806e-02 4.07812983e-01 2.81312287e-01
-3.01067024e-01 -9.22535121e-01 3.41564745e-01 -6.55549228e-01
-1.21751571e+00 -4.12476212e-01 -3.14004153e-01 1.19077109e-01
1.09748197e+00 3.06288153e-03 -7.30449557e-01 2.02349260e-01
-6.07520282e-01 4.60189879e-02 9.03749824e-01 2.86409110e-01
2.17664242e-01 -1.14309037e+00 -3.36745590e-01 7.23427087e-02
4.76324975e-01 -3.52047861e-01 -3.17747325e-01 6.21183515e-01
-8.29464257e-01 6.07223332e-01 -2.80667871e-01 -2.42118925e-01
-1.31802559e+00 8.96086574e-01 3.44733626e-01 -4.62823957e-01
-3.85160625e-01 3.40676665e-01 -3.18002641e-01 -9.37915921e-01
-2.59790011e-02 -4.79694277e-01 -2.53978163e-01 -1.92186028e-01
5.55005074e-01 1.94404647e-01 2.93480724e-01 -5.19876070e-02
-3.36342424e-01 5.39134145e-01 2.45189339e-01 -4.41894233e-02
1.30598104e+00 6.02088630e-01 -4.35607433e-01 3.89715910e-01
7.27257907e-01 4.55568917e-02 -5.37100494e-01 -5.01669526e-01
4.23668683e-01 -4.01078701e-01 -2.43756533e-01 -8.95317554e-01
-8.57454658e-01 6.92110181e-01 2.11836427e-01 4.69758421e-01
8.62835109e-01 -3.55114453e-02 2.17050642e-01 1.08252406e+00
6.65817261e-01 -1.02827573e+00 -1.18236996e-01 6.37540042e-01
7.40522206e-01 -1.11880040e+00 -4.43013087e-02 -6.75543904e-01
-3.99046004e-01 1.58604240e+00 6.69046760e-01 -1.03453413e-01
4.02498960e-01 5.10928273e-01 -2.13706702e-01 -4.00281131e-01
-7.30108857e-01 -3.96036088e-01 2.13907123e-01 1.05520415e+00
5.10492146e-01 -4.07610275e-02 -4.40771490e-01 6.98877215e-01
-2.61588633e-01 7.05211043e-01 9.15293336e-01 1.36188281e+00
-4.16056246e-01 -1.79160345e+00 -2.49160245e-01 4.47504282e-01
-2.50015080e-01 -2.75582135e-01 -5.19292235e-01 1.12111437e+00
6.38531148e-02 6.49166584e-01 -4.56570148e-01 -2.64608413e-01
6.16415858e-01 3.63258630e-01 8.59184384e-01 -1.19672799e+00
-1.01358438e+00 -6.72579050e-01 6.58202469e-01 -3.92473042e-01
-7.46650517e-01 -2.47406676e-01 -1.04448485e+00 -3.25778723e-01
-2.96131849e-01 4.67131525e-01 6.49880230e-01 7.95185566e-01
2.53746808e-01 5.59220612e-01 -2.03030944e-01 -5.37006259e-01
-2.30445236e-01 -4.12092865e-01 -2.71597475e-01 2.13598251e-01
3.07159461e-02 -2.81099319e-01 -2.03148890e-02 6.88200057e-01] | [8.27916145324707, 7.448019027709961] |
76eb9cbd-8e1c-457e-9711-33172003e134 | spatr-mocap-3d-human-action-recognition-based | 2306.17574 | null | https://arxiv.org/abs/2306.17574v1 | https://arxiv.org/pdf/2306.17574v1.pdf | SpATr: MoCap 3D Human Action Recognition based on Spiral Auto-encoder and Transformer Network | Recent advancements in technology have expanded the possibilities of human action recognition by leveraging 3D data, which offers a richer representation of actions through the inclusion of depth information, enabling more accurate analysis of spatial and temporal characteristics. However, 3D human action recognition is a challenging task due to the irregularity and Disarrangement of the data points in action sequences. In this context, we present our novel model for human action recognition from fixed topology mesh sequences based on Spiral Auto-encoder and Transformer Network, namely SpATr. The proposed method first disentangles space and time in the mesh sequences. Then, an auto-encoder is utilized to extract spatial geometrical features, and tiny transformer is used to capture the temporal evolution of the sequence. Previous methods either use 2D depth images, sample skeletons points or they require a huge amount of memory leading to the ability to process short sequences only. In this work, we show competitive recognition rate and high memory efficiency by building our auto-encoder based on spiral convolutions, which are light weight convolution directly applied to mesh data with fixed topologies, and by modeling temporal evolution using a attention, that can handle large sequences. The proposed method is evaluated on on two 3D human action datasets: MoVi and BMLrub from the Archive of Motion Capture As Surface Shapes (AMASS). The results analysis shows the effectiveness of our method in 3D human action recognition while maintaining high memory efficiency. The code will soon be made publicly available. | ['Lahoucine Ballihi', 'Hamza Bouzid'] | 2023-06-30 | null | null | null | null | ['action-recognition-in-videos', 'action-recognition'] | ['computer-vision', 'computer-vision'] | [ 2.68564969e-01 -4.35206205e-01 -2.40387544e-02 9.06328931e-02
-1.31413981e-01 -3.57618153e-01 6.28752887e-01 -3.67787510e-01
-4.97249871e-01 3.11529905e-01 2.29566500e-01 -1.11818239e-01
-4.51569036e-02 -8.28773320e-01 -6.87171221e-01 -6.98570371e-01
-2.55540252e-01 4.21464026e-01 3.82968903e-01 -1.45525441e-01
2.90722370e-01 9.22863543e-01 -1.67995155e+00 3.34426314e-01
5.29554665e-01 1.31107152e+00 1.12303317e-01 8.22671771e-01
-2.03783885e-01 7.91477442e-01 -4.56503183e-01 -1.35084480e-01
4.27610070e-01 -3.31209093e-01 -7.88471878e-01 2.65737832e-01
2.23905131e-01 -4.64952558e-01 -8.22068810e-01 5.90520263e-01
4.42836314e-01 2.37129569e-01 6.46338403e-01 -9.29502964e-01
-4.29784447e-01 7.53664374e-02 -5.08618176e-01 3.90984029e-01
5.23246467e-01 2.90156007e-01 7.04369247e-01 -7.01965332e-01
6.76128864e-01 1.31244874e+00 8.64630938e-01 5.14063239e-01
-8.85040879e-01 -5.45170844e-01 -1.28321037e-01 6.68449819e-01
-1.31412280e+00 -2.77511299e-01 8.66482735e-01 -5.46607196e-01
1.37458622e+00 1.26158237e-01 1.21774697e+00 1.22900641e+00
3.38585138e-01 9.32922781e-01 6.75176084e-01 -2.13062912e-01
1.35129839e-01 -6.57880425e-01 -2.19668090e-01 7.14202464e-01
-2.94413924e-01 1.80516466e-01 -5.47086418e-01 9.93878022e-02
1.34418356e+00 4.14305687e-01 -2.74335593e-01 -4.98037159e-01
-1.34708154e+00 5.28077364e-01 3.16724479e-01 4.23093289e-01
-5.86347580e-01 9.07886028e-02 6.73862696e-01 2.53974169e-01
4.19490337e-01 3.98100615e-02 -2.12763458e-01 -7.68239558e-01
-7.08298802e-01 1.91156536e-01 5.06705940e-01 7.00034678e-01
3.59606773e-01 9.16266516e-02 -6.30877316e-02 7.10047841e-01
-4.00241911e-02 6.99143589e-01 8.58001888e-01 -9.10309255e-01
4.55084682e-01 9.39848006e-01 -7.66139627e-02 -1.27896619e+00
-4.86604422e-01 2.18032300e-02 -1.06248426e+00 3.89239818e-01
3.70468944e-01 4.12968010e-01 -9.42294180e-01 1.47206926e+00
3.13289344e-01 2.43852869e-01 -5.11553921e-02 9.72683132e-01
4.61961478e-01 5.07280767e-01 -1.99932560e-01 9.46079865e-02
1.26517069e+00 -7.98969209e-01 -5.26253819e-01 2.77083546e-01
6.71805382e-01 -3.10645193e-01 8.92400742e-01 3.78147036e-01
-9.99382913e-01 -7.37381101e-01 -1.12938786e+00 -1.33530051e-01
-3.56331527e-01 7.65587017e-02 4.81944978e-01 2.97082901e-01
-7.92006493e-01 9.34767842e-01 -1.35694301e+00 -3.60617757e-01
6.39128208e-01 3.81103307e-01 -6.16918206e-01 5.92629239e-02
-1.10332501e+00 7.00314581e-01 3.35299015e-01 1.78574830e-01
-6.17637575e-01 -3.95630658e-01 -9.66974676e-01 -1.12372272e-01
1.75587952e-01 -6.04864895e-01 9.32798386e-01 -9.28261936e-01
-1.66883075e+00 6.89376712e-01 1.36943236e-01 -6.90166116e-01
7.27511704e-01 -2.23343253e-01 -3.18547100e-01 4.79390115e-01
-1.49955541e-01 4.60634828e-01 8.45342696e-01 -3.85030508e-01
-4.36247349e-01 -6.15545273e-01 1.26838461e-01 2.53657579e-01
-3.78340751e-01 -1.83496654e-01 -3.31856638e-01 -8.84455681e-01
9.97023508e-02 -9.10896599e-01 -2.24223416e-02 1.70366183e-01
6.74127117e-02 -1.41408786e-01 1.06126714e+00 -7.58051455e-01
1.18056512e+00 -2.27920318e+00 4.58965033e-01 -2.39109881e-02
1.01313360e-01 5.02983689e-01 -1.53580070e-01 3.39582175e-01
-6.57611042e-02 -1.43329754e-01 -3.99676919e-01 -2.23721310e-01
-1.89395517e-01 4.95919079e-01 -9.43978280e-02 5.78766584e-01
1.88484430e-01 1.01083481e+00 -7.62463808e-01 -4.76998091e-01
4.84394848e-01 5.98082244e-01 -4.67086196e-01 1.14751458e-01
4.08532508e-02 4.54398811e-01 -5.06549954e-01 5.87032080e-01
4.63904023e-01 -2.19850466e-01 -1.25032082e-01 -2.84172118e-01
-2.08837777e-01 1.08610496e-01 -1.15371799e+00 2.04222727e+00
-4.67117012e-01 5.25616527e-01 -2.15076402e-01 -1.22778785e+00
8.77898335e-01 3.69802624e-01 8.99913013e-01 -1.02082217e+00
1.90872416e-01 2.43302658e-01 -1.85739294e-01 -5.98570883e-01
2.62844950e-01 2.85017751e-02 6.77956268e-03 5.38994491e-01
-1.17855221e-01 7.05932006e-02 1.48669943e-01 -1.68696061e-01
1.37813258e+00 4.50725615e-01 2.67353296e-01 1.07736789e-01
7.17883468e-01 -9.61084738e-02 4.50026691e-01 2.87845463e-01
-5.71023934e-02 5.19303918e-01 2.81045169e-01 -9.17325974e-01
-1.30525386e+00 -9.03554976e-01 9.13641751e-02 4.08491045e-01
8.47834796e-02 -3.75094146e-01 -5.15822232e-01 -6.30631685e-01
-7.20987171e-02 1.57155052e-01 -9.10333276e-01 -1.83152065e-01
-1.07796419e+00 -2.83132881e-01 6.78639293e-01 8.72719347e-01
8.32449675e-01 -1.27386975e+00 -1.30314708e+00 2.23550409e-01
-6.94857314e-02 -1.15131915e+00 -5.42285740e-01 -1.63894698e-01
-1.13293517e+00 -1.16752803e+00 -8.62975240e-01 -4.19685543e-01
3.64598513e-01 9.64167193e-02 6.75761819e-01 1.19202891e-02
-5.86603165e-01 5.26368082e-01 -4.59985912e-01 1.12168873e-02
-1.81330010e-01 -4.24014218e-02 1.31084874e-01 2.61819750e-01
2.40341797e-01 -1.00575519e+00 -6.12770677e-01 4.37188417e-01
-1.04815435e+00 1.65366143e-01 5.97618282e-01 9.28084433e-01
6.44806027e-01 1.75705813e-02 1.04687005e-01 -2.43829593e-01
1.15833081e-01 -2.00262964e-01 -3.35886836e-01 3.17203403e-02
-4.09324840e-02 1.51448026e-01 6.54285967e-01 -6.62560999e-01
-8.59627664e-01 2.55304456e-01 -1.33905962e-01 -1.07766688e+00
-2.02397630e-01 1.91703871e-01 -8.25281143e-02 4.56881709e-02
4.15366709e-01 5.58489501e-01 4.36159670e-01 -6.67139351e-01
5.88084273e-02 5.23447037e-01 5.17716587e-01 -3.47369999e-01
4.12080377e-01 8.28442812e-01 2.45890707e-01 -1.07035100e+00
-2.93268889e-01 -3.56644601e-01 -1.04690278e+00 -3.11677575e-01
1.01343620e+00 -5.37004411e-01 -7.92205334e-01 9.70843792e-01
-1.20090795e+00 -4.04391170e-01 -3.93064409e-01 5.29861987e-01
-9.61802661e-01 7.07434416e-01 -7.27504909e-01 -6.61337197e-01
-3.49260420e-01 -1.06588125e+00 1.15128434e+00 -1.99176356e-01
-2.51581192e-01 -9.43436801e-01 1.14675306e-01 1.82650238e-01
1.78092644e-01 6.13647282e-01 7.42662430e-01 -3.51994008e-01
-5.47530532e-01 -2.77468115e-01 4.49702726e-04 3.47520411e-01
2.67529547e-01 -2.17517599e-01 -5.02665102e-01 -8.21605176e-02
9.58780423e-02 -2.53842950e-01 7.07526505e-01 2.13140711e-01
1.39458907e+00 -3.00054789e-01 -2.85646677e-01 6.48739994e-01
1.03679895e+00 3.31762910e-01 8.68881047e-01 3.06705236e-01
8.37719619e-01 4.50545907e-01 4.96819049e-01 7.26343155e-01
7.04057366e-02 9.88750398e-01 3.74392241e-01 1.10700808e-01
-2.38875583e-01 -2.90128738e-01 3.94582599e-01 9.55682278e-01
-6.22306526e-01 4.47409302e-02 -9.78601396e-01 4.11744326e-01
-1.93970037e+00 -1.24081802e+00 2.22099405e-02 2.03365874e+00
4.99710590e-01 3.88558544e-02 3.40284318e-01 5.39354205e-01
3.98047626e-01 2.11953208e-01 -6.50538862e-01 -2.45291725e-01
-3.09560932e-02 4.04365480e-01 3.75203073e-01 2.10919321e-01
-1.01964378e+00 6.60511255e-01 5.41704369e+00 9.62358117e-01
-1.20352399e+00 -9.13209245e-02 1.38713762e-01 -3.29834402e-01
1.94101200e-01 -4.26262468e-01 -4.10729259e-01 4.85077918e-01
6.97948992e-01 3.35022025e-02 2.62964189e-01 6.44638240e-01
1.94593683e-01 1.49277344e-01 -1.10340738e+00 1.32185364e+00
1.17221706e-01 -1.35103250e+00 2.18909755e-01 2.34980345e-01
2.68548489e-01 -1.70935571e-01 -1.29059136e-01 1.37896650e-02
-1.96578354e-01 -1.11896062e+00 7.72249103e-01 8.65371406e-01
9.59049463e-01 -7.56817639e-01 5.75491130e-01 4.49820548e-01
-1.48856843e+00 -2.66333044e-01 -1.67284489e-01 -2.19427675e-01
2.74101108e-01 1.92577764e-01 -4.69106287e-01 6.28458023e-01
8.37768435e-01 1.20761180e+00 -3.05487454e-01 8.53716135e-01
1.88611820e-01 2.36468405e-01 -3.31366301e-01 -8.28250498e-02
2.55694002e-01 -1.48194194e-01 6.40503287e-01 1.04818273e+00
5.04832029e-01 4.76145267e-01 1.27956510e-01 6.16249561e-01
3.61783713e-01 -5.96441850e-02 -7.86396503e-01 -1.87717214e-01
-5.16509078e-03 7.08559692e-01 -7.29589403e-01 -3.11549306e-01
-3.76823187e-01 1.27236140e+00 1.73884168e-01 4.97147366e-02
-8.04555178e-01 -2.82298326e-01 6.64721429e-01 1.92941502e-01
5.58929861e-01 -7.14051902e-01 -1.75928503e-01 -1.15759611e+00
1.57514751e-01 -9.09004509e-01 3.84284645e-01 -6.04312778e-01
-9.39015388e-01 5.89658022e-01 2.01602206e-02 -1.59446490e+00
-3.52934152e-01 -8.17949831e-01 -2.70955205e-01 5.48073411e-01
-8.64450693e-01 -9.96129036e-01 -3.35169613e-01 8.24222505e-01
8.07894111e-01 -9.59085599e-02 7.18388498e-01 3.85912478e-01
-2.79027402e-01 4.00359839e-01 -1.16888545e-01 2.74027169e-01
9.99110714e-02 -9.11052644e-01 6.09379351e-01 4.21792597e-01
2.69988447e-01 2.05329522e-01 2.45279789e-01 -7.07058847e-01
-1.52420545e+00 -7.09020615e-01 7.31261730e-01 -5.11320591e-01
4.86390442e-01 -3.35711569e-01 -9.93870258e-01 3.96775573e-01
-2.31892645e-01 2.65175086e-02 4.19277340e-01 -3.46564382e-01
-1.63133517e-01 6.93157613e-02 -8.97956967e-01 5.92480600e-01
1.63157558e+00 -3.86239171e-01 -6.76863432e-01 5.85416481e-02
4.42114830e-01 -5.22656441e-01 -9.43511665e-01 5.38824201e-01
8.58522534e-01 -1.11981297e+00 1.13599551e+00 -5.02976000e-01
5.60882330e-01 -2.58710116e-01 -1.31597310e-01 -9.21974063e-01
-2.76085496e-01 -3.18362057e-01 -4.10902947e-01 5.43112457e-01
-1.36006474e-01 -4.18379992e-01 8.54386568e-01 2.34936729e-01
-1.77235901e-01 -1.00898516e+00 -1.20205569e+00 -9.23832595e-01
-1.49419740e-01 -5.74663162e-01 6.38208032e-01 6.20455801e-01
-8.06788132e-02 -1.33999527e-01 -3.84573936e-01 -2.05683410e-01
4.47055548e-01 1.43650919e-01 7.23214209e-01 -1.09632599e+00
-3.76082599e-01 -5.71683645e-01 -9.86127853e-01 -1.38450730e+00
1.84458926e-01 -6.59796774e-01 -2.09384397e-01 -1.21437383e+00
-1.08226292e-01 -1.80011034e-01 2.67619062e-02 5.00360131e-01
4.02492583e-01 4.23558265e-01 9.10706818e-02 2.53845602e-01
-4.14015770e-01 9.57076013e-01 1.48607779e+00 -2.44384155e-01
-1.04985170e-01 -3.83547470e-02 3.17400992e-01 7.92034805e-01
5.63890874e-01 -4.25472409e-02 -3.53361011e-01 -4.51707065e-01
-1.45503998e-01 2.39370570e-01 6.67711675e-01 -1.30896306e+00
1.62878856e-01 2.81017330e-02 5.63492179e-01 -7.43707359e-01
6.64038599e-01 -8.59759033e-01 3.12752604e-01 6.29562259e-01
-1.38804018e-01 3.76270831e-01 2.34773710e-01 6.39836550e-01
-2.48216242e-01 7.38872774e-03 5.93124568e-01 -2.87644148e-01
-7.99373448e-01 6.49454951e-01 -3.45576942e-01 -2.38429204e-01
9.67407644e-01 -7.40705848e-01 4.21462148e-01 -1.57684609e-01
-8.82797241e-01 -1.83263302e-01 6.37618840e-01 5.08901894e-01
9.12129641e-01 -1.70514059e+00 -4.32397574e-01 5.62093914e-01
-1.04718804e-01 -4.84285392e-02 5.85673094e-01 9.33068216e-01
-6.29234135e-01 4.33359385e-01 -7.11313248e-01 -7.24301577e-01
-1.29955971e+00 5.19465864e-01 4.43481743e-01 -3.01793933e-01
-1.27460170e+00 4.20058489e-01 7.25592896e-02 -7.95100555e-02
1.96036562e-01 -3.65853995e-01 -2.62323290e-01 -3.82201187e-02
6.24788821e-01 6.22223079e-01 -4.82215248e-02 -9.02101457e-01
-3.28262627e-01 9.78947341e-01 2.51141667e-01 -8.79772976e-02
1.30702841e+00 8.05898383e-02 1.18455380e-01 5.08862436e-01
1.31254172e+00 -4.36737865e-01 -1.51228130e+00 -2.12127402e-01
-1.99524328e-01 -7.19755054e-01 -2.17606023e-01 -2.53818601e-01
-1.11748409e+00 1.15486717e+00 5.53509831e-01 7.26788044e-02
1.18112051e+00 -2.42372945e-01 1.18122387e+00 2.10815132e-01
4.88863021e-01 -9.92716491e-01 4.95868534e-01 7.58755088e-01
1.08229935e+00 -7.24212348e-01 -2.34513223e-01 -1.36029899e-01
-4.26033437e-01 1.41402900e+00 4.84193414e-01 -2.42490694e-01
6.23761415e-01 1.40211642e-01 -2.73847491e-01 -3.36251944e-01
-4.86362517e-01 -3.05354651e-02 2.20115617e-01 5.76799393e-01
9.74029824e-02 -1.28606290e-01 -2.67189622e-01 3.77714843e-01
-2.09246114e-01 1.38272300e-01 2.05566540e-01 1.05017269e+00
-1.13235883e-01 -1.05674577e+00 -2.08061129e-01 3.21565330e-01
-2.37236485e-01 2.26333410e-01 -2.15070397e-01 7.58949161e-01
1.14806026e-01 3.65174204e-01 3.60781759e-01 -5.72459638e-01
5.62083840e-01 1.67954490e-01 7.99811244e-01 -2.31970161e-01
-2.54733562e-01 -2.37035170e-01 -1.21259183e-01 -8.69892359e-01
-5.93572199e-01 -9.14428711e-01 -1.23082948e+00 -3.03330898e-01
2.16362253e-01 -1.86124638e-01 4.38238829e-01 8.65452349e-01
5.81222117e-01 4.38617766e-01 4.53423470e-01 -1.17097068e+00
-5.04272103e-01 -8.73376787e-01 -5.29201686e-01 6.40166640e-01
2.81495273e-01 -1.11892557e+00 -2.24467471e-01 7.14974627e-02] | [7.842621326446533, 0.3511888086795807] |
644651ab-247d-445f-ab05-5ea5b0c48d21 | benchmarking-joint-lexical-and-syntactic | null | null | https://aclanthology.org/W17-1725 | https://aclanthology.org/W17-1725.pdf | Benchmarking Joint Lexical and Syntactic Analysis on Multiword-Rich Data | This article evaluates the extension of a dependency parser that performs joint syntactic analysis and multiword expression identification. We show that, given sufficient training data, the parser benefits from explicit multiword information and improves overall labeled accuracy score in eight of the ten evaluation cases. | ["H{\\'e}ctor Martinez Alonso", 'Matthieu Constant'] | 2017-04-01 | null | null | null | ws-2017-4 | ['lexical-analysis'] | ['natural-language-processing'] | [ 8.06342252e-03 2.45404050e-01 -6.48838103e-01 -1.02979970e+00
-1.37894952e+00 -7.90519655e-01 1.03756785e-01 3.13002527e-01
-7.40386248e-01 1.10988283e+00 3.66996795e-01 -3.94059956e-01
3.04381430e-01 -2.18875036e-01 -9.61769596e-02 -3.22580367e-01
-2.29096785e-01 2.70211667e-01 1.83890879e-01 -2.70857930e-01
-1.46773577e-01 2.68238932e-01 -1.07351851e+00 5.98435223e-01
5.08574307e-01 7.18990266e-01 -2.55432755e-01 9.41763818e-01
-5.60626924e-01 1.11433208e+00 -8.88983846e-01 -8.77268016e-01
-2.23302290e-01 -1.08066604e-01 -1.14979768e+00 -2.13033274e-01
1.37150913e-01 9.64528322e-02 4.29177582e-01 8.55850458e-01
4.94568229e-01 -1.23699486e-01 2.63104439e-01 -8.65502894e-01
-3.29749316e-01 7.73253381e-01 -3.52179915e-01 6.40623271e-01
7.67243564e-01 -4.23015743e-01 1.64430439e+00 -7.66513050e-01
8.77323091e-01 1.19235325e+00 7.50472605e-01 6.47749782e-01
-1.22764730e+00 -6.95094764e-01 6.24222681e-02 -3.83650064e-01
-1.20409489e+00 -3.21303844e-01 6.72180116e-01 -1.23294093e-01
1.79699874e+00 -1.53355105e-02 1.07677199e-01 1.00617206e+00
3.04139555e-01 8.06894958e-01 1.33805478e+00 -8.62700045e-01
-1.70617566e-01 -3.09561640e-02 1.22925127e+00 8.75231206e-01
1.54434219e-01 -6.24987036e-02 -3.05756092e-01 -3.41727406e-01
-5.88831678e-02 -9.25796211e-01 1.89920083e-01 4.53117937e-01
-5.59284866e-01 9.41121936e-01 -3.51723194e-01 6.02342010e-01
-1.46907136e-01 6.34890348e-02 9.03394043e-01 4.29946393e-01
7.82500148e-01 5.49239099e-01 -7.96268404e-01 -2.96605557e-01
-5.44136763e-01 -2.37812415e-01 1.15557349e+00 8.52480173e-01
5.62981606e-01 -3.78187560e-02 -4.89014946e-02 9.88755286e-01
5.45081720e-02 5.04838347e-01 5.30601799e-01 -8.17995310e-01
5.59268951e-01 5.54051936e-01 -1.46987051e-01 -2.98691362e-01
-8.17652881e-01 -2.01053709e-01 1.78091899e-01 -1.45214662e-01
2.58429378e-01 -7.12656677e-01 -7.13503957e-01 1.80943441e+00
2.66452342e-01 -3.89876932e-01 5.03543854e-01 3.62003982e-01
9.62331593e-01 3.98427993e-01 1.23955464e+00 -4.90739584e-01
1.77564323e+00 -7.07628548e-01 -1.05329108e+00 -4.73211050e-01
1.55031955e+00 -8.72951567e-01 5.93037963e-01 1.69548437e-01
-9.06050980e-01 -2.58732706e-01 -9.44789886e-01 -1.29794300e-01
-2.91613430e-01 1.74492136e-01 1.14113116e+00 1.04610980e+00
-9.15552557e-01 2.63165057e-01 -6.19572341e-01 -2.49480069e-01
1.07855596e-01 7.01789081e-01 -9.31909084e-01 -5.89356050e-02
-1.57830930e+00 1.26993084e+00 5.20487964e-01 -2.33972311e-01
1.10591715e-02 -1.84598818e-01 -9.39101219e-01 -1.24912135e-01
-1.22335188e-01 -1.21415541e-01 1.62035763e+00 -1.05801404e+00
-1.38562584e+00 1.31838822e+00 -3.83211553e-01 -1.84609387e-02
-2.18758672e-01 -3.17946583e-01 -6.35717988e-01 1.09497361e-01
2.52819031e-01 3.43648791e-01 1.63254648e-01 -7.89687157e-01
-8.55946839e-01 -4.15427476e-01 1.30791381e-01 -5.44549618e-03
-3.29777449e-01 1.05763268e+00 -3.04567009e-01 -5.39034069e-01
-3.73405486e-01 -7.67893672e-01 -1.67766392e-01 -1.03272939e+00
-1.47523507e-01 -7.61430681e-01 3.00813913e-01 -8.74473870e-01
1.40556216e+00 -2.13930392e+00 4.70914580e-02 1.72345772e-01
-9.87039804e-02 4.79354560e-02 -2.66646117e-01 1.34372517e-01
-7.02476501e-01 4.17326272e-01 -9.03321281e-02 -5.25533199e-01
-1.45761624e-01 9.63051677e-01 1.20259970e-01 2.49025524e-01
5.63300788e-01 9.85568643e-01 -7.84625769e-01 -8.80591035e-01
-3.01481605e-01 2.42201135e-01 -1.82691887e-01 2.25573987e-01
1.43318713e-01 5.08466847e-02 -7.29607761e-01 8.65024209e-01
4.39796090e-01 -9.17441472e-02 8.98445606e-01 -1.04581274e-01
-7.19970465e-02 2.45048925e-01 -5.87319553e-01 1.59213412e+00
-7.78041184e-01 3.60749692e-01 2.06768185e-01 -9.46491301e-01
7.31134951e-01 5.93500018e-01 2.89641082e-01 -6.72669470e-01
3.24540704e-01 3.39562386e-01 -6.98964447e-02 -7.04326570e-01
3.18954170e-01 -5.57191670e-01 -7.81051993e-01 4.55301851e-01
4.35061663e-01 1.93491578e-01 2.66867191e-01 1.45723104e-01
1.46580303e+00 -1.25313792e-02 7.89802790e-01 -2.93874204e-01
6.07694507e-01 1.11066982e-01 7.32146502e-01 2.54647017e-01
-2.28320733e-01 -1.44124046e-01 7.56594360e-01 -3.55397552e-01
-7.39959419e-01 -5.84439754e-01 -2.08688095e-01 1.91543460e+00
-4.58542764e-01 -6.72446072e-01 -7.31393218e-01 -1.32603991e+00
-1.51820526e-01 9.70849931e-01 -5.63835919e-01 3.12150925e-01
-9.25943315e-01 -1.18258572e+00 1.08372343e+00 9.98083174e-01
-1.53237984e-01 -8.92296791e-01 -4.44793522e-01 3.36789459e-01
-2.14795440e-01 -1.40622163e+00 -4.15507071e-02 1.09723639e+00
-8.15590620e-01 -1.11931241e+00 -1.34102270e-01 -1.23415792e+00
5.64822376e-01 -3.43926400e-01 1.27332258e+00 1.62813932e-01
-2.14482933e-01 2.03112349e-01 -7.02990651e-01 -9.78618562e-02
-8.22226226e-01 1.87600151e-01 -2.69416153e-01 -6.31147146e-01
8.03434432e-01 -1.84805110e-01 4.40862983e-01 3.50749977e-02
-4.75153595e-01 -4.38097507e-01 7.16107488e-01 9.14609373e-01
2.48786122e-01 -4.78221565e-01 5.44650197e-01 -1.43016493e+00
8.01752925e-01 -4.56843168e-01 -3.68855268e-01 5.61563194e-01
-4.67722267e-01 5.85929990e-01 4.11248952e-01 -1.19909376e-01
-1.40503061e+00 5.35757840e-01 -7.89213955e-01 2.38313094e-01
-2.12569997e-01 6.83317304e-01 -1.45549774e-01 -1.29362315e-01
5.82654297e-01 -4.83436793e-01 -3.71547997e-01 -5.59064448e-01
3.78289461e-01 7.24801719e-01 2.44732976e-01 -7.69856453e-01
5.77668957e-02 -2.01534942e-01 -6.79313016e-05 -5.52035868e-01
-1.09418619e+00 -5.77788234e-01 -9.13238645e-01 4.53553386e-02
1.10125792e+00 -8.64499092e-01 -4.66871500e-01 1.16462506e-01
-1.58953428e+00 -4.92133573e-02 -6.77988008e-02 5.00485063e-01
-2.79330704e-02 6.59997165e-01 -8.82191718e-01 -8.10746253e-01
-3.61697823e-01 -8.69549632e-01 1.18474519e+00 -3.46669316e-01
-6.83293045e-01 -1.27079821e+00 2.62597740e-01 1.09266140e-01
-1.97811291e-01 1.63104609e-01 1.08428061e+00 -1.17581224e+00
6.23769343e-01 -7.35134900e-01 -1.52762249e-01 4.30832207e-01
2.22997889e-02 -9.03915018e-02 -1.03440154e+00 1.26578743e-02
-2.07018867e-01 -7.02652693e-01 6.62530303e-01 4.93648835e-02
6.49393559e-01 1.42760545e-01 -3.46863866e-01 1.44286692e-01
1.17336130e+00 1.53913170e-01 1.03440128e-01 3.13125141e-02
5.05957484e-01 8.98701191e-01 5.05154312e-01 -2.62938929e-03
3.04035515e-01 4.80838031e-01 -2.51338668e-02 3.74470912e-02
2.48882025e-02 1.92140415e-01 4.82804149e-01 1.06010532e+00
-1.13463745e-01 -2.77905792e-01 -9.24239397e-01 4.74661797e-01
-1.42394447e+00 -5.12672722e-01 -2.80264467e-01 1.38961518e+00
9.74112868e-01 9.83962268e-02 -7.55085126e-02 1.35257259e-01
8.52951050e-01 7.76735544e-02 1.15393132e-01 -9.99519050e-01
-4.49962109e-01 1.03407395e+00 8.23815882e-01 9.06255245e-01
-1.17489851e+00 1.24816501e+00 8.76547432e+00 6.11799121e-01
-4.98808920e-01 5.75823069e-01 4.28142190e-01 3.59669596e-01
-1.99613854e-01 -2.43728608e-02 -1.17878330e+00 -5.24171963e-02
1.55203390e+00 -3.33455652e-02 -3.65067959e-01 1.10821271e+00
-3.53103280e-01 -1.99147329e-01 -9.58437085e-01 5.35266519e-01
1.42735943e-01 -8.88523579e-01 -3.14123482e-01 -1.88467532e-01
2.11952031e-01 1.17997833e-01 -6.49802864e-01 4.25195336e-01
7.64768124e-01 -7.96251953e-01 3.00353348e-01 -2.32352421e-01
7.32796729e-01 -8.58271062e-01 1.14199162e+00 4.12832759e-02
-1.06409848e+00 -1.45992860e-01 -5.29616289e-02 -1.08389340e-01
3.35549504e-01 4.86574352e-01 -5.74906826e-01 5.50199568e-01
2.73424089e-01 3.79923433e-01 -5.78159392e-01 2.16251299e-01
-1.04701638e+00 1.01144218e+00 -2.50612885e-01 -2.83197492e-01
1.16341077e-01 4.01158571e-01 3.12932968e-01 1.93359518e+00
-1.25872925e-01 6.40317798e-01 3.46630812e-01 -3.26469749e-01
-1.34457350e-01 8.07999134e-01 -3.82691413e-01 4.59838361e-02
2.60197520e-01 1.48419619e+00 -7.07298279e-01 -4.26163465e-01
-7.69523501e-01 7.81918049e-01 8.33365977e-01 -2.91470550e-02
-5.62534332e-01 -4.32525426e-01 2.72896737e-01 -4.84822750e-01
3.10590059e-01 -1.42063066e-01 -1.58215150e-01 -1.00998485e+00
-1.79808125e-01 -5.48541963e-01 1.08037543e+00 -5.59453368e-01
-1.40428627e+00 8.76453459e-01 -6.65012188e-03 -4.74001169e-01
-5.89000463e-01 -1.24012673e+00 -4.90948975e-01 7.69600630e-01
-1.50981140e+00 -1.39070272e+00 1.92483395e-01 1.03377558e-01
2.36788332e-01 -1.50342956e-02 1.63005352e+00 4.55181420e-01
-9.26633298e-01 9.53803241e-01 -4.31223631e-01 5.31928539e-01
6.31170213e-01 -1.35340321e+00 3.51188093e-01 6.94743693e-01
-3.70298885e-02 4.88896430e-01 5.82466304e-01 -4.73389179e-01
-1.06056511e+00 -8.28013778e-01 1.71088314e+00 -4.46736604e-01
9.90661383e-01 -2.37710118e-01 -7.58137405e-01 8.99781346e-01
2.48138264e-01 1.58273369e-01 1.36564970e+00 6.89243376e-01
-5.14974773e-01 3.12297016e-01 -1.12230754e+00 -2.21563742e-01
8.31163466e-01 -6.23541415e-01 -9.22353923e-01 3.96022826e-01
6.16763353e-01 -3.70473236e-01 -1.37358224e+00 2.42556483e-01
7.40123272e-01 -3.65215689e-01 5.21964252e-01 -1.30187309e+00
3.92434537e-01 2.92383224e-01 -5.78687131e-01 -1.09814239e+00
-3.29557955e-01 -3.14984530e-01 1.33025721e-01 1.22104394e+00
9.52392578e-01 -5.17873943e-01 4.95065540e-01 1.00691438e+00
-2.58054435e-01 -3.52245569e-01 -1.24247277e+00 -7.28676558e-01
2.91415602e-01 -7.25920379e-01 2.26799935e-01 9.29885149e-01
7.58745968e-01 9.61115718e-01 -8.42594132e-02 -4.16024327e-02
3.71681392e-01 8.01741034e-02 1.92181140e-01 -1.04372489e+00
-4.01041925e-01 1.16024148e-02 -5.98492920e-01 -7.30370581e-01
1.10902214e+00 -1.00308156e+00 -1.69369020e-03 -1.27322507e+00
-3.38723068e-03 -2.46758372e-01 -4.60873723e-01 8.93011153e-01
-4.00826514e-01 3.45223099e-01 -2.26325750e-01 -1.42249316e-01
-8.17300022e-01 -1.12743273e-01 5.22949219e-01 2.63179809e-01
2.25052938e-01 -3.55763406e-01 -7.57404983e-01 8.00382495e-01
6.67489052e-01 -7.57426381e-01 1.70372590e-01 -5.20698428e-01
3.30746740e-01 1.89017102e-01 -2.71186411e-01 -3.70403260e-01
-2.11633146e-01 -7.71342590e-03 2.66433090e-01 -8.07070807e-02
-1.95627231e-02 -4.83028978e-01 -2.29513064e-01 3.63966376e-01
-4.93586481e-01 4.24831599e-01 4.93126273e-01 1.89417869e-01
-2.75559038e-01 -5.80564618e-01 7.62769103e-01 -1.79392189e-01
-1.13224804e+00 -3.79666269e-01 -6.17893636e-01 2.40208939e-01
1.17021334e+00 2.80150145e-01 -3.91062528e-01 1.12350799e-01
-1.11016583e+00 1.99404016e-01 -8.36339965e-02 1.73130199e-01
7.34529644e-02 -1.12666786e+00 -4.75860387e-01 4.17195354e-03
3.61562699e-01 -6.36091828e-01 -3.40122432e-01 5.73628187e-01
-4.05175656e-01 5.53052604e-01 6.85852673e-03 -1.64360881e-01
-1.86688876e+00 3.14016759e-01 9.37371999e-02 -7.24989891e-01
-1.23756662e-01 1.16795743e+00 -3.28845978e-01 -3.09785724e-01
-1.39779106e-01 -4.42045540e-01 -4.84938681e-01 2.71148890e-01
4.19668585e-01 -4.13463153e-02 1.17954008e-01 -9.79543924e-01
-6.81581020e-01 5.41645348e-01 -7.75663778e-02 -1.26636744e-01
1.23571634e+00 1.32894799e-01 -3.09835464e-01 3.43557566e-01
1.37248766e+00 2.34659463e-01 -2.73034066e-01 -6.73494041e-02
6.33270741e-01 -2.38331798e-02 1.94339469e-01 -9.58574951e-01
-7.50794113e-01 6.53985023e-01 1.38276890e-01 1.13602005e-01
1.06405151e+00 3.28779757e-01 6.79814160e-01 6.14559591e-01
2.85259783e-01 -1.27659857e+00 -3.78167927e-01 6.61518514e-01
2.79638469e-01 -1.10087597e+00 1.19608358e-01 -9.09512639e-01
-7.80152500e-01 1.17904603e+00 4.69867140e-01 -6.56187311e-02
7.17556357e-01 8.76271904e-01 4.53529209e-01 -3.68665159e-01
-8.86974275e-01 -2.18318969e-01 -1.26402035e-01 4.33189034e-01
1.08276272e+00 1.97499350e-01 -8.32233310e-01 1.08178985e+00
-3.56790200e-02 -2.33537316e-01 6.87299818e-02 1.29760647e+00
-4.79398906e-01 -1.61439085e+00 1.76355448e-02 3.23927045e-01
-1.19911444e+00 -2.32790247e-01 -4.89222735e-01 9.39282417e-01
-9.69521776e-02 9.57404315e-01 1.37160784e-02 -2.67602772e-01
5.90804815e-01 8.21783662e-01 6.09077096e-01 -7.95354903e-01
-8.67945373e-01 -2.86478810e-02 1.24619079e+00 -4.79447842e-01
-1.09426069e+00 -7.64173090e-01 -1.66653359e+00 -4.12565805e-02
-6.99327409e-01 5.29542506e-01 7.23604381e-01 1.31873238e+00
-1.55169377e-02 4.02888596e-01 6.95445240e-01 -1.88031614e-01
-4.52792555e-01 -9.98602688e-01 -5.31101286e-01 3.56699497e-01
-6.47134483e-02 -6.67759776e-01 -2.58539081e-01 1.00980140e-01] | [10.333008766174316, 9.77993106842041] |
bdaded89-1d9c-4b9f-97a9-1132e76382a8 | rethinking-the-defocus-blur-detection-problem | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1182_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123550613.pdf | Rethinking the Defocus Blur Detection Problem and A Real-Time Deep DBD Model | Defocus blur detection (DBD) is a classical low level vision task. It has recently attracted attention focusing on designing complex convolutional neural networks (CNN) which make full use of both low level features and high level semantic information. The heavy networks used in these methods lead to low processing speed, resulting difficulty in applying to real-time applications. In this work, we propose novel perspectives on the DBD problem and design convenient approach to build a real-time cost-effective DBD model. First, we observe that the semantic information does not always relate to and sometimes mislead the blur detection. We start from the essential characteristics of the DBD problem and propose a data augmentation method accordingly to destroy the semantic information and enforce the model to learn image blur related features rather than the semantic features. A novel self-supervision training objective is proposed to enhance the model training consistency and stability. Second, by rethinking the relationship between defocus blur detection and salience detection, we identify two previously ignored but common scenarios, based on which we design a hard mining strategy to enhance the DBD model. By using the proposed techniques, our model that uses a slightly modified U-Net as backbone, improves the processing speed by more than 3 times and performs competitively against state of the art methods. Ablation study is also conducted to verify the effectiveness of each part of our proposed methods. | ['Junchi Yan', 'Ning Zhang'] | null | null | null | null | eccv-2020-8 | ['defocus-blur-detection'] | ['computer-vision'] | [ 4.14278388e-01 -8.51627067e-02 3.83061878e-02 -3.41915488e-01
2.91389555e-01 8.15282390e-03 5.19833028e-01 -6.64879307e-02
-4.04513478e-01 7.10607648e-01 1.56838268e-01 -2.14885905e-01
-4.37834084e-01 -5.23210943e-01 -5.70712149e-01 -6.53564811e-01
9.94455889e-02 -2.15605900e-01 4.92148936e-01 -1.08272247e-01
4.37152177e-01 4.98190582e-01 -1.95763063e+00 7.43565410e-02
1.22169232e+00 1.04707003e+00 7.33756542e-01 3.42852265e-01
1.11303195e-01 9.12695527e-01 -6.07512236e-01 -1.30314089e-03
4.52895939e-01 -3.32468808e-01 -5.76111734e-01 2.09546283e-01
5.25880754e-01 -4.20189798e-01 -3.73539299e-01 1.27233219e+00
6.25609040e-01 5.78666851e-02 4.92087573e-01 -1.15343702e+00
-5.85262775e-01 3.89796436e-01 -8.04204047e-01 5.76552749e-01
2.09749416e-02 1.86657235e-01 5.28168380e-01 -6.28531992e-01
3.31701189e-01 1.02915478e+00 6.90834761e-01 4.84654069e-01
-8.18324327e-01 -5.21834731e-01 1.06868379e-01 5.79719722e-01
-1.08285022e+00 -3.94855201e-01 9.70412850e-01 -3.79942209e-01
8.36968839e-01 1.41022697e-01 3.94311041e-01 1.01076245e+00
1.32780790e-01 5.41167140e-01 1.40576077e+00 -6.10359550e-01
1.22939028e-01 2.86317587e-01 4.74897861e-01 8.52671921e-01
5.21496356e-01 2.11641088e-01 -4.61088300e-01 3.15687150e-01
7.02337205e-01 8.67834017e-02 -5.88492393e-01 -2.30660036e-01
-9.47116017e-01 4.67999786e-01 5.47512472e-01 4.15621221e-01
-2.36878708e-01 -1.84157118e-01 2.03181714e-01 2.72194982e-01
2.31508821e-01 5.34851491e-01 -4.03277129e-01 -1.58042014e-01
-8.33194792e-01 7.04400092e-02 4.88291711e-01 8.54113102e-01
8.86027932e-01 -8.54222178e-02 -3.69872600e-01 8.00996542e-01
7.71648735e-02 1.86867625e-01 6.30967438e-01 -6.79965377e-01
1.36641279e-01 7.73204684e-01 1.43801615e-01 -1.31584430e+00
-4.81985718e-01 -8.08625638e-01 -1.16682291e+00 4.37044382e-01
2.25169271e-01 -9.96313244e-02 -1.07532287e+00 1.69936907e+00
1.83698207e-01 3.98976415e-01 1.98470452e-03 1.26217568e+00
6.40266418e-01 3.99668038e-01 -2.61481285e-01 -4.53561008e-01
1.27516389e+00 -1.17112088e+00 -8.97019267e-01 -9.86527950e-02
2.15884209e-01 -9.20478284e-01 1.03912652e+00 5.63987017e-01
-9.54071403e-01 -1.00111604e+00 -1.37581313e+00 -2.58174501e-02
-3.61099571e-01 2.53187776e-01 6.53566599e-01 5.47956884e-01
-9.37783778e-01 6.25430644e-01 -5.32575011e-01 -3.89351785e-01
4.50767100e-01 3.82736713e-01 -5.35273217e-02 -1.64958030e-01
-1.23644948e+00 1.08637166e+00 5.09401143e-01 3.79451364e-01
-8.37581873e-01 -5.18019855e-01 -5.16253412e-01 2.09726058e-02
4.93493021e-01 -6.90636277e-01 1.08387733e+00 -1.06554389e+00
-1.29695809e+00 5.91503680e-01 -1.02961540e-01 -6.71132326e-01
5.97653508e-01 -3.64285827e-01 -4.55420345e-01 1.31934479e-01
-1.06550060e-01 5.33201694e-01 1.17478967e+00 -1.27127373e+00
-8.51429403e-01 -2.21208215e-01 2.47168645e-01 2.10252047e-01
-5.86085439e-01 -2.74027020e-01 -3.16807508e-01 -6.24359488e-01
5.96306175e-02 -5.96826494e-01 -3.73796262e-02 -7.13875815e-02
-2.89237678e-01 -1.96856335e-01 1.09036756e+00 -4.10223275e-01
1.39087486e+00 -2.19083738e+00 -1.43049806e-02 8.29201341e-02
5.22787571e-01 6.82618737e-01 1.57156765e-01 1.19864587e-02
-9.78244022e-02 -2.45978624e-01 -3.63073081e-01 -3.15408707e-01
-2.02859148e-01 2.79199779e-01 -3.10448091e-02 3.82591188e-01
2.99749553e-01 4.86434400e-01 -7.42785215e-01 -3.86508197e-01
3.91637117e-01 3.16898942e-01 -4.74645466e-01 4.54878896e-01
-1.29404947e-01 1.30826980e-01 -2.64219880e-01 4.90208864e-01
1.05603468e+00 -2.66195744e-01 -2.88646549e-01 -8.05851877e-01
-4.31056052e-01 -1.22125760e-01 -1.25360215e+00 1.62154770e+00
-3.48018169e-01 5.59688210e-01 7.61955753e-02 -1.27243412e+00
1.01185822e+00 -5.33155836e-02 2.05489784e-01 -6.69590354e-01
3.37819308e-01 1.60918400e-01 1.37568623e-01 -9.64493752e-01
3.18585545e-01 -1.52940350e-03 4.63787138e-01 2.54055321e-01
-5.14602363e-02 1.39327675e-01 6.21819682e-03 -6.86656535e-02
9.05434728e-01 8.64722207e-02 2.61727124e-01 -5.15412569e-01
6.99627459e-01 -5.66412043e-03 5.44522941e-01 8.57917011e-01
-4.23452586e-01 5.14750779e-01 2.14565516e-01 -4.53877747e-01
-8.70546877e-01 -5.19946575e-01 -2.43690357e-01 5.44851065e-01
5.91345072e-01 -6.62594140e-02 -8.62097025e-01 -6.06658518e-01
-1.02752661e-02 4.94545519e-01 -7.97905862e-01 -4.50920224e-01
-2.86628515e-01 -1.02014410e+00 2.71864891e-01 1.72472909e-01
1.17887008e+00 -8.01763237e-01 -8.84960294e-01 9.38599464e-03
8.46252888e-02 -1.19426632e+00 -3.04547489e-01 3.86862129e-01
-7.59113729e-01 -1.12810338e+00 -5.72626352e-01 -9.13517356e-01
7.02966332e-01 7.75865734e-01 7.04215586e-01 1.86573923e-01
-4.12838280e-01 -1.06081367e-01 -3.79629850e-01 -5.11336148e-01
-3.05413544e-01 3.15448865e-02 2.39170521e-01 2.19481379e-01
5.16312599e-01 -6.93575442e-01 -8.30038607e-01 1.98552206e-01
-9.81455803e-01 4.40103471e-01 9.59415913e-01 8.60315204e-01
1.01477444e-01 4.20526773e-01 6.25971079e-01 -8.03147674e-01
7.33658135e-01 -2.67895967e-01 -5.12811244e-01 1.25302091e-01
-1.09354234e+00 1.54819697e-01 5.39629579e-01 -3.57533693e-01
-1.22291648e+00 -9.96532291e-02 1.04916938e-01 -8.20346713e-01
-3.21953446e-01 3.72368336e-01 2.12440658e-02 -2.76765406e-01
7.43507147e-01 2.33665824e-01 2.75513798e-01 -6.20145380e-01
1.13049321e-01 9.34436321e-01 5.87960243e-01 -2.22250342e-01
8.68316650e-01 3.58508945e-01 9.86194164e-02 -7.18749583e-01
-1.15696812e+00 -3.94183278e-01 -6.09381199e-01 -1.99583903e-01
8.68102074e-01 -8.50947082e-01 -7.47597277e-01 7.73549139e-01
-1.17201650e+00 8.30299258e-02 7.33729899e-02 4.64455605e-01
-2.10737124e-01 6.33350253e-01 -3.60584974e-01 -8.21307659e-01
-2.00944319e-01 -1.08369410e+00 6.01945698e-01 5.12259483e-01
2.59208143e-01 -7.25905597e-01 -2.48759583e-01 4.54751581e-01
5.68979442e-01 2.20543906e-01 8.78354073e-01 -3.02763134e-01
-7.07542419e-01 3.34877551e-01 -7.84390092e-01 7.75980532e-01
5.38962901e-01 -2.22589120e-01 -1.22000480e+00 -1.94226488e-01
5.04314959e-01 -1.82603613e-01 1.06723583e+00 3.67816508e-01
1.40247703e+00 -2.00236112e-01 -1.25702962e-01 6.91575289e-01
1.59173381e+00 2.15097636e-01 5.58352351e-01 4.95393842e-01
6.24113500e-01 5.12224615e-01 6.28738701e-01 4.98410106e-01
9.12787244e-02 5.06474733e-01 7.29113877e-01 -4.59040791e-01
-3.69634360e-01 -1.24236479e-01 9.12509859e-02 7.05986500e-01
-1.97695836e-01 1.48549378e-01 -6.03085756e-01 4.93623465e-01
-2.03146219e+00 -8.77666414e-01 -1.53951094e-01 2.01997685e+00
1.02777100e+00 5.30448258e-01 -1.27093092e-01 3.03356081e-01
7.53633380e-01 6.48348555e-02 -4.17326242e-01 -1.73685461e-01
-2.17674062e-01 1.50014162e-01 4.38135028e-01 4.50020015e-01
-1.09059143e+00 7.81062722e-01 5.46026468e+00 7.56379187e-01
-1.28378069e+00 6.20469190e-02 6.18244350e-01 9.46864039e-02
-9.71320923e-03 -1.96191788e-01 -6.88938022e-01 6.49783194e-01
3.46733570e-01 4.93382011e-03 5.04373133e-01 6.73070908e-01
6.46370232e-01 -3.34675759e-01 -1.00997806e+00 1.23326802e+00
2.84107924e-01 -1.24737036e+00 7.74117857e-02 -2.25790441e-01
7.12453485e-01 -3.05535614e-01 -1.14293814e-01 -8.83906987e-03
-1.53343946e-01 -1.01383555e+00 5.47105372e-01 6.85666203e-01
4.93305475e-01 -6.31613970e-01 9.91251945e-01 4.79699641e-01
-7.31107771e-01 -2.73123592e-01 -5.63677907e-01 -4.67707336e-01
-2.70374507e-01 9.55218792e-01 -8.34109306e-01 7.29299963e-01
8.46483827e-01 8.87618959e-01 -7.59559333e-01 1.24514008e+00
-2.31429599e-02 4.87401515e-01 -5.63079901e-02 -3.12847234e-02
4.15958315e-01 -1.29933953e-01 5.09083569e-01 1.20659971e+00
2.48923078e-01 2.79551856e-02 -7.63423368e-02 1.03211617e+00
1.37871161e-01 -2.59030342e-01 -6.16841435e-01 3.86008263e-01
2.86149740e-01 1.17971194e+00 -4.17198002e-01 -3.05067420e-01
-4.49515045e-01 9.19355273e-01 2.46536314e-01 2.53002405e-01
-9.71965671e-01 -4.45919037e-01 5.74748337e-01 1.20017501e-02
2.96974123e-01 -1.27154514e-01 -4.24339861e-01 -1.15291059e+00
1.60113588e-01 -7.52902687e-01 7.34960660e-02 -7.93304026e-01
-1.27402449e+00 6.32259607e-01 -1.22609794e-01 -1.28701758e+00
2.89209843e-01 -6.87212348e-01 -5.37857771e-01 9.41414714e-01
-2.20950174e+00 -8.84333193e-01 -9.82953787e-01 8.13305318e-01
7.34494328e-01 -2.57930189e-01 4.15616900e-01 4.16943818e-01
-7.73869514e-01 3.49574059e-01 -1.11819655e-01 -2.63540119e-01
8.05175424e-01 -1.14800584e+00 -2.12147936e-01 1.08262444e+00
-2.87030041e-01 5.86664736e-01 1.01165354e+00 -4.77743536e-01
-1.20455778e+00 -1.07096732e+00 7.06076920e-01 4.33001593e-02
4.10892516e-01 -2.47693136e-01 -8.60095739e-01 4.19641621e-02
4.67901349e-01 -1.66374400e-01 7.65275285e-02 -5.24648577e-02
-1.12868667e-01 -5.96725821e-01 -1.10190988e+00 4.81503367e-01
1.15997672e+00 -2.87559539e-01 -1.02650356e+00 2.06956685e-01
8.43216419e-01 -2.20322102e-01 -3.57080519e-01 6.25960350e-01
1.69848904e-01 -1.45726991e+00 7.29213774e-01 -3.69441360e-01
5.85723996e-01 -5.68002582e-01 1.35753393e-01 -1.15837920e+00
-4.98650461e-01 -6.47541344e-01 -3.19585860e-01 1.06804800e+00
2.46781129e-02 -5.09941757e-01 6.42469049e-01 1.65023357e-01
-3.90135497e-01 -7.78576076e-01 -6.90212727e-01 -6.76652610e-01
-5.36969364e-01 -1.51844084e-01 3.97624522e-01 1.05281126e+00
-2.46013492e-01 3.68336886e-01 -6.01913273e-01 3.32789272e-01
7.63322711e-01 1.83223754e-01 4.63247806e-01 -1.20887089e+00
-2.37222806e-01 -4.46573466e-01 -3.03996772e-01 -1.20265710e+00
-2.94853836e-01 -4.31225777e-01 6.34372979e-02 -1.30279231e+00
3.02686930e-01 -5.20484030e-01 -6.30906224e-01 3.74796569e-01
-1.99568436e-01 1.63582951e-01 2.96591464e-02 3.16875011e-01
-3.45213175e-01 5.36743402e-01 1.50973988e+00 -7.92374983e-02
-1.45498201e-01 1.95142522e-01 -8.89110744e-01 6.47646010e-01
6.54415905e-01 -8.39803368e-02 -8.03358734e-01 -5.46500742e-01
-3.79927792e-02 -2.59966016e-01 5.32812119e-01 -1.36524940e+00
4.20013100e-01 -2.00070247e-01 3.95724475e-01 -4.59005207e-01
9.59443226e-02 -1.14979148e+00 -2.38394916e-01 5.80433190e-01
-1.38273418e-01 -1.09854579e-01 1.66032508e-01 5.76482654e-01
-3.24041158e-01 -4.97054219e-01 1.11503184e+00 -2.57528037e-01
-9.63434458e-01 4.98791002e-02 -7.81358182e-02 -2.88645983e-01
1.06955791e+00 -4.58869904e-01 -4.77835089e-01 -1.07744247e-01
-4.73939300e-01 1.25054836e-01 2.59274364e-01 5.36171496e-01
5.93003452e-01 -9.94936943e-01 -3.44562531e-01 3.80707324e-01
1.68637298e-02 6.52506724e-02 2.51760811e-01 9.73771870e-01
-5.06485343e-01 3.48400235e-01 -5.54840326e-01 -6.06451154e-01
-1.15578580e+00 7.86117852e-01 4.58944559e-01 1.32785514e-01
-5.44724762e-01 7.58904338e-01 1.50418103e-01 1.93091661e-01
5.82572162e-01 -5.76417089e-01 -4.60019499e-01 -1.02125227e-01
4.26032662e-01 2.76841015e-01 1.69572711e-01 -6.95635080e-02
-2.62968659e-01 4.75539982e-01 -9.59493071e-02 4.15171742e-01
1.42156243e+00 -3.49693209e-01 -2.05692068e-01 1.43676117e-01
9.61885870e-01 -1.71061561e-01 -1.40102220e+00 -2.15556979e-01
-7.01904893e-02 -6.43274605e-01 3.27447057e-01 -1.04010594e+00
-1.04777217e+00 8.76832724e-01 9.79470015e-01 3.20447832e-01
1.64689243e+00 -3.96756321e-01 6.43498242e-01 3.21022332e-01
1.40563503e-01 -1.11275578e+00 1.62117973e-01 3.13403845e-01
6.98512495e-01 -1.40016472e+00 -8.00600201e-02 -5.33787906e-01
-4.14471358e-01 1.23539400e+00 1.07437932e+00 -1.01876810e-01
6.99123800e-01 2.64317662e-01 -5.20687811e-02 -7.85092413e-02
-6.51682258e-01 -2.26009592e-01 1.91270977e-01 5.35723627e-01
-1.24960234e-02 -4.41659480e-01 -4.03850198e-01 5.72321355e-01
1.33788818e-02 2.41396055e-01 6.40225708e-01 7.57393837e-01
-8.91939878e-01 -9.18689370e-01 -2.62599140e-01 4.46678072e-01
-1.31966248e-01 -1.28793895e-01 -1.31939754e-01 7.07031906e-01
7.31664777e-01 8.59283268e-01 -2.54404806e-02 -5.59783995e-01
2.97169864e-01 -2.16227755e-01 3.98677647e-01 -2.51652509e-01
-3.08200836e-01 -4.93498407e-02 -6.14173822e-02 -3.50025207e-01
-8.14549267e-01 -3.67421001e-01 -7.72471309e-01 -1.86160967e-01
-3.35320860e-01 3.01307887e-02 4.38168675e-01 1.02275074e+00
4.32354867e-01 6.76057339e-01 7.86534190e-01 -5.41800499e-01
-7.30713308e-01 -1.08282781e+00 -4.89007831e-01 5.40476561e-01
6.41364157e-01 -9.17751908e-01 -5.05465746e-01 1.06648900e-01] | [11.238838195800781, -2.6792984008789062] |
9ad31758-6d56-47c8-8cec-af9a596f039e | projection-inpainting-using-partial | 2005.00762 | null | https://arxiv.org/abs/2005.00762v1 | https://arxiv.org/pdf/2005.00762v1.pdf | Projection Inpainting Using Partial Convolution for Metal Artifact Reduction | In computer tomography, due to the presence of metal implants in the patient body, reconstructed images will suffer from metal artifacts. In order to reduce metal artifacts, metals are typically removed in projection images. Therefore, the metal corrupted projection areas need to be inpainted. For deep learning inpainting methods, convolutional neural networks (CNNs) are widely used, for example, the U-Net. However, such CNNs use convolutional filter responses on both valid and corrupted pixel values, resulting in unsatisfactory image quality. In this work, partial convolution is applied for projection inpainting, which only relies on valid pixels values. The U-Net with partial convolution and conventional convolution are compared for metal artifact reduction. Our experiments demonstrate that the U-Net with partial convolution is able to inpaint the metal corrupted areas better than that with conventional convolution. | ['Yixing Huang', 'Lin Yuan', 'Andreas Maier'] | 2020-05-02 | null | null | null | null | ['metal-artifact-reduction'] | ['medical'] | [ 3.95427942e-01 1.87687695e-01 2.65688062e-01 -1.58532307e-01
-3.76255423e-01 1.38984606e-01 -7.69857736e-03 -1.78229198e-01
-3.84422481e-01 8.60682309e-01 2.62205362e-01 -1.16419271e-02
2.47672707e-01 -1.04829907e+00 -8.79400909e-01 -6.32851481e-01
6.34619415e-01 -2.26015057e-02 3.44679326e-01 -1.22969300e-02
-1.93719968e-01 4.95866150e-01 -9.00670588e-01 6.56365335e-01
8.76390994e-01 1.19736350e+00 3.34188074e-01 -1.10130861e-01
-2.73495764e-01 7.90637970e-01 -4.27229464e-01 -2.28403121e-01
5.71349025e-01 -5.79539299e-01 -4.06025082e-01 1.55906245e-01
2.33496279e-01 -7.25070775e-01 -7.81601846e-01 1.29919732e+00
3.85643184e-01 -1.85900271e-01 4.79352117e-01 -6.94154501e-01
-5.57936192e-01 7.52891839e-01 -5.76697946e-01 -1.75843239e-01
5.36136292e-02 1.39546022e-01 7.72988945e-02 -8.60600650e-01
5.97032309e-01 9.93384063e-01 9.54033494e-01 7.74338186e-01
-1.28743947e+00 -7.92202115e-01 -3.95756453e-01 -5.49069680e-02
-1.01228762e+00 -3.04629430e-02 1.16709840e+00 -3.93236727e-01
5.42851031e-01 3.76833349e-01 7.63927639e-01 7.35778391e-01
7.06053972e-01 6.14958227e-01 1.15267372e+00 -2.98602551e-01
9.78253782e-03 9.93667468e-02 -2.74956763e-01 6.39291167e-01
3.10045809e-01 2.12333113e-01 -1.87955141e-01 -5.67603633e-02
1.31381476e+00 7.38196254e-01 -8.53432953e-01 -1.17949389e-01
-1.05996048e+00 5.43669522e-01 9.59982634e-01 3.97761583e-01
-5.76018929e-01 3.19397509e-01 3.32808435e-01 2.70790190e-01
4.01235461e-01 4.81402457e-01 -9.28551108e-02 3.48960161e-01
-1.06080353e+00 9.90524739e-02 1.58110783e-01 5.80088198e-01
6.33083403e-01 1.45812348e-01 -3.86943847e-01 9.67411280e-01
5.36996759e-02 1.41184598e-01 7.50465870e-01 -7.29667068e-01
3.36338907e-01 7.30837762e-01 1.58204269e-02 -8.13090086e-01
-4.60222900e-01 -2.76759505e-01 -1.31792819e+00 7.14748502e-01
2.69564509e-01 -7.46583343e-02 -1.36751866e+00 1.36358714e+00
1.14146702e-01 1.10189490e-01 -1.40918717e-02 1.06362319e+00
9.27588820e-01 5.23636281e-01 -2.03222290e-01 -2.39633814e-01
1.19125688e+00 -9.12067413e-01 -1.02699947e+00 -2.52660453e-01
3.93465430e-01 -9.58033741e-01 1.10333502e+00 4.17497486e-01
-1.33730090e+00 -5.12480736e-01 -1.21483123e+00 -1.36913285e-01
2.18931779e-01 -1.19476892e-01 4.42105770e-01 4.16123658e-01
-6.86835170e-01 1.16241133e+00 -9.94755030e-01 1.96638346e-01
7.90449560e-01 2.73690492e-01 -4.25620288e-01 -3.39542001e-01
-1.14381731e+00 8.40196550e-01 -3.11337840e-02 2.35058933e-01
-5.52430630e-01 -8.13551903e-01 -7.50857055e-01 7.26997852e-02
1.04433790e-01 -5.82640052e-01 1.23294032e+00 -1.13809252e+00
-1.63771963e+00 4.13402408e-01 4.76817369e-01 -3.72290462e-01
1.00779974e+00 -3.00685257e-01 -1.71207160e-01 2.08847612e-01
-1.25512958e-01 4.29517478e-01 8.56734455e-01 -1.17643380e+00
8.50360766e-02 -2.82047749e-01 -2.81661779e-01 1.29501969e-02
-3.03986847e-01 -1.76846962e-02 -4.33243334e-01 -1.15487599e+00
7.63331234e-01 -7.61460781e-01 -5.03772259e-01 7.85548210e-01
-3.64192396e-01 4.45931584e-01 9.57270622e-01 -1.00794995e+00
8.99755955e-01 -2.10195804e+00 -3.12940001e-01 6.41101524e-02
2.24883258e-01 5.22989929e-02 1.25312358e-01 -6.31368980e-02
-2.74308175e-01 -2.69656718e-01 -7.49520957e-01 -3.86163026e-01
-7.26207852e-01 3.31204832e-01 1.50012210e-01 4.85238403e-01
1.23943342e-02 7.86575675e-01 -5.19473791e-01 -4.65235502e-01
3.84375602e-01 6.63820446e-01 -4.54929769e-01 1.11058667e-01
-1.88158289e-01 6.70602500e-01 -2.61196911e-01 5.87380052e-01
1.11886489e+00 -1.58629082e-02 -7.26871043e-02 -5.22194803e-01
-3.49304937e-02 1.35314781e-02 -8.24315608e-01 1.89931667e+00
-6.88291669e-01 5.11200488e-01 2.75769144e-01 -5.55887103e-01
7.14369953e-01 5.78660369e-01 8.34850669e-01 -8.99231672e-01
3.63110036e-01 5.41631520e-01 7.96509758e-02 -4.08655584e-01
1.79758936e-01 -6.05216563e-01 4.68168855e-01 3.22338104e-01
-4.82619882e-01 -2.61243820e-01 -3.59455079e-01 -1.85180813e-01
1.09214997e+00 -2.36866958e-02 -3.04099292e-01 -7.57064372e-02
2.08354354e-01 -1.45808175e-01 6.89402580e-01 2.68191993e-01
1.59756914e-01 1.43531680e+00 2.29691207e-01 -8.02972496e-01
-1.08795583e+00 -7.73944676e-01 -3.58118832e-01 -1.81372091e-01
2.57457465e-01 5.31093739e-02 -8.41545939e-01 -5.69198668e-01
-2.54322261e-01 2.52578050e-01 -5.49151361e-01 -4.33924049e-01
-9.86215472e-01 -6.04404271e-01 2.13305727e-01 7.79012322e-01
7.48910904e-01 -1.48957217e+00 -7.98606396e-01 6.61015391e-01
-1.66818306e-01 -7.29879498e-01 -5.66095054e-01 3.33602309e-01
-1.40860546e+00 -9.78264809e-01 -1.25029147e+00 -6.98533118e-01
9.90812421e-01 1.98028073e-01 8.04544628e-01 3.76841992e-01
-3.43480945e-01 -5.29514730e-01 -1.70829624e-01 -1.59679770e-01
-5.59867799e-01 -4.81490999e-01 -3.01661462e-01 -1.01139374e-01
-2.16351509e-01 -7.41344392e-01 -1.00355577e+00 1.86639547e-01
-1.37521482e+00 5.30809760e-01 8.84030223e-01 1.06346262e+00
5.94932139e-01 8.59329104e-02 -2.39959341e-02 -9.85098720e-01
5.12967825e-01 -3.17137339e-03 -5.40925741e-01 7.87275136e-02
-4.99379694e-01 1.02368824e-01 7.63590932e-01 -5.60080707e-01
-1.10360301e+00 3.12406927e-01 -3.97741407e-01 -7.93216407e-01
1.18713513e-01 3.66853923e-01 -1.51121736e-01 -3.49899620e-01
6.32141709e-01 1.85182542e-01 2.17563912e-01 -6.05700552e-01
-2.61828899e-01 3.22578341e-01 4.69365895e-01 -1.08704671e-01
4.58317131e-01 5.70045888e-01 1.41135722e-01 -5.57961583e-01
-3.23936105e-01 1.87301099e-01 -2.70348608e-01 -3.02617341e-01
8.18821907e-01 -6.62030578e-01 -1.49480805e-01 6.66649520e-01
-1.41979825e+00 -3.87897730e-01 -4.60664600e-01 8.01505923e-01
-3.25173974e-01 5.28915942e-01 -1.25565171e+00 -2.76556551e-01
-6.84121192e-01 -1.49116468e+00 6.98009849e-01 3.22603643e-01
-7.26784300e-03 -3.45153928e-01 -1.11448914e-01 8.47009569e-02
4.26565617e-01 2.93214530e-01 1.08025432e+00 2.23655090e-01
-4.75760579e-01 -3.18543881e-01 -2.59879053e-01 6.58151865e-01
3.70217472e-01 -4.51750726e-01 -8.90825212e-01 -6.15293756e-02
4.42620158e-01 -9.37712938e-03 9.67867434e-01 6.69011772e-01
1.59888721e+00 -2.45525032e-01 -2.30529830e-01 8.40193927e-01
1.47328746e+00 4.32578743e-01 1.22355580e+00 3.19868684e-01
6.31994188e-01 3.65975887e-01 1.75749004e-01 2.88998365e-01
-4.65443164e-01 5.28510392e-01 5.50096452e-01 -4.63911980e-01
-5.09409606e-01 -7.22346604e-02 5.14599346e-02 4.69316900e-01
-9.81423189e-04 1.20529063e-01 -6.12660468e-01 3.11794668e-01
-1.65909886e+00 -4.84186023e-01 -4.31779981e-01 2.10621166e+00
9.46224689e-01 7.87492245e-02 -4.63484824e-01 3.20731342e-01
6.22417092e-01 -1.46140888e-01 -4.98990715e-01 -3.36641297e-02
5.68645857e-02 6.04566693e-01 6.35626674e-01 3.81870806e-01
-7.31372297e-01 3.69955301e-01 5.57566071e+00 6.72069371e-01
-1.39940250e+00 4.40124154e-01 6.26429975e-01 -3.79230864e-02
-2.58511513e-01 -3.10343325e-01 7.61854574e-02 6.87082589e-01
2.01745689e-01 6.23835504e-01 -3.08858720e-03 7.05494106e-01
2.01103941e-01 -3.02860677e-01 -7.94619679e-01 1.17522371e+00
-2.09664136e-01 -1.36770844e+00 -4.83684801e-02 -7.07729459e-02
9.27378595e-01 -3.68396521e-01 -3.92581429e-03 -1.41884133e-01
-1.25894904e-01 -1.00867569e+00 6.50332749e-01 5.02602994e-01
7.80210257e-01 -8.46544564e-01 1.02385008e+00 2.06606966e-02
-6.07874691e-01 1.83519065e-01 -6.20088637e-01 8.80979225e-02
4.11471784e-01 1.17819607e+00 -5.11761248e-01 4.65298563e-01
7.82476902e-01 2.76055813e-01 -1.25636861e-01 1.29143488e+00
-1.96879983e-01 1.96907178e-01 -3.19948971e-01 4.29562867e-01
2.00289134e-02 -3.43513966e-01 1.01589061e-01 7.70118475e-01
5.76762974e-01 3.73586178e-01 9.73336585e-03 1.16979039e+00
-4.37954068e-01 5.59921609e-03 -4.42927271e-01 4.12987739e-01
-1.52351111e-01 1.07249427e+00 -8.99931192e-01 -4.13568318e-01
-5.36274612e-01 1.37688756e+00 -2.03458726e-01 1.58329174e-01
-7.23214090e-01 -2.49773502e-01 5.10791779e-01 6.47042632e-01
-6.20472431e-02 3.61264460e-02 -8.28895330e-01 -8.11699748e-01
1.42823145e-01 -5.20044565e-01 -4.39886190e-02 -8.65800977e-01
-1.13169944e+00 9.85383868e-01 -4.63603914e-01 -1.48176050e+00
4.75409627e-01 -4.52780008e-01 -8.36272657e-01 9.48179364e-01
-1.22871399e+00 -8.80962133e-01 -6.19983017e-01 6.64473593e-01
5.24508834e-01 3.32608700e-01 3.78317147e-01 7.02230990e-01
-2.60847300e-01 4.11034644e-01 3.12004033e-02 1.02046534e-01
7.28828728e-01 -7.69704223e-01 3.26990597e-02 6.23318315e-01
-3.83547574e-01 2.91280270e-01 5.71384609e-01 -1.08435321e+00
-9.97398078e-01 -1.16344762e+00 2.82772094e-01 3.11784744e-01
2.85628270e-02 3.88941206e-02 -1.08885825e+00 5.11211216e-01
2.20094159e-01 4.14596677e-01 2.08247051e-01 -8.07494760e-01
3.94197740e-02 -8.03245381e-02 -1.58109832e+00 6.09784961e-01
5.20178795e-01 -2.53848493e-01 -3.75164062e-01 3.53760213e-01
6.07311368e-01 -5.68191290e-01 -7.25157022e-01 6.32957339e-01
5.84341705e-01 -1.11827445e+00 9.70357358e-01 2.13740133e-02
9.09105659e-01 -2.97160417e-01 1.84587911e-01 -1.33946717e+00
-2.59032279e-01 -1.63133442e-02 1.37467146e-01 5.63639283e-01
3.85738045e-01 -3.54660124e-01 1.19739497e+00 7.55374789e-01
-6.11107469e-01 -7.83629119e-01 -8.79776061e-01 -5.93502760e-01
2.17433155e-01 -2.40028873e-01 4.24495459e-01 9.06290233e-01
-3.83134604e-01 -2.75945842e-01 -4.35554713e-01 2.03358289e-02
4.52590644e-01 -2.60261502e-02 2.49568954e-01 -1.17637897e+00
-3.75779331e-01 -2.54762292e-01 -7.63748661e-02 -8.57752442e-01
-3.93893301e-01 -6.55115485e-01 3.00368816e-01 -1.67319763e+00
1.64737523e-01 -4.06887084e-01 -2.45413780e-01 5.11123180e-01
-8.52823555e-02 8.37815642e-01 1.50053293e-01 2.77140737e-01
2.91074425e-01 6.20508909e-01 1.85759580e+00 -4.79347706e-01
-3.50576520e-01 1.17716916e-01 -1.40535325e-01 7.71734715e-01
8.20460677e-01 -6.37569427e-01 -5.26465923e-02 -5.92428863e-01
8.41137990e-02 2.77287245e-01 3.25633526e-01 -1.42347133e+00
2.04630196e-01 1.10407636e-01 9.20292318e-01 -7.01118410e-01
4.47032452e-01 -1.16833723e+00 6.60658002e-01 9.90241945e-01
3.88975081e-04 -2.24146657e-02 1.91311717e-01 9.15646106e-02
-4.53561962e-01 -3.77108276e-01 1.29385221e+00 -6.91825151e-01
-5.42943776e-02 4.07915056e-01 -2.58269519e-01 -7.83517182e-01
6.77580535e-01 -4.21999276e-01 3.09700400e-01 -3.20655525e-01
-8.19175601e-01 -3.00149292e-01 5.96556008e-01 -7.79078603e-02
1.13898468e+00 -1.55543947e+00 -4.15475309e-01 2.99672931e-01
-2.55462646e-01 3.54807824e-01 6.89739764e-01 1.12543237e+00
-1.07152915e+00 -1.16533972e-01 -4.76658255e-01 -3.10770720e-01
-1.11679471e+00 5.61287284e-01 5.48926592e-01 -4.01492000e-01
-1.14363253e+00 7.45310426e-01 4.94771451e-01 -2.09023863e-01
9.86366197e-02 -7.26180792e-01 2.73337871e-01 -5.09057581e-01
5.60260713e-01 1.66802689e-01 4.99974877e-01 -9.16342065e-02
1.62155461e-02 4.94529068e-01 -1.82413206e-01 1.50956884e-02
1.45217931e+00 2.11237207e-01 -2.63867050e-01 -1.11475959e-01
9.93181348e-01 -1.20947972e-01 -1.29469490e+00 -2.33637244e-01
-4.42776680e-01 -5.58135748e-01 3.47177744e-01 -7.09874988e-01
-1.86474860e+00 9.12084699e-01 9.57062304e-01 -2.52147824e-01
1.22024298e+00 -3.25797379e-01 1.36551607e+00 -9.14189219e-02
3.95582885e-01 -9.42383945e-01 2.16755152e-01 -4.07246426e-02
1.12711775e+00 -1.05155897e+00 9.77062732e-02 -6.44364178e-01
-3.34558845e-01 1.35062301e+00 8.84322524e-01 -2.46083736e-01
7.07784235e-01 5.51898420e-01 1.34890467e-01 -2.11219609e-01
1.19214714e-01 2.89395213e-01 -1.98488310e-02 3.81726027e-01
4.20488566e-01 -2.61970144e-02 -7.27610350e-01 6.72384918e-01
3.52289639e-02 1.08288802e-01 6.30706847e-01 1.14389241e+00
-2.38272369e-01 -1.04007626e+00 -8.18844855e-01 6.72155917e-01
-5.45084417e-01 -1.42475948e-01 -2.07453802e-01 4.96991456e-01
3.18254560e-01 6.07278883e-01 -4.65814397e-02 -2.52784520e-01
5.08355796e-01 -2.62526661e-01 6.30003631e-01 -5.69695473e-01
-8.76180410e-01 4.32613313e-01 -4.52164501e-01 -6.70934677e-01
-1.15853831e-01 -1.87036037e-01 -1.35025609e+00 -2.20837012e-01
-3.82610828e-01 -3.02527010e-01 6.26276910e-01 4.85403955e-01
-1.47902623e-01 9.53575373e-01 3.61878246e-01 -8.44520867e-01
-2.13632897e-01 -1.24618280e+00 -8.42798829e-01 5.38943529e-01
1.95116192e-01 -5.95971823e-01 -1.05401538e-01 -2.15475053e-01] | [13.499804496765137, -2.5542221069335938] |
a746ffa7-1544-4e67-9ce5-f4e0b5bdbe5a | a-heat-jarrow-morton-framework-for-energy | 2305.01485 | null | https://arxiv.org/abs/2305.01485v2 | https://arxiv.org/pdf/2305.01485v2.pdf | A Heath-Jarrow-Morton framework for energy markets: a pragmatic approach | In this article we discuss the application of the Heath-Jarrow-Morton framework Heath et al. [26] to energy markets. The goal of the article is to give a detailed overview of the topic, focusing on practical aspects rather than on theory, which has been widely studied in literature. This work aims to be a guide for practitioners and for all those who deal with the practical issues of this approach to energy market. In particular, we focus on the markets' structure, model calibration by dimension reduction with Principal Component Analysis (PCA), Monte Carlo simulations and derivatives pricing. As application, we focus on European power and gas markets: we calibrate the model on historical futures quotations, we perform futures and spot simulations and we analyze the results. | ['Edoardo Santilli', 'Matteo Gardini'] | 2023-05-02 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [-5.55122614e-01 -1.93590328e-01 -3.36276330e-02 1.46988437e-01
-4.25620943e-01 -1.03875089e+00 1.00234711e+00 -3.74150485e-01
-1.10684946e-01 6.27363384e-01 7.57200047e-02 -7.28861153e-01
-4.63406265e-01 -8.95492733e-01 -1.94520783e-02 -7.61725903e-01
-2.61596859e-01 5.36945820e-01 -4.01358694e-01 -1.62448660e-01
2.53640234e-01 7.60289550e-01 -8.84013176e-01 -5.24872303e-01
3.50589544e-01 1.09688032e+00 -1.68194622e-01 4.97738630e-01
1.01829082e-01 7.48670161e-01 -2.79609382e-01 -7.33343720e-01
8.40196073e-01 -5.21501005e-01 -5.03901482e-01 1.39064983e-01
-8.44629943e-01 -3.71298969e-01 1.53600657e-02 1.20050848e+00
3.80850673e-01 3.13569993e-01 9.07617092e-01 -1.12153184e+00
-6.05148494e-01 9.41536784e-01 -5.99510312e-01 3.35179865e-01
-1.11804187e-01 -4.79823500e-02 1.09439647e+00 -8.86845112e-01
1.18974212e-03 1.07593381e+00 5.22921085e-01 2.51077473e-01
-1.17181647e+00 -1.19369119e-01 -2.75562406e-01 2.90418506e-01
-1.12971687e+00 -1.67080611e-01 1.05562484e+00 -6.61000073e-01
1.00194383e+00 7.06537962e-01 8.45125616e-01 4.51857448e-01
2.47662053e-01 7.25709915e-01 1.26030838e+00 -6.34222150e-01
9.30028856e-01 6.03097584e-03 4.03909944e-02 -5.22608101e-01
2.37719923e-01 7.86584139e-01 6.93868771e-02 -3.89731050e-01
6.97095037e-01 -8.29463899e-02 2.36029193e-01 -4.69200432e-01
-7.18066573e-01 1.32272804e+00 -3.03503275e-02 7.42281795e-01
-9.88865137e-01 1.45651042e-01 2.01261297e-01 2.70866603e-01
6.11952662e-01 -3.50961611e-02 -7.44808853e-01 -3.54179621e-01
-9.90167499e-01 3.88351142e-01 9.94327307e-01 4.53578889e-01
1.82501063e-01 5.62856317e-01 2.72351027e-01 4.74169254e-01
7.63526797e-01 9.50374246e-01 6.01423800e-01 -9.26170409e-01
9.35231894e-02 -4.13357615e-02 2.82240242e-01 -5.74630916e-01
-3.67317408e-01 -1.03351615e-01 -7.89633274e-01 4.21427637e-01
4.04278412e-02 -7.12903976e-01 -1.88312694e-01 1.09045863e+00
6.63748831e-02 -2.53157001e-02 3.14842910e-01 6.62068665e-01
-7.95801207e-02 7.15813220e-01 -6.55728253e-03 -7.07872391e-01
1.24908197e+00 -7.27993250e-01 -7.72523940e-01 4.94765073e-01
8.51728320e-02 -9.92329180e-01 4.69761252e-01 4.54440206e-01
-1.18428516e+00 -1.10084556e-01 -5.45183778e-01 6.56126201e-01
-4.69025433e-01 -1.91679616e-02 7.12833166e-01 1.29343009e+00
-1.01469684e+00 7.95222104e-01 -1.07992768e+00 -1.49872750e-01
-2.23225638e-01 -1.22381011e-02 4.39854860e-01 1.05489540e+00
-1.18781531e+00 1.24270260e+00 3.79686594e-01 2.86859125e-01
-3.92243445e-01 -6.26157403e-01 -6.03303850e-01 2.65163541e-01
3.16605717e-02 -5.34262955e-01 1.67048824e+00 -6.51595414e-01
-1.94885838e+00 8.08832571e-02 3.62537086e-01 -8.76715183e-01
8.15122247e-01 2.37902403e-02 -7.34221160e-01 1.10570015e-02
-3.02268714e-01 -3.04075748e-01 4.49244946e-01 -8.05340886e-01
-6.02193773e-01 -2.58866221e-01 -4.43650216e-01 -1.13075696e-01
2.13166907e-01 4.70321178e-01 4.55213904e-01 -1.12893522e+00
7.25730062e-02 -9.51248825e-01 -3.46454561e-01 -9.15012836e-01
-6.25245199e-02 -2.93723375e-01 4.22511399e-01 -9.05824900e-01
1.38540030e+00 -1.78121150e+00 -2.53644027e-02 8.05049658e-01
-6.19165301e-01 -2.12947890e-01 4.28083003e-01 1.09178030e+00
-6.94796324e-01 1.92162067e-01 -6.53244495e-01 -7.47351944e-02
7.59421885e-01 2.44283378e-01 -7.70583451e-01 6.90610945e-01
7.53803402e-02 9.86279726e-01 -4.76046234e-01 2.52711087e-01
7.10345507e-01 1.42535180e-01 -2.79944688e-02 -3.15774232e-01
-2.57579479e-02 2.87339892e-02 -4.68632191e-01 5.72892487e-01
9.04286981e-01 2.75498599e-01 1.66781262e-01 1.25022769e-01
-6.47498965e-01 -3.16225216e-02 -1.31773746e+00 1.02070940e+00
-1.55835420e-01 2.86759555e-01 -3.79120298e-02 -1.15919948e+00
6.84094012e-01 4.14980501e-01 8.82810473e-01 -6.42643213e-01
1.93976611e-01 3.68402213e-01 -9.11594927e-02 -9.13751423e-02
3.70878518e-01 -6.52410150e-01 -6.61555454e-02 9.12713587e-01
-3.59384030e-01 -5.11262715e-01 4.99448687e-01 -2.12966695e-01
5.18099129e-01 7.31455535e-02 4.29481030e-01 -6.60937369e-01
5.41508257e-01 -8.10077414e-02 2.65216261e-01 1.63605567e-02
1.70444194e-02 1.78591445e-01 6.72755241e-01 -3.54087144e-01
-1.32378066e+00 -9.35117960e-01 -4.87846315e-01 4.87065226e-01
-6.00031257e-01 -6.60848692e-02 -7.64965832e-01 -3.20356995e-01
1.94347277e-01 1.47501063e+00 -6.27509058e-01 4.83383477e-01
-2.10975841e-01 -1.74867642e+00 -1.92573190e-01 3.25724065e-01
2.59801656e-01 -9.59372759e-01 -1.00076532e+00 2.54637420e-01
4.64573473e-01 -4.81538892e-01 7.17234164e-02 6.04158863e-02
-7.92978942e-01 -1.02476108e+00 -9.98044908e-01 1.18958674e-01
1.60739407e-01 -3.89509201e-01 1.29197454e+00 -5.28553009e-01
1.26395494e-01 7.27198005e-01 -4.61633772e-01 -6.64382219e-01
-4.89128649e-01 -3.30018550e-01 -6.16534576e-02 1.98448356e-02
7.02077448e-01 -5.80246389e-01 -5.86722493e-01 3.31629068e-01
-1.04998112e+00 -3.62688541e-01 3.49569023e-01 6.02379501e-01
3.82460147e-01 5.74084163e-01 3.40584785e-01 -6.45875394e-01
9.26650882e-01 -4.93494600e-01 -1.50250649e+00 3.09785247e-01
-1.33355212e+00 -1.56541392e-01 8.82964954e-02 1.64265022e-01
-1.08360684e+00 2.09843531e-01 -1.92353979e-01 -1.14058480e-01
2.81632543e-01 6.52656972e-01 5.07756919e-02 -7.57992594e-03
2.05447733e-01 4.73082364e-02 -4.32833768e-02 -9.80652809e-01
6.13485873e-01 3.66731048e-01 2.66227245e-01 -5.37467718e-01
1.11882210e+00 3.51939440e-01 3.20332944e-01 -4.86498207e-01
4.79417145e-01 -2.83323050e-01 -6.24723792e-01 -3.17409858e-02
6.12154722e-01 -5.87264657e-01 -6.71056747e-01 4.67860311e-01
-7.48408318e-01 -3.44836742e-01 -6.81736946e-01 6.59194112e-01
-1.02797747e+00 3.43013257e-01 -8.78770947e-01 -1.55708170e+00
-3.47070575e-01 -9.53303158e-01 5.07362723e-01 2.66521007e-01
-5.29882405e-03 -1.51915538e+00 7.00246811e-01 -2.13316292e-01
6.62770212e-01 4.40389454e-01 7.23210931e-01 -8.44569981e-01
-3.49307984e-01 -4.46863994e-02 3.18110138e-01 5.47676444e-01
3.85313779e-02 3.29919457e-01 -9.93107378e-01 -1.34828538e-01
6.68438733e-01 5.63594103e-01 4.30326819e-01 4.93730813e-01
3.75690162e-01 -3.89060527e-01 2.91927308e-01 6.94325566e-02
1.68633080e+00 5.25169790e-01 6.23231530e-01 6.53169572e-01
-1.69087306e-01 8.30528200e-01 3.61741126e-01 8.49566698e-01
6.08956665e-02 3.73405010e-01 2.18948707e-01 7.13866204e-02
5.21488965e-01 1.09186329e-01 3.29542458e-01 9.38528240e-01
-3.10968995e-01 1.40616089e-01 -8.99353027e-01 4.54566360e-01
-2.00590301e+00 -1.01348376e+00 -1.69322550e-01 1.88430285e+00
3.47355813e-01 -2.65210241e-01 7.09522247e-01 2.09376603e-01
4.71374422e-01 -9.56240296e-02 -2.52215594e-01 -4.49164629e-01
-4.15376782e-01 1.67714715e-01 9.74590063e-01 4.51489776e-01
-1.01246226e+00 2.54028916e-01 7.36244917e+00 8.16312969e-01
-6.11937046e-01 7.77815357e-02 8.84862244e-01 1.42808944e-01
-6.75274193e-01 3.22702527e-01 -6.96735978e-02 7.26997852e-01
1.31016338e+00 -7.69640505e-01 6.99001849e-01 8.45819652e-01
7.07643926e-01 -2.56317347e-01 -7.99433708e-01 9.28220630e-01
-3.50162745e-01 -1.05797207e+00 -4.26711440e-01 4.50138211e-01
9.59253371e-01 9.76656601e-02 2.62450308e-01 -2.64538010e-03
3.93103093e-01 -6.59118950e-01 8.64407659e-01 7.90453672e-01
-2.04006359e-01 -1.08764088e+00 9.69093263e-01 1.66355833e-01
-1.02188337e+00 -2.13949114e-01 -5.39254248e-01 -1.61072567e-01
6.19040668e-01 7.49524117e-01 -1.16724089e-01 9.44674730e-01
6.73671663e-01 2.53664106e-01 -1.22450992e-01 9.25168395e-01
2.10501440e-03 4.86073673e-01 -5.97474098e-01 -3.13992091e-02
1.29459277e-01 -1.09912574e+00 1.73168331e-01 9.60223079e-01
6.78583384e-01 4.11729425e-01 -4.92267013e-01 1.06916952e+00
6.08310461e-01 3.50104600e-01 -1.93975970e-01 -3.04038465e-01
2.07974985e-01 1.01754999e+00 -9.30366397e-01 -2.83217579e-01
-9.57211077e-01 5.27002156e-01 -8.93465161e-01 4.60165411e-01
-3.94078821e-01 -5.83086982e-02 5.14209330e-01 -1.58718660e-01
4.43512797e-01 -1.34089395e-01 -5.47613323e-01 -1.07964873e+00
-2.11425200e-01 -6.20226085e-01 5.64798772e-01 -5.16277850e-01
-1.59647727e+00 -6.66622147e-02 5.70028126e-01 -9.91649568e-01
-1.03386676e+00 -9.51088965e-01 -8.26507866e-01 1.34149611e+00
-1.51931727e+00 -5.16200781e-01 8.07475388e-01 2.45171323e-01
3.86937529e-01 -6.03298724e-01 6.90927863e-01 2.06453446e-02
-6.98009729e-01 -1.98416352e-01 1.08983362e+00 1.69328988e-01
-3.21141541e-01 -1.61221910e+00 1.04725504e+00 9.81478095e-01
2.78322637e-01 4.16041225e-01 9.22131956e-01 -6.45317972e-01
-1.41999149e+00 -3.64897639e-01 6.84449434e-01 -5.81661940e-01
1.41882885e+00 2.33344018e-01 -3.65254402e-01 5.88813424e-01
9.25073147e-01 -5.82632780e-01 7.37469435e-01 -4.81046647e-01
3.94332111e-01 7.24356845e-02 -1.26813769e+00 2.12629452e-01
-1.14807047e-01 -4.79754508e-01 -7.33738899e-01 3.04024994e-01
-1.07722834e-01 1.09428041e-01 -1.28141379e+00 1.48501992e-01
4.57631409e-01 -1.20271254e+00 9.37896132e-01 -2.10372433e-01
-4.30918783e-01 -3.43005449e-01 -1.58777192e-01 -1.17350948e+00
-2.14636862e-01 -1.20542634e+00 -7.41170496e-02 1.34985006e+00
3.79237026e-01 -9.49948549e-01 5.64667583e-01 1.35579169e+00
4.09403503e-01 -2.01041400e-01 -1.42554426e+00 -8.39858711e-01
7.27357984e-01 -6.14387810e-01 1.20158863e+00 8.23740602e-01
3.06242019e-01 -4.86732535e-02 -2.50053763e-01 -1.34179771e-01
7.83880115e-01 3.41097742e-01 2.44355336e-01 -1.11734676e+00
-3.47358704e-01 -9.34502363e-01 1.01647750e-01 -5.33116817e-01
-1.88694119e-01 -4.12686676e-01 -4.73458260e-01 -1.05923808e+00
7.66565874e-02 -6.74139783e-02 -3.92874271e-01 -1.89212024e-01
5.85691094e-01 3.69926402e-03 6.15952730e-01 2.00622588e-01
3.45134586e-01 5.65479040e-01 4.56071496e-01 3.72053981e-02
7.30406120e-02 4.95881379e-01 -3.99090916e-01 5.62155068e-01
1.12748265e+00 -1.85005352e-01 -3.59829754e-01 1.89925984e-01
6.39398456e-01 3.56508605e-02 2.84242034e-01 -2.90801674e-01
-6.56268895e-02 -3.51662904e-01 2.62454152e-01 -1.01942766e+00
-1.80589393e-01 -1.30015075e+00 1.01765239e+00 6.52186096e-01
1.39869630e-01 8.96175325e-01 3.78139205e-02 3.42492014e-01
-1.46984920e-01 -8.76873434e-01 3.82998466e-01 -2.41831914e-01
-5.71425319e-01 3.46266702e-02 -4.03214186e-01 -4.75944847e-01
1.29890180e+00 2.91774541e-01 -8.68727863e-02 -3.92900705e-01
-9.58054483e-01 3.93465430e-01 6.16324723e-01 4.40245233e-02
7.86988437e-02 -1.60324490e+00 -7.92847574e-01 2.48005949e-02
-6.89062297e-01 -6.91957533e-01 2.28719041e-01 7.41096914e-01
-7.79063284e-01 7.79336452e-01 7.98861012e-02 9.14507639e-03
-4.20901924e-01 1.05591369e+00 4.61980343e-01 1.45105805e-04
-3.54671419e-01 4.77806330e-02 -1.08295344e-01 1.17529668e-01
-3.00170690e-01 -5.07804215e-01 -3.38143945e-01 4.53628540e-01
4.52971637e-01 9.68435884e-01 -1.18747000e-02 -8.53947341e-01
-3.42373073e-01 8.25718164e-01 8.40474606e-01 -8.05932224e-01
1.43565083e+00 -4.60472584e-01 -4.24218237e-01 7.56765842e-01
8.17013502e-01 -5.28044663e-02 -1.25519443e+00 3.45577687e-01
6.73926532e-01 -1.98260337e-01 2.56116152e-01 -5.34115195e-01
-1.32294965e+00 6.43045008e-01 8.32083166e-01 1.10704553e+00
1.16551900e+00 -2.69878060e-01 4.46959823e-01 7.64216930e-02
1.10077247e-01 -1.71586847e+00 -8.87711227e-01 -5.45323603e-02
1.03104734e+00 -7.06768870e-01 4.39402834e-02 2.43033636e-02
-6.84343755e-01 1.38335896e+00 -8.21455240e-01 -2.66100645e-01
1.40493989e+00 2.27483034e-01 4.45872247e-02 2.14673743e-01
-5.31947732e-01 -1.53391317e-01 1.22582652e-01 4.30029511e-01
1.62998423e-01 4.71111804e-01 -9.27837491e-01 7.19911277e-01
-5.33552825e-01 -1.00220963e-01 5.46571255e-01 8.39879215e-01
-1.16142407e-01 -1.41424751e+00 -6.72321737e-01 8.60307068e-02
-7.02552140e-01 -6.71755746e-02 -2.53883779e-01 8.23896945e-01
-4.35162842e-01 9.13572073e-01 1.65980682e-01 1.51432142e-01
5.21948874e-01 3.05937022e-01 2.91428361e-02 4.75747278e-03
-5.23268342e-01 9.70970213e-01 7.51585048e-03 -1.04915828e-01
-6.69187129e-01 -1.49303019e+00 -6.94212139e-01 -8.02557051e-01
-1.95176825e-01 6.54870152e-01 9.87443686e-01 6.00451112e-01
-2.88627028e-01 1.00137606e-01 1.07689381e+00 -6.28297031e-01
-1.28244269e+00 -9.26210999e-01 -1.43551266e+00 -3.46265472e-02
1.47845224e-01 -3.35265249e-01 -6.53059185e-01 -2.48151226e-03] | [5.285648345947266, 3.9043285846710205] |
abb42e2a-ba6f-47c8-80d2-7b79ddfb1c8d | active-learning-strategies-for-weakly | 2207.12112 | null | https://arxiv.org/abs/2207.12112v1 | https://arxiv.org/pdf/2207.12112v1.pdf | Active Learning Strategies for Weakly-supervised Object Detection | Object detectors trained with weak annotations are affordable alternatives to fully-supervised counterparts. However, there is still a significant performance gap between them. We propose to narrow this gap by fine-tuning a base pre-trained weakly-supervised detector with a few fully-annotated samples automatically selected from the training set using ``box-in-box'' (BiB), a novel active learning strategy designed specifically to address the well-documented failure modes of weakly-supervised detectors. Experiments on the VOC07 and COCO benchmarks show that BiB outperforms other active learning techniques and significantly improves the base weakly-supervised detector's performance with only a few fully-annotated images per class. BiB reaches 97% of the performance of fully-supervised Fast RCNN with only 10% of fully-annotated images on VOC07. On COCO, using on average 10 fully-annotated images per class, or equivalently 1% of the training set, BiB also reduces the performance gap (in AP) between the weakly-supervised detector and the fully-supervised Fast RCNN by over 70%, showing a good trade-off between performance and data efficiency. Our code is publicly available at https://github.com/huyvvo/BiB. | ['Jean Ponce', 'Patrick Pérez', 'Andrei Bursuc', 'Spyros Gidaris', 'Oriane Siméoni', 'Huy V. Vo'] | 2022-07-25 | null | null | null | null | ['weakly-supervised-object-detection'] | ['computer-vision'] | [-2.66192462e-02 5.02648413e-01 -6.15786731e-01 -3.29419762e-01
-1.42272067e+00 -7.52455413e-01 5.54451287e-01 2.33138025e-01
-9.48382735e-01 3.91454011e-01 6.86531365e-02 5.98298721e-02
4.20205891e-01 -3.70134860e-01 -7.34568655e-01 -7.99857318e-01
1.12389095e-01 5.73893368e-01 8.22995424e-01 3.05895925e-01
-2.01817006e-01 1.90305918e-01 -1.49187624e+00 3.64183515e-01
5.02583623e-01 1.00118303e+00 2.72927821e-01 7.81106770e-01
9.24620628e-02 8.31415057e-01 -4.10996944e-01 -3.65502954e-01
3.80024493e-01 -1.14010341e-01 -6.59331918e-01 1.04013495e-01
6.16011918e-01 -2.82479256e-01 -1.19516216e-01 8.99078488e-01
4.99013901e-01 -2.62672573e-01 5.05860507e-01 -1.09124279e+00
-3.00098628e-01 6.43013060e-01 -4.63936746e-01 3.25303197e-01
-7.29584973e-03 4.30990756e-01 1.30503607e+00 -1.52671683e+00
4.96662080e-01 8.26903880e-01 8.74623179e-01 6.70768976e-01
-1.25967026e+00 -7.46193349e-01 2.62445509e-01 8.40891711e-03
-1.55695379e+00 -6.87817633e-01 3.89746666e-01 -4.19957608e-01
8.47215116e-01 3.83177027e-02 6.20608389e-01 9.36311007e-01
-5.07216513e-01 1.46151090e+00 8.14133584e-01 -5.79180717e-01
3.95988971e-01 5.92301130e-01 4.57338572e-01 8.26098084e-01
4.05540407e-01 2.12650776e-01 -5.04425228e-01 -4.21192557e-01
5.07686138e-01 -1.60958156e-01 -8.07078108e-02 -6.71831012e-01
-1.11248946e+00 8.55573416e-01 4.56871778e-01 5.03567718e-02
1.42246326e-02 7.01547787e-02 5.13460398e-01 -3.37376520e-02
5.90662479e-01 5.67477107e-01 -5.85743189e-01 1.60720095e-01
-9.54225957e-01 -2.28583634e-01 8.08655858e-01 1.20193756e+00
8.52180779e-01 -1.71177126e-02 -1.99812949e-01 8.47180128e-01
3.52379322e-01 4.59653020e-01 1.99383527e-01 -7.80269682e-01
2.77063549e-01 8.82740140e-01 1.58098280e-01 -1.66718081e-01
-5.70663959e-02 -8.06982517e-01 -3.97320300e-01 2.70755470e-01
5.76762795e-01 -1.46544129e-01 -1.03602695e+00 1.27243161e+00
2.48875588e-01 -1.09158762e-01 -1.82782605e-01 7.63488591e-01
9.09745753e-01 5.11862218e-01 3.39937747e-01 -3.15060586e-01
1.03178155e+00 -1.39530504e+00 -2.91967601e-01 -5.80502808e-01
8.34749520e-01 -5.97486615e-01 1.05497861e+00 2.27127224e-01
-9.15667653e-01 -6.15996420e-01 -1.13605928e+00 1.17108904e-01
-2.47562468e-01 6.46330476e-01 5.32777488e-01 5.52374065e-01
-9.86813009e-01 1.51919844e-02 -9.04439449e-01 -2.55396068e-01
9.16144252e-01 4.29676384e-01 -3.70373279e-01 -1.54845774e-01
-6.97898984e-01 7.17544079e-01 6.51084781e-01 -2.77152024e-02
-1.32230461e+00 -8.45224619e-01 -7.56510973e-01 -2.97238622e-02
7.25113928e-01 -1.29659846e-01 1.58584750e+00 -1.20486474e+00
-1.07265329e+00 1.32038057e+00 1.52353542e-02 -6.61504447e-01
6.82948232e-01 -4.25209433e-01 -2.18803108e-01 8.94334987e-02
2.48456553e-01 1.09989369e+00 6.27379775e-01 -1.21090519e+00
-9.61421669e-01 -8.06975588e-02 3.38455066e-02 1.56356037e-01
-3.86829257e-01 1.65934548e-01 -8.06768417e-01 -3.04136068e-01
-1.31713629e-01 -1.01494205e+00 -3.56460333e-01 5.53894520e-01
-2.96932191e-01 -6.47296190e-01 9.31796491e-01 -2.16014683e-02
8.43026936e-01 -2.31872892e+00 -4.30439800e-01 -2.07392752e-01
4.45708305e-01 8.41041446e-01 -2.67143756e-01 1.21949643e-01
-1.73090808e-02 -9.78591740e-02 -2.88319111e-01 -6.38583899e-01
-2.61803746e-01 2.36283168e-01 -2.01885834e-01 6.53099298e-01
4.78357613e-01 9.56899703e-01 -1.30436599e+00 -6.68589413e-01
1.47252336e-01 2.44473130e-01 -3.64590675e-01 3.58294249e-01
-2.56524295e-01 2.76560456e-01 -1.18666336e-01 8.92745197e-01
4.87740457e-01 -4.06729281e-01 7.25282803e-02 -1.42133003e-03
-1.74820036e-01 1.96935773e-01 -1.04555368e+00 1.50914454e+00
-1.02627963e-01 8.63175213e-01 3.40659395e-02 -1.05747449e+00
8.88438761e-01 3.49164128e-01 3.03407043e-01 -5.39199352e-01
-4.11078259e-02 4.45851326e-01 -4.22065780e-02 -6.82781637e-02
2.83586252e-02 3.48947793e-01 -9.07377899e-02 3.42437059e-01
3.95318568e-01 2.33929485e-01 5.05738497e-01 3.41836363e-01
1.31704164e+00 7.79294223e-02 5.70588768e-01 -2.54822105e-01
3.99654180e-01 3.35913658e-01 8.31474364e-01 1.16860068e+00
-6.38255596e-01 8.16411018e-01 4.58617568e-01 -3.06914419e-01
-1.10361660e+00 -1.18790269e+00 -3.31172705e-01 1.51238525e+00
2.11015325e-02 -4.80669171e-01 -6.41642630e-01 -1.26985395e+00
-6.36207834e-02 3.89624029e-01 -6.06932521e-01 -8.93663149e-03
-4.69859749e-01 -8.78547311e-01 6.50253117e-01 8.72927189e-01
6.25875354e-01 -9.52580392e-01 -4.05323386e-01 5.42111322e-03
2.20134422e-01 -1.13912892e+00 -2.16312677e-01 6.55611277e-01
-7.86037743e-01 -1.15604043e+00 -7.96636939e-01 -8.90002966e-01
9.45824742e-01 4.20639336e-01 1.35294569e+00 1.45640105e-01
-3.66353691e-01 2.85094857e-01 -5.27379155e-01 -5.32089233e-01
-4.14663792e-01 2.40620345e-01 -1.01636909e-01 -1.80620044e-01
5.50390959e-01 1.26026079e-01 -6.33378565e-01 6.08060539e-01
-5.15003026e-01 -1.21880770e-01 8.59517753e-01 8.86664271e-01
7.04141915e-01 -3.92394096e-01 4.65104192e-01 -1.16442657e+00
-2.13587910e-01 -4.08833295e-01 -8.13160062e-01 7.04039037e-02
-6.59268439e-01 -1.95070058e-01 3.96057785e-01 -6.55422688e-01
-9.55047607e-01 8.89000773e-01 -8.80507231e-02 -2.97985792e-01
-1.67689145e-01 -6.52566627e-02 -9.06103253e-02 5.35921827e-02
1.12487197e+00 1.80580840e-02 -3.38429600e-01 -4.95849431e-01
3.30010355e-01 7.47069180e-01 6.10508621e-01 -1.76774129e-01
9.51176405e-01 6.14740491e-01 -6.99855983e-01 -4.87323165e-01
-1.55483174e+00 -1.27354586e+00 -9.40747261e-01 -2.33876809e-01
5.75708866e-01 -1.33175838e+00 -1.24428943e-01 2.37095341e-01
-9.42937672e-01 -6.46570921e-01 -6.57105684e-01 5.10267198e-01
-3.17704231e-01 2.43810918e-02 -5.19148350e-01 -9.15794253e-01
-4.61648792e-01 -9.62803721e-01 1.12677884e+00 1.15694046e-01
-2.70862460e-01 -8.10111165e-01 2.53797948e-01 4.79653835e-01
1.08987257e-01 -6.08069189e-02 1.61466286e-01 -9.62852776e-01
-5.20834267e-01 -6.98174238e-01 -2.81592339e-01 7.21323788e-01
-2.56769806e-01 -1.40426576e-01 -1.36467707e+00 -5.18074930e-01
-3.87259066e-01 -7.62820423e-01 1.28355503e+00 5.07684387e-02
8.69692028e-01 8.90467595e-03 -5.79957128e-01 1.64349645e-01
1.36288750e+00 -2.61249542e-02 4.72345859e-01 1.55273557e-01
5.69309413e-01 2.07938999e-01 8.39473963e-01 1.64262697e-01
-3.40236872e-02 5.40206134e-01 3.27803910e-01 -5.36296248e-01
-3.21280688e-01 -3.38056773e-01 5.70163310e-01 4.38822806e-01
4.17322479e-02 -9.49956253e-02 -1.19966924e+00 7.27096200e-01
-1.93044770e+00 -8.36719155e-01 -2.18504295e-01 2.38992190e+00
9.19245839e-01 6.54918969e-01 3.88477653e-01 2.36103654e-01
7.82951593e-01 -6.99887350e-02 -6.27393067e-01 1.75047770e-01
-1.75192177e-01 4.44897264e-02 8.16834748e-01 3.26819837e-01
-1.58131266e+00 9.25169826e-01 6.71813726e+00 9.17287648e-01
-8.86680722e-01 5.24831235e-01 6.68179750e-01 -2.89231062e-01
4.66092885e-01 2.35750616e-01 -1.40874660e+00 3.63934964e-01
8.42984140e-01 1.26632765e-01 -3.77578139e-01 1.53369808e+00
-1.81604140e-02 -3.73213649e-01 -1.37319899e+00 8.96339893e-01
9.69879031e-02 -1.42101765e+00 -4.08867538e-01 2.17775516e-02
9.09853995e-01 5.40376365e-01 -2.14485884e-01 6.29791737e-01
4.53880757e-01 -7.14529812e-01 8.17313492e-01 -6.45277202e-02
7.49189377e-01 -4.55337286e-01 8.93136084e-01 7.14167893e-01
-1.20091510e+00 -2.71391451e-01 -5.05517960e-01 4.80723344e-02
-1.76934183e-01 8.13835561e-01 -1.06482279e+00 -1.55901745e-01
8.86639893e-01 6.98830009e-01 -9.80475426e-01 1.40096903e+00
-4.89210784e-01 1.16290069e+00 -4.51061457e-01 6.17068969e-02
1.58972993e-01 4.66322571e-01 4.60071057e-01 1.57541323e+00
-3.80711287e-01 -4.27936986e-02 3.99347037e-01 6.95482075e-01
-2.55773425e-01 -1.44813105e-01 -3.40855688e-01 1.08730599e-01
5.90807557e-01 1.46948349e+00 -9.25524116e-01 -6.41478121e-01
-5.49340367e-01 7.32677579e-01 6.31395936e-01 1.42444894e-01
-8.18630099e-01 -2.26873718e-02 8.68720412e-02 3.03470314e-01
4.06223625e-01 2.56249048e-02 -1.16441131e-01 -9.91101325e-01
-4.55300920e-02 -5.64094245e-01 4.97248203e-01 -5.46464741e-01
-1.27747726e+00 4.21865433e-01 -8.66044015e-02 -1.47235799e+00
7.29847401e-02 -7.71925092e-01 -6.07424498e-01 4.03265178e-01
-1.21142483e+00 -1.15298080e+00 -4.11073864e-01 2.07972556e-01
7.63142288e-01 -1.81156635e-01 7.41069078e-01 3.59787673e-01
-7.53432691e-01 7.98882663e-01 1.95606306e-01 6.56786919e-01
8.68111730e-01 -1.34148669e+00 4.31534857e-01 8.72757792e-01
3.95479262e-01 3.72626215e-01 3.96903276e-01 -3.06495279e-01
-9.88892555e-01 -1.19546866e+00 6.95332289e-01 -7.49548495e-01
4.90450531e-01 -7.75241494e-01 -9.02545571e-01 8.00417364e-01
-7.03903139e-02 6.87257707e-01 7.44936764e-01 1.91148371e-01
-7.19207585e-01 -1.45192236e-01 -9.62290764e-01 2.10049748e-01
1.09074581e+00 -6.14525259e-01 -5.38926482e-01 7.10666358e-01
4.96261120e-01 -2.19035536e-01 -4.76377130e-01 5.85942984e-01
3.07104200e-01 -9.35159147e-01 8.83246660e-01 -1.91320807e-01
1.84617322e-02 -3.47106934e-01 1.79133445e-01 -8.77698004e-01
-4.02028412e-01 -2.91772336e-01 -2.45844916e-01 1.18879199e+00
8.61308575e-01 -5.41238487e-01 1.18717289e+00 2.47640952e-01
-1.98668659e-01 -8.61958563e-01 -6.79642200e-01 -9.06436384e-01
-6.31144792e-02 -3.73016536e-01 -3.78937036e-01 7.48188913e-01
-1.16301522e-01 4.66665834e-01 3.93339172e-02 -4.41466309e-02
7.62563765e-01 -2.55347550e-01 7.59209812e-01 -1.24484801e+00
-4.15795356e-01 -1.21342562e-01 -4.26042408e-01 -1.31701946e+00
-1.04715824e-01 -8.75435472e-01 4.37024802e-01 -1.18173814e+00
8.24331880e-01 -7.02939808e-01 -2.52924591e-01 8.78591239e-01
-2.09430039e-01 8.39740813e-01 3.73063534e-02 6.70524657e-01
-1.12010288e+00 1.40647560e-01 7.61353672e-01 -1.95659354e-01
-1.55467659e-01 1.05692014e-01 -5.06900609e-01 1.00123322e+00
6.72468781e-01 -8.43917310e-01 -3.03866476e-01 -2.41762757e-01
4.80825044e-02 -3.67238373e-01 4.13536698e-01 -1.05012667e+00
4.29765671e-01 2.97674835e-01 5.39000988e-01 -7.74165869e-01
2.31554836e-01 -4.59821552e-01 -3.42075557e-01 5.27576983e-01
-6.00136399e-01 -3.91408354e-01 5.86724132e-02 7.41443753e-01
-2.58390665e-01 -3.41786712e-01 1.22885239e+00 -2.38463059e-01
-1.01527655e+00 2.37242281e-01 -3.90709817e-01 1.81713223e-01
1.17778766e+00 -3.92079264e-01 -3.91162574e-01 8.12602788e-02
-7.88437188e-01 2.35963374e-01 2.71861464e-01 2.41350427e-01
2.32559845e-01 -1.06569982e+00 -6.29021049e-01 -2.34938003e-02
6.26523733e-01 3.64437759e-01 -2.04442024e-01 8.15264583e-01
-3.80062610e-01 5.40191889e-01 2.61078119e-01 -1.04859030e+00
-1.34253824e+00 2.92393476e-01 3.57609004e-01 -2.96039015e-01
-5.20388782e-01 1.11645389e+00 3.77285302e-01 -5.27880788e-01
6.05168819e-01 6.57339320e-02 6.71086982e-02 1.49512842e-01
6.02007091e-01 3.24784875e-01 1.03114143e-01 -4.23426121e-01
-5.56444407e-01 8.31800848e-02 -3.45418572e-01 6.38563111e-02
9.97984111e-01 2.89097548e-01 3.50052208e-01 4.78668272e-01
1.21678662e+00 -4.53865677e-02 -1.67113411e+00 -5.73212445e-01
3.07177544e-01 -2.48916745e-01 2.46583343e-01 -7.44546473e-01
-9.72686529e-01 6.52342677e-01 8.08404207e-01 -7.17584044e-02
7.78398395e-01 6.43912971e-01 3.18800896e-01 5.75216055e-01
3.84631038e-01 -1.23096144e+00 6.30336463e-01 3.16222072e-01
6.76583767e-01 -1.67534018e+00 2.73589283e-01 -6.58201516e-01
-6.61249518e-01 6.88852966e-01 8.87007177e-01 -3.54388952e-01
4.97260422e-01 5.02589285e-01 2.21667901e-01 -4.89631202e-03
-8.17819536e-01 -4.88026410e-01 2.70573527e-01 4.79281932e-01
4.81753498e-01 5.32806059e-03 2.49797665e-02 2.21850365e-01
2.93578565e-01 -1.02682263e-01 1.70112804e-01 1.00339866e+00
-7.52964854e-01 -8.45566034e-01 -3.28182697e-01 3.99074763e-01
-1.63959369e-01 -9.04474873e-03 -6.63960755e-01 9.07430470e-01
2.83581376e-01 8.99492502e-01 3.02514493e-01 2.80795936e-02
2.33827487e-01 -6.30729785e-03 2.23178074e-01 -1.15022242e+00
-5.90404153e-01 1.20747954e-01 2.65353680e-01 -4.44752961e-01
-5.08592546e-01 -6.79201543e-01 -1.11490774e+00 1.11732505e-01
-8.23085129e-01 1.98402591e-02 2.34210595e-01 8.37836504e-01
1.32255912e-01 -2.82797273e-02 5.23326874e-01 -7.65985847e-01
-6.15071476e-01 -1.05222166e+00 -1.33225128e-01 2.01905578e-01
1.80277124e-01 -5.76982379e-01 -7.15660989e-01 9.28722546e-02] | [9.243380546569824, 1.2538433074951172] |
1962c496-b2ae-4e51-b3b8-9afcdd1bb18c | structural-break-detection-in-quantile | 2302.05193 | null | https://arxiv.org/abs/2302.05193v1 | https://arxiv.org/pdf/2302.05193v1.pdf | Structural Break Detection in Quantile Predictive Regression Models with Persistent Covariates | We propose an econometric environment for structural break detection in nonstationary quantile predictive regressions. We establish the limit distributions for a class of Wald and fluctuation type statistics based on both the ordinary least squares estimator and the endogenous instrumental regression estimator proposed by Phillips and Magdalinos (2009a, Econometric Inference in the Vicinity of Unity. Working paper, Singapore Management University). Although the asymptotic distribution of these test statistics appears to depend on the chosen estimator, the IVX based tests are shown to be asymptotically nuisance parameter-free regardless of the degree of persistence and consistent under local alternatives. The finite-sample performance of both tests is evaluated via simulation experiments. An empirical application to house pricing index returns demonstrates the practicality of the proposed break tests for regression quantiles of nonstationary time series data. | ['Christis Katsouris'] | 2023-02-10 | null | null | null | null | ['unity'] | ['computer-vision'] | [-1.88819617e-01 -3.84518981e-01 -6.74156368e-01 -1.85714468e-01
-1.04091358e+00 -8.47188473e-01 4.43135947e-01 3.06249224e-02
3.18608806e-02 1.09514165e+00 1.53837860e-01 -9.82025504e-01
-9.46194887e-01 -6.66799784e-01 -4.56832886e-01 -8.64086866e-01
-3.56890172e-01 2.63221741e-01 -2.54381090e-01 1.60320252e-01
5.10796785e-01 3.48329246e-01 -1.25268507e+00 -8.03275049e-01
1.13163710e+00 1.03049195e+00 -4.11332577e-01 7.56656468e-01
5.04752755e-01 4.79069650e-01 -3.92240793e-01 -1.37520969e-01
3.33743900e-01 -3.68887067e-01 -9.14257914e-02 -6.85185045e-02
1.59461886e-01 -1.98392555e-01 1.07513711e-01 1.02018583e+00
3.68479371e-01 3.52261662e-01 1.01165092e+00 -1.26532817e+00
-8.00688922e-01 7.50994861e-01 -6.51304483e-01 6.50604784e-01
1.02588676e-01 5.01593351e-02 1.04590106e+00 -6.83465719e-01
5.28976955e-02 1.25493956e+00 8.93224537e-01 -4.05962676e-01
-1.37419963e+00 -8.29819739e-01 -2.90213168e-01 -2.43222103e-01
-1.26589406e+00 -5.21173954e-01 3.87835026e-01 -5.38974702e-01
5.23839593e-01 1.92122832e-01 2.57618308e-01 9.85277176e-01
7.43619204e-01 9.62890834e-02 1.46835470e+00 -5.00442326e-01
3.89511108e-01 -1.89784691e-01 5.61727285e-01 3.28938887e-02
8.49968791e-01 4.75742191e-01 -7.67310783e-02 -8.91005695e-01
1.33372366e+00 -3.99690643e-02 1.41433239e-01 9.43263993e-02
-8.14059854e-01 1.22959197e+00 -3.19959521e-01 1.69389531e-01
-7.95515060e-01 6.95942119e-02 3.96643937e-01 6.35926664e-01
9.06656921e-01 -1.59001097e-01 -7.01719880e-01 -1.09464884e-01
-9.95012999e-01 4.00751621e-01 9.81304049e-01 7.34895766e-01
4.59683686e-02 6.67724252e-01 -3.26637983e-01 5.10008454e-01
4.27793622e-01 1.26043248e+00 4.00511324e-01 -1.08044040e+00
2.08926097e-01 -2.24101126e-01 5.37822664e-01 -5.69032371e-01
-2.85727233e-01 -6.46939337e-01 -7.28281617e-01 -8.63711536e-02
7.05769718e-01 -5.01190305e-01 -4.23971534e-01 1.66606438e+00
-2.52135154e-02 5.21241724e-01 -5.80965802e-02 3.25997084e-01
-1.13880582e-01 6.55559838e-01 1.54582649e-01 -1.00537968e+00
9.88205314e-01 -4.39874440e-01 -7.83320367e-01 2.50754118e-01
3.55611891e-01 -6.62971616e-01 7.77338982e-01 2.44907632e-01
-1.16473866e+00 -4.76070970e-01 -4.95749712e-01 6.80306137e-01
3.21149319e-01 -4.19692904e-01 3.30792308e-01 8.15072894e-01
-7.73915529e-01 5.95948696e-01 -8.06336403e-01 -6.39365688e-02
-8.20674449e-02 2.14266270e-01 6.72157109e-02 3.55541438e-01
-1.06160533e+00 5.60578048e-01 -1.58931360e-01 7.60919228e-02
-3.97010982e-01 -8.12427461e-01 -5.54825306e-01 4.69255239e-01
2.17795726e-02 -5.63429892e-01 1.21981633e+00 -1.23705077e+00
-1.40916562e+00 2.62946010e-01 -1.42809242e-01 -3.70024055e-01
4.74976540e-01 -6.17035925e-02 -5.86175442e-01 -1.04545489e-01
5.79383731e-01 -6.09639883e-01 7.66050816e-01 -5.91624618e-01
-7.49680459e-01 -3.22190344e-01 -4.73786533e-01 -2.95302719e-01
2.79411376e-01 2.67344147e-01 4.68177617e-01 -9.84117091e-01
-1.78894177e-02 -1.00720191e+00 -1.86674103e-01 -9.66839254e-01
-1.92877769e-01 -3.28144521e-01 1.87566265e-01 -9.52108145e-01
1.41796923e+00 -2.12633681e+00 -5.21375120e-01 8.13138425e-01
-6.44033730e-01 -6.31794214e-01 3.41445133e-02 5.96990526e-01
-4.95571971e-01 2.59545088e-01 -2.98559487e-01 1.57041237e-01
3.65779489e-01 -4.63775285e-02 -6.76216900e-01 1.08914769e+00
-2.37395391e-01 5.81469893e-01 -4.10771370e-01 -4.69749719e-02
6.14109635e-02 -1.09384373e-01 -3.51037771e-01 -1.04967788e-01
5.84321737e-01 3.57562631e-01 -3.67296576e-01 8.55947971e-01
7.15723455e-01 -5.98278344e-02 -6.99899793e-02 4.99528855e-01
-7.27999628e-01 2.96038359e-01 -1.33760679e+00 5.26570022e-01
-9.80080739e-02 5.43209493e-01 -1.02779955e-01 -1.26406848e+00
8.76398563e-01 5.19309163e-01 2.29365245e-01 -6.85324550e-01
-6.61450475e-02 3.90310913e-01 1.54690355e-01 -3.70870084e-01
2.12960422e-01 -6.17288053e-01 -4.18555647e-01 4.84250426e-01
-2.90760100e-01 2.11872578e-01 1.13650590e-01 -4.50389206e-01
9.72536564e-01 1.17945321e-01 8.37907016e-01 -1.04704034e+00
1.94703102e-01 -3.06877285e-01 9.24825609e-01 1.31370735e+00
-2.69202441e-01 1.93961963e-01 8.06700170e-01 5.01128584e-02
-1.11109984e+00 -1.30893838e+00 -9.11095500e-01 1.06493425e+00
-3.70342553e-01 3.48492414e-01 -1.42378137e-01 -8.35565850e-02
6.59831762e-01 1.13381338e+00 -6.60669446e-01 5.42446598e-02
-2.93987483e-01 -8.33499968e-01 4.22155201e-01 6.05521679e-01
1.03590667e-01 -6.84950590e-01 -3.19740355e-01 3.91173810e-01
1.77700445e-01 -5.78242183e-01 -1.88285008e-01 3.37401360e-01
-1.14193165e+00 -9.42334592e-01 -4.30942535e-01 -2.69138336e-01
8.12815428e-02 1.75706372e-01 9.91203487e-01 -4.93362695e-01
5.08708298e-01 7.61624217e-01 -1.03683472e-01 -5.02763271e-01
-4.02074754e-01 -3.00093085e-01 2.48597041e-01 -1.33546457e-01
2.39094481e-01 -8.32343042e-01 -6.78478777e-01 4.35472101e-01
-7.58128107e-01 -8.46769631e-01 2.07254454e-01 8.77009153e-01
4.04330730e-01 1.19683817e-01 1.31735337e+00 -6.29486442e-01
6.31320953e-01 -9.29295480e-01 -1.14579415e+00 2.32052699e-01
-8.48972857e-01 -2.70506799e-01 3.39113325e-01 -6.29185617e-01
-1.16320813e+00 -7.63744235e-01 4.47059453e-01 -1.47758320e-01
-9.12703052e-02 1.21030009e+00 3.40727478e-01 2.91743964e-01
1.58613697e-01 -1.47560716e-01 -9.18985456e-02 -3.10562193e-01
-2.57685006e-01 4.59025502e-01 6.06706440e-01 -7.07747579e-01
8.89707983e-01 2.04622298e-01 2.74271041e-01 -8.50130796e-01
-5.12208521e-01 -5.88501513e-01 -3.76546413e-01 8.66595134e-02
4.57351208e-01 -1.14123857e+00 -7.98375845e-01 2.50331819e-01
-4.33881134e-01 -3.96493196e-01 -4.65673953e-01 9.73966837e-01
-8.35572958e-01 2.47185647e-01 -6.23127997e-01 -1.42000830e+00
-1.11329466e-01 -7.54313350e-01 7.06502020e-01 3.23910713e-01
-3.23896468e-01 -1.48377180e+00 4.97201651e-01 -1.62330922e-02
3.72183383e-01 7.59110153e-01 1.04578388e+00 -8.50870967e-01
-7.99001157e-02 -5.62382638e-01 7.91002437e-02 1.91768959e-01
1.66828349e-01 6.96195066e-01 -5.25524914e-01 -4.48611021e-01
2.38432854e-01 4.47620660e-01 5.17144501e-01 1.38983512e+00
5.71435153e-01 -5.48225760e-01 7.74100944e-02 2.21680373e-01
1.56103003e+00 3.79096329e-01 6.35627687e-01 5.78760803e-01
-2.04443224e-02 4.21610713e-01 7.07633436e-01 9.78545010e-01
1.99277639e-01 2.84776717e-01 7.38053769e-02 1.01574101e-01
8.74379635e-01 6.92995265e-03 6.70243144e-01 7.48189926e-01
-3.28343272e-01 -1.84405103e-01 -8.68960559e-01 5.69986224e-01
-1.90960526e+00 -1.41121614e+00 -7.18502283e-01 2.77831268e+00
6.30143642e-01 1.34274349e-01 5.39898872e-01 -1.06435835e-01
8.31023037e-01 -1.69658870e-01 -3.61214191e-01 -7.61688709e-01
-4.92224813e-01 5.90749502e-01 9.87842441e-01 4.73702043e-01
-1.06861377e+00 4.02869374e-01 7.26691961e+00 5.31534791e-01
-5.73887408e-01 1.12468652e-01 5.59548199e-01 2.68889278e-01
-3.90901744e-01 5.41446030e-01 -6.14475965e-01 5.57710707e-01
1.61374772e+00 -7.77604163e-01 -1.38960145e-02 5.76427877e-01
8.74449372e-01 -5.85432589e-01 -5.02784073e-01 4.42371517e-01
-5.36126375e-01 -7.43303537e-01 -6.69716179e-01 4.72335815e-01
1.15877128e+00 -7.10144341e-02 4.64056849e-01 5.17938495e-01
4.93722707e-01 -9.64480340e-01 7.42627025e-01 8.50093484e-01
3.70833635e-01 -1.17239952e+00 9.18354154e-01 2.81223357e-01
-8.00022364e-01 -4.94311303e-01 -5.08089185e-01 -4.26593035e-01
-1.65864918e-02 8.94553602e-01 -3.34101170e-02 3.69501293e-01
6.53271616e-01 4.29372728e-01 -3.24992418e-01 1.30422556e+00
1.94433834e-02 1.53388417e+00 -3.46771330e-01 5.05909979e-01
2.06373617e-01 -8.53649914e-01 6.18846774e-01 1.01399302e+00
9.00059998e-01 1.72811225e-01 -1.79499641e-01 4.97092128e-01
5.50590336e-01 1.35534033e-01 -6.83450222e-01 1.74919933e-01
6.76021039e-01 6.66395485e-01 -6.49977207e-01 -1.77055284e-01
-8.11249435e-01 1.91704944e-01 -4.74886656e-01 7.77530551e-01
-9.32721555e-01 -1.50930658e-01 5.42761624e-01 1.48094356e-01
5.34940422e-01 -3.61660063e-01 -5.10512888e-01 -1.10969973e+00
-3.30949426e-01 -8.69651914e-01 7.45472610e-01 -3.89252931e-01
-1.53666079e+00 -5.63368738e-01 4.80904132e-01 -1.14466524e+00
-6.85424268e-01 -3.45815271e-01 -1.07468522e+00 1.03444290e+00
-1.29045582e+00 -5.55383861e-01 3.93225282e-01 3.49579722e-01
2.83335000e-01 4.35808999e-03 5.11903822e-01 -1.23773433e-01
-7.97364950e-01 3.24035168e-01 1.21880555e+00 -3.13034475e-01
8.03680062e-01 -1.26603103e+00 2.05074444e-01 8.10910642e-01
-4.23048198e-01 8.13989520e-01 1.19001317e+00 -8.84838343e-01
-1.22198105e+00 -7.09840357e-01 7.60500193e-01 -1.88648701e-01
1.22241116e+00 2.63432086e-01 -9.19855595e-01 1.16469669e+00
2.94154316e-01 -3.88162285e-01 1.05155027e+00 3.42205882e-01
-1.08086407e-01 -1.43871054e-01 -8.22655022e-01 3.93896699e-01
3.74403358e-01 -2.96606123e-01 -6.07767582e-01 2.52833039e-01
2.08379060e-01 3.22808594e-01 -1.32671499e+00 6.17902160e-01
9.50287223e-01 -1.09513998e+00 6.41709387e-01 -6.19879842e-01
7.13305324e-02 9.84262601e-02 -4.91390854e-01 -9.76380885e-01
-6.92021966e-01 -7.52115905e-01 3.46110731e-01 1.51041889e+00
1.63158417e-01 -1.22623575e+00 -8.01593065e-02 4.32463378e-01
2.50595719e-01 9.74496379e-02 -1.53980911e+00 -1.23192716e+00
5.80228686e-01 -4.29694563e-01 4.54052627e-01 1.10383856e+00
-1.19259998e-01 9.67788510e-03 -2.16533706e-01 3.31072360e-01
1.04871857e+00 5.37247062e-01 8.40335965e-01 -1.73834419e+00
-4.22126114e-01 -3.55569333e-01 -1.27750829e-01 -5.54140270e-01
6.18410945e-01 -3.49732637e-01 -2.23004982e-01 -7.42982090e-01
2.84862846e-01 -1.83397338e-01 -5.87388039e-01 5.28639965e-02
-6.00287840e-02 -1.07555062e-01 -3.65828425e-01 3.29008043e-01
5.94648905e-02 4.74023610e-01 5.98406017e-01 2.38655686e-01
-4.06457543e-01 6.42755032e-01 -7.88953722e-01 5.61977506e-01
8.61308634e-01 -6.74688160e-01 -3.88950825e-01 5.35370827e-01
4.42309707e-01 8.99575114e-01 7.46602833e-01 -4.03338492e-01
-2.45424807e-01 -7.96225011e-01 2.26418301e-01 -7.39205420e-01
-5.52126765e-01 -6.00140691e-01 5.68917871e-01 4.31242496e-01
-1.87753335e-01 5.44554532e-01 3.02196592e-01 7.03682959e-01
-1.82374492e-01 -2.36722082e-01 6.97517097e-01 5.52854061e-01
1.28254980e-01 1.16394004e-02 -6.61226273e-01 9.52534974e-02
7.48480916e-01 -3.22681963e-01 -4.09342170e-01 -9.07921195e-01
-5.68444788e-01 1.25780618e-02 2.14825660e-01 -1.14534132e-01
4.62827794e-02 -1.36864471e+00 -1.02107131e+00 4.50443327e-02
-3.34830910e-01 -8.62196445e-01 1.26025289e-01 1.55380607e+00
-1.18668601e-01 6.41152263e-01 2.86786947e-02 -5.03280997e-01
-7.59409606e-01 3.85366261e-01 1.42843470e-01 -6.74730912e-02
-3.39069605e-01 2.98484461e-03 4.18604106e-01 1.44309863e-01
-3.83774698e-01 -4.39493060e-01 -4.41758595e-02 3.60496461e-01
2.56596565e-01 8.78240824e-01 -3.03686827e-01 -5.69822550e-01
-3.02015662e-01 4.68202859e-01 5.75657010e-01 -3.01937401e-01
1.44457626e+00 -5.87650061e-01 -3.19103897e-01 1.06132400e+00
9.52991545e-01 3.01586330e-01 -1.24025393e+00 -1.58366024e-01
4.59878653e-01 -2.39267021e-01 -7.90102631e-02 -2.16364637e-01
-5.26782930e-01 1.91003710e-01 5.02671719e-01 7.36045361e-01
9.21041965e-01 -4.00848866e-01 1.57789841e-01 -1.30451750e-03
-6.97756261e-02 -1.00577080e+00 -6.46526873e-01 5.00820041e-01
7.69161463e-01 -9.57545638e-01 1.59509853e-01 4.18123789e-02
-3.45858157e-01 1.26798713e+00 -2.46553257e-01 -6.12224579e-01
1.11440885e+00 1.68001384e-01 -1.85940549e-01 2.85985887e-01
-8.63438725e-01 -4.13118452e-02 2.83607632e-01 4.53764558e-01
4.10329103e-01 3.76275718e-01 -9.29325223e-01 7.45320439e-01
-2.57541329e-01 -4.37516451e-01 8.07455540e-01 6.92436755e-01
-3.56883526e-01 -5.49530387e-01 -7.18199909e-01 4.42444503e-01
-9.83146667e-01 -1.52376994e-01 3.38689178e-01 1.38632989e+00
-6.47955418e-01 1.18428850e+00 3.49184990e-01 5.66164196e-01
3.48837972e-01 3.73215646e-01 3.04145701e-02 -2.16996342e-01
-4.60601062e-01 8.95458460e-01 3.14623602e-02 2.61958153e-03
-5.99204659e-01 -1.38822973e+00 -6.94984794e-01 -8.15766692e-01
-6.99093878e-01 2.30140284e-01 1.98837072e-01 1.13760173e+00
-1.11097127e-01 8.06780905e-02 1.26160634e+00 -5.05475283e-01
-1.09354019e+00 -1.04186463e+00 -1.27457011e+00 -4.28902619e-02
4.33371097e-01 -6.49659157e-01 -1.00915396e+00 -1.99041963e-01] | [6.369168758392334, 4.201462268829346] |
ae34e19d-1e2e-4e3c-b359-34a6bf8a07b5 | planning-in-stochastic-environments-with-a | null | null | https://openreview.net/forum?id=X6D9bAHhBQ1 | https://openreview.net/pdf?id=X6D9bAHhBQ1 | Planning in Stochastic Environments with a Learned Model | Model-based reinforcement learning has proven highly successful. However, learning a model in isolation from its use during planning is problematic in complex environments. To date, the most effective techniques have instead combined value-equivalent model learning with powerful tree-search methods. This approach is exemplified by MuZero, which has achieved state-of-the-art performance in a wide range of domains, from board games to visually rich environments, with discrete and continuous action spaces, in online and offline settings. However, previous instantiations of this approach were limited to the use of deterministic models. This limits their performance in environments that are inherently stochastic, partially observed, or so large and complex that they appear stochastic to a finite agent. In this paper we extend this approach to learn and plan with stochastic models. Specifically, we introduce a new algorithm, Stochastic MuZero, that learns a stochastic model incorporating afterstates, and uses this model to perform a stochastic tree search. Stochastic MuZero matched or exceeded the state of the art in a set of canonical single and multi-agent environments, including 2048 and backgammon, while maintaining the same performance as standard MuZero in the game of Go. | ['David Silver', 'Thomas K Hubert', 'Sherjil Ozair', 'Julian Schrittwieser', 'Ioannis Antonoglou'] | 2021-09-29 | null | null | null | iclr-2022-4 | ['game-of-go', 'board-games', '2048'] | ['playing-games', 'playing-games', 'playing-games'] | [ 6.41980991e-02 1.38891011e-01 -1.89738184e-01 1.34741277e-01
-8.23646903e-01 -6.82660341e-01 8.60418737e-01 1.20835572e-01
-7.25804627e-01 1.27423310e+00 -1.36780873e-01 -3.88459861e-01
-3.94229323e-01 -8.48976731e-01 -5.06939709e-01 -6.14634395e-01
-4.92827207e-01 1.02890074e+00 4.93751943e-01 -4.29888666e-01
3.27020764e-01 1.69410229e-01 -1.70217001e+00 -3.77147533e-02
6.51676893e-01 6.18615866e-01 5.20009100e-01 8.90896440e-01
-8.40891339e-03 1.07059526e+00 -6.82962060e-01 1.24925926e-01
3.55279803e-01 -5.37818432e-01 -6.97735965e-01 8.17327052e-02
-2.62915254e-01 -3.58992040e-01 -1.95284456e-01 6.41763330e-01
5.82955301e-01 2.81188130e-01 3.17370385e-01 -1.21642804e+00
9.70508084e-02 6.24973297e-01 -2.98907578e-01 -8.46453309e-02
6.30704522e-01 5.47547460e-01 7.81638145e-01 -1.45594060e-01
6.90011442e-01 1.39129639e+00 4.45320249e-01 7.46450782e-01
-1.51001239e+00 -4.30198282e-01 3.54696542e-01 3.22814256e-01
-1.17763567e+00 -2.68612772e-01 3.87370706e-01 -3.52635354e-01
1.47886360e+00 2.95084976e-02 1.05160844e+00 1.04151392e+00
3.57693225e-01 8.49771380e-01 1.60378504e+00 -7.01998115e-01
8.50686073e-01 -2.24109307e-01 -5.19269228e-01 7.60469735e-01
7.43338168e-02 8.03107083e-01 -5.35075665e-01 -2.05779821e-01
7.73476779e-01 -4.01441544e-01 1.38292447e-01 -7.91710258e-01
-1.05055118e+00 9.45888758e-01 -1.56110883e-01 -2.70547569e-02
-4.42635059e-01 5.42879760e-01 2.57565290e-01 4.99569148e-01
-1.45335197e-02 7.37942874e-01 -4.62117642e-01 -7.13489890e-01
-7.29875147e-01 8.55284929e-01 1.17982280e+00 8.33793521e-01
8.13561797e-01 1.65490061e-01 1.59823492e-01 2.91825026e-01
2.69095778e-01 2.81529874e-01 4.29035127e-01 -1.40289378e+00
2.34073386e-01 1.43499836e-01 6.08055651e-01 -3.95762026e-01
-6.68731749e-01 -2.78737456e-01 -4.00344491e-01 1.02806473e+00
5.87838233e-01 -3.68951350e-01 -8.75822842e-01 1.76675248e+00
4.66011465e-01 1.20784797e-01 3.26643467e-01 5.24289131e-01
1.06390983e-01 4.78295237e-01 -1.86605498e-01 -2.86293566e-01
7.98349261e-01 -1.09291708e+00 -5.28240442e-01 -5.42047679e-01
6.72774792e-01 -2.95233339e-01 1.19241595e+00 8.39774966e-01
-1.09687400e+00 -7.19884261e-02 -1.05528343e+00 6.20276749e-01
-2.09646240e-01 -5.70079446e-01 9.21958387e-01 8.47935438e-01
-1.23094690e+00 8.90324533e-01 -1.16036975e+00 -3.61243755e-01
1.47902384e-01 5.54775894e-01 -1.28481269e-01 -1.06327735e-01
-1.03025985e+00 1.18744099e+00 5.24814546e-01 -3.62354070e-01
-1.45165884e+00 -2.51750857e-01 -8.48688185e-01 -9.36882272e-02
1.12304306e+00 -6.07670784e-01 1.87875378e+00 -7.82759488e-01
-2.16838956e+00 3.26376349e-01 1.49272025e-01 -8.55082989e-01
7.71634817e-01 1.18224816e-02 2.01002788e-02 -2.22528294e-01
-3.02510019e-02 4.87722516e-01 5.47698379e-01 -1.41629064e+00
-8.44582915e-01 -1.67292014e-01 6.34345770e-01 4.08650070e-01
1.86578378e-01 -6.71416968e-02 -7.07068145e-02 -1.04570791e-01
-2.78708249e-01 -1.01973796e+00 -8.51668954e-01 -3.69979590e-01
1.14482129e-02 -1.16165049e-01 3.49523753e-01 -8.51906240e-02
1.18364954e+00 -1.78219163e+00 2.29893818e-01 1.49389759e-01
-1.58305056e-02 -1.93156555e-01 -1.55253097e-01 1.04155099e+00
4.26819295e-01 -2.85720825e-03 -1.27218410e-01 -4.76460457e-01
3.80792439e-01 6.80621028e-01 1.29930079e-01 2.48142898e-01
-2.87252873e-01 6.93998873e-01 -1.23257446e+00 -5.66674590e-01
3.14376056e-01 -2.11648956e-01 -7.90269017e-01 8.00624490e-02
-7.18450904e-01 3.58741194e-01 -4.21270728e-01 4.64266479e-01
1.30570799e-01 6.38171136e-02 5.22432208e-01 9.67122376e-01
-3.79822910e-01 1.17217168e-01 -1.53719747e+00 1.80341554e+00
-6.36335313e-01 3.36813182e-01 1.65193424e-01 -7.57223189e-01
6.75295293e-01 2.18274787e-01 6.29116118e-01 -6.20608687e-01
-1.46248817e-01 3.06323528e-01 2.62080044e-01 -4.84286159e-01
4.74380881e-01 -3.85214150e-01 -2.11270660e-01 4.94942695e-01
-1.66035011e-01 -8.45130086e-01 5.95204055e-01 -1.41480431e-01
1.39840305e+00 6.37449861e-01 5.82115829e-01 -1.50223553e-01
-2.75991019e-02 3.41505229e-01 5.92478216e-01 1.42598295e+00
-2.61967868e-01 3.21861118e-01 6.29884779e-01 -4.68844652e-01
-8.43815565e-01 -8.81444633e-01 4.26225692e-01 1.05035722e+00
2.76131988e-01 -7.07496345e-01 -6.91806853e-01 -4.15283471e-01
-9.30457413e-02 1.12247825e+00 -6.34215415e-01 -4.26543169e-02
-5.62967956e-01 -4.46627438e-01 3.10774773e-01 1.98728427e-01
3.99038523e-01 -1.35938680e+00 -1.33609331e+00 6.76849604e-01
1.46043524e-01 -7.86099017e-01 1.50230364e-03 5.67550302e-01
-7.03171492e-01 -1.09414482e+00 -2.98494816e-01 -3.93675625e-01
7.33660907e-02 -1.01361543e-01 1.23260629e+00 -1.65868640e-01
-1.42045289e-01 5.97378433e-01 -3.89168322e-01 -6.76115513e-01
-5.52024662e-01 4.71310429e-02 1.44598171e-01 -4.32784468e-01
-6.46071061e-02 -6.12201154e-01 -1.55562520e-01 5.48647270e-02
-7.59271145e-01 1.80982649e-01 5.11360824e-01 1.17877173e+00
4.08454418e-01 3.41043800e-01 4.14598733e-01 -6.74078941e-01
8.25151622e-01 -2.18416691e-01 -9.71901059e-01 1.26085728e-01
-6.25103116e-01 1.42111316e-01 6.04023337e-01 -4.73107904e-01
-8.73288929e-01 2.09407806e-01 1.39084747e-02 -1.56372964e-01
-2.86496937e-01 7.92009950e-01 2.03620046e-02 -7.62570873e-02
6.79902434e-01 3.76397461e-01 1.03956267e-01 -1.85242191e-01
4.33915630e-02 2.52488971e-01 5.36392517e-02 -8.77241969e-01
5.69141269e-01 1.26066431e-01 1.51094794e-01 -6.71186984e-01
-2.49310195e-01 -2.29404625e-02 -1.96539223e-01 -2.72181064e-01
5.81327140e-01 -4.85760510e-01 -8.89940560e-01 6.08977258e-01
-7.56537557e-01 -1.38704324e+00 -7.25971878e-01 4.99199867e-01
-1.32821238e+00 2.03971952e-01 -3.07108849e-01 -1.23784316e+00
4.11736876e-01 -1.25824714e+00 7.42694855e-01 2.71634579e-01
-6.85073063e-02 -8.90481830e-01 3.65592033e-01 -2.03567192e-01
4.93024349e-01 2.70445824e-01 9.55438435e-01 -5.06752312e-01
-6.88368797e-01 -9.62957516e-02 5.93431711e-01 -2.19801992e-01
-2.94671822e-02 -1.50798649e-01 -6.11448884e-01 -4.27188843e-01
-2.32086569e-01 -6.89999878e-01 4.31557059e-01 4.71337676e-01
8.77117097e-01 -1.02518164e-01 -2.61247784e-01 2.60504156e-01
1.48860800e+00 6.91882253e-01 5.04479825e-01 1.08316326e+00
-1.19067580e-01 5.09279013e-01 8.08808625e-01 9.33622777e-01
5.99661291e-01 9.36208844e-01 9.18202996e-01 1.99192330e-01
2.84649760e-01 -3.00257176e-01 4.56436932e-01 1.03168420e-01
-2.50832289e-01 -3.05609584e-01 -1.09661555e+00 4.21360224e-01
-2.21648908e+00 -1.18370295e+00 4.87986416e-01 2.31800675e+00
7.68299937e-01 5.58550715e-01 4.06264365e-01 -1.02579873e-02
2.49833018e-01 1.55302674e-01 -6.24399364e-01 -5.07876575e-01
4.62361537e-02 1.97372869e-01 4.54304636e-01 7.26187587e-01
-8.60849559e-01 1.17917979e+00 7.01390934e+00 8.51960838e-01
-7.43430257e-01 9.32658687e-02 2.56217539e-01 -3.02143931e-01
-1.59405068e-01 3.24282080e-01 -5.02709508e-01 2.37858981e-01
9.21390176e-01 -3.36557895e-01 1.05447090e+00 9.53643024e-01
4.82351363e-01 -7.23540902e-01 -1.17695594e+00 8.37904632e-01
-3.98294389e-01 -1.27282667e+00 -5.11575997e-01 2.88972437e-01
8.37667882e-01 4.10885103e-02 -1.15693705e-02 8.37323427e-01
1.05183733e+00 -1.26379240e+00 9.78684962e-01 3.63568008e-01
4.32218283e-01 -8.93410027e-01 5.99291623e-01 9.74897921e-01
-9.19446111e-01 -4.79081094e-01 -9.51935276e-02 -7.14405894e-01
1.59233093e-01 -1.48570016e-01 -7.97775328e-01 4.52860773e-01
6.01848722e-01 1.86059818e-01 -1.69839814e-01 1.36460578e+00
-1.96383938e-01 4.07116205e-01 -5.47001004e-01 -4.58219171e-01
6.74204230e-01 -1.37250230e-01 5.16344488e-01 7.87336171e-01
2.25409791e-01 4.38255928e-02 8.20014536e-01 3.35936159e-01
5.38723469e-01 -1.67873979e-01 -8.44042361e-01 1.62555248e-01
5.57139337e-01 7.32088268e-01 -7.16277838e-01 -2.21000463e-01
-2.47357070e-01 6.19987190e-01 3.09078813e-01 3.58605057e-01
-7.59249926e-01 -1.78999528e-01 5.46289086e-01 -7.70742726e-03
2.99688190e-01 -6.25512242e-01 -6.73250109e-02 -8.43810797e-01
-1.74829543e-01 -1.25966418e+00 3.08789849e-01 -4.21156496e-01
-6.35855198e-01 5.07572711e-01 4.10445035e-01 -1.12063718e+00
-9.24105823e-01 -3.22688967e-01 -4.37204808e-01 5.61400414e-01
-1.39888597e+00 -8.77431095e-01 -1.48723930e-01 5.12938023e-01
9.41168368e-01 -3.87038946e-01 1.16021454e+00 -2.16520682e-01
-2.50712425e-01 2.10332528e-01 4.18289363e-01 -6.45478070e-01
3.38110000e-01 -1.65253377e+00 3.62350494e-01 5.66300571e-01
4.63721119e-02 1.48010835e-01 1.01345849e+00 -5.26875794e-01
-1.66540456e+00 -5.49679935e-01 3.28836113e-01 -1.51111931e-01
6.95797503e-01 -3.76874506e-01 -2.99906015e-01 6.67140484e-01
2.68171191e-01 -1.70249775e-01 2.75683999e-01 1.55361161e-01
2.11628929e-01 1.62467107e-01 -1.11882699e+00 9.79487181e-01
1.10656738e+00 -1.68702871e-01 -3.05410266e-01 2.81106293e-01
4.85461324e-01 -7.09585607e-01 -3.77241492e-01 1.18955657e-01
5.76935530e-01 -1.27565944e+00 6.30293310e-01 -6.45203948e-01
1.73567176e-01 -3.19019169e-01 -2.30008811e-01 -1.73744345e+00
-1.95383027e-01 -1.13950324e+00 -2.93606371e-01 5.98192453e-01
2.80455470e-01 -7.72705853e-01 1.07792032e+00 4.41287667e-01
-1.96744591e-01 -9.31287944e-01 -1.23188865e+00 -1.14210474e+00
1.63809404e-01 -4.29277301e-01 7.09216714e-01 3.19856912e-01
1.20342992e-01 -8.38145539e-02 -5.49558640e-01 -8.68202820e-02
6.18583083e-01 5.45374788e-02 1.02703476e+00 -9.37911332e-01
-9.10998404e-01 -5.40452361e-01 -3.08150351e-01 -1.08309042e+00
1.66955903e-01 -3.48858416e-01 4.34579581e-01 -1.72082651e+00
-1.31170884e-01 -6.73325539e-01 -3.93049978e-02 3.29291493e-01
2.22995028e-01 -4.23535168e-01 5.10074019e-01 -4.49189134e-02
-7.35124648e-01 6.81121886e-01 1.25176680e+00 -1.46298289e-01
-5.00423551e-01 2.98376292e-01 -3.11379254e-01 8.58048081e-01
9.18205798e-01 -4.10097152e-01 -6.47119820e-01 -3.50922465e-01
1.90426499e-01 5.23893356e-01 5.89112379e-02 -1.32942879e+00
3.49875033e-01 -8.60157907e-01 -1.06636308e-01 -1.84343129e-01
5.60826957e-01 -7.94529140e-01 2.99782336e-01 8.22048903e-01
-4.11076546e-01 2.74803787e-01 2.43356928e-01 6.69822216e-01
-1.28900170e-01 -5.82779646e-01 4.74730790e-01 -6.36590123e-01
-1.07520807e+00 5.47784604e-02 -8.77024710e-01 7.39396736e-02
1.38247752e+00 -4.93539006e-01 2.14749165e-02 -7.77623117e-01
-9.85076368e-01 3.86490673e-01 6.56466961e-01 5.58230951e-02
5.30304313e-01 -7.93444455e-01 -6.18454933e-01 -3.22734267e-02
-6.54183030e-02 3.69339213e-02 7.03817755e-02 5.38316965e-01
-5.02305984e-01 2.08938047e-01 -3.51191133e-01 -3.27710837e-01
-1.12884021e+00 5.53965569e-01 6.08438373e-01 -8.50401103e-01
-5.09452641e-01 4.75023150e-01 -1.19778395e-01 -5.50724447e-01
3.24588746e-01 -2.09254593e-01 -5.17948829e-02 -2.54596174e-01
3.16928267e-01 2.52783030e-01 -2.22069308e-01 1.11104734e-01
-2.15658218e-01 2.54600614e-01 1.44468009e-01 -7.34441400e-01
1.40524566e+00 -4.70564328e-02 3.66835266e-01 6.06367528e-01
1.55534804e-01 -1.41832769e-01 -1.80329299e+00 -4.23311293e-02
2.56583214e-01 -4.71976489e-01 -1.56027749e-01 -9.82026219e-01
-3.62224907e-01 6.47151530e-01 4.77648914e-01 3.98158342e-01
9.09500420e-01 -2.97681063e-01 2.33868241e-01 8.91211450e-01
1.34527302e+00 -1.31164408e+00 2.23716900e-01 9.53780830e-01
6.34265602e-01 -1.15912867e+00 -1.07138157e-01 8.80128220e-02
-7.83740699e-01 1.03623235e+00 7.57386565e-01 -6.01775572e-02
1.82869598e-01 5.32488167e-01 -1.38563782e-01 1.20707594e-01
-1.15791547e+00 -5.17888248e-01 -6.80072486e-01 1.13072789e+00
-2.17751965e-01 2.07160711e-01 -4.75549512e-02 2.95195907e-01
-2.34086886e-01 1.98710710e-01 9.52109873e-01 1.57527494e+00
-6.56986654e-01 -1.37185788e+00 -3.95523638e-01 2.61227399e-01
-2.13873684e-01 8.16706717e-02 -1.45853266e-01 1.19232976e+00
-1.11627020e-01 1.12046695e+00 -1.75274327e-01 -2.97111988e-01
3.02152425e-01 -8.69009569e-02 8.67158651e-01 -7.36475289e-01
-7.09140062e-01 3.36266793e-02 5.06791174e-01 -7.59604692e-01
-2.77294934e-01 -8.48949730e-01 -1.19927955e+00 -1.86914697e-01
-1.01119146e-01 2.79169768e-01 5.48480093e-01 9.35028553e-01
-1.49166714e-02 4.71177638e-01 4.96192306e-01 -1.20521259e+00
-1.18935943e+00 -6.95338249e-01 -7.83104479e-01 -2.56383628e-01
2.87631124e-01 -9.72612619e-01 1.03978086e-02 -4.11883891e-01] | [3.921502113342285, 1.6609264612197876] |
bc44cb0c-a428-4078-b50d-731035b304a2 | communication-efficient-robust-federated-1 | 2206.05558 | null | https://arxiv.org/abs/2206.05558v1 | https://arxiv.org/pdf/2206.05558v1.pdf | Communication-Efficient Robust Federated Learning with Noisy Labels | Federated learning (FL) is a promising privacy-preserving machine learning paradigm over distributed located data. In FL, the data is kept locally by each user. This protects the user privacy, but also makes the server difficult to verify data quality, especially if the data are correctly labeled. Training with corrupted labels is harmful to the federated learning task; however, little attention has been paid to FL in the case of label noise. In this paper, we focus on this problem and propose a learning-based reweighting approach to mitigate the effect of noisy labels in FL. More precisely, we tuned a weight for each training sample such that the learned model has optimal generalization performance over a validation set. More formally, the process can be formulated as a Federated Bilevel Optimization problem. Bilevel optimization problem is a type of optimization problem with two levels of entangled problems. The non-distributed bilevel problems have witnessed notable progress recently with new efficient algorithms. However, solving bilevel optimization problems under the Federated Learning setting is under-investigated. We identify that the high communication cost in hypergradient evaluation is the major bottleneck. So we propose \textit{Comm-FedBiO} to solve the general Federated Bilevel Optimization problems; more specifically, we propose two communication-efficient subroutines to estimate the hypergradient. Convergence analysis of the proposed algorithms is also provided. Finally, we apply the proposed algorithms to solve the noisy label problem. Our approach has shown superior performance on several real-world datasets compared to various baselines. | ['Heng Huang', 'Jian Pei', 'Junyi Li'] | 2022-06-11 | null | null | null | null | ['learning-with-noisy-labels', 'learning-with-noisy-labels'] | ['computer-vision', 'natural-language-processing'] | [-1.22632131e-01 -8.15136209e-02 -2.99415171e-01 -5.40233552e-01
-1.15572608e+00 -9.10034001e-01 2.43731260e-01 2.04469115e-01
-4.65941548e-01 8.25724125e-01 6.76123286e-03 -3.35852265e-01
-2.28507996e-01 -5.93992829e-01 -7.95567334e-01 -1.28312349e+00
-5.01687564e-02 2.18808651e-01 -4.13480908e-01 3.02645862e-01
-9.76496562e-02 4.09848154e-01 -1.05187356e+00 1.59018740e-01
7.20359564e-01 1.15059733e+00 -2.31447503e-01 4.85039622e-01
-7.13414103e-02 7.98041224e-01 -5.19525766e-01 -7.26384282e-01
7.81669319e-01 -2.88529456e-01 -9.57376182e-01 -1.61932692e-01
4.80263948e-01 -3.00506443e-01 -1.54440999e-01 1.46025181e+00
5.92157483e-01 1.04202721e-02 5.94076551e-02 -1.58577406e+00
-3.69673908e-01 6.66919112e-01 -3.74444991e-01 -1.23342015e-01
-9.85670611e-02 6.69478625e-02 1.07533216e+00 -7.10910201e-01
5.30400038e-01 8.25248718e-01 5.99053085e-01 5.37747324e-01
-1.06954420e+00 -6.81262434e-01 8.10476243e-02 2.28302300e-01
-1.26594734e+00 -4.81497765e-01 5.86132109e-01 -1.73684672e-01
2.35983893e-01 8.10720980e-01 3.13307643e-01 7.71015406e-01
-1.00490805e-02 8.47578824e-01 1.46321249e+00 -1.88936770e-01
5.22951484e-01 3.58861834e-01 3.07881385e-01 7.71463752e-01
4.39789891e-01 1.20963857e-01 -6.88347161e-01 -8.37099552e-01
-6.91557229e-02 4.04389054e-01 -5.71013510e-01 -6.55091584e-01
-1.07660007e+00 8.26553226e-01 4.67513800e-01 -1.30678434e-02
-1.76629528e-01 1.40626222e-01 6.12507522e-01 5.60077608e-01
6.55768275e-01 1.53619155e-01 -7.58890927e-01 1.72805846e-01
-7.43422806e-01 1.83249831e-01 1.21553075e+00 7.92063951e-01
9.72621322e-01 -3.77807468e-01 -1.75270125e-01 6.26599550e-01
3.72652173e-01 3.70284736e-01 2.01713145e-01 -1.04135656e+00
4.74620759e-01 4.75986451e-01 9.17355716e-02 -1.03339303e+00
-1.76606521e-01 -3.77770931e-01 -1.01564586e+00 1.02370434e-01
5.95209658e-01 -3.83824527e-01 -1.47917390e-01 1.76291943e+00
1.06167471e+00 2.66355008e-01 1.73866123e-01 1.31463540e+00
3.81887704e-01 6.22207761e-01 -1.00259237e-01 -5.27994275e-01
1.36602914e+00 -1.18073547e+00 -8.60874176e-01 3.46831650e-01
8.72803450e-01 -5.43509364e-01 7.33630121e-01 5.00870228e-01
-6.94755137e-01 4.18037295e-01 -8.19706380e-01 -1.08513974e-01
-4.25863624e-01 -2.17241809e-01 8.16420019e-01 1.01584053e+00
-8.52123022e-01 5.26344597e-01 -5.90140760e-01 -1.20230131e-01
6.42986119e-01 4.63336408e-01 -6.15401804e-01 -3.73005122e-01
-9.52226341e-01 4.06719536e-01 1.02068387e-01 8.01195949e-02
-1.14723110e+00 -6.58315659e-01 -6.41522467e-01 1.49440917e-03
5.90069652e-01 -7.01532543e-01 1.29768264e+00 -6.25738084e-01
-1.32349300e+00 9.07657266e-01 -1.41291125e-02 -4.11784858e-01
7.75238991e-01 5.13398759e-02 -1.38619885e-01 -1.78195700e-01
-1.76389068e-01 -3.27490568e-01 8.14732015e-01 -1.32968509e+00
-9.31525409e-01 -8.31935406e-01 -7.27160051e-02 1.31253630e-01
-8.00941348e-01 8.49397629e-02 -3.38776231e-01 -4.73048180e-01
-1.78103328e-01 -8.90589476e-01 -3.20445865e-01 2.57513553e-01
-3.85938883e-01 -2.57634908e-01 1.00183630e+00 -5.32340229e-01
1.09676468e+00 -2.23460507e+00 -1.70118004e-01 3.73670518e-01
5.46505868e-01 1.77309543e-01 5.25633655e-02 3.67721617e-01
2.91799635e-01 2.44510219e-01 -3.61341834e-01 -6.23400450e-01
2.61187404e-01 3.06109846e-01 -3.77046764e-01 1.19317007e+00
-7.45882094e-01 7.18882203e-01 -8.77900243e-01 -5.67411780e-01
-2.70076692e-01 2.86394268e-01 -4.28964853e-01 4.35379177e-01
-2.49866933e-01 4.08353508e-01 -6.58174276e-01 8.36048365e-01
9.36349154e-01 -3.55694473e-01 3.75423014e-01 -1.34771854e-01
-2.23843586e-02 -9.21561643e-02 -1.27491260e+00 1.75086975e+00
-4.67472583e-01 6.88856691e-02 8.72267842e-01 -9.35447276e-01
6.96394920e-01 4.74266678e-01 7.16082513e-01 -2.51795501e-01
1.85259476e-01 3.79954308e-01 -5.85131466e-01 -6.26072049e-01
1.94151580e-01 -1.79779530e-02 -1.54315978e-01 9.08739269e-01
-2.31059223e-01 3.29922616e-01 -3.56397778e-01 1.81171104e-01
1.23801923e+00 -2.76144832e-01 1.94660842e-01 -2.38681406e-01
5.63513041e-01 6.93785846e-02 9.58793998e-01 7.64379919e-01
-4.81973201e-01 2.13577271e-01 3.03188711e-01 -6.91678047e-01
-6.17862463e-01 -5.95032036e-01 -8.04294720e-02 1.38355744e+00
3.96659374e-01 -6.08343661e-01 -7.94640124e-01 -1.31755865e+00
3.20559084e-01 3.60123605e-01 -3.11098337e-01 6.96087331e-02
-4.34069574e-01 -1.08125174e+00 5.87823451e-01 -2.67076045e-01
5.53800762e-01 -5.92638969e-01 -4.44599897e-01 -5.09486459e-02
-2.22856328e-01 -6.60087764e-01 -8.64397347e-01 3.79801303e-01
-5.76404393e-01 -1.38277650e+00 -3.32926035e-01 -6.46054626e-01
7.43249416e-01 3.34403932e-01 6.23746634e-01 7.97584727e-02
-2.36453429e-01 2.75018960e-01 -1.86776772e-01 -2.11561635e-01
-1.29301012e-01 4.26075608e-02 7.12635741e-02 6.63426816e-01
2.01890215e-01 -2.68313110e-01 -7.61477947e-01 4.80067760e-01
-1.03532410e+00 -3.58898640e-01 1.45947099e-01 1.06063366e+00
8.03958297e-01 -6.70366874e-03 4.65205848e-01 -1.31974840e+00
7.08545268e-01 -7.61906087e-01 -7.69677520e-01 4.59865749e-01
-9.11216617e-01 1.61702707e-01 1.10963500e+00 -2.51369953e-01
-9.30003643e-01 2.62189090e-01 8.92891735e-02 -4.34081554e-01
2.59534921e-02 2.79017448e-01 -4.66869235e-01 -4.66493577e-01
3.62778097e-01 1.14308521e-01 1.96854576e-01 -7.85290658e-01
4.99373108e-01 1.11806333e+00 3.08657765e-01 -9.11926091e-01
5.21747887e-01 7.26690829e-01 1.36928298e-02 -1.81251943e-01
-8.60273838e-01 -5.06556153e-01 8.01379979e-02 -1.29922163e-02
3.74816686e-01 -7.73169994e-01 -1.31139553e+00 2.96474725e-01
-1.03730822e+00 -2.27148682e-01 -4.99922782e-01 2.62383401e-01
-4.83764887e-01 4.82736379e-01 -5.80248773e-01 -7.96798289e-01
-8.32057655e-01 -1.23750579e+00 7.51747429e-01 -1.48757538e-02
3.03318024e-01 -9.17601824e-01 2.95233876e-01 6.27141893e-01
5.60498416e-01 4.26477164e-01 7.14317083e-01 -1.01378036e+00
-6.95322752e-01 -2.97263741e-01 -1.08366080e-01 2.42317051e-01
8.15743208e-02 -6.47559345e-01 -1.07916903e+00 -8.28254342e-01
3.74547452e-01 -5.27176082e-01 5.04817188e-01 -2.99846083e-01
1.63543952e+00 -1.05525374e+00 -2.28997603e-01 1.17216063e+00
1.42348468e+00 -3.40134978e-01 -1.16409354e-01 2.91009724e-01
5.55001855e-01 4.99062061e-01 5.70826173e-01 8.65164876e-01
5.05590379e-01 4.65874761e-01 7.94771075e-01 3.05595007e-02
2.24040329e-01 -2.88629591e-01 2.21055180e-01 8.11091483e-01
5.17820001e-01 -1.78524494e-01 -5.99025905e-01 4.15240020e-01
-2.12628794e+00 -6.59297824e-01 -2.71486640e-01 2.33131695e+00
1.01403439e+00 -8.50133359e-01 -9.22484174e-02 -1.17375895e-01
6.59849524e-01 2.83195138e-01 -8.57407153e-01 -3.86701763e-01
-1.71760738e-01 -1.20975211e-01 8.63513827e-01 2.50084728e-01
-1.07745266e+00 7.11188138e-01 5.41106081e+00 8.76049399e-01
-1.13221490e+00 7.80012310e-01 7.55729258e-01 -3.07789654e-01
-3.34023833e-01 1.52799651e-01 -6.59212589e-01 5.96804142e-01
6.89292789e-01 -4.46341574e-01 9.29514706e-01 1.05758643e+00
5.33593781e-02 2.51035303e-01 -1.12360990e+00 1.31994259e+00
-6.64495826e-02 -1.23055565e+00 -3.82109165e-01 3.71938676e-01
8.84219170e-01 1.93672672e-01 1.00890622e-01 1.29346147e-01
7.00529933e-01 -8.54212046e-01 3.99810731e-01 2.63384581e-01
8.68962348e-01 -7.18649268e-01 5.59850097e-01 7.68509030e-01
-7.63224721e-01 -3.62299949e-01 -4.17487890e-01 2.06564710e-01
-1.08471349e-01 9.57574069e-01 -5.25084674e-01 6.58160269e-01
9.36268806e-01 2.89245546e-01 -2.57765055e-01 1.21179771e+00
-6.47212118e-02 6.42735541e-01 -5.22953033e-01 5.98078221e-02
4.50707600e-02 -3.09867531e-01 4.45831239e-01 9.90489602e-01
2.03248307e-01 7.32342228e-02 2.79267192e-01 5.43889165e-01
-7.10889041e-01 5.23410320e-01 -5.81207991e-01 2.45345533e-01
5.28081059e-01 1.62241089e+00 -1.51300719e-02 -3.15038413e-02
-3.66784662e-01 1.00188208e+00 7.80597925e-01 3.37084323e-01
-5.23314774e-01 -8.29177573e-02 1.09429991e+00 -2.87098348e-01
3.00871255e-03 1.29718915e-01 -3.27728271e-01 -1.43691325e+00
5.17880358e-02 -1.19376910e+00 1.08459830e+00 3.15703563e-02
-1.58933830e+00 4.52156961e-01 -5.29877424e-01 -7.52370775e-01
2.27942631e-01 -1.59401298e-01 -3.12690884e-01 7.15933084e-01
-1.71347308e+00 -9.42533553e-01 -2.88224190e-01 1.10208261e+00
-5.21914065e-02 -1.36375606e-01 1.00382936e+00 3.93479377e-01
-8.11176062e-01 9.53133821e-01 6.69683635e-01 -4.78848889e-02
9.60742414e-01 -1.07101953e+00 -3.00018221e-01 8.82459641e-01
1.92207575e-01 6.05795681e-01 3.66370946e-01 -4.39932883e-01
-2.06502819e+00 -1.45798290e+00 9.57303822e-01 -1.88719720e-01
3.61828119e-01 -6.31624699e-01 -6.61810338e-01 6.53055489e-01
5.81083372e-02 9.80515838e-01 1.01568878e+00 -9.74632576e-02
-5.86899817e-01 -5.88889539e-01 -1.90719438e+00 2.33836725e-01
9.15854692e-01 -4.61211264e-01 5.49339578e-02 8.66704524e-01
8.62825751e-01 -4.01511461e-01 -1.00245857e+00 4.16155271e-02
2.01652914e-01 -8.02421153e-01 6.16683066e-01 -7.61111617e-01
-2.93389529e-01 -5.27082145e-01 -3.27668548e-01 -1.21332645e+00
-8.85594413e-02 -1.21293473e+00 -6.09680653e-01 1.26327407e+00
8.46782848e-02 -1.01527154e+00 1.08530784e+00 9.33903635e-01
3.83338600e-01 -7.40021408e-01 -1.32480943e+00 -5.86135983e-01
-1.04959957e-01 -2.14446634e-01 1.03014946e+00 1.38951671e+00
1.37271851e-01 -3.16969514e-01 -6.95859969e-01 4.39749181e-01
1.14684904e+00 4.24157649e-01 9.69052613e-01 -8.81331384e-01
-3.24777037e-01 7.65841305e-02 8.32591876e-02 -8.16264153e-01
4.13270354e-01 -1.37694120e+00 -1.19588308e-01 -1.00281549e+00
2.89713591e-01 -9.69923496e-01 -4.25128698e-01 8.85064721e-01
-6.96070418e-02 2.05133576e-02 4.30400699e-01 4.49963033e-01
-9.39046860e-01 5.25716722e-01 8.92043650e-01 -2.76128381e-01
1.80684417e-01 1.35470048e-01 -8.37602913e-01 1.35483816e-01
7.97476113e-01 -8.78135860e-01 4.65227515e-02 -5.32528996e-01
1.67981565e-01 1.50597125e-01 2.79261857e-01 -4.37900364e-01
7.46136606e-01 -3.66170347e-01 -1.75687492e-01 -9.63190794e-02
-5.03762811e-03 -1.52388871e+00 4.47402954e-01 4.39586282e-01
-4.21057314e-01 -8.92507359e-02 -4.65120465e-01 7.57661223e-01
-1.43199772e-01 -1.18019752e-01 7.06791222e-01 -7.14805946e-02
-2.18263924e-01 7.79779673e-01 7.54788369e-02 1.02651775e-01
1.33914375e+00 4.33350503e-01 -5.05033731e-01 -1.84024081e-01
-4.79212224e-01 7.00286984e-01 5.93890667e-01 1.17481686e-01
3.41557354e-01 -1.17201138e+00 -6.71036065e-01 2.59873629e-01
9.10641029e-02 1.36959001e-01 6.78737387e-02 8.99089277e-01
-2.85283625e-01 1.91412851e-01 4.77171481e-01 -2.44778305e-01
-1.35564005e+00 9.24627662e-01 4.31808025e-01 -3.59991759e-01
-5.09420097e-01 7.33907044e-01 -2.12875441e-01 -9.64319468e-01
7.16390014e-01 2.05415174e-01 4.25893843e-01 -1.12527266e-01
6.20399117e-01 6.54076338e-01 2.85449237e-01 -3.97117198e-01
-3.39451760e-01 1.81897245e-02 1.14369048e-02 2.89952815e-01
1.41656351e+00 -2.20405996e-01 -6.59277439e-01 1.32967532e-02
1.72166240e+00 1.04089119e-01 -1.06992269e+00 -6.56797051e-01
2.23889798e-02 -7.89747059e-01 2.13376671e-01 -8.69325697e-01
-1.44181097e+00 3.43373269e-01 6.26252055e-01 2.93328688e-02
1.04247499e+00 -2.69471347e-01 1.09156299e+00 2.99878389e-01
7.57021427e-01 -9.62221801e-01 -4.83634621e-01 1.62549064e-01
3.84014547e-01 -1.29891765e+00 -4.47459221e-02 -3.62701654e-01
-3.55463207e-01 9.15658534e-01 3.73214662e-01 2.55721271e-01
7.24779844e-01 3.65071774e-01 1.66728541e-01 -1.20158203e-01
-8.54828358e-01 3.46662670e-01 -3.14202607e-01 2.54626960e-01
-2.14984119e-01 2.00085208e-01 -5.30250788e-01 8.14704061e-01
3.76933962e-02 6.51226714e-02 1.45286709e-01 9.97066498e-01
-8.39532986e-02 -1.23139393e+00 -7.94950664e-01 2.97266841e-01
-8.22979987e-01 1.72965974e-01 -3.22531402e-01 1.14596158e-01
2.30497643e-01 1.06254590e+00 -4.04271513e-01 -1.70991957e-01
1.38625160e-01 1.72363132e-01 -1.02871031e-01 -2.21018553e-01
-1.20298040e+00 -2.86085725e-01 -2.33634949e-01 -8.02452385e-01
-1.22763276e-01 -5.80204666e-01 -1.12304497e+00 -5.19266725e-01
-3.42804551e-01 5.81927598e-01 1.06551075e+00 5.85518122e-01
6.14248097e-01 -2.30986223e-01 1.21744561e+00 -2.29753822e-01
-1.32346094e+00 -2.49754682e-01 -7.76216388e-01 5.50546706e-01
5.10688901e-01 2.34211870e-02 -6.16091549e-01 -2.68207937e-01] | [5.870334148406982, 6.373134136199951] |
a83aa72a-2cf0-4735-a0b8-fde49ffae25d | improving-point-cloud-semantic-segmentation | 2009.10569 | null | https://arxiv.org/abs/2009.10569v3 | https://arxiv.org/pdf/2009.10569v3.pdf | Improving Point Cloud Semantic Segmentation by Learning 3D Object Detection | Point cloud semantic segmentation plays an essential role in autonomous driving, providing vital information about drivable surfaces and nearby objects that can aid higher level tasks such as path planning and collision avoidance. While current 3D semantic segmentation networks focus on convolutional architectures that perform great for well represented classes, they show a significant drop in performance for underrepresented classes that share similar geometric features. We propose a novel Detection Aware 3D Semantic Segmentation (DASS) framework that explicitly leverages localization features from an auxiliary 3D object detection task. By utilizing multitask training, the shared feature representation of the network is guided to be aware of per class detection features that aid tackling the differentiation of geometrically similar classes. We additionally provide a pipeline that uses DASS to generate high recall proposals for existing 2-stage detectors and demonstrate that the added supervisory signal can be used to improve 3D orientation estimation capabilities. Extensive experiments on both the SemanticKITTI and KITTI object datasets show that DASS can improve 3D semantic segmentation results of geometrically similar classes up to 37.8% IoU in image FOV while maintaining high precision bird's-eye view (BEV) detection results. | ['Luc van Gool', 'Ozan Unal', 'Dengxin Dai'] | 2020-09-22 | null | null | null | null | ['object-proposal-generation'] | ['computer-vision'] | [ 7.96603635e-02 1.70362785e-01 -9.68482122e-02 -7.84090102e-01
-8.33408713e-01 -7.36560166e-01 6.50840521e-01 2.54749686e-01
-4.98755872e-01 -1.33545890e-01 -3.68341416e-01 -5.20640373e-01
1.94591954e-01 -8.22343886e-01 -8.68795395e-01 -2.92767763e-01
-9.94610637e-02 8.89936924e-01 1.01674867e+00 -6.14107788e-01
3.78572136e-01 9.16206300e-01 -1.87457299e+00 1.14489779e-01
7.22840309e-01 1.24543369e+00 3.32717657e-01 6.40509605e-01
-2.71420449e-01 1.70814484e-01 -4.19964641e-01 2.78800160e-01
6.87013984e-01 3.60612631e-01 -5.89872837e-01 -9.36478525e-02
1.01648259e+00 -4.85825509e-01 7.28000477e-02 8.09583724e-01
2.59721130e-01 3.67660642e-01 7.74888039e-01 -1.33057773e+00
3.02563071e-01 -2.25622207e-01 -5.81718385e-01 4.67455566e-01
1.36335179e-01 2.99990833e-01 8.29785109e-01 -8.24481070e-01
4.34222281e-01 1.40321195e+00 7.54319429e-01 2.99943149e-01
-7.51707792e-01 -8.37494969e-01 4.72575545e-01 -1.56126305e-01
-1.18968034e+00 -3.48674178e-01 5.59126616e-01 -3.55720133e-01
1.42594087e+00 4.78346981e-02 4.28501040e-01 5.07302701e-01
1.90010265e-01 9.01955605e-01 7.27981031e-01 7.56525323e-02
1.85338065e-01 -3.80507596e-02 3.04163247e-01 9.23270166e-01
3.56844217e-01 8.11995119e-02 -4.28823799e-01 3.95654917e-01
6.16764486e-01 -1.42096713e-01 2.80830443e-01 -8.33751917e-01
-9.44286644e-01 8.55539799e-01 8.18485320e-01 -3.09877425e-01
-7.86849633e-02 3.32934976e-01 1.05599321e-01 -7.78877586e-02
5.54347456e-01 4.27605450e-01 -5.40908933e-01 1.50271505e-01
-7.97118187e-01 4.80809748e-01 5.52528322e-01 1.28803825e+00
1.04916275e+00 3.77963930e-02 4.54243869e-02 4.61881518e-01
4.11566585e-01 9.62732494e-01 -1.83482349e-01 -1.24639261e+00
4.75963593e-01 9.09812868e-01 1.74281880e-01 -7.19330311e-01
-8.94136250e-01 -5.56692600e-01 1.34258002e-01 8.08942735e-01
2.62492567e-01 2.68172413e-01 -1.69329274e+00 1.19662690e+00
7.26587832e-01 -9.40777808e-02 -2.57684160e-02 1.18095016e+00
1.16155314e+00 4.03309137e-01 -5.25789186e-02 9.20593500e-01
1.37538469e+00 -9.20157790e-01 1.76222399e-01 -9.18287933e-01
6.79518878e-01 -5.67388058e-01 7.21633375e-01 1.83980197e-01
-8.01857710e-01 -5.51755965e-01 -1.36064816e+00 -6.08682990e-01
-6.49681211e-01 2.65949909e-02 8.10767949e-01 6.44786656e-01
-1.14140570e+00 1.56784236e-01 -1.03643882e+00 -5.91058910e-01
9.52093363e-01 4.48041707e-01 -1.43072948e-01 -1.81703746e-01
-6.83284819e-01 1.00184035e+00 4.18241143e-01 -2.55441248e-01
-1.22433043e+00 -9.12419975e-01 -1.06949389e+00 -5.20919450e-02
5.24082422e-01 -6.14341736e-01 1.29246938e+00 -4.27812219e-01
-1.08700168e+00 1.13953674e+00 -1.94665164e-01 -6.15640283e-01
4.39768463e-01 -3.48760068e-01 -7.36099854e-02 4.03367758e-01
5.60277522e-01 1.38076651e+00 7.50812054e-01 -1.22862077e+00
-1.36073291e+00 -5.78167260e-01 1.77366078e-01 5.41272104e-01
3.43057901e-01 -4.36751723e-01 -7.05778122e-01 6.36803359e-03
5.85522711e-01 -9.19489324e-01 -4.29038733e-01 2.43722469e-01
-3.79824162e-01 -2.80256838e-01 1.34436703e+00 -1.11199714e-01
1.42832279e-01 -2.03583694e+00 -4.83053774e-01 2.13507444e-01
2.75572360e-01 1.66735083e-01 -1.16387129e-01 -8.22922215e-02
4.37217325e-01 -1.17762350e-01 -2.92059213e-01 -5.13936758e-01
-4.96702343e-02 2.47355148e-01 -1.76730275e-01 4.23709363e-01
6.30947828e-01 9.04159307e-01 -8.21502268e-01 -1.82725132e-01
6.58624411e-01 4.11686271e-01 -6.85282826e-01 1.66547131e-02
-4.38938767e-01 2.48644784e-01 -7.09870934e-01 9.13077176e-01
8.48032832e-01 -1.36574969e-01 -4.96791542e-01 -9.94533077e-02
-2.20132470e-01 5.82870066e-01 -1.07773876e+00 1.84785962e+00
-3.03655356e-01 7.64207900e-01 3.47261041e-01 -7.69979239e-01
1.05096900e+00 -4.97112155e-01 3.70133013e-01 -8.28925252e-01
3.04651201e-01 2.46313214e-01 -2.29941785e-01 -4.95711677e-02
8.36790025e-01 1.61893755e-01 -1.87489048e-01 1.39343172e-01
-1.71011448e-01 -8.04851174e-01 -1.48807764e-01 3.08499664e-01
1.06393015e+00 3.32882494e-01 -2.16408566e-01 -5.14870465e-01
1.38889641e-01 8.79924238e-01 2.43936285e-01 9.30004179e-01
-3.63867015e-01 8.18140447e-01 1.37223855e-01 -4.95708406e-01
-9.64339972e-01 -1.29046738e+00 -3.30007195e-01 9.59815860e-01
9.82306957e-01 -2.63863010e-04 -5.32257736e-01 -7.44910002e-01
4.31439340e-01 7.38771200e-01 -4.94982839e-01 -1.41087279e-01
-4.05101418e-01 -5.25027812e-01 4.22030896e-01 8.98134649e-01
6.84826851e-01 -5.47368228e-01 -1.35512388e+00 1.08448170e-01
2.08260223e-01 -1.49445057e+00 4.80363108e-02 7.25313425e-01
-7.78562009e-01 -1.22056210e+00 -3.31165045e-01 -8.07398140e-01
4.81322467e-01 1.03491759e+00 1.19591093e+00 -1.70899168e-01
-5.77577651e-01 4.85216409e-01 -3.06462169e-01 -9.04255092e-01
-2.23813921e-01 3.65384579e-01 -7.36453980e-02 -5.55681288e-01
6.80711746e-01 -7.32825324e-02 -8.33643913e-01 5.21049619e-01
-3.21661234e-01 1.01026945e-01 4.58550483e-01 2.72499979e-01
7.21549273e-01 -2.73264170e-01 2.29029149e-01 -6.70273483e-01
-1.76994383e-01 -3.35840613e-01 -9.12823975e-01 -3.97970289e-01
-5.89692414e-01 3.96590345e-02 -3.47479433e-03 1.91905379e-01
-9.20067787e-01 3.96908849e-01 -2.18873397e-01 -5.36382437e-01
-4.42855507e-01 -3.40495169e-01 -7.46379942e-02 -4.85977888e-01
6.57612145e-01 -1.89411461e-01 9.73796099e-02 -3.40087831e-01
3.50188076e-01 4.80545908e-01 5.76646447e-01 -3.98475617e-01
7.58708537e-01 1.12169087e+00 2.74968117e-01 -8.01777601e-01
-9.74385917e-01 -9.85130429e-01 -6.20573580e-01 -4.74395528e-02
1.18186378e+00 -1.25694311e+00 -5.45682669e-01 2.87657171e-01
-8.42449427e-01 -5.85372388e-01 -7.83959478e-02 1.98193729e-01
-5.03694654e-01 -1.10101566e-01 -1.32633805e-01 -6.27900898e-01
-2.18143061e-01 -1.43168569e+00 1.87608647e+00 2.77261555e-01
-6.30316278e-03 -6.17237926e-01 -5.11922419e-01 5.65643728e-01
3.71956676e-01 1.89539403e-01 6.75750077e-01 -6.72412753e-01
-1.04742837e+00 -2.80714989e-01 -4.57032442e-01 2.43900884e-02
-1.93446413e-01 -3.02402496e-01 -1.29431152e+00 -5.34782745e-02
-4.20111805e-01 -1.08501643e-01 1.33361900e+00 3.98838818e-01
8.49072278e-01 6.93752766e-01 -7.37307489e-01 8.09039712e-01
1.18489265e+00 8.10571909e-02 1.14700966e-01 4.12779570e-01
9.68227088e-01 6.11394346e-01 9.45554137e-01 -1.11655546e-02
7.04870760e-01 5.36393642e-01 1.08344293e+00 -4.78744030e-01
-3.58301699e-01 -1.23249076e-01 4.35127430e-02 -3.23842853e-01
4.36460674e-01 -2.42544457e-01 -1.27171314e+00 6.62206233e-01
-1.57672787e+00 -4.08005238e-01 -5.49212694e-01 1.94004273e+00
-2.85074506e-02 7.47291028e-01 8.51285383e-02 -1.13911413e-01
3.63804907e-01 1.10092133e-01 -9.16137815e-01 -3.47478330e-01
-2.83679750e-04 1.52145594e-01 1.29665399e+00 6.42772794e-01
-1.40382504e+00 1.48479497e+00 5.83919001e+00 6.15505993e-01
-1.02710736e+00 -8.50973129e-02 5.45515776e-01 -2.29089215e-01
-2.61867523e-01 -6.05271272e-02 -1.30458856e+00 -1.32216230e-01
5.79128563e-01 4.46637899e-01 -2.14149684e-01 1.07992089e+00
2.06893831e-01 -5.84094882e-01 -9.59341705e-01 8.38549793e-01
3.14622074e-02 -1.39142692e+00 4.46043015e-02 6.63679373e-03
6.63456380e-01 8.66325855e-01 1.35432005e-01 2.25665003e-01
5.61187148e-01 -9.08715546e-01 9.41151440e-01 3.16310488e-02
6.18619919e-01 -8.69660854e-01 5.07983088e-01 5.11690438e-01
-1.29376948e+00 9.44930222e-03 -3.88823211e-01 -3.76310199e-02
9.62976888e-02 3.72813016e-01 -1.45468938e+00 3.04414332e-01
9.82937515e-01 7.62846231e-01 -7.10696042e-01 1.03573394e+00
-2.59939075e-01 2.77073830e-01 -8.47586870e-01 5.73802441e-02
8.09180677e-01 7.85940886e-02 8.15157175e-01 1.03563011e+00
1.13304891e-01 1.43547982e-01 3.54030490e-01 8.27299595e-01
1.85270324e-01 -3.27569723e-01 -6.59555435e-01 3.67964745e-01
3.70285511e-01 1.31028283e+00 -1.28434420e+00 -3.20173502e-01
-1.75257504e-01 8.30050349e-01 4.90472242e-02 2.41147965e-01
-7.21842885e-01 -4.08541352e-01 1.23626924e+00 4.15889055e-01
7.09531486e-01 -6.60601795e-01 -6.96391881e-01 -7.04068244e-01
-1.06683865e-01 -1.21881433e-01 1.14711545e-01 -8.42366636e-01
-5.54567933e-01 2.71374226e-01 3.68122309e-02 -1.03405833e+00
-3.66931222e-02 -8.55688691e-01 -4.31956530e-01 8.14935744e-01
-1.95088637e+00 -1.29916334e+00 -8.53112221e-01 3.50750923e-01
7.96660960e-01 1.24821223e-01 3.12770873e-01 1.68930486e-01
-1.85512245e-01 2.48028576e-01 -2.29258806e-01 -1.28444821e-01
5.08019865e-01 -1.34291410e+00 9.20387626e-01 8.83404016e-01
3.47084813e-02 5.57685494e-02 5.01356661e-01 -7.47328758e-01
-1.35328913e+00 -1.34834480e+00 3.20058286e-01 -9.51605558e-01
9.39446986e-02 -6.54911458e-01 -7.20567405e-01 4.36853021e-01
-5.47480941e-01 1.85173780e-01 7.19894990e-02 -1.20295033e-01
-4.16805148e-01 -8.54820907e-02 -1.16984141e+00 3.61659765e-01
1.48052931e+00 -3.06064069e-01 -3.29805881e-01 2.94199079e-01
8.34600985e-01 -9.66578066e-01 -2.20551133e-01 6.79911911e-01
3.50559592e-01 -1.00977600e+00 1.27253604e+00 -4.10907686e-01
1.41558915e-01 -5.94895363e-01 -2.70186663e-01 -9.29613590e-01
-5.08942232e-02 -1.17071837e-01 2.18431085e-01 5.27703881e-01
3.69653583e-01 -4.80891675e-01 1.31916845e+00 4.77755994e-01
-8.82839680e-01 -3.84514987e-01 -9.78396595e-01 -8.63029301e-01
-1.02596954e-01 -8.63625586e-01 5.34304261e-01 5.26851952e-01
-7.81776547e-01 4.04451340e-01 5.46915591e-01 7.53217757e-01
6.97902620e-01 3.43883425e-01 1.03864193e+00 -1.32071912e+00
3.60513031e-01 -5.80229938e-01 -8.46836269e-01 -1.52914047e+00
-6.78859279e-03 -1.08713830e+00 4.73523229e-01 -1.88804007e+00
-4.07313973e-01 -9.56209660e-01 -2.30158935e-03 5.43607771e-01
-1.13472752e-02 5.58143139e-01 -1.32187223e-02 3.89655419e-02
-8.77320409e-01 3.57358903e-01 1.02515483e+00 -1.03482954e-01
-2.58800387e-01 -1.02637492e-01 -4.87913817e-01 7.36670852e-01
7.12122977e-01 -2.98590213e-01 -5.18414617e-01 -7.84845114e-01
-1.61663871e-02 -4.77567196e-01 7.53093600e-01 -1.42665327e+00
1.61678419e-01 9.58323479e-03 5.49934268e-01 -1.25956476e+00
6.21173441e-01 -7.57509708e-01 -5.36550045e-01 4.40820634e-01
2.89028045e-02 -3.71998191e-01 6.77565336e-01 5.96108317e-01
1.93441078e-01 1.71837464e-01 8.36873829e-01 -1.01392306e-01
-1.42920077e+00 3.84055823e-01 -3.98126811e-01 1.04609564e-01
1.20513439e+00 -7.87226379e-01 -1.34563029e-01 -1.65086433e-01
-3.91872048e-01 7.94580758e-01 6.93232775e-01 7.22241223e-01
6.72661543e-01 -7.46125638e-01 -4.48402405e-01 4.47786063e-01
4.15778488e-01 8.23248088e-01 -4.68286499e-02 5.51158965e-01
-7.36446142e-01 6.00552857e-01 -9.18188840e-02 -1.30444431e+00
-9.25410330e-01 1.02591656e-01 4.92922455e-01 5.32700479e-01
-8.77988994e-01 1.18954456e+00 5.48594892e-01 -4.83697772e-01
1.57539263e-01 -7.70795643e-01 1.59008689e-02 -7.10972846e-02
1.74532071e-01 3.78610283e-01 5.49791276e-01 -5.46106994e-01
-7.13219464e-01 6.00628912e-01 -2.55125493e-01 1.06856398e-01
1.16311848e+00 -2.68806726e-01 6.64737880e-01 2.05247492e-01
1.17543185e+00 -3.01398605e-01 -1.84877276e+00 2.27593798e-02
-6.43551052e-02 -3.36965472e-01 4.74593341e-01 -8.03442657e-01
-9.00413632e-01 1.06698585e+00 9.46759164e-01 -1.90393940e-01
5.81038296e-01 3.76142263e-01 8.29769850e-01 5.53992093e-01
6.76036000e-01 -9.42091584e-01 -9.58891734e-02 8.43757212e-01
3.06291431e-01 -1.80220413e+00 2.17750464e-02 -7.33776212e-01
-3.86527359e-01 1.02260756e+00 8.66765022e-01 -3.06826293e-01
4.17489916e-01 2.65621036e-01 3.01171541e-01 -7.43875802e-01
-2.17881814e-01 -6.09874189e-01 4.28955168e-01 7.45257914e-01
-2.03773931e-01 -7.90744927e-03 4.42543715e-01 1.72092468e-01
-2.40072772e-01 -5.83882332e-01 2.46397495e-01 1.03522182e+00
-1.04710627e+00 -5.61904192e-01 -2.40669057e-01 6.38137937e-01
1.77554220e-01 5.14717922e-02 -3.11291635e-01 1.02681756e+00
1.86950654e-01 8.09128881e-01 6.63446307e-01 -2.81126410e-01
3.98800492e-01 -1.22708276e-01 3.12692106e-01 -9.53627944e-01
-2.77931094e-01 -1.50267363e-01 2.75576502e-01 -1.07851660e+00
-1.20472550e-01 -6.76858664e-01 -1.81235015e+00 2.18172893e-01
-1.17584124e-01 -2.72978812e-01 1.04575169e+00 9.78700459e-01
5.45629680e-01 5.02159715e-01 3.38892400e-01 -1.35382807e+00
-1.03039533e-01 -3.47661883e-01 -2.00409770e-01 4.79217097e-02
5.56230724e-01 -1.05824709e+00 -2.65797466e-01 -3.98451835e-01] | [7.7774763107299805, -2.6534321308135986] |
64b971c1-9988-41e9-8d58-e301bd9fe8c1 | multi-spectral-class-center-network-for-face | 2305.10794 | null | https://arxiv.org/abs/2305.10794v1 | https://arxiv.org/pdf/2305.10794v1.pdf | Multi-spectral Class Center Network for Face Manipulation Detection and Localization | As Deepfake contents continue to proliferate on the internet, advancing face manipulation forensics has become a pressing issue. To combat this emerging threat, previous methods mainly focus on studying how to distinguish authentic and manipulated face images. Despite impressive, image-level classification lacks explainability and is limited to some specific application scenarios. Existing forgery localization methods suffer from imprecise and inconsistent pixel-level annotations. To alleviate these problems, this paper first re-constructs the FaceForensics++ dataset by introducing pixel-level annotations, then builds an extensive benchmark for localizing tampered regions. Next, a novel Multi-Spectral Class Center Network (MSCCNet) is proposed for face manipulation detection and localization. Specifically, inspired by the power of frequency-related forgery traces, we design Multi-Spectral Class Center (MSCC) module to learn more generalizable and semantic-agnostic features. Based on the features of different frequency bands, the MSCC module collects multispectral class centers and computes pixel-to-class relations. Applying multi-spectral class-level representations suppresses the semantic information of the visual concepts, which is insensitive to manipulations. Furthermore, we propose a Multi-level Features Aggregation (MFA) module to employ more low-level forgery artifacts and structure textures. Experimental results quantitatively and qualitatively indicate the effectiveness and superiority of the proposed MSCCNet on comprehensive localization benchmarks. We expect this work to inspire more studies on pixel-level face manipulation localization. The annotations and code will be available. | ['Nenghai Yu', 'Honggang Hu', 'Bin Liu', 'Yue Wu', 'Wanyi Zhuang', 'Zhenchao Jin', 'Zhentao Tan', 'Qi Chu', 'Changtao Miao'] | 2023-05-18 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [ 3.85736346e-01 -5.51779449e-01 -2.67010003e-01 -2.80016929e-01
-6.26933157e-01 -4.91730422e-01 5.37039042e-01 -1.63238332e-01
1.67361051e-01 3.31540495e-01 8.03908557e-02 1.13013551e-01
-3.00343454e-01 -8.47122967e-01 -5.06560922e-01 -7.84863710e-01
-6.69977739e-02 -6.02617919e-01 -9.37120989e-02 -1.83845133e-01
4.62042421e-01 8.30763698e-01 -1.79527688e+00 6.17106855e-01
7.83824325e-01 1.30666268e+00 -2.09048204e-03 -2.90762186e-02
1.74763706e-02 5.01815975e-01 -8.03441048e-01 -5.27220964e-01
4.05510694e-01 -4.48573887e-01 -4.77938592e-01 2.89897233e-01
7.49901235e-01 -3.60523194e-01 -3.85311186e-01 1.57407033e+00
4.26114857e-01 -6.63561374e-02 3.66664648e-01 -1.52541137e+00
-9.16437507e-01 1.47376582e-01 -9.25393701e-01 4.01796997e-01
4.02418464e-01 3.64346087e-01 5.06105721e-01 -1.08476353e+00
3.59907150e-01 1.50479555e+00 9.36138451e-01 2.79750258e-01
-9.58253741e-01 -1.04741657e+00 3.77683938e-02 6.83312178e-01
-1.72545862e+00 -4.98591214e-01 1.27910447e+00 -9.13729444e-02
4.24882829e-01 3.18864703e-01 2.93766558e-01 1.24173212e+00
-9.13067982e-02 5.23643553e-01 1.41543543e+00 -3.74080420e-01
-7.64449760e-02 2.77690236e-02 8.62060487e-02 9.19055998e-01
4.40856248e-01 1.46765247e-01 -8.38268578e-01 -2.01642826e-01
6.80218935e-01 2.51030177e-01 -6.39852941e-01 8.89028888e-03
-7.78675377e-01 6.35163069e-01 4.54149544e-01 5.82295656e-01
-2.09057391e-01 9.25539657e-02 3.11643720e-01 1.16511174e-01
4.73603278e-01 2.30341926e-02 -9.78589132e-02 3.25758904e-01
-1.13801801e+00 -7.61335669e-03 2.83671439e-01 7.36293554e-01
1.05222404e+00 1.69689238e-01 -1.64578259e-01 1.03113163e+00
1.66887522e-01 5.43211401e-01 3.52537215e-01 -1.03888583e+00
3.09812009e-01 4.85887408e-01 -2.47393087e-01 -1.76584625e+00
-2.21309140e-01 -4.59076762e-01 -8.91602874e-01 1.35148883e-01
1.87081948e-01 3.14920425e-01 -5.70774674e-01 1.48274994e+00
3.04544419e-01 5.43586195e-01 -3.65180582e-01 1.07037449e+00
6.35690153e-01 2.71765888e-01 8.46628100e-02 -4.57920790e-01
1.50635886e+00 -6.29697204e-01 -8.05489838e-01 6.37174398e-02
2.52204597e-01 -6.96188509e-01 9.98091042e-01 4.10361528e-01
-7.28158176e-01 -6.76413715e-01 -9.13691878e-01 1.17796153e-01
-6.15483880e-01 3.23393017e-01 6.74355984e-01 1.26444006e+00
-9.18639362e-01 6.47115946e-01 -3.51275325e-01 -2.30232790e-01
9.57183957e-01 1.92727894e-01 -5.87885618e-01 -1.86854199e-01
-1.16334200e+00 5.55165827e-01 3.08268964e-01 5.01666248e-01
-7.84340084e-01 -7.79098213e-01 -8.04542065e-01 1.92221943e-02
3.70004207e-01 -3.89926345e-03 4.53497291e-01 -1.12768757e+00
-1.26475179e+00 8.18516850e-01 -5.88084105e-03 -7.22099468e-02
1.90075949e-01 1.18453480e-01 -9.02135015e-01 5.89781106e-01
2.09123805e-01 2.41986290e-01 1.46055317e+00 -1.57087481e+00
-4.33086634e-01 -7.34373629e-01 -1.29381314e-01 -2.93329209e-01
-9.28328097e-01 3.73887509e-01 -4.63661045e-01 -1.11308670e+00
2.08532006e-01 -4.18225497e-01 4.57434863e-01 3.58657986e-01
-3.53312582e-01 -1.00727789e-01 1.35616016e+00 -8.62139463e-01
1.24406421e+00 -2.32744932e+00 -3.17794532e-01 2.57455528e-01
1.47918731e-01 5.19564390e-01 -3.43916446e-01 1.68259218e-01
-3.87664437e-01 1.11526079e-01 -4.61107850e-01 -2.19259575e-01
-8.56864974e-02 -1.47026673e-01 -3.95585239e-01 1.10090899e+00
3.47950161e-01 5.81442773e-01 -8.32988441e-01 -5.19228995e-01
3.98266852e-01 5.99213481e-01 -5.26968002e-01 -2.13635236e-01
9.47963595e-02 2.72517949e-01 -3.59923393e-01 1.44205356e+00
1.43268943e+00 7.70340711e-02 -7.42343515e-02 -7.49880195e-01
1.17653355e-01 -5.23494959e-01 -1.03403783e+00 1.65476179e+00
-4.82510030e-01 5.34080327e-01 3.94009620e-01 -1.10715115e+00
7.65853643e-01 4.67730686e-02 5.47444224e-01 -7.28474498e-01
1.16983011e-01 1.58226311e-01 -3.55322689e-01 -7.00706244e-01
2.00696379e-01 3.35855298e-02 1.23582706e-01 1.94008097e-01
2.10814908e-01 2.61486024e-01 -1.53568730e-01 -1.16173074e-01
9.54409122e-01 1.73833594e-01 -5.08403741e-02 -2.80676335e-01
1.01277506e+00 -4.16936249e-01 6.19534969e-01 4.12165463e-01
-6.49501562e-01 5.05498886e-01 2.36580849e-01 -3.75669658e-01
-3.79644990e-01 -1.01884210e+00 -4.71737236e-01 9.90608037e-01
4.70535189e-01 -4.71614361e-01 -9.45933700e-01 -9.89366889e-01
3.05534154e-01 4.00755733e-01 -6.47879243e-01 -2.90149599e-01
-5.35148859e-01 -8.81667614e-01 9.57313061e-01 2.24669039e-01
1.05258787e+00 -8.40380490e-01 -9.38627943e-02 -2.56398559e-01
-3.73429179e-01 -1.25009823e+00 -3.83808821e-01 -4.65335429e-01
-4.49512720e-01 -1.30394375e+00 -4.17524129e-01 -5.73424101e-01
5.72176158e-01 7.69182682e-01 5.64368129e-01 4.51802582e-01
-7.51914799e-01 4.72305208e-01 -3.77311110e-01 -4.79902737e-02
7.26126134e-02 -4.71356124e-01 1.16068088e-01 6.69776499e-01
4.69204485e-01 -7.24446833e-01 -7.46120632e-01 3.52757454e-01
-9.60870862e-01 -6.48699880e-01 5.16540408e-01 7.37728775e-01
2.28743404e-01 4.58459586e-01 6.90826893e-01 -4.97417867e-01
4.84340578e-01 -5.96129179e-01 -3.03369671e-01 3.19812506e-01
-4.46922392e-01 -4.05905902e-01 6.38582230e-01 -3.36644202e-01
-1.24116182e+00 -1.03402920e-01 2.03587353e-01 -8.65212262e-01
-2.70874500e-01 1.73557103e-02 -6.04001105e-01 -7.12953627e-01
7.46746480e-01 4.39239413e-01 -8.08309168e-02 -3.61854643e-01
4.08869445e-01 7.71734715e-01 8.40080082e-01 -5.76055586e-01
1.20018184e+00 8.11049342e-01 1.08117998e-01 -1.19896388e+00
-5.93544483e-01 -4.77869868e-01 -4.10242587e-01 -5.30021727e-01
5.76665521e-01 -8.81723344e-01 -9.60964501e-01 6.55514657e-01
-1.18227124e+00 3.11529815e-01 1.12494461e-01 4.56965193e-02
-3.09610367e-01 1.20405102e+00 -5.28494000e-01 -1.02538240e+00
-1.34919405e-01 -1.01057363e+00 1.41621590e+00 1.87149927e-01
3.28435123e-01 -7.13230729e-01 -5.65082610e-01 6.87489450e-01
3.55460793e-01 3.01723987e-01 6.90408051e-01 -1.74725890e-01
-6.03641093e-01 -1.50270641e-01 -6.22859657e-01 6.11047804e-01
4.23868716e-01 -2.42925137e-01 -1.46056831e+00 -2.90370107e-01
3.54530245e-01 -1.40223548e-01 1.06454241e+00 -2.38359254e-02
1.98074245e+00 -3.47588807e-01 -3.23478162e-01 8.70327532e-01
1.41538119e+00 -2.58015513e-01 6.45860672e-01 1.06968403e-01
7.96974301e-01 9.17979538e-01 4.05893117e-01 5.84157228e-01
1.24887293e-02 7.15840340e-01 8.16710651e-01 5.20705059e-02
-4.04591143e-01 -4.41869646e-02 4.84519720e-01 2.31419057e-01
-1.66268021e-01 -3.41972187e-02 -5.47971666e-01 4.22721326e-01
-1.43884814e+00 -1.36066377e+00 -1.75726742e-01 1.90803564e+00
5.81191421e-01 -5.27445555e-01 -1.48922399e-01 4.32405114e-01
1.08711731e+00 4.64706033e-01 -2.78289229e-01 2.61469126e-01
-6.09418154e-01 4.39310819e-01 4.73396748e-01 1.03174329e-01
-1.23910165e+00 1.02243197e+00 4.94016552e+00 1.59739649e+00
-9.76864576e-01 4.63377297e-01 5.56554079e-01 1.83573455e-01
-6.30204827e-02 -1.19512655e-01 -5.88376403e-01 8.20823908e-01
5.10577261e-01 3.74408990e-01 6.16047442e-01 6.86262369e-01
2.91044384e-01 -6.24998892e-03 -5.83903670e-01 1.44590664e+00
6.31661236e-01 -1.26487827e+00 1.12117097e-01 8.78528506e-02
5.10745943e-01 -6.72461927e-01 2.65982449e-01 1.23925926e-03
-3.44604731e-01 -9.98246908e-01 6.00483596e-01 4.88405108e-01
8.78583550e-01 -9.10449684e-01 4.90779191e-01 2.98401862e-02
-1.32489169e+00 -5.22216201e-01 -4.23487335e-01 3.16768527e-01
-1.63291901e-01 6.36584401e-01 -2.80752778e-01 8.23300302e-01
9.84216034e-01 8.04807484e-01 -7.27196991e-01 7.69254327e-01
-1.52925581e-01 4.41109002e-01 -1.05767876e-01 5.64007998e-01
-1.22327879e-01 -2.18438372e-01 3.81449491e-01 1.23683679e+00
3.59523684e-01 5.00906557e-02 7.79447556e-02 9.90040898e-01
-2.85346240e-01 -1.47852162e-02 -5.50187051e-01 -4.99269832e-03
5.27227342e-01 1.58189654e+00 -7.09696949e-01 -3.30339894e-02
-5.23879409e-01 1.39863563e+00 4.37762924e-02 3.91212761e-01
-1.00696564e+00 -3.86296928e-01 9.44500685e-01 1.23200484e-01
1.98912755e-01 -5.65865971e-02 -2.54335254e-01 -1.18921113e+00
1.99046031e-01 -9.33063388e-01 3.79340410e-01 -6.42830253e-01
-1.56773603e+00 3.70672524e-01 -1.23152867e-01 -1.16407728e+00
4.09797907e-01 -7.49632299e-01 -5.45464218e-01 6.89692020e-01
-1.84491205e+00 -1.58400118e+00 -7.31257498e-01 1.03186679e+00
4.82661217e-01 -3.09670627e-01 6.22879922e-01 5.05696237e-01
-7.98455656e-01 8.93877089e-01 -1.04838416e-01 3.89200002e-01
7.82847106e-01 -6.36042297e-01 -1.99718758e-01 9.80205476e-01
6.58818558e-02 7.76814401e-01 2.93461829e-01 -7.39027798e-01
-1.58454990e+00 -1.35164332e+00 3.71244401e-01 -3.18447977e-01
6.37457430e-01 -2.66277701e-01 -9.51337457e-01 2.54919469e-01
7.70265376e-03 5.81758142e-01 5.08530080e-01 -3.07247967e-01
-8.02011847e-01 -4.83308494e-01 -1.58956027e+00 2.71685332e-01
1.32477081e+00 -1.05714691e+00 -1.34156749e-01 6.15984440e-01
2.21454248e-01 2.16527328e-01 -7.60592699e-01 5.67898810e-01
3.42926234e-01 -1.35828519e+00 1.28063130e+00 -9.47793350e-02
1.90082893e-01 -4.56930965e-01 -3.61005604e-01 -8.47397566e-01
-1.80937693e-01 -5.56028605e-01 -1.93665698e-01 1.41250432e+00
-3.24747533e-01 -5.30448854e-01 7.65332162e-01 -2.10959643e-01
-9.42536891e-02 -3.97564203e-01 -1.01635885e+00 -9.72632885e-01
-2.40639701e-01 -5.37230313e-01 7.63699472e-01 1.26050103e+00
-1.87719181e-01 -4.81600434e-01 -4.32080746e-01 6.62140012e-01
1.07694256e+00 2.21670434e-01 3.99241269e-01 -1.13373542e+00
2.86918432e-02 -5.06363571e-01 -6.98202729e-01 -4.08906162e-01
5.92964828e-01 -8.46195579e-01 -1.57366559e-01 -6.96350932e-01
1.01446956e-01 -1.68369815e-01 -3.10029387e-01 5.45542479e-01
-1.41321987e-01 8.64102006e-01 9.50509086e-02 2.52011836e-01
-3.31956416e-01 6.55620635e-01 1.00696456e+00 -2.82335460e-01
3.82090092e-01 -2.76429743e-01 -5.62970877e-01 8.02175939e-01
6.10733747e-01 -1.79491118e-01 -3.21706593e-01 -1.48645714e-01
-3.86516988e-01 -2.20965847e-01 9.08145010e-01 -1.28707945e+00
2.47863024e-01 -1.91071793e-01 8.22808683e-01 -3.68467689e-01
4.34644520e-01 -9.33858454e-01 -1.20445453e-01 3.54237080e-01
2.33846784e-01 -3.36206555e-01 1.55674323e-01 8.02787960e-01
-3.73736948e-01 -8.76544416e-02 1.02276397e+00 -9.10443738e-02
-8.31783891e-01 4.19012338e-01 -1.17916830e-01 -3.17248434e-01
1.11937964e+00 -6.18782878e-01 -5.37163138e-01 -9.71016362e-02
-3.50509942e-01 -3.02532285e-01 4.87234086e-01 5.07831395e-01
7.92424142e-01 -1.39420712e+00 -4.45699841e-01 5.00785291e-01
2.30609164e-01 -6.20906711e-01 7.55929172e-01 6.32486880e-01
-4.33055907e-01 1.62744895e-01 -3.63683164e-01 -5.23384452e-01
-1.17138934e+00 7.38241434e-01 4.51635420e-01 3.46844941e-01
-2.33215332e-01 8.15801382e-01 1.25304326e-01 -2.08312571e-01
8.18940923e-02 4.54238392e-02 -1.40675008e-01 2.45137036e-01
7.68263459e-01 6.57134831e-01 2.02626422e-01 -1.08520472e+00
-5.76314211e-01 7.43254840e-01 2.20343322e-01 3.42612147e-01
1.15334749e+00 -1.33375660e-01 -4.96248245e-01 -3.86000067e-01
1.30915058e+00 1.13906115e-01 -1.10202515e+00 -1.80420294e-01
1.92005299e-02 -1.03093541e+00 1.81442156e-01 -5.98498404e-01
-1.53640926e+00 8.87631297e-01 9.72564280e-01 5.81374839e-02
1.67517006e+00 -2.66895711e-01 6.54497683e-01 4.70401198e-02
5.21227598e-01 -1.11400104e+00 4.11596000e-01 6.29325062e-02
9.26313818e-01 -1.12519264e+00 1.12419590e-01 -9.47816908e-01
-7.44649991e-02 1.27010262e+00 6.42727673e-01 3.92024480e-02
7.05675423e-01 1.32263331e-02 -2.09059298e-01 -2.93921828e-01
1.49621546e-01 -1.89709395e-01 -4.48608166e-03 1.00651956e+00
4.73074578e-02 -1.56103417e-01 -6.81337938e-02 6.62552893e-01
2.10475922e-03 -2.22637251e-01 1.18170872e-01 7.60042250e-01
-4.21579301e-01 -1.07716739e+00 -8.16906512e-01 1.63232952e-01
-4.82759178e-01 -8.23645890e-02 -3.61693740e-01 5.29600501e-01
7.85136998e-01 1.25077200e+00 -2.35727742e-01 -4.41346318e-01
9.82189178e-02 -4.47893813e-02 5.62149405e-01 -2.69208431e-01
-6.35029733e-01 -2.07771108e-01 -3.71431112e-01 -9.89186406e-01
-5.46285391e-01 -5.76153100e-01 -8.72363329e-01 -5.05415380e-01
-4.75227863e-01 -1.92689210e-01 6.32410288e-01 7.36225903e-01
4.30370063e-01 1.91352740e-01 8.98715854e-01 -9.69394445e-01
-5.17059505e-01 -8.69367719e-01 -7.92555273e-01 5.00608861e-01
3.42996746e-01 -8.79214048e-01 -5.08573949e-01 8.83918479e-02] | [12.764852523803711, 1.07510244846344] |
2710b07a-df8f-45cc-803f-d3074d00680a | a-cautionary-tale-on-fitting-decision-trees | 2110.09626 | null | https://arxiv.org/abs/2110.09626v1 | https://arxiv.org/pdf/2110.09626v1.pdf | A cautionary tale on fitting decision trees to data from additive models: generalization lower bounds | Decision trees are important both as interpretable models amenable to high-stakes decision-making, and as building blocks of ensemble methods such as random forests and gradient boosting. Their statistical properties, however, are not well understood. The most cited prior works have focused on deriving pointwise consistency guarantees for CART in a classical nonparametric regression setting. We take a different approach, and advocate studying the generalization performance of decision trees with respect to different generative regression models. This allows us to elicit their inductive bias, that is, the assumptions the algorithms make (or do not make) to generalize to new data, thereby guiding practitioners on when and how to apply these methods. In this paper, we focus on sparse additive generative models, which have both low statistical complexity and some nonparametric flexibility. We prove a sharp squared error generalization lower bound for a large class of decision tree algorithms fitted to sparse additive models with $C^1$ component functions. This bound is surprisingly much worse than the minimax rate for estimating such sparse additive models. The inefficiency is due not to greediness, but to the loss in power for detecting global structure when we average responses solely over each leaf, an observation that suggests opportunities to improve tree-based algorithms, for example, by hierarchical shrinkage. To prove these bounds, we develop new technical machinery, establishing a novel connection between decision tree estimation and rate-distortion theory, a sub-field of information theory. | ['Bin Yu', 'Abhineet Agarwal', 'Yan Shuo Tan'] | 2021-10-18 | null | null | null | null | ['additive-models'] | ['methodology'] | [ 6.16554379e-01 3.82766843e-01 -4.91360486e-01 -6.21418357e-01
-8.68654013e-01 -6.52512074e-01 4.55760241e-01 9.30878818e-02
3.68045568e-02 9.71881747e-01 2.98408329e-01 -5.88227034e-01
-3.95598650e-01 -8.81173253e-01 -6.20168269e-01 -9.17134345e-01
-9.38773453e-02 6.02514207e-01 -2.70519167e-01 1.75687790e-01
1.15363836e-01 2.63263404e-01 -1.40941346e+00 -3.28201870e-03
1.17663717e+00 9.23417032e-01 -2.18486831e-01 5.62500358e-01
2.16921661e-02 7.05995858e-01 -2.22370163e-01 -8.47120464e-01
1.21693492e-01 -4.90284473e-01 -7.15490997e-01 3.93396057e-02
3.41823995e-01 -3.15976113e-01 7.06254244e-02 9.16989982e-01
3.47439110e-01 -1.37135340e-02 1.17583060e+00 -1.33524489e+00
-5.30011714e-01 1.05899405e+00 -6.79654002e-01 1.08413458e-01
1.04811065e-01 -2.57176667e-01 1.35252249e+00 -5.22930682e-01
2.80445963e-01 1.24751878e+00 9.67892647e-01 4.12526399e-01
-1.75966966e+00 -8.72327566e-01 3.79736334e-01 -3.56247909e-02
-1.31194329e+00 -6.87322021e-01 5.57534754e-01 -6.58767462e-01
2.13568434e-01 6.01115644e-01 2.12680593e-01 1.27762771e+00
1.28227875e-01 8.12875867e-01 1.50590563e+00 -5.54273725e-01
5.88212609e-01 4.03417885e-01 2.53140599e-01 5.98568559e-01
6.95825696e-01 2.09330484e-01 -3.93909365e-01 -3.56488228e-01
6.88337028e-01 7.18353242e-02 -1.22933805e-01 -5.22454798e-01
-6.44857228e-01 1.31064236e+00 2.90388614e-01 1.34582102e-01
-4.11328793e-01 1.68321177e-01 1.29928887e-01 1.63708702e-01
7.93563366e-01 -4.59785834e-02 -4.79392767e-01 2.15145513e-01
-9.97408867e-01 3.65373194e-01 1.03134906e+00 8.53463650e-01
5.11108160e-01 -3.47804576e-02 -1.78096339e-01 7.34463871e-01
3.26714903e-01 4.63116944e-01 1.07032903e-01 -9.02913272e-01
3.68342131e-01 2.34493673e-01 -3.13008726e-02 -9.52344835e-01
-3.49657416e-01 -7.89196849e-01 -1.25466609e+00 6.71558306e-02
5.20437837e-01 -1.21737726e-01 -7.01501429e-01 2.01411939e+00
3.09653729e-01 -3.01131129e-01 -4.25978690e-01 4.69706833e-01
1.55017048e-01 2.87348896e-01 2.46115342e-01 -5.28870881e-01
1.12940180e+00 -3.48649532e-01 -4.76299226e-01 -3.33920121e-01
8.01962912e-01 -3.31935167e-01 8.82828295e-01 4.62357998e-01
-9.15825307e-01 -4.64064516e-02 -7.59654820e-01 9.74049643e-02
7.21145123e-02 -1.88333109e-01 1.20817733e+00 1.12031054e+00
-9.10623610e-01 4.56010878e-01 -7.02878773e-01 -2.58616209e-01
6.90710306e-01 2.62402624e-01 -1.26068667e-01 -2.99799532e-01
-8.09241533e-01 8.12365651e-01 -4.99925837e-02 9.09795240e-02
-5.13461590e-01 -5.60220599e-01 -5.78906000e-01 1.59432039e-01
4.87642765e-01 -1.04992998e+00 1.25871038e+00 -1.02167416e+00
-1.16440547e+00 7.18165994e-01 -4.69504267e-01 -6.62360013e-01
7.50186503e-01 3.30503732e-02 6.92692399e-02 -3.88427496e-01
2.61100769e-01 2.22567946e-01 8.68635535e-01 -1.07733476e+00
-7.43347108e-01 -8.70134711e-01 3.15694772e-02 -1.25533715e-02
-2.34162305e-02 -1.34240344e-01 3.42777789e-01 -7.68443048e-01
3.74426693e-01 -9.70927715e-01 -6.98190153e-01 -2.34782338e-01
-5.79671919e-01 -2.88863301e-01 1.73321739e-01 -5.69974244e-01
1.35432005e+00 -1.89110672e+00 1.91724766e-02 3.91983390e-01
3.89737129e-01 -5.56627095e-01 1.96895123e-01 1.99326649e-01
-4.07242551e-02 4.74327922e-01 -5.37262499e-01 -1.55006066e-01
2.74728298e-01 9.77713913e-02 -6.53850257e-01 4.49078977e-01
-6.24693371e-02 7.92842090e-01 -4.77622330e-01 -2.80886203e-01
-5.75472340e-02 9.05182585e-02 -6.36869848e-01 -3.33596319e-01
2.52791271e-02 1.46650732e-01 -5.33077955e-01 6.81215525e-01
7.51767814e-01 -2.20336467e-01 3.81704360e-01 2.37101197e-01
6.30877167e-02 5.05045056e-01 -1.18267775e+00 1.18712640e+00
-5.48906267e-01 3.61089677e-01 2.55422056e-01 -1.57966685e+00
6.80233896e-01 -4.78405952e-02 3.95374686e-01 -2.35530227e-01
3.19337696e-02 2.76544660e-01 -3.18662785e-02 5.66255264e-02
2.77193207e-02 -6.99319005e-01 -2.30447754e-01 5.02757847e-01
-1.49306566e-01 -4.44313176e-02 -2.50115134e-02 2.91101672e-02
1.09781468e+00 -1.66807726e-01 6.48404241e-01 -4.78219539e-01
-2.11599879e-02 -1.49611458e-01 4.69863415e-01 1.28510141e+00
1.94559604e-01 1.65876314e-01 6.04611456e-01 -3.72839183e-01
-8.72251809e-01 -1.19240451e+00 -6.11142039e-01 1.35247457e+00
-4.53231454e-01 -1.45136490e-01 -5.61998546e-01 -6.96906567e-01
2.52510130e-01 1.02782261e+00 -1.00110900e+00 7.36594945e-02
-5.73508367e-02 -1.19363034e+00 2.16064394e-01 7.13954091e-01
1.96816683e-01 -4.12673771e-01 -3.51619601e-01 3.10975224e-01
-1.48483589e-01 -8.23898494e-01 -1.93921715e-01 8.40875983e-01
-1.37419033e+00 -9.07350123e-01 -5.31845987e-01 -2.35945746e-01
5.24106920e-01 2.30809361e-01 1.20299792e+00 -3.23400617e-01
1.38155492e-02 3.15127343e-01 -1.76026538e-01 -6.80408657e-01
-2.90260047e-01 2.90663391e-01 9.67477262e-02 -8.94185081e-02
4.07468557e-01 -1.00613725e+00 -3.45002353e-01 4.46979493e-01
-6.29408121e-01 1.62878148e-02 8.33508313e-01 8.02954614e-01
4.44298774e-01 5.33886030e-02 4.94312555e-01 -1.15614092e+00
5.04243255e-01 -6.35892808e-01 -6.67776704e-01 1.20254248e-01
-8.34998965e-01 2.69535720e-01 3.35971296e-01 -2.58049160e-01
-8.93511295e-01 8.27100500e-02 6.41516149e-02 1.39294505e-01
5.90413325e-02 5.88520646e-01 -1.61528111e-01 1.31905407e-01
7.68607438e-01 6.13632053e-02 -2.29648337e-01 -3.75473917e-01
4.80197072e-01 7.08340228e-01 2.45188534e-01 -7.89755166e-01
8.22656512e-01 4.45586264e-01 2.81377167e-01 -6.87661648e-01
-1.19729972e+00 -1.54393747e-01 -4.99774784e-01 2.58004636e-01
5.46794057e-01 -7.95212328e-01 -7.74111986e-01 -6.89023659e-02
-7.61496186e-01 -2.89007634e-01 -4.19989854e-01 3.84040594e-01
-7.86428452e-01 1.61135614e-01 -1.41474977e-01 -1.42154872e+00
-1.83329329e-01 -6.96610034e-01 9.87279177e-01 -2.56077707e-01
-4.39321578e-01 -1.04850149e+00 4.11123671e-02 4.47522521e-01
3.10658038e-01 2.31093615e-01 1.03896713e+00 -7.16829360e-01
-5.65144062e-01 -2.77320355e-01 -1.33043647e-01 2.28883535e-01
-6.02178685e-02 -9.19492766e-02 -1.07605493e+00 -7.09659979e-02
3.42103809e-01 -1.05015464e-01 1.17290914e+00 9.87976730e-01
1.45860004e+00 -8.98342669e-01 -4.91254002e-01 5.80919445e-01
1.20969069e+00 -1.12583771e-01 4.31471914e-01 2.70049404e-02
3.64249200e-01 7.66620398e-01 6.96392953e-02 5.36601186e-01
4.89234209e-01 7.07454681e-01 1.80264041e-01 1.09564491e-01
4.08261836e-01 -5.13914764e-01 3.07943791e-01 5.73419631e-01
-3.05298805e-01 -9.33459625e-02 -5.73797226e-01 3.36032510e-01
-1.75749755e+00 -1.21067226e+00 -2.42741287e-01 2.68184471e+00
9.35974360e-01 3.02187562e-01 5.13281226e-01 2.86525488e-01
6.27842307e-01 -1.22977071e-01 -5.01037836e-01 -6.15176141e-01
-1.31646693e-01 4.72232938e-01 7.56469607e-01 4.78959203e-01
-1.15339816e+00 3.37307960e-01 7.13450098e+00 1.03792739e+00
-6.51192904e-01 1.83588684e-01 1.04853106e+00 -4.24384512e-02
-5.93964040e-01 2.49379858e-01 -8.48855078e-01 4.33123797e-01
1.03145421e+00 -3.40466440e-01 4.22542930e-01 1.11417973e+00
1.19741298e-01 -2.53892958e-01 -1.38442385e+00 6.37668908e-01
-2.89433360e-01 -1.16573179e+00 -2.58661032e-01 5.86912334e-01
6.82101786e-01 -2.70869076e-01 1.37620673e-01 3.83788645e-01
1.13467646e+00 -1.35338044e+00 5.51250815e-01 2.37394243e-01
8.15582156e-01 -5.44919968e-01 5.44236183e-01 6.07273817e-01
-8.41406524e-01 -2.17179775e-01 -3.88675541e-01 -3.11857015e-01
-2.44162738e-01 1.10097694e+00 -7.54856467e-01 3.65153700e-01
5.51423252e-01 3.06891531e-01 -3.39258909e-01 7.67098129e-01
-1.94495633e-01 1.08362353e+00 -8.35639179e-01 -9.22014564e-02
-1.19113445e-01 -3.01224113e-01 3.29205781e-01 1.13227534e+00
3.99959117e-01 1.00885704e-01 -2.92698424e-02 8.94232869e-01
-6.91051558e-02 1.45227000e-01 -8.30296636e-01 9.98146459e-02
4.93042588e-01 1.01210749e+00 -5.58850050e-01 -9.08122286e-02
-3.94771367e-01 5.31608462e-01 1.35968789e-01 2.02368572e-01
-7.00740635e-01 1.08791254e-01 4.78897035e-01 4.32045460e-01
3.52452457e-01 -2.93810368e-02 -9.46396887e-01 -1.10982478e+00
1.50480211e-01 -8.14070523e-01 5.04642844e-01 -3.49571437e-01
-1.53786516e+00 -1.80444811e-02 1.63205564e-01 -8.23673010e-01
-4.44950908e-01 -4.33389217e-01 -3.92820299e-01 9.42902982e-01
-1.01299119e+00 -1.06623149e+00 2.08319519e-02 3.44908506e-01
3.73578429e-01 3.83674353e-01 8.37709725e-01 -1.26365557e-01
-4.58216876e-01 8.28804493e-01 4.37958300e-01 -9.83239189e-02
2.35274687e-01 -1.28922939e+00 1.43493623e-01 6.40391469e-01
3.10373724e-01 6.07994378e-01 9.21256900e-01 -3.43903124e-01
-1.27742696e+00 -9.26508486e-01 8.51322770e-01 -7.93642163e-01
7.58502126e-01 -4.96845484e-01 -7.17526019e-01 8.93829763e-01
-4.19982851e-01 -1.91337287e-01 9.83070493e-01 1.01133287e+00
-5.13987601e-01 -2.28780866e-01 -1.21217513e+00 4.35490161e-01
1.25170815e+00 -2.15171278e-01 -2.46358141e-01 3.52384597e-01
3.49934340e-01 8.02273229e-02 -7.46405482e-01 6.03576362e-01
8.85755897e-01 -1.06458545e+00 8.79072487e-01 -8.84061515e-01
6.18817747e-01 8.36858749e-02 -6.41341448e-01 -1.19675028e+00
-6.54513240e-01 -6.03916705e-01 -1.50121048e-01 1.18045413e+00
5.14439940e-01 -8.49838734e-01 8.23705792e-01 6.69680715e-01
2.24255070e-01 -7.82789767e-01 -1.00042653e+00 -7.84420848e-01
4.96211767e-01 -7.29231238e-01 2.75710702e-01 1.01300812e+00
-1.41827092e-01 5.84300995e-01 -2.79863626e-01 -1.13696039e-01
9.44375277e-01 1.49645880e-01 8.44770133e-01 -1.74965918e+00
-4.95388865e-01 -6.66178286e-01 -4.74997282e-01 -1.04287434e+00
5.44462316e-02 -8.95454407e-01 -8.88701901e-02 -1.23287916e+00
5.55682898e-01 -7.68383026e-01 -1.07269242e-01 3.75890046e-01
-2.50020146e-01 -1.40916212e-02 -8.67712721e-02 2.44156361e-01
-1.60511419e-01 3.76488298e-01 8.18305492e-01 -9.14338604e-02
-1.31702438e-01 7.31723607e-01 -1.44537413e+00 9.08470452e-01
7.02693701e-01 -6.68782890e-01 -3.32559913e-01 -1.23367175e-01
3.60435694e-01 1.73796684e-01 6.10247016e-01 -6.14191055e-01
-6.04610108e-02 -4.27731693e-01 5.04983604e-01 -4.25732583e-01
-8.40718821e-02 -7.37479806e-01 3.58280420e-01 3.46266896e-01
-5.65339029e-01 -2.45692313e-01 -1.88696697e-01 8.13128769e-01
1.20136477e-01 -3.67843807e-01 6.41297877e-01 -1.94053445e-02
-5.01822196e-02 2.26989374e-01 -3.45184118e-01 1.63032845e-01
7.24184215e-01 -2.94035524e-01 -6.67845532e-02 -6.21129632e-01
-8.14942360e-01 -1.44034415e-01 3.12705874e-01 -1.65575042e-01
7.64641464e-02 -1.16788495e+00 -1.01955640e+00 -7.90872872e-02
-7.15082884e-02 -5.42559773e-02 1.43916637e-01 1.16711676e+00
2.41974518e-01 4.35257822e-01 2.83523560e-01 -5.65546691e-01
-9.06898856e-01 5.12243509e-01 7.59164169e-02 -5.44250071e-01
-3.64794910e-01 8.98461938e-01 6.93080783e-01 -1.98031709e-01
1.58333912e-01 -3.38267237e-01 3.46809596e-01 1.69132799e-01
3.34367365e-01 4.32751328e-01 7.72395879e-02 -4.55576144e-02
-4.56333220e-01 2.65362948e-01 -6.79405704e-02 -7.45247751e-02
1.29696071e+00 -1.11882210e-01 2.96422448e-02 5.11319578e-01
8.00481200e-01 2.29107156e-01 -9.53999400e-01 -2.66579509e-01
8.31689239e-02 -4.04407352e-01 4.96021286e-02 -8.41765285e-01
-6.66081667e-01 8.14637601e-01 4.04288143e-01 6.68708086e-01
1.19737542e+00 1.21310458e-01 8.75306353e-02 2.57003397e-01
4.64849055e-01 -7.87427604e-01 -7.19501913e-01 9.05643180e-02
8.67259085e-01 -1.18707132e+00 3.16527516e-01 -5.17011523e-01
-3.88580292e-01 8.19088936e-01 -8.90527517e-02 1.30262328e-02
6.43392265e-01 3.93528432e-01 -5.41170418e-01 1.65168136e-01
-9.83977556e-01 -2.73757309e-01 2.09058017e-01 8.42571318e-01
5.18878400e-01 5.32305300e-01 -3.02930385e-01 8.96818876e-01
-5.86255252e-01 1.48061141e-01 3.30288023e-01 4.33320522e-01
-6.16885304e-01 -8.84195983e-01 -3.22742462e-01 1.01167786e+00
-5.51908672e-01 -3.05181801e-01 -4.94650751e-01 7.68472612e-01
-1.97130948e-01 1.04970002e+00 -4.09570895e-02 -2.65291423e-01
9.69627053e-02 2.34709814e-01 6.20938718e-01 -4.54357028e-01
-2.29645986e-02 -1.04719229e-01 1.78620964e-01 -2.25951299e-01
-3.53124440e-01 -1.01591527e+00 -2.19420224e-01 -7.45430708e-01
-5.05177915e-01 1.27289310e-01 7.10768104e-01 9.10190463e-01
1.74392298e-01 1.61724001e-01 8.68552744e-01 -4.86306757e-01
-1.16724491e+00 -1.01885688e+00 -6.27389789e-01 -7.27462247e-02
7.24554434e-02 -7.30911493e-01 -6.45963371e-01 -6.31158426e-02] | [8.11760425567627, 4.756628513336182] |
5f1ec1d7-bc7c-41c2-aa89-9c025f1971a4 | ae-net-adjoint-enhancement-network-for | null | null | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9835116 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9835116 | AE-Net:Adjoint Enhancement Network for Efficient Action Recognition in Video Understanding | Action recognition in video understanding is a challenging task, largely because of the complexity and difficulty in
temporal modeling, making it suffer from motion information loss
and misalignment of temporal attention in spatial dimensions. To
overcome these difficulties, we propose a novel temporal modeling
method called Adjoint Enhancement Network (AE-Net), which
can fully explore clues of motion and time in the long-range
structure. The AE-Net mainly consists of two new modules:
the Initial Adjoint Enhancement Module (IAE-Module), which
deals with shallow features; and the Global Adjoint Enhancement
Module (GAE-Module), which deals with global features. With
a novel mechanism of parallel spatio-temporal convolution and
difference fusion, the IAE-Module is to enhance the degree of
motion transformation in shallow network features, exciting the
potential of motion flow and avoiding motion information loss.
