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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
eaae0e60-d4d5-4d02-a84d-4a445d4b08f7 | melanoma-diagnosis-with-spatio-temporal | 2006.10950 | null | https://arxiv.org/abs/2006.10950v1 | https://arxiv.org/pdf/2006.10950v1.pdf | Melanoma Diagnosis with Spatio-Temporal Feature Learning on Sequential Dermoscopic Images | Existing studies for automated melanoma diagnosis are based on single-time point images of lesions. However, melanocytic lesions de facto are progressively evolving and, moreover, benign lesions can progress into malignant melanoma. Ignoring cross-time morphological changes of lesions thus may lead to misdiagnosis in borderline cases. Based on the fact that dermatologists diagnose ambiguous skin lesions by evaluating the dermoscopic changes over time via follow-up examination, in this study, we propose an automated framework for melanoma diagnosis using sequential dermoscopic images. To capture the spatio-temporal characterization of dermoscopic evolution, we construct our model in a two-stream network architecture which capable of simultaneously learning appearance representations of individual lesions while performing temporal reasoning on both raw pixels difference and abstract features difference. We collect 184 cases of serial dermoscopic image data, which consists of histologically confirmed 92 benign lesions and 92 melanoma lesions, to evaluate the effectiveness of the proposed method. Our model achieved AUC of 74.34%, which is ~8% higher than that of only using single images and ~6% higher than the widely used sequence learning model based on LSTM. | ['ZongYuan Ge', 'Xiaojun Chang', 'Jennifer Nguyen', 'Zhen Yu', 'Lei Zhang', 'John Kelly', 'Victoria Mar', 'Catriona Mclean'] | 2020-06-19 | null | null | null | null | ['melanoma-diagnosis'] | ['computer-vision'] | [ 8.61583889e-01 -3.16651583e-01 -2.05586955e-01 -7.51955733e-02
-4.59397644e-01 -5.09203970e-01 5.10284305e-01 2.25746363e-01
-6.53891265e-01 6.47820592e-01 -3.94892573e-01 -4.32642132e-01
-2.03822792e-01 -7.79850662e-01 -2.83657253e-01 -9.48875010e-01
-1.17922708e-01 -4.05036137e-02 1.94480270e-01 1.30966261e-01
1.18239336e-01 5.13360620e-01 -1.34972334e+00 6.63309038e-01
9.93030846e-01 9.66849327e-01 1.14674583e-01 1.21791720e+00
-1.81566700e-01 8.90986860e-01 -4.81714636e-01 -5.49723923e-01
1.22198202e-01 -3.32904398e-01 -7.83115804e-01 5.18703878e-01
6.52878106e-01 -5.97666800e-01 -3.92190963e-01 1.28247702e+00
1.65543929e-01 -2.48330012e-01 7.95867383e-01 -8.36754680e-01
-7.16700017e-01 -1.41667938e-02 -9.35813546e-01 4.75166351e-01
6.25005960e-02 5.00377953e-01 5.45828581e-01 -2.47234806e-01
9.51680481e-01 7.18382716e-01 7.56076515e-01 7.58027434e-01
-8.39818001e-01 -1.08954415e-01 2.82608598e-01 4.92299914e-01
-1.06523073e+00 -2.18048811e-01 2.99409986e-01 -5.47540605e-01
7.34432518e-01 5.40872931e-01 9.39004600e-01 1.32282984e+00
7.35641360e-01 7.96465278e-01 1.51308239e+00 -3.36305887e-01
-1.40937537e-01 9.26216319e-03 8.16096514e-02 1.01456439e+00
4.50581089e-02 2.96574980e-01 -2.41755188e-01 -5.50949909e-02
1.01174045e+00 4.47613060e-01 4.08812389e-02 3.10307264e-01
-1.07612157e+00 3.36622328e-01 5.02145827e-01 5.62491380e-02
-6.83341861e-01 1.87178865e-01 4.32374090e-01 4.43643332e-01
5.09865046e-01 -8.14118907e-02 5.17465323e-02 2.42528822e-02
-8.04104209e-01 -3.90816003e-01 1.44641355e-01 -4.55027334e-02
1.72606170e-01 -1.34504735e-01 -1.80162758e-01 8.65825415e-01
-6.84225038e-02 3.72289151e-01 8.03316712e-01 -3.39851469e-01
-5.93239367e-02 7.37585545e-01 -3.14424857e-02 -6.58560872e-01
-3.73228520e-01 -1.85403720e-01 -1.15904224e+00 2.73931593e-01
3.73145223e-01 -2.50135064e-01 -1.56841850e+00 1.22558987e+00
3.55858356e-01 3.00266802e-01 6.90553263e-02 7.72514045e-01
4.87751335e-01 3.21362883e-01 3.12892854e-01 -5.60186148e-01
1.22654152e+00 -9.13931906e-01 -7.64036655e-01 3.34735841e-01
3.46288353e-01 -4.91781414e-01 6.84764564e-01 5.35079658e-01
-9.47175086e-01 -3.39009792e-01 -8.41048241e-01 3.22651297e-01
-2.92109430e-01 4.53484088e-01 7.77354062e-01 4.05474842e-01
-1.21154594e+00 5.12380600e-01 -1.17838311e+00 -8.19046497e-01
3.66864264e-01 2.40898579e-01 -3.35822284e-01 -7.01830834e-02
-1.07172692e+00 6.85300708e-01 1.01426013e-01 4.83629316e-01
-1.00747323e+00 -6.45012200e-01 -4.25365746e-01 -6.38253093e-01
2.74963111e-01 -7.54511893e-01 1.10149515e+00 -1.34891725e+00
-1.36561418e+00 1.02232158e+00 -4.52943534e-01 -6.30584061e-01
9.05318379e-01 2.17415273e-01 -9.73165274e-01 4.09870416e-01
-3.73956233e-01 6.26325846e-01 1.07211244e+00 -8.32084715e-01
-1.15106976e+00 -3.90984952e-01 8.04124251e-02 3.53750318e-01
-5.44138312e-01 -2.63856798e-01 -3.02830309e-01 -3.01407814e-01
-3.08384657e-01 -9.52285826e-01 -4.90109026e-01 5.31933129e-01
-4.88179564e-01 6.56991899e-02 5.53016365e-01 -1.15119302e+00
1.19255149e+00 -2.07150126e+00 -3.40102345e-01 1.17273591e-02
5.29821575e-01 5.77918708e-01 -1.65693387e-01 2.48446301e-01
-3.66134681e-02 2.12483197e-01 -1.96642429e-01 5.83011173e-02
-6.89058721e-01 -1.61053225e-01 -9.66440979e-03 5.40717661e-01
5.89804292e-01 1.09726608e+00 -9.80841398e-01 -6.00940347e-01
3.10437292e-01 2.05642700e-01 4.32814300e-01 -2.48884380e-01
-2.29582250e-01 -4.23197672e-02 -2.15737626e-01 1.02304924e+00
5.93029320e-01 -5.69660246e-01 3.09898198e-01 -3.24222520e-02
8.60919207e-02 -4.37195152e-01 -5.40677965e-01 1.51977777e+00
-2.92212039e-01 9.10354435e-01 -2.56339371e-01 -3.76576155e-01
4.47015107e-01 4.15649861e-01 6.28735662e-01 -6.78896129e-01
-1.24480486e-01 -3.35276723e-02 4.05135900e-01 -1.02820539e+00
1.55966133e-01 -2.25680918e-01 5.94181836e-01 4.30511266e-01
-4.69674736e-01 2.91804492e-01 3.29787523e-01 -4.74500954e-02
1.47651255e+00 -8.88397694e-02 4.13092792e-01 3.83222133e-01
5.67934513e-01 2.70530403e-01 1.96973249e-01 5.57625413e-01
-5.88414729e-01 1.65065661e-01 3.33858758e-01 -6.15548730e-01
-9.04770672e-01 -1.47544491e+00 -2.62227982e-01 3.93209845e-01
3.03564191e-01 3.63192439e-01 -4.25119728e-01 -8.12414467e-01
9.06494632e-02 1.37027264e-01 -1.20551920e+00 2.80369446e-03
-1.44484103e-01 -1.25020361e+00 5.76239288e-01 3.84363472e-01
6.06716573e-01 -8.86205196e-01 -4.98588592e-01 1.97994754e-01
1.41407087e-01 -7.70632923e-01 -2.75223285e-01 -3.27767879e-01
-8.83923352e-01 -1.44535244e+00 -8.92671049e-01 -6.50486946e-01
1.04842436e+00 3.95743549e-01 3.81357819e-01 1.71205238e-01
-1.27481747e+00 3.25611472e-01 -3.05133592e-02 -3.66648018e-01
-5.50243080e-01 -2.84315526e-01 -1.52356431e-01 4.08924937e-01
1.98349640e-01 -1.98802873e-01 -7.53052592e-01 -1.39894590e-01
-1.20942700e+00 3.40615094e-01 1.09788549e+00 1.00233769e+00
7.36064196e-01 2.22616374e-01 4.27656293e-01 -1.03110933e+00
8.52975130e-01 -5.29354632e-01 -2.55393595e-01 8.50241721e-01
-5.74382663e-01 -3.44680578e-01 4.96153563e-01 -7.46208608e-01
-1.33196235e+00 7.80010223e-02 2.12190494e-01 -5.77777922e-01
-1.56167746e-01 5.00917852e-01 9.65963423e-01 -2.26700455e-01
7.16834486e-01 6.58365488e-01 4.81945634e-01 1.49991333e-01
-8.64882860e-03 6.83895528e-01 6.39865577e-01 2.74574429e-01
4.14832562e-01 9.75648463e-01 1.89962670e-01 -8.94074023e-01
-5.96890986e-01 -5.07084131e-01 -6.29192650e-01 -6.30196154e-01
6.57673180e-01 -6.60803378e-01 -6.75864041e-01 9.87155378e-01
-7.59137034e-01 -2.98130155e-01 -1.53113499e-01 2.34702691e-01
-8.13693181e-02 6.92464173e-01 -1.10907161e+00 -1.05768824e+00
-5.17739415e-01 -7.76952326e-01 8.73861194e-01 3.25895309e-01
-3.41801420e-02 -1.46049750e+00 1.07946388e-01 6.67208135e-02
2.31623575e-01 5.98929107e-01 1.01118779e+00 -4.13550995e-02
-2.83950746e-01 -4.99119461e-01 -3.66491914e-01 2.37833753e-01
7.51582086e-01 6.96857750e-01 -8.90232146e-01 -3.26228797e-01
-1.77785560e-01 -1.80123031e-01 1.04595280e+00 6.35014534e-01
1.11702669e+00 -1.89151037e-02 -5.48876822e-01 4.63094085e-01
1.71171904e+00 5.48752964e-01 5.17789841e-01 4.24931981e-02
5.16707957e-01 6.84036314e-01 5.86897254e-01 2.23806798e-01
1.72306597e-01 9.79944542e-02 5.75415134e-01 -4.38236654e-01
-4.12804365e-01 -3.20589580e-02 3.33210051e-01 6.52972281e-01
-2.66170681e-01 -3.65830123e-01 -8.77128065e-01 8.47605526e-01
-1.61058569e+00 -9.93579805e-01 -2.95659006e-01 2.09532237e+00
8.10760200e-01 -8.84235948e-02 1.05921701e-01 -2.26064965e-01
1.15159988e+00 -2.19757147e-02 -1.12522423e+00 -4.01853293e-01
-1.01886608e-01 -3.82088386e-02 6.26025856e-01 3.54819536e-01
-9.69429672e-01 6.48091018e-01 6.82172966e+00 7.90970445e-01
-1.91109669e+00 -1.72360206e-03 7.51335800e-01 -4.37672824e-01
-2.06076071e-01 -5.09431243e-01 -3.61797363e-01 5.58084309e-01
7.79826999e-01 -2.65591443e-01 2.77145445e-01 3.96809494e-03
2.15597257e-01 -3.90713066e-01 -7.82652438e-01 8.89698744e-01
2.57669967e-02 -1.37746441e+00 1.61580577e-01 2.08605289e-01
9.79066849e-01 -9.24669057e-02 6.79745972e-01 -2.75411725e-01
4.12442505e-01 -1.15695858e+00 -1.34559199e-01 1.10533893e+00
1.27406597e+00 -3.12246025e-01 6.69964314e-01 2.36860931e-01
-8.42097044e-01 -6.87935054e-02 -9.00881812e-02 2.29349077e-01
-3.24593857e-02 4.77187067e-01 -1.53750062e+00 6.75948799e-01
2.19977483e-01 9.70253706e-01 -8.27893496e-01 1.13091612e+00
-3.58632184e-03 5.63344717e-01 -1.62295345e-02 -2.29216382e-01
5.36405206e-01 -2.11963672e-02 2.58788228e-01 9.15288150e-01
3.33338678e-01 -2.66042382e-01 -2.47750938e-01 6.13165498e-01
3.42512310e-01 -7.40089566e-02 -5.28523803e-01 -3.62268060e-01
1.92683086e-01 1.47066021e+00 -6.58137679e-01 -2.73325443e-01
-2.65077889e-01 1.30455160e+00 1.19986266e-01 5.35712123e-01
-8.65061998e-01 -1.20180212e-01 5.04547417e-01 -1.85377542e-02
-5.46752326e-02 1.76316500e-01 -7.27623254e-02 -8.96988273e-01
1.56907111e-01 -5.26462555e-01 5.36651969e-01 -5.57323515e-01
-1.49456227e+00 6.56770706e-01 -6.55157089e-01 -1.27416241e+00
-3.25738877e-01 -7.63504326e-01 -8.93006980e-01 7.02260315e-01
-1.74164224e+00 -1.41946447e+00 -6.77641451e-01 6.64816618e-01
6.00453138e-01 -3.48222093e-03 8.00829828e-01 -1.80590644e-01
-7.80454755e-01 4.54071105e-01 2.22469792e-01 -8.61356184e-02
8.26326072e-01 -1.22400522e+00 3.42413396e-01 9.01192784e-01
-1.94303080e-01 3.24424446e-01 6.90112114e-02 -9.71794486e-01
-1.14253080e+00 -1.23699331e+00 4.22181100e-01 -1.57747701e-01
9.90622163e-01 2.95634776e-01 -6.68435216e-01 2.73022026e-01
1.79966122e-01 6.33944795e-02 7.28414237e-01 -2.93034881e-01
-1.07663749e-02 -8.10782611e-02 -1.29892874e+00 9.11952436e-01
9.79776323e-01 -5.97419918e-01 3.50843184e-02 5.38570464e-01
3.30143064e-01 -3.54788244e-01 -1.07996356e+00 5.44086277e-01
7.00408101e-01 -6.74363494e-01 6.92969620e-01 -7.35344529e-01
6.61722481e-01 -4.85732183e-02 3.14924598e-01 -1.13911867e+00
-1.94915444e-01 -3.17080200e-01 -1.86944157e-02 5.95120728e-01
4.10493225e-01 -6.67257428e-01 9.45698321e-01 2.50095755e-01
2.76843488e-01 -1.15653348e+00 -7.78030336e-01 -4.86831218e-01
-1.96937069e-01 -6.43895343e-02 4.54119146e-02 7.64163673e-01
-2.89882213e-01 -3.32462758e-01 -2.41746053e-01 2.39334211e-01
5.36561251e-01 -2.77116392e-02 5.42757399e-02 -8.47589970e-01
-1.42349944e-01 -5.42690992e-01 -6.13529623e-01 -3.33367705e-01
-2.07089156e-01 -7.69668519e-01 -2.11450055e-01 -1.57107925e+00
4.93284762e-01 -1.95764348e-01 -6.63213313e-01 4.59630460e-01
-3.99439722e-01 2.92252272e-01 -3.05698037e-01 3.21810722e-01
-3.68340522e-01 -1.10454708e-01 1.71058404e+00 -5.23824990e-01
1.14447915e-03 4.34606336e-02 -3.31261486e-01 6.17947042e-01
6.06397450e-01 1.64373174e-01 -5.20104349e-01 -4.32661980e-01
-4.08576801e-02 3.30113232e-01 5.58049858e-01 -9.05574560e-01
6.27227664e-01 -5.50532699e-01 6.31699264e-01 -4.38955665e-01
4.91852283e-01 -4.36357975e-01 1.74401522e-01 1.07814145e+00
-3.30351114e-01 -1.09145083e-01 2.63251275e-01 1.00458288e+00
-3.43463987e-01 -2.32713833e-01 7.39022613e-01 -3.09276909e-01
-1.27811265e+00 7.39733756e-01 -6.88100696e-01 -6.23965800e-01
1.61528432e+00 -5.09673715e-01 -6.05218887e-01 -1.27749920e-01
-8.42242420e-01 -2.76918001e-02 4.76328194e-01 3.01325232e-01
9.57671046e-01 -1.00107837e+00 -7.38806307e-01 -4.70206365e-02
1.28603771e-01 -4.04127985e-01 1.10656905e+00 1.16514456e+00
-8.18814516e-01 2.85437822e-01 -4.48499531e-01 -7.44471192e-01
-1.68174219e+00 4.52001929e-01 5.92171848e-01 -4.46616203e-01
-4.92195934e-01 1.02725351e+00 9.40564461e-03 7.08549693e-02
-8.62551033e-02 -1.66028917e-01 -5.01187295e-02 3.66025679e-02
4.98155326e-01 2.31520385e-01 6.99631125e-02 6.64648190e-02
-5.08427024e-02 3.73874217e-01 -6.34820819e-01 -4.46274281e-02
9.75598931e-01 -1.44638404e-01 -2.06738457e-01 6.36179149e-01
8.58954191e-01 -3.30447346e-01 -1.34617651e+00 -2.63616592e-01
-2.93428689e-01 -3.25895727e-01 -3.75801437e-02 -1.22240889e+00
-9.73888814e-01 8.45512331e-01 1.00702763e+00 2.80588508e-01
1.35589802e+00 -6.47688210e-01 8.60026896e-01 2.00488955e-01
1.45964891e-01 -9.70099986e-01 6.28646165e-02 -7.31884018e-02
4.06396270e-01 -1.29456651e+00 -6.83391169e-02 -3.89072984e-01
-7.93496370e-01 1.27346599e+00 7.21579850e-01 -1.83638468e-01
2.39583805e-01 1.79327801e-01 3.52415830e-01 -1.45897731e-01
-1.29322946e+00 -7.00924844e-02 2.16647208e-01 6.70229554e-01
9.82659087e-02 2.51273394e-01 -1.56297147e-01 -1.58492744e-01
3.92721951e-01 1.70312941e-01 6.42392159e-01 7.10381687e-01
-1.14160188e-01 -9.19412732e-01 -2.57421639e-02 1.04344964e+00
-4.74816501e-01 -1.28356665e-01 -5.57284296e-01 9.20040667e-01
1.68857962e-01 6.61020279e-01 2.72345692e-01 -4.32821780e-01
-2.06245020e-01 -5.30444011e-02 7.53937304e-01 -3.63321543e-01
-3.74114513e-01 -2.29246452e-01 4.56310026e-02 -1.76484272e-01
-5.63510001e-01 -7.93678403e-01 -8.23348522e-01 -1.42423704e-01
-1.05845489e-01 -5.72672367e-01 7.60983169e-01 8.16559494e-01
6.23813644e-02 9.16809201e-01 7.81506360e-01 -1.83030255e-02
-6.42071009e-01 -9.18676615e-01 -8.26817870e-01 3.39962363e-01
5.61648905e-01 -2.72135645e-01 -5.71271330e-02 4.18707758e-01] | [15.655463218688965, -3.000121593475342] |
c2bed743-5ee2-44d7-b69c-100d8926188d | a-step-towards-the-applicability-of | 2304.02286 | null | https://arxiv.org/abs/2304.02286v1 | https://arxiv.org/pdf/2304.02286v1.pdf | A step towards the applicability of algorithms based on invariant causal learning on observational data | Machine learning can benefit from causal discovery for interpretation and from causal inference for generalization. In this line of research, a few invariant learning algorithms for out-of-distribution (OOD) generalization have been proposed by using multiple training environments to find invariant relationships. Some of them are focused on causal discovery as Invariant Causal Prediction (ICP), which finds causal parents of a variable of interest, and some directly provide a causal optimal predictor that generalizes well in OOD environments as Invariant Risk Minimization (IRM). This group of algorithms works under the assumption of multiple environments that represent different interventions in the causal inference context. Those environments are not normally available when working with observational data and real-world applications. Here we propose a method to generate them in an efficient way. We assess the performance of this unsupervised learning problem by implementing ICP on simulated data. We also show how to apply ICP efficiently integrated with our method for causal discovery. Finally, we proposed an improved version of our method in combination with ICP for datasets with multiple covariates where ICP and other causal discovery methods normally degrade in performance. | ['Borja Guerrero Santillan'] | 2023-04-05 | null | null | null | null | ['causal-inference', 'causal-discovery', 'causal-inference'] | ['knowledge-base', 'knowledge-base', 'miscellaneous'] | [ 5.36737025e-01 4.46444303e-01 -4.41906512e-01 -5.05434990e-01
-5.55937946e-01 -2.07621902e-01 8.69622588e-01 4.18006927e-01
-1.13614634e-01 1.12289500e+00 5.94570220e-01 -4.20569241e-01
-9.50672746e-01 -1.11319661e+00 -1.21441400e+00 -7.36414373e-01
-7.96988368e-01 6.45131230e-01 3.95477600e-02 1.92959204e-01
2.27248162e-01 3.13682139e-01 -1.68022346e+00 2.40241930e-01
1.15458226e+00 6.28303811e-02 1.21425048e-01 5.04720151e-01
1.07688695e-01 4.72051352e-01 -2.61502266e-01 -1.75332651e-01
1.57465041e-01 -5.42312145e-01 -8.48086178e-01 -3.99691999e-01
5.21774709e-01 -9.79463235e-02 -4.07975875e-02 7.40770519e-01
4.62416977e-01 1.09036215e-01 1.12717485e+00 -1.55454540e+00
-5.04032493e-01 1.07353926e+00 -4.45793301e-01 5.47965169e-02
4.10483897e-01 -3.01944256e-01 9.86120641e-01 -5.20357192e-01
6.51059449e-01 1.89056945e+00 4.48146582e-01 2.96439201e-01
-1.78618872e+00 -7.53662944e-01 1.78713351e-01 3.40783626e-01
-9.40882683e-01 -1.80693388e-01 4.61693615e-01 -6.80484593e-01
5.84097326e-01 4.53968257e-01 1.49071068e-01 1.41097319e+00
3.06555480e-01 7.02246547e-01 1.11515105e+00 -6.91442370e-01
5.82646906e-01 -4.03665425e-03 2.16583759e-01 3.49279135e-01
3.49029332e-01 6.02953434e-01 -6.21207356e-01 -7.34849691e-01
6.11213624e-01 -1.45554021e-01 -2.89518684e-01 -5.02582967e-01
-1.16405892e+00 1.08907938e+00 2.96832293e-01 2.04598326e-02
-3.93819451e-01 2.54988939e-01 2.88250864e-01 3.20000380e-01
6.55851424e-01 3.37447256e-01 -7.00370431e-01 3.78921062e-01
-5.27045429e-01 7.15338349e-01 8.01134706e-01 7.66956270e-01
4.86617625e-01 -3.88078123e-01 -3.10593724e-01 6.47819877e-01
2.38990679e-01 5.47377706e-01 5.69350384e-02 -7.00408936e-01
3.63423258e-01 4.52524036e-01 2.85962194e-01 -8.88886034e-01
-5.58131456e-01 -3.19263071e-01 -8.79750669e-01 7.38800466e-02
4.56463307e-01 -2.13185370e-01 -6.20507538e-01 2.08708310e+00
6.53206229e-01 8.13527584e-01 5.37106320e-02 5.35373867e-01
2.65610695e-01 5.52569449e-01 3.68434042e-01 -7.20343351e-01
7.83749938e-01 -2.62057126e-01 -6.98587418e-01 3.03768635e-01
6.83580458e-01 -4.83468771e-01 1.03254187e+00 5.63259065e-01
-5.64396381e-01 -3.18818897e-01 -7.57045209e-01 4.91023898e-01
-2.30795383e-01 -3.39887440e-01 9.38819110e-01 7.07294285e-01
-8.58596921e-01 9.60785449e-01 -6.10882819e-01 -4.93584663e-01
1.95099786e-01 4.56487685e-01 -2.12474704e-01 -2.00443380e-02
-1.50649095e+00 7.34041989e-01 3.95457774e-01 -3.84915769e-01
-1.25524235e+00 -1.29811788e+00 -5.70902526e-01 -2.67523490e-02
5.22708654e-01 -8.77335846e-01 9.22555327e-01 -6.72736824e-01
-9.59337831e-01 3.19810599e-01 -2.20329419e-01 -4.58500504e-01
6.27211690e-01 -5.22790730e-01 -3.09406757e-01 -2.97568500e-01
3.61569732e-01 3.15670460e-01 7.11962998e-01 -1.30991781e+00
-6.92336917e-01 -6.00603342e-01 -1.45937994e-01 -2.60879576e-01
-2.36904427e-01 1.05220638e-01 3.03811282e-01 -6.60563707e-01
-1.78997088e-02 -8.30353320e-01 -4.93421793e-01 -5.25105894e-01
-5.30628562e-01 -7.03346491e-01 7.85720944e-01 -4.33316290e-01
1.08925796e+00 -1.74491966e+00 3.89317095e-01 2.80031860e-01
-1.55052915e-01 -3.35755527e-01 -4.82476782e-03 6.49148107e-01
-6.13622010e-01 3.34829241e-01 -5.05706191e-01 3.87798063e-02
2.04132628e-02 3.67534041e-01 -6.86255991e-01 5.62060237e-01
2.19931543e-01 4.07402217e-01 -1.12112319e+00 -5.65660715e-01
4.65051532e-02 1.61034957e-01 -8.94932866e-01 3.88033599e-01
-4.96056378e-01 6.01931214e-01 -4.62392390e-01 2.30961427e-01
5.37039936e-01 9.25828442e-02 4.62426454e-01 2.20890731e-01
-3.17704707e-01 1.97155833e-01 -1.66263235e+00 1.32399249e+00
-4.15183187e-01 1.22598417e-01 -4.00528103e-01 -1.41438365e+00
8.68634701e-01 6.11173451e-01 5.06817758e-01 -1.25987500e-01
-2.38352835e-01 1.75416023e-01 -1.40669599e-01 -7.69004881e-01
-7.36203566e-02 -3.41903865e-01 2.34037209e-02 4.02052015e-01
3.20441574e-02 3.64184082e-01 1.30396662e-02 1.43264085e-01
1.45726752e+00 2.00401187e-01 5.15182197e-01 -6.27435923e-01
2.18015864e-01 -1.40991107e-01 7.91312039e-01 1.15992475e+00
3.90173525e-01 1.64480552e-01 7.53515840e-01 -3.35901320e-01
-9.33528543e-01 -1.47976279e+00 -6.20772362e-01 9.76843476e-01
-7.50610083e-02 -2.12140739e-01 -4.27864999e-01 -7.83807337e-01
1.71372131e-01 1.20941663e+00 -8.23906183e-01 -1.09174594e-01
-3.88955832e-01 -1.33289516e+00 2.31181711e-01 3.37000310e-01
4.56233211e-02 -8.15414727e-01 -4.06869113e-01 2.66005099e-01
-3.14179100e-02 -3.49731505e-01 2.93265730e-01 5.42936623e-01
-1.03447604e+00 -1.24773872e+00 -2.45460987e-01 -2.04227716e-01
4.44026828e-01 -2.58344918e-01 9.76636708e-01 -2.91047007e-01
-2.41181433e-01 2.26198465e-01 -4.54652846e-01 -6.71261728e-01
-7.21067071e-01 -1.58984676e-01 5.08089900e-01 7.83553496e-02
9.91507918e-02 -1.02831197e+00 -1.49821460e-01 2.85037756e-01
-8.95311296e-01 -2.23302141e-01 5.73802769e-01 9.57945049e-01
2.73942679e-01 3.49621981e-01 9.56327975e-01 -1.22503591e+00
4.79020149e-01 -9.97731149e-01 -7.28427291e-01 2.82623500e-01
-7.29330897e-01 4.54169869e-01 5.46682358e-01 -3.84379953e-01
-1.36990893e+00 -7.81029835e-03 2.09718391e-01 -1.57416061e-01
-6.15631819e-01 4.64232266e-01 -5.95820546e-01 5.37359118e-01
9.58812594e-01 -4.06036109e-01 -4.34120178e-01 -6.45638824e-01
4.37050134e-01 5.15365541e-01 3.36565971e-01 -9.76399541e-01
8.16732168e-01 4.53858197e-01 4.98736471e-01 -5.38181067e-01
-7.88282752e-01 -3.29435408e-01 -7.82061100e-01 -5.94878010e-02
8.41982067e-01 -5.48315644e-01 -7.09919155e-01 -1.62726790e-01
-1.16230690e+00 -4.06498224e-01 -1.18753664e-01 8.87527883e-01
-7.00663805e-01 5.19731231e-02 2.00167540e-02 -1.09142673e+00
1.53619200e-01 -7.84727693e-01 1.01914930e+00 -1.42632097e-01
-2.63880759e-01 -1.18431568e+00 4.90800619e-01 -2.04230249e-01
-2.01257229e-01 5.04981339e-01 1.37017155e+00 -6.64730012e-01
-2.77092367e-01 2.00274944e-01 -3.63342427e-02 -1.97609186e-01
1.82454661e-01 -5.87388128e-02 -9.96839285e-01 -7.83226490e-02
-2.56207794e-01 1.13736495e-01 9.28721249e-01 8.21388602e-01
1.17866504e+00 -5.13495564e-01 -9.39163864e-01 3.64705622e-01
1.42316782e+00 3.69730920e-01 4.76872742e-01 1.59736156e-01
5.59833586e-01 1.10677528e+00 7.83235908e-01 5.77789783e-01
-4.24656086e-03 8.33075523e-01 5.48445702e-01 -1.49320573e-01
2.98375517e-01 -4.71027732e-01 3.15624624e-01 -7.11144209e-02
-2.40148142e-01 -2.19397858e-01 -9.26713109e-01 7.26400435e-01
-2.31062579e+00 -1.32372892e+00 -8.33695173e-01 2.41708684e+00
9.84472334e-01 -1.36503369e-01 2.33792290e-01 1.09670751e-01
8.32660198e-01 -5.03691137e-01 -3.92637819e-01 -3.60101342e-01
-1.85153133e-03 3.00020754e-01 5.72384536e-01 6.53936148e-01
-1.30727446e+00 4.82105881e-01 6.48335075e+00 6.26164913e-01
-4.31416303e-01 2.75472790e-01 5.26154160e-01 2.97758222e-01
-6.08231246e-01 4.10372883e-01 -7.75493503e-01 1.66480824e-01
1.06722164e+00 -1.66226476e-01 5.03384247e-02 8.26698780e-01
7.68146753e-01 -8.60062689e-02 -1.58685887e+00 2.86220431e-01
-4.90347207e-01 -1.13979733e+00 -4.45357524e-02 2.07297489e-01
9.22114372e-01 -4.75076169e-01 -1.68970048e-01 4.34439257e-02
7.84552157e-01 -1.00651276e+00 3.34450245e-01 6.46120191e-01
3.53793144e-01 -7.60685086e-01 5.75374067e-01 4.16896939e-01
-5.96338809e-01 -2.89661348e-01 -5.44075966e-01 -3.99011374e-01
1.22234657e-01 1.19816685e+00 -1.12762082e+00 8.91567945e-01
7.94926703e-01 7.60899842e-01 -2.93418735e-01 1.07444954e+00
-4.07152921e-01 1.20509040e+00 -2.74197459e-01 3.42133231e-02
-1.32185981e-01 -6.53525069e-02 8.99535418e-01 1.25326991e+00
3.76512647e-01 1.71662673e-01 2.32631620e-02 1.08805084e+00
3.46619248e-01 1.84024274e-01 -8.29334617e-01 5.69235981e-01
4.57611769e-01 6.03941798e-01 -4.97527331e-01 -2.31131062e-01
-7.19209611e-02 4.95431423e-01 8.05934742e-02 3.28464687e-01
-9.27515984e-01 1.00313172e-01 5.49228847e-01 1.40195534e-01
-1.81768358e-01 1.20632492e-01 -2.81196058e-01 -8.72772336e-01
-4.45355505e-01 -6.89405620e-01 8.90118539e-01 -3.92137855e-01
-1.66507864e+00 -1.28594786e-01 8.66821170e-01 -9.31822121e-01
-1.79252148e-01 -3.43767911e-01 -5.97359598e-01 7.89444745e-01
-1.02585673e+00 -8.72246087e-01 9.67459232e-02 6.20917082e-01
4.30284202e-01 2.96882708e-02 8.89820457e-01 2.35435829e-01
-5.07717192e-01 3.27945828e-01 1.50426179e-01 -5.07963061e-01
1.01494014e+00 -1.68091857e+00 -8.23238567e-02 8.04192364e-01
6.44469112e-02 7.28303075e-01 1.16616082e+00 -1.02488434e+00
-9.53069985e-01 -1.34846127e+00 1.00342619e+00 -3.75288934e-01
7.52177358e-01 -3.33261311e-01 -8.87168229e-01 8.29533994e-01
4.16345634e-02 -5.20985901e-01 5.83275199e-01 1.01446235e+00
-2.34235466e-01 -1.58205643e-01 -1.01106477e+00 7.93182135e-01
1.45665598e+00 1.63019478e-01 -9.21561778e-01 7.28434861e-01
1.08816040e+00 1.36250317e-01 -8.53389621e-01 5.40090799e-01
1.49024546e-01 -9.03134227e-01 9.85641122e-01 -1.03868365e+00
7.01518655e-01 -3.91720682e-01 -1.70846120e-01 -1.45241487e+00
-3.20718467e-01 -5.68750739e-01 2.76488453e-01 1.47534847e+00
3.62603009e-01 -7.37472236e-01 1.64257258e-01 3.66545171e-01
-3.64800282e-02 -2.57497817e-01 -8.93457413e-01 -8.69190335e-01
2.16855571e-01 -6.75976634e-01 5.44432282e-01 1.23640215e+00
3.03908177e-02 4.16103035e-01 -5.00803232e-01 7.10652769e-01
1.21441567e+00 2.23073587e-01 8.29642355e-01 -1.61328852e+00
-5.59931636e-01 4.33152877e-02 -3.87152016e-01 -2.83054590e-01
4.71223593e-01 -8.97420347e-01 -1.40993416e-01 -1.22345233e+00
5.72177589e-01 -7.76665211e-01 -3.67055088e-02 6.43870354e-01
-3.93790334e-01 -5.10366738e-01 -3.57208937e-01 -2.53117569e-02
1.06291816e-01 4.30392414e-01 7.63861537e-01 -1.45791769e-01
-4.70690817e-01 3.14150602e-01 -4.11258727e-01 7.28974938e-01
7.92916059e-01 -9.08371389e-01 -6.75074637e-01 1.18661739e-01
2.42941469e-01 1.66272804e-01 9.38095510e-01 -5.19342065e-01
-1.37677565e-01 -5.63369393e-01 2.30292603e-01 -4.30485636e-01
-4.51309621e-01 -6.63383007e-01 5.93483627e-01 5.08020222e-01
-7.29000390e-01 -2.48887509e-01 -1.36114936e-02 8.44681859e-01
5.99021986e-02 -3.45405906e-01 4.82541084e-01 8.33990499e-02
-5.02505720e-01 -4.07983027e-02 -1.58541292e-01 -1.20260656e-01
9.55468535e-01 3.70849252e-01 -2.92390645e-01 -2.14387283e-01
-9.41168904e-01 1.98797345e-01 -2.96192784e-02 4.05718833e-01
3.64148766e-01 -1.17462850e+00 -1.15931475e+00 -2.80345321e-01
8.28916486e-03 -2.47022301e-01 1.00551888e-01 9.12921429e-01
2.10876390e-01 5.04437089e-01 9.17273387e-02 -6.55008316e-01
-1.12993896e+00 9.74754989e-01 1.11966856e-01 -3.06463748e-01
-5.87940693e-01 5.92382967e-01 7.21377015e-01 -5.49691260e-01
1.18637502e-01 -1.07969381e-01 -1.83438540e-01 -2.65931617e-02
4.40802604e-01 6.13051116e-01 -1.86291233e-01 8.47273916e-02
-3.14701676e-01 1.99938536e-01 3.81624132e-01 -1.90732211e-01
1.67378891e+00 1.11002408e-01 -4.15000767e-01 7.13896632e-01
1.00618994e+00 1.56375453e-01 -1.03359914e+00 1.19224936e-01
4.38219637e-01 -5.17866671e-01 -8.32463130e-02 -8.33808243e-01
-4.48241621e-01 5.07949889e-01 6.97684884e-01 3.26943934e-01
1.23132658e+00 2.94445008e-01 -2.88770258e-01 8.17386881e-02
3.72280985e-01 -6.99271619e-01 -2.38115638e-01 -3.10143661e-02
1.13775051e+00 -1.12728226e+00 7.29607046e-02 -6.79273129e-01
-3.36128771e-02 1.01960111e+00 3.49067122e-01 -9.20985192e-02
7.16390848e-01 3.45315307e-01 -4.20919061e-01 -1.99246988e-01
-9.17915463e-01 -1.57628909e-01 2.63216048e-01 8.31301510e-01
4.56836939e-01 3.66870344e-01 -9.05876100e-01 4.06084597e-01
-1.57162338e-01 -1.56085476e-01 5.53946376e-01 3.79926354e-01
-1.99146509e-01 -1.31751359e+00 -7.52701581e-01 5.06040812e-01
-2.50220656e-01 -1.14367783e-01 -2.70289540e-01 1.00056326e+00
4.03812081e-01 1.27918911e+00 -5.23261689e-02 -2.52856433e-01
4.13861930e-01 5.36958240e-02 3.78287077e-01 -5.77354968e-01
-1.72838554e-01 4.34773490e-02 1.87129691e-01 -5.88224173e-01
-7.03353107e-01 -1.13963687e+00 -1.13091576e+00 -9.34130102e-02
-3.00900042e-01 2.20845833e-01 3.37012768e-01 1.05220532e+00
-2.42094155e-02 6.07501507e-01 7.06751406e-01 -5.27170837e-01
-4.89574462e-01 -8.38305831e-01 -6.78203106e-01 3.57775807e-01
1.60283178e-01 -1.10944593e+00 -5.31535387e-01 3.11538160e-01] | [7.8523030281066895, 5.28935432434082] |
3ec77407-fa19-4cfd-992e-5c5346c57693 | split-brain-autoencoders-unsupervised | 1611.09842 | null | http://arxiv.org/abs/1611.09842v3 | http://arxiv.org/pdf/1611.09842v3.pdf | Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Prediction | We propose split-brain autoencoders, a straightforward modification of the
traditional autoencoder architecture, for unsupervised representation learning.
The method adds a split to the network, resulting in two disjoint sub-networks.
Each sub-network is trained to perform a difficult task -- predicting one
subset of the data channels from another. Together, the sub-networks extract
features from the entire input signal. By forcing the network to solve
cross-channel prediction tasks, we induce a representation within the network
which transfers well to other, unseen tasks. This method achieves
state-of-the-art performance on several large-scale transfer learning
benchmarks. | ['Alexei A. Efros', 'Richard Zhang', 'Phillip Isola'] | 2016-11-29 | split-brain-autoencoders-unsupervised-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Zhang_Split-Brain_Autoencoders_Unsupervised_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Zhang_Split-Brain_Autoencoders_Unsupervised_CVPR_2017_paper.pdf | cvpr-2017-7 | ['self-supervised-image-classification'] | ['computer-vision'] | [ 4.08842236e-01 4.04695004e-01 -1.02711417e-01 -4.18377161e-01
-7.36292064e-01 -4.46119845e-01 3.45945507e-01 -4.52639878e-01
-3.63478392e-01 8.04946542e-01 2.27216542e-01 -1.26232682e-02
2.89561003e-01 -5.91060042e-01 -1.03057098e+00 -7.44488895e-01
-1.70277700e-01 5.30612051e-01 5.14649153e-02 1.08465813e-01
-1.90953568e-01 2.48376727e-01 -1.12702954e+00 8.09043407e-01
1.85437262e-01 1.28664041e+00 1.75287977e-01 5.89724660e-01
1.62163571e-01 1.08791018e+00 -5.90140104e-01 -8.94189253e-02
1.78462416e-01 -4.37300652e-01 -9.49277341e-01 1.24118201e-01
2.57641971e-01 -1.98652342e-01 -9.08897281e-01 8.51746380e-01
3.15061450e-01 1.81041002e-01 9.74875867e-01 -1.18618608e+00
-7.54634321e-01 7.63846159e-01 -2.75508046e-01 3.62884462e-01
-6.27506375e-02 -1.98021919e-01 1.18853831e+00 -8.28434765e-01
4.75454271e-01 1.05693173e+00 5.93698502e-01 8.21942151e-01
-1.87196708e+00 -1.07180858e+00 1.47296548e-01 -3.66199166e-02
-1.16182160e+00 -5.85519433e-01 7.56034493e-01 -4.70854074e-01
1.21377897e+00 -3.09003472e-01 5.09854734e-01 1.61866844e+00
2.22878382e-01 9.46827710e-01 1.09811699e+00 -2.24393547e-01
3.50437582e-01 1.44933954e-01 1.86372027e-01 2.56647915e-01
-2.05532104e-01 1.30090222e-01 -4.82713729e-01 3.37608345e-02
8.67559791e-01 1.89319789e-01 -3.14814717e-01 -3.48000169e-01
-1.09415329e+00 9.79547560e-01 6.72046542e-01 1.74283013e-01
-4.48190689e-01 3.31203640e-01 4.28367555e-01 7.30960190e-01
3.55321795e-01 4.88574386e-01 -8.15627575e-01 1.36909798e-01
-7.02381432e-01 -7.83143714e-02 8.25985074e-01 8.08096051e-01
8.71529937e-01 2.72770166e-01 -8.81949589e-02 6.69417381e-01
4.03529108e-02 -3.99001315e-02 8.25566232e-01 -6.96660399e-01
5.04697800e-01 4.00177479e-01 -4.11556333e-01 -3.61001790e-01
-3.56571555e-01 -6.15909398e-01 -1.00477338e+00 1.90474078e-01
4.15237039e-01 -4.35060948e-01 -1.09538054e+00 1.65787649e+00
-3.47903699e-01 4.88405347e-01 3.35668564e-01 7.91454554e-01
6.92287266e-01 8.67148280e-01 1.01231635e-01 2.35232487e-01
1.07273138e+00 -1.19818652e+00 -3.08270812e-01 -6.10321403e-01
1.84358209e-01 -2.85637021e-01 6.11741662e-01 5.96518278e-01
-9.30610478e-01 -6.97252512e-01 -1.09285355e+00 -1.54140946e-02
-4.68404353e-01 1.53738663e-01 7.76403844e-01 9.12105069e-02
-1.01529539e+00 4.46394056e-01 -6.05262399e-01 9.02084783e-02
8.14568996e-01 6.37099206e-01 -6.89877510e-01 -8.60010237e-02
-1.09592676e+00 5.61907291e-01 6.32527113e-01 -2.53414780e-01
-1.60455108e+00 -6.12054348e-01 -9.22163844e-01 3.60742778e-01
-1.53367296e-01 -3.38391870e-01 1.40572011e+00 -1.59861946e+00
-1.45211446e+00 7.84639657e-01 -2.13262156e-01 -7.36985445e-01
-8.07338879e-02 -1.64639413e-01 -6.00969613e-01 6.63128421e-02
-7.33492598e-02 9.50269938e-01 1.35639358e+00 -1.18291950e+00
-4.69325244e-01 -2.63068795e-01 -3.52385938e-01 -1.52215371e-02
-5.06112874e-01 -1.09257642e-02 -4.27533627e-01 -9.43100333e-01
-1.16481006e-01 -9.11588907e-01 -1.58174545e-01 -3.16162795e-01
-5.59743881e-01 -1.75830647e-01 8.33643854e-01 -4.70226526e-01
8.17201078e-01 -2.52696300e+00 3.23746562e-01 4.15764451e-01
4.88048524e-01 -9.03241932e-02 -4.99445230e-01 2.67288322e-03
-7.73644984e-01 -3.81174892e-01 -1.52637929e-01 -4.55690742e-01
-3.19280803e-01 4.02448118e-01 -7.73716390e-01 3.39015931e-01
4.74658579e-01 9.77314174e-01 -5.55019677e-01 2.94443574e-02
-3.26571316e-01 3.54032367e-01 -5.53691208e-01 5.00619352e-01
-1.60466105e-01 4.21439648e-01 -2.94805527e-01 2.95515209e-01
3.17621410e-01 -5.29994667e-01 1.45011723e-01 1.43700942e-01
3.70879829e-01 4.33754265e-01 -7.67244339e-01 1.81084013e+00
-4.47882950e-01 9.50014710e-01 -5.44532016e-02 -1.62670779e+00
9.15092111e-01 5.41583121e-01 2.90913105e-01 -6.54592097e-01
3.48673880e-01 -8.42230320e-02 2.38312244e-01 -2.19490424e-01
2.15915544e-03 -2.86972314e-01 -3.79652642e-02 7.15863764e-01
1.00367129e+00 3.45651358e-01 -2.35775933e-01 1.00683264e-01
1.25424707e+00 -6.91273063e-02 2.32356146e-01 -5.47750331e-02
1.52023584e-01 -3.47239852e-01 4.35764998e-01 7.01978385e-01
-1.76855907e-01 8.01417947e-01 5.65275311e-01 -5.77856779e-01
-9.42649126e-01 -1.30184734e+00 4.97962162e-03 1.51348114e+00
-3.15289348e-01 -2.60856420e-01 -6.94914818e-01 -9.42205071e-01
2.62033284e-01 3.55094314e-01 -1.19151771e+00 -4.17025715e-01
-1.45368114e-01 -5.06954849e-01 7.33395636e-01 1.06145775e+00
2.86677957e-01 -1.39860129e+00 -4.09183234e-01 2.77870148e-01
1.11293808e-01 -1.15845895e+00 -1.42066434e-01 1.13925707e+00
-1.03412974e+00 -8.74566078e-01 -7.20571518e-01 -1.27159488e+00
7.85802424e-01 -2.55571362e-02 1.30270195e+00 -2.07074821e-01
-2.01399177e-01 1.42619088e-01 -1.07919753e-01 -5.04330218e-01
-1.00223154e-01 1.99168652e-01 -5.12750782e-02 3.23386550e-01
7.33856797e-01 -8.37806344e-01 -3.58034700e-01 2.13402376e-01
-6.56306684e-01 -4.23191309e-01 8.28023136e-01 1.18976963e+00
8.12094033e-01 1.87134221e-02 7.71733284e-01 -1.04436433e+00
6.31429434e-01 -8.07385147e-01 -2.39099234e-01 -1.74615338e-01
-3.03086400e-01 2.99456209e-01 9.47074056e-01 -8.64620984e-01
-7.27447808e-01 4.88673359e-01 1.69665769e-01 -8.46565902e-01
-5.16317666e-01 3.56822848e-01 2.32500695e-02 1.99149437e-02
7.20173419e-01 5.46271503e-01 1.13121629e-01 -6.46191478e-01
2.88682669e-01 6.56413615e-01 7.46558309e-01 -4.11491573e-01
6.11305654e-01 3.28564227e-01 -5.04855096e-01 -3.86638701e-01
-9.23953831e-01 -4.00228053e-01 -6.77147508e-01 1.89234331e-01
1.03448308e+00 -1.35206366e+00 -4.23179418e-01 2.97846586e-01
-1.15073454e+00 -8.16720724e-01 -4.37720209e-01 5.48906565e-01
-5.36000967e-01 -3.88868749e-01 -6.62420750e-01 -3.80696386e-01
-9.61507857e-02 -1.03852868e+00 8.51092577e-01 8.67038444e-02
-2.88000464e-01 -9.66481924e-01 2.62538493e-01 1.27496002e-02
3.28382283e-01 -3.45380992e-01 8.33352923e-01 -1.13563418e+00
-2.24940360e-01 -2.36815050e-01 -2.68563807e-01 8.42662871e-01
-1.93918034e-01 -6.43494666e-01 -1.55249047e+00 -2.83507556e-01
-1.01100698e-01 -9.17432368e-01 1.46068966e+00 3.50167990e-01
1.95718014e+00 -2.12508187e-01 -3.96952868e-01 6.97624445e-01
1.12269521e+00 1.17425084e-01 9.08832371e-01 2.00796902e-01
3.67598236e-01 2.74390161e-01 -3.01387727e-01 2.55473644e-01
1.97281942e-01 2.51342118e-01 3.55228841e-01 -2.38563776e-01
-1.54484054e-02 -3.98184031e-01 5.42536080e-01 6.82977438e-01
-9.11098272e-02 5.47611602e-02 -6.51861191e-01 7.16724515e-01
-1.60610330e+00 -8.29637945e-01 4.27430034e-01 1.82660425e+00
7.99775958e-01 1.24067016e-01 1.91204146e-01 2.10357681e-01
4.48204339e-01 9.49937403e-02 -6.84719980e-01 -3.26908737e-01
-9.74906012e-02 7.73288667e-01 4.91867840e-01 6.41428027e-03
-1.31340086e+00 9.49337244e-01 7.75634909e+00 5.92017829e-01
-1.32319546e+00 1.68731242e-01 8.22468758e-01 -2.66345982e-02
-7.72177801e-03 -9.24853310e-02 -5.98980606e-01 4.08396304e-01
1.05061769e+00 1.85332775e-01 8.48176420e-01 1.04656923e+00
-3.79427195e-01 2.22643346e-01 -1.47312009e+00 9.32026327e-01
2.13056505e-01 -1.17296338e+00 2.59079710e-02 -4.30536754e-02
8.39225173e-01 5.07762253e-01 2.07819194e-01 7.50850260e-01
6.06887877e-01 -1.53074527e+00 4.81114775e-01 4.15290773e-01
8.32456708e-01 -7.61496484e-01 5.27227759e-01 2.74187118e-01
-9.50223327e-01 -4.50770289e-01 -5.70926607e-01 -3.22748363e-01
-3.39207023e-01 4.96620059e-01 -6.42020822e-01 -7.66573548e-02
8.73727977e-01 8.97209823e-01 -3.65818322e-01 7.89153695e-01
-4.19813305e-01 9.70995963e-01 3.48115675e-02 1.65488064e-01
2.20923647e-01 7.75558501e-02 1.90280586e-01 1.39160550e+00
-4.96485718e-02 8.29776376e-02 2.12025449e-01 1.05772412e+00
-8.34635973e-01 -2.62078464e-01 -8.78436983e-01 -2.39463449e-01
3.14179629e-01 1.23927665e+00 -2.39590928e-01 -4.18505579e-01
-7.78708816e-01 1.26485145e+00 1.00646353e+00 9.62116897e-01
-4.19738084e-01 -6.00917161e-01 7.66508758e-01 -1.71853706e-01
7.59186029e-01 2.43156329e-02 -2.41835818e-01 -1.25608778e+00
-2.82905340e-01 -1.00971568e+00 5.96902013e-01 -8.01326990e-01
-1.57431340e+00 6.95278943e-01 -4.87431586e-01 -1.16600049e+00
-3.48279506e-01 -1.07223642e+00 -8.45836997e-01 8.29804540e-01
-1.62145543e+00 -8.28481257e-01 -1.38799846e-01 1.09230018e+00
5.08192003e-01 -7.31854498e-01 1.32488072e+00 1.71877399e-01
-5.30004978e-01 7.25750089e-01 1.05470292e-01 8.10126185e-01
6.27972901e-01 -1.23336792e+00 2.24633470e-01 4.47443962e-01
4.92880791e-01 6.10679746e-01 7.21396878e-02 -8.55854750e-02
-1.09443021e+00 -1.28759718e+00 6.26738906e-01 -3.47919285e-01
8.08008373e-01 -7.71702945e-01 -9.19174492e-01 1.37246513e+00
3.75066549e-01 5.48697352e-01 1.12459302e+00 4.92462754e-01
-8.38223934e-01 -1.37390658e-01 -7.57831454e-01 4.48885486e-02
6.31947458e-01 -1.00500906e+00 -9.87596929e-01 -5.60213104e-02
5.04350066e-01 -1.39568776e-01 -8.76488328e-01 -1.50258571e-01
6.21610641e-01 -4.55860734e-01 9.70648170e-01 -1.38342917e+00
9.74238575e-01 1.80961356e-01 -1.93092644e-01 -1.83922029e+00
-6.35538042e-01 -5.53860068e-01 -5.86526811e-01 6.28759801e-01
8.62381041e-01 -6.29996777e-01 8.01894188e-01 2.73887813e-01
-1.08907893e-01 -8.00449312e-01 -7.78771937e-01 -7.31072068e-01
2.55848914e-01 -5.30806839e-01 2.15087175e-01 1.01698565e+00
3.01931173e-01 8.57914865e-01 -4.37866986e-01 -1.57931652e-02
4.61540282e-01 5.40244691e-02 6.21979833e-01 -1.44541132e+00
-7.44966209e-01 -2.29800195e-01 -4.75534171e-01 -1.39624417e+00
6.45447850e-01 -1.17979050e+00 1.35567188e-01 -1.06076992e+00
2.83070773e-01 -1.46304831e-01 -9.56393421e-01 9.92593825e-01
1.04391307e-01 4.70728844e-01 -3.50363925e-02 2.08560616e-01
-5.99010110e-01 6.54672921e-01 9.85432804e-01 -3.18545341e-01
-2.66336113e-01 -1.04857646e-02 -9.11907852e-01 5.94639659e-01
7.77031541e-01 -7.97429264e-01 -3.72379422e-01 -6.22486532e-01
-9.15819332e-02 -1.77904308e-01 3.46862376e-01 -1.10721338e+00
1.88763663e-01 3.40720773e-01 1.16594887e+00 1.62068848e-02
2.70130157e-01 -9.41327274e-01 -4.75725770e-01 5.50125092e-02
-7.80150056e-01 -2.46692017e-01 3.49205196e-01 7.71203756e-01
-5.84971547e-01 -4.47353497e-02 6.74365222e-01 -1.05302699e-01
-7.59630203e-01 4.35916156e-01 -5.11049509e-01 1.18750528e-01
8.55082095e-01 -1.77683428e-01 -7.01876283e-02 -4.82084692e-01
-1.20980966e+00 2.52476513e-01 -1.85749725e-01 4.57463384e-01
7.78764963e-01 -1.32029486e+00 -5.41337609e-01 6.02865756e-01
2.55419225e-01 5.83176985e-02 -2.17393078e-02 6.60905540e-01
2.77352035e-01 3.50815654e-01 -5.38088858e-01 -3.16896766e-01
-7.07656264e-01 6.11862779e-01 4.67228442e-01 -2.95764387e-01
-7.13470697e-01 1.11685789e+00 6.39792979e-01 -3.75088632e-01
4.21461612e-01 -1.47847399e-01 -3.09355348e-01 2.33365837e-02
6.84673488e-01 -2.29727492e-01 -1.04330420e-01 -4.13228363e-01
5.77590056e-03 7.98411295e-02 -3.83225977e-01 -1.14650555e-01
1.60302246e+00 2.75748372e-01 3.24744806e-02 6.05833769e-01
1.57316542e+00 -5.09127319e-01 -1.50891078e+00 -4.87203032e-01
-2.73369193e-01 1.43746600e-01 3.38029772e-01 -8.86909962e-01
-1.34388173e+00 1.12006462e+00 4.17033494e-01 -6.30093273e-03
1.15502775e+00 2.74946600e-01 6.23755574e-01 8.57264221e-01
5.63649572e-02 -1.29506266e+00 2.92836279e-01 8.45467567e-01
8.88118863e-01 -1.17959368e+00 -4.24774051e-01 2.30841152e-02
-8.07574511e-01 1.30923223e+00 6.71865940e-01 -6.98462129e-01
8.85614991e-01 5.44050992e-01 1.54294423e-03 -2.69163281e-01
-9.57091808e-01 -1.33777549e-02 4.18226808e-01 7.68195987e-01
3.95953149e-01 1.32164419e-01 6.86754107e-01 1.31200683e+00
-1.92543902e-02 1.84967011e-01 1.16694465e-01 6.68794513e-01
-3.58553648e-01 -7.85884023e-01 -1.64855286e-01 8.32899988e-01
-3.29561025e-01 -2.41547272e-01 -5.72128713e-01 5.83695412e-01
-4.14055698e-02 4.83995706e-01 6.10280991e-01 -6.82848513e-01
1.58119857e-01 4.81796443e-01 3.40177566e-01 -7.67207623e-01
-4.53846097e-01 1.22355931e-01 -1.92639843e-01 -5.90658963e-01
-1.98352382e-01 -2.40281403e-01 -1.29077339e+00 3.02804440e-01
-1.11377418e-01 1.28664628e-01 3.42662632e-01 1.02647710e+00
4.17710871e-01 8.25410604e-01 7.53885269e-01 -9.56107140e-01
-5.99764526e-01 -1.23716307e+00 -9.04763162e-01 4.78030175e-01
6.72667980e-01 -6.15612507e-01 -2.74795502e-01 3.31373274e-01] | [9.433745384216309, 2.6292405128479004] |
e6ac3ccb-5b5f-4c60-987f-3e4c491c86c7 | neural-network-based-on-chip-spectroscopy | 2012.00878 | null | https://arxiv.org/abs/2012.00878v1 | https://arxiv.org/pdf/2012.00878v1.pdf | Neural network-based on-chip spectroscopy using a scalable plasmonic encoder | Conventional spectrometers are limited by trade-offs set by size, cost, signal-to-noise ratio (SNR), and spectral resolution. Here, we demonstrate a deep learning-based spectral reconstruction framework, using a compact and low-cost on-chip sensing scheme that is not constrained by the design trade-offs inherent to grating-based spectroscopy. The system employs a plasmonic spectral encoder chip containing 252 different tiles of nanohole arrays fabricated using a scalable and low-cost imprint lithography method, where each tile has a unique geometry and, thus, a unique optical transmission spectrum. The illumination spectrum of interest directly impinges upon the plasmonic encoder, and a CMOS image sensor captures the transmitted light, without any lenses, gratings, or other optical components in between, making the entire hardware highly compact, light-weight and field-portable. A trained neural network then reconstructs the unknown spectrum using the transmitted intensity information from the spectral encoder in a feed-forward and non-iterative manner. Benefiting from the parallelization of neural networks, the average inference time per spectrum is ~28 microseconds, which is orders of magnitude faster compared to other computational spectroscopy approaches. When blindly tested on unseen new spectra (N = 14,648) with varying complexity, our deep-learning based system identified 96.86% of the spectral peaks with an average peak localization error, bandwidth error, and height error of 0.19 nm, 0.18 nm, and 7.60%, respectively. This system is also highly tolerant to fabrication defects that may arise during the imprint lithography process, which further makes it ideal for applications that demand cost-effective, field-portable and sensitive high-resolution spectroscopy tools. | ['Aydogan Ozcan', 'Yair Rivenson', 'Yunzhe Qiu', 'Ashley Clemens', 'Mason Fordham', 'Zachary Ballard', 'Artem Goncharov', 'Calvin Brown'] | 2020-12-01 | null | null | null | null | ['spectral-reconstruction'] | ['computer-vision'] | [ 8.58548701e-01 -3.71511906e-01 5.24415433e-01 6.93319738e-03
-6.23018682e-01 -6.04806721e-01 -1.00184083e-01 1.64237931e-01
-6.39699697e-01 8.05421591e-01 -3.56692195e-01 -7.20449165e-02
8.69759843e-02 -8.74467194e-01 -1.01730525e+00 -1.08489490e+00
1.82184488e-01 3.78547817e-01 2.54599929e-01 3.08895290e-01
1.92720920e-01 3.58157545e-01 -1.47460485e+00 5.76232076e-02
7.73879707e-01 1.33619392e+00 6.77812576e-01 4.83286649e-01
3.02069504e-02 3.47901076e-01 -3.29839349e-01 -1.27943426e-01
2.08226144e-01 -3.02986205e-01 3.06278970e-02 -2.16339156e-01
2.60817915e-01 -4.82412100e-01 -4.31936502e-01 1.18252432e+00
4.78083879e-01 -1.81710258e-01 4.66908693e-01 -4.06932414e-01
-9.12762284e-01 2.63980538e-01 -6.18316412e-01 -2.97730565e-01
4.03864533e-01 3.90871823e-01 6.16845071e-01 -8.01978946e-01
4.61269140e-01 6.50542855e-01 9.37147260e-01 3.76643449e-01
-1.47833979e+00 -9.11368668e-01 -5.92175424e-01 -1.30576327e-01
-1.18075991e+00 -6.46724641e-01 6.69377267e-01 -5.73496222e-01
8.97671163e-01 2.22305395e-02 7.45172858e-01 1.06105745e+00
5.44935524e-01 -3.56964707e-01 1.33421135e+00 -2.82897145e-01
3.70287210e-01 1.02187313e-01 -4.70202655e-01 1.00478804e+00
4.95023310e-01 1.31170213e-01 -5.71642935e-01 -1.17482126e-01
8.75146389e-01 3.84144634e-01 -4.45857912e-01 -1.27134442e-01
-1.14581037e+00 2.95007616e-01 4.59754109e-01 -8.15157667e-02
-2.22053424e-01 8.21082219e-02 1.55535201e-02 1.01188272e-01
1.98374704e-01 9.30606306e-01 -1.93219200e-01 1.58794090e-01
-7.22173214e-01 -4.64030921e-01 7.72288382e-01 5.46918035e-01
1.06065083e+00 -5.53530417e-02 2.27692232e-01 5.83992243e-01
2.67034352e-01 1.07666862e+00 1.89934403e-01 -9.83768821e-01
-2.76097953e-02 5.99980533e-01 4.64753658e-01 -6.70047581e-01
-6.19812548e-01 -4.52581316e-01 -7.89958298e-01 1.27113044e-01
4.83746320e-01 -3.57639343e-01 -7.25803435e-01 1.56009316e+00
2.47376606e-01 -6.62214607e-02 -2.80200057e-02 1.10665870e+00
5.65467060e-01 5.05786061e-01 -4.84586120e-01 -5.73620737e-01
1.19848478e+00 -6.13013446e-01 -4.99694467e-01 -2.63309360e-01
-4.00030129e-02 -1.00867629e+00 1.14601517e+00 2.52950817e-01
-9.71768320e-01 -8.44083279e-02 -1.44571638e+00 -2.20985651e-01
-1.49338618e-01 3.23421001e-01 5.70295215e-01 6.24396861e-01
-5.97917616e-01 6.62946403e-01 -7.93201625e-01 -1.33088306e-01
2.94131696e-01 4.76179004e-01 -8.50877725e-03 -3.37584913e-02
-5.83987117e-01 2.89339483e-01 1.14812233e-01 -2.67775118e-01
-4.82629269e-01 -1.16991925e+00 -2.19538420e-01 -1.10824015e-02
2.78286695e-01 -6.37108505e-01 9.64984417e-01 -5.45070291e-01
-2.17330837e+00 7.28102744e-01 -2.39991352e-01 -3.46357405e-01
2.22847596e-01 1.78003713e-01 -6.19541049e-01 4.68807876e-01
1.73307136e-01 5.85298017e-02 7.11746037e-01 -8.76032948e-01
-1.01126306e-01 -4.58922029e-01 -4.38278407e-01 -3.84669065e-01
-4.08006787e-01 -3.36604491e-02 1.21706784e-01 2.09999844e-01
3.82770389e-01 -8.21686029e-01 7.86294267e-02 9.55754966e-02
-4.21569467e-01 6.21231616e-01 8.10901344e-01 -2.41675884e-01
6.13722324e-01 -2.22361493e+00 -4.56882238e-01 1.51768491e-01
3.71520728e-01 3.70666496e-02 -4.55320925e-02 7.16362000e-01
3.31894606e-01 -2.39358962e-01 -3.66888553e-01 1.33836851e-01
-2.38560423e-01 -4.26865190e-01 -1.09237703e-02 5.43511093e-01
-9.63656530e-02 6.28683209e-01 -9.92045403e-01 2.41447121e-01
1.94629878e-01 3.75085860e-01 -2.88537145e-01 1.74122170e-01
-2.66687453e-01 5.27801335e-01 -1.72998235e-01 9.61251676e-01
7.14669228e-01 -7.60158479e-01 4.61047560e-01 -6.95071042e-01
-7.92753160e-01 3.27349544e-01 -7.49302506e-01 1.73419583e+00
-3.11769307e-01 5.75628996e-01 3.15593570e-01 -5.74144125e-01
1.14877760e+00 4.62150723e-02 6.86809480e-01 -1.08478272e+00
5.55363633e-02 6.97246671e-01 -4.52265590e-02 -6.97605968e-01
1.64705336e-01 -5.32140791e-01 1.69734493e-01 6.23112500e-01
-6.05933331e-02 1.54404968e-01 -2.94224676e-02 -3.04289550e-01
1.40826559e+00 -1.83395877e-01 -9.93615296e-03 -3.70548934e-01
6.26346990e-02 -1.83148365e-02 3.82315010e-01 8.42956543e-01
8.63575041e-02 5.23985684e-01 6.36366904e-02 -5.31043291e-01
-1.49716544e+00 -1.00163054e+00 -2.00799629e-01 7.65742600e-01
3.32828850e-01 -1.22949377e-01 -6.00954115e-01 2.80052572e-01
1.68532327e-01 -1.60445254e-02 -6.52840063e-02 -3.97632532e-02
-4.77731675e-02 -8.16620767e-01 3.47203016e-01 1.38663754e-01
6.65196419e-01 -4.94617969e-01 -9.69032168e-01 6.22528017e-01
5.76600209e-02 -1.47135425e+00 -1.78831324e-01 1.74753994e-01
-4.59479392e-01 -1.16751397e+00 -1.23973079e-01 -4.78764683e-01
6.22839153e-01 5.36217213e-01 6.10750794e-01 -5.65822780e-01
-7.91917205e-01 3.90847534e-01 1.13826506e-01 -5.25163949e-01
-1.59449369e-01 -1.87540978e-01 2.98088908e-01 1.24831811e-01
3.24816048e-01 -1.01591980e+00 -1.02506709e+00 1.44380629e-01
-3.53291065e-01 2.92952895e-01 8.99485588e-01 7.94459283e-01
8.92850637e-01 -9.11655277e-02 4.49579775e-01 -6.13140941e-01
3.42080206e-01 -1.38918892e-01 -1.07389402e+00 1.34074658e-01
-4.92729723e-01 5.42898709e-03 9.22318220e-01 -4.28430706e-01
-9.16437685e-01 4.03311074e-01 1.86759695e-01 -2.51035035e-01
7.02569857e-02 3.63264561e-01 3.20046358e-02 -6.84747279e-01
9.30028141e-01 2.53848046e-01 3.02386880e-01 -1.97499260e-01
-2.47584790e-01 6.71068311e-01 6.40997291e-01 -4.33108062e-01
6.62889063e-01 6.01588964e-01 2.32766688e-01 -1.29932702e+00
-7.24049330e-01 -4.14970547e-01 -1.68243781e-01 -3.05503815e-01
7.18847096e-01 -1.08594120e+00 -1.43672585e+00 5.91861784e-01
-9.77654934e-01 -2.90047914e-01 7.54154921e-02 6.14071190e-01
-8.87817293e-02 1.41353011e-01 -7.66645312e-01 -8.65522325e-01
-6.24394655e-01 -1.18512273e+00 1.02489400e+00 2.62143642e-01
1.17036164e-01 -5.20191014e-01 -3.03413123e-01 5.30669332e-01
6.03841066e-01 4.75657344e-01 9.64975834e-01 1.73153564e-01
-1.07427943e+00 -1.94046006e-01 -6.80466235e-01 5.88948689e-02
3.89671385e-01 -2.33699635e-01 -1.14366961e+00 -3.89501303e-01
2.12034389e-01 -5.68929732e-01 8.29668522e-01 3.57100487e-01
1.18204665e+00 -1.96649715e-01 -3.87540877e-01 7.45640218e-01
1.76454020e+00 2.45309651e-01 4.05310273e-01 8.93844441e-02
8.10541511e-01 2.56508570e-02 1.35909561e-02 5.66108823e-01
-5.81022911e-02 2.93696731e-01 4.37328398e-01 6.09894237e-03
2.93185823e-02 -5.10388846e-03 4.21054661e-01 7.60237992e-01
-7.19693005e-02 -1.67568773e-01 -7.31509089e-01 9.41297505e-03
-1.47134197e+00 -8.72049928e-01 -3.21427852e-01 2.39286900e+00
9.51225817e-01 -2.13847563e-01 -1.72847405e-01 -2.60870904e-01
6.72456801e-01 -5.88816553e-02 -1.01538324e+00 6.73268661e-02
-3.87143455e-02 5.03256321e-01 9.90503788e-01 4.05390739e-01
-7.73280323e-01 5.06283462e-01 6.24062777e+00 2.11315274e-01
-1.51006198e+00 3.52251343e-02 1.16180532e-01 -3.76031548e-01
-3.91062200e-01 -2.52036303e-01 -5.18640876e-01 6.51649117e-01
9.64888692e-01 1.79862767e-01 9.77660537e-01 3.14368993e-01
1.85194209e-01 -3.05829972e-01 -9.52867806e-01 1.17968917e+00
-3.94586831e-01 -1.72178411e+00 -3.26454818e-01 2.29529142e-01
5.76199532e-01 3.03696990e-01 2.22433165e-01 -6.33568227e-01
-1.19925320e-01 -7.38321006e-01 5.73459268e-01 6.51961446e-01
1.36696315e+00 -5.11729121e-01 3.48488778e-01 2.22722813e-01
-8.28968346e-01 -2.91130960e-01 -7.31382132e-01 -4.55974549e-01
-3.15793872e-01 1.43135464e+00 -8.99475455e-01 1.74515873e-01
6.78395748e-01 3.20550948e-01 2.35416099e-01 5.24867713e-01
6.66773617e-02 5.60107291e-01 -5.12053072e-01 -4.07738000e-01
5.75158857e-02 -6.65932775e-01 4.49390441e-01 1.15008736e+00
7.30918705e-01 1.70592487e-01 2.14098841e-01 1.44143438e+00
-4.84062612e-01 -3.94665688e-01 -3.05538177e-01 -3.58222336e-01
8.86989772e-01 1.53428805e+00 -7.10162759e-01 1.37240037e-01
-5.82580805e-01 7.69130111e-01 1.67706877e-01 4.08917040e-01
-5.98461092e-01 -4.76672947e-01 8.05584729e-01 3.61315429e-01
2.39396676e-01 -4.20650154e-01 -3.27809066e-01 -1.06450343e+00
2.25170434e-01 -4.21348542e-01 -4.61049378e-01 -7.54846513e-01
-1.32677138e+00 -1.23853408e-01 -9.85303342e-01 -7.59836853e-01
6.01696312e-01 -8.91445875e-01 -1.29335821e-01 8.80794168e-01
-1.56816435e+00 -6.05207384e-01 -5.13027966e-01 3.56470853e-01
-1.66360661e-01 -4.43982799e-03 1.12197590e+00 3.33494067e-01
-4.95770931e-01 2.52770156e-01 6.93415821e-01 3.24695334e-02
8.58182847e-01 -9.20907855e-01 1.26391008e-01 6.50324047e-01
-3.07093531e-01 6.16658866e-01 4.72752035e-01 -5.88022232e-01
-2.22876740e+00 -1.26187134e+00 3.02874476e-01 -2.05508643e-03
9.55229402e-01 -6.35652721e-01 -7.31866002e-01 8.57635066e-02
2.72518191e-02 1.69034526e-01 1.11760199e+00 -2.59617329e-01
-4.98555303e-01 -5.49820423e-01 -1.29428840e+00 2.87226766e-01
1.27076066e+00 -8.41461182e-01 7.62124285e-02 6.73664987e-01
6.12406850e-01 -5.43489695e-01 -7.44794011e-01 2.57796884e-01
1.02885962e+00 -1.13276291e+00 8.69994223e-01 2.11844951e-01
4.65705752e-01 -2.06266686e-01 -1.37357995e-01 -8.25473189e-01
-7.02678978e-01 -7.24104822e-01 1.87306538e-01 7.73049474e-01
5.64369321e-01 -1.18039262e+00 7.87085593e-01 3.10307950e-01
-4.02460128e-01 -5.09414375e-01 -8.07965159e-01 -7.74080276e-01
-5.86122930e-01 5.99297211e-02 6.21082842e-01 7.49627411e-01
-4.41967137e-02 4.02429163e-01 -1.84275225e-01 5.62766612e-01
1.03054154e+00 4.63575184e-01 1.97732463e-01 -1.42676198e+00
-2.75042236e-01 -2.99113411e-02 -8.27588215e-02 -8.37401152e-01
-2.25542128e-01 -6.54444873e-01 1.12146311e-01 -1.19636822e+00
2.69069403e-01 -2.64663607e-01 -3.10533464e-01 3.70949775e-01
3.68552923e-01 2.92876333e-01 -1.32551357e-01 3.41432035e-01
-5.33719540e-01 2.07810029e-01 1.21976209e+00 -2.07170218e-01
-2.72325039e-01 -3.61577839e-01 -6.39711261e-01 4.30623442e-01
8.20902944e-01 -3.84030998e-01 -3.59543204e-01 -5.35438538e-01
5.06262183e-01 4.73102815e-02 5.73716879e-01 -1.56624818e+00
6.35261416e-01 -1.20439261e-01 6.44943893e-01 -5.32832928e-02
4.01762098e-01 -7.50791013e-01 4.80598241e-01 5.46576917e-01
1.11255303e-01 -6.61686897e-01 2.04080905e-04 6.05191231e-01
2.49086618e-01 1.17797228e-02 8.36049557e-01 -4.71949548e-01
-2.86493838e-01 2.67945558e-01 -5.04173458e-01 -2.55340636e-01
7.14598298e-01 -3.49651664e-01 -5.75087667e-01 1.38218388e-01
-8.68425965e-02 -3.53925526e-01 8.19050789e-01 -4.92964059e-01
6.24168158e-01 -1.00656080e+00 -1.63534477e-01 4.57534909e-01
4.79532182e-02 -2.00874463e-01 4.25425500e-01 7.20394373e-01
-5.71188152e-01 2.59872735e-01 -1.83721468e-01 -6.81289434e-01
-7.55497336e-01 2.03863174e-01 4.45108384e-01 3.13028485e-01
-4.50735927e-01 7.84054399e-01 -1.88017160e-01 -3.15433681e-01
-3.19608092e-01 -2.03432202e-01 7.73438692e-01 -3.40267986e-01
5.60050964e-01 4.25689787e-01 3.49424064e-01 1.51983961e-01
-2.53201485e-01 6.55965149e-01 8.67670104e-02 1.82819724e-01
1.29383636e+00 6.95818216e-02 -3.91425550e-01 3.69928032e-01
1.14049649e+00 9.47997347e-02 -1.55500531e+00 -3.24445158e-01
-4.98216033e-01 -2.27491468e-01 1.30816847e-01 -9.86777842e-01
-7.40654647e-01 6.11946404e-01 5.85283458e-01 2.60448754e-01
1.05931747e+00 -1.44741893e-01 1.02146852e+00 6.82369232e-01
4.82310176e-01 -1.31123579e+00 9.78334323e-02 4.74870324e-01
3.96851450e-01 -1.04142809e+00 5.90717532e-02 -3.73642862e-01
2.22446471e-01 1.39257431e+00 2.99768597e-01 1.22300275e-01
1.65643707e-01 6.91201866e-01 -1.31111443e-01 -4.08099890e-01
-4.22401756e-01 1.73111811e-01 -2.17489123e-01 6.38117433e-01
2.64207512e-01 4.67351556e-01 2.86751956e-01 3.66671383e-01
-1.30575851e-01 -5.98765463e-02 3.97524148e-01 6.80601656e-01
-9.35616136e-01 -6.94963396e-01 -2.04315230e-01 6.93063200e-01
-5.80193400e-02 -8.19763988e-02 -3.27877522e-01 1.96869999e-01
1.95970312e-01 8.18950474e-01 -2.87854066e-03 -5.78885138e-01
2.28431255e-01 7.53715560e-02 6.62410676e-01 -5.33725917e-01
1.24700889e-02 1.62036762e-01 -1.34785265e-01 -6.62918329e-01
-5.75791597e-01 -5.00821710e-01 -1.35652733e+00 -3.87414426e-01
-2.15400681e-01 -1.97936639e-01 7.54068255e-01 9.19635534e-01
8.77068341e-01 3.84251922e-01 6.42391741e-01 -7.43105292e-01
-3.07690471e-01 -5.23429394e-01 -7.50877380e-01 -1.72630563e-01
4.57313508e-01 -5.11407614e-01 -2.76715040e-01 -1.79436341e-01] | [10.224894523620605, -2.476801872253418] |
61bedacb-c440-4049-b474-ea7b291aa2a9 | unibuckernel-geolocating-swiss-german-jodels | 2102.09379 | null | https://arxiv.org/abs/2102.09379v3 | https://arxiv.org/pdf/2102.09379v3.pdf | UnibucKernel: Geolocating Swiss German Jodels Using Ensemble Learning | In this work, we describe our approach addressing the Social Media Variety Geolocation task featured in the 2021 VarDial Evaluation Campaign. We focus on the second subtask, which is based on a data set formed of approximately 30 thousand Swiss German Jodels. The dialect identification task is about accurately predicting the latitude and longitude of test samples. We frame the task as a double regression problem, employing an XGBoost meta-learner with the combined power of a variety of machine learning approaches to predict both latitude and longitude. The models included in our ensemble range from simple regression techniques, such as Support Vector Regression, to deep neural models, such as a hybrid neural network and a neural transformer. To minimize the prediction error, we approach the problem from a few different perspectives and consider various types of features, from low-level character n-grams to high-level BERT embeddings. The XGBoost ensemble resulted from combining the power of the aforementioned methods achieves a median distance of 23.6 km on the test data, which places us on the third place in the ranking, at a difference of 6.05 km and 2.9 km from the submissions on the first and second places, respectively. | ['Radu Tudor Ionescu', 'Sebastian Cojocariu', 'Mihaela Gaman'] | 2021-02-18 | null | https://aclanthology.org/2021.vardial-1.10 | https://aclanthology.org/2021.vardial-1.10.pdf | eacl-vardial-2021-4 | ['dialect-identification'] | ['natural-language-processing'] | [-6.26176715e-01 6.90668002e-02 -2.54118949e-01 -5.13835013e-01
-1.04300129e+00 -7.12508738e-01 1.22794700e+00 4.95124131e-01
-7.56658375e-01 7.26455748e-01 4.32006776e-01 -3.23633462e-01
-3.77847075e-01 -9.66217160e-01 -5.40179789e-01 -6.38337076e-01
-3.93339666e-03 5.86471438e-01 -1.89844653e-01 -3.69860530e-01
4.60413277e-01 3.65929931e-01 -1.51186728e+00 5.20846024e-02
1.01186597e+00 1.13912189e+00 -4.86141622e-01 5.47092319e-01
-2.92890072e-01 4.95711148e-01 -6.06678426e-01 -7.17673481e-01
2.66769230e-01 -2.73604848e-04 -6.02581143e-01 -6.49941444e-01
5.58825612e-01 -6.34527579e-02 -3.71882245e-02 8.19030404e-01
7.63967633e-01 1.18587084e-01 8.78522038e-01 -1.13936806e+00
-3.47598910e-01 7.14914978e-01 -4.08497542e-01 4.42278646e-02
4.53536302e-01 -4.15395975e-01 1.11684716e+00 -8.94537091e-01
3.14913690e-01 8.69832098e-01 8.77233267e-01 1.03745103e-01
-9.64851558e-01 -3.43811780e-01 5.27900010e-02 1.18040681e-01
-1.52696931e+00 -4.00458753e-01 6.19541943e-01 -8.76912534e-01
7.49801278e-01 4.30088878e-01 5.55858910e-01 1.01327741e+00
1.06249005e-01 5.17866075e-01 1.32285261e+00 -5.01133978e-01
4.12124991e-01 6.50340497e-01 1.39716521e-01 2.37315774e-01
9.95805189e-02 4.09026537e-03 -4.45154846e-01 -4.70601499e-01
-1.84489429e-01 -1.03211418e-01 -4.03845496e-02 -1.14984691e-01
-1.00248325e+00 1.06167781e+00 6.16725326e-01 3.61978382e-01
-2.65916973e-01 -2.09427699e-01 1.91935822e-01 5.59348986e-02
9.71422017e-01 6.62630320e-01 -5.91948986e-01 1.81826726e-02
-1.11167383e+00 3.63447964e-01 9.27096367e-01 4.50053543e-01
7.72819281e-01 -3.25698465e-01 5.28318696e-02 8.26493621e-01
4.29923624e-01 4.32526261e-01 6.71128333e-01 -1.96124181e-01
5.14157116e-01 4.19698000e-01 1.35607734e-01 -1.13497615e+00
-6.47326112e-01 -8.52908790e-01 -5.55226684e-01 1.12241007e-01
6.23259127e-01 -7.67242432e-01 -5.98986804e-01 1.58613074e+00
5.14605224e-01 7.43228570e-02 -1.42665327e-01 8.73413444e-01
6.70842409e-01 7.67868519e-01 7.03779906e-02 2.46383578e-01
1.26436424e+00 -1.04149902e+00 -1.97264791e-01 -1.56850561e-01
8.86036694e-01 -4.99080211e-01 6.08021200e-01 3.47560883e-01
-7.54734993e-01 -5.16488433e-01 -1.22512364e+00 2.13769764e-01
-1.18701947e+00 3.56855318e-02 3.18912774e-01 7.35714138e-01
-1.34633982e+00 6.28748059e-01 -1.86407119e-01 -4.42939013e-01
-1.98521942e-01 2.19395205e-01 -2.18976736e-01 2.25737751e-01
-1.54343760e+00 1.00803983e+00 8.76223892e-02 1.34640962e-01
6.03038759e-04 -6.77827716e-01 -5.89224577e-01 5.11054918e-02
-2.07046211e-01 -5.50821483e-01 7.08386302e-01 -9.24060106e-01
-1.26060212e+00 8.82144928e-01 7.08366930e-02 -5.37914395e-01
8.94535601e-01 -4.55098689e-01 -7.87445426e-01 -3.49385142e-01
2.99366832e-01 3.99450958e-01 3.75297755e-01 -9.91951764e-01
-9.46355999e-01 -5.65312028e-01 -1.40186563e-01 2.22957432e-01
-4.37639087e-01 -1.21629775e-01 -2.71627247e-01 -4.76682454e-01
-7.39579126e-02 -1.08310866e+00 -1.22902855e-01 -7.01005280e-01
-5.03039300e-01 -2.32807845e-01 2.04718068e-01 -7.64220715e-01
1.40419674e+00 -1.94537377e+00 9.05968342e-03 5.79716682e-01
2.27316961e-01 2.47280137e-03 2.14430824e-01 6.58856153e-01
-5.38174920e-02 2.47538522e-01 2.34670013e-01 -2.82217622e-01
1.45155728e-01 -4.95365888e-01 -4.62321490e-01 6.49572074e-01
-1.44643903e-01 6.57728910e-01 -8.73137116e-01 -8.81365985e-02
4.71757874e-02 3.92301023e-01 -2.94782430e-01 -6.12097830e-02
-4.91039194e-02 2.81336874e-01 -3.77776355e-01 4.99652237e-01
7.85664201e-01 1.34071723e-01 1.83870316e-01 -5.44371782e-03
-5.07296085e-01 4.64417636e-01 -7.38621354e-01 1.40875745e+00
-4.76906210e-01 6.44767761e-01 -3.61859560e-01 -8.04745555e-01
1.20764339e+00 1.62002593e-02 5.43998301e-01 -8.62746596e-01
1.81765277e-02 4.64163095e-01 -2.71885514e-01 -3.25492561e-01
6.81089818e-01 1.15482494e-01 -5.56554377e-01 2.62897432e-01
-1.18620820e-01 4.13346201e-01 1.80633813e-02 -1.84103131e-01
7.22890913e-01 1.41132891e-01 2.52587348e-01 -5.46519697e-01
6.16672337e-01 9.21601057e-02 2.78101355e-01 7.94952691e-01
-1.38066903e-01 5.57985246e-01 6.53625786e-01 -7.68632591e-01
-8.40974510e-01 -8.23200703e-01 -3.31923336e-01 1.13738954e+00
1.70232076e-02 -4.75940883e-01 -7.67803609e-01 -8.48072171e-01
3.97346914e-01 7.49459445e-01 -8.03370833e-01 -2.05477178e-02
-3.86777163e-01 -1.03014529e+00 6.74500644e-01 1.79870725e-01
3.15339625e-01 -2.99632996e-01 -9.19233542e-03 -2.70289127e-02
-3.76565903e-02 -6.42032683e-01 -9.85964015e-03 3.10523957e-01
-3.89768898e-01 -5.99782825e-01 -8.99110377e-01 -6.11617744e-01
2.22083628e-01 -1.65872723e-01 1.13600063e+00 -3.26111972e-01
1.57751322e-01 5.92911709e-03 -4.10404384e-01 -3.77460182e-01
1.42942294e-01 6.90286875e-01 1.93390921e-01 4.49937224e-01
6.06938004e-01 -4.83236402e-01 -6.31217182e-01 3.02505434e-01
-3.34073424e-01 -2.00497985e-01 3.47565264e-01 6.42853498e-01
2.80284196e-01 -2.45411113e-01 6.13368928e-01 -9.12435949e-01
4.71926570e-01 -1.09121013e+00 -5.58245778e-01 1.75807834e-01
-7.69888103e-01 -1.36400172e-02 6.45771086e-01 -1.20440513e-01
-5.10295689e-01 -1.87960207e-01 -3.54058653e-01 2.80914694e-01
-1.05667807e-01 8.83631706e-01 -2.07048744e-01 1.10217266e-01
7.84118950e-01 7.75771812e-02 -2.64282197e-01 -6.58956051e-01
4.11885113e-01 1.04046237e+00 1.88134924e-01 -3.32995117e-01
7.80460298e-01 1.22808091e-01 -3.21852148e-01 -7.50251651e-01
-8.58244538e-01 -4.22965258e-01 -5.42920947e-01 -3.30607921e-01
9.19400573e-01 -9.99288976e-01 -4.90228862e-01 4.72747564e-01
-8.15144122e-01 -2.36004248e-01 9.64360014e-02 6.96202278e-01
-2.99895734e-01 -2.90004075e-01 -2.00920984e-01 -6.24140382e-01
-1.33729592e-01 -6.75111234e-01 9.75212157e-01 2.48435751e-01
-3.12011093e-01 -1.17089367e+00 6.02119327e-01 2.32120857e-01
5.55743754e-01 4.88922626e-01 9.30650949e-01 -1.21160793e+00
-1.72192991e-01 -5.38849473e-01 -1.68998055e-02 1.75865814e-02
-1.38652399e-01 -1.12181768e-01 -1.25070632e+00 -3.31298769e-01
-1.87939763e-01 1.22791436e-02 9.14129913e-01 3.21308076e-01
8.49273205e-01 -2.50999480e-01 -4.25862730e-01 7.31108010e-01
1.67497301e+00 -6.84749102e-03 4.07333136e-01 8.18830669e-01
5.54401100e-01 7.92951465e-01 6.20426059e-01 3.90685648e-01
7.82468021e-01 1.04410803e+00 4.75728363e-01 -9.73931774e-02
3.11506480e-01 -3.91216964e-01 1.15905836e-01 5.95235765e-01
-3.49226743e-02 -4.25181031e-01 -1.42762804e+00 4.21076030e-01
-1.84736454e+00 -8.04922283e-01 -3.51759745e-03 2.46381259e+00
3.22798967e-01 1.19454540e-01 4.04914737e-01 2.30456799e-01
7.69650817e-01 4.66925889e-01 -9.33726430e-02 -5.84366202e-01
-6.32797182e-02 -3.76809418e-01 7.35818803e-01 5.72202563e-01
-1.38351285e+00 6.63804889e-01 5.88824463e+00 5.40996015e-01
-1.47504616e+00 -1.78760156e-01 8.29388976e-01 -2.17026949e-01
-3.21616828e-01 -9.66726765e-02 -1.02205503e+00 8.25881362e-01
1.41261125e+00 -1.97571695e-01 2.64861137e-01 8.36889327e-01
-5.53139746e-02 -1.16266936e-01 -8.37704539e-01 8.85490179e-01
3.82324815e-01 -1.12950957e+00 -2.33602747e-01 2.89590925e-01
7.63163090e-01 5.15642107e-01 3.04916590e-01 4.75376010e-01
1.87559277e-01 -1.05011511e+00 9.18624640e-01 8.35868359e-01
5.31585217e-01 -8.00111353e-01 9.72380102e-01 4.50000703e-01
-8.15087676e-01 -1.75876051e-01 -5.06114624e-02 -4.50321659e-02
9.16477889e-02 8.93394709e-01 -8.43676329e-01 7.92659223e-01
8.07553649e-01 6.63671255e-01 -7.54912794e-01 1.20674217e+00
8.18954259e-02 4.32168663e-01 -3.19601089e-01 -5.35629034e-01
3.58370751e-01 -2.03532502e-01 4.05665815e-01 1.15064561e+00
6.40909135e-01 -5.00959992e-01 -9.66896489e-02 2.67931074e-01
-3.15801241e-02 4.92190599e-01 -6.32426262e-01 2.13514715e-01
4.39280778e-01 1.30756605e+00 -3.51228058e-01 -3.30198139e-01
-3.47272336e-01 5.53053081e-01 5.67431390e-01 3.85698617e-01
-1.00958359e+00 -5.10454237e-01 6.08357251e-01 3.65835845e-01
1.47983879e-01 -3.73268016e-02 -4.57152069e-01 -1.09868813e+00
-4.76931594e-02 -4.90783095e-01 2.54008740e-01 -6.13456786e-01
-1.20873320e+00 9.96362627e-01 -1.95030153e-01 -1.46864462e+00
-5.55926919e-01 -5.74983478e-01 -5.47213137e-01 1.10357893e+00
-1.51925063e+00 -9.57950830e-01 -1.92099646e-01 1.61846280e-01
-9.34268907e-02 -4.38218653e-01 8.63168001e-01 6.04542613e-01
-4.79863226e-01 6.66118681e-01 7.90846646e-01 1.80876017e-01
8.98898542e-01 -1.51592398e+00 3.74124616e-01 4.10124034e-01
3.44005190e-02 4.73024637e-01 6.70134485e-01 -3.68313849e-01
-1.06926680e+00 -1.16576862e+00 1.77651107e+00 -5.06635010e-01
7.23590553e-01 -5.93239009e-01 -3.54544520e-01 4.48430538e-01
8.37608520e-03 -3.19857784e-02 1.01070929e+00 6.90317333e-01
-2.34567657e-01 -3.81467611e-01 -1.04771900e+00 4.62966412e-01
4.28084165e-01 -6.50680780e-01 -2.77376562e-01 2.39929765e-01
1.84540927e-01 -2.01413795e-01 -9.28548872e-01 3.93682688e-01
8.19863260e-01 -1.07097363e+00 8.59045446e-01 -5.81184208e-01
5.76193154e-01 -1.40233114e-01 -3.44755024e-01 -1.66792822e+00
-3.14360201e-01 -1.32643700e-01 2.47039080e-01 1.53387499e+00
8.86170208e-01 -8.88801813e-01 8.30252588e-01 4.67052460e-01
1.69387043e-01 -8.87316823e-01 -8.92960191e-01 -5.03428698e-01
3.25061917e-01 -3.01378667e-01 7.87097096e-01 1.05081880e+00
-2.23392546e-02 4.18720096e-01 -5.07153630e-01 2.04371139e-02
3.11007529e-01 1.84674785e-01 8.66554499e-01 -1.58359420e+00
-2.29130015e-01 -5.99616647e-01 -4.84986991e-01 -8.44737887e-01
1.26442328e-01 -1.16837096e+00 -2.39642069e-01 -1.34766948e+00
-2.31561124e-01 -6.13814592e-01 -4.17569280e-01 -5.38947247e-02
-3.25943753e-02 2.27766544e-01 6.43915758e-02 1.65817603e-01
-4.50814754e-01 3.43404293e-01 3.68586183e-01 -1.97767064e-01
-2.11178780e-01 3.32449436e-01 -9.69887674e-01 7.29960203e-01
8.53572845e-01 -3.58199000e-01 -8.49424824e-02 -6.38190985e-01
6.67775333e-01 -7.14651942e-02 1.54863000e-01 -9.68299508e-01
2.22393140e-01 1.83388472e-01 6.02102160e-01 -5.09414434e-01
2.52660483e-01 -5.20210087e-01 6.62842318e-02 2.69069672e-01
-3.63696665e-01 2.98725873e-01 2.83582928e-03 4.84360367e-01
-4.09333467e-01 1.75294038e-02 3.41908574e-01 2.05161944e-01
-4.47018474e-01 1.90346673e-01 -2.41884857e-01 -9.09667462e-02
9.20189679e-01 -1.96229905e-01 -4.71267700e-01 -3.76732498e-01
-9.65667188e-01 1.89615265e-01 3.50076228e-01 6.05659068e-01
-7.12936297e-02 -1.51990616e+00 -8.75232220e-01 1.70779884e-01
2.64044285e-01 -5.67189813e-01 1.71134621e-01 7.75132954e-01
-5.78348637e-01 5.61675549e-01 -1.19158924e-01 -3.48527014e-01
-8.17872465e-01 2.97158331e-01 3.74474168e-01 -4.03215200e-01
1.07943937e-02 6.88339472e-01 -2.83941001e-01 -8.87426555e-01
1.02197930e-01 1.66452110e-01 -6.35285318e-01 7.65652299e-01
5.32614648e-01 6.06315613e-01 4.54173923e-01 -8.36066306e-01
-5.75760126e-01 6.87840760e-01 5.22284023e-02 -4.03966337e-01
1.36725962e+00 -7.84513429e-02 -1.38692588e-01 6.27113581e-01
1.55539763e+00 5.39774001e-01 -6.65599287e-01 -1.49942815e-01
4.94762182e-01 -2.40353763e-01 1.63290024e-01 -1.10240006e+00
-7.03980744e-01 6.57276273e-01 6.41133904e-01 6.87095523e-01
6.75750434e-01 -1.06924668e-01 4.65933681e-01 2.76662242e-02
2.55409539e-01 -1.28855753e+00 -6.09347761e-01 5.56349993e-01
7.49239087e-01 -1.24600017e+00 -8.04055333e-02 7.71028772e-02
-3.47540885e-01 9.64233041e-01 2.94250995e-01 -2.93584287e-01
8.87002528e-01 3.32441255e-02 1.95138648e-01 1.46609977e-01
-6.77564740e-01 -9.20143723e-02 5.31608224e-01 2.90119350e-01
5.46188951e-01 2.56232679e-01 -5.57270348e-01 7.56331205e-01
-6.68347120e-01 -3.17423761e-01 1.10561498e-01 4.70063955e-01
-3.26883405e-01 -1.00506055e+00 -3.72158557e-01 4.77764249e-01
-4.52590704e-01 -2.22575858e-01 -4.96879339e-01 6.59076273e-01
1.72036037e-01 8.84956002e-01 2.70783663e-01 -9.28259671e-01
2.58082598e-01 2.21607968e-01 -1.76108882e-01 -1.38489679e-01
-6.75754011e-01 -1.69418991e-01 3.76260608e-01 -3.15136582e-01
-1.27103209e-01 -7.70339847e-01 -6.56143546e-01 -5.91381490e-01
-1.16812415e-01 3.23942423e-01 1.21048796e+00 8.70573342e-01
3.19341302e-01 1.29179567e-01 1.07679200e+00 -8.89717996e-01
-7.15540707e-01 -9.56591487e-01 -6.82049692e-01 4.41582352e-02
4.18924481e-01 -3.38434815e-01 -5.22742748e-01 -7.25392818e-01] | [9.613792419433594, 10.28873348236084] |
426e89fa-14d2-479c-ad2b-7e39aea268bf | curve-your-enthusiasm-concurvity | 2305.11475 | null | https://arxiv.org/abs/2305.11475v1 | https://arxiv.org/pdf/2305.11475v1.pdf | Curve Your Enthusiasm: Concurvity Regularization in Differentiable Generalized Additive Models | Generalized Additive Models (GAMs) have recently experienced a resurgence in popularity due to their interpretability, which arises from expressing the target value as a sum of non-linear transformations of the features. Despite the current enthusiasm for GAMs, their susceptibility to concurvity - i.e., (possibly non-linear) dependencies between the features - has hitherto been largely overlooked. Here, we demonstrate how concurvity can severly impair the interpretability of GAMs and propose a remedy: a conceptually simple, yet effective regularizer which penalizes pairwise correlations of the non-linearly transformed feature variables. This procedure is applicable to any differentiable additive model, such as Neural Additive Models or NeuralProphet, and enhances interpretability by eliminating ambiguities due to self-canceling feature contributions. We validate the effectiveness of our regularizer in experiments on synthetic as well as real-world datasets for time-series and tabular data. Our experiments show that concurvity in GAMs can be reduced without significantly compromising prediction quality, improving interpretability and reducing variance in the feature importances. | ['Martin Genzel', 'Johannes S. Otterbach', 'Maximilian Schambach', 'Alma Lindborg', 'Winfried Ripken', 'Konstantin Ditschuneit', 'Julien Siems'] | 2023-05-19 | null | null | null | null | ['additive-models'] | ['methodology'] | [ 2.94162273e-01 1.58810303e-01 2.28172481e-01 -7.24174798e-01
-4.24251646e-01 -5.76272666e-01 6.38943791e-01 -1.09105734e-02
-1.96499228e-01 7.58031249e-01 2.77857989e-01 -3.01173031e-01
-5.56451082e-01 -6.81193769e-01 -5.11117220e-01 -6.65070415e-01
-2.02608705e-01 2.94662654e-01 -3.13602805e-01 -3.05348575e-01
3.87946337e-01 4.00353260e-02 -1.59656954e+00 7.40330666e-02
1.21822274e+00 7.24311352e-01 -3.00429612e-01 3.58636171e-01
2.45560482e-01 9.29490864e-01 -5.83288848e-01 -8.08517158e-01
4.12057430e-01 -3.55332226e-01 -2.79044420e-01 5.82704432e-02
3.33403021e-01 -2.67731249e-01 -1.45674393e-01 9.44758952e-01
1.56264126e-01 2.99576461e-01 1.00721991e+00 -1.64206731e+00
-9.49517488e-01 5.80522120e-01 -7.04710960e-01 -1.98368251e-01
2.73618139e-02 2.60700136e-01 1.45634413e+00 -8.00840378e-01
-9.12209228e-02 1.40026176e+00 1.04140246e+00 2.61817068e-01
-1.65243876e+00 -4.64097589e-01 1.62448227e-01 6.63834438e-02
-1.36803424e+00 -2.30402082e-01 6.25610650e-01 -5.30201018e-01
9.38425303e-01 7.12605059e-01 3.06023210e-01 5.82140267e-01
2.92218000e-01 5.91531754e-01 1.00355291e+00 -3.91522527e-01
1.91295952e-01 8.77996907e-02 3.59670162e-01 5.25348485e-01
4.27660018e-01 2.07764015e-01 -5.20171583e-01 -5.16922712e-01
7.68231809e-01 3.14391702e-02 -1.37309015e-01 -2.87588984e-01
-9.19407904e-01 1.08053410e+00 3.27222735e-01 -3.80680934e-02
-4.92908180e-01 1.76710963e-01 1.94466367e-01 4.24446285e-01
7.83864141e-01 8.41988683e-01 -5.84350705e-01 -3.38893980e-01
-5.40992618e-01 4.22674686e-01 7.45948672e-01 5.45884192e-01
5.88833988e-01 2.29605004e-01 -1.32819200e-02 1.01298153e+00
3.81155849e-01 2.54172355e-01 5.23213804e-01 -9.49677765e-01
1.91227794e-01 7.32141674e-01 2.26188317e-01 -1.47734606e+00
-5.91811717e-01 -4.20787275e-01 -8.67696583e-01 5.29152691e-01
7.73564398e-01 -2.40178481e-01 -6.81399107e-01 2.16864920e+00
7.76313618e-02 2.22968776e-02 -3.41473877e-01 9.22616184e-01
3.78222466e-01 3.25653046e-01 2.94680715e-01 -9.63718295e-02
1.14335358e+00 -6.82753086e-01 -4.76669818e-01 -2.66519845e-01
4.65863913e-01 -6.32192254e-01 1.27671945e+00 3.72281551e-01
-1.04188454e+00 -1.50174975e-01 -9.27187145e-01 -1.77246571e-01
-2.55143523e-01 -1.26821339e-01 9.38448608e-01 8.59087288e-01
-8.90325189e-01 9.26569462e-01 -8.23954046e-01 9.61382091e-02
2.16182634e-01 6.56279683e-01 -3.09718192e-01 3.13885033e-01
-1.17239547e+00 1.04375577e+00 -1.21412277e-02 1.28816038e-01
2.22428739e-01 -1.14183271e+00 -6.99228346e-01 3.76259983e-01
3.82010609e-01 -9.18212354e-01 1.18424702e+00 -1.33596182e+00
-1.43242502e+00 3.26774776e-01 -3.47702093e-02 -4.69027638e-01
5.60495138e-01 -2.90102988e-01 -1.88008755e-01 -4.76321608e-01
-7.01235309e-02 4.90532517e-01 8.78179193e-01 -1.00300634e+00
-3.56687099e-01 -3.52643013e-01 1.65980488e-01 3.86757463e-01
-3.85948598e-01 -2.06731886e-01 1.13059357e-01 -9.63290572e-01
2.08184287e-01 -9.85756516e-01 -5.22694886e-01 1.45769835e-01
-5.59675217e-01 6.99179024e-02 2.99863160e-01 -6.80747449e-01
1.39854372e+00 -2.06111550e+00 -5.80488443e-02 2.69357741e-01
4.38883573e-01 -9.77636650e-02 -1.77379683e-01 3.91409278e-01
-4.02994365e-01 3.00818652e-01 -4.45802033e-01 -3.73479396e-01
4.49099243e-01 2.05846146e-01 -2.54150808e-01 2.90337831e-01
4.80310500e-01 8.31744432e-01 -8.27651203e-01 2.44298354e-02
3.37309897e-01 6.19960546e-01 -8.46569061e-01 -8.52346644e-02
8.47661495e-02 -1.96637027e-02 -1.41542763e-01 2.31359571e-01
6.08463764e-01 -3.61180782e-01 1.60510391e-01 -1.20698381e-03
1.15621418e-01 5.67753494e-01 -1.20134664e+00 7.05766380e-01
-3.83487284e-01 5.88088095e-01 -3.13880503e-01 -9.88931715e-01
7.67241597e-01 -6.29216209e-02 6.83133304e-01 -4.18193698e-01
-4.78145108e-02 8.27823058e-02 2.48780027e-01 -1.09612457e-01
7.40557611e-01 -3.30108255e-01 -2.23214552e-02 5.53213418e-01
-3.01374763e-01 -1.64905954e-02 5.95705584e-02 -7.05068633e-02
9.14257407e-01 -6.47298917e-02 7.06555188e-01 -5.14198542e-01
8.66132528e-02 -1.61033556e-01 4.31785494e-01 8.03047299e-01
1.11517921e-01 7.07298458e-01 7.06253469e-01 -4.43916708e-01
-9.97789502e-01 -1.04646790e+00 -1.83596030e-01 1.07169104e+00
-1.98341221e-01 -4.92140263e-01 -6.12610579e-01 -4.30176049e-01
4.88446444e-01 1.12875080e+00 -8.93349469e-01 -5.41643798e-01
-3.31313610e-01 -1.41480207e+00 2.75293171e-01 5.57255924e-01
1.42812684e-01 -5.15918255e-01 -5.75807929e-01 2.70135224e-01
5.89102879e-02 -4.86409307e-01 -5.74950993e-01 3.54887605e-01
-7.43357062e-01 -9.76821601e-01 -3.28690857e-01 -3.89215983e-02
7.97145665e-01 3.44707012e-01 1.27424335e+00 -4.43784744e-02
-5.62120005e-02 2.50975460e-01 1.36193493e-02 -4.26783979e-01
-2.31050670e-01 -3.28357458e-01 1.01210754e-02 1.16056502e-01
5.83770037e-01 -8.94941747e-01 -5.49732387e-01 3.94686043e-01
-7.49634504e-01 1.92985177e-01 3.13130498e-01 1.15582883e+00
1.26502186e-01 -5.65084256e-02 6.40663922e-01 -9.85185027e-01
1.04659665e+00 -5.42527199e-01 -4.76550430e-01 -1.78731158e-02
-9.78823125e-01 3.03929985e-01 6.33124709e-01 -5.63102424e-01
-9.59079921e-01 -9.64936763e-02 3.01165372e-01 -6.33477569e-02
2.13609487e-01 8.26140344e-01 -1.39646009e-02 -4.50534374e-02
7.26461530e-01 -1.49581268e-01 1.27838820e-01 -4.57838684e-01
4.83097821e-01 4.95298028e-01 4.59513187e-01 -5.47033429e-01
7.58236527e-01 1.21013060e-01 2.26796955e-01 -7.04871118e-01
-4.97159034e-01 -4.39131483e-02 -3.12404305e-01 7.11718500e-02
3.85658920e-01 -7.38116562e-01 -9.05706882e-01 3.34675223e-01
-8.91530156e-01 -7.07317218e-02 -3.84627789e-01 3.78762543e-01
-5.41889012e-01 3.27305168e-01 -4.39375877e-01 -1.02251053e+00
-1.81573272e-01 -9.63102341e-01 6.90803647e-01 1.20259754e-01
-8.68227601e-01 -1.22700250e+00 -7.65556395e-02 9.49246511e-02
5.64875245e-01 2.71179140e-01 1.27072835e+00 -6.98487222e-01
-2.59199142e-01 -2.96654344e-01 -3.29620332e-01 3.33997875e-01
2.77212560e-01 3.02002788e-01 -8.79482567e-01 1.49631828e-01
-7.24783093e-02 1.32287502e-01 8.02067220e-01 6.01557314e-01
8.86041164e-01 -7.05357969e-01 1.86366811e-01 3.48103374e-01
1.08155012e+00 3.70243460e-01 4.16100264e-01 1.50922060e-01
6.38173640e-01 6.01276994e-01 2.69916922e-01 7.36547410e-01
4.07381922e-01 8.13882947e-01 3.71439666e-01 -1.43704221e-01
3.72308582e-01 -2.42956549e-01 2.03070715e-01 4.37434196e-01
-3.16443890e-01 -2.06222117e-01 -7.27254331e-01 2.48676971e-01
-2.11871982e+00 -9.30622280e-01 -4.82156932e-01 2.46174216e+00
7.81555057e-01 -7.24212453e-02 3.17337126e-01 1.35694236e-01
5.11888325e-01 -4.16926742e-02 -7.06399083e-01 -5.23013592e-01
-1.09452099e-01 7.15141743e-02 4.60380346e-01 6.72013521e-01
-1.19995379e+00 3.74083370e-01 7.15564394e+00 6.49378002e-01
-9.87815082e-01 -1.80743113e-01 1.04183435e+00 -3.00432742e-01
-3.78040016e-01 -1.90833673e-01 -7.49437138e-02 5.55571735e-01
7.25143969e-01 -8.24051917e-01 7.33986616e-01 1.03178871e+00
3.78479570e-01 5.05783297e-02 -1.22733271e+00 8.81464541e-01
-1.40373781e-01 -1.00590897e+00 -6.80769086e-02 4.92514521e-02
6.10111713e-01 -2.92251676e-01 4.78358537e-01 4.07576501e-01
6.74246311e-01 -1.24676740e+00 7.86906600e-01 4.28909600e-01
3.57167989e-01 -8.88747573e-01 4.96915758e-01 2.26204783e-01
-7.70620406e-01 -1.71860069e-01 -3.08462113e-01 -6.57741308e-01
-8.26270133e-02 7.93069839e-01 -7.82884002e-01 1.83239758e-01
1.56058609e-01 5.39367199e-01 -4.51905340e-01 9.32812870e-01
3.21560279e-02 5.50015807e-01 -5.59222460e-01 -1.61778554e-01
6.74112961e-02 -4.31369066e-01 4.73183364e-01 9.30257440e-01
3.48373562e-01 2.58227229e-01 -2.17817426e-01 1.19730318e+00
2.74123341e-01 1.97370604e-01 -3.31767499e-01 -6.98209628e-02
2.77519226e-01 9.95909154e-01 -4.53548312e-01 -1.82417065e-01
-4.83842522e-01 6.42370582e-01 1.26064211e-01 4.26438600e-01
-9.07675207e-01 -8.10791701e-02 1.13054907e+00 2.13242099e-01
2.17881054e-02 -3.76427770e-02 -1.05917418e+00 -1.10257363e+00
1.15925603e-01 -9.79406834e-01 2.53915906e-01 -7.35048711e-01
-1.70977640e+00 4.65953141e-01 8.95630047e-02 -1.26559305e+00
-5.65429509e-01 -6.44217908e-01 -5.64287126e-01 1.04489636e+00
-9.24598455e-01 -8.83461177e-01 -7.00393468e-02 2.76245177e-01
2.28335455e-01 7.38273114e-02 8.15982044e-01 1.97611332e-01
-3.56706887e-01 7.32782841e-01 2.17385948e-01 -3.94447505e-01
5.89118481e-01 -1.46812332e+00 5.36396205e-01 4.92348850e-01
8.18880796e-02 1.00914299e+00 1.20572996e+00 -3.79250020e-01
-1.04413652e+00 -7.95480311e-01 8.69355261e-01 -4.16033268e-01
9.26704943e-01 -2.40171552e-01 -9.80934381e-01 6.69596791e-01
-1.89765289e-01 -3.52834463e-01 7.50751317e-01 5.13832867e-01
-5.79689562e-01 -7.35413143e-03 -1.17632866e+00 9.72861588e-01
8.12421679e-01 -3.43483686e-01 -3.90470058e-01 1.92464456e-01
4.09793764e-01 -1.73135728e-01 -7.26072490e-01 2.62465954e-01
7.88093865e-01 -1.04495704e+00 1.02264941e+00 -8.62243593e-01
6.98759973e-01 -1.24644183e-01 -1.54445171e-01 -1.63106871e+00
-5.51951945e-01 -8.04507852e-01 1.17578708e-01 1.00505471e+00
4.85230535e-01 -8.98766577e-01 6.38803363e-01 1.58877683e+00
1.66219193e-02 -7.73721397e-01 -8.83285582e-01 -7.74028957e-01
1.22633219e-01 -4.42076594e-01 7.65646040e-01 1.17085814e+00
5.98182082e-01 3.90826046e-01 -6.08100533e-01 -9.88715515e-02
4.47090060e-01 -4.56671529e-02 7.79440463e-01 -1.29814649e+00
-7.28502035e-01 -8.46673310e-01 -5.70221543e-01 -9.09060776e-01
-2.34421909e-01 -5.11181056e-01 -6.37042671e-02 -8.61311018e-01
1.74632385e-01 -4.93695021e-01 -2.69212961e-01 8.48611832e-01
-6.90919757e-01 1.95445374e-01 3.40713143e-01 -2.29973383e-02
-1.74678251e-01 4.74043012e-01 9.22331631e-01 8.19699913e-02
-4.08608139e-01 1.22271575e-01 -1.24254668e+00 9.60813999e-01
6.43716872e-01 -4.16727602e-01 -7.09197223e-01 -2.08695248e-01
4.35781032e-01 -2.71085836e-02 5.06721616e-01 -6.27823412e-01
-1.82194605e-01 -3.82536352e-01 1.89858526e-01 -2.47142278e-02
5.27494550e-01 -7.35992849e-01 6.57084465e-01 1.48494765e-01
-5.56039035e-01 2.78828800e-01 2.14548633e-01 3.44705850e-01
-3.91303934e-02 -8.21460113e-02 4.58496302e-01 1.36813745e-01
-1.60041466e-01 -1.19999021e-01 -2.55818546e-01 -1.35859653e-01
8.53669345e-01 -1.23944432e-01 -4.30366427e-01 -8.54746699e-01
-5.73361993e-01 -5.60764894e-02 3.64426404e-01 4.01483446e-01
3.06331873e-01 -1.26202941e+00 -7.50148594e-01 3.57725322e-01
-2.17340708e-01 -3.88216734e-01 3.79428774e-01 1.00446522e+00
-1.25071198e-01 1.91797048e-01 -9.77165848e-02 -2.43732780e-01
-1.30661368e+00 4.33297902e-01 2.76644021e-01 -3.99072707e-01
-2.66783923e-01 6.20692492e-01 5.41985989e-01 -2.84953743e-01
-7.17290714e-02 -4.98745114e-01 9.31569636e-02 -1.97385445e-01
5.92715919e-01 4.61166888e-01 -6.68889377e-03 -4.24383938e-01
-2.99087316e-01 1.36779115e-01 -1.78966895e-01 -1.64335296e-02
1.55659854e+00 4.89160269e-02 -6.05215225e-03 5.60245872e-01
9.56714153e-01 -6.15176857e-02 -1.41107070e+00 1.40186310e-01
-3.84677313e-02 -6.21521294e-01 2.70334482e-02 -9.25851524e-01
-9.34842765e-01 6.52324498e-01 1.33175194e-01 7.15265632e-01
1.08802235e+00 -5.43517113e-01 2.90726990e-01 3.46098155e-01
2.57917732e-01 -6.57496274e-01 -2.88016558e-01 2.98394173e-01
9.37107444e-01 -1.04854226e+00 1.51747316e-01 -5.94773173e-01
-8.49690735e-01 9.79640722e-01 3.08204412e-01 -2.76775658e-01
7.01560020e-01 3.20045382e-01 -5.22144027e-02 1.61905482e-01
-9.16235209e-01 2.16872796e-01 3.41529131e-01 6.36412621e-01
5.83450496e-01 2.78369099e-01 -5.36322653e-01 1.03566825e+00
-4.73946959e-01 -1.36102065e-01 5.34420133e-01 4.14329976e-01
-1.01424500e-01 -9.36521113e-01 -4.13462371e-01 7.10750759e-01
-3.46665949e-01 -4.72495198e-01 -4.06725198e-01 1.01465952e+00
-1.71679437e-01 9.52689350e-01 2.26499513e-01 -6.06473267e-01
3.06851000e-01 3.51436511e-02 1.59452379e-01 -1.74298614e-01
-6.88957870e-01 -1.18888468e-02 1.79536581e-01 -4.48840022e-01
-5.24779521e-02 -9.42961395e-01 -9.11157787e-01 -7.78377950e-01
-3.28169316e-01 4.84011509e-03 4.24307883e-01 8.76213312e-01
6.01510584e-01 4.52403635e-01 5.70055902e-01 -6.73382163e-01
-1.12109160e+00 -8.45520020e-01 -6.21387303e-01 6.93090677e-01
1.84276104e-01 -8.96373034e-01 -6.86587214e-01 3.57495509e-02] | [8.683663368225098, 5.4773406982421875] |
2e256162-44d7-4c44-b9d9-2f2df346b0ce | vitaa-visual-textual-attributes-alignment-in | 2005.07327 | null | https://arxiv.org/abs/2005.07327v2 | https://arxiv.org/pdf/2005.07327v2.pdf | ViTAA: Visual-Textual Attributes Alignment in Person Search by Natural Language | Person search by natural language aims at retrieving a specific person in a large-scale image pool that matches the given textual descriptions. While most of the current methods treat the task as a holistic visual and textual feature matching one, we approach it from an attribute-aligning perspective that allows grounding specific attribute phrases to the corresponding visual regions. We achieve success as well as the performance boosting by a robust feature learning that the referred identity can be accurately bundled by multiple attribute visual cues. To be concrete, our Visual-Textual Attribute Alignment model (dubbed as ViTAA) learns to disentangle the feature space of a person into subspaces corresponding to attributes using a light auxiliary attribute segmentation computing branch. It then aligns these visual features with the textual attributes parsed from the sentences by using a novel contrastive learning loss. Upon that, we validate our ViTAA framework through extensive experiments on tasks of person search by natural language and by attribute-phrase queries, on which our system achieves state-of-the-art performances. Code will be publicly available upon publication. | ['Jun Wang', 'Zhiyuan Fang', 'Zhe Wang', 'Yezhou Yang'] | 2020-05-15 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1593_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123570392.pdf | eccv-2020-8 | ['nlp-based-person-retrival', 'person-search'] | ['computer-vision', 'computer-vision'] | [ 3.00442457e-01 -5.35261407e-02 -3.47003073e-01 -8.34729671e-01
-1.24085462e+00 -8.56528759e-01 1.05376482e+00 1.98962405e-01
-6.89145088e-01 3.54604900e-01 3.17242056e-01 2.27842659e-01
-7.52358511e-02 -4.07636136e-01 -5.92135608e-01 -5.46293974e-01
3.41871649e-01 1.23940742e+00 -2.87705421e-01 1.08270101e-01
3.06465905e-02 2.16586232e-01 -1.64823294e+00 4.70455110e-01
5.62830150e-01 1.16528070e+00 -2.58971125e-01 3.76240015e-01
-1.11249611e-01 3.93623680e-01 -2.91292280e-01 -9.92599726e-01
4.34051275e-01 -3.07864130e-01 -1.15068436e+00 5.06113827e-01
1.21445477e+00 -1.01966344e-01 -2.74810612e-01 9.35395420e-01
5.14601648e-01 5.59597276e-02 9.41404164e-01 -1.54799461e+00
-8.66582632e-01 8.30056667e-02 -5.01689017e-01 3.48038562e-02
9.26458716e-01 1.79484800e-01 1.63183284e+00 -1.23355401e+00
7.87156641e-01 1.49496365e+00 4.35631216e-01 5.93202770e-01
-1.59951270e+00 -4.34648633e-01 3.55889261e-01 3.79751474e-01
-1.69510400e+00 -3.87257993e-01 6.68051362e-01 -5.90836287e-01
7.06318974e-01 3.67174417e-01 8.32839310e-01 1.35610735e+00
-2.37157136e-01 1.14425492e+00 1.28778553e+00 -3.23188066e-01
-2.67780006e-01 3.99895310e-01 1.88465312e-01 8.25638056e-01
-4.23474833e-02 2.30365572e-03 -7.81332195e-01 -3.20553094e-01
2.96685368e-01 -5.58927581e-02 2.98737548e-02 -8.70852888e-01
-1.25077426e+00 8.83856952e-01 4.73504663e-01 -1.81789622e-01
-3.47861856e-01 9.05656293e-02 3.64056915e-01 1.41576663e-01
3.69196683e-01 3.19963276e-01 -1.14237763e-01 3.86286706e-01
-9.19941902e-01 6.59804821e-01 5.84837019e-01 1.28059292e+00
8.83226991e-01 -5.57576239e-01 -6.36624336e-01 6.89208984e-01
3.55933905e-01 8.01925600e-01 1.21724449e-01 -1.03807688e+00
4.86992389e-01 7.93391228e-01 5.66063588e-03 -7.23565400e-01
-7.72108510e-02 -2.92555779e-01 -5.09576678e-01 6.14853352e-02
5.62013328e-01 3.16159993e-01 -8.59572589e-01 1.91615105e+00
4.25247043e-01 -2.61721760e-01 -5.59209883e-02 1.12756896e+00
8.25179577e-01 2.74620116e-01 5.83011150e-01 1.55799478e-01
2.16264510e+00 -8.12664211e-01 -5.85377336e-01 -7.34336734e-01
6.79588169e-02 -5.21626949e-01 1.22220600e+00 -1.15981542e-01
-1.12324369e+00 -5.24210751e-01 -8.91040266e-01 -4.36844915e-01
-4.55035955e-01 9.48259309e-02 4.21840787e-01 5.67821264e-01
-9.89607394e-01 4.57776487e-02 -3.97273362e-01 -7.54767179e-01
4.95810360e-01 4.08024698e-01 -7.27347672e-01 -3.96569958e-03
-1.13167346e+00 8.13104272e-01 1.38399616e-01 -3.11814249e-01
-8.51186931e-01 -5.06418049e-01 -1.10933399e+00 -1.15123644e-01
5.00549853e-01 -1.50374174e+00 1.09268248e+00 -1.14190888e+00
-6.10780656e-01 1.89688909e+00 -5.45355141e-01 -4.40660685e-01
7.33560264e-01 -4.15471345e-01 -5.82325868e-02 4.31831956e-01
6.75493717e-01 8.51096332e-01 1.01429713e+00 -1.42311203e+00
-6.43494904e-01 -9.89346385e-01 -1.84029397e-02 4.92977977e-01
-3.50211293e-01 3.28676730e-01 -8.89863253e-01 -6.56069756e-01
-4.89736572e-02 -1.14866936e+00 -2.88674533e-02 3.60594809e-01
-5.51369786e-01 -5.02249897e-01 3.97136927e-01 -6.41673386e-01
7.20211864e-01 -1.96975505e+00 3.20167929e-01 2.90677935e-01
3.13670397e-01 -2.49939680e-01 -9.40290838e-02 2.46629104e-01
1.02770977e-01 7.35127181e-02 -1.93991497e-01 -7.62522519e-01
3.49354267e-01 4.48220111e-02 -3.31145644e-01 6.81373656e-01
1.87623531e-01 1.19300568e+00 -7.95742571e-01 -1.14661348e+00
-1.58185855e-01 4.27287102e-01 -3.03082675e-01 3.57702076e-01
-9.19376314e-02 3.67568403e-01 -7.00453103e-01 9.74496841e-01
2.73676276e-01 -3.20460856e-01 -1.87516570e-01 -2.73074061e-01
3.01576525e-01 -2.11339608e-01 -8.65404308e-01 1.84885371e+00
-1.22281713e-02 5.36927640e-01 2.46724505e-02 -7.15608478e-01
7.75294900e-01 2.00989664e-01 5.38487077e-01 -6.82305634e-01
-2.19946876e-01 1.30349761e-02 -6.36658728e-01 -4.62007433e-01
5.09291530e-01 6.91142529e-02 -6.52769566e-01 4.26472485e-01
2.32340083e-01 1.33491620e-01 2.63041258e-02 4.68633324e-01
4.68975544e-01 3.92721534e-01 3.89025182e-01 -1.64695963e-01
6.29681528e-01 1.95976153e-01 3.18449765e-01 1.05404472e+00
-4.22624350e-01 6.63748264e-01 1.13610156e-01 -6.03363931e-01
-1.43374884e+00 -1.40819252e+00 -8.45345780e-02 1.46430707e+00
3.03488433e-01 -4.64214653e-01 -8.12146962e-01 -7.40998030e-01
1.79311737e-01 4.50735211e-01 -7.79267192e-01 2.00576290e-01
-3.39407682e-01 -4.06067580e-01 5.99834800e-01 5.69097281e-01
5.50918102e-01 -1.03855419e+00 -4.03000563e-01 -2.97094822e-01
-5.93691647e-01 -1.48216641e+00 -9.11223531e-01 -1.75423250e-01
-1.20799795e-01 -9.86113071e-01 -8.94772291e-01 -1.05896235e+00
5.55761456e-01 -1.22970015e-01 1.46757591e+00 9.65176001e-02
-5.42893231e-01 1.06028390e+00 -8.67906809e-02 -3.57101232e-01
1.27129965e-02 1.46505982e-01 1.63416758e-01 3.99442911e-01
9.10063863e-01 -1.81980580e-01 -7.70467639e-01 2.64888972e-01
-4.03320163e-01 -1.03045247e-01 6.67344391e-01 7.01903462e-01
1.10805047e+00 -7.91916132e-01 3.73664498e-02 -7.28675187e-01
4.59135920e-01 -2.11239502e-01 -3.07062268e-01 5.76173425e-01
-7.42967069e-01 2.45927870e-01 1.40309274e-01 -3.28825951e-01
-9.15632248e-01 6.84876919e-01 2.80266434e-01 -3.81946594e-01
-4.72032338e-01 -6.08455092e-02 -3.49269480e-01 2.00740904e-01
4.10571665e-01 5.34080565e-01 -1.05308853e-01 -2.26039872e-01
7.49316931e-01 5.36788344e-01 1.02415371e+00 -8.23502243e-01
1.04175413e+00 6.72891140e-01 3.38224880e-02 -5.71974337e-01
-1.18699193e+00 -8.99971664e-01 -9.69161391e-01 -2.76202857e-01
1.39949322e+00 -1.17947257e+00 -9.56001759e-01 1.10945739e-01
-1.12047780e+00 2.31773689e-01 -1.44344717e-01 1.27801150e-01
-9.78093326e-01 4.27871406e-01 -3.05396974e-01 -7.69772172e-01
-4.74629253e-01 -1.02716899e+00 1.55063558e+00 1.05732724e-01
-5.36761642e-01 -6.89479828e-01 2.12324262e-02 8.27247381e-01
-1.49225384e-01 2.58756280e-01 8.98254871e-01 -1.14403796e+00
-7.53799081e-01 -4.01796550e-01 -3.63467544e-01 -2.87210703e-01
-4.36873019e-01 -5.09323001e-01 -1.19241202e+00 -4.49691176e-01
-1.90674424e-01 -5.15099347e-01 9.98279512e-01 1.06012195e-01
9.80760157e-01 -4.99697298e-01 -7.06878781e-01 8.47689629e-01
1.39523327e+00 -4.53866363e-01 2.91116804e-01 6.08695090e-01
7.43818581e-01 8.84960532e-01 7.06089377e-01 3.57070178e-01
5.87195873e-01 1.10152245e+00 3.91648978e-01 -2.06827000e-01
-1.42829344e-01 -4.84364212e-01 -5.31966053e-02 -1.17420286e-01
-6.48161918e-02 -1.85423195e-01 -8.54195833e-01 7.03039825e-01
-1.89365363e+00 -1.24716139e+00 9.52392071e-02 2.17620468e+00
8.63365591e-01 -2.38701090e-01 5.43706417e-01 -3.41425568e-01
7.73668170e-01 1.75833374e-01 -5.44613898e-01 -5.17443977e-02
-3.44400525e-01 -3.14057350e-01 2.78196305e-01 3.96606237e-01
-1.69396651e+00 9.40608203e-01 5.81094074e+00 5.96009552e-01
-2.16543511e-01 3.50022465e-02 6.72530532e-01 -4.80429158e-02
-2.90945262e-01 3.31301615e-02 -1.12731707e+00 1.60177037e-01
3.89336318e-01 -2.91547686e-01 2.33482093e-01 9.41010535e-01
-2.16292307e-01 3.01728100e-01 -1.71581459e+00 1.25044739e+00
4.36172694e-01 -9.80438828e-01 5.80670655e-01 3.46248537e-01
3.60918790e-01 -3.90602112e-01 4.73212510e-01 1.35518402e-01
1.28139034e-01 -1.26322103e+00 9.24721777e-01 7.53686368e-01
9.64884341e-01 -5.69830894e-01 5.51096559e-01 2.61353869e-02
-1.40977001e+00 -1.27266183e-01 -5.90546466e-02 3.38838607e-01
6.36163577e-02 -1.96348801e-01 -8.00699651e-01 3.83357853e-01
7.62499452e-01 4.81274545e-01 -8.53342414e-01 9.17960823e-01
-1.85840040e-01 6.88857362e-02 9.17077884e-02 9.60917622e-02
1.31423041e-01 -2.42786005e-01 8.14489663e-01 1.36176383e+00
-7.00164139e-02 5.06899208e-02 5.81802070e-01 1.14163065e+00
-1.97879031e-01 3.34723443e-01 -7.29314446e-01 1.79296449e-01
4.96348679e-01 1.17501342e+00 -3.73758137e-01 -3.45171720e-01
-4.77513313e-01 1.37955439e+00 4.22915161e-01 5.07593989e-01
-5.50674677e-01 -1.61326811e-01 7.82707155e-01 1.57978028e-01
1.23819873e-01 1.04688272e-01 -2.67779917e-01 -1.00339508e+00
2.96049178e-01 -7.79693484e-01 8.65912378e-01 -9.23326492e-01
-1.76685166e+00 7.63496280e-01 1.10253550e-01 -1.18206584e+00
-6.34913683e-01 -4.89092648e-01 -1.02899469e-01 1.02371454e+00
-1.22019875e+00 -1.87698269e+00 -3.85089099e-01 1.00859118e+00
6.90761805e-01 -4.65402812e-01 1.04580712e+00 2.14408860e-02
-1.36586189e-01 9.27642345e-01 -3.58628482e-01 5.75172961e-01
1.06730509e+00 -1.53585517e+00 4.28309321e-01 7.60678291e-01
5.45098126e-01 6.21407092e-01 9.24918354e-01 -5.37017167e-01
-1.19238997e+00 -8.84429216e-01 1.29161286e+00 -1.07002401e+00
5.13832629e-01 -6.49334133e-01 -7.07012296e-01 8.63275588e-01
4.59979504e-01 -4.28034142e-02 6.61048651e-01 1.11848190e-01
-7.40113437e-01 3.12564038e-02 -9.35663581e-01 5.77289999e-01
1.15228117e+00 -9.89967406e-01 -6.69832289e-01 5.53163528e-01
5.42005658e-01 -1.28670558e-01 -8.39447081e-01 3.77926044e-02
5.94650388e-01 -5.88094771e-01 1.64689004e+00 -9.66862619e-01
3.07316303e-01 -1.48878947e-01 -3.45118403e-01 -8.07418168e-01
-3.81864041e-01 -3.29513043e-01 2.81316370e-01 1.45179760e+00
4.18140858e-01 -2.07171172e-01 7.54710793e-01 9.48236108e-01
3.17004859e-01 -4.68689561e-01 -9.52626765e-01 -7.08085716e-01
-1.13244541e-01 -7.15439171e-02 4.53321308e-01 7.28570104e-01
-2.39859641e-01 6.63236141e-01 -4.09507364e-01 2.63180166e-01
1.20205605e+00 4.50730234e-01 7.44611025e-01 -1.48653626e+00
-1.09691992e-01 -4.34997618e-01 -5.66007555e-01 -1.04085541e+00
6.10743880e-01 -1.11372542e+00 1.73277762e-02 -1.51396835e+00
1.05977595e+00 -8.96913931e-02 1.48711633e-02 5.34008920e-01
-3.58676702e-01 4.64780718e-01 3.27603787e-01 6.34497941e-01
-9.83927250e-01 4.64237154e-01 8.24403107e-01 -7.45002806e-01
2.71584541e-01 1.53704256e-01 -6.91576600e-01 6.06725514e-01
3.63983095e-01 -4.58964437e-01 -1.78979635e-01 -3.14139724e-01
1.05680756e-01 -2.22454146e-02 8.59714508e-01 -7.52426088e-01
4.58428383e-01 -5.50623424e-02 8.04570794e-01 -4.94874120e-01
7.56777585e-01 -8.18064213e-01 -1.69240475e-01 1.05056062e-01
-8.16433728e-01 4.48893994e-01 -2.02016473e-01 8.57629418e-01
-2.54057944e-01 2.81618740e-02 4.51058984e-01 -2.83162057e-01
-1.04437375e+00 7.04111934e-01 -1.78065281e-02 3.86777192e-01
9.50922668e-01 -4.54392970e-01 -2.40681604e-01 -6.60627067e-01
-9.67539668e-01 4.86011714e-01 7.98416734e-01 2.99552560e-01
7.40981400e-01 -1.51911080e+00 -1.05567503e+00 1.05060481e-01
7.18329728e-01 -5.69608450e-01 -2.31660809e-03 5.87088764e-01
3.54099050e-02 4.69907373e-01 -1.81324407e-01 -8.65304649e-01
-1.73185909e+00 9.75161076e-01 4.26154971e-01 -7.83317536e-02
-5.28217554e-01 8.57211649e-01 5.26372373e-01 -2.63245136e-01
3.57871085e-01 7.12330520e-01 -2.89243907e-01 1.85115591e-01
4.05618012e-01 -3.81859615e-02 -3.06924254e-01 -1.25466847e+00
-6.97924972e-01 9.26192880e-01 -6.21962249e-02 -5.33910751e-01
8.85757923e-01 -4.75088090e-01 6.63974672e-04 1.78683504e-01
1.25448096e+00 -6.14330024e-02 -1.20919216e+00 -6.99076116e-01
1.30143762e-01 -3.69963259e-01 -5.60315847e-01 -7.56198049e-01
-7.31518149e-01 7.39151835e-01 8.18341553e-01 -1.26420766e-01
9.94739294e-01 7.82936573e-01 4.89141852e-01 3.89348596e-01
7.88563341e-02 -1.04464257e+00 2.84597546e-01 1.21678837e-01
8.74341249e-01 -1.40786612e+00 1.19008511e-01 -4.05905366e-01
-9.76647258e-01 7.69276083e-01 5.74086547e-01 3.52453850e-02
1.32034659e-01 -2.54138231e-01 1.68518469e-01 -3.63685220e-01
-4.66341317e-01 -6.12805903e-01 8.80579829e-01 8.08369458e-01
1.96370080e-01 3.41272168e-02 -2.38946844e-02 5.96655846e-01
-2.76027173e-01 -6.55776620e-01 -8.28778297e-02 4.67191249e-01
-2.81111509e-01 -1.09815860e+00 -3.51751059e-01 2.21027002e-01
-4.69175398e-01 -2.00793922e-01 -9.30003524e-01 8.40256393e-01
-1.19206585e-01 8.11620235e-01 2.65183955e-01 -1.82156786e-01
3.24940264e-01 3.83251607e-01 4.41099197e-01 -3.89859855e-01
-4.83888716e-01 -9.39928070e-02 1.01936154e-01 -6.78921700e-01
-5.29434681e-01 -1.00014973e+00 -8.45001519e-01 4.17732857e-02
3.36434960e-01 5.35280444e-02 2.79612958e-01 1.12056851e+00
8.84566978e-02 -8.35214853e-02 2.33425975e-01 -4.88129556e-01
-5.86530864e-01 -5.97823739e-01 -3.82217556e-01 1.15749311e+00
3.40255946e-01 -5.63538015e-01 -5.57528399e-02 4.19856906e-01] | [14.558941841125488, 0.9375434517860413] |
34b85369-c382-4646-917b-96dc1581cd1e | hunsum-1-an-abstractive-summarization-dataset | 2302.00455 | null | https://arxiv.org/abs/2302.00455v1 | https://arxiv.org/pdf/2302.00455v1.pdf | HunSum-1: an Abstractive Summarization Dataset for Hungarian | We introduce HunSum-1: a dataset for Hungarian abstractive summarization, consisting of 1.14M news articles. The dataset is built by collecting, cleaning and deduplicating data from 9 major Hungarian news sites through CommonCrawl. Using this dataset, we build abstractive summarizer models based on huBERT and mT5. We demonstrate the value of the created dataset by performing a quantitative and qualitative analysis on the models' results. The HunSum-1 dataset, all models used in our experiments and our code are available open source. | ['Judit Ács', 'Milán Konor Nyist', 'Attila Nagy', 'Dorina Lakatos', 'Botond Barta'] | 2023-02-01 | null | null | null | null | ['abstractive-text-summarization'] | ['natural-language-processing'] | [-3.57773215e-01 3.70297462e-01 -8.79459620e-01 -1.17257744e-01
-1.37112832e+00 -6.51280820e-01 1.08035254e+00 8.40915024e-01
-4.77822721e-01 1.10668266e+00 1.54084277e+00 2.67461818e-02
-9.95583236e-02 -6.38011336e-01 -8.58346879e-01 -5.45342825e-02
8.13325867e-03 7.21368253e-01 1.01661466e-01 -4.36058253e-01
6.79407299e-01 -7.98441917e-02 -1.21817696e+00 7.84280121e-01
1.19585776e+00 6.71066940e-01 -6.16296493e-02 8.87216091e-01
-3.27274591e-01 1.21731484e+00 -1.01888931e+00 -8.73629868e-01
3.87456007e-02 -5.64924836e-01 -1.07452822e+00 -5.63406825e-01
5.62080026e-01 -2.54749745e-01 -7.08100796e-01 8.55307460e-01
6.32609546e-01 1.47397429e-01 8.55652094e-01 -9.09804702e-01
-9.31741953e-01 1.84124827e+00 -2.74951547e-01 5.00337183e-01
3.95559579e-01 -3.22375566e-01 1.39034855e+00 -5.78322589e-01
1.18959975e+00 9.44213688e-01 6.45394444e-01 3.42420071e-01
-8.58694851e-01 -3.66762221e-01 -2.34847918e-01 3.95400137e-01
-9.58021998e-01 -8.91891837e-01 6.38830960e-01 -9.95952636e-02
1.26824141e+00 7.32105136e-01 8.91184092e-01 1.32581675e+00
3.87137055e-01 1.26022792e+00 4.21245337e-01 -4.68320459e-01
3.19146752e-01 -2.68749427e-03 9.06324565e-01 4.80405092e-01
8.67423594e-01 -6.14503264e-01 -9.58576262e-01 -2.82698244e-01
-1.64533481e-01 -1.35648698e-01 -2.74380773e-01 1.62810445e-01
-1.05237567e+00 7.91201055e-01 2.73704797e-01 1.97855040e-01
-6.09692991e-01 2.21633792e-01 7.99398899e-01 4.87187356e-01
9.17001307e-01 7.86577582e-01 -3.77601802e-01 -3.82498980e-01
-1.44902647e+00 5.95461071e-01 1.13903058e+00 1.43600810e+00
2.58495092e-01 -2.12489784e-01 -5.57038605e-01 8.80004466e-01
-1.56130701e-01 7.13285565e-01 7.98841119e-01 -8.88404608e-01
8.99893522e-01 6.31618261e-01 -9.04803351e-02 -7.32047856e-01
-2.94484556e-01 -4.08299267e-01 -8.25595975e-01 -8.43876004e-01
-5.07641435e-01 -2.14087665e-01 -1.15578485e+00 1.06471670e+00
-4.13082063e-01 -4.55988467e-01 4.95587766e-01 3.73699129e-01
1.74840915e+00 1.03263867e+00 -2.55200744e-01 -5.60366035e-01
1.16438305e+00 -1.37067699e+00 -1.28634059e+00 -7.64757022e-02
7.56399751e-01 -6.14663899e-01 7.47427702e-01 2.61847168e-01
-1.31707394e+00 1.16769694e-01 -1.05867517e+00 -5.64980388e-01
-6.65347517e-01 1.64293081e-01 4.62438256e-01 2.14665502e-01
-8.71364474e-01 4.70841765e-01 -7.09108412e-01 -7.66728163e-01
4.71971184e-01 -4.64440227e-01 -3.65497857e-01 1.42029762e-01
-1.17239642e+00 9.36666131e-01 7.28355408e-01 -5.47211289e-01
-6.66707933e-01 -9.93919253e-01 -7.77189136e-01 2.80188918e-01
3.55596662e-01 -1.01068127e+00 1.65219116e+00 -7.62944221e-02
-1.32385778e+00 6.00521982e-01 -5.96960038e-02 -1.14023507e+00
4.85424817e-01 -7.00853169e-01 -2.88221240e-01 1.15538403e-01
2.35778645e-01 1.24619566e-01 1.08682826e-01 -1.10994637e+00
-6.39080226e-01 -2.85271764e-01 -4.38601583e-01 1.23843387e-01
-1.04326390e-01 1.74515560e-01 -4.41837758e-01 -1.02755570e+00
-9.39177051e-02 -4.10680979e-01 2.84952987e-02 -1.12074757e+00
-9.41338480e-01 8.24812576e-02 4.68946725e-01 -1.21419787e+00
2.03226185e+00 -1.64651752e+00 1.61088154e-01 -1.63709059e-01
2.15423107e-01 1.52949825e-01 -5.64049892e-02 1.20254290e+00
4.02701408e-01 6.52775407e-01 -3.16665858e-01 -4.49625105e-01
6.77318275e-02 4.50743660e-02 -8.34000528e-01 8.59058350e-02
-3.56979311e-01 1.17779744e+00 -9.53198791e-01 -7.61820316e-01
-4.53706272e-02 -2.71884769e-01 -5.67838609e-01 3.53703648e-02
-3.57862204e-01 -5.31473517e-01 -4.64882642e-01 6.74430430e-01
3.65894556e-01 1.65799052e-01 -8.98551121e-02 -8.60328972e-02
-3.39751750e-01 6.96611583e-01 -3.15588713e-01 2.02919245e+00
-4.57829610e-02 7.47961521e-01 -2.46028677e-01 -6.12276793e-01
7.17431903e-01 9.32435915e-02 2.44666964e-01 -6.99683905e-01
2.34987155e-01 1.85750112e-01 -6.85387969e-01 -4.70299959e-01
1.42946672e+00 4.34078544e-01 -7.20022619e-01 4.97401476e-01
4.28336322e-01 -6.46800578e-01 8.53828549e-01 9.58974063e-01
1.43821931e+00 -6.10377491e-01 7.27968931e-01 -4.26731527e-01
7.11629391e-02 6.49762273e-01 4.27340150e-01 1.26111531e+00
1.38831437e-01 7.42521882e-01 4.66932893e-01 -3.74030530e-01
-1.16537333e+00 -1.07538879e+00 5.35701215e-02 8.41688991e-01
-2.42982045e-01 -1.23606801e+00 -6.06174469e-01 -6.82241201e-01
1.12350516e-01 1.45965338e+00 -7.61877239e-01 -3.06900799e-01
-4.17942017e-01 -9.03617740e-01 9.45845127e-01 9.26135629e-02
6.81520104e-01 -1.01313233e+00 -6.09444976e-01 -6.87308051e-03
-7.25437284e-01 -6.81255579e-01 -3.83950740e-01 3.14402312e-01
-8.53981674e-01 -7.68202841e-01 -4.04548764e-01 -5.08643329e-01
-2.91365609e-02 2.24361390e-01 1.50499594e+00 -3.49832237e-01
-5.78730926e-02 2.81276494e-01 -8.85615766e-01 -8.50128949e-01
-6.46875560e-01 9.99352336e-01 -1.18568040e-01 -8.75943422e-01
2.24866271e-01 -4.05492187e-01 -2.39531249e-01 -6.30151629e-01
-9.44579005e-01 1.62667707e-01 7.30468273e-01 5.67930698e-01
4.36359972e-01 -3.69397789e-01 7.35762239e-01 -1.29320645e+00
1.21055233e+00 -7.78733611e-01 -2.12175995e-01 6.30945563e-01
-6.90761149e-01 2.78151423e-01 4.42165345e-01 1.90527782e-01
-1.08570457e+00 -6.08272791e-01 4.27866951e-02 7.23988935e-02
4.49445963e-01 1.10985792e+00 2.46558622e-01 7.85710216e-01
9.04584289e-01 2.66018569e-01 -3.17812681e-01 -9.14704859e-01
9.57140326e-01 1.00686252e+00 1.05140412e+00 -2.86520004e-01
4.04796332e-01 1.51205391e-01 -6.82500124e-01 -1.13460994e+00
-1.09953570e+00 -6.13921940e-01 -3.13400984e-01 -5.75965755e-02
5.47051966e-01 -1.09749651e+00 8.79716575e-02 2.78511047e-01
-1.08044541e+00 2.14275401e-02 -8.02772641e-01 3.27961266e-01
-6.05528295e-01 2.97943503e-01 -8.23681235e-01 -6.39005840e-01
-1.36547160e+00 -2.06569418e-01 8.52512598e-01 1.98615134e-01
-7.11505413e-01 -6.52580082e-01 6.87312901e-01 3.55553329e-01
3.63436580e-01 1.59668833e-01 1.07571578e+00 -1.16742694e+00
-1.61425155e-02 -2.65400767e-01 1.04272841e-02 1.56102955e-01
-2.05148105e-02 3.91307056e-01 -3.50439310e-01 1.25064105e-01
-1.70715109e-01 -4.15083855e-01 1.65441895e+00 7.68184841e-01
8.92372787e-01 -1.01060438e+00 -2.22842544e-01 3.07931691e-01
1.20045519e+00 -2.80814052e-01 5.79578280e-01 7.79994249e-01
4.39695746e-01 3.27868462e-01 3.75787020e-01 7.62963355e-01
5.44187188e-01 1.07522190e-01 3.39080393e-02 3.84919405e-01
-3.92690837e-01 -7.59540558e-01 3.62739891e-01 1.89191854e+00
3.68228406e-02 -6.48104429e-01 -8.23203921e-01 7.48042285e-01
-2.25663447e+00 -1.29321492e+00 -8.78211856e-02 1.84846413e+00
1.26894248e+00 5.74946553e-02 -9.42283496e-02 -1.08407438e-01
3.81332099e-01 5.15768886e-01 -2.29360715e-01 -6.96052372e-01
-6.40476584e-01 1.92888990e-01 8.04255247e-01 4.14817125e-01
-9.39651906e-01 1.22009778e+00 7.58677578e+00 8.22343767e-01
-4.58378345e-01 -1.46888405e-01 2.26356611e-01 -6.29422784e-01
-5.43529749e-01 3.75470007e-03 -5.12554884e-01 4.15325671e-01
1.36000299e+00 -1.22549903e+00 1.24100208e-01 6.78166091e-01
1.69244245e-01 -4.05546457e-01 -7.33878314e-01 6.59135222e-01
4.69831109e-01 -2.06348395e+00 6.16555691e-01 -4.55739766e-01
9.29344594e-01 3.90371859e-01 -2.83596218e-01 7.36584187e-01
9.32231128e-01 -5.97370684e-01 1.07496905e+00 7.01070249e-01
4.44766909e-01 -8.52686942e-01 1.00527835e+00 4.66867357e-01
-3.60008627e-01 1.90297049e-02 -5.77932894e-01 1.58344284e-01
1.05326712e-01 9.17576730e-01 -6.01348221e-01 1.01005030e+00
8.49034488e-01 1.01966143e+00 -8.98605883e-01 1.05181205e+00
-2.77189285e-01 8.92139196e-01 -1.32652789e-01 -2.46826902e-01
8.56972784e-02 -1.06362969e-01 7.93402314e-01 1.72595763e+00
2.92092562e-01 1.16445035e-01 -3.12352419e-01 5.67003310e-01
-7.13603914e-01 4.12796557e-01 -5.55534482e-01 -3.30309838e-01
8.26322675e-01 1.02697206e+00 -3.34463656e-01 -8.80391240e-01
9.58938375e-02 7.24695027e-01 3.56042147e-01 1.67964637e-01
-6.93231106e-01 -8.45521927e-01 3.28438729e-01 -1.67780921e-01
7.44095817e-02 5.96668087e-02 -5.82552671e-01 -1.59037483e+00
-3.71384546e-02 -8.67557108e-01 6.85381055e-01 -7.68261671e-01
-1.18862522e+00 3.63808125e-01 5.57403386e-01 -5.83963633e-01
-3.22952718e-01 2.13729590e-01 -6.61631882e-01 1.33041337e-01
-1.10636425e+00 -1.16891372e+00 -1.20526947e-01 -7.33721210e-03
8.48289371e-01 -3.85815620e-01 5.84601760e-01 1.41366944e-02
-7.34811425e-01 3.64862621e-01 7.03042507e-01 3.37610245e-02
8.85048449e-01 -1.33530521e+00 8.50113034e-01 9.26098108e-01
1.66791424e-01 6.97840512e-01 1.16673481e+00 -1.02323163e+00
-1.51726937e+00 -1.20463812e+00 1.45342565e+00 -6.96615934e-01
7.84293592e-01 -2.81633347e-01 -7.03071475e-01 8.69145513e-01
8.27121079e-01 -7.80500352e-01 6.51846111e-01 1.80033401e-01
-3.20521414e-01 -3.34334262e-02 -1.05375946e+00 5.79689741e-01
1.10809326e+00 -1.43389404e-01 -1.50314224e+00 3.78147334e-01
1.11896014e+00 -1.14250019e-01 -8.21035206e-01 2.47991920e-01
5.08506179e-01 -5.68945825e-01 4.68318820e-01 -7.69020855e-01
1.05177951e+00 1.56695679e-01 -2.48955995e-01 -1.90866673e+00
-3.22575778e-01 -5.92762530e-01 -5.63690186e-01 1.51439667e+00
8.18195343e-01 -3.99768859e-01 4.20041203e-01 2.29532883e-01
-5.16785979e-01 -1.17256530e-01 -8.56425583e-01 -7.63782084e-01
3.98848206e-01 -3.72629166e-02 6.21455610e-01 8.05581808e-01
6.24260962e-01 6.98250949e-01 -2.52444535e-01 -4.91055667e-01
4.09048080e-01 1.81357607e-01 9.69462812e-01 -1.09081602e+00
3.38950902e-01 -6.57883465e-01 4.56269123e-02 -6.87553406e-01
4.03705686e-02 -1.07414663e+00 1.50689036e-01 -2.27446938e+00
7.49537110e-01 3.46940458e-01 1.33108288e-01 4.74767238e-01
-1.66005865e-01 -3.70281100e-01 3.03706364e-03 5.23506999e-01
-9.88904059e-01 8.38258266e-01 5.55532634e-01 -3.62494707e-01
-2.16236502e-01 -3.42474759e-01 -9.87230659e-01 2.98330218e-01
9.20406401e-01 -6.06075466e-01 -2.43448943e-01 -4.83143181e-01
3.57646823e-01 -3.87316234e-02 -2.68562198e-01 -9.70831990e-01
3.81799132e-01 -6.96635768e-02 1.34703487e-01 -1.11309135e+00
-1.53099954e-01 -1.78149328e-01 2.20467955e-01 4.44004923e-01
-8.15622926e-01 3.36546659e-01 1.50054991e-01 4.39929426e-01
-4.14844394e-01 -3.01210165e-01 2.85327047e-01 -2.41842076e-01
-4.17999655e-01 -7.84602463e-02 -5.25106668e-01 5.18734813e-01
5.71768165e-01 4.92383063e-01 -1.19836545e+00 -6.42934740e-01
-2.10667998e-01 3.57618839e-01 4.50112522e-01 3.85465711e-01
4.11214709e-01 -1.10068476e+00 -1.29636502e+00 -2.12703437e-01
4.19164866e-01 -1.97781369e-01 -1.76131744e-02 4.61853296e-01
-8.78854513e-01 7.33786047e-01 -2.55189598e-01 1.19169101e-01
-9.93099749e-01 5.04607022e-01 -4.09103513e-01 -4.75317031e-01
-8.29731464e-01 3.59326422e-01 -6.36273682e-01 -2.81809509e-01
-9.90169588e-03 -7.24278808e-01 -3.88061851e-01 4.47130829e-01
6.40261769e-01 8.01226020e-01 3.68260592e-01 -2.82581389e-01
-1.57535017e-01 -2.47210443e-01 -5.05362988e-01 -5.14309824e-01
1.63615704e+00 -2.03780264e-01 -4.37644571e-01 7.58020222e-01
1.12059021e+00 4.18459624e-01 -2.12783664e-01 -1.17914461e-01
3.35490316e-01 7.09460080e-02 1.59796491e-01 -9.76113141e-01
-6.54864192e-01 9.47011337e-02 -3.63125831e-01 2.67971426e-01
7.80512512e-01 1.36222824e-01 8.54656458e-01 7.95696676e-01
7.13423714e-02 -1.44703209e+00 -3.63977432e-01 9.37259257e-01
1.13986659e+00 -9.80280936e-01 6.28816783e-01 1.07261360e-01
-7.34726787e-01 8.56586397e-01 4.54538129e-02 -3.52888890e-02
4.05216753e-01 1.17213622e-01 -6.11799099e-02 -4.30704504e-01
-1.22912645e+00 4.86297943e-02 2.57428378e-01 1.97714150e-01
4.21602041e-01 2.18856484e-01 -8.99518788e-01 1.23474360e+00
-1.06303561e+00 -1.66110933e-01 1.09077084e+00 1.14151525e+00
-7.73647666e-01 -5.90668201e-01 8.64786580e-02 9.46752191e-01
-6.14413679e-01 -2.89132565e-01 -1.11538696e+00 8.50456238e-01
-7.09202707e-01 9.97916818e-01 -2.67465729e-02 -4.05492663e-01
5.93626022e-01 7.37216920e-02 1.75910726e-01 -6.41141951e-01
-8.25694621e-01 -3.65913749e-01 6.12353504e-01 -1.26353592e-01
-7.89502189e-02 -7.14741230e-01 -8.72976065e-01 -8.62817228e-01
-1.04010120e-01 7.04936445e-01 7.65099764e-01 4.01103258e-01
6.19971812e-01 4.94489402e-01 3.35674852e-01 -3.32829386e-01
-5.28482199e-01 -1.45617831e+00 -4.75882500e-01 1.65619224e-01
2.56565183e-01 -1.77384657e-03 -5.52653551e-01 1.82987481e-01] | [12.484601020812988, 9.493569374084473] |
71a79836-b7b7-47b2-825c-a4c217c7621b | a-fast-and-accurate-vietnamese-word-segmenter | 1709.06307 | null | http://arxiv.org/abs/1709.06307v2 | http://arxiv.org/pdf/1709.06307v2.pdf | A Fast and Accurate Vietnamese Word Segmenter | We propose a novel approach to Vietnamese word segmentation. Our approach is
based on the Single Classification Ripple Down Rules methodology (Compton and
Jansen, 1990), where rules are stored in an exception structure and new rules
are only added to correct segmentation errors given by existing rules.