The GAE-Module is proposed to improve the local temporal
representation in long-range structures by feeding the enhanced
feature differences into a spatial cascade module with residuals
to resolve the misalignment of temporal attention in the spatial
dimension.The experimental results show that our AE-Net can
achieve state-of-the-art results in Something-Something V1, UCF101 and HMDB-51 datasets. | ['Wenqian Wang and Nanjun Li', 'Faliang Chang', 'Chunsheng Liu', 'Bin Wang'] | 2022-07-21 | null | null | null | tmm-2022-7 | ['video-understanding'] | ['computer-vision'] | [-4.81258556e-02 -4.33137506e-01 -2.95751207e-02 -1.76333591e-01
-2.17706352e-01 -1.06163330e-01 5.34556031e-01 -6.77161038e-01
-4.78109926e-01 3.41603339e-01 7.19883323e-01 -1.37974545e-02
-2.28184357e-01 -5.98044038e-01 -4.52505827e-01 -9.37824488e-01
-2.18315154e-01 -4.73213583e-01 6.52378559e-01 -2.68529862e-01
2.41189241e-01 2.20127597e-01 -1.24191725e+00 7.98622549e-01
7.87308455e-01 1.25234258e+00 1.71005249e-01 4.96820956e-01
-1.54667541e-01 1.26197231e+00 -2.02469975e-01 9.43377838e-02
1.64271697e-01 -5.31218529e-01 -8.09581339e-01 3.87396477e-02
8.08347389e-02 -6.62083507e-01 -1.10608506e+00 8.19381833e-01
4.54925090e-01 3.61283869e-01 3.78651619e-01 -1.34475243e+00
-6.29480243e-01 2.28393018e-01 -7.34332502e-01 8.84042978e-01
3.28468606e-02 5.68228960e-01 6.36528730e-01 -8.78957510e-01
4.35329288e-01 1.58893335e+00 6.75592959e-01 5.13912857e-01
-6.14935398e-01 -5.14165580e-01 5.74113429e-01 7.86203861e-01
-1.15390074e+00 -1.68150082e-01 8.50018799e-01 -5.34107447e-01
1.07849085e+00 3.88817750e-02 7.31883466e-01 1.13153887e+00
4.09885913e-01 9.85653341e-01 7.91497827e-01 7.68043026e-02
-6.63183630e-02 -4.33070183e-01 1.15763642e-01 7.26903319e-01
-4.71190631e-01 4.10218537e-01 -7.22972274e-01 4.58867669e-01
1.17980182e+00 2.00328708e-01 -4.40851361e-01 -1.35640815e-01
-1.46036577e+00 5.72431743e-01 7.90676534e-01 6.10782027e-01
-5.39188206e-01 1.02373257e-01 5.39787114e-01 1.67916998e-01
4.68563080e-01 -6.89739436e-02 -5.92928112e-01 -4.06947672e-01
-7.28309333e-01 -1.91455428e-02 2.62614992e-02 4.42935437e-01
4.80846852e-01 5.97257577e-02 -6.62583470e-01 7.07595110e-01
4.04580325e-01 1.28632873e-01 9.17151093e-01 -9.50028956e-01
7.86507308e-01 6.66138351e-01 -5.84137924e-02 -1.07108080e+00
-5.67598879e-01 -5.01307070e-01 -1.12927532e+00 1.43318415e-01
3.38632971e-01 -9.17079896e-02 -9.56545711e-01 1.86677110e+00
3.60751927e-01 6.91089451e-01 3.34109343e-03 1.13503981e+00
9.35342312e-01 9.28912818e-01 2.54199207e-01 -1.22383460e-01
1.36309004e+00 -1.36505663e+00 -8.40583682e-01 -2.58389950e-01
8.02208960e-01 -4.58614379e-01 7.31322706e-01 -6.24271296e-02
-9.13172722e-01 -1.04297531e+00 -8.19733620e-01 -1.96483612e-01
-3.12019080e-01 2.32728720e-01 6.73911750e-01 7.52059789e-03
-8.20682049e-01 9.19402003e-01 -8.30496788e-01 -2.46730193e-01
5.64636469e-01 1.49196073e-01 -5.05579829e-01 -1.10217527e-01
-1.52530980e+00 7.26827562e-01 4.19895142e-01 6.40140951e-01
-6.16652250e-01 -8.66177142e-01 -1.01412845e+00 9.13209543e-02
1.72255948e-01 -7.12639391e-01 9.24489737e-01 -1.03168452e+00
-1.33818865e+00 2.64892757e-01 -2.28573531e-01 -2.59559214e-01
6.20445311e-01 -2.51389444e-01 -6.42499030e-01 3.33737135e-01
1.70613468e-01 9.10815418e-01 9.04222906e-01 -5.35145283e-01
-8.11162949e-01 -4.74606335e-01 -7.78769553e-02 3.80568951e-01
-3.49040955e-01 -1.94553167e-01 -4.61576968e-01 -1.02948916e+00
2.50731468e-01 -6.88946664e-01 -2.36121595e-01 1.60087004e-01
1.89980775e-01 -1.44311190e-01 1.26855433e+00 -9.89049137e-01
1.47480893e+00 -2.32569242e+00 2.16291234e-01 -3.31982493e-01
3.62830199e-02 6.84504628e-01 -4.69665587e-01 1.78449214e-01
-5.29857934e-01 -1.09756544e-01 -1.25924870e-01 -5.76822348e-02
-3.45717072e-01 2.11218879e-01 -4.11414653e-01 3.34281981e-01
6.71072781e-01 1.24452293e+00 -9.23590183e-01 -3.15019578e-01
6.03289604e-01 6.58430278e-01 -4.25858110e-01 1.00272015e-01
6.53640851e-02 7.69544899e-01 -7.49273539e-01 4.33362007e-01
8.01731765e-01 -2.12843835e-01 -4.23336565e-01 -4.96555775e-01
-2.98184097e-01 1.79783627e-01 -1.26617062e+00 1.95374286e+00
-9.09232199e-02 6.03972733e-01 -2.18484372e-01 -9.12559986e-01
6.19603813e-01 2.72123158e-01 7.48743296e-01 -1.17613554e+00
1.51714817e-01 1.43303489e-02 1.29808456e-01 -8.99061203e-01
2.95933247e-01 -7.39928782e-02 3.20291489e-01 1.24439090e-01
-9.82228369e-02 6.02935076e-01 1.74254198e-02 -8.11172090e-03
1.19808972e+00 5.63835561e-01 -2.42656469e-01 -7.34562352e-02
9.47990477e-01 -4.15450633e-01 1.04962969e+00 2.00609386e-01
-6.14978552e-01 5.03165185e-01 4.08489466e-01 -8.24394405e-01
-8.18002522e-01 -7.07184315e-01 1.49651781e-01 8.20619047e-01
2.80620426e-01 -3.14242870e-01 -4.73314524e-01 -8.38598251e-01
-6.51648939e-02 2.73810506e-01 -9.88639891e-01 -5.89572191e-01
-8.02693188e-01 -6.83573484e-01 2.14922577e-01 1.13063681e+00
1.27147853e+00 -1.26225221e+00 -6.15407884e-01 4.00100768e-01
-5.00038564e-01 -1.08202100e+00 -9.22793210e-01 -8.84545669e-02
-1.04920781e+00 -9.71070707e-01 -9.94581103e-01 -5.41670740e-01
3.94101053e-01 5.94901025e-01 6.60135388e-01 -6.71174552e-04
-3.04760993e-01 1.35349065e-01 -5.17870486e-01 6.94674700e-02
2.63900995e-01 -3.64945471e-01 -3.57701302e-01 5.09851694e-01
4.47420180e-01 -7.00833142e-01 -9.88378227e-01 6.61024809e-01
-9.81451273e-01 1.62290648e-01 6.38417721e-01 1.00299764e+00
1.56341687e-01 1.77848577e-01 3.79622549e-01 -9.39285010e-02
6.72874525e-02 -5.27372599e-01 -3.07808608e-01 8.48920569e-02
-1.53611019e-01 2.41975859e-01 3.57960075e-01 -6.56139195e-01
-1.33924365e+00 -4.98721674e-02 -2.93017268e-01 -7.77278304e-01
5.25071472e-02 3.08652669e-01 -3.23411018e-01 1.05627934e-02
1.60204351e-01 4.59224313e-01 -6.12883344e-02 -4.43426758e-01
2.38416344e-01 2.30794713e-01 5.41538715e-01 -7.41082430e-02
5.73225319e-01 6.59150243e-01 4.15449403e-02 -7.03784168e-01
-8.94744873e-01 -5.75608194e-01 -9.71514523e-01 -4.19363439e-01
1.27873921e+00 -1.03005195e+00 -5.64265728e-01 1.07695556e+00
-1.13364184e+00 -3.76793265e-01 -2.68925428e-01 8.50650251e-01
-4.51754808e-01 5.11650741e-01 -7.84782469e-01 -3.94575149e-01
-1.42176062e-01 -1.11602151e+00 1.03719306e+00 4.96314615e-01
8.04063752e-02 -9.56340909e-01 2.19591781e-02 3.46679747e-01
3.80592346e-01 1.21142283e-01 6.97837651e-01 1.79839283e-02
-7.94155657e-01 2.21320346e-01 -5.19889474e-01 3.84747982e-01
4.54801619e-02 -1.39027312e-01 -9.59748685e-01 -1.04726091e-01
1.49688169e-01 -5.09149395e-02 1.34901440e+00 8.69731188e-01
1.14410508e+00 -1.16687514e-01 -2.39447922e-01 9.31513429e-01
1.05036116e+00 4.78954881e-01 1.21625781e+00 4.88067031e-01
8.91070306e-01 5.16348243e-01 9.34751570e-01 4.10408229e-01
2.89303601e-01 6.84221566e-01 3.86255085e-01 -3.13082606e-01
-2.02510566e-01 -1.92695558e-01 5.13015270e-01 6.24892890e-01
-1.66487455e-01 1.35277182e-01 -3.47884446e-01 6.00466430e-01
-2.25322890e+00 -1.29803395e+00 -2.57707566e-01 1.77831614e+00
3.97376090e-01 -2.59152632e-02 9.19757411e-02 2.65226662e-02
6.57047629e-01 6.42143071e-01 -6.53212428e-01 -5.52241411e-03
-3.12078178e-01 -2.24932492e-01 1.60859555e-01 1.43590301e-01
-1.33020520e+00 8.24483991e-01 4.78906822e+00 9.11261976e-01
-1.12863064e+00 7.96108842e-02 7.16103613e-01 6.34259032e-03
1.01774417e-01 -1.04375392e-01 -6.43227398e-01 7.29506373e-01
6.42995715e-01 3.49272609e-01 7.15052634e-02 3.78906906e-01
5.29863477e-01 -7.08454177e-02 -9.10170913e-01 1.03592730e+00
-2.05879301e-01 -1.26305711e+00 -1.88315399e-02 3.99093665e-02
4.97979760e-01 -6.57845438e-02 3.77514847e-02 5.37803769e-01
-3.54086846e-01 -9.37241912e-01 7.22926378e-01 7.79327631e-01
4.38045532e-01 -6.26286983e-01 6.83363497e-01 2.87148446e-01
-1.81402802e+00 -4.78684694e-01 -3.01267892e-01 -6.87333345e-02
3.58206540e-01 3.35938156e-01 6.43106326e-02 6.47261083e-01
1.01241326e+00 1.36806953e+00 -5.08349240e-01 1.08495772e+00
-7.13485107e-02 2.78281122e-01 5.79006113e-02 3.99127722e-01
7.32442260e-01 -2.93392599e-01 6.05173469e-01 1.14983106e+00
4.04209226e-01 4.75645751e-01 -8.75087678e-02 7.51164436e-01
3.23054135e-01 -3.22209716e-01 -3.47633004e-01 4.30016927e-02
-4.54891734e-02 1.27906251e+00 -5.44148624e-01 -2.96545148e-01
-4.84915286e-01 1.21099949e+00 1.31061301e-01 4.62850392e-01
-1.05059421e+00 -1.54603601e-01 8.64157975e-01 -6.80377707e-02
6.75739110e-01 -1.97496891e-01 1.42854854e-01 -1.23076510e+00
3.13171387e-01 -7.46356547e-01 6.86900377e-01 -9.94020283e-01
-1.13818526e+00 5.18061161e-01 -1.51746586e-01 -1.59474838e+00
6.64835349e-02 -6.60443604e-01 -9.01900291e-01 9.84611332e-01
-1.52780855e+00 -1.14310837e+00 -4.91751581e-01 9.60406482e-01
8.52828741e-01 -1.46573214e-02 2.24049389e-01 5.42760491e-01
-7.06736207e-01 2.99836248e-01 -1.07576601e-01 1.67518035e-01
6.24850214e-01 -8.65179360e-01 3.34964663e-01 9.32226062e-01
-1.59277335e-01 2.31633082e-01 1.89874485e-01 -6.54238462e-01
-1.13725543e+00 -1.42335367e+00 7.05685258e-01 -2.42777422e-01
5.91089487e-01 4.87090871e-02 -1.21657789e+00 5.56137681e-01
6.99504539e-02 3.70446384e-01 1.56328663e-01 -3.86143625e-01
-3.00795645e-01 -5.11445031e-02 -7.66764581e-01 4.39749390e-01
1.22286367e+00 -5.83894789e-01 -5.67257226e-01 1.52697936e-01
7.44842529e-01 -3.05035830e-01 -8.27013254e-01 6.78971827e-01
5.00143766e-01 -1.08675051e+00 1.04048944e+00 -6.51917219e-01
6.32296860e-01 -5.78063846e-01 5.89256920e-02 -1.06579649e+00
-8.03726614e-01 -5.25146127e-01 -1.54674351e-01 1.26853490e+00
-4.97832559e-02 -4.35211480e-01 6.47510827e-01 3.97065967e-01
-2.29886279e-01 -9.00892437e-01 -1.06185162e+00 -7.07694352e-01
-2.70241825e-03 -4.27505672e-01 3.45551640e-01 8.82050276e-01
-2.62855589e-01 3.47210944e-01 -5.92224181e-01 1.10013500e-01
1.26602277e-01 -1.62208155e-01 3.88296694e-01 -8.56602073e-01
-3.11572582e-01 -5.77496231e-01 -4.43066984e-01 -1.46633935e+00
-1.49126723e-02 -3.85915726e-01 -9.22635794e-02 -1.47913849e+00
2.01599315e-01 1.51700228e-01 -4.75992650e-01 4.63709086e-01
-4.09701794e-01 3.95797342e-02 1.86689183e-01 1.23222016e-01
-6.79449081e-01 1.22565091e+00 1.68390691e+00 -1.16548896e-01
-2.39465714e-01 -1.14480861e-01 -2.90435076e-01 8.38777900e-01
2.78122008e-01 -1.71319738e-01 -5.48961163e-01 -4.81444418e-01
-3.80754918e-01 2.22580016e-01 7.16273487e-01 -1.16392410e+00
2.03467950e-01 -1.67230330e-02 8.87876332e-01 -9.55742896e-01
4.09328848e-01 -7.14477420e-01 -2.47707479e-02 6.29248857e-01
-1.95605978e-01 2.70532042e-01 3.25055420e-01 7.29730487e-01
-5.66166103e-01 2.00350523e-01 7.88644493e-01 -1.01064235e-01
-1.38367677e+00 6.58360183e-01 -2.85864025e-01 -9.97802913e-02
1.02446854e+00 -3.51658314e-01 -4.29905951e-01 -2.57715970e-01
-7.77789176e-01 5.51666081e-01 2.22264347e-03 8.14658225e-01
7.42368162e-01 -1.66679215e+00 -5.74682534e-01 3.73801023e-01
-1.04622088e-01 -2.43592814e-01 1.11161160e+00 1.32686520e+00
-1.66601896e-01 5.08240700e-01 -4.46263075e-01 -7.04516768e-01
-1.09971321e+00 6.28113031e-01 5.70083261e-01 -5.19105613e-01
-8.94778430e-01 1.09683621e+00 8.18605840e-01 1.08537488e-02
1.72440454e-01 -3.64017457e-01 -5.38366854e-01 7.91939795e-02
8.16688597e-01 3.33993256e-01 -1.82045385e-01 -9.33881700e-01
-5.24083436e-01 8.85325313e-01 9.91435815e-03 -9.13082436e-02
1.31087863e+00 -3.22334290e-01 -2.44851001e-02 2.02014685e-01
1.37425733e+00 -5.75759172e-01 -1.88298333e+00 -3.26524287e-01
-2.89163500e-01 -5.79815030e-01 1.89695939e-01 -6.12069726e-01
-1.34449124e+00 1.19135046e+00 8.57643902e-01 -3.20737928e-01
1.54684937e+00 -3.36843193e-01 9.28425133e-01 -9.80943590e-02
1.32344132e-02 -9.44417834e-01 4.44305450e-01 7.48767734e-01
9.31758642e-01 -1.16820204e+00 -3.33467722e-01 -2.62891501e-01
-8.25810552e-01 1.20636225e+00 8.71096969e-01 8.71367231e-02
7.57554471e-01 -1.90592781e-01 -6.51139691e-02 -2.07015529e-01
-8.26873064e-01 -1.63261354e-01 7.32118666e-01 3.92170399e-01
2.62259424e-01 -5.50378442e-01 -1.66033849e-01 6.65323377e-01
3.50054711e-01 8.58691782e-02 -7.80516043e-02 8.77942264e-01
-1.81187078e-01 -8.38608384e-01 -1.13762036e-01 1.22613519e-01
-2.83826649e-01 -1.25897169e-01 -1.01788118e-01 8.79194438e-01
4.30658072e-01 9.00860488e-01 2.43545860e-01 -6.86601460e-01
3.59702826e-01 -1.03655592e-01 2.97213495e-01 -1.65758319e-02
-4.80825692e-01 3.37837517e-01 -2.87470780e-02 -1.10384929e+00
-5.74355543e-01 -5.93866229e-01 -1.31428909e+00 -9.16453172e-03
-2.02560928e-02 -9.64389369e-03 1.37655750e-01 1.14754295e+00
4.86553520e-01 9.88704085e-01 6.25431538e-01 -8.32165778e-01
-2.96567500e-01 -1.03639269e+00 -5.95842898e-01 4.92134035e-01
4.62929070e-01 -8.23147953e-01 -2.92655617e-01 9.03812572e-02] | [8.619685173034668, 0.4687407612800598] |
0cf096fc-b2a9-4582-81c8-a60180714038 | mask-conditioned-latent-diffusion-for | 2304.05233 | null | https://arxiv.org/abs/2304.05233v1 | https://arxiv.org/pdf/2304.05233v1.pdf | Mask-conditioned latent diffusion for generating gastrointestinal polyp images | In order to take advantage of AI solutions in endoscopy diagnostics, we must overcome the issue of limited annotations. These limitations are caused by the high privacy concerns in the medical field and the requirement of getting aid from experts for the time-consuming and costly medical data annotation process. In computer vision, image synthesis has made a significant contribution in recent years as a result of the progress of generative adversarial networks (GANs) and diffusion probabilistic models (DPM). Novel DPMs have outperformed GANs in text, image, and video generation tasks. Therefore, this study proposes a conditional DPM framework to generate synthetic GI polyp images conditioned on given generated segmentation masks. Our experimental results show that our system can generate an unlimited number of high-fidelity synthetic polyp images with the corresponding ground truth masks of polyps. To test the usefulness of the generated data, we trained binary image segmentation models to study the effect of using synthetic data. Results show that the best micro-imagewise IOU of 0.7751 was achieved from DeepLabv3+ when the training data consists of both real data and synthetic data. However, the results reflect that achieving good segmentation performance with synthetic data heavily depends on model architectures. | ['Vajira Thambawita', 'Michael A. Riegler', 'Pål Halvorsen', 'Sravanthi Parasa', 'Zahra Sepasdar', 'Leila Mozaffari', 'Roman Macháček'] | 2023-04-11 | null | null | null | null | ['video-generation'] | ['computer-vision'] | [ 5.91649532e-01 6.30139768e-01 1.42611936e-01 -1.81315109e-01
-8.16731632e-01 -4.66121525e-01 5.85060656e-01 -3.29913110e-01
-4.20120329e-01 8.30427229e-01 1.30511463e-01 -2.09485471e-01
4.19628263e-01 -8.38418186e-01 -8.62843812e-01 -6.98542416e-01
9.88951623e-02 3.13596874e-01 -7.52199849e-04 1.61955222e-01
-1.84211418e-01 -4.42916714e-03 -1.04019678e+00 4.63696986e-01
1.16589689e+00 7.98901320e-01 2.51394778e-01 8.73250663e-01
1.80403084e-01 7.00949371e-01 -9.41187084e-01 -6.22285664e-01
6.04959130e-01 -8.80655527e-01 -5.87058008e-01 1.54292688e-01
-8.61895978e-02 -5.26332319e-01 -2.87196904e-01 1.15874147e+00
6.78835511e-01 -2.46393770e-01 5.06377459e-01 -1.31620264e+00
-6.22454464e-01 6.21680558e-01 -3.79828691e-01 -1.27858549e-01
2.76666641e-01 5.11225224e-01 3.82131517e-01 -2.50528634e-01
7.25726962e-01 8.43111873e-01 5.01497030e-01 9.08293784e-01
-1.14937031e+00 -5.83692551e-01 -3.59864801e-01 -4.07676369e-01
-1.20061398e+00 -1.87430054e-01 5.23630977e-01 -4.25611168e-01
5.00418067e-01 4.51457947e-01 8.80719006e-01 1.54128158e+00
4.50283408e-01 7.56683290e-01 1.15851557e+00 -4.94120777e-01
2.70080060e-01 4.30801511e-01 -5.64502299e-01 7.68989742e-01
4.35398310e-01 2.69773126e-01 -1.09698484e-02 -7.20919669e-02
1.11960375e+00 -1.95501670e-01 -6.11325264e-01 -3.19122151e-02
-1.18484402e+00 9.79469240e-01 4.33987349e-01 3.49354118e-01
-4.24085885e-01 2.07181603e-01 2.25406483e-01 1.10030785e-01
8.74999315e-02 7.34025240e-01 -8.50141607e-03 6.94836164e-03
-9.86697197e-01 -1.48452476e-01 8.47315967e-01 1.08639526e+00
3.27917710e-02 2.04422459e-01 -2.89862216e-01 4.95693982e-01
2.65329123e-01 4.41147804e-01 1.10941422e+00 -8.69840860e-01
2.64252245e-01 4.90004361e-01 1.52567104e-01 -8.74417305e-01
9.70011428e-02 -5.09035885e-01 -9.69770432e-01 1.76260322e-01
5.32437682e-01 -4.73621398e-01 -1.41550481e+00 1.52949965e+00
1.49193048e-01 9.00867283e-02 2.55960107e-01 8.73078823e-01
9.04701173e-01 5.34870565e-01 1.44088998e-01 6.82622707e-03
1.15897095e+00 -1.02027059e+00 -7.83968210e-01 -2.44745865e-01
5.55245578e-01 -8.77294421e-01 1.02848923e+00 3.38636130e-01
-1.31965315e+00 -4.90554988e-01 -1.11777103e+00 2.06420928e-01
-2.87780594e-02 1.84377000e-01 6.55496120e-01 9.63041067e-01
-1.00439453e+00 2.83231854e-01 -1.16367793e+00 -2.38106102e-01
7.13799417e-01 3.59182894e-01 -2.24617109e-01 -2.37982154e-01
-1.12127936e+00 7.37572193e-01 3.61870348e-01 1.21746488e-01
-1.04868329e+00 -5.48488438e-01 -8.36798072e-01 -1.59991935e-01
7.77851343e-02 -1.11279321e+00 1.11767125e+00 -1.30594325e+00
-1.51727498e+00 9.69846487e-01 3.34154487e-01 -8.30123961e-01
1.14120829e+00 1.04061112e-01 -1.64073840e-01 2.29799405e-01
-6.20861202e-02 1.13621294e+00 8.05021524e-01 -1.21084619e+00
-4.35255706e-01 9.63538885e-02 -7.67513886e-02 3.07908226e-02
3.33469249e-02 -3.58920187e-01 -3.77931267e-01 -6.65022135e-01
-2.67860979e-01 -1.31875443e+00 -6.70687020e-01 -6.23649172e-02
-5.59075892e-01 6.19762480e-01 4.00447816e-01 -7.72177517e-01
8.42875719e-01 -2.13272309e+00 -2.18157768e-01 3.36123519e-02
-7.87585601e-02 4.55797434e-01 4.85043935e-02 7.37169608e-02
5.36747985e-02 5.05196035e-01 -3.38450581e-01 -1.47511452e-01
-2.67021537e-01 3.11365515e-01 -1.38826817e-01 2.38675714e-01
6.89956993e-02 1.14759135e+00 -8.68864238e-01 -7.88770020e-01
2.99686372e-01 5.88684559e-01 -6.58812106e-01 3.84273648e-01
-2.56796300e-01 1.05568576e+00 -4.65669274e-01 5.94456375e-01
5.45662701e-01 -4.23310131e-01 3.94041687e-02 -2.05837548e-01
3.25772136e-01 -2.54443705e-01 -7.49577224e-01 1.92019439e+00
-3.79985869e-01 3.50341439e-01 -6.14323318e-02 -7.11808324e-01
5.14784038e-01 7.08233595e-01 4.68117028e-01 -6.16684377e-01
5.13271153e-01 2.14856967e-01 3.93397033e-01 -5.88135242e-01
1.61872029e-01 -1.43655851e-01 3.04526254e-03 5.19705452e-02
-8.84865969e-02 -5.83380938e-01 2.88739447e-02 1.73890851e-02
1.03590691e+00 1.06784113e-01 7.82582313e-02 -7.37882629e-02
2.95267433e-01 2.80820906e-01 4.94913489e-01 6.58950806e-01
-2.58655727e-01 1.10382760e+00 3.95744801e-01 -2.84372389e-01
-1.08658874e+00 -8.57077420e-01 2.38675997e-02 3.43875960e-02
-1.85134877e-02 -3.00834998e-02 -1.06279838e+00 -7.64711618e-01
-5.96704483e-01 8.22581470e-01 -6.47879124e-01 -2.72208333e-01
-5.07966638e-01 -1.17345440e+00 6.84570312e-01 3.98252696e-01
6.66785479e-01 -1.21708572e+00 -9.47195828e-01 2.46219859e-01
-3.30243558e-01 -1.16810167e+00 -4.22930360e-01 -2.16227069e-01
-8.24690342e-01 -9.31236625e-01 -1.18073344e+00 -9.13589895e-01
1.12620318e+00 -2.42667839e-01 1.04196215e+00 1.90530922e-02
-7.00442195e-01 3.00943255e-01 -2.81993002e-01 -6.84613347e-01
-1.01086879e+00 -8.99988338e-02 -4.56647456e-01 -1.23246841e-01
-2.10708529e-01 -4.23644692e-01 -1.12840796e+00 3.63356173e-01
-1.12096786e+00 4.34128821e-01 7.97096670e-01 1.13131273e+00
7.81537831e-01 3.01776100e-02 4.29695755e-01 -1.14382267e+00
5.70436060e-01 -2.91550070e-01 -6.17665708e-01 1.33899912e-01
-5.83210647e-01 -7.33989179e-02 5.55545390e-01 -6.37785017e-01
-1.10282958e+00 2.46056572e-01 -2.24666253e-01 -4.20958340e-01
-4.78866138e-02 3.94439250e-01 3.97890387e-03 3.44389677e-02
8.56483519e-01 2.17570454e-01 2.56766170e-01 1.44933447e-01
2.98579216e-01 5.34358263e-01 4.61671621e-01 -3.82754356e-01
4.12747920e-01 4.75120991e-01 -1.28300935e-01 -4.82825875e-01
-5.45271993e-01 2.55154431e-01 -5.80020323e-02 -5.42919002e-02
1.20929432e+00 -1.01322746e+00 -4.09315825e-01 4.31387365e-01
-8.49384546e-01 -4.59715813e-01 -5.25053382e-01 6.56184912e-01
-4.92156833e-01 2.42557094e-01 -8.89572799e-01 -5.50908208e-01
-6.27390981e-01 -1.63998747e+00 7.57185280e-01 3.84202480e-01
-3.62747371e-01 -8.84485602e-01 -2.20567673e-01 5.18423498e-01
5.88479102e-01 8.89198959e-01 7.18931198e-01 -3.91227812e-01
-9.16112840e-01 -5.72768629e-01 5.68752736e-02 4.74498600e-01
3.22286367e-01 -1.77966654e-01 -8.16087902e-01 -3.54007930e-01
2.73081928e-01 -2.64468849e-01 4.55861479e-01 6.69719577e-01
1.14609718e+00 -4.03473854e-01 -3.39866042e-01 8.13427150e-01
1.38121879e+00 4.69165862e-01 9.05613542e-01 5.23155890e-02
6.32686019e-01 4.01249290e-01 5.56835413e-01 1.07927017e-01
-5.38267531e-02 1.62904307e-01 6.00987434e-01 -4.43219572e-01
-4.42067474e-01 -4.58206415e-01 1.18389074e-02 5.89378178e-01
8.53972211e-02 -5.55370450e-01 -7.14352727e-01 6.59153581e-01
-1.56640732e+00 -5.73967934e-01 -4.66020033e-02 2.17686033e+00
9.20667410e-01 1.49756402e-01 -1.58734962e-01 -1.79942891e-01
6.59162879e-01 -2.66738027e-01 -4.95557725e-01 -1.81995064e-01
9.76540446e-02 3.10296088e-01 8.67463529e-01 2.47317865e-01
-9.53155220e-01 6.87186778e-01 6.45276117e+00 7.31115878e-01
-1.28408706e+00 4.51960601e-02 1.08507442e+00 -3.27770635e-02
-2.67071694e-01 -3.58924091e-01 -2.95494825e-01 8.07464480e-01
1.06514072e+00 -1.33326910e-02 1.82917461e-01 7.72028983e-01
4.14163480e-03 -1.95509404e-01 -8.73649716e-01 8.85705829e-01
-1.60735399e-02 -1.30584419e+00 9.28352475e-02 3.26143324e-01
1.10146558e+00 -4.15708087e-02 4.30508316e-01 6.93174303e-02
5.29582381e-01 -1.34718859e+00 3.71390194e-01 4.48609531e-01
1.03580928e+00 -4.96080965e-01 1.04213989e+00 3.06710601e-01
-3.80818933e-01 2.80574918e-01 -2.02588141e-01 4.32090163e-01
1.81068078e-01 4.57933486e-01 -1.59078658e+00 4.13102508e-01
3.35033774e-01 1.16006106e-01 -3.16652805e-01 1.06706035e+00
-3.58567476e-01 7.55721569e-01 -2.91717708e-01 2.16175690e-01
2.58071810e-01 -1.90940008e-01 2.49744728e-01 9.51922059e-01
6.35255158e-01 -8.90818909e-02 -1.14759067e-02 9.54163790e-01
-5.46686761e-02 -1.40577834e-02 -5.50236285e-01 -3.04847777e-01
5.59381992e-02 1.22393525e+00 -1.00103438e+00 -2.55721480e-01
-2.46941045e-01 1.17881024e+00 -4.41897422e-01 1.47505865e-01
-1.17123306e+00 2.29182374e-02 8.87438431e-02 1.95131108e-01
9.70011503e-02 5.03385067e-02 -5.10352015e-01 -9.32219326e-01
-1.32108852e-01 -1.03980625e+00 2.70623207e-01 -7.19674766e-01
-9.68661070e-01 1.00458705e+00 -3.41255367e-01 -1.45886135e+00
-4.90089387e-01 -2.57783741e-01 -4.94736105e-01 8.62439215e-01
-1.17680061e+00 -1.26800466e+00 -5.41625321e-01 3.98231596e-01
4.46606904e-01 -1.01176180e-01 1.15505469e+00 2.72405833e-01
-3.40803295e-01 8.03732514e-01 -1.26783192e-01 2.20230162e-01
6.44981980e-01 -1.18953121e+00 3.65207016e-01 8.63975704e-01
1.07043944e-01 3.40505302e-01 7.18540728e-01 -6.85907066e-01
-9.97477710e-01 -1.17859256e+00 2.99144298e-01 -4.02455449e-01
1.08290888e-01 -1.02895550e-01 -5.73935151e-01 6.97799146e-01
4.27961826e-01 9.36855748e-02 6.74169600e-01 -8.18673074e-01
4.33113933e-01 1.95857629e-01 -1.64692032e+00 7.48823583e-01
6.89141393e-01 -9.29660443e-03 -2.63663262e-01 3.98791850e-01
8.35148811e-01 -8.05665195e-01 -1.01451325e+00 4.88313586e-01
3.31008375e-01 -7.68930018e-01 9.02866721e-01 -2.02259630e-01
7.45167911e-01 -2.71178126e-01 1.58550754e-01 -1.28509796e+00
1.87635094e-01 -6.97952926e-01 3.94861788e-01 9.56380606e-01
6.84734762e-01 -6.71532094e-01 1.05014551e+00 1.05224586e+00
-9.16967615e-02 -6.64437771e-01 -6.98049009e-01 -3.84094685e-01
-1.01053372e-01 -8.19156840e-02 4.70597118e-01 9.42509770e-01
-3.01802099e-01 5.36265336e-02 -4.83390391e-01 5.68068251e-02
3.03106040e-01 9.00409557e-03 7.77639389e-01 -5.12617350e-01
-6.01519704e-01 -9.60406885e-02 -4.03786808e-01 -7.07082629e-01
-3.10569495e-01 -8.27174962e-01 -2.60381475e-02 -1.58737123e+00
-1.29265323e-01 -4.98025775e-01 -6.08696342e-02 2.78241307e-01
-1.71940506e-01 5.32801807e-01 1.79723546e-01 1.79839924e-01
-6.63475022e-02 2.59625822e-01 1.88167214e+00 -2.20903307e-01
-1.96894914e-01 2.17511967e-01 -6.85827136e-01 6.65897846e-01
8.55832994e-01 -4.98720407e-01 -7.33709574e-01 -4.29175377e-01
-6.05541356e-02 3.51026088e-01 2.45088115e-01 -1.09597015e+00
7.19113573e-02 1.69027597e-01 5.98707139e-01 -2.35027492e-01
3.60848308e-01 -9.02282894e-01 6.55698299e-01 9.48494375e-01
-1.85817614e-01 -3.29767406e-01 2.37267286e-01 4.52706605e-01
-4.15788025e-01 -2.15033785e-01 8.09625328e-01 -6.02842510e-01
-1.70355678e-01 2.04314679e-01 -2.16909215e-01 4.44805846e-02
1.15645444e+00 -3.80963236e-01 -4.32082899e-02 -5.08216560e-01
-7.69464076e-01 8.14595148e-02 6.36036754e-01 1.56842083e-01
5.30441523e-01 -9.57658708e-01 -6.90357804e-01 4.56921548e-01
-2.01954067e-01 4.55437303e-01 3.36979717e-01 7.94813871e-01
-1.00854301e+00 3.12410086e-01 -2.75061399e-01 -6.60228670e-01
-9.53286767e-01 6.82100713e-01 3.80849719e-01 -4.29573029e-01
-6.75265610e-01 9.68839943e-01 5.13073146e-01 -4.10830975e-02
6.38232306e-02 -5.96557260e-01 1.73703298e-01 -5.07724702e-01
2.65337378e-01 -6.14576004e-02 -5.19430600e-02 -1.49308473e-01
5.17124943e-02 5.22037074e-02 -1.25848770e-01 -2.29569271e-01
9.47103262e-01 1.31112605e-01 3.99692684e-01 -5.85277043e-02
8.85795414e-01 -2.87390407e-02 -1.38836372e+00 3.55431825e-01
-4.86756951e-01 -5.06099582e-01 -1.24648333e-01 -1.32599103e+00
-1.35342741e+00 5.71445048e-01 7.38696456e-01 5.60180470e-02
1.18456197e+00 -2.59091377e-01 1.09855330e+00 -3.60404819e-01
4.94784504e-01 -6.61203027e-01 -1.46523088e-01 -2.82142550e-01
8.50289822e-01 -1.23170924e+00 -2.36209407e-01 -6.41007721e-01
-9.22733784e-01 6.20644271e-01 5.08395016e-01 -2.31522530e-01
3.46231222e-01 3.52067202e-01 4.24038798e-01 2.57579889e-02
-2.95401305e-01 2.46659711e-01 1.93385854e-01 6.51641905e-01
4.77791548e-01 2.05058873e-01 -4.31772649e-01 4.99005079e-01
-4.90306795e-01 2.84293860e-01 5.55515349e-01 9.12900925e-01
2.58293718e-01 -1.24531233e+00 -2.63825446e-01 3.98947954e-01
-1.02287483e+00 -4.94583249e-02 -1.08620547e-01 8.84232104e-01
2.23939553e-01 8.08382154e-01 -2.43722260e-01 -1.29904538e-01
9.49319452e-02 -2.21892104e-01 5.38800418e-01 -4.17983383e-01
-6.57770216e-01 2.08124578e-01 6.86038435e-02 -3.51974428e-01
-3.50817084e-01 -3.79635930e-01 -1.17412937e+00 2.07331061e-01
-3.11433017e-01 1.57266945e-01 8.80495548e-01 4.73714292e-01
3.43188554e-01 9.73051131e-01 2.59957284e-01 -4.27079350e-01
-4.53637958e-01 -1.07487702e+00 -2.35780910e-01 5.57206392e-01
-3.95073518e-02 -2.79313624e-02 -3.18806022e-01 4.47672784e-01] | [14.187469482421875, -1.9212177991867065] |
3974c409-4107-4637-9340-fc364f428039 | adapting-membership-inference-attacks-to-gnn | 2110.08760 | null | https://arxiv.org/abs/2110.08760v1 | https://arxiv.org/pdf/2110.08760v1.pdf | Adapting Membership Inference Attacks to GNN for Graph Classification: Approaches and Implications | Graph Neural Networks (GNNs) are widely adopted to analyse non-Euclidean data, such as chemical networks, brain networks, and social networks, modelling complex relationships and interdependency between objects. Recently, Membership Inference Attack (MIA) against GNNs raises severe privacy concerns, where training data can be leaked from trained GNN models. However, prior studies focus on inferring the membership of only the components in a graph, e.g., an individual node or edge. How to infer the membership of an entire graph record is yet to be explored. In this paper, we take the first step in MIA against GNNs for graph-level classification. Our objective is to infer whether a graph sample has been used for training a GNN model. We present and implement two types of attacks, i.e., training-based attacks and threshold-based attacks from different adversarial capabilities. We perform comprehensive experiments to evaluate our attacks in seven real-world datasets using five representative GNN models. Both our attacks are shown effective and can achieve high performance, i.e., reaching over 0.7 attack F1 scores in most cases. Furthermore, we analyse the implications behind the MIA against GNNs. Our findings confirm that GNNs can be even more vulnerable to MIA than the models with non-graph structures. And unlike the node-level classifier, MIAs on graph-level classification tasks are more co-related with the overfitting level of GNNs rather than the statistic property of their training graphs. | ['Xingliang Yuan', 'Shirui Pan', 'Xiangwen Yang', 'Bang Wu'] | 2021-10-17 | null | null | null | null | ['membership-inference-attack'] | ['computer-vision'] | [ 5.43465734e-01 3.08562785e-01 -1.37906857e-02 -1.49095789e-01
-5.32733202e-02 -9.37482536e-01 4.46304381e-01 4.59777445e-01
-3.07349056e-01 8.16841424e-01 -6.66871369e-01 -8.03852439e-01
-2.16339558e-01 -1.25484788e+00 -1.01007414e+00 -6.81598306e-01
-6.12700939e-01 1.68580309e-01 3.51239234e-01 8.86705518e-03
9.04264450e-02 8.61306548e-01 -7.25041270e-01 -1.00464500e-01
6.51198685e-01 7.55589545e-01 -6.96238160e-01 6.53815091e-01
2.93454975e-01 5.15627444e-01 -8.91339183e-01 -8.92337978e-01
4.68611985e-01 -3.16998214e-01 -6.31465733e-01 -3.41536820e-01
3.60724270e-01 -6.96817040e-02 -6.82745934e-01 1.62575948e+00
4.55944091e-01 -2.94413328e-01 6.25505328e-01 -1.88543725e+00
-2.54753768e-01 9.54496205e-01 -5.32916665e-01 1.60541743e-01
2.97683060e-01 3.08991462e-01 8.97460163e-01 -1.26981393e-01
5.03435314e-01 1.14395344e+00 7.44373024e-01 5.05056679e-01
-1.28545201e+00 -1.23725963e+00 3.72447222e-02 -1.36222839e-01
-1.45243108e+00 -1.49736539e-01 6.66888654e-01 -2.60080725e-01
6.54175460e-01 4.40365821e-01 3.74561757e-01 1.41395736e+00
5.34010231e-01 2.31522486e-01 9.97206509e-01 4.10512723e-02
3.60735327e-01 -2.16714099e-01 9.02840421e-02 6.33223712e-01
9.65222120e-01 2.08957046e-01 -1.90141886e-01 -6.81033790e-01
6.89845741e-01 6.72330335e-02 -2.34448791e-01 -3.18610251e-01
-6.78844273e-01 9.53862548e-01 7.33541548e-01 1.13298915e-01
-1.66737884e-01 4.80686694e-01 4.58724380e-01 4.32068080e-01
1.19696788e-01 5.51742435e-01 -3.26652825e-01 4.29570884e-01
-5.62550068e-01 -1.00644208e-01 1.19444895e+00 9.27363575e-01
4.68690425e-01 8.76681060e-02 1.89171851e-01 -1.82826519e-02
2.78584093e-01 3.18284273e-01 -3.82876466e-03 -2.25451648e-01
3.25551540e-01 6.64027214e-01 -6.58612013e-01 -1.41486633e+00
-5.12729824e-01 -4.74573165e-01 -1.34638250e+00 1.07243232e-01
6.19191647e-01 -4.79040116e-01 -9.91665661e-01 2.01474857e+00
2.71392882e-01 4.55728114e-01 8.19542110e-02 5.07905185e-01
8.79235566e-01 2.39517510e-01 2.58721650e-01 7.64294788e-02
1.28108299e+00 -2.52431333e-01 -4.99062538e-01 -2.76231557e-01
7.70456135e-01 -1.35919079e-01 3.56808960e-01 3.40757132e-01
-6.28302395e-01 2.62462888e-02 -1.26024520e+00 4.55954403e-01
-7.78585732e-01 -5.54581642e-01 9.72750366e-01 1.20560002e+00
-8.53724062e-01 8.54125202e-01 -7.19287872e-01 -2.43381098e-01
8.49029541e-01 7.40157902e-01 -6.31863117e-01 -1.33092880e-01
-1.61594713e+00 5.08235812e-01 5.28089046e-01 -3.02393408e-03
-9.89000738e-01 -6.40529692e-01 -8.26147318e-01 9.64925811e-02
5.54351091e-01 -3.99177343e-01 4.09236014e-01 -5.74759662e-01
-9.25204277e-01 5.92642009e-01 4.34307426e-01 -8.51769745e-01
4.79216635e-01 5.34490824e-01 -7.36147165e-01 1.81679025e-01
-2.57912755e-01 9.82779860e-02 6.78289413e-01 -1.04557896e+00
-4.50474806e-02 -4.36489731e-01 3.79150927e-01 -3.47132057e-01
-1.61295012e-01 -1.94607526e-01 -5.02183996e-02 -4.25584495e-01
1.17848255e-01 -9.68920350e-01 -4.32918340e-01 7.10126981e-02
-1.14646375e+00 5.91480359e-02 9.13992941e-01 -2.38676161e-01
1.25529909e+00 -2.07084632e+00 -3.26964855e-01 1.02721107e+00
7.58251548e-01 5.89533746e-01 -7.78762102e-02 4.10205275e-01
-2.48387545e-01 7.02101946e-01 -3.37021708e-01 2.65756994e-01
-5.22900075e-02 1.48361087e-01 -2.36332938e-01 8.26711059e-01
1.42233536e-01 1.02356935e+00 -8.48239422e-01 -1.21446900e-01
-1.27847239e-01 2.87894726e-01 -3.65076572e-01 -1.58214733e-01
-1.23369493e-01 7.78231472e-02 -4.78903472e-01 6.96937144e-01
9.22166049e-01 -4.74695623e-01 5.99607766e-01 -2.65146911e-01
8.20585608e-01 6.86382428e-02 -9.31405127e-01 9.50522482e-01
1.32192671e-01 5.32313824e-01 5.72227053e-02 -8.74442458e-01
8.96547377e-01 1.85360134e-01 1.52577385e-01 -3.58595967e-01
3.50593925e-01 1.82055607e-02 6.74851418e-01 5.05451746e-02
-2.78018080e-02 1.02605037e-01 -2.62475848e-01 5.22515118e-01
-4.50840173e-03 1.57019868e-01 2.56266538e-02 5.43515384e-01
1.75456631e+00 -6.83016002e-01 4.43881303e-01 -2.35835552e-01
4.19458658e-01 -4.62597132e-01 3.79589558e-01 9.71704781e-01
-4.05720741e-01 2.17338666e-01 1.17542148e+00 -2.32026324e-01
-5.87557554e-01 -1.09063339e+00 -2.10500900e-02 4.40531552e-01
2.64910936e-01 -5.51711321e-01 -7.86140740e-01 -1.20188797e+00
2.14361265e-01 4.16060030e-01 -7.00405478e-01 -6.36199772e-01
-2.33739674e-01 -9.81271803e-01 1.37140775e+00 1.57511041e-01
4.90037143e-01 -8.31929982e-01 -3.34595330e-02 -7.01412908e-04
2.37712055e-01 -1.27190709e+00 -3.82716447e-01 2.50194758e-01
-6.02198184e-01 -1.62278521e+00 1.71416968e-01 -3.96545172e-01
1.02054191e+00 -1.14283167e-01 1.03155768e+00 4.97474611e-01
-2.35066816e-01 2.19804063e-01 -4.53138798e-02 -5.34057736e-01
-5.99028051e-01 1.73457205e-01 1.78196758e-01 7.01168552e-02
3.75349462e-01 -9.77237999e-01 -3.80761534e-01 4.95171547e-01
-1.13796914e+00 -5.41463137e-01 5.94827950e-01 3.73549193e-01
3.50028098e-01 4.69949692e-01 6.60632551e-01 -1.60873854e+00
9.16990340e-01 -6.93665802e-01 -7.01851666e-01 3.35590541e-01
-6.33026659e-01 -7.25991428e-02 8.28638196e-01 -8.35025787e-01
9.74608064e-02 5.51448762e-02 -5.61129265e-02 -4.29438055e-01
-4.09863703e-03 6.08573496e-01 -5.88055372e-01 -6.38231874e-01
8.03971589e-01 -1.00560181e-01 8.79912376e-02 1.83632687e-01
1.30312338e-01 2.14409217e-01 3.04863244e-01 -5.75703502e-01
1.14659762e+00 2.58821994e-01 7.54386187e-01 -7.48731971e-01
-1.73932955e-01 3.72874252e-02 -2.31669381e-01 -1.10388458e-01
4.37734336e-01 -4.63721365e-01 -1.25724244e+00 6.25224710e-01
-8.73902082e-01 -2.10294589e-01 3.05088479e-02 2.30909765e-01
1.59241632e-01 6.20246112e-01 -7.12434411e-01 -6.78141713e-01
-4.40352201e-01 -1.06526554e+00 3.12844962e-01 1.28275931e-01
-1.29547432e-01 -1.26681626e+00 -3.66265655e-01 -6.21079095e-02
2.07451388e-01 9.70604420e-01 9.09627318e-01 -1.43822014e+00
-5.86735427e-01 -7.07331598e-01 -2.59370685e-01 1.20019965e-01
1.65524736e-01 1.04621105e-01 -9.50232208e-01 -5.99605560e-01
-4.09593061e-02 -2.80389249e-01 5.95271111e-01 1.19573906e-01
1.31990695e+00 -7.20507741e-01 -5.04561126e-01 7.34006405e-01
1.37122381e+00 2.09438637e-01 9.11741793e-01 -1.45362318e-01
9.68265235e-01 3.50236326e-01 1.07389554e-01 1.12054110e-01
-1.96693152e-01 1.62491173e-01 9.29406643e-01 -1.03897706e-01
3.81896079e-01 -4.87391084e-01 2.12050453e-01 1.72964051e-01
2.52673745e-01 -8.74412000e-01 -7.65236616e-01 -1.32067949e-01
-1.44837821e+00 -6.97582126e-01 -3.54869753e-01 2.11872721e+00
6.06637239e-01 5.44012129e-01 8.81074555e-03 1.58783987e-01
9.48588371e-01 2.20838904e-01 -9.80755210e-01 -3.05545866e-01
-8.54344442e-02 3.54336619e-01 1.00754595e+00 3.23305815e-01
-1.04783475e+00 8.43766689e-01 5.95601559e+00 9.09196973e-01
-9.52487886e-01 -2.92513907e-01 7.21289873e-01 2.16070607e-01
-3.49822909e-01 1.74151167e-01 -4.31780547e-01 4.88060951e-01
1.01251948e+00 -3.99377257e-01 4.94526178e-01 5.71992338e-01
-4.76017296e-01 2.80209422e-01 -1.17330933e+00 6.80712819e-01
-9.09854248e-02 -1.15781379e+00 1.87029287e-01 7.43992150e-01
4.47104543e-01 -1.25299409e-01 2.27447636e-02 2.36096710e-01
9.28379178e-01 -1.46684480e+00 1.12579815e-01 9.96350050e-02
9.25642788e-01 -1.00118971e+00 7.07499921e-01 3.14118654e-01
-1.02516496e+00 1.99687779e-01 -3.18909764e-01 1.34496570e-01
-3.04140449e-01 6.81588769e-01 -9.02163386e-01 8.69363964e-01
2.43295252e-01 3.79300267e-01 -6.86352134e-01 7.33620286e-01
-4.35747057e-01 7.56459057e-01 -5.66928923e-01 -1.42088667e-01
1.82945222e-01 -9.44229439e-02 5.75125754e-01 8.69444609e-01
-1.90598182e-02 3.48188318e-02 4.27567102e-02 9.47753310e-01
-5.47237039e-01 -2.37899780e-01 -1.00012147e+00 -4.68181044e-01
6.47966981e-01 1.33048236e+00 -1.00126755e+00 1.20598771e-01
-2.00055867e-01 7.19836771e-01 3.34984332e-01 3.72190833e-01
-8.51564884e-01 -6.09213471e-01 7.35877693e-01 2.15121210e-01
1.14477739e-01 5.67066781e-02 1.80914777e-03 -8.52637410e-01
-2.45478496e-01 -1.09302306e+00 6.54100597e-01 -2.81454533e-01
-1.52852011e+00 5.55908620e-01 -2.47432724e-01 -1.01696575e+00
-2.35013012e-02 -7.05933571e-01 -8.37531686e-01 6.95968330e-01
-1.05179143e+00 -9.20647442e-01 -8.40643700e-03 6.87697113e-01
-6.37526155e-01 -7.79639333e-02 8.56675923e-01 2.10477903e-01
-7.06305325e-01 1.08563530e+00 -3.39739799e-01 9.07399952e-01
3.05064172e-01 -9.95287061e-01 7.00938702e-01 1.32611167e+00
2.15132520e-01 8.30833435e-01 5.08809745e-01 -9.47869301e-01
-1.47770417e+00 -1.22282016e+00 3.53699565e-01 -3.93621266e-01
8.24634612e-01 -8.27273071e-01 -9.65552628e-01 7.88813353e-01
-4.74844277e-01 4.67641950e-01 8.42009604e-01 2.09537894e-02
-7.35713661e-01 3.88538418e-03 -1.70226109e+00 6.47349715e-01
1.19715822e+00 -6.80726886e-01 2.08926499e-01 2.87392825e-01
7.88279951e-01 -4.11765277e-01 -1.00132084e+00 5.51219165e-01
3.45465809e-01 -8.54908943e-01 1.10567713e+00 -8.37856293e-01
1.13217145e-01 -2.15848476e-01 6.53660297e-02 -1.15771449e+00
-1.21456131e-01 -8.18656683e-01 -2.34373480e-01 1.14553595e+00
6.09754682e-01 -1.20311785e+00 9.88009453e-01 7.22891986e-01
4.02320832e-01 -5.97912252e-01 -1.00679553e+00 -8.08771610e-01
-1.11630401e-02 -3.11254412e-01 9.34074104e-01 1.20876670e+00
-1.46289811e-01 3.69315326e-01 -2.76710719e-01 7.51796722e-01
8.68003547e-01 -4.00038123e-01 8.87185037e-01 -1.37262714e+00
-1.80364043e-01 -4.72786307e-01 -1.10145462e+00 -2.56471574e-01
2.48168081e-01 -1.23877037e+00 -4.48946059e-01 -1.12013328e+00
-1.84892759e-01 -3.49487782e-01 -4.01868820e-01 5.36980033e-01
-7.17380177e-03 2.50063002e-01 -1.31881936e-02 -1.89870581e-01
-4.04101908e-01 7.03154504e-03 1.12437379e+00 -2.99305379e-01
6.36460409e-02 2.89807677e-01 -8.25304508e-01 6.53752625e-01
1.06260931e+00 -7.31346607e-01 -7.58154631e-01 3.75895947e-01
4.21769619e-01 7.15103224e-02 5.94281256e-01 -9.94063377e-01
4.52806294e-01 -1.40816970e-02 3.40916097e-01 -2.31171906e-01
-4.54199687e-02 -1.13923895e+00 5.87404132e-01 8.02703679e-01
-1.83615938e-01 -1.73209026e-01 2.22639218e-01 9.20878708e-01
1.78395629e-01 -1.09543398e-01 5.95045149e-01 -2.94924770e-02
-9.18527916e-02 8.86239469e-01 -6.64738640e-02 2.46153414e-01
1.07947135e+00 -2.96361566e-01 -7.73198783e-01 -4.24945682e-01
-5.94497263e-01 2.70501643e-01 3.55042815e-01 5.69905341e-02
6.73277438e-01 -1.07840157e+00 -5.19402802e-01 5.01008868e-01
1.25007465e-01 1.69124734e-02 5.23763290e-03 5.12152493e-01
-4.16835278e-01 -3.20059322e-02 -1.84484199e-01 -3.09420496e-01
-1.39005899e+00 8.04797351e-01 4.32905108e-01 -4.72209364e-01
-3.40819418e-01 7.49272406e-01 2.68318474e-01 -4.59331393e-01
2.83411860e-01 -2.18161389e-01 -2.34514475e-02 -3.02849770e-01
2.61890978e-01 2.66316444e-01 6.79311808e-03 -3.89645606e-01
-6.13950133e-01 -5.51211473e-04 -3.08043182e-01 4.52248126e-01
8.63536000e-01 4.51711684e-01 -2.58377373e-01 -4.82051708e-02
1.21441102e+00 2.69781239e-02 -9.21310484e-01 -2.15003744e-01
-9.79841650e-02 -2.50798494e-01 -3.55528206e-01 -6.39865279e-01
-1.40706801e+00 6.79506481e-01 1.70435444e-01 7.39295483e-01
9.80178475e-01 -1.50660187e-01 5.97567856e-01 3.91875088e-01
5.32228112e-01 -3.55960578e-01 -1.58395261e-01 1.78233191e-01
4.37027901e-01 -9.39104557e-01 3.43485534e-01 -7.36536145e-01
-1.57332987e-01 8.36132765e-01 6.57442451e-01 -1.90157548e-01
9.56086099e-01 3.29935551e-01 -4.13049519e-01 -5.05324066e-01
-5.55897176e-01 2.36660242e-01 1.63582802e-01 7.97029316e-01
-1.48692265e-01 2.47928038e-01 1.03320487e-01 6.98537111e-01
-2.71230608e-01 -3.51740539e-01 6.74692690e-01 6.98398232e-01
-3.11745238e-03 -9.67267692e-01 1.75461285e-02 7.60753632e-01
-8.06414545e-01 -1.39167815e-01 -8.96593511e-01 8.62178981e-01
-2.15952218e-01 8.83717716e-01 -2.72620678e-01 -9.07352090e-01
2.70080358e-01 -2.42967606e-01 2.32213110e-01 -4.13058966e-01
-7.10391045e-01 -5.70447385e-01 2.18215212e-01 -5.57895124e-01
-5.91984093e-02 -2.78724551e-01 -1.19200802e+00 -8.95198047e-01
-5.50892949e-01 8.68966430e-02 4.96040285e-01 5.91944098e-01
3.55221629e-01 5.41903019e-01 6.44675374e-01 -1.69533476e-01
-4.70789373e-01 -6.27934992e-01 -7.94649482e-01 3.73364627e-01
1.77701309e-01 -3.46569330e-01 -8.81853402e-01 -7.09055543e-01] | [6.035338878631592, 7.281796932220459] |
353fa992-7760-4fac-846a-8205e9b63edb | a-survey-of-applications-of-artificial | 2107.06179 | null | https://arxiv.org/abs/2107.06179v2 | https://arxiv.org/pdf/2107.06179v2.pdf | Application of artificial intelligence techniques for automated detection of myocardial infarction: A review | Myocardial infarction (MI) results in heart muscle injury due to receiving insufficient blood flow. MI is the most common cause of mortality in middle-aged and elderly individuals around the world. To diagnose MI, clinicians need to interpret electrocardiography (ECG) signals, which requires expertise and is subject to observer bias. Artificial intelligence-based methods can be utilized to screen for or diagnose MI automatically using ECG signals. In this work, we conducted a comprehensive assessment of artificial intelligence-based approaches for MI detection based on ECG as well as other biophysical signals, including machine learning (ML) and deep learning (DL) models. The performance of traditional ML methods relies on handcrafted features and manual selection of ECG signals, whereas DL models can automate these tasks. The review observed that deep convolutional neural networks (DCNNs) yielded excellent classification performance for MI diagnosis, which explains why they have become prevalent in recent years. To our knowledge, this is the first comprehensive survey of artificial intelligence techniques employed for MI diagnosis using ECG and other biophysical signals. | ['U Rajendra Acharya', 'Ru-San Tan', 'Hui Wen Loh', 'Edris Hassannatajjeloudari', 'Mitra Akbari Kohnehshari', 'Samiyeh Khosravi', 'Tahereh Tamadon', 'Amir Mosavi', 'Danial Sharifrazi', 'Roohallah Alizadehsani', 'Sahar Khanjani Shirkharkolaie', 'Zeynab Kiani Zadegan', 'Amir Mashmool', 'Issa Nodehi', 'Sanaz Mojrian', 'Javad Hassannataj Joloudari'] | 2021-07-05 | null | null | null | null | ['electrocardiography-ecg'] | ['methodology'] | [ 4.41397339e-01 -3.40421647e-01 -2.53109783e-01 -4.25517336e-02
-4.61414844e-01 -2.48707980e-01 -3.27069849e-01 3.19991767e-01
-5.62074721e-01 8.80231619e-01 -2.54419029e-01 -5.89033306e-01
-8.07277560e-02 -6.84475541e-01 -8.72602761e-02 -6.14860296e-01
-4.91783321e-01 3.25254619e-01 -2.81541675e-01 1.10629395e-01
1.77712515e-01 6.42252922e-01 -1.21180725e+00 3.69514316e-01
8.15090954e-01 1.11173809e+00 -3.58888239e-01 1.04201484e+00
2.75287926e-01 6.61433995e-01 -7.98004270e-01 -1.52090853e-02
-8.91684741e-02 -8.94296169e-01 -5.98075747e-01 -4.70294118e-01
8.25557974e-04 -3.13246578e-01 -2.43502200e-01 4.74743724e-01
1.28657234e+00 -4.87018406e-01 8.97639930e-01 -9.01838958e-01
-1.52903795e-01 4.82501566e-01 -2.20116943e-01 8.48586142e-01
2.25228757e-01 1.34444982e-01 3.00466388e-01 -8.66936326e-01
1.37041613e-01 4.96218801e-01 1.27627492e+00 5.31641960e-01
-1.09133303e+00 -4.89140093e-01 -3.84296864e-01 3.95773739e-01
-1.35182750e+00 -2.87532866e-01 1.03684425e+00 -5.38569093e-01
8.26231599e-01 3.79834890e-01 1.03165650e+00 8.64231408e-01
6.58106625e-01 5.77066362e-01 9.22511876e-01 -5.36786377e-01
2.81986535e-01 -2.86041424e-02 3.06731731e-01 4.12115037e-01
7.48582125e-01 1.48644775e-01 -3.31313729e-01 -5.11669517e-01
8.95162046e-01 2.28293046e-01 -3.19106251e-01 1.13785632e-01
-1.38461411e+00 6.89541519e-01 4.28652242e-02 4.47276324e-01
-6.88929498e-01 1.24006867e-01 7.39971280e-01 4.86678809e-01
4.91160266e-02 5.77118456e-01 -7.01255262e-01 -2.88595587e-01
-8.61072719e-01 1.02176391e-01 7.16352642e-01 1.13311879e-01
1.86384827e-01 2.36770138e-01 -7.63088465e-02 7.56307006e-01
1.76348582e-01 7.05722928e-01 7.98299015e-01 -1.10916448e+00
1.96490481e-01 7.05360532e-01 -1.34952530e-01 -1.12271202e+00
-8.64879429e-01 -7.80617118e-01 -1.47571206e+00 4.00602728e-01
4.27398652e-01 -6.21776938e-01 -5.81677139e-01 1.23963857e+00
-1.79887980e-01 -5.40381707e-02 3.53436694e-02 7.93635309e-01
1.12301004e+00 -3.91582474e-02 2.25323021e-01 -4.03746694e-01
1.32936823e+00 9.16209258e-03 -7.18169153e-01 -1.73789099e-01
6.80710495e-01 -6.54713884e-02 4.43096697e-01 6.61781430e-01
-8.54093730e-01 -5.81088006e-01 -1.21396554e+00 2.98255473e-01
-1.48508087e-01 3.29825640e-01 6.67358279e-01 9.61382151e-01
-8.30237269e-01 7.08371639e-01 -9.37638342e-01 -4.39411581e-01
5.07367074e-01 6.26615524e-01 -2.04069734e-01 3.77307862e-01
-1.52211237e+00 1.02876854e+00 1.30407363e-01 3.96913767e-01
-4.03097212e-01 -4.65837240e-01 -7.02326298e-01 -4.06201333e-01
-3.75807136e-01 -9.59047019e-01 7.90274441e-01 -1.08219838e+00
-1.31510842e+00 1.10126078e+00 -1.12182818e-01 -6.25934124e-01
4.34796214e-01 -3.39802414e-01 -3.45477223e-01 6.23947680e-01
-1.64794207e-01 2.18746528e-01 6.68487191e-01 -8.06812286e-01
-3.00708592e-01 -5.42870760e-01 -3.30048621e-01 -5.56550883e-02
-5.42569421e-02 3.84560823e-02 3.06389689e-01 -7.79535055e-01
3.93716425e-01 -7.91007817e-01 -3.69262069e-01 -1.03638075e-01
-1.85861513e-02 -1.04064286e-01 4.43593502e-01 -8.93828154e-01
1.49834740e+00 -1.85664070e+00 -1.44484758e-01 2.06585154e-01
7.71211982e-01 5.66973031e-01 4.70860809e-01 3.14805388e-01
-1.59009725e-01 3.52885783e-01 -2.10465178e-01 3.05620432e-01
-3.77051324e-01 1.52612804e-02 3.48149389e-01 5.83113372e-01
1.49676979e-01 1.14420640e+00 -6.33067906e-01 -6.36182189e-01
4.08428431e-01 4.60923731e-01 -5.13372011e-03 9.96457934e-02
5.47049224e-01 7.40398765e-01 -3.77368331e-01 7.52300739e-01
3.62129897e-01 -1.47709012e-01 2.79954165e-01 -4.13769633e-01
3.14491421e-01 2.39796154e-02 -8.94062817e-01 1.41418958e+00
8.35329369e-02 6.70787036e-01 -4.13803250e-01 -1.53522599e+00
1.09754097e+00 9.34609711e-01 8.75921249e-01 -4.93598402e-01
5.29063821e-01 2.89124817e-01 3.05720598e-01 -9.79926169e-01
-7.33232915e-01 -2.66366214e-01 7.13066980e-02 4.22613621e-01
-1.60802230e-01 2.91447461e-01 6.26401305e-02 -2.74481744e-01
1.22117603e+00 -1.14743307e-01 6.83855712e-01 -1.69763297e-01
6.03685617e-01 -9.54684690e-02 7.51663327e-01 1.06371331e+00
-5.86546659e-01 7.97371149e-01 3.04565996e-01 -1.19688725e+00
-5.14507473e-01 -9.55208659e-01 -3.58360261e-01 2.83253968e-01
-1.06786154e-01 4.83075827e-02 -7.07875013e-01 -3.59943837e-01
-9.98737812e-02 7.31829107e-02 -6.00699604e-01 -5.47065198e-01
-8.63567173e-01 -1.23401260e+00 9.99094248e-01 9.21175003e-01
5.89867473e-01 -1.30685663e+00 -1.32948577e+00 6.70543551e-01
-4.44600821e-01 -6.63099825e-01 3.12522024e-01 4.12599564e-01
-1.46177185e+00 -1.20753050e+00 -1.00136209e+00 -7.70079315e-01
2.87409335e-01 -3.62643093e-01 1.26641059e+00 3.41685802e-01
-7.28220463e-01 1.08970203e-01 -2.29690850e-01 -1.00178277e+00
-3.06327313e-01 2.13069126e-01 1.57789126e-01 -3.69289666e-02
7.71949708e-01 -9.21526432e-01 -1.11651945e+00 2.13875752e-02
-3.42475414e-01 -1.17936440e-01 9.06877279e-01 7.56453931e-01
4.16169614e-01 -1.57897800e-01 1.19768715e+00 -7.87337303e-01
6.50221586e-01 -3.46145660e-01 2.56421268e-01 -1.04862414e-01
-7.59700298e-01 -4.22000855e-01 3.28346074e-01 -5.35001159e-01
-4.39858586e-01 1.66313007e-01 -2.16097414e-01 3.68427783e-02
-6.03421748e-01 7.18006015e-01 -2.81699989e-02 -5.01088314e-02
1.02731442e+00 5.08828796e-02 2.33563438e-01 -2.06865087e-01
-4.67910320e-01 9.43856955e-01 6.94879055e-01 -2.29153588e-01
1.98955685e-01 3.77216935e-01 1.58042729e-01 -8.14172208e-01
-6.33863509e-01 -3.94175559e-01 -9.90809262e-01 -4.23830092e-01
9.68062341e-01 -7.09361851e-01 -7.24916458e-01 7.16334939e-01
-9.22417164e-01 -2.36053556e-01 -3.84263620e-02 8.86446774e-01
-3.60429674e-01 3.62051636e-01 -7.61343658e-01 -9.92857695e-01
-1.05207777e+00 -6.77290678e-01 6.69222772e-01 2.82207113e-02
-7.57189810e-01 -9.03603435e-01 1.56599745e-01 1.77341238e-01
4.87233788e-01 9.80394185e-01 9.06331480e-01 -4.59607333e-01
2.88071752e-01 -7.23180115e-01 -8.50713253e-03 3.79061073e-01
2.95021236e-01 -1.91777155e-01 -1.01982713e+00 7.49386176e-02
2.57947445e-01 6.92740455e-02 6.98456824e-01 9.70116138e-01
8.88389945e-01 1.76262870e-01 -4.41534758e-01 5.18538535e-01
1.12148905e+00 7.74487376e-01 9.35752630e-01 4.45092410e-01
5.61088204e-01 6.22157976e-02 -4.23461162e-02 3.74348521e-01
2.23127589e-01 2.01262951e-01 1.25795543e-01 -6.70653999e-01
4.72504906e-02 5.35591960e-01 1.69444419e-02 4.42233294e-01
-6.48947001e-01 1.48335978e-01 -9.89161074e-01 3.08106542e-01
-1.69248867e+00 -1.08973467e+00 -5.11166930e-01 2.24558949e+00
8.94197881e-01 2.54562020e-01 2.66194344e-01 1.17137253e+00
5.67129552e-01 -6.19327188e-01 -8.06526065e-01 -3.59531313e-01
-1.24836899e-01 4.45036381e-01 1.75128594e-01 -1.18966646e-01
-1.41620958e+00 -3.59778367e-02 7.63757563e+00 -1.51097611e-01
-1.40186596e+00 -5.04430849e-03 8.52326751e-01 3.76287699e-01
5.03281415e-01 -5.38960636e-01 -1.69218004e-01 4.33603525e-01
1.12061441e+00 2.25053772e-01 -1.46697909e-01 4.38686401e-01
5.81904650e-01 -2.07547955e-02 -1.11044478e+00 1.38190734e+00
1.18103430e-01 -1.20501196e+00 -3.78823817e-01 -1.46756485e-01
1.38662666e-01 -1.75525516e-01 -3.10592532e-01 1.37282163e-01
-6.09113991e-01 -1.09231877e+00 1.42056227e-01 8.60274017e-01
9.41435397e-01 -5.66674769e-01 1.46656525e+00 1.19142115e-01
-8.74517024e-01 -1.50095269e-01 1.16321601e-01 -5.89336216e-01
-9.11751413e-05 8.66642714e-01 -4.47730690e-01 2.39992127e-01
7.89993525e-01 8.26456845e-01 -4.51756418e-01 1.16520762e+00
-4.36244160e-02 1.10026300e+00 -3.24630708e-01 1.61718175e-01
-5.09662986e-01 9.37192664e-02 4.27544326e-01 1.18546033e+00
8.88281316e-02 1.40905261e-01 2.61991978e-01 7.09340215e-01
3.80699456e-01 2.85455808e-02 -5.10636508e-01 1.02140881e-01
1.24983191e-01 9.20945585e-01 -8.06749582e-01 -5.83902955e-01
-4.27726716e-01 8.59989643e-01 -2.85367221e-01 4.08987969e-01
-5.90097070e-01 -6.82690024e-01 3.37131768e-01 2.82278836e-01
-4.74648446e-01 -8.81427433e-03 -1.02352393e+00 -1.08158743e+00
4.85220999e-02 -8.93887877e-01 6.38462365e-01 -3.74936432e-01
-1.05504453e+00 4.16194171e-01 -1.01142891e-01 -1.32789099e+00
-2.42508247e-01 -4.51831520e-01 -6.77358091e-01 9.19314325e-01
-1.21054554e+00 -5.42618573e-01 -4.29402232e-01 3.12567890e-01
2.73178995e-01 -8.11939165e-02 1.28512549e+00 5.38254678e-01
-7.25921631e-01 4.92377162e-01 -3.86152565e-01 6.34044170e-01
5.18480778e-01 -1.40584242e+00 4.30520177e-02 5.28204203e-01
-2.79999942e-01 4.97612536e-01 4.05494392e-01 -4.72049534e-01
-1.15626466e+00 -8.05405021e-01 1.12888563e+00 -3.97031546e-01
-8.23967308e-02 2.19469249e-01 -6.64137900e-01 2.76640236e-01
-1.98935881e-01 1.06983222e-01 1.05174303e+00 -2.07592517e-01
1.99911162e-01 -2.67457098e-01 -1.05573726e+00 2.61898249e-01
6.58305347e-01 -2.39333019e-01 -7.57616341e-01 -1.97935104e-02
-2.16761112e-01 -2.07153320e-01 -1.23020637e+00 9.26126599e-01
1.13574994e+00 -8.30609024e-01 1.21509874e+00 -4.51013446e-01
8.15712288e-02 -2.89192140e-01 5.38223624e-01 -8.54694664e-01
-5.33929706e-01 -4.13324028e-01 -4.10999507e-01 6.07780516e-01
3.75740826e-01 -7.52582252e-01 6.95062518e-01 5.67481875e-01
7.74403811e-02 -8.08201015e-01 -6.82560623e-01 -3.40535462e-01
-6.09426983e-02 -3.46218348e-01 1.03156626e-01 6.81901872e-01
3.31546754e-01 3.37076575e-01 -2.32030958e-01 -1.00327253e-01
7.03084707e-01 -3.71387564e-02 3.93879443e-01 -1.79863465e+00
-8.10587779e-02 -5.15812755e-01 -9.40300822e-01 -1.65634334e-01
-3.90841573e-01 -6.28351212e-01 -3.20335835e-01 -1.81460381e+00
1.32515997e-01 -3.04784924e-01 -7.94924200e-01 3.87662262e-01
-4.86741364e-01 8.25678051e-01 -3.59702259e-01 3.04083675e-01
-9.38005224e-02 -2.44519353e-01 7.89402187e-01 -1.11174613e-01
-5.77939153e-01 3.41943115e-01 -7.62647271e-01 8.09762478e-01
1.39543688e+00 -5.63404918e-01 -2.97519535e-01 -3.07876375e-02
4.52743083e-01 1.69798255e-01 4.34259951e-01 -1.42852521e+00
-2.85345227e-01 1.31333306e-01 1.17720032e+00 -5.64583898e-01
-2.38026362e-02 -5.60787082e-01 3.83800715e-01 1.08694696e+00
-2.82976717e-01 5.49668036e-02 1.57981608e-02 1.61601782e-01
-2.51772970e-01 -9.84995998e-03 7.69211233e-01 -4.05639142e-01
-3.76681238e-01 1.47433192e-01 -1.21713364e+00 4.94453162e-02
7.40325928e-01 -6.01677239e-01 2.22623438e-01 -3.06922078e-01
-1.06502604e+00 -3.22118372e-01 -2.32167944e-01 -5.87507291e-03
8.73870254e-01 -1.24259067e+00 -8.42126846e-01 1.95699811e-01
6.68999404e-02 -2.41143465e-01 -6.91787899e-03 1.62215936e+00
-1.01374161e+00 1.87785223e-01 -3.48935872e-01 -7.38599718e-01
-1.27974308e+00 3.78864050e-01 8.55128527e-01 -8.04214478e-02
-1.06827390e+00 4.10128623e-01 -2.19860971e-01 1.11084759e-01
4.32613194e-01 -4.72670496e-01 -7.55231261e-01 -6.85965316e-03
8.69415700e-01 5.68757653e-01 3.16776037e-01 -1.79860637e-01
-6.00853503e-01 6.36790454e-01 3.77893925e-01 1.24865465e-01
9.32289898e-01 -2.38008559e-01 -7.62543827e-03 7.33869612e-01
8.12174141e-01 -7.05238163e-01 -4.71361876e-01 1.38247237e-01
-5.68726510e-02 1.90351307e-01 2.43411556e-01 -1.02280462e+00
-1.16153240e+00 1.14744747e+00 1.21020162e+00 1.52477607e-01
1.27718604e+00 -4.67736691e-01 1.01218927e+00 4.14209545e-01
2.90465206e-01 -9.50471997e-01 -1.41888648e-01 -2.77482927e-01
5.73412180e-01 -1.11388636e+00 -8.26266930e-02 -1.49642199e-01
-4.72254932e-01 1.51333189e+00 2.96446115e-01 -2.63665825e-01
1.08937299e+00 2.57526636e-01 6.10713899e-01 -2.20504209e-01
-3.04345638e-01 -9.39277560e-02 1.11892402e-01 9.51294005e-01
9.39388454e-01 2.75682867e-01 -8.24714303e-01 7.82864869e-01
1.86471239e-01 6.79975152e-01 2.28463456e-01 1.21586299e+00
-4.96249944e-01 -9.30921495e-01 -3.80267680e-01 9.30515110e-01
-9.86874819e-01 -9.02751237e-02 -4.38367695e-01 2.36916468e-01
4.30380195e-01 1.25891781e+00 -3.42263252e-01 -2.15105161e-01
1.80986777e-01 3.61067235e-01 4.16132361e-01 -3.01181585e-01
-7.85713017e-01 2.43681259e-02 1.01377517e-01 -1.15374990e-01
-6.22056603e-01 -5.63796699e-01 -1.30325472e+00 2.65658975e-01
-6.22224621e-02 -5.58237098e-02 3.86318296e-01 1.01811373e+00
3.76569271e-01 6.69411838e-01 3.28792751e-01 -7.57028699e-01
-5.85333407e-02 -1.28058302e+00 -7.88565576e-01 3.74593645e-01
3.66080970e-01 -4.84905541e-01 -3.52023482e-01 5.83121657e-01] | [14.310415267944336, 3.270939350128174] |
f15fdea0-e174-4c19-820c-87c78c44e82a | look-remember-and-reason-visual-reasoning | 2306.17778 | null | https://arxiv.org/abs/2306.17778v1 | https://arxiv.org/pdf/2306.17778v1.pdf | Look, Remember and Reason: Visual Reasoning with Grounded Rationales | Large language models have recently shown human level performance on a variety of reasoning tasks. However, the ability of these models to perform complex visual reasoning has not been studied in detail yet. A key challenge in many visual reasoning tasks is that the visual information needs to be tightly integrated in the reasoning process. We propose to address this challenge by drawing inspiration from human visual problem solving which depends on a variety of low-level visual capabilities. It can often be cast as the three step-process of ``Look, Remember, Reason'': visual information is incrementally extracted using low-level visual routines in a step-by-step fashion until a final answer is reached. We follow the same paradigm to enable existing large language models, with minimal changes to the architecture, to solve visual reasoning problems. To this end, we introduce rationales over the visual input that allow us to integrate low-level visual capabilities, such as object recognition and tracking, as surrogate tasks. We show competitive performance on diverse visual reasoning tasks from the CLEVR, CATER, and ACRE datasets over state-of-the-art models designed specifically for these tasks. | ['Roland Memisevic', 'Pulkit Madan', 'Reza Pourreza', 'Mingu Lee', 'Sunny Panchal', 'Apratim Bhattacharyya'] | 2023-06-30 | null | null | null | null | ['object-recognition', 'visual-reasoning', 'visual-reasoning'] | ['computer-vision', 'computer-vision', 'reasoning'] | [-1.16045333e-01 1.32602826e-01 4.88499813e-02 -3.47012997e-01
-4.63467181e-01 -8.80709946e-01 7.98929036e-01 3.10673594e-01
-3.07360023e-01 1.89916089e-01 2.35418305e-01 -7.73890913e-01
6.88467845e-02 -5.93784690e-01 -7.20791161e-01 -1.88684851e-01
2.77267724e-01 5.98266482e-01 4.61776853e-01 -4.05725330e-01
3.07078600e-01 6.78478181e-01 -1.40670693e+00 1.07642806e+00
3.38753581e-01 5.46908617e-01 9.74434391e-02 9.51723695e-01
-5.60790002e-01 1.48929274e+00 -4.80382979e-01 -4.77361590e-01
-3.96111794e-02 -3.82639170e-01 -9.57006514e-01 1.10661173e-02
5.31393945e-01 -1.38845652e-01 3.93755138e-02 7.94236660e-01
1.55568689e-01 -2.23972686e-02 8.17344487e-01 -1.26362944e+00
-1.05048299e+00 4.70580459e-01 -8.42213511e-01 3.35061461e-01
5.90794027e-01 4.41489130e-01 9.77400422e-01 -1.27585673e+00
7.66790748e-01 1.57834554e+00 5.70824683e-01 6.12073421e-01
-1.50453806e+00 -3.52209002e-01 4.66519266e-01 1.96659580e-01
-1.03615832e+00 -4.37296033e-01 5.91098547e-01 -6.54797018e-01
1.34711719e+00 3.54460180e-01 8.86285067e-01 8.60967159e-01
6.08928874e-02 1.02204859e+00 1.10586274e+00 -6.61168993e-01
2.13795200e-01 2.76265860e-01 4.39465255e-01 1.03359890e+00
2.39226520e-01 -9.64059085e-02 -5.72091997e-01 2.29402632e-01
6.02132261e-01 -3.19196694e-02 1.07699148e-02 -9.11500871e-01
-1.11865497e+00 6.70251489e-01 9.23900187e-01 3.03291589e-01
-2.83123523e-01 4.28400904e-01 3.64366263e-01 2.82840341e-01
4.24792245e-02 3.96892130e-01 2.93678734e-02 2.27105543e-01
-8.20932269e-01 3.47479045e-01 7.57244587e-01 8.07856441e-01
5.40324748e-01 8.42593517e-03 -6.94469333e-01 6.26722097e-01
7.31021464e-01 3.33886862e-01 -1.07366249e-01 -9.03022647e-01
5.73808193e-01 9.77236331e-01 9.06049013e-02 -8.00176263e-01
-5.63059092e-01 -1.68379530e-01 -3.98779780e-01 1.03691733e+00
6.53435469e-01 2.67046273e-01 -1.25217211e+00 1.40896606e+00
3.46407324e-01 -5.06633580e-01 1.21583812e-01 9.75279808e-01
1.16198945e+00 5.00688553e-01 5.94952285e-01 4.37359959e-01
1.80626297e+00 -1.18490160e+00 -4.77238536e-01 -6.42737985e-01
4.19266284e-01 -6.77245975e-01 1.41119277e+00 3.18227738e-01
-1.47476935e+00 -6.79330289e-01 -1.10614455e+00 -6.03549063e-01
-7.51389086e-01 1.03540465e-01 6.40351951e-01 5.56558609e-01
-1.27251911e+00 -8.54924098e-02 -6.71549976e-01 -6.66302264e-01
5.73560417e-01 -1.04672156e-01 -3.18708807e-01 -1.26249000e-01
-7.57401586e-01 1.20669365e+00 4.03792530e-01 1.97292089e-01
-8.45088184e-01 -6.58048272e-01 -9.57208812e-01 2.57713825e-01
5.18267989e-01 -1.13419116e+00 1.48468268e+00 -7.71816432e-01
-9.23765659e-01 1.30368555e+00 -8.21691230e-02 -3.85548264e-01
7.68122494e-01 -1.33905813e-01 -8.12533498e-02 3.27440083e-01
-1.04993306e-01 8.03479016e-01 7.73746967e-01 -1.76705658e+00
-4.14131701e-01 -3.53067964e-01 4.96149808e-01 1.76604360e-01
1.50239021e-01 1.72565863e-01 -8.92171204e-01 -4.29259837e-01
-1.99987471e-01 -7.48599529e-01 -9.73704457e-02 5.88996649e-01
-2.06328422e-01 -1.27325416e-01 7.62930214e-01 -5.20447552e-01
1.01446748e+00 -2.08181381e+00 3.79669011e-01 1.26864582e-01
6.59232259e-01 3.52366060e-01 -8.10918659e-02 4.96864647e-01
-6.39569908e-02 1.24373890e-01 2.84463428e-02 -5.02725780e-01
2.88277268e-01 1.07389845e-01 -4.39542085e-01 1.38890475e-01
3.48274797e-01 1.52055240e+00 -8.66275847e-01 -6.88982069e-01
4.75155711e-01 4.79682416e-01 -5.47324955e-01 1.67185366e-01
-5.19718647e-01 4.48099934e-02 -1.73436508e-01 7.58199155e-01
3.55968297e-01 -5.91656864e-01 2.69589812e-01 -3.38579148e-01
-2.82917339e-02 -2.10343935e-02 -9.23661709e-01 1.66580653e+00
-4.03468043e-01 7.93519616e-01 3.51666063e-01 -6.84171796e-01
5.51352680e-01 2.64935568e-02 -2.71252960e-01 -9.98966992e-01
-5.65885790e-02 -3.01214010e-01 2.77268700e-02 -5.16100645e-01
4.12501752e-01 -1.99413359e-01 -1.69470519e-01 5.57708621e-01
-1.50450096e-01 -2.15485245e-01 5.18344879e-01 6.96180642e-01
9.04696107e-01 4.80516315e-01 5.27566969e-01 4.31928225e-02
4.74536806e-01 5.60488820e-01 -1.25814304e-01 8.98067057e-01
-1.12063199e-01 4.99231130e-01 7.97843337e-01 -5.42778075e-01
-1.07656860e+00 -1.18331921e+00 2.65660465e-01 1.35501015e+00
-5.41480593e-02 -5.74075401e-01 -5.29825270e-01 -7.00901747e-01
2.44123071e-01 7.59581983e-01 -9.85309899e-01 -7.52903372e-02
-5.06728232e-01 -2.83588797e-01 5.39599061e-01 9.88996863e-01
2.50503808e-01 -1.49096346e+00 -1.25099874e+00 -1.67557597e-01
4.38743383e-01 -9.60753858e-01 5.63900024e-02 2.61145979e-01
-3.91667515e-01 -1.07635820e+00 -6.33269846e-01 -6.32491529e-01
6.85355604e-01 4.01131213e-01 1.55759907e+00 4.12488192e-01
-6.09053373e-01 6.68538034e-01 -1.33156970e-01 -4.52119917e-01
-4.41059679e-01 -3.56463552e-01 -5.94635129e-01 -3.31461698e-01
3.22726786e-01 -8.35536867e-02 -5.07375658e-01 -2.33474541e-02
-8.60800087e-01 5.99283099e-01 7.66963780e-01 5.91621518e-01
3.99755061e-01 -5.08362651e-01 -4.48172577e-02 -9.24829423e-01
8.28180254e-01 -2.59623468e-01 -4.41125035e-01 6.99773788e-01
-2.60091424e-01 3.94217044e-01 5.42471468e-01 -2.76099950e-01
-1.02487075e+00 4.92564030e-02 1.28848910e-01 -5.56975901e-01
-1.41349524e-01 4.55208212e-01 1.39761284e-01 9.19298902e-02
6.19521022e-01 1.88299809e-02 -2.46514305e-01 -2.78752148e-01
1.04026937e+00 2.52388209e-01 5.93749583e-01 -6.89886987e-01
8.47191989e-01 5.02181232e-01 7.51169622e-02 -5.61889648e-01
-8.81189346e-01 -3.61715645e-01 -7.12436557e-01 -4.26414877e-01
1.26161230e+00 -8.63530815e-01 -1.17718279e+00 4.51736599e-02
-1.26287007e+00 -6.04906678e-01 -2.63045847e-01 -1.09714538e-01
-4.36845720e-01 1.35039285e-01 -5.44762671e-01 -9.35309947e-01
-2.95728415e-01 -1.04452145e+00 1.08947825e+00 3.87379199e-01
-4.62515712e-01 -9.27073717e-01 -9.58687589e-02 4.68048751e-01
4.44250464e-01 1.55553803e-01 1.30518293e+00 -3.91539931e-01
-8.28554034e-01 1.06816934e-02 -9.08273697e-01 -1.74697787e-01
-4.14353073e-01 4.14658077e-02 -9.85443532e-01 -1.59888655e-01
-3.33181113e-01 -7.54532874e-01 9.73665297e-01 -8.17967281e-02
8.90143037e-01 2.05000073e-01 -4.46329206e-01 5.28121948e-01
1.31031096e+00 1.00531533e-01 4.47592169e-01 2.72160381e-01
1.01239836e+00 5.10226488e-01 3.33464921e-01 -1.98714174e-02
6.81023359e-01 7.03828871e-01 3.84099215e-01 -4.36345488e-01
-5.57258248e-01 -2.92072803e-01 8.18280205e-02 1.11781865e-01
-1.15777720e-02 -1.28351972e-01 -1.51573217e+00 4.20843244e-01
-2.10890222e+00 -1.19604623e+00 -1.10892534e-01 1.67056620e+00
5.82557619e-01 3.90019625e-01 1.89655602e-01 -7.83044379e-04
1.10862412e-01 3.29403788e-01 -3.95928651e-01 -7.98639238e-01
2.84492224e-01 -5.36617599e-02 4.63421941e-02 4.95499998e-01
-7.40453899e-01 1.06800795e+00 6.88378572e+00 2.15079963e-01
-1.03566337e+00 -9.42260101e-02 2.99064308e-01 -2.18870372e-01
-4.01013613e-01 1.16546437e-01 -6.73897922e-01 -1.34689540e-01
6.14861012e-01 1.73571482e-01 5.45806766e-01 7.48747349e-01
-1.32731348e-01 -2.55707115e-01 -1.32168531e+00 1.05785823e+00
4.42871660e-01 -1.47035336e+00 3.50979984e-01 -2.64341384e-01
6.07050508e-02 -3.96182656e-01 4.97535542e-02 7.30485439e-01
7.86250591e-01 -1.33454466e+00 1.12583029e+00 6.42305136e-01
6.33747578e-01 -5.30533075e-01 9.12282430e-03 3.00578237e-01
-1.54743588e+00 -2.21153215e-01 -5.87252937e-02 -4.32546996e-02
3.16912681e-01 -1.74325034e-02 -8.23523343e-01 4.29026276e-01
8.29585314e-01 5.48856914e-01 -1.09244180e+00 7.80308127e-01
-5.19866943e-01 -5.87080233e-02 1.16980091e-01 -1.39188468e-01
2.47356713e-01 1.38973668e-01 1.15578599e-01 1.36799455e+00
-2.39614636e-01 7.18469237e-05 2.19803855e-01 1.26211703e+00
1.53193370e-01 -1.43873796e-01 -6.59405231e-01 -1.41144633e-01
1.01932302e-01 1.30213594e+00 -9.97475386e-01 -4.10614312e-01
-7.64346063e-01 9.97759521e-01 1.02416372e+00 4.91741300e-01
-9.65183079e-01 -1.28802210e-01 3.33506405e-01 2.72359431e-01
4.89201188e-01 -3.04078519e-01 -3.65175724e-01 -1.06415868e+00
3.38922255e-02 -1.04422688e+00 6.22469544e-01 -1.62566268e+00
-1.00579119e+00 5.27586281e-01 2.46887337e-02 -7.40932167e-01
-1.72828957e-01 -1.07374454e+00 -4.41175371e-01 9.49086070e-01
-1.35834694e+00 -1.63745224e+00 -5.73674500e-01 7.87042677e-01
5.74728370e-01 1.50304928e-01 6.75598145e-01 -6.12350218e-02
-2.80621678e-01 1.93447128e-01 -6.58784211e-01 3.23069572e-01
5.05195618e-01 -1.48188472e+00 6.77586675e-01 8.98267448e-01
4.61713314e-01 8.99036825e-01 6.61629856e-01 -5.34663796e-01
-1.64724851e+00 -6.19573593e-01 7.31922090e-01 -1.09666610e+00
6.72878742e-01 -9.40611839e-01 -8.48547876e-01 9.31162715e-01
4.92996812e-01 2.43581921e-01 3.54600221e-01 2.52385885e-01
-8.96824956e-01 1.45940751e-01 -8.13561201e-01 1.02730310e+00
1.03604352e+00 -8.73684049e-01 -1.14785874e+00 4.56961151e-03
4.23693627e-01 -4.99980360e-01 -2.85927176e-01 5.01548685e-02
7.19725490e-01 -1.05506623e+00 1.56825209e+00 -1.13515890e+00
5.19278646e-01 -5.72055876e-01 -3.68598700e-02 -9.84804034e-01
-4.86669868e-01 -2.21450403e-01 -3.64355832e-01 1.06111884e+00
4.35278684e-01 -1.03756770e-01 3.52098405e-01 6.57993376e-01
1.31544262e-01 -5.20009637e-01 -2.43138596e-01 -4.09563720e-01
-8.19768310e-02 -4.58727717e-01 2.17323542e-01 6.23260617e-01
2.03025132e-01 6.66914940e-01 -2.12679759e-01 2.68746652e-02
4.34127927e-01 4.44985390e-01 1.06064105e+00 -1.21390939e+00
-3.63791555e-01 -6.58779621e-01 -1.46901906e-01 -9.13957477e-01
1.92363677e-03 -1.02209127e+00 -9.17417407e-02 -2.36797428e+00
4.10015643e-01 -1.09755285e-01 -1.87860221e-01 7.19376981e-01
-2.56767839e-01 2.37907216e-01 7.82391548e-01 1.05923600e-01
-1.00055635e+00 -1.82202123e-02 1.30422580e+00 -2.48421013e-01
-7.06697349e-03 -5.01968622e-01 -7.93266714e-01 7.69933224e-01
3.64496976e-01 -1.82473525e-01 -6.07025921e-01 -5.80677271e-01
6.23537660e-01 1.38742596e-01 6.66055024e-01 -9.57029939e-01
4.03474867e-01 -2.42158070e-01 8.82197201e-01 -6.26798272e-01
5.99965096e-01 -8.85644794e-01 2.57600658e-02 4.23498541e-01
-4.36065078e-01 5.37386179e-01 5.45967519e-01 4.97341990e-01
6.08823914e-03 9.95433331e-02 7.57291317e-01 -4.58717763e-01
-1.15710104e+00 -2.96893775e-01 -3.65268081e-01 1.44410893e-01
1.32026744e+00 -2.92925954e-01 -8.04402888e-01 -1.35518789e-01
-9.27720964e-01 4.05154198e-01 6.79659128e-01 6.46893322e-01
5.74935257e-01 -1.27271640e+00 -4.19628441e-01 1.34764910e-01
5.65613329e-01 -1.87698215e-01 1.56371444e-01 7.45091438e-01
-7.51677334e-01 5.32087028e-01 -4.85683531e-01 -6.49733663e-01
-1.49058819e+00 1.31412482e+00 3.14729154e-01 -2.98979938e-01
-8.44060600e-01 7.56278634e-01 4.28480834e-01 -2.70701885e-01
2.96650767e-01 -5.98246813e-01 -2.82293111e-01 3.30584049e-01
8.50505769e-01 -4.85430993e-02 -1.37318954e-01 -4.59232420e-01
-6.72497272e-01 7.62636185e-01 -3.44572030e-02 -3.23075742e-01
1.04098988e+00 -7.81608000e-02 8.56062993e-02 6.14585698e-01
6.65013433e-01 7.17736781e-02 -1.20543122e+00 1.79096982e-02
6.42961562e-02 -2.62487471e-01 -2.45427638e-01 -1.19894731e+00
-8.81606102e-01 1.19718468e+00 2.61852741e-01 3.71425062e-01
7.37765729e-01 3.29699308e-01 -1.21362932e-01 3.74766737e-01
1.33788124e-01 -7.88282335e-01 3.71659517e-01 3.88293236e-01
1.12819743e+00 -1.20976424e+00 1.10138237e-01 -1.21610045e-01
-9.81352329e-01 1.10977387e+00 7.19412386e-01 1.29135117e-01
2.40096375e-01 5.12628615e-01 4.42537516e-01 -7.39015520e-01
-1.03756726e+00 -3.84786844e-01 6.14824951e-01 5.04518569e-01
5.83613813e-01 -1.70493796e-01 1.49044782e-01 2.03522995e-01
2.49755591e-01 5.04617542e-02 9.02059823e-02 1.18438876e+00
-4.63500500e-01 -8.42288256e-01 -5.35301208e-01 5.47161028e-02
-2.66912542e-02 -2.55782515e-01 -7.07145214e-01 1.06193614e+00
-2.51036227e-01 7.18724430e-01 3.62561233e-02 1.57248989e-01
5.44999182e-01 3.35191399e-01 6.69269741e-01 -7.09857106e-01
-8.43586445e-01 -3.62136334e-01 2.05579504e-01 -7.86592185e-01
-2.80411124e-01 -3.55049521e-01 -1.52338076e+00 -2.39680707e-01
4.82720464e-01 -3.24528068e-01 4.79066700e-01 8.93002331e-01
2.07815066e-01 8.93875360e-01 -6.18303001e-01 -8.70000839e-01
-2.45701119e-01 -5.19973874e-01 -1.35057405e-01 7.40062892e-01
4.40358460e-01 -5.56208491e-01 2.38728393e-02 8.07324126e-02] | [10.821497917175293, 1.9917234182357788] |
6f9ed357-08ea-44b7-b7e5-e9be3a6a1817 | a-hierarchical-pose-based-approach-to-complex | 1606.04992 | null | http://arxiv.org/abs/1606.04992v1 | http://arxiv.org/pdf/1606.04992v1.pdf | A Hierarchical Pose-Based Approach to Complex Action Understanding Using Dictionaries of Actionlets and Motion Poselets | In this paper, we introduce a new hierarchical model for human action
recognition using body joint locations. Our model can categorize complex
actions in videos, and perform spatio-temporal annotations of the atomic
actions that compose the complex action being performed.That is, for each
atomic action, the model generates temporal action annotations by estimating
its starting and ending times, as well as, spatial annotations by inferring the
human body parts that are involved in executing the action. our model includes
three key novel properties: (i) it can be trained with no spatial supervision,
as it can automatically discover active body parts from temporal action
annotations only; (ii) it jointly learns flexible representations for motion
poselets and actionlets that encode the visual variability of body parts and
atomic actions; (iii) a mechanism to discard idle or non-informative body parts
which increases its robustness to common pose estimation errors. We evaluate
the performance of our method using multiple action recognition benchmarks. Our
model consistently outperforms baselines and state-of-the-art action
recognition methods. | ['Ivan Lillo', 'Juan Carlos Niebles', 'Alvaro Soto'] | 2016-06-15 | a-hierarchical-pose-based-approach-to-complex-1 | http://openaccess.thecvf.com/content_cvpr_2016/html/Lillo_A_Hierarchical_Pose-Based_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Lillo_A_Hierarchical_Pose-Based_CVPR_2016_paper.pdf | cvpr-2016-6 | ['action-understanding'] | ['computer-vision'] | [ 3.12842757e-01 2.78983712e-01 -5.46038568e-01 -2.05312863e-01
-5.31391382e-01 -4.14073527e-01 6.87485874e-01 -1.45455480e-01
-1.92422062e-01 5.28880417e-01 9.67589915e-01 4.29822475e-01
6.23052381e-02 -2.53151059e-01 -7.60393918e-01 -6.04664624e-01
-5.32478154e-01 6.61521554e-01 6.34801447e-01 -6.67491853e-02
-1.91697665e-02 5.62984169e-01 -1.58121598e+00 4.60729808e-01
3.61698538e-01 1.13064659e+00 -1.55662820e-01 8.53715122e-01
3.96228373e-01 1.35166526e+00 -5.65179527e-01 -2.91014969e-01
2.29674816e-01 -6.86673224e-01 -1.14348352e+00 5.38942695e-01
5.32031894e-01 -6.41008139e-01 -6.33051872e-01 3.34500641e-01
2.30096087e-01 3.80575985e-01 6.39017344e-01 -1.22433162e+00
-4.96622622e-02 3.00291747e-01 -5.66357315e-01 1.61503643e-01
7.13721514e-01 4.14960474e-01 9.89524961e-01 -4.52539563e-01
8.41783524e-01 1.26216900e+00 7.23096907e-01 6.56380415e-01
-1.08194602e+00 -1.90002471e-01 4.91762608e-01 1.76452771e-01
-1.11420691e+00 -5.98959565e-01 5.10115743e-01 -6.98732853e-01
1.07031047e+00 -3.03044394e-02 1.03276646e+00 1.20874941e+00
3.80226433e-01 1.29607451e+00 4.12569791e-01 -2.25062281e-01
1.05523393e-01 -1.02032948e+00 -1.84456497e-01 1.05158353e+00
-2.45242909e-01 -1.35202095e-01 -8.72643292e-01 -1.62852079e-01
9.96573746e-01 1.05008923e-01 7.47185946e-02 -7.09470451e-01
-1.67825484e+00 3.89773518e-01 8.20376426e-02 4.15353589e-02
-5.98290026e-01 6.98955417e-01 5.69888592e-01 -2.12393790e-01
2.74086505e-01 7.95726106e-02 -5.02022862e-01 -5.61629891e-01
-7.59374380e-01 3.60130966e-01 5.93817055e-01 8.05886388e-01
4.38877821e-01 -8.67234766e-02 -6.77227616e-01 4.34473604e-01
1.36793673e-01 3.53403926e-01 3.70826900e-01 -1.35288763e+00
5.01854658e-01 7.22692072e-01 3.00419062e-01 -7.54052520e-01
-5.03207386e-01 1.58910334e-01 -5.01677334e-01 -1.36076771e-02
4.60158855e-01 1.25266194e-01 -1.00389862e+00 1.64385927e+00
6.09354973e-01 2.74251074e-01 -3.28830808e-01 7.93353677e-01
6.23453915e-01 1.95014268e-01 3.37904304e-01 4.55618054e-02
1.37119985e+00 -1.27078366e+00 -6.81882799e-01 -4.53856915e-01
6.83237731e-01 -3.43293518e-01 5.66089749e-01 2.13334873e-01
-1.12942398e+00 -7.37631261e-01 -7.22377896e-01 -1.89126983e-01
1.14675917e-01 4.95645493e-01 8.45465839e-01 2.56425679e-01
-6.52634919e-01 7.16435373e-01 -1.50051129e+00 -3.63184810e-01
3.68733644e-01 3.43637615e-01 -6.13909364e-01 2.09589303e-01
-7.93094575e-01 6.67906821e-01 3.46892655e-01 4.89342175e-02
-1.22896385e+00 -2.29586095e-01 -1.29725218e+00 -2.35445812e-01
6.65216863e-01 -7.15394855e-01 1.41364193e+00 -1.20742333e+00
-1.64598036e+00 7.40536332e-01 -3.73393565e-01 -4.35117662e-01
3.94404680e-01 -7.42132783e-01 -2.41437584e-01 5.09406924e-01
2.10032463e-01 8.90528440e-01 8.63922417e-01 -6.41648948e-01
-6.38138771e-01 -4.73371208e-01 -1.81142930e-02 3.34038407e-01
2.12992460e-01 7.19765127e-02 -9.69877541e-01 -8.79991114e-01
3.39317441e-01 -1.07797694e+00 -2.10630327e-01 2.99975932e-01
-3.92622560e-01 -2.86217093e-01 6.69900358e-01 -8.92401636e-01
1.28514755e+00 -2.11367083e+00 6.56238914e-01 -4.78964224e-02
1.27538562e-01 9.13435686e-03 -6.51581064e-02 3.96144271e-01
1.90021530e-01 -3.01030278e-01 -9.10408124e-02 -4.52546716e-01
6.96680620e-02 5.29846549e-01 -2.40749821e-01 6.51736081e-01
1.31704822e-01 1.22755301e+00 -9.90280569e-01 -7.63678968e-01
4.19751078e-01 2.81452596e-01 -5.02681613e-01 2.57276416e-01
-3.81276369e-01 8.67579162e-01 -5.77907860e-01 1.00726366e+00
-6.43692017e-02 -3.35569859e-01 4.05435830e-01 -2.48608276e-01
1.42891362e-01 3.91239077e-01 -1.29772127e+00 1.86376715e+00
1.26199722e-01 2.84495085e-01 -8.73628855e-02 -7.36756325e-01
3.53557676e-01 4.13361818e-01 1.17995882e+00 -3.61117661e-01
-1.08924270e-01 -1.99881464e-01 -2.83927441e-01 -5.96124291e-01
2.01325923e-01 2.33209476e-01 -2.60441184e-01 5.68983197e-01
3.76057267e-01 3.36867779e-01 2.81405360e-01 7.88506586e-03
1.37353528e+00 9.59941506e-01 7.69106746e-01 2.15823978e-01
3.75713319e-01 -1.62490532e-01 8.24275136e-01 5.95248699e-01
-4.21624511e-01 4.34352607e-01 4.89300281e-01 -8.83590698e-01
-7.08174050e-01 -1.13102305e+00 5.05734146e-01 1.46608257e+00
1.61612242e-01 -7.30544984e-01 -4.72853661e-01 -9.37802315e-01
-7.12733418e-02 -2.30840389e-02 -8.44358861e-01 -3.45939286e-02
-8.97125244e-01 -2.49852035e-02 6.64468825e-01 1.13899624e+00
6.02283061e-01 -1.28530753e+00 -1.09961855e+00 1.39253795e-01
-5.39248407e-01 -1.16191733e+00 -8.40130806e-01 -6.33169338e-02
-1.01129150e+00 -1.42660570e+00 -5.72263956e-01 -4.62312460e-01
5.72286844e-01 -1.23161770e-01 1.22535598e+00 4.72517572e-02
-4.18787628e-01 7.37334490e-01 -4.82450396e-01 -2.36700103e-02
-9.14808661e-02 -2.38744259e-01 1.57857999e-01 2.13942811e-01
1.86284825e-01 -3.97635847e-01 -8.12088072e-01 6.84225857e-01
-5.25042832e-01 2.31108770e-01 5.82692921e-01 6.16626918e-01
8.99663568e-01 -2.77794182e-01 -2.19827309e-01 -3.63599688e-01
-1.87663689e-01 -9.19912532e-02 -1.71846226e-01 3.74505818e-01
3.10919493e-01 1.68176368e-01 2.07579762e-01 -4.64367032e-01
-1.01695228e+00 7.74898887e-01 1.07615881e-01 -4.68945116e-01
-4.14140850e-01 6.41808957e-02 -2.58824557e-01 2.24549040e-01
4.39687401e-01 3.39884192e-01 -9.01738461e-03 -5.29209375e-01
5.00703633e-01 -2.85860896e-02 8.04209530e-01 -6.84626400e-01
5.08341014e-01 8.64029884e-01 2.02043325e-01 -8.39916408e-01
-9.03985500e-01 -6.89919591e-01 -1.44432855e+00 -5.53983629e-01
1.34234059e+00 -9.12975192e-01 -8.89302731e-01 6.43632948e-01
-1.01025057e+00 -6.02260768e-01 -3.55220437e-01 6.53480113e-01
-9.88192379e-01 7.45472074e-01 -7.72806704e-01 -9.01339889e-01
1.12224957e-02 -9.21867192e-01 1.75544274e+00 -1.29941270e-01
-7.65345335e-01 -7.10171163e-01 2.02479601e-01 4.58551884e-01
-3.59802991e-01 6.14191651e-01 3.88659984e-01 -2.49622181e-01
-6.15712106e-01 -1.01858035e-01 3.42291266e-01 1.45314515e-01
2.04072192e-01 4.27463911e-02 -4.28055108e-01 -2.03337520e-01
-6.25400662e-01 -6.94583535e-01 8.45772684e-01 6.37333095e-01
1.16047084e+00 -5.44765651e-01 -5.60616910e-01 6.25669301e-01
4.65628207e-01 1.73351821e-02 8.26237082e-01 2.12549735e-02
8.50171804e-01 6.37136102e-01 9.15038526e-01 6.22160971e-01
3.21418196e-01 1.14836717e+00 4.65543300e-01 2.38388795e-02
-1.54530570e-01 -5.33670962e-01 6.29736245e-01 2.76274711e-01
-6.90794945e-01 2.23892741e-02 -6.72294974e-01 4.60682720e-01
-2.23920226e+00 -1.33430338e+00 1.09563246e-01 2.24475479e+00
6.89870358e-01 -1.88216511e-02 7.92865396e-01 -4.42304946e-02
3.72346729e-01 5.17812669e-01 -6.07012570e-01 1.18041657e-01
2.03183830e-01 2.59778500e-01 4.26987767e-01 1.20899864e-01
-1.66745567e+00 9.89612222e-01 6.83456755e+00 3.59438181e-01
-6.06780887e-01 -4.98589762e-02 2.12950364e-01 -2.73996174e-01
3.56827527e-01 -2.44171932e-01 -6.38719499e-01 2.53801137e-01
2.72617012e-01 4.80729759e-01 5.16085066e-02 8.84603381e-01
7.07440600e-02 -1.62483424e-01 -1.35460985e+00 9.14315045e-01
2.05995455e-01 -1.10359025e+00 1.80348884e-02 1.07591771e-01
6.22815609e-01 -1.90428048e-01 -4.30528134e-01 1.95827648e-01
4.30889308e-01 -9.61014509e-01 8.93567979e-01 7.38507986e-01
6.27692401e-01 -4.56573963e-01 3.36647213e-01 3.42826992e-01
-1.57538724e+00 -2.17463255e-01 1.87057465e-01 -2.83739746e-01
4.18377131e-01 1.32566355e-02 -3.63056362e-01 4.71778840e-01
6.26460373e-01 1.12306285e+00 -5.76976955e-01 8.58726025e-01
-5.89155257e-01 5.51475883e-01 -2.96484530e-01 3.67151082e-01
5.09653129e-02 5.20147085e-02 3.99199545e-01 1.00413930e+00
1.34634286e-01 2.84385324e-01 6.86714947e-01 3.66171807e-01
3.13390017e-01 -2.51496017e-01 -5.32738090e-01 -2.25511119e-01
1.48370162e-01 8.08466971e-01 -7.46252537e-01 -2.87439436e-01
-3.82215798e-01 1.50532520e+00 2.54323632e-01 2.49941692e-01
-1.01121795e+00 2.23748833e-01 9.56848502e-01 2.09694639e-01
5.42111576e-01 -4.51321840e-01 1.71401367e-01 -1.16278005e+00
2.42767230e-01 -8.59311283e-01 8.54207218e-01 -6.82605088e-01
-8.06863964e-01 6.11837171e-02 1.31787449e-01 -1.36201453e+00
-5.74813545e-01 -6.12118483e-01 -3.49644691e-01 2.83420593e-01
-4.54505116e-01 -1.44788134e+00 -1.84310198e-01 6.90345585e-01
7.34136879e-01 1.09001234e-01 7.84249246e-01 1.12751694e-02
-4.14734036e-01 2.82985896e-01 -5.76679170e-01 5.13522089e-01
5.00241995e-01 -1.02834892e+00 5.90220451e-01 8.34150136e-01
5.81990600e-01 4.37534750e-01 3.69072944e-01 -9.95778501e-01
-1.23926210e+00 -9.44350481e-01 8.29137385e-01 -9.05976176e-01
4.12494391e-01 -2.75454134e-01 -3.86572123e-01 1.13177538e+00
-4.20284957e-01 1.13437526e-01 6.26416266e-01 2.00378761e-01
-3.87809277e-01 3.80604640e-02 -5.74304938e-01 6.44302011e-01
1.61528003e+00 -4.52664077e-01 -7.81349599e-01 4.48151588e-01
3.42569768e-01 -6.81489944e-01 -6.31027162e-01 5.55867434e-01
9.91171956e-01 -1.02862346e+00 1.31136358e+00 -9.84820247e-01
5.65502822e-01 -4.10236806e-01 7.78952539e-02 -6.22920513e-01
-5.60499847e-01 -5.82701504e-01 -8.92173707e-01 6.39101505e-01
-4.96951006e-02 4.05677706e-02 8.19386661e-01 4.73852545e-01
-1.62980214e-01 -8.70051742e-01 -1.11040437e+00 -8.54147375e-01
-7.09808350e-01 -3.40074837e-01 3.39672536e-01 5.53298414e-01
-2.63339840e-02 -1.24896690e-02 -8.63380790e-01 -8.02468695e-03
3.88324291e-01 1.71007678e-01 1.14243233e+00 -1.12412357e+00
-7.92528510e-01 -2.33714879e-01 -8.11115205e-01 -1.54671788e+00
2.22785026e-01 -2.99675971e-01 2.46526420e-01 -1.53499973e+00
2.13860869e-01 3.66922140e-01 -1.12173222e-01 1.03197038e+00
-3.47788967e-02 3.08829814e-01 5.20975403e-02 3.47814441e-01
-1.20594716e+00 4.85903651e-01 1.22985458e+00 -9.71462578e-02
-1.53449431e-01 2.59276778e-01 -2.86815669e-02 9.83110368e-01
3.54299009e-01 -2.24029019e-01 -3.60103577e-01 -2.35111296e-01
1.98151870e-03 3.37213755e-01 7.67801464e-01 -1.26935649e+00
-4.20062207e-02 -3.16538632e-01 6.36511922e-01 -7.09140122e-01
5.68602502e-01 -5.67934513e-01 2.65377134e-01 6.41414762e-01
-3.29341948e-01 -1.03887543e-01 -1.28781602e-01 8.19229424e-01
-2.10415106e-02 2.92579681e-01 4.63045418e-01 -3.96052271e-01
-1.11702728e+00 4.83911425e-01 -4.13099796e-01 9.94711444e-02
1.18745542e+00 -4.55671757e-01 8.19102004e-02 -4.92939591e-01
-1.05738902e+00 1.78692415e-01 4.89551365e-01 6.30271256e-01
3.26350302e-01 -1.42928720e+00 -3.28680426e-01 -1.67346746e-02
3.64372939e-01 -2.55555391e-01 2.92359501e-01 1.12070334e+00
-5.11405289e-01 2.99936235e-01 -2.76409626e-01 -7.29135573e-01
-1.65758359e+00 5.24683058e-01 3.46637011e-01 -3.64390582e-01
-1.13348198e+00 8.25354040e-01 3.74155879e-01 -1.16948336e-01
5.83228827e-01 -4.13564205e-01 -1.55290648e-01 -1.88938484e-01
5.01665175e-01 5.56897759e-01 -3.86480570e-01 -1.13662636e+00
-6.25294864e-01 5.50821662e-01 3.69946241e-01 -1.69599399e-01
1.04391062e+00 4.82433811e-02 -2.34714411e-02 4.90584731e-01
7.48713911e-01 -1.99415177e-01 -1.82979429e+00 -9.24680233e-02
9.95189510e-03 -5.47237396e-01 -5.14798462e-01 -7.91446269e-01
-8.04390430e-01 6.28587663e-01 1.93659246e-01 -4.37940508e-01
1.10376179e+00 4.10246760e-01 7.90201008e-01 3.34518373e-01
5.72157145e-01 -1.42374539e+00 5.31128109e-01 5.96171677e-01
8.48473012e-01 -9.06456709e-01 2.23185807e-01 -4.51388508e-01
-9.17605281e-01 9.39578414e-01 6.37549996e-01 1.38752371e-01
3.88485283e-01 2.57159751e-02 -1.88947663e-01 -4.83693749e-01
-5.38564026e-01 -5.51798522e-01 7.99267769e-01 6.29333615e-01
5.03226340e-01 1.10352315e-01 -2.58434862e-01 4.69696194e-01
1.49282172e-01 1.02137677e-01 -2.33901680e-01 1.17980719e+00
-2.98062325e-01 -9.35500562e-01 -2.47515902e-01 2.41115659e-01
-3.59956026e-01 4.29595053e-01 -8.71881723e-01 9.26567912e-01
3.64320397e-01 6.07225895e-01 4.01561558e-02 -3.98299724e-01
3.29337806e-01 3.62024069e-01 8.79023969e-01 -7.15879798e-01
-2.97975272e-01 9.67855603e-02 2.36642152e-01 -1.44215131e+00
-7.01004148e-01 -1.01052356e+00 -1.34330928e+00 2.31490046e-01
2.75800139e-01 -2.06645742e-01 1.64211225e-02 1.17333484e+00
3.73268098e-01 4.58541542e-01 1.01967022e-01 -1.26681840e+00
-3.65990520e-01 -8.87983382e-01 -5.32199919e-01 7.72488713e-01
2.08026245e-01 -9.88199711e-01 -9.88187641e-03 5.20297587e-01] | [8.120256423950195, 0.45924386382102966] |
395c747f-f5e7-406d-b40b-c07d8b44eb68 | hierarchical-strategies-for-cooperative-multi | 2212.07397 | null | https://arxiv.org/abs/2212.07397v1 | https://arxiv.org/pdf/2212.07397v1.pdf | Hierarchical Strategies for Cooperative Multi-Agent Reinforcement Learning | Adequate strategizing of agents behaviors is essential to solving cooperative MARL problems. One intuitively beneficial yet uncommon method in this domain is predicting agents future behaviors and planning accordingly. Leveraging this point, we propose a two-level hierarchical architecture that combines a novel information-theoretic objective with a trajectory prediction model to learn a strategy. To this end, we introduce a latent policy that learns two types of latent strategies: individual $z_A$, and relational $z_R$ using a modified Graph Attention Network module to extract interaction features. We encourage each agent to behave according to the strategy by conditioning its local $Q$ functions on $z_A$, and we further equip agents with a shared $Q$ function that conditions on $z_R$. Additionally, we introduce two regularizers to allow predicted trajectories to be accurate and rewarding. Empirical results on Google Research Football (GRF) and StarCraft (SC) II micromanagement tasks show that our method establishes a new state of the art being, to the best of our knowledge, the first MARL algorithm to solve all super hard SC II scenarios as well as the GRF full game with a win rate higher than $95\%$, thus outperforming all existing methods. Videos and brief overview of the methods and results are available at: https://sites.google.com/view/hier-strats-marl/home. | ['Ammar Fayad', 'Majd Ibrahim'] | 2022-12-14 | null | null | null | null | ['starcraft'] | ['playing-games'] | [-3.83388132e-01 3.26709241e-01 -3.84271950e-01 -4.33033891e-02
-4.61527675e-01 -4.84231532e-01 5.71999490e-01 -1.50995031e-01
-4.89262700e-01 9.56010640e-01 5.82306720e-02 -3.71395618e-01
-6.63644671e-01 -9.26792026e-01 -8.34877968e-01 -6.77776992e-01
-7.11710572e-01 8.15940142e-01 6.08429983e-02 -6.37556374e-01
2.19488040e-01 1.39828786e-01 -1.34010673e+00 -6.73176795e-02
1.04552650e+00 8.15121889e-01 2.40573287e-01 6.43042624e-01
4.22495246e-01 1.43282318e+00 -3.65318328e-01 -7.39774823e-01
6.58132136e-01 -2.51897603e-01 -7.78946579e-01 -6.05946109e-02
-2.74374157e-01 -3.90251458e-01 -7.50199020e-01 9.58840191e-01
1.10019527e-01 5.90665340e-01 5.46664953e-01 -1.46381903e+00
-5.98068893e-01 1.11482239e+00 -4.97639447e-01 2.27133632e-01
4.26481694e-01 7.48768151e-01 1.42043161e+00 -2.36295775e-01
6.11537337e-01 9.88749683e-01 3.43828499e-01 5.87783635e-01
-1.13433647e+00 -5.31675398e-01 6.63862705e-01 3.14334869e-01
-1.22623110e+00 -1.27875313e-01 6.90882862e-01 -3.92346084e-01
1.08340061e+00 6.37709796e-02 8.83497953e-01 1.09009182e+00
1.21444732e-01 1.03406811e+00 7.81675875e-01 9.29804370e-02
3.29601139e-01 -3.01899582e-01 8.09762329e-02 9.03066754e-01
8.35431814e-02 3.58295709e-01 -6.02423370e-01 -1.37227684e-01
7.11497784e-01 9.85065475e-02 -1.53890520e-01 -4.58114505e-01
-8.25599790e-01 1.01672614e+00 5.51464617e-01 1.70562398e-02
-6.11712277e-01 6.31762147e-01 4.58747298e-02 4.67509270e-01
1.10723853e-01 8.71622086e-01 -3.14642996e-01 -4.36251462e-01
-3.87987494e-01 6.15173280e-01 7.92409658e-01 9.32771325e-01
8.31227422e-01 2.30887011e-02 -1.34397283e-01 3.43687803e-01
2.18701884e-01 1.03412330e-01 1.02176115e-01 -1.45319295e+00
6.26834333e-01 8.69124949e-01 5.75984120e-01 -1.14684892e+00
-5.78251064e-01 -6.90141499e-01 -4.34813261e-01 3.04588079e-01
3.46236259e-01 -2.11470827e-01 -4.00867850e-01 2.18186784e+00
-5.18293492e-02 4.26381350e-01 8.80618244e-02 8.12701106e-01
3.10399204e-01 7.02542961e-01 -1.02626577e-01 -2.70677805e-01
8.74869287e-01 -1.05517435e+00 -3.83120984e-01 -4.69930023e-01
7.57156849e-01 2.93819681e-02 1.19565213e+00 3.62322032e-01
-1.45539296e+00 -1.22357316e-01 -7.54568219e-01 2.84947097e-01
-1.09063819e-01 -5.72657101e-02 1.07532763e+00 4.93045807e-01
-1.35392582e+00 5.78387082e-01 -1.02680099e+00 -9.54079032e-02
5.06911457e-01 6.83797061e-01 -9.15695950e-02 1.20747387e-01
-1.22415340e+00 8.42898130e-01 2.96242535e-01 -3.88333052e-02
-1.49767351e+00 -4.26314265e-01 -7.04606295e-01 1.65068775e-01
1.11033809e+00 -7.46399045e-01 1.35364211e+00 -7.91021526e-01
-1.89434397e+00 6.16177678e-01 2.89960682e-01 -8.92248631e-01
6.11298144e-01 -1.89641446e-01 2.14092866e-01 2.65715038e-03
2.95075566e-01 4.21709001e-01 5.02119124e-01 -1.05140674e+00
-7.10786045e-01 -1.88314766e-01 6.94691241e-01 6.80654705e-01
-2.12345004e-01 -2.51967251e-01 -3.61158431e-01 -2.87487656e-01
-5.14099777e-01 -9.09780800e-01 -8.11176002e-01 -6.38252914e-01
-3.32154214e-01 -5.58558166e-01 1.61887646e-01 -3.76591325e-01
1.29445767e+00 -1.77109170e+00 6.71099663e-01 3.80042374e-01
5.76563895e-01 2.03366354e-02 -3.45671207e-01 7.12947965e-01
4.41575080e-01 8.64205062e-02 -1.08330147e-02 -4.90098149e-01
2.80823231e-01 8.48802999e-02 -1.69121668e-01 5.23192406e-01
-3.03431451e-01 1.10161197e+00 -9.86490786e-01 1.69202745e-01
5.86428382e-02 -6.92397207e-02 -1.10486746e+00 3.63036960e-01
-6.02548778e-01 3.53051871e-01 -7.54518926e-01 1.51870772e-01
9.37377289e-02 -3.71171087e-01 5.85567057e-01 5.98664820e-01
-1.36166111e-01 4.46321249e-01 -9.50536251e-01 1.55743241e+00
-2.11166516e-01 2.55709499e-01 3.35539460e-01 -1.30402267e+00
7.51487017e-01 4.75376695e-02 6.63349688e-01 -4.92252797e-01
4.41951007e-01 1.63681973e-02 1.22951418e-01 -1.62088498e-01
4.66479808e-01 2.80116141e-01 -5.29774189e-01 4.53979760e-01
4.55680527e-02 -1.38486177e-02 4.64999378e-01 7.09273160e-01
1.38047087e+00 1.11110479e-01 2.50557274e-01 -3.11596215e-01
3.30111772e-01 7.77090788e-02 7.45321393e-01 1.29973674e+00
-2.63049603e-01 -3.30872148e-01 1.04882002e+00 -4.05450761e-01
-8.52005601e-01 -7.69150257e-01 9.11796749e-01 1.31564784e+00
4.76814598e-01 -7.53327668e-01 -7.07084239e-01 -6.12262249e-01
-8.91777128e-02 9.59952354e-01 -8.75904024e-01 -3.32039565e-01
-7.33074307e-01 -4.15331095e-01 1.34027541e-01 4.34128642e-01
5.58451533e-01 -1.07734549e+00 -6.79307997e-01 2.32594803e-01
-9.28284004e-02 -7.70323575e-01 -6.75337315e-01 -4.66544367e-02
-3.11104327e-01 -1.16507816e+00 -2.20361188e-01 -5.37853301e-01
3.99948090e-01 1.35314077e-01 1.11953664e+00 1.71499088e-01
-3.83327827e-02 6.46390438e-01 -4.06465083e-01 1.06767425e-02
-6.41631559e-02 3.19483429e-01 3.23848575e-01 -1.01533756e-01
1.39376894e-01 -7.07750976e-01 -8.34893942e-01 2.95716107e-01
-2.60791659e-01 2.44649798e-01 5.15800536e-01 6.86993778e-01
6.97219744e-02 2.33021051e-01 3.47294033e-01 -6.00930154e-01
8.55904818e-01 -6.91621006e-01 -1.02992499e+00 -9.92958434e-04
-4.21441138e-01 -4.65201661e-02 8.81463826e-01 -3.19512010e-01
-8.52203488e-01 -1.26950681e-01 6.86518475e-02 -3.88831705e-01
8.25476497e-02 5.53484261e-01 -9.15033370e-02 2.61168361e-01
5.57538450e-01 2.33200997e-01 1.46019146e-01 -7.83602297e-02
5.35300255e-01 1.46758020e-01 2.06610799e-01 -7.20143080e-01
7.92334914e-01 3.54579955e-01 -1.50310665e-01 -4.61019933e-01
-6.83351219e-01 -9.79813635e-02 -2.74699833e-02 -4.79772151e-01
7.48741329e-01 -7.60210395e-01 -1.82611144e+00 4.12287176e-01
-6.88254535e-01 -1.09671021e+00 -2.90970534e-01 3.77332509e-01
-1.12954390e+00 1.04510203e-01 -7.15145171e-01 -1.07695436e+00
7.69990757e-02 -1.26099622e+00 4.09306258e-01 3.07828188e-01
1.82484582e-01 -8.17191958e-01 1.82995558e-01 5.37448943e-01
4.35106009e-01 -1.07039355e-01 3.88015032e-01 -6.05044365e-01
-1.05380762e+00 1.11624494e-01 2.15986714e-01 -3.08175862e-01
-2.54759699e-01 -3.10409188e-01 -8.95900950e-02 -5.71145415e-01
-1.52894214e-01 -4.92702007e-01 8.74777675e-01 6.21582985e-01
9.05850410e-01 -7.41216838e-01 -4.53114510e-01 5.77812195e-01
1.01296604e+00 5.42582691e-01 3.49284053e-01 3.06100667e-01
3.42345208e-01 5.34024060e-01 5.77833116e-01 7.73945391e-01
8.87326598e-01 8.91440928e-01 7.94208646e-01 3.96716774e-01
4.41659033e-01 -5.37217796e-01 7.44826198e-01 2.56619453e-01
-5.14265120e-01 -4.42860246e-01 -9.03303742e-01 2.84826994e-01
-2.33698630e+00 -1.33596456e+00 2.77208716e-01 2.03169942e+00
6.25559270e-01 2.39228770e-01 6.46815717e-01 -4.90318269e-01
5.17883539e-01 1.10882595e-01 -7.67638147e-01 -8.27587023e-02
1.16727062e-01 -1.15153559e-01 5.84482849e-01 9.98627245e-01
-9.38527226e-01 1.40778995e+00 5.54854822e+00 9.48646903e-01
-6.30823076e-01 7.71583766e-02 7.83795774e-01 -4.82301950e-01
-2.85233498e-01 2.45621409e-02 -8.52671146e-01 5.02329171e-01
7.27801383e-01 -2.86552131e-01 1.21351969e+00 9.40963686e-01
4.96574908e-01 2.39783153e-02 -8.86510909e-01 7.92069197e-01
-1.01339221e-01 -1.58805263e+00 -2.20811516e-01 3.79300535e-01
6.77996457e-01 -5.85768232e-03 2.44859993e-01 7.65713274e-01
1.42778301e+00 -1.22759604e+00 8.08052957e-01 5.33832788e-01
3.22763830e-01 -9.72650886e-01 4.72103745e-01 7.58138657e-01
-1.24935436e+00 -6.66748703e-01 -6.74659982e-02 -6.13691509e-01
1.78866848e-01 -2.39086419e-01 -5.24935067e-01 5.20447314e-01
6.22512639e-01 9.25158620e-01 -2.08998799e-01 6.22855246e-01
-4.41741109e-01 4.03292000e-01 -3.35558981e-01 -3.37139785e-01
6.39830470e-01 -5.40112793e-01 6.96606219e-01 4.96087462e-01
2.41899893e-01 5.83306313e-01 6.76351786e-01 8.59600425e-01
-5.91903105e-02 -1.96189675e-02 -6.34469807e-01 -1.28885701e-01
5.02282321e-01 9.15706456e-01 -5.24116516e-01 -9.51874703e-02
-2.46361911e-01 7.47251987e-01 8.81569266e-01 3.74557406e-01
-9.73144233e-01 8.46018866e-02 9.62870479e-01 9.63426009e-03
1.92888990e-01 -3.72011781e-01 3.68651226e-02 -1.38807201e+00
-3.65838140e-01 -8.52960348e-01 6.10974371e-01 -3.95704716e-01
-1.23245108e+00 5.72282135e-01 -6.03392944e-02 -9.58971560e-01
-4.33213383e-01 -5.60194612e-01 -7.06000924e-01 3.33031297e-01
-1.18317223e+00 -1.00301373e+00 -1.75938495e-02 7.34564900e-01
5.22843361e-01 -5.88356793e-01 3.12387764e-01 -2.46031463e-01
-8.90024185e-01 4.88743037e-01 3.46409492e-02 -1.06688716e-01
4.28261049e-03 -1.24653101e+00 1.85507983e-01 8.08424473e-01
-5.04673533e-02 6.05573416e-01 8.27852130e-01 -6.68084979e-01
-1.51892126e+00 -7.40967929e-01 1.85440004e-01 -4.25239742e-01
8.75295818e-01 -3.98545891e-01 -3.16090375e-01 1.15167236e+00
3.03868055e-01 -4.05875057e-01 5.81278563e-01 2.10986122e-01
4.18564938e-02 -1.38476808e-02 -7.74623394e-01 1.12492287e+00
1.46856308e+00 -8.16173255e-02 -3.02581906e-01 3.63968551e-01
7.06333458e-01 -2.99386770e-01 -5.12513578e-01 1.24734960e-01
1.49420857e-01 -1.23651731e+00 9.89317775e-01 -7.18432546e-01
3.26855898e-01 -4.48393971e-02 -1.05963141e-01 -1.40946960e+00
-4.88137066e-01 -1.26646101e+00 -2.34844595e-01 6.49720967e-01
4.41378027e-01 -7.99472332e-01 1.27044034e+00 6.30784810e-01
-3.03630203e-01 -8.68087173e-01 -9.29831088e-01 -7.43400574e-01
2.16801703e-01 -5.12608171e-01 4.19665158e-01 6.45116508e-01
4.91241664e-01 2.08363861e-01 -8.97078693e-01 2.21536875e-01
7.60444462e-01 1.49268031e-01 9.36749876e-01 -8.26554418e-01
-8.18176270e-01 -9.35447752e-01 -1.35138994e-02 -1.50020790e+00
4.69180405e-01 -8.10568869e-01 -1.20595835e-01 -1.48365939e+00
2.36089230e-01 -5.09888828e-01 -1.49116442e-01 3.90476912e-01
7.69817317e-03 -5.12682609e-02 5.00197530e-01 2.69470215e-01
-1.15306818e+00 8.21634948e-01 1.03872311e+00 6.52354956e-02
-5.83673000e-01 3.79004836e-01 -8.81880403e-01 7.80062854e-01
1.11934924e+00 -4.29248475e-02 -5.82302630e-01 -1.82364076e-01
5.85757256e-01 4.90811259e-01 1.90642148e-01 -6.56266272e-01
4.28158015e-01 -7.59243608e-01 -4.20215666e-01 -2.37991810e-01
7.04838574e-01 -4.68536586e-01 8.03691521e-03 7.67738461e-01
-6.33936703e-01 2.60922965e-02 -3.78301412e-01 6.65706754e-01
2.29480132e-01 6.73411116e-02 6.16960108e-01 -3.79056782e-01
-6.02789760e-01 6.94378436e-01 -6.78409457e-01 -1.23911165e-02
1.52902949e+00 7.25162700e-02 -3.48668367e-01 -1.03266907e+00
-6.35123849e-01 9.67233121e-01 4.14604485e-01 4.26214747e-02
4.89206731e-01 -1.03022897e+00 -6.73344374e-01 -2.91019548e-02
-2.26012602e-01 -2.46851236e-01 4.42068517e-01 9.04955804e-01
-3.71661037e-01 4.27845687e-01 -2.64253080e-01 4.00623120e-02
-7.92952180e-01 7.38298535e-01 6.03358746e-01 -7.88417757e-01
-4.49087709e-01 1.04493952e+00 2.60838002e-01 -4.00209606e-01
3.02640855e-01 4.41898480e-02 -4.19328123e-01 -2.43478045e-01
3.38986576e-01 5.24255276e-01 -6.69396222e-01 -4.51992810e-01
-3.19057763e-01 1.61080658e-01 -1.70848414e-01 -3.20296913e-01
1.47657073e+00 -2.35656008e-01 1.26216531e-01 -4.14864160e-02
4.89382178e-01 -6.82314783e-02 -1.72181892e+00 -2.41818503e-01
-6.98229745e-02 -5.15548527e-01 -2.50974000e-01 -6.17691576e-01
-1.08745754e+00 4.59712505e-01 -1.85351729e-01 6.04824603e-01
8.41608405e-01 2.65461653e-01 3.31095278e-01 5.23231745e-01
7.65264452e-01 -9.35239375e-01 5.59359550e-01 7.45185614e-01
7.53482640e-01 -9.10913408e-01 -3.58723760e-01 -9.59570929e-02
-1.04384112e+00 5.62910795e-01 8.92664671e-01 -6.22984350e-01
4.75306451e-01 1.49315953e-01 -3.44404995e-01 -3.67554754e-01
-1.26134074e+00 -4.46253210e-01 -2.22797886e-01 4.27228898e-01
-1.09004945e-01 1.89587414e-01 -4.31391597e-01 9.26186979e-01
-4.21646982e-01 -1.78714752e-01 6.04168952e-01 7.07538128e-01
-5.23202598e-01 -1.08858776e+00 9.63447914e-02 5.70892096e-01
-3.25535059e-01 1.17827870e-01 -3.38645816e-01 5.03011882e-01
-2.09697202e-01 1.05433226e+00 3.19203176e-02 -5.16043484e-01
1.27491042e-01 -4.07208860e-01 2.33417392e-01 -5.09195447e-01
-7.40012228e-01 -9.68681797e-02 7.87911639e-02 -1.08685529e+00
-1.80021048e-01 -6.14947736e-01 -9.85337734e-01 -7.55784929e-01
6.68835342e-02 4.08084929e-01 -1.23084851e-01 7.43029118e-01
4.41354662e-01 3.37774545e-01 7.26015151e-01 -1.13414609e+00
-6.89646959e-01 -5.95278680e-01 -6.11146867e-01 3.44432235e-01
6.24820851e-02 -1.02076960e+00 -4.35781538e-01 -5.33400536e-01] | [3.7915091514587402, 1.8117890357971191] |
6cdbe47a-db5d-467b-a22c-073449645a8c | ppt-token-pruned-pose-transformer-for | 2209.08194 | null | https://arxiv.org/abs/2209.08194v1 | https://arxiv.org/pdf/2209.08194v1.pdf | PPT: token-Pruned Pose Transformer for monocular and multi-view human pose estimation | Recently, the vision transformer and its variants have played an increasingly important role in both monocular and multi-view human pose estimation. Considering image patches as tokens, transformers can model the global dependencies within the entire image or across images from other views. However, global attention is computationally expensive. As a consequence, it is difficult to scale up these transformer-based methods to high-resolution features and many views. In this paper, we propose the token-Pruned Pose Transformer (PPT) for 2D human pose estimation, which can locate a rough human mask and performs self-attention only within selected tokens. Furthermore, we extend our PPT to multi-view human pose estimation. Built upon PPT, we propose a new cross-view fusion strategy, called human area fusion, which considers all human foreground pixels as corresponding candidates. Experimental results on COCO and MPII demonstrate that our PPT can match the accuracy of previous pose transformer methods while reducing the computation. Moreover, experiments on Human 3.6M and Ski-Pose demonstrate that our Multi-view PPT can efficiently fuse cues from multiple views and achieve new state-of-the-art results. | ['Xiaohui Xie', 'Hao Tang', 'Xiangyi Yan', 'Xingwei Liu', 'Liangjian Chen', 'Deying Kong', 'Yifei Chen', 'Zhe Wang', 'Haoyu Ma'] | 2022-09-16 | null | null | null | null | ['3d-human-pose-estimation', '2d-human-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-1.49640650e-01 -3.31393480e-01 1.20476313e-01 -1.22923128e-01
-8.04226995e-01 -2.67733842e-01 2.37976119e-01 -4.10840720e-01
-3.64936382e-01 4.90226865e-01 1.71220288e-01 5.24122179e-01
2.14701846e-01 -4.12649751e-01 -7.60209203e-01 -5.65235376e-01
3.10614705e-01 5.00144541e-01 6.99993551e-01 -1.71721995e-01
-2.75498420e-01 3.25564355e-01 -1.60170281e+00 3.14782560e-01
5.65942705e-01 9.74053979e-01 4.06487107e-01 4.87870544e-01
6.22657239e-01 5.90975165e-01 -6.32054090e-01 -4.84652519e-01
2.57910788e-01 -1.11959316e-01 -5.52330077e-01 4.99058694e-01
7.76221812e-01 -4.92798120e-01 -5.11720955e-01 8.27929378e-01
8.08621526e-01 5.57896420e-02 8.87194350e-02 -1.26427245e+00
-2.62537718e-01 5.45372814e-02 -1.10546649e+00 3.07587057e-01
8.07540774e-01 2.05355316e-01 5.61149776e-01 -1.17290771e+00
5.70264995e-01 1.58058619e+00 6.92669153e-01 3.23787510e-01
-8.14476550e-01 -5.89407861e-01 5.93064010e-01 4.63372797e-01
-1.39761353e+00 -1.90318629e-01 6.80681050e-01 -2.23750249e-01
8.13102305e-01 3.44048440e-01 1.06085312e+00 9.81076717e-01
2.16077521e-01 1.20019138e+00 1.26564133e+00 -2.09597915e-01
-5.13579845e-01 -1.70963973e-01 -2.98819959e-01 1.07922268e+00
3.47452760e-01 -3.40543948e-02 -8.20240915e-01 -4.14050519e-02
9.13834155e-01 2.35830441e-01 -5.87424099e-01 -6.86577559e-01
-1.58345699e+00 2.96583831e-01 4.11200434e-01 -7.28857964e-02
-3.62118274e-01 2.31309369e-01 2.27381915e-01 -3.46717238e-01
5.42990088e-01 -4.04880680e-02 -4.66375113e-01 1.04692079e-01
-6.38572037e-01 3.75523508e-01 1.89960733e-01 9.78270233e-01
5.49808621e-01 -2.46971264e-01 -1.63152009e-01 7.32124746e-01
2.80548304e-01 9.48345304e-01 1.54894799e-01 -8.90998304e-01
6.17936075e-01 6.56869769e-01 3.37557614e-01 -1.22355354e+00
-5.70449412e-01 -6.70641124e-01 -7.59126723e-01 -1.68418631e-01
4.05147612e-01 4.76086996e-02 -8.85405123e-01 1.45618153e+00
9.04209375e-01 -1.83459129e-02 -4.73746479e-01 1.36292195e+00
7.36614943e-01 5.40134907e-01 -1.78480208e-01 -8.86460170e-02
1.60260820e+00 -1.29454815e+00 -5.89105010e-01 -6.69740200e-01
9.14105251e-02 -5.70721209e-01 7.56327927e-01 5.53473651e-01
-1.17918062e+00 -8.75636160e-01 -7.60175884e-01 -1.85837224e-01
-2.30869446e-02 4.57517743e-01 3.92849386e-01 4.07308042e-01
-7.26107955e-01 1.18877493e-01 -9.17054653e-01 -3.71218592e-01
1.89322740e-01 2.21578464e-01 -5.54476380e-01 -3.04536521e-01
-1.09968770e+00 8.94675851e-01 2.22147778e-01 5.44934273e-01
-8.80885065e-01 -3.86854231e-01 -9.45627928e-01 -1.21182360e-01
9.50164080e-01 -1.05359960e+00 1.01854944e+00 -4.88565832e-01
-1.17708981e+00 8.38386953e-01 -3.96114081e-01 -1.86892658e-01
6.31053150e-01 -8.67098331e-01 -2.71801144e-01 5.05570650e-01
2.78161347e-01 6.86444879e-01 8.71803522e-01 -1.40085912e+00
-9.79134500e-01 -8.02673757e-01 8.42890795e-03 6.45301759e-01
-8.71927813e-02 2.30694357e-02 -1.20592844e+00 -5.84161103e-01
3.78742337e-01 -1.08215916e+00 -2.08612308e-01 2.76370179e-02
-4.48601812e-01 -1.88034326e-01 7.43222594e-01 -9.30683315e-01
1.19873548e+00 -1.78059781e+00 7.10652590e-01 8.65513235e-02
4.80840683e-01 5.79568520e-02 2.75880933e-01 9.78060290e-02
2.53453851e-01 -4.84823674e-01 1.11324228e-01 -4.92783129e-01
-1.58323258e-01 1.32995635e-01 2.41202131e-01 7.76470006e-01
-1.29521728e-01 1.03111386e+00 -7.20455170e-01 -8.73547256e-01
4.96832311e-01 4.93368924e-01 -4.97822016e-01 2.63882279e-01
1.30306765e-01 5.51306307e-01 -3.45858157e-01 9.09170866e-01
7.28478968e-01 -5.41440725e-01 8.18342716e-02 -6.22559369e-01
-2.88445819e-02 -3.30785453e-01 -1.12456667e+00 1.81814182e+00
-2.25234181e-01 2.11447656e-01 2.24642426e-01 -6.42168403e-01
4.68598872e-01 1.59247726e-01 6.15938246e-01 -3.80369931e-01
7.88818225e-02 -5.66722527e-02 -3.31729203e-01 -4.40326363e-01
5.21136820e-01 2.65016913e-01 -1.67114854e-01 1.32039431e-02
-3.80354673e-02 3.98005188e-01 1.51002049e-01 7.18192086e-02
6.16903126e-01 4.61614519e-01 4.90106136e-01 7.50439763e-02
6.93046987e-01 -2.57228136e-01 8.66175294e-01 3.53348106e-01
-4.92892295e-01 9.19136405e-01 1.66988209e-01 -5.19326806e-01
-6.72284722e-01 -1.14233112e+00 3.92671049e-01 1.11732876e+00
5.90942740e-01 -5.51614702e-01 -8.42636108e-01 -8.48194957e-01
-2.56111138e-02 -1.54269487e-01 -6.42561316e-01 7.23273531e-02
-8.98938537e-01 -5.28546095e-01 2.50535667e-01 7.77300537e-01
8.09522271e-01 -6.50527716e-01 -1.09834194e+00 9.56005696e-03
-1.03257453e+00 -1.41767156e+00 -9.58083212e-01 -3.24033111e-01
-5.51132441e-01 -1.31418133e+00 -1.22415376e+00 -7.00564742e-01
6.15968525e-01 7.87547946e-01 9.80097592e-01 -1.61423475e-01
-3.88761312e-01 7.32457638e-01 -2.53003567e-01 -8.47480074e-02
3.94037902e-01 -1.91856608e-01 2.37845451e-01 2.71508396e-01
1.98693916e-01 -2.01694056e-01 -1.04085612e+00 7.04439223e-01
-2.81691164e-01 1.91209063e-01 7.74806559e-01 9.05312121e-01
7.29140580e-01 -7.89642259e-02 1.01896018e-01 -5.51513076e-01
-1.28489971e-01 -6.82863966e-02 -3.94249320e-01 5.43994367e-01
-1.63554460e-01 -3.77357662e-01 1.43176392e-01 -4.02340502e-01
-1.27121961e+00 3.74351770e-01 2.09169745e-01 -9.04745460e-01
-3.84336291e-03 1.20417990e-01 -6.32840395e-01 -1.12091258e-01
2.69813895e-01 2.87019610e-01 -5.39611578e-02 -4.48978156e-01
2.80837297e-01 3.25723529e-01 7.17549562e-01 -3.98516983e-01
7.95103610e-01 7.46020019e-01 -1.34359255e-01 -8.08950424e-01
-1.06932843e+00 -7.71568358e-01 -7.99429774e-01 -6.15821302e-01
1.12604642e+00 -1.37645805e+00 -7.85907507e-01 6.51597500e-01
-1.16752851e+00 1.82270840e-01 2.17605576e-01 4.26509917e-01
-4.53339309e-01 7.06504405e-01 -5.25596499e-01 -8.26590061e-01
-1.98329017e-01 -1.24232900e+00 1.70817828e+00 1.58139363e-01
-5.31396307e-02 -6.09927058e-01 -2.93612838e-01 9.61825430e-01
-1.55426160e-01 3.29898506e-01 9.19135362e-02 4.16880883e-02
-8.34136248e-01 -1.41902819e-01 -1.85123414e-01 8.38509481e-03
-2.85278242e-02 -5.30723214e-01 -8.88562322e-01 -5.13925552e-01
-6.14325926e-02 -1.83640659e-01 8.68212163e-01 5.60534239e-01
1.02572179e+00 1.10257072e-02 -7.52967596e-01 6.22237146e-01
9.23053741e-01 -1.04220562e-01 4.50504422e-01 3.16747993e-01
1.27140903e+00 6.66182816e-01 1.08525395e+00 5.48562407e-01
6.19999170e-01 1.13042414e+00 3.38575304e-01 -2.75554627e-01
-1.14369571e-01 -3.11318845e-01 4.63993281e-01 6.58383310e-01
-4.34349775e-01 -9.72086936e-02 -7.14317262e-01 3.96572381e-01
-1.97049189e+00 -9.69169855e-01 1.76805034e-02 2.11547899e+00
3.27967703e-01 -4.76974100e-02 4.93297189e-01 9.76224616e-03
9.16377544e-01 2.05413222e-01 -4.60647047e-01 6.17860436e-01
7.55617116e-03 -2.54271954e-01 5.34371734e-01 2.97451645e-01
-1.47532439e+00 7.30066776e-01 5.59396267e+00 8.42054486e-01
-6.25331044e-01 2.81411469e-01 4.54830080e-01 -4.06866193e-01
1.64608091e-01 -3.59285593e-01 -9.99288678e-01 3.86601955e-01
6.42638654e-02 2.13760406e-01 1.34189427e-01 7.99686790e-01
-7.43102329e-03 -3.98129344e-01 -8.78047824e-01 1.51319778e+00
6.12351000e-01 -6.45979643e-01 -6.56881556e-02 9.94818881e-02
6.22150838e-01 -4.38241512e-01 9.21386778e-02 1.46146938e-01
-2.15041310e-01 -6.89766586e-01 9.60532725e-01 4.75947171e-01
5.66599309e-01 -9.81164277e-01 6.68793261e-01 4.49425042e-01
-1.83644080e+00 -1.81146130e-01 -3.12572539e-01 1.83151215e-01
5.24594784e-01 4.56560671e-01 -3.90671909e-01 8.83383930e-01
1.18074954e+00 8.00429523e-01 -8.52068186e-01 8.60207677e-01
-8.79078209e-02 -1.38384566e-01 -3.67809743e-01 2.10925594e-01
-1.47525147e-01 8.28054994e-02 6.18912280e-01 8.88388038e-01
3.17381829e-01 2.81654596e-01 7.01023698e-01 4.86052155e-01
3.36550295e-01 -1.76558226e-01 -3.44948798e-01 4.98649895e-01
2.68966496e-01 1.31573880e+00 -7.92386115e-01 -4.30132180e-01
-4.22368020e-01 1.41501498e+00 3.34801644e-01 3.27218294e-01
-1.25794208e+00 4.32642661e-02 5.01193285e-01 3.31940114e-01
6.58098459e-01 -1.76039860e-01 2.01221496e-01 -1.44734061e+00
4.14134711e-01 -1.03747022e+00 4.61690038e-01 -8.68269563e-01
-9.86189723e-01 5.63832223e-01 3.11218679e-01 -1.46013856e+00
-5.64926155e-02 -5.41266382e-01 6.79091811e-02 5.16948581e-01
-1.03412807e+00 -1.54939473e+00 -5.97618103e-01 6.83750331e-01
6.46264613e-01 9.26362723e-02 2.55087346e-01 4.28573489e-01
-5.68910658e-01 6.12208188e-01 -4.58801299e-01 1.08672537e-01
8.04237127e-01 -1.09091556e+00 3.83646280e-01 1.02367926e+00
1.05731107e-01 6.07086301e-01 4.95977849e-01 -8.24101865e-01
-1.46732461e+00 -1.06912339e+00 5.80297828e-01 -7.30960965e-01
-9.59258247e-03 -2.59466410e-01 -5.95460534e-01 7.89594173e-01
1.15556709e-01 1.34303689e-01 2.43472189e-01 4.95540649e-02
-9.80477780e-02 -2.29718357e-01 -8.32514167e-01 4.71488178e-01
1.39362097e+00 -2.17868492e-01 -4.78764772e-01 2.58957446e-01
5.20884275e-01 -9.67755377e-01 -7.87235677e-01 5.57050526e-01
7.39487529e-01 -1.26554072e+00 1.49419641e+00 -6.87359050e-02
1.03345439e-01 -5.43255687e-01 -1.09421581e-01 -1.13383877e+00
-4.57213968e-01 -4.51181233e-01 -4.21363056e-01 9.52218235e-01
-1.39509976e-01 -5.53419769e-01 7.98017502e-01 2.51615018e-01
-9.47460309e-02 -8.91426504e-01 -9.74380493e-01 -7.45935082e-01
-5.80237269e-01 -8.41581598e-02 3.22741330e-01 4.52627212e-01
-2.05552787e-01 2.44558930e-01 -8.86322558e-01 4.51084882e-01
1.07553554e+00 2.91473180e-01 1.07666731e+00 -1.06376398e+00
-5.73770046e-01 9.36863571e-03 -4.53252167e-01 -1.55811107e+00
-1.27919614e-01 -3.21149796e-01 1.59378394e-01 -1.48473251e+00
5.37260234e-01 1.20580822e-01 -4.51299995e-02 2.39305437e-01
-6.72499418e-01 4.28382695e-01 4.01628494e-01 1.41898945e-01
-1.00348735e+00 6.48106933e-01 1.48482144e+00 3.15426700e-02
2.28319347e-01 -7.43235275e-03 -4.52024579e-01 1.05927300e+00
4.01211113e-01 -2.04676747e-01 -3.01626086e-01 -3.69172662e-01
-2.37238985e-02 1.42333731e-01 8.19501519e-01 -1.38453436e+00
3.31863046e-01 -7.95112476e-02 8.93340647e-01 -1.26368594e+00
8.61727476e-01 -7.53099918e-01 2.67512351e-01 6.58195853e-01
2.02226311e-01 3.25565070e-01 -7.89172500e-02 9.42627788e-01
-1.36377037e-01 5.30763507e-01 7.32827246e-01 -4.91023242e-01
-9.21328843e-01 5.54762959e-01 8.89464468e-02 2.47698128e-01
1.11254168e+00 -2.99308896e-01 -9.49063599e-02 -3.42539966e-01
-7.12219715e-01 5.73228717e-01 5.24117231e-01 4.84196037e-01
1.00012398e+00 -1.49395621e+00 -6.04324222e-01 1.89269453e-01
1.72309294e-01 1.97178200e-01 5.98532081e-01 1.13976979e+00
-4.84834552e-01 5.86281180e-01 -7.81538934e-02 -1.05723929e+00
-1.74471807e+00 6.44771397e-01 3.74133855e-01 -3.13438833e-01
-7.58650541e-01 9.76412177e-01 7.06291735e-01 -2.18569115e-01
2.03569517e-01 -1.95118845e-01 -1.56391904e-01 -4.57700565e-02
6.04793191e-01 6.40096903e-01 -3.04661512e-01 -9.82034922e-01
-6.06657863e-01 1.11864471e+00 5.06692603e-02 -1.59895927e-01
9.96544003e-01 -5.48053086e-01 3.31411026e-02 3.37140441e-01
1.01254165e+00 1.04487352e-01 -1.50307250e+00 -2.84490556e-01
-5.54851174e-01 -9.04708803e-01 -3.17406774e-01 -5.17707586e-01
-1.29946589e+00 9.24140036e-01 7.11751461e-01 -3.83524567e-01
1.18466997e+00 1.13699898e-01 1.07851589e+00 1.96765125e-01
7.86514997e-01 -1.12048519e+00 4.12727237e-01 1.29240662e-01
8.72907460e-01 -1.23386276e+00 3.46738636e-01 -7.91723490e-01
-7.84484863e-01 6.97889090e-01 1.19653559e+00 8.11610296e-02
2.81068146e-01 -1.43103558e-03 -1.08003192e-01 -2.71853536e-01
-5.49044669e-01 -3.77794862e-01 6.18863821e-01 6.60547316e-01
1.94783673e-01 -3.21425237e-02 -3.52298841e-02 4.39675242e-01
1.20018139e-01 -1.80887446e-01 9.83363762e-02 8.73249710e-01
-5.44016361e-01 -7.06554174e-01 -9.23640788e-01 1.61750704e-01
-3.75643611e-01 2.33955860e-01 -3.46668422e-01 9.87259805e-01
3.47127467e-01 8.92836213e-01 -2.72412419e-01 -5.79300523e-01
6.33535445e-01 -1.20667011e-01 8.79179418e-01 -4.33913887e-01
-4.33224559e-01 5.18214643e-01 5.69754951e-02 -1.03040290e+00
-7.11167276e-01 -8.87745440e-01 -8.97354007e-01 -1.17296413e-01
-4.46698695e-01 -1.69765934e-01 -3.21961120e-02 9.67921019e-01
2.86368430e-01 6.16731584e-01 3.51974279e-01 -1.25322115e+00
-4.42859411e-01 -7.89831579e-01 -4.39723253e-01 3.86963904e-01
1.83141768e-01 -1.25183237e+00 -5.85744232e-02 9.72770974e-02] | [7.174656867980957, -0.8458951711654663] |
4b3b98a3-b07b-4054-b2ba-b42b5fcad70d | enhancing-unsupervised-sentence-similarity | null | null | https://aclanthology.org/R19-1115 | https://aclanthology.org/R19-1115.pdf | Enhancing Unsupervised Sentence Similarity Methods with Deep Contextualised Word Representations | Calculating Semantic Textual Similarity (STS) plays a significant role in many applications such as question answering, document summarisation, information retrieval and information extraction. All modern state of the art STS methods rely on word embeddings one way or another. The recently introduced contextualised word embeddings have proved more effective than standard word embeddings in many natural language processing tasks. This paper evaluates the impact of several contextualised word embeddings on unsupervised STS methods and compares it with the existing supervised/unsupervised STS methods for different datasets in different languages and different domains | ['Tharindu Ranasinghe', 'Constantin Orasan', 'Ruslan Mitkov'] | 2019-09-01 | null | null | null | ranlp-2019-9 | ['contextualised-word-representations'] | ['natural-language-processing'] | [ 2.73197651e-01 -5.97313568e-02 -3.58713299e-01 -1.10157885e-01