Experimental results on the benchmark Vietnamese treebank show that our
approach outperforms previous state-of-the-art approaches JVnSegmenter,
vnTokenizer, DongDu and UETsegmenter in terms of both accuracy and performance
speed. Our code is open-source and available at:
https://github.com/datquocnguyen/RDRsegmenter. | ['Mark Johnson', 'Mark Dras', 'Thanh Vu', 'Dat Quoc Nguyen', 'Dai Quoc Nguyen'] | 2017-09-19 | a-fast-and-accurate-vietnamese-word-segmenter-2 | https://aclanthology.org/L18-1410 | https://aclanthology.org/L18-1410.pdf | lrec-2018-5 | ['vietnamese-word-segmentation'] | ['natural-language-processing'] | [-2.88216680e-01 3.01138729e-01 -7.09512591e-01 -4.69465643e-01
-7.38947928e-01 -9.58016574e-01 3.71059865e-01 1.17034182e-01
-7.01906264e-01 1.15746629e+00 1.77664161e-01 -6.28411531e-01
3.04004312e-01 -5.70023775e-01 -3.47060561e-01 -1.46459401e-01
4.38186705e-01 7.17754722e-01 7.10857093e-01 -5.71543515e-01
2.77977496e-01 1.39393151e-01 -1.01728892e+00 3.78916800e-01
1.07025945e+00 2.47662082e-01 3.99324410e-02 6.38190746e-01
-1.51966766e-01 7.37403512e-01 -6.42671168e-01 -9.89957869e-01
1.65239260e-01 -4.65110838e-01 -1.15625644e+00 -4.21792507e-01
-1.20584369e-01 -1.17797606e-01 -1.06906250e-01 1.11860490e+00
2.36865804e-01 1.41462982e-01 3.94831687e-01 -7.43699551e-01
-5.62730432e-01 1.47067583e+00 -3.15847307e-01 5.42263627e-01
4.09517884e-01 -1.98081166e-01 1.29053855e+00 -1.02490008e+00
1.17079926e+00 1.06922233e+00 5.19323766e-01 8.10557842e-01
-8.90055120e-01 -5.14781654e-01 4.36975539e-01 4.15732414e-01
-1.28392518e+00 -5.99123776e-01 3.48802179e-01 -1.03625085e-03
1.53652906e+00 5.18381298e-01 5.88187933e-01 8.59360635e-01
9.05201733e-02 1.09762979e+00 9.39005315e-01 -7.36369371e-01
1.03785537e-01 -9.39016491e-02 6.44284964e-01 6.71488583e-01
6.18750393e-01 1.20912090e-01 -3.25514495e-01 4.67925966e-02
4.69440609e-01 -7.36661673e-01 -6.40934780e-02 3.48404408e-01
-1.07396853e+00 1.08653879e+00 2.99214981e-02 7.02051580e-01
-1.50418445e-01 -1.87975943e-01 7.03152061e-01 2.66350597e-01
4.01546091e-01 2.67142028e-01 -9.90473092e-01 -3.27765346e-01
-8.18591416e-01 5.46997964e-01 1.05952334e+00 1.29509974e+00
3.74631196e-01 2.59195089e-01 1.29594296e-01 9.13985670e-01
4.25893486e-01 1.55984372e-01 6.59588575e-01 -1.04713583e+00
6.31464720e-01 4.05414492e-01 -1.67968720e-02 -2.97219992e-01
-1.60792083e-01 1.61649317e-01 -1.52855426e-01 -1.56712830e-01
6.83304608e-01 -2.47682646e-01 -1.11648810e+00 1.31116545e+00
4.90839213e-01 -5.46062708e-01 3.45517784e-01 7.11978912e-01
8.75185609e-01 9.29018795e-01 2.55654126e-01 -4.08970922e-01
1.46352673e+00 -9.66198444e-01 -1.12437379e+00 -3.34110528e-01
9.44194198e-01 -1.03188550e+00 9.66587126e-01 6.57481313e-01
-1.29751611e+00 -3.33966166e-01 -9.40811813e-01 -3.56504887e-01
-6.83666170e-01 1.78561546e-02 8.15965772e-01 9.79165018e-01
-6.13737524e-01 4.25902516e-01 -9.33881819e-01 -4.08902675e-01
9.64871570e-02 2.33570457e-01 -1.07633844e-01 1.08447812e-01
-1.23913145e+00 8.92371655e-01 1.24894750e+00 3.13575804e-01
-3.47019136e-01 -4.83694851e-01 -9.60378349e-01 -4.40665215e-01
7.33047545e-01 -8.83318186e-02 1.84573770e+00 -8.16032887e-01
-1.72389996e+00 1.05272400e+00 -2.69629478e-01 -3.95539135e-01
6.16273284e-01 -3.90301049e-01 -6.54300392e-01 -1.48736879e-01
3.62718850e-01 3.35453272e-01 9.13178027e-02 -1.21238136e+00
-1.00421739e+00 -3.25206727e-01 -3.33164155e-01 1.62415862e-01
3.54505450e-01 7.11383700e-01 -6.63906813e-01 -1.01436973e+00
1.16315149e-01 -7.34273851e-01 -6.75734580e-02 -1.01421762e+00
-6.14793956e-01 -5.75544834e-01 6.36064649e-01 -1.29727173e+00
1.85005593e+00 -1.70760000e+00 -8.02434795e-03 3.29349130e-01
-2.16250837e-01 6.98346972e-01 2.23386884e-01 5.27744591e-01
-5.39070331e-02 5.33011913e-01 -5.86457968e-01 1.62387863e-01
1.45566255e-01 7.69865453e-01 8.39376003e-02 2.01777741e-01
-9.21559334e-02 9.64865804e-01 -9.22580600e-01 -6.37778223e-01
3.82027626e-02 -3.12704816e-02 -4.25347924e-01 8.98299180e-03
-3.91771317e-01 -8.02244693e-02 -2.51813293e-01 1.09631991e+00
5.07667840e-01 5.94316304e-01 9.51955140e-01 2.97246993e-01
-6.26863360e-01 7.99311221e-01 -1.17646205e+00 1.51302230e+00
-9.74654779e-02 8.09591040e-02 -2.19467133e-02 -7.47126937e-01
7.39193976e-01 6.15996778e-01 -2.25112393e-01 -5.45559227e-01
3.40973169e-01 8.34741473e-01 3.26685011e-01 -3.49143565e-01
7.99942911e-01 2.20402822e-01 -5.48709869e-01 5.92584796e-02
1.15999766e-01 -2.21280441e-01 1.01857996e+00 2.98265725e-01
5.96190810e-01 5.61300814e-01 1.01173019e+00 -4.12645549e-01
7.04338372e-01 6.42991066e-01 1.20550263e+00 5.44252872e-01
-4.54857945e-01 2.43643686e-01 7.68326461e-01 -2.75048077e-01
-1.22633910e+00 -9.75419939e-01 -8.37429389e-02 1.29542685e+00
-3.98739636e-01 -5.65366268e-01 -1.07903254e+00 -1.17305243e+00
-3.92193228e-01 1.16814625e+00 -3.17507118e-01 6.11186981e-01
-1.42796922e+00 -5.46910048e-01 8.59183252e-01 5.71907103e-01
4.70461696e-01 -1.41074347e+00 -5.92949629e-01 5.77555537e-01
-1.13450766e-01 -1.25750172e+00 -6.79066241e-01 1.13764308e-01
-6.88791215e-01 -1.14295030e+00 -3.54138643e-01 -1.17210209e+00
2.74259567e-01 -3.96337628e-01 1.06639624e+00 3.47925216e-01
-3.88735719e-02 -2.71140546e-01 -8.55980754e-01 -3.85524571e-01
-6.34532869e-01 4.45493162e-01 -3.84212792e-01 -6.18585229e-01
6.16110861e-01 1.57865047e-01 -1.10814184e-01 -9.40970406e-02
-5.86853027e-01 -3.61399978e-01 1.47615418e-01 8.65734041e-01
6.70664549e-01 -3.72298658e-01 5.79453826e-01 -1.74418533e+00
6.36646390e-01 -4.89661098e-01 -8.58682096e-01 2.96913892e-01
-7.60594666e-01 -2.22678706e-01 5.98994672e-01 -5.75252771e-02
-1.52196383e+00 -9.95557848e-03 -7.60944545e-01 2.57509947e-01
-3.94515693e-01 6.06299758e-01 -4.62468565e-01 2.43404418e-01
2.54003316e-01 3.53329140e-03 -5.31583071e-01 -5.03898025e-01
6.00801647e-01 7.16844916e-01 5.72044909e-01 -6.58056974e-01
2.62201011e-01 6.65370896e-02 -7.96217322e-01 -7.56672084e-01
-4.81970549e-01 -4.60349232e-01 -9.43754911e-01 -1.28270552e-01
9.24342811e-01 -6.83978021e-01 -1.92639634e-01 4.32577640e-01
-1.45399988e+00 -4.82524723e-01 -3.54418695e-01 2.40048081e-01
-2.94964492e-01 4.94792759e-01 -1.05406260e+00 -5.91314912e-01
-4.59761918e-01 -8.55975151e-01 4.62984979e-01 2.89831519e-01
-5.54869413e-01 -1.12809312e+00 1.24619424e-01 3.36811185e-01
-1.76734433e-01 1.08403236e-01 1.01449025e+00 -1.11044812e+00
-6.51829466e-02 7.84458872e-03 9.57010314e-02 2.92973310e-01
-2.36716628e-01 4.25870150e-01 -4.00725394e-01 1.54124260e-01
-3.00910264e-01 -8.27587619e-02 7.36567914e-01 2.09482506e-01
4.63953108e-01 -6.44695938e-01 -2.79834449e-01 3.50144655e-01
1.51414573e+00 7.57450283e-01 8.08056533e-01 6.49075568e-01
4.75232065e-01 6.03369653e-01 9.69553411e-01 2.28886873e-01
7.34105825e-01 4.76024926e-01 3.37870233e-03 1.82527766e-01
-2.46853039e-01 -2.56681234e-01 3.82416368e-01 1.28141880e+00
-4.03881639e-01 -4.90903229e-01 -1.22901082e+00 8.95113230e-01
-2.03835940e+00 -8.91772509e-01 -6.94617927e-01 1.89231622e+00
1.04238653e+00 1.63468674e-01 2.77927727e-01 3.90825450e-01
7.67869294e-01 6.53597414e-02 -2.43093856e-02 -1.35985792e+00
-1.53294265e-01 6.40131533e-01 5.85337341e-01 8.84218037e-01
-9.98007715e-01 2.08684492e+00 6.32059145e+00 7.83070803e-01
-5.50968647e-01 5.58392227e-01 3.08763057e-01 3.70400667e-01
-2.55312264e-01 2.47552216e-01 -1.16636610e+00 3.96633863e-01
8.92014265e-01 -8.44017714e-02 2.42525354e-01 5.43960571e-01
5.38857915e-02 -1.90936729e-01 -5.82442939e-01 4.46583241e-01
4.87824529e-03 -1.29095244e+00 -1.50463641e-01 -2.14151546e-01
7.04479039e-01 8.14796612e-02 -3.94967228e-01 3.95006120e-01
6.93940163e-01 -3.72698188e-01 1.19695461e+00 -7.26932660e-02
4.41455752e-01 -8.63299847e-01 8.30942035e-01 6.39802590e-02
-1.21414471e+00 2.21181870e-01 -2.47748107e-01 7.22834319e-02
3.27294141e-01 1.97250202e-01 -7.69061148e-01 8.79357219e-01
6.46689773e-01 4.18776214e-01 -2.23359555e-01 6.57330692e-01
-1.05616772e+00 1.09132564e+00 -2.22560972e-01 -1.94986597e-01
3.31239551e-01 -4.09872234e-01 5.65421402e-01 1.69744754e+00
6.85294569e-02 6.28620982e-01 1.84320390e-01 4.90925670e-01
2.42770072e-02 4.98019278e-01 -2.17805386e-01 1.46079630e-01
6.53095543e-01 9.02034283e-01 -1.13813949e+00 -6.55472457e-01
-4.62767810e-01 1.02308154e+00 3.56449574e-01 4.05791968e-01
-7.37762749e-01 -7.77651012e-01 4.54917431e-01 -1.34700298e-01
7.09611595e-01 -4.13697273e-01 -4.28871214e-01 -1.22002769e+00
-3.53004411e-02 -1.19967353e+00 9.72771585e-01 -1.57072365e-01
-8.21651280e-01 6.86678052e-01 1.94017157e-01 -4.91711885e-01
-2.95139968e-01 -7.24851906e-01 -5.38369417e-01 3.91544878e-01
-1.49997878e+00 -1.08659613e+00 4.01180565e-01 6.50270224e-01
1.06567681e+00 -4.83859666e-02 7.27374256e-01 1.62922576e-01
-9.72579300e-01 7.37501025e-01 -5.88216931e-02 4.93572265e-01
3.37311238e-01 -1.28344250e+00 6.34612560e-01 1.34405994e+00
9.38738808e-02 4.24053699e-01 5.71583629e-01 -1.07167339e+00
-9.28734303e-01 -8.55518222e-01 1.63054669e+00 -1.92095682e-01
8.79581869e-01 -4.61663753e-01 -9.57666039e-01 1.18363798e+00
7.04251766e-01 -3.12889010e-01 5.85533917e-01 -1.01755084e-02
-3.36360842e-01 1.78531766e-01 -1.14947653e+00 5.39901614e-01
1.06996012e+00 7.27508590e-02 -1.04596448e+00 2.80183882e-01
6.01778388e-01 -5.91357946e-01 -1.04524648e+00 1.74928665e-01
4.29337561e-01 -8.32904816e-01 2.41067365e-01 -6.37845635e-01
6.02174178e-02 -1.99771076e-02 -5.34675270e-02 -1.15851414e+00
6.92153275e-02 -8.67653370e-01 2.68568918e-02 1.52667153e+00
1.13496995e+00 -8.80849898e-01 5.40398657e-01 2.89760351e-01
-5.57282388e-01 -4.36357379e-01 -1.24912119e+00 -9.55423534e-01
4.43492830e-01 -6.10759556e-01 6.45930648e-01 1.10975790e+00
3.74559432e-01 9.48293060e-02 -5.35109043e-02 -3.00781913e-02
5.25414526e-01 -1.98450506e-01 2.29970887e-01 -6.93907738e-01
-1.27449080e-01 -3.22809428e-01 8.33477080e-03 -5.63164473e-01
3.20888430e-01 -1.01166368e+00 1.51837930e-01 -1.63762653e+00
-3.88680160e-01 -4.39517796e-01 1.10257633e-01 9.09797847e-01
1.73545734e-03 1.66954279e-01 4.02698487e-01 2.74995565e-02
-3.49405557e-01 -8.99576989e-04 8.05049837e-01 2.35334083e-01
-5.80312967e-01 -2.96141326e-01 -5.04971921e-01 8.70031357e-01
1.37142658e+00 -7.87707627e-01 2.43642151e-01 -7.60223567e-01
1.47889704e-01 -1.24479547e-01 -2.86968261e-01 -6.14997566e-01
9.18119103e-02 -1.62823260e-01 -9.10710264e-03 -5.65214574e-01
-1.21096291e-01 -2.21530199e-01 -6.26677275e-02 7.78124332e-01
1.05769649e-01 4.81288344e-01 1.48640484e-01 -1.90907925e-01
-2.85605073e-01 -6.71468318e-01 6.53539777e-01 -4.33831066e-01
-1.11265290e+00 -3.07898186e-02 -8.00766766e-01 3.55550081e-01
1.29074180e+00 -3.68791401e-01 -5.40611506e-01 3.04845512e-01
-7.78190255e-01 1.59477696e-01 1.37709618e-01 4.22710568e-01
6.19546354e-01 -8.33701849e-01 -8.14280212e-01 1.76529497e-01
-1.03756011e-01 -3.55343968e-01 -3.76356244e-01 4.83844072e-01
-1.21110249e+00 6.32208765e-01 1.09771807e-02 8.16448033e-02
-1.10757554e+00 3.13649833e-01 9.26147252e-02 -4.31891203e-01
-3.70154530e-01 8.72850776e-01 -6.84587717e-01 -6.18129134e-01
-1.60711363e-01 -4.94742215e-01 -2.92578131e-01 1.49527848e-01
1.91359326e-01 6.34499609e-01 2.17637584e-01 -8.79180610e-01
-6.44141555e-01 2.70250946e-01 -4.88874823e-01 -4.54989165e-01
9.55925882e-01 5.23572080e-02 -2.52555490e-01 3.34346741e-01
5.26805878e-01 3.16185415e-01 -4.75420356e-01 1.93160772e-01
5.71137428e-01 -3.04069698e-01 -1.80040941e-01 -1.01848412e+00
-8.15873861e-01 2.48469174e-01 2.03792769e-02 2.26343796e-01
8.56100440e-01 1.59205124e-01 9.47862983e-01 9.46902111e-02
4.01876152e-01 -1.72479188e+00 -6.43433928e-01 1.00570607e+00
6.46833718e-01 -9.91975307e-01 -1.26508579e-01 -8.89332652e-01
-9.86643255e-01 1.01839912e+00 6.47529602e-01 -3.69271547e-01
7.49570131e-01 3.83290142e-01 3.16709638e-01 8.32786039e-02
-7.24026620e-01 -6.18257284e-01 -1.47508129e-01 6.79411054e-01
7.47151375e-01 4.61448222e-01 -1.51312506e+00 8.34455490e-01
-4.98354644e-01 -2.17079282e-01 6.48288131e-01 1.22001529e+00
-4.10445035e-01 -1.76038241e+00 -3.17025483e-01 9.99697298e-02
-9.47305858e-01 -4.58025515e-01 -5.66431224e-01 1.60869753e+00
3.53701532e-01 9.95341659e-01 -9.26449448e-02 -1.48154993e-03
6.95247412e-01 2.16225058e-01 4.23439354e-01 -9.51380730e-01
-1.13777757e+00 4.40225810e-01 8.45207095e-01 -4.67926741e-01
-3.73077184e-01 -1.11524236e+00 -1.68908894e+00 -3.23865414e-01
-3.53521466e-01 4.04011369e-01 4.82219845e-01 8.19602668e-01
-1.18768126e-01 2.65439302e-01 1.01545960e-01 -2.20333204e-01
-2.88930565e-01 -8.32132161e-01 -5.40006280e-01 -6.83777332e-02
-2.53371418e-01 -9.55634937e-02 -1.20526552e-01 1.69294745e-01] | [10.38459587097168, 10.073356628417969] |
cdd4f445-3cd7-4fef-ad3f-9ae424c5cf5a | geometric-feature-based-face-sketch | 1312.1462 | null | https://arxiv.org/abs/1312.1462v1 | https://arxiv.org/pdf/1312.1462v1.pdf | Geometric Feature Based Face-Sketch Recognition | This paper presents a novel facial sketch image or face-sketch recognition approach based on facial feature extraction. To recognize a face-sketch, we have concentrated on a set of geometric face features like eyes, nose, eyebrows, lips, etc and their length and width ratio because it is difficult to match photos and sketches because they belong to two different modalities. In this system, first the facial features/components from training images are extracted, then ratios of length, width, and area etc. are calculated and those are stored as feature vectors for individual images. After that the mean feature vectors are computed and subtracted from each feature vector for centering of the feature vectors. In the next phase, feature vector for the incoming probe face-sketch is also computed in similar fashion. Here, K-NN classifier is used to recognize probe face-sketch. It is experimentally verified that the proposed method is robust against faces are in a frontal pose, with normal lighting and neutral expression and have no occlusions. The experiment has been conducted with 80 male and female face images from different face databases. It has useful applications for both law enforcement and digital entertainment. | ['Debotosh Bhattacharjee', 'Sourav Pramanik'] | 2013-12-05 | null | null | null | null | ['sketch-recognition'] | ['computer-vision'] | [ 3.04834694e-01 -2.33758032e-01 -1.88247472e-01 -5.13645470e-01
-5.80795519e-02 -5.11048079e-01 5.52899957e-01 -3.65266979e-01
-1.77259594e-01 4.90417093e-01 4.46362831e-02 2.42798343e-01
-2.97495365e-01 -6.01401150e-01 -6.40388727e-02 -8.25343490e-01
2.64763921e-01 9.28289071e-03 -1.21991029e-02 -4.60130423e-02
6.27076387e-01 1.31911743e+00 -2.02557707e+00 7.11454451e-03
-1.01160802e-01 1.13925350e+00 -3.32512468e-01 5.22845328e-01
-2.64984101e-01 2.01152071e-01 -4.45213288e-01 -3.53268147e-01
4.75684911e-01 -2.83555031e-01 -1.00183338e-01 5.59752345e-01
6.87610805e-01 -5.74008167e-01 -1.86412990e-01 9.87297475e-01
5.45111418e-01 5.53735495e-02 1.11832869e+00 -1.41261053e+00
-3.86303306e-01 -1.63014159e-01 -9.91620481e-01 -1.56789750e-01
5.08062124e-01 -1.58465207e-01 3.05912644e-01 -1.27975953e+00
5.60646951e-01 1.43587112e+00 4.80007589e-01 7.12591350e-01
-9.16402161e-01 -1.09748018e+00 -2.19603792e-01 7.99160674e-02
-1.82361996e+00 -9.28264976e-01 1.11432016e+00 -2.60266602e-01
4.92047727e-01 3.15857798e-01 5.05258203e-01 4.12751615e-01
9.80255604e-02 2.94441819e-01 1.17492855e+00 -6.56934857e-01
-1.83341920e-01 5.33930957e-01 6.88451529e-02 9.20707107e-01
-3.36017722e-04 3.13875601e-02 -2.68515110e-01 -2.46487752e-01
7.45423138e-01 3.62506509e-01 -1.23649217e-01 -1.68979034e-01
-6.95700407e-01 5.25540948e-01 -2.65084207e-01 6.19105518e-01
-4.48849976e-01 -4.98335510e-01 1.70447677e-01 3.02664280e-01
-1.68281332e-01 -4.45135355e-01 -1.21603059e-02 2.23216116e-01
-1.02357149e+00 2.46379614e-01 8.04625392e-01 9.24791813e-01
6.84016109e-01 -1.32938966e-01 1.14424102e-01 1.03772473e+00
5.83312750e-01 7.86819339e-01 4.21298206e-01 -6.47284687e-01
1.35756895e-01 7.73727953e-01 -6.69798488e-03 -1.48652244e+00
-2.00332642e-01 5.30965328e-01 -7.46489704e-01 4.53026474e-01
2.17529058e-01 7.06259906e-02 -9.42456245e-01 1.34819555e+00
4.02236164e-01 1.12710536e-01 1.02211751e-01 7.26693213e-01
1.14826715e+00 6.25286520e-01 -2.99482327e-02 -4.44476724e-01
1.60785985e+00 -4.91321385e-01 -9.48441386e-01 2.38675222e-01
-3.22760195e-01 -1.26424325e+00 7.09690392e-01 5.80402792e-01
-9.71775115e-01 -7.57337391e-01 -1.10352349e+00 1.74530253e-01
-4.90573347e-01 7.68460393e-01 7.55032301e-02 1.10112417e+00
-7.03647733e-01 3.99773985e-01 -2.78161883e-01 -4.72632825e-01
2.33435541e-01 5.92050850e-01 -8.44811976e-01 8.38535503e-02
-4.69562620e-01 6.32710218e-01 -1.65830106e-01 2.41814852e-01
-3.07691455e-01 -8.49583969e-02 -6.68006480e-01 -1.07441776e-01
-2.05640018e-01 1.31060600e-01 6.81945562e-01 -1.20352995e+00
-1.72253561e+00 9.60493565e-01 -4.42230999e-01 5.25294900e-01
1.26128435e-01 3.74740541e-01 -9.77689683e-01 2.45121479e-01
-1.98216647e-01 4.18396205e-01 1.40033758e+00 -9.51357126e-01
-3.63309115e-01 -9.05425727e-01 -5.44131100e-01 1.52102858e-01
-2.94516385e-01 4.40366954e-01 -4.55046952e-01 -3.47431093e-01
4.72802818e-01 -8.12073529e-01 5.82980096e-01 2.46594697e-01
-1.03266910e-01 -3.07483912e-01 1.44777250e+00 -6.70605838e-01
1.03980410e+00 -2.20327640e+00 -2.79613078e-01 8.01407933e-01
-2.33411923e-01 4.70181584e-01 -1.73247486e-01 4.72693652e-01
-1.97246596e-01 -2.23025799e-01 8.88901576e-02 1.03178360e-01
-2.97169030e-01 6.50420934e-02 -8.64072591e-02 5.91143548e-01
3.99844319e-01 1.71718568e-01 -1.92226857e-01 -7.87959039e-01
2.98309803e-01 6.91769421e-01 -1.30095422e-01 1.57274589e-01
4.77580547e-01 7.55108222e-02 -5.67835510e-01 1.04326224e+00
1.28640032e+00 6.01390421e-01 1.94111377e-01 -5.54400682e-01
1.53167108e-02 -6.09151185e-01 -1.74918139e+00 9.14033711e-01
-1.58488333e-01 6.66406453e-01 3.68141145e-01 -8.42027903e-01
1.48745632e+00 5.94101071e-01 3.05620939e-01 -4.46966857e-01
6.05445743e-01 1.79011196e-01 -1.00045636e-01 -8.62027168e-01
3.03495586e-01 -2.51069695e-01 4.57104236e-01 4.34682518e-01
1.47912726e-01 -3.48320678e-02 2.32509568e-01 -4.14039850e-01
3.75010967e-01 -7.77220652e-02 3.97933543e-01 -1.33288190e-01
1.44202137e+00 -6.17060959e-01 2.14653403e-01 -5.75085767e-02
-1.73787296e-01 2.48002157e-01 4.16576117e-01 -3.92001003e-01
-9.47153807e-01 -9.05033290e-01 -3.33840460e-01 6.24748349e-01
-7.22846687e-02 2.56178342e-02 -7.54849553e-01 -4.97261733e-01
9.55326557e-02 -6.17824681e-02 -4.76486593e-01 1.10402338e-01
-4.24495429e-01 -9.03256536e-02 5.66567719e-01 1.73720792e-01
4.08304423e-01 -1.10744822e+00 -3.27698290e-01 -3.11884172e-02
5.78408062e-01 -8.03268313e-01 -3.74825269e-01 -6.65534556e-01
-7.03817189e-01 -1.35317171e+00 -8.76385212e-01 -1.07334054e+00
1.14385176e+00 2.43803978e-01 3.29477936e-01 3.07636172e-01
-7.29671538e-01 4.52135205e-01 -9.27433595e-02 -5.86913884e-01
-2.02797234e-01 -5.63200533e-01 2.28012100e-01 6.01053774e-01
6.72519207e-01 -5.13939023e-01 -5.71423590e-01 5.85835457e-01
-7.68685222e-01 -5.89351535e-01 4.21174616e-01 6.41346812e-01
2.92884797e-01 1.64217874e-01 4.90977734e-01 -5.87534606e-01
8.33186150e-01 -1.07493296e-01 -6.59438252e-01 4.03591692e-01
-1.67145491e-01 -1.94564596e-01 5.34840703e-01 -3.95993739e-01
-1.20897627e+00 2.21166357e-01 -1.49924219e-01 -3.38367641e-01
-4.58864808e-01 -1.17179640e-01 -3.47948670e-01 -3.67897600e-01
4.31443721e-01 3.15516323e-01 4.89806414e-01 -3.04237664e-01
9.06246826e-02 1.13599420e+00 5.13216913e-01 -3.40681076e-01
8.80711615e-01 4.36854720e-01 2.89759487e-01 -1.35552263e+00
7.51222894e-02 -4.75274891e-01 -6.96596980e-01 -4.97345984e-01
6.68296397e-01 -5.05098701e-01 -1.06860781e+00 6.48477077e-01
-9.61384177e-01 8.14684868e-01 4.83958304e-01 5.11148453e-01
-1.91808745e-01 5.43623149e-01 -1.79655552e-01 -1.33552468e+00
-5.74622810e-01 -9.92826879e-01 8.62946391e-01 7.22242296e-01
4.94842120e-02 -6.53595746e-01 -2.18027532e-01 4.74914424e-02
4.36457545e-01 1.33049339e-01 7.57494986e-01 -4.43861723e-01
-1.84833199e-01 -7.46667564e-01 -3.21131200e-01 5.26442051e-01
7.13754475e-01 7.09968626e-01 -1.09374559e+00 -8.74986798e-02
-4.64525707e-02 -1.90931424e-01 4.66980040e-01 6.81858659e-02
9.68589664e-01 -2.92815298e-01 -3.29275340e-01 3.81132782e-01
1.40873027e+00 6.46012604e-01 5.85417867e-01 -4.22827691e-01
1.17052659e-01 6.26620531e-01 6.48785532e-01 4.91569996e-01
-1.46467403e-01 3.81897688e-01 -4.57191095e-02 9.46237072e-02
1.30306065e-01 -2.43796054e-02 2.20287278e-01 5.67562878e-01
-2.87983835e-01 8.60261843e-02 -3.57358247e-01 2.09392503e-01
-1.15964496e+00 -1.24452889e+00 1.88884094e-01 2.22406983e+00
3.97263765e-01 -3.38233858e-01 1.06559768e-01 7.04105914e-01
8.65336657e-01 -5.73634990e-02 -1.50320306e-01 -6.31192148e-01
9.39120203e-02 6.12045705e-01 1.99426800e-01 4.26586092e-01
-9.28750515e-01 6.98861718e-01 5.79439068e+00 6.59865022e-01
-1.61973631e+00 -3.35932910e-01 3.42467487e-01 3.87308821e-02
4.95559722e-02 -1.92468390e-01 -9.88475144e-01 3.08276922e-01
4.14898455e-01 3.56281139e-02 5.15056670e-01 6.31613910e-01
-3.90528962e-02 -3.62063289e-01 -8.30361843e-01 1.35265565e+00
3.80562127e-01 -8.53790283e-01 1.53839663e-01 -9.46333036e-02
3.03479105e-01 -6.90683246e-01 8.07523653e-02 5.43124564e-02
-7.54374564e-01 -1.09548140e+00 3.09900880e-01 8.05006802e-01
9.43005085e-01 -8.70989621e-01 6.05531573e-01 1.58700392e-01
-1.30969977e+00 -3.76656726e-02 -5.60865700e-01 1.25842869e-01
-1.82800129e-01 -1.23100758e-01 -8.58101368e-01 2.26029485e-01
2.20235899e-01 1.57677874e-01 -3.07756484e-01 6.70706272e-01
3.35396111e-01 3.13812308e-02 -5.04043519e-01 -2.23269090e-01
-5.54015450e-02 -5.38466275e-01 1.71435446e-01 1.19302642e+00
6.11432135e-01 4.12374854e-01 -2.22146839e-01 7.24555433e-01
-6.75071171e-03 6.86268151e-01 -8.07167947e-01 -2.88457274e-01
7.32421339e-01 1.72410142e+00 -7.44103312e-01 -3.78671199e-01
-4.93275464e-01 7.20076919e-01 -4.29167032e-01 1.71516150e-01
-5.44000864e-01 -7.08027303e-01 4.89330977e-01 1.88608125e-01
1.33662298e-01 9.76974964e-02 3.57221544e-01 -5.84313393e-01
7.55517036e-02 -7.64970005e-01 1.62603989e-01 -5.63754618e-01
-9.68970418e-01 5.59464633e-01 5.81664145e-02 -1.05121469e+00
-1.20957635e-01 -9.13931429e-01 -8.45581651e-01 1.23603439e+00
-1.29014862e+00 -1.19840097e+00 -5.50634503e-01 1.04659748e+00
3.61851364e-01 -7.41641223e-01 9.14159536e-01 4.45284545e-01
-5.08561730e-01 7.13616014e-01 -5.30047752e-02 4.06404674e-01
6.41854823e-01 -5.00110984e-01 -1.28654435e-01 3.37264985e-01
-6.14373498e-02 8.66531491e-01 1.67605594e-01 -4.69507307e-01
-1.52535188e+00 -5.37239492e-01 8.75834703e-01 8.25106725e-02
3.05396318e-02 -8.10007527e-02 -5.90972364e-01 3.12327176e-01
1.46112084e-01 3.24111998e-01 6.86812937e-01 -5.10458767e-01
-1.31379098e-01 -3.23120654e-01 -1.58924413e+00 3.64468932e-01
5.25621891e-01 -5.22756994e-01 -4.53716993e-01 3.32796510e-04
-4.35123026e-01 -1.03617691e-01 -1.01104259e+00 3.56638104e-01
1.30901527e+00 -1.14169586e+00 7.42321908e-01 -4.43023592e-01
-1.94410443e-01 -5.37029028e-01 -1.52271137e-01 -5.13444126e-01
-1.34427056e-01 -3.66294503e-01 5.91805756e-01 1.55904448e+00
1.53251484e-01 -5.87124944e-01 7.93462455e-01 4.48057652e-01
6.36164188e-01 -5.71485937e-01 -6.54577553e-01 -4.87691402e-01
-5.74888527e-01 1.80024803e-01 6.72329724e-01 7.72840917e-01
-2.20717907e-01 2.12343365e-01 -2.82105476e-01 6.68627489e-03
5.97540021e-01 5.09906001e-02 9.04350698e-01 -1.38168168e+00
3.05919468e-01 -3.37386698e-01 -6.58935785e-01 -2.13389084e-01
1.99956656e-01 -5.88261664e-01 -5.47816217e-01 -9.56104219e-01
5.75166307e-02 -3.53863448e-01 -6.31430447e-02 3.66127044e-01
4.43972945e-01 4.81305689e-01 1.31207824e-01 -1.31084770e-01
3.96283299e-01 1.38947412e-01 9.65124965e-01 -7.81231821e-02
-1.89127877e-01 1.80911526e-01 -1.88881233e-01 7.93336034e-01
7.22078681e-01 -2.10443854e-01 -5.19090354e-01 2.91888714e-01
-5.75905800e-01 2.45135546e-01 1.41564205e-01 -9.69757855e-01
4.16786782e-02 -2.57877886e-01 8.90103042e-01 -7.00704813e-01
6.66163862e-01 -1.16107750e+00 3.69736046e-01 3.07805717e-01
2.79855598e-02 1.77858919e-01 1.41603112e-01 7.76520371e-02
-2.06280902e-01 -5.45791149e-01 9.28288043e-01 -1.62287802e-01
-5.24380088e-01 2.51383930e-01 -2.84070492e-01 -5.79338372e-01
1.19381392e+00 -7.76239753e-01 8.84374827e-02 -2.44837269e-01
-6.69585764e-01 -4.31185544e-01 3.72659832e-01 5.15983403e-01
9.95313168e-01 -1.62018394e+00 -6.42684639e-01 1.10657561e+00
1.53597653e-01 -8.53818238e-01 2.83507317e-01 7.87483692e-01
-5.88486314e-01 3.25458705e-01 -9.24626052e-01 -3.15627277e-01
-2.09770918e+00 5.11059403e-01 2.27014318e-01 5.90274036e-01
-1.39475456e-02 5.49304307e-01 -3.02424788e-01 -1.31646663e-01
3.19517732e-01 -7.95750171e-02 -5.16013980e-01 3.23892206e-01
7.38821685e-01 3.01506072e-01 6.47855476e-02 -1.11576617e+00
-4.23058659e-01 1.32736742e+00 6.20802417e-02 -1.13510221e-01
1.02420795e+00 1.34573489e-01 -3.18611503e-01 2.84294190e-04
1.46959949e+00 6.53637886e-01 -6.77618742e-01 -2.03167275e-01
-2.08025753e-01 -8.76192987e-01 -3.11638802e-01 -3.89919370e-01
-1.18467391e+00 9.20011044e-01 9.15750384e-01 -1.26788437e-01
1.20932758e+00 -4.24480021e-01 3.75306278e-01 3.16207111e-01
1.38444275e-01 -1.26152313e+00 -3.03570926e-01 6.53233454e-02
1.13383996e+00 -1.00779009e+00 1.41328499e-01 -3.26845586e-01
-4.28280145e-01 1.65348971e+00 4.50988412e-01 -2.23205864e-01
1.14453280e+00 3.45058471e-01 -4.80982056e-03 -9.69552845e-02
-4.85012472e-01 -5.14706187e-02 5.78176141e-01 5.80461383e-01
6.27299011e-01 -1.16532773e-01 -5.58137059e-01 3.59984636e-01
-1.60057954e-02 3.29600006e-01 2.72763669e-01 9.82425213e-01
-6.89034462e-01 -1.10548997e+00 -7.22576201e-01 4.85626996e-01
-5.39019763e-01 3.64358127e-01 -5.44144511e-01 9.30034220e-01
3.31414789e-01 8.60175669e-01 1.71170905e-01 -3.63308519e-01
3.50847900e-01 4.30018216e-01 1.02842641e+00 -9.01207030e-02
-3.29926461e-01 1.11475900e-01 -3.76686603e-02 -3.27527881e-01
-5.14698148e-01 -5.96565723e-01 -1.16177630e+00 -2.86462158e-01
-3.72533113e-01 1.16038948e-01 1.05730319e+00 4.74894971e-01
-5.35518804e-04 -3.57165754e-01 8.34127724e-01 -9.38555717e-01
-4.33665663e-01 -1.03407574e+00 -7.47095525e-01 4.03004557e-01
1.68549389e-01 -9.01052296e-01 -1.55949354e-01 1.56459168e-01] | [13.176111221313477, 0.7015678882598877] |
e4e6a8c7-6bdd-478c-b8ef-e44ced4c69bd | few-shot-controllable-style-transfer-for-low-1 | null | null | https://openreview.net/forum?id=iU7nNeWLbA7 | https://openreview.net/pdf?id=iU7nNeWLbA7 | Few-shot Controllable Style Transfer for Low-Resource Multilinugal Settings | Style transfer is the task of rewriting an input sentence into a target style while approximately preserving its content. While most prior literature assumes access to large style-labelled corpora, recent work (Riley et al. 2021) has attempted "few-shot" style transfer using only 3-10 sentences at inference for extracting the target style. In this work we study a relevant low-resource setting: style transfer for languages where no style-labelled corpora are available. We find that existing few-shot methods perform this task poorly, with a strong tendency to copy inputs verbatim. We push the state-of-the-art for few-shot style transfer with a new method modeling the stylistic difference between paraphrases. When compared to prior work using automatic and human evaluations, our model achieves 2-3x better performance and output diversity in formality transfer and code-mixing addition across seven languages. Moreover, our method is better able to control the amount of style transfer using an input scalar knob. We report promising qualitative results for several attribute transfer directions, including sentiment transfer, text simplification, gender neutralization and text anonymization, all without retraining the model. Finally we found model evaluation to be difficult due to the lack of evaluation datasets and metrics for many languages. To facilitate further research in formality transfer for Indic languages, we crowdsource annotations for 4000 sentence pairs in four languages, and use this dataset to design our automatic evaluation suite. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['text-anonymization'] | ['natural-language-processing'] | [ 3.52235287e-01 2.47342527e-01 -3.60871479e-02 -6.54541969e-01
-1.19990969e+00 -1.02602088e+00 7.04955101e-01 -1.40404522e-01
-6.63819611e-01 1.02216518e+00 4.88432914e-01 -2.96251774e-01
3.62440318e-01 -4.77505326e-01 -7.67731667e-01 -6.36565462e-02
6.99705780e-01 7.51924217e-01 -2.45264873e-01 -7.43515491e-01
2.45766178e-01 1.73361138e-01 -8.81449103e-01 4.66973275e-01
1.16065550e+00 3.55976611e-01 -1.46681339e-01 7.23881185e-01
-2.32991368e-01 6.07919633e-01 -9.64092553e-01 -1.17715764e+00
3.83699566e-01 -5.90255380e-01 -1.19798446e+00 -2.93086797e-01
8.14862549e-01 -2.32699811e-01 -2.79350895e-02 1.00061858e+00
8.43860507e-01 -5.54077290e-02 6.86574042e-01 -1.17871344e+00
-1.07537138e+00 8.90109718e-01 -4.19235975e-01 -2.86949240e-02
5.26077449e-01 3.68310571e-01 1.04172516e+00 -6.93938792e-01
1.00731361e+00 1.51305950e+00 7.68003643e-01 1.12849379e+00
-1.76072764e+00 -7.31877744e-01 -1.37330815e-01 -3.71636927e-01
-1.10481858e+00 -6.18556619e-01 6.36103630e-01 -3.84989381e-01
9.58656311e-01 4.00175124e-01 3.47158045e-01 1.55571485e+00
-1.63273830e-02 6.70751274e-01 1.39798045e+00 -6.02927327e-01
4.92549269e-03 6.24468446e-01 4.64240387e-02 3.45907331e-01
1.23026930e-01 -2.09545121e-01 -6.38818502e-01 -1.62850603e-01
3.21956128e-01 -6.37948394e-01 4.90302593e-02 -8.96107703e-02
-1.11953008e+00 8.16684365e-01 2.10464433e-01 2.55776137e-01
3.56627911e-01 -2.43353546e-02 7.92634964e-01 9.04047549e-01
8.79140496e-01 1.03630221e+00 -5.31583488e-01 -4.28787291e-01
-1.01224446e+00 5.61225951e-01 1.14218426e+00 1.56404984e+00
7.93604016e-01 -3.55246244e-03 -6.58574879e-01 1.05928564e+00
-3.54014188e-01 6.89757228e-01 3.92775238e-01 -9.43159163e-01
9.26594198e-01 3.36960435e-01 1.08057380e-01 -4.25089926e-01
-1.24442667e-01 -6.42629527e-03 -5.91729224e-01 9.85910296e-02
6.64405465e-01 -6.42929733e-01 -6.85021222e-01 2.01350284e+00
-2.06365332e-01 -7.44961917e-01 -1.30864218e-01 4.83746260e-01
6.30444407e-01 3.36606503e-01 1.97983906e-01 1.99283019e-01
1.33302486e+00 -8.51374865e-01 -5.78286052e-01 -4.83574986e-01
9.92594123e-01 -1.21951389e+00 1.92101181e+00 1.08716294e-01
-1.29038525e+00 -2.94373184e-01 -7.70315826e-01 -5.40246427e-01
-3.49478334e-01 2.61143930e-02 5.60098410e-01 1.03196657e+00
-1.14333034e+00 7.90169477e-01 -4.73954439e-01 -7.32164919e-01
6.29794180e-01 3.55178624e-01 -4.58478659e-01 -3.32079157e-02
-1.18707812e+00 1.18318582e+00 4.44149971e-02 -5.17833650e-01
-2.96123415e-01 -1.21423137e+00 -8.72188270e-01 -1.54274955e-01
1.19366042e-01 -1.07414007e+00 1.40139592e+00 -1.40895736e+00
-1.72558999e+00 1.29452777e+00 -2.22210363e-01 -3.26489091e-01
8.37961018e-01 -5.17320156e-01 -2.16287032e-01 -3.59070748e-01
3.72621924e-01 1.03927433e+00 6.51410401e-01 -9.50390577e-01
-3.11485410e-01 -5.36789261e-02 2.37528026e-01 3.63558888e-01
-6.79317832e-01 5.80239773e-01 -1.30874887e-01 -1.06715155e+00
-8.45480978e-01 -1.06329799e+00 5.85974380e-03 -2.24708378e-01
-4.15810943e-01 2.76917079e-03 4.28480387e-01 -7.86807954e-01
1.16771352e+00 -2.03766513e+00 4.76859450e-01 -1.53753415e-01
-7.46646523e-02 3.05276573e-01 -2.93481469e-01 4.53174531e-01
5.08302301e-02 5.24842978e-01 -4.34417218e-01 -8.11621666e-01
2.09914461e-01 -5.29329814e-02 -4.34556186e-01 7.73840621e-02
4.09926116e-01 1.10317445e+00 -7.45582283e-01 -5.66757023e-01
-2.18000308e-01 1.67051837e-01 -1.13586259e+00 1.90493315e-01
-5.22569045e-02 2.71837473e-01 9.66713056e-02 4.45651323e-01
5.80800772e-01 2.29500890e-01 -5.11551350e-02 7.68060386e-02
1.14916846e-01 3.74501318e-01 -8.35171044e-01 2.02790999e+00
-8.08325708e-01 7.07333505e-01 -3.38720717e-02 -1.80637538e-01
7.90150642e-01 2.90425140e-02 -1.49148285e-01 -4.86127615e-01
1.85411289e-01 3.13576579e-01 2.73524672e-02 -1.57187268e-01
8.58172417e-01 -6.33934021e-01 -6.16076052e-01 6.30647421e-01
3.83719325e-01 -7.91677296e-01 4.27249819e-01 3.33897471e-01
9.37576056e-01 6.06632493e-02 -4.42199111e-02 -6.50278866e-01
3.51100087e-01 -1.04263853e-02 4.88082886e-01 7.78383732e-01
-1.80416390e-01 8.96513522e-01 6.96425438e-01 -7.10119829e-02
-1.57279897e+00 -9.96788800e-01 9.89195555e-02 1.55471730e+00
-3.66779476e-01 -5.03581405e-01 -1.37061608e+00 -6.30404294e-01
2.08834419e-03 1.17245889e+00 -6.88693583e-01 -3.19390893e-01
-6.82126403e-01 -4.62991148e-01 1.16742313e+00 4.50915873e-01
2.42267966e-01 -1.20416403e+00 -5.12062833e-02 -1.43189346e-02
-2.91023761e-01 -7.50811756e-01 -1.02893937e+00 -8.92344639e-02
-6.53480113e-01 -4.62756276e-01 -8.44044685e-01 -7.93610990e-01
5.47169566e-01 -3.29556137e-01 1.51662016e+00 -1.85718939e-01
-1.66376293e-01 1.70588657e-01 -8.97046477e-02 -9.04523432e-01
-8.32444906e-01 8.09921086e-01 2.28791609e-02 -3.39534998e-01
5.48686445e-01 -3.68993551e-01 -2.79414922e-01 7.91865140e-02
-8.24476302e-01 -2.97045782e-02 3.60755265e-01 9.96048152e-01
-6.73266500e-02 -8.33459795e-01 5.26442111e-01 -1.53926432e+00
1.15682101e+00 -5.61175235e-02 -2.51376957e-01 2.85802156e-01
-4.73574013e-01 1.99844256e-01 8.41402650e-01 -4.93413001e-01
-1.39142549e+00 -3.06691051e-01 -2.23693997e-02 -1.53842703e-01
-2.07806095e-01 -1.05699748e-01 -3.48316193e-01 2.17522997e-02
1.11264431e+00 -2.98599154e-01 1.96036071e-01 -4.81586456e-01
7.42496729e-01 7.62752056e-01 3.33785713e-01 -9.99470592e-01
9.44627702e-01 2.47835591e-01 -4.15249705e-01 -6.42655909e-01
-8.87786984e-01 3.95699628e-02 -7.02195346e-01 3.16361398e-01
7.63928115e-01 -9.85190511e-01 -3.14984232e-01 4.58255082e-01
-1.20249236e+00 -5.85756779e-01 -6.46568656e-01 1.65913776e-01
-7.19853044e-01 1.59309357e-01 -8.41032982e-01 -2.93762654e-01
-7.00256169e-01 -9.95575130e-01 1.13654602e+00 -1.03120208e-01
-1.19037735e+00 -1.16063583e+00 3.45264167e-01 4.58438963e-01
7.01751530e-01 -9.52274054e-02 1.03128147e+00 -5.95844388e-01
2.84072533e-02 3.57822026e-03 -1.09470904e-01 4.94899601e-01
1.85559705e-01 -1.72519386e-02 -1.09116137e+00 -4.51768994e-01
-1.51437297e-01 -7.03094304e-01 6.12046480e-01 -3.61252762e-02
7.37362206e-01 -3.32936406e-01 4.30902056e-02 8.19523871e-01
1.03072619e+00 -2.66591370e-01 6.88751817e-01 2.81438529e-01
1.00620544e+00 7.77152956e-01 4.38730747e-01 9.97302979e-02
2.66187698e-01 5.14708877e-01 -5.12531877e-01 -2.71259248e-01
-3.11932564e-01 -3.83817077e-01 4.25574541e-01 8.36954713e-01
1.21541880e-01 -1.38819173e-01 -7.60199726e-01 5.98681450e-01
-1.38442814e+00 -9.35389817e-01 2.22994387e-01 2.00250387e+00
1.45620847e+00 3.96871030e-01 3.07237476e-01 -2.92557210e-01
6.59683049e-01 -8.15028623e-02 -3.31948251e-01 -1.10493135e+00
-2.76431382e-01 4.17995840e-01 6.08602285e-01 7.70990551e-01
-9.09421027e-01 1.48249984e+00 6.66924858e+00 9.51020122e-01
-9.76638138e-01 5.34677245e-02 8.18779945e-01 -5.83978951e-01
-7.60176837e-01 1.57190889e-01 -9.08507884e-01 5.52447140e-01
8.42339396e-01 -5.04095912e-01 8.01582098e-01 6.07419252e-01
-1.99432727e-02 1.62482351e-01 -1.59279096e+00 7.98695743e-01
2.46215791e-01 -1.04449403e+00 2.64738202e-01 -2.18881816e-01
9.83156383e-01 -1.39294758e-01 1.77619591e-01 5.79639137e-01
5.65351069e-01 -1.19281960e+00 9.79839623e-01 5.27973294e-01
1.46322751e+00 -8.76734793e-01 4.68736887e-01 1.16387665e-01
-5.14467120e-01 8.08796287e-02 -3.55699003e-01 -3.47098202e-01
1.35438710e-01 3.21813762e-01 -6.97925150e-01 3.15121114e-01
6.01825178e-01 5.85064232e-01 -9.95862126e-01 3.76449883e-01
-1.97743833e-01 6.24675810e-01 -2.26794168e-01 -2.73632467e-01
-1.09906308e-01 -1.83962390e-01 5.89833140e-01 1.42706466e+00
8.13617408e-02 -2.23497361e-01 -3.00979763e-01 1.07458293e+00
-4.88104552e-01 3.19338590e-01 -9.49041665e-01 -2.70307492e-02
4.92433518e-01 1.20430386e+00 -3.77458543e-01 -4.11324263e-01
-4.03889090e-01 1.24196994e+00 6.80768013e-01 2.85277814e-01
-5.49042344e-01 -9.45870578e-01 8.26946616e-01 -2.32587699e-02
-3.13906968e-02 5.91568276e-02 -1.13767123e+00 -1.33815956e+00
1.52427137e-01 -1.27152562e+00 8.74681473e-02 -6.84443772e-01
-1.62698781e+00 5.57908475e-01 2.61887852e-02 -9.71393168e-01
-3.39143515e-01 -4.13767010e-01 -7.15907514e-01 1.36689246e+00
-8.35081875e-01 -1.37551475e+00 1.75987706e-01 3.31170976e-01
6.13404155e-01 -4.77420330e-01 8.81604314e-01 4.03743654e-01
-3.76767039e-01 1.35572183e+00 9.37627815e-03 1.40334457e-01
1.53835630e+00 -1.47192991e+00 9.77274120e-01 7.08512306e-01
-2.58270860e-01 1.04037762e+00 1.00396895e+00 -8.23451340e-01
-1.08520329e+00 -1.02975535e+00 1.25742376e+00 -1.09304571e+00
7.57976830e-01 -1.06101429e+00 -7.65983582e-01 8.80639493e-01
6.07521236e-01 -5.41589200e-01 6.42690480e-01 3.88266593e-01
-4.22509581e-01 -7.85010606e-02 -1.24803483e+00 1.10982084e+00
1.34632003e+00 -8.28185678e-01 -7.59365141e-01 2.35990688e-01
1.01293790e+00 -3.89938802e-01 -8.74407649e-01 5.19367829e-02
3.30369413e-01 -5.99670947e-01 5.82789063e-01 -8.68240178e-01
8.21366906e-01 1.36868164e-01 7.59964213e-02 -1.79162431e+00
-3.76057565e-01 -9.53610778e-01 6.45694137e-01 1.98457170e+00
7.86288321e-01 -4.49479252e-01 7.06143618e-01 9.82320189e-01
2.59100944e-02 -3.23265612e-01 -5.84954560e-01 -7.58216023e-01
9.87240732e-01 -1.73137169e-02 7.24705696e-01 9.79186594e-01
-1.91710424e-02 8.50681007e-01 -4.12506521e-01 -6.77268386e-01
3.63037199e-01 -2.55379379e-01 1.04527295e+00 -8.43226016e-01
-3.00055653e-01 -7.04136908e-01 9.82268602e-02 -3.64894092e-01
5.87643266e-01 -1.29147971e+00 -1.71956480e-01 -1.19433582e+00
3.08020055e-01 -2.44006708e-01 3.61463279e-01 3.85701090e-01
-3.83940399e-01 3.57178062e-01 4.85803336e-01 -3.98297980e-02
-3.26277047e-01 5.98044038e-01 1.16301155e+00 -1.21740922e-01
-1.90764263e-01 -1.85886860e-01 -1.24345982e+00 5.33454955e-01
8.83656800e-01 -4.72958028e-01 -3.77821982e-01 -7.65450776e-01
3.27528775e-01 -5.24483204e-01 4.99170199e-02 -6.43376589e-01
-1.68756917e-01 -4.58167195e-02 3.15163642e-01 2.47775335e-02
1.89773124e-02 -4.09153372e-01 -1.76431283e-01 1.40005633e-01
-7.43579984e-01 2.12174296e-01 4.69835818e-01 3.96526568e-02
6.76544309e-02 -3.31575215e-01 7.95444846e-01 -1.10394061e-01
-1.40061766e-01 -2.31380444e-02 -3.10305178e-01 8.85057688e-01
6.72507465e-01 1.06414240e-02 -3.59131873e-01 -4.05717909e-01
-3.31406087e-01 1.10092148e-01 1.00705218e+00 5.89342892e-01
-1.07424699e-01 -1.47883773e+00 -9.52792823e-01 1.82356700e-01
2.62615740e-01 -1.92478612e-01 -6.06741123e-02 3.28497976e-01
-6.44034922e-01 1.49445698e-01 -4.94234145e-01 -2.54143894e-01
-1.21477616e+00 5.34472048e-01 1.76817015e-01 -2.27763459e-01
-2.29264766e-01 9.25533235e-01 1.82592034e-01 -1.05217695e+00
-3.91021706e-02 -9.64461640e-02 2.79658496e-01 1.23741046e-01
3.91902983e-01 5.40920317e-01 1.92878500e-01 -3.80026340e-01
-1.98381945e-01 4.70646828e-01 -3.51898551e-01 -6.36285484e-01
1.03834188e+00 -8.35245624e-02 -2.68269986e-01 6.29086852e-01
1.37146151e+00 3.41321379e-01 -1.09294116e+00 -1.55720204e-01
4.83714044e-03 -3.98482293e-01 -6.78123295e-01 -8.47401679e-01
-4.60874110e-01 9.58245277e-01 1.70138553e-01 -1.68677315e-01
7.91241407e-01 -1.86111316e-01 7.66456187e-01 3.57868582e-01
2.32472137e-01 -1.52745605e+00 -6.70869350e-02 9.74122882e-01
9.48761284e-01 -1.12591517e+00 -8.52196068e-02 -3.45714986e-01
-1.07125533e+00 6.43415213e-01 8.32335651e-01 -4.85955650e-04
2.26866171e-01 3.73220742e-01 2.92490661e-01 2.97527283e-01
-6.39613092e-01 3.08771223e-01 2.70054787e-01 6.37393653e-01
8.84089112e-01 1.68403327e-01 -3.38432014e-01 6.22447431e-01
-1.10760999e+00 -1.08044885e-01 5.47113478e-01 8.22894633e-01
1.46243334e-01 -1.46617389e+00 -2.75867194e-01 6.01664484e-01
-6.34263992e-01 -5.32607377e-01 -9.20045376e-01 6.93292558e-01
-2.33123913e-01 8.24325085e-01 1.35784313e-01 -1.99762508e-01
7.11726665e-01 3.58945906e-01 6.69131458e-01 -9.79518414e-01
-1.21149039e+00 -2.30403602e-01 3.89613390e-01 -2.68146724e-01
1.31138325e-01 -7.62949288e-01 -7.26207614e-01 -8.42547596e-01
1.73654128e-02 -1.55204341e-01 3.77265275e-01 9.05104518e-01
3.39083195e-01 3.96716535e-01 4.54975128e-01 -8.18138301e-01
-7.24590957e-01 -1.20944011e+00 -4.34612393e-01 1.04823923e+00
-3.88427228e-02 -1.29718348e-01 -2.71067113e-01 1.96108386e-01] | [11.509256362915039, 9.658592224121094] |
989520ce-e504-4138-ab9c-cfdf2ef7a2f0 | deep-fusion-an-attention-guided-factorized | 1901.04889 | null | http://arxiv.org/abs/1901.04889v1 | http://arxiv.org/pdf/1901.04889v1.pdf | Deep Fusion: An Attention Guided Factorized Bilinear Pooling for Audio-video Emotion Recognition | Automatic emotion recognition (AER) is a challenging task due to the abstract
concept and multiple expressions of emotion. Although there is no consensus on
a definition, human emotional states usually can be apperceived by auditory and
visual systems. Inspired by this cognitive process in human beings, it's
natural to simultaneously utilize audio and visual information in AER. However,
most traditional fusion approaches only build a linear paradigm, such as
feature concatenation and multi-system fusion, which hardly captures complex
association between audio and video. In this paper, we introduce factorized
bilinear pooling (FBP) to deeply integrate the features of audio and video.
Specifically, the features are selected through the embedded attention
mechanism from respective modalities to obtain the emotion-related regions. The
whole pipeline can be completed in a neural network. Validated on the AFEW
database of the audio-video sub-challenge in EmotiW2018, the proposed approach
achieves an accuracy of 62.48%, outperforming the state-of-the-art result. | ['Zi-Rui Wang', 'Yuanyuan Zhang', 'Jun Du'] | 2019-01-15 | null | null | null | null | ['video-emotion-recognition'] | ['computer-vision'] | [ 8.56160745e-03 -5.27769685e-01 4.48993355e-01 -4.27793264e-01
-9.09220338e-01 -2.42885917e-01 3.52213055e-01 1.71386033e-01
-6.04931653e-01 2.73723125e-01 2.65913397e-01 3.00134867e-01
1.47562519e-01 -3.02323490e-01 -4.36253846e-01 -6.66998506e-01
3.81892808e-02 -3.63237858e-01 -1.13646835e-01 -1.84111878e-01
3.65116149e-02 2.98591584e-01 -1.95230520e+00 7.49249876e-01
7.57742226e-01 1.68949139e+00 -4.11077514e-02 5.96125066e-01
-8.59415904e-02 7.26672292e-01 -5.15654922e-01 -5.05941093e-01
-2.50966012e-01 -4.16627765e-01 -4.16874051e-01 -2.50391122e-02
8.53102356e-02 -1.53970063e-01 -3.50487411e-01 1.01432753e+00
7.87022293e-01 2.42097899e-01 3.44495237e-01 -1.41055822e+00
-5.70071518e-01 3.35797876e-01 -5.37456870e-01 1.45508721e-01
5.46435297e-01 -1.08384453e-01 1.00500083e+00 -1.29942691e+00
2.72343248e-01 1.15372992e+00 3.91250581e-01 2.65275657e-01
-8.35724771e-01 -8.10755491e-01 2.83882588e-01 7.82891035e-01
-1.64320982e+00 -6.28865361e-01 8.69494200e-01 -3.25142413e-01
8.14355195e-01 3.88807446e-01 9.56366181e-01 1.08149457e+00
8.10719356e-02 1.14443958e+00 1.11651778e+00 -1.64110214e-01
-3.97270657e-02 2.71307230e-01 -5.16750058e-03 5.10622203e-01
-5.54843128e-01 -1.39485076e-01 -8.97599876e-01 -4.11095880e-02
1.89902484e-01 1.67891771e-01 -4.02176321e-01 1.69394538e-01
-1.22894967e+00 4.50308442e-01 4.75526154e-01 4.70121950e-01
-7.63386488e-01 -5.67316748e-02 5.83985984e-01 2.96089321e-01
4.13590729e-01 9.39249769e-02 -2.13396341e-01 -2.44732261e-01
-1.01084435e+00 1.11623868e-01 4.58039582e-01 3.77145261e-01
6.13528430e-01 -1.07731624e-02 -3.02953064e-01 8.80729437e-01
4.47883278e-01 2.69940048e-01 4.58162457e-01 -7.05889583e-01
7.08211139e-02 5.61651587e-01 3.61977257e-02 -1.26455224e+00
-3.91431004e-01 -3.60799283e-01 -8.98434997e-01 7.18860105e-02
8.00636187e-02 -2.66033947e-01 -5.39097250e-01 1.77219415e+00
3.46166730e-01 4.45307970e-01 3.36780213e-02 1.15823591e+00
1.18149412e+00 8.36366236e-01 2.37809047e-01 -1.84377357e-01
1.61108625e+00 -1.08054113e+00 -1.05077863e+00 -9.07457843e-02
1.27733782e-01 -7.48005629e-01 8.65650594e-01 6.96125925e-01
-9.59844530e-01 -6.91035509e-01 -1.10810423e+00 -7.36399293e-02
-4.34048057e-01 4.26921934e-01 5.72933316e-01 4.19880182e-01
-8.49772036e-01 3.02923918e-01 -6.30696654e-01 -2.09351435e-01
4.06912208e-01 2.71105409e-01 -6.87298357e-01 9.57891569e-02
-1.37846696e+00 8.45507383e-01 2.84838378e-01 7.02458978e-01
-8.82112086e-01 -4.17506903e-01 -7.79946923e-01 2.86012262e-01
3.38594168e-01 -4.76764202e-01 9.65943933e-01 -1.32053912e+00
-1.62557697e+00 5.21558046e-01 -2.22994328e-01 -2.16555774e-01
1.75812438e-01 -5.47807634e-01 -7.21060991e-01 3.91442567e-01
-4.05529380e-01 6.39784813e-01 9.67174053e-01 -9.95862484e-01
-6.74462020e-01 -4.77582902e-01 -1.06770128e-01 3.37790221e-01
-7.39473820e-01 5.45081496e-01 -5.63104153e-01 -4.70182985e-01
-2.00678408e-02 -5.44528961e-01 1.79130156e-02 -1.11110091e-01
-2.82269299e-01 -2.85783768e-01 7.88555145e-01 -9.77100492e-01
1.45432937e+00 -2.63687062e+00 3.02247435e-01 4.23568264e-02
1.86763078e-01 8.91125128e-02 -1.89246282e-01 1.54614747e-01
-2.39614502e-01 -1.03613101e-01 1.45013526e-01 -6.10519171e-01
3.81350100e-01 -3.34141254e-01 -3.07357371e-01 4.27224725e-01
4.45867628e-01 8.86254013e-01 -7.03727663e-01 -6.20425105e-01
2.30662778e-01 9.40463185e-01 -4.39495414e-01 3.74272048e-01
3.12659591e-01 3.55767488e-01 -3.79399210e-01 7.70836592e-01
5.31327963e-01 2.73580644e-02 -1.50954410e-01 -5.52306116e-01
-1.30359188e-01 -9.35935453e-02 -1.20784545e+00 1.87836051e+00
-4.07416284e-01 5.57634175e-01 4.07904804e-01 -8.71308982e-01
1.03051937e+00 7.35784173e-01 5.65863192e-01 -4.64100689e-01
4.93225604e-01 1.18485540e-01 -5.49701415e-02 -7.25666940e-01
4.06859398e-01 -2.51374483e-01 -3.71589661e-01 2.82742530e-02
4.85142320e-01 3.23242173e-02 -7.20148608e-02 -4.57953056e-03
9.53304529e-01 5.00579104e-02 1.49502113e-01 2.25561246e-01
7.44887412e-01 -5.49880326e-01 6.60567403e-01 2.99934804e-01
-4.72060680e-01 5.64623594e-01 4.19679284e-01 -1.60943002e-01
-5.13014495e-01 -8.64404082e-01 3.44725512e-02 1.20727515e+00
6.11493103e-02 -6.02491617e-01 -6.43225670e-01 -4.25462574e-01
-3.38248044e-01 3.32888037e-01 -7.16111660e-01 -3.18209201e-01
3.15071568e-02 -6.08198524e-01 5.57153404e-01 3.28052789e-01
5.85293412e-01 -1.26677275e+00 -7.32067645e-01 3.10320377e-01
-4.37360972e-01 -1.28750539e+00 -3.06786478e-01 -3.87330204e-02
-1.69070646e-01 -7.12641895e-01 -6.67278111e-01 -4.61284578e-01
8.29648823e-02 -1.32912099e-02 7.30248332e-01 -8.37607607e-02
-2.78706074e-01 4.49725449e-01 -5.19247234e-01 -5.16987026e-01
1.19642526e-01 -1.42313287e-01 -7.69523978e-02 7.93793619e-01
5.58306634e-01 -5.72708547e-01 -7.22717166e-01 1.42357230e-01
-7.54900992e-01 1.66928172e-02 5.68163931e-01 7.69462466e-01
5.06204128e-01 -2.66617560e-03 6.86384261e-01 1.02905653e-01
6.81125164e-01 -4.93153036e-01 -4.72373739e-02 4.29279000e-01
1.12066641e-01 -3.90163839e-01 4.58684415e-01 -5.60607255e-01
-1.28637815e+00 2.15232715e-01 -3.80985379e-01 -6.99330032e-01
-4.31428254e-01 6.11290991e-01 -4.74246889e-01 6.02281019e-02
1.02466732e-01 1.56486884e-01 -2.03303039e-01 -2.75639981e-01
3.06017131e-01 9.18102801e-01 6.63672090e-01 -4.15129870e-01
5.03924750e-02 3.57113510e-01 -3.96651030e-01 -8.52838278e-01
-7.06569791e-01 -3.80323827e-01 -2.76229978e-01 -7.43104517e-01
1.15454614e+00 -1.12788105e+00 -1.03860354e+00 6.39892042e-01
-1.22949326e+00 3.14904630e-01 9.81292054e-02 7.92351127e-01
-3.09712052e-01 3.46153826e-01 -5.90089202e-01 -1.06096351e+00
-3.87051404e-01 -1.26224184e+00 1.07729578e+00 4.39460874e-01
-1.77038729e-01 -3.39859813e-01 -1.09072492e-01 2.62672216e-01
5.36356628e-01 2.08740860e-01 3.72976840e-01 -6.48195982e-01
1.04446923e-02 -4.18915987e-01 -3.05733234e-01 5.85888088e-01
-1.89029679e-01 1.87448621e-01 -1.39373982e+00 1.45210186e-02
1.06563598e-01 -6.73455060e-01 1.08585262e+00 1.45725876e-01
1.28380251e+00 1.06100753e-01 5.82078993e-02 3.32830876e-01
1.08172059e+00 1.64817035e-01 6.95171595e-01 -4.37395088e-02
3.84588093e-01 8.09011102e-01 6.00823104e-01 7.73828864e-01
5.13885081e-01 6.01140559e-01 6.09682262e-01 -1.25803620e-01
2.68458098e-01 5.51002212e-02 6.20480359e-01 1.05303776e+00
-1.15772821e-01 -9.17506218e-02 -6.10738993e-01 3.86116058e-01
-1.84954548e+00 -1.10222912e+00 7.88808465e-02 1.81572330e+00
6.18519247e-01 -1.45859987e-01 -3.43374722e-02 2.29166985e-01
7.98455000e-01 2.04462767e-01 -3.08684736e-01 -3.95871490e-01
-3.38621378e-01 2.78626889e-01 -3.56253982e-01 2.26018250e-01
-1.42715085e+00 6.69665098e-01 5.29157066e+00 9.40217376e-01
-1.34543049e+00 3.12501043e-01 5.49700081e-01 -3.89550626e-01
-1.64850224e-02 -3.83718133e-01 -4.19793040e-01 3.86211783e-01
1.04337728e+00 -7.24391639e-02 4.11881983e-01 4.79007304e-01
3.81988771e-02 7.29149859e-03 -7.68251359e-01 1.33037651e+00
2.49391839e-01 -6.93328321e-01 -3.05276185e-01 -2.86644876e-01
2.95674384e-01 -2.24786714e-01 1.57427683e-01 6.70385599e-01
-3.35499823e-01 -1.10878181e+00 7.98070848e-01 9.72525120e-01
3.92140746e-01 -9.48133886e-01 8.91527355e-01 8.46716538e-02
-1.38820887e+00 -2.95479804e-01 -1.21959113e-01 8.71579275e-02
3.14826608e-01 5.57153046e-01 -5.08863777e-02 7.78239906e-01
1.14932561e+00 7.48305440e-01 -4.00257319e-01 1.10374880e+00
-9.67320353e-02 5.43904126e-01 -4.78534520e-01 -2.08511651e-02
1.04499191e-01 -8.02424252e-02 4.17824984e-01 1.33282220e+00
6.71425879e-01 1.73100099e-01 -2.02788562e-02 6.69990778e-01
-3.16988565e-02 4.68823254e-01 -3.01238835e-01 -2.89391756e-01
2.35641927e-01 1.92551243e+00 -4.18050706e-01 -3.98391396e-01
-5.23726940e-01 1.07354665e+00 1.86423585e-01 3.18009824e-01
-1.22537470e+00 -5.79257190e-01 6.13337696e-01 -5.55597603e-01
4.51232702e-01 5.79873100e-02 1.69190302e-01 -1.19384098e+00
7.18711019e-02 -9.73151743e-01 2.54667580e-01 -1.02436268e+00
-1.18037343e+00 1.11041081e+00 -3.96397442e-01 -1.17753243e+00
1.61884844e-01 -4.99606818e-01 -5.91803133e-01 8.23806524e-01
-1.19019926e+00 -1.09050560e+00 -5.71818650e-01 7.22881615e-01
1.57777175e-01 3.63381691e-02 1.01224697e+00 7.50584304e-01
-8.43617439e-01 5.66754222e-01 -2.86608845e-01 -1.58943273e-02
9.78985906e-01 -8.05635393e-01 -3.97719324e-01 7.20753670e-01
1.81117252e-01 3.89119864e-01 6.22964799e-01 -1.55486047e-01
-1.32235742e+00 -7.66306579e-01 7.46857941e-01 3.20512317e-02
7.36265481e-01 -3.65985930e-01 -1.05979848e+00 2.81453043e-01
6.37067854e-01 2.49597982e-01 9.05794322e-01 -5.28250821e-02
-4.21558291e-01 -2.03122169e-01 -8.45626712e-01 4.25327539e-01
5.60514927e-01 -7.46488631e-01 -5.79598725e-01 -2.30946928e-01
6.01497173e-01 -1.08965077e-01 -1.08291817e+00 6.32143557e-01
8.09139788e-01 -1.08113468e+00 8.32005918e-01 -5.20523190e-01
5.07881701e-01 -4.83888954e-01 -5.18128395e-01 -1.25789905e+00
-9.43519324e-02 -4.92530406e-01 -3.40555906e-02 1.49604475e+00
1.34047970e-01 -3.64019305e-01 8.22212622e-02 3.65952790e-01
-2.04164296e-01 -8.91145587e-01 -1.09025133e+00 -1.48041904e-01
-5.08857608e-01 -7.70186841e-01 5.43005705e-01 8.96047294e-01
3.41830939e-01 3.81372005e-01 -5.48910737e-01 1.98859110e-01
2.45988339e-01 1.16993770e-01 4.77919936e-01 -1.04821587e+00
-1.47683501e-01 -6.58306003e-01 -5.06533265e-01 -6.34126842e-01
3.96365166e-01 -6.47083282e-01 8.19195881e-02 -1.33433199e+00
1.47877187e-01 1.47739887e-01 -9.22171831e-01 6.05068147e-01
-2.36914396e-01 4.91752833e-01 3.91460508e-01 -3.59990865e-01
-1.05429149e+00 1.08570278e+00 1.01561904e+00 -1.59289196e-01
-4.94118370e-02 -4.57838148e-01 -6.91825628e-01 7.35872567e-01
6.97339773e-01 -7.05946982e-02 1.44864572e-02 -1.44109681e-01
3.06587994e-01 2.04293221e-01 6.37261152e-01 -1.08778620e+00
3.66141498e-01 2.51789212e-01 4.82930541e-01 -6.58798277e-01
8.52958620e-01 -8.77085030e-01 2.53601879e-01 -5.84914256e-03
-2.01576233e-01 2.90865898e-02 4.01512563e-01 3.66543949e-01
-8.68359387e-01 1.06177427e-01 5.15257359e-01 2.13261411e-01
-7.86145091e-01 3.12806666e-01 -4.77114707e-01 -3.86660397e-01
9.36841846e-01 4.64840010e-02 -8.46785977e-02 -4.87522483e-01
-1.02730012e+00 2.70165354e-01 -1.29399240e-01 5.30222833e-01
9.00122285e-01 -1.65119720e+00 -7.90773213e-01 1.97598785e-01
1.57664225e-01 -4.45739895e-01 9.92897451e-01 1.27508652e+00
1.09844431e-01 -1.01435359e-03 -4.22607422e-01 -4.88671064e-01
-1.40623212e+00 5.09000003e-01 3.98664296e-01 -8.30982476e-02
-1.05293475e-01 9.11572576e-01 1.19774774e-01 -2.29566365e-01
2.62419790e-01 1.00359861e-02 -5.76638699e-01 5.60822666e-01
8.63307476e-01 1.75717279e-01 8.90821815e-02 -9.92379963e-01
-5.06969929e-01 4.83967841e-01 1.79418087e-01 -2.37035766e-01
1.35847783e+00 -1.67819843e-01 -2.64783084e-01 6.44958794e-01
1.29113340e+00 -1.09799251e-01 -1.01466811e+00 -1.83400407e-01
-3.44512790e-01 -2.42810398e-01 2.63129443e-01 -7.03472137e-01
-1.18492293e+00 1.41230583e+00 8.31019759e-01 9.77185145e-02
1.57558441e+00 -1.42251030e-01 6.27151072e-01 1.44972876e-01
8.21633115e-02 -1.09484506e+00 8.06048512e-02 3.85477513e-01
1.15904939e+00 -1.10589337e+00 -3.09560597e-01 -1.21483669e-01
-8.71995211e-01 1.06452215e+00 5.88760376e-01 -2.85269953e-02
8.65977585e-01 2.34678879e-01 9.84362364e-02 -3.59125957e-02
-1.02160513e+00 -1.78045765e-01 6.38950706e-01 1.02443285e-01
5.09202361e-01 -8.53309929e-02 -1.55343533e-01 1.39347935e+00
-5.85363470e-02 8.17897022e-02 3.60946078e-03 6.87755704e-01
-2.97878623e-01 -6.87363207e-01 -4.36750621e-01 2.16193423e-01
-7.56262183e-01 -2.76102815e-02 -2.20981091e-01 4.62964654e-01
2.34056681e-01 1.22361600e+00 1.20713674e-01 -8.15548480e-01
4.15816516e-01 4.38061774e-01 4.12002742e-01 -1.14232302e-01
-9.50575948e-01 3.72567087e-01 -6.73546940e-02 -8.48883629e-01
-6.88495159e-01 -6.01336539e-01 -1.23338270e+00 2.54684333e-02
-3.18285078e-01 2.09903240e-01 6.50289476e-01 9.63064671e-01
5.24958193e-01 7.74921477e-01 5.68488657e-01 -1.09170997e+00
-1.18357860e-01 -1.12735128e+00 -6.68201447e-01 4.44773972e-01
2.08157465e-01 -7.92897761e-01 -3.73654038e-01 -1.13011122e-01] | [13.303610801696777, 5.040400505065918] |
12db982e-ec75-410e-b43d-410c37535321 | lid-2020-the-learning-from-imperfect-data | 2010.11724 | null | https://arxiv.org/abs/2010.11724v1 | https://arxiv.org/pdf/2010.11724v1.pdf | LID 2020: The Learning from Imperfect Data Challenge Results | Learning from imperfect data becomes an issue in many industrial applications after the research community has made profound progress in supervised learning from perfectly annotated datasets. The purpose of the Learning from Imperfect Data (LID) workshop is to inspire and facilitate the research in developing novel approaches that would harness the imperfect data and improve the data-efficiency during training. A massive amount of user-generated data nowadays available on multiple internet services. How to leverage those and improve the machine learning models is a high impact problem. We organize the challenges in conjunction with the workshop. The goal of these challenges is to find the state-of-the-art approaches in the weakly supervised learning setting for object detection, semantic segmentation, and scene parsing. There are three tracks in the challenge, i.e., weakly supervised semantic segmentation (Track 1), weakly supervised scene parsing (Track 2), and weakly supervised object localization (Track 3). In Track 1, based on ILSVRC DET, we provide pixel-level annotations of 15K images from 200 categories for evaluation. In Track 2, we provide point-based annotations for the training set of ADE20K. In Track 3, based on ILSVRC CLS-LOC, we provide pixel-level annotations of 44,271 images for evaluation. Besides, we further introduce a new evaluation metric proposed by \cite{zhang2020rethinking}, i.e., IoU curve, to measure the quality of the generated object localization maps. This technical report summarizes the highlights from the challenge. The challenge submission server and the leaderboard will continue to open for the researchers who are interested in it. More details regarding the challenge and the benchmarks are available at https://lidchallenge.github.io | ['Jun He', 'Huanqing Yan', 'Chen Gong', 'Zhenyuan Chen', 'Zhendong Wang', 'Oles Dobosevych', 'Ostap Viniavskyi', 'Mariia Dobko', 'Yao Zhao', 'Shikui Wei', 'Guanghua Gu', 'Tao Ruan', 'Chuangchuang Tan', 'Li Zhang', 'Yurong Chen', 'Yiwen Guo', 'Anbang Yao', 'Ming Lu', 'Hao Zhao', 'Gunhee Kim', 'Jinhwan Seo', 'Junhyug Noh', 'Wonho Bae', 'Luc van Gool', 'Wenguan Wang', 'Guolei Sun', 'Ting Liu', 'Antonio Torralba', 'Yi Yang', 'Errui Ding', 'LiWei Wang', 'Hang Zhao', 'Ming-Ming Cheng', 'Shuai Zheng', 'Yunchao Wei'] | 2020-10-17 | null | null | null | null | ['scene-parsing'] | ['computer-vision'] | [ 2.04348654e-01 3.08061242e-02 -3.38697225e-01 -6.12095535e-01
-1.41052735e+00 -9.50171232e-01 1.45061255e-01 -2.09243566e-01
-4.14789379e-01 6.12863362e-01 -2.07291275e-01 -3.14819783e-01
1.51239604e-01 -3.04092586e-01 -1.22500026e+00 -6.44293427e-01
1.91372424e-01 4.45669264e-01 5.82540393e-01 2.37246603e-01
2.00181812e-01 2.03246295e-01 -1.40883589e+00 6.92771971e-01
5.81082642e-01 1.15851116e+00 5.89837253e-01 8.52203131e-01
-1.94207266e-01 6.76994026e-01 -5.16378701e-01 -2.69312769e-01
3.93997759e-01 -2.67196327e-01 -1.17754924e+00 -4.29555811e-02
6.38226748e-01 -2.20780119e-01 -4.15431596e-02 1.23202765e+00
4.38567519e-01 -1.12187810e-01 2.34250948e-01 -1.38736701e+00
-4.08660501e-01 5.62797844e-01 -4.76437271e-01 1.21159926e-01
-4.50704470e-02 2.84243286e-01 1.02160692e+00 -9.71089244e-01
6.22115076e-01 1.02318585e+00 7.56406128e-01 5.34654438e-01
-8.17758322e-01 -5.73766053e-01 2.70202368e-01 3.25472236e-01
-1.13525105e+00 -4.94981945e-01 7.44263530e-01 -3.61683816e-01
8.10166121e-01 2.28356600e-01 9.41011235e-02 9.55933154e-01
-5.06219864e-01 1.28835368e+00 1.18587208e+00 -5.90934277e-01
2.91700840e-01 2.92380810e-01 4.63316321e-01 7.77326286e-01
1.89104185e-01 1.30335480e-01 -5.21561503e-01 1.88550502e-01
5.78248024e-01 -3.66107553e-01 8.99310857e-02 -3.97984266e-01
-1.09579420e+00 5.52951872e-01 5.04417121e-01 2.24296123e-01
1.91205591e-02 1.60166681e-01 4.15538460e-01 2.48522963e-03
4.39949483e-01 3.61596167e-01 -8.69819283e-01 -3.70231807e-01
-6.64440036e-01 1.63540646e-01 7.38897085e-01 1.31286156e+00
6.25848949e-01 -2.02725366e-01 9.73825976e-02 1.02036631e+00
2.22662538e-01 6.56194627e-01 2.05504045e-01 -1.33731842e+00
7.07937539e-01 5.52425981e-01 3.41385007e-02 -4.08971280e-01
-3.67917269e-01 -2.01230451e-01 -4.55880255e-01 1.48387924e-01
7.09256232e-01 -9.12045687e-02 -1.23627710e+00 1.44920254e+00
3.62689078e-01 3.17411005e-01 -4.04877309e-03 9.77804601e-01
1.02250898e+00 5.04204869e-01 5.18685021e-02 3.70562464e-01
1.14557517e+00 -1.34730971e+00 -5.07270634e-01 -2.52951294e-01
9.75128353e-01 -8.67524147e-01 1.32846177e+00 3.79164696e-01
-9.29064393e-01 -8.47639680e-01 -9.21154559e-01 -1.36592597e-01
-6.50088847e-01 3.34525943e-01 5.52209616e-01 5.77683806e-01
-9.49568391e-01 5.24664938e-01 -9.71543968e-01 -4.38561231e-01
8.11304867e-01 2.20736340e-01 -2.73704469e-01 -4.88734215e-01
-7.59513795e-01 4.55698818e-01 4.67909366e-01 1.17336690e-01
-9.14905131e-01 -7.84425855e-01 -8.40274572e-01 -4.89899576e-01
6.12428784e-01 -1.57786697e-01 1.52193642e+00 -8.17006171e-01
-1.21423531e+00 1.26609790e+00 -6.20418563e-02 -4.07975107e-01
4.72723663e-01 -3.64258319e-01 -9.71707329e-02 7.55811995e-03
4.21768218e-01 7.71355450e-01 3.52661788e-01 -1.55079627e+00
-1.09545338e+00 -5.77256382e-01 5.50113097e-02 1.27925649e-01
2.30930559e-02 1.32795423e-01 -1.05276561e+00 -3.92569184e-01
2.20062509e-01 -1.12003529e+00 -3.21292728e-01 -1.23296633e-01
-7.80879319e-01 -2.77045101e-01 1.15517867e+00 -5.87516963e-01
4.38767344e-01 -2.40974569e+00 -2.27051467e-01 -3.52179378e-01
-1.37850180e-01 3.61536533e-01 -2.07597494e-01 -6.10505156e-02
1.53512005e-02 2.30344966e-01 -2.46426404e-01 -7.31437981e-01
1.82898730e-01 3.37880760e-01 -2.85131216e-01 3.59269679e-01
2.63268650e-01 1.07847273e+00 -8.81823599e-01 -4.71558869e-01
4.86425459e-01 -7.74669275e-02 -2.19615534e-01 3.09855849e-01
-4.11428958e-01 5.09615660e-01 -5.17370760e-01 9.69258785e-01
9.35755551e-01 -3.59767824e-01 -3.94178666e-02 -5.01412809e-01
-2.14625552e-01 5.80182113e-02 -1.21759081e+00 1.98248494e+00
-2.36830443e-01 5.05503535e-01 3.07656795e-01 -1.18369544e+00
7.30192780e-01 3.64706852e-02 3.85010958e-01 -7.88011074e-01
-9.62055326e-02 3.05713803e-01 -4.69449669e-01 -4.45112824e-01
4.05869633e-01 1.72136635e-01 -4.00636137e-01 7.66957626e-02
4.09364849e-01 -3.83676738e-01 2.81593591e-01 1.58566937e-01
1.22864032e+00 6.07645869e-01 -8.57304782e-02 -1.14520609e-01
2.41748795e-01 3.88365686e-01 6.63352191e-01 1.01021767e+00
-6.82586491e-01 9.05276358e-01 3.12836856e-01 -5.10852695e-01
-1.03017628e+00 -1.20539463e+00 -1.83630273e-01 1.13011181e+00
4.36649829e-01 -1.63185373e-01 -9.36679900e-01 -1.13630486e+00
-1.55743942e-01 2.97107846e-01 -4.27355677e-01 1.36692256e-01
-6.91382825e-01 -6.84440553e-01 6.55888677e-01 8.76742840e-01
8.83841395e-01 -1.22372639e+00 -1.77312806e-01 -1.48416787e-01
-4.00486916e-01 -1.82348156e+00 -8.96153077e-02 4.99404609e-01
-7.11588919e-01 -1.28086388e+00 -5.37008107e-01 -1.06265152e+00
4.27906424e-01 3.10202807e-01 1.33901978e+00 7.81752355e-03
-4.79456246e-01 3.97790045e-01 -6.05930984e-01 -6.43899977e-01
-4.02305685e-02 1.56679556e-01 -2.07690839e-02 -4.21608776e-01
3.72954726e-01 6.02165870e-02 -3.23135197e-01 5.74190021e-01
-7.14254677e-01 5.83995841e-02 5.50460935e-01 6.28788054e-01
1.13598371e+00 2.02422012e-02 3.97586435e-01 -1.23527634e+00
-2.46792912e-01 -2.97775209e-01 -8.80210698e-01 1.81434959e-01
-9.16961953e-02 -1.75553739e-01 4.85333353e-01 1.64747834e-02
-1.17866242e+00 6.27205372e-01 -4.79985386e-01 -2.68887550e-01
-6.99139655e-01 -2.33313087e-02 -4.52783704e-01 -1.33646086e-01
5.45774877e-01 -3.07674780e-02 -7.66667068e-01 -8.04079354e-01
5.70989609e-01 8.68865728e-01 8.77562165e-01 -8.43460739e-01
7.51750171e-01 5.98975003e-01 -2.59467632e-01 -9.11872745e-01
-1.49119663e+00 -9.87185180e-01 -8.44699025e-01 -1.17901579e-01
1.11494303e+00 -1.03575802e+00 -3.05824816e-01 8.13544095e-01
-1.12224996e+00 -8.87960255e-01 -5.99908113e-01 3.97231579e-01
-7.96375275e-01 9.14273113e-02 -6.28469825e-01 -5.84938228e-01
-3.66989262e-02 -1.30968761e+00 1.43340051e+00 3.17246795e-01
3.15534323e-01 -7.80776262e-01 -6.43128157e-02 1.01713789e+00
-5.34822419e-03 2.06997290e-01 4.77527618e-01 -7.04280972e-01
-9.64236915e-01 -3.11543643e-01 -6.02386296e-01 6.82210326e-01
-1.36196211e-01 -3.42616707e-01 -1.36797023e+00 -4.34323177e-02
-2.58661091e-01 -8.03453743e-01 8.51821661e-01 4.12485749e-01
1.53758669e+00 2.39709094e-01 -4.12054747e-01 7.31070697e-01
1.42387152e+00 -6.78592995e-02 4.78882521e-01 3.45890790e-01
1.02322650e+00 7.43282735e-01 9.22089577e-01 -3.04636378e-02
3.76128197e-01 9.39931989e-01 6.24135733e-01 -2.55921721e-01
-5.15031099e-01 -3.40072900e-01 7.50459209e-02 5.99975169e-01
2.28311066e-02 -2.01261505e-01 -1.04888785e+00 6.22199893e-01
-1.91902280e+00 -5.15288472e-01 -5.76695979e-01 1.97348368e+00
6.29381597e-01 3.93615037e-01 1.72129199e-02 2.45873071e-02
7.26600349e-01 1.14374094e-01 -5.94486117e-01 -1.45136252e-01
-1.34993196e-01 3.52470316e-02 7.97094107e-01 3.19612950e-01
-1.51949823e+00 1.31634843e+00 5.50866270e+00 9.18366313e-01
-6.79137051e-01 3.86873990e-01 7.89887309e-01 1.55173019e-01
3.12026113e-01 -1.10578820e-01 -1.17027283e+00 4.47690725e-01
9.83484089e-01 2.76551515e-01 1.87505513e-01 1.22461021e+00
1.19599245e-01 -1.45941466e-01 -9.90557313e-01 1.01745474e+00
7.88864791e-02 -1.38243556e+00 -5.48864186e-01 -5.41412085e-02
1.00951254e+00 7.34093904e-01 1.38705462e-01 5.51337302e-01
6.72919035e-01 -8.32322478e-01 6.65368259e-01 1.69535920e-01
7.38003194e-01 -3.96974236e-01 1.00663388e+00 4.47293073e-01
-1.35582590e+00 -8.29713941e-02 -4.49819386e-01 2.32967243e-01
2.24833816e-01 7.42793262e-01 -6.86060011e-01 6.21865988e-01
1.23349500e+00 6.78365052e-01 -6.09077692e-01 1.22220385e+00
-3.72797400e-01 8.68510842e-01 -3.40281874e-01 3.87111515e-01
2.69316316e-01 1.71969458e-01 2.88353324e-01 1.36158967e+00
-1.25675783e-01 1.91264192e-03 5.57762682e-01 7.75253415e-01
-3.88577104e-01 -1.64236933e-01 -5.86049497e-01 1.24961212e-01
2.93097317e-01 1.35151684e+00 -1.04708064e+00 -2.35991642e-01
-3.48086178e-01 1.14899945e+00 2.68761098e-01 2.66807020e-01
-7.04368114e-01 -1.24011286e-01 4.81070995e-01 -1.74875513e-01
3.26838911e-01 -2.10722193e-01 -6.28982186e-01 -7.18739212e-01
2.76090920e-01 -7.17030227e-01 4.10486907e-01 -8.17102015e-01
-1.32286406e+00 4.56019819e-01 -4.09280844e-02 -9.52359557e-01
9.92470011e-02 -9.47608590e-01 -3.84109378e-01 6.11064315e-01
-1.66157424e+00 -1.47711694e+00 -5.17340660e-01 2.85917342e-01
9.66244638e-01 1.46600187e-01 6.19526982e-01 3.58660072e-01
-6.12144232e-01 5.98273933e-01 1.82857573e-01 4.55055416e-01
6.61183596e-01 -1.54132164e+00 5.87571204e-01 7.19974279e-01
5.17179668e-01 -5.01797460e-02 3.84220272e-01 -4.59842414e-01
-1.26236737e+00 -1.49630916e+00 6.03856444e-01 -1.08391225e+00
5.15088022e-01 -5.75925529e-01 -6.98147893e-01 7.48964548e-01
-2.89546221e-01 6.46832466e-01 3.96931708e-01 1.23493187e-02
-2.85419464e-01 1.36281960e-02 -1.15758467e+00 1.03272296e-01
1.21279919e+00 -4.06977475e-01 -2.20331877e-01 7.76667058e-01
1.11048865e+00 -7.30109692e-01 -4.92228061e-01 5.77690721e-01
2.28612602e-01 -8.47641051e-01 9.63216722e-01 -4.67029303e-01
1.80837259e-01 -3.37903798e-01 -4.99621063e-01 -9.75331724e-01
-1.18396757e-02 -1.11572221e-01 1.16130903e-01 1.38660014e+00
5.79795778e-01 -1.46491528e-01 1.42421591e+00 3.09366643e-01
-6.16038501e-01 -6.39836609e-01 -8.77985060e-01 -9.96909678e-01
1.21027626e-01 -8.59360814e-01 2.10178301e-01 6.05477750e-01
-5.33755004e-01 1.18565068e-01 -2.01801538e-01 3.84189218e-01
9.76886153e-01 1.56871781e-01 9.34654355e-01 -9.28830981e-01
-3.35973352e-01 8.70713685e-03 -4.17466611e-01 -1.22030687e+00
1.44141823e-01 -8.59215558e-01 5.96945822e-01 -1.68391383e+00
2.82640010e-01 -7.40635872e-01 -4.11372721e-01 6.36752784e-01
-1.39032573e-01 6.58349037e-01 3.58435363e-01 1.54691339e-01
-1.28975284e+00 7.26295933e-02 1.16818571e+00 -5.19560575e-02
1.41101861e-02 2.35620573e-01 -5.35164714e-01 1.02656591e+00
1.08536196e+00 -4.81243491e-01 -1.73892856e-01 -6.19075239e-01
-2.98082918e-01 -4.52878922e-01 5.95872343e-01 -1.18609762e+00
1.49676368e-01 -5.73445158e-03 3.57182831e-01 -9.02468562e-01
4.49853033e-01 -7.45671451e-01 -4.49640274e-01 8.41475651e-02
-1.87838405e-01 -2.28403538e-01 2.70995349e-01 5.17179668e-01
-2.16109306e-01 -3.76651376e-01 8.42575788e-01 -3.93591791e-01
-1.28592825e+00 1.93919331e-01 3.45682085e-01 3.75091732e-01
1.09227610e+00 -6.17119297e-02 -4.64423299e-01 1.76013745e-02
-7.05089986e-01 2.84402370e-01 4.04710025e-01 6.03622198e-01
3.85594457e-01 -1.04503679e+00 -5.96575975e-01 1.96374729e-02
2.62352437e-01 4.09056634e-01 3.55597109e-01 6.09298468e-01
-3.46500635e-01 6.46432638e-01 1.84378654e-01 -8.71847689e-01
-1.30960274e+00 2.64518827e-01 1.94035694e-01 -4.15825605e-01
-4.98120248e-01 1.18066072e+00 2.28678346e-01 -8.78647745e-01
4.90658283e-01 -2.88881063e-01 1.29105538e-01 -4.47548568e-01
5.28528571e-01 3.61488312e-01 3.43307823e-01 -5.71770310e-01
-5.01905024e-01 5.71133256e-01 1.98229626e-02 5.79634048e-02
1.26467335e+00 9.83106252e-03 2.21301556e-01 4.27586228e-01
1.25266159e+00 -3.75174314e-01 -1.50524259e+00 -2.90354192e-01
3.34667087e-01 -3.56395662e-01 -1.73928458e-02 -1.16949117e+00
-1.08538556e+00 8.99322331e-01 9.98749375e-01 1.96212344e-02
8.17759633e-01 6.20660961e-01 8.43928397e-01 4.14854854e-01
5.89654624e-01 -1.27442193e+00 2.93622673e-01 7.09180415e-01
5.30891120e-01 -1.67503631e+00 -3.95882607e-01 -6.89968646e-01
-5.94701290e-01 6.98190689e-01 8.66577983e-01 1.60391018e-01
5.27448475e-01 6.54482782e-01 3.12460750e-01 -1.49534509e-01
-2.49147221e-01 -4.10566300e-01 1.09248966e-01 1.00286329e+00
2.71043271e-01 1.37463719e-01 2.26644993e-01 8.56827736e-01
-3.56309563e-02 7.33561963e-02 1.41718462e-01 9.77509737e-01
-5.86098194e-01 -1.18403554e+00 -3.22478563e-01 3.14304501e-01
-3.86931121e-01 5.43673225e-02 -4.30138260e-01 6.00112677e-01
5.36350608e-01 1.16385770e+00 -8.95840153e-02 -3.62328976e-01
7.52695739e-01 -9.52330530e-02 3.98802161e-01 -9.51591074e-01
-1.26294836e-01 -2.71463394e-01 2.08963856e-01 -8.74182880e-01
-2.99321145e-01 -5.36428869e-01 -1.43690193e+00 1.56585798e-01
-3.55345577e-01 9.03098062e-02 1.14693928e+00 9.49690878e-01
2.08248273e-01 5.43582618e-01 4.22473520e-01 -1.00075400e+00
-4.75221962e-01 -7.13981628e-01 -4.38869178e-01 3.97791922e-01
6.86700419e-02 -4.49336588e-01 -3.81061375e-01 3.50776166e-01] | [9.485994338989258, 0.5862497091293335] |
a9991cac-421c-40ca-aaf5-df35ae77aae2 | on-predictive-information-sub-optimality-of | null | null | https://openreview.net/forum?id=HklsHyBKDr | https://openreview.net/pdf?id=HklsHyBKDr | On Predictive Information Sub-optimality of RNNs | Certain biological neurons demonstrate a remarkable capability to optimally compress the history of sensory inputs while being maximally informative about the future. In this work, we investigate if the same can be said of artificial neurons in recurrent neural networks (RNNs) trained with maximum likelihood. In experiments on two datasets, restorative Brownian motion and a hand-drawn sketch dataset, we find that RNNs are sub-optimal in the information plane. Instead of optimally compressing past information, they extract additional information that is not relevant for predicting the future. Overcoming this limitation may require alternative training procedures and architectures, or objectives beyond maximum likelihood estimation. | ['Alexander A. Alemi', 'Ben Poole', 'Deniz Oktay', 'Zhe Dong'] | 2019-09-25 | null | null | null | null | ['information-plane'] | ['methodology'] | [ 6.10679686e-01 3.49619657e-01 -1.91287100e-01 -3.93078953e-01
-1.10891007e-01 -2.92742461e-01 8.06746423e-01 -1.03717536e-01
-7.90828347e-01 9.93934631e-01 5.14560819e-01 -4.00579572e-01
-2.95781583e-01 -8.04703355e-01 -6.80134714e-01 -6.63420916e-01
-3.95755976e-01 2.34034657e-01 9.06332210e-02 -5.70459338e-03
5.10667205e-01 7.77215302e-01 -1.46565843e+00 1.96530983e-01
2.22099036e-01 8.72746527e-01 3.70048672e-01 9.09228384e-01
-6.96079209e-02 9.95272398e-01 -5.94326735e-01 -1.87448651e-01
9.67903584e-02 -5.49587309e-01 -7.55413532e-01 -3.51940989e-01
3.37592587e-02 -2.31026977e-01 -8.44968021e-01 7.81037390e-01
2.60088742e-01 2.56960422e-01 7.98516631e-01 -7.59693980e-01
-7.38467336e-01 8.76414537e-01 9.81432423e-02 5.55530369e-01
-1.21240281e-01 1.79952398e-01 9.38250899e-01 -6.01714313e-01
7.88616419e-01 1.21831560e+00 5.94476402e-01 9.67790425e-01
-1.15865219e+00 -3.04188877e-01 2.59644061e-01 1.70665368e-01
-1.16255069e+00 -7.77818024e-01 1.17075121e+00 -5.94888292e-02
1.33537495e+00 1.21047989e-01 5.87186396e-01 1.50600946e+00
5.01951158e-01 9.66511548e-01 6.93688869e-01 -2.70080805e-01
5.84604561e-01 -2.09067255e-01 9.90869179e-02 3.55400175e-01
2.17702746e-01 3.87696177e-01 -6.82711840e-01 -1.63114313e-02
7.19815135e-01 -8.66246130e-03 -2.73269802e-01 -5.45275658e-02
-1.28774023e+00 5.50134063e-01 1.60329893e-01 5.57839811e-01
-5.57903767e-01 4.95252907e-01 2.39594206e-01 4.35027391e-01
3.05106312e-01 6.18349254e-01 -5.08588672e-01 -1.48606628e-01
-9.92437541e-01 -6.23148121e-02 1.01099885e+00 5.20297527e-01
5.11888802e-01 4.78400767e-01 1.57101691e-01 4.85210061e-01
3.06568537e-02 3.56660038e-01 8.87475252e-01 -1.16284931e+00
2.71782786e-01 2.34012619e-01 5.02879918e-02 -7.89786696e-01
-5.26593745e-01 -5.60890079e-01 -9.19811487e-01 8.81097242e-02
4.68957990e-01 -3.50321084e-01 -9.49311137e-01 2.03889704e+00
-4.15505618e-01 2.82168478e-01 4.90203112e-01 5.15717983e-01
2.59962887e-01 8.65300834e-01 1.97282843e-02 -4.20892596e-01
8.66624653e-01 -4.43835884e-01 -7.04432726e-01 -6.40662253e-01
3.92316461e-01 -3.30778629e-01 5.43121755e-01 4.59622651e-01
-1.14443040e+00 -4.55426812e-01 -1.36400700e+00 1.64525323e-02
-4.09455091e-01 7.34201521e-02 6.09645605e-01 3.78644258e-01
-1.11443079e+00 1.27732873e+00 -1.16050446e+00 -1.01979673e-01
2.57676542e-01 2.48767391e-01 -1.21137209e-03 4.32214439e-01
-1.22562706e+00 1.00143194e+00 8.52132738e-01 1.78330243e-01
-8.33544075e-01 -2.23965272e-01 -6.13097906e-01 1.61329463e-01
4.92470041e-02 -5.03203452e-01 1.05763543e+00 -7.12442577e-01
-1.47231555e+00 4.38258320e-01 -1.86609015e-01 -1.13572681e+00
2.59945899e-01 -1.98268339e-01 -3.73033971e-01 1.48302957e-01
-6.41819477e-01 9.23590481e-01 8.22919548e-01 -8.00084054e-01
-1.53993577e-01 -4.46610004e-01 -2.91500211e-01 -6.82104900e-02
-4.29534048e-01 -4.20917690e-01 9.04277042e-02 -8.20492029e-01
3.05489600e-01 -1.05313408e+00 -3.35448414e-01 -8.08075592e-02
-3.12949270e-01 -2.74560809e-01 6.75653875e-01 -5.62700093e-01
1.29219127e+00 -1.84490526e+00 1.89166769e-01 1.22725097e-02
-1.67589158e-01 3.00450027e-01 -1.16656795e-01 4.23674017e-01
7.50444084e-02 1.60699412e-01 -2.18534261e-01 -3.76326412e-01
-2.83245832e-01 7.09793448e-01 -9.24561799e-01 4.07717973e-01
5.61326921e-01 1.02063417e+00 -6.96912825e-01 -2.78203219e-01
-3.55291702e-02 5.27211964e-01 -1.31861672e-01 -1.58408418e-01
-5.39114833e-01 2.23316878e-01 -4.35215026e-01 3.67402971e-01
1.15939371e-01 -4.15156215e-01 1.77936897e-01 2.88042963e-01
2.31925379e-02 3.65739912e-01 -8.73168945e-01 1.69847846e+00
-4.15605485e-01 1.32965183e+00 -4.61554497e-01 -1.32150960e+00
1.25325930e+00 2.73835689e-01 1.71721295e-01 -7.38060176e-01
9.36629027e-02 1.04307987e-01 -5.77562377e-02 -2.59245217e-01
6.06233120e-01 -2.16115445e-01 1.74933881e-01 7.71417260e-01
9.17273462e-02 1.74749985e-01 -4.15858924e-02 -4.75380868e-02
1.18810427e+00 -2.12307796e-02 4.08912078e-03 -6.26152232e-02
3.13705325e-01 -4.28856790e-01 5.82031965e-01 1.05304921e+00
-1.03185005e-01 5.14355302e-01 4.50651348e-01 -5.84132135e-01
-1.49658275e+00 -1.15088725e+00 -1.23838320e-01 7.36189663e-01
-2.59335756e-01 -1.87188819e-01 -3.68455946e-01 -1.03287116e-01
-3.39989215e-01 1.14162719e+00 -5.91700256e-01 -5.32987475e-01
-8.29130232e-01 -6.84225142e-01 8.27816725e-01 6.38684809e-01
2.61615038e-01 -1.43604624e+00 -1.26346695e+00 4.92210060e-01
-1.19771317e-01 -1.03310537e+00 7.18935579e-02 3.47251326e-01
-1.47722399e+00 -5.16017318e-01 -9.55013990e-01 -4.31587785e-01
4.11213756e-01 -1.39547631e-01 1.05729353e+00 -1.54788449e-01
-1.76282614e-01 3.57185602e-01 7.60078654e-02 -6.67948842e-01
-4.75252092e-01 2.32996941e-01 1.50089785e-01 -3.59662205e-01
2.62551576e-01 -1.03016746e+00 -6.96196020e-01 4.59447466e-02
-9.13863122e-01 6.71312585e-02 8.37288320e-01 9.37824607e-01
3.68353128e-01 2.19816924e-03 7.61415124e-01 -3.53338659e-01
7.12904632e-01 -5.42630613e-01 -2.90260673e-01 3.08834821e-01
-6.70708477e-01 7.95520723e-01 8.63507390e-01 -8.28166604e-01
-1.05418169e+00 1.06201120e-01 -1.31141126e-01 -4.45302218e-01
-1.06507197e-01 2.85260409e-01 4.06035960e-01 3.76615375e-01
7.18154669e-01 6.16270244e-01 -7.56445602e-02 -5.39795756e-01
4.78059538e-02 2.31247097e-01 6.65073872e-01 -4.89788681e-01
8.46155658e-02 7.08216846e-01 2.55641043e-01 -8.52831900e-01
-7.79837668e-01 1.40108140e-02 -6.34825766e-01 -1.40009448e-01
4.67051536e-01 -2.94996530e-01 -5.64458311e-01 3.06298912e-01
-1.27612531e+00 -2.07234919e-01 -5.75718045e-01 7.17367351e-01
-9.30296719e-01 6.29825294e-02 -7.44308472e-01 -1.34484160e+00
-3.61815959e-01 -5.95338881e-01 5.95157623e-01 3.24989200e-01
-3.84531975e-01 -9.48666453e-01 2.37416327e-01 -3.73514235e-01
5.46379745e-01 1.14473231e-01 8.25596690e-01 -8.97590160e-01
-6.57513916e-01 -2.50759989e-01 1.38442069e-01 3.58554900e-01
-2.49563739e-01 -3.18494928e-03 -1.11866832e+00 -2.26607502e-01
3.33472550e-01 -2.87512600e-01 1.59126115e+00 4.82380033e-01
9.87128556e-01 -5.99354029e-01 -3.24097604e-01 4.97086495e-01
1.21454644e+00 4.62011725e-01 7.52032995e-01 1.84692383e-01
-7.84015059e-02 8.57192755e-01 1.44647747e-01 5.14228165e-01
-1.18299626e-01 6.07670732e-02 3.89174670e-01 6.53173745e-01
-5.33509650e-04 -5.05747736e-01 3.47231030e-01 9.11513507e-01
-6.65367395e-02 -3.96861106e-01 -9.32359636e-01 6.88780487e-01
-1.73518598e+00 -1.28620481e+00 6.05940163e-01 1.94130015e+00
8.42772722e-01 5.45133114e-01 -1.68467805e-01 3.49764556e-01
5.85772097e-01 4.17016268e-01 -1.13532615e+00 -4.27214712e-01
-5.44494629e-01 4.45020497e-02 3.76389652e-01 2.20985323e-01
-7.80804932e-01 5.72767973e-01 7.73420620e+00 5.35238028e-01
-1.10420203e+00 -3.78512621e-01 9.01372373e-01 -3.31555903e-01
-4.18876946e-01 1.60453096e-01 -8.61583591e-01 3.82048160e-01
1.63436103e+00 -1.87560692e-01 4.36753839e-01 5.99649489e-01
-1.41617522e-01 -9.10417512e-02 -1.01190400e+00 6.78563297e-01
-1.63539857e-01 -1.65946937e+00 1.15823328e-01 9.04458389e-02
4.34147865e-01 3.16563606e-01 3.48091632e-01 1.49803504e-01
2.23201171e-01 -1.16686869e+00 6.43500745e-01 1.27520084e+00
3.73184472e-01 -8.64832342e-01 3.87300551e-01 8.38678122e-01
-6.39056921e-01 -2.94617116e-01 -8.65385234e-01 -2.60601729e-01
2.21719772e-01 4.69151527e-01 -9.98190343e-01 -7.55663067e-02
4.41880137e-01 7.22999811e-01 -5.67812026e-01 7.80435324e-01
-8.27793255e-02 6.94484711e-01 -5.31840801e-01 -5.37435949e-01
3.39081913e-01 3.35521959e-02 8.49911451e-01 1.19794321e+00
4.80540931e-01 4.69233021e-02 -6.55896068e-01 1.07200551e+00
4.66351621e-02 -3.72377306e-01 -9.76069033e-01 -4.82307553e-01
4.90662634e-01 6.27432287e-01 -9.26163733e-01 -2.28785202e-01
5.35949655e-02 8.23958516e-01 3.08234304e-01 5.30977547e-01
-3.06206614e-01 -3.13810229e-01 3.98927987e-01 -3.76447707e-01
4.42488432e-01 -6.13912761e-01 -3.35032314e-01 -1.03961647e+00
-2.74093505e-02 -2.16638625e-01 2.75310218e-01 -8.96924078e-01
-9.32849884e-01 5.96568346e-01 -8.54353011e-02 -8.47029805e-01
-9.77608085e-01 -7.28041887e-01 -6.12086058e-01 6.09606385e-01
-1.25946295e+00 -7.67071605e-01 5.63699663e-01 3.10107380e-01
5.44165969e-01 -1.88090116e-01 8.14258218e-01 -3.93734515e-01
-3.85719657e-01 2.96657264e-01 3.68574649e-01 3.86471413e-02
-3.89709137e-02 -1.03662229e+00 8.51919234e-01 7.03617156e-01
5.17070115e-01 7.29345441e-01 9.72341597e-01 -5.97115278e-01
-1.42198575e+00 -7.13704765e-01 8.03667068e-01 -2.01936811e-01
8.24865282e-01 -9.71192867e-02 -1.25695598e+00 5.91145575e-01
-5.70075884e-02 -2.78698653e-01 4.09095258e-01 -1.86179392e-02
-2.76700407e-01 1.56382412e-01 -8.18261206e-01 7.76757777e-01
1.16661310e+00 -8.46347690e-01 -9.90985215e-01 -1.37339130e-01
8.77779543e-01 1.28129274e-01 -6.67943418e-01 4.56259847e-01
9.54329967e-01 -8.70924175e-01 1.15649283e+00 -8.39946210e-01
3.65906507e-01 3.52332413e-01 -1.97522774e-01 -1.11667109e+00
5.51590621e-02 -7.43584037e-01 -6.93101943e-01 8.70412409e-01
4.41805869e-01 -4.82350558e-01 1.10335839e+00 6.80618465e-01
1.52310088e-01 -9.08822596e-01 -1.21656060e+00 -9.74675119e-01
3.19647819e-01 -7.03776240e-01 3.31141502e-01 4.72827107e-01
-7.19582662e-02 1.96212858e-01 -5.50075412e-01 -1.77515373e-01
5.38166940e-01 -7.80931786e-02 1.06603593e-01 -1.27450979e+00
-2.48467878e-01 -6.58622146e-01 -3.85534376e-01 -1.14935863e+00
4.99223530e-01 -7.28570759e-01 -1.53859451e-01 -1.29515254e+00
-1.05317697e-01 6.63005784e-02 -6.11242354e-01 2.45491982e-01
4.19457227e-01 -2.24156216e-01 1.49603531e-01 4.05663341e-01
-4.20309842e-01 6.34816408e-01 1.03884995e+00 -8.04048479e-02
-2.19018906e-01 3.23356509e-01 -2.68604308e-01 8.67358148e-01
1.05953181e+00 -5.12595952e-01 -5.24073303e-01 -2.68081576e-01
5.77380538e-01 3.39586496e-01 4.64446962e-01 -1.18363893e+00
6.76163852e-01 -1.27040803e-01 6.24694526e-01 -9.81147170e-01
8.44447434e-01 -5.31096876e-01 6.52133003e-02 7.12896168e-01
-9.10577476e-01 1.53554916e-01 1.87735185e-01 9.84893084e-01
5.47901131e-02 -6.87913120e-01 5.59342146e-01 -4.82297212e-01
-6.32769942e-01 5.22278622e-02 -7.34008372e-01 -3.03835925e-02
5.33062220e-01 -2.55844980e-01 -3.53653997e-01 -5.94301045e-01
-7.68633664e-01 -7.04994053e-02 3.56921583e-01 3.32709908e-01
1.05910003e+00 -9.97953534e-01 -5.05498648e-01 1.22248910e-01
-3.00722361e-01 -5.13641298e-01 1.25497878e-01 2.94455379e-01
-3.87159735e-02 8.86505663e-01 -3.78427923e-01 -3.72155309e-01
-7.33492672e-01 6.18745148e-01 3.39752018e-01 -2.31604293e-01
-7.48629808e-01 6.25030637e-01 -2.49326020e-01 -3.55532728e-02
3.82862777e-01 -4.49798524e-01 -3.17383647e-01 1.25419617e-01
6.74852192e-01 5.04697442e-01 -2.61771441e-01 -3.18295240e-01
-4.07204926e-02 1.47966638e-01 -6.29831105e-02 -6.16834939e-01
1.55610740e+00 -1.89830139e-01 9.71425399e-02 1.04363477e+00
1.25817919e+00 -5.62520862e-01 -1.58605838e+00 -4.70495343e-01
5.76197505e-01 3.23127210e-02 8.64287466e-02 -5.69458842e-01
-8.09833407e-01 1.23642528e+00 2.74599403e-01 4.46071118e-01
1.11070096e+00 -8.24856609e-02 9.22733188e-01 1.09264553e+00
2.94584453e-01 -1.23386061e+00 1.51891589e-01 6.99782729e-01
8.68011117e-01 -7.44056940e-01 -6.37942106e-02 6.07088864e-01
-4.47884917e-01 1.54569709e+00 1.18691094e-01 -4.26185936e-01
8.61977458e-01 2.16509610e-01 -5.83634794e-01 9.87989642e-03
-1.56712365e+00 -5.05137965e-02 3.22184950e-01 4.71966445e-01
8.91013294e-02 -1.62117705e-01 -1.49084795e-02 4.54805553e-01
-3.85810286e-01 -1.85417756e-02 5.94660103e-01 1.09593594e+00
-6.46128953e-01 -5.63391805e-01 -7.46689290e-02 5.69873214e-01
-5.39720476e-01 -4.00194339e-02 -3.94205838e-01 5.04017830e-01
-4.81607229e-01 7.07511067e-01 3.55715722e-01 -3.68191391e-01
-1.58032686e-01 4.03690100e-01 4.06444699e-01 -1.73731387e-01
1.65551994e-02 -2.25179493e-01 1.35509983e-01 -3.03530037e-01
-3.75573844e-01 -6.33157194e-01 -1.20078146e+00 -1.42907187e-01
-2.29017869e-01 -1.25359312e-01 8.76766980e-01 1.08830202e+00
3.68542641e-01 4.36655432e-01 3.72371942e-01 -6.73314929e-01
-7.91588664e-01 -8.87804508e-01 -2.84122527e-01 4.87865619e-02
6.13349319e-01 -2.91050643e-01 -6.05209827e-01 -2.20476706e-02] | [7.7275896072387695, 3.49001407623291] |
2867de83-12ca-474a-9893-abebead939ee | personalized-substitution-ranking-for-lexical | null | null | https://aclanthology.org/W19-8634 | https://aclanthology.org/W19-8634.pdf | Personalized Substitution Ranking for Lexical Simplification | A lexical simplification (LS) system substitutes difficult words in a text with simpler ones to make it easier for the user to understand. In the typical LS pipeline, the Substitution Ranking step determines the best substitution out of a set of candidates. Most current systems do not consider the user{'}s vocabulary proficiency, and always aim for the simplest candidate. This approach may overlook less-simple candidates that the user can understand, and that are semantically closer to the original word. We propose a personalized approach for Substitution Ranking to identify the candidate that is the closest synonym and is non-complex for the user. In experiments on learners of English at different proficiency levels, we show that this approach enhances the semantic faithfulness of the output, at the cost of a relatively small increase in the number of complex words. | ['John Lee', 'Chak Yan Yeung'] | 2019-10-01 | null | null | null | ws-2019-10 | ['lexical-simplification'] | ['natural-language-processing'] | [ 6.46373406e-02 2.02690616e-01 6.77075163e-02 -2.41106957e-01