-4.08981532e-01 -3.54114503e-01 1.22555685e+00 1.27857697e+00
-1.18084812e+00 5.15555084e-01 1.10696328e+00 -2.04382360e-01
-4.76685047e-01 -6.85984433e-01 2.75883198e-01 -3.07187825e-01
1.88411660e-02 6.40244424e-01 6.78091705e-01 -6.60164654e-01
1.04375815e+00 7.67402947e-02 -1.54520822e+00 1.99344265e-03
6.29530549e-01 6.47840440e-01 9.04986635e-02 7.01062739e-01
-1.04383790e+00 4.52110589e-01 -8.10873747e-01 -4.12446737e-01
-2.83317864e-01 -1.32206082e-01 -1.05775857e+00 -4.49329108e-01
-2.41207536e-02 5.91144145e-01 -3.18253666e-01 8.81832838e-01
7.00563252e-01 5.92164457e-01 8.64762723e-01 -9.37914431e-01
-9.30148304e-01 5.10302901e-01 -2.97831595e-01 8.22461784e-01
6.64846718e-01 -4.15838927e-01 1.29946578e+00 -9.17202175e-01
7.01227248e-01 1.34894550e+00 6.48033142e-01 3.38222496e-02
-8.66506577e-01 -1.38866678e-01 -2.29163960e-01 7.35567808e-01
-1.22474134e+00 -6.68386370e-02 5.95666647e-01 -2.10976705e-01
1.58839869e+00 1.64941952e-01 4.11026627e-01 9.86623108e-01
3.53446811e-01 5.29813945e-01 7.43328989e-01 -8.28993380e-01
1.44480959e-01 1.91996917e-01 4.93838608e-01 1.88999429e-01
3.81527811e-01 -5.38857341e-01 -5.88247478e-01 -3.26954424e-01
6.90122172e-02 -4.40331642e-03 -3.33011895e-02 -4.94309515e-02
-1.32805383e+00 1.43852687e+00 5.46777546e-02 1.15031278e+00
-1.60870403e-01 1.20719157e-01 1.12914729e+00 3.13675404e-01
8.54555547e-01 8.38235855e-01 -6.06686294e-01 -3.84917587e-01
-7.55760550e-01 2.38733649e-01 5.35579085e-01 5.93201399e-01
4.24480766e-01 -3.72538194e-02 -4.68138039e-01 1.27800643e+00
3.47189665e-01 2.72330612e-01 1.49286151e+00 -3.03765833e-01
3.37295473e-01 9.14294004e-01 -1.39332458e-01 -1.25540233e+00
-2.91738153e-01 -4.65859957e-02 -4.24667805e-01 -5.03046691e-01
-2.13608339e-01 1.81568503e-01 -8.92964125e-01 1.14675915e+00
2.79018849e-01 4.13647071e-02 3.76193494e-01 5.69406867e-01
1.35038650e+00 9.25266802e-01 2.91871220e-01 -1.42490283e-01
1.67263675e+00 -7.34363079e-01 -1.24410486e+00 -3.76474857e-01
8.77625942e-01 -1.27701652e+00 9.07647848e-01 1.49048373e-01
-5.43854117e-01 -2.56857216e-01 -1.11618268e+00 -4.84861463e-01
-1.14792156e+00 -3.64294142e-01 2.90031433e-01 7.38393605e-01
-8.36078882e-01 5.34790814e-01 -9.59503502e-02 -1.14687777e+00
3.40712339e-01 1.01604760e-01 -5.88339329e-01 -1.22316360e-01
-1.54486895e+00 1.55553353e+00 8.49482536e-01 -5.00601292e-01
3.26599292e-02 -4.20832604e-01 -1.30667925e+00 6.74649701e-02
3.18675309e-01 -3.67840260e-01 1.00018334e+00 -3.88541818e-01
-1.17104709e+00 1.00548089e+00 -2.81176358e-01 -7.22049415e-01
-1.53604656e-01 -4.84119684e-01 -7.03848600e-01 1.52046785e-01
2.17066079e-01 2.24487484e-01 5.74786007e-01 -2.77306616e-01
-1.42073616e-01 -2.37842739e-01 -4.87392157e-01 3.13283503e-01
-8.97983730e-01 4.52013165e-01 4.62299883e-02 -1.02789974e+00
-2.67413557e-01 -2.82751083e-01 -1.12619437e-01 -2.53769934e-01
4.43135723e-02 -1.12758875e+00 1.15798199e+00 -3.68142039e-01
1.39820969e+00 -1.94064391e+00 2.38688096e-01 -1.58213079e-01
9.22303572e-02 7.95216680e-01 -3.74632739e-02 1.29098427e+00
-9.60085727e-03 2.44669527e-01 -2.36744776e-01 -1.66345030e-01
1.54795408e-01 3.93868089e-01 -2.32633978e-01 3.09821218e-01
1.66086510e-01 8.91910493e-01 -1.14090788e+00 -8.22903275e-01
5.02536714e-01 3.04700315e-01 -1.36014402e-01 -2.50909617e-03
-7.22397640e-02 -3.90237659e-01 -6.04064941e-01 4.62384000e-02
4.29381765e-02 2.68903583e-01 1.82387039e-01 -2.04601325e-02
-1.55374184e-01 4.82425958e-01 -8.13074470e-01 1.90959752e+00
-6.49230957e-01 1.22160065e+00 -1.03083158e+00 -1.44989741e+00
1.12606406e+00 6.10489011e-01 3.95699352e-01 -8.38180423e-01
4.72739130e-01 1.18535243e-01 -2.37309083e-01 -8.21244717e-01
1.13513398e+00 -4.91557539e-01 -2.46289283e-01 5.54222763e-01
4.80812937e-01 -3.53461355e-01 2.82210588e-01 5.32139361e-01
1.10313141e+00 -3.73451173e-01 9.74946141e-01 -4.98531729e-01
6.18248761e-01 1.35184765e-01 1.36163965e-01 1.93291217e-01
-2.61740416e-01 6.74391985e-01 3.82291317e-01 -3.14905196e-01
-1.26872241e+00 -8.96886706e-01 -1.09157965e-01 1.09498417e+00
-7.62037709e-02 -9.21633780e-01 -2.47425437e-01 -4.80047613e-01
1.23865671e-01 1.04487622e+00 -8.68389606e-01 -4.86766189e-01
-2.47215986e-01 -4.35328156e-01 6.39399648e-01 6.00097537e-01
1.58177406e-01 -1.17936349e+00 -6.51860118e-01 4.36768800e-01
2.36676745e-02 -1.09146559e+00 -3.59555513e-01 1.31476507e-01
-9.16150272e-01 -8.83137286e-01 -7.45851934e-01 -8.95698607e-01
1.49214104e-01 3.92708182e-01 9.46227908e-01 -7.04815611e-02
-5.56864142e-01 4.33557540e-01 -1.08234608e+00 -7.31634438e-01
-7.11496845e-02 2.76698768e-01 3.29871923e-01 -3.21229309e-01
9.69909966e-01 -2.06432000e-01 -3.57512474e-01 -2.51922637e-01
-1.45476425e+00 -6.91060364e-01 1.43303752e-01 8.54859829e-01
-5.56215271e-02 -2.15684753e-02 8.34357858e-01 -9.39790428e-01
1.41621554e+00 -6.72567606e-01 1.49787292e-01 3.70970249e-01
-5.46486557e-01 5.47361434e-01 3.42328966e-01 -4.22580898e-01
-7.89874136e-01 -7.87311852e-01 -2.59474844e-01 1.23323739e-01
1.01482213e-01 8.75318468e-01 5.16257703e-01 4.17565405e-01
7.64714003e-01 1.98350057e-01 -2.00840935e-01 -4.94393766e-01
5.44026077e-01 9.63176489e-01 8.51900280e-02 -2.22912714e-01
5.99014521e-01 1.64174035e-01 -3.55377585e-01 -1.37092972e+00
-6.43899560e-01 -1.25216329e+00 -5.20141661e-01 -3.28596234e-02
1.15041780e+00 -2.13763043e-01 -1.87979601e-02 -4.12349626e-02
-1.22623086e+00 3.87809306e-01 -5.73261082e-01 5.03266752e-01
-1.94111597e-02 7.08521843e-01 -3.85408849e-02 -5.92325509e-01
-5.53124249e-01 -4.92544919e-01 9.59038675e-01 1.87088326e-01
-7.89296031e-01 -1.52895272e+00 7.20692515e-01 1.49507135e-01
7.55864143e-01 1.69503987e-01 1.17893434e+00 -1.32247341e+00
7.79087603e-01 -6.32167637e-01 -8.34549032e-03 2.69019336e-01
4.29600149e-01 -1.36554744e-02 -7.18076229e-01 7.55999982e-02
-1.67182654e-01 -3.30110401e-01 1.02705741e+00 1.67306229e-01
4.87399250e-01 -2.14931324e-01 -2.74923235e-01 -2.50667959e-01
1.53909838e+00 -1.11662727e-02 7.77890682e-01 5.16786158e-01
4.08179522e-01 7.37049580e-01 6.86649024e-01 4.61058676e-01
-3.98757756e-02 3.51544201e-01 -1.04956597e-01 4.41307843e-01
2.02790394e-01 -4.81516458e-02 1.63148567e-01 1.52334654e+00
2.99969703e-01 -5.04580736e-01 -1.05181134e+00 1.08278334e+00
-1.90579689e+00 -9.26994860e-01 -2.43442699e-01 2.03752708e+00
5.87832391e-01 1.67986482e-01 -2.24567711e-01 6.08221769e-01
6.76472664e-01 7.65949070e-01 1.83206916e-01 -1.40304971e+00
-1.20039478e-01 8.91274869e-01 3.23989928e-01 2.71886855e-01
-7.56207108e-01 1.07779229e+00 6.36241579e+00 1.07690072e+00
-7.45386124e-01 4.08004045e-01 -2.49897107e-01 2.49856263e-01
-4.37905580e-01 -9.84187145e-03 -1.68857172e-01 1.88018724e-01
1.09182024e+00 -8.38712394e-01 -4.41783935e-01 6.16663694e-01
9.62567516e-03 -5.63930690e-01 -7.30861187e-01 8.77294660e-01
6.16738319e-01 -1.42254198e+00 3.87942225e-01 -2.82012671e-01
6.24730349e-01 -3.69611531e-02 -1.23047166e-01 3.20342988e-01
5.15118018e-02 -1.13825369e+00 1.32764816e-01 2.30874047e-01
6.60034120e-01 -7.75098264e-01 1.38666975e+00 -7.69476034e-03
-1.19147813e+00 1.45208538e-01 -5.61655521e-01 -2.39681661e-01
2.71081120e-01 5.30247271e-01 -8.89916241e-01 6.97096407e-01
5.35308719e-01 8.93066764e-01 -5.74294984e-01 8.51956248e-01
-3.25360954e-01 5.08431792e-01 -8.31921324e-02 -6.55789793e-01
7.44850576e-01 -8.53892565e-02 6.70693755e-01 1.56093955e+00
1.37656584e-01 6.74496964e-02 -3.05334628e-01 -2.46221628e-02
-3.52698155e-02 6.91887856e-01 -9.38477457e-01 -6.37257874e-01
3.65990907e-01 9.57739234e-01 -9.35317874e-01 -6.33402169e-01
-3.88765305e-01 8.72880936e-01 3.18440609e-02 -5.64237870e-02
-4.36779201e-01 -1.03420615e+00 5.75643837e-01 -9.80995744e-02
3.27235699e-01 -4.68613356e-01 1.51406884e-01 -8.37496936e-01
-2.01340318e-01 -3.40889245e-01 7.74836659e-01 -6.90865755e-01
-1.40811837e+00 5.05832016e-01 3.49807382e-01 -1.18207204e+00
-2.14739755e-01 -5.78246355e-01 -8.56822193e-01 5.22117198e-01
-1.44486678e+00 -5.38830101e-01 3.35122012e-02 4.20434505e-01
1.00119543e+00 -2.92987227e-01 1.03332770e+00 1.80252045e-01
-2.68042356e-01 1.41174316e-01 3.98155510e-01 -3.99060734e-02
9.82846916e-01 -1.32346380e+00 4.72518086e-01 5.95974028e-01
5.86904109e-01 5.62041283e-01 9.49942112e-01 -3.63032520e-01
-1.14534080e+00 -6.92449808e-01 1.67257738e+00 -3.46890002e-01
9.96322751e-01 -9.50598195e-02 -1.03763425e+00 2.02213183e-01
8.66946042e-01 -9.17219222e-02 1.00742197e+00 1.46883419e-02
-3.76703113e-01 1.11018136e-01 -1.00762475e+00 6.68303430e-01
6.13181114e-01 -6.01087689e-01 -1.61182189e+00 4.20456886e-01
9.20864880e-01 3.77338201e-01 -7.75475204e-01 1.40554368e-01
2.98077255e-01 -6.08504474e-01 8.40852439e-01 -8.52443278e-01
5.00035107e-01 -1.04920872e-01 -1.21932931e-01 -1.59073365e+00
-7.12485686e-02 -3.65695089e-01 1.05330691e-01 1.25483060e+00
7.00412840e-02 -9.78607416e-01 2.64608264e-01 -1.09891690e-01
-1.08224355e-01 -5.57023644e-01 -1.14761662e+00 -8.55855882e-01
1.90542981e-01 -4.35192227e-01 3.25999916e-01 1.16215920e+00
4.15101975e-01 7.12006032e-01 1.24413423e-01 -5.61574280e-01
1.15216002e-01 -3.83267939e-01 3.89519840e-01 -1.36554778e+00
3.72674137e-01 -6.54002190e-01 -1.28553033e+00 -3.50677609e-01
1.44018397e-01 -9.58590746e-01 -1.23478554e-01 -2.11846137e+00
3.11408639e-02 7.50436559e-02 -5.30792952e-01 6.85174763e-02
-3.20772946e-01 1.92808807e-01 -1.95128411e-01 -3.49702597e-01
-6.84846759e-01 9.36895490e-01 5.93683362e-01 -2.95832068e-01
2.85169691e-01 -8.00486803e-01 -4.72666711e-01 5.77761889e-01
8.72312844e-01 -7.99732327e-01 -5.53444564e-01 -5.38101733e-01
4.93243873e-01 -6.24382019e-01 -1.93718046e-01 -7.70401359e-01
2.59315848e-01 -1.23322114e-01 -1.63350746e-01 -3.32091838e-01
1.10847734e-01 -6.13678873e-01 -5.13061583e-01 3.88185591e-01
-4.95709151e-01 4.24730271e-01 1.87639996e-01 5.44218957e-01
-6.67370617e-01 -7.81030059e-01 3.98849696e-01 2.65388284e-02
-9.97327149e-01 -9.19663459e-02 -7.34554470e-01 6.34446442e-01
9.68609035e-01 -5.69504201e-01 -1.18265435e-01 -2.12019458e-01
-8.07301477e-02 2.24729896e-01 1.19389854e-01 9.99639869e-01
8.93632412e-01 -1.41120708e+00 -7.57686198e-01 -3.42822254e-01
7.54850268e-01 -2.50309706e-01 -1.83951527e-01 6.88211918e-01
-7.61572897e-01 6.46065474e-01 -6.31910749e-03 -2.01793447e-01
-1.50604451e+00 4.90300387e-01 -4.26043630e-01 -3.71591955e-01
-5.94290257e-01 6.47757113e-01 -4.52924818e-01 -2.29258478e-01
-7.19784945e-02 -2.36113206e-01 -8.96903098e-01 5.27929366e-01
4.59294260e-01 4.50292975e-01 1.67650610e-01 -8.11715424e-01
-5.42361856e-01 8.53542805e-01 -2.05216601e-01 -3.65912795e-01
1.39995265e+00 1.97576433e-01 -1.56730145e-01 9.30593312e-01
1.57066917e+00 -1.82601497e-01 2.39027828e-01 -5.52016139e-01
7.63354719e-01 -6.28090262e-01 2.37177297e-01 -1.66750550e-01
-3.16036940e-01 1.12099040e+00 3.22342634e-01 5.50138772e-01
6.69927895e-01 -1.57633517e-02 1.13420117e+00 6.17143333e-01
1.06542148e-01 -1.67782199e+00 2.61155367e-01 6.66010559e-01
8.55303824e-01 -1.19280803e+00 5.02077222e-01 9.79105011e-02
-8.69140744e-01 1.33698797e+00 -1.44893736e-01 -3.65262568e-01
6.20727181e-01 -2.00758249e-01 -8.64543207e-03 -4.74089772e-01
-7.16641963e-01 -4.29345936e-01 5.36950886e-01 6.26673877e-01
7.64514506e-01 -1.95602238e-01 -1.04571509e+00 5.68298548e-02
-1.15052968e-01 -3.49014401e-01 4.19628114e-01 1.29157829e+00
-8.20454240e-01 -1.38973999e+00 -1.75783575e-01 6.61607146e-01
-6.27056062e-01 -3.07350963e-01 -7.77122736e-01 6.48204386e-01
-1.76596373e-01 1.05998588e+00 7.66028985e-02 -9.49506834e-02
2.16468796e-01 3.70994568e-01 3.36934328e-01 -1.09580410e+00
-8.36792707e-01 -5.04672885e-01 3.60763848e-01 -1.10800497e-01
-7.74404824e-01 -4.99307007e-01 -1.17223597e+00 -1.33680567e-01
-6.01232708e-01 6.27882004e-01 7.55818367e-01 1.38754272e+00
9.78147388e-02 5.28418303e-01 4.37974393e-01 -4.51853991e-01
-1.70286506e-01 -1.38477290e+00 -3.81726921e-01 5.17104387e-01
2.65411921e-02 -7.89310396e-01 -4.06799167e-01 -1.60465017e-01] | [10.499785423278809, 8.675131797790527] |
9199d65c-0a27-4767-9766-da566bd225b5 | biobart-pretraining-and-evaluation-of-a | 2204.03905 | null | https://arxiv.org/abs/2204.03905v2 | https://arxiv.org/pdf/2204.03905v2.pdf | BioBART: Pretraining and Evaluation of A Biomedical Generative Language Model | Pretrained language models have served as important backbones for natural language processing. Recently, in-domain pretraining has been shown to benefit various domain-specific downstream tasks. In the biomedical domain, natural language generation (NLG) tasks are of critical importance, while understudied. Approaching natural language understanding (NLU) tasks as NLG achieves satisfying performance in the general domain through constrained language generation or language prompting. We emphasize the lack of in-domain generative language models and the unsystematic generative downstream benchmarks in the biomedical domain, hindering the development of the research community. In this work, we introduce the generative language model BioBART that adapts BART to the biomedical domain. We collate various biomedical language generation tasks including dialogue, summarization, entity linking, and named entity recognition. BioBART pretrained on PubMed abstracts has enhanced performance compared to BART and set strong baselines on several tasks. Furthermore, we conduct ablation studies on the pretraining tasks for BioBART and find that sentence permutation has negative effects on downstream tasks. | ['Sheng Yu', 'Yutao Xie', 'Jiaxing Zhang', 'Ruyi Gan', 'Zheng Yuan', 'Hongyi Yuan'] | 2022-04-08 | null | https://aclanthology.org/2022.bionlp-1.9 | https://aclanthology.org/2022.bionlp-1.9.pdf | bionlp-acl-2022-5 | ['nested-named-entity-recognition'] | ['natural-language-processing'] | [ 6.52407229e-01 7.47035801e-01 -8.68935660e-02 -5.02705693e-01
-1.24612761e+00 -4.96807098e-01 6.78798079e-01 2.91138202e-01
-4.59250748e-01 1.32223058e+00 9.23721313e-01 -5.73461235e-01
8.29135776e-02 -6.94005668e-01 -7.47577548e-01 -4.13677394e-01
3.96394283e-02 8.19367170e-01 -3.10729563e-01 -5.11180758e-01
-4.48901355e-02 -1.78733654e-02 -5.99516988e-01 7.30638742e-01
1.08379924e+00 2.78980225e-01 7.62927085e-02 6.57482445e-01
-1.82726368e-01 6.42053366e-01 -9.63192105e-01 -4.67304975e-01
-2.94834405e-01 -9.07010794e-01 -1.08863521e+00 -4.05689359e-01
2.60803223e-01 -4.93952259e-03 -1.47970036e-01 6.13328040e-01
1.01136422e+00 -8.40596780e-02 8.87536466e-01 -1.02198732e+00
-1.01974034e+00 1.13135862e+00 -3.33858430e-01 2.34801576e-01
3.70570034e-01 2.21097425e-01 1.02804625e+00 -7.38392174e-01
1.14237642e+00 1.47552323e+00 6.41854763e-01 1.06864834e+00
-1.51954412e+00 -5.85546494e-01 1.10814631e-01 -3.22287410e-01
-1.01857984e+00 -6.17284894e-01 3.09747010e-01 -2.57825226e-01
1.39946520e+00 -6.97088093e-02 1.37053117e-01 1.70139575e+00
6.56480193e-01 7.57420480e-01 6.94672763e-01 -1.99874148e-01
1.28239691e-01 -2.43652403e-01 2.16279924e-01 4.81144130e-01
5.83238423e-01 -1.09227687e-01 -6.06878996e-01 -1.71210349e-01
2.41385952e-01 -5.00627816e-01 -4.33202058e-01 4.25028920e-01
-1.46560729e+00 1.00550997e+00 3.36788505e-01 3.16982150e-01
-4.03422385e-01 2.14250997e-01 5.41134298e-01 2.41582811e-01
7.91802406e-01 1.06427956e+00 -4.23165858e-01 -5.94885461e-02
-8.81150007e-01 2.47431993e-01 9.84770238e-01 1.20299292e+00
1.44807383e-01 3.58747095e-02 -7.38010943e-01 7.29086757e-01
-1.92735307e-02 3.54174674e-01 6.24421656e-01 -3.80308002e-01
6.94903493e-01 4.22069013e-01 -3.64298820e-01 -5.98373413e-01
-4.91568208e-01 -5.56525767e-01 -1.20570934e+00 -6.07448399e-01
3.05455387e-01 -8.08466733e-01 -1.04158652e+00 2.10394812e+00
-2.63991766e-02 -6.05805516e-02 6.60488069e-01 4.81727749e-01
1.35354888e+00 8.43711555e-01 6.09047413e-01 -9.73200575e-02
1.43298650e+00 -5.98419309e-01 -7.71721363e-01 -6.31341815e-01
8.19483221e-01 -7.55834818e-01 6.70751750e-01 1.46170646e-01
-1.11628842e+00 -4.09204215e-01 -6.76588535e-01 -5.84646821e-01
-3.49103481e-01 -4.97253239e-02 6.21575892e-01 2.13762909e-01
-1.28888059e+00 3.91755432e-01 -8.16468298e-01 -5.08077264e-01
5.64633608e-01 1.61390558e-01 -3.41660678e-01 -2.15364814e-01
-1.60416365e+00 9.01888847e-01 7.38355815e-01 -3.34181748e-02
-8.60643744e-01 -1.08001995e+00 -1.08625197e+00 2.03870967e-01
5.77094890e-02 -1.44313121e+00 1.21983194e+00 -1.80491120e-01
-1.06274986e+00 8.97210836e-01 -1.17207542e-01 -9.01189029e-01
4.45713073e-01 -2.62364179e-01 -2.02692375e-01 -2.25688517e-02
4.27443027e-01 1.21588147e+00 2.12567046e-01 -6.67925954e-01
-1.21851251e-01 -1.36737347e-01 -3.77170563e-01 2.13393986e-01
-4.26087677e-02 -3.50233316e-01 -6.95172250e-02 -8.42877746e-01
-3.72042567e-01 -7.10263968e-01 -4.27354902e-01 -6.99945450e-01
-8.99685919e-01 -4.87848401e-01 3.24884504e-01 -7.58142352e-01
1.10639238e+00 -1.86902308e+00 -6.56575635e-02 -3.38824123e-01
1.16671324e-01 1.08627409e-01 -4.32988077e-01 7.34295845e-01
-4.02479023e-01 4.00613010e-01 -3.56046170e-01 -2.24931091e-01
-2.72928886e-02 4.10161689e-02 -4.99592543e-01 -3.12727839e-02
7.73785830e-01 1.41298616e+00 -1.12969828e+00 -4.66435939e-01
-4.90696073e-01 4.92418885e-01 -7.72203028e-01 1.93299726e-01
-5.98819137e-01 4.98227060e-01 -6.03660226e-01 3.54404122e-01
1.90000519e-01 -5.32753825e-01 2.01053798e-01 -1.64903581e-01
3.09563756e-01 7.77097464e-01 -2.66336143e-01 2.01140714e+00
-2.88462728e-01 3.94688040e-01 -2.78200120e-01 -9.45994496e-01
8.06094289e-01 5.70715189e-01 1.97082326e-01 -4.02643591e-01
-9.07661468e-02 2.93240324e-02 3.40992451e-01 -4.64135826e-01
4.43223208e-01 -4.79808301e-01 -5.52033782e-01 6.54091120e-01
4.21226174e-01 -2.89791763e-01 3.57245564e-01 7.27141321e-01
1.31541908e+00 1.44169152e-01 5.98176599e-01 -3.95565778e-01
9.32552759e-03 4.03829098e-01 3.05031538e-01 8.28736007e-01
2.40513742e-01 5.57476759e-01 6.65665269e-01 1.19650975e-01
-7.00690269e-01 -9.76131499e-01 -9.02898610e-02 1.06564248e+00
-4.72421885e-01 -5.03398299e-01 -7.92175651e-01 -6.32311404e-01
-3.90002243e-02 1.09902084e+00 -5.10025799e-01 -6.68324709e-01
-5.82353354e-01 -1.34780288e+00 1.00388706e+00 5.54077089e-01
3.17096233e-01 -1.24297249e+00 -1.11281378e-02 4.51713353e-01
-3.99206281e-01 -1.24219978e+00 -7.24862933e-01 1.32109031e-01
-8.57924819e-01 -7.52976954e-01 -9.41516221e-01 -8.68411481e-01
7.77536631e-01 -3.10397327e-01 1.57435107e+00 -5.68765938e-01
-3.96264255e-01 2.14957088e-01 -1.68729573e-01 -6.51484609e-01
-1.08688700e+00 6.49485886e-01 -3.00036550e-01 -5.91633976e-01
3.41723979e-01 -4.38689053e-01 -6.54854178e-01 -1.62490696e-01
-9.34938967e-01 3.89758378e-01 9.30096567e-01 1.02614856e+00
3.53713304e-01 -5.96720278e-01 1.36857474e+00 -1.50100780e+00
1.31998420e+00 -7.14229167e-01 3.93030941e-02 2.50558615e-01
-5.95500886e-01 4.61226344e-01 5.29362619e-01 -2.09851295e-01
-1.29438567e+00 -3.51654112e-01 -3.59695554e-01 3.64984751e-01
-3.15331787e-01 8.98865163e-01 -2.03148216e-01 7.34561682e-01
1.04027390e+00 2.63243645e-01 -1.10016996e-02 -3.24950069e-01
5.86055160e-01 3.97175014e-01 5.80608606e-01 -7.19301879e-01
3.13196778e-01 -5.29659986e-02 -1.99759118e-02 -7.75964618e-01
-9.25487399e-01 -1.08378083e-01 -1.80459678e-01 5.66403985e-01
1.15012312e+00 -1.08137906e+00 -3.11675519e-01 -2.10345369e-02
-1.46526825e+00 -4.30268914e-01 -3.54895234e-01 1.98648930e-01
-3.57594937e-01 2.16658294e-01 -8.38451922e-01 -2.86562592e-01
-1.02300227e+00 -8.66498411e-01 1.35299885e+00 6.43302314e-03
-8.49186122e-01 -1.18306541e+00 2.04809174e-01 2.87075400e-01
2.85764098e-01 4.67869371e-01 1.27136254e+00 -1.12604463e+00
-2.38273278e-01 -3.49197611e-02 -3.69492397e-02 5.63380420e-02
2.09040388e-01 -5.23592770e-01 -8.07676375e-01 -1.14931881e-01
-3.39582920e-01 -4.54642594e-01 1.17883182e+00 4.52316493e-01
9.32282329e-01 -4.23184842e-01 -6.66066110e-01 3.96690667e-01
8.58472586e-01 2.05280736e-01 4.64265019e-01 -2.48214528e-01
4.80843157e-01 6.30366027e-01 2.43266180e-01 6.21142685e-02
3.75854135e-01 1.04344361e-01 -3.60290736e-01 -4.22739953e-01
-3.06738019e-01 -4.80696499e-01 4.08115685e-01 5.72049201e-01
3.69557738e-01 -8.37073326e-01 -1.10589361e+00 6.44955814e-01
-1.70626521e+00 -8.68769944e-01 -1.20573878e-01 1.62124395e+00
1.57397544e+00 6.27138764e-02 -1.85418293e-01 -6.53821290e-01
6.09429181e-01 -1.13245584e-01 -5.70718229e-01 -3.84604067e-01
-2.41758466e-01 4.92794961e-01 9.01708566e-03 4.23926651e-01
-7.35197544e-01 1.13500154e+00 6.28831768e+00 5.74896812e-01
-9.39317644e-01 1.79338586e-02 1.23470080e+00 9.89550799e-02
-3.95448834e-01 -2.01704502e-01 -1.15777504e+00 3.75207692e-01
1.19236636e+00 -6.83031142e-01 -2.72517055e-01 5.96727312e-01
3.59798610e-01 1.00671694e-01 -1.70597208e+00 5.55860758e-01
1.04788085e-02 -1.49216795e+00 5.28980672e-01 3.27574201e-02
8.08565021e-01 1.81651160e-01 -9.88659933e-02 5.86291492e-01
7.28283703e-01 -1.33256316e+00 5.10218628e-02 4.82024789e-01
9.59176838e-01 -4.51595545e-01 8.18942189e-01 2.58175194e-01
-4.02065128e-01 5.27457833e-01 -2.61629671e-01 3.69453788e-01
5.32293677e-01 9.22685027e-01 -1.73973930e+00 5.17130196e-01
8.45466182e-02 6.84308648e-01 -5.15992224e-01 5.96698999e-01
-4.24785316e-01 8.57666016e-01 -2.68918909e-02 -7.32182711e-02
2.46339858e-01 1.13773182e-01 4.59196955e-01 1.77127874e+00
3.53032291e-01 2.20869616e-01 1.31195217e-01 1.15879977e+00
-5.97747564e-01 4.81593981e-02 -8.11598420e-01 -8.31987977e-01
1.11184843e-01 1.09319687e+00 -4.43202734e-01 -5.55304825e-01
1.28894811e-02 9.03129876e-01 1.11942865e-01 4.37953353e-01
-5.74994504e-01 -4.04857308e-01 5.62487364e-01 2.23997235e-02
-1.84040964e-01 -1.15853886e-03 -3.27583969e-01 -1.14878869e+00
-5.29507697e-01 -9.68622029e-01 6.86695933e-01 -7.25037456e-01
-1.55816066e+00 7.71838605e-01 1.06169239e-01 -6.65163100e-01
-9.09225464e-01 -3.83742303e-01 -6.35382950e-01 1.11434019e+00
-1.26373827e+00 -9.94908690e-01 8.69188905e-02 2.57567763e-01
6.74909949e-01 -1.42858818e-01 9.90726054e-01 8.55005980e-02
-4.20672655e-01 6.17313981e-01 -2.64519721e-01 2.83670545e-01
1.10340834e+00 -1.31383598e+00 9.67323244e-01 7.25541353e-01
-2.37183407e-01 1.13483691e+00 6.12544954e-01 -9.87246692e-01
-1.02084434e+00 -1.56976974e+00 1.43491089e+00 -4.95852262e-01
5.26940823e-01 -6.17715299e-01 -8.67769063e-01 8.98128271e-01
6.26451850e-01 -5.87512255e-01 1.03577054e+00 1.60888031e-01
6.12455532e-02 2.45452434e-01 -8.45206738e-01 6.48423016e-01
1.22752750e+00 -2.79560924e-01 -7.85507500e-01 7.22396970e-01
1.08044839e+00 -4.33127284e-01 -9.32072282e-01 3.29358429e-01
8.02522302e-02 -1.60927072e-01 9.33378637e-01 -1.19217908e+00
1.05257869e+00 2.72034287e-01 5.12017548e-01 -1.65441263e+00
-2.34114796e-01 -8.14869881e-01 1.82611406e-01 1.36977863e+00
1.03559458e+00 -6.15044713e-01 5.72753549e-01 5.06878674e-01
-4.86252725e-01 -5.32791972e-01 -4.74827051e-01 -4.27669287e-01
5.08325338e-01 -5.81168048e-02 3.01030606e-01 9.07823861e-01
2.54158765e-01 1.36721087e+00 -2.14865543e-02 -3.06914866e-01
2.26510450e-01 1.84718445e-01 6.35179043e-01 -9.70811486e-01
-4.25866842e-01 -3.58968347e-01 6.79207966e-02 -9.47192848e-01
3.38758558e-01 -1.43851411e+00 2.82600611e-01 -2.04729152e+00
2.93326199e-01 -4.38979752e-02 1.05351262e-01 7.83875644e-01
-5.92452943e-01 -8.92814025e-02 -2.76448578e-01 -2.69303322e-01
-3.85561854e-01 5.32025874e-01 1.32806230e+00 -3.17516804e-01
-7.30708912e-02 -1.68242082e-01 -1.40872490e+00 1.84117258e-01
8.52519810e-01 -3.98313284e-01 -7.25086689e-01 -5.83877504e-01
3.54516894e-01 1.98130310e-01 5.90137579e-02 -5.40377021e-01
4.41171862e-02 1.92363411e-02 4.28959727e-01 -3.59197915e-01
-1.25448806e-02 7.08257705e-02 -8.11986476e-02 5.98159552e-01
-8.71894956e-01 2.76351750e-01 5.07721901e-01 5.19260168e-01
-1.34307832e-01 4.04344946e-02 5.20578742e-01 -3.88848186e-01
-3.74110878e-01 2.53011912e-01 -4.48412895e-01 7.88318872e-01
5.31727135e-01 1.65115908e-01 -6.05708182e-01 -5.56049466e-01
-9.45541859e-01 3.73212725e-01 -4.13094647e-02 4.26194012e-01
6.14864409e-01 -8.17414045e-01 -1.29571903e+00 -3.55903544e-02
1.26558498e-01 3.86806816e-01 1.73071295e-01 7.99561262e-01
-2.82970130e-01 9.21295226e-01 1.08100340e-01 -4.26262468e-01
-1.03154743e+00 3.45120162e-01 3.56003605e-02 -7.01699018e-01
-4.19770539e-01 9.97723222e-01 8.00002337e-01 -4.30273563e-01
-1.26876235e-01 -6.13514602e-01 -5.67664430e-02 4.42065485e-02
3.91301244e-01 -3.82358232e-03 2.23342255e-01 -6.14216365e-02
-2.30601415e-01 -2.88011342e-01 -4.42649066e-01 -7.20951781e-02
1.47950578e+00 2.18019336e-01 -1.06552519e-01 7.62553737e-02
1.20769525e+00 -3.16072136e-01 -4.88125592e-01 -1.66151347e-03
3.11139107e-01 5.65625966e-01 -1.89674750e-01 -1.25955260e+00
-4.86861616e-01 8.63042414e-01 -6.82829097e-02 -2.58103341e-01
7.98876584e-01 2.31148556e-01 1.02768624e+00 5.32393277e-01
5.37474640e-02 -6.77063286e-01 -3.49562764e-02 6.44170880e-01
1.06560326e+00 -1.13335824e+00 -1.35751233e-01 -3.74458164e-01
-6.63279533e-01 8.38158965e-01 6.60291314e-01 1.54965654e-01
3.57784867e-01 4.90787029e-01 -1.52051121e-01 -3.42797250e-01
-1.12346733e+00 1.14935860e-01 4.56065774e-01 5.14087260e-01
1.16974807e+00 -9.53916982e-02 -6.79798067e-01 7.47633338e-01
-5.67150116e-01 1.09210148e-01 3.58896762e-01 7.18768597e-01
-1.01364128e-01 -1.17700946e+00 -4.70305085e-02 5.17988145e-01
-6.19710326e-01 -7.82325089e-01 -7.57072210e-01 7.63460815e-01
-1.78307235e-01 8.68417323e-01 -4.71772887e-02 1.90138191e-01
1.12628691e-01 3.87745500e-01 3.97096962e-01 -1.24835348e+00
-6.22276545e-01 1.84408739e-01 5.69067717e-01 -1.57039121e-01
-9.80406106e-02 -4.35670882e-01 -1.30112469e+00 -2.51984168e-02
1.03793465e-01 5.68132162e-01 2.06733033e-01 7.80643225e-01
9.90328610e-01 9.61428761e-01 -2.29682669e-01 -2.36875683e-01
-5.12097657e-01 -1.25893700e+00 -3.49801383e-03 3.56230199e-01
1.74507067e-01 1.62671823e-02 1.34904236e-01 4.45157707e-01] | [8.677449226379395, 8.752117156982422] |
42bd5446-76e4-4404-ab6c-5f10621ad08a | vidlankd-improving-language-understanding-via | 2107.02681 | null | https://arxiv.org/abs/2107.02681v2 | https://arxiv.org/pdf/2107.02681v2.pdf | VidLanKD: Improving Language Understanding via Video-Distilled Knowledge Transfer | Since visual perception can give rich information beyond text descriptions for world understanding, there has been increasing interest in leveraging visual grounding for language learning. Recently, vokenization (Tan and Bansal, 2020) has attracted attention by using the predictions of a text-to-image retrieval model as labels for language model supervision. Despite its success, the method suffers from approximation error of using finite image labels and the lack of vocabulary diversity of a small image-text dataset. To overcome these limitations, we present VidLanKD, a video-language knowledge distillation method for improving language understanding. We train a multi-modal teacher model on a video-text dataset, and then transfer its knowledge to a student language model with a text dataset. To avoid approximation error, we propose to use different knowledge distillation objectives. In addition, the use of a large-scale video-text dataset helps learn diverse and richer vocabularies. In our experiments, VidLanKD achieves consistent improvements over text-only language models and vokenization models, on several downstream language understanding tasks including GLUE, SQuAD, and SWAG. We also demonstrate the improved world knowledge, physical reasoning, and temporal reasoning capabilities of our model by evaluating on the GLUE-diagnostics, PIQA, and TRACIE datasets. Lastly, we present comprehensive ablation studies as well as visualizations of the learned text-to-video grounding results of our teacher and student language models. Our code and models are available at: https://github.com/zinengtang/VidLanKD | ['Mohit Bansal', 'Hao Tan', 'Jaemin Cho', 'Zineng Tang'] | 2021-07-06 | null | http://proceedings.neurips.cc/paper/2021/hash/ccdf3864e2fa9089f9eca4fc7a48ea0a-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/ccdf3864e2fa9089f9eca4fc7a48ea0a-Paper.pdf | neurips-2021-12 | ['video-grounding'] | ['computer-vision'] | [-2.02129722e-01 1.87045261e-01 -4.56625998e-01 -4.22630310e-01
-7.36016631e-01 -6.86871827e-01 6.61590695e-01 -2.77680196e-02
-4.36739057e-01 4.22875106e-01 2.86449224e-01 -5.20874381e-01
5.55302985e-02 -5.85208058e-01 -9.89895463e-01 -2.84572542e-01
3.51191223e-01 5.35747826e-01 2.10849330e-01 -1.80581123e-01
-1.13678001e-01 1.50595710e-01 -1.46725488e+00 6.04413748e-01
9.26411152e-01 6.70049965e-01 4.26791906e-01 7.30290055e-01
-1.92970023e-01 1.36643100e+00 -3.13670307e-01 -5.05038381e-01
-4.75080982e-02 -2.77938992e-01 -9.60235000e-01 -1.41224399e-01
1.05044198e+00 -7.44359612e-01 -6.19610131e-01 7.81514347e-01
2.70651728e-01 3.58688653e-01 5.18820167e-01 -1.53840220e+00
-1.33993554e+00 7.05628693e-01 -3.76333296e-01 6.38255924e-02
4.98830616e-01 3.37261111e-01 1.07608676e+00 -9.30667281e-01
7.32632339e-01 1.41329062e+00 3.04898500e-01 7.80176222e-01
-1.24333358e+00 -9.59951401e-01 5.00722885e-01 6.34296179e-01
-1.48285651e+00 -2.16474771e-01 6.47126794e-01 -5.69948375e-01
1.11932790e+00 -1.32425189e-01 9.19985950e-01 1.38008809e+00
-2.70334661e-01 1.46659195e+00 1.07826650e+00 -4.89181459e-01
-1.45739108e-01 2.92721480e-01 1.94890022e-01 1.28571594e+00
-4.66492288e-02 1.62190810e-01 -1.19933534e+00 4.03227329e-01
7.28772223e-01 -1.30692393e-01 -3.50516409e-01 -6.25016570e-01
-1.13609397e+00 6.67794466e-01 5.37063897e-01 -4.79114316e-02
2.14213893e-01 5.55934429e-01 1.52663544e-01 3.35089356e-01
4.00862783e-01 2.57978320e-01 -3.22826922e-01 -2.25933105e-01
-7.09897399e-01 4.78767753e-02 5.93940914e-01 1.06542337e+00
8.05766702e-01 1.60526678e-01 -3.28147188e-02 6.34905040e-01
5.17536283e-01 8.36889446e-01 4.39700842e-01 -1.02248919e+00
5.46551228e-01 5.47040284e-01 -2.13976100e-01 -6.53463304e-01
-2.67850697e-01 -8.84019136e-02 -3.48181635e-01 1.60692260e-01
3.96049082e-01 1.67147934e-01 -1.35410237e+00 1.94711852e+00
3.08059156e-01 5.68286300e-01 3.46621424e-01 8.84622574e-01
1.20541298e+00 7.63815761e-01 3.89220864e-01 2.20011026e-01
1.25269818e+00 -1.34823823e+00 -5.29559255e-01 -2.77075440e-01
1.00229311e+00 -4.18882906e-01 1.64826512e+00 5.99270225e-01
-9.76802826e-01 -5.75416982e-01 -7.70751059e-01 -6.50525570e-01
-5.55524230e-01 -6.57558721e-03 6.62135720e-01 2.70895660e-01
-1.23412764e+00 1.39531847e-02 -1.01356828e+00 -4.79550987e-01
4.49849606e-01 1.22455373e-01 -4.36029106e-01 -2.94052243e-01
-1.08492148e+00 9.10479724e-01 4.84025300e-01 -3.31624925e-01
-1.35076547e+00 -1.06219125e+00 -1.20468795e+00 9.75325331e-02
5.46654046e-01 -8.10936153e-01 1.47475743e+00 -1.00155389e+00
-1.39793646e+00 9.91378307e-01 -5.83101884e-02 -3.80807817e-01
3.60286146e-01 -4.14970785e-01 -1.66214064e-01 4.56110388e-01
-8.46318752e-02 1.30231035e+00 6.88115597e-01 -1.17720771e+00
-3.96243423e-01 -1.16980955e-01 4.70541477e-01 4.35444534e-01
-4.55338299e-01 -3.34726125e-01 -7.13388801e-01 -6.20988488e-01
-9.64310020e-02 -8.33296657e-01 3.08245391e-01 2.32233673e-01
-1.95528671e-01 -4.34147447e-01 8.04721236e-01 -7.06971467e-01
9.96641278e-01 -2.16460633e+00 2.20737427e-01 -8.04438889e-02
3.57993037e-01 1.57369122e-01 -5.85738957e-01 2.42989093e-01
-1.69133320e-01 3.24093759e-01 3.19964886e-01 -3.43901157e-01
-8.44248943e-03 3.89163822e-01 -5.48733711e-01 1.68034323e-02
2.17095651e-02 1.21916080e+00 -1.01554418e+00 -6.22376323e-01
4.04929638e-01 6.13747954e-01 -8.15823197e-01 3.35135162e-01
-7.43060112e-01 3.81247163e-01 -3.73994440e-01 6.08725846e-01
2.36065358e-01 -6.07035458e-01 2.03739163e-02 -3.79630625e-01
2.65948206e-01 2.75515646e-01 -7.72471905e-01 2.03934240e+00
-6.56126976e-01 8.12351048e-01 -3.01899523e-01 -8.77328515e-01
3.80145013e-01 3.20877343e-01 1.64020449e-01 -1.07570136e+00
-1.56940311e-01 -2.24315733e-01 -2.40973562e-01 -7.53301740e-01
3.38453650e-01 -1.58922106e-01 2.00851299e-02 5.99512994e-01
4.25634325e-01 -5.82677841e-01 2.46532083e-01 8.55941474e-01
7.43760765e-01 4.80475485e-01 -3.86751480e-02 1.75449416e-01
1.76562950e-01 1.92310214e-01 9.75609049e-02 6.04457915e-01
-1.30988628e-01 3.12338501e-01 3.73830855e-01 -3.57256174e-01
-7.87557542e-01 -1.38471472e+00 2.37205774e-01 1.34909594e+00
3.15764129e-01 -7.14910328e-01 -4.75271612e-01 -7.89638817e-01
1.29754543e-01 1.09934497e+00 -4.57855999e-01 -2.90082395e-01
-3.76825720e-01 9.83051807e-02 7.31807590e-01 8.19988430e-01
4.98584807e-01 -1.01900423e+00 -2.64572769e-01 -2.57399380e-01
-2.79990613e-01 -1.51855695e+00 -4.31496382e-01 -1.11640215e-01
-5.64568162e-01 -1.08816254e+00 -3.48548174e-01 -8.35480571e-01
6.63854599e-01 3.77432376e-01 1.29819107e+00 3.03895295e-01
-2.66266465e-01 1.26092780e+00 -2.97233045e-01 -4.43227381e-01
-4.16271538e-01 -1.89984411e-01 1.64203987e-01 -4.76461411e-01
3.67229313e-01 -3.79023463e-01 -3.53319049e-01 7.81330913e-02
-8.35453391e-01 9.08186615e-01 3.19492221e-01 8.22260082e-01
5.17483234e-01 -3.25354725e-01 2.21034914e-01 -5.24897277e-01
3.21214020e-01 -1.49246499e-01 -8.38523090e-01 6.12389565e-01
-5.78599870e-01 2.51619279e-01 3.56356531e-01 -6.75249219e-01
-1.29395807e+00 -9.09993574e-02 1.95189834e-01 -8.96110237e-01
-2.39012204e-02 6.62667453e-01 6.90673515e-02 -5.32499142e-02
4.55166817e-01 2.85422295e-01 -1.41940981e-01 -3.82933706e-01
9.21340823e-01 3.53439867e-01 5.93424797e-01 -1.06437850e+00
8.39014411e-01 3.94077331e-01 -4.08064276e-01 -7.67231107e-01
-1.21832216e+00 -3.78697246e-01 -5.25850415e-01 -2.87829340e-01
1.16712928e+00 -1.46789193e+00 -9.84849155e-01 3.40535581e-01
-1.05723250e+00 -9.77224827e-01 -2.36568287e-01 6.38144970e-01
-6.28704727e-01 2.60214567e-01 -5.74727654e-01 -5.98384559e-01
-1.80417355e-02 -1.08458102e+00 1.00033629e+00 2.81304896e-01
-3.11057512e-02 -1.15995288e+00 -5.58959469e-02 1.02198863e+00
1.72609668e-02 -3.25559855e-01 1.15056825e+00 -5.52715123e-01
-1.10559142e+00 4.74590242e-01 -4.01846498e-01 3.23035538e-01
-1.97509363e-01 -1.01177275e-01 -9.98567104e-01 -2.74793953e-01
-4.67634380e-01 -1.29355109e+00 9.71534908e-01 -2.30654385e-02
1.31344104e+00 -2.23948568e-01 -2.18718812e-01 8.23715866e-01
1.25357151e+00 6.07833490e-02 1.61889449e-01 2.27485225e-01
1.04784131e+00 3.54360312e-01 4.99270827e-01 1.77437626e-02
9.21141744e-01 5.48123240e-01 3.10840964e-01 -1.02976918e-01
-6.22277081e-01 -8.94237876e-01 6.17875099e-01 1.02262878e+00
2.20978901e-01 -4.86345619e-01 -1.37082601e+00 8.31754386e-01
-1.69565654e+00 -8.25050175e-01 3.71955395e-01 1.72319603e+00
1.06230235e+00 -5.79579473e-02 -2.01952577e-01 -4.88126904e-01
8.85728002e-02 9.02069360e-02 -7.25478768e-01 -2.58151293e-01
1.43381478e-02 6.91330433e-02 1.20644234e-01 8.11458290e-01
-8.52219224e-01 1.53026223e+00 5.85700941e+00 8.40155959e-01
-1.19022560e+00 1.14471413e-01 4.13390398e-01 -3.16495568e-01
-6.06705785e-01 -1.12054246e-02 -8.54366899e-01 -7.52161369e-02
8.40964794e-01 -1.70798719e-01 5.89018226e-01 6.55853808e-01
6.06215112e-02 -2.40744263e-01 -1.36670935e+00 1.10198247e+00
4.58079994e-01 -1.43432403e+00 5.44838548e-01 -3.71388972e-01
7.35528409e-01 2.26401001e-01 2.38011599e-01 7.22652912e-01
7.40947723e-01 -1.25861990e+00 8.17530274e-01 2.91244686e-01
9.38880324e-01 -3.12444597e-01 3.39487158e-02 3.93060863e-01
-1.18222558e+00 1.01173542e-01 4.25191559e-02 6.45668283e-02
2.35387180e-02 -7.06138313e-02 -1.06336522e+00 2.93187201e-01
7.71929622e-01 8.41238201e-01 -7.02092648e-01 5.25457025e-01
-7.14854896e-01 8.33109677e-01 -3.81505907e-01 9.89027768e-02
2.70806700e-01 6.86037540e-02 1.65762663e-01 9.41462696e-01
4.80585434e-02 3.36687624e-01 4.78655279e-01 1.05654180e+00
-3.19818109e-01 -8.28801841e-02 -6.30251765e-01 -4.48698223e-01
3.46029967e-01 9.30695415e-01 -3.91576916e-01 -5.10355592e-01
-8.87871265e-01 7.66810238e-01 7.01775551e-01 7.89874315e-01
-8.51599097e-01 1.41274720e-01 6.70055985e-01 -3.62331197e-02
1.16240904e-01 -4.92823601e-01 -1.42108975e-02 -1.62349665e+00
-2.78421700e-01 -1.04330480e+00 5.86297750e-01 -1.57536173e+00
-1.01022279e+00 1.12187624e-01 4.81224239e-01 -7.88647711e-01
-2.58587837e-01 -8.76948714e-01 -2.49273419e-01 5.51123738e-01
-1.59740949e+00 -1.65983748e+00 -5.14426827e-01 9.21756148e-01
8.15676510e-01 -1.58493400e-01 5.73726535e-01 4.27034358e-03
-4.13286030e-01 6.00951433e-01 -3.18998814e-01 3.78522396e-01
9.03465629e-01 -1.27709472e+00 2.79743344e-01 6.55319333e-01
7.12903023e-01 4.72495854e-01 4.01775658e-01 -7.13400543e-01
-1.58135283e+00 -8.93206596e-01 5.24072170e-01 -8.08309913e-01
1.04320121e+00 -5.04876256e-01 -9.58423734e-01 1.30656660e+00
4.25937742e-01 -1.06038928e-01 6.10711873e-01 2.03092381e-01
-8.45120251e-01 5.65747023e-02 -6.99690521e-01 8.20334792e-01
1.12085891e+00 -9.86144185e-01 -8.07527184e-01 4.60587204e-01
1.15031981e+00 -6.35994196e-01 -7.65692353e-01 2.31211901e-01
5.73301971e-01 -6.22965693e-01 1.12574577e+00 -8.36897850e-01
4.76411581e-01 -2.66444266e-01 -1.54294623e-02 -1.23255324e+00
6.01804852e-02 -2.80844688e-01 -2.06537843e-01 1.03620815e+00
3.58663976e-01 -2.00744629e-01 8.71249974e-01 6.57523930e-01
3.74330883e-03 -6.12762332e-01 -7.50408173e-01 -7.47955441e-01
2.20092744e-01 -7.81440139e-01 2.12711096e-01 1.11511016e+00
1.90959632e-01 5.11988044e-01 -2.47339606e-01 3.40848386e-01
4.46934223e-01 9.78054255e-02 9.51181889e-01 -7.61268377e-01
-3.38939220e-01 -1.84722394e-01 -2.31214955e-01 -1.49484420e+00
5.89375436e-01 -1.22732234e+00 -1.59746051e-01 -1.62667942e+00
4.26314890e-01 -2.26766735e-01 -1.75620601e-01 1.12139916e+00
-9.80222225e-03 2.22791836e-01 4.90240872e-01 2.89657079e-02
-9.18643594e-01 6.55583084e-01 1.54169488e+00 -5.48416555e-01
-5.79770990e-02 -6.78112805e-01 -4.29654121e-01 8.95452380e-01
5.21430790e-01 -9.83525366e-02 -9.84371662e-01 -1.02095079e+00
4.70413029e-01 1.10528842e-01 7.23048329e-01 -7.31140316e-01
4.44200993e-01 -3.79752874e-01 3.18139076e-01 -6.45115316e-01
5.78136384e-01 -6.54558182e-01 -2.94911683e-01 2.19775885e-01
-4.99341905e-01 1.20685007e-02 6.04292870e-01 4.36674118e-01
-2.29691550e-01 1.16427109e-01 5.71001887e-01 -4.37243562e-03
-1.28620768e+00 2.39004061e-01 -1.71939924e-01 3.63783985e-01
9.57439244e-01 -1.03154052e-02 -7.74669230e-01 -7.33800948e-01
-8.24367404e-01 8.22053075e-01 3.91514719e-01 7.52906561e-01
8.17766666e-01 -1.23778617e+00 -4.90655601e-01 3.06308597e-01
4.27992404e-01 1.57658741e-01 2.95806259e-01 6.87753379e-01
-4.52819914e-01 4.98590082e-01 3.86197381e-02 -8.45438838e-01
-1.51460147e+00 7.64092326e-01 3.94522816e-01 -2.09151171e-02
-6.49548650e-01 1.11443079e+00 8.50607097e-01 -5.41599214e-01
4.52868223e-01 -6.67112708e-01 2.23889463e-02 -1.84995353e-01
4.56183523e-01 7.42106873e-04 -3.29558462e-01 -4.44690049e-01
-2.03159019e-01 7.31786072e-01 -1.87189817e-01 -2.59390652e-01
9.95303035e-01 -2.07830310e-01 2.57870853e-01 3.94697517e-01
8.88303578e-01 -5.22424802e-02 -1.27934742e+00 -4.14904773e-01
-1.38842255e-01 -1.82201505e-01 1.11117341e-01 -1.19986439e+00
-9.64944005e-01 1.14534950e+00 4.85919297e-01 -3.49643171e-01
1.03826368e+00 4.49960232e-01 6.14578426e-01 8.27873230e-01
2.18057483e-01 -9.24508274e-01 6.46871448e-01 7.11076975e-01
9.04783130e-01 -1.51166749e+00 2.65292712e-02 -2.62484342e-01
-7.65876889e-01 9.11218286e-01 1.26478577e+00 4.49247509e-01
5.07679939e-01 1.95709541e-01 4.28846747e-01 -2.17189223e-01
-1.18549478e+00 -3.26458454e-01 4.96540248e-01 6.06545746e-01
4.06621754e-01 -7.26002902e-02 4.46069807e-01 3.94663572e-01
-3.94490719e-01 1.62398905e-01 4.34424579e-01 6.47569656e-01
-3.09953570e-01 -1.00875759e+00 -1.51425391e-01 1.99430324e-02
6.83163255e-02 -4.19946760e-01 -3.44915926e-01 1.17088246e+00
3.00822547e-03 7.74860799e-01 3.82908136e-02 -2.66962200e-01
1.49680346e-01 3.12444866e-01 6.81549132e-01 -8.15232873e-01
-2.46613786e-01 -2.46393710e-01 3.58090661e-02 -8.11415076e-01
-4.14284617e-01 -1.34224117e-01 -1.67103231e+00 -2.67816603e-01
-1.65709451e-01 1.09528087e-01 4.32526201e-01 1.04862821e+00
1.57461867e-01 4.42790657e-01 -1.46251470e-01 -4.04714167e-01
-2.78961390e-01 -6.27529800e-01 -2.55771101e-01 5.42413354e-01
4.11660016e-01 -7.05977082e-01 -2.75614262e-01 3.33423078e-01] | [10.738512992858887, 1.689260721206665] |
e2f93982-ba9f-4094-b90e-19c880036e49 | biorex-improving-biomedical-relation | 2306.11189 | null | https://arxiv.org/abs/2306.11189v1 | https://arxiv.org/pdf/2306.11189v1.pdf | BioREx: Improving Biomedical Relation Extraction by Leveraging Heterogeneous Datasets | Biomedical relation extraction (RE) is the task of automatically identifying and characterizing relations between biomedical concepts from free text. RE is a central task in biomedical natural language processing (NLP) research and plays a critical role in many downstream applications, such as literature-based discovery and knowledge graph construction. State-of-the-art methods were used primarily to train machine learning models on individual RE datasets, such as protein-protein interaction and chemical-induced disease relation. Manual dataset annotation, however, is highly expensive and time-consuming, as it requires domain knowledge. Existing RE datasets are usually domain-specific or small, which limits the development of generalized and high-performing RE models. In this work, we present a novel framework for systematically addressing the data heterogeneity of individual datasets and combining them into a large dataset. Based on the framework and dataset, we report on BioREx, a data-centric approach for extracting relations. Our evaluation shows that BioREx achieves significantly higher performance than the benchmark system trained on the individual dataset, setting a new SOTA from 74.4% to 79.6% in F-1 measure on the recently released BioRED corpus. We further demonstrate that the combined dataset can improve performance for five different RE tasks. In addition, we show that on average BioREx compares favorably to current best-performing methods such as transfer learning and multi-task learning. Finally, we demonstrate BioREx's robustness and generalizability in two independent RE tasks not previously seen in training data: drug-drug N-ary combination and document-level gene-disease RE. The integrated dataset and optimized method have been packaged as a stand-alone tool available at https://github.com/ncbi/BioREx. | ['Zhiyong Lu', 'Qingyu Chen', 'Ling Luo', 'Chih-Hsuan Wei', 'Po-Ting Lai'] | 2023-06-19 | null | null | null | null | ['graph-construction', 'multi-task-learning', 'transfer-learning', 'relation-extraction'] | ['graphs', 'methodology', 'miscellaneous', 'natural-language-processing'] | [ 3.75786573e-01 3.24078314e-02 -4.48164850e-01 -1.03105359e-01
-1.22312665e+00 -5.09677708e-01 5.05361497e-01 8.50042045e-01
-4.84779000e-01 1.16465807e+00 1.02114551e-01 -6.00803077e-01
-4.84523177e-01 -5.39634466e-01 -6.99525654e-01 -6.43524885e-01
-1.61340386e-01 7.69057274e-01 9.85922106e-03 -6.58310279e-02
-9.50514376e-02 4.42126602e-01 -9.13342476e-01 4.42879468e-01
7.94154823e-01 4.85178411e-01 8.12246054e-02 5.55771410e-01
1.14440415e-02 4.82053250e-01 -4.71850038e-01 -3.06049615e-01
-3.11129391e-01 -4.79300320e-01 -1.17615211e+00 -5.77877045e-01
-9.53282043e-02 4.04311091e-01 -3.57033461e-02 6.68308258e-01
7.24625647e-01 -2.63432264e-01 7.43738294e-01 -1.09212863e+00
-4.24152523e-01 5.47263741e-01 -3.44330877e-01 2.88435280e-01
4.21654224e-01 -4.13253158e-02 1.02393770e+00 -7.71120667e-01
1.16277850e+00 9.71710622e-01 7.18303621e-01 6.44127905e-01
-1.42102480e+00 -7.83335865e-01 -4.47598100e-01 8.33666623e-02
-1.58192623e+00 -5.37507534e-01 2.70341724e-01 -5.85051954e-01
1.54304087e+00 4.13790613e-01 1.88467339e-01 1.12748528e+00
3.89118940e-01 4.49316591e-01 1.08003545e+00 -5.37956953e-01
-4.93049398e-02 -1.22667171e-01 2.15771824e-01 7.59591162e-01
3.03303033e-01 -2.91331232e-01 -5.05494356e-01 -5.89706421e-01
1.90521106e-01 -1.57637343e-01 -4.15015757e-01 7.12014735e-02
-1.46444464e+00 6.23058915e-01 1.46520332e-01 6.58159614e-01
-3.35756779e-01 -1.12477146e-01 5.68829536e-01 3.36111456e-01
7.89969087e-01 7.70524800e-01 -1.11415088e+00 -2.86539970e-03
-6.68287158e-01 -8.27436522e-03 9.91906881e-01 8.45353544e-01
3.73451799e-01 -7.87751853e-01 -2.52842546e-01 1.11856890e+00
5.58748394e-02 4.97630611e-02 5.98638415e-01 -3.05096388e-01
2.89549977e-01 6.73930526e-01 -3.05273443e-01 -8.15628231e-01
-8.25215816e-01 -3.84530187e-01 -1.02152777e+00 -5.08245707e-01
4.81376350e-01 -1.81890756e-01 -8.38622212e-01 1.73885465e+00
4.21990097e-01 3.12109262e-01 3.55048120e-01 4.42194372e-01
1.24660909e+00 2.52554089e-01 6.17654264e-01 -2.84667343e-01
1.72772050e+00 -5.84796369e-01 -8.47396612e-01 7.24840239e-02
1.32147765e+00 -8.18077326e-01 6.12881422e-01 2.66074717e-01
-6.58578277e-01 -5.27931787e-02 -7.80230522e-01 -3.43644351e-01
-9.30031896e-01 3.38861227e-01 7.22661257e-01 1.91532806e-01
-9.18527544e-01 6.09654307e-01 -6.89907849e-01 -6.31558716e-01
6.99005008e-01 6.08556926e-01 -8.43186796e-01 -1.15903713e-01
-1.48071146e+00 1.18096197e+00 5.74639916e-01 5.60394861e-02
-6.02681756e-01 -1.19129944e+00 -6.63250208e-01 -6.64638355e-02
5.41902661e-01 -9.91802990e-01 8.86723995e-01 -3.46362800e-03
-1.11464190e+00 1.22892427e+00 -2.17739508e-01 -3.83466214e-01
2.34270856e-01 -2.83544451e-01 -5.20811617e-01 5.64574823e-02
1.75482795e-01 3.57454777e-01 -2.05708131e-01 -7.64289141e-01
-2.94217467e-01 -3.25014979e-01 -4.21899736e-01 -2.20995992e-01
-1.72675967e-01 2.35997006e-01 -5.88132024e-01 -3.98897111e-01
-3.88064116e-01 -9.26304519e-01 -3.79990578e-01 -3.00802022e-01
-7.18812108e-01 -5.61075449e-01 4.89141524e-01 -6.15668714e-01
1.26460564e+00 -1.80031025e+00 1.94205284e-01 1.30328745e-01
4.98969555e-01 4.89003599e-01 -9.81580466e-02 6.94649160e-01
-5.66192031e-01 4.16798830e-01 -1.82796046e-01 -4.26100427e-03
-3.57290119e-01 4.72014099e-02 3.48568767e-01 5.02986312e-01
4.56302196e-01 1.15440571e+00 -1.09829175e+00 -6.67376578e-01
-1.78626463e-01 5.05848765e-01 -2.18454778e-01 -2.77833617e-03
-3.11570495e-01 5.15007555e-01 -6.11722767e-01 7.07643390e-01
2.28528813e-01 -6.94388092e-01 5.59449792e-01 -3.98486376e-01
2.09184676e-01 4.58238751e-01 -5.23743391e-01 1.78641629e+00
-3.47911775e-01 3.02358538e-01 -2.57032454e-01 -1.21279967e+00
8.68752241e-01 6.24469221e-01 8.61897945e-01 -3.48425835e-01
4.48842831e-02 3.10516387e-01 2.19248235e-01 -6.92283869e-01
-1.19692266e-01 -1.27777398e-01 8.46999362e-02 1.31435528e-01
3.04446906e-01 1.87017456e-01 2.91243702e-01 3.18527609e-01
1.56795382e+00 5.32645471e-02 1.00363696e+00 -4.32483792e-01
6.20129347e-01 1.98485062e-01 4.90076214e-01 2.96185344e-01
1.07778773e-01 2.25774884e-01 7.86417723e-01 -1.08339816e-01
-6.20137513e-01 -4.60789919e-01 -5.50598741e-01 8.60318542e-01
-4.24842477e-01 -8.27828467e-01 -4.24316138e-01 -8.24756861e-01
9.29291397e-02 3.93559605e-01 -7.44052947e-01 -1.19087949e-01
-3.25216383e-01 -1.44600916e+00 9.74478006e-01 6.97413012e-02
1.27169102e-01 -6.96394265e-01 -1.03834970e-02 4.27193761e-01
-3.01623255e-01 -1.40333915e+00 -2.67066926e-01 4.54652905e-01
-6.23007298e-01 -1.60190547e+00 -5.93918085e-01 -6.27167940e-01
5.18692970e-01 -2.01972619e-01 1.20206010e+00 5.79086579e-02
-6.02529824e-01 1.05234168e-01 -2.87691832e-01 -6.03944063e-01
-5.60951948e-01 3.82557809e-01 -1.64231852e-01 -4.40756857e-01
7.26645231e-01 -3.25368732e-01 -5.48389316e-01 7.53383562e-02
-7.27806985e-01 1.49070367e-01 8.09548080e-01 1.06385231e+00
8.04752827e-01 -2.33525202e-01 1.13365734e+00 -1.49869621e+00
7.57233977e-01 -8.48061442e-01 -1.87473536e-01 4.98787820e-01
-8.58869314e-01 7.48765171e-02 4.00789678e-01 -2.99979120e-01
-7.00392306e-01 2.40724310e-01 -2.66993731e-01 1.56661086e-02
-1.87768131e-01 1.03212857e+00 -1.30199730e-01 1.16568595e-01
7.58137941e-01 -6.10027537e-02 2.44988054e-01 -5.10475576e-01
1.38716355e-01 8.03509057e-01 2.84312338e-01 -4.76648837e-01
1.94880620e-01 -1.03984568e-02 5.32072902e-01 -6.60226762e-01
-8.44196558e-01 -8.39718401e-01 -6.21678531e-01 2.71363646e-01
1.00194883e+00 -9.37166214e-01 -8.09879303e-01 2.44141489e-01
-1.08291543e+00 -4.12820816e-01 2.50441998e-01 5.04573882e-01
-1.52405620e-01 1.80425316e-01 -7.68857896e-01 -2.59741426e-01
-7.10786581e-01 -1.07032931e+00 1.34938598e+00 -3.15414727e-01
-4.90749806e-01 -1.30042624e+00 2.65649050e-01 4.70548391e-01
-5.43935150e-02 5.75322926e-01 1.37108648e+00 -1.27320385e+00
-1.66273378e-02 -1.88374713e-01 -2.45104194e-01 -1.50781095e-01
5.20919442e-01 -5.68366237e-02 -8.55583251e-01 -1.04683833e-02
-5.48198223e-01 -2.92929292e-01 9.26224709e-01 1.37545586e-01
1.16494787e+00 -3.79809290e-02 -1.08952105e+00 3.02994370e-01
1.43745124e+00 6.24603331e-02 5.06040990e-01 1.70966029e-01
7.11587965e-01 5.95142841e-01 5.37956715e-01 4.54850607e-02
1.80766612e-01 7.12983012e-01 -1.65122077e-01 -3.88471097e-01
-1.63165405e-01 -2.24132519e-02 -1.16359778e-01 6.39154017e-01
-1.38789743e-01 -3.32477450e-01 -1.36942101e+00 4.39425528e-01
-1.87544584e+00 -5.60399711e-01 -5.26594162e-01 1.92622948e+00
1.50900614e+00 -1.08826458e-01 -4.23171222e-02 -3.19991447e-02
5.54480493e-01 -5.73663473e-01 -2.74810135e-01 -1.67010546e-01
-1.47141725e-01 5.52812994e-01 4.23131406e-01 3.80388767e-01
-1.09089899e+00 1.02251577e+00 6.02111483e+00 9.91426885e-01
-9.54752326e-01 2.36078680e-01 7.20362008e-01 1.00008830e-01
2.55943239e-02 -2.52539039e-01 -8.81672084e-01 2.28827327e-01
1.27508271e+00 -4.31907237e-01 -1.48860857e-01 4.83749390e-01
3.46915871e-01 2.16020755e-02 -1.44421005e+00 8.04456651e-01
-1.60614580e-01 -1.64763713e+00 -1.48523435e-01 2.71405458e-01
2.93969363e-01 2.67222315e-01 -4.33208495e-01 2.29825065e-01
2.86080360e-01 -1.24821687e+00 -2.80051500e-01 3.97022128e-01
1.22159052e+00 -2.47794285e-01 1.02069974e+00 8.44198987e-02
-9.36210334e-01 3.92320544e-01 -1.23990551e-02 3.52322996e-01
-9.40566212e-02 1.12324882e+00 -1.24192965e+00 1.05130231e+00
5.04071474e-01 9.43195045e-01 -6.19369090e-01 8.56901705e-01
9.42336582e-03 6.13545120e-01 -1.84606478e-01 -4.56746630e-02
-8.45601112e-02 1.29651487e-01 2.35230908e-01 1.70409262e+00
8.86983573e-02 3.05593491e-01 2.37293229e-01 5.27767479e-01
-3.76986176e-01 4.85988021e-01 -5.93207121e-01 -1.93909138e-01
2.83278495e-01 1.48582435e+00 -6.75929666e-01 -3.90184402e-01
-4.18511182e-01 6.11624718e-01 5.30077100e-01 2.82799035e-01
-6.88909411e-01 -3.38279754e-01 4.92450058e-01 -5.39291799e-02
-5.12747467e-02 1.42406881e-01 -1.26339942e-01 -9.76951301e-01
-4.31632131e-01 -8.71628225e-01 7.43270278e-01 -3.32163870e-01
-1.52613723e+00 5.50626397e-01 8.79166871e-02 -8.80270123e-01
-1.97831213e-01 -7.54731774e-01 -4.49512117e-02 9.18803394e-01
-1.52729321e+00 -1.20451224e+00 4.69972473e-03 4.18602973e-01
1.03537396e-01 -1.01505302e-01 1.43911517e+00 6.49895489e-01
-8.39885414e-01 6.53461456e-01 1.90310746e-01 8.00733790e-02
1.19062328e+00 -1.22804320e+00 6.26225397e-02 2.43186086e-01
-9.97406170e-02 8.59084249e-01 4.80260760e-01 -7.07301855e-01
-1.35809064e+00 -1.30818963e+00 1.31340504e+00 -5.12073338e-01
8.32964361e-01 -5.13788283e-01 -1.10684717e+00 6.74726903e-01
-7.84335583e-02 1.27854198e-01 1.42219937e+00 5.84503829e-01
-3.41979384e-01 1.41356409e-01 -1.09311116e+00 4.37880218e-01
1.05332625e+00 -4.41950470e-01 -3.97991389e-01 9.47354078e-01
4.86649066e-01 -3.99382263e-01 -1.49042547e+00 5.13955116e-01
3.38290155e-01 -2.06132337e-01 9.69298720e-01 -1.05078638e+00
5.35349727e-01 -6.36138022e-02 2.45712191e-01 -1.21314144e+00
-3.88578475e-01 -6.64670110e-01 -8.12818669e-03 1.29939318e+00
9.86604095e-01 -8.69911671e-01 4.19075817e-01 3.42366666e-01
-1.22264452e-01 -1.21557653e+00 -9.27483797e-01 -5.86948574e-01
2.46549278e-01 -1.61489770e-01 2.72009045e-01 1.33143461e+00
4.34421003e-01 9.54020858e-01 1.20597184e-02 -3.73024866e-02
2.27449641e-01 -6.56734109e-02 5.45029104e-01 -1.34483480e+00
-1.73601508e-01 -3.32438976e-01 -3.31134856e-01 -3.17977756e-01
3.33816320e-01 -1.48165667e+00 -1.11974940e-01 -1.74973416e+00
5.55610895e-01 -4.86972451e-01 -5.56142449e-01 1.01242793e+00
-4.66455072e-01 1.39780104e-01 -4.47714418e-01 2.14552879e-01
-4.86784607e-01 2.91491039e-02 1.10454309e+00 -1.64914280e-01
-1.81652397e-01 -2.83108920e-01 -9.15477872e-01 3.51030052e-01
9.24410820e-01 -6.83581233e-01 -3.07765156e-01 3.43172289e-02
1.68515801e-01 2.17736163e-03 2.15077505e-01 -4.81108159e-01
3.11734647e-01 8.92182440e-03 2.88888216e-01 -2.40042776e-01
-2.98117995e-02 -4.57196981e-01 4.60979193e-01 6.35390043e-01
-4.19431865e-01 -1.72580421e-01 4.45498794e-01 5.05599797e-01
-5.71963862e-02 6.14159219e-02 4.82333094e-01 -4.97452244e-02
-3.29025507e-01 3.10407102e-01 -2.39860088e-01 5.01130000e-02
1.02203894e+00 3.24797630e-01 -7.05978394e-01 1.02998346e-01
-7.33926713e-01 3.51675421e-01 -2.09948465e-01 2.86262691e-01
3.69626909e-01 -8.33399355e-01 -1.05093658e+00 -3.57268155e-01
4.80523229e-01 -8.74252319e-02 1.32393315e-01 1.08378959e+00
-2.96548784e-01 7.93441176e-01 2.43527755e-01 -5.05653322e-01
-1.73906708e+00 7.13142276e-01 2.33808160e-01 -8.12322080e-01
-5.30794024e-01 6.38076603e-01 6.24939017e-02 -3.78752530e-01
-1.87656701e-01 -2.80245632e-01 -4.04204428e-01 5.76400794e-02
3.21942985e-01 5.50905429e-02 4.60451812e-01 -3.62152010e-01
-7.62469769e-01 3.60436916e-01 -2.85358071e-01 3.77521962e-01
1.53141916e+00 3.96680236e-01 -5.62727571e-01 3.09661984e-01
1.43802989e+00 -3.47433351e-02 -1.46485567e-01 -8.95636901e-02
5.51269233e-01 1.61762834e-01 -9.65626389e-02 -1.19494414e+00
-7.45362043e-01 4.13385510e-01 1.65922180e-01 -1.98233679e-01
9.79759753e-01 3.83458376e-01 4.75792527e-01 4.78411198e-01
9.12095979e-02 -6.32052243e-01 -4.01351660e-01 2.96335161e-01
8.35548878e-01 -1.26799715e+00 3.78430039e-01 -9.84098792e-01
-2.69981503e-01 9.23311651e-01 3.32981557e-01 4.89431262e-01
8.07454109e-01 4.48926598e-01 -3.44780535e-02 -6.24001682e-01
-1.07206774e+00 -1.90565720e-01 5.35842836e-01 5.66815853e-01
1.11068428e+00 6.40759692e-02 -7.89036870e-01 7.63973951e-01
8.52983147e-02 4.64285135e-01 1.10129170e-01 7.72662818e-01
2.23416910e-01 -1.59857726e+00 1.20611459e-01 7.68667936e-01
-1.05100346e+00 -4.84830856e-01 -5.31728685e-01 8.89864981e-01
1.29306287e-01 1.00268877e+00 -4.84852850e-01 -2.79320091e-01
2.23177224e-01 3.39121789e-01 4.48718011e-01 -8.84746492e-01
-6.80541694e-01 1.47951171e-01 6.61170542e-01 -4.20285583e-01
-6.48671627e-01 -5.83587885e-01 -1.45626867e+00 -2.04491615e-02
-4.12226260e-01 3.82712662e-01 4.20025080e-01 1.02387917e+00
8.09854031e-01 9.05421138e-01 -1.18554734e-01 -8.96756202e-02
-1.40573354e-02 -1.04800344e+00 -1.73964575e-01 3.59030545e-01
4.26938087e-02 -5.60778260e-01 -5.59322052e-02 2.52967805e-01] | [8.47409725189209, 8.732340812683105] |
505794dd-8a2c-4ac3-9cfb-53c50c210aa6 | hcr-net-a-deep-learning-based-script | 2108.06663 | null | https://arxiv.org/abs/2108.06663v3 | https://arxiv.org/pdf/2108.06663v3.pdf | HCR-Net: A deep learning based script independent handwritten character recognition network | Despite being studied extensively for a few decades, handwritten character recognition (HCR) is still considered a challenging learning problem in pattern recognition, and there is very limited research on script independent models. This is mainly because of similarity in structure of characters, different handwriting styles, noisy datasets, diversity of scripts, focus of the conventional research on handcrafted feature extraction techniques, and unavailability of public datasets and code-repositories to reproduce the results. On the other hand, deep learning has witnessed huge success in different areas of pattern recognition, including HCR, and provides an end-to-end learning. However, deep learning techniques are computationally expensive, need large amount of data for training and have been developed for specific scripts only. To address the above limitations, we have proposed a novel generic deep learning architecture for script independent handwritten character recognition, called HCR-Net. HCR-Net is based on a novel transfer learning approach for HCR, which partly utilizes feature extraction layers of a pre-trained network. Due to transfer learning and image-augmentation, HCR-Net provides faster and computationally efficient training, better performance and better generalizations, and can work with small datasets. HCR-Net is extensively evaluated on 40 publicly available datasets of Bangla, Punjabi, Hindi, English, Swedish, Urdu, Farsi, Tibetan, Kannada, Malayalam, Telugu, Marathi, Nepali and Arabic languages, and established 26 new benchmark results while performed close to the best results in the rest cases. HCR-Net showed performance improvements up to 11% against the existing results and achieved a fast convergence rate showing up to 99% of final performance in the very first epoch. HCR-Net significantly outperformed the state-of-the-art transfer learning techniques... | ['Anuj Sharma', 'Sukhdeep Singh', 'Vinod Kumar Chauhan'] | 2021-08-15 | null | null | null | null | ['image-augmentation'] | ['computer-vision'] | [-2.32967660e-02 -7.56865859e-01 -1.22142985e-01 -3.61251831e-01
-4.51430142e-01 -6.26531005e-01 5.36800683e-01 -4.31393385e-01
-5.55497289e-01 7.95632124e-01 -9.04280022e-02 -3.35498184e-01
-8.11532885e-02 -6.82706892e-01 -5.90626776e-01 -8.10631812e-01
8.51376653e-02 5.65380454e-01 3.16278517e-01 -3.92913401e-01
4.71473217e-01 8.46161723e-01 -1.05530512e+00 3.38678658e-01
7.32922792e-01 7.31589317e-01 1.45077214e-01 7.95798242e-01
1.65500373e-01 1.19437587e+00 -7.43768036e-01 -6.59415901e-01
3.84435281e-02 -4.36828643e-01 -8.18890750e-01 1.51370600e-01
1.91246644e-01 -4.43488210e-01 -8.25546443e-01 5.97349703e-01
7.36693382e-01 2.40682717e-03 7.19058454e-01 -7.16487467e-01
-1.10506535e+00 5.21440327e-01 -6.03772104e-01 1.07983351e-01
-4.76588793e-02 1.12533629e-01 5.30920208e-01 -9.53923702e-01
5.45770764e-01 8.69739413e-01 1.02905393e+00 6.83753133e-01
-8.53020966e-01 -4.44279850e-01 -2.17762008e-01 3.00658554e-01
-1.33064306e+00 4.21906039e-02 6.23520494e-01 -3.63716662e-01
1.10292768e+00 2.56979704e-01 2.96579093e-01 1.33058107e+00
1.25769109e-01 1.06143594e+00 1.39658868e+00 -4.64799553e-01
6.76594004e-02 1.43353015e-01 2.43471116e-01 7.19118476e-01
-5.81608936e-02 -3.36951464e-01 -3.60203028e-01 2.24481761e-01
1.08907425e+00 4.01973762e-02 -2.91361660e-01 1.28304094e-01
-1.08782518e+00 8.31619859e-01 1.20550431e-01 4.74259883e-01
-1.23518176e-01 -2.97978640e-01 6.65448189e-01 4.32399720e-01
9.55562294e-02 2.26282045e-01 -4.84940469e-01 -6.31982565e-01
-8.70197713e-01 1.72181353e-01 8.09059918e-01 8.66992652e-01
5.63730001e-01 4.54147190e-01 -1.31248981e-01 1.51105583e+00
-8.59577283e-02 5.14681756e-01 8.21612000e-01 -1.18614398e-01
7.39894629e-01 6.01431072e-01 -2.19189197e-01 -9.32198644e-01
-1.84888095e-01 -2.64472932e-01 -1.19215453e+00 1.11200355e-01
5.14834225e-01 -2.33592615e-01 -1.34973800e+00 1.00328302e+00
-4.50578660e-01 -2.90010333e-01 2.92929322e-01 1.10441542e+00
6.66930854e-01 9.51200426e-01 -2.81694710e-01 3.06006253e-01
8.76164734e-01 -1.33175611e+00 -4.21855092e-01 -1.06217347e-01
1.87944606e-01 -8.60091567e-01 1.07153165e+00 6.98445678e-01
-7.16517627e-01 -7.45407581e-01 -1.07706988e+00 1.23675995e-01
-4.21249628e-01 6.92904532e-01 6.64827108e-01 7.66971350e-01
-6.29126549e-01 7.47044861e-01 -7.13821352e-01 -5.25281191e-01
5.52704155e-01 3.06972563e-01 -5.95916808e-01 -4.50231761e-01
-7.69610584e-01 8.54922116e-01 6.74237370e-01 2.71517545e-01
-9.48537827e-01 -4.59872007e-01 -4.01271433e-01 -2.26544306e-01
1.05774596e-01 2.84692973e-01 8.75554204e-01 -1.07262671e+00
-2.03315330e+00 7.14514077e-01 2.93499917e-01 -2.50437737e-01
9.71539855e-01 -1.77577630e-01 -7.14298427e-01 -2.84992363e-02
-3.35608184e-01 1.34081049e-02 7.62126267e-01 -1.02310240e+00
-3.40866059e-01 -3.10455024e-01 -5.42582095e-01 -9.42511577e-03
-6.42376065e-01 2.58389413e-01 -7.91222692e-01 -8.68611693e-01
-1.19265877e-01 -8.57112110e-01 2.79077087e-02 -5.63132286e-01
-2.34120339e-01 -1.33890942e-01 1.17358279e+00 -1.00402665e+00
9.32827115e-01 -2.21063495e+00 6.24051541e-02 3.87130789e-02
-4.41473365e-01 9.55651283e-01 -3.55114013e-01 6.99440181e-01
1.03138024e-02 -1.80709109e-01 -5.40545523e-01 1.25197470e-01
-1.54879898e-01 4.44804460e-01 -4.57827270e-01 5.07919312e-01
4.03161436e-01 7.80614197e-01 -6.02016330e-01 -2.59803027e-01
3.42588335e-01 4.37790066e-01 5.43423705e-02 3.08681518e-01
8.78607407e-02 2.52053529e-01 -2.07558960e-01 1.04909575e+00
7.45261967e-01 -9.66692120e-02 8.28536451e-02 3.38795036e-02
-2.00694531e-01 -2.85443068e-01 -1.12419558e+00 1.29274666e+00
-3.68106097e-01 7.62864530e-01 -4.35004652e-01 -1.21103430e+00
1.47683537e+00 2.18333527e-01 -4.49967869e-02 -6.94233954e-01
1.92056894e-01 7.39160240e-01 1.65957183e-01 -5.57351887e-01
4.79339629e-01 2.54472017e-01 2.13234961e-01 2.74796754e-01
2.71112442e-01 1.61694847e-02 3.74570727e-01 -2.80338824e-01
8.87053728e-01 2.23397076e-01 -6.51440397e-02 4.04825695e-02
7.59473324e-01 3.92898060e-02 4.97914553e-01 8.27063918e-01
5.07846251e-02 1.09741616e+00 2.29157507e-01 -8.27754974e-01
-1.20272505e+00 -7.39484489e-01 -2.92778224e-01 1.05244684e+00
-1.65772095e-01 4.24863920e-02 -4.91316319e-01 -6.48951530e-01
-2.96783783e-02 4.32329804e-01 -6.33037865e-01 1.73275873e-01
-1.01555860e+00 -1.18804526e+00 1.02645528e+00 8.80444348e-01
1.27397096e+00 -1.35311186e+00 -1.85654178e-01 4.27795827e-01
2.05204844e-01 -1.16546190e+00 -1.63253292e-01 2.97098756e-01
-9.39674497e-01 -8.38302374e-01 -1.38961160e+00 -9.72553134e-01
5.28833508e-01 -1.05396740e-01 7.54551172e-01 -1.73863620e-01
-5.68078935e-01 -8.19690302e-02 -5.96991241e-01 -3.47688228e-01
-5.12180567e-01 2.74431944e-01 -2.67804056e-01 1.65808678e-01
3.27803850e-01 -2.07146838e-01 -2.46081471e-01 6.83747172e-01
-7.55045354e-01 -2.73642749e-01 1.03716540e+00 1.19069588e+00
1.87766954e-01 1.09094782e-02 6.16246104e-01 -8.23919594e-01
4.91699308e-01 -1.28203645e-01 -5.10395050e-01 5.56983769e-01
-4.46629941e-01 -2.41280764e-01 1.05133212e+00 -6.00419819e-01
-1.14890397e+00 3.32226977e-03 -2.59273887e-01 -1.76982492e-01
-2.70324409e-01 5.52444637e-01 -1.48794934e-01 -3.25403988e-01
6.66600943e-01 9.77800906e-01 -8.06160271e-02 -6.86153531e-01
-8.94709229e-02 1.21129000e+00 7.85160840e-01 -7.14707136e-01
7.31379211e-01 1.58903345e-01 -2.39436761e-01 -1.13759983e+00
-4.40254182e-01 -2.46717051e-01 -8.56072843e-01 -6.92630559e-02
5.62063634e-01 -7.36615360e-01 -4.06151742e-01 1.54627788e+00
-9.30880785e-01 -5.06593406e-01 1.56823024e-01 5.13278902e-01
-2.58540869e-01 6.64896190e-01 -8.79615724e-01 -6.28355265e-01
-4.82313395e-01 -1.06109488e+00 4.69120204e-01 4.14934754e-01
1.73756823e-01 -1.04946828e+00 1.13461763e-01 3.38129252e-01
6.52212560e-01 2.00796589e-01 9.32202399e-01 -8.73049676e-01
-3.07864010e-01 -7.85441995e-01 -4.61103231e-01 9.56864774e-01
4.04705107e-01 3.37018013e-01 -8.46002758e-01 -4.58251864e-01
-3.83893549e-01 -7.93488860e-01 9.88838553e-01 -2.06246644e-01
1.30125964e+00 -1.67860672e-01 9.59113017e-02 6.84135795e-01
1.57467639e+00 4.24697131e-01 1.07817674e+00 6.92045152e-01
8.34896088e-01 1.94386780e-01 3.42710286e-01 4.99363035e-01
-7.14078313e-04 4.47567582e-01 -5.12859784e-02 -1.37911364e-01
-1.36992410e-01 -5.14219068e-02 3.65778148e-01 1.09547210e+00
-7.50635326e-01 -1.81394652e-01 -1.12828112e+00 5.77037394e-01
-1.72002816e+00 -8.56381476e-01 -1.58835918e-01 2.07476425e+00
1.02951264e+00 4.43105735e-02 1.31325021e-01 3.86400878e-01
5.97517967e-01 4.08942625e-02 -4.69918996e-01 -6.55110300e-01
-5.58954298e-01 4.12107706e-01 6.85939014e-01 -1.00624673e-01
-1.29652619e+00 1.16272700e+00 5.74674892e+00 9.52804983e-01
-1.63920915e+00 -2.47836575e-01 5.30159712e-01 5.89023530e-01
4.75457788e-01 -3.74793559e-01 -8.81896019e-01 4.76205498e-01
7.04923689e-01 2.96447575e-01 3.88672888e-01 9.30994213e-01
-2.44644210e-01 1.83168590e-01 -9.18130815e-01 1.12973452e+00
3.84033114e-01 -1.39893234e+00 1.55778572e-01 -1.65728316e-01
9.86972749e-01 2.11225167e-01 7.53787905e-02 4.35508758e-01
6.26972765e-02 -1.37429547e+00 4.52295631e-01 2.47418240e-01
8.00412893e-01 -9.63645577e-01 1.16084194e+00 2.52917886e-01
-5.86188436e-01 -2.33427078e-01 -7.28157401e-01 -2.23336965e-02
-3.64070177e-01 2.56678075e-01 -5.88612914e-01 7.61553705e-01
7.03874052e-01 8.81553829e-01 -7.75846004e-01 1.03925872e+00
-2.43079573e-01 1.13863111e+00 3.88321429e-02 -3.25831115e-01
6.24816120e-01 -1.65384546e-01 2.18166243e-02 1.72059667e+00
1.82837099e-01 -4.05659378e-02 -3.76153737e-02 4.69426215e-01
-1.65345352e-02 2.44056687e-01 -2.70932615e-01 -3.02406579e-01
-8.36146623e-02 1.31999147e+00 -8.59534621e-01 -2.43931636e-01
-4.66425538e-01 1.32288265e+00 2.39715695e-01 2.56039202e-01
-9.22149599e-01 -1.04426658e+00 8.41149241e-02 -3.87678266e-01
5.57496607e-01 -3.63405943e-01 -2.13321835e-01 -1.30212820e+00
1.29166082e-01 -1.35102868e+00 1.91702709e-01 -2.22300932e-01
-1.47860157e+00 8.64749491e-01 -4.91744667e-01 -1.10503149e+00
2.76208706e-02 -1.26623571e+00 -8.52886856e-01 9.92435038e-01
-1.46489990e+00 -1.26628673e+00 -4.04314905e-01 7.33634651e-01
6.99540496e-01 -9.86987352e-01 9.51927125e-01 3.52866501e-01
-9.49284494e-01 1.06445348e+00 6.96990311e-01 9.56885040e-01
7.63527274e-01 -1.14463317e+00 4.70277429e-01 8.73986363e-01
3.07650864e-02 3.29989105e-01 1.39281765e-01 -3.55880946e-01
-1.69948399e+00 -1.18435502e+00 5.10558665e-01 -1.67350218e-01
5.62716663e-01 -3.94132346e-01 -1.09865797e+00 5.28988004e-01
1.84532285e-01 1.51958629e-01 6.25187278e-01 -1.96876764e-01
-6.33833885e-01 -1.88286677e-01 -9.39153016e-01 4.54541624e-01
3.96876901e-01 -3.31652969e-01 -5.47172189e-01 3.16523433e-01
-1.96567193e-01 -4.36838567e-01 -9.95342970e-01 2.97550917e-01
6.92519069e-01 -8.22001755e-01 6.16413474e-01 -6.16675735e-01
8.66602123e-01 -8.63462761e-02 -2.48134300e-01 -1.06961060e+00
-4.03800815e-01 -3.29654634e-01 1.47668749e-01 1.49773228e+00
5.08022070e-01 -6.04641080e-01 8.47014725e-01 3.31453502e-01
-2.23005891e-01 -7.51402855e-01 -6.69267714e-01 -1.21415901e+00
3.87177706e-01 -1.97110474e-02 3.58212382e-01 1.03264475e+00
-4.88152415e-01 7.18879774e-02 -7.64380872e-01 -6.48994520e-02
4.98789102e-01 3.60816807e-01 8.67610991e-01 -8.40081632e-01
-4.38749343e-01 -4.41187501e-01 -6.13851845e-01 -7.67421484e-01
4.03315835e-02 -7.63141394e-01 -3.08013055e-02 -1.52756000e+00
3.12953651e-01 -4.20708716e-01 -1.14220403e-01 9.14595485e-01
-4.37390953e-02 6.16652966e-01 2.85354584e-01 3.87663782e-01
-3.09818625e-01 3.68213356e-01 1.24441695e+00 -4.11660671e-01
-1.52986422e-01 8.07931870e-02 -1.50206208e-01 5.93543172e-01
9.21417534e-01 -3.28232124e-02 7.07791820e-02 -7.55257666e-01
-2.85971999e-01 -2.61412561e-01 1.49409339e-01 -1.11646748e+00
1.37426063e-01 -5.48316725e-02 8.49862337e-01 -5.42563200e-01
8.75919461e-02 -4.98523206e-01 -1.39484733e-01 4.95229542e-01
-1.41617820e-01 -1.94636300e-01 2.58340389e-01 1.73259616e-01
-4.51280683e-01 -5.05127251e-01 9.91436303e-01 -1.97883919e-01
-1.12096763e+00 8.17191824e-02 -5.91126084e-01 -5.52120209e-02
8.44061434e-01 -5.73518515e-01 -2.16899246e-01 1.12246715e-01
-2.73889750e-01 -2.63996094e-01 1.54300913e-01 6.09017909e-01
6.83940232e-01 -1.05156422e+00 -1.18139815e+00 1.67397499e-01
1.47466421e-01 -2.35251263e-01 2.36155152e-01 5.40809274e-01
-1.16960990e+00 4.89397019e-01 -7.53280878e-01 -5.69900036e-01
-1.16225922e+00 1.72713637e-01 1.89956546e-01 -2.15475231e-01
-5.74971139e-01 6.92794383e-01 -5.29341400e-01 -7.39163101e-01
3.85828406e-01 5.62625714e-02 -6.31682649e-02 -1.97484553e-01
4.97820348e-01 5.72126269e-01 4.18648094e-01 -5.81225693e-01
-3.33288789e-01 7.73345768e-01 -4.69858438e-01 2.82342196e-01
1.64944911e+00 6.07537150e-01 -2.32649878e-01 3.29947025e-01
1.26014292e+00 -1.72389358e-01 -1.19382620e+00 -6.90380931e-02
8.72930661e-02 -4.69471216e-01 -2.78864622e-01 -1.22457433e+00
-1.18419302e+00 9.16965425e-01 5.48854351e-01 -2.92331219e-01
1.07333541e+00 -5.98125935e-01 6.36322856e-01 8.85286033e-01
3.49113494e-01 -1.26902604e+00 1.30403876e-01 1.04797423e+00
8.92223120e-01 -1.32762206e+00 -5.60170449e-02 5.61325550e-02
-9.22559381e-01 1.91097617e+00 6.34362698e-01 -2.10056245e-01
2.61179417e-01 2.22746149e-01 3.14522594e-01 3.87044400e-01
-2.25218445e-01 3.93553138e-01 1.12460338e-01 5.43404996e-01
6.38344944e-01 -1.04131058e-01 -3.00034136e-01 6.34843051e-01
3.20652276e-02 1.61816791e-01 7.21527338e-01 1.22175300e+00
7.65084475e-03 -1.21588981e+00 -3.45763296e-01 3.44266027e-01
-6.83159828e-01 1.00564577e-01 -4.34457213e-01 1.08221889e+00
-1.91420197e-01 6.00632310e-01 -2.71619141e-01 -4.11334693e-01
3.42363983e-01 -7.54082063e-03 5.74242592e-01 -1.25093818e-01
-7.60952711e-01 -2.15480074e-01 -1.47242159e-01 -1.60571616e-02
-1.71459630e-01 -4.95129436e-01 -8.74681771e-01 -2.67999053e-01
-1.53129235e-01 2.02338919e-02 7.09257066e-01 8.43981743e-01
1.00172505e-01 1.94965258e-01 8.91332269e-01 -6.72966599e-01
-9.23903525e-01 -1.14134645e+00 -6.88291013e-01 3.40689093e-01
-1.24027029e-01 -1.51734591e-01 1.04198111e-02 1.53283447e-01] | [11.884960174560547, 2.6029372215270996] |
665cc9c3-6c87-4e67-ab84-ba9676aaa95a | image-captioning-with-unseen-objects | 1908.00047 | null | https://arxiv.org/abs/1908.00047v1 | https://arxiv.org/pdf/1908.00047v1.pdf | Image Captioning with Unseen Objects | Image caption generation is a long standing and challenging problem at the intersection of computer vision and natural language processing. A number of recently proposed approaches utilize a fully supervised object recognition model within the captioning approach. Such models, however, tend to generate sentences which only consist of objects predicted by the recognition models, excluding instances of the classes without labelled training examples. In this paper, we propose a new challenging scenario that targets the image captioning problem in a fully zero-shot learning setting, where the goal is to be able to generate captions of test images containing objects that are not seen during training. The proposed approach jointly uses a novel zero-shot object detection model and a template-based sentence generator. Our experiments show promising results on the COCO dataset. | ['Nazli Ikizler-Cinbis', 'Ramazan Gokberk Cinbis', 'Berkan Demirel'] | 2019-07-31 | null | null | null | null | ['zero-shot-object-detection'] | ['computer-vision'] | [ 9.40455496e-01 6.32708430e-01 1.45243220e-02 -3.65669489e-01
-9.90036964e-01 -3.82572234e-01 8.66290748e-01 -1.07406996e-01
-2.55161494e-01 8.00968289e-01 -5.47440313e-02 -1.04178965e-01
3.90026033e-01 -7.87804782e-01 -1.16789949e+00 -5.89902520e-01
5.18574655e-01 6.44073308e-01 2.52329886e-01 -9.42254625e-03
2.26711240e-02 -3.61482240e-02 -1.84184229e+00 4.85015929e-01
6.49515033e-01 7.76691318e-01 4.43795413e-01 7.97825038e-01
-2.11070091e-01 8.13438952e-01 -5.92747152e-01 -5.46433628e-01
1.67263404e-01 -6.75790846e-01 -5.76993167e-01 7.63228953e-01
7.57894099e-01 -1.03007607e-01 -6.34393767e-02 1.16440535e+00
3.66511017e-01 -5.15368544e-02 6.95868731e-01 -1.40640783e+00
-6.75267220e-01 3.59673291e-01 -3.13488275e-01 2.02675685e-02
3.18224013e-01 5.67209125e-01 7.94172585e-01 -1.05133140e+00
7.76288331e-01 1.02716708e+00 1.18710492e-02 1.00973928e+00
-1.30142081e+00 -3.47603440e-01 -4.99415807e-02 1.95760965e-01
-1.11740172e+00 -5.90547442e-01 8.88971746e-01 -6.48437917e-01
5.65503597e-01 2.41976157e-01 1.03205264e-01 1.31743109e+00
-2.46118814e-01 8.19267750e-01 9.55440640e-01 -8.07998478e-01
5.23145676e-01 6.23245478e-01 9.87418294e-02 3.65997612e-01
3.11877459e-01 -4.69815172e-02 -3.39961261e-01 -2.67700050e-02
4.24837172e-01 -1.64535582e-01 -3.18916947e-01 -7.45915830e-01
-1.28729475e+00 9.01616693e-01 3.49337518e-01 3.76157612e-01
-4.94505078e-01 5.82257919e-02 1.29309282e-01 -1.64400086e-01
4.05081838e-01 3.33203107e-01 -4.21821028e-02 3.02375108e-01
-8.49439383e-01 1.81840718e-01 6.88032568e-01 1.25521743e+00
7.36163318e-01 3.39509845e-02 -8.42893541e-01 7.20741212e-01
1.88395396e-01 6.14694655e-01 5.38957298e-01 -5.65099359e-01
6.94154024e-01 4.36013252e-01 3.68063629e-01 -6.31024182e-01
2.32469663e-01 -4.08432990e-01 -5.48398256e-01 1.38029337e-01
3.20007384e-01 -1.54080391e-01 -1.26444030e+00 1.72810555e+00
3.34146559e-01 4.39705491e-01 5.54809690e-01 1.01941812e+00
8.33035469e-01 8.79560828e-01 1.55597419e-01 -2.79920459e-01
1.38360596e+00 -1.35471439e+00 -6.63611412e-01 -6.61596000e-01
4.38188314e-01 -6.89265430e-01 1.04557526e+00 -2.42825225e-02
-9.97197270e-01 -6.31331980e-01 -1.06559539e+00 2.26755589e-01
-2.66640186e-01 2.48787954e-01 1.22696146e-01 6.69721901e-01
-8.25241685e-01 2.81831734e-02 -2.24387854e-01 -5.84172249e-01
5.43574631e-01 6.98321387e-02 -1.47742152e-01 -3.64090502e-01
-9.41220582e-01 8.36526930e-01 6.16297603e-01 -6.54220358e-02
-1.18340993e+00 -1.55341297e-01 -1.15046823e+00 2.74319977e-01
5.74911535e-01 -6.95598304e-01 1.47046900e+00 -1.36748314e+00
-1.12552357e+00 1.01328170e+00 -2.07203478e-01 -6.31537735e-01
5.39201260e-01 9.86293554e-02 -2.28030756e-02 2.28994444e-01
2.65850812e-01 1.06441903e+00 1.09279466e+00 -1.65400565e+00
-3.64996523e-01 -1.33406177e-01 -8.40782002e-02 2.20752820e-01
-1.94793209e-01 -6.42409101e-02 -3.00848871e-01 -3.62582147e-01
-3.18534404e-01 -9.18539166e-01 -4.49057370e-01 3.77677791e-02
-4.47684497e-01 -1.33650005e-01 7.28612661e-01 -2.60990024e-01
4.06559050e-01 -1.96755135e+00 -1.14172533e-01 -2.53366530e-01
-3.09256822e-01 5.64134598e-01 -5.28465450e-01 2.94605851e-01
-5.89193329e-02 1.10612875e-02 -5.88377714e-01 -4.39616382e-01
-1.42227739e-01 2.20075920e-01 -6.59933567e-01 1.22780383e-01
7.10013688e-01 1.11804330e+00 -1.00560248e+00 -6.66676760e-01
3.48128378e-01 2.79806674e-01 -1.60709441e-01 6.28484845e-01
-7.68965781e-01 3.93259972e-01 -3.19431394e-01 3.78519207e-01
6.42954707e-01 -4.07058030e-01 1.77730378e-02 1.37837663e-01
7.30912238e-02 -2.59853363e-01 -9.35719430e-01 1.60760546e+00
-3.03022385e-01 4.70281154e-01 -2.70223767e-01 -1.03989005e+00
9.14568603e-01 5.66499770e-01 8.17738101e-02 -4.44065481e-01
2.22699881e-01 5.78379370e-02 3.43619995e-02 -7.37948298e-01
3.41580838e-01 -3.74113977e-01 -1.43787771e-01 4.43875790e-01
3.81789267e-01 -2.19460756e-01 4.34850693e-01 2.14568511e-01
9.96017158e-01 1.44122019e-01 2.75487989e-01 1.77693978e-01
5.36382496e-01 2.52854735e-01 3.76632482e-01 1.09195900e+00
-2.17279539e-01 1.30727899e+00 3.63656878e-01 -2.89335042e-01
-1.45197511e+00 -9.30395782e-01 9.80544016e-02 6.74987555e-01
2.18947068e-01 1.15648828e-01 -9.70149338e-01 -7.19459414e-01
-4.99329925e-01 1.06128407e+00 -6.74355626e-01 -1.18455619e-01
-2.57668704e-01 -4.62317824e-01 6.01079762e-02 1.01799436e-01
2.19228432e-01 -1.60295630e+00 -8.36972177e-01 1.66052744e-01
-3.28447878e-01 -1.52468991e+00 -3.11794907e-01 -5.19283973e-02
-5.95175266e-01 -9.34395909e-01 -1.23888862e+00 -1.18046892e+00
1.11758864e+00 4.01845098e-01 1.00115561e+00 6.77088127e-02
-4.71101969e-01 4.49695110e-01 -5.77557087e-01 -7.12224305e-01
-7.79476464e-01 -1.17127821e-01 -3.02026302e-01 6.59458280e-01
2.64433473e-01 -4.54813167e-02 -3.01032871e-01 1.92533046e-01
-1.21017754e+00 5.07648587e-01 9.57098961e-01 9.07593369e-01
4.19843346e-01 -5.56087673e-01 7.40542650e-01 -8.29756498e-01
3.10016006e-01 -5.61262786e-01 -5.11387765e-01 5.82567811e-01
-1.22248746e-01 6.08443953e-02 6.40530765e-01 -6.40583336e-01
-1.12168241e+00 6.20225549e-01 8.76215324e-02 -6.74235344e-01
-6.36124849e-01 6.88686222e-02 -3.80845606e-01 1.26899108e-01
6.91766083e-01 7.29665041e-01 3.54160145e-02 -1.37348115e-01
3.63131374e-01 8.32189918e-01 5.60251355e-01 -3.21796477e-01
9.90070760e-01 3.93643051e-01 -3.49978507e-01 -8.91385198e-01
-1.05972528e+00 -4.67688143e-01 -4.38958287e-01 -4.58535522e-01
1.02276552e+00 -8.89717400e-01 -1.14862628e-01 2.19287872e-01
-1.72400796e+00 1.86797343e-02 -5.27582586e-01 2.94640958e-01
-9.36627328e-01 1.81514189e-01 4.42491062e-02 -1.10742640e+00
-4.52481449e-01 -1.11210418e+00 1.33816862e+00 4.72217351e-01
6.89533800e-02 -4.35598433e-01 1.01602197e-01 5.24931133e-01
2.37562031e-01 3.13371599e-01 6.57513022e-01 -8.45094323e-01
-8.54440928e-01 -3.20502758e-01 -2.83733308e-01 4.51039940e-01
-7.73597583e-02 -5.21154642e-01 -1.10553563e+00 -2.95998771e-02
1.60215631e-01 -6.22678101e-01 6.89706862e-01 4.94918972e-02
8.49516451e-01 -3.66895825e-01 -2.72337347e-01 -5.14463894e-02
1.60913932e+00 1.35929689e-01 7.93638766e-01 -6.01268299e-02
3.77048522e-01 8.48790884e-01 6.34664595e-01 1.47971109e-01
-2.94371713e-02 6.95970416e-01 5.74859500e-01 -1.05566561e-01
-1.77946672e-01 -2.69625008e-01 2.30447277e-01 2.56311864e-01
5.20588458e-01 -7.69769728e-01 -9.66631353e-01 9.78946567e-01
-2.10590649e+00 -1.14848745e+00 -1.97337031e-01 2.15167260e+00
6.84776783e-01 1.69447333e-01 -3.19144607e-01 -6.11785129e-02
1.22207248e+00 1.00817233e-01 -4.77729201e-01 -3.26913625e-01
-4.11593802e-02 6.19684979e-02 2.29933366e-01 2.52507776e-01
-1.13268411e+00 8.67991984e-01 5.60025024e+00 4.43558425e-01
-7.89452076e-01 2.89326847e-01 8.36812615e-01 7.88834020e-02
-6.80184439e-02 6.32103579e-03 -7.96018720e-01 5.54861844e-01
1.00457668e+00 -3.84534597e-01 -1.28847584e-01 1.02001405e+00
1.11903533e-01 -2.33438984e-01 -1.20802057e+00 9.82802689e-01
8.59699547e-01 -1.30339682e+00 3.35532457e-01 -1.15072861e-01
9.75858450e-01 -2.14818269e-01 3.36969495e-02 3.29317689e-01
-7.68050477e-02 -8.49297762e-01 6.51763916e-01 6.39799535e-01
5.96708179e-01 -3.44974458e-01 7.87414849e-01 6.88578904e-01
-4.75148171e-01 -7.79646039e-02 -3.19790572e-01 -3.76479663e-02
4.44383532e-01 4.28651750e-01 -1.24942040e+00 3.00483465e-01
1.87670752e-01 3.23796213e-01 -6.23501956e-01 1.45963657e+00
-2.83920258e-01 5.38125038e-01 7.44213909e-02 -2.47077405e-01
3.12447041e-01 9.87496898e-02 6.47335470e-01 9.40523088e-01
3.84871006e-01 9.68888998e-02 1.84948623e-01 1.12186182e+00
-1.90906405e-01 2.07933426e-01 -9.78443027e-01 -3.03441528e-02
-1.23839937e-01 1.32739770e+00 -8.18751156e-01 -6.99731410e-01
-4.58744526e-01 1.05482841e+00 1.08940125e-01 2.62068868e-01
-9.02851999e-01 -2.78599709e-01 1.42717198e-01 6.37964457e-02
5.62842131e-01 5.08032329e-02 -9.98688340e-02 -1.34533298e+00
2.69499958e-01 -6.21909261e-01 -5.62885776e-02 -1.19991481e+00
-1.19464695e+00 7.07518280e-01 -1.07563786e-01 -1.50402069e+00
-5.07675767e-01 -5.43392301e-01 -8.99789393e-01 6.84698522e-01
-1.49007177e+00 -1.22483802e+00 -4.07177985e-01 2.15236366e-01
1.16753483e+00 -2.70956457e-02 7.49368906e-01 1.38823492e-02
-2.40969971e-01 1.71580687e-01 -3.91377546e-02 1.29481822e-01
4.33375835e-01 -9.36418951e-01 3.74830693e-01 1.02002978e+00
4.41478729e-01 8.46092775e-02 1.10514200e+00 -6.08405948e-01
-1.04274809e+00 -1.49894011e+00 1.06176400e+00 -3.62739980e-01
2.61515141e-01 -7.54304349e-01 -8.62970471e-01 4.86058056e-01
5.30143082e-01 2.63325959e-01 4.55108345e-01 -5.27486384e-01
-5.25452122e-02 1.52113855e-01 -1.09232759e+00 4.53809500e-01
6.31115556e-01 -1.33216470e-01 -7.82786071e-01 6.41941011e-01
7.22611845e-01 -5.77604696e-02 5.46957878e-03 3.59869331e-01
8.13640580e-02 -4.49034572e-01 6.69476509e-01 -7.28765905e-01
7.43766487e-01 -3.47001284e-01 7.61405006e-02 -9.99776602e-01
2.04127952e-01 -3.38181674e-01 6.25906372e-03 1.27753508e+00
5.51012695e-01 -2.62937415e-02 9.21148419e-01 6.21618032e-01
3.96190211e-02 -2.57409126e-01 -8.62323284e-01 -9.34624791e-01
-3.63466114e-01 -2.06745252e-01 1.88740164e-01 5.58491349e-01
-3.01331103e-01 8.42378676e-01 -6.49792969e-01 -9.04809404e-03
8.55138242e-01 2.10235491e-01 9.43364143e-01 -1.10587037e+00
-3.05331200e-01 1.48395091e-01 -5.69803774e-01 -6.44557416e-01
1.39232993e-01 -9.02886629e-01 8.35626066e-01 -1.77805090e+00
5.75013995e-01 9.67452303e-03 1.47374989e-02 3.14601541e-01
-1.72053367e-01 4.58759010e-01 4.50111866e-01 8.02467763e-02
-8.37856770e-01 7.94953346e-01 1.13897908e+00 -4.70756233e-01
6.17616661e-02 2.54883245e-02 -5.30926168e-01 5.09483159e-01
7.20696807e-01 -7.05304742e-01 -4.93243873e-01 -4.30143982e-01
-2.33865738e-01 1.65657982e-01 6.21032178e-01 -1.20751774e+00
1.51963040e-01 -2.26085320e-01 2.39607349e-01 -4.24490124e-01
5.05530417e-01 -7.15823233e-01 -9.09498781e-02 3.93226296e-01
-5.16040623e-01 -4.30553734e-01 -3.79619859e-02 7.21759140e-01
-1.83770061e-01 -6.44824207e-01 8.98711741e-01 -3.13564837e-01
-6.60137594e-01 1.98105097e-01 -2.63677657e-01 -1.41023949e-01
1.68420172e+00 -1.12922482e-01 -3.08236629e-01 -3.91039461e-01
-8.61682713e-01 1.44876704e-01 2.81198949e-01 8.49686444e-01
8.29527438e-01 -1.23639870e+00 -9.87238765e-01 6.30728006e-02
6.86179459e-01 7.84484521e-02 1.77895993e-01 3.53458881e-01
-1.48822591e-01 5.31588137e-01 -2.02296183e-01 -7.25397050e-01
-1.12946999e+00 9.70312655e-01 1.37081742e-01 -4.79362644e-02
-4.37128305e-01 5.14821887e-01 4.72933561e-01 -2.59009242e-01
1.14202552e-01 2.09350169e-01 -3.01713735e-01 -2.20614269e-01
7.46194720e-01 -3.47826481e-01 -6.81139901e-02 -9.19616282e-01
-1.16540426e-02 1.37580916e-01 -1.52981132e-01 -3.88103187e-01
1.23097348e+00 1.44493729e-02 2.20541984e-01 4.58446026e-01
1.04595268e+00 -6.66292727e-01 -1.16725206e+00 -2.79670328e-01
2.19981149e-01 -3.65058213e-01 -4.35849518e-01 -7.54837990e-01
-6.61218405e-01 9.48126376e-01 6.64673507e-01 4.51911092e-02
8.03951383e-01 3.68809968e-01 6.74004853e-01 2.45796934e-01
3.45231563e-01 -8.67059171e-01 3.94858420e-01 2.92086035e-01
9.29301322e-01 -1.54745889e+00 -5.65881312e-01 -4.77630138e-01
-7.72730529e-01 7.91715503e-01 1.02342272e+00 -1.62906215e-01
6.42211363e-02 -1.42076835e-01 -1.82990823e-02 1.17848895e-01
-1.03113651e+00 -5.81915498e-01 1.99078396e-01 6.52337849e-01
9.98505652e-02 -2.61430949e-01 -4.78613228e-01 5.20581961e-01
2.59992540e-01 1.49853170e-01 8.97168458e-01 1.12258303e+00
-7.13927567e-01 -1.08643997e+00 -5.50199270e-01 3.05632263e-01
-1.29917681e-01 -1.62141562e-01 -4.64713514e-01 2.71693200e-01
1.34246454e-01 1.03114617e+00 2.09329948e-01 3.34032923e-02
1.64948225e-01 4.23552305e-01 3.54933590e-01 -1.15221989e+00
-2.61638850e-01 -1.94595531e-01 -6.82114810e-02 -1.09640069e-01
-4.09423977e-01 -5.50479412e-01 -8.56152833e-01 8.48862648e-01
-4.02679175e-01 3.25768799e-01 9.54817057e-01 9.82674062e-01
3.01384658e-01 3.31804752e-01 7.33429551e-01 -8.32013428e-01
-5.95280945e-01 -1.02920485e+00 -1.67186543e-01 6.63269818e-01
3.09645832e-01 -3.60654712e-01 -2.11921304e-01 6.11179888e-01] | [10.936067581176758, 1.0347152948379517] |
374cba95-d958-4497-90c2-bbb1d14a1d05 | functional-nanomaterials-design-in-the | 2108.13171 | null | https://arxiv.org/abs/2108.13171v1 | https://arxiv.org/pdf/2108.13171v1.pdf | Functional Nanomaterials Design in the Workflow of Building Machine-Learning Models | Machine-learning (ML) techniques have revolutionized a host of research fields of chemical and materials science with accelerated, high-efficiency discoveries in design, synthesis, manufacturing, characterization and application of novel functional materials, especially at the nanometre scale. The reason is the time efficiency, prediction accuracy and good generalization abilities, which gradually replaces the traditional experimental or computational work. With enormous potentiality to tackle more real-world problems, ML provides a more comprehensive insight into combinations with molecules/materials under the fundamental procedures for constructing ML models, like predicting properties or functionalities from given parameters, nanoarchitecture design and generating specific models for other purposes. The key to the advances in nanomaterials discovery is how input fingerprints and output values can be linked quantitatively. Finally, some great opportunities and technical challenges are concluded in this fantastic field. | ['Zhexu Xi'] | 2021-08-16 | null | null | null | null | ['design-synthesis'] | ['adversarial'] | [ 5.33878684e-01 -5.20517826e-01 -5.59945822e-01 -6.11285083e-02
-4.91139978e-01 -5.48006535e-01 4.79793847e-01 6.07309759e-01
-1.86457112e-01 1.28473961e+00 -1.77336738e-01 -3.34920138e-01
-2.31749237e-01 -1.10073090e+00 -5.30763745e-01 -1.18777144e+00
-6.24788590e-02 4.36220765e-01 1.99067160e-01 -1.88598216e-01
5.93212724e-01 8.21542680e-01 -1.69437850e+00 2.42470011e-01
9.16473925e-01 1.37353444e+00 4.72442180e-01 2.85917491e-01
-5.00253081e-01 2.47042805e-01 -2.91568780e-04 -3.58972132e-01
-1.70330852e-01 -4.85815071e-02 -2.75012761e-01 -3.32265586e-01
3.90180200e-02 4.64725614e-01 3.75539586e-02 1.00091314e+00
4.52123404e-01 -7.12525696e-02 1.07496870e+00 -7.08413482e-01
-8.76369894e-01 3.87465328e-01 -2.07992554e-01 -4.32971656e-01
5.32618344e-01 2.24505812e-01 7.88562894e-01 -9.18241382e-01
4.49748009e-01 1.10735965e+00 3.99258643e-01 5.78320742e-01
-1.23069513e+00 -8.71153414e-01 -2.95088083e-01 1.41186908e-01
-9.75245595e-01 -3.87803674e-01 6.03661180e-01 -7.67166138e-01
8.07325423e-01 3.40488642e-01 3.29803258e-01 8.70283127e-01
7.00031459e-01 2.57807881e-01 1.19006824e+00 -2.18738183e-01
3.17570865e-01 3.06578577e-01 -1.98516279e-01 5.56230485e-01
5.04286170e-01 2.23481178e-01 -1.87942550e-01 1.50661051e-01
6.60909712e-01 4.91121113e-01 1.60531059e-01 -2.46754423e-01
-1.15785944e+00 6.44966364e-01 4.24768418e-01 6.96423888e-01
-4.64520305e-01 2.62654442e-02 2.13535234e-01 2.67601609e-01
1.23938076e-01 1.18419003e+00 -8.13415170e-01 1.72560781e-01
-5.84861517e-01 8.21268559e-02 6.51904523e-01 6.88625216e-01
9.90540624e-01 1.98596194e-01 1.29135251e-01 5.07311881e-01
2.01027915e-02 5.70190132e-01 1.68783367e-01 -6.23171628e-01
1.62374154e-01 6.50743067e-01 2.57287472e-01 -4.92553622e-01
-7.50226259e-01 -3.90082717e-01 -8.13230038e-01 9.01192054e-02
1.23062603e-01 -2.35077157e-03 -6.53000891e-01 1.27866006e+00
3.05365890e-01 -7.15047777e-01 -5.04562110e-02 3.40325624e-01
9.14282143e-01 8.17130625e-01 4.63875055e-01 -4.01244581e-01
1.18379629e+00 -6.86387181e-01 -4.34020251e-01 1.25775442e-01
5.01051188e-01 -8.56135309e-01 8.89713407e-01 7.06950545e-01
-9.68591571e-01 -5.23104012e-01 -1.01726174e+00 1.13352060e-01
-9.28299785e-01 7.88456723e-02 1.35031676e+00 7.22005308e-01
-3.39602977e-01 1.19180822e+00 -4.74294007e-01 -2.87519485e-01
5.45320868e-01 8.88252676e-01 -4.75060821e-01 -1.95109755e-01
-9.44520891e-01 7.94645965e-01 4.72938210e-01 -2.69691616e-01
-7.18830705e-01 -1.02116609e+00 -4.14659232e-01 -2.15071529e-01
3.38663489e-01 -6.44123018e-01 6.88977063e-01 -4.10836965e-01
-1.69288385e+00 4.89403635e-01 -9.12798792e-02 -1.86684355e-01
1.39324799e-01 1.07386447e-01 -6.36032939e-01 -3.17790568e-01
-3.00108315e-03 6.26592278e-01 3.74478310e-01 -1.04877949e+00
-3.53252888e-01 -3.71133059e-01 -4.03460741e-01 -5.32066286e-01
-3.53598863e-01 -1.34181231e-01 4.84144211e-01 -3.12177658e-01
1.14581421e-01 -6.30459785e-01 -6.08478248e-01 -1.16223387e-01
-7.21073151e-01 -3.17639470e-01 5.06560981e-01 -1.85832456e-01
8.61038089e-01 -1.70458293e+00 2.57005900e-01 3.53231728e-01
3.64882052e-01 2.79501557e-01 -5.27954400e-02 1.21882570e+00
-1.42253995e-01 2.72331476e-01 -3.26206572e-02 4.78474230e-01
-2.90515006e-01 -4.51007962e-01 -4.27844897e-02 2.40337685e-01
1.93489075e-01 1.07966876e+00 -1.09693372e+00 1.62091300e-01
5.31905115e-01 3.46956909e-01 -2.99786776e-01 -3.34717594e-02
-8.24968815e-01 9.69023585e-01 -8.62352908e-01 1.05716646e+00
4.10765380e-01 -3.44577193e-01 4.42784689e-02 -4.02672172e-01
-7.09978223e-01 2.52866149e-01 -6.09719753e-01 1.32832551e+00
-4.45077688e-01 2.63697386e-01 -1.41904593e-01 -8.66198719e-01
1.17450833e+00 1.90199465e-01 9.15357232e-01 -1.00589514e+00
3.02540570e-01 4.41450655e-01 5.78685068e-02 -8.01170409e-01
1.10468892e-02 -6.18671536e-01 1.40803605e-01 1.19408771e-01
-1.92608610e-01 9.86947306e-03 2.50586450e-01 -5.22009671e-01
6.62906468e-01 -2.00194597e-01 3.96037996e-01 -4.11359489e-01
6.95003510e-01 -4.83250208e-02 1.00199562e-02 3.91049951e-01
2.44419754e-01 3.48100960e-02 1.01925820e-01 -4.61341947e-01
-1.22928512e+00 -1.08459556e+00 -2.86495388e-01 1.16164362e+00
-2.60525127e-03 -3.27610016e-01 -4.40204471e-01 -5.98340295e-02
2.39586785e-01 2.40160912e-01 -5.97375929e-01 -1.99104100e-01
-2.95592338e-01 -8.62794995e-01 2.03828700e-02 3.76752436e-01
1.38325304e-01 -9.03059423e-01 1.96078017e-01 6.81825399e-01
6.85998201e-01 -1.12635124e+00 8.59588012e-02 3.64910156e-01
-9.21031475e-01 -1.14449120e+00 -4.26931500e-01 -8.56344461e-01
6.26069367e-01 2.57253170e-01 6.84657097e-01 -2.25123331e-01
-8.01976383e-01 -2.21569583e-01 -3.50300856e-02 -7.64506340e-01
-7.30763137e-01 8.45621899e-02 2.34115824e-01 -2.12104425e-01
2.00478900e-02 -9.71397519e-01 -8.92039061e-01 1.95007369e-01
-6.69469774e-01 -1.42269894e-01 1.23764801e+00 5.37768364e-01
8.60111058e-01 2.62311459e-01 1.24195945e+00 -1.02030265e+00
4.38820124e-01 -3.60430062e-01 -5.71118772e-01 3.75137061e-01
-7.93983877e-01 4.51225013e-01 1.09098363e+00 -5.66229641e-01
-5.73733032e-01 7.29259849e-02 -5.54980636e-01 3.37997377e-01
-1.85055837e-01 4.76495147e-01 -7.94326663e-01 -2.73536682e-01
6.54573798e-01 2.44775698e-01 2.86458973e-02 -5.33269525e-01
3.14536184e-01 3.79096121e-01 1.00289032e-01 -8.88199508e-01
7.37848997e-01 2.91751027e-01 6.42346203e-01 -1.15308893e+00
-6.42427087e-01 -3.53105515e-01 -6.62615180e-01 3.68328206e-02
8.71583700e-01 -4.80293810e-01 -1.15071833e+00 6.79833665e-02
-7.33393431e-01 1.33415964e-02 -8.76852050e-02 3.53232592e-01
-1.90875918e-01 6.89387098e-02 -2.43044958e-01 -8.58364344e-01
-4.62974638e-01 -1.31822836e+00 8.18469226e-01 2.92224735e-01
-2.09158435e-01 -1.15572679e+00 -2.08901688e-01 2.70643413e-01
5.84945083e-01 6.00320220e-01 1.63309574e+00 -4.59157169e-01
-7.47094095e-01 -5.41806281e-01 -1.99260756e-01 7.51436595e-03
7.64411986e-01 1.09626643e-01 -9.14672852e-01 -1.26357391e-01
-3.46136004e-01 -2.86033303e-01 9.17413056e-01 4.06084716e-01
1.47335947e+00 -2.35487252e-01 -7.02818632e-01 2.69566417e-01
1.53649175e+00 7.33044744e-01 5.91498375e-01 1.71737950e-02
8.40247631e-01 7.21998572e-01 3.42360318e-01 2.05630705e-01
-2.52146840e-01 2.62391448e-01 3.98888230e-01 -7.86202401e-02
-6.09156713e-02 -3.91115934e-01 1.82820097e-01 6.58071935e-01
-2.39689633e-01 -7.55396932e-02 -5.56896985e-01 -1.51864126e-01
-1.29775453e+00 -1.15708792e+00 -2.33752564e-01 2.54341865e+00
8.52938652e-01 2.30394766e-01 3.10207695e-01 2.43424490e-01
6.28089130e-01 -2.15989143e-01 -8.73749316e-01 -2.51508027e-01
-1.63225174e-01 4.71276939e-01 5.73386490e-01 1.61363706e-01
-9.40366566e-01 8.58877063e-01 7.43904543e+00 9.73822951e-01
-1.63853633e+00 -2.84011245e-01 6.04596317e-01 3.52893054e-01
-5.21537840e-01 -2.69864529e-01 -9.43085074e-01 2.97685057e-01
8.06025147e-01 -1.23315454e-01 4.36762005e-01 5.49900055e-01
3.65677327e-01 2.59218439e-02 -1.35225391e+00 7.31954038e-01
-7.07346499e-01 -2.15688443e+00 2.94114351e-01 1.99877918e-01
9.66019869e-01 -1.89663440e-01 3.35088938e-01 -9.89560112e-02
-3.57531011e-01 -1.17375839e+00 4.12584603e-01 6.30281270e-01
1.12041438e+00 -6.62322044e-01 2.62452126e-01 1.76675826e-01
-1.04954410e+00 -3.17328274e-01 -4.35709506e-01 -1.63186029e-01
-3.96248966e-01 1.00836480e+00 -9.98391151e-01 5.21214306e-01
6.64906725e-02 7.55124032e-01 -2.75737107e-01 1.14483547e+00
1.24573529e-01 2.10061073e-01 -1.35174654e-02 -7.30682313e-01
3.70359756e-02 -5.56739032e-01 1.14213496e-01 1.07989562e+00
2.61289626e-01 -2.03014001e-01 1.36170626e-01 9.07507420e-01
-1.98873639e-01 2.20883474e-01 -5.81028759e-01 -6.73551381e-01
6.44633830e-01 1.16747832e+00 -9.64093149e-01 1.20573491e-01
-3.93211991e-01 1.38377100e-01 -2.45742202e-01 1.59902871e-01
-5.15669405e-01 -3.57354075e-01 6.21053338e-01 5.46484828e-01
1.49072111e-01 -5.07666469e-01 -3.91524643e-01 -5.13182104e-01
-4.39305872e-01 -5.71775258e-01 -1.69489786e-01 -1.25931054e-01
-1.39740086e+00 1.01845905e-01 -2.42807806e-01 -9.37158644e-01
2.88594633e-01 -1.24339008e+00 -7.21501231e-01 3.73149633e-01
-1.39719594e+00 -9.54750299e-01 1.71928227e-01 -4.06493386e-03
3.78351092e-01 -6.75813630e-02 9.60180700e-01 3.68902892e-01
-6.39283478e-01 2.58553296e-01 8.10977399e-01 -4.16484594e-01
7.51919389e-01 -1.04067445e+00 1.25710994e-01 2.52779108e-02
-2.23419070e-01 5.81866264e-01 8.31385076e-01 -4.35008258e-01
-1.99779391e+00 -1.18440330e+00 5.06118894e-01 -3.43634576e-01
1.00966227e+00 -5.97732186e-01 -5.16919374e-01 -1.34691820e-01
-4.02762324e-01 -3.68376613e-01 1.08325446e+00 7.76815712e-02
-2.61180401e-01 -3.68784636e-01 -1.16026521e+00 5.98334014e-01
8.36181819e-01 -4.37823117e-01 9.64638516e-02 7.56390214e-01
7.80033290e-01 5.70307747e-02 -1.40630424e+00 4.38019544e-01
1.03915381e+00 -6.80884838e-01 1.23292029e+00 -8.02897573e-01
5.03284276e-01 -1.51440814e-01 -2.57219404e-01 -7.30977595e-01
-4.17187393e-01 -7.27499008e-01 1.12301879e-01 9.52140570e-01
9.55933869e-01 -7.27868736e-01 8.77830923e-01 5.25395572e-01
-2.79316485e-01 -1.27077055e+00 -6.30611598e-01 -8.68780255e-01
2.92038739e-01 -1.81191981e-01 8.25893998e-01 5.81825256e-01
4.43979464e-02 6.00182235e-01 -1.89635515e-01 -2.98454314e-01
5.23479700e-01 4.08144563e-01 2.81580746e-01 -1.68874717e+00
-8.35528299e-02 -7.30728030e-01 -4.47663963e-01 -6.66315615e-01
-4.17989269e-02 -7.91222870e-01 -3.23913336e-01 -1.58452892e+00
3.06273215e-02 -5.98356783e-01 -5.54862320e-01 3.43323164e-02
1.80404067e-01 1.36860654e-01 -1.31920546e-01 -3.80462804e-03
-5.36909163e-01 4.05630589e-01 1.74493659e+00 -5.06049991e-01
-3.96854967e-01 4.07139540e-01 -1.12122452e+00 2.87421733e-01
8.94867122e-01 -3.07734728e-01 -2.97360688e-01 1.17031716e-01
5.98152280e-01 -1.66601222e-02 -1.27080113e-01 -1.16934478e+00
6.88120723e-02 -5.83948433e-01 6.67253613e-01 -3.53491426e-01
3.01752865e-01 -7.64119029e-01 2.75896370e-01 6.26916766e-01
-2.71470904e-01 -2.37952411e-01 -3.05692330e-02 4.46803302e-01
1.89932868e-01 -1.82186648e-01 8.11331630e-01 -1.89678758e-01
-4.67943698e-01 7.85361648e-01 -4.07492995e-01 -5.43226004e-01
1.20403934e+00 -5.27912080e-01 -3.25592488e-01 1.26280144e-01
-8.16060066e-01 -5.29420525e-02 5.54183602e-01 1.08212031e-01
4.98457581e-01 -1.17586076e+00 -2.59721071e-01 1.09330364e-01
1.92671269e-01 -2.28938758e-01 9.68964994e-02 7.60271430e-01
-2.16010526e-01 7.64370739e-01 -2.94307590e-01 -3.59061301e-01
-8.05852950e-01 8.06812942e-01 1.77012742e-01 -1.35670006e-01
1.08230419e-01 5.86655140e-01 3.52052003e-01 -1.03222840e-01
-6.00696448e-03 -3.19840819e-01 -1.32814914e-01 -5.76025099e-02
5.16652465e-01 6.52527153e-01 3.05359721e-01 -9.92132276e-02
-2.80258596e-01 7.07205951e-01 -9.24679264e-02 6.07943594e-01
1.71767771e+00 3.99342209e-01 -2.56783128e-01 5.75438380e-01
1.18713129e+00 1.05669744e-01 -1.10538876e+00 2.29615971e-01
3.16264272e-01 1.20925449e-01 -1.10853367e-01 -8.31352651e-01
-5.94998956e-01 7.93166399e-01 5.04607439e-01 3.65487933e-01
6.75570667e-01 2.43058398e-01 8.48168731e-01 7.96214402e-01
5.15876830e-01 -9.83485222e-01 2.67353922e-01 2.18299448e-01
8.04451287e-01 -1.17284787e+00 3.84559631e-01 -5.86251080e-01
-3.43317911e-03 1.40939355e+00 2.48666823e-01 1.56490117e-01
7.97115684e-01 2.47307882e-01 -4.03867632e-01 -1.46095529e-01
-6.60330236e-01 -4.71120775e-02 2.07826987e-01 1.00143707e+00
1.00322616e+00 2.76347280e-01 -2.28063300e-01 5.91242611e-01
-2.08357861e-03 -4.11479056e-01 9.89035219e-02 8.82847965e-01
-1.02213585e+00 -1.84030426e+00 -1.23638339e-01 7.85397828e-01
-5.41992307e-01 -9.73267257e-02 -5.38655698e-01 7.03224659e-01
3.15697193e-01 8.89962554e-01 -4.14957106e-01 -6.85441136e-01
2.41074845e-01 -8.23114067e-02 9.09377575e-01 -5.61330914e-01
-3.75126332e-01 2.73176748e-02 2.14235969e-02 -2.61304766e-01
-3.15358520e-01 -5.31418920e-01 -1.49150765e+00 -2.48499930e-01
-5.92459977e-01 5.66918075e-01 1.13794148e+00 1.08875000e+00
4.84711826e-01 3.81336540e-01 1.13496268e+00 -1.05824351e+00
-1.07661746e-01 -5.17680168e-01 -6.09830379e-01 -1.09661892e-01
2.22116619e-01 -6.51910365e-01 1.88163090e-02 -1.74334928e-01] | [5.1822381019592285, 5.468028545379639] |
e49d2903-3d9b-44ae-aa24-56c846d01cb4 | nlpositionality-characterizing-design-biases | 2306.01943 | null | https://arxiv.org/abs/2306.01943v1 | https://arxiv.org/pdf/2306.01943v1.pdf | NLPositionality: Characterizing Design Biases of Datasets and Models | Design biases in NLP systems, such as performance differences for different populations, often stem from their creator's positionality, i.e., views and lived experiences shaped by identity and background. Despite the prevalence and risks of design biases, they are hard to quantify because researcher, system, and dataset positionality is often unobserved. We introduce NLPositionality, a framework for characterizing design biases and quantifying the positionality of NLP datasets and models. Our framework continuously collects annotations from a diverse pool of volunteer participants on LabintheWild, and statistically quantifies alignment with dataset labels and model predictions. We apply NLPositionality to existing datasets and models for two tasks -- social acceptability and hate speech detection. To date, we have collected 16,299 annotations in over a year for 600 instances from 1,096 annotators across 87 countries. We find that datasets and models align predominantly with Western, White, college-educated, and younger populations. Additionally, certain groups, such as non-binary people and non-native English speakers, are further marginalized by datasets and models as they rank least in alignment across all tasks. Finally, we draw from prior literature to discuss how researchers can examine their own positionality and that of their datasets and models, opening the door for more inclusive NLP systems. | ['Maarten Sap', 'Katharina Reinecke', 'Ronan Le Bras', 'Jenny T. Liang', 'Sebastin Santy'] | 2023-06-02 | null | null | null | null | ['hate-speech-detection'] | ['natural-language-processing'] | [-7.68992901e-02 3.36371183e-01 -3.96744847e-01 -5.77161491e-01
-5.52444518e-01 -1.07053864e+00 6.16138458e-01 2.55565524e-01
-5.40234745e-01 4.67005759e-01 1.08001685e+00 -2.73974866e-01
9.42897424e-03 -3.53833064e-02 -5.39133132e-01 -1.84570059e-01
6.14870965e-01 2.06949994e-01 -6.36656821e-01 2.78092802e-01
3.14185917e-01 1.86154515e-01 -1.00640869e+00 2.55843014e-01
1.09689116e+00 2.99381793e-01 -3.34130436e-01 3.17486376e-01
1.33090898e-01 8.60957205e-01 -6.76748455e-01 -8.43480945e-01
4.68341522e-02 -1.01502150e-01 -7.07714379e-01 -3.68923634e-01
1.13563478e+00 -5.96445687e-02 -1.92974955e-01 8.46078336e-01
9.72290933e-01 -3.13834488e-01 7.99179018e-01 -1.20615625e+00
-1.65060389e+00 1.10240579e+00 -5.30185103e-01 7.73153678e-02
5.40146172e-01 6.66337192e-01 1.11397505e+00 -8.15775275e-01
1.08174527e+00 1.49547720e+00 1.24983001e+00 5.46798170e-01
-1.77170455e+00 -1.08758795e+00 1.78794861e-01 -5.07196426e-01
-1.28958082e+00 -9.28638279e-01 4.79905844e-01 -1.25468123e+00
6.49664521e-01 2.48274535e-01 5.64662397e-01 1.84517574e+00
-1.44583449e-01 5.08247972e-01 1.52077746e+00 -1.64410830e-01
-1.22690080e-02 5.02849340e-01 5.68263233e-01 3.75022143e-01
6.25556290e-01 -7.87600949e-02 -8.87255013e-01 -8.25329840e-01
9.45571810e-02 -3.73515964e-01 -3.10503721e-01 1.02643892e-01
-1.30196452e+00 5.97983062e-01 1.37839764e-01 3.49473417e-01
-9.46807414e-02 -8.36596414e-02 2.63043582e-01 -8.50476231e-03
6.70166135e-01 1.03277183e+00 -6.11277521e-01 -3.95861894e-01
-8.29191387e-01 5.75683653e-01 1.15353799e+00 9.17518556e-01
4.46581185e-01 -2.25227177e-01 -3.02212268e-01 9.77686942e-01
2.01630920e-01 8.06345940e-01 4.31574851e-01 -1.15771306e+00
6.45847440e-01 4.98853296e-01 4.85898376e-01 -1.26297534e+00
-8.64158809e-01 -2.84601897e-01 -2.87511647e-01 -1.48042917e-01
7.17685342e-01 -6.21250808e-01 -5.32561898e-01 2.11156082e+00
2.97462735e-02 -5.58659434e-01 -2.52612919e-01 8.30358386e-01
7.77796626e-01 2.68768877e-01 6.22259915e-01 -1.06707662e-01
1.41334879e+00 -3.98593456e-01 -5.34256577e-01 -6.82227671e-01
8.60830367e-01 -9.48371351e-01 1.54085839e+00 2.87184238e-01
-9.05590057e-01 -4.43519264e-01 -4.88490880e-01 -4.06470865e-01
-2.56096810e-01 -1.47736340e-03 3.98743182e-01 1.17233288e+00
-8.85432541e-01 4.16117132e-01 -1.95656762e-01 -7.40913630e-01
6.74841404e-01 2.84358338e-02 -3.55650783e-01 2.66248405e-01
-8.10421050e-01 1.00878584e+00 -1.37288541e-01 -4.21080031e-02
-4.21868414e-01 -1.20334601e+00 -6.66376233e-01 -4.31155324e-01
-9.16410685e-02 -3.75554502e-01 1.12342000e+00 -1.02462459e+00
-8.00305426e-01 1.13334513e+00 -2.02963650e-01 8.58367980e-02
5.04249692e-01 -5.15164375e-01 -5.49583733e-01 -7.25837290e-01
3.77152979e-01 6.23260260e-01 2.98026174e-01 -1.24226725e+00
-3.24337006e-01 -3.93508762e-01 -1.47278354e-01 7.75301456e-02
-5.84225416e-01 5.91029763e-01 3.53994459e-01 -6.62322760e-01
-1.56998217e-01 -1.04753196e+00 1.45137519e-01 -1.32306769e-01
-5.82555890e-01 -8.64880681e-02 4.18603450e-01 -1.15049148e+00
1.55435777e+00 -2.32746410e+00 -7.55499825e-02 6.07909523e-02
5.59494078e-01 -6.26936629e-02 -2.15750132e-02 3.14390212e-01
-4.43350673e-02 1.04484177e+00 7.91398063e-02 -5.04096985e-01
4.62197572e-01 -2.74560034e-01 -1.90392077e-01 7.10681796e-01
3.10058985e-02 7.45724440e-01 -7.73516476e-01 -5.52585840e-01
-2.05345273e-01 2.99619853e-01 -6.89287066e-01 -1.25333726e-01
6.78842664e-02 4.89429146e-01 3.81024294e-02 7.16597795e-01
5.17577112e-01 3.55420709e-02 2.77282089e-01 -7.27298036e-02
-7.06931949e-01 4.84581262e-01 -5.43965161e-01 1.32890296e+00
-3.07736933e-01 1.20084870e+00 3.34355026e-01 2.16042064e-02
8.86509240e-01 2.17753515e-01 -1.22202925e-01 -6.44695580e-01
1.22515790e-01 4.61564869e-01 3.97871107e-01 -8.02728832e-01
8.45247507e-01 1.66794844e-02 -3.99023712e-01 6.50381029e-01
-3.70882124e-01 4.14426846e-04 -1.99495211e-01 -1.64365813e-01
9.59607184e-01 5.41926734e-02 7.74464384e-02 -5.88315129e-01
-3.44574332e-01 1.55421374e-02 1.01864493e+00 8.95524621e-01
-7.71955311e-01 5.93822300e-01 8.21432054e-01 -3.62252802e-01
-1.44698727e+00 -8.85453105e-01 -4.10874605e-01 1.52253163e+00
-4.47941124e-01 -1.53583542e-01 -6.51463628e-01 -5.66466331e-01
2.63736725e-01 8.27336371e-01 -7.41418183e-01 -5.82020245e-02
-4.25045311e-01 -6.57364607e-01 9.80716646e-01 3.21048111e-01
-3.86211872e-02 -7.75830209e-01 -2.97346145e-01 -1.08153209e-01
-4.63319749e-01 -1.14099216e+00 -4.71919715e-01 -3.40451479e-01
-4.09542352e-01 -8.39169800e-01 -4.67833102e-01 -6.33013904e-01
3.82486969e-01 -2.07862839e-01 1.25606358e+00 -3.22908685e-02
4.20298353e-02 2.01415867e-01 -1.78253233e-01 -8.10692668e-01
-4.32046354e-01 3.23042065e-01 3.09288174e-01 -3.87776107e-01
7.77397454e-01 -3.59142631e-01 -1.69851243e-01 2.94713616e-01
-2.05531716e-01 -1.44388571e-01 3.56242776e-01 3.53026718e-01
-1.23619497e-01 -5.85281909e-01 3.87115180e-01 -1.25739896e+00
6.46220922e-01 -7.97749758e-01 -9.12793800e-02 2.66929358e-01
-5.70008278e-01 -4.14330631e-01 2.95741677e-01 -7.15425432e-01
-9.35133517e-01 -3.63123655e-01 4.01601493e-01 1.86024636e-01
-3.73147011e-01 2.47934297e-01 -1.17590733e-01 1.10247068e-01
1.15960681e+00 -7.18912661e-01 -1.04190752e-01 -5.48264265e-01
3.30181271e-01 1.17261696e+00 3.09956014e-01 -9.13906336e-01
6.95615828e-01 2.01116711e-01 -9.83404398e-01 -9.68642354e-01
-8.71436477e-01 -3.05333789e-02 -3.85700583e-01 -1.64150372e-01
9.10033107e-01 -1.13388348e+00 -9.29056704e-01 4.43948984e-01
-1.21171296e+00 -6.31660461e-01 4.65504304e-02 6.28599226e-01
3.82138528e-02 6.57076528e-03 -5.53159356e-01 -1.05994892e+00
-2.45562524e-01 -9.49196577e-01 7.25806594e-01 2.65257895e-01
-1.38647068e+00 -1.03295207e+00 2.17322156e-01 7.22963870e-01
2.84049064e-01 4.97580379e-01 1.02513874e+00 -9.31848288e-01
1.47235885e-01 7.48252198e-02 -4.69447672e-01 9.76330042e-02
-1.22646146e-01 3.97891641e-01 -1.32254732e+00 -6.76673371e-03
-1.42716676e-01 -4.45806593e-01 2.77525127e-01 3.12233478e-01
9.15451705e-01 -4.48571950e-01 -4.36834455e-01 2.24924758e-01
1.07521808e+00 -5.28889261e-02 1.12404138e-01 1.72296852e-01
1.07317150e+00 1.05682755e+00 2.05378816e-01 4.22879606e-01
5.77676058e-01 3.25413406e-01 -2.84365892e-01 7.55961984e-02
1.51986763e-01 -5.75879812e-01 6.26303136e-01 5.01843631e-01
1.68272331e-02 -1.60654977e-01 -1.59038401e+00 7.23925769e-01
-1.62647402e+00 -9.48717415e-01 -3.24642956e-01 1.93907595e+00
9.23965156e-01 -6.77272528e-02 3.13104063e-01 -4.88733232e-01
1.01492894e+00 4.74156171e-01 -4.82948571e-01 -6.40569746e-01
-4.16537136e-01 -2.41704419e-01 6.10938072e-01 4.81746018e-01
-8.09942901e-01 7.39983499e-01 7.42189693e+00 4.66685027e-01
-9.53485489e-01 3.70026052e-01 9.05338824e-01 -4.89542067e-01
-7.42072999e-01 4.91560921e-02 -7.69181609e-01 7.25410700e-01
1.09751487e+00 -1.28754407e-01 4.56618845e-01 6.43842161e-01
3.95691156e-01 1.14708379e-01 -1.37089896e+00 8.42434406e-01
2.31445283e-01 -8.61076415e-01 -6.04192197e-01 3.47108454e-01
1.01039875e+00 1.73185676e-01 4.10611838e-01 2.28853360e-01
8.16480041e-01 -1.31961679e+00 1.50603819e+00 4.13520098e-01
6.56169116e-01 -2.68041253e-01 4.86989915e-01 1.73830643e-01
-3.16668600e-01 -6.51404977e-01 -3.00471336e-01 -5.83617091e-01
8.03650916e-02 6.86654270e-01 -4.81565893e-01 -3.04373771e-01
9.43875790e-01 5.02121687e-01 -5.96948087e-01 5.49971998e-01
-9.22259092e-02 8.11159670e-01 -2.34614804e-01 -1.55064622e-02
-2.43938938e-02 -5.05202264e-03 4.11345094e-01 1.38197947e+00
2.15010166e-01 3.72200869e-02 -1.66437984e-01 1.18972206e+00
-2.30845422e-01 6.64846301e-02 -9.20947254e-01 -6.87508583e-01
1.01679289e+00 1.31928408e+00 -2.27009460e-01 1.03798591e-01
-6.25686347e-01 4.51966733e-01 5.93072355e-01 5.78436136e-01
-9.54558432e-01 -4.36722487e-02 9.50874269e-01 2.03058034e-01
-5.83269238e-01 -4.10253853e-01 -8.75306606e-01 -8.56325328e-01
1.81018226e-02 -1.24883890e+00 9.18629467e-02 -8.75604928e-01
-1.66143370e+00 1.21815585e-01 -2.82999873e-01 -5.98268449e-01
3.65666479e-01 -5.95507622e-01 -1.47042766e-01 8.94283116e-01
-8.16753805e-01 -1.27823842e+00 -2.62872815e-01 -4.99965511e-02
-7.63719827e-02 -1.43235669e-01 6.15971267e-01 2.86404669e-01
-9.57856297e-01 7.20563710e-01 -1.37225568e-01 4.16406840e-01
1.43208003e+00 -1.13635957e+00 6.91204190e-01 4.79650497e-01
-9.01855156e-02 1.00967574e+00 7.32420266e-01 -8.61976206e-01
-8.58764410e-01 -6.12803161e-01 1.41331828e+00 -1.29359400e+00
9.27601159e-01 -7.13936090e-01 -6.62017107e-01 9.93681610e-01
2.68751562e-01 -4.48846996e-01 1.29211080e+00 9.78200853e-01
-8.65726829e-01 2.46571004e-01 -1.23524058e+00 7.84815013e-01
1.58589888e+00 -9.95924592e-01 -7.27666616e-01 2.98223138e-01
8.74423921e-01 -4.48821634e-01 -1.04026556e+00 -3.24602872e-02
1.05649245e+00 -9.08046544e-01 4.83637959e-01 -6.67363524e-01
5.57324409e-01 1.79163858e-01 -1.49636537e-01 -1.18765867e+00
-5.96369326e-01 -8.57850730e-01 4.30495858e-01 1.91213906e+00
8.71242583e-01 -7.27454901e-01 4.94163424e-01 1.69566107e+00
-5.42011186e-02 -3.83325756e-01 -4.92618293e-01 -5.10343969e-01
5.95404208e-01 -4.08367485e-01 8.27267289e-01 1.63037932e+00
2.64183916e-02 2.57573724e-01 -1.91923425e-01 1.22467853e-01
4.24411267e-01 -3.52610409e-01 8.04004550e-01 -1.46317077e+00
-2.12101769e-02 -5.74090660e-01 4.06709723e-02 -3.34368527e-01
6.18631542e-01 -6.95318699e-01 -5.02871163e-02 -1.18959641e+00
5.11032820e-01 -9.11037982e-01 2.26658508e-01 6.06393635e-01
-1.03819549e-01 1.30896583e-01 3.53212208e-01 3.34236711e-01
-1.82449013e-01 -9.26477686e-02 8.48487258e-01 1.03353336e-01
-4.63705957e-01 -6.28162563e-01 -1.35270274e+00 9.66562450e-01
7.16797054e-01 -5.26372075e-01 -1.10760838e-01 -1.06503654e+00
8.18843365e-01 -6.88132584e-01 5.21696985e-01 -9.01139021e-01
-4.07290868e-02 -3.67686540e-01 5.76152980e-01 1.85657337e-01
7.85349086e-02 -6.57474399e-01 3.51109028e-01 1.62819669e-01
-6.54997170e-01 1.73570409e-01 6.67189211e-02 5.64717799e-02
3.54425550e-01 -1.04076684e-01 4.32763994e-01 1.44321203e-01
-5.58065623e-02 -4.94928360e-02 -2.76176721e-01 7.44353056e-01
6.62408471e-01 -2.73284167e-01 -1.10963678e+00 -7.92810991e-02
-5.49994886e-01 2.10022882e-01 1.00694013e+00 5.46551585e-01
-1.61819831e-01 -1.27615643e+00 -7.76651561e-01 -1.94164261e-01
3.10883909e-01 -3.91387582e-01 9.21734273e-02 9.64313388e-01
-3.45715821e-01 1.12203173e-01 -3.81189771e-02 -6.36624768e-02
-1.04552555e+00 1.84424177e-01 1.28667459e-01 4.30469573e-01
-9.21256915e-02 8.33376706e-01 2.93496639e-01 -7.17981398e-01
-2.47628298e-02 -3.11726332e-01 1.38565600e-01 5.88676572e-01
2.73016423e-01 7.55193472e-01 -6.91469371e-01 -9.51216102e-01
-3.78792971e-01 4.50835049e-01 4.61458862e-02 -3.86131942e-01
1.10597348e+00 -1.49335504e-01 -1.76351920e-01 7.66561329e-01
1.06609595e+00 5.26175499e-01 -9.20838177e-01 -5.25788441e-02
7.61582628e-02 -5.36797941e-01 -4.76257682e-01 -9.00100112e-01
-4.24140066e-01 6.23917103e-01 2.28096053e-01 2.58615851e-01
2.47771874e-01 6.52527064e-02 6.01968169e-01 2.79865554e-03
1.51632309e-01 -1.68114507e+00 -1.10120468e-01 4.35754269e-01
9.71668482e-01 -1.01973987e+00 6.18372634e-02 -2.37573281e-01
-8.96075070e-01 4.85471517e-01 8.41681182e-01 3.69890869e-01
5.53211808e-01 1.13083422e-01 2.77610242e-01 -1.14047699e-01
-5.26600182e-01 1.56948000e-01 -1.06250606e-02 8.39956343e-01
1.11642730e+00 5.28417706e-01 -5.30273080e-01 1.05475223e+00
-8.37622225e-01 -2.54324406e-01 4.55200702e-01 5.42812467e-01
-1.89283758e-01 -7.04051495e-01 -5.83595991e-01 2.90904462e-01
-6.56734526e-01 -1.70797765e-01 -1.06937158e+00 8.75518382e-01
6.44051433e-01 1.05478847e+00 3.72300059e-01 -4.92016405e-01
3.69333029e-01 3.95793259e-01 1.55769780e-01 -6.83625937e-01
-1.13294756e+00 -5.15115559e-01 7.54486084e-01 -3.99026364e-01
-2.36100793e-01 -1.17632341e+00 -7.70131528e-01 -8.49849641e-01
-2.11909965e-01 -2.42294058e-01 5.20481527e-01 7.52467513e-01
5.59357941e-01 -2.38160059e-01 3.30841184e-01 -4.42624837e-01
-2.70960242e-01 -1.08674777e+00 -4.18437243e-01 5.40533125e-01
2.49246895e-01 -4.01876211e-01 -5.21725178e-01 -1.88107491e-01] | [9.245162010192871, 10.010473251342773] |