-4.90420401e-01 -7.21541882e-01 2.13699967e-01 1.08410048e+00
-8.64994526e-01 4.41886067e-01 4.19290870e-01 -3.37876260e-01
-1.47680238e-01 -8.30494940e-01 -3.94565344e-01 -1.09443709e-01
8.27540517e-01 5.96442163e-01 5.03484786e-01 -7.13948488e-01
2.46762976e-01 2.87726343e-01 -1.71584380e+00 4.22364444e-01
1.43311882e+00 1.85948715e-01 7.01297045e-01 1.19464852e-01
-3.37354094e-01 3.30683410e-01 -7.36024916e-01 -7.33438373e-01
6.71820939e-02 -5.77439606e-01 -8.00953388e-01 -4.47924346e-01
5.30499935e-01 -1.52903929e-01 6.24331199e-02 1.18566132e+00
2.50703007e-01 3.74402195e-01 4.99699235e-01 -4.06972289e-01
-5.07637858e-01 1.13653374e+00 1.62920117e-01 2.19425067e-01
5.79605579e-01 -3.78527284e-01 1.13475108e+00 -1.03862703e+00
7.87648261e-01 1.19180202e+00 6.38392091e-01 5.58991373e-01
-1.26871479e+00 -4.41588521e-01 3.46036136e-01 1.61006793e-01
-1.51280355e+00 -2.59597033e-01 4.49082673e-01 -1.97863176e-01
1.24639928e+00 4.22167659e-01 7.25295365e-01 5.47795475e-01
-1.42843589e-01 4.64210421e-01 8.17733288e-01 -9.03929293e-01
-4.35039140e-02 2.66201496e-01 4.52277333e-01 6.76479816e-01
4.67585504e-01 -3.56074542e-01 -7.17563927e-01 2.26161405e-01
2.81394899e-01 -1.16347298e-01 -4.01254803e-01 -3.69753689e-01
-9.22908306e-01 6.90524399e-01 1.32933050e-01 4.85503405e-01
-2.44898930e-01 -3.56856227e-01 2.77975976e-01 5.10191500e-01
1.06394313e-01 1.05865693e+00 -5.82724154e-01 -1.58147275e-01
-8.22958231e-01 4.26503867e-01 7.69154429e-01 9.61149096e-01
6.95670068e-01 -2.66171515e-01 -1.68926373e-01 9.81693566e-01
1.53993264e-01 4.00860667e-01 7.55904675e-01 -5.97366810e-01
3.22407275e-01 8.80602956e-01 6.90350682e-02 -6.35890841e-01
-2.84447968e-01 -6.91640615e-01 5.18237380e-03 1.09558418e-01
5.10874033e-01 3.27877253e-02 -7.16250479e-01 1.75040007e+00
1.78679168e-01 -5.09582758e-01 6.29120916e-02 5.80077112e-01
9.60596621e-01 4.15790290e-01 3.50501627e-01 -2.37690479e-01
1.50783074e+00 -7.46215105e-01 -5.68206906e-01 -5.64730465e-01
1.04182959e+00 -9.59014714e-01 1.67266810e+00 3.86076719e-01
-1.00980723e+00 -5.95620096e-01 -1.08882678e+00 -4.30956304e-01
-5.81205070e-01 5.96257597e-02 2.88396835e-01 8.02107453e-01
-7.88778543e-01 8.79484951e-01 -2.27821112e-01 -4.03218716e-01
-2.19847977e-01 2.46221408e-01 -2.08363190e-01 -1.86926708e-01
-1.37653780e+00 1.44479990e+00 5.02228498e-01 -4.81263280e-01
-2.92530537e-01 -8.90776396e-01 -8.50896060e-01 3.51408482e-01
3.36315215e-01 -6.96426153e-01 1.32669067e+00 -1.25478196e+00
-1.26370466e+00 6.96654618e-01 -3.16527337e-01 -1.48491725e-01
1.98056474e-01 -4.58275825e-01 -1.97783917e-01 -2.17348978e-01
8.83386657e-02 5.00186205e-01 6.74011648e-01 -8.07954073e-01
-8.70725513e-01 -1.45629123e-01 3.49490672e-01 7.82642841e-01
-4.31812197e-01 1.30300343e-01 -4.05037999e-01 -8.92794371e-01
1.91293985e-01 -8.48125935e-01 1.50241300e-01 -3.88414979e-01
4.05365489e-02 -6.25315249e-01 1.38442650e-01 -7.60118783e-01
1.65518975e+00 -2.00485492e+00 3.15379500e-01 2.73931175e-01
3.26052457e-02 5.23375273e-01 -4.82868142e-02 5.59429526e-01
-1.20513357e-01 3.33745778e-01 1.83267087e-01 -4.23947036e-01
-1.15838915e-01 1.74154975e-02 -1.39988706e-01 -1.78898409e-01
-1.62567124e-01 6.82370842e-01 -1.23179138e+00 -2.63564438e-01
1.19634628e-01 2.75936395e-01 -6.67442501e-01 4.92947400e-02
-1.93995953e-01 -3.30611542e-02 -4.62585501e-02 1.82913706e-01
1.09495744e-01 2.87295669e-01 2.34851226e-01 -5.97665757e-02
-2.03275010e-01 1.07204294e+00 -1.19715118e+00 1.40298522e+00
-8.21126342e-01 4.76363033e-01 -4.41234589e-01 -4.84543920e-01
6.83482111e-01 2.64313251e-01 -1.01900078e-01 -7.54786313e-01
-3.63950767e-02 4.51098174e-01 3.82979840e-01 -2.01024801e-01
7.47257650e-01 -2.50811130e-01 -5.72356991e-02 4.82963234e-01
-8.03589672e-02 -1.77667171e-01 5.34547687e-01 2.24622384e-01
8.27473700e-01 1.68271527e-01 4.97976243e-01 -4.90084678e-01
3.39147955e-01 3.78033891e-02 3.24171394e-01 8.30134034e-01
3.04598004e-01 1.78364053e-01 1.90941781e-01 -2.42604792e-01
-9.11034286e-01 -1.02630150e+00 1.50207907e-01 1.52962184e+00
2.97490150e-01 -1.02820945e+00 -8.27008665e-01 -5.62680483e-01
-1.45399272e-01 1.35394764e+00 -2.13463858e-01 -3.94614846e-01
-7.96539307e-01 1.34299308e-01 3.82519782e-01 1.97493479e-01
-1.38578832e-01 -1.08623648e+00 -8.44688296e-01 1.99304402e-01
-3.15156251e-01 -6.91327274e-01 -4.98164862e-01 -1.47824967e-03
-8.22792232e-01 -6.74682319e-01 -4.22350109e-01 -9.71842349e-01
8.45705092e-01 3.16603005e-01 1.19322300e+00 6.21120095e-01
2.27669477e-01 -2.14716215e-02 -6.56597495e-01 -5.85252702e-01
-6.32714808e-01 2.34123722e-01 6.02743700e-02 -5.60881674e-01
6.58941984e-01 -1.07936420e-01 -2.67149687e-01 -6.60156831e-03
-7.42911935e-01 1.81854278e-01 3.51186782e-01 6.47963941e-01
6.49222314e-01 2.79206544e-01 3.48298550e-01 -1.09870803e+00
9.33688998e-01 -2.20673963e-01 -2.46350124e-01 3.60300630e-01
-9.59153891e-01 3.20501924e-01 1.07699120e+00 -4.72157359e-01
-9.14189994e-01 1.00000031e-01 -1.40583307e-01 2.75406614e-03
-2.47671455e-02 7.73149252e-01 -1.34351656e-01 -1.78911593e-02
8.79972994e-01 1.37332231e-01 -2.88086563e-01 -7.94817269e-01
6.54128313e-01 3.17349911e-01 2.92811692e-01 -5.64023137e-01
6.48528576e-01 -2.25989550e-01 -4.21128899e-01 -8.52845728e-01
-8.70667636e-01 -5.08821189e-01 -4.99585956e-01 -8.47276002e-02
3.25946599e-01 -7.29901016e-01 -3.25352818e-01 3.28414552e-02
-1.18173444e+00 -5.67300431e-02 -3.88094932e-01 4.40603495e-01
7.11264238e-02 4.68446791e-01 -1.78138077e-01 -3.13628554e-01
-1.85958892e-01 -1.31177318e+00 6.11993372e-01 3.51047873e-01
-1.10711968e+00 -7.57860780e-01 -2.95379877e-01 8.67927372e-02
4.62948948e-01 -5.14291465e-01 1.59716606e+00 -1.03049862e+00
-2.04552114e-01 -2.32610658e-01 2.42134884e-01 3.32813442e-01
1.08687624e-01 -5.29691167e-02 -5.46213806e-01 -7.81100895e-03
-1.05614729e-01 1.80853546e-01 7.19374061e-01 8.71689171e-02
8.02728117e-01 -2.32057944e-01 -1.74692333e-01 4.22651112e-01
1.17513824e+00 1.86615497e-01 5.45461178e-01 2.82188773e-01
8.21949303e-01 7.25800276e-01 5.60322702e-01 4.02073897e-02
4.62359637e-01 8.31957102e-01 -1.05604112e-01 7.00784326e-02
-3.48510832e-01 -5.74180663e-01 5.17826498e-01 9.48365927e-01
2.31860667e-01 -3.52345616e-01 -8.93844306e-01 6.38842225e-01
-1.41120505e+00 -5.80922127e-01 -3.42869163e-02 2.57497525e+00
1.12164402e+00 3.78123313e-01 7.39002749e-02 2.62033790e-01
5.73493004e-01 -2.35112637e-01 -1.06308959e-01 -8.17652166e-01
-6.63461490e-03 8.29329491e-01 2.77383566e-01 1.09456277e+00
-6.32178903e-01 1.52096069e+00 6.19684315e+00 9.09876406e-01
-1.01517141e+00 -1.24779612e-01 1.14950970e-01 -1.47839919e-01
-8.06241870e-01 3.13085347e-01 -9.61170077e-01 3.14897180e-01
9.18171763e-01 -6.40779912e-01 6.14442289e-01 6.09939694e-01
-2.75800899e-02 -1.84163794e-01 -1.26991427e+00 6.95909142e-01
1.21907353e-01 -9.91258025e-01 4.29828376e-01 -5.01248717e-01
3.23340118e-01 -1.90389991e-01 -2.02890679e-01 3.54086667e-01
4.98482324e-02 -1.10767329e+00 9.94172156e-01 5.15771747e-01
4.47845817e-01 -9.72307324e-01 5.21080673e-01 4.23106790e-01
-9.65615928e-01 -4.21327801e-04 -1.45528361e-01 -3.49712878e-01
-1.42154455e-01 2.25040033e-01 -7.22476959e-01 7.16398954e-02
4.14788604e-01 1.17637239e-01 -8.87385845e-01 8.60752404e-01
-8.07890654e-01 3.93510967e-01 -3.52042168e-01 -5.92353702e-01
-1.06178239e-01 -1.37507603e-01 8.81600380e-01 1.30071449e+00
3.97330880e-01 1.14250556e-01 1.34251893e-01 4.83688712e-01
-5.08840084e-02 9.53489840e-01 -4.99099672e-01 -5.75301128e-05
1.12142706e+00 9.40508485e-01 -6.33354843e-01 -5.26517153e-01
-4.10633832e-01 1.16240263e+00 5.49424589e-01 3.31933424e-02
-1.97784543e-01 -7.38451123e-01 5.10795236e-01 2.85086513e-01
2.60735542e-01 -2.31979370e-01 -4.96757001e-01 -9.03878212e-01
1.86117455e-01 -1.18955100e+00 4.91725564e-01 -6.76840127e-01
-9.03096616e-01 5.28202713e-01 1.16714150e-01 -8.95731211e-01
-1.95412084e-01 -5.92462003e-01 -3.40486258e-01 1.18011916e+00
-1.07560861e+00 -6.18586838e-01 -5.87484911e-02 1.59218729e-01
7.42808521e-01 5.08778393e-02 9.82654691e-01 2.00358555e-01
-3.67062777e-01 7.01214373e-01 -1.44217595e-01 -2.69284308e-01
6.16690397e-01 -1.39698756e+00 5.06061852e-01 1.20602608e+00
3.81624311e-01 1.32515085e+00 1.13172102e+00 -8.69115114e-01
-9.21290219e-01 -6.72119856e-01 1.74620390e+00 -3.93859595e-01
4.24513191e-01 4.31282520e-02 -1.06195092e+00 2.92643189e-01
6.77720606e-02 -7.21641898e-01 7.25369573e-01 2.16702908e-01
-4.30330873e-01 -1.10289324e-02 -8.31657767e-01 1.13370788e+00
8.88810039e-01 -6.42999232e-01 -1.14300168e+00 2.55538791e-01
7.46929467e-01 -3.33228260e-01 -6.16497099e-01 1.18946387e-02
5.55976450e-01 -6.87131107e-01 7.96946645e-01 -6.02642119e-01
2.23621368e-01 -4.88922387e-01 1.28136814e-01 -1.86886108e+00
-4.44321424e-01 -7.49458611e-01 1.28153011e-01 9.82584536e-01
7.27076471e-01 -3.25452447e-01 3.41385722e-01 7.02593684e-01
-3.65764707e-01 -6.74923241e-01 -8.73816609e-01 -7.85794616e-01
2.13153750e-01 -7.62519687e-02 6.88368917e-01 8.49876046e-01
3.64400536e-01 6.75052941e-01 7.95928091e-02 -1.11098476e-01
1.64017916e-01 -1.58157889e-02 3.71897846e-01 -1.51164770e+00
-2.57048249e-01 -8.22839499e-01 -6.41099811e-02 -9.87950027e-01
1.46316975e-01 -1.17440355e+00 1.92610711e-01 -1.55307937e+00
-1.61297694e-01 -3.88176888e-01 -2.57663250e-01 4.82174933e-01
-5.59028268e-01 -1.78088732e-02 2.19939813e-01 -5.58763929e-02
-3.58091593e-01 2.53788270e-02 8.37670684e-01 2.97807515e-01
-4.18434143e-01 -1.16136014e-01 -1.00770974e+00 8.74356329e-01
7.75407910e-01 -7.55328298e-01 -5.35965085e-01 -8.13963175e-01
5.96947551e-01 -4.92637485e-01 -2.33844265e-01 -8.21918964e-01
1.40009746e-01 -1.92210615e-01 1.58812732e-01 -1.46401539e-01
-2.08151527e-02 -7.21846581e-01 1.13424368e-01 5.89384496e-01
-4.70631301e-01 5.53732038e-01 2.42897496e-01 -1.45442411e-01
1.31970108e-01 -9.62794781e-01 7.96516776e-01 -1.13661371e-01
-8.41982603e-01 -2.61057079e-01 -4.35321152e-01 2.29482092e-02
6.40476048e-01 -2.68752068e-01 -1.65667713e-01 -2.42399886e-01
-5.24932742e-01 -4.96456325e-02 7.52414167e-01 5.87385535e-01
4.91398126e-01 -1.05743349e+00 -6.27161801e-01 7.52147585e-02
3.53478074e-01 -2.60195792e-01 -3.22432339e-01 3.37852389e-01
-7.17960298e-01 3.02511722e-01 -4.73047607e-02 1.73559055e-01
-1.66169667e+00 4.08011734e-01 4.32175905e-01 -7.67660812e-02
-6.55092120e-01 1.06242013e+00 -1.04404680e-01 -2.64793694e-01
3.22782040e-01 -5.05847573e-01 -4.57240790e-01 3.62936378e-01
6.05984211e-01 3.15834224e-01 5.18456697e-01 -8.53123546e-01
-2.45191991e-01 3.80271554e-01 -3.18515450e-01 -1.67316198e-01
1.00087214e+00 -2.36139998e-01 -9.71574336e-02 4.68448818e-01
7.34126508e-01 5.98822594e-01 -5.12183428e-01 -4.54507023e-01
4.29199040e-01 -5.23643792e-01 7.37934262e-02 -8.62845540e-01
-5.34014642e-01 5.39268911e-01 2.66389430e-01 1.97177023e-01
9.68394041e-01 -1.26423463e-01 7.41637945e-01 7.04306364e-01
1.79483175e-01 -1.28870928e+00 -2.40563601e-01 9.41695511e-01
7.72134900e-01 -6.22354805e-01 1.25023752e-01 -5.70279598e-01
-6.56943202e-01 1.20331478e+00 7.22231805e-01 1.53283909e-01
3.18907917e-01 -3.46300274e-01 1.85666203e-01 -6.39143363e-02
-5.99205315e-01 -3.27703208e-01 4.57750142e-01 3.42924863e-01
9.33549285e-01 1.56855553e-01 -9.79396284e-01 7.55215883e-01
-6.97883844e-01 -5.73277175e-01 5.34165025e-01 7.08748400e-01
-7.85623312e-01 -1.46061409e+00 3.93114947e-02 5.69503486e-01
-3.78698409e-01 -7.23847926e-01 -6.85507894e-01 5.15101910e-01
3.91452968e-01 7.64079928e-01 7.14371204e-02 -1.28215164e-01
8.55357111e-01 4.94019717e-01 5.94862103e-01 -1.18058133e+00
-9.20547664e-01 -2.13720068e-01 1.98163033e-01 -4.08265412e-01
2.06561223e-01 -6.50274575e-01 -1.50692415e+00 -2.10325375e-01
-3.04929376e-01 1.59808099e-01 4.80555385e-01 1.23515749e+00
1.59302473e-01 2.66440421e-01 1.00026518e-01 -2.16711715e-01
-3.39198172e-01 -9.00582492e-01 -1.51200995e-01 7.45570958e-01
-2.22577099e-02 -6.27690852e-01 6.51339293e-02 -3.03100869e-02] | [10.863243103027344, 10.272683143615723] |
7e45e9fb-b395-4a2c-bc36-00332fc90d2a | a-unifying-framework-for-differentially | 2307.04733 | null | https://arxiv.org/abs/2307.04733v1 | https://arxiv.org/pdf/2307.04733v1.pdf | A unifying framework for differentially private quantum algorithms | Differential privacy is a widely used notion of security that enables the processing of sensitive information. In short, differentially private algorithms map "neighbouring" inputs to close output distributions. Prior work proposed several quantum extensions of differential privacy, each of them built on substantially different notions of neighbouring quantum states. In this paper, we propose a novel and general definition of neighbouring quantum states. We demonstrate that this definition captures the underlying structure of quantum encodings and can be used to provide exponentially tighter privacy guarantees for quantum measurements. Our approach combines the addition of classical and quantum noise and is motivated by the noisy nature of near-term quantum devices. Moreover, we also investigate an alternative setting where we are provided with multiple copies of the input state. In this case, differential privacy can be ensured with little loss in accuracy combining concentration of measure and noise-adding mechanisms. En route, we prove the advanced joint convexity of the quantum hockey-stick divergence and we demonstrate how this result can be applied to quantum differential privacy. Finally, we complement our theoretical findings with an empirical estimation of the certified adversarial robustness ensured by differentially private measurements. | ['Elham Kashefi', 'Mina Doosti', 'Armando Angrisani'] | 2023-07-10 | null | null | null | null | ['adversarial-robustness'] | ['adversarial'] | [ 5.66185117e-01 2.26793289e-01 3.94994915e-01 -2.79980689e-01
-1.11864471e+00 -1.21346366e+00 1.15602866e-01 2.15159774e-01
-5.41315913e-01 6.95245445e-01 -5.58739193e-02 -4.31709796e-01
-4.56807554e-01 -1.07767975e+00 -8.02090824e-01 -1.27463126e+00
5.55301942e-02 -3.13788466e-02 -1.06855750e-01 -5.74958444e-01
3.81872803e-01 6.41326785e-01 -1.14213204e+00 7.12740943e-02
5.69308400e-01 7.60805964e-01 -4.63503361e-01 9.26596224e-01
4.74572331e-01 1.85774297e-01 -3.34954202e-01 -7.34901309e-01
8.76776159e-01 -6.94549024e-01 -8.27970207e-01 -4.36416626e-01
-1.05784405e-02 -3.07489336e-01 -9.62605357e-01 1.45607483e+00
7.41840065e-01 -5.33950888e-02 3.25374752e-01 -1.16020417e+00
-6.54985130e-01 7.49252319e-01 9.46287960e-02 8.82583037e-02
3.32309932e-01 1.27072766e-01 9.96570468e-01 -1.61300153e-01
5.71855068e-01 6.82263196e-01 6.79454386e-01 7.96002567e-01
-1.72331560e+00 -6.14557683e-01 -8.92387807e-01 -5.96776009e-02
-1.75796402e+00 -7.07449615e-01 4.06423509e-01 -6.43057823e-02
5.17878115e-01 6.83180511e-01 3.85640003e-02 7.05695927e-01
3.59180927e-01 2.99814910e-01 1.39858651e+00 -5.88585019e-01
3.23808938e-01 4.90613192e-01 8.46491680e-02 3.66633087e-01
4.57305491e-01 6.46961093e-01 -3.42083931e-01 -6.08711839e-01
3.46987724e-01 5.80973700e-02 -6.43115044e-01 -6.70170903e-01
-9.99339402e-01 8.20992470e-01 3.86382401e-01 3.97169113e-01
2.19534384e-03 2.43216410e-01 2.34518632e-01 1.01450086e+00
8.41400251e-02 3.54374826e-01 -1.39167085e-01 5.48044173e-03
-3.87542754e-01 3.30779076e-01 1.23232913e+00 1.24751341e+00
9.53914702e-01 -8.64074409e-01 -1.86363593e-01 -1.81992501e-01
-8.55937973e-02 7.54903793e-01 -1.90592527e-01 -1.11160719e+00
3.76937926e-01 -1.25375450e-01 5.09676158e-01 -7.27166831e-01
-2.32542679e-01 -1.39045760e-01 -8.56320381e-01 -1.25738189e-01
5.57370663e-01 -2.60751814e-01 -3.16253245e-01 2.09527349e+00
1.54402152e-01 -1.24688640e-01 6.73238158e-01 6.63191020e-01
-1.10604607e-01 5.13722837e-01 -3.97642285e-01 -2.46443987e-01
1.06504750e+00 -5.20917997e-02 -6.33241236e-01 3.63064498e-01
8.83548796e-01 -4.08280432e-01 4.65004742e-01 1.14692755e-01
-1.08851886e+00 9.05908719e-02 -1.17449963e+00 -2.72081435e-01
-2.97488004e-01 -8.07591259e-01 6.90567434e-01 1.55145478e+00
-1.26904786e+00 1.19126260e+00 -6.28122211e-01 -2.62880892e-01
2.60282248e-01 7.46071756e-01 -5.38490891e-01 -7.03032464e-02
-1.45480132e+00 5.50597250e-01 5.06810248e-01 -7.48536438e-02
-3.64342600e-01 -4.56437439e-01 -6.61374629e-01 3.67629156e-02
1.86550930e-01 -9.74884152e-01 1.04900718e+00 -4.28939998e-01
-1.55009103e+00 1.11951697e+00 -1.47932349e-02 -6.95471227e-01
5.17130494e-01 4.79463160e-01 -2.24229574e-01 3.42848241e-01
-1.07513733e-01 -2.18733564e-01 3.82718533e-01 -7.87286580e-01
-4.91298884e-01 -1.00541174e+00 4.50661898e-01 9.64925811e-02
-7.93519467e-02 -1.45049259e-01 2.34422803e-01 6.88568503e-02
3.20755243e-01 -1.14512742e+00 -3.94951254e-01 -1.96528081e-02
-2.68318623e-01 3.69518757e-01 2.86716461e-01 1.06134564e-02
9.31241393e-01 -2.55567050e+00 2.17817754e-01 5.18936217e-01
2.80746251e-01 -1.41660079e-01 7.53157139e-02 9.29751873e-01
2.52858460e-01 3.54932249e-01 -5.56965292e-01 -5.09637058e-01
6.49896204e-01 3.36486369e-01 -4.29370224e-01 1.22906244e+00
-1.57237768e-01 9.50971484e-01 -8.90243948e-01 6.36672005e-02
-8.69303271e-02 2.53185809e-01 -5.89425743e-01 -1.16700701e-01
3.81703168e-01 6.73652947e-01 -4.56882089e-01 3.48910511e-01
1.31170595e+00 6.40154928e-02 2.12355107e-01 2.05230474e-01
-9.96635929e-02 3.00784051e-01 -1.32763207e+00 1.93478525e+00
-1.26977414e-01 2.10722327e-01 4.89867508e-01 -8.35569024e-01
4.23039287e-01 4.36018288e-01 1.30286947e-01 -7.84021437e-01
5.37815511e-01 5.09038270e-01 -1.21354766e-01 -2.38562986e-01
5.68890631e-01 -8.47192585e-01 -7.26982176e-01 6.97998881e-01
-6.28308505e-02 -3.22271883e-01 -5.16490340e-01 1.94335967e-01
1.23532295e+00 -2.59721935e-01 1.42637834e-01 -5.57576776e-01
3.76814127e-01 -5.11522114e-01 5.66636860e-01 1.14132237e+00
-5.72782218e-01 3.52468997e-01 6.05417132e-01 2.11535662e-01
-1.21519911e+00 -1.26303971e+00 -5.87505758e-01 7.60939240e-01
6.60720289e-01 -3.52960795e-01 -7.99654424e-01 -2.97537059e-01
2.27594346e-01 6.05824471e-01 -6.74568295e-01 -4.19237107e-01
2.89712660e-02 -7.74816453e-01 1.02799261e+00 3.24433623e-03
3.40260506e-01 -3.48950118e-01 -2.42463276e-01 -1.24978736e-01
1.70654565e-01 -1.00518012e+00 -2.76488543e-01 6.02033079e-01
-7.22484708e-01 -7.97769845e-01 -2.47935206e-01 -3.65916193e-01
4.46666986e-01 3.17724794e-01 6.03432953e-01 -1.45835161e-01
2.70133078e-01 6.35408401e-01 -3.43792588e-01 -3.83138597e-01
-7.48104453e-01 -8.23427290e-02 3.15035552e-01 2.71137804e-01
5.99595487e-01 -7.30444312e-01 -6.08191013e-01 -7.76267648e-02
-1.33945465e+00 -4.43407446e-01 3.47787321e-01 7.70753264e-01
5.55105388e-01 2.08949119e-01 -1.57289766e-02 -1.05853331e+00
5.11789620e-01 -4.71385211e-01 -6.50116265e-01 9.43369716e-02
-3.94969732e-01 5.80573738e-01 8.21376681e-01 2.42272779e-01
-7.37746656e-01 2.77577136e-02 -2.22658232e-01 9.64510217e-02
-7.63630643e-02 -1.36401266e-01 -3.55197161e-01 -8.35086286e-01
6.95303798e-01 3.96612376e-01 1.30588859e-01 -1.60494566e-01
7.13409603e-01 9.98612702e-01 5.12305379e-01 -7.11664379e-01
1.11938131e+00 9.62250412e-01 6.45811260e-01 -5.51894009e-01
-4.31947440e-01 -3.42475891e-01 -7.34206557e-01 6.68137193e-01
5.91163099e-01 -6.01178586e-01 -1.19585669e+00 4.79280084e-01
-8.03184092e-01 -1.90139264e-02 -4.30668145e-01 2.89125293e-01
-9.90618408e-01 8.25947881e-01 -7.77260005e-01 -1.29197812e+00
-6.24285340e-02 -1.11305344e+00 1.16207480e+00 9.36195776e-02
3.49204510e-01 -9.87367094e-01 5.94669767e-02 1.36555070e-02
5.01293361e-01 3.43138844e-01 3.94942284e-01 -6.25796258e-01
-6.72612786e-01 -5.67196965e-01 3.17052333e-03 2.31786147e-01
-1.32462725e-01 -5.95936179e-01 -1.27733874e+00 -6.21731460e-01
5.40426731e-01 -1.33984178e-01 8.51985991e-01 -8.33060071e-02
9.46317255e-01 -2.03725651e-01 -1.85038835e-01 9.99957561e-01
1.77443886e+00 -8.05976391e-02 9.80180860e-01 1.87169105e-01
3.57997417e-01 4.70011294e-01 2.21385375e-01 5.86979389e-01
2.20404625e-01 4.89162177e-01 2.65525520e-01 3.59516829e-01
9.37092483e-01 -1.41572252e-01 2.26019174e-01 5.89885294e-01
1.37336865e-01 -8.44738334e-02 -3.03560197e-01 3.95516425e-01
-1.62057459e+00 -1.18320632e+00 -2.63216525e-01 2.73238206e+00
9.06237304e-01 -9.56668928e-02 -4.94710840e-02 2.21749857e-01
6.51992261e-01 -2.46392965e-01 -1.96385309e-01 -8.02264631e-01
-4.79034007e-01 7.28638530e-01 1.41942596e+00 6.58328831e-01
-1.11907518e+00 5.15874326e-01 6.23160982e+00 5.12075365e-01
-6.62899137e-01 4.60830182e-01 2.31209129e-01 9.44532827e-02
-8.59236777e-01 4.20911938e-01 -4.27752674e-01 5.94640315e-01
1.38705444e+00 -4.77641433e-01 5.93614459e-01 2.52886832e-01
-3.47825795e-01 -5.73686883e-02 -1.50365365e+00 1.02948833e+00
-3.77111286e-01 -1.04878449e+00 -4.60677892e-01 5.62555432e-01
6.88609540e-01 -3.24035645e-01 3.02390486e-01 1.79482009e-02
4.40247357e-02 -7.30820596e-01 4.85512733e-01 6.79768860e-01
9.38576341e-01 -1.11258757e+00 8.22409928e-01 4.65437412e-01
-7.72367001e-01 -2.93292224e-01 -6.62243545e-01 -3.14770222e-01
-3.36137116e-02 2.21069604e-01 -1.95915252e-01 1.06727338e+00
3.74969393e-01 8.15938134e-03 -2.96392497e-02 6.78327560e-01
1.86154414e-02 1.87428519e-01 -5.79517841e-01 -6.33001849e-02
2.20618770e-01 -6.20167315e-01 6.37291074e-01 1.08717239e+00
4.22467858e-01 5.15332997e-01 -4.07401890e-01 1.15649807e+00
-3.00417483e-01 -1.83466598e-01 -1.01924872e+00 -4.72809486e-02
6.81980610e-01 7.66796172e-01 -1.59660786e-01 -1.44808076e-03
-2.23517627e-01 1.76459610e+00 1.11211301e-03 1.07502453e-01
-4.79132593e-01 -7.92494357e-01 9.68135953e-01 -3.53447676e-01
3.12685221e-01 -1.66509226e-01 -3.34824800e-01 -1.35079551e+00
2.01636091e-01 -4.11496878e-01 2.89731026e-01 9.84115601e-02
-1.46710885e+00 -1.31382927e-01 -4.47889537e-01 -1.06920803e+00
3.34115028e-02 -4.83707845e-01 -2.41168052e-01 1.23387611e+00
-1.43806589e+00 -5.77991188e-01 3.30245167e-01 6.63510442e-01
-8.87916565e-01 4.12848204e-01 1.36466217e+00 2.78665572e-01
-2.79987633e-01 9.83474791e-01 1.03935146e+00 1.09219961e-01
6.16211414e-01 -1.62871659e+00 6.13469183e-01 1.02302301e+00
1.91422403e-02 1.04090714e+00 7.45052516e-01 -1.55668139e-01
-2.16487670e+00 -7.98769951e-01 1.02782595e+00 -8.77557576e-01
7.68528581e-01 -7.01779664e-01 -6.67467237e-01 7.74082839e-01
-1.42333359e-01 7.62216523e-02 1.09333968e+00 -2.19497547e-01
-4.08930510e-01 -1.22779407e-01 -1.82117069e+00 2.35820696e-01
1.04045868e+00 -1.41533768e+00 -4.06903207e-01 3.12504590e-01
8.03468883e-01 -5.38917124e-01 -9.43483770e-01 -2.51466502e-02
5.28436959e-01 -1.17472172e+00 6.59267128e-01 -5.07680297e-01
-2.82368988e-01 -3.67793560e-01 -7.55452991e-01 -8.40177655e-01
-1.96621761e-01 -1.35548043e+00 1.61705971e-01 9.25160587e-01
2.25658864e-01 -1.06791139e+00 5.77525973e-01 1.02462256e+00
2.06189170e-01 -2.38810137e-01 -1.27642703e+00 -8.07524741e-01
5.45251787e-01 -4.55930561e-01 6.50900364e-01 8.42731535e-01
3.71258825e-01 -1.85268328e-01 -3.97908151e-01 6.38143122e-01
1.02998698e+00 1.59137161e-03 7.20992506e-01 -9.10573125e-01
-6.77044094e-01 -1.72973469e-01 -1.04838121e+00 -1.20140576e+00
-9.98260006e-02 -1.10491002e+00 -9.59817171e-02 -7.16856837e-01
3.48003805e-01 -4.48950320e-01 -5.07764399e-01 -2.48880148e-01
-1.27750356e-02 4.47507113e-01 1.17888190e-01 1.66619103e-02
-5.91062725e-01 3.97889942e-01 8.52189600e-01 2.30941877e-01
-4.97772768e-02 3.53081793e-01 -1.01853776e+00 -4.75672111e-02
6.95950627e-01 -5.65770984e-01 -1.32855967e-01 -6.18940108e-02
5.46951175e-01 9.61657241e-02 5.58125019e-01 -9.24364209e-01
4.94413525e-01 1.55607894e-01 -3.75666708e-01 1.22242622e-01
2.62811095e-01 -7.59855151e-01 2.90479928e-01 5.63712597e-01
-2.17505306e-01 -4.44045335e-01 -1.50164038e-01 9.23380017e-01
2.32413441e-01 -3.23140353e-01 9.16883588e-01 -2.46275198e-02
-3.98011118e-01 3.38336617e-01 -9.33647994e-03 -9.39062685e-02
1.15876842e+00 -2.60910600e-01 -2.10245445e-01 -3.92300606e-01
-7.89037883e-01 2.65875421e-02 1.07571948e+00 -2.86258191e-01
2.79954046e-01 -1.21096742e+00 -6.14894867e-01 3.23975563e-01
3.08022648e-01 -5.11452258e-01 4.75476265e-01 1.13362706e+00
-5.29515445e-01 4.68406767e-01 -2.81071868e-02 -2.14983493e-01
-8.65015030e-01 7.52868176e-01 6.12307966e-01 3.69671911e-01
-4.55537260e-01 9.22058761e-01 -4.47965926e-03 -3.58129770e-01
4.22209278e-02 -1.71257749e-01 8.32262635e-01 -5.11313260e-01
6.47563994e-01 1.27866045e-01 1.45963570e-02 -7.30426371e-01
-3.01243931e-01 3.77235740e-01 2.28963211e-01 -3.77784789e-01
9.48391080e-01 -7.08735943e-01 -3.96441489e-01 3.91418248e-01
1.70726991e+00 3.60781848e-01 -8.92046452e-01 -4.00549948e-01
-2.95626260e-02 -5.03393233e-01 -3.24978352e-01 -1.49940342e-01
-5.36689460e-01 8.49715352e-01 7.84488976e-01 6.08692467e-01
1.27953076e+00 -3.46610160e-03 9.12685812e-01 5.89743316e-01
1.03305519e+00 -1.00998533e+00 -1.09326029e+00 2.65334576e-01
7.13853762e-02 -1.14209282e+00 -3.82596076e-01 -2.67787308e-01
-1.21626496e-01 9.96944368e-01 -4.80804741e-01 -5.77796362e-02
5.06018519e-01 2.35783577e-01 -2.17463762e-01 4.75181546e-03
-2.62987345e-01 -3.60036373e-01 -2.23304316e-01 6.48668408e-01
-1.20797800e-02 4.72018689e-01 -5.97974360e-01 5.08829534e-01
-3.02828640e-01 -3.06050569e-01 7.17308700e-01 1.12335813e+00
-2.08601877e-01 -1.62368643e+00 -3.14641654e-01 1.47748459e-02
-8.79142880e-01 -2.15130016e-01 -4.60700780e-01 4.33242321e-01
7.03851357e-02 9.79560792e-01 -1.51976034e-01 -5.24579763e-01
2.31020078e-01 4.20948043e-02 7.60365665e-01 -4.62517887e-01
-3.79587620e-01 -4.89465773e-01 -3.12871546e-01 -6.93088949e-01
-4.41445619e-01 -5.93372047e-01 -1.14938629e+00 -1.04437888e+00
-4.15481895e-01 4.65632498e-01 5.89440286e-01 8.54861557e-01
6.11736655e-01 -1.14009738e-01 1.10904682e+00 -3.80257756e-01
-1.35011923e+00 -3.40123236e-01 -1.40356863e+00 5.92810035e-01
7.60321498e-01 7.63238594e-02 -8.03898394e-01 -5.51272810e-01] | [5.788342475891113, 5.983330249786377] |
99508462-e746-4d98-a72d-ed8c2406d09c | mining-mid-level-features-for-action | 1409.4014 | null | http://arxiv.org/abs/1409.4014v1 | http://arxiv.org/pdf/1409.4014v1.pdf | Mining Mid-level Features for Action Recognition Based on Effective Skeleton Representation | Recently, mid-level features have shown promising performance in computer
vision. Mid-level features learned by incorporating class-level information are
potentially more discriminative than traditional low-level local features. In
this paper, an effective method is proposed to extract mid-level features from
Kinect skeletons for 3D human action recognition. Firstly, the orientations of
limbs connected by two skeleton joints are computed and each orientation is
encoded into one of the 27 states indicating the spatial relationship of the
joints. Secondly, limbs are combined into parts and the limb's states are
mapped into part states. Finally, frequent pattern mining is employed to mine
the most frequent and relevant (discriminative, representative and
non-redundant) states of parts in continuous several frames. These parts are
referred to as Frequent Local Parts or FLPs. The FLPs allow us to build
powerful bag-of-FLP-based action representation. This new representation yields
state-of-the-art results on MSR DailyActivity3D and MSR ActionPairs3D. | ['Pichao Wang', 'Zhimin Gao', 'Philip Ogunbona', 'Wanqing Li', 'Hanling Zhang'] | 2014-09-14 | null | null | null | null | ['3d-human-action-recognition'] | ['computer-vision'] | [ 1.48506895e-01 -2.34925568e-01 -8.46544921e-01 -4.34301764e-01
-7.21094966e-01 -1.01342112e-01 6.91198111e-01 -6.74435273e-02
-2.15029478e-01 5.35289407e-01 7.26692140e-01 6.35847509e-01
-2.29551524e-01 -4.56322789e-01 -3.60402942e-01 -7.61603117e-01
-4.73423690e-01 1.93160713e-01 6.57480717e-01 -3.81336734e-02
3.10326874e-01 8.60453010e-01 -2.03204727e+00 6.56509042e-01
9.37403813e-02 1.19684291e+00 9.55755264e-02 4.58530307e-01
-1.48574233e-01 9.58327830e-01 -4.31138426e-01 1.79959819e-01
2.19094083e-01 -7.21696317e-01 -7.45020688e-01 6.07419908e-01
2.89326519e-01 -3.51608008e-01 -3.68305653e-01 6.54766083e-01
3.90423089e-01 3.41838509e-01 7.16900468e-01 -1.12446952e+00
2.95356691e-01 1.50719017e-03 -9.26118851e-01 2.56957620e-01
1.17615986e+00 1.03474028e-01 1.04386425e+00 -9.90924060e-01
8.64270806e-01 1.37872148e+00 3.05157602e-01 3.45869273e-01
-8.45134377e-01 -2.39865765e-01 2.54619092e-01 8.03516388e-01
-1.18832421e+00 -2.40519017e-01 8.83856535e-01 -3.63294303e-01
1.33565390e+00 1.24087386e-01 1.34413064e+00 1.06300426e+00
5.93105435e-01 1.18638563e+00 1.01300871e+00 -3.88732076e-01
2.89360166e-01 -6.96953237e-01 8.54959264e-02 9.69454288e-01
-6.64288774e-02 -7.37430602e-02 -1.26464760e+00 -1.80581093e-01
9.33259010e-01 4.45684612e-01 1.93241805e-01 -6.11024559e-01
-1.40962112e+00 5.93967915e-01 1.39833152e-01 3.38623554e-01
-8.79489005e-01 3.04468989e-01 3.76256794e-01 -4.70758714e-02
2.04507023e-01 -2.40494773e-01 -4.30929810e-01 -7.40188360e-01
-5.88773727e-01 4.04699355e-01 2.03657031e-01 7.15039372e-01
8.59205484e-01 -3.05248916e-01 -3.59103322e-01 8.48350048e-01
3.99914294e-01 3.01096052e-01 6.41911149e-01 -1.27469683e+00
3.95302385e-01 1.16376865e+00 -1.58468038e-01 -9.46057975e-01
-5.52128613e-01 3.09173584e-01 -6.20493412e-01 2.45620102e-01
1.39196843e-01 4.01774347e-01 -9.60765004e-01 1.12942171e+00
7.31232285e-01 8.28567892e-02 -2.98190832e-01 8.51143420e-01
5.27441263e-01 4.03770119e-01 -2.48911511e-02 -2.10716799e-01
1.54576755e+00 -8.01946759e-01 -7.25176096e-01 -2.27353707e-01
3.72333139e-01 -5.10382771e-01 7.31868207e-01 3.83334845e-01
-9.44212258e-01 -7.06019282e-01 -7.60800838e-01 1.76984921e-01
-1.74216807e-01 1.65144369e-01 7.88533628e-01 1.82852745e-01
-5.31749547e-01 6.45366549e-01 -1.29782856e+00 -3.62946928e-01
4.30734694e-01 1.91344142e-01 -8.16631496e-01 -9.34304297e-02
-9.44057643e-01 9.32697654e-01 2.75154442e-01 1.38456998e-02
-8.94000351e-01 1.69654429e-01 -1.21334946e+00 -5.55628300e-01
3.32430273e-01 -3.92492831e-01 1.01909447e+00 -6.06449127e-01
-1.57214534e+00 9.65221167e-01 -3.79898816e-01 -1.42449319e-01
9.87964794e-02 -2.48621270e-01 -2.04263210e-01 6.97887421e-01
1.96368009e-01 6.39239192e-01 1.06905246e+00 -6.78305745e-01
-9.49862421e-01 -7.61996567e-01 -7.07582608e-02 5.30345142e-01
-1.59428284e-01 1.56354114e-01 -7.28815913e-01 -6.69310927e-01
7.47370481e-01 -8.74040246e-01 -3.01033765e-01 2.65430331e-01
-1.16335422e-01 -6.94861650e-01 6.08787179e-01 -4.98344392e-01
1.04121542e+00 -2.02098489e+00 4.75019902e-01 3.15404028e-01
-1.91566914e-01 -1.72610983e-01 -1.37457894e-02 2.21822023e-01
-1.27275676e-01 -5.78446925e-01 1.00574367e-01 -1.89113781e-01
-4.92658764e-02 7.25018799e-01 3.91806722e-01 6.96958244e-01
1.50667563e-01 7.59549320e-01 -9.34263170e-01 -8.71740341e-01
7.53503919e-01 3.04958999e-01 -1.13324761e-01 9.08813477e-02
9.19317529e-02 3.58691990e-01 -8.32330942e-01 1.01589048e+00
5.74265681e-02 1.75931484e-01 8.40779990e-02 -2.00296268e-01
2.74341330e-02 3.13339442e-01 -1.40603912e+00 2.09803581e+00
8.14259574e-02 6.24686256e-02 -4.11894292e-01 -9.59864795e-01
9.49737132e-01 2.03573704e-01 1.24341643e+00 -6.82341635e-01
-1.58428226e-03 -1.30984977e-01 -4.74549711e-01 -6.62419260e-01
2.39053786e-01 2.36249361e-02 -3.72700274e-01 3.95575732e-01
1.64954767e-01 -5.64458817e-02 3.51171374e-01 -1.10329956e-01
1.32004154e+00 6.99110568e-01 8.45090210e-01 1.41116634e-01
5.76172233e-01 -7.67359212e-02 6.85556889e-01 3.10769349e-01
-5.12649655e-01 5.45937121e-01 3.72786105e-01 -5.72067201e-01
-3.35863024e-01 -1.15975618e+00 2.76182503e-01 8.13412130e-01
1.30094662e-01 -6.76263630e-01 -3.40133071e-01 -7.61396945e-01
1.35399383e-02 1.22609273e-01 -7.43332565e-01 -3.78537983e-01
-6.92235708e-01 -1.99296743e-01 3.19063008e-01 9.18001473e-01
5.85661590e-01 -1.21331513e+00 -1.22619617e+00 2.14063287e-01
-3.27442050e-01 -9.19350445e-01 -3.13077956e-01 3.73322666e-01
-1.23831379e+00 -1.25237870e+00 -7.06465304e-01 -6.61308408e-01
5.07277608e-01 3.16197395e-01 8.46054614e-01 -4.31600899e-01
-6.14868581e-01 5.33009291e-01 -8.47801685e-01 3.52606140e-02
7.23811090e-02 -5.27560771e-01 2.66149521e-01 1.69526771e-01
5.36766827e-01 -6.70230269e-01 -6.38503969e-01 4.83682066e-01
-4.88423079e-01 -2.15911031e-01 8.18947375e-01 5.37907004e-01
1.20278895e+00 3.42384838e-02 -4.43117581e-02 1.06363907e-01
4.52132896e-02 -1.39560223e-01 1.00429170e-01 1.44231305e-01
1.30568653e-01 1.21145301e-01 1.04122087e-01 -4.43254173e-01
-9.51941907e-01 7.33866274e-01 8.43046512e-03 -5.76700926e-01
-5.11906147e-01 5.07351875e-01 -1.98148832e-01 2.31506169e-01
5.45899630e-01 4.93393689e-01 3.65799516e-02 -6.08297110e-01
3.07014197e-01 4.70760763e-01 3.85110229e-01 -6.11121297e-01
3.61786634e-01 6.80929422e-01 3.36012214e-01 -9.21808064e-01
-7.43438601e-01 -9.47254539e-01 -1.16098714e+00 -7.62017310e-01
1.01342320e+00 -8.87739837e-01 -5.09196341e-01 7.48202562e-01
-7.80599058e-01 -2.04074271e-02 -6.38398409e-01 7.78032541e-01
-1.08509278e+00 7.20247447e-01 -6.18717432e-01 -9.06913817e-01
5.81465922e-02 -8.44428599e-01 1.42272055e+00 9.96254459e-02
-5.39490163e-01 -2.45265990e-01 3.16853702e-01 3.97007078e-01
-4.42836881e-01 4.78715092e-01 5.27627945e-01 -1.10244825e-01
-1.80212364e-01 -4.85748738e-01 4.01343495e-01 3.06755662e-01
5.82455039e-01 -1.65555980e-02 -4.28978711e-01 1.02695242e-01
-1.39087871e-01 -5.16195834e-01 7.85066485e-01 5.77372134e-01
1.00636899e+00 5.66721745e-02 -4.03648853e-01 1.28320098e-01
8.95275593e-01 1.47392288e-01 7.78019607e-01 2.76124179e-01
5.17810822e-01 5.25902987e-01 1.30907023e+00 8.97698879e-01
1.46337420e-01 9.91779149e-01 3.24987471e-01 3.05108637e-01
-1.83829486e-01 -1.33969396e-01 8.10046136e-01 5.48430204e-01
-8.09871376e-01 5.47832489e-01 -6.25847399e-01 3.91456783e-01
-2.12775087e+00 -1.32010424e+00 -6.51095016e-03 1.83536959e+00
9.41796720e-01 3.13039958e-01 4.69861835e-01 4.45914537e-01
4.85794425e-01 4.75298882e-01 -4.43471789e-01 -1.73601046e-01
-8.97443481e-03 2.92792946e-01 1.24432787e-01 1.83174480e-02
-1.23865879e+00 7.32399881e-01 6.01224184e+00 9.16314304e-01
-4.48901325e-01 -1.04990907e-01 5.75816520e-02 -2.72613645e-01
3.00198019e-01 -6.83322549e-02 -8.80725086e-01 2.14770034e-01
5.72699130e-01 4.23384011e-01 -1.69657499e-01 1.10342407e+00
3.05054337e-01 -7.28618503e-01 -1.03005934e+00 1.15129483e+00
2.32289985e-01 -9.05460656e-01 5.15407287e-02 7.61206150e-02
5.68866730e-01 -2.09349558e-01 -3.75777066e-01 1.39679372e-01
-3.51235474e-04 -6.11856997e-01 7.33088315e-01 8.40602338e-01
5.60653687e-01 -8.95056844e-01 3.45106483e-01 3.72545719e-01
-1.74031150e+00 -2.19674017e-02 -3.25149059e-01 -2.70215750e-01
2.80761391e-01 4.51104879e-01 -4.23768699e-01 7.00333595e-01
8.79306912e-01 1.36243474e+00 -5.58575273e-01 7.31319547e-01
-3.89339983e-01 3.00289422e-01 -4.29830492e-01 -5.16697317e-02
2.35370502e-01 -1.54229596e-01 5.11419833e-01 8.98793697e-01
1.07926041e-01 2.43939355e-01 4.06316489e-01 1.94795579e-01
6.11191094e-01 -1.89934149e-01 -6.11454308e-01 -3.35801044e-03
-1.06870681e-01 1.10835779e+00 -8.59295964e-01 -3.06190819e-01
-5.35795510e-01 1.22950506e+00 3.70716415e-02 -3.15866470e-02
-7.44104207e-01 -8.35161433e-02 1.07639670e+00 -9.16992500e-03
3.41717809e-01 -5.45525908e-01 1.46772951e-01 -1.17091537e+00
2.78655589e-01 -8.44534338e-01 7.73742735e-01 -5.84002912e-01
-1.15577412e+00 5.28438203e-02 2.87351757e-01 -1.73059404e+00
-5.70267498e-01 -5.23862541e-01 -2.44440198e-01 3.67338806e-01
-1.02227712e+00 -1.09905517e+00 -2.93080986e-01 1.13607872e+00
1.03800046e+00 -4.75134999e-02 1.05736923e+00 -3.60197313e-02
-2.71570832e-01 1.21288776e-01 -2.63368875e-01 2.02855498e-01
3.04471254e-01 -9.98268485e-01 1.57035977e-01 6.18819654e-01
4.94184107e-01 3.72629285e-01 3.62261772e-01 -9.03097272e-01
-1.42019522e+00 -6.93434536e-01 8.78436744e-01 -4.18709457e-01
4.02932793e-01 -6.76617846e-02 -2.40253583e-01 3.99068773e-01
-5.62283099e-01 3.00569713e-01 5.97183287e-01 -1.91153988e-01
-1.28755033e-01 -1.28696278e-01 -9.40967441e-01 5.44973969e-01
1.49803090e+00 -4.89067554e-01 -1.00153983e+00 2.69047618e-01
-7.17219487e-02 -2.38897055e-01 -9.10391271e-01 3.80013436e-01
8.65887046e-01 -1.07761478e+00 1.24639750e+00 -6.79037809e-01
3.42026591e-01 -4.02046531e-01 -3.97731841e-01 -7.81054497e-01
-3.35278630e-01 -2.21442774e-01 -7.50177264e-01 6.64014876e-01
-2.02772826e-01 4.19537462e-02 1.20170736e+00 4.99434993e-02
-1.89376295e-01 -1.02958584e+00 -1.42051125e+00 -1.01391649e+00
-6.56520247e-01 -4.42411125e-01 1.36557356e-01 4.58101869e-01
4.05263424e-01 1.90378875e-02 -3.55942041e-01 -4.25313979e-01
7.21976995e-01 4.60161179e-01 6.78964198e-01 -1.14973092e+00
-2.63911456e-01 -2.19288662e-01 -1.15770698e+00 -1.31205463e+00
1.58182517e-01 -5.84609926e-01 2.48156488e-01 -1.59394372e+00
2.82777339e-01 1.92607194e-01 -4.70197886e-01 8.83922637e-01
1.88892968e-02 2.52298892e-01 -6.02657385e-02 1.79668665e-01
-8.46025169e-01 6.91041231e-01 1.37967288e+00 -1.19829670e-01
-1.25437304e-01 2.63486534e-01 1.30846187e-01 1.05280912e+00
5.50553203e-01 -4.03874725e-01 -2.73499995e-01 2.52858639e-01
-1.31589159e-01 2.76513249e-01 3.58380884e-01 -1.20669568e+00
5.44380620e-02 -6.10259175e-01 7.93240666e-01 -1.22869265e+00
9.09481764e-01 -8.37910652e-01 1.02032460e-01 7.76276171e-01
-1.78430453e-01 -2.33365983e-01 -3.28266472e-01 6.70955777e-01
-4.12715375e-01 -3.58515009e-02 5.39570212e-01 -5.39584458e-01
-1.32579172e+00 3.65196854e-01 -8.30830216e-01 -2.79608399e-01
1.34469604e+00 -8.91243815e-01 4.20535088e-01 -1.75981045e-01
-1.02235341e+00 -1.07374437e-01 1.25940323e-01 6.27045572e-01
1.01914012e+00 -1.70795977e+00 -2.81827867e-01 2.04671428e-01
3.81891578e-01 -1.83534011e-01 3.36965650e-01 8.92966449e-01
-1.17147945e-01 3.51435006e-01 -6.90865219e-01 -8.40215683e-01
-1.65876925e+00 1.17110208e-01 6.34949505e-02 -2.86903441e-01
-8.77418756e-01 7.78049111e-01 -3.99890184e-01 8.35756883e-02
3.40959191e-01 -4.23995465e-01 -3.79153699e-01 4.45453048e-01
5.47076821e-01 6.73971117e-01 -7.87090808e-02 -1.05562162e+00
-9.02816176e-01 1.00301516e+00 2.90002912e-01 -1.37221277e-01
1.51498103e+00 -3.21257929e-03 -2.06365567e-02 7.06767082e-01
1.10155344e+00 -5.14238238e-01 -1.49776101e+00 -1.81084633e-01
6.92103878e-02 -6.98983073e-01 -3.57958943e-01 -4.55004692e-01
-9.49044168e-01 6.92346573e-01 6.58869147e-01 -2.99852729e-01
1.17454350e+00 4.24152106e-01 7.74566412e-01 4.17740077e-01
8.84583652e-01 -1.41311073e+00 7.41065860e-01 5.22046089e-01
8.59987855e-01 -1.00355852e+00 3.71685058e-01 -2.63789564e-01
-7.56606698e-01 1.15505612e+00 4.62431520e-01 -1.97516724e-01
6.83832347e-01 6.90500438e-02 -2.19010100e-01 -4.73552018e-01
-4.96299088e-01 -6.56462669e-01 5.19259453e-01 6.74800932e-01
5.38636297e-02 7.99072161e-02 -4.99347568e-01 4.73179221e-01
2.74949580e-01 1.74783207e-02 5.43609858e-02 1.59113503e+00
-7.16456890e-01 -1.14899921e+00 -3.18788767e-01 5.09835720e-01
-2.36777812e-01 5.90191841e-01 -5.43437302e-01 6.74016953e-01
2.53159523e-01 8.61043453e-01 -1.82209704e-02 -6.26470149e-01
6.40313268e-01 2.51357585e-01 1.04311132e+00 -7.71682203e-01
-2.73963928e-01 2.33717918e-01 2.22888410e-01 -1.42006660e+00
-1.05923200e+00 -1.26713431e+00 -1.62471855e+00 2.81344771e-01
-5.42930514e-02 -3.01404506e-01 3.27998668e-01 1.05278993e+00
1.21372566e-01 1.34040862e-01 5.51450610e-01 -1.30879807e+00
-4.82133985e-01 -8.25398922e-01 -9.09916997e-01 6.92608058e-01
6.18296266e-02 -1.25694704e+00 -1.01113819e-01 2.65590906e-01] | [7.798377990722656, 0.31536900997161865] |
b9d9374a-1a14-43d6-9330-257d923fe0d1 | surface-representation-for-point-clouds | 2205.05740 | null | https://arxiv.org/abs/2205.05740v2 | https://arxiv.org/pdf/2205.05740v2.pdf | Surface Representation for Point Clouds | Most prior work represents the shapes of point clouds by coordinates. However, it is insufficient to describe the local geometry directly. In this paper, we present \textbf{RepSurf} (representative surfaces), a novel representation of point clouds to \textbf{explicitly} depict the very local structure. We explore two variants of RepSurf, Triangular RepSurf and Umbrella RepSurf inspired by triangle meshes and umbrella curvature in computer graphics. We compute the representations of RepSurf by predefined geometric priors after surface reconstruction. RepSurf can be a plug-and-play module for most point cloud models thanks to its free collaboration with irregular points. Based on a simple baseline of PointNet++ (SSG version), Umbrella RepSurf surpasses the previous state-of-the-art by a large margin for classification, segmentation and detection on various benchmarks in terms of performance and efficiency. With an increase of around \textbf{0.008M} number of parameters, \textbf{0.04G} FLOPs, and \textbf{1.12ms} inference time, our method achieves \textbf{94.7\%} (+0.5\%) on ModelNet40, and \textbf{84.6\%} (+1.8\%) on ScanObjectNN for classification, while \textbf{74.3\%} (+0.8\%) mIoU on S3DIS 6-fold, and \textbf{70.0\%} (+1.6\%) mIoU on ScanNet for segmentation. For detection, previous state-of-the-art detector with our RepSurf obtains \textbf{71.2\%} (+2.1\%) mAP$\mathit{_{25}}$, \textbf{54.8\%} (+2.0\%) mAP$\mathit{_{50}}$ on ScanNetV2, and \textbf{64.9\%} (+1.9\%) mAP$\mathit{_{25}}$, \textbf{47.7\%} (+2.5\%) mAP$\mathit{_{50}}$ on SUN RGB-D. Our lightweight Triangular RepSurf performs its excellence on these benchmarks as well. The code is publicly available at \url{https://github.com/hancyran/RepSurf}. | ['Chengjie Wang', 'Jun Liu', 'Haoxi Ran'] | 2022-05-11 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Ran_Surface_Representation_for_Point_Clouds_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Ran_Surface_Representation_for_Point_Clouds_CVPR_2022_paper.pdf | cvpr-2022-1 | ['3d-point-cloud-classification'] | ['computer-vision'] | [ 9.05342773e-02 1.60586491e-01 2.27838993e-01 -7.28518665e-02
-1.00015843e+00 -7.29095578e-01 4.75435734e-01 1.17913619e-01
-4.11003470e-01 3.68272036e-01 -8.64001095e-01 -4.94224131e-01
-6.27934653e-03 -1.29510164e+00 -1.28351080e+00 -2.49192417e-01
-2.69304752e-01 6.96119368e-01 8.86083484e-01 -1.75659209e-01
3.14690202e-01 7.32753396e-01 -1.87504256e+00 1.19766042e-01
8.08950663e-01 1.42751122e+00 2.63078421e-01 5.63632309e-01
-3.99108052e-01 3.00547406e-02 -4.78890061e-01 -3.18822980e-01
5.35062671e-01 1.42203748e-01 -3.84472340e-01 -4.91652876e-01
8.47300529e-01 -2.04052448e-01 -5.57535887e-02 1.27088058e+00
2.49377042e-01 7.44879022e-02 9.99600947e-01 -1.23150158e+00
-1.45133197e-01 8.05619955e-02 -1.11711681e+00 1.50148213e-01
9.20422450e-02 2.60917664e-01 6.89420044e-01 -1.14163661e+00
4.76592779e-01 1.07628655e+00 8.55324507e-01 2.40235284e-01
-1.16295874e+00 -1.29807091e+00 1.14895195e-01 -4.53471571e-01
-1.70401847e+00 8.57068673e-02 2.13307545e-01 -5.92477381e-01
9.92640853e-01 7.46573269e-01 4.96495724e-01 3.70508820e-01
1.43576786e-01 3.58180106e-01 1.01762438e+00 -7.03804493e-02
1.88018501e-01 -2.44167030e-01 2.77033806e-01 8.15334558e-01
3.27567160e-01 1.36331603e-01 -1.04360856e-01 -2.66897142e-01
1.33311272e+00 2.47584097e-03 -1.13102838e-01 2.21275389e-01
-1.07127166e+00 6.43753886e-01 6.83713973e-01 -2.12350354e-01
-1.19090483e-01 6.39869273e-01 -1.08144730e-01 -6.50792420e-02
5.57586074e-01 9.60718766e-02 -2.89231688e-01 -6.09496087e-02
-1.19869268e+00 3.82832944e-01 5.05728245e-01 1.48019660e+00
1.25584006e+00 1.05389394e-01 7.69038796e-02 8.96522641e-01
3.26521724e-01 1.32084441e+00 -3.56624842e-01 -1.22889531e+00
4.02535379e-01 7.01150417e-01 1.68736607e-01 -9.30909932e-01
-5.18078804e-01 -4.38294709e-01 -9.38987017e-01 6.86919034e-01
4.15630460e-01 1.33723971e-02 -1.11777020e+00 1.19287920e+00
2.63960421e-01 2.28404969e-01 -5.49023390e-01 4.78819370e-01
9.02808070e-01 1.03528714e+00 -5.47269732e-02 6.20658807e-02
1.50383425e+00 -5.04146576e-01 2.72356540e-01 -1.61403656e-01
4.71610278e-01 -1.03335094e+00 1.19082916e+00 5.10196805e-01
-1.21610618e+00 -6.16777599e-01 -8.59710336e-01 1.15429282e-01
-8.89112949e-02 -2.25239933e-01 2.32032567e-01 6.58791840e-01
-1.16119361e+00 5.80325186e-01 -9.98862505e-01 -8.67335498e-02
4.49546874e-01 5.43171287e-01 9.74290222e-02 -1.57585114e-01
-4.11209971e-01 2.84236193e-01 5.57913706e-02 -3.37882161e-01
-5.95748663e-01 -1.05998147e+00 -6.03341222e-01 -6.87637478e-02
3.86270016e-01 -5.30029058e-01 8.55062008e-01 -1.77371621e-01
-8.18035245e-01 9.68690872e-01 -3.65016572e-02 -3.28881562e-01
5.03392696e-01 -1.04731238e-02 -4.47553433e-02 1.46106929e-02
2.45585397e-01 9.75116134e-01 5.26284873e-01 -1.42489839e+00
-7.38012850e-01 -5.71036875e-01 -2.68359259e-02 -1.58129141e-01
1.48000449e-01 6.05329052e-02 -7.50529945e-01 -6.89494491e-01
4.97731000e-01 -1.09551609e+00 -1.65735558e-01 3.25917631e-01
-4.13721889e-01 -6.73120469e-02 7.81612277e-01 -3.07838619e-01
1.00414515e+00 -2.03294563e+00 -4.02755380e-01 5.78764498e-01
3.73531759e-01 2.13100255e-01 3.18517327e-01 9.44960117e-02
2.08557382e-01 4.86407638e-01 -4.83736217e-01 -4.32092249e-01
-6.69059157e-02 1.76583111e-01 -9.68042612e-02 5.38988948e-01
-1.92289144e-01 5.58022082e-01 -6.14716172e-01 -2.80701488e-01
5.81417739e-01 5.24810672e-01 -5.58027923e-01 -2.48644441e-01
-2.48752296e-01 1.86180681e-01 -7.60199606e-01 8.82329404e-01
1.31303692e+00 -1.69879168e-01 -4.68218446e-01 -2.51966715e-01
-5.06832898e-01 -1.51091859e-01 -1.46931136e+00 1.49323559e+00
-1.35916159e-01 2.06553742e-01 5.03982008e-01 -3.12396944e-01
1.17014813e+00 -1.14314377e-01 6.88961208e-01 -6.71978235e-01
2.61810333e-01 4.21048403e-01 -3.93718302e-01 2.60194510e-01
5.45725465e-01 8.54845047e-02 -8.43819082e-02 4.82727647e-01
-3.57990026e-01 -6.91079736e-01 -9.06719565e-02 1.06890537e-01
1.23411536e+00 2.83629060e-01 -1.54579744e-01 -6.21645212e-01
2.20747411e-01 1.25011295e-01 3.25654387e-01 8.28255773e-01
1.91964641e-01 9.52258348e-01 2.41811574e-01 -2.47738585e-01
-9.62423146e-01 -1.20267653e+00 -6.87236190e-01 8.99991095e-01
4.06812012e-01 -4.73927617e-01 -1.00178385e+00 -2.37411156e-01
2.05230996e-01 6.93557620e-01 -3.11825693e-01 2.10234612e-01
-6.84536815e-01 -8.54381680e-01 7.27726460e-01 7.16234803e-01
6.25154674e-01 -9.39397275e-01 -8.30514789e-01 -1.16345799e-02
2.29032084e-01 -8.57472122e-01 -2.34431967e-01 -1.07887737e-01
-9.00812030e-01 -9.39023554e-01 -7.49745011e-01 -2.24121198e-01
6.94195628e-01 2.88412392e-01 1.24165332e+00 3.69596601e-01
-2.46006429e-01 2.11549804e-01 -3.84492874e-01 -6.80533290e-01
-1.46987841e-01 -8.43578428e-02 -1.47506669e-01 -3.90176922e-01
1.51369780e-01 -6.60601437e-01 -8.86713326e-01 7.69040704e-01
-9.17248011e-01 1.90043271e-01 1.83660865e-01 2.71692038e-01
1.34469879e+00 -3.52634966e-01 -3.62458639e-02 -6.77331567e-01
-2.11654201e-01 -4.08148319e-01 -1.05234945e+00 -2.90575296e-01
-5.34036040e-01 -2.91184485e-01 3.17599952e-01 -5.50655872e-02
-3.89478087e-01 -1.46439195e-01 -4.15092915e-01 -9.05625820e-01
-1.82836995e-01 -4.09082174e-02 1.58718631e-01 -1.77514225e-01
7.99128830e-01 1.14060491e-02 -2.85366476e-01 -7.44415224e-01
1.67504326e-01 4.44337577e-01 5.92941701e-01 -7.95826554e-01
7.53629088e-01 8.67223263e-01 1.99321985e-01 -7.98051119e-01
-3.73251349e-01 -2.62625039e-01 -3.24105442e-01 -1.13415435e-01
9.32300925e-01 -7.70248771e-01 -9.74746943e-01 3.56770635e-01
-9.56266284e-01 -4.46200252e-01 -3.85323107e-01 1.14045806e-01
-5.43349028e-01 3.56233984e-01 -4.28095222e-01 -9.91167188e-01
-4.40976083e-01 -1.45195019e+00 1.42449808e+00 7.21498877e-02
-1.03102021e-01 -3.64899665e-01 -5.44347823e-01 3.46932054e-01
2.40770787e-01 6.98084116e-01 6.78345680e-01 -2.68317282e-01
-1.00206447e+00 -2.10635245e-01 -7.93998837e-01 4.44710106e-01
-1.90119922e-01 1.68116570e-01 -9.58456516e-01 -2.38189831e-01
-2.12058127e-01 4.50019866e-01 6.46842718e-01 3.50285858e-01
1.54884863e+00 1.25218540e-01 -6.71595693e-01 7.90334940e-01
1.39126611e+00 3.67767900e-01 7.51499414e-01 2.10936487e-01
7.94543028e-01 6.31109700e-02 5.35450697e-01 5.48576593e-01
4.43668097e-01 7.21377432e-01 1.04127455e+00 -5.77474982e-02
-2.83181190e-01 1.30392909e-01 7.37845898e-02 3.66784513e-01
-5.19578636e-01 -1.97214261e-01 -1.42606556e+00 3.42491686e-01
-1.50922751e+00 -5.15501022e-01 -8.72667909e-01 2.47708392e+00
4.96261090e-01 4.75224257e-01 -2.28018034e-04 -2.48953514e-02
7.88075030e-01 -5.88945150e-02 -4.20829594e-01 -2.45247692e-01
7.97309875e-02 8.18706512e-01 9.32996273e-01 3.43631178e-01
-9.40583289e-01 9.81915176e-01 3.07365274e+00 1.27557421e+00
-1.08991838e+00 6.27070218e-02 6.48082376e-01 -2.09942609e-01
-2.36604795e-01 5.28667122e-02 -9.73174334e-01 8.06545734e-01
7.11533189e-01 1.37751803e-01 1.35491610e-01 8.65323126e-01
-1.67602062e-01 -5.49379170e-01 -8.88527751e-01 1.26718092e+00
-1.94344819e-01 -1.45482111e+00 -2.68315166e-01 2.88940907e-01
6.04479790e-01 6.18911982e-01 4.61431080e-03 2.46366024e-01
4.12645251e-01 -1.14997625e+00 1.11278629e+00 5.11574149e-01
1.53933358e+00 -5.95998228e-01 3.63202572e-01 6.06002331e-01
-1.51865685e+00 4.64309961e-01 -3.44313711e-01 1.18529387e-01
1.96441442e-01 7.74254084e-01 -5.34563661e-01 6.68769598e-01
1.26604939e+00 2.24850208e-01 -3.39689851e-01 9.10035491e-01
1.19400144e-01 5.15375793e-01 -1.15315950e+00 1.01763934e-01
2.32364714e-01 -1.85929209e-01 6.87193513e-01 1.26156962e+00
5.67650199e-01 3.62035573e-01 5.56633353e-01 9.64179754e-01
-2.14222908e-01 4.60681878e-02 -2.43076026e-01 5.80177307e-01
5.59226155e-01 8.72832775e-01 -1.19302559e+00 -2.73438305e-01
-7.38786310e-02 5.10848105e-01 -3.08223832e-02 1.29866794e-01
-1.01086986e+00 -4.50120330e-01 6.35916114e-01 1.02417707e+00
3.78219068e-01 -3.85085881e-01 -6.48023963e-01 -8.36975574e-01
2.29501069e-01 -2.45664150e-01 2.50181079e-01 -9.02775049e-01
-9.92839873e-01 8.47961903e-01 3.42665374e-01 -1.24623442e+00
1.47432923e-01 -5.81881881e-01 -2.82652229e-01 1.15851629e+00
-1.17893910e+00 -1.09050369e+00 -6.92708075e-01 4.33793604e-01
3.97785246e-01 2.30181441e-01 7.23491192e-01 3.61997843e-01
-2.15452105e-01 5.52507639e-01 1.50655419e-01 -6.50274456e-02
2.72855431e-01 -9.21077907e-01 8.02523851e-01 4.31202471e-01
-5.15869819e-02 4.83461380e-01 4.70263153e-01 -8.49347830e-01
-1.24418688e+00 -1.18106651e+00 1.58585131e-01 -6.31326795e-01
3.09471190e-01 -4.26003397e-01 -1.09515643e+00 7.44969964e-01
-3.47197384e-01 4.25049424e-01 -3.89563572e-03 -2.71187663e-01
-3.25112522e-01 -1.36643171e-01 -1.58416414e+00 4.51269418e-01
1.12833619e+00 -5.01081571e-02 -1.56033337e-01 3.22834790e-01
5.71200252e-01 -8.87506485e-01 -1.06747782e+00 8.37388396e-01
3.48741502e-01 -1.24343073e+00 1.09489274e+00 3.33286464e-01
3.76427267e-03 -5.21870315e-01 -4.27119374e-01 -5.38872600e-01
-1.62136063e-01 -5.60991526e-01 1.71015471e-01 1.01770425e+00
3.37837428e-01 -8.27440202e-01 9.55843091e-01 4.84283447e-01
-7.17358708e-01 -8.64828587e-01 -1.24809468e+00 -6.03799462e-01
3.70673120e-01 -1.03180850e+00 6.44540966e-01 7.65706658e-01
-8.74460578e-01 -3.80940020e-01 3.95058602e-01 2.75318384e-01
6.54919922e-01 -8.86535272e-02 8.10971200e-01 -1.41949272e+00
-5.98453730e-02 -5.21827221e-01 -3.00857097e-01 -1.15860534e+00
-5.71933031e-01 -9.12123322e-01 -3.85588109e-01 -1.62529695e+00
-1.67874843e-01 -1.34467769e+00 1.00120716e-03 6.50167406e-01
3.08076203e-01 6.65990472e-01 3.39254677e-01 4.87711370e-01
-1.72512025e-01 8.20506439e-02 9.35641646e-01 1.23539515e-01
-5.34371361e-02 3.95261683e-04 -3.38124156e-01 1.11659670e+00
7.28590369e-01 -5.58244109e-01 -5.05273044e-02 -6.89880073e-01
4.21033621e-01 6.11065254e-02 7.60461688e-01 -1.26249576e+00
7.83250034e-02 -1.00373767e-01 3.14010054e-01 -1.11760259e+00
8.99731457e-01 -6.32027388e-01 4.54582185e-01 1.68940261e-01
7.52589405e-01 2.01483041e-01 6.49804950e-01 2.94499844e-01
2.17068851e-01 -1.51815087e-01 9.65373397e-01 -3.88053507e-01
-4.51657444e-01 4.33980703e-01 4.60871086e-02 8.35024565e-02
1.25016594e+00 -7.97385812e-01 -3.54996771e-01 -8.01955983e-02
-7.08663523e-01 2.95483261e-01 8.20140302e-01 6.81072623e-02
4.19615269e-01 -9.02539909e-01 -6.11182749e-01 2.30572730e-01
-1.07042417e-01 9.65366900e-01 4.84528869e-01 6.02130175e-01
-9.66501713e-01 -5.08574620e-02 1.71891525e-01 -1.13075805e+00
-9.35772777e-01 1.33444995e-01 2.41595268e-01 1.77450031e-01
-6.26767874e-01 1.24230683e+00 3.89865845e-01 -2.64944941e-01
-1.09826308e-02 -6.61380231e-01 5.82507372e-01 -9.08074901e-02
2.69844502e-01 8.63819838e-01 4.08300728e-01 -7.31850982e-01
-4.56898332e-01 9.27447915e-01 3.69831324e-02 -4.08136845e-02
1.15075803e+00 3.52515221e-01 -2.54163951e-01 1.27668530e-01
1.00635028e+00 1.20808758e-01 -1.34236193e+00 7.45006800e-02
-3.39101315e-01 -4.00430113e-01 -1.38456345e-01 -7.67645180e-01
-1.02535880e+00 9.58185196e-01 6.93655610e-01 1.64883673e-01
8.74382615e-01 3.54398876e-01 8.16445112e-01 -1.02766171e-01
1.17040253e+00 -8.10641050e-01 -3.57006073e-01 6.00101590e-01
9.40396547e-01 -7.50569761e-01 2.76016235e-01 -9.30968642e-01
-1.37795538e-01 8.69462848e-01 5.55862904e-01 -5.29284477e-01
8.66533399e-01 5.15529275e-01 -3.19971502e-01 -5.33084989e-01
-8.76992010e-03 7.69172609e-02 2.35818014e-01 1.80535495e-01
1.06391266e-01 2.38108888e-01 -1.24890767e-01 5.45937181e-01
-6.45404577e-01 -3.28219116e-01 2.43713915e-01 1.08365643e+00
-5.70182323e-01 -9.27872539e-01 -7.88320422e-01 7.20371008e-01
-4.06298727e-01 -1.42159805e-01 1.18353993e-01 8.83137226e-01
5.24409771e-01 6.72623336e-01 4.82071996e-01 -2.34196886e-01
6.17059112e-01 -1.54259652e-01 3.20501000e-01 -8.27489853e-01
-8.09595942e-01 4.33700264e-01 -1.03777103e-01 -7.62418032e-01
1.80138528e-01 -6.95876300e-01 -1.91064620e+00 -4.04590994e-01
-1.74575627e-01 -1.61463082e-01 8.94105613e-01 3.30953032e-01
5.23109734e-01 3.73038232e-01 1.22544900e-01 -1.17380226e+00
-2.00861901e-01 -7.24630237e-01 -5.33924282e-01 7.60819912e-02
-2.82704085e-01 -8.04510236e-01 -2.80988872e-01 -2.25209042e-01] | [7.980206489562988, -3.2728707790374756] |
31b31d19-db07-4b32-8893-10600c653284 | distributed-stochastic-bandit-learning-with | 2207.14391 | null | https://arxiv.org/abs/2207.14391v2 | https://arxiv.org/pdf/2207.14391v2.pdf | Distributed Stochastic Bandit Learning with Delayed Context Observation | We consider the problem where M agents collaboratively interact with an instance of a stochastic K-armed contextual bandit, where K>>M. The goal of the agents is to simultaneously minimize the cumulative regret over all the agents over a time horizon T. We consider a setting where the exact context is observed after a delay and at the time of choosing the action the agents are unaware of the context and only a distribution on the set of contexts is available. Such a situation arises in different applications where at the time of the decision the context needs to be predicted (e.g., weather forecasting or stock market prediction), and the context can be estimated once the reward is obtained. We propose an Upper Confidence Bound (UCB)-based distributed algorithm and prove the regret and communications bounds for linearly parametrized reward functions. We validated the performance of our algorithm via numerical simulations on synthetic data and real-world Movielens data. | ['Shana Moothedath', 'Jiabin Lin'] | 2022-07-28 | null | null | null | null | ['weather-forecasting', 'stock-market-prediction'] | ['miscellaneous', 'time-series'] | [-1.34753156e-02 -3.93472202e-02 -2.18092009e-01 -2.89667726e-01
-5.92505395e-01 -9.77643311e-01 4.64722574e-01 4.87827450e-01
-9.44424927e-01 9.09205675e-01 5.44710457e-02 -3.53509516e-01
-5.07276237e-01 -6.03935778e-01 -7.71057427e-01 -9.37691450e-01
-3.77658337e-01 7.89740324e-01 -1.05564892e-01 2.61720121e-01
2.36050338e-01 1.53881878e-01 -8.31386268e-01 -4.15299870e-02
5.58619499e-01 1.36372459e+00 5.01454175e-01 9.99944806e-01
1.39169171e-01 7.11679101e-01 -6.05303466e-01 -2.62716621e-01
6.82158768e-01 -3.55575979e-01 -5.37696719e-01 4.58706439e-01
-6.09026551e-01 -5.42426348e-01 -3.49390619e-02 1.16654563e+00
2.50408947e-01 5.17580092e-01 3.75754774e-01 -9.81622458e-01
-3.74432914e-02 7.52214193e-01 -5.19737303e-01 2.59828925e-01
1.50419220e-01 -9.17774290e-02 1.03490865e+00 -1.49733350e-01
4.14530993e-01 8.05424094e-01 -5.95293753e-02 5.24020731e-01
-1.11971116e+00 -1.43770084e-01 8.51202905e-01 2.31622934e-01
-7.56038785e-01 -1.03509732e-01 4.03374642e-01 -3.42634410e-01
2.70730346e-01 2.99923152e-01 6.32977307e-01 5.13965726e-01
1.25365052e-02 1.02146876e+00 1.16295445e+00 -4.91517037e-01
1.21576738e+00 1.54618114e-01 5.37111610e-02 1.17637917e-01
1.73103049e-01 2.38253623e-01 -6.07896686e-01 -5.87703645e-01
4.80271608e-01 4.20699388e-01 -1.80605665e-01 -2.28768140e-01
-1.08780766e+00 7.93714225e-01 9.34852511e-02 -1.98755786e-01
-1.05798352e+00 4.63866070e-02 5.74984774e-02 6.94746315e-01
5.60338318e-01 1.54911965e-01 -3.02029014e-01 -2.02146783e-01
-5.87431669e-01 3.71063799e-01 9.55784976e-01 8.57731640e-01
2.01763898e-01 -3.14130396e-01 -3.87679398e-01 3.74770314e-01
-3.45396250e-02 5.32407880e-01 -1.90935582e-02 -1.14355183e+00
8.90333056e-01 -9.24699679e-02 1.47328663e+00 -3.30743670e-01
-1.75043315e-01 -3.86022091e-01 -5.00253320e-01 2.61624604e-01
6.73769295e-01 -8.37303877e-01 -4.64648783e-01 1.78886127e+00
5.90427220e-01 4.08645481e-01 2.31182367e-01 1.34857631e+00
-3.24973792e-01 7.00286031e-01 -1.63869560e-01 -8.45807493e-01
8.41189742e-01 -6.79427207e-01 -6.08631611e-01 -3.44788253e-01
2.22381458e-01 -6.02759898e-01 2.88393766e-01 6.49361968e-01
-1.11026633e+00 2.50978172e-01 -7.55685091e-01 8.05180132e-01
8.83558467e-02 -9.66180786e-02 4.75816756e-01 4.80814487e-01
-4.25259799e-01 4.08592105e-01 -7.06032574e-01 -5.73528819e-02
7.67890364e-02 2.33103231e-01 1.11787565e-01 -8.65104496e-02
-7.98497856e-01 4.39126432e-01 2.52574742e-01 3.78287494e-01
-1.23911119e+00 -6.54555410e-02 1.22291088e-01 1.16785511e-01
1.12384880e+00 -5.40616691e-01 1.78313196e+00 -1.34527028e+00
-1.56436920e+00 1.88209429e-01 6.57617897e-02 -6.30265117e-01
7.81832099e-01 -1.48899063e-01 7.91992545e-02 5.56854159e-02
4.09300365e-02 -3.69223267e-01 7.28237152e-01 -9.40271914e-01
-1.31790209e+00 -3.82906139e-01 7.31700003e-01 5.81727386e-01
6.37402683e-02 -2.79680807e-02 -6.86352327e-02 -4.91354972e-01
-9.74372104e-02 -1.39273536e+00 -6.92312717e-01 -3.91644508e-01
-1.60701081e-01 8.62362757e-02 2.65744150e-01 -3.29549223e-01
6.60506248e-01 -1.80444527e+00 1.06826276e-01 3.18031847e-01
-1.83356434e-01 -1.61601737e-01 7.96078071e-02 7.68942535e-01
5.25420845e-01 -2.50913445e-02 8.93731564e-02 -5.16725779e-01
1.32759899e-01 3.23526353e-01 -4.63975936e-01 5.97919106e-01
-8.06474209e-01 3.02036405e-01 -7.47441947e-01 1.62319914e-01
-1.09053276e-01 -3.96387607e-01 -4.43441719e-01 5.88637888e-01
-8.61718476e-01 4.81210619e-01 -1.09257770e+00 4.05597612e-02
3.86111349e-01 -3.63825709e-01 4.73803014e-01 7.28473783e-01
-5.37972525e-02 -9.01705027e-03 -1.58439624e+00 1.33356237e+00
-7.18560696e-01 1.59413755e-01 5.37610114e-01 -9.73374248e-01
5.82843572e-02 4.96686786e-01 3.23989749e-01 -3.49964529e-01
2.80832350e-01 -8.84741843e-02 -1.28229171e-01 -2.36752868e-01
4.51879591e-01 -2.29374945e-01 -1.47149801e-01 9.35652971e-01
-5.13185203e-01 5.20863712e-01 -1.71977729e-01 2.64445812e-01
9.26567495e-01 -4.20919180e-01 3.27401996e-01 6.07203133e-02
5.37117459e-02 -6.64265677e-02 7.09488332e-01 1.40967309e+00
-2.90505141e-02 -1.65018663e-02 6.06042206e-01 -3.86428028e-01
-8.08452308e-01 -5.33268154e-01 5.33751667e-01 1.21545327e+00
3.99533391e-01 2.62566298e-01 -5.04600823e-01 -5.89827180e-01
1.28278822e-01 8.88015509e-01 -9.94097829e-01 5.09257376e-01
4.15581763e-02 -5.89990199e-01 -8.39751244e-01 1.83494061e-01
3.99759680e-01 -6.92225397e-01 -1.00125194e+00 5.97034514e-01
-8.10819566e-02 -1.15195298e+00 -8.31243932e-01 4.71854024e-02
-6.26643300e-01 -9.03251708e-01 -5.12962461e-01 -7.91385025e-02
6.88861907e-01 3.14388126e-01 6.53749883e-01 -3.55084777e-01
3.84910464e-01 7.70244360e-01 -2.36023843e-01 -5.78431249e-01
-4.27777320e-02 -2.56617039e-01 5.47609515e-02 5.95443547e-01
-2.75932044e-01 -1.93187997e-01 -1.00710642e+00 3.42444241e-01
-9.43671644e-01 1.08260803e-01 2.22768992e-01 5.70521057e-01
6.05490625e-01 3.02038908e-01 6.69854701e-01 -1.03979886e+00
7.88821995e-01 -7.04640865e-01 -1.32549512e+00 5.68930089e-01
-2.79240280e-01 -1.64487168e-01 6.85518026e-01 -6.93683445e-01
-1.29858220e+00 9.32291597e-02 7.90567219e-01 -1.53477982e-01
9.54730660e-02 6.54494226e-01 7.21770823e-02 2.73934066e-01
1.66396201e-01 1.84211940e-01 -2.75423288e-01 -3.81470948e-01
1.88624382e-01 7.08561480e-01 1.40643090e-01 -9.16581392e-01
2.18972176e-01 3.31433237e-01 7.56365061e-02 -4.34566349e-01
-8.23928893e-01 -3.56369466e-01 8.32991302e-02 -4.44180965e-01
2.89519787e-01 -9.18666482e-01 -1.39147913e+00 2.76994407e-01
-1.01311874e+00 -4.97892499e-01 -2.49813706e-01 6.91574931e-01
-8.46983790e-01 1.07679799e-01 -2.30026364e-01 -1.67945147e+00
2.56739464e-02 -8.44581068e-01 3.31167787e-01 5.35393059e-01
4.22741979e-01 -8.22157860e-01 -1.20863236e-01 3.47901165e-01
3.87451261e-01 1.10719621e-01 5.37404120e-01 -8.06965411e-01
-9.75475013e-01 -3.16378027e-01 3.43535602e-01 -8.63834023e-02
3.10544465e-02 -5.89158297e-01 -5.49580395e-01 -4.17923093e-01
5.75254671e-02 -5.78615405e-02 1.75353885e-01 5.67600608e-01
9.35258031e-01 -1.16865718e+00 -3.36306512e-01 -2.15449333e-01
1.23880363e+00 7.03019083e-01 -2.99701124e-01 3.73252749e-01
-1.53424367e-01 4.09493983e-01 1.04486418e+00 1.22684073e+00
3.98404598e-01 7.89287746e-01 7.42646873e-01 7.17360914e-01
9.58220065e-01 -3.73078287e-02 2.04424739e-01 -1.03589110e-01
3.91521715e-02 -6.40952468e-01 -4.53838944e-01 8.10860634e-01
-2.44841647e+00 -8.33685517e-01 6.87147975e-01 2.89793706e+00
7.42557645e-01 5.68305515e-02 4.54638213e-01 -2.75549263e-01
8.38950038e-01 -1.20465286e-01 -1.27406168e+00 -3.42027575e-01
1.19620614e-01 -5.48475623e-01 8.10863972e-01 8.14053953e-01
-8.18392158e-01 5.07245958e-01 5.16242838e+00 5.78207612e-01
-8.59984517e-01 2.26116925e-01 9.95065689e-01 -9.45120335e-01
7.40158334e-02 9.75699648e-02 -5.38277507e-01 6.91876769e-01
1.02102423e+00 -4.90989447e-01 1.06541038e+00 7.34885156e-01
4.23362195e-01 -7.94948220e-01 -1.27352858e+00 7.31203258e-01
-6.55719757e-01 -1.26673496e+00 -7.27119505e-01 1.96211725e-01
9.07013893e-01 1.09050855e-01 -4.08429466e-02 -1.68871790e-01
9.62077498e-01 -4.36011910e-01 8.86438906e-01 7.58735299e-01
1.19959414e-01 -1.18898451e+00 6.37971997e-01 1.03775620e+00
-7.61337519e-01 -6.67019725e-01 -1.01300918e-01 -3.44362706e-01
2.01774254e-01 5.35419106e-01 -8.60710263e-01 3.95808548e-01
2.74212897e-01 2.08611842e-02 5.07917225e-01 1.32382309e+00
-8.23777989e-02 7.33101189e-01 -5.97542942e-01 -5.68437994e-01
4.60472554e-01 -4.39125419e-01 6.16889000e-01 6.74924433e-01
2.07168981e-01 8.54143798e-01 6.55427992e-01 3.04169804e-01
-3.64758410e-02 1.57819867e-01 -8.60538781e-02 7.09669963e-02
6.58311725e-01 8.96308184e-01 -7.77994156e-01 -3.71864349e-01
-2.77707160e-01 9.25132394e-01 1.87651262e-01 6.58171415e-01
-6.26374483e-01 1.86187863e-01 7.45625496e-01 -4.38463569e-01
5.46862304e-01 -1.11833483e-01 -1.53052673e-01 -8.84429693e-01
7.76928291e-02 -4.10768390e-01 7.65175283e-01 -5.50725043e-01
-1.33874726e+00 5.76975569e-02 -1.12643406e-01 -9.98643517e-01
-3.75172824e-01 3.96624804e-02 -4.80790913e-01 8.30227971e-01
-1.39555573e+00 -3.82128388e-01 4.03081030e-01 4.88270104e-01
4.07118946e-01 -1.22431740e-01 5.65946579e-01 -4.03128415e-01
-5.16109705e-01 -5.04261740e-02 8.63208830e-01 -1.42904416e-01
1.20410815e-01 -1.28497171e+00 -1.66057348e-01 7.16956377e-01
-4.43115495e-02 1.96025237e-01 1.02455878e+00 -4.25492227e-01
-1.53739738e+00 -9.89037752e-01 3.04554284e-01 1.11218505e-01
6.92510605e-01 -2.19235495e-01 -3.00546467e-01 7.08454132e-01
9.16179344e-02 2.66271591e-01 6.51902378e-01 -5.93397617e-02
1.64639294e-01 -3.18903029e-01 -1.41992080e+00 4.96640325e-01
6.26470923e-01 -2.17056394e-01 -8.07566270e-02 6.15006924e-01
5.11209786e-01 -5.09494185e-01 -5.56401253e-01 -2.62735665e-01
5.29424131e-01 -6.59302354e-01 2.44995445e-01 -9.11902189e-01
-2.08960593e-01 -2.11283430e-01 -3.88929188e-01 -1.65670335e+00
-1.02368347e-01 -1.13018990e+00 -1.49618044e-01 8.64368081e-01
4.98590946e-01 -8.84121358e-01 9.54008579e-01 1.40954101e+00
6.34401262e-01 -4.47786897e-01 -1.45641768e+00 -6.33893371e-01
-2.69616753e-01 -3.02731067e-01 7.44707763e-01 4.86692637e-01
1.27148569e-01 3.46212415e-03 -6.62351966e-01 7.46566236e-01
7.37258375e-01 8.58320296e-01 6.28931999e-01 -6.49666607e-01
-7.86556005e-01 -7.41286650e-02 4.31060880e-01 -1.32143283e+00
5.61419204e-02 -7.66453296e-02 1.10398695e-01 -1.23834097e+00
3.93062621e-01 -7.99620152e-01 -4.67646718e-01 2.55323470e-01
-1.06481269e-01 -6.55697703e-01 5.08806825e-01 1.13600031e-01
-1.14261973e+00 4.02780771e-01 9.87053454e-01 -2.49023493e-02
-2.74599284e-01 1.05040920e+00 -5.87868690e-01 3.53164822e-01
8.29703331e-01 -5.60424387e-01 -6.59377396e-01 -2.82110125e-01
4.00483310e-01 1.29336083e+00 9.22939777e-02 -3.58694077e-01
5.45850396e-01 -1.04408288e+00 -7.85237849e-02 -5.77269614e-01
4.04953599e-01 -1.31141913e+00 3.55550885e-01 3.27373892e-01
-8.99033308e-01 -1.72086820e-01 -2.54026502e-01 1.32029784e+00
3.54690760e-01 -3.82931471e-01 3.85782301e-01 -1.14445500e-01
-2.06111521e-01 5.67128837e-01 -5.34419119e-01 -6.65572956e-02
1.30649149e+00 3.01934391e-01 -1.77809775e-01 -1.20372212e+00
-1.08558869e+00 6.95225298e-01 1.76319078e-01 2.60622632e-02
3.72477293e-01 -8.63639057e-01 -5.39116800e-01 -4.35879827e-01
-1.80421934e-01 -1.17534295e-01 4.29351568e-01 4.97371137e-01
3.73204798e-01 4.48315263e-01 3.39295089e-01 -1.76514432e-01
-1.24249017e+00 7.74552584e-01 4.58122700e-01 -1.44288927e-01
-1.20546572e-01 5.04797697e-01 2.66556680e-01 3.79281461e-01
5.04305720e-01 -3.57830077e-01 -4.49506454e-02 -4.73405048e-02
7.74381220e-01 4.08893079e-01 -2.29810122e-02 -2.81857830e-02
6.36348352e-02 -1.37151793e-01 -2.84373730e-01 -7.33860552e-01
1.55443656e+00 -4.97938246e-01 2.01046556e-01 4.67324138e-01
4.09518570e-01 -2.64602737e-03 -1.70493925e+00 -9.75794256e-01
-6.66200044e-03 -8.24602187e-01 1.67844325e-01 -1.12481225e+00
-1.18259430e+00 1.73073098e-01 3.90199542e-01 8.07061195e-01
1.09422481e+00 -1.19186454e-01 3.99068207e-01 4.80464995e-01
8.33616555e-01 -1.27615666e+00 -2.09781423e-01 2.95010209e-01
6.12853467e-01 -9.54437077e-01 -7.13215470e-02 2.04032853e-01
-1.02750862e+00 7.00148463e-01 2.45544352e-02 -7.88855702e-02
7.85123467e-01 1.21632755e-01 -2.41570130e-01 2.97074825e-01
-1.36536455e+00 -1.46286905e-01 -1.92638665e-01 1.94242522e-01
-3.65607858e-01 7.57353604e-01 -2.82764912e-01 7.81248569e-01
-1.41582461e-02 7.95805827e-02 8.85319471e-01 1.04318976e+00
-5.27073503e-01 -8.76728654e-01 -5.54796159e-01 4.56073463e-01
-1.02393138e+00 3.02944392e-01 -1.71656266e-01 1.44320820e-02
-4.02039915e-01 1.40495288e+00 -1.17418744e-01 2.46302664e-01
2.16621727e-01 -2.37569094e-01 3.32286447e-01 -4.39466357e-01
-4.20714438e-01 2.31295303e-01 2.79165804e-01 -3.80643010e-01
-4.85191315e-01 -7.59045541e-01 -9.41131115e-01 -2.87140787e-01
-3.07612956e-01 7.30257571e-01 8.83073032e-01 1.10551786e+00
3.90311062e-01 -5.56184538e-02 1.39355612e+00 -6.16510272e-01
-7.09196568e-01 -6.77456141e-01 -7.94385493e-01 2.33556386e-02
7.76415884e-01 -4.09259617e-01 -3.33355695e-01 -2.83803999e-01] | [4.4788689613342285, 3.20560359954834] |
b83c954e-7130-4bc2-9992-5ab45233daf1 | logdet-rank-minimization-with-application-to | 1507.00908 | null | http://arxiv.org/abs/1507.00908v1 | http://arxiv.org/pdf/1507.00908v1.pdf | LogDet Rank Minimization with Application to Subspace Clustering | Low-rank matrix is desired in many machine learning and computer vision
problems. Most of the recent studies use the nuclear norm as a convex surrogate
of the rank operator. However, all singular values are simply added together by
the nuclear norm, and thus the rank may not be well approximated in practical
problems. In this paper, we propose to use a log-determinant (LogDet) function
as a smooth and closer, though non-convex, approximation to rank for obtaining
a low-rank representation in subspace clustering. Augmented Lagrange
multipliers strategy is applied to iteratively optimize the LogDet-based
non-convex objective function on potentially large-scale data. By making use of
the angular information of principal directions of the resultant low-rank
representation, an affinity graph matrix is constructed for spectral
clustering. Experimental results on motion segmentation and face clustering
data demonstrate that the proposed method often outperforms state-of-the-art
subspace clustering algorithms. | ['Qiang Chen', 'Zhao Kang', 'Chong Peng', 'Jie Cheng'] | 2015-07-03 | null | null | null | null | ['face-clustering'] | ['computer-vision'] | [-1.03541970e-01 -1.86670601e-01 -1.69242412e-01 -3.06792885e-01
-6.94329202e-01 -6.22802556e-01 1.79064810e-01 -3.10646832e-01
-3.28740090e-01 4.88724649e-01 1.55686110e-01 -3.94002981e-02
-3.80257219e-01 -8.50359872e-02 -4.26143706e-01 -1.01673269e+00
-5.55360988e-02 3.12242121e-01 4.95805293e-02 2.83868790e-01
1.89514279e-01 5.23885190e-01 -1.23574257e+00 4.35169004e-02
1.25141811e+00 6.74052238e-01 1.39656037e-01 1.73586398e-01
2.37512335e-01 1.63912937e-01 -3.80212558e-03 -4.54012603e-02
4.02807087e-01 -3.58840048e-01 -5.20557046e-01 5.09945571e-01
5.00010788e-01 -3.94775234e-02 -3.06976497e-01 1.48944342e+00
3.71735036e-01 4.94858533e-01 7.07879961e-01 -1.01770222e+00
-4.05665636e-01 4.27153260e-01 -1.09050274e+00 -1.32245585e-01
4.42640394e-01 -4.61633474e-01 8.75244319e-01 -1.43196785e+00
6.06137931e-01 1.37350106e+00 5.01675606e-01 8.70929584e-02
-1.42415404e+00 -3.91397089e-01 7.14345369e-03 4.37998593e-01
-1.81926119e+00 -4.22702491e-01 1.08885300e+00 -5.63636243e-01
3.35229248e-01 4.45718706e-01 4.63250726e-01 5.03028810e-01
-3.34640086e-01 6.52925551e-01 1.23418260e+00 -2.31249347e-01
1.39541268e-01 -1.20030688e-02 2.04235986e-01 8.19995642e-01
4.70835805e-01 -1.33121178e-01 -3.35769892e-01 -6.94627643e-01
7.87629604e-01 -4.02325615e-02 -4.86155629e-01 -8.62527966e-01
-1.34229207e+00 7.52222359e-01 2.95064330e-01 3.08590919e-01
-4.42884237e-01 -1.51250720e-01 -1.59569979e-02 -2.38024116e-01
2.39670992e-01 3.85946594e-02 9.31839198e-02 5.76241165e-02
-1.08424437e+00 -1.14080861e-01 6.56421661e-01 8.04969251e-01
8.73178601e-01 2.59945095e-01 9.91170704e-02 1.09409428e+00
6.14680290e-01 5.79588771e-01 2.04540431e-01 -1.21345580e+00
1.66960374e-01 6.17663860e-01 1.08184151e-01 -1.31841552e+00
-3.83088410e-01 -3.57768953e-01 -1.14090872e+00 -3.82368360e-03
6.36080682e-01 1.06198087e-01 -5.57986438e-01 1.44949436e+00
6.44449115e-01 5.70755363e-01 -2.91270427e-02 1.44548273e+00
4.35933262e-01 6.05408370e-01 -3.55745047e-01 -7.81331003e-01
1.02409816e+00 -6.87795341e-01 -6.98964715e-01 -3.30398395e-03
2.36044079e-01 -9.28202689e-01 8.97229493e-01 4.46081221e-01
-8.62717807e-01 -3.06100845e-01 -9.18100476e-01 1.61125496e-01
2.18222395e-01 5.38907588e-01 5.16682386e-01 5.37241697e-01
-8.37114155e-01 5.59024930e-01 -8.73194158e-01 -3.66277009e-01
-9.56643373e-02 3.88756096e-01 -4.79367644e-01 -2.22264588e-01
-6.96700573e-01 3.19113523e-01 4.13331896e-01 5.49258113e-01
-5.67098320e-01 -3.14785361e-01 -5.54619193e-01 -2.16978326e-01
3.67515266e-01 -3.79199296e-01 4.90595102e-01 -5.63419461e-01
-1.37306333e+00 7.06050754e-01 -3.86756122e-01 2.72789180e-01
3.36705089e-01 -1.92835748e-01 -2.89209247e-01 3.58437240e-01
1.30680889e-01 1.11639835e-01 1.11952949e+00 -1.50858736e+00
-2.68518656e-01 -5.79479098e-01 -3.96693289e-01 5.21228433e-01
-4.89086777e-01 -1.08304329e-01 -8.36885691e-01 -5.36467612e-01
8.83808196e-01 -1.17574501e+00 -4.39799458e-01 -2.38808379e-01
-3.78555268e-01 -1.48820952e-01 9.34076011e-01 -8.28651071e-01
1.48553693e+00 -2.47130394e+00 6.59926534e-01 7.56401241e-01
1.21975653e-01 1.93400327e-02 -4.30104695e-02 2.74673134e-01
-1.91628531e-01 -2.34047800e-01 -5.15661716e-01 1.77186921e-01
-1.75019026e-01 8.41898173e-02 1.44100450e-02 1.06752038e+00
-2.67693311e-01 1.72523305e-01 -1.05991757e+00 -6.56142294e-01
2.01009735e-01 5.45283735e-01 -4.74045843e-01 3.08404043e-02
3.58133852e-01 4.84703839e-01 -5.86441398e-01 5.46793163e-01
8.44298959e-01 -1.79410815e-01 4.32456940e-01 -7.28223383e-01
-6.41935840e-02 -6.23648047e-01 -1.86211991e+00 1.83218241e+00
1.92161232e-01 3.44741046e-01 6.60564840e-01 -1.14811039e+00
7.46736288e-01 1.54646054e-01 9.71018672e-01 8.21806192e-02
7.20032528e-02 2.82562852e-01 -1.92479268e-02 -2.84182876e-01
3.15704644e-01 1.08506858e-01 3.30465823e-01 1.50317788e-01
-3.65426183e-01 2.33153582e-01 4.20238942e-01 2.96912909e-01
7.28408873e-01 6.97232932e-02 1.85718745e-01 -5.92736661e-01
1.03719091e+00 -1.37696296e-01 9.09257054e-01 1.82027429e-01
-8.90185311e-02 6.39694214e-01 3.32835972e-01 4.26092744e-02
-8.98074508e-01 -1.02285588e+00 -4.49685872e-01 9.38161433e-01
2.25969389e-01 -2.64243037e-01 -9.41787660e-01 -3.90846074e-01
-1.11640871e-01 2.34086379e-01 -2.19676942e-02 2.69848071e-02
-6.72593057e-01 -8.86181295e-01 2.05165595e-01 1.98122635e-01
3.44545037e-01 -3.75371337e-01 -1.73618458e-02 1.23047017e-01
-2.40294501e-01 -1.05865681e+00 -9.26040888e-01 -2.14288056e-01
-1.17914033e+00 -1.19917226e+00 -8.03053081e-01 -9.29523885e-01
1.28008163e+00 6.19590640e-01 4.38889205e-01 -7.90378898e-02
-8.65574926e-02 5.21408200e-01 -3.49413991e-01 4.64758962e-01
1.91365704e-02 -3.99964690e-01 5.86784005e-01 6.05706871e-01
2.42782265e-01 -4.73532379e-01 -6.47556543e-01 5.93742430e-01
-8.68096590e-01 -2.35266820e-01 5.95415950e-01 1.02269340e+00
8.96499276e-01 2.86416501e-01 2.54002690e-01 -6.73542798e-01
6.79662526e-01 -1.89544633e-01 -7.69207656e-01 2.53513277e-01
-7.44122386e-01 8.09902176e-02 6.73855722e-01 -7.18018591e-01
-1.01928914e+00 6.08690321e-01 4.57673550e-01 -9.34261203e-01
2.16540158e-01 7.70676076e-01 -2.86641061e-01 -2.97941476e-01
5.43093026e-01 3.16392750e-01 -1.92648359e-02 -5.77376306e-01
6.99930489e-01 6.43657446e-01 6.30217016e-01 -7.24358499e-01
1.13585782e+00 6.97337449e-01 3.72936606e-01 -1.24063861e+00
-5.97908974e-01 -1.00929844e+00 -8.96739542e-01 -2.83764511e-01
7.30701447e-01 -8.95688593e-01 -5.80009460e-01 2.00005874e-01
-7.61629939e-01 3.19173008e-01 4.53098901e-02 9.70178306e-01
-5.12391567e-01 1.04373503e+00 -5.03130794e-01 -9.68099415e-01
-1.04064882e-01 -1.01486838e+00 7.58049965e-01 9.86661986e-02
1.86446652e-01 -8.09859097e-01 -3.69854420e-02 6.04919791e-01
-2.16884926e-01 4.14781779e-01 7.92337954e-01 -2.90473759e-01
-4.78008628e-01 -2.34910935e-01 -2.21074119e-01 3.03321660e-01
2.14372319e-03 9.64242063e-05 -5.78942478e-01 -6.80421710e-01
1.63226128e-01 5.13540320e-02 6.76291823e-01 4.75547701e-01
8.79581571e-01 -4.26765442e-01 -2.87657976e-01 7.37363636e-01
1.32818913e+00 1.22921735e-01 2.37691030e-01 -1.64687242e-02
1.21050715e+00 7.20077574e-01 9.62453723e-01 6.30473554e-01
8.52225348e-02 5.39307237e-01 1.59159809e-01 2.38680299e-02
2.31156215e-01 3.88131700e-02 4.74479139e-01 1.34888387e+00
-2.40937188e-01 5.10646343e-01 -9.68671978e-01 4.36495960e-01
-2.33815432e+00 -8.99906456e-01 -6.02784991e-01 2.40073514e+00
6.95974648e-01 -4.08537328e-01 8.08946341e-02 9.48476642e-02
1.04730844e+00 4.13855910e-02 -6.54389441e-01 1.51574194e-01
-2.13249117e-01 -5.32522202e-01 5.17059386e-01 5.48867345e-01
-1.11683464e+00 8.79115760e-01 6.07214832e+00 9.85899150e-01
-7.85737097e-01 -7.60058910e-02 2.32247457e-01 1.31166041e-01
-1.32367834e-01 2.04964012e-01 -3.49847257e-01 3.27418804e-01
4.61269736e-01 -2.80691266e-01 9.44986105e-01 1.00896609e+00
5.57485402e-01 8.76532868e-02 -9.98354197e-01 1.53923929e+00
6.31783083e-02 -7.10218906e-01 -1.20126836e-01 3.03945810e-01
1.04909992e+00 -3.15741539e-01 1.85444683e-01 -2.92885274e-01
5.15760994e-03 -7.89938927e-01 4.29326326e-01 5.42166710e-01
7.35233545e-01 -9.71977592e-01 4.84333724e-01 2.95162112e-01
-1.41647172e+00 -3.42880152e-02 -6.34769857e-01 4.08555061e-01
8.10187086e-02 9.44863737e-01 -5.49214840e-01 4.48920578e-01
5.94124436e-01 8.91176105e-01 -3.44851851e-01 1.07407916e+00
1.63699780e-02 5.53572655e-01 -5.27989149e-01 3.10834795e-01
1.77020639e-01 -1.26938593e+00 9.44553852e-01 1.04299998e+00
3.48980844e-01 5.73635340e-01 5.77243388e-01 6.18504882e-01
7.66749904e-02 4.33135241e-01 -5.04787624e-01 -2.34389931e-01
3.43965083e-01 1.61418772e+00 -7.71175623e-01 -1.50536805e-01
-5.86262167e-01 1.00639069e+00 6.19046912e-02 7.67916560e-01
-5.28019249e-01 -3.01878210e-02 5.95393300e-01 1.34781888e-02
1.18902989e-01 -6.69837058e-01 -1.61363617e-01 -1.29904449e+00
1.14904001e-01 -9.05888498e-01 3.89880955e-01 -4.66158658e-01
-1.32680118e+00 2.43067190e-01 6.93855062e-02 -1.45572150e+00
-1.50039792e-01 -5.42033195e-01 -4.41596955e-01 6.27893269e-01
-9.42687988e-01 -9.76512372e-01 -1.43577620e-01 8.59279811e-01
1.65288538e-01 -2.88533092e-01 3.41708541e-01 3.86757672e-01
-8.21349025e-01 2.86388546e-01 7.93073237e-01 9.44116861e-02
7.54707277e-01 -1.26906943e+00 -6.18355513e-01 1.04682195e+00
1.69512615e-01 8.67489517e-01 7.55016923e-01 -7.15895951e-01
-1.96254325e+00 -9.40727830e-01 2.14915678e-01 -5.78080863e-02
8.17294896e-01 -1.34164840e-02 -9.67666566e-01 2.63655037e-01
-8.65874439e-02 9.34461057e-02 8.71600509e-01 6.71267509e-02
-2.19816610e-01 -2.15441331e-01 -1.08201146e+00 6.00962460e-01
9.73563552e-01 -5.95182478e-01 -3.53004247e-01 4.53268439e-01
2.38603413e-01 -1.30217120e-01 -1.21685541e+00 4.76846367e-01
5.01459062e-01 -6.81913853e-01 1.06821704e+00 -2.61235386e-01
-2.95138270e-01 -9.89057064e-01 -3.52223426e-01 -1.11800623e+00
-7.97083080e-01 -7.37496674e-01 -2.62327284e-01 1.19394302e+00
3.43936980e-02 -3.13558787e-01 1.04374671e+00 5.08933544e-01
-1.49364546e-02 -5.70224285e-01 -1.05133641e+00 -9.28366482e-01
-3.55931997e-01 -3.73308398e-02 -1.88354906e-02 1.23958552e+00
1.96782112e-01 4.11907554e-01 -5.11920989e-01 4.55383599e-01
1.35354912e+00 1.64858356e-01 5.69977403e-01 -1.28862536e+00
-1.22989148e-01 -2.91206688e-01 -3.70123863e-01 -1.02001166e+00
2.78989255e-01 -1.05451345e+00 6.81884885e-02 -1.34904146e+00
4.98410374e-01 -9.69548374e-02 -3.61097544e-01 3.64964083e-02
-1.87281236e-01 1.51808023e-01 9.90776196e-02 6.80985928e-01
-5.41525126e-01 6.64449811e-01 1.17251921e+00 -1.07265785e-01
-4.38539416e-01 -2.80703872e-01 -4.23307687e-01 1.02439809e+00
4.66418862e-01 -2.63751805e-01 -4.92243588e-01 -2.01452728e-02
-6.25672713e-02 1.94807947e-01 -6.45980835e-02 -8.54228735e-01
4.69221205e-01 -3.79612327e-01 3.11071932e-01 -7.88465679e-01
4.26903188e-01 -1.19580805e+00 3.23888570e-01 1.90892354e-01
1.22490279e-01 -1.77623570e-01 -3.05541784e-01 8.02846372e-01
-3.68899137e-01 -1.60250515e-01 9.08603311e-01 1.96997046e-01
-6.21110976e-01 4.29834485e-01 -1.59713358e-01 -1.90349102e-01
9.80830371e-01 -5.00134170e-01 3.99563238e-02 -2.83685595e-01
-8.41526330e-01 1.43822134e-01 5.79706073e-01 1.66427121e-01
9.10173714e-01 -1.70030892e+00 -7.23640382e-01 8.16653669e-02
-5.87000065e-02 -7.74583742e-02 1.32038429e-01 1.10635448e+00
-5.29209614e-01 3.07892114e-01 6.51681870e-02 -8.43142450e-01
-1.47104025e+00 6.96963608e-01 -9.41020474e-02 2.59669800e-03
-2.73144245e-01 4.67774391e-01 2.45976359e-01 -2.60796756e-01
1.80852905e-01 2.41859108e-01 -2.95731634e-01 1.73694715e-01
2.13728368e-01 9.00879323e-01 -4.86434728e-01 -1.26357758e+00
-6.08194113e-01 1.13911927e+00 3.11083049e-01 -3.93174529e-01
1.21108353e+00 -3.90205353e-01 -5.82644761e-01 3.44271868e-01
1.36784041e+00 9.60799679e-02 -1.22874784e+00 -3.15949112e-01
2.15096653e-01 -6.16942465e-01 8.77411216e-02 -9.67960954e-02
-1.15870917e+00 5.15274525e-01 6.71207726e-01 -1.44061983e-01
1.23099339e+00 -2.07896963e-01 3.75950843e-01 5.80851436e-01
3.60778153e-01 -1.42321479e+00 7.05479011e-02 3.18721324e-01
9.20510292e-01 -1.13975251e+00 4.02826339e-01 -8.37865531e-01
-5.54658532e-01 1.05069661e+00 5.24955869e-01 -2.15063408e-01
7.73208380e-01 -2.80696362e-01 -5.93907759e-02 2.10289052e-03
-1.22750029e-01 -1.27535731e-01 7.78130293e-01 4.55814600e-01
4.45568711e-01 1.35453299e-01 -7.77824759e-01 2.64195323e-01
5.61474357e-03 -5.17133951e-01 3.74962270e-01 6.12794995e-01
-5.85188687e-01 -1.07124007e+00 -9.95064080e-01 3.56240302e-01
-3.41413200e-01 1.01004235e-01 -4.16809589e-01 2.38068059e-01
-9.97570530e-02 1.22214532e+00 -3.21883410e-01 -3.49221438e-01
1.68221846e-01 -1.26532346e-01 2.81984717e-01 -4.26060528e-01
2.31432989e-02 7.92455494e-01 -1.60928681e-01 -7.55604684e-01
-4.58974898e-01 -8.49084318e-01 -1.41552079e+00 -1.99167296e-01
-4.15524006e-01 4.57334310e-01 6.54163837e-01 4.93255407e-01
1.71183079e-01 -9.97351333e-02 8.50118935e-01 -7.52517343e-01
-6.27200246e-01 -7.10844457e-01 -9.87522602e-01 6.62432551e-01
2.16914844e-02 -5.29173017e-01 -5.23475289e-01 2.24493161e-01] | [7.714892864227295, 4.456671237945557] |
e1a80d99-50b6-44d9-b9d3-8d037aec9721 | solving-soft-clustering-ensemble-via-k-sparse | null | null | http://proceedings.neurips.cc/paper/2021/hash/07a4e20a7bbeeb7a736682b26b16ebe8-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/07a4e20a7bbeeb7a736682b26b16ebe8-Paper.pdf | Solving Soft Clustering Ensemble via $k$-Sparse Discrete Wasserstein Barycenter | Clustering ensemble is one of the most important problems in ensemble learning. Though it has been extensively studied in the past decades, the existing methods often suffer from the issues like high computational complexity and the difficulty on understanding the consensus. In this paper, we study the more general soft clustering ensemble problem where each individual solution is a soft clustering. We connect it to the well-known discrete Wasserstein barycenter problem in geometry. Based on some novel geometric insights in high dimensions, we propose the sampling-based algorithms with provable quality guarantees. We also provide the systematical analysis on the consensus of our model. Finally, we conduct the experiments to evaluate our proposed algorithms. | ['Hu Ding', 'Mengying Li', 'Ruizhe Qin'] | 2021-12-01 | null | https://openreview.net/forum?id=yAIYc7YjGbd | https://openreview.net/pdf?id=yAIYc7YjGbd | neurips-2021-12 | ['clustering-ensemble'] | ['graphs'] | [-9.81624201e-02 -1.70174956e-01 3.52930158e-01 -3.33580136e-01
-8.94524515e-01 -5.72050214e-01 2.12214857e-01 8.29051733e-02
-8.05089995e-02 6.93726957e-01 1.43516599e-03 -1.37603477e-01
-7.77186155e-01 -4.40092117e-01 -3.92601073e-01 -1.51206172e+00
-9.07934830e-02 5.61572373e-01 2.25089910e-03 -5.29011376e-02
4.38328028e-01 1.33646101e-01 -1.47008646e+00 -1.61966443e-01
1.56805074e+00 1.10681617e+00 -2.40280718e-01 5.75597644e-01
3.86002868e-01 1.69596225e-01 -4.22070771e-01 -3.14976513e-01
3.86241466e-01 -4.46945369e-01 -5.95762014e-01 2.70336181e-01
3.70502979e-01 2.28421390e-01 -2.16128901e-02 1.34595883e+00
8.44281375e-01 8.17517415e-02 9.31463420e-01 -1.68573773e+00
-6.31302357e-01 5.98478496e-01 -8.98930848e-01 -1.11097589e-01
-2.09921841e-02 -2.36491486e-01 1.07541418e+00 -9.67292786e-01
2.58226067e-01 1.20136976e+00 6.57666802e-01 2.72415191e-01
-1.02861190e+00 -5.47263741e-01 2.17139602e-01 5.04993558e-01
-1.54843771e+00 -3.64854753e-01 8.65559101e-01 -4.41745579e-01
-2.62853950e-02 3.17694992e-01 2.71973252e-01 6.23072028e-01
3.99504341e-02 7.36053109e-01 1.52392864e+00 -4.49519604e-01
6.77573025e-01 -1.63350761e-01 2.56769240e-01 6.46814466e-01
5.17060101e-01 -1.08246423e-01 -1.24023281e-01 -3.55724782e-01
4.68723923e-01 1.29798725e-01 -3.79468262e-01 -6.35993361e-01
-8.89972627e-01 7.41628647e-01 2.41418406e-01 2.12882087e-01
3.44830230e-02 2.12846398e-02 5.63620664e-02 8.82409811e-02
7.10963905e-01 3.97884995e-02 -2.57848322e-01 1.02027588e-01
-8.39341521e-01 8.13167170e-03 7.78735459e-01 9.35337126e-01
5.53199649e-01 -1.71118602e-01 2.30527595e-01 6.02630377e-01
4.72950280e-01 6.48390949e-01 -1.47118196e-01 -1.18736517e+00
2.59996772e-01 4.55118328e-01 3.07852834e-01 -1.36401486e+00
-2.44138241e-01 -5.58065116e-01 -1.59885788e+00 1.78107932e-01
4.40360516e-01 -3.53550941e-01 -4.36340600e-01 1.48671067e+00
5.16701460e-01 5.57130158e-01 -7.13370517e-02 9.03383970e-01
3.78186643e-01 4.52837050e-01 -4.64443654e-01 -6.29395008e-01
6.12171829e-01 -8.49687636e-01 -6.49875820e-01 5.78532398e-01
4.25600946e-01 -6.49587154e-01 4.47288454e-01 7.55898178e-01
-9.51046765e-01 -4.69517022e-01 -9.79483068e-01 4.59740043e-01
-1.00128690e-03 -9.30405483e-02 3.53940010e-01 7.65723169e-01
-1.18309784e+00 7.35733688e-01 -7.56905615e-01 -4.40470397e-01
1.23107098e-01 3.94909948e-01 3.13294120e-02 -7.93624595e-02
-7.92879939e-01 6.89769208e-01 2.70971566e-01 3.20158333e-01
-3.97179097e-01 -2.42228433e-01 -4.57693458e-01 -1.79313615e-01
5.42563617e-01 -9.58827317e-01 9.20808494e-01 -8.20417702e-01
-1.58555305e+00 4.69052374e-01 -3.82294714e-01 -8.84618610e-02
8.47012103e-01 -2.12812364e-01 -3.29756916e-01 1.04173245e-02
-1.39736488e-01 -6.86230659e-02 8.00935447e-01 -1.55464625e+00
-6.72802150e-01 -8.46861184e-01 -2.18496516e-01 1.22035950e-01
-2.29276732e-01 -1.89718097e-01 -3.48433778e-02 -6.07205749e-01
4.83445138e-01 -1.05180609e+00 -6.47992134e-01 -4.97092962e-01
-5.16218603e-01 -5.71229398e-01 7.78855145e-01 -2.24124789e-01
1.46582007e+00 -1.94596410e+00 6.22008801e-01 6.09620273e-01
4.69636947e-01 -2.07607478e-01 1.96716905e-01 5.34586608e-01
3.51269059e-02 3.41202199e-01 -4.53197151e-01 -4.80114430e-01
2.23453596e-01 1.26770914e-01 -3.39572549e-01 8.11259508e-01
-4.42068309e-01 5.95259249e-01 -1.02296996e+00 -6.27244651e-01
2.15390339e-01 2.60920316e-01 -2.69660175e-01 3.30355503e-02
8.71234462e-02 6.53481901e-01 -4.52409565e-01 4.27464783e-01
9.09435749e-01 -5.28345525e-01 2.83359498e-01 -2.11556822e-01
8.61334652e-02 -5.73675394e-01 -1.81008530e+00 1.52011263e+00
-1.11700278e-02 -1.61442682e-02 3.50196511e-01 -1.35386324e+00
7.39738703e-01 3.37467998e-01 7.02726185e-01 1.07072361e-01
2.70045578e-01 4.49101001e-01 -3.58614773e-02 -3.32342386e-01
4.52106476e-01 -2.01313809e-01 -1.63032994e-01 7.75999367e-01
-3.56720597e-01 1.79667156e-02 -7.11281970e-03 3.29656571e-01
7.79472113e-01 -1.76452473e-01 2.09885091e-01 -6.11091793e-01
5.74677527e-01 -3.63183290e-01 7.30449378e-01 8.14529836e-01
-2.09490329e-01 7.13260472e-01 5.99988736e-02 -7.38775879e-02
-1.12114489e+00 -1.18428767e+00 -1.44942507e-01 7.87361860e-01
3.55901569e-01 -6.31298721e-01 -9.33175504e-01 -6.12466276e-01
-1.43298477e-01 3.98997635e-01 -5.47561228e-01 6.45339023e-03
-2.47061685e-01 -9.13009405e-01 1.65949062e-01 5.60422897e-01
6.41481519e-01 -1.49946570e-01 5.54415258e-03 1.03628905e-02
-1.75479323e-01 -7.51721382e-01 -4.79335845e-01 -2.61582196e-01
-1.06696832e+00 -1.26040375e+00 -6.91059589e-01 -4.87207562e-01
9.48033512e-01 5.95835805e-01 9.91169035e-01 2.90858537e-01
-7.90517870e-03 6.97785437e-01 -4.27154154e-01 -4.39352959e-01
-1.49918139e-01 1.06732443e-01 6.28901601e-01 2.98840463e-01
3.38406175e-01 -6.74189806e-01 -7.89021194e-01 5.83966911e-01
-7.82032967e-01 -1.04834616e-01 3.36806804e-01 8.35631907e-01
6.60974085e-01 5.43655634e-01 8.03383231e-01 -6.96428597e-01
6.50459051e-01 -4.43702370e-01 -5.75886607e-01 4.98115867e-01
-6.60521269e-01 1.49840131e-01 6.39966428e-01 1.26073897e-01
-8.43115091e-01 1.00623094e-01 2.77631551e-01 -3.01133692e-01
-1.37279987e-01 4.94238049e-01 -3.23264956e-01 -2.16409620e-02
2.37072825e-01 1.31335676e-01 -1.25629380e-01 -4.10359293e-01
5.44509113e-01 9.07157123e-01 2.98311204e-01 -1.06082392e+00
1.07136202e+00 8.31737757e-01 1.09705582e-01 -8.70274127e-01
-1.16675580e+00 -5.94384968e-01 -9.98677969e-01 -4.07137513e-01
7.23557234e-01 -4.98355538e-01 -9.67343032e-01 4.67781484e-01
-8.91772568e-01 1.55243173e-01 1.80475160e-01 2.85845131e-01
-6.51049674e-01 7.69421279e-01 -1.99036852e-01 -1.19683087e+00
-3.40258837e-01 -9.72445011e-01 1.00886440e+00 2.08021209e-01
4.23880331e-02 -1.40669119e+00 3.76983821e-01 3.45632344e-01
1.21329591e-01 5.29796660e-01 5.23766696e-01 -5.24442971e-01
-6.35822356e-01 -2.68938225e-02 -8.25677291e-02 4.28717881e-01
6.99551031e-02 5.07096350e-01 -6.47074223e-01 -5.80727518e-01
2.28086635e-01 -3.66258994e-02 1.02993536e+00 4.94150370e-01
1.61063886e+00 -7.88359493e-02 -3.82333606e-01 4.25995171e-01
1.46499908e+00 -1.92485064e-01 2.43634865e-01 -1.56907246e-01
5.49807012e-01 4.40686554e-01 5.93328893e-01 5.29084027e-01
4.85321671e-01 3.34618062e-01 3.00888807e-01 2.03917041e-01
6.91097081e-01 4.49536033e-02 6.79690810e-03 1.59320819e+00
-7.00342953e-01 -7.90589824e-02 -9.68997300e-01 2.16690511e-01
-2.57525325e+00 -1.15142179e+00 -3.78613204e-01 2.17847943e+00
6.70675099e-01 -1.61014304e-01 1.85471356e-01 4.75878656e-01
1.13944483e+00 1.05952825e-02 -3.72992933e-01 -2.10101139e-02
-2.48823389e-01 7.33982921e-02 3.50247085e-01 6.33399367e-01
-1.16076267e+00 4.30459321e-01 6.98054075e+00 8.91683042e-01
-5.95124662e-01 5.83767593e-02 7.18577325e-01 8.84426013e-02
-2.62397468e-01 -2.03213375e-02 -6.87545598e-01 6.77390873e-01
4.28661466e-01 -3.50157619e-01 2.14170784e-01 7.42533863e-01
2.20778540e-01 -1.94039211e-01 -1.13916314e+00 1.13267040e+00
3.32722455e-01 -1.13912213e+00 -2.92476088e-01 9.51556116e-02
1.47898495e+00 -2.02960774e-01 2.58677185e-01 -1.35645315e-01
7.46135056e-01 -1.26207268e+00 3.36635113e-01 6.50904655e-01
5.11507154e-01 -1.00654840e+00 8.30644786e-01 6.54643893e-01
-1.12641311e+00 -1.28514305e-01 -5.82265675e-01 -3.28775615e-01
1.56016961e-01 1.18295813e+00 -1.73655555e-01 1.06029367e+00
7.49055684e-01 9.67658997e-01 -5.10831594e-01 1.27855241e+00
-2.13582262e-01 6.97608829e-01 -6.24611259e-01 -1.51039809e-01
1.10358179e-01 -8.24853718e-01 6.44820333e-01 7.98652112e-01
6.18956745e-01 5.17868102e-01 3.47786903e-01 5.89795589e-01
-3.44210342e-02 -3.99540700e-02 -5.82752943e-01 5.43733180e-01
6.36063159e-01 1.39375818e+00 -7.80214727e-01 -3.11884135e-01
2.02605017e-02 6.63304090e-01 2.65377700e-01 4.12038445e-01
-8.37688148e-01 -1.21731631e-01 3.64863783e-01 -2.98441559e-01
1.55471265e-01 -4.67892170e-01 -5.80675960e-01 -1.37336004e+00
-1.69944782e-02 -7.14396179e-01 5.69387972e-01 -5.33288836e-01
-1.92462254e+00 6.81043491e-02 -1.47234902e-01 -1.29801464e+00
2.26982519e-01 -5.49994648e-01 -8.99174571e-01 4.35120821e-01
-8.20090652e-01 -7.68161476e-01 -4.15011793e-01 5.71444392e-01
1.60357043e-01 1.27727404e-01 7.69932270e-01 8.71306211e-02
-8.42936158e-01 4.10211891e-01 1.04445231e+00 4.44049612e-02
9.63424802e-01 -1.70749915e+00 -2.63318032e-01 6.74231589e-01
-4.30534855e-02 6.30342185e-01 8.51476073e-01 -4.66766775e-01
-1.25573742e+00 -1.02610600e+00 4.87835884e-01 -7.38759875e-01
7.28230655e-01 -1.95588470e-01 -7.84475923e-01 5.70919991e-01
4.07507539e-01 -1.20302811e-01 1.02569211e+00 3.37088794e-01
-2.32095346e-01 -2.09190890e-01 -1.11288977e+00 5.09867847e-01
1.13361800e+00 -2.47601289e-02 -7.36519754e-01 3.45202148e-01
2.61445582e-01 -1.14870824e-01 -1.01648128e+00 5.60531318e-01
4.34717506e-01 -1.19015408e+00 7.35762119e-01 -6.19127631e-01
3.06717247e-01 -6.76082134e-01 -3.30251276e-01 -1.78640842e+00
-4.77181017e-01 -6.67723715e-01 -1.62695780e-01 1.20713401e+00
3.06887664e-02 -8.00229073e-01 7.42197037e-01 3.49861205e-01
9.42615196e-02 -8.65551174e-01 -9.89597142e-01 -9.47597086e-01
4.61188436e-01 -2.91655928e-01 5.33651710e-01 1.16109133e+00
2.72355109e-01 3.15836847e-01 -3.97821099e-01 2.29238540e-01
1.45347464e+00 3.29895318e-01 8.03941607e-01 -1.63135970e+00
-2.21860871e-01 -4.57797021e-01 -3.53710413e-01 -9.10695434e-01
6.80115744e-02 -9.16906416e-01 -1.80766955e-02 -1.45893133e+00
5.79816878e-01 -6.43195868e-01 -4.78886515e-01 -3.66774589e-01
-4.58006501e-01 6.06768578e-02 -1.13062393e-02 2.45033428e-01
-1.43146896e+00 5.65196276e-01 1.10457873e+00 3.54933888e-02
2.00958714e-01 5.63784987e-02 -7.50806570e-01 1.08588135e+00
9.90367353e-01 -2.72546291e-01 -3.74261707e-01 -3.64591986e-01
4.53412414e-01 -1.28243923e-01 3.25323284e-01 -8.67156744e-01
5.67576051e-01 -3.05962682e-01 2.22823486e-01 -9.75502014e-01
-5.18576652e-02 -9.83356893e-01 1.45348176e-01 3.59923393e-01
-1.05785094e-01 1.28010854e-01 -5.27331889e-01 1.03630495e+00
-1.83480591e-01 1.21091707e-02 9.90083694e-01 9.87184271e-02
-2.06304997e-01 5.41103840e-01 1.69605333e-02 2.63814032e-01
1.41041696e+00 -1.57927841e-01 -3.24520051e-01 -4.71359104e-01
-8.27528775e-01 6.66651011e-01 7.88208902e-01 4.11396706e-03
5.15447140e-01 -1.38134515e+00 -9.73087847e-01 -2.30559871e-01
-4.36659157e-02 3.25864851e-01 3.77568334e-01 1.29369771e+00
-2.29273558e-01 1.45742923e-01 2.63092279e-01 -1.04069495e+00
-1.23934019e+00 5.52127481e-01 3.34450394e-01 -2.19199043e-02
-2.43258998e-01 6.46554053e-01 -4.81077982e-03 -5.04361153e-01
3.62920970e-01 2.68085778e-01 1.72071934e-01 -1.56822905e-01
4.07180935e-01 1.09427023e+00 -2.47482464e-01 -4.82002765e-01
-2.82801390e-01 9.54101682e-01 2.06805006e-01 4.64658737e-02
1.42835355e+00 -4.06335056e-01 -6.17920578e-01 7.40728319e-01
8.66974235e-01 2.39526685e-02 -8.62608612e-01 -3.07431549e-01
2.37645701e-01 -3.31672609e-01 -2.07073599e-01 -4.89806324e-01
-9.52677190e-01 9.89719808e-01 4.30096596e-01 5.13055384e-01
1.00902510e+00 4.62797731e-02 6.19224548e-01 3.51481199e-01
7.09842205e-01 -1.42179263e+00 -3.01189977e-03 5.90811908e-01
6.72857642e-01 -1.32579088e+00 1.56549573e-01 -4.28663105e-01
-3.06900352e-01 1.05567896e+00 4.89218116e-01 -4.27113473e-01
1.06416702e+00 1.45574525e-01 -1.55914932e-01 1.60804519e-03
-7.00270236e-01 5.97808771e-02 2.15480685e-01 4.86951619e-01
3.21154624e-01 2.65066504e-01 -3.29997569e-01 7.36310720e-01
-1.87081590e-01 -2.47873247e-01 4.11487818e-01 5.33558071e-01
-7.50957131e-01 -1.08238935e+00 -6.71047628e-01 5.16491175e-01
-2.93649763e-01 2.67635405e-01 -5.65599144e-01 2.83777267e-01
1.70201540e-01 1.20232582e+00 -1.71098232e-01 -5.24465561e-01
-1.14720710e-01 3.72509398e-02 6.02256000e-01 -4.04018551e-01
-1.24854527e-01 -6.50253817e-02 -4.31100756e-01 -2.31560081e-01
-8.76419842e-01 -8.49329233e-01 -1.08412552e+00 -7.44169235e-01
-5.64004898e-01 5.74967444e-01 2.66241521e-01 1.03014231e+00
4.02881414e-01 1.00207090e-01 9.07614708e-01 -7.07056403e-01
-9.35094893e-01 -9.04335141e-01 -9.06656027e-01 2.82845110e-01
-7.39046419e-03 -7.12264597e-01 -8.37722838e-01 -1.02988511e-01] | [7.9371490478515625, 4.627134799957275] |
08ab96c6-044b-41ea-93a7-86d411be3117 | dehazed-image-quality-evaluation-from-partial | 2211.12636 | null | https://arxiv.org/abs/2211.12636v1 | https://arxiv.org/pdf/2211.12636v1.pdf | Dehazed Image Quality Evaluation: From Partial Discrepancy to Blind Perception | Image dehazing aims to restore spatial details from hazy images. There have emerged a number of image dehazing algorithms, designed to increase the visibility of those hazy images. However, much less work has been focused on evaluating the visual quality of dehazed images. In this paper, we propose a Reduced-Reference dehazed image quality evaluation approach based on Partial Discrepancy (RRPD) and then extend it to a No-Reference quality assessment metric with Blind Perception (NRBP). Specifically, inspired by the hierarchical characteristics of the human perceiving dehazed images, we introduce three groups of features: luminance discrimination, color appearance, and overall naturalness. In the proposed RRPD, the combined distance between a set of sender and receiver features is adopted to quantify the perceptually dehazed image quality. By integrating global and local channels from dehazed images, the RRPD is converted to NRBP which does not rely on any information from the references. Extensive experiment results on several dehazed image quality databases demonstrate that our proposed methods outperform state-of-the-art full-reference, reduced-reference, and no-reference quality assessment models. Furthermore, we show that the proposed dehazed image quality evaluation methods can be effectively applied to tune parameters for potential image dehazing algorithms. | ['Huiyan Chen', 'Hantao Liu', 'Leida Li', 'Ruizeng Zhang', 'Wei Zhou'] | 2022-11-22 | null | null | null | null | ['image-dehazing'] | ['computer-vision'] | [ 3.25904816e-01 -6.03985012e-01 4.18510884e-01 -2.44261011e-01
-5.70565522e-01 -1.46401018e-01 5.39126158e-01 -1.03647090e-01
-2.16455415e-01 5.64307034e-01 3.15668374e-01 -1.14020415e-01
-3.17136019e-01 -1.05904329e+00 -3.09440613e-01 -1.15186191e+00
1.13828436e-01 -8.03382337e-01 4.17577684e-01 -4.62752163e-01
4.81747180e-01 3.06532502e-01 -1.88321149e+00 1.46466956e-01
1.26017845e+00 1.24834383e+00 1.64599821e-01 8.68179560e-01
4.25458908e-01 7.03702927e-01 -1.02412653e+00 -2.90201038e-01
4.74123895e-01 -6.74811244e-01 -2.68433273e-01 2.89459825e-01
4.00101811e-01 -7.32399821e-01 -3.11916113e-01 1.41199052e+00
6.15962148e-01 2.24420041e-01 6.98106468e-01 -1.18309176e+00
-1.30214202e+00 -2.83445150e-01 -4.09458488e-01 6.39808238e-01
2.30795696e-01 3.21294338e-01 5.16148984e-01 -9.51265216e-01
-1.60569832e-01 1.19968522e+00 2.88541764e-01 7.70051405e-02
-9.01206791e-01 -6.16668344e-01 -2.38955051e-01 6.62236691e-01
-1.57428384e+00 -4.06892717e-01 8.39374483e-01 -3.60689312e-01
4.56362098e-01 3.22023392e-01 4.16216403e-01 1.92441836e-01
4.91131276e-01 2.11120054e-01 1.55921221e+00 -6.12756670e-01
2.49124900e-01 1.73725218e-01 -9.94531736e-02 5.98168314e-01
4.84032124e-01 4.98725027e-01 -2.52046436e-01 1.82025805e-01
5.42975485e-01 6.30464107e-02 -7.36308396e-01 -1.02669261e-01
-8.16279829e-01 5.77443242e-01 5.19517422e-01 2.98557401e-01
-2.94476509e-01 -3.03383172e-01 -2.51988769e-01 4.21637177e-01
5.66959798e-01 3.75939727e-01 2.50361562e-01 2.13061839e-01
-8.49855363e-01 -1.48226395e-02 3.04995298e-01 6.31097198e-01
1.03598833e+00 1.09547086e-01 -3.14502329e-01 1.00010562e+00
2.63227135e-01 6.63742840e-01 3.92277062e-01 -9.68738973e-01
2.33061314e-01 4.44529623e-01 3.78174156e-01 -1.46641982e+00
2.31455892e-01 -1.73319235e-01 -1.00117612e+00 8.24842691e-01
-4.49170135e-02 1.68512970e-01 -9.56593752e-01 8.94849062e-01
1.32580683e-01 -4.38996218e-02 2.53393352e-01 1.28501904e+00
7.46614337e-01 1.12933016e+00 -2.56989092e-01 -2.12747142e-01
1.08129883e+00 -9.02003288e-01 -8.67045224e-01 1.06194966e-01
-8.11718479e-02 -1.04000366e+00 9.48810637e-01 7.31753051e-01
-1.26026511e+00 -9.65221584e-01 -1.56370997e+00 -3.22985346e-03
-4.94961649e-01 -1.09253630e-01 -1.45874187e-01 1.14008975e+00
-1.42220521e+00 3.32140446e-01 -2.78314412e-01 1.89653803e-02
1.65345058e-01 9.70001221e-02 -1.64271533e-01 -6.67693317e-01
-1.24205649e+00 1.06377482e+00 3.11858058e-01 2.01443315e-01
-9.72247005e-01 -5.63304603e-01 -7.18272567e-01 2.20117494e-02
7.00496882e-02 -4.86240774e-01 7.09823012e-01 -9.36983287e-01
-1.56005478e+00 5.69037974e-01 1.10363677e-01 -2.76947707e-01
1.24770617e-02 1.94859747e-02 -1.05994558e+00 5.98772407e-01
-2.30438828e-01 2.83216059e-01 1.33092582e+00 -1.71800363e+00
-7.90237784e-01 -2.83262789e-01 2.21162453e-01 4.41739082e-01
-2.94270992e-01 1.13068663e-01 -3.29391539e-01 -8.16045344e-01
-9.40817595e-02 -2.37356380e-01 1.68959111e-01 2.95567334e-01
-1.26781926e-01 2.60324210e-01 9.88659322e-01 -9.81816053e-01
1.36744142e+00 -2.50834084e+00 -1.39169097e-01 2.10772961e-01
3.38187426e-01 7.21453071e-01 -2.55852312e-01 2.42823049e-01
1.98067397e-01 1.21196665e-01 -5.60896099e-01 1.57153413e-01
-1.38382554e-01 5.69283543e-03 3.13830450e-02 6.33244216e-01
2.05846742e-01 3.47934008e-01 -7.21272290e-01 -5.53615987e-01
5.82576990e-01 7.77853906e-01 -1.78881511e-01 6.22859001e-01
4.06205595e-01 2.12711871e-01 -1.08175846e-02 6.22038126e-01
1.23404026e+00 1.97029233e-01 -6.98347867e-01 -2.08438993e-01
-3.87160629e-01 -1.56352684e-01 -9.30406272e-01 9.23632145e-01
-5.79721212e-01 6.67977393e-01 1.33142695e-01 -6.09153092e-01
1.07396352e+00 1.94194362e-01 -8.55450854e-02 -1.00668609e+00
4.60886322e-02 8.22439939e-02 -1.24522768e-01 -6.33029222e-01
6.56117797e-01 -1.62789047e-01 5.34407318e-01 8.51274580e-02
-1.21046036e-01 -4.30409908e-01 2.20962968e-02 -1.00059681e-01
7.60726452e-01 -3.70919406e-01 3.80498350e-01 -1.56883329e-01
1.00464880e+00 -4.02790993e-01 1.35077804e-01 5.31387508e-01
-5.09530663e-01 9.99066710e-01 -8.47823024e-02 -1.37030527e-01
-1.09682405e+00 -1.27463055e+00 -2.59092122e-01 8.04605126e-01
8.15027952e-01 -6.38743415e-02 -9.08700228e-01 -1.33030806e-02
-3.58720869e-01 6.58227682e-01 -5.16763330e-01 -4.52524722e-01
-1.76999852e-01 -7.85740435e-01 2.75701612e-01 -2.28704929e-01
1.27905166e+00 -6.87242031e-01 -3.96816701e-01 -7.46716335e-02
-2.77702719e-01 -8.36430311e-01 -6.08381808e-01 -5.51754534e-01
-5.50326943e-01 -9.04937267e-01 -1.09948933e+00 -7.70288706e-01
5.84027767e-01 1.12161005e+00 6.75581753e-01 4.02260959e-01
-2.32277751e-01 4.43700641e-01 -7.00327277e-01 -2.82399178e-01
-4.59698617e-01 -7.36371815e-01 -2.69997329e-01 5.73960483e-01
1.24173261e-01 -5.62062383e-01 -1.19702852e+00 5.89256942e-01
-1.40670621e+00 -2.54461080e-01 8.85525703e-01 6.21877491e-01
4.20404434e-01 9.66753125e-01 1.86365783e-01 -1.58644840e-01
7.37223029e-01 -4.27179039e-01 -6.39494479e-01 2.04034194e-01
-9.02323484e-01 -1.42398492e-01 4.29132134e-01 -2.30563328e-01
-1.43912041e+00 -8.61533880e-01 9.29834023e-02 -2.69807279e-01
-2.31385723e-01 2.51598597e-01 -4.21862215e-01 -5.71548462e-01
5.55909932e-01 5.56572556e-01 7.12184906e-02 -3.68083358e-01
3.40092897e-01 1.03415811e+00 8.32548320e-01 -1.40220597e-01
1.16002595e+00 5.74945033e-01 -1.55759737e-01 -1.09717250e+00
-3.93039823e-01 -6.08810246e-01 -1.79323405e-01 -3.23016852e-01
1.04183877e+00 -9.55074191e-01 -4.46294516e-01 8.20316970e-01
-1.00552094e+00 -7.87756313e-03 1.51916221e-01 4.97235894e-01
-1.25810593e-01 8.46076667e-01 -4.10736710e-01 -9.77650881e-01
-3.27180028e-01 -1.13917971e+00 9.15800571e-01 3.38480949e-01
6.24035358e-01 -9.04817045e-01 -9.94800255e-02 3.74398470e-01
8.79140854e-01 1.49818048e-01 9.90174472e-01 2.22163454e-01
-8.10488462e-01 -3.15000266e-02 -5.89029014e-01 1.09311926e+00
5.74691832e-01 -1.34698406e-01 -9.00498867e-01 -3.00955355e-01
4.00648206e-01 2.11872503e-01 7.21879303e-01 3.07573110e-01
9.55327749e-01 -4.99798149e-01 3.08318585e-01 7.93677628e-01
1.64751458e+00 4.70854640e-01 1.34056675e+00 5.31329691e-01
3.55960667e-01 5.36830485e-01 7.68405735e-01 2.85289824e-01
1.46709979e-01 5.56399226e-01 6.54931188e-01 -4.97152001e-01
-3.55976582e-01 8.51612687e-02 4.09212321e-01 7.04126179e-01
-2.67999738e-01 -4.69965994e-01 -5.32698095e-01 5.59644163e-01
-1.06146967e+00 -8.55886161e-01 -5.74079640e-02 2.32731414e+00
6.91206515e-01 -7.42039233e-02 -1.58106267e-01 4.29692954e-01
9.43127632e-01 3.34447384e-01 -1.07545078e-01 -3.43574107e-01
-4.13126290e-01 1.65198147e-01 5.79679549e-01 5.61345100e-01
-1.01050079e+00 5.16867816e-01 5.78631067e+00 8.38097155e-01
-8.80520344e-01 7.52407387e-02 5.71184933e-01 3.25249910e-01
-1.83730766e-01 -1.14479423e-01 -2.44131312e-01 6.96248531e-01
8.64700258e-01 -4.49611664e-01 6.73669636e-01 3.72134954e-01
3.73104364e-01 -4.20648903e-01 -5.18809974e-01 1.22199881e+00
4.94844109e-01 -8.41454566e-01 3.63698930e-01 1.01914562e-01
8.11088979e-01 -7.10976601e-01 6.58952355e-01 -2.24516541e-01
7.89966509e-02 -1.05798268e+00 5.26401818e-01 6.60650849e-01
7.87793279e-01 -7.82747984e-01 1.00434577e+00 2.53913701e-02
-1.05847383e+00 -9.32913199e-02 -6.45382106e-01 -1.45595707e-02
-4.73730154e-02 5.29322267e-01 -3.40892702e-01 8.65703762e-01
1.04675138e+00 4.63338882e-01 -8.68993521e-01 1.37737405e+00
-2.41890743e-01 4.97413397e-01 2.91878074e-01 2.88085014e-01
1.49274752e-01 -3.86951536e-01 3.70215625e-01 8.62404644e-01
6.63900435e-01 6.92739904e-01 -4.75825936e-01 8.13068569e-01
2.67827839e-01 4.11341749e-02 -4.31329221e-01 3.19629490e-01
3.58420700e-01 8.83134842e-01 -1.36399820e-01 -2.82669872e-01
-5.07731616e-01 1.04092801e+00 -5.55316091e-01 5.95611751e-01
-6.35989845e-01 -9.44382191e-01 7.74425030e-01 -5.29064983e-02
3.33367735e-01 -2.08828717e-01 3.18724401e-02 -8.36974740e-01
5.44969141e-02 -9.83372688e-01 1.03983819e-01 -1.39494336e+00
-1.30636060e+00 6.49980009e-01 8.14150274e-02 -1.58035421e+00
3.93812060e-01 -6.48226976e-01 -5.20988643e-01 1.23034692e+00
-2.03264403e+00 -7.40308881e-01 -8.70504677e-01 8.71687472e-01
4.60246086e-01 1.61409695e-02 4.54635978e-01 3.97239506e-01
-3.30237985e-01 4.94079918e-01 2.33880460e-01 1.32060498e-02
7.42047548e-01 -1.08922553e+00 -3.88373323e-02 1.36715972e+00
-5.16506910e-01 3.77832055e-01 8.27500403e-01 -2.85453975e-01
-1.03571057e+00 -1.15311372e+00 3.12643826e-01 -1.40639752e-01
2.59041458e-01 2.49516785e-01 -1.19125283e+00 2.64756046e-02
5.61463296e-01 1.28764454e-02 5.19369125e-01 -8.29044044e-01
-3.95073116e-01 -5.08222461e-01 -1.30989897e+00 5.19031227e-01
5.80106795e-01 -4.71315533e-01 -7.20805407e-01 -1.63721442e-01
9.15734172e-01 6.88408539e-02 -8.85664642e-01 3.68747562e-01
3.45041364e-01 -1.44991696e+00 1.16534197e+00 3.84102196e-01
3.53706986e-01 -7.73549676e-01 -3.86690915e-01 -1.49588561e+00
-4.66994941e-01 -5.22732973e-01 1.83110133e-01 1.13044441e+00
5.84938154e-02 -6.88337922e-01 1.35198846e-01 1.33155122e-01
-1.93886891e-01 -2.39050224e-01 -5.97239196e-01 -8.65094960e-01
-1.64896071e-01 1.11176625e-01 8.72510374e-01 6.38819337e-01
-3.84708345e-01 -2.25110888e-01 -6.17087841e-01 8.77778709e-01
9.79522049e-01 -2.01616675e-01 5.29580235e-01 -9.24714625e-01
-8.73839259e-02 -2.85455436e-01 -8.10917437e-01 -8.23705554e-01
-4.71230358e-01 -2.19395190e-01 2.49542326e-01 -1.64983082e+00
-4.78112735e-02 -2.29062978e-02 -5.52111268e-01 -3.06705851e-02
-5.25884449e-01 6.05312228e-01 6.71206489e-02 4.35164720e-01
-3.43714982e-01 7.89236844e-01 1.35936606e+00 -4.26008135e-01
-1.65066555e-01 -9.75216106e-02 -7.18176365e-01 2.86426216e-01
8.69944692e-01 -1.89723119e-01 -5.27479231e-01 -4.46879417e-01
-2.70744652e-01 4.52535823e-02 5.47316432e-01 -1.41620743e+00
2.82437444e-01 -3.26813310e-01 3.74369830e-01 -3.41879755e-01
4.05852735e-01 -8.59034956e-01 -6.04515662e-04 2.67887563e-01
-4.80917729e-02 1.92534905e-02 -1.00894973e-01 7.14116573e-01
-7.49583781e-01 -1.94220394e-02 1.16157675e+00 -3.74583043e-02
-8.84318113e-01 9.39393416e-02 -4.64371234e-01 -3.36740613e-01
9.73671317e-01 -7.53027678e-01 -8.12198341e-01 -5.08975267e-01
-2.10569337e-01 -2.87947595e-01 6.31967008e-01 2.65612185e-01
1.30477297e+00 -1.16310978e+00 -8.46009910e-01 2.93139607e-01
4.14782465e-01 -3.80010515e-01 5.85865796e-01 6.24246955e-01
-8.86070848e-01 2.03686401e-01 -5.07313132e-01 -3.51311356e-01
-1.35544717e+00 1.01444149e+00 4.73201334e-01 2.73774296e-01
-4.07815516e-01 5.18119633e-01 4.91456449e-01 3.07154655e-01
1.84344146e-02 -2.42716432e-01 -4.52867359e-01 -5.05551994e-01
9.42883968e-01 7.55055428e-01 1.83361992e-01 -8.39363754e-01
1.76348817e-02 7.07185090e-01 1.91155195e-01 -2.27423638e-01
1.01736736e+00 -7.68663764e-01 -3.98583919e-01 8.41627866e-02
1.26511252e+00 6.11946695e-02 -1.22007382e+00 -2.38102466e-01
-3.57257515e-01 -1.12750363e+00 4.93104577e-01 -8.10004711e-01
-9.86815035e-01 1.04499960e+00 1.24275303e+00 5.19440830e-01
1.89776957e+00 -4.71892953e-01 6.16438806e-01 3.75928450e-03
3.08049321e-01 -7.03099549e-01 3.21827888e-01 -9.13036242e-02
8.94185066e-01 -1.10060620e+00 9.71147791e-02 -4.89888370e-01
-3.71050209e-01 8.40171993e-01 4.82859492e-01 5.29495738e-02
6.03310525e-01 -2.78141975e-01 3.19031686e-01 3.03642750e-02
-3.08468074e-01 -3.73759657e-01 4.90670741e-01 9.83514249e-01
-5.89663573e-02 -2.24767402e-01 -1.79893479e-01 8.98793191e-02
-8.63890797e-02 -2.04645067e-01 8.32474470e-01 6.76017761e-01
-9.43839192e-01 -7.06095815e-01 -9.16074693e-01 1.08316861e-01
-3.75539094e-01 -3.25902522e-01 -5.12109213e-02 5.86073458e-01
3.48263115e-01 1.80493343e+00 -1.78163186e-01 -4.98213291e-01
3.93816948e-01 -4.86003190e-01 3.60052258e-01 -3.31742704e-01
-2.91312318e-02 -1.14452191e-01 -4.61409092e-01 -2.85604537e-01
-7.25604355e-01 1.49890840e-01 -5.86929858e-01 -4.70286399e-01
-3.31751972e-01 3.35963368e-01 5.63009799e-01 5.62339306e-01
1.87038243e-01 4.73297864e-01 1.00848842e+00 -8.49431157e-01
-1.69419870e-01 -9.03907418e-01 -1.01033056e+00 3.80932242e-01
9.87467527e-01 -6.34529591e-01 -8.60022366e-01 2.36462757e-01] | [10.94222640991211, -3.045412540435791] |
3e5919d3-5e26-4318-a039-43675f02d6a2 | emt-nas-transferring-architectural-knowledge | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Liao_EMT-NASTransferring_Architectural_Knowledge_Between_Tasks_From_Different_Datasets_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Liao_EMT-NASTransferring_Architectural_Knowledge_Between_Tasks_From_Different_Datasets_CVPR_2023_paper.pdf | EMT-NAS:Transferring Architectural Knowledge Between Tasks From Different Datasets | The success of multi-task learning (MTL) can largely be attributed to the shared representation of related tasks, allowing the models to better generalise. In deep learning, this is usually achieved by sharing a common neural network architecture and jointly training the weights. However, the joint training of weighting parameters on multiple related tasks may lead to performance degradation, known as negative transfer. To address this issue, this work proposes an evolutionary multi-tasking neural architecture search (EMT-NAS) algorithm to accelerate the search process by transferring architectural knowledge across multiple related tasks. In EMT-NAS, unlike the traditional MTL, the model for each task has a personalised network architecture and its own weights, thus offering the capability of effectively alleviating negative transfer. A fitness re-evaluation method is suggested to alleviate fluctuations in performance evaluations resulting from parameter sharing and the mini-batch gradient descent training method, thereby avoiding losing promising solutions during the search process. To rigorously verify the performance of EMT-NAS, the classification tasks used in the empirical assessments are derived from different datasets, including the CIFAR-10 and CIFAR-100, and four MedMNIST datasets. Extensive comparative experiments on different numbers of tasks demonstrate that EMT-NAS takes 8% and up to 40% on CIFAR and MedMNIST, respectively, less time to find competitive neural architectures than its single-task counterparts. | ['Wenli Du', 'Yaochu Jin', 'Peng Liao'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['architecture-search'] | ['methodology'] | [ 2.20702186e-01 -3.44224691e-01 2.98692554e-01 -3.41265768e-01
-6.30097091e-01 -2.72571027e-01 2.21275195e-01 -2.34564468e-02
-7.81517565e-01 6.86487317e-01 -2.43030384e-01 2.43373085e-02
-5.47688365e-01 -2.96225786e-01 -5.67775607e-01 -8.52768004e-01
1.58430729e-02 4.89647359e-01 8.08294341e-02 -1.66949719e-01
1.94053292e-01 1.04552492e-01 -1.55584037e+00 2.29361683e-01
1.14774263e+00 1.02466893e+00 6.46726668e-01 3.06896091e-01
-2.22560354e-02 3.77919972e-01 -9.48709965e-01 -6.60422027e-01
2.52527595e-01 -1.62254006e-01 -7.15512156e-01 -2.05592662e-01
1.24043830e-01 2.09341466e-01 1.70488790e-01 8.93957734e-01
7.57703125e-01 5.03220260e-01 6.00015461e-01 -1.33483541e+00
-5.46823025e-01 5.09854555e-01 -4.91491139e-01 2.28323355e-01
-4.72653508e-01 -1.53687537e-01 1.04242563e+00 -9.83766377e-01
5.48501674e-04 1.11379635e+00 6.99577153e-01 6.25650465e-01
-1.32881832e+00 -7.89755046e-01 2.56900191e-01 1.12958416e-01
-1.56527281e+00 -2.90333033e-01 7.30136633e-01 -2.93576688e-01
9.69479322e-01 2.02565581e-01 4.80577916e-01 1.04529011e+00
3.11956495e-01 7.61074483e-01 8.28037083e-01 -4.59011972e-01
4.22691375e-01 3.48142415e-01 7.73482844e-02 4.20691907e-01
4.51146394e-01 -1.06689915e-01 -6.94373071e-01 -3.69799882e-01
3.85629505e-01 -1.52864486e-01 -2.78333426e-01 -4.12290484e-01
-8.58202577e-01 7.69509554e-01 4.43788707e-01 5.00049710e-01
-5.63379765e-01 1.07257843e-01 6.58272564e-01 5.21958828e-01
5.97089469e-01 7.73053944e-01 -6.46626234e-01 1.11393772e-01
-6.70379162e-01 1.18793406e-01 3.74243647e-01 3.52500528e-01
7.55866408e-01 2.51715392e-01 -2.93290317e-01 1.26264811e+00
2.36018673e-01 1.51229963e-01 1.09292138e+00 -7.51900434e-01
5.70138216e-01 6.69547737e-01 -2.46440277e-01 -7.65713394e-01
-4.09666479e-01 -1.14667654e+00 -9.73585725e-01 3.30634534e-01
1.89748108e-01 -4.06408519e-01 -6.17202997e-01 2.05999303e+00
9.13056433e-02 6.50734007e-02 3.06437284e-01 7.56484330e-01
3.81711334e-01 3.97099763e-01 2.73824334e-01 -3.78086530e-02
1.27589357e+00 -1.00426245e+00 -3.94313663e-01 -7.74628997e-01
8.58790696e-01 -6.12913847e-01 1.06144369e+00 4.70664471e-01
-8.38507771e-01 -8.08276594e-01 -1.06084371e+00 4.59432513e-01
-1.74285471e-01 3.52370113e-01 5.10309219e-01 6.81774378e-01
-9.71534669e-01 5.43968558e-01 -4.39188093e-01 1.16176412e-01
5.53065419e-01 5.09535611e-01 -8.14093873e-02 1.95619836e-02
-1.25179243e+00 9.81856942e-01 7.75138795e-01 1.39456928e-01
-8.02107275e-01 -8.76769543e-01 -3.06595325e-01 4.74194139e-01
1.82719752e-01 -7.19479978e-01 1.01207316e+00 -1.30270839e+00
-1.28012097e+00 5.54353595e-01 -2.34129392e-02 -4.28829342e-01
2.52049237e-01 -2.93535024e-01 -2.94856668e-01 -5.01151681e-01
-1.70919806e-01 7.80913532e-01 9.79455173e-01 -1.14739799e+00
-5.88420630e-01 -4.03185874e-01 -1.59725219e-01 5.11582196e-01
-1.07956553e+00 1.57172792e-02 -1.76840395e-01 -8.78214777e-01
-1.75031707e-01 -1.09617865e+00 -2.08331823e-01 -3.74108613e-01
-5.33237457e-02 -3.70748609e-01 6.63017154e-01 -2.40545481e-01
1.36867940e+00 -2.27149367e+00 4.99892533e-01 1.52114376e-01
1.62993386e-01 6.50636137e-01 -5.28485477e-01 -7.81316981e-02
-2.33494490e-01 -3.39279063e-02 -1.46962032e-01 -6.93657339e-01
-2.23940760e-01 2.68819898e-01 6.57930747e-02 -8.43330920e-02
1.20402552e-01 9.22991335e-01 -4.21316922e-01 -1.62353128e-01
-2.98569858e-01 4.13485914e-01 -3.09856325e-01 -1.61786497e-01
-3.08883134e-02 2.14197934e-01 -5.40163577e-01 3.31646591e-01
3.86438370e-01 -4.35798526e-01 2.78891563e-01 -1.01465873e-01
2.84481198e-01 -7.80769717e-03 -1.02421474e+00 1.73859584e+00
-7.27668941e-01 5.48675001e-01 -5.54849356e-02 -1.25005245e+00
1.20212579e+00 4.39553529e-01 2.97204494e-01 -9.79453981e-01
1.37408227e-01 4.39024061e-01 4.39110547e-01 5.11427410e-03
2.46704400e-01 1.24287777e-01 4.15564552e-02 6.28125787e-01
5.92031190e-03 3.15383583e-01 2.41472628e-02 -3.17922413e-01
7.50248849e-01 -2.13824153e-01 -1.26772821e-01 -4.54159498e-01
6.11315846e-01 -1.65074944e-01 8.04238975e-01 5.30479789e-01
-1.60689652e-01 6.96206018e-02 -3.13986763e-02 -6.43856764e-01
-8.22375655e-01 -7.07458258e-01 -1.78263307e-01 1.64393950e+00
-1.30084038e-01 -2.36720189e-01 -5.96869826e-01 -6.79043710e-01
-2.79055573e-02 7.80890405e-01 -6.19534969e-01 -6.36878967e-01
-6.09642923e-01 -1.35578132e+00 6.13338351e-01 4.13347304e-01
4.98740047e-01 -1.10693622e+00 -9.79401052e-01 3.51751685e-01
-1.98266447e-01 -6.88819706e-01 -4.21727389e-01 6.23297274e-01
-1.09778738e+00 -7.37322152e-01 -1.04021418e+00 -8.62385213e-01
6.89358473e-01 3.08636665e-01 1.05783212e+00 2.09778920e-01
-3.23991895e-01 -1.32824704e-01 -1.51908025e-01 -5.61369121e-01
-2.68848598e-01 4.58175570e-01 9.99682322e-02 1.78175673e-01
8.03681388e-02 -7.09769785e-01 -3.81373018e-01 6.15158498e-01
-7.20982492e-01 8.98082629e-02 8.65084052e-01 1.13278508e+00
5.34333050e-01 2.79484838e-01 1.01989484e+00 -4.65895504e-01
1.00916791e+00 -4.14964795e-01 -5.08945107e-01 7.26688862e-01
-9.98754919e-01 2.10989237e-01 5.50941765e-01 -8.44478130e-01
-1.21713591e+00 -9.86678004e-02 3.22467327e-01 -7.40789771e-01
3.22903365e-01 6.25286520e-01 1.42767921e-01 -2.78906047e-01
8.28483403e-01 2.42693380e-01 -1.01141974e-01 -6.13224804e-01
-7.33097568e-02 4.00495648e-01 2.12874353e-01 -6.08412385e-01
4.56926733e-01 -1.50697097e-01 -9.87678468e-02 -3.24110657e-01
-8.03088427e-01 -6.65090680e-02 -2.70637721e-01 -1.65163547e-01
6.35565460e-01 -8.42743993e-01 -5.71143866e-01 5.81416130e-01
-1.04056776e+00 -4.65528518e-01 -8.96989182e-02 6.05435252e-01
-7.60093555e-02 -1.06323436e-01 -2.49566689e-01 -6.63412035e-01
-5.96384346e-01 -1.31117392e+00 4.80212659e-01 2.22376198e-01
-2.48481587e-01 -1.15875733e+00 3.23949405e-03 4.19519275e-01
7.97871530e-01 -3.33552092e-01 1.35867548e+00 -1.01900661e+00
-1.51284067e-02 1.53875783e-01 -8.83575976e-02 5.59604406e-01
1.63464889e-01 -4.46960837e-01 -1.05965400e+00 -6.25145614e-01
1.88754290e-01 -6.40975595e-01 8.02052438e-01 4.30138618e-01
1.14105558e+00 -5.25843538e-02 -3.49308372e-01 3.77037019e-01
1.28515315e+00 5.75169384e-01 3.01552534e-01 6.64611578e-01
4.13572788e-01 7.48136342e-01 4.59342301e-01 3.65688741e-01
1.91383421e-01 8.11318219e-01 4.52187091e-01 -7.39290491e-02
-4.22211923e-02 3.96882534e-01 2.43237853e-01 7.11321294e-01
-2.00582966e-02 -3.07022482e-01 -1.02470851e+00 3.98562878e-01
-1.83220816e+00 -5.10414481e-01 8.82267058e-02 2.24468827e+00
7.40669429e-01 1.37972072e-01 -1.25714853e-01 1.41536027e-01
7.69575357e-01 -2.01883614e-01 -8.77343953e-01 -3.97736162e-01
-1.58696279e-01 1.11575775e-01 7.75867403e-02 -3.35674500e-04
-9.13184822e-01 5.96434176e-01 5.83555317e+00 1.07917297e+00
-1.25117970e+00 3.78331721e-01 9.86937582e-01 -4.14321721e-01
-1.44618213e-01 -3.73732626e-01 -7.72638917e-01 4.45125461e-01
8.51047218e-01 -4.17291105e-01 5.26929319e-01 8.25688541e-01
-5.91956228e-02 1.36347428e-01 -8.46153855e-01 1.03388727e+00
-5.56629077e-02 -1.14490998e+00 6.24710955e-02 -2.70617683e-03
8.84585619e-01 2.55543441e-01 2.52618253e-01 5.04861712e-01
2.66762495e-01 -9.82682467e-01 7.12159097e-01 2.77010024e-01
5.47754169e-01 -8.86943698e-01 7.11201310e-01 3.98923486e-01
-1.11642241e+00 -6.62117243e-01 -4.73503768e-01 2.87554651e-01
-8.14305618e-02 4.94303524e-01 -5.07395566e-01 7.26377308e-01
9.27999616e-01 1.71929270e-01 -7.50438392e-01 1.04543161e+00
1.86965883e-01 4.14441377e-01 -1.00333609e-01 -4.96493764e-02
2.27914453e-01 -1.96013562e-02 3.88049930e-01 7.56650627e-01
5.97531974e-01 -4.34822857e-01 9.72893387e-02 7.11347461e-01
-1.58816531e-01 1.23747408e-01 -1.10283338e-01 1.76155359e-01
6.35442376e-01 1.19456029e+00 -5.03584445e-01 -1.34988040e-01
-2.03718081e-01 8.79331291e-01 6.23585761e-01 3.69001806e-01
-8.68226767e-01 -3.10339838e-01 6.00485802e-01 -3.68580222e-01
2.06463531e-01 6.27795011e-02 -5.40014684e-01 -6.93120122e-01
9.40929279e-02 -9.73895192e-01 4.06251758e-01 -5.39463758e-01
-1.17847824e+00 1.09563839e+00 -1.57144293e-01 -9.75781798e-01
-2.10863441e-01 -2.99598485e-01 -7.11648047e-01 1.11586761e+00
-1.49582887e+00 -7.39998996e-01 -2.07136974e-01 6.15112722e-01
7.86985338e-01 -8.99351478e-01 7.90726781e-01 5.27645111e-01
-9.48221326e-01 9.22685385e-01 1.91548660e-01 -3.59470665e-01
8.06500435e-01 -8.48818123e-01 2.50135601e-01 4.10884947e-01
1.38170674e-01 4.48898852e-01 3.70016843e-01 -2.97158688e-01
-9.97927904e-01 -1.17168748e+00 6.46821380e-01 3.52316722e-02
3.72665882e-01 -1.93082988e-01 -1.28184068e+00 3.42489928e-01
1.26377530e-02 -2.35241726e-01 8.48958433e-01 2.74961650e-01
-2.34132260e-01 -3.68173748e-01 -8.49216342e-01 3.92119199e-01
9.60546911e-01 -1.42561212e-01 -4.36251163e-01 2.91158229e-01
4.52981472e-01 2.58101765e-02 -8.05293918e-01 4.87325221e-01
4.26171213e-01 -7.56280005e-01 9.69014108e-01 -4.80889797e-01
1.59864247e-01 7.69808330e-03 -2.04017907e-01 -1.88617730e+00
-4.84548658e-01 -2.05366537e-01 9.02778953e-02 1.20434916e+00
6.84072614e-01 -9.76852417e-01 6.69064283e-01 6.04300439e-01
-2.22424328e-01 -1.03107595e+00 -1.20298409e+00 -9.83987689e-01
1.39898375e-01 -2.19637096e-01 5.80620408e-01 9.23339605e-01
-5.94084620e-01 5.92733383e-01 -2.85370201e-01 -1.42949834e-01
5.64255297e-01 -1.60093293e-01 2.78703988e-01 -1.75890589e+00
-6.03743672e-01 -8.86138499e-01 2.02282131e-01 -6.17416978e-01
3.13021272e-01 -9.52954948e-01 -4.86885756e-02 -9.92353499e-01
1.15551323e-01 -1.01870573e+00 -7.16438890e-01 8.32187891e-01
-2.57916957e-01 5.39547093e-02 3.18835199e-01 5.63062072e-01
-4.65817690e-01 8.29305291e-01 1.10762405e+00 -6.76167980e-02
-2.75494754e-01 2.05055773e-01 -7.22954631e-01 5.94828844e-01
9.57797825e-01 -8.53000998e-01 -8.67299139e-01 -9.33138847e-01
3.12878847e-01 -1.98659286e-01 2.58714199e-01 -1.18627763e+00
3.20222557e-01 1.01103872e-01 4.91087288e-01 -3.55516560e-02
4.61354375e-01 -6.89854205e-01 3.89859915e-01 7.49539077e-01
-4.31863904e-01 3.71533632e-01 5.30109465e-01 5.03515005e-01
-1.19871870e-01 -6.06683969e-01 7.37818420e-01 5.57184070e-02
-5.94145358e-01 4.18976927e-03 -3.34628135e-01 -2.18441803e-02
1.04327321e+00 -2.64442861e-01 -2.08805174e-01 1.48304895e-01
-5.72202742e-01 4.02752429e-01 5.05620353e-02 7.18076110e-01
5.72043896e-01 -1.23234951e+00 -8.69623959e-01 1.95904106e-01
-2.11006440e-02 -1.03245206e-01 5.08385122e-01 7.27306366e-01
1.09952077e-01 4.65835899e-01 -3.32922488e-01 -4.75436747e-01
-1.32027483e+00 2.25557029e-01 6.65911317e-01 -6.58028245e-01
-1.46096811e-01 1.03354228e+00 4.70566660e-01 -4.60555315e-01
3.84323329e-01 3.17662030e-01 -2.17602491e-01 2.21463710e-01
2.53103703e-01 4.37140763e-01 4.79274511e-01 -2.34007448e-01
-4.48655248e-01 3.51174682e-01 -3.58271003e-01 4.80514579e-02
1.26145959e+00 -6.76306188e-02 1.64919659e-01 2.25982293e-01
9.68374550e-01 -5.69144487e-01 -1.14482522e+00 -5.02521098e-01
1.63270384e-01 -9.07155499e-02 3.56632799e-01 -1.11258280e+00
-1.37602198e+00 6.68394089e-01 7.86225557e-01 3.07761412e-02
1.38782418e+00 -2.75141150e-01 6.04349136e-01 6.51368558e-01
3.86519343e-01 -1.08487284e+00 4.63455588e-01 5.15262842e-01
9.54504669e-01 -1.07576752e+00 -2.76214421e-01 1.13871694e-01
-8.63044798e-01 9.20862794e-01 9.26452875e-01 3.26405793e-01
3.25390995e-01 5.94178140e-02 -3.61253682e-04 -1.32080182e-01
-1.09304917e+00 3.59321982e-01 3.96075398e-01 2.78758973e-01
2.76248038e-01 -2.04337642e-01 -1.03527114e-01 7.08789468e-01
1.27105460e-01 -1.34233326e-01 -8.66172537e-02 8.08550656e-01
-4.35013801e-01 -1.25248313e+00 -3.31767291e-01 4.38565880e-01
-4.65454876e-01 -1.33481905e-01 -1.42661303e-01 5.52126706e-01
7.13707432e-02 8.50812435e-01 1.38390347e-01 -4.48118865e-01
3.29077899e-01 2.96973348e-01 1.57454103e-01 -2.83453524e-01
-1.08886373e+00 -5.44484630e-02 -5.80768734e-02 -1.91017896e-01
-1.65993810e-01 -3.31191510e-01 -1.09348762e+00 -6.77191392e-02
-4.16793585e-01 3.24983060e-01 7.95521796e-01 8.68519843e-01
7.22346127e-01 9.94829774e-01 6.60641611e-01 -7.23770499e-01
-8.15776110e-01 -8.80288184e-01 -3.98463875e-01 2.75709182e-01
-1.24964491e-01 -1.00410330e+00 -2.07566708e-01 -3.94362688e-01] | [9.27874755859375, 3.370607376098633] |
ba31bb2c-9e1b-4088-90a2-c9a4f68d1fdd | towards-reasoning-aware-explainable-vqa | 2211.05190 | null | https://arxiv.org/abs/2211.05190v1 | https://arxiv.org/pdf/2211.05190v1.pdf | Towards Reasoning-Aware Explainable VQA | The domain of joint vision-language understanding, especially in the context of reasoning in Visual Question Answering (VQA) models, has garnered significant attention in the recent past. While most of the existing VQA models focus on improving the accuracy of VQA, the way models arrive at an answer is oftentimes a black box. As a step towards making the VQA task more explainable and interpretable, our method is built upon the SOTA VQA framework by augmenting it with an end-to-end explanation generation module. In this paper, we investigate two network architectures, including Long Short-Term Memory (LSTM) and Transformer decoder, as the explanation generator. Our method generates human-readable textual explanations while maintaining SOTA VQA accuracy on the GQA-REX (77.49%) and VQA-E (71.48%) datasets. Approximately 65.16% of the generated explanations are approved by humans as valid. Roughly 60.5% of the generated explanations are valid and lead to the correct answers. | ['Govind Thattai', 'Abhinav Mathur', 'Feng Gao', 'Rakesh Vaideeswaran'] | 2022-11-09 | null | null | null | null | ['explanation-generation'] | ['natural-language-processing'] | [ 2.32764423e-01 7.85510898e-01 1.06743105e-01 -8.14563394e-01
-8.51191640e-01 -5.76543391e-01 7.25681782e-01 -2.37060994e-01
9.99448355e-03 6.62973046e-01 5.87477088e-01 -7.97953129e-01
1.85304463e-01 -8.10533285e-01 -9.00941133e-01 -1.16007254e-02
6.21401608e-01 7.60554016e-01 -8.28465819e-02 -2.09129408e-01
2.01272175e-01 -9.17119831e-02 -1.24188530e+00 1.00144899e+00
1.18294764e+00 1.02951419e+00 2.47744638e-02 9.99984384e-01
-6.33585393e-01 1.31872070e+00 -6.45745456e-01 -1.04155123e+00
-5.72655164e-02 -8.37100446e-01 -1.42789567e+00 1.26378596e-01
7.14097321e-01 -5.16967416e-01 -2.65212208e-01 6.48350000e-01
-1.66127663e-02 -1.75519019e-01 7.20330715e-01 -1.50879896e+00
-1.63191652e+00 8.29406500e-01 -2.58745253e-01 6.44352436e-02
4.08603519e-01 5.61655402e-01 1.42005146e+00 -1.16065323e+00
5.96147060e-01 1.59012496e+00 1.67790428e-01 9.39798892e-01
-9.86008942e-01 -3.97471607e-01 4.13394243e-01 6.89106703e-01
-7.18211591e-01 -2.16963097e-01 4.10229087e-01 -3.03512573e-01
1.29005659e+00 2.76268840e-01 7.37501025e-01 1.21454537e+00
1.99659809e-01 1.22193480e+00 8.50167334e-01 -2.56980300e-01
6.02402985e-02 -7.62891844e-02 3.93857986e-01 7.75199771e-01
1.36941418e-01 -9.03837234e-02 -6.50173903e-01 1.90081492e-01
6.70181155e-01 -1.45555303e-01 -2.61282116e-01 -3.01059097e-01
-1.20514572e+00 1.06856382e+00 1.06209731e+00 -2.33370468e-01
-6.51922226e-01 5.06050467e-01 2.01493278e-01 2.03451619e-01
3.29012454e-01 6.55451715e-01 -3.78660321e-01 1.01117097e-01
-4.43041116e-01 5.10859549e-01 4.92272943e-01 9.00717974e-01
5.88943005e-01 3.81250799e-01 -6.81252897e-01 4.07194525e-01
7.84839511e-01 7.45281637e-01 3.62700671e-02 -1.18132424e+00
6.62073851e-01 9.05124903e-01 1.00914232e-01 -6.92397773e-01
-3.44019476e-03 -4.97056037e-01 -6.11502111e-01 2.55145341e-01
5.49890161e-01 1.01213984e-01 -1.32504725e+00 1.63266075e+00
-7.73388427e-03 -2.80075312e-01 4.54217196e-01 1.25920439e+00
1.38329160e+00 9.26992834e-01 4.14682388e-01 2.31912956e-01
1.44736338e+00 -1.50353634e+00 -8.47363353e-01 -7.47182071e-01
1.69164911e-01 -5.92594862e-01 1.47081649e+00 1.96831197e-01
-1.20641804e+00 -7.38836944e-01 -7.27014899e-01 -6.93115532e-01
4.77807745e-02 1.94357941e-03 5.12342215e-01 2.10117653e-01
-1.34322941e+00 -5.07929102e-02 -3.30129981e-01 -1.06757604e-01
6.62199259e-01 1.18458889e-01 -1.89586759e-01 -3.16823155e-01
-1.10658121e+00 9.98857737e-01 -8.95968378e-02 1.68979838e-01
-1.20990288e+00 -5.51247895e-01 -9.64179873e-01 2.98314333e-01
3.70709687e-01 -1.56986916e+00 1.73904407e+00 -1.36879194e+00
-1.14427805e+00 6.26991689e-01 -6.65299475e-01 -6.79760277e-01
3.15831155e-01 -3.78426313e-01 -1.28458291e-01 1.74683437e-01
2.86343664e-01 1.32992101e+00 7.82010198e-01 -1.28814304e+00
-4.11803305e-01 -4.09558326e-01 4.32185501e-01 1.42380998e-01
3.74947101e-01 -3.91876847e-01 -3.23250502e-01 -3.24989200e-01
-2.14911392e-03 -7.80672550e-01 -1.94409356e-01 1.66388229e-01
-4.60034400e-01 -5.75765371e-01 5.06667256e-01 -1.01143098e+00
7.60295212e-01 -1.53120875e+00 2.72751242e-01 -9.08272862e-02
6.60611331e-01 2.40858749e-01 -4.82170790e-01 3.68670076e-01
3.88252214e-02 1.64361492e-01 -2.84493029e-01 -3.75636250e-01
2.35533237e-01 4.09140706e-01 -1.03577673e+00 -2.74999976e-01
6.16835058e-01 1.69786632e+00 -9.23774183e-01 -2.21082658e-01
1.94000140e-01 5.21250248e-01 -6.48575068e-01 4.78282601e-01
-7.71149933e-01 4.30187404e-01 -5.24410486e-01 5.05412519e-01
4.04627025e-01 -7.01259494e-01 -2.32544899e-01 3.21168564e-02
1.61084205e-01 4.28091675e-01 -3.62471968e-01 1.58960390e+00
-4.46005493e-01 1.01883447e+00 -5.60405612e-01 -4.00320888e-01
8.62725675e-01 3.21303189e-01 -3.88904363e-01 -1.10634470e+00
4.74822754e-03 1.03678316e-01 1.39733374e-01 -5.77615321e-01
4.74906951e-01 -3.33057344e-02 3.17573607e-01 5.49271047e-01
7.04631135e-02 -7.35878646e-02 2.14862805e-02 7.60668457e-01
8.79229963e-01 3.11271310e-01 3.12113702e-01 2.25501731e-01
5.48350751e-01 4.22139913e-01 6.81125075e-02 8.80531847e-01
-1.57361522e-01 8.82234037e-01 5.58931112e-01 -7.75427282e-01
-1.13050544e+00 -1.12836790e+00 5.51202416e-01 6.97838724e-01
2.80269291e-02 -4.39347506e-01 -8.34303916e-01 -9.97168064e-01
-4.78150509e-03 1.43412423e+00 -7.96054363e-01 -1.84861839e-01
-2.36867562e-01 2.94748485e-01 3.90067428e-01 8.47151518e-01
7.29534566e-01 -1.57762480e+00 -7.37260461e-01 3.42838764e-02
-6.49159610e-01 -1.07802367e+00 -2.98492819e-01 -4.41936672e-01
-8.15983653e-01 -1.02759969e+00 -6.58825696e-01 -3.41492981e-01
6.36573493e-01 4.60699260e-01 1.85763609e+00 4.31982875e-01
1.74102128e-01 5.81387103e-01 -4.63446647e-01 -7.22443402e-01
-5.33756614e-01 -2.02272609e-01 -5.85256338e-01 -1.28318593e-01
5.44479311e-01 2.27674749e-02 -7.02612996e-01 -3.95880416e-02
-7.20748901e-01 7.06558168e-01 6.01952314e-01 7.42794752e-01
6.32638276e-01 -1.06367540e+00 6.84598982e-01 -8.11615467e-01
5.90222538e-01 -3.62981826e-01 -3.05274695e-01 4.37467068e-01
-7.62081027e-01 3.96975189e-01 5.72426975e-01 2.61831805e-02
-1.26104426e+00 -1.04545578e-01 -3.79753530e-01 -3.37248266e-01
-7.91085660e-02 3.63277227e-01 -1.68962449e-01 3.65471065e-01
7.61614501e-01 1.51344568e-01 7.58177554e-03 2.04976406e-02
1.02040279e+00 4.26956296e-01 6.59547746e-01 -1.78607821e-01
7.74900556e-01 3.29828531e-01 -4.13033873e-01 -2.47267023e-01
-1.29838157e+00 8.63914102e-05 -1.47682458e-01 -3.82123739e-01
1.23740947e+00 -8.20919156e-01 -7.73594797e-01 -1.50468107e-02
-1.71476972e+00 -3.82189512e-01 -3.25215071e-01 8.15364346e-02
-5.26813865e-01 1.59161955e-01 -2.37297446e-01 -8.57033134e-01
-8.02397251e-01 -1.24495876e+00 1.09863043e+00 4.44133401e-01
-5.33026993e-01 -7.46580124e-01 -1.61456674e-01 1.00962961e+00
4.66854721e-01 -5.25783449e-02 1.14574623e+00 -3.77617002e-01
-1.06951845e+00 2.75845557e-01 -7.06536174e-01 3.88295725e-02
-2.61415511e-01 -1.08182006e-01 -1.15827131e+00 2.04062045e-01
-2.82751232e-01 -4.99549091e-01 1.11525309e+00 5.43598413e-01
1.23173547e+00 -4.49063927e-01 1.01698056e-01 2.76253670e-01
1.02883887e+00 1.20650880e-01 7.78362811e-01 1.79696634e-01
7.94174552e-01 6.37331069e-01 6.01869106e-01 9.46892425e-03
1.06408799e+00 4.14750695e-01 1.13017416e+00 -3.08476895e-01
-6.12325788e-01 -5.82294226e-01 3.08344394e-01 3.57628375e-01
1.34181634e-01 -7.06718206e-01 -1.08943284e+00 8.64517748e-01
-2.04736066e+00 -9.30300415e-01 -6.97732806e-01 1.51659644e+00
5.42604029e-01 -1.98109880e-01 -2.40943328e-01 -2.63099909e-01
2.09085301e-01 9.67537835e-02 -7.75343597e-01 -7.95159400e-01
-8.36554021e-02 1.17691942e-01 -1.93744585e-01 8.28687966e-01
-4.96191531e-01 1.11104095e+00 5.95239687e+00 6.05072714e-02
-6.76050127e-01 -9.89664495e-02 8.23723078e-01 1.61385819e-01
-1.04657042e+00 2.36744225e-01 -2.87710100e-01 -1.60648763e-01
8.77675235e-01 5.98295704e-02 3.59632343e-01 8.67851377e-01
2.00011358e-01 2.14093495e-02 -1.09101081e+00 9.39509928e-01
2.74918020e-01 -1.60482478e+00 8.90748084e-01 -2.53866315e-01
6.06748879e-01 -1.15308665e-01 2.95549154e-01 5.89019239e-01
3.56705993e-01 -1.51408184e+00 1.08665955e+00 6.38114810e-01
6.87677443e-01 -5.87693334e-01 8.25461090e-01 1.99348629e-01
-9.47617769e-01 -1.07933477e-01 -5.42961776e-01 -4.48124558e-01
5.40710032e-01 3.40385705e-01 -1.10144711e+00 3.70659441e-01
6.12943232e-01 6.54339910e-01 -9.21420634e-01 7.57743001e-01
-1.03504479e+00 8.39857757e-01 3.92563611e-01 -1.73321500e-01
5.08815527e-01 -3.32586304e-03 3.57374549e-01 6.51778102e-01
1.65508017e-01 3.94148380e-01 -3.45547795e-01 1.37130916e+00
-2.85080433e-01 -1.58570364e-01 -5.51185071e-01 -1.54461488e-01
4.25622948e-02 1.03340316e+00 -1.97950497e-01 -5.86483538e-01
-4.89330500e-01 1.10576034e+00 5.70428610e-01 6.16903007e-01
-9.91644382e-01 3.23188566e-02 6.05408549e-01 -2.18348011e-01
2.33096600e-01 1.25400469e-01 -6.78108335e-01 -1.14150703e+00
1.58390794e-02 -1.18722868e+00 4.13466334e-01 -1.80897474e+00
-1.27007437e+00 9.58697855e-01 -3.53527904e-01 -7.38048971e-01
-6.49840176e-01 -7.19486535e-01 -5.67057133e-01 1.02494395e+00
-1.64841521e+00 -1.52467799e+00 -7.41243243e-01 4.52372789e-01
1.04759514e+00 -1.27740741e-01 8.40385497e-01 -2.31956258e-01
-4.40406986e-02 3.05159986e-01 -7.36961067e-01 -9.96686965e-02
4.52739924e-01 -1.53173077e+00 1.19209945e+00 9.96583939e-01
7.64668941e-01 6.69814348e-01 1.00240946e+00 -6.14807129e-01
-1.27258205e+00 -1.08304238e+00 1.29453933e+00 -1.05323923e+00
2.22171366e-01 -1.56919762e-01 -1.21336818e+00 1.00181174e+00
8.23247969e-01 -2.88922369e-01 4.02347565e-01 5.48976660e-02
-6.92184091e-01 1.22352727e-01 -7.68717289e-01 8.31987739e-01
1.00861955e+00 -7.43897498e-01 -8.85774672e-01 2.22046182e-01
1.00796604e+00 -3.21653306e-01 -9.80708525e-02 1.55573592e-01
4.08032745e-01 -9.88757551e-01 9.21493649e-01 -9.81454313e-01
1.01590586e+00 -5.94915390e-01 1.17722847e-01 -1.30721056e+00
-3.28837216e-01 -2.38876879e-01 -2.78830260e-01 9.38485086e-01
7.97895014e-01 -2.40008757e-01 5.82508743e-01 8.01805437e-01
-3.74403685e-01 -5.87735474e-01 -7.51766622e-01 -1.41092120e-02
-1.30497679e-01 -5.70599258e-01 7.78437436e-01 4.36496675e-01
-3.95386428e-01 8.08926463e-01 -3.19531113e-01 1.24084748e-01
5.20846844e-01 3.91839653e-01 1.09339035e+00 -9.08844829e-01
-2.43182078e-01 -2.24778995e-01 9.82581154e-02 -1.33587337e+00
2.20899105e-01 -9.35256839e-01 1.03516534e-01 -2.58148193e+00
3.96716535e-01 3.71957004e-01 8.04017410e-02 7.23981559e-01
-5.47200441e-01 2.13544458e-01 3.26432914e-01 -2.50326749e-02
-7.65520632e-01 7.73166001e-01 1.52249801e+00 -3.19044620e-01
2.51499057e-01 -1.69836074e-01 -1.07705319e+00 5.70439279e-01
6.00555301e-01 -8.90659764e-02 -8.32240999e-01 -1.25040007e+00
5.34614027e-01 2.22804010e-01 8.39751124e-01 -5.10015428e-01
-9.36965570e-02 -1.12631515e-01 5.22379816e-01 -6.83603942e-01
2.35994831e-01 -5.17593920e-01 5.29273078e-02 3.51617932e-01
-6.34311438e-01 4.57170844e-01 2.27614135e-01 5.36890924e-01
-3.06306392e-01 1.28110349e-01 2.97715813e-01 -1.36745244e-01
-9.21184659e-01 2.76669949e-01 -2.80443460e-01 4.73871566e-02
6.43214166e-01 -1.90670133e-01 -8.76665235e-01 -1.06243777e+00
-4.31757390e-01 6.46098733e-01 2.06198514e-01 6.22469485e-01
1.19629610e+00 -1.27361250e+00 -1.08278716e+00 -2.20119968e-01
4.92180496e-01 -7.58494660e-02 4.92180884e-01 3.86297852e-01
-3.42437238e-01 8.50242615e-01 -2.07941189e-01 -4.81601000e-01
-1.23718214e+00 4.26624358e-01 3.84312212e-01 -2.78722979e-02
-6.53146207e-01 1.11427248e+00 5.50620317e-01 -2.78657109e-01
-9.48990360e-02 -4.85317737e-01 -5.68453729e-01 -3.62355113e-01
6.89700961e-01 -2.88482732e-03 -4.67475533e-01 -5.20317554e-01
-2.83697963e-01 3.42447460e-01 4.90187816e-02 -2.92054981e-01
1.14432478e+00 -2.12015823e-01 -1.85469016e-02 2.50255048e-01
3.86566371e-01 -2.29708210e-01 -1.33661878e+00 -9.36690867e-02
-1.45755604e-01 -4.31783795e-01 -2.59749949e-01 -1.33519995e+00
-9.88940895e-01 1.49308383e+00 2.09357455e-01 7.80594349e-02
8.94489408e-01 4.21715558e-01 8.40405345e-01 3.73447239e-01
3.62908468e-02 -2.89995790e-01 4.45124954e-01 6.61753476e-01
1.42560184e+00 -1.43987882e+00 -3.93693954e-01 -2.93036431e-01
-1.25480902e+00 1.02676952e+00 1.04473507e+00 8.86561349e-02
-4.60132986e-01 -7.31769800e-01 5.30962288e-01 -4.57272559e-01
-1.24017310e+00 -1.77055717e-01 8.21931124e-01 6.85981691e-01
5.37113249e-01 3.58997434e-02 4.00647148e-02 8.78471255e-01
-3.42843294e-01 -1.34977907e-01 6.42757535e-01 4.94435169e-02
-4.49627131e-01 -7.13882744e-01 -6.17114492e-02 3.15522581e-01
-1.71249248e-02 -3.18659425e-01 -8.99038255e-01 5.84450185e-01
-2.57791787e-01 1.24862194e+00 2.18663383e-02 -8.18589404e-02
3.24124545e-01 2.40911692e-01 2.32577503e-01 -5.44256866e-01
-7.01032639e-01 -6.57908261e-01 3.01249892e-01 -7.47138858e-01
-1.89823538e-01 -1.04970470e-01 -1.69015229e+00 -2.53229201e-01
-3.41106579e-02 4.90266755e-02 3.84169966e-01 1.29825222e+00
5.49564600e-01 8.32613766e-01 -1.59354135e-01 -1.10312566e-01
-3.89776796e-01 -8.07934225e-01 1.92378059e-01 5.52696705e-01
4.58755463e-01 -2.81079441e-01 -6.20850548e-02 1.14656918e-01] | [10.785514831542969, 1.8441369533538818] |
fe8540ef-0b69-4549-93fe-028626032590 | audio-visual-contrastive-learning-with | 2302.07702 | null | https://arxiv.org/abs/2302.07702v1 | https://arxiv.org/pdf/2302.07702v1.pdf | Audio-Visual Contrastive Learning with Temporal Self-Supervision | We propose a self-supervised learning approach for videos that learns representations of both the RGB frames and the accompanying audio without human supervision. In contrast to images that capture the static scene appearance, videos also contain sound and temporal scene dynamics. To leverage the temporal and aural dimension inherent to videos, our method extends temporal self-supervision to the audio-visual setting and integrates it with multi-modal contrastive objectives. As temporal self-supervision, we pose playback speed and direction recognition in both modalities and propose intra- and inter-modal temporal ordering tasks. Furthermore, we design a novel contrastive objective in which the usual pairs are supplemented with additional sample-dependent positives and negatives sampled from the evolving feature space. In our model, we apply such losses among video clips and between videos and their temporally corresponding audio clips. We verify our model design in extensive ablation experiments and evaluate the video and audio representations in transfer experiments to action recognition and retrieval on UCF101 and HMBD51, audio classification on ESC50, and robust video fingerprinting on VGG-Sound, with state-of-the-art results. | ['John Collomosse', 'Alexander Black', 'Simon Jenni'] | 2023-02-15 | null | null | null | null | ['audio-classification'] | ['audio'] | [ 4.57067907e-01 -3.48857194e-01 -2.81360596e-01 -4.34004515e-01
-1.02451360e+00 -7.59741783e-01 6.73155129e-01 -3.83614510e-01
-4.67660576e-01 3.57169718e-01 3.77758086e-01 4.85011637e-01
-1.40224323e-01 -1.99850500e-01 -1.05089462e+00 -6.61082327e-01
-7.15959251e-01 -1.17810957e-01 2.49721527e-01 1.42336205e-01
-3.20870802e-02 1.19100310e-01 -2.05183005e+00 7.92361796e-01
1.83027565e-01 1.65333223e+00 -4.43061292e-02 1.13391089e+00
5.12356520e-01 1.22925973e+00 -2.76442200e-01 -1.28790185e-01
4.66062099e-01 -3.56498659e-01 -7.91926622e-01 3.24648291e-01
1.02139699e+00 -5.77448845e-01 -8.82866561e-01 6.93446517e-01
5.36235273e-01 2.60999024e-01 4.28698689e-01 -1.62113059e+00
-5.78678548e-01 4.81459916e-01 -5.50386786e-01 3.65585923e-01
7.55559087e-01 2.54184395e-01 1.10659480e+00 -5.87231040e-01
7.81660855e-01 1.12732613e+00 6.49149597e-01 5.66814244e-01
-1.22620630e+00 -7.39242911e-01 4.32683885e-01 6.19264722e-01
-1.14473546e+00 -8.47265482e-01 7.83816874e-01 -4.58877414e-01
8.34673107e-01 2.94367075e-01 8.56598437e-01 1.75173867e+00
-1.67506427e-01 1.07415974e+00 7.83964694e-01 -2.31666565e-01
1.13433540e-01 -1.09703414e-01 -3.30012918e-01 5.70003092e-01
-7.36939549e-01 4.56341684e-01 -1.40743947e+00 1.77406706e-02
6.32596016e-01 6.52134567e-02 -4.62205678e-01 -6.13788366e-01
-1.31494403e+00 4.67558265e-01 1.41167834e-01 9.36128199e-02
-1.35489702e-01 6.28896534e-01 5.55678546e-01 6.92111075e-01
2.22213998e-01 1.01253860e-01 -4.25969630e-01 -5.51738679e-01
-1.09485972e+00 -2.88802781e-03 4.50955629e-01 9.46926832e-01
4.36678827e-01 1.99305847e-01 -4.01577771e-01 5.20113468e-01
5.46653569e-02 5.26132286e-01 7.41617382e-01 -1.47764218e+00
3.08590233e-01 -6.51399195e-02 -1.28624449e-02 -8.97711754e-01
-2.08305895e-01 -1.79073587e-01 -3.98727894e-01 -1.66628152e-01
3.31420958e-01 2.59729236e-01 -9.04104888e-01 2.14207530e+00
1.61524013e-01 7.70455480e-01 -2.23086685e-01 9.52943921e-01
5.35212040e-01 4.06697452e-01 -4.64902967e-02 -3.72116297e-01
1.11309493e+00 -9.78885114e-01 -6.23379529e-01 8.72412696e-02
3.58245164e-01 -4.84504968e-01 1.10135782e+00 6.23419166e-01
-1.08137894e+00 -6.54163301e-01 -9.40104544e-01 -1.17482133e-01
-7.03071058e-02 2.11434796e-01 4.10659969e-01 2.40237713e-01
-1.13736367e+00 8.20939720e-01 -1.16111386e+00 -2.82408506e-01
3.39693457e-01 3.21196765e-01 -5.93546867e-01 1.33410081e-01
-1.28234649e+00 1.65206224e-01 -5.45078367e-02 -1.22778364e-01
-1.62185705e+00 -8.21885645e-01 -8.66712213e-01 -2.69556880e-01
3.96576405e-01 -3.94864827e-01 1.27945137e+00 -1.45317006e+00
-1.52669024e+00 9.86080587e-01 1.60568357e-02 -6.40910387e-01
7.50785887e-01 -3.62230092e-01 -6.69484258e-01 9.05514061e-01
1.07259259e-01 8.75167251e-01 1.57309055e+00 -9.07276869e-01
-6.55126274e-01 -2.03156382e-01 2.90241897e-01 1.47551104e-01
-7.13523507e-01 -1.26455724e-01 -6.15906179e-01 -7.87967503e-01
-1.83878645e-01 -9.86268699e-01 4.53313261e-01 4.18989122e-01
-1.93042099e-01 3.33837420e-01 1.15050209e+00 -6.08314097e-01
1.01439583e+00 -2.53375816e+00 4.32488650e-01 3.02029657e-03
1.41706842e-03 -3.50831896e-01 -4.58295733e-01 1.70602694e-01
-2.54326671e-01 -3.67343992e-01 3.27751040e-02 -5.37307084e-01
7.61758760e-02 1.23922020e-01 -5.63064754e-01 7.59537876e-01
2.27419838e-01 6.86672866e-01 -1.14368284e+00 -6.51814699e-01
1.04136661e-01 6.94497287e-01 -7.89793730e-01 2.91906893e-01
-1.86896935e-01 5.17132342e-01 -2.41906807e-01 9.00622964e-01
1.01703100e-01 -3.19426417e-01 1.93261057e-02 -4.30733770e-01
1.33891046e-01 4.13555317e-02 -1.05183542e+00 2.34937930e+00
-4.14728671e-01 8.91118407e-01 1.82647198e-01 -9.70352948e-01
1.55767858e-01 4.84988272e-01 7.80279756e-01 -8.93007457e-01
-4.75992262e-02 -1.28403738e-01 -4.61412698e-01 -8.10452044e-01
5.35953403e-01 7.04743713e-02 5.68165863e-03 3.86593282e-01
6.86776519e-01 6.95139617e-02 1.34502545e-01 3.28394562e-01
1.38432133e+00 5.87154567e-01 -3.04201126e-01 2.15763509e-01
1.52010500e-01 -5.44656157e-01 3.43764782e-01 5.92914343e-01
-4.33698863e-01 8.45999777e-01 4.34715718e-01 2.92327274e-02
-7.00668395e-01 -1.09687328e+00 -6.72167465e-02 1.50450397e+00
4.08981815e-02 -5.95327914e-01 -5.14722884e-01 -6.88896060e-01
1.35187328e-01 2.47723222e-01 -1.09442818e+00 -3.90240163e-01
-2.39799738e-01 -1.19817145e-01 6.42092288e-01 5.80891848e-01
2.96606511e-01 -7.52971649e-01 -8.43662679e-01 2.51673218e-02
-2.58325309e-01 -1.37870550e+00 -8.16217780e-01 1.61654085e-01
-5.57045519e-01 -1.15461946e+00 -7.60561526e-01 -5.03753424e-01
1.29093174e-02 1.46087915e-01 9.18420136e-01 -4.57860857e-01
-6.71894372e-01 1.35191870e+00 -4.71662462e-01 2.31774867e-01
4.44336236e-03 -3.06776822e-01 3.85652065e-01 5.26119053e-01
-2.91122288e-01 -9.67306733e-01 -8.31459820e-01 4.31078821e-01
-1.04072797e+00 -2.61260003e-01 1.00768127e-01 9.19304192e-01
6.38863504e-01 -2.51663297e-01 2.42451921e-01 -2.77790785e-01
-2.03906253e-01 -4.18104827e-01 -2.95457631e-01 2.81747431e-01
-9.27267149e-02 -1.25923559e-01 4.10335243e-01 -8.13056290e-01
-7.81542361e-01 2.77175039e-01 3.94032240e-01 -1.35478377e+00
7.86291435e-02 6.07333742e-02 -1.29254147e-01 -1.56568989e-01
6.87926412e-01 2.42511794e-01 6.47063786e-03 -1.82090685e-01
5.06890237e-01 3.69535089e-01 9.36562359e-01 -5.69630563e-01
6.48232341e-01 8.97877216e-01 -1.49957359e-01 -8.62534940e-01
-7.77741790e-01 -3.58348370e-01 -4.53612149e-01 -6.18233323e-01
8.52731645e-01 -1.29083753e+00 -8.46876204e-01 5.05208790e-01
-7.44390309e-01 -5.99460423e-01 -6.01545811e-01 6.45373940e-01
-1.13126755e+00 4.49448705e-01 -8.00744057e-01 -8.04748237e-01
1.93856284e-01 -1.01542521e+00 1.53654802e+00 -2.51275122e-01
-1.93863019e-01 -6.12804711e-01 8.68100747e-02 1.73915237e-01
1.73899874e-01 2.57306069e-01 2.40088016e-01 -2.61488944e-01
-7.33200431e-01 -1.21146649e-01 8.48807693e-02 5.43155789e-01
3.37112360e-02 7.76386857e-02 -1.44607472e+00 -4.56975490e-01
-1.31881431e-01 -9.57998753e-01 1.20805991e+00 3.97024393e-01
1.43688750e+00 -3.18021685e-01 -1.18611872e-01 1.02577162e+00
1.07911992e+00 -3.45877782e-02 3.02244604e-01 3.15324634e-01
6.58245802e-01 6.90266311e-01 7.22094953e-01 6.42132103e-01
1.41091287e-01 9.88458753e-01 5.61682701e-01 2.28654891e-01
-1.21800207e-01 -4.04311687e-01 9.41819370e-01 4.76547241e-01
-3.57746109e-02 -1.41046509e-01 -4.33139443e-01 4.93589580e-01
-1.85160899e+00 -1.38347602e+00 6.95573092e-01 2.14364576e+00
8.42337251e-01 -1.03329837e-01 4.78579909e-01 2.82915294e-01
4.44552600e-01 3.76292050e-01 -7.00294733e-01 4.08529252e-01
-4.19416696e-01 7.99873173e-02 3.13068837e-01 1.77357942e-01
-1.49938440e+00 5.50327599e-01 6.19717026e+00 7.19054878e-01
-1.40924227e+00 1.96895376e-01 4.96669143e-01 -9.68448877e-01
-2.60046870e-01 -2.66756028e-01 -1.77428350e-01 4.73535210e-01
1.00568569e+00 1.24963216e-01 7.75695920e-01 5.74343801e-01
1.00692809e-02 1.24037795e-01 -1.74073839e+00 1.26454294e+00
2.26922110e-01 -1.13902569e+00 -1.81307673e-01 -6.47414401e-02
5.79218268e-01 -3.63876522e-02 5.52128732e-01 1.73078507e-01
-1.67439468e-02 -9.06544030e-01 1.31177032e+00 5.34210026e-01
1.30383098e+00 -3.42958689e-01 -1.14085495e-01 -3.42410922e-01
-1.20741093e+00 -4.82074291e-01 2.58298457e-01 1.51119441e-01
1.43210143e-01 1.43471789e-02 -8.95648450e-02 4.25277799e-01
1.29217958e+00 1.48179209e+00 -5.08339047e-01 7.15014219e-01
7.65742138e-02 6.33378625e-01 -3.87847453e-01 6.15016937e-01
6.66149557e-02 1.93283319e-01 7.24690616e-01 1.18048203e+00
3.39827687e-01 -1.49297357e-01 -2.96012908e-02 4.13806081e-01
4.59821522e-03 -2.81202585e-01 -4.87866133e-01 -3.43253851e-01
2.97108144e-01 8.59476089e-01 -3.76776248e-01 -1.64509848e-01
-3.46246094e-01 1.29547715e+00 4.75050695e-02 6.98290586e-01
-1.02438319e+00 -8.09790120e-02 8.54372919e-01 1.66711599e-01
6.64430320e-01 7.15714023e-02 6.20589435e-01 -1.42825043e+00
2.65563697e-01 -9.94651735e-01 6.49940073e-01 -8.21918726e-01
-1.14914846e+00 5.95885277e-01 8.51666927e-02 -1.80496693e+00
-6.38543904e-01 -4.44642395e-01 -2.17403114e-01 3.98745015e-03
-1.37530744e+00 -1.17286229e+00 -2.65389889e-01 1.07463777e+00
3.13916385e-01 -2.55341023e-01 6.99350297e-01 6.40273154e-01
-3.17553908e-01 1.01368773e+00 6.88722506e-02 8.56027976e-02
1.12387156e+00 -1.00187969e+00 -1.45454168e-01 6.00854158e-01
4.43616092e-01 1.98690116e-01 7.25626826e-01 -1.40828520e-01
-1.69068849e+00 -9.39555526e-01 -1.49108078e-02 -3.59107792e-01
1.07968402e+00 -5.64111352e-01 -6.04926348e-01 7.54805386e-01
3.43499891e-02 5.38321257e-01 5.81163347e-01 -7.46987760e-02
-8.28820646e-01 -4.26715016e-01 -8.11714292e-01 1.88994184e-01
1.39097786e+00 -1.26528668e+00 -1.98101878e-01 3.37913543e-01
8.93673837e-01 -4.76830304e-01 -7.61717379e-01 2.92882770e-01
1.04376805e+00 -1.03447533e+00 1.11432469e+00 -8.06011021e-01
6.56032324e-01 -2.70412624e-01 -6.38531506e-01 -1.01457262e+00
2.66164225e-02 -8.39965940e-01 -4.70494866e-01 1.25913894e+00
2.39059720e-02 1.14728976e-02 8.11230898e-01 1.39486417e-01
-5.70833404e-03 -3.66259992e-01 -1.26069629e+00 -1.00671339e+00
-5.05709946e-01 -7.62271821e-01 1.93624914e-01 9.63978291e-01
9.45951343e-02 -1.97716895e-02 -8.17869604e-01 1.04541250e-01
7.54812717e-01 1.96618930e-01 4.55070585e-01 -6.93849504e-01
-9.15441632e-01 -1.79546505e-01 -7.91585147e-01 -1.03213811e+00
3.52897912e-01 -5.67561269e-01 4.26454172e-02 -5.03876567e-01
1.81548700e-01 1.18504740e-01 -6.66154385e-01 3.78894866e-01
4.09212589e-01 4.28924412e-01 3.00431520e-01 2.29637384e-01
-1.20992279e+00 9.05212879e-01 8.15405130e-01 -4.20962244e-01
-3.49568836e-02 -2.73136973e-01 -1.25082821e-01 3.96428019e-01
1.04689807e-01 -3.03924978e-01 -6.32840574e-01 -5.43833256e-01
1.38074681e-01 4.55017149e-01 8.07586253e-01 -1.03233516e+00
1.73094094e-01 7.22184926e-02 4.46738482e-01 -2.55013317e-01
8.14919889e-01 -8.31194520e-01 1.02984227e-01 6.64682537e-02
-8.81273746e-01 -9.82288346e-02 1.58242017e-01 1.15566421e+00
-5.24402678e-01 5.24586916e-01 5.59123933e-01 1.25541180e-01
-7.77016878e-01 5.68376064e-01 -1.11448817e-01 1.95744097e-01
8.48017871e-01 -2.34958425e-01 -1.93951219e-01 -7.87655413e-01
-1.14942706e+00 3.50322843e-01 5.66577792e-01 6.08775198e-01
6.03108525e-01 -1.58793890e+00 -3.28220993e-01 1.43928677e-01
4.40091610e-01 -4.72675025e-01 6.83068037e-01 1.07089555e+00
1.75800491e-02 1.39188409e-01 -4.39150333e-01 -9.88940418e-01
-1.17008460e+00 5.70890546e-01 3.47250670e-01 2.10742056e-01
-4.20750588e-01 1.11994588e+00 4.59653258e-01 1.30236104e-01
9.06665981e-01 -2.10350171e-01 1.92550942e-02 3.70173961e-01
6.04845405e-01 2.42384300e-01 -6.34650365e-02 -6.90327823e-01
-3.98574978e-01 4.02010709e-01 6.41768873e-02 -5.16280949e-01
1.25577605e+00 -2.65930563e-01 3.95311862e-01 9.06312943e-01
1.70997000e+00 -2.10854843e-01 -1.89497042e+00 -3.05058837e-01
-3.17330927e-01 -5.54777801e-01 1.56069607e-01 -5.79845786e-01
-1.36016631e+00 8.99219453e-01 1.00601017e+00 -1.12146288e-02
1.45276046e+00 2.03514434e-02 6.17500067e-01 1.75164014e-01
2.87911028e-01 -1.29097712e+00 6.68016493e-01 1.54648721e-01
9.50969636e-01 -1.21601593e+00 -2.25190356e-01 -2.39178017e-02
-5.61722577e-01 9.06373560e-01 4.17293280e-01 2.88724657e-02
6.94441795e-01 2.78438717e-01 -1.18276179e-01 9.61448345e-03
-1.13620710e+00 -6.08048029e-02 2.84852892e-01 5.46007037e-01
2.12330192e-01 -1.82612807e-01 4.99276906e-01 5.64710736e-01
-9.09051448e-02 2.15814576e-01 1.92439258e-01 1.05323124e+00
2.36286953e-01 -7.14840829e-01 -9.79339108e-02 3.06515098e-01
-4.09962773e-01 2.91342027e-02 -2.79047042e-01 6.47357464e-01
-7.45918006e-02 7.66120851e-01 3.02365929e-01 -9.11061347e-01
2.30889708e-01 -5.84447645e-02 8.16201150e-01 -2.04749152e-01
-3.14133346e-01 2.69093990e-01 -2.98829400e-03 -1.30357325e+00
-7.55068660e-01 -1.10400271e+00 -9.33543921e-01 1.56945825e-01
-5.25592305e-02 1.22517031e-02 3.26601148e-01 6.66164279e-01
3.63853425e-01 5.17707825e-01 9.16081667e-01 -1.18521404e+00
-5.66956341e-01 -7.38876760e-01 -8.78924012e-01 6.04434550e-01
1.01997435e+00 -8.55201304e-01 -6.51558816e-01 4.58774477e-01] | [9.48346996307373, 1.0403611660003662] |
0597a1a1-ebd7-4aa7-91bf-032fe35f27cf | container-few-shot-named-entity-recognition | 2109.07589 | null | https://arxiv.org/abs/2109.07589v2 | https://arxiv.org/pdf/2109.07589v2.pdf | CONTaiNER: Few-Shot Named Entity Recognition via Contrastive Learning | Named Entity Recognition (NER) in Few-Shot setting is imperative for entity tagging in low resource domains. Existing approaches only learn class-specific semantic features and intermediate representations from source domains. This affects generalizability to unseen target domains, resulting in suboptimal performances. To this end, we present CONTaiNER, a novel contrastive learning technique that optimizes the inter-token distribution distance for Few-Shot NER. Instead of optimizing class-specific attributes, CONTaiNER optimizes a generalized objective of differentiating between token categories based on their Gaussian-distributed embeddings. This effectively alleviates overfitting issues originating from training domains. Our experiments in several traditional test domains (OntoNotes, CoNLL'03, WNUT '17, GUM) and a new large scale Few-Shot NER dataset (Few-NERD) demonstrate that on average, CONTaiNER outperforms previous methods by 3%-13% absolute F1 points while showing consistent performance trends, even in challenging scenarios where previous approaches could not achieve appreciable performance. | ['Rui Zhang', 'Rebecca J. Passonneau', 'Arzoo Katiyar', 'Sarkar Snigdha Sarathi Das'] | 2021-09-15 | null | https://aclanthology.org/2022.acl-long.439 | https://aclanthology.org/2022.acl-long.439.pdf | acl-2022-5 | ['few-shot-ner'] | ['natural-language-processing'] | [-3.10104400e-01 -9.93754566e-02 -7.62102827e-02 -4.62268412e-01
-1.25478160e+00 -8.71186018e-01 5.37554026e-01 4.04175043e-01
-9.48710680e-01 8.86204898e-01 3.20020288e-01 1.72027498e-02
7.56166577e-02 -6.05811775e-01 -1.74479023e-01 -3.50045085e-01
-2.06745982e-01 4.62930709e-01 2.53431886e-01 3.94574180e-03
6.34498149e-02 2.75026709e-01 -1.20767689e+00 -3.23111750e-02
7.84658194e-01 6.44766986e-01 5.25215827e-02 6.02016091e-01
-5.30633569e-01 5.09080231e-01 -7.38398671e-01 -6.00378811e-01
1.36126727e-01 -4.50895019e-02 -9.09681797e-01 -3.69771510e-01
3.38967890e-01 -6.40590638e-02 -1.19658619e-01 1.00041234e+00
9.64690566e-01 5.17535567e-01 8.42357635e-01 -1.05304170e+00
-1.07644892e+00 4.11190271e-01 -1.42593339e-01 4.73548800e-01
6.47638068e-02 -2.40123078e-01 1.28730166e+00 -9.69058394e-01
7.48904467e-01 9.14110422e-01 1.12116933e+00 7.21057296e-01
-1.16017377e+00 -5.37880540e-01 -7.13341385e-02 -1.87931404e-01
-1.53692448e+00 -4.61956263e-01 1.27143323e-01 -2.18979076e-01
1.26703703e+00 -1.63323600e-02 -2.10260659e-01 1.08446324e+00
-2.39647791e-01 6.82399154e-01 9.42778528e-01 -4.16131884e-01
5.44696987e-01 3.20925057e-01 5.69874287e-01 3.07159632e-01
3.66290689e-01 -2.07236573e-01 -3.25091094e-01 -4.57394242e-01
2.95965970e-01 -4.07711156e-02 3.90315661e-04 -1.77388176e-01
-1.01012671e+00 8.89728904e-01 1.32195249e-01 6.89766228e-01
-3.44059139e-01 -2.49674305e-01 5.08933187e-01 2.73771882e-01
7.48569548e-01 7.54279196e-01 -1.06720114e+00 -5.01918793e-01
-7.47906625e-01 -6.56832457e-02 1.03562856e+00 1.10494184e+00
7.21625805e-01 1.64597169e-01 -3.18316221e-01 1.22142994e+00
1.20497704e-03 3.82718384e-01 7.12201059e-01 -5.30302227e-01
5.12336910e-01 3.25754315e-01 3.99891585e-01 -4.56133991e-01
-4.39366996e-01 -2.47434318e-01 -3.45229417e-01 -2.41880670e-01
4.44259733e-01 -6.34872854e-01 -1.09450471e+00 1.78047729e+00
3.92822921e-01 4.74937975e-01 4.66858178e-01 6.80404961e-01
9.69581425e-01 4.99274552e-01 8.66043806e-01 1.35554969e-01
1.59724331e+00 -7.62630939e-01 -5.58319807e-01 -2.77798891e-01
1.08008265e+00 -6.81288064e-01 1.09349656e+00 -2.32745126e-01
-5.15815377e-01 -3.38209391e-01 -7.41094053e-01 -1.72889978e-01
-9.65171814e-01 1.94818109e-01 6.22977734e-01 8.28944623e-01
-7.55413949e-01 7.12753117e-01 -5.64952970e-01 -8.36331129e-01
3.72804523e-01 2.58822531e-01 -6.00052953e-01 -1.06147416e-01
-1.43987978e+00 9.24317181e-01 6.07758224e-01 -5.13706148e-01
-4.95626986e-01 -1.12944460e+00 -1.00821376e+00 3.02990437e-01
2.60739569e-02 -2.67195314e-01 1.31142426e+00 -3.84004682e-01
-1.25190949e+00 1.00749195e+00 8.59803706e-02 -4.97662991e-01
6.85286075e-02 -4.98277426e-01 -9.33877409e-01 -1.59458354e-01
4.33719337e-01 6.55175090e-01 2.11710930e-01 -9.68028486e-01
-5.81547916e-01 -3.24013121e-02 -5.94989536e-03 1.10806793e-01
-8.40519130e-01 2.45126843e-01 -6.49736151e-02 -6.44047856e-01
-3.82897347e-01 -6.38411283e-01 -2.51291871e-01 -4.22662497e-01
-2.44633898e-01 -4.56080556e-01 6.16470933e-01 -4.45526481e-01
1.17473137e+00 -2.30151010e+00 -5.19552767e-01 -3.03987712e-01
-3.50043066e-02 6.54931426e-01 -4.58881408e-01 4.91706997e-01
-1.23032913e-01 3.55022609e-01 1.44121554e-02 -3.25911403e-01
3.35088283e-01 1.75646245e-01 -9.68168825e-02 3.50098431e-01
5.24830222e-01 7.71963179e-01 -1.11573172e+00 -5.17579675e-01
1.29238114e-01 5.19893706e-01 -4.23969358e-01 3.53057086e-01
2.28041574e-01 -1.60850644e-01 -4.34423029e-01 7.74062812e-01
6.24701738e-01 -2.23814428e-01 1.38253912e-01 -4.03194316e-02
-1.63417324e-01 3.21904063e-01 -1.16964626e+00 1.69359255e+00
-6.24960482e-01 3.61469001e-01 -2.69987136e-01 -8.46479416e-01
1.04097188e+00 5.44563830e-01 3.75226408e-01 -6.57592595e-01
1.14760287e-01 1.95769697e-01 -3.97111118e-01 -3.06818664e-01
6.59237385e-01 -5.20494103e-01 -6.82017982e-01 1.29799962e-01
7.55861998e-01 4.18004304e-01 1.87363192e-01 4.60097082e-02
1.56050003e+00 -1.41985893e-01 6.87663198e-01 -3.05759996e-01
-1.02055065e-01 1.48052424e-01 9.59733129e-01 7.75563359e-01
-6.41492069e-01 7.52010345e-01 2.05084339e-01 -2.59040743e-01
-1.01835847e+00 -1.41465878e+00 -3.58405381e-01 1.54937863e+00
1.82870701e-02 -4.18767780e-01 -5.58920145e-01 -1.09231079e+00
-8.94083977e-02 1.05849493e+00 -6.09476626e-01 2.16279216e-02
-2.89534062e-01 -9.25922871e-01 9.50445235e-01 7.10477352e-01
3.26336861e-01 -1.00381327e+00 -2.51687914e-01 4.04074520e-01
1.22131538e-02 -1.33765018e+00 -5.01010835e-01 7.08277822e-01
-7.69267142e-01 -8.31390977e-01 -1.03274417e+00 -9.24741983e-01
4.00738150e-01 2.22289503e-01 1.29272187e+00 -4.66363072e-01
-4.24731076e-01 3.78568828e-01 -6.48674548e-01 -3.18963915e-01
4.83900942e-02 4.20072466e-01 2.00502664e-01 -4.56088424e-01
1.09816837e+00 -4.96556044e-01 -2.86751449e-01 1.87530369e-01
-7.12974072e-01 -8.48567665e-01 6.54669344e-01 1.09693348e+00
2.51117676e-01 -1.50278166e-01 8.76528144e-01 -1.12562382e+00
5.63257754e-01 -1.06781483e+00 -2.18672216e-01 4.67158496e-01
-4.62128371e-01 5.61701208e-02 7.67866194e-01 -6.37778759e-01
-1.33030581e+00 -1.34799242e-01 -5.57700619e-02 -1.91978335e-01
-5.84683537e-01 1.99621737e-01 -2.64089614e-01 3.61355901e-01
8.96682739e-01 -8.62594396e-02 -8.38653386e-01 -8.54280829e-01
3.35714370e-01 8.55467856e-01 4.33052838e-01 -5.91685951e-01
7.31982112e-01 5.46745025e-02 -6.21665418e-01 -8.64228487e-01
-1.11823750e+00 -1.06632268e+00 -8.18531215e-01 4.03872907e-01
9.87521887e-01 -1.18778348e+00 -1.09716624e-01 4.66798060e-02
-1.07003915e+00 -9.34749991e-02 -5.41267335e-01 7.19777226e-01
-4.82075214e-02 1.08506791e-01 -7.91373134e-01 -7.28209913e-01
-3.98204416e-01 -3.93325359e-01 1.13707614e+00 6.11976683e-01
-3.78864169e-01 -1.33081365e+00 4.92301911e-01 -8.30882564e-02
3.54663938e-01 8.67285132e-02 7.89532542e-01 -1.67797863e+00
2.24314839e-01 -3.18684578e-01 -3.58433753e-01 3.27822506e-01
1.35064170e-01 -5.05472660e-01 -1.14415073e+00 -1.47065550e-01
-4.81748134e-01 -5.56254566e-01 6.33951128e-01 -2.71160295e-03
5.04233181e-01 -1.85480993e-02 -3.09335440e-01 3.99829775e-01
1.72477770e+00 1.46234915e-01 5.27664125e-01 5.58754504e-01
4.58846658e-01 3.32569510e-01 7.57682502e-01 5.82088709e-01
3.37720960e-01 3.20277870e-01 -2.04859167e-01 6.05842005e-03
-1.43575639e-01 -2.83903986e-01 3.93522173e-01 8.88342500e-01
1.24370024e-01 -3.67091537e-01 -1.10007977e+00 1.03589201e+00
-1.51421583e+00 -1.09782839e+00 9.95622799e-02 2.04471755e+00
8.19542766e-01 -4.89479750e-02 -2.05697101e-02 -5.06346762e-01
1.05185688e+00 1.16973184e-01 -5.62247992e-01 -4.64422911e-01
-8.84125233e-02 5.19158840e-01 6.20757580e-01 -4.45546210e-02
-1.47963834e+00 1.20367372e+00 6.51825333e+00 9.16334927e-01
-7.54952073e-01 5.56606472e-01 3.59948695e-01 7.29918480e-02
2.53931303e-02 3.01591419e-02 -1.23278534e+00 4.96641278e-01
1.32245922e+00 -3.63544613e-01 -1.51108801e-01 1.24057376e+00
-3.44446063e-01 2.68962473e-01 -8.28669131e-01 7.16190159e-01
-7.39839971e-02 -9.40245986e-01 -4.19909596e-01 -5.82750887e-02
7.68353879e-01 4.70537186e-01 -1.96377575e-01 1.07119441e+00
8.47391367e-01 -7.64621556e-01 2.43886203e-01 1.56861022e-01
8.64798903e-01 -7.77271867e-01 1.04916990e+00 1.77509040e-01
-1.30327237e+00 1.13170356e-01 -8.63350511e-01 2.11260870e-01
2.82400519e-01 5.93007684e-01 -9.94701207e-01 4.81416792e-01
7.24216759e-01 4.93696660e-01 -4.40126657e-01 1.09635746e+00
6.12736642e-02 7.49337912e-01 -2.57599473e-01 -6.29285797e-02
3.61709684e-01 2.84970373e-01 3.18087369e-01 1.68436313e+00
5.60363293e-01 2.99391061e-01 1.98141783e-01 4.85076189e-01
-4.85838532e-01 3.09439898e-01 -6.80264533e-01 -4.06050861e-01
9.29774880e-01 1.56760752e+00 -7.39426136e-01 -3.73751372e-01
-7.78371334e-01 1.23974872e+00 7.02545881e-01 3.40576857e-01
-9.37460482e-01 -1.04181790e+00 1.01383436e+00 -2.28466496e-01
7.00684071e-01 -2.29036376e-01 5.66411036e-05 -1.40853703e+00
-3.44569564e-01 -4.04673219e-01 7.42380500e-01 -3.26472372e-01
-1.91626966e+00 7.90058434e-01 -2.51303703e-01 -1.40364349e+00
2.98038609e-02 -6.23812079e-01 -6.91069782e-01 6.62421286e-01
-1.70009387e+00 -1.04118884e+00 3.46684232e-02 3.86122286e-01
5.45650423e-01 -2.69328624e-01 1.39931858e+00 5.72905362e-01
-5.22505701e-01 8.90010357e-01 5.06157160e-01 6.00451767e-01
1.19366872e+00 -1.52411997e+00 5.66612959e-01 6.30564213e-01
2.59482145e-01 6.63195968e-01 5.77257752e-01 -5.41521132e-01
-8.58811259e-01 -1.36201751e+00 1.31928802e+00 -4.66395259e-01
8.59376967e-01 -3.43305349e-01 -8.66445541e-01 6.86376035e-01
1.56249404e-01 4.91237789e-01 1.33071148e+00 5.41842937e-01
-6.59158885e-01 2.13395819e-01 -1.52992475e+00 3.98708105e-01
1.16477633e+00 -5.45193553e-01 -1.01528049e+00 2.06163868e-01
6.96437001e-01 6.35334104e-02 -1.34994471e+00 8.13513026e-02
2.32357964e-01 -5.51705837e-01 9.28728342e-01 -1.13821638e+00
-2.90105194e-02 -1.20657682e-01 -4.16992724e-01 -1.56898165e+00
-4.13205951e-01 -3.42056394e-01 1.10833742e-01 1.95497739e+00
4.81357038e-01 -5.48457742e-01 5.68326175e-01 8.27284634e-01
-1.29632667e-01 -1.56914905e-01 -9.14842367e-01 -1.26315594e+00
2.93977201e-01 -2.39383563e-01 4.72712338e-01 1.48199201e+00
8.55396688e-02 4.99625474e-01 -2.74108857e-01 3.05901498e-01
5.74069977e-01 -2.60266185e-01 3.85083556e-01 -1.40137923e+00
4.62551937e-02 8.97712931e-02 -6.10309482e-01 -6.59594476e-01
4.83385742e-01 -9.50674295e-01 3.17040861e-01 -1.30602086e+00
2.76262850e-01 -7.40823686e-01 -7.55785048e-01 7.89451897e-01
-3.48813236e-01 2.18515173e-01 1.01045310e-01 -1.20552359e-02
-1.10892463e+00 4.88419652e-01 4.44302738e-01 1.26009062e-01
1.34216901e-03 -4.15648878e-01 -7.32124865e-01 5.70200264e-01
8.31270278e-01 -9.37594056e-01 -1.33833781e-01 -4.82827157e-01
-2.59939015e-01 -4.14445370e-01 -7.28016952e-03 -1.06579435e+00
1.66772544e-01 -9.08586234e-02 4.26478654e-01 -9.12546739e-02
1.57990158e-01 -5.30555904e-01 -2.37176031e-01 -4.61198427e-02
-3.52455437e-01 7.97794312e-02 1.92312717e-01 5.98739862e-01
-3.60860258e-01 -5.69138587e-01 7.50753164e-01 -9.39234197e-02
-1.39504480e+00 3.21524590e-01 -2.52895266e-01 8.50382209e-01
1.03726125e+00 -9.29200053e-02 -4.04160589e-01 1.19962484e-01
-7.22480595e-01 1.01352204e-02 3.90472531e-01 5.35547793e-01
1.78569391e-01 -1.58179784e+00 -5.63705623e-01 -1.54984698e-01
5.25870800e-01 -4.90183234e-01 4.14020181e-01 4.50710148e-01
-1.36825144e-01 3.29668164e-01 -9.83786583e-02 -1.23704150e-01
-1.13109601e+00 3.98285985e-01 -1.79583412e-02 -5.24759233e-01
-4.74048316e-01 8.87196660e-01 2.49220966e-03 -8.89921010e-01
-1.37582067e-02 -3.56931277e-02 -2.57028997e-01 2.78719813e-01
6.48087859e-01 4.38170016e-01 6.38221279e-02 -5.81940472e-01
-6.42847538e-01 2.48537004e-01 -2.18120769e-01 -3.00415847e-02
1.73831022e+00 -8.03688318e-02 5.80170453e-01 5.32928109e-01
1.69777262e+00 7.16773868e-02 -9.78748024e-01 -3.30758929e-01
6.95991516e-01 -2.83083797e-01 3.78311402e-03 -7.13969707e-01
-6.60787642e-01 7.98244655e-01 6.76667571e-01 1.54223591e-01
7.15028048e-01 9.58299264e-02 1.16625059e+00 5.45300066e-01
4.87319350e-01 -1.31604075e+00 -1.10911772e-01 8.73734891e-01
-5.99795431e-02 -1.37397814e+00 -3.30148309e-01 2.44484171e-02
-9.30722952e-01 9.71894741e-01 6.50721908e-01 -1.69455647e-01
6.86024547e-01 4.02699292e-01 1.72709957e-01 -9.09359753e-02
-6.02327585e-01 -6.72693729e-01 1.51439577e-01 8.69336784e-01
7.74261534e-01 3.06372736e-02 -2.31509700e-01 9.93428171e-01
1.55724913e-01 -1.21570922e-01 1.65877283e-01 1.09923398e+00
-5.43506086e-01 -1.21799421e+00 -3.49427648e-02 3.82695317e-01
-9.12465215e-01 -4.25569922e-01 -7.11569414e-02 1.01808059e+00
1.45688653e-01 8.56627643e-01 7.87950382e-02 -1.31338552e-01
4.88100052e-01 4.92563158e-01 -2.45924313e-02 -1.02926016e+00
-5.79693913e-01 -3.69010091e-01 5.42637885e-01 -1.96126744e-01
-2.68474221e-01 -6.60672903e-01 -1.42066920e+00 -1.72213897e-01
-5.27760684e-01 5.75789571e-01 4.74179387e-01 8.06357443e-01
6.43189251e-01 2.11299568e-01 3.98802489e-01 -4.16055828e-01
-7.85847902e-01 -1.05809283e+00 -9.61326838e-01 7.06777394e-01
-1.52688652e-01 -8.26738417e-01 -4.50851142e-01 -1.62017830e-02] | [9.6933012008667, 9.408905029296875] |
d6f34da9-b3a3-406f-90ea-907660bde98c | single-image-reflection-removal-beyond | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Wen_Single_Image_Reflection_Removal_Beyond_Linearity_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Wen_Single_Image_Reflection_Removal_Beyond_Linearity_CVPR_2019_paper.pdf | Single Image Reflection Removal Beyond Linearity | Due to the lack of paired data, the training of image reflection removal relies heavily on synthesizing reflection images. However, existing methods model reflection as a linear combination model, which cannot fully simulate the real-world scenarios. In this paper, we inject non-linearity into reflection removal from two aspects. First, instead of synthesizing reflection with a fixed combination factor or kernel, we propose to synthesize reflection images by predicting a non-linear alpha blending mask. This enables a free combination of different blurry kernels, leading to a controllable and diverse reflection synthesis. Second, we design a cascaded network for reflection removal with three tasks: predicting the transmission layer, reflection layer, and the non-linear alpha blending mask. The former two tasks are the fundamental outputs, while the latter one being the side output of the network. This side output, on the other hand, making the training a closed loop, so that the separated transmission and reflection layers can be recombined together for training with a reconstruction loss. Extensive quantitative and qualitative experiments demonstrate the proposed synthesis and removal approaches outperforms state-of-the-art methods on two standard benchmarks, as well as in real-world scenarios.