7a1d52ef-3de9-434d-83e3-1d78deb581b0 | conditional-diffusion-feature-refinement-for | 2305.03614 | null | https://arxiv.org/abs/2305.03614v2 | https://arxiv.org/pdf/2305.03614v2.pdf | Conditional Diffusion Feature Refinement for Continuous Sign Language Recognition | In this work, we are dedicated to leveraging the denoising diffusion models' success and formulating feature refinement as the autoencoder-formed diffusion process, which is a mask-and-predict scheme. The state-of-the-art CSLR framework consists of a spatial module, a visual module, a sequence module, and a sequence learning function. However, this framework has faced sequence module overfitting caused by the objective function and small-scale available benchmarks, resulting in insufficient model training. To overcome the overfitting problem, some CSLR studies enforce the sequence module to learn more visual temporal information or be guided by more informative supervision to refine its representations. In this work, we propose a novel autoencoder-formed conditional diffusion feature refinement~(ACDR) to refine the sequence representations to equip desired properties by learning the encoding-decoding optimization process in an end-to-end way. Specifically, for the ACDR, a noising Encoder is proposed to progressively add noise equipped with semantic conditions to the sequence representations. And a denoising Decoder is proposed to progressively denoise the noisy sequence representations with semantic conditions. Therefore, the sequence representations can be imbued with the semantics of provided semantic conditions. Further, a semantic constraint is employed to prevent the denoised sequence representations from semantic corruption. Extensive experiments are conducted to validate the effectiveness of our ACDR, benefiting state-of-the-art methods and achieving a notable gain on three benchmarks. | ['ShengYong Chen', 'Tiantian Yuan', 'Yuxi Zhou', 'Qing Guo', 'Wanli Xue', 'Leming Guo'] | 2023-05-05 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 6.33511007e-01 1.75869502e-02 4.41925488e-02 -3.73723030e-01
-4.81228530e-01 -1.93426609e-02 7.27441847e-01 -3.80380958e-01
-3.25214595e-01 4.14172620e-01 2.87443727e-01 -5.87741286e-02
7.33374357e-02 -6.87220871e-01 -7.75352836e-01 -1.04575586e+00
2.79181480e-01 7.31712505e-02 1.19547307e-01 -1.76772520e-01
1.78758278e-01 1.23589180e-01 -1.64136481e+00 5.36451995e-01
1.08763945e+00 1.18634224e+00 8.29376400e-01 4.74206448e-01
-4.12727296e-01 9.49735761e-01 -4.48957860e-01 -1.19665705e-01
5.20897545e-02 -7.39051282e-01 -4.91916358e-01 5.42304635e-01
-1.58600226e-01 -2.04777464e-01 -3.62223595e-01 1.10838580e+00
4.35378134e-01 2.71044046e-01 5.55884540e-01 -9.78230596e-01
-1.13821828e+00 4.53750879e-01 -5.71198881e-01 1.74104616e-01
2.60248005e-01 3.50231051e-01 8.64172041e-01 -9.11334515e-01
6.43599987e-01 1.30074573e+00 3.68032217e-01 7.61461973e-01
-1.22853112e+00 -5.15045345e-01 5.38663805e-01 5.02590179e-01
-1.32653105e+00 -4.85954314e-01 1.06130636e+00 -3.44037861e-01
1.01838338e+00 1.87915042e-01 6.76023662e-01 1.39802349e+00
9.58782509e-02 9.36812460e-01 9.36117887e-01 -3.70819420e-01
4.89554822e-01 -5.15126288e-02 -2.32028320e-01 7.02731133e-01
-4.02092874e-01 2.96577424e-01 -4.63825196e-01 2.76341468e-01
8.49504709e-01 2.31612533e-01 -3.04872483e-01 -1.99408472e-01
-1.01915097e+00 7.05680788e-01 6.80256784e-01 4.53503728e-01
-6.10256493e-01 2.41887271e-02 2.67902106e-01 2.13745743e-01
6.98112369e-01 -2.01406423e-02 -2.76170939e-01 1.71181411e-02
-9.78613913e-01 -1.63627177e-01 3.18771660e-01 8.18376362e-01
7.63326645e-01 3.77347797e-01 -3.33805472e-01 8.67208838e-01
5.78369737e-01 1.62075177e-01 7.76470244e-01 -8.20312798e-01
3.46898854e-01 6.12366557e-01 -2.91790634e-01 -8.41064453e-01
4.97032441e-02 -6.06103539e-01 -1.25270355e+00 1.38285652e-01
-2.57102311e-01 1.72039047e-01 -1.36093640e+00 1.70570731e+00
2.48954177e-01 7.57084250e-01 1.07396074e-01 1.07093668e+00
4.40387577e-01 9.36242044e-01 1.37952566e-01 -4.28359151e-01
1.07431114e+00 -1.12687874e+00 -8.98938894e-01 -2.26960570e-01
5.63032091e-01 -5.67776561e-01 9.78401005e-01 4.25582170e-01
-1.03244209e+00 -9.89814758e-01 -1.18714678e+00 6.07010908e-02
-4.50071432e-02 1.20388642e-01 2.18358621e-01 2.51017779e-01
-1.08314526e+00 6.88850343e-01 -9.83493090e-01 -2.30342537e-01
4.09108758e-01 -5.88662997e-02 -1.59283191e-01 -3.75185847e-01
-1.11328697e+00 7.45779276e-01 5.87177157e-01 4.54408914e-01
-1.29226100e+00 -5.28110385e-01 -9.78604615e-01 3.23407724e-02
1.62176341e-01 -8.64230096e-01 9.05501962e-01 -1.44477606e+00
-1.62717676e+00 3.80289555e-01 -2.69678950e-01 -4.76852328e-01
5.11996508e-01 -1.88564196e-01 -6.12158954e-01 1.90701321e-01
-7.75619294e-04 6.44332111e-01 1.33782053e+00 -1.43491578e+00
-6.16362154e-01 -2.40299001e-01 -3.20050418e-01 2.73011148e-01
-3.74117434e-01 -9.30109471e-02 -6.29951179e-01 -1.14067566e+00
1.87019527e-01 -4.65742320e-01 -4.14843172e-01 -8.87836590e-02
-1.95480898e-01 -5.34816366e-03 9.87018347e-01 -8.99857223e-01
1.46978235e+00 -2.56190133e+00 6.42271519e-01 2.08355337e-01
4.14222628e-02 3.02817225e-01 -3.75228971e-01 1.67538032e-01
-1.64130986e-01 1.02293938e-01 -6.99609816e-01 -7.48684764e-01
-1.57936513e-01 5.79465628e-01 -2.66763091e-01 4.69553798e-01
4.96690303e-01 5.75095236e-01 -7.78893828e-01 -2.60738134e-01
2.40187824e-01 7.53105581e-01 -5.39897978e-01 6.14955187e-01
-5.52367568e-01 6.26513302e-01 -4.32293028e-01 4.33691919e-01
6.31913602e-01 -1.99933022e-01 7.30875460e-03 -1.19913265e-01
-1.12188555e-01 7.70648709e-03 -1.14465976e+00 2.09164667e+00
-2.93600470e-01 1.14200190e-01 2.17533380e-01 -1.22055459e+00
1.21379149e+00 3.34383130e-01 2.70151079e-01 -8.13767672e-01
4.08289656e-02 1.10405266e-01 -1.90678716e-01 -5.80569088e-01
2.22693756e-01 -2.67762631e-01 3.52728158e-01 7.04440624e-02
5.84555157e-02 5.22261187e-02 -1.22748129e-01 1.76274568e-01
1.07941771e+00 5.73925257e-01 -9.66158435e-02 7.93317631e-02
7.41902828e-01 -1.70499638e-01 7.14335740e-01 3.51537049e-01
2.92696133e-02 7.82914996e-01 1.75398946e-01 -1.97377011e-01
-1.11112952e+00 -8.58129740e-01 1.69992685e-01 9.28901196e-01
1.50999531e-01 -3.32435906e-01 -8.04305434e-01 -7.39208996e-01
-3.38184446e-01 8.77138078e-01 -7.19560862e-01 -5.33916950e-01
-4.77828413e-01 -5.28577149e-01 3.76881599e-01 6.14374578e-01
6.52689755e-01 -1.16668081e+00 -2.26124182e-01 4.87996042e-01
-1.29743621e-01 -8.23918104e-01 -5.76116025e-01 2.93611944e-01
-7.49395132e-01 -6.92148268e-01 -7.88647592e-01 -9.04651940e-01
7.27663577e-01 9.85288620e-02 5.82250059e-01 1.49260625e-01
-4.45630215e-02 1.48267876e-02 -6.34236872e-01 2.86662757e-01
-4.96342748e-01 -1.93439603e-01 -1.47595957e-01 4.25624341e-01
1.88699656e-03 -7.23132908e-01 -7.87170231e-01 3.23815823e-01
-1.29864419e+00 1.90588102e-01 8.32533300e-01 9.62284207e-01
7.40612030e-01 1.56385392e-01 5.09741843e-01 -5.46498299e-01
5.00997484e-01 -6.42657101e-01 -2.59530395e-01 1.88907385e-01
-7.14746892e-01 3.79813612e-01 7.51082003e-01 -4.93948430e-01
-1.43719673e+00 1.41384155e-01 -5.30993164e-01 -8.77625108e-01
-3.04695159e-01 5.75087130e-01 -3.82566690e-01 3.17734808e-01
5.45176268e-01 8.18045080e-01 1.37668207e-01 -7.20848441e-01
5.68110585e-01 5.58741570e-01 5.60089827e-01 -5.73708057e-01
6.28046095e-01 3.20834190e-01 -4.38079536e-01 -7.19222248e-01
-6.85843825e-01 -3.65891963e-01 -4.36519295e-01 -9.59578827e-02
1.00092614e+00 -9.64295566e-01 -1.14637472e-01 5.07978439e-01
-1.15587807e+00 -2.35420361e-01 -2.51123518e-01 2.21409112e-01
-6.36123717e-01 6.42356217e-01 -6.63791120e-01 -6.67838693e-01
-2.60030657e-01 -1.31340063e+00 9.34340477e-01 1.38976514e-01
2.81460006e-02 -8.00678968e-01 -4.53195814e-03 2.68488318e-01
4.39930350e-01 -4.55474257e-02 8.53306532e-01 -4.89661783e-01
-5.76925695e-01 2.37520635e-01 -2.63721108e-01 8.80849004e-01
-1.51626614e-03 -1.21448621e-01 -9.47145820e-01 -2.67593652e-01
2.86015749e-01 -1.31494179e-01 1.20105517e+00 2.83812225e-01
1.31282759e+00 -4.49484408e-01 -2.33949259e-01 8.27476799e-01
1.33088577e+00 2.91837692e-01 9.43389595e-01 4.66213018e-01
6.57683194e-01 3.06122512e-01 4.69220698e-01 5.88623464e-01
3.03491145e-01 4.66096461e-01 4.33194071e-01 7.79661685e-02
-3.75193745e-01 -4.32154864e-01 6.04476333e-01 1.00670063e+00
-6.20282851e-02 -2.51804054e-01 -4.80362624e-01 4.13134664e-01
-2.03659987e+00 -8.71295869e-01 1.15896493e-01 1.73822296e+00
1.01480675e+00 5.42357638e-02 -1.93530113e-01 2.61161625e-01
7.02195168e-01 4.82109934e-01 -4.68750775e-01 -1.14193775e-01
-9.43134129e-02 7.34980851e-02 -1.27753615e-01 3.95080566e-01
-8.19175184e-01 9.73828316e-01 4.79657698e+00 1.15850115e+00
-1.15617847e+00 1.89690500e-01 5.16924500e-01 2.77400583e-01
-6.53810024e-01 6.74765483e-02 -7.05085516e-01 7.16464520e-01
7.17572451e-01 4.42022651e-01 6.81993186e-01 6.71720624e-01
4.18345422e-01 1.38153702e-01 -7.31958151e-01 8.70860755e-01
1.85885146e-01 -1.18005252e+00 1.94536522e-01 -1.97255224e-01
5.33130467e-01 -2.57395446e-01 8.85105282e-02 4.13712591e-01
1.56066477e-01 -9.43482459e-01 9.19160366e-01 9.46518540e-01
6.59496665e-01 -6.64711177e-01 4.46548522e-01 5.83276629e-01
-1.16648400e+00 -1.23890683e-01 -3.43914539e-01 3.54549685e-03
2.87508756e-01 6.75957024e-01 -5.69575608e-01 7.12297142e-01
6.21809304e-01 1.04093921e+00 -3.18026900e-01 8.94406378e-01
-5.08947730e-01 6.68843210e-01 1.24214133e-02 2.63333112e-01
4.00619566e-01 -3.01636159e-01 4.26147103e-01 1.22593498e+00
4.75703865e-01 1.05824746e-01 2.68741101e-01 9.43934560e-01
1.44640952e-01 -1.45468533e-01 -3.21084529e-01 -3.11965495e-02
1.76844582e-01 1.07067049e+00 -4.17141348e-01 -3.07293028e-01
-3.03810060e-01 1.52458024e+00 3.19493264e-01 7.62054265e-01
-9.02739942e-01 -1.02706023e-01 6.21190786e-01 -2.66364403e-02
7.43235230e-01 -2.50067383e-01 -2.60169864e-01 -1.19499218e+00
-3.81253175e-02 -8.10377121e-01 2.12979883e-01 -8.83801281e-01
-1.30660605e+00 7.81994402e-01 -3.67125005e-01 -1.11390221e+00
-2.41961718e-01 -2.28131369e-01 -4.48081940e-01 9.39740479e-01
-1.73350739e+00 -1.30870426e+00 -2.26742864e-01 8.28077137e-01
8.78699362e-01 -1.49018362e-01 5.24132669e-01 3.40318322e-01
-7.46783614e-01 3.08869123e-01 8.61149430e-02 -2.04350784e-01
4.10335511e-01 -1.06806827e+00 1.04297541e-01 1.03080869e+00
4.03572135e-02 3.05883259e-01 6.49837017e-01 -8.45900059e-01
-1.29595077e+00 -1.17836952e+00 6.23275638e-01 5.53145073e-02
7.17740119e-01 -2.98125863e-01 -1.27723300e+00 5.35464406e-01
2.24394515e-01 2.15159431e-01 3.31330568e-01 -3.38709265e-01
-2.74565101e-01 -7.15480447e-02 -8.60415101e-01 4.31937724e-01
1.27441704e+00 -5.47016621e-01 -6.48501694e-01 -3.44517152e-03
8.87160361e-01 -3.17770451e-01 -7.38649070e-01 3.96566361e-01
1.56238675e-01 -7.49400616e-01 8.89179587e-01 -3.82634193e-01
6.28045678e-01 -6.23997450e-01 -2.53774375e-01 -1.58736694e+00
-5.69885731e-01 -3.76643032e-01 -4.22690839e-01 1.43139231e+00
3.01828563e-01 -3.04760069e-01 4.70245868e-01 1.95346043e-01
-5.72585881e-01 -7.61531115e-01 -8.94578338e-01 -6.92129076e-01
-2.91697621e-01 -5.81992865e-01 5.05720794e-01 8.13627481e-01
-3.02749336e-01 3.45208734e-01 -5.50241709e-01 2.75097668e-01
4.96354550e-01 -1.62564158e-01 3.10650796e-01 -6.91226661e-01
-5.36095858e-01 -3.61866921e-01 -1.07866704e-01 -1.44975233e+00
2.52563059e-01 -7.45311737e-01 3.91149014e-01 -1.50026178e+00
1.17513910e-02 -2.71487027e-01 -4.78547215e-01 4.55046058e-01
-3.71360391e-01 -1.56247854e-01 1.69003755e-01 3.62505227e-01
-5.27230084e-01 1.17006934e+00 1.43625283e+00 -3.15902799e-01
-8.89542177e-02 -1.36389047e-01 -5.04132509e-01 5.13059258e-01
5.67973614e-01 -3.82748246e-01 -6.24700129e-01 -6.02702498e-01
-6.79081157e-02 9.70979556e-02 4.51023847e-01 -8.51458013e-01
3.09271455e-01 -8.14016163e-02 4.61105913e-01 -4.63699341e-01
3.96103591e-01 -1.01425326e+00 2.33763561e-01 3.56406093e-01
-2.95019865e-01 -1.93423226e-01 -5.31727746e-02 9.45658326e-01
-4.76350725e-01 -1.42065972e-01 7.45358765e-01 -1.14287451e-01
-1.07560611e+00 2.14043871e-01 -3.19199204e-01 -2.66783744e-01
9.99435782e-01 -2.93327630e-01 -1.30591556e-01 -1.40694857e-01
-7.86705732e-01 1.10788956e-01 3.74293089e-01 4.10265237e-01
1.06430948e+00 -1.23523402e+00 -7.41676152e-01 6.39261305e-01
3.33339237e-02 1.55481219e-01 4.42061961e-01 7.31332839e-01
-1.33984938e-01 -1.04514398e-02 -3.50881834e-03 -6.45743072e-01
-8.28851819e-01 8.56077731e-01 1.30109206e-01 -2.23728761e-01
-7.59178698e-01 8.35848451e-01 2.75946915e-01 -7.61909932e-02
3.06893468e-01 -2.78353155e-01 -1.56052023e-01 -4.22728769e-02
4.75231618e-01 7.03581646e-02 1.20645249e-02 -5.88817716e-01
-2.04102337e-01 2.66594797e-01 -1.59505993e-01 -1.69512242e-01
1.43077576e+00 -3.41232210e-01 2.02080086e-02 5.76163195e-02
1.13891625e+00 -4.65628117e-01 -1.80876005e+00 -3.35335463e-01
-4.95237634e-02 -2.87115335e-01 2.81785697e-01 -7.67479718e-01
-1.19632339e+00 9.08026040e-01 6.02719665e-01 3.23652616e-03
1.48700976e+00 -7.43265077e-02 8.81049335e-01 -2.31677637e-01
2.71301456e-02 -9.96129572e-01 3.03195357e-01 5.43931127e-01
1.02499259e+00 -8.81624281e-01 -3.80725920e-01 -2.45871201e-01
-9.37433481e-01 9.73114550e-01 5.80756009e-01 -1.08837008e-01
5.57755828e-01 2.79828101e-01 -1.15176484e-01 -5.85977063e-02
-8.24371994e-01 -2.64295638e-01 3.64117593e-01 5.44021785e-01
1.48366973e-01 -2.46756107e-01 -9.76772085e-02 8.24549973e-01
2.84106046e-01 2.36753896e-01 -8.51901844e-02 9.24387395e-01
-6.47848368e-01 -9.76063430e-01 -2.61841953e-01 1.38487155e-02
-4.48299684e-02 -2.56073684e-01 3.00199743e-02 3.10215205e-01
3.02889794e-01 8.85367692e-01 5.03922366e-02 -6.02539480e-01
3.42109650e-01 1.59612194e-01 1.69063061e-01 -4.58084375e-01
-4.88304526e-01 5.21288514e-01 -5.10335937e-02 -6.43502414e-01
-4.93185848e-01 -5.01017451e-01 -1.26004279e+00 -9.01611447e-02
-1.98005572e-01 -4.31907289e-02 4.57951337e-01 1.19299650e+00
5.22259712e-01 9.30537224e-01 7.28061974e-01 -7.30570555e-01
-6.14721715e-01 -1.08272207e+00 -5.44764161e-01 5.19873440e-01
3.09962630e-01 -5.12510359e-01 -2.98828453e-01 3.71349514e-01] | [11.150444030761719, -1.218219518661499] |
1987f947-d262-4273-abed-68177a2d2cce | joint-passage-ranking-for-diverse-multi | 2104.08445 | null | https://arxiv.org/abs/2104.08445v2 | https://arxiv.org/pdf/2104.08445v2.pdf | Joint Passage Ranking for Diverse Multi-Answer Retrieval | We study multi-answer retrieval, an under-explored problem that requires retrieving passages to cover multiple distinct answers for a given question. This task requires joint modeling of retrieved passages, as models should not repeatedly retrieve passages containing the same answer at the cost of missing a different valid answer. In this paper, we introduce JPR, the first joint passage retrieval model for multi-answer retrieval. JPR makes use of an autoregressive reranker that selects a sequence of passages, each conditioned on previously selected passages. JPR is trained to select passages that cover new answers at each timestep and uses a tree-decoding algorithm to enable flexibility in the degree of diversity. Compared to prior approaches, JPR achieves significantly better answer coverage on three multi-answer datasets. When combined with downstream question answering, the improved retrieval enables larger answer generation models since they need to consider fewer passages, establishing a new state-of-the-art. | ['Hannaneh Hajishirzi', 'Kristina Toutanova', 'Ming-Wei Chang', 'Kenton Lee', 'Sewon Min'] | 2021-04-17 | null | https://aclanthology.org/2021.emnlp-main.560 | https://aclanthology.org/2021.emnlp-main.560.pdf | emnlp-2021-11 | ['passage-ranking'] | ['natural-language-processing'] | [ 1.06175505e-01 -1.39955372e-01 -2.65764028e-01 1.22371033e-01
-1.85876560e+00 -8.01696360e-01 4.93666798e-01 5.05575538e-01
-4.51583296e-01 9.47780550e-01 7.06162155e-01 -3.32711011e-01
-3.75864118e-01 -8.48778605e-01 -6.53926373e-01 -1.11611933e-01
2.82961637e-01 8.50138903e-01 5.98526657e-01 -5.61311543e-01
6.56559825e-01 9.18494910e-02 -1.59291887e+00 6.59576118e-01
8.99489641e-01 3.10047179e-01 3.34123224e-01 1.34090781e+00
-5.02560854e-01 1.13284397e+00 -7.76547551e-01 -3.36982280e-01
-4.34029698e-02 -6.41591251e-01 -1.33288574e+00 -4.70372707e-01
6.99763060e-01 -4.45623606e-01 -5.50600529e-01 2.57757276e-01
4.65759188e-01 5.08462727e-01 8.44789207e-01 -5.22930622e-01
-9.71576452e-01 4.11294252e-01 -1.40799955e-01 6.41377926e-01
1.14373147e+00 -1.86190277e-01 1.68860412e+00 -8.55600238e-01
8.09975803e-01 1.05761099e+00 4.35336620e-01 5.29809594e-01
-1.25017643e+00 -6.94787577e-02 1.27537265e-01 1.10261627e-01
-1.12629175e+00 -3.66645396e-01 2.93179244e-01 -1.05619900e-01
1.39108455e+00 8.91011894e-01 4.95327026e-01 6.96443141e-01
1.86791912e-01 1.08891964e+00 4.66529816e-01 -4.40116405e-01
8.68143514e-02 -2.14251667e-01 4.79467392e-01 2.88754076e-01
4.97242734e-02 -4.79904503e-01 -6.59898400e-01 -5.90047061e-01
2.96671957e-01 5.70073500e-02 -4.99706268e-01 2.81520545e-01
-8.98911357e-01 8.49112749e-01 7.54731223e-02 2.97449857e-01
-4.16456819e-01 9.06981379e-02 1.58481032e-01 7.82588124e-01
3.71078908e-01 1.10754514e+00 -5.55085421e-01 -1.95640102e-01
-1.22799647e+00 7.25136995e-01 1.11643744e+00 8.37057412e-01
8.96085322e-01 -5.30205548e-01 -1.03723669e+00 1.01226056e+00
1.30721465e-01 5.11151731e-01 4.23496097e-01 -1.19640803e+00
5.43840170e-01 5.52062929e-01 4.37288165e-01 -9.11138833e-01
-7.08976313e-02 -5.97964406e-01 -3.39519411e-01 -8.15027118e-01
5.20805344e-02 1.66275233e-01 -6.94924533e-01 1.40102851e+00
1.22190416e-01 -8.71799141e-02 1.20973304e-01 3.80926341e-01
9.90595818e-01 1.08522558e+00 -5.59064485e-02 -2.63729393e-01
9.39792216e-01 -1.26334751e+00 -3.64374906e-01 -1.00300953e-01
7.01956093e-01 -9.62516487e-01 9.98766005e-01 1.08304650e-01
-1.37632024e+00 -4.84625340e-01 -4.20540780e-01 -4.23972428e-01
-4.54704799e-02 -9.18920487e-02 1.16543069e-01 1.23587467e-01
-1.43691528e+00 3.94679606e-01 -7.77910352e-02 -2.93849289e-01
-1.23623081e-01 -9.81530361e-03 5.75139150e-02 -4.16796893e-01
-1.38465893e+00 9.10982668e-01 -8.08963031e-02 -2.80201912e-01
-6.27063036e-01 -9.57521856e-01 -6.25392079e-01 2.44704187e-01
1.39083192e-01 -1.29114664e+00 1.48788416e+00 -3.39266866e-01
-1.21076751e+00 5.90719402e-01 -8.51713121e-01 -4.61495340e-01
1.05569132e-01 -5.47555327e-01 -3.78828019e-01 6.75931633e-01
3.57708961e-01 6.74001992e-01 9.90469575e-01 -1.10242879e+00
-5.95677555e-01 1.77525446e-01 2.70303130e-01 4.22952235e-01
-6.24613650e-02 -6.71805143e-02 -9.23294961e-01 -4.18364018e-01
-7.19731823e-02 -6.16012394e-01 -2.61981428e-01 -6.59744442e-01
-2.32261106e-01 -4.71816361e-01 1.89377069e-01 -8.70998383e-01
1.88684726e+00 -1.62501132e+00 3.12879086e-01 -5.30865192e-02
2.62835860e-01 1.29833996e-01 -7.47995973e-01 1.15549242e+00
5.03068984e-01 3.96558523e-01 -2.59845760e-02 -3.18871588e-01
-1.16504803e-01 6.74148127e-02 -9.43182170e-01 -1.18431062e-01
1.58948824e-01 1.12890840e+00 -8.71404111e-01 -6.18221879e-01
-3.15628827e-01 1.20230839e-01 -8.50241423e-01 1.88611761e-01
-7.43179500e-01 4.14094388e-01 -6.98342323e-01 3.68807197e-01
1.72102019e-01 -6.58665121e-01 -4.32584226e-01 4.64760303e-01
2.05923244e-01 6.86110735e-01 -6.84498906e-01 1.64482713e+00
-6.21033788e-01 5.18261492e-01 -2.96232432e-01 -2.34846786e-01
7.54603744e-01 3.88514876e-01 3.76805961e-01 -9.64929640e-01
-3.89401346e-01 4.67712253e-01 -5.87295115e-01 -6.97526813e-01
1.35881627e+00 1.64062455e-01 -1.20758094e-01 6.91027045e-01
-2.39530772e-01 -3.66560668e-01 7.25973725e-01 7.63041735e-01
1.57981575e+00 -1.10064693e-01 -4.15006354e-02 5.91566861e-02
6.75606370e-01 2.36428529e-01 1.25759929e-01 1.54188180e+00
3.30599874e-01 8.28004420e-01 4.05411422e-02 2.66139358e-02
-9.13969338e-01 -1.03372872e+00 2.13108063e-02 1.26918876e+00
1.25867963e-01 -4.78162885e-01 -1.58203810e-01 -5.36962748e-01
3.14112753e-01 9.50391889e-01 -2.50711083e-01 -1.57282606e-01
-9.01960194e-01 -1.12643965e-01 3.80028248e-01 8.49524960e-02
-5.00817318e-03 -1.08380115e+00 -4.42827761e-01 4.62256968e-01
-7.68785655e-01 -5.71190953e-01 -7.42320180e-01 -2.72987515e-01
-9.52237904e-01 -8.99573147e-01 -1.21817005e+00 -8.11410367e-01
4.07142162e-01 5.73748410e-01 1.86478508e+00 6.73739791e-01
-3.83376516e-02 1.12285554e+00 -7.57674515e-01 1.98348343e-01
-4.69176441e-01 7.73552179e-01 -5.31582534e-01 -5.14545143e-01
2.80972779e-01 -3.32327992e-01 -9.06423390e-01 1.06967360e-01
-1.13314581e+00 -3.73882860e-01 4.89387274e-01 8.80384505e-01
7.75989294e-01 -5.40682137e-01 1.26923966e+00 -7.37859666e-01
1.16020012e+00 -8.31224620e-01 -8.06326643e-02 8.20347130e-01
-3.79220366e-01 3.01042259e-01 6.14136875e-01 -3.11597675e-01
-9.02445376e-01 -8.27943206e-01 -4.24819618e-01 1.95694081e-02
2.32125074e-01 6.30629838e-01 7.90383995e-01 2.61902243e-01
6.23599708e-01 6.01048052e-01 -3.34303498e-01 -6.54874921e-01
6.81631684e-01 5.34861326e-01 1.10670149e-01 -5.49302876e-01
4.69177127e-01 9.98244584e-02 -2.60727048e-01 -7.67487466e-01
-1.02464437e+00 -1.22726142e+00 -2.44941086e-01 -1.98634639e-01
5.15617490e-01 -8.95631969e-01 -3.35603565e-01 -2.10137684e-02
-1.21368587e+00 -6.96291700e-02 -5.42991817e-01 1.68144807e-01
-3.71507734e-01 5.35996258e-01 -9.20707226e-01 -6.39669180e-01
-7.88205504e-01 -6.32147491e-01 1.20450294e+00 2.25577652e-01
-6.22890234e-01 -8.93115342e-01 6.43458605e-01 6.29442632e-01
6.35405838e-01 -4.39147800e-01 1.18264306e+00 -8.81728172e-01
-9.59137261e-01 -4.54224885e-01 9.32579339e-02 -6.20682351e-02
-1.82311937e-01 -1.89348057e-01 -5.17791748e-01 -4.76127297e-01
-3.30931880e-02 -5.90489566e-01 1.41851366e+00 1.36553094e-01
7.53462672e-01 -4.53279614e-01 -2.00550988e-01 -1.52740642e-01
1.33867586e+00 -1.07699111e-01 6.14511013e-01 2.73943275e-01
1.92027941e-01 4.57919419e-01 4.59634990e-01 3.35536569e-01
5.03065467e-01 2.56021291e-01 3.09994686e-02 3.72337431e-01
-3.27620119e-01 -4.48162645e-01 1.58546716e-01 1.40674412e+00
6.13933563e-01 -6.41733170e-01 -7.55515993e-01 9.90453124e-01
-1.64862072e+00 -1.07870185e+00 -3.33542407e-01 2.29334378e+00
9.71912622e-01 -2.03482330e-01 -3.67856845e-02 -1.67318061e-01
3.56457084e-01 3.82465839e-01 -5.50625801e-01 -3.90496224e-01
-1.40771359e-01 5.85581362e-01 -2.84953061e-02 9.40244555e-01
-4.63367134e-01 8.27669322e-01 7.41035604e+00 8.34522724e-01
-3.54703456e-01 -1.24419995e-01 3.57723594e-01 -4.42431062e-01
-1.10001504e+00 2.15473980e-01 -1.09773135e+00 2.07743034e-01
8.88070107e-01 -2.36460671e-01 4.58089411e-01 2.67407447e-01
-1.79588079e-01 -5.15926123e-01 -8.86189222e-01 3.81039709e-01
4.62405473e-01 -1.46070147e+00 6.64872944e-01 -5.25947213e-01
1.00883186e+00 2.20265593e-02 8.53930861e-02 8.64652574e-01
3.53895038e-01 -9.07698452e-01 3.37209314e-01 1.07650030e+00
3.11658025e-01 -6.48930907e-01 3.16486061e-01 5.11640072e-01
-1.10411119e+00 -1.45277515e-01 -5.00400782e-01 -9.17164385e-02
3.98693502e-01 5.03549755e-01 -5.99068761e-01 7.26265848e-01
5.10609269e-01 5.64383686e-01 -7.28017271e-01 1.35013723e+00
-9.21595618e-02 7.38641143e-01 -2.45586276e-01 -4.48906153e-01
1.88731715e-01 1.28741205e-01 8.20547402e-01 1.11786485e+00
4.01922226e-01 -1.87959168e-02 1.23330727e-01 4.81587827e-01
-2.69508272e-01 3.54462206e-01 -3.79470140e-01 1.61993653e-01
7.99762666e-01 7.29215324e-01 -3.22263502e-02 -4.03563410e-01
-3.50501209e-01 1.09845889e+00 5.46663880e-01 7.50838101e-01
-4.23706651e-01 -4.05428410e-01 2.48798862e-01 -5.50617762e-02
4.88459408e-01 -4.96180877e-02 -9.13929939e-03 -1.27785170e+00
1.90169960e-01 -1.09971356e+00 7.17804432e-01 -7.58675933e-01
-1.37242901e+00 4.34696168e-01 -1.60950288e-01 -9.91739869e-01
-6.16061628e-01 9.92892012e-02 -4.32142615e-01 1.08647549e+00
-1.89003932e+00 -6.02621019e-01 3.55768681e-01 2.27289274e-01
9.11913514e-01 1.39432833e-01 8.48646045e-01 3.20431113e-01
-2.06127122e-01 4.97409165e-01 4.23163950e-01 -3.32925171e-01
8.10568452e-01 -1.08443260e+00 2.82379121e-01 6.23579621e-01
3.86564285e-01 9.88324285e-01 6.67643309e-01 -6.69511497e-01
-1.34853935e+00 -9.36599970e-01 1.55548322e+00 -9.15632665e-01
4.51362371e-01 3.75409335e-01 -1.17775178e+00 4.91033912e-01
4.20753688e-01 -6.58800662e-01 8.96477580e-01 3.28393340e-01
-4.03175682e-01 2.92921066e-02 -7.66089916e-01 8.80036235e-01
6.31839573e-01 -8.60007405e-01 -9.78065312e-01 2.35762298e-01
1.23941481e+00 -1.56708941e-01 -7.93832839e-01 2.89472222e-01
3.22716683e-01 -7.42684484e-01 1.14690077e+00 -7.36725271e-01
4.84491795e-01 -2.21617252e-01 -1.47173122e-01 -1.00870800e+00
-4.60314661e-01 -6.95174515e-01 -6.94173098e-01 1.10802674e+00
8.57290149e-01 -5.20411074e-01 5.51406264e-01 6.35977089e-01
4.11345363e-02 -1.14909005e+00 -9.49660301e-01 -5.72007954e-01
5.10657489e-01 -7.77707025e-02 6.73227608e-01 3.07550222e-01
-1.42427742e-01 6.90214932e-01 -2.66562283e-01 -1.13240138e-01
1.69676453e-01 5.86560845e-01 6.76733851e-01 -1.03411031e+00
-5.43307841e-01 -5.00143230e-01 4.29444253e-01 -2.01364732e+00
-1.34612219e-02 -9.82058644e-01 1.33704171e-01 -2.18412447e+00
4.08883512e-01 -1.45842761e-01 -9.42248777e-02 -5.44041842e-02
-8.12508762e-01 -4.67424020e-02 1.05849855e-01 6.26349211e-01
-1.29735792e+00 7.39122748e-01 1.38907719e+00 -2.07619891e-01
-2.91075230e-01 2.30065569e-01 -8.01626027e-01 1.13108553e-01
6.75437748e-01 -6.08497322e-01 -5.52142382e-01 -7.88953483e-01
9.51150954e-01 5.67144990e-01 7.68637508e-02 -7.25515783e-01
5.99993706e-01 1.82566315e-01 1.11596532e-01 -9.49352205e-01
3.58546555e-01 -3.54613900e-01 -9.96966437e-02 2.07775921e-01
-9.55158293e-01 4.29481953e-01 1.83626577e-01 9.93880391e-01
-3.73636931e-01 -6.77903414e-01 1.53219804e-01 -2.99621791e-01
-4.07253116e-01 1.62343323e-01 -7.56769478e-01 6.87435627e-01
3.98135990e-01 2.81206276e-02 -4.23135757e-01 -7.32559025e-01
-3.99956286e-01 7.29583681e-01 3.02615613e-01 4.77363110e-01
8.54463995e-01 -1.14501953e+00 -1.10839093e+00 -3.02174449e-01
2.85489261e-01 -1.09546511e-02 5.16453862e-01 4.55482066e-01
-3.26040030e-01 7.76023209e-01 7.70653844e-01 -3.83029699e-01
-1.16793203e+00 3.28885466e-01 9.46228132e-02 -1.07569742e+00
-4.80634838e-01 1.00729191e+00 -4.12107229e-01 -5.42918205e-01
8.34119841e-02 -1.13027759e-01 -5.16406596e-01 1.61137506e-01
6.15652502e-01 4.92639422e-01 -8.46165931e-04 -1.95409477e-01
1.17955610e-01 5.83869338e-01 -5.86357415e-01 -3.34787965e-01
8.48649681e-01 -5.33219576e-01 -4.40950602e-01 5.25863171e-01
1.37084293e+00 2.34825969e-01 -7.62546837e-01 -4.33850884e-01
1.43021166e-01 -4.51927096e-01 -3.30548346e-01 -9.43712294e-01
-2.91306078e-01 3.60830665e-01 -2.21379787e-01 2.59032428e-01
9.96036232e-01 1.10150896e-01 1.35008228e+00 7.64433980e-01
5.26018083e-01 -9.08871114e-01 4.03974533e-01 1.15952802e+00
1.03140450e+00 -7.99478412e-01 -1.36991218e-01 1.31573126e-01
-4.32477087e-01 8.44166696e-01 3.46001476e-01 2.29947064e-02
3.77315372e-01 -4.64703977e-01 -4.23867889e-02 -1.99606791e-01
-1.27733171e+00 -2.68019587e-01 5.93207777e-01 8.00078139e-02
3.57571959e-01 -4.42075044e-01 -4.79592949e-01 1.03298806e-01
-8.43027756e-02 -9.07750353e-02 2.65014917e-01 9.67490196e-01
-8.83998156e-01 -1.13516939e+00 -1.95694074e-01 9.20565486e-01
-5.96449614e-01 -4.95551318e-01 -3.87544543e-01 2.55312711e-01
-5.92648268e-01 1.48318481e+00 -5.10373414e-02 -1.64484322e-01
1.58560261e-01 3.60642940e-01 3.85941863e-01 -8.50585282e-01
-1.05348873e+00 -4.06344712e-01 2.06626892e-01 -1.89840868e-01
-7.16462955e-02 -5.24191320e-01 -9.78687227e-01 -2.10996911e-01
-4.49729055e-01 7.50901341e-01 -1.17956914e-01 8.38833272e-01
9.09093499e-01 5.13214171e-01 6.16947472e-01 -5.88928349e-02
-7.24800408e-01 -8.86975706e-01 -1.85136020e-01 2.79832393e-01
7.08425343e-01 1.73601642e-01 -4.39524442e-01 -3.07240248e-01] | [11.47359848022461, 7.771096229553223] |
c72fb3c3-b698-4f70-b5b6-46c9865c6ec6 | add-net-an-effective-deep-learning-model-for | null | null | https://ieeexplore.ieee.org/abstract/document/9877809 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9877809 | ADD-Net: An Effective Deep Learning Model for Early Detection of Alzheimer Disease in MRI Scans | Alzheimer's Disease (AD) is a neurological brain disorder marked by dementia and neurological dysfunction that affects memory, behavioral patterns, and reasoning. Alzheimer's disease is an incurable disease that primarily affects people over 40. Alzheimer's disease is diagnosed through a manual evaluation of a patient's MRI scan and neuro-psychological examinations. Deep Learning (DL), a type of Artificial Intelligence (AI), has pioneered new approaches to automate medical image diagnosis. This study aims to create a reliable and ef cient system for classifying AD using MRI by applying the deep Convolutional Neural Network (CNN). In this paper, we propose a new CNN architecture for detecting AD with relatively few parameters, and the proposed solution is ideal for training a smaller dataset. This proposed model successfully distinguishes the early stages of Alzheimer's disease and shows class activation maps as a heat
map on the brain. The proposed Alzheimer's Disease Detection Network (ADD-Net) is built from scratch to precisely classify the stages of AD by decreasing parameters and calculation costs. The Kaggle MRI image dataset has a signi cant class imbalance problem, and we exploited a synthetic oversampling technique to evenly distribute the image among the classes to prevent the problem of class imbalance. The proposed ADD-Net is extensively evaluated against DenseNet169, VGG19, and InceptionResNet V2 using precision,
recall, F1-score, Area Under the Curve (AUC), and loss. The ADD-Net achieved the following values for evaluation metrics: 98.63%, 99.76%, 98.61%, 98.63%, 98.58%, and 0.0549% accuracy, AUC, F1-score, precision, recall, and loss, respectively. The simulation results show that the proposed ADD-Net outperforms other state-of-the-art models in all the evaluation metrics. | ['Muhammad Asad', 'AHMAD MOUSTAFA', 'SYEDA FIZZAH JILLANI', 'Muhammad Aslam', 'SAQIB MAHMOOD', 'MUI-ZZUD-DIN', 'GULNAZ AHMED', 'SHAHID ZIKRIA', 'MIAN MUHAMMAD SADIQ FAREED'] | 2022-09-19 | null | null | null | journal-2022-9 | ['alzheimer-s-disease-detection'] | ['medical'] | [-1.01880215e-01 -4.16285684e-03 -4.23271023e-02 -4.87020522e-01
-2.12703809e-01 7.94409513e-02 2.94744194e-01 1.40883967e-01
-7.11957991e-01 8.88796031e-01 -2.04975829e-01 -1.51761904e-01
-1.92544505e-01 -8.73302460e-01 -1.45433858e-01 -5.55521727e-01
-6.16362870e-01 6.04835808e-01 4.83851314e-01 -6.15729077e-04
2.69216411e-02 5.54708362e-01 -1.28469837e+00 5.24847448e-01
1.14628589e+00 1.42030692e+00 2.63851374e-01 3.14717561e-01
-3.62421013e-02 7.51541853e-01 -6.48168325e-01 -3.23204041e-01
1.04536429e-01 -1.63525760e-01 -5.99830687e-01 -1.23693079e-01
2.60322422e-01 -6.68854296e-01 -2.32843935e-01 1.27250564e+00
7.11962938e-01 -4.19752508e-01 6.98451519e-01 -1.21477771e+00
-4.50140774e-01 -1.18040117e-02 -5.45606971e-01 8.66932154e-01
-3.57956856e-01 1.33338690e-01 3.59676063e-01 -5.95740676e-01
3.66641968e-01 1.19667184e+00 8.84753823e-01 7.29732335e-01
-8.62837553e-01 -9.26513672e-01 -1.37285069e-01 7.02422619e-01
-1.21566927e+00 -1.06947914e-01 5.27248979e-01 -5.99364936e-01
7.58278906e-01 1.55949116e-01 9.91811633e-01 7.81411946e-01
7.44910359e-01 5.72153926e-01 1.28167582e+00 -2.28742048e-01
4.64770585e-01 -9.87212285e-02 6.37417793e-01 8.04633200e-01
4.82555419e-01 -1.13046095e-01 1.94339693e-01 -4.03975368e-01
8.65020275e-01 4.11810905e-01 -1.68399796e-01 -6.81706890e-02
-9.68289196e-01 8.23938012e-01 8.49517345e-01 2.79253244e-01
-4.97378528e-01 -1.98808238e-01 5.65939128e-01 1.88769758e-01
3.59162748e-01 3.51523049e-02 -4.24033254e-01 4.73424554e-01
-8.73877704e-01 3.92983884e-01 3.81496400e-01 4.19101387e-01
1.61350772e-01 -6.48122188e-03 -1.66241705e-01 1.13440406e+00
4.32077110e-01 5.24779975e-01 1.00226712e+00 -7.08463252e-01
3.19067717e-01 1.00803304e+00 -3.42089593e-01 -8.51587892e-01
-7.78447509e-01 -7.06780732e-01 -1.45360816e+00 4.35102850e-01
2.64220715e-01 -1.43078715e-01 -1.23273587e+00 1.46532214e+00
-4.67586108e-02 -1.74265593e-01 -5.59641048e-02 8.95873129e-01
8.51607859e-01 3.51677209e-01 2.94930428e-01 -2.37584442e-01
1.59948814e+00 -5.58346987e-01 -6.61688685e-01 -4.02376026e-01
3.75505507e-01 -2.09241033e-01 8.73169720e-01 3.02949280e-01
-8.48913372e-01 -4.15941447e-01 -1.20704699e+00 2.16131240e-01
-2.98012495e-01 4.26381618e-01 7.03593612e-01 7.25336015e-01
-1.07843208e+00 4.52207327e-01 -9.97079253e-01 -2.85008401e-01
1.16741621e+00 6.41875982e-01 -2.77769536e-01 -1.89800724e-01
-1.23647034e+00 8.24416578e-01 2.61991918e-01 1.53045058e-01
-9.27351236e-01 -6.96591020e-01 -1.99963242e-01 -5.13435379e-02
-2.37531826e-01 -8.03233027e-01 1.01261425e+00 -1.01998615e+00
-7.39819825e-01 9.82816696e-01 1.91140175e-01 -7.38786161e-01
4.48376089e-01 -1.75439566e-01 -6.26389325e-01 2.62983084e-01
8.54517743e-02 9.43742514e-01 1.97623327e-01 -7.44340658e-01
-5.83545268e-01 -1.18701744e+00 -1.30596295e-01 -1.44325376e-01
-3.29167992e-01 9.46325138e-02 1.04900986e-01 -6.68933094e-01
2.46292248e-01 -7.77869046e-01 -3.40643555e-01 4.11949128e-01
-2.26822376e-01 -1.80150017e-01 7.23524153e-01 -1.00671613e+00
8.98597062e-01 -1.87802684e+00 -5.38302422e-01 1.38285503e-01
6.80432200e-01 6.86277390e-01 2.63919294e-01 -5.18400967e-01
-2.46909335e-01 4.68818918e-02 -3.38769317e-01 2.55739152e-01
-3.90153617e-01 4.16668281e-02 4.39971119e-01 3.82032394e-01
5.62730543e-02 6.93626344e-01 -4.31875855e-01 -4.05256808e-01
-9.09505934e-02 6.07415020e-01 -5.97445965e-01 1.32399127e-01
3.08222752e-02 1.35430880e-02 -5.16240239e-01 8.21026683e-01
7.06334770e-01 -2.73681313e-01 2.31423259e-01 -2.89220452e-01
1.85916200e-01 -9.76401716e-02 -8.42202842e-01 1.05756819e+00
9.43802074e-02 3.59919518e-01 -1.19557224e-01 -1.27420723e+00
1.12832212e+00 2.73055196e-01 4.13481534e-01 -1.14586878e+00
4.30951715e-01 4.08649772e-01 4.74753648e-01 -7.49220431e-01
-5.58864117e-01 -4.01972234e-03 2.74063557e-01 2.14751571e-01
-8.55991617e-02 8.49648356e-01 9.57328528e-02 -6.38763383e-02
1.26686716e+00 -7.51499653e-01 4.69121248e-01 -4.57478911e-01
6.15722120e-01 -9.20739099e-02 7.15154052e-01 4.80777442e-01
-6.60963833e-01 3.87938827e-01 4.96309102e-01 -1.05172467e+00
-9.79915202e-01 -1.14641118e+00 -4.04713482e-01 4.32446212e-01
-2.75330901e-01 1.50290653e-01 -9.92699742e-01 -6.36434495e-01
-1.06552035e-01 2.80687332e-01 -5.95658362e-01 -4.48864520e-01
-5.68364561e-01 -1.49167109e+00 6.88808560e-01 6.70566201e-01
1.45600903e+00 -1.07009029e+00 -5.89455187e-01 5.13222888e-02
-4.41600755e-03 -6.16494060e-01 4.35432419e-02 -4.80906339e-03
-1.44257236e+00 -1.29935205e+00 -1.02624619e+00 -1.19578540e+00
8.32137108e-01 -1.31458968e-01 9.39254999e-01 1.85926296e-02
-5.29696345e-01 -1.02944046e-01 8.85989070e-02 -4.10680801e-01
-1.10961735e-01 -1.29248723e-02 4.21268716e-02 -2.10244462e-01
5.90509117e-01 -8.80553722e-01 -1.16496217e+00 2.19617352e-01
-5.81896424e-01 -1.61380589e-01 9.99032378e-01 6.93066001e-01
6.66521192e-01 1.69107169e-01 9.50284481e-01 -6.95040822e-01
6.65671945e-01 -3.96899045e-01 -2.48404950e-01 1.69164851e-01
-8.71645451e-01 -3.44250232e-01 3.86258274e-01 -3.60462725e-01
-6.95550144e-01 4.26689722e-02 -1.60184339e-01 -1.29335165e-01
-3.05617362e-01 1.59190103e-01 -5.47819614e-01 6.61404505e-02
5.77427030e-01 1.46553725e-01 3.53865504e-01 -5.82796991e-01
-2.93127984e-01 9.00554061e-01 5.72239876e-01 -7.21637458e-02
-3.19275223e-02 2.69563705e-01 -1.15225606e-01 -6.71468437e-01
-5.96430540e-01 -1.09972090e-01 -4.80152875e-01 -2.60615945e-01
8.52840662e-01 -9.68522966e-01 -5.39160311e-01 8.46084535e-01
-1.12388241e+00 2.09850864e-03 1.95464611e-01 5.76953590e-01
-7.00066090e-02 6.45415038e-02 -6.22440577e-01 -4.47604477e-01
-1.09006619e+00 -1.14966249e+00 4.91317153e-01 2.25007787e-01
-2.05621257e-01 -7.02827990e-01 -1.43431544e-01 3.16017061e-01
5.59970260e-01 3.04189622e-01 1.33159900e+00 -8.75735939e-01
-2.21117675e-01 -2.67747402e-01 -7.96075225e-01 6.36391580e-01
3.55030112e-02 -6.84259355e-01 -8.02949011e-01 -1.97441146e-01
2.23380685e-01 -1.93453655e-02 8.66845906e-01 5.78544199e-01
1.22977757e+00 -5.07273197e-01 -4.07891810e-01 3.43050331e-01
1.28568459e+00 9.40814734e-01 1.05042434e+00 8.19724619e-01
4.83456105e-01 1.59756511e-01 -4.95385239e-03 1.88560709e-01
3.62494171e-01 4.19285148e-01 5.51883340e-01 -1.66003436e-01
-3.60491395e-01 7.26238132e-01 9.64973196e-02 5.85780740e-01
-2.97913551e-01 1.33804277e-01 -1.14371288e+00 5.00658274e-01
-1.49706233e+00 -8.84157538e-01 -2.84103692e-01 2.04439235e+00
8.49727631e-01 4.33083862e-01 1.43383831e-01 4.24494237e-01
9.70555484e-01 -2.77695954e-01 -6.89369261e-01 -6.75692931e-02
1.38407629e-02 4.27893549e-01 3.86966139e-01 -1.80498466e-01
-1.18614066e+00 2.07901761e-01 5.33533812e+00 3.56062233e-01
-1.20670855e+00 3.91999006e-01 1.09413719e+00 -9.25634801e-02
4.01260465e-01 -7.63454199e-01 -7.54840732e-01 7.66169846e-01
8.40052843e-01 4.45402004e-02 1.98719010e-01 9.33659732e-01
2.10067391e-01 5.26460335e-02 -7.53721535e-01 8.66811037e-01
2.11180151e-02 -1.01742375e+00 1.92322060e-01 -5.39078331e-03
3.44361067e-01 1.19503550e-01 -1.08617265e-02 2.53152847e-01
-1.20580927e-01 -9.45272207e-01 2.76668400e-01 6.30106211e-01
7.56494999e-01 -9.05651450e-01 1.26806974e+00 -1.00833736e-02
-7.26744473e-01 -2.30503649e-01 -2.85267442e-01 -2.82595642e-02
-3.20543855e-01 9.95021164e-01 -8.48822176e-01 -9.18762013e-02
1.09100473e+00 4.67467636e-01 -7.95858264e-01 1.40619373e+00
1.65596232e-01 6.59811556e-01 -2.00732257e-02 -7.90699869e-02
1.01205938e-01 2.21628636e-01 2.42550418e-01 1.06913257e+00
4.15986598e-01 2.34652176e-01 2.00864106e-01 7.82967985e-01
-4.34528403e-02 8.23080838e-02 -1.35555834e-01 2.08856478e-01
2.31793046e-01 9.33221757e-01 -8.86517763e-01 -4.21049058e-01
-1.91389918e-01 5.74471831e-01 3.60130295e-02 -3.94114777e-02
-6.95332944e-01 -5.53395987e-01 4.84984785e-01 4.53263044e-01
3.48412432e-03 1.10994071e-01 -6.12552166e-01 -7.63093472e-01
1.45069078e-01 -8.67187023e-01 4.91214961e-01 -7.12614357e-01
-1.24048746e+00 8.13221574e-01 -1.98490515e-01 -9.42688942e-01
1.09906681e-01 -7.10595191e-01 -6.14893615e-01 7.24087536e-01
-1.34680223e+00 -7.91049123e-01 -6.36758864e-01 6.54701710e-01
3.66720468e-01 -6.22551084e-01 8.74200463e-01 8.31837893e-01
-7.16816604e-01 4.38135475e-01 -8.92611817e-02 4.77129877e-01
4.36885566e-01 -1.10120535e+00 1.29681289e-01 4.25003052e-01
-8.69436681e-01 5.43557107e-01 2.08171278e-01 -7.95105517e-01
-6.55109704e-01 -1.51722503e+00 8.19030285e-01 3.56591463e-01
2.97675818e-01 -4.37349156e-02 -9.89241123e-01 4.34638768e-01
-1.70102790e-01 1.79135814e-01 4.90690261e-01 -2.87871003e-01
-1.34088412e-01 -5.43880165e-01 -1.65193772e+00 2.69480973e-01
8.36465240e-01 -2.91206427e-02 -6.95515871e-01 3.76379520e-01
3.82904857e-01 1.44651923e-02 -8.69279683e-01 6.56007290e-01
7.52560735e-01 -9.61136103e-01 1.20210326e+00 -4.84922707e-01
4.26662415e-01 -1.86368842e-02 -1.29219577e-01 -1.07580709e+00
-5.20523250e-01 3.33266288e-01 -1.28177509e-01 9.15301740e-01
2.63026834e-01 -7.43614614e-01 7.49572754e-01 4.04840291e-01
-2.52299696e-01 -1.08649838e+00 -8.65259647e-01 -7.82743335e-01
8.40443522e-02 -1.38838947e-01 5.70418596e-01 8.52360904e-01
-5.22684038e-01 1.93041101e-01 2.03327060e-01 1.90848589e-01
7.75894463e-01 -5.58855891e-01 -6.93894550e-02 -1.85136378e+00
3.69409740e-01 -3.34031790e-01 -9.39089298e-01 -7.75460079e-02
-3.37858737e-01 -9.41529989e-01 -4.37849224e-01 -1.71716917e+00
5.85731924e-01 -4.75593448e-01 -7.52222180e-01 5.98296225e-01
1.38717622e-01 3.62132013e-01 -1.62088871e-01 3.23972851e-01
-3.33135784e-01 2.60378540e-01 1.30179656e+00 -4.85451013e-01
-1.38019085e-01 5.57576306e-02 -5.32661974e-01 8.61592233e-01
1.29778194e+00 -4.17492360e-01 -4.57304716e-01 -2.83994466e-01
-1.47455230e-01 -3.07532489e-01 8.42247427e-01 -1.63358259e+00
8.13110992e-02 3.35546374e-01 1.15352893e+00 -5.33747315e-01
1.36007771e-01 -6.10305130e-01 1.17327325e-01 1.03872383e+00
-1.32228792e-01 1.33554131e-01 7.24470839e-02 2.11371139e-01
2.16398239e-02 -3.10037993e-02 1.11202586e+00 -4.13534194e-01
-7.01581240e-01 4.19720948e-01 -4.37033921e-01 -5.53412922e-02
1.06510270e+00 -1.74680158e-01 -5.80571115e-01 2.86671072e-01
-9.79746878e-01 -6.57969862e-02 -1.99279591e-01 1.87469006e-01
8.29970121e-01 -1.51834214e+00 -5.36591351e-01 1.79527909e-01
-9.10613984e-02 -2.12676376e-01 2.58900613e-01 9.00094092e-01
-7.61879802e-01 4.01237279e-01 -8.35235417e-01 -5.32113731e-01
-1.37254441e+00 8.81661251e-02 7.31387615e-01 -4.14723516e-01
-7.40551531e-01 4.85787451e-01 2.12599307e-01 -2.31595084e-01
4.50966805e-01 -4.57697511e-01 -6.16794944e-01 3.02929096e-02
9.71313953e-01 6.67992473e-01 4.93755251e-01 -2.67312169e-01
-5.63117802e-01 1.94521338e-01 -5.70581377e-01 3.04414153e-01
1.58157623e+00 1.57496452e-01 -3.60731274e-01 -2.05467623e-02
1.11176705e+00 -9.04702544e-01 -6.62766397e-01 1.26832306e-01
-6.49064556e-02 1.23140663e-01 4.23504561e-01 -1.37747848e+00
-1.54516888e+00 9.12163854e-01 1.80156541e+00 1.17705241e-01
1.08460891e+00 -9.82661620e-02 1.21050501e+00 4.14457083e-01
2.61328697e-01 -8.42974603e-01 5.14779016e-02 3.03394139e-01
8.46356273e-01 -1.02088892e+00 3.54109071e-02 -7.59240612e-02
-2.69787431e-01 1.21693146e+00 8.07208061e-01 -3.17353487e-01
8.57927442e-01 1.19230188e-01 3.70675884e-02 -4.91653532e-01
-4.58382905e-01 2.28480592e-01 1.96204334e-01 8.24001372e-01
2.64415532e-01 1.79216787e-01 -5.67945540e-01 1.25361383e+00
-5.32096811e-02 5.59445322e-01 1.33419603e-01 9.26705897e-01
-9.50351894e-01 -5.57818115e-01 -3.54880273e-01 1.08539248e+00
-7.04413593e-01 4.64223288e-02 -4.18678850e-01 7.80202389e-01
6.57863915e-01 5.00301480e-01 3.40638250e-01 -3.03737551e-01
3.56438637e-01 5.94938882e-02 3.20756853e-01 -1.61647737e-01
-2.80144066e-01 -2.31178477e-01 3.52778845e-02 -3.92423630e-01
-3.93983126e-01 -3.85283142e-01 -1.40488601e+00 -1.31997094e-01
1.12038106e-01 -1.39501140e-01 6.01551414e-01 9.10532713e-01
4.58596826e-01 9.33832884e-01 2.89255381e-01 -5.11489153e-01
-2.18812197e-01 -1.03618014e+00 -5.21387577e-01 -5.02350777e-02
9.39659327e-02 -8.94431412e-01 -1.88451082e-01 9.91102904e-02] | [14.166129112243652, -1.7507214546203613] |
b3b71e28-2f8f-4389-92ae-0fa228b145bb | can-gamification-reduce-the-burden-of-self | 2302.03616 | null | https://arxiv.org/abs/2302.03616v2 | https://arxiv.org/pdf/2302.03616v2.pdf | Can gamification reduce the burden of self-reporting in mHealth applications? A feasibility study using machine learning from smartwatch data to estimate cognitive load | The effectiveness of digital treatments can be measured by requiring patients to self-report their state through applications, however, it can be overwhelming and causes disengagement. We conduct a study to explore the impact of gamification on self-reporting. Our approach involves the creation of a system to assess cognitive load (CL) through the analysis of photoplethysmography (PPG) signals. The data from 11 participants is utilized to train a machine learning model to detect CL. Subsequently, we create two versions of surveys: a gamified and a traditional one. We estimate the CL experienced by other participants (13) while completing surveys. We find that CL detector performance can be enhanced via pre-training on stress detection tasks. For 10 out of 13 participants, a personalized CL detector can achieve an F1 score above 0.7. We find no difference between the gamified and non-gamified surveys in terms of CL but participants prefer the gamified version. | ['Aneta Lisowska', 'Maciej Malawski', 'Arkadiusz Sitek', 'Tomasz Trzciński', 'Rosmary Blanco', 'M. Patrycja Lelujko', 'Maciej Kuś', 'Ryszard Pręcikowski', 'Sylwia Marek', 'Paulina Adamczyk', 'Michal K. Grzeszczyk'] | 2023-02-07 | null | null | null | null | ['photoplethysmography-ppg'] | ['medical'] | [ 2.01480202e-02 3.48165512e-01 -2.78456300e-01 -4.42881554e-01
-5.01380026e-01 -3.40501040e-01 4.60513793e-02 -1.17584892e-01
-5.58526635e-01 5.29602110e-01 3.29377592e-01 -2.41164997e-01
2.07449004e-01 -7.01339781e-01 2.66869128e-01 -1.61894009e-01
1.34954244e-01 -4.43578139e-02 -2.54228503e-01 9.51596200e-02
6.32184327e-01 -1.31188646e-01 -1.15956950e+00 3.92442048e-01
1.07568896e+00 7.96874940e-01 8.74493718e-02 5.20469844e-01
1.08351380e-01 6.27872944e-01 -6.74499452e-01 -3.30035329e-01
2.27865428e-01 -6.90028727e-01 -2.31600925e-01 -1.24702536e-01
2.43482068e-01 -8.91023993e-01 -1.49178103e-01 8.87382448e-01
7.78973877e-01 2.12623164e-01 2.13477343e-01 -1.44742739e+00
-4.06747669e-01 2.47093827e-01 -3.28192443e-01 7.97279254e-02
9.88253713e-01 5.83509803e-01 3.13239574e-01 -3.46853435e-01
2.17647597e-01 9.09475088e-01 8.50226283e-01 7.79150367e-01
-1.55015731e+00 -1.07443941e+00 -5.66925347e-01 1.43844783e-01
-1.06722891e+00 -5.52021921e-01 8.54008734e-01 -6.26896322e-01
8.42294395e-01 3.50766003e-01 7.45496333e-01 1.16941082e+00
-1.74667668e-02 3.30711156e-01 1.88737249e+00 -2.71128297e-01
8.93214166e-01 5.12582004e-01 4.35370505e-01 3.41352165e-01
3.17478865e-01 3.36556464e-01 -7.30885029e-01 -5.93972921e-01
8.35985243e-01 -1.88716561e-01 -2.15025961e-01 -3.49103063e-02
-6.35770380e-01 7.73037791e-01 -1.10376738e-01 -5.83978882e-03
-8.00626457e-01 -2.71645814e-01 5.08605182e-01 4.68238205e-01
2.55800873e-01 5.67902207e-01 -1.03733279e-01 -7.28757858e-01
-9.41338003e-01 2.38198638e-01 1.05495262e+00 1.31662756e-01
2.90940732e-01 -5.52237295e-02 -5.85846007e-01 9.54847932e-01
1.78244248e-01 1.37442708e-01 7.86024988e-01 -1.11523116e+00
4.37728688e-02 7.52164662e-01 4.39935893e-01 -1.00337899e+00
-5.45185089e-01 3.16197425e-02 -4.17435735e-01 4.85427320e-01
4.53819007e-01 -5.93895912e-01 -4.60928887e-01 1.78366005e+00
5.72438445e-03 3.68495762e-01 -4.45565164e-01 1.04641831e+00
5.43789923e-01 -5.05901538e-02 7.72045732e-01 -5.00269115e-01
1.43278062e+00 -2.15655163e-01 -8.82010460e-01 -5.98922074e-01
7.59550810e-01 -4.66642290e-01 1.52006555e+00 5.97652376e-01
-1.10362303e+00 -4.56713945e-01 -1.02827621e+00 3.13649327e-01
3.47388908e-02 2.50094533e-02 8.66476536e-01 1.80812430e+00
-1.08377814e+00 7.88009405e-01 -6.87296212e-01 -4.81304854e-01
3.87212068e-01 3.23109031e-01 1.24805430e-02 7.73789883e-02
-1.14763856e+00 1.13262737e+00 3.69811803e-02 -5.56430817e-01
-4.02098626e-01 -9.05047536e-01 -6.52441084e-01 3.14077407e-01
2.63830461e-02 -6.30088627e-01 1.20375860e+00 -9.81729388e-01
-2.01993060e+00 1.01490963e+00 2.23692693e-02 -9.71085951e-02
6.12911880e-01 9.90603715e-02 -3.55742782e-01 2.87270290e-03
-3.66952680e-02 2.95474470e-01 5.40955544e-01 -3.98405820e-01
-9.86968428e-02 -5.60846806e-01 -6.67710006e-02 4.04486477e-01
-4.60131764e-01 2.26807579e-01 4.73409712e-01 -1.13593109e-01
6.59241974e-02 -7.89335668e-01 -1.41131163e-01 -1.18879214e-01
-1.08947009e-01 -6.17413269e-03 3.39171618e-01 -9.75530207e-01
1.24748576e+00 -2.04769182e+00 -9.18449759e-01 1.64863899e-01
3.60010982e-01 4.14504766e-01 -3.65832187e-02 2.20492810e-01
-2.28526279e-01 2.54603863e-01 3.94812942e-01 -2.26094753e-01
3.14767420e-01 -3.69398028e-01 1.43095374e-01 3.54753047e-01
-2.07427189e-01 8.07267189e-01 -7.87466824e-01 -6.95177019e-02
4.57322925e-01 2.45621860e-01 -7.05582500e-01 5.04630804e-01
3.68231267e-01 5.51808663e-02 -3.03274188e-02 5.21879911e-01
9.56031442e-01 9.41636488e-02 4.80660230e-01 2.82690763e-01
1.29634663e-01 6.60972297e-01 -8.55848551e-01 1.15096152e+00
-2.76725739e-01 4.60088640e-01 -1.37061372e-01 -6.78001583e-01
1.16770267e+00 5.53852320e-01 3.75183880e-01 -9.95100677e-01
1.76334485e-01 -1.70640036e-01 7.02915415e-02 -7.50280440e-01
3.50519419e-01 -7.28703141e-01 -7.37849548e-02 9.30200636e-01
-2.79458046e-01 7.28294477e-02 -2.39921674e-01 -4.78388779e-02
1.39686167e+00 -4.56796259e-01 7.10826874e-01 -1.37989938e-01
-9.05095115e-02 -3.98707032e-01 4.27644402e-01 9.94322479e-01
-1.01252425e+00 6.80599511e-02 8.46950293e-01 -1.55789182e-01
-7.48549342e-01 -1.09816635e+00 1.43255442e-01 1.05024576e+00
-3.56895953e-01 -2.00524688e-01 -9.30760682e-01 -4.26990211e-01
9.93855372e-02 1.33616304e+00 -3.67849350e-01 -6.35491788e-01
3.04033726e-01 -7.75594294e-01 4.48454320e-01 6.46017015e-01
7.14198351e-01 -1.10758293e+00 -9.84480798e-01 7.55418912e-02
-3.70761573e-01 -8.29335630e-01 -4.64939296e-01 -3.98017138e-01
-1.05487394e+00 -7.84563243e-01 -3.45961124e-01 -1.96944833e-01
1.97352320e-01 3.88310254e-01 9.84205365e-01 -3.72921638e-02
-1.80165216e-01 5.31241834e-01 -7.53914788e-02 -4.20925409e-01
-1.54201701e-01 -2.85267770e-01 1.74569324e-01 -2.38388076e-01
1.27663291e+00 -8.08204353e-01 -1.01390910e+00 4.17065322e-01
-2.02339441e-01 -5.12005575e-02 2.90655643e-01 2.30581075e-01
-2.38174930e-01 -4.60529983e-01 9.68924642e-01 -7.53910661e-01
1.40696990e+00 -5.06518841e-01 -1.82469144e-01 -3.15086782e-01
-8.53228450e-01 -5.73689103e-01 4.80584837e-02 -8.53513718e-01
-1.00721574e+00 1.22597761e-01 1.72177821e-01 3.30533981e-02
-3.79439771e-01 5.43033838e-01 8.47046077e-02 -1.95552394e-01
9.73345637e-01 -1.04642309e-01 9.57501158e-02 -1.76115394e-01
-3.32359166e-04 1.03865409e+00 3.26598048e-01 -2.08531678e-01
-6.58922922e-03 -1.10090099e-01 -4.62603062e-01 -8.56074393e-01
-4.51820791e-01 -2.86367238e-01 -6.78744391e-02 -6.09822512e-01
5.28663576e-01 -8.25266182e-01 -1.54633963e+00 3.72556239e-01
-6.02379441e-01 -6.16451919e-01 1.01122282e-01 8.05013657e-01
-7.27609098e-01 3.44889402e-01 -6.73977256e-01 -1.43722939e+00
-4.76558536e-01 -4.49992150e-01 4.07346606e-01 5.88503778e-01
-1.13301277e+00 -7.00951695e-01 3.19332212e-01 8.12353671e-01
6.77091956e-01 2.43884921e-01 5.21375239e-01 -5.66959321e-01
-2.45227553e-02 -7.60088027e-01 -1.59695059e-01 1.59850374e-01
6.85786083e-02 -7.36666024e-01 -1.35727692e+00 -2.29327902e-01
6.31972313e-01 -4.80313063e-01 3.05651844e-01 7.03508258e-01
9.81278300e-01 -1.42079338e-01 7.87993222e-02 1.64879635e-01
1.26625550e+00 4.62913096e-01 1.43638992e+00 5.12287430e-02
-4.61361110e-02 3.74415427e-01 3.08225185e-01 5.96402884e-01
4.22295988e-01 4.94876176e-01 -1.37088358e-01 2.19741583e-01
3.29691499e-01 -3.86549622e-01 7.29908943e-01 2.34912321e-01
-1.53857172e-01 2.61979878e-01 -6.01156235e-01 1.44877121e-01
-1.39744556e+00 -1.18110490e+00 -2.56166956e-03 2.74137926e+00
6.72245443e-01 8.87804404e-02 5.18004417e-01 2.03591451e-01
7.42920876e-01 -2.50440359e-01 -6.22561097e-01 -7.54287481e-01
6.15755916e-01 4.49650168e-01 2.64667183e-01 3.42370212e-01
-6.42855108e-01 4.30948198e-01 7.51892424e+00 1.55746356e-01
-9.55334783e-01 1.43444672e-01 7.91599691e-01 -2.66826063e-01
1.70460388e-01 -1.03166685e-01 -7.65615329e-02 8.40082586e-01
1.52750278e+00 -6.45479500e-01 5.75356960e-01 8.55998814e-01
8.02562356e-01 -7.35850334e-01 -9.78152871e-01 1.25496769e+00
8.81685093e-02 -6.81989789e-01 -8.32602739e-01 1.03420377e-01
2.51630217e-01 -5.25538623e-01 -1.00802809e-01 8.42428744e-01
1.48280412e-01 -9.77066755e-01 -1.00267574e-01 7.14461267e-01
8.07753086e-01 -4.10717994e-01 5.13219059e-01 4.73343313e-01
-3.41978997e-01 -7.31270108e-03 -3.78310829e-01 -9.52410758e-01
-1.24634296e-01 5.08433759e-01 -1.23195016e+00 -2.51627922e-01
1.57736480e-01 -7.51283467e-02 -2.02940255e-01 1.28193021e+00
-1.74992368e-01 9.16314960e-01 -1.49894312e-01 -2.15266138e-01
-5.06251812e-01 -1.90700591e-01 4.96313870e-02 7.57335305e-01
5.15557289e-01 5.80811977e-01 5.00398278e-02 1.24145269e+00
1.33898705e-01 1.18127890e-01 -3.81104112e-01 -2.09731072e-01
4.72712427e-01 1.26710367e+00 -5.25998771e-01 -3.82031441e-01
-3.02673727e-01 9.17033792e-01 -1.14648882e-02 1.30779207e-01
-4.06991810e-01 -3.52485567e-01 7.11216569e-01 3.14825088e-01
-4.86441314e-01 8.28357190e-02 -1.06736326e+00 -1.02688241e+00
-1.86541826e-01 -6.35639906e-01 4.48458016e-01 -1.08781409e+00
-1.33458686e+00 -9.44596305e-02 -3.77500772e-01 -8.59378934e-01
-4.26854104e-01 -3.26972872e-01 -1.03835595e+00 1.36656117e+00
-7.81406522e-01 -4.13202226e-01 -9.35125709e-01 3.47161919e-01
2.16170788e-01 2.82148689e-01 1.08178449e+00 4.05115604e-01
-5.14183640e-01 6.33367121e-01 -7.73244560e-01 -1.59309193e-01
1.00652599e+00 -1.13896120e+00 1.13429241e-01 3.03863943e-01
-7.51459181e-01 4.86288846e-01 6.60860837e-01 -1.00783253e+00
-1.07621813e+00 -4.48284239e-01 9.42973435e-01 -5.92666566e-01
2.41489112e-01 -2.14786261e-01 -9.34858263e-01 6.17696345e-01
-9.96364877e-02 -5.82442522e-01 1.12636662e+00 5.40560365e-01
-1.22964524e-01 3.63688648e-01 -1.79347646e+00 6.89500391e-01
7.13282824e-01 -8.62002194e-01 -5.94607770e-01 6.41305149e-02
5.81391566e-02 -2.72043049e-01 -8.18299770e-01 -2.07495153e-01
9.50397849e-01 -1.25472033e+00 7.85169184e-01 -6.11113489e-01
4.72908616e-01 4.57681775e-01 2.94754833e-01 -1.59311712e+00
-3.68255764e-01 -7.02327490e-01 1.02678351e-01 9.60680604e-01
-1.55558422e-01 -9.01407123e-01 1.32542193e+00 1.86108339e+00
2.59018153e-01 -8.68491083e-02 -5.41365385e-01 -5.38606524e-01
-1.20026812e-01 -4.98379022e-01 1.62638560e-01 1.28827798e+00
1.42215800e+00 1.62769645e-01 -4.26502168e-01 -1.30736262e-01
6.08907044e-01 -3.60104382e-01 6.16409123e-01 -1.19915104e+00
-3.04498374e-01 -2.77728111e-01 -6.16538823e-01 -3.64466578e-01
-3.09767455e-01 -7.07636297e-01 -3.47005367e-01 -1.09537983e+00
5.99189579e-01 1.38624925e-02 -3.40614110e-01 6.26576602e-01
-3.72670949e-01 1.64033473e-01 1.66249245e-01 -2.85494536e-01
-9.68441665e-02 1.57517567e-01 7.32738793e-01 5.12480617e-01
-7.70828307e-01 1.55992672e-01 -1.23247683e+00 4.95640367e-01
1.27736390e+00 -3.01534265e-01 -4.40947950e-01 5.06244898e-01
1.62866022e-02 5.86125016e-01 6.11425996e-01 -1.00032842e+00
7.85820186e-02 -3.40546519e-01 4.89156127e-01 -1.27280191e-01
4.77631092e-01 -2.74253249e-01 1.80233419e-01 5.93478680e-01
-3.09515089e-01 -4.23025161e-01 3.53682339e-01 -5.97593859e-02
5.14870405e-01 -4.78089541e-01 7.93453872e-01 -2.28673026e-01
-2.76625901e-01 -2.15678841e-01 -6.65776491e-01 -1.27824888e-01
7.92057455e-01 -4.00188804e-01 -4.07125115e-01 -8.88507783e-01
-8.97757947e-01 -1.14455625e-01 2.93771058e-01 -4.19683494e-02
4.79455531e-01 -1.26535368e+00 -3.24934930e-01 3.58990043e-01
-1.57434165e-01 -1.21247327e+00 8.96881819e-01 9.99191701e-01
-1.22981109e-01 1.81207433e-01 -7.53757477e-01 5.43077197e-03
-1.23691261e+00 3.46258938e-01 4.53557342e-01 -1.16041973e-01
-2.95935363e-01 1.65905729e-01 -1.49907544e-01 -1.98451385e-01
-3.69817913e-02 8.75148848e-02 1.23713734e-02 -9.26872641e-02
1.06994402e+00 8.37847114e-01 2.03229506e-02 2.96842426e-01
-2.83795279e-02 -2.64062107e-01 3.70488279e-02 -3.04685533e-01
9.84516919e-01 -2.02264890e-01 2.71227688e-01 1.75309524e-01
6.34995162e-01 -3.32321256e-01 -9.66359615e-01 1.89403608e-01
-4.53282893e-03 -1.07591856e+00 4.32187796e-01 -1.26754057e+00
-6.01840079e-01 6.98313892e-01 1.16000867e+00 3.06730390e-01
1.11316931e+00 -5.84004402e-01 5.01685679e-01 1.51546866e-01
3.32907498e-01 -1.29550231e+00 2.11971119e-01 -1.65223315e-01
4.68957663e-01 -1.15741301e+00 -2.77355134e-01 -3.24063867e-01
-9.25726473e-01 6.71340704e-01 6.75356567e-01 -2.87077218e-01
3.97169054e-01 2.05077142e-01 2.26802886e-01 -2.38686427e-01
-7.65820026e-01 1.66382506e-01 -5.34344018e-02 1.04022682e+00
6.08418524e-01 5.78321278e-01 -9.48976636e-01 1.21619403e+00
-4.65401322e-01 9.72524166e-01 1.03159511e+00 5.59845686e-01
-4.78603184e-01 -7.81304121e-01 -3.71447831e-01 1.05676341e+00
-1.73290268e-01 1.40627354e-01 -4.91866171e-01 3.43729973e-01
-3.60089242e-01 1.20326817e+00 1.00396760e-01 -6.16092205e-01
5.71318269e-01 7.84095168e-01 2.37630829e-01 -6.05600238e-01
-6.32078528e-01 -3.03193897e-01 2.98368603e-01 -8.01748812e-01
2.98317405e-03 -8.26696515e-01 -6.08738005e-01 -8.43351245e-01
-4.74387445e-02 -3.61499846e-01 6.34153485e-01 6.55895591e-01
3.79373491e-01 3.28845143e-01 6.36694193e-01 -6.98853970e-01
-7.40002751e-01 -1.16876209e+00 -9.88361061e-01 5.18300354e-01
-2.90927291e-01 -5.81432104e-01 -4.73924935e-01 -3.50133896e-01] | [13.653636932373047, 3.133635997772217] |
ceea79f0-5c9f-44e5-b526-7a667c6563ea | improving-patent-mining-and-relevance | 2105.03979 | null | https://arxiv.org/abs/2105.03979v2 | https://arxiv.org/pdf/2105.03979v2.pdf | Improving Patent Mining and Relevance Classification using Transformers | Patent analysis and mining are time-consuming and costly processes for companies, but nevertheless essential if they are willing to remain competitive. To face the overload induced by numerous patents, the idea is to automatically filter them, bringing only few to read to experts. This paper reports a successful application of fine-tuning and retraining on pre-trained deep Natural Language Processing models on patent classification. The solution that we propose combines several state-of-the-art treatments to achieve our goal - decrease the workload while preserving recall and precision metrics. | ['Binbin Xu', 'Sylvie Ranwez', 'Walter Vermeiren', 'Théo Ding'] | 2021-05-09 | null | null | null | null | ['classification'] | ['methodology'] | [ 2.52591133e-01 5.79754375e-02 -3.14343333e-01 -2.18171731e-01
-6.70564532e-01 -7.84650743e-01 3.24595690e-01 3.46559912e-01
-6.42996311e-01 8.12642694e-01 -3.05201203e-01 -9.12747979e-01
-4.26457018e-01 -8.63253057e-01 -5.98211527e-01 -1.35629326e-01
1.49541825e-01 5.07669747e-01 3.18506733e-02 -8.65013450e-02
9.00113285e-01 9.36754525e-01 -1.68567491e+00 4.56803709e-01
9.00344312e-01 1.07155478e+00 1.81828603e-01 4.88203973e-01
-6.08597517e-01 9.68459785e-01 -6.91148996e-01 -7.29865968e-01
6.52779877e-01 2.12676585e-01 -9.28409159e-01 -2.17505276e-01
5.13795197e-01 8.71125385e-02 -9.29948390e-02 1.26847839e+00
3.61718565e-01 -1.26127973e-01 5.04648447e-01 -6.48907423e-01
-7.45660841e-01 6.08233511e-01 -6.57460690e-01 5.07002592e-01
1.69883162e-01 4.04450707e-02 9.79338825e-01 -5.91504574e-01
5.73655427e-01 9.48294222e-01 5.66927075e-01 1.44893765e-01
-1.07116592e+00 -7.46992886e-01 3.61756921e-01 -2.72659753e-02
-1.26460707e+00 -5.42527795e-01 4.40648079e-01 -5.94634593e-01
1.51463223e+00 -1.24575300e-02 2.79812515e-01 9.83258307e-01
5.28056920e-01 3.78375769e-01 9.52765405e-01 -4.67580050e-01
2.41995022e-01 1.91007793e-01 4.44757968e-01 4.77976471e-01
5.82448125e-01 -1.52029067e-01 -4.28529054e-01 -7.57048950e-02
5.60087740e-01 2.34459937e-01 2.20335834e-02 -6.82531893e-02
-8.52847755e-01 8.49210978e-01 -1.06917895e-01 6.00336313e-01
-8.61497402e-01 -1.23472251e-01 4.36756104e-01 9.67259467e-01
6.71366870e-01 1.22723222e+00 -1.22982955e+00 -4.67785448e-01
-1.28664458e+00 2.24805251e-01 8.61513972e-01 9.73794818e-01
5.82417786e-01 -4.00391906e-01 -2.00258464e-01 3.55225921e-01
2.83148158e-02 1.67400260e-02 5.57386935e-01 -6.57542348e-01
4.42872643e-01 8.95589054e-01 1.09455995e-01 -6.42855048e-01
-2.59735465e-01 -1.07842398e+00 -6.34680152e-01 1.96906552e-01
2.37782687e-01 -1.04544863e-01 -1.21588027e+00 1.14162421e+00
-3.89724374e-01 -1.55526623e-01 -8.70835409e-02 1.95788026e-01
3.72986048e-01 4.63715732e-01 2.76666641e-01 -4.20945108e-01
1.43331707e+00 -9.38930035e-01 -5.76223731e-01 -4.13474083e-01
6.21924818e-01 -1.14551950e+00 1.38722312e+00 9.83204186e-01
-1.10926735e+00 -6.29890859e-01 -1.15607274e+00 3.46209444e-02
-9.07094181e-01 2.80214280e-01 6.93918884e-01 9.25527632e-01
-6.49527490e-01 1.09287179e+00 -4.97522354e-01 -1.12775723e-02
9.50207591e-01 8.36637974e-01 -1.36236235e-01 -7.34473243e-02
-1.06095064e+00 9.32528079e-01 3.03453714e-01 -1.51556611e-01
-3.84389549e-01 -7.78004944e-01 -2.17036024e-01 4.87311304e-01
8.03475857e-01 -7.17233598e-01 1.47192144e+00 -7.40332544e-01
-1.62236130e+00 6.58077657e-01 -2.17711702e-01 -5.28845608e-01
5.25378406e-01 -6.99262500e-01 -4.11963582e-01 -2.08770722e-01
1.70967177e-01 4.53264296e-01 8.46086740e-01 -5.75821280e-01
-9.81234491e-01 -4.65006888e-01 1.13185309e-02 -3.32335293e-01
-6.33714557e-01 3.20916101e-02 -2.99552143e-01 -7.23123789e-01
-2.46458054e-01 -6.71899140e-01 -4.93645340e-01 -8.05389702e-01
-2.98682213e-01 -5.14373243e-01 6.18751168e-01 -3.97237241e-01
1.39382708e+00 -2.01516676e+00 -3.33669245e-01 3.52183342e-01
2.74908185e-01 6.87672317e-01 -3.71241748e-01 2.33906463e-01
-2.42246166e-01 5.10804296e-01 1.69568762e-01 -1.47785574e-01
-3.19979875e-03 -1.79193407e-01 -6.25264645e-01 1.36912078e-01
4.98768568e-01 9.55671251e-01 -6.97535038e-01 -1.69488713e-01
4.74639013e-02 2.43478373e-01 -1.82208523e-01 -9.58665013e-02
-5.05914032e-01 9.94706079e-02 -6.99519813e-01 5.81313252e-01
6.80306494e-01 -2.62218475e-01 -1.68789685e-01 1.57734409e-01
-4.33927685e-01 7.31077969e-01 -1.12930059e+00 1.72534609e+00
-5.44802606e-01 5.71986020e-01 -1.61931604e-01 -9.64774311e-01
7.98994422e-01 2.45904177e-02 4.27832007e-01 -8.61487329e-01
3.75130981e-01 4.12591577e-01 -6.33785352e-02 -4.87937152e-01
5.65668225e-01 2.88316403e-02 5.45559116e-02 2.44000301e-01
-1.21842198e-01 7.74654076e-02 4.28687781e-01 -1.23804189e-01
1.21759665e+00 -4.94028851e-02 4.48223680e-01 -3.15303534e-01
7.27255881e-01 1.76804841e-01 4.27229613e-01 9.75301802e-01
1.18743852e-01 -7.91572854e-02 6.78846300e-01 -8.19677770e-01
-8.19581330e-01 -5.78765452e-01 1.75772294e-01 1.08675241e+00
-4.76941824e-01 -3.79977405e-01 -6.31648958e-01 -9.57738400e-01
4.83251661e-02 6.30795777e-01 -3.58156979e-01 7.42667392e-02
-2.62272120e-01 -2.41233826e-01 2.84916550e-01 3.48855376e-01
2.05662340e-01 -1.03055143e+00 -7.35081851e-01 5.57566226e-01
3.41717094e-01 -9.80906904e-01 -1.46234095e-01 6.43152952e-01
-7.51749516e-01 -8.85443926e-01 -8.17587674e-01 -8.99452209e-01
2.13292435e-01 2.16990650e-01 1.36825669e+00 -1.26632556e-01
-3.74637574e-01 -1.24809854e-01 -1.29782096e-01 -1.17274988e+00
-3.43843549e-02 8.34364295e-01 -1.80928171e-01 -3.79322678e-01
9.65713024e-01 -5.24140418e-01 -4.03174073e-01 4.58724098e-03
-6.34875417e-01 -6.80259705e-01 1.21507347e+00 3.99454266e-01
7.32563555e-01 6.93929315e-01 7.70382762e-01 -1.20033348e+00
1.28725040e+00 1.15885153e-01 -8.79293382e-01 5.27937353e-01
-1.18782628e+00 2.00431079e-01 7.44958878e-01 -5.25004804e-01
-9.08435881e-01 3.27879101e-01 -7.25893676e-02 -3.24708581e-01
-3.20719481e-01 6.56393826e-01 -1.40057579e-01 -2.26965174e-01
6.24513924e-01 -2.16572762e-01 -4.01076198e-01 -4.83558446e-01
6.33171082e-01 9.78079736e-01 2.72828311e-01 -2.68669963e-01
4.95604515e-01 1.57819279e-02 8.03300831e-03 -6.47656322e-01
-1.05730724e+00 -7.30484247e-01 -6.43192708e-01 2.39186540e-01
5.32310367e-01 -8.41581583e-01 -8.26386929e-01 8.86412039e-02
-1.44339287e+00 1.09487489e-01 -5.62886178e-01 3.73963505e-01
-1.51269123e-01 3.31706762e-01 -4.83456463e-01 -6.06012285e-01
-6.32051945e-01 -9.85901535e-01 9.23361659e-01 2.11465061e-01
-4.32184428e-01 -4.77092743e-01 1.31673664e-01 3.54789019e-01
5.97486854e-01 -3.57131153e-01 9.31461394e-01 -9.57258880e-01
-4.87344205e-01 -5.35246253e-01 -3.44683260e-01 5.23582101e-01
1.91418618e-01 -2.13254616e-01 -1.09565830e+00 -8.26661289e-02
1.38229162e-01 -7.25724101e-02 1.08506382e+00 6.44088447e-01
1.48326707e+00 -9.69652981e-02 -3.72671723e-01 1.87760547e-01
1.15797210e+00 5.03455997e-01 5.84417999e-01 4.54538167e-01
4.09344256e-01 8.74103129e-01 7.24763811e-01 1.53976545e-01
-2.32562765e-01 9.36763361e-02 -6.89520389e-02 -2.72341277e-02
3.83912116e-01 1.24966465e-01 -9.38055217e-02 3.69613022e-01
-2.85372853e-01 -3.63827884e-01 -8.68514180e-01 3.65676045e-01
-1.86891234e+00 -8.02101433e-01 1.15919240e-01 2.08988738e+00
6.27380788e-01 1.02788925e+00 3.33813801e-02 5.72626777e-02
4.16433871e-01 -1.11149877e-01 -5.32196522e-01 -5.71893692e-01
1.20468512e-02 1.02241361e+00 8.80978167e-01 2.24354833e-01
-1.30578709e+00 1.10613024e+00 6.68722773e+00 1.04040623e+00
-1.17295825e+00 -8.95556733e-02 5.92918634e-01 -1.10118575e-01
-1.89783126e-01 -1.58828832e-02 -5.95162690e-01 2.28657454e-01
1.25787663e+00 -1.35975853e-01 3.01870495e-01 1.09157538e+00
1.23523846e-01 7.30238259e-02 -1.21192908e+00 1.12861466e+00
-2.84084767e-01 -1.60173810e+00 1.60472453e-01 3.80092889e-01
4.77449983e-01 5.93834408e-02 1.79625615e-01 4.27637637e-01
2.36459419e-01 -1.18280005e+00 4.99452829e-01 4.07317758e-01
4.77983207e-01 -9.18530643e-01 9.65042353e-01 2.33870208e-01
-1.24606037e+00 -4.66200590e-01 -4.57857281e-01 -3.56140375e-01
-7.97387287e-02 9.87266421e-01 -5.90758801e-01 6.66882396e-01
5.45408010e-01 5.67902327e-01 -4.53974605e-01 7.92851627e-01
-2.00163171e-01 7.75526017e-02 1.78429008e-01 -1.38172761e-01
4.43553925e-01 2.81633134e-03 9.77775529e-02 1.29814065e+00
4.22749490e-01 -1.68495268e-01 7.46711418e-02 6.99241936e-01
-3.13185096e-01 2.75982499e-01 -6.84160590e-01 -6.82425439e-01
4.56520528e-01 1.07414639e+00 -9.33667302e-01 -4.27173376e-01
-6.40965164e-01 8.02682221e-01 2.10879460e-01 -1.49534233e-02
-4.71926749e-01 -8.95956337e-01 4.65122223e-01 3.87562335e-01
5.61801314e-01 -2.18529448e-01 -7.81849027e-01 -9.64649856e-01
2.13837072e-01 -1.07346272e+00 2.79482454e-01 -2.37903669e-01
-1.43410337e+00 7.20506668e-01 -5.64424455e-01 -1.09172344e+00
7.30729550e-02 -1.07714605e+00 -4.84451443e-01 9.47828531e-01
-1.79156065e+00 -8.51455808e-01 3.11732870e-02 3.73353183e-01
7.92988241e-01 -5.95139325e-01 6.97102427e-01 6.29072964e-01
-4.66024965e-01 5.09356678e-01 -1.70492247e-01 -3.45976502e-01
5.50994396e-01 -1.16253459e+00 6.76525950e-01 7.71253169e-01
3.38740796e-01 1.08769369e+00 4.96901393e-01 -6.11105025e-01
-1.14455724e+00 -1.02815652e+00 1.38533258e+00 -3.13622922e-01
7.57229745e-01 -1.22410998e-01 -7.77971029e-01 2.48082861e-01
3.88446450e-01 -6.57677114e-01 7.42418647e-01 4.97642189e-01
-3.21302980e-01 -3.08148146e-01 -8.84347498e-01 3.96059096e-01
9.12677765e-01 -7.05091536e-01 -7.68088758e-01 3.22434783e-01
6.99832082e-01 7.61250928e-02 -7.50343204e-01 3.13336402e-01
5.25293946e-01 -5.52977502e-01 9.48901355e-01 -7.91823626e-01
1.95326626e-01 -1.59841433e-01 3.94338340e-01 -1.06998968e+00
-5.49013674e-01 -9.29954529e-01 1.77873522e-01 1.07733428e+00
9.38528717e-01 -5.76104641e-01 1.15914869e+00 6.91257596e-01
2.24292856e-02 -8.01344693e-01 -6.75619543e-01 -1.04066837e+00
5.49533106e-02 -3.28241140e-01 7.54930496e-01 1.01692867e+00
-1.97917297e-01 9.17380035e-01 2.66585629e-02 -1.60314485e-01
1.87060535e-01 2.69438088e-01 6.30373538e-01 -1.83462918e+00
-1.36587828e-01 -8.58336151e-01 -1.90816894e-01 -8.72905016e-01
1.37026440e-02 -5.38399398e-01 -6.62489086e-02 -1.53901505e+00
-9.75575969e-02 -5.02459034e-02 -7.97543287e-01 4.63863015e-01
1.98450476e-01 -2.91714877e-01 -1.05536193e-01 2.82536924e-01
-5.86814106e-01 -5.58525026e-02 1.05298710e+00 -4.22122538e-01
-6.22812629e-01 2.71905482e-01 -1.27451682e+00 5.65744638e-01
9.56760645e-01 -7.85615444e-01 -6.64916217e-01 -3.40498179e-01
7.82283425e-01 -4.64429587e-01 -1.66676164e-01 -8.40866387e-01
5.06216586e-01 -7.82884005e-03 4.09673482e-01 -8.19293141e-01
-1.58716485e-01 -8.91413987e-01 -1.55201882e-01 5.79520345e-01
-3.17990750e-01 3.05445135e-01 2.91449994e-01 5.08187890e-01
-3.94393206e-01 -3.41801971e-01 5.35448790e-01 -4.10752535e-01
-6.62969649e-01 1.05167083e-01 -5.82144797e-01 -2.45955572e-01
1.07191443e+00 2.10507377e-03 -4.78110075e-01 1.02214083e-01
-5.22968650e-01 1.00957640e-01 1.26146823e-01 3.88828397e-01
3.68707150e-01 -6.05515957e-01 -4.40089822e-01 2.06203386e-01
3.80578935e-02 -4.15900797e-02 -1.32861897e-01 4.73699421e-01
-5.08242190e-01 1.08251655e+00 -2.78321534e-01 1.99082475e-02
-1.07124376e+00 9.48701441e-01 -3.36758904e-02 -8.64366889e-01
-2.62751132e-01 1.10635197e+00 -2.56764382e-01 -1.40356526e-01
6.18264258e-01 -7.35920370e-01 -3.40854645e-01 5.05077302e-01
5.07966876e-01 3.31177741e-01 6.15387678e-01 2.64660925e-01
-5.53201020e-01 6.96180820e-01 -6.65782928e-01 1.09668717e-01
1.48624504e+00 2.36814156e-01 -2.58372873e-01 -4.73724790e-02
8.45775604e-01 2.33794197e-01 -6.22738957e-01 -7.86958411e-02
6.78094864e-01 -3.15834612e-01 3.14228088e-01 -1.01468217e+00
-9.46306884e-01 9.09562349e-01 6.85376048e-01 5.93985796e-01
1.09884167e+00 -1.20743997e-01 7.23606467e-01 9.59982932e-01
7.87746236e-02 -1.48080492e+00 -2.36814514e-01 4.79830593e-01
5.09365857e-01 -1.02097368e+00 2.29896411e-01 -5.95203400e-01
-2.75342971e-01 1.17952871e+00 1.64532274e-01 2.07228865e-02
8.45497131e-01 2.70170957e-01 1.49058299e-02 -5.77144325e-01
-5.95399618e-01 -1.76679194e-01 3.34974438e-01 3.35611373e-01
6.58699930e-01 -3.31931859e-02 -5.00885844e-01 7.36730814e-01
-7.86879510e-02 3.94449681e-01 2.44615167e-01 1.18753898e+00
-5.05269110e-01 -1.65945721e+00 4.68321331e-02 8.08111250e-01
-1.02477539e+00 -2.83616215e-01 -9.18729305e-01 6.30039930e-01
3.93721461e-01 1.12954021e+00 -1.20443784e-01 -4.09418464e-01
8.07189107e-01 9.03562829e-02 3.18147719e-01 -7.32376814e-01
-7.09661484e-01 2.70491749e-01 2.52925634e-01 -4.17475879e-01
-4.55691487e-01 -3.30676019e-01 -7.28936493e-01 -2.17306241e-01
-4.21549916e-01 2.70051271e-01 7.39025116e-01 8.52566183e-01
9.48334634e-01 8.79326940e-01 5.12780309e-01 -3.71049821e-01
-7.71423340e-01 -1.01758122e+00 -5.89185238e-01 1.13251489e-02
2.97212034e-01 -5.19482791e-01 -1.41929746e-01 -8.88414457e-02] | [9.86742115020752, 8.188393592834473] |
dfe63395-3457-4a5e-81f6-9ef245fda7fa | ask2transformers-zero-shot-domain-labelling-1 | null | null | https://aclanthology.org/2021.gwc-1.6 | https://aclanthology.org/2021.gwc-1.6.pdf | Ask2Transformers: Zero-Shot Domain labelling with Pretrained Language Models | In this paper we present a system that exploits different pre-trained Language Models for assigning domain labels to WordNet synsets without any kind of supervision. Furthermore, the system is not restricted to use a particular set of domain labels. We exploit the knowledge encoded within different off-the-shelf pre-trained Language Models and task formulations to infer the domain label of a particular WordNet definition. The proposed zero-shot system achieves a new state-of-the-art on the English dataset used in the evaluation. | ['German Rigau', 'Oscar Sainz'] | null | null | null | null | eacl-gwc-2021-1 | ['domain-labelling'] | ['natural-language-processing'] | [ 1.05611585e-01 4.08288166e-02 -5.41941881e-01 -6.20445907e-01
-3.96326125e-01 -6.19562447e-01 7.32581437e-01 4.00188297e-01
-9.75023091e-01 8.84219229e-01 -1.02732129e-01 -1.79072097e-01
5.70961982e-02 -9.12268400e-01 4.95789805e-03 -1.12894788e-01
3.57770056e-01 8.82821798e-01 5.84559262e-01 -6.78857982e-01
1.66044146e-01 1.65720329e-01 -1.58958638e+00 1.76766783e-01
5.31078458e-01 8.31179500e-01 3.37034464e-01 4.63942170e-01
-7.57534504e-01 6.93002820e-01 -4.54172730e-01 -4.11201090e-01
-7.56779686e-02 -1.58166274e-01 -9.14381742e-01 -9.44961328e-03
2.32392371e-01 1.13337517e-01 -5.53835370e-02 1.06214631e+00
3.49290609e-01 4.60589319e-01 7.78297663e-01 -7.60356605e-01
-5.58372021e-01 4.92398322e-01 -2.42967834e-03 4.16063011e-01
2.47632012e-01 -1.14775963e-01 1.30020654e+00 -1.04577851e+00
8.74553859e-01 1.01906288e+00 3.03627282e-01 6.85629427e-01
-1.11247349e+00 -8.25420558e-01 1.92936763e-01 3.40897560e-01
-1.72463238e+00 -2.48148456e-01 6.65899515e-01 -6.56566679e-01
1.57682645e+00 -3.95839065e-01 2.17917338e-01 1.15582991e+00
-2.83827595e-02 2.65542008e-02 1.09983575e+00 -9.88290846e-01
4.84719187e-01 4.60582376e-01 1.03894520e+00 7.30263114e-01
1.74642906e-01 -2.53630638e-01 -4.41802710e-01 -2.91298211e-01
1.99861333e-01 -2.24961475e-01 2.44880512e-01 -2.12022349e-01
-6.92143857e-01 1.00856793e+00 -1.93967789e-01 6.64925575e-01
-2.63778538e-01 -6.94843829e-02 7.69103289e-01 2.77113855e-01
8.77097428e-01 5.53167820e-01 -6.64353311e-01 1.15584448e-01
-1.01073468e+00 -2.02470183e-01 9.82306182e-01 9.58684564e-01
1.11852098e+00 -8.32496807e-02 -1.36362508e-01 1.13818383e+00
3.38423491e-01 1.91701293e-01 6.68005526e-01 -5.35642326e-01
2.17121676e-01 4.42994177e-01 2.27111295e-01 -3.87549251e-01
-2.94263542e-01 -2.46965483e-01 -2.40978263e-02 -1.29766777e-01
5.87338619e-02 -1.55287459e-01 -1.22138369e+00 1.74074614e+00
6.47259206e-02 6.34417236e-01 3.74013454e-01 5.75007498e-01
1.01834738e+00 7.23695099e-01 7.91501284e-01 4.00566831e-02
1.60112882e+00 -8.22898924e-01 -7.74174154e-01 -5.88758111e-01
7.31041193e-01 -6.72370613e-01 1.27295840e+00 2.45956168e-01
-4.73975688e-01 -6.12308323e-01 -1.21220446e+00 -8.78570303e-02
-1.02447069e+00 -2.52374187e-02 2.85300136e-01 7.10801125e-01
-7.65732110e-01 3.37297499e-01 -4.38943774e-01 -9.29944694e-01
-9.40510184e-02 8.63641649e-02 -2.12745517e-01 -1.96933270e-01
-1.87125850e+00 1.34711564e+00 9.83372986e-01 -8.25720966e-01
-1.07039654e+00 -5.05022109e-01 -1.10882521e+00 -8.08578264e-03
4.88062978e-01 -3.16255808e-01 1.34285653e+00 -8.25790107e-01
-1.52969813e+00 1.53960896e+00 -2.55317390e-01 -3.29830050e-01
-1.90786868e-01 -1.49410561e-01 -9.14842427e-01 2.69868486e-02
3.09287846e-01 5.43788850e-01 6.35165036e-01 -9.94539976e-01
-6.50414765e-01 3.92606184e-02 4.23529625e-01 9.23241675e-02
-6.36949539e-01 3.73994738e-01 -2.36057505e-01 -4.61262733e-01
-6.77831948e-01 -5.84896028e-01 -1.22460939e-01 -4.00274366e-01
-7.15762097e-03 -8.45585585e-01 5.56729615e-01 -3.59217465e-01
1.34882379e+00 -2.05345106e+00 -2.06343502e-01 1.60650104e-01
-1.95459589e-01 7.47735381e-01 -3.37540388e-01 6.38603926e-01
-4.58040694e-03 1.72494829e-01 -1.82342499e-01 -4.55775023e-01
2.71359622e-01 7.14626610e-01 -5.28711617e-01 2.84418136e-01
2.05869779e-01 3.82013619e-01 -1.38704860e+00 -6.58777952e-01
5.76381922e-01 3.02053183e-01 -2.14986295e-01 3.87944162e-01
-5.00504613e-01 -1.65263548e-01 -4.17648315e-01 2.33636618e-01
4.12407547e-01 5.32920193e-03 3.63234460e-01 1.04516849e-01
-5.08997068e-02 7.85680890e-01 -1.00639153e+00 2.16820717e+00
-8.77952218e-01 3.73744160e-01 -5.65945208e-01 -1.12147522e+00
1.20782542e+00 7.45097220e-01 9.26730111e-02 -2.96780288e-01
3.09410125e-01 4.22826290e-01 -7.85994753e-02 -4.55414414e-01
4.52416986e-01 -6.77977741e-01 -3.19518179e-01 4.34863776e-01
1.00284731e+00 -8.23902190e-02 7.04248190e-01 -1.55142441e-01
9.13594425e-01 2.61469483e-01 9.41054106e-01 -5.28976679e-01
7.89536536e-01 2.33497471e-01 4.42403048e-01 4.91453320e-01
-1.45649269e-01 1.05484545e-01 4.69638743e-02 -2.97357529e-01
-1.09158015e+00 -1.02352154e+00 -4.16156292e-01 1.74326849e+00
1.22709632e-01 -4.61955577e-01 -4.68231916e-01 -7.87110925e-01
-2.75228888e-01 1.48077738e+00 -3.67469996e-01 -4.45389077e-02
-2.34754562e-01 -1.53286666e-01 8.20538342e-01 2.71250457e-01
2.17328206e-01 -1.21296418e+00 -5.42098939e-01 3.56136471e-01
4.40034382e-02 -1.36749244e+00 -3.78225237e-01 4.38451290e-01
-3.43189269e-01 -8.42606783e-01 -4.73908901e-01 -1.18045855e+00
4.01245058e-01 -2.56662499e-02 1.57414508e+00 -2.46575817e-01
-2.78964229e-02 3.31897020e-01 -6.09603286e-01 -5.42201757e-01
-3.13120097e-01 4.51647073e-01 3.70159606e-03 -1.50329858e-01
1.35478723e+00 -4.22334254e-01 5.12716211e-02 1.22283757e-01
-9.54238594e-01 -4.93109852e-01 -1.73546246e-03 5.32119155e-01
3.93331587e-01 -7.18493536e-02 7.68302023e-01 -1.13853097e+00
8.54478121e-01 -7.60213494e-01 -5.11759222e-01 5.29519320e-01
-4.94284302e-01 3.86896878e-01 7.14575052e-01 -5.73683619e-01
-1.24790347e+00 -4.53364756e-03 -2.64802158e-01 -1.68458089e-01
-5.82008779e-01 6.80034697e-01 2.62034386e-02 1.55169025e-01
7.96286821e-01 -5.88231124e-02 -7.19889879e-01 -7.33173072e-01
7.88754106e-01 9.32334244e-01 1.72907680e-01 -6.43478036e-01
3.82943422e-01 2.94093192e-01 -3.69392753e-01 -9.70023334e-01
-1.47357523e+00 -1.15354550e+00 -9.63097215e-01 -2.14139391e-02
1.02990210e+00 -9.78753924e-01 7.35056819e-03 4.95749786e-02
-1.50285983e+00 -7.40799308e-02 -3.74107689e-01 4.10841137e-01
-2.63496757e-01 3.53697926e-01 -3.61363679e-01 -7.19883502e-01
-4.95593995e-01 -5.95596671e-01 7.87208796e-01 1.03726760e-01
-5.04448175e-01 -1.46399403e+00 6.33603871e-01 -2.03138351e-01
1.99236646e-01 -2.40885302e-01 9.30888236e-01 -1.41773653e+00
3.42064112e-01 -1.65303990e-01 -1.18721537e-01 6.58652723e-01
1.84887260e-01 -4.27187890e-01 -1.12306738e+00 -6.37668669e-02
-7.94721246e-02 -6.96183980e-01 7.46252418e-01 -4.20768373e-02
6.44966662e-01 9.87364799e-02 -2.54890949e-01 2.47686282e-01
1.81312895e+00 8.82633030e-02 2.86273539e-01 2.60887355e-01
2.94402122e-01 7.22279727e-01 7.80578077e-01 3.92522186e-01
2.89749533e-01 6.76712751e-01 -1.79122314e-01 8.67671892e-02
-6.22221157e-02 -3.07886094e-01 2.73312569e-01 8.67785275e-01
3.32456321e-01 -3.92100751e-01 -1.19808924e+00 9.62328434e-01
-1.70950007e+00 -6.84630036e-01 2.72680551e-01 2.01197171e+00
1.23416221e+00 2.93889850e-01 -2.16881275e-01 -3.18468511e-01
9.54510987e-01 2.83095211e-01 -2.99270779e-01 -8.49641323e-01
1.43901929e-01 1.00582492e+00 6.38979793e-01 8.30649614e-01
-1.13706732e+00 1.49582899e+00 6.61383104e+00 1.04631531e+00
-1.04820693e+00 6.65833712e-01 -4.92489040e-02 2.52354443e-01
-1.23279318e-01 8.52041543e-02 -1.12259221e+00 1.94077119e-01
1.34774923e+00 -5.93976140e-01 2.32902139e-01 1.04236662e+00
-2.16044858e-01 8.73349383e-02 -1.11723757e+00 7.01782346e-01
3.62836480e-01 -9.52176869e-01 1.31369516e-01 -2.94803262e-01
6.35456264e-01 3.19946796e-01 -4.67439771e-01 5.92005491e-01
7.60738850e-01 -9.69340861e-01 4.25807625e-01 4.77395743e-01
1.12829006e+00 -7.14604259e-01 9.87670481e-01 4.73149836e-01
-1.17334247e+00 3.79720330e-01 -7.91791260e-01 -3.55352432e-01
3.28960031e-01 5.93369484e-01 -1.10217035e+00 3.26387525e-01
9.99801457e-02 7.37048149e-01 -2.53413498e-01 5.55743098e-01
-7.39180744e-01 7.41833925e-01 -3.68299842e-01 -2.04501599e-01
4.95150208e-01 1.07288346e-01 2.71530300e-01 1.79624140e+00
1.91446751e-01 1.37355596e-01 6.36260211e-01 6.91499233e-01
-1.91115275e-01 4.22369152e-01 -7.54838586e-01 -2.37549394e-01
6.22587204e-01 1.16580057e+00 -5.02377450e-01 -8.75432789e-01
-7.88473845e-01 7.80912995e-01 5.89652598e-01 3.44180733e-01
-6.02651119e-01 -7.58661091e-01 9.02914882e-01 8.90164152e-02
3.18913996e-01 -2.56550968e-01 -1.48256555e-01 -1.30737126e+00
-5.48609734e-01 -3.56182158e-01 6.06200218e-01 -5.57818472e-01
-1.77623177e+00 6.00255847e-01 4.12667185e-01 -8.55727732e-01
-5.34434974e-01 -8.95020604e-01 -5.43494165e-01 1.16900814e+00
-1.89842498e+00 -9.25884843e-01 -1.33600309e-01 5.54879844e-01
6.74244463e-01 -3.72630239e-01 1.42722166e+00 2.96268672e-01
-1.30991608e-01 2.57092714e-01 -6.28157854e-02 2.29567930e-01
1.02281523e+00 -1.12335646e+00 4.38318133e-01 6.63042367e-01
4.30456460e-01 5.97084701e-01 8.42949152e-01 -6.56940699e-01
-3.63245785e-01 -1.10529947e+00 1.56161380e+00 -3.48506093e-01
1.10261881e+00 -4.57978308e-01 -9.09545541e-01 6.75911069e-01
5.87962270e-01 5.99072427e-02 1.21614587e+00 2.40083188e-01
-6.79901719e-01 2.23682486e-02 -1.22749686e+00 1.16651155e-01
8.48315477e-01 -9.28350925e-01 -1.49687004e+00 3.65695119e-01
9.81548011e-01 2.00096101e-01 -7.19369233e-01 2.26327494e-01
1.33397698e-01 -1.83492407e-01 9.39734042e-01 -9.82244134e-01
1.91229060e-01 -2.93005854e-01 -3.05581897e-01 -1.38282728e+00
-3.25095266e-01 -9.84704718e-02 1.39836878e-01 1.38520825e+00
5.59271157e-01 -6.99900150e-01 3.70932221e-01 3.90068024e-01
8.79602283e-02 -2.39980802e-01 -9.73757327e-01 -9.28881824e-01
1.89824656e-01 -5.88925898e-01 1.94948792e-01 1.19477355e+00
2.73715973e-01 8.95603418e-01 -1.76271245e-01 -4.04470153e-02
4.71551359e-01 -3.26171011e-01 1.07903883e-01 -1.40426898e+00
-1.17582068e-01 -1.05568478e-02 -4.80971783e-01 -6.84719384e-01
8.52850318e-01 -1.18697155e+00 1.38691589e-01 -1.54316092e+00
-3.43551151e-02 -2.54355699e-01 -1.00361955e+00 7.26500809e-01
-4.89453301e-02 3.36288780e-01 2.49888487e-02 -1.05753191e-01
-8.67386401e-01 3.40186864e-01 4.43919957e-01 -1.43623531e-01
1.08988602e-02 -4.73992556e-01 -3.60233694e-01 8.20351720e-01
9.39906955e-01 -9.25230920e-01 -6.70629025e-01 -3.79532665e-01
1.22128114e-01 -5.64392388e-01 -1.29117593e-01 -8.00574183e-01
1.76715717e-01 -3.31003398e-01 -3.58892471e-01 -2.00813800e-01
2.53970504e-01 -7.41324961e-01 -3.52070451e-01 2.57638395e-01
-5.29904485e-01 -2.07212970e-01 3.80718589e-01 3.13361675e-01
-2.76893944e-01 -1.02562952e+00 9.59953249e-01 -4.99655962e-01
-1.50452006e+00 -1.25101358e-01 -5.00802815e-01 5.73762238e-01
1.14902055e+00 2.29493067e-01 -5.50292656e-02 4.81764264e-02
-7.83788681e-01 6.30236790e-02 3.80750030e-01 5.78469753e-01
3.10376674e-01 -1.20080841e+00 -5.75088203e-01 -8.99392813e-02
8.02542984e-01 -5.60155511e-01 -1.99187577e-01 -1.83137983e-01
-3.01270306e-01 7.07447886e-01 -2.28445143e-01 -2.08777577e-01
-1.07091486e+00 6.98267341e-01 2.13663876e-01 -6.74593151e-01
-3.33344817e-01 7.38718331e-01 -4.43317965e-02 -6.64251864e-01
6.22460581e-02 -7.36145228e-02 -5.86891413e-01 2.82468140e-01
5.26701212e-01 6.63392097e-02 1.00628305e-02 -6.61371768e-01
-5.99778473e-01 5.50221384e-01 4.75707315e-02 -4.21927124e-01
1.13293982e+00 -1.08206253e-02 2.02856399e-02 1.01165962e+00
1.08965480e+00 -3.59280527e-01 -6.02164924e-01 -8.04119408e-01
4.54557568e-01 -1.94065318e-01 1.64663628e-01 -9.22918439e-01
-3.18580568e-01 1.06567824e+00 4.37562704e-01 -1.16863720e-01
8.95204961e-01 5.25489897e-02 7.68399477e-01 5.53231537e-01
7.22105920e-01 -1.56553566e+00 -1.04727879e-01 1.17604887e+00
2.99355209e-01 -1.05146074e+00 -1.65693790e-01 -4.12279099e-01
-6.94785714e-01 1.12370431e+00 5.56926847e-01 -5.91237485e-01
9.98117208e-01 1.22609548e-01 2.28891954e-01 -2.05400899e-01
-9.58228111e-01 -8.52705300e-01 4.56552237e-01 6.82079613e-01
7.88736403e-01 1.24059692e-01 -9.60494876e-01 6.04990840e-01
5.07824905e-02 2.77226299e-01 3.26665640e-01 7.49101162e-01
-8.53657901e-01 -1.48698473e+00 1.19232975e-01 8.49900022e-02
-4.50294197e-01 -6.16325438e-01 -3.54534268e-01 5.52195013e-01
5.70141435e-01 1.06048453e+00 2.66238786e-02 -1.08135663e-01
3.91289711e-01 8.74301851e-01 2.04006389e-01 -1.50814104e+00
-4.78694558e-01 -2.86857456e-01 5.55413485e-01 -1.10895529e-01
-6.42408848e-01 -2.02218607e-01 -1.39512396e+00 1.26369409e-02
-1.96366444e-01 2.86890149e-01 6.38042629e-01 1.32747018e+00
1.06690258e-01 4.66069788e-01 1.14685856e-01 -4.91791874e-01
-4.44952965e-01 -1.34136021e+00 -6.68492198e-01 6.54233098e-01
-1.05814144e-01 -9.06410456e-01 -1.99335188e-01 1.34914443e-01] | [10.47866439819336, 8.810250282287598] |
0d02a7e0-bcee-426d-8fec-c8d9d4d28140 | probabilistic-3d-segmentation-for-aleatoric | 2305.00950 | null | https://arxiv.org/abs/2305.00950v1 | https://arxiv.org/pdf/2305.00950v1.pdf | Probabilistic 3D segmentation for aleatoric uncertainty quantification in full 3D medical data | Uncertainty quantification in medical images has become an essential addition to segmentation models for practical application in the real world. Although there are valuable developments in accurate uncertainty quantification methods using 2D images and slices of 3D volumes, in clinical practice, the complete 3D volumes (such as CT and MRI scans) are used to evaluate and plan the medical procedure. As a result, the existing 2D methods miss the rich 3D spatial information when resolving the uncertainty. A popular approach for quantifying the ambiguity in the data is to learn a distribution over the possible hypotheses. In recent work, this ambiguity has been modeled to be strictly Gaussian. Normalizing Flows (NFs) are capable of modelling more complex distributions and thus, better fit the embedding space of the data. To this end, we have developed a 3D probabilistic segmentation framework augmented with NFs, to enable capturing the distributions of various complexity. To test the proposed approach, we evaluate the model on the LIDC-IDRI dataset for lung nodule segmentation and quantify the aleatoric uncertainty introduced by the multi-annotator setting and inherent ambiguity in the CT data. Following this approach, we are the first to present a 3D Squared Generalized Energy Distance (GED) of 0.401 and a high 0.468 Hungarian-matched 3D IoU. The obtained results reveal the value in capturing the 3D uncertainty, using a flexible posterior distribution augmented with a Normalizing Flow. Finally, we present the aleatoric uncertainty in a visual manner with the aim to provide clinicians with additional insight into data ambiguity and facilitating more informed decision-making. | ['Fons van der Sommen', 'Peter H. N. de With', 'Amaan M. M. Valiuddin', 'Christiaan G. A. Viviers'] | 2023-05-01 | null | null | null | null | ['medical-procedure', 'lung-nodule-segmentation'] | ['medical', 'medical'] | [-3.71934064e-02 4.03455585e-01 1.05837788e-02 -4.21380192e-01
-9.43688214e-01 -5.56074619e-01 4.81145084e-01 3.95478934e-01
-5.38718402e-01 6.99699342e-01 1.31180674e-01 -2.40461066e-01
-6.39656186e-01 -5.42121708e-01 -4.41312224e-01 -7.95795143e-01
-1.05678573e-01 1.08554053e+00 2.59780318e-01 4.42540556e-01
-4.12236564e-02 6.47969246e-01 -1.37136710e+00 -4.88150306e-02
1.04804397e+00 1.05550349e+00 3.95897835e-01 3.15255165e-01
-2.28505164e-01 8.76016915e-02 -4.33238328e-01 -4.88482088e-01
1.70666859e-01 -8.72571021e-02 -5.83621323e-01 1.00464009e-01
4.08641174e-02 -3.18221509e-01 7.36580640e-02 1.14851117e+00
4.90062803e-01 -1.24637242e-02 1.19871330e+00 -8.63189578e-01
-4.88057435e-02 8.37825954e-01 -3.32272142e-01 1.26441091e-01
1.92956239e-01 4.16454449e-02 7.43370652e-01 -6.78370297e-01
7.02963948e-01 1.05405831e+00 4.22444522e-01 3.47838163e-01
-1.05848849e+00 -9.83240157e-02 -2.98188627e-01 5.10773435e-02
-1.50861013e+00 2.32463032e-01 7.28941083e-01 -7.84750581e-01
2.19036534e-01 4.28892612e-01 6.03355408e-01 1.11145020e+00
5.40803134e-01 4.50827599e-01 1.30627942e+00 -3.02443683e-01
4.08383191e-01 5.16296983e-01 -9.57083479e-02 4.77570146e-01
5.46826243e-01 2.06521861e-02 3.07428483e-02 -2.23471835e-01
8.03590655e-01 -1.61565803e-02 -3.82043362e-01 -6.07742012e-01
-1.03561616e+00 7.46452689e-01 6.41781032e-01 6.53049946e-01
-3.28862876e-01 4.26774248e-02 2.41026729e-01 -6.29676700e-01
4.62435067e-01 3.82624537e-01 -8.59101340e-02 -1.07613064e-01
-1.19156826e+00 1.45130634e-01 6.22691631e-01 6.16367161e-01
3.41104984e-01 -2.65559375e-01 -4.81330872e-01 4.04313564e-01
6.93085849e-01 4.54352379e-01 5.98665893e-01 -7.72261798e-01
2.27781296e-01 5.77359438e-01 6.47320598e-02 -9.49813366e-01
-5.86829185e-01 -6.62716031e-01 -9.47315395e-01 3.89366388e-01
5.90954542e-01 1.90251678e-01 -1.04129064e+00 1.52408457e+00
5.17067313e-01 2.89266948e-02 -2.78360784e-01 1.06921804e+00
6.17002964e-01 1.08826399e-01 1.28263505e-02 -1.16148815e-01
1.68295646e+00 -3.60800743e-01 -8.10049653e-01 1.03745654e-01
3.93648803e-01 -7.02706814e-01 9.18628871e-01 4.12874520e-01
-1.01116657e+00 -4.14593309e-01 -8.70547116e-01 2.83784896e-01
-1.02223910e-01 2.53275722e-01 1.45968586e-01 8.20688426e-01
-7.19888568e-01 6.69997752e-01 -1.19131279e+00 -8.09436068e-02
4.59967047e-01 1.24480106e-01 -3.04706603e-01 -1.86436564e-01
-1.30682456e+00 1.22180808e+00 7.22451568e-01 2.87322849e-01
-7.50545681e-01 -7.68169880e-01 -8.33380520e-01 4.09643650e-02
5.92402399e-01 -6.77029014e-01 9.24931169e-01 -1.64634421e-01
-1.16961956e+00 6.82815075e-01 1.92457721e-01 -4.86492932e-01
1.21016061e+00 1.09310597e-02 -1.29493594e-01 2.46267840e-01
-3.11546568e-02 3.67134899e-01 7.93646932e-01 -1.31014919e+00
-1.71643274e-03 -5.22300124e-01 -1.87818199e-01 7.06980601e-02
4.33477238e-02 -3.87484998e-01 -4.35072720e-01 -5.53177476e-01
4.14630473e-01 -9.66071010e-01 -4.12577629e-01 1.53600484e-01
-4.57362980e-01 1.80693284e-01 3.08173835e-01 -8.48642945e-01
1.24697840e+00 -2.01783109e+00 4.20411497e-01 5.77265739e-01
5.26184738e-01 2.44069509e-02 5.99966049e-01 -1.17674008e-01
1.59383073e-01 3.34574640e-01 -9.04177189e-01 -5.54758668e-01
1.98976994e-01 4.33296859e-01 3.58495325e-01 4.75480646e-01
1.78914264e-01 7.85691917e-01 -7.71148860e-01 -9.20882046e-01
6.21931612e-01 7.85154819e-01 -3.88866127e-01 1.36966392e-01
-1.43778011e-01 7.68334985e-01 -5.87959588e-01 4.10661072e-01
9.37706172e-01 -2.73952395e-01 1.60979666e-02 -4.91190255e-01
2.18988303e-02 -2.53676713e-01 -1.47037816e+00 1.78924525e+00
-3.73901278e-01 3.23864445e-02 1.25021815e-01 -4.42665040e-01
7.01733708e-01 4.52584445e-01 7.57074177e-01 -1.74164504e-01
4.31327254e-01 6.30639076e-01 2.46264845e-01 -3.56097221e-01
2.64797062e-01 -3.10978144e-01 -2.00592149e-02 1.31115794e-01
8.04993659e-02 -7.42905140e-01 -2.09351326e-03 1.05836615e-01
7.25986958e-01 3.44784446e-02 4.92635876e-01 -4.71280277e-01
6.24325037e-01 -2.40683109e-01 1.07813440e-01 4.73392636e-01
-4.02614355e-01 1.01812911e+00 5.70490718e-01 -1.81867033e-02
-9.26959872e-01 -1.20348489e+00 -8.39561641e-01 -3.34291160e-01
-3.69725041e-02 -7.10855126e-02 -8.95305634e-01 -7.13509560e-01
7.29977712e-02 8.61356437e-01 -7.50303090e-01 6.41700253e-02
-1.50258660e-01 -7.12106466e-01 3.73817146e-01 3.08694184e-01
2.28992537e-01 -6.24047637e-01 -9.42054749e-01 1.33482993e-01
-3.27313364e-01 -1.20288122e+00 -3.19992006e-01 2.76515782e-01
-9.80472565e-01 -1.20303977e+00 -8.81539166e-01 1.51864246e-01
6.05522871e-01 -5.60483277e-01 1.09715295e+00 -2.60135412e-01
-7.31729388e-01 5.34958184e-01 -2.45003864e-01 -2.23741487e-01
-6.20325446e-01 -9.75182950e-02 6.17011636e-03 -2.03775197e-01
-8.10981244e-02 -2.77282417e-01 -5.96997619e-01 3.30011219e-01
-1.42041242e+00 -1.62578017e-01 6.91272378e-01 7.76621163e-01
9.47918534e-01 2.71260023e-01 -6.45094225e-03 -8.90918374e-01
5.44788837e-01 -5.00856400e-01 -6.77201569e-01 2.32825309e-01
-6.54503524e-01 3.75978768e-01 2.23201707e-01 -2.16396570e-01
-1.26904845e+00 3.12503092e-02 -4.04493988e-01 -6.36930108e-01
-3.01092386e-01 5.97680569e-01 -6.13110848e-02 4.29303339e-03
7.11355805e-01 -1.88199371e-01 7.05742538e-02 -3.88873518e-01
2.35361427e-01 4.79880333e-01 2.34747693e-01 -6.82852745e-01
5.60495496e-01 3.77100766e-01 3.21381152e-01 -4.39021528e-01
-5.52489460e-01 -4.56520557e-01 -8.16152275e-01 -2.69651264e-01
1.15016675e+00 -4.88514483e-01 -4.82667357e-01 9.01620612e-02
-1.11516798e+00 2.13155419e-01 -6.74633682e-01 9.60295856e-01
-5.76224446e-01 7.14775145e-01 -3.10532302e-01 -9.48620617e-01
-1.55232236e-01 -1.78396058e+00 1.14347506e+00 1.80052921e-01
-1.68338746e-01 -1.07789063e+00 5.16258590e-02 3.39411110e-01
4.59441751e-01 5.03658473e-01 9.20291543e-01 -7.18953252e-01
-4.90505517e-01 -1.20491073e-01 -1.57789007e-01 4.43379104e-01
1.34790704e-01 -6.06079921e-02 -9.30141628e-01 1.45830214e-01
4.44000065e-01 7.25438744e-02 6.39531970e-01 7.91246295e-01
1.27983248e+00 2.37005562e-01 -2.53703296e-01 3.19700420e-01
1.53095031e+00 -1.20372489e-01 4.98881638e-01 -1.77714303e-01
4.84732717e-01 6.93956256e-01 6.04147851e-01 7.07371116e-01
2.56097373e-02 7.01441705e-01 9.74895298e-01 1.26947999e-01
-1.11211061e-01 -3.49600390e-02 -3.00314903e-01 6.27514362e-01
-1.74559310e-01 -2.76123047e-01 -1.11337793e+00 3.82370085e-01
-1.60475540e+00 -4.02623177e-01 -2.11031824e-01 2.29655123e+00
6.72548890e-01 3.41508508e-01 -3.02703947e-01 9.83803719e-02
6.24566674e-01 1.09567689e-02 -3.70850891e-01 -1.11422263e-01
1.06386267e-01 9.29774344e-02 5.64689100e-01 6.85129225e-01
-9.36325848e-01 2.63947010e-01 5.36336279e+00 1.09968865e+00
-8.99395585e-01 4.42580320e-02 5.79000115e-01 1.44742683e-01
-7.57297218e-01 -2.71619380e-01 -5.90265751e-01 5.97483754e-01
7.62833893e-01 8.59897286e-02 2.39658654e-02 7.80292988e-01
1.35513142e-01 -5.14494121e-01 -9.66043830e-01 1.01197553e+00
-5.06545082e-02 -9.47547317e-01 -9.15565938e-02 2.75276124e-01
4.88246828e-01 -1.99920982e-01 1.32289901e-01 -8.80152136e-02
-1.19174793e-01 -1.15467906e+00 5.63377023e-01 9.13508236e-01
7.58629620e-01 -5.41545510e-01 1.30328631e+00 4.48662221e-01
-8.60494614e-01 2.62188792e-01 -2.44324520e-01 6.54600263e-01
4.59030926e-01 1.17381167e+00 -1.26339650e+00 1.09794903e+00
5.63253045e-01 1.53161138e-01 -5.40669322e-01 1.17443991e+00
-1.41816586e-01 2.55083323e-01 -6.39399886e-01 1.47421435e-01
9.96727720e-02 -3.30515087e-01 7.19825685e-01 9.80979919e-01
7.80571938e-01 -1.13598527e-02 -2.38517493e-01 1.35308278e+00
1.94352746e-01 1.09686948e-01 -3.76769006e-01 -7.43008107e-02
1.64283603e-01 1.37000155e+00 -1.00696743e+00 1.16646767e-01
7.55887199e-03 8.07231605e-01 -1.15559578e-01 -1.66931842e-02
-1.00465262e+00 1.58306211e-01 2.84636438e-01 2.71673977e-01
1.03446469e-01 -1.98371619e-01 -2.91663885e-01 -1.02465940e+00
1.04769610e-01 -3.64182949e-01 2.72004485e-01 -7.60298550e-01
-1.40193462e+00 8.95851314e-01 4.72504795e-01 -1.14259970e+00
-5.01834512e-01 -8.91608417e-01 -8.84521976e-02 1.13810956e+00
-1.34220099e+00 -9.37392235e-01 -3.29281092e-01 2.97928721e-01
1.52283177e-01 3.98454428e-01 8.41331482e-01 3.71136516e-01
-1.49625808e-01 1.46941856e-01 1.03672788e-01 -3.79550815e-01
6.77439094e-01 -1.54575050e+00 -3.11026275e-01 6.26980484e-01
2.50201877e-02 3.95249695e-01 8.81028533e-01 -7.70181060e-01
-7.61382699e-01 -6.38282955e-01 4.41333920e-01 -7.28718400e-01
5.72370172e-01 -1.47239685e-01 -9.96752858e-01 3.32039088e-01
-2.22725585e-01 4.23408091e-01 5.94796658e-01 -5.07977486e-01
1.34665459e-01 2.91584313e-01 -1.72170186e+00 1.68428570e-01
6.28026307e-01 -3.34765226e-01 -6.34875476e-01 -4.48743813e-02
6.52652562e-01 -8.09691608e-01 -1.33453727e+00 6.59642935e-01
4.17686939e-01 -1.19985032e+00 8.60603511e-01 6.97936788e-02
3.58438164e-01 -3.44842434e-01 -1.36217475e-01 -1.25431347e+00
5.68081252e-02 1.17529534e-01 -3.33994068e-02 9.02555823e-01
3.62066180e-01 -4.13965791e-01 7.43630290e-01 1.03316593e+00
-2.18977854e-01 -8.29029262e-01 -1.47218704e+00 -3.55042934e-01
5.80654256e-02 -8.15823555e-01 4.23620403e-01 4.32542890e-01
-3.59747946e-01 -5.86347699e-01 -5.58444262e-02 2.50067800e-01
7.13267446e-01 -2.89030164e-01 -7.04700034e-03 -1.53933704e+00
-4.67524618e-01 -4.39193726e-01 -5.70225716e-01 -4.50654328e-01
-1.42029360e-01 -8.41841340e-01 2.26200037e-02 -1.51826239e+00
1.98979393e-01 -5.61005235e-01 -3.07874352e-01 -1.39731914e-01
-7.80007467e-02 -8.39359611e-02 5.78719899e-02 2.72650540e-01
-1.31521150e-01 6.25037313e-01 1.56581140e+00 -6.49851840e-03
1.04895115e-01 2.55733937e-01 -2.88586617e-01 7.22882152e-01
5.04107356e-01 -5.93681455e-01 -5.34887493e-01 -8.20677206e-02
1.37093902e-01 1.14543580e-01 4.17775631e-01 -9.39273238e-01
6.90734163e-02 2.44446605e-01 3.22280258e-01 -7.89063573e-01
3.50830644e-01 -1.52909946e+00 4.47615445e-01 5.05994678e-01
-1.42271146e-02 -4.10766214e-01 3.41912359e-01 6.39304459e-01
-3.63579422e-01 -7.31634617e-01 7.54727006e-01 -2.73112357e-01
-2.48831034e-01 2.76525646e-01 -9.94165689e-02 -3.32242390e-03
1.06161332e+00 -8.09514374e-02 2.32548624e-01 -1.01931974e-01
-1.20141518e+00 -1.88691076e-02 4.45522070e-01 -7.02772588e-02
5.68246722e-01 -1.14026761e+00 -5.24761379e-01 9.25445557e-02
1.09671429e-01 4.80784982e-01 7.50362694e-01 1.06441498e+00
-7.06869006e-01 5.27210116e-01 -1.91418380e-01 -1.04866195e+00
-9.70219135e-01 4.15192455e-01 5.42049766e-01 -7.21198916e-01
-2.69363731e-01 7.10214198e-01 2.06886418e-02 -5.04627705e-01
1.26681849e-01 -8.89577210e-01 -2.49581128e-01 2.38046929e-01
4.23656665e-02 3.83478343e-01 2.95016468e-01 -6.07884407e-01
-5.01775324e-01 6.83016062e-01 3.09082091e-01 -3.93967330e-01
8.81646454e-01 -1.85341686e-01 3.75727713e-02 4.81840998e-01
8.74098599e-01 2.75148779e-01 -1.14419353e+00 -3.99932489e-02
2.72876024e-02 -3.52084845e-01 1.87149227e-01 -8.93614650e-01
-8.66959572e-01 9.99291420e-01 9.49364662e-01 9.20980945e-02
8.68099213e-01 1.85392097e-01 1.17178157e-01 -2.80766696e-01
4.55451429e-01 -7.97296882e-01 -2.05654591e-01 -7.02039823e-02
9.26562607e-01 -1.45017934e+00 9.35102999e-02 -6.65449977e-01
-8.38158965e-01 1.16554272e+00 2.91217893e-01 3.18117440e-01
8.41510653e-01 3.19603324e-01 -1.80064552e-02 -1.37217000e-01
8.63437578e-02 -1.60154536e-01 7.75251806e-01 4.38606560e-01
4.64766502e-01 2.63437212e-01 -5.52301109e-01 9.61015105e-01
2.95168422e-02 3.89019288e-02 4.60624129e-01 5.91496706e-01
-1.03869699e-01 -1.13733423e+00 -6.26644254e-01 3.29443097e-01
-4.88484770e-01 1.29712269e-01 1.04061544e-01 8.70835125e-01
3.11925828e-01 4.53740656e-01 -1.50463283e-01 2.89625232e-03
3.25157583e-01 2.15454027e-01 5.76682746e-01 -6.16466641e-01
-4.04983193e-01 2.21872374e-01 -2.71712393e-01 -4.41255897e-01
-3.49333823e-01 -5.90721250e-01 -1.32588148e+00 2.99760044e-01
-4.76843745e-01 2.68223286e-01 1.17304158e+00 7.78394341e-01
3.96768888e-03 9.38119113e-01 1.03656426e-01 -8.05949807e-01
-8.43425930e-01 -8.78281713e-01 -8.07828307e-01 2.93993205e-01
6.13117814e-02 -9.85055566e-01 -7.44215608e-01 -4.20445591e-01] | [14.3719482421875, -2.1162331104278564] |
65afcf36-b986-4378-8656-f8b78022f248 | neural-network-accelerator-for-quantum | 2208.02645 | null | https://arxiv.org/abs/2208.02645v2 | https://arxiv.org/pdf/2208.02645v2.pdf | Neural network accelerator for quantum control | Efficient quantum control is necessary for practical quantum computing implementations with current technologies. Conventional algorithms for determining optimal control parameters are computationally expensive, largely excluding them from use outside of the simulation. Existing hardware solutions structured as lookup tables are imprecise and costly. By designing a machine learning model to approximate the results of traditional tools, a more efficient method can be produced. Such a model can then be synthesized into a hardware accelerator for use in quantum systems. In this study, we demonstrate a machine learning algorithm for predicting optimal pulse parameters. This algorithm is lightweight enough to fit on a low-resource FPGA and perform inference with a latency of 175 ns and pipeline interval of 5 ns with $~>~$0.99 gate fidelity. In the long term, such an accelerator could be used near quantum computing hardware where traditional computers cannot operate, enabling quantum control at a reasonable cost at low latencies without incurring large data bandwidths outside of the cryogenic environment. | ['Farah Fahim', 'Luca Carloni', 'Gabriel N. Perdue', 'Nhan Tran', 'Giuseppe Di Guglielmo', 'A. Barış Özgüler', 'David Xu'] | 2022-08-04 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 1.28068626e-01 -1.55254379e-01 -2.47241095e-01 -4.05899972e-01
-5.95214665e-01 -6.22843266e-01 3.14613134e-01 5.81510544e-01
-6.02843165e-01 8.95416737e-01 -5.88352561e-01 -1.05079365e+00
2.33964577e-01 -1.02650404e+00 -7.30307341e-01 -6.52872503e-01
-2.14408070e-01 5.27014315e-01 2.05424607e-01 -3.12375158e-01
7.27957726e-01 7.33175039e-01 -1.26485598e+00 8.74419659e-02
6.34288907e-01 7.75095403e-01 -2.54071467e-02 8.28658581e-01
4.13238525e-01 6.95161700e-01 -5.93802989e-01 -4.77571855e-04
4.33143288e-01 -6.12792790e-01 -5.10498941e-01 -8.11440706e-01
8.70246813e-02 -6.87645376e-01 -9.76569712e-01 1.05912781e+00
6.61622226e-01 1.01613412e-02 3.59007150e-01 -7.12739825e-01
2.16729850e-01 7.94523239e-01 4.43583280e-02 3.85623097e-01
1.71680897e-01 6.43414319e-01 5.00989676e-01 -7.24796206e-02
4.59203452e-01 5.45793891e-01 5.20366609e-01 1.95174232e-01
-1.49433148e+00 -8.00451517e-01 -9.52826679e-01 4.31194752e-01
-1.66639078e+00 -6.76810145e-01 3.22753102e-01 -3.83654758e-02
1.61885953e+00 1.91444069e-01 7.57385075e-01 3.68799537e-01
1.05680156e+00 -1.88682437e-01 1.22319770e+00 -6.94092274e-01
7.92762518e-01 -6.60350919e-02 -1.54889181e-01 7.87474632e-01
3.69057178e-01 7.29393005e-01 -4.94016051e-01 -2.25772917e-01
6.83059871e-01 -4.49389726e-01 4.42975909e-02 1.86680466e-01
-1.27127731e+00 8.36971939e-01 5.09878278e-01 2.67666996e-01
-1.95638761e-01 7.79738486e-01 6.60057008e-01 4.16782707e-01
-2.10875988e-01 9.54232574e-01 -3.25174183e-01 -4.37503606e-01
-1.15980422e+00 5.25835276e-01 8.43026340e-01 8.74922991e-01
8.01669776e-01 2.93946892e-01 -6.62793592e-02 -3.56192738e-01
-6.84252605e-02 5.36970198e-01 1.30105346e-01 -9.88244295e-01
-8.15319642e-02 -8.75383988e-02 2.24190280e-01 -1.66976050e-01
-7.42873847e-01 -1.75264820e-01 -5.61384916e-01 3.51249903e-01
4.03520852e-01 -2.33470634e-01 -7.61135697e-01 1.05246830e+00
2.79367626e-01 3.71596098e-01 1.25165805e-01 8.34339499e-01
8.61079395e-02 1.06732464e+00 -1.14693850e-01 -3.91453683e-01
1.29420388e+00 -5.42320013e-01 -7.09348321e-01 -4.64761928e-02
1.22146821e+00 -9.22223926e-01 5.89304388e-01 4.46844161e-01
-8.48680377e-01 -4.83043224e-01 -1.79521966e+00 -1.82412475e-01
-1.11503221e-01 -2.34235182e-01 1.13016105e+00 9.30016994e-01
-9.93746400e-01 1.50369346e+00 -1.09922743e+00 2.24319741e-01
-1.33349046e-01 8.32402706e-01 2.03328028e-01 5.62359877e-02
-1.27547026e+00 1.18705964e+00 8.55137050e-01 2.25819107e-02
-7.03930974e-01 -9.80429053e-01 -4.54082727e-01 -8.58066157e-02
-8.97434354e-03 -6.98558927e-01 1.53119493e+00 6.96092844e-02
-1.77929306e+00 2.07947075e-01 5.57918102e-02 -9.92627263e-01
-1.30340636e-01 3.07902783e-01 -5.79252660e-01 2.17162803e-01
-1.85824350e-01 2.79578149e-01 6.81789577e-01 -1.01510331e-01
-3.42219979e-01 -1.81781814e-01 -3.20435083e-03 -2.21063837e-01
2.50033945e-01 -5.13944142e-02 9.04603750e-02 2.63128936e-01
1.13706209e-01 -1.27597058e+00 -6.47894561e-01 -3.62631828e-01
-3.40228789e-02 1.53880715e-01 5.66968799e-01 -2.40352690e-01
1.11921561e+00 -1.58683503e+00 -1.54668942e-01 2.70480335e-01
-2.81657189e-01 2.58837938e-01 6.76334918e-01 5.84091246e-01
1.37405053e-01 -2.81031519e-01 7.12719187e-02 4.88664031e-01
-5.66804521e-02 -5.74302785e-02 -3.67112458e-01 7.40129650e-01
-9.66157243e-02 6.26520157e-01 -9.47787166e-01 -1.55856952e-01
4.25664127e-01 1.35357544e-01 -7.68619359e-01 1.69260040e-01
-2.59337246e-01 4.99511868e-01 -4.27192897e-01 6.25247777e-01
4.73233581e-01 5.92574961e-02 3.74027818e-01 -5.80843925e-01
-7.78181672e-01 6.28205538e-01 -8.76223981e-01 1.77264595e+00
-3.84554654e-01 6.00120068e-01 7.88539574e-02 -8.13806772e-01
9.34596181e-01 1.48779213e-01 2.35010803e-01 -1.02318931e+00
6.29565775e-01 5.42466283e-01 4.44684923e-01 -1.69089228e-01
9.37572122e-01 -5.03276706e-01 -4.28892285e-01 5.31986296e-01
-7.09680887e-03 -1.00065506e+00 3.34966272e-01 1.52761489e-01
1.31727684e+00 6.64302185e-02 1.61484733e-01 -7.10994065e-01
1.85616702e-01 5.70141494e-01 3.75565052e-01 8.53226900e-01
-3.41528773e-01 -2.24844977e-01 7.66025782e-02 -7.40866184e-01
-1.67659271e+00 -7.36813188e-01 -7.41146922e-01 5.57835400e-01
1.99360177e-01 -9.15303946e-01 -5.60971618e-01 2.51475841e-01
-8.32466409e-02 8.73889267e-01 4.37706143e-01 -4.42561835e-01
-6.94562793e-01 -8.64129007e-01 6.75620973e-01 2.87437975e-01
1.95478499e-01 -6.67432547e-01 -8.97382259e-01 6.01728737e-01
6.30811691e-01 -1.10225856e+00 -9.79985390e-03 6.83244109e-01
-1.04421127e+00 -4.21548337e-01 4.15580899e-01 -1.59820497e-01
3.74217182e-01 -6.24369867e-02 7.97494352e-01 1.62655294e-01
-6.63027644e-01 -5.14665782e-01 -1.68929294e-01 -2.10400611e-01
-7.20257580e-01 -9.02423486e-02 7.77171493e-01 -9.29628789e-01
3.73341769e-01 -5.78902304e-01 -7.43321776e-01 -2.63081998e-01
-3.41268033e-01 1.42856553e-01 5.82866907e-01 1.01417327e+00
5.60491562e-01 3.16460252e-01 3.13418388e-01 -6.60834849e-01
1.53650418e-02 -1.60822660e-01 -1.30519736e+00 -2.52818525e-01
-9.80122805e-01 5.96006155e-01 1.14172947e+00 -6.64054826e-02
-6.77284181e-01 3.52181315e-01 -3.95489424e-01 -1.45696357e-01
2.50186473e-01 2.39085749e-01 3.02658617e-01 -7.54396558e-01
1.00444996e+00 -1.30171273e-02 -5.52504696e-02 5.34982085e-02
3.04605693e-01 6.16876721e-01 6.76929653e-01 -8.29645514e-01
9.02590156e-01 1.56922638e-01 6.00065112e-01 -9.72461104e-01
-5.10964990e-01 1.36708230e-01 -6.28744006e-01 -1.27667887e-02
4.50419813e-01 -1.09898984e+00 -1.04852533e+00 -8.21152404e-02
-7.68537045e-01 -2.12255239e-01 4.45711426e-02 9.20169055e-01
-6.28803790e-01 3.66950184e-01 -8.21484864e-01 -6.94859445e-01
-5.33308148e-01 -1.41506636e+00 7.32656837e-01 7.41145015e-01
-2.85899401e-01 -4.89788145e-01 -1.77443296e-01 1.38604417e-01
6.08615577e-01 7.10012019e-02 7.55784273e-01 -1.92021295e-01
-1.11179972e+00 -4.57583845e-01 2.77474313e-03 -2.48074546e-01
-4.73540932e-01 1.18610322e-01 -9.53699112e-01 -6.82731092e-01
1.12353176e-01 -4.93555725e-01 3.57517898e-01 1.36628121e-01
1.16604030e+00 1.74083084e-01 -4.92176741e-01 7.95887828e-01
1.49100113e+00 5.29821455e-01 5.42281210e-01 1.13086492e-01
3.55292469e-01 -1.13641471e-01 9.14676547e-01 4.40020025e-01
-2.26621568e-01 6.35656714e-01 9.65536013e-02 4.96859968e-01
5.88655233e-01 -1.87324390e-01 3.50088090e-01 1.12474167e+00
2.71791555e-02 4.24783170e-01 -1.01859629e+00 -1.83631569e-01
-1.25276148e+00 -1.02112055e+00 -1.78147092e-01 2.53201771e+00
9.65993226e-01 5.86917520e-01 -3.24675739e-01 8.79367143e-02
4.35123414e-01 -3.04698110e-01 -6.70971274e-01 -1.22963011e+00
7.65563726e-01 9.39593911e-01 1.13572931e+00 3.36448878e-01
-7.64481246e-01 9.46407974e-01 6.91625071e+00 9.30255294e-01
-1.36059153e+00 -2.09390931e-03 2.86298573e-01 -1.96367368e-01
8.47995747e-03 9.20506775e-01 -1.05727768e+00 5.19354701e-01
2.34582305e+00 -4.41577017e-01 6.03325367e-01 8.35434198e-01
1.60274044e-01 -4.01976734e-01 -1.27251780e+00 8.70125532e-01
-5.89710653e-01 -1.95070624e+00 -5.96162736e-01 7.99175575e-02
3.93820673e-01 4.63251472e-02 -4.90261525e-01 6.88137174e-01
-5.25609627e-02 -9.41402853e-01 5.89243412e-01 3.19738805e-01
1.00717580e+00 -1.26710415e+00 6.59986854e-01 5.94532669e-01
-8.45456898e-01 3.76992635e-02 -8.19697142e-01 -6.96823895e-01
3.93977433e-01 4.93597180e-01 -1.13710558e+00 4.11016315e-01
3.32645565e-01 -8.77430588e-02 -2.08236739e-01 8.68828058e-01
1.67389661e-01 5.90243101e-01 -6.46305025e-01 -4.80368376e-01
2.31994748e-01 -3.93614799e-01 7.89182037e-02 9.53789949e-01
6.22729063e-01 3.39560181e-01 1.88792914e-01 6.55893683e-01
3.18580791e-02 -4.21618730e-01 -5.79568624e-01 -4.36902910e-01
9.90757048e-01 1.25202286e+00 -9.19682562e-01 -3.19808245e-01
-1.79483399e-01 6.92602873e-01 5.81573583e-02 -5.13630569e-01
-1.09726751e+00 -7.71196544e-01 4.67374355e-01 -5.30943200e-02
-3.39585021e-02 -6.89145863e-01 -3.99878442e-01 -9.99419451e-01
-6.16783977e-01 -6.67317510e-01 -1.46189705e-01 -4.14432764e-01
-6.34192944e-01 4.64229465e-01 -1.16950706e-01 -1.03432262e+00
-5.63141167e-01 -6.94256365e-01 -4.94959384e-01 9.41907227e-01
-1.04145730e+00 -4.69210744e-01 2.14358166e-01 1.10321850e-01
-1.19989917e-01 -6.75783902e-02 1.24321294e+00 1.86802953e-01
-5.49625814e-01 5.25466979e-01 4.82914358e-01 -5.27198374e-01
5.47611654e-01 -1.05540276e+00 6.67064965e-01 8.60365689e-01
-1.11410115e-02 8.40184748e-01 1.40445530e+00 -7.25858331e-01
-2.60797477e+00 -7.26034880e-01 5.40469766e-01 -2.10076600e-01
9.73487437e-01 -4.88620222e-01 -7.15413928e-01 2.94100195e-01
2.09359214e-01 2.05289304e-01 6.50556982e-01 1.31623462e-01
-2.07231026e-02 -1.67918324e-01 -1.13342297e+00 4.34617668e-01
4.65130448e-01 -9.50199783e-01 -1.20635718e-01 7.64594316e-01
7.48015642e-01 -9.47989225e-01 -1.28706193e+00 2.78120875e-01
5.32733798e-01 -7.18343019e-01 8.44158053e-01 -2.36291587e-01
-3.53842936e-02 -6.52672887e-01 1.30155003e-02 -1.01625335e+00
-2.08078980e-01 -1.18825114e+00 -2.84239650e-01 4.92094219e-01
3.41420054e-01 -3.16628098e-01 6.99702382e-01 7.39605069e-01
-5.10588467e-01 -4.13778603e-01 -1.22974777e+00 -7.29274571e-01
1.56332895e-01 -3.64345819e-01 5.44580877e-01 5.84315538e-01
5.18342555e-01 4.88450557e-01 -2.36052588e-01 5.02825499e-01
7.16059089e-01 4.21494246e-01 3.02894205e-01 -6.31573021e-01
-5.26737034e-01 6.39027730e-02 -7.83470690e-01 -6.17956758e-01
6.91478252e-02 -1.01092958e+00 2.16519400e-01 -5.92983425e-01
-1.87403426e-01 -7.43037343e-01 2.85558645e-02 8.53198096e-02
3.72648686e-01 2.56610066e-01 -7.65738115e-02 2.28948351e-02
-3.40635963e-02 2.95105100e-01 8.88967395e-01 4.20623086e-02
-5.39305322e-02 -3.11530292e-01 -8.96426961e-02 3.83713841e-01
1.03235841e+00 -6.42529309e-01 -2.90499628e-01 2.80540641e-02
2.58237094e-01 4.03645396e-01 -7.88149983e-02 -1.52002537e+00
7.99810767e-01 2.73627862e-02 4.84578788e-01 -5.33824503e-01
4.08086330e-01 -3.59108776e-01 4.13721979e-01 9.97240067e-01
3.09200555e-01 1.78682625e-01 2.79353052e-01 2.20895931e-01
1.98131695e-01 -6.48538113e-01 9.93026912e-01 -1.05650099e-02
-6.82291150e-01 1.61342785e-01 -6.20054603e-01 -5.17949164e-01
1.17683816e+00 2.53191322e-01 -5.29766619e-01 3.43146920e-01
-4.42068547e-01 5.77392615e-02 7.26761401e-01 -2.94744164e-01
2.26417303e-01 -1.09767807e+00 -1.27857745e-01 3.80373031e-01
-2.62362868e-01 -3.79629463e-01 4.24387157e-01 6.08445883e-01
-1.54197311e+00 8.97434592e-01 -5.04772604e-01 -6.54469430e-01
-7.26877630e-01 6.25714898e-01 1.71665132e-01 7.80902281e-02
-7.84390271e-01 5.53434908e-01 -8.95335734e-01 -9.27164778e-02
-4.89921004e-01 -2.83515871e-01 5.38258016e-01 -4.67091084e-01
8.88118386e-01 1.17580622e-01 5.49140275e-01 -2.45638683e-01
-2.95284003e-01 5.99199384e-02 -1.63311034e-01 -4.43785265e-02
9.82898414e-01 4.54403192e-01 -3.09208751e-01 1.07694693e-01
9.58088994e-01 -2.66593277e-01 -8.92681539e-01 1.51863337e-01
-1.58118214e-02 -4.07598525e-01 8.07829916e-01 -2.38171369e-01
-3.00121278e-01 8.16846669e-01 8.14456344e-01 -7.52731115e-02
8.78412187e-01 -4.68972385e-01 8.85699153e-01 1.05677807e+00
1.13462102e+00 -1.32132840e+00 -5.35232127e-01 6.15334272e-01
-5.51305674e-02 -7.82964528e-01 5.83838761e-01 -9.53606442e-02
-1.11369908e-01 1.55481958e+00 3.47473353e-01 -2.48718768e-01
4.22586530e-01 1.00474107e+00 -4.44154024e-01 -1.45516500e-01
-7.67029345e-01 1.26168907e-01 -3.89674008e-01 1.86009616e-01
5.05573511e-01 4.08613652e-01 -5.63466251e-01 -5.80091029e-03
-4.88163888e-01 2.69488133e-02 1.15155351e+00 1.38665581e+00
-7.52608299e-01 -1.33552456e+00 -3.24624300e-01 5.46936989e-01
-3.24959159e-01 -1.95416465e-01 5.51879704e-01 4.98414665e-01
-2.17114016e-02 7.44662166e-01 2.43726462e-01 -6.92640841e-01
8.75087902e-02 1.82510093e-01 1.02069569e+00 -4.63749707e-01
-6.38942420e-01 -1.83680236e-01 3.71712863e-01 -8.78703892e-01
5.71078658e-02 -5.17347157e-01 -1.81585515e+00 -1.15299165e+00
-4.11491483e-01 4.72810090e-01 1.08672214e+00 8.27158689e-01
4.00683790e-01 4.35203940e-01 6.53623104e-01 -1.17982984e+00
-9.83055353e-01 -6.10542238e-01 -6.48762226e-01 -6.54584348e-01
-3.66917066e-02 -5.34171522e-01 -9.07507911e-02 -3.19059283e-01] | [5.582597255706787, 4.923130989074707] |
8e856cd3-063e-42bd-a060-5e7d43489a8e | visual-speech-recognition-aligning | 1710.01292 | null | http://arxiv.org/abs/1710.01292v1 | http://arxiv.org/pdf/1710.01292v1.pdf | Visual speech recognition: aligning terminologies for better understanding | We are at an exciting time for machine lipreading. Traditional research
stemmed from the adaptation of audio recognition systems. But now, the computer
vision community is also participating. This joining of two previously
disparate areas with different perspectives on computer lipreading is creating
opportunities for collaborations, but in doing so the literature is
experiencing challenges in knowledge sharing due to multiple uses of terms and
phrases and the range of methods for scoring results.