| [' Shengfeng He', ' Guoqiang Han', ' Wenxi Liu', ' Jing Qin', ' Yinjie Tan', 'Qiang Wen'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['reflection-removal'] | ['computer-vision'] | [ 5.70123792e-01 9.30180103e-02 4.22157258e-01 -8.31593871e-02
-3.30066800e-01 -2.22168133e-01 5.46618223e-01 -7.34445512e-01
-1.93318591e-01 4.40032661e-01 2.17548773e-01 -3.43940645e-01
2.46316224e-01 -7.56769180e-01 -9.19685423e-01 -1.09062994e+00
5.18313527e-01 -7.77793974e-02 9.74918157e-02 -2.62197375e-01
-6.92213550e-02 3.59338403e-01 -1.21664727e+00 5.75377285e-01
1.14305961e+00 9.96555746e-01 2.49497548e-01 5.55731177e-01
1.64127365e-01 8.27008963e-01 -5.92732370e-01 -4.06232864e-01
7.33177423e-01 -6.82248533e-01 1.02778517e-01 1.95297822e-01
5.87492287e-01 -6.57992363e-01 -5.60022652e-01 1.00038934e+00
4.08967227e-01 6.00336753e-02 5.79054594e-01 -8.38185728e-01
-9.93111432e-01 5.30598342e-01 -8.66995573e-01 -4.94486660e-01
6.11109138e-02 4.81085420e-01 7.14506388e-01 -8.57172191e-01
1.97846159e-01 1.13592935e+00 7.27319300e-01 4.90923584e-01
-1.12740612e+00 -8.50919604e-01 1.39568925e-01 -2.08245918e-01
-1.14726889e+00 -6.92780554e-01 9.79264081e-01 -3.46803546e-01
5.97769022e-01 3.20845276e-01 6.60247922e-01 1.02090943e+00
2.64495522e-01 5.28056204e-01 1.12867498e+00 -3.58058989e-01
-1.59468621e-01 4.13129628e-01 -2.05072924e-01 7.36901700e-01
2.08906353e-01 2.81805307e-01 -2.89486885e-01 2.42787063e-01
1.00215816e+00 3.75539303e-01 -7.52035081e-01 -1.30704284e-01
-1.04208696e+00 3.69239181e-01 6.96231723e-01 2.25967765e-02
-2.73191392e-01 1.16291985e-01 -1.51101470e-01 4.10767823e-01
5.39256990e-01 4.21396881e-01 -2.38628030e-01 6.63803279e-01
-9.45643127e-01 1.56638082e-02 6.58830822e-01 7.82901824e-01
8.55650961e-01 9.63645428e-02 -5.71201205e-01 9.59642649e-01
4.11576509e-01 5.97462893e-01 9.39253941e-02 -6.90217912e-01
6.06257737e-01 5.19364059e-01 2.63958663e-01 -7.20601499e-01
-2.62709647e-01 -8.02319467e-01 -1.33456957e+00 5.15458226e-01
3.91466498e-01 -2.31752068e-01 -1.13436806e+00 1.64626598e+00
7.46009052e-02 4.53852564e-01 1.48148194e-01 1.23217380e+00
8.44309747e-01 6.66802466e-01 -5.43124199e-01 -2.42658794e-01
1.14906716e+00 -1.52469218e+00 -6.29610538e-01 -1.93738624e-01
5.89725189e-02 -1.05412281e+00 1.03729272e+00 5.93456984e-01
-1.29099166e+00 -6.11090362e-01 -1.10717344e+00 -2.94459820e-01
1.03426032e-01 5.69936991e-01 4.94778335e-01 2.43369669e-01
-7.71077871e-01 3.93011063e-01 -6.40649199e-01 4.05642480e-01
3.55291724e-01 -4.66036387e-02 -6.84170425e-02 -3.23850125e-01
-1.05225980e+00 7.47719586e-01 -2.84688622e-01 6.29553735e-01
-7.95920551e-01 -9.08665359e-01 -5.19593000e-01 2.31030613e-01
3.52237016e-01 -9.28952932e-01 8.40434968e-01 -1.19230318e+00
-1.84973705e+00 5.43080449e-01 2.03747917e-02 -2.04501420e-01
1.00686276e+00 -3.66054565e-01 -3.53105694e-01 -3.56206149e-02
-4.15140122e-01 1.81537539e-01 1.35066485e+00 -1.83546925e+00
-3.31428945e-01 -3.07659563e-02 7.75627233e-03 4.12300915e-01
-1.64109588e-01 -3.81270528e-01 -6.09896779e-01 -5.51607788e-01
1.18823521e-01 -8.25255215e-01 -1.18315980e-01 1.90480113e-01
-6.32903278e-01 5.66770673e-01 8.74954164e-01 -7.82275319e-01
9.12455142e-01 -2.30925274e+00 1.49072617e-01 8.61135051e-02
4.65277761e-01 4.77984697e-01 -4.50447530e-01 2.24142700e-01
-2.46390790e-01 -1.44813254e-01 -3.24793965e-01 -6.73202336e-01
-1.43837258e-01 -2.41688266e-01 -6.00172222e-01 4.90872949e-01
1.83607846e-01 7.31308699e-01 -7.47482181e-01 1.09939598e-01
4.51252550e-01 9.98137772e-01 -6.36915624e-01 4.04577732e-01
-1.90193072e-01 7.20325410e-01 -2.28417560e-01 1.91489100e-01
1.04102612e+00 -1.55325353e-01 1.69864763e-02 -5.70518911e-01
-1.81938827e-01 1.39396667e-01 -1.01985836e+00 1.38668454e+00
-9.27969694e-01 5.55219114e-01 1.68065861e-01 -6.47937715e-01
1.09773242e+00 2.28451014e-01 3.66489738e-01 -1.03959131e+00
9.47853550e-03 2.06503898e-01 2.06468850e-02 -2.59629935e-01
2.74625957e-01 -1.29032001e-01 3.47773761e-01 5.46868205e-01
-2.83618003e-01 -1.47538140e-01 -2.34706521e-01 -4.92210239e-02
1.06218708e+00 1.59194246e-01 -1.47555560e-01 3.26636076e-01
6.02946162e-01 -7.47301042e-01 5.50202906e-01 6.49032474e-01
3.97207409e-01 1.19951057e+00 3.79252344e-01 -2.91338891e-01
-9.97971535e-01 -1.23815405e+00 1.74617544e-01 6.43950880e-01
4.09321904e-01 -6.49140552e-02 -7.48015940e-01 -3.62672061e-01
-2.37469465e-01 5.97233474e-01 -3.87737662e-01 -4.29577887e-01
-8.85578156e-01 -7.19149768e-01 1.96659952e-01 -5.31143602e-03
9.84742105e-01 -1.04029405e+00 -5.59612453e-01 -4.93679531e-02
-2.33241305e-01 -1.20833337e+00 -6.47843242e-01 3.57653983e-02
-5.40125251e-01 -1.08890080e+00 -8.94035041e-01 -5.00532091e-01
8.72399390e-01 8.07619095e-01 1.12016678e+00 5.61913490e-01
-4.16000992e-01 -1.81114256e-01 -6.03786446e-02 -2.30258837e-01
-3.27820718e-01 -2.20692381e-01 -5.52152097e-01 3.31906557e-01
-2.75773734e-01 -8.40435922e-01 -1.09038877e+00 3.03076774e-01
-1.03618288e+00 7.22117424e-01 1.07328391e+00 8.93102527e-01
3.80800754e-01 1.27881065e-01 9.97435302e-02 -9.86229181e-01
5.33539593e-01 -4.68862690e-02 -6.40856326e-01 4.27145690e-01
-3.67162526e-01 -8.51096213e-02 9.91865575e-01 -5.82707107e-01
-1.50971580e+00 -6.38971254e-02 -3.60798091e-02 -6.40826225e-01
1.66502804e-01 -8.86979327e-02 -2.81182140e-01 1.04779694e-02
7.28907883e-01 1.89161420e-01 -2.65887771e-02 -4.91325855e-01
4.77781743e-01 4.50840920e-01 4.60630566e-01 -2.15140402e-01
1.25270069e+00 6.71178281e-01 -3.20876502e-02 -9.03104067e-01
-1.05422556e+00 -1.85302734e-01 -2.17925832e-01 -1.19172111e-01
6.85288906e-01 -1.12077141e+00 -7.05362618e-01 8.29018950e-01
-1.11979699e+00 -7.52125502e-01 -3.27727973e-01 3.15891534e-01
-2.29140058e-01 1.31053299e-01 -7.06229091e-01 -6.95073009e-01
-4.29674476e-01 -1.25741625e+00 1.07230806e+00 2.92260259e-01
4.41712499e-01 -6.05202675e-01 -5.65630645e-02 4.71556276e-01
5.87139487e-01 -1.20599233e-01 7.12235212e-01 1.74010143e-01
-9.94538546e-01 2.49346510e-01 -7.57544637e-01 7.78218210e-01
2.95822531e-01 -3.35961357e-02 -1.10535467e+00 -3.23903680e-01
3.34604740e-01 2.72452012e-02 1.45842385e+00 2.49542296e-01
1.11813390e+00 -4.41170096e-01 -3.48649058e-03 1.16192961e+00
1.37926400e+00 -8.11892375e-02 1.17388308e+00 5.86307384e-02
9.64614153e-01 4.40369546e-01 2.28265107e-01 2.26687834e-01
1.60909593e-01 6.46071136e-01 4.40739155e-01 -7.21414864e-01
-8.48320484e-01 -2.00790823e-01 3.85348767e-01 6.87478840e-01
-1.79252118e-01 -5.02436399e-01 -4.22556400e-01 -1.37020806e-02
-1.78383625e+00 -9.55726027e-01 -9.50349644e-02 2.36542869e+00
9.31907296e-01 -1.62528232e-02 -3.31887484e-01 -1.32920936e-01
3.88393283e-01 3.65727633e-01 -5.20592809e-01 3.85575145e-02
-5.61151095e-02 2.44964749e-01 5.40326536e-01 7.43506432e-01
-7.22006083e-01 7.33680427e-01 5.09872198e+00 4.55146313e-01
-1.54969108e+00 -8.94384682e-02 5.78997493e-01 -1.61974415e-01
-6.46418035e-01 6.15509078e-02 -6.75061345e-01 5.14503300e-01
2.62053162e-01 4.82710987e-01 9.48427320e-01 1.54308341e-02
3.87061030e-01 -3.42542864e-03 -8.90469193e-01 8.45464230e-01
8.37801024e-02 -9.33384717e-01 1.89331070e-01 -2.29051694e-01
8.41210544e-01 -1.85423598e-01 3.06218237e-01 7.18891993e-02
1.86393425e-01 -1.07433379e+00 7.10624516e-01 1.02444732e+00
8.18543136e-01 -4.87454534e-01 4.26724792e-01 2.95459062e-01
-8.05608213e-01 -6.79076314e-02 -2.37111270e-01 2.08180636e-01
1.90412626e-01 1.07209706e+00 -3.75569314e-01 6.50206745e-01
5.06553710e-01 6.65997744e-01 -2.00306401e-01 1.00257218e+00
-6.83992505e-01 3.61698657e-01 -1.46926746e-01 6.71766102e-01
-1.36514634e-01 -8.03568006e-01 4.51458663e-01 1.23797381e+00
3.27473968e-01 1.90608366e-03 -1.44810490e-02 1.12069404e+00
-3.15074772e-01 -4.73267466e-01 -2.52321929e-01 2.72680432e-01
1.47006795e-01 1.39927363e+00 -3.16906720e-01 2.12776549e-02
-5.41386664e-01 1.14752162e+00 1.89571321e-01 8.75038207e-01
-1.07956886e+00 -1.98576897e-01 7.04103529e-01 2.24249870e-01
2.13237733e-01 -5.22348657e-02 -3.55226785e-01 -1.40775979e+00
2.55637079e-01 -1.07696307e+00 -3.19560677e-01 -8.76978517e-01
-1.26711261e+00 5.17332911e-01 -3.76978338e-01 -1.30446410e+00
3.89281660e-01 -4.69016403e-01 -8.73707891e-01 1.22097278e+00
-2.09675312e+00 -1.33017087e+00 -7.61158288e-01 5.56907833e-01
2.04783037e-01 7.80187696e-02 2.86460727e-01 5.12299538e-01
-6.74030125e-01 5.27183771e-01 -2.66906843e-02 2.75138784e-02
1.05497479e+00 -8.48618686e-01 2.40298614e-01 9.71006751e-01
-1.18853219e-01 6.68171287e-01 6.49032712e-01 -3.57449532e-01
-1.28242195e+00 -1.09811163e+00 4.18342233e-01 1.56162813e-01
2.51737535e-01 -6.34636760e-01 -1.08174479e+00 5.21884024e-01
1.36443675e-01 1.36765733e-01 1.72736064e-01 -3.52298617e-01
-7.31883049e-01 -5.86677432e-01 -8.70190859e-01 8.57903123e-01
1.16999722e+00 -4.47723478e-01 -1.15470469e-01 3.43777418e-01
8.11897397e-01 -5.31427324e-01 -3.41507137e-01 3.69957387e-01
6.83861554e-01 -1.39208496e+00 1.33939970e+00 6.91662058e-02
9.83085573e-01 -4.63128805e-01 1.50769502e-01 -1.55315197e+00
-9.44881663e-02 -7.00068355e-01 -3.91073488e-02 1.02558959e+00
4.80030715e-01 -8.87708724e-01 5.76744020e-01 3.78413856e-01
-4.59237754e-01 -6.21199727e-01 -1.88430727e-01 -3.91792506e-01
-1.54835880e-01 -4.20806408e-02 5.07121980e-01 6.79075599e-01
-7.85972714e-01 4.34301019e-01 -8.00601065e-01 2.10478559e-01
7.28812695e-01 3.41080666e-01 9.31483746e-01 -8.68804276e-01
-7.95035243e-01 -5.08909822e-01 3.93422037e-01 -1.51041675e+00
-1.16221167e-01 -4.66915697e-01 4.43538010e-01 -1.57359028e+00
1.49383411e-01 -7.56457567e-01 -9.19843689e-02 2.71608382e-01
-2.45791584e-01 3.14415544e-01 2.92383075e-01 2.80915409e-01
-2.50018537e-01 7.36609817e-01 1.93656778e+00 -8.05583745e-02
-4.37553585e-01 1.91495806e-01 -9.67965901e-01 7.51010597e-01
6.70366108e-01 -1.17475614e-01 -7.25769579e-01 -9.15785849e-01
3.44254911e-01 1.00081325e-01 5.57019055e-01 -8.06862533e-01
2.17213377e-01 -1.77295759e-01 4.66137409e-01 -2.60471612e-01
5.22533059e-01 -9.37963903e-01 3.37615341e-01 1.78670585e-01
-2.77513504e-01 -5.75433969e-01 -4.52854186e-02 4.62436885e-01
-9.53449234e-02 1.13907501e-01 1.04273224e+00 -1.93271905e-01
9.79780219e-03 4.04002517e-01 2.23125875e-01 -1.39654920e-01
6.30037248e-01 -1.36898220e-01 -8.04552317e-01 -4.45950806e-01
-2.35419765e-01 -4.37917635e-02 6.33623719e-01 3.04590017e-01
6.85324669e-01 -9.02905345e-01 -7.14157641e-01 5.77598512e-01
-2.70403415e-01 4.07527119e-01 5.02888978e-01 9.59146321e-01
-4.04900908e-01 -1.28099203e-01 -6.40488788e-02 -2.38919467e-01
-1.10547340e+00 4.05952275e-01 6.01036429e-01 -4.55702484e-01
-9.15845573e-01 6.59682333e-01 9.28145170e-01 -3.14813644e-01
3.23666304e-01 -1.10973388e-01 -2.62477994e-02 -3.28408927e-01
6.31179512e-01 1.43466681e-01 9.16696042e-02 -2.18117297e-01
9.82334316e-02 5.66690624e-01 3.95294726e-02 2.88924687e-02
1.47164237e+00 -1.57125980e-01 -2.30231702e-01 2.00530469e-01
1.17282593e+00 2.76895791e-01 -1.67593217e+00 -3.82066488e-01
-8.29488099e-01 -5.64524353e-01 8.81295577e-02 -9.87290025e-01
-1.56380212e+00 9.64615107e-01 2.96036422e-01 1.43995006e-02
1.43502390e+00 -5.47627389e-01 9.73041594e-01 1.47165567e-01
-1.07473135e-01 -7.53559113e-01 3.59564632e-01 3.87289286e-01
1.10184264e+00 -9.76031899e-01 1.07304424e-01 -7.23570287e-01
-4.57818002e-01 9.84445870e-01 6.63169503e-01 -3.07432622e-01
5.35082340e-01 5.32591283e-01 2.43693888e-01 5.61364926e-02
-7.32713103e-01 6.34114295e-02 5.27895033e-01 4.15784895e-01
4.53644127e-01 -1.36209816e-01 -3.42327356e-02 2.50261128e-01
-1.06306836e-01 -5.40862381e-02 4.82607096e-01 3.34280223e-01
-2.55992949e-01 -9.16510761e-01 -2.57230788e-01 4.64203954e-01
-1.74544871e-01 -3.24521303e-01 -3.20315510e-01 5.77049494e-01
1.22114606e-01 8.14568579e-01 3.93645167e-02 -2.76923537e-01
4.13155854e-01 -5.85324168e-01 5.47461450e-01 -5.79773009e-01
-7.15792358e-01 2.52162874e-01 -1.74482882e-01 -4.43124115e-01
-1.74093097e-01 -1.11051120e-01 -9.09494340e-01 -3.54177237e-01
-4.64930385e-01 -2.81687766e-01 4.91558313e-01 7.16396153e-01
1.55147225e-01 9.38510776e-01 8.12452912e-01 -8.69251609e-01
-6.36536896e-01 -8.50110531e-01 -4.46247220e-01 5.30726910e-01
7.32047379e-01 -3.83973271e-01 -7.19209731e-01 -9.30789933e-02] | [10.645275115966797, -2.762925148010254] |
0ab977c5-789d-4043-9bd5-8399c96c5bec | gaussian-process-gradient-maps-for-loop | 2009.00221 | null | https://arxiv.org/abs/2009.00221v1 | https://arxiv.org/pdf/2009.00221v1.pdf | Gaussian Process Gradient Maps for Loop-Closure Detection in Unstructured Planetary Environments | The ability to recognize previously mapped locations is an essential feature for autonomous systems. Unstructured planetary-like environments pose a major challenge to these systems due to the similarity of the terrain. As a result, the ambiguity of the visual appearance makes state-of-the-art visual place recognition approaches less effective than in urban or man-made environments. This paper presents a method to solve the loop closure problem using only spatial information. The key idea is to use a novel continuous and probabilistic representations of terrain elevation maps. Given 3D point clouds of the environment, the proposed approach exploits Gaussian Process (GP) regression with linear operators to generate continuous gradient maps of the terrain elevation information. Traditional image registration techniques are then used to search for potential matches. Loop closures are verified by leveraging both the spatial characteristic of the elevation maps (SE(2) registration) and the probabilistic nature of the GP representation. A submap-based localization and mapping framework is used to demonstrate the validity of the proposed approach. The performance of this pipeline is evaluated and benchmarked using real data from a rover that is equipped with a stereo camera and navigates in challenging, unstructured planetary-like environments in Morocco and on Mt. Etna. | ['Teresa Vidal-Calleja', 'Wolfgang Stürzl', 'Rudolph Triebel', 'Mallikarjuna Vayugundla', 'Cedric Le Gentil', 'Riccardo Giubilato'] | 2020-09-01 | null | null | null | null | ['loop-closure-detection'] | ['computer-vision'] | [ 9.79619622e-02 -1.42556936e-01 4.85542595e-01 -3.36229771e-01
-7.06054330e-01 -7.64577627e-01 9.90514815e-01 2.91596204e-01
-6.48034215e-01 6.69552863e-01 -5.29379189e-01 -7.62469321e-02
-3.47361833e-01 -1.02765191e+00 -4.93715227e-01 -5.84849775e-01
-4.20728058e-01 1.02804995e+00 5.49638212e-01 -4.06089962e-01
4.70831990e-01 9.31803107e-01 -1.96681309e+00 -4.66034293e-01
1.16329026e+00 8.77517104e-01 5.53184330e-01 4.83551145e-01
-8.41718987e-02 -1.06390454e-01 -2.43194893e-01 1.90620542e-01
6.58926547e-01 2.37930477e-01 -2.61640042e-01 -6.79732254e-03
6.88179374e-01 1.87220141e-01 -1.30195513e-01 1.06486988e+00
3.20174485e-01 5.18171072e-01 9.28625584e-01 -1.25767338e+00
9.07474607e-02 -3.99029434e-01 -4.99188066e-01 3.07058524e-02
3.28911781e-01 -1.83713175e-02 6.24051809e-01 -9.75022018e-01
5.04694283e-01 1.01192832e+00 1.03897405e+00 -3.65413040e-01
-1.27277315e+00 -2.36210734e-01 1.91602670e-02 3.03244144e-01
-1.97058427e+00 -9.71759036e-02 3.79772961e-01 -6.22925282e-01
8.19176733e-01 1.07498519e-01 5.95705271e-01 4.89921838e-01
3.92060310e-01 -2.78831609e-02 1.49707973e+00 -4.96181846e-01
4.14511114e-01 1.31456912e-01 -2.56979644e-01 7.14029670e-01
4.65648264e-01 2.92382687e-01 -7.12934256e-01 -3.62911940e-01
6.08069777e-01 -1.95374270e-03 -2.36090884e-01 -9.45266068e-01
-9.75477517e-01 6.51471198e-01 6.08213544e-01 7.00798631e-02
-6.15983605e-01 -1.94731485e-02 -2.24361762e-01 -6.41383454e-02
1.87401012e-01 2.37250850e-01 1.18492462e-01 -2.23851010e-01
-1.19806278e+00 3.79179984e-01 8.51488233e-01 9.16410625e-01
1.15096247e+00 3.45104150e-02 3.39468837e-01 5.00561535e-01
7.03041434e-01 1.12761533e+00 1.95531785e-01 -4.65321034e-01
4.60588545e-01 5.50055146e-01 4.19034183e-01 -1.47561657e+00
-4.06152755e-01 -1.93088219e-01 -4.34677809e-01 8.25152814e-01
3.75267118e-01 1.44956455e-01 -1.21478391e+00 1.08529270e+00
4.99059200e-01 9.72616598e-02 2.57007837e-01 6.79156780e-01
3.18832219e-01 6.11313164e-01 -8.23717639e-02 5.66893160e-01
1.17359197e+00 -2.61557579e-01 -3.62633675e-01 -6.12399340e-01
1.25561371e-01 -7.42480576e-01 5.46534240e-01 4.07549292e-02
-2.96119660e-01 -4.80938494e-01 -1.36787379e+00 1.72611251e-01
-7.39884436e-01 2.11976513e-01 2.03539699e-01 5.43163478e-01
-1.21740341e+00 4.09932494e-01 -8.76823246e-01 -8.38180244e-01
8.83990303e-02 2.91074634e-01 -6.00655437e-01 1.24063298e-01
-1.04343951e+00 1.22778678e+00 4.65085357e-01 3.44179600e-01
-6.62739277e-01 -3.92186224e-01 -1.14385498e+00 -4.74836677e-02
-8.28559399e-02 -2.59549588e-01 5.67787945e-01 -4.93900746e-01
-1.05693889e+00 8.95528078e-01 -4.03238326e-01 -7.63705015e-01
8.63692820e-01 -2.46246114e-01 -1.81375444e-01 4.89767157e-02
4.62627262e-01 5.27243853e-01 8.29586327e-01 -1.40705514e+00
-1.17168498e+00 -5.13259470e-01 -4.40758228e-01 7.20200777e-01
4.66351181e-01 -4.59996879e-01 -2.80956388e-01 -2.58902311e-01
9.20430183e-01 -1.23675931e+00 -3.93692374e-01 1.47997975e-01
3.39198858e-02 3.45037222e-01 9.50340748e-01 -7.70594776e-01
5.94400465e-01 -2.24198341e+00 -1.76277623e-01 9.99147952e-01
-1.30343840e-01 -2.04284310e-01 4.21225369e-01 5.82229316e-01
5.33863664e-01 -1.56088486e-01 -5.70874631e-01 -2.82725900e-01
8.19090605e-02 2.11315557e-01 -4.20119017e-01 8.14464986e-01
-7.16281682e-02 3.77268255e-01 -6.99811935e-01 -3.54991287e-01
5.85682988e-01 5.43914020e-01 2.36933559e-01 1.02755511e-02
-1.97652895e-02 5.45325994e-01 -4.44667369e-01 4.81668144e-01
1.04074550e+00 4.24841702e-01 -5.78476824e-02 5.04695550e-02
-7.63486862e-01 5.82372136e-02 -1.58380330e+00 1.37704706e+00
-2.93940842e-01 8.15046847e-01 -5.97373396e-02 -2.96223938e-01
1.60778153e+00 -1.03714705e-01 9.98884216e-02 -7.63336539e-01
-2.75594115e-01 4.60488141e-01 -3.41955245e-01 -9.10645127e-02
9.32159603e-01 3.70111875e-02 -9.65585932e-02 -2.39214450e-02
-3.10146064e-01 -7.94298708e-01 -2.59994000e-01 -1.47854507e-01
7.14602828e-01 4.25963491e-01 5.61845899e-01 -4.10552293e-01
5.08567095e-01 6.50601625e-01 4.03341860e-01 9.33992445e-01
-7.89536759e-02 8.90915632e-01 8.26012809e-03 -4.70596999e-01
-1.10867667e+00 -1.20222461e+00 -2.98396856e-01 2.83294499e-01
5.16732752e-01 -6.65884167e-02 -2.07520828e-01 -3.19951266e-01
2.44222552e-01 3.81227493e-01 -6.96250737e-01 2.07557708e-01
-2.09623769e-01 -5.78379631e-01 5.18449545e-01 8.87153074e-02
7.62333810e-01 -6.10933304e-01 -1.25256634e+00 3.03323835e-01
-1.86092824e-01 -1.00816464e+00 3.90792310e-01 2.25680321e-01
-8.12176466e-01 -8.25885713e-01 -4.32529658e-01 -4.68199760e-01
6.27985179e-01 4.76707101e-01 6.12635195e-01 -3.67572606e-01
-2.16353983e-01 4.78410006e-01 -1.98003188e-01 -5.09947479e-01
-5.34406267e-02 -1.88120887e-01 1.12675615e-01 1.91836119e-01
1.91407889e-01 -7.47929156e-01 -2.12156445e-01 5.41010678e-01
-2.48500317e-01 -2.45407507e-01 6.40628099e-01 6.47070110e-01
9.77126539e-01 1.87492251e-01 -3.00071567e-01 -2.24835709e-01
2.98374921e-01 -5.95801294e-01 -1.21575582e+00 1.57762393e-01
-4.05668676e-01 1.14249110e-01 7.84310102e-02 -1.43441215e-01
-1.01029432e+00 5.48419654e-01 3.15407276e-01 -2.40182802e-01
-4.06075269e-01 8.13393712e-01 -2.17151456e-03 -6.36034966e-01
8.71252775e-01 4.05973107e-01 3.56265195e-02 -3.49174738e-01
2.38734186e-01 6.78594410e-01 8.30176473e-01 -4.43596423e-01
1.27271497e+00 1.05855441e+00 5.37804127e-01 -1.21118140e+00
-1.64434731e-01 -8.59171629e-01 -9.71138537e-01 -2.40693390e-01
6.32233799e-01 -9.49106097e-01 -9.26027671e-02 4.40053165e-01
-7.92014778e-01 -2.08994925e-01 -4.90896702e-02 5.82661748e-01
-8.13836753e-01 4.51778978e-01 4.01352853e-01 -1.06263137e+00
-1.86137125e-01 -8.43522310e-01 9.27443862e-01 4.53569353e-01
-6.22341447e-02 -9.47538435e-01 5.47960699e-01 -2.09528849e-01
4.17073041e-01 6.37008250e-01 2.76872307e-01 -3.27402949e-01
-7.40809619e-01 -4.49910402e-01 -1.72044083e-01 -3.28461558e-01
2.15714633e-01 -1.41584232e-01 -9.96423781e-01 -1.19458616e-01
2.29830723e-02 2.55902022e-01 5.66874504e-01 2.18968749e-01
1.28191456e-01 1.63431078e-01 -4.58245516e-01 7.51514018e-01
1.62320197e+00 1.86327457e-01 7.76423514e-01 1.04728615e+00
5.62760055e-01 7.65079141e-01 1.27373493e+00 5.27484119e-01
5.72968006e-01 9.77994144e-01 6.27671063e-01 -5.45763113e-02
2.13413343e-01 -4.71738636e-01 1.58054233e-01 -2.84651905e-01
-2.17360124e-01 3.02634239e-01 -1.51223421e+00 7.34000862e-01
-2.05551958e+00 -7.67790735e-01 -4.45764482e-01 2.78573728e+00
3.13528590e-02 3.61953750e-02 -5.88806272e-01 -1.14215598e-01
7.19354928e-01 1.87223759e-02 -1.12753466e-01 -2.78246012e-02
-2.32939765e-01 1.43512025e-01 1.23713720e+00 7.76902914e-01
-1.51690984e+00 1.01511919e+00 5.12458992e+00 6.73937082e-01
-1.24570286e+00 -1.55248955e-01 -2.28378385e-01 6.23899043e-01
1.28736541e-01 2.84137130e-01 -8.86593819e-01 2.26559073e-01
8.20891440e-01 -1.33797362e-01 1.37057945e-01 1.04499328e+00
2.91340530e-01 -7.36998260e-01 -5.44228911e-01 1.07245433e+00
1.74585998e-01 -1.11046124e+00 -3.43860447e-01 2.85577893e-01
6.49281740e-01 4.94182885e-01 2.22489148e-01 -1.07866406e-01
2.25837201e-01 -9.89447653e-01 1.14178050e+00 8.12927723e-01
5.99932611e-01 -6.21315777e-01 7.27477670e-01 1.97614267e-01
-1.58999515e+00 1.48379073e-01 -3.92442971e-01 -5.75412400e-02
2.32643023e-01 2.04027623e-01 -1.47678077e+00 8.95612955e-01
1.06110883e+00 4.72414672e-01 -8.41412783e-01 1.83701110e+00
-5.53355217e-01 2.99403090e-02 -9.95065808e-01 2.19500065e-01
5.73677659e-01 -4.66668874e-01 9.02593076e-01 9.19507205e-01
8.26141536e-01 -3.13455373e-01 2.50544876e-01 7.60334074e-01
6.37147129e-01 1.08463630e-01 -1.09239435e+00 5.49506664e-01
5.68859041e-01 1.30947280e+00 -8.37533534e-01 -1.08708501e-01
-5.94974086e-02 7.98357010e-01 9.31684822e-02 1.62250981e-01
-4.45692867e-01 -4.69580054e-01 4.95568216e-01 2.89050400e-01
3.35966051e-01 -8.47089171e-01 -3.24446827e-01 -5.44229984e-01
1.57475725e-01 -5.39571524e-01 2.95521729e-02 -8.26524377e-01
-8.52600038e-01 6.61670268e-01 1.96184993e-01 -1.78732657e+00
-1.90605983e-01 -2.97934681e-01 -6.20007336e-01 1.50527894e+00
-1.52390659e+00 -1.31408250e+00 -8.84398401e-01 3.93535167e-01
1.92514107e-01 -7.65425041e-02 7.74693131e-01 -2.52792686e-01
3.52012962e-01 -1.99628636e-01 5.37484467e-01 -2.76442349e-01
6.92383409e-01 -1.44799793e+00 7.72882164e-01 1.20619285e+00
3.35885048e-01 3.63459766e-01 1.06411290e+00 -1.12536466e+00
-1.05937040e+00 -1.09211934e+00 9.19537008e-01 -4.16132390e-01
6.27928138e-01 -3.79662097e-01 -9.93579447e-01 4.91134018e-01
-3.82930070e-01 -1.51893243e-01 1.08664311e-01 -2.23622262e-01
-5.09604402e-02 -3.81323919e-02 -1.44008064e+00 5.84625602e-01
4.61923897e-01 -5.48297465e-01 -8.34752142e-01 -2.40519904e-02
-1.51254371e-01 -6.01785004e-01 -4.97521788e-01 4.98599738e-01
5.11622012e-01 -8.91834319e-01 8.08026671e-01 3.32554907e-01
-4.90813971e-01 -9.89694715e-01 -4.30369437e-01 -1.32926691e+00
-4.06498313e-02 -3.30314845e-01 3.81480038e-01 9.14403558e-01
2.83760309e-01 -9.25043941e-01 8.26405168e-01 5.75365603e-01
4.63013276e-02 5.29123358e-02 -1.32179713e+00 -9.84336317e-01
-4.59076077e-01 -2.45946422e-01 3.36767107e-01 5.36574185e-01
-3.02520752e-01 -2.85548538e-01 -1.35793209e-01 1.06073380e+00
8.96108747e-01 4.20194156e-02 1.05971229e+00 -1.69397736e+00
1.49765626e-01 1.65013224e-02 -1.12746429e+00 -4.93386090e-01
-1.02828160e-01 -5.36299765e-01 5.93870044e-01 -1.55673647e+00
-4.83470261e-01 -1.13888156e+00 2.84443051e-01 2.58740485e-01
1.35855526e-01 3.36933106e-01 6.91450909e-02 5.96321404e-01
-2.23733664e-01 8.04603636e-01 1.58738494e-01 -1.33736590e-02
-4.07559425e-01 5.61699197e-02 3.11418355e-01 7.07467020e-01
8.94965470e-01 -5.46237111e-01 -1.05414934e-01 -1.72359973e-01
2.19552204e-01 -3.62220377e-01 5.25030196e-01 -1.74480796e+00
5.57232678e-01 -5.49965091e-02 3.96594554e-01 -1.10650289e+00
6.61700189e-01 -1.12263584e+00 6.12195015e-01 3.00074339e-01
4.42369401e-01 2.62844682e-01 3.79963905e-01 7.58413970e-01
-3.92914593e-01 -4.29266989e-01 6.40799582e-01 -5.93619086e-02
-1.44977188e+00 1.85136557e-01 -4.98695612e-01 -3.85259688e-01
1.21656704e+00 -6.76461041e-01 -1.74758397e-02 -3.79685938e-01
-6.63614750e-01 2.40008280e-01 1.10901201e+00 3.00625950e-01
8.22537839e-01 -8.68062854e-01 -5.19002020e-01 4.32333082e-01
5.23873806e-01 8.77585709e-02 1.66832715e-01 7.80007899e-01
-1.04680133e+00 2.30747074e-01 -5.83439648e-01 -9.82266188e-01
-1.39359379e+00 1.21793412e-02 5.63100457e-01 6.75586518e-03
-7.51771033e-01 3.27604502e-01 -7.54743889e-02 -5.22410512e-01
-2.39149049e-01 -5.46091348e-02 -3.66147220e-01 6.23391615e-03
2.82584995e-01 3.63605261e-01 2.50095755e-01 -1.20711052e+00
-3.55315506e-01 8.82794201e-01 4.05804664e-01 -8.05315137e-01
9.96191382e-01 -4.32600766e-01 1.09017566e-01 4.36552584e-01
4.06799495e-01 1.70686767e-01 -1.24345469e+00 -1.28800347e-01
3.41670752e-01 -8.17410707e-01 1.36784390e-01 -4.45277274e-01
-2.07585976e-01 8.18793774e-01 1.09486425e+00 -2.22623631e-01
4.73768175e-01 -2.57384658e-01 -2.19355136e-01 6.52798414e-01
1.03150797e+00 -9.54594791e-01 -7.33464658e-01 7.87235379e-01
8.59849453e-01 -1.21715415e+00 5.07825799e-02 -4.80959326e-01
-6.88622236e-01 1.11750877e+00 1.22003295e-01 -1.21589698e-01
5.40648103e-01 -1.20190838e-02 2.44824618e-01 -1.87639669e-01
1.84038103e-01 -5.87887764e-01 2.52886236e-01 1.14364028e+00
-3.21216106e-01 1.09759733e-01 -3.67087089e-02 -5.47032617e-02
-4.36984360e-01 -2.80832767e-01 5.51291406e-01 1.28689492e+00
-7.15960562e-01 -8.29755306e-01 -1.13605988e+00 1.00446135e-01
9.70522836e-02 -7.32352510e-02 -1.85495049e-01 8.71702492e-01
5.56345545e-02 7.43557692e-01 2.50015289e-01 -1.30919829e-01
3.74737054e-01 -5.72404414e-02 2.33935922e-01 -4.51237679e-01
-2.58549631e-01 -2.33825281e-01 8.91278088e-02 -5.82338274e-01
-1.49535045e-01 -8.81109595e-01 -1.16954303e+00 4.07786258e-02
7.62510970e-02 3.40023279e-01 1.25448370e+00 6.97839022e-01
4.18843597e-01 -2.31931031e-01 4.15145934e-01 -1.34125018e+00
-3.19983721e-01 -7.53926754e-01 -7.18909979e-01 -2.87949052e-02
1.37650073e-01 -1.24110854e+00 -2.39741892e-01 -6.89945891e-02] | [7.326570510864258, -2.087146043777466] |
fc4af598-14bb-4117-afaf-00d70efc8f90 | novel-features-for-the-detection-of-bearing | 2304.08249 | null | https://arxiv.org/abs/2304.08249v1 | https://arxiv.org/pdf/2304.08249v1.pdf | Novel features for the detection of bearing faults in railway vehicles | {In this paper, we address the challenging problem of detecting bearing faults from vibration signals. For this, several time- and frequency-domain features have been proposed already in the past. However, these features are usually evaluated on data originating from relatively simple scenarios and a significant performance loss can be observed if more realistic scenarios are considered. To overcome this, we introduce Mel-Frequency Cepstral Coefficients (MFCCs) and features extracted from the Amplitude Modulation Spectrogram (AMS) as features for the detection of bearing faults. Both AMS and MFCCs were originally introduced in the context of audio signal processing but it is demonstrated that a significantly improved classification performance can be obtained by using these features. Furthermore, to tackle the characteristic data imbalance problem in the context of bearing fault detection, i.e., typically much more data from healthy bearings than from damaged bearings is available, we propose to train a One-class \ac{SVM} with data from healthy bearings only. Bearing faults are then classified by the detection of outliers. Our approach is evaluated with data measured in a highly challenging scenario comprising a state-of-the-art commuter railway engine which is supplied by an industrial power converter and coupled to a load machine. | ['Walter Kellermann', 'Alexander Schmidt', 'Matthias Kreuzer'] | 2023-04-14 | null | null | null | null | ['audio-signal-processing', 'fault-detection'] | ['audio', 'miscellaneous'] | [ 5.74534118e-01 7.47948512e-02 3.36332142e-01 -1.61733747e-01
-8.05902421e-01 -4.43643890e-02 3.84224862e-01 7.07029164e-01
-3.02165419e-01 6.85990870e-01 -4.19587970e-01 2.53367648e-02
-4.87148404e-01 -5.33013403e-01 -4.03150350e-01 -8.87800395e-01
-3.23090911e-01 1.72514036e-01 2.95328051e-01 -4.65637207e-01
3.43343347e-01 6.58779979e-01 -2.29495525e+00 -2.39923410e-03
8.43400419e-01 1.25388300e+00 2.40026250e-01 5.91053367e-01
3.97199750e-01 2.85240501e-01 -1.26650703e+00 1.47626042e-01
1.11703843e-01 -4.11516458e-01 -5.20218670e-01 3.61385047e-01
1.15066350e-01 1.08399257e-01 3.01967915e-02 8.64368498e-01
6.46020174e-01 1.58555642e-01 6.09402180e-01 -1.29551744e+00
4.04819518e-01 2.34584019e-01 -1.92366332e-01 4.23181534e-01
4.82811332e-01 -1.25821471e-01 9.27578330e-01 -9.83902693e-01
2.00119868e-01 6.45952880e-01 4.49193895e-01 4.89056408e-02
-1.22898173e+00 -2.19882190e-01 -1.88885257e-01 6.14858270e-01
-1.24792874e+00 -4.01229501e-01 1.07197690e+00 -5.02327263e-01
8.37486148e-01 4.39090490e-01 5.47466636e-01 8.36546481e-01
3.11250716e-01 5.84451973e-01 9.68852878e-01 -6.07047558e-01
3.41606915e-01 -8.09721798e-02 1.25809759e-01 -3.44860107e-02
1.03954613e-01 1.44046038e-01 -5.80831468e-01 -1.01832906e-02
1.83846295e-01 -1.21796481e-01 -5.33707857e-01 -1.13884896e-01
-9.92902219e-01 6.88088179e-01 5.27622737e-02 7.47543752e-01
-5.48780739e-01 -2.72795022e-01 6.75336421e-01 8.03068697e-01
6.29287541e-01 2.66608030e-01 -4.03117388e-01 -3.55595142e-01
-9.73375857e-01 2.70986140e-01 7.89373696e-01 3.25542688e-01
4.15501207e-01 4.37136263e-01 3.33105654e-01 7.43847668e-01
-3.08173988e-02 4.46586043e-01 6.81455791e-01 -2.03808933e-01
3.73835444e-01 2.83067852e-01 7.88877085e-02 -9.26666081e-01
-6.12123787e-01 -5.23861527e-01 -7.31882989e-01 3.37031662e-01
3.93271476e-01 -1.02164254e-01 -4.90321189e-01 1.17540467e+00
2.94643402e-01 1.06502689e-01 2.22646073e-01 8.55433524e-01
2.08579555e-01 2.82544285e-01 -4.15970564e-01 -5.03703833e-01
1.24931276e+00 -1.76922396e-01 -1.07344937e+00 4.28717583e-02
5.55569172e-01 -9.59763288e-01 9.49157119e-01 1.05231702e+00
-8.74867797e-01 -8.08379352e-01 -1.49138045e+00 6.92624748e-01
-5.18291116e-01 3.15952480e-01 -9.26153511e-02 7.37023473e-01
-4.72167432e-01 1.00125897e+00 -8.37813973e-01 -1.99487936e-02
-2.84285456e-01 3.43175650e-01 -4.06356603e-01 1.86355978e-01
-1.40876174e+00 9.64055359e-01 3.51535439e-01 4.42103535e-01
-4.62368935e-01 -3.40870589e-01 -7.96108663e-01 3.83239090e-02
4.55926478e-01 1.97889641e-01 9.65185702e-01 -7.24515498e-01
-1.42717052e+00 2.93195099e-01 2.47727364e-01 -4.97344404e-01
4.80929464e-01 -4.66450363e-01 -8.61173093e-01 3.06228399e-01
-2.80501276e-01 -5.10208547e-01 1.21706319e+00 -9.81092870e-01
-6.79646313e-01 -3.15327108e-01 -3.82562578e-01 -8.49172845e-02
-3.88978094e-01 -9.60930735e-02 4.89946961e-01 -7.96654999e-01
1.18885286e-01 -8.28163147e-01 1.12031385e-01 -6.99791849e-01
-3.91074508e-01 -2.14935273e-01 1.01274061e+00 -8.06737661e-01
1.09453535e+00 -2.38676214e+00 1.51796252e-01 4.80489403e-01
-2.30647609e-01 3.68412942e-01 3.74490649e-01 7.49925733e-01
-3.40296537e-01 -4.92582083e-01 -3.15381318e-01 -7.32511133e-02
-6.98747486e-02 1.94804117e-01 -2.52085567e-01 6.74738944e-01
6.83755577e-01 -1.64446030e-02 -5.99948645e-01 -1.22125365e-01
5.02707481e-01 4.36981618e-01 -2.08175063e-01 2.88371295e-01
2.89351672e-01 5.14212728e-01 -1.17925629e-01 4.10372466e-01
4.65868592e-01 5.14778376e-01 -1.29794300e-01 -2.83243716e-01
-6.16934672e-02 4.08033840e-03 -1.59566379e+00 1.30743706e+00
-6.19572699e-01 5.48251569e-01 1.18991874e-01 -1.60333991e+00
1.33377004e+00 7.75898397e-01 5.73648155e-01 -4.62753534e-01
2.29748473e-01 6.45394683e-01 2.79139787e-01 -6.34826303e-01
5.12618184e-01 -2.25978300e-01 -1.32736325e-01 -1.16559803e-01
1.06420122e-01 -3.30383986e-01 4.51588720e-01 -4.23002839e-01
1.07280970e+00 -4.05881517e-02 2.78926175e-02 -2.52610445e-01
8.78703177e-01 -2.01342523e-01 4.23488200e-01 1.99329704e-01
8.16891864e-02 6.60840213e-01 4.18272138e-01 3.88212735e-04
-6.43633962e-01 -9.66144681e-01 -4.25211370e-01 6.87600195e-01
-1.63014024e-01 -2.59464830e-01 -5.22247136e-01 -4.35492367e-01
9.82492641e-02 5.52264810e-01 -2.90852278e-01 -5.20346463e-01
-7.75774658e-01 -8.08363855e-01 3.66487771e-01 4.07157838e-01
-2.20105723e-02 -8.22707236e-01 -1.03859389e+00 5.74121296e-01
5.29042147e-02 -1.11888373e+00 1.13203682e-01 8.15746546e-01
-9.74261880e-01 -1.22830784e+00 -5.66614330e-01 -5.71874261e-01
4.29458976e-01 -1.64173678e-01 8.83413732e-01 8.52204263e-02
-6.24047935e-01 2.83144712e-01 -7.50708222e-01 -5.24864674e-01
-5.70783794e-01 -5.47054969e-02 2.88139045e-01 3.94538015e-01
1.38246164e-01 -6.14357173e-01 -3.00732017e-01 4.88733649e-01
-1.10678458e+00 -6.44795954e-01 6.59479737e-01 1.13091624e+00
1.46486059e-01 8.07166278e-01 1.17985988e+00 -5.70425332e-01
6.16193533e-01 -3.96113127e-01 -3.77414286e-01 -2.95003742e-01
-5.31126261e-01 -2.09644958e-01 1.08734894e+00 -4.17611927e-01
-8.48755479e-01 -4.63402420e-02 -4.79126006e-01 -2.73990810e-01
-5.80002367e-01 6.95381343e-01 -2.87937045e-01 9.13621858e-02
5.02967060e-01 1.59428477e-01 1.91597536e-01 -7.69522130e-01
-8.23686197e-02 1.01192021e+00 8.72663736e-01 -5.83079636e-01
8.06800485e-01 9.81322750e-02 3.24165702e-01 -1.31253254e+00
-2.90033758e-01 -7.32966602e-01 -7.49647319e-01 -4.27225471e-01
2.91814446e-01 -6.11617625e-01 -6.00071549e-01 7.14721918e-01
-7.58209467e-01 2.16378585e-01 -6.99307501e-01 7.45206654e-01
-6.92409098e-01 4.46735889e-01 -5.20884395e-01 -1.21107733e+00
-1.03517473e-01 -1.01167727e+00 1.03529704e+00 -6.74139857e-02
-1.72975376e-01 -7.84597814e-01 -2.33433023e-01 -2.34545413e-02
3.39628190e-01 8.21247220e-01 7.89416015e-01 -9.11764562e-01
2.66342849e-01 -8.24303269e-01 5.58620930e-01 8.31561923e-01
3.39453280e-01 -1.09836154e-01 -1.11167204e+00 -6.60994649e-01
4.89601016e-01 -1.41383097e-01 5.16416669e-01 -5.51849864e-02
7.05015659e-01 1.48097649e-01 -1.49866892e-02 -2.07480252e-01
1.29162276e+00 2.24954262e-01 3.66629750e-01 1.02143951e-01
1.71730593e-01 9.06403303e-01 1.22131956e+00 5.90831995e-01
-3.42891425e-01 9.23097610e-01 5.43374479e-01 -2.49305032e-02
1.71790972e-01 3.96418035e-01 5.41300595e-01 1.11593652e+00
-2.55850881e-01 2.74841134e-02 -6.60112798e-01 6.91647053e-01
-1.46000206e+00 -5.98875761e-01 -4.35619742e-01 2.39653802e+00
6.51138186e-01 6.99170411e-01 2.75031239e-01 1.63365710e+00
6.97033703e-01 -2.23672949e-02 -1.59242317e-01 -2.27125168e-01
-7.59726167e-02 4.91031885e-01 2.21745282e-01 1.18523322e-01
-9.98166919e-01 -1.30196691e-01 4.74540854e+00 8.73118877e-01
-1.26446235e+00 -1.36646271e-01 -2.19768994e-02 -5.59887476e-03
4.04809952e-01 -2.08981767e-01 -4.05267775e-01 5.49960256e-01
1.31459880e+00 8.73849075e-03 -1.42685190e-01 7.10074902e-01
1.81550175e-01 -4.84904945e-01 -1.02675664e+00 7.54685938e-01
1.26285449e-01 -4.41580504e-01 -7.12818563e-01 1.98321752e-02
1.30295530e-01 -3.77785325e-01 -5.15986746e-03 1.11617379e-01
-7.34568834e-01 -7.12595701e-01 8.08193803e-01 4.88420218e-01
5.57549775e-01 -1.03886425e+00 1.21397793e+00 3.88636231e-01
-1.22970283e+00 -2.43239865e-01 -1.74183384e-01 -1.52322993e-01
3.37555885e-01 1.16533196e+00 -7.59659171e-01 1.28936446e+00
6.27481639e-01 7.04741061e-01 -4.22749668e-01 1.21881187e+00
1.13083236e-01 6.93544447e-01 -4.11159277e-01 2.86333501e-01
-2.23222762e-01 -1.82891858e-03 7.14824200e-01 1.06719720e+00
5.88838041e-01 -3.18644911e-01 1.10352375e-01 1.47523671e-01
6.48025751e-01 2.15575218e-01 -4.97097194e-01 2.51499116e-01
2.03010395e-01 1.13559079e+00 -8.15011978e-01 -1.10433362e-01
-3.91490370e-01 6.71614110e-01 -3.80549282e-01 5.00055738e-02
-5.96781313e-01 -9.25184727e-01 5.02121031e-01 2.06836075e-01
3.78455549e-01 -1.83937415e-01 2.15071008e-01 -8.52430284e-01
3.68737072e-01 -7.18510985e-01 2.89675564e-01 -3.63732338e-01
-1.29664862e+00 5.94457209e-01 1.67168736e-01 -1.73029160e+00
-5.72562456e-01 -6.99695230e-01 -6.03188396e-01 7.47036457e-01
-1.55591047e+00 -4.59596902e-01 -1.90890223e-01 5.19545496e-01
7.31666923e-01 -1.30919486e-01 7.85986125e-01 6.64522469e-01
-4.01599079e-01 3.18269610e-01 2.55224437e-01 -1.32977039e-01
6.49885356e-01 -1.51960087e+00 9.38904062e-02 7.68464386e-01
1.23088630e-02 4.08064313e-02 1.20239639e+00 -3.77731442e-01
-1.36710954e+00 -7.94593453e-01 7.53094018e-01 -9.88388062e-02
6.41861677e-01 -2.69575387e-01 -1.21618092e+00 5.05907536e-02
-4.34317738e-02 2.11485803e-01 5.35591543e-01 -3.00384879e-01
3.58107895e-01 -8.56373683e-02 -1.10623789e+00 -1.06081270e-01
3.74572486e-01 -6.20774686e-01 -9.63150859e-01 2.39324585e-01
1.04546987e-01 -4.23625320e-01 -1.44398880e+00 6.06807172e-01
2.26665214e-01 -9.38568592e-01 8.36765587e-01 -1.02542609e-01
-1.48820579e-01 -5.81345141e-01 -1.72355995e-01 -1.69865620e+00
3.61188591e-01 -5.30530632e-01 -2.10450381e-01 1.21513689e+00
1.24625610e-02 -7.44542480e-01 5.07342875e-01 -2.45412365e-01
-4.01642352e-01 -5.76150954e-01 -1.39577496e+00 -1.11671543e+00
-1.87958896e-01 -5.07070422e-01 3.81539911e-01 6.38292551e-01
2.57574618e-01 2.26111591e-01 -4.04170394e-01 2.43801549e-01
3.97368133e-01 -4.81349230e-02 5.07485867e-01 -1.76768327e+00
-3.50447536e-01 -9.71867889e-02 -9.68333960e-01 -4.63170767e-01
-7.04760402e-02 -4.96219009e-01 4.46314901e-01 -8.52970004e-01
-8.05442452e-01 -1.96668535e-01 -4.34772134e-01 -4.11814712e-02
-9.26782005e-03 3.87643993e-01 1.72638774e-01 5.30919619e-02
1.44664347e-01 5.10877728e-01 7.23439693e-01 -4.39091027e-02
1.97371915e-02 6.06148243e-01 2.61199802e-01 6.18045151e-01
6.40248716e-01 -4.13676172e-01 -2.69509763e-01 4.34330523e-01
1.58813372e-02 3.98190945e-01 3.64034355e-01 -1.56279385e+00
5.50454035e-02 3.65737796e-01 3.15765738e-01 -7.95397222e-01
4.18205291e-01 -1.21902442e+00 2.13110059e-01 7.48294413e-01
2.48339400e-02 1.81602359e-01 2.35892944e-02 6.05643034e-01
-7.88706243e-01 -4.15611506e-01 4.97580558e-01 3.81311983e-01
-5.15182197e-01 -6.00673854e-01 -7.13696420e-01 -3.91126692e-01
9.72508192e-01 -2.58017719e-01 -2.12842710e-02 -1.05271295e-01
-1.07770741e+00 -2.41451517e-01 1.60234645e-01 4.94913816e-01
6.09674633e-01 -1.14727008e+00 -7.97220647e-01 5.96984267e-01
2.71108389e-01 -3.96365486e-02 2.39678994e-01 1.38426697e+00
-1.70680016e-01 3.60092402e-01 -2.54989594e-01 -7.94722676e-01
-1.39568210e+00 4.06219780e-01 8.46095979e-02 -5.16919717e-02
-5.55857360e-01 2.76839286e-01 -5.63036382e-01 5.45425117e-02
1.64631993e-01 -5.86322427e-01 -5.37655473e-01 6.04950428e-01
2.49642730e-01 6.56538546e-01 1.02465534e+00 -9.06258941e-01
-3.14122111e-01 6.10933721e-01 3.43369186e-01 8.72305930e-02
1.33830428e+00 -6.10745102e-02 1.94780678e-01 8.90199602e-01
1.10080254e+00 2.01786980e-02 -9.83073294e-01 2.53660157e-02
4.22547042e-01 -3.56374502e-01 -8.96403417e-02 -2.80325770e-01
-1.09834814e+00 8.11292350e-01 8.60203803e-01 1.01594901e+00
1.41104972e+00 -4.06525493e-01 6.86988831e-01 2.96817571e-01
4.73248899e-01 -1.29920793e+00 -3.51390131e-02 2.00166166e-01
8.33029628e-01 -9.12081540e-01 -7.87267238e-02 -3.37796301e-01
-2.80136108e-01 1.54682803e+00 4.75962460e-02 -3.29689831e-01
8.15461338e-01 3.86705279e-01 -3.24312858e-02 -9.44502205e-02
-4.73995954e-01 -4.38905895e-01 1.91412032e-01 5.41913271e-01
3.99929404e-01 -1.55291229e-01 -6.18604481e-01 4.72202241e-01
-3.30034852e-01 -1.93975627e-01 7.20695734e-01 1.33356452e+00
-4.50057685e-01 -1.13871598e+00 -8.39970708e-01 5.74857771e-01
-7.58682013e-01 4.17747915e-01 1.64301664e-01 9.51946378e-01
9.91476253e-02 1.35693967e+00 6.13974296e-02 -4.35100257e-01
8.57851624e-01 3.30397874e-01 2.53276139e-01 -4.08814043e-01
-7.15636611e-01 2.06894472e-01 2.18790978e-01 -3.63280356e-01
-5.67028284e-01 -8.32251966e-01 -1.09999943e+00 2.47683465e-01
-9.11497533e-01 6.33691967e-01 9.16115224e-01 1.05164504e+00
-1.41411334e-01 1.16493154e+00 1.05069339e+00 -1.04666495e+00
-9.34002519e-01 -1.11153674e+00 -1.19041193e+00 6.10329151e-01
6.40081048e-01 -8.85843933e-01 -7.84535050e-01 1.43101498e-01] | [6.734221458435059, 2.4240124225616455] |
b3ac9184-c514-480b-8812-c1ca1e8dc412 | affordancenet-an-end-to-end-deep-learning | 1709.07326 | null | http://arxiv.org/abs/1709.07326v3 | http://arxiv.org/pdf/1709.07326v3.pdf | AffordanceNet: An End-to-End Deep Learning Approach for Object Affordance Detection | We propose AffordanceNet, a new deep learning approach to simultaneously
detect multiple objects and their affordances from RGB images. Our
AffordanceNet has two branches: an object detection branch to localize and
classify the object, and an affordance detection branch to assign each pixel in
the object to its most probable affordance label. The proposed framework
employs three key components for effectively handling the multiclass problem in
the affordance mask: a sequence of deconvolutional layers, a robust resizing
strategy, and a multi-task loss function. The experimental results on the
public datasets show that our AffordanceNet outperforms recent state-of-the-art
methods by a fair margin, while its end-to-end architecture allows the
inference at the speed of 150ms per image. This makes our AffordanceNet well
suitable for real-time robotic applications. Furthermore, we demonstrate the
effectiveness of AffordanceNet in different testing environments and in real
robotic applications. The source code is available at
https://github.com/nqanh/affordance-net | ['Ian Reid', 'Thanh-Toan Do', 'Anh Nguyen'] | 2017-09-21 | null | null | null | null | ['affordance-detection'] | ['computer-vision'] | [-7.43003935e-02 -9.10933688e-02 -1.03337526e-01 -4.15776551e-01
-2.66927958e-01 -2.76591778e-01 3.72367293e-01 -1.02733582e-01
-5.45937240e-01 2.62161255e-01 -7.28932247e-02 -3.58949229e-02
-1.09750733e-01 -4.20746714e-01 -8.68713796e-01 -6.57011509e-01
-2.00558960e-01 1.54829726e-01 4.15458113e-01 -3.87357026e-02
3.14514339e-01 6.32506192e-01 -1.67524958e+00 -1.12753831e-01
8.61736417e-01 1.28811431e+00 8.15006196e-01 5.89792192e-01
5.47303915e-01 7.85826147e-01 -1.77122295e-01 1.78668916e-01
5.08118570e-01 1.45662546e-01 -7.40339339e-01 8.04213062e-02
6.92353845e-01 -9.05751526e-01 -6.71696842e-01 1.07073736e+00
4.63763595e-01 4.48283315e-01 4.05630141e-01 -1.52025342e+00
-9.01882708e-01 2.27756172e-01 -3.93113106e-01 3.31520736e-01
3.69773179e-01 5.42229652e-01 1.27226639e+00 -8.36897254e-01
3.79393190e-01 1.45263135e+00 1.61212176e-01 5.14464974e-01
-1.11678410e+00 -5.77342451e-01 3.53902608e-01 2.42095485e-01
-1.30717599e+00 -3.30878347e-01 5.45178533e-01 -2.99968690e-01
1.31247699e+00 1.37049899e-01 7.56485403e-01 9.27644551e-01
1.52712047e-01 1.28698754e+00 9.35034335e-01 -7.00312555e-02
3.43619101e-03 -5.96838892e-01 1.08941585e-01 1.07606208e+00
6.56064302e-02 1.59292236e-01 -1.73038632e-01 -4.19267379e-02
1.35317731e+00 5.96617758e-01 -3.58838439e-01 -4.83739465e-01
-1.60591829e+00 4.21913654e-01 1.33408225e+00 -6.18601292e-02
-3.97408694e-01 6.33959770e-01 1.06810022e-03 1.46944389e-01
2.34047756e-01 3.35558087e-01 -5.13411820e-01 3.47392797e-01
-1.57055870e-01 5.99697709e-01 4.83345807e-01 1.26605237e+00
7.19961703e-01 -4.74927247e-01 -1.04288563e-01 6.77559495e-01
8.46570730e-01 4.41180319e-01 3.60696226e-01 -9.50850666e-01
4.61415082e-01 6.92698598e-01 3.33324909e-01 -6.74506664e-01
-6.48362935e-01 -6.67354465e-02 -3.60062569e-01 3.34287196e-01
3.79665554e-01 2.40548611e-01 -1.05186677e+00 1.75832415e+00
6.58680141e-01 3.22338670e-01 -3.13608617e-01 1.74208736e+00
8.29533100e-01 3.61052752e-01 2.07613289e-01 5.92188060e-01
1.52920771e+00 -1.49882746e+00 -4.16129202e-01 -3.80171210e-01
3.77084523e-01 -7.23140538e-01 1.63322580e+00 1.49063602e-01
-5.65980554e-01 -5.76322079e-01 -1.44413984e+00 -7.57656217e-01
-1.43806964e-01 5.76019943e-01 1.28288734e+00 1.01379575e-02
-6.33342445e-01 7.73577988e-01 -1.23011506e+00 -2.53630072e-01
6.15047395e-01 3.94315839e-01 -4.95092720e-01 -1.42117947e-01
-6.38132095e-01 7.63738334e-01 2.71788210e-01 3.42477053e-01
-1.33713257e+00 -3.68334621e-01 -1.23470354e+00 1.12602837e-01
5.70836961e-01 -8.74413490e-01 1.35580981e+00 -5.24604261e-01
-1.23764193e+00 7.67523408e-01 -2.65943236e-03 -1.30412862e-01
4.61328536e-01 -6.45525992e-01 -3.62169445e-02 6.67979419e-02
3.95105749e-01 8.82472932e-01 9.58935201e-01 -9.29914236e-01
-6.52994871e-01 -6.26986504e-01 5.57271004e-01 3.30988228e-01
1.18462354e-01 -5.57445884e-02 -3.98177594e-01 -7.10654020e-01
7.73169696e-01 -1.03899086e+00 -2.18474194e-01 6.21747911e-01
-6.81529880e-01 -6.96918964e-01 5.06584287e-01 -3.35596740e-01
4.82552499e-01 -2.27904987e+00 3.56195122e-01 -3.23729485e-01
3.57412100e-01 -2.72894144e-01 -2.71116585e-01 8.23296905e-02
6.39855117e-02 -2.31555581e-01 -2.04648599e-01 -4.39889282e-01
3.30112964e-01 2.15197101e-01 -2.46054187e-01 8.20413768e-01
3.84739101e-01 8.58191371e-01 -1.02448690e+00 -2.17548132e-01
4.09182429e-01 4.59809572e-01 -5.43977618e-01 5.27531147e-01
-3.68568003e-01 4.47421849e-01 -6.08098090e-01 8.52949619e-01
5.97673059e-01 -2.40691200e-01 -2.22070679e-01 -1.94655553e-01
-2.07587127e-02 6.70249581e-01 -1.37353230e+00 2.41447115e+00
-1.65210515e-01 2.88856626e-01 -5.29541001e-02 -2.97049552e-01
6.63756192e-01 -2.58585010e-02 1.80301636e-01 -4.60284501e-01
3.58706027e-01 4.07400548e-01 2.58185863e-01 -7.11820066e-01
3.41071099e-01 1.59294620e-01 -8.18865225e-02 6.09524190e-01
4.64496225e-01 -2.14752689e-01 2.10604653e-01 -5.73340096e-02
8.63881826e-01 8.01950216e-01 1.94969833e-01 -4.64414805e-01
3.31825495e-01 -2.49931604e-01 5.50871551e-01 4.48805839e-01
-5.80150902e-01 5.26849031e-01 2.15741217e-01 -8.69423032e-01
-9.02156711e-01 -1.13699508e+00 -2.54973412e-01 1.36096537e+00
6.83679998e-01 -2.22040024e-02 -7.01787770e-01 -4.90768790e-01
3.77285033e-01 3.96225512e-01 -6.85272157e-01 -1.39448732e-01
-4.09649312e-01 -6.00977123e-01 1.36274546e-01 6.42510533e-01
6.49906933e-01 -1.36958158e+00 -1.12082350e+00 -1.34880483e-01
-2.19691321e-02 -1.10698807e+00 -7.24129796e-01 1.05593279e-01
-8.38541746e-01 -1.28346205e+00 -6.19258821e-01 -7.70004332e-01
7.33221769e-01 5.22036135e-01 6.60015881e-01 3.30803841e-01
-6.45984948e-01 1.27823710e-01 -1.41854942e-01 -3.36352646e-01
1.70682356e-01 1.65932506e-01 1.42289773e-01 -3.25282849e-02
4.06420261e-01 -4.67808455e-01 -1.08633208e+00 3.40363622e-01
-8.89884293e-01 5.74739240e-02 6.48902118e-01 6.26692295e-01
7.60540485e-01 -5.62322438e-01 2.20334351e-01 -1.73614286e-02
3.02997649e-01 -2.92865187e-01 -6.43287182e-01 -1.89630836e-02
-3.19393277e-01 8.54182467e-02 4.63110656e-02 -4.75673467e-01
-6.78016901e-01 3.56403291e-01 -1.01481579e-01 -3.52001190e-01
-3.96166265e-01 -7.27621512e-03 -1.95015863e-01 -1.30943418e-01
6.24756813e-01 -1.96873605e-01 -2.92739104e-02 -7.40241349e-01
5.05745709e-01 5.80012083e-01 3.40543330e-01 -6.23056173e-01
4.17570889e-01 6.12386525e-01 -9.34895575e-02 -5.46074331e-01
-9.60029662e-01 -5.97881854e-01 -1.04344308e+00 -6.42670542e-02
9.11624253e-01 -1.11768866e+00 -9.53231514e-01 6.05064750e-01
-1.14277518e+00 -5.98740757e-01 -6.32571802e-02 6.27734184e-01
-7.58881330e-01 5.14418520e-02 -8.03829610e-01 -5.15255690e-01
-3.24434966e-01 -1.46410823e+00 1.33742630e+00 3.90974700e-01
6.34496957e-02 -2.65739292e-01 -4.08463627e-01 2.76840150e-01
-5.45279421e-02 3.25695992e-01 6.36287749e-01 -2.96698481e-01
-1.08565187e+00 -2.19519600e-01 -7.13562250e-01 1.16788454e-01
2.74894595e-01 -4.05198067e-01 -1.05493033e+00 -3.21947247e-01
-1.52315423e-01 -7.40098774e-01 1.10349369e+00 1.68728799e-01
1.11012983e+00 -2.36506537e-02 -3.14803421e-01 7.45269179e-01
1.37719011e+00 -3.05159032e-01 3.53188336e-01 3.87727499e-01
1.15312564e+00 3.96260262e-01 7.91436732e-01 2.56539166e-01
8.23152959e-01 4.47778136e-01 1.08264315e+00 2.57781558e-02
-2.37242728e-01 -2.67703027e-01 -5.62481731e-02 2.77557611e-01
-2.79752463e-01 1.95627645e-01 -8.01617086e-01 4.65125114e-01
-2.12156415e+00 -4.96025056e-01 -8.88161287e-02 2.00230670e+00
5.69154918e-01 -7.18930885e-02 1.76249936e-01 9.53626633e-03
3.41425180e-01 2.63815254e-01 -1.15632200e+00 1.21513464e-01
1.51749000e-01 -2.78816015e-01 3.59199941e-01 3.52094620e-01
-1.51447320e+00 1.06239676e+00 5.30915499e+00 2.42860645e-01
-9.50528741e-01 1.86850443e-01 2.53276199e-01 -1.60907015e-01
6.87513724e-02 -1.27767369e-01 -5.53592622e-01 -2.56846081e-02
1.59696519e-01 3.89463753e-01 9.24919426e-01 1.13569987e+00
-4.53372672e-02 -3.04888517e-01 -1.28219032e+00 8.97569299e-01
9.08839181e-02 -7.74406374e-01 -1.15603760e-01 -5.89132458e-02
3.85813057e-01 6.56993568e-01 -1.16481818e-01 2.86805211e-03
-6.00107340e-03 -7.78870046e-01 1.20953596e+00 4.62018639e-01
5.08831978e-01 -4.15854126e-01 2.67543018e-01 3.58086050e-01
-1.29980671e+00 -5.15007377e-01 -6.70909107e-01 -4.88318115e-01
-1.54308692e-01 1.44487485e-01 -4.45914805e-01 2.11564362e-01
1.04887283e+00 8.72432113e-01 -5.07425010e-01 1.13279092e+00
-7.67389476e-01 -3.15101683e-01 -4.58220571e-01 -1.76590785e-01
4.20147836e-01 -1.15251489e-01 5.10089159e-01 6.53595448e-01
1.31219208e-01 6.76572323e-02 3.60324025e-01 1.23778880e+00
-8.70121121e-02 -3.06399241e-02 -3.25121313e-01 4.85290885e-02
4.61526096e-01 1.37896013e+00 -7.85597801e-01 -1.06865801e-01
-3.95879269e-01 1.21639681e+00 7.52988994e-01 3.97359341e-01
-7.54655659e-01 -5.71766138e-01 1.06808126e+00 -1.39017001e-01
3.27529967e-01 -6.06471241e-01 -2.70581335e-01 -1.32424068e+00
4.57261056e-01 -5.64165056e-01 2.76579171e-01 -1.00133824e+00
-9.90536869e-01 5.91156244e-01 -1.70952842e-01 -1.16199040e+00
2.36849472e-01 -1.16618955e+00 -1.47545516e-01 8.71190310e-01
-1.82950628e+00 -1.41478503e+00 -5.53431213e-01 6.56691670e-01
4.91275489e-01 2.18933657e-01 8.31956089e-01 1.08877465e-01
-4.69172478e-01 9.33066830e-02 -5.56401193e-01 2.18417153e-01
5.31790137e-01 -1.44997513e+00 6.37273371e-01 6.87641203e-01
1.15730606e-01 8.72906685e-01 5.55989087e-01 -3.06643337e-01
-1.77515733e+00 -7.95326710e-01 5.40069520e-01 -6.69827044e-01
7.04517245e-01 -5.65588892e-01 -6.75799787e-01 9.49222624e-01
5.10563441e-02 5.77390134e-01 3.87347579e-01 7.82123655e-02
-6.15139127e-01 4.52046692e-02 -1.20638645e+00 6.98211074e-01
1.57917571e+00 -5.33763587e-01 -7.86092341e-01 5.92453361e-01
1.00152671e+00 -8.28251362e-01 -8.49680245e-01 3.81615758e-01
7.89647460e-01 -9.82810497e-01 9.93586302e-01 -4.80114222e-01
3.55175912e-01 -6.36552989e-01 -4.00011450e-01 -1.02235508e+00
-4.08840925e-01 -3.31820905e-01 -2.74609357e-01 4.68967229e-01
2.56757051e-01 -6.93933785e-01 1.54851362e-01 5.69450676e-01
-2.91275442e-01 -8.72338295e-01 -8.99928033e-01 -6.13775671e-01
-4.13669616e-01 -4.34921265e-01 1.04425824e+00 5.42585909e-01
-1.10179581e-01 3.66044164e-01 -3.18071470e-02 4.61251915e-01
7.92147636e-01 5.59585690e-01 8.12255144e-01 -1.27321053e+00
-2.44331822e-01 -4.49980795e-01 -4.43671137e-01 -1.53993833e+00
2.04931617e-01 -9.14034486e-01 3.53847623e-01 -1.54808652e+00
2.06667006e-01 -6.40422225e-01 -5.48065424e-01 8.77473891e-01
-2.85278022e-01 2.01603860e-01 4.08022016e-01 4.63377267e-01
-6.49116993e-01 7.45079517e-01 1.76051271e+00 -1.94932565e-01
-2.91680127e-01 -1.94830403e-01 -4.89219964e-01 6.59646988e-01
6.46341383e-01 -4.12527204e-01 -2.63461083e-01 -8.59869361e-01
-1.05341241e-01 -3.36292446e-01 9.06522036e-01 -8.72109652e-01
-5.05452231e-02 1.21763526e-02 3.14427733e-01 -3.34871888e-01
4.46152985e-01 -8.78812611e-01 -5.16753495e-01 7.33838022e-01
-4.44414206e-02 -2.59904545e-02 1.76415414e-01 5.71080804e-01
2.71057010e-01 3.76839042e-02 5.11242688e-01 -1.54473186e-01
-1.03535736e+00 7.69949436e-01 1.76141083e-01 -4.47080821e-01
9.11819756e-01 2.40739003e-01 -3.36263061e-01 2.93361872e-01
-6.32644653e-01 3.51431817e-01 4.79344249e-01 9.51093018e-01
7.49435544e-01 -1.39288020e+00 -4.13265795e-01 4.93573695e-01
4.94512826e-01 3.65717173e-01 1.55564085e-01 7.92723894e-01
-6.64608538e-01 1.78282306e-01 -3.80591214e-01 -9.37773347e-01
-7.16012537e-01 6.54382288e-01 4.90558445e-01 4.37899500e-01
-9.43391025e-01 1.10403979e+00 1.71371996e-01 -9.83619809e-01
5.27274072e-01 -9.85993385e-01 5.03952168e-02 -5.09002030e-01
5.62781215e-01 2.39281952e-01 -3.05592835e-01 -7.80021906e-01
-4.71989989e-01 4.63825285e-01 1.75599724e-01 1.05014391e-01
1.39183462e+00 -1.40252516e-01 -3.11122715e-01 6.54943645e-01
9.59945560e-01 -9.73079443e-01 -1.96092761e+00 -2.21699014e-01
6.13030454e-04 -5.86566389e-01 1.71087682e-01 -6.66931748e-01
-7.71972954e-01 8.26000452e-01 5.87604284e-01 5.47362119e-02
1.01684833e+00 2.05726236e-01 7.28760183e-01 7.20003128e-01
6.70699120e-01 -7.31884122e-01 8.72416869e-02 4.83355641e-01
1.40595663e+00 -1.58208454e+00 2.60791313e-02 -4.53785777e-01
-2.93816805e-01 8.37947905e-01 6.84031665e-01 -7.20270157e-01
7.84416854e-01 -3.42289582e-02 1.51566625e-01 -2.43296504e-01
-4.16632533e-01 -4.02940482e-01 4.46273267e-01 3.22714776e-01
2.62284726e-01 3.16002071e-01 5.28790653e-02 4.90791023e-01
-5.48355915e-02 -1.06465630e-01 1.43332989e-03 8.03363979e-01
-4.77063179e-01 -5.31653047e-01 -1.66335955e-01 2.89131761e-01
-2.32652605e-01 9.49457008e-03 -3.92718092e-02 6.86576605e-01
9.81629565e-02 8.29577923e-01 1.19408958e-01 -2.07888737e-01
5.53808331e-01 -1.98928550e-01 9.31128383e-01 -6.69959307e-01
-2.94811308e-01 2.59862989e-01 -2.40491927e-01 -1.24957788e+00
-5.54414570e-01 -7.79251695e-01 -1.61440361e+00 -2.21511386e-02
-3.38998526e-01 -5.63323557e-01 7.76232302e-01 1.07422698e+00
3.84459019e-01 6.48841202e-01 3.34125757e-01 -1.55692554e+00
-7.52755702e-01 -1.44734097e+00 -6.95605099e-01 3.64588916e-01
7.27420330e-01 -1.04815829e+00 -1.61664903e-01 -1.27117321e-01] | [5.161718845367432, -0.11632885783910751] |
105bbcc7-9f29-4178-a1be-f6e109a44035 | gci-a-g-raph-c-oncept-i-nterpretation | 2302.04899 | null | https://arxiv.org/abs/2302.04899v1 | https://arxiv.org/pdf/2302.04899v1.pdf | GCI: A (G)raph (C)oncept (I)nterpretation Framework | Explainable AI (XAI) underwent a recent surge in research on concept extraction, focusing on extracting human-interpretable concepts from Deep Neural Networks. An important challenge facing concept extraction approaches is the difficulty of interpreting and evaluating discovered concepts, especially for complex tasks such as molecular property prediction. We address this challenge by presenting GCI: a (G)raph (C)oncept (I)nterpretation framework, used for quantitatively measuring alignment between concepts discovered from Graph Neural Networks (GNNs) and their corresponding human interpretations. GCI encodes concept interpretations as functions, which can be used to quantitatively measure the alignment between a given interpretation and concept definition. We demonstrate four applications of GCI: (i) quantitatively evaluating concept extractors, (ii) measuring alignment between concept extractors and human interpretations, (iii) measuring the completeness of interpretations with respect to an end task and (iv) a practical application of GCI to molecular property prediction, in which we demonstrate how to use chemical functional groups to explain GNNs trained on molecular property prediction tasks, and implement interpretations with a 0.76 AUCROC completeness score. | ['Pietro Lio', 'Mateja Jamnik', 'Pietro Barbiero', 'Lucie Charlotte Magister', 'Botty Dimanov', 'Dmitry Kazhdan'] | 2023-02-09 | null | null | null | null | ['molecular-property-prediction'] | ['miscellaneous'] | [ 1.14684749e+00 7.87539840e-01 -2.68046021e-01 -3.22298914e-01
-2.40971819e-01 -9.65719044e-01 6.71870291e-01 9.39806223e-01
-4.72551771e-02 1.09923768e+00 2.15169877e-01 -8.24096918e-01
-5.99958062e-01 -9.17676568e-01 -8.65625679e-01 -4.80276704e-01
-5.01693785e-01 8.12273502e-01 -3.85936081e-01 -8.54231566e-02
3.11736405e-01 4.68217045e-01 -1.37809575e+00 5.58824241e-01
9.98422861e-01 9.66708779e-01 -1.50703490e-01 6.60750926e-01
-4.46503842e-03 8.20637763e-01 -8.56654167e-01 -5.25390744e-01
-2.21451908e-01 -6.84079885e-01 -1.26039422e+00 -2.71928281e-01
3.76904666e-01 7.82438815e-02 1.22861415e-01 1.10805583e+00
-1.61929980e-01 1.45416288e-02 1.09035707e+00 -1.55630386e+00
-8.98886144e-01 1.00881422e+00 2.64429972e-02 5.69015220e-02
5.53928256e-01 -4.22740392e-02 1.60765982e+00 -6.66602910e-01
7.81112790e-01 1.23094988e+00 1.95012465e-01 6.27239227e-01
-1.36249208e+00 -5.29808760e-01 1.79134399e-01 2.62770444e-01
-1.11328185e+00 3.64406258e-02 2.74898529e-01 -3.28200221e-01
1.52074468e+00 4.49761987e-01 6.13489807e-01 1.07062125e+00
3.24380279e-01 6.92355514e-01 7.31202424e-01 -2.30527580e-01
7.25835145e-01 -2.49289751e-01 6.17566049e-01 8.04655313e-01
8.20278227e-01 -3.81827392e-02 -8.38370800e-01 -3.89048398e-01
5.06604373e-01 -1.93441045e-02 -4.20645565e-01 1.74510583e-01
-1.29132926e+00 9.43955660e-01 7.31915474e-01 9.19938534e-02
-3.74678195e-01 2.20120311e-01 2.40496784e-01 3.08782846e-01
4.33255166e-01 1.32996070e+00 -8.53077173e-01 2.06198707e-01
-4.93033648e-01 3.38773429e-02 9.01524723e-01 1.06357419e+00
6.20933115e-01 -2.25025993e-02 1.37225017e-01 1.09401181e-01
-6.35814213e-05 2.99361259e-01 3.17192137e-01 -4.10664052e-01
2.29746237e-01 9.31734025e-01 -4.93284047e-01 -8.96070778e-01
-5.93116105e-01 -5.33048809e-01 -9.64072943e-01 -3.11473999e-02
-5.24191596e-02 1.78810675e-02 -1.03106475e+00 1.72338569e+00
-1.85936406e-01 -4.52222154e-02 2.46380031e-01 8.17922354e-01
1.18032432e+00 5.10449409e-01 5.28692365e-01 -1.67714640e-01
1.39251924e+00 -5.43034315e-01 -2.88861871e-01 -8.73559117e-02
7.97194302e-01 -5.55296540e-02 5.89723945e-01 6.04078472e-01
-8.03032875e-01 -4.23733294e-01 -1.52042913e+00 -2.03363508e-01
-7.78957844e-01 -1.83060393e-01 1.29746103e+00 2.78294086e-01
-1.04498446e+00 8.94887984e-01 -4.66086149e-01 -9.19511914e-02
7.50047386e-01 9.89389479e-01 -6.86386824e-01 1.20668396e-01
-1.26452315e+00 6.69555485e-01 9.50864851e-01 -6.75077587e-02
-9.75826204e-01 -8.72018337e-01 -1.01871359e+00 6.13938570e-01
6.67712748e-01 -1.04087615e+00 7.70936668e-01 -1.09876394e+00
-8.23525548e-01 4.40913558e-01 -8.05427805e-02 -6.86440527e-01
-1.13232598e-01 2.38954015e-02 -4.32381481e-01 3.33041340e-01
1.42483562e-01 9.58648801e-01 4.48280036e-01 -1.17129314e+00
-6.64251029e-01 -4.42042142e-01 4.27859485e-01 -6.54902160e-02
-2.72132427e-01 -5.78829050e-01 1.95678324e-01 -3.39437664e-01
2.48764604e-01 -7.23015070e-01 -2.00485736e-01 -2.00845450e-01
-1.04449832e+00 -2.95209587e-01 5.69881260e-01 -4.41751122e-01
9.00609314e-01 -1.45389915e+00 4.86782372e-01 6.98367059e-01
1.24072742e+00 1.83565304e-01 -3.98295820e-01 5.14253259e-01
-7.83373356e-01 8.10459495e-01 -4.98332232e-01 2.18033433e-01
-1.28423452e-01 2.28467673e-01 -4.49218661e-01 -4.03858237e-02
7.10269332e-01 1.28164005e+00 -1.29026103e+00 5.70497736e-02
-9.74333212e-02 3.10035974e-01 -5.24568796e-01 -1.00923842e-02
-8.14461946e-01 3.12548608e-01 -4.28845525e-01 7.90538847e-01
2.90670037e-01 -6.65918231e-01 6.65371299e-01 -2.02167317e-01
4.72449780e-01 4.14545506e-01 -3.52489859e-01 1.29515445e+00
7.64836594e-02 8.55395794e-01 -7.20618427e-01 -9.39736724e-01
9.07686651e-01 3.95868748e-01 3.60938996e-01 -4.89980727e-01
-1.50702327e-01 8.07884783e-02 2.06633940e-01 1.44003835e-02
4.92071420e-01 -2.13034511e-01 2.54283518e-01 5.19715667e-01
3.33297014e-01 2.72480488e-01 2.88071334e-01 6.26743317e-01
1.43135095e+00 -6.83202548e-03 9.29565370e-01 -4.45109218e-01
6.12763941e-01 2.30164275e-01 1.17869943e-01 8.02783430e-01
4.27721798e-01 3.44962150e-01 1.16966915e+00 -8.59920442e-01
-9.26282942e-01 -8.09601486e-01 2.07551360e-01 9.65253592e-01
-3.40485610e-02 -8.10230196e-01 -6.80933356e-01 -9.08816755e-01
5.55155575e-02 7.14720011e-01 -8.21544051e-01 -3.49299729e-01
-9.97597203e-02 -6.43499434e-01 6.60197198e-01 6.53832912e-01
2.37333059e-01 -1.07335639e+00 -2.77512908e-01 8.23733211e-02
-2.74991006e-01 -1.00709438e+00 4.57298458e-02 5.72354794e-01
-9.13595498e-01 -1.38140678e+00 -3.06247193e-02 -5.60263455e-01
1.10122907e+00 1.43039599e-01 1.45952356e+00 6.16138518e-01
-2.63335466e-01 1.69993594e-01 -2.45497212e-01 -7.21436560e-01
-5.54117918e-01 3.70595723e-01 4.42507491e-02 -4.64477897e-01
4.52837288e-01 -6.78141356e-01 -5.68679929e-01 1.77178442e-01
-1.42253482e+00 3.87422800e-01 6.78578436e-01 7.08637357e-01
6.56132460e-01 -1.19005062e-01 6.41035259e-01 -1.28281415e+00
9.50109363e-01 -4.14218247e-01 -3.97138894e-01 6.31359816e-01
-9.05035436e-01 5.13776958e-01 7.56475091e-01 -1.43887788e-01
-3.82274359e-01 -3.01762428e-02 2.32367986e-03 1.63193747e-01
-1.20284885e-01 1.20577323e+00 -2.80520201e-01 1.45993546e-01
1.00036168e+00 -8.06166530e-02 -1.91657320e-01 6.15528673e-02
5.94245911e-01 3.31219494e-01 6.12661242e-01 -6.90349102e-01
6.00936651e-01 2.25310177e-01 6.58223391e-01 -7.46268034e-01
-8.92592847e-01 -2.30894893e-01 -9.05736148e-01 1.61199868e-01
8.96760046e-01 -6.17082417e-01 -1.40530145e+00 -4.10321206e-01
-1.32590008e+00 8.83283913e-02 1.01411594e-02 2.09674567e-01
-4.45587993e-01 4.85837638e-01 -3.45532328e-01 -5.44571161e-01
-8.83093476e-01 -9.30997908e-01 1.14206111e+00 1.21194236e-01
-6.31658852e-01 -1.24638402e+00 -1.88474804e-01 2.74527878e-01
-8.15959051e-02 6.92390859e-01 1.77727497e+00 -1.16790533e+00
-6.69008493e-01 -2.27084979e-02 -4.17474777e-01 4.28455658e-02
2.69832820e-01 -2.24661604e-01 -1.09605265e+00 -1.55473515e-01
-5.86704671e-01 -2.47649312e-01 8.68560910e-01 1.36928856e-01
1.25972545e+00 -6.51721954e-01 -4.29554731e-01 4.11274999e-01
1.38562036e+00 1.27984375e-01 7.57420301e-01 1.23123089e-02
7.29163349e-01 3.84924233e-01 2.59278953e-01 -4.38140221e-02
-1.77337348e-01 3.26506913e-01 5.46517134e-01 -2.07366303e-01
1.35700181e-01 -5.74135959e-01 9.13990960e-02 4.40374553e-01
-6.54504955e-01 -7.06003904e-01 -9.96060431e-01 8.89729038e-02
-2.02691650e+00 -8.60967457e-01 -1.80300966e-01 1.98104882e+00
6.92416787e-01 2.15208143e-01 -1.17056683e-01 8.28520805e-02
2.97838807e-01 -4.26561713e-01 -7.31578708e-01 -4.63554054e-01
-1.26657769e-01 6.51618719e-01 2.36492619e-01 4.22391325e-01
-7.78439462e-01 9.36957061e-01 6.07790852e+00 6.60032570e-01
-8.44389439e-01 -3.36874366e-01 8.89145136e-01 4.70981956e-01
-6.95012689e-01 2.55586922e-01 -3.17927778e-01 -2.20930064e-03
1.17355967e+00 -3.73752266e-01 3.90094995e-01 8.90077412e-01
-1.70894355e-01 2.42283776e-01 -1.74512041e+00 6.94842756e-01
7.30980188e-02 -1.77382708e+00 9.99983966e-01 3.52860779e-01
4.83010650e-01 -4.43986446e-01 -2.30931982e-01 1.79035649e-01
1.02990299e-01 -1.66529238e+00 2.94762939e-01 4.75492567e-01
9.72229660e-01 -8.12116802e-01 7.95108914e-01 -1.39104337e-01
-1.15779650e+00 1.33405328e-01 -6.75369143e-01 -3.10611367e-01
-4.00262356e-01 4.68519926e-01 -1.42942345e+00 9.48238194e-01
4.81724367e-02 9.14964557e-01 -6.92865252e-01 6.20919168e-01
-6.85156703e-01 6.29367650e-01 9.77862477e-02 -4.06610489e-01
3.28941822e-01 -3.18059027e-01 2.57025659e-01 1.11734045e+00
-9.90312621e-02 3.32403421e-01 -8.60798433e-02 1.31000280e+00
-3.58646452e-01 -2.38888860e-01 -5.92207313e-01 -7.89860785e-01
1.10261120e-01 1.25412798e+00 -1.07066655e+00 -6.58124208e-01
4.82517630e-02 9.47188020e-01 3.48692596e-01 5.28134406e-01
-8.22504282e-01 -3.01663250e-01 8.11935425e-01 -7.24260062e-02
-3.56182717e-02 -2.12633356e-01 -3.84683430e-01 -1.02705133e+00
-2.41268888e-01 -9.69981253e-01 4.88841534e-01 -1.07459450e+00
-1.22003090e+00 7.76639700e-01 1.16849452e-01 -9.75690424e-01
-1.71350539e-01 -1.21084332e+00 -3.53628069e-01 7.28195190e-01
-1.16329932e+00 -9.70057964e-01 -3.33633989e-01 5.76893389e-02
3.60827863e-01 -1.26709223e-01 1.22042608e+00 -1.75915226e-01
-2.64059007e-01 4.12703335e-01 -3.95483851e-01 2.59295888e-02
-2.39719376e-02 -1.46105397e+00 7.99144626e-01 3.80901635e-01
5.64704776e-01 1.22803259e+00 5.69122076e-01 -7.61052489e-01
-1.23521626e+00 -1.17145872e+00 8.85896087e-01 -7.32871473e-01
4.95299250e-01 -2.97711074e-01 -9.54812527e-01 7.03032792e-01
-2.28610814e-01 -5.92066944e-01 1.25441742e+00 3.44894022e-01
-4.94095206e-01 4.10774887e-01 -8.30292523e-01 7.58504391e-01
1.46032643e+00 -5.94331145e-01 -4.16081339e-01 7.90136278e-01
1.10496593e+00 -1.38821304e-01 -1.03674495e+00 4.35789585e-01
5.60963154e-01 -4.94070292e-01 9.87349451e-01 -1.52951825e+00
1.01378047e+00 -4.13478941e-01 1.04804695e-01 -1.08950567e+00
-3.09402257e-01 -5.88874161e-01 -2.40975365e-01 3.71382475e-01
1.20828593e+00 -5.97817779e-01 7.19155252e-01 8.47110033e-01
-7.67986551e-02 -9.20948923e-01 -5.39273024e-01 -5.04896760e-01
-2.53857374e-01 -2.41860822e-01 7.86898136e-01 9.12725627e-01
3.59069318e-01 1.05683184e+00 -1.28702164e-01 2.00753227e-01
4.10484284e-01 7.57690594e-02 5.21403849e-01 -1.56408894e+00
-4.45716858e-01 -3.29742372e-01 -9.84074235e-01 -8.65770459e-01
6.33108914e-02 -1.31577754e+00 -2.60526210e-01 -1.66881967e+00
6.11283839e-01 3.82140458e-01 -3.83969426e-01 1.01776528e+00
-2.40955248e-01 1.58813726e-02 -4.86043505e-02 2.90860951e-01
-6.51121855e-01 3.26927930e-01 1.06007206e+00 -4.71634448e-01
5.07082744e-03 -3.32188398e-01 -1.17216241e+00 4.40290242e-01
6.18140697e-01 -4.91142392e-01 -8.80798757e-01 -1.23338893e-01
8.38205457e-01 -1.15774468e-01 5.25063038e-01 -8.83436322e-01
1.41036093e-01 -1.89389095e-01 5.71487486e-01 -3.37052792e-01
1.18057709e-02 -5.24721980e-01 2.91868776e-01 6.29762232e-01
-6.58359170e-01 2.01366231e-01 2.15742752e-01 9.11717474e-01
-1.53379738e-01 -2.11188704e-01 7.79030249e-02 -2.37893522e-01
-7.43903220e-01 4.45420504e-01 -1.91142514e-01 -2.72877008e-01
4.36675817e-01 -2.39018351e-01 -8.96622241e-01 -5.79751849e-01
-8.29003870e-01 6.66588470e-02 1.93474516e-01 3.33668888e-01
1.17464054e+00 -8.94331098e-01 -4.19735521e-01 1.54258013e-02
4.51289326e-01 -1.72299832e-01 -4.02053952e-01 3.65590155e-01
-7.80394733e-01 7.26055741e-01 -1.33088171e-01 -5.08807540e-01
-1.39615393e+00 5.51346421e-01 2.43452474e-01 -2.78495222e-01
-3.79621148e-01 6.75430059e-01 6.11536205e-01 -2.09593222e-01
-4.71007489e-02 -6.42904162e-01 -2.11989284e-01 -3.84233177e-01
5.58030605e-01 -4.01816405e-02 6.47746027e-02 -4.92356829e-02
-4.97428149e-01 3.84724699e-02 -2.93744862e-01 1.84680313e-01
1.54632056e+00 6.77184880e-01 -4.28787112e-01 -1.36330619e-01
9.39889073e-01 -5.99488378e-01 -5.96211910e-01 5.97611479e-02
4.24855202e-01 1.30025357e-01 -3.91354531e-01 -1.18531609e+00
-5.50331533e-01 7.35483408e-01 6.85393214e-02 1.70791209e-01
9.86920655e-01 7.61535019e-02 4.36800808e-01 1.06078529e+00
3.21476996e-01 -7.14427531e-01 2.40047351e-01 3.62404734e-01
9.07863677e-01 -9.47709322e-01 5.96133471e-01 -7.65444636e-01
-5.67279935e-01 1.62919533e+00 2.85073102e-01 4.32341009e-01
2.29540184e-01 -3.76029640e-01 -4.87454206e-01 -9.56057072e-01
-7.57869363e-01 -3.04072350e-01 7.51762271e-01 6.47240818e-01
6.43809557e-01 3.62502933e-01 -2.71790773e-01 6.22516215e-01
-3.80307943e-01 -1.85553059e-01 5.75308263e-01 7.60491371e-01
-5.28269529e-01 -9.99273539e-01 2.47426420e-01 6.79918587e-01
-9.12489146e-02 -6.15751743e-01 -1.27338254e+00 6.92899764e-01
4.08145376e-02 1.02905381e+00 -4.57540065e-01 -6.29563093e-01
3.38388458e-02 7.59960059e-03 1.76328465e-01 -9.42806959e-01
-3.68389368e-01 -5.40620208e-01 3.97995353e-01 -4.04866338e-01
-3.90127808e-01 -1.56300794e-02 -1.58676279e+00 -3.33892435e-01
-3.23367685e-01 3.79089236e-01 6.79426134e-01 1.13502371e+00
5.69694579e-01 6.24004841e-01 1.90638360e-02 -3.30169052e-01
7.35730156e-02 -7.04365313e-01 -3.95132571e-01 6.14621460e-01
1.12288013e-01 -3.69471759e-01 -3.18905890e-01 9.41894352e-02] | [7.762241363525391, 6.195655345916748] |
dbf5f118-dca3-4691-ac2e-dfa7416689a5 | scribble-supervised-lidar-semantic | 2203.08537 | null | https://arxiv.org/abs/2203.08537v2 | https://arxiv.org/pdf/2203.08537v2.pdf | Scribble-Supervised LiDAR Semantic Segmentation | Densely annotating LiDAR point clouds remains too expensive and time-consuming to keep up with the ever growing volume of data. While current literature focuses on fully-supervised performance, developing efficient methods that take advantage of realistic weak supervision have yet to be explored. In this paper, we propose using scribbles to annotate LiDAR point clouds and release ScribbleKITTI, the first scribble-annotated dataset for LiDAR semantic segmentation. Furthermore, we present a pipeline to reduce the performance gap that arises when using such weak annotations. Our pipeline comprises of three stand-alone contributions that can be combined with any LiDAR semantic segmentation model to achieve up to 95.7% of the fully-supervised performance while using only 8% labeled points. Our scribble annotations and code are available at github.com/ouenal/scribblekitti. | ['Luc van Gool', 'Dengxin Dai', 'Ozan Unal'] | 2022-03-16 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Unal_Scribble-Supervised_LiDAR_Semantic_Segmentation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Unal_Scribble-Supervised_LiDAR_Semantic_Segmentation_CVPR_2022_paper.pdf | cvpr-2022-1 | ['lidar-semantic-segmentation'] | ['computer-vision'] | [ 2.11618856e-01 2.95033365e-01 -2.34889880e-01 -6.61055803e-01
-1.00860322e+00 -8.32303286e-01 3.04298759e-01 4.13431704e-01
-3.89653444e-01 5.93042195e-01 -3.21058661e-01 -3.09415370e-01
2.40787142e-03 -6.71141565e-01 -9.85725105e-01 -9.88627821e-02
2.08116278e-01 1.16156125e+00 7.71298110e-01 2.27289468e-01
6.15804456e-02 5.66103935e-01 -1.76659286e+00 1.38464764e-01
9.79071379e-01 7.74461865e-01 2.57105321e-01 4.76498991e-01
-4.99661833e-01 1.19597279e-01 -2.50969142e-01 -3.15990746e-01
5.00482202e-01 1.56285077e-01 -8.47268581e-01 -1.91602558e-01
1.03097081e+00 -1.53144732e-01 3.62828761e-01 7.95167923e-01
3.29613566e-01 -2.74585217e-01 4.36151356e-01 -1.47026432e+00
-1.33182555e-01 6.57091558e-01 -5.28382480e-01 -2.42178347e-02
3.28316949e-02 -5.44431619e-02 1.05281007e+00 -1.12992680e+00
5.89609027e-01 1.04016638e+00 9.08654511e-01 2.89123416e-01
-1.22747076e+00 -9.88584220e-01 9.55676958e-02 -1.81060042e-02
-1.64406252e+00 -2.89028376e-01 5.19252837e-01 -6.52922392e-01
8.65368187e-01 1.13716602e-01 6.91013515e-01 5.35605550e-01
-4.59643990e-01 7.96656311e-01 1.06523752e+00 -3.24106544e-01
1.25692502e-01 1.57890901e-01 5.08495510e-01 8.93806517e-01
6.35964453e-01 -4.98757176e-02 -4.48245704e-01 -1.26973853e-01
4.96237367e-01 -3.38310599e-01 2.24197045e-01 -7.18069255e-01
-9.69700217e-01 7.72309065e-01 5.24901748e-01 -6.28617033e-02
1.12830549e-01 3.24247211e-01 2.75549233e-01 -1.99217677e-01
6.66197717e-01 2.14378864e-01 -6.90088868e-01 -3.21142077e-01
-1.29976392e+00 2.16964632e-01 6.62774980e-01 1.30126476e+00
1.32407546e+00 -3.45325351e-01 5.80937266e-01 7.35188842e-01
2.86692411e-01 5.48155189e-01 -3.04353863e-01 -1.35272777e+00
2.91656852e-01 7.39819586e-01 1.06684595e-01 -4.61401403e-01
-3.38377982e-01 -4.81899194e-02 -4.47369106e-02 3.81136566e-01
4.51072007e-01 -6.02068603e-02 -1.00300562e+00 1.01609325e+00
4.83332515e-01 2.84817994e-01 -2.44282603e-01 7.57367551e-01
9.92067337e-01 3.05316836e-01 8.80567133e-02 3.83952975e-01
1.24446118e+00 -1.00526285e+00 -2.56264031e-01 -4.10135925e-01
6.87191010e-01 -8.84656131e-01 1.23513877e+00 2.97988266e-01
-9.19224858e-01 -4.49496299e-01 -1.15314353e+00 -4.26713377e-01
-6.07610881e-01 -1.25515968e-01 7.23622680e-01 6.24509752e-01
-1.10281241e+00 4.04354334e-01 -1.14986348e+00 -5.64183891e-01
8.81550550e-01 5.80716848e-01 -2.20758617e-01 -9.29955542e-02
-5.54560959e-01 9.11776662e-01 4.96107757e-01 -2.99277812e-01
-5.30311465e-01 -1.08507574e+00 -7.72942781e-01 -3.12419951e-01
6.59926057e-01 -5.18088222e-01 1.40817249e+00 -4.56434667e-01
-9.30906892e-01 1.09003866e+00 -2.83813417e-01 -3.38621467e-01
5.82631946e-01 -4.72791761e-01 2.66179234e-01 3.40945870e-02
3.81937444e-01 1.42811072e+00 2.72030085e-01 -1.60219133e+00
-8.14089537e-01 -3.67793322e-01 -1.16089717e-01 5.61969914e-03
-1.49217948e-01 -7.14212656e-02 -5.04707575e-01 -1.75976679e-01
1.35559529e-01 -1.19335294e+00 -1.28493980e-01 2.55309254e-01
-2.20916480e-01 -3.59251529e-01 1.38198173e+00 -2.22210124e-01
8.14061105e-01 -2.03855944e+00 -2.94673979e-01 1.42695019e-02
-2.11339667e-02 4.31061536e-01 -3.67565304e-02 4.68486935e-01
-2.10481957e-02 3.63529563e-01 -8.03625226e-01 -7.98435390e-01
5.68485484e-02 5.88993371e-01 -3.29734594e-01 1.22663140e-01
3.31042022e-01 1.03857422e+00 -1.11817110e+00 -8.26196313e-01
5.86308837e-01 3.97507459e-01 -5.62357843e-01 -9.75516364e-02
-4.29780632e-01 4.49896753e-01 -3.23474854e-01 9.77092743e-01
7.68594265e-01 -6.47022799e-02 -2.10085779e-01 -1.96378082e-01
-3.34562689e-01 2.90112138e-01 -1.12273109e+00 2.03143620e+00
-3.11090559e-01 4.72808748e-01 2.00727627e-01 -5.73875248e-01
8.47625732e-01 -7.17197731e-02 7.73876667e-01 -5.65521717e-02
-4.20676842e-02 6.13879859e-01 -3.11491013e-01 -1.17414325e-01
5.20681679e-01 -7.58539885e-02 -1.21407710e-01 4.05878037e-01
3.13893817e-02 -8.33097279e-01 1.50985196e-01 1.95932075e-01
9.32497919e-01 6.28059864e-01 1.93678454e-01 -2.56044030e-01
2.19795972e-01 6.40578926e-01 3.72435927e-01 4.12743002e-01
-1.27608836e-01 8.88671100e-01 1.93633735e-02 -3.39271247e-01
-1.05477858e+00 -1.14189184e+00 -5.78665614e-01 8.42008650e-01
1.02991059e-01 -5.49890995e-01 -7.91847110e-01 -8.00106883e-01
3.23448867e-01 7.58927763e-01 -6.60384540e-03 4.52359676e-01
-4.58668202e-01 -3.11749995e-01 7.56451964e-01 6.51358545e-01
4.10906285e-01 -9.60289419e-01 -8.40296447e-01 -8.32567960e-02
5.39641418e-02 -1.42472911e+00 -7.88739547e-02 3.15621674e-01
-9.56845105e-01 -1.09038353e+00 -1.09311067e-01 -6.11112654e-01
5.45944273e-01 2.91618645e-01 1.35007691e+00 2.66825169e-01
-3.95128340e-01 5.13342202e-01 -3.29238057e-01 -8.82464707e-01
-2.49427091e-02 5.44345558e-01 -2.98226804e-01 -7.06789315e-01
6.97030306e-01 -7.22530127e-01 -3.43539387e-01 1.98025227e-01
-7.61404157e-01 1.57326639e-01 2.34834820e-01 2.30393589e-01
8.37355435e-01 -2.46778309e-01 3.88882786e-01 -1.19260144e+00
1.00100972e-02 -2.78730899e-01 -8.72978330e-01 -1.91699371e-01
-6.18011892e-01 -2.79619277e-01 7.09550455e-02 1.45767599e-01
-6.74468100e-01 6.83337092e-01 -3.59960973e-01 -5.61581850e-01
-5.45258880e-01 3.76257747e-01 -2.21414939e-02 -1.99689075e-01
3.43591422e-01 -4.58155781e-01 -2.07100615e-01 -6.63911402e-01
7.65820503e-01 8.18418503e-01 6.40097499e-01 -8.02900016e-01
9.38013554e-01 7.77356982e-01 -3.90827842e-02 -7.63517380e-01
-1.17999566e+00 -7.59767950e-01 -1.31705797e+00 -1.23535462e-01
8.39547992e-01 -9.35812473e-01 -3.45817000e-01 9.10369456e-02
-1.09959435e+00 -4.35143054e-01 -5.49408853e-01 6.36753216e-02
-4.99819219e-01 5.04860759e-01 -1.98376253e-01 -7.57422149e-01
-2.19303504e-01 -1.13325191e+00 1.53953230e+00 -8.57547969e-02
-3.75895739e-01 -7.51889110e-01 1.06241159e-01 8.25038195e-01
-1.65330857e-01 2.87830323e-01 6.26261532e-01 -6.33557320e-01
-9.43811834e-01 -2.28216067e-01 -4.63860333e-01 3.20955366e-01
-7.12635517e-02 2.08749220e-01 -1.20902050e+00 -6.19471595e-02
-4.38059419e-01 -4.17373568e-01 7.03345060e-01 4.13524173e-02
1.23828614e+00 2.18026519e-01 -4.98714000e-01 6.02991402e-01
1.36308432e+00 -1.39679626e-01 3.36164057e-01 3.23253214e-01
1.03644145e+00 6.63219512e-01 8.68943572e-01 1.36479780e-01
7.73971558e-01 6.48568928e-01 6.75012410e-01 -2.12439179e-01
-2.49120891e-01 -4.69563961e-01 -6.28168061e-02 8.19643199e-01
9.16912314e-03 -2.31550112e-02 -1.47313368e+00 8.00553083e-01
-1.84499717e+00 -4.79959577e-01 -7.21222103e-01 2.11087632e+00
7.96091855e-01 2.23014131e-01 5.81143834e-02 1.24154858e-01
4.03401971e-01 -3.12955417e-02 -4.92203027e-01 -2.88417190e-01
2.34826744e-01 6.66969717e-01 7.89642096e-01 7.84404278e-01
-1.14025664e+00 1.44645965e+00 6.13497353e+00 6.98484004e-01
-8.57047141e-01 4.59898561e-01 8.83348063e-02 -1.47040129e-01
-2.09248573e-01 4.66080815e-01 -1.05427635e+00 2.98275381e-01
9.53006089e-01 2.01544404e-01 6.76613748e-02 9.32835937e-01
2.63364017e-01 -3.43120068e-01 -1.06750274e+00 8.26414466e-01
-9.69225839e-02 -1.10751617e+00 -2.70490557e-01 1.41125843e-01
7.81750679e-01 6.15817726e-01 -2.64031053e-01 7.13406354e-02
7.44246483e-01 -9.67304587e-01 7.34961271e-01 2.22519070e-01
8.23924541e-01 -6.67345345e-01 5.89524686e-01 6.06582105e-01
-1.42406452e+00 3.62680554e-01 -3.97105277e-01 -1.27660975e-01
3.19907427e-01 9.05558646e-01 -1.07815373e+00 6.16752863e-01
8.52651000e-01 8.03676248e-01 -7.75906265e-01 1.08324623e+00
-3.59229863e-01 7.88092017e-01 -7.40458190e-01 4.06792372e-01
1.91163346e-01 -4.12480742e-01 2.38062114e-01 1.36721802e+00
2.45851547e-01 1.34382251e-04 6.72112763e-01 7.92995632e-01
-5.93230128e-02 1.93998776e-02 -7.11595476e-01 8.22842941e-02
8.17209542e-01 1.40000904e+00 -9.66055751e-01 -3.63973320e-01
-4.25265908e-01 7.97797084e-01 4.93816257e-01 -4.60225455e-02
-8.29615831e-01 5.41342050e-02 6.74182653e-01 4.40462589e-01
2.56422281e-01 -5.77015162e-01 -9.26401913e-01 -8.74764144e-01
3.93195227e-02 -2.73932099e-01 3.41971040e-01 -1.14161313e+00
-1.24133003e+00 2.72084147e-01 5.09083092e-01 -9.23219085e-01
1.82845727e-01 -5.33744812e-01 -2.03223050e-01 7.44428813e-01
-1.81978655e+00 -1.69428504e+00 -7.09925473e-01 1.30182162e-01
6.46629095e-01 4.65003401e-01 8.23358774e-01 4.69896853e-01
-3.66039991e-01 1.28580809e-01 -4.32798713e-01 -6.15582727e-02
7.39618242e-01 -1.42408955e+00 5.37107706e-01 5.96786559e-01
4.01136309e-01 5.47018588e-01 3.37793231e-01 -8.22626770e-01
-8.84822309e-01 -1.20484424e+00 8.79296422e-01 -1.10781407e+00
7.47555196e-01 -6.44330084e-01 -1.03291440e+00 9.83547390e-01
6.42876104e-02 3.40407073e-01 5.99776983e-01 1.58113927e-01
-4.02224213e-01 -2.30682380e-02 -1.03503430e+00 2.52115756e-01
1.33487153e+00 -4.78110224e-01 -6.75769269e-01 3.90667319e-01
6.25294030e-01 -6.13749206e-01 -9.60182130e-01 7.67671585e-01
3.09351891e-01 -1.01214135e+00 1.08493078e+00 2.52892785e-02
2.83002317e-01 -6.71890974e-01 -3.49159129e-02 -7.62658179e-01
2.60953665e-01 -1.69139370e-01 2.06600979e-01 1.41734993e+00
5.20411313e-01 -5.39944649e-01 1.25489604e+00 5.16836822e-01
-5.25533915e-01 -6.50359094e-01 -1.00465119e+00 -7.55022168e-01
2.29734153e-01 -7.48157263e-01 4.60004091e-01 1.02716267e+00
-1.75234914e-01 3.21673065e-01 1.62206948e-01 2.39772499e-01
6.62799478e-01 1.67354435e-01 1.10009849e+00 -1.68292582e+00
3.21785986e-01 -2.15645015e-01 -3.31709206e-01 -9.78657842e-01
2.65594125e-01 -1.36457360e+00 2.55944043e-01 -1.91538334e+00
1.25762552e-01 -9.69621301e-01 2.51270115e-01 1.01315355e+00
-6.18765764e-02 6.99708104e-01 2.73962319e-01 4.67127949e-01
-4.49839383e-01 2.64382184e-01 6.88610673e-01 8.96987990e-02
1.12331830e-01 -1.48409128e-01 -4.28583384e-01 1.02270329e+00
8.76571238e-01 -6.34885609e-01 -4.59645897e-01 -6.91546202e-01
-3.99621651e-02 -6.07859313e-01 4.71748620e-01 -1.17959034e+00
1.63575828e-01 4.47187871e-02 -9.86712202e-02 -1.26916325e+00
5.55479825e-01 -1.07366610e+00 2.96274841e-01 1.36509374e-01
8.99098590e-02 1.31319597e-01 4.19242591e-01 3.60285431e-01
-9.67774838e-02 -5.47902822e-01 8.16508114e-01 -2.66702622e-01
-7.56176293e-01 3.42650682e-01 1.46187097e-03 3.18262801e-02
1.17852390e+00 -5.26771784e-01 -1.10142298e-01 1.75893396e-01
-5.70039392e-01 3.86942387e-01 1.00325644e+00 2.96933472e-01
3.03920031e-01 -9.12704825e-01 -4.71437305e-01 -1.06635131e-01
1.97855830e-01 8.43516827e-01 -2.85637200e-01 4.01499987e-01
-8.57333839e-01 5.27403891e-01 5.39062433e-02 -9.40495074e-01
-1.46590889e+00 2.80295521e-01 5.57079762e-02 -9.36371908e-02
-6.82772398e-01 8.42449605e-01 -1.66595772e-01 -9.34372902e-01
7.15399459e-02 -4.12569225e-01 1.64999232e-01 9.85711366e-02
-2.78292418e-01 3.53965163e-01 3.64956737e-01 -5.80230653e-01
-5.62108159e-01 9.25744116e-01 2.97676951e-01 -1.33581787e-01
1.22523069e+00 6.99509978e-02 -1.99671805e-01 6.83660030e-01
8.68169129e-01 9.71894637e-02 -1.23302865e+00 -4.61543314e-02
5.47997236e-01 -5.25165975e-01 9.21268016e-02 -8.64211440e-01
-8.71025681e-01 1.07507050e+00 4.68278110e-01 -4.03135307e-02
7.55811155e-01 3.06834310e-01 8.30004692e-01 2.29903430e-01
7.49618173e-01 -8.37261438e-01 -1.68847397e-01 5.34888327e-01
6.05337203e-01 -1.40976858e+00 3.60299200e-01 -1.11236548e+00
-3.94250065e-01 7.56822586e-01 7.48961449e-01 -1.37727812e-01
6.98532999e-01 6.07600510e-01 2.61305541e-01 -4.15891737e-01
-3.92090440e-01 -6.51905835e-01 9.38972235e-02 7.84098506e-01
4.61523622e-01 1.90430414e-02 -3.00192356e-01 1.46069065e-01
-5.35546064e-01 1.76635072e-01 3.08466882e-01 1.08385909e+00
-6.62130833e-01 -1.52002847e+00 -3.87722790e-01 3.22694391e-01
-5.85297272e-02 -2.32494369e-01 -7.16220021e-01 8.92187774e-01
5.06964147e-01 7.93533325e-01 2.35186249e-01 -1.53429627e-01
4.82460767e-01 3.21978360e-01 2.60420829e-01 -1.28903604e+00
-4.14323807e-01 -3.37715745e-02 2.65622318e-01 -4.55075413e-01
-7.17261016e-01 -9.24732208e-01 -1.62630510e+00 -1.50789574e-01
-3.45897973e-01 9.46232080e-02 8.42663050e-01 7.75658488e-01
4.40718085e-01 1.92627385e-01 -7.12272301e-02 -1.07459843e+00
-4.27022837e-02 -7.82812774e-01 -1.87814068e-02 1.13389850e-01
-8.37946609e-02 -7.63117433e-01 -1.41427457e-01 1.63052320e-01] | [8.049899101257324, -3.023550510406494] |
b07139c6-ecc7-4e58-be40-f9f9dc7a9d57 | using-supervised-bigram-based-ilp-for | null | null | https://aclanthology.org/P13-1099 | https://aclanthology.org/P13-1099.pdf | Using Supervised Bigram-based ILP for Extractive Summarization | null | ['Chen Li', 'Yang Liu', 'Xian Qian'] | 2013-08-01 | null | null | null | acl-2013-8 | ['extractive-document-summarization'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.353213787078857, 3.7914838790893555] |
5ddb14b9-cf87-46fd-9b29-f5aaa6f000db | graph-attention-with-hierarchies-for-multi | 2301.11792 | null | https://arxiv.org/abs/2301.11792v1 | https://arxiv.org/pdf/2301.11792v1.pdf | Graph Attention with Hierarchies for Multi-hop Question Answering | Multi-hop QA (Question Answering) is the task of finding the answer to a question across multiple documents. In recent years, a number of Deep Learning-based approaches have been proposed to tackle this complex task, as well as a few standard benchmarks to assess models Multi-hop QA capabilities. In this paper, we focus on the well-established HotpotQA benchmark dataset, which requires models to perform answer span extraction as well as support sentence prediction. We present two extensions to the SOTA Graph Neural Network (GNN) based model for HotpotQA, Hierarchical Graph Network (HGN): (i) we complete the original hierarchical structure by introducing new edges between the query and context sentence nodes; (ii) in the graph propagation step, we propose a novel extension to Hierarchical Graph Attention Network GATH (Graph ATtention with Hierarchies) that makes use of the graph hierarchy to update the node representations in a sequential fashion. Experiments on HotpotQA demonstrate the efficiency of the proposed modifications and support our assumptions about the effects of model related variables. | ['Pontus Stenetorp', 'Ieva Staliunaite', 'Philip John Gorinski', 'Yunjie He'] | 2023-01-27 | null | null | null | null | ['multi-hop-question-answering'] | ['knowledge-base'] | [ 9.88087356e-02 6.49305105e-01 2.46886969e-01 -4.38363463e-01
-9.13232744e-01 -5.67950964e-01 4.70417351e-01 6.31877065e-01
-1.85502648e-01 6.21442020e-01 4.15537030e-01 -6.23969257e-01
-3.10385436e-01 -1.09205830e+00 -7.01501548e-01 -2.75249451e-01
3.12839337e-02 9.59773183e-01 6.71873271e-01 -6.58325613e-01
2.33669728e-01 1.44172162e-01 -8.66471767e-01 3.44860822e-01
1.08956552e+00 8.81123543e-01 -1.35524422e-01 1.13344145e+00
-2.88920611e-01 1.53124881e+00 -6.33653402e-01 -9.27946985e-01
-8.47015977e-02 -4.42332059e-01 -1.60521650e+00 -2.54567623e-01
7.00699151e-01 -3.40112895e-01 -5.25782287e-01 7.76861131e-01
4.33783054e-01 3.44108582e-01 3.71606082e-01 -1.25000644e+00
-9.60602462e-01 8.88045311e-01 -4.35875952e-01 3.63695353e-01
3.35820675e-01 9.37675461e-02 1.78784108e+00 -4.56476182e-01
6.55488670e-01 1.41532898e+00 6.65000319e-01 4.10296321e-01
-8.47052097e-01 -1.08785503e-01 2.32238263e-01 7.57241309e-01
-9.03742671e-01 9.80460830e-03 8.47184241e-01 -1.33979648e-01
1.39220786e+00 2.22762913e-01 3.24265808e-01 6.21664524e-01
3.85268182e-01 6.51374519e-01 7.08691299e-01 -4.67800856e-01
3.59201282e-02 -3.95574540e-01 9.81955886e-01 1.18730998e+00
-1.40230820e-01 -6.12926781e-01 -2.14384541e-01 -3.90318811e-01
1.38032645e-01 -5.90850413e-01 -2.71802068e-01 -3.23533535e-01
-8.25386763e-01 1.39423847e+00 9.59519267e-01 4.15867507e-01
-3.24826956e-01 3.93747538e-01 5.34276187e-01 6.66954041e-01
3.17992806e-01 7.25067496e-01 -4.59122002e-01 2.77967751e-01
-3.65148723e-01 3.83639127e-01 1.07730258e+00 8.54130745e-01
7.65689611e-01 -2.38559365e-01 -7.75305212e-01 6.00751400e-01
2.68271238e-01 -3.35961655e-02 1.63613856e-01 -8.71247053e-01
1.04400623e+00 9.28002834e-01 -2.43498027e-01 -1.25150371e+00
-7.36788452e-01 -4.73647952e-01 -7.55437613e-01 -6.13220632e-01
3.07898968e-01 -9.46042314e-02 -8.51908207e-01 1.61002505e+00
4.24930245e-01 -1.28597826e-01 1.88868283e-03 7.03845501e-01
1.34795380e+00 7.12140501e-01 1.44460589e-01 1.23201475e-01
1.17128670e+00 -1.59866762e+00 -7.61126339e-01 -1.91556573e-01
9.96387899e-01 -2.56155044e-01 1.25845480e+00 -1.31653160e-01
-1.25086701e+00 -4.29407448e-01 -7.81699538e-01 -7.50192106e-01
-5.43437362e-01 -5.72902918e-01 5.55822968e-01 4.35529888e-01
-1.65813196e+00 4.07604277e-01 -3.58779430e-01 -2.99587220e-01
2.55233735e-01 5.38476288e-01 -1.32876039e-01 -3.06557089e-01
-1.75441229e+00 1.13393116e+00 3.90028566e-01 2.64205605e-01
-8.61965120e-01 -3.69399458e-01 -1.18169022e+00 4.71537977e-01
6.45999670e-01 -1.26720870e+00 1.38334572e+00 -4.84171212e-01
-1.33023882e+00 4.78845358e-01 -1.90707982e-01 -6.74231887e-01
7.33869802e-03 9.21831559e-03 -1.87918574e-01 5.19269526e-01
8.48580003e-02 6.03670061e-01 4.20333743e-01 -1.06393683e+00
-2.26265416e-01 -5.51412523e-01 7.07972467e-01 3.41901213e-01
-1.52683929e-01 2.85338946e-02 -4.60762173e-01 -6.21078163e-02
-1.76111653e-01 -6.53909624e-01 -3.52197617e-01 -6.04304612e-01
-7.36108184e-01 -7.69944251e-01 6.41414881e-01 -1.08461607e+00
1.33843589e+00 -1.49142468e+00 5.05308270e-01 1.34395450e-01
7.80387521e-01 1.87344566e-01 -5.68508387e-01 7.35941052e-01
2.01850429e-01 4.52835523e-02 -4.98455852e-01 -3.46753329e-01
2.17572927e-01 3.07072699e-01 -1.11355335e-01 3.10067777e-02
3.81037593e-01 1.63756192e+00 -8.04805994e-01 -6.60986841e-01
-3.00954431e-01 1.78956196e-01 -6.42970800e-01 1.86318487e-01
-7.23955691e-01 1.98734343e-01 -4.47158515e-01 3.77013147e-01
3.48756045e-01 -8.23072493e-01 1.84604049e-01 -1.59088820e-01
4.85064238e-01 7.12580264e-01 -5.33615768e-01 1.54086590e+00
-2.19448775e-01 4.88366872e-01 6.78998902e-02 -9.39719021e-01
6.57450855e-01 2.37627387e-01 -3.90266664e-02 -9.27694082e-01
1.69028997e-01 -1.77050903e-01 3.36354315e-01 -5.26216507e-01
7.90878534e-01 2.55598456e-01 -4.23675030e-02 4.19500530e-01
2.96449989e-01 -2.23844990e-01 4.52103049e-01 8.65062237e-01
1.50585306e+00 -3.27537119e-01 1.42945737e-01 -6.76756501e-02
7.98325181e-01 1.38863415e-01 -3.39934565e-02 9.50069845e-01
-1.45145103e-01 3.34464490e-01 9.71047461e-01 -2.48680443e-01
-7.18548179e-01 -7.01847672e-01 6.89968467e-01 1.37853742e+00
-1.41058117e-01 -5.50117850e-01 -9.01574910e-01 -1.11695194e+00
-1.58517435e-01 8.62501919e-01 -7.85657048e-01 -1.49034753e-01
-1.04655814e+00 -5.16135812e-01 5.60789704e-01 4.90045756e-01
7.18668401e-01 -1.24277055e+00 -1.90807015e-01 1.96265817e-01
-4.33182061e-01 -1.29045391e+00 -4.54635799e-01 -1.09693356e-01
-7.24696159e-01 -1.13273203e+00 -5.15630960e-01 -8.94134104e-01
2.53921449e-01 1.60253525e-01 1.72309875e+00 6.55831039e-01
1.59592032e-01 8.53533089e-01 -5.27188003e-01 -9.73997265e-02
-4.34049994e-01 6.38179123e-01 -7.41059482e-01 -5.54151461e-02
5.19855544e-02 -2.75884092e-01 -4.80516195e-01 -1.72115088e-01
-8.15125763e-01 -2.17848159e-02 4.01166886e-01 7.46419966e-01
4.06724632e-01 -1.87893182e-01 8.94593775e-01 -1.30748832e+00
1.29910362e+00 -6.43879056e-01 -5.42366266e-01 7.91091204e-01
-5.87494791e-01 7.09021837e-02 6.44713700e-01 2.23089471e-01
-8.11802626e-01 -4.49719846e-01 -4.04791623e-01 -3.60427648e-02
1.08749054e-01 1.05256736e+00 -1.30715117e-01 -2.77170986e-01
4.54911113e-01 6.93180114e-02 -1.57571331e-01 -1.42105386e-01
8.19583654e-01 3.12014371e-01 3.09684604e-01 -3.43561858e-01
7.74614751e-01 3.08296606e-02 1.59090802e-01 -6.05450153e-01
-1.13267112e+00 -4.25776422e-01 -7.33685136e-01 1.67710381e-03
1.20455503e+00 -3.45501959e-01 -8.89227509e-01 1.30592495e-01
-1.43384290e+00 -5.85625112e-01 7.53852120e-03 -1.87885135e-01
-3.76835883e-01 6.72377706e-01 -1.05336249e+00 -4.77779448e-01
-9.34319615e-01 -9.14656162e-01 9.96360064e-01 1.26048446e-01
2.73417644e-02 -1.51746941e+00 2.21289113e-01 9.62269306e-01
4.50970352e-01 1.38876036e-01 1.71540082e+00 -9.46301460e-01
-6.92924917e-01 -1.67365167e-02 -6.16272211e-01 7.71841854e-02
-2.54939348e-01 -3.80717158e-01 -6.79834783e-01 -3.02685320e-01
-6.84001967e-02 -4.98887211e-01 9.83543813e-01 1.80789620e-01
1.03518891e+00 -5.00535011e-01 5.80247585e-03 3.39954086e-02
1.28957665e+00 -1.04586549e-01 5.50911546e-01 2.68554300e-01
1.10975564e+00 8.26720178e-01 1.46424755e-01 -2.79075086e-01
1.11656141e+00 5.77941060e-01 8.27197552e-01 -4.09717374e-02
-2.52639502e-01 -1.61671951e-01 -5.52964993e-02 1.30670297e+00
1.97454914e-01 -7.34632552e-01 -1.16708589e+00 6.14349008e-01
-1.90645790e+00 -5.47907710e-01 -6.39504492e-01 1.56627107e+00
4.78696316e-01 -2.59015430e-02 1.13935098e-01 -6.49927258e-02
5.69782853e-01 3.94529641e-01 -4.63045746e-01 -5.66643417e-01
-1.72845423e-01 4.28067625e-01 1.94250256e-01 9.76910651e-01
-8.97658467e-01 1.11726582e+00 6.07108641e+00 3.91658932e-01
-4.75531310e-01 2.66614527e-01 4.57310945e-01 6.22484148e-01
-5.75358748e-01 -3.83948572e-02 -6.02958441e-01 -1.09726705e-01
1.15282691e+00 -6.22704811e-02 3.53076637e-01 3.50649953e-01
-3.08113635e-01 1.92959845e-01 -8.97351384e-01 3.88238162e-01
2.95894295e-01 -1.48206890e+00 4.98991847e-01 -3.10714900e-01
5.21469355e-01 1.21690407e-01 -1.70867562e-01 7.03152955e-01
5.62563181e-01 -1.03008974e+00 6.43941760e-02 3.92476052e-01
1.41620338e-01 -6.90256357e-01 9.92174268e-01 2.56397873e-01
-1.14463377e+00 -3.85564268e-01 -4.35678661e-01 -1.42812490e-01
2.68650234e-01 1.73617959e-01 -1.11963952e+00 1.17751098e+00
5.65025032e-01 3.82947177e-01 -1.09774077e+00 8.98340523e-01
-5.41252375e-01 8.98598731e-01 1.99691519e-01 -2.54744053e-01
6.93648934e-01 -3.85177583e-02 3.83287072e-01 9.20795798e-01
-6.01781793e-02 1.02770053e-01 -2.77800597e-02 7.49327600e-01
-4.67137307e-01 3.03751320e-01 -3.73404235e-01 -1.40164450e-01
1.07613213e-01 1.29037619e+00 -4.77104127e-01 -3.39351684e-01
-6.23784602e-01 9.54470515e-01 1.02511585e+00 4.41017330e-01
-7.57600188e-01 -4.56171572e-01 -1.72107950e-01 -1.09530449e-01
3.51480097e-01 -2.35278010e-01 1.07069954e-01 -9.35434520e-01
-4.06856164e-02 -1.04944468e+00 1.22828448e+00 -1.10584021e+00
-1.50405943e+00 8.71948183e-01 -2.21213609e-01 -2.46003568e-01
-2.40754813e-01 -4.46441948e-01 -6.93506718e-01 9.00367379e-01
-1.78759885e+00 -1.41764319e+00 -3.54457229e-01 9.56067085e-01
3.34951431e-01 1.02359734e-01 8.66555870e-01 2.68362164e-01
-3.98474813e-01 5.97267032e-01 -3.81420135e-01 3.93464416e-01
2.83276051e-01 -1.51575494e+00 8.05722952e-01 6.70990288e-01
3.81826937e-01 4.59938109e-01 4.72558796e-01 -6.03060186e-01
-1.31626523e+00 -1.17394245e+00 1.33364260e+00 -6.15797818e-01
9.91754353e-01 -3.07357967e-01 -1.28960955e+00 1.03906775e+00
7.38837481e-01 -3.53392333e-01 3.99929076e-01 1.45531684e-01
-3.93413901e-01 6.20309263e-02 -8.84613395e-01 5.13992548e-01
7.05744088e-01 -5.23973227e-01 -7.85063565e-01 7.02220559e-01
1.56036007e+00 -4.50732380e-01 -8.55796099e-01 4.14913535e-01
-1.39433578e-01 -8.34121048e-01 7.65642583e-01 -1.00647902e+00
4.38322276e-01 -9.49693620e-02 -1.98794175e-02 -1.24258971e+00
-5.85824490e-01 -6.24351501e-01 -5.22478223e-01 1.01071429e+00
6.68844819e-01 -7.34013617e-01 7.14985907e-01 4.84208673e-01
-4.75445420e-01 -1.03778052e+00 -1.01499641e+00 -3.28468621e-01
3.03052932e-01 7.14685991e-02 6.69483662e-01 8.34307909e-01
-2.70531893e-01 1.14927804e+00 -1.99507520e-01 4.29412514e-01
4.16786432e-01 2.63491809e-01 6.59595907e-01 -1.22797382e+00
-4.39316005e-01 -3.47010702e-01 -1.27418309e-01 -1.08524394e+00
3.03315908e-01 -9.82083797e-01 -2.55447865e-01 -2.36845064e+00
1.96415409e-01 2.29295157e-02 -1.99671909e-01 2.45661244e-01
-4.19193953e-01 -9.04631149e-03 1.30503729e-01 -6.46140948e-02
-1.20263731e+00 5.72888136e-01 1.53784442e+00 -4.64305401e-01
5.57929762e-02 -1.04042795e-02 -7.75365174e-01 2.47576952e-01
7.56974220e-01 -2.82992154e-01 -7.30257094e-01 -9.81870234e-01
7.19549358e-01 3.50530952e-01 3.79351586e-01 -6.20481372e-01
5.27769506e-01 2.85691768e-01 -1.85212493e-01 -5.73630452e-01
2.63689399e-01 -5.18132985e-01 -5.80152750e-01 4.20878500e-01
-5.84464073e-01 6.05708838e-01 7.35670328e-02 6.53223634e-01
-3.76098573e-01 -2.89839059e-01 3.77706945e-01 -2.94921458e-01
-4.73643094e-01 4.59233910e-01 -9.03884992e-02 4.39908177e-01
6.25324905e-01 2.35913262e-01 -9.80628610e-01 -9.11443770e-01
-7.25566804e-01 7.50926018e-01 -2.57577419e-01 3.25867981e-01
5.81284881e-01 -9.43649948e-01 -8.58264983e-01 -5.15790641e-01
1.13671899e-01 4.19331854e-03 4.05308276e-01 9.31605816e-01
-5.92400193e-01 8.37974668e-01 2.17232615e-01 -2.17980728e-01
-1.28888655e+00 6.95410669e-01 5.65334976e-01 -1.09203696e+00
-4.01758522e-01 9.80245411e-01 3.80819999e-02 -8.03778291e-01
3.84664014e-02 -2.60153234e-01 -8.22065055e-01 -5.94099686e-02
2.30946437e-01 3.28450531e-01 3.96760106e-01 -5.99272311e-01
-7.94095173e-02 3.24699104e-01 -3.15455377e-01 1.93729863e-01
1.21034122e+00 -2.54753113e-01 -7.26563931e-01 2.57703543e-01
1.02581108e+00 -3.18266213e-01 -3.86037856e-01 -3.65152299e-01
3.41758251e-01 1.19057477e-01 -1.43278390e-01 -8.34778786e-01
-8.57430220e-01 1.15832734e+00 8.41741636e-02 7.87509143e-01
9.13052678e-01 3.19267303e-01 1.11693192e+00 7.56598055e-01
7.79911652e-02 -7.37492979e-01 3.61251473e-01 9.57732320e-01
1.10894716e+00 -1.02898431e+00 -2.35054463e-01 -5.04654527e-01
-4.86462116e-01 9.19213355e-01 8.28140914e-01 7.27987140e-02
2.66464412e-01 -4.58338261e-01 4.13601696e-02 -8.19809258e-01
-9.50926125e-01 -2.60826826e-01 5.07077932e-01 4.52764899e-01
2.91896343e-01 -2.34903336e-01 -2.40717649e-01 4.40591276e-01
-2.53572524e-01 -4.95277971e-01 5.18205166e-01 6.26519740e-01
-4.27625984e-01 -9.07468677e-01 1.32519141e-01 5.65547526e-01
-3.41024578e-01 -5.27683675e-01 -9.55210865e-01 9.16052759e-01
-5.11878252e-01 1.42088819e+00 -1.21546455e-01 -3.20608258e-01
4.44870383e-01 2.93414503e-01 3.92879337e-01 -7.20766485e-01
-1.03239298e+00 -7.28068948e-01 4.78409827e-01 -5.21588147e-01
-2.29768425e-01 -9.77031216e-02 -1.18492067e+00 -2.11117506e-01
-4.97143954e-01 5.85852146e-01 1.99756458e-01 1.09647799e+00
5.99433541e-01 8.84251297e-01 1.68851584e-01 -1.27508655e-01
-4.87984627e-01 -1.24618340e+00 -2.10880563e-01 3.48986089e-01
3.05605978e-01 -2.29690447e-01 -2.64859438e-01 -4.79941756e-01] | [10.770428657531738, 7.947452068328857] |
465fb2e6-e569-4a9b-a3f3-abc144efe8b2 | federated-ensemble-directed-offline | 2305.03097 | null | https://arxiv.org/abs/2305.03097v1 | https://arxiv.org/pdf/2305.03097v1.pdf | Federated Ensemble-Directed Offline Reinforcement Learning | We consider the problem of federated offline reinforcement learning (RL), a scenario under which distributed learning agents must collaboratively learn a high-quality control policy only using small pre-collected datasets generated according to different unknown behavior policies. Naively combining a standard offline RL approach with a standard federated learning approach to solve this problem can lead to poorly performing policies. In response, we develop the Federated Ensemble-Directed Offline Reinforcement Learning Algorithm (FEDORA), which distills the collective wisdom of the clients using an ensemble learning approach. We develop the FEDORA codebase to utilize distributed compute resources on a federated learning platform. We show that FEDORA significantly outperforms other approaches, including offline RL over the combined data pool, in various complex continuous control environments and real world datasets. Finally, we demonstrate the performance of FEDORA in the real-world on a mobile robot. | ['Srinivas Shakkottai', 'Dileep Kalathil', 'Nitin Ragothaman', 'Desik Rengarajan'] | 2023-05-04 | null | null | null | null | ['offline-rl', 'continuous-control'] | ['playing-games', 'playing-games'] | [-5.32810450e-01 7.58852884e-02 -2.78210584e-02 -1.94954559e-01
-8.68610799e-01 -7.71358311e-01 6.20999157e-01 -1.81565836e-01
-4.36004221e-01 1.29532230e+00 5.81163839e-02 -2.31869802e-01
-2.42898494e-01 -8.50990057e-01 -1.00042641e+00 -9.29743946e-01
-4.92930561e-01 8.05819392e-01 -1.20899200e-01 -1.59142911e-01
-3.54758054e-02 1.90106407e-01 -1.89806747e+00 7.41831139e-02
8.70674074e-01 9.66369092e-01 1.80724323e-01 8.25341702e-01
-1.87747683e-02 1.51557016e+00 -5.18862963e-01 -7.96734169e-02
6.44513726e-01 -3.29556227e-01 -6.44778907e-01 -1.29774228e-01
3.21859419e-01 -6.80314243e-01 -3.17985922e-01 9.17356908e-01
5.91304183e-01 4.06563580e-01 1.62716240e-01 -1.56285584e+00
-1.20547615e-01 7.53491402e-01 1.66766956e-01 -3.43600482e-01
1.66644126e-01 7.50440717e-01 7.90575802e-01 -3.63244832e-01
6.89368486e-01 1.25202894e+00 4.90478873e-01 7.83455908e-01
-9.81293082e-01 -6.48119390e-01 3.52743059e-01 2.94360489e-01
-6.04455352e-01 -4.71031308e-01 2.84804940e-01 -1.80029050e-01
1.03116810e+00 -2.08441958e-01 6.37844265e-01 1.24024916e+00
4.05013353e-01 9.23688889e-01 1.37497413e+00 -1.62703335e-01
1.05801702e+00 -2.85830408e-01 -3.01449448e-01 9.56378460e-01
4.04300570e-01 7.60078490e-01 -6.59811497e-01 -6.59151852e-01
3.65421355e-01 1.26906067e-01 1.12137057e-01 -6.44735575e-01
-9.31705952e-01 6.55264795e-01 2.01826707e-01 -2.38107845e-01
-8.01876962e-01 5.18386841e-01 6.26601458e-01 9.50497389e-01
8.90951678e-02 4.75752145e-01 -9.80664253e-01 -3.06459665e-01
-6.37005419e-02 5.47284663e-01 1.38009095e+00 7.53957450e-01
1.17803729e+00 4.26554978e-01 5.47442846e-02 4.70370233e-01
2.39600182e-01 6.90058172e-01 7.47994721e-01 -1.92556298e+00
1.57292277e-01 3.81963015e-01 7.09515333e-01 -4.36844558e-01
-1.89464286e-01 -1.23880676e-03 -2.43297577e-01 9.40158546e-01
4.92685348e-01 -9.67521369e-01 -4.84778374e-01 1.73586714e+00
8.47086608e-01 1.88488454e-01 5.76547325e-01 9.60596621e-01
-3.24708335e-02 3.84937495e-01 -8.00970122e-02 -3.00312966e-01
5.21352112e-01 -1.32678628e+00 -6.06443703e-01 8.08031037e-02
5.99064827e-01 2.59394906e-02 7.23753929e-01 8.24038863e-01
-7.67051160e-01 -1.06747985e-01 -1.08345580e+00 6.57064140e-01
-3.62675130e-01 -3.31896365e-01 7.73910701e-01 2.83165365e-01
-1.04835904e+00 9.65940773e-01 -1.07392502e+00 -2.82858223e-01
2.91362494e-01 5.37619650e-01 -2.12821916e-01 -2.05899730e-01
-7.03019857e-01 8.22960436e-01 2.45652154e-01 -4.98566240e-01
-1.87336552e+00 -3.84530872e-01 -5.11618674e-01 -2.15137497e-01
9.29047823e-01 -8.47262800e-01 1.98617947e+00 -1.34110761e+00
-2.20764470e+00 8.03908426e-03 4.28741276e-01 -7.03753531e-01
4.89343584e-01 -3.42538863e-01 -2.07165718e-01 -2.88590305e-02
-1.21590093e-01 -1.01071127e-01 9.06520188e-01 -1.42586184e+00
-1.17919874e+00 -6.24275148e-01 1.29897073e-01 3.47847372e-01
-1.75739616e-01 -5.06365240e-01 3.66160423e-01 -3.14985923e-02
-6.73666835e-01 -9.85468626e-01 -6.66742504e-01 -2.76036829e-01
2.59665579e-01 -4.23113257e-01 9.67564702e-01 -1.85305983e-01
4.45396215e-01 -1.79899144e+00 1.20511808e-01 1.59536034e-01
1.30109295e-01 2.51808539e-02 -5.10495543e-01 7.39929557e-01
6.19618833e-01 -3.17300141e-01 1.72441214e-01 -7.09694326e-02
4.05548960e-01 7.61169791e-01 -3.05985451e-01 4.59776878e-01
-4.86503303e-01 5.22411942e-01 -1.38439941e+00 -3.26135188e-01
-1.67431325e-01 -1.28331110e-01 -6.48263812e-01 8.03669930e-01
-1.01080239e+00 6.11076117e-01 -8.54096591e-01 6.45357609e-01
1.86309710e-01 -2.60334849e-01 9.57304299e-01 4.64480817e-01
-1.62484095e-01 1.65902693e-02 -1.10831892e+00 1.91189408e+00
-6.10903919e-01 -2.64093024e-03 8.15145552e-01 -8.69362533e-01
8.50243449e-01 3.88329297e-01 1.09540355e+00 -6.54046655e-01
1.44808769e-01 3.33561391e-01 -2.33980983e-01 -4.96950626e-01
2.60886133e-01 2.56311029e-01 -1.58352718e-01 1.20697248e+00
6.03206336e-01 1.65967777e-01 1.27177835e-02 2.15805620e-01
1.76346624e+00 4.93238717e-01 3.79840255e-01 -5.32231620e-03
1.67760283e-01 4.38715607e-01 8.15054536e-01 1.18963361e+00
-6.63546383e-01 -2.85439163e-01 1.38520703e-01 -7.50865698e-01
-7.88130999e-01 -9.15958703e-01 6.96923435e-01 1.60575509e+00
5.34327440e-02 -4.47118074e-01 -5.58554888e-01 -1.30122077e+00
6.14040494e-01 4.38298464e-01 -4.48995441e-01 1.24160744e-01
-5.06153047e-01 -2.81422317e-01 3.34059834e-01 9.63818058e-02
5.60638368e-01 -1.24430478e+00 -1.19263625e+00 5.65568507e-01
1.75201207e-01 -5.99744976e-01 -1.70616373e-01 4.55224693e-01
-6.28092170e-01 -1.57264733e+00 -2.56983973e-02 -3.70070219e-01
2.52138823e-01 1.81216523e-01 1.17490602e+00 -1.72023457e-02
-3.74722667e-02 1.18412519e+00 -3.98517460e-01 -3.56654376e-01
-6.36813164e-01 -3.27858239e-01 6.50806487e-01 -5.00367954e-02
4.20308523e-02 -6.33917749e-01 -5.56126416e-01 1.05526142e-01
-5.95526338e-01 -4.30510014e-01 3.06114614e-01 1.28276384e+00
3.45559329e-01 -3.33738849e-02 9.63725984e-01 -7.58029044e-01
7.59451747e-01 -8.95383000e-01 -9.32854354e-01 3.80269796e-01
-9.93520379e-01 2.38473877e-01 1.25414538e+00 -4.89761442e-01
-1.23062456e+00 2.89874315e-01 4.26005512e-01 -6.12332880e-01
-2.40693733e-01 2.39047438e-01 1.98503002e-03 -1.22535951e-01
8.04588377e-01 2.65243888e-01 5.95002472e-01 -5.03588915e-01
6.73699915e-01 8.17170560e-01 4.78290111e-01 -1.25948274e+00
2.79304028e-01 2.63407439e-01 -2.90728867e-01 -2.40953621e-02
-3.79239351e-01 -6.39395714e-02 9.23847705e-02 -4.62917864e-01
1.16992906e-01 -1.04458952e+00 -1.21101987e+00 4.35937911e-01
-7.73300409e-01 -1.16290760e+00 -6.07407331e-01 3.81769717e-01
-1.39170980e+00 -8.80746692e-02 -6.41399682e-01 -1.00039864e+00
-4.53405529e-01 -9.05855834e-01 4.68763798e-01 2.98925579e-01
3.45566332e-01 -7.63907850e-01 1.00829339e+00 9.20932158e-04
6.32185459e-01 1.80094063e-01 5.80623448e-01 -9.16491926e-01
-4.51591283e-01 3.79861653e-01 4.61399406e-01 1.17488347e-01
1.75612435e-01 -2.39337400e-01 -8.98215830e-01 -7.68296599e-01
-2.66052604e-01 -1.24682629e+00 3.35033059e-01 -2.11061850e-01
9.44995403e-01 -9.57039416e-01 -1.25328779e-01 3.60667378e-01
1.63171375e+00 2.01172501e-01 -1.09588750e-01 4.05165076e-01
1.31237820e-01 4.50453550e-01 7.92461693e-01 1.16114712e+00
5.56652546e-01 1.70742467e-01 8.93002689e-01 5.84260046e-01
1.46289483e-01 -5.62204719e-01 9.52864885e-01 6.71848178e-01
3.53557505e-02 7.88182169e-02 -6.62359118e-01 4.56863374e-01
-2.57993031e+00 -1.10148203e+00 8.26973617e-01 2.00859070e+00
8.53937328e-01 -6.40012205e-01 4.27847445e-01 -5.08956909e-01
3.03415388e-01 -1.85678259e-01 -1.50689590e+00 -4.69615340e-01
4.64819074e-02 1.64990440e-01 6.15655601e-01 1.34020612e-01
-8.03510189e-01 8.38534355e-01 6.66220474e+00 4.04580921e-01
-1.17045820e+00 3.77036363e-01 2.34018311e-01 -2.07376301e-01
2.44679991e-02 -9.38663781e-02 -3.52684706e-01 4.02947843e-01
1.37062097e+00 -5.35260201e-01 1.36549640e+00 1.45092845e+00
7.41101354e-02 -2.96067521e-02 -9.97806013e-01 7.76984274e-01
-3.65365595e-01 -1.38959014e+00 -4.27446753e-01 6.03384674e-02
1.16921830e+00 1.00171721e+00 -2.25997820e-01 8.31962705e-01
1.97497249e+00 -7.91118443e-01 5.67501426e-01 5.73075533e-01
4.25615430e-01 -7.94992149e-01 2.36703128e-01 7.95956433e-01
-8.16452026e-01 -9.80679631e-01 -2.07216725e-01 -2.03012601e-01
-3.72538418e-01 9.22157895e-03 -7.95254707e-01 5.52364886e-01
9.85565782e-01 6.06108785e-01 -1.55401081e-01 9.13896263e-01
4.09350395e-02 6.07743680e-01 -2.78993994e-01 -2.08526999e-01
2.49294192e-01 -1.97467566e-01 4.77968901e-01 4.92930144e-01
1.37046635e-01 1.09783314e-01 5.07671893e-01 4.89847511e-01
-9.12207887e-02 -1.75661054e-02 -9.55941439e-01 -1.41091282e-02
6.09436512e-01 1.35053515e+00 8.77029523e-02 -4.90777314e-01
-3.93790632e-01 7.43039250e-01 9.54914749e-01 5.77069044e-01
-2.92562813e-01 -2.59211808e-02 1.16801620e+00 -7.07449436e-01
3.73942733e-01 -3.13745260e-01 3.39502305e-01 -1.27878737e+00
-4.05380905e-01 -1.66304588e+00 6.77009821e-01 -3.75377327e-01
-1.86882472e+00 3.51551116e-01 -5.29895067e-01 -1.12481916e+00
-9.25016344e-01 -4.97594297e-01 -5.23426056e-01 4.24048267e-02
-1.34104025e+00 -7.24892139e-01 -2.20876768e-01 1.03748405e+00
4.69830692e-01 -6.47236884e-01 1.18938911e+00 -2.35092506e-01
-5.70039630e-01 1.37674153e-01 7.80510187e-01 -3.52962822e-01
8.36047173e-01 -1.35559177e+00 -2.20451787e-01 2.73213953e-01
-9.36249569e-02 2.03808367e-01 5.21978438e-01 -5.27672648e-01
-2.42781878e+00 -1.37371683e+00 -2.08039984e-01 -2.73782253e-01
8.08718145e-01 -1.79201029e-02 -5.09165943e-01 7.11942375e-01
5.90483010e-01 4.80560720e-01 4.94015962e-01 -6.40494749e-02
-3.34206045e-01 -4.86115545e-01 -1.45266986e+00 4.98398989e-01
9.65452194e-01 -2.24213421e-01 -2.96306670e-01 3.96341056e-01
7.62756348e-01 -1.38357878e-01 -9.88587379e-01 7.50333741e-02
5.97035408e-01 -9.00105715e-01 2.42408931e-01 -1.26761997e+00
5.42946830e-02 -2.23046646e-01 -4.69047517e-01 -1.93657255e+00
-4.14934196e-02 -1.27350318e+00 -7.01805174e-01 5.88329911e-01
-3.90542746e-02 -1.05894327e+00 7.87182987e-01 6.07173085e-01
1.34292662e-01 -5.47490537e-01 -1.02240646e+00 -9.46345210e-01
2.24372268e-01 -1.81018591e-01 1.02051711e+00 7.32796371e-01
4.90463674e-01 -1.58140793e-01 -3.02789152e-01 -7.81729072e-02
1.04819894e+00 3.26583117e-01 1.34438884e+00 -7.72679210e-01
-6.76516652e-01 8.35635066e-02 1.89187899e-01 -6.72193766e-01
6.80978000e-01 -6.38015509e-01 4.05717850e-01 -1.09521472e+00
-3.52738202e-02 -8.70714068e-01 -5.50094426e-01 8.76454651e-01
3.48196805e-01 -5.72745860e-01 3.85291874e-01 3.52259159e-01
-1.30271077e+00 8.82133603e-01 1.01722884e+00 -2.60538399e-01
-1.99311554e-01 -2.31105238e-01 -4.56419855e-01 5.81271529e-01
1.03179705e+00 -3.43236595e-01 -3.88801664e-01 -4.79401290e-01
-3.96788232e-02 5.26928663e-01 1.45935491e-01 -9.92921472e-01
6.33437991e-01 -7.49383271e-01 -1.71887483e-02 4.19401377e-02
-1.16726205e-01 -9.94857073e-01 7.79150054e-02 6.73869252e-01
-3.34359765e-01 2.86920786e-01 -1.77250445e-01 1.01174223e+00
9.92817506e-02 2.67421305e-01 4.65821594e-01 -6.10186279e-01
-8.32964182e-01 3.85820895e-01 -5.88436484e-01 1.10128745e-02
1.35807848e+00 5.78454733e-01 -7.17761099e-01 -5.04992723e-01
-5.72769582e-01 8.33224714e-01 6.56122267e-01 3.05089474e-01
5.03198624e-01 -1.01824474e+00 -6.14549875e-01 2.06082731e-01
-1.31341010e-01 -3.33842158e-01 -9.28724334e-02 5.16085148e-01
-2.99652927e-02 -7.76443910e-03 -4.86233741e-01 -2.31898963e-01
-6.55434191e-01 6.28841400e-01 7.76715398e-01 -3.02644193e-01
-5.71998298e-01 3.48954536e-02 -5.89188814e-01 -1.07912123e+00
5.24969339e-01 1.90559447e-01 2.01998785e-01 -4.41055954e-01
5.27741015e-01 7.30074942e-01 6.88815564e-02 -3.71918939e-02
-1.12800188e-01 -2.08176106e-01 2.56466687e-01 -3.51994276e-01
1.39497888e+00 -2.34061554e-02 8.80326517e-03 3.01732391e-01
6.39083862e-01 -2.50743181e-01 -1.89733958e+00 -3.16760361e-01
1.15043536e-01 -2.86279261e-01 -7.06249475e-02 -1.18945312e+00
-1.10233462e+00 -4.34517637e-02 7.12017536e-01 -5.70064969e-02
8.86006832e-01 -4.37214106e-01 5.93022764e-01 1.14871597e+00
1.33972466e+00 -1.55382192e+00 1.85719937e-01 7.36180723e-01
5.91630757e-01 -1.31459343e+00 -1.66286528e-01 7.44312346e-01
-7.60140657e-01 1.14683950e+00 1.05889881e+00 -5.22914827e-01
5.19250989e-01 6.22847199e-01 3.09048206e-01 6.69630989e-02
-1.89755750e+00 -1.84864521e-01 -7.67353714e-01 9.00375426e-01
-4.45763975e-01 2.41993293e-01 -1.74842123e-02 6.92983687e-01
1.03495754e-01 3.93570125e-01 4.70013410e-01 1.44567764e+00
-5.52266479e-01 -1.11092234e+00 -4.39641237e-01 3.73158336e-01
-2.20613554e-01 8.23327422e-01 -3.49066079e-01 3.90314341e-01
1.03138566e-01 1.22232187e+00 4.56321090e-02 -5.65666616e-01
-8.13990235e-02 -1.49060162e-02 5.70776939e-01 -2.20121905e-01
-9.50172186e-01 -3.09987456e-01 1.76102445e-01 -1.34877992e+00
-4.41463143e-01 -5.53069830e-01 -1.55825508e+00 -1.41820639e-01
1.33457199e-01 3.96809131e-01 8.70876253e-01 8.05207491e-01
7.48233080e-01 9.06141326e-02 1.15827382e+00 -8.86034429e-01
-1.62065756e+00 -6.50769413e-01 -5.58191001e-01 2.19672129e-01
5.51904976e-01 -6.14262283e-01 -3.57854426e-01 -2.16701940e-01] | [3.954127073287964, 1.9103909730911255] |
abbd3517-fd6c-4dd8-81f7-0a2153191621 | feature-rich-named-entity-recognition-for | 2109.15121 | null | https://arxiv.org/abs/2109.15121v1 | https://arxiv.org/pdf/2109.15121v1.pdf | Feature-Rich Named Entity Recognition for Bulgarian Using Conditional Random Fields | The paper presents a feature-rich approach to the automatic recognition and categorization of named entities (persons, organizations, locations, and miscellaneous) in news text for Bulgarian. We combine well-established features used for other languages with language-specific lexical, syntactic and morphological information. In particular, we make use of the rich tagset annotation of the BulTreeBank (680 morpho-syntactic tags), from which we derive suitable task-specific tagsets (local and nonlocal). We further add domain-specific gazetteers and additional unlabeled data, achieving F1=89.4%, which is comparable to the state-of-the-art results for English. | ['Kiril Ivanov Simov', 'Petya Osenova', 'Kuzman Ganchev', 'Preslav Nakov', 'Georgi Georgiev'] | 2021-09-26 | null | null | null | null | ['miscellaneous'] | ['miscellaneous'] | [-6.16896391e-01 6.33039400e-02 -2.17076853e-01 -5.45704424e-01
-8.21906328e-01 -7.99713016e-01 9.54038918e-01 7.49553144e-01
-1.06439590e+00 1.12895107e+00 6.40262365e-01 -2.19184682e-01
-4.43480089e-02 -7.93482363e-01 -1.53384104e-01 -5.33599317e-01
-4.99085821e-02 7.14096069e-01 3.44953269e-01 -2.65836298e-01
3.29604983e-01 6.51071429e-01 -1.30269516e+00 2.11649001e-01
8.74646306e-01 6.27461076e-01 1.21835880e-01 2.85194665e-01
-4.89766061e-01 5.80464900e-01 -6.52088940e-01 -8.82345319e-01
-2.30181381e-01 -2.36779395e-02 -1.22407007e+00 8.26237425e-02
3.71493697e-01 2.15424269e-01 -1.37921736e-01 1.02265549e+00
3.18997413e-01 1.89578399e-01 8.80978584e-01 -3.98714483e-01
-6.29192948e-01 8.53160024e-01 -3.14424366e-01 3.96239579e-01
2.26993248e-01 -4.58118975e-01 1.17683256e+00 -1.01485717e+00
1.19922137e+00 9.68442917e-01 4.52635169e-01 2.94102669e-01
-7.43572533e-01 -5.29125512e-01 -8.26204475e-03 -1.07293792e-01
-1.76408374e+00 -5.13660669e-01 1.09025538e-01 -7.29475260e-01
9.42880094e-01 1.32495105e-01 2.08959028e-01 7.11663842e-01
1.28772080e-01 2.93351024e-01 1.12122381e+00 -7.82094002e-01
8.63361172e-03 5.74943841e-01 4.67449784e-01 7.64611781e-01
6.80799782e-01 -2.92137235e-01 -2.82961696e-01 -2.76224822e-01
6.00642502e-01 -2.38254339e-01 1.33857071e-01 -2.31624469e-01
-1.29181719e+00 6.43438101e-01 1.78844690e-01 1.02383983e+00
-5.43175042e-01 -3.73629004e-01 6.51631236e-01 1.05198912e-01
7.79644668e-01 5.70060909e-01 -9.13130105e-01 -4.12218347e-02
-8.88577700e-01 -1.56188207e-02 9.67584193e-01 1.13295543e+00
1.01766682e+00 -6.38516098e-02 2.09607603e-03 1.03812730e+00
1.94515854e-01 4.31130737e-01 5.21537542e-01 -3.54618639e-01
6.52388215e-01 7.50275254e-01 3.10739309e-01 -8.65716696e-01
-6.46337867e-01 -4.34066743e-01 -4.71263617e-01 -4.80862826e-01
6.26361132e-01 -5.77889860e-01 -9.72876132e-01 1.39102721e+00
4.04942989e-01 -4.74262595e-01 3.80416870e-01 4.86380041e-01
1.28674233e+00 5.69240808e-01 5.44401884e-01 -3.24916452e-01
1.79072022e+00 -5.51018476e-01 -7.76902676e-01 1.47274667e-02
8.92085493e-01 -8.05346251e-01 3.66163015e-01 -6.16442114e-02
-7.07581222e-01 -3.16370904e-01 -3.47856462e-01 1.23680383e-01
-1.01438081e+00 5.01843631e-01 7.58782566e-01 8.73739898e-01
-7.72449374e-01 1.15021773e-01 -5.09802580e-01 -7.33266056e-01
-7.31033459e-02 3.27415705e-01 -9.23653364e-01 1.91036627e-01
-1.18423021e+00 1.12377799e+00 8.99924576e-01 -1.01172052e-01
-4.06351805e-01 -2.42186755e-01 -1.18317473e+00 -1.70423001e-01
6.79861367e-01 -2.09193584e-03 9.70904112e-01 -5.92212558e-01
-1.11847162e+00 1.56394839e+00 -1.28344968e-01 -3.05014461e-01
1.00700192e-01 -5.14920577e-02 -1.20098829e+00 2.02009067e-01
5.83397329e-01 2.72008955e-01 1.95356846e-01 -8.76941383e-01
-1.05576360e+00 -3.42850834e-01 -1.31879538e-01 5.51033802e-02
-5.26614964e-01 8.39938283e-01 -4.88856047e-01 -7.98425615e-01
1.37834698e-01 -7.02762365e-01 -1.24542132e-01 -8.60790491e-01
-2.92860091e-01 -5.52518308e-01 2.10460261e-01 -1.28926229e+00
1.49444687e+00 -2.08246803e+00 -8.04295316e-02 4.51925635e-01
9.45626125e-02 2.98619360e-01 2.58799046e-01 5.91142356e-01
-5.73539212e-02 2.23870441e-01 2.04500735e-01 3.67358653e-03
-5.59753701e-02 1.70186520e-01 8.76625720e-03 5.30628324e-01
1.10512666e-01 7.14921176e-01 -1.01535368e+00 -7.38090515e-01
1.31437719e-01 -9.86103192e-02 -1.90867946e-01 -4.65748578e-01
3.21155965e-01 2.41487905e-01 -7.48450875e-01 8.16206694e-01
3.95226836e-01 2.73610741e-01 5.68777740e-01 1.66074187e-01
-7.83797979e-01 7.03963459e-01 -1.19040704e+00 1.27095139e+00
-4.51384813e-01 4.66400564e-01 1.33634984e-01 -7.83190310e-01
1.11373961e+00 4.95129168e-01 2.16893882e-01 -1.93269432e-01
3.15060109e-01 5.11749685e-01 -1.23182006e-01 -5.48453927e-01
9.65630233e-01 -2.36618742e-02 -8.55989575e-01 2.47281194e-02
6.97062731e-01 2.48369053e-01 5.35650492e-01 8.08659866e-02
7.14282155e-01 7.19033107e-02 9.39874768e-01 -8.82484376e-01
7.27635622e-01 1.39568463e-01 5.85258603e-01 5.12702763e-01
-1.59948453e-01 2.38387525e-01 4.74265784e-01 -4.45709914e-01
-9.73237991e-01 -5.75471640e-01 -2.69492239e-01 1.46677089e+00
-3.42428833e-01 -6.62970781e-01 -7.81377912e-01 -1.01038122e+00
-2.80731976e-01 7.97250867e-01 -5.85208297e-01 5.98459482e-01
-7.54548192e-01 -6.59315765e-01 6.70065761e-01 3.56475770e-01
3.99315238e-01 -1.05682611e+00 -5.85893244e-02 4.44173068e-01
-7.28165656e-02 -1.29967988e+00 -2.36426756e-01 2.69787490e-01
-4.38315362e-01 -1.01089692e+00 -7.51698256e-01 -1.05076909e+00
4.88862038e-01 -2.05451056e-01 1.15095520e+00 -1.87020928e-01
9.30803418e-02 4.60333042e-02 -5.75740576e-01 -2.90906847e-01
-4.87617701e-01 5.07369459e-01 -1.13870809e-02 -9.92843136e-02
5.18094778e-01 4.97443564e-02 2.65299648e-01 3.23138565e-01
-6.18048847e-01 -5.23678541e-01 4.81351435e-01 5.64337611e-01
2.48347223e-01 -7.35901967e-02 5.35196424e-01 -1.39863336e+00
1.51909888e-01 -5.14288306e-01 -4.61838096e-01 3.33896279e-01
-1.93934605e-01 -1.13884963e-01 3.92136723e-01 -2.66503990e-01
-1.40503633e+00 7.89200589e-02 -3.27923656e-01 4.63927150e-01
-7.63525605e-01 7.14909196e-01 -3.97523850e-01 -4.35488075e-02
6.13785505e-01 4.25125100e-02 -6.70972645e-01 -6.96282387e-01
3.73767465e-01 1.00948274e+00 4.79645520e-01 -6.30195796e-01
5.55013418e-01 1.24544472e-01 -3.95605028e-01 -1.21128595e+00
-8.38351071e-01 -1.06871963e+00 -1.18235755e+00 -4.19332832e-02
9.94842291e-01 -1.04421103e+00 -1.95226610e-01 2.90405899e-01
-1.02343380e+00 1.23681940e-01 -2.11025089e-01 8.39929581e-01
-1.45718038e-01 3.12537730e-01 -7.68503189e-01 -8.68259668e-01
-1.08872116e-01 -4.22934234e-01 7.93491006e-01 1.35974303e-01
-3.34737271e-01 -1.27816916e+00 6.81423023e-02 1.80377916e-01
-8.78958479e-02 2.60504276e-01 6.54069602e-01 -1.45752382e+00
2.37106800e-01 -2.46939719e-01 -2.09156379e-01 2.13172376e-01
-3.88484411e-02 -1.12110987e-01 -5.91255844e-01 7.82711431e-02
-3.79104942e-01 -7.32624754e-02 8.72881532e-01 4.21168543e-02
2.61105657e-01 -4.48626846e-01 -5.91998935e-01 -2.98815630e-02
1.27994251e+00 1.87966302e-01 4.74062115e-01 5.86984396e-01
5.53304851e-01 8.10875654e-01 7.60282993e-01 2.68999636e-01
5.75630724e-01 5.64622343e-01 -2.39710912e-01 -2.42566839e-02
8.15494955e-02 -1.26585111e-01 3.00036818e-01 8.70169044e-01
-5.29338300e-01 -2.03414813e-01 -1.39550269e+00 1.05979955e+00
-1.60964620e+00 -1.04798913e+00 -3.91596913e-01 1.99779046e+00
8.09387922e-01 1.18429236e-01 2.00737059e-01 -1.47084519e-01
1.15310085e+00 1.72373593e-01 5.33277571e-01 -3.33000720e-01
-1.65397435e-01 3.03133965e-01 9.91540492e-01 2.81599492e-01
-1.64397860e+00 1.48904479e+00 6.86965084e+00 1.22508264e+00
-6.93802536e-01 2.41528884e-01 9.43773091e-02 5.24944603e-01
3.95744182e-02 -1.07924491e-01 -1.46042430e+00 1.97804257e-01
1.21880972e+00 -6.93630278e-02 -1.55316323e-01 8.14718366e-01
7.54538402e-02 -1.49759561e-01 -4.74455208e-01 6.03049278e-01
2.69790828e-01 -1.20413268e+00 -9.43717733e-02 2.45763376e-01
7.72483587e-01 1.36296034e-01 -4.51503158e-01 4.58843976e-01
5.01825511e-01 -4.01739001e-01 9.72562492e-01 2.57231116e-01
9.42393184e-01 -8.20488095e-01 1.09328949e+00 2.96113849e-01
-1.20413208e+00 6.34715185e-02 -4.51207578e-01 -1.18902009e-02
2.63522089e-01 4.62183535e-01 -8.87011290e-01 9.06346202e-01
6.54004335e-01 6.30810499e-01 -6.57603681e-01 1.09238899e+00
-5.24711072e-01 8.06690037e-01 -2.07102641e-01 -3.37588042e-01
5.86025238e-01 -1.76910087e-01 5.09453058e-01 1.79447389e+00
2.86158979e-01 1.39906153e-01 2.25488454e-01 1.81691438e-01
-2.07982004e-01 8.29462349e-01 -5.39238930e-01 -1.02278687e-01
5.11590898e-01 1.47457266e+00 -1.13751113e+00 -7.24388361e-01
-5.49300671e-01 7.38363802e-01 3.77604842e-01 8.95582065e-02
-3.37379396e-01 -8.38123739e-01 3.88349652e-01 4.79724854e-02
4.47417676e-01 -3.79046232e-01 2.15413328e-02 -1.31962192e+00
-3.29920888e-01 -3.43832523e-01 8.57631207e-01 -3.19036275e-01
-1.24310684e+00 7.53594160e-01 3.19557607e-01 -8.84129286e-01
-3.55798125e-01 -7.68373191e-01 -6.67501315e-02 7.55157173e-01
-1.20978653e+00 -1.37183928e+00 3.18409562e-01 2.75922656e-01
2.98776329e-01 -4.87947464e-01 1.05253839e+00 4.43409830e-01
-4.04967874e-01 2.78025985e-01 2.05356658e-01 9.11160171e-01
8.53774905e-01 -1.19970572e+00 4.88366663e-01 7.39815295e-01
2.71000147e-01 8.95435989e-01 5.67808032e-01 -7.68783212e-01
-4.80631560e-01 -9.75644648e-01 2.08841920e+00 -6.03370786e-01
1.10600710e+00 -3.62563491e-01 -6.53010130e-01 7.62357473e-01
2.49538124e-01 -3.37213814e-01 9.47588384e-01 5.96099198e-01
-3.26498687e-01 2.82251805e-01 -1.25872779e+00 2.67805129e-01
1.00195169e+00 -4.86122489e-01 -1.12022531e+00 3.92489970e-01
4.45416331e-01 -2.15791628e-01 -8.98591757e-01 -1.22906052e-01
3.01421106e-01 -5.16489685e-01 7.16017008e-01 -7.37220049e-01
-2.46412992e-01 -9.20725018e-02 -6.06452748e-02 -1.14959598e+00
-4.24309552e-01 -4.31542993e-01 7.70183206e-01 1.78186798e+00
7.31282294e-01 -7.78247118e-01 3.25100273e-01 4.60784793e-01
-3.66955966e-01 9.76189002e-02 -9.77725983e-01 -8.02235305e-01
8.79889727e-02 -2.18041554e-01 4.07285452e-01 1.42658091e+00
3.66327047e-01 4.61665899e-01 -3.15191001e-01 2.62574613e-04
1.08308949e-01 -7.29233176e-02 3.39754790e-01 -1.61394012e+00
2.36493349e-01 -2.52361566e-01 -5.27068794e-01 -4.78724271e-01
5.94357789e-01 -8.16311300e-01 1.58709120e-02 -1.41968966e+00
2.05025852e-01 -7.10682631e-01 -1.90203473e-01 7.44424164e-01
1.17995299e-01 3.63003194e-01 2.90699862e-02 2.13823423e-01
-7.93345869e-01 2.90708672e-02 7.54319429e-01 2.29455397e-01
-2.71598667e-01 -3.25660966e-02 -6.53356135e-01 1.02419806e+00
6.35790169e-01 -6.98803127e-01 6.35775089e-01 -1.98628008e-01
1.62949085e-01 -2.60415971e-01 -1.91192925e-01 -7.12872744e-01
3.78888771e-02 -3.55938971e-01 1.27240077e-01 -5.38719893e-01
3.45863551e-02 -6.88979685e-01 -5.45961224e-03 3.33542585e-01
-1.12774543e-01 -5.44267595e-02 5.05842566e-02 2.97100216e-01
-3.48553181e-01 -7.32108057e-01 5.39867520e-01 -3.83370727e-01
-1.10858262e+00 1.73834097e-02 -8.92930567e-01 1.51006758e-01
1.01707733e+00 -1.04899071e-02 -2.68061221e-01 -6.85519278e-02
-9.66004074e-01 -1.85221747e-01 4.93382454e-01 1.81407690e-01
-4.24095057e-02 -1.06796324e+00 -7.99121439e-01 -4.21832055e-02
3.76947284e-01 -6.82051957e-01 1.04968309e-01 5.74321985e-01
-7.52139747e-01 7.14108229e-01 -3.14354539e-01 1.55349523e-01
-1.22276878e+00 4.10244823e-01 -1.97059155e-01 -2.88462996e-01
-2.46217623e-01 4.88972545e-01 -2.36932337e-01 -6.90412879e-01
-1.98010921e-01 -1.91358939e-01 -9.94221389e-01 6.90599501e-01
3.87770504e-01 4.11827624e-01 2.18058765e-01 -1.37687743e+00
-7.19919920e-01 3.24301571e-01 4.63370122e-02 -3.56500149e-01
1.32469380e+00 -1.78531677e-01 -2.77012825e-01 4.94657397e-01
8.16677749e-01 7.16986537e-01 -2.02327013e-01 -4.23600197e-01
6.70225620e-01 -8.50473940e-02 -1.15406290e-01 -6.27288103e-01
-5.58675647e-01 5.59624434e-01 -1.93542227e-01 3.84406298e-01
5.04466891e-01 3.63884985e-01 2.97717631e-01 3.82205486e-01
6.14627540e-01 -1.35693932e+00 -6.12115622e-01 1.07820833e+00
4.84601140e-01 -7.59682238e-01 3.58960144e-02 -8.60605478e-01
-7.68470526e-01 9.77309346e-01 1.55004233e-01 -2.76925229e-02
7.40908325e-01 6.65912256e-02 8.46815631e-02 -6.47547245e-02
-3.47692817e-01 -8.19414079e-01 5.07176518e-01 5.86266756e-01
6.15493655e-01 2.71579325e-01 -1.11365342e+00 8.26820850e-01
-1.86047733e-01 -3.68362695e-01 4.42795545e-01 8.67509842e-01
-7.29879797e-01 -1.11464989e+00 -3.17049861e-01 3.82254869e-01
-1.18700457e+00 -3.01457465e-01 -3.45989317e-01 1.32001686e+00
4.15947169e-01 8.74479175e-01 1.11280099e-01 1.76372733e-02
2.79180884e-01 3.23369712e-01 2.60617405e-01 -1.14294374e+00
-1.01992452e+00 3.58984321e-01 9.72164154e-01 -1.29460460e-02
-7.99008071e-01 -9.92122948e-01 -1.11443663e+00 -3.04620899e-02
-5.47514260e-01 6.45070791e-01 6.27529979e-01 1.05957019e+00
5.68350814e-02 6.82828873e-02 1.10987529e-01 -5.72170675e-01
-7.66209746e-03 -1.21331882e+00 -9.69713211e-01 3.82552892e-01
-2.43112981e-01 -7.08031952e-01 -1.19537741e-01 1.84180468e-01] | [9.750544548034668, 9.620210647583008] |
d0fcf07c-7d82-4abf-9556-ebc74ff67c62 | acoustic-gait-based-person-identification | 1406.2895 | null | http://arxiv.org/abs/1406.2895v1 | http://arxiv.org/pdf/1406.2895v1.pdf | Acoustic Gait-based Person Identification using Hidden Markov Models | We present a system for identifying humans by their walking sounds. This
problem is also known as acoustic gait recognition. The goal of the system is
to analyse sounds emitted by walking persons (mostly the step sounds) and
identify those persons. These sounds are characterised by the gait pattern and
are influenced by the movements of the arms and legs, but also depend on the
type of shoe. We extract cepstral features from the recorded audio signals and
use hidden Markov models for dynamic classification. A cyclic model topology is
employed to represent individual gait cycles. This topology allows to model and
detect individual steps, leading to very promising identification rates. For
experimental validation, we use the publicly available TUM GAID database, which
is a large gait recognition database containing 3050 recordings of 305 subjects
in three variations. In the best setup, an identification rate of 65.5 % is
achieved out of 155 subjects. This is a relative improvement of almost 30 %
compared to our previous work, which used various audio features and support
vector machines. | ['Björn Schuller', 'Maximilian Kneißl', 'Jürgen T. Geiger', 'Gerhard Rigoll'] | 2014-06-11 | null | null | null | null | ['person-identification'] | ['computer-vision'] | [ 1.32852942e-01 -4.79851931e-01 1.72563940e-01 7.26970583e-02
-2.82298237e-01 -1.13634929e-01 3.85546684e-01 -8.58879834e-02
-4.99814451e-01 4.63485569e-01 8.69607404e-02 5.35503387e-01
-1.65658221e-02 -7.35920846e-01 -1.30832657e-01 -8.88024628e-01
-5.00684440e-01 4.55320626e-01 6.78628623e-01 -3.21457148e-01
3.97497267e-02 3.84552866e-01 -2.12377071e+00 2.12257504e-01
3.70710433e-01 8.00185978e-01 1.47942066e-01 1.09846401e+00
2.34895259e-01 4.26757216e-01 -8.25178564e-01 -1.38443291e-01
-3.04429799e-01 -6.22729361e-01 -5.94454706e-01 3.43146138e-02
-1.95205599e-01 1.45614192e-01 -5.47190458e-02 5.42130709e-01
7.63663888e-01 3.06011662e-02 7.72523046e-01 -1.03081131e+00
3.44521329e-02 4.15821701e-01 -1.48336112e-01 2.38434225e-01
9.16162193e-01 -1.29946068e-01 7.76221514e-01 -6.02476835e-01
3.98384422e-01 1.01296437e+00 9.99792695e-01 5.33101797e-01
-1.23709142e+00 -3.58249903e-01 -6.68755591e-01 6.27821267e-01
-1.58143020e+00 -6.47595584e-01 8.07546675e-01 -5.95669031e-01
9.20320570e-01 3.50173980e-01 9.59042430e-01 1.28276503e+00
3.01516503e-01 4.21921790e-01 8.33188057e-01 -8.37518156e-01
3.83377105e-01 -1.80883303e-01 1.90326214e-01 6.83423579e-01
2.65320152e-01 1.30501511e-02 -6.79801881e-01 -3.14194977e-01
3.89354378e-01 -2.65878677e-01 -1.14109404e-01 -1.26556918e-01
-1.10626841e+00 6.16851628e-01 -1.92864463e-01 5.99860430e-01
-6.72584295e-01 1.77720279e-01 4.78428215e-01 1.36530340e-01
-7.12465867e-02 -3.40889469e-02 5.37728099e-03 -6.76462650e-01
-8.56399000e-01 2.93553799e-01 1.10328424e+00 2.18070835e-01
3.03933024e-01 3.47434968e-01 2.28077412e-01 1.03951287e+00
4.38798487e-01 9.16151702e-01 7.32436478e-01 -8.06425512e-01
1.64866745e-01 1.20756827e-01 1.75303414e-01 -1.17918468e+00
-6.67330980e-01 -1.03758067e-01 -6.90069914e-01 -9.24502686e-02
4.19212669e-01 -1.70942247e-01 -6.53499961e-01 1.35489190e+00
1.35882407e-01 2.83906013e-01 -2.27581412e-02 6.33862495e-01
6.78607404e-01 4.69734877e-01 1.09069787e-01 -3.21838140e-01
1.61318982e+00 -1.88459128e-01 -9.70744550e-01 -2.33433638e-02
4.14041191e-01 -1.00132644e+00 8.95314455e-01 8.01255524e-01
-7.80029774e-01 -8.39024127e-01 -1.20265007e+00 6.96192324e-01
-2.09326491e-01 4.37644929e-01 1.77617475e-01 1.22101820e+00
-8.32915783e-01 7.28130758e-01 -9.89256024e-01 -6.48423612e-01
-2.01919436e-01 3.61166596e-01 -2.39854902e-01 4.75096792e-01
-1.34390998e+00 5.93973041e-01 2.15844885e-01 1.21403925e-01
-2.57771462e-01 -8.62665325e-02 -8.35529745e-01 -1.99167982e-01
-1.14321895e-01 -5.59530497e-01 9.38230336e-01 -3.87752205e-01
-1.70619130e+00 7.88410008e-01 -1.90020695e-01 -5.22888541e-01
5.87440014e-01 -1.56049505e-01 -9.13329124e-01 4.48643565e-01
1.82122253e-02 -8.09842199e-02 1.19811964e+00 -8.32916081e-01
-4.22851026e-01 -1.95665851e-01 -7.78543711e-01 -3.07958156e-01
-2.93777913e-01 2.95045912e-01 -2.76099622e-01 -5.62437832e-01
1.13100205e-02 -1.18995965e+00 5.83496392e-02 -6.64047897e-01
-2.84534574e-01 5.90300150e-02 7.38421679e-01 -9.16774452e-01
1.53447509e+00 -2.31974506e+00 3.41753811e-02 4.21322912e-01
-2.04883665e-01 1.89555317e-01 2.59490550e-01 6.82523429e-01
1.30493090e-01 -2.80014217e-01 -2.80809939e-01 -3.55421096e-01
-1.49871502e-02 4.90535825e-01 -8.75536799e-02 4.59991753e-01
3.53017449e-02 2.86233634e-01 -4.56496000e-01 -6.08231902e-01
4.73084629e-01 6.89032316e-01 -2.97239602e-01 2.55447123e-02
5.14441729e-01 6.13684915e-02 -4.21915442e-01 5.66272199e-01
2.78595686e-01 4.55858886e-01 1.06422387e-01 -6.22442104e-02
4.00017910e-02 -7.44392574e-02 -1.57629395e+00 1.06699288e+00
-4.08176363e-01 4.65841740e-01 -9.49595869e-02 -1.19101870e+00
1.28371906e+00 6.22725189e-01 6.96632326e-01 -3.65336686e-01
1.18245371e-01 3.38003099e-01 1.40844071e-02 -7.95039713e-01
2.72232562e-01 -1.86697528e-01 -3.28482419e-01 8.37740749e-02
1.96041688e-01 1.86013058e-01 4.80599761e-01 -3.97462696e-01
1.11086810e+00 -1.17415488e-01 4.27144855e-01 -1.82972938e-01
7.31694698e-01 -3.53097320e-01 4.78057116e-01 5.47977567e-01
-3.10172766e-01 5.59543014e-01 2.44966626e-01 -3.65741253e-01
-5.90585589e-01 -1.19377398e+00 -1.02846213e-01 7.88094521e-01
-2.28755251e-01 -5.85062683e-01 -7.10957289e-01 8.36395323e-02
-1.33530408e-01 3.04049343e-01 -4.77053195e-01 -3.10023516e-01
-8.74004722e-01 -9.26696777e-01 9.01903868e-01 6.09347641e-01
4.99324441e-01 -1.33763421e+00 -1.01003957e+00 7.11456597e-01
-5.17231405e-01 -1.32272589e+00 -3.48589802e-03 1.39462337e-01
-6.19294524e-01 -1.09761608e+00 -7.98462212e-01 -7.52704620e-01
-1.40402660e-01 -3.37668538e-01 9.58167791e-01 -1.19936384e-01
-7.59025753e-01 5.62187552e-01 -5.65385163e-01 -4.21454519e-01
-6.29027188e-01 -1.56133026e-01 5.11568785e-01 3.66966397e-01
3.12610120e-01 -7.64828861e-01 -2.56185561e-01 5.82128227e-01
-3.62453431e-01 -6.53022051e-01 3.29246074e-01 7.39158869e-01
3.22493523e-01 4.32861090e-01 4.33870047e-01 -3.23618531e-01
6.53188646e-01 -1.14107490e-01 -1.04693532e-01 -7.06252456e-02
-4.75026257e-02 -2.57087141e-01 6.63861275e-01 -4.62733150e-01
-7.15461850e-01 3.72856766e-01 -5.48727930e-01 -5.81336655e-02
-6.43895090e-01 1.40385553e-01 -1.22583941e-01 -1.34175867e-01
6.57922924e-01 4.82696772e-01 8.88902321e-02 -5.63363314e-01
-1.70649916e-01 8.23117256e-01 8.93605947e-01 -4.32983160e-01
5.12253284e-01 3.77396584e-01 1.97245419e-01 -1.80989003e+00
1.31777734e-01 -7.29353964e-01 -7.38821864e-01 -7.06682861e-01
1.04647648e+00 -4.82255191e-01 -9.08208370e-01 8.22380006e-01
-8.17678750e-01 6.35918882e-03 -1.69521809e-01 7.14813828e-01
-9.33303833e-01 6.20124638e-01 -7.90997446e-01 -1.27207696e+00
-2.76054978e-01 -7.49978483e-01 1.05894780e+00 -1.01061147e-02
-7.30065882e-01 -8.98701429e-01 2.23446757e-01 2.35097393e-01
2.20370770e-01 5.84267557e-01 6.11470401e-01 -4.33621794e-01
3.89445037e-01 -5.37618160e-01 6.92041397e-01 2.39789397e-01
2.61000603e-01 3.38094056e-01 -9.95886683e-01 -5.01415543e-02
-9.21053067e-03 3.26347463e-02 8.44633758e-01 5.58586836e-01
3.50083143e-01 5.86376600e-02 -2.64830083e-01 6.26980886e-02
1.07200611e+00 5.21547496e-01 7.56149471e-01 2.36721307e-01
3.54065269e-01 6.10424936e-01 5.44870675e-01 6.31565809e-01
-6.98060542e-02 1.09527087e+00 -1.35205183e-02 2.04744428e-01
-7.90643394e-02 -2.21772376e-03 7.47950792e-01 9.38577592e-01
-6.49676085e-01 3.50317918e-04 -1.14361918e+00 4.00992572e-01
-1.50124657e+00 -1.26901555e+00 -4.07069206e-01 2.22193861e+00
3.04877192e-01 3.30384433e-01 8.82691860e-01 1.16866469e+00
7.47933388e-01 1.71154495e-02 3.31349462e-01 -5.78291595e-01
9.97409746e-02 4.84359473e-01 1.82919130e-01 3.67689013e-01
-1.35231555e+00 4.78964657e-01 6.86570597e+00 6.15742028e-01
-1.05340290e+00 -1.80530697e-01 -1.93762511e-01 1.96422100e-01
6.64366007e-01 -6.27942443e-01 -7.07201898e-01 7.83983648e-01
1.46428514e+00 -8.66886824e-02 7.08541321e-03 5.80550790e-01
4.61147070e-01 -1.66041672e-01 -5.99932551e-01 1.04954255e+00
1.35309458e-01 -5.86011529e-01 -3.04356843e-01 9.54707861e-02
3.73053364e-03 -4.66005981e-01 -1.27739623e-01 8.69055241e-02
-2.12669030e-01 -7.07786381e-01 6.92551434e-01 7.03275800e-01
4.76772934e-01 -9.83068466e-01 9.14421260e-01 2.59488821e-01
-1.72285414e+00 -2.03432500e-01 -1.80809602e-01 -1.49178684e-01
5.12775540e-01 5.49450696e-01 -6.77470982e-01 5.74440122e-01
9.67322588e-01 6.21357441e-01 -4.27679360e-01 1.23618007e+00
6.86080288e-03 1.17272937e+00 -5.69009662e-01 -2.26620078e-01
-2.27222681e-01 -1.26061335e-01 7.60782123e-01 1.64700508e+00
5.85674286e-01 -2.08750680e-01 7.79383257e-02 3.04328799e-01
7.98941314e-01 2.47277245e-01 -6.60750151e-01 -2.73563694e-02
2.32068792e-01 8.23720574e-01 -8.30300927e-01 -1.42601982e-01
2.78804377e-02 9.81614232e-01 -6.70483053e-01 8.17065965e-03
-6.81920409e-01 -7.18884766e-01 5.17441154e-01 2.21043304e-01
3.87172759e-01 -3.30795050e-01 3.19194406e-01 -7.86311090e-01
1.06862597e-01 -5.38177252e-01 5.67222714e-01 -3.81788880e-01
-9.98921335e-01 4.98509228e-01 2.77086366e-02 -1.43889272e+00
-8.58743906e-01 -5.01063526e-01 -6.10495627e-01 5.81305265e-01
-6.57178760e-01 -8.65645826e-01 -1.33974805e-01 6.88801885e-01
3.44498962e-01 -2.17661932e-01 1.26310694e+00 5.19832194e-01
-3.19682091e-01 3.77491951e-01 -1.51462659e-01 1.61070436e-01
4.89453405e-01 -1.13783550e+00 3.38272005e-01 6.42030776e-01
3.61020088e-01 3.91054213e-01 1.12042832e+00 -5.25753856e-01
-1.10306132e+00 -7.19997168e-01 1.35949135e+00 -2.60063291e-01
6.09682381e-01 -2.60693014e-01 -8.55034113e-01 1.02317728e-01
-2.76210517e-01 -2.09945306e-01 9.26030815e-01 -7.23608285e-02
8.53415728e-02 -1.64510757e-01 -9.84996796e-01 1.91437572e-01
8.17735195e-01 -3.87809008e-01 -8.60048890e-01 -7.52926841e-02
3.85411619e-03 7.72407502e-02 -8.78555238e-01 2.53978223e-01
9.14400756e-01 -1.17302954e+00 1.26999211e+00 -8.54568481e-02
-1.32535949e-01 -2.71404535e-01 -1.55874923e-01 -1.15052354e+00
-4.62900609e-01 -5.02890527e-01 -9.06520933e-02 1.34918606e+00
-3.79846171e-02 -7.16217041e-01 7.02455878e-01 -9.06588361e-02
2.61790186e-01 -1.31503195e-01 -1.12479544e+00 -1.18078911e+00
-4.84739393e-01 -5.78324616e-01 1.23022199e-01 5.98614275e-01
4.24970165e-02 2.06358850e-01 -7.30305314e-01 3.62694599e-02
8.99915338e-01 -1.50676712e-01 4.88960087e-01 -1.53319514e+00
-4.63326335e-01 -2.42568254e-01 -1.25420952e+00 -2.01583311e-01
-1.40602496e-02 -3.48303050e-01 1.14026748e-01 -1.15208459e+00
-2.82796144e-01 8.42232034e-02 -1.91955358e-01 3.01429719e-01
1.87125966e-01 5.99381030e-01 1.58709899e-01 6.25111312e-02
-1.56519696e-01 2.71715403e-01 5.32301188e-01 1.41901802e-02
-4.28055555e-01 5.65626860e-01 1.17441155e-01 9.31644261e-01
8.05993915e-01 -3.40141863e-01 -1.12324253e-01 3.27155054e-01
-2.76589483e-01 2.30614588e-01 5.29415131e-01 -1.77982903e+00
1.57237891e-02 2.15523750e-01 4.02391016e-01 -5.47737539e-01
7.36091256e-01 -6.47291541e-01 6.58522666e-01 1.18323851e+00
8.65391716e-02 -6.29968150e-03 4.75103967e-02 5.59343815e-01
-3.79535764e-01 -1.19589217e-01 8.33320200e-01 5.32091074e-02
-9.01112199e-01 -3.19839031e-01 -1.07996345e+00 -3.69700998e-01
8.20100963e-01 -4.60747719e-01 2.29979098e-01 -4.53450233e-01
-1.35491753e+00 -3.85298193e-01 8.47532749e-02 2.78352261e-01
6.10721469e-01 -1.32802582e+00 -6.42491102e-01 4.75724161e-01
7.37473592e-02 -1.01642668e+00 2.86902815e-01 8.12620878e-01
-6.05775356e-01 4.05420691e-01 -5.12450159e-01 -7.38327682e-01
-1.93856704e+00 1.06335737e-01 3.40881497e-01 -2.70348229e-02
-7.88749218e-01 4.48471606e-01 -7.40323722e-01 1.66693740e-02
3.62450108e-02 -1.08530551e-01 -8.34793687e-01 4.76297677e-01
6.38569593e-01 7.13240325e-01 2.09825844e-01 -1.07968473e+00
-7.59897768e-01 1.04087651e+00 5.01711547e-01 -2.57612646e-01
1.12755620e+00 1.42125534e-02 2.23485917e-01 8.53236377e-01
1.04094434e+00 8.25502053e-02 -3.52528006e-01 1.12611979e-01
1.42720118e-01 -2.73781955e-01 -3.90792161e-01 -2.10767657e-01
-7.04674006e-01 9.32843745e-01 9.50492501e-01 5.73387682e-01
1.14742994e+00 -2.83631533e-01 8.38396966e-01 2.65723795e-01
7.85067558e-01 -1.09938836e+00 -4.66113008e-04 2.66899824e-01
7.34470546e-01 -5.00979483e-01 -3.77042562e-01 -5.74016929e-01
-6.23794079e-01 1.26909137e+00 -7.46318623e-02 -2.91793495e-01
7.20535934e-01 4.44601417e-01 4.34594415e-02 1.98799178e-01
-5.25190175e-01 -4.87434357e-01 1.16856903e-01 1.00734043e+00
2.10835382e-01 3.06640655e-01 -5.11609137e-01 8.01772237e-01
-6.19049609e-01 1.74873978e-01 3.67751330e-01 1.02171052e+00
-6.73513472e-01 -1.05403960e+00 -8.28938663e-01 2.70435810e-01
-5.53668320e-01 4.76779252e-01 -3.68694335e-01 6.91192865e-01
6.90783784e-02 1.28104043e+00 -1.66651905e-01 -7.38484979e-01
7.38987088e-01 4.53722209e-01 4.14399058e-01 -2.31121078e-01
-3.94865125e-01 1.10663839e-01 4.10790652e-01 -6.02585077e-01
-5.71526468e-01 -1.05969059e+00 -1.27847099e+00 -2.19946504e-01
7.15530217e-02 3.39581013e-01 5.20837903e-01 7.14033484e-01
-1.25901327e-01 5.05167246e-01 5.95729411e-01 -1.01193917e+00
-4.22052920e-01 -7.22854078e-01 -1.14520323e+00 4.45371211e-01
1.72390208e-01 -1.00002992e+00 -4.75886762e-01 5.35371780e-01] | [14.143410682678223, 1.511596441268921] |
f486384a-10ac-454d-909f-cdd171ef271e | semantic-instance-segmentation-of-3d-scenes | 2206.01203 | null | https://arxiv.org/abs/2206.01203v2 | https://arxiv.org/pdf/2206.01203v2.pdf | Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes | Current 3D segmentation methods heavily rely on large-scale point-cloud datasets, which are notoriously laborious to annotate. Few attempts have been made to circumvent the need for dense per-point annotations. In this work, we look at weakly-supervised 3D semantic instance segmentation. The key idea is to leverage 3D bounding box labels which are easier and faster to annotate. Indeed, we show that it is possible to train dense segmentation models using only bounding box labels. At the core of our method, \name{}, lies a deep model, inspired by classical Hough voting, that directly votes for bounding box parameters, and a clustering method specifically tailored to bounding box votes. This goes beyond commonly used center votes, which would not fully exploit the bounding box annotations. On ScanNet test, our weakly supervised model attains leading performance among other weakly supervised approaches (+18 mAP@50). Remarkably, it also achieves 97% of the mAP@50 score of current fully supervised models. To further illustrate the practicality of our work, we train Box2Mask on the recently released ARKitScenes dataset which is annotated with 3D bounding boxes only, and show, for the first time, compelling 3D instance segmentation masks. | ['Gerard Pons-Moll', 'Tuan Anh Tran', 'Francis Engelmann', 'Julian Chibane'] | 2022-06-02 | null | null | null | null | ['3d-instance-segmentation-1', '3d-semantic-instance-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.54062840e-02 4.46455568e-01 -2.45389044e-01 -4.02536511e-01
-1.16804850e+00 -1.05697155e+00 6.41873121e-01 6.33912235e-02
-3.34401727e-01 2.36502618e-01 -2.18525320e-01 -5.14576375e-01
2.82302320e-01 -5.76115549e-01 -8.77773523e-01 -4.33467627e-01
6.81106374e-02 9.78984296e-01 6.78374648e-01 -4.46623489e-02
2.78034300e-01 6.55281901e-01 -1.35477245e+00 -6.30306080e-02
7.50162125e-01 1.07005572e+00 -9.69072524e-03 5.02164781e-01
-3.28994930e-01 1.95794091e-01 -3.90099525e-01 -4.53232706e-01
6.37808442e-01 2.00708974e-02 -9.07328010e-01 2.27636039e-01
7.36227274e-01 -3.59732807e-01 7.99317509e-02 8.25744450e-01
1.84450045e-01 -6.66133761e-02 7.75052309e-01 -1.01429200e+00
-1.19485736e-01 3.49261105e-01 -8.88178587e-01 -5.38742542e-02
-3.07089072e-02 9.15669054e-02 1.21329963e+00 -9.43767786e-01
5.82710147e-01 9.01521206e-01 9.24002111e-01 4.65957344e-01
-1.24614942e+00 -5.58670640e-01 1.95865020e-01 -5.09565771e-01
-1.33372879e+00 -3.79611254e-02 7.84068704e-01 -5.57377815e-01
9.05013978e-01 2.42730573e-01 4.69910443e-01 7.05430865e-01
-5.07650971e-01 9.07265902e-01 1.22146535e+00 -3.57225865e-01
2.89149225e-01 7.47359032e-03 2.75921822e-01 7.90294349e-01
3.13348807e-02 -1.29356887e-02 -7.87323937e-02 -1.56939272e-02
1.04848897e+00 -3.87616875e-03 1.47109658e-01 -9.09009218e-01
-1.24516058e+00 8.44659567e-01 6.52943611e-01 7.40915090e-02
1.55821905e-01 2.82706082e-01 2.26051211e-01 -2.49412417e-01
8.67703557e-01 5.02002478e-01 -7.62372732e-01 -2.43384857e-02
-1.25641954e+00 2.47436821e-01 6.59363329e-01 1.14689851e+00
9.46240306e-01 -4.56698716e-01 1.96724907e-02 7.16735363e-01
2.21528232e-01 4.56222117e-01 -5.90401366e-02 -1.20613003e+00
4.24943507e-01 8.11756372e-01 8.39845389e-02 -4.95397806e-01
-5.79295456e-01 -3.29141676e-01 -3.72458398e-01 3.85250330e-01
7.77962685e-01 1.05126962e-01 -1.30268741e+00 1.23480189e+00
5.92665672e-01 1.03259474e-01 -3.34473968e-01 9.93139088e-01
7.05614209e-01 2.89653242e-01 1.24761295e-02 4.51786280e-01
1.28876591e+00 -1.05852699e+00 7.75230452e-02 -2.33767331e-01
9.28045332e-01 -7.02444077e-01 1.29099703e+00 2.20862478e-01
-1.02651823e+00 -3.91306221e-01 -8.48914504e-01 -2.99731195e-01
-3.54658008e-01 1.40490741e-01 8.07634234e-01 7.07683921e-01
-8.54808450e-01 6.06192827e-01 -1.07653952e+00 -2.66486526e-01
8.09389234e-01 3.04303348e-01 -3.36653948e-01 7.72880092e-02
-5.26505172e-01 7.37654090e-01 3.16822112e-01 -1.24096349e-01
-6.96490288e-01 -8.37749004e-01 -8.01045418e-01 -2.36917436e-01
6.24092817e-01 -4.30732489e-01 1.40678668e+00 -5.52150071e-01
-1.25930357e+00 1.43548453e+00 -7.48736188e-02 -5.26592970e-01
8.46396327e-01 -3.28190565e-01 2.80532032e-01 2.60998994e-01
3.05174768e-01 1.24695778e+00 6.20662630e-01 -1.39771867e+00
-5.90744615e-01 -4.59443212e-01 7.92254582e-02 -1.13734137e-02
4.39679883e-02 -2.25826338e-01 -8.52260709e-01 -4.99336571e-01
4.18855697e-01 -9.88892913e-01 -5.36549389e-01 2.15056732e-01
-7.41406739e-01 -4.83663678e-01 8.08755696e-01 -3.56044412e-01
7.09006250e-01 -2.11318994e+00 -2.12429285e-01 4.86546189e-01
3.42650235e-01 7.95936137e-02 2.07450494e-01 6.62050322e-02
7.65749216e-02 3.65760148e-01 -7.48804033e-01 -6.07723117e-01
3.53496969e-01 3.61557245e-01 -3.91915768e-01 5.82980812e-01
3.64291459e-01 1.13866615e+00 -7.24539757e-01 -6.23745799e-01
5.34546018e-01 2.04181254e-01 -7.82704353e-01 -3.96650992e-02
-5.65506518e-01 4.79247689e-01 -4.29335028e-01 8.82421017e-01
7.48026431e-01 -5.59705138e-01 -2.31721431e-01 -1.69891387e-01
-2.50907779e-01 1.50290146e-01 -1.01973224e+00 2.07204723e+00
-1.69616014e-01 4.29955453e-01 3.02018598e-03 -9.11151350e-01
8.74962687e-01 -5.59898168e-02 4.99563813e-01 -4.07331645e-01
4.90657091e-02 5.63810647e-01 -3.22526634e-01 -5.73624708e-02
3.15055877e-01 -2.37807751e-01 -3.83227825e-01 3.51835728e-01
5.11781909e-02 -8.76579225e-01 -6.90972656e-02 1.58307359e-01
1.04227877e+00 5.91227531e-01 1.49771169e-01 -2.94782341e-01
2.25242555e-01 5.31639218e-01 2.86347032e-01 7.51941741e-01
-2.18466207e-01 1.12420917e+00 6.16256654e-01 -4.51448023e-01
-1.46731031e+00 -1.12803340e+00 -4.15346503e-01 8.84817958e-01
3.98690671e-01 -3.29888344e-01 -9.46629822e-01 -1.13533688e+00
1.22347280e-01 3.72136891e-01 -6.10935688e-01 4.02821392e-01
-5.91782928e-01 -4.45706517e-01 6.39845848e-01 8.22396636e-01
5.52559376e-01 -5.94757378e-01 -5.19852996e-01 -1.16125882e-01
2.49436677e-01 -1.36860108e+00 -3.06964517e-01 4.14096028e-01
-1.14894986e+00 -1.13261175e+00 -9.64120388e-01 -6.39197826e-01
7.44228959e-01 2.17880905e-01 1.21438384e+00 1.40179917e-01
-2.25564688e-01 2.69130975e-01 -3.93664509e-01 -3.03562462e-01
6.42102957e-02 4.10746753e-01 -3.61381948e-01 -4.68736231e-01
4.88358527e-01 -4.84960288e-01 -6.76610172e-01 5.59767127e-01
-5.81794500e-01 1.87162280e-01 6.34717107e-01 6.41462564e-01
9.34525251e-01 -3.49525034e-01 1.24032490e-01 -1.22947431e+00
-1.99777126e-01 -2.24800676e-01 -6.87157512e-01 -1.84968263e-01
-3.71191204e-01 -1.44196033e-01 3.11864823e-01 2.69528013e-04
-7.34326780e-01 5.02315521e-01 -5.61768293e-01 -4.86702263e-01
-6.91410184e-01 -6.80848360e-02 -4.21281159e-02 -1.32201672e-01
5.93516946e-01 -2.36361891e-01 -2.37695754e-01 -7.52585888e-01
7.96642303e-01 5.09685278e-01 6.91407204e-01 -8.25418353e-01
1.20503759e+00 9.82947886e-01 -2.75222193e-02 -5.91929436e-01
-1.22333753e+00 -7.31923699e-01 -1.12413573e+00 8.18946958e-02
1.16878343e+00 -8.72433782e-01 -4.83757287e-01 1.39117286e-01
-1.07916713e+00 -7.15358794e-01 -4.32158232e-01 1.82079822e-01
-6.81016326e-01 3.97940397e-01 -5.12805402e-01 -6.39725447e-01
-5.43021336e-02 -9.87914383e-01 1.76430428e+00 -1.00300863e-01
-3.42575848e-01 -7.84957051e-01 -3.64140458e-02 7.11549282e-01
-3.05578858e-03 4.95432496e-01 7.20629454e-01 -8.70718539e-01
-7.80898809e-01 -3.21980476e-01 -6.11740589e-01 2.81991899e-01
-2.91475594e-01 -4.67410460e-02 -1.21664989e+00 1.09919474e-01
-3.43116641e-01 -3.23584765e-01 9.31363225e-01 3.48245293e-01
1.44504547e+00 2.36112133e-01 -4.70909238e-01 7.90040612e-01
1.18782806e+00 -2.90979505e-01 3.50381106e-01 2.76008844e-01
1.01383352e+00 5.97846866e-01 6.15678549e-01 1.80260003e-01
4.14935470e-01 6.90939188e-01 5.79788983e-01 -5.08419454e-01
-6.79459870e-02 -5.94569325e-01 -2.24560753e-01 3.77457172e-01
-1.79164454e-01 3.43450950e-03 -1.21468687e+00 5.37872434e-01
-1.74091446e+00 -4.70371753e-01 -4.65877473e-01 1.92819250e+00
6.51312232e-01 6.40547752e-01 2.17162862e-01 1.46670729e-01
5.27750194e-01 1.19949564e-01 -6.10191822e-01 -1.07504681e-01
2.46036872e-02 5.65358877e-01 8.46835136e-01 4.95622516e-01
-1.54822826e+00 1.29114318e+00 6.02387190e+00 9.49296534e-01
-7.03943431e-01 2.39428848e-01 7.48995185e-01 -5.44145443e-02
-2.85232723e-01 2.43742496e-01 -7.98810244e-01 3.78796816e-01
4.62820619e-01 6.48071766e-01 -8.30822214e-02 1.17670798e+00
1.74403004e-02 -1.88861966e-01 -1.26423907e+00 8.66264105e-01
-1.13399208e-01 -1.36979806e+00 -1.80605114e-01 2.12881625e-01
9.45521235e-01 3.40936661e-01 -3.54694985e-02 3.53426278e-01
4.79796231e-01 -1.04534364e+00 8.00927520e-01 -3.67904715e-02
8.68238270e-01 -6.57999039e-01 5.64234495e-01 5.39725184e-01
-1.15681887e+00 2.54797071e-01 -2.52507508e-01 -4.04962115e-02
3.13280642e-01 7.03588963e-01 -7.93673754e-01 2.77582794e-01
8.73653173e-01 6.11353397e-01 -5.12800336e-01 1.02694035e+00
-5.04105926e-01 6.19737864e-01 -7.97027767e-01 2.14931786e-01
6.49214447e-01 -2.05424353e-02 3.99127841e-01 1.21784055e+00
1.23062320e-01 1.53931499e-01 3.94076049e-01 1.03237581e+00
-2.82860458e-01 -8.92917961e-02 -5.27678847e-01 2.89837807e-01
3.42577010e-01 1.32132828e+00 -1.24181354e+00 -2.70172834e-01
-3.86952758e-01 9.00599062e-01 3.56137633e-01 9.92829055e-02
-8.48268509e-01 -1.85001418e-01 3.93265694e-01 1.96087822e-01
4.71444458e-01 -5.39303303e-01 -8.63672256e-01 -1.02505469e+00
4.05544974e-02 -3.07779551e-01 2.40882620e-01 -7.22885847e-01
-1.36800683e+00 1.67181239e-01 4.70729843e-02 -1.06076574e+00
7.54199475e-02 -7.82960415e-01 -5.84576190e-01 6.49199605e-01
-1.63220489e+00 -1.47094965e+00 -3.56858611e-01 4.11137760e-01
5.29466987e-01 3.39461923e-01 4.37737465e-01 3.15714180e-01
-3.92904162e-01 3.37259769e-01 -1.84267730e-01 3.09313148e-01
5.75206339e-01 -1.65099597e+00 7.98022866e-01 6.20093405e-01
4.80908126e-01 4.38305855e-01 4.35618550e-01 -5.88752508e-01
-9.40359831e-01 -1.07401705e+00 6.48789823e-01 -1.06423974e+00
6.74079478e-01 -7.07681835e-01 -8.89869213e-01 6.83257818e-01
-4.37164217e-01 3.11012566e-01 5.50101101e-01 9.33800489e-02
-4.73950446e-01 2.27960974e-01 -1.07089615e+00 4.05957848e-01
1.35683990e+00 -4.67865884e-01 -7.04805970e-01 5.35151005e-01
8.49548221e-01 -7.92459190e-01 -6.97100520e-01 6.99531496e-01
3.30987871e-01 -1.02895224e+00 1.20986938e+00 -4.16135490e-01
4.54891324e-01 -5.09800076e-01 -1.37858868e-01 -7.09853768e-01
2.41309553e-01 -3.57369810e-01 -4.09575738e-02 1.10450363e+00
6.01770639e-01 -2.87318766e-01 1.40719414e+00 7.67785072e-01
-4.92563695e-01 -7.99607992e-01 -9.54032898e-01 -8.14662755e-01
3.91877562e-01 -9.01982903e-01 5.98852634e-01 9.31822658e-01
-3.73854965e-01 4.12690230e-02 7.68319592e-02 1.92222267e-01
7.50404358e-01 3.51902336e-01 1.09539068e+00 -1.39419138e+00
-2.98018992e-01 -6.14306450e-01 -2.77924269e-01 -1.62613809e+00
9.73313078e-02 -9.87154305e-01 1.80087328e-01 -1.46024227e+00
8.71161297e-02 -9.41437125e-01 1.87471241e-01 6.42611802e-01
-7.89961405e-03 9.58276212e-01 9.44248438e-02 4.55681682e-01
-7.80812740e-01 2.36654878e-01 1.21151352e+00 4.53445911e-02
-2.56770909e-01 6.92545027e-02 -4.37302649e-01 1.19332910e+00
7.07676589e-01 -3.93128574e-01 -1.23492572e-02 -5.96918941e-01
6.64621741e-02 -3.40754718e-01 7.21157551e-01 -7.75363207e-01
2.25721851e-01 1.32545620e-01 4.24847543e-01 -1.08827019e+00
4.22341645e-01 -9.11152244e-01 -3.20474744e-01 2.65954584e-02
-7.45905787e-02 -3.62819403e-01 1.08028315e-01 5.75655162e-01
1.63909327e-02 -2.13640630e-01 6.88822865e-01 -2.52018571e-01
-7.56597221e-01 5.05668044e-01 2.27058336e-01 1.49048209e-01
1.11532521e+00 -4.39236015e-01 -4.36712839e-02 1.50146410e-01
-7.92754531e-01 3.47071826e-01 8.20583642e-01 3.46212387e-02
3.19569468e-01 -9.58793700e-01 -3.61168981e-01 5.65388650e-02
2.22520500e-01 8.11453938e-01 1.88319404e-02 7.96717644e-01
-7.56767333e-01 4.37747002e-01 4.04265910e-01 -1.08197141e+00
-8.81908000e-01 2.97081769e-01 3.17050099e-01 -1.38363451e-01
-8.93502712e-01 9.79953647e-01 3.09438527e-01 -8.51176381e-01
1.67065591e-01 -5.79366922e-01 2.17350274e-01 -1.06118828e-01
-2.16803886e-02 2.03262180e-01 1.48497373e-01 -4.58046168e-01
-4.64064837e-01 1.06958032e+00 5.80747388e-02 -1.92142814e-01
1.33458614e+00 1.28249556e-01 1.69271350e-01 2.57134289e-01
1.09251416e+00 -6.71363845e-02 -1.70142281e+00 -1.53097555e-01
2.71945298e-01 -5.24362922e-01 5.09664975e-03 -7.71378934e-01
-8.79107416e-01 1.20658159e+00 3.00992191e-01 2.50140160e-01
5.50346971e-01 6.00512147e-01 9.56498384e-01 3.56089473e-01
4.36173677e-01 -1.01053405e+00 -3.79264802e-02 4.57351089e-01
4.02675986e-01 -1.56533468e+00 -5.25127091e-02 -6.16265059e-01
-3.92641068e-01 1.04036295e+00 6.43882275e-01 -4.86384809e-01
4.96343523e-01 2.64120817e-01 -1.35930553e-02 -4.36746895e-01
-9.84206498e-02 -5.73629022e-01 4.27260756e-01 5.44279993e-01
1.84300244e-01 -7.59172067e-03 1.52520267e-02 6.09677434e-01
-3.23515892e-01 -3.63064855e-01 -2.28265859e-02 7.40131974e-01
-6.53933406e-01 -1.09091425e+00 -4.14461017e-01 4.08007830e-01
-2.95609117e-01 2.17692219e-02 -5.58147848e-01 1.09208226e+00
2.27907047e-01 5.83922744e-01 2.98123956e-01 -1.11181617e-01
4.40261781e-01 9.05102640e-02 3.93729895e-01 -7.99414277e-01
-3.29731882e-01 1.90514341e-01 -1.16530105e-01 -5.80941737e-01
-4.41732556e-01 -5.64051390e-01 -1.57727194e+00 -2.25107059e-01
-3.12555730e-01 1.97945815e-02 7.14310050e-01 1.00295711e+00
2.11300235e-02 1.08542442e-01 3.82913649e-01 -1.29856181e+00
-3.28862578e-01 -6.59737110e-01 -4.21630621e-01 3.77919704e-01
1.64992899e-01 -6.98793292e-01 -5.32430828e-01 -3.66331562e-02] | [8.084687232971191, -3.103752374649048] |
b693b11e-ba82-490c-b20c-4db9da025102 | projb-an-improved-bilinear-biased-proje-model | 2209.02390 | null | https://arxiv.org/abs/2209.02390v2 | https://arxiv.org/pdf/2209.02390v2.pdf | ProjB: An Improved Bilinear Biased ProjE model for Knowledge Graph Completion | Knowledge Graph Embedding (KGE) methods have gained enormous attention from a wide range of AI communities including Natural Language Processing (NLP) for text generation, classification and context induction. Embedding a huge number of inter-relationships in terms of a small number of dimensions, require proper modeling in both cognitive and computational aspects. Recently, numerous objective functions regarding cognitive and computational aspects of natural languages are developed. Among which are the state-of-the-art methods of linearity, bilinearity, manifold-preserving kernels, projection-subspace, and analogical inference. However, the major challenge of such models lies in their loss functions that associate the dimension of relation embeddings to corresponding entity dimension. This leads to inaccurate prediction of corresponding relations among entities when counterparts are estimated wrongly. ProjE KGE, published by Bordes et al., due to low computational complexity and high potential for model improvement, is improved in this work regarding all translative and bilinear interactions while capturing entity nonlinearity. Experimental results on benchmark Knowledge Graphs (KGs) such as FB15K and WN18 show that the proposed approach outperforms the state-of-the-art models in entity prediction task using linear and bilinear methods and other recent powerful ones. In addition, a parallel processing structure is proposed for the model in order to improve the scalability on large KGs. The effects of different adaptive clustering and newly proposed sampling approaches are also explained which prove to be effective in improving the accuracy of knowledge graph completion. | ['Farhana Zulkernine', 'Sahar Vahdati', 'Mojtaba Moattari'] | 2022-08-15 | null | null | null | null | ['knowledge-graph-embedding'] | ['graphs'] | [-1.73058659e-01 2.60632455e-01 -1.06386051e-01 -1.35610044e-01
-6.18111640e-02 -4.35020238e-01 6.64767921e-01 5.32001197e-01
-4.92054611e-01 8.25944841e-01 1.67549655e-01 -2.09140852e-01
-7.72778511e-01 -1.07278609e+00 -6.41709328e-01 -4.69083160e-01
-2.28218362e-01 6.83190584e-01 1.11457154e-01 -3.00656468e-01
2.19418667e-02 3.70946795e-01 -1.41474342e+00 1.85353458e-01
1.33940721e+00 8.85117590e-01 3.97816077e-02 3.42645943e-01
-3.95296007e-01 7.96559095e-01 -1.26332983e-01 -1.06654477e+00
-2.69460734e-02 -2.41241399e-02 -8.83720994e-01 -4.67701107e-01
2.19793946e-01 2.28436545e-01 -4.69959408e-01 1.06436038e+00
5.26994109e-01 2.65761822e-01 8.58940125e-01 -1.31933033e+00
-1.25660813e+00 8.20906997e-01 -3.04871291e-01 -1.82610434e-02
2.63915598e-01 -4.77704644e-01 1.14948058e+00 -1.22907364e+00
6.07021451e-01 1.35926557e+00 9.13740218e-01 2.76292153e-02
-1.29607272e+00 -5.78118801e-01 7.12244064e-02 9.29188490e-01
-1.60144544e+00 6.31169528e-02 9.08209324e-01 -5.17319560e-01
1.07474577e+00 2.49639124e-01 5.72524488e-01 8.75257850e-01
5.37624233e-04 4.55997199e-01 1.14021766e+00 -6.01180732e-01
-1.96193419e-02 6.87409997e-01 5.18866777e-01 9.31456089e-01
4.93994743e-01 -1.43728808e-01 -6.91714644e-01 -2.83852071e-01
3.83533984e-01 -2.10660577e-01 -3.38765889e-01 -5.40923297e-01
-1.23692632e+00 8.43893468e-01 6.46651089e-01 4.56682563e-01
-3.77789795e-01 -2.20619351e-01 3.21931988e-01 9.28400308e-02
5.74115276e-01 5.23339450e-01 -5.34305632e-01 1.73554957e-01
-5.45613706e-01 2.14344114e-01 9.08074677e-01 1.10450244e+00
7.33408332e-01 -1.24466732e-01 -1.91078752e-01 9.00545597e-01
1.13412842e-01 3.30275148e-01 4.20526206e-01 -2.89398819e-01
6.44221842e-01 1.18151069e+00 -1.29935443e-01 -1.71698785e+00
-6.56276882e-01 -6.50343120e-01 -1.28972816e+00 -3.14922988e-01
1.51508436e-01 2.00728074e-01 -4.34140652e-01 1.59669840e+00
4.98508453e-01 3.40907693e-01 3.47238570e-01 6.05341196e-01
8.97249401e-01 5.20765841e-01 2.44157478e-01 -4.90825325e-02
1.54945946e+00 -8.89831722e-01 -9.23557281e-01 5.50380489e-03
7.50658512e-01 -6.74374580e-01 1.00862920e+00 9.24477354e-02
-7.06538856e-01 -5.44082165e-01 -8.25009823e-01 -2.15223953e-01
-1.06866145e+00 4.48093414e-01 1.08602405e+00 6.70602858e-01
-9.81257677e-01 5.09748101e-01 -4.90457326e-01 -4.32909667e-01
2.35988796e-01 3.58938783e-01 -5.67963481e-01 -1.05004027e-01
-1.70927978e+00 1.05326271e+00 9.69128788e-01 2.39471927e-01
-2.56932788e-02 -8.66153598e-01 -8.67999077e-01 2.86653936e-01
4.52125907e-01 -8.70389998e-01 1.03407808e-01 -1.82482824e-01
-1.16113484e+00 5.58492541e-01 1.53840527e-01 -5.82796752e-01
3.06272447e-01 -2.83289820e-01 -6.77256942e-01 1.35519532e-02
-2.25548148e-01 4.61723596e-01 4.93202835e-01 -9.96245623e-01
-2.96160161e-01 -6.22914791e-01 1.90338492e-01 3.69854659e-01
-9.51008379e-01 -3.32696259e-01 -3.07338625e-01 -6.25156999e-01
1.39512807e-01 -1.07266319e+00 7.40814954e-02 -4.07794058e-01
-5.16086161e-01 -4.91287231e-01 6.33559883e-01 -9.87353146e-01
1.41790831e+00 -1.74594319e+00 5.10437608e-01 2.77343243e-01
2.14649126e-01 4.07279342e-01 1.01054594e-01 5.26871324e-01
-5.73196299e-02 1.44295871e-01 -1.09199263e-01 -1.79431677e-01
1.69538438e-01 2.22248077e-01 -2.90928811e-01 1.73240453e-01
1.30626187e-01 1.13997006e+00 -7.68113017e-01 -6.62454247e-01
2.26975977e-01 8.28760743e-01 -6.38303638e-01 9.63619947e-02
8.19061045e-03 1.78611577e-01 -3.71595949e-01 4.55632746e-01
5.86755693e-01 -2.75860339e-01 3.20684433e-01 -7.17128158e-01
2.34872952e-01 6.42550513e-02 -1.26451981e+00 1.54671514e+00
-4.48625267e-01 2.18854085e-01 -4.58653778e-01 -1.22925067e+00
9.76498961e-01 2.41665035e-01 1.64676979e-01 -2.81298697e-01
-1.51951715e-01 4.04844657e-02 2.94717718e-02 -4.35903907e-01
7.04394996e-01 1.62338421e-01 1.59522936e-01 -6.01403378e-02
2.13518858e-01 2.27652684e-01 3.13482285e-01 4.55018520e-01
8.23330641e-01 1.08857371e-01 3.57185364e-01 -3.42344970e-01
8.33015323e-01 -1.79468706e-01 4.18492347e-01 4.07815307e-01
1.92489102e-01 -1.55780375e-01 4.74940330e-01 -2.07661673e-01
-8.89948547e-01 -9.19099152e-01 -2.41156906e-01 9.26578820e-01
7.43619204e-02 -6.10051751e-01 -5.47187984e-01 -6.03491008e-01
1.95420623e-01 9.84548748e-01 -7.04755425e-01 -4.22925591e-01
-2.64000803e-01 -9.11045194e-01 7.17478454e-01 5.42320669e-01
6.08808577e-01 -9.76139367e-01 1.53372064e-01 1.26553953e-01
-2.61777282e-01 -1.49944842e+00 -7.50196129e-02 -9.03754681e-02
-8.54851365e-01 -1.15725899e+00 -3.50568146e-01 -7.92416155e-01
6.92500412e-01 -9.60673392e-02 9.63380575e-01 -2.47396782e-01
-2.79469222e-01 5.10059714e-01 -5.17365098e-01 -2.95189530e-01
-2.41724253e-01 1.54347226e-01 2.75742680e-01 1.49882630e-01
5.06093621e-01 -8.87157559e-01 -3.52684230e-01 1.23173833e-01
-8.43853831e-01 1.24382779e-01 7.89211333e-01 1.00427437e+00
6.93992019e-01 3.81290734e-01 5.55925012e-01 -1.09392953e+00
8.38335395e-01 -5.29827476e-01 -5.17856777e-01 7.63059497e-01
-1.18046820e+00 4.40069139e-01 8.07481587e-01 -3.52828711e-01
-1.10458374e+00 -3.52457970e-01 3.93582761e-01 -4.06058848e-01
7.88064227e-02 8.40079844e-01 -1.82016984e-01 -1.80016994e-01
6.23823106e-01 4.33715552e-01 -3.46005410e-01 -4.75432843e-01
7.76077271e-01 2.96499968e-01 2.74158865e-01 -6.90809429e-01
8.84261072e-01 1.31434679e-01 3.70400578e-01 -8.57147872e-01
-7.34026015e-01 -3.41507792e-01 -7.35964775e-01 8.83253291e-02
8.24659288e-01 -8.01809788e-01 -8.78610134e-01 1.33490175e-01
-1.30605125e+00 4.00886625e-01 -5.30394688e-02 7.27501333e-01
-6.80435449e-02 5.57090700e-01 -4.78386223e-01 -8.30682695e-01
-5.72500527e-01 -7.90013671e-01 6.40169144e-01 1.42878667e-01
-3.28333001e-03 -1.31789052e+00 -6.47900701e-02 6.87694073e-01
3.55856955e-01 1.53366610e-01 1.50098002e+00 -8.32763672e-01
-5.93444228e-01 -1.23626500e-01 -3.79557520e-01 3.84484440e-01
-8.72741491e-02 -2.95464784e-01 -7.29833663e-01 -2.41507608e-02
-3.49556893e-01 -2.28681281e-01 7.75199652e-01 -1.83680113e-02
9.10461426e-01 -4.55607653e-01 -4.22930121e-01 4.98172760e-01
1.44883561e+00 2.66655963e-02 4.59052473e-01 1.88654795e-01
1.09661901e+00 7.12120116e-01 4.56926703e-01 2.11858466e-01
7.96417117e-01 6.55365765e-01 1.39741272e-01 3.08914185e-01
-4.96783778e-02 -4.29699957e-01 1.56915247e-01 1.41450858e+00
-3.77852350e-01 -6.31394759e-02 -9.88940239e-01 5.88335216e-01
-1.98514247e+00 -8.82519901e-01 -3.34528655e-01 2.12310338e+00
9.55096900e-01 -2.51517296e-02 -3.15081269e-01 2.41234139e-01
6.52351558e-01 -3.23112011e-02 -3.18378299e-01 -2.46351197e-01
-4.13389921e-01 2.40707070e-01 3.98411214e-01 5.24581075e-01
-9.80969489e-01 1.13312161e+00 4.39142704e+00 9.75515842e-01
-5.35591960e-01 1.21331669e-01 3.20977241e-01 3.85982156e-01
-3.86014462e-01 1.25734851e-01 -9.75025475e-01 1.84930354e-01
7.96486199e-01 -2.47078925e-01 7.18801558e-01 8.26337755e-01
-3.55127871e-01 1.08589485e-01 -1.01590359e+00 1.12990034e+00
3.15614223e-01 -1.17343116e+00 3.37031662e-01 -7.31039494e-02
8.09940875e-01 -5.10472596e-01 -3.78800854e-02 6.48615241e-01
-8.65879282e-03 -9.44877803e-01 2.18536839e-01 7.84786642e-01
3.22845161e-01 -8.16582322e-01 8.56398702e-01 2.10765392e-01
-1.33274984e+00 8.00378397e-02 -6.25553310e-01 6.77114502e-02
-1.18040547e-01 8.33278120e-01 -1.00523496e+00 1.09716010e+00
5.04952967e-01 5.14766514e-01 -9.69582200e-01 5.18619955e-01
-4.02767032e-01 5.07213533e-01 -2.61652350e-01 -3.36740643e-01
-5.49071617e-02 -5.84604502e-01 5.18006325e-01 1.13823807e+00
1.90907106e-01 1.75000787e-01 -4.50245896e-03 1.03883576e+00
-9.47344825e-02 5.99589348e-01 -6.69341624e-01 -2.33329102e-01
6.19384587e-01 1.46289968e+00 -3.95899236e-01 -2.93864489e-01
-3.54215145e-01 8.09389353e-01 8.37684155e-01 5.06083548e-01
-8.46273482e-01 -4.32373136e-01 3.01854312e-01 -1.75869521e-02
2.43371293e-01 -2.80820489e-01 -3.21203679e-01 -1.16965663e+00
3.48454803e-01 -4.25932407e-01 4.50339854e-01 -5.99554241e-01
-1.38357973e+00 6.17569387e-01 8.79937857e-02 -9.47645664e-01
-3.90482657e-02 -7.68131495e-01 -8.93114209e-02 9.05089736e-01
-1.58565485e+00 -1.42246222e+00 -3.11615199e-01 8.31609249e-01
-1.35247976e-01 -4.49080080e-01 1.10787499e+00 5.72119474e-01
-5.61195076e-01 8.58043194e-01 2.14855239e-01 1.47144735e-01
5.67557335e-01 -1.35945010e+00 -7.09792078e-02 5.43837667e-01
4.97609824e-01 7.59899139e-01 4.47907627e-01 -6.99452817e-01
-1.51999724e+00 -1.20143092e+00 1.20745337e+00 -3.79632920e-01
9.16094422e-01 -4.04820740e-01 -1.01840401e+00 6.34654403e-01
-9.27278306e-03 1.05162784e-01 7.58306623e-01 5.02149522e-01
-5.23460269e-01 -2.79928625e-01 -7.96961427e-01 6.96959615e-01
1.11386061e+00 -7.52047539e-01 -6.69607580e-01 5.38350523e-01
7.55578876e-01 -1.83087632e-01 -1.27516317e+00 4.49466497e-01
3.85082006e-01 -8.16004753e-01 1.19069386e+00 -8.14228714e-01
2.86430657e-01 -3.10819119e-01 -1.10161938e-01 -1.37693739e+00
-4.38096702e-01 -1.97045222e-01 -5.44874668e-01 1.52410948e+00
4.57445234e-01 -8.62058401e-01 6.96217418e-01 6.76276386e-01
1.83320552e-01 -1.01443315e+00 -8.26035619e-01 -9.32374477e-01
-1.22822687e-01 -3.10190201e-01 3.45858097e-01 1.43705416e+00
1.98968738e-01 9.31965828e-01 -4.46517020e-01 3.50173891e-01
7.46976912e-01 1.55673295e-01 6.76580906e-01 -1.41255379e+00
-2.49529094e-01 -2.02625796e-01 -9.22663093e-01 -4.74326670e-01
4.18149501e-01 -1.39080358e+00 -8.33549380e-01 -1.56390977e+00
1.42990276e-01 -4.82784688e-01 -4.35716271e-01 4.53600556e-01
-4.87663984e-01 -2.64277786e-01 1.27349198e-01 6.82177991e-02
-5.16052186e-01 9.26230550e-01 1.01974154e+00 -2.42021427e-01
-1.45273775e-01 -3.18237513e-01 -4.75135714e-01 6.51563823e-01
7.41391063e-01 -3.99521053e-01 -6.94349706e-01 -1.61507279e-01
5.95922828e-01 -1.06947117e-01 4.68902081e-01 -9.00617778e-01
5.99957466e-01 2.93914557e-01 3.38268578e-01 -5.10027587e-01
5.00868678e-01 -1.07360363e+00 3.27599496e-01 1.76512048e-01
-2.20764130e-01 6.23853281e-02 1.40191272e-01 9.27295864e-01
-3.91285151e-01 -2.56717373e-02 3.26887190e-01 1.96691751e-01
-7.48658121e-01 2.82919973e-01 4.16513175e-01 2.09372491e-01
9.77548540e-01 -1.55015318e-02 -2.52407134e-01 -8.90784636e-02
-7.93147624e-01 9.08731297e-02 -1.99776649e-01 5.55967569e-01
5.50171435e-01 -1.67724538e+00 -7.70507991e-01 -1.01620041e-01
2.48636872e-01 -1.02013946e-01 4.80602682e-01 1.05628788e+00
-2.34082252e-01 7.31610179e-01 5.57326563e-02 -2.11170912e-01
-1.12942898e+00 8.07254195e-01 -7.80924112e-02 -8.72897148e-01
-3.53215307e-01 7.34942973e-01 1.01186775e-01 -7.56117105e-01
1.85050011e-01 -3.32327962e-01 -4.59652215e-01 1.65956631e-01
2.36644492e-01 7.53271461e-01 1.40909627e-01 -6.08249962e-01
-4.46271837e-01 6.03691638e-01 -2.12670952e-01 3.15144897e-01
1.22648430e+00 1.07323751e-01 -4.52944517e-01 3.46220762e-01
1.09421420e+00 -3.01876459e-02 -3.73042464e-01 -6.42056108e-01
2.15745673e-01 -2.00907111e-01 5.48532195e-02 -9.15465474e-01
-7.57049680e-01 8.57121706e-01 5.65376401e-01 1.49982244e-01
9.55353379e-01 -1.21393502e-01 5.05189478e-01 7.34722733e-01
5.14490664e-01 -1.06567538e+00 -2.31771469e-01 5.75731099e-01
1.05828869e+00 -1.08908904e+00 2.31769979e-01 -7.53749669e-01
-6.32304907e-01 1.04431581e+00 5.29096365e-01 1.42583415e-01
8.63018334e-01 -2.15723634e-01 -5.57565331e-01 -2.33773410e-01
-4.71381783e-01 -3.02642174e-02 7.41747499e-01 6.27803206e-01
4.64711428e-01 3.73891741e-01 -6.71677828e-01 1.02881598e+00
-4.20690238e-01 -2.39094406e-01 2.16605933e-03 3.61876667e-01
7.35123083e-02 -1.11071920e+00 -2.63222963e-01 4.57807124e-01
-1.45416215e-01 -5.91481566e-01 -4.24134076e-01 9.58632290e-01
1.85179889e-01 8.16084385e-01 -5.18936813e-01 -5.26253521e-01
5.08410811e-01 3.68400574e-01 4.24214900e-01 -2.78841257e-01
-3.86606812e-01 -5.93101323e-01 2.15309471e-01 -3.18288893e-01
-4.10155267e-01 -3.19876075e-01 -1.19390333e+00 -2.46142954e-01
-6.03606820e-01 3.33789259e-01 6.07491970e-01 8.19305062e-01
6.60170913e-01 5.93337417e-01 2.22289219e-01 -3.40821952e-01
-5.58848441e-01 -1.05963838e+00 -8.15654933e-01 6.01686299e-01
-5.38279474e-01 -8.94490540e-01 -1.81669056e-01 3.22650522e-02] | [8.718070030212402, 7.854051113128662] |
1c9def60-da36-4de6-9206-90386dc2f1ef | seo-safety-aware-energy-optimization | 2302.12493 | null | https://arxiv.org/abs/2302.12493v1 | https://arxiv.org/pdf/2302.12493v1.pdf | SEO: Safety-Aware Energy Optimization Framework for Multi-Sensor Neural Controllers at the Edge | Runtime energy management has become quintessential for multi-sensor autonomous systems at the edge for achieving high performance given the platform constraints. Typical for such systems, however, is to have their controllers designed with formal guarantees on safety that precede in priority such optimizations, which in turn limits their application in real settings. In this paper, we propose a novel energy optimization framework that is aware of the autonomous system's safety state, and leverages it to regulate the application of energy optimization methods so that the system's formal safety properties are preserved. In particular, through the formal characterization of a system's safety state as a dynamic processing deadline, the computing workloads of the underlying models can be adapted accordingly. For our experiments, we model two popular runtime energy optimization methods, offloading and gating, and simulate an autonomous driving system (ADS) use-case in the CARLA simulation environment with performance characterizations obtained from the standard Nvidia Drive PX2 ADS platform. Our results demonstrate that through a formal awareness of the perceived risks in the test case scenario, energy efficiency gains are still achieved (reaching 89.9%) while maintaining the desired safety properties. | ['Mohammad Abdullah Al Faruque', 'Yasser Shoukry', 'James Ferlez', 'Mohanad Odema'] | 2023-02-24 | null | null | null | null | ['energy-management'] | ['time-series'] | [ 7.00061694e-02 4.33571011e-01 -4.72077906e-01 -2.31951416e-01
-2.97192454e-01 -4.92084742e-01 4.19158787e-01 1.80593103e-01
-3.51358622e-01 4.53731596e-01 -1.65735006e-01 -7.01651275e-01
3.05161793e-02 -7.11486399e-01 -6.16541564e-01 -5.39480507e-01
-1.80778816e-01 2.83434242e-02 4.25466597e-01 -2.19054922e-01
-7.28051066e-02 4.94974852e-01 -1.98363698e+00 -4.01092410e-01
5.53445280e-01 1.16044497e+00 -3.16147655e-01 8.38623345e-01
5.05189538e-01 5.61327398e-01 -5.28092802e-01 2.88997501e-01
3.34165365e-01 -1.69090112e-03 -5.51314831e-01 -2.26301238e-01
-2.92693041e-02 -5.73263228e-01 -2.54888862e-01 9.08222258e-01
1.20858036e-01 -2.43565947e-01 1.01742618e-01 -2.02898431e+00
3.71793240e-01 5.61046779e-01 -4.33864027e-01 -2.79528759e-02
-2.83037603e-01 4.22028154e-01 8.01270545e-01 1.78620070e-01
2.78646827e-01 7.61174679e-01 1.90096498e-01 6.61988020e-01
-1.30344510e+00 -5.35479724e-01 2.42767990e-01 -8.45431685e-02
-1.47561729e+00 -8.03962231e-01 4.63731229e-01 -1.53672010e-01
1.44732821e+00 7.40912616e-01 5.44333875e-01 7.62992799e-01
9.92415428e-01 1.67569190e-01 1.01136935e+00 -3.50398690e-01
8.85631561e-01 1.86217323e-01 5.26287675e-01 2.39463314e-01
6.00260139e-01 5.42898595e-01 -4.22150284e-01 -4.61762697e-01
-1.37793720e-01 -4.79444593e-01 -3.74620222e-02 -6.16944790e-01
-8.45850408e-01 3.35722715e-01 -1.39724791e-01 -1.44684300e-01
-4.59437929e-02 8.01719308e-01 8.59358788e-01 -1.29729971e-01
-1.11023553e-01 3.79213005e-01 -4.91925776e-01 -4.65201229e-01
-7.06923008e-01 1.59439668e-01 9.95651901e-01 9.63087261e-01
5.30794621e-01 1.12897836e-01 -1.21201072e-02 -4.00985509e-01
4.20865089e-01 6.44771695e-01 8.21848437e-02 -1.31570387e+00
-1.24038681e-01 5.00169575e-01 3.67931426e-01 -4.41506624e-01
-6.90541029e-01 -2.10512266e-01 -3.89236152e-01 7.60954976e-01
-1.70160756e-01 -1.50906667e-02 -6.14018559e-01 2.04344368e+00
3.91123950e-01 -9.19729248e-02 3.72637719e-01 7.87653208e-01
-3.47419739e-01 7.00881243e-01 2.13192478e-01 -5.01596510e-01
1.40683866e+00 -4.54348296e-01 -6.04747117e-01 -3.01630348e-01
6.34122908e-01 -2.04266056e-01 9.60360348e-01 4.27784920e-01
-9.63493705e-01 -2.69198984e-01 -1.87939548e+00 2.63578117e-01
1.43982451e-02 -2.07146376e-01 4.44788516e-01 9.91634607e-01
-1.15756404e+00 3.87063652e-01 -1.64910316e+00 -4.47637349e-01
-7.32753053e-02 7.03091383e-01 4.19718921e-01 6.11135840e-01
-8.84272575e-01 1.07715034e+00 3.08717877e-01 -2.63196677e-01
-1.23281717e+00 -8.65693569e-01 -5.25850654e-01 2.66694665e-01
5.47994792e-01 -6.25125229e-01 1.46572781e+00 -5.86192310e-01
-1.65744758e+00 4.75088686e-01 3.19839716e-01 -6.40248120e-01
2.83105910e-01 -9.30091143e-02 -7.13067651e-01 -5.87290078e-02
-2.42161542e-01 3.08617413e-01 6.84182346e-01 -1.11511385e+00
-6.83438718e-01 -3.66158932e-01 3.66488069e-01 -1.77386180e-01
-6.70597851e-01 -1.56858966e-01 1.13089584e-01 3.42727453e-01
-6.60047174e-01 -1.26557970e+00 -3.82328719e-01 -2.22256765e-01
-4.55162615e-01 1.89461082e-01 9.55462039e-01 5.97492307e-02
1.35518849e+00 -2.35030794e+00 -8.70616511e-02 4.17578638e-01
2.19062746e-01 -5.98450601e-02 2.43082181e-01 2.41496280e-01
3.25108171e-01 1.25757426e-01 -5.23079075e-02 -1.33181825e-01
4.30272818e-01 3.55115920e-01 -4.67853457e-01 6.45790517e-01
-6.26759678e-02 3.14283818e-01 -5.37508905e-01 -2.62584955e-01
4.16043013e-01 2.05284774e-01 -5.91824472e-01 3.17569941e-01
-3.89706731e-01 -3.91415134e-02 -6.55719101e-01 2.43886828e-01
3.98082852e-01 -1.68611854e-03 6.04718685e-01 -4.43603307e-01
-5.80334127e-01 8.79032612e-02 -1.08407032e+00 1.18734443e+00
-7.73311377e-01 3.91543120e-01 6.33898258e-01 -3.20833415e-01
4.36848730e-01 -2.14110926e-01 5.53391039e-01 -8.83650362e-01
4.12915647e-01 1.08861431e-01 1.06597476e-01 -7.20391423e-02
8.65182817e-01 1.50410429e-01 -5.12170911e-01 4.65706229e-01
-7.92448401e-01 -1.50727168e-01 -2.30915010e-01 1.21403314e-01
1.62744308e+00 1.23667113e-01 2.27177620e-01 -1.07407057e+00
3.04637730e-01 5.03928900e-01 6.88143730e-01 3.95950437e-01
-4.34887439e-01 -2.91739374e-01 6.42821074e-01 -7.44630471e-02
-1.21125841e+00 -9.45888817e-01 -2.00001791e-01 7.48504877e-01
4.13985044e-01 -7.43288219e-01 -1.02869868e+00 -2.68738091e-01
-1.57748923e-01 1.31837308e+00 -3.96517068e-01 -7.87962198e-01
-5.24458468e-01 -5.12613535e-01 6.61214113e-01 4.29082036e-01
2.85901666e-01 -2.30912745e-01 -2.16660571e+00 1.30960315e-01
5.39976358e-01 -1.24493718e+00 -3.82411480e-01 5.78612626e-01
-5.33450127e-01 -9.63808119e-01 5.74906111e-01 1.17261536e-01
3.89081389e-01 1.68239668e-01 1.08338690e+00 9.20042545e-02
-4.42331642e-01 5.22773266e-01 1.13873906e-01 -2.32467309e-01
-8.66947532e-01 1.38063272e-02 3.73545527e-01 -3.13535094e-01
5.33037148e-02 -5.05577564e-01 -7.27318048e-01 3.01728368e-01
-1.00788116e+00 2.05682054e-01 2.14948997e-01 3.24917287e-01
7.39449441e-01 2.99124330e-01 3.30361545e-01 -4.32669789e-01
2.21784204e-01 -3.34689051e-01 -1.33496928e+00 2.41957933e-01
-1.37113762e+00 3.94765377e-01 9.81159329e-01 -1.53561890e-01
-7.31523454e-01 2.57266492e-01 2.44787753e-01 -4.02297854e-01
8.66782665e-02 2.58091670e-02 -6.95176542e-01 -3.99904884e-02
5.41520476e-01 -7.90473297e-02 4.26843353e-02 3.57605368e-01
2.04518110e-01 4.96181011e-01 7.07743943e-01 -9.36855793e-01
6.14283562e-01 5.76608121e-01 5.10273397e-01 -5.40032327e-01
-1.51261792e-01 2.69561917e-01 8.46421123e-02 -2.86583245e-01
7.68960297e-01 -8.55845213e-01 -1.43418658e+00 2.06279084e-01
-6.25312567e-01 -5.64097703e-01 -3.30090255e-01 1.16728343e-01
-7.31313825e-01 3.88827808e-02 -1.71203643e-01 -1.13088560e+00
-4.48573679e-01 -1.57838535e+00 1.13218880e+00 4.59611595e-01
-4.24603343e-01 -4.39531952e-01 5.20617142e-02 -2.34067902e-01
8.00825417e-01 5.17441571e-01 8.92555654e-01 -3.82693291e-01
-8.22333813e-01 1.95802018e-01 1.84211791e-01 1.41233191e-01
-1.43314704e-01 3.39540958e-01 -1.12652385e+00 -6.90847218e-01
2.07320496e-01 5.05689755e-02 1.43214330e-01 1.16005152e-01
1.21630728e+00 -2.88782865e-01 -5.25285780e-01 4.95354205e-01
1.72450936e+00 2.26223558e-01 3.65333706e-01 3.18949729e-01
2.66363651e-01 2.82552242e-01 7.65036702e-01 8.49549353e-01
4.66625065e-01 9.99652684e-01 1.00661707e+00 4.36096579e-01
3.95128816e-01 9.49975103e-02 8.33918989e-01 3.90246123e-01
5.65526843e-01 -1.20479397e-01 -8.58009934e-01 3.38223010e-01
-1.77386558e+00 -2.89182454e-01 3.81997973e-02 2.65909195e+00
6.01750553e-01 5.89008987e-01 1.15516856e-01 -2.88617350e-02
3.05263698e-01 -4.91567142e-03 -1.17690873e+00 -1.00466681e+00
3.89652193e-01 1.99887842e-01 1.28157985e+00 4.09304082e-01
-4.89856958e-01 4.95674998e-01 5.77795553e+00 6.11880422e-01
-1.24251390e+00 1.15782000e-01 4.14267659e-01 -5.35793006e-01
-3.75320524e-01 4.03518230e-01 -9.29000974e-01 4.99523669e-01
2.18306899e+00 -9.43988800e-01 5.47704935e-01 1.05447662e+00
5.90583563e-01 -5.85796654e-01 -1.61956263e+00 5.75881183e-01
-3.84199977e-01 -1.10592198e+00 -6.91862047e-01 2.93977141e-01
1.61575094e-01 -3.56058665e-02 -8.71310309e-02 3.70340616e-01
3.50235492e-01 -8.56646895e-01 1.11205971e+00 1.61368445e-01
8.51543665e-01 -1.28368700e+00 3.74723822e-01 3.34670812e-01
-9.85491812e-01 -1.96235441e-02 9.63971112e-03 -1.13493517e-01
3.56970698e-01 5.64707279e-01 -5.62201083e-01 2.87062943e-01
7.98633993e-01 -1.68989435e-01 -4.59142923e-01 2.07870901e-01
2.08740383e-01 5.69912612e-01 -6.57826543e-01 -3.92002240e-02
-1.30650371e-01 2.89937165e-02 4.49597478e-01 7.19168782e-01
1.72319368e-01 1.39322668e-01 2.17536598e-01 8.87809873e-01
2.50213027e-01 -5.03162801e-01 -4.72848952e-01 -1.27725586e-01
6.44333601e-01 1.51622510e+00 -7.04513252e-01 -5.49289696e-02
-2.44734064e-01 4.59909618e-01 -2.21084714e-01 -1.87139586e-02
-1.46588540e+00 -1.35208622e-01 1.52093673e+00 1.69719934e-01
-2.38773525e-01 -4.93408144e-01 -7.22407699e-01 -6.20061934e-01
7.68049061e-02 -6.47437811e-01 3.15547064e-02 -5.38872182e-01
-4.39768881e-01 4.54012603e-01 3.08820069e-01 -8.25386107e-01
-2.38213256e-01 -4.36278850e-01 -3.84030730e-01 4.11272347e-01
-1.37857008e+00 -7.81931400e-01 -3.26565802e-01 2.07468092e-01
3.65735143e-02 3.51246357e-01 8.11554372e-01 2.06806064e-01
-8.57777178e-01 6.78648472e-01 -6.78559393e-02 -9.12709713e-01
3.92141730e-01 -9.18946207e-01 3.98889810e-01 1.08844149e+00
-8.39923501e-01 4.13613111e-01 1.37636793e+00 -5.58574975e-01
-2.41152096e+00 -1.17668188e+00 3.46880034e-02 -2.92208314e-01
9.55426097e-01 -4.63767201e-01 -6.58892512e-01 5.57929516e-01
4.90406692e-01 -2.12128125e-02 3.27168316e-01 -3.09333652e-01
-7.06312526e-03 -4.15723413e-01 -1.12097609e+00 7.28382051e-01
8.99026513e-01 -2.97869772e-01 1.81998536e-01 2.31483281e-01
9.29279625e-01 -5.82399309e-01 -1.08454967e+00 5.02293766e-01
4.89893675e-01 -7.85128951e-01 3.00376803e-01 -3.61940652e-01
-3.16828303e-02 -6.97245419e-01 -5.89956462e-01 -9.61072028e-01
1.87251642e-01 -9.06338930e-01 -5.25728226e-01 1.17150617e+00
1.05175942e-01 -5.56318820e-01 6.31139994e-01 1.25301385e+00
-5.75838327e-01 -5.70852876e-01 -1.26103020e+00 -1.03703833e+00
-1.34489626e-01 -6.28662884e-01 7.27759302e-01 3.54881048e-01
2.90980905e-01 1.64068460e-01 -1.03673302e-02 7.54550576e-01
7.32019186e-01 5.42915426e-02 6.23160660e-01 -5.19788325e-01
-5.51565230e-01 -3.24009538e-01 -3.22664171e-01 -3.41065228e-01
5.56017697e-01 -4.48088169e-01 3.28014553e-01 -7.31389225e-01
1.93819374e-01 -4.81740624e-01 -2.50963092e-01 4.21730071e-01
3.62345278e-01 -5.32993078e-01 1.51240915e-01 -7.18846321e-02
-7.05200315e-01 6.15329862e-01 2.34244466e-01 -8.60006884e-02
-2.64459759e-01 -1.99514166e-01 -6.03329122e-01 3.60552967e-01
7.64920473e-01 -2.67811090e-01 -7.69765913e-01 -1.25002205e-01
3.77818167e-01 1.97119787e-01 4.18423355e-01 -1.28756988e+00
2.97350764e-01 -6.13180339e-01 -5.04419923e-01 -1.69721872e-01
2.14821279e-01 -1.22735023e+00 8.59571874e-01 8.32216322e-01
-1.55923441e-01 3.46136063e-01 7.91047335e-01 5.19540966e-01
3.83896232e-01 3.90141398e-01 8.98316741e-01 6.78025126e-01
-9.38915193e-01 7.42063969e-02 -6.02663636e-01 -9.18332860e-02
1.76143968e+00 3.53403576e-02 -7.71836460e-01 1.56386524e-01
-3.72372061e-01 7.17801213e-01 1.30232787e+00 3.87208939e-01
2.52206653e-01 -1.04595935e+00 -2.23668620e-01 1.18077412e-01
4.47633415e-01 -2.55132496e-01 2.35992581e-01 8.51134062e-01
-3.56640697e-01 2.83511668e-01 -4.10663515e-01 -6.48767233e-01
-1.17657936e+00 8.70089114e-01 6.41002297e-01 -1.10286407e-01
-4.95208979e-01 -3.52522135e-02 2.51489550e-01 2.43586272e-01
-1.90091819e-01 -6.07143462e-01 5.18581629e-01 -5.18640220e-01
2.68345028e-01 4.05625761e-01 4.36436117e-01 -1.70655593e-01
-9.54971313e-01 6.55409023e-02 2.28506401e-01 -2.99429387e-01
9.57099974e-01 -2.68351525e-01 -6.82848841e-02 2.87591070e-01
8.26197743e-01 6.88704923e-02 -1.28750336e+00 5.34668624e-01
5.78124784e-02 -6.42781854e-02 4.56499606e-01 -7.30776250e-01
-1.02414238e+00 2.47799486e-01 8.82275283e-01 2.91151196e-01
1.55714035e+00 -2.22094551e-01 7.74469733e-01 1.19687364e-01
9.08803582e-01 -1.46698177e+00 -5.07019281e-01 1.43950611e-01
1.97802618e-01 -5.42728484e-01 2.17504576e-02 -3.50348681e-01
-1.39723137e-01 8.93306196e-01 9.48637486e-01 2.58320626e-02
6.09567344e-01 1.22218239e+00 -3.56597692e-01 1.10406308e-02
-1.18033969e+00 3.40771586e-01 -6.30625129e-01 2.46387437e-01
-1.66815147e-01 5.46198964e-01 -2.68255621e-01 6.45106256e-01
9.47477471e-04 -7.52367526e-02 1.10990167e+00 1.19857299e+00
-5.56720853e-01 -9.97751057e-01 -3.45670998e-01 1.88827634e-01
-2.43680939e-01 4.22972798e-01 5.38011640e-02 1.08068311e+00
-1.88008606e-01 1.06926215e+00 1.02575921e-01 -6.61806941e-01
4.47781682e-01 -7.64483139e-02 2.39617303e-01 -4.02452767e-01
-5.26461065e-01 -2.93875217e-01 5.91764033e-01 -1.31268954e+00
-5.65915592e-02 -5.44652283e-01 -1.78317201e+00 -4.65253323e-01
-3.62438671e-02 1.74314052e-01 1.05722678e+00 8.55121613e-01
9.29497361e-01 1.08560479e+00 7.85630107e-01 -4.19713497e-01
-9.62337136e-01 -1.42339602e-01 -5.00219345e-01 -2.44597867e-01
3.17945778e-01 -6.98278725e-01 -3.68286937e-01 -2.35928729e-01] | [5.645238876342773, 2.916300058364868] |
ca735186-a4e7-4157-907d-2de731103338 | square-a-large-scale-dataset-of-sensitive | 2305.17696 | null | https://arxiv.org/abs/2305.17696v1 | https://arxiv.org/pdf/2305.17696v1.pdf | SQuARe: A Large-Scale Dataset of Sensitive Questions and Acceptable Responses Created Through Human-Machine Collaboration | The potential social harms that large language models pose, such as generating offensive content and reinforcing biases, are steeply rising. Existing works focus on coping with this concern while interacting with ill-intentioned users, such as those who explicitly make hate speech or elicit harmful responses. However, discussions on sensitive issues can become toxic even if the users are well-intentioned. For safer models in such scenarios, we present the Sensitive Questions and Acceptable Response (SQuARe) dataset, a large-scale Korean dataset of 49k sensitive questions with 42k acceptable and 46k non-acceptable responses. The dataset was constructed leveraging HyperCLOVA in a human-in-the-loop manner based on real news headlines. Experiments show that acceptable response generation significantly improves for HyperCLOVA and GPT-3, demonstrating the efficacy of this dataset. | ['Jung-Woo Ha', 'Sangchul Park', 'Alice Oh', 'Yong Lim', 'Eun-Ju Lee', 'Gunhee Kim', 'Byoung Pil Kim', 'Yejin Choi', 'Meeyoung Cha', 'Takyoung Kim', 'Joonsuk Park', 'Seokhee Hong', 'Hwaran Lee'] | 2023-05-28 | null | null | null | null | ['response-generation'] | ['natural-language-processing'] | [-7.22412392e-02 3.68858218e-01 -2.27740929e-01 -2.96888232e-01
-9.22851503e-01 -7.84296453e-01 4.67165381e-01 9.98546258e-02
-3.40078682e-01 6.76125109e-01 9.26529109e-01 -2.63658047e-01
1.42792061e-01 -5.12000799e-01 -1.23189270e-01 -2.19946966e-01
3.09684336e-01 -6.31809756e-02 -5.59778325e-02 -6.60691261e-01
4.91886497e-01 -1.33150458e-01 -8.15355361e-01 6.23679519e-01
1.04755700e+00 4.13295209e-01 -5.75114131e-01 7.24458396e-01
1.89149812e-01 1.38834465e+00 -6.95492923e-01 -9.72186446e-01
-6.54827803e-02 -3.55748117e-01 -9.57007408e-01 -4.38115776e-01
6.44100606e-01 -6.29502356e-01 -2.57367700e-01 1.07532895e+00
8.08849752e-01 7.85650834e-02 4.56356943e-01 -1.22888482e+00
-1.40635598e+00 8.58820379e-01 -4.20448214e-01 1.92561880e-01
7.92836666e-01 5.03934205e-01 9.11878109e-01 -7.38668263e-01
3.85750175e-01 1.75975978e+00 7.72753537e-01 1.06887460e+00
-1.19621325e+00 -9.92679417e-01 1.54865012e-01 -2.49536142e-01
-7.86219954e-01 -3.75858098e-01 8.11789751e-01 -6.08598590e-01
6.77431941e-01 9.29900408e-01 4.57153767e-01 2.04297543e+00
2.54150659e-01 5.74910343e-01 1.29999065e+00 2.21745953e-01
1.73378542e-01 6.37058198e-01 7.39762127e-01 1.79296598e-01
1.54563919e-01 1.16018057e-02 -6.04533851e-01 -8.77242327e-01
-2.18236119e-01 -1.52127683e-01 -2.74724215e-01 6.46676719e-01
-8.11174989e-01 1.53723836e+00 5.36998212e-01 3.44452858e-02
-4.02493864e-01 -3.64177912e-01 4.77841765e-01 2.86594212e-01
9.81375754e-01 8.47449481e-01 -3.96907687e-01 -2.87369877e-01
-4.26187724e-01 5.63866735e-01 1.19968736e+00 7.88655818e-01
2.58478552e-01 -1.40981913e-01 -7.09364772e-01 9.98557687e-01
2.50563741e-01 8.76183450e-01 3.70775431e-01 -9.31125402e-01
4.32710618e-01 5.56002438e-01 2.88955420e-01 -1.64899254e+00
-4.20536578e-01 -2.56969362e-01 -7.28589416e-01 -4.18116421e-01
3.22477639e-01 -7.87204385e-01 -2.00486034e-01 1.94839191e+00
3.39092463e-01 -5.56913435e-01 -1.03102617e-01 7.12530315e-01
8.12244475e-01 8.25411439e-01 6.41113937e-01 -4.78155255e-01
1.39311695e+00 -6.90006435e-01 -1.16108096e+00 -5.33998549e-01
1.12825930e+00 -9.26433623e-01 1.78803396e+00 2.07846418e-01
-9.24997747e-01 1.01176597e-01 -3.97627503e-01 -2.96342283e-01
-3.66065770e-01 -5.60458720e-01 1.92608207e-01 1.02012599e+00
-8.90690804e-01 -1.82045288e-02 4.83868569e-01 -5.08277893e-01
2.82982796e-01 -9.94344875e-02 -1.25718698e-01 2.31911391e-01
-1.90141332e+00 1.12075734e+00 -2.07003921e-01 5.15127294e-02
-5.58611512e-01 -1.01833320e+00 -6.21171296e-01 -1.65197790e-01
5.67321181e-01 -1.95960850e-01 1.46584475e+00 -7.92443037e-01
-1.30519509e+00 7.70008683e-01 2.91117847e-01 -4.40589450e-02
6.11623526e-01 -7.77915061e-01 -5.21719038e-01 -2.39569247e-01
1.15120269e-01 4.03980970e-01 8.32179606e-01 -1.00254154e+00
1.05871245e-01 -4.22285825e-01 3.84408295e-01 5.95946535e-02
-9.14071977e-01 8.28613281e-01 7.05440938e-01 -6.87353611e-01
-6.01735830e-01 -1.13308609e+00 -2.48294055e-01 -3.66357148e-01
-8.76275182e-01 -4.22414511e-01 1.05115914e+00 -1.02612865e+00
1.84070623e+00 -2.03868771e+00 -3.47697467e-01 -3.64031494e-02
5.12169540e-01 5.55817544e-01 -9.49200168e-02 7.37159193e-01
1.47861931e-02 9.08268213e-01 1.60732135e-01 1.44719481e-01
1.66143283e-01 -3.85590762e-01 -4.83038872e-01 2.93922544e-01
7.80511424e-02 9.51378942e-01 -9.07469213e-01 -4.30363059e-01
-3.53681415e-01 3.13857764e-01 -1.03807259e+00 4.92989033e-01
-7.76368156e-02 1.23811044e-01 -6.47385180e-01 2.59080529e-01
7.21744120e-01 -6.73792660e-02 -1.44341484e-01 1.87433824e-01
-2.15587318e-02 4.87211972e-01 -3.15524697e-01 6.09249771e-01
-4.58331674e-01 5.24130881e-01 2.52422065e-01 1.61317736e-01
8.70693982e-01 9.71295238e-02 -1.71890259e-01 -7.81433821e-01
1.49386421e-01 -9.60366875e-02 1.72534529e-02 -8.69933665e-01
6.43320680e-01 -3.73986095e-01 -6.60756707e-01 7.50180244e-01
-6.56110823e-01 -1.84010103e-01 -1.28850624e-01 8.70947897e-01
1.03783500e+00 -5.98506689e-01 2.41583318e-01 -5.31732619e-01
5.70475996e-01 -2.82081783e-01 2.41620168e-01 8.18572462e-01
-7.30202496e-01 1.52856588e-01 1.02558815e+00 -2.57953852e-01
-6.40922308e-01 -4.30999845e-01 2.16456920e-01 1.65460277e+00
-2.16295958e-01 -6.40191793e-01 -1.30926943e+00 -9.88091946e-01
-1.45043612e-01 1.39540493e+00 -6.83292687e-01 -5.20204902e-01
-3.04532975e-01 -6.47608042e-01 6.71239495e-01 2.40754392e-02
4.56686676e-01 -1.20120430e+00 -2.83500254e-01 1.16651945e-01
-7.10903883e-01 -9.61699009e-01 -8.60544682e-01 -3.80267024e-01
-1.49404809e-01 -9.23731446e-01 -3.76326889e-01 -2.83902198e-01
5.14754117e-01 3.24231178e-01 9.25089180e-01 3.10523599e-01
2.83741057e-01 2.46291518e-01 -3.44620198e-01 -5.28286040e-01
-8.15307975e-01 1.01003192e-01 1.56791329e-01 4.06865291e-02
5.40910125e-01 -5.92923537e-02 -4.40982550e-01 4.51916903e-01
-9.55963910e-01 -8.94775465e-02 1.56323577e-03 4.41808999e-01
-6.98923945e-01 -5.75020909e-01 9.34431195e-01 -1.52681148e+00
1.45589137e+00 -9.25605595e-01 7.02902535e-03 1.13486864e-01
-3.15511584e-01 -4.20655817e-01 7.78155208e-01 -7.39527822e-01
-1.37914705e+00 -5.99164546e-01 -3.91920537e-01 5.29949307e-01
-1.55914441e-01 2.44854152e-01 3.49588394e-02 -1.00608967e-01
1.17965019e+00 -4.03518558e-01 -1.77529067e-01 -2.88262159e-01
3.59469533e-01 1.28614581e+00 -1.31253004e-01 -4.00121480e-01
1.13349342e+00 -7.50645846e-02 -7.49940038e-01 -1.05097556e+00
-1.40899754e+00 -2.49620989e-01 -8.41286033e-03 -6.43006742e-01
7.49918759e-01 -6.84048116e-01 -1.20941949e+00 7.29601383e-01
-1.30238962e+00 -3.49628299e-01 2.18907639e-01 6.74400628e-02
-1.31714240e-01 3.09489578e-01 -1.04519188e+00 -1.13368487e+00
-6.70835257e-01 -6.50435328e-01 5.06376326e-01 1.84939995e-01
-1.08790505e+00 -9.85894680e-01 1.68377608e-01 8.82541120e-01
8.91843021e-01 2.67723918e-01 1.18404090e+00 -1.28558278e+00
6.21816292e-02 -5.20135581e-01 -6.83411434e-02 3.03491652e-01
-1.11264594e-01 7.10281283e-02 -1.05325365e+00 -1.10420182e-01
5.29560864e-01 -1.11815917e+00 1.35301709e-01 -2.31474191e-01
8.76797497e-01 -1.44772673e+00 1.83887064e-01 -2.80725360e-01
8.31973135e-01 -3.07215005e-01 6.03100061e-01 -3.30552049e-02
6.01297617e-01 1.22992694e+00 4.88580137e-01 7.33855724e-01
5.13564646e-01 3.04752469e-01 3.13431084e-01 1.10605575e-01
3.56466025e-01 -7.53097951e-01 7.68391013e-01 9.63205755e-01
7.16690600e-01 -5.51072776e-01 -8.19797456e-01 2.36366957e-01
-1.61430204e+00 -1.11995125e+00 -7.74803400e-01 1.89790595e+00
1.23642182e+00 7.65930116e-02 1.27385691e-01 -1.14679284e-01
7.74555147e-01 6.06424272e-01 -2.59964794e-01 -1.05908513e+00
-1.08365990e-01 -5.86294234e-01 -2.82149483e-02 9.55553651e-01
-7.81698346e-01 9.36786890e-01 6.47166061e+00 6.80737555e-01
-8.97683859e-01 3.59777838e-01 1.11597121e+00 -1.75199449e-01
-7.85831034e-01 -1.99704319e-01 -7.23829687e-01 7.31388688e-01
1.35457969e+00 -3.77600104e-01 2.38835648e-01 8.87236655e-01
6.26761734e-01 -1.27371952e-01 -7.67023325e-01 5.47582746e-01
4.12611097e-01 -8.34980965e-01 -7.96890333e-02 -3.66531760e-02
5.98937511e-01 -4.73705381e-01 4.55805451e-01 7.90648639e-01
3.79395187e-01 -8.04346740e-01 6.52500749e-01 1.61753103e-01
3.05043370e-01 -4.89391685e-01 5.13072252e-01 9.85172451e-01
-7.22929835e-02 -2.54429787e-01 -2.24408656e-01 -5.17544925e-01
1.26039445e-01 4.06731546e-01 -6.36500180e-01 -4.39773470e-01
4.87799346e-01 4.57953334e-01 -8.56590271e-01 2.23119244e-01
-3.94727349e-01 1.01903069e+00 -1.15919158e-01 -7.30018795e-01
3.29855740e-01 9.35319290e-02 4.98298168e-01 1.42197275e+00
-1.54226288e-01 5.70262671e-01 -1.14861824e-01 8.02155674e-01
-2.13594079e-01 3.31476599e-01 -9.97753739e-01 -1.46179050e-01
4.31174040e-01 1.57723463e+00 -4.00735997e-02 -1.54867187e-01
-3.09194148e-01 5.70000648e-01 5.06251574e-01 4.34116274e-01
-1.09295511e+00 -1.26456648e-01 4.49151516e-01 3.95572305e-01
-7.29385257e-01 2.61632532e-01 -4.49453652e-01 -9.54956830e-01
-2.41256669e-01 -1.49943757e+00 3.85129839e-01 -8.62811089e-01
-1.66957200e+00 3.31059784e-01 -1.06748216e-01 -5.57424128e-01
-3.68436165e-02 -1.85971722e-01 -5.73027730e-01 6.97855115e-01
-1.11104393e+00 -1.02865207e+00 -2.04738483e-01 4.15970564e-01
5.62393188e-01 4.47785765e-01 6.29207850e-01 1.47985682e-01
-8.46301079e-01 7.30515420e-01 -4.89138842e-01 -9.82847214e-02
1.08507800e+00 -7.29573786e-01 2.09363133e-01 7.36109078e-01
-7.49617040e-01 9.28791106e-01 9.10366654e-01 -9.35366094e-01
-1.13410485e+00 -1.05718112e+00 1.55566990e+00 -1.05922890e+00
9.71498728e-01 -8.09876382e-01 -1.16681480e+00 6.39294922e-01
6.40218496e-01 -5.17592788e-01 1.21425295e+00 7.21460357e-02
-7.06252456e-01 4.08650696e-01 -1.35731006e+00 1.06415153e+00
9.34025049e-01 -7.28538275e-01 -5.32490730e-01 8.05580497e-01
1.01036072e+00 -8.85186493e-02 -5.85582972e-01 -3.80858220e-02
2.52693236e-01 -1.02484787e+00 5.89831412e-01 -1.12772250e+00
8.05005908e-01 4.30533975e-01 8.69036019e-02 -1.39294934e+00
-7.66931415e-01 -1.39407122e+00 3.15140896e-02 1.52967918e+00
5.82284510e-01 -5.37510753e-01 3.42402518e-01 1.60966563e+00
2.31426418e-01 -4.39339221e-01 -4.09258097e-01 -1.90114349e-01
3.48234147e-01 -2.23075867e-01 2.29375258e-01 1.24650311e+00
4.59224075e-01 7.85822570e-01 -1.02467453e+00 -4.62410375e-02
4.35181916e-01 -6.72370732e-01 7.84743369e-01 -1.00643241e+00
3.29449177e-01 -1.05905779e-01 5.29572546e-01 -6.32102430e-01
4.18947488e-01 -4.30877328e-01 -3.93179730e-02 -7.87570715e-01
6.74806595e-01 2.78579127e-02 2.40942180e-01 4.27277416e-01
-7.89795280e-01 1.91139221e-01 3.88053507e-01 -9.20510292e-02
-7.21949279e-01 4.91457641e-01 1.12434900e+00 7.12137148e-02
-2.51725558e-02 -7.60269389e-02 -1.37736952e+00 1.00212646e+00
8.82317841e-01 -7.37113893e-01 -3.04824084e-01 7.74303898e-02
8.48997056e-01 -2.28596497e-02 3.02772164e-01 -3.52055132e-01
-7.24346712e-02 -6.26586139e-01 -9.61456373e-02 -1.46623522e-01
6.92306086e-02 -6.42967403e-01 -3.07382107e-01 5.88084280e-01
-1.29249322e+00 1.73932061e-01 -5.16825356e-03 5.74015737e-01
2.14917451e-01 -1.28448442e-01 8.82725894e-01 -2.37887241e-02
-1.23634517e-01 1.81339651e-01 -8.26617777e-01 9.16276276e-01
1.01301885e+00 1.21045835e-01 -1.11743784e+00 -1.13983405e+00
-2.40125433e-01 2.45706245e-01 4.93381247e-02 8.62889111e-01
5.72751999e-01 -1.24315655e+00 -1.03264213e+00 2.44594198e-02
3.86471242e-01 -7.88005114e-01 5.94670296e-01 8.23605537e-01
-1.12555414e-01 3.29039961e-01 6.26252890e-02 1.78349331e-01
-1.26361835e+00 7.12374151e-01 1.22332796e-01 -7.58525208e-02
3.48260105e-02 8.68665040e-01 4.89200205e-01 -6.74252152e-01
6.22167327e-02 1.47519439e-01 -4.50851172e-01 4.28323925e-01
8.27078283e-01 6.96150064e-01 -3.32084328e-01 -8.09730530e-01
-3.65740240e-01 -1.18209407e-01 -3.93985391e-01 8.33743215e-02
7.97565222e-01 -2.55978793e-01 -3.31768781e-01 1.74309537e-01
1.29032350e+00 5.53022861e-01 -6.84056938e-01 -1.31446689e-01
1.24653369e-01 -8.17642748e-01 -2.10097775e-01 -1.10308957e+00
-4.74572301e-01 6.75584793e-01 -8.16256404e-02 8.13734710e-01
5.59455097e-01 -2.59090215e-01 1.07723320e+00 3.27151626e-01
-9.69579518e-02 -1.33380556e+00 6.83804095e-01 6.93117976e-01
1.51577687e+00 -1.24049628e+00 -3.70973200e-01 -4.79406655e-01
-1.22540724e+00 7.16130674e-01 1.21424103e+00 2.41362661e-01
5.73954403e-01 4.23072614e-02 6.62394166e-01 -1.24009922e-01
-1.12066412e+00 6.11199737e-01 -8.98069665e-02 6.04409158e-01
7.98914731e-01 1.17136903e-01 -6.24062717e-01 7.93090463e-01
-4.04936045e-01 -4.49117631e-01 1.15312040e+00 5.90500653e-01
-4.15261924e-01 -4.38422620e-01 -3.92357498e-01 2.78056651e-01
-8.15028429e-01 -3.46260637e-01 -1.22031593e+00 5.52873850e-01
-4.68660653e-01 1.44649756e+00 -6.10410750e-01 -6.03953362e-01
5.70045948e-01 3.35279852e-02 -4.98193771e-01 -6.82550192e-01
-1.44313169e+00 -2.17223898e-01 6.63164556e-01 -6.62010789e-01
7.08231032e-02 -2.89170235e-01 -6.63935721e-01 -1.04267716e+00
-3.61610711e-01 1.06667310e-01 1.29658803e-01 6.80803537e-01
4.05632734e-01 -5.45338988e-02 1.04957700e+00 -1.13136172e-01
-1.31951463e+00 -1.24027574e+00 -3.90598118e-01 8.82852852e-01
5.82119465e-01 5.05354144e-02 -8.36979866e-01 -5.15457690e-01] | [8.836612701416016, 10.355944633483887] |
5afc6724-c4db-453b-a569-73bdaa5eaa3f | improved-counting-and-localization-from | 2203.15691 | null | https://arxiv.org/abs/2203.15691v1 | https://arxiv.org/pdf/2203.15691v1.pdf | Improved Counting and Localization from Density Maps for Object Detection in 2D and 3D Microscopy Imaging | Object counting and localization are key steps for quantitative analysis in large-scale microscopy applications. This procedure becomes challenging when target objects are overlapping, are densely clustered, and/or present fuzzy boundaries. Previous methods producing density maps based on deep learning have reached a high level of accuracy for object counting by assuming that object counting is equivalent to the integration of the density map. However, this model fails when objects show significant overlap regarding accurate localization. We propose an alternative method to count and localize objects from the density map to overcome this limitation. Our procedure includes the following three key aspects: 1) Proposing a new counting method based on the statistical properties of the density map, 2) optimizing the counting results for those objects which are well-detected based on the proposed counting method, and 3) improving localization of poorly detected objects using the proposed counting method as prior information. Validation includes processing of microscopy data with known ground truth and comparison with other models that use conventional processing of the density map. Our results show improved performance in counting and localization of objects in 2D and 3D microscopy data. Furthermore, the proposed method is generic, considering various applications that rely on the density map approach. Our code will be released post-review. | ['Guido Gerig', 'Thomas Ach', 'Shijie Li'] | 2022-03-29 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [ 1.03020236e-01 -3.87864202e-01 2.55827725e-01 -3.09555858e-01
-5.56486845e-01 -4.10266250e-01 5.51401436e-01 7.09284902e-01
-1.18612444e+00 9.33520794e-01 -4.28245038e-01 1.72893003e-01
-6.58036843e-02 -8.69248986e-01 -7.20487475e-01 -9.07036960e-01
-1.17653064e-01 1.01921606e+00 5.88448882e-01 6.28086388e-01
5.36525667e-01 8.13106418e-01 -1.46639609e+00 -3.39464075e-03
5.84922016e-01 9.08437490e-01 7.78676629e-01 9.96646404e-01
-3.51877421e-01 6.32575274e-01 -9.39843416e-01 -1.03882909e-01
-9.28087309e-02 -1.75534099e-01 -7.39929259e-01 -1.50049075e-01
4.36539859e-01 -6.25260413e-01 4.10050184e-01 1.19199932e+00
4.85877514e-01 -1.29831120e-01 1.10306108e+00 -8.61958086e-01
-4.80501026e-01 8.74224678e-02 -9.36409891e-01 6.87171340e-01
2.70672869e-02 -2.15589926e-01 3.67052674e-01 -8.06706607e-01
4.41705644e-01 1.07270396e+00 8.17985833e-01 4.25757289e-01
-1.35615098e+00 -5.59799194e-01 -3.08179885e-01 7.46609420e-02
-1.75710547e+00 -6.82317615e-02 9.99329835e-02 -7.56932676e-01
9.92212355e-01 -1.00033075e-01 6.31153643e-01 3.01199436e-01
3.09198409e-01 3.70822608e-01 1.26320887e+00 -5.12946248e-01
5.38815498e-01 2.25317821e-01 7.02471957e-02 7.62492299e-01
5.97828448e-01 -3.37704241e-01 -1.99482143e-01 -1.14463449e-01
1.12554562e+00 1.09507777e-01 1.48967609e-01 -1.06078535e-01
-1.33716750e+00 7.81355798e-01 3.32664132e-01 5.42224109e-01
-3.84302408e-01 9.37626287e-02 1.39591590e-01 -5.12253940e-01
7.60636866e-01 2.74840474e-01 -1.36703208e-01 1.55430175e-02
-1.43610215e+00 2.09486336e-01 6.55330122e-01 7.16442287e-01
9.35287952e-01 -4.62480962e-01 -2.99791932e-01 6.42956555e-01
1.12106517e-01 6.85046017e-01 1.24210745e-01 -9.89066660e-01
3.96620110e-02 7.53433585e-01 3.36523116e-01 -9.68672872e-01
-6.28698409e-01 1.00868881e-01 -9.31302845e-01 3.48981768e-01
8.19046021e-01 9.12473649e-02 -1.01143718e+00 1.37681890e+00
4.88451093e-01 1.99765414e-01 -3.17257732e-01 7.76323676e-01
7.86980450e-01 5.86241364e-01 2.05005720e-01 -2.99042374e-01
1.35269964e+00 -3.48230451e-01 -7.69638181e-01 1.88006967e-01
4.65233266e-01 -5.74597239e-01 7.18678713e-01 1.18840545e-01
-1.02034771e+00 -3.95260245e-01 -1.01191676e+00 -2.73856342e-01
-7.41870105e-01 4.61583078e-01 4.88084137e-01 4.31979507e-01
-1.20777845e+00 5.08535922e-01 -1.09749341e+00 -7.17296898e-01
8.84690523e-01 7.22457886e-01 -2.39467159e-01 1.76222846e-01
-4.51887339e-01 1.02787447e+00 5.48301816e-01 -1.67758867e-01
-6.38690054e-01 -7.22352862e-01 -6.66935205e-01 9.27727763e-03
-3.46702367e-01 -6.70943379e-01 9.39899683e-01 -2.36935198e-01
-1.07819569e+00 1.14755523e+00 -5.85898757e-01 -4.06149566e-01
4.82438564e-01 1.81767732e-01 3.96641791e-01 4.21634674e-01
4.23854649e-01 9.17982996e-01 8.84766430e-02 -1.10527790e+00
-9.15563226e-01 -5.53159237e-01 -8.10291395e-02 -2.60308534e-02
-2.14461222e-01 -5.77332228e-02 -2.24988252e-01 -2.25364771e-02
-1.15484439e-01 -3.28609407e-01 -3.46245281e-02 2.67578781e-01
-1.43813059e-01 -2.08319277e-01 8.89143527e-01 -4.76021320e-01
8.25271964e-01 -1.77091181e+00 -1.83163390e-01 -5.98151423e-02
6.21068001e-01 1.94486335e-01 2.26702183e-01 7.43638650e-02
5.73693931e-01 3.56393643e-02 -2.72732615e-01 -6.23388410e-01
-3.30139697e-01 4.10397118e-03 2.28948265e-01 8.16387951e-01
4.21578377e-01 6.73199952e-01 -1.07819200e+00 -1.20219147e+00
7.21307456e-01 8.55650842e-01 -3.49570304e-01 1.28093630e-01
8.48395526e-02 7.79568911e-01 -8.62097666e-02 7.89095640e-01
1.09119689e+00 -6.15940094e-01 -3.06185167e-02 -3.13699692e-01
-4.30399746e-01 -1.81467965e-01 -1.06992853e+00 1.10214496e+00
-3.68839383e-01 5.03597736e-01 1.35836914e-01 -8.54133844e-01
7.22977459e-01 -1.08331606e-01 6.84999943e-01 -2.86530346e-01
4.37091023e-01 4.23987508e-01 -1.21571653e-01 -2.62547106e-01
3.81683111e-01 -4.62459803e-01 3.00118685e-01 3.13009799e-01
6.24508977e-01 -4.48940963e-01 8.28089476e-01 6.99805617e-02
1.07747722e+00 -8.33872631e-02 6.82852566e-01 -5.10037720e-01
5.34570813e-01 -1.56609400e-03 2.14405254e-01 7.70296931e-01
-4.83625799e-01 6.65307105e-01 4.86042440e-01 -3.92626017e-01
-1.17613399e+00 -1.06630898e+00 -7.24273622e-01 5.39361417e-01
3.11168641e-01 2.11499199e-01 -8.72296214e-01 -3.94341141e-01
4.92394110e-03 2.01015994e-01 -8.73236120e-01 5.00544190e-01
-3.90522450e-01 -1.31089282e+00 4.02792603e-01 4.19224918e-01
5.30022442e-01 -1.04937208e+00 -7.12629616e-01 1.61040992e-01
-2.82113105e-01 -1.22870052e+00 -4.55948487e-02 4.93573934e-01
-7.65829265e-01 -1.21679771e+00 -1.04346204e+00 -7.71483302e-01
1.07451212e+00 3.50762725e-01 1.01074684e+00 2.87387580e-01
-4.56764758e-01 1.65658325e-01 -5.22189848e-02 -6.84116721e-01
-2.28387684e-01 -8.55475944e-03 2.46225208e-01 -3.99869204e-01
8.87836397e-01 -4.86352205e-01 -6.15210414e-01 2.32872218e-01
-8.56460154e-01 -2.43214056e-01 4.21364099e-01 6.91426277e-01
1.05477309e+00 2.07722738e-01 5.58200717e-01 -6.30921125e-01
3.63830686e-01 -4.24974799e-01 -9.76580262e-01 1.51158273e-01
-1.52311370e-01 -2.04216838e-01 5.18141329e-01 -4.44905996e-01
-6.09587133e-01 1.76939473e-01 -1.40536025e-01 -3.71097803e-01
-2.11343482e-01 -8.94075707e-02 1.80250078e-01 -4.57811713e-01
5.65996647e-01 2.05714226e-01 2.09453270e-01 -1.74666435e-01
-2.27462366e-01 4.99734461e-01 4.47137773e-01 -4.22219366e-01
3.62427860e-01 1.09401095e+00 3.66380900e-01 -8.32622409e-01
-7.28941739e-01 -9.94643807e-01 -1.04230607e+00 -1.68343186e-01
1.00321412e+00 -5.86049795e-01 -1.17546058e+00 5.95575809e-01
-1.43130541e+00 -2.00285882e-01 -2.56063432e-01 8.29781413e-01
-6.41441345e-01 4.68918353e-01 -7.32112408e-01 -1.06926644e+00
-3.24941278e-01 -1.12934577e+00 1.50567555e+00 5.29019237e-01
-4.11472544e-02 -1.34185064e+00 3.36938024e-01 -1.36504069e-01
4.05320525e-01 1.22243509e-01 7.06871808e-01 -5.14483750e-01
-4.46249366e-01 -3.91546100e-01 -6.31636262e-01 1.88361123e-01
2.85471380e-01 2.37813950e-01 -9.69918132e-01 -1.24033742e-01
-4.13464010e-02 -2.28938282e-01 8.99160981e-01 1.05798268e+00
1.10782635e+00 -7.21079414e-05 -6.31522357e-01 5.68419397e-01
1.71429360e+00 -5.10623120e-02 5.59689879e-01 2.40785599e-01
3.19878876e-01 6.54256523e-01 4.03054625e-01 4.24500883e-01
2.79949248e-01 4.52438831e-01 4.69391942e-01 -1.70941219e-01
-9.17719379e-02 2.45411508e-02 -2.16480255e-01 6.30676031e-01
-2.89541245e-01 -2.06360489e-01 -9.10360634e-01 6.59328759e-01
-1.52086580e+00 -1.05909455e+00 -2.33139932e-01 2.19997621e+00
7.31533766e-01 -1.18084453e-01 3.63856703e-01 7.46466266e-03
9.60846961e-01 -4.49022800e-01 -3.86673599e-01 -4.04943861e-02
-7.69218877e-02 3.34612101e-01 6.88725352e-01 6.63836360e-01
-1.28786945e+00 6.25552416e-01 7.00647926e+00 8.91671598e-01
-9.37335849e-01 2.13696942e-01 8.00988376e-01 -1.36035085e-01
3.91484141e-01 -3.87423605e-01 -1.36970782e+00 7.44348884e-01
5.71920455e-01 -2.68234424e-02 1.02732703e-01 6.31467462e-01
2.42186889e-01 -8.39344203e-01 -9.91168559e-01 9.56644177e-01
7.26103177e-03 -1.35979998e+00 1.51640922e-01 3.22833002e-01
4.95101660e-01 -6.08102791e-02 -1.78304702e-01 8.59993845e-02
2.30539754e-01 -1.00779414e+00 5.32696009e-01 6.26428962e-01