In particular we highlight three areas with the intention to improve
communication between those researching lipreading; the effects of
interchanging between speech reading and lipreading; speaker dependence across
train, validation, and test splits; and the use of accuracy, correctness,
errors, and varying units (phonemes, visemes, words, and sentences) to measure
system performance. We make recommendations as to how we can be more
consistent. | ['Helen L. Bear', 'Sarah Taylor'] | 2017-10-03 | null | null | null | null | ['lipreading'] | ['computer-vision'] | [ 3.11552227e-01 4.61906083e-02 -4.08671737e-01 -2.78166711e-01
-9.98773396e-01 -3.91142607e-01 5.58494151e-01 3.62074040e-02
-5.21207035e-01 4.15553212e-01 9.76622581e-01 -4.06309605e-01
1.21074751e-01 1.97840575e-02 -2.86061734e-01 -2.28175074e-01
5.26119471e-01 2.64838934e-02 8.63860250e-02 1.24089316e-01
7.06793487e-01 1.44401491e-01 -1.77098250e+00 2.43465379e-01
5.58389306e-01 4.54976082e-01 5.39343320e-02 7.82920361e-01
-4.09656018e-01 5.59155405e-01 -6.29200518e-01 -5.70815623e-01
-7.52318725e-02 -5.20668268e-01 -7.78244793e-01 2.03929573e-01
5.80995739e-01 -3.11168462e-01 -9.82224122e-02 7.87599981e-01
9.52284098e-01 9.81365368e-02 5.11320353e-01 -1.34350741e+00
-7.74628222e-01 3.29255968e-01 -2.46345744e-01 2.48903066e-01
6.75845683e-01 4.55511361e-01 5.82146585e-01 -8.40243399e-01
2.79041350e-01 1.12379503e+00 8.33502591e-01 6.54773355e-01
-8.93527567e-01 -6.37689233e-01 -8.71649534e-02 3.74255449e-01
-1.35642171e+00 -1.35585260e+00 3.33600014e-01 -7.21272707e-01
1.00486922e+00 1.59556851e-01 6.40749097e-01 1.10569048e+00
-3.90581280e-01 4.67484951e-01 1.05110228e+00 -8.75750542e-01
7.73307234e-02 6.75470650e-01 1.70207128e-01 3.37978184e-01
7.34402984e-02 9.84762330e-03 -9.02210355e-01 -1.21385984e-01
3.43726039e-01 -4.34155554e-01 -5.99153161e-01 -2.75232419e-02
-1.18483305e+00 6.62137866e-01 -5.12524843e-01 2.62908936e-01
4.21985332e-03 -1.67715102e-01 6.43635571e-01 9.85099524e-02
2.15491131e-01 2.32264340e-01 -2.08375812e-01 -8.55536401e-01
-9.30999219e-01 -1.37330711e-01 1.02465284e+00 6.18386745e-01
2.17718199e-01 -1.46894872e-01 -6.61197603e-02 1.34398425e+00
4.76306945e-01 1.93967268e-01 7.15542555e-01 -9.70867336e-01
3.28956574e-01 7.10739568e-02 7.16734007e-02 -7.15318561e-01
-1.76209286e-01 2.10653335e-01 -1.16122942e-02 3.79989713e-01
6.76155210e-01 -2.50044614e-01 -7.59483039e-01 1.78205204e+00
-1.55164525e-01 -1.90312602e-02 -1.06136419e-01 6.33738995e-01
6.43518686e-01 3.25479865e-01 3.98317575e-01 -3.44808131e-01
1.24641013e+00 -8.09594393e-01 -7.73963869e-01 -4.03963983e-01
2.54153401e-01 -1.23851252e+00 1.20528769e+00 4.71742004e-01
-1.36238933e+00 -4.10966039e-01 -8.99345338e-01 -1.58693165e-01
-3.42269450e-01 -6.81515038e-02 1.65963337e-01 1.01870739e+00
-1.38155878e+00 2.12846413e-01 -3.75829071e-01 -7.81274617e-01
2.56684035e-01 8.71879160e-02 -3.39437336e-01 1.24291189e-01
-7.86813200e-01 1.25619209e+00 -2.19621360e-01 -1.44021615e-01
-1.97431557e-02 -5.74444950e-01 -6.34895980e-01 -1.21561773e-01
1.20582290e-01 -5.49103975e-01 1.37276077e+00 -1.07181084e+00
-1.55403113e+00 1.04066956e+00 -4.48624372e-01 -4.82884273e-02
3.18597257e-01 -2.15696031e-03 -6.13720357e-01 -1.03115171e-01
-1.35529920e-01 7.68666029e-01 7.86560118e-01 -8.48730087e-01
-7.83550620e-01 -2.85781026e-01 -4.52985078e-01 4.44168627e-01
-3.25990945e-01 4.72837657e-01 -2.70318180e-01 -4.19969261e-01
8.74755606e-02 -6.60998940e-01 6.88416243e-01 2.01583996e-01
1.50587320e-01 -3.96398902e-01 6.56429887e-01 -9.98613775e-01
9.49950933e-01 -2.23794460e+00 -2.08863258e-01 -2.88361788e-01
1.30943879e-01 3.22645813e-01 6.05577566e-02 5.55540383e-01
1.54584497e-01 4.39818323e-01 1.96190596e-01 -4.01791483e-01
1.89120352e-01 -1.61088213e-01 -2.09354132e-01 2.52679676e-01
-1.65187031e-01 6.50067449e-01 -6.09709561e-01 -5.27663291e-01
3.18070352e-01 8.50809574e-01 -1.48855582e-01 -2.38230214e-01
3.82703722e-01 -6.86393306e-02 3.04118276e-01 7.03548253e-01
3.68782312e-01 -2.86592124e-03 3.47390361e-02 1.78149998e-01
-4.49176908e-01 6.49883151e-01 -1.00786853e+00 1.35601759e+00
-4.36771780e-01 1.35913241e+00 3.60588729e-01 -5.25517344e-01
6.23272300e-01 6.03237689e-01 1.78685606e-01 -3.50922734e-01
9.33833420e-02 2.73019522e-01 1.01387985e-01 -8.49902511e-01
4.21813160e-01 -3.13852191e-01 5.52120566e-01 4.41329867e-01
-1.02590926e-01 -3.55743557e-01 -1.73348859e-01 -5.34869954e-02
6.09940171e-01 -2.06386879e-01 1.74164876e-01 2.36080773e-02
1.89998493e-01 -3.50918561e-01 1.71516404e-01 4.93478328e-01
-8.60492349e-01 5.95819294e-01 3.01917374e-01 2.71790117e-01
-9.31431830e-01 -1.15484953e+00 -3.84325206e-01 1.13854468e+00
-1.41132906e-01 -2.64983296e-01 -7.03018606e-01 -1.30158424e-01
-2.29402706e-02 7.52217948e-01 -2.48082966e-01 -1.30996376e-01
3.69477645e-02 -8.04068893e-02 8.55155528e-01 4.86754119e-01
1.71979725e-01 -1.14365411e+00 -6.62223816e-01 -4.02022190e-02
-3.69633228e-01 -9.37068999e-01 -7.67097592e-01 -3.06309909e-01
-5.09830296e-01 -8.37820649e-01 -9.70315218e-01 -1.05985296e+00
-4.70332056e-02 5.09561360e-01 8.45456958e-01 5.61527461e-02
-1.22331411e-01 1.05257356e+00 -1.51709691e-01 -6.86113536e-01
-8.48996520e-01 -1.45471111e-01 6.80853501e-02 -3.74614656e-01
6.93412840e-01 -2.82110244e-01 -2.55885422e-01 3.41338366e-01
-5.59526026e-01 1.74073756e-01 4.48760152e-01 6.22843325e-01
-6.88075423e-02 -4.89327848e-01 7.63059556e-01 -1.11872323e-01
1.07676721e+00 -3.96390557e-01 4.85594152e-03 4.63896185e-01
-8.79053831e-01 -4.69269812e-01 -2.14707553e-01 -6.53175235e-01
-8.45584452e-01 -5.79438388e-01 -3.73380296e-02 -1.97466444e-02
-4.47321087e-01 5.68160117e-02 -3.83135974e-02 -1.38136119e-01
6.06794953e-01 1.57216266e-01 6.35896087e-01 -3.29778790e-01
1.68160558e-01 1.23977613e+00 4.01754856e-01 -2.26552397e-01
1.01081878e-01 8.93585682e-02 -7.39895046e-01 -1.20365715e+00
1.88431725e-01 -5.32754719e-01 -2.24403784e-01 -5.16184568e-01
6.59935713e-01 -6.90201044e-01 -7.06437886e-01 8.04465652e-01
-9.44423556e-01 -4.03503329e-01 -3.09575379e-01 8.80535841e-01
-4.67649013e-01 6.06582701e-01 -4.33381379e-01 -9.76129174e-01
-1.29435003e-01 -1.29997778e+00 6.41020775e-01 4.89313304e-01
-7.10783601e-01 -9.20155108e-01 1.18390605e-01 8.74914348e-01
7.51736403e-01 -4.97630388e-01 8.63335669e-01 -3.66405517e-01
-1.10798597e-01 -5.60746118e-02 -3.30352992e-01 6.61867976e-01
5.77395797e-01 1.35967791e-01 -1.17089212e+00 -3.26045342e-02
-9.26019475e-02 -5.59858799e-01 5.41946709e-01 5.49086630e-01
6.20261967e-01 -3.08807224e-01 -2.68189669e-01 1.18176844e-02
1.00904572e+00 4.01098341e-01 6.29268467e-01 2.05972552e-01
1.60396904e-01 9.87739444e-01 5.09442016e-02 7.78029040e-02
5.70444942e-01 7.25525796e-01 -2.47041136e-01 8.75054076e-02
-5.72775781e-01 -3.38412583e-01 5.32892644e-01 7.33115554e-01
1.90014437e-01 -1.69732630e-01 -1.21580791e+00 7.46948481e-01
-1.23065531e+00 -1.15274668e+00 1.82575360e-01 2.47439480e+00
7.32674539e-01 7.69706666e-02 4.33903128e-01 3.17459792e-01
9.62295949e-01 5.89615628e-02 -3.95099491e-01 -6.53175652e-01
1.23000763e-01 8.36110786e-02 2.32901856e-01 9.38925326e-01
-4.52431470e-01 7.59991586e-01 7.50937176e+00 4.98985589e-01
-1.39964056e+00 6.30922765e-02 4.89151746e-01 -1.81498885e-01
-1.29287183e-01 -4.74075079e-02 -5.52894354e-01 5.45780003e-01
8.26704621e-01 -3.20899874e-01 6.39356852e-01 2.88359165e-01
4.18664843e-01 -4.83339816e-01 -7.75112569e-01 1.10560775e+00
6.03807747e-01 -9.93673563e-01 -3.12770665e-01 -1.47200644e-01
1.51549235e-01 3.36275369e-01 2.63998061e-01 -5.16038761e-02
-2.06034616e-01 -1.03207064e+00 7.27847517e-01 5.11708498e-01
8.64903390e-01 -1.46683946e-01 2.04382718e-01 6.26522303e-02
-6.85181975e-01 -5.15434705e-02 6.46491274e-02 -1.58107653e-01
9.35211927e-02 -1.19999118e-01 -1.27303183e+00 -2.33011708e-01
4.72005129e-01 2.61258453e-01 -2.85762906e-01 1.25056875e+00
9.26048309e-02 5.90724945e-01 -1.29229352e-01 -2.90745378e-01
-3.40692341e-01 9.60929915e-02 5.29808283e-01 1.24789512e+00
1.61392674e-01 -1.35701045e-01 -3.35266739e-01 3.10873270e-01
2.25397885e-01 3.32940429e-01 -6.16328001e-01 -2.63553888e-01
7.17133820e-01 9.40339029e-01 -4.90623772e-01 -3.81733254e-02
-6.86688244e-01 6.45932496e-01 -5.33394888e-02 4.97314364e-01
-5.16226113e-01 -3.66954058e-01 8.33889544e-01 3.84332031e-01
-2.04525232e-01 -2.09935486e-01 -5.53884327e-01 -6.68880105e-01
8.93598422e-02 -1.09942091e+00 1.43571962e-02 -1.03691304e+00
-9.32691455e-01 1.15845114e-01 -5.43595143e-02 -6.62190616e-01
-3.15441728e-01 -4.55105603e-01 -4.52781796e-01 1.07585979e+00
-1.23493803e+00 -6.81573153e-01 -1.74138963e-01 2.73159534e-01
6.78407013e-01 -2.13281840e-01 7.81234801e-01 1.38613641e-01
-2.10013539e-01 9.44730878e-01 -1.84098426e-02 -1.79562382e-02
1.20538592e+00 -5.57244837e-01 2.95923918e-01 2.43519738e-01
1.64364234e-01 5.38300574e-01 5.07752240e-01 -1.52929097e-01
-1.06109846e+00 -2.38205627e-01 1.25441504e+00 -5.62916934e-01
4.27528173e-01 -3.27067412e-02 -7.11578369e-01 5.36063373e-01
4.45992321e-01 -8.55730474e-01 9.36927676e-01 3.65605235e-01
-3.80454659e-01 -1.21592628e-02 -1.24535382e+00 7.62960076e-01
9.07308280e-01 -1.01103616e+00 -5.88092268e-01 1.27581194e-01
3.85066330e-01 -1.91418201e-01 -4.79758292e-01 1.75581127e-01
1.06640458e+00 -1.10115886e+00 7.26262510e-01 -2.89425552e-01
-5.53709120e-02 1.53402090e-01 -1.24484211e-01 -9.87791657e-01
8.41066912e-02 -7.59585559e-01 4.54476863e-01 1.45001948e+00
5.04423082e-01 -9.45540130e-01 7.04346657e-01 8.66628528e-01
-1.19467326e-01 -6.90024495e-01 -1.10195994e+00 -3.65030885e-01
2.07983673e-01 -6.75804079e-01 3.08801979e-01 8.62796068e-01
6.76193833e-01 2.70188302e-01 -1.20697424e-01 -2.48706684e-01
2.12911144e-01 -7.18541443e-01 7.16117501e-01 -1.04550695e+00
-2.63242006e-01 -9.61949050e-01 -6.40126288e-01 -8.01744759e-01
-8.01918209e-02 -5.53375423e-01 -1.95443835e-02 -1.37113833e+00
3.14315915e-01 -1.58389404e-01 3.79314902e-03 3.81677270e-01
-7.23244995e-02 1.09346092e-01 4.08588618e-01 1.22712374e-01
2.33887676e-02 2.76454743e-02 8.36622834e-01 -2.92556249e-02
-3.93270224e-01 1.56556204e-01 -1.05692363e+00 6.90323651e-01
9.81566489e-01 1.57764982e-02 -5.93437850e-01 -4.60116386e-01
-1.25600517e-01 8.54059458e-02 4.95519936e-01 -1.14608753e+00
5.01033723e-01 -1.32339567e-01 -3.71052586e-02 -2.47708902e-01
7.56646574e-01 -3.07130784e-01 7.97744468e-02 1.95511892e-01
-5.32383740e-01 6.07218891e-02 5.21034181e-01 1.76678345e-01
-3.48257162e-02 -3.33315760e-01 8.20557535e-01 5.05664609e-02
-5.59297979e-01 -2.36715928e-01 -6.04206085e-01 3.39155942e-01
6.67469144e-01 -8.00567567e-01 -2.75850892e-01 -9.45815980e-01
-6.27618611e-01 -1.10360824e-01 7.55128920e-01 6.25181496e-01
5.25353968e-01 -7.84205735e-01 -5.93971729e-01 2.97196209e-01
1.77007187e-02 -8.11915517e-01 9.12364423e-02 9.56892967e-01
-1.85882933e-02 4.85320061e-01 -1.25180244e-01 -2.99141020e-01
-1.74119329e+00 -1.62873622e-02 4.02311981e-01 5.58150232e-01
-3.63368124e-01 9.80791211e-01 -2.09632680e-01 -7.12428764e-02
8.38748693e-01 -1.86511669e-02 9.56198499e-02 2.18981698e-01
5.47552168e-01 7.91152418e-01 -5.25663272e-02 -5.44074953e-01
-2.32189506e-01 5.05763531e-01 -3.28709073e-02 -8.04167807e-01
5.95761776e-01 -3.31091344e-01 1.48509204e-01 7.98267186e-01
1.02543867e+00 4.30325031e-01 -1.13299680e+00 1.56152785e-01
-1.68040544e-01 -4.13800508e-01 -1.16594456e-01 -1.05160475e+00
-3.14265281e-01 1.04350030e+00 7.28295505e-01 2.00550541e-01
9.43120420e-01 1.32532403e-01 4.51593935e-01 -1.24586217e-01
3.35151702e-02 -1.48964870e+00 -1.43293634e-01 2.68213958e-01
8.37266684e-01 -1.10423970e+00 -2.15859354e-01 -1.51895627e-01
-7.83773482e-01 9.79305148e-01 2.24725291e-01 8.84223521e-01
6.03804231e-01 1.68123960e-01 5.63487172e-01 3.05914998e-01
-5.37245452e-01 -1.40552759e-01 4.09849249e-02 9.38942432e-01
7.46095896e-01 4.88449670e-02 -5.88523924e-01 1.36347741e-01
-2.65289783e-01 3.36405843e-01 4.17381197e-01 5.48310101e-01
-5.99910259e-01 -1.23518062e+00 -5.09627223e-01 4.70110744e-01
-4.02995139e-01 -1.89548850e-01 -7.08765924e-01 8.95487845e-01
9.14849043e-02 1.37839079e+00 -4.37077880e-02 -5.09415388e-01
4.95937690e-02 7.38174140e-01 4.44766849e-01 -5.03723741e-01
-2.85455734e-01 -9.51003134e-02 2.46622995e-01 6.93363079e-04
-1.23261802e-01 -1.16872334e+00 -8.12269807e-01 -4.83827114e-01
-5.16942441e-01 5.33854067e-02 1.15912056e+00 7.81213880e-01
5.12292266e-01 -1.25424787e-01 3.44574511e-01 -4.62030947e-01
-6.76776826e-01 -1.17314625e+00 -3.58319938e-01 6.60777315e-02
5.09571433e-01 -4.70706880e-01 -6.08731329e-01 -2.10181866e-02] | [14.282767295837402, 4.962830066680908] |
151eb1fd-4a1d-450f-bb94-88114367269f | collaborative-and-distributed-bayesian | 2306.14348 | null | https://arxiv.org/abs/2306.14348v1 | https://arxiv.org/pdf/2306.14348v1.pdf | Collaborative and Distributed Bayesian Optimization via Consensus: Showcasing the Power of Collaboration for Optimal Design | Optimal design is a critical yet challenging task within many applications. This challenge arises from the need for extensive trial and error, often done through simulations or running field experiments. Fortunately, sequential optimal design, also referred to as Bayesian optimization when using surrogates with a Bayesian flavor, has played a key role in accelerating the design process through efficient sequential sampling strategies. However, a key opportunity exists nowadays. The increased connectivity of edge devices sets forth a new collaborative paradigm for Bayesian optimization. A paradigm whereby different clients collaboratively borrow strength from each other by effectively distributing their experimentation efforts to improve and fast-track their optimal design process. To this end, we bring the notion of consensus to Bayesian optimization, where clients agree (i.e., reach a consensus) on their next-to-sample designs. Our approach provides a generic and flexible framework that can incorporate different collaboration mechanisms. In lieu of this, we propose transitional collaborative mechanisms where clients initially rely more on each other to maneuver through the early stages with scant data, then, at the late stages, focus on their own objectives to get client-specific solutions. Theoretically, we show the sub-linear growth in regret for our proposed framework. Empirically, through simulated datasets and a real-world collaborative material discovery experiment, we show that our framework can effectively accelerate and improve the optimal design process and benefit all participants. | ['Blake N. Johnson', 'Kevin Edgar', 'Zhenghao Zai', 'Yang Liu', 'Albert S. Berahas', 'Raed Al Kontar', 'Xubo Yue'] | 2023-06-25 | null | null | null | null | ['bayesian-optimization'] | ['methodology'] | [-9.77956578e-02 -1.60372213e-01 -3.32335502e-01 -3.44912708e-01
-8.57280910e-01 -6.38082743e-01 1.04415379e-01 -1.26073271e-01
-3.66613179e-01 7.32315898e-01 1.24363415e-01 -4.80849177e-01
-7.02126980e-01 -6.35975122e-01 -7.08190322e-01 -8.40104401e-01
-6.22749180e-02 6.53824389e-01 -2.88420856e-01 2.57712543e-01
2.71344692e-01 3.03991139e-01 -1.07130778e+00 -1.10259190e-01
9.61128473e-01 9.73875284e-01 1.78533733e-01 4.51626033e-01
1.47649525e-02 2.84701109e-01 -2.00330928e-01 -5.32931387e-01
4.06713247e-01 -1.59842476e-01 -3.76396626e-01 5.88103272e-02
-2.47611612e-01 -5.74073792e-01 -9.81797948e-02 9.57710385e-01
8.04562986e-01 6.91567957e-02 3.32360476e-01 -1.30467951e+00
-2.39191920e-01 1.07289386e+00 -9.64651346e-01 -1.44281790e-01
3.69704127e-01 3.43696505e-01 1.35239768e+00 -7.47477174e-01
3.95441830e-01 1.17921066e+00 5.67529976e-01 2.85199523e-01
-1.41802955e+00 -8.09391975e-01 6.00117683e-01 2.00318083e-01
-1.47301567e+00 -4.33428675e-01 7.96725273e-01 -1.40690133e-01
1.12202361e-01 4.03614074e-01 6.74394906e-01 1.14138663e+00
-1.71238139e-01 1.06679249e+00 1.03042948e+00 -1.77786067e-01
7.63557017e-01 2.05814108e-01 9.54474062e-02 2.92115837e-01
5.66284180e-01 2.27345690e-01 -7.33407557e-01 -4.54988271e-01
4.82512653e-01 3.11937600e-01 -4.83298182e-01 -5.27224898e-01
-8.01053822e-01 6.73063934e-01 2.99796849e-01 -2.50849605e-01
-7.55700648e-01 3.42933863e-01 -5.78232370e-02 1.26494095e-01
1.08420938e-01 9.07567963e-02 -3.31958562e-01 -3.25564295e-01
-8.13130856e-01 3.82485509e-01 9.45840955e-01 8.27830017e-01
4.98203009e-01 -5.90700269e-01 -3.45552176e-01 5.41781485e-01
6.81434810e-01 5.06572485e-01 -2.87132621e-01 -1.12795365e+00
6.83574855e-01 4.50548083e-01 6.22055769e-01 -9.26549733e-01
-9.81736779e-02 -7.07439065e-01 -8.32041144e-01 1.95399627e-01
5.68713486e-01 -4.98699337e-01 -3.94226611e-01 1.84065795e+00
7.02093840e-01 3.03765923e-01 -2.08226874e-01 1.20242786e+00
-1.84815004e-01 4.32596147e-01 -1.54765785e-01 -4.81241286e-01
1.19952214e+00 -4.95715618e-01 -5.09820938e-01 -3.52937579e-02
2.03531295e-01 -6.11905992e-01 8.33012223e-01 7.25879371e-01
-1.01584864e+00 1.48524731e-01 -9.10168588e-01 6.75876200e-01
2.15667784e-01 -2.33274400e-01 7.83514261e-01 1.04715776e+00
-7.17030704e-01 5.51181912e-01 -6.92252755e-01 -1.30094558e-01
8.70309711e-01 5.13192713e-01 2.03775555e-01 -4.44746703e-01
-6.41802549e-01 2.90333629e-01 -2.68394798e-02 4.32577908e-01
-9.24725473e-01 -1.02336287e+00 2.20976118e-02 2.77964801e-01
1.03998053e+00 -1.11216581e+00 1.33247554e+00 -5.64480007e-01
-1.79411626e+00 9.37741920e-02 -1.59976985e-02 -2.38204435e-01
9.38842118e-01 -3.39306682e-01 3.29147935e-01 -3.36611658e-01
-2.59725004e-01 1.71839207e-01 7.02301443e-01 -1.21165955e+00
-7.45052695e-01 -5.19191384e-01 2.94076413e-01 2.25350097e-01
-5.58847606e-01 3.84528711e-02 -5.28825104e-01 -3.31743330e-01
1.00968182e-01 -1.10262084e+00 -6.74476087e-01 2.39338607e-01
-3.39475304e-01 -1.51993766e-01 4.34129626e-01 -1.21930309e-01
1.29991102e+00 -1.90332031e+00 2.00862780e-01 6.74620688e-01
4.44228321e-01 -1.51650563e-01 1.78572357e-01 4.28430647e-01
5.21707535e-01 2.67860085e-01 8.51953700e-02 -4.76139694e-01
3.40240866e-01 9.75248665e-02 -1.06070541e-01 4.27028090e-01
-2.60292083e-01 6.94509029e-01 -9.45970595e-01 -3.17575812e-01
-3.12549807e-02 2.13943988e-01 -9.86800194e-01 2.39632934e-01
-2.47204319e-01 5.91351271e-01 -1.02595246e+00 6.74025118e-01
9.01643515e-01 -6.73855841e-01 6.22030139e-01 -1.81802665e-03
-6.21840246e-02 -9.16982144e-02 -1.84842753e+00 1.53130615e+00
-6.89077795e-01 1.84184015e-01 6.95154130e-01 -1.09187269e+00
4.17389363e-01 2.38987237e-01 6.08382761e-01 -3.36336404e-01
2.51760989e-01 4.36852604e-01 -9.83390361e-02 -2.28565112e-01
1.48993656e-01 2.67408311e-01 3.55936848e-02 9.42929447e-01
-6.28681839e-01 3.41271043e-01 -1.32409632e-01 1.18237250e-01
1.18221498e+00 -1.16437435e-01 1.03631482e-01 -7.20532760e-02
1.51213497e-01 -3.28934669e-01 7.49431610e-01 1.11918354e+00
-3.18182670e-02 2.17320547e-01 4.98100638e-01 -2.46930540e-01
-7.36636102e-01 -1.05463910e+00 6.31223768e-02 8.76927674e-01
2.64382660e-01 -2.33861968e-01 -5.02038836e-01 -6.33995116e-01
2.85307348e-01 3.76863003e-01 -4.92207468e-01 8.48224759e-02
-1.85678884e-01 -8.36578786e-01 -1.73780039e-01 3.90295297e-01
4.54727381e-01 -4.71066117e-01 -7.23549128e-01 4.54867810e-01
1.85980275e-02 -6.71590149e-01 -6.54562831e-01 7.32479319e-02
-9.76514220e-01 -7.89701283e-01 -8.27658772e-01 1.03435032e-01
7.30402231e-01 4.76540416e-01 7.77262628e-01 -1.94638688e-02
-2.22397074e-01 3.56533259e-01 -2.75224775e-01 -3.45088869e-01
1.06475867e-01 4.45519090e-02 1.08168595e-01 3.58429492e-01
-1.38672069e-01 -9.14510906e-01 -1.07607472e+00 7.36859679e-01
-5.64885736e-01 2.68126950e-02 8.78111124e-01 6.71237886e-01
3.64686221e-01 3.61723267e-02 5.32017469e-01 -7.05579042e-01
7.73677289e-01 -6.91822648e-01 -1.06003904e+00 4.03502822e-01
-8.81738305e-01 2.91310281e-01 4.02475655e-01 -5.95225394e-01
-1.03280151e+00 -1.03769891e-01 3.36217701e-01 -5.04495919e-01
3.48248243e-01 7.57115543e-01 -4.27774251e-01 -3.95518029e-03
4.39665735e-01 -2.36310095e-01 -1.45205662e-01 -5.52947760e-01
4.72146213e-01 9.76216733e-01 3.68566960e-02 -1.05370820e+00
7.23312795e-01 4.90062594e-01 1.50487227e-02 -2.02850923e-01
-7.79241860e-01 -2.18706474e-01 1.35119529e-02 -5.68671644e-01
2.53861755e-01 -6.79547668e-01 -1.60170388e+00 2.70452559e-01
-9.51755166e-01 -3.06467772e-01 -4.30589989e-02 5.96454084e-01
-1.58091187e-01 7.30894953e-02 -9.45978463e-02 -1.35225439e+00
-2.79900968e-01 -1.07155299e+00 7.03949511e-01 6.91163778e-01
-9.28919315e-02 -6.81322217e-01 -5.31999618e-02 5.43235779e-01
5.64689696e-01 -5.82223237e-02 3.97794574e-01 -3.96745652e-01
-1.26755500e+00 -2.42553547e-01 -2.94306040e-01 -2.15814948e-01
2.35797977e-03 -6.62117004e-02 -6.66478217e-01 -3.36663693e-01
-1.97157457e-01 -2.93133333e-02 3.41588765e-01 5.43939650e-01
1.18586826e+00 -4.25952494e-01 -5.86904407e-01 4.63205248e-01
1.30680358e+00 1.85466751e-01 2.15353578e-01 1.38074070e-01
3.46827239e-01 5.09400427e-01 5.88890910e-01 1.18216908e+00
4.66476500e-01 9.43489075e-01 4.61884767e-01 2.99511552e-01
4.10817802e-01 -2.73834348e-01 1.83822915e-01 3.67021799e-01
-5.56954294e-02 -3.87460172e-01 -7.19008029e-01 2.63439506e-01
-2.31229782e+00 -6.98297799e-01 -7.96610583e-03 2.56242490e+00
8.07475448e-01 1.61613211e-01 2.15403333e-01 -7.78734684e-02
7.50360906e-01 -2.07324475e-01 -8.60413134e-01 -1.78550575e-02
3.29524368e-01 -9.76644754e-02 7.45525718e-01 2.21586138e-01
-4.87865597e-01 3.32880020e-01 5.58964968e+00 9.13230121e-01
-8.17120969e-01 1.27678126e-01 8.74467671e-01 -4.80260938e-01
-4.46367025e-01 3.12854499e-01 -7.59967268e-01 7.17392683e-01
8.37768078e-01 -4.86739159e-01 9.60614681e-01 8.19519579e-01
7.28684723e-01 -3.35754186e-01 -1.20840216e+00 1.13474202e+00
-4.81354713e-01 -1.49034464e+00 -5.73788941e-01 6.00406706e-01
9.02799070e-01 -3.28288317e-01 3.28027904e-02 3.25967334e-02
8.31650972e-01 -7.63512492e-01 7.06848681e-01 5.97385764e-01
2.33050987e-01 -7.89104223e-01 6.29698396e-01 4.31875318e-01
-1.07841182e+00 -3.81988436e-01 -9.38632637e-02 -3.76797356e-02
6.05459452e-01 1.02175581e+00 -6.66044056e-01 7.72499859e-01
6.92030251e-01 3.05176377e-01 9.89343449e-02 1.53098607e+00
-1.18888326e-01 5.40705919e-01 -7.06459403e-01 -3.83456081e-01
-6.31806105e-02 -4.98089641e-01 5.16716182e-01 4.67004269e-01
4.97285664e-01 1.88495874e-01 3.93306375e-01 9.61162984e-01
-2.11526275e-01 6.78461120e-02 -1.55464724e-01 -7.63971433e-02
8.99727881e-01 1.21462822e+00 -6.51838183e-01 -5.74140288e-02
-1.93242669e-01 5.20054042e-01 1.11157916e-01 3.52357298e-01
-1.04644859e+00 5.90250967e-03 7.85153449e-01 -1.53535455e-01
3.54697168e-01 -1.68488249e-01 -6.39084637e-01 -7.73501337e-01
2.98024952e-01 -7.55564511e-01 3.93060178e-01 -4.30649877e-01
-1.34054244e+00 -1.60484165e-01 -2.73212716e-02 -9.83570993e-01
1.01984091e-01 -2.81266302e-01 -7.10587323e-01 8.20282519e-01
-1.12288952e+00 -7.06556082e-01 -3.15833539e-01 3.23357642e-01
3.17822218e-01 2.79028326e-01 2.93045551e-01 4.73120332e-01
-8.30340028e-01 7.22343981e-01 4.14879531e-01 -3.83807421e-01
5.09712994e-01 -9.21010792e-01 -1.89500540e-01 5.96465349e-01
6.83862418e-02 7.66006172e-01 8.72825265e-01 -5.62515795e-01
-1.91781294e+00 -5.59389532e-01 2.58499116e-01 -1.29259348e-01
7.44110227e-01 -2.98534244e-01 -4.72591937e-01 1.61996186e-01
-7.22714961e-02 -1.63100258e-01 6.67392492e-01 5.72223544e-01
5.77247068e-02 -5.64598918e-01 -1.03356886e+00 9.87927020e-01
1.30664492e+00 -9.73516256e-02 1.09993868e-01 3.27513814e-01
4.23948526e-01 -1.78579852e-01 -7.21571147e-01 2.98423737e-01
8.56579006e-01 -9.00898457e-01 7.01745152e-01 -4.63065267e-01
-6.12606816e-02 -3.16385865e-01 -1.83020681e-01 -1.19337904e+00
9.80373174e-02 -1.35459375e+00 -2.85845160e-01 1.27708733e+00
4.65493262e-01 -7.87686586e-01 1.23790121e+00 1.04749405e+00
2.21174404e-01 -9.16218996e-01 -9.56820846e-01 -7.38906443e-01
-3.08043540e-01 -8.07306707e-01 6.85666025e-01 3.48557681e-01
-2.23060548e-01 7.30013922e-02 -5.72590768e-01 3.59286338e-01
1.04345715e+00 3.02980572e-01 1.06698394e+00 -1.24831998e+00
-7.39897490e-01 -5.66811264e-01 1.61667019e-01 -1.44530654e+00
-3.05631220e-01 -4.87587601e-01 8.54739696e-02 -1.24166083e+00
5.68963706e-01 -8.79469573e-01 -3.22338551e-01 2.45144352e-01
-2.42659464e-01 -2.44982690e-01 1.03086405e-01 2.39149526e-01
-8.17565620e-01 5.46955705e-01 1.01146734e+00 -1.72276061e-03
-4.65967506e-01 5.88015437e-01 -1.00106049e+00 2.85274267e-01
4.68886822e-01 -7.30343521e-01 -5.06182313e-01 -1.99847013e-01
6.63529992e-01 3.12084347e-01 3.45536828e-01 -6.25559270e-01
5.42876065e-01 -4.92674500e-01 -1.42923459e-01 -4.17830586e-01
8.88315663e-02 -1.06279516e+00 6.65535867e-01 4.97976810e-01
-1.14528529e-01 -2.50505209e-01 -3.29685032e-01 1.10475397e+00
4.27430123e-01 -2.04609588e-01 4.67650712e-01 1.77341670e-01
9.20209754e-03 4.37999010e-01 -2.12354183e-01 -2.98584122e-02
1.12829924e+00 -1.13763593e-01 -8.42918679e-02 -5.91871381e-01
-6.75116301e-01 8.14726412e-01 4.02970046e-01 5.17232753e-02
2.96815783e-01 -1.17747378e+00 -6.43547118e-01 -2.67063439e-01
-1.81141719e-01 -4.37768959e-02 4.22304034e-01 1.38694191e+00
-1.18705193e-02 1.40905574e-01 2.95474231e-01 -8.31511557e-01
-1.05977964e+00 3.68184537e-01 1.43119633e-01 -5.25691688e-01
-3.48339081e-01 6.77203238e-01 -8.21531266e-02 -1.49841709e-02
5.15912771e-01 -1.49791613e-01 3.61608237e-01 1.51920810e-01
4.95637089e-01 5.58760822e-01 -1.41920477e-01 3.52992117e-01
-4.18355286e-01 4.28332537e-01 -8.85465741e-02 -4.51444596e-01
1.51063156e+00 -3.83327276e-01 1.81272656e-01 1.03041917e-01
8.33926380e-01 2.19618157e-01 -1.53030753e+00 -3.48115444e-01
1.82344034e-01 -9.43260610e-01 3.16315293e-01 -8.19366038e-01
-1.19796503e+00 2.76279837e-01 6.44446850e-01 1.33350149e-01
1.00184858e+00 -1.46356449e-02 5.13310671e-01 4.41000372e-01
7.83375502e-01 -1.10217392e+00 1.07992455e-01 -1.26355037e-01
4.17033911e-01 -1.16363800e+00 6.09698407e-02 -2.89625257e-01
-3.70375127e-01 7.06725061e-01 1.32486850e-01 1.63348556e-01
7.04640627e-01 1.72440886e-01 -4.07180727e-01 1.97140183e-02
-9.45330024e-01 -1.65725891e-02 -1.42796099e-01 2.62144893e-01
-7.29594529e-02 1.79520354e-01 -3.51926506e-01 9.17094290e-01
3.50899100e-01 2.20789567e-01 3.03033948e-01 9.92542624e-01
-2.74333894e-01 -1.51474833e+00 -6.63099051e-01 5.53773761e-01
-3.80403906e-01 1.92728266e-01 -5.67708462e-02 5.39672196e-01
-2.55127519e-01 1.07956707e+00 -2.87413806e-01 -1.92063212e-01
3.35973173e-01 -2.51857072e-01 3.57010812e-01 -3.21324229e-01
-4.55985785e-01 1.52810380e-01 -1.24412905e-02 -6.45352006e-01
-1.74118459e-01 -9.05772924e-01 -7.04927266e-01 -6.88892245e-01
-5.11765420e-01 3.26893091e-01 9.22877312e-01 9.38928902e-01
6.02177501e-01 3.22554320e-01 1.17509651e+00 -8.78783762e-01
-1.02647161e+00 -5.93060613e-01 -4.13372338e-01 -6.50938824e-02
-5.36563508e-02 -8.51874769e-01 -4.70527798e-01 -5.00176907e-01] | [4.893179893493652, 3.4494359493255615] |
8e61c68e-92c3-4d09-801b-c3809f87ac2a | stain-normalized-breast-histopathology-image | 2201.00957 | null | https://arxiv.org/abs/2201.00957v1 | https://arxiv.org/pdf/2201.00957v1.pdf | Stain Normalized Breast Histopathology Image Recognition using Convolutional Neural Networks for Cancer Detection | Computer assisted diagnosis in digital pathology is becoming ubiquitous as it can provide more efficient and objective healthcare diagnostics. Recent advances have shown that the convolutional Neural Network (CNN) architectures, a well-established deep learning paradigm, can be used to design a Computer Aided Diagnostic (CAD) System for breast cancer detection. However, the challenges due to stain variability and the effect of stain normalization with such deep learning frameworks are yet to be well explored. Moreover, performance analysis with arguably more efficient network models, which may be important for high throughput screening, is also not well explored.To address this challenge, we consider some contemporary CNN models for binary classification of breast histopathology images that involves (1) the data preprocessing with stain normalized images using an adaptive colour deconvolution (ACD) based color normalization algorithm to handle the stain variabilities; and (2) applying transfer learning based training of some arguably more efficient CNN models, namely Visual Geometry Group Network (VGG16), MobileNet and EfficientNet. We have validated the trained CNN networks on a publicly available BreaKHis dataset, for 200x and 400x magnified histopathology images. The experimental analysis shows that pretrained networks in most cases yield better quality results on data augmented breast histopathology images with stain normalization, than the case without stain normalization. Further, we evaluated the performance and efficiency of popular lightweight networks using stain normalized images and found that EfficientNet outperforms VGG16 and MobileNet in terms of test accuracy and F1 Score. We observed that efficiency in terms of test time is better in EfficientNet than other networks; VGG Net, MobileNet, without much drop in the classification accuracy. | ['Arnav Bhavsar', 'Shivsubramani Krishnamoorthy', 'Suganthi S. S', 'Sruthi Krishna'] | 2022-01-04 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 2.12451711e-01 2.74973810e-01 1.30578503e-01 -1.83599219e-01
-6.54849768e-01 -3.81981999e-01 2.95591921e-01 2.99787879e-01
-6.58198655e-01 5.90462863e-01 -1.87114149e-01 -8.43282461e-01
-3.18138480e-01 -7.43822575e-01 -5.05492628e-01 -9.90795255e-01
-3.60327542e-01 5.25837600e-01 2.48642430e-01 -1.25193745e-01
-1.01124784e-02 1.06628180e+00 -1.32400537e+00 3.30799520e-01
4.00426090e-01 1.16525137e+00 -1.38324648e-01 1.28497779e+00
6.93737119e-02 8.59540641e-01 -3.48716259e-01 -4.88645792e-01
2.06559867e-01 -2.10508198e-01 -1.04072690e+00 -1.68204382e-01
2.39103779e-01 -4.47282612e-01 2.22974066e-02 8.62671077e-01
8.05247724e-01 -3.90198082e-01 6.72232747e-01 -9.64778960e-01
-5.31210244e-01 2.09170461e-01 -4.30182964e-01 5.83556175e-01
-3.68735790e-01 4.37058151e-01 3.94509733e-01 -6.11397445e-01
9.28379059e-01 1.03752065e+00 1.17243671e+00 6.30386055e-01
-9.67050374e-01 -5.35889387e-01 -6.68542266e-01 1.82593763e-01
-1.15890062e+00 -7.71212205e-02 3.19238126e-01 -2.78412431e-01
9.81910825e-01 3.53361785e-01 8.47520530e-01 6.04699910e-01
6.29823446e-01 2.49437854e-01 1.09442592e+00 -6.10643148e-01
1.49769887e-01 2.13782862e-01 -4.45236132e-04 1.09284961e+00
5.24181068e-01 2.54067183e-01 5.58610484e-02 -1.32157251e-01
9.45208073e-01 1.14085592e-01 -2.08877891e-01 -2.40667149e-01
-1.01536942e+00 9.17315304e-01 8.35961401e-01 5.47475994e-01
-1.11344472e-01 2.92713076e-01 6.58530295e-01 3.80716383e-01
5.78983903e-01 5.56211710e-01 -4.13654208e-01 1.29537061e-01
-7.04392731e-01 -9.83346137e-04 6.26521051e-01 3.85281205e-01
7.11287379e-01 -9.51338932e-02 -1.66591287e-01 7.68864930e-01
2.38795683e-01 2.88018942e-01 4.10749793e-01 -6.32853150e-01
-1.01317674e-01 9.20199990e-01 -4.85487729e-01 -9.17528629e-01
-9.24992263e-01 -6.72514617e-01 -1.26669955e+00 5.22835910e-01
5.82320929e-01 -7.55679086e-02 -1.33715189e+00 1.16815972e+00
3.76080304e-01 2.11170316e-02 7.00273216e-02 9.61238563e-01
1.04685712e+00 1.24012552e-01 3.55168581e-01 2.28059173e-01
1.45379305e+00 -5.76470971e-01 -5.97674131e-01 2.20714331e-01
1.38034391e+00 -7.88620591e-01 7.27935791e-01 2.46497706e-01
-1.01240027e+00 -2.82357275e-01 -1.31516254e+00 -2.86585480e-01
-8.19075465e-01 2.66692191e-01 9.47524905e-01 9.06525314e-01
-1.50907743e+00 7.20535874e-01 -1.07925475e+00 -8.22797596e-01
1.05108082e+00 8.06301117e-01 -7.07058847e-01 -4.10840511e-01
-7.03124702e-01 8.51559699e-01 2.02124178e-01 1.80172041e-01
-7.85462379e-01 -1.16381299e+00 -7.63748229e-01 -8.14557895e-02
-1.61328793e-01 -6.28026426e-01 1.10918069e+00 -9.68757451e-01
-1.20848596e+00 1.32355368e+00 4.53786880e-01 -5.20097196e-01
5.80325007e-01 5.27496040e-01 -2.54098356e-01 4.22735274e-01
-2.88314134e-01 9.24255013e-01 6.70481101e-03 -9.72691238e-01
-5.00107884e-01 -4.40374196e-01 -1.80694118e-01 -9.32435989e-02
-4.04774338e-01 -5.16995192e-02 -3.30627918e-01 -4.05451447e-01
-8.69658217e-02 -8.29282403e-01 -2.89275557e-01 8.22416842e-01
-3.47120941e-01 5.86074516e-02 1.03124845e+00 -7.69206405e-01
5.52879214e-01 -2.12551141e+00 -4.21561569e-01 4.53646600e-01
4.18944836e-01 5.30585527e-01 -1.59507081e-01 -1.61752909e-01
-3.31546903e-01 3.95772725e-01 -6.95663914e-02 -7.89845884e-02
-3.44858587e-01 2.13921160e-01 7.00827420e-01 8.72474492e-01
4.39060003e-01 1.10888839e+00 -5.70485055e-01 -7.50995159e-01
3.15307051e-01 8.58807981e-01 -5.15628994e-01 1.15480661e-01
2.90804595e-01 2.63489187e-02 3.96111049e-02 1.03917110e+00
7.73960233e-01 -5.07471204e-01 2.11768568e-01 -5.32765448e-01
1.36981189e-01 -4.48815018e-01 -6.11217976e-01 1.23832536e+00
-2.10888088e-01 8.85060191e-01 2.18265191e-01 -1.04888332e+00
6.71433389e-01 3.24677855e-01 4.76655930e-01 -8.64220858e-01
5.25810182e-01 3.81225169e-01 4.81234252e-01 -6.84096456e-01
-5.14533967e-02 -2.57028669e-01 6.27803028e-01 2.44065702e-01
2.36774787e-01 5.39824329e-02 6.87332302e-02 4.74485643e-02
1.40914142e+00 -4.79229897e-01 2.34643683e-01 -5.20160317e-01
3.37837487e-01 3.44408542e-01 7.16069192e-02 3.23374122e-01
-5.18278182e-01 6.77055120e-01 7.49386609e-01 -7.42936730e-01
-1.26127672e+00 -6.93914533e-01 -2.89662361e-01 7.47339487e-01
-2.13544562e-01 3.38497669e-01 -7.06113279e-01 -6.01437747e-01
-7.93203264e-02 -1.90270931e-01 -1.11861730e+00 -1.03538349e-01
-6.36698484e-01 -1.36067200e+00 8.05802107e-01 7.25001693e-01
4.61583555e-01 -1.00475907e+00 -6.87433362e-01 -2.16892660e-02
4.65746403e-01 -8.17371964e-01 1.82113826e-01 5.04642844e-01
-9.68410134e-01 -1.58140206e+00 -1.03128290e+00 -1.06851339e+00
1.09353161e+00 -5.64953163e-02 1.01829052e+00 8.71585488e-01
-1.03812575e+00 1.39263466e-01 -4.08004820e-01 -7.30798662e-01
-5.88876605e-01 1.58451229e-01 -6.94253027e-01 -3.60961318e-01
3.02983373e-01 -2.07939640e-01 -9.61232603e-01 -1.28608318e-02
-1.27806783e+00 3.86121520e-03 1.17721879e+00 1.01259506e+00
5.91588974e-01 -1.67831287e-01 2.00661853e-01 -1.26529443e+00
3.77573729e-01 -2.89710850e-01 -1.07903183e-01 1.31375507e-01
-5.90696752e-01 -2.63737410e-01 4.43812668e-01 -1.85700938e-01
-7.06196249e-01 -9.43880454e-02 -3.36240232e-01 -1.03304699e-01
-1.50550917e-01 5.99721372e-01 3.27151835e-01 -7.57485926e-01
9.58708107e-01 -3.51737380e-01 5.80259740e-01 -3.00706439e-02
-1.96484089e-01 4.57958341e-01 4.16527152e-01 -1.97187122e-02
6.24408007e-01 9.91127849e-01 5.49551010e-01 -6.12575173e-01
-3.94424111e-01 -4.86215353e-01 -4.80571598e-01 -3.08733493e-01
1.14821780e+00 -5.79491854e-01 -6.49927676e-01 4.83667403e-01
-8.83668661e-01 -6.18509531e-01 -4.33972627e-02 3.43903661e-01
-1.06396735e-01 1.29116446e-01 -9.65490699e-01 -3.71560752e-01
-6.77248001e-01 -1.09024525e+00 8.95714819e-01 3.70135367e-01
-9.64434072e-02 -1.50234866e+00 -1.44995764e-01 1.86042860e-01
7.63225198e-01 8.83035541e-01 1.50818634e+00 -7.07956970e-01
-2.64949650e-01 -4.41259086e-01 -7.93257892e-01 3.42244923e-01
8.49808604e-02 2.85258591e-01 -9.55151856e-01 -4.87093478e-01
-4.74986792e-01 -3.23854238e-01 7.38020778e-01 5.85110724e-01
1.34240055e+00 -6.77090883e-02 -5.55673003e-01 1.03002107e+00
1.83879566e+00 2.22716779e-01 9.49725151e-01 6.26123190e-01
5.09634733e-01 5.48907816e-01 1.73512667e-01 -1.04763582e-01
-1.53436646e-01 1.11197844e-01 7.73249686e-01 -1.08506966e+00
-5.73738158e-01 4.44969893e-01 -4.75621223e-01 2.71018833e-01
-3.54674906e-01 -1.07454583e-01 -1.20930135e+00 7.37107098e-01
-1.34330082e+00 -4.83087897e-01 -2.72678733e-01 1.43289673e+00
5.33356309e-01 -5.03537543e-02 -1.56894103e-01 4.53624159e-01
4.45619166e-01 -6.54127061e-01 -5.48905849e-01 -4.34861690e-01
-2.83595413e-01 7.91721344e-01 7.87349224e-01 -8.21447298e-02
-1.01102984e+00 2.53636718e-01 5.98256207e+00 6.72016025e-01
-1.57252634e+00 2.37775430e-01 1.43250775e+00 1.72699928e-01
2.12142408e-01 -6.56959772e-01 -3.11783433e-01 4.11122404e-02
9.30243075e-01 3.71725380e-01 -7.29846433e-02 7.88532972e-01
8.19793493e-02 -1.53478339e-01 -9.45274711e-01 8.39213729e-01
-6.49866015e-02 -1.73512697e+00 7.63564184e-02 3.87057394e-01
6.31462872e-01 -6.37999326e-02 1.85642138e-01 1.11257382e-01
1.56917185e-01 -1.52593839e+00 9.92768109e-02 4.01505679e-01
1.09756529e+00 -1.01524913e+00 1.65315771e+00 -2.96659738e-01
-7.30738699e-01 1.20495846e-02 -3.48633945e-01 2.62833983e-01
-6.35606885e-01 4.64131951e-01 -1.35311830e+00 4.44609433e-01
1.08787608e+00 4.77939278e-01 -1.03922546e+00 1.25478303e+00
5.57490647e-01 6.77854717e-01 -2.67590862e-02 -2.50370651e-01
5.05613148e-01 4.47651535e-01 -1.20998509e-01 1.59848702e+00
3.70426387e-01 -1.20242245e-01 -3.93335044e-01 5.16071677e-01
5.49503975e-02 1.91824153e-01 -1.83865353e-01 8.86050612e-02
-6.87872544e-02 1.80070448e+00 -1.47024345e+00 -2.48700693e-01
-1.24811269e-01 4.84750122e-01 2.06681296e-01 1.91657946e-01
-6.86695278e-01 -4.03147370e-01 4.33268785e-01 4.36237484e-01
2.10157990e-01 3.48389775e-01 -3.15186977e-01 -3.49625885e-01
-3.90034348e-01 -8.41641009e-01 5.15972853e-01 -8.17334294e-01
-1.40956700e+00 4.88425523e-01 -3.56916696e-01 -9.60411072e-01
1.45081356e-01 -1.44053960e+00 -5.91506064e-01 7.81771541e-01
-1.76848936e+00 -1.41052437e+00 -5.97249806e-01 5.41349411e-01
-2.08723713e-02 -1.26330197e-01 9.52438891e-01 4.20345634e-01
-3.98775160e-01 6.49791837e-01 9.50567052e-02 5.71813345e-01
6.43759549e-01 -1.40795016e+00 -3.07139486e-01 4.41251844e-01
-7.21829355e-01 3.73047352e-01 3.72705638e-01 -2.11807117e-01
-1.26509142e+00 -1.38248527e+00 5.39461076e-01 -8.51122439e-02
4.12925661e-01 -8.46112370e-02 -8.08011234e-01 3.44629139e-01
2.23954827e-01 5.01041353e-01 1.02271128e+00 -3.48011225e-01
-4.16618660e-02 -7.80652687e-02 -1.76662695e+00 3.89360070e-01
6.62741005e-01 -2.22101286e-01 2.79089868e-01 4.55312192e-01
3.26289773e-01 -5.39855182e-01 -8.77658486e-01 6.26803279e-01
5.68170011e-01 -1.17396104e+00 8.57453227e-01 -4.39797878e-01
5.43705106e-01 -1.95408747e-01 1.71274781e-01 -1.04428124e+00
-3.36407363e-01 9.81179103e-02 3.80079895e-01 8.14613640e-01
4.68487114e-01 -7.46275663e-01 1.15521085e+00 2.61096895e-01
-3.34827989e-01 -1.33325565e+00 -1.06082165e+00 -4.67294484e-01
2.56529450e-01 -1.75422043e-01 4.85230833e-01 8.32278550e-01
-4.60602075e-01 -3.70924503e-01 4.28754807e-01 2.93279197e-02
3.83699685e-01 -6.82010114e-01 4.35903460e-01 -1.28186989e+00
-3.69844362e-02 -5.57259262e-01 -1.03820491e+00 1.17799290e-01
-3.26996624e-01 -8.60681117e-01 -2.24183396e-01 -1.59890509e+00
4.72134382e-01 -3.86307836e-01 -3.15868586e-01 6.11972511e-01
1.25913799e-01 1.09149873e+00 -2.42962569e-01 -9.29333642e-03
-2.91373283e-01 -3.84913445e-01 1.57359648e+00 -3.58457059e-01
2.08538324e-01 -5.38505018e-01 -8.14275980e-01 5.15459716e-01
7.52029955e-01 -3.53638530e-01 -1.15825288e-01 -2.51591057e-01
2.93498248e-01 -3.30867738e-01 8.63213599e-01 -1.31875980e+00
2.38514289e-01 3.08486998e-01 1.08864093e+00 -3.67530763e-01
2.01570541e-02 -8.12282085e-01 2.35161483e-01 1.02489734e+00
-1.49901673e-01 2.02835258e-02 4.58293438e-01 2.73758650e-01
-6.99611455e-02 4.12480235e-02 1.02091861e+00 -1.85713246e-01
-4.18349206e-01 3.01190913e-01 -4.55607086e-01 -6.67288125e-01
1.10084260e+00 -7.72903383e-01 -5.64468801e-01 -1.11947441e-03
-6.28101587e-01 -1.31160632e-01 3.79049987e-01 -1.82857618e-01
4.87663627e-01 -1.14290488e+00 -5.37807345e-01 3.07176989e-02
1.18498065e-01 2.53646433e-01 5.86114049e-01 1.19232082e+00
-1.50008225e+00 3.80219638e-01 -4.47893620e-01 -8.54298711e-01
-1.50944531e+00 1.78504452e-01 9.61952984e-01 -8.12548697e-01
-1.89777732e-01 1.07803142e+00 6.91757575e-02 -5.41328132e-01
8.87643546e-02 -5.87552607e-01 -3.69108468e-01 -2.72750825e-01
5.16055882e-01 4.70533639e-01 7.45738149e-01 -2.68456280e-01
-3.48298222e-01 4.06670541e-01 -2.75695652e-01 4.49740827e-01
1.68254066e+00 3.56472611e-01 -2.68995732e-01 -1.19340740e-01
1.55106294e+00 -7.62053907e-01 -8.66023064e-01 2.96881765e-01
-1.04041636e-01 3.63055281e-02 3.41997653e-01 -1.14369166e+00
-1.43713164e+00 8.73031735e-01 1.34734476e+00 2.61486053e-01
1.18254900e+00 -1.67824909e-01 4.61591899e-01 1.69443280e-01
-1.71842068e-01 -7.75606930e-01 2.52818525e-01 1.52294829e-01
5.24941325e-01 -1.34785163e+00 1.34408951e-01 -5.39358675e-01
-4.43764217e-03 1.58839011e+00 7.04674423e-01 -2.82104552e-01
9.08383071e-01 5.21352291e-01 5.35290480e-01 -7.44351923e-01
-4.61830527e-01 -5.85907064e-02 2.02545360e-01 8.97261620e-01
7.30103433e-01 -6.73353374e-02 -1.29791340e-02 2.50048310e-01
-2.23696828e-01 7.19143227e-02 4.24188137e-01 1.08602774e+00
-8.06820393e-02 -6.15272641e-01 -4.02963430e-01 8.29596400e-01
-8.11071157e-01 2.09341347e-01 -4.76994962e-01 1.36966491e+00
5.42026699e-01 5.35853982e-01 2.91785926e-01 -7.58222938e-02
2.13308930e-01 -2.55773693e-01 4.07571018e-01 -5.80158174e-01
-9.24594522e-01 -1.44256398e-01 -2.00168297e-01 -3.00642461e-01
-6.49206221e-01 -2.34229177e-01 -1.15499330e+00 -6.22060835e-01
-3.48428100e-01 -1.91907898e-01 8.11084330e-01 8.81152213e-01
1.42252788e-01 1.02444029e+00 7.79432133e-02 -7.99044251e-01
5.40862372e-03 -8.88299525e-01 -6.61888838e-01 2.88782001e-01
4.11536038e-01 -3.21216136e-01 -3.57747585e-01 2.09690228e-01] | [15.184981346130371, -2.9449448585510254] |
89620684-f9ac-41e3-96b2-4dec1aaea261 | share-a-system-for-hierarchical-assistive | 2105.08185 | null | https://arxiv.org/abs/2105.08185v2 | https://arxiv.org/pdf/2105.08185v2.pdf | SHARE: a System for Hierarchical Assistive Recipe Editing | The large population of home cooks with dietary restrictions is under-served by existing cooking resources and recipe generation models. To help them, we propose the task of controllable recipe editing: adapt a base recipe to satisfy a user-specified dietary constraint. This task is challenging, and cannot be adequately solved with human-written ingredient substitution rules or existing end-to-end recipe generation models. We tackle this problem with SHARE: a System for Hierarchical Assistive Recipe Editing, which performs simultaneous ingredient substitution before generating natural-language steps using the edited ingredients. By decoupling ingredient and step editing, our step generator can explicitly integrate the available ingredients. Experiments on the novel RecipePairs dataset -- 83K pairs of similar recipes where each recipe satisfies one of seven dietary constraints -- demonstrate that SHARE produces convincing, coherent recipes that are appropriate for a target dietary constraint. We further show through human evaluations and real-world cooking trials that recipes edited by SHARE can be easily followed by home cooks to create appealing dishes. | ['Julian McAuley', 'Jianmo Ni', 'Yufei Li', 'Shuyang Li'] | 2021-05-17 | null | null | null | null | ['recipe-generation'] | ['miscellaneous'] | [ 2.60025471e-01 3.79900455e-01 -2.02212408e-01 -5.52507818e-01
-4.52131957e-01 -9.42313850e-01 2.05590382e-01 4.12023544e-01
1.00751974e-01 4.99625921e-01 8.58814716e-01 -8.17214027e-02
1.62210673e-01 -9.43727732e-01 -8.43337297e-01 -1.25052303e-01
2.09212929e-01 5.96594870e-01 -4.04000580e-01 -7.49490440e-01
-1.75805405e-01 -3.03821951e-01 -1.45963085e+00 5.44971764e-01
1.34686220e+00 -1.95192605e-01 3.23775202e-01 7.76940286e-01
-3.63969266e-01 7.32045114e-01 -2.99409121e-01 -2.83257276e-01
4.05392259e-01 -1.18061173e+00 -2.19481245e-01 6.17525466e-02
5.15383065e-01 -6.20968938e-01 1.56019822e-01 7.99177706e-01
5.60596704e-01 3.41091126e-01 7.20935285e-01 -1.18997979e+00
-1.43853641e+00 1.44947302e+00 5.53348102e-02 -7.96053052e-01
8.65706325e-01 4.37772930e-01 8.82467926e-01 -5.75086057e-01
5.86458206e-01 1.08784008e+00 9.44327414e-01 1.05246258e+00
-1.67300212e+00 -5.55894315e-01 2.21332103e-01 -5.41963756e-01
-1.05861866e+00 -5.14109015e-01 7.11336970e-01 -3.06412101e-01
9.31230128e-01 3.66464704e-01 1.24070096e+00 9.42762017e-01
-2.97174037e-01 7.45129287e-01 7.33666420e-01 -5.43966949e-01
2.98721969e-01 5.40940166e-02 9.97928455e-02 7.59727418e-01
3.53357404e-01 2.77772844e-01 -4.44261521e-01 -8.66854638e-02
8.10634792e-01 -2.60658443e-01 -5.23265339e-02 -4.54323143e-01
-1.50457418e+00 9.20777261e-01 2.32140124e-01 -1.49349883e-01
-6.77244484e-01 1.53957188e-01 2.76037574e-01 3.72230053e-01
4.57357824e-01 1.08332384e+00 -7.71520674e-01 3.81195456e-01
-9.41146076e-01 1.00464249e+00 1.19440949e+00 1.50652790e+00
4.97802973e-01 -1.58512946e-02 -5.19662201e-01 9.76328313e-01
3.55727941e-01 8.84653509e-01 -1.30043387e-01 -1.20389819e+00
2.21761242e-02 4.09448802e-01 6.59884334e-01 -8.88118804e-01
-5.43344676e-01 3.34432751e-01 -7.20481336e-01 2.85029292e-01
6.22718334e-01 -4.82056826e-01 -7.70178914e-01 2.12483311e+00
5.33061802e-01 -2.69970745e-01 -4.96330559e-02 8.31576884e-01
1.28074837e+00 7.19779551e-01 8.50149870e-01 -1.00018736e-03
1.45120084e+00 -1.14887416e+00 -8.92072320e-01 -1.66284740e-01
6.25561655e-01 -7.48516560e-01 1.31723762e+00 2.95554936e-01
-1.30297387e+00 -6.24637842e-01 -9.53239381e-01 -3.73713464e-01
-3.88540596e-01 3.78102154e-01 9.69401479e-01 6.80219054e-01
-9.24679279e-01 6.63809836e-01 -5.08436978e-01 -5.02219915e-01
-7.11084083e-02 1.09623186e-02 -2.59646028e-01 -3.86363178e-01
-1.22240937e+00 7.64126718e-01 6.21273518e-01 -2.35930055e-01
-6.97593451e-01 -1.49660861e+00 -1.46324658e+00 -1.11223392e-01
4.41739976e-01 -1.33257437e+00 1.53357339e+00 -1.11455643e+00
-1.81426382e+00 5.97777545e-01 2.92957515e-01 1.36372954e-01
3.54740798e-01 -4.89388257e-01 -5.59887588e-01 -2.77520806e-01
4.17012870e-01 1.19168687e+00 4.63651389e-01 -9.94959414e-01
-2.83981025e-01 2.16794387e-01 2.62190819e-01 4.19255584e-01
2.59826183e-01 2.09235579e-01 -3.02256439e-02 -9.86927927e-01
-3.06960195e-01 -8.34233999e-01 -7.49490738e-01 3.89369100e-01
-5.76419950e-01 -1.29673317e-01 -2.03870147e-01 -1.02939904e+00
1.11536980e+00 -1.83311200e+00 2.18706936e-01 -8.45337808e-02
-1.50557682e-01 3.73828337e-02 -5.25299668e-01 6.12391174e-01
-1.20029733e-01 1.02071159e-01 -1.99772075e-01 -1.81736365e-01
6.65360212e-01 7.30284000e-06 2.25555047e-01 6.56782314e-02
1.47121355e-01 9.16323841e-01 -1.47104406e+00 -3.25035989e-01
6.57041848e-01 5.24981380e-01 -1.26892245e+00 5.72767258e-01
-1.14199758e+00 3.18376511e-01 -2.71364212e-01 3.15575838e-01
5.44784784e-01 8.12185779e-02 5.28012633e-01 -5.90906918e-01
-1.32229164e-01 3.01681906e-01 -1.47339392e+00 2.41849542e+00
-3.86764467e-01 -1.79265276e-01 1.11378349e-01 -1.75690860e-01
8.05024803e-01 3.29766661e-01 3.36977780e-01 -4.44981158e-01
-2.57198483e-01 1.47527918e-01 -4.98348683e-01 -4.98656362e-01
9.00170207e-01 -2.32186645e-01 -4.57334161e-01 7.59623885e-01
9.48293041e-03 -6.16256475e-01 8.01495612e-01 2.53468901e-01
7.04601586e-01 8.50806773e-01 6.73860371e-01 -6.76207185e-01
2.13177074e-02 4.77598250e-01 3.45553875e-01 5.60640514e-01
1.97351635e-01 6.40375555e-01 -1.44240916e-01 -4.41552609e-01
-1.51750326e+00 -1.13264120e+00 6.60925031e-01 1.61588430e+00
-3.64732146e-01 -6.89497769e-01 -7.45792031e-01 -4.87687379e-01
8.78238380e-02 1.42272627e+00 -5.97673118e-01 -7.67468438e-02
-7.83886790e-01 -5.49303055e-01 4.06850666e-01 5.02492666e-01
4.09218669e-01 -1.38640559e+00 -6.57887101e-01 6.81995571e-01
-4.65179890e-01 -4.55227762e-01 -1.41505003e+00 4.69892062e-02
-3.53739738e-01 -1.18351448e+00 -7.61822760e-01 -1.08148003e+00
7.37748623e-01 2.42545322e-01 1.89399910e+00 2.16558605e-01
-3.22277725e-01 3.38149399e-01 -4.40400481e-01 -3.20073128e-01
-1.31764662e+00 -2.46561896e-02 -1.23118259e-01 -7.99047410e-01
5.15261412e-01 -4.14952219e-01 -7.71323621e-01 1.71343908e-01
-9.01935279e-01 1.01950192e+00 3.07433195e-02 4.98809963e-01
4.41992998e-01 -2.62846798e-01 7.00528860e-01 -9.94962454e-01
6.69876337e-01 -5.85154831e-01 -2.72597551e-01 4.40815866e-01
-4.23613995e-01 3.82639885e-01 7.96920121e-01 -7.09573150e-01
-1.16654480e+00 5.39417505e-01 -1.15059987e-01 3.50221813e-01
-5.17701864e-01 3.05187970e-01 -4.29864436e-01 8.26539814e-01
9.27293956e-01 -1.58624873e-02 -1.85155734e-01 -7.39042580e-01
1.57850838e+00 1.86157390e-01 8.52063060e-01 -1.06053603e+00
5.40074587e-01 -3.23951095e-01 -5.34242928e-01 -4.23957974e-01
-6.05236173e-01 -3.54779363e-02 -2.77104586e-01 3.15296985e-02
1.17453301e+00 -1.15086865e+00 -7.12442815e-01 3.37728709e-01
-1.03651106e+00 -1.13821054e+00 -5.48045576e-01 2.55318016e-01
-6.53026283e-01 1.86056215e-02 -7.57428765e-01 -3.26840490e-01
-5.33760548e-01 -6.84130609e-01 9.29953039e-01 3.51021349e-01
-8.78113866e-01 -9.04821038e-01 4.28099871e-01 -3.88407670e-02
5.25579929e-01 5.43009043e-01 1.19478917e+00 -3.19327414e-01
-1.18261084e-01 2.99665511e-01 -2.95669194e-02 -3.80625203e-02
7.11891055e-01 6.25871792e-02 -2.23680928e-01 7.79362097e-02
-7.53446221e-01 -2.97924697e-01 3.79676104e-01 5.38423121e-01
3.02409977e-01 -4.23730493e-01 -8.23060051e-02 3.55898678e-01
1.27957904e+00 -1.26696080e-01 3.28029871e-01 -1.46289349e-01
7.00137436e-01 6.85185134e-01 4.80931222e-01 5.16034663e-01
6.60653710e-01 7.33877361e-01 -2.42121473e-01 -4.00894523e-01
-5.93320012e-01 -1.05902112e+00 5.19754350e-01 4.86060977e-01
2.00704545e-01 -3.05416077e-01 -2.23164633e-01 7.30795324e-01
-1.85505939e+00 -1.20129585e+00 -3.23563576e-01 1.96971643e+00
1.54873312e+00 -6.96476102e-01 4.54337895e-01 -3.98835540e-01
5.25348783e-01 -1.20023482e-01 -5.62399566e-01 -3.46598327e-01
6.86953813e-02 2.55899578e-01 4.07689303e-01 6.32873952e-01
-1.07035661e+00 9.35825646e-01 7.16286993e+00 1.84071466e-01
-1.99003458e-01 -1.57693192e-01 3.84448439e-01 -1.78545445e-01
-8.46042871e-01 -4.04683053e-01 -4.78423059e-01 3.62841129e-01
8.00290644e-01 -4.35445420e-02 1.26577950e+00 6.53485775e-01
6.42690778e-01 1.32971570e-01 -1.69264781e+00 3.92401129e-01
-1.70816239e-02 -1.35086143e+00 -4.92249615e-02 -6.18829370e-01
8.51632476e-01 -6.03092849e-01 -3.43143702e-01 6.94281757e-01
1.61032248e+00 -1.03435624e+00 1.14262605e+00 3.92177343e-01
8.75084460e-01 -6.25584900e-01 -1.67016715e-01 2.50857413e-01
-1.48411524e+00 3.90438855e-01 -3.81172568e-01 -6.07992616e-03
5.34620404e-01 4.67446119e-01 -5.65986574e-01 5.15101731e-01
3.04750174e-01 7.57463038e-01 -3.94173831e-01 7.28544414e-01
-5.73231697e-01 3.87727708e-01 -5.30424416e-01 7.34025287e-03
-1.57744229e-01 -4.24134254e-01 2.41554603e-01 1.46947229e+00
5.92078865e-01 4.78713274e-01 4.53163326e-01 1.51008844e+00
5.34239225e-02 4.82701927e-01 -5.92369020e-01 -2.36265466e-01
3.05622637e-01 1.12463057e+00 -4.46741045e-01 -4.53552216e-01
-2.85289288e-01 1.38159466e+00 1.76218763e-01 3.45412463e-01
-8.79718244e-01 -5.31345746e-03 8.71378005e-01 -1.84595332e-01
2.29156643e-01 7.90273994e-02 -1.17195420e-01 -1.17809534e+00
-5.24575830e-01 -1.48005426e+00 3.59504730e-01 -1.24622118e+00
-1.47550321e+00 -2.61971448e-02 2.74584234e-01 -7.34968305e-01
-3.25339198e-01 -1.64886251e-01 -5.01806080e-01 9.56195414e-01
-9.12553549e-01 -1.51168919e+00 -2.43218765e-01 4.57895190e-01
9.22602952e-01 3.37197989e-01 1.30915821e+00 5.37994325e-01
-2.17392951e-01 4.65364516e-01 -2.80889422e-01 -1.89287197e-02
9.35295820e-01 -1.52689910e+00 9.18286204e-01 6.22683406e-01
-7.30155408e-02 7.34549165e-01 1.39087522e+00 -1.13878560e+00
-1.29101717e+00 -1.32728994e+00 1.04607618e+00 -5.79258323e-01
4.03132141e-02 -3.74340981e-01 -5.12868524e-01 6.94592535e-01
4.42099988e-01 -5.43633461e-01 9.06010926e-01 4.83906083e-02
-6.56297147e-01 4.68723178e-01 -1.36669052e+00 9.35340941e-01
1.46587920e+00 -1.31131321e-01 -5.70821583e-01 4.96837318e-01
1.10530102e+00 -4.54175770e-01 -9.14696038e-01 -1.47579253e-01
6.21731281e-01 -5.39355457e-01 1.34683585e+00 -9.67218220e-01
7.12741792e-01 -5.72498381e-01 -1.75196826e-01 -2.07892013e+00
-6.94480121e-01 -7.31878221e-01 2.36616418e-01 1.14498079e+00
6.31618917e-01 2.25319400e-01 5.07818520e-01 1.24815881e+00
-4.10736412e-01 2.83369303e-01 2.16607854e-01 -4.66058046e-01
1.68416202e-01 -1.58413425e-01 1.37066364e+00 1.19822478e+00
4.11234051e-01 2.75185108e-01 -9.39617157e-01 -4.68198508e-02
6.04350805e-01 1.51059199e-02 9.97122824e-01 -7.05140948e-01
-8.09104085e-01 -4.47002947e-01 9.68315899e-01 -9.51871097e-01
-1.35024175e-01 -1.07004726e+00 7.65486598e-01 -2.07561684e+00
2.80188978e-01 -5.09724140e-01 1.42253950e-01 7.51286685e-01
-6.74817264e-01 -6.26285374e-03 4.54058677e-01 -5.00372410e-01
-4.40586120e-01 3.97554338e-01 1.27573061e+00 -3.24692786e-01
-8.12210500e-01 -4.13545936e-01 -1.20901835e+00 3.89665186e-01
8.61511648e-01 -5.79009891e-01 -6.71672165e-01 -3.74260962e-01
6.80433691e-01 8.78403410e-02 1.69497311e-01 -6.18106663e-01
-1.56684622e-01 -5.21575034e-01 1.83948517e-01 -2.11169034e-01
-2.62947440e-01 -6.63408756e-01 1.11541927e+00 4.05698359e-01
-9.47601020e-01 2.67972592e-02 3.40737134e-01 2.92449206e-01
7.57993698e-01 -8.64495039e-02 2.88627654e-01 -4.32633102e-01
-5.57837129e-01 -2.93914825e-02 -5.02693057e-01 5.39943501e-02
6.53914928e-01 2.14304253e-01 -2.99586624e-01 -2.96805590e-01
-7.63069451e-01 2.83745885e-01 7.40477800e-01 6.01279914e-01
1.70139909e-01 -1.78749871e+00 -1.32247388e+00 1.80968747e-01
3.76344293e-01 -1.98665425e-01 3.61186713e-01 1.99969307e-01
-7.02589989e-01 -1.80432364e-01 -2.83351511e-01 1.35109797e-01
-1.00230229e+00 1.17552674e+00 3.64046276e-01 -4.06643122e-01
-8.54754746e-01 4.15183365e-01 2.47127056e-01 -1.23904884e+00
-1.05732627e-01 -8.77001584e-01 -1.69790331e-02 -4.24127817e-01
8.71850252e-01 2.23653272e-01 -5.57173252e-01 -5.63465767e-02
7.55527541e-02 3.21606755e-01 3.54626864e-01 1.15245022e-01
1.31871319e+00 -1.10626206e-01 1.40781552e-01 3.71800996e-02
5.01540601e-01 5.38144894e-02 -1.33598018e+00 -1.74031574e-02
-3.97674710e-01 -1.68015704e-01 -4.69802231e-01 -1.53499103e+00
-7.32048273e-01 6.78880960e-02 2.87348717e-01 -4.36729416e-02
9.90751743e-01 -2.62368619e-01 9.62757051e-01 3.16219181e-01
4.07878280e-01 -9.23438728e-01 -1.82946593e-01 -1.25198007e-01
8.67636323e-01 -1.13941729e+00 -1.12097241e-01 -6.04463696e-01
-7.21858144e-01 7.71158218e-01 4.69438821e-01 -1.75126791e-01
5.01917064e-01 2.17322513e-01 1.85675815e-01 -1.03606910e-01
-5.96885443e-01 1.34166047e-01 3.29488128e-01 1.07847333e+00
7.66813934e-01 6.78304076e-01 -4.79694009e-01 8.45493913e-01
-2.01770976e-01 6.70407712e-01 3.21639806e-01 5.52820981e-01
-3.63410383e-01 -1.36818027e+00 -2.10577503e-01 2.16018528e-01
1.59817398e-01 -5.99525750e-01 -1.41075581e-01 6.09321296e-01
3.85946989e-01 1.28753948e+00 -2.78229982e-01 2.15829965e-02
5.06387055e-01 1.19768493e-01 6.85463369e-01 -9.28743899e-01
-1.26176155e+00 -1.58505011e-02 7.23186970e-01 -4.91219372e-01
-5.37218094e-01 -6.07919097e-01 -1.33163667e+00 -7.99640119e-01
1.34446830e-01 5.84345534e-02 3.31783026e-01 4.96943772e-01
2.15044320e-01 6.36673748e-01 2.15974927e-01 -8.20538223e-01
-4.66182560e-01 -7.88428307e-01 -5.90148330e-01 9.45267797e-01
-1.09574199e-01 -7.32321590e-02 9.69504938e-02 6.87779009e-01] | [11.509529113769531, 4.547859191894531] |
aa374b35-ae87-44d4-99e4-4ed2cfed332e | word-sense-disambiguation-of-french | null | null | https://aclanthology.org/2022.textgraphs-1.8 | https://aclanthology.org/2022.textgraphs-1.8.pdf | Word Sense Disambiguation of French Lexicographical Examples Using Lexical Networks | This paper focuses on the task of word sense disambiguation (WSD) on lexicographic examples relying on the French Lexical Network (fr-LN). For this purpose, we exploit the lexical and relational properties of the network, that we integrated in a feedforward neural WSD model on top of pretrained French BERT embeddings. We provide a comparative study with various models and further show the impact of our approach regarding polysemic units. | ['Mathieu Constant', 'Sandrine Ollinger', 'Aman Sinha'] | null | null | null | null | coling-textgraphs-2022-10 | ['word-sense-disambiguation'] | ['natural-language-processing'] | [-9.40683112e-02 2.72909433e-01 -2.70216942e-01 -3.21722776e-01
-1.10712629e-02 -6.14889920e-01 8.13702583e-01 5.67959130e-01
-1.09965670e+00 7.92571664e-01 7.69238234e-01 -4.43880945e-01
-2.74100304e-01 -9.70047891e-01 -2.07027301e-01 -3.70599441e-02
-6.59079924e-02 3.85079652e-01 1.37663737e-01 -8.29242289e-01
6.82435632e-02 5.13654172e-01 -1.05948591e+00 4.05226797e-01
6.37243390e-01 8.66666079e-01 2.95440078e-01 -9.44447070e-02
-5.06233037e-01 6.61378920e-01 -6.73398495e-01 -4.81452286e-01
1.29940748e-01 -8.05885792e-02 -9.98054683e-01 -5.94684303e-01
4.04556870e-01 1.69740751e-01 -2.76503414e-01 1.13447678e+00
3.84976745e-01 4.93090808e-01 6.76596224e-01 -4.88045961e-01
-9.21105623e-01 1.41710865e+00 2.21651599e-01 6.48012578e-01
6.11019850e-01 -2.78964430e-01 1.67968380e+00 -8.94645691e-01
1.25593793e+00 1.32758057e+00 7.45254338e-01 5.60215235e-01
-1.38976812e+00 -3.89664143e-01 4.25818086e-01 2.32940853e-01
-1.47227705e+00 -2.27545485e-01 7.79392958e-01 -2.19909847e-01
1.77441609e+00 -1.86216012e-01 8.33995759e-01 1.22883809e+00
1.16876237e-01 6.40938640e-01 1.06843984e+00 -6.90593123e-01
9.67949331e-02 8.69484469e-02 6.98611975e-01 5.56991160e-01
4.47710633e-01 2.05669880e-01 -4.71417516e-01 1.31484419e-01
7.71028757e-01 -3.34506691e-01 -1.55216709e-01 -2.31444135e-01
-9.81626332e-01 9.83558834e-01 8.46995294e-01 1.27407777e+00
-4.59752172e-01 -6.18687421e-02 5.73728383e-01 5.62799335e-01
5.18267632e-01 1.10934484e+00 -7.23796904e-01 5.41075647e-01
-8.83013964e-01 1.16963647e-01 1.03863525e+00 7.69543886e-01
8.98660481e-01 1.27378866e-01 -5.34205697e-02 8.71982157e-01
2.75695652e-01 2.55759597e-01 9.37752903e-01 -3.19009274e-01
1.14513397e-01 7.51205087e-01 -3.43356401e-01 -1.20451820e+00
-7.24833250e-01 -3.73956978e-01 -2.95729727e-01 -3.29679996e-01
-8.81479457e-02 -1.35068689e-02 -7.62466609e-01 1.93682373e+00
-1.01002328e-01 -4.20925301e-03 3.55139375e-01 8.26391280e-01
1.00667906e+00 3.34212273e-01 6.56608045e-01 -2.72625655e-01
1.31640136e+00 -4.80362296e-01 -9.80020523e-01 -3.08116317e-01
7.16817975e-01 -4.83962268e-01 1.33546627e+00 9.76340100e-02
-1.07757771e+00 -6.24657631e-01 -1.16860747e+00 -2.32130229e-01
-1.28743851e+00 3.49003784e-02 5.34305394e-01 5.17800152e-01
-1.29607439e+00 6.47333622e-01 -5.89961767e-01 -8.47959876e-01
7.51567706e-02 3.80166918e-01 -3.61697972e-01 1.37465194e-01
-2.14428782e+00 1.40458572e+00 1.12435067e+00 -8.58018734e-03
-3.98459077e-01 -5.74957073e-01 -1.35932362e+00 -1.72935858e-01
5.09223163e-01 -7.52870202e-01 8.75425875e-01 -9.86213267e-01
-1.31093335e+00 1.24953008e+00 8.10628291e-03 -8.05637181e-01
-2.11301506e-01 -3.67556274e-01 -9.43668544e-01 -5.04509062e-02
2.26519927e-01 6.43791020e-01 1.79386914e-01 -1.08115339e+00
-3.06501687e-01 -1.79904968e-01 5.40196061e-01 1.51122019e-01
-3.53659183e-01 1.14470862e-01 1.90842167e-01 -1.07242441e+00
-2.46739641e-01 -5.45827687e-01 -1.09164827e-01 -8.96984637e-01
-4.14356530e-01 -7.70867407e-01 2.84609318e-01 -4.63827819e-01
1.75365520e+00 -2.23925114e+00 4.24375355e-01 4.22866225e-01
9.76563394e-02 4.53787357e-01 -3.34301919e-01 6.63889468e-01
-4.81959313e-01 3.40533525e-01 -1.47534475e-01 -2.06197370e-02
3.39534074e-01 5.95529139e-01 -3.61981750e-01 4.65973988e-02
5.40865660e-01 1.08083987e+00 -1.10963595e+00 -2.57705122e-01
3.00788701e-01 2.78033078e-01 -6.40947759e-01 4.03705277e-02
-4.18356717e-01 -3.67654294e-01 -2.94591874e-01 3.32298279e-01
1.99744552e-01 2.35749230e-01 8.13532591e-01 -4.69401270e-01
-2.78311014e-01 1.29006565e+00 -1.18885696e+00 2.09874153e+00
-8.49741697e-01 1.49287686e-01 -3.94425362e-01 -8.10312033e-01
8.23190689e-01 3.94064546e-01 -3.13704349e-02 -8.19572449e-01
3.37134719e-01 4.42078203e-01 1.39376774e-01 -1.26321092e-01
8.06285501e-01 -5.14971018e-01 -3.65614921e-01 4.33040917e-01
8.95308137e-01 9.18102190e-02 6.89597249e-01 3.10852170e-01
8.69263649e-01 3.16781133e-01 1.05716670e+00 -9.41507518e-01
7.54346848e-01 -5.44124842e-02 1.52097359e-01 4.89753038e-01
1.51941672e-01 6.13333145e-03 4.80633140e-01 -5.18245637e-01
-5.65617502e-01 -1.13642943e+00 -3.99208248e-01 1.24329329e+00
-1.49924845e-01 -9.83473718e-01 -6.58464968e-01 -9.77627814e-01
1.19392574e-03 1.14149773e+00 -7.96571791e-01 -1.81170449e-01
-9.54698741e-01 -5.73759615e-01 5.93486428e-01 5.74834466e-01
4.73856106e-02 -1.46062660e+00 -2.64627963e-01 3.69164556e-01
3.30909938e-01 -1.05665147e+00 -6.35082200e-02 6.92437708e-01
-4.01035279e-01 -9.61561203e-01 -8.54123756e-03 -1.01399839e+00
1.11236842e-02 -3.18521619e-01 1.71278238e+00 -7.96750113e-02
1.06499651e-02 2.64003664e-01 -4.60788906e-01 -2.18582332e-01
-2.95037240e-01 6.62281513e-01 5.06782830e-01 -4.62244719e-01
1.13585138e+00 -5.97186804e-01 7.05044344e-02 -2.08068416e-01
-1.28242719e+00 -7.22879410e-01 2.07893133e-01 7.65120685e-01
4.44218606e-01 -3.08691800e-01 5.45767844e-01 -1.51424289e+00
1.00722277e+00 -5.88619173e-01 -5.55297375e-01 4.24901992e-02
-5.71461678e-01 3.58299255e-01 5.12321234e-01 -1.49914280e-01
-9.60617065e-01 -1.74261108e-01 -4.06401187e-01 1.84951108e-02
-2.52097338e-01 8.85522068e-01 -4.23504472e-01 1.16909750e-01
7.92863131e-01 -3.64259064e-01 -6.10621214e-01 -5.91655910e-01
9.49420393e-01 3.26250911e-01 8.65238979e-02 -5.45657933e-01
3.52607399e-01 -1.30430788e-01 -2.28183702e-01 -9.19894814e-01
-1.24182904e+00 -2.70472139e-01 -1.08291423e+00 2.85322666e-01
1.10753739e+00 -6.78254843e-01 -2.63053983e-01 -1.80733845e-01
-1.30762374e+00 -1.78857259e-02 -6.77209735e-01 3.86809379e-01
-3.43437791e-01 -3.12127788e-02 -7.67546237e-01 -2.67487735e-01
6.79438561e-02 -7.18598962e-01 7.20108628e-01 -1.67358980e-01
-7.66851008e-01 -1.69193375e+00 3.63001943e-01 -5.58011949e-01
5.52344561e-01 -5.58628840e-03 1.34485006e+00 -1.50592875e+00
-1.45266041e-01 1.73964843e-01 -2.08783627e-01 2.86572844e-01
1.24476761e-01 -5.70372760e-01 -9.78162408e-01 3.28247696e-02
-1.99865289e-02 -2.45411694e-01 1.22978687e+00 2.17117399e-01
3.99234861e-01 -8.25883076e-02 -2.72906363e-01 4.58448023e-01
1.80474055e+00 -7.50977099e-02 1.57077298e-01 5.00660539e-01
6.38971210e-01 5.83842099e-01 3.89448822e-01 -2.99923904e-02
2.88965583e-01 3.64319205e-01 1.56745479e-01 2.84521610e-01
-4.61964965e-01 -3.40226531e-01 4.03130740e-01 1.24439287e+00
1.13913104e-01 -4.02393252e-01 -9.45572436e-01 6.09854460e-01
-1.45245159e+00 -4.27548289e-01 2.86568075e-01 1.72720540e+00
1.03019333e+00 2.82075524e-01 -2.90484745e-02 -4.13850993e-02
4.91111368e-01 7.65094340e-01 3.18282209e-02 -7.73249209e-01
-5.74346721e-01 9.67070222e-01 2.00127602e-01 9.54347253e-01
-1.17800725e+00 1.74946666e+00 7.02010918e+00 7.37078726e-01
-8.45763445e-01 2.91843057e-01 -1.77512497e-01 1.38073444e-01
-4.81062800e-01 -2.65058875e-01 -9.65142787e-01 -1.07022591e-01
1.11559451e+00 8.09997767e-02 2.49881923e-01 6.07612908e-01
-4.80815291e-01 1.56185761e-01 -1.26624918e+00 4.62842345e-01
3.26375932e-01 -1.22730839e+00 5.91894686e-01 -2.76684314e-01
4.27129537e-01 1.89268410e-01 -3.88459116e-01 3.69091213e-01
5.41044474e-01 -8.57439458e-01 3.74614209e-01 4.19807911e-01
7.22909570e-01 -7.59273350e-01 9.63174582e-01 -3.29138845e-01
-1.16848707e+00 1.18785165e-01 -5.97883821e-01 -1.42578647e-01
3.14562857e-01 4.86499697e-01 -8.23643327e-01 8.90180349e-01
1.89952135e-01 1.36237252e+00 -5.69608629e-01 3.21054101e-01
-8.49521279e-01 3.94827366e-01 -1.60018072e-01 -3.50985378e-01
5.77654004e-01 -1.06748298e-01 8.41327488e-01 1.78528011e+00
-1.28599554e-01 -2.92866472e-02 3.14633399e-01 9.71430719e-01
-1.45422831e-01 5.47786951e-01 -7.78479338e-01 -2.94945925e-01
4.06126469e-01 1.01036799e+00 -5.91252267e-01 -5.17120540e-01
-2.94557363e-01 7.84287095e-01 6.72369897e-01 1.85547486e-01
-2.10544452e-01 -5.07431984e-01 8.01393747e-01 -2.86765974e-02
3.72079104e-01 -3.42555314e-01 -4.76732850e-03 -1.30918109e+00
-3.07074159e-01 -4.40226823e-01 6.34706140e-01 -4.43621039e-01
-1.81434429e+00 1.09450901e+00 -2.02351399e-02 -5.01803756e-01
-3.92137647e-01 -1.23605108e+00 -5.94632886e-02 1.00882447e+00
-1.78255737e+00 -1.03034973e+00 5.60176611e-01 5.62427282e-01
2.02666372e-01 -3.28486264e-01 1.28799021e+00 4.32443649e-01
-2.75637686e-01 2.89238155e-01 -5.60297549e-01 3.02085787e-01
6.55029476e-01 -1.66163933e+00 5.25737286e-01 6.99027598e-01
5.26997387e-01 1.06285930e+00 4.12591249e-01 -8.36530507e-01
-6.65441096e-01 -8.96172285e-01 1.76361120e+00 -5.71173668e-01
1.29210734e+00 -4.65223312e-01 -8.88941824e-01 9.86429155e-01
9.06180203e-01 -2.60351002e-01 9.04747725e-01 5.81468344e-01
-5.53356051e-01 8.10787156e-02 -7.48979032e-01 4.57501322e-01
1.40641451e+00 -8.50124180e-01 -1.63415265e+00 3.82709913e-02
1.19429052e+00 9.27466806e-03 -9.14725244e-01 2.31255755e-01
3.74068096e-02 -6.41708732e-01 1.05123568e+00 -1.18631232e+00
1.93200961e-01 -1.30023003e-01 -4.28366214e-01 -1.88184249e+00
-5.44613302e-01 -2.71836400e-01 2.49829531e-01 1.21386409e+00
7.11660743e-01 -5.97508371e-01 9.04535037e-03 1.15099335e-02
-2.56559730e-01 -2.96022952e-01 -8.04612696e-01 -5.74979007e-01
2.94379979e-01 -6.09457731e-01 5.24997294e-01 1.28959560e+00
3.08022052e-01 9.00588095e-01 8.29331949e-02 -2.11495399e-01
-8.28246679e-03 -3.59908454e-02 -3.37199420e-01 -1.37920654e+00
-7.00191185e-02 -5.83221853e-01 -6.21881545e-01 -8.07242990e-01
8.74056756e-01 -1.54456675e+00 -2.77884841e-01 -1.26796746e+00
-4.12084937e-01 -3.23317647e-01 -9.24718082e-01 6.19718134e-01
-1.81368634e-01 3.56255710e-01 3.11196059e-01 -3.97009850e-01
-6.85157716e-01 2.28099003e-01 6.41237557e-01 1.31992102e-01
-6.82663620e-02 -7.18964994e-01 -6.60890520e-01 7.17929602e-01
6.77970648e-01 -5.71472108e-01 -7.49042928e-02 -8.37634623e-01
9.64874864e-01 -5.18266439e-01 -1.04803443e-01 -7.26051211e-01
3.58053185e-02 2.71135837e-01 1.95228964e-01 -1.36855453e-01
1.83042943e-01 -9.52726603e-01 -4.90883172e-01 5.07406771e-01
-5.67829370e-01 4.79675204e-01 3.44689250e-01 2.30079368e-01
-5.79492807e-01 -3.85493666e-01 4.02532935e-01 -4.20852393e-01
-1.01984441e+00 -1.13455512e-01 -5.44816911e-01 3.57271314e-01
5.08343875e-01 3.71243685e-01 -7.33372569e-02 3.69404614e-01
-1.11326504e+00 -7.09922612e-02 3.35307419e-01 7.39332378e-01
3.89213115e-01 -1.54230392e+00 -3.95287246e-01 4.66577798e-01
3.11695486e-01 -6.09931886e-01 -3.43014926e-01 4.05185968e-01
-2.53922462e-01 7.57265627e-01 -3.17436427e-01 7.64528736e-02
-5.91093063e-01 1.03549802e+00 3.27193916e-01 -3.53425324e-01
-2.30148301e-01 8.43490362e-01 -1.24814600e-01 -7.25374639e-01
1.41569460e-02 -7.25219131e-01 -8.69434774e-01 6.15122795e-01
4.15539443e-01 1.06054351e-01 2.91175336e-01 -7.11920321e-01
-5.88959634e-01 4.23847407e-01 1.03796884e-01 -1.27201751e-01
1.37035513e+00 -1.39585078e-01 -5.74848890e-01 9.35697198e-01
1.04615319e+00 4.65069890e-01 -9.82578248e-02 -6.04276419e-01
7.13265538e-01 3.88967544e-02 -1.00754544e-01 -1.04743588e+00
-7.60489702e-01 7.71721303e-01 2.23808661e-01 5.30278563e-01
7.46279657e-01 1.19201533e-01 8.08206379e-01 5.58432639e-01
1.37234107e-01 -1.06338167e+00 -4.86127257e-01 1.14937854e+00
6.33045197e-01 -8.46651435e-01 -4.50049758e-01 -3.24916750e-01
-4.83639807e-01 1.32877815e+00 3.11914831e-01 -9.56543565e-01
1.06225955e+00 2.98017472e-01 2.51949787e-01 -4.59402531e-01
-5.95906615e-01 -6.91377103e-01 4.91567403e-01 4.77737397e-01
9.62218642e-01 1.86168343e-01 -1.02158761e+00 9.19989169e-01
-4.81642514e-01 -1.41627610e-01 2.52682686e-01 6.97292745e-01
-3.54544908e-01 -1.48084211e+00 3.10520947e-01 7.07641467e-02
-4.58201706e-01 -8.91855419e-01 -8.36395621e-01 9.41915929e-01
4.55753982e-01 4.70857918e-01 2.01176539e-01 -2.29899168e-01
6.80404603e-01 4.85864848e-01 4.40905392e-01 -1.08945668e+00
-1.13304055e+00 -9.44791138e-02 5.13266563e-01 -8.08599293e-01
-9.56522346e-01 -5.47117829e-01 -1.08727837e+00 2.63729215e-01
-1.80115715e-01 1.96207657e-01 4.03123826e-01 1.06487131e+00
1.00154832e-01 7.60483325e-01 1.14119202e-01 -5.97464919e-01
-2.26196870e-01 -1.17188871e+00 -8.12604189e-01 4.75861907e-01
2.36871138e-01 -7.40869462e-01 -1.99471727e-01 -3.42493594e-01] | [10.537095069885254, 9.386603355407715] |
10c84895-5443-4e01-a30d-7191fd22f442 | trimming-the-sail-a-second-order-learning | 2002.06878 | null | http://arxiv.org/abs/2002.06878v1 | http://arxiv.org/pdf/2002.06878v1.pdf | Trimming the Sail: A Second-order Learning Paradigm for Stock Prediction | Nowadays, machine learning methods have been widely used in stock prediction.
Traditional approaches assume an identical data distribution, under which a
learned model on the training data is fixed and applied directly in the test
data. Although such assumption has made traditional machine learning techniques
succeed in many real-world tasks, the highly dynamic nature of the stock market
invalidates the strict assumption in stock prediction. To address this
challenge, we propose the second-order identical distribution assumption, where
the data distribution is assumed to be fluctuating over time with certain
patterns. Based on such assumption, we develop a second-order learning paradigm
with multi-scale patterns. Extensive experiments on real-world Chinese stock
data demonstrate the effectiveness of our second-order learning paradigm in
stock prediction. | [] | 2020-02-17 | null | null | null | null | ['stock-prediction'] | ['time-series'] | [-5.47021925e-01 -6.28623903e-01 -7.05086768e-01 -2.93619305e-01
1.14591956e-01 -6.84060633e-01 6.14915848e-01 -1.00179002e-01
-3.15999359e-01 9.73185539e-01 -3.32223415e-01 -5.68917513e-01
-5.45716658e-02 -1.21264815e+00 -6.22931480e-01 -6.99392259e-01
-2.26649418e-01 5.23856997e-01 4.90192831e-01 -3.04281712e-01
4.78414059e-01 9.00894627e-02 -1.46662128e+00 -1.02689959e-01
6.55661345e-01 1.29153323e+00 9.07273218e-02 4.45782691e-02
-5.32520294e-01 1.16940343e+00 -8.08532119e-01 -4.18650597e-01
9.10715699e-01 -4.09647584e-01 -7.45688006e-02 9.00906771e-02
-9.41163823e-02 -4.85289752e-01 -6.48704946e-01 1.27379036e+00
5.93139306e-02 -1.61069065e-01 6.26747131e-01 -1.55606806e+00
-7.16792881e-01 8.10292482e-01 -7.58817554e-01 7.29384780e-01
-2.81161606e-01 -1.75016876e-02 1.10538447e+00 -5.49218774e-01
-4.58916137e-03 7.10580587e-01 5.08887231e-01 3.24221313e-01
-9.76807475e-01 -1.09865308e+00 5.27855158e-01 2.81928986e-01
-1.14669430e+00 3.27050149e-01 9.89491761e-01 -4.45549846e-01
6.14581704e-01 -6.66638985e-02 1.08031201e+00 6.83074832e-01
6.38007939e-01 1.13052177e+00 1.57998633e+00 1.37081081e-02
3.77894491e-01 3.07540089e-01 1.18264712e-01 -7.65811428e-02
7.47299016e-01 2.21280590e-01 -2.99405396e-01 -1.74498633e-01
8.03641975e-01 6.18472040e-01 -7.66456425e-02 -5.16192578e-02
-6.71879351e-01 9.64392602e-01 7.74759576e-02 5.84357560e-01
-5.88509858e-01 -2.02765986e-01 3.39702249e-01 1.00250614e+00
7.09545135e-01 -1.78117808e-02 -1.04418373e+00 -7.60316178e-02
-9.94491398e-01 5.84825814e-01 8.90511394e-01 7.99781978e-01
6.45940423e-01 4.96466041e-01 3.63308668e-01 4.33134347e-01
1.60256252e-01 5.53791046e-01 1.29314709e+00 4.54845093e-02
3.67091268e-01 6.21870041e-01 2.73657769e-01 -9.63105202e-01
-2.93449163e-01 -8.19885910e-01 -8.44410062e-01 4.18079376e-01
6.25010312e-01 -1.17783174e-01 -6.54140294e-01 1.43578398e+00
1.72904193e-01 8.55076790e-01 2.67401904e-01 5.98398030e-01
-1.01423830e-01 8.52576256e-01 -3.46057057e-01 -6.83933854e-01
6.80834055e-01 -5.22206485e-01 -6.27723515e-01 1.87435985e-01
2.82408625e-01 -3.14528584e-01 8.43023419e-01 6.89330757e-01
-6.05658889e-01 -3.49220037e-01 -9.36644197e-01 1.01032138e+00
-3.46769005e-01 -5.33712447e-01 4.94838208e-01 5.71548760e-01
-6.42916739e-01 8.22774708e-01 -4.80040669e-01 4.80645150e-01
1.09995119e-01 2.92289168e-01 4.52184752e-02 3.36502612e-01
-1.42696035e+00 6.34074569e-01 7.23485172e-01 5.11624524e-03
-2.81417221e-01 -9.47145820e-01 -2.55144745e-01 -7.53751844e-02
4.92554188e-01 -1.77119657e-01 1.50684524e+00 -1.21413505e+00
-1.60699987e+00 3.64379048e-01 4.50385690e-01 -7.46987581e-01
8.46218824e-01 -1.48610294e-01 -6.72007978e-01 -4.11434025e-01
-1.67707771e-01 -5.00998616e-01 1.02416873e+00 -9.91986871e-01
-1.14568698e+00 -1.33922249e-01 -8.36792886e-02 -7.96935260e-02
-3.89759928e-01 -3.58039379e-01 9.35449228e-02 -9.86894965e-01
2.15523094e-01 -7.81889677e-01 -3.55766639e-02 -5.75312316e-01
4.30444703e-02 -4.90991682e-01 9.57212627e-01 -4.11049664e-01
1.49316931e+00 -1.90260148e+00 -4.54168797e-01 4.96456981e-01
-1.60201982e-01 3.26384455e-02 1.22298062e-01 4.80816722e-01
-3.06872398e-01 9.25962627e-03 -1.56373635e-01 -2.67611966e-02
1.24839760e-01 4.58756834e-01 -9.48553562e-01 4.53893006e-01
-4.51517366e-02 7.59009898e-01 -8.67533922e-01 1.92496125e-02
-1.09019734e-01 -2.35429168e-01 -1.12143867e-01 2.08803222e-01
-4.31431472e-01 3.55341770e-02 -5.44434607e-01 5.50971031e-01
8.02815259e-01 -2.78392166e-01 1.81784868e-01 3.26580554e-01
-2.60372788e-01 -9.36127231e-02 -1.46321845e+00 5.84038734e-01
-1.25815660e-01 1.87398821e-01 -5.92813015e-01 -1.40906978e+00
1.02687943e+00 2.25931317e-01 8.08296502e-01 -8.68667603e-01
-1.06014509e-03 4.60771620e-01 3.14593762e-01 -1.45331383e-01
7.96459466e-02 -8.27632964e-01 -9.31961313e-02 7.02910364e-01
-2.92097241e-01 -9.68832225e-02 1.79989442e-01 -3.46111059e-01
8.38694930e-01 -1.01959981e-01 4.68432128e-01 -4.69914526e-01
2.76735038e-01 -9.87998620e-02 1.07711565e+00 4.81025547e-01
-1.99337438e-01 7.18932599e-03 6.07571483e-01 -1.00369096e+00
-1.04208255e+00 -8.40052485e-01 -2.36607313e-01 7.01630235e-01
6.40161484e-02 7.88546056e-02 -1.73082799e-01 -9.38233256e-01
7.12372065e-01 6.81597233e-01 -7.82648981e-01 1.41932992e-02
-2.49569088e-01 -1.20810747e+00 9.42631066e-02 4.75519896e-01
6.29134417e-01 -9.05394256e-01 -5.03334939e-01 5.36691606e-01
5.30356348e-01 -6.64764762e-01 -2.39840761e-01 3.06285657e-02
-9.26278412e-01 -1.13148761e+00 -7.20711708e-01 -3.41602117e-01
3.33621740e-01 1.95058584e-01 1.26768076e+00 2.62126356e-01
1.89117238e-01 2.38026202e-01 -4.04503345e-01 -1.00240362e+00
-1.42416000e-01 -1.12519987e-01 4.27313626e-01 3.38324636e-01
5.86490452e-01 -6.73723221e-01 -5.36819160e-01 3.02629471e-01
-1.32747507e+00 -4.82787669e-01 7.72522628e-01 9.47668314e-01
4.73823637e-01 1.11448526e+00 1.19825208e+00 -8.13658357e-01
5.00249267e-01 -8.82298470e-01 -1.24476051e+00 3.91072035e-01
-1.19026351e+00 1.15695722e-01 7.50734925e-01 -8.97080302e-01
-8.05707872e-01 -3.21380556e-01 3.31673294e-01 -6.71445072e-01
3.86524618e-01 1.11450696e+00 -7.19205216e-02 3.05890769e-01
-2.45950907e-01 9.01920021e-01 -2.96710953e-02 -2.85750270e-01
-4.49429184e-01 3.94868374e-01 3.12548459e-01 -4.82977897e-01
1.40188789e+00 1.95339248e-01 6.79594502e-02 -4.43631858e-01
-8.54684234e-01 -2.61251360e-01 -7.89215922e-01 -1.01159818e-01
2.12775767e-01 -8.65705132e-01 -5.68722486e-01 1.08237505e+00
-3.93310040e-01 -6.29622817e-01 -3.23708713e-01 5.81522644e-01
-4.71629232e-01 -1.66839790e-02 -5.34759402e-01 -1.39279890e+00
-1.96006820e-02 -4.03556824e-01 3.67773920e-01 3.30141127e-01
3.33996087e-01 -1.39042473e+00 5.69564104e-01 -2.81269878e-01
4.50939894e-01 1.14944048e-01 9.14289236e-01 -1.09378839e+00
-6.64274991e-01 -5.67416608e-01 9.75138396e-02 4.56814110e-01
6.34238839e-01 -7.65216723e-02 -4.13015217e-01 -1.86798528e-01
3.22943091e-01 -2.83774614e-01 7.17982829e-01 1.99019983e-01
9.54486012e-01 -3.59893173e-01 9.96134654e-02 1.41591981e-01
1.58202815e+00 6.28687799e-01 4.85286623e-01 7.51177788e-01
-4.82388213e-03 3.92134011e-01 8.01480174e-01 8.31716776e-01
2.36439005e-01 2.85212427e-01 3.78568739e-01 4.01973933e-01
9.24809515e-01 -4.94681269e-01 2.60467112e-01 1.07829142e+00
3.31221186e-02 -5.66461422e-02 -7.82479703e-01 3.22250843e-01
-1.95686376e+00 -1.35105979e+00 3.31355721e-01 2.28693056e+00
1.01699364e+00 6.23364270e-01 5.02692163e-01 4.13578302e-01
3.68002862e-01 1.02127589e-01 -1.01544118e+00 4.83871758e-01
-4.22232985e-01 -8.69577676e-02 6.57001793e-01 5.47279529e-02
-1.19453204e+00 4.92457986e-01 6.24428082e+00 6.91434681e-01
-1.63116562e+00 -3.43254596e-01 7.24485874e-01 -2.54992992e-02
-3.53877783e-01 -2.53818065e-01 -9.11093414e-01 1.15002000e+00
8.01674604e-01 -8.76463592e-01 1.09272428e-01 9.31214631e-01
1.49852782e-01 1.43851131e-01 -8.13798726e-01 9.59875345e-01
-2.34202459e-01 -1.21681368e+00 9.79336724e-02 3.05672348e-01
8.82510185e-01 -7.44586289e-02 3.91030550e-01 7.04274535e-01
3.20744693e-01 -5.35275102e-01 8.69791210e-01 7.82150149e-01
3.31942737e-01 -9.72429633e-01 9.09813106e-01 7.72926092e-01
-1.41922104e+00 -4.79872942e-01 -4.90292251e-01 -3.60828102e-01
-1.06137872e-01 7.42237747e-01 -3.64662737e-01 6.89538002e-01
5.39701462e-01 8.24041605e-01 -1.16923667e-01 1.09755909e+00
1.10971630e-01 8.35411668e-01 -3.69824499e-01 -2.18259916e-01
3.11296552e-01 -4.91038561e-01 3.74958962e-02 6.67140841e-01
5.08915782e-01 2.27462724e-01 4.96710181e-01 3.82559955e-01
1.04630522e-01 2.48010635e-01 -5.88745415e-01 -6.44976869e-02
3.59175980e-01 6.46150410e-01 -9.03623402e-01 -4.52163130e-01
-9.22714770e-01 5.58153927e-01 -1.91107523e-02 2.94257671e-01
-6.80980742e-01 -6.86859414e-02 6.14400804e-01 3.06034118e-01
4.70395178e-01 -2.94631332e-01 -1.15250818e-01 -1.54047763e+00
2.96007544e-01 -7.56999075e-01 5.64318180e-01 9.65560973e-02
-2.09382701e+00 3.24295908e-01 5.62410280e-02 -1.76564479e+00
-4.29713905e-01 -9.32898760e-01 -1.01024151e+00 6.33930624e-01
-1.87408531e+00 -6.23953640e-01 4.08667058e-01 7.07036197e-01
6.22320116e-01 -7.48079181e-01 3.07966322e-01 2.32911646e-01
-6.11074269e-01 3.04126948e-01 5.74399471e-01 2.80181020e-01
2.92076141e-01 -1.52888298e+00 1.81442648e-01 6.02987885e-01
2.75008738e-01 4.12180036e-01 6.70267403e-01 -1.04745710e+00
-1.23435211e+00 -1.14065135e+00 4.93811637e-01 -8.89373571e-02
1.47176921e+00 -3.80643606e-02 -1.38402867e+00 6.86634362e-01
1.38011128e-01 2.51341343e-01 8.02088320e-01 -2.56761074e-01
-4.30945694e-01 -4.08240616e-01 -1.02185476e+00 1.68649241e-01
4.96029407e-01 -2.15189785e-01 -7.76949525e-01 1.30672097e-01
4.12202120e-01 5.32412976e-02 -8.18994999e-01 4.89084393e-01
5.60556471e-01 -6.96696162e-01 4.81721610e-01 -8.48877132e-01
9.32084769e-02 -2.39250749e-01 -1.90792546e-01 -1.46729076e+00
-3.31396669e-01 -5.45524061e-01 -4.29681212e-01 9.33346748e-01
4.16349888e-01 -1.16771507e+00 9.04396236e-01 7.79771984e-01
4.29440320e-01 -7.95495093e-01 -1.06014514e+00 -1.34014642e+00
5.46632290e-01 -1.42156109e-01 9.40212369e-01 1.33208561e+00
-1.24093331e-03 -1.65911138e-01 -5.36814690e-01 1.34407401e-01
5.92312634e-01 6.73210025e-01 7.72422731e-01 -1.63267577e+00
-5.96161604e-01 -6.68199956e-01 -4.87922877e-01 -9.51020122e-01
5.95773816e-01 -3.30413938e-01 -9.65671912e-02 -8.46552193e-01
1.03582770e-01 -6.16299808e-01 -1.01383686e+00 1.70131892e-01
-4.58851516e-01 -1.34432316e-01 2.04184681e-01 6.93682194e-01
-1.99269131e-01 6.60299122e-01 7.93744862e-01 -1.46996260e-01
-1.99022993e-01 8.16065609e-01 -4.33383077e-01 6.32185400e-01
1.07337534e+00 -3.58438909e-01 -7.77163863e-01 -5.24448976e-03
2.03040257e-01 -2.64707834e-01 -1.28494322e-01 -6.88469648e-01
1.69006988e-01 -8.29006195e-01 3.96965384e-01 -9.69391286e-01
-1.61743715e-01 -1.07670093e+00 2.73566633e-01 9.36964393e-01
1.83182463e-01 4.55668420e-01 1.30457848e-01 6.01477087e-01
-5.46490908e-01 -2.15990186e-01 5.02868831e-01 -9.27165449e-02
-8.28134358e-01 7.92590678e-01 -1.20765544e-01 -1.24192782e-01
1.41490257e+00 -9.15982500e-02 -7.25271879e-03 -5.92782080e-01
-4.29905176e-01 2.75050312e-01 1.17612623e-01 4.03119534e-01
5.36776125e-01 -1.48226678e+00 -8.75361264e-01 2.95383364e-01
-6.63083792e-02 -3.79924476e-01 -7.94581510e-03 5.05964041e-01
-1.05866410e-01 3.50012660e-01 -5.79661690e-02 -2.55031079e-01
-5.91685891e-01 9.56153989e-01 4.02315468e-01 -5.09627402e-01
-6.07123017e-01 2.67693639e-01 1.75793797e-01 3.33325155e-02
-5.48077673e-02 -3.88836712e-01 -7.87287503e-02 3.52445126e-01
8.58306885e-01 7.70382758e-04 -1.36319444e-01 -3.23634714e-01
5.85067980e-02 4.03096080e-01 -1.22589730e-01 -6.78557605e-02
1.70350838e+00 2.06237007e-02 6.02300167e-02 1.23978662e+00
6.71566308e-01 -8.09703097e-02 -1.50619817e+00 -6.69178307e-01
6.48159027e-01 -6.33106947e-01 -1.14983298e-01 -3.85887384e-01
-1.31242263e+00 2.96553314e-01 2.95244515e-01 9.21395600e-01
1.04311895e+00 -3.84340644e-01 7.82617331e-01 5.09484053e-01
7.59947956e-01 -1.20175087e+00 4.06970307e-02 4.76307958e-01
7.23650634e-01 -1.25142896e+00 -1.66252181e-01 8.83501768e-02
-4.75475729e-01 1.08416402e+00 6.21969402e-01 -4.87654716e-01
1.38057232e+00 5.77501893e-01 2.43908703e-01 3.40164453e-01
-9.60485518e-01 1.87392626e-02 7.08425120e-02 2.14549214e-01
1.09626502e-01 1.36729656e-02 -2.17214435e-01 7.44687378e-01
-3.20559472e-01 2.13498622e-01 4.32650268e-01 1.21723902e+00
-4.77584004e-01 -1.23735905e+00 -3.05426121e-01 7.25002170e-01
-6.89195871e-01 1.12010024e-01 -1.87195554e-01 1.14339483e+00
-1.34481356e-01 3.74225199e-01 4.67080146e-01 -3.46145689e-01
4.56157178e-01 2.18994007e-01 1.63360357e-01 -5.18400490e-01
-2.97886521e-01 1.73096165e-01 -6.57209396e-01 -1.48076981e-01
-3.66653442e-01 -1.04779983e+00 -1.09470117e+00 -4.11088437e-01
-1.76809326e-01 4.05629635e-01 2.51517177e-01 1.07556546e+00
-1.70776606e-01 1.40553072e-01 1.31953692e+00 -4.47479665e-01
-1.38602638e+00 -8.77703488e-01 -1.35455000e+00 3.76097292e-01
3.98758709e-01 -9.36354458e-01 -4.90255058e-01 1.40537128e-01] | [4.499125957489014, 4.186222076416016] |
dac06c87-d6c1-44fb-927e-052ae03d96e9 | large-context-conversational-representation | 2102.08147 | null | https://arxiv.org/abs/2102.08147v1 | https://arxiv.org/pdf/2102.08147v1.pdf | Large-Context Conversational Representation Learning: Self-Supervised Learning for Conversational Documents | This paper presents a novel self-supervised learning method for handling conversational documents consisting of transcribed text of human-to-human conversations. One of the key technologies for understanding conversational documents is utterance-level sequential labeling, where labels are estimated from the documents in an utterance-by-utterance manner. The main issue with utterance-level sequential labeling is the difficulty of collecting labeled conversational documents, as manual annotations are very costly. To deal with this issue, we propose large-context conversational representation learning (LC-CRL), a self-supervised learning method specialized for conversational documents. A self-supervised learning task in LC-CRL involves the estimation of an utterance using all the surrounding utterances based on large-context language modeling. In this way, LC-CRL enables us to effectively utilize unlabeled conversational documents and thereby enhances the utterance-level sequential labeling. The results of experiments on scene segmentation tasks using contact center conversational datasets demonstrate the effectiveness of the proposed method. | ['Shota Orihashi', 'Tomohiro Tanaka', 'Akihiko Takashima', 'Mana Ihori', 'Naoki Makishima', 'Ryo Masumura'] | 2021-02-16 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 5.75769663e-01 2.79253304e-01 -1.56569645e-01 -7.96645105e-01
-1.00428903e+00 -6.32762074e-01 6.75886452e-01 2.76253253e-01
-2.67403036e-01 5.32244146e-01 5.02156019e-01 -2.01929346e-01
2.96597242e-01 -4.26839054e-01 -2.33359724e-01 -5.68665087e-01
1.42938539e-01 6.81748271e-01 1.26379162e-01 -7.91085213e-02
4.12902385e-01 2.70018160e-01 -1.42674851e+00 8.04665804e-01
7.95909226e-01 6.34256184e-01 7.74779201e-01 8.74099970e-01
-9.49122310e-01 1.11048651e+00 -7.65861273e-01 1.97772354e-01
-3.44073385e-01 -7.62265742e-01 -1.39023125e+00 9.94544566e-01
4.00341935e-02 -2.00322926e-01 2.50600606e-01 7.26054907e-01
6.09046174e-03 5.30051291e-01 6.54577374e-01 -1.06196475e+00
1.15148760e-01 7.86411285e-01 -5.16346872e-01 1.34516358e-01
7.03116775e-01 -4.71361607e-01 8.45779121e-01 -8.75453532e-01
5.75773895e-01 1.64078224e+00 2.45222032e-01 4.83592331e-01
-7.26824224e-01 -2.14745909e-01 5.90061843e-01 -1.36431614e-02
-1.11642170e+00 -3.08750182e-01 8.72983098e-01 -3.66953880e-01
9.48327303e-01 2.98105419e-01 3.25920612e-01 6.37123346e-01
-9.05915648e-02 1.28142428e+00 1.14102435e+00 -8.47784102e-01
3.16250235e-01 4.05963659e-01 7.22916305e-01 6.08917892e-01
-5.62770367e-01 -6.87888920e-01 -2.62481153e-01 -3.49979848e-01
5.22137821e-01 7.20232427e-02 -2.06055446e-03 1.76323589e-03
-8.48350644e-01 7.16107547e-01 -2.12453872e-01 6.74809575e-01
-2.55318403e-01 -3.15000325e-01 7.99626768e-01 7.62000978e-02
7.86891639e-01 -1.25514332e-03 -3.08695376e-01 -2.60153323e-01
-7.18147099e-01 1.22088425e-01 9.45589662e-01 1.21354151e+00
8.87693048e-01 -1.95825875e-01 5.97959198e-02 1.45191276e+00
3.14538181e-01 1.78243816e-01 6.03069961e-01 -9.49482262e-01
4.75870043e-01 7.31045306e-01 9.66433138e-02 -7.11746871e-01
-5.00407815e-01 1.29133627e-01 -5.48940301e-01 -5.28022170e-01
2.17500046e-01 -2.93930352e-01 -5.68276942e-01 1.34809303e+00
6.10982239e-01 7.48100802e-02 5.51218271e-01 6.84293807e-01
9.63558078e-01 1.11908448e+00 2.68595755e-01 -8.28119218e-01
1.43603313e+00 -1.48765540e+00 -1.13795269e+00 -3.10769767e-01
7.99564362e-01 -9.32135880e-01 1.18525577e+00 1.06482528e-01
-6.55714631e-01 -6.56135976e-01 -6.02238953e-01 -2.64422782e-02
-3.53497326e-01 3.26016963e-01 6.47247136e-01 5.01409292e-01
-8.14135671e-01 -1.52916625e-01 -5.33986688e-01 -5.73261261e-01
-2.09551096e-01 1.48296505e-01 -8.49934891e-02 -1.67897657e-01
-9.23986316e-01 5.39820075e-01 4.33893323e-01 6.86093792e-02
-7.14758575e-01 9.00311098e-02 -1.14368224e+00 4.50868905e-02
6.75273538e-01 1.96580067e-01 1.80364144e+00 -1.06905627e+00
-1.50416553e+00 8.97748649e-01 -8.78824592e-01 -2.43638128e-01
1.89497575e-01 -2.88566738e-01 -3.88323992e-01 4.06529963e-01
3.29758018e-01 5.20638406e-01 8.61400545e-01 -1.65360487e+00
-7.94888318e-01 -2.89323717e-01 5.48660830e-02 6.81926310e-01
-1.15235001e-01 2.25336403e-02 -7.18619108e-01 -3.44002306e-01
4.12026852e-01 -1.00516737e+00 -3.97884071e-01 -9.74153280e-01
-4.20510978e-01 -9.01645958e-01 1.48704875e+00 -5.69477081e-01
8.17035913e-01 -2.08794379e+00 -2.82481700e-01 -6.56575412e-02
-2.41408590e-02 -2.18497980e-02 -1.40715152e-01 6.81061327e-01
7.33100250e-02 -2.12464854e-01 -3.77888918e-01 -7.80626297e-01
-3.13405007e-01 2.26670757e-01 -4.22482163e-01 4.55280878e-02
-1.09341316e-01 5.86346388e-01 -1.03422618e+00 -9.76466835e-01
3.79744112e-01 2.21055791e-01 -1.47492304e-01 7.28387177e-01
-5.74295521e-01 8.89160991e-01 -7.12354362e-01 5.16842961e-01
4.55371529e-01 -7.37435520e-02 5.50471604e-01 1.42212827e-02
-2.55856216e-01 4.39656764e-01 -9.70830202e-01 1.68651772e+00
-8.30400169e-01 6.46975696e-01 2.87696242e-01 -1.09612119e+00
9.65175390e-01 5.11480689e-01 3.88100415e-01 -1.95719704e-01
1.37022272e-01 -1.73837125e-01 -5.48238158e-01 -8.33319187e-01
8.52262914e-01 -8.65175426e-02 -5.52235544e-01 1.03434074e+00
5.69038000e-03 -4.61154491e-01 2.45540887e-01 5.06687760e-01
5.80841541e-01 -3.75530161e-02 5.46963394e-01 -5.73505089e-02
9.32433784e-01 1.56773537e-01 4.94315416e-01 6.12387955e-01
-2.71660954e-01 3.67419958e-01 4.28903043e-01 -1.98864043e-01
-5.86612284e-01 -4.26704079e-01 1.42657802e-01 1.49547577e+00
7.85089955e-02 -3.43696862e-01 -1.22683024e+00 -7.49246180e-01
-6.50620162e-01 8.13585877e-01 -1.19214676e-01 4.49903011e-01
-6.61386490e-01 -3.67388636e-01 1.11817054e-01 4.51790780e-01
7.02336788e-01 -1.29078531e+00 -2.11897492e-01 3.54158461e-01
-5.87312937e-01 -1.69963479e+00 -6.24382138e-01 -9.43153054e-02
-8.21868718e-01 -9.10361469e-01 -4.81466442e-01 -1.39774191e+00
9.72811699e-01 7.61877716e-01 7.00161159e-01 -5.29471226e-03
-1.86914504e-01 7.86556184e-01 -6.37138724e-01 -2.04442874e-01
-8.07383060e-01 -4.92172800e-02 4.06835862e-02 2.35068113e-01
5.24465680e-01 -3.31005484e-01 -8.07289854e-02 4.59142745e-01
-7.80638456e-01 1.24241538e-01 2.15500981e-01 7.13007092e-01
5.58028340e-01 4.38030839e-01 1.03261423e+00 -1.46618414e+00
9.29937840e-01 -1.95906505e-01 -3.80337119e-01 4.69063073e-01
2.73618083e-02 -2.99757659e-01 5.36143899e-01 -3.24938774e-01
-1.86870635e+00 2.99535006e-01 -5.21063432e-02 9.85344946e-02
-6.17584586e-01 4.04694140e-01 -3.74956042e-01 4.16096687e-01
1.93575010e-01 3.77625495e-01 -2.32936725e-01 -3.69123697e-01
6.19647026e-01 1.41597366e+00 5.85620940e-01 -9.10802007e-01
2.05419045e-02 3.45422477e-01 -7.01714516e-01 -1.47151566e+00
-1.03704667e+00 -1.14862823e+00 -8.84589255e-01 -5.22697091e-01
1.11058569e+00 -9.40485120e-01 -7.36055136e-01 4.68883991e-01
-1.27037179e+00 -3.77893925e-01 -7.64694512e-02 5.65340519e-01
-6.38269961e-01 7.87394881e-01 -9.37103033e-01 -1.53138971e+00
-2.85600692e-01 -1.11035943e+00 1.11919272e+00 4.02484775e-01
-5.09090960e-01 -1.12619650e+00 -6.70413747e-02 1.08989108e+00
-9.96270254e-02 -8.50581527e-02 8.66626620e-01 -1.00513089e+00
-3.02232355e-01 -2.49933541e-01 -1.36081859e-01 3.50463271e-01
5.35457194e-01 -3.21098596e-01 -1.23891640e+00 -1.69197351e-01
4.05835956e-01 -6.73006713e-01 5.27208567e-01 2.08998948e-01
1.10241485e+00 -2.63762295e-01 -2.89991587e-01 -4.29343849e-01
9.41107631e-01 6.51555300e-01 1.64449945e-01 -2.35750347e-01
7.89809048e-01 1.23674333e+00 1.15181935e+00 2.73527741e-01
5.23320317e-01 2.24977791e-01 -2.19385594e-01 -7.17634931e-02
3.10626864e-01 -2.58861601e-01 1.04414053e-01 1.37166488e+00
2.40515947e-01 -3.68707806e-01 -8.53184044e-01 6.61262751e-01
-2.06093526e+00 -6.81432605e-01 -2.78509483e-02 1.55546486e+00
8.03394973e-01 1.27388358e-01 3.72396186e-02 -1.20724849e-02
1.07543135e+00 3.39317501e-01 -4.58585054e-01 -5.64273715e-01
2.14259639e-01 -5.31752348e-01 -1.76688612e-01 8.51043463e-01
-1.20424926e+00 1.33438349e+00 6.09443569e+00 7.83892751e-01
-7.72657692e-01 3.18175524e-01 7.88612306e-01 4.69725162e-01
-3.50157022e-02 -2.95028221e-02 -9.97927070e-01 8.04856718e-02
8.46783817e-01 9.93521363e-02 1.38829768e-01 1.26128006e+00
6.00655317e-01 -6.62940860e-01 -1.06562495e+00 8.81681085e-01
4.79251742e-01 -1.24148631e+00 -9.44863483e-02 -3.81738782e-01
8.19445252e-01 -4.14642066e-01 -4.61258084e-01 5.36291182e-01
2.14589819e-01 -5.40873110e-01 9.02201906e-02 -3.68498228e-02
4.79214370e-01 -8.55552793e-01 8.22337568e-01 7.51393199e-01
-1.40485466e+00 1.06509194e-01 -3.31735224e-01 -4.91380915e-02
4.84825104e-01 5.43819308e-01 -1.39487123e+00 2.82353908e-01
5.42953134e-01 4.63637918e-01 6.25652000e-02 2.52664775e-01
-1.69475332e-01 7.47156680e-01 -6.72142655e-02 -7.15268254e-02
4.26905274e-01 -5.79394162e-01 3.55961710e-01 1.82396317e+00
-2.91057140e-01 5.35071671e-01 8.78453732e-01 4.04203415e-01
-7.75010064e-02 4.82767284e-01 -5.93854010e-01 -3.56486678e-01
5.31619132e-01 1.27639759e+00 -1.21593606e+00 -7.81533897e-01
-4.17475224e-01 9.72049296e-01 1.70434758e-01 3.31633538e-01
-2.89934516e-01 -1.74141571e-01 7.53938481e-02 -2.89548755e-01
-5.23868501e-02 -3.81078035e-01 -4.83528711e-02 -8.37457776e-01
1.07475091e-02 -8.41475606e-01 2.87182599e-01 -5.87460637e-01
-1.14030921e+00 7.69415259e-01 2.22195178e-01 -1.04471076e+00
-5.86268425e-01 -1.19034611e-01 -7.66150951e-01 4.73741025e-01
-1.17138696e+00 -1.22394347e+00 -2.65771866e-01 4.37777013e-01
1.56468904e+00 -2.05351233e-01 9.92804527e-01 -1.10247180e-01
-6.37573421e-01 8.46970379e-02 -1.45106360e-01 1.81139588e-01
5.86047232e-01 -1.10543907e+00 1.88312665e-01 4.50184315e-01
4.86388318e-02 6.39892101e-01 4.81517732e-01 -7.75401235e-01
-1.02294624e+00 -9.62551296e-01 1.08565855e+00 -1.05490103e-01
3.99512082e-01 -5.94266653e-01 -1.08139849e+00 7.59522557e-01
4.56661195e-01 -5.62161386e-01 1.08286810e+00 1.84300855e-01
5.80830425e-02 8.79162401e-02 -1.06560457e+00 4.43997294e-01
6.06220126e-01 -1.05843687e+00 -9.88782585e-01 7.88757920e-01
1.23293853e+00 -3.93462896e-01 -5.65115929e-01 1.49308801e-01
2.66204812e-02 -6.14791930e-01 5.60753942e-01 -1.58249378e-01
1.07386179e-01 -1.45433396e-01 -1.86865076e-01 -1.00199687e+00
4.80062813e-01 -6.11133933e-01 2.50532836e-01 1.61762786e+00
2.05193996e-01 -3.79904300e-01 7.91243851e-01 6.53973758e-01
-1.89135894e-01 -4.92700934e-01 -4.96029288e-01 -4.00184333e-01
-4.15369600e-01 -5.83613932e-01 3.21599096e-01 1.20907211e+00
6.07093036e-01 8.22343349e-01 -3.54036748e-01 1.39033645e-01
5.19110322e-01 4.85325456e-01 8.28887165e-01 -8.02947223e-01
-7.97139853e-02 2.03410059e-01 1.55420676e-01 -1.62513268e+00
6.71208620e-01 -3.21170628e-01 6.36498511e-01 -1.63798177e+00
1.90275073e-01 -3.89059037e-01 3.27204525e-01 2.37472519e-01
-1.59741879e-01 -3.04842979e-01 1.10740829e-02 4.85423863e-01
-9.32068944e-01 4.15339470e-01 1.06586134e+00 -2.43944347e-01
-7.36254632e-01 3.76822829e-01 -2.22608894e-01 1.04618812e+00
8.35320473e-01 -2.37278491e-01 -8.17539752e-01 8.59824866e-02
-7.16823578e-01 4.59864438e-01 -3.93450141e-01 -5.07515490e-01
4.37843353e-01 -2.83646494e-01 -1.75349578e-01 -1.25390494e+00
5.58326304e-01 -6.18260324e-01 -5.53488016e-01 5.70266694e-02
-7.30684638e-01 -3.00405145e-01 7.19847679e-02 7.19648600e-01
-5.83996236e-01 -3.48404527e-01 5.63013613e-01 -3.70159328e-01
-1.07778895e+00 -3.08348089e-01 -1.08721817e+00 -1.12140343e-01
1.02036929e+00 -3.44342389e-03 -5.65319955e-02 -7.98567951e-01
-9.03881371e-01 4.78789449e-01 1.34460330e-02 4.14785266e-01
6.39064491e-01 -7.59455442e-01 -4.48293507e-01 -3.51024745e-03
2.06883326e-01 4.29594725e-01 5.15209556e-01 3.00545871e-01
-4.17345047e-01 7.70141304e-01 3.55118513e-01 -8.87479424e-01
-1.57762969e+00 3.54201436e-01 -8.48955214e-02 -3.96744847e-01
-5.44828951e-01 7.45596230e-01 5.16977191e-01 -9.10513282e-01
6.23942196e-01 -2.42562056e-01 -7.43072033e-01 1.92457736e-01
7.43400216e-01 1.00704826e-01 -1.78670987e-01 -8.19604814e-01
-1.40619725e-01 3.91268075e-01 -4.43057626e-01 -3.76667082e-01
9.80935335e-01 -7.11772859e-01 -3.74608994e-01 9.91405129e-01
1.18973064e+00 -1.94198452e-02 -1.17273784e+00 -5.69235504e-01
2.26533353e-01 -1.30165920e-01 -2.69275904e-01 -3.58035862e-01
-5.18257737e-01 7.31855094e-01 2.19605997e-01 3.26622993e-01
8.56994927e-01 1.69142693e-01 9.31562066e-01 9.32749569e-01
5.25430024e-01 -1.40796280e+00 4.13725317e-01 6.76010191e-01
7.56039739e-01 -1.47446573e+00 -6.31851852e-02 -1.01845706e+00
-1.06010854e+00 1.00691378e+00 7.64425397e-01 3.28057051e-01
5.32566607e-01 1.97423741e-01 5.45826554e-01 -9.13981423e-02
-5.37132859e-01 -2.44032502e-01 -1.01885788e-01 3.72422397e-01
5.26768446e-01 1.11910254e-01 -3.20326000e-01 4.06489104e-01
-4.07265179e-04 -4.42320198e-01 6.10779643e-01 1.25961208e+00
-7.71856964e-01 -1.15098357e+00 -5.34775972e-01 6.73593208e-02
-1.65143296e-01 1.70061782e-01 -6.25552237e-01 4.90997434e-01
-2.42974341e-01 1.70224655e+00 2.62787670e-01 -1.48527905e-01
-1.28869014e-03 3.94521922e-01 -3.23628671e-02 -1.14518368e+00
-5.05751252e-01 6.87412500e-01 4.96197671e-01 -1.19760692e-01
-9.44857597e-01 -5.67439616e-01 -1.86664748e+00 2.22118229e-01
-4.34311986e-01 8.95670056e-01 8.51166070e-01 1.47198558e+00
-1.71639591e-01 4.51432139e-01 9.51492667e-01 -8.92350852e-01
5.59247397e-02 -1.11879838e+00 -5.63020170e-01 3.81976545e-01
2.13402003e-01 -1.19107708e-01 -2.94447631e-01 3.99160653e-01] | [12.613600730895996, 7.58472204208374] |
c892551d-9a48-493b-aec2-3a35607fd935 | laplace-redux-effortless-bayesian-deep | 2106.14806 | null | https://arxiv.org/abs/2106.14806v3 | https://arxiv.org/pdf/2106.14806v3.pdf | Laplace Redux -- Effortless Bayesian Deep Learning | Bayesian formulations of deep learning have been shown to have compelling theoretical properties and offer practical functional benefits, such as improved predictive uncertainty quantification and model selection. The Laplace approximation (LA) is a classic, and arguably the simplest family of approximations for the intractable posteriors of deep neural networks. Yet, despite its simplicity, the LA is not as popular as alternatives like variational Bayes or deep ensembles. This may be due to assumptions that the LA is expensive due to the involved Hessian computation, that it is difficult to implement, or that it yields inferior results. In this work we show that these are misconceptions: we (i) review the range of variants of the LA including versions with minimal cost overhead; (ii) introduce "laplace", an easy-to-use software library for PyTorch offering user-friendly access to all major flavors of the LA; and (iii) demonstrate through extensive experiments that the LA is competitive with more popular alternatives in terms of performance, while excelling in terms of computational cost. We hope that this work will serve as a catalyst to a wider adoption of the LA in practical deep learning, including in domains where Bayesian approaches are not typically considered at the moment. | ['Philipp Hennig', 'Matthias Bauer', 'Runa Eschenhagen', 'Alexander Immer', 'Agustinus Kristiadi', 'Erik Daxberger'] | 2021-06-28 | laplace-redux-effortless-bayesian-deep-1 | http://proceedings.neurips.cc/paper/2021/hash/a7c9585703d275249f30a088cebba0ad-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/a7c9585703d275249f30a088cebba0ad-Paper.pdf | neurips-2021-12 | ['misconceptions'] | ['miscellaneous'] | [-1.95319295e-01 5.36134802e-02 1.77595124e-01 -5.21418393e-01
-9.52996790e-01 -5.36663532e-01 7.06503570e-01 5.36192618e-02
-6.04771197e-01 9.54221010e-01 -1.74217373e-01 -5.78860700e-01
-4.02235925e-01 -4.77144480e-01 -8.55758429e-01 -9.59411979e-01
-1.71794802e-01 5.10463893e-01 3.02510440e-01 1.71941444e-01
2.54725575e-01 5.88904202e-01 -1.74344671e+00 -3.07969987e-01
5.97703397e-01 1.17299163e+00 4.82411645e-02 3.41591299e-01
8.32993090e-02 6.01483285e-01 -2.11713701e-01 -7.05875576e-01
8.26933309e-02 -1.73504800e-01 -6.25208139e-01 -4.42647725e-01
3.91445160e-01 -5.64452350e-01 -9.73779559e-02 1.21178865e+00
5.26621401e-01 1.12304300e-01 8.05270076e-01 -1.12764716e+00
-1.08156493e-02 6.06562614e-01 -5.67236900e-01 2.92547584e-01
-1.60526529e-01 -8.74752775e-02 1.02644897e+00 -7.78732002e-01
1.44942671e-01 1.26461744e+00 1.01949477e+00 4.43213612e-01
-1.53065908e+00 -6.34462595e-01 3.79306376e-02 1.12729058e-01
-1.49602580e+00 -7.33211637e-01 4.79906052e-01 -4.44769382e-01
9.82615888e-01 9.48808566e-02 4.95109051e-01 1.24808204e+00
4.24264252e-01 8.71495187e-01 1.14847529e+00 -3.67121786e-01
6.61419034e-01 3.60234171e-01 2.92927176e-01 6.31861269e-01
3.77022177e-01 1.67164549e-01 -4.29300189e-01 -5.40876985e-01
5.51751316e-01 -3.86682034e-01 -2.15925828e-01 -5.00210524e-01
-6.28078818e-01 1.00528359e+00 2.21951291e-01 5.64188585e-02
-4.22265649e-01 6.66208506e-01 3.95035177e-01 -7.93755893e-03
6.53131068e-01 9.04148892e-02 -2.81853616e-01 -5.05847573e-01
-1.21099198e+00 7.32093930e-01 9.83606279e-01 6.63711190e-01
4.35183734e-01 6.30489364e-02 2.96657294e-01 8.46758127e-01
7.51962006e-01 1.20672196e-01 1.50237575e-01 -1.37301338e+00
-1.72291044e-02 -3.02309453e-01 2.90229172e-01 -6.96623743e-01
-4.05364931e-01 -5.92255890e-01 -7.49458015e-01 4.38194662e-01
5.81726789e-01 -1.64443448e-01 -5.17990470e-01 1.82082808e+00
1.08021202e-04 7.79892355e-02 -8.50461870e-02 6.79898679e-01
6.13448203e-01 5.47881544e-01 1.16714373e-01 -1.91893622e-01
1.16991985e+00 -2.14583412e-01 -5.73412836e-01 -2.08495960e-01
2.57357717e-01 -5.75831950e-01 7.68415034e-01 7.68831730e-01
-1.23154402e+00 -1.59455150e-01 -1.13275027e+00 9.05883759e-02
-2.29005128e-01 -1.10403411e-01 9.80767548e-01 1.04131174e+00
-1.16216218e+00 1.04680395e+00 -1.38400960e+00 -2.97020197e-01
6.89998686e-01 3.29019397e-01 5.00324070e-02 1.60336807e-01
-1.06111062e+00 1.05235744e+00 3.77231985e-01 1.91716149e-01
-8.12030554e-01 -7.84214735e-01 -7.81599104e-01 2.26945758e-01
2.94559479e-01 -5.89425743e-01 1.60283625e+00 -6.23615563e-01
-1.57575309e+00 5.18981218e-01 -3.34592611e-02 -6.58875644e-01
8.30845296e-01 -4.29811507e-01 1.94814697e-01 1.79889463e-02
-2.17294976e-01 6.33180559e-01 6.48059130e-01 -9.62558687e-01
-3.64278704e-01 -3.95445794e-01 8.87179747e-02 8.63822475e-02
-1.26088262e-01 -7.92859048e-02 -3.25033575e-01 -4.49131697e-01
2.13878065e-01 -9.07464445e-01 -2.20035017e-01 1.74843580e-01
-2.50362307e-01 -4.42744404e-01 2.79566169e-01 -4.15839046e-01
9.00481939e-01 -2.03318262e+00 -2.26573259e-01 1.78213686e-01
2.03647494e-01 1.81388840e-01 3.70394170e-01 3.93746734e-01
9.70485583e-02 1.50001228e-01 -4.34982777e-01 -4.67425555e-01
3.97689939e-01 2.90462554e-01 -3.50347668e-01 7.31793463e-01
1.07574821e-01 5.79759300e-01 -7.84497380e-01 -3.64780873e-01
3.43890309e-01 7.81603336e-01 -5.02007544e-01 -2.82810092e-01
-1.38321221e-01 3.71187776e-02 -3.69215548e-01 2.44413629e-01
7.32566953e-01 -1.25502780e-01 1.14914834e-01 -2.43959993e-01
-1.77948371e-01 6.22371376e-01 -1.39320421e+00 1.30547392e+00
-1.51544228e-01 8.96698534e-01 7.48219118e-02 -1.06781852e+00
5.45890808e-01 3.19689065e-01 2.36286134e-01 8.56986793e-04
1.39119521e-01 3.29275787e-01 -1.25919702e-02 -1.01197988e-01
2.49015182e-01 -3.04742038e-01 3.03078681e-01 5.15167713e-01
2.35383913e-01 -3.22106749e-01 1.17809184e-01 1.87492028e-01
8.46459270e-01 5.60750365e-01 3.28315884e-01 -7.61184692e-01
3.05417906e-02 -3.22329044e-01 3.68459791e-01 1.10084009e+00
-2.82097578e-01 4.37950313e-01 7.75560856e-01 -1.86355397e-01
-9.09877002e-01 -1.27839887e+00 -7.15615690e-01 8.62339616e-01
-3.08177561e-01 -1.56486258e-01 -8.41296315e-01 -1.49836600e-01
-9.15556699e-02 9.80339170e-01 -4.13400024e-01 7.65047073e-02
-4.89112325e-02 -1.12358093e+00 6.44031405e-01 6.78111255e-01
3.41560066e-01 -6.73903644e-01 -8.36515069e-01 1.27472579e-01
2.42821380e-01 -8.60783935e-01 2.13211760e-01 6.00976825e-01
-1.06734872e+00 -7.01313198e-01 -7.47389615e-01 -1.62787870e-01
2.15750024e-01 -1.01191677e-01 1.00363088e+00 -2.50147492e-01
6.54568970e-02 3.33548486e-01 1.80722065e-02 -5.42609811e-01
-2.94017822e-01 -8.10815841e-02 2.92730480e-01 -3.92507702e-01
5.82065403e-01 -7.51627743e-01 -4.56823200e-01 -1.61284029e-01
-8.97196770e-01 -1.77273020e-01 4.06453878e-01 7.65889168e-01
3.67810428e-01 2.63566762e-01 4.94170338e-01 -9.34695661e-01
7.79325306e-01 -4.92426366e-01 -1.05651581e+00 -5.29778153e-02
-8.49994302e-01 2.41798252e-01 2.49496460e-01 -8.62363800e-02
-1.19864416e+00 -9.14741978e-02 -5.46186209e-01 -3.19925398e-01
-1.92928419e-01 7.43105710e-01 1.62617475e-01 -9.98698026e-02
7.14304566e-01 -6.71042595e-03 2.54456382e-02 -5.76871157e-01
1.52122691e-01 5.70984244e-01 3.63009572e-01 -7.91992128e-01
4.08412784e-01 3.31699312e-01 1.97265372e-01 -8.87056828e-01
-8.03011715e-01 -6.80551752e-02 -4.13299680e-01 -7.24709332e-02
5.56084931e-01 -7.24261940e-01 -7.80373693e-01 4.29235876e-01
-1.07524538e+00 -3.58411759e-01 5.04344031e-02 6.38399959e-01
-7.26730108e-01 4.00951266e-01 -7.68934309e-01 -1.13391709e+00
-1.27292216e-01 -1.33783388e+00 6.96489036e-01 4.25317317e-01
-7.64082149e-02 -1.10754228e+00 -1.41072452e-01 2.27411557e-02
4.12512332e-01 9.16369706e-02 8.49211693e-01 -6.27264917e-01
-2.66303509e-01 -3.05954516e-01 -2.73904115e-01 5.71660936e-01
-3.53267282e-01 4.54813719e-01 -1.48650479e+00 -2.15078443e-01
1.23225756e-01 -3.85008514e-01 1.07055807e+00 8.34931791e-01
1.27541554e+00 -4.94557433e-02 -1.53150976e-01 4.92036670e-01
1.47416770e+00 9.11711156e-02 5.59737921e-01 2.66329825e-01
1.40646279e-01 7.95642018e-01 1.63882747e-01 5.78183413e-01
2.56694824e-01 5.63269556e-01 4.23856556e-01 4.03732181e-01
3.00482512e-01 4.80070477e-04 3.54564905e-01 6.38243735e-01
-1.69085518e-01 1.23430118e-02 -9.65396404e-01 3.44207942e-01
-1.85387862e+00 -9.68262851e-01 -2.29898736e-01 2.30652404e+00
7.66472042e-01 5.25929093e-01 7.24186525e-02 6.71344697e-02
2.90505916e-01 -2.31123827e-02 -5.47448695e-01 -5.13326705e-01
1.35350406e-01 3.49949330e-01 5.28951824e-01 5.12194932e-01
-1.09966266e+00 6.02210641e-01 7.02675581e+00 1.09055638e+00
-7.23309815e-01 3.91687185e-01 6.79896355e-01 -2.31656224e-01
-2.19956979e-01 -3.40455137e-02 -8.89084160e-01 3.66638660e-01
1.18166268e+00 -5.44493832e-02 3.65180820e-01 9.42497611e-01
3.78038436e-02 -4.12399143e-01 -1.30324578e+00 7.77976215e-01
-3.24301630e-01 -1.36038959e+00 -4.51644719e-01 2.07006171e-01
4.55549926e-01 5.74137449e-01 1.04131840e-01 3.79035056e-01
3.41717452e-01 -1.04721129e+00 9.30575252e-01 4.55654889e-01
3.33205968e-01 -1.04031277e+00 8.79803896e-01 4.34357345e-01
-6.58959448e-01 2.26871297e-01 -7.78307498e-01 -4.26364802e-02
6.63249269e-02 8.85987043e-01 -3.47646236e-01 2.72389948e-01
9.31356311e-01 4.18667585e-01 -3.46164912e-01 1.12145567e+00
-8.43748674e-02 7.82580853e-01 -7.49038279e-01 -2.24964291e-01
4.85682696e-01 -1.13507912e-01 5.88245451e-01 1.34472394e+00
2.92575419e-01 -1.32123232e-01 -3.60749036e-01 1.24831402e+00
2.51866311e-01 -3.11335415e-01 -6.30599022e-01 -1.51132286e-01
6.18235350e-01 1.04727745e+00 -7.79331505e-01 -1.55772805e-01
-5.20994484e-01 4.76362795e-01 3.19007367e-01 3.87311012e-01
-7.76042640e-01 -3.60135198e-01 6.64182842e-01 -1.44447938e-01
3.83836299e-01 -1.12234019e-01 -3.04792553e-01 -8.84492338e-01
-1.81551293e-01 -7.25117803e-01 1.40126377e-01 -7.97891021e-01
-1.39841020e+00 4.90657300e-01 5.18171430e-01 -7.34291852e-01
-3.92850339e-01 -1.08276892e+00 -3.11933905e-01 9.49785888e-01
-1.23793101e+00 -4.76975530e-01 1.72740564e-01 2.08308175e-01
2.22722024e-01 5.70724383e-02 8.30492318e-01 1.73945129e-01
-5.37196577e-01 5.74346066e-01 5.27785003e-01 -2.82424212e-01
2.88968921e-01 -1.36366713e+00 2.12973133e-01 6.34271562e-01
2.15354607e-01 8.77230287e-01 1.21333551e+00 -2.24324927e-01
-1.19287360e+00 -5.36571681e-01 4.30957228e-01 -4.30116266e-01
7.69899487e-01 -2.57363081e-01 -1.00502229e+00 7.64689982e-01
9.33230668e-03 -2.80817151e-01 7.00399280e-01 3.57194543e-01
-2.44982824e-01 -4.96189333e-02 -1.20221460e+00 5.53403258e-01
4.28130537e-01 -4.59246159e-01 -5.53869784e-01 2.69857615e-01
2.15130076e-01 -2.42953911e-01 -9.69971716e-01 4.27527755e-01
1.01265287e+00 -1.46330440e+00 7.88145125e-01 -2.31317878e-01
3.31051886e-01 6.69274330e-02 -4.62142050e-01 -1.07509542e+00
-6.23459183e-02 -5.75005829e-01 -2.44572997e-01 1.22039843e+00
3.36277753e-01 -9.46274400e-01 7.06286252e-01 9.35274601e-01
-2.28790417e-02 -1.01812458e+00 -1.30268741e+00 -7.50134528e-01
4.96286571e-01 -1.03869474e+00 2.83451885e-01 5.21825612e-01
-1.66742787e-01 6.33812323e-02 -1.68382317e-01 6.24599569e-02
8.80903304e-01 -2.64594972e-01 3.76140326e-01 -1.52409458e+00
-5.00109076e-01 -9.50857341e-01 -3.79816592e-01 -1.05660510e+00
5.10006323e-02 -5.90732396e-01 2.44776145e-01 -1.21609318e+00
3.30963731e-01 -5.01086593e-01 -4.17101324e-01 2.48400614e-01
1.32270947e-01 3.55528653e-01 -1.39540687e-01 1.59386218e-01
-2.29919121e-01 5.13087809e-01 5.29657841e-01 2.34292850e-01
1.46801203e-01 2.59244144e-01 -7.82310128e-01 1.22351933e+00
8.80819023e-01 -5.64234734e-01 -4.01932836e-01 -2.99603224e-01
4.19454068e-01 -1.22829393e-01 6.02057278e-01 -1.00634038e+00
3.46302837e-02 8.67009461e-02 3.69724005e-01 -5.93161404e-01
6.43098056e-01 -5.31353235e-01 1.45128682e-01 3.81144851e-01
-3.41015965e-01 -3.05289984e-01 5.29743254e-01 5.78766763e-01
1.71726510e-01 -9.02000368e-01 1.02711248e+00 -7.83188939e-02
-4.59882885e-01 1.01484746e-01 -5.77936053e-01 -1.47564858e-01
6.42382383e-01 -8.51592049e-02 -1.51440352e-01 -5.98516762e-01
-5.01094997e-01 -1.73160270e-01 3.10696721e-01 -3.30874547e-02
3.10004473e-01 -9.90402460e-01 -5.13932467e-01 -6.47409484e-02
-2.02969790e-01 -1.00644231e-01 7.91870803e-02 1.11304402e+00
-6.44867301e-01 5.92441857e-01 8.92144814e-02 -7.02060938e-01
-8.76394510e-01 1.07363656e-01 4.89264309e-01 1.47749204e-02
-7.86843777e-01 1.19374287e+00 1.43678382e-01 -1.87947914e-01
6.98903918e-01 -8.37419555e-02 1.31537272e-02 -2.35746950e-02
5.78913212e-01 4.58484173e-01 3.45675647e-02 -3.57411355e-01
-4.61936325e-01 1.62903637e-01 -5.64969555e-02 -4.27099377e-01
1.43070900e+00 -8.17402750e-02 -7.97954947e-02 8.11164260e-01
8.03982139e-01 -3.32718283e-01 -1.53635228e+00 -1.60738498e-01
2.10330170e-02 -2.17823178e-01 6.22741818e-01 -5.96647024e-01
-8.29643667e-01 1.17732739e+00 7.75341809e-01 3.22090924e-01
8.88567805e-01 2.84300242e-02 2.85935372e-01 5.88901281e-01
5.03052473e-01 -9.04157519e-01 -5.43163121e-01 2.38282278e-01
6.51591539e-01 -1.01022100e+00 3.60654294e-01 1.17084913e-01
-4.11089659e-01 1.13783181e+00 2.34909251e-01 -2.25720733e-01
9.78976488e-01 4.39649671e-01 -8.24688748e-02 -1.00542024e-01
-8.96801174e-01 1.15391286e-02 1.61697581e-01 4.83434677e-01
6.64431691e-01 -1.29142061e-01 -3.01282406e-01 5.31237721e-01
-2.79329866e-01 5.84213063e-02 3.74290913e-01 5.76217651e-01
-3.88298094e-01 -8.48647594e-01 -2.32293382e-01 7.48308241e-01
-7.95502245e-01 -1.91391990e-01 1.39083236e-01 7.30808854e-01
-3.13438848e-02 9.32272136e-01 -3.83140445e-02 -4.62460704e-03
-1.92290753e-01 1.79262102e-01 7.89318919e-01 -3.25970352e-01
-1.28752425e-01 2.03548670e-01 7.93335736e-02 -5.35170376e-01
-4.09284204e-01 -8.19353044e-01 -8.88061881e-01 -6.39755368e-01
-4.34500396e-01 1.19122416e-01 9.76671875e-01 1.19603324e+00
5.05333692e-02 2.45213822e-01 -4.04265597e-02 -1.22757566e+00
-1.15902507e+00 -1.03291738e+00 -8.57211053e-01 -3.83682698e-01
3.56623411e-01 -1.01165295e+00 -5.15553534e-01 -2.80024052e-01] | [7.271490097045898, 3.830002546310425] |
d9afdba1-043c-4bf0-a941-05687f63c1e3 | carla-bsp-a-simulated-dataset-with | 2305.00204 | null | https://arxiv.org/abs/2305.00204v1 | https://arxiv.org/pdf/2305.00204v1.pdf | CARLA-BSP: a simulated dataset with pedestrians | We present a sample dataset featuring pedestrians generated using the ARCANE framework, a new framework for generating datasets in CARLA (0.9.13). We provide use cases for pedestrian detection, autoencoding, pose estimation, and pose lifting. We also showcase baseline results. For more information, visit https://project-arcane.eu/. | ['Muhammad Naveed Riaz', 'Antonio M. López', 'Maciej Wielgosz'] | 2023-04-29 | null | null | null | null | ['pedestrian-detection'] | ['computer-vision'] | [-2.58278847e-01 -7.23435879e-02 4.25813347e-01 -4.49328840e-01
-6.43780768e-01 -7.63617754e-01 9.20434535e-01 -3.21756035e-01
-3.99056882e-01 8.32769930e-01 3.28912079e-01 -7.47329518e-02
7.03354657e-01 -8.36229384e-01 -1.11120808e+00 -4.89396125e-01
-2.43712559e-01 3.45520765e-01 1.99543819e-01 -2.36876398e-01
-2.13430256e-01 5.61478674e-01 -1.57954443e+00 5.74625313e-01
8.22104588e-02 6.86656535e-01 -1.09762564e-01 1.20838833e+00
5.43281138e-01 1.37018234e-01 -8.54780316e-01 -1.10990322e+00
4.69509035e-01 -1.43770337e-01 -6.29872084e-01 2.75187910e-01
1.02105021e+00 -5.00413299e-01 -6.96335077e-01 7.37272441e-01
9.05558586e-01 1.18546836e-01 6.88494921e-01 -1.57972383e+00
-2.42662698e-01 5.68015158e-01 -2.35089615e-01 1.29407614e-01
7.94355214e-01 9.50704396e-01 4.13129508e-01 -1.18153465e+00
6.73686147e-01 1.63797450e+00 1.01660848e+00 7.57473052e-01
-1.11978364e+00 -4.79598314e-01 -6.34592474e-02 2.80817151e-01
-1.50429356e+00 -7.81726956e-01 3.75500083e-01 -5.69266319e-01
8.27056408e-01 6.39769912e-01 1.00083673e+00 2.11068678e+00
-8.78809311e-04 9.52823281e-01 9.65175271e-01 -4.11643684e-02
-7.54550146e-03 6.04024436e-03 4.85612676e-02 6.40262842e-01
1.75169691e-01 6.36166036e-01 -2.81840473e-01 4.51463908e-02
6.01857841e-01 -5.87507188e-01 2.01498806e-01 -4.90146428e-01
-1.40885067e+00 5.55923343e-01 4.52368945e-01 -5.97507834e-01
-1.72920361e-01 3.06385547e-01 8.26722085e-01 -6.19538166e-02
-4.58801091e-02 2.40780443e-01 -1.86407700e-01 -7.40718320e-02
-5.46723723e-01 1.12877858e+00 6.51347458e-01 1.63908637e+00
2.34500542e-01 1.48193523e-01 -4.19907153e-01 4.03683811e-01
4.40034837e-01 7.87652075e-01 -6.03948459e-02 -1.24013722e+00
5.56426883e-01 -1.45392612e-01 3.13246936e-01 -7.31356561e-01
-2.67912418e-01 -2.96086311e-01 -5.40966988e-01 2.24858761e-01
5.91611862e-01 -5.61388195e-01 -8.29021871e-01 1.18171191e+00
3.74433964e-01 1.64210051e-02 1.05999857e-01 9.90752459e-01
1.24704814e+00 5.42642891e-01 3.80244702e-01 6.66670322e-01
1.47662795e+00 -1.03338468e+00 -2.73122847e-01 -2.45852977e-01
1.89515889e-01 -9.40320313e-01 6.18453801e-01 4.01915401e-01
-1.15651000e+00 -8.32333744e-01 -8.55110884e-01 -1.21726424e-01
-7.00012028e-01 6.16995871e-01 5.96412659e-01 1.03488541e+00
-1.02460682e+00 4.49229360e-01 -7.83789277e-01 -6.35078251e-01
5.95573246e-01 6.31033182e-02 -5.22507906e-01 1.01676382e-01
-9.70045507e-01 7.71249831e-01 5.59858441e-01 8.35438967e-02
-1.23985541e+00 -3.97158325e-01 -1.07295787e+00 -6.28554404e-01
-2.95998901e-02 -9.47136223e-01 1.63793385e+00 -4.90262806e-02
-1.13954568e+00 9.61056173e-01 5.04411347e-02 -9.94770110e-01
1.21376610e+00 -6.79877043e-01 -4.09036279e-01 -1.16791636e-01
3.55993330e-01 1.59310627e+00 6.54204726e-01 -1.12546742e+00
-7.56196856e-01 -8.71319622e-02 7.81915486e-02 -1.72080752e-02
4.24586296e-01 3.02785009e-01 -5.24410963e-01 -7.26866364e-01
-6.71903431e-01 -8.80974114e-01 -4.79484409e-01 3.03187728e-01
-1.01173878e+00 -8.33894499e-03 8.64381254e-01 -9.81770694e-01
5.86418748e-01 -1.99484015e+00 -1.29552767e-01 3.89322527e-02
1.12597629e-01 3.03651243e-01 -3.00509542e-01 3.91195416e-01
-3.10850501e-01 -1.86295539e-01 -2.82269776e-01 -7.67602324e-01
2.89616108e-01 1.53544815e-02 -1.10246897e-01 4.94053423e-01
4.10378188e-01 8.57335865e-01 -7.85496831e-01 -5.26397645e-01
8.30264270e-01 7.89190412e-01 -6.05248868e-01 1.01249374e-01
-6.40290827e-02 6.44976914e-01 1.08373992e-01 8.78686666e-01
8.77789795e-01 3.93488526e-01 -4.28128570e-01 -4.53984439e-01
-4.53118652e-01 2.98817065e-02 -1.44339192e+00 1.45989823e+00
-1.01370208e-01 8.96238863e-01 -9.21354741e-02 -3.32249850e-01
7.18210161e-01 6.29821643e-02 2.33132899e-01 -1.92742676e-01
2.35510647e-01 -1.25414953e-01 -2.37856835e-01 -4.07318860e-01
6.75432324e-01 7.98873007e-01 -2.49841392e-01 -2.62345701e-01
2.61446178e-01 -4.57676649e-02 9.17647243e-01 2.46655896e-01
8.67466927e-01 5.33192694e-01 2.56349981e-01 -1.79405436e-01
3.79971296e-01 3.21840793e-01 1.62553340e-01 6.97116733e-01
-4.67137516e-01 9.36466277e-01 2.40863383e-01 -7.89027452e-01
-1.57791972e+00 -1.59413528e+00 -5.19503593e-01 6.54973209e-01
-2.89214671e-01 -7.99253285e-01 -1.22682977e+00 -7.31531382e-01
1.18329234e-01 6.38526201e-01 -6.10621989e-01 1.60726562e-01
-7.40554035e-01 -6.21251166e-01 1.01114297e+00 7.82880127e-01
8.16289008e-01 -1.08991838e+00 -8.76963019e-01 -8.99406001e-02
-1.92536741e-01 -1.35647213e+00 -5.80633581e-01 -2.89149284e-01
-9.87524837e-02 -1.07208610e+00 -9.45860624e-01 -5.52031875e-01
4.60328639e-01 -3.58118005e-02 1.36557305e+00 -3.80562365e-01
-8.17990899e-01 5.43630838e-01 -2.07702532e-01 -5.05258799e-01
-3.77290696e-01 -1.09867090e-02 1.21209987e-01 -4.28383172e-01
1.21501721e-01 -1.11499615e-01 -7.75973618e-01 3.98631662e-01
-2.30634317e-01 2.76027977e-01 2.40774781e-01 3.32262039e-01
4.13032025e-01 -4.06831652e-01 3.13202515e-02 -5.50458491e-01
2.23492831e-01 -3.35661560e-01 -9.46947992e-01 -3.37405831e-01
2.35542849e-01 -1.55381024e-01 2.79324859e-01 1.88850388e-02
-9.58291173e-01 7.11156964e-01 -7.75680959e-01 -1.09933652e-01
-1.00009131e+00 -4.41913247e-01 -6.29688978e-01 9.29215699e-02
9.99294519e-01 4.56226021e-02 -2.55297661e-01 -3.97169769e-01
7.72325218e-01 5.65250933e-01 1.07866240e+00 -7.19527781e-01
6.75987899e-01 4.85611856e-01 -3.43072824e-02 -9.11662698e-01
-4.00086582e-01 -2.68165886e-01 -8.62990320e-01 -4.77623492e-01
8.72872412e-01 -1.19906056e+00 -7.35660374e-01 7.56067395e-01
-1.62715578e+00 -3.98622304e-01 -4.49180603e-01 3.01440001e-01
-7.89799333e-01 1.15680203e-01 -4.85985905e-01 -6.67836070e-01
-2.28449985e-01 -1.23462534e+00 1.37866068e+00 1.46339521e-01
-4.26692516e-01 -6.38496995e-01 6.09504012e-03 9.94954258e-02
1.24198370e-01 8.55819285e-01 -5.45506954e-01 -1.79772839e-01
-5.08353174e-01 -1.54508144e-01 -3.82343858e-01 1.58288807e-01
-3.39063019e-01 3.54044974e-01 -9.29596543e-01 -3.52119446e-01
-9.51144159e-01 -9.58731845e-02 8.72816026e-01 3.36462379e-01
1.06065166e+00 -5.05992651e-01 -4.82723773e-01 7.53514171e-01
8.11943173e-01 -1.82309926e-01 7.05569208e-01 4.03352231e-01
8.12559307e-01 7.82090783e-01 5.27670741e-01 8.26686442e-01
5.85843563e-01 9.51325774e-01 4.21608925e-01 1.30912960e-01
-6.74871445e-01 -3.68474603e-01 4.78348076e-01 -1.04999095e-01
-2.01851800e-01 -1.38968289e-01 -1.28165936e+00 6.72042251e-01
-1.60875177e+00 -1.21849597e+00 -7.25461662e-01 1.92769694e+00
6.63534880e-01 -5.89504316e-02 9.99820650e-01 2.32190229e-02
8.93277943e-01 -1.71509981e-01 -1.17404036e-01 -2.52956778e-01
-2.25627407e-01 1.76960319e-01 6.97235227e-01 4.50320214e-01
-1.55411100e+00 1.15353310e+00 6.84872961e+00 5.03862143e-01
-6.24096930e-01 5.98632880e-02 4.52356309e-01 -8.15103725e-02
4.43373948e-01 -2.32947484e-01 -1.36214745e+00 5.94682276e-01
1.00382447e+00 -3.99586186e-02 1.80079386e-01 1.12481844e+00
2.64791071e-01 7.51649216e-02 -1.12090981e+00 9.72446203e-01
-9.92115885e-02 -1.24548078e+00 1.14682401e-02 1.96539596e-01
4.72052127e-01 9.49544832e-02 1.18045866e-01 3.33490521e-01
5.86871147e-01 -6.74688995e-01 1.20132673e+00 4.44452465e-01
8.56708467e-01 -9.47787523e-01 5.56770742e-01 -8.93519297e-02
-1.31787610e+00 2.41416529e-01 -4.22192603e-01 3.40641320e-01
4.91475791e-01 2.97907859e-01 -9.87140238e-01 3.82927299e-01
1.00833166e+00 8.36544454e-01 -1.24787927e+00 1.60045063e+00
-2.85962880e-01 3.43014508e-01 -4.76040840e-01 1.87640205e-01
1.24482438e-01 1.15135208e-01 9.36719716e-01 2.06205106e+00
3.89035761e-01 -4.87313896e-01 2.13336900e-01 6.80234611e-01
3.07446361e-01 -1.84298977e-01 -9.24976230e-01 5.17571807e-01
4.67437357e-01 1.19649160e+00 -4.68198299e-01 -4.74629283e-01
-2.77672876e-02 1.23720908e+00 -9.74677801e-02 1.03969783e-01
-1.26119244e+00 -4.01136518e-01 7.39199996e-01 5.00614345e-01
2.35025972e-01 -4.27570701e-01 -1.05923861e-01 -8.62336993e-01
-7.53048882e-02 -7.89762735e-01 2.40719482e-01 -7.80726731e-01
-1.16987729e+00 5.77921808e-01 4.51622814e-01 -1.14896274e+00
-4.53000367e-01 -9.28351283e-01 -5.93711972e-01 6.08587503e-01
-6.82116151e-01 -1.56062579e+00 -4.75936532e-01 6.04218662e-01
7.15494394e-01 -2.23327041e-01 7.12232292e-01 4.19226319e-01
-5.84302008e-01 7.30895102e-01 -4.73845631e-01 4.86143231e-01
5.29768586e-01 -1.58759654e+00 1.60572338e+00 9.44883645e-01
-2.51299471e-01 3.21048796e-01 1.00416505e+00 -6.43605232e-01
-1.13188326e+00 -1.64230072e+00 3.96132052e-01 -9.06135082e-01
4.19867188e-01 -7.27095425e-01 -1.76804557e-01 8.84845674e-01
5.29963017e-01 1.94020167e-01 3.14934701e-01 -2.75772631e-01
-1.60604432e-01 1.54937014e-01 -1.30890846e+00 9.54353034e-01
1.49596238e+00 1.07731484e-01 -1.07139714e-01 5.00066757e-01
4.25665528e-01 -9.28823888e-01 -7.40431130e-01 2.53745466e-01
6.95859134e-01 -1.09442651e+00 1.38117659e+00 -4.09731120e-01
2.64530718e-01 -4.33592200e-01 -3.39602023e-01 -1.30835176e+00
-2.60582954e-01 -7.16979265e-01 -3.10572684e-01 1.27829766e+00
3.95841300e-01 -3.56155902e-01 7.74774909e-01 2.51898587e-01
-3.34026426e-01 -1.58220366e-01 -8.91521096e-01 -9.25256312e-01
1.41103894e-01 -5.67902148e-01 7.50671208e-01 3.41637790e-01
-4.78108108e-01 1.37534544e-01 -4.29705143e-01 2.90303975e-01
1.21524239e+00 -6.13398314e-01 1.43558967e+00 -6.08940303e-01
-2.54242122e-01 -2.88187832e-01 -7.18231499e-01 -8.35515201e-01
-8.22439641e-02 -5.16202092e-01 7.11767673e-02 -1.35133576e+00
-1.09835565e-01 -2.35619232e-01 5.51869035e-01 2.84179479e-01
5.97096048e-02 6.09117925e-01 3.73851359e-01 -2.45379657e-01
-6.15955889e-01 3.49108994e-01 8.80325496e-01 -1.52464896e-01
1.97668940e-01 3.23804736e-01 -2.39838764e-01 6.61048710e-01
1.81184804e+00 -2.24581063e-01 2.22072080e-01 -3.35798472e-01
-2.40311041e-01 -3.78896534e-01 1.21319842e+00 -1.87494457e+00
-1.64133638e-01 2.41647840e-01 9.34388220e-01 -9.74390626e-01
7.29927480e-01 -4.15954143e-01 3.12973738e-01 5.58402717e-01
-9.50758979e-02 5.01068294e-01 4.35133696e-01 1.05836719e-01
3.05147886e-01 5.62217869e-02 8.70373189e-01 -2.46662989e-01
-1.01290691e+00 3.12864959e-01 -3.86496842e-01 -1.34918854e-01
1.08361256e+00 -9.25076604e-02 -4.88815248e-01 -1.25228733e-01
-9.44170594e-01 1.39808834e-01 4.30929512e-01 6.94899201e-01
6.95818543e-01 -1.72869575e+00 -1.25175798e+00 2.71411955e-01
2.72797346e-01 -1.71529099e-01 8.58825371e-02 4.13580596e-01
-9.25456166e-01 5.50058305e-01 -6.48950994e-01 -7.41350353e-01
-1.37863696e+00 4.71551776e-01 3.67877752e-01 3.14915657e-01
-7.10180998e-01 7.20735073e-01 3.12565789e-02 -6.09113097e-01
3.33628729e-02 8.49406123e-02 4.13960777e-03 -4.12086844e-01
8.09517682e-01 9.20117259e-01 -4.30557951e-02 -1.07802451e+00
-6.05295181e-01 1.78046584e-01 1.16684429e-01 -5.16786575e-01
7.86482990e-01 -1.06616067e-02 4.61856395e-01 -1.78947434e-01
8.41941237e-01 -1.32723629e-01 -1.57799745e+00 3.59745234e-01
-3.29494029e-01 -5.83553374e-01 -3.60318929e-01 -7.67158329e-01
-8.36120903e-01 9.16845322e-01 9.25646722e-01 -2.32521996e-01
5.06136239e-01 1.48820326e-01 6.68102503e-01 3.04048955e-01
4.76210862e-01 -8.57270896e-01 -2.23754585e-01 5.26619375e-01
1.24911594e+00 -1.11809611e+00 -2.54615277e-01 -5.62853515e-01
-5.70372701e-01 8.64714205e-01 9.30880249e-01 -2.05431029e-01
3.46977025e-01 5.66153884e-01 -3.57933119e-02 7.55770653e-02
-3.72357428e-01 -4.93851244e-01 1.19342901e-01 1.44286895e+00
6.34303629e-01 4.90330875e-01 3.79006863e-02 3.73512894e-01
-9.31494057e-01 -3.01392645e-01 5.02947092e-01 8.37607861e-01
-1.13412991e-01 -1.13831699e+00 -9.07421231e-01 -3.21892127e-02
-8.84233937e-02 -1.04763724e-01 -6.48207545e-01 7.91827798e-01
5.07999063e-01 8.43140423e-01 1.20725282e-01 -3.33875775e-01
7.05578208e-01 -2.10327446e-01 8.12289715e-01 -4.79908198e-01
-4.41717625e-01 -3.02834034e-01 5.64732909e-01 -7.27387190e-01
4.56687137e-02 -1.18131506e+00 -7.76322722e-01 -8.05456758e-01
3.11120719e-01 -1.18995108e-01 8.34035397e-01 2.03595698e-01
4.55680102e-01 3.43694597e-01 1.75927088e-01 -1.22542882e+00
-9.53253955e-02 -1.03126466e+00 4.67305072e-02 2.19333902e-01
2.09085464e-01 -6.15871549e-01 1.39745757e-01 4.76568431e-01] | [7.696247100830078, -0.8630622029304504] |
1f467a41-b227-4cd0-a95c-7676b028549d | pseudo-pair-based-self-similarity-learning | 2207.13035 | null | https://arxiv.org/abs/2207.13035v1 | https://arxiv.org/pdf/2207.13035v1.pdf | Pseudo-Pair based Self-Similarity Learning for Unsupervised Person Re-identification | Person re-identification (re-ID) is of great importance to video surveillance systems by estimating the similarity between a pair of cross-camera person shorts. Current methods for estimating such similarity require a large number of labeled samples for supervised training. In this paper, we present a pseudo-pair based self-similarity learning approach for unsupervised person re-ID without human annotations. Unlike conventional unsupervised re-ID methods that use pseudo labels based on global clustering, we construct patch surrogate classes as initial supervision, and propose to assign pseudo labels to images through the pairwise gradient-guided similarity separation. This can cluster images in pseudo pairs, and the pseudos can be updated during training. Based on pseudo pairs, we propose to improve the generalization of similarity function via a novel self-similarity learning:it learns local discriminative features from individual images via intra-similarity, and discovers the patch correspondence across images via inter-similarity. The intra-similarity learning is based on channel attention to detect diverse local features from an image. The inter-similarity learning employs a deformable convolution with a non-local block to align patches for cross-image similarity. Experimental results on several re-ID benchmark datasets demonstrate the superiority of the proposed method over the state-of-the-arts. | ['Jialie Shen', 'Mohammed Bennamoun', 'Farid Boussaid', 'ZongYuan Ge', 'Dapeng Chen', 'Wenying Zhang', 'Deyin Liu', 'Lin Wu'] | 2022-07-09 | null | null | null | null | ['unsupervised-person-re-identification'] | ['computer-vision'] | [ 3.14666808e-01 -3.88958871e-01 -9.59428251e-02 -7.28894889e-01
-5.92554390e-01 -3.84705484e-01 7.98525751e-01 1.22036979e-01
-6.03294611e-01 3.23422432e-01 3.24080288e-01 5.40569484e-01
-1.69368804e-01 -5.10435045e-01 -7.91160166e-01 -6.20902717e-01
1.40228301e-01 4.23203170e-01 2.55862474e-01 1.21476747e-01
2.48569831e-01 9.79885906e-02 -1.60014558e+00 3.52717310e-01
8.64395201e-01 8.39046240e-01 1.47541061e-01 4.36648369e-01
2.43418694e-01 4.25484210e-01 -4.63299751e-01 -5.26666403e-01
5.93396008e-01 -5.20316422e-01 -7.70398498e-01 5.41012764e-01
1.04049909e+00 -2.96238750e-01 -3.72103930e-01 1.38051212e+00
4.58806574e-01 4.15461630e-01 6.21438563e-01 -1.29272890e+00
-9.41069067e-01 5.27162813e-02 -6.34044349e-01 4.35622782e-01
6.55169666e-01 2.21484294e-03 7.14705825e-01 -7.82443285e-01
3.04651409e-01 1.44701278e+00 1.07224214e+00 7.25840390e-01
-1.26483750e+00 -8.46245646e-01 3.15153778e-01 5.11469662e-01
-1.79038775e+00 -3.95870745e-01 7.76880860e-01 -5.54191828e-01
5.92717707e-01 1.83972254e-01 6.46764278e-01 9.02555525e-01
-2.45076239e-01 6.25235736e-01 9.84257698e-01 -3.26941460e-01
-7.57282749e-02 2.78513581e-01 2.27877572e-01 7.00873673e-01
2.12218404e-01 2.38814414e-01 -3.00471693e-01 -2.49697506e-01
7.70479977e-01 6.99271977e-01 -4.31048930e-01 -4.98812646e-01
-1.26920116e+00 5.58267772e-01 5.04923940e-01 3.53005737e-01
3.83214243e-02 -1.19837217e-01 3.92473489e-01 1.59890130e-01
4.01963949e-01 2.03874901e-01 -6.79813921e-02 2.42811233e-01
-9.24579799e-01 1.34650767e-01 2.97162533e-01 1.00114346e+00
9.23803329e-01 -7.24058926e-01 -4.11081642e-01 1.12314487e+00
1.96566060e-01 6.16918564e-01 7.90904224e-01 -7.05141902e-01
3.76252413e-01 6.80337369e-01 2.72073627e-01 -1.32420087e+00
-1.72993034e-01 -1.41344011e-01 -1.16892159e+00 -2.23365754e-01
4.67986763e-01 1.47997513e-01 -7.49809444e-01 1.74655676e+00
3.65034282e-01 8.85499120e-01 -7.20245615e-02 9.97093260e-01
8.33892584e-01 4.58467722e-01 2.85074729e-02 -1.26853138e-01
1.22874439e+00 -1.23297846e+00 -3.97969872e-01 -9.83512178e-02
3.72367203e-01 -5.00732839e-01 6.44652009e-01 -5.87060153e-02
-7.28047192e-01 -1.28981996e+00 -8.42243910e-01 3.06089163e-01
-3.88035178e-01 1.67656153e-01 -2.88940333e-02 7.12409377e-01
-1.09146750e+00 5.88678241e-01 -4.65148449e-01 -8.16985071e-01
3.90197337e-01 4.31647569e-01 -6.96692288e-01 -2.09071562e-01
-1.02503312e+00 5.54282010e-01 4.75456268e-01 9.23815742e-02
-7.39069343e-01 -6.32394373e-01 -1.01836812e+00 -3.48295756e-02
1.07843820e-02 -7.29134500e-01 7.41780698e-01 -1.37631214e+00
-1.22854221e+00 1.24753845e+00 -3.64049643e-01 -3.63909185e-01
2.62609929e-01 -3.37007418e-02 -7.43087232e-01 3.35089654e-01
5.99123001e-01 7.14633286e-01 1.15389311e+00 -1.40905237e+00
-1.01640213e+00 -4.47510540e-01 -1.41474888e-01 3.83113295e-01
-5.38557351e-01 8.32224190e-02 -7.93120563e-01 -7.92416573e-01
-4.97226864e-02 -1.06267965e+00 -1.33072779e-01 -1.27215207e-01
-2.68219173e-01 -4.65019256e-01 6.26323819e-01 -6.93733990e-01
9.47654366e-01 -2.08644152e+00 -1.19411293e-02 3.96684349e-01
2.88970292e-01 4.15418565e-01 -2.19385639e-01 2.84152161e-02
-3.32375914e-01 -2.92137057e-01 -1.49190083e-01 -5.79151571e-01
-2.11171493e-01 -1.86102271e-01 3.90516147e-02 7.94171154e-01
7.01115951e-02 8.04563642e-01 -1.05560231e+00 -6.35261714e-01
3.68922085e-01 3.90920877e-01 -3.04002136e-01 6.41384602e-01
5.48201263e-01 5.85047305e-01 -2.07888380e-01 4.78502035e-01
8.00403535e-01 -3.48682791e-01 -5.87574206e-02 -5.35945296e-01
2.55886409e-02 -3.70199323e-01 -1.20821500e+00 1.73555267e+00
-8.20076168e-02 3.71744752e-01 -2.66528666e-01 -1.52875686e+00
9.80026245e-01 1.24769077e-01 6.14637077e-01 -5.61519861e-01
-9.23789367e-02 -1.49717078e-01 -4.35839415e-01 -5.12425900e-01
2.22420320e-01 2.99709678e-01 1.20214019e-02 4.64309752e-01
1.61605820e-01 5.53613901e-01 2.08158642e-01 8.83065313e-02
5.67370832e-01 -1.47827625e-01 1.71636909e-01 -3.67916584e-01
1.20438719e+00 -3.16777647e-01 6.63383245e-01 8.61084521e-01
-4.78088439e-01 7.40446806e-01 -3.71135861e-01 -8.02719414e-01
-1.03041124e+00 -1.12021363e+00 -2.14155465e-01 9.09234047e-01
8.44725370e-01 -3.73746991e-01 -9.58360970e-01 -1.01117468e+00
-1.37144504e-02 -8.55645984e-02 -8.36775482e-01 -2.27173194e-01
-4.95828420e-01 -7.38430560e-01 3.49217385e-01 5.61151624e-01
1.07742262e+00 -6.98323548e-01 -9.07161459e-03 -6.87526390e-02
-4.29140270e-01 -1.19196630e+00 -1.31609702e+00 -4.51126516e-01
-4.84885544e-01 -1.24740875e+00 -1.20709205e+00 -1.40068018e+00
1.16345549e+00 9.10443127e-01 8.24025154e-01 7.64751360e-02
-3.31558496e-01 1.14821613e+00 -3.22086096e-01 2.33524308e-01
3.31836566e-02 -2.53126413e-01 5.16589820e-01 7.73125350e-01
8.58611584e-01 -3.76321912e-01 -8.22319388e-01 8.00370514e-01
-4.52861845e-01 -1.61784530e-01 4.13551062e-01 1.12009799e+00
7.13183463e-01 9.50626433e-02 4.80117083e-01 -5.63415170e-01
3.74322593e-01 -3.39897603e-01 -3.16017926e-01 6.12924278e-01
-4.74437207e-01 -3.22029777e-02 5.57382286e-01 -7.31053948e-01
-1.08873665e+00 1.97422445e-01 4.33009237e-01 -5.06732821e-01
-5.31896830e-01 -1.64562643e-01 -2.61013955e-01 -3.37111056e-01
4.16810393e-01 5.60529351e-01 2.98685450e-02 -3.15720618e-01
2.31347680e-01 9.30909634e-01 8.74942422e-01 -4.90706861e-01
9.12581146e-01 7.76993752e-01 -4.88825828e-01 -7.76377618e-01
-8.49525332e-01 -1.15820408e+00 -9.60139632e-01 -3.14866662e-01
1.12434828e+00 -1.22698021e+00 -7.58166671e-01 6.22997224e-01
-9.68453705e-01 -3.73705477e-02 -2.07302839e-01 5.67155898e-01
-3.24051887e-01 9.34449792e-01 -2.93078005e-01 -4.64250982e-01
-3.80304426e-01 -9.16506290e-01 1.21036255e+00 6.66632593e-01
-9.72563624e-02 -1.15022683e+00 2.68751830e-01 5.23442149e-01
3.45729254e-02 -1.65354967e-01 9.90320835e-03 -6.93769991e-01
-2.29681075e-01 -4.17008400e-01 -5.92566490e-01 3.80199671e-01
6.80187225e-01 -7.24366605e-01 -8.49229097e-01 -7.87612140e-01
-2.24002391e-01 -1.33756921e-01 7.79786289e-01 2.91665733e-01
1.28071749e+00 -3.09811324e-01 -7.53035128e-01 9.07241940e-01
1.15242136e+00 3.97958010e-02 3.35464418e-01 3.67460966e-01
1.06510830e+00 6.60394907e-01 5.90160191e-01 4.38871354e-01
6.64856374e-01 8.96487772e-01 -1.16389655e-01 -2.42259890e-01
-1.90518424e-01 -3.81348312e-01 3.79372030e-01 5.04803240e-01
-2.22297117e-01 2.51909971e-01 -5.99087417e-01 6.72276795e-01
-2.09679914e+00 -1.35696411e+00 1.76658034e-01 2.59702349e+00
6.70071065e-01 -2.94491857e-01 4.87536013e-01 -1.49606302e-01
1.60030115e+00 -1.18881352e-01 -4.60045725e-01 2.85511047e-01
-4.92984615e-02 -3.02065074e-01 5.90543032e-01 4.41079766e-01
-1.59256208e+00 7.61142492e-01 5.39855909e+00 8.88269186e-01
-7.09705293e-01 1.47571087e-01 6.26823366e-01 3.34727615e-01
3.12751591e-01 -2.38748044e-01 -1.17146385e+00 9.52734292e-01
4.97911572e-01 -1.37112557e-03 4.38921362e-01 7.14635491e-01
-4.33681421e-02 1.03551541e-02 -1.28439498e+00 1.76709032e+00
8.07762444e-01 -1.05053806e+00 1.29657507e-01 -2.12578937e-01
1.03822505e+00 -3.96769524e-01 5.45384921e-02 -8.17839354e-02
2.90420026e-01 -8.14791501e-01 4.14266020e-01 7.91025639e-01
7.82740593e-01 -7.20542490e-01 7.90279269e-01 3.68561074e-02
-1.71780229e+00 -2.03907862e-01 -6.63485229e-01 1.64382458e-01
-8.78795758e-02 1.52965486e-01 -4.58122164e-01 4.25499052e-01
1.33527780e+00 1.48317134e+00 -9.44510281e-01 1.13657546e+00
1.33626491e-01 1.70224681e-01 -9.17042792e-02 2.95802683e-01
-1.12843186e-01 -3.93905431e-01 3.81738722e-01 1.45982730e+00
1.44141495e-01 5.61804101e-02 6.22655511e-01 6.19004309e-01
1.95254222e-01 -9.07045528e-02 -4.02092904e-01 5.73313117e-01
3.70081246e-01 1.04019141e+00 -4.28842932e-01 -6.28019333e-01
-5.91310501e-01 1.77239895e+00 3.37945431e-01 4.14875418e-01
-7.32938051e-01 -4.59216386e-01 6.51589811e-01 4.58483249e-02
4.20806140e-01 1.64995510e-02 3.12365234e-01 -1.33930469e+00
1.21859424e-01 -6.79047048e-01 8.01845431e-01 -4.42302674e-01
-2.03598118e+00 4.71673757e-01 1.97070196e-01 -1.67606139e+00
-2.48041172e-02 -4.23531026e-01 -6.66258872e-01 6.87004447e-01
-1.43322778e+00 -1.53721952e+00 -7.90545046e-01 1.23366892e+00
4.46014673e-01 -5.87738216e-01 6.96221054e-01 3.91885817e-01
-6.37646616e-01 1.14417899e+00 3.07503104e-01 6.57376647e-01
1.18865442e+00 -1.13886881e+00 3.68862689e-01 8.93449485e-01
8.13398436e-02 7.40852416e-01 1.78403467e-01 -7.84109294e-01
-8.45609546e-01 -1.43476510e+00 7.82960415e-01 -4.42620486e-01
2.17977971e-01 -2.43136510e-01 -7.04002142e-01 6.51739717e-01
6.16191030e-02 4.02574360e-01 9.53495741e-01 -1.14778690e-01
-6.13179266e-01 -5.25668502e-01 -1.17692316e+00 3.16471905e-01
1.36652207e+00 -6.99536920e-01 -6.24278426e-01 4.88141656e-01
3.26446354e-01 1.11376837e-01 -9.02181923e-01 2.18875110e-01
5.10449469e-01 -8.68884504e-01 1.42782795e+00 -4.26793575e-01
-1.89144164e-01 -7.00790405e-01 5.26736230e-02 -1.22031045e+00
-8.11112285e-01 -4.29859906e-01 3.37959796e-01 1.58471155e+00
-2.96898812e-01 -6.76751852e-01 7.44198322e-01 6.12714827e-01
1.26226902e-01 -7.39563853e-02 -7.47376323e-01 -1.03030741e+00
-3.58017921e-01 1.82207495e-01 5.32375932e-01 1.13083184e+00
4.76052538e-02 9.49418396e-02 -6.44388795e-01 5.98313391e-01
1.19280767e+00 1.67413995e-01 9.06940222e-01 -1.23705041e+00
-3.60778332e-01 -2.54494458e-01 -8.73558581e-01 -1.27178347e+00
4.05853122e-01 -9.70881462e-01 1.06991231e-01 -1.10786855e+00
7.96724439e-01 -5.40290475e-01 -4.00023282e-01 2.38699883e-01
-6.12173557e-01 5.61509371e-01 6.77015483e-02 6.41990125e-01
-1.26038992e+00 6.03380799e-01 7.80320466e-01 -6.88575685e-01
-1.79045051e-01 -4.24663676e-03 -4.94068831e-01 6.71961784e-01
6.40240967e-01 -3.17877561e-01 -2.38145635e-01 -2.96459824e-01
-5.18400967e-01 -5.56792021e-01 7.24077880e-01 -1.41197371e+00
5.96377611e-01 3.70465219e-02 7.84016430e-01 -3.48985732e-01
1.69641674e-01 -9.77494001e-01 -3.67682017e-02 3.14027429e-01
-4.54959244e-01 3.68378833e-02 -1.65727139e-01 8.69114399e-01
-3.73147190e-01 -1.39236480e-01 9.33813930e-01 -2.10396454e-01
-1.00023532e+00 6.02919042e-01 5.50247580e-02 8.99645016e-02
1.17154002e+00 -5.40712237e-01 9.29866508e-02 -2.56169587e-01
-5.98646939e-01 3.18576247e-01 6.50797129e-01 4.86110598e-01
7.71990597e-01 -1.67555273e+00 -6.80938125e-01 4.23845083e-01
5.53828001e-01 -2.52402097e-01 5.22872031e-01 5.19034326e-01
3.31016853e-02 2.56088793e-01 -4.52895433e-01 -9.48972106e-01
-1.61376631e+00 8.32545996e-01 5.23140907e-01 -5.99471256e-02
-5.26523590e-01 9.12816882e-01 6.31945491e-01 -4.24926400e-01
2.94677496e-01 2.37336278e-01 -5.29238462e-01 -2.31470585e-01
1.07407343e+00 3.86211663e-01 -3.85629416e-01 -1.13656890e+00
-3.85474533e-01 1.25027049e+00 -3.54239851e-01 1.44957364e-01
9.70371008e-01 -5.08310497e-01 -1.39469966e-01 1.59845501e-01
1.44354379e+00 -3.57070625e-01 -1.51305056e+00 -6.91363931e-01
-2.29674369e-01 -8.08131635e-01 -3.57254177e-01 -2.35797584e-01
-1.01177955e+00 5.60660779e-01 1.19378507e+00 -1.45128310e-01
1.11827230e+00 1.04026511e-01 9.43401158e-01 3.75968784e-01
2.49722958e-01 -1.19640779e+00 4.75313902e-01 5.52574918e-02
5.99493265e-01 -1.67010963e+00 -5.09667508e-02 -4.14510220e-01
-6.07293487e-01 8.86695027e-01 7.61885762e-01 -3.25652897e-01
7.65903175e-01 -3.91088039e-01 -1.66258842e-01 1.83618873e-01
1.63799033e-01 -3.33887011e-01 5.46036065e-01 1.07794714e+00
-4.56710719e-02 -5.39304800e-02 7.63454139e-02 6.21201217e-01
2.38381520e-01 -1.90081343e-01 -9.17871520e-02 6.00746453e-01
-2.49208152e-01 -1.14281356e+00 -7.31871128e-01 3.08442652e-01
-2.43003182e-02 2.39913054e-02 -2.79348224e-01 2.78923571e-01
6.29545093e-01 1.08281231e+00 4.47441608e-01 -6.98765159e-01
1.52416602e-01 -3.79809678e-01 4.95489061e-01 -5.04446149e-01
-5.22151470e-01 -2.85796523e-01 -4.29059088e-01 -5.00108182e-01
-8.99701655e-01 -8.59422088e-01 -7.14318991e-01 -1.28096074e-01
-2.05092847e-01 2.70104796e-01 5.81893548e-02 8.63121629e-01
4.50719476e-01 -4.97623160e-02 9.29309845e-01 -1.05162203e+00
-3.11129391e-01 -8.55805278e-01 -4.09345388e-01 1.21874237e+00
3.13743919e-01 -7.49408126e-01 -3.29459012e-01 6.33047283e-01] | [14.773555755615234, 0.9880939722061157] |
413fa7cd-b859-4c57-8109-61cd10792b9a | a-survey-on-sentiment-analysis-in-persian-a | 2104.14751 | null | https://arxiv.org/abs/2104.14751v1 | https://arxiv.org/pdf/2104.14751v1.pdf | A Survey on sentiment analysis in Persian: A Comprehensive System Perspective Covering Challenges and Advances in Resources, and Methods | Social media has been remarkably grown during the past few years. Nowadays, posting messages on social media websites has become one of the most popular Internet activities. The vast amount of user-generated content has made social media the most extensive data source of public opinion. Sentiment analysis is one of the techniques used to analyze user-generated data. The Persian language has specific features and thereby requires unique methods and models to be adopted for sentiment analysis, which are different from those in English language. Sentiment analysis in each language has specified prerequisites; hence, the direct use of methods, tools, and resources developed for English language in Persian has its limitations. The main target of this paper is to provide a comprehensive literature survey for state-of-the-art advances in Persian sentiment analysis. In this regard, the present study aims to investigate and compare the previous sentiment analysis studies on Persian texts and describe contributions presented in articles published in the last decade. First, the levels, approaches, and tasks for sentiment analysis are described. Then, a detailed survey of the sentiment analysis methods used for Persian texts is presented, and previous relevant works on Persian Language are discussed. Moreover, we present in this survey the authentic and published standard sentiment analysis resources and advances that have been done for Persian sentiment analysis. Finally, according to the state-of-the-art development of English sentiment analysis, some issues and challenges not being addressed in Persian texts are listed, and some guidelines and trends are provided for future research on Persian texts. The paper provides information to help new or established researchers in the field as well as industry developers who aim to deploy an operational complete sentiment analysis system. | ['MohammadReza Valavi', 'Zeinab Rajabi'] | 2021-04-30 | null | null | null | null | ['persian-sentiment-anlysis'] | ['natural-language-processing'] | [-2.00353786e-01 -6.46272674e-02 -4.42169815e-01 -3.86252493e-01
-2.44855002e-01 -6.71700180e-01 4.60741013e-01 5.80610633e-01
-7.86492825e-01 6.93885565e-01 9.14003551e-02 -1.40852019e-01
2.38236934e-02 -6.77441359e-01 2.98946142e-01 -4.72425222e-01
2.23375291e-01 4.66377437e-01 -2.13008612e-01 -9.79929030e-01
8.27263534e-01 3.09976697e-01 -1.57806468e+00 1.21497028e-01
8.10948193e-01 5.62025130e-01 3.55887741e-01 4.02988166e-01
-5.72545409e-01 8.57021749e-01 -1.04351866e+00 -9.90935802e-01
-2.07652152e-01 -2.08088815e-01 -1.01942801e+00 1.34457842e-01
-1.67293668e-01 3.46140504e-01 2.09421232e-01 1.03148735e+00
3.80113989e-01 1.53204292e-01 6.34996772e-01 -1.05043697e+00
-6.93166316e-01 8.47124517e-01 -4.75664735e-01 1.44935280e-01
5.49569070e-01 -6.90688252e-01 1.17456841e+00 -8.79497051e-01
6.39582992e-01 1.09839261e+00 4.46815044e-01 3.88227165e-01
-3.69002134e-01 -5.34036279e-01 1.22235648e-01 2.86594987e-01
-9.73220170e-01 9.01235733e-03 9.34950888e-01 -4.32130605e-01
1.08369553e+00 1.75252825e-01 9.14951086e-01 7.34689355e-01
7.87393034e-01 8.41269255e-01 1.29627645e+00 -6.33698106e-01
2.18872964e-01 6.40846550e-01 4.75380450e-01 1.64834142e-01
4.07604545e-01 -8.52791846e-01 -5.38484395e-01 -9.70560536e-02
-1.26842573e-01 4.87625189e-02 2.41137788e-01 1.91517130e-01
-7.17676759e-01 1.03450799e+00 -2.08590031e-01 9.79495764e-01
-3.93266648e-01 -7.17647314e-01 8.59195650e-01 3.22757572e-01
8.31683159e-01 5.05275369e-01 -7.86422551e-01 -5.53979933e-01
-9.76543188e-01 2.62283653e-01 9.92340267e-01 6.94531858e-01
5.27890444e-01 6.38035014e-02 4.44241911e-01 1.15287566e+00
5.56366444e-01 9.82453465e-01 9.30540383e-01 -5.58925688e-01
3.95583153e-01 9.90359783e-01 -1.75526366e-02 -1.52248609e+00
-4.25115436e-01 -6.25779033e-02 -4.73654300e-01 -1.81061164e-01
-7.17597306e-02 -3.19956392e-01 -3.53972167e-01 9.27472115e-01
2.03752995e-01 -1.00550747e+00 6.25557482e-01 4.34269994e-01
1.20427895e+00 7.60493219e-01 1.72449946e-01 -4.38702434e-01
1.87328851e+00 -7.98724651e-01 -1.02477312e+00 -5.84726334e-01
6.55245900e-01 -1.41251171e+00 1.08669543e+00 6.84000254e-01
-1.00965953e+00 -2.32879847e-01 -9.31668282e-01 -1.52288005e-02
-9.13869321e-01 1.18211187e-01 6.02158487e-01 1.16651785e+00
-6.63577259e-01 3.21700126e-01 -6.94502473e-01 -9.06169295e-01
1.17846318e-01 3.55235994e-01 -2.30742559e-01 4.33182925e-01
-1.26841164e+00 1.08984506e+00 1.87925920e-01 2.23885421e-02
2.62066960e-01 -8.21376666e-02 -8.84034038e-01 -4.29589897e-01
1.41033918e-01 7.69480690e-02 1.30283821e+00 -1.31145203e+00
-1.59023273e+00 1.23547602e+00 -3.64067435e-01 -1.50258094e-01
-5.63044392e-04 -3.02154869e-01 -8.25030327e-01 1.56915873e-01
5.41650474e-01 5.10502532e-02 3.93041670e-01 -8.79251301e-01
-1.09099102e+00 -5.92841029e-01 7.10732043e-02 2.73807436e-01
-8.10199082e-01 8.84540796e-01 -3.74796987e-01 -6.89812958e-01
-3.80648822e-02 -9.22070861e-01 -1.78893134e-01 -1.10600162e+00
-4.37599346e-02 -4.57598329e-01 9.34034646e-01 -5.55255115e-01
1.53631914e+00 -2.14650655e+00 -1.66045442e-01 2.09412068e-01
-7.85031915e-02 2.71540374e-01 2.52514690e-01 1.12275338e+00
-1.52240470e-02 3.26117873e-01 -3.53721529e-02 -3.28379631e-01
-2.03082338e-02 2.31411248e-01 -4.27599639e-01 4.20922399e-01
-2.45389462e-01 4.40963745e-01 -9.31294858e-01 -4.43502456e-01
2.99284488e-01 5.71560502e-01 -9.64074060e-02 -2.93508202e-01
3.57020110e-01 7.73158595e-02 -6.55499101e-01 6.73460186e-01
5.08994102e-01 9.61715635e-03 4.82910305e-01 -1.38077680e-02
-6.04945481e-01 2.94129372e-01 -9.32647347e-01 1.05500364e+00
-5.63452482e-01 6.03056848e-01 -1.39916033e-01 -9.47763443e-01
1.11428142e+00 3.25631440e-01 5.87854028e-01 -5.09512842e-01
6.46658659e-01 5.17738044e-01 -1.51201472e-01 -3.51195961e-01
1.26455653e+00 -1.55037239e-01 -5.51754296e-01 4.28853869e-01
-4.58040647e-02 -5.29882550e-01 8.78173113e-01 2.85447121e-01
3.30211371e-01 -3.41326952e-01 7.61161983e-01 -4.01573807e-01
1.14131796e+00 2.56533623e-01 3.41434658e-01 6.19553439e-02
-2.60141671e-01 1.71358615e-01 5.40863574e-01 -3.76690537e-01
-6.07522488e-01 -1.21176660e-01 -2.30635479e-01 1.07270777e+00
-8.04585889e-02 -1.18059206e+00 -7.52648056e-01 -6.43834114e-01
-3.38098109e-01 5.64628482e-01 -5.32834291e-01 5.27416706e-01
-3.20571840e-01 -9.88511384e-01 2.61400610e-01 1.96975678e-01
5.67737639e-01 -1.38815773e+00 -6.85283542e-01 1.82647452e-01
-3.30196649e-01 -1.15340340e+00 2.46538460e-01 7.20532238e-02
-8.41866434e-01 -1.18905044e+00 -5.75353563e-01 -9.73477721e-01
4.77490693e-01 4.20802534e-01 9.77136433e-01 -1.77959353e-02
3.35090011e-01 4.42508191e-01 -9.80479419e-01 -9.86714602e-01
-2.49361113e-01 3.80999982e-01 9.40269753e-02 -4.91777629e-01
1.15830207e+00 -2.71056872e-02 -1.95692509e-01 2.00145811e-01
-8.61407995e-01 -4.59365040e-01 1.63615093e-01 5.01423359e-01
2.42899284e-01 6.12092972e-01 6.32669091e-01 -1.27922189e+00
9.36695039e-01 -4.05052781e-01 -3.68971020e-01 -2.70579696e-01
-8.98757935e-01 -5.23405790e-01 7.88057566e-01 2.75510728e-01
-1.07399225e+00 -4.68786091e-01 -3.59241217e-01 6.59102142e-01
-5.63540049e-02 1.05626404e+00 1.10529393e-01 1.47732288e-01
5.44640779e-01 -1.40707716e-01 -9.88587155e-04 -1.61997631e-01
1.16764670e-02 1.22392333e+00 -9.68367830e-02 -3.12224299e-01
3.34586471e-01 5.99865615e-01 -2.82253772e-01 -1.35989940e+00
-8.77789915e-01 -8.03147793e-01 -4.29820508e-01 -4.34663892e-01
7.18178272e-01 -8.22358429e-01 -6.11763656e-01 7.74684489e-01
-7.68112719e-01 7.56806359e-02 -2.60127306e-01 3.09890151e-01
-3.11239183e-01 5.89683414e-01 -7.77326345e-01 -1.09408951e+00
-7.01646447e-01 -1.13215232e+00 6.72262847e-01 4.00616109e-01
-9.50487375e-01 -1.35903931e+00 1.37595415e-01 8.12703550e-01
1.38811767e-01 -2.35232234e-01 5.03164232e-01 -8.33886743e-01
6.76117599e-01 -4.62500811e-01 3.12728733e-01 6.13794863e-01
4.03116822e-01 4.85132188e-01 -6.96499646e-01 -1.69874460e-01
4.03276592e-01 -3.00966561e-01 2.69360304e-01 4.37096268e-01
4.89846706e-01 2.35598311e-02 -1.46289423e-01 -2.46688485e-01
1.24912107e+00 5.67060590e-01 6.92179918e-01 1.08193970e+00
2.37650976e-01 1.13947725e+00 1.40316164e+00 6.69182479e-01
6.98044717e-01 6.30715862e-02 -8.78340229e-02 2.44525105e-01
6.05620861e-01 7.31719136e-02 7.08398521e-01 1.39291918e+00
-4.16663140e-01 -3.04587066e-01 -9.46712971e-01 4.83651191e-01
-1.68254769e+00 -8.28279734e-01 -4.34646428e-01 1.77844584e+00
4.67718869e-01 1.12299360e-01 2.29868680e-01 7.73701370e-01
5.08517087e-01 3.00563961e-01 2.00595230e-01 -8.46239090e-01
-4.06053960e-01 4.46284562e-01 3.34984452e-01 3.76306564e-01
-1.16380799e+00 1.31506646e+00 6.55461979e+00 6.63791060e-01
-1.26334524e+00 -5.05943187e-02 2.63213485e-01 2.40429044e-01
-1.72239929e-01 -1.89182386e-02 -9.34915543e-01 3.15761805e-01
1.02194810e+00 -2.96489894e-01 -1.19086683e-01 1.04254925e+00
5.03813028e-01 -3.65641594e-01 -1.57390490e-01 1.04578340e+00
5.25648057e-01 -8.02890301e-01 -1.40166298e-01 8.36562663e-02
7.63588428e-01 1.80383503e-01 1.84009552e-01 1.96512938e-01
6.21863194e-02 -5.61608970e-01 7.49257624e-01 -2.51382500e-01
2.38445505e-01 -1.09835958e+00 1.40073860e+00 5.79118319e-02
-1.11246395e+00 1.63742267e-02 -4.41740334e-01 -5.82410634e-01
2.51462609e-01 7.11918533e-01 -3.94238949e-01 7.01198399e-01
1.09955454e+00 1.29247141e+00 -5.08466601e-01 2.82543629e-01
-5.11341870e-01 6.80469573e-01 5.47652654e-02 -5.85423708e-01
4.31140900e-01 -7.33270645e-01 4.45130646e-01 1.40634096e+00
2.58065969e-01 -9.66228694e-02 -1.91418864e-02 -2.87975259e-02
5.38121402e-01 1.20626652e+00 -5.65949500e-01 -5.92243791e-01
2.43835717e-01 1.50485718e+00 -1.27004039e+00 -3.14731866e-01
-8.66671383e-01 7.31344640e-01 -1.28697500e-01 6.05609454e-02
-3.17047656e-01 -6.22610629e-01 4.33065474e-01 -2.43529268e-02
-6.29554614e-02 -1.29150584e-01 -5.31889439e-01 -1.26700497e+00
-2.03982100e-01 -1.16644263e+00 6.82649314e-01 -6.03544891e-01
-1.36752045e+00 7.02653229e-01 -6.52259588e-02 -1.21441257e+00
-9.51939747e-02 -1.07009542e+00 -1.73406184e-01 6.19355083e-01
-1.32571065e+00 -1.03336132e+00 1.87573135e-01 4.07118767e-01
9.02967572e-01 -7.10771263e-01 1.12096167e+00 1.79376319e-01
-5.23050129e-01 -4.03722608e-03 -1.88886859e-02 3.10240742e-02
1.06154108e+00 -1.09802067e+00 5.56375384e-02 6.27795994e-01
-3.19670320e-01 8.06071043e-01 8.14897895e-01 -1.01101828e+00
-1.21589124e+00 -3.82767767e-01 1.90179765e+00 -3.47337306e-01
1.10466421e+00 1.13911316e-01 -2.77143359e-01 6.90640807e-01
6.45609319e-01 -1.03878403e+00 1.39670146e+00 4.06526536e-01
1.64120242e-01 -2.45467518e-02 -1.14443970e+00 8.64871383e-01
1.25671297e-01 -3.45516533e-01 -9.47517872e-01 1.22020051e-01
-1.52141005e-01 -1.23655334e-01 -7.04562485e-01 -8.77970532e-02
6.22383118e-01 -8.87077808e-01 4.83456969e-01 -2.39555448e-01
6.28235459e-01 -2.31713772e-01 -3.44190270e-01 -1.28032553e+00
-1.28110098e-02 -5.20904243e-01 7.01937616e-01 1.31539607e+00
4.79818553e-01 -9.34802413e-01 8.90619874e-01 4.94745195e-01
-7.47548565e-02 -4.44188654e-01 -2.47807294e-01 -1.90155119e-01
3.28617036e-01 -7.17673004e-01 2.31728658e-01 9.89472270e-01
8.97234738e-01 5.81427336e-01 -1.07541278e-01 -4.22611475e-01
1.01937845e-01 -2.97000147e-02 8.29009295e-01 -1.30446529e+00
3.74041885e-01 -3.77395004e-01 -5.67599237e-01 -6.14966691e-01
3.83453459e-01 -6.67151213e-01 -4.23980445e-01 -1.66698396e+00
8.06971565e-02 3.68303172e-02 4.40454073e-02 2.58012831e-01
3.02991699e-02 4.45168346e-01 3.00354302e-01 2.77285546e-01
-2.18465358e-01 1.45233735e-01 1.10007143e+00 -1.65338591e-01
-3.12400818e-01 -5.26193380e-02 -1.28657126e+00 1.30405939e+00
9.93040025e-01 -3.71676207e-01 -4.62582350e-01 1.51772514e-01
1.19433343e+00 -5.01456499e-01 -5.94297945e-01 -6.11471117e-01
1.15345225e-01 -3.14711213e-01 3.23362239e-02 -1.03441417e+00
2.35222921e-01 -9.59347010e-01 -2.09318087e-01 1.61639825e-01
2.79717177e-01 5.19096494e-01 2.17065811e-01 -2.37327278e-01
-7.44335771e-01 -6.29559934e-01 6.52872384e-01 -1.38478026e-01
-9.38982189e-01 -2.60824174e-01 -1.30266809e+00 -2.42708456e-02
1.06573772e+00 -3.36074471e-01 -5.78369908e-02 -4.51930314e-01
-5.24944544e-01 1.09911717e-01 5.45277596e-01 4.88999039e-01
3.48785639e-01 -8.12835991e-01 -5.44131041e-01 -5.81040457e-02
2.25780010e-01 -5.31255186e-01 5.07200718e-01 7.25425661e-01
-8.38563085e-01 4.48353231e-01 -3.50636244e-01 -9.40993950e-02
-1.61443770e+00 2.57928193e-01 -3.32416594e-01 -3.45744997e-01
-1.93277195e-01 2.56340176e-01 -3.96612465e-01 -4.18079436e-01
-3.24709773e-01 -1.33400299e-02 -1.24050415e+00 7.94722319e-01
6.66348279e-01 5.07294774e-01 3.41811955e-01 -1.43761635e+00
-4.69519436e-01 7.21135736e-01 -2.40123436e-01 -3.16831261e-01
1.26747417e+00 -4.92396623e-01 -7.36575842e-01 9.67687607e-01
5.62323034e-01 6.56671345e-01 2.37497807e-01 3.78239542e-01
-8.79391208e-02 -2.34912753e-01 -3.48759219e-02 -7.56307602e-01
-1.01680958e+00 4.54853773e-01 4.11550514e-02 6.83961570e-01
1.05710948e+00 -1.43829659e-01 6.36337757e-01 3.96585792e-01
4.13941145e-01 -1.89436066e+00 -3.35828125e-01 1.14682853e+00
6.47226453e-01 -1.12459970e+00 1.33389011e-01 -7.17495263e-01
-1.10213375e+00 1.31120408e+00 1.50784925e-01 -8.90726224e-02
1.14847374e+00 4.00296569e-01 7.94950247e-01 -3.24090183e-01
-1.51149362e-01 -4.34001684e-02 -2.06016321e-02 3.90635192e-01
1.30699098e+00 2.03910649e-01 -1.41562474e+00 1.03345478e+00
-1.11835992e+00 -1.20019712e-01 8.76375437e-01 1.39964235e+00
-3.62865299e-01 -1.76912737e+00 -6.28530145e-01 4.17035997e-01
-1.19764996e+00 -2.99187992e-02 -7.55294800e-01 9.48566914e-01
-1.84460431e-01 1.73712480e+00 -1.26187220e-01 -4.43646908e-01
4.54361051e-01 -7.98722357e-02 2.76577044e-02 -7.79849708e-01
-9.65481043e-01 1.43033236e-01 3.56846333e-01 -2.64731105e-02
-1.12180543e+00 -9.79659498e-01 -1.43469644e+00 -6.49991751e-01
-2.60291606e-01 8.23712528e-01 9.39051092e-01 1.02920187e+00
1.54656582e-02 2.07521409e-01 4.48731959e-01 -6.30910814e-01
1.91527605e-01 -9.97241735e-01 -1.15102983e+00 1.62949905e-01
-2.89581716e-01 -4.96674895e-01 -5.37117362e-01 1.32437319e-01] | [11.060519218444824, 6.917701721191406] |
bdd2db54-d68c-4c62-ba4a-c05687add30f | multilayer-graph-contrastive-clustering | 2112.14021 | null | https://arxiv.org/abs/2112.14021v1 | https://arxiv.org/pdf/2112.14021v1.pdf | Multilayer Graph Contrastive Clustering Network | Multilayer graph has garnered plenty of research attention in many areas due to their high utility in modeling interdependent systems. However, clustering of multilayer graph, which aims at dividing the graph nodes into categories or communities, is still at a nascent stage. Existing methods are often limited to exploiting the multiview attributes or multiple networks and ignoring more complex and richer network frameworks. To this end, we propose a generic and effective autoencoder framework for multilayer graph clustering named Multilayer Graph Contrastive Clustering Network (MGCCN). MGCCN consists of three modules: (1)Attention mechanism is applied to better capture the relevance between nodes and neighbors for better node embeddings. (2)To better explore the consistent information in different networks, a contrastive fusion strategy is introduced. (3)MGCCN employs a self-supervised component that iteratively strengthens the node embedding and clustering. Extensive experiments on different types of real-world graph data indicate that our proposed method outperforms state-of-the-art techniques. | ['Xixu He', 'Wenbo Xu', 'Ling Tian', 'Zhao Kang', 'Liang Liu'] | 2021-12-28 | null | null | null | null | ['graph-clustering'] | ['graphs'] | [-1.78201839e-01 4.31481376e-02 -1.84558198e-01 -3.09972495e-01
1.58570260e-01 -1.84314653e-01 7.41244733e-01 4.93479192e-01
4.42876294e-02 2.03355327e-01 2.62263805e-01 -1.17248222e-01
-3.43426079e-01 -8.47332537e-01 -3.23348433e-01 -7.66532004e-01
-2.27645114e-01 2.12858588e-01 2.50358373e-01 -2.90382858e-02
-3.17539414e-03 1.15948983e-01 -1.12038434e+00 -2.35921759e-02
8.26566160e-01 7.07034945e-01 2.01635301e-01 3.04269582e-01
-1.95676208e-01 9.41658616e-01 -2.43511051e-01 -5.05255640e-01
2.90383715e-02 -4.17983800e-01 -6.32821560e-01 4.74122345e-01
-6.03220519e-03 1.74972013e-01 -4.54290599e-01 1.39018512e+00
3.20204824e-01 2.15851013e-02 5.23446977e-01 -1.49166310e+00
-7.91389704e-01 1.07995713e+00 -7.13400364e-01 -1.13620766e-01
2.66830828e-02 -1.88032184e-02 1.39202595e+00 -7.17020452e-01
3.06731761e-01 1.32901180e+00 4.72064644e-01 1.69461235e-01
-1.15780985e+00 -5.49965441e-01 6.96874797e-01 3.27482253e-01
-1.41609120e+00 9.66960788e-02 1.30637097e+00 -5.31588018e-01
5.30164421e-01 -3.25671434e-02 8.16283464e-01 8.36902082e-01
9.06765759e-02 6.91810250e-01 8.62423658e-01 -2.07243651e-01
2.09299028e-01 1.68732136e-01 3.19774598e-01 7.26568103e-01
3.75380129e-01 -3.97518039e-01 -1.26336321e-01 -7.13885427e-02
6.49788857e-01 4.48087484e-01 -3.21972370e-01 -6.21951222e-01
-1.15907228e+00 9.18494940e-01 9.63855505e-01 6.78047717e-01
-6.10219181e-01 -1.17899910e-01 3.58108401e-01 2.29605630e-01
4.90552872e-01 1.54450774e-01 -1.01263858e-01 5.31035185e-01
-5.23721993e-01 -2.70841837e-01 6.74102664e-01 7.61008859e-01
1.04092216e+00 7.86442980e-02 2.98041254e-01 8.33799481e-01
7.20144987e-01 -1.43356949e-01 3.98125410e-01 -5.01098812e-01
3.64261955e-01 1.34147072e+00 -5.62982976e-01 -1.64408588e+00
-3.12794060e-01 -7.56758630e-01 -1.41333449e+00 -6.93229362e-02
-2.48834923e-01 -1.54171437e-01 -6.08044028e-01 1.52722645e+00
3.94860953e-01 4.46443796e-01 9.74165723e-02 8.48702013e-01
1.10358834e+00 6.09090924e-01 1.53280720e-01 8.26383010e-03
1.18539536e+00 -1.05541766e+00 -7.12003708e-01 -1.74956977e-01
2.69106776e-01 -2.68735826e-01 5.87656915e-01 -4.13503647e-02
-7.91940570e-01 -5.84810913e-01 -1.21576834e+00 3.12050015e-01
-4.45287913e-01 -2.23791450e-01 8.00713420e-01 1.82028979e-01
-1.19127512e+00 4.20083731e-01 -7.96617031e-01 -5.01368761e-01
2.92595237e-01 3.78190786e-01 -6.09439015e-01 -1.97507262e-01
-1.27811563e+00 3.15913379e-01 7.03068435e-01 3.23371291e-01
-5.95669925e-01 -1.82768583e-01 -1.19694042e+00 3.16274166e-01
5.58518589e-01 -6.67577505e-01 4.30025846e-01 -1.12145555e+00
-1.18132043e+00 4.61427480e-01 1.56814784e-01 -2.34587297e-01
4.87680361e-02 2.32934967e-01 -5.90304613e-01 3.11353892e-01
-5.17073683e-02 5.24436831e-01 9.21743751e-01 -1.39115536e+00
-3.41038018e-01 -5.08162796e-01 -1.10280262e-02 5.12940288e-01
-8.20333779e-01 -1.74613178e-01 -7.06905901e-01 -6.63569152e-01
3.86737496e-01 -8.01303387e-01 -3.61516446e-01 -3.92171234e-01
-6.33255601e-01 -4.22293425e-01 1.02466893e+00 -4.27773118e-01
1.42350960e+00 -1.95597792e+00 6.27338648e-01 3.52825344e-01
9.10593867e-01 2.12395459e-01 -6.23372197e-02 7.57323623e-01
-2.34196901e-01 2.84919351e-01 -1.69421077e-01 -4.39360827e-01
-4.71739247e-02 1.54782787e-01 3.50028664e-01 4.16837811e-01
1.93837941e-01 7.76328683e-01 -8.11412573e-01 -6.72910690e-01
4.70665991e-01 5.61845124e-01 -3.46263915e-01 2.00997457e-01
9.84508768e-02 3.40695202e-01 -5.01861334e-01 6.33672297e-01
7.10540891e-01 -8.37295532e-01 6.55006111e-01 -4.54143524e-01
2.35589728e-01 -3.63979071e-01 -1.26951647e+00 1.29230917e+00
1.78648829e-01 4.92984891e-01 4.24785286e-01 -1.27703679e+00
8.06074142e-01 2.37702012e-01 5.37312925e-01 -1.78954378e-01
3.38658661e-01 -1.81279451e-01 2.79811144e-01 -3.80811930e-01
2.42450714e-01 1.21439390e-01 8.26848075e-02 3.14977467e-01
2.28453875e-01 3.63794893e-01 1.29996672e-01 5.29094517e-01
1.12384105e+00 -4.25028533e-01 4.81266439e-01 -3.31495643e-01
7.90615201e-01 -5.44598341e-01 7.42846429e-01 1.38573498e-01
-4.40334290e-01 1.73143923e-01 5.89376807e-01 -2.78837979e-01
-6.37013435e-01 -7.72302747e-01 3.20278436e-01 7.38993943e-01
4.19174999e-01 -5.46092451e-01 -6.59060717e-01 -7.46564388e-01
1.94439311e-02 1.74490884e-02 -7.52442002e-01 -1.47350296e-01
-3.43463182e-01 -6.79657996e-01 4.57930416e-02 5.11913478e-01
6.95921957e-01 -1.09327555e+00 1.13686860e-01 4.37767208e-02
-7.09796101e-02 -1.04187083e+00 -3.67847025e-01 -2.89053079e-02
-8.82005274e-01 -1.26205420e+00 -5.06654918e-01 -1.15049660e+00
8.57357681e-01 6.30704880e-01 1.01915574e+00 4.27085102e-01
1.97745651e-01 4.24869329e-01 -5.69109261e-01 -8.86678249e-02
-1.95919484e-01 2.98151553e-01 -9.89233926e-02 4.97078747e-01
2.76488543e-01 -8.61578047e-01 -6.07952356e-01 2.33021989e-01
-9.02256370e-01 1.81822181e-01 8.78333390e-01 6.88139856e-01
3.96724373e-01 4.58108544e-01 6.26906633e-01 -9.80092525e-01
9.49488759e-01 -8.25822353e-01 -3.71804327e-01 2.15587333e-01
-7.30969012e-01 -4.56168354e-02 6.98878586e-01 -4.04913753e-01
-7.29688346e-01 -3.05252075e-01 1.01994872e-01 -5.74020207e-01
-1.53261587e-01 1.03189933e+00 -4.70232964e-01 2.77134161e-02
7.60688633e-02 3.94808464e-02 4.30126972e-02 -3.24425340e-01
3.35215896e-01 5.78775048e-01 2.81614184e-01 -1.60411313e-01
8.08068633e-01 3.52531582e-01 -7.57126510e-02 -7.48354614e-01
-2.72518814e-01 -5.45291126e-01 -6.51166618e-01 -4.81414139e-01
8.72506797e-01 -9.24519658e-01 -8.55865717e-01 5.07421851e-01
-7.61608541e-01 -8.69965032e-02 3.16044271e-01 5.84344447e-01
1.60062149e-01 5.29443860e-01 -8.39907885e-01 -6.86202705e-01
-3.23128402e-01 -1.08280945e+00 6.84138179e-01 4.11694020e-01
1.68425843e-01 -1.42844486e+00 1.35437306e-02 2.28428707e-01
2.67822981e-01 3.90818834e-01 9.46394563e-01 -8.75469029e-01
-6.17834270e-01 -2.95804918e-01 -4.19490099e-01 3.36163789e-01
4.03339028e-01 6.59893900e-02 -5.70749462e-01 -4.54815626e-01
-2.29263127e-01 -9.54221860e-02 9.02061105e-01 2.58550972e-01
1.06288648e+00 -3.66706401e-01 -5.34813821e-01 3.13464701e-01
1.50946879e+00 1.18453860e-01 3.04512173e-01 5.48093021e-02
1.29630411e+00 7.82439470e-01 2.60180365e-02 2.74707198e-01
7.99038768e-01 1.94366127e-01 8.41210008e-01 -3.55035245e-01
1.93694249e-01 -2.13229239e-01 1.60246611e-01 1.62666571e+00
1.29268140e-01 -4.48275864e-01 -8.30933332e-01 6.05625629e-01
-2.16039777e+00 -8.57465863e-01 -9.30012092e-02 1.66970074e+00
2.78871596e-01 1.97005987e-01 1.37773693e-01 1.76909640e-01
1.31507015e+00 6.19953990e-01 -5.06108940e-01 -9.70961410e-04
-7.54149929e-02 -4.02380705e-01 -4.72943764e-03 3.83903652e-01
-1.14328206e+00 7.95496821e-01 4.83760881e+00 5.55679083e-01
-9.64789629e-01 -1.41915485e-01 6.28887713e-01 6.17538154e-01
-3.49120378e-01 1.93420351e-01 -2.83486217e-01 4.56470430e-01
4.52382714e-01 -8.98928288e-03 3.30988765e-01 8.11383009e-01
-1.46838248e-01 3.79734039e-01 -8.12404811e-01 8.18033874e-01
1.54399842e-01 -1.15883255e+00 6.47701994e-02 1.50122806e-01
7.00622261e-01 3.78805632e-03 -1.59502164e-01 4.31861430e-01
5.58836579e-01 -8.51155698e-01 2.72747099e-01 5.57575524e-01
1.47111684e-01 -8.37337136e-01 9.49174583e-01 4.35214669e-01
-1.66274619e+00 -8.00302550e-02 -1.52943462e-01 -9.26126540e-02
-8.72109234e-02 7.13137269e-01 -6.12124860e-01 1.04048264e+00
7.18347192e-01 1.39699340e+00 -7.92110205e-01 7.40918219e-01
-1.83857068e-01 6.14286423e-01 4.82337214e-02 -1.77136615e-01
3.66645962e-01 -5.11679471e-01 4.33279842e-01 9.31376398e-01
-1.48004040e-01 -6.68308735e-02 5.23214996e-01 6.75238907e-01
-3.13383758e-01 1.97584391e-01 -5.34894466e-01 -3.82224470e-01
6.44170523e-01 1.67423999e+00 -9.10909235e-01 -3.07184607e-01
-6.30477667e-01 8.56711686e-01 6.36622369e-01 5.15343428e-01
-6.22226596e-01 -4.44307297e-01 3.60905915e-01 -1.33822948e-01
4.47153151e-01 -2.08429977e-01 1.68416217e-01 -1.31961310e+00
-9.55633894e-02 -8.34815621e-01 6.28612697e-01 -5.02858222e-01
-1.56799507e+00 7.18769431e-01 -1.59859002e-01 -1.13843739e+00
5.26811928e-02 -2.99555331e-01 -7.94571936e-01 5.42976975e-01
-1.32639384e+00 -1.31284046e+00 -6.20591283e-01 6.49320483e-01
2.75413901e-01 -1.98017433e-01 6.15116477e-01 3.43442082e-01
-9.57063258e-01 3.09380680e-01 -3.93523164e-02 5.15211582e-01
4.50389087e-01 -1.28648508e+00 2.23710105e-01 9.39017117e-01
5.92506528e-02 9.10630882e-01 3.87690842e-01 -7.66933560e-01
-1.35371208e+00 -1.31359982e+00 5.84663153e-01 1.77783117e-01
7.67985523e-01 -4.51659411e-01 -1.09591186e+00 7.79473901e-01
6.24816298e-01 9.43176150e-02 7.03233600e-01 2.53615826e-01
-3.00931543e-01 -3.30325067e-01 -8.85457039e-01 7.73099363e-01
7.36307919e-01 -5.63950002e-01 -2.17703834e-01 2.32059937e-02
8.27400744e-01 1.70233414e-01 -1.17631590e+00 5.28362513e-01
3.25299561e-01 -9.76460159e-01 7.30825782e-01 -2.67174959e-01
3.04009140e-01 -5.02304137e-01 -9.65324864e-02 -1.47214007e+00
-6.40277445e-01 -3.97485375e-01 -2.73095369e-01 1.46037269e+00
5.67859598e-03 -8.03587437e-01 6.85578048e-01 1.93850845e-01
-2.71376781e-02 -9.33019221e-01 -3.99171770e-01 -2.73856938e-01
-3.25813800e-01 -1.41214728e-01 6.42661452e-01 1.45550454e+00
7.31829777e-02 7.49386370e-01 -3.03947240e-01 2.45250002e-01
7.93992579e-01 1.95254922e-01 6.79000735e-01 -1.64027047e+00
-2.30641991e-01 -6.34185076e-01 -7.41166353e-01 -7.95590699e-01
3.24286133e-01 -9.91591990e-01 -2.82793730e-01 -1.81536496e+00
4.13809448e-01 -2.34440759e-01 -5.82399189e-01 3.61158401e-01
-5.15037894e-01 -6.58021048e-02 1.54563308e-01 2.60766804e-01
-1.07273698e+00 8.23033452e-01 9.60859954e-01 -3.16643119e-01
-8.65706503e-02 -1.93002075e-01 -9.95759070e-01 7.92515278e-01
7.51090765e-01 -1.74117044e-01 -6.70964956e-01 -2.57598698e-01
9.15670022e-02 1.26503393e-01 1.96997404e-01 -1.04113102e+00
5.53378403e-01 1.40527815e-01 1.81522548e-01 -5.92115462e-01
6.31435439e-02 -9.86935675e-01 3.18195552e-01 3.57275367e-01
-2.09915236e-01 2.55928814e-01 -2.06595749e-01 1.07788682e+00
-4.22223508e-01 1.78503320e-01 6.13801003e-01 -1.09388813e-01
-7.32009768e-01 5.59312463e-01 -2.98113585e-01 -2.51376241e-01
9.19640243e-01 -8.30418840e-02 -2.96276480e-01 -5.05005062e-01
-6.82993710e-01 6.50680661e-01 4.82302099e-01 5.45851231e-01
7.19208062e-01 -1.48336875e+00 -6.16041303e-01 1.09857030e-01
1.89955756e-01 5.89415431e-02 3.08939099e-01 8.02682698e-01
-1.48572803e-01 8.63599926e-02 3.53193507e-02 -7.13052273e-01
-1.33152330e+00 7.69367337e-01 2.32643664e-01 -5.03524482e-01
-5.44203460e-01 5.90026975e-01 2.93642163e-01 -5.10299921e-01
2.67470479e-01 4.69549652e-03 -6.41196609e-01 1.34027243e-01
1.44140527e-01 2.76933938e-01 -1.73611730e-01 -9.10991609e-01
-4.09688294e-01 4.62419242e-01 -2.10387200e-01 3.09163928e-01
1.34188902e+00 -3.34434718e-01 -5.02446651e-01 5.45375764e-01
1.19824195e+00 -2.47258782e-01 -1.03294015e+00 -3.45507711e-01
3.77343483e-02 4.40521017e-02 2.16961190e-01 -2.56529897e-01
-1.41971016e+00 8.20705891e-01 3.50189269e-01 6.22493505e-01
1.21537554e+00 7.20794946e-02 5.26129961e-01 3.25887114e-01
1.41279399e-01 -6.93191469e-01 3.28165412e-01 2.18754902e-01
6.03189766e-01 -1.39914083e+00 2.03898922e-01 -5.18022776e-01
-5.95814049e-01 9.01606262e-01 6.97922051e-01 -3.08386981e-01
9.79993224e-01 -1.09975934e-01 -1.92983925e-01 -5.84812045e-01
-6.46131516e-01 -1.61678642e-01 3.51365983e-01 4.18583691e-01
3.63654196e-01 1.69075087e-01 -1.07385293e-01 6.95492923e-01
3.47991645e-01 -5.93945086e-01 2.06881776e-01 6.99535131e-01
-3.70422006e-01 -1.01318717e+00 -2.88446341e-02 3.99426907e-01
-3.03999513e-01 1.97173096e-03 -6.31468832e-01 9.06569421e-01
-5.42599298e-02 1.16411138e+00 -1.15119919e-01 -9.01835918e-01
-9.56094489e-02 -3.50998431e-01 -7.14337900e-02 -5.04984558e-01
-7.27897346e-01 2.91792274e-01 -2.01300040e-01 -4.00128603e-01
-8.06284726e-01 -5.09881496e-01 -1.04340255e+00 -2.67392069e-01
-5.26086509e-01 3.71943384e-01 3.82853299e-01 8.18752110e-01
5.30479193e-01 8.58427823e-01 7.08334446e-01 -8.21675599e-01
1.51558146e-01 -1.04981244e+00 -6.71726167e-01 4.44212228e-01
2.32610703e-01 -6.23341441e-01 -4.38660115e-01 -1.57327011e-01] | [7.329298973083496, 6.025482177734375] |
f863bacb-be8e-493f-8e13-4cf4fb8e86bf | beyond-labels-empowering-human-with-natural | 2305.12710 | null | https://arxiv.org/abs/2305.12710v1 | https://arxiv.org/pdf/2305.12710v1.pdf | Beyond Labels: Empowering Human with Natural Language Explanations through a Novel Active-Learning Architecture | Data annotation is a costly task; thus, researchers have proposed low-scenario learning techniques like Active-Learning (AL) to support human annotators; Yet, existing AL works focus only on the label, but overlook the natural language explanation of a data point, despite that real-world humans (e.g., doctors) often need both the labels and the corresponding explanations at the same time. This work proposes a novel AL architecture to support and reduce human annotations of both labels and explanations in low-resource scenarios. Our AL architecture incorporates an explanation-generation model that can explicitly generate natural language explanations for the prediction model and for assisting humans' decision-making in real-world. For our AL framework, we design a data diversity-based AL data selection strategy that leverages the explanation annotations. The automated AL simulation evaluations demonstrate that our data selection strategy consistently outperforms traditional data diversity-based strategy; furthermore, human evaluation demonstrates that humans prefer our generated explanations to the SOTA explanation-generation system. | ['Dakuo Wang', 'James Hendler', 'Shashank Srivastava', 'Yuxuan Lu', 'Lihong He', 'Sayan Ghosh', 'Yannis Katsis', 'Lucian Popa', 'Ishan Jindal', 'Bingsheng Yao'] | 2023-05-22 | null | null | null | null | ['explanation-generation'] | ['natural-language-processing'] | [ 1.59719080e-01 1.04186797e+00 -6.29752815e-01 -8.17057133e-01
-7.47419775e-01 -1.50446698e-01 3.53757769e-01 6.61743522e-01
-1.56187579e-01 6.76573217e-01 3.90128642e-01 -5.23385167e-01
-2.09689096e-01 -5.46888530e-01 -2.76502490e-01 -2.95042276e-01
-2.32408047e-02 9.42068458e-01 1.34930713e-02 1.10254101e-01
4.11268383e-01 2.09397776e-03 -1.79330087e+00 4.88599926e-01
1.30504084e+00 5.64934134e-01 -1.65670067e-02 6.26112163e-01
-3.94167513e-01 1.12741482e+00 -6.06348813e-01 -2.96264887e-01
2.67543465e-01 -4.28129107e-01 -9.11331177e-01 8.76201242e-02
1.12352841e-01 -2.57442564e-01 2.86724091e-01 4.19767559e-01
5.54467797e-01 2.60747463e-01 4.67126012e-01 -1.74124098e+00
-8.40926826e-01 8.09291601e-01 -1.54627204e-01 -1.85790271e-01
7.61711717e-01 3.19560766e-01 1.09588730e+00 -1.00097489e+00
5.04983306e-01 1.15755343e+00 6.41426444e-01 8.71682763e-01
-9.36953068e-01 -7.32402503e-01 4.93760794e-01 4.89127576e-01
-9.33439493e-01 -3.32628578e-01 6.80163383e-01 -4.58964974e-01
1.00130153e+00 4.86664951e-01 6.78756654e-01 8.29444885e-01
1.26828030e-01 9.64027405e-01 5.96599698e-01 -7.10591257e-01
6.31394923e-01 3.23267758e-01 5.53067267e-01 9.10109878e-01
4.05830353e-01 9.35801025e-03 -8.64746630e-01 -4.19997334e-01
4.25636530e-01 2.09346533e-01 4.04457301e-02 -3.41580272e-01
-1.08984911e+00 1.05957103e+00 2.88383514e-01 -3.28869104e-01
-4.17289823e-01 -1.65944964e-01 2.65193403e-01 1.58150733e-01
3.98675501e-01 9.69561100e-01 -7.17427731e-01 1.46833763e-01
-3.96099627e-01 3.63409072e-01 8.47060680e-01 1.31455505e+00
7.57629871e-01 -2.91206837e-01 -3.09211969e-01 5.54998934e-01
7.50565529e-01 2.96117235e-02 4.21062469e-01 -8.34928036e-01
3.58419567e-01 1.51690459e+00 2.98536181e-01 -8.82170141e-01
-4.79176849e-01 -2.23643094e-01 -6.08145535e-01 2.82937825e-01
-1.55905530e-01 -1.70249149e-01 -7.57573426e-01 1.44126678e+00
7.39926934e-01 1.07022703e-01 2.86493987e-01 9.61709142e-01
1.23213625e+00 2.29171410e-01 4.21087742e-01 -4.78850186e-01
9.69880044e-01 -1.35446656e+00 -9.06536758e-01 -3.68959308e-01
8.81620049e-01 -4.41133529e-01 1.27930319e+00 4.95480478e-01
-8.62191379e-01 -7.13576376e-01 -8.46861601e-01 -2.85528809e-01
-1.23740926e-01 1.52686648e-02 1.10163951e+00 4.29624826e-01
-6.53265357e-01 1.91977486e-01 -7.74237454e-01 -2.75430948e-01
3.88515174e-01 4.20626253e-01 -1.49075583e-01 1.90572172e-01
-1.13650024e+00 9.02240813e-01 4.36266541e-01 -1.98026314e-01
-8.21316600e-01 -7.08507299e-01 -9.86038864e-01 -3.18648592e-02
5.28629601e-01 -8.50098491e-01 1.45962489e+00 -1.02447379e+00
-1.02545965e+00 5.71129978e-01 -2.21761465e-01 -5.52368164e-01
4.91060376e-01 -5.74281037e-01 -4.80407178e-01 -3.20488870e-01
2.32690498e-01 9.18369055e-01 2.27613941e-01 -1.51965666e+00
-7.73872197e-01 -2.70796195e-02 3.72822553e-01 4.13876474e-01
-2.04299137e-01 -3.81190389e-01 -1.30730554e-01 -3.97147000e-01
1.31919399e-01 -8.36452663e-01 -7.65305996e-01 2.68543780e-01
-7.03048050e-01 -6.29576862e-01 8.14904451e-01 -1.65321350e-01
1.51372039e+00 -1.97624195e+00 -3.80089939e-01 1.90547004e-01
5.80097377e-01 1.54747382e-01 9.60776210e-02 2.32936874e-01
-2.08233610e-01 2.39113748e-01 -2.40326636e-02 -5.29595554e-01
7.31454343e-02 3.91163141e-01 -4.16223347e-01 -2.31655631e-02
1.18163079e-01 6.30502462e-01 -1.14015543e+00 -9.13604379e-01
2.66649812e-01 3.78262885e-02 -9.53013837e-01 7.88273096e-01
-6.00278616e-01 5.85099220e-01 -5.57715893e-01 4.95574802e-01
2.01039970e-01 -5.46372294e-01 1.65936381e-01 1.78072944e-01
-8.55803266e-02 3.55851054e-01 -1.12880957e+00 1.64356804e+00
-4.34403598e-01 2.42387325e-01 -6.03533328e-01 -4.96597141e-01
1.23280919e+00 5.28667808e-01 5.32188416e-01 -5.87807000e-01
-1.06175557e-01 1.17751040e-01 1.38227850e-01 -8.74862373e-01
4.47549552e-01 3.75972062e-01 -1.06547825e-01 9.25145030e-01
-4.20292944e-01 3.12589973e-01 1.74626336e-02 4.39538568e-01
1.16420197e+00 2.01622084e-01 9.23307598e-01 -1.80140622e-02
4.15061951e-01 4.56983447e-01 8.42552364e-01 9.24656570e-01
-1.64982289e-01 4.38197404e-01 1.27417371e-01 -1.01255250e+00
-8.36352646e-01 -6.67695284e-01 2.49743968e-01 1.14644194e+00
4.50893164e-01 -7.56598830e-01 -5.58553874e-01 -1.33141518e+00
-1.59509014e-02 1.31696951e+00 -6.14250600e-01 -1.51794448e-01
-1.57243520e-01 -3.27919602e-01 1.98519975e-01 4.67930704e-01
5.33996224e-01 -1.25785923e+00 -1.12731445e+00 1.05710328e-01
-1.35221794e-01 -4.38792318e-01 -3.72558981e-01 3.50750953e-01
-6.41607225e-01 -1.27104700e+00 3.71755630e-01 -5.24910808e-01
9.81576204e-01 8.15374553e-02 1.30990767e+00 6.63209498e-01
-2.77474970e-01 4.18933272e-01 -6.99418485e-01 -1.00905442e+00
-4.82020020e-01 -1.98138468e-02 -6.62468374e-02 -4.54228491e-01
6.95554852e-01 -1.89552948e-01 -7.51842856e-01 5.75568855e-01
-5.28159320e-01 5.18755198e-01 5.54495692e-01 7.56448507e-01
7.39581406e-01 -1.50103882e-01 8.37465465e-01 -1.59056675e+00
5.97255230e-01 -7.32615888e-01 -2.19515443e-01 5.98038852e-01
-1.36212969e+00 2.38752812e-01 6.59647644e-01 -3.92613649e-01
-1.21991527e+00 4.72401112e-01 -5.86013011e-05 -7.70211667e-02
-4.19265747e-01 3.26612949e-01 -2.78099388e-01 3.87611538e-01
1.05286074e+00 -4.10216302e-01 -2.58077443e-01 -1.91914171e-01
4.08089846e-01 7.77936757e-01 3.29029560e-01 -2.86058336e-01
5.98487616e-01 2.99068503e-02 -2.68382490e-01 7.24431202e-02
-1.31447077e+00 -5.15088439e-01 -5.92659056e-01 -2.86819875e-01
6.38165712e-01 -7.43307531e-01 -9.37627375e-01 -4.53532845e-01
-1.37688220e+00 -2.15013430e-01 -5.20159483e-01 3.99866730e-01
-5.64755976e-01 -3.49100493e-02 7.46157244e-02 -9.83597100e-01
-6.16372883e-01 -1.12332606e+00 8.93086135e-01 1.46243602e-01
-1.11824834e+00 -1.11005580e+00 1.82848319e-01 6.20584488e-01
2.80478954e-01 3.28464419e-01 1.23705316e+00 -1.34230137e+00
-6.74431145e-01 -9.34672803e-02 7.75700435e-02 -4.59978223e-01
2.39033163e-01 -1.57369107e-01 -1.01385343e+00 1.70082852e-01
-1.81145400e-01 -2.84240246e-01 2.42704049e-01 -1.50972055e-02
1.46725059e+00 -7.67216265e-01 -4.68063176e-01 6.69147745e-02
9.26895142e-01 2.46367887e-01 3.59205425e-01 2.30559632e-01
7.32634842e-01 7.52030015e-01 1.16151583e+00 6.13076746e-01
6.88812077e-01 5.06269574e-01 4.92420107e-01 -5.57586730e-01
-1.69357941e-01 -4.20531332e-01 -8.28806013e-02 6.67449653e-01
1.50353342e-01 -5.09775102e-01 -1.18104374e+00 2.60583252e-01
-2.53471684e+00 -7.63684452e-01 -4.93785858e-01 2.10838532e+00
8.63942325e-01 1.21220924e-01 -2.35611781e-01 3.04136127e-01
5.51734984e-01 -4.73036945e-01 -8.85658562e-01 -4.57248092e-01
3.73065770e-01 -3.59314680e-01 1.12428308e-01 6.13397241e-01
-8.52383018e-01 5.14985859e-01 6.89020014e+00 3.28842461e-01
-6.77040279e-01 1.99576691e-01 6.14564657e-01 6.93797618e-02
-6.60789788e-01 1.97408378e-01 -6.57854736e-01 3.28653812e-01
7.42072761e-01 -3.84008139e-01 -9.91298854e-02 1.48528051e+00
3.78363729e-01 -7.26713091e-02 -1.64932823e+00 1.00629640e+00
2.48528436e-01 -1.38234770e+00 4.40218508e-01 5.38801914e-03
8.44575644e-01 -7.20043123e-01 -1.76143453e-01 3.10092360e-01
7.52337813e-01 -1.09633303e+00 6.77379668e-01 6.84263587e-01
4.67055827e-01 -5.71781635e-01 1.01798236e+00 7.35417783e-01
-9.46767271e-01 -4.53284055e-01 -1.39483884e-01 -2.99916625e-01
2.16608614e-01 6.20491982e-01 -1.63377333e+00 4.45576549e-01
2.74996608e-01 5.61862886e-01 -7.91073978e-01 9.52619672e-01
-6.97098255e-01 5.41690052e-01 3.05384576e-01 4.23271693e-02
-3.07663590e-01 4.15438473e-01 2.38438323e-01 1.07028627e+00
1.21985942e-01 4.38721091e-01 7.20522881e-01 7.08260477e-01
3.83777358e-03 2.24471301e-01 -6.04129970e-01 2.50692338e-01
8.58880162e-01 1.10860980e+00 -5.82453430e-01 -3.82490158e-01
-2.73786783e-01 5.98179460e-01 3.84362340e-01 -4.89291474e-02
-7.60863364e-01 5.61413988e-02 1.92917913e-01 1.60307765e-01
-6.91980243e-01 2.54277140e-01 -9.63475704e-01 -6.99618340e-01
-3.71208668e-01 -1.01265609e+00 7.58639932e-01 -9.98233497e-01
-1.23583174e+00 6.34832859e-01 -6.21790588e-02 -1.46660388e+00
-5.76333940e-01 -1.36401236e-01 -7.58389771e-01 5.77487588e-01
-1.06069529e+00 -1.13286054e+00 -8.50198805e-01 4.02346879e-01
1.08928037e+00 -4.69155341e-01 1.24284339e+00 2.43847415e-01
-3.96959722e-01 6.68119669e-01 -7.83056676e-01 -2.04208612e-01
7.79794157e-01 -1.46154690e+00 4.34653401e-01 6.93788707e-01
1.11718349e-01 9.89848197e-01 9.41697121e-01 -1.03587449e+00
-1.13309240e+00 -1.11356413e+00 1.16140783e+00 -7.62104988e-01
5.18769920e-02 -2.09082305e-01 -1.00387001e+00 7.32567132e-01
9.46251377e-02 -2.70552307e-01 1.24257016e+00 4.87011850e-01
-1.11597694e-01 1.47308871e-01 -1.38185775e+00 5.22989333e-01
1.24845934e+00 -1.60090372e-01 -6.97445452e-01 5.12683153e-01
9.33343470e-01 -4.53567982e-01 -2.81378478e-01 3.00588697e-01
5.04352868e-01 -9.39965785e-01 6.32828057e-01 -7.81346321e-01
5.78833759e-01 -5.35906672e-01 2.77963549e-01 -1.08054101e+00
-4.36793119e-01 -4.02695000e-01 -3.13857615e-01 1.13347638e+00
9.00746822e-01 -3.44234049e-01 7.98630297e-01 1.29698277e+00
-4.23410445e-01 -8.24030995e-01 -3.39119494e-01 -2.26671517e-01
-8.20347369e-01 -4.13229853e-01 9.92191076e-01 1.21993458e+00
2.39790335e-01 5.79134703e-01 -2.76238918e-01 3.61680657e-01
4.63422954e-01 5.13727143e-02 1.05204201e+00 -1.63673615e+00
-2.73318708e-01 6.73926175e-02 -1.49173647e-01 -7.35236764e-01
2.03976557e-01 -7.27355361e-01 3.96819085e-01 -1.93902922e+00
3.49620342e-01 -8.83208096e-01 -6.91659823e-02 9.49847102e-01
-5.14857113e-01 -2.31611162e-01 -2.76151240e-01 4.31361765e-01
-1.18439615e+00 2.93481350e-01 9.60284054e-01 6.10687584e-02
-3.65032434e-01 -1.91813372e-02 -9.11308527e-01 9.62032676e-01
6.40311897e-01 -7.34350026e-01 -1.06258488e+00 -4.68833923e-01
5.34125328e-01 -8.42409730e-02 2.11380839e-01 -7.12947667e-01
6.99391961e-01 -5.85427582e-01 2.90196896e-01 -5.48764586e-01
-2.86287487e-01 -1.00273168e+00 3.15045267e-01 4.53523308e-01
-1.03818345e+00 1.74258620e-01 -1.07285179e-01 6.24904752e-01
5.87131418e-02 -2.06540301e-01 3.87649655e-01 -9.56997871e-02
-8.25254321e-01 2.69543618e-01 -1.09692387e-01 -2.90359497e-01
1.21529675e+00 -3.45748574e-01 -5.28952360e-01 -3.60778958e-01
-8.15114439e-01 6.16077900e-01 2.27393761e-01 4.00204599e-01
6.64691150e-01 -1.18791997e+00 -6.10079527e-01 1.81303471e-01
5.91368496e-01 5.83479464e-01 1.77986436e-02 4.28488106e-01
-3.40618849e-01 1.52869120e-01 -1.89558379e-02 -4.73642230e-01
-1.39290667e+00 6.32329583e-01 5.57652861e-03 -1.80083185e-01
-6.00124538e-01 7.10227907e-01 1.80562183e-01 -8.04758549e-01
5.39524257e-01 5.21708615e-02 -4.83668983e-01 -2.56567508e-01
5.68694115e-01 5.55929899e-01 -8.58677085e-03 5.56103280e-03
-1.72847793e-01 -1.05263174e-01 -1.26794547e-01 1.25168532e-01
1.17666757e+00 -2.68282622e-01 2.44813830e-01 7.57678330e-01
1.88799500e-01 1.08342310e-02 -1.08163989e+00 -1.46300063e-01
4.32167739e-01 -5.73028028e-01 -2.72458285e-01 -1.28038824e+00
-5.38559020e-01 6.12809658e-01 5.47911465e-01 2.77103484e-01
9.45412040e-01 5.44565991e-02 2.89971232e-01 6.13740683e-01
4.44586992e-01 -1.04251659e+00 4.63204116e-01 3.77722830e-02
1.05256701e+00 -1.51716495e+00 3.53633016e-01 -7.47149885e-01
-1.03227305e+00 8.92853081e-01 1.19322133e+00 3.92627418e-01
1.07044809e-01 2.08429068e-01 4.77862418e-01 -4.37838376e-01
-1.37253356e+00 1.97812885e-01 3.33138883e-01 5.78841567e-01
8.02948833e-01 7.02985674e-02 -3.32961053e-01 7.83888996e-01
-3.70957285e-01 1.20814266e-02 4.51352477e-01 8.36946070e-01
-4.70967084e-01 -1.16289914e+00 -2.32863560e-01 5.22929490e-01
1.60269573e-01 1.10681020e-01 -7.56906748e-01 5.37930965e-01
7.34253824e-01 1.26772857e+00 4.27979305e-02 -2.40680531e-01
3.06226611e-01 4.37176265e-02 -1.58588648e-01 -1.26690555e+00
-8.22689652e-01 -4.10918027e-01 2.15986565e-01 -4.84949201e-01
-4.92363721e-01 -3.96306425e-01 -1.89237368e+00 -2.00497270e-01
-5.31161189e-01 5.47421157e-01 3.59011859e-01 1.15403509e+00
5.13091266e-01 5.76777279e-01 6.01354361e-01 -9.72143263e-02
-4.29252237e-02 -8.08975041e-01 -1.19831674e-01 7.19282508e-01
2.63305098e-01 -6.54681265e-01 -2.51557678e-01 3.70773077e-01] | [8.958307266235352, 5.86250114440918] |
8638862e-5a48-4fcf-a960-b218f53e3aa0 | conflict-aware-pseudo-labeling-via-optimal | 2209.01847 | null | https://arxiv.org/abs/2209.01847v2 | https://arxiv.org/pdf/2209.01847v2.pdf | Conflict-Aware Pseudo Labeling via Optimal Transport for Entity Alignment | Entity alignment aims to discover unique equivalent entity pairs with the same meaning across different knowledge graphs (KGs). Existing models have focused on projecting KGs into a latent embedding space so that inherent semantics between entities can be captured for entity alignment. However, the adverse impacts of alignment conflicts have been largely overlooked during training, thereby limiting the entity alignment performance. To address this issue, we propose a novel Conflict-aware Pseudo Labeling via Optimal Transport model (CPL-OT) for entity alignment. The key idea is to iteratively pseudo-label alignment pairs empowered with conflict-aware optimal transport (OT) modeling to boost the precision of entity alignment. CPL-OT is composed of two key components -- entity embedding learning with global-local aggregation and iterative conflict-aware pseudo labeling -- that mutually reinforce each other. To mitigate alignment conflicts during pseudo labeling, we propose to use optimal transport as an effective means to warrant one-to-one entity alignment between two KGs with the minimal overall transport cost. Extensive experiments on benchmark datasets validate the superiority of CPL-OT over state-of-the-art baselines under both settings with and without prior alignment seeds. | ['Jie Yin', 'Daokun Zhang', 'Qijie Ding'] | 2022-09-05 | null | null | null | null | ['entity-alignment', 'entity-embeddings', 'entity-alignment'] | ['knowledge-base', 'methodology', 'natural-language-processing'] | [-8.66925046e-02 4.56423670e-01 -5.90691209e-01 -5.92650354e-01
-5.63103318e-01 -5.62168300e-01 4.94896561e-01 4.97292459e-01
-2.75732040e-01 5.77568591e-01 3.59073997e-01 -7.21248388e-02
-4.44068789e-01 -9.83165026e-01 -9.19284225e-01 -4.27414745e-01
-1.81716811e-02 5.33315957e-01 1.25904515e-01 -3.96598037e-03
-9.03493166e-02 1.23172857e-01 -8.82466614e-01 7.88201541e-02
1.32302332e+00 5.48632264e-01 -3.63571048e-02 -2.22745091e-02
-4.75271821e-01 3.61942798e-01 -7.78752044e-02 -9.32178020e-01
4.94183511e-01 3.59364972e-02 -1.00739968e+00 -2.49144122e-01
4.91515458e-01 -2.31768548e-01 -2.38595575e-01 1.09910846e+00
4.00735199e-01 -1.66425914e-01 6.55778587e-01 -1.67060411e+00
-9.63275969e-01 1.04148424e+00 -6.65549338e-01 -2.80902609e-02
1.28008515e-01 -1.24364898e-01 1.53810894e+00 -9.87979949e-01
7.14103222e-01 9.65695977e-01 7.85570502e-01 2.33984575e-01
-1.39309108e+00 -6.33836925e-01 4.90322709e-01 3.31896275e-01
-1.64875281e+00 -1.67984262e-01 6.41418517e-01 -3.84121865e-01
9.99721527e-01 -3.77778569e-03 6.05773032e-01 7.80201256e-01
-1.35273188e-01 9.03760612e-01 5.43672144e-01 -2.28674680e-01
-1.30772859e-01 2.11534843e-01 3.09441358e-01 7.80925095e-01
7.34204412e-01 -3.47824812e-01 -6.70832217e-01 -2.28583291e-01
4.35747147e-01 -1.49342746e-01 -2.32426941e-01 -8.61326754e-01
-1.60367119e+00 5.14082432e-01 8.27626467e-01 1.34840280e-01
-5.13912976e-01 1.18876137e-01 3.44759792e-01 1.19463891e-01
5.28256416e-01 7.64600396e-01 -6.19591832e-01 2.08550617e-01
-6.95074022e-01 2.82247305e-01 5.70213974e-01 1.39329147e+00
9.60959435e-01 -3.20510566e-01 -2.64877230e-01 5.65992475e-01
6.53649569e-01 2.16845691e-01 2.40335260e-02 -4.82075185e-01
9.21285748e-01 1.17009985e+00 1.65517792e-01 -1.25372672e+00
-3.32030475e-01 -6.49382234e-01 -5.84216774e-01 -4.83897537e-01
8.97485390e-02 2.07524304e-03 -5.59675992e-01 2.04008269e+00
7.71920383e-01 6.64933622e-01 2.88710058e-01 7.17813551e-01
7.05185533e-01 5.66711307e-01 4.06180143e-01 1.19838908e-01
1.38335907e+00 -1.33965707e+00 -6.91674113e-01 -2.34445348e-01
1.26648021e+00 -6.39705479e-01 9.38319743e-01 -6.16813540e-01
-7.59682000e-01 -2.97463268e-01 -1.22483158e+00 -6.19065985e-02
-5.72552383e-01 1.61924690e-01 7.38752306e-01 4.84730661e-01
-8.59693706e-01 5.42994261e-01 -7.31538475e-01 -3.01092029e-01
2.91322351e-01 5.19547224e-01 -7.37054229e-01 1.10425055e-01
-1.53609645e+00 8.90663207e-01 7.41670787e-01 2.85333753e-01
-1.04558922e-01 -1.08314943e+00 -1.01940572e+00 1.71930656e-01
5.78349948e-01 -1.05868864e+00 8.71810794e-01 -3.20093364e-01
-9.62025583e-01 6.71326220e-01 -2.69468457e-01 -3.06123316e-01
2.65560627e-01 -4.68588531e-01 -5.88274300e-01 -1.73581749e-01
5.84823132e-01 8.67110491e-01 4.14841354e-01 -1.36481583e+00
-6.77688360e-01 -1.67973444e-01 1.45487458e-01 5.61470270e-01
-7.68982232e-01 -3.99201930e-01 -8.20779204e-01 -5.38545072e-01
3.17502081e-01 -9.36262071e-01 -1.34398594e-01 -2.38674626e-01
-6.97720408e-01 -3.75468224e-01 6.26126170e-01 -5.76093853e-01
1.58585048e+00 -1.88307774e+00 3.76153558e-01 2.60937482e-01
4.20854568e-01 1.85772255e-01 -1.55695975e-01 5.40403306e-01
-3.76229547e-02 2.10310757e-01 -1.32997409e-01 -4.67587948e-01
3.33668113e-01 2.22937956e-01 -3.39103788e-01 2.64254957e-01
4.31143880e-01 1.32972920e+00 -1.22136211e+00 -7.15970039e-01
1.04570888e-01 1.16586693e-01 -4.69944984e-01 1.26732364e-01
-1.35491133e-01 2.53665090e-01 -6.58558846e-01 8.42247427e-01
7.89174676e-01 -5.66886604e-01 4.92665648e-01 -8.04132700e-01
1.65493667e-01 2.81341046e-01 -1.35157585e+00 1.73097634e+00
-2.68028408e-01 1.43270850e-01 -4.10870016e-01 -7.14973390e-01
6.43664479e-01 2.18196616e-01 6.98281229e-01 -4.10197526e-01
-2.10976854e-01 3.79048020e-01 -2.62015969e-01 -2.61912555e-01
9.67465997e-01 3.50682318e-01 -2.77923137e-01 4.25465375e-01
5.80159575e-02 6.46583915e-01 1.36628598e-01 4.63419735e-01
8.31427574e-01 3.11018944e-01 2.71499395e-01 -2.38646239e-01
5.38636327e-01 -2.15792432e-02 9.01328743e-01 4.37374651e-01
-1.38227537e-01 6.78685531e-02 3.42192650e-01 -3.45150322e-01
-1.00028777e+00 -1.10746169e+00 -2.43444331e-02 7.56165266e-01
9.37383354e-01 -7.75318205e-01 -5.77275634e-01 -1.18852329e+00
2.63846397e-01 7.04951406e-01 -5.02128124e-01 -2.87361175e-01
-7.13831365e-01 -8.49016249e-01 7.11685121e-01 8.07885170e-01
5.46149671e-01 -6.60778821e-01 -1.13872178e-02 4.47100729e-01
-3.67979646e-01 -1.41271985e+00 -7.58039236e-01 -1.29806995e-01
-6.29049838e-01 -9.81507480e-01 -4.24697369e-01 -8.30616772e-01
1.13948822e+00 2.40386352e-01 9.82363939e-01 -2.36859992e-02
1.70074075e-01 2.30779886e-01 -2.79124439e-01 7.17777237e-02
-2.82733664e-02 6.68751776e-01 2.42682621e-01 8.08234513e-02
5.63195705e-01 -5.31600714e-01 -6.41799450e-01 5.88679135e-01
-6.99141622e-01 4.91102010e-01 7.59835780e-01 7.86810279e-01
7.51137078e-01 -1.52992979e-01 5.53404927e-01 -1.13463497e+00
4.47475702e-01 -7.00806558e-01 -3.67872357e-01 8.57881486e-01
-1.10739255e+00 4.38368201e-01 4.60044235e-01 -3.28304470e-01
-8.27809751e-01 -1.35653913e-01 1.76050082e-01 -4.84117538e-01
2.96695679e-01 6.39725208e-01 -5.61836958e-01 -7.20422389e-03
1.00949548e-01 8.72327611e-02 -6.65007293e-01 -3.06764752e-01
8.35903645e-01 4.64797795e-01 5.84726095e-01 -8.84378672e-01
1.02116215e+00 2.60289580e-01 -5.36179505e-02 -1.14126347e-01
-9.11954701e-01 -6.69045985e-01 -8.59717786e-01 9.44948643e-02
8.21477056e-01 -1.10024881e+00 -2.81426430e-01 3.44178289e-01
-1.04036772e+00 4.89393212e-02 -2.30505347e-01 4.67118621e-01
-1.81662053e-01 5.07785499e-01 -3.59117448e-01 -2.81751275e-01
-5.26018679e-01 -1.08275831e+00 1.08023322e+00 4.31667358e-01
-2.29393423e-01 -1.13304329e+00 1.10275283e-01 2.91716605e-01
2.13205293e-01 1.50962442e-01 1.00406051e+00 -7.73304999e-01
-9.18245375e-01 -1.00371696e-01 -5.17984211e-01 -1.49282947e-01
4.87060428e-01 -4.87891473e-02 -6.97057605e-01 -2.30703562e-01
-8.35450828e-01 1.03376009e-01 5.46204984e-01 -4.01206128e-02
5.84582865e-01 -3.49929810e-01 -8.64621162e-01 7.64098763e-01
1.50071287e+00 -2.98172742e-01 3.00075561e-01 6.96125269e-01
1.18336034e+00 5.06451964e-01 9.70882118e-01 -4.43217671e-03
1.04986620e+00 1.08930087e+00 3.28366160e-01 -7.44076744e-02
-3.86680327e-02 -7.94134498e-01 1.25349596e-01 1.08842599e+00
2.48155028e-01 -3.48951578e-01 -8.03949654e-01 9.73949194e-01
-2.22781992e+00 -6.73106790e-01 -2.60473341e-01 2.06329274e+00
8.58492851e-01 8.21447223e-02 -2.00219482e-01 -3.25250775e-01
7.99125493e-01 2.36090481e-01 -4.08358097e-01 1.02049261e-01
-9.63782594e-02 -3.86216104e-01 7.75355458e-01 4.27925140e-01
-1.19863677e+00 1.22476053e+00 4.84619808e+00 7.08646119e-01
-8.27206492e-01 9.04943794e-02 1.55325875e-01 3.72732997e-01
-6.98883832e-01 4.09060895e-01 -1.24397528e+00 6.19931340e-01
4.47095394e-01 -4.20963138e-01 -1.69376340e-02 8.78067195e-01
-2.87378311e-01 4.01672482e-01 -1.11758566e+00 5.28548539e-01
-3.63131642e-01 -1.49607623e+00 3.20038706e-01 1.19567968e-01
9.01530683e-01 -1.31801292e-01 -5.55403084e-02 4.54528153e-01
3.34699929e-01 -4.97865736e-01 6.62382424e-01 3.30831379e-01
7.20240474e-01 -5.57979882e-01 8.37622046e-01 -3.62856798e-02
-1.85037374e+00 2.96726078e-01 -2.99233168e-01 6.24071836e-01
5.45274436e-01 4.94935125e-01 -1.05806577e+00 1.46564615e+00
4.76644933e-01 9.44919586e-01 -5.42510986e-01 8.73100996e-01
-4.63796049e-01 3.11326504e-01 -4.24643099e-01 2.75419474e-01
3.42878580e-01 -2.77042449e-01 6.48393571e-01 1.20448470e+00
2.53369331e-01 -1.80413127e-01 2.01264203e-01 8.46076906e-01
-4.79112446e-01 3.59283119e-01 -4.82824773e-01 -2.26083770e-01
1.11705637e+00 1.33651519e+00 -6.03329897e-01 -3.16958755e-01
-5.67271471e-01 1.11912036e+00 6.54860795e-01 1.40954360e-01
-1.15752733e+00 -3.31311822e-01 8.67863417e-01 -3.57013866e-02
2.55198866e-01 -8.77394453e-02 -3.06082398e-01 -1.27849388e+00
3.61743987e-01 -3.83202910e-01 5.76821566e-01 -3.61225665e-01
-1.30910313e+00 4.32059437e-01 2.37778515e-01 -1.28677356e+00
2.23742872e-01 -1.45371452e-01 -5.77173829e-01 8.28628004e-01
-1.86516273e+00 -1.70724046e+00 -2.82234013e-01 2.25677297e-01
2.35020101e-01 -4.63576578e-02 8.33473027e-01 6.16587698e-01
-8.40472102e-01 1.05125356e+00 -4.76213358e-02 9.44016352e-02
9.00132298e-01 -1.51285136e+00 8.20494115e-01 1.06308937e+00
3.02175909e-01 1.09059858e+00 4.29742426e-01 -9.46664929e-01
-1.09068644e+00 -1.46873486e+00 1.24729824e+00 -4.65363652e-01
6.54873729e-01 -1.85705945e-01 -1.05036426e+00 9.46387112e-01
5.02266809e-02 7.38176778e-02 8.64101648e-01 4.58149046e-01
-6.58721626e-01 -1.75839677e-01 -9.02890205e-01 8.10682356e-01
1.31322014e+00 -5.09099185e-01 -6.45164549e-01 2.41178840e-01
1.13680494e+00 -4.69141573e-01 -1.27521503e+00 7.82723665e-01
5.48531055e-01 -3.83792937e-01 1.13145244e+00 -7.06718206e-01
2.07505837e-01 -8.37186396e-01 -1.68025434e-01 -1.24160063e+00
-4.07871842e-01 -5.00100434e-01 -5.11304200e-01 1.66881192e+00
7.05744326e-01 -7.40979433e-01 7.18995154e-01 7.28781402e-01
-1.58236340e-01 -1.03984034e+00 -7.21304178e-01 -9.05672312e-01
-3.17263961e-01 3.44135985e-02 1.14338446e+00 1.51993716e+00
-3.04928049e-02 3.74241233e-01 -4.38411653e-01 9.71343637e-01
6.06134057e-01 3.65283936e-01 1.04381287e+00 -9.49818134e-01
-1.78934410e-01 -3.34163994e-01 -5.59436679e-01 -1.17128551e+00
3.26284528e-01 -1.15034056e+00 -3.37313831e-01 -1.83050334e+00
2.74859965e-01 -8.92135262e-01 -6.06558323e-01 7.43826151e-01
-7.22362041e-01 -1.82259813e-01 4.27613705e-02 4.36308175e-01
-8.27033579e-01 7.61299372e-01 9.53851521e-01 -1.23902775e-01
-1.15369081e-01 -3.10161471e-01 -8.89245450e-01 4.40710187e-01
5.03439128e-01 -7.34176517e-01 -6.09006226e-01 -6.35737419e-01
5.26699960e-01 -4.54724312e-01 1.20541453e-01 -6.63943172e-01
5.68435729e-01 -1.71025693e-01 -3.72047573e-02 -6.22264862e-01
-1.82603858e-03 -9.68842208e-01 4.74404633e-01 1.74970761e-01
-2.29557827e-01 1.44461393e-01 1.12329237e-01 8.51257861e-01
-2.30484650e-01 -8.36922452e-02 2.61233777e-01 4.57634896e-01
-1.10028720e+00 5.87366939e-01 5.92380285e-01 -4.12837230e-02
1.32323110e+00 -1.67139828e-01 -5.93767166e-01 7.33567104e-02
-4.94747639e-01 9.01137471e-01 5.88064075e-01 6.43720925e-01
1.92170858e-01 -1.71335006e+00 -6.50014281e-01 1.08289123e-01
6.28623843e-01 2.64556020e-01 3.35082680e-01 9.89322126e-01
-3.91735345e-01 4.51660961e-01 4.19824347e-02 -4.86821622e-01
-1.24465144e+00 3.28339994e-01 2.10167795e-01 -7.69905329e-01
-6.79067731e-01 9.46160495e-01 2.87367970e-01 -9.53685701e-01
9.43323597e-03 -2.88879164e-02 -7.94655904e-02 -1.31226227e-01
3.03627253e-02 2.12704480e-01 -6.88914508e-02 -6.87304854e-01
-5.32096028e-01 6.36030495e-01 -5.42256653e-01 2.60161340e-01
1.25680196e+00 -2.96309114e-01 -1.75388604e-01 1.05905540e-01
1.03684270e+00 2.65668988e-01 -1.04057550e+00 -5.63297331e-01
4.06449407e-01 -6.11998081e-01 -1.98568612e-01 -5.52983344e-01
-8.90812635e-01 4.14113820e-01 2.24737689e-01 -3.20683792e-02
8.03868175e-01 -6.45028502e-02 1.03234911e+00 4.00161713e-01
4.71418202e-01 -9.41209376e-01 -1.74099848e-01 1.46325767e-01
2.47037038e-01 -1.07596183e+00 1.36788219e-01 -8.63434911e-01
-6.94505334e-01 7.15381801e-01 1.00237787e+00 2.84660429e-01
3.42167407e-01 -8.91484469e-02 -1.44974738e-01 -4.03043121e-01
-5.19475162e-01 -1.27825215e-01 4.63240594e-01 2.68336773e-01
1.02038756e-01 3.29327695e-02 -3.09992015e-01 5.77894628e-01
1.15679413e-01 -2.98903316e-01 3.99374450e-03 9.80447531e-01
-2.37667598e-02 -1.54095769e+00 3.65342468e-01 3.50002497e-01
-1.06774308e-01 -1.97916940e-01 -3.40432286e-01 8.15115631e-01
1.72973916e-01 3.96587729e-01 -2.59477705e-01 -5.25292933e-01
3.24659646e-01 1.45186633e-01 1.88886106e-01 -5.49644589e-01
-4.15352285e-01 -2.88543969e-01 1.36414513e-01 -4.38597143e-01
-3.71988624e-01 -4.81307089e-01 -1.29327226e+00 -3.27165037e-01
-9.09191251e-01 2.63272345e-01 3.95009995e-01 8.50184202e-01
8.66036654e-01 5.15798032e-01 5.15253425e-01 -2.06649795e-01
-3.90511751e-01 -6.83446467e-01 -4.65800792e-01 4.86321032e-01
-9.40199345e-02 -8.44606161e-01 -5.30310534e-03 -7.23398253e-02] | [8.726268768310547, 7.984524726867676] |
3df5db4d-a4d9-4841-9596-0b94808e9cb7 | a-central-asian-food-dataset-for-personalized | 2305.07257 | null | https://arxiv.org/abs/2305.07257v1 | https://arxiv.org/pdf/2305.07257v1.pdf | A Central Asian Food Dataset for Personalized Dietary Interventions, Extended Abstract | Nowadays, it is common for people to take photographs of every beverage, snack, or meal they eat and then post these photographs on social media platforms. Leveraging these social trends, real-time food recognition and reliable classification of these captured food images can potentially help replace some of the tedious recording and coding of food diaries to enable personalized dietary interventions. Although Central Asian cuisine is culturally and historically distinct, there has been little published data on the food and dietary habits of people in this region. To fill this gap, we aim to create a reliable dataset of regional foods that is easily accessible to both public consumers and researchers. To the best of our knowledge, this is the first work on creating a Central Asian Food Dataset (CAFD). The final dataset contains 42 food categories and over 16,000 images of national dishes unique to this region. We achieved a classification accuracy of 88.70\% (42 classes) on the CAFD using the ResNet152 neural network model. The food recognition models trained on the CAFD demonstrate computer vision's effectiveness and high accuracy for dietary assessment. | ['Mei-Yen Chan', 'Huseyin Atakan Varol', 'Arman Bolatov', 'Aknur Karabay'] | 2023-05-12 | null | null | null | null | ['food-recognition'] | ['computer-vision'] | [ 1.75287545e-01 -2.52759397e-01 -4.26438332e-01 -4.62026596e-01
-3.83515000e-01 -5.03145874e-01 1.07758410e-01 9.73619998e-01
-3.61056209e-01 1.83806866e-01 5.97053349e-01 1.49063766e-02
4.58221853e-01 -1.15549612e+00 -6.77337348e-01 -6.22669637e-01
-2.00676098e-01 -3.28862667e-01 -2.42991909e-01 -4.84546348e-02
-2.09264606e-01 -1.66591123e-01 -1.55816519e+00 4.51873988e-01
9.30622518e-01 1.04339695e+00 5.84950745e-01 4.11601126e-01
-1.20887803e-02 6.17199600e-01 -2.68116832e-01 -2.39588231e-01
2.73325682e-01 -5.97596407e-01 -1.12002976e-01 1.58472165e-01
6.85837924e-01 -7.20497429e-01 -1.71151891e-01 1.54305696e+00
3.90569806e-01 -8.90337750e-02 4.16678309e-01 -4.83438611e-01
-1.92559516e+00 1.16782045e+00 -4.13610250e-01 2.04520553e-01
6.13258123e-01 3.78528573e-02 5.41719317e-01 -5.42002618e-01
3.11841965e-01 9.26290512e-01 8.53420436e-01 1.48031592e-01
-1.01709640e+00 -9.20374274e-01 1.13190621e-01 1.77150577e-01
-1.28979254e+00 -2.99830019e-01 5.73958099e-01 -5.78875542e-01
6.13146782e-01 1.37075022e-01 1.39581025e+00 1.02112317e+00
2.39174321e-01 5.90237617e-01 1.21941113e+00 -4.99337405e-01
1.64176315e-01 2.27140650e-01 4.71783102e-01 6.73877776e-01
6.66911662e-01 7.69113377e-02 -8.95700604e-02 1.99195117e-01
5.89915693e-01 6.51356041e-01 -1.23773247e-01 -6.93452731e-02
-1.28881085e+00 1.10721529e+00 8.64938855e-01 3.16591173e-01
-6.47864640e-01 -4.12715346e-01 3.00289840e-01 4.02323514e-01
8.58673990e-01 5.54371327e-02 -1.54202506e-01 5.68746805e-01
-6.00170434e-01 -1.49887964e-01 7.29441524e-01 6.41114295e-01
2.98418880e-01 9.55707133e-02 3.31622750e-01 1.03583193e+00
6.74347043e-01 1.17394459e+00 4.96945649e-01 -4.58511055e-01
4.03371118e-02 7.64678419e-01 1.66246995e-01 -1.85205650e+00
-6.44811928e-01 -1.14910103e-01 -1.15855372e+00 -3.99111249e-02
5.89691818e-01 -8.02784711e-02 -7.02201724e-01 1.38560462e+00
5.39159894e-01 -3.69437665e-01 -2.28366669e-04 1.13414669e+00
1.39629352e+00 7.02412367e-01 6.39127493e-01 -5.64600565e-02
1.83062553e+00 -7.94770777e-01 -7.95370758e-01 7.01980740e-02
4.05991226e-01 -9.08542633e-01 7.95341611e-01 2.88353533e-01
-7.41077304e-01 -7.71507561e-01 -1.34521806e+00 -4.02723067e-02
-6.73226357e-01 3.02388191e-01 7.14387596e-01 1.00309491e+00
-6.75618827e-01 4.13088799e-01 -8.05547118e-01 -7.78844774e-01
3.78255755e-01 -1.70546060e-03 -3.98327738e-01 -4.32127923e-01
-1.11124277e+00 6.75530732e-01 3.09598416e-01 2.26288676e-01
-5.15246511e-01 -8.82842004e-01 -1.23941100e+00 -3.41690570e-01
8.79417509e-02 -4.08887506e-01 1.16216004e+00 -1.40047932e+00
-1.34283662e+00 1.01762378e+00 3.41608465e-01 -4.89222139e-01
-5.77570498e-02 -2.51864374e-01 -1.07495582e+00 1.83681846e-01
2.37969398e-01 7.04957008e-01 2.01427370e-01 -7.51256466e-01
-5.74512899e-01 -7.10679531e-01 -8.61397199e-03 -4.51273508e-02
-3.87894660e-01 2.44878128e-01 4.91928428e-01 -7.38132834e-01
-3.87838073e-02 -7.84130216e-01 -1.70456946e-01 4.56785977e-01
-2.81829834e-01 -2.01028194e-02 1.81843385e-01 -1.20746386e+00
9.89156306e-01 -2.32780671e+00 -4.89772707e-01 -9.79090929e-02
1.58947241e-02 4.13356900e-01 -4.57128771e-02 2.01222926e-01
9.75237936e-02 -2.10107653e-03 2.28417069e-01 3.98904771e-01
1.58973143e-01 -7.23683015e-02 3.13351810e-01 8.93028617e-01
-3.13930094e-01 8.33440006e-01 -1.10860419e+00 -1.76910833e-01
5.01421809e-01 7.78431654e-01 -3.86360884e-01 -2.09292099e-01
-3.72623086e-01 6.26070574e-02 -4.72193599e-01 7.95342922e-01
8.72918665e-01 -3.06973130e-01 4.13915545e-01 -5.92986107e-01
-4.97900486e-01 -5.15058264e-03 -1.08238888e+00 1.66424048e+00
1.02763161e-01 1.05766714e-01 1.25233531e-01 -6.81003392e-01
9.43631232e-01 1.84235528e-01 5.69043756e-01 -9.13801134e-01
3.99546087e-01 -1.40747922e-02 -6.82664886e-02 -7.52879798e-01
4.45107400e-01 1.94057778e-01 6.72788499e-03 4.57390398e-01
-2.42714509e-01 6.81667984e-01 2.34062359e-01 -4.15388077e-01
3.01185578e-01 9.64485388e-03 6.99481368e-01 -6.89249039e-01
2.63220936e-01 4.49481130e-01 2.19028339e-01 3.46378773e-01
-5.44113457e-01 2.86105782e-01 -6.76860392e-01 -1.01344657e+00
-1.15243554e+00 -8.87832284e-01 -1.49066001e-01 1.10147774e+00
-6.00095801e-02 -1.69943333e-01 -6.12163603e-01 -2.50712395e-01
1.40654489e-01 4.17274266e-01 -7.06534445e-01 -2.52949484e-02
-3.71526599e-01 -9.39612508e-01 -1.83723662e-02 4.34549332e-01
1.03893387e+00 -9.29076612e-01 -3.68417114e-01 4.39820588e-01
-1.98043182e-01 -7.05934405e-01 -1.05445242e+00 -2.72717148e-01
-4.30542260e-01 -1.51824558e+00 -8.18878114e-01 -1.26365709e+00
5.40895343e-01 1.00058818e+00 1.07063365e+00 -2.29264021e-01
-3.72976035e-01 1.25362411e-01 -6.37972534e-01 -7.35247433e-01
-6.59238577e-01 -5.30668311e-02 -5.75371087e-03 -6.12378493e-02
1.24837649e+00 -2.09311023e-01 -1.30342126e+00 2.43699625e-01
-7.35809207e-01 1.52033433e-01 2.04811588e-01 2.28160113e-01
8.26675475e-01 -2.16846764e-02 5.67393959e-01 -5.58977425e-01
1.10284492e-01 -9.15333092e-01 -4.65665907e-01 1.77450910e-01
-3.91975671e-01 -6.46058738e-01 5.56960940e-01 -7.64472008e-01
-7.18557537e-01 3.28992993e-01 -2.22096652e-01 4.55146074e-01
-4.88028646e-01 6.18788838e-01 2.93665469e-01 1.75603285e-01
8.34063768e-01 -5.15770987e-02 1.99383706e-01 -7.20946074e-01
3.05157959e-01 9.42857802e-01 3.38266462e-01 6.11124747e-02
2.15474233e-01 4.06273037e-01 -6.15620434e-01 -9.85303104e-01
-1.21720338e+00 -4.84128237e-01 -3.96015853e-01 -3.49761069e-01
1.43588030e+00 -1.39266849e+00 -8.13713968e-01 6.70492947e-01
-2.81078011e-01 -3.68125349e-01 -4.20202166e-02 9.29954588e-01
1.93576247e-01 3.13767463e-01 -6.65328145e-01 -4.08510506e-01
-7.92317927e-01 -7.37930834e-01 5.50510764e-01 3.64022940e-01
-2.04688668e-01 -8.49811733e-01 1.58678606e-01 3.46999973e-01
4.86627460e-01 6.76857352e-01 4.66453880e-01 -2.56117553e-01
-4.89181578e-02 -9.43480358e-02 -4.38416213e-01 2.04012260e-01
7.41013169e-01 -3.50464098e-02 -6.67761445e-01 -2.47346178e-01
2.74504181e-02 -2.56751597e-01 7.78125525e-01 7.70913601e-01
6.12653852e-01 -3.75136375e-01 -5.64749613e-02 2.59126395e-01
1.64638186e+00 5.37214756e-01 3.57053459e-01 3.51891190e-01
5.90314031e-01 2.50678152e-01 1.96480334e-01 1.50757849e-01
1.03649139e+00 5.34012377e-01 2.18298510e-01 -3.02466184e-01
-5.54786205e-01 -3.24638605e-01 6.76408112e-01 1.09818709e+00
1.17893137e-01 7.79715776e-02 -5.00864089e-01 5.01921952e-01
-1.23660481e+00 -1.12402368e+00 -3.42390090e-01 1.90102816e+00
7.67031074e-01 -6.34276986e-01 5.57721853e-01 4.40387987e-03
8.04412961e-01 -6.87209377e-03 -5.15400827e-01 -1.26769289e-01
-1.59844682e-01 -1.53665483e-01 6.23010099e-01 4.22146842e-02
-1.67400825e+00 2.48202324e-01 6.74211550e+00 1.00893363e-01
-9.05974746e-01 1.48396671e-01 6.40906513e-01 6.65831640e-02
1.82665601e-01 -1.04207826e+00 -7.31549501e-01 5.26591718e-01
1.22334218e+00 1.98176906e-01 6.08018756e-01 7.79595077e-01
3.24626446e-01 -1.98993504e-01 -6.84537113e-01 6.30668819e-01
3.97299141e-01 -1.19912100e+00 -3.86351198e-01 -2.44073179e-02
7.78485596e-01 2.17556685e-01 -5.94278565e-03 4.60686795e-02
8.37814867e-01 -7.28628337e-01 9.26370084e-01 4.23843443e-01
6.06337249e-01 -5.72379470e-01 4.05783772e-01 2.59351954e-02
-1.50179505e+00 -2.64771998e-01 -8.14990461e-01 -2.08495319e-01
-8.38839039e-02 8.31447899e-01 -3.00870925e-01 3.59328359e-01
9.61376429e-01 1.17841876e+00 -5.83399653e-01 1.02590394e+00
2.18397096e-01 5.66528022e-01 -3.85477722e-01 -1.68607220e-01
8.76233056e-02 -4.63439435e-01 -1.32903188e-01 9.73852754e-01
7.74251282e-01 2.65744537e-01 7.16646850e-01 5.40940940e-01
6.43074289e-02 6.32331133e-01 -5.85561752e-01 -2.14487657e-01
5.80990948e-02 1.11294699e+00 -8.62660706e-01 -1.66108519e-01
-9.15121377e-01 5.83629310e-01 1.18084863e-01 -8.99469852e-02
-5.71787536e-01 3.60444546e-01 6.61258638e-01 -2.49185562e-02
3.08036089e-01 -2.04812035e-01 2.02403933e-01 -1.17852247e+00
-2.47136831e-01 -9.33381081e-01 7.35255182e-01 -5.11832833e-01
-1.69932961e+00 1.24099396e-01 -2.30296507e-01 -9.23227668e-01
3.13921064e-01 -4.87894803e-01 8.94783996e-03 3.63978505e-01
-1.61110306e+00 -1.31163394e+00 -4.25086200e-01 4.75180775e-01
4.30001080e-01 -1.51938740e-02 1.38498902e+00 6.98925674e-01
-7.97314823e-01 3.84010166e-01 5.12321174e-01 4.13510233e-01
5.01759350e-01 -1.01690388e+00 1.86625615e-01 3.31689417e-01
-2.03639176e-02 3.64738375e-01 5.37331104e-01 -8.66292119e-01
-1.64064717e+00 -1.54811049e+00 8.28418374e-01 -3.09989844e-02
5.71053147e-01 -1.20449483e-01 -6.30593657e-01 6.37063980e-01
4.25440848e-01 -3.86726171e-01 1.40005612e+00 -7.18916059e-02
-4.95894939e-01 -3.28094542e-01 -1.45330536e+00 4.39459264e-01
8.92454684e-01 -2.27461323e-01 -3.42772156e-01 6.45476341e-01
5.40406168e-01 -1.20450377e-01 -1.53053463e+00 -2.15605080e-01
9.21290278e-01 -6.68771207e-01 1.32223856e+00 1.78402979e-02
4.65420485e-01 -3.15169424e-01 -4.32078898e-01 -1.45749569e+00
-6.75500631e-01 3.09473753e-01 2.20181122e-01 1.09044337e+00
1.52196065e-01 -6.65115416e-01 4.43694234e-01 4.99161363e-01
-1.91612780e-01 5.34461159e-03 -2.33369932e-01 -3.80522132e-01
5.79758957e-02 1.24963760e-01 1.06434751e+00 1.22652459e+00
5.29543981e-02 -2.55443573e-01 -3.70480418e-01 2.46130854e-01
7.18479633e-01 4.41140056e-01 3.76602262e-01 -1.02393401e+00
3.89986932e-02 -1.35156423e-01 -1.46298811e-01 -5.65709233e-01
-2.99400300e-01 -1.20086575e+00 -1.68860912e-01 -1.54927766e+00
4.69228327e-01 -2.56591529e-01 -3.11932236e-01 5.11559904e-01
-4.90681157e-02 8.11959684e-01 1.65622190e-01 6.63643405e-02
-3.24362934e-01 -1.99525245e-02 1.54405928e+00 -5.86564362e-01
-2.82604277e-01 -5.54187953e-01 -1.21433127e+00 5.90488672e-01
1.16336370e+00 -3.23480695e-01 -2.43805632e-01 -5.27977228e-01
1.83285281e-01 -3.94477934e-01 3.33942711e-01 -9.66100693e-01
-5.67614741e-04 -2.97305137e-01 8.23051035e-01 -5.35611331e-01
-2.46770367e-01 -9.52986121e-01 9.01315212e-01 8.10316265e-01
-4.12938952e-01 -1.78885043e-01 -1.17012374e-02 4.62738901e-01
8.86186361e-02 1.46436527e-01 6.29624367e-01 -6.33928835e-01
-6.09269857e-01 7.02552050e-02 -5.96058726e-01 -5.66015840e-01
1.15657508e+00 -1.32605731e-01 -6.49571300e-01 1.22044057e-01
-8.34663332e-01 -3.47946696e-02 4.71310586e-01 4.23190445e-01
2.64983118e-01 -1.72309744e+00 -9.12241399e-01 5.73506355e-01
2.65250713e-01 -5.47907770e-01 7.42552638e-01 6.04944050e-01
-7.47654438e-01 3.27228904e-01 -4.95711327e-01 -4.05938596e-01
-1.04605591e+00 1.08547831e+00 1.42490491e-01 2.04509437e-01
-1.17082179e+00 7.84638301e-02 -1.42998677e-02 -2.47508079e-01
6.38149828e-02 -6.24431193e-01 -7.94749200e-01 2.92922765e-01
1.39015818e+00 2.60894001e-01 -1.70471147e-01 -9.21159565e-01
5.35413921e-02 4.80514675e-01 3.69146909e-03 9.91881609e-01
1.64250338e+00 -5.16215086e-01 6.53579757e-02 4.53424692e-01
1.04572177e+00 -2.62735099e-01 -1.05176771e+00 -1.93443313e-01
-5.20751894e-01 -4.81012583e-01 1.47326797e-01 -9.95509684e-01
-1.46247149e+00 3.10041845e-01 1.31566846e+00 7.08925605e-01
1.10874689e+00 -2.69599020e-01 1.18202877e+00 2.19859739e-05
3.44353527e-01 -7.98456132e-01 -4.00613129e-01 -1.21199101e-01
7.80826926e-01 -1.48085320e+00 3.82019095e-02 -5.22248805e-01
-2.00978741e-01 9.88477111e-01 -7.44718090e-02 -3.69060129e-01
1.06545472e+00 -1.21688649e-01 4.97842401e-01 -2.32138142e-01
-4.72937897e-02 -3.44987493e-03 3.31664234e-01 9.40764308e-01
5.89995205e-01 7.28513837e-01 -3.45886737e-01 6.38804674e-01
-3.21040273e-01 5.48387110e-01 1.70933351e-01 7.14551806e-01
-6.98464930e-01 -8.86084080e-01 -5.07812083e-01 5.85598350e-01
-6.25801027e-01 1.86149345e-03 1.56742215e-01 5.71636856e-01
3.00231427e-01 1.36801577e+00 1.04536638e-01 -1.13345563e-01
2.70445287e-01 1.42111316e-01 4.18420196e-01 -4.02809620e-01
-9.36469138e-01 1.70192540e-01 3.21571290e-01 -3.40260357e-01
-1.19143593e+00 -8.74812484e-01 -7.28550136e-01 -8.21543336e-01
-1.07064806e-02 -3.19930047e-01 6.80943787e-01 4.49522555e-01
1.82533398e-01 1.68108359e-01 5.54077506e-01 -8.86948168e-01
-2.78810889e-01 -9.61536825e-01 -8.50926936e-01 7.12589204e-01
1.11532345e-01 -4.30485874e-01 9.56972092e-02 4.33232516e-01] | [11.556360244750977, 4.422477722167969] |
eb5ea4f0-aab0-42bd-9e92-573fc6daf0b3 | cosqa-20000-web-queries-for-code-search-and | 2105.13239 | null | https://arxiv.org/abs/2105.13239v1 | https://arxiv.org/pdf/2105.13239v1.pdf | CoSQA: 20,000+ Web Queries for Code Search and Question Answering | Finding codes given natural language query isb eneficial to the productivity of software developers. Future progress towards better semantic matching between query and code requires richer supervised training resources. To remedy this, we introduce the CoSQA dataset.It includes 20,604 labels for pairs of natural language queries and codes, each annotated by at least 3 human annotators. We further introduce a contrastive learning method dubbed CoCLR to enhance query-code matching, which works as a data augmenter to bring more artificially generated training instances. We show that evaluated on CodeXGLUE with the same CodeBERT model, training on CoSQA improves the accuracy of code question answering by 5.1%, and incorporating CoCLR brings a further improvement of 10.5%. | ['Nan Duan', 'Ming Zhou', 'Daxin Jiang', 'Ke Xu', 'Ming Gong', 'Linjun Shou', 'Duyu Tang', 'JunJie Huang'] | 2021-05-27 | null | https://aclanthology.org/2021.acl-long.442 | https://aclanthology.org/2021.acl-long.442.pdf | acl-2021-5 | ['code-search', 'code-search'] | ['computer-code', 'computer-vision'] | [-2.02320829e-01 2.87110090e-01 -3.95927042e-01 -3.53276134e-01
-1.36351514e+00 -9.50068772e-01 2.44719967e-01 3.30835909e-01
-1.80630818e-01 2.94087231e-01 7.79683217e-02 -5.86866796e-01
3.12998772e-01 -4.59920108e-01 -8.75587344e-01 2.41425544e-01
1.52915403e-01 2.77978659e-01 4.77690816e-01 -2.02920079e-01
2.44759738e-01 -1.53898865e-01 -1.26532006e+00 7.88102865e-01
1.23644125e+00 6.46095157e-01 2.86237508e-01 6.79270804e-01
-5.96025646e-01 1.53141105e+00 -5.51144361e-01 -9.65184212e-01
-3.55741978e-02 -2.65623271e-01 -1.26969314e+00 -2.30259433e-01
6.47972524e-01 9.28642824e-02 -1.11785149e-02 1.14259970e+00
6.44859448e-02 -3.52896035e-01 3.55805099e-01 -1.29271090e+00
-1.03031909e+00 7.24963009e-01 -4.79008853e-01 1.44579008e-01
8.84235322e-01 -1.13999575e-01 1.38492596e+00 -1.14721274e+00
7.39863217e-01 9.29136395e-01 8.57034683e-01 7.26121306e-01
-1.28785408e+00 -6.57209992e-01 -3.87662083e-01 9.80366468e-02
-1.40913737e+00 -4.19575334e-01 4.18269455e-01 -7.39966035e-01
1.39074552e+00 2.14939147e-01 8.60147700e-02 5.07552326e-01
-6.59165084e-02 6.98418140e-01 5.06847501e-01 -6.92153037e-01
-8.26133415e-02 4.07436073e-01 1.67225644e-01 1.13939178e+00
-4.07314114e-02 -2.36540556e-01 -1.28194511e-01 -6.54243171e-01
2.11422607e-01 -2.81439483e-01 -1.45536244e-01 -4.57883656e-01
-8.70609045e-01 8.55703592e-01 4.62576121e-01 3.89343172e-01
1.56981558e-01 5.49627542e-01 5.59056997e-01 6.29384995e-01
3.33108187e-01 1.36318707e+00 -8.44313264e-01 -3.46295655e-01
-6.33801043e-01 3.73439729e-01 9.42553818e-01 1.58454919e+00
1.11577463e+00 -2.33410433e-01 -1.14368014e-01 1.06842542e+00
4.58514810e-01 7.56056249e-01 2.60361612e-01 -1.35055542e+00
7.09037066e-01 1.10295439e+00 4.28200141e-02 -8.58820319e-01
-1.03907488e-01 -4.81156826e-01 1.49437174e-01 -1.00913629e-01
3.89549464e-01 -3.87337655e-02 -4.25120711e-01 1.50102198e+00
-2.23371804e-01 -1.89822599e-01 9.45447758e-02 5.28568864e-01
6.94534183e-01 4.26448137e-01 1.41460329e-01 4.35415208e-01
1.38792741e+00 -1.18396997e+00 -4.60027814e-01 -4.59953606e-01
1.40644217e+00 -1.16885412e+00 1.37315714e+00 8.11312571e-02
-8.76235962e-01 -4.38481480e-01 -8.29935551e-01 -3.05032462e-01
-4.05035228e-01 3.21775794e-01 5.17373443e-01 5.82784772e-01
-1.21298659e+00 -1.19390823e-01 -6.91486835e-01 -2.61643827e-01
4.38952267e-01 5.14495894e-02 -1.10804945e-01 -4.94102269e-01
-7.54442036e-01 7.62578309e-01 2.02836115e-02 -6.84030890e-01
-8.65489602e-01 -1.04640448e+00 -1.02892864e+00 9.48891267e-02
5.40139735e-01 -2.55396456e-01 1.72457623e+00 -1.10206842e+00
-6.38701022e-01 1.34980512e+00 -3.58671874e-01 -2.45744601e-01
1.86198652e-01 -3.89328808e-01 -5.22658408e-01 5.14644347e-02
6.69624746e-01 7.00597942e-01 4.02080536e-01 -1.29250944e+00
-4.33576524e-01 -3.88862081e-02 6.40848756e-01 -2.64828831e-01
-2.33149394e-01 5.00116527e-01 -7.25599885e-01 -6.67867601e-01
-3.13354343e-01 -1.02051008e+00 4.17732634e-02 2.84986317e-01
3.37366089e-02 -5.55133581e-01 5.39180458e-01 -7.86824882e-01
1.21201336e+00 -2.24930096e+00 -1.30790368e-01 -1.53197348e-01
4.03277218e-01 2.23831803e-01 -4.62013930e-01 4.83828008e-01
-1.65738583e-01 3.00919741e-01 -4.92671877e-01 -1.28077477e-01
-2.05340143e-02 1.08042777e-01 -3.42865765e-01 8.37631375e-02
6.26432955e-01 1.14588022e+00 -1.25168955e+00 -5.83099663e-01
-5.59213996e-01 9.03248508e-03 -1.12968528e+00 4.25780267e-01
-5.93516052e-01 -1.69126689e-01 -4.96559530e-01 6.72166884e-01
2.96439320e-01 -5.81510186e-01 1.48340482e-02 3.84280473e-01
1.02873296e-01 4.46568489e-01 -5.17266750e-01 2.25447416e+00
-9.18667495e-01 9.89960492e-01 -1.12821482e-01 -5.45004189e-01
9.88887072e-01 4.41789120e-01 2.59485185e-01 -7.73699760e-01
-3.34095508e-01 5.08374393e-01 -1.36244878e-01 -1.11805558e+00
4.17735666e-01 1.32590458e-01 -3.10049623e-01 5.11344790e-01
-2.02700845e-03 -3.71030122e-01 1.37697086e-01 5.23918927e-01
1.57509387e+00 6.35534972e-02 -6.89894184e-02 -5.31553149e-01
7.52863824e-01 5.94673455e-01 3.47351432e-01 6.68014824e-01
-6.48132265e-02 3.35629195e-01 6.80254698e-01 -3.66884470e-01
-9.96085227e-01 -7.50743210e-01 -1.76446840e-01 1.51257539e+00
4.75466549e-02 -6.90565050e-01 -8.08680534e-01 -1.24936652e+00
1.16092257e-01 8.32793295e-01 -3.99208724e-01 -2.42290676e-01
-7.10829794e-01 -2.70119786e-01 1.04197133e+00 6.68672919e-01
2.58899331e-01 -7.86740005e-01 -5.45255482e-01 3.62729318e-02
-4.37775910e-01 -1.12172449e+00 -6.94755018e-01 3.59262228e-02
-4.77749854e-01 -1.46120536e+00 -5.37023842e-01 -1.24366307e+00
6.75734580e-01 1.63602129e-01 2.08976102e+00 9.00137544e-01
-1.84598848e-01 4.47274834e-01 -7.80157208e-01 -1.69480503e-01
-1.00915861e+00 3.91570836e-01 -7.00647712e-01 -7.39892066e-01
7.81337678e-01 -6.95922822e-02 -1.66100889e-01 2.25624204e-01
-8.91016483e-01 -2.35921785e-01 2.64043331e-01 7.46107697e-01
-1.07450314e-01 -4.89995778e-01 6.15190268e-01 -1.13682973e+00
6.17352545e-01 -8.18965554e-01 -7.91488886e-01 5.99099100e-01
-7.16036618e-01 3.86318892e-01 6.84192777e-01 -2.37332545e-02
-1.15530455e+00 9.56291705e-02 -3.04324567e-01 -8.93621296e-02
4.81306110e-03 6.90627635e-01 2.50196517e-01 -3.59234512e-01
1.28329611e+00 -2.22718179e-01 -3.42696935e-01 -5.66767931e-01
5.75220168e-01 8.63119543e-01 5.50937116e-01 -1.04440093e+00
7.28143990e-01 -3.45573872e-02 -7.06208348e-01 -2.54917651e-01
-7.26417363e-01 -6.81523800e-01 -3.77083868e-01 1.08414985e-01
9.81144071e-01 -1.10840988e+00 -3.17839921e-01 3.80715318e-02
-1.50925672e+00 -4.80139047e-01 1.43782878e-02 2.17358276e-01
-3.91688079e-01 2.15331063e-01 -5.72531641e-01 -3.95802200e-01
-1.09876990e-02 -1.21506035e+00 1.24207473e+00 -2.18388304e-01
-2.16756746e-01 -8.72733891e-01 2.97587544e-01 7.08192527e-01
4.51791197e-01 -1.65887207e-01 1.21561706e+00 -6.48923695e-01
-6.07828021e-01 -2.79131383e-01 -6.04550302e-01 3.66907269e-01
1.14891138e-02 -1.34333432e-01 -8.74861300e-01 -3.37691009e-01
-1.65621892e-01 -8.39348912e-01 3.86879712e-01 -5.02890170e-01
1.02701306e+00 -1.28058538e-01 -3.08069915e-01 2.68159449e-01
1.76539993e+00 3.18557680e-01 4.41913098e-01 7.26189166e-02
7.65776575e-01 6.21129513e-01 5.91918051e-01 1.59775496e-01
7.58167684e-01 7.16214180e-01 4.39535469e-01 -9.89159849e-03
-3.86135042e-01 -1.71429425e-01 1.41347036e-01 1.09874916e+00
4.01947081e-01 5.44908643e-02 -1.65890801e+00 9.99194145e-01
-1.59079611e+00 -4.02037859e-01 -3.10325146e-01 1.64824843e+00
1.14809430e+00 -1.70665547e-01 -1.80742890e-01 -4.16562617e-01
4.05847341e-01 -2.77035505e-01 -4.95323777e-01 -2.35453576e-01
2.24183336e-01 2.37865418e-01 3.72590929e-01 5.98916471e-01
-6.86299860e-01 9.86032188e-01 6.38917303e+00 6.11909151e-01
-6.66086376e-01 3.39633942e-01 2.01314077e-01 5.05552948e-01
-7.76626945e-01 3.10257256e-01 -2.78477222e-01 2.96232015e-01
8.37963343e-01 -2.55979031e-01 8.19692314e-01 1.18124664e+00
-2.98765749e-01 6.58717006e-02 -1.22520316e+00 8.09693515e-01
4.00596023e-01 -1.33023489e+00 -3.87341976e-01 -3.87584478e-01
1.07624435e+00 4.09614861e-01 -4.15087461e-01 7.76263475e-01
6.79276824e-01 -8.22833478e-01 6.96852505e-01 4.55400586e-01
9.00248945e-01 -4.38196450e-01 8.89591157e-01 3.58658671e-01
-1.23877454e+00 -1.95235983e-01 -3.08023810e-01 -1.45600243e-02
-4.93495792e-01 1.84963614e-01 -9.56568360e-01 3.31842542e-01
7.71097183e-01 7.58646607e-01 -1.21491289e+00 1.06392443e+00
-2.63602436e-01 5.52912235e-01 3.45509320e-01 -1.18953608e-01
1.57178834e-01 2.65823215e-01 1.49945781e-01 1.40980768e+00
8.94281343e-02 -5.68513237e-02 2.49510825e-01 1.38341784e+00
-4.91598696e-01 1.30364746e-01 -6.32913172e-01 -3.48018467e-01
6.62575841e-01 8.65467727e-01 -3.75128955e-01 -5.59715807e-01
-1.06612170e+00 8.09418082e-01 5.62065661e-01 4.65014726e-01
-8.20491076e-01 -8.65761995e-01 5.55908740e-01 6.73568249e-02
-2.52587162e-02 -9.30045247e-02 -5.00050634e-02 -1.41054153e+00
3.22265178e-01 -1.31525362e+00 3.06437939e-01 -1.00298929e+00
-1.20676804e+00 6.18007362e-01 -2.75381446e-01 -1.05366719e+00
-3.44871014e-01 -5.10406673e-01 -2.52763778e-01 7.99831748e-01
-1.51911902e+00 -9.17127550e-01 -4.50470418e-01 2.29866326e-01
7.24902689e-01 -2.88255900e-01 8.87857437e-01 7.51706183e-01
-8.17091540e-02 7.19715714e-01 -4.78431284e-02 5.29075921e-01
7.36267090e-01 -1.56176126e+00 7.03913927e-01 8.36409330e-01
3.05630535e-01 8.39224458e-01 5.31448781e-01 -4.97073293e-01
-1.30858004e+00 -1.23859775e+00 1.26189733e+00 -1.06871903e+00
8.57273817e-01 -4.74678129e-01 -1.20604098e+00 7.54855216e-01
4.06508058e-01 2.09195822e-01 6.58318758e-01 3.38360108e-02
-1.01599753e+00 1.07610926e-01 -9.88798857e-01 1.13340214e-01
1.03208601e+00 -1.39332247e+00 -8.52335453e-01 5.67042351e-01
1.33224571e+00 -4.00959224e-01 -1.15967679e+00 2.16373295e-01
1.34833679e-01 -7.68112302e-01 7.18454361e-01 -7.92438745e-01
6.83946073e-01 -4.74193424e-01 -4.18701679e-01 -9.60750341e-01
-2.59374063e-02 -2.76883125e-01 2.71322548e-01 1.18114603e+00
7.45311379e-01 -3.30927014e-01 7.17112482e-01 7.98091531e-01
-4.92781490e-01 -6.23638988e-01 -5.32329202e-01 -9.46421921e-01
3.81107539e-01 -7.25185633e-01 7.42642701e-01 1.33088279e+00
5.03618240e-01 3.13746601e-01 2.09247336e-01 -4.38494794e-02
4.66524400e-02 1.44007474e-01 6.17517412e-01 -8.32477927e-01
-4.61816728e-01 -2.25485548e-01 -2.07989931e-01 -1.23196399e+00
5.16194046e-01 -1.26712179e+00 2.12770209e-01 -1.24261892e+00
3.97792608e-01 -7.75877774e-01 1.52695239e-01 7.43019640e-01
-2.83662856e-01 2.05048308e-01 1.43573537e-01 1.93650812e-01
-9.11265373e-01 2.10596308e-01 8.83567929e-01 -3.42691481e-01
2.51548469e-01 -1.80057272e-01 -6.68010354e-01 4.40279484e-01
4.70449805e-01 -8.64401102e-01 -6.84039295e-01 -1.16800046e+00
8.51480961e-01 2.99529344e-01 1.96513817e-01 -8.90693843e-01
2.03086630e-01 2.94418603e-01 -4.66044992e-01 6.12764619e-02
-2.56296545e-01 -8.85717034e-01 -1.94219917e-01 5.05827844e-01
-7.46448457e-01 4.92759645e-01 3.75464559e-01 3.80841941e-01
-4.42644924e-01 -8.44776034e-01 4.38985348e-01 -3.54154021e-01
-6.05462790e-01 -2.11691007e-01 -2.92105079e-01 8.64198983e-01
7.35663533e-01 4.49993193e-01 -7.11874962e-01 -1.79426223e-01
-2.34667614e-01 4.41242158e-01 8.25184286e-01 6.74498916e-01
2.00315133e-01 -1.27713430e+00 -6.53565645e-01 -2.13452745e-02
8.97034883e-01 -3.68917972e-01 -4.41967905e-01 3.90806586e-01
-7.83787787e-01 6.36078417e-01 3.67430449e-01 -4.43294227e-01
-1.02773631e+00 6.94325507e-01 4.96351868e-01 -8.82391818e-03
9.06315967e-02 1.13743508e+00 -1.28045067e-01 -8.87872696e-01
1.14316814e-01 -4.77917552e-01 2.57655862e-03 -3.75778556e-01
2.94016749e-01 3.02469492e-01 2.66695201e-01 -5.47311544e-01
-5.97477019e-01 4.78189886e-01 9.00567397e-02 2.01819822e-01
9.98558462e-01 1.97790772e-01 -5.04790783e-01 2.07739890e-01
1.64942873e+00 4.16223496e-01 -8.60310137e-01 -4.39038366e-01
7.83514261e-01 -6.75444961e-01 -3.53646666e-01 -9.79989707e-01
-1.01921928e+00 7.49622703e-01 4.67806518e-01 4.93082374e-01
9.06766832e-01 5.66715002e-01 6.53732538e-01 6.69775784e-01
5.37557781e-01 -5.65295279e-01 4.53071058e-01 4.80481297e-01
7.85075903e-01 -1.51060033e+00 -4.23972994e-01 -5.46654761e-01
-3.27231258e-01 1.06012535e+00 9.38885391e-01 -1.73258819e-02
3.74357015e-01 2.78098762e-01 3.39666158e-01 -6.63479865e-01
-8.19824338e-01 -1.19008757e-01 3.77464771e-01 5.31048238e-01
1.02182782e+00 -1.39266849e-01 -1.54381722e-01 3.39255601e-01
-2.16945633e-02 -6.86213700e-03 5.33350766e-01 1.03029072e+00
-2.38088042e-01 -1.36885488e+00 -7.67066032e-02 4.83979672e-01
-5.69909036e-01 -5.18225670e-01 -3.98597002e-01 5.00358462e-01
3.37629527e-01 1.23110020e+00 5.55423126e-02 -3.40731889e-01
4.08272177e-01 4.91753109e-02 1.94128186e-01 -1.22891963e+00
-7.41614819e-01 -4.96684879e-01 3.35928857e-01 -6.22804999e-01
-4.08625573e-01 -5.36893487e-01 -1.46099138e+00 2.06027344e-01
-5.78363776e-01 5.44578731e-01 4.44608897e-01 8.22609305e-01
5.22445560e-01 4.54718500e-01 4.15627271e-01 3.33555937e-01
-4.21528965e-01 -6.26825035e-01 8.08424726e-02 4.70011413e-01
4.78633285e-01 -2.71491051e-01 -2.89592087e-01 5.12143791e-01] | [7.547796249389648, 8.062687873840332] |
fb32f9d4-6b46-414d-8f89-05b7980982b1 | disentangling-bipartite-and-core-periphery | 1511.08830 | null | http://arxiv.org/abs/1511.08830v1 | http://arxiv.org/pdf/1511.08830v1.pdf | Disentangling bipartite and core-periphery structure in financial networks | A growing number of systems are represented as networks whose architecture
conveys significant information and determines many of their properties.
Examples of network architecture include modular, bipartite, and core-periphery
structures. However inferring the network structure is a non trivial task and
can depend sometimes on the chosen null model. Here we propose a method for
classifying network structures and ranking its nodes in a statistically
well-grounded fashion. The method is based on the use of Belief Propagation for
learning through Entropy Maximization on both the Stochastic Block Model (SBM)
and the degree-corrected Stochastic Block Model (dcSBM). As a specific
application we show how the combined use of the two ensembles -SBM and dcSBM-
allows to disentangle the bipartite and the core-periphery structure in the
case of the e-MID interbank network. Specifically we find that, taking into
account the degree, this interbank network is better described by a bipartite
structure, while using the SBM the core-periphery structure emerges only when
data are aggregated for more than a week. | [] | 2015-11-25 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [-7.42766112e-02 5.92815220e-01 -1.19258896e-01 -2.24880129e-01
4.07768823e-02 -7.61721969e-01 9.89902675e-01 3.26818764e-01
-1.45871369e-02 8.32580924e-01 3.57686281e-01 -5.41862071e-01
-5.98558486e-01 -8.98506403e-01 -3.04982960e-01 -8.77393305e-01
-2.94674695e-01 9.31248128e-01 4.56201822e-01 -4.20130402e-01
1.46989041e-04 4.99209255e-01 -1.06882858e+00 1.58181801e-01
5.55768967e-01 6.50532722e-01 1.03262276e-03 5.79798877e-01
-1.03187650e-01 7.32139111e-01 -4.06064034e-01 -6.96636677e-01
2.13889718e-01 -3.78163576e-01 -9.73838389e-01 -1.92973703e-01
-7.84976184e-02 3.07641149e-01 -3.04111212e-01 1.13143229e+00
1.97769552e-02 -1.23171858e-01 8.27122033e-01 -1.23965144e+00
3.84319201e-02 1.09934556e+00 -3.18730950e-01 1.27429618e-02
3.26533347e-01 -2.54782796e-01 1.63271821e+00 -4.12006438e-01
7.59168565e-01 1.25145757e+00 7.23652959e-01 -3.31820659e-02
-1.89774740e+00 -2.13086963e-01 -9.18200165e-02 3.07023916e-02
-1.21163344e+00 -3.51453096e-01 9.48134124e-01 -8.29600453e-01
6.32832646e-01 2.94518024e-01 7.45796442e-01 8.32823277e-01
7.82061294e-02 2.43140385e-01 1.37800848e+00 -5.08258939e-01
2.79006243e-01 1.67932883e-01 3.34092796e-01 7.18332529e-01
6.42313302e-01 4.05489542e-02 -1.76692858e-01 -3.67490619e-01
6.12400115e-01 -3.66874158e-01 -1.58931538e-01 -6.10059738e-01
-1.08679223e+00 8.28684986e-01 5.08894563e-01 7.93509007e-01
-2.78224915e-01 -2.39318922e-01 3.19364488e-01 4.66145992e-01
4.00881052e-01 6.27838790e-01 -4.81234759e-01 1.29754379e-01
-1.24161565e+00 -1.05005860e-01 1.43448734e+00 4.75719392e-01
1.06966329e+00 -4.47445810e-01 2.77181089e-01 6.15632653e-01
4.28919017e-01 2.31478363e-02 2.00966578e-02 -8.54587674e-01
2.82271475e-01 8.64585459e-01 -1.26797423e-01 -1.04981756e+00
-6.20701134e-01 -4.89287734e-01 -1.34563279e+00 1.44933745e-01
7.92733252e-01 -1.35309160e-01 -3.43674064e-01 1.78466928e+00
1.94375724e-01 -1.91444099e-01 -1.75189912e-01 5.77137947e-01
5.74663579e-01 3.20886195e-01 -1.82373270e-01 -2.86077619e-01
1.19199526e+00 -5.54186702e-01 -4.07264590e-01 1.61879376e-01
5.86906075e-01 -4.27794069e-01 1.90329552e-01 4.46577311e-01
-1.08164501e+00 -3.00699651e-01 -1.14098287e+00 3.95291805e-01
-3.04430395e-01 -4.46728803e-02 4.47693050e-01 7.83264160e-01
-1.37476504e+00 8.37938905e-01 -6.98786020e-01 -3.80499810e-01
6.66976534e-03 3.65123689e-01 -7.91161299e-01 1.82119817e-01
-1.25361252e+00 1.08866346e+00 5.87907016e-01 2.07661390e-01
-4.88243282e-01 -1.44397020e-01 -6.70008540e-01 2.06205830e-01
2.65434776e-02 -8.27402771e-01 7.30984211e-01 -9.17643487e-01
-1.40862477e+00 7.86039352e-01 7.28600994e-02 -3.27831864e-01
4.27183062e-01 7.49764681e-01 -3.27124596e-01 2.11528972e-01
-2.35319421e-01 1.75072759e-01 6.20453179e-01 -1.35179675e+00
-6.57035504e-03 -4.03465420e-01 2.39055321e-01 -2.75628865e-01
-6.43757060e-02 -2.14440197e-01 1.50853008e-01 -3.05310130e-01
5.94053268e-01 -9.36236560e-01 -2.13339463e-01 -5.54769754e-01
-8.03959191e-01 -1.94595039e-01 3.16815317e-01 -6.31845415e-01
1.56818795e+00 -1.77326560e+00 5.79834223e-01 9.92447972e-01
6.41554415e-01 -9.64403674e-02 1.27929211e-01 9.63691890e-01
-4.34381396e-01 2.60091811e-01 -3.47971439e-01 -1.43770382e-01
2.75674671e-01 2.95528233e-01 2.54972219e-01 5.87230504e-01
-3.72748636e-02 5.21472096e-01 -7.40671098e-01 -5.09606421e-01
6.70015588e-02 1.58472329e-01 -5.16807139e-01 1.92643870e-02
-3.13896053e-02 3.59452397e-01 -1.69301808e-01 1.88706726e-01
5.52097917e-01 -4.93787408e-01 1.09291267e+00 -3.58623862e-01
-3.39252725e-02 3.54516655e-01 -1.28311408e+00 1.23766196e+00
-1.51385367e-01 7.68437743e-01 5.10976613e-01 -1.40649259e+00
9.75015163e-01 3.74187648e-01 4.46066052e-01 -2.10902736e-01
1.21810094e-01 3.07169139e-01 7.42065072e-01 -1.42504767e-01
2.20059201e-01 -3.61891836e-01 3.39467004e-02 5.98463356e-01
3.95153999e-01 1.93247199e-01 5.74332952e-01 5.80950260e-01
1.40054691e+00 -3.73791844e-01 4.52217638e-01 -7.85508871e-01
5.62520981e-01 -3.58257473e-01 4.40562129e-01 6.43242121e-01
-1.14117237e-02 3.44905615e-01 1.37502706e+00 -1.20031081e-01
-1.13807178e+00 -1.13459504e+00 -2.42550299e-01 6.82443619e-01
-2.03081578e-01 -4.66912091e-01 -6.87903762e-01 -4.94798869e-01
9.88185704e-02 3.73096704e-01 -6.02451861e-01 -2.54453391e-01
-2.84380794e-01 -7.83058107e-01 4.12358999e-01 -5.01369499e-02
1.28826886e-01 -7.78707862e-01 -9.68524441e-02 -2.55104899e-02
-1.98473766e-01 -8.99488509e-01 6.04998460e-03 3.25468570e-01
-1.01082206e+00 -1.23167694e+00 -4.55130011e-01 -4.96475816e-01
6.14131927e-01 -3.01563025e-01 1.45444000e+00 1.17467836e-01
1.45879880e-01 8.59492570e-02 -2.99359828e-01 2.98649102e-01
-8.52385044e-01 3.11959952e-01 3.14230286e-02 2.18718201e-01
-6.83249906e-03 -9.74656105e-01 -4.56131935e-01 3.37091446e-01
-7.06533194e-01 7.33419731e-02 7.20956028e-01 7.30427921e-01
-1.50689408e-01 2.86881208e-01 4.82624948e-01 -1.03414989e+00
6.07640386e-01 -7.30296135e-01 -4.11251813e-01 3.67959648e-01
-7.71487653e-01 3.49509031e-01 2.65163302e-01 5.58966137e-02
-7.48278320e-01 -1.26966834e-01 -2.27988437e-02 2.02908561e-01
-2.20911689e-02 9.00604248e-01 -3.67859483e-01 -5.17117325e-04
4.54126030e-01 -1.45594971e-02 2.26479605e-01 -5.34939051e-01
3.76530766e-01 5.06869256e-01 7.51255080e-02 -6.83507144e-01
5.17606080e-01 3.84978086e-01 3.27848881e-01 -9.94064450e-01
-3.20498645e-01 -4.27462906e-01 -1.01215947e+00 -4.02082086e-01
6.72406673e-01 -5.25134504e-01 -7.62641311e-01 1.90704763e-01
-1.05992949e+00 -2.46629536e-01 -1.23516716e-01 4.74736959e-01
-4.04212117e-01 5.65728486e-01 -7.74674535e-01 -9.30967629e-01
2.39182357e-02 -1.01767409e+00 6.54960871e-01 -3.94732475e-01
-3.72513622e-01 -1.45421636e+00 5.13758123e-01 2.41792455e-01
2.44097114e-01 3.27979714e-01 1.37615037e+00 -8.07864547e-01
-3.62807155e-01 -2.40981832e-01 -2.23045275e-01 1.95128843e-01
-9.16169286e-02 2.54491121e-01 -6.25036240e-01 -3.60936075e-01
-2.48851836e-01 1.89165547e-01 1.00037122e+00 3.75606984e-01
3.75955671e-01 -2.77950257e-01 -1.98371574e-01 8.13915879e-02
1.34707952e+00 -2.57785916e-01 5.81997454e-01 4.03033905e-02
5.00728726e-01 9.36032951e-01 -2.28751510e-01 3.11976612e-01
4.81093705e-01 5.79864085e-01 5.07318854e-01 1.45265684e-01
1.74007669e-01 -1.43219888e-01 3.15698206e-01 1.34053648e+00
-2.85083592e-01 5.46738738e-03 -1.00067151e+00 4.45834100e-01
-1.85565305e+00 -1.08040690e+00 -4.47241753e-01 2.28732324e+00
8.82080972e-01 5.21506011e-01 4.41634178e-01 3.16473842e-01
8.76480877e-01 2.25916028e-01 2.01067701e-02 -3.65303546e-01
-2.83490807e-01 2.11996928e-01 2.99389005e-01 7.22975612e-01
-6.47500455e-01 3.03026855e-01 6.56136227e+00 6.83364093e-01
-5.70107460e-01 -2.69258358e-02 5.55978358e-01 2.83726752e-01
-4.17261183e-01 5.32227576e-01 -3.81780237e-01 5.95947623e-01
1.09594548e+00 2.09194943e-01 2.07619548e-01 3.47121745e-01
5.98019958e-02 -3.66044223e-01 -1.03037202e+00 6.34530246e-01
-4.30271417e-01 -1.21667433e+00 -2.29630083e-01 4.18397009e-01
6.12642348e-01 2.27813497e-02 -4.67255831e-01 -2.95556467e-02
6.98492050e-01 -8.69136512e-01 7.04137921e-01 6.64491951e-01
7.01480567e-01 -3.96549046e-01 7.56754518e-01 5.09715736e-01
-1.15435266e+00 -3.14819976e-03 -1.67458639e-01 -2.09855005e-01
5.06059490e-02 1.23201060e+00 -5.87215900e-01 8.23173702e-01
2.82036960e-01 7.49319613e-01 -3.98527592e-01 8.55282784e-01
-1.71937242e-01 8.15882385e-01 -3.76370370e-01 -5.35373315e-02
6.75126016e-02 -8.37396741e-01 7.77766526e-01 1.01567960e+00
1.46226123e-01 -2.45216683e-01 4.18432197e-03 8.00345778e-01
9.68576148e-02 -2.55603611e-01 -4.65765655e-01 -1.75346449e-01
4.48381245e-01 1.51285493e+00 -9.94546413e-01 -2.47331247e-01
-2.92183727e-01 3.82127523e-01 5.90832233e-01 2.48862252e-01
-2.17764974e-01 -2.73056477e-01 3.65436971e-01 2.86010295e-01
2.67093271e-01 -3.10407817e-01 -1.72243252e-01 -1.13736427e+00
-2.18921825e-01 -7.09399402e-01 3.11926603e-01 -5.26710153e-01
-1.43410194e+00 6.38254285e-01 6.37586489e-02 -8.09094369e-01
-2.59856343e-01 -7.80689120e-01 -6.19777918e-01 8.53899717e-01
-9.36984539e-01 -8.89987767e-01 1.30827293e-01 4.66901213e-01
-3.79818350e-01 -1.38945311e-01 8.80171657e-01 2.20347852e-01
-8.06519747e-01 3.90342101e-02 3.70644957e-01 1.30452529e-01
3.47926021e-01 -1.72499979e+00 -8.16094726e-02 4.94397759e-01
4.94238883e-01 6.39812946e-01 9.87688184e-01 -5.76732755e-01
-9.27281260e-01 -5.39729714e-01 1.12345207e+00 -4.46072161e-01
1.00442362e+00 -5.92696905e-01 -6.85908437e-01 5.09932816e-01
4.19235438e-01 -2.36596584e-01 7.11768627e-01 6.32915139e-01
-5.60111940e-01 -1.53127566e-01 -8.81539404e-01 3.61774951e-01
7.94350803e-01 -6.89476073e-01 -6.13572598e-01 2.53893942e-01
1.36845052e-01 4.07866597e-01 -1.18314910e+00 1.95937395e-01
5.17427504e-01 -1.59289086e+00 6.05626523e-01 -4.78983313e-01
3.91131282e-01 -1.97862536e-01 -1.86309487e-01 -1.49576962e+00
-6.23714626e-01 -7.59628654e-01 -1.30217046e-01 1.04085660e+00
6.73363805e-01 -7.79550076e-01 7.49715149e-01 1.35102034e-01
2.80252576e-01 -6.58777773e-01 -1.00474799e+00 -5.88787675e-01
5.05016558e-02 -4.77159545e-02 3.19581598e-01 1.08168578e+00
4.99500841e-01 6.92330658e-01 -8.28298181e-02 -2.47888654e-01
5.37440598e-01 1.49324238e-01 4.09532696e-01 -1.84356558e+00
-5.20588934e-01 -7.59840310e-01 -6.29806280e-01 -8.63480985e-01
1.83513492e-01 -1.08735502e+00 -1.58588737e-01 -1.36521578e+00
2.76713878e-01 -4.72881198e-01 -2.41040319e-01 -1.27193525e-01
3.80106330e-01 -8.28519389e-02 -1.33753363e-02 3.64866853e-01
-5.30104637e-01 1.51637420e-01 8.56251240e-01 2.14627646e-02
1.89367324e-01 1.49233863e-01 -5.32658100e-01 6.23127997e-01
7.42642224e-01 -3.29902321e-01 -1.66049898e-01 4.87295352e-02
9.08644080e-01 2.74428248e-01 3.37377995e-01 -7.33111382e-01
4.04792309e-01 4.52670157e-01 3.25782485e-02 -4.14849818e-01
3.25933576e-01 -8.48563135e-01 4.77694422e-01 5.88884950e-01
-2.83799887e-01 3.55430320e-02 -3.33311468e-01 8.09848666e-01
-3.84404898e-01 -4.80919659e-01 5.73846281e-01 -2.70248950e-01
-4.90243100e-02 -2.86768768e-02 -3.80089790e-01 -9.01204050e-02
5.76416135e-01 -1.86176628e-01 -2.60565132e-01 -5.74673951e-01
-1.18458474e+00 -1.72428396e-02 3.21171433e-01 -1.56712949e-01
1.67129546e-01 -1.20664096e+00 -7.88443029e-01 1.68724716e-01
-2.97926039e-01 -4.75990295e-01 3.91966961e-02 1.37902820e+00
-4.69145268e-01 3.85921925e-01 -1.57767817e-01 -5.65337479e-01
-1.14176536e+00 1.88290969e-01 4.00082827e-01 -8.47055614e-01
-1.56811252e-01 6.20710373e-01 -7.72249103e-02 -5.47964394e-01
-1.33871704e-01 -1.13563687e-01 -3.14424783e-01 4.91040528e-01
-2.73081630e-01 5.44106364e-01 2.46858131e-03 -8.92877758e-01
-2.39651069e-01 2.08929911e-01 1.56017497e-01 -2.55437523e-01
1.21763945e+00 -2.06210762e-01 -9.33156013e-01 6.00239277e-01
1.14294386e+00 4.08294164e-02 -7.62036920e-01 -2.63104022e-01
5.09974957e-01 1.35489091e-01 -1.19299538e-01 -4.82912123e-01
-9.95037854e-01 8.19227695e-01 -7.31431022e-02 1.28742886e+00
7.72775233e-01 1.54908061e-01 1.31080359e-01 2.29065329e-01
3.71777534e-01 -8.77085149e-01 -2.17151657e-01 5.55807412e-01
7.21721172e-01 -9.66269553e-01 1.37149304e-01 -3.67992938e-01
-1.95198536e-01 1.26121318e+00 -6.64666817e-02 -4.04586852e-01
1.34734273e+00 -4.03276645e-02 -4.34875995e-01 -4.22684014e-01
-8.31475973e-01 -1.80956259e-01 3.73058885e-01 2.17612192e-01
4.66110766e-01 4.28001255e-01 -2.91806221e-01 5.40477037e-01
-3.77650857e-01 -6.96878314e-01 5.26538551e-01 4.28547174e-01
-3.66425693e-01 -1.30016458e+00 -1.25784755e-01 7.10518897e-01
-2.82970786e-01 -1.28153786e-01 -8.15605283e-01 8.81228447e-01
-1.44713847e-02 1.01512098e+00 1.27098337e-01 -6.08484685e-01
5.28721660e-02 1.85600668e-01 6.63862765e-01 -5.06033301e-01
-6.60368502e-01 1.13762394e-01 5.19058347e-01 -2.12691098e-01
-6.55196488e-01 -6.23513997e-01 -4.84184474e-01 -6.27684653e-01
-6.52416706e-01 5.51968575e-01 5.73573291e-01 1.01169074e+00
2.38366097e-01 4.06875581e-01 5.06663263e-01 -8.78397644e-01
-5.66207826e-01 -1.25919557e+00 -1.08675432e+00 3.52747351e-01
3.06019485e-01 -7.31251180e-01 -6.11037612e-01 -2.42455855e-01] | [6.984835147857666, 5.295896530151367] |
4bb13756-3029-43fb-83af-6e61cb3b25aa | learning-deep-temporal-representations-for | 1412.7522 | null | http://arxiv.org/abs/1412.7522v4 | http://arxiv.org/pdf/1412.7522v4.pdf | Learning Deep Temporal Representations for Brain Decoding | Functional magnetic resonance imaging produces high dimensional data, with a
less then ideal number of labelled samples for brain decoding tasks (predicting
brain states). In this study, we propose a new deep temporal convolutional
neural network architecture with spatial pooling for brain decoding which aims
to reduce dimensionality of feature space along with improved classification
performance. Temporal representations (filters) for each layer of the
convolutional model are learned by leveraging unlabelled fMRI data in an
unsupervised fashion with regularized autoencoders. Learned temporal
representations in multiple levels capture the regularities in the temporal
domain and are observed to be a rich bank of activation patterns which also
exhibit similarities to the actual hemodynamic responses. Further, spatial
pooling layers in the convolutional architecture reduce the dimensionality
without losing excessive information. By employing the proposed temporal
convolutional architecture with spatial pooling, raw input fMRI data is mapped
to a non-linear, highly-expressive and low-dimensional feature space where the
final classification is conducted. In addition, we propose a simple heuristic
approach for hyper-parameter tuning when no validation data is available.
Proposed method is tested on a ten class recognition memory experiment with
nine subjects. The results support the efficiency and potential of the proposed
model, compared to the baseline multi-voxel pattern analysis techniques. | ['Fatos T. Yarman Vural', 'Emre Aksan', 'Orhan Firat', 'Ilke Oztekin'] | 2014-12-23 | null | null | null | null | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 3.64229143e-01 -5.95272109e-02 1.58125497e-02 -5.23754478e-01
-3.05219274e-02 -3.25416297e-01 5.73107541e-01 -3.09355985e-02
-6.09853864e-01 7.92360842e-01 4.87174332e-01 -2.41587609e-02
-4.70953256e-01 -5.39732158e-01 -4.23473209e-01 -7.84219146e-01
-6.78543687e-01 -3.46266888e-02 3.61562520e-01 1.90225929e-01
4.15490001e-01 6.55345798e-01 -1.43522310e+00 7.67735362e-01
6.44455791e-01 1.18034637e+00 2.76977390e-01 4.53963429e-01
2.67079413e-01 6.61712945e-01 -8.82300436e-02 2.36795872e-01
2.72255987e-01 -3.56724292e-01 -7.62454748e-01 -1.63702637e-01
-9.97947454e-02 -2.31504798e-01 -4.02953416e-01 6.45132899e-01
4.89816785e-01 4.80734229e-01 8.67761254e-01 -5.57578206e-01
-5.34276664e-01 2.82547802e-01 -4.07759100e-01 1.04883337e+00
-1.07326634e-01 3.54602337e-01 4.42559451e-01 -9.48409557e-01
4.56028283e-01 1.03352201e+00 4.60897297e-01 3.40839088e-01
-1.31558120e+00 -6.66962802e-01 -9.51950401e-02 3.19141299e-01
-1.10374856e+00 -5.29930055e-01 5.66548228e-01 -7.40441203e-01
1.40531838e+00 3.13243642e-02 8.20895314e-01 1.10765862e+00
7.61818886e-01 1.16983652e-01 1.71186793e+00 -4.29203082e-03
1.83149099e-01 6.80600200e-03 4.93444264e-01 6.10039175e-01
-5.67068197e-02 5.71728945e-01 -6.35744154e-01 1.06940744e-02
9.10340607e-01 5.15331142e-02 -1.16784453e-01 -3.91585901e-02
-1.44397175e+00 7.06876457e-01 7.17985690e-01 7.13578880e-01
-9.98678088e-01 1.66765064e-01 6.44340217e-01 1.28744408e-01
2.76476562e-01 2.99546033e-01 -4.44032729e-01 1.94506600e-01
-1.03496075e+00 -1.57086626e-01 1.69445127e-01 2.38778904e-01
4.59813684e-01 3.79725993e-01 -5.47643542e-01 6.33093715e-01
2.36988291e-01 3.46705355e-02 1.29769695e+00 -6.43456042e-01
1.77328408e-01 4.72550869e-01 -2.20837086e-01 -1.05063009e+00
-9.67907608e-01 -5.48537731e-01 -1.01081955e+00 1.61758915e-01
9.39628333e-02 -2.08217591e-01 -1.03901148e+00 1.72252214e+00
2.96930429e-02 2.75287092e-01 1.89181075e-01 1.02437854e+00
4.84961182e-01 6.10737503e-01 4.18648630e-01 -4.87195283e-01
1.61529708e+00 -5.76284468e-01 -7.17495799e-01 -2.16248855e-01
3.72547239e-01 -2.36959115e-01 5.76786220e-01 2.77756095e-01
-9.68869328e-01 -6.24306202e-01 -1.14715064e+00 6.97473064e-02
-4.56149787e-01 1.30735531e-01 9.48987067e-01 5.27148247e-01
-9.73226786e-01 8.46945524e-01 -1.03080785e+00 -3.71839374e-01
6.53362691e-01 7.62461126e-01 -5.84169149e-01 3.12746882e-01
-1.22109497e+00 1.01628613e+00 9.31237817e-01 6.35242403e-01
-9.70231414e-01 -6.08706713e-01 -4.57742959e-01 1.13791585e-01
-2.44941354e-01 -2.90486544e-01 7.12667346e-01 -9.78218973e-01
-1.57093847e+00 5.35486460e-01 -2.18591630e-01 -5.67433774e-01
-5.53657524e-02 1.23014756e-01 -4.29637134e-01 3.61286700e-01
-1.95615411e-01 7.02712774e-01 8.56939316e-01 -5.41669726e-01
-7.82177746e-02 -6.27939880e-01 -4.12799925e-01 1.02876015e-01
-4.56695020e-01 -1.83978751e-02 3.93625349e-01 -5.39955318e-01
4.36640173e-01 -6.76848650e-01 -2.76841581e-01 -3.21862817e-01
5.54789742e-03 5.19957654e-02 5.49804270e-01 -7.80602694e-01
1.11707807e+00 -2.09999418e+00 2.13970706e-01 4.04625624e-01
3.63169223e-01 -3.67590459e-03 3.68051715e-02 1.21868409e-01
-5.72283506e-01 3.76024246e-02 -2.72121549e-01 4.84982789e-01
-3.54493558e-01 7.22465813e-02 -4.02428620e-02 7.11776495e-01
3.31445515e-01 9.34671938e-01 -6.19545519e-01 -2.34652758e-01
1.21455915e-01 4.29373145e-01 -5.40440023e-01 6.64828494e-02
1.58489481e-01 8.80218625e-01 -5.06708920e-01 2.83653021e-01
4.76403564e-01 -1.10459477e-01 2.30643108e-01 -4.87533867e-01
-3.67866218e-01 1.65512115e-01 -8.65578294e-01 1.71189666e+00
-1.55644968e-01 5.91147423e-01 -4.85532641e-01 -1.46165061e+00
9.10914838e-01 7.15974689e-01 4.70345467e-01 -1.20288360e+00
4.69910234e-01 2.17714548e-01 6.73229277e-01 -8.13160837e-01
-1.51128694e-01 -4.86148685e-01 1.72881186e-01 5.36016226e-01
5.15210211e-01 6.36691093e-01 -1.35836825e-01 -4.05968547e-01
1.08956110e+00 9.99304429e-02 2.48167321e-01 -6.59949303e-01
5.08234262e-01 -3.96945357e-01 4.49988633e-01 4.16815341e-01
-4.02412355e-01 5.02859764e-02 6.00803673e-01 -6.89282238e-01
-1.18811536e+00 -8.82656217e-01 -3.97976726e-01 7.78369009e-01
-4.91309375e-01 2.35214323e-01 -5.62322676e-01 -9.09845978e-02
-3.30954641e-01 3.00269485e-01 -1.16297185e+00 -3.79694313e-01
-6.84019744e-01 -1.13314378e+00 4.79551524e-01 6.17502868e-01
5.96085966e-01 -1.42742288e+00 -1.42305875e+00 4.80515718e-01
3.07797313e-01 -1.00935459e+00 -1.15808927e-01 7.17753232e-01
-1.48899651e+00 -8.70495260e-01 -4.91374373e-01 -7.11304069e-01
5.41119099e-01 -2.36326367e-01 4.19840068e-01 -2.50145435e-01
-5.66583395e-01 -1.07544266e-01 -1.97069332e-01 -5.13110533e-02
3.13516736e-01 -1.07650153e-01 1.93340898e-01 4.09209520e-01
4.83850509e-01 -1.03123081e+00 -1.12406540e+00 1.46699315e-02
-7.88599670e-01 5.85334189e-02 9.12213445e-01 1.03609741e+00
3.91672492e-01 -1.60276309e-01 9.17830586e-01 -6.27247155e-01
7.39620686e-01 -7.97649860e-01 -3.12410951e-01 -6.96291551e-02
-4.49887037e-01 5.28921366e-01 5.79070866e-01 -6.81689322e-01
-1.19931042e+00 -7.25159198e-02 3.25086862e-01 -2.34310880e-01
-1.98266834e-01 7.80426204e-01 5.68449259e-01 -7.07029626e-02
8.53828967e-01 4.83132720e-01 -7.92467669e-02 -3.25014710e-01
1.13297768e-01 2.37384647e-01 3.79188597e-01 -4.87361133e-01
2.52043724e-01 2.84905910e-01 1.80287331e-01 -6.69745982e-01
-3.96291107e-01 -1.85098231e-01 -1.16656244e+00 -2.33415619e-01
1.13070178e+00 -6.42742872e-01 -8.42599511e-01 1.26412138e-01
-1.02434123e+00 -1.93639010e-01 5.33228591e-02 9.36375022e-01
-5.46247602e-01 -1.39582390e-02 -6.05149567e-01 -8.51386428e-01
-4.08851564e-01 -9.54336345e-01 4.25922722e-01 1.13957658e-01
-7.14882016e-02 -7.52830327e-01 -8.34699534e-03 -1.74762025e-01
5.87421656e-01 4.89512175e-01 1.27972996e+00 -6.93655372e-01
-4.30423647e-01 -1.53468981e-01 -2.58768201e-01 9.12691876e-02
-1.73370272e-01 -5.08192003e-01 -1.19013000e+00 -2.92874333e-02
3.50205570e-01 -3.21399897e-01 8.86538923e-01 5.54250479e-01
1.27564561e+00 -3.05161625e-01 -1.22040607e-01 4.71186996e-01
1.42958164e+00 5.89601696e-01 7.26241469e-01 9.82894301e-02
2.78741866e-01 8.08933437e-01 -1.55948415e-01 4.46751356e-01
-1.79675728e-01 2.41814882e-01 1.22949570e-01 1.84167996e-01
5.22770844e-02 2.22702563e-01 1.62925318e-01 1.00099480e+00
-3.52420151e-01 3.87599111e-01 -9.15928304e-01 5.60407043e-01
-1.69635546e+00 -1.12239635e+00 -1.45208150e-01 2.21814895e+00
5.87805510e-01 8.07791203e-02 -4.88335937e-02 1.41529009e-01
4.17859703e-01 3.60246375e-02 -6.06415033e-01 -7.12419689e-01
-5.47481105e-02 7.83351004e-01 3.96334440e-01 1.54441580e-01
-8.19551706e-01 8.50905299e-01 6.40656042e+00 2.03295916e-01
-1.41769218e+00 4.76046294e-01 5.09336174e-01 -1.86478794e-01
2.18694508e-01 -1.96413577e-01 -1.04069412e-01 2.09855288e-01
1.63016069e+00 -1.69351593e-01 5.51355064e-01 2.28523910e-01
6.52327478e-01 -1.56819820e-01 -9.61506844e-01 9.16234553e-01
-2.48203620e-01 -1.28300488e+00 -1.72641233e-01 3.05650812e-02
5.03381968e-01 -4.11549024e-03 2.66045868e-01 1.70997307e-01
-3.66652697e-01 -1.49051297e+00 3.92116517e-01 1.06183779e+00
6.56868637e-01 -6.24668419e-01 5.98942459e-01 2.44708195e-01
-1.00762665e+00 -5.99202454e-01 -4.06108975e-01 -2.66127110e-01
-1.42498359e-01 1.94459870e-01 -6.32656693e-01 1.82032600e-01
7.07161367e-01 6.51549995e-01 -4.48666215e-01 1.09579599e+00
3.05312127e-01 5.26440859e-01 -1.58920601e-01 -1.38493240e-01
4.71348405e-01 1.73673898e-01 1.34354874e-01 1.30866325e+00
3.18335384e-01 5.86612761e-01 -1.34313941e-01 1.02412295e+00
1.27854556e-01 3.11112642e-01 -6.52218223e-01 -3.47014666e-01
1.28225759e-01 1.18661737e+00 -8.95800233e-01 -3.02285165e-01
-3.08307499e-01 9.18165386e-01 3.49000573e-01 4.88955498e-01
-4.82512116e-01 -1.05343238e-01 1.48041397e-01 1.51391821e-02
1.85404897e-01 -3.55649799e-01 -3.85298699e-01 -9.73015368e-01
-9.87244099e-02 -4.38655406e-01 2.25214392e-01 -6.21294618e-01
-1.03486061e+00 9.35066581e-01 2.00662225e-01 -8.31187665e-01
-2.09461004e-01 -8.34503531e-01 -4.93979782e-01 1.33061182e+00
-1.34436631e+00 -8.16836238e-01 -1.12640355e-02 9.08000886e-01
3.94541174e-01 -1.74723163e-01 1.11434591e+00 3.18825126e-01
-5.97849846e-01 1.56326041e-01 -4.62828390e-02 -4.60878760e-02
2.93727905e-01 -8.04870188e-01 -2.82555372e-01 8.42545986e-01
-3.94493043e-01 9.44955409e-01 3.38036954e-01 -5.83310127e-01
-1.24023044e+00 -9.05869782e-01 5.59530318e-01 6.19655624e-02
6.88838840e-01 -4.12080973e-01 -1.13969755e+00 5.00007868e-01
1.87872663e-01 3.03375363e-01 8.09625566e-01 -1.46317616e-01
-2.03411594e-01 -1.07549010e-02 -1.23790836e+00 2.54689485e-01
6.50791943e-01 -5.54108262e-01 -8.11535060e-01 4.14404809e-01
2.03269705e-01 -1.55578451e-02 -1.08441389e+00 1.50582507e-01
8.87531936e-01 -8.26600790e-01 6.84660316e-01 -9.33149934e-01
2.79094219e-01 -5.96870258e-02 -1.42911077e-01 -1.12614977e+00
-7.30382919e-01 -2.52315611e-01 -1.35958418e-01 4.33934718e-01
4.20554698e-01 -5.72551370e-01 6.26377523e-01 6.49486780e-01
-1.84418857e-01 -7.44407117e-01 -1.11847806e+00 -4.96950358e-01
1.41336218e-01 -2.65347123e-01 1.51627913e-01 7.90538132e-01
2.18026757e-01 2.59713054e-01 -4.08356220e-01 1.19866811e-01
6.01957500e-01 -3.45875859e-01 -2.97612518e-01 -1.25059223e+00
-1.25080302e-01 -3.37795734e-01 -6.45222247e-01 -3.66926163e-01
1.21871434e-01 -1.29750204e+00 -1.32912546e-01 -1.08182561e+00
3.57073069e-01 -1.30471423e-01 -8.14023972e-01 5.62803209e-01
3.19021761e-01 2.51959890e-01 -1.54050350e-01 1.44673005e-01
4.82558571e-02 4.44319576e-01 1.22834074e+00 1.33642644e-01
-2.92807490e-01 -4.31493193e-01 -3.02626163e-01 2.35309854e-01
8.92645478e-01 -3.88118505e-01 -5.54851353e-01 -4.42497015e-01
-2.24377811e-01 4.13823605e-01 5.67585409e-01 -1.31873417e+00
1.81037679e-01 7.01825246e-02 1.16905832e+00 -2.07418650e-01
3.42794955e-01 -7.32381284e-01 2.08718717e-01 6.71818614e-01
-4.41382825e-01 2.32958719e-01 3.73708218e-01 4.02715325e-01
-2.05819346e-02 -3.70835960e-02 9.60949421e-01 -4.10364181e-01
-7.39668429e-01 4.11455661e-01 -6.25231266e-01 -4.62750703e-01
9.77138758e-01 -6.01975739e-01 -8.10441598e-02 4.28887725e-01
-1.28995311e+00 -1.42520502e-01 -5.36544979e-01 2.78620481e-01
7.92116106e-01 -1.43118286e+00 -4.73600566e-01 4.47344214e-01
-2.41003454e-01 -7.75965989e-01 8.08702946e-01 1.32093620e+00
-2.93717802e-01 8.46447229e-01 -1.08409214e+00 -4.95076478e-01
-5.14114499e-01 6.93298757e-01 6.56118929e-01 -8.40206370e-02
-9.52454865e-01 5.42314410e-01 2.19853476e-01 2.84427013e-02
-1.54604182e-01 -4.36645836e-01 -6.33128703e-01 3.25443923e-01
5.42162359e-01 3.37219462e-02 6.71996176e-02 -6.49998426e-01
-4.14828509e-01 2.51487225e-01 -5.28508984e-02 -2.99658477e-01
1.82340086e+00 1.26895145e-01 -3.25316161e-01 6.94153845e-01
1.29269481e+00 -8.35430264e-01 -1.13513124e+00 -1.63539350e-01
1.77775353e-01 -1.60418436e-01 4.13224936e-01 -8.87315691e-01
-1.06320298e+00 1.20365119e+00 1.30413043e+00 -2.23268121e-01
1.18564165e+00 -5.01016378e-01 3.43154550e-01 3.46248895e-01
9.78208855e-02 -1.11756659e+00 2.55716145e-01 4.40967828e-01
1.04397118e+00 -9.55852211e-01 -2.26647198e-01 2.42303431e-01
-5.73357701e-01 1.55150092e+00 6.05782866e-01 -5.88465512e-01
9.87282455e-01 -3.52873467e-02 -4.12077665e-01 -6.93747044e-01
-9.14672911e-01 1.25978231e-01 3.78087312e-01 4.59762454e-01
6.88145876e-01 9.50156376e-02 -6.28214002e-01 9.16565955e-01
2.68954467e-02 2.29842961e-01 3.36062729e-01 7.96396911e-01
-6.63147271e-01 -5.78146577e-01 -2.49364719e-01 8.57930481e-01
-4.66662437e-01 -1.78845733e-01 2.49159709e-01 6.10021710e-01
1.26348972e-01 5.65762639e-01 3.69428359e-02 -3.02767307e-01
2.89483577e-01 6.84918106e-01 6.39111340e-01 -5.28819263e-01
-7.61373937e-01 2.28961825e-01 -2.00064018e-01 -7.04241395e-01
-3.33042264e-01 -6.79542005e-01 -1.29350233e+00 1.52556360e-01
1.25868410e-01 -4.95938584e-02 5.97129822e-01 1.10559797e+00
2.59682000e-01 9.31372344e-01 5.01322925e-01 -9.12882745e-01
-3.01660329e-01 -1.30336785e+00 -5.08644521e-01 1.79217368e-01
4.01161373e-01 -1.02804792e+00 8.41593891e-02 7.25560039e-02] | [12.658693313598633, 3.3963143825531006] |
eb6f603c-e60f-45bf-b49e-0ba2d8110b19 | finding-dataset-shortcuts-with-grammar | 2210.11560 | null | https://arxiv.org/abs/2210.11560v1 | https://arxiv.org/pdf/2210.11560v1.pdf | Finding Dataset Shortcuts with Grammar Induction | Many NLP datasets have been found to contain shortcuts: simple decision rules that achieve surprisingly high accuracy. However, it is difficult to discover shortcuts automatically. Prior work on automatic shortcut detection has focused on enumerating features like unigrams or bigrams, which can find only low-level shortcuts, or relied on post-hoc model interpretability methods like saliency maps, which reveal qualitative patterns without a clear statistical interpretation. In this work, we propose to use probabilistic grammars to characterize and discover shortcuts in NLP datasets. Specifically, we use a context-free grammar to model patterns in sentence classification datasets and use a synchronous context-free grammar to model datasets involving sentence pairs. The resulting grammars reveal interesting shortcut features in a number of datasets, including both simple and high-level features, and automatically identify groups of test examples on which conventional classifiers fail. Finally, we show that the features we discover can be used to generate diagnostic contrast examples and incorporated into standard robust optimization methods to improve worst-group accuracy. | ['Danqi Chen', 'Alexander Wettig', 'Dan Friedman'] | 2022-10-20 | null | null | null | null | ['sentence-classification'] | ['natural-language-processing'] | [ 8.04034531e-01 6.14877880e-01 -2.36260608e-01 -8.58557403e-01
-1.01691651e+00 -6.38656676e-01 6.05031013e-01 6.73043430e-01
-8.75447392e-02 6.84797406e-01 3.02955925e-01 -3.89305919e-01
-5.95148861e-01 -4.44083303e-01 -4.47870016e-01 -5.46453297e-01
-1.00074194e-01 5.82382321e-01 2.29225591e-01 -1.81718647e-01
6.69140637e-01 1.54952064e-01 -1.85777617e+00 7.46285081e-01
9.75900471e-01 6.21395469e-01 2.58284032e-01 6.64890885e-01
-2.50716597e-01 4.18901205e-01 -7.19929338e-01 -4.67117548e-01
-5.80335297e-02 -7.42856026e-01 -9.03801680e-01 3.83356065e-02
4.23560083e-01 3.61123383e-01 2.18554869e-01 1.25015676e+00
2.91076511e-01 -1.79176480e-01 6.03070080e-01 -1.23846710e+00
-4.80324447e-01 9.10195827e-01 -4.13808942e-01 4.55541939e-01
8.15672219e-01 9.32143927e-02 1.54514325e+00 -5.84644973e-01
8.96842957e-01 1.53270400e+00 5.93291104e-01 5.22136986e-01
-1.67109942e+00 -3.89838308e-01 1.20423540e-01 1.42116278e-01
-8.68135750e-01 -3.30898136e-01 7.80722737e-01 -3.13675672e-01
1.06592751e+00 6.28729641e-01 4.84588623e-01 1.20233774e+00
3.14421952e-01 1.01776719e+00 1.43047190e+00 -8.67695212e-01
1.80355012e-01 1.02489173e-01 6.48929954e-01 8.24860096e-01
1.48581594e-01 2.71514773e-01 -6.99253440e-01 -2.70053983e-01
1.57229230e-01 -1.03244521e-01 -2.59166598e-01 -1.13604762e-01
-1.30640495e+00 8.56715798e-01 2.67308533e-01 5.36131084e-01
-1.34613872e-01 -1.89213127e-01 1.25889942e-01 5.39209545e-01
4.68541294e-01 1.03706515e+00 -6.71988785e-01 -2.07858801e-01
-9.37276304e-01 4.78332907e-01 7.35109448e-01 6.84552968e-01
9.88316953e-01 -7.31257617e-01 -5.46416104e-01 8.93785417e-01
1.03261352e-01 8.67017359e-02 6.35937691e-01 -7.80213535e-01
4.93630737e-01 9.68668401e-01 -2.53720611e-01 -1.19272900e+00
-8.32636058e-01 -2.00505078e-01 -1.88716888e-01 -9.44966227e-02
3.29361260e-01 1.52181968e-01 -1.00358653e+00 1.78830934e+00
5.21217696e-02 -2.32173637e-01 -3.55618484e-02 8.27913702e-01
6.20446265e-01 1.05492540e-01 -4.51227315e-02 -2.29774222e-01
1.42754436e+00 -4.47884977e-01 -5.82531750e-01 -4.18867081e-01
1.26281750e+00 -5.53554237e-01 1.72606409e+00 2.28450343e-01
-7.50318766e-01 -6.53578267e-02 -1.00449216e+00 8.12972784e-02
-3.73930126e-01 -2.38785893e-01 7.46753573e-01 6.16128385e-01
-7.97044694e-01 9.29332256e-01 -5.90055048e-01 -4.28830296e-01
4.18152213e-01 3.06056738e-01 -1.47911564e-01 -8.94855633e-02
-1.19103158e+00 8.16592336e-01 4.90916848e-01 -2.17331111e-01
-5.48582196e-01 -7.72154152e-01 -1.00536633e+00 4.83425036e-02
6.22972965e-01 -5.65558255e-01 1.10093331e+00 -1.19164729e+00
-9.69053209e-01 1.18516099e+00 -2.64606029e-01 -4.14780676e-01
9.99137238e-02 1.88145906e-01 -3.21399033e-01 1.97446570e-01
4.00675803e-01 7.86703885e-01 6.14085078e-01 -8.81804407e-01
-4.42086190e-01 -3.76295567e-01 1.84238702e-02 1.90304786e-01
-1.20248884e-01 3.64745915e-01 9.49348137e-02 -6.68512225e-01
4.03147429e-01 -7.59383142e-01 -4.15494174e-01 -5.06792545e-01
-1.01484013e+00 -3.62156063e-01 5.25585771e-01 -3.10232162e-01
1.09788609e+00 -1.90301931e+00 -2.17927853e-03 3.42413008e-01
4.11561430e-01 -2.57246405e-01 -1.18138239e-01 2.72652745e-01
-2.16788620e-01 4.45596337e-01 -5.58590770e-01 1.71687797e-01
1.73782542e-01 3.91193599e-01 -2.77279556e-01 1.42810851e-01
6.64464116e-01 9.79839087e-01 -1.19276476e+00 -5.80862522e-01
-3.15132439e-02 -3.55787098e-01 -6.41199648e-01 -8.89024362e-02
-4.78209287e-01 -8.51172581e-02 -4.85474110e-01 7.99010992e-01
1.25421897e-01 -2.50354022e-01 2.02290162e-01 2.33634442e-01
1.97307050e-01 9.85272408e-01 -8.18373680e-01 1.56618023e+00
-9.47003216e-02 7.39912152e-01 -4.28712517e-01 -1.08541870e+00
9.34866607e-01 -2.69814059e-02 1.56146944e-01 -5.77282012e-01
9.45519507e-02 3.07544410e-01 3.15229595e-01 -8.11420798e-01
3.87556255e-01 -3.89111072e-01 -2.93135762e-01 5.58696687e-01
5.13286591e-02 -1.77897602e-01 4.26492214e-01 2.67260969e-01
1.54592872e+00 -2.87977546e-01 3.31076950e-01 -5.45381427e-01
6.97278678e-02 2.46940494e-01 6.90450013e-01 1.07156229e+00
-1.30409338e-02 9.21567023e-01 1.28142178e+00 -3.93549293e-01
-6.92180395e-01 -8.63245964e-01 -2.91088670e-01 1.04188180e+00
-4.86846641e-02 -7.37895489e-01 -7.66506910e-01 -1.12753844e+00
-1.45567041e-02 1.02461231e+00 -7.95891821e-01 -3.54894608e-01
-2.05944851e-01 -1.15379524e+00 5.41788518e-01 3.10609579e-01
-1.15811639e-01 -1.50427079e+00 -8.77683938e-01 6.56802952e-02
-3.60851325e-02 -9.37328935e-01 -2.08908081e-01 7.16830194e-01
-8.89845371e-01 -1.45514345e+00 -1.89704463e-01 -8.00717473e-01
8.88760448e-01 -8.21427256e-02 1.40905881e+00 2.20789745e-01
-6.37750804e-01 7.00649619e-02 -4.76700246e-01 -5.80824912e-01
-5.58803380e-01 1.31518304e-01 1.33608505e-02 -2.02081665e-01
7.44902253e-01 -4.52051550e-01 -1.07825555e-01 1.60753563e-01
-6.33626819e-01 2.26971075e-01 7.88732588e-01 1.18191040e+00
5.86355388e-01 -2.68737733e-01 7.41331995e-01 -1.43549228e+00
1.00743175e+00 -3.09664160e-01 -2.89860100e-01 4.10973400e-01
-8.49761665e-01 4.17385399e-01 4.11770374e-01 -3.21615905e-01
-6.09614253e-01 -3.39853531e-03 1.49972931e-01 -1.96472704e-01
-4.53208923e-01 7.52590954e-01 -1.51120290e-01 2.41719007e-01
9.60188687e-01 1.32980883e-01 -2.01021045e-01 -3.89575243e-01
2.39928320e-01 6.01437211e-01 3.91262978e-01 -6.01878703e-01
4.00951952e-01 1.44895585e-02 -1.06312022e-01 -6.93812191e-01
-1.25347614e+00 -2.78164327e-01 -6.45502746e-01 7.23833665e-02
6.23815715e-01 -2.10799158e-01 -3.17795604e-01 -1.37937143e-01
-9.89962578e-01 -3.73527378e-01 -3.89313817e-01 2.61235625e-01
-7.01505244e-01 1.49157673e-01 -3.49384308e-01 -5.62396407e-01
7.56019130e-02 -1.00745392e+00 1.32025027e+00 1.79937884e-01
-6.82518303e-01 -8.36783350e-01 9.68082696e-02 1.87365875e-01
-1.82006150e-01 3.05559486e-01 1.25477719e+00 -1.17232120e+00
-3.36621255e-01 -1.50051937e-01 -1.41670093e-01 -1.04163818e-01
1.60607144e-01 3.35189817e-03 -9.45800900e-01 6.90828785e-02
-1.02266997e-01 -4.42892760e-01 1.02638865e+00 4.17571872e-01
1.48741722e+00 -4.54406440e-01 -6.69918180e-01 4.69560623e-01
9.24394131e-01 -3.96329723e-02 3.28344315e-01 1.38877854e-01
4.14432347e-01 1.02077627e+00 6.47264898e-01 2.06498485e-02
1.68218151e-01 5.87322652e-01 -2.17257533e-02 1.23841815e-01
1.88292399e-01 -3.02803785e-01 2.27651224e-01 5.79237401e-01
4.07108247e-01 8.75778347e-02 -1.20015311e+00 7.03683555e-01
-1.80067122e+00 -8.18743646e-01 -2.00977474e-01 1.87879467e+00
9.66242969e-01 6.73419595e-01 -1.54336005e-01 4.19708312e-01
7.31522918e-01 1.02979265e-01 -1.93615422e-01 -5.27274549e-01
-2.81217366e-01 1.93133086e-01 9.15009305e-02 3.76041025e-01
-1.05936635e+00 8.68650496e-01 6.86822844e+00 5.79957724e-01
-7.78162599e-01 -1.58399060e-01 7.09186912e-01 -2.79364079e-01
-7.69769192e-01 2.60455966e-01 -6.63161993e-01 3.61096650e-01
8.73302877e-01 -2.47858226e-01 1.71678647e-01 6.20475829e-01
2.31435776e-01 -2.00805143e-01 -1.50687158e+00 8.34785044e-01
1.77408218e-01 -1.40713930e+00 -5.40917329e-02 -1.15948208e-01
7.01453447e-01 -1.31351292e-01 -1.63098752e-01 1.32803276e-01
2.68186331e-01 -1.13399112e+00 4.75481212e-01 4.67743844e-01
5.23456812e-01 -6.02797627e-01 6.79030597e-01 3.18202913e-01
-4.99843597e-01 -1.97597355e-01 -4.02477384e-01 -2.97072023e-01
-1.35055348e-01 9.71884310e-01 -1.24160898e+00 2.30193689e-01
5.43014288e-01 7.28979886e-01 -9.62150872e-01 1.08671403e+00
-5.25833011e-01 7.62041688e-01 -2.95635045e-01 -5.83285272e-01
2.64190912e-01 2.31753901e-01 8.28696489e-01 1.43110669e+00
3.00102364e-02 -6.37143292e-03 2.09621608e-01 1.06783831e+00
-1.33170169e-02 1.11863397e-01 -9.71077442e-01 -9.81611386e-02
2.49982759e-01 1.24144876e+00 -1.19231892e+00 -2.33651847e-01
-1.46618918e-01 6.17085636e-01 4.44612831e-01 1.62538245e-01
-4.39863354e-01 -8.05968642e-01 5.34596503e-01 3.10146455e-02
2.62574796e-02 3.44569862e-01 -6.90745533e-01 -1.15956128e+00
2.66271204e-01 -9.72477436e-01 6.57814443e-01 -7.11746216e-01
-1.38254249e+00 3.14343303e-01 9.21143070e-02 -7.51091301e-01
-5.58382213e-01 -6.10517263e-01 -8.27969611e-01 6.30589008e-01
-1.06119382e+00 -6.79233074e-01 4.66754287e-02 2.17127964e-01
6.00581646e-01 -4.41314653e-02 8.16561699e-01 -3.03055495e-01
-5.56011140e-01 5.79083204e-01 -3.09015960e-01 1.12472236e-01
4.47373450e-01 -1.70542359e+00 3.15870672e-01 9.04636800e-01
8.10763240e-01 6.00582838e-01 9.33093071e-01 -6.99609458e-01
-8.49570811e-01 -8.56136262e-01 1.41393209e+00 -7.38828540e-01
5.65206826e-01 -6.42120063e-01 -1.13889849e+00 5.89134157e-01
-1.55129761e-01 -6.08038418e-02 6.52603149e-01 6.72240615e-01
-3.59657228e-01 1.64530352e-01 -1.08701229e+00 6.60962582e-01
1.14697218e+00 -6.57853842e-01 -1.19528532e+00 5.49851537e-01
7.69662023e-01 -2.15352863e-01 -2.15775549e-01 3.42262924e-01
2.45708033e-01 -9.25920427e-01 5.97571909e-01 -1.17314470e+00
6.47890568e-01 5.87442238e-03 -7.58416504e-02 -1.43756175e+00
-1.96385652e-01 -7.53503323e-01 1.84637696e-01 1.13651359e+00
1.02096653e+00 -4.30267364e-01 8.13024104e-01 7.08021045e-01
-3.81403774e-01 -1.00662994e+00 -8.97830665e-01 -6.66692197e-01
-2.63765812e-01 -7.24381924e-01 4.83325809e-01 1.02566803e+00
7.05181181e-01 4.33036089e-01 1.45513818e-01 2.00383589e-02
4.36665446e-01 4.59138244e-01 1.70186073e-01 -1.25913608e+00
-3.02599967e-01 -6.92703187e-01 -5.68711638e-01 -4.94063646e-01
3.61734748e-01 -1.27209544e+00 4.73454207e-01 -1.23878133e+00
4.72585648e-01 -4.37309891e-01 -3.70635122e-01 7.23055840e-01
-4.96518880e-01 -1.13870047e-01 -2.98892409e-02 -1.68311790e-01
-7.16591358e-01 3.46333474e-01 9.14599955e-01 -2.53437430e-01
-4.24484193e-01 -6.25973940e-02 -1.08013654e+00 8.81465316e-01
7.70790339e-01 -9.64667261e-01 -3.74271750e-01 -1.16482377e-01
4.39283490e-01 -1.18766680e-01 5.61137795e-01 -6.40255272e-01
3.32517885e-02 -3.63052934e-01 2.84278393e-01 -3.89958948e-01
-1.82767183e-01 -1.08804241e-01 -3.97840440e-01 3.36683840e-01
-9.95573878e-01 -8.37248378e-03 -5.17682619e-02 4.93709207e-01
-2.01820806e-01 -6.06316864e-01 5.00858247e-01 -1.00779116e-01
-4.37268704e-01 -2.38505736e-01 -3.26704770e-01 4.88626868e-01
7.98283577e-01 3.70231755e-02 -3.73042196e-01 -4.32260707e-02
-6.72798634e-01 2.89240927e-01 3.08647662e-01 5.15287757e-01
7.58500218e-01 -1.03742468e+00 -7.84866273e-01 3.25924575e-01
2.94010878e-01 -1.47968590e-01 -1.58298522e-01 8.59882057e-01
-1.67203546e-01 4.85946506e-01 4.76844162e-02 -7.90850103e-01
-1.31633747e+00 5.60287416e-01 2.89868504e-01 -3.07085156e-01
-6.97115183e-01 9.15422261e-01 2.00116307e-01 -6.62282526e-01
-6.40344545e-02 -7.61945963e-01 -9.05115679e-02 1.89066559e-01
4.75538015e-01 -1.67311996e-01 1.43172100e-01 -1.54297546e-01
-5.96195459e-01 2.47556359e-01 -1.29956454e-01 -4.89109606e-02
1.33411491e+00 1.27292112e-01 -1.53904483e-01 6.92864418e-01
9.71990287e-01 -2.44882509e-01 -8.41074824e-01 4.58501913e-02
8.49455655e-01 -5.81300020e-01 -1.70482993e-01 -9.89298403e-01
-5.86579025e-01 8.71082246e-01 1.64771006e-01 5.57730913e-01
9.90376770e-01 5.16818464e-01 2.98823506e-01 7.03465104e-01
2.10726902e-01 -1.05060494e+00 -1.85452998e-02 4.04117554e-01
8.77018988e-01 -1.34779763e+00 -1.90322563e-01 -5.73420525e-01
-4.53836769e-01 1.09193122e+00 3.89934480e-01 -3.51245925e-02
4.02559996e-01 2.65291721e-01 -8.17746017e-03 -5.32900155e-01
-1.23769403e+00 -3.39995772e-01 4.20184433e-01 6.51143610e-01
4.26381022e-01 2.08997816e-01 -5.15041590e-01 9.88061070e-01
-6.34713829e-01 -2.20372379e-01 6.60546064e-01 7.22290039e-01
-4.69131351e-01 -1.03911710e+00 -1.24299653e-01 1.23255181e+00
-5.29197574e-01 -2.91396350e-01 -7.74984241e-01 5.22285759e-01
9.07745771e-03 8.69399548e-01 7.46540874e-02 -5.98204911e-01
2.09759519e-01 4.13678765e-01 4.13574606e-01 -9.43117738e-01
-5.96734941e-01 -3.34335834e-01 4.19573516e-01 -6.60680115e-01
-2.90284455e-01 -1.03457367e+00 -1.12234795e+00 2.76572764e-01
-3.00401539e-01 1.31712660e-01 3.15680116e-01 1.25480211e+00
3.01090211e-01 4.44995880e-01 4.45589393e-01 -4.42089617e-01
-4.33264345e-01 -9.49777305e-01 -4.62470263e-01 6.18021846e-01
2.49980301e-01 -6.95503235e-01 -4.92244780e-01 -1.53989851e-01] | [10.901677131652832, 8.608402252197266] |
f7950636-13f5-496a-affc-485084545773 | sketching-out-the-details-sketch-based-image | null | null | https://doi.org/10.1016/j.cag.2017.12.006 | https://doi.org/10.1016/j.cag.2017.12.006 | Sketching out the Details: Sketch-based Image Retrieval using Convolutional Neural Networks with Multi-stage Regression | We propose and evaluate several deep network architectures for measuring the similarity between sketches and photographs, within the context of the sketch based image retrieval (SBIR) task. We study the ability of our networks to generalize across diverse object categories from limited training data, and explore in detail strategies for weight sharing, pre-processing, data augmentation and dimensionality reduction. In addition to a detailed comparative study of network configurations, we contribute by describing a hybrid multi-stage training network that exploits both contrastive and triplet networks to exceed state of the art performance on several SBIR benchmarks by a significant margin.