8.96858811e-01 -5.52836299e-01 1.37580979e+00 5.50190389e-01
-1.08089972e+00 3.63046378e-01 -8.30055118e-01 -1.24828435e-01
2.75676638e-01 1.04446316e+00 -1.05017769e+00 -3.03740613e-02
7.89261997e-01 3.70383948e-01 -6.40825868e-01 1.37357485e+00
2.39243835e-01 1.68087974e-01 -7.20009744e-01 -2.56258667e-01
-4.47842814e-02 -2.80123740e-01 2.00923175e-01 1.53876472e+00
5.43930352e-01 -2.10255697e-01 -7.17093199e-02 1.33157086e+00
2.83911861e-02 2.98016537e-02 -5.40000439e-01 -4.19924501e-04
4.62044448e-01 1.56239271e+00 -1.58755636e+00 -4.62438405e-01
-1.77541837e-01 5.46258509e-01 6.19072735e-01 -3.28759439e-02
-9.04706240e-01 -2.50413269e-01 1.36810303e-01 4.87962067e-01
2.51362234e-01 -1.87855855e-01 -2.66524225e-01 -6.81977987e-01
-7.56498873e-02 6.20549694e-02 1.46230787e-01 -5.55317760e-01
-1.50468445e+00 3.28159958e-01 3.05853546e-01 -1.00460291e+00
3.69148068e-02 -8.88540506e-01 -5.18379867e-01 8.17276239e-01
-1.45533872e+00 -1.12996531e+00 -4.01871532e-01 2.48047814e-01
1.48467556e-01 2.71840453e-01 8.02443266e-01 6.14657521e-01
-8.22724178e-02 1.71229184e-01 1.07826047e-01 -3.20926271e-02
3.42641264e-01 -1.49503577e+00 -2.28209540e-01 5.31616092e-01
-6.42482713e-02 6.95093334e-01 4.13635075e-01 -6.90576851e-01
-5.75554729e-01 -9.50704694e-01 8.40492845e-01 -7.15609789e-01
4.92244303e-01 -5.03502071e-01 -8.45263243e-01 2.91141748e-01
-3.40617985e-01 2.98981100e-01 6.56686127e-01 -1.86993241e-01
2.19389930e-01 2.90287405e-01 -1.74205661e+00 -2.01555546e-02
5.42036533e-01 -2.26577565e-01 -1.51132450e-01 5.97341061e-01
1.06182612e-01 -2.29167268e-01 -7.59067476e-01 4.02241826e-01
5.08230805e-01 -9.70280409e-01 9.20965850e-01 2.11795375e-01
3.90911192e-01 -6.07267797e-01 -8.82219374e-02 -8.46857667e-01
-3.88132870e-01 2.90771484e-01 -1.36460811e-01 1.02845204e+00
5.35930172e-02 -2.98733264e-01 8.89279842e-01 -6.28089085e-02
2.79917121e-01 -5.78523338e-01 -1.24805641e+00 -7.98345029e-01
1.88321769e-01 1.02221526e-01 2.76076436e-01 6.67751610e-01
-1.08471796e-01 1.20652966e-01 1.10991992e-01 5.20324782e-02
9.18677807e-01 -2.93045104e-01 7.51995385e-01 -1.32215393e+00
1.32983878e-01 -4.54192191e-01 -8.81158888e-01 -8.53456914e-01
-1.33630604e-01 -7.56899655e-01 2.58783668e-01 -1.89449906e+00
9.22282696e-01 -2.49322489e-01 -3.50859426e-02 8.62884521e-02
-2.23085850e-01 8.00221086e-01 -1.17255270e-01 4.90661561e-01
-9.37812507e-01 5.94231933e-02 1.27622700e+00 -1.25756055e-01
1.39519081e-01 -1.19494654e-01 -1.45219862e-01 8.90920937e-01
7.29498506e-01 -5.80660284e-01 8.23801309e-02 -3.10466588e-01
8.73035789e-02 -3.33624601e-01 6.74801171e-01 -1.36116993e+00
3.42964858e-01 -4.76992019e-02 8.91584277e-01 -1.00133550e+00
4.81622159e-01 -7.11126745e-01 -1.84248060e-01 4.19438720e-01
1.47221699e-01 -2.89129525e-01 3.02838862e-01 5.59942842e-01
-9.83053818e-02 -5.24493396e-01 1.30731153e+00 -4.13841486e-01
-3.57716352e-01 1.49215296e-01 -5.46030760e-01 -2.65474617e-01
1.04778290e+00 -6.54114783e-01 -4.23433870e-01 -1.13810152e-01
-6.55496895e-01 -3.22558545e-02 6.50405884e-01 -3.72183114e-01
3.93321812e-01 -1.16004574e+00 -6.35174096e-01 -1.19413674e-01
-5.61123379e-02 2.87467182e-01 9.90879908e-02 1.17449701e+00
-1.02890503e+00 5.76957464e-01 -3.62530857e-01 -1.08857393e+00
-1.13046169e+00 5.05865276e-01 4.67854470e-01 -6.23451114e-01
-2.03485861e-01 9.25482631e-01 5.34255564e-01 -5.78115761e-01
-3.79287358e-03 -5.26124001e-01 -3.98668557e-01 -7.28140622e-02
7.92921364e-01 5.99926770e-01 7.77656287e-02 -6.84588432e-01
-6.43678904e-01 8.24066281e-01 1.19575355e-02 3.09669506e-02
1.28633320e+00 -1.03907928e-01 -3.82796437e-01 7.34358251e-01
1.18705523e+00 -2.49028698e-01 -1.18610954e+00 8.64535496e-02
-1.80861712e-01 -5.67747295e-01 -1.97184980e-02 -5.20459771e-01
-7.32965827e-01 1.12595046e+00 8.48174274e-01 2.10779130e-01
6.67969882e-01 1.90196678e-01 2.77382642e-01 1.38836920e-01
6.96555912e-01 -1.02097464e+00 5.94299212e-02 5.42388320e-01
4.08169597e-01 -1.45590103e+00 4.82938349e-01 -2.69448936e-01
8.95214453e-02 1.05474734e+00 6.48291409e-01 -2.09901854e-01
7.04463124e-01 7.14681149e-01 -2.83501357e-01 -4.52410579e-01
-2.28854731e-01 -3.58676106e-01 -1.01772882e-02 1.02131927e+00
4.96533811e-01 -4.62341383e-02 -4.17919844e-01 2.50959307e-01
1.15587331e-01 3.36539537e-01 5.01413763e-01 9.44490492e-01
-7.31412232e-01 -4.19458866e-01 -6.10116661e-01 7.65729010e-01
-7.08940864e-01 2.56069917e-02 -3.04818898e-01 8.22293282e-01
4.52894658e-01 7.53527701e-01 6.60032034e-01 5.76907732e-02
-5.31664453e-02 -3.32048833e-01 9.01337981e-01 -6.90499246e-01
-1.48002312e-01 -1.44233080e-02 -6.77521527e-01 -9.07806158e-02
-9.06988919e-01 -4.30159062e-01 -1.51196027e+00 -6.59619153e-01
-6.94107056e-01 -1.69773325e-02 8.62745464e-01 8.54716182e-01
-3.72771472e-02 3.50265563e-01 -3.58137228e-02 -1.29817486e+00
2.26302329e-03 -1.25020432e+00 -9.71050382e-01 1.28409103e-01
1.71409488e-01 -1.11825132e+00 -6.39086485e-01 3.13119531e-01] | [14.655692100524902, -3.1959433555603027] |
aabea63f-aa77-4571-a9ba-5215cc4a97ad | forecasting-localized-weather-impacts-on | 2303.16198 | null | https://arxiv.org/abs/2303.16198v1 | https://arxiv.org/pdf/2303.16198v1.pdf | Forecasting localized weather impacts on vegetation as seen from space with meteo-guided video prediction | We present a novel approach for modeling vegetation response to weather in Europe as measured by the Sentinel 2 satellite. Existing satellite imagery forecasting approaches focus on photorealistic quality of the multispectral images, while derived vegetation dynamics have not yet received as much attention. We leverage both spatial and temporal context by extending state-of-the-art video prediction methods with weather guidance. We extend the EarthNet2021 dataset to be suitable for vegetation modeling by introducing a learned cloud mask and an appropriate evaluation scheme. Qualitative and quantitative experiments demonstrate superior performance of our approach over a wide variety of baseline methods, including leading approaches to satellite imagery forecasting. Additionally, we show how our modeled vegetation dynamics can be leveraged in a downstream task: inferring gross primary productivity for carbon monitoring. To the best of our knowledge, this work presents the first models for continental-scale vegetation modeling at fine resolution able to capture anomalies beyond the seasonal cycle, thereby paving the way for predictive assessments of vegetation status. | ['Markus Reichstein', 'Mélanie Weynants', 'Nora Linscheid', 'Zhihan Gao', 'José Cortés', 'Lazaro Alonso', 'Claire Robin', 'Christian Requena-Mesa', 'Vitus Benson'] | 2023-03-28 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [ 2.81694949e-01 -3.79743159e-01 -2.44019434e-01 -2.75998533e-01
-1.73536703e-01 -1.04938090e+00 8.83204401e-01 1.41213670e-01
-9.17015299e-02 6.71421766e-01 2.17222765e-01 -6.41698182e-01
1.21223137e-01 -1.27868307e+00 -5.99478245e-01 -6.94987178e-01
-5.51526129e-01 -1.22220524e-01 2.10662082e-01 -7.16152787e-01
-2.22523227e-01 6.90892279e-01 -1.88878345e+00 2.28258148e-01
8.26362789e-01 7.40116954e-01 5.81368446e-01 8.08200002e-01
9.06657502e-02 5.17446458e-01 -1.20298937e-01 2.46999159e-01
3.78069848e-01 -5.66079468e-02 -5.75015783e-01 2.63869643e-01
6.18167281e-01 -4.76600498e-01 1.02216443e-02 9.27850842e-01
8.13862532e-02 -3.81531790e-02 4.59618777e-01 -1.02473426e+00
-3.55932087e-01 3.48668814e-01 -7.42664635e-01 3.37397665e-01
-8.60055834e-02 4.06080037e-01 1.09209430e+00 -6.00406647e-01
7.85897255e-01 9.32615876e-01 7.13946760e-01 1.00382924e-01
-1.63654065e+00 -5.16378760e-01 6.38159692e-01 -4.48170453e-02
-1.26123595e+00 -2.68457681e-01 3.74229819e-01 -1.02710664e+00
9.91874218e-01 7.07163811e-01 9.65832293e-01 9.16124761e-01
-2.28678882e-02 5.09947538e-01 1.36213624e+00 -2.62886286e-01
2.81560570e-02 -1.03979312e-01 -1.77955300e-01 4.39029396e-01
1.35214284e-01 6.74469590e-01 -1.22751296e-01 1.99841708e-01
5.49936712e-01 4.90555018e-02 -4.16112632e-01 -2.10420802e-01
-1.23160970e+00 5.95148683e-01 7.54141808e-01 2.07489058e-01
-8.00444841e-01 1.48509040e-01 1.80855229e-01 1.60669193e-01
1.09058845e+00 9.21740010e-02 -7.70599782e-01 1.95948645e-01
-1.59119868e+00 4.68629837e-01 3.80268216e-01 5.20060897e-01
6.78784609e-01 3.67791921e-01 6.08418649e-03 5.21628022e-01
3.99249643e-01 1.05707037e+00 -3.35760325e-01 -1.02271736e+00
1.48525223e-01 3.97125781e-01 5.66493154e-01 -1.12402093e+00
-3.74490708e-01 -4.73680705e-01 -8.27712774e-01 6.03884637e-01
1.60154790e-01 -2.60421634e-01 -8.21019709e-01 1.57993913e+00
2.96338111e-01 4.54736501e-01 1.18832178e-01 9.93134558e-01
6.08631611e-01 1.00130546e+00 4.18029249e-01 -2.40046993e-01
1.16413987e+00 -7.15011895e-01 -4.57871675e-01 -2.99377263e-01
5.81752837e-01 -5.04077196e-01 8.29065621e-01 1.04753107e-01
-4.44960773e-01 -4.31880862e-01 -6.94620728e-01 6.91010237e-01
-7.31415272e-01 2.45046064e-01 7.99384773e-01 4.94832754e-01
-1.33738339e+00 5.79157472e-01 -9.19998586e-01 -7.48234570e-01
2.24129856e-01 -2.80764580e-01 -2.48299971e-01 2.08780080e-01
-1.12952793e+00 9.57221389e-01 4.39188749e-01 4.44996923e-01
-9.00605679e-01 -9.62833703e-01 -8.14400911e-01 1.23286143e-01
1.19720735e-01 -3.49734336e-01 1.02886546e+00 -1.40030062e+00
-1.07365060e+00 1.00652146e+00 -7.91856498e-02 -7.72811651e-01
4.18617457e-01 1.16179764e-01 -6.62293494e-01 8.38010479e-03
1.61301494e-01 1.08287466e+00 6.77506089e-01 -1.53729343e+00
-8.70580733e-01 -2.87960708e-01 1.77101329e-01 1.76013827e-01
-1.77835226e-01 1.11440122e-01 1.11203060e-01 -9.11700487e-01
-3.02330971e-01 -9.97335076e-01 -4.31776553e-01 3.61674398e-01
2.04524413e-01 5.52975297e-01 8.42974663e-01 -9.44172323e-01
1.19686759e+00 -1.90625000e+00 1.65170819e-01 1.19805820e-01
-9.33644250e-02 4.63819385e-01 -4.66339827e-01 5.58261454e-01
-7.15129152e-02 5.00543952e-01 -5.19580126e-01 -7.53338123e-03
-2.28864163e-01 5.31907976e-01 -8.35610688e-01 3.03762972e-01
3.32317561e-01 7.84488082e-01 -8.27391028e-01 -3.06244032e-03
5.16953051e-01 4.98008728e-01 -2.08693549e-01 -6.28789961e-02
-6.67039573e-01 4.93985772e-01 -1.42832667e-01 7.19644606e-01
1.13140225e+00 -2.74777040e-02 4.01026517e-01 1.13758378e-01
-8.24093282e-01 -6.34927824e-02 -9.64928150e-01 1.24465752e+00
-3.73690009e-01 7.59731233e-01 4.44997281e-01 -7.84407318e-01
7.29356885e-01 3.53828073e-01 3.79732788e-01 -5.20980835e-01
-3.35068047e-01 1.80459172e-02 -1.54446840e-01 -3.38185787e-01
5.90656877e-01 -1.12218313e-01 4.93039578e-01 -3.48905995e-02
-3.12198251e-01 -2.96338677e-01 7.13663995e-02 -9.39724371e-02
3.69697332e-01 7.35791028e-01 1.56018406e-01 -5.83629429e-01
3.43586594e-01 6.54566705e-01 3.23095858e-01 5.68450630e-01
-1.94666654e-01 3.82272214e-01 2.33455852e-01 -6.61444366e-01
-1.01666701e+00 -5.85931480e-01 -3.16760421e-01 1.05018747e+00
-1.90583333e-01 -3.72795641e-01 -4.08184916e-01 -2.79518306e-01
2.79367179e-01 6.88803852e-01 -8.46374035e-01 5.01667261e-01
8.89182761e-02 -1.21980846e+00 6.49866164e-01 4.69920069e-01
5.11541009e-01 -9.46982920e-01 -7.52019286e-01 3.09858471e-01
-2.11988837e-01 -1.18122983e+00 3.60112876e-01 -1.62001371e-01
-1.04646873e+00 -8.83237541e-01 -6.23218000e-01 -1.42813504e-01
1.36159047e-01 6.49225652e-01 1.32206106e+00 1.90639555e-01
-1.88263372e-01 4.82060254e-01 -3.24977219e-01 -3.82959902e-01
-2.42823392e-01 3.15908454e-02 -1.06399946e-01 -3.88894528e-02
4.38362472e-02 -6.35812759e-01 -7.04542994e-01 2.92668641e-01
-1.06474662e+00 4.19390649e-01 2.02866167e-01 4.13373977e-01
4.96250004e-01 -1.32574514e-01 -1.52304759e-02 -5.45970201e-01
-6.21353500e-02 -7.90786207e-01 -1.14394212e+00 1.56891614e-01
-6.82260931e-01 -3.92203510e-01 1.81198835e-01 5.71589358e-02
-1.23422348e+00 3.86181384e-01 5.88375069e-02 -8.67459625e-02
-7.27281690e-01 1.17801750e+00 3.89037997e-01 -1.94572821e-01
6.42941117e-01 3.42577010e-01 -1.28225580e-01 -4.04854596e-01
4.41643924e-01 5.24768353e-01 4.86412287e-01 -3.21276575e-01
9.16328132e-01 9.15692210e-01 1.98738605e-01 -1.33010721e+00
-7.39096522e-01 -6.03004038e-01 -7.41187334e-01 -5.32918751e-01
6.25121295e-01 -1.43653905e+00 -3.29741925e-01 5.77086985e-01
-9.35988665e-01 -6.49512827e-01 -8.93808305e-02 2.97627479e-01
-1.75516054e-01 3.12912524e-01 -1.53728291e-01 -1.15294790e+00
-2.58179128e-01 -6.46040976e-01 1.04901969e+00 -7.74229839e-02
1.46764606e-01 -1.23882580e+00 4.62634116e-01 8.07482749e-03
7.95650303e-01 8.16651404e-01 2.65499860e-01 3.82086039e-01
-7.57040203e-01 2.17762768e-01 -4.96773541e-01 1.50307029e-01
1.94535583e-01 7.32846320e-01 -1.22160888e+00 -1.96927875e-01
-5.81657469e-01 4.76398505e-02 1.46752322e+00 8.03230286e-01
8.55699897e-01 -1.20773688e-01 -2.56214768e-01 8.79196346e-01
1.75920320e+00 -2.28031978e-01 7.17400074e-01 7.67993212e-01
5.63545287e-01 7.87305176e-01 7.80035198e-01 5.44798374e-01
5.19701779e-01 5.95850527e-01 1.14882421e+00 -6.08337641e-01
8.08964074e-02 2.58538783e-01 3.40504467e-01 1.04625456e-01
-5.58041573e-01 -1.62605301e-01 -1.54600930e+00 7.97965229e-01
-2.02383590e+00 -1.47116959e+00 -5.59934974e-01 2.10485840e+00
4.30473119e-01 -2.49246448e-01 -1.56737566e-02 -2.69771338e-01
4.42067862e-01 8.77170265e-01 -2.82987028e-01 -2.99535673e-02
-6.28574491e-01 8.51331726e-02 9.97613728e-01 6.07229590e-01
-1.48686135e+00 1.36512589e+00 6.68449545e+00 1.33225739e-01
-1.41325033e+00 3.97620238e-02 4.05931473e-01 5.60460091e-02
-4.41445500e-01 3.55270237e-01 -6.33875489e-01 7.07355738e-02
1.10794508e+00 3.98662478e-01 4.77944970e-01 4.74402398e-01
9.44672883e-01 -2.87438244e-01 -1.78597838e-01 2.31906325e-01
-5.20917237e-01 -1.63498044e+00 -5.37078567e-02 1.60617247e-01
8.79259467e-01 6.33373499e-01 -7.30947852e-02 2.89294496e-02
4.97382492e-01 -8.49678218e-01 6.32703245e-01 8.99213612e-01
7.21864641e-01 -2.59391755e-01 3.68801504e-01 3.17099005e-01
-1.64773607e+00 4.54675145e-02 -1.83196813e-01 -4.62539554e-01
-9.80554670e-02 4.40674245e-01 -3.56208414e-01 6.90231085e-01
9.27877247e-01 1.28320253e+00 -6.22981727e-01 9.60445642e-01
-1.23267323e-02 1.24428260e+00 -5.24640143e-01 5.18732131e-01
6.03565514e-01 -2.57924139e-01 7.04490364e-01 1.33896470e+00
5.21734059e-01 1.92125008e-01 3.48651797e-01 6.78967834e-01
2.64146060e-01 2.29275584e-01 -8.23590636e-01 -1.11659154e-01
2.01269522e-01 1.29041171e+00 -5.17598927e-01 -2.06212372e-01
-5.95099688e-01 7.17830896e-01 -1.25349596e-01 4.57105666e-01
-5.51405311e-01 3.47908020e-01 1.08165765e+00 2.53646374e-01
5.54112732e-01 -4.88199979e-01 -8.22334960e-02 -1.51831305e+00
-2.26182699e-01 -7.20043123e-01 1.82905331e-01 -1.24358511e+00
-7.85082996e-01 2.91554391e-01 9.34368670e-02 -1.45632493e+00
-9.77171063e-02 -6.99016988e-01 -7.10816026e-01 1.20387292e+00
-2.40664339e+00 -1.70440555e+00 -6.88434541e-01 1.77046329e-01
3.38751465e-01 -1.46407983e-03 1.30370879e+00 -3.35827231e-01
-4.58786309e-01 -6.61013544e-01 5.39872169e-01 -3.54099423e-01
4.88557458e-01 -1.24671888e+00 8.02237034e-01 1.33397102e+00
-9.05284882e-02 6.18444406e-04 7.81865001e-01 -6.69771910e-01
-1.04968858e+00 -1.49939990e+00 7.59766638e-01 -3.44744176e-02
9.53195751e-01 1.02170341e-01 -9.56195414e-01 7.31970310e-01
2.03374535e-01 2.92557955e-01 5.07658780e-01 2.30156258e-01
-4.40195292e-01 -4.07599866e-01 -8.34832489e-01 5.11038899e-01
7.17599034e-01 -4.42405462e-01 3.05791367e-02 5.18035293e-01
3.70359838e-01 -1.64499909e-01 -8.69618297e-01 5.25642157e-01
7.18113661e-01 -9.35853124e-01 8.69017303e-01 -6.86397493e-01
6.59823596e-01 -6.10682130e-01 -7.11260319e-01 -1.53989971e+00
-7.29025066e-01 -2.44609088e-01 3.30392160e-02 1.16948962e+00
2.99299657e-01 -2.66305149e-01 4.65798080e-01 2.98278153e-01
2.36502141e-01 7.28119910e-02 -5.66243529e-01 -7.40229309e-01
2.92132884e-01 -4.63801235e-01 7.16133893e-01 1.08528304e+00
-5.90557456e-01 -3.67037088e-01 -5.90244114e-01 1.01264346e+00
4.38787550e-01 5.45398772e-01 6.97478950e-01 -1.73360729e+00
-6.57687783e-02 -5.34644067e-01 -2.77908742e-01 -5.28520048e-01
1.89008087e-01 -5.13424158e-01 -2.25671530e-01 -1.37336278e+00
-3.83744352e-02 -3.15343887e-01 -2.11814120e-01 7.17932343e-01
-5.52878492e-02 3.64155740e-01 2.36335203e-01 2.31906876e-01
1.72879010e-01 5.19833088e-01 8.17455471e-01 -2.46258989e-01
-4.37919460e-02 -1.09325685e-01 -2.97216386e-01 5.34887612e-01
1.23403919e+00 -3.27994913e-01 -7.44375661e-02 -5.75335503e-01
3.25430065e-01 1.20012812e-01 9.21541035e-01 -8.97749662e-01
-5.39017618e-01 -8.83465230e-01 2.98478961e-01 -5.64368784e-01
6.20237291e-02 -1.02452803e+00 4.94942218e-01 3.43063116e-01
-8.14509615e-02 -5.86478077e-02 8.46426666e-01 5.56585968e-01
-1.11261308e-01 9.14952010e-02 5.63296080e-01 -1.38936907e-01
-1.24308324e+00 5.28749824e-01 -8.72375429e-01 -3.83440882e-01
8.60462964e-01 -1.40847964e-03 -3.80723923e-01 -3.03214997e-01
-8.82858217e-01 4.47740912e-01 7.13751733e-01 5.05749047e-01
4.85958159e-02 -9.40413117e-01 -1.05796874e+00 1.23398051e-01
4.09865171e-01 -6.42350554e-01 2.28214726e-01 6.50064766e-01
-6.96055353e-01 4.83613193e-01 -3.94281358e-01 -8.40485096e-01
-1.43564773e+00 2.94086576e-01 7.72246540e-01 -1.70620099e-01
-4.04932052e-01 4.36314791e-01 -5.12771606e-02 -7.42276192e-01
-3.03697675e-01 -5.36396205e-01 -4.14574295e-01 4.44061697e-01
4.26132709e-01 -1.49439171e-01 9.08337608e-02 -1.06061161e+00
-3.46285790e-01 3.44541281e-01 5.62589526e-01 -3.68495077e-01
1.68164790e+00 -3.64885300e-01 -4.29076180e-02 6.06383324e-01
3.85900885e-01 -1.96954131e-01 -1.69665682e+00 -2.75098830e-01
-6.17014915e-02 -6.57581091e-01 7.98464239e-01 -1.11792457e+00
-1.26464033e+00 9.63214934e-01 9.69188750e-01 3.93127233e-01
1.26336908e+00 -6.62131965e-01 4.50513028e-02 3.08675766e-01
1.69208765e-01 -7.38580287e-01 -9.12728667e-01 4.56468076e-01
9.93513405e-01 -1.56751955e+00 2.92853620e-02 -2.89027631e-01
-6.55961096e-01 1.15702677e+00 3.16975951e-01 1.25784069e-01
1.02136934e+00 3.18323791e-01 2.92279661e-01 -6.71001375e-02
-8.75195146e-01 -8.62486243e-01 3.46908450e-01 7.62306929e-01
4.92793411e-01 7.05531836e-01 1.01621270e-01 -2.92380042e-02
2.14661539e-01 3.16525817e-01 4.97959226e-01 8.50545466e-01
-5.79639256e-01 -9.50101137e-01 -5.26765347e-01 3.74473721e-01
-1.99235260e-01 -5.26405215e-01 -1.28690511e-01 6.60005450e-01
-3.14002633e-02 7.32758880e-01 9.98566896e-02 -6.23436347e-02
2.00041562e-01 -5.17392680e-02 1.92149922e-01 -5.62028766e-01
-4.94382441e-01 7.95883909e-02 1.43485323e-01 -4.72244292e-01
-1.05711913e+00 -1.14073300e+00 -5.07364631e-01 -6.66357100e-01
-2.02845946e-01 -3.30628067e-01 7.25376964e-01 8.69455040e-01
5.29107511e-01 3.42952132e-01 4.78618920e-01 -1.44865358e+00
2.90885866e-02 -9.18033063e-01 -8.23229671e-01 -2.10949332e-02
6.53779685e-01 -5.85148335e-01 -1.24767125e-01 2.74525404e-01] | [9.5361328125, -1.5697654485702515] |
87ac747e-37df-4858-bb55-bb1ee25cfee4 | contactless-human-activity-recognition-using | 2304.09756 | null | https://arxiv.org/abs/2304.09756v1 | https://arxiv.org/pdf/2304.09756v1.pdf | Contactless Human Activity Recognition using Deep Learning with Flexible and Scalable Software Define Radio | Ambient computing is gaining popularity as a major technological advancement for the future. The modern era has witnessed a surge in the advancement in healthcare systems, with viable radio frequency solutions proposed for remote and unobtrusive human activity recognition (HAR). Specifically, this study investigates the use of Wi-Fi channel state information (CSI) as a novel method of ambient sensing that can be employed as a contactless means of recognizing human activity in indoor environments. These methods avoid additional costly hardware required for vision-based systems, which are privacy-intrusive, by (re)using Wi-Fi CSI for various safety and security applications. During an experiment utilizing universal software-defined radio (USRP) to collect CSI samples, it was observed that a subject engaged in six distinct activities, which included no activity, standing, sitting, and leaning forward, across different areas of the room. Additionally, more CSI samples were collected when the subject walked in two different directions. This study presents a Wi-Fi CSI-based HAR system that assesses and contrasts deep learning approaches, namely convolutional neural network (CNN), long short-term memory (LSTM), and hybrid (LSTM+CNN), employed for accurate activity recognition. The experimental results indicate that LSTM surpasses current models and achieves an average accuracy of 95.3% in multi-activity classification when compared to CNN and hybrid techniques. In the future, research needs to study the significance of resilience in diverse and dynamic environments to identify the activity of multiple users. | ['Qammer H. Abbasi', 'Anis Koubaa', 'Syed Aziz Shah', 'Matthew Broadbent', 'Wadii Boulila', 'Jawad Ahmad', 'Muhammad Zakir Khan'] | 2023-04-18 | null | null | null | null | ['human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'time-series'] | [ 4.93200421e-01 -1.94327012e-01 9.10854191e-02 -1.44452751e-01
-5.48572421e-01 -2.05232114e-01 1.46823332e-01 -1.49943650e-01
-5.16868591e-01 1.09234703e+00 3.78491879e-01 -3.88228565e-01
-2.47562006e-01 -6.12200975e-01 -1.46190539e-01 -9.64623511e-01
-3.63099486e-01 -5.78302383e-01 -1.46649048e-01 1.63056731e-01
-1.85369238e-01 2.84221292e-01 -1.40529716e+00 1.76979855e-01
3.33324730e-01 1.55421424e+00 6.73158169e-02 1.02616644e+00
2.97765255e-01 1.02491844e+00 -1.03448009e+00 2.79005766e-01
8.47551823e-02 -2.84426302e-01 -3.91677409e-01 -6.14792883e-01
1.55914590e-01 -3.79109532e-01 -4.55920815e-01 5.20542562e-01
9.90990400e-01 2.90646225e-01 -2.14937944e-02 -1.11569440e+00
-1.16583601e-01 1.04070678e-01 -6.70607239e-02 5.36279023e-01
1.09402192e+00 -1.71640441e-01 1.34374872e-01 -3.46026480e-01
-9.80575308e-02 3.05276304e-01 1.05554926e+00 4.07555461e-01
-4.00713503e-01 -9.11398232e-01 -2.69088864e-01 3.16092908e-01
-1.56052065e+00 -5.31729221e-01 7.75255620e-01 -8.21557567e-02
1.24639606e+00 5.44204712e-01 8.29014182e-01 1.81983685e+00
5.16721189e-01 4.19558287e-01 1.01244545e+00 -5.41573644e-01
5.23000419e-01 1.32680923e-01 1.73467696e-01 6.62586868e-01
6.12416089e-01 2.77662992e-01 -8.49616587e-01 -4.27993476e-01
7.15863585e-01 4.64155912e-01 -5.03490448e-01 1.00797825e-01
-1.35753143e+00 7.08163232e-02 2.22093806e-01 1.01729822e+00
-6.89171612e-01 1.82124913e-01 3.07336748e-01 4.37112583e-04
1.12437002e-01 2.88762122e-01 -5.37478507e-01 -9.30271566e-01
-8.27414811e-01 -2.25181580e-01 1.09747088e+00 8.88310015e-01
8.57549980e-02 1.90398902e-01 -3.63014817e-01 4.88279760e-01
3.88192803e-01 6.93115413e-01 7.00364351e-01 -8.04292083e-01
4.28953111e-01 1.24436982e-01 4.77830380e-01 -1.06762183e+00
-7.61931896e-01 -6.72521412e-01 -1.07571447e+00 -2.91626692e-01
2.38289729e-01 -6.45275176e-01 -5.45070052e-01 1.54636300e+00
7.67065734e-02 5.76462567e-01 6.26455154e-03 4.46592897e-01
6.90511823e-01 3.18336844e-01 2.68526495e-01 -3.02810580e-01
1.43914413e+00 -6.93310678e-01 -1.20187652e+00 -2.61663109e-01
4.19460684e-01 -2.46685520e-01 6.03005171e-01 5.49991488e-01
-6.61628962e-01 -7.12841153e-01 -1.55725777e+00 6.22021675e-01
-7.38840044e-01 -3.71970013e-02 8.11053991e-01 1.54866326e+00
-9.53284740e-01 4.47691977e-01 -1.07322407e+00 -5.21424711e-01
3.54963034e-01 5.39791763e-01 -3.13052446e-01 1.17709547e-01
-1.41803074e+00 7.13584185e-01 -6.64668111e-03 4.37476665e-01
-5.37667274e-01 -2.46025279e-01 -8.76716495e-01 -1.01027219e-02
-3.19975540e-02 -6.20781839e-01 1.10294592e+00 -8.47429752e-01
-1.60507858e+00 2.95382559e-01 -2.72433549e-01 -5.64793289e-01
3.57995182e-01 -4.10659045e-01 -1.15750384e+00 -7.64514282e-02
-2.35545769e-01 -3.27031732e-01 4.50792104e-01 -7.79057443e-01
-5.43497860e-01 -5.85072279e-01 -2.59101409e-02 -8.33715498e-02
-4.86992866e-01 -1.14515603e-01 2.46423006e-01 -3.82805854e-01
2.48810381e-01 -9.89668489e-01 -3.69630903e-02 -1.69928044e-01
-1.87376574e-01 2.85369784e-01 6.84636891e-01 -8.94922435e-01
1.52393341e+00 -2.05518365e+00 -9.14231956e-01 4.55550164e-01
1.54902712e-01 5.55640757e-01 6.70914114e-01 1.87857017e-01
1.98044762e-01 -1.00226678e-01 -1.92630049e-02 -4.58439410e-01
-1.90553144e-01 2.54066557e-01 3.56727093e-01 5.66031337e-01
-6.61510050e-01 7.65877366e-01 -1.02727413e+00 -1.46851584e-01
4.60360199e-01 6.67028427e-01 1.97215408e-01 2.33955428e-01
9.41435158e-01 6.50652409e-01 -4.29409176e-01 7.94712961e-01
4.76666033e-01 -3.10354196e-02 2.77435154e-01 -2.18715146e-01
-2.16156051e-01 1.19937077e-01 -1.27618611e+00 1.81900930e+00
-7.62169659e-01 6.71763241e-01 -1.09685034e-01 -7.84167528e-01
7.90553808e-01 9.58382726e-01 5.91680646e-01 -9.84847128e-01
4.96932000e-01 6.50245249e-02 -2.72247493e-01 -8.81065071e-01
1.62839428e-01 3.29307675e-01 -2.03006893e-01 2.45859519e-01
-1.74891219e-01 9.99062300e-01 -4.73606318e-01 -3.93242061e-01
1.75893891e+00 1.15453675e-02 6.71505809e-01 -1.84583470e-01
6.42609417e-01 -6.17328465e-01 3.65221977e-01 1.17087817e+00
-9.48370576e-01 2.09482074e-01 -6.62621558e-01 -4.00185347e-01
-5.89120425e-02 -1.28300202e+00 6.75448328e-02 1.02721465e+00
1.07437298e-01 7.25483820e-02 -6.73809707e-01 -3.59235078e-01
-4.57387745e-01 5.29160619e-01 -6.66830182e-01 -2.71145076e-01
-5.73172510e-01 -5.91818988e-01 1.24403012e+00 8.81752133e-01
1.45034981e+00 -1.06224751e+00 -1.24560285e+00 3.08602899e-01
-6.11984134e-01 -1.19476008e+00 3.77243720e-02 6.18995786e-01
-6.14274442e-01 -8.77224207e-01 -6.78904533e-01 -5.48525214e-01
2.77397394e-01 5.48087895e-01 7.08884478e-01 -2.62499660e-01
-3.52670938e-01 7.40474999e-01 -1.75336242e-01 -7.70894706e-01
2.35528126e-01 -8.95501077e-02 3.04259449e-01 5.34963496e-02
7.80726731e-01 -7.54228294e-01 -8.53343308e-01 2.32877612e-01
-1.61257207e-01 -4.64858621e-01 4.76159483e-01 4.54253554e-01
8.41545314e-02 9.12935194e-03 7.64013648e-01 -4.13254619e-01
5.97605109e-01 -7.10494101e-01 1.31075159e-01 2.42957130e-01
-5.15125215e-01 -3.90509456e-01 4.36740339e-01 -3.06601673e-01
-1.31721699e+00 2.13762317e-02 -3.02432686e-01 4.90146056e-02
-6.33405507e-01 2.68942237e-01 -3.27016234e-01 -2.89585382e-01
8.04456890e-01 2.75944293e-01 -3.61890644e-01 -2.02620372e-01
-4.18146312e-01 1.15586317e+00 6.02549791e-01 -1.23966493e-01
3.13764751e-01 5.26772738e-01 -1.35188207e-01 -1.22728229e+00
-7.18099892e-01 -7.93726504e-01 -4.91089165e-01 -5.87727249e-01
1.04386914e+00 -1.06302464e+00 -9.53282535e-01 6.67018414e-01
-8.45821142e-01 -3.35412979e-01 2.77024060e-01 6.57369375e-01
-2.39419311e-01 2.43740723e-01 -3.85050386e-01 -1.35975909e+00
-7.05019951e-01 -4.76076007e-01 6.15722001e-01 4.38540727e-01
-7.16815591e-01 -1.05222619e+00 2.34597370e-01 8.06213081e-01
9.93304372e-01 6.85953438e-01 -3.54103372e-02 -3.43253374e-01
-5.73444180e-02 -6.47970259e-01 2.32746050e-01 1.10948838e-01
5.97590089e-01 -8.03838313e-01 -1.66940665e+00 -1.62877053e-01
3.02647948e-01 1.63966641e-01 -4.76463474e-02 5.25513291e-01
1.15704477e+00 -4.87440079e-01 -5.81417918e-01 5.95534861e-01
1.28347087e+00 8.15162241e-01 1.09219420e+00 4.09277022e-01
5.41363478e-01 -1.18357375e-01 4.61163700e-01 5.62965214e-01
2.43205652e-01 3.80234718e-01 1.49998486e-01 -2.14473426e-01
1.50629148e-01 -6.82952330e-02 3.21081430e-01 5.19011676e-01
-5.75604856e-01 -2.43693218e-01 -6.76307797e-01 1.66555479e-01
-1.77519369e+00 -1.23343277e+00 7.42571577e-02 2.47882080e+00
1.28714055e-01 1.37472495e-01 -1.13808340e-03 7.26072252e-01
4.54164565e-01 1.23712048e-01 -3.43416035e-01 -4.11453098e-01
3.35538268e-01 6.32962227e-01 9.39519882e-01 1.12051755e-01
-1.30028296e+00 1.59088388e-01 6.24190807e+00 6.46786764e-02
-1.13481820e+00 4.36679780e-01 3.51844281e-01 -6.70903847e-02
5.14647365e-01 -8.81543219e-01 -2.24216715e-01 5.61788201e-01
1.49649668e+00 4.95254368e-01 3.56449932e-01 1.03701448e+00
5.17000079e-01 -5.42693973e-01 -7.70640850e-01 1.45815635e+00
9.39358324e-02 -1.00627959e+00 -8.82611752e-01 2.27655750e-02
3.41898620e-01 -2.96050422e-02 -1.76665083e-01 3.91471654e-01
-2.65169024e-01 -9.11358595e-01 3.26695651e-01 8.08218539e-01
6.11225367e-01 -5.97052157e-01 1.15108204e+00 3.09792131e-01
-1.61403155e+00 -4.12784368e-01 2.23855302e-01 -5.74973702e-01
-3.38407047e-02 2.11904496e-01 -7.12454617e-01 5.91125011e-01
1.12286353e+00 1.83302239e-01 -4.00527745e-01 8.03596616e-01
1.86696455e-01 6.61298692e-01 -2.71114290e-01 -5.13729751e-01
-4.41188924e-02 2.91779310e-01 9.95856225e-02 1.30960763e+00
4.92117733e-01 9.01410356e-02 -1.63407810e-02 2.83354312e-01
4.18330997e-01 -5.08528829e-01 -1.03958440e+00 2.87067890e-01
5.82361162e-01 1.00522411e+00 -5.09458542e-01 -6.45903274e-02
-5.55263162e-01 1.12717247e+00 -4.52647716e-01 5.89918137e-01
-9.59344268e-01 -6.02115750e-01 5.52808940e-01 5.02541289e-02
-1.85023129e-01 -6.06523573e-01 -5.19999862e-01 -6.63882554e-01
1.52856871e-01 -3.38178575e-01 2.69264698e-01 -6.40727758e-01
-5.75953424e-01 4.03051645e-01 -9.57242101e-02 -1.33385634e+00
-2.50416815e-01 -3.15472782e-01 -5.53848684e-01 6.85662270e-01
-1.03553355e+00 -1.01386857e+00 -1.09771740e+00 9.99065280e-01
2.87790269e-01 1.44905284e-01 1.45396626e+00 8.35536242e-01
-6.59393907e-01 9.49669600e-01 7.94554949e-02 3.62871349e-01
2.41817698e-01 -9.65031803e-01 2.69340109e-02 8.66156042e-01
-1.09407745e-01 9.86671746e-01 5.61494350e-01 -3.80618393e-01
-1.38226175e+00 -1.18613696e+00 8.26432824e-01 -3.08987498e-01
7.38850757e-02 -4.43991512e-01 -3.23329568e-01 6.74692392e-01
-2.83489143e-03 1.53582469e-01 1.29625452e+00 4.89805155e-02
3.60755622e-01 -3.29416424e-01 -1.68847120e+00 4.75419521e-01
1.18116140e+00 -6.07093513e-01 -2.97417045e-01 -1.47008821e-01
1.14374682e-01 -1.02911375e-01 -9.11019981e-01 1.72777191e-01
1.18790364e+00 -9.45398390e-01 1.07278335e+00 9.68428403e-02
-6.02802396e-01 -1.73920453e-01 -3.44220757e-01 -8.64231825e-01
-4.58542258e-01 -5.49188018e-01 -7.58050740e-01 8.40882361e-01
8.52387324e-02 -9.95607078e-01 8.79655898e-01 8.85782599e-01
-1.35728016e-01 -5.25392592e-01 -1.17676556e+00 -7.71615386e-01
-9.98430312e-01 -7.22218573e-01 4.45335031e-01 7.31520414e-01
2.56098598e-01 -1.30237952e-01 -5.97496212e-01 2.69281745e-01
3.95258993e-01 -8.92890453e-01 6.81655943e-01 -1.03600454e+00
-1.82055026e-01 1.01642258e-01 -6.52525187e-01 -6.69075966e-01
-5.79289436e-01 -2.33397663e-01 8.24586973e-02 -1.71528685e+00
-4.36631054e-01 -2.40306854e-01 -9.94542301e-01 4.93914962e-01
2.06770197e-01 2.79100388e-01 -3.88466597e-01 -3.26224864e-01
-6.11265063e-01 1.63754255e-01 6.51713014e-01 -1.14986569e-01
-3.93295437e-01 5.78847229e-01 -5.46155751e-01 7.24805355e-01
1.10287499e+00 -1.36567324e-01 -5.65862715e-01 1.24827817e-01
-4.80402261e-02 2.97035813e-01 3.56340259e-01 -2.08601356e+00
4.37439114e-01 5.93696274e-02 1.02853930e+00 -2.27796882e-01
6.93085849e-01 -1.19265664e+00 4.76743668e-01 8.11430275e-01
1.85805280e-02 -5.81322946e-02 1.64656624e-01 5.72743595e-01
2.77915120e-01 1.77324429e-01 4.26220685e-01 -3.63543779e-01
-6.48439884e-01 1.08480215e-01 -9.77115989e-01 -5.98480701e-01
9.48342383e-01 -8.03177357e-01 -1.03745423e-01 -4.93139893e-01
-6.67506754e-01 -4.01191264e-01 -2.99810201e-01 4.32576478e-01
5.51019430e-01 -1.23166895e+00 2.67496437e-01 2.77647138e-01
1.49689421e-01 -3.88311893e-01 5.13939023e-01 8.43374491e-01
-3.84558856e-01 7.00269461e-01 -3.61452997e-01 -3.50999922e-01
-1.41918075e+00 6.43927380e-02 6.22092426e-01 -1.51753590e-01
-4.77737337e-01 5.30866206e-01 -5.58609247e-01 -6.76924363e-02
7.51625478e-01 -4.43347335e-01 -3.57492387e-01 -3.75371873e-01
9.07568991e-01 9.49253023e-01 3.54319394e-01 -3.21760803e-01
-6.80604041e-01 1.54207826e-01 5.26603281e-01 -8.82191435e-02
7.18186140e-01 -3.17205936e-01 6.43074453e-01 7.71434367e-01
1.17683196e+00 -1.34005427e-01 -8.60367596e-01 5.56709385e-03
1.35675594e-01 -3.58679742e-01 1.62936568e-01 -1.12042689e+00
-6.66434646e-01 5.85511208e-01 1.70650291e+00 2.64574319e-01
1.11318731e+00 -5.38445532e-01 1.06796825e+00 8.28323841e-01
1.09441102e+00 -1.17407155e+00 -4.90825437e-02 1.95795909e-01
3.21993589e-01 -1.04009998e+00 -3.07014793e-01 1.32312924e-01
-2.47575730e-01 9.10633028e-01 4.28278029e-01 2.42598444e-01
1.04932833e+00 3.20781052e-01 1.86256826e-01 -1.05134100e-01
2.33619679e-02 -8.97851065e-02 -1.41365528e-01 1.15797949e+00
5.82038403e-01 4.20534402e-01 1.01100966e-01 8.05398226e-01
-1.89632818e-01 4.17776316e-01 2.36263126e-01 1.63817215e+00
-4.14370537e-01 -3.72196406e-01 -4.31012750e-01 7.07148194e-01
-8.41970682e-01 3.79528731e-01 1.16455823e-01 3.48724872e-01
5.02827168e-01 1.60295248e+00 -2.44384170e-01 -6.71353519e-01
3.60714585e-01 1.74211293e-01 4.28257972e-01 -1.33115456e-01
-7.93160796e-01 -5.21066487e-01 3.37703615e-01 -8.25959861e-01
-6.17227316e-01 -4.60253328e-01 -1.02788377e+00 -1.55658916e-01
-1.63435832e-01 -9.62360427e-02 1.04850769e+00 1.22783351e+00
3.56282294e-01 1.10048389e+00 5.03571630e-01 -7.78934002e-01
1.00756958e-01 -1.12017918e+00 -5.88101983e-01 -6.32194132e-02
4.34708148e-01 -7.27863431e-01 -3.67092788e-01 -1.92255795e-01] | [6.951259136199951, 0.6648915410041809] |
4fdf690f-5f7a-40bc-8f1e-7b58ff804c81 | partial-counterfactual-identification-of | 2306.01424 | null | https://arxiv.org/abs/2306.01424v1 | https://arxiv.org/pdf/2306.01424v1.pdf | Partial Counterfactual Identification of Continuous Outcomes with a Curvature Sensitivity Model | Counterfactual inference aims to answer retrospective ''what if'' questions and thus belongs to the most fine-grained type of inference in Pearl's causality ladder. Existing methods for counterfactual inference with continuous outcomes aim at point identification and thus make strong and unnatural assumptions about the underlying structural causal model. In this paper, we relax these assumptions and aim at partial counterfactual identification of continuous outcomes, i.e., when the counterfactual query resides in an ignorance interval with informative bounds. We prove that, in general, the ignorance interval of the counterfactual queries has non-informative bounds, already when functions of structural causal models are continuously differentiable. As a remedy, we propose a novel sensitivity model called Curvature Sensitivity Model. This allows us to obtain informative bounds by bounding the curvature of level sets of the functions. We further show that existing point counterfactual identification methods are special cases of our Curvature Sensitivity Model when the bound of the curvature is set to zero. We then propose an implementation of our Curvature Sensitivity Model in the form of a novel deep generative model, which we call Augmented Pseudo-Invertible Decoder. Our implementation employs (i) residual normalizing flows with (ii) variational augmentations. We empirically demonstrate the effectiveness of our Augmented Pseudo-Invertible Decoder. To the best of our knowledge, ours is the first partial identification model for Markovian structural causal models with continuous outcomes. | ['Stefan Feuerriegel', 'Dennis Frauen', 'Valentyn Melnychuk'] | 2023-06-02 | null | null | null | null | ['counterfactual-inference'] | ['miscellaneous'] | [ 1.61841869e-01 3.33429873e-01 -5.91531694e-01 -5.66274188e-02
-7.17337668e-01 -7.10366249e-01 8.41082633e-01 -1.51773870e-01
-2.11279720e-01 1.18043971e+00 6.01726711e-01 -8.28959227e-01
-3.95307094e-01 -9.22110438e-01 -1.32975137e+00 -5.94547212e-01
-2.55379230e-01 5.73123693e-01 -2.77217507e-01 -7.60758147e-02
2.70514488e-01 2.09579661e-01 -1.13580799e+00 -7.14892298e-02
9.46677089e-01 5.29402852e-01 -1.80140018e-01 6.03245616e-01
1.75357267e-01 7.22748399e-01 -3.64989281e-01 -7.06672013e-01
2.96368122e-01 -3.61575484e-01 -8.78800869e-01 -5.04079461e-01
3.31536472e-01 -6.18068218e-01 -2.79642224e-01 1.18237889e+00
2.50648499e-01 1.33337369e-02 7.61091650e-01 -1.38851953e+00
-7.31079280e-01 1.21510017e+00 -6.44585907e-01 2.38320619e-01
1.30682334e-01 9.89613235e-02 1.09703672e+00 -5.20131111e-01
4.63770986e-01 1.38933003e+00 7.24277496e-01 5.71697950e-01
-1.56371450e+00 -7.13675439e-01 2.30620071e-01 4.80423644e-02
-1.14389372e+00 -4.03249711e-01 9.29211617e-01 -4.56983626e-01
4.62248057e-01 4.13920254e-01 4.33114737e-01 1.47248352e+00
8.53460729e-02 7.08849132e-01 1.15546823e+00 -2.39107072e-01
4.27252233e-01 -2.24253833e-01 5.84352687e-02 4.34936285e-01
5.34984529e-01 6.26310706e-01 -1.74528390e-01 -4.53051627e-01
9.55840528e-01 -3.07251438e-02 -4.81885701e-01 -5.15291095e-01
-1.29053319e+00 1.18191421e+00 2.93316990e-01 1.11202680e-01
-4.28649187e-01 6.77637517e-01 2.03219801e-01 2.02732772e-01
3.02220821e-01 3.51407856e-01 -4.80145574e-01 -8.08534995e-02
-7.28695333e-01 4.62738156e-01 1.01138496e+00 7.12385952e-01
3.90723586e-01 -8.25032294e-02 -3.96720290e-01 3.50692004e-01
2.80215800e-01 5.80440938e-01 1.73122361e-01 -1.20222616e+00
5.51522791e-01 1.43895328e-01 5.90687752e-01 -5.88265002e-01
-2.55059600e-01 -5.10549784e-01 -7.48967052e-01 -7.02729225e-02
7.93245018e-01 -4.09771800e-01 -6.11971319e-01 2.37375689e+00
2.94828832e-01 5.74365795e-01 -2.24354696e-02 9.28339660e-01
8.78575817e-02 4.20733184e-01 1.96726829e-01 -6.07326806e-01
1.12923431e+00 -8.47941041e-02 -6.73848331e-01 2.72050321e-01
5.56473732e-01 -3.97960246e-01 1.16079378e+00 -6.85891733e-02
-9.08976138e-01 4.67181094e-02 -8.12495828e-01 7.41052777e-02
6.26687109e-02 -2.53970772e-01 8.65592122e-01 9.45363760e-01
-7.69999087e-01 5.35994768e-01 -7.24249840e-01 2.95657992e-01
3.87523115e-01 2.25104704e-01 9.64742005e-02 2.67550528e-01
-1.49599278e+00 6.68477714e-01 1.64721221e-01 -4.97300066e-02
-1.07435572e+00 -1.16309440e+00 -6.84784174e-01 1.96691960e-01
8.26250315e-01 -1.13366997e+00 1.40231216e+00 -9.60553586e-01
-1.23446906e+00 5.31722069e-01 -1.74111456e-01 -8.15558851e-01
9.77129698e-01 -1.34826571e-01 -1.92412794e-01 -1.89388782e-01
3.64333987e-01 1.88467294e-01 7.60275126e-01 -1.28405523e+00
-3.16859603e-01 -5.34766495e-01 5.00659704e-01 1.02805175e-01
2.81829238e-01 -2.97469914e-01 1.25590146e-01 -7.15485752e-01
-3.58174920e-01 -8.06951106e-01 -1.10692278e-01 -2.73276001e-01
-7.35130012e-01 -1.50595665e-01 4.61468428e-01 -4.44593370e-01
1.30986643e+00 -1.91095304e+00 -1.50159091e-01 1.09632291e-01
1.72534019e-01 -2.02811733e-01 2.10313380e-01 1.27437338e-01
-3.42891812e-01 5.13209462e-01 -6.54814959e-01 -1.38393819e-01
3.84721339e-01 6.47042990e-02 -7.38164246e-01 7.03493357e-01
-2.14429244e-01 1.00487542e+00 -9.68132019e-01 -2.12984219e-01
2.97073424e-02 2.50435680e-01 -7.62960315e-01 -1.19120516e-01
-5.46358347e-01 3.22593689e-01 -3.41287285e-01 1.50410205e-01
6.96915746e-01 -8.81080255e-02 2.81478316e-01 1.47797959e-03
-3.66730303e-01 5.49994707e-01 -1.26034355e+00 1.38637853e+00
-4.98360753e-01 2.21857727e-01 -1.70175314e-01 -1.00889218e+00
2.18559876e-01 4.43804950e-01 2.64180809e-01 -3.53333533e-01
1.09726995e-01 1.68614641e-01 9.43006501e-02 -2.46512353e-01
3.05264056e-01 -8.30048978e-01 -3.39053810e-01 4.37076718e-01
-3.03681999e-01 3.13425839e-01 -5.07224277e-02 1.47117898e-01
7.29262054e-01 1.60363629e-01 4.79582578e-01 -3.91409755e-01
1.08179241e-01 -2.24682435e-01 6.82390094e-01 1.12946272e+00
-1.06777072e-01 4.60090697e-01 1.02916002e+00 -1.59246743e-01
-1.09736419e+00 -1.48764277e+00 -2.11862057e-01 6.26500249e-01
-1.24136843e-01 2.41267711e-01 -6.46358907e-01 -8.92357111e-01
2.15603665e-01 1.36959696e+00 -9.33102071e-01 -8.29562768e-02
-5.71841359e-01 -9.31384802e-01 5.72901487e-01 4.81130749e-01
6.02702141e-01 -6.25811040e-01 -4.05894190e-01 6.24693222e-02
-4.25433129e-01 -5.03495991e-01 -7.82704592e-01 -3.24858248e-01
-8.76308441e-01 -1.33989608e+00 -5.61915815e-01 8.75253510e-03
2.24698246e-01 -1.66269511e-01 9.14305389e-01 -4.64267612e-01
4.01877254e-01 3.08351874e-01 2.79917836e-01 -4.94195402e-01
-3.07467580e-01 -3.45342010e-01 -5.24327345e-03 1.83746159e-01
5.80132939e-02 -6.32876813e-01 -8.00777733e-01 -3.38604599e-02
-8.89747500e-01 5.02147861e-02 3.55697095e-01 9.06688452e-01
4.89199311e-01 -6.23409338e-02 7.83100665e-01 -1.11793852e+00
7.05125809e-01 -7.74330854e-01 -9.61949885e-01 1.97942674e-01
-4.97545123e-01 5.48846126e-01 7.51091301e-01 -4.27958965e-01
-1.53400564e+00 -3.00510317e-01 -8.94945338e-02 -4.49594706e-01
-1.00097008e-01 5.27996540e-01 -4.30756688e-01 6.01952672e-01
5.24411082e-01 -1.33494407e-01 -2.99917817e-01 -4.88115668e-01
6.80526793e-01 3.09234977e-01 6.79960132e-01 -6.90630674e-01
5.23719490e-01 9.33308065e-01 2.08495632e-01 -4.14799184e-01
-8.97033572e-01 -1.41218174e-02 -1.71597019e-01 1.17144346e-01
5.57861447e-01 -5.42514622e-01 -1.38820016e+00 1.14322983e-01
-1.16525245e+00 -3.94296855e-01 -5.63839376e-01 6.96600318e-01
-8.89603734e-01 2.05816880e-01 -3.76036644e-01 -1.21268845e+00
1.20498044e-02 -8.40310454e-01 7.87187874e-01 -2.22381279e-01
-2.63010561e-01 -1.30875754e+00 2.27438763e-01 -7.05706561e-03
5.39752990e-02 4.45091784e-01 8.83039713e-01 -3.65405649e-01
-6.06413305e-01 1.22668833e-01 -1.14910737e-01 -9.28020477e-02
-1.51514351e-01 -2.00572461e-01 -7.61874914e-01 8.73685926e-02
3.69213939e-01 3.50683779e-01 1.06997657e+00 9.30650294e-01
1.08390093e+00 -1.08605242e+00 -4.58379209e-01 5.84963799e-01
1.47179651e+00 2.05563843e-01 6.56379282e-01 1.71325371e-01
6.15114212e-01 4.46708560e-01 2.27598727e-01 4.94987071e-01
5.99929273e-01 6.12756014e-01 4.75251526e-01 2.55101264e-01
2.88033962e-01 -7.64432609e-01 2.86509454e-01 -4.84180124e-03
-1.71111912e-01 -5.92825599e-02 -6.17350459e-01 9.07301247e-01
-1.84259367e+00 -1.43994486e+00 -3.18451226e-01 2.56297588e+00
1.07655394e+00 1.09122708e-01 2.60651946e-01 -9.83457640e-02
7.39000201e-01 -2.52413582e-02 -6.49957120e-01 -5.26212156e-01
1.06286239e-02 8.32642391e-02 8.96324396e-01 8.68761420e-01
-1.09991419e+00 4.11656886e-01 5.79799461e+00 6.28792226e-01
-7.18737423e-01 3.80125165e-01 5.69482386e-01 -2.08083257e-01
-8.84511650e-01 3.91287416e-01 -5.86466134e-01 9.18055236e-01
9.82924163e-01 -4.32269961e-01 3.18858504e-01 5.16253889e-01
5.27812541e-01 4.31098156e-02 -1.37673652e+00 4.97151166e-01
-4.57583427e-01 -1.54748201e+00 4.64577116e-02 2.14664251e-01
8.09138536e-01 -3.37504268e-01 2.93289304e-01 6.85927048e-02
9.15903091e-01 -9.49431002e-01 9.06285584e-01 7.78371334e-01
8.72088134e-01 -8.95785749e-01 5.34190893e-01 4.71656650e-01
-6.82807505e-01 -3.14089000e-01 7.93559402e-02 -2.41439134e-01
3.62566829e-01 7.01067567e-01 -6.60920620e-01 5.08666694e-01
1.84076965e-01 2.29101241e-01 2.66244084e-01 9.36377227e-01
-3.76744539e-01 9.58121598e-01 -4.58583206e-01 6.58289194e-02
1.05310135e-01 -9.28525627e-02 8.16677332e-01 1.00268793e+00
3.27987403e-01 1.38984263e-01 -4.67847705e-01 1.39483595e+00
-3.40678632e-01 -3.15827787e-01 -7.94072270e-01 2.65844971e-01
4.74918962e-01 2.41213486e-01 -2.22697675e-01 -2.49770269e-01
-4.03263241e-01 5.72490454e-01 1.73757538e-01 5.16750276e-01
-1.08184767e+00 6.04160614e-02 7.96308398e-01 4.78951074e-02
1.64514199e-01 1.40941441e-01 -6.67532682e-01 -1.36773086e+00
-7.98767433e-02 -5.97872734e-01 7.63995826e-01 -3.52524132e-01
-1.24837124e+00 -5.33205867e-01 3.14294219e-01 -6.76091433e-01
-4.38896924e-01 -2.29579076e-01 -6.42903209e-01 9.10129309e-01
-1.39590228e+00 -9.02065158e-01 6.05149031e-01 5.81009746e-01
6.44592941e-02 4.73380089e-01 5.72762012e-01 -8.03425312e-02
-4.47518557e-01 4.71906662e-01 2.85181165e-01 6.57529831e-02
1.94431826e-01 -1.55054903e+00 3.00042659e-01 9.28398013e-01
-2.50745043e-02 9.79721308e-01 9.90373850e-01 -8.06272984e-01
-1.28155196e+00 -1.04653406e+00 9.37204480e-01 -4.90762234e-01
8.93664479e-01 -2.48358428e-01 -7.29771376e-01 1.09730792e+00
-7.02992529e-02 -5.32048106e-01 3.91740054e-01 3.52416575e-01
-3.86916220e-01 -1.32228434e-02 -1.14328790e+00 9.67182457e-01
9.55271304e-01 -4.42487180e-01 -8.93904924e-01 3.08274597e-01
9.67246354e-01 -1.95297301e-01 -6.03750527e-01 3.60698402e-01
5.89291632e-01 -7.91868865e-01 1.12952721e+00 -6.19840741e-01
5.57956934e-01 -2.46477127e-01 -5.94959110e-02 -1.18736029e+00
-9.38212201e-02 -8.62645209e-01 -3.46468240e-01 1.03644419e+00
3.77746254e-01 -9.46186185e-01 4.91874486e-01 5.85621297e-01
1.53494909e-01 -4.56779420e-01 -1.21134293e+00 -7.16646910e-01
4.98314261e-01 -7.83817887e-01 9.45718229e-01 1.12509835e+00
3.78506333e-02 2.80699760e-01 -4.50781554e-01 3.44412595e-01
1.05557430e+00 4.51835692e-01 4.13752079e-01 -9.77682114e-01
-5.45701683e-01 -5.85482895e-01 1.64915025e-01 -9.81561005e-01
3.06084841e-01 -6.95679784e-01 -3.35456580e-01 -1.16544521e+00
4.70076233e-01 -4.41616386e-01 -7.35222474e-02 1.59779221e-01
-2.36179620e-01 -9.05592591e-02 -3.32185254e-02 1.91429481e-02
-9.31567103e-02 5.31776428e-01 1.20691323e+00 1.54808545e-02
-2.50060499e-01 3.07086587e-01 -1.02522933e+00 8.06264341e-01
9.45887029e-01 -4.53066081e-01 -5.00411749e-01 -1.10025987e-01
4.68684733e-01 6.78167760e-01 9.87113297e-01 -1.22089863e-01
8.34813341e-02 -4.73479331e-01 5.31899966e-02 -5.01235962e-01
9.68174357e-03 -4.15174544e-01 4.72426534e-01 4.60716099e-01
-5.19685626e-01 -3.50469828e-01 5.43425567e-02 8.39885116e-01
2.60424733e-01 -1.68401599e-01 5.67425966e-01 -1.47451043e-01
-3.02148581e-01 1.43467352e-01 -1.01828694e-01 2.07431495e-01
7.76282310e-01 1.36444300e-01 -3.80966783e-01 -5.48257589e-01
-7.53815353e-01 2.30661094e-01 3.55611801e-01 5.40885441e-02
4.89200711e-01 -1.29749835e+00 -7.86188662e-01 -2.29646266e-01
-3.39621514e-01 -3.49782407e-01 4.00449067e-01 1.12327576e+00
8.37971866e-02 6.67551935e-01 3.29879969e-01 -8.63141716e-02
-6.28549159e-01 9.66236413e-01 5.71969569e-01 -4.01018083e-01
-3.57312411e-01 4.78426307e-01 7.37731040e-01 -2.33257115e-01
-1.93004340e-01 -3.55895162e-01 -3.33875827e-02 -9.10282508e-02
2.96826303e-01 6.23887897e-01 -3.55350643e-01 -3.84360194e-01
-2.50942171e-01 -1.81607287e-02 1.30950525e-01 -5.10585785e-01
1.05411530e+00 -2.45249957e-01 -1.43438295e-01 6.27394021e-01
1.17939723e+00 2.29597032e-01 -1.40432930e+00 9.93396342e-02
-9.48616676e-03 -3.72193754e-01 -1.58307612e-01 -9.42702591e-01
-6.77303433e-01 7.27421761e-01 2.34215796e-01 2.96723813e-01
1.00662756e+00 1.31701574e-01 3.44925225e-01 -9.22328681e-02
2.35177755e-01 -7.61088669e-01 -6.96038127e-01 2.16449827e-01
8.90032470e-01 -9.53907669e-01 -3.46901417e-01 -1.98578551e-01
-2.73298323e-01 5.76541841e-01 -1.93411484e-02 -7.27083981e-02
5.34130633e-01 2.63963908e-01 -4.58210588e-01 -9.52227786e-02
-7.35559642e-01 -1.17268592e-01 1.86988696e-01 4.27213490e-01
4.25074250e-01 7.23966658e-01 -7.35026658e-01 7.73456275e-01
-5.22720993e-01 1.19310126e-01 6.35961473e-01 3.21278095e-01
1.70828328e-02 -7.26139903e-01 -5.99448204e-01 4.81332898e-01
-8.98365438e-01 -4.32785422e-01 -1.14446968e-01 1.00523520e+00
-1.57637060e-01 8.63746047e-01 1.29352719e-01 2.83823848e-01
3.20461512e-01 8.90185311e-03 5.28680265e-01 -2.17219442e-01
-8.32308531e-02 -7.76759386e-02 4.60791700e-02 -4.93919522e-01
-3.08483571e-01 -9.03897405e-01 -1.09217727e+00 -7.13663340e-01
-2.58000463e-01 2.14710742e-01 3.65817279e-01 1.12653697e+00
8.92086998e-02 3.94721091e-01 5.11829138e-01 -2.38271445e-01
-1.02802408e+00 -8.61105978e-01 -5.14818907e-01 3.89211655e-01
6.97021306e-01 -7.24108517e-01 -4.89797175e-01 -9.31430832e-02] | [8.110467910766602, 5.4063920974731445] |
682ea29a-4707-4c39-816a-9fdaac3bfb14 | unsupervised-action-segmentation-with-self | 2105.14158 | null | https://arxiv.org/abs/2105.14158v3 | https://arxiv.org/pdf/2105.14158v3.pdf | SSCAP: Self-supervised Co-occurrence Action Parsing for Unsupervised Temporal Action Segmentation | Temporal action segmentation is a task to classify each frame in the video with an action label. However, it is quite expensive to annotate every frame in a large corpus of videos to construct a comprehensive supervised training dataset. Thus in this work we propose an unsupervised method, namely SSCAP, that operates on a corpus of unlabeled videos and predicts a likely set of temporal segments across the videos. SSCAP leverages Self-Supervised learning to extract distinguishable features and then applies a novel Co-occurrence Action Parsing algorithm to not only capture the correlation among sub-actions underlying the structure of activities, but also estimate the temporal path of the sub-actions in an accurate and general way. We evaluate on both classic datasets (Breakfast, 50Salads) and the emerging fine-grained action dataset (FineGym) with more complex activity structures and similar sub-actions. Results show that SSCAP achieves state-of-the-art performance on all datasets and can even outperform some weakly-supervised approaches, demonstrating its effectiveness and generalizability. | ['Charless Fowlkes', 'Joseph Tighe', 'Yuanjun Xiong', 'Chunhui Liu', 'Xinyu Li', 'Hao Chen', 'Zhe Wang'] | 2021-05-29 | null | null | null | null | ['action-parsing'] | ['natural-language-processing'] | [ 4.98775989e-01 5.64374216e-03 -7.46544480e-01 -5.05809069e-01
-8.22415709e-01 -7.99922109e-01 6.19198859e-01 -1.31026268e-01
-1.94861740e-01 5.24098694e-01 6.17404401e-01 1.00945912e-01
-6.98394999e-02 -2.77977288e-01 -6.89914525e-01 -4.71892506e-01
-4.89356220e-01 2.33407721e-01 7.07980573e-01 2.00957999e-01
7.77376071e-02 8.48532245e-02 -1.54525709e+00 8.00227284e-01
4.48312491e-01 1.07032073e+00 5.22932783e-02 6.82044208e-01
8.34380314e-02 1.50695395e+00 -2.47900158e-01 -1.76465049e-01
2.62646794e-01 -9.19789493e-01 -1.28283978e+00 6.84842944e-01
5.01424134e-01 -3.41794103e-01 -3.74878079e-01 6.29553258e-01
-2.51557469e-01 2.38198936e-01 5.70068240e-01 -1.35132241e+00
1.07376408e-02 7.56336093e-01 -5.82288146e-01 5.24842560e-01
6.34200275e-01 2.45989829e-01 1.33692789e+00 -2.36251161e-01
9.65870321e-01 9.73012388e-01 6.88994467e-01 3.38041246e-01
-9.09109473e-01 -3.61697853e-01 5.04494369e-01 2.98741162e-01
-9.48608100e-01 -3.67494524e-01 5.18111587e-01 -6.50569260e-01
9.16337609e-01 4.13142592e-02 8.20054293e-01 1.36093485e+00
-7.55798742e-02 1.21832561e+00 1.04901576e+00 -1.32457316e-01
2.92499989e-01 -6.32562459e-01 2.19591498e-01 7.94466138e-01
-3.00266713e-01 -1.31337434e-01 -7.61210084e-01 1.02402255e-01
5.38270354e-01 1.29827365e-01 -5.92981465e-02 -3.68342340e-01
-1.37336636e+00 4.45425808e-01 -3.81954946e-02 3.32444638e-01
-4.12194252e-01 2.37569377e-01 6.03177249e-01 5.69688715e-02
2.96838492e-01 2.94738948e-01 -7.39060760e-01 -8.50520968e-01
-1.02821910e+00 9.74977463e-02 7.96319783e-01 8.57226968e-01
6.86220407e-01 -3.90488923e-01 -3.21160555e-01 4.64456320e-01
-2.23840084e-02 -5.36310263e-02 5.04229128e-01 -1.53743541e+00
4.92238939e-01 7.97495723e-01 8.12223330e-02 -6.36092424e-01
-4.33139354e-01 1.20591834e-01 -2.91855723e-01 -1.02340713e-01
6.60663962e-01 -1.65105999e-01 -9.22365904e-01 1.65136266e+00
3.98542881e-01 5.91760576e-01 -1.00643411e-01 7.68028438e-01
4.30915952e-01 5.89292824e-01 4.28988010e-01 -3.27678919e-01
1.13540006e+00 -1.44823718e+00 -5.81619918e-01 -4.04693604e-01
7.49803305e-01 -2.83834249e-01 9.25111294e-01 3.97922218e-01
-8.31829190e-01 -6.44016027e-01 -7.97110915e-01 5.81322722e-02
-1.04084492e-01 1.26090452e-01 8.11446905e-01 3.24267000e-01
-5.67233980e-01 8.52412343e-01 -1.30270946e+00 -5.58767974e-01
8.55233669e-01 1.34639964e-01 -5.96354127e-01 -1.77334711e-01
-6.64738536e-01 3.26226681e-01 6.86137497e-01 -2.63206929e-01
-1.17867255e+00 -5.11898279e-01 -9.73234057e-01 -1.59068272e-01
8.57919037e-01 -1.66003138e-01 1.31913376e+00 -1.44934416e+00
-1.39170468e+00 7.46344745e-01 -1.94428444e-01 -7.31474221e-01
4.68190014e-01 -5.24058461e-01 -3.06314647e-01 5.16516089e-01
3.45759898e-01 8.51373017e-01 7.11056590e-01 -7.88399935e-01
-1.16437781e+00 -2.08046511e-01 3.00382197e-01 1.89118698e-01
-1.03082500e-01 1.77615479e-01 -8.37206304e-01 -6.47827387e-01
-6.66872412e-02 -1.09488940e+00 -2.11396173e-01 -3.42455417e-01
-4.04294252e-01 -3.88072103e-01 7.61256337e-01 -6.38913989e-01
1.36087561e+00 -2.30368352e+00 1.59364402e-01 -2.21189618e-01
-7.16278329e-03 -3.52968313e-02 -1.34719312e-01 5.24479210e-01
-1.52209410e-02 -1.61779270e-01 -2.84358293e-01 -2.48108536e-01
-4.34875898e-02 5.60130656e-01 -8.95170346e-02 4.62381899e-01
2.17193633e-01 9.64320660e-01 -1.21578705e+00 -6.61519468e-01
1.03445001e-01 2.65045613e-02 -5.64881682e-01 1.92180499e-01
-5.58409572e-01 6.96178555e-01 -6.76061034e-01 8.04278672e-01
-5.41486666e-02 -5.30434251e-01 5.45925796e-01 -1.78527050e-02
9.69069451e-03 2.61106074e-01 -1.03874612e+00 2.05528212e+00
1.44135550e-01 7.56164968e-01 -4.66300249e-01 -1.27321327e+00
2.94486910e-01 1.92305356e-01 1.17921722e+00 -4.78405297e-01
-9.22233835e-02 -4.14362624e-02 -1.86766371e-01 -9.38629746e-01
2.94996917e-01 1.97270513e-01 -4.10965621e-01 5.86390674e-01
3.84984314e-01 5.10566175e-01 7.35942066e-01 2.50219941e-01
1.57899964e+00 8.98918331e-01 4.01803136e-01 -8.28411728e-02
3.74363005e-01 2.59709924e-01 8.85038614e-01 7.42513835e-01
-4.74296033e-01 6.17899537e-01 7.81396031e-01 -6.31153882e-01
-6.48701191e-01 -7.90441394e-01 3.35957736e-01 1.41087234e+00
2.23732546e-01 -8.92638862e-01 -9.74285245e-01 -1.27286112e+00
-2.90905148e-01 2.06327856e-01 -8.01176727e-01 1.03506424e-01
-6.97743237e-01 -4.90648687e-01 4.99594867e-01 9.58372355e-01
6.66781843e-01 -1.28069937e+00 -8.46050501e-01 1.37919798e-01
-5.00361860e-01 -1.70875919e+00 -5.04864991e-01 1.84002817e-01
-7.51724780e-01 -1.57839715e+00 -1.12137407e-01 -6.85301721e-01
5.19532442e-01 8.71964395e-02 1.19471800e+00 -1.85517922e-01
-1.11584872e-01 4.88679439e-01 -7.90490866e-01 8.53075609e-02
-2.00926691e-01 4.46676351e-02 -1.74480379e-01 3.26634496e-01
5.22942424e-01 -4.18443531e-01 -5.88286698e-01 5.07924795e-01
-6.66167021e-01 1.05045140e-01 5.29892087e-01 3.43265444e-01
9.40504611e-01 2.99727947e-01 3.25169474e-01 -9.67505336e-01
-1.90376639e-01 -4.69330490e-01 -3.27113330e-01 2.81250715e-01
-1.95102304e-01 -9.99350920e-02 5.22706807e-01 -5.17103076e-01
-1.08626425e+00 6.44933224e-01 2.66614258e-01 -2.94942021e-01
-5.35328031e-01 3.45452189e-01 -2.08663613e-01 4.22760874e-01
3.48618954e-01 3.33739579e-01 -3.45903873e-01 -5.18636525e-01
3.54684919e-01 2.88339555e-01 7.85584211e-01 -4.88105148e-01
5.35055816e-01 7.92745709e-01 -3.91077921e-02 -5.80823541e-01
-1.38296843e+00 -7.84742892e-01 -1.09717381e+00 -4.29509401e-01
1.34454024e+00 -1.08751297e+00 -3.59630555e-01 5.18821180e-01
-6.24616146e-01 -8.85475874e-01 -3.92151237e-01 6.43436193e-01
-9.83238935e-01 4.62903798e-01 -7.28822231e-01 -3.99256378e-01
2.55116135e-01 -8.85390937e-01 1.18362474e+00 8.73911381e-02
-4.89531398e-01 -7.62042046e-01 1.68168008e-01 8.63002062e-01
-5.02947807e-01 4.73648876e-01 3.53323489e-01 -6.95141733e-01
-7.17765272e-01 -8.22567940e-02 9.92458835e-02 3.51200491e-01
3.60584825e-01 1.47049442e-01 -6.72837734e-01 -1.39758252e-02
-3.69736731e-01 -5.57116568e-01 8.98706019e-01 4.85297889e-01
1.22721767e+00 -1.56931028e-01 -3.60888720e-01 4.85567033e-01
1.07526195e+00 1.99558645e-01 7.69459188e-01 3.45466346e-01
7.19572783e-01 6.53681576e-01 1.21641684e+00 3.96882862e-01
3.44125271e-01 7.09714949e-01 2.06625804e-01 2.68326461e-01
-2.44524237e-02 -4.45387989e-01 6.62491441e-01 4.72769350e-01
-4.15766329e-01 -1.97979107e-01 -6.92390263e-01 5.73831975e-01
-2.17885160e+00 -1.30555511e+00 -2.14403365e-02 1.78339088e+00
7.65707910e-01 2.42468432e-01 6.19796395e-01 3.44853476e-02
5.13968349e-01 4.66240942e-01 -5.18798590e-01 1.19270399e-01
4.30396236e-02 -6.34549633e-02 4.12900895e-01 1.34103540e-02
-1.69031465e+00 1.02995288e+00 6.68778515e+00 6.76116824e-01
-6.65418565e-01 1.22897923e-01 7.57382095e-01 -2.15160683e-01
4.45521951e-01 1.30053267e-01 -6.66147053e-01 6.18866980e-01
1.01383889e+00 3.22575569e-01 2.36667827e-01 8.44866455e-01
4.04722869e-01 -3.75878364e-01 -1.27785289e+00 8.12132478e-01
6.47763684e-02 -1.33267951e+00 -2.68333375e-01 -3.58115137e-02
1.01643395e+00 2.10809261e-01 -5.75989842e-01 3.83472711e-01
3.49691868e-01 -7.80105948e-01 8.38418067e-01 4.09941345e-01
4.37683553e-01 -3.80485177e-01 4.53150839e-01 3.56797665e-01
-1.47118175e+00 -3.86560977e-01 1.98178411e-01 -2.73605376e-01
3.44198048e-01 1.23151384e-01 -2.76262790e-01 4.81052518e-01
9.83948529e-01 1.49960923e+00 -6.75737739e-01 8.15706670e-01
-3.66714329e-01 8.89512181e-01 -1.28132328e-01 4.35603380e-01
5.37914515e-01 -1.40498832e-01 1.03074826e-01 1.29644918e+00
1.07491938e-02 3.15741539e-01 7.39901543e-01 1.40359983e-01
-1.32950963e-02 -1.86569065e-01 -2.81328470e-01 -4.73393023e-01
1.67575449e-01 1.21789622e+00 -1.15584052e+00 -5.25051415e-01
-7.73994803e-01 1.05422521e+00 1.85445145e-01 2.62311757e-01
-1.16644931e+00 3.14978868e-01 7.28797793e-01 1.58917859e-01
6.87806726e-01 -9.13038179e-02 1.79893017e-01 -1.29567266e+00
1.07232690e-01 -9.79342282e-01 9.05089915e-01 -7.29700327e-01
-1.02328444e+00 4.38531399e-01 1.37571886e-01 -1.59941411e+00
-4.08199608e-01 -3.42628598e-01 -4.60100234e-01 -5.06031178e-02
-1.26381075e+00 -1.21423614e+00 -4.63814706e-01 6.85096383e-01
1.03035736e+00 8.46082792e-02 4.39397633e-01 1.71780378e-01
-7.74926484e-01 7.83910230e-02 -1.80722475e-01 5.44876635e-01
5.35022080e-01 -1.12377512e+00 2.55619168e-01 1.02831304e+00
5.99928141e-01 3.03276740e-02 3.55461299e-01 -6.81885779e-01
-1.22942352e+00 -1.27173436e+00 6.18785739e-01 -7.19690621e-01
8.52719963e-01 -1.86135992e-01 -6.61996424e-01 1.10784507e+00
7.27002323e-02 3.82526010e-01 8.21084619e-01 -9.54933092e-02
-3.96671265e-01 1.16679259e-01 -6.21443450e-01 2.80626059e-01
1.67001975e+00 -5.34966111e-01 -7.31971264e-01 4.02967840e-01
5.15136600e-01 -3.25999647e-01 -8.69210720e-01 4.02947456e-01
6.72962070e-01 -1.23896229e+00 8.05801749e-01 -9.84369040e-01
8.46592367e-01 -3.56332600e-01 -1.76010057e-01 -7.15575814e-01
-2.56213158e-01 -7.84197092e-01 -5.30012131e-01 1.21715033e+00
2.20494792e-01 -5.65311685e-02 1.04047692e+00 3.43482047e-01
-1.61952808e-01 -7.66715467e-01 -6.59116328e-01 -9.97291803e-01
-5.80032766e-01 -6.35992110e-01 2.28364915e-01 9.25491989e-01
1.85157299e-01 1.81424558e-01 -4.45976138e-01 3.99890728e-02
4.75668132e-01 3.61381412e-01 7.68170834e-01 -9.00538564e-01
-4.78196770e-01 -1.75624788e-01 -7.08801925e-01 -1.27483630e+00
5.04652023e-01 -4.51432437e-01 2.20150590e-01 -1.48095310e+00
4.28334296e-01 -1.12704702e-01 -4.03613746e-01 7.96779513e-01
-1.08953886e-01 4.81983900e-01 2.81923804e-02 4.28954870e-01
-1.54824030e+00 2.03926712e-01 1.00037920e+00 -5.63325807e-02
-3.14751595e-01 1.56317987e-02 -3.87303919e-01 1.05029297e+00
5.33123374e-01 -3.95600885e-01 -6.63107038e-01 -2.21117482e-01
-1.34242460e-01 1.21349186e-01 3.16521674e-01 -1.28795063e+00
-1.08048871e-01 -4.48817790e-01 2.52812445e-01 -5.38528800e-01
9.15170312e-02 -6.88883364e-01 2.07374662e-01 2.25342885e-01
-5.20101964e-01 -3.16938549e-01 -1.55052081e-01 9.11776841e-01
-4.56369579e-01 4.99896370e-02 4.95962948e-01 -2.03099340e-01
-1.48150432e+00 4.89567399e-01 -4.04376805e-01 3.03373247e-01
1.32223141e+00 -4.02615458e-01 -1.26291856e-01 -2.82363415e-01
-8.38401318e-01 2.38821417e-01 4.95045990e-01 6.23294830e-01
1.77079439e-01 -1.18048692e+00 -3.38169456e-01 -1.16953231e-01
2.68786848e-01 -1.70865640e-01 3.06458533e-01 1.01315153e+00
-3.15631032e-01 2.01963544e-01 -2.78342754e-01 -6.74456418e-01
-1.34169996e+00 5.42814732e-01 1.64721414e-01 -4.19792950e-01
-9.20840621e-01 6.52058005e-01 2.11205423e-01 1.19113341e-01
1.75874993e-01 -3.97368908e-01 -4.58784193e-01 1.97298959e-01
6.06204152e-01 3.04243028e-01 -3.88616830e-01 -9.14979756e-01
-4.19680208e-01 4.69586372e-01 2.53967643e-01 1.69588536e-01
1.36585402e+00 -1.91293776e-01 7.38269836e-02 4.40410048e-01
1.20477951e+00 -2.72992820e-01 -2.12925529e+00 -2.06418306e-01
3.93560231e-01 -4.74857748e-01 -3.01907688e-01 -6.32330418e-01
-1.19481957e+00 3.30885887e-01 2.12738663e-01 1.37791142e-01
1.31898022e+00 4.54738110e-01 8.96564007e-01 2.48742282e-01
4.53970850e-01 -1.23292577e+00 4.46475387e-01 4.59876686e-01
2.06500188e-01 -1.32440746e+00 6.87320903e-02 -5.87062299e-01
-9.77013111e-01 9.64010656e-01 6.33445024e-01 6.30268967e-03
4.04557943e-01 1.14867233e-01 -2.68749446e-02 -3.43919754e-01
-7.31788039e-01 -4.74003613e-01 2.53804028e-01 4.08937544e-01
1.90670595e-01 -1.12991326e-01 -1.95131585e-01 6.15878522e-01
2.59967446e-01 3.86986643e-01 3.23433697e-01 1.28359473e+00
-2.89699048e-01 -1.08384907e+00 1.50029257e-01 4.89063323e-01
-6.60444260e-01 3.21670026e-01 -3.92705292e-01 7.70141304e-01
4.02461380e-01 9.57851589e-01 1.91388220e-01 -4.24564481e-01
1.04577124e-01 1.49976328e-01 5.62846959e-01 -6.26092911e-01
-3.13308001e-01 1.45911738e-01 4.00339365e-01 -1.23120117e+00
-1.30688882e+00 -1.11545146e+00 -1.49215913e+00 2.42641032e-01
2.39039555e-01 -3.98742110e-02 -5.30601479e-02 1.35774720e+00
3.39706093e-01 4.32710558e-01 5.63712955e-01 -8.18814218e-01
-4.26309854e-02 -7.82430232e-01 -6.13252044e-01 9.77438092e-01
1.05060779e-01 -7.43205249e-01 -1.32561386e-01 8.23012054e-01] | [8.42308521270752, 0.5774403214454651] |
e134e020-1092-4577-9156-6a688c482127 | eclare-extreme-classification-with-label | 2108.00261 | null | https://arxiv.org/abs/2108.00261v1 | https://arxiv.org/pdf/2108.00261v1.pdf | ECLARE: Extreme Classification with Label Graph Correlations | Deep extreme classification (XC) seeks to train deep architectures that can tag a data point with its most relevant subset of labels from an extremely large label set. The core utility of XC comes from predicting labels that are rarely seen during training. Such rare labels hold the key to personalized recommendations that can delight and surprise a user. However, the large number of rare labels and small amount of training data per rare label offer significant statistical and computational challenges. State-of-the-art deep XC methods attempt to remedy this by incorporating textual descriptions of labels but do not adequately address the problem. This paper presents ECLARE, a scalable deep learning architecture that incorporates not only label text, but also label correlations, to offer accurate real-time predictions within a few milliseconds. Core contributions of ECLARE include a frugal architecture and scalable techniques to train deep models along with label correlation graphs at the scale of millions of labels. In particular, ECLARE offers predictions that are 2 to 14% more accurate on both publicly available benchmark datasets as well as proprietary datasets for a related products recommendation task sourced from the Bing search engine. Code for ECLARE is available at https://github.com/Extreme-classification/ECLARE. | ['Manik Varma', 'Purushottam Kar', 'Sumeet Agarwal', 'Sheshansh Agrawal', 'Noveen Sachdeva', 'Anshul Mittal'] | 2021-07-31 | null | null | null | null | ['extreme-multi-label-classification', 'product-recommendation', 'short-text-clustering'] | ['methodology', 'miscellaneous', 'natural-language-processing'] | [-2.30251282e-01 -1.37709498e-01 -5.04491270e-01 -9.82333958e-01
-8.47964227e-01 -8.31875861e-01 4.19962615e-01 3.78302634e-01
2.58130562e-02 4.01603341e-01 9.00272802e-02 -2.21685156e-01
-2.84575373e-01 -6.30113542e-01 -5.19865334e-01 -4.77123499e-01
-9.86428708e-02 7.89129615e-01 -3.50223035e-01 -1.55700579e-01
1.00758724e-01 3.50748688e-01 -1.57900929e+00 8.49332273e-01
4.48888540e-01 1.68337429e+00 -2.19700232e-01 4.43987429e-01
-1.89662039e-01 1.05543602e+00 -2.18714669e-01 -5.09298861e-01
3.47287357e-01 1.32471308e-01 -8.31252754e-01 -2.23130822e-01
9.16545630e-01 -3.12686265e-01 -1.39549524e-02 6.92332804e-01
6.14451647e-01 9.67486650e-02 8.12228858e-01 -1.51228130e+00
-8.09723675e-01 8.80604982e-01 -4.82248157e-01 5.23954928e-02
2.63260186e-01 -2.74562091e-02 1.70460057e+00 -9.10613298e-01
3.95600289e-01 7.42232442e-01 1.15893972e+00 5.32288074e-01
-1.20828879e+00 -7.69897938e-01 1.80827722e-01 1.67273730e-01
-1.41457474e+00 -1.31964356e-01 6.47038937e-01 -4.78513598e-01
1.09227371e+00 4.09199089e-01 1.74723968e-01 1.20687401e+00
-7.01093152e-02 7.16976404e-01 9.34385717e-01 -1.07262485e-01
1.84914008e-01 1.77694708e-01 5.92100143e-01 6.02397025e-01
-9.66210514e-02 -4.08579074e-02 -4.77202445e-01 -3.16879272e-01
-1.04443312e-01 5.96748412e-01 1.50612324e-01 -3.87273878e-02
-8.97648990e-01 8.79310071e-01 6.18786752e-01 1.48879871e-01
-4.20398802e-01 4.10440862e-01 7.69276738e-01 3.67836118e-01
9.22064185e-01 7.93504298e-01 -9.69932973e-01 -7.64181539e-02
-8.03569436e-01 2.97127128e-01 1.03747976e+00 1.05620635e+00
6.41831398e-01 -4.17918116e-01 -3.02783698e-01 8.78943741e-01
1.01556286e-01 2.36854449e-01 4.71686244e-01 -1.07879305e+00
1.06735796e-01 5.88620007e-01 1.20559007e-01 -1.00473845e+00
-7.98367977e-01 -1.05550778e+00 -6.06255114e-01 1.11696981e-01
2.55514681e-01 -9.30396020e-02 -8.03823709e-01 1.60900640e+00
1.59128383e-01 1.08557172e-01 -4.28184748e-01 6.84570730e-01
1.11364198e+00 5.61221898e-01 2.00893462e-01 1.62057981e-01
1.14249241e+00 -1.11688852e+00 -4.12369996e-01 -9.88731533e-02
1.37023151e+00 -6.93294644e-01 1.34231615e+00 6.01316333e-01
-7.18707740e-01 -4.62059766e-01 -8.09828341e-01 -3.12759727e-01
-7.35450208e-01 1.95622191e-01 8.62496734e-01 2.60693610e-01
-1.15518212e+00 8.85637581e-01 -4.03483689e-01 -2.10626543e-01
7.47735202e-01 5.72657168e-01 1.05183078e-02 -1.94349095e-01
-1.08330977e+00 5.69073737e-01 3.39557566e-02 -2.08941996e-01
-5.60396194e-01 -9.45215940e-01 -5.07312357e-01 4.79711413e-01
3.68909061e-01 -2.36851841e-01 1.68370533e+00 -1.02653182e+00
-9.89119351e-01 8.02314222e-01 1.07972510e-01 -4.96551484e-01
3.72208238e-01 -3.96101803e-01 -5.96560299e-01 -3.07778955e-01
1.54608950e-01 8.33146214e-01 2.83550173e-01 -1.03648019e+00
-8.52106571e-01 -1.14003308e-01 -8.04863572e-02 -1.22058010e-02
-6.22962415e-01 -1.72417551e-01 6.59073964e-02 -3.36410522e-01
-2.53144830e-01 -1.04277194e+00 -8.50898102e-02 -1.08938918e-01
-4.56637681e-01 -9.77538705e-01 7.66974509e-01 -3.20995063e-01
1.13889420e+00 -2.21096802e+00 -4.97459024e-01 8.35745875e-03
7.03415871e-01 1.93384647e-01 -2.02882662e-01 5.91079712e-01
-2.54987746e-01 8.66031200e-02 5.73325694e-01 -5.82541227e-01
4.67360914e-01 -1.33672029e-01 -6.53972149e-01 3.18260610e-01
-1.66598395e-01 1.09753144e+00 -1.03573871e+00 -1.32371947e-01
-4.81500439e-02 4.94308650e-01 -4.08661693e-01 2.08994914e-02
-7.98713028e-01 5.01797609e-02 -2.47670993e-01 8.54389668e-01
4.26633537e-01 -1.02309465e+00 -8.34966600e-02 -2.53897041e-01
1.30883634e-01 6.75286353e-01 -7.02093542e-01 1.37116349e+00
-5.60625851e-01 5.20729005e-01 -6.42262638e-01 -8.20319891e-01
8.99028957e-01 1.63531467e-01 8.66661131e-01 -6.93832397e-01
3.95601660e-01 4.02097315e-01 -4.10901785e-01 -1.51324153e-01
4.52163398e-01 7.18654618e-02 -1.76752537e-01 9.05614853e-01
-2.96801655e-03 4.77919161e-01 2.17556939e-01 2.66612440e-01
1.23165309e+00 -4.27848324e-02 2.13875771e-02 -8.58768001e-02
-3.13105173e-02 1.21617064e-01 5.03702223e-01 7.83454061e-01
2.09804084e-02 4.46230292e-01 5.19527376e-01 -1.20052397e+00
-1.04237473e+00 -6.31782115e-01 -1.37416229e-01 1.69605589e+00
-2.19829589e-01 -8.85425806e-01 -3.00616413e-01 -1.08372653e+00
2.90962815e-01 9.14080203e-01 -7.85012960e-01 -2.73207247e-01
-1.40737802e-01 -3.79686564e-01 1.26458958e-01 6.65097177e-01
2.07191348e-01 -9.17610943e-01 -1.13847964e-01 1.06747940e-01
-2.80350130e-02 -9.15088117e-01 -4.99427289e-01 5.08733988e-01
-5.13114929e-01 -8.94441903e-01 -2.93846667e-01 -7.52266228e-01
3.40743810e-01 2.26461977e-01 1.63856089e+00 1.54778168e-01
-1.92197144e-01 1.72175933e-02 -6.17355049e-01 -4.33059752e-01
-2.57917285e-01 4.64846492e-01 7.84194171e-02 -1.04462981e-01
1.08416331e+00 -6.27785802e-01 -7.96486557e-01 4.19824272e-01
-3.56386423e-01 2.56839637e-02 3.45343500e-01 7.44818628e-01
7.10452974e-01 4.78213206e-02 8.29152405e-01 -1.57444704e+00
4.46980029e-01 -1.05180621e+00 -4.98846710e-01 1.51188195e-01
-1.21367514e+00 -1.41261175e-01 1.01090443e+00 -4.28852528e-01
-5.02519429e-01 1.93356946e-01 -3.21643919e-01 -3.07985425e-01
-2.89964378e-01 4.53785419e-01 5.73515356e-01 1.12811647e-01
1.01832712e+00 -2.45830134e-01 -2.99952090e-01 -7.28334367e-01
6.51946008e-01 8.64645243e-01 2.27097660e-01 -3.44281256e-01
9.32467133e-02 3.17539722e-01 3.92465219e-02 1.82154149e-01
-1.79597628e+00 -7.48072207e-01 -4.23101783e-01 -1.78314433e-01
4.39063042e-01 -8.06925535e-01 -1.11144042e+00 1.10413142e-01
-6.62594616e-01 -5.74215770e-01 -4.36421156e-01 1.35013655e-01
-4.94700968e-01 -1.37437209e-01 -9.15651143e-01 -2.08106518e-01
-7.89500475e-01 -7.51082659e-01 1.08581841e+00 -2.30583977e-02
-5.72032332e-01 -9.20733869e-01 -6.15163222e-02 3.07711691e-01
4.92210269e-01 2.89466679e-01 9.74817693e-01 -1.24244678e+00
-3.52048546e-01 -6.35036290e-01 -3.17399472e-01 3.70480955e-01
-4.92704771e-02 -1.23491943e-01 -1.08329153e+00 -4.64873612e-01
-4.56674218e-01 -8.53782594e-01 8.64927709e-01 1.96961910e-01
1.45377171e+00 -1.92131862e-01 -4.81996775e-01 6.49944842e-01
1.31243491e+00 -1.29491627e-01 -7.40641132e-02 1.79774076e-01
8.74334514e-01 3.23139846e-01 6.36692643e-01 5.42143166e-01
4.32347000e-01 7.27235079e-01 5.50924480e-01 -1.15999885e-01
-2.26713225e-01 -1.19860776e-01 -1.57414109e-01 6.96307361e-01
4.28134233e-01 -3.75515878e-01 -1.19134939e+00 2.80969471e-01
-1.84630716e+00 -7.85092652e-01 -2.65531868e-01 1.95213962e+00
7.53842473e-01 2.51896501e-01 2.04423726e-01 -8.80240556e-03
4.84217286e-01 7.32482085e-03 -8.49808216e-01 -4.80338186e-01
2.17931315e-01 -1.04628436e-01 5.67917466e-01 1.91744998e-01
-1.26453078e+00 7.34332800e-01 5.98854113e+00 8.04666758e-01
-1.28044701e+00 2.83784926e-01 1.16396892e+00 -6.50639057e-01
-2.74986684e-01 -2.73745418e-01 -1.14659953e+00 6.79447412e-01
1.31789052e+00 4.01332863e-02 2.61364967e-01 1.39225614e+00
-2.18831748e-01 3.89868021e-01 -1.60830152e+00 9.87877071e-01
5.47941588e-03 -1.53274524e+00 -2.41484687e-01 3.53185594e-01
1.13309908e+00 6.52365565e-01 7.24053979e-01 6.05884969e-01
8.33632708e-01 -1.13436842e+00 6.29239142e-01 4.82016146e-01
1.08745122e+00 -8.11753154e-01 9.30084825e-01 3.49898428e-01
-8.31717432e-01 -6.46120071e-01 -4.76992369e-01 -1.71320796e-01
-1.19471930e-01 9.05080080e-01 -1.01440358e+00 -1.27321910e-02
8.04525077e-01 1.23829305e+00 -5.97612560e-01 8.63369942e-01
9.40366685e-02 7.85866857e-01 -3.33504528e-01 -1.52655393e-01
4.61939514e-01 3.12425196e-01 -1.86082765e-01 1.13807786e+00
4.61910516e-01 -7.43374811e-04 2.68534839e-01 7.13719249e-01
-4.98008430e-01 2.90540636e-01 -5.57813227e-01 -1.04882419e-01
6.10664129e-01 1.50454473e+00 -6.20710731e-01 -3.49184632e-01
-7.24939704e-01 7.01401114e-01 8.96092176e-01 2.10530356e-01
-6.62939191e-01 -2.06326023e-02 6.26887560e-01 2.50514865e-01
1.58129111e-01 9.72943753e-02 -5.85070074e-01 -7.58336425e-01
-3.26450109e-01 -7.24420369e-01 5.14633715e-01 -7.41735935e-01
-1.71300364e+00 6.23856962e-01 -5.22986829e-01 -1.27023900e+00
-4.49781358e-01 -6.47438228e-01 -4.93395597e-01 7.48255551e-01
-1.28724241e+00 -1.16476107e+00 -4.92911756e-01 2.49222666e-01
5.28043747e-01 -2.30702296e-01 1.10062230e+00 5.93622565e-01
-3.23235303e-01 9.44878340e-01 6.73994243e-01 -5.79773001e-02
9.64558542e-01 -1.54572690e+00 6.51083112e-01 1.22455806e-02
4.49193060e-01 3.26911449e-01 5.31834722e-01 -3.20388794e-01
-7.86466062e-01 -1.38725889e+00 1.23171091e+00 -8.06061447e-01
5.80646634e-01 -5.80960155e-01 -6.36936784e-01 9.22059774e-01
-1.97904423e-01 4.51499373e-01 1.25804174e+00 7.56205082e-01
-7.96176612e-01 -3.83686960e-01 -8.66441488e-01 3.86186451e-01
9.52054858e-01 -5.22718489e-01 8.22883770e-02 1.08534825e+00
7.63404667e-01 -2.89434612e-01 -9.05893683e-01 2.96805918e-01
5.75762212e-01 -8.43498051e-01 6.85506344e-01 -6.78447545e-01
7.90749907e-01 2.07443148e-01 -1.16476394e-01 -1.34007084e+00
-8.11862648e-01 -2.94068336e-01 -3.42589915e-01 1.05469310e+00
7.44263411e-01 -3.99619401e-01 1.15641475e+00 7.93619215e-01
-1.83570281e-01 -1.32784176e+00 -2.41236791e-01 -5.77432752e-01
-6.88518211e-02 -4.77484137e-01 7.52251685e-01 1.20550430e+00
3.00882235e-02 4.78649825e-01 -4.38978702e-01 -1.97414100e-01
4.63841528e-01 4.85625356e-01 4.66915518e-01 -1.66386998e+00
-3.10347974e-01 -3.47764671e-01 -2.42898360e-01 -8.27050030e-01
3.24158996e-01 -1.35171902e+00 -2.20171392e-01 -1.62632227e+00
2.57834047e-01 -1.00475192e+00 -8.01505089e-01 8.80139768e-01
1.77346498e-01 6.93074346e-01 -4.13821340e-02 4.19832081e-01
-1.14261568e+00 1.43766269e-01 7.18362272e-01 -6.55664504e-02
1.98956996e-01 2.87721097e-01 -9.80737329e-01 6.03272617e-01
9.96881068e-01 -6.93838358e-01 -5.61769903e-01 -4.54341233e-01
9.46798444e-01 -4.42032605e-01 1.33137312e-02 -8.62871945e-01
1.91885829e-01 1.10624097e-01 5.28818488e-01 -6.75826073e-01
3.17722887e-01 -8.75111222e-01 3.88884209e-02 1.86720505e-01
-1.14397681e+00 -9.69556272e-02 -2.56979674e-01 5.04903257e-01
7.35924542e-02 -2.29549140e-01 5.28630197e-01 -1.22935034e-01
-6.45041764e-01 6.01285756e-01 2.14373797e-01 -2.84349583e-02
1.00321138e+00 3.46509039e-01 -6.76264286e-01 -5.93452871e-01
-8.72369647e-01 2.93458730e-01 2.77081966e-01 4.75126356e-01
1.40082642e-01 -1.56709278e+00 -6.20656013e-01 -6.81846822e-03
4.02255684e-01 -9.04269591e-02 1.33456707e-01 5.30839682e-01
-3.44102353e-01 6.27124071e-01 -3.86252180e-02 -4.19727176e-01
-1.09335649e+00 8.34443510e-01 3.41499120e-01 -2.78601259e-01
-6.25568986e-01 1.22544932e+00 -1.07434159e-02 -4.61574078e-01
4.56692308e-01 -7.93815628e-02 -2.25836486e-01 9.93639827e-02
7.24303186e-01 4.07078296e-01 1.83998987e-01 -5.04725933e-01
-1.98844239e-01 2.62997627e-01 -6.79037631e-01 4.88388896e-01
1.41731656e+00 -4.78483103e-02 8.52259919e-02 6.32467031e-01
1.68838799e+00 -3.99042189e-01 -1.35380435e+00 -7.71024227e-01
1.13948680e-01 -2.96798199e-01 2.33701691e-01 -1.34781420e+00
-1.16343606e+00 8.64558935e-01 6.62525892e-01 4.58615303e-01
8.85343254e-01 3.13381255e-01 1.22118282e+00 5.80024779e-01
1.61700130e-01 -1.06854510e+00 2.93002546e-01 4.32964832e-01
5.73000252e-01 -1.55122054e+00 -6.56899111e-03 -2.51350999e-01
-4.65127975e-01 9.27142501e-01 6.22359514e-01 -5.42949662e-02
8.73365223e-01 1.48417518e-01 1.88748747e-01 -5.36925077e-01
-1.27759969e+00 1.52354926e-01 3.62768680e-01 2.20292151e-01
7.95323789e-01 2.87672073e-01 -1.45012811e-01 7.05025077e-01
-1.63757518e-01 1.53936800e-02 1.71720460e-01 5.76025486e-01
-4.70348179e-01 -9.42881227e-01 4.00221854e-01 1.16853917e+00
-7.38804340e-01 -3.63997608e-01 -3.03040236e-01 1.67266220e-01
2.21114084e-01 9.15670633e-01 2.12016970e-01 -7.34206915e-01
-6.55552521e-02 2.54448682e-01 -2.61139214e-01 -7.62961447e-01
-6.50074482e-01 -1.51800171e-01 1.42227054e-01 -7.78293848e-01
1.55021608e-01 -5.31511605e-01 -1.21025777e+00 -7.37001300e-01
-1.56143546e-01 1.65797114e-01 7.87630320e-01 7.40500093e-01
7.20903158e-01 4.46559519e-01 8.70020866e-01 -6.08207464e-01
-9.49449480e-01 -1.04239464e+00 -7.81701088e-01 6.61936164e-01
2.18676150e-01 -5.51348627e-01 -3.66524965e-01 -2.11912692e-01] | [9.596043586730957, 4.408268928527832] |
8f3ba328-f596-4a65-9f77-a2eb92d69562 | on-the-role-of-depth-predictions-for-3d-human | 2103.02521 | null | https://arxiv.org/abs/2103.02521v1 | https://arxiv.org/pdf/2103.02521v1.pdf | On the role of depth predictions for 3D human pose estimation | Following the successful application of deep convolutional neural networks to 2d human pose estimation, the next logical problem to solve is 3d human pose estimation from monocular images. While previous solutions have shown some success, they do not fully utilize the depth information from the 2d inputs. With the goal of addressing this depth ambiguity, we build a system that takes 2d joint locations as input along with their estimated depth value and predicts their 3d positions in camera coordinates. Given the inherent noise and inaccuracy from estimating depth maps from monocular images, we perform an extensive statistical analysis showing that given this noise there is still a statistically significant correlation between the predicted depth values and the third coordinate of camera coordinates. We further explain how the state-of-the-art results we achieve on the H3.6M validation set are due to the additional input of depth. Notably, our results are produced on neural network that accepts a low dimensional input and be integrated into a real-time system. Furthermore, our system can be combined with an off-the-shelf 2d pose detector and a depth map predictor to perform 3d pose estimation in the wild. | ['Stephen Baek', 'Dmitriy Shin', 'Mitchell Messmore', 'Alec Diaz-Arias'] | 2021-03-03 | null | null | null | null | ['2d-human-pose-estimation'] | ['computer-vision'] | [-1.99609622e-01 1.46944672e-01 1.95371971e-01 -5.09269118e-01
-4.68860835e-01 -3.57387900e-01 3.74510020e-01 -5.02759106e-02
-8.83761585e-01 4.75735068e-01 1.74996123e-01 1.62061587e-01
2.35963136e-01 -5.68294644e-01 -7.49910891e-01 -1.55994356e-01
-6.91652894e-02 8.35387111e-01 3.51617217e-01 -3.81923556e-01
-9.15426854e-03 5.89720607e-01 -1.62788975e+00 -2.38809623e-02
3.33347879e-02 1.21210504e+00 -2.21289054e-01 8.64511609e-01
3.40799451e-01 2.25263223e-01 -6.97262049e-01 -1.77744880e-01
6.54957056e-01 -1.07348941e-01 -5.14276981e-01 7.32512921e-02
8.17707598e-01 -8.62472177e-01 -5.34770787e-01 4.90873754e-01
8.15818071e-01 -1.22770689e-01 4.03185397e-01 -1.18614984e+00
1.26178011e-01 -2.27124110e-01 -2.57550776e-01 -1.16997048e-01
8.55650961e-01 3.70884329e-01 5.01267254e-01 -9.22217250e-01
7.72630334e-01 1.15897596e+00 9.51624691e-01 3.17863792e-01
-8.96692634e-01 -3.02654237e-01 -5.75479642e-02 -1.62183955e-01
-1.44435239e+00 -1.57250702e-01 5.96471190e-01 -6.35593295e-01
1.20353353e+00 -2.75047459e-02 9.84307528e-01 1.12103784e+00
2.68097490e-01 6.72543228e-01 7.61956811e-01 -4.44871962e-01
1.10957518e-01 7.97136426e-02 -2.26443022e-01 6.61849558e-01
2.00050786e-01 3.22501600e-01 -5.98008096e-01 2.29086220e-01
8.73023510e-01 -8.52497853e-03 -1.50189444e-01 -8.30724120e-01
-1.14954495e+00 6.27288282e-01 8.17074895e-01 -7.22038150e-02
-2.46535271e-01 3.05236608e-01 1.44240588e-01 6.97809234e-02
4.03012455e-01 4.57059264e-01 -6.61328256e-01 -3.57195139e-01
-7.92832196e-01 6.32630110e-01 5.12581587e-01 8.25905740e-01
5.61128080e-01 -3.03563565e-01 3.32471281e-01 2.54017770e-01
1.95483029e-01 5.62530339e-01 3.05846304e-01 -1.08327591e+00
5.77346385e-01 7.24998534e-01 2.68673360e-01 -1.10703361e+00
-1.01119471e+00 -4.27583158e-01 -3.00244093e-01 5.75730920e-01
8.16206515e-01 -3.84970754e-01 -8.41403723e-01 1.55573761e+00
3.11028183e-01 -5.81176400e-01 -2.85317987e-01 1.35040128e+00
5.85195780e-01 -2.10316022e-04 -3.29355955e-01 5.39620638e-01
1.09763360e+00 -5.16495347e-01 -2.67099828e-01 -7.23506927e-01
6.70201123e-01 -5.01586080e-01 6.21463776e-01 5.66775084e-01
-8.98677707e-01 -7.66457498e-01 -1.30675614e+00 -3.50623250e-01
-4.58298326e-01 3.60357463e-01 4.62554038e-01 6.33305907e-01
-9.20640469e-01 6.81435049e-01 -9.85951602e-01 -6.42523408e-01
3.07406709e-02 7.52314210e-01 -8.79816532e-01 2.03500450e-01
-1.17701113e+00 1.29172802e+00 3.18954855e-01 2.74858296e-01
-2.89753169e-01 -3.91786277e-01 -1.04922915e+00 -2.53401130e-01
5.00814319e-01 -8.02853227e-01 1.22842073e+00 -3.89427572e-01
-1.26218843e+00 9.59556341e-01 5.27211651e-02 -3.48328590e-01
1.00784934e+00 -7.14133739e-01 3.24484855e-01 1.93909153e-01
5.05556259e-03 1.08694565e+00 3.48113298e-01 -1.01283932e+00
-6.20728016e-01 -8.18991363e-01 1.82209238e-01 4.23545569e-01
3.78634073e-02 -5.89622259e-01 -6.78414583e-01 -8.89421329e-02
4.91351038e-01 -1.22406101e+00 -1.61530182e-01 4.40601945e-01
-4.55441505e-01 2.67101169e-01 4.40336138e-01 -7.98537016e-01
6.43548489e-01 -1.93509030e+00 4.05858010e-01 2.29657426e-01
1.37100026e-01 5.28084747e-02 2.46309057e-01 3.34168613e-01
-3.84980515e-02 -2.63598084e-01 1.61094829e-01 -6.67551041e-01
-1.13211377e-02 -1.26761734e-01 1.79369867e-01 6.05000436e-01
2.65774429e-01 9.54130888e-01 -6.66605175e-01 -2.64704943e-01
6.88164651e-01 7.43900418e-01 -6.34087801e-01 2.14605629e-01
-3.11212260e-02 4.18462694e-01 -1.42496955e-02 4.75019723e-01
4.17981982e-01 3.35002691e-02 1.49647608e-01 -1.88694790e-01
-2.19212519e-03 2.55323648e-01 -1.28111827e+00 1.98402417e+00
-2.76762784e-01 8.23210359e-01 -1.03682419e-02 -4.49203134e-01
8.92809272e-01 1.79324716e-01 6.45318329e-01 -6.10202134e-01
4.45980787e-01 2.28174478e-01 -1.91736415e-01 -2.43519843e-01
7.19337285e-01 8.91611504e-04 -1.60243049e-01 1.36807263e-01
1.86653603e-02 -2.56582022e-01 -2.42820829e-01 -9.46235880e-02
1.06845164e+00 6.17253423e-01 2.93077141e-01 1.75941750e-01
9.26339775e-02 2.32871354e-01 1.70578986e-01 3.91812503e-01
-2.46754676e-01 1.20863497e+00 5.72352946e-01 -7.38098383e-01
-1.31574857e+00 -1.07066560e+00 1.50024846e-01 4.24637020e-01
1.74741782e-02 -4.50808346e-01 -7.13697433e-01 -4.66066748e-01
3.08255494e-01 1.66528136e-01 -8.04587424e-01 1.72037247e-03
-6.25311017e-01 -3.88590634e-01 4.36753035e-01 8.40128124e-01
5.17430186e-01 -5.04343987e-01 -1.37435758e+00 -8.46863165e-03
5.93913160e-02 -1.29141140e+00 -5.55570945e-02 3.86812240e-01
-6.00949645e-01 -1.18477428e+00 -7.77059913e-01 -4.29090708e-01
4.87017959e-01 3.08970753e-02 9.69897330e-01 -1.76800713e-01
-4.11588430e-01 3.69169325e-01 -1.76455438e-01 -3.72857451e-01
1.39802247e-01 2.74893224e-01 2.53939301e-01 -5.84860027e-01
4.88087893e-01 -3.24200124e-01 -7.04953253e-01 3.35534334e-01
-3.59521478e-01 7.86145702e-02 4.45780575e-01 6.05895579e-01
2.38459438e-01 -2.32716888e-01 2.13870034e-02 -5.33119082e-01
9.48802829e-02 -1.51244402e-02 -6.89882755e-01 -3.29721004e-01
-2.76576877e-01 4.16805819e-02 2.01234192e-01 -8.85759369e-02
-5.70973575e-01 7.72916734e-01 -3.51081192e-01 -3.95464897e-01
-4.36402291e-01 3.80710334e-01 -6.85512833e-03 1.31039515e-01
8.15754712e-01 -3.03632349e-01 4.86692041e-01 -4.39344198e-01
2.38501936e-01 5.55781007e-01 7.85245061e-01 -1.80423796e-01
6.24810994e-01 6.35458171e-01 3.21451813e-01 -7.46065497e-01
-7.63234317e-01 -4.43119884e-01 -1.32037306e+00 -2.96604127e-01
1.07761431e+00 -1.21774530e+00 -1.07036030e+00 6.50262237e-01
-1.23288321e+00 -1.71187237e-01 6.91373497e-02 6.38587058e-01
-7.27720976e-01 2.95352221e-01 -2.74202883e-01 -8.76385093e-01
5.11392616e-02 -1.21127474e+00 1.46212256e+00 -1.24030113e-01
-6.68411911e-01 -6.39013767e-01 -1.10592976e-01 3.80707383e-01
8.02236497e-02 6.58222437e-01 4.01074320e-01 -2.68067211e-01
-4.67337370e-01 -9.28823650e-01 4.18623257e-03 1.17160961e-01
-1.55735269e-01 -3.05889189e-01 -9.95900214e-01 -2.68829554e-01
-2.34796882e-01 -4.22837168e-01 5.03705859e-01 5.36994219e-01
5.20749271e-01 3.88873190e-01 -2.80075073e-01 6.48023546e-01
1.09408033e+00 -2.70961672e-01 3.89066428e-01 6.70800328e-01
7.99608290e-01 8.21945727e-01 6.57305479e-01 5.09807646e-01
4.85426217e-01 1.26754892e+00 5.61368883e-01 -1.69358954e-01
-8.92056972e-02 -4.79101509e-01 4.63036001e-02 1.06299132e-01
4.29610945e-02 -2.39149816e-02 -1.10878968e+00 2.71982431e-01
-1.83863473e+00 -5.79672873e-01 1.35508284e-03 2.37157035e+00
3.80784631e-01 5.85979342e-01 4.56728250e-01 2.75097638e-01
2.25092158e-01 -1.11108795e-01 -5.02052903e-01 -1.38794005e-01
7.97254294e-02 3.33333872e-02 8.33781838e-01 5.56843340e-01
-1.21489847e+00 6.16951525e-01 6.89017773e+00 1.90872059e-03
-1.23386550e+00 -5.01020610e-01 3.48222405e-01 -6.98033929e-01
3.32453161e-01 -2.95624852e-01 -8.39432657e-01 2.69146830e-01
8.12760651e-01 4.57902163e-01 4.72569726e-02 9.37097430e-01
1.48487031e-01 -4.90725815e-01 -1.34299242e+00 1.14518452e+00
8.12921375e-02 -8.16967249e-01 -4.84908491e-01 3.08630586e-01
3.87507319e-01 9.42711011e-02 2.08946131e-02 1.33238420e-01
-1.12741210e-01 -1.17101896e+00 8.63150179e-01 4.74076658e-01
7.12729990e-01 -8.99840474e-01 1.03152001e+00 7.50717044e-01
-8.67066324e-01 -2.71793064e-02 -2.55366921e-01 -7.47056901e-01
1.91051319e-01 3.82391572e-01 -1.25499356e+00 4.18122023e-01
7.34721899e-01 6.22635782e-01 -8.70183110e-01 8.52669775e-01
-3.20108414e-01 -2.74966657e-01 -7.25415170e-01 -1.44467726e-02
4.47811671e-02 2.70255208e-01 1.87515363e-01 6.83379471e-01
2.18812674e-01 -1.78490460e-01 1.53238714e-01 6.06850505e-01
3.33134472e-01 -3.38856608e-01 -7.06093907e-01 4.13079172e-01
1.27947494e-01 8.57664526e-01 -5.49696147e-01 1.45070637e-02
-2.33845085e-01 1.04394305e+00 3.41987014e-01 -8.65514576e-02
-6.77797139e-01 -2.97898650e-01 7.12779343e-01 4.34886634e-01
3.48781765e-01 -7.22952664e-01 -6.16695642e-01 -1.00516999e+00
4.42904294e-01 -7.13202238e-01 2.01456733e-02 -1.09459007e+00
-8.31937134e-01 5.46691060e-01 2.23944694e-01 -1.26837015e+00
-8.13709319e-01 -1.18909824e+00 -2.58363653e-02 8.64318848e-01
-9.87014234e-01 -9.29933786e-01 -5.46182632e-01 9.83400196e-02
1.83826804e-01 -4.49949205e-02 7.48643219e-01 2.21551269e-01
-4.06855568e-02 5.34335792e-01 -4.17084247e-01 3.91339064e-01
7.75091827e-01 -1.29770255e+00 8.26938868e-01 6.66041315e-01
-5.73436394e-02 5.48189163e-01 8.79881918e-01 -6.50612473e-01
-1.36379588e+00 -4.34610128e-01 9.87093627e-01 -1.17683876e+00
1.58399075e-01 -5.86467087e-01 -2.24648103e-01 7.01574743e-01
-3.86446863e-01 -5.45308851e-02 2.20203534e-01 2.49259964e-01
-1.21951863e-01 1.33560911e-01 -1.08659482e+00 4.24339503e-01
1.14604104e+00 -4.47814614e-01 -6.14710450e-01 7.85782933e-02
4.42833394e-01 -9.97298658e-01 -6.25932097e-01 4.82129782e-01
9.71220076e-01 -1.26838791e+00 1.09059143e+00 -3.28209132e-01
5.07991731e-01 -2.38372862e-01 -2.49173433e-01 -1.12705731e+00
1.38668686e-01 -1.17967494e-01 2.63919771e-01 4.63378280e-01
1.63889557e-01 -7.73077011e-02 1.43389130e+00 8.77584636e-01
1.46513909e-01 -7.20073164e-01 -1.15019190e+00 -6.13125861e-01
1.30165443e-02 -5.81856787e-01 2.21094921e-01 9.73530114e-02
7.68252686e-02 3.03760707e-01 -5.25417984e-01 1.26169726e-01
5.48519611e-01 -2.74067193e-01 1.38670051e+00 -1.24316323e+00
-2.92423040e-01 -8.40131789e-02 -1.03808725e+00 -1.33887529e+00
-6.50036931e-02 -2.97287196e-01 2.50171989e-01 -1.32774639e+00
-1.55245394e-01 3.41934562e-02 3.42466742e-01 3.56504947e-01
1.67604927e-02 4.95660007e-01 3.64889741e-01 -2.37965584e-02
-3.62427801e-01 3.89493525e-01 9.37159717e-01 3.11768591e-01
-4.51479740e-02 -3.39551941e-02 -2.40413398e-01 7.43297100e-01
5.94076097e-01 -1.76953420e-01 -2.66962022e-01 -5.83698392e-01
3.85214716e-01 1.45320103e-01 7.49635100e-01 -1.54286504e+00
1.86120972e-01 4.48005080e-01 1.07671666e+00 -1.03080463e+00
1.02704132e+00 -9.53632891e-01 9.82486978e-02 7.22808719e-01
-2.25226343e-01 2.72071570e-01 2.07067907e-01 2.13098064e-01
2.58551911e-02 1.62134036e-01 6.41641438e-01 -3.76035005e-01
-7.37874627e-01 7.00502917e-02 -2.30742246e-01 -1.91626117e-01
8.56351435e-01 -5.38702548e-01 -8.95952210e-02 -6.65986836e-01
-7.32634962e-01 7.22274780e-02 8.23792934e-01 5.28172553e-01
5.28217494e-01 -1.32018173e+00 -4.21419382e-01 5.39108396e-01
2.74815895e-02 2.84744889e-01 2.69163717e-02 5.43392181e-01
-8.93091917e-01 8.22775781e-01 -5.05058050e-01 -8.58191013e-01
-1.11888015e+00 2.59115249e-01 5.26129007e-01 -3.26158665e-02
-5.41336536e-01 7.97340453e-01 -2.26018295e-01 -6.35979891e-01
3.53213727e-01 -3.79766911e-01 2.87134171e-01 -1.19708903e-01
2.06738800e-01 2.60130167e-01 2.61758357e-01 -6.86346471e-01
-5.96742988e-01 7.06170678e-01 2.63196051e-01 -5.36912203e-01
1.31557465e+00 -2.16620266e-01 3.93731952e-01 4.94614631e-01
1.28308785e+00 -1.61158696e-01 -1.69842744e+00 6.16628379e-02
-1.78126007e-01 -6.51502132e-01 -1.14142284e-01 -8.39325964e-01
-8.01163912e-01 1.06503010e+00 7.66093552e-01 -2.66664177e-01
5.49393892e-01 -1.67420268e-01 6.95404172e-01 4.56429899e-01
4.97241527e-01 -1.14516270e+00 1.36292636e-01 7.33050644e-01
6.76587343e-01 -1.49377823e+00 3.54438096e-01 -3.97590011e-01
-4.90639865e-01 1.17344713e+00 7.79439986e-01 -2.32641384e-01
4.55605417e-01 2.64025569e-01 4.33235109e-01 -1.55316919e-01
-4.25644010e-01 -2.16434181e-01 3.28390807e-01 7.74518609e-01
6.01673603e-01 -6.95515648e-02 1.60513818e-02 2.14632168e-01
-7.30575800e-01 2.79146768e-02 4.33530390e-01 9.79560137e-01
-4.75555718e-01 -8.77594590e-01 -4.83981222e-01 -1.05312236e-01
-2.54100055e-01 3.93207788e-01 -6.70138061e-01 1.20784497e+00
1.30130574e-01 7.49336243e-01 1.87742651e-01 -8.06338131e-01
6.69902384e-01 1.14593804e-01 6.93370163e-01 -6.67509556e-01
-4.97434229e-01 -5.07289805e-02 5.57510443e-02 -8.25344503e-01
-6.35955408e-02 -5.90544224e-01 -1.08038127e+00 -3.18693101e-01
4.37134057e-02 -4.30984557e-01 1.01439905e+00 1.04122221e+00
2.88729340e-01 3.81257057e-01 6.23624995e-02 -1.48447466e+00
-5.03264785e-01 -8.05178106e-01 -4.12894607e-01 4.39320505e-01
5.01922846e-01 -8.04564655e-01 -1.82976291e-01 -3.50679398e-01] | [6.91553258895874, -0.9779030680656433] |
37d02c04-999d-49f2-88d5-d67c71b85e27 | semi-supervised-image-captioning-with-clip | 2306.15111 | null | https://arxiv.org/abs/2306.15111v1 | https://arxiv.org/pdf/2306.15111v1.pdf | Semi-Supervised Image Captioning with CLIP | Image captioning, a fundamental task in vision-language understanding, seeks to generate accurate natural language descriptions for provided images. The CLIP model, with its rich semantic features learned from a large corpus of image-text pairs, is well-suited for this task. In this paper, we present a two-stage semi-supervised image captioning approach that exploits the potential of CLIP encoding. Our model comprises a CLIP visual encoder, a mapping network, and a language model for text generation. In the initial stage, we train the model using a small labeled dataset by contrasting the generated captions with the ground truth captions. In the subsequent stage, we continue the training using unlabeled images, aiming to maximize the image-caption similarity based on CLIP embeddings. Remarkably, despite utilizing less than 2% of the COCO-captions, our approach delivers a performance comparable to state-of-the-art models trained on the complete dataset. Furthermore, the captions generated by our approach are more distinctive, informative, and in line with human preference. | ['Chuanyang Jin'] | 2023-06-26 | null | null | null | null | ['image-captioning', 'text-generation'] | ['computer-vision', 'natural-language-processing'] | [ 6.40172064e-01 5.53090513e-01 -1.35105669e-01 -4.80486065e-01
-1.19909072e+00 -5.33325493e-01 1.01470208e+00 1.55326482e-02
-3.27662081e-01 8.55136216e-01 5.28967977e-01 1.52892008e-01
5.72913706e-01 -3.15796643e-01 -1.14341998e+00 -3.14194560e-01
4.44314539e-01 7.24721909e-01 -1.18029229e-02 4.91800234e-02
2.70833403e-01 5.02053462e-02 -1.58648193e+00 6.39060557e-01
7.11756527e-01 1.17393792e+00 5.76949120e-01 5.51735878e-01
-1.99580818e-01 1.13702118e+00 -2.73728907e-01 -6.62013948e-01
2.21702270e-02 -7.56935596e-01 -8.89554918e-01 3.91659498e-01
8.00337195e-01 -3.36065501e-01 -2.52810299e-01 1.03420269e+00
2.45585784e-01 -2.67862588e-01 8.01001370e-01 -1.31505322e+00
-1.14368916e+00 6.07802629e-01 -4.17840660e-01 -2.77121186e-01
5.86418927e-01 1.13758393e-01 1.13202143e+00 -1.26922262e+00
1.06649125e+00 1.02121842e+00 3.09163809e-01 9.14405823e-01
-1.39323866e+00 -2.63954699e-01 2.10556854e-02 -2.08253833e-03
-1.24274790e+00 -6.07991874e-01 8.62006187e-01 -6.07359409e-01
5.29623091e-01 -3.05124652e-02 3.97792488e-01 1.49790823e+00
-2.68783122e-01 9.76174414e-01 7.86207557e-01 -7.63617039e-01
2.22327441e-01 7.04173982e-01 -3.15826058e-01 3.87734324e-01
4.38992046e-02 5.67345973e-03 -4.42669898e-01 -2.57204529e-02
6.23553097e-01 -2.27125689e-01 -4.72016454e-01 -7.82680929e-01
-1.37362409e+00 9.88822103e-01 5.52330732e-01 2.80272305e-01
-4.83825862e-01 3.50979537e-01 4.04799134e-01 -1.72249600e-01
4.67175245e-01 6.52146518e-01 8.52765590e-02 2.59575006e-02
-1.18935823e+00 1.97903961e-01 6.83293641e-01 1.32750392e+00
5.60576797e-01 -2.08066761e-01 -5.81953585e-01 7.40202367e-01
3.29764992e-01 5.68880796e-01 3.85518581e-01 -8.25232089e-01
6.22473538e-01 2.93857425e-01 3.87316585e-01 -7.21038222e-01
3.00748259e-01 -2.48482928e-01 -6.30932093e-01 -1.55840376e-02
1.97610736e-01 1.04751185e-01 -9.97586668e-01 1.89233303e+00
-1.72904506e-01 1.00277551e-01 3.93430144e-01 9.29924965e-01
6.72557771e-01 7.71616995e-01 3.07456285e-01 -1.38279006e-01
1.30590963e+00 -1.38929665e+00 -7.28980660e-01 -5.28359592e-01
2.83492267e-01 -8.78494680e-01 1.08735228e+00 -6.52881339e-02
-1.23464036e+00 -6.92942262e-01 -9.05781567e-01 -2.14245498e-01
-3.13268065e-01 3.51288050e-01 1.76338762e-01 2.20672816e-01
-1.33253527e+00 1.51830748e-01 -3.30135584e-01 -4.95439142e-01
5.86876869e-01 -1.93781734e-01 -4.32100326e-01 -3.73645812e-01
-1.04209590e+00 9.35537338e-01 5.61265707e-01 -1.27127662e-01
-1.17394590e+00 -6.72569513e-01 -1.12950635e+00 1.57259740e-02
4.36936952e-02 -7.45572805e-01 1.38592660e+00 -1.48393953e+00
-1.17380583e+00 1.25631547e+00 -2.00398102e-01 -6.68419600e-01
6.92305505e-01 -1.35511801e-01 -7.84656405e-02 5.96973538e-01
3.09444278e-01 1.64310908e+00 1.10219574e+00 -1.88102913e+00
-3.90912533e-01 4.41707484e-02 -8.88209045e-03 2.32787296e-01
-2.11478949e-01 -9.79615897e-02 -6.35094047e-01 -5.76740146e-01
-5.37247896e-01 -9.43241537e-01 -2.85927743e-01 3.84648681e-01
-3.81898820e-01 4.80555855e-02 5.88593125e-01 -5.78938544e-01
7.55630791e-01 -2.25459957e+00 2.23310560e-01 -1.95465237e-01
-3.42277549e-02 3.51932824e-01 -5.05804896e-01 7.17309177e-01
-1.44557238e-01 -1.18157556e-02 -5.61004877e-01 -8.52820158e-01
7.33745564e-03 4.28555049e-02 -7.17505455e-01 1.62666112e-01
7.17993200e-01 1.29895103e+00 -1.08002496e+00 -8.18370342e-01
3.70421708e-01 5.71657896e-01 -3.65375996e-01 6.84806168e-01
-7.81082749e-01 4.11233783e-01 -3.59979689e-01 3.06340963e-01
5.69871902e-01 -4.88233179e-01 8.48216377e-03 -3.48380208e-01
1.06457770e-01 -2.39227653e-01 -4.29490596e-01 2.20429468e+00
-6.00770831e-01 1.12622762e+00 -2.58344620e-01 -9.93536532e-01
9.40433264e-01 5.12352407e-01 3.16829652e-01 -7.59852469e-01
1.01042867e-01 2.78147668e-01 -6.71869516e-01 -7.38474488e-01
5.80270052e-01 -5.42824194e-02 -1.18926257e-01 4.84847873e-01
2.79174894e-01 -3.55307668e-01 3.87756765e-01 4.22320664e-01
5.35573721e-01 4.77755547e-01 1.04419112e-01 5.70493340e-02
5.62265992e-01 2.80491948e-01 -1.81678653e-01 6.02969706e-01
-1.07704289e-01 1.32927036e+00 3.60109568e-01 -3.16184402e-01
-1.62127709e+00 -1.02730978e+00 -1.41331777e-01 6.03955507e-01
1.27199933e-01 -1.67627543e-01 -9.96314645e-01 -5.19308269e-01
-2.69738287e-01 8.77375901e-01 -8.77444565e-01 1.72176268e-02
-2.99695909e-01 1.97652970e-02 3.34968597e-01 6.51517093e-01
4.36199367e-01 -1.42886305e+00 -6.40487790e-01 6.91789463e-02
-5.24809003e-01 -1.69157422e+00 -7.19228148e-01 -2.38208190e-01
-4.91189480e-01 -8.91011894e-01 -1.28210104e+00 -1.31547129e+00
1.06319284e+00 2.01729879e-01 1.44386828e+00 -5.72786480e-02
-3.63067269e-01 6.43615663e-01 -5.88865995e-01 -3.33097875e-01
-6.75242186e-01 -1.94415614e-01 -4.98728216e-01 3.35379273e-01
1.69876978e-01 -2.97190964e-01 -5.58531284e-01 -1.10731483e-01
-1.27176964e+00 5.45038998e-01 7.52560198e-01 9.79175746e-01
5.80636382e-01 -8.06249201e-01 4.20045167e-01 -8.18046331e-01
4.55889672e-01 -3.91398221e-01 -3.70424896e-01 4.89664227e-01
-2.78956383e-01 1.25056565e-01 8.07892203e-01 -4.45414901e-01
-1.09801304e+00 5.83124697e-01 1.07461967e-01 -8.52551460e-01
-2.41238430e-01 2.73677975e-01 1.98774301e-02 1.02684870e-01
7.26873994e-01 5.14336288e-01 8.26384500e-02 -2.12055549e-01
9.86630142e-01 8.45429540e-01 8.79628897e-01 -3.49762827e-01
8.49921346e-01 5.50995588e-01 -3.86558115e-01 -6.34004116e-01
-1.11962068e+00 -3.82022321e-01 -7.23393440e-01 -2.96542317e-01
1.26326036e+00 -1.20015657e+00 -7.35157430e-02 7.45527968e-02
-1.57767546e+00 -6.91896975e-02 -4.18956578e-01 4.01142359e-01
-1.12166023e+00 1.91872105e-01 -2.24077955e-01 -7.52307773e-01
-4.92484510e-01 -1.08417678e+00 1.54156184e+00 1.88253850e-01
-9.45691243e-02 -9.64894235e-01 1.25919849e-01 5.08053064e-01
3.92544210e-01 5.45101225e-01 5.66078067e-01 -6.27694130e-01
-6.83940470e-01 -2.14445874e-01 -6.74184561e-01 4.88279253e-01
-1.24391884e-01 -2.77646691e-01 -1.17526174e+00 -2.60514438e-01
-2.46127367e-01 -9.99408066e-01 8.16653073e-01 2.48230726e-01
1.02735841e+00 -2.51319528e-01 -2.74855047e-01 3.33460063e-01
1.75754952e+00 -5.36449775e-02 7.70027220e-01 1.27452672e-01
5.72366774e-01 8.01218450e-01 6.20797038e-01 1.77917376e-01
2.94349372e-01 6.14873707e-01 5.96798539e-01 -3.77911299e-01
-4.01400775e-01 -9.00446832e-01 2.66086251e-01 3.94584298e-01
5.14884710e-01 -4.99390572e-01 -6.92427635e-01 9.60970461e-01
-2.00727725e+00 -9.89407003e-01 1.19661689e-01 1.96558416e+00
8.64457309e-01 7.71803735e-03 -2.39944950e-01 -3.33757341e-01
8.45703363e-01 2.25239694e-01 -4.72881138e-01 -2.96546787e-01
-1.38008714e-01 -4.89052981e-02 2.22117409e-01 3.73137206e-01
-9.59330618e-01 9.17801917e-01 6.39128208e+00 6.19402468e-01
-9.79872286e-01 -8.19354653e-02 7.97807992e-01 1.56970963e-01
-4.84133452e-01 8.33033770e-02 -4.12778258e-01 4.58882123e-01
9.11561489e-01 -1.60620376e-01 2.59901494e-01 9.36269045e-01
1.31320968e-01 -2.58682333e-02 -1.49867666e+00 1.20754611e+00
8.57145905e-01 -1.67241120e+00 5.41075110e-01 -2.54055738e-01
9.50042009e-01 -8.87283236e-02 3.11085731e-02 -3.15040387e-02
1.41348122e-02 -1.17670453e+00 1.08249044e+00 5.58259130e-01
1.16976309e+00 -3.51342022e-01 7.35821187e-01 9.17603523e-02
-8.60782325e-01 1.58453003e-01 -2.75819838e-01 2.69609392e-01
6.72972262e-01 3.72457623e-01 -7.41275370e-01 5.11690795e-01
3.04064453e-01 8.61474276e-01 -6.67331457e-01 8.92819345e-01
-2.56229281e-01 2.46622145e-01 1.53777018e-01 -1.64235067e-02
4.73215252e-01 -5.76306768e-02 2.61008710e-01 1.21385717e+00
1.37353450e-01 -3.05179030e-01 1.61562294e-01 1.23853731e+00
-3.13628167e-01 2.54673958e-01 -8.92368913e-01 -4.27639574e-01
2.39480883e-01 1.19206119e+00 -4.57943738e-01 -6.21506155e-01
-4.58881140e-01 1.27368808e+00 4.83309239e-01 3.61074656e-01
-8.32792342e-01 -3.40812385e-01 1.26089111e-01 -7.63252825e-02
3.90495509e-01 1.61188290e-01 -5.17595606e-03 -1.23906732e+00
2.66581059e-01 -7.32267439e-01 -9.97429788e-02 -1.54384422e+00
-1.33297873e+00 1.10405385e+00 -6.68614581e-02 -1.48518217e+00
-6.12447560e-01 -5.30909896e-01 -6.35409892e-01 7.19616771e-01
-1.71440387e+00 -1.61349487e+00 -5.24419844e-01 4.73991215e-01
9.16152358e-01 1.42152933e-02 9.15788114e-01 1.64586470e-01
-5.95075265e-02 4.42433804e-01 -4.69744354e-02 2.48787090e-01
7.43924320e-01 -1.07466388e+00 4.70341623e-01 7.70519853e-01
4.41086173e-01 2.13916242e-01 9.90458846e-01 -3.32507700e-01
-1.00254631e+00 -1.28006399e+00 1.05099428e+00 -4.07926798e-01
5.23356080e-01 -5.40369928e-01 -7.42393374e-01 6.54404700e-01
7.95854986e-01 1.04029827e-01 5.98886430e-01 -6.55381978e-01
-5.19839644e-01 2.38148972e-01 -9.15554702e-01 5.60932100e-01
8.74210835e-01 -7.21228004e-01 -7.14669228e-01 5.60337186e-01
9.78662789e-01 -2.06310555e-01 -5.65161645e-01 -3.01186219e-02
4.44471627e-01 -7.01076746e-01 8.06816161e-01 -5.49812376e-01
1.09761357e+00 -2.50708580e-01 -1.51600137e-01 -1.15771985e+00
-3.03308275e-02 -5.46736360e-01 3.35697591e-01 1.34544885e+00
7.26610541e-01 4.26658392e-02 6.60127819e-01 4.88980502e-01
-1.31583691e-01 -3.97182286e-01 -5.89080393e-01 -6.51539385e-01
-5.08629195e-02 -1.43888861e-01 2.46682316e-01 7.17030704e-01
1.96766090e-02 5.62029600e-01 -6.59311473e-01 -2.48122603e-01
8.95609081e-01 2.95550883e-01 7.49706149e-01 -8.96785200e-01
-1.30669162e-01 -9.17957872e-02 -2.49328956e-01 -1.05565572e+00
3.48196149e-01 -9.56980109e-01 4.83581126e-01 -1.69645011e+00
5.41185915e-01 -9.63325128e-02 1.10287085e-01 2.82665640e-01
-5.46867140e-02 6.45113289e-01 4.69249487e-01 2.99153775e-01
-7.94049680e-01 8.04352462e-01 1.29176211e+00 -3.21509063e-01
1.09621324e-01 -5.75921237e-01 -7.09994614e-01 3.97023916e-01
5.53719938e-01 -3.85896534e-01 -5.66820264e-01 -6.97137594e-01
-3.07455678e-02 1.00605346e-01 4.93589818e-01 -9.16845381e-01
1.82040513e-01 -3.58094983e-02 3.49378645e-01 -4.00862366e-01
5.86774111e-01 -7.88811803e-01 8.37756917e-02 3.13848466e-01