Datasets and models are available at http://www.cvssp.org. | ['J. Collomosse', 'M. Ponti', 'L. Ribeiro', 'T. Bui'] | 2017-12-01 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 2.24209368e-01 -5.33215582e-01 -1.43811092e-01 -6.49557769e-01
-6.53851509e-01 -7.78527379e-01 1.00786483e+00 -2.50290811e-01
-4.54819441e-01 2.34044760e-01 1.49178773e-01 -7.24977404e-02
-6.70602202e-01 -5.40806293e-01 -3.49870771e-01 -2.28156701e-01
5.10623232e-02 5.63377976e-01 -5.57115525e-02 -3.28416288e-01
4.77231771e-01 1.24012351e+00 -1.53692913e+00 6.06006145e-01
9.27358419e-02 1.56081414e+00 -1.79091081e-01 6.45836294e-01
-2.21616447e-01 4.51458216e-01 -5.96958995e-01 -7.24068880e-01
6.17015004e-01 2.39115775e-01 -7.78312743e-01 -1.39699996e-01
1.56715262e+00 -7.87043035e-01 -8.18517983e-01 6.78825200e-01
7.66172409e-01 3.76614243e-01 8.61876130e-01 -1.56427920e+00
-1.29698038e+00 3.74465883e-01 -1.10425793e-01 3.92013073e-01
5.61096594e-02 5.33296280e-02 1.17032015e+00 -1.41522515e+00
8.19045246e-01 1.42902839e+00 6.87432587e-01 7.59196520e-01
-1.26215339e+00 -7.65632212e-01 1.08407281e-01 2.08029166e-01
-1.46923685e+00 -7.10914135e-01 1.06419289e+00 -1.56353310e-01
1.04516733e+00 2.10927323e-01 5.30541182e-01 1.37733626e+00
-3.06375712e-01 1.04226243e+00 6.85623527e-01 -2.63748437e-01
-1.17129274e-01 -6.57209307e-02 2.61443824e-01 5.72870076e-01
1.00837164e-01 1.12044543e-01 -6.58799052e-01 -2.61552632e-01
1.13484442e+00 1.96056977e-01 6.93114921e-02 -7.39722490e-01
-1.16727400e+00 7.17483461e-01 7.45492995e-01 4.94610310e-01
-1.80788428e-01 6.12876356e-01 5.63086390e-01 5.73751092e-01
4.29702133e-01 7.50252187e-01 -4.49424416e-01 2.98299164e-01
-9.74394858e-01 4.44955200e-01 6.22930050e-01 1.27314901e+00
4.28223997e-01 1.05990738e-01 -5.04875600e-01 1.26306438e+00
-4.42276224e-02 3.90784264e-01 8.34501088e-02 -1.18994176e+00
3.30869198e-01 3.47761631e-01 -3.16477492e-02 -1.16696393e+00
-1.64537951e-01 -2.04372659e-01 -8.71683478e-01 2.67481267e-01
1.61263987e-01 3.67652118e-01 -1.12370241e+00 1.54790950e+00
-3.16386908e-01 -1.51136294e-01 -3.24932396e-01 1.09943962e+00
1.54239404e+00 1.89587891e-01 1.94172099e-01 5.81963480e-01
1.14457226e+00 -9.67248619e-01 -4.39421803e-01 -4.08582836e-02
7.94036686e-02 -7.97898233e-01 1.15220273e+00 1.65401369e-01
-1.41808259e+00 -7.17461526e-01 -1.04820442e+00 -6.26419187e-01
-8.37871552e-01 4.62863356e-01 8.31375659e-01 3.14171731e-01
-1.56814659e+00 1.11523211e+00 -2.74580956e-01 -6.66544855e-01
8.65027547e-01 4.01041746e-01 -5.26032448e-01 -2.33539030e-01
-9.98428226e-01 8.99832129e-01 1.02882691e-01 1.06447898e-01
-8.25478613e-01 -8.81109655e-01 -5.83804250e-01 3.21117818e-01
9.15285721e-02 -7.29740977e-01 1.01257420e+00 -9.89595294e-01
-1.22333586e+00 1.05016279e+00 2.05309734e-01 -2.35448912e-01
4.30595785e-01 1.51098205e-03 -2.63786763e-01 3.08617473e-01
-4.16902691e-01 1.48053241e+00 8.95736575e-01 -1.15137422e+00
-1.16904862e-01 -1.87005013e-01 3.58079880e-01 1.39534220e-01
-6.00778699e-01 2.07761273e-01 -5.95585644e-01 -9.64004397e-01
-1.27981454e-01 -9.44733679e-01 1.16956003e-01 9.58610356e-01
-1.12535022e-01 -5.66497326e-01 8.95195723e-01 -3.01294655e-01
7.69497812e-01 -2.00220227e+00 1.28609985e-01 1.73713088e-01
4.18830693e-01 5.56321204e-01 -1.02610290e+00 6.09807730e-01
-2.42716581e-01 2.79069990e-01 -6.45926520e-02 -6.36746168e-01
2.20469594e-01 9.31039900e-02 -4.77852583e-01 2.09097400e-01
5.04389763e-01 1.35667408e+00 -5.10410428e-01 -2.67972946e-01
3.10214520e-01 7.07961619e-01 -2.51690060e-01 4.78939228e-02
-5.68283796e-02 -1.58219352e-01 -7.36804605e-02 9.23702121e-01
7.46937692e-01 -3.35660130e-01 6.84349313e-02 -5.99162102e-01
1.99286520e-01 2.56733477e-01 -9.15153563e-01 2.13242483e+00
-4.64171857e-01 1.24020207e+00 3.36824246e-02 -9.42468703e-01
1.05440152e+00 1.85775220e-01 3.05313230e-01 -8.87972713e-01
8.84559080e-02 2.18725204e-02 -1.91551939e-01 -2.15607546e-02
6.43360794e-01 1.21412247e-01 2.43158937e-01 5.39661467e-01
4.82521564e-01 5.19095571e-04 3.20569068e-01 3.47784787e-01
8.87438893e-01 -9.81198624e-03 -2.00787932e-01 -3.97662371e-01
4.85815048e-01 -2.65225440e-01 -1.36681765e-01 9.27641869e-01
-2.00239301e-01 7.54844487e-01 2.51401871e-01 -1.01093626e+00
-1.43814993e+00 -1.30015481e+00 -2.30199620e-01 1.44467032e+00
9.21598449e-02 -3.15956414e-01 -1.20912954e-01 -6.17380083e-01
4.88772810e-01 2.60325849e-01 -9.47116137e-01 -1.15355939e-01
-4.01242912e-01 -2.46331379e-01 6.85949326e-01 8.04192007e-01
5.35451353e-01 -1.39110100e+00 -2.94879754e-03 -2.17544466e-01
3.32599938e-01 -9.85592365e-01 -5.28970778e-01 -3.10455680e-01
-8.42224777e-01 -1.01857889e+00 -9.51247513e-01 -8.19195032e-01
3.94316971e-01 6.25822067e-01 1.58667457e+00 5.66082001e-01
-7.27776945e-01 8.42033446e-01 -5.53525388e-02 -2.90272921e-01
-1.09755434e-02 3.28148842e-01 -9.27315280e-02 -4.28835511e-01
4.41258371e-01 -5.47827721e-01 -9.17593777e-01 3.00833642e-01
-1.03906929e+00 -2.10759342e-01 6.58272088e-01 7.55810499e-01
3.36947352e-01 -7.50155270e-01 5.08033574e-01 -3.95392299e-01
1.00850523e+00 -1.09187871e-01 -5.26896656e-01 6.26922250e-01
-5.79825938e-01 8.97036940e-02 2.14956984e-01 -7.15924978e-01
-6.33272052e-01 -2.03361616e-01 -1.28454184e-02 -9.26062107e-01
-2.08374634e-01 5.62939197e-02 3.19820374e-01 -6.31572068e-01
6.64392173e-01 8.24241117e-02 3.02065127e-02 -6.04142725e-01
6.34532094e-01 4.71919537e-01 2.68017501e-01 -7.26093888e-01
6.84505582e-01 5.82048476e-01 1.76121712e-01 -7.23024011e-01
-7.52136111e-01 -3.62578124e-01 -8.11760902e-01 -5.49888425e-02
4.39149886e-01 -6.03245497e-01 -7.58937538e-01 2.87129641e-01
-1.25461435e+00 -2.38286704e-01 -2.49171183e-01 5.09988256e-02
-4.64291692e-01 1.97251305e-01 -7.80551016e-01 -5.20139039e-01
-9.16271091e-01 -8.63762081e-01 1.16799748e+00 -6.53260201e-02
7.93514773e-03 -9.86419916e-01 5.02899736e-02 5.00894673e-02
9.02972937e-01 2.78068520e-02 7.35382438e-01 -8.61010849e-01
-6.99056506e-01 -2.34169587e-01 -8.92684162e-01 3.98098201e-01
-9.83254388e-02 3.05367678e-01 -9.63301718e-01 -4.21911985e-01
-6.74766839e-01 -7.99996972e-01 1.22760415e+00 1.32146657e-01
1.48925674e+00 -2.95572430e-01 -1.80474415e-01 8.43830407e-01
1.41066670e+00 -9.57381800e-02 6.08190298e-01 2.44051933e-01
6.36704504e-01 5.21713376e-01 1.51428983e-01 2.84781847e-02
5.53929470e-02 7.40290701e-01 2.24727824e-01 -1.19061299e-01
-6.59918547e-01 -5.41677289e-02 -3.29309762e-01 3.83236170e-01
1.93469711e-02 -2.83784926e-01 -8.00336599e-01 6.27234578e-01
-1.60106683e+00 -9.99194384e-01 4.90486979e-01 1.88958967e+00
3.89567196e-01 -2.35717222e-01 2.55917788e-01 -1.95749491e-01
5.72229862e-01 4.52969104e-01 -6.36985064e-01 -4.07551110e-01
-1.81472659e-01 4.53583419e-01 3.78582031e-01 2.02104837e-01
-1.19109023e+00 1.19046891e+00 7.51481676e+00 8.42979848e-01
-1.11518431e+00 -4.88064066e-02 5.25080562e-01 -4.92044330e-01
-2.02818662e-01 -4.65966433e-01 -4.02404428e-01 4.48971838e-02
5.03267348e-01 1.40582070e-01 9.06262934e-01 8.59645009e-01
-5.69671869e-01 5.31062961e-01 -1.39421821e+00 1.17756128e+00
2.72450656e-01 -1.77093124e+00 5.94915330e-01 -1.27198771e-01
7.60612071e-01 3.78465861e-01 4.42150056e-01 2.33540118e-01
8.39838982e-02 -1.02812958e+00 5.79563916e-01 7.17565954e-01
1.16444337e+00 -5.81788898e-01 3.71143699e-01 -3.88095170e-01
-1.16927505e+00 -1.56357199e-01 -6.81514323e-01 1.56312376e-01
-3.73351485e-01 -4.99408245e-02 -3.89628857e-01 2.82116085e-01
8.34122837e-01 9.08071697e-01 -8.19286168e-01 9.83276129e-01
1.01741984e-01 -1.45375401e-01 -1.82616904e-01 -1.09181352e-01
3.85173231e-01 -6.74509630e-02 3.81800801e-01 1.29350877e+00
1.98848367e-01 1.74208004e-02 -2.21487477e-01 1.10034752e+00
-5.31826377e-01 -1.81905091e-01 -7.80013025e-01 -3.21041524e-01
6.08335018e-01 1.51552570e+00 -4.07161802e-01 -5.86134374e-01
-1.92954347e-01 8.73243093e-01 6.70466423e-01 5.43298066e-01
-3.63349140e-01 -5.87094188e-01 1.11136913e+00 -2.63138533e-01
4.34941143e-01 -3.94307375e-01 -2.97585875e-01 -1.10233474e+00
1.07423374e-02 -5.98955214e-01 3.85618627e-01 -1.03184998e+00
-1.80235314e+00 7.01030493e-01 4.78020832e-02 -1.03242040e+00
8.31809640e-02 -1.11899137e+00 -6.89810753e-01 7.93044388e-01
-1.65055740e+00 -1.33490586e+00 -5.42316616e-01 5.59437990e-01
4.84718502e-01 -5.24187505e-01 8.68345916e-01 4.35979456e-01
-3.26648593e-01 9.36485469e-01 1.05510399e-01 3.63096535e-01
8.89545560e-01 -1.02167153e+00 1.02047980e+00 1.91286802e-01
5.32522500e-01 8.13982487e-01 1.95176542e-01 -1.49904102e-01
-1.39329183e+00 -7.77535677e-01 5.87021649e-01 -5.35292327e-01
7.43780792e-01 -6.36947095e-01 -8.84350419e-01 5.43289900e-01
2.93869227e-01 2.67314792e-01 4.12207782e-01 3.32863122e-01
-1.00044346e+00 -3.27073216e-01 -1.28097343e+00 6.31056428e-01
1.54482365e+00 -9.47122931e-01 -3.88338417e-01 4.42802757e-01
4.46503252e-01 -2.58978903e-01 -1.04576826e+00 2.50714600e-01
1.14646542e+00 -9.65588212e-01 1.55320346e+00 -1.09757054e+00
6.73517525e-01 3.00013900e-01 -1.64153308e-01 -9.98371124e-01
-7.43778884e-01 -1.70079052e-01 -9.60313808e-03 9.73598957e-01
1.94515705e-01 -3.56262535e-01 8.35184157e-01 8.46479714e-01
2.14233950e-01 -7.56539881e-01 -7.48010159e-01 -9.30758953e-01
2.35811308e-01 -2.55904764e-01 6.07133627e-01 9.01108801e-01
-3.53100240e-01 3.33562195e-01 -2.91609228e-01 -3.48395526e-01
6.28085375e-01 2.71171838e-01 6.18222415e-01 -1.28318477e+00
2.80109048e-01 -1.12191093e+00 -5.01056910e-01 -8.63091409e-01
3.24921817e-01 -1.05394411e+00 -4.19977337e-01 -1.70943034e+00
1.81993887e-01 -5.60006261e-01 -6.40544534e-01 5.17835319e-01
1.20900467e-01 8.63737047e-01 7.28846550e-01 3.95092309e-01
-7.65135705e-01 5.20096064e-01 1.15426791e+00 -5.20759165e-01
2.27973014e-01 -3.52798194e-01 -4.64307159e-01 2.40704224e-01
8.69883299e-01 -2.06939161e-01 -4.42689717e-01 -9.42058921e-01
9.53799039e-02 -3.78873616e-01 6.43830895e-01 -8.73903930e-01
1.57593310e-01 6.46269992e-02 7.54050493e-01 -5.44579387e-01
8.57923627e-01 -8.54482770e-01 -1.20204575e-01 4.10858206e-02
-9.67697024e-01 2.95426905e-01 5.54615140e-01 3.79534811e-01
-1.36061916e-02 -1.73386917e-01 5.92784703e-01 -1.68999165e-01
-7.25416839e-01 7.86984503e-01 6.82936236e-02 -2.96916932e-01
4.01408613e-01 -6.17818683e-02 -7.75827467e-01 -4.94867921e-01
-7.02180266e-01 6.89964443e-02 4.15878922e-01 8.14919174e-01
8.48217189e-01 -1.84714794e+00 -5.59014678e-01 1.41038641e-01
3.77650291e-01 -6.66677535e-01 2.80776113e-01 3.77690852e-01
-4.62976426e-01 7.71315515e-01 -7.21117556e-01 -3.41532618e-01
-1.25897145e+00 8.14176679e-01 4.09166574e-01 3.28790955e-02
-3.38083357e-01 8.61547887e-01 8.14861208e-02 -6.07663274e-01
4.27837580e-01 -2.01971680e-02 -6.65952936e-02 -1.27661511e-01
5.44493377e-01 5.08822262e-01 1.14362955e-01 -3.54212224e-01
-4.74301428e-01 5.93818188e-01 -1.66018292e-01 -7.82116801e-02
1.41381919e+00 1.89347833e-01 -1.36020631e-01 1.93062752e-01
1.65876222e+00 -7.49102175e-01 -1.12538338e+00 -3.96890610e-01
-4.43933606e-02 -6.06928110e-01 1.26881599e-01 -1.11828887e+00
-1.36016178e+00 9.76022065e-01 7.49149501e-01 1.55869067e-01
1.00338411e+00 -1.19010964e-02 5.73805928e-01 1.11501181e+00
7.61695486e-03 -1.00673163e+00 3.83007973e-01 4.25100952e-01
1.53162718e+00 -1.37350965e+00 1.96618974e-01 8.32337365e-02
-4.43384588e-01 1.34999001e+00 5.27321637e-01 -7.91807413e-01
7.10939884e-01 1.44425640e-02 8.16978216e-02 -4.38281298e-01
-8.94448519e-01 -5.16059399e-02 9.25009489e-01 5.29899895e-01
3.81202698e-01 -2.62895226e-01 -3.50690149e-02 -2.18644254e-02
2.55842537e-01 1.96317621e-02 -2.28613541e-02 7.98800290e-01
-4.66825776e-02 -1.10210061e+00 -1.67551767e-02 6.73813224e-01
-1.63298726e-01 -3.46318215e-01 -1.01643407e+00 9.28100646e-01
-3.10353994e-01 3.87376875e-01 5.28909028e-01 -2.20800638e-01
3.95085156e-01 9.76345465e-02 7.93168843e-01 -1.81320444e-01
-6.95696950e-01 -4.27697212e-01 8.08745623e-02 -7.28015661e-01
-4.60651487e-01 -3.05180073e-01 -2.89534658e-01 -5.47644734e-01
-9.85403284e-02 -4.75177407e-01 9.30979490e-01 5.64497828e-01
7.42489517e-01 1.82530314e-01 3.16777468e-01 -1.33242393e+00
-5.25830030e-01 -1.12555039e+00 -3.65118653e-01 4.46800470e-01
2.59178370e-01 -7.02528298e-01 -2.58407325e-01 -4.57228661e-01] | [11.627283096313477, 0.4905932545661926] |
28fd6c92-a866-4e0a-8482-804263f9a255 | enhanced-dynamic-sign-language-recognition | null | null | https://ieeexplore.ieee.org/document/9698904 | https://ieeexplore.ieee.org/document/9698904 | Enhanced dynamic sign language recognition using slowfast networks | In this paper, we use the SlowFast Networks developed by the Facebook research team to enhance the accuracy of dynamic sign language recognition. Firstly, we prepared the Word-Level American Sign Language (WLASL) dataset so each sign can be considered an action. We used the pre-trained SLOWFAST_8×8_R50 model provided on the official PySlowFast Github repository to initialize the weights of our model and fine-tune using the WLASL dataset and performed a parameter sweeping to fit the Dynamic Sign Language Recognition task. Through this transfer learning approach, we introduced a new state-of-the-art accuracy on the WLASL300 (300 words e.g., 300 classes) dataset with an improvement of 23.2 % top-1 accuracy compared to the previous state-of-the-art introduced in the WLASL paper using an I3D model. The top-1 accuracy was improved from 56.14% to 79.34% and the top-5 accuracy from 79.94% to 90.31%. | ['Elsayed Hemayed', 'Ahmed Elgabry', 'Ahmed Hassan'] | 2021-12-30 | null | null | null | ieee-2021-12 | ['sign-language-recognition'] | ['computer-vision'] | [-8.66563544e-02 -2.40030989e-01 -3.35995734e-01 -2.89673746e-01
-8.32583487e-01 -4.58817124e-01 5.43189406e-01 -9.26006913e-01
-8.17312121e-01 5.09832919e-01 5.65537870e-01 -1.69407427e-01
-8.09187070e-02 -4.13836300e-01 -6.26563251e-01 -7.34395325e-01
-1.04104742e-01 4.61602390e-01 5.42791963e-01 -1.96270794e-01
1.15506902e-01 5.31538785e-01 -1.80695117e+00 5.57264090e-01
8.36280346e-01 8.60933065e-01 -2.33098730e-01 9.41254139e-01
-9.32980180e-02 9.48978245e-01 -5.47221303e-01 -2.60933787e-01
4.00473744e-01 -1.07264057e-01 -5.79877317e-01 -5.27878940e-01
1.03111422e+00 -6.62027895e-01 -5.77412069e-01 7.70776689e-01
9.94523942e-01 9.54435691e-02 5.08581400e-01 -1.27483428e+00
-7.84790933e-01 7.90703297e-01 -4.58466023e-01 9.61978137e-02
1.38107955e-01 6.77387238e-01 7.60990560e-01 -9.27593470e-01
9.34000134e-01 1.38431275e+00 5.16553521e-01 1.06545472e+00
-7.22933829e-01 -1.33549905e+00 4.08042938e-01 5.96248925e-01
-1.32867956e+00 -3.89270961e-01 2.19271332e-01 -4.44318503e-01
1.21411538e+00 -2.57420186e-02 1.05343282e+00 1.51027191e+00
-3.90157431e-01 1.29442549e+00 1.40341091e+00 -4.96006608e-01
7.54701942e-02 -4.41935688e-01 6.80478871e-01 6.16374016e-01
9.81525928e-02 1.96077213e-01 -7.53864348e-01 1.16634101e-01
7.02837884e-01 -4.34261143e-01 8.44349638e-02 -2.00627353e-02
-1.23564601e+00 3.53815585e-01 5.33890128e-01 3.83116752e-01
-2.77293622e-01 3.44445616e-01 4.85936999e-01 4.44259495e-01
9.63155702e-02 -1.84950177e-02 -5.39007604e-01 -6.85849130e-01
-5.06883442e-01 2.46757597e-01 7.56830454e-01 7.12612808e-01
-3.10578365e-02 -1.05414828e-02 -5.48786700e-01 1.04485285e+00
4.48466122e-01 8.69707525e-01 6.41452551e-01 -8.82404685e-01
5.95269859e-01 5.94048381e-01 8.73530954e-02 -7.71674067e-02
-4.93173093e-01 -4.62089688e-01 -6.14410818e-01 7.26004899e-01
1.03078496e+00 -5.69255538e-02 -1.77025878e+00 1.73537791e+00
-5.70101924e-02 2.67557383e-01 -8.30346048e-02 9.37186956e-01
1.08591866e+00 3.88206929e-01 4.54777360e-01 5.88885128e-01
9.79943812e-01 -1.27758205e+00 -3.93125862e-01 1.74325585e-01
7.80656099e-01 -4.29726779e-01 1.56274712e+00 6.65972888e-01
-6.55803680e-01 -3.89537334e-01 -8.36398482e-01 2.89839189e-02
-5.65567851e-01 4.76228714e-01 5.12858212e-01 7.83713996e-01
-1.27254820e+00 2.84390002e-01 -9.66241360e-01 -5.05407453e-01
6.78702533e-01 3.33121032e-01 -3.35009158e-01 -2.38177747e-01
-9.85362649e-01 1.00357211e+00 1.38547122e-01 1.57631233e-01
-6.05404675e-01 -8.90143096e-01 -3.79263818e-01 -4.42684948e-01
9.79786664e-02 -4.87676740e-01 1.29278183e+00 -8.47004414e-01
-2.04745865e+00 1.20268893e+00 -7.12306425e-02 -3.75927538e-01
1.18632829e+00 -4.59628642e-01 -6.07030392e-01 -2.73381680e-01
-3.11927527e-01 7.83084452e-01 6.04402304e-01 -7.05308735e-01
-6.41406715e-01 -2.27926880e-01 -5.88587821e-02 3.52904089e-02
-3.96703631e-02 2.99014479e-01 -5.85235298e-01 -4.54952806e-01
-2.43806347e-01 -1.13688493e+00 1.46852478e-01 4.42104489e-01
-1.78497434e-01 -3.67254823e-01 5.78893661e-01 -9.05674160e-01
1.10720372e+00 -2.12732768e+00 -8.29798505e-02 5.08186221e-01
1.68267950e-01 9.97219324e-01 -6.96857870e-01 5.75379580e-02
-7.75407329e-02 -8.61949250e-02 1.92071386e-02 -3.68932664e-01
1.24131031e-01 1.95270583e-01 7.67335147e-02 3.78818840e-01
-1.35553226e-01 1.21672726e+00 -8.10539544e-01 -1.29841477e-01
4.20482367e-01 8.43599916e-01 -5.70935369e-01 -6.35826141e-02
1.70341522e-01 3.27386647e-01 -1.88819408e-01 8.56802702e-01
5.91117024e-01 -2.12206572e-01 -2.84835815e-01 -2.11094040e-02
-2.27675602e-01 5.63432388e-02 -1.06713891e+00 1.64750516e+00
-6.04575157e-01 7.53973365e-01 -1.56832784e-01 -3.89843285e-01
8.04706275e-01 -1.12910621e-01 5.26170194e-01 -1.00178754e+00
3.05052757e-01 3.94650906e-01 3.43306452e-01 -5.91806293e-01
-6.25230446e-02 3.44872117e-01 1.73375368e-01 2.35599741e-01
4.07960452e-02 3.77150178e-01 2.22029328e-01 -1.80830419e-01
1.19722497e+00 2.17954561e-01 9.55967605e-02 -6.30615354e-02
5.87624729e-01 -3.13368261e-01 8.45005661e-02 9.65763152e-01
-5.67749977e-01 4.92811888e-01 1.62010878e-01 -4.08241749e-01
-6.29351318e-01 -1.12090564e+00 -2.56402224e-01 1.45290267e+00
-4.06810641e-01 -1.54157877e-01 -7.96259940e-01 -6.53913796e-01
2.09142342e-01 8.38866770e-01 -7.51074910e-01 4.26654294e-02
-7.26466119e-01 -6.77730858e-01 9.27255213e-01 8.13445508e-01
9.52449560e-01 -1.41699779e+00 -4.77401584e-01 -6.20292053e-02
1.30765796e-01 -1.04521632e+00 -5.90618312e-01 -3.24992776e-01
-3.54086071e-01 -1.20124674e+00 -1.13286924e+00 -6.94994867e-01
4.78793204e-01 -4.06237721e-01 6.92411840e-01 4.49537896e-02
-2.37931579e-01 3.54287267e-01 -5.78526437e-01 -4.04110014e-01
-4.47882652e-01 2.76054442e-01 -1.08479224e-02 1.70023918e-01
6.18460000e-01 -3.38001937e-01 -5.84008873e-01 2.96963930e-01
-5.58772326e-01 1.80080310e-01 6.74946785e-01 8.28622937e-01
3.37797046e-01 -9.56396997e-01 4.63958591e-01 -3.71641964e-01
4.40887243e-01 4.93144132e-02 -8.03580403e-01 3.10162991e-01
-6.70491397e-01 1.39912337e-01 1.25206664e-01 -8.33658218e-01
-8.70653391e-01 2.28794605e-01 -5.44821084e-01 -2.60408461e-01
-1.12252161e-01 1.98297977e-01 -9.41341296e-02 -3.79989684e-01
7.57618070e-01 2.47474983e-01 1.51210770e-01 -5.57012498e-01
6.63536668e-01 1.14385188e+00 5.41847289e-01 -2.43123427e-01
5.00169039e-01 3.28201532e-01 -3.43268216e-01 -7.33481765e-01
-6.86539292e-01 -5.29824555e-01 -8.80734563e-01 -5.10468841e-01
7.07812607e-01 -8.06725681e-01 -9.85396981e-01 1.49284482e+00
-7.67948270e-01 -1.16761410e+00 -5.29938519e-01 3.07216138e-01
-4.68412399e-01 1.06958024e-01 -3.38988304e-01 -7.21088111e-01
-6.80430949e-01 -8.31895113e-01 9.59532738e-01 2.11344719e-01
-3.07980597e-01 -4.71072346e-01 2.18917668e-01 3.20348382e-01
8.32858145e-01 5.88329546e-02 2.51725197e-01 -5.21493018e-01
-3.74631703e-01 -4.02278781e-01 -6.68888509e-01 5.08487463e-01
4.86539900e-02 1.17967837e-01 -1.10737276e+00 -7.18428567e-02
-1.08347595e+00 -5.08903861e-01 1.08047485e+00 4.59407389e-01
1.13205314e+00 -5.84986731e-02 -1.30974770e-01 6.84566259e-01
7.99275398e-01 1.10745661e-01 8.62363636e-01 4.96914506e-01
6.61330581e-01 5.94149083e-02 2.10084662e-01 2.83017099e-01
5.28348923e-01 8.81708205e-01 -2.48532053e-02 1.20853186e-01
-9.75270689e-01 -5.11092126e-01 4.95029181e-01 6.65744960e-01
-4.07259017e-01 -2.00303435e-01 -1.10696638e+00 5.48216939e-01
-1.86185551e+00 -1.01084042e+00 -1.89722497e-02 2.06680965e+00
7.89180815e-01 -7.59487376e-02 4.60290998e-01 9.85163152e-02
3.67960721e-01 1.08617224e-01 -8.01729202e-01 -1.08413883e-01
-3.76501650e-01 2.04980224e-01 6.94189131e-01 7.22828031e-01
-9.78342295e-01 1.42528975e+00 6.34392786e+00 7.30476856e-01
-1.55246902e+00 1.84222609e-01 1.77078933e-01 -3.95210922e-01
2.95744836e-01 -5.77311218e-01 -1.16286087e+00 5.05631447e-01
1.09941745e+00 -1.45589747e-02 7.00335860e-01 8.05789292e-01
3.53206486e-01 -3.66400816e-02 -6.36083543e-01 1.24190032e+00
2.76140243e-01 -8.89499307e-01 1.31027460e-01 -3.14025879e-02
7.05492795e-01 9.24261093e-01 5.95446266e-02 6.47856891e-01
6.49937749e-01 -1.08277035e+00 7.10807860e-01 9.26069498e-01
1.31420958e+00 -2.12804049e-01 6.48597836e-01 3.84719819e-01
-1.09721696e+00 -2.81560332e-01 1.97282985e-01 -1.39550082e-02
3.91558915e-01 1.06177695e-01 -5.64074457e-01 -2.50128657e-01
8.40511620e-01 9.19834137e-01 -6.92949474e-01 1.38115156e+00
-4.95948613e-01 9.30177033e-01 -7.66266823e-01 -5.56358039e-01
2.19400227e-01 1.47329494e-01 4.58075106e-01 1.42022419e+00
2.50465542e-01 -3.58174220e-02 -2.07238585e-01 5.23397505e-01
-1.90298572e-01 2.41366223e-01 -2.28088930e-01 1.43692732e-01
-2.01655142e-02 8.06438327e-01 -1.00624301e-01 -5.36599159e-01
-1.85261413e-01 9.03636754e-01 2.92509317e-01 4.12182897e-01
-7.18078017e-01 -2.54168987e-01 8.47957850e-01 7.48406053e-02
2.11763248e-01 -2.77688026e-01 -1.73259959e-01 -1.29375720e+00
6.64267018e-02 -9.15821612e-01 4.37732369e-01 -9.73631322e-01
-1.37574232e+00 5.48210979e-01 4.62956121e-03 -1.25878870e+00
-3.30245018e-01 -1.14638305e+00 -3.46398205e-01 8.53576303e-01
-1.72401047e+00 -1.60201597e+00 -7.57974148e-01 6.43202722e-01
3.61099094e-01 -3.87408435e-01 8.65798235e-01 4.26128209e-01
-3.39719057e-01 1.12326312e+00 6.48111403e-02 4.86990750e-01
7.78542995e-01 -9.48080480e-01 6.44722342e-01 6.97117209e-01
-2.85943747e-02 3.49714488e-01 1.39015988e-01 -5.18403292e-01
-1.00358069e+00 -9.71755207e-01 1.06054306e+00 -6.96935773e-01
1.04622102e+00 -3.32028061e-01 -6.66805029e-01 5.41273534e-01
-2.23002210e-01 4.93472099e-01 2.75012374e-01 -1.02360934e-01
-7.10222244e-01 -1.35854170e-01 -1.18774796e+00 8.57370079e-01
1.88822138e+00 -4.36993241e-01 -2.82275319e-01 2.28885084e-01
2.53982455e-01 -7.88160324e-01 -6.25869215e-01 1.43139347e-01
1.46672642e+00 -5.59311032e-01 8.47593367e-01 -8.56683493e-01
4.69093435e-02 -2.17303470e-01 -1.76565140e-01 -1.28543186e+00
-2.95223594e-01 -3.78315002e-01 -3.37598354e-01 7.39256561e-01
5.64018846e-01 -1.01081228e+00 7.59119034e-01 7.38488257e-01
5.67281358e-02 -4.81582671e-01 -1.41088676e+00 -9.78817999e-01
2.71541357e-01 -8.87549520e-01 3.81812483e-01 3.82690400e-01
-2.52096623e-01 -1.90981910e-01 -3.36289555e-01 -3.62391263e-01
6.55542850e-01 -3.95220906e-01 1.11357903e+00 -1.11130381e+00
-1.28301442e-01 -8.02120864e-01 -5.66737294e-01 -1.15535343e+00
2.45150536e-01 -1.17840040e+00 -1.11399628e-01 -1.51148689e+00
1.86661142e-03 -3.42591792e-01 -7.03699648e-01 1.08326149e+00
1.24481805e-01 3.12344402e-01 4.90660727e-01 -4.69456911e-02
-4.25413579e-01 4.48954582e-01 1.33120644e+00 -3.11946064e-01
-4.52478826e-01 1.86567172e-01 -4.15494204e-01 4.86810952e-01
6.83455467e-01 -4.62661982e-02 5.34558669e-02 -3.94225389e-01
-1.28637657e-01 -6.41241550e-01 5.83817542e-01 -1.06892169e+00
1.83688760e-01 2.16335937e-01 6.08293004e-02 -5.89440227e-01
4.02091891e-01 -6.14764273e-01 -2.50121683e-01 6.95423961e-01
-4.49823081e-01 -4.88841325e-01 4.44709092e-01 1.06330380e-01
2.20406279e-01 2.90180057e-01 9.68516588e-01 3.58406365e-01
-1.19943857e+00 2.86354482e-01 -3.52699339e-01 1.39707521e-01
6.70384347e-01 -2.48197898e-01 -5.71149707e-01 -4.86326873e-01
-7.69637704e-01 1.89392775e-01 -2.64408737e-02 9.53023255e-01
4.34973329e-01 -1.23150086e+00 -8.56803417e-01 5.18658876e-01
5.89009345e-01 -5.74075639e-01 2.75835007e-01 8.18172812e-01
-4.71668839e-01 3.46797615e-01 -5.20549715e-01 -5.85784197e-01
-1.50029242e+00 -2.93811172e-01 6.71927452e-01 -1.82674557e-01
-9.78414834e-01 9.07809854e-01 -6.35343254e-01 -8.17856431e-01
6.49761379e-01 -8.53999197e-01 -3.54237318e-01 -7.24601895e-02
8.67535293e-01 5.34387052e-01 -1.06842093e-01 -7.77399242e-01
-5.13664305e-01 1.12750947e+00 5.01350388e-02 -3.15187931e-01
1.45667112e+00 4.58212584e-01 3.15592498e-01 4.54845250e-01
1.12793005e+00 -1.33278981e-01 -1.31915033e+00 -2.87072957e-01
-2.27542758e-01 -2.95417875e-01 3.08346599e-02 -1.62302446e+00
-1.02815139e+00 6.44500017e-01 1.28755951e+00 -6.31043851e-01
8.59198332e-01 -4.22926210e-02 5.70520759e-01 5.88791490e-01
4.80911285e-01 -1.24108005e+00 -2.41526783e-01 1.00833404e+00
1.37428081e+00 -1.34447443e+00 -4.43852484e-01 3.55582386e-02
-5.44348180e-01 9.48897421e-01 5.69842041e-01 -2.58440405e-01
9.67526197e-01 3.87771040e-01 6.87954426e-01 1.50754720e-01
-3.74785691e-01 -2.70783186e-01 6.46554291e-01 7.42989242e-01
2.70719349e-01 2.67890006e-01 -3.60948801e-01 5.68459749e-01
-4.48765695e-01 8.67225468e-01 1.18897349e-01 6.29760444e-01
-1.38840929e-01 -9.90287483e-01 -3.33692022e-02 6.70946360e-01
1.36230350e-01 -7.85605609e-03 -5.16911566e-01 9.06295836e-01
7.88311139e-02 6.08551621e-01 -1.72476172e-01 -5.41764438e-01
8.28097820e-01 3.51248443e-01 5.02042711e-01 -9.59967524e-02
-6.31199777e-01 -4.79028463e-01 2.94923961e-01 -9.11597729e-01
-3.42691869e-01 -8.66229117e-01 -1.36884046e+00 -1.42212480e-01
1.00220278e-01 -7.57004797e-01 7.83087254e-01 9.84379888e-01
3.64961028e-01 3.87651533e-01 -6.69250190e-02 -7.97309160e-01
-6.31778955e-01 -1.34426022e+00 -3.53496164e-01 4.13442910e-01
1.78927392e-01 -6.75842404e-01 -4.28630561e-01 -2.12869734e-01] | [9.156557083129883, -6.4785943031311035] |
f704b3fc-d8cc-4257-8b8b-c1708932cb5a | pay-attention-to-the-atlas-atlas-guided-test | 2307.00676 | null | https://arxiv.org/abs/2307.00676v1 | https://arxiv.org/pdf/2307.00676v1.pdf | Pay Attention to the Atlas: Atlas-Guided Test-Time Adaptation Method for Robust 3D Medical Image Segmentation | Convolutional neural networks (CNNs) often suffer from poor performance when tested on target data that differs from the training (source) data distribution, particularly in medical imaging applications where variations in imaging protocols across different clinical sites and scanners lead to different imaging appearances. However, re-accessing source training data for unsupervised domain adaptation or labeling additional test data for model fine-tuning can be difficult due to privacy issues and high labeling costs, respectively. To solve this problem, we propose a novel atlas-guided test-time adaptation (TTA) method for robust 3D medical image segmentation, called AdaAtlas. AdaAtlas only takes one single unlabeled test sample as input and adapts the segmentation network by minimizing an atlas-based loss. Specifically, the network is adapted so that its prediction after registration is aligned with the learned atlas in the atlas space, which helps to reduce anatomical segmentation errors at test time. In addition, different from most existing TTA methods which restrict the adaptation to batch normalization blocks in the segmentation network only, we further exploit the use of channel and spatial attention blocks for improved adaptability at test time. Extensive experiments on multiple datasets from different sites show that AdaAtlas with attention blocks adapted (AdaAtlas-Attention) achieves superior performance improvements, greatly outperforming other competitive TTA methods. | ['Chen Chen', 'Daniel Rueckert', 'Matthew Sinclair', 'Weitong Zhang', 'Jingjie Guo'] | 2023-07-02 | null | null | null | null | ['medical-image-segmentation', 'unsupervised-domain-adaptation'] | ['medical', 'methodology'] | [ 3.02878737e-01 5.68088032e-02 -2.29388714e-01 -9.33637679e-01
-9.77227092e-01 -6.85524106e-01 5.60653210e-02 1.23278648e-01
-7.47200668e-01 6.19561493e-01 -5.21355532e-02 -3.84939522e-01
3.84124406e-02 -3.76821041e-01 -7.39201188e-01 -8.81759763e-01
5.77766262e-02 6.89935982e-01 3.39180112e-01 2.63220519e-01
-2.06746042e-01 5.24133027e-01 -4.56862032e-01 1.57344967e-01
1.02050161e+00 1.07925403e+00 2.87205100e-01 9.38651040e-02
-1.18725352e-01 3.26961547e-01 -5.69025397e-01 -2.89571136e-01
4.47693110e-01 -4.71069694e-01 -9.51729834e-01 9.66710746e-02
4.64760929e-01 -3.91140848e-01 -2.72175133e-01 1.31117678e+00
6.80541396e-01 2.02207327e-01 4.67559487e-01 -9.38561976e-01
-5.09021461e-01 5.12415946e-01 -6.37328684e-01 3.64807606e-01
-5.11164665e-01 2.51928359e-01 2.77100205e-01 -5.76673806e-01
5.61209500e-01 7.53839612e-01 6.98707402e-01 7.44345307e-01
-1.38967478e+00 -8.58992815e-01 6.98426738e-02 9.52709913e-02
-1.44278991e+00 -1.78397328e-01 6.71527982e-01 -4.77630913e-01
4.43929136e-01 2.85908133e-01 4.77976710e-01 1.06836331e+00
3.22076917e-01 7.40243912e-01 1.12673974e+00 -6.66634366e-02
2.98970431e-01 -4.43330314e-03 5.71037717e-02 5.36969960e-01
5.39131574e-02 -2.01640964e-01 1.07632741e-01 -9.49510485e-02
9.26547408e-01 -4.09139208e-02 -3.76893491e-01 -7.34840631e-01
-1.43574023e+00 6.59306049e-01 9.69591498e-01 3.74304533e-01
-5.14566362e-01 -3.07217270e-01 7.14747369e-01 1.38934597e-01
2.60978132e-01 3.44293088e-01 -5.47892928e-01 3.53018552e-01
-9.22840595e-01 -1.48413241e-01 2.39132017e-01 8.32762122e-01
4.17390972e-01 -6.53125569e-02 -5.72869062e-01 1.08342934e+00
2.54571401e-02 1.78623989e-01 1.16141629e+00 -5.34275532e-01
3.22503120e-01 5.79631925e-01 -3.29795480e-01 -5.12799799e-01
-7.25605965e-01 -8.02589417e-01 -1.25327575e+00 1.36728555e-01
6.15507960e-01 -2.00424165e-01 -1.58025169e+00 1.78447402e+00
5.18279731e-01 6.43256754e-02 -2.80545652e-01 1.14581072e+00
7.68490970e-01 1.40794665e-01 3.07679087e-01 -1.53372243e-01
1.14133871e+00 -9.44176316e-01 -4.58602279e-01 -3.75361234e-01
8.91224384e-01 -4.25385565e-01 1.24080193e+00 1.18538283e-01
-8.59971821e-01 -4.06871796e-01 -1.02991021e+00 1.56596243e-01
-1.41528875e-01 4.23705839e-02 3.89622480e-01 5.50256312e-01
-8.37059975e-01 3.33369970e-01 -1.18157077e+00 -1.56131655e-01
9.74983454e-01 7.25450158e-01 -5.60422778e-01 -1.88892677e-01
-1.03398693e+00 8.62300992e-01 5.39613545e-01 3.24059844e-01
-8.97948265e-01 -8.52594674e-01 -8.01305711e-01 -2.17371240e-01
5.86920083e-01 -5.60681403e-01 1.33559000e+00 -1.24289644e+00
-1.52851021e+00 9.98149633e-01 9.83050540e-02 -5.18398166e-01
7.12668478e-01 2.04684019e-01 -3.62668931e-01 -4.90345508e-02
4.20153886e-01 6.29200935e-01 7.36923814e-01 -9.34359550e-01
-2.09158406e-01 -6.18319511e-01 -2.99220324e-01 1.05930418e-01
-1.72191024e-01 -1.86700881e-01 -8.36884677e-01 -9.45208848e-01
3.11076254e-01 -1.03067613e+00 -7.76623607e-01 8.53992552e-02
-6.60495043e-01 2.70401478e-01 6.82689130e-01 -6.48787141e-01
9.18357313e-01 -2.26713777e+00 6.17541112e-02 5.50597012e-01
4.04464185e-01 4.08499867e-01 -1.43314674e-01 -6.20746017e-01
-4.54180121e-01 -6.47970736e-02 -6.88824296e-01 -3.90802026e-02
-4.27796602e-01 3.13794732e-01 3.39658171e-01 5.13928950e-01
-2.62465701e-02 8.90902758e-01 -7.42674828e-01 -6.25629842e-01
2.06340462e-01 2.17887595e-01 -5.79918146e-01 1.17229797e-01
1.10731311e-01 1.29271829e+00 -6.87383354e-01 6.16776526e-01
7.15774775e-01 -4.20951396e-01 2.00535521e-01 -3.94611686e-01
3.40406954e-01 -2.88189910e-02 -7.19558120e-01 1.98062646e+00
-4.20767933e-01 2.42605031e-01 1.20191306e-01 -1.28819418e+00
6.75835669e-01 2.96790391e-01 8.03966522e-01 -9.17650521e-01
3.85563880e-01 4.24562395e-01 4.75248992e-01 -5.04698575e-01
-2.42015690e-01 -2.25699827e-01 -3.08499727e-02 1.04902461e-01
1.20398447e-01 2.32730463e-01 -1.65941447e-01 -7.02723414e-02
1.06687975e+00 -1.33281857e-01 1.16418563e-01 -3.82829726e-01
4.72380161e-01 2.64511611e-02 9.06473100e-01 7.34759986e-01
-6.89485013e-01 7.36008584e-01 3.53498816e-01 -5.85879028e-01
-9.22750354e-01 -1.02857625e+00 -4.56385255e-01 7.48777688e-01
1.99132785e-02 -5.19710742e-02 -1.02888381e+00 -1.39026022e+00
-2.38821134e-01 5.05181849e-01 -9.35808361e-01 -3.41088504e-01
-6.98919535e-01 -9.59513009e-01 3.84690762e-01 7.87003338e-01
6.14799798e-01 -1.00268507e+00 -4.87256587e-01 3.04214001e-01
-1.27354622e-01 -1.10787523e+00 -9.15929675e-01 4.18545008e-01
-1.12114227e+00 -1.04587793e+00 -1.14365208e+00 -6.90634847e-01
1.16298544e+00 -1.59945458e-01 8.04709196e-01 -5.47113046e-02
-2.71461934e-01 5.15743829e-02 -2.23934397e-01 -3.33489388e-01
-3.48286718e-01 2.85015494e-01 -2.15243056e-01 1.97396770e-01
2.76200026e-01 -3.45886320e-01 -8.19999814e-01 4.39674020e-01
-1.00565565e+00 8.10097978e-02 8.20366383e-01 1.15556633e+00
1.10593724e+00 -2.34407604e-01 5.21134496e-01 -1.30093491e+00
2.82337576e-01 -4.72039580e-01 -5.45688450e-01 3.80388021e-01
-5.51517248e-01 -8.99617895e-02 4.96520996e-01 -6.08268619e-01
-8.55355203e-01 4.24155921e-01 -2.27170050e-01 -6.52347803e-01
-1.48919180e-01 6.49384022e-01 -2.08957151e-01 -3.67533952e-01
7.40771234e-01 1.56018972e-01 1.74426466e-01 -3.64640862e-01
-1.00444838e-01 3.72486472e-01 6.18286490e-01 -3.51494521e-01
5.74082971e-01 3.19617987e-01 -7.34740719e-02 -3.48670393e-01
-8.57311368e-01 -1.88523173e-01 -1.01347470e+00 -2.17581242e-01
1.06020856e+00 -5.54767728e-01 -2.20875219e-01 5.30049682e-01
-6.96306646e-01 -7.16837227e-01 -3.39435279e-01 7.71070182e-01
-3.83129776e-01 9.28869098e-02 -3.63599747e-01 4.98997644e-02
-5.17009735e-01 -1.78080583e+00 7.29999065e-01 2.55106747e-01
-1.44041553e-01 -9.61632729e-01 -2.38150075e-01 2.03860700e-01
5.51023066e-01 4.61562246e-01 1.05290532e+00 -1.04286003e+00
-3.69599551e-01 -3.84612739e-01 -2.29285702e-01 5.05403459e-01
3.98318827e-01 -6.55424416e-01 -6.79309964e-01 -4.57958221e-01
1.20095178e-01 -1.80265173e-01 5.95969021e-01 9.29746211e-01
1.82320917e+00 -9.28006172e-02 -3.79112542e-01 1.01637340e+00
1.10334361e+00 4.22642171e-01 4.67415422e-01 4.28280860e-01
9.24526811e-01 2.53210664e-01 4.91632938e-01 1.06120989e-01
1.49104238e-01 7.65827775e-01 3.86769682e-01 -5.10006487e-01
-1.73742667e-01 1.57719970e-01 -1.65700018e-01 6.01325154e-01
1.08899638e-01 1.53254479e-01 -1.09274125e+00 5.72321832e-01
-1.64532292e+00 -2.64840394e-01 8.60697962e-03 2.53085709e+00
1.04789245e+00 9.28755924e-02 9.85170826e-02 -4.89251792e-01
7.90480614e-01 -1.15365289e-01 -1.09844005e+00 1.66821461e-02
1.92154244e-01 4.14886177e-01 9.21798587e-01 1.89312965e-01
-1.22147572e+00 7.45883822e-01 6.05448151e+00 9.42320049e-01
-1.49773514e+00 6.10457599e-01 1.10103726e+00 -1.90754339e-01
8.05490240e-02 -3.07400793e-01 -3.10019046e-01 4.43178952e-01
6.45421565e-01 4.73388396e-02 1.88897908e-01 8.17233324e-01
1.01767071e-01 1.20412268e-01 -9.96348619e-01 8.47150087e-01
-1.23894677e-01 -1.19089282e+00 -1.32262751e-01 -7.31003135e-02
6.24086380e-01 3.81667554e-01 1.14911236e-01 3.42750460e-01
1.96080372e-01 -8.86784077e-01 4.20356870e-01 1.74108952e-01
1.11397409e+00 -5.97723663e-01 8.64237189e-01 4.62326333e-02
-7.88843930e-01 2.07605049e-01 -1.70340389e-01 8.01087797e-01
-1.14604505e-02 3.10745150e-01 -1.05845320e+00 4.37167376e-01
1.01551867e+00 4.72570956e-01 -7.23778725e-01 1.26543033e+00
-3.17428149e-02 6.18618011e-01 -2.34890953e-01 3.07186872e-01
4.14835691e-01 5.74389771e-02 5.46641111e-01 9.64218974e-01
2.07102060e-01 1.47711039e-02 3.52479577e-01 8.28491569e-01
-2.71125436e-01 3.09575289e-01 -2.12580413e-01 3.69264722e-01
2.35345319e-01 1.30794501e+00 -1.01911736e+00 -3.36330801e-01
-2.78712064e-01 1.03871548e+00 1.93407670e-01 4.18909550e-01
-8.72409821e-01 -9.49647501e-02 2.02038407e-01 2.60631323e-01
1.69563636e-01 4.14140597e-02 -3.86608362e-01 -1.05676842e+00
-5.90871833e-02 -8.24876249e-01 7.18130708e-01 -4.76286620e-01
-1.35986841e+00 8.75256658e-01 5.13785183e-02 -1.33723855e+00
5.95993064e-02 -4.31097597e-01 -4.83179748e-01 9.76660728e-01
-1.22215581e+00 -1.11061442e+00 -2.96311587e-01 9.22591805e-01
5.48337817e-01 -1.67943746e-01 7.43447363e-01 5.56761086e-01
-7.67184496e-01 1.07471550e+00 2.07292110e-01 4.57988143e-01
1.00819838e+00 -1.17988908e+00 1.65439755e-01 7.61073589e-01
-1.68579623e-01 5.34519017e-01 1.84797898e-01 -6.25431478e-01
-8.31132591e-01 -1.40735400e+00 2.26111293e-01 -1.47451624e-01
4.95323688e-01 -1.75267801e-01 -1.29151034e+00 9.44838643e-01
4.20890003e-02 6.82328045e-01 7.80526340e-01 -9.35000479e-02
-1.46532446e-01 -2.55602509e-01 -1.36567676e+00 5.64639270e-01
5.81384361e-01 -1.42317072e-01 -1.86890930e-01 4.76819515e-01
4.71384883e-01 -1.01154673e+00 -1.12928140e+00 6.39424622e-01
2.90754199e-01 -4.57458466e-01 7.90181518e-01 -6.45878613e-01
-6.77398294e-02 -2.07738385e-01 1.22579187e-01 -1.40306509e+00
-4.46791828e-01 -2.73834735e-01 3.89315516e-01 9.06211615e-01
4.13608044e-01 -7.11066484e-01 8.31948817e-01 1.01652277e+00
-4.50557083e-01 -7.64515936e-01 -1.28789246e+00 -6.58114374e-01
3.52727324e-01 -2.46528625e-01 6.31990790e-01 1.27566755e+00
-4.70499337e-01 1.89307518e-02 -4.79793437e-02 2.63891727e-01
6.33416533e-01 -1.31712452e-01 4.91013438e-01 -8.77768576e-01
-2.47176319e-01 -5.93146324e-01 -4.45521623e-01 -6.29072249e-01
5.63360676e-02 -1.26764309e+00 2.56664872e-01 -1.19510627e+00
2.47701302e-01 -7.68876731e-01 -8.33005309e-01 7.99915612e-01
-2.13679701e-01 4.95611995e-01 -1.04145601e-01 2.48327285e-01
-3.92917573e-01 4.07471836e-01 1.62642407e+00 -3.11566919e-01
-2.65607387e-01 2.61500269e-01 -5.83636642e-01 6.53698683e-01
7.67975569e-01 -7.13646114e-01 -4.07826602e-01 -5.71208954e-01
-5.65378010e-01 2.72228848e-02 4.50292259e-01 -9.33093548e-01
1.79536045e-01 -1.04596773e-02 6.18898749e-01 -3.15413028e-01
-1.61486492e-01 -9.82575476e-01 4.30770367e-02 5.85804343e-01
-5.08526087e-01 -1.91703126e-01 4.73428220e-01 4.48832005e-01
-1.58226386e-01 -2.46578887e-01 1.16069603e+00 -1.63836703e-01
-5.43400824e-01 8.88836563e-01 -1.87608629e-01 6.20078817e-02
9.54321325e-01 -1.68199912e-01 7.25687221e-02 1.80751830e-02
-1.17022240e+00 3.60133886e-01 1.79310277e-01 3.34917188e-01
3.99691552e-01 -1.46891522e+00 -5.90090096e-01 2.81100601e-01
1.69111654e-01 4.17021483e-01 7.22247422e-01 1.34838617e+00
-3.73327583e-01 2.00398564e-01 -3.25369358e-01 -8.55839491e-01
-9.64440644e-01 5.32393515e-01 7.97532380e-01 -3.87824655e-01
-7.98362732e-01 9.30078804e-01 7.51527011e-01 -5.98022699e-01
1.19210929e-01 -3.75095427e-01 3.96504663e-02 -4.85135525e-01
3.27213764e-01 -2.81316370e-01 4.06722069e-01 -6.14940524e-01
-4.36859936e-01 3.22379142e-01 -6.10020816e-01 2.50021636e-01
1.23574710e+00 5.50971553e-02 6.04019091e-02 1.66054487e-01
1.18839490e+00 -2.52604008e-01 -1.38674998e+00 -5.85493624e-01
-1.58033863e-01 -3.36914361e-01 3.48076731e-01 -1.22157192e+00
-1.68271673e+00 7.45438576e-01 1.08166909e+00 -2.17956319e-01
1.28832614e+00 2.99869850e-02 9.00486946e-01 5.94801083e-02
1.77918062e-01 -8.62868786e-01 -2.23899245e-01 1.41235366e-01
9.15557742e-01 -1.52507579e+00 -3.30036849e-01 -1.33192912e-01
-8.71476293e-01 9.20970917e-01 9.80250180e-01 1.19797491e-01
6.74381018e-01 1.18643917e-01 3.25163394e-01 -1.95565954e-01
1.70693398e-02 1.85632244e-01 4.96646374e-01 6.76126420e-01
2.75989592e-01 1.27465069e-01 -9.66306925e-02 8.02519202e-01
1.52131200e-01 -5.78237921e-02 4.53111492e-02 8.16168427e-01
4.06915918e-02 -1.17368066e+00 -3.03778142e-01 7.56311357e-01
-6.28324211e-01 1.40405423e-03 -1.06962197e-01 9.62009311e-01
7.90046901e-02 2.17131585e-01 -3.18748578e-02 -2.63575822e-01
4.86373544e-01 -7.96844512e-02 3.86025459e-01 -6.63830876e-01
-7.43410289e-01 4.01434690e-01 -5.14339507e-01 -5.95657408e-01
-2.38265827e-01 -7.19360352e-01 -1.45711827e+00 1.81463927e-01
-2.82718003e-01 7.43759004e-03 5.09110868e-01 1.01486146e+00
3.60141456e-01 9.35694933e-01 5.42612612e-01 -6.03159666e-01
-3.98693949e-01 -9.02145922e-01 -4.37570900e-01 4.35753554e-01
1.68764621e-01 -7.24938393e-01 2.80143917e-01 -6.18192460e-03] | [14.5600004196167, -2.0592892169952393] |
728f0cf5-5ed8-4637-aaff-502d2ff95df1 | enolp-musk-smm4h22-leveraging-pre-trained | null | null | https://aclanthology.org/2022.smm4h-1.42 | https://aclanthology.org/2022.smm4h-1.42.pdf | Enolp musk@SMM4H’22 : Leveraging Pre-trained Language Models for Stance And Premise Classification | This paper covers our approaches for the Social Media Mining for Health (SMM4H) Shared Tasks 2a and 2b. Apart from the baseline architectures, we experiment with Parts of Speech (PoS), dependency parsing, and Tf-Idf features. Additionally, we perform contrastive pretraining on our best models using a supervised contrastive loss function. In both the tasks, we outperformed the mean and median scores and ranked first on the validation set. For stance classification, we achieved an F1-score of 0.636 using the CovidTwitterBERT model, while for premise classification, we achieved an F1-score of 0.664 using BART-base model on test dataset. | ['Sohan Patnaik', 'Manav Kapadnis', 'Ishan Manchanda', 'Archit Mangrulkar', 'Millon Das'] | null | null | null | null | smm4h-coling-2022-10 | ['dependency-parsing'] | ['natural-language-processing'] | [ 3.19535196e-01 8.12859654e-01 -3.34847957e-01 -5.61650336e-01
-1.31697965e+00 -1.46555498e-01 6.21112645e-01 7.53065228e-01
-8.51762950e-01 8.08718085e-01 1.68695882e-01 -4.15286273e-01
-1.53806955e-01 -4.64646965e-01 -6.65267110e-01 -4.43983108e-01
-4.66352910e-01 5.56146622e-01 3.80123407e-01 -1.30672276e-01
4.92785051e-02 -2.60159671e-01 -1.09871328e+00 9.04686272e-01
6.46740139e-01 1.14075685e+00 -6.01041988e-02 8.51350963e-01
1.28986910e-01 7.86073685e-01 -7.18092144e-01 -6.64343297e-01
-2.95210630e-01 -3.05276304e-01 -1.08340037e+00 -4.86121237e-01
8.30562145e-04 3.54225874e-01 1.26263127e-01 7.30433345e-01
5.59622169e-01 -1.20645002e-01 7.48695970e-01 -9.19679105e-01
-1.54856518e-01 1.03864944e+00 -5.28352797e-01 3.25717330e-01
5.72717845e-01 -3.31221551e-01 1.36494935e+00 -8.37085605e-01
7.18895257e-01 1.08601296e+00 7.63685763e-01 5.05415201e-01
-1.14829004e+00 -4.83526349e-01 7.08461404e-02 -1.17794629e-02
-9.74751413e-01 -3.89416784e-01 1.48142502e-01 -1.54979169e-01
1.27030790e+00 2.95392275e-01 2.64581919e-01 1.34403121e+00
4.74077463e-01 1.00103676e+00 1.17957592e+00 -6.01798117e-01
-1.62251145e-01 1.33038685e-01 5.59237421e-01 6.47585630e-01
-6.97773620e-02 -2.67255511e-02 -5.61520398e-01 -3.89023781e-01
-5.06535685e-03 -2.08133489e-01 2.39213035e-01 6.04461610e-01
-1.17674506e+00 7.88727462e-01 2.36718372e-01 3.77311349e-01
-5.10435522e-01 -2.31345713e-01 5.58052897e-01 3.96827966e-01
7.42222309e-01 2.75234014e-01 -9.48376715e-01 1.03119031e-01
-7.37488449e-01 1.08046927e-01 1.02634668e+00 7.16599941e-01
6.90183118e-02 -6.77876830e-01 -6.03717029e-01 9.72278535e-01
2.83785850e-01 3.91069949e-01 5.02027929e-01 -5.60426772e-01
5.51128805e-01 2.53522128e-01 -3.39584708e-01 -4.98158276e-01
-8.39603484e-01 -5.87432146e-01 -7.29318500e-01 -4.53599066e-01
2.64164090e-01 -5.27698159e-01 -1.00934279e+00 1.58318210e+00
2.12917849e-01 -3.95547450e-02 2.13807538e-01 1.48885250e-01
1.15101492e+00 3.02165031e-01 3.63112867e-01 -3.86853158e-01
1.61093855e+00 -1.04414439e+00 -8.97804558e-01 -1.88619196e-01
8.37741196e-01 -9.02053833e-01 6.58857405e-01 4.79396343e-01
-1.38371944e+00 -4.00907278e-01 -9.05354321e-01 1.86331064e-01
-2.36613989e-01 -3.43515188e-01 3.00605237e-01 4.50600296e-01
-9.56054986e-01 8.72410595e-01 -8.87450755e-01 -3.34219337e-01
3.34488034e-01 3.85306418e-01 -3.88755530e-01 1.74650460e-01
-1.43267465e+00 1.03086579e+00 3.94078881e-01 -4.08677965e-01
-6.37852788e-01 -6.30084097e-01 -7.28208542e-01 -2.18629733e-01
-1.88461524e-02 -5.51087379e-01 1.36218894e+00 -4.04298276e-01
-1.25400591e+00 1.16054368e+00 7.46527240e-02 -9.01450157e-01
4.87514079e-01 -3.67181182e-01 -6.09337568e-01 -2.63488609e-02
1.68904886e-01 6.85734868e-01 3.61923426e-01 -7.33856499e-01
-8.40935528e-01 -1.26167566e-01 -2.49708459e-01 -5.31340390e-03
-7.65161514e-02 4.11795408e-01 -1.69382095e-01 -4.84417915e-01
-1.02397047e-01 -8.96626711e-01 -3.01310390e-01 -6.62761509e-01
-6.71673477e-01 -4.85569775e-01 2.82768101e-01 -1.00025141e+00
1.46713364e+00 -2.31150508e+00 -7.84505531e-02 3.89006585e-01
9.33349058e-02 1.05201907e-01 -1.09033383e-01 4.83822227e-01
-2.51624823e-01 2.70110488e-01 -2.20499843e-01 -5.16583562e-01
-2.83426344e-01 2.24962398e-01 3.18948030e-01 3.74629766e-01
3.57669681e-01 7.06153572e-01 -8.79090011e-01 -7.40544438e-01
-3.70058089e-01 2.57310420e-01 -7.43078232e-01 1.77993938e-01
-9.87986699e-02 3.39518338e-01 -1.70623317e-01 4.83396977e-01
1.15574054e-01 -4.01114672e-01 4.69089448e-01 3.84182483e-02
1.14428088e-01 7.97759354e-01 -6.31819248e-01 1.44430888e+00
-5.23657262e-01 1.41671598e-01 3.19878571e-02 -1.02574599e+00
8.48301172e-01 6.67349100e-01 7.15234458e-01 -7.58191228e-01
1.37858212e-01 3.67726535e-01 3.45078975e-01 -5.91687441e-01
-1.22257374e-01 -1.34912550e-01 -3.15815955e-01 2.98867851e-01
2.19618127e-01 3.89046192e-01 2.20239460e-01 7.80304670e-02
1.36719596e+00 -2.04984620e-02 5.18785119e-01 -4.13002431e-01
3.90514523e-01 -2.73062050e-01 4.74375933e-01 6.60417438e-01
-3.15000303e-02 6.03164434e-01 6.33819580e-01 -1.69457704e-01
-5.16716540e-01 -7.69407272e-01 -6.34563506e-01 1.34478760e+00
-6.43375993e-01 -6.83777094e-01 -6.74213171e-01 -1.26512432e+00
-4.51090299e-02 6.97620451e-01 -6.90556645e-01 1.83876097e-01
-7.24786758e-01 -1.14940870e+00 7.18132854e-01 5.66528916e-01
2.95205295e-01 -1.15025592e+00 -2.45296791e-01 2.47742906e-01
-3.99181247e-01 -1.11262548e+00 -4.15833384e-01 7.57239640e-01
-6.19121075e-01 -1.10493207e+00 -4.92681473e-01 -9.55043435e-01
1.12822883e-01 -4.79080200e-01 1.36876225e+00 9.22810584e-02
-4.88083363e-02 -1.18012495e-01 -5.40813386e-01 -8.12963903e-01
-4.31848705e-01 2.99919665e-01 -2.38169611e-01 -2.69518822e-01
3.35697591e-01 -3.24610621e-01 -4.76560414e-01 7.88608938e-02
-5.50406992e-01 -1.17536940e-01 4.91784453e-01 1.03231192e+00
6.27201736e-01 -2.87384599e-01 9.38625991e-01 -1.45640397e+00
3.66495907e-01 -7.90779352e-01 6.38150275e-02 1.12936519e-01
-7.05237567e-01 -2.61583149e-01 2.58173525e-01 -1.74442187e-01
-6.90903008e-01 -2.55694807e-01 -6.54721081e-01 3.07947040e-01
-1.76679790e-02 7.93446064e-01 1.48835346e-01 7.53160834e-01
7.06360102e-01 -3.11177492e-01 3.09822429e-02 -6.00698888e-01
1.56539679e-02 8.83710027e-01 1.35905921e-01 -1.96915954e-01
2.45806132e-03 1.35653496e-01 -2.26065174e-01 -7.41790771e-01
-1.53230727e+00 -6.19345486e-01 -4.58574384e-01 3.78862828e-01
1.34673023e+00 -8.77946317e-01 -5.59075236e-01 3.48914623e-01
-1.08417606e+00 -2.31712624e-01 -9.79875848e-02 5.98875821e-01
-2.50975609e-01 1.90974578e-01 -9.55421746e-01 -7.89556861e-01
-7.09917247e-01 -8.26532543e-01 8.74691546e-01 -2.93416172e-01
-8.28389287e-01 -1.11714268e+00 1.05941221e-01 6.59818351e-01
9.22953859e-02 4.42605466e-01 1.19579744e+00 -1.55288529e+00
5.09602427e-01 -2.37010792e-01 2.40101460e-02 5.87982595e-01
-2.45051458e-02 -2.92869389e-01 -1.09846127e+00 -1.87720463e-01
-2.09667280e-01 -4.14361477e-01 1.08799791e+00 5.17806113e-01
1.22051084e+00 -3.12955320e-01 -5.03536582e-01 2.98737377e-01
8.56686652e-01 2.68752486e-01 5.51149487e-01 3.03372234e-01
3.12869668e-01 7.10282028e-01 6.19412243e-01 2.85934329e-01
4.32090402e-01 4.50201780e-01 1.26209736e-01 -1.73946157e-01
-3.53797227e-02 -2.05266610e-01 3.20411086e-01 8.36108267e-01
-8.81527737e-02 -2.87045687e-01 -1.10320306e+00 4.92222816e-01
-1.90315294e+00 -7.15816259e-01 -3.04867357e-01 1.95174193e+00
1.04435349e+00 7.22394407e-01 4.06926990e-01 3.15963268e-01
5.35216987e-01 1.23394795e-01 -5.72848916e-02 -4.93542701e-01
-3.68899889e-02 7.29564548e-01 4.12742943e-01 4.64930177e-01
-1.50539029e+00 5.74934959e-01 6.86852598e+00 7.70481586e-01
-6.07049704e-01 4.41982955e-01 8.54598820e-01 -6.88335970e-02
-2.46820115e-02 -2.98156798e-01 -9.83006001e-01 5.88479340e-01
1.70110250e+00 3.91826093e-01 -1.63914204e-01 5.41840076e-01
-1.04099669e-01 -5.52197881e-02 -1.17663205e+00 4.71362710e-01
3.83161195e-02 -1.01793337e+00 -4.32881683e-01 1.40766829e-01
7.22279251e-01 5.32205045e-01 -2.07195245e-02 6.57034338e-01
3.43281925e-01 -1.16216648e+00 4.08833295e-01 3.11876953e-01
6.78817749e-01 -6.54859781e-01 1.16266036e+00 4.85075861e-01
-7.27684140e-01 -4.84666638e-02 1.04738586e-01 1.20621786e-01
1.28166452e-01 8.25678766e-01 -1.16211545e+00 6.55581057e-01
8.22541535e-01 5.83614171e-01 -2.07137808e-01 7.61210322e-01
-2.33067527e-01 8.61691654e-01 -2.26985142e-01 -1.02585390e-01
1.75140128e-01 3.85044783e-01 3.26527059e-01 1.71028709e+00
-1.07263848e-01 -1.48047298e-01 3.08813900e-01 -1.63157478e-01
-2.71399975e-01 3.34482729e-01 -2.46190116e-01 1.14422880e-01
3.11084211e-01 9.39189970e-01 -4.36409742e-01 -4.17637080e-01
-4.19195384e-01 6.82702839e-01 2.05788150e-01 -1.39969438e-01
-9.66286242e-01 -4.44524795e-01 1.70260385e-01 2.71136343e-01
2.66514271e-01 3.28326553e-01 -2.83856660e-01 -8.01593125e-01
-2.49779165e-01 -1.03822160e+00 1.04767704e+00 3.93131236e-03
-1.58854783e+00 7.70431936e-01 1.63903326e-01 -6.98088646e-01
-3.02491009e-01 -5.26946664e-01 -6.36529386e-01 8.28584313e-01
-1.44273210e+00 -1.16245294e+00 3.20017874e-01 2.67215997e-01
5.17138898e-01 -1.83818579e-01 1.22710848e+00 6.04393661e-01
-5.64528883e-01 7.89022446e-01 -9.22346488e-03 2.83198357e-01
8.82573366e-01 -1.27834296e+00 4.80585963e-01 6.40461519e-02
-1.24203876e-01 4.95570809e-01 5.22429883e-01 -7.50007093e-01
-6.36907339e-01 -1.14122605e+00 1.72510266e+00 -4.71079320e-01
5.33149719e-01 -8.34867507e-02 -8.15296054e-01 7.05641985e-01
3.58974129e-01 -1.62936866e-01 1.16978049e+00 7.61506915e-01
-3.07336897e-01 1.28560215e-01 -1.34482825e+00 6.14940710e-02
1.04213679e+00 -3.35014760e-01 -6.76202059e-01 8.55677783e-01
7.13511646e-01 -2.92108148e-01 -1.30238390e+00 7.13091612e-01
6.50607705e-01 -6.50795400e-01 8.39702785e-01 -1.10639894e+00
6.00215673e-01 3.28550160e-01 -4.02643174e-01 -1.17116582e+00
-2.97547400e-01 -3.12001646e-01 7.12177856e-03 1.01223731e+00
1.28879118e+00 -5.28567076e-01 4.80645806e-01 4.00319770e-02
-2.90136009e-01 -1.29246628e+00 -7.14061379e-01 -6.61469638e-01
2.40923449e-01 -3.70726109e-01 9.60250348e-02 8.88046145e-01
3.20401758e-01 8.17562461e-01 -2.68560320e-01 -1.43000782e-02
4.85714316e-01 -1.17115974e-01 6.16385043e-02 -1.21146607e+00
-7.58958697e-01 -2.06091240e-01 1.21432263e-02 -4.91643399e-01
1.38922423e-01 -1.16837823e+00 3.19332406e-02 -1.62643576e+00
4.54646915e-01 -2.97127187e-01 -7.38258243e-01 5.94586432e-01
-2.15228796e-01 1.86017007e-01 -1.78022146e-01 -2.29555871e-02
-7.06907928e-01 -1.32000357e-01 8.78521442e-01 -1.28251970e-01
-2.16942593e-01 3.88002634e-01 -5.93599319e-01 8.31668735e-01
9.40755844e-01 -7.94173479e-01 -2.54822075e-02 -2.60796547e-01
2.49945104e-01 2.88983226e-01 -6.03058971e-02 -6.13647878e-01
1.35033568e-02 8.33167415e-03 2.92980552e-01 -7.75648355e-01
2.55488575e-01 -2.76327461e-01 -9.41726640e-02 7.80965388e-01
-6.13201797e-01 1.40718699e-01 6.85802894e-04 3.69838089e-01
-1.43453225e-01 -2.23655015e-01 8.04917455e-01 -1.16772585e-01
6.00234121e-02 -4.43457291e-02 -3.38700712e-01 3.26793551e-01
7.65383184e-01 3.46366256e-01 -2.65385628e-01 1.78792439e-02
-1.27089846e+00 3.83985966e-01 -3.50735813e-01 1.75880045e-01
4.51426268e-01 -8.26005816e-01 -1.19872069e+00 1.38611630e-01
7.70044625e-02 -3.01557202e-02 1.06663897e-03 1.30983388e+00
-2.51350671e-01 3.41657579e-01 1.70914426e-01 -5.72836757e-01
-1.58311546e+00 2.16622651e-01 1.35532767e-01 -8.62428427e-01
-2.77239233e-01 1.02750027e+00 9.25902184e-03 -4.36071664e-01
4.49793369e-01 -2.74060845e-01 -5.49406946e-01 3.00475448e-01
4.87160057e-01 4.44920361e-01 5.73438585e-01 -2.70655602e-01
-6.77363336e-01 5.08718230e-02 -4.15205926e-01 -1.84911281e-01
1.59315228e+00 2.81802833e-01 5.22056362e-03 3.89372826e-01
1.38099968e+00 -6.59432784e-02 -4.74113941e-01 -1.39935479e-01
6.19362533e-01 3.92993301e-01 -8.26483592e-02 -1.10445881e+00
-6.20564938e-01 5.06451130e-01 3.56012583e-01 4.58006769e-01
7.37407029e-01 3.18680912e-01 7.46251047e-01 4.15570498e-01
7.08683580e-02 -1.00770438e+00 3.44995670e-02 9.48836029e-01
6.53855622e-01 -1.32766283e+00 -1.18638746e-01 -4.10917670e-01
-8.27687323e-01 6.21905029e-01 1.44377455e-01 -6.70398697e-02
1.18849492e+00 4.67045188e-01 7.63035640e-02 -3.01499397e-01
-1.04601717e+00 -1.35213494e-01 4.86069053e-01 3.80566806e-01
1.15324271e+00 3.42252761e-01 -7.20536351e-01 9.22056198e-01
-2.63459176e-01 -3.38869363e-01 -1.00245111e-01 1.01043200e+00
-4.69997883e-01 -1.32430708e+00 -8.28018133e-03 7.16037989e-01
-1.37970495e+00 -2.56784618e-01 -2.47640476e-01 7.89378822e-01
2.19678804e-01 1.08924997e+00 3.49205583e-02 -5.53163946e-01
4.98725533e-01 4.22034442e-01 3.37574363e-01 -1.05824149e+00
-1.10464251e+00 3.03584784e-01 9.67994809e-01 -3.36146742e-01
-4.09847140e-01 -7.25013614e-01 -1.41813242e+00 -4.10376787e-02
-3.36793810e-01 2.75242746e-01 5.52665293e-01 8.93075705e-01
1.02004647e-01 7.21204281e-01 5.29815853e-01 -1.98990759e-02
-7.33866394e-01 -1.40318537e+00 -3.79799545e-01 4.02855039e-01
1.99731797e-01 -2.50506580e-01 -1.76673859e-01 -4.94104289e-02] | [8.54227066040039, 8.895301818847656] |
e35230b8-4776-4b8e-84ff-c4de1bf09e99 | building-multimodal-simulations-for-natural | null | null | https://aclanthology.org/E17-5006 | https://aclanthology.org/E17-5006.pdf | Building Multimodal Simulations for Natural Language | In this tutorial, we introduce a computational framework and modeling language (VoxML) for composing multimodal simulations of natural language expressions within a 3D simulation environment (VoxSim). We demonstrate how to construct voxemes, which are visual object representations of linguistic entities. We also show how to compose events and actions over these objects, within a restricted domain of dynamics. This gives us the building blocks to simulate narratives of multiple events or participate in a multimodal dialogue with synthetic agents in the simulation environment. To our knowledge, this is the first time such material has been presented as a tutorial within the CL community.This will be of relevance to students and researchers interested in modeling actionable language, natural language communication with agents and robots, spatial and temporal constraint solving through language, referring expression generation, embodied cognition, as well as minimal model creation.Multimodal simulation of language, particularly motion expressions, brings together a number of existing lines of research from the computational linguistic, semantics, robotics, and formal logic communities, including action and event representation (Di Eugenio, 1991), modeling gestural correlates to NL expressions (Kipp et al., 2007; Neff et al., 2008), and action event modeling (Kipper and Palmer, 2000; Yang et al., 2015). We combine an approach to event modeling with a scene generation approach akin to those found in work by (Coyne and Sproat, 2001; Siskind, 2011; Chang et al., 2015). Mapping natural language expressions through a formal model and a dynamic logic interpretation into a visualization of the event described provides an environment for grounding concepts and referring expressions that is interpretable by both a computer and a human user. This opens a variety of avenues for humans to communicate with computerized agents and robots, as in (Matuszek et al., 2013; Lauria et al., 2001), (Forbes et al., 2015), and (Deits et al., 2013; Walter et al., 2013; Tellex et al., 2014). Simulation and automatic visualization of events from natural language descriptions and supplementary modalities, such as gestures, allows humans to use their native capabilities as linguistic and visual interpreters to collaborate on tasks with an artificial agent or to put semantic intuitions to the test in an environment where user and agent share a common context.In previous work (Pustejovsky and Krishnaswamy, 2014; Pustejovsky, 2013a), we introduced a method for modeling natural language expressions within a 3D simulation environment built on top of the game development platform Unity (Goldstone, 2009). The goal of that work was to evaluate, through explicit visualizations of linguistic input, the semantic presuppositions inherent in the different lexical choices of an utterance. This work led to two additional lines of research: an explicit encoding for how an object is itself situated relative to its environment; and an operational characterization of how an object changes its location or how an agent acts on an object over time, e.g., its affordance structure. The former has developed into a semantic notion of situational context, called a habitat (Pustejovsky, 2013a; McDonald and Pustejovsky, 2014), while the latter is addressed by dynamic interpretations of event structure (Pustejovsky and Moszkowicz, 2011; Pustejovsky and Krishnaswamy, 2016b; Pustejovsky, 2013b).The requirements on building a visual simulation from language include several components. We require a rich type system for lexical items and their composition, as well as a language for modeling the dynamics of events, based on Generative Lexicon (GL). Further, a minimal embedding space (MES) for the simulation must be determined. This is the 3D region within which the state is configured or the event unfolds. Object-based attributes for participants in a situation or event also need to be specified; e.g., orientation, relative size, default position or pose, etc. The simulation establishes an epistemic condition on the object and event rendering, imposing an implicit point of view (POV). Finally, there must be some sort of agent-dependent embodiment; this determines the relative scaling of an agent and its event participants and their surroundings, as it engages in the environment.In order to construct a robust simulation from linguistic input, an event and its participants must be embedded within an appropriate minimal embedding space. This must sufficiently enclose the event localization, while optionally including space enough for a frame of reference for the event (the viewer{\^a}€{\mbox{$^\mbox{TM}$}}s perspective).We first describe the formal multimodal foundations for the modeling language, VoxML, which creates a minimal simulation from the linguistic input interpreted by the multimodal language, DITL. We then describe VoxSim, the compositional modeling and simulation environment, which maps the minimal VoxML model of the linguistic utterance to a simulation in Unity. This knowledge includes specification of object affordances, e.g., what actions are possible or enabled by use an object.VoxML (Pustejovsky and Krishnaswamy, 2016b; Pustejovsky and Krishnaswamy, 2016a) encodes semantic knowledge of real-world objects represented as 3D models, and of events and attributes related to and enacted over these objects. VoxML goes beyond the limitations of existing 3D visual markup languages by allowing for the encoding of a broad range of semantic knowledge that can be exploited by a simulation platform such as VoxSim.VoxSim (Krishnaswamy and Pustejovsky, 2016a; Krishnaswamy and Pustejovsky, 2016b) uses object and event semantic knowledge to generate animated scenes in real time without a complex animation interface. It uses the Unity game engine for graphics and I/O processing and takes as input a simple natural language utterance. The parsed utterance is semantically interpreted and transformed into a hybrid dynamic logic representation (DITL), and used to generate a minimal simulation of the event when composed with VoxML knowledge. 3D assets and VoxML-modeled nominal objects and events are created with other Unity-based tools, and VoxSim uses the entirety of the composed information to render a visualization of the described event.The tutorial participants will learn how to build simulatable objects, compose dynamic event structures, and simulate the events running over the objects. The toolkit consists of object and program (event) composers and the runtime environment, which allows for the user to directly manipulate the objects, or interact with synthetic agents in VoxSim. As a result of this tutorial, the student will acquire the following skill set: take a novel object geometry from a library and model it in VoxML; apply existing library behaviors (actions or events) to the new VoxML object; model attributes of new objects as well as introduce novel attributes; model novel behaviors over objects.The tutorial modules will be conducted within a build image of the software. Access to libraries will be provided by the instructors. No knowledge of 3D modeling or the Unity platform will be required. | ['Nikhil Krishnaswamy', 'James Pustejovsky'] | 2017-04-01 | null | null | null | eacl-2017-4 | ['referring-expression-generation', 'scene-generation', 'formal-logic'] | ['computer-vision', 'computer-vision', 'reasoning'] | [ 9.95238125e-03 5.33616364e-01 9.58237946e-02 -6.12231232e-02
1.49761587e-01 -9.37532902e-01 1.43362081e+00 1.83983862e-01
-1.68212801e-01 6.01514816e-01 3.63974541e-01 -4.85787332e-01
-6.17746934e-02 -8.41534138e-01 -3.36109221e-01 -2.33680397e-01
-2.69440055e-01 3.79025877e-01 1.80493221e-01 -4.59940672e-01
3.89744453e-02 8.28684509e-01 -1.91820049e+00 3.23175848e-01
4.26909745e-01 3.62285405e-01 4.54976231e-01 8.18049192e-01
-1.97440162e-01 1.34953880e+00 -3.19056123e-01 -1.10845774e-01
-2.94916451e-01 -6.34698629e-01 -1.02912438e+00 8.17051679e-02
-3.44462961e-01 3.56098376e-02 5.53653836e-02 8.21798801e-01
5.58822751e-02 2.76091814e-01 6.67632997e-01 -1.62267828e+00
-3.65285218e-01 7.58501172e-01 4.23091501e-02 -5.96332192e-01
1.16083157e+00 5.51916242e-01 4.48713124e-01 -7.59736121e-01
1.28423500e+00 1.61441314e+00 4.25868928e-01 7.41680741e-01
-1.27829123e+00 -2.60166466e-01 1.68879300e-01 3.39708984e-01
-1.19941187e+00 -1.57346487e-01 6.28162026e-01 -5.79517007e-01
1.06067097e+00 4.13282633e-01 1.22993672e+00 1.21058118e+00
-1.20716691e-01 8.22963655e-01 1.05985045e+00 -9.33540702e-01
3.47835124e-01 3.51685941e-01 -5.51051572e-02 7.65389621e-01
-3.33121091e-01 3.18069190e-01 -5.41831374e-01 -1.62655324e-01
1.11670411e+00 -2.88013279e-01 -9.27536339e-02 -5.51512241e-01
-1.67591524e+00 8.34039807e-01 6.10320345e-02 6.65016711e-01
-6.27245545e-01 4.06197846e-01 4.55019593e-01 2.10957646e-01
-1.68829143e-01 4.89602864e-01 -1.43155158e-01 -5.70817828e-01
-3.22207570e-01 7.28501201e-01 1.16090250e+00 8.57258201e-01
3.92954528e-01 -6.68775216e-02 1.35030046e-01 2.38140166e-01
8.79851699e-01 3.47013414e-01 3.08168530e-01 -1.40642643e+00
-2.94619560e-01 6.59832358e-01 5.05228043e-01 -9.23070431e-01
-5.61952829e-01 5.44910431e-01 -1.79813370e-01 8.21511686e-01
5.95811129e-01 -6.60938993e-02 -5.65771282e-01 2.11133361e+00
5.49461842e-01 1.12372853e-01 5.90942383e-01 1.10765469e+00
9.61011171e-01 9.41368639e-01 5.71540236e-01 -3.38024020e-01
1.54982042e+00 -7.72467017e-01 -8.62965226e-01 6.11233488e-02
8.93994689e-01 -6.73118830e-01 1.24856162e+00 4.31795895e-01
-1.43747020e+00 -1.80035532e-01 -8.55911672e-01 -8.66898149e-02
-6.20469689e-01 -1.92345083e-01 1.13489914e+00 3.24885130e-01
-1.12962449e+00 2.74895340e-01 -1.14332986e+00 -1.18302560e+00
-6.64482862e-02 4.40567732e-02 -1.79861188e-01 3.39263141e-01
-1.26726341e+00 1.31894326e+00 4.11955208e-01 -1.92694038e-01
-1.06317770e+00 -4.01928693e-01 -1.12046599e+00 -4.00755972e-01
3.17292035e-01 -8.24385822e-01 1.49756932e+00 -1.37085664e+00
-1.94022083e+00 1.06143284e+00 1.38931237e-02 -4.62862492e-01
5.69083452e-01 1.34620011e-01 -4.00092393e-01 3.99509408e-02
3.31073534e-03 9.07293200e-01 -3.28297317e-02 -1.54118145e+00
-1.67721301e-01 -2.65516937e-01 7.73450017e-01 4.85921353e-01
4.66130257e-01 5.77603281e-01 3.38975117e-02 -4.69596356e-01
-2.58315373e-02 -7.44065642e-01 -2.35665232e-01 3.74296963e-01
-2.24902585e-01 -2.30045289e-01 7.17056870e-01 -2.87340403e-01
8.34106922e-01 -1.89063072e+00 6.11479104e-01 1.49920136e-01
1.84329435e-01 9.53430310e-03 1.16398809e-02 1.23542023e+00
-1.55688792e-01 1.54555589e-01 -3.68116289e-01 -1.30748957e-01
6.01537824e-01 2.59196311e-01 -3.48753214e-01 3.91746044e-01
-1.73326656e-01 8.23418379e-01 -9.90984201e-01 -4.53911424e-01
7.80813515e-01 7.62237072e-01 -1.16035707e-01 2.60078967e-01
-5.58188975e-01 9.02398944e-01 -3.33370596e-01 3.83900166e-01
7.38663450e-02 -8.17615986e-02 4.71395731e-01 -5.17506897e-02
-7.44624734e-01 4.10079151e-01 -1.70091450e+00 1.66194296e+00
-8.50751102e-01 6.93223715e-01 2.24067941e-01 -6.08252048e-01
5.00637770e-01 9.62574482e-01 1.51511133e-01 -2.91879058e-01
4.01486218e-01 6.84414357e-02 -1.99194863e-01 -6.81526423e-01
3.26660991e-01 -3.31314415e-01 -3.24395180e-01 9.00016308e-01
-2.29688406e-01 -6.58299744e-01 4.28485781e-01 4.03121382e-01
5.75653076e-01 9.21220720e-01 8.26218069e-01 -1.23265058e-01
6.45909071e-01 3.90392184e-01 -8.49550366e-02 4.59067583e-01
-9.62015539e-02 3.39660123e-02 4.01562870e-01 -5.09656191e-01
-9.86181259e-01 -1.19459307e+00 -4.97333854e-02 1.10905123e+00
3.45312685e-01 -5.56822419e-01 -8.83936286e-01 2.17236012e-01
-4.61812586e-01 1.10440803e+00 -5.65798402e-01 3.89250852e-02
-6.53880477e-01 1.28637940e-01 7.52422690e-01 1.22356296e-01
1.74409717e-01 -1.67056131e+00 -1.58346856e+00 2.71459758e-01
-8.44372809e-02 -9.55942690e-01 3.51691544e-01 -1.15051478e-01
-5.61932206e-01 -8.28250885e-01 -2.49621332e-01 -7.39990115e-01
2.87924677e-01 -2.80731440e-01 9.78816867e-01 6.75595775e-02
-4.02111828e-01 8.86741221e-01 -4.65535283e-01 -8.17351282e-01
-9.41839337e-01 -7.35856295e-01 -5.05123660e-02 -3.85975301e-01
-6.13035634e-02 -7.15920329e-01 -9.00549591e-02 1.04308262e-01
-1.11522698e+00 8.42558265e-01 -1.83120579e-01 3.55333090e-01
3.46334219e-01 -5.50327837e-01 8.86542574e-02 -2.75336981e-01
6.86814249e-01 -4.25457954e-01 -5.21910071e-01 5.80621436e-02
-4.71440926e-02 -2.04429805e-01 5.12835264e-01 -7.81460404e-01
-1.19737899e+00 -1.66199714e-01 2.64283299e-01 -8.66070762e-02
-6.71527326e-01 6.03448749e-01 -2.63802886e-01 6.02126122e-02
6.77683413e-01 2.97965873e-02 6.60771504e-02 -2.38465834e-02
1.04103541e+00 2.75230765e-01 3.79058182e-01 -8.82479429e-01
5.41574240e-01 7.77311325e-01 1.18021637e-01 -6.03482127e-01
1.21021017e-01 1.44417480e-01 -6.86900020e-01 -6.57102585e-01
1.04005265e+00 -6.33679032e-01 -1.17438507e+00 3.23526412e-01
-1.44929564e+00 -8.37868929e-01 -6.00140810e-01 7.02611923e-01
-1.16222858e+00 1.98274046e-01 -5.22050381e-01 -1.21274114e+00
3.99997860e-01 -1.11430728e+00 9.13530529e-01 2.76179105e-01
-1.03375459e+00 -1.24080896e+00 5.10198250e-02 -3.28135550e-01
2.02661812e-01 8.14951658e-01 8.53776574e-01 -3.74942422e-01
-5.50952315e-01 2.47335732e-01 2.49318793e-01 -4.84534204e-01
-1.06702492e-01 2.88699031e-01 -6.59037113e-01 2.78804660e-01
-3.02004311e-02 -5.30372620e-01 -1.49491772e-01 1.00535929e-01
4.99138623e-01 -3.12520534e-01 -4.12930012e-01 1.34530202e-01
1.14393497e+00 7.91973591e-01 8.37257385e-01 5.34667492e-01
3.62936914e-01 1.10902870e+00 6.82538748e-01 4.69901770e-01
7.45536447e-01 8.44833732e-01 2.98441708e-01 3.98086049e-02
-4.06325161e-02 -2.21577421e-01 5.08081079e-01 2.76868761e-01
-4.55965489e-01 -1.65300027e-01 -1.07620406e+00 4.67732459e-01
-2.28861928e+00 -1.40213788e+00 -3.70304316e-01 1.87364435e+00
8.06661367e-01 -4.54890847e-01 1.89761534e-01 -1.99846923e-02
6.05415940e-01 1.85569629e-01 9.74837244e-02 -8.67607713e-01
3.21855322e-02 6.94814920e-02 -1.69865996e-01 8.36832047e-01
-7.66666532e-01 1.31060410e+00 4.64644623e+00 3.43954682e-01
-9.56992388e-01 1.25384808e-01 3.66328023e-02 6.02953546e-02
-4.51067895e-01 3.53545427e-01 -6.94538578e-02 7.90996403e-02
1.02221727e+00 -4.01570439e-01 8.93485308e-01 3.41198266e-01
7.16708541e-01 -4.07206386e-01 -1.38145447e+00 8.85785878e-01
5.59530556e-02 -1.35718703e+00 -1.94544822e-01 -2.97804356e-01
3.47089231e-01 -4.42328602e-01 -4.47445571e-01 9.02235359e-02
7.92167187e-01 -1.33755827e+00 1.51398802e+00 7.90474057e-01
3.73317361e-01 -4.41446185e-01 1.89896449e-01 5.65983713e-01
-1.24227691e+00 1.85922489e-01 5.65399289e-01 -6.21127844e-01
8.96095991e-01 -2.53843963e-01 -3.98135126e-01 3.53209496e-01
4.71966505e-01 2.48827606e-01 2.38433555e-01 3.95154774e-01
-4.47240829e-01 1.80026934e-01 -4.75293845e-01 -5.35759389e-01
1.26543581e-01 -3.31249416e-01 9.55141068e-01 1.17958164e+00
2.88213193e-02 3.98934931e-01 2.53801972e-01 1.29772162e+00
5.92607737e-01 1.48150846e-01 -9.25270736e-01 -2.00409129e-01
4.59459066e-01 1.07706022e+00 -1.04820204e+00 -3.31593394e-01
-2.50438869e-01 7.10124016e-01 -1.77164391e-01 6.42386794e-01
-9.75179791e-01 -2.31024459e-01 5.44754565e-01 -4.41554077e-02
-1.91027179e-01 -5.50799489e-01 -3.30657899e-01 -8.76287520e-01
-2.49997973e-01 -9.54617620e-01 -4.15091142e-02 -1.33163571e+00
-6.76346600e-01 3.51460516e-01 7.25802362e-01 -1.04501069e+00
-6.90844178e-01 -4.30911183e-01 -6.67911351e-01 7.93557048e-01
-6.81188226e-01 -1.71880329e+00 -3.33862066e-01 5.63358843e-01
2.56695718e-01 1.66073561e-01 1.15219927e+00 -1.27905339e-01
4.52619419e-02 -4.39145207e-01 -6.40616298e-01 -2.20626667e-01
1.12672612e-01 -9.58059967e-01 2.13998288e-01 4.97588813e-01
6.39814138e-02 7.55721211e-01 1.05932415e+00 -6.09423339e-01
-1.76151013e+00 -4.50757504e-01 8.39688182e-01 -5.71712017e-01
8.01515281e-01 -3.37082207e-01 -7.65285432e-01 1.11304259e+00
6.36117160e-01 -4.24831152e-01 4.86027688e-01 -6.31657362e-01
-3.34657654e-02 6.47035599e-01 -1.28533280e+00 1.32201743e+00
1.13900924e+00 -6.04560614e-01 -7.59394765e-01 2.66179562e-01
7.41512060e-01 -6.11070633e-01 -7.45712876e-01 1.15689725e-01
6.77774787e-01 -1.05811489e+00 9.02193189e-01 -6.65269077e-01
9.81005356e-02 -7.75415063e-01 -2.07453638e-01 -9.05732989e-01
2.43885472e-01 -8.06173086e-01 1.99784741e-01 1.29103112e+00
2.75802702e-01 -6.42045140e-01 -4.21609320e-02 7.03393877e-01
-2.11248040e-01 -3.28187972e-01 -7.11403966e-01 -4.38184857e-01
7.24725425e-02 -1.15696406e+00 6.29019201e-01 1.10131562e+00
7.20706940e-01 3.07266191e-02 1.67083383e-01 7.24733323e-02
1.43399537e-01 -7.66635388e-02 1.02493429e+00 -1.24353135e+00
-3.11604261e-01 -5.97099781e-01 -4.35662985e-01 -6.24752820e-01
3.36961806e-01 -9.71185744e-01 -2.98545677e-02 -2.10354853e+00
-1.44229531e-01 -8.32099915e-02 6.61567271e-01 3.81440341e-01
7.74254739e-01 -2.19960958e-01 6.79054856e-01 2.59835541e-01
-4.53397542e-01 3.19548249e-01 1.22570419e+00 3.77815276e-01
-4.92951274e-01 -6.54659629e-01 -2.37419948e-01 1.25894046e+00
5.05436957e-01 -1.39393240e-01 -4.85174537e-01 -5.72982430e-02
5.16046405e-01 3.46566916e-01 1.07912493e+00 -7.17988133e-01
3.61220092e-01 -8.25040877e-01 -1.25787348e-01 -6.61476180e-02
4.90833461e-01 -9.05136466e-01 8.21459293e-01 5.51696539e-01
-3.61558110e-01 2.07640380e-01 5.74308991e-01 6.16370626e-02
-6.72207326e-02 -2.97000825e-01 5.27443886e-01 -4.46256757e-01
-8.68101954e-01 -5.48391461e-01 -1.15665007e+00 -3.00298184e-01
1.52646446e+00 -4.32849705e-01 -2.15436384e-01 -8.04186702e-01
-1.02462077e+00 2.36479998e-01 7.20065653e-01 2.30595127e-01
4.61827904e-01 -1.29435110e+00 -2.15732113e-01 -1.47221714e-01
-1.01088054e-01 5.93814105e-02 7.63725340e-02 8.68815184e-01
-8.71078849e-01 1.49432138e-01 -3.35003376e-01 -3.78634751e-01
-1.04255986e+00 3.45463693e-01 4.35007066e-01 5.75589612e-02
-5.06114364e-01 5.25727451e-01 3.76341581e-01 -6.31851315e-01
2.70640142e-02 -5.13862371e-01 -2.94120967e-01 7.74675906e-02
6.25766993e-01 3.15480053e-01 -6.97020054e-01 -9.65590119e-01
-4.94115800e-01 6.24110222e-01 7.23208904e-01 -8.97110343e-01
1.02528119e+00 -3.16891402e-01 -3.51035357e-01 1.06913555e+00
6.33283019e-01 -1.01272151e-01 -8.24513674e-01 2.17764676e-01
-4.43964824e-02 1.56638809e-02 -4.96128738e-01 -1.00745189e+00
-7.64794499e-02 7.84483016e-01 5.09366632e-01 6.79052293e-01
8.52017045e-01 5.20823479e-01 2.25594163e-01 1.48776472e-01
7.16504812e-01 -8.84680212e-01 1.05524613e-02 5.97050786e-01
1.30499411e+00 -6.81320667e-01 -2.40209758e-01 -5.10230184e-01
-9.13357198e-01 1.44254601e+00 4.89930540e-01 2.50975192e-01
3.41217935e-01 6.59235954e-01 1.35246053e-01 -4.64024276e-01
-7.48718262e-01 -4.99606967e-01 -1.72009647e-01 8.52303863e-01
3.64263594e-01 3.44358563e-01 -7.01642156e-01 5.40392876e-01
-2.77861863e-01 3.39061111e-01 8.03407907e-01 1.31509256e+00
-2.72507757e-01 -1.08915234e+00 -7.31166244e-01 -4.57146764e-01
-6.05253540e-02 8.80582854e-02 -5.63217402e-01 1.32240796e+00
2.53992736e-01 1.08835363e+00 2.65033156e-01 4.54051420e-02
5.08257151e-01 3.02552640e-01 7.85563409e-01 -6.84115291e-01
-7.76621819e-01 -9.38703194e-02 5.35493135e-01 -4.78274792e-01
-9.96025980e-01 -8.55312824e-01 -2.13778710e+00 -3.70787024e-01
2.29623318e-01 1.40859298e-02 7.44275510e-01 1.10132349e+00
-8.10574144e-02 4.63683933e-01 2.46297792e-02 -1.24572611e+00
2.29990929e-01 -8.02271962e-01 -2.68944561e-01 3.12421679e-01
-9.28622633e-02 -7.78221369e-01 -3.92906725e-01 6.39905930e-01] | [5.089273929595947, 0.5207586884498596] |
78cdd9b2-8c1f-4925-9b7b-ed109a284f40 | time-frequency-warped-waveforms-for-well | 2305.01113 | null | https://arxiv.org/abs/2305.01113v1 | https://arxiv.org/pdf/2305.01113v1.pdf | Time-Frequency Warped Waveforms for Well-Contained Massive Machine Type Communications | This paper proposes a novel time-frequency warped waveform for short symbols, massive machine-type communication (mMTC), and internet of things (IoT) applications. The waveform is composed of asymmetric raised cosine (RC) pulses to increase the signal containment in time and frequency domains. The waveform has low power tails in the time domain, hence better performance in the presence of delay spread and time offsets. The time-axis warping unitary transform is applied to control the waveform occupancy in time-frequency space and to compensate for the usage of high roll-off factor pulses at the symbol edges. The paper explains a step-by-step analysis for determining the roll-off factors profile and the warping functions. Gains are presented over the conventional Zero-tail Discrete Fourier Transform-spread-Orthogonal Frequency Division Multiplexing (ZT-DFT-s-OFDM), and Cyclic prefix (CP) DFT-s-OFDM schemes in the simulations section. | ['Sabit Ekin', 'Hakan Ali Cirpan', 'Huseyin Arslan', 'Mostafa Ibrahim'] | 2023-05-01 | null | null | null | null | ['type'] | ['speech'] | [ 6.64768696e-01 -2.69239604e-01 -4.96765167e-01 2.30415296e-02
-1.29050732e-01 -5.30215442e-01 5.80874622e-01 -3.96029890e-01
-3.36154625e-02 9.49290991e-01 1.34618014e-01 -7.26829648e-01
-6.15699947e-01 -3.28645736e-01 2.65762806e-01 -7.73794830e-01
-7.15080082e-01 -3.68935347e-01 1.05988584e-01 1.87394228e-02
3.33260089e-01 5.17827868e-01 -1.11012065e+00 -6.72930703e-02
9.36807752e-01 1.11984694e+00 2.23811358e-01 7.81798840e-01
3.18450004e-01 3.09319943e-01 -1.00874782e+00 9.40840989e-02
2.11949959e-01 -6.17380679e-01 1.29964381e-01 -7.17762485e-02
-5.15857458e-01 -6.04003994e-03 -6.37603223e-01 6.59081995e-01
6.12873018e-01 -1.97384823e-02 9.16646004e-01 -1.37319839e+00
-4.01015490e-01 5.61313212e-01 -8.66091192e-01 4.36039031e-01
1.20250322e-01 -4.76096600e-01 6.63293377e-02 -6.19758904e-01
4.33070749e-01 7.81653285e-01 6.64217949e-01 -2.43216634e-01
-7.72975922e-01 -9.56975818e-01 -9.08793926e-01 2.82316357e-01
-1.37250304e+00 -2.36093774e-01 6.94954753e-01 -2.09043294e-01
6.95473492e-01 2.93601960e-01 2.56902546e-01 4.47946131e-01
6.85925484e-01 -1.29677445e-01 7.21857190e-01 -1.02102983e+00
4.68586423e-02 -3.70299339e-01 -4.19143647e-01 1.45807028e-01
3.93777341e-01 5.38031042e-01 -2.92243570e-01 -4.95056242e-01
8.49229753e-01 -1.97412148e-01 -5.98737419e-01 1.58690214e-02
-1.23351848e+00 6.64187670e-01 -2.25129217e-01 8.80469620e-01
-4.67284918e-01 1.38457596e-01 1.91369295e-01 5.70239782e-01
2.36636773e-01 2.18722552e-01 -2.59138912e-01 -3.29504341e-01
-1.35296500e+00 -7.33120218e-02 7.23102093e-01 1.19463539e+00
1.08030029e-01 8.82764220e-01 -4.86230493e-01 3.51175874e-01
7.53480196e-02 7.60036945e-01 6.92928135e-01 -6.91422224e-01
2.53843546e-01 -5.74247539e-01 1.57832637e-01 -7.25315094e-01
-4.16179806e-01 -8.49620879e-01 -7.25411296e-01 -2.48500690e-01
3.48284096e-01 -6.64779127e-01 -5.90705931e-01 1.17109466e+00
1.89881809e-02 3.24348122e-01 1.49580166e-01 3.95570934e-01
-7.98094869e-02 8.18392873e-01 -4.17935431e-01 -7.61545479e-01
1.59662950e+00 -5.35981178e-01 -1.30058062e+00 1.37395933e-01
3.83095473e-01 -1.23438263e+00 5.22953570e-02 2.57103801e-01
-7.38239408e-01 -4.44419593e-01 -1.38164711e+00 6.05078399e-01
1.29883006e-01 2.58768290e-01 -1.21439318e-03 1.38982594e+00
-4.95469213e-01 7.05138683e-01 -6.90738335e-02 1.72100529e-01
2.67457757e-02 -9.53120645e-03 1.76493227e-01 7.41380826e-02
-1.36907220e+00 7.27811992e-01 5.73099442e-02 -4.84158188e-01
1.96168765e-01 -9.23261821e-01 -6.16664708e-01 1.98780179e-01
-2.10216746e-01 -8.88972431e-02 1.16797495e+00 -5.60811698e-01
-1.48428011e+00 1.60998572e-03 -9.95441824e-02 -7.54227400e-01
1.12767443e-01 2.28428826e-01 -1.57171690e+00 5.92356205e-01
-1.12921469e-01 -2.92341530e-01 1.55963373e+00 -5.62314868e-01
-6.12302542e-01 2.99853669e-03 -7.43079364e-01 -3.66104364e-01
-1.21210150e-01 -7.87717178e-02 2.67589748e-01 -1.21041453e+00
1.69869423e-01 -7.67288625e-01 -9.02554095e-02 -5.92607737e-01
-4.65854071e-02 5.49167991e-01 1.35359502e+00 -6.37046158e-01
1.74344003e+00 -2.37039924e+00 -4.43534881e-01 2.72973597e-01
-3.77686828e-01 5.34891129e-01 1.18920371e-01 1.00487077e+00
-2.07528204e-01 -2.96834528e-01 -1.64880440e-01 3.72768551e-01
-3.38404715e-01 -4.69652750e-02 -4.64171916e-01 9.08300281e-01
1.48362726e-01 3.61330241e-01 -5.93687594e-01 -4.01545987e-02
1.01083286e-01 5.53070068e-01 -2.22708419e-01 -4.17105705e-01
3.53762150e-01 4.87133563e-01 -2.49875456e-01 5.20506322e-01
1.00081325e+00 3.32115918e-01 1.00312248e-01 -4.33820009e-01
-9.08238351e-01 -1.15494072e-01 -8.52976203e-01 1.09278083e+00
-4.91858482e-01 8.96336794e-01 2.05405548e-01 -7.13259816e-01
1.23396099e+00 7.84993291e-01 6.61919415e-01 -1.01576340e+00
4.83819544e-01 5.20633698e-01 2.56076247e-01 -3.77536923e-01
4.55451995e-01 -1.55825689e-01 4.52752784e-03 4.60750878e-01
9.28368568e-02 -8.34882036e-02 2.50929207e-01 -2.35673890e-01
7.30677962e-01 -2.49723569e-01 6.05464697e-01 -5.60678184e-01
5.70663691e-01 -4.55491155e-01 2.51908094e-01 8.55357274e-02
-1.04904816e-01 1.40382215e-01 1.38846606e-01 1.71297580e-01
-1.23235250e+00 -5.39612889e-01 -4.43401396e-01 7.77075410e-01
2.38096625e-01 -1.49533942e-01 -2.35131606e-01 3.36491704e-01
1.68673471e-01 1.22526240e+00 2.16015682e-01 -2.82407016e-01
-5.16504049e-01 -4.87020403e-01 8.49455178e-01 -8.37466195e-02
4.47302520e-01 -2.01742962e-01 -8.61458361e-01 6.63208008e-01
7.75936469e-02 -1.43763065e+00 -8.42170894e-01 2.78435111e-01
-6.39953554e-01 -6.32956684e-01 -1.04425979e+00 -4.93800968e-01
4.06471968e-01 4.15563226e-01 2.05633238e-01 -8.52965117e-01
-3.15431207e-01 3.47966701e-01 -7.52967536e-01 -6.55382395e-01
-2.32934356e-01 -4.81785357e-01 8.04536417e-02 2.08332166e-01
7.61546865e-02 -8.87851417e-01 -5.68874955e-01 7.28707969e-01
-7.14812040e-01 -5.24018526e-01 6.15380228e-01 6.00326538e-01
-4.14366752e-01 4.75435972e-01 9.54702914e-01 8.80535096e-02
9.17504013e-01 -4.78746653e-01 -5.87614238e-01 -2.06385747e-01
-4.81235087e-01 -3.27786952e-02 7.64027238e-01 -8.43415439e-01
-8.77656758e-01 -5.35434783e-01 2.25882307e-01 -2.50470757e-01
4.94614720e-01 4.37167227e-01 5.11349797e-01 -5.02769470e-01
8.22336137e-01 4.53382969e-01 1.91583842e-01 -6.31896555e-02
3.78284067e-01 1.20014048e+00 7.03435123e-01 -2.91950375e-01
1.15295219e+00 5.49040556e-01 2.35151321e-01 -1.41084421e+00
-8.49690959e-02 -4.50070441e-01 -1.01422995e-01 -1.08468369e-01
4.15887564e-01 -5.92941523e-01 -1.94317520e-01 2.28456542e-01
-1.02185404e+00 1.06615819e-01 -1.50540814e-01 1.15920973e+00
-4.43236232e-01 5.71747959e-01 -2.28783011e-01 -1.20460093e+00
-5.23732662e-01 -3.54628474e-01 3.93905312e-01 5.12460649e-01
-3.51304501e-01 -5.45999944e-01 -2.69784510e-01 -4.37638730e-01
1.02222979e+00 4.29970384e-01 8.38509142e-01 -1.76845029e-01
-2.24718586e-01 -4.59167242e-01 1.22024722e-01 -1.92633886e-02
9.81248356e-03 -5.23190983e-02 -4.87436265e-01 -2.98651576e-01
3.48863930e-01 6.62000716e-01 1.49412483e-01 8.86090636e-01
4.69751477e-01 -2.68681139e-01 -5.48549652e-01 7.00493991e-01
1.36968291e+00 8.58559489e-01 9.23339546e-01 -1.57516718e-01
-2.95645803e-01 1.26764208e-01 7.73752034e-01 8.97174597e-01
-2.23965704e-01 6.42595947e-01 -8.46986845e-02 2.07442001e-01
1.61117449e-01 1.01527661e-01 2.48528481e-01 7.08840430e-01
-1.06640309e-01 -4.74656701e-01 -2.67635196e-01 4.48562771e-01
-1.16278064e+00 -1.18477786e+00 -4.67346668e-01 2.42539191e+00
4.57549572e-01 3.49672288e-02 5.08711748e-02 7.31669068e-01
1.25337124e+00 1.14994287e-01 -9.31765363e-02 -9.13319945e-01
3.61691974e-02 5.05307913e-01 1.44314718e+00 2.99792379e-01
-6.67758346e-01 -4.73962054e-02 5.88623857e+00 1.37062621e+00
-1.51141083e+00 8.84418264e-02 -1.78377375e-01 3.58080834e-01
-1.37939215e-01 -3.48974541e-02 -3.27663481e-01 7.36321390e-01
1.23609662e+00 -9.54567790e-01 2.45605528e-01 3.17559212e-01
3.70129883e-01 -1.44242197e-01 -3.52724642e-01 1.05469918e+00
-2.79548794e-01 -1.07085896e+00 -5.98124862e-01 2.23245651e-01
6.64616764e-01 -3.01259011e-01 1.63170725e-01 1.53438225e-01
-6.15501463e-01 -8.04672301e-01 6.58342540e-01 2.30286181e-01
1.42911327e+00 -8.39573562e-01 2.86721915e-01 4.89005655e-01
-1.59834647e+00 -5.06076336e-01 -8.47737342e-02 -2.43215531e-01
6.69296622e-01 1.21160746e+00 -9.03413773e-01 9.48078394e-01
-4.75147039e-01 -2.35721603e-01 5.74147820e-01 1.34115171e+00
2.00563088e-01 7.30624080e-01 -3.89487147e-01 -1.56710297e-01
1.21308997e-01 -1.92190424e-01 8.32525313e-01 1.16830373e+00
1.37823296e+00 3.59540373e-01 -2.76843578e-01 2.43578970e-01
4.41578358e-01 -3.59686166e-01 -2.82425612e-01 -4.56098855e-01
1.13678062e+00 7.72593617e-01 -4.20832157e-01 -1.71420556e-02
-4.99207497e-01 5.10477543e-01 -1.18010712e+00 4.86261934e-01
-9.80745971e-01 -1.27591467e+00 2.73561478e-01 2.97371060e-01
7.84375072e-01 -9.48491573e-01 -2.25048885e-01 -4.09034759e-01
-2.18133658e-01 -3.99063468e-01 -1.27704293e-02 -2.81520694e-01
-6.49924695e-01 3.03474188e-01 -1.26027260e-02 -2.04119897e+00
-5.24772227e-01 2.96221348e-03 -5.54474652e-01 9.60576057e-01
-1.60575151e+00 -5.68370819e-01 2.74486125e-01 5.17113924e-01
1.97473302e-01 -2.90987909e-01 8.77781391e-01 6.84637964e-01
4.56779987e-01 4.63422090e-01 6.57752335e-01 -3.53407085e-01
3.65921766e-01 -5.08650243e-01 2.16180712e-01 7.91616380e-01
-3.98067206e-01 4.91601527e-01 1.15153694e+00 -6.40351236e-01
-1.16941750e+00 -9.41829324e-01 1.02269638e+00 5.68603814e-01
4.38340545e-01 4.94563282e-02 -2.17833161e-01 -1.23749729e-02
4.13190484e-01 -7.64209777e-02 1.03220499e+00 -9.31877255e-01
-1.68113783e-01 8.56271535e-02 -1.51239526e+00 4.57172006e-01
2.69066960e-01 -2.26207629e-01 -4.66140896e-01 -2.06824362e-01
5.66794395e-01 -3.08195680e-01 -7.83190191e-01 4.07306045e-01
9.58117843e-01 -8.05484951e-01 8.75936449e-01 2.95848608e-01
-7.28511065e-03 -5.19688129e-01 -3.97560954e-01 -1.45174563e+00
-5.85866094e-01 -1.56613970e+00 -2.61435270e-01 7.56542504e-01
1.22356474e-01 -8.63775790e-01 8.86397436e-02 -4.71381634e-01
-1.96103558e-01 -4.68141884e-01 -1.36136067e+00 -1.32467055e+00
-4.96705055e-01 -1.30097315e-01 4.36193109e-01 7.81032860e-01
6.47308886e-01 2.52889156e-01 -6.58723414e-01 4.28147167e-01
7.23354161e-01 -7.33639374e-02 3.40010524e-01 -9.85427618e-01
-3.74837339e-01 -1.68528974e-01 -3.90907168e-01 -1.07404387e+00
-8.16587269e-01 -5.04843473e-01 -4.20412540e-01 -1.04358482e+00
-1.08511543e+00 -1.36977509e-01 1.04125179e-01 -3.21426481e-01
4.69198108e-01 5.28455712e-02 1.81531742e-01 -2.00514272e-01
5.01677215e-01 5.23342490e-01 1.36327040e+00 3.26775789e-01
-1.35391310e-01 5.42488694e-01 -4.17379625e-02 3.91969174e-01
8.46362233e-01 -3.78317237e-01 -7.36427009e-01 2.54531503e-01
-2.88908243e-01 7.66302168e-01 -1.93131134e-01 -1.39687109e+00
2.55111724e-01 -5.65948598e-02 1.51857227e-01 -9.42211211e-01
4.13233221e-01 -1.04817164e+00 4.29330587e-01 1.05492520e+00
2.86889642e-01 -1.95738345e-01 8.75174329e-02 5.08434117e-01
-8.74455124e-02 -1.55970737e-01 1.00394583e+00 4.75615144e-01
-2.72486985e-01 2.75682774e-03 -8.58182728e-01 -1.41644165e-01
1.34000182e+00 -6.17549837e-01 -5.61422646e-01 -6.62193000e-01
-8.41773450e-02 -7.88081363e-02 -2.26388305e-01 2.28622541e-01
3.03534716e-01 -1.40322185e+00 -6.55682683e-01 3.83314520e-01
-3.05872887e-01 -7.86464751e-01 4.38671649e-01 1.18466949e+00
-5.28921664e-01 9.78493094e-01 -4.16289806e-01 -4.22157288e-01
-1.06829882e+00 1.81877330e-01 8.22014660e-02 3.31605077e-02
-4.94636565e-01 5.01752138e-01 -4.76596177e-01 4.87641335e-01
-2.17260346e-02 -2.64426261e-01 -1.50150672e-01 1.58814877e-01
9.45128381e-01 7.11245537e-01 -1.92463715e-02 -5.16028345e-01
-2.55047411e-01 7.73678899e-01 6.38916790e-01 -3.76355708e-01
7.74328411e-01 -1.80459231e-01 -3.52773862e-03 7.76123106e-02
1.29234028e+00 5.70699990e-01 -6.48882270e-01 -1.61663383e-01
-2.69409008e-02 -7.15224683e-01 -5.49535006e-02 -3.57810855e-01
-4.51608032e-01 3.88373494e-01 4.31060642e-01 7.39990413e-01
1.02179718e+00 -6.19353294e-01 1.07708561e+00 -4.17208701e-01
8.51755440e-01 -1.05559075e+00 -1.36461124e-01 5.94737649e-01
6.24260545e-01 3.46401900e-01 3.41565609e-01 -5.65545082e-01
-2.19139338e-01 1.60399818e+00 -5.83486915e-01 -8.76337811e-02
8.67747664e-01 6.51642263e-01 -1.45576283e-01 3.88466746e-01
-4.15675223e-01 -1.08627416e-01 -4.69360165e-02 1.05008829e+00
3.50168675e-01 1.64875552e-01 -1.14827657e+00 4.45263416e-01
-5.47032416e-01 2.21403807e-01 8.15989256e-01 9.98823583e-01
-1.03715348e+00 -1.18934453e+00 -1.06000412e+00 4.98759031e-01
-5.50174475e-01 3.88916321e-02 4.40180480e-01 4.62853402e-01
1.52727500e-01 1.30479300e+00 2.93639928e-01 -4.09678668e-01
2.75826544e-01 1.83433697e-01 6.35499299e-01 4.54512937e-03
4.04718705e-02 4.61840034e-01 3.53688538e-01 -3.84510048e-02
-1.40690848e-01 -3.11264247e-01 -1.40207207e+00 -4.10048157e-01
-6.24956787e-01 2.83601522e-01 8.60583544e-01 8.39258790e-01
3.09417874e-01 9.51985776e-01 1.13175189e+00 -7.86724925e-01
-7.24030197e-01 -7.84308314e-01 -9.82802153e-01 -3.00059378e-01
5.77654958e-01 -5.40199280e-01 -4.77233559e-01 -1.12705164e-01] | [6.43930196762085, 1.3129463195800781] |
a0785568-8011-4815-bff5-b358c07fccfe | a-link-recognizing-disguised-faces-via-active | null | null | http://iab-rubric.org/papers/2019_BTAS_ALINK.pdf | http://iab-rubric.org/papers/2019_BTAS_ALINK.pdf | A-LINK: Recognizing Disguised Faces via Active Learning based Inter-Domain Knowledge | Recent advancements in deep learning have significantly increased the capabilities of face recognition. However, face recognition in an unconstrained environment is still an active research challenge. Covariates such as pose and low resolution have received significant attention, but “disguise” is considered an onerous covariate of face recognition. One primary reason for this is the unavailability of large and representative databases. To address the problem of recognizing disguised faces, we propose an active learning framework A-LINK, that intelligently selects training samples from the target domain data, such that the decision boundary does not overfit to a particular set of variations, and better generalizes to encode variability. The framework further applies domain adaptation with the actively selected training samples to fine-tune the network. We demonstrate the effectiveness of the proposed framework on DFW and Multi-PIE datasets with state-of-the-art models such as LCSSE and DenseNet. | ['Mayank Vatsa', 'Richa Singh', 'Anshuman Suri'] | 2019-09-23 | null | null | null | ieee-international-conference-on-biometrics | ['heterogeneous-face-recognition'] | ['computer-vision'] | [ 2.8319436e-01 4.9551083e-03 -4.3370917e-01 -9.2907202e-01
-5.4832971e-01 -2.4215306e-01 4.9215522e-01 -5.7909673e-01
-3.3851713e-01 6.7651302e-01 -7.7780262e-02 3.1492522e-01
-4.5058063e-01 -6.5632480e-01 -6.5004957e-01 -9.5342153e-01
-1.4804359e-01 4.4884676e-01 -1.3305920e-01 9.6424609e-02
3.4943789e-02 9.0201294e-01 -1.6009533e+00 1.5462837e-02
9.1167158e-01 1.3229465e+00 -6.8592258e-02 -3.7119385e-02
-2.5741103e-01 2.1862653e-01 -6.6906369e-01 -3.8024062e-01
3.7509269e-01 -1.6258200e-01 -2.2528680e-01 2.0482282e-01
8.8436377e-01 -3.3960980e-01 -9.5768578e-02 9.3569523e-01
7.8264940e-01 1.6902898e-01 6.1224943e-01 -1.1760178e+00
-6.4182860e-01 2.5488678e-01 -5.5008703e-01 1.8396217e-01
-2.7934830e-03 2.5466043e-01 4.2913496e-01 -1.0415286e+00
3.6396354e-01 1.5320178e+00 4.5280957e-01 9.1852498e-01
-1.3291320e+00 -1.2979507e+00 3.3764297e-01 2.0417912e-01
-1.6978791e+00 -1.1532811e+00 1.0215691e+00 -2.3677491e-01
5.2534282e-01 8.9469992e-02 3.2840407e-01 1.5714686e+00
-1.7084645e-02 5.7505423e-01 9.5551926e-01 -2.6482803e-01
4.6330315e-01 1.4557379e-01 -3.3011299e-01 3.6179173e-01
3.6914316e-01 2.5860330e-01 -7.4952692e-01 -1.6726479e-01
8.4228599e-01 -5.4534070e-02 -2.6282844e-01 -6.5773588e-01
-5.1179963e-01 6.4831382e-01 3.7290221e-01 1.5946867e-02
-2.6733860e-01 -3.6844492e-01 4.7584880e-02 7.0339277e-02
6.6860539e-01 3.8924208e-01 -4.5130709e-01 7.8744300e-02
-1.1492531e+00 8.1036441e-02 6.3151425e-01 6.6215640e-01
6.5276319e-01 2.9297116e-01 -1.0046448e-01 1.1057267e+00
5.7164305e-01 5.5399603e-01 2.0299031e-01 -6.0453647e-01
4.1696131e-01 7.7398223e-01 -4.4736624e-02 -1.0792432e+00
-1.2651022e-01 -6.3928801e-01 -8.2769948e-01 5.1282316e-01
3.1733501e-01 -1.8933599e-01 -1.1451253e+00 1.8563004e+00
2.9860535e-01 3.9210099e-01 -2.4790585e-01 9.1520911e-01
9.1589934e-01 2.7867976e-01 4.3148208e-01 -4.8381761e-01
6.8030864e-01 -5.7443660e-01 -5.9812868e-01 -4.6226239e-01
-1.4246839e-01 -4.8655561e-01 6.3794816e-01 3.9643022e-01
-7.1240383e-01 -7.3594755e-01 -1.0857136e+00 3.6697185e-01
-1.8284032e-01 2.0273386e-01 4.3958938e-01 9.1852140e-01
-8.0172861e-01 4.6303049e-01 -7.5826675e-01 -2.8802580e-01
1.1066761e+00 8.0987322e-01 -5.0865495e-01 -1.7363290e-01
-1.0606467e+00 6.0320818e-01 2.8750268e-01 5.1070446e-01
-1.0714549e+00 -6.7909443e-01 -6.0418600e-01 -3.9126851e-02
6.8607223e-01 -1.7958903e-01 6.4621103e-01 -1.4840097e+00
-1.5296999e+00 8.4385997e-01 5.0757915e-02 -1.4342441e-01
5.1860571e-01 -2.7955645e-01 -8.0229700e-01 -1.4874415e-01
-3.7561801e-01 7.1140349e-01 1.3259181e+00 -1.2572136e+00
1.0654966e-01 -8.6404395e-01 -1.0478755e-01 2.2465643e-01
-6.4174449e-01 6.2911958e-02 -4.5030609e-01 -5.2185053e-01
-5.9267115e-02 -7.0676130e-01 6.6883743e-02 4.9401349e-01
-1.0949325e-01 -2.2510198e-01 1.0951117e+00 -4.0689266e-01
8.8812387e-01 -2.2918849e+00 1.7449632e-01 2.2732666e-01
1.8935037e-01 7.2067708e-01 -2.7148044e-01 -2.4280013e-01
-2.7100328e-01 1.8062290e-02 -1.5994623e-01 -2.6225561e-01
-2.5572824e-01 2.8093451e-01 -1.6606830e-01 4.2886680e-01
6.9481266e-01 6.5221983e-01 -4.9268189e-01 -5.9626245e-01
-5.1308203e-02 6.8469352e-01 -6.2339175e-01 5.6004161e-01
-3.6311758e-01 7.8546721e-01 -5.0738454e-01 7.7232379e-01
1.1748047e+00 2.4549223e-02 1.9313607e-01 -3.1100839e-01
2.2562717e-03 -1.0571275e-01 -1.2874285e+00 1.6258222e+00
-1.6817611e-01 2.3556036e-01 2.8600216e-01 -1.0521321e+00
1.4639288e+00 2.2206233e-01 4.1955975e-01 -8.2344937e-01
2.5893003e-01 -1.3706652e-02 1.1410940e-01 -6.7864013e-01
-1.3921399e-01 1.8608154e-01 4.4222835e-01 2.2510771e-02
3.0427289e-01 4.5713142e-01 -1.6728583e-01 -4.2319149e-01
6.5427196e-01 5.0524265e-01 2.0154020e-01 -2.8295884e-01
5.5919272e-01 -5.6232280e-01 1.0279009e+00 3.3823055e-01
-5.5444574e-01 4.0588620e-01 4.3021926e-01 -3.5645372e-01
-6.5845764e-01 -1.1370598e+00 -4.1421828e-01 1.1341209e+00
-1.3952138e-01 1.3964528e-01 -6.5759152e-01 -9.7675180e-01
2.2862236e-01 3.5043854e-01 -7.6384485e-01 -1.6742362e-01
-5.9127587e-01 -8.6265194e-01 5.5530155e-01 4.2750061e-01
8.0274230e-01 -1.0409774e+00 -2.4799332e-01 -2.2279103e-01
3.4352827e-01 -7.6538819e-01 -5.9760075e-02 8.7784506e-02
-8.0460304e-01 -8.9433402e-01 -5.8250892e-01 -6.6452754e-01
8.7329727e-01 -1.9703853e-01 1.1230441e+00 -8.0299906e-02
-3.5419649e-01 -8.4779650e-02 -2.1022281e-01 -2.0743096e-01
6.2505454e-02 1.9362384e-02 1.8060701e-01 4.8023599e-01
5.4908347e-01 -5.6158185e-01 -6.5202355e-01 5.0866199e-01
-7.5786602e-01 -4.1864890e-01 7.5161940e-01 6.9862074e-01
4.3165675e-01 -1.8309461e-02 8.7980878e-01 -1.0830073e+00
3.2523850e-01 -4.8297429e-01 -5.0705636e-01 3.1651795e-01
-4.9452400e-01 -7.2191786e-03 5.4563391e-01 -7.3184937e-01
-1.3457323e+00 2.9114470e-01 -7.4756309e-02 -6.0460097e-01
-4.3723264e-01 1.3141866e-01 -1.0374700e+00 -3.8994411e-01
8.0514663e-01 -3.3003781e-02 1.6499941e-01 -4.2081583e-01
1.5971687e-01 7.2384888e-01 2.9731321e-01 -7.9092848e-01
8.0608577e-01 2.5966442e-01 -7.6157384e-02 -9.3273002e-01
-7.2250140e-01 -3.1648040e-02 -7.7711362e-01 -3.6404598e-01
3.5417560e-01 -8.2247722e-01 -6.1889690e-01 4.7661954e-01
-9.1610843e-01 -5.9584063e-02 -3.7992090e-02 2.6061821e-01
-1.6473511e-01 6.2427856e-02 2.1861322e-01 -1.0490910e+00
-1.7233332e-01 -1.2355337e+00 9.6039879e-01 6.3508469e-01
-1.7173567e-01 -7.1213359e-01 -1.3372548e-01 3.8605464e-01
6.2498635e-01 2.5711867e-01 7.8581363e-01 -8.9663976e-01
-6.9537205e-01 2.4723003e-02 -1.7780812e-01 5.3282708e-01
3.7124464e-01 6.4106710e-02 -1.3242279e+00 -4.9145287e-01
6.7377524e-03 -6.7053860e-01 6.3972378e-01 1.0102834e-01
1.5194634e+00 -3.2779405e-01 -3.3509350e-01 9.0118986e-01
1.1253389e+00 5.4949689e-01 8.7190342e-01 -1.1066171e-01
3.8702777e-01 6.7360860e-01 5.0256354e-01 3.6278674e-01
-2.8583175e-01 7.2422212e-01 4.9927655e-01 1.2108947e-02
-2.6094610e-02 -4.3385107e-02 6.9948239e-03 4.1202801e-01
-5.6552652e-02 -1.6690062e-01 -8.3907771e-01 2.7006698e-01
-1.4426821e+00 -9.2689800e-01 6.7623723e-01 2.3220844e+00
8.0583918e-01 1.5838771e-01 1.3437261e-02 2.3980094e-02
8.2708389e-01 2.9757625e-01 -1.0884554e+00 -3.7573084e-02
-1.0337321e-01 5.4162556e-01 2.3797408e-01 8.2112914e-03
-1.1119323e+00 7.9609102e-01 5.5752821e+00 8.8278538e-01
-1.6642916e+00 -1.5825716e-01 9.8817724e-01 -7.9259358e-02
1.7867051e-01 -4.0311196e-01 -9.9069232e-01 5.0893253e-01
5.5605811e-01 7.0666403e-02 4.4936812e-01 1.0794637e+00
2.1941055e-01 2.3137736e-01 -1.2794106e+00 1.0840569e+00
3.8537711e-01 -8.9391851e-01 1.5224704e-01 -1.4589710e-02
5.6621170e-01 -5.2750468e-01 5.4684299e-01 4.0691987e-01
-2.8715998e-01 -1.4387436e+00 3.3690754e-01 6.6093147e-01
1.1495123e+00 -7.4204701e-01 3.6813632e-01 2.4535581e-01
-8.8380718e-01 -1.9502005e-01 -4.5471677e-01 3.0684650e-01
-4.1392219e-01 4.5855296e-01 -8.8498038e-01 3.2834551e-01
5.7552284e-01 5.8827734e-01 -8.3007765e-01 1.1580832e+00
3.2035995e-02 6.3060850e-01 -2.5864866e-01 2.1924065e-01
-3.4842992e-01 -6.7042440e-02 3.8829970e-01 7.5821495e-01
1.8122374e-01 -9.6588559e-02 1.9118096e-01 1.0167992e+00
-3.2638338e-01 5.3292926e-02 -3.6450872e-01 -3.1228769e-01
7.8425950e-01 1.1794846e+00 -3.9288470e-01 1.5683128e-01
-2.4080381e-01 7.0741248e-01 4.8676369e-01 4.1652489e-01
-6.1164796e-01 -8.5334145e-02 7.5429147e-01 2.4406926e-01
2.7722082e-01 -1.5410767e-01 -2.9915860e-01 -9.6336764e-01
3.8004451e-02 -1.0310833e+00 2.3499648e-01 -3.8049302e-01
-1.4021991e+00 7.4906319e-01 -7.3023170e-02 -1.1071099e+00
-1.4550166e-01 -6.8843240e-01 -5.1459670e-01 1.1001133e+00
-1.3367703e+00 -1.0281513e+00 -5.8340502e-01 5.8705246e-01
4.1335726e-01 -7.1232402e-01 9.2139208e-01 5.3567922e-01
-8.1983179e-01 9.9865836e-01 2.5421750e-02 3.5722020e-01
1.0257256e+00 -6.0638863e-01 2.8231797e-01 7.3546255e-01
-1.4580999e-01 9.2234224e-01 5.1040721e-01 -6.7508519e-01
-1.4169145e+00 -1.2479477e+00 3.3049124e-01 -1.8259555e-01
9.8531060e-02 -5.8086544e-01 -1.0777820e+00 3.4852478e-01
-2.2355606e-01 3.7014309e-01 6.6402572e-01 2.0211057e-01
-5.7275200e-01 -7.6153558e-01 -1.5169400e+00 3.7151471e-01
1.1849487e+00 -3.4748754e-01 -3.9142075e-01 -3.6164641e-02
2.2740650e-01 -3.2241178e-01 -5.4158956e-01 8.4432113e-01
7.0748222e-01 -8.8927799e-01 1.0493010e+00 -5.2995133e-01
1.0738288e-01 -2.7792593e-02 -2.5812255e-02 -1.3254404e+00
-4.0450752e-01 -4.0015569e-01 -9.8785020e-02 1.6034478e+00
2.1847159e-01 -4.6418419e-01 1.3038172e+00 7.2032273e-01
2.8790846e-01 -7.6816332e-01 -1.1957510e+00 -6.6177279e-01
-3.7369262e-02 -3.8553365e-02 7.2468019e-01 8.9532763e-01
-6.7597347e-01 2.4330072e-01 -3.2711282e-01 1.1905245e-01
9.6315205e-01 -2.3468676e-01 6.2144905e-01 -1.6527098e+00
-1.5917492e-01 -1.5029754e-01 -5.9704208e-01 -9.4952685e-01
4.0452322e-01 -6.7546731e-01 1.5728209e-02 -8.6772490e-01
7.0510909e-02 -7.3393112e-01 -3.0552185e-01 5.2112645e-01
-5.2335556e-03 2.4103811e-01 -1.1145929e-01 -1.1978978e-03
-4.2607424e-01 7.5559598e-01 1.2342286e+00 -1.8341060e-01
-1.0989891e-01 -2.5734723e-02 -7.4266523e-01 4.3399659e-01
6.4510393e-01 -2.3871845e-01 -6.2240201e-01 -4.5200402e-01
-1.1597789e-01 -2.8206581e-02 1.0279818e-01 -1.0456728e+00
1.5080380e-01 -4.0885314e-01 1.0738163e+00 -2.7763113e-01
6.8486619e-01 -1.0236466e+00 2.2477044e-01 -6.9468871e-02
-3.9812085e-01 -4.2538965e-01 2.8485331e-01 5.4531640e-01
3.9644100e-02 3.2262783e-02 1.1873204e+00 2.1163349e-01
-8.2909113e-01 7.8561497e-01 2.5900945e-01 1.1077768e-01
9.3373311e-01 -3.8110751e-01 -2.6824224e-01 8.8613585e-02
-6.4759064e-01 1.7172815e-02 2.0725644e-01 7.6236534e-01
6.3895434e-01 -1.3137401e+00 -6.7928898e-01 8.5948581e-01
1.7478883e-01 1.1871583e-01 3.3133340e-01 3.3548900e-01
-1.0958727e-01 2.1575707e-01 -5.8435297e-01 -7.5767601e-01
-1.2248894e+00 2.8254223e-01 4.3181595e-01 3.1385177e-01
-1.5169261e-01 9.6016967e-01 2.8493288e-01 -3.8784930e-01
5.4598141e-01 4.0671128e-01 -4.7590205e-01 4.7405191e-02
5.9334069e-01 1.5512788e-01 2.5876343e-01 -7.4793851e-01
-6.2018043e-01 4.2875198e-01 -2.6297802e-01 4.0573317e-01
1.3980566e+00 6.9444679e-02 7.0222378e-02 9.0044841e-02
1.1916171e+00 -1.7537601e-01 -1.5378854e+00 -2.5271076e-01
-2.8667089e-02 -8.5485047e-01 5.3470705e-02 -1.0244107e+00
-1.3448292e+00 6.8163270e-01 1.0640272e+00 -3.1487608e-01
1.1431446e+00 -3.3866259e-01 1.2354939e-01 2.9144451e-01
3.8031834e-01 -1.0643015e+00 1.8999685e-01 3.0292478e-01
1.0725298e+00 -1.2971531e+00 7.9556540e-02 -5.7340187e-01
-3.4109232e-01 1.0317340e+00 1.2374616e+00 -4.8019268e-02
7.4194467e-01 3.2812190e-01 1.2201815e-01 6.4708151e-02
-5.9651762e-01 1.5690942e-01 4.0272123e-01 7.9316533e-01
5.0386840e-01 -4.3885674e-02 1.0500594e-02 6.7500454e-01
4.9963634e-02 2.8563389e-01 -3.4152398e-01 7.6868182e-01
-3.4864122e-01 -1.0463668e+00 -6.6388100e-01 4.6106225e-01
-2.2829032e-01 4.2884272e-01 -3.9714757e-01 6.4769912e-01
6.1724567e-01 7.5673085e-01 2.5841641e-01 -3.2065883e-01
2.2974777e-01 2.7465826e-01 7.1426034e-01 -7.0828444e-01
-2.4135038e-01 -1.1206580e-01 -9.2457712e-02 -3.8050032e-01
-3.4156051e-01 -4.8003185e-01 -5.8710003e-01 -1.4569373e-01
-1.9079418e-01 -2.2243708e-02 6.0997790e-01 9.3211865e-01
4.2572361e-01 3.3631471e-01 8.7303054e-01 -7.2550905e-01
-7.4297410e-01 -9.8754817e-01 -4.9483183e-01 4.2100713e-01
1.5624240e-01 -1.2035584e+00 -3.3637735e-01 -9.5829904e-02] | [13.221677780151367, 0.73390793800354] |
4bde510b-ca35-4086-8d09-29d181a7e5b2 | deep-conversational-recommender-systems-a-new | 2004.13245 | null | https://arxiv.org/abs/2004.13245v1 | https://arxiv.org/pdf/2004.13245v1.pdf | Deep Conversational Recommender Systems: A New Frontier for Goal-Oriented Dialogue Systems | In recent years, the emerging topics of recommender systems that take advantage of natural language processing techniques have attracted much attention, and one of their applications is the Conversational Recommender System (CRS). Unlike traditional recommender systems with content-based and collaborative filtering approaches, CRS learns and models user's preferences through interactive dialogue conversations. In this work, we provide a summarization of the recent evolution of CRS, where deep learning approaches are applied to CRS and have produced fruitful results. We first analyze the research problems and present key challenges in the development of Deep Conversational Recommender Systems (DCRS), then present the current state of the field taken from the most recent researches, including the most common deep learning models that benefit DCRS. Finally, we discuss future directions for this vibrant area. | ['Nguyen Lu Dang Khoa', 'Nguyen H. Tran', 'Lina Yao', 'Salma Abdalla Hamad', 'Munazza Zaib', 'Wei Emma Zhang', 'Quan Z. Sheng', 'Dai Hoang Tran'] | 2020-04-28 | null | null | null | null | ['goal-oriented-dialogue-systems'] | ['natural-language-processing'] | [-1.31907552e-01 -8.02075714e-02 -1.18581943e-01 -5.24128199e-01
-2.53565848e-01 -3.54392409e-01 8.36902142e-01 -2.20083803e-01
-1.97233886e-01 4.96303439e-01 9.10917819e-01 -3.85290116e-01
-4.93654698e-01 -9.81598854e-01 8.60649943e-02 -5.15030384e-01
-7.36270249e-02 5.15705347e-01 -3.63980383e-02 -1.09954154e+00
7.29166448e-01 3.35476577e-01 -1.63394880e+00 9.55777705e-01
9.00998116e-01 9.65627491e-01 2.35619411e-01 7.01122165e-01
-5.20370364e-01 1.01696622e+00 -4.88216370e-01 -5.99968851e-01
-1.22768402e-01 -8.06206018e-02 -9.83738959e-01 -5.01613557e-01
7.65082538e-02 -2.82402039e-01 -6.76415384e-01 3.38262349e-01
7.32655764e-01 1.19905317e+00 7.24900782e-01 -1.02009726e+00
-1.42703128e+00 1.02146053e+00 1.21154867e-01 2.57336169e-01
7.48226166e-01 -5.83355427e-01 1.43632674e+00 -1.13212359e+00
3.57327163e-01 1.31478405e+00 6.05605304e-01 8.54111969e-01
-4.86739069e-01 -4.37032312e-01 7.26237774e-01 4.68003273e-01
-8.89512956e-01 -1.07289702e-01 6.01185739e-01 -4.31186408e-01
1.27893829e+00 3.50039363e-01 6.06374979e-01 1.34639597e+00
1.70559585e-01 1.21832442e+00 6.07763708e-01 -2.04331860e-01
1.29485562e-01 2.02060491e-01 6.26210272e-01 1.21333666e-01
-5.30101776e-01 1.36825591e-02 -2.85080999e-01 -3.52709919e-01
4.10869956e-01 5.94261110e-01 -3.81697305e-02 2.81469561e-02
-6.92985773e-01 1.33856142e+00 4.65942860e-01 5.97184777e-01
-2.20687225e-01 -4.76976126e-01 2.41262585e-01 5.83801031e-01
7.37819672e-01 7.51294672e-01 -6.36440217e-01 -1.11502700e-01
-4.98636305e-01 4.05299067e-01 1.13283002e+00 8.46163392e-01
1.22905716e-01 -8.82342905e-02 -2.77100414e-01 1.54049110e+00
6.85346663e-01 2.45942622e-01 6.21220648e-01 -8.12557936e-01
-4.95756119e-02 3.69640291e-01 1.09138347e-01 -1.04114091e+00
-3.99690121e-01 -3.75102431e-01 -7.72556305e-01 -4.48669255e-01
-1.06756225e-01 -6.50369763e-01 -9.05572027e-02 8.69293213e-01
7.42371380e-02 1.52090296e-01 4.35202360e-01 7.88957894e-01
1.73001885e+00 1.00054419e+00 -6.17948808e-02 -9.44150761e-02
1.02811730e+00 -1.43253863e+00 -6.44941866e-01 1.28497943e-01
2.16752008e-01 -8.84007275e-01 9.61927891e-01 6.84072912e-01
-9.16716933e-01 -6.96586013e-01 -5.54455757e-01 -1.97058365e-01
-6.00541651e-01 1.66072294e-01 1.06381226e+00 6.47958755e-01
-1.37336493e+00 6.01467013e-01 -3.13425660e-01 -5.35665214e-01
8.05104598e-02 5.69008410e-01 2.13181838e-01 -5.66545054e-02
-1.62013102e+00 9.82362211e-01 -5.31848669e-01 3.23420435e-01
-7.16498017e-01 -5.59333384e-01 -3.30885530e-01 2.36135378e-01
3.17636877e-01 -4.41289902e-01 1.78485680e+00 -7.04232633e-01
-2.41772413e+00 2.35275224e-01 3.42676975e-02 -3.57937068e-01
-3.35235223e-02 -7.15412915e-01 -8.05491388e-01 -3.62259954e-01
-6.88618481e-01 1.91746101e-01 2.87503839e-01 -8.35925519e-01
-9.22105134e-01 9.73561928e-02 7.13487089e-01 4.21609223e-01
-4.88128901e-01 4.46932435e-01 -2.67714679e-01 -2.77717769e-01
-6.37100995e-01 -9.72652912e-01 -7.90982783e-01 -6.31970704e-01
1.35704517e-01 -1.01473486e+00 3.69640231e-01 2.35220091e-03
1.17670369e+00 -1.82015717e+00 2.27926433e-01 -2.02207312e-01
2.45688438e-01 5.48392355e-01 -2.83015281e-01 1.17710173e+00
3.32667589e-01 -9.72058773e-02 4.47230488e-01 -3.93495828e-01
1.41008720e-01 9.57773924e-02 -7.43106782e-01 1.79484412e-01
-4.30109650e-01 7.71181285e-01 -9.79373932e-01 1.60295188e-01
4.14830953e-01 7.60768712e-01 -8.68031025e-01 7.75989771e-01
-3.25051546e-01 2.93672770e-01 -7.48265326e-01 1.02273710e-02
4.58327830e-01 -1.82539627e-01 3.44627917e-01 3.87128033e-02
-4.53494668e-01 1.05835629e+00 -8.78885686e-01 1.44068480e+00
-8.81549120e-01 6.28520966e-01 -6.43469840e-02 -9.81472373e-01
1.17883897e+00 4.79970694e-01 5.72677314e-01 -5.81253052e-01
2.38651857e-01 -2.48774648e-01 1.40864253e-02 -5.17225683e-01
1.12806821e+00 1.39401808e-01 -2.50690384e-03 9.18022275e-01
1.01309076e-01 1.99468240e-01 -2.09246814e-01 4.43045616e-01
8.65016162e-01 -1.82438910e-01 2.77888387e-01 -1.83588430e-01
8.92178893e-01 -3.64576757e-01 2.36089930e-01 9.78779316e-01
1.12333097e-01 4.83822674e-01 -1.07099637e-01 -6.04291260e-01
-2.22041190e-01 -6.86795890e-01 2.25793850e-02 2.02242589e+00
-2.26577949e-02 -5.43277800e-01 -2.18664423e-01 -7.40221262e-01
-2.57947952e-01 6.85116529e-01 -6.45631552e-01 7.59992450e-02
-4.23859924e-01 -7.48629510e-01 7.32670948e-02 6.56863511e-01
4.69793119e-02 -1.62162387e+00 2.94808120e-01 1.94987118e-01
-1.58970490e-01 -6.22813761e-01 -3.93547028e-01 -3.37483764e-01
-7.46164262e-01 -8.21198165e-01 -7.79705226e-01 -7.94643760e-01
1.04165152e-01 9.53417718e-01 1.17837024e+00 2.87624061e-01
4.33240116e-01 7.03146279e-01 -1.09314251e+00 -1.08724967e-01
-2.16380954e-01 4.32245702e-01 2.10972652e-01 -1.14526615e-01
5.39240241e-01 -5.76070070e-01 -7.81397521e-01 4.80359882e-01
-3.64236593e-01 -2.87442178e-01 3.36810380e-01 5.20594299e-01
-1.09941781e-01 -3.59186172e-01 1.09548533e+00 -1.57158315e+00
1.37288654e+00 -9.48267043e-01 -2.29084566e-01 1.81114167e-01
-6.19709611e-01 -4.93392944e-01 8.70538652e-01 -4.84777808e-01
-1.61367989e+00 -5.51934898e-01 -8.95938516e-01 3.26265275e-01
-2.87358731e-01 7.95375347e-01 1.44991934e-01 6.70464560e-02
7.90867329e-01 5.55370301e-02 -4.43105638e-01 -7.99592078e-01
9.60544884e-01 1.32154131e+00 1.58030994e-03 -5.83530545e-01
7.40971267e-02 1.72418669e-01 -1.06220806e+00 -8.62459004e-01
-1.05033755e+00 -8.68289292e-01 -3.68223757e-01 -3.90751034e-01
4.59044844e-01 -8.85133386e-01 -1.04156649e+00 8.81547108e-02
-1.04860103e+00 -2.62999862e-01 -6.43830225e-02 7.00291574e-01
-3.37144107e-01 2.49380574e-01 -1.09914339e+00 -9.84772205e-01
-9.42593277e-01 -1.16273069e+00 4.27562267e-01 6.30883396e-01
-2.45065346e-01 -1.20606971e+00 6.28370881e-01 6.50220454e-01
1.03266394e+00 -6.87364578e-01 5.30907869e-01 -1.34304094e+00
-2.48048753e-01 -8.54748189e-02 5.32176532e-02 2.22949460e-01
4.85053807e-02 9.97398570e-02 -1.04264343e+00 -1.58611283e-01
-1.56059027e-01 -1.12377524e-01 6.60611928e-01 4.20827240e-01
1.10165060e+00 -6.14799522e-02 -4.69372630e-01 6.72047585e-02
9.92221177e-01 7.29494452e-01 6.20691955e-01 -2.69118380e-02
4.94676799e-01 6.56177342e-01 5.93024373e-01 3.18112880e-01
8.74434292e-01 3.56084764e-01 1.84624776e-01 2.48222277e-01
1.88313350e-01 5.75818075e-03 6.59233212e-01 1.64753425e+00
-5.20878494e-01 -7.90275335e-01 -3.53129596e-01 2.63549954e-01
-2.13045883e+00 -9.95432138e-01 -9.88098383e-02 1.71525407e+00
4.28867459e-01 -4.36039567e-01 3.19198549e-01 -4.51656044e-01
5.97719193e-01 2.83281326e-01 -3.87787819e-01 -9.35724616e-01
2.25143865e-01 4.15058553e-01 -3.15859795e-01 5.24895310e-01
-9.19258356e-01 1.07962108e+00 7.59717274e+00 4.40442264e-01
-8.81948233e-01 7.77148008e-02 1.72438234e-01 -2.70679910e-02
-3.46701652e-01 -2.96191722e-01 -9.13794577e-01 1.13014966e-01
1.29698288e+00 -1.86684176e-01 6.82060480e-01 1.31944025e+00
1.10584609e-01 3.80088001e-01 -1.00855911e+00 6.69373930e-01
3.52221638e-01 -1.74851179e+00 7.29232207e-02 8.03335011e-02
9.66520309e-01 6.63562596e-01 2.40188122e-01 9.03794110e-01
1.18892527e+00 -8.60178173e-01 -1.87650427e-01 8.60873997e-01
6.05267696e-02 -7.66846180e-01 9.88127708e-01 1.67216972e-01
-9.54354703e-01 -4.87367690e-01 -9.44411874e-01 -4.46780741e-01
5.40688336e-02 3.77340257e-01 -5.03163278e-01 4.69297528e-01
7.62424648e-01 1.29763126e+00 9.84806791e-02 1.19574666e+00
-1.07419670e-01 6.38935626e-01 2.57180065e-01 -7.02935398e-01
2.13686824e-01 -5.75392485e-01 2.43689895e-01 1.40383196e+00
3.06094468e-01 6.14704132e-01 3.13935906e-01 3.75975072e-01
-1.97256982e-01 5.04557490e-01 -3.91863734e-01 -1.33351879e-02
5.26872516e-01 1.50136638e+00 -2.97620416e-01 -2.08216444e-01
-9.08912897e-01 4.36215758e-01 3.73335749e-01 1.60015881e-01
-4.55118418e-01 -1.55356184e-01 1.01718950e+00 -2.49261171e-01
3.40839565e-01 -4.83454987e-02 1.89273879e-01 -1.25864697e+00
-8.79283667e-01 -1.14231646e+00 4.65401620e-01 -5.61699808e-01
-1.92509449e+00 9.79003549e-01 -4.13828135e-01 -1.11057651e+00
-1.45905361e-01 -4.87808138e-01 -1.03656757e+00 5.43039739e-01
-1.27016401e+00 -7.68325806e-01 5.02526015e-02 6.60849273e-01
1.09603214e+00 -7.61405528e-01 1.20268035e+00 5.51783681e-01
-4.72655565e-01 6.84767842e-01 6.21096432e-01 -3.87910791e-02
8.10022712e-01 -1.30234301e+00 6.56026781e-01 1.02712318e-01
2.39707440e-01 1.15359318e+00 4.76676792e-01 -1.57920256e-01
-1.47849178e+00 -8.15016925e-01 9.62276638e-01 -4.22198027e-01
6.42643392e-01 -3.81043464e-01 -6.98217332e-01 5.80731332e-01
7.46425986e-01 -5.27628958e-01 1.40364599e+00 1.07034457e+00
-7.58919120e-02 -1.76119544e-02 -9.57979202e-01 7.72602081e-01
9.78154600e-01 -5.50272048e-01 -9.13299382e-01 4.76799935e-01
8.79274011e-01 -1.76686615e-01 -9.99507487e-01 -1.31566748e-01
8.70818257e-01 -9.09142435e-01 1.08409929e+00 -9.18511808e-01
4.65282023e-01 9.31166708e-02 -1.94368273e-01 -1.74203527e+00
-8.01917076e-01 -6.61020696e-01 -1.27198368e-01 9.71306384e-01
4.34062272e-01 -3.61448467e-01 6.64763689e-01 5.25868535e-01
-5.78718483e-01 -8.10919046e-01 -2.02427372e-01 2.27410465e-01
4.39627171e-01 -5.73866189e-01 6.07268035e-01 1.06628394e+00
6.86279118e-01 8.51630807e-01 -6.94714487e-01 -2.74670541e-01
-2.31355488e-01 4.63400155e-01 7.98861921e-01 -1.72833347e+00
-1.69667825e-01 -4.19948936e-01 3.52899313e-01 -1.71646678e+00
1.37230858e-01 -8.30837250e-01 4.94304262e-02 -1.91897023e+00
-8.10088739e-02 -5.08968174e-01 -6.52417600e-01 3.67118721e-03
1.19323395e-01 -3.33308019e-02 1.40991837e-01 -4.87652197e-02
-1.05094779e+00 7.02259421e-01 1.15725839e+00 -1.06810838e-01
-4.65619743e-01 8.08968723e-01 -1.48035204e+00 7.02303410e-01
7.70404339e-01 -1.58743829e-01 -7.60735154e-01 -3.17028701e-01
7.15888083e-01 1.90334707e-01 -7.08314419e-01 -4.04682100e-01
4.49809819e-01 -3.00836802e-01 -2.22238861e-02 -7.74196863e-01
3.80010128e-01 -5.38662791e-01 -3.04961979e-01 -2.70364195e-01
-1.11641300e+00 4.28718962e-02 -1.70463353e-01 6.75050795e-01
5.62953167e-02 -2.08316837e-02 4.22118098e-01 -2.01460898e-01
-6.00801766e-01 4.73926485e-01 -1.40216279e+00 -3.54276776e-01
3.97630215e-01 4.45853651e-01 -4.82920527e-01 -7.19795704e-01
-9.83483255e-01 3.47683758e-01 -3.47079009e-01 1.20587432e+00
8.38330567e-01 -9.99674559e-01 -6.68644488e-01 -2.54141361e-01
-3.12518552e-02 -5.35260558e-01 5.93352139e-01 4.54258233e-01
4.01057974e-02 6.01060867e-01 -6.93168864e-02 7.54444301e-02
-1.17031920e+00 6.47897601e-01 4.14263278e-01 -2.55142897e-01
-7.51352668e-01 1.00187504e+00 1.77723348e-01 -1.11015546e+00
5.28297007e-01 -1.55525416e-01 -1.65556753e+00 1.67452931e-01
1.08108079e+00 5.12949526e-01 1.28536254e-01 -5.47863662e-01
-1.92273557e-01 7.28351697e-02 -7.49300003e-01 2.86808670e-01
1.74257958e+00 -3.72623891e-01 -1.26353711e-01 3.72281909e-01
8.97085607e-01 5.56205958e-02 -5.39337993e-01 -6.80428088e-01
-1.67196929e-01 -1.83216408e-01 2.42740273e-01 -1.05472052e+00
-1.37896216e+00 8.80625606e-01 4.44306374e-01 4.99955863e-01
6.72605395e-01 -1.58545464e-01 1.05230403e+00 9.25616860e-01
2.10643128e-01 -1.18863738e+00 -8.74996930e-03 1.09027994e+00
9.55551028e-01 -1.06536400e+00 -3.94888856e-02 -5.47503419e-02
-9.57372427e-01 1.20975482e+00 5.10050416e-01 -7.16366053e-01
1.41977155e+00 3.30130979e-02 2.04520747e-01 -2.75965959e-01
-1.25881934e+00 -2.14145616e-01 4.27479297e-01 7.84637451e-01
1.18216312e+00 1.38775513e-01 -5.67353666e-01 1.30030560e+00
-4.09569412e-01 -4.34360281e-03 6.68268561e-01 4.56480354e-01
-5.86888909e-01 -1.42733121e+00 1.98677152e-01 5.91540277e-01
-3.49735856e-01 -3.34693193e-01 -6.95694327e-01 1.14041068e-01
-2.21417457e-01 1.77869344e+00 -2.69767046e-02 -9.87853706e-01
3.84680301e-01 -3.01008165e-01 -3.12300660e-02 -9.33494508e-01
-1.24354303e+00 -1.13934018e-01 3.10340643e-01 -3.57213527e-01
-6.49546564e-01 -4.09060061e-01 -9.80038047e-01 -6.35837972e-01
-5.85821867e-01 9.09215569e-01 3.85962099e-01 1.31145537e+00
4.94941026e-01 5.25969326e-01 9.00883436e-01 -9.09196496e-01
-1.60787642e-01 -1.14525652e+00 -4.99710381e-01 -4.08475190e-01
1.25111401e-01 -6.63521171e-01 -1.46105736e-01 -4.43324387e-01] | [10.212749481201172, 5.8230133056640625] |
4db6d474-81e4-4b1a-a325-91b55c6bd1fb | one-shot-learning-for-channel-estimation-in | 2306.05759 | null | https://arxiv.org/abs/2306.05759v1 | https://arxiv.org/pdf/2306.05759v1.pdf | One-shot Learning for Channel Estimation in Massive MIMO Systems | In conventional supervised deep learning based channel estimation algorithms, a large number of training samples are required for offline training. However, in practical communication systems, it is difficult to obtain channel samples for every signal-to-noise ratio (SNR). Furthermore, the generalization ability of these deep neural networks (DNN) is typically poor. In this work, we propose a one-shot self-supervised learning framework for channel estimation in multi-input multi-output (MIMO) systems. The required number of samples for offline training is small and our approach can be directly deployed to adapt to variable channels. Our framework consists of a traditional channel estimation module and a denoising module. The denoising module is designed based on the one-shot learning method Self2Self and employs Bernoulli sampling to generate training labels. Besides,we further utilize a blind spot strategy and dropout technique to avoid overfitting. Simulation results show that the performance of the proposed one-shot self-supervised learning method is very close to the supervised learning approach while obtaining improved generalization ability for different channel environments. | ['Yonina C. Eldar', 'Yunlong Cai', 'Qiyu Hu', 'Kai Kang'] | 2023-06-09 | null | null | null | null | ['one-shot-learning'] | ['methodology'] | [ 1.56256452e-01 -2.61829495e-01 -1.22873280e-02 -3.72960120e-01
-9.32051957e-01 -2.82959454e-02 3.03074326e-02 -1.49711877e-01
-5.29574513e-01 9.11508739e-01 -3.12313467e-01 -5.30781686e-01
6.49252981e-02 -7.92768180e-01 -6.75591111e-01 -1.00850022e+00
-2.27531970e-01 -1.95373371e-01 -7.01777935e-02 3.17810923e-02
-3.38814706e-02 1.96502179e-01 -8.48421395e-01 -3.67717713e-01
8.46265078e-01 1.12222707e+00 3.35293263e-01 9.47225630e-01
3.23072895e-02 6.44500613e-01 -6.39873385e-01 -1.25263929e-01
1.91155121e-01 -8.94074440e-01 -4.18989547e-02 2.83181489e-01
-2.35541046e-01 -6.71759069e-01 -8.80208194e-01 1.25570440e+00
1.08666658e+00 -4.45792675e-02 4.71268624e-01 -1.04167807e+00
-2.67298341e-01 6.96583509e-01 -4.45908308e-01 1.24177054e-01
-1.75229788e-01 -7.65007660e-02 4.69337463e-01 -7.61260152e-01
-1.20466493e-01 1.03538942e+00 9.11825836e-01 4.86202627e-01
-9.34728980e-01 -1.14477408e+00 -2.78989132e-03 2.25197807e-01
-1.52053428e+00 -8.14592719e-01 7.25736499e-01 -3.11033651e-02
2.11787879e-01 -3.71045381e-01 4.68654007e-01 1.04288018e+00
6.10687435e-02 8.73210788e-01 1.06423748e+00 -5.14545441e-01
7.82441735e-01 4.21915919e-01 -1.16145909e-01 4.86957103e-01
9.42442268e-02 1.64127514e-01 -4.80445325e-01 -1.75012469e-01
1.11838520e+00 2.73508698e-01 -2.92561144e-01 -2.26911321e-01
-8.68344724e-01 7.84034193e-01 3.00323695e-01 9.24989954e-02
-1.50833637e-01 4.37472433e-01 3.24633807e-01 8.40669036e-01
1.43541276e-01 -1.83732584e-01 -3.72509658e-01 -1.64019272e-01
-9.01445985e-01 -2.50560135e-01 1.00348544e+00 1.22588277e+00
8.26884747e-01 5.95380366e-01 -3.66327651e-02 7.36536682e-01
4.51123476e-01 6.57468259e-01 3.85873556e-01 -6.07938290e-01
5.21886885e-01 -2.55164862e-01 8.39644205e-03 -8.48961890e-01
-2.65895873e-01 -1.11190712e+00 -1.43151343e+00 -2.79141683e-02
2.57421404e-01 -9.83603597e-01 -8.29432905e-01 1.47339010e+00
-6.76469281e-02 5.19786477e-01 5.21330476e-01 8.34631920e-01
5.14002144e-01 6.59726739e-01 -1.28621638e-01 -5.05028248e-01
5.54166436e-01 -9.24085498e-01 -9.01729941e-01 -2.51290768e-01
7.14060664e-01 -5.18076420e-01 6.40572846e-01 5.04304528e-01
-7.23364770e-01 -4.31571424e-01 -1.41181314e+00 6.98788643e-01
-9.66971833e-03 2.34211996e-01 5.52883327e-01 1.21978617e+00
-7.10364282e-01 4.96428907e-01 -5.41605651e-01 -5.17904200e-02
5.76493084e-01 4.44536388e-01 -9.81245860e-02 -2.73475140e-01
-1.32553148e+00 4.52676304e-02 4.64806557e-01 4.79385257e-01
-1.32439208e+00 2.17810590e-02 -8.26248646e-01 1.70023724e-01
4.13000405e-01 -2.00398058e-01 1.28515887e+00 -1.25078094e+00
-1.80619323e+00 -2.61822671e-01 -2.20760599e-01 -6.36035264e-01
3.89299333e-01 9.68688652e-02 -5.98031461e-01 1.46071509e-01
-2.67687052e-01 -9.87297948e-03 1.35465205e+00 -1.27948213e+00
-4.88274157e-01 -2.06942603e-01 -2.03344107e-01 2.97032654e-01
-6.02686763e-01 -2.85328150e-01 -4.82024610e-01 -8.14241230e-01
3.89842153e-01 -6.24294519e-01 -6.37506247e-01 -2.37929061e-01
-3.83290321e-01 7.03206837e-01 8.02966774e-01 -6.17253006e-01
1.30547500e+00 -2.37276459e+00 -4.31198478e-01 7.40791261e-01
1.53672963e-01 2.40936607e-01 -1.54253179e-02 3.97376776e-01
2.07986623e-01 -1.73096493e-01 -2.19234198e-01 -3.19849879e-01
-1.72056273e-01 1.61581904e-01 5.58546633e-02 5.94398201e-01
-2.49495566e-01 4.29875702e-01 -9.73238468e-01 -2.94401377e-01
8.55205879e-02 3.96328926e-01 -4.33896452e-01 2.75652528e-01
3.12538475e-01 7.24271774e-01 -3.53007674e-01 6.67666376e-01
8.10868859e-01 -2.49318838e-01 3.46927851e-01 5.14244940e-03
3.14151883e-01 -6.56414628e-02 -1.44673407e+00 1.42924964e+00
-9.31670189e-01 6.73440635e-01 3.06507587e-01 -1.15822232e+00
9.42144394e-01 7.07400441e-01 2.85522081e-02 -7.80983746e-01
5.46296656e-01 2.67296523e-01 1.57246292e-01 -3.87873203e-01
-1.21446839e-03 -1.56673595e-01 -4.34005484e-02 3.93905044e-01
2.89853841e-01 4.69453543e-01 -3.03606361e-01 2.54040003e-01
1.05966806e+00 -4.70568299e-01 4.09467936e-01 1.96438611e-01
4.98690546e-01 -4.89668041e-01 7.94268608e-01 1.16783011e+00
-2.31283560e-01 3.11708272e-01 3.87632966e-01 1.41038179e-01
-9.23521101e-01 -8.73131275e-01 4.07201005e-03 9.14962173e-01
4.06893343e-01 -3.80144315e-03 -7.68040597e-01 -2.94322103e-01
-1.25404671e-01 4.16215479e-01 -1.37439832e-01 -2.10847527e-01
1.72625259e-02 -9.79520679e-01 7.88114011e-01 5.54162562e-01
9.06166911e-01 -6.51404858e-01 7.88062997e-03 5.81181824e-01
2.67095417e-01 -9.44046497e-01 -9.12962258e-02 5.63857317e-01
-9.07316208e-01 -6.56071901e-01 -1.06411612e+00 -1.01125753e+00
8.69966090e-01 4.01601642e-01 3.43517184e-01 -1.71438500e-01
9.03835297e-02 2.28633583e-01 -6.42515779e-01 -2.86117613e-01
-3.57773572e-01 -9.34667811e-02 2.04532206e-01 4.08656925e-01
2.86708146e-01 -6.41409457e-01 -6.08057916e-01 3.05423081e-01
-6.24377489e-01 -4.51087356e-01 1.13670254e+00 1.25627100e+00
2.97215968e-01 6.93951726e-01 1.02498102e+00 -8.88988197e-01
7.10606277e-01 -5.57203948e-01 -6.29478514e-01 -5.04433885e-02
-5.91500401e-01 -1.29649699e-01 1.02771115e+00 -4.86466229e-01
-1.01548958e+00 2.28977114e-01 -2.60737896e-01 -2.21253306e-01
-6.62835911e-02 5.44193864e-01 -4.91108567e-01 -6.15139723e-01
4.00854439e-01 3.99507761e-01 1.71095729e-01 -3.92884761e-01
1.19883470e-01 1.37463236e+00 3.33416224e-01 -8.80200416e-02
9.74936545e-01 1.19994141e-01 -2.06863299e-01 -1.01002109e+00
-5.25562108e-01 -4.36392188e-01 -3.53288174e-01 -1.87404573e-01
2.20908657e-01 -1.43053722e+00 -6.37616813e-01 8.96037161e-01
-6.76117778e-01 -5.44708014e-01 3.68213713e-01 8.79276037e-01
-2.28800327e-01 4.62282270e-01 -7.29122281e-01 -1.14520383e+00
-3.66815954e-01 -1.05641615e+00 5.94287395e-01 3.27190876e-01
4.15329009e-01 -1.16672719e+00 -4.72211242e-01 -1.61723644e-01
6.22980118e-01 -2.50402987e-01 5.28638840e-01 -6.46694899e-01
-5.78332424e-01 -5.71646392e-01 -2.39282057e-01 6.32914245e-01
7.90405646e-02 -7.51287758e-01 -1.08752394e+00 -6.84281766e-01
1.94036946e-01 -2.61487752e-01 7.90518939e-01 5.07377446e-01
1.35281503e+00 -3.41969997e-01 -9.84876454e-02 7.23875403e-01
1.63108432e+00 2.95789301e-01 7.89236426e-01 1.28629655e-01
6.35338485e-01 -4.59574312e-02 5.55651009e-01 7.75166571e-01
1.94512993e-01 2.01944560e-01 2.27030978e-01 -2.12554038e-01
3.13005000e-01 -1.07456401e-01 3.49493742e-01 9.68185902e-01
4.79612827e-01 -2.04195544e-01 -5.83533525e-01 3.78760219e-01
-1.69721639e+00 -8.18169951e-01 -3.78908329e-02 2.59845734e+00
7.82731295e-01 3.65312517e-01 -1.47310793e-01 6.38261676e-01
7.12950945e-01 -7.78056234e-02 -6.87841833e-01 -1.39772566e-02
-2.32086498e-02 7.01543987e-02 1.07925081e+00 4.00813639e-01
-1.19632614e+00 6.68368757e-01 5.72984409e+00 9.68617976e-01
-1.21930695e+00 3.38313431e-01 6.71407223e-01 1.77775085e-01
2.15305954e-01 7.10026845e-02 -6.98687911e-01 5.87638676e-01
1.01448894e+00 1.15350462e-01 5.16434848e-01 7.45534718e-01
4.01224285e-01 -2.53079772e-01 -7.64723539e-01 1.57309341e+00
1.77232083e-02 -9.12164867e-01 -4.08937603e-01 -2.49492705e-01
6.33032560e-01 -2.73693919e-01 -5.12451958e-03 3.76116246e-01
-4.63559516e-02 -8.83074820e-01 3.07718962e-01 4.72689360e-01
8.20275366e-01 -9.07421470e-01 1.02280807e+00 4.49037373e-01
-1.01583123e+00 -6.07238352e-01 -4.28831875e-01 -9.73142534e-02
1.43199816e-01 1.05874467e+00 -7.20514417e-01 3.49808753e-01
4.77078587e-01 5.97745061e-01 -2.26542398e-01 1.59020221e+00
-1.06033631e-01 9.88217056e-01 6.09146943e-03 -1.97584480e-01
2.41841346e-01 -6.25632256e-02 4.16353405e-01 1.03103113e+00
4.67245877e-01 -8.33144337e-02 3.38303655e-01 1.62405483e-02
-1.80000499e-01 7.97949582e-02 -4.55475271e-01 -1.37872081e-02
9.48078454e-01 9.14252639e-01 -4.29529458e-01 -3.30457151e-01
-4.90947574e-01 1.20388865e+00 -3.01083982e-01 8.31772804e-01
-5.42426229e-01 -1.09336841e+00 7.73061290e-02 -2.99758404e-01
5.11793315e-01 -3.51060390e-01 -3.19482207e-01 -9.03201699e-01
-1.80425316e-01 -8.18199933e-01 -1.90541759e-01 -3.73215586e-01
-1.17224443e+00 2.46877640e-01 -6.82842493e-01 -1.69407892e+00
-1.01152778e-01 -3.07572126e-01 -5.80559075e-01 8.14518154e-01
-1.69435799e+00 -5.78878939e-01 -3.86990905e-01 7.13351846e-01
3.47511172e-01 -6.09474003e-01 8.78422558e-01 6.89860284e-01
-7.75476456e-01 1.20476353e+00 8.45220745e-01 5.37870347e-01
7.17549443e-01 -1.05125725e+00 7.01515377e-02 9.40990925e-01
-4.16823834e-01 6.03576422e-01 5.65271437e-01 -4.57067668e-01
-1.53917277e+00 -1.37882972e+00 1.56464547e-01 6.82746291e-01
5.33641338e-01 -4.16767210e-01 -7.51881897e-01 3.67394239e-01
6.91666231e-02 1.06136672e-01 9.65665698e-01 -4.20164019e-02
-4.19169292e-02 -3.92377257e-01 -1.06934190e+00 4.48367059e-01
6.16872430e-01 -5.01706064e-01 2.44504631e-01 2.73172438e-01
2.36752108e-01 -4.05179232e-01 -7.36635327e-01 -3.27698626e-02
5.32242954e-01 -7.49856293e-01 5.13754964e-01 7.51579925e-02
-1.66476965e-01 -2.43668422e-01 -2.84363806e-01 -1.48312545e+00
-9.09926817e-02 -9.97530341e-01 -3.91361803e-01 1.01256108e+00
5.77106535e-01 -7.31104374e-01 9.18827891e-01 -1.54787879e-02
4.04523872e-02 -4.93211389e-01 -7.40477383e-01 -9.33538318e-01
-3.50991249e-01 -3.69804680e-01 3.05400431e-01 6.30357563e-01
-1.38400093e-01 4.95748818e-01 -8.51836741e-01 5.01750410e-01
1.14763963e+00 -4.61800665e-01 9.61521566e-01 -1.03963327e+00
-8.36402833e-01 1.76010564e-01 -4.55027491e-01 -1.45705771e+00
-1.81271583e-01 -4.56279844e-01 3.91891718e-01 -1.21195734e+00
-1.88042130e-02 -7.27047086e-01 -4.73626643e-01 1.44263089e-01
-6.65814504e-02 7.61518329e-02 -1.98066428e-01 1.66927174e-01
-6.73855901e-01 7.79080033e-01 8.90206635e-01 -2.37019025e-02
-4.99138415e-01 5.85576475e-01 -7.36856461e-01 4.89539564e-01
1.18073499e+00 -5.04029632e-01 -4.19073611e-01 -3.64137739e-01
-8.32768008e-02 4.02962834e-01 2.35262707e-01 -1.44099021e+00
4.03195560e-01 2.97376901e-01 6.04732156e-01 -3.36248159e-01
2.47607559e-01 -9.36045408e-01 -2.38164306e-01 4.87970352e-01
-1.52274728e-01 -6.74748540e-01 -1.78358182e-01 1.23133361e+00
-2.73410320e-01 -2.95019388e-01 6.97786689e-01 -4.97111566e-02
-9.39402997e-01 3.97782028e-01 -6.26586199e-01 -2.81212360e-01
8.68057907e-01 -2.81436086e-01 1.82102442e-01 -9.54641879e-01
-7.99268186e-01 3.64965975e-01 1.32476166e-01 -1.31125852e-01
8.58395100e-01 -1.37941325e+00 -3.01219732e-01 3.19790125e-01
4.85621318e-02 -1.54802978e-01 3.45910788e-01 9.51857924e-01
-3.47140819e-01 2.31852576e-01 1.09260365e-01 -4.00594980e-01
-9.20622766e-01 7.09848776e-02 3.74708325e-01 2.89740920e-01
-4.59833860e-01 8.53785992e-01 -4.59945261e-01 -3.88695449e-01
9.49661195e-01 2.59281069e-01 -6.64886609e-02 -2.67576814e-01
7.39375591e-01 2.90791780e-01 -5.48976995e-02 -1.69459671e-01
5.41855879e-02 5.18502258e-02 -1.08561777e-01 -2.06672996e-01
9.82027888e-01 -4.38067824e-01 9.38379169e-02 5.30550301e-01
1.44940102e+00 -1.30427793e-01 -1.39801753e+00 -6.52873933e-01
-3.92777741e-01 -4.20457214e-01 4.70975131e-01 -3.94885540e-01
-1.10450315e+00 1.00222433e+00 1.14969766e+00 4.67389869e-03
1.30238140e+00 -5.93573272e-01 1.12190998e+00 8.11034739e-01
5.37364721e-01 -1.37616336e+00 -5.91791049e-02 5.06869137e-01
-1.70632340e-02 -1.44887948e+00 -2.98293680e-01 -1.13592118e-01
-4.82673615e-01 1.17498231e+00 2.54435509e-01 -5.56442514e-02
1.04552364e+00 3.94437402e-01 2.96234727e-01 3.38904917e-01
-3.64950895e-01 -2.26880923e-01 -4.06690150e-01 7.64161944e-01
2.27839112e-01 7.13605136e-02 -2.25941855e-02 6.46151841e-01
1.97568789e-01 -2.14814786e-02 8.17952931e-01 1.02531362e+00
-8.12625527e-01 -1.05808210e+00 -6.31757021e-01 7.69648433e-01
-4.04929012e-01 -3.43289435e-01 1.18136913e-01 2.42003605e-01
-3.15877587e-01 1.32601917e+00 -1.07438855e-01 -6.42783284e-01
-6.77001700e-02 -3.90928179e-01 2.16299891e-01 -5.95395684e-01
1.07211396e-01 3.85313302e-01 -3.78061309e-02 -2.25777835e-01
-2.84846842e-01 -3.49241763e-01 -1.30071187e+00 -1.73802331e-01
-5.61697245e-01 2.47847587e-01 7.43978143e-01 1.07147276e+00
1.43490702e-01 7.36197472e-01 1.24090707e+00 -7.45102763e-01
-7.49491930e-01 -9.83431518e-01 -1.12157118e+00 -2.35468075e-01
6.38254106e-01 -4.84895229e-01 -2.95661002e-01 -6.66265786e-02] | [6.395370960235596, 1.4813352823257446] |
dda1ef2b-27ce-4eef-998a-90302c28cc89 | sharp-shape-regularized-multidimensional | 2306.00554 | null | https://arxiv.org/abs/2306.00554v1 | https://arxiv.org/pdf/2306.00554v1.pdf | ShaRP: Shape-Regularized Multidimensional Projections | Projections, or dimensionality reduction methods, are techniques of choice for the visual exploration of high-dimensional data. Many such techniques exist, each one of them having a distinct visual signature - i.e., a recognizable way to arrange points in the resulting scatterplot. Such signatures are implicit consequences of algorithm design, such as whether the method focuses on local vs global data pattern preservation; optimization techniques; and hyperparameter settings. We present a novel projection technique - ShaRP - that provides users explicit control over the visual signature of the created scatterplot, which can cater better to interactive visualization scenarios. ShaRP scales well with dimensionality and dataset size, generically handles any quantitative dataset, and provides this extended functionality of controlling projection shapes at a small, user-controllable cost in terms of quality metrics. | ['Michael Behrisch', 'Alexandru Telea', 'Alister Machado'] | 2023-06-01 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [-2.20442861e-01 -2.03317434e-01 -1.04876637e-01 -2.66269833e-01
-1.34772509e-01 -1.06307256e+00 7.06945300e-01 3.40187848e-01
-4.56571952e-02 1.98808461e-01 4.83821154e-01 -6.25943840e-01
-6.26336932e-01 -6.28629148e-01 -4.74270545e-02 -7.44743943e-01
-4.08847153e-01 6.42603695e-01 1.89352915e-01 -9.44692791e-02
5.81946015e-01 8.76717865e-01 -1.73894715e+00 2.75599301e-01
6.26246333e-01 6.80814087e-01 -6.57629818e-02 7.68158317e-01
-2.99718052e-01 -1.73893534e-02 -5.30132115e-01 2.87600365e-02
4.39894497e-01 -3.11290175e-01 -2.32051373e-01 2.62662284e-02
2.66745329e-01 1.10541679e-01 2.09700555e-01 9.38927948e-01
3.33013684e-01 -2.41202161e-01 7.68943727e-01 -1.25487578e+00
-6.03544354e-01 1.26357466e-01 -6.90155983e-01 1.68049425e-01
5.17332733e-01 5.03140330e-01 6.86532855e-01 -1.03009152e+00
9.63649452e-01 1.34058309e+00 4.09786165e-01 3.16098109e-02
-2.06358671e+00 -2.86098123e-01 -1.58179536e-01 -2.43514150e-01
-1.25275111e+00 -3.11520994e-01 8.30296159e-01 -8.91987324e-01
4.74076450e-01 1.14226365e+00 8.12713265e-01 6.60983205e-01
8.14069659e-02 2.48249531e-01 1.52582383e+00 -3.77936691e-01
6.06108785e-01 5.18351853e-01 4.85799879e-01 2.44488999e-01
4.18777764e-01 5.52837141e-02 -4.77099419e-01 -4.78227317e-01
8.11143517e-01 8.23425427e-02 -3.77176017e-01 -1.46041656e+00
-1.25857735e+00 6.12260640e-01 1.39420986e-01 9.27776098e-03
-6.53755516e-02 -3.18252593e-01 4.25037771e-01 4.59662050e-01
9.48706344e-02 9.76754963e-01 -1.78127691e-01 -3.77688646e-01
-6.89833760e-01 5.17308593e-01 6.10371470e-01 7.93126464e-01
7.06570864e-01 -3.72161984e-01 -2.57881939e-01 5.82608402e-01
1.85060337e-01 2.65389264e-01 3.69962424e-01 -6.60785496e-01
3.17564964e-01 1.18978715e+00 2.94752747e-01 -1.14030528e+00
-8.66886437e-01 5.64595172e-03 -5.54761291e-01 1.14283907e+00
5.52201390e-01 1.34914428e-01 -9.38127041e-01 1.43800354e+00
7.58397698e-01 -8.66833866e-01 -2.73427457e-01 1.07289445e+00
4.62178618e-01 4.45681185e-01 7.92556256e-02 -1.23285100e-01
1.48084259e+00 -1.92494169e-01 -6.33098781e-01 1.23390719e-01
6.10444725e-01 -5.13003945e-01 1.91204226e+00 5.32918572e-01
-1.01471615e+00 -1.25059932e-01 -1.14197183e+00 -1.02976620e-01
-7.43315816e-01 1.81374550e-01 4.71233994e-01 7.96676040e-01
-1.04481602e+00 6.62823796e-01 -7.50415981e-01 -3.92635703e-01
1.93366230e-01 3.60912502e-01 -3.93004239e-01 3.43648225e-01
-5.58738291e-01 7.19891548e-01 3.27846676e-01 -2.26375580e-01
-1.46377489e-01 -1.02873921e+00 -3.65728080e-01 3.79636973e-01
5.62245362e-02 -5.97961485e-01 7.69484103e-01 -5.78057945e-01
-1.29986405e+00 9.77963746e-01 1.16143465e-01 1.40194774e-01
8.44280839e-01 -3.67437117e-02 -3.68359447e-01 -1.06993221e-01
-3.17539275e-01 3.70780289e-01 7.65701473e-01 -1.46589696e+00
-1.56618997e-01 -7.91754246e-01 -3.25642735e-01 2.74527013e-01
-4.56367493e-01 1.16829112e-01 -3.70531559e-01 -5.72828531e-01
3.27825040e-01 -8.66308093e-01 -3.57988089e-01 6.51345015e-01
-6.32247925e-01 1.88294619e-01 1.34323239e+00 -3.02069902e-01
1.78039861e+00 -2.46642828e+00 1.87848121e-01 7.48966396e-01
5.14775455e-01 2.62579262e-01 2.47852400e-01 8.49928260e-01
-4.80905920e-01 2.99515665e-01 -1.13841154e-01 -1.96435094e-01
1.65800601e-01 -2.04342678e-01 -2.41729051e-01 5.69009662e-01
2.81004366e-02 5.34101844e-01 -5.92812181e-01 -1.76715210e-01
3.74433517e-01 4.79793638e-01 -2.90595144e-01 1.27426445e-01
4.36335467e-02 4.05537426e-01 -2.97645092e-01 3.50241572e-01
9.87836659e-01 -3.76283556e-01 5.84780812e-01 6.52218983e-02
-5.98746181e-01 1.86913282e-01 -1.75992465e+00 1.43861842e+00
2.07070820e-02 7.46863186e-01 1.30352020e-01 -2.66088963e-01
1.27000976e+00 -1.53965145e-01 1.71334356e-01 -5.91684699e-01
-1.84990257e-01 -4.31449749e-02 -1.37147352e-01 -3.72039676e-01
5.58459759e-01 2.17124462e-01 2.66375721e-01 8.09335530e-01
-5.82194507e-01 1.43865168e-01 3.49334061e-01 1.21519111e-01
9.03886437e-01 -2.68244781e-02 6.07617974e-01 -7.52565742e-01
1.32187307e-01 1.18245877e-01 4.41324860e-02 3.65490377e-01
3.33055913e-01 7.72692680e-01 1.05081904e+00 -7.01282978e-01
-1.32233107e+00 -1.19590664e+00 -5.16432941e-01 1.03193533e+00
7.89269209e-02 -7.38285661e-01 -4.00685459e-01 -4.17232692e-01
3.11731607e-01 5.86195469e-01 -1.01056063e+00 1.28694385e-01
-4.42605585e-01 -5.56129754e-01 -1.45870417e-01 1.74751654e-01
-3.48386228e-01 -9.67594147e-01 -1.36030352e+00 -2.71668732e-01
7.00906277e-01 -1.18576944e-01 -2.49908596e-01 3.23744148e-01
-1.38709080e+00 -1.06356597e+00 -4.73641068e-01 -2.55997300e-01
9.87449825e-01 2.67368585e-01 1.17037821e+00 -1.57165498e-01
-5.14223397e-01 1.22628257e-01 -2.08593816e-01 -1.02584697e-01
-2.67879874e-01 -2.37045929e-01 -1.57351851e-01 -1.83898762e-01
2.27572054e-01 -9.07593071e-01 -7.87165940e-01 5.73017478e-01
-9.40872788e-01 2.33933419e-01 3.04430813e-01 4.45705712e-01
6.27875328e-01 -2.55056322e-01 -1.18156634e-01 -1.11686540e+00
1.13064158e+00 -2.53144175e-01 -6.57269180e-01 3.16647768e-01
-8.82391810e-01 3.61328542e-01 5.79626381e-01 -3.63624036e-01
-5.43549120e-01 -4.13213186e-02 5.72122395e-01 -5.06096244e-01
-3.61142874e-01 2.83084899e-01 -3.98593903e-01 -4.84924428e-02
1.03865099e+00 -1.60753764e-02 2.37817243e-01 -9.29218650e-01
7.20591545e-01 3.53101403e-01 3.41511428e-01 -2.68800080e-01
8.09581578e-01 5.50039589e-01 2.29505748e-01 -5.41132510e-01
2.61528432e-01 -4.09906298e-01 -8.66150677e-01 -1.75852105e-01
4.16210264e-01 -9.49878320e-02 -8.13749790e-01 -1.17326401e-01
-6.66587532e-01 -8.18942413e-02 -4.08955485e-01 1.18353546e-01
-3.89272273e-01 1.32533506e-01 1.72662482e-01 -7.81605899e-01
-1.13587901e-01 -1.14800119e+00 7.74768710e-01 -1.45728886e-03
-6.73168004e-01 -9.09724593e-01 4.34943587e-01 -5.39157987e-01
4.63859886e-01 7.16233730e-01 1.25716281e+00 -4.20485675e-01
-4.54956144e-01 -1.75521493e-01 -3.24874818e-01 -3.16002965e-01
2.42153078e-01 3.59025240e-01 -9.10894871e-01 -4.00696427e-01
-3.24970663e-01 1.79390997e-01 2.84241140e-01 2.37408400e-01
1.28852224e+00 -6.60326362e-01 -5.34523010e-01 7.45763004e-01
1.29729903e+00 2.18339682e-01 6.44170582e-01 4.61243570e-01
5.64257085e-01 7.48569071e-01 3.86634886e-01 6.27612472e-01
-2.01485902e-01 9.34407890e-01 3.28675449e-01 -5.57445884e-01
1.03327688e-02 -2.82630265e-01 -2.91477621e-01 2.97496282e-02
1.44254819e-01 1.04920626e-01 -1.07521129e+00 4.45703343e-02
-1.81977427e+00 -7.37642407e-01 -3.88392270e-01 2.68788171e+00
5.02527833e-01 1.56033665e-01 6.37881398e-01 3.08613867e-01
5.09372175e-01 2.00616315e-01 -5.91500163e-01 -6.53499424e-01
1.34980604e-01 -2.51215160e-01 3.38305950e-01 1.82308137e-01
-7.75407612e-01 1.63302660e-01 7.09173870e+00 4.08954203e-01
-1.27254522e+00 -4.68303770e-01 4.76754487e-01 -3.00172925e-01
-6.22321784e-01 6.18341379e-02 -2.98945427e-01 5.31154811e-01
5.81536770e-01 -2.96530545e-01 3.18440348e-01 1.07213390e+00
6.21866763e-01 -1.26218975e-01 -1.33610201e+00 1.08343709e+00
-4.66924191e-01 -1.56175375e+00 2.29671761e-01 4.16387647e-01
1.61678597e-01 -7.29622960e-01 3.79100591e-01 -4.49434102e-01
1.08383536e-01 -1.10386252e+00 6.25411630e-01 5.35926461e-01
1.13828063e+00 -9.08463657e-01 -1.18197896e-01 2.12413706e-02
-9.08199906e-01 -2.54162490e-01 -3.59473750e-02 2.96326540e-02
8.09204727e-02 5.88738501e-01 -8.02404702e-01 8.16737860e-02
8.62139881e-01 2.69801021e-01 -9.24342275e-01 1.24768889e+00
2.01377064e-01 1.74488500e-01 -3.34207803e-01 -8.50398317e-02
-1.36933178e-01 -5.51303029e-01 8.88561249e-01 1.51171374e+00
2.85745859e-01 -1.05243824e-01 -1.30638540e-01 1.06538332e+00
6.79662883e-01 3.03386509e-01 -6.89572215e-01 -5.01867458e-02
6.58772171e-01 1.26135468e+00 -9.21603560e-01 -2.25409046e-02
1.04656001e-03 4.20190036e-01 2.02397868e-01 4.30427551e-01
-1.67954683e-01 -3.91666472e-01 1.00817037e+00 7.00046659e-01
1.81273952e-01 -4.05872256e-01 -1.07903254e+00 -6.86623096e-01
2.20147580e-01 -1.08611429e+00 5.14212251e-01 -7.63738155e-01
-9.01517689e-01 4.94635016e-01 1.21648356e-01 -1.51860619e+00
-8.67362469e-02 -7.66854465e-01 -7.65239239e-01 1.01006508e+00
-6.96293175e-01 -7.32041478e-01 -3.90795946e-01 3.79546463e-01
-1.02777563e-01 1.04629241e-01 9.16285992e-01 1.26393912e-02
-3.91418636e-01 3.25357735e-01 3.73620689e-01 -5.51668525e-01
5.93067527e-01 -1.61365342e+00 5.30321777e-01 6.77213788e-01
3.76143828e-02 1.03122234e+00 1.20732009e+00 -3.96640956e-01
-1.59751117e+00 -3.06527913e-01 4.55549717e-01 -7.02006161e-01
5.93774676e-01 -7.21825480e-01 -1.24973321e+00 1.97768405e-01
-1.66694120e-01 -2.76072919e-01 9.15492833e-01 6.04467690e-01
-4.39242691e-01 -5.32142036e-02 -1.17868459e+00 8.31534386e-01
8.71589661e-01 -2.24655107e-01 -2.45474502e-01 9.08795446e-02
3.56767267e-01 -4.98722374e-01 -8.06387544e-01 9.49708670e-02
7.27540612e-01 -1.54187870e+00 1.06362474e+00 -7.86542296e-01
1.47457868e-01 -5.82017779e-01 2.29171365e-01 -1.28240240e+00
-6.51495636e-01 -8.00517440e-01 -1.67053565e-01 8.88682246e-01
6.02555394e-01 -5.95456064e-01 8.31177711e-01 1.07015288e+00
1.69511929e-01 -8.16288769e-01 -5.54008663e-01 -6.35451913e-01
-3.51917148e-01 -4.33964357e-02 9.82427776e-01 9.54976857e-01
5.32068014e-01 -3.09408568e-02 -4.66470905e-02 -3.62515450e-02
4.12808597e-01 3.90271813e-01 1.22235274e+00 -1.43891132e+00
-8.50112662e-02 -9.20220375e-01 -6.20314419e-01 -7.30698824e-01
-9.98901546e-01 -4.35745806e-01 -5.36940634e-01 -1.25946772e+00
-5.19286282e-02 -7.80387998e-01 -3.85705531e-02 1.98891714e-01
-9.74251777e-02 -7.20935315e-02 3.23783010e-01 5.33391535e-01
-1.01251960e-01 1.88028738e-01 1.25060833e+00 3.98012161e-01
-8.23981285e-01 -3.01622339e-02 -6.48954153e-01 3.97441685e-01
6.17743373e-01 -5.39658368e-01 -6.59307480e-01 -1.03193320e-01
3.17965686e-01 -1.48427933e-01 3.17293435e-01 -8.13623071e-01
7.48974532e-02 -3.15983623e-01 5.45286298e-01 -5.22855043e-01
3.78763489e-02 -8.79166424e-01 6.48676693e-01 3.80646765e-01
-4.27473247e-01 6.60257161e-01 2.53135741e-01 5.39819717e-01
1.78851247e-01 9.85065848e-02 8.16191494e-01 7.29622766e-02
-4.60732371e-01 -1.03770331e-01 -4.94103357e-02 -1.69384405e-01
1.01134837e+00 -7.82334030e-01 -4.00782019e-01 -1.56779259e-01
-9.16119754e-01 1.77523866e-01 9.99254882e-01 3.90833557e-01
3.50829393e-01 -1.33173645e+00 -2.57551640e-01 5.81806958e-01
3.49966139e-01 -2.49429554e-01 8.95313695e-02 6.06438220e-01
-5.08041739e-01 1.64784491e-01 -3.94918382e-01 -6.89811826e-01
-1.54197538e+00 9.54716325e-01 1.23777509e-01 -1.21456519e-01
-1.15860784e+00 3.41894805e-01 3.76855046e-01 -3.50558877e-01
2.93064982e-01 -3.36986065e-01 -2.65610635e-01 2.49992132e-01
8.71756196e-01 6.71773732e-01 1.22295499e-01 -7.72332326e-02
-3.76389414e-01 6.14122272e-01 1.18598543e-01 -1.88586444e-01
1.29060376e+00 -2.53512025e-01 7.27025140e-03 6.49946094e-01
1.04269767e+00 1.17968775e-01 -1.40734076e+00 1.61802009e-01
2.00504631e-01 -1.04800272e+00 -2.82519639e-01 -1.06549764e+00
-6.88232899e-01 8.28550458e-01 8.60065818e-01 8.65183055e-01
1.01199806e+00 -1.42518416e-01 -2.10227728e-01 5.31139262e-02
5.44508100e-02 -8.19914997e-01 -4.12684828e-02 -6.28037676e-02
1.31918740e+00 -9.25979137e-01 4.36410099e-01 -2.56515831e-01
-7.90932536e-01 1.46048784e+00 3.92167389e-01 -2.87285931e-02
6.61888123e-01 5.00250340e-01 3.07161123e-01 -7.78701842e-01
-5.55049777e-01 3.03438842e-01 5.69218755e-01 8.14146042e-01
3.29636812e-01 2.30782107e-01 -4.07688528e-01 -1.45149883e-02
-4.10876572e-01 -2.90809691e-01 1.73680559e-01 9.73226726e-01
-5.59875369e-01 -9.86818314e-01 -7.26257443e-01 5.55592060e-01
2.61719704e-01 2.54854858e-01 -7.19609141e-01 9.09705520e-01
-2.45632127e-01 3.48282278e-01 2.09233910e-01 -1.45368412e-01
5.93841672e-01 1.51883811e-02 -8.35313648e-02 -3.98906171e-01
-5.29000938e-01 3.90934236e-02 -1.20241307e-01 -8.46122921e-01
2.97128916e-01 -8.22122753e-01 -9.02061164e-01 -3.39261711e-01
-2.75725313e-02 1.35615200e-01 7.77843297e-01 2.47509092e-01
8.57188523e-01 3.09820240e-03 5.72682083e-01 -9.05048490e-01
-4.17138964e-01 -6.82392418e-01 -8.69104981e-01 5.55340052e-01
4.10129845e-01 -7.41410136e-01 -4.31066036e-01 -1.41708329e-01] | [7.993043899536133, 4.55886697769165] |
0ecd2587-8f5a-41ad-88c0-a8bcb8fd7c83 | multi-person-3d-pose-and-shape-estimation-via | 2210.13529 | null | https://arxiv.org/abs/2210.13529v2 | https://arxiv.org/pdf/2210.13529v2.pdf | Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement | Estimating 3D poses and shapes in the form of meshes from monocular RGB images is challenging. Obviously, it is more difficult than estimating 3D poses only in the form of skeletons or heatmaps. When interacting persons are involved, the 3D mesh reconstruction becomes more challenging due to the ambiguity introduced by person-to-person occlusions. To tackle the challenges, we propose a coarse-to-fine pipeline that benefits from 1) inverse kinematics from the occlusion-robust 3D skeleton estimation and 2) Transformer-based relation-aware refinement techniques. In our pipeline, we first obtain occlusion-robust 3D skeletons for multiple persons from an RGB image. Then, we apply inverse kinematics to convert the estimated skeletons to deformable 3D mesh parameters. Finally, we apply the Transformer-based mesh refinement that refines the obtained mesh parameters considering intra- and inter-person relations of 3D meshes. Via extensive experiments, we demonstrate the effectiveness of our method, outperforming state-of-the-arts on 3DPW, MuPoTS and AGORA datasets. | ['Seungryul Baek', 'Mingyu Shin', 'GeonU Kim', 'Muhammad Saqlain', 'Junuk Cha'] | 2022-10-24 | null | null | null | null | ['3d-human-pose-estimation', '3d-multi-person-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-5.78956082e-02 -1.03149503e-01 4.11972463e-01 -3.89579684e-01
-7.84256279e-01 -4.21834528e-01 3.58158052e-01 -5.74590638e-02
-2.42863044e-01 5.20935178e-01 2.57099420e-01 2.83855379e-01
3.70366648e-02 -8.31489444e-01 -6.87774122e-01 -2.97473639e-01
2.90369064e-01 1.14696538e+00 4.32767838e-01 1.02993073e-02
2.20657196e-02 6.49282634e-01 -1.65581024e+00 -1.27375454e-01
6.78663373e-01 8.99636447e-01 -3.51952225e-01 5.41813493e-01
-1.77161783e-01 6.13840111e-02 -3.29016536e-01 -4.76342559e-01
4.42371666e-01 -1.94358170e-01 -6.32529676e-01 5.67746222e-01
6.79217458e-01 -4.60237890e-01 -7.03748390e-02 7.57075429e-01
4.89332825e-01 4.86055464e-02 4.68480617e-01 -1.01664364e+00
7.37943724e-02 -9.80106518e-02 -1.06019020e+00 -4.60985571e-01
9.11849678e-01 8.41444731e-02 6.14913404e-01 -9.94785130e-01
8.29406798e-01 1.75808227e+00 7.91398644e-01 4.28160340e-01
-1.25846720e+00 -5.13180673e-01 2.49943018e-01 -1.01665549e-01
-1.52728188e+00 -2.74430275e-01 8.93363655e-01 -5.91688395e-01
3.22463214e-01 2.64760137e-01 1.22193635e+00 9.45141733e-01
-3.58164907e-01 4.99565840e-01 1.15249324e+00 -1.58897415e-01
6.21571280e-02 -4.64132935e-01 -1.16695717e-01 8.82577360e-01
1.87232733e-01 -2.48716205e-01 -7.23102510e-01 -3.72989178e-01
1.28188467e+00 1.21838070e-01 -5.04812151e-02 -6.69866502e-01
-1.31389952e+00 2.83852428e-01 4.24155712e-01 -2.29916900e-01
-5.75557709e-01 2.00487733e-01 -4.05169316e-02 -2.66562939e-01
6.21094465e-01 -1.34215012e-01 -4.92730916e-01 -2.17361346e-01
-9.30246234e-01 6.18017972e-01 6.67829871e-01 9.85791028e-01
9.08586264e-01 -4.30740386e-01 9.31871682e-02 6.20051801e-01
5.39463818e-01 6.24618053e-01 -1.98047116e-01 -1.08432114e+00
4.77485299e-01 1.06721985e+00 3.43878090e-01 -9.32014465e-01
-5.00584364e-01 -1.38364598e-01 -7.13554025e-01 1.25769779e-01
7.82762706e-01 5.30101508e-02 -9.43889320e-01 1.31914127e+00
1.20616138e+00 1.91237360e-01 -5.15050113e-01 1.26229072e+00
7.62914598e-01 1.47351757e-01 -1.47031564e-02 6.01295009e-02
1.66670370e+00 -7.91811228e-01 -4.78963137e-01 -8.95494670e-02
-6.60143979e-03 -7.17520118e-01 9.20218885e-01 2.64650285e-01
-1.35215759e+00 -5.16569793e-01 -6.38180792e-01 -4.36581761e-01
6.45432174e-02 1.50986046e-01 1.81574568e-01 4.40793604e-01
-4.39726740e-01 5.75125635e-01 -9.78712022e-01 -5.05548120e-01
3.83741856e-01 3.92522514e-01 -5.30899942e-01 -3.52093205e-02
-7.21533120e-01 6.92922831e-01 2.04562731e-02 3.11686248e-01
-5.29989541e-01 -7.30518162e-01 -9.52738941e-01 -3.58573198e-01
5.65553248e-01 -1.22350514e+00 8.85283172e-01 -3.14675629e-01
-1.60226786e+00 1.05228007e+00 -2.47109190e-01 1.22614965e-01
1.09916401e+00 -6.52656257e-01 3.11347127e-01 1.85527071e-01
-1.40131405e-02 4.46912050e-01 8.53744090e-01 -1.36827707e+00
-4.45361912e-01 -8.97639632e-01 1.41897798e-01 5.37215292e-01
5.96881285e-02 -1.77665457e-01 -1.06681883e+00 -3.99407327e-01
6.86044812e-01 -9.85615790e-01 -2.19151154e-01 7.78899610e-01
-6.76280975e-01 -1.37258619e-01 7.94533610e-01 -9.77083385e-01
7.52435386e-01 -2.00497746e+00 7.48757124e-01 3.44966263e-01
3.20483685e-01 -1.16395086e-01 2.21649930e-01 -7.59496018e-02
4.07908916e-01 -1.22001499e-01 -2.97284871e-01 -9.88605022e-01
2.65869945e-01 1.67015567e-01 2.82254666e-01 6.97256386e-01
7.23236278e-02 7.33640373e-01 -8.09166908e-01 -8.69543254e-01
4.33542967e-01 9.48817849e-01 -3.58593464e-01 3.94482255e-01
-2.07170531e-01 9.04025435e-01 -6.88102663e-01 1.07957876e+00
8.10375154e-01 -9.86885875e-02 2.20097192e-02 -5.21884203e-01
-1.55389264e-01 -3.10131405e-02 -1.69695354e+00 2.19778371e+00
-2.69164115e-01 -2.12852627e-01 4.33437556e-01 -4.29832429e-01
8.16394746e-01 2.39541247e-01 5.90798259e-01 -1.02563456e-01
2.56976575e-01 2.65945256e-01 -6.18987322e-01 -3.12883019e-01
1.63178638e-01 2.75002662e-02 -1.07459635e-01 4.13399816e-01
-3.07410151e-01 -4.85778660e-01 2.84068175e-02 -1.45976424e-01
6.74336612e-01 9.72981095e-01 7.12154657e-02 1.30230794e-03
6.26091003e-01 -1.59995377e-01 6.02773607e-01 7.61999935e-02
1.45633176e-01 8.71315837e-01 2.15619087e-01 -7.77787387e-01
-1.06088090e+00 -1.29153955e+00 -6.55038804e-02 5.37253201e-01
3.07623804e-01 -4.86079901e-01 -9.31851029e-01 -5.69077551e-01
2.04414174e-01 5.34058502e-03 -6.29467964e-01 4.20895308e-01
-6.59240305e-01 -4.88181591e-01 2.34751150e-01 3.90992612e-01
6.32866800e-01 -5.34532011e-01 -8.47754061e-01 1.02484189e-01
-3.78466666e-01 -1.34855962e+00 -6.06127679e-01 -6.23397350e-01
-9.90620077e-01 -1.33506441e+00 -9.48834240e-01 -2.63007641e-01
9.10036385e-01 5.10831326e-02 9.45634723e-01 2.98477530e-01
-3.42361808e-01 4.56110567e-01 -1.34168327e-01 -9.39132422e-02
1.02192096e-01 -3.24676782e-02 1.91443488e-01 2.71010786e-01
4.81460541e-02 -8.36440206e-01 -8.76357853e-01 5.48162103e-01
-4.75864649e-01 3.99370879e-01 2.80059010e-01 2.25277066e-01
1.06249213e+00 -8.81422218e-03 -2.11802438e-01 -6.99355066e-01
1.27123222e-01 -3.73292714e-02 -6.01016641e-01 1.70784026e-01
2.11787835e-01 -1.86050594e-01 -2.06761267e-02 -4.68106240e-01
-9.47736263e-01 6.13652229e-01 -4.48797308e-02 -5.28200328e-01
-3.03873301e-01 7.59772807e-02 -3.60893667e-01 -7.87154436e-02
4.55680847e-01 -9.34135839e-02 -1.53525501e-01 -8.63523424e-01
3.72465581e-01 4.73514616e-01 6.36282027e-01 -9.28258955e-01
1.23186672e+00 9.40279067e-01 2.73930132e-01 -7.63976097e-01
-9.50124562e-01 -2.77821839e-01 -1.31826258e+00 -4.93251324e-01
1.28267074e+00 -1.03732002e+00 -9.27374125e-01 6.06219649e-01
-1.40314138e+00 -1.67924747e-01 -2.43911117e-01 3.28055233e-01
-4.66297716e-01 4.72401768e-01 -6.58198833e-01 -8.86502147e-01
-3.87949228e-01 -1.22592938e+00 1.60450912e+00 1.39591813e-01
-3.75330001e-01 -6.02614880e-01 4.63231504e-02 7.33859420e-01
-1.24314472e-01 9.71233010e-01 5.00190735e-01 1.87779024e-01
-7.29053438e-01 -1.87798724e-01 -1.70708358e-01 -2.36669049e-01
8.32250044e-02 6.79809153e-02 -7.42735565e-01 -1.30513366e-02
-3.68760437e-01 -1.65225118e-01 1.97560549e-01 1.77705348e-01
9.33526039e-01 -2.53024381e-02 -2.28308901e-01 7.92786181e-01
1.03776670e+00 -6.47894442e-01 4.73680049e-01 2.23345086e-01
1.20805359e+00 8.87545645e-01 6.35191917e-01 6.83875084e-01
9.61854577e-01 1.05270541e+00 4.09739196e-01 -5.69727235e-02
-3.05844873e-01 -4.94635552e-01 -3.79263912e-03 7.85539448e-01
-7.23541617e-01 5.39066136e-01 -8.01416576e-01 2.66186953e-01
-1.82872653e+00 -2.95308411e-01 -5.52495897e-01 2.41032052e+00
8.12655926e-01 7.97567889e-03 5.18799067e-01 7.25236908e-02
7.68518806e-01 -1.94709361e-01 -6.49679959e-01 3.47578913e-01
1.94283187e-01 1.52127981e-01 9.77187604e-02 6.45464957e-01
-9.16257739e-01 9.09118176e-01 4.79641771e+00 3.85963827e-01
-5.42472899e-01 1.03804938e-01 2.77347177e-01 -2.61400789e-01
-1.28453553e-01 7.49821737e-02 -7.60822892e-01 2.92725235e-01
8.02756846e-02 4.72388238e-01 4.16599691e-01 6.53488994e-01
1.65917143e-01 -1.93782389e-01 -1.15446508e+00 1.30088997e+00
-9.53731537e-02 -8.01628113e-01 -9.84017625e-02 3.03373337e-01
6.30883694e-01 -4.85185534e-01 -4.20419306e-01 -4.39435095e-02
8.00935999e-02 -7.13344634e-01 1.10603118e+00 8.11800361e-01
7.67081678e-01 -6.55869305e-01 3.95413458e-01 4.68851894e-01
-1.58411741e+00 4.99953240e-01 -3.38662952e-01 -9.01788697e-02
5.57398140e-01 7.73441255e-01 -3.11810941e-01 7.48664558e-01
8.46126914e-01 5.81699550e-01 -3.67903888e-01 8.85964155e-01
-5.31885207e-01 -1.53819278e-01 -5.98523617e-01 3.81338805e-01
-4.31669325e-01 -4.91019130e-01 4.51525033e-01 9.12388623e-01
3.39799047e-01 3.36904436e-01 3.61252725e-01 9.14520979e-01
-3.64905931e-02 6.53936267e-02 -7.16290325e-02 4.76053834e-01
5.89431047e-01 1.41000926e+00 -8.66542995e-01 -2.80716509e-01
-6.99804649e-02 1.29812098e+00 3.89604151e-01 1.79946691e-01
-8.05741549e-01 1.90359354e-01 6.22332096e-01 4.92727488e-01
9.56463367e-02 -4.96861905e-01 -1.82547271e-01 -1.31374216e+00
3.23047698e-01 -7.17540383e-01 1.64042756e-01 -8.03615630e-01
-1.11143553e+00 4.36254472e-01 1.04855813e-01 -1.06217861e+00
7.81279355e-02 -1.73846006e-01 -2.65541643e-01 8.90234709e-01
-1.24309087e+00 -1.55172288e+00 -7.72874236e-01 6.49077952e-01
4.42606539e-01 6.00621164e-01 6.78102553e-01 2.97449142e-01
-5.18286288e-01 1.55053392e-01 -7.40978181e-01 5.23262024e-02
5.79858661e-01 -1.14382422e+00 6.58666432e-01 6.23326778e-01
3.51251066e-02 4.80465829e-01 4.34834838e-01 -9.60197389e-01
-1.62987757e+00 -8.22952986e-01 8.19354117e-01 -8.06096554e-01
2.26812497e-01 -6.81596994e-01 -6.59877002e-01 6.74828887e-01
-5.22987425e-01 2.64521658e-01 2.72038221e-01 9.89430174e-02
-3.11448127e-01 6.28817305e-02 -1.36208248e+00 5.78058839e-01
1.54673076e+00 -2.85845637e-01 -5.14431596e-01 2.51063913e-01
4.32834655e-01 -1.18515897e+00 -1.17884243e+00 2.58835644e-01
9.31579471e-01 -8.83512855e-01 1.40720367e+00 -2.12950349e-01
1.76249623e-01 -7.82887340e-01 -7.31218681e-02 -9.32245493e-01
5.23676798e-02 -6.60798728e-01 -2.83904165e-01 1.20403254e+00
-2.05269247e-01 -1.74190626e-01 9.63493228e-01 9.87013817e-01
2.95358807e-01 -6.96695626e-01 -1.06218433e+00 -3.46546352e-01
-4.73255932e-01 -4.00289148e-01 8.28151762e-01 7.00707257e-01
-4.63940978e-01 1.16898544e-01 -4.09819245e-01 3.68066281e-01
1.13167715e+00 1.94051221e-01 1.44790280e+00 -1.59428263e+00
-2.29757026e-01 7.87244923e-03 -3.65602732e-01 -1.29468799e+00
-6.70877099e-02 -3.15614134e-01 2.67695617e-02 -1.65653372e+00
1.31803751e-01 -4.91493434e-01 4.12394434e-01 3.39278698e-01
-2.08024070e-01 6.39112830e-01 3.83045882e-01 1.97184548e-01
-7.05725431e-01 5.34589827e-01 1.51973248e+00 1.76291659e-01
-2.14561015e-01 -4.50278074e-02 -1.82476327e-01 1.20308185e+00
2.47398347e-01 -3.74095857e-01 1.74547210e-02 -6.32170081e-01
2.00410604e-01 1.43030688e-01 7.56572723e-01 -9.67665613e-01
4.97036427e-02 -3.76757532e-02 5.86698234e-01 -1.00320339e+00
9.01230812e-01 -9.53991354e-01 5.59899986e-01 4.13881898e-01
2.45747477e-01 1.22921811e-02 -8.51340517e-02 4.68816578e-01
4.47630793e-01 1.42691895e-01 5.57134926e-01 -5.02267480e-01
-6.89480752e-02 6.81761980e-01 2.03381687e-01 2.32088678e-02
8.70080829e-01 -3.58394563e-01 2.48233154e-01 -2.15807825e-01
-9.80210602e-01 2.59780109e-01 8.75157535e-01 1.21694945e-01
7.47873545e-01 -1.43969929e+00 -8.22288215e-01 1.30840510e-01
-1.11045375e-01 7.88421571e-01 2.01428264e-01 1.01192713e+00
-6.80903375e-01 -2.93794870e-01 -2.62107216e-02 -9.30190444e-01
-1.52955699e+00 6.62682652e-02 2.30389565e-01 -1.23564184e-01
-7.73768485e-01 7.73787618e-01 5.85895814e-02 -7.48803854e-01
2.96780199e-01 -2.81929702e-01 1.65565416e-01 -2.65006572e-02
4.83509421e-01 7.49279201e-01 -2.04131171e-01 -9.31191385e-01
-5.55985212e-01 1.49861503e+00 3.81137401e-01 -2.60235667e-01
1.47279167e+00 -3.39946330e-01 -2.06935540e-01 2.40969151e-01
7.29628980e-01 1.69599667e-01 -1.75486338e+00 -2.40518287e-01
-3.05293679e-01 -7.62838900e-01 -3.58098179e-01 -3.75708759e-01
-1.06048894e+00 9.14429128e-01 2.23868281e-01 -2.21956164e-01
7.62561083e-01 8.17872062e-02 1.02480936e+00 -2.00045686e-02
7.16152966e-01 -9.67274249e-01 5.99291958e-02 2.71857500e-01
1.00508332e+00 -8.24096680e-01 4.51947063e-01 -9.06504154e-01
-1.89995721e-01 1.09691346e+00 4.87739772e-01 -2.00389773e-01
5.34830570e-01 5.44215813e-02 5.24782874e-02 -3.34629923e-01
3.55792008e-02 -1.80668220e-01 4.77007747e-01 5.34206331e-01
1.86243579e-01 2.10326631e-02 -1.76312700e-01 3.03768605e-01
-3.96552026e-01 5.19003570e-02 9.95941833e-02 7.70878315e-01
-1.43593490e-01 -1.27714705e+00 -9.59208429e-01 3.98955867e-02
-1.49286196e-01 4.19096798e-01 -4.38211739e-01 7.29422510e-01
3.69725734e-01 7.54661441e-01 -3.62782041e-03 -1.89239249e-01
7.14889228e-01 -1.63238756e-02 8.83281231e-01 -6.21947527e-01
-4.28040385e-01 4.35164213e-01 6.19895831e-02 -7.93333173e-01
-6.09973013e-01 -1.00151336e+00 -1.28686833e+00 -3.57366174e-01
-1.10832311e-01 -2.86949366e-01 8.68772447e-01 1.01358998e+00
2.40879565e-01 2.85936803e-01 1.58766344e-01 -1.48273921e+00
-1.63489535e-01 -8.30082774e-01 -4.10226196e-01 8.44278097e-01
3.17681693e-02 -9.35191333e-01 -1.27699912e-01 1.82855681e-01] | [7.070488452911377, -1.114454984664917] |
d8ed1dc6-8639-4855-857d-2de08559c308 | time-series-prediction-by-multi-task-gpr-with | 2204.12085 | null | https://arxiv.org/abs/2204.12085v1 | https://arxiv.org/pdf/2204.12085v1.pdf | Time Series Prediction by Multi-task GPR with Spatiotemporal Information Transformation | Making an accurate prediction of an unknown system only from a short-term time series is difficult due to the lack of sufficient information, especially in a multi-step-ahead manner. However, a high-dimensional short-term time series contains rich dynamical information, and also becomes increasingly available in many fields. In this work, by exploiting spatiotemporal information (STI) transformation scheme that transforms such high-dimensional/spatial information to temporal information, we developed a new method called MT-GPRMachine to achieve accurate prediction from a short-term time series. Specifically, we first construct a specific multi-task GPR which is multiple linked STI mappings to transform high dimensional/spatial information into temporal/dynamical information of any given target variable, and then makes multi step-ahead prediction of the target variable by solving those STI mappings. The multi-step-ahead prediction results on various synthetic and real-world datasets clearly validated that MT-GPRMachine outperformed other existing approaches. | ['Luonan Chen', 'Jie Cheng', 'Xiaohu Hao', 'Peng Tao'] | 2022-04-26 | null | null | null | null | ['gpr', 'gpr', 'time-series-prediction'] | ['computer-vision', 'miscellaneous', 'time-series'] | [-1.60966478e-02 -7.30640531e-01 7.78250545e-02 -1.14254519e-01
-5.65715015e-01 -2.93338656e-01 5.78901052e-01 -6.97638988e-02
8.75412151e-02 9.59747553e-01 -1.22644797e-01 -3.26188087e-01
-7.53953695e-01 -7.86572397e-01 -3.80029589e-01 -8.94883454e-01
-4.03287619e-01 4.24133867e-01 3.68426204e-01 -3.37852985e-01
2.14639530e-01 3.17665607e-01 -1.40654254e+00 -1.84562936e-01
8.72546732e-01 1.21302724e+00 4.35218096e-01 4.05255258e-01
8.49807262e-02 5.36415160e-01 -2.83833504e-01 4.83828306e-01
1.06966704e-01 -3.22843671e-01 -2.78623313e-01 -3.35195363e-01
-6.31424904e-01 -9.93440300e-03 -6.28551543e-02 5.79415917e-01
2.67909080e-01 5.12921393e-01 4.76362526e-01 -1.30790675e+00
-3.11434567e-01 -7.41181225e-02 -5.32130420e-01 5.24714410e-01
5.09763733e-02 3.45493048e-01 5.93770862e-01 -7.32937574e-01
3.87872338e-01 1.09260738e+00 8.25572252e-01 -4.62603644e-02
-1.15726805e+00 -7.21408844e-01 3.49002659e-01 1.91588506e-01
-1.29803431e+00 -1.44728599e-02 1.01887023e+00 -7.79920220e-01
7.74664104e-01 1.36768043e-01 7.57683158e-01 9.02334273e-01
8.78782153e-01 1.58574864e-01 1.21133041e+00 1.59063041e-01
6.45610085e-03 -2.78492779e-01 9.71479490e-02 2.31018916e-01
-3.05026084e-01 5.21192551e-01 -3.21279258e-01 -1.63945600e-01
8.67445290e-01 5.46968222e-01 -1.05285503e-01 -6.51052571e-04
-1.61324048e+00 4.73109424e-01 2.30414093e-01 4.77106839e-01
-5.53005934e-01 -1.83872283e-01 5.14140129e-01 4.75047767e-01
9.28768396e-01 4.15487766e-01 -8.55879724e-01 -5.05288720e-01
-8.60882521e-01 1.54670790e-01 5.00985324e-01 4.46219027e-01
8.01020861e-01 1.33275717e-01 -1.60004824e-01 5.11529326e-01
1.90427274e-01 6.45096302e-01 8.54617238e-01 -5.78255117e-01
6.95087790e-01 3.90430987e-01 3.96848619e-01 -1.42906439e+00
-6.71982527e-01 -7.01010585e-01 -1.38861573e+00 -2.22360775e-01
1.74478501e-01 -4.46581066e-01 -5.96922159e-01 1.88029075e+00
5.94786942e-01 7.94567347e-01 5.98021299e-02 1.17916501e+00
1.30771413e-01 1.35500574e+00 -1.52101904e-01 -1.00293219e+00
9.51551199e-01 -7.89679229e-01 -8.00599396e-01 8.83410349e-02
5.80538929e-01 -5.59014142e-01 8.25320065e-01 2.39644080e-01
-3.50013882e-01 -8.00004542e-01 -7.14341581e-01 4.35104698e-01
-3.31672281e-01 1.85291264e-02 3.14071387e-01 -2.23032445e-01
-4.36095268e-01 7.95206666e-01 -1.00218797e+00 -1.10834904e-01
-3.87519270e-01 8.81578326e-02 -3.22842896e-01 3.46276522e-01
-1.55704010e+00 6.63117111e-01 3.66303205e-01 3.94053549e-01
-8.47222447e-01 -1.00013721e+00 -5.85124552e-01 -1.45690888e-01
4.52912956e-01 -5.42394221e-01 9.15301681e-01 -4.58833456e-01
-1.52495277e+00 -1.46368239e-02 -4.27637368e-01 -2.15105206e-01
2.95348674e-01 -3.48182805e-02 -8.79066408e-01 -3.27064157e-01
1.87089413e-01 -2.73229808e-01 1.02390397e+00 -7.08793521e-01
-5.30572057e-01 -4.86111343e-01 -4.34478879e-01 2.21230656e-01
-3.47855538e-01 -8.16567019e-02 -3.15251462e-02 -8.28887939e-01
2.63428092e-01 -9.44000304e-01 -3.39756966e-01 -3.83892715e-01
-3.27915013e-01 -4.02404852e-02 1.23093534e+00 -8.97344589e-01
1.67058313e+00 -2.11003971e+00 3.85777295e-01 1.30671382e-01
2.01174185e-01 2.69261777e-01 1.35227397e-01 6.87616229e-01
-3.59438807e-01 -3.40688303e-02 -3.22872013e-01 -2.28908077e-01
-3.98108631e-01 1.26743153e-01 -7.39206433e-01 2.21341863e-01
2.69573443e-02 6.44518137e-01 -9.93782163e-01 -3.30324113e-01
5.31227946e-01 3.77024174e-01 1.99240968e-01 3.62948716e-01
-2.59358078e-01 1.08198667e+00 -9.59352970e-01 3.81700188e-01
5.33196390e-01 -4.67323214e-01 -2.48654246e-01 -3.13416719e-02
-5.45074284e-01 -1.75692677e-01 -8.50831211e-01 1.38619018e+00
-5.86043656e-01 3.89185220e-01 -3.80975187e-01 -1.22495854e+00
1.32195544e+00 5.07933557e-01 1.01025617e+00 -9.98068869e-01
-1.36387423e-01 2.02693716e-01 -3.05973083e-01 -6.80491865e-01
2.55187303e-01 -3.99269223e-01 -1.86799407e-01 3.28431755e-01
-4.72946435e-01 2.76742596e-02 -7.95126259e-02 -3.39247555e-01
9.85123873e-01 1.86826691e-01 2.78738558e-01 -8.97767097e-02
8.38086247e-01 1.41503364e-01 1.06041086e+00 1.09738246e-01
-1.44514591e-01 4.30718154e-01 3.49491030e-01 -6.86239600e-01
-9.36642170e-01 -6.35693669e-01 -9.03047621e-02 5.29178858e-01
2.06053287e-01 -4.25801963e-01 5.82863092e-02 -3.14944535e-01
1.27200171e-01 2.53850758e-01 -6.34944916e-01 -2.67865896e-01
-6.11632824e-01 -9.47737694e-01 5.27373403e-02 2.62091845e-01
3.29663187e-01 -9.09795165e-01 -3.92798871e-01 7.81228364e-01
-1.79419488e-01 -8.46789658e-01 -4.55706686e-01 -1.60758197e-02
-9.62243617e-01 -8.38764250e-01 -8.98750424e-01 -3.53532374e-01
2.85160333e-01 2.61994064e-01 5.19143164e-01 -3.05474937e-01
-3.21156383e-02 -3.43018830e-01 -2.64722705e-01 -1.94108263e-01
1.42884431e-02 4.30693962e-02 3.59111100e-01 3.35973084e-01
-1.60735041e-01 -8.99133861e-01 -4.96275395e-01 7.49350131e-01
-5.76847851e-01 4.73258257e-01 3.83769542e-01 9.11006808e-01
1.03247714e+00 5.10160983e-01 8.04655850e-01 -3.24970752e-01
5.89793086e-01 -9.04851258e-01 -8.31926346e-01 1.90915555e-01
-8.08535874e-01 5.15987761e-02 1.39441144e+00 -6.92505896e-01
-9.75578845e-01 -1.34834349e-01 3.82577367e-02 -8.95768225e-01
2.23771885e-01 1.17962110e+00 1.60603434e-01 2.71079391e-01
2.30311975e-01 6.97440207e-01 2.56168656e-02 -8.34072590e-01
-2.73564011e-01 4.13924038e-01 4.08637077e-01 -4.51071084e-01
8.46845627e-01 1.14373066e-01 4.03276324e-01 -5.33201933e-01
-6.63698792e-01 -3.86934817e-01 -6.27393365e-01 -3.05782586e-01
6.80822670e-01 -8.58964443e-01 -6.60036206e-01 9.05864656e-01
-7.95792103e-01 -4.43853348e-01 -6.11466430e-02 7.87225723e-01
-4.40675259e-01 7.59245604e-02 -4.84985113e-01 -6.53169394e-01
-2.53521591e-01 -8.85673404e-01 9.72843885e-01 1.01920083e-01
4.51927073e-02 -1.19203234e+00 5.97420514e-01 -1.08895592e-01
5.06396055e-01 7.00686455e-01 7.38328993e-01 -3.34292501e-01
-3.17918777e-01 -1.64882481e-01 -1.52804226e-01 3.29374522e-02
3.91316265e-01 1.81930840e-01 -3.67755234e-01 -3.04321170e-01
3.04549843e-01 2.37989336e-01 3.64396483e-01 4.52502340e-01
1.22576118e+00 -4.11124200e-01 -5.67983687e-01 7.13060319e-01
1.42863846e+00 5.09626031e-01 1.31971285e-01 3.46379459e-01
8.45058918e-01 4.11850065e-01 1.46734214e+00 7.10137486e-01
5.31205952e-01 8.37960064e-01 1.62470028e-01 4.72981893e-02
4.59785730e-01 -3.96983802e-01 1.67665020e-01 1.28713822e+00
-5.54756150e-02 -6.26265109e-02 -1.14699590e+00 6.83052957e-01
-2.22566366e+00 -9.94665921e-01 -1.69142663e-01 2.07218599e+00
6.64513946e-01 -9.48496908e-03 7.41600171e-02 2.27808818e-01
7.25673854e-01 3.76348674e-01 -1.05339205e+00 1.76339626e-01
4.30214256e-02 -2.70969689e-01 2.85657018e-01 1.82189941e-01
-1.03610766e+00 6.01870120e-01 5.79993200e+00 9.20585692e-01
-1.66048670e+00 -1.72275547e-02 5.92737615e-01 5.77390604e-02
-5.10849319e-02 4.86158021e-02 -6.46455705e-01 1.00823104e+00
1.35302091e+00 -8.62478077e-01 3.05738539e-01 6.71827555e-01
8.71953309e-01 1.65268078e-01 -6.43263042e-01 1.05977941e+00
-4.57805008e-01 -1.14308393e+00 -2.55169302e-01 -1.86638050e-02
6.30980909e-01 -1.05893373e-01 1.14088871e-01 3.12757909e-01
-4.43902090e-02 -8.24147940e-01 2.46168509e-01 1.15191221e+00
7.87980497e-01 -5.73478401e-01 5.72667360e-01 9.28750098e-01
-1.89660263e+00 -1.71168685e-01 -3.81066687e-02 -3.26640010e-01
4.97729182e-01 9.17498171e-01 -4.10866916e-01 8.55537057e-01
5.81206143e-01 1.59145188e+00 -2.65365064e-01 9.91580486e-01
2.61246681e-01 5.52590013e-01 -4.36248899e-01 1.67168349e-01
1.34718329e-01 -5.62203467e-01 6.59346819e-01 4.63323921e-01
9.35203969e-01 2.56208479e-01 3.29771936e-01 4.98916060e-01
6.46231294e-01 -1.52450457e-01 -5.95391989e-01 -1.01677284e-01
4.11554039e-01 1.07892525e+00 -2.59757429e-01 -3.12592238e-01
-1.28065497e-01 5.87614238e-01 -1.28055587e-01 4.14866149e-01
-1.08531439e+00 -3.72049987e-01 6.49962127e-01 -1.23677449e-02
7.66711459e-02 -5.18922508e-01 -1.34743288e-01 -1.29371977e+00
2.76533902e-01 -6.49837375e-01 3.92902911e-01 -1.06054938e+00
-1.17748678e+00 6.56092525e-01 -3.01616844e-02 -1.90103734e+00
-6.70288563e-01 -2.64919605e-02 -7.62880623e-01 1.16560066e+00
-1.45645475e+00 -9.82390881e-01 -3.50185126e-01 7.83046365e-01
4.97230887e-01 -2.86725350e-02 7.55756438e-01 2.89208174e-01
-9.87309337e-01 1.14548635e-02 6.05062246e-01 -4.85383689e-01
7.12209702e-01 -7.85004437e-01 3.93264145e-01 7.65778065e-01
-6.14474118e-01 4.23852235e-01 8.48763287e-01 -7.05546975e-01
-1.46037459e+00 -1.33579862e+00 8.10352206e-01 -1.24534303e-02
9.35209036e-01 1.02596261e-01 -1.13847911e+00 6.77006066e-01
-5.20745158e-01 1.96547315e-01 1.56330630e-01 -2.16974095e-01
1.13480709e-01 -5.29913902e-01 -9.00668859e-01 2.05030784e-01
8.37681413e-01 -6.48550153e-01 -5.08742273e-01 3.64683330e-01
9.21297550e-01 -3.92492026e-01 -1.29913151e+00 7.63905287e-01
4.33988929e-01 -6.00368202e-01 9.63072836e-01 -3.41341019e-01
3.49469006e-01 -5.83346844e-01 -1.65340379e-02 -1.68333232e+00
-4.66118574e-01 -7.25253105e-01 -2.52955586e-01 9.99337137e-01
1.20557278e-01 -1.13479578e+00 3.76528949e-01 4.85156953e-01
-1.80665493e-01 -9.31431115e-01 -1.01329446e+00 -9.85307455e-01
-1.19022116e-01 -3.58591467e-01 9.26265240e-01 1.09813309e+00
-2.05419101e-02 1.23777658e-01 -8.22565079e-01 3.13767880e-01
2.82727897e-01 5.86131215e-01 7.06343949e-01 -1.37880743e+00
-1.97922155e-01 -1.86765287e-02 -3.24117988e-01 -9.77946877e-01
4.43858914e-02 -2.34716609e-01 -3.59534584e-02 -1.18416762e+00
-7.55311772e-02 -6.88551009e-01 -6.30717933e-01 4.15361911e-01
-1.83783010e-01 -1.31100550e-01 -9.56599563e-02 7.62072086e-01
-1.93682313e-01 9.92205441e-01 1.43105936e+00 1.62209123e-01
-3.44463199e-01 2.54378855e-01 -3.34372930e-02 2.11041972e-01
7.99273372e-01 -4.09400105e-01 -6.27594411e-01 -1.56678319e-01
1.73847412e-03 1.08208573e+00 3.26252908e-01 -1.27692318e+00
2.10537583e-01 -7.89710343e-01 1.09436065e-01 -1.04793966e+00
6.47638559e-01 -8.45253050e-01 7.60581076e-01 3.80999237e-01
5.24260178e-02 3.45528871e-01 3.17434192e-01 6.82517409e-01
-6.51950717e-01 6.55906558e-01 4.24471289e-01 2.01446220e-01
-9.19617653e-01 7.79952109e-01 -1.69767648e-01 -1.48576468e-01
1.32046640e+00 -7.64583200e-02 -2.67037004e-01 -2.54126251e-01
-9.54586506e-01 5.61223745e-01 2.77824193e-01 5.23544490e-01
4.19093132e-01 -1.32501864e+00 -2.82686174e-01 4.64210898e-01
-5.36353104e-02 -5.63279465e-02 7.61942029e-01 1.15822506e+00
-9.33060721e-02 7.00490355e-01 -3.70259017e-01 -7.79082477e-01
-1.14364886e+00 6.92352474e-01 4.54637915e-01 -5.64494729e-01
-8.73224735e-01 2.44606972e-01 1.90240934e-01 -4.75797832e-01
-4.67917264e-01 -4.42895770e-01 -2.54130274e-01 7.34574869e-02
4.53804880e-01 2.92514205e-01 -1.20634802e-01 -8.82877350e-01
-3.82310778e-01 1.16048515e+00 3.02875549e-01 -1.40956014e-01
1.53187573e+00 -4.13493723e-01 -1.85961917e-03 9.96255338e-01
1.46666431e+00 -5.25114536e-01 -1.53958166e+00 -3.42939317e-01
-2.26115331e-01 -5.23928463e-01 -1.06373793e-02 -3.88901889e-01
-1.07583904e+00 8.68770778e-01 3.55874330e-01 4.60303247e-01
1.31442750e+00 -6.37173235e-01 1.24426699e+00 2.54846752e-01
5.70203185e-01 -8.14777970e-01 -5.90837710e-02 8.44561636e-01
7.65820920e-01 -1.03747261e+00 -2.43860751e-01 -2.91318625e-01
-5.57879984e-01 1.10489035e+00 5.48948944e-01 -6.04190864e-02
1.16217816e+00 -6.51919693e-02 -1.75226226e-01 -4.11876142e-02
-1.15174901e+00 9.37194601e-02 2.86516398e-01 2.26504132e-01
4.65347618e-03 9.52492878e-02 -2.85683274e-01 7.56151438e-01
-1.65450126e-01 3.33728850e-01 -5.30609302e-03 7.73963153e-01
-4.33694869e-02 -9.03856277e-01 -5.16702354e-01 5.44026852e-01
1.54890912e-02 3.32566559e-01 1.72458470e-01 6.33530438e-01
-1.03186376e-01 8.52205575e-01 8.42375606e-02 -8.98368061e-01
1.74902588e-01 -7.99809620e-02 -3.13133717e-01 -3.15136701e-01
-3.05030674e-01 1.05670720e-01 -2.63694733e-01 -7.11169600e-01
-2.01279402e-01 -6.83725297e-01 -1.28227401e+00 -3.95531207e-01
-1.00439698e-01 2.74557084e-01 5.72118282e-01 1.18004012e+00
7.92490125e-01 6.94036365e-01 1.19556475e+00 -9.44337666e-01
-3.68431330e-01 -8.98802102e-01 -6.17507637e-01 3.29868644e-01
6.61438048e-01 -9.43709314e-01 -6.25668943e-01 -1.80588871e-01] | [6.929501533508301, 3.033017635345459] |
0749e0a7-8c9d-4583-b415-4583711dd1a9 | gett-qa-graph-embedding-based-t2t-transformer | 2303.13284 | null | https://arxiv.org/abs/2303.13284v3 | https://arxiv.org/pdf/2303.13284v3.pdf | GETT-QA: Graph Embedding based T2T Transformer for Knowledge Graph Question Answering | In this work, we present an end-to-end Knowledge Graph Question Answering (KGQA) system named GETT-QA. GETT-QA uses T5, a popular text-to-text pre-trained language model. The model takes a question in natural language as input and produces a simpler form of the intended SPARQL query. In the simpler form, the model does not directly produce entity and relation IDs. Instead, it produces corresponding entity and relation labels. The labels are grounded to KG entity and relation IDs in a subsequent step. To further improve the results, we instruct the model to produce a truncated version of the KG embedding for each entity. The truncated KG embedding enables a finer search for disambiguation purposes. We find that T5 is able to learn the truncated KG embeddings without any change of loss function, improving KGQA performance. As a result, we report strong results for LC-QuAD 2.0 and SimpleQuestions-Wikidata datasets on end-to-end KGQA over Wikidata. | ['Chris Biemann', 'Ricardo Usbeck', 'Pranav Ajit Nair', 'Debayan Banerjee'] | 2023-03-23 | null | null | null | null | ['graph-question-answering'] | ['graphs'] | [-2.74231791e-01 8.00162971e-01 -8.61804858e-02 -5.45876265e-01
-1.14900041e+00 -7.13427067e-01 3.40030462e-01 5.13568640e-01
-4.15875971e-01 6.26830220e-01 2.89157689e-01 -5.29058635e-01
-8.10579807e-02 -1.24140036e+00 -7.82553613e-01 -9.45412815e-02
2.16422230e-02 1.08317137e+00 4.55026060e-01 -3.97228807e-01
-2.42883429e-01 -1.40422180e-01 -1.07904685e+00 5.46494663e-01
1.15345788e+00 1.04023612e+00 -2.62508895e-02 6.96316838e-01
-4.70712602e-01 1.12488639e+00 -4.56724048e-01 -1.24011362e+00
1.70269877e-01 -2.73738056e-01 -1.46678078e+00 -5.45062125e-01
1.03293347e+00 -2.79952109e-01 -2.80798882e-01 7.19836593e-01
5.09575248e-01 5.73946945e-02 5.89985192e-01 -1.21190226e+00
-1.06443715e+00 9.30277944e-01 1.31442314e-02 -1.51613906e-01
4.25520957e-01 -5.99529818e-02 1.93152153e+00 -1.01235557e+00
7.40105331e-01 1.41956258e+00 5.68409741e-01 5.16574621e-01
-9.72242475e-01 -2.67863125e-01 5.60682118e-02 4.20334876e-01
-1.30707824e+00 -1.16998620e-01 4.27941531e-01 7.86825046e-02
1.40878928e+00 2.50220031e-01 4.18315083e-01 4.43769276e-01
-2.58938093e-02 8.32180500e-01 6.83688700e-01 -6.03869557e-01
1.06068000e-01 3.25798662e-03 4.45347309e-01 1.15120530e+00
2.69881159e-01 -4.19350058e-01 -4.13706303e-01 -3.85285318e-01
2.22728983e-01 -4.83807117e-01 -2.80778319e-01 -6.93107963e-01
-1.10083258e+00 9.16493237e-01 8.31494451e-01 -1.52459368e-02
-2.47186139e-01 4.79432404e-01 3.81679118e-01 5.05710006e-01
3.78152966e-01 8.79541814e-01 -7.83450007e-01 1.59467340e-01
-4.90589768e-01 5.34327030e-01 1.16147041e+00 1.04660571e+00
1.02593696e+00 -4.68099624e-01 -5.83701253e-01 6.40383363e-01
3.16836894e-01 5.65938652e-01 4.20471514e-03 -1.04378498e+00
8.20659578e-01 7.58719146e-01 -1.73755027e-02 -8.23654652e-01
-2.42429674e-01 -4.81488109e-01 -1.45967323e-02 -2.84239769e-01
6.15660608e-01 -1.12121746e-01 -1.03609395e+00 1.71742594e+00
4.78181481e-01 -8.13682601e-02 6.17420554e-01 9.22124803e-01
1.29044044e+00 6.13148451e-01 3.91104579e-01 3.67981315e-01
1.72860074e+00 -1.34107339e+00 -8.10722291e-01 -3.57775837e-01
1.31301212e+00 -3.44076216e-01 1.41404796e+00 -3.02063208e-02
-8.27712297e-01 -2.57511437e-01 -6.85771286e-01 -8.94713640e-01
-8.19576204e-01 7.20908046e-02 7.23016798e-01 3.67498696e-01
-1.22417295e+00 3.08907926e-01 -5.94342411e-01 -5.60209394e-01
2.39507675e-01 6.73432350e-02 -3.49236071e-01 -4.41658825e-01
-1.69388771e+00 1.13129389e+00 6.12517893e-01 -1.58612698e-01
-6.28003895e-01 -9.46144521e-01 -1.09975374e+00 2.18729898e-01
8.25220525e-01 -1.31354380e+00 1.55127585e+00 -2.56112635e-01
-1.21754503e+00 9.00450528e-01 -3.33897710e-01 -5.72753191e-01
1.48665002e-02 -4.39830840e-01 -4.83197480e-01 2.99775273e-01
3.33261102e-01 7.89883018e-01 5.34830809e-01 -1.05667853e+00
-5.91815293e-01 -4.43461031e-01 4.83164161e-01 3.64453048e-01
-8.08084533e-02 -2.67707080e-01 -6.92748904e-01 -3.29639792e-01
-1.72364503e-01 -5.87445617e-01 7.55189136e-02 -1.50792273e-02
-4.70489293e-01 -8.89467418e-01 4.62105364e-01 -8.67678940e-01
1.21594977e+00 -1.82630289e+00 2.51904111e-02 5.57289571e-02
5.05983412e-01 2.98766106e-01 -3.39537501e-01 8.26232195e-01
2.16012537e-01 1.39316797e-01 -2.83127546e-01 -3.07136387e-01
3.25083643e-01 5.41377544e-01 -3.90148371e-01 -2.35195488e-01
5.55985928e-01 1.56335735e+00 -1.31373799e+00 -6.76061392e-01
-2.46070608e-01 5.83040714e-02 -6.40659451e-01 2.28948504e-01
-9.26068485e-01 -4.03351545e-01 -6.35068297e-01 6.32234871e-01
3.94673586e-01 -3.34937125e-01 7.42865950e-02 -5.09371281e-01
4.59830731e-01 8.72606516e-01 -5.83385766e-01 1.95132005e+00
-5.91288626e-01 3.78490955e-01 -2.69432753e-01 -6.35668695e-01
7.14652717e-01 1.58442840e-01 -1.04230218e-01 -8.04454923e-01
-1.25098556e-01 3.77725989e-01 -1.94964558e-01 -6.38393879e-01
7.56932735e-01 -1.76989082e-02 -2.79979378e-01 4.35412854e-01
3.70512754e-01 -4.44415897e-01 2.48681694e-01 8.17198038e-01
1.28498518e+00 2.24918686e-02 9.56878215e-02 -7.02218637e-02
1.98918104e-01 4.62981999e-01 7.81082958e-02 6.60564780e-01
2.52054036e-01 2.35377148e-01 5.42150438e-01 -1.77588612e-01
-7.33544111e-01 -1.31513917e+00 1.46365434e-01 1.25047040e+00
1.59077626e-02 -8.17848325e-01 -6.13269508e-01 -1.42286468e+00
4.41511780e-01 1.23361719e+00 -6.07676089e-01 -2.30601236e-01
-5.75675488e-01 -1.29112214e-01 7.31480122e-01 4.66614455e-01
5.00514209e-01 -1.15173054e+00 -2.19268903e-01 3.14877480e-01
-2.36648351e-01 -1.17254579e+00 -5.13464034e-01 1.21131435e-01
-6.66206777e-01 -1.17760324e+00 -3.89788687e-01 -9.96859729e-01
4.98300076e-01 -3.72125469e-02 1.79637611e+00 2.44212691e-02
6.49318546e-02 6.68354034e-01 -5.85149288e-01 -3.21016312e-01
-2.06464052e-01 3.87562901e-01 -5.10475934e-01 -2.39901543e-01
6.84263825e-01 -2.70785958e-01 -4.42965358e-01 -2.18800455e-02
-8.60291541e-01 -1.19105503e-01 5.50644755e-01 7.03536212e-01
5.94111919e-01 -1.21620536e-01 7.00651824e-01 -1.10157144e+00
9.97863054e-01 -4.10986513e-01 -4.89450693e-01 8.06547582e-01
-7.89295077e-01 6.08959913e-01 6.98843658e-01 -1.74859073e-02
-8.73037457e-01 -2.87247866e-01 -2.91054696e-01 -2.48694763e-01
2.27327347e-01 1.03040636e+00 -3.35941255e-01 1.16152018e-01
7.94894099e-01 -3.79818752e-02 -2.49574631e-01 -5.47185838e-01
1.27735388e+00 4.72864002e-01 7.17974007e-01 -7.89996326e-01
9.39281464e-01 6.89669698e-02 -3.33833903e-01 -2.63714492e-01
-1.26038206e+00 -5.58161259e-01 -2.39210233e-01 1.74949855e-01
9.01918232e-01 -9.09630179e-01 -6.64389133e-01 1.24054542e-02
-1.17696464e+00 -5.31781554e-01 -5.94359875e-01 7.71597102e-02
-2.27656201e-01 2.54833430e-01 -5.52861214e-01 -4.31735188e-01
-7.43554294e-01 -5.20326436e-01 1.25766408e+00 2.29249895e-02
-1.45052835e-01 -1.17337418e+00 2.54819870e-01 6.60222769e-01
3.60799760e-01 -9.17556360e-02 1.39875245e+00 -8.78550231e-01
-6.96698308e-01 -1.15461439e-01 -5.26563883e-01 3.64714831e-01
8.07865039e-02 -3.40728104e-01 -7.51153052e-01 -1.54221073e-01
-6.10592127e-01 -7.29737103e-01 1.01177764e+00 -3.55852783e-01
7.80348003e-01 -5.87009132e-01 -1.30974963e-01 3.29732239e-01
1.44541371e+00 -1.95352048e-01 6.08056962e-01 2.60485739e-01
9.01342273e-01 6.41502500e-01 7.39527225e-01 -2.59826541e-01
1.14836574e+00 5.21069348e-01 5.41905880e-01 2.08291076e-02
-4.31063890e-01 -9.28379834e-01 2.31655568e-01 7.74644971e-01
4.75703478e-01 -5.05449474e-01 -9.93679106e-01 9.94051456e-01
-1.70482016e+00 -6.57920897e-01 -1.52708426e-01 1.79780149e+00
1.26018381e+00 -2.14102909e-01 -1.36775762e-01 -2.99377292e-01
1.25620082e-01 6.44233599e-02 -7.11645782e-01 -3.63837183e-01
-1.05009615e-01 6.62367463e-01 4.53890651e-01 7.73869872e-01
-6.81156933e-01 1.52010250e+00 5.37788439e+00 9.07643855e-01
-7.71215022e-01 1.14487790e-01 1.07730269e-01 3.06288987e-01
-7.82452285e-01 2.82383502e-01 -8.66433799e-01 1.96787678e-02
8.12692046e-01 -3.03413451e-01 3.79871517e-01 8.32098246e-01
-5.28862238e-01 2.98235267e-02 -1.12948215e+00 5.84870100e-01
4.62800413e-02 -1.43728793e+00 6.23068452e-01 -2.89309651e-01
2.70477265e-01 2.85608135e-02 -2.00100094e-01 9.47676897e-01
6.04580939e-01 -1.07023501e+00 5.04616201e-01 3.18113565e-01
7.63401389e-01 -6.83602333e-01 8.30270469e-01 2.38391161e-01
-9.80010271e-01 1.03568673e-01 -4.89087522e-01 7.25948587e-02
3.05599153e-01 5.73182225e-01 -1.42378640e+00 1.12642956e+00
5.40475667e-01 3.88461769e-01 -7.94079542e-01 7.16650486e-01
-7.60075212e-01 6.71648860e-01 -3.12850446e-01 -2.59949356e-01
2.59471148e-01 -4.41130921e-02 3.33453268e-01 1.03830731e+00
2.56550372e-01 2.23020881e-01 1.84333995e-02 8.86595547e-01
-6.47655964e-01 2.81430990e-01 -4.54284608e-01 -3.49205881e-01
5.51466227e-01 1.29611981e+00 6.68493956e-02 -6.06722832e-01
-4.28217560e-01 1.04696155e+00 1.03309858e+00 5.09019017e-01
-5.42742848e-01 -6.95625603e-01 5.90339720e-01 1.72477439e-01
5.14486194e-01 4.53631431e-02 1.53612599e-01 -1.21460569e+00
2.09464774e-01 -9.59928036e-01 9.78195250e-01 -1.08327222e+00
-1.53230381e+00 5.51854968e-01 -1.89359505e-02 -4.60934430e-01
-3.57510895e-01 -6.63985550e-01 -3.84240001e-01 9.96910214e-01
-1.80467927e+00 -1.31683445e+00 -2.18682885e-01 5.94088376e-01
2.51170862e-02 2.17654839e-01 1.04725730e+00 2.54973412e-01
-2.01334402e-01 8.42568576e-01 -2.99555331e-01 4.83611137e-01
9.57647860e-01 -1.66596699e+00 6.80449307e-01 6.21650994e-01
4.70851839e-01 7.24160731e-01 6.91086948e-01 -7.16917098e-01
-1.45458364e+00 -1.39751732e+00 1.53260016e+00 -8.67869556e-01
8.91348779e-01 -3.70676368e-01 -1.03237677e+00 8.81772757e-01
5.36378086e-01 1.37293222e-03 5.80711007e-01 4.36452180e-01
-7.82298088e-01 -1.84064239e-01 -1.06291866e+00 5.31169295e-01
1.03200471e+00 -8.21087897e-01 -1.09652948e+00 3.57012808e-01
1.48346233e+00 -4.22639191e-01 -1.15226448e+00 4.54627305e-01
1.72507569e-01 -3.27830553e-01 9.39278603e-01 -8.89968157e-01
2.02535719e-01 -5.74187636e-01 -1.25645474e-01 -1.50920308e+00
-1.10909596e-01 -3.90686393e-01 -5.43091178e-01 1.03770709e+00
1.03611386e+00 -8.71124983e-01 8.31672966e-01 7.12257326e-01
-4.22694802e-01 -9.94275212e-01 -1.01108050e+00 -8.08192909e-01
3.20579946e-01 -3.47679526e-01 7.81746924e-01 1.04512095e+00
1.44467920e-01 8.72258902e-01 9.16630104e-02 4.57903534e-01
6.19985640e-01 5.21977127e-01 8.36500049e-01 -1.09680390e+00
-3.51907700e-01 5.77402934e-02 -8.35844036e-03 -1.38836813e+00
1.06473610e-01 -1.14011824e+00 -1.62321419e-01 -2.12764621e+00
-1.26005679e-01 -6.38500631e-01 -1.01528168e-01 1.02928638e+00
-5.40770233e-01 -3.31464894e-02 3.55849974e-02 -2.22554460e-01
-9.26548243e-01 7.02168763e-01 1.37535071e+00 -3.92664850e-01
7.20252693e-02 -4.51238632e-01 -8.87209356e-01 8.61021969e-03
8.69786680e-01 -5.48364222e-01 -7.41046131e-01 -8.94451797e-01
6.65177166e-01 4.81899492e-02 5.01001775e-01 -5.02166450e-01
3.14908594e-01 1.79117277e-01 -2.21469238e-01 -5.66049635e-01
3.09134603e-01 -5.96311510e-01 -2.82337874e-01 8.26233178e-02
-3.72391462e-01 7.26989433e-02 1.93711877e-01 3.90014499e-01
-4.26452845e-01 -2.30135113e-01 1.70630857e-01 -4.59322631e-02
-7.00043559e-01 4.46557879e-01 2.27831125e-01 6.15065753e-01
5.25994360e-01 2.04159707e-01 -8.25814307e-01 -5.36495745e-01
-6.49741530e-01 9.92848396e-01 3.78964454e-01 3.62423360e-01
4.98217613e-01 -1.29909265e+00 -8.36972535e-01 -2.22735092e-01
6.83028340e-01 3.42626572e-01 -1.33660771e-02 5.72683156e-01
-6.30090058e-01 7.31578529e-01 3.33584368e-01 -7.80005381e-02
-1.03633761e+00 5.49958408e-01 4.22218651e-01 -7.04225719e-01
-3.59397173e-01 9.43037510e-01 4.21086699e-02 -1.01794457e+00
1.30214870e-01 -7.52546489e-01 -2.32942581e-01 3.25816832e-02
3.04667175e-01 -2.21372265e-02 2.06208780e-01 -1.53722256e-01
-2.63596982e-01 2.58315802e-01 -2.17769071e-01 -1.04217634e-01
1.05303359e+00 8.15589726e-02 -3.41017455e-01 -2.25194450e-02
1.20968652e+00 2.33309776e-01 -5.93773961e-01 -5.82096517e-01
2.69834280e-01 -2.24849299e-01 -2.84026526e-02 -1.31133115e+00
-9.25488234e-01 7.89716184e-01 -3.85611318e-02 2.52160400e-01
9.57701921e-01 4.92193907e-01 1.20293558e+00 8.89984429e-01
5.56050420e-01 -6.58770144e-01 -2.93971449e-02 7.42258608e-01
9.65627849e-01 -9.05434072e-01 -3.44344735e-01 -6.31676555e-01
-5.40328324e-01 8.60449731e-01 6.86350048e-01 2.11598411e-01
1.57493591e-01 -2.45950520e-01 3.21999252e-01 -7.37791359e-01
-9.04417217e-01 -6.40054405e-01 5.35308778e-01 6.75586939e-01
1.57559127e-01 -2.13763583e-02 -1.43437117e-01 5.84878027e-01
-5.29010713e-01 -2.03095526e-01 2.90727951e-02 6.99657202e-01
-4.10883188e-01 -1.33399010e+00 1.58598885e-01 5.22121191e-01
-2.17329979e-01 -6.17225528e-01 -7.16675282e-01 8.55492353e-01
-7.83285350e-02 9.40188050e-01 -1.97665229e-01 -2.42453814e-01
7.05025136e-01 4.56985652e-01 4.41365212e-01 -8.29554260e-01
-6.04700923e-01 -8.37680757e-01 8.08007002e-01 -5.59547305e-01
2.98921671e-02 -1.73391268e-01 -1.56216371e+00 -1.57714099e-01
-4.38330799e-01 8.01090002e-01 3.12624127e-01 6.77509069e-01
7.25377083e-01 4.03083056e-01 6.01087026e-02 2.09335580e-01
-5.13478398e-01 -1.09644520e+00 -3.16426396e-01 4.35159713e-01
2.20564738e-01 -2.80337065e-01 -2.48891383e-01 -2.08760023e-01] | [10.4762601852417, 7.948594570159912] |
8c665a2a-8e8f-4f71-ba87-7ccacf22f2b7 | posterior-ratio-estimation-for-latent | 2002.06410 | null | https://arxiv.org/abs/2002.06410v2 | https://arxiv.org/pdf/2002.06410v2.pdf | Posterior Ratio Estimation of Latent Variables | Density Ratio Estimation has attracted attention from the machine learning community due to its ability to compare the underlying distributions of two datasets. However, in some applications, we want to compare distributions of random variables that are \emph{inferred} from observations. In this paper, we study the problem of estimating the ratio between two posterior probability density functions of a latent variable. Particularly, we assume the posterior ratio function can be well-approximated by a parametric model, which is then estimated using observed information and prior samples. We prove the consistency of our estimator and the asymptotic normality of the estimated parameters as the number of prior samples tending to infinity. Finally, we validate our theories using numerical experiments and demonstrate the usefulness of the proposed method through some real-world applications. | ['Yulong Zhang', 'Mingxuan Yi', 'Song Liu', 'Mladen Kolar'] | 2020-02-15 | null | null | null | null | ['density-ratio-estimation'] | ['methodology'] | [-4.36702892e-02 2.69848271e-04 -3.80496502e-01 -4.02522147e-01
-6.74557626e-01 -2.58111715e-01 4.23091680e-01 6.91315159e-02
-2.53670692e-01 1.07571244e+00 -3.12657952e-01 -2.88183928e-01
-2.24100217e-01 -7.82632411e-01 -6.29634738e-01 -7.83021033e-01
7.72798955e-02 6.10651851e-01 -1.50434345e-01 5.35238981e-01
8.31567124e-02 2.02043757e-01 -1.12485409e+00 -7.19306171e-01
1.14170873e+00 1.08365917e+00 -4.92116511e-02 4.59766179e-01
8.17649364e-02 3.11247110e-01 -6.20186985e-01 -3.01951885e-01
2.41910383e-01 -3.26773137e-01 -3.54672998e-01 1.61456212e-01
1.16552114e-01 -2.85421133e-01 5.48037589e-02 1.47725558e+00
2.35195621e-03 1.59923062e-01 1.40387690e+00 -1.63741517e+00
-4.54859823e-01 3.34881216e-01 -1.11845803e+00 1.66331902e-01
1.16929233e-01 -3.79637331e-01 6.25902355e-01 -5.51923573e-01
-4.03667092e-02 1.17944825e+00 4.55621362e-01 8.83540511e-02
-1.41730320e+00 -7.96866953e-01 -1.83830455e-01 1.03094000e-02
-1.73257852e+00 -2.52197832e-01 5.68450451e-01 -6.21088684e-01
3.45686115e-02 -1.28182128e-01 9.70641449e-02 7.49323606e-01
2.33611390e-01 5.92791617e-01 9.15813446e-01 -3.87308627e-01
4.40006644e-01 5.07590353e-01 8.54236633e-02 3.74661535e-01
6.46873176e-01 -2.16734082e-01 -5.98079786e-02 -5.59527874e-01
9.45748508e-01 1.60500318e-01 -4.26385492e-01 -3.35884959e-01
-6.77154541e-01 9.11045253e-01 -1.76297456e-01 5.20263426e-02
-4.22023475e-01 8.16484615e-02 -1.33223236e-01 -3.68241817e-02
7.08419919e-01 -2.66025752e-01 -6.23230487e-02 -1.30643725e-01
-8.65469992e-01 1.31577834e-01 8.44447076e-01 1.19452918e+00
6.60304487e-01 -1.73245803e-01 7.58823603e-02 7.96498954e-01
4.77467090e-01 8.17767799e-01 2.49930829e-01 -8.62932324e-01
4.20466512e-01 1.93394169e-01 6.29554927e-01 -8.87714505e-01
1.66973159e-01 -2.90196866e-01 -1.05777490e+00 -2.56374419e-01
1.04184878e+00 -2.26396337e-01 -3.56527418e-01 1.83012331e+00
2.31267229e-01 2.21795589e-01 -2.25383043e-02 6.19768858e-01
-9.74563882e-02 8.01595747e-01 -5.77183776e-02 -5.90954840e-01
9.61827099e-01 -2.89993376e-01 -8.48344326e-01 5.36218956e-02
1.21670343e-01 -6.16824925e-01 9.39964712e-01 3.39449793e-01
-1.03953230e+00 -4.06299978e-01 -7.43157864e-01 3.71806353e-01
2.85781085e-01 2.93148994e-01 1.57255396e-01 5.94776869e-01
-4.62822676e-01 3.09922010e-01 -9.33167815e-01 -9.19786915e-02
3.43667537e-01 2.20251996e-02 -1.92775548e-01 2.56875400e-02
-7.76546657e-01 2.34621570e-01 3.88596296e-01 8.20295662e-02
-4.46978271e-01 -6.39714003e-01 -7.25072145e-01 3.76406521e-01
2.20357388e-01 -4.81141478e-01 1.25580299e+00 -7.85200000e-01
-1.36857331e+00 3.89739543e-01 -4.78772610e-01 -2.10335135e-01
7.21849740e-01 -1.98342472e-01 -2.92823792e-01 1.51870772e-01
2.74643302e-01 -8.44606236e-02 9.73443866e-01 -8.95473361e-01
-6.27290845e-01 -4.67644840e-01 -4.10156667e-01 -2.07390979e-01
-2.88327783e-01 -2.93739110e-01 -2.86168814e-01 -3.76983434e-01
1.52146563e-01 -6.99155867e-01 -3.97906415e-02 1.69045091e-01
-3.95849645e-01 -2.24516168e-01 6.24150217e-01 -4.97438431e-01
1.03814232e+00 -2.52462101e+00 -2.10849062e-01 5.06158829e-01
1.28965884e-01 -3.90046686e-01 3.74820769e-01 2.35899597e-01
1.08214952e-01 1.43785298e-01 -3.80588442e-01 -1.53040335e-01
-7.61980750e-03 -1.01805158e-01 -4.63179976e-01 1.13124526e+00
-1.26192674e-01 2.17956141e-01 -7.77260542e-01 -5.10825515e-01
7.61022046e-02 2.50386149e-01 -2.59198099e-01 5.00322521e-01
1.27619445e-01 2.99129933e-01 -4.60998029e-01 2.31463552e-01
1.11888027e+00 -4.92041677e-01 3.72983694e-01 5.68767786e-02
1.45134300e-01 -2.36696437e-01 -1.40654933e+00 5.97708404e-01
-4.31096584e-01 5.86400568e-01 -4.45082076e-02 -1.21615171e+00
1.07112443e+00 2.38209173e-01 3.25822979e-01 -2.30329320e-01
4.24904585e-01 2.70006239e-01 -7.18910024e-02 -2.30631679e-01
1.45356178e-01 -5.05192339e-01 -6.51407912e-02 3.60904187e-01
-1.63128182e-01 6.84450334e-03 3.03027362e-01 -1.26869455e-01
4.33017582e-01 -1.91010356e-01 8.31879139e-01 -3.85685682e-01
4.40593213e-01 -5.86988628e-01 4.29546654e-01 7.45118797e-01
-8.25707167e-02 3.96333277e-01 9.37413275e-01 1.92060739e-01
-1.13858163e+00 -1.50165999e+00 -6.88396871e-01 2.93857068e-01
3.28170210e-01 1.03741586e-01 -6.42852068e-01 -4.66184527e-01
1.23790249e-01 9.04109478e-01 -5.89014411e-01 -1.74849574e-02
-7.81694129e-02 -9.59146798e-01 3.09996843e-01 5.17928600e-01
5.54606676e-01 -2.52990186e-01 -3.79893214e-01 2.38722116e-02
-3.16906929e-01 -1.19691670e+00 -4.92592961e-01 -6.99769035e-02
-8.30078721e-01 -1.19756794e+00 -1.07449186e+00 -4.64342564e-01
7.23172128e-01 3.77591103e-02 7.15848625e-01 -3.65093738e-01
2.40781352e-01 3.40223402e-01 9.69890952e-02 -3.47203761e-01
-5.16948402e-01 -1.29402190e-01 2.10910901e-01 3.24809641e-01
3.09359312e-01 -6.48304403e-01 -2.54033118e-01 6.99708939e-01
-8.89875710e-01 -2.90209532e-01 3.69809091e-01 4.45778280e-01
5.22441685e-01 4.54906970e-01 7.15757966e-01 -5.99838138e-01
6.41536415e-01 -9.00012195e-01 -1.09257555e+00 3.81329149e-01
-3.49087685e-01 2.38373950e-01 7.47099936e-01 -6.75750256e-01
-1.26356685e+00 -3.90705019e-01 4.20837879e-01 -4.52699661e-01
-1.97470769e-01 4.83620644e-01 -4.19102371e-01 4.77320075e-01
1.66152522e-01 9.29362550e-02 1.28999233e-01 -5.20324528e-01
2.62711365e-02 9.48938489e-01 7.72890508e-01 -7.43481874e-01
7.02697575e-01 5.73096514e-01 2.95003027e-01 -9.24339890e-01
-9.21161532e-01 -3.91811460e-01 -3.07570517e-01 -3.28825079e-02
4.66880143e-01 -6.06987774e-01 -9.63802934e-01 5.11101902e-01
-8.23857307e-01 -6.03498034e-02 -1.45508885e-01 8.57839286e-01
-6.51891589e-01 5.44645309e-01 -3.40718299e-01 -1.47217703e+00
8.88130292e-02 -8.61822605e-01 8.29684913e-01 5.20270407e-01
-1.24110721e-01 -1.30418766e+00 2.18703568e-01 -1.58079743e-01
6.86605200e-02 8.30704942e-02 1.00953007e+00 -6.26972854e-01
-1.46163762e-01 -4.05731052e-01 -4.30829287e-01 3.51599216e-01
4.58032221e-01 1.59850821e-01 -6.10142946e-01 -2.15988517e-01
2.97631025e-01 8.19505472e-03 4.56036717e-01 8.45984399e-01
1.32227647e+00 -4.60319519e-01 -5.11093915e-01 4.32904303e-01
1.38128507e+00 7.50456378e-02 5.18110096e-01 -9.89545956e-02
1.56554639e-01 4.22627807e-01 6.98009610e-01 8.42319965e-01
5.92034347e-02 4.18058306e-01 1.68289080e-01 4.46309656e-01
6.78370416e-01 -3.18542540e-01 1.08891360e-01 4.90525961e-01
1.55353919e-01 -5.81061840e-01 -7.54035830e-01 5.63822031e-01
-1.70638883e+00 -9.04723406e-01 -2.38078788e-01 2.68712139e+00
8.35019171e-01 9.60088819e-02 3.27756643e-01 -1.01430602e-02
1.20772433e+00 -3.75801772e-01 -6.78348780e-01 1.03424311e-01
1.03761941e-01 -6.97473511e-02 5.61967313e-01 5.14811337e-01
-7.70641148e-01 1.16496287e-01 6.86544323e+00 1.06502354e+00
-8.54289412e-01 -2.12207139e-01 9.10315752e-01 2.69067675e-01
-2.15274528e-01 8.47380310e-02 -7.90704608e-01 8.72049510e-01
8.32731843e-01 -6.22014582e-01 4.44089659e-02 7.87232339e-01
1.34714246e-01 -6.48041129e-01 -1.15522587e+00 9.13263202e-01
-2.75648206e-01 -6.33025348e-01 -2.87160486e-01 4.77775961e-01
5.66861987e-01 -6.16311312e-01 2.91075110e-01 1.55883640e-01
1.99467137e-01 -9.26272035e-01 5.83708823e-01 8.43811631e-01
8.12047064e-01 -8.36459517e-01 1.00944817e+00 6.83916509e-01
-9.89688873e-01 -1.09722549e-02 -5.58564603e-01 -2.09232375e-01
2.56410807e-01 1.03468001e+00 -9.98954594e-01 1.78985387e-01
4.25752431e-01 2.82051980e-01 -1.92152381e-01 1.33743715e+00
-4.79012839e-02 8.75414371e-01 -6.83629096e-01 -5.08281067e-02
-3.27857077e-01 -6.96339130e-01 3.41775835e-01 6.26628399e-01
5.72022915e-01 -8.70649964e-02 -1.46295223e-02 1.11623228e+00
-5.51152490e-02 4.10338007e-02 -3.04013342e-01 -3.74090970e-02
8.05106461e-01 9.01617527e-01 -7.18585372e-01 -2.95387626e-01
-4.11516756e-01 4.46103692e-01 3.17942411e-01 6.31588340e-01
-1.09358966e+00 -4.25885648e-01 3.79479468e-01 1.70432433e-01
3.20227623e-01 -2.42737770e-01 -6.71942011e-02 -1.10336208e+00
1.34625077e-01 -2.50824451e-01 1.75314054e-01 -4.56494480e-01
-1.58110332e+00 6.01095594e-02 6.48587167e-01 -1.10869062e+00
-5.29126942e-01 -3.05871010e-01 -7.03553438e-01 9.11433697e-01
-1.08259928e+00 -3.92336607e-01 -1.38810441e-01 3.10601324e-01
5.54454327e-02 -2.98395213e-02 2.85817742e-01 1.43466815e-01
-6.50985122e-01 6.05289936e-01 6.25834823e-01 1.23928219e-01
5.18402219e-01 -1.18185318e+00 -2.59506080e-04 5.43754816e-01
-2.32954204e-01 6.06628835e-01 1.03980339e+00 -6.62612140e-01
-6.96476161e-01 -7.70135343e-01 4.32074189e-01 8.75619128e-02
7.35792041e-01 -1.58302903e-01 -9.69911695e-01 7.45743930e-01
-8.43295231e-02 2.05546811e-01 7.31118619e-01 2.18489673e-02
-1.31100073e-01 1.69021171e-02 -1.38268697e+00 2.83930272e-01
2.94782728e-01 -1.34925559e-01 -2.74925321e-01 -3.74906361e-02
1.65402308e-01 -5.33042476e-03 -9.59355175e-01 3.68430495e-01
7.91102886e-01 -8.50787997e-01 6.85470462e-01 -2.93346494e-01
2.33059183e-01 -1.72487646e-01 -4.82062787e-01 -1.21623957e+00
3.97574008e-02 -2.63578117e-01 -2.27148533e-01 1.45102870e+00
9.95916799e-02 -8.00895810e-01 5.68556607e-01 4.96655673e-01
6.92249775e-01 -4.77528721e-01 -9.23180938e-01 -9.94730294e-01
7.49933720e-02 -3.67873430e-01 5.03171146e-01 5.40670753e-01
-8.54598284e-02 2.47857496e-01 -2.99033612e-01 3.59255284e-01
9.50335264e-01 1.28232390e-01 8.17898989e-01 -1.31914103e+00
-5.43537259e-01 -2.22624421e-01 -3.64424616e-01 -1.33640444e+00
3.34278584e-01 -3.36897105e-01 2.03230649e-01 -9.86720145e-01
6.13645852e-01 -3.78002137e-01 1.02406949e-01 -8.67430270e-02
-2.45211527e-01 -3.11109293e-02 -2.10327223e-01 3.29254687e-01
-3.67436796e-01 9.83828902e-01 9.35799241e-01 1.62694335e-01
-9.37259421e-02 5.78422368e-01 -4.59723026e-01 1.01445365e+00
7.56895125e-01 -6.14623547e-01 -5.47946453e-01 1.49404287e-01
2.18712628e-01 4.08956617e-01 4.88748252e-01 -7.88682044e-01
-1.72153980e-01 -4.69871581e-01 2.95686930e-01 -6.45676136e-01
1.21698178e-01 -8.35610330e-01 1.61273405e-02 3.77842262e-02
-1.47428274e-01 -3.25799078e-01 -6.95995614e-02 7.81598270e-01
-2.04726890e-01 -5.59077144e-01 1.00007665e+00 3.58130068e-01
2.04457134e-01 2.05279216e-01 -5.26431024e-01 3.57927620e-01
1.00041986e+00 1.49563616e-02 -2.38374263e-01 -9.35283840e-01
-4.66369927e-01 1.49601579e-01 3.22336406e-01 -2.38613725e-01
4.88638520e-01 -1.31707799e+00 -5.56110084e-01 2.24327788e-01
-1.02397315e-02 1.36251422e-02 1.50732860e-01 1.07905805e+00
-2.13883728e-01 1.72332585e-01 1.54064670e-01 -7.82294393e-01
-7.28494704e-01 5.19442797e-01 1.98807180e-01 2.16664169e-02
-1.23531818e-01 1.22515954e-01 3.96260679e-01 -1.73703790e-01
1.43734828e-01 -3.52111250e-01 2.45576110e-02 -2.02000946e-01
5.02416193e-01 7.73452580e-01 -4.92275894e-01 -3.73019785e-01
-2.33213052e-01 4.02310640e-01 -1.70690138e-02 -3.23581547e-01
1.03722799e+00 -4.03765917e-01 4.78015691e-02 7.93662488e-01
1.49383473e+00 2.62515485e-01 -1.41610551e+00 -2.86158323e-01
-5.17390668e-02 -6.81153238e-01 -2.37743601e-01 -7.92679042e-02
-8.19078326e-01 7.20429838e-01 4.79979277e-01 5.94900787e-01
9.73294497e-01 2.56039798e-01 4.02037710e-01 2.26670370e-01
1.33303538e-01 -7.76313007e-01 -1.01462536e-01 6.95117936e-02
3.89575005e-01 -1.10266817e+00 5.88387176e-02 -4.53722388e-01
-2.54208863e-01 9.43436801e-01 4.51827794e-01 -2.74277180e-01
9.81025398e-01 2.16255963e-01 -3.67856413e-01 3.01490098e-01
-4.19708729e-01 2.16751099e-01 3.19663078e-01 6.03857219e-01
4.15405869e-01 1.62327498e-01 -2.39659816e-01 6.72761858e-01
-3.09559941e-01 2.16165762e-02 7.63605297e-01 4.47562575e-01
-3.97222608e-01 -8.48961771e-01 -5.16803324e-01 5.44902265e-01
-5.23005009e-01 1.56729296e-01 2.68639922e-01 8.51617694e-01
-3.84434372e-01 8.97148252e-01 4.53402609e-01 4.05225426e-01
-1.58240974e-01 -1.78492740e-01 5.95102668e-01 -2.76208133e-01
7.29039669e-01 3.98252338e-01 -4.77313995e-01 4.46917936e-02
-3.58705431e-01 -1.05666339e+00 -9.33259070e-01 -4.02048945e-01
-5.70596516e-01 3.64391178e-01 3.89259428e-01 1.18150043e+00
-1.65425971e-01 1.39400110e-01 7.11735904e-01 -2.59303004e-01
-8.73758197e-01 -1.13900411e+00 -1.14063907e+00 1.01906374e-01
3.08490187e-01 -7.44905293e-01 -6.36146843e-01 -3.40517350e-02] | [7.192014694213867, 4.0993266105651855] |
409c3ec9-9964-4a30-a0ef-e408cada3e61 | three-sentences-are-all-you-need-local-path | 2106.01793 | null | https://arxiv.org/abs/2106.01793v1 | https://arxiv.org/pdf/2106.01793v1.pdf | Three Sentences Are All You Need: Local Path Enhanced Document Relation Extraction | Document-level Relation Extraction (RE) is a more challenging task than sentence RE as it often requires reasoning over multiple sentences. Yet, human annotators usually use a small number of sentences to identify the relationship between a given entity pair. In this paper, we present an embarrassingly simple but effective method to heuristically select evidence sentences for document-level RE, which can be easily combined with BiLSTM to achieve good performance on benchmark datasets, even better than fancy graph neural network based methods. We have released our code at https://github.com/AndrewZhe/Three-Sentences-Are-All-You-Need. | ['Dongyan Zhao', 'Yuxuan Lai', 'Yuan Ye', 'Yansong Feng', 'Shengqi Zhu', 'Quzhe Huang'] | 2021-06-03 | null | https://aclanthology.org/2021.acl-short.126 | https://aclanthology.org/2021.acl-short.126.pdf | acl-2021-5 | ['document-level-relation-extraction'] | ['natural-language-processing'] | [-3.14469673e-02 4.17195559e-01 -3.43027502e-01 -4.65632409e-01
-1.15412271e+00 -5.08679092e-01 2.77408749e-01 6.27531290e-01
-4.59188282e-01 1.04477644e+00 2.52009243e-01 -6.43893242e-01
-5.37160374e-02 -9.40749824e-01 -6.80063128e-01 -1.42084166e-01
1.18223708e-02 5.37856102e-01 2.81361431e-01 -2.91747659e-01
1.30085021e-01 1.93146154e-01 -9.39565122e-01 3.52503389e-01
6.43178403e-01 6.97807372e-01 5.52248321e-02 8.05481374e-01
-1.76287130e-01 1.18278837e+00 -6.13975286e-01 -8.18988025e-01
-2.95920908e-01 -4.88017142e-01 -1.41691339e+00 -2.95873702e-01
4.27900344e-01 -1.91105247e-01 -2.68523425e-01 1.11278617e+00
4.21943784e-01 1.13000646e-01 4.02904749e-01 -7.94548810e-01
-6.44798100e-01 1.25477016e+00 -4.02667612e-01 7.62135088e-01
5.66089749e-01 -1.98169917e-01 1.71568859e+00 -1.02016401e+00
6.27588332e-01 1.03958666e+00 4.75056916e-01 5.09451747e-01
-8.97383332e-01 -5.36255896e-01 5.99008463e-02 5.26873827e-01
-1.43116796e+00 -4.72619772e-01 9.00264084e-01 -5.36373332e-02
1.53453374e+00 3.31060797e-01 3.78541768e-01 1.00429916e+00
7.19419122e-02 9.03495729e-01 6.97172880e-01 -5.23708165e-01
-2.24360749e-01 -2.10730404e-01 4.30323660e-01 8.00783277e-01
3.69963974e-01 -5.40559232e-01 -3.53924185e-01 -4.38337326e-02
1.71099380e-01 -1.70264959e-01 -3.32495630e-01 4.27221239e-01
-1.10561264e+00 8.73653293e-01 5.63055813e-01 6.16491973e-01
-3.21066082e-01 2.87028879e-01 5.33262551e-01 2.80487686e-01
6.04963362e-01 6.77510858e-01 -6.13315344e-01 -2.17858180e-01
-8.21470976e-01 1.10626198e-01 8.88155878e-01 8.67673576e-01
4.38520908e-01 -3.21376085e-01 -3.98268133e-01 9.28677499e-01
1.07764997e-01 1.47987247e-01 2.29725882e-01 -5.53379357e-01
8.13253045e-01 5.09908915e-01 -1.69653922e-01 -1.31740034e+00
-3.75488639e-01 -2.61256993e-01 -1.00835836e+00 -5.21517456e-01
1.35486260e-01 -4.71262217e-01 -6.15554035e-01 1.56959069e+00
1.08261883e-01 -1.00745104e-01 1.20025329e-01 6.21635437e-01
1.48535824e+00 5.96697986e-01 2.04933926e-01 -1.26088306e-01
1.51069903e+00 -1.04292142e+00 -9.24008012e-01 -5.92098713e-01
8.12065780e-01 -6.05226219e-01 9.46084857e-01 1.08881004e-01
-1.18896103e+00 -1.74207345e-01 -1.04401994e+00 -4.35149014e-01
-3.40550125e-01 2.48075143e-01 5.01423717e-01 9.80872959e-02
-9.89955902e-01 6.73464179e-01 -4.98774797e-01 -2.14185178e-01
4.58286017e-01 3.04445148e-01 -3.88693422e-01 6.24694973e-02
-1.74493051e+00 1.22628307e+00 7.74751425e-01 3.92930627e-01
-3.76717508e-01 -8.53490308e-02 -1.11110532e+00 2.59639561e-01
8.76274645e-01 -6.19952679e-01 1.41866744e+00 -3.32672119e-01
-1.11898291e+00 9.42690015e-01 -4.87643927e-01 -5.47386527e-01
1.57830998e-01 -3.47821474e-01 -3.26944679e-01 1.50402263e-01
1.01624630e-01 2.35006750e-01 2.46611595e-01 -9.99987900e-01
-4.30896699e-01 -3.73720825e-02 1.64462104e-01 1.43418863e-01
-1.13205254e-01 3.71617049e-01 -3.47885013e-01 -4.78333652e-01
-8.85819644e-02 -6.31490707e-01 -2.07501128e-01 -6.00905895e-01
-1.07689273e+00 -9.34469700e-01 3.67397726e-01 -9.18212414e-01
1.61356640e+00 -1.75694120e+00 9.21505131e-03 -2.04229638e-01
5.05216360e-01 2.96587378e-01 -5.72692566e-02 7.21452296e-01
-8.87204185e-02 6.15921259e-01 -2.23138943e-01 -1.58214688e-01
-1.31886423e-01 -9.46018547e-02 -1.90061331e-01 1.19698524e-01
5.42171896e-01 1.20550823e+00 -1.14249337e+00 -8.15452576e-01
-1.89324498e-01 1.56236157e-01 2.72638659e-04 1.62473366e-01
-3.61007690e-01 4.46536243e-02 -5.28845370e-01 5.25071323e-01
2.18244702e-01 -6.78994238e-01 4.40758944e-01 -7.81717300e-02
4.24654454e-01 1.09017181e+00 -6.70496643e-01 1.12014949e+00
-4.15308237e-01 8.46687913e-01 -3.18593979e-01 -9.68248844e-01
9.39858496e-01 2.90465683e-01 -3.60732675e-02 -4.54008847e-01
2.85384685e-01 1.30889982e-01 8.04650038e-02 -7.46447742e-01
4.90895003e-01 -1.78687498e-01 -1.56815499e-01 3.37900281e-01
-1.41105230e-03 -1.81754306e-01 7.47590780e-01 4.07076180e-01
1.38177598e+00 -1.89365789e-01 7.61402249e-01 1.42805383e-01
5.89942575e-01 -3.78050432e-02 6.38122320e-01 7.32747138e-01
-2.39952863e-03 3.97032320e-01 8.10246646e-01 -2.97405571e-01
-8.17218423e-01 -6.21620893e-01 1.51479691e-01 7.92369545e-01
-8.76986235e-03 -8.23163867e-01 -4.16908205e-01 -1.12155735e+00
-3.82166147e-01 1.10241354e+00 -3.86084557e-01 -6.24089539e-02
-8.18344355e-01 -5.68846881e-01 5.16561210e-01 6.13260150e-01
5.23628473e-01 -1.32242131e+00 -1.16961740e-01 2.25014552e-01
-6.03904963e-01 -1.40259504e+00 -4.15093631e-01 3.01954091e-01
-5.51131487e-01 -1.00303555e+00 -3.32488537e-01 -8.75003159e-01
5.90875387e-01 1.14461638e-01 1.47345376e+00 5.26617825e-01
5.69328154e-03 -1.71423689e-01 -6.70238733e-01 -2.88857788e-01
-4.20528650e-01 3.28629494e-01 -3.16573024e-01 -5.28526366e-01
7.61570275e-01 -3.46301615e-01 -3.89902055e-01 -2.02029780e-01
-5.67813396e-01 6.49837628e-02 5.11352301e-01 7.66512215e-01
6.04230404e-01 1.94261044e-01 6.14590645e-01 -1.09893310e+00
1.21118188e+00 -5.04757285e-01 -2.56592780e-01 5.63260555e-01
-5.55721700e-01 8.55241790e-02 6.50636911e-01 -1.72979921e-01
-6.22556448e-01 -3.18023562e-01 -3.47237676e-01 1.58056110e-01
-4.98485602e-02 1.09294558e+00 1.11874387e-01 4.99644518e-01
5.07684350e-01 -4.33215126e-02 -4.20681059e-01 -2.28629753e-01
3.08359295e-01 7.81271458e-01 1.88756093e-01 -5.17245710e-01
6.49382532e-01 -1.16249062e-01 -3.61386538e-02 -4.93670702e-01
-1.54346776e+00 -1.97238877e-01 -5.45172572e-01 -1.48041919e-01
9.22352314e-01 -6.31075144e-01 -7.65324414e-01 1.47059575e-01
-1.46337390e+00 -4.77662861e-01 -1.28851449e-02 1.64049685e-01
8.66773576e-02 4.28882152e-01 -1.00700772e+00 -6.00789905e-01
-8.82430553e-01 -7.47148216e-01 7.48049676e-01 2.98999071e-01
-4.18540597e-01 -1.00195765e+00 -6.28948137e-02 4.92424130e-01
6.11958019e-02 1.81303635e-01 9.36322391e-01 -9.43648040e-01
-3.11631083e-01 -4.69707489e-01 -3.35172534e-01 3.17304224e-01
2.10685313e-01 2.97446012e-01 -6.38654470e-01 -4.19624075e-02
-3.00056875e-01 -4.91227329e-01 9.32647169e-01 1.87839866e-01
1.31956446e+00 -6.03325486e-01 -3.94118428e-01 -5.36565445e-02
1.24713874e+00 5.03408723e-02 4.65749949e-01 3.47045451e-01
8.20963264e-01 4.77674484e-01 5.93015730e-01 4.86357063e-02
8.75993073e-01 5.24330378e-01 5.25324605e-02 -5.79888336e-02
-2.26141274e-01 -2.64296263e-01 1.56421155e-01 1.04604948e+00
-4.75733075e-03 -6.59967899e-01 -1.12845898e+00 5.91682315e-01
-1.91445899e+00 -1.12857914e+00 -3.66624475e-01 1.75982022e+00
1.25007522e+00 4.88713592e-01 -1.35519385e-01 6.57875091e-02
8.97264063e-01 2.64182925e-01 -1.53694451e-01 -6.57880485e-01
-1.14248976e-01 1.17702626e-01 3.22766244e-01 8.67731333e-01
-1.15384924e+00 1.24315476e+00 5.94358110e+00 8.46135795e-01
-8.28205228e-01 1.56985059e-01 8.77498031e-01 9.50151384e-02
-4.11730886e-01 1.17735304e-01 -1.15358794e+00 2.92135417e-01
1.18582284e+00 -3.23617995e-01 1.38325810e-01 7.15216041e-01
1.05264366e-01 -2.25852519e-01 -1.08732855e+00 5.98117709e-01
1.36186451e-01 -1.47344995e+00 -1.61238164e-01 -3.12147021e-01
2.66296178e-01 1.30763710e-01 -4.77695197e-01 4.64116693e-01
5.40167391e-01 -1.11970556e+00 4.30037498e-01 2.47804448e-01
6.34841800e-01 -4.91265625e-01 1.02092469e+00 3.22115809e-01
-1.13607717e+00 1.24439336e-01 -2.91056335e-01 -1.23691000e-01
3.80987346e-01 1.11510730e+00 -9.86509860e-01 5.87128043e-01
7.80144215e-01 6.80112362e-01 -7.35863090e-01 7.82167196e-01
-1.01494229e+00 7.75954723e-01 -2.34274283e-01 -6.65054202e-01
7.66230077e-02 1.03660271e-01 3.93173635e-01 1.42576122e+00
2.16522783e-01 2.04528809e-01 5.56286424e-02 6.61924899e-01
-5.15490353e-01 1.41781837e-01 -7.50520825e-01 -2.55574942e-01
6.10103488e-01 1.38789248e+00 -7.54787445e-01 -5.05231619e-01
-3.94777536e-01 7.49455631e-01 9.94842410e-01 2.70471811e-01
-7.73424804e-01 -6.71346188e-01 -2.05579791e-02 -1.50908276e-01
3.68334055e-01 -2.80712992e-01 -1.59011111e-01 -1.16090512e+00
4.40761715e-01 -5.57604730e-01 6.89288914e-01 -8.77049029e-01
-1.48736286e+00 9.26966667e-01 -6.48097694e-02 -1.00526071e+00
-3.65538627e-01 -4.53202635e-01 -7.68393338e-01 6.62258446e-01
-1.50055003e+00 -9.46460485e-01 -5.28090121e-03 8.96298140e-02
5.11581182e-01 1.75607935e-01 7.42695391e-01 2.84765393e-01
-9.44636643e-01 5.05865574e-01 -3.93269330e-01 7.51099527e-01
3.62512976e-01 -1.40303504e+00 5.86940646e-01 1.06587553e+00
4.42350388e-01 6.61249161e-01 6.60014749e-01 -7.26420999e-01
-1.04083037e+00 -8.90742481e-01 1.76519549e+00 -3.63508046e-01
1.03252316e+00 -4.59010154e-01 -9.37606454e-01 9.28446412e-01
5.97121596e-01 -9.22791734e-02 6.65406704e-01 4.20460701e-01
-2.79257894e-01 -6.73036799e-02 -7.84567535e-01 7.52145529e-01
8.65244150e-01 -6.58628702e-01 -9.15567636e-01 7.24149585e-01
8.59355807e-01 -4.87906903e-01 -9.20186400e-01 5.70282936e-01
-1.47819057e-01 -5.09457111e-01 7.29446888e-01 -5.98933101e-01
8.31893444e-01 -1.93404317e-01 1.57627285e-01 -1.14328039e+00
-3.50215226e-01 -4.71759677e-01 -3.77214938e-01 1.44942546e+00
1.06873989e+00 -5.93004704e-01 4.65641707e-01 5.71597576e-01
1.50109634e-01 -1.18462241e+00 -8.77120972e-01 -7.10278571e-01
1.26778781e-01 -2.92788774e-01 3.35349917e-01 9.48063791e-01
4.13577974e-01 1.09984469e+00 -1.93523437e-01 1.00026436e-01
2.31856406e-01 2.02370107e-01 2.62158900e-01 -9.37584043e-01
-2.31229827e-01 -6.51612580e-01 -4.81039621e-02 -7.91534364e-01
5.69543183e-01 -1.22668993e+00 3.15842122e-01 -2.20955610e+00
4.02981043e-01 -9.65013653e-02 -3.38658094e-01 8.83118391e-01
-5.45110345e-01 9.00835991e-02 -8.42389390e-02 -5.32488041e-02
-8.37757766e-01 5.25161624e-01 1.22518945e+00 -3.62155020e-01
7.55886585e-02 5.25160227e-03 -9.29086924e-01 4.86003190e-01
1.10441113e+00 -6.64840162e-01 -1.84521005e-01 -4.24641430e-01
8.05905461e-01 1.26310408e-01 1.57222942e-01 -5.35433948e-01
2.36960381e-01 -1.70598820e-01 2.23456800e-01 -5.59686542e-01
4.41086851e-02 -2.49624535e-01 -3.18001449e-01 2.61412442e-01
-6.50738895e-01 1.76781774e-01 1.44491762e-01 2.53230661e-01
-4.54735994e-01 -6.17866755e-01 3.09903711e-01 -2.38823131e-01
-5.73798716e-01 2.68638849e-01 -3.76405090e-01 3.46013129e-01
7.27695465e-01 4.01520103e-01 -6.58548236e-01 -6.54570520e-01
-5.98502517e-01 4.02462840e-01 -1.85200155e-01 4.25478220e-01
9.18034077e-01 -1.08801103e+00 -9.82761621e-01 -5.38850784e-01
2.68516019e-02 3.88185501e-01 2.63016075e-02 7.32511282e-01
-4.64161366e-01 3.89936149e-01 4.03852254e-01 -7.22075105e-02
-1.45146942e+00 4.82983530e-01 2.09297910e-01 -5.81235707e-01
-8.59808922e-01 1.14825535e+00 -6.08613670e-01 -2.98933536e-01
-4.83351983e-02 -2.94235617e-01 -4.74373639e-01 4.82065789e-02
3.75569612e-01 -3.08723841e-02 2.50182790e-03 -3.94081891e-01
-5.69757998e-01 2.64330536e-01 -3.01773816e-01 4.77782711e-02
1.40812588e+00 -4.79105227e-02 -6.51032746e-01 4.64565277e-01
1.07542312e+00 5.21684475e-02 -6.01848781e-01 -2.76902050e-01
5.25331020e-01 -1.54354116e-02 3.52502917e-03 -7.37267077e-01
-8.64733338e-01 7.56832659e-01 -1.70008942e-01 6.87080860e-01
8.55722427e-01 4.97986525e-01 8.81705463e-01 7.20030606e-01
6.19094670e-02 -1.08361578e+00 -8.03215355e-02 7.13443935e-01
1.02432299e+00 -1.50663316e+00 3.74352068e-01 -5.69421351e-01
-6.84698939e-01 1.07272196e+00 6.69744730e-01 -1.31522968e-01
6.12794280e-01 3.22478563e-01 1.02106156e-02 -5.14304996e-01
-9.54554498e-01 -3.62799376e-01 4.34661239e-01 1.34993598e-01
8.97752106e-01 1.37917429e-01 -6.68279111e-01 5.98468542e-01
-2.54495353e-01 -1.77239403e-01 6.65381730e-01 6.81478143e-01
-2.62130827e-01 -1.18688607e+00 3.19952428e-01 7.37029910e-01
-7.60662079e-01 -6.58976257e-01 -5.72046041e-01 6.35644674e-01
-3.51751059e-01 1.29713893e+00 -2.47203156e-01 -4.37395364e-01
1.43842772e-01 9.17713717e-02 3.73372406e-01 -8.20363343e-01
-5.33673346e-01 -1.49384797e-01 9.35722947e-01 -3.16070050e-01
-5.37032604e-01 -5.27204990e-01 -1.57363427e+00 -3.28415483e-01
-3.53835940e-01 2.23252550e-01 3.14528257e-01 1.13439894e+00
3.99905652e-01 5.06854534e-01 3.54383469e-01 -2.01458946e-01
-1.38859972e-01 -1.33833790e+00 -1.38015896e-01 2.54103243e-01
2.38131106e-01 -3.44611943e-01 -4.30161208e-01 -2.61099577e-01] | [9.373714447021484, 8.697059631347656] |
867b92f1-ba39-4be9-8e25-3763643161ab | talking-heads-signing-avatars-and-social | null | null | https://aclanthology.org/W15-5101 | https://aclanthology.org/W15-5101.pdf | Talking Heads, Signing Avatars and Social Robots | null | ['Jonas Beskow'] | 2015-09-01 | null | null | null | ws-2015-9 | ['lipreading'] | ['computer-vision'] | [-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] | [-7.283622741699219, 3.7357709407806396] |
1bc5e069-5292-4fd3-945f-675245735ecc | perceptual-speech-enhancement-via-generative | 1910.12620 | null | https://arxiv.org/abs/1910.12620v3 | https://arxiv.org/pdf/1910.12620v3.pdf | AeGAN: Time-Frequency Speech Denoising via Generative Adversarial Networks | Automatic speech recognition (ASR) systems are of vital importance nowadays in commonplace tasks such as speech-to-text processing and language translation. This created the need for an ASR system that can operate in realistic crowded environments. Thus, speech enhancement is a valuable building block in ASR systems and other applications such as hearing aids, smartphones and teleconferencing systems. In this paper, a generative adversarial network (GAN) based framework is investigated for the task of speech enhancement, more specifically speech denoising of audio tracks. A new architecture based on CasNet generator and an additional feature-based loss are incorporated to get realistically denoised speech phonetics. Finally, the proposed framework is shown to outperform other learning and traditional model-based speech enhancement approaches. | ['Jayasankar T. Sajeev', 'Karim Armanious', 'Karim Guirguis', 'Sherif Abdulatif', 'Bin Yang'] | 2019-10-21 | null | null | null | null | ['speech-denoising'] | ['speech'] | [ 4.76459116e-01 9.64550972e-02 6.13236606e-01 -2.20900491e-01
-9.87434387e-01 -2.55107611e-01 7.96042442e-01 -1.50516063e-01
-4.29793626e-01 6.99864209e-01 4.04607475e-01 -5.50575972e-01
1.51273429e-01 -5.46268284e-01 -4.74782795e-01 -8.64498019e-01
4.21970695e-01 3.60853188e-02 3.29361064e-03 -5.89270771e-01
-8.15564319e-02 6.30201817e-01 -1.52335227e+00 8.19409788e-02
8.80612612e-01 7.99472392e-01 4.28333730e-01 9.15003181e-01
-1.14975549e-01 6.47413731e-01 -9.83429253e-01 -6.06638610e-01
1.85538128e-01 -3.61021221e-01 -2.01599583e-01 1.24166712e-01
-9.44256708e-02 -2.15686008e-01 -6.29128456e-01 1.15608501e+00
9.38664436e-01 4.84915018e-01 3.39639038e-01 -9.09989893e-01
-4.94910210e-01 5.98804176e-01 -4.56862926e-01 4.39143740e-02
1.76793531e-01 2.03158617e-01 4.46214348e-01 -7.71359444e-01
7.26726204e-02 1.41421044e+00 2.63061106e-01 7.76453733e-01
-7.36092269e-01 -6.90151989e-01 5.22825532e-02 2.70504802e-01
-1.14800513e+00 -9.27846491e-01 9.01581109e-01 -4.29247469e-02
6.95057213e-01 3.42539996e-01 4.04027730e-01 1.26200032e+00
-5.07958010e-02 7.49449015e-01 1.15908659e+00 -4.66693193e-01
2.46960789e-01 1.60713151e-01 -5.74916184e-01 2.55674850e-02
-1.32854357e-01 2.90978372e-01 -4.89356607e-01 2.49785691e-01
7.42823720e-01 -1.68370217e-01 -1.30786553e-01 4.27182257e-01
-8.32759261e-01 6.53432131e-01 2.16053039e-01 4.11351919e-01
-5.81626713e-01 9.71907601e-02 3.96600485e-01 3.48114997e-01
7.32724190e-01 1.32270023e-01 -8.73111859e-02 -4.01974946e-01
-1.06656194e+00 -4.08879071e-02 2.32099876e-01 8.32612336e-01
1.35296449e-01 8.26875627e-01 -3.86447571e-02 1.18176651e+00
5.38625896e-01 7.11868584e-01 6.83272719e-01 -6.76021039e-01
4.80035454e-01 -6.06255047e-02 -9.70825553e-02 -6.57102168e-01
-6.22851867e-03 -7.34478116e-01 -1.32803011e+00 2.94206023e-01
-2.22345799e-01 -2.37630203e-01 -1.12651205e+00 1.53207242e+00
3.47669810e-01 5.58814824e-01 2.88377613e-01 8.40955377e-01
7.62116373e-01 9.97411609e-01 -1.80130988e-01 -2.25793257e-01
1.16718018e+00 -8.61558080e-01 -1.03714168e+00 -1.16329268e-01
-2.40575708e-02 -1.17696440e+00 9.26377118e-01 4.66043532e-01
-1.22660089e+00 -7.27323294e-01 -8.60359967e-01 6.08710758e-02
-1.37306586e-01 -8.48426111e-03 1.81196854e-01 9.24296081e-01
-1.10140252e+00 1.42761439e-01 -8.26977730e-01 -2.11372718e-01
1.79225340e-01 3.48247379e-01 -3.09763700e-01 -1.17150880e-02
-1.22564209e+00 6.53700709e-01 8.89970064e-02 5.16475379e-01
-1.00666559e+00 -3.16900104e-01 -1.00529754e+00 1.62810609e-01
2.62716059e-02 -5.58869839e-01 1.35700929e+00 -1.00535333e+00
-2.13520932e+00 5.40554702e-01 -2.52392709e-01 -6.13479614e-01
6.38323188e-01 -2.34647557e-01 -7.57264197e-01 -1.06680982e-01
-4.87362206e-01 3.02281976e-01 1.33999944e+00 -1.12044525e+00
-4.25504059e-01 -2.79322952e-01 -2.72631258e-01 3.77173930e-01
-2.46749073e-01 3.63090277e-01 -2.45852321e-01 -1.07084298e+00
2.01733839e-02 -6.37444496e-01 -2.52109200e-01 -4.49354857e-01
-4.11159128e-01 1.52477562e-01 1.03508937e+00 -1.30247521e+00
9.67090905e-01 -2.17892694e+00 -4.29630885e-03 1.57250866e-01
-1.53296709e-01 7.73942947e-01 -1.09082162e-01 3.36377770e-01
-3.16489525e-02 -1.46505639e-01 -2.79794008e-01 -7.33852684e-01
-8.29265267e-02 -9.89329070e-03 -3.33375454e-01 3.23506236e-01
1.93602487e-01 4.70334560e-01 -5.65132797e-01 -2.06961751e-01
6.76167786e-01 1.06813347e+00 -4.09244627e-01 4.06119019e-01
7.50987381e-02 8.72463882e-01 -1.16727747e-01 4.34824228e-01
6.58941925e-01 5.99704802e-01 -1.87222898e-01 5.39389513e-02
-8.11226889e-02 3.37999642e-01 -1.16995239e+00 1.37942743e+00
-1.03048337e+00 7.91578650e-01 6.80399001e-01 -9.58192289e-01
1.29104745e+00 8.83267522e-01 1.44393280e-01 -6.62065327e-01
2.37502903e-01 2.63495266e-01 6.54642284e-02 -1.99085250e-01
6.02864683e-01 -2.37873927e-01 2.43698120e-01 5.31509593e-02
-1.23198144e-01 -4.26838934e-01 -3.47216755e-01 2.92159449e-02
1.09365392e+00 -3.01299095e-01 5.00168540e-02 2.01281786e-01
8.67844105e-01 -5.73756099e-01 4.22668993e-01 3.33014697e-01
-1.53731495e-01 7.55303919e-01 -2.38184169e-01 1.94858596e-01
-1.30683017e+00 -9.40901220e-01 2.11423263e-01 6.15841329e-01
-1.48997843e-01 1.31983936e-01 -8.96096230e-01 -2.10651055e-01
-4.36624050e-01 6.74022853e-01 1.79881626e-03 -1.76077768e-01
-6.34155929e-01 -3.49421799e-01 6.71459019e-01 4.19076294e-01
6.06635928e-01 -1.13695991e+00 1.89534575e-01 3.74706328e-01
-2.02082291e-01 -1.27803135e+00 -4.98712987e-01 2.33546346e-02
-6.29556894e-01 -2.47885823e-01 -1.11685479e+00 -1.06594789e+00
4.54011589e-01 2.94106215e-01 5.63989222e-01 -1.98338434e-01
-3.93785611e-02 3.10820222e-01 -5.06409168e-01 -4.98921275e-01
-1.06577075e+00 7.55218640e-02 2.17092171e-01 4.50840145e-01
-3.38498019e-02 -8.07766020e-01 -5.73290765e-01 2.69375712e-01
-1.03535235e+00 -1.52056351e-01 6.60769224e-01 1.11688137e+00
4.14436936e-01 3.86836797e-01 1.15680122e+00 -4.53226537e-01
1.03810751e+00 -2.05130070e-01 -7.70786583e-01 1.56752393e-01
-9.94558558e-02 -3.03936899e-01 9.71469879e-01 -4.35767740e-01
-1.59997869e+00 -8.20813254e-02 -1.11602736e+00 -4.18181002e-01
-3.92213941e-01 3.49661767e-01 -7.52664685e-01 5.71533851e-02
3.83019865e-01 4.28675860e-01 -4.21067588e-02 -6.14518166e-01
2.70391524e-01 1.47881866e+00 7.74250627e-01 -2.06165627e-01
9.38387632e-01 1.60717860e-01 -1.91017181e-01 -1.45615387e+00
-1.21527523e-01 -6.90080583e-01 -1.52844250e-01 -4.23562787e-02
6.46905363e-01 -1.06588340e+00 -6.17820740e-01 8.15630376e-01
-1.40230525e+00 -2.57226765e-01 -1.71280414e-01 7.10975409e-01
-4.92842346e-01 3.33125740e-01 -6.87121391e-01 -1.22703254e+00
-6.21871352e-01 -1.40883338e+00 1.07535887e+00 2.72279888e-01
2.65389472e-01 -8.52067530e-01 4.10616994e-02 5.90832174e-01
7.61281908e-01 -2.27661282e-01 4.54248577e-01 -6.20848298e-01
-4.34441745e-01 -2.03336269e-01 1.30080149e-01 9.58470881e-01
3.45013201e-01 -1.51265666e-01 -1.12587583e+00 -3.66532624e-01
1.41115919e-01 1.62227545e-02 5.92035234e-01 4.57171381e-01
9.57767189e-01 -2.89061040e-01 1.93945736e-01 5.33485889e-01
9.97159898e-01 6.76608860e-01 1.12144876e+00 -2.25183871e-02
6.45869970e-01 4.85035837e-01 4.46621388e-01 3.70249093e-01
-1.99780300e-01 8.21739018e-01 2.43614301e-01 -3.56832981e-01
-4.17591691e-01 -2.88995951e-01 5.42375207e-01 1.25870299e+00
8.78152624e-02 -5.18662810e-01 -5.68906724e-01 6.80657089e-01
-1.40495360e+00 -9.87881660e-01 1.45301282e-01 2.21672440e+00
6.68721735e-01 -5.89462742e-03 -1.17239416e-01 5.64042807e-01
1.13170707e+00 7.63721094e-02 -4.80099320e-01 -6.54027402e-01
-1.77567646e-01 6.80508077e-01 2.21937969e-01 7.04046071e-01
-8.63603473e-01 1.03920841e+00 5.14637232e+00 1.09542418e+00
-1.32541442e+00 3.44933182e-01 5.29539347e-01 3.21171284e-01
-2.61306405e-01 -5.51049650e-01 -4.72200543e-01 4.59527522e-01
8.98880184e-01 5.12395985e-03 6.23842895e-01 5.96752465e-01
8.95836473e-01 2.45218173e-01 -4.83536750e-01 1.18307221e+00
-4.65028547e-02 -8.22613239e-01 -2.58983038e-02 -6.03399873e-02
6.75612628e-01 -1.78201452e-01 3.24046671e-01 3.95873412e-02
2.56034136e-01 -1.06403172e+00 4.98235106e-01 5.05054630e-02
8.37602973e-01 -9.44226861e-01 1.01495981e+00 2.41710365e-01
-1.10062349e+00 9.87193733e-02 -3.29231322e-02 1.83622524e-01
5.94841301e-01 6.61039770e-01 -1.00792468e+00 5.04826486e-01
4.42001343e-01 2.46261984e-01 -2.55346633e-02 1.21459138e+00
-4.85015422e-01 1.05311215e+00 -2.15401456e-01 1.46198422e-01
-4.03678752e-02 -1.44524544e-01 9.03327286e-01 1.17897153e+00
6.87191427e-01 -9.55317728e-03 -4.12915438e-01 3.11142713e-01
-3.00276250e-01 1.58648029e-01 -6.45573676e-01 -9.08794180e-02
5.70588946e-01 1.15355349e+00 -5.28810859e-01 -1.70174152e-01
-1.64375573e-01 1.13527274e+00 -3.60726118e-01 3.88856322e-01
-8.94831121e-01 -6.64798617e-01 8.50392640e-01 2.32765734e-01
2.69469440e-01 -3.97487491e-01 -6.48486242e-02 -1.06329596e+00
5.94665073e-02 -1.22862828e+00 -3.17509174e-01 -8.41808617e-01
-1.06577551e+00 9.10311997e-01 -5.81042588e-01 -1.15667570e+00
-3.30621779e-01 -3.50031674e-01 -8.02429318e-01 1.10580087e+00
-1.66845179e+00 -1.20379615e+00 -2.95016319e-01 7.81743646e-01
8.61819267e-01 -4.64352101e-01 5.67703068e-01 7.42754102e-01
-4.74701673e-01 8.83284748e-01 4.30884421e-01 8.78601298e-02
6.25772893e-01 -9.32716548e-01 6.72048688e-01 1.30600607e+00
2.23869354e-01 4.63279396e-01 8.44278216e-01 -5.36446035e-01
-9.88695383e-01 -1.37147820e+00 5.44384897e-01 3.47552776e-01
3.13853145e-01 -4.33597088e-01 -7.69232213e-01 4.72815543e-01
5.58330119e-01 -2.63575554e-01 3.92388672e-01 -4.43001777e-01
1.64493188e-01 -4.76665407e-01 -1.17520070e+00 6.03078723e-01
6.89572752e-01 -6.26766384e-01 -2.69470304e-01 9.77185816e-02
9.18930650e-01 -3.74164015e-01 -4.77944195e-01 2.67945409e-01
3.55182171e-01 -6.85857654e-01 9.14370239e-01 -2.82690376e-01
1.60532042e-01 -4.31498796e-01 -1.69108704e-01 -1.69297826e+00
2.16823131e-01 -1.28751600e+00 8.44104588e-02 1.69642544e+00
2.70495564e-01 -7.92062640e-01 7.36474037e-01 7.25186020e-02
-4.98947740e-01 -2.84734398e-01 -1.08685017e+00 -7.11205065e-01
-3.00469130e-01 -5.57491004e-01 5.82717657e-01 5.06185472e-01
-4.78928924e-01 2.77535111e-01 -6.94889069e-01 4.23367620e-01
5.36413431e-01 -7.55688310e-01 7.25665808e-01 -8.06548357e-01
-5.10868073e-01 -1.40506238e-01 -5.41828632e-01 -1.14909303e+00
1.23292156e-01 -4.99160558e-01 4.08858120e-01 -1.57661116e+00
-4.72605199e-01 -3.80897999e-01 -2.87446231e-01 -9.12802741e-02
-1.00668661e-01 1.48152113e-01 1.00719228e-01 -2.98973531e-01
-6.32506013e-02 9.98116195e-01 1.14566493e+00 -2.68434286e-01
-2.68465102e-01 5.90468526e-01 -5.41813493e-01 5.84517360e-01
1.00577521e+00 -3.53919983e-01 -5.74358702e-01 -4.52256382e-01
-1.97331876e-01 2.89544225e-01 2.50109166e-01 -9.34532762e-01
3.02857757e-01 1.42879292e-01 -6.93410309e-03 -4.09855157e-01
7.52393603e-01 -9.17789936e-01 1.57713249e-01 2.19198272e-01
-3.65245610e-01 6.55641109e-02 1.40134618e-01 3.91899049e-01
-7.01529801e-01 -7.82792550e-03 8.70270193e-01 1.62151650e-01
-2.92533249e-01 9.56029296e-02 -6.43409848e-01 -1.90174952e-01
7.38372803e-01 -5.75507917e-02 -1.69444486e-01 -9.79405999e-01
-6.73559248e-01 -2.71459579e-01 3.53699401e-02 4.45182472e-01
1.02089179e+00 -1.07958961e+00 -8.25893104e-01 3.02511513e-01
-3.30738068e-01 8.47706124e-02 6.30499661e-01 5.47527730e-01
-4.92129743e-01 2.66058445e-01 3.51063646e-02 -2.76825577e-01
-1.52148426e+00 2.53318250e-01 2.59949088e-01 -2.08288193e-01
-3.53214353e-01 7.47720778e-01 1.60940096e-01 -4.40514445e-01
5.75185359e-01 -1.57109603e-01 -1.29735380e-01 -4.82996225e-01
5.83276868e-01 4.64602262e-01 4.58350122e-01 -7.51959622e-01
-9.07745212e-02 1.56066015e-01 1.09137118e-01 -5.57328939e-01
1.23477483e+00 -2.68813074e-01 1.15901604e-01 -7.02513335e-03
9.21447158e-01 2.79961109e-01 -1.00351989e+00 -9.61506218e-02
-4.12417203e-01 -4.00738776e-01 3.46881539e-01 -7.74734318e-01
-1.08126664e+00 1.34011543e+00 7.60410130e-01 1.08074836e-01
1.43538547e+00 -2.46941686e-01 1.10524344e+00 2.39193797e-01
2.34211817e-01 -1.04758477e+00 1.16995662e-01 4.30348724e-01
9.29853857e-01 -1.05130494e+00 -5.25400937e-01 -2.77617276e-01
-6.65332139e-01 7.37862527e-01 3.09656888e-01 2.81295836e-01
5.61501205e-01 5.56748033e-01 3.02549988e-01 4.41842735e-01
-3.94128233e-01 -3.32529157e-01 1.13737218e-01 9.62918878e-01
3.88042867e-01 -6.52926937e-02 -7.95129240e-02 5.71876943e-01
-3.25883180e-01 -2.86258996e-01 4.38377380e-01 7.02583015e-01
-2.84486592e-01 -1.26984215e+00 -7.46940136e-01 1.19635411e-01
-7.56261885e-01 -2.64239520e-01 -1.04455262e-01 2.77348399e-01
-1.01805598e-01 1.34430671e+00 -9.80858654e-02 -5.30389845e-01
3.05832505e-01 -1.53066516e-01 3.10379505e-01 -4.94717628e-01
-6.30922675e-01 3.54081511e-01 6.20958433e-02 4.80185412e-02
-3.81653398e-01 -3.96245658e-01 -9.30511951e-01 -3.00354093e-01
-5.53981364e-01 1.50297239e-01 1.20668924e+00 7.70845234e-01
3.79364818e-01 8.34946275e-01 1.00736868e+00 -6.90209031e-01
-4.81218934e-01 -1.11404514e+00 -6.05242670e-01 3.26500058e-01
3.94418657e-01 -3.00437868e-01 -3.65238786e-01 1.21743180e-01] | [14.963175773620605, 6.086564540863037] |
6db26fc9-225f-4d1c-8f13-c3f320beb519 | dynamic-conditional-networks-for-few-shot | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Fang_Zhao_Dynamic_Conditional_Networks_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Fang_Zhao_Dynamic_Conditional_Networks_ECCV_2018_paper.pdf | Dynamic Conditional Networks for Few-Shot Learning | This paper proposes a novel Dynamic Conditional Convolutional Network (DCCN) to handle conditional few-shot learning, i.e, only a few training samples are available for each condition. DCCN consists of dual subnets: DyConvNet contains a dynamic convolutional layer with a bank of basis filters; CondiNet predicts a set of adaptive weights from conditional inputs to linearly combine the basis filters. In this manner, a specific convolutional kernel can be dynamically obtained for each conditional input. The filter bank is shared between all conditions thus only a low-dimension weight vector needs to be learned. This significantly facilitates the parameter learning across different conditions when training data are limited. We evaluate DCCN on four tasks which can be formulated as conditional model learning, including specific object counting, multi-modal image classification, phrase grounding and identity based face generation. Extensive experiments demonstrate the superiority of the proposed model in the conditional few-shot learning setting. | ['Jian Zhao', 'Fang Zhao', 'Jiashi Feng', 'Shuicheng Yan'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['object-counting', 'phrase-grounding'] | ['computer-vision', 'natural-language-processing'] | [ 3.14952165e-01 -6.39546812e-02 -4.85195100e-01 -5.24095476e-01
-4.53101248e-01 -8.24027359e-02 6.83368266e-01 -1.79464698e-01
-6.19656563e-01 5.94738722e-01 -1.27367646e-01 1.45402968e-01
1.62535682e-01 -8.90325546e-01 -7.19896019e-01 -8.14633131e-01
3.36915731e-01 4.98280406e-01 2.82575816e-01 2.38812655e-01
-6.47405609e-02 2.69940108e-01 -1.76621294e+00 3.24125528e-01
5.15719712e-01 1.14686644e+00 5.65301716e-01 7.77482688e-01
-1.57040685e-01 7.22544074e-01 -3.62477481e-01 -3.55593175e-01
-3.89989242e-02 -4.00256425e-01 -4.73979473e-01 2.57324100e-01
2.64562666e-01 -3.58474165e-01 1.72915701e-02 9.35932577e-01
7.45620787e-01 5.14811218e-01 8.05261195e-01 -1.21928370e+00
-5.96829772e-01 5.79600513e-01 -5.50367534e-01 1.70219019e-01
1.29384950e-01 4.68843579e-01 1.12933540e+00 -1.05489278e+00
6.03903353e-01 1.35473442e+00 3.56082946e-01 1.18331552e+00
-1.41805232e+00 -9.53649819e-01 3.99812490e-01 3.37437928e-01
-1.23187971e+00 -5.77218592e-01 5.57426274e-01 -4.16939348e-01
8.66143107e-01 -2.26455152e-01 6.05812728e-01 1.25890088e+00
-1.37292936e-01 7.52582073e-01 8.04762542e-01 -5.74613452e-01
4.94787991e-01 1.60568967e-01 1.05177276e-01 8.68715763e-01
-1.75369918e-01 -8.03091656e-03 -8.59613895e-01 -4.98321690e-02
6.90859318e-01 7.77093023e-02 -1.03133850e-01 -1.73058376e-01
-9.14061010e-01 8.74066830e-01 1.28628820e-01 1.86675459e-01
-2.53610492e-01 4.04699057e-01 3.97647232e-01 1.85773149e-01
2.03642368e-01 -1.96542174e-01 -6.09877884e-01 1.96955234e-01
-7.48810887e-01 1.97072238e-01 6.86764181e-01 1.15725040e+00
1.11730218e+00 1.53813198e-01 -8.26794565e-01 1.09955263e+00
8.29876661e-02 5.32473981e-01 5.19725025e-01 -6.43076599e-01
4.11115557e-01 2.73205131e-01 -1.51494607e-01 -5.97633660e-01
-1.66624606e-01 -1.10630095e-01 -8.29141438e-01 -3.34307365e-02
9.47609171e-02 -5.40669084e-01 -1.17598116e+00 2.05061817e+00
3.45573992e-01 9.05322969e-01 -1.37989819e-01 5.90032876e-01
8.92608941e-01 6.54405475e-01 4.39685583e-01 -3.10105175e-01
1.31098747e+00 -8.43550086e-01 -7.50999928e-01 -1.48725271e-01
3.37005764e-01 -5.97026587e-01 9.18638170e-01 5.34957945e-02
-8.86530042e-01 -6.93526089e-01 -1.00596380e+00 2.23853141e-01
-4.46981937e-01 3.10670733e-01 7.57874429e-01 6.10381782e-01
-7.87446439e-01 3.98038894e-01 -7.28843808e-01 -9.65011343e-02
6.06089294e-01 3.86915118e-01 -1.63416699e-01 -3.61067802e-01
-1.19896078e+00 5.18161118e-01 7.13830054e-01 5.83059639e-02
-1.26493430e+00 -7.97579825e-01 -1.03082001e+00 4.41518515e-01
4.18608636e-01 -4.82898116e-01 1.15503430e+00 -8.99673522e-01
-1.47940719e+00 7.65484989e-01 -3.24230045e-01 -2.93852121e-01
1.72319785e-01 7.96466321e-02 -4.11430478e-01 1.56158254e-01
1.23343952e-01 8.03016186e-01 1.38943696e+00 -1.04768264e+00
-7.82783866e-01 -1.78247187e-02 3.76465246e-02 -5.71102314e-02
-5.36655128e-01 1.47583306e-01 -9.47685838e-01 -5.95246971e-01
-2.85550773e-01 -8.66821706e-01 -6.58603683e-02 1.74295902e-03
-5.13988137e-01 -4.49324578e-01 6.63294137e-01 -1.11801736e-01
1.17459667e+00 -2.01034665e+00 9.39708725e-02 1.63928062e-01
-1.01882041e-01 2.97404587e-01 -2.96017528e-01 -6.10190295e-02
-1.33017510e-01 -1.56603500e-01 -2.63717473e-01 -4.54189450e-01
-8.41162354e-02 2.39353076e-01 9.61794611e-03 2.11964041e-01
5.34337282e-01 8.60704064e-01 -7.43391156e-01 -7.34208226e-01
3.43125135e-01 5.08194685e-01 -8.08751166e-01 4.07498360e-01
-6.02678657e-01 1.67988494e-01 -3.52589488e-01 7.21552730e-01
8.80346358e-01 -2.83549070e-01 2.15815455e-01 -4.64849174e-02
1.55207798e-01 -4.21090364e-01 -1.46704018e+00 1.72847354e+00
-7.93873072e-01 2.74700701e-01 -1.46808863e-01 -1.01649320e+00
6.08050644e-01 4.46778476e-01 2.46537149e-01 -4.24648315e-01
3.62002134e-01 -3.00178796e-01 -1.70259550e-01 -5.00489533e-01
3.03840283e-02 -3.45592797e-01 -1.35312468e-01 4.91603374e-01
9.52615499e-01 3.56446892e-01 3.63833994e-01 1.32140487e-01
8.20754588e-01 3.97492312e-02 2.46990919e-01 -8.24620575e-02
8.18236470e-01 -6.19721174e-01 1.17256868e+00 6.26303613e-01
-9.51171517e-02 5.34278512e-01 5.47267914e-01 -2.52421468e-01
-6.41159654e-01 -9.95387614e-01 -1.35475218e-01 1.56456208e+00
-1.05005624e-02 -3.53330970e-01 -8.18382204e-01 -6.09845221e-01
-9.32106823e-02 7.51775503e-01 -1.05262125e+00 -3.17346543e-01
-4.25006002e-01 -7.27409005e-01 2.28696793e-01 6.34006262e-01
4.33319211e-01 -1.39798295e+00 -4.68715906e-01 4.97638471e-02
1.04920588e-01 -1.24577391e+00 -6.67489409e-01 3.42259705e-01
-5.83724022e-01 -1.24188113e+00 -6.12265944e-01 -9.83717084e-01
7.75295317e-01 1.13279529e-01 9.25979197e-01 1.38524324e-01
-6.10526025e-01 4.21215445e-01 -2.73206204e-01 -4.29487258e-01
-1.17173135e-01 -1.53189395e-02 5.37519306e-02 6.39994919e-01
5.45779645e-01 -5.61302900e-01 -3.87633413e-01 4.52908538e-02
-8.70232582e-01 1.81725062e-02 5.89444101e-01 1.28356147e+00
6.65329635e-01 -1.21790908e-01 5.75729191e-01 -1.25602531e+00
3.84920508e-01 -5.88951647e-01 -6.36883795e-01 5.48047185e-01
-1.54674247e-01 1.93695992e-01 6.46980941e-01 -7.71234035e-01
-1.64934635e+00 3.22795689e-01 2.33941339e-02 -6.85381413e-01
-2.16390431e-01 2.42679983e-01 -1.98265642e-01 1.91815197e-01
4.80565280e-01 2.05019429e-01 -2.36112833e-01 -5.01017511e-01
5.86989582e-01 4.18287843e-01 4.15529370e-01 -7.61802197e-01
8.21945131e-01 3.95526171e-01 -2.35724956e-01 -8.63758206e-01
-9.37137723e-01 -4.39685345e-01 -8.11329246e-01 -3.54044378e-01
1.17611516e+00 -1.07122099e+00 -8.68898869e-01 5.37493289e-01
-1.14778233e+00 -3.65433723e-01 -2.63072461e-01 3.86866480e-01
-5.81493139e-01 -2.25964189e-01 -6.07614875e-01 -7.94653118e-01
-3.58355284e-01 -9.03147638e-01 7.78494298e-01 5.90102434e-01
9.07545686e-02 -9.35912788e-01 -6.17648140e-02 -7.50789419e-03
1.46422297e-01 9.80181247e-02 1.13195205e+00 -4.84845012e-01
-3.92401963e-01 -1.99615970e-01 -1.39697745e-01 4.04923350e-01
1.68416966e-02 -2.70074327e-03 -1.28020275e+00 -3.59376192e-01
-2.72067547e-01 -9.27581191e-01 1.10694814e+00 3.64960253e-01
1.51845241e+00 -4.68355529e-02 -2.83338755e-01 7.43400097e-01
1.52459419e+00 -3.21289413e-02 3.45284283e-01 -3.96308035e-01
6.36034966e-01 3.71008962e-01 5.55197299e-01 6.45869374e-01
5.57293463e-03 4.54093307e-01 2.56721735e-01 1.98498830e-01
-2.80947775e-01 -2.92976916e-01 8.92018825e-02 6.96444094e-01
-2.11448565e-01 8.77491310e-02 -7.18228400e-01 6.10539377e-01
-1.71859610e+00 -1.11315107e+00 5.45929134e-01 2.16465926e+00
8.51134956e-01 1.18320689e-01 -1.95747808e-01 -1.14343859e-01
1.03586674e+00 2.34322608e-01 -7.56960869e-01 -7.55388066e-02
4.10449654e-02 7.72770822e-01 1.31033137e-01 3.09945464e-01
-1.05204844e+00 1.31368315e+00 5.72991371e+00 1.04054594e+00
-1.06923604e+00 2.64558733e-01 5.67964017e-01 -2.50082940e-01
-8.19366276e-02 -5.66328876e-02 -1.14577019e+00 4.14066583e-01
7.33400583e-01 -7.46506974e-02 3.02859426e-01 1.01741314e+00
-1.68285221e-01 -1.49292484e-01 -1.03043032e+00 1.13804221e+00
1.79575786e-01 -1.42597103e+00 9.35380682e-02 -5.11488914e-01
8.85874510e-01 -1.29806429e-01 2.01318607e-01 6.57374024e-01
3.99194717e-01 -9.02766287e-01 4.30506140e-01 6.14325643e-01
1.24890745e+00 -1.05621207e+00 2.94512302e-01 4.93246794e-01
-1.35653877e+00 -3.72518659e-01 -6.77107692e-01 2.94564039e-01
-5.27418405e-02 2.43874043e-01 -6.58993661e-01 1.29372090e-01
6.47063255e-01 5.60612261e-01 -3.68726283e-01 8.43939781e-01
-3.46352041e-01 6.93311334e-01 9.40421224e-02 -4.31470722e-02
1.78832393e-02 1.62823722e-01 2.07004830e-01 1.32093644e+00
1.26607656e-01 3.70205730e-01 2.63141900e-01 8.28262091e-01
-3.86064321e-01 1.44077629e-01 -1.67573854e-01 1.08518332e-01
5.69934964e-01 1.50903261e+00 -7.11710870e-01 -3.72463822e-01
-6.67809725e-01 1.04611981e+00 6.62392497e-01 5.03156722e-01
-8.17581236e-01 -3.14098209e-01 6.95185184e-01 -2.48017177e-01
8.23232412e-01 1.58502638e-01 5.86260669e-02 -1.46223617e+00
-3.06361735e-01 -4.59381193e-01 6.27733767e-01 -4.51231509e-01
-1.26173508e+00 3.77269715e-01 -1.57765374e-02 -1.14492571e+00
-3.04844707e-01 -6.25822484e-01 -1.05668485e+00 8.81089628e-01
-1.61674035e+00 -1.33735061e+00 -2.62152404e-01 1.02977538e+00
6.72406137e-01 -5.30457020e-01 1.20648718e+00 1.37187555e-01
-9.39373553e-01 8.56978953e-01 -2.40889713e-01 3.29161584e-01
7.71263838e-01 -1.07962751e+00 4.02618721e-02 7.31182098e-01
3.18456553e-02 6.76304519e-01 4.09693420e-01 -5.00743628e-01
-1.17490411e+00 -1.19175088e+00 7.89545298e-01 1.25191987e-01
3.37417334e-01 -5.92564285e-01 -7.93276966e-01 5.61458051e-01
1.02551065e-01 4.58906949e-01 9.77504194e-01 3.49141657e-01
-5.74737608e-01 -2.18303367e-01 -1.06012130e+00 4.21103030e-01
7.54690409e-01 -5.88404953e-01 -2.67218888e-01 3.95468444e-01
8.13134313e-01 -2.39121795e-01 -5.68338811e-01 1.03488237e-01
4.34284240e-01 -6.23023272e-01 7.57628083e-01 -9.51216817e-01
4.29100901e-01 -5.61349653e-02 -6.22476377e-02 -1.15915227e+00
-4.49991733e-01 -4.57992405e-01 -4.75368649e-01 1.16447794e+00
5.14339030e-01 -2.18864605e-01 9.05295789e-01 7.38405287e-01
2.14386433e-01 -8.20409060e-01 -8.38525474e-01 -5.47994077e-01
-1.63558125e-01 -6.24980330e-01 4.89772081e-01 9.42146003e-01
-2.87630320e-01 6.41208112e-01 -6.55802846e-01 -1.56886224e-02
7.44801402e-01 7.33935088e-02 6.02072299e-01 -1.19524479e+00
-6.51265502e-01 2.00281525e-03 -3.03177685e-01 -8.42490137e-01
5.77135563e-01 -9.04562056e-01 1.47580966e-01 -1.02672327e+00
5.32293737e-01 -4.60991204e-01 -5.84231317e-01 7.36533940e-01
-5.09470582e-01 -6.87527806e-02 3.22918296e-01 -2.67772973e-01
-6.71190202e-01 6.20538175e-01 1.06667757e+00 -2.59769261e-01
-1.18019596e-01 2.64175922e-01 -3.71116430e-01 6.46916211e-01
7.20277011e-01 -4.55538839e-01 -6.57648683e-01 -2.71399915e-01
-9.63233113e-02 9.99332368e-02 1.94071025e-01 -1.08155942e+00
4.85816747e-01 -3.59887093e-01 8.53439927e-01 -5.12779832e-01
5.10164976e-01 -5.62297404e-01 -2.56717324e-01 3.05694729e-01
-4.85763788e-01 -5.12510657e-01 1.50060862e-01 9.03940320e-01
-2.63676904e-02 -4.23587322e-01 1.15792096e+00 -2.32915521e-01
-1.07389402e+00 9.16245639e-01 -1.71032280e-01 1.44742563e-01
1.15955150e+00 1.14659190e-01 1.17103331e-01 -1.75056309e-01
-1.07007420e+00 3.92008007e-01 -5.18828295e-02 4.61586654e-01
8.82468998e-01 -1.63570690e+00 -5.56560755e-01 2.64463216e-01
3.13998312e-01 1.22037316e-02 7.73250818e-01 4.46396023e-01
3.91214602e-02 2.83873647e-01 -3.30232531e-01 -3.87923568e-01
-1.21789324e+00 7.46275723e-01 1.79046631e-01 -1.44706637e-01
-5.20806134e-01 1.30654013e+00 2.31375933e-01 -4.38600332e-01
4.79572386e-01 1.07914306e-01 -5.09080112e-01 4.30619270e-01
9.12517548e-01 1.11725628e-01 -2.78377354e-01 -6.31738424e-01
-4.13269624e-02 3.57486099e-01 -3.59167248e-01 -2.05868855e-01
1.39279675e+00 1.87438145e-01 -1.13158685e-03 6.23204172e-01
1.32895398e+00 -5.70950925e-01 -1.49409461e+00 -6.41354024e-01
-2.27103949e-01 -3.77749383e-01 4.40014713e-02 -5.33706486e-01
-1.21399486e+00 9.99442041e-01 5.50238132e-01 -4.68001246e-01
1.26045597e+00 -1.47875950e-01 6.08856201e-01 5.09408295e-01
7.53617212e-02 -1.64221156e+00 3.76437187e-01 5.31983554e-01
5.09834349e-01 -1.36053789e+00 -2.65357196e-01 -3.54187220e-01
-7.07533062e-01 1.15442204e+00 9.67276275e-01 -8.57838541e-02
1.02865744e+00 1.59748197e-01 -3.69919628e-01 -9.40760523e-02
-1.14968145e+00 -5.10493100e-01 1.25878647e-01 6.62411451e-01
3.17439616e-01 1.46606505e-01 -5.24548106e-02 8.00023854e-01
2.75681317e-01 5.54776825e-02 2.13041738e-01 8.63403022e-01
-4.67599392e-01 -9.85272348e-01 -4.39595096e-02 6.16469741e-01
-2.62539536e-01 -1.87408209e-01 7.17983171e-02 4.85713363e-01
5.22938788e-01 7.27144778e-01 1.14978097e-01 -3.39191973e-01
9.74388942e-02 4.37855363e-01 4.64013577e-01 -1.08739924e+00
-3.64719361e-01 1.12021655e-01 -1.89857602e-01 -5.71542501e-01
-4.49970931e-01 -6.57099485e-01 -1.25527966e+00 -1.03835143e-01
-4.24228281e-01 -1.51179835e-01 3.50724399e-01 9.40909982e-01
-4.05166298e-02 7.58216262e-01 6.16935015e-01 -6.46556854e-01
-3.94226044e-01 -1.05560863e+00 -6.97158992e-01 3.63153458e-01
2.19057098e-01 -7.80534387e-01 -8.28012750e-02 3.59602690e-01] | [9.951421737670898, 2.904409170150757] |
b7577098-da51-4094-8b5f-eb2b78acf773 | uncertainty-in-extreme-multi-label | 2210.10160 | null | https://arxiv.org/abs/2210.10160v1 | https://arxiv.org/pdf/2210.10160v1.pdf | Uncertainty in Extreme Multi-label Classification | Uncertainty quantification is one of the most crucial tasks to obtain trustworthy and reliable machine learning models for decision making. However, most research in this domain has only focused on problems with small label spaces and ignored eXtreme Multi-label Classification (XMC), which is an essential task in the era of big data for web-scale machine learning applications. Moreover, enormous label spaces could also lead to noisy retrieval results and intractable computational challenges for uncertainty quantification. In this paper, we aim to investigate general uncertainty quantification approaches for tree-based XMC models with a probabilistic ensemble-based framework. In particular, we analyze label-level and instance-level uncertainty in XMC, and propose a general approximation framework based on beam search to efficiently estimate the uncertainty with a theoretical guarantee under long-tail XMC predictions. Empirical studies on six large-scale real-world datasets show that our framework not only outperforms single models in predictive performance, but also can serve as strong uncertainty-based baselines for label misclassification and out-of-distribution detection, with significant speedup. Besides, our framework can further yield better state-of-the-art results based on deep XMC models with uncertainty quantification. | ['Hsiang-Fu Yu', 'Cho-Jui Hsieh', 'Jiong Zhong', 'Wei-Cheng Chang', 'Jyun-Yu Jiang'] | 2022-10-18 | null | null | null | null | ['extreme-multi-label-classification'] | ['methodology'] | [-8.35000202e-02 -1.47652730e-01 -3.21951620e-02 -6.84824824e-01
-1.57211292e+00 -5.55450618e-01 3.16158682e-01 3.62287372e-01
-1.30326048e-01 1.04526818e+00 -3.67067605e-01 -4.94883388e-01
-3.90441060e-01 -7.02734888e-01 -6.99725389e-01 -9.46187496e-01
1.20959960e-01 9.59449172e-01 8.79729614e-02 4.11456496e-01
2.14743197e-01 2.91017257e-02 -1.44953632e+00 1.66016579e-01
1.16477931e+00 1.58406019e+00 -3.21925282e-01 2.18458459e-01
-3.69952470e-01 6.65599287e-01 -5.63023686e-01 -9.66141582e-01
-3.50519791e-02 1.34292832e-02 -6.41834080e-01 -2.72484690e-01
2.60846883e-01 -2.06530124e-01 2.85975307e-01 1.43067551e+00
5.32699645e-01 -7.58808777e-02 1.12696970e+00 -1.49531794e+00
-2.91702986e-01 1.05861032e+00 -9.60044086e-01 -3.05921733e-01
-1.55101478e-01 -2.28048086e-01 1.18315327e+00 -6.89779460e-01
2.18528524e-01 1.38354826e+00 7.14458704e-01 5.54415345e-01
-1.04108000e+00 -8.30362260e-01 1.25969678e-01 4.19335902e-01
-1.45310724e+00 6.76422119e-02 4.13169771e-01 -2.34468713e-01
4.09195721e-01 2.37697735e-01 -1.88352734e-01 9.95359302e-01
2.84248918e-01 8.96510422e-01 1.27836955e+00 -3.92670065e-01
7.31622040e-01 1.63123325e-01 3.51172596e-01 4.34232652e-01
2.98305184e-01 -8.37456901e-03 -3.39267015e-01 -5.04252911e-01
-3.13139660e-03 -3.82743403e-02 -2.42224544e-01 -1.08853713e-01
-9.79074299e-01 1.00348270e+00 3.03541392e-01 -1.71885788e-01
-1.00481905e-01 5.75293660e-01 4.34135258e-01 4.79079001e-02
1.00785422e+00 9.52536613e-02 -6.55853808e-01 -2.18477294e-01
-8.95879924e-01 3.30870241e-01 7.09065437e-01 9.05341625e-01
3.22969645e-01 -4.23352659e-01 -4.85692441e-01 8.37368011e-01
6.50680184e-01 7.99596667e-01 4.00348417e-02 -1.12115908e+00
6.55339479e-01 3.05286348e-01 2.50583768e-01 -8.02680910e-01
-4.84441340e-01 -7.56871283e-01 -8.90517592e-01 1.32356942e-01
5.13352990e-01 1.04563106e-02 -6.76386714e-01 1.68110001e+00
6.18195236e-01 2.77415663e-01 -2.21949980e-01 8.79314601e-01
5.11596978e-01 5.61808288e-01 1.85942352e-01 -3.90946031e-01
1.48236907e+00 -7.63776064e-01 -6.84408844e-01 9.13376659e-02
8.40634167e-01 -5.00370622e-01 8.94072473e-01 5.47733724e-01
-6.23494387e-01 7.60067999e-02 -8.40863705e-01 1.79563731e-01
-3.59609872e-01 -2.10672468e-01 6.75606966e-01 8.48528087e-01
-5.26538908e-01 7.10354924e-01 -5.71402192e-01 2.43007734e-01
5.50012946e-01 -1.15883291e-01 1.65519286e-02 -5.05627036e-01
-1.53453410e+00 7.67001808e-01 6.80396914e-01 1.72025636e-01
-6.24186099e-01 -6.02780640e-01 -6.20445490e-01 1.98213920e-01
8.32700610e-01 -4.61974472e-01 1.47980988e+00 -5.28366603e-02
-1.05554080e+00 4.72409725e-01 1.16779178e-01 -5.44404507e-01
8.03393126e-01 -1.68238923e-01 -2.62123764e-01 -1.56962469e-01
1.10258952e-01 5.43559968e-01 7.73074746e-01 -1.38923442e+00
-8.17177773e-01 -7.22064435e-01 -2.67560363e-01 3.27558778e-02
-3.04160770e-02 6.16441369e-02 -3.58334392e-01 -4.47754383e-01
3.77631068e-01 -8.85988474e-01 -1.26161754e-01 -5.10961935e-02
-5.41498184e-01 -6.09060168e-01 4.35096145e-01 -4.45906252e-01
1.49154246e+00 -1.78384995e+00 -1.49700940e-01 2.41363838e-01
2.67687261e-01 -8.17296281e-03 2.05725372e-01 4.83367890e-02
3.49568427e-01 3.50083917e-01 -5.23730218e-01 -6.27334893e-01
5.35127699e-01 1.43251896e-01 -5.25253475e-01 3.31037968e-01
-1.21164825e-02 1.02988780e+00 -1.00648844e+00 -8.56376112e-01
2.33195219e-02 2.35898674e-01 -1.72841892e-01 -1.94776624e-01
-7.79215932e-01 1.81360558e-01 -3.49052489e-01 1.06600213e+00
9.75748360e-01 -5.91054499e-01 -2.53728386e-02 3.52610648e-02
4.55043405e-01 -2.20940500e-01 -1.25173700e+00 1.35419452e+00
-5.88801146e-01 6.27071783e-02 -8.59211907e-02 -7.54075289e-01
4.70252246e-01 1.26474917e-01 3.45744252e-01 -2.40331471e-01
3.89658779e-01 5.49272299e-01 -4.69205797e-01 1.21330932e-01
2.81590939e-01 -2.22666621e-01 -4.70803082e-01 7.05594242e-01
-1.75204605e-01 -2.35267252e-01 3.72161493e-02 3.24403495e-01
8.08765471e-01 -6.38433322e-02 7.03568310e-02 -1.61811009e-01
2.72979856e-01 -3.63749713e-01 6.41168058e-01 9.08133984e-01
-4.09304500e-01 5.54462790e-01 6.90348387e-01 -1.13985717e-01
-6.36533558e-01 -8.92851889e-01 -8.12795103e-01 9.95066643e-01
1.60824463e-01 -4.20785993e-01 -7.62394011e-01 -1.03022242e+00
4.72966731e-01 1.16114175e+00 -4.81175542e-01 -8.56252462e-02
3.16535980e-01 -1.22406256e+00 4.76126075e-01 4.41918135e-01
4.97545838e-01 -5.80971837e-01 -1.10374436e-01 2.09848896e-01
-5.72540700e-01 -1.05336475e+00 -2.84397393e-01 3.05388570e-01
-7.23582506e-01 -1.05989659e+00 -5.90526640e-01 1.57557055e-01
2.50480562e-01 -1.42257303e-01 1.39593840e+00 -3.05068582e-01
-1.28326744e-01 1.12777330e-01 -3.10258240e-01 -6.86068356e-01
-2.76749879e-01 -2.18817830e-01 1.61192462e-01 -1.32273540e-01
5.57578325e-01 -2.04916507e-01 -4.11667079e-01 5.82438171e-01
-9.49106693e-01 -2.69076049e-01 3.00585300e-01 1.01305187e+00
8.47646952e-01 6.47664726e-01 7.99576879e-01 -1.09455192e+00
7.13476121e-01 -6.22817397e-01 -8.84410262e-01 7.40639389e-01
-1.30792177e+00 3.15370262e-01 2.18471035e-01 -1.34216204e-01
-1.14623094e+00 -2.86864072e-01 -5.04364669e-02 -3.15834701e-01
1.86891362e-01 5.65891385e-01 -2.95398653e-01 1.53401271e-01
4.83944923e-01 -1.19882666e-01 -3.38595480e-01 -5.40647030e-01
5.36168456e-01 1.17113161e+00 3.69838804e-01 -1.02087426e+00
2.52959341e-01 3.20262879e-01 2.31169090e-01 1.19838387e-01
-1.56347728e+00 -4.67442036e-01 -3.35874557e-01 -1.79237455e-01
4.71842021e-01 -9.28295314e-01 -9.11152959e-01 6.67535543e-01
-1.11020362e+00 1.15282787e-02 5.83738275e-02 2.29920581e-01
-3.93843234e-01 4.72009689e-01 -6.84696138e-01 -1.36675799e+00
-5.84201992e-01 -1.09792829e+00 1.63813567e+00 7.86786228e-02
2.57974207e-01 -8.40557992e-01 -7.25682825e-02 5.91325819e-01
3.69217664e-01 1.36666656e-01 1.02367604e+00 -5.79651773e-01
-6.55608475e-01 -3.86917442e-01 -6.39074326e-01 4.27069843e-01
-5.75500906e-01 -5.19593358e-02 -1.06464088e+00 -8.84768218e-02
-1.15213372e-01 -7.59835541e-01 1.23746419e+00 3.29171151e-01
1.63993275e+00 -3.45368520e-03 -4.36763287e-01 3.23031694e-01
1.47219431e+00 -3.17724437e-01 4.57946926e-01 -2.17939541e-02
3.39755803e-01 4.68569905e-01 9.03112292e-01 7.51373708e-01
4.49671239e-01 6.14423871e-01 6.71017110e-01 7.27815151e-01
2.74312943e-01 -2.10844293e-01 -7.46671110e-03 6.22393847e-01
2.50300080e-01 -4.80835736e-01 -1.05682671e+00 1.22656055e-01
-2.11994672e+00 -6.50661349e-01 -2.22921774e-01 2.29774547e+00
1.14103591e+00 2.21824840e-01 -3.30226302e-01 1.50959864e-01
9.07189667e-01 -1.69235229e-01 -8.10487568e-01 5.27950935e-02
-2.24130210e-02 -1.96593240e-01 4.83176112e-01 2.90378392e-01
-1.29688215e+00 6.12629771e-01 5.76359367e+00 1.58537340e+00
-5.32438695e-01 3.32653105e-01 1.04369104e+00 -2.44389459e-01
-5.36454320e-01 -2.54737258e-01 -1.14601886e+00 1.06356525e+00
1.08249772e+00 9.86041799e-02 2.07499772e-01 1.13026559e+00
-4.28936422e-01 -3.63653600e-01 -1.07483995e+00 1.21044540e+00
-1.62969977e-01 -8.70121658e-01 -4.49907005e-01 1.53187230e-01
1.06847978e+00 3.85148712e-02 1.15373142e-01 7.80729890e-01
5.89932799e-01 -9.42481756e-01 5.31641006e-01 6.04135871e-01
9.90246236e-01 -9.72133875e-01 1.30163324e+00 6.51628554e-01
-6.82289541e-01 -1.64276257e-01 -7.31227696e-01 2.90636957e-01
2.43971601e-01 1.63023436e+00 -3.61310720e-01 6.01589084e-01
6.78612590e-01 1.16138555e-01 -5.24630129e-01 9.99715626e-01
-2.76787847e-01 7.34424412e-01 -7.96608150e-01 -3.39556515e-01
7.77782425e-02 -2.35163588e-02 1.56714857e-01 8.22442710e-01
4.64400649e-01 1.26986802e-02 -6.60496652e-02 9.57036257e-01
-4.60377336e-01 5.91669045e-02 -1.75069675e-01 6.87145516e-02
8.20964098e-01 1.29092371e+00 -6.81949139e-01 -2.54226804e-01
-1.37352601e-01 7.96126485e-01 5.49875200e-01 7.99799189e-02
-1.22945130e+00 -2.29039401e-01 2.69416690e-01 -4.44204956e-01
-4.94172201e-02 2.59749562e-01 -5.05152822e-01 -1.32473743e+00
3.15627381e-02 -6.17462814e-01 6.64191842e-01 -6.03695095e-01
-1.86949837e+00 3.35257620e-01 -4.10118476e-02 -9.76401329e-01
-3.67784649e-01 -7.74592876e-01 -2.92101055e-01 8.59847128e-01
-1.51266670e+00 -9.56510901e-01 -1.65309057e-01 1.92650557e-01
1.59828663e-01 1.38625637e-01 9.13064003e-01 1.41711950e-01
-6.21782124e-01 7.98037231e-01 1.01277816e+00 -3.39314491e-01
9.51109529e-01 -1.56872416e+00 -2.71270536e-02 4.60538983e-01
2.71516860e-01 1.98919401e-01 6.09013796e-01 -6.62360132e-01
-7.49330223e-01 -1.20554304e+00 7.69263208e-01 -9.25098777e-01
7.05854237e-01 -3.09801012e-01 -7.68142998e-01 1.31800517e-01
-4.46488589e-01 3.21934730e-01 7.91093826e-01 5.69246471e-01
-6.81683898e-01 -1.72413111e-01 -1.36398661e+00 2.05211878e-01
7.56429970e-01 -4.80706960e-01 -1.67107493e-01 6.86979592e-01
9.27414119e-01 -5.07255673e-01 -9.84080553e-01 7.55476415e-01
6.29746675e-01 -9.02502656e-01 6.74038112e-01 -6.04207754e-01
6.15109801e-01 -1.26109332e-01 -5.08030236e-01 -1.35477698e+00
-2.20029242e-02 3.22493142e-03 -7.37940967e-01 1.44524443e+00
5.58555961e-01 -5.99869847e-01 6.67146266e-01 1.03772545e+00
2.66920686e-01 -9.66707170e-01 -1.04424310e+00 -9.16435361e-01
4.18380797e-01 -1.03468084e+00 8.66366327e-01 4.97935414e-01
-8.32840353e-02 1.31904602e-01 -3.72760892e-01 1.87930956e-01
1.26209688e+00 3.48284215e-01 1.50821432e-01 -1.63594389e+00
-2.84665674e-01 -3.57553929e-01 -2.70955503e-01 -6.00721657e-01
4.43593085e-01 -7.12475777e-01 2.97983706e-01 -1.36754429e+00
5.52474558e-01 -8.21121275e-01 -6.28805935e-01 2.75859058e-01
-4.97010887e-01 1.51624739e-01 5.03865853e-02 3.06830883e-01
-1.28710604e+00 6.48355544e-01 8.35603237e-01 -3.18813026e-01
5.70702314e-01 4.55886662e-01 -7.58566141e-01 5.19908369e-01
6.66375518e-01 -6.41284943e-01 -4.13270205e-01 -1.17133006e-01
6.63038075e-01 1.31585449e-01 2.08516389e-01 -8.63204420e-01
1.37362689e-01 -1.19442731e-01 1.93784967e-01 -9.37003434e-01
-3.10844276e-02 -1.01145494e+00 -3.27320993e-02 4.80035841e-02
-4.96535987e-01 -5.63708305e-01 -4.49282564e-02 1.17167687e+00
-1.53969929e-01 -7.11817682e-01 6.27392530e-01 -8.86085555e-02
-3.82314324e-01 3.63904506e-01 2.53394037e-01 1.97163329e-01
1.15827286e+00 7.80309141e-01 -5.44275761e-01 -3.90957415e-01
-5.25681913e-01 8.59005511e-01 1.67188182e-01 6.09146729e-02
2.52268761e-01 -1.58697474e+00 -8.99321973e-01 -3.44359875e-01
5.03544748e-01 3.69272918e-01 4.61686581e-01 5.49839139e-01
-9.97466221e-02 8.40590656e-01 5.76002479e-01 -8.19890618e-01
-8.83459091e-01 7.09211648e-01 1.95217311e-01 -7.98045099e-01
1.10333994e-01 1.14959049e+00 -1.65175259e-01 -7.52897620e-01
6.79890931e-01 -4.23549972e-02 2.97223419e-01 -1.45596545e-03
7.68849790e-01 5.49599290e-01 2.61895061e-01 -1.25780612e-01
-3.58809680e-01 2.38674253e-01 -3.91027220e-02 -9.84340236e-02
7.84391224e-01 -3.36511791e-01 -2.04785019e-01 5.62774062e-01
1.03582346e+00 -4.22441661e-01 -1.10488129e+00 -6.04652882e-01
2.91004479e-01 -4.57022786e-01 3.96579385e-01 -1.50209105e+00
-7.83626974e-01 1.29296052e+00 5.60973585e-01 1.75528228e-02
9.26996469e-01 2.07769230e-01 6.96873248e-01 4.80240822e-01
9.15564120e-01 -1.33350897e+00 -1.63182363e-01 2.61530310e-01
7.85779238e-01 -1.79573917e+00 -1.71248373e-02 -2.98257411e-01
-6.57721817e-01 8.48950148e-01 5.23373961e-01 8.21007609e-01
9.04956877e-01 1.83037162e-01 -1.51264042e-01 -9.21575949e-02
-9.95982587e-01 -2.82416642e-02 3.06405336e-01 3.31451982e-01
2.34489188e-01 4.15720910e-01 -2.76979387e-01 1.08278489e+00
2.74291962e-01 -1.32439821e-03 5.48025221e-02 3.50823969e-01
-5.09523690e-01 -1.04869735e+00 -4.22604769e-01 8.99693549e-01
-6.44457817e-01 -2.55139381e-01 4.15626206e-02 2.90584892e-01
8.05175975e-02 9.97234285e-01 -2.20979333e-01 -2.54431814e-01
-1.87946081e-01 4.84003931e-01 4.13953424e-01 -2.46415108e-01
1.96701601e-01 -9.97210592e-02 -1.57655682e-02 -4.61190760e-01
-5.37769347e-02 -6.00906312e-01 -1.19998515e+00 -5.21093905e-01
-9.37557995e-01 3.27198684e-01 9.43268597e-01 8.56634676e-01
3.50944072e-01 2.68517733e-01 5.27371526e-01 -3.30859751e-01
-1.66919541e+00 -1.25685143e+00 -1.10573995e+00 2.32566640e-01
-1.58493966e-01 -9.34620500e-01 -6.07874274e-01 -5.25659263e-01] | [8.897016525268555, 4.190561771392822] |
5dbae2a8-1f5d-45ae-b8eb-2d5c6d7fb1f9 | action-improving-semi-supervised-medical | 2304.02689 | null | https://arxiv.org/abs/2304.02689v2 | https://arxiv.org/pdf/2304.02689v2.pdf | ACTION++: Improving Semi-supervised Medical Image Segmentation with Adaptive Anatomical Contrast | Medical data often exhibits long-tail distributions with heavy class imbalance, which naturally leads to difficulty in classifying the minority classes (i.e., boundary regions or rare objects). Recent work has significantly improved semi-supervised medical image segmentation in long-tailed scenarios by equipping them with unsupervised contrastive criteria. However, it remains unclear how well they will perform in the labeled portion of data where class distribution is also highly imbalanced. In this work, we present ACTION++, an improved contrastive learning framework with adaptive anatomical contrast for semi-supervised medical segmentation. Specifically, we propose an adaptive supervised contrastive loss, where we first compute the optimal locations of class centers uniformly distributed on the embedding space (i.e., off-line), and then perform online contrastive matching training by encouraging different class features to adaptively match these distinct and uniformly distributed class centers. Moreover, we argue that blindly adopting a constant temperature $\tau$ in the contrastive loss on long-tailed medical data is not optimal, and propose to use a dynamic $\tau$ via a simple cosine schedule to yield better separation between majority and minority classes. Empirically, we evaluate ACTION++ on ACDC and LA benchmarks and show that it achieves state-of-the-art across two semi-supervised settings. Theoretically, we analyze the performance of adaptive anatomical contrast and confirm its superiority in label efficiency. | ['Jasjeet S. Sekhon', 'James S. Duncan', 'Lawrence Staib', 'Yifei Min', 'Weicheng Dai', 'Chenyu You'] | 2023-04-05 | null | null | null | null | ['semi-supervised-medical-image-segmentation'] | ['computer-vision'] | [ 3.85753632e-01 5.19458801e-02 -5.13927639e-01 -5.07349551e-01
-1.12574482e+00 -4.75032806e-01 1.34497344e-01 5.56752205e-01
-7.12305844e-01 5.70494056e-01 -1.44568190e-01 -2.57136524e-01
-1.69761583e-01 -5.70717573e-01 -5.43201387e-01 -1.18573737e+00
1.07639041e-02 8.59044611e-01 3.47576916e-01 2.33737364e-01
1.30947873e-01 3.64744216e-01 -1.26071668e+00 1.74662188e-01
1.14178085e+00 1.01790166e+00 -7.54702240e-02 6.20303690e-01
-2.95144133e-02 6.16247654e-01 -5.30636430e-01 -4.30045962e-01
2.81926364e-01 -6.81240022e-01 -8.81701171e-01 1.78492978e-01
5.56282401e-01 -2.22552806e-01 1.52548373e-01 1.10851681e+00
7.90429056e-01 -1.04382239e-01 1.08816314e+00 -1.13939846e+00
-1.75930306e-01 3.04358661e-01 -1.06528878e+00 5.47950804e-01
-2.20590264e-01 1.41096383e-01 1.07066488e+00 -5.93217492e-01
5.00738978e-01 5.92394352e-01 6.96790278e-01 5.06978452e-01
-1.31449723e+00 -7.68872380e-01 5.22176810e-02 1.13126010e-01
-1.43129098e+00 -1.34270594e-01 8.53926063e-01 -4.35660213e-01
3.39192390e-01 4.33459908e-01 4.10715133e-01 5.06453156e-01
1.59817398e-01 9.01374578e-01 1.21101093e+00 -4.26920146e-01
3.77499670e-01 1.49346277e-01 2.57264763e-01 7.42228031e-01
2.28942767e-01 -1.46152914e-01 -3.98455173e-01 -4.29832369e-01
2.30064556e-01 1.81316640e-02 -1.70436636e-01 -6.21694684e-01
-9.81450856e-01 8.74108911e-01 5.83009183e-01 1.29270986e-01
-3.13203245e-01 -9.52920988e-02 4.85432208e-01 7.69342333e-02
6.64639175e-01 1.60164177e-01 -3.03283960e-01 1.45872340e-01
-1.19973791e+00 8.51460248e-02 4.16851640e-01 5.45066714e-01
5.27905345e-01 -5.32680273e-01 -3.00538689e-01 1.00819325e+00
8.12548399e-02 4.72796857e-01 4.15538311e-01 -8.89671147e-01
3.53525192e-01 5.80267251e-01 -1.98559344e-01 -6.22771144e-01
-5.43570101e-01 -7.84951627e-01 -9.47283745e-01 1.29562885e-01
7.49155760e-01 1.22082215e-02 -1.09125960e+00 1.72229993e+00
6.81968510e-01 -8.86805877e-02 -2.51394838e-01 9.19636071e-01
3.90780270e-01 3.21897417e-01 1.88754529e-01 -2.93430924e-01
1.46928799e+00 -9.63208556e-01 -5.92627406e-01 -1.50660127e-01
9.32034016e-01 -5.67309201e-01 1.19987381e+00 3.74008298e-01
-1.04844403e+00 -1.59603536e-01 -9.63030815e-01 1.48741126e-01
-1.07218951e-01 2.55007949e-02 3.35969210e-01 8.68331194e-01
-7.42967129e-01 6.43096685e-01 -1.03631210e+00 1.25798956e-01
9.53564227e-01 4.09884185e-01 -7.70629942e-02 -2.47470573e-01
-9.85153794e-01 3.95991892e-01 7.79749453e-02 -2.21166328e-01
-5.56875348e-01 -9.19608772e-01 -7.86977828e-01 -2.90876955e-01
4.47973192e-01 -5.58738828e-01 9.28153455e-01 -1.00791156e+00
-9.78383482e-01 1.27880323e+00 -4.17464115e-02 -4.97351617e-01
7.14288950e-01 2.31456667e-01 1.06504774e-02 4.53174591e-01
2.96301603e-01 8.19831669e-01 7.01427281e-01 -1.31300545e+00
-6.72552288e-01 -7.75380731e-01 -1.46341637e-01 3.56020987e-01
-5.39525986e-01 -5.86411320e-02 -2.57494748e-01 -7.00523078e-01
3.13816279e-01 -1.14569116e+00 -3.90147924e-01 5.29456794e-01
-4.94067281e-01 -4.52628843e-02 3.18422288e-01 -4.25247580e-01
1.37777030e+00 -2.29175901e+00 -2.09119067e-01 5.95229745e-01
5.00452220e-01 -9.95874777e-03 2.89591908e-01 -3.04162532e-01
-1.34930573e-02 6.78235218e-02 -8.18378031e-01 -2.94612139e-01
-1.40650183e-01 2.24061236e-01 1.17532402e-01 9.31471467e-01
9.62416828e-02 5.49941838e-01 -8.99605215e-01 -9.58153963e-01
-4.96840384e-03 2.70775825e-01 -8.85277867e-01 -2.46541947e-02
1.04427330e-01 3.37480336e-01 -3.63850892e-01 7.20725715e-01
8.48083735e-01 -5.45299411e-01 1.43588573e-01 -2.06425458e-01
2.99806833e-01 -4.65451106e-02 -1.00931728e+00 1.35763669e+00
-2.61523604e-01 2.16017097e-01 1.90749481e-01 -1.17185247e+00
5.51111817e-01 -7.86502138e-02 8.31590176e-01 -7.92777717e-01
6.44482896e-02 4.75326151e-01 1.60330564e-01 -2.24546045e-01
1.01771623e-01 -6.15975916e-01 3.03108525e-02 4.59569335e-01
-2.46079236e-01 -2.36784235e-01 -1.41121885e-02 1.05733767e-01
1.00280511e+00 -2.90761977e-01 1.14252515e-01 -4.03554738e-01
3.03274602e-01 -4.59958278e-02 5.59862614e-01 7.00212777e-01
-7.19008446e-01 9.52244401e-01 6.74613535e-01 -7.68932253e-02
-8.05721760e-01 -1.27380002e+00 -6.33542180e-01 1.12155700e+00
3.93592685e-01 -1.73745938e-02 -9.13239837e-01 -1.28157032e+00
-8.19775313e-02 4.65004057e-01 -7.92688549e-01 -2.65321463e-01
-5.59511423e-01 -1.44074833e+00 5.22884250e-01 5.29696405e-01
1.86848104e-01 -7.62068927e-01 -7.67664194e-01 3.20111811e-02
-2.57461786e-01 -8.33061039e-01 -7.30642438e-01 7.54044831e-01
-9.19852793e-01 -1.10344410e+00 -1.08310688e+00 -7.54556000e-01
1.16871548e+00 5.71958460e-02 1.21605504e+00 1.86278597e-01
-4.69863892e-01 3.75863467e-03 -2.75218964e-01 -3.08034420e-01
-1.49674192e-01 2.26749837e-01 -3.26751649e-01 -3.09514324e-03
1.07471406e-01 -2.62386084e-01 -1.16996765e+00 5.80278158e-01
-1.09478867e+00 -2.49525160e-01 3.86015445e-01 1.05459750e+00
1.04229963e+00 9.13669616e-02 5.18221378e-01 -1.25678515e+00
8.86424854e-02 -3.99757326e-01 -2.71452516e-01 2.22749576e-01
-8.18594038e-01 -8.69174078e-02 4.72617596e-01 -2.85884917e-01
-6.20040357e-01 5.26518486e-02 -3.31229329e-01 -1.86988175e-01
8.98037404e-02 1.48321927e-01 8.80915374e-02 -4.20743041e-02
6.70117319e-01 5.11734895e-02 8.97688121e-02 -3.76401916e-02
9.46771652e-02 7.83045173e-01 3.32171202e-01 -5.95475376e-01
4.81609792e-01 1.05063474e+00 -6.53183162e-02 -4.37704027e-01
-1.15543449e+00 -7.06624091e-01 -3.95127177e-01 -1.00215487e-01
7.85052359e-01 -6.60557985e-01 -3.52659613e-01 5.62086105e-01
-4.49786156e-01 -6.13707840e-01 -6.09395564e-01 4.67565715e-01
-6.32865727e-01 3.64946604e-01 -7.05208719e-01 -5.97277403e-01
-4.93752450e-01 -1.27191567e+00 1.31538963e+00 1.91687867e-01
-2.02137440e-01 -9.94276404e-01 8.57790411e-02 7.02213168e-01
2.10170418e-01 2.69369572e-01 1.08503950e+00 -8.27682734e-01
-6.50770068e-02 -1.33633733e-01 -2.24770129e-01 3.96372736e-01
1.80380702e-01 -1.46509305e-01 -8.13356459e-01 -4.07817155e-01
-1.50866047e-01 -4.60123867e-01 1.01870596e+00 6.42513692e-01
1.60788918e+00 5.48366345e-02 -4.79500592e-01 7.40141451e-01
1.17021310e+00 -7.88276270e-02 4.98207271e-01 2.23547935e-01
4.40554738e-01 7.55934000e-01 8.47853601e-01 5.01772046e-01
3.80378842e-01 6.23105228e-01 4.72776771e-01 -4.61537898e-01
-1.85712397e-01 4.78870384e-02 -6.36153445e-02 7.35647559e-01
4.45433438e-01 -2.15043813e-01 -9.76223469e-01 7.03288078e-01
-1.32963300e+00 -5.19726157e-01 2.06588835e-01 2.32505918e+00
1.42643857e+00 4.80246991e-01 3.51645291e-01 3.89041901e-01
7.16771066e-01 9.51168537e-02 -7.83769369e-01 -3.67780663e-02
-1.12057872e-01 3.61472726e-01 7.21319020e-01 4.05347347e-01
-1.23423409e+00 4.89264548e-01 5.70095873e+00 1.37602043e+00
-1.23984241e+00 2.66260415e-01 1.37412441e+00 -3.41752619e-01
-3.55971217e-01 -2.99944311e-01 -6.72938168e-01 8.04427683e-01
6.16089225e-01 2.57376373e-01 -7.57134184e-02 5.22610664e-01
-5.67534715e-02 -2.68897504e-01 -8.83451223e-01 7.90735126e-01
1.93888605e-01 -1.06509197e+00 -1.37072831e-01 1.02339827e-01
7.70246446e-01 -7.91441277e-02 2.92990446e-01 5.43876700e-02
8.11893791e-02 -7.76209772e-01 5.86230695e-01 -5.48582934e-02
9.32896852e-01 -7.57835448e-01 7.26562142e-01 3.21225554e-01
-7.53690124e-01 -9.54197068e-03 -9.81767252e-02 5.96968055e-01
1.72160100e-02 1.00652993e+00 -6.89900994e-01 3.19352448e-01
7.13881671e-01 3.98771793e-01 -6.30988181e-01 1.18704677e+00
2.02147454e-01 7.94984043e-01 -5.23004413e-01 1.31519496e-01
2.55016088e-01 -1.06425419e-01 2.45871291e-01 1.11318946e+00
2.22350415e-02 -1.85084827e-02 2.35169604e-01 3.52846652e-01
-1.53981477e-01 3.32742155e-01 2.88037919e-02 3.77921760e-01
2.31205836e-01 1.16957235e+00 -1.26234829e+00 -3.74340326e-01
-1.80104356e-02 7.76845813e-01 1.25460893e-01 2.92043388e-02
-1.04916167e+00 -2.90726721e-01 1.64436907e-01 5.06910682e-01
2.03655928e-01 2.82246917e-01 -5.73616564e-01 -9.12570059e-01
7.50182122e-02 -7.17612267e-01 1.02745068e+00 -3.44839305e-01
-1.37363982e+00 4.78794724e-01 -4.53214087e-02 -1.26072991e+00
1.84913054e-01 -4.87530261e-01 -3.42628211e-01 2.80584127e-01
-1.83389783e+00 -8.35623682e-01 -2.43631035e-01 5.62358439e-01
2.82785207e-01 4.65347886e-01 3.82968456e-01 7.08706558e-01
-4.62294608e-01 1.11458826e+00 2.23084465e-01 1.81046978e-01
1.00519860e+00 -1.56115878e+00 -3.38086545e-01 3.83115441e-01
-3.32480147e-02 1.67026699e-01 5.87226450e-01 -4.51029897e-01
-7.28712618e-01 -1.12308455e+00 4.70114291e-01 -1.76019132e-01
3.82999718e-01 -2.27304935e-01 -9.19239342e-01 1.99903071e-01
-2.03769192e-01 6.12385094e-01 9.38164234e-01 -2.83683866e-01
-1.57484129e-01 -4.17801291e-01 -1.72314298e+00 4.85260278e-01
9.08987999e-01 -2.60906219e-01 -1.85908049e-01 6.67537928e-01
4.47679013e-01 -4.21892017e-01 -8.08924198e-01 8.01108062e-01
4.36577082e-01 -8.91303182e-01 9.41985011e-01 -3.83485526e-01
4.17712778e-01 -2.31767669e-01 -1.00619443e-01 -1.11468077e+00
8.22883621e-02 -2.57230878e-01 9.17419270e-02 8.79162729e-01
6.05911851e-01 -7.33472586e-01 1.15413380e+00 2.97144741e-01
-1.72419474e-01 -1.42160928e+00 -1.13038349e+00 -6.10345602e-01
5.16842186e-01 -1.51727080e-01 9.58773270e-02 9.37787414e-01
-3.09506536e-01 -9.90562215e-02 9.36368927e-02 -3.57525200e-02
8.48833799e-01 2.81250060e-01 2.46444866e-01 -1.03071380e+00
-2.90782779e-01 -5.55988252e-01 -2.64740944e-01 -9.38025951e-01
5.53316735e-02 -9.72166896e-01 1.07443884e-01 -1.19374979e+00
6.62579238e-01 -1.02324033e+00 -5.95509291e-01 4.27519888e-01
-5.24999738e-01 7.26568878e-01 -2.43022963e-01 2.31222942e-01
-8.44006658e-01 4.20302153e-01 1.29352689e+00 -3.10052723e-01
-3.84681225e-02 1.54640853e-01 -6.92738235e-01 6.76674306e-01
6.85481727e-01 -7.73485184e-01 -2.62589246e-01 -4.66593206e-02
2.38092303e-01 -2.56728351e-01 3.89921278e-01 -8.33469570e-01
1.48385718e-01 -1.29768094e-02 2.86340117e-01 -5.91615081e-01
5.23323044e-02 -7.19369531e-01 -4.87168312e-01 6.84611976e-01
-5.57520449e-01 -2.15828270e-01 -6.51095212e-02 4.96819854e-01
-1.67607516e-01 -2.72228032e-01 1.39400029e+00 1.08940013e-01
3.82607579e-02 4.54154581e-01 -3.32447998e-02 6.69457495e-01
1.17533743e+00 -2.05431908e-01 -2.94271171e-01 -8.64803791e-02
-7.25029290e-01 4.39610302e-01 5.35118878e-01 -1.92506462e-01
3.48717391e-01 -1.02329791e+00 -6.45592093e-01 8.81695002e-02
3.06526124e-01 2.72500336e-01 5.00069737e-01 1.32836187e+00
-6.34556592e-01 -1.68108165e-01 1.87629715e-01 -9.55152869e-01
-1.23657870e+00 4.24482942e-01 5.94406426e-01 -5.81287503e-01
-4.49570775e-01 9.72938538e-01 4.00813341e-01 -5.26973486e-01
2.09473461e-01 -1.72657073e-01 2.88280402e-03 2.08219931e-01
2.92615235e-01 2.83841848e-01 3.17135990e-01 -4.33224767e-01
-6.56006694e-01 6.88206613e-01 -2.95693487e-01 -2.99768732e-03
1.04428256e+00 -6.79439306e-02 2.45518759e-02 2.69658536e-01
1.38728452e+00 1.86090082e-01 -1.37980831e+00 -3.08051407e-01
-1.26032725e-01 -3.37785780e-01 1.44171700e-01 -7.40570426e-01
-1.39400733e+00 9.35051501e-01 9.64856625e-01 9.87494513e-02
1.28683054e+00 3.36983979e-01 9.96228695e-01 -1.78046197e-01
1.58789754e-02 -1.30182123e+00 1.79594561e-01 -2.77904049e-02
2.67545819e-01 -1.29207408e+00 2.21289724e-01 -5.30624568e-01
-7.06958771e-01 6.72124207e-01 4.45961744e-01 -1.07948817e-02
8.13272119e-01 5.21104753e-01 1.86135620e-01 -3.24089557e-01
-4.40200150e-01 -8.15661550e-02 2.15495378e-01 2.98100591e-01
2.80343413e-01 1.93432674e-01 -3.99702102e-01 4.00872082e-01
-1.34982824e-01 -2.38765225e-01 2.42118508e-01 9.87768769e-01
-3.64303976e-01 -8.39467049e-01 -3.06564391e-01 9.53525066e-01
-9.00125682e-01 -9.32833627e-02 -2.19855662e-02 4.47881371e-01
2.61177242e-01 6.13629282e-01 1.89591572e-01 6.32897113e-03
1.27506539e-01 -1.77984998e-01 4.74983126e-01 -4.36589569e-01
-5.91470897e-01 1.69093862e-01 -3.42915297e-01 -2.76955336e-01
-3.98309410e-01 -6.96501791e-01 -1.47083032e+00 3.92564386e-02
-5.08936584e-01 1.09173089e-01 5.14852762e-01 8.19957435e-01
5.68177737e-02 4.74006355e-01 9.72647130e-01 -3.34219962e-01
-8.40613306e-01 -4.38585252e-01 -6.08225524e-01 5.55724621e-01
4.55685705e-01 -7.05443203e-01 -7.87203193e-01 -1.19161345e-01] | [14.719650268554688, -2.3124208450317383] |
6a95e8a5-0c8c-48d1-9d68-3640076a4b9d | a-predictive-model-for-the-identification-of | null | null | https://link.springer.com/article/10.1007%2Fs10586-019-02992-4 | https://link.springer.com/article/10.1007%2Fs10586-019-02992-4 | A predictive model for the identification of learning styles in MOOC environments | Massive online open course (MOOC) platform generates a large amount of data, which provides many opportunities for studying the behaviors of learners. In parallel, recent advancements in machine learning techniques and big data analysis have created new opportunities for a better understanding of how learners behave and learn in environments known for their massiveness and openness. The work is about predicting learners’ learning styles based on their learning traces. The Felder Silverman learning style model (FSLSM) is adopted since it is one of the most commonly used models in technology-enhanced learning. In order to attend our objective, we analyzed data collected from the edX course “statistical learning” (session Winter 2015 and Winter 2016), administered via Stanford’s Logunita platform. The results show that decision tree performs best for all 3 dimensions, with an accuracy of higher than 98% and a reduced risk of overfitting the training data. | ['Omar Baz', 'Ali El Mezouary', 'Brahim Hmedna'] | 2019-10-12 | null | null | null | null | ['event-data-classification', 'automatic-machine-learning-model-selection', 'clustering-algorithms-evaluation'] | ['computer-vision', 'methodology', 'methodology'] | [-6.02659583e-01 -1.67299688e-01 -4.59980726e-01 -3.36040735e-01
-3.48736316e-01 -4.27083433e-01 1.89962134e-01 5.54288805e-01
-1.44955114e-01 5.23634851e-01 2.29787394e-01 -6.49723947e-01
-5.91247916e-01 -9.84435141e-01 -6.14883482e-01 -3.51468295e-01
2.94633843e-02 -1.22678339e-01 1.25871211e-01 -3.21752429e-01
4.00967360e-01 2.63292879e-01 -2.14828229e+00 3.07180017e-01
1.38048863e+00 7.85387397e-01 1.91463307e-01 6.36979997e-01
-4.21000093e-01 1.36445951e+00 -3.93865287e-01 -5.79710841e-01
1.61158130e-01 -2.53540486e-01 -1.08149254e+00 -1.93500355e-01
7.90067434e-01 -3.13844681e-02 -2.69688427e-01 8.78010809e-01
1.22947134e-01 4.06114012e-01 4.09058958e-01 -1.07691753e+00
-5.76713145e-01 7.63821304e-01 -1.40611142e-01 2.08133906e-01
3.82542104e-01 -1.94005638e-01 9.47026432e-01 -4.89479721e-01
5.72598100e-01 1.02268231e+00 6.06724083e-01 5.60106218e-01
-9.39298689e-01 -5.86639404e-01 -9.64866132e-02 2.57797629e-01
-1.05299819e+00 4.26035970e-02 2.88950175e-01 -7.59817898e-01
2.42282823e-01 3.13638359e-01 8.09412837e-01 1.05055392e+00
5.19279003e-01 8.98320913e-01 1.46429741e+00 -7.50289917e-01
3.95413935e-01 6.34288192e-01 7.94718206e-01 6.67722940e-01
2.77692348e-01 -3.97540033e-02 -7.34076858e-01 1.04821928e-01
2.95881987e-01 6.30388319e-01 1.91755548e-01 -1.79707438e-01
-6.80673957e-01 1.01784968e+00 2.21258253e-01 3.34866315e-01
-1.16862901e-01 -4.46895629e-01 -2.56128281e-01 9.84690964e-01
1.83108822e-01 3.89394104e-01 -6.20981038e-01 -8.21813345e-01
-3.03335488e-01 4.52661097e-01 9.84417558e-01 1.07418597e+00
6.37109756e-01 -1.87285200e-01 9.63457022e-03 8.90019596e-01
5.35879791e-01 -1.17416464e-01 1.01125371e+00 -8.84453177e-01
3.92936677e-01 1.27441716e+00 -3.09892625e-01 -5.84169984e-01
-2.31164888e-01 -6.09787345e-01 -2.17386782e-01 3.38908672e-01
8.75485599e-01 -3.99473846e-01 -3.61580074e-01 1.23166442e+00
2.91673988e-01 -4.55662906e-02 -7.57210925e-02 2.78166860e-01
7.55013168e-01 3.17938268e-01 1.58094049e-01 -2.16662623e-02
1.41556811e+00 -6.74268484e-01 -8.06272268e-01 -7.72092864e-02
1.23063946e+00 -7.32832193e-01 1.52600133e+00 9.01664317e-01
-1.06983030e+00 -7.47318327e-01 -8.99982274e-01 -1.12400288e-02
-5.65906703e-01 -5.97607255e-01 1.02384090e+00 1.12709963e+00
-7.55675614e-01 7.38667667e-01 -5.45227230e-01 -3.61483783e-01
6.19088769e-01 2.25268409e-01 -3.58193487e-01 -2.17718273e-01
-7.15471864e-01 3.79448920e-01 4.28434312e-02 -7.52243519e-01
-6.89125657e-01 -1.14889348e+00 -4.03726488e-01 3.68461370e-01
1.43341586e-01 -1.77300453e-01 1.28704822e+00 -9.34440255e-01
-1.59217501e+00 6.75050318e-01 3.33818614e-01 -7.94076398e-02
4.26456690e-01 -5.15327692e-01 -2.98475921e-01 -5.55696547e-01
-4.40111369e-01 -5.95407886e-03 1.96296424e-01 -6.17826223e-01
-9.67896819e-01 -8.47649693e-01 -3.93196866e-02 -1.19350202e-01
-1.19234502e+00 1.34266257e-01 1.95075288e-01 -3.62732500e-01
1.34868205e-01 -9.71201420e-01 -7.76979402e-02 -7.94599771e-01
2.85006911e-01 -7.11351335e-01 6.20972157e-01 -4.73704785e-01
1.80604422e+00 -2.01948786e+00 -2.80506372e-01 2.31883064e-01
5.05149245e-01 1.50736451e-01 2.19468206e-01 4.71283346e-01
2.07787409e-01 1.93900526e-01 5.45558035e-01 -5.61727323e-02
-1.18345901e-01 -1.21054657e-01 -1.84352268e-02 -1.50879964e-01
-5.34936607e-01 4.52234030e-01 -8.77642274e-01 -3.49108785e-01
-5.29547669e-02 1.89045802e-01 -8.33800256e-01 6.14625037e-01
1.67656913e-01 2.49227434e-01 -4.39715594e-01 5.62074721e-01
2.12697312e-01 -1.12170346e-01 7.14031532e-02 1.00722408e+00
-4.93671417e-01 3.85797381e-01 -1.25127530e+00 1.51983333e+00
-5.22281468e-01 7.30486333e-01 -2.97388107e-01 -4.68509227e-01
1.27662110e+00 2.63859987e-01 4.44505900e-01 -6.67630076e-01
4.89155315e-02 1.95002511e-01 2.06301808e-01 -8.08136642e-01
3.62184197e-01 4.37342972e-02 1.84216216e-01 4.78892446e-01
3.76814961e-01 2.35516757e-01 1.41182169e-01 1.93482876e-01
1.16508257e+00 -8.48228708e-02 9.34871882e-02 -5.56368053e-01
1.92056969e-01 -2.23582596e-01 6.95153236e-01 5.32804012e-01
-4.67313439e-01 1.12895906e-01 4.93416607e-01 -6.43002152e-01
-4.03265417e-01 -8.47535014e-01 -2.22649619e-01 1.62538540e+00
-5.53934574e-01 -7.25358546e-01 -8.60822856e-01 -6.37373745e-01
1.70394167e-01 7.17619359e-01 -3.94901574e-01 -2.34651849e-01
-2.29150459e-01 -3.26136023e-01 -2.24750582e-02 3.09659511e-01
2.98579186e-01 -9.94970679e-01 -3.68206799e-01 7.37226456e-02
4.39591631e-02 -7.67517567e-01 -3.24367620e-02 1.37717634e-01
-1.40761387e+00 -9.52274978e-01 -1.92792609e-01 -8.71549428e-01
5.24709702e-01 2.20134035e-01 1.10721385e+00 2.56282717e-01
-3.36048573e-01 3.21444690e-01 -4.43205208e-01 -1.06389141e+00
-3.53320658e-01 4.40007240e-01 3.44261885e-01 5.46207726e-02
1.04427242e+00 -5.87622523e-01 -3.57411593e-01 1.48551747e-01
-6.89670503e-01 -2.72981048e-01 5.58002532e-01 3.45094889e-01
1.42435851e-02 3.26644659e-01 8.16920161e-01 -1.25237048e+00
7.74236083e-01 -7.56878793e-01 -8.05284262e-01 1.05604313e-01
-1.34096074e+00 -5.78554720e-02 5.59832871e-01 -4.30096418e-01
-1.22437239e+00 -3.19005638e-01 -1.64142847e-02 3.55159002e-03
-5.67035735e-01 6.00924492e-01 -6.05495647e-02 -3.11651677e-01
8.94074976e-01 -3.83789718e-01 -5.19080721e-02 -8.97844911e-01
-3.24425578e-01 1.11006582e+00 -6.03997335e-02 -5.23428082e-01
6.31242573e-01 -1.54619262e-01 -5.87730110e-02 -9.05059636e-01
-1.16094303e+00 -4.46964830e-01 -8.40838611e-01 -4.30242032e-01
6.42591059e-01 -8.51475656e-01 -1.29295254e+00 4.45403934e-01
-1.05082005e-01 -6.17141068e-01 -2.25551695e-01 1.01872325e+00
-1.45423681e-01 -2.82451719e-01 -6.76031768e-01 -8.88739705e-01
2.08611056e-01 -9.08016026e-01 -1.64920241e-01 9.56172049e-01
-2.27914646e-01 -1.36793506e+00 1.46777555e-01 1.20213985e+00
5.21680653e-01 1.37510365e-02 1.04498327e+00 -9.62430537e-01
-5.51013470e-01 -3.09416443e-01 5.34848869e-01 2.93992251e-01
-1.47578761e-01 2.63314158e-01 -1.18182683e+00 -1.73258856e-01
1.36411205e-01 -4.92648333e-01 3.82562220e-01 1.16638057e-01
1.41077852e+00 -2.72268295e-01 1.44373894e-01 4.63204890e-01
1.34044909e+00 4.20411266e-02 4.12678212e-01 7.92683482e-01
7.03644872e-01 8.76839280e-01 4.01473075e-01 4.98516321e-01
6.56153560e-01 2.69098371e-01 2.44272426e-01 7.24781394e-01
1.95844531e-01 -5.82119882e-01 6.96220577e-01 1.18031573e+00
-2.33490199e-01 2.03010038e-01 -1.20869374e+00 1.97489157e-01
-1.52626514e+00 -8.74888122e-01 -7.46879339e-01 2.57218242e+00
7.78852701e-01 1.90855443e-01 4.46174115e-01 2.56685048e-01
8.35227668e-02 -1.76459700e-01 1.23819776e-01 -8.98693085e-01
3.29319298e-01 2.61239320e-01 2.51867414e-01 3.41240138e-01
-5.75525701e-01 6.79450512e-01 6.49703074e+00 5.74754536e-01
-8.62086892e-01 -7.40174800e-02 3.76545042e-01 -1.22272491e-01
-3.22099358e-01 -4.82363477e-02 -1.23533022e+00 4.06394303e-01
1.65699685e+00 -6.93422407e-02 4.79028314e-01 9.69865739e-01
1.34527579e-01 -2.61761360e-02 -9.87443686e-01 7.28992462e-01
-3.07218224e-01 -1.18685269e+00 -2.74157047e-01 6.14691436e-01
1.17614341e+00 -1.81504413e-01 1.88342243e-01 1.03152800e+00
6.28774822e-01 -1.10102916e+00 3.36712182e-01 5.33018768e-01
-9.49273929e-02 -8.93327892e-01 5.28913677e-01 9.25509751e-01
-5.14068425e-01 -8.58317971e-01 -4.34753984e-01 -8.43475997e-01
-8.95282924e-01 4.18464780e-01 -6.94471121e-01 8.01530555e-02
9.17771876e-01 5.68922818e-01 -8.57416928e-01 1.00544524e+00
-8.82789120e-02 1.32645369e+00 2.23780543e-01 -2.88286179e-01
-2.26645786e-02 -5.16528487e-01 2.61248294e-02 6.46715939e-01
4.17626172e-01 1.64338633e-01 7.86639005e-02 6.47080481e-01
-2.63982534e-01 5.05797207e-01 -6.19552135e-01 -1.92441776e-01
5.87286413e-01 1.47452319e+00 -3.27414483e-01 4.76640426e-02
-9.74137783e-01 1.78882867e-01 5.28995633e-01 5.85617386e-02
-9.89287347e-03 -4.35958028e-01 7.35664189e-01 4.71335441e-01
-3.80542427e-01 -1.55806541e-01 -5.98203182e-01 -1.13222086e+00
-2.58014083e-01 -1.17509460e+00 6.74121082e-01 -7.29365945e-02
-1.16084552e+00 -4.51621786e-02 -5.30327797e-01 -1.16226697e+00
4.55339849e-02 -6.95177853e-01 -1.00493121e+00 6.50785387e-01
-1.19346166e+00 -5.33841610e-01 -8.12804043e-01 7.00462341e-01
5.27990103e-01 -6.71528280e-01 9.35302973e-01 3.27640295e-01
-9.41007972e-01 7.22401440e-01 5.57940662e-01 1.98749881e-02
7.91834891e-01 -1.51875770e+00 -1.82550117e-01 4.71468538e-01
2.23137289e-01 6.97311401e-01 5.23106575e-01 -4.44958061e-01
-1.85370326e+00 -7.09380925e-01 1.18388271e+00 -8.97024393e-01
6.31284595e-01 -3.41246843e-01 -9.92704570e-01 9.36447382e-01
1.67917963e-02 -1.59380585e-01 1.49305475e+00 7.13933110e-01
-1.03016563e-01 -4.28012550e-01 -8.87557507e-01 4.42938775e-01
6.70448124e-01 -2.30094239e-01 -5.34277737e-01 1.37761593e-01
3.35705012e-01 -1.72384068e-01 -1.64854360e+00 -2.40256507e-02
6.56580865e-01 -1.39596355e+00 5.21141946e-01 -1.04732776e+00
6.55943334e-01 5.46947658e-01 -7.41988495e-02 -1.13496852e+00
-3.16325158e-01 -8.59463096e-01 -4.70269859e-01 1.54807329e+00
1.72454059e-01 -4.93270308e-01 1.13596916e+00 1.13377512e+00
-1.99665815e-01 -8.75529766e-01 -2.33220801e-01 -8.30174148e-01
3.97072583e-01 -3.18807781e-01 5.20853758e-01 1.35022533e+00
4.67625916e-01 3.25650901e-01 1.94754392e-01 -4.10610318e-01
7.50165701e-01 -2.15699128e-03 1.08492053e+00 -2.06307077e+00
-9.49047226e-03 -5.52146971e-01 -4.27655041e-01 -5.96533477e-01
7.80500770e-02 -9.58987951e-01 -6.35014653e-01 -1.06004179e+00
2.54076183e-01 -5.10742664e-01 -4.99540895e-01 1.96411595e-01
-2.91073561e-01 -2.76210994e-01 8.20368081e-02 -4.91673984e-02
-5.07452428e-01 1.80241272e-01 1.28223681e+00 5.76723278e-01
-5.68967164e-01 5.14493525e-01 -1.02157092e+00 9.29949820e-01
7.04239070e-01 -3.31704229e-01 -3.74327421e-01 -2.18742862e-01
3.94409746e-01 -3.91659066e-02 -8.05290192e-02 -1.22873402e+00
1.55340508e-01 -6.57368839e-01 4.64526087e-01 -3.58050689e-02
-2.38161251e-01 -8.31110895e-01 -3.30841243e-01 3.89114469e-01
-6.99414134e-01 -4.99006994e-02 -9.48641598e-02 7.93677643e-02
-3.18037063e-01 -5.08487761e-01 7.29033947e-01 -2.61905137e-02
-5.54972053e-01 4.34454203e-01 -4.00136381e-01 2.03801319e-01
1.32776868e+00 -1.28563687e-01 -3.30432385e-01 -2.89337426e-01
-4.88681704e-01 3.24637443e-01 2.79313326e-01 9.14665043e-01
3.48504364e-01 -1.31051493e+00 -6.03635848e-01 5.58951914e-01
1.60209551e-01 -1.03094809e-01 3.05898011e-01 7.79358387e-01
-5.13293147e-01 4.13409323e-01 -5.30758262e-01 -2.78842986e-01
-1.43777299e+00 3.10872316e-01 4.64886278e-02 -2.07524136e-01
-4.40182179e-01 7.85325587e-01 -3.11095208e-01 -6.08419001e-01
4.84001875e-01 -1.43438652e-02 -6.42247260e-01 2.12752968e-01
8.70891213e-01 1.20608127e+00 1.91510528e-01 9.92868003e-03
2.63508141e-01 6.65186942e-02 -1.96022928e-01 1.83779344e-01
1.63682222e+00 -1.39964996e-02 2.63481110e-01 1.13374424e+00
1.01794088e+00 5.32961965e-01 -9.64290380e-01 -2.00719193e-01
6.07411027e-01 -7.75527596e-01 2.32698441e-01 -6.44990981e-01
-7.09703267e-01 1.19371605e+00 6.91610038e-01 5.42601049e-01
5.94688356e-01 -2.69722283e-01 4.72470254e-01 3.82876605e-01
3.49851370e-01 -1.37698972e+00 2.93523401e-01 8.21811497e-01
2.10360438e-01 -1.48565257e+00 -3.01654071e-01 1.19181447e-01
-3.70256841e-01 1.29041970e+00 1.13623738e+00 1.93023056e-01
9.60150182e-01 -3.53831947e-02 3.13186973e-01 -7.96867013e-02
-1.07592320e+00 3.49470794e-01 3.27416897e-01 2.88012743e-01
1.20989859e+00 2.99548179e-01 -3.00125301e-01 1.04916978e+00
-6.20912731e-01 8.44808891e-02 8.36535573e-01 1.03834534e+00
-9.42589879e-01 -1.32875669e+00 -4.85606492e-01 7.13616490e-01
-7.65722275e-01 2.22204670e-01 -3.09614629e-01 6.06036127e-01
-8.43945798e-03 1.03043413e+00 -1.07802160e-01 -6.58858418e-01
2.83480555e-01 6.08773649e-01 4.36408013e-01 -7.45776474e-01
-1.04208648e+00 -4.18065161e-01 -4.28154320e-01 -4.57409143e-01
3.15361679e-01 -9.92759228e-01 -1.27592158e+00 -7.76991963e-01
-2.68053353e-01 4.05490309e-01 8.38645041e-01 9.26124096e-01
2.73915023e-01 5.76466978e-01 9.35394287e-01 1.05524801e-01
-9.14144993e-01 -1.09892619e+00 -9.99439180e-01 5.49960196e-01
-8.75611696e-03 -3.60444278e-01 -4.59577471e-01 -7.16937035e-02] | [10.127315521240234, 7.205688953399658] |
af2d812d-5112-44fa-8b7f-c592090d6010 | mawps-a-math-word-problem-repository | null | null | https://aclanthology.org/N16-1136 | https://aclanthology.org/N16-1136.pdf | MAWPS: A Math Word Problem Repository | null | ['Rik Koncel-Kedziorski', 'Nate Kushman', 'Subhro Roy', 'Hannaneh Hajishirzi', 'Aida Amini'] | 2016-06-01 | null | null | null | naacl-2016-6 | ['math-word-problem-solving', 'math-word-problem-solving', 'math-word-problem-solving'] | ['knowledge-base', 'reasoning', 'time-series'] | [-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] | [-7.498989105224609, 3.5796220302581787] |
0bd90b69-b02c-4ee6-8800-edcb3c2b8adc | toward-subgraph-guided-knowledge-graph | 2004.06015 | null | https://arxiv.org/abs/2004.06015v4 | https://arxiv.org/pdf/2004.06015v4.pdf | Toward Subgraph-Guided Knowledge Graph Question Generation with Graph Neural Networks | Knowledge graph (KG) question generation (QG) aims to generate natural language questions from KGs and target answers. Previous works mostly focus on a simple setting which is to generate questions from a single KG triple. In this work, we focus on a more realistic setting where we aim to generate questions from a KG subgraph and target answers. In addition, most of previous works built on either RNN-based or Transformer based models to encode a linearized KG sugraph, which totally discards the explicit structure information of a KG subgraph. To address this issue, we propose to apply a bidirectional Graph2Seq model to encode the KG subgraph. Furthermore, we enhance our RNN decoder with node-level copying mechanism to allow directly copying node attributes from the KG subgraph to the output question. Both automatic and human evaluation results demonstrate that our model achieves new state-of-the-art scores, outperforming existing methods by a significant margin on two QG benchmarks. Experimental results also show that our QG model can consistently benefit the Question Answering (QA) task as a mean of data augmentation. | ['Lingfei Wu', 'Yu Chen', 'Mohammed J. Zaki'] | 2020-04-13 | null | null | null | null | ['kg-to-text'] | ['natural-language-processing'] | [ 3.74878019e-01 8.77279520e-01 1.65558457e-01 -3.17471385e-01
-1.15467691e+00 -6.04335606e-01 5.97942531e-01 -9.44624543e-02
8.01333413e-03 7.07149804e-01 4.43564117e-01 -5.90737224e-01
2.14356720e-01 -1.27724469e+00 -1.13274372e+00 -2.54484564e-01
5.52292645e-01 5.83164096e-01 3.27271163e-01 -6.71479702e-01
-2.84358542e-02 -2.57745802e-01 -1.15248883e+00 3.64081711e-01
1.28897917e+00 6.58963144e-01 1.01998955e-01 8.25511992e-01
-4.79687721e-01 1.18737769e+00 -7.53609300e-01 -1.01016140e+00
1.80090129e-01 -9.13263083e-01 -1.22519886e+00 -2.47911870e-01
7.97755182e-01 -3.29945505e-01 -5.26118040e-01 8.51337433e-01
6.02591932e-01 4.07231087e-03 3.37967813e-01 -1.24428999e+00
-1.16847932e+00 1.09065759e+00 -8.93203393e-02 -1.51950434e-01
5.30232668e-01 2.69085824e-01 1.49605715e+00 -6.49306297e-01
5.89412093e-01 1.25186431e+00 3.56914133e-01 7.43987679e-01
-9.71175790e-01 -3.65664124e-01 2.11863428e-01 4.14060831e-01
-1.12683737e+00 -2.55450189e-01 7.87577569e-01 7.97409564e-02
9.56157863e-01 2.62061149e-01 4.04370904e-01 8.60688269e-01
-1.18786901e-01 9.05009031e-01 9.39351320e-01 -5.07175684e-01
6.01072311e-02 -2.84668237e-01 9.67879519e-02 9.41752791e-01
9.20853689e-02 -3.08829218e-01 -4.51113045e-01 -5.94207756e-02
4.44882482e-01 -6.90312624e-01 -5.13300121e-01 -2.86293656e-01
-1.18936121e+00 9.85782981e-01 6.36025429e-01 -6.23381510e-02
-2.63045043e-01 4.21570331e-01 9.63213295e-02 4.90176320e-01
7.58122280e-02 6.14372730e-01 -4.54837441e-01 8.99519846e-02
-4.18033451e-01 5.12727797e-01 1.21933424e+00 1.28290021e+00
8.49856198e-01 1.55140571e-02 -8.74064386e-01 4.86587316e-01
1.67198673e-01 4.64292914e-01 2.68388152e-01 -8.54488909e-01
8.74304593e-01 7.90975690e-01 -1.28047496e-01 -7.74528503e-01
-2.28211910e-01 -5.65113008e-01 -6.22680664e-01 -5.67270517e-01
4.91221607e-01 -3.38355541e-01 -1.00016844e+00 1.96224141e+00
5.75310946e-01 1.43523976e-01 2.27497339e-01 8.46089959e-01
1.45512640e+00 5.94269395e-01 -4.75537851e-02 3.58110189e-01
1.40410435e+00 -1.34849441e+00 -6.05277717e-01 -2.52050757e-01
1.00023687e+00 -3.43803495e-01 1.60216689e+00 1.98511430e-03
-1.08917058e+00 -5.31797647e-01 -7.06737161e-01 -5.25016248e-01
-2.38742203e-01 -2.19104253e-02 3.30194324e-01 6.34444475e-01
-1.34519863e+00 1.63995013e-01 -2.95092344e-01 -2.50621647e-01
-1.41020510e-02 1.35127500e-01 -1.41246572e-01 -4.21287179e-01
-1.62978137e+00 6.16545022e-01 5.08669198e-01 5.12496457e-02
-6.97776258e-01 -8.28641593e-01 -1.07625949e+00 1.87639058e-01
7.04662144e-01 -1.58086514e+00 1.53915167e+00 -4.69161421e-01
-1.62336743e+00 5.51281869e-01 -1.66056484e-01 -5.16197264e-01
2.10968643e-01 4.53792920e-04 -1.96927667e-01 1.90388113e-01
6.46764273e-03 8.80407870e-01 6.46965981e-01 -1.02529979e+00
-2.96049148e-01 -2.66699851e-01 6.26915395e-01 3.86906236e-01
-1.72798991e-01 -3.13018769e-01 -6.23290718e-01 -6.07067287e-01
-3.79515216e-02 -7.94349134e-01 -9.93670970e-02 -6.31980598e-01
-7.73545980e-01 -5.69854617e-01 3.96562815e-01 -9.08994079e-01
1.19941235e+00 -1.53585160e+00 1.13765992e-01 -3.85287553e-02
2.71722794e-01 2.57430315e-01 -4.04172093e-01 7.29385674e-01
2.60361224e-01 1.81886360e-01 -2.89849848e-01 -1.58115074e-01
3.74943674e-01 3.06578368e-01 -5.09197533e-01 -2.97171086e-01
3.30507994e-01 1.77393663e+00 -1.02802992e+00 -4.32438105e-01
-3.82642448e-01 1.04418345e-01 -8.44398201e-01 4.20702785e-01
-8.61828208e-01 2.54255474e-01 -3.94312561e-01 3.92730087e-01
5.46844959e-01 -4.96668667e-01 2.38294259e-01 -1.71254918e-01
5.21345258e-01 9.55841541e-01 -6.43024147e-01 1.89029074e+00
-5.34229219e-01 6.77691549e-02 -1.64777637e-01 -7.19495296e-01
8.96331072e-01 1.85079455e-01 -3.31661999e-01 -8.28513920e-01
-1.25500932e-01 8.47427100e-02 5.82630523e-02 -4.46870625e-01
7.70640790e-01 -9.32648554e-02 -2.78424621e-01 5.19606113e-01
1.03540488e-01 -3.23181987e-01 2.18599826e-01 6.82198405e-01
1.42607141e+00 1.44494534e-01 -6.72787055e-02 8.89571682e-02
4.99600083e-01 -2.98538972e-02 2.87660837e-01 9.67822552e-01
2.33857378e-01 7.23318577e-01 8.74524891e-01 2.75456011e-01
-6.51199758e-01 -1.05681205e+00 6.95588410e-01 1.07186246e+00
-1.33888468e-01 -4.71440047e-01 -1.01310670e+00 -1.17661142e+00
-6.30218387e-02 1.09200621e+00 -6.17152691e-01 -4.28398907e-01
-8.45373690e-01 -4.46034968e-01 8.34439635e-01 5.80025613e-01
7.18909025e-01 -1.10757387e+00 5.85298091e-02 3.30969453e-01
-6.54437542e-01 -1.43290091e+00 -7.56438196e-01 -4.35143381e-01
-5.98865569e-01 -1.03988302e+00 -5.78902066e-01 -9.16037977e-01
6.70853198e-01 1.81511626e-01 1.52067876e+00 1.08911060e-01
2.78107315e-01 6.28601789e-01 -6.59122646e-01 -2.11279914e-01
-5.66111326e-01 6.11972690e-01 -7.16493666e-01 3.61322495e-03
1.56006590e-01 -3.74272019e-01 -6.14822507e-01 1.60389036e-01
-1.09102714e+00 3.62193763e-01 6.79969072e-01 6.27919018e-01
4.32318598e-01 -3.98311049e-01 1.34579492e+00 -1.28784370e+00
1.01922607e+00 -4.99740034e-01 -4.34253633e-01 6.59869671e-01
-6.19045734e-01 4.30395842e-01 9.94436860e-01 3.57202739e-02
-1.03787184e+00 -2.97618449e-01 -5.28402627e-01 -5.79242520e-02
1.35111645e-01 7.37391472e-01 -6.32419527e-01 8.52143690e-02
5.43541670e-01 4.13656801e-01 -2.46151000e-01 -3.39767814e-01
9.27741826e-01 4.69981253e-01 6.92275405e-01 -6.05182290e-01
9.81098354e-01 7.30050728e-02 -6.36895820e-02 -3.66907418e-01
-9.31403339e-01 -2.75360167e-01 -2.70440012e-01 -2.04775520e-02
7.22013354e-01 -7.07060516e-01 -6.25129580e-01 3.24996799e-01
-1.20187044e+00 -5.02542615e-01 -3.80126715e-01 -1.25261033e-02
-4.58909601e-01 4.86520618e-01 -5.45438588e-01 -3.78110975e-01
-6.80153608e-01 -7.35104322e-01 1.16390884e+00 1.18007191e-01
1.03432670e-01 -9.36928511e-01 5.03207333e-02 1.01486492e+00
4.03369963e-01 1.20031685e-01 1.27342725e+00 -7.36617208e-01
-1.00279284e+00 2.63714083e-02 -2.75159955e-01 2.81666487e-01
1.31447405e-01 -5.99937379e-01 -7.22804964e-01 -9.78635773e-02
-2.16161549e-01 -5.14623106e-01 9.30963397e-01 -3.48552763e-01
1.03304768e+00 -7.46840119e-01 1.12542734e-01 4.33337092e-01
1.21265900e+00 -1.97353959e-01 8.03328514e-01 -7.63806775e-02
1.07502401e+00 7.83872962e-01 2.05193192e-01 -2.35885695e-01
1.24590874e+00 6.21215641e-01 6.02488101e-01 1.63321704e-01
-6.08007669e-01 -9.43713427e-01 3.77135962e-01 1.13367558e+00
5.08807182e-01 -6.81278050e-01 -8.12503219e-01 8.18230808e-01
-1.67564750e+00 -5.28804660e-01 -4.30083185e-01 2.01098442e+00
9.00260210e-01 -1.53016001e-01 1.04716811e-02 -1.07719406e-01
5.79438508e-01 1.18571579e-01 -5.61149299e-01 -2.42276385e-01
-1.78126320e-01 4.61314261e-01 3.16493958e-01 5.54197431e-01
-5.56125462e-01 1.15194821e+00 5.69280577e+00 6.92814887e-01
-6.41426980e-01 -5.06884307e-02 3.08938891e-01 1.54418766e-01
-1.03557706e+00 8.48407373e-02 -8.97910774e-01 3.79742265e-01
1.01452005e+00 -3.37227136e-01 4.08978939e-01 3.75199854e-01
-1.36723146e-01 3.24911267e-01 -9.06273723e-01 5.77906907e-01
3.37894589e-01 -1.26960087e+00 6.28137350e-01 -2.00737879e-01
8.42402875e-01 -2.46646971e-01 -8.46622288e-02 8.21185112e-01
6.27927542e-01 -1.02708125e+00 4.06252146e-01 4.76542503e-01
5.97036302e-01 -6.88167393e-01 6.79408133e-01 4.24565226e-01
-1.06047106e+00 2.02593356e-01 -3.02123159e-01 7.35446587e-02
2.97721028e-01 4.11398768e-01 -1.47603631e+00 1.20260811e+00
3.09682816e-01 5.86360842e-02 -8.87195230e-01 8.22360098e-01
-9.07981992e-01 1.05825543e+00 -1.06556214e-01 -2.14563861e-01
3.00749958e-01 -4.26315404e-02 3.23487759e-01 9.38665092e-01
3.71876538e-01 1.37950346e-01 -7.34051540e-02 1.04767025e+00
-6.76745176e-01 2.56755918e-01 -6.06440187e-01 -1.48204351e-02
3.71534437e-01 1.29221857e+00 -2.64658421e-01 -3.32314581e-01
-5.75855672e-01 1.15507925e+00 5.93863845e-01 2.63932914e-01
-7.06292570e-01 -7.32895851e-01 1.89834774e-01 2.26116493e-01
4.97887015e-01 -4.42776904e-02 3.76302823e-02 -1.25060928e+00
3.27667385e-01 -1.13430560e+00 6.64201677e-01 -1.10265088e+00
-1.19219828e+00 5.17414927e-01 -1.67913258e-01 -6.57199383e-01
-6.77074730e-01 -1.24378689e-01 -6.29796326e-01 9.47870374e-01
-1.79494798e+00 -1.45451152e+00 -3.83170485e-01 6.54770434e-01
7.12227374e-02 3.02924007e-01 5.90459883e-01 1.64011315e-01
-2.87725419e-01 9.78250384e-01 -2.54424155e-01 3.34235907e-01
5.72115421e-01 -1.55417562e+00 1.10415292e+00 9.30585146e-01
2.30586693e-01 5.00163674e-01 4.27923262e-01 -7.25959241e-01
-1.60649943e+00 -1.51626897e+00 1.27053857e+00 -6.38325214e-01
5.63678741e-01 -5.12398243e-01 -1.15040481e+00 8.21971118e-01
4.38664585e-01 -2.57486194e-01 4.64292258e-01 -1.20444238e-01
-3.92975897e-01 2.52721319e-03 -9.70165074e-01 7.59014904e-01
1.19540119e+00 -4.97890979e-01 -6.85711503e-01 2.74891436e-01
1.44429600e+00 -5.43865204e-01 -7.18888700e-01 5.60471475e-01
9.85055566e-02 -7.13742614e-01 7.29933858e-01 -8.41262937e-01
4.46972311e-01 -4.35205102e-01 7.35649168e-02 -1.45095789e+00
-1.10885262e-01 -8.22542071e-01 -4.01209921e-01 1.40051162e+00
6.15250528e-01 -7.85610795e-01 1.07412183e+00 4.98537838e-01
-4.65420663e-01 -8.51730943e-01 -7.85661578e-01 -6.87931478e-01
1.97004870e-01 -2.23763585e-01 1.15420294e+00 7.19755948e-01
-1.93439737e-01 8.83595526e-01 -2.07500592e-01 2.38327205e-01
4.12428766e-01 3.98179799e-01 1.02316689e+00 -8.17787111e-01
-4.89399940e-01 -1.92392558e-01 -1.34465238e-02 -1.57667708e+00
2.73751527e-01 -1.29287052e+00 4.29565199e-02 -2.21759510e+00
-1.03364058e-01 -4.06558588e-02 1.12683617e-01 4.30052876e-01
-7.98453093e-01 2.12193608e-01 1.64374650e-01 -3.25717628e-01
-6.93056583e-01 9.12227690e-01 1.65706921e+00 -1.05822265e-01
6.99369758e-02 -1.49067879e-01 -1.36504602e+00 2.34375030e-01
9.24708128e-01 -2.25388914e-01 -8.92195940e-01 -8.11983585e-01
6.62419498e-01 2.70761818e-01 4.98143584e-01 -6.22347057e-01
3.98001760e-01 8.51492807e-02 -2.96793967e-01 -4.24634457e-01
1.29288644e-01 -2.05133095e-01 -2.17227906e-01 1.59088969e-01
-4.22592163e-01 9.68979374e-02 -2.54966002e-02 5.99525630e-01
-2.91354090e-01 -2.34487131e-01 3.16903412e-01 -2.26860255e-01
-5.05489349e-01 2.91762322e-01 7.55112022e-02 6.93371713e-01
5.30421615e-01 4.70738783e-02 -8.46323192e-01 -1.02698243e+00
-3.57523888e-01 6.55889034e-01 1.40711531e-01 4.60201770e-01
6.84531450e-01 -1.40047574e+00 -1.06076992e+00 -9.63485800e-04
2.79468179e-01 3.22019607e-01 2.91733712e-01 5.44595242e-01
-4.71477449e-01 6.75247788e-01 3.03533614e-01 2.07770970e-02
-9.16794956e-01 4.17434961e-01 4.24041063e-01 -6.81141376e-01
-3.72440845e-01 9.04630125e-01 4.04259920e-01 -1.03753436e+00
-5.48362546e-02 -5.08859813e-01 -2.52652317e-01 -2.54643708e-01
1.92688584e-01 3.03637356e-01 3.58525634e-01 -4.04225469e-01
7.20722079e-02 2.61801958e-01 -6.43446892e-02 -1.36287101e-02
8.87391806e-01 -1.30629931e-02 -1.61075234e-01 -1.38892427e-01
1.03175700e+00 7.89858028e-02 -7.16984272e-01 -4.16807592e-01
-3.93843465e-02 -8.74529108e-02 -3.77006173e-01 -1.09977782e+00
-9.57860887e-01 7.58963764e-01 -2.51035333e-01 2.92952389e-01
1.15924442e+00 3.19968045e-01 1.32278645e+00 6.97197974e-01
3.29145491e-01 -6.51726127e-01 2.24541709e-01 7.64645338e-01
1.07989669e+00 -8.50434065e-01 -6.47290587e-01 -6.49250150e-01
-5.26395082e-01 6.40836179e-01 9.54037070e-01 1.12045012e-01
7.42054507e-02 -3.93465906e-01 -5.65943941e-02 -2.14365035e-01
-9.08071220e-01 -5.20505130e-01 4.45684880e-01 6.34699047e-01
2.78941959e-01 -6.76273033e-02 -1.24476455e-01 7.36570299e-01
-8.67631793e-01 5.50024305e-03 6.47517562e-01 6.46764696e-01
-3.20349842e-01 -1.36992645e+00 -9.13044810e-02 6.26702666e-01
-5.62817514e-01 -5.37050307e-01 -7.78821051e-01 8.21389794e-01
-3.11073720e-01 1.13598299e+00 -3.07495892e-01 -3.72099012e-01
7.03195453e-01 4.11535025e-01 6.93620503e-01 -8.65384102e-01
-8.03522348e-01 -6.62911296e-01 4.12796885e-01 -2.64369547e-01
-1.57794764e-03 -2.30538622e-01 -1.27412760e+00 -2.56984442e-01
-3.17672551e-01 5.15666902e-01 2.51223534e-01 7.35447288e-01
6.85830414e-01 6.28648639e-01 3.01665902e-01 2.68992186e-01
-7.33667314e-01 -1.19237375e+00 -4.12348449e-01 3.93690377e-01
1.05159409e-01 3.71129997e-02 -2.38569260e-01 -6.88472912e-02] | [11.25779914855957, 8.120773315429688] |
cf681bab-9990-428d-b5b5-28604237038c | generalized-inter-class-loss-for-gait | 2210.06779 | null | https://arxiv.org/abs/2210.06779v1 | https://arxiv.org/pdf/2210.06779v1.pdf | Generalized Inter-class Loss for Gait Recognition | Gait recognition is a unique biometric technique that can be performed at a long distance non-cooperatively and has broad applications in public safety and intelligent traffic systems. Previous gait works focus more on minimizing the intra-class variance while ignoring the significance in constraining inter-class variance. To this end, we propose a generalized inter-class loss which resolves the inter-class variance from both sample-level feature distribution and class-level feature distribution. Instead of equal penalty strength on pair scores, the proposed loss optimizes sample-level inter-class feature distribution by dynamically adjusting the pairwise weight. Further, in class-level distribution, generalized inter-class loss adds a constraint on the uniformity of inter-class feature distribution, which forces the feature representations to approximate a hypersphere and keep maximal inter-class variance. In addition, the proposed method automatically adjusts the margin between classes which enables the inter-class feature distribution to be more flexible. The proposed method can be generalized to different gait recognition networks and achieves significant improvements. We conduct a series of experiments on CASIA-B and OUMVLP, and the experimental results show that the proposed loss can significantly improve the performance and achieves the state-of-the-art performances. | ['Liang Wang', 'Yan Huang', 'Hongyuan Yu', 'Weichen Yu'] | 2022-10-13 | null | null | null | null | ['gait-recognition'] | ['computer-vision'] | [-7.28730187e-02 -2.64397532e-01 -3.12083423e-01 -5.52034259e-01
-1.92189604e-01 -8.81425440e-02 8.22754949e-02 -1.89954892e-01
-4.28910643e-01 6.61139071e-01 -1.15061603e-01 3.17357123e-01
-5.39530694e-01 -9.29145277e-01 -1.69783548e-01 -1.09366620e+00
-2.45400429e-01 2.83375084e-01 4.31716114e-01 -1.42309204e-01
2.00209677e-01 4.09539521e-01 -1.40658307e+00 -9.58076864e-02
1.14245558e+00 1.09137249e+00 -1.49753064e-01 6.23515286e-02
2.95667574e-02 1.53886050e-01 -7.66754091e-01 -3.31781656e-01
2.54449219e-01 -4.03794199e-01 -2.52093136e-01 1.68744728e-01
1.85936809e-01 5.94273731e-02 -2.94002026e-01 1.30869710e+00
8.33543181e-01 1.95637375e-01 7.52523184e-01 -1.70140719e+00
-6.20463789e-01 2.75643259e-01 -1.17894506e+00 -4.37518544e-02
1.66506946e-01 1.17639154e-01 8.68394911e-01 -6.20459497e-01
2.58390550e-02 1.35043228e+00 6.91735566e-01 4.26844329e-01
-1.00999415e+00 -1.04490888e+00 3.45985830e-01 4.76902783e-01
-1.83328724e+00 -3.22058573e-02 1.06926107e+00 -1.23180375e-01
4.45928127e-01 2.93313235e-01 7.20923483e-01 6.30318880e-01
1.48560494e-01 7.47237325e-01 8.13685656e-01 -1.69879809e-01
1.62229612e-01 -1.89042896e-01 2.75491089e-01 6.74792945e-01
6.40990555e-01 -1.34458676e-01 -4.06685442e-01 -2.27914616e-01
7.48940408e-01 2.65817773e-02 -4.44835454e-01 -7.28472829e-01
-1.01878536e+00 6.11291289e-01 3.65027189e-01 2.41137832e-01
-4.21280824e-02 -5.59831709e-02 3.64014387e-01 2.32656356e-02
1.32543102e-01 -4.11569357e-01 -8.98932666e-02 -2.91535765e-01
-8.71941507e-01 4.53462191e-02 5.52942514e-01 7.70751357e-01
6.51929021e-01 -8.53125453e-02 -3.53075564e-01 1.10347176e+00
4.32240754e-01 7.33294070e-01 4.97494370e-01 -5.67923546e-01
6.73810601e-01 6.54107869e-01 -1.20208755e-01 -1.59110677e+00
-3.71491462e-01 -5.13993144e-01 -1.21335137e+00 -9.41810291e-03
5.76753795e-01 -1.32057369e-01 -6.34534061e-01 1.94891477e+00
3.00888479e-01 2.71718770e-01 -2.90151745e-01 9.18963075e-01
5.04038274e-01 2.63960212e-01 -1.64713934e-01 -4.54505026e-01
1.19900250e+00 -5.68965018e-01 -6.79737091e-01 -3.75722013e-02
2.67500341e-01 -4.77828175e-01 9.81826723e-01 1.55714825e-01
-6.47548676e-01 -4.91319150e-01 -1.10280311e+00 5.88770270e-01
-1.40404359e-01 3.20598334e-01 2.05901429e-01 9.21608865e-01
-5.86318672e-01 4.56733912e-01 -8.21536243e-01 -3.80974799e-01
4.92556006e-01 6.20600343e-01 -3.20395648e-01 4.24867123e-03
-1.24204588e+00 4.37211990e-01 3.47077340e-01 3.57485175e-01
2.07131043e-01 -4.10305172e-01 -7.36407638e-01 1.37222737e-01
1.68275163e-01 -2.75290012e-01 3.99307549e-01 -5.15878618e-01
-1.40641451e+00 5.49274504e-01 -1.37147859e-01 -5.65626398e-02
5.93045533e-01 3.28053594e-01 -6.59576714e-01 -1.51729167e-01
1.88716441e-01 3.93272191e-01 5.90568185e-01 -8.31428349e-01
-5.13336241e-01 -5.22681892e-01 -3.37464184e-01 2.79443949e-01
-7.26555169e-01 -2.96060979e-01 -5.82559288e-01 -6.80486143e-01
3.26712966e-01 -9.06996906e-01 4.88210879e-02 7.97772557e-02
-3.21684808e-01 -3.39721680e-01 1.08470404e+00 -1.54499695e-01
1.34258902e+00 -2.18295240e+00 -9.79470387e-02 6.73695385e-01
6.42626733e-03 2.66442209e-01 -1.91564083e-01 1.03576578e-01
7.22699016e-02 -1.21225946e-01 -4.02582854e-01 8.85508582e-03
3.54061984e-02 2.52970576e-01 2.56695241e-01 6.70538962e-01
8.19836929e-02 3.53824705e-01 -7.51608968e-01 -5.71073532e-01
2.13167116e-01 4.56462801e-01 -3.89792919e-01 -1.78319067e-01
7.14239538e-01 1.97035357e-01 -6.30621016e-01 6.47728801e-01
1.20859885e+00 2.23956779e-02 1.81781024e-01 -4.32094336e-01
3.06867719e-01 -3.65357280e-01 -1.66534853e+00 1.36992252e+00
1.15413554e-01 3.35691035e-01 -6.83392286e-02 -1.34821808e+00
1.27832842e+00 -1.84921339e-01 7.70522296e-01 -5.69725513e-01
1.95501104e-01 1.48603946e-01 3.08529854e-01 -3.33609790e-01
2.87064284e-01 9.56655368e-02 -1.94982320e-01 1.73905134e-01
-1.84993416e-01 3.99312109e-01 2.52175599e-01 -1.69568077e-01
6.22408390e-01 -1.11981342e-02 2.09111840e-01 -5.06057262e-01
9.93710935e-01 -7.99266875e-01 1.22143090e+00 5.16734421e-01
-8.06660831e-01 5.51258802e-01 3.80267084e-01 -1.93621442e-01
-3.95563513e-01 -1.10110819e+00 -3.11465889e-01 5.30513823e-01
7.22735941e-01 -1.21001832e-01 -7.15433776e-01 -8.96667480e-01
3.18137109e-01 9.35992599e-02 -6.31000996e-01 -4.68587399e-01
-5.74310899e-01 -1.11053753e+00 8.12335849e-01 5.86013913e-01
1.02594995e+00 -6.29736662e-01 -8.12839195e-02 1.18901506e-02
-4.10832077e-01 -8.19037259e-01 -9.24575448e-01 -3.70620608e-01
-4.81369048e-01 -1.13046837e+00 -8.17866325e-01 -8.00242245e-01
6.82307959e-01 2.45351896e-01 5.99318922e-01 8.16427916e-02
-2.96340615e-01 2.47663289e-01 -1.76321223e-01 -1.24303110e-01
4.73201573e-01 1.07028857e-01 4.11693960e-01 4.61241573e-01
7.28897631e-01 -7.74448633e-01 -7.17206717e-01 8.43032897e-01
-4.27007645e-01 -4.74188954e-01 3.37033272e-01 1.04648530e+00
4.53627884e-01 3.50072414e-01 8.27211261e-01 -3.51261231e-03
6.56146049e-01 -4.37613934e-01 -2.06516460e-01 3.89674187e-01
-6.46695912e-01 -8.41900855e-02 4.00051683e-01 -5.46918571e-01
-8.05976689e-01 -2.63065487e-01 2.15781957e-01 -1.83674559e-01
1.68934599e-01 3.06833237e-01 -7.20360637e-01 -3.48048508e-01
1.67123541e-01 3.66951942e-01 2.28997231e-01 -1.14460938e-01
2.23621912e-02 7.90822566e-01 4.65536773e-01 -7.11985767e-01
9.15325284e-01 4.58542436e-01 3.94148618e-01 -8.46098423e-01
-2.45858952e-01 -3.71515185e-01 -5.53691089e-01 -4.45816249e-01
5.47187746e-01 -6.51489377e-01 -1.06908345e+00 8.58370245e-01
-6.07859671e-01 2.31978297e-01 -1.55084908e-01 6.78133428e-01
-2.48595685e-01 9.55675423e-01 -3.87036502e-01 -7.99537897e-01
-3.26909065e-01 -1.14887989e+00 8.28196585e-01 6.76716506e-01
-8.82407243e-04 -8.96122158e-01 -1.86172515e-01 1.48279950e-01
1.83055907e-01 1.41348809e-01 3.37840080e-01 -4.97160167e-01
-2.16731459e-01 -3.39555979e-01 -3.35467517e-01 2.82634169e-01
6.79946721e-01 4.31107879e-02 -5.54859340e-01 -5.17299175e-01
-4.59941983e-01 -4.94504496e-02 7.60601461e-01 4.36166465e-01
1.24421179e+00 2.48023402e-02 -4.09783721e-01 5.23123920e-01
9.29117978e-01 3.52123886e-01 8.61088693e-01 2.29981109e-01
7.07771361e-01 6.74037218e-01 8.69078040e-01 4.68227535e-01
5.41621447e-01 9.43756104e-01 1.34743080e-01 1.35420799e-01
1.37964368e-01 -5.17757749e-03 2.38918617e-01 7.91942954e-01
-1.26123562e-01 -1.68808982e-01 -6.27408087e-01 3.85063022e-01
-2.19259167e+00 -1.23678517e+00 8.69564638e-02 2.38994265e+00
6.28879845e-01 8.42411965e-02 4.40719843e-01 3.41026455e-01
1.25792181e+00 1.93965986e-01 -5.67255497e-01 -5.57001792e-02
-2.66539484e-01 -1.27563804e-01 4.87868369e-01 3.49394232e-01
-1.10113966e+00 5.12054682e-01 5.17688131e+00 1.32022285e+00
-1.06828392e+00 -3.73295844e-01 4.12447214e-01 -1.88781738e-01
2.60379285e-01 -4.58932519e-01 -9.18844461e-01 9.31864798e-01
1.93977818e-01 -3.10266107e-01 1.79512858e-01 6.78249776e-01
2.93159753e-01 1.20393366e-01 -7.63249874e-01 1.26156354e+00
1.04965664e-01 -8.52310419e-01 -2.91634537e-02 3.69910181e-01
2.26465836e-01 -3.70724440e-01 3.62277180e-02 1.29525855e-01
-1.44232273e-01 -5.71654499e-01 3.58415395e-01 4.16716278e-01
7.21197307e-01 -1.24507368e+00 7.95675933e-01 2.32767433e-01
-1.80383062e+00 -1.08436039e-02 -5.42707205e-01 -7.60606974e-02
-9.21401829e-02 6.68955922e-01 -9.70785022e-02 7.82444715e-01
7.77134240e-01 9.06731308e-01 -5.04866958e-01 1.24964380e+00
-1.65629099e-05 3.05193245e-01 -4.52143043e-01 -3.22834760e-01
-6.13206178e-02 -5.57376981e-01 6.90901041e-01 1.13841236e+00
3.68682981e-01 -1.08806215e-01 5.41485667e-01 5.22054017e-01
1.14200301e-01 1.95235193e-01 -1.94073051e-01 3.89564842e-01
9.07908738e-01 1.11472988e+00 -7.32260942e-01 -3.06582730e-02
-2.06159458e-01 8.37128520e-01 5.46423607e-02 2.24827096e-01
-9.32798803e-01 -1.00877178e+00 9.88945842e-01 -2.89208051e-02
3.33275616e-01 1.06443195e-02 -3.42752039e-01 -1.38247371e+00
4.82091069e-01 -4.79079545e-01 5.31879961e-01 -2.38065630e-01
-1.55491400e+00 2.38722682e-01 -1.37550503e-01 -1.67076492e+00
-4.54142988e-02 -4.02933627e-01 -9.34421420e-01 8.13409984e-01
-1.20567107e+00 -1.14281213e+00 -4.07269448e-01 7.58058548e-01
-4.38989699e-02 -2.58226424e-01 5.00995219e-01 6.09044850e-01
-9.02177691e-01 1.31109393e+00 1.21487215e-01 4.82526720e-01
8.92610371e-01 -7.89706051e-01 -4.67302054e-02 8.70744228e-01
-4.29705441e-01 6.65372849e-01 4.24756825e-01 -6.54907882e-01
-1.00238574e+00 -1.07474744e+00 5.70975065e-01 1.50861993e-01
4.31000233e-01 -2.38685161e-01 -7.99185216e-01 2.27952421e-01
-2.17484742e-01 3.57729882e-01 8.45114410e-01 1.04455784e-01
-2.96198905e-01 -6.23570442e-01 -1.49836433e+00 5.47872603e-01
1.29013145e+00 -2.25678816e-01 -3.37007701e-01 -3.07565387e-02
8.21205229e-02 -1.95852995e-01 -9.36583877e-01 5.75409055e-01
9.72047269e-01 -7.57377863e-01 9.33321655e-01 -2.33378023e-01
-2.93767363e-01 -9.17720437e-01 -1.92644358e-01 -1.19376051e+00
-5.12407959e-01 -3.74045432e-01 1.02560602e-01 1.56983328e+00
-1.43077113e-02 -9.97481406e-01 9.89710569e-01 6.34127438e-01
1.45639449e-01 -9.01266158e-01 -1.26159048e+00 -1.06488848e+00
-5.75596690e-02 -1.68825969e-01 7.72499859e-01 9.67440069e-01
2.49967247e-01 -1.08022436e-01 -5.73987484e-01 1.58690080e-01
1.12885213e+00 8.89469460e-02 7.13661253e-01 -1.42045701e+00
-3.03373575e-01 -5.48731029e-01 -1.08583248e+00 -9.33085442e-01
1.58800974e-01 -7.10107148e-01 -2.10500695e-02 -1.20564997e+00
2.88292497e-01 -5.44486403e-01 -5.78392804e-01 5.64159095e-01
-2.72424728e-01 4.17899817e-01 8.64059478e-02 2.73069561e-01
-5.06048262e-01 7.87523329e-01 1.29055750e+00 -2.04609722e-01
-3.24683547e-01 -4.01865579e-02 -6.56614244e-01 6.77466631e-01
7.64379144e-01 -1.75469637e-01 -4.52342004e-01 1.03212059e-01
-4.31513906e-01 -2.58240223e-01 8.23062360e-02 -1.16705239e+00
3.66134763e-01 -3.44184279e-01 5.14717340e-01 -6.67293549e-01
2.81538755e-01 -7.39061475e-01 -1.62058115e-01 6.03637695e-01
4.96531390e-02 -3.72502744e-01 -2.43626237e-02 8.19768727e-01
-1.86595693e-01 3.52932066e-01 9.27340567e-01 4.40577000e-01
-4.42035019e-01 5.41984618e-01 -1.97550878e-01 2.25711260e-02
1.09529305e+00 -8.46128225e-01 -3.13415170e-01 -2.78480619e-01
-5.47886968e-01 6.76235259e-01 3.13072473e-01 5.97551227e-01
5.32992184e-01 -1.93374455e+00 -6.44333065e-01 4.91644382e-01
2.64294118e-01 -4.39597070e-01 4.71516132e-01 9.36866641e-01
1.17287040e-02 4.61313874e-02 -3.66296232e-01 -9.34335828e-01
-1.47931230e+00 -8.72572884e-02 6.17292166e-01 -1.52135223e-01
-2.43293256e-01 7.75965393e-01 7.12590963e-02 -5.64534903e-01
2.16202527e-01 7.07038641e-02 -3.35617691e-01 8.01363885e-02
4.86405104e-01 5.92229664e-01 -2.18969867e-01 -9.72739637e-01
-7.78638601e-01 1.16862059e+00 -6.53576255e-02 2.63504773e-01
1.00740850e+00 -2.92212278e-01 -1.71400473e-01 8.24371874e-02
1.18520105e+00 7.53074279e-03 -1.09675121e+00 -4.19452637e-02
-3.56502771e-01 -7.44494557e-01 -2.31149271e-01 -4.20198202e-01
-1.39982557e+00 7.05777884e-01 8.48040104e-01 -2.43939348e-02
1.25202715e+00 -5.87492287e-01 8.11986864e-01 2.97648877e-01
6.36218607e-01 -1.39671183e+00 1.93980455e-01 3.45912099e-01
5.00555694e-01 -9.98482943e-01 3.43137980e-02 -6.89840794e-01
-4.75952029e-01 1.12505853e+00 8.59804451e-01 -3.15045208e-01
8.92690718e-01 1.19965144e-01 -3.18840504e-01 1.96922749e-01
-9.50388685e-02 -3.31193618e-02 5.14676571e-01 9.41882491e-01
4.28063691e-01 2.80849814e-01 -8.15405548e-01 8.62066388e-01
-2.84559876e-02 4.32410836e-02 2.67914325e-01 7.69952476e-01
-4.09795552e-01 -1.28181946e+00 -4.33206379e-01 4.71777827e-01
-2.02355132e-01 2.63032466e-01 -1.89299464e-01 7.03778982e-01
3.96765172e-01 1.08174157e+00 1.91031009e-01 -7.83797443e-01
3.21218103e-01 -2.12374523e-01 5.30469239e-01 -1.02141954e-01
-3.96874174e-02 1.05629779e-01 -8.43627751e-02 -4.78915393e-01
-5.04351974e-01 -8.60878706e-01 -1.41298866e+00 -6.18750513e-01
-5.60031712e-01 3.69845450e-01 8.56672004e-02 8.64528418e-01
4.77799028e-01 2.50404119e-01 8.34719121e-01 -6.21619403e-01
-5.28226197e-01 -8.31648111e-01 -9.36118960e-01 2.62187481e-01
-6.59662932e-02 -1.03422427e+00 -3.85689408e-01 -3.76729935e-01] | [14.295713424682617, 1.3435463905334473] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.