-9.02347624e-01 -7.29063973e-02 8.12806934e-02 6.19388342e-01
-5.55234253e-01 -4.77811545e-01 7.67352581e-01 -1.53726831e-01
-5.79635620e-01 2.82354951e-01 -4.26166095e-02 2.80959129e-01
1.24990284e+00 -1.06457114e-01 -3.28576624e-01 -5.88811636e-01
-5.53014338e-01 2.97022730e-01 7.27716863e-01 8.06067824e-01
7.84498990e-01 -1.51007533e+00 -9.77075458e-01 8.53670761e-02
8.45707536e-01 -6.94360882e-02 4.19756286e-02 3.37191671e-01
-5.03768325e-01 6.58506870e-01 -3.75276446e-01 -6.57548547e-01
-8.62581313e-01 9.30823624e-01 8.63516554e-02 -1.45689756e-01
-4.48069364e-01 5.49724877e-01 3.71406615e-01 1.31931141e-01
2.58795232e-01 -6.81553409e-02 -1.96137682e-01 -6.33136928e-02
7.62279689e-01 -4.25208569e-01 -3.79326075e-01 -8.82582784e-01
-7.03899339e-02 6.30569577e-01 -1.41906559e-01 -4.97863561e-01
1.18743682e+00 -2.42976338e-01 1.20228261e-01 2.47611016e-01
1.47391403e+00 -3.91508371e-01 -1.51237023e+00 -2.59335965e-01
-2.11104769e-02 -3.68064165e-01 -2.04893142e-01 -6.94047451e-01
-9.23475742e-01 9.28788781e-01 4.44886714e-01 -1.65746994e-02
1.01649749e+00 2.90306211e-01 7.16884494e-01 8.44139978e-02
2.44587138e-01 -8.11154366e-01 4.75472808e-01 4.88196574e-02
1.23127651e+00 -1.58752465e+00 -3.46243203e-01 -3.68829072e-01
-1.02046549e+00 9.38989222e-01 5.89387238e-01 -1.45369962e-01
1.79811463e-01 -1.90701470e-01 1.89649105e-01 2.14215746e-04
-9.10712421e-01 -2.87580878e-01 2.11704075e-01 9.43918407e-01
2.91727275e-01 -1.83817670e-01 -1.38824120e-01 2.16016576e-01
-5.70735410e-02 2.33842373e-01 6.95487380e-01 7.01742232e-01
-3.59885305e-01 -1.03651392e+00 -1.79697201e-01 1.38955921e-01
-1.50285751e-01 -3.44413340e-01 -5.38829684e-01 4.41809326e-01
-2.39496589e-01 8.57716918e-01 2.90322185e-01 -5.50765684e-03
2.01566309e-01 2.24492792e-02 4.15381640e-01 -6.18778288e-01
-1.60026804e-01 -2.33370572e-01 -7.80995330e-03 -4.91821587e-01
-6.78561926e-01 -4.15558070e-01 -9.68193114e-01 4.21586305e-01
-1.27940625e-01 3.10315788e-01 8.93682837e-01 7.66439557e-01
4.35528904e-01 1.48986399e-01 7.54858255e-01 -1.03474689e+00
-4.13004875e-01 -9.25340176e-01 -2.75499314e-01 9.68040824e-01
3.45230818e-01 -3.58071864e-01 -3.39985460e-01 7.94955790e-01] | [10.989714622497559, 1.018776535987854] |
a887da77-260b-4d24-aded-b26a8edfd223 | person-recognition-using-smartphones | 1711.04689 | null | http://arxiv.org/abs/1711.04689v1 | http://arxiv.org/pdf/1711.04689v1.pdf | Person Recognition using Smartphones' Accelerometer Data | Smartphones have become quite pervasive in various aspects of our daily
lives. They have become important links to a host of important data and
applications, which if compromised, can lead to disastrous results. Due to
this, today's smartphones are equipped with multiple layers of authentication
modules. However, there still lies the need for a viable and unobtrusive layer
of security which can perform the task of user authentication using resources
which are cost-efficient and widely available on smartphones. In this work, we
propose a method to recognize users using data from a phone's embedded
accelerometer sensors. Features encapsulating information from both time and
frequency domains are extracted from walking data samples, and are used to
build a Random Forest ensemble classification model. Based on the experimental
results, the resultant model delivers an accuracy of 0.9679 and Area under
Curve (AUC) of 0.9822. | ['Rajsekhar Kumar Nath', 'A. V. Narsimhadhan', 'Thingom Bishal Singha'] | 2017-11-13 | null | null | null | null | ['person-recognition'] | ['computer-vision'] | [ 1.12649798e-01 -2.12988287e-01 -3.70827556e-01 -2.00583562e-01
-4.34759915e-01 -2.57905871e-01 4.46421176e-01 3.65144253e-01
-4.69805896e-01 7.61941314e-01 -6.21635579e-02 -4.23465401e-01
-9.71780624e-03 -1.00045323e+00 -5.85062392e-02 -5.51945448e-01
-6.95194677e-02 -4.04025972e-01 3.51260126e-01 -1.56632781e-01
2.89237797e-01 3.19047183e-01 -1.76962030e+00 -1.64041713e-01
1.02311802e+00 1.39585698e+00 -2.72770792e-01 3.73802066e-01
3.41745079e-01 1.41936764e-01 -7.04075933e-01 -6.80903256e-01
-1.41109630e-01 3.35236080e-02 -2.45447248e-01 -3.16398263e-01
-3.96930635e-01 -4.01102990e-01 9.38039571e-02 9.64357972e-01
3.72433335e-01 -6.52255416e-02 3.11058760e-01 -1.41352654e+00
3.64432409e-02 2.57829800e-02 -6.93298221e-01 -2.47862451e-02
7.52622008e-01 -2.23106012e-01 6.01921737e-01 -6.21047556e-01
-1.91770434e-01 7.50568688e-01 9.45159316e-01 1.22801133e-01
-9.63435888e-01 -9.53731656e-01 -2.69019395e-01 2.74633497e-01
-1.51367259e+00 -5.48951864e-01 7.05648601e-01 -2.98319161e-01
6.20747089e-01 5.37843347e-01 7.46594012e-01 1.05347848e+00
4.62752759e-01 3.20463657e-01 1.06952608e+00 -3.18497688e-01
1.95857555e-01 5.29652655e-01 9.73136202e-02 4.61774409e-01
9.73352432e-01 -1.44915119e-01 -5.03879547e-01 -4.11400974e-01
2.54174650e-01 4.68662739e-01 -6.13712929e-02 -2.93442160e-02
-8.45504403e-01 6.07635677e-01 2.61399508e-01 3.45599532e-01
-4.33890164e-01 -3.56476635e-01 2.78112978e-01 -1.88133687e-01
1.33358821e-01 2.89501213e-02 -2.97599882e-01 -4.42249268e-01
-6.69579804e-01 -1.36586532e-01 5.83462715e-01 4.36776012e-01
5.08900881e-01 -3.17714930e-01 4.32245672e-01 6.81039870e-01
6.75373614e-01 6.70209289e-01 7.61897445e-01 -3.49908382e-01
3.63336474e-01 1.04895663e+00 2.79161096e-01 -1.48539484e+00
-5.10645807e-01 -2.80515671e-01 -1.14409959e+00 -1.76518232e-01
2.27368072e-01 -6.49662539e-02 -1.61938339e-01 1.47348833e+00
5.55029154e-01 2.75544375e-01 -2.10084856e-01 3.35289299e-01
4.66943979e-01 3.75971913e-01 2.40248039e-01 -2.21934661e-01
1.67356646e+00 -2.08218604e-01 -7.78502762e-01 5.09918341e-03
2.94215381e-01 -6.19315684e-01 1.30583251e+00 6.75875425e-01
-6.39249504e-01 -6.60654485e-01 -1.49318767e+00 2.81425625e-01
-5.87816298e-01 1.88177794e-01 6.12918556e-01 1.45920241e+00
-5.49300790e-01 4.86268133e-01 -1.08890390e+00 -3.17641526e-01
3.98242921e-01 6.39220715e-01 -3.56712639e-01 4.03666981e-02
-1.28929210e+00 6.49829865e-01 1.96081236e-01 1.24725595e-01
-1.36360765e-01 4.11820821e-02 -6.78818405e-01 -6.23046607e-02
1.20171532e-01 -3.68549198e-01 9.18670595e-01 -2.00179175e-01
-1.30210900e+00 4.90587890e-01 -6.75293580e-02 -3.57498765e-01
3.42277497e-01 -4.19295520e-01 -9.70927536e-01 7.20776841e-02
2.46961877e-01 -1.01506777e-01 8.73609483e-01 -7.17523456e-01
-8.29675853e-01 -6.83695018e-01 -7.17289969e-02 -1.62042722e-01
-9.88484442e-01 -4.65388829e-03 -2.18465105e-01 -4.10254091e-01
8.02086368e-02 -9.60978389e-01 -6.48864657e-02 -3.66040438e-01
-5.29584110e-01 1.33636415e-01 9.69727933e-01 -1.03514004e+00
1.76660550e+00 -2.12439275e+00 -4.00015146e-01 3.53778213e-01
1.67817950e-01 6.10061586e-01 8.14490497e-01 2.93273032e-01
3.48186910e-01 2.37409353e-01 -7.56398812e-02 -2.88745999e-01
-1.72993779e-01 -2.33929306e-02 -1.21344000e-01 3.75503540e-01
-1.90113425e-01 5.97559571e-01 -7.48416960e-01 -3.18627715e-01
4.20632452e-01 7.31226861e-01 -1.24065965e-01 2.57591039e-01
5.88899314e-01 4.16057855e-01 -7.37322986e-01 9.00119603e-01
4.78740335e-01 -1.73580021e-01 2.81135559e-01 1.17156200e-01
2.14769635e-02 3.90256315e-01 -1.16525459e+00 1.12519920e+00
-4.04695928e-01 1.45082816e-01 -2.31033221e-01 -8.15223992e-01
8.11122417e-01 3.26821327e-01 2.86814094e-01 -4.52314049e-01
4.79594350e-01 2.31805280e-01 -2.94159055e-01 -4.85280246e-01
4.38466102e-01 1.57105848e-01 -4.31053698e-01 3.12680185e-01
-4.76927221e-01 3.42244297e-01 -1.36989534e-01 -7.72893652e-02
1.15538275e+00 -1.14499740e-01 7.04203367e-01 1.09541498e-01
1.07750869e+00 -5.30766785e-01 3.87525439e-01 3.31645787e-01
-3.40204477e-01 6.63225874e-02 7.43412599e-02 -2.47021645e-01
-3.67681772e-01 -9.15247858e-01 -2.02174425e-01 5.91130137e-01
2.20060438e-01 -6.69750988e-01 -7.79885232e-01 -5.39067626e-01
-4.50575650e-02 4.51092392e-01 -4.31070954e-01 -4.41199332e-01
-1.12331539e-01 -8.65962267e-01 8.34887266e-01 3.76678288e-01
1.09720588e+00 -5.57744920e-01 -8.92273128e-01 2.20272467e-01
-3.67514729e-01 -1.22767639e+00 -8.46871510e-02 -3.06849241e-01
-9.05364335e-01 -1.13544977e+00 -2.60513067e-01 -1.12201318e-01
4.56372857e-01 6.41005754e-01 5.47737062e-01 3.16711426e-01
-1.25488758e-01 1.59076720e-01 -3.77150953e-01 -5.57346225e-01
2.87956558e-02 3.39111298e-01 4.63165522e-01 3.99989128e-01
5.70920467e-01 -8.26712251e-01 -7.08182633e-01 5.59168935e-01
-6.93641961e-01 -3.05893064e-01 3.59438360e-01 4.73785281e-01
2.20394120e-01 4.70197469e-01 8.09589088e-01 -6.24703109e-01
6.80977643e-01 -5.93874574e-01 -5.14808059e-01 -2.13069525e-02
-8.79125357e-01 -5.75738490e-01 7.13543236e-01 -3.00895095e-01
-7.28986263e-01 -9.65976492e-02 -2.97901064e-01 3.66214305e-01
-2.57319778e-01 4.50467110e-01 -6.16643906e-01 -1.83608830e-02
3.78384739e-01 6.50412217e-02 -1.02751233e-01 -5.03397763e-01
-2.00093985e-01 1.28707981e+00 5.03393054e-01 -3.25833052e-01
8.25330794e-01 5.15505731e-01 4.08895239e-02 -1.14955568e+00
-6.48955524e-01 -5.45795262e-01 -4.08832312e-01 -2.06698388e-01
7.24369109e-01 -7.60428071e-01 -1.02316701e+00 8.29384208e-01
-7.19904244e-01 3.86434972e-01 3.47717166e-01 3.65310580e-01
7.27158040e-02 6.12843513e-01 -2.47844428e-01 -1.26921391e+00
-3.59310925e-01 -8.69328141e-01 7.58314788e-01 6.18691742e-01
-6.63914382e-01 -6.82076991e-01 -1.59981474e-01 6.59983933e-01
3.86146218e-01 5.63512504e-01 3.84435296e-01 -5.00021577e-01
-2.22986206e-01 -9.31912899e-01 1.49946334e-02 1.97571114e-01
6.05921388e-01 -8.40946287e-02 -1.27271271e+00 -2.80636907e-01
2.45476037e-01 -3.22417263e-03 2.67565578e-01 6.65512159e-02
1.29422987e+00 -4.96173084e-01 -5.32146394e-01 5.32747388e-01
1.10590577e+00 4.63258654e-01 9.18251336e-01 4.33458239e-01
5.01626790e-01 4.37662125e-01 7.47369766e-01 6.29322231e-01
6.30257010e-01 7.02683449e-01 3.96150112e-01 2.90985227e-01
5.84528565e-01 -4.07091767e-01 3.10683697e-01 6.78644121e-01
-2.75830090e-01 -5.34544978e-03 -8.50296438e-01 2.89812237e-01
-1.63738418e+00 -9.48652148e-01 -1.69860527e-01 2.86214256e+00
4.75210130e-01 4.54583526e-01 4.29530442e-01 1.05137205e+00
7.00876296e-01 -1.60865337e-01 -4.05459464e-01 -1.77020505e-01
4.12017077e-01 2.79818028e-01 3.66844475e-01 7.05150366e-02
-1.20245445e+00 2.33410597e-01 5.41319561e+00 5.17820299e-01
-1.12838411e+00 9.55007151e-02 6.49948537e-01 3.68499577e-01
2.36415505e-01 -2.83276558e-01 -7.42023587e-01 1.03412426e+00
1.35236657e+00 -1.01538703e-01 1.77348778e-01 8.12819183e-01
3.31644267e-01 -5.51644146e-01 -6.24892414e-01 1.05262852e+00
-1.59047842e-01 -7.35432327e-01 -5.18738270e-01 5.28558552e-01
2.02052966e-01 -4.43394870e-01 2.31518194e-01 1.73758686e-01
-1.65428907e-01 -9.51496243e-01 1.13399632e-01 2.07479864e-01
8.60590577e-01 -1.02341747e+00 9.68447089e-01 7.12635040e-01
-1.29623771e+00 -2.05599159e-01 7.37930015e-02 -5.02714574e-01
1.85891613e-01 8.34125161e-01 -7.48698533e-01 6.68107331e-01
8.24895859e-01 5.07259727e-01 -6.32789552e-01 9.99663651e-01
-2.91911989e-01 6.32737994e-01 -4.13335145e-01 -2.39945173e-01
-4.65040505e-01 -2.13210598e-01 -1.40538923e-02 6.94277287e-01
7.18481421e-01 8.75989795e-02 -1.05406269e-01 5.66764213e-02
3.56217548e-02 1.36783049e-01 -8.20546806e-01 -4.60819993e-03
7.82035708e-01 1.49761426e+00 -9.15914655e-01 -2.01866269e-01
-6.08062267e-01 7.01109946e-01 -2.82996088e-01 -2.31595099e-01
-8.85837853e-01 -6.95726156e-01 6.34114623e-01 4.03622448e-01
-1.70749143e-01 -2.48712823e-01 -5.28368115e-01 -1.22774410e+00
1.55067280e-01 -8.71895373e-01 3.61188233e-01 -3.06808263e-01
-1.03864145e+00 2.67072886e-01 -2.75103807e-01 -1.43996704e+00
-2.26528019e-01 -4.78932112e-01 -5.84418058e-01 6.08821988e-01
-1.21188915e+00 -1.04171431e+00 -5.83359778e-01 6.51288271e-01
-1.47252247e-01 -4.09389883e-02 1.15595102e+00 6.29117906e-01
-7.48368561e-01 7.17249155e-01 -6.25284761e-02 1.49468377e-01
3.71661961e-01 -8.01118612e-01 2.82028854e-01 9.33376014e-01
5.03516048e-02 1.10010421e+00 5.60570657e-01 -5.84440768e-01
-1.29168165e+00 -7.96116590e-01 9.18677509e-01 -4.42894012e-01
4.71104920e-01 -2.34276175e-01 -6.36519134e-01 3.33787084e-01
-4.16530967e-01 -2.21568108e-01 1.31602108e+00 2.61143833e-01
-1.08148202e-01 -2.93535233e-01 -1.48528361e+00 4.23018277e-01
5.54130793e-01 -5.83425581e-01 -4.07644629e-01 -2.06551060e-01
1.93287387e-01 -9.16959643e-02 -9.14420784e-01 3.62097651e-01
1.07103324e+00 -1.27410197e+00 1.13891268e+00 -4.96998616e-02
-4.03252132e-02 -3.65092456e-01 -2.42073357e-01 -8.23281825e-01
2.09611833e-01 -6.49249852e-01 -5.52053392e-01 1.41331470e+00
1.15510739e-01 -1.09500408e+00 7.59753466e-01 8.58147740e-01
3.91379267e-01 -6.84654117e-01 -1.01023149e+00 -5.58552802e-01
-6.15787625e-01 -6.92907929e-01 9.39757943e-01 8.76685202e-01
3.26673150e-01 3.58180374e-01 -4.36906457e-01 1.92893058e-01
6.41249955e-01 -3.52918625e-01 8.73276353e-01 -1.70515585e+00
-1.22461274e-01 -1.17679805e-01 -6.51086867e-01 -5.91908813e-01
-3.89659226e-01 -3.03069800e-01 -4.70403254e-01 -1.16519737e+00
-4.36615292e-03 -4.78608400e-01 -5.98442018e-01 4.24116015e-01
-2.69240260e-01 5.89507103e-01 -2.50268638e-01 1.48095414e-01
-4.04184997e-01 2.53006518e-01 4.61397678e-01 9.53084305e-02
-2.93812633e-01 7.17752457e-01 -1.04550219e+00 1.06175756e+00
1.15870190e+00 -1.35343522e-01 -5.50503492e-01 1.65984198e-01
3.20208639e-01 -5.59232049e-02 2.12580502e-01 -1.17165005e+00
-1.02558181e-01 -2.03572270e-02 5.60704410e-01 -4.35890108e-01
6.79297149e-01 -1.11895335e+00 1.50045171e-01 7.57827520e-01
3.55954260e-01 1.05683975e-01 6.01720326e-02 4.82676804e-01
4.55762120e-03 4.77083065e-02 6.37054086e-01 3.15126628e-01
-2.11222321e-01 1.37804756e-02 -2.88627505e-01 -4.04522985e-01
1.18934679e+00 -4.31891859e-01 -1.94513202e-01 -4.00180280e-01
-3.96088988e-01 -1.67323828e-01 5.69728911e-01 4.20123279e-01
6.08940244e-01 -1.36865425e+00 1.24759421e-01 5.64329386e-01
1.01085819e-01 -3.97719145e-01 1.26554579e-01 7.58228779e-01
-1.28590092e-01 4.03321832e-01 -2.67090976e-01 -4.96742964e-01
-1.37811124e+00 2.91261554e-01 -2.41463259e-01 -3.16922188e-01
-2.78299332e-01 3.04988831e-01 -6.94473267e-01 -2.06340089e-01
1.14038236e-01 -2.09079072e-01 -4.62784916e-01 2.64238209e-01
9.80393529e-01 7.60934412e-01 3.31914842e-01 -8.85302067e-01
-6.04630768e-01 4.52279359e-01 1.58658743e-01 -8.13950524e-02
1.11560166e+00 -2.90749997e-01 2.87120771e-02 4.37026739e-01
7.90965319e-01 5.27978480e-01 -6.94534481e-01 2.24608704e-01
1.89403042e-01 -7.45248556e-01 -2.64129639e-01 -4.37864602e-01
-6.45766735e-01 9.16963041e-01 8.44227493e-01 7.52080500e-01
1.18264568e+00 -6.09384179e-01 1.17761064e+00 1.97826788e-01
9.53724921e-01 -9.73119557e-01 -1.43690899e-01 3.60227302e-02
1.34209963e-03 -1.31013882e+00 1.57504588e-01 -3.70848179e-01
-2.56607264e-01 6.61800086e-01 3.36025178e-01 2.32927963e-01
8.55733097e-01 1.13195948e-01 -1.65578991e-01 2.60843992e-01
-1.31703198e-01 2.94243731e-02 7.96677992e-02 9.12213564e-01
3.33102107e-01 2.68961251e-01 -6.02827728e-01 1.14175010e+00
-3.07579011e-01 1.39341831e-01 3.38505328e-01 9.81652379e-01
-6.17323756e-01 -1.40932751e+00 -6.17440164e-01 6.49358332e-01
-8.72896969e-01 3.40581179e-01 -1.31322637e-01 6.18746877e-01
1.09086663e-01 1.46632373e+00 -4.59533602e-01 -9.91669059e-01
1.81852818e-01 1.23379827e-01 -5.03809899e-02 -3.23985964e-01
-2.50035971e-01 -2.32215211e-01 1.24213763e-01 -3.91034126e-01
-2.97272265e-01 -6.79198146e-01 -1.01029098e+00 -6.11904442e-01
-2.32517302e-01 1.72830746e-01 7.50082612e-01 9.36073780e-01
3.46556067e-01 2.97140390e-01 8.96531820e-01 -8.87098253e-01
-3.98536921e-01 -7.64956474e-01 -4.99386907e-01 1.52957097e-01
1.65932968e-01 -9.22941685e-01 -7.29412511e-02 -2.16096491e-01] | [13.817909240722656, 1.4381335973739624] |
a025def9-3371-4f14-9e27-68a7d27950c0 | optic-disc-cup-and-fovea-detection-from | null | null | https://doi.org/10.1007/978-3-030-63419-3_10 | https://link.springer.com/content/pdf/10.1007%2F978-3-030-63419-3.pdf | Optic Disc, Cup and Fovea Detection from Retinal Images Using U-Net++ with EfficientNet Encoder | The accurate detection of retinal structures like an optic disc (OD), cup, and fovea is crucial for the analysis of Age-related Macular Degeneration (AMD), Glaucoma, and other retinal conditions. Most segmentation methods rely on separate detection of these retinal structures due to which a combined analysis for computer-aided ophthalmic diagnosis and screening is challenging. To address this issue, the paper introduces an approach incorporating OD, cup, and fovea analysis together. The paper presents a novel method for the detection of OD with a cup and fovea using modified U-Net++ architecture with the EfficientNet-B4 model as a backbone. The extracted features from the EfficientNet are utilized using skip connections in U-Net++ for precise segmentation. Datasets from ADAM and REFUGE challenges are used for evaluating the performance. The proposed method achieved a success rate of 94.74% and 95.73% dice value for OD segmentation on ADAM and REFUGE data, respectively. For fovea detection, the average Euclidean distance of 26.17 pixels is achieved for the ADAM dataset. The proposed method stood first for OD detection and segmentation tasks in ISBI ADAM 2020 challenge. | ['Nitin Singhal', 'Pranab Samanta', 'Ravi Kamble'] | 2020-11-20 | null | null | null | null | ['optic-disc-detection', 'optic-cup-segmentation', 'fovea-detection', 'optic-cup-detection'] | ['medical', 'medical', 'medical', 'medical'] | [-1.57298267e-01 4.98812497e-02 1.84649780e-01 -1.98812857e-01
-4.12689716e-01 -1.61265105e-01 1.15956567e-01 -8.52623209e-02
-7.49888897e-01 6.07807994e-01 -1.85937986e-01 -4.24806178e-01
-1.02690637e-01 -4.69373584e-01 -2.16366559e-01 -5.52886546e-01
-1.94415078e-01 1.24518834e-01 6.66978359e-01 1.99765265e-01
4.95309174e-01 7.49091983e-01 -1.77930617e+00 2.20464677e-01
1.14625025e+00 1.27415574e+00 2.24856585e-01 1.01396585e+00
9.22023654e-02 2.61706263e-01 -4.41323549e-01 -3.86504769e-01
6.98814929e-01 -1.46499798e-01 -8.62691224e-01 9.36631858e-02
1.11666989e+00 -7.67398953e-01 1.34166270e-01 9.91995931e-01
1.04253972e+00 -2.26375207e-01 7.06056654e-01 -5.96222460e-01
8.13645348e-02 -1.57153115e-01 -9.67148483e-01 7.04400361e-01
-2.87649989e-01 4.66421545e-01 5.64146101e-01 -7.46019304e-01
6.99652195e-01 9.74313796e-01 6.21840656e-01 4.02346045e-01
-7.16917694e-01 -4.01183814e-01 -4.36063260e-01 4.10582185e-01
-1.23676980e+00 -3.43126327e-01 -7.59658962e-02 -7.75784373e-01
1.09787786e+00 8.56216550e-02 1.15342498e+00 -6.89798146e-02
1.18017286e-01 7.36215889e-01 1.17213440e+00 -3.52735370e-01
-5.20048523e-03 -2.64549196e-01 4.12872970e-01 8.93988609e-01
4.98525530e-01 1.18032135e-01 8.26956928e-02 1.85989872e-01
7.75212407e-01 -5.05114794e-01 -2.18285963e-01 1.74714342e-01
-7.70198822e-01 5.04868805e-01 7.72574604e-01 -2.78151810e-01
-5.37146688e-01 -1.00144468e-01 3.47325176e-01 -8.54497552e-02
3.02603900e-01 4.61322516e-01 -2.19254106e-01 -3.23420539e-02
-1.00980282e+00 2.60949045e-01 2.30739325e-01 6.42700434e-01
1.89969718e-01 -3.73235345e-01 -3.49866480e-01 8.65714192e-01
4.19957250e-01 1.91685334e-01 4.48452115e-01 -1.02032328e+00
2.49960691e-01 8.77627254e-01 1.72068149e-01 -4.45685238e-01
-6.59347951e-01 -5.55258036e-01 -6.48196936e-01 8.59238863e-01
8.42533946e-01 -5.77065825e-01 -1.62704444e+00 7.58921742e-01
5.82173884e-01 2.69885421e-01 -3.34993958e-01 1.19266498e+00
1.00523829e+00 1.96129397e-01 -2.38043204e-01 -3.22147086e-02
1.50253868e+00 -1.05042553e+00 -3.38164568e-01 -7.58600533e-02
7.24243283e-01 -9.67274666e-01 5.63131034e-01 3.01618665e-01
-1.37512875e+00 -4.68653738e-01 -1.01364183e+00 -4.67682779e-01
-1.49759933e-01 8.38231504e-01 4.40527409e-01 6.07026994e-01
-1.43634701e+00 2.86460280e-01 -8.28284979e-01 -4.49944258e-01
8.82789075e-01 7.65110731e-01 -1.33311585e-01 -9.42551196e-02
-3.85512263e-01 9.11448836e-01 3.08913141e-01 3.21332723e-01
-3.16129208e-01 -4.91371751e-01 -3.98361295e-01 -3.81557345e-01
-1.80048376e-01 -1.03414834e+00 1.11669481e+00 -5.19609451e-01
-1.15197384e+00 1.17742038e+00 -3.16723973e-01 -1.17547572e+00
8.44351172e-01 -3.44102502e-01 -1.11992188e-01 5.72873235e-01
-7.47453570e-02 1.00147462e+00 7.54938543e-01 -5.45776963e-01
-1.34279013e+00 -6.78222120e-01 -2.24976311e-03 1.46289960e-01
2.30919272e-01 3.76783907e-01 -5.59716284e-01 -3.87188345e-02
3.24451208e-01 -7.50310481e-01 -1.78160042e-01 4.58679736e-01
-6.22346759e-01 -2.26716772e-01 6.72406316e-01 -9.00340736e-01
1.21930134e+00 -1.92076719e+00 -4.57878977e-01 2.41335452e-01
8.89830232e-01 1.18284810e+00 -3.72644961e-02 -5.82553387e-01
4.18486074e-02 1.92244947e-02 3.43954451e-02 -3.00938100e-01
-7.52249002e-01 -2.88819075e-01 4.40265656e-01 5.67146063e-01
4.03034151e-01 7.60839522e-01 -5.35097361e-01 -5.86352468e-01
2.12404430e-01 5.98001838e-01 -4.83546823e-01 -1.14782691e-01
1.57976653e-02 1.86105043e-01 -1.54783279e-01 8.92875671e-01
7.83253133e-01 -1.82300016e-01 -2.62104690e-01 -3.88082474e-01
-4.83414888e-01 -5.11928499e-02 -1.14679253e+00 9.26024973e-01
-5.83566688e-02 1.03994286e+00 5.97163402e-02 -2.67127335e-01
7.27193654e-01 8.48091915e-02 2.78194487e-01 -6.15738571e-01
4.69937921e-01 7.59055376e-01 7.94792473e-01 -5.95011711e-01
2.07572132e-01 4.54068393e-01 1.01592660e+00 -1.84146594e-02
-2.54972190e-01 2.90402949e-01 6.25997782e-01 -1.49688318e-01
8.40457141e-01 -6.66575357e-02 4.27014232e-01 -2.59024128e-02
6.85107350e-01 -1.12091554e-02 4.98827904e-01 4.48103637e-01
-7.49626398e-01 9.45249259e-01 6.07357442e-01 -5.78831673e-01
-1.01336205e+00 -7.06924677e-01 -6.97922170e-01 -6.78199381e-02
4.35168110e-02 -2.94250280e-01 -8.28665853e-01 -5.01597404e-01
1.27301872e-01 -2.29402661e-01 -5.07153451e-01 3.44535619e-01
-4.21483964e-01 -8.96166503e-01 3.87258679e-01 3.21703404e-01
1.14089060e+00 -5.53611755e-01 -9.35949922e-01 1.32490337e-01
2.00744838e-01 -9.67800617e-01 -2.17185065e-01 -5.82460701e-01
-1.06522262e+00 -1.66531837e+00 -1.19183350e+00 -1.02637649e+00
7.23129749e-01 2.89557725e-01 6.02800965e-01 7.27457181e-02
-1.04800379e+00 -2.21793905e-01 1.55759454e-01 -5.78318059e-01
5.92660233e-02 -1.78817511e-01 -2.14503676e-01 1.43236354e-01
4.63669032e-01 -1.95771217e-01 -1.41916311e+00 3.92824858e-01
-2.66320974e-01 -6.86089173e-02 8.32488060e-01 5.51687300e-01
8.49182367e-01 -2.28206933e-01 2.68556297e-01 -4.48769450e-01
5.83447278e-01 -1.65397212e-01 -9.49174643e-01 -5.00171334e-02
-6.39120281e-01 -4.72168505e-01 -1.69715896e-01 -5.29454313e-02
-6.75877213e-01 2.29332745e-01 5.95110394e-02 -2.73744613e-01
-1.51932880e-01 1.84717894e-01 2.97702521e-01 -6.44720316e-01
8.34572375e-01 -2.92668939e-01 5.88450432e-01 -5.25704801e-01
6.72843084e-02 1.22310758e+00 6.29831016e-01 1.30063653e-01
2.68388651e-02 7.19464421e-01 3.33074898e-01 -9.40771580e-01
-6.97495282e-01 -9.17725205e-01 -5.59755862e-01 -2.71435618e-01
1.01012993e+00 -7.93194592e-01 -6.36047542e-01 1.15866041e+00
-1.30049872e+00 -1.22251371e-02 -6.36038929e-02 7.84625947e-01
-1.11952476e-01 3.63396794e-01 -4.30834055e-01 -5.93615830e-01
-7.77089059e-01 -1.48383677e+00 7.58193433e-01 7.80981779e-01
3.64469215e-02 -5.42479455e-01 -3.21411908e-01 6.93729579e-01
3.86727303e-01 2.43648082e-01 6.97368324e-01 -2.48838469e-01
-6.37304664e-01 -3.55426580e-01 -9.73499894e-01 7.27670550e-01
-1.00871678e-02 5.64137995e-01 -8.47855926e-01 -6.43432327e-03
-4.24260676e-01 1.79656222e-01 1.09194088e+00 1.13948464e+00
8.16050529e-01 1.05193265e-01 -2.64556676e-01 7.85959780e-01
1.42471671e+00 2.47591063e-01 9.41991568e-01 4.47318912e-01
4.61277634e-01 6.21670425e-01 5.89407086e-01 2.98034728e-01
2.54875124e-01 5.14571905e-01 7.49162316e-01 -3.38450283e-01
-8.15025568e-01 6.92689419e-01 -1.51213542e-01 5.84359430e-02
-4.54716563e-01 1.52367398e-01 -1.27307105e+00 9.66056406e-01
-1.42981088e+00 -4.85891819e-01 -8.21354210e-01 2.30647254e+00
6.95026696e-01 1.05396463e-02 3.26970428e-01 -1.19709149e-01
9.25964177e-01 -5.43057561e-01 -6.67409241e-01 -1.83417246e-01
-1.67070925e-01 4.89810914e-01 7.83854187e-01 4.30204451e-01
-1.35127699e+00 7.42162943e-01 5.69274092e+00 5.38099527e-01
-1.20537388e+00 -1.18902802e-01 6.57147765e-01 -5.50474703e-01
7.80133009e-01 -2.96043664e-01 -1.31267273e+00 4.05528098e-01
6.52008533e-01 1.68164462e-01 -1.44454017e-01 2.27242678e-01
5.34610212e-01 -6.31690800e-01 -4.45284784e-01 1.17045438e+00
-2.79213011e-01 -1.45057499e+00 -7.80627877e-03 4.10011739e-01
7.37507701e-01 5.30256033e-01 1.32994831e-01 -5.06577075e-01
-2.85670102e-01 -1.11581218e+00 -1.27393663e-01 7.35105455e-01
1.00392079e+00 -6.88793957e-01 1.19092906e+00 -2.40925372e-01
-7.54706740e-01 -1.28290519e-01 -3.53877574e-01 1.13177739e-01
-8.88168961e-02 5.28886735e-01 -1.21466172e+00 3.59821580e-02
7.38464117e-01 8.96870911e-01 -8.82555068e-01 2.48774505e+00
-5.99055886e-02 5.17837346e-01 -5.32221258e-01 3.69687408e-01
3.83514404e-01 -4.15792108e-01 9.98874426e-01 7.71683693e-01
3.86157691e-01 -2.11968884e-01 -3.87075841e-01 7.34734774e-01
-1.31212389e-02 3.10409069e-01 -2.20646515e-01 2.67649949e-01
2.78564900e-01 1.27327585e+00 -6.09679878e-01 -1.36974111e-01
-4.36032891e-01 3.07276785e-01 -2.30692163e-01 3.03615928e-01
-4.40410614e-01 -5.47594190e-01 9.04783607e-01 6.05161130e-01
1.83287129e-01 -1.11846134e-01 -4.95530456e-01 -4.47551697e-01
2.41550639e-01 -6.02086365e-01 2.21546426e-01 -7.58639872e-01
-7.38085449e-01 5.31005740e-01 -4.57690060e-01 -1.45901203e+00
1.25486255e-01 -9.68509972e-01 -7.97337532e-01 1.34414780e+00
-1.82615614e+00 -9.69860733e-01 -6.47740901e-01 4.11754578e-01
1.52879640e-01 -5.88039696e-01 2.29272202e-01 3.93434763e-01
-9.12494779e-01 3.56752694e-01 1.08336218e-01 2.36940205e-01
6.97314560e-01 -1.34318125e+00 3.91722888e-01 1.10469234e+00
-6.58779204e-01 5.77957809e-01 1.50334015e-01 -7.73536980e-01
-6.28624678e-01 -1.30230093e+00 6.62257254e-01 -3.49013843e-02
2.94142187e-01 6.71610236e-01 -5.98804653e-01 3.07039857e-01
4.36771177e-02 3.83917332e-01 4.13110256e-01 -3.28959167e-01
1.01251915e-01 -1.14586979e-01 -1.29442430e+00 6.18582487e-01
6.80650592e-01 3.57964300e-02 -2.02706844e-01 3.11890423e-01
2.89613694e-01 -7.66078472e-01 -8.15747023e-01 3.57441664e-01
6.19958818e-01 -1.28026247e+00 8.97948623e-01 -5.95398784e-01
2.63624847e-01 -4.57889557e-01 2.79896915e-01 -7.09443808e-01
2.45269060e-01 -7.97083080e-01 -2.49584809e-01 6.97834969e-01
4.63670701e-01 -8.70337248e-01 9.12861884e-01 4.41631287e-01
-2.87513345e-01 -9.28771317e-01 -9.30428028e-01 -4.14427698e-01
-1.44546524e-01 -1.86777294e-01 1.88012525e-01 2.06152484e-01
-9.22440767e-01 -6.11215644e-02 2.59865135e-01 2.75932640e-01
7.28152573e-01 -2.06228256e-01 7.20734417e-01 -1.50395799e+00
2.76186228e-01 -7.14527726e-01 -1.14399588e+00 -8.23278785e-01
-6.50811434e-01 -6.88435256e-01 -4.52500403e-01 -1.90102088e+00
-2.19197810e-01 -1.20556429e-01 -1.25133008e-01 1.45414591e-01
-1.97843313e-02 4.99538541e-01 -1.12056814e-01 3.55240077e-01
3.78444628e-03 -8.98470283e-02 1.67683864e+00 -7.53973518e-03
-6.36813521e-01 4.49543804e-01 -3.38218153e-01 9.47676420e-01
9.24256325e-01 7.29779527e-02 -4.08129930e-01 -3.15574050e-01
-2.52396241e-02 -2.46374458e-01 8.00780773e-01 -1.31717801e+00
4.29950118e-01 5.54215908e-01 3.05087715e-01 -7.84357369e-01
2.48764724e-01 -3.33418518e-01 -5.21822214e-01 5.90133786e-01
2.20967054e-01 -4.56989020e-01 3.86145562e-01 4.01341379e-01
-3.46317887e-01 -9.04285461e-02 1.29781175e+00 -3.17062810e-02
-8.15500021e-01 5.15137553e-01 -1.17258553e-03 1.31929308e-01
1.24649096e+00 -7.47671247e-01 -7.40595222e-01 1.26118824e-01
-9.81711447e-01 5.46507716e-01 1.41190395e-01 2.45788619e-02
9.72987950e-01 -5.44318378e-01 -1.05243933e+00 3.04275125e-01
4.85937484e-02 3.64924610e-01 2.30306819e-01 1.57338524e+00
-1.25790393e+00 4.65464830e-01 -5.62605619e-01 -6.92266762e-01
-1.86732078e+00 -5.00131428e-01 1.17775917e+00 2.00572506e-01
-6.71717107e-01 1.04983342e+00 -2.81932950e-01 3.41425627e-01
3.10093701e-01 -6.00780308e-01 -6.25531197e-01 3.99102271e-01
8.05109441e-01 8.95540833e-01 2.87432104e-01 -4.99153763e-01
-9.65037271e-02 8.22081208e-01 -6.15804851e-01 2.47867614e-01
9.43054676e-01 -3.54446352e-01 -4.94269907e-01 -3.25308830e-01
8.95489693e-01 -2.98234910e-01 -1.11786377e+00 -2.52770215e-01
-1.56058356e-01 -5.30606151e-01 6.42463207e-01 -1.02275276e+00
-1.26279140e+00 1.09450233e+00 1.35564482e+00 -1.39412090e-01
1.02956700e+00 -3.85482728e-01 9.74388242e-01 5.19201159e-02
2.27485802e-02 -8.76145840e-01 -5.67688882e-01 3.09175074e-01
7.72758603e-01 -1.40408897e+00 6.41789986e-04 -4.64787364e-01
-3.61539364e-01 1.38934433e+00 8.72358680e-01 -1.85079381e-01
6.99054122e-01 -2.53138065e-01 2.27538496e-01 -2.73517907e-01
-1.97490275e-01 -9.15822685e-01 9.40571785e-01 8.21442187e-01
3.93057108e-01 -5.29765747e-02 -7.17056513e-01 9.93624553e-02
-8.61827284e-02 2.66432226e-01 8.60276937e-01 6.26501203e-01
-7.66453087e-01 -6.61995888e-01 -7.67082870e-02 1.20354986e+00
-6.78403199e-01 -2.73367196e-01 -4.55630153e-01 8.89019072e-01
5.17055035e-01 7.95524299e-01 3.34208995e-01 5.86961247e-02
1.02660462e-01 -2.35692650e-01 3.05606216e-01 -7.16831446e-01
-5.79025030e-01 1.41099229e-01 3.62625867e-01 -7.98701525e-01
-4.28949803e-01 -5.12889206e-01 -1.26097226e+00 7.83856027e-03
-7.26166442e-02 -3.68036151e-01 7.59519637e-01 6.65871263e-01
7.28904307e-01 2.26837963e-01 1.30794480e-01 -5.04528701e-01
-2.33391330e-01 -9.92030144e-01 -8.60552013e-01 -2.25054234e-01
6.80037439e-01 -5.43313622e-01 -2.40011245e-01 -6.70104250e-02] | [15.819473266601562, -3.9910285472869873] |
69709c54-1c7d-453c-95db-e86bb6b4cdbc | proto-program-guided-transformer-for-program | 2110.00804 | null | https://arxiv.org/abs/2110.00804v2 | https://arxiv.org/pdf/2110.00804v2.pdf | ProTo: Program-Guided Transformer for Program-Guided Tasks | Programs, consisting of semantic and structural information, play an important role in the communication between humans and agents. Towards learning general program executors to unify perception, reasoning, and decision making, we formulate program-guided tasks which require learning to execute a given program on the observed task specification. Furthermore, we propose the Program-guided Transformer (ProTo), which integrates both semantic and structural guidance of a program by leveraging cross-attention and masked self-attention to pass messages between the specification and routines in the program. ProTo executes a program in a learned latent space and enjoys stronger representation ability than previous neural-symbolic approaches. We demonstrate that ProTo significantly outperforms the previous state-of-the-art methods on GQA visual reasoning and 2D Minecraft policy learning datasets. Additionally, ProTo demonstrates better generalization to unseen, complex, and human-written programs. | ['Le Song', 'Binghong Chen', 'Karan Samel', 'Zelin Zhao'] | 2021-10-02 | null | http://proceedings.neurips.cc/paper/2021/hash/8d34201a5b85900908db6cae92723617-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/8d34201a5b85900908db6cae92723617-Paper.pdf | neurips-2021-12 | ['learning-to-execute'] | ['computer-code'] | [ 2.26421744e-01 1.32808924e-01 -5.20355225e-01 -4.77587521e-01
-2.00242892e-01 -6.45986915e-01 1.09090304e+00 3.68703246e-01
-1.03744909e-01 -6.47959113e-02 2.67952532e-01 -6.52010679e-01
2.45282978e-01 -1.01653624e+00 -1.18939435e+00 -1.99584424e-01
-6.31352980e-03 5.27124047e-01 1.64240524e-01 -1.11974046e-01
3.28493774e-01 1.55815735e-01 -1.48223066e+00 8.39365900e-01
8.06688130e-01 6.80479109e-01 3.75267982e-01 7.03647852e-01
-1.45199597e-01 1.69619954e+00 -2.66559571e-01 -8.19418356e-02
-2.83279151e-01 -1.94382772e-01 -9.41134512e-01 -1.35029778e-01
2.18677908e-01 -4.77047414e-01 -5.32938123e-01 1.08853018e+00
-2.61386693e-01 2.89295167e-01 6.37122571e-01 -1.48841703e+00
-1.22494876e+00 1.01832020e+00 -2.75264859e-01 -1.71157449e-01
7.02381849e-01 7.92013705e-01 1.22218144e+00 -4.80378419e-01
3.76261204e-01 1.75483060e+00 2.68747598e-01 8.03445876e-01
-1.63774991e+00 -4.15190339e-01 4.42924500e-01 2.54972801e-02
-7.84322202e-01 -3.26843560e-01 7.21735895e-01 -9.83810008e-01
1.61731029e+00 -1.67640015e-01 7.12764680e-01 1.40839314e+00
1.46435961e-01 1.09561718e+00 9.13526893e-01 -2.77019769e-01
3.34361613e-01 -2.97252461e-02 5.46019495e-01 1.49101269e+00
-1.50587365e-01 6.83059990e-01 -5.81358492e-01 -4.66130614e-01
8.71188104e-01 2.38134772e-01 -2.24008322e-01 -8.11619699e-01
-1.41026044e+00 8.19017947e-01 4.68626469e-01 -1.95561070e-02
-1.83761686e-01 7.55067647e-01 6.56385362e-01 4.85156894e-01
-1.31065384e-01 8.25652063e-01 -3.93816084e-01 -1.86641693e-01
-5.67156017e-01 4.03079122e-01 9.06595051e-01 1.15536642e+00
7.32118845e-01 3.69146913e-01 -5.11330366e-01 1.32383883e-01
5.06061196e-01 6.44111812e-01 3.74534041e-01 -1.03607798e+00
4.72613066e-01 1.17351556e+00 -9.00017321e-02 -9.40263033e-01
-7.55357072e-02 2.10694596e-02 -3.84602278e-01 3.18507940e-01
3.29834968e-02 2.48799458e-01 -8.34191144e-01 1.75044739e+00
-1.45860687e-01 6.11387640e-02 4.04544264e-01 7.51407266e-01
9.64351118e-01 7.19630003e-01 3.19327146e-01 3.05664659e-01
1.28572738e+00 -1.50037801e+00 -1.78346679e-01 -5.59590757e-01
5.91025352e-01 7.64663666e-02 1.62727737e+00 9.57909301e-02
-9.23471093e-01 -9.02206957e-01 -9.19045210e-01 -9.37808529e-02
-2.56118625e-01 2.65838057e-01 1.05317068e+00 7.54418075e-02
-1.08878279e+00 5.04355967e-01 -1.29053724e+00 -1.70715392e-01
6.16367519e-01 2.79710889e-01 -1.97109923e-01 1.77537367e-01
-5.33792436e-01 5.49158096e-01 6.54898703e-01 -4.21800166e-01
-1.93840015e+00 -6.37857199e-01 -1.50007010e+00 3.33523482e-01
7.20833361e-01 -1.05624950e+00 1.73198009e+00 -1.03045189e+00
-1.94741631e+00 1.06221688e+00 -2.77338058e-01 -6.76108003e-01
5.00682928e-02 -2.17209518e-01 -1.65620502e-02 1.02592967e-01
-1.06562942e-01 5.65264940e-01 1.11972237e+00 -1.02827632e+00
-3.62960726e-01 -2.85390168e-01 8.15464795e-01 -1.02450132e-01
2.51717895e-01 1.47677846e-02 -2.82920092e-01 -4.63604480e-01
-2.69808084e-01 -9.32842612e-01 4.73675616e-02 -1.99504178e-02
-3.12255204e-01 -6.95335388e-01 7.37883031e-01 -4.75011736e-01
8.81858885e-01 -2.40281367e+00 7.48384655e-01 -6.93876967e-02
3.94895583e-01 -2.79543344e-02 -1.92570031e-01 3.63417745e-01
2.84525976e-02 -1.27141222e-01 -1.07855819e-01 1.88710243e-02
7.06763804e-01 4.47819531e-01 -1.00460219e+00 3.78844440e-01
2.78307170e-01 1.42040169e+00 -1.15198529e+00 -2.56361365e-01
1.00805789e-01 -6.23493344e-02 -9.59450543e-01 8.44851077e-01
-1.26661360e+00 4.36286926e-01 -8.92462969e-01 7.01084375e-01
-1.65688023e-01 -7.30496466e-01 2.56563067e-01 7.57282823e-02
1.18884608e-01 5.21659493e-01 -4.41609561e-01 2.04437995e+00
-5.84405482e-01 4.73919690e-01 -7.34341070e-02 -1.21791458e+00
7.74097264e-01 2.05493495e-01 -3.67409825e-01 -6.18997931e-01
-1.02066867e-01 -2.34639883e-01 -1.14470668e-01 -8.01546276e-01
2.16008052e-01 3.71957093e-01 -4.31871444e-01 7.85071194e-01
2.34785661e-01 -2.94103682e-01 -1.84644554e-02 4.05843347e-01
1.13185215e+00 8.04841399e-01 4.59174424e-01 -1.66186899e-01
4.96035218e-01 1.55295447e-01 1.72147334e-01 1.11976862e+00
-7.81717151e-02 -3.27079266e-01 1.00187755e+00 -5.99526346e-01
-7.77610600e-01 -7.87308097e-01 7.76768267e-01 1.81684220e+00
-6.29860088e-02 -4.76696789e-01 -5.48318386e-01 -8.95714343e-01
2.83651441e-01 1.01580501e+00 -7.91318953e-01 -4.74150300e-01
-6.70403242e-01 1.68393895e-01 6.43542647e-01 8.14413309e-01
4.81893033e-01 -1.44615066e+00 -1.13857317e+00 -1.09109238e-01
1.36281475e-01 -9.76152420e-01 -5.06709397e-01 1.77609220e-01
-7.02867270e-01 -1.26236844e+00 -4.05375212e-02 -8.60189021e-01
8.91050577e-01 2.58797240e-02 1.49927247e+00 2.02425703e-01
2.20702648e-01 8.84845912e-01 -1.00404926e-01 -3.15411128e-02
-7.48154640e-01 2.39309594e-02 -3.08142811e-01 -4.04302686e-01
1.88375130e-01 -4.19173747e-01 -5.02265245e-03 -1.23354122e-01
-4.49335188e-01 5.01941144e-01 4.69774693e-01 8.98396075e-01
2.08065212e-01 -4.61422950e-02 -1.47062927e-01 -9.65548754e-01
6.95705354e-01 -2.82082200e-01 -1.20176685e+00 4.26050544e-01
-3.41376692e-01 6.93416417e-01 6.94351673e-01 -4.15313452e-01
-1.15936911e+00 2.67446581e-02 5.36453784e-01 -6.26192153e-01
-4.85693067e-01 5.40593863e-01 -7.40544721e-02 9.27843377e-02
7.53890812e-01 5.40805399e-01 -3.57987992e-02 -1.06390975e-01
6.68941617e-01 4.46217619e-02 7.34535813e-01 -1.50674295e+00
7.54573226e-01 3.38278890e-01 -1.66065082e-01 -3.49003732e-01
-5.87725878e-01 -9.80527475e-02 -2.60551840e-01 3.84878069e-01
1.00522494e+00 -8.78663123e-01 -1.37920451e+00 3.09411258e-01
-1.50383198e+00 -1.11364949e+00 -9.18163806e-02 1.70631837e-02
-1.16369438e+00 -5.33971675e-02 -6.61788940e-01 -6.15742326e-01
-4.68079560e-02 -1.73609531e+00 1.22458482e+00 -6.30097911e-02
-4.70872104e-01 -9.04428542e-01 1.63772345e-01 2.22854484e-02
2.09968045e-01 9.41340029e-02 1.71562397e+00 -7.11915612e-01
-1.03193676e+00 3.28941375e-01 -2.66110688e-01 -3.91317625e-03
4.17184122e-02 7.22564161e-02 -8.17580819e-01 -3.58554065e-01
-1.17656462e-01 -8.07689250e-01 7.11160302e-01 6.57164529e-02
1.42330921e+00 -6.51468813e-01 -5.57375550e-01 8.22865248e-01
1.08229983e+00 3.87740642e-01 5.69136962e-02 1.76137045e-01
6.53181016e-01 5.23847580e-01 2.60037869e-01 2.05099210e-01
4.65446979e-01 4.15329427e-01 5.45635283e-01 3.28053176e-01
1.01626679e-01 -7.43414283e-01 7.50692904e-01 1.50385261e-01
1.21068843e-01 2.53805906e-01 -1.16463232e+00 3.37589324e-01
-2.15025735e+00 -1.10203874e+00 5.48482656e-01 1.57635915e+00
8.78831863e-01 6.05423488e-02 1.89427033e-01 -4.58710760e-01
2.98978060e-01 3.92676175e-01 -8.32220614e-01 -4.30298716e-01
4.43964452e-01 2.05225989e-01 -7.97882825e-02 5.33725202e-01
-1.02065027e+00 1.20410788e+00 6.48924446e+00 2.32834294e-01
-1.09096718e+00 -1.25846416e-01 9.87198427e-02 3.49439353e-01
-3.59260768e-01 1.10197142e-01 -5.49173415e-01 1.48823306e-01
8.86031210e-01 -3.73915225e-01 9.71081495e-01 1.42466688e+00
-2.37818673e-01 8.90118331e-02 -2.03174376e+00 6.65827453e-01
8.21748674e-02 -1.56464958e+00 1.77556977e-01 -3.62644404e-01
6.35781586e-01 2.03407947e-02 3.00169826e-01 1.04491830e+00
1.09053802e+00 -1.16698039e+00 9.89925742e-01 5.82289159e-01
5.08045018e-01 -3.53518218e-01 7.66928270e-02 4.88205075e-01
-1.07296002e+00 -3.29812557e-01 -5.07488735e-02 -2.47822538e-01
-5.30354440e-01 -1.96152776e-01 -6.24596179e-01 1.78735375e-01
5.02390921e-01 1.04210043e+00 -4.26076710e-01 2.62400568e-01
-9.05883253e-01 3.80798340e-01 3.22078586e-01 -1.53211415e-01
4.62864131e-01 6.79022670e-02 4.47995186e-01 9.49095488e-01
-2.62757540e-02 2.43087597e-02 5.91136158e-01 1.94773471e+00
7.33230859e-02 -5.96701205e-01 -9.83279288e-01 -8.02339375e-01
9.89288315e-02 5.82053185e-01 -2.99604625e-01 -7.35396802e-01
-6.68637991e-01 8.77427697e-01 6.99874341e-01 8.60586107e-01
-9.73899126e-01 -5.43568842e-03 6.43284380e-01 -4.32267785e-01
5.44091702e-01 -4.23697829e-01 1.10022143e-01 -1.27398515e+00
-1.65395796e-01 -1.46620572e+00 2.63293087e-01 -8.93618941e-01
-7.64150679e-01 5.02622545e-01 2.36438081e-01 -6.33805811e-01
-4.72828507e-01 -8.45732510e-01 -6.31620407e-01 6.73855066e-01
-1.29946733e+00 -1.25954354e+00 -1.59723058e-01 6.48117602e-01
5.24976790e-01 -7.26859927e-01 1.12679851e+00 -3.90232593e-01
-3.82108390e-01 2.84353465e-01 -6.42658830e-01 2.57693738e-01
1.00558311e-01 -1.21095824e+00 6.70982897e-01 6.32256567e-01
2.00440273e-01 1.20824575e+00 5.05630851e-01 -6.37244046e-01
-2.05707908e+00 -1.17093432e+00 2.41439387e-01 -6.79577827e-01
9.60007250e-01 -3.56651902e-01 -8.62317502e-01 1.36746264e+00
3.93676937e-01 -1.31628543e-01 2.93978393e-01 1.50662556e-01
-1.07906115e+00 2.77649522e-01 -6.50261939e-01 6.49465680e-01
9.12526190e-01 -1.38520360e+00 -1.28890049e+00 1.92356586e-01
1.25435400e+00 -6.79975748e-01 -3.42290014e-01 1.16437860e-01
4.32305813e-01 -8.19025099e-01 8.78024340e-01 -1.15667093e+00
6.91928327e-01 -4.09595132e-01 -2.82833159e-01 -1.18418109e+00
-2.76855201e-01 -7.25084126e-01 -6.14040077e-01 7.98617125e-01
1.21542357e-01 -4.85309869e-01 6.72413230e-01 4.84686166e-01
-2.83422500e-01 -4.97559100e-01 -1.56831026e-01 -5.45632362e-01
-1.21107966e-01 -4.13697928e-01 8.86490047e-01 8.10518980e-01
4.40094441e-01 3.42142940e-01 -6.00306243e-02 4.22125548e-01
5.16258001e-01 1.00495708e+00 1.01734281e+00 -9.91724014e-01
-1.06406546e+00 -7.51134455e-01 3.60450521e-02 -1.44027662e+00
1.14431036e+00 -1.41415668e+00 4.99560349e-02 -1.18674767e+00
3.77515018e-01 4.61957455e-02 -6.52624592e-02 1.01088262e+00
1.08036302e-01 -8.16149950e-01 8.65193829e-02 2.23459765e-01
-7.52680540e-01 6.39952183e-01 1.20563293e+00 -7.55211592e-01
-1.68171510e-01 -9.17774513e-02 -5.84722459e-01 9.21482623e-01
4.40799892e-01 -3.95696133e-01 -7.18364954e-01 -8.39798570e-01
3.82437110e-01 4.29960936e-01 9.17583108e-01 -7.93865025e-01
4.15587336e-01 -6.94569767e-01 -3.01197022e-01 -2.65886962e-01
8.05238709e-02 -7.83108115e-01 -2.53395081e-01 8.15892935e-01
-9.59638417e-01 1.62273496e-01 3.08320135e-01 7.18383551e-01
-1.65888891e-01 2.78571304e-02 4.36005414e-01 -6.19282663e-01
-1.25860560e+00 3.10641408e-01 -4.01723891e-01 -9.64212194e-02
9.10393953e-01 3.02783370e-01 -7.15092123e-01 -1.62271857e-01
-7.13445246e-01 5.23379683e-01 6.61791384e-01 4.41377938e-01
6.27910018e-01 -1.05411720e+00 -1.42033780e-02 5.18496454e-01
5.32492220e-01 -1.12244517e-01 -9.60929394e-02 5.39386332e-01
-3.76211673e-01 6.90301716e-01 -3.22088510e-01 -6.67050421e-01
-1.17909825e+00 9.85699236e-01 3.58702034e-01 -1.25556052e-01
-6.24964476e-01 8.72106552e-01 6.75554931e-01 -8.21601868e-01
5.42336464e-01 -1.04313076e+00 5.75475432e-02 -6.66857123e-01
4.33467776e-01 -1.33441329e-01 -6.58914030e-01 -1.43742546e-01
-2.30825767e-01 3.02970856e-01 -3.04849818e-02 1.27239525e-01
1.16778576e+00 6.32699549e-01 -6.34959698e-01 5.51199734e-01
7.25161195e-01 -4.59692925e-01 -1.44516790e+00 -6.23474896e-01
9.76970121e-02 -2.17030913e-01 -1.52786061e-01 -5.98258495e-01
-7.09380269e-01 1.18412757e+00 -6.65320680e-02 7.47069390e-03
5.24433434e-01 3.96943122e-01 1.40671447e-01 1.06673634e+00
5.50052702e-01 -5.49176931e-01 8.28899026e-01 8.16504896e-01
9.97390926e-01 -1.23290300e+00 -2.87642300e-01 -5.82972132e-02
-6.15176260e-01 1.05911362e+00 1.02838624e+00 -2.87814617e-01
1.27502874e-01 3.03767174e-01 -5.66191018e-01 -4.22187984e-01
-1.29800260e+00 -1.45116039e-02 2.90013224e-01 7.46936440e-01
2.07805380e-01 7.20475614e-02 7.29976892e-01 7.86150515e-01
8.15231353e-02 2.79253989e-01 9.51884463e-02 1.13714457e+00
-2.89956212e-01 -7.75541365e-01 -1.79929845e-02 1.31599769e-01
7.88749978e-02 -1.15932114e-01 -2.30048120e-01 7.50023782e-01
-1.00473166e-01 4.81489748e-01 7.09063262e-02 -2.61234254e-01
2.85463363e-01 2.71131009e-01 7.41968930e-01 -1.08130801e+00
-4.48699683e-01 -3.68543625e-01 -1.08571552e-01 -1.15380049e+00
-4.37557727e-01 -3.45629960e-01 -1.48216808e+00 -2.12650821e-01
3.41857940e-01 -2.17367485e-02 5.49426675e-02 8.93447459e-01
4.22183543e-01 9.67256188e-01 5.00806905e-02 -6.40948117e-01
-9.21553254e-01 -3.86179745e-01 -1.22283943e-01 3.07027668e-01
7.68260598e-01 -5.70375025e-01 -8.30485746e-02 3.28625917e-01] | [8.602160453796387, 7.105854511260986] |
e193d5ed-3e42-4ae9-a689-dfed9124c307 | hand-segmentation-and-fingertip-tracking-from | 1901.03465 | null | https://arxiv.org/abs/1901.03465v3 | https://arxiv.org/pdf/1901.03465v3.pdf | Hand Segmentation and Fingertip Tracking from Depth Camera Images Using Deep Convolutional Neural Network and Multi-task SegNet | Hand segmentation and fingertip detection play an indispensable role in hand gesture-based human-machine interaction systems. In this study, we propose a method to discriminate hand components and to locate fingertips in RGB-D images. The system consists of three main steps: hand detection using RGB images providing regions which are considered as promising areas for further processing, hand segmentation, and fingertip detection using depth image and our modified SegNet, a single lightweight architecture that can process two independent tasks at the same time. The experimental results show that our system is a promising method for hand segmentation and fingertip detection which achieves a comparable performance while model complexity is suitable for real-time applications. | ['Soo-Hyung Kim', 'In-Seop Na', 'Tai Nhu Do', 'Duong Hai Nguyen'] | 2019-01-11 | null | null | null | null | ['hand-detection', 'fingertip-detection', 'hand-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.52814046e-01 -5.05948126e-01 -2.95057148e-01 -1.48215130e-01
-5.07491827e-02 -7.40972936e-01 3.22650850e-01 -4.01073426e-01
-8.10714960e-01 1.34517372e-01 -2.92907983e-01 -4.61690903e-01
-4.95193452e-02 -6.32192612e-01 1.57945171e-01 -6.50230467e-01
1.74518332e-01 6.74584627e-01 8.77075613e-01 -8.86095315e-02
6.00880146e-01 1.51782763e+00 -1.45486712e+00 1.02508746e-01
7.97487944e-02 9.50658619e-01 2.68830568e-01 1.04801178e+00
-2.68710554e-01 2.16321312e-02 -4.77309912e-01 1.12556957e-01
5.55475831e-01 -1.89223737e-01 -7.10537076e-01 2.92911497e-03
2.28498727e-01 -9.20616150e-01 -2.57634401e-01 7.23732948e-01
1.16960609e+00 4.14565913e-02 4.33482885e-01 -9.92753386e-01
2.01671883e-01 2.45892942e-01 -6.30511343e-01 1.93259656e-01
4.98315811e-01 4.39927906e-01 4.44544226e-01 -7.45339692e-01
4.34601218e-01 1.31630707e+00 1.98025420e-01 9.44869876e-01
-5.49810052e-01 -4.51478899e-01 1.23693176e-01 1.54659718e-01
-1.48468244e+00 4.34679091e-02 7.48908877e-01 -2.06467628e-01
8.88058424e-01 6.59965456e-01 6.65479898e-01 7.00236201e-01
-7.16464221e-02 1.18890476e+00 1.00414371e+00 -6.19317830e-01
1.29216358e-01 -1.18365780e-01 4.42458391e-01 6.35795951e-01
8.98190886e-02 2.98839621e-02 -2.68566340e-01 1.72385752e-01
1.65849841e+00 2.99850196e-01 -3.71488780e-02 1.26808703e-01
-1.25214303e+00 4.06607203e-02 5.52278042e-01 7.53572762e-01
-7.31745243e-01 1.97418943e-01 4.18374203e-02 -1.08260103e-01
-1.33753896e-01 -1.71347365e-01 -1.96051359e-01 -3.41076910e-01
-9.97977912e-01 2.87510067e-01 7.18148470e-01 9.50627685e-01
-1.09686395e-02 -5.38456440e-01 -4.79811072e-01 7.22489133e-02
5.63248694e-01 6.01273417e-01 3.51657234e-02 -4.99096453e-01
6.34549260e-01 8.28686476e-01 4.15377885e-01 -4.67026830e-01
-6.19128466e-01 2.70021290e-01 -5.48963547e-01 8.09899330e-01
9.54893112e-01 8.76622871e-02 -1.22341764e+00 5.44786572e-01
3.47385049e-01 -2.54713029e-01 -7.43342102e-01 1.39210057e+00
8.81360769e-01 1.10282816e-01 1.97696790e-01 3.61470312e-01
1.52593064e+00 -7.10976303e-01 -6.27658725e-01 9.27386433e-02
-2.48535767e-01 -9.35775340e-01 1.28765047e+00 7.74387479e-01
-1.07168508e+00 -6.95739329e-01 -6.22139394e-01 -1.27789602e-01
-5.98587751e-01 5.98116577e-01 6.97729528e-01 1.22022843e+00
-8.28978896e-01 4.79171962e-01 -1.08327615e+00 -3.61090451e-01
2.28993416e-01 7.52888799e-01 2.54772534e-03 3.40227515e-01
-4.77497727e-01 4.68849361e-01 3.98268551e-01 6.21989012e-01
-1.24712057e-01 1.16651371e-01 2.78882980e-02 -1.20290101e-01
1.41149893e-01 -2.48861045e-01 1.00355840e+00 -1.84273437e-01
-1.76437306e+00 1.00485229e+00 -4.94893640e-01 2.47517496e-01
8.58547807e-01 -2.21589774e-01 -8.43218267e-02 3.47409159e-01
-5.07115364e-01 3.43962133e-01 9.62624192e-01 -9.77097988e-01
-8.54902208e-01 -1.05633795e+00 -1.34281859e-01 1.39873639e-01
-2.66670018e-01 4.85337794e-01 -1.05411136e+00 -3.63089174e-01
7.16185868e-01 -7.42216170e-01 3.31269391e-02 -1.22993113e-02
-6.59878135e-01 -8.06836903e-01 9.03566301e-01 -9.49052155e-01
1.35212994e+00 -1.45960438e+00 7.60034174e-02 8.79231513e-01
2.46133506e-01 7.15752840e-01 5.03006689e-02 6.49825856e-02
4.80711430e-01 1.92081649e-02 1.78602979e-01 -2.77438134e-01
-5.81479520e-02 -3.38229723e-02 1.79398477e-01 9.58564356e-02
-1.06753267e-01 1.03023338e+00 -6.75120711e-01 -7.58621752e-01
8.67370069e-01 7.03452468e-01 2.15037465e-01 2.63948470e-01
3.35118249e-02 7.58035362e-01 -8.14734817e-01 1.36841106e+00
6.93833292e-01 4.30436432e-01 6.87284470e-02 -2.92617619e-01
-3.50489736e-01 -1.47796735e-01 -1.58723366e+00 1.52364933e+00
-1.14555307e-01 3.95989001e-01 3.14450145e-01 -1.68186456e-01
8.45900774e-01 3.20820719e-01 3.54082525e-01 -5.23074687e-01
6.87640250e-01 4.22460139e-01 -1.38497487e-01 -7.60314584e-01
2.96510726e-01 4.59563881e-01 5.37939310e-01 7.15269446e-01
-3.66780907e-01 -1.15443366e-02 -4.57797647e-02 -3.81348222e-01
6.14903033e-01 4.00554806e-01 -4.87046279e-02 1.66148886e-01
5.54122627e-01 -1.13816150e-01 -1.79041654e-01 7.78450727e-01
-5.26479006e-01 5.65292478e-01 8.78162086e-02 -5.13562858e-01
-7.72730708e-01 -9.24408197e-01 5.65818809e-02 1.06025088e+00
2.38443583e-01 2.13934898e-01 -1.04243052e+00 -6.04257524e-01
-1.42391488e-01 -1.29632294e-01 -4.02989149e-01 7.89170027e-01
-9.92572188e-01 -3.63648355e-01 6.08816445e-01 1.04837394e+00
7.64367819e-01 -1.36736465e+00 -1.35966170e+00 1.14336058e-01
1.59109488e-01 -9.13916707e-01 -2.78384238e-01 -6.18978702e-02
-1.14271677e+00 -1.37507749e+00 -1.54078174e+00 -8.52553606e-01
6.39405072e-01 2.45797619e-01 6.21433556e-01 3.13006729e-01
-7.80298233e-01 4.32578981e-01 -4.80245113e-01 -3.43544960e-01
2.28088215e-01 3.77782732e-01 -2.43116811e-01 -1.77117765e-01
6.86030269e-01 -3.22238743e-01 -1.14285600e+00 4.72363144e-01
-5.74660122e-01 -4.20404822e-01 6.20533288e-01 1.15943462e-01
4.99692708e-01 -4.36793566e-01 -3.10250282e-01 -1.64399669e-01
8.98691058e-01 3.22701454e-01 -4.55922604e-01 5.68596184e-01
-3.07461828e-01 -1.85370162e-01 -4.14834768e-02 -2.05783412e-01
-8.62920403e-01 5.34018695e-01 -3.75239253e-01 -1.60499290e-01
-7.41241693e-01 -2.99762577e-01 -2.03109413e-01 -4.97007370e-01
4.84157652e-01 2.19702080e-01 -9.02653858e-02 -1.08802199e+00
3.69755328e-01 1.35397923e+00 5.59821010e-01 -3.94174993e-01
3.34019303e-01 4.43428278e-01 2.64042497e-01 -9.78343904e-01
1.72509804e-01 -1.01110530e+00 -1.63374579e+00 -5.95151842e-01
9.41422284e-01 -8.08907747e-02 -1.48741567e+00 9.00563717e-01
-1.45773494e+00 -2.74085283e-01 2.09564939e-01 4.33752745e-01
-2.80720592e-02 4.75739807e-01 -4.86294538e-01 -1.50413013e+00
-7.41177261e-01 -7.57988930e-01 1.42734194e+00 5.63415766e-01
-7.59104034e-03 -6.57161117e-01 -2.71876425e-01 7.39993230e-02
2.69124985e-01 2.59724170e-01 5.15420377e-01 -3.21378380e-01
-7.03341901e-01 -8.16745520e-01 -4.98936236e-01 -5.64007424e-02
3.95851463e-01 1.75576478e-01 -1.03772283e+00 -6.84682950e-02
-5.95108986e-01 1.97029009e-01 6.66828036e-01 4.49042201e-01
1.39602137e+00 4.34631646e-01 -5.12011528e-01 2.30744511e-01
1.19944263e+00 4.21522945e-01 7.58663595e-01 2.19562411e-01
8.73240232e-01 6.34498894e-01 6.81808174e-01 5.34872234e-01
-4.62141372e-02 6.63645148e-01 4.34129179e-01 -3.29211384e-01
-3.34936559e-01 1.71553761e-01 -3.22602481e-01 2.55891353e-01
-1.27304506e+00 -2.09502921e-01 -1.23501265e+00 2.63850421e-01
-1.64504325e+00 -6.89936638e-01 -5.48460782e-01 2.25708532e+00
3.95332545e-01 -4.76673953e-02 8.79142404e-01 7.48127759e-01
7.14245021e-01 -4.54294562e-01 -5.25343776e-01 -1.02627546e-01
3.17453265e-01 1.15477026e+00 7.05258787e-01 3.38443160e-01
-1.18818235e+00 1.16137028e+00 6.26947594e+00 3.49588394e-01
-1.20430231e+00 -9.97809693e-02 2.79813260e-02 3.67720537e-02
4.81337935e-01 -6.12538159e-01 -8.24705362e-01 2.25285918e-01
8.94395187e-02 7.16662228e-01 4.33031112e-01 5.79185426e-01
3.42604429e-01 -2.68881798e-01 -8.89202952e-01 1.24762499e+00
-1.99770793e-01 -6.19529307e-01 3.69913206e-02 -8.39970335e-02
1.28664747e-02 -2.54462838e-01 -2.03075767e-01 -4.76407796e-01
-4.55786169e-01 -1.00045538e+00 5.68697155e-01 6.80204630e-01
8.05012524e-01 -6.18761837e-01 6.87478304e-01 5.11008263e-01
-1.47150350e+00 1.29278421e-01 -1.27997667e-01 -2.91289777e-01
1.42070383e-01 4.40687835e-02 -9.70095992e-01 1.17304169e-01
6.12119496e-01 -4.31886762e-02 -4.23934042e-01 1.19210839e+00
-4.25247699e-01 1.07563883e-01 -5.59185863e-01 -7.16120839e-01
7.66078755e-02 -3.58157009e-02 3.59037906e-01 1.38189721e+00
1.86089218e-01 4.67856973e-01 9.25615653e-02 9.02431071e-01
4.47578371e-01 2.64836568e-02 -1.02953725e-01 -1.17705725e-01
5.03420770e-01 1.19032788e+00 -1.57214010e+00 -3.42090756e-01
3.80088165e-02 1.54471469e+00 -6.71127856e-01 3.42087239e-01
-3.86790693e-01 -1.07725132e+00 3.42909485e-01 2.58367807e-01
1.08316235e-01 -7.76637375e-01 -8.06548715e-01 -7.26194084e-01
3.32089186e-01 -2.58220792e-01 4.43605892e-02 -3.78788531e-01
-5.31860888e-01 4.66937661e-01 -3.05422544e-01 -1.10115635e+00
-2.26222605e-01 -1.49516594e+00 -5.82477868e-01 1.33048046e+00
-1.27638280e+00 -1.33483934e+00 -1.08713198e+00 1.09982479e+00
4.53123212e-01 -3.39041539e-02 9.27886069e-01 1.19895838e-01
-5.51294088e-01 5.68129122e-01 -3.93975019e-01 5.60535192e-01
2.60427058e-01 -1.34190261e+00 7.90748715e-01 7.38864958e-01
3.44113782e-02 8.63083243e-01 1.45229727e-01 -7.68311441e-01
-1.57618117e+00 -3.46799880e-01 6.29515171e-01 -5.63970089e-01
-1.80274263e-01 -1.07862666e-01 -2.18273103e-01 4.15396899e-01
-1.78254694e-01 -1.59561470e-01 3.78408819e-01 -9.91535112e-02
1.47804935e-02 2.49104217e-01 -1.61340654e+00 4.81312096e-01
1.28971720e+00 -5.73840499e-01 -3.35728317e-01 2.38257408e-01
-2.68058211e-01 -5.94512045e-01 -6.88108623e-01 7.90680796e-02
1.30248189e+00 -1.08659852e+00 1.18373632e+00 -2.89232194e-01
-4.26351279e-01 -2.73740560e-01 2.34051555e-01 -3.15244526e-01
-3.69359106e-02 -4.61994857e-01 -2.73819476e-01 8.47724378e-01
-8.68166089e-02 -4.77776080e-01 1.35557127e+00 1.15151393e+00
5.93593717e-01 -4.48227882e-01 -8.77979934e-01 -7.61056483e-01
-4.06132817e-01 -5.74405313e-01 6.77293479e-01 1.75382882e-01
-9.80640650e-02 -3.08890283e-01 1.85428619e-01 1.37576237e-01
1.01650071e+00 1.26444548e-01 7.15780616e-01 -1.44798696e+00
1.36664599e-01 -9.63140011e-01 -4.47593123e-01 -1.24625528e+00
-5.98908126e-01 -3.29546809e-01 -6.46357834e-02 -1.64022720e+00
-1.16649918e-01 -4.05751914e-01 -3.63235950e-01 6.23710155e-01
1.26475453e-01 4.88264978e-01 3.33319068e-01 3.60661745e-01
-1.10740714e-01 -4.66383219e-01 1.34136534e+00 1.63021199e-02
-6.91034913e-01 7.11725473e-01 2.23220289e-01 6.58425212e-01
7.74589539e-01 9.59208906e-02 8.51317868e-02 -1.51624918e-01
-4.80060160e-01 -1.46620095e-01 7.21592069e-01 -1.04253912e+00
3.44739258e-01 -1.72599718e-01 6.92459524e-01 -1.03659320e+00
1.63779452e-01 -8.58396649e-01 -5.11593878e-01 1.00592136e+00
6.47554174e-02 -3.44556868e-01 -1.27192065e-01 -1.80664122e-01
1.60096794e-01 -2.42638946e-01 3.00331980e-01 -4.69302922e-01
-1.13307714e+00 2.48155862e-01 -2.58912563e-01 -7.85769999e-01
9.42613363e-01 -7.98303723e-01 2.91505903e-01 1.39559492e-01
-1.12688708e+00 -1.70772463e-01 -3.38470265e-02 5.36218822e-01
8.50766003e-01 -9.76726770e-01 -4.00793135e-01 2.85495669e-01
-2.64906526e-01 -4.03887741e-02 -1.98482379e-01 6.36189699e-01
-1.08725953e+00 8.67692590e-01 -5.75377643e-01 -5.95152974e-01
-1.79288435e+00 -1.72713343e-02 2.56046623e-01 1.48511708e-01
-4.31952208e-01 1.12990689e+00 -8.92265141e-01 -1.08735859e-01
1.06130981e+00 -8.55069041e-01 -3.02792579e-01 -6.26877993e-02
7.87000179e-01 1.04620075e+00 1.95878029e-01 -2.22874805e-01
-6.69692636e-01 1.12739217e+00 5.02873302e-01 -2.47875750e-01
7.84400523e-01 -1.46221416e-03 -1.88192919e-01 1.47105396e-01
6.32096529e-01 -2.23538503e-01 -1.11673152e+00 6.25134334e-02
8.30920935e-02 -8.02992344e-01 1.16607301e-01 -1.17210400e+00
-8.54027092e-01 1.42687738e+00 1.30455649e+00 1.88325599e-01
1.17807162e+00 5.18577658e-02 8.58677089e-01 5.11542380e-01
9.12982821e-01 -1.26969635e+00 -8.88733640e-02 2.34839559e-01
9.51782465e-01 -1.00819254e+00 -1.66858077e-01 -6.45536542e-01
-7.05064386e-02 1.74446869e+00 4.96975541e-01 -8.82316902e-02
6.00236654e-01 3.00895125e-01 5.80569319e-02 -1.61442146e-01
6.68743968e-01 -8.40037882e-01 6.54091835e-01 6.87367857e-01
5.40281296e-01 4.09534305e-01 -5.67194283e-01 5.98298073e-01
2.87023604e-01 4.35735643e-01 -4.41575587e-01 1.33447123e+00
-5.45581818e-01 -1.45159829e+00 -7.30819762e-01 1.57064423e-01
-1.82511300e-01 2.88580239e-01 -8.72023642e-01 7.33434975e-01
2.64982849e-01 7.18905568e-01 -6.12476729e-02 -7.46395826e-01
5.92339039e-01 3.16854566e-01 1.22488141e+00 -3.91974509e-01
-8.46924782e-01 8.40017423e-02 -6.21709585e-01 -8.30784380e-01
-4.39248770e-01 -3.34049046e-01 -1.55829191e+00 -2.87494570e-01
-1.49508044e-01 -5.20449758e-01 1.23227012e+00 8.68606985e-01
1.37129694e-01 3.26173723e-01 3.81119996e-01 -1.70311320e+00
-5.56696296e-01 -1.07976627e+00 -8.12725663e-01 -7.14974478e-02
1.34554937e-01 -4.20424193e-01 3.28424484e-01 -7.35188425e-02] | [6.464465618133545, -0.35355231165885925] |
a0eb0212-4b58-4ef3-8b82-529086e83d6d | experts-in-the-loop-conditional-variable | 2209.15249 | null | https://arxiv.org/abs/2209.15249v1 | https://arxiv.org/pdf/2209.15249v1.pdf | Experts in the Loop: Conditional Variable Selection for Accelerating Post-Silicon Analysis Based on Deep Learning | Post-silicon validation is one of the most critical processes in modern semiconductor manufacturing. Specifically, correct and deep understanding in test cases of manufactured devices is key to enable post-silicon tuning and debugging. This analysis is typically performed by experienced human experts. However, with the fast development in semiconductor industry, test cases can contain hundreds of variables. The resulting high-dimensionality poses enormous challenges to experts. Thereby, some recent prior works have introduced data-driven variable selection algorithms to tackle these problems and achieved notable success. Nevertheless, for these methods, experts are not involved in training and inference phases, which may lead to bias and inaccuracy due to the lack of prior knowledge. Hence, this work for the first time aims to design a novel conditional variable selection approach while keeping experts in the loop. In this way, we expect that our algorithm can be more efficiently and effectively trained to identify the most critical variables under certain expert knowledge. Extensive experiments on both synthetic and real-world datasets from industry have been conducted and shown the effectiveness of our method. | ['Bin Yang', 'Raphaël Latty', 'Yiwen Liao'] | 2022-09-30 | null | null | null | null | ['variable-selection'] | ['methodology'] | [ 3.73598158e-01 -1.29440621e-01 -2.45865837e-01 -4.88152623e-01
-4.06445384e-01 -3.88582736e-01 9.71335396e-02 1.59351885e-01
9.07646418e-02 1.00326467e+00 -7.70626664e-01 -4.96765018e-01
-2.83615410e-01 -7.45953143e-01 -5.98764002e-01 -5.53230286e-01
2.38991916e-01 4.94106591e-01 2.12466151e-01 1.46611646e-01
6.29208386e-01 2.95463294e-01 -1.39275801e+00 -1.36717111e-01
1.18496692e+00 1.05908227e+00 1.54451236e-01 1.81644827e-01
3.66377346e-02 5.62197626e-01 -5.88786840e-01 -3.42368692e-01
2.17325673e-01 -4.55213994e-01 -3.29985410e-01 2.79025167e-01
-2.32499436e-01 -6.63586035e-02 8.26008096e-02 1.31572926e+00
3.48439693e-01 -1.53074175e-01 2.40570158e-01 -1.19059372e+00
-1.17588798e-02 6.82852268e-01 -7.55563080e-01 -1.68597251e-01
-1.98679753e-02 2.05785945e-01 7.03079939e-01 -7.55004883e-01
2.32370079e-01 8.44583094e-01 1.87607840e-01 2.60722727e-01
-1.00655949e+00 -9.48130906e-01 2.42878526e-01 3.85830440e-02
-1.33136916e+00 -1.51146173e-01 9.36148345e-01 -4.32280034e-01
6.16777778e-01 4.22075354e-02 5.30693769e-01 8.73930991e-01
3.84291261e-01 5.09674013e-01 1.11256635e+00 -3.68727177e-01
5.62038720e-01 6.58601820e-01 1.69890467e-02 5.58123529e-01
6.78974688e-01 7.70545229e-02 -4.25848246e-01 -1.45172263e-02
6.95621908e-01 1.50983140e-01 -1.35428771e-01 -5.43637633e-01
-1.02029085e+00 7.17233121e-01 3.44288424e-02 2.69660264e-01
-2.58955747e-01 -1.28629550e-01 4.70827281e-01 3.59735876e-01
1.37242928e-01 7.86542475e-01 -7.13721633e-01 -4.01175112e-01
-9.36431289e-01 5.79504762e-03 7.50175059e-01 9.44162607e-01
6.22844160e-01 2.10996613e-01 1.48502454e-01 4.74506438e-01
1.96565777e-01 3.43058258e-01 1.08029865e-01 -4.60514188e-01
7.20276177e-01 9.60599601e-01 2.02319130e-01 -1.00569940e+00
-1.98178619e-01 -9.56762850e-01 -9.34092224e-01 2.34776661e-01
5.64543754e-02 -2.93451875e-01 -8.55943680e-01 1.23923588e+00
2.68813640e-01 1.98535696e-01 -2.09720194e-01 8.59761298e-01
-1.12428017e-01 3.60451758e-01 -1.16785496e-01 -4.26080137e-01
9.57973599e-01 -6.22283280e-01 -8.13088715e-01 -4.13295656e-01
3.76186758e-01 -7.01411545e-01 9.03141022e-01 9.34600234e-01
-8.59095573e-01 -5.76446176e-01 -1.54304838e+00 6.72752738e-01
-1.07383378e-01 3.93942297e-01 8.09644639e-01 7.56818652e-01
-1.76928490e-01 7.43511736e-01 -1.07925224e+00 2.09408805e-01
5.65816224e-01 5.68377793e-01 -5.70216700e-02 -3.05525064e-01
-1.02520478e+00 6.26469493e-01 2.75533646e-01 5.71294308e-01
-8.20255101e-01 -6.21631980e-01 -5.36989987e-01 1.38078053e-02
8.99015009e-01 -4.49439019e-01 1.04047871e+00 -7.45070755e-01
-1.48693180e+00 8.87458622e-02 -1.00737706e-01 -2.75668740e-01
4.69546378e-01 -3.09866935e-01 -3.74244630e-01 -3.11424553e-01
-1.31939590e-01 -2.00876445e-01 9.34233904e-01 -1.33357263e+00
-6.23219490e-01 -3.89903963e-01 1.18206188e-01 -3.03219944e-01
-4.06570584e-01 -1.56693488e-01 -4.90934879e-01 -3.05954695e-01
2.08141789e-01 -8.63077462e-01 -5.08535743e-01 -4.38528031e-01
-4.51119751e-01 1.29911274e-01 6.28289461e-01 -5.34077823e-01
1.42219102e+00 -2.05852652e+00 1.21350229e-01 5.71262896e-01
2.48267829e-01 2.64357716e-01 5.70066810e-01 1.97074085e-01
-1.04836971e-01 1.52295321e-01 -2.41707504e-01 -7.54563808e-02
-5.05138189e-02 -1.77644808e-02 -1.94805861e-01 3.71568084e-01
6.14437699e-01 5.62477291e-01 -8.07645559e-01 -4.51508671e-01
4.24914360e-01 1.32034108e-01 -4.53419179e-01 2.12065518e-01
-3.98343235e-01 6.35820627e-01 -9.87206280e-01 9.91154194e-01
5.95226884e-01 -3.45732689e-01 4.13967758e-01 6.11899421e-03
-5.28966188e-02 1.35915369e-01 -1.25605154e+00 1.28564191e+00
-7.64053881e-01 3.66219252e-01 5.44708176e-03 -1.21970439e+00
1.12161505e+00 3.09575051e-01 2.43171498e-01 -6.89016998e-01
4.71579492e-01 4.10625845e-01 2.15891093e-01 -4.31989968e-01
2.14341521e-01 -1.42417535e-01 -2.12895021e-01 5.96815236e-02
-5.06549895e-01 -2.80016690e-01 2.14135677e-01 -2.96155512e-01
1.03443897e+00 2.55096275e-02 2.53802866e-01 1.12827020e-02
7.27292597e-01 1.07890755e-01 1.16904056e+00 3.97234917e-01
9.25824977e-03 2.36320138e-01 7.26719320e-01 -9.18264985e-02
-7.71230876e-01 -7.47698784e-01 -4.62916605e-02 1.76967606e-01
1.06368303e-01 -3.14677030e-01 -4.83657628e-01 -6.53348327e-01
-1.20822839e-01 7.29891717e-01 -3.00542593e-01 -3.71078193e-01
-3.49327356e-01 -5.84688425e-01 -4.77191620e-02 7.24451363e-01
1.82228431e-01 -7.85439730e-01 -6.68005049e-01 3.69392931e-01
4.19304341e-01 -1.07801926e+00 1.59217611e-01 4.82925743e-01
-9.70010281e-01 -1.09784567e+00 -4.12308216e-01 -5.89323223e-01
9.99851108e-01 -8.16161409e-02 1.05306959e+00 2.80883610e-01
-5.68522930e-01 -5.03079355e-01 -4.25160348e-01 -6.03691638e-01
-2.77109474e-01 6.14632070e-02 -6.51488528e-02 -7.20724538e-02
3.53825837e-01 -5.02804816e-01 -5.36474407e-01 6.06597662e-01
-5.81710696e-01 -9.63417515e-02 1.04780889e+00 1.02736533e+00
6.70420945e-01 1.05334949e+00 7.06254900e-01 -1.19165218e+00
3.47360134e-01 -3.89602512e-01 -1.34431922e+00 3.22073877e-01
-8.38818133e-01 5.53178377e-02 9.12129521e-01 -2.90778041e-01
-9.65494394e-01 6.17146455e-02 2.31306255e-01 -3.77136588e-01
-4.84984182e-02 9.16809499e-01 -6.52744412e-01 -4.75507118e-02
1.78297088e-01 -7.07434788e-02 -3.35425675e-01 -2.38778844e-01
-4.32385385e-01 6.45050824e-01 3.38646322e-02 -5.97210586e-01
1.09228015e+00 2.25486755e-02 3.43987197e-01 -4.03944463e-01
-5.91588259e-01 -1.67003438e-01 -4.21106011e-01 -1.46748185e-01
3.94856989e-01 -7.36105144e-01 -6.12842023e-01 3.02000254e-01
-6.41530752e-01 -6.83894306e-02 2.22323775e-01 4.63645548e-01
-1.34038255e-01 2.33614787e-01 -6.32862076e-02 -9.77195561e-01
-6.97494894e-02 -1.51929176e+00 7.82901585e-01 4.61018443e-01
-3.47291440e-01 -8.29854846e-01 -4.53303069e-01 3.95284474e-01
3.33500981e-01 2.69762665e-01 1.07573843e+00 -5.10280907e-01
-9.00155663e-01 -6.40802503e-01 -2.42989957e-02 4.65306193e-01
3.83952677e-01 3.63197446e-01 -6.60861135e-01 -2.43722618e-01
2.97803611e-01 -1.74843952e-01 4.03381854e-01 2.60686368e-01
1.39321303e+00 4.39198911e-01 -4.79894459e-01 2.26288721e-01
1.44996023e+00 5.01614988e-01 5.79906285e-01 1.24573797e-01
5.09048760e-01 6.42081380e-01 1.52607441e+00 7.71614611e-01
-1.99120760e-01 3.84386301e-01 4.29413348e-01 -6.92282915e-02
6.26226127e-01 -9.92717445e-02 7.94015080e-02 7.90340364e-01
1.47904500e-01 -4.16946143e-01 -9.97309208e-01 6.08360708e-01
-1.66075051e+00 -4.01421368e-01 5.02846427e-02 2.40069962e+00
7.72522390e-01 8.12537551e-01 -3.43770891e-01 7.57429183e-01
6.62285209e-01 -3.45485866e-01 -7.84963787e-01 -8.94609019e-02
4.16364312e-01 5.20464182e-01 4.81067866e-01 3.41111282e-03
-9.20312524e-01 5.97745299e-01 4.66859818e+00 7.73078740e-01
-1.19591630e+00 -4.56766009e-01 4.97238517e-01 -5.40647954e-02
-8.72995108e-02 3.23558301e-01 -8.47744584e-01 5.04545391e-01
8.65488887e-01 -2.07771793e-01 2.10152954e-01 1.05235660e+00
2.60802954e-01 -3.03744555e-01 -1.22321844e+00 9.13917482e-01
-3.96285892e-01 -9.78914380e-01 -4.58323419e-01 1.08383812e-01
8.40106130e-01 -6.56406105e-01 1.93509847e-01 3.66131902e-01
5.30845411e-02 -1.13702393e+00 3.87945592e-01 3.11925620e-01
6.04174733e-01 -1.24934423e+00 1.13906217e+00 3.31851721e-01
-9.72973824e-01 -2.43039623e-01 -3.39733660e-01 -6.70000166e-02
1.32220268e-01 1.15242517e+00 -1.11933398e+00 8.40754926e-01
3.27266961e-01 5.53833008e-01 -3.73154551e-01 1.03307796e+00
-4.56415266e-01 7.19486952e-01 -5.31878322e-02 -3.09789598e-01
-1.54209122e-01 -1.18481368e-02 7.54219368e-02 4.87613380e-01
3.29751253e-01 -3.51588160e-01 2.59118229e-01 9.16561604e-01
2.16137096e-02 -2.03502357e-01 -2.60526836e-01 -5.50554037e-01
5.83736777e-01 1.23893940e+00 -1.00048506e+00 7.47562153e-03
-3.76565218e-01 5.67194879e-01 -5.27409166e-02 1.36178270e-01
-1.12226975e+00 -5.89745283e-01 6.58295453e-01 3.37535650e-01
4.43217307e-01 -4.42052960e-01 -7.58803666e-01 -6.45478070e-01
2.58379489e-01 -1.12954497e+00 -1.14770234e-01 -1.51549146e-01
-1.09224761e+00 3.35351020e-01 -2.07102656e-01 -1.33382058e+00
-2.53488809e-01 -6.75487638e-01 -4.36679542e-01 8.67527843e-01
-1.44804537e+00 -4.94097501e-01 -2.35226795e-01 1.59567624e-01
6.80610776e-01 -2.09780425e-01 5.21804094e-01 4.14394140e-01
-9.72166359e-01 6.46424055e-01 -6.20370992e-02 -1.38824299e-01
6.48827970e-01 -1.12620533e+00 2.64024496e-01 9.23375487e-01
-2.54623920e-01 8.54107440e-01 8.82480323e-01 -8.83119345e-01
-1.78575492e+00 -1.05475783e+00 3.34162384e-01 -1.08490385e-01
7.67125189e-01 -4.04149026e-01 -8.06022525e-01 3.53142262e-01
-1.18871823e-01 -2.35142976e-01 5.85804820e-01 2.99144536e-01
2.72131711e-01 -3.46959025e-01 -1.03057551e+00 4.41666514e-01
5.78531086e-01 -3.24358195e-01 -1.60071373e-01 2.07175344e-01
4.82279480e-01 -4.68065023e-01 -8.92905474e-01 7.51352429e-01
2.62459487e-01 -8.12879801e-01 4.72867876e-01 -7.57452622e-02
4.94183630e-01 -4.96744186e-01 1.69708431e-01 -1.12155354e+00
1.59700111e-01 -4.13474083e-01 -1.27698407e-01 1.47667480e+00
6.22360468e-01 -6.54652297e-01 9.31403518e-01 7.65380502e-01
-3.82153578e-02 -1.02708042e+00 -4.76871192e-01 -6.74410164e-01
-3.95786822e-01 -3.79647225e-01 5.97833753e-01 6.35365605e-01
-2.72847176e-01 2.18844369e-01 -3.74691248e-01 4.30694938e-01
5.49145877e-01 2.76348382e-01 7.61902750e-01 -1.33237970e+00
-4.86079305e-01 -9.31989923e-02 -4.67956603e-01 -6.48078084e-01
1.54463559e-01 -2.89829433e-01 2.24313855e-01 -1.21110129e+00
8.84421468e-02 -7.38944948e-01 -4.08141792e-01 1.78694472e-01
-3.64319354e-01 -2.59471923e-01 -3.77846777e-01 -3.71158153e-01
-3.61860007e-01 4.93854284e-01 1.26801383e+00 -8.25617909e-02
-1.41447023e-01 4.21149939e-01 -7.26541638e-01 6.13397121e-01
1.03563154e+00 -5.41511655e-01 -7.74682760e-01 -1.22221902e-01
5.38950443e-01 1.78950265e-01 1.91045418e-01 -1.13914514e+00
2.40724057e-01 -4.63688821e-01 3.56517583e-01 -8.16708148e-01
1.35513470e-01 -1.13498747e+00 2.78156817e-01 4.72217977e-01
2.00918958e-01 -1.01296186e-01 1.87734723e-01 5.32863855e-01
-5.48980951e-01 -3.51072580e-01 5.28025508e-01 1.17209569e-01
-7.68734992e-01 1.47550449e-01 -2.03970224e-02 -1.72604755e-01
1.21870470e+00 -1.57044157e-01 2.98986882e-02 9.40366760e-02
-2.22146615e-01 5.40460467e-01 4.77229953e-01 2.49803886e-01
6.31562293e-01 -7.55820870e-01 -3.33585531e-01 3.68388504e-01
6.92017302e-02 3.11212212e-01 3.38175058e-01 8.02889585e-01
-3.58961165e-01 5.68262100e-01 -1.16155446e-02 -5.66633463e-01
-1.10577834e+00 6.00897133e-01 -1.65414438e-01 -5.02571702e-01
-1.26804784e-01 7.85556614e-01 -2.21994907e-01 -1.39886262e-02
1.65866196e-01 -3.90597761e-01 1.99699044e-01 -1.44387931e-01
3.14425558e-01 2.47456267e-01 2.58622646e-01 1.91371992e-01
-5.05374610e-01 4.36721146e-01 -4.46120709e-01 1.76655352e-01
1.33742046e+00 1.98598430e-02 7.31626973e-02 4.93280888e-01
6.81908667e-01 1.24951519e-01 -1.12751782e+00 5.09153754e-02
3.06511223e-01 -6.25603855e-01 1.98633894e-01 -6.76999569e-01
-1.12081885e+00 1.09398949e+00 3.59746784e-01 -4.78775352e-02
1.27737164e+00 -4.90250170e-01 7.36459672e-01 3.22558761e-01
8.80465984e-01 -1.26845706e+00 1.07910670e-02 1.29373610e-01
3.43786508e-01 -1.26876235e+00 2.05929711e-01 -7.44744420e-01
-5.67257106e-01 9.92849708e-01 8.56371880e-01 9.99080986e-02
5.33063412e-01 5.27567387e-01 -1.28213421e-01 -9.30392966e-02
-7.70673096e-01 2.60451853e-01 5.06614614e-03 3.49304736e-01
4.58146304e-01 2.53719185e-03 -1.88930199e-01 8.00013959e-01
-9.00111198e-02 2.15109795e-01 2.48766467e-01 1.39299107e+00
-2.24411920e-01 -1.50389481e+00 -2.20640540e-01 5.25945723e-01
-4.19655532e-01 1.89580619e-01 -2.53756076e-01 9.33879137e-01
1.82300836e-01 1.00308466e+00 -4.45120186e-01 -4.99025881e-01
4.67173755e-01 -6.23889714e-02 4.83730853e-01 -8.05987120e-01
-3.54574323e-01 3.98905249e-03 2.90139206e-02 -3.54044795e-01
4.14275425e-03 -5.66033781e-01 -1.39099622e+00 -5.30998632e-02
-7.61302829e-01 2.98041791e-01 8.35098028e-01 8.86116624e-01
1.56592965e-01 1.20361984e+00 8.79720569e-01 -4.07754987e-01
-8.76516402e-01 -8.20549190e-01 -6.28595114e-01 -2.33128145e-02
1.27045000e-02 -1.08486390e+00 -2.69599050e-01 -3.00859928e-01] | [7.0144853591918945, 2.3962554931640625] |
24b08a1f-b769-40c5-809b-642ac82b28dd | a-meta-evaluation-of-c-w-l-a-metrics-system | 2307.02936 | null | https://arxiv.org/abs/2307.02936v1 | https://arxiv.org/pdf/2307.02936v1.pdf | A Meta-Evaluation of C/W/L/A Metrics: System Ranking Similarity, System Ranking Consistency and Discriminative Power | Recently, Moffat et al. proposed an analytic framework, namely C/W/L/A, for offline evaluation metrics. This framework allows information retrieval (IR) researchers to design evaluation metrics through the flexible combination of user browsing models and user gain aggregations. However, the statistical stability of C/W/L/A metrics with different aggregations is not yet investigated. In this study, we investigate the statistical stability of C/W/L/A metrics from the perspective of: (1) the system ranking similarity among aggregations, (2) the system ranking consistency of aggregations and (3) the discriminative power of aggregations. More specifically, we combined various aggregation functions with the browsing model of Precision, Discounted Cumulative Gain (DCG), Rank-Biased Precision (RBP), INST, Average Precision (AP) and Expected Reciprocal Rank (ERR), examing their performances in terms of system ranking similarity, system ranking consistency and discriminative power on two offline test collections. Our experimental result suggests that, in terms of system ranking consistency and discriminative power, the aggregation function of expected rate of gain (ERG) has an outstanding performance while the aggregation function of maximum relevance usually has an insufficient performance. The result also suggests that Precision, DCG, RBP, INST and AP with their canonical aggregation all have favourable performances in system ranking consistency and discriminative power; but for ERR, replacing its canonical aggregation with ERG can further strengthen the discriminative power while obtaining a system ranking list similar to the canonical version at the same time. | ['Tetsuya Sakai', 'Nuo Chen'] | 2023-07-06 | null | null | null | null | ['information-retrieval'] | ['natural-language-processing'] | [-3.66981000e-01 -3.32373261e-01 -1.97036028e-01 -1.17124766e-01
-5.92210054e-01 -8.81806016e-01 7.34075665e-01 6.11276090e-01
-4.91893381e-01 5.23507714e-01 -1.43101573e-01 -3.43071997e-01
-1.22413540e+00 -7.25990713e-01 -1.38239309e-01 -5.94301462e-01
-3.78791988e-01 3.33588332e-01 5.75243771e-01 -4.46642399e-01
4.11840528e-01 4.98535454e-01 -2.21899557e+00 2.27326885e-01
1.16531861e+00 1.33749175e+00 2.08848804e-01 3.35060328e-01
-1.00178681e-01 2.39991352e-01 -6.20795906e-01 -4.56116617e-01
2.91834712e-01 -2.89607316e-01 -6.98536575e-01 -5.00329435e-01
8.80378559e-02 -5.24026632e-01 5.49423927e-03 8.44400287e-01
4.71138090e-01 3.35136056e-03 7.14034557e-01 -1.34707057e+00
-5.86099088e-01 2.99208969e-01 -3.32435369e-01 5.28572015e-02
8.00545990e-01 -1.79087549e-01 1.25526989e+00 -5.15852571e-01
3.83779079e-01 1.07707572e+00 1.75471678e-01 -1.68297857e-01
-1.13879979e+00 -3.94039601e-01 1.63505659e-01 1.85597345e-01
-1.63837528e+00 2.42034465e-01 2.72695094e-01 -3.85737151e-01
4.27490234e-01 7.78724015e-01 6.66490316e-01 3.57925594e-01
2.55072623e-01 6.66797817e-01 1.16821647e+00 -5.74481964e-01
2.57644773e-01 4.91423875e-01 6.29092395e-01 2.76109368e-01
6.48246288e-01 -1.48994207e-01 -2.48051867e-01 -3.15431535e-01
4.82798636e-01 -1.20531581e-01 -3.36106956e-01 -3.61516804e-01
-6.84909165e-01 6.02750063e-01 3.96695942e-01 5.42630911e-01
-8.20153058e-02 -3.48586321e-01 1.20807730e-01 6.17847860e-01
-1.59603562e-02 6.35245323e-01 -4.67719048e-01 -1.85493939e-02
-6.75716937e-01 4.26886231e-01 9.20884430e-01 1.03618097e+00
5.60394943e-01 -6.70820832e-01 -7.78883159e-01 9.77052152e-01
4.24133539e-01 5.58055401e-01 6.45030856e-01 -8.03357244e-01
2.09194109e-01 1.12168777e+00 3.95878524e-01 -8.62082779e-01
-4.87340182e-01 -4.96053696e-01 -5.01341879e-01 -9.13393348e-02
6.24797225e-01 4.25619364e-01 -2.77148336e-01 1.74695015e+00
-3.69314402e-02 -9.53175366e-01 -1.04343735e-01 8.20612311e-01
7.53196478e-01 4.73875135e-01 9.47903022e-02 -5.99168420e-01
1.48881471e+00 -4.14798826e-01 -6.52235687e-01 3.70063037e-01
5.07803977e-01 -9.31697667e-01 1.47712195e+00 5.33462286e-01
-1.06268442e+00 -6.89791620e-01 -9.85187531e-01 2.87525356e-01
-4.97593850e-01 3.18452775e-01 5.40775955e-01 7.46806860e-01
-1.12034976e+00 5.72499812e-01 -2.93221682e-01 -3.79964501e-01
-5.04652858e-01 3.68138582e-01 8.36632922e-02 1.94824263e-01
-1.34831345e+00 8.02783310e-01 4.92677927e-01 -1.00028068e-01
-4.18741167e-01 -4.14725244e-01 -1.40046805e-01 3.09216559e-01
2.64162004e-01 -3.95024121e-01 8.98212314e-01 -5.35437822e-01
-1.35969472e+00 1.80711627e-01 3.52326959e-01 1.86512712e-03
6.10447347e-01 -3.61516565e-01 -8.45168591e-01 2.96028033e-02
-3.47014926e-02 2.09913060e-01 3.18685658e-02 -1.20090997e+00
-7.02837050e-01 -5.39333522e-01 4.25289452e-01 4.00656044e-01
-5.42131305e-01 9.57430806e-03 -4.18093890e-01 -7.92544037e-02
3.01367342e-01 -8.00147653e-01 1.54064432e-01 -3.55485559e-01
-8.17030370e-02 -4.98147339e-01 4.04666215e-01 -5.70267439e-01
2.13245344e+00 -1.99204779e+00 -1.59450084e-01 8.62975299e-01
-1.04701808e-02 3.82471770e-01 -6.84137642e-02 7.44811058e-01
3.64226133e-01 2.35602692e-01 2.23493576e-01 5.58680236e-01
1.54668212e-01 -3.83466706e-02 1.04936309e-01 2.56003216e-02
-2.29642153e-01 4.81335282e-01 -1.00272286e+00 -4.82824922e-01
-4.54659723e-02 2.19324261e-01 -5.28282762e-01 2.06803352e-01
8.69338959e-02 -1.59917712e-01 -6.64663553e-01 7.37712026e-01
4.94309247e-01 -2.29008451e-01 3.96974862e-01 -3.17677855e-01
-4.31187272e-01 1.09293610e-01 -1.51746273e+00 8.54352117e-01
-3.17873001e-01 1.67433783e-01 -2.99964696e-01 -4.17359084e-01
1.17218781e+00 4.94207479e-02 5.41080892e-01 -1.16614234e+00
-2.17112467e-01 6.33953929e-01 1.30120248e-01 -3.82778019e-01
8.94612551e-01 5.82753122e-01 2.61008162e-02 4.42684799e-01
-2.01255575e-01 2.15952903e-01 6.44506693e-01 4.28610861e-01
8.37740779e-01 1.00980923e-02 4.07140106e-01 -6.39168441e-01
7.81702697e-01 -5.18653095e-01 -1.30739078e-01 7.07450688e-01
8.86928737e-02 3.08129996e-01 8.06015670e-01 1.26143973e-02
-1.04736018e+00 -1.25594699e+00 -6.64150655e-01 1.10516703e+00
5.96802235e-01 -6.31650686e-01 -4.46677476e-01 -4.68766272e-01
9.95536745e-02 7.33048320e-01 -5.17976470e-02 -2.97787249e-01
-2.22433895e-01 -7.20022142e-01 4.26993042e-01 1.24056898e-01
6.65443897e-01 -6.50295973e-01 -5.38882256e-01 -1.13964938e-01
-3.24324131e-01 -4.99849468e-01 -3.11607301e-01 -1.30574536e-02
-9.16656673e-01 -1.03736186e+00 -6.60686314e-01 -1.98860511e-01
5.12430906e-01 4.24634874e-01 1.06874144e+00 4.35667276e-01
2.44362876e-01 4.06928778e-01 -6.82383597e-01 -1.15924135e-01
-1.01253912e-01 6.61503943e-03 2.61691988e-01 -2.38262072e-01
1.05168698e-02 -2.93083638e-01 -8.61889005e-01 1.00750160e+00
-1.18292296e+00 -6.31338060e-01 7.67343760e-01 4.72836316e-01
5.11683345e-01 -1.56824407e-03 5.67072570e-01 -3.81602973e-01
1.12567747e+00 -2.99637228e-01 -6.23838603e-01 7.82278180e-01
-1.51365995e+00 2.01265007e-01 3.35009307e-01 -4.20027554e-01
-8.75158429e-01 -6.25828207e-01 1.76563755e-01 2.34665379e-01
3.81516248e-01 6.27402604e-01 -3.23484987e-01 1.30068839e-01
9.07453597e-01 1.77362204e-01 5.30274957e-02 -5.83240807e-01
1.87280104e-01 8.79378080e-01 1.60604239e-01 -8.21552157e-01
3.84100199e-01 -1.36685133e-01 7.33868452e-03 -5.60317874e-01
-6.17136240e-01 -6.58960998e-01 -4.07428712e-01 -3.45049798e-01
3.13997418e-01 -5.83882749e-01 -1.04133976e+00 1.81935742e-01
-7.96444356e-01 2.65847623e-01 -1.56353235e-01 6.02636099e-01
-3.60855550e-01 6.90847754e-01 -2.58914351e-01 -1.04478645e+00
-2.90260375e-01 -1.16022456e+00 8.76430213e-01 2.65202343e-01
-3.31742465e-01 -5.85925460e-01 -1.19644336e-01 1.76380262e-01
4.80512470e-01 -2.68710367e-02 1.07234132e+00 -7.92663634e-01
-5.98317444e-01 -4.17817324e-01 -4.71970737e-01 4.72655922e-01
2.37597600e-02 2.31385499e-01 -4.18467999e-01 -4.11660463e-01
-2.69855052e-01 1.43201619e-01 3.34886193e-01 4.15565252e-01
1.00428963e+00 -5.17213523e-01 -1.20694526e-01 9.49487612e-02
1.49625540e+00 6.48707569e-01 1.04431820e+00 6.62944973e-01
-7.88823888e-02 6.41512871e-01 8.78493309e-01 4.99749064e-01
1.09940924e-01 1.00013280e+00 1.59300223e-01 3.42402726e-01
2.13914573e-01 -1.25535458e-01 4.50821668e-01 6.03460848e-01
-4.87968743e-01 -3.19718510e-01 -6.79525256e-01 1.34807363e-01
-1.73991120e+00 -8.15061450e-01 -2.00062677e-01 2.94726992e+00
4.66355801e-01 3.39488834e-01 4.06845659e-01 4.43823099e-01
6.90585434e-01 -2.77926177e-01 4.86533046e-02 -5.59942126e-01
-6.80543436e-03 -1.76776260e-01 4.18901891e-01 2.62927562e-01
-4.22892630e-01 6.80225194e-02 5.97128534e+00 1.08744705e+00
-6.11229122e-01 -1.59311473e-01 4.02749807e-01 -8.58063102e-02
-5.50447106e-01 2.01507002e-01 -1.07421303e+00 8.33349764e-01
9.35189664e-01 -5.49382150e-01 3.33860070e-01 7.91959286e-01
-9.29900333e-02 -4.05279487e-01 -1.09851098e+00 5.53041041e-01
-1.96343184e-01 -5.24643898e-01 3.44356149e-01 3.40382755e-01
3.97311628e-01 -6.72088921e-01 5.39000630e-02 3.06394547e-01
2.36238968e-02 -5.83975792e-01 6.41690493e-01 7.76356697e-01
6.79148555e-01 -6.11580491e-01 1.16433299e+00 2.57841289e-01
-1.08004570e+00 -2.74966866e-01 -1.78364009e-01 4.49876487e-02
-2.87831631e-02 7.47688830e-01 -2.88865715e-01 1.00251877e+00
6.66204512e-01 -8.61376897e-02 -9.18009043e-01 1.20574641e+00
-3.88716869e-02 1.05095811e-01 -4.75757152e-01 -5.27632117e-01
1.17785834e-01 -5.08834958e-01 4.57924664e-01 9.82321203e-01
6.47595108e-01 1.24379769e-01 -6.27598539e-02 6.41876340e-01
4.35749650e-01 5.01702070e-01 -1.10379659e-01 3.27779539e-02
8.31016362e-01 1.17944920e+00 -6.87029302e-01 -1.56077549e-01
-3.83925959e-02 1.56294987e-01 -2.29962572e-01 2.49750108e-01
-8.41022611e-01 -5.55284739e-01 2.22196773e-01 4.64155257e-01
-1.06294043e-01 -1.02132589e-01 -1.02508634e-01 -7.29816735e-01
4.36537415e-01 -6.33885860e-01 6.65752292e-01 -6.05521202e-01
-1.06811786e+00 6.29234791e-01 6.29481852e-01 -1.71377599e+00
-6.13054447e-02 -4.48255301e-01 -2.60486931e-01 7.99992800e-01
-9.05572772e-01 -6.98157310e-01 -4.04312879e-01 5.58415353e-01
-4.33311723e-02 -2.21970677e-01 5.09706140e-01 4.19214576e-01
-3.83145779e-01 8.96652997e-01 4.86125737e-01 -5.66642880e-01
4.82178211e-01 -1.20112312e+00 -6.24843180e-01 3.79056036e-01
-2.01290384e-01 8.87094736e-01 7.11547732e-01 -4.18051511e-01
-1.05820155e+00 -4.43038851e-01 8.54991317e-01 -4.79148775e-01
5.13290703e-01 1.54212847e-01 -8.24448705e-01 -7.89402425e-02
-2.97008306e-01 -7.53050506e-01 5.36140740e-01 5.23693562e-01
-2.37308115e-01 -6.12249136e-01 -1.16218197e+00 6.33098423e-01
7.97031224e-01 -2.93277681e-01 -2.37401158e-01 2.92221040e-01
7.37743437e-01 1.44388363e-01 -1.35979640e+00 4.88781989e-01
1.18432438e+00 -1.22037101e+00 1.03732634e+00 -3.59139666e-02
3.38948995e-01 -3.85250896e-01 -4.38676596e-01 -8.45951974e-01
-5.29615164e-01 -4.47834954e-02 -1.47330925e-01 1.42302346e+00
6.23000681e-01 -6.87693417e-01 8.62500370e-02 8.33170950e-01
3.00805476e-02 -9.20701265e-01 -7.30984449e-01 -1.30239832e+00
-2.52848148e-01 -2.85157673e-02 7.79577672e-01 4.03095394e-01
1.37209848e-01 2.76757956e-01 -1.19109377e-01 -2.31646657e-01
2.62438893e-01 4.75042351e-02 4.90126491e-01 -1.50609994e+00
-3.41107398e-01 -6.93472147e-01 -3.39512080e-01 -7.02789903e-01
-7.65769362e-01 -5.63390851e-01 -4.88959879e-01 -1.61893439e+00
1.99211851e-01 -5.05675495e-01 -7.34106004e-01 8.99230391e-02
-5.04519511e-03 -1.58277452e-01 3.19539636e-01 7.61686981e-01
-7.84283876e-01 8.95354971e-02 1.16331708e+00 1.77684665e-01
-7.24954188e-01 1.98802337e-01 -7.59970129e-01 3.70995790e-01
5.08901060e-01 3.44152041e-02 -5.70446134e-01 -1.13551699e-01
6.61890686e-01 1.82841718e-01 5.76501265e-02 -8.81858826e-01
2.06715956e-01 -1.38286635e-01 2.49248609e-01 -5.50056338e-01
1.21558858e-02 -8.39779973e-01 5.52611649e-01 6.01265490e-01
-3.92768055e-01 2.28705645e-01 -1.14751667e-01 3.48979294e-01
-2.37272710e-01 -4.49550718e-01 4.08216268e-01 2.42301479e-01
-6.22094452e-01 -2.01343909e-01 -2.84809880e-02 -4.44633245e-01
8.50384235e-01 -4.52092379e-01 -6.05065942e-01 -3.76451969e-01
-4.77957308e-01 3.16526324e-01 4.08937454e-01 4.08533365e-01
2.23375082e-01 -1.41238105e+00 -4.91331160e-01 6.54717162e-02
4.44750071e-01 -5.56597412e-01 1.85257465e-01 1.08798301e+00
-4.32820946e-01 5.71223736e-01 -2.03749970e-01 -5.62569320e-01
-1.16640878e+00 5.42683303e-01 -7.64171109e-02 -6.94173813e-01
1.47251099e-01 4.27316129e-01 7.48036429e-02 -6.07407764e-02
2.99595177e-01 -1.86565354e-01 -5.21113515e-01 4.20521259e-01
4.44414616e-01 9.17993963e-01 2.92936116e-01 -3.26606959e-01
-3.69348913e-01 5.15921772e-01 -7.92438090e-02 -1.55335605e-01
8.00882518e-01 -3.32849801e-01 -2.44441181e-01 2.37344548e-01
9.04875040e-01 1.35070294e-01 -4.14797962e-01 8.75542313e-02
2.73360133e-01 -6.61377966e-01 -1.12770557e-01 -1.18472922e+00
-5.90027452e-01 2.62624055e-01 9.30829346e-01 1.04573989e+00
1.40401280e+00 4.30534035e-02 2.36159101e-01 3.14637303e-01
5.58171511e-01 -1.33503187e+00 4.79649007e-02 4.66517955e-01
1.02666461e+00 -8.38627517e-01 2.42137030e-01 -3.28292787e-01
-6.30684376e-01 1.05515265e+00 5.36967695e-01 2.21570864e-01
6.76606417e-01 -1.32192343e-01 -4.25286919e-01 -1.67379275e-01
-5.98729670e-01 -3.28423828e-01 8.15370679e-01 1.06595911e-01
8.64121258e-01 2.60262549e-01 -1.53787076e+00 8.81926537e-01
-3.64224911e-01 -1.24797650e-01 1.57472104e-01 4.86026764e-01
-6.55454516e-01 -1.18942475e+00 -2.89246768e-01 6.50509536e-01
-1.61035284e-01 7.77997226e-02 -3.00918519e-01 1.07635689e+00
3.06740478e-02 1.04941642e+00 -1.02805115e-01 -6.93047822e-01
7.74054110e-01 4.94847596e-02 3.20901036e-01 -2.63743345e-02
-5.25904238e-01 1.73846126e-01 2.03076825e-01 -5.08793712e-01
-3.21372390e-01 -5.73369622e-01 -7.13952780e-01 -4.38994676e-01
-7.92279541e-01 6.39868915e-01 6.62836611e-01 7.51550913e-01
2.27894753e-01 2.34275624e-01 8.46991599e-01 -1.01391941e-01
-8.84902179e-01 -1.07306206e+00 -8.81186068e-01 5.01750529e-01
-2.73061723e-01 -8.46099079e-01 -4.69094217e-01 -6.56605840e-01] | [9.740206718444824, 5.760509490966797] |
9202230e-a913-44d4-890c-13a1f419ef7a | deep-filter-banks-for-texture-recognition-and | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Cimpoi_Deep_Filter_Banks_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Cimpoi_Deep_Filter_Banks_2015_CVPR_paper.pdf | Deep Filter Banks for Texture Recognition and Segmentation | Research in texture recognition often concentrates on the problem of material recognition in uncluttered conditions, an assumption rarely met by applications. In this work we conduct a first study of material and describable texture attributes recognition in clutter, using a new dataset derived from the OpenSurface texture repository. Motivated by the challenge posed by this problem, we propose a new texture descriptor, \dcnn, obtained by Fisher Vector pooling of a Convolutional Neural Network (CNN) filter bank. \dcnn substantially improves the state-of-the-art in texture, material and scene recognition. Our approach achieves 79.8\% accuracy on Flickr material dataset and 81\% accuracy on MIT indoor scenes, providing absolute gains of more than 10\% over existing approaches. \dcnn easily transfers across domains without requiring feature adaptation as for methods that build on the fully-connected layers of CNNs. Furthermore, \dcnn can seamlessly incorporate multi-scale information and describe regions of arbitrary shapes and sizes. Our approach is particularly suited at localizing ``stuff'' categories and obtains state-of-the-art results on MSRC segmentation dataset, as well as promising results on recognizing materials and surface attributes in clutter on the OpenSurfaces dataset. | ['Subhransu Maji', 'Mircea Cimpoi', 'Andrea Vedaldi'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['material-recognition'] | ['computer-vision'] | [ 5.05263567e-01 -3.28263044e-01 2.05442324e-01 -4.33977604e-01
-9.31062400e-01 -4.98801619e-01 6.20495498e-01 4.04612236e-02
-1.79994002e-01 4.26218390e-01 -1.78487763e-01 9.49609801e-02
-2.14337677e-01 -9.86989737e-01 -9.17506158e-01 -9.96459007e-01
-1.81369185e-01 4.41978782e-01 1.85877651e-01 -1.23357125e-01
1.67109057e-01 9.08581257e-01 -1.89868283e+00 7.88505375e-01
2.14824930e-01 2.07288480e+00 6.00984469e-02 5.69308579e-01
-1.91633493e-01 5.40030897e-01 -1.35725647e-01 -1.77312061e-01
3.93411219e-01 3.69508922e-01 -9.77883399e-01 2.26120785e-01
1.29150617e+00 -7.85954073e-02 -3.97452056e-01 8.15459371e-01
3.81099552e-01 5.98116256e-02 9.79901314e-01 -7.61445522e-01
-9.07905638e-01 9.65483710e-02 -3.86382014e-01 1.81446642e-01
2.73475498e-01 -2.79634651e-02 9.32052910e-01 -1.06432652e+00
7.13465810e-01 1.24011886e+00 1.00148761e+00 2.67269403e-01
-1.38959384e+00 -2.41608590e-01 1.08221486e-01 2.02301238e-02
-1.59914422e+00 -4.50082242e-01 7.49800205e-01 -4.93679106e-01
1.04903400e+00 4.81046498e-01 4.80045766e-01 1.21658993e+00
3.18609983e-01 1.02530849e+00 1.44692111e+00 -2.71914780e-01
4.39768210e-02 -1.55086443e-01 -9.49898511e-02 7.19219327e-01
6.97073340e-03 -9.82109159e-02 -5.43441176e-01 4.48665291e-04
1.11545348e+00 -1.23748332e-01 5.26644178e-02 -6.12175226e-01
-1.28874958e+00 4.42013532e-01 4.50966150e-01 1.79206550e-01
-2.17835382e-01 3.73419464e-01 2.79123247e-01 4.81902063e-01
9.14496899e-01 3.01063627e-01 -5.04983783e-01 -6.24362156e-02
-8.01232219e-01 2.82337755e-01 8.15439522e-01 1.00741780e+00
8.00721526e-01 1.86253205e-01 -4.96990196e-02 1.21049511e+00
-5.87827526e-02 1.14252687e+00 -7.32990131e-02 -6.31783068e-01
1.99693784e-01 3.52308601e-01 5.82005568e-02 -1.17798388e+00
-3.22583079e-01 -4.04671043e-01 -8.80312204e-01 2.20923632e-01
5.74819863e-01 4.62672502e-01 -1.19450402e+00 1.05468035e+00
1.49641663e-01 -3.81204545e-01 -2.43412420e-01 8.93456757e-01
1.03314817e+00 1.96983099e-01 -2.28292063e-01 5.44449210e-01
1.28990018e+00 -7.17416525e-01 -1.82583928e-01 -1.04809485e-01
2.13938981e-01 -1.11403334e+00 1.03864646e+00 7.56792903e-01
-7.21975684e-01 -6.85115635e-01 -8.41982305e-01 8.40212181e-02
-6.41907096e-01 1.02189764e-01 9.79796290e-01 6.53611898e-01
-1.00998867e+00 8.09517801e-01 -7.73460686e-01 -5.83636999e-01
9.03451264e-01 3.58062297e-01 -6.89816654e-01 -3.89761478e-01
-5.71651876e-01 6.11822248e-01 -4.29671705e-02 3.21343720e-01
-7.90436804e-01 -6.07612908e-01 -8.70098233e-01 -2.93294460e-01
3.95473003e-01 -3.51509243e-01 9.20834124e-01 -1.15293515e+00
-1.31519163e+00 1.07733440e+00 8.85398686e-02 -1.25198290e-01
5.88046789e-01 -5.02169371e-01 -6.53368235e-01 1.18564904e-01
2.53302697e-02 5.85281551e-01 9.92105424e-01 -1.29818225e+00
-3.12144727e-01 -1.77612007e-01 -2.04704523e-01 -1.12075076e-01
-1.65751725e-01 -1.32585272e-01 -4.62164164e-01 -8.57745588e-01
3.90998125e-01 -9.43738461e-01 -4.26810347e-02 3.65435183e-01
-4.75126266e-01 -4.00466621e-02 9.64264810e-01 -4.64760989e-01
6.17808878e-01 -2.21727705e+00 -1.05232403e-01 4.88176614e-01
-1.56117780e-02 -2.97413170e-02 -3.86447549e-01 2.74708241e-01
3.74972150e-02 -3.48334424e-02 -1.96023494e-01 -1.35813830e-02
2.82675564e-01 1.30104765e-01 -2.65592426e-01 6.59879088e-01
6.16068423e-01 9.38879609e-01 -4.90761429e-01 -2.36579731e-01
3.68267506e-01 5.40105939e-01 -4.48177010e-01 -2.64205597e-02
-2.65462190e-01 3.08315247e-01 -4.49189693e-01 1.48147762e+00
9.03483272e-01 -2.07662553e-01 -1.83010250e-01 -5.09752989e-01
-7.82971233e-02 -3.03215653e-01 -1.19239891e+00 1.62849593e+00
-3.98798794e-01 7.67294884e-01 1.99331149e-01 -1.06768179e+00
1.22607887e+00 -2.37610228e-02 6.04287565e-01 -8.87381375e-01
1.76327169e-01 5.02919316e-01 -2.96721876e-01 -4.52650577e-01
4.79367465e-01 8.30877721e-02 -2.77957857e-01 1.90445423e-01
9.76315886e-02 -4.39085186e-01 -2.05031067e-01 -5.55113792e-01
1.03510785e+00 4.39839095e-01 -2.10550293e-01 -8.03946257e-01
4.51674223e-01 7.21843541e-02 -5.67443743e-02 1.08532274e+00
7.20154941e-02 9.90020573e-01 1.17654033e-01 -1.07491195e+00
-1.14214873e+00 -1.46410131e+00 -6.69002473e-01 1.19450223e+00
2.11632829e-02 -1.80701527e-03 -6.15018070e-01 -5.56777656e-01
4.90483552e-01 -3.33856553e-01 -1.10174739e+00 2.72045851e-01
-5.63551605e-01 -6.36737585e-01 4.16035265e-01 8.21029007e-01
8.64296794e-01 -1.10574770e+00 -2.40641430e-01 2.70523608e-01
-1.19629152e-01 -1.29731488e+00 -2.86621228e-02 4.01365578e-01
-5.74544191e-01 -1.01606309e+00 -7.65631497e-01 -7.79847622e-01
5.40025771e-01 1.59102783e-01 1.25306463e+00 -1.18593968e-01
-7.55999506e-01 6.31177723e-01 -3.45568895e-01 -4.01796013e-01
1.49939299e-01 1.98717967e-01 1.28698405e-02 4.23057318e-01
3.94602925e-01 -3.76763463e-01 -6.35429204e-01 5.98348737e-01
-9.05260801e-01 -1.14612475e-01 7.14003384e-01 8.23775232e-01
9.93829072e-01 1.95235182e-02 2.45986655e-01 -7.38304079e-01
1.45608753e-01 -2.31144488e-01 -3.14241856e-01 2.06253886e-01
-9.54420790e-02 -1.11684920e-02 4.88549024e-01 -4.98879790e-01
-8.48893464e-01 7.56233931e-02 -2.11648777e-01 -4.16088641e-01
-6.82845294e-01 2.49887526e-01 -1.00746818e-01 -7.17455685e-01
6.62929714e-01 2.08264932e-01 -1.94312498e-01 -7.57894814e-01
2.77243555e-01 6.63560212e-01 6.54751062e-01 -1.09603488e+00
7.15188086e-01 8.83153081e-01 6.91004619e-02 -1.16913819e+00
-9.08185840e-01 -4.61180657e-01 -9.73981977e-01 -3.23172510e-01
8.00603986e-01 -8.75207901e-01 -5.64778268e-01 9.39818263e-01
-4.90810812e-01 -6.12386405e-01 -3.55861992e-01 1.54411457e-02
-9.15854156e-01 1.17319219e-01 -4.81958061e-01 -5.48129022e-01
-2.33080417e-01 -9.59061623e-01 1.47669339e+00 -1.74606025e-01
-3.04513946e-02 -9.23230886e-01 -4.08539176e-01 2.37306669e-01
1.00801659e+00 6.39104009e-01 7.49525964e-01 -1.94794700e-01
-7.71902621e-01 -2.98686802e-01 -6.16468966e-01 4.38797414e-01
1.99707985e-01 -1.79796413e-01 -1.48170722e+00 -3.18225563e-01
-4.66781467e-01 -5.35443544e-01 1.20483994e+00 2.96662360e-01
1.61633372e+00 -1.98374819e-02 -7.55718797e-02 8.28879654e-01
1.53663945e+00 -2.33873948e-01 6.74272239e-01 5.65531790e-01
9.01953936e-01 5.05321205e-01 5.64492583e-01 2.93945730e-01
1.29502878e-01 7.80367792e-01 3.63352329e-01 -4.08913910e-01
-4.20861781e-01 2.47058585e-01 8.68553370e-02 4.48868692e-01
-4.82238173e-01 4.37683724e-02 -1.10587919e+00 4.39230204e-01
-1.53174675e+00 -6.53156996e-01 -2.48691197e-02 1.93266797e+00
4.61496890e-01 2.51407862e-01 -1.29532620e-01 1.38714924e-01
4.07256097e-01 2.22157151e-01 -3.79225940e-01 -4.98700261e-01
-5.79126358e-01 8.59249115e-01 6.30323827e-01 7.75304437e-02
-1.92577422e+00 1.11030245e+00 6.75480461e+00 1.23422062e+00
-1.46977222e+00 -2.00760484e-01 8.49241197e-01 9.79395136e-02
8.70308746e-03 -6.64179862e-01 -5.34140348e-01 7.01833293e-02
6.29795611e-01 5.97714841e-01 5.20946324e-01 9.74084377e-01
-2.67146617e-01 -2.99332589e-01 -1.03352261e+00 1.03803122e+00
1.55438840e-01 -1.36744380e+00 -3.14152166e-02 1.08396806e-01
7.45825112e-01 2.62527287e-01 4.02371556e-01 -2.90121790e-03
7.97088742e-02 -1.36956906e+00 1.02368975e+00 8.18969011e-01
1.40727174e+00 -5.28384626e-01 8.96327019e-01 -3.30157518e-01
-1.54835367e+00 8.81837979e-02 -5.50938487e-01 5.71499057e-02
-3.65544021e-01 6.97787702e-01 -4.10974026e-01 6.01680756e-01
1.25994408e+00 8.40134799e-01 -8.88740003e-01 9.60562587e-01
2.96619773e-01 4.09339845e-01 -5.45981467e-01 3.23308073e-02
2.37040401e-01 1.50853589e-01 1.47175759e-01 1.61098301e+00
2.44171664e-01 -2.76674688e-01 3.76743168e-01 7.83187807e-01
1.34605616e-01 1.20229200e-01 -7.93917239e-01 -1.37419492e-01
3.59506086e-02 1.31659365e+00 -1.07911539e+00 -2.52546817e-01
-2.32396364e-01 8.88766050e-01 2.96501726e-01 2.97870100e-01
-5.32191753e-01 -3.26298594e-01 5.83403111e-01 2.35769823e-01
7.59340763e-01 -4.13235068e-01 -3.97689670e-01 -1.00405049e+00
1.69255868e-01 -8.14252257e-01 -6.02241233e-02 -7.68235147e-01
-1.84014118e+00 7.44885266e-01 -2.05037475e-01 -1.37384915e+00
4.55634058e-01 -1.28761196e+00 -1.71671584e-01 7.80564129e-01
-1.44621003e+00 -1.71019995e+00 -6.47032082e-01 6.40053213e-01
2.94817626e-01 -1.35318786e-02 1.17339182e+00 2.50894576e-01
-2.24004924e-01 4.38332289e-01 2.85268724e-01 2.06449524e-01
7.26748943e-01 -1.28714085e+00 5.84601939e-01 3.31020772e-01
-6.08876608e-02 3.44616771e-01 3.90294760e-01 -4.54300731e-01
-1.66502678e+00 -1.20235074e+00 4.33905959e-01 -7.01689303e-01
7.27049589e-01 -8.19591701e-01 -8.52354109e-01 1.93782687e-01
-1.91631228e-01 4.96900529e-01 6.37546480e-01 2.77265012e-01
-8.01882744e-01 -2.08548024e-01 -1.18538809e+00 2.19515070e-01
1.31627250e+00 -8.01670551e-01 -2.32317492e-01 4.85056967e-01
6.60891756e-02 -5.97290099e-01 -1.31134057e+00 7.24842131e-01
1.00214458e+00 -8.24299037e-01 1.24106431e+00 -3.80295902e-01
4.12029862e-01 3.20998169e-02 -8.21324408e-01 -8.49276364e-01
-5.95742047e-01 -1.32102758e-01 1.46331578e-01 8.27316463e-01
2.57781863e-01 -4.12418187e-01 9.28007424e-01 1.79609478e-01
-3.09755445e-01 -8.51764441e-01 -8.94337058e-01 -1.01532912e+00
1.58399940e-01 -7.47762859e-01 4.97244656e-01 8.05179656e-01
-7.08154082e-01 -4.17039722e-01 -2.26552650e-01 2.02387921e-04
7.28845954e-01 4.16240841e-01 7.21585035e-01 -1.33899140e+00
-6.79683015e-02 -4.87487376e-01 -6.14253581e-01 -8.40600789e-01
3.80311757e-02 -8.49150956e-01 2.29664952e-01 -1.36601555e+00
9.49242562e-02 -9.40965891e-01 -4.61149842e-01 6.29516065e-01
3.59408081e-01 9.90038335e-01 -6.79325983e-02 1.54558763e-01
-6.99382961e-01 3.95209253e-01 1.37558842e+00 -5.06365597e-01
3.27346385e-01 -2.95804769e-01 -4.24250931e-01 6.85708106e-01
7.43256152e-01 5.66642173e-02 2.24636540e-01 -6.04872763e-01
-2.93423776e-02 -4.51871037e-01 7.68297613e-01 -1.35536814e+00
-2.50199765e-01 -1.97048590e-01 9.43883359e-01 -6.11837447e-01
5.19160151e-01 -8.30044985e-01 2.37077534e-01 -6.20945804e-02
9.21792015e-02 -1.13055542e-01 5.13782799e-01 4.79152471e-01
-2.03679785e-01 3.97946894e-01 7.96407580e-01 -1.75497621e-01
-1.00525916e+00 4.89851564e-01 -4.01118845e-01 -1.77442998e-01
4.90430087e-01 -5.18185914e-01 -4.84498829e-01 5.42380139e-02
-1.04863071e+00 -3.86041880e-01 7.15015292e-01 6.44248664e-01
6.63663208e-01 -1.59507287e+00 -3.75999391e-01 6.52109027e-01
3.81498039e-01 -2.98185879e-03 4.83281910e-01 8.02056432e-01
-8.00828397e-01 4.23455894e-01 -5.50162554e-01 -1.02914906e+00
-9.24671531e-01 8.82476047e-02 3.69143784e-01 8.08366686e-02
-7.54402399e-01 9.61299956e-01 1.97484702e-01 -5.04414916e-01
1.39084205e-01 -4.60549593e-01 4.11003083e-02 -6.02704622e-02
2.39463955e-01 2.47320049e-02 5.81336558e-01 -7.39792347e-01
-4.34010088e-01 1.08064187e+00 1.08339638e-02 2.00159445e-01
1.53282046e+00 1.39536500e-01 -1.90705717e-01 4.80229884e-01
1.15022516e+00 -4.99360301e-02 -1.39407825e+00 -4.87091988e-01
-7.12331831e-02 -6.34413540e-01 2.10813768e-02 -9.94061828e-01
-1.10531533e+00 9.08905268e-01 8.89596283e-01 1.65008321e-01
9.60531592e-01 1.21768236e-01 6.14217043e-01 8.00841689e-01
8.19780231e-01 -1.22983193e+00 2.83693671e-01 7.74236560e-01
1.17898405e+00 -1.35836840e+00 -2.33065849e-03 -6.66195631e-01
-2.75983363e-01 1.31783903e+00 4.54338789e-01 -4.92630184e-01
1.04139400e+00 3.60280633e-01 2.22147390e-01 -5.22620201e-01
-3.16388309e-01 -2.03348204e-01 7.05006778e-01 6.94265008e-01
3.06955487e-01 4.15717572e-01 5.06399989e-01 4.77633893e-01
-3.06628883e-01 -2.86061555e-01 5.57755046e-02 1.18728089e+00
-4.98833835e-01 -9.15927529e-01 -4.31400567e-01 7.99651146e-01
-4.81516719e-01 -1.73499033e-01 -3.18682551e-01 8.41822147e-01
2.38395259e-01 5.83748996e-01 3.17264885e-01 -5.43306053e-01
3.78550828e-01 -8.97582918e-02 8.19917142e-01 -4.85227168e-01
-4.13955033e-01 3.92966211e-01 3.73006076e-01 -9.14201498e-01
-5.76422334e-01 -7.75894701e-01 -5.98948956e-01 -3.92410994e-01
-2.16199294e-01 -3.34841877e-01 5.87527096e-01 7.61350513e-01
8.30939114e-02 4.30758268e-01 4.76310402e-01 -1.32292879e+00
-7.07895383e-02 -8.95866752e-01 -8.48788142e-01 7.22208798e-01
2.63091594e-01 -1.00498950e+00 -1.36532351e-01 7.42783248e-02] | [10.194026947021484, -0.19641128182411194] |
44fc4f1e-ca6d-432d-8700-700859c2d286 | deep-convolutional-poses-for-human | 1612.03982 | null | http://arxiv.org/abs/1612.03982v1 | http://arxiv.org/pdf/1612.03982v1.pdf | Deep Convolutional Poses for Human Interaction Recognition in Monocular Videos | Human interaction recognition is a challenging problem in computer vision and
has been researched over the years due to its important applications. With the
development of deep models for the human pose estimation problem, this work
aims to verify the effectiveness of using the human pose in order to recognize
the human interaction in monocular videos. This paper developed a method based
on 5 steps: detect each person in the scene, track them, retrieve the human
pose, extract features based on the pose and finally recognize the interaction
using a classifier. The Two-Person interaction dataset was used for the
development of this methodology. Using a whole sequence evaluation approach it
achieved 87.56% of average accuracy of all interaction. Yun, et at achieved
91.10% using the same dataset, however their methodology used the depth sensor
to recognize the interaction. The methodology developed in this paper shows
that an RGB camera can be as effective as depth cameras to recognize the
interaction between two persons using the recent development of deep models to
estimate the human pose. | ['Neil M. Robertson', 'Sankha Mukherjee', 'Marcel Sheeny de Moraes'] | 2016-12-13 | null | null | null | null | ['human-interaction-recognition'] | ['computer-vision'] | [-5.19096106e-02 -3.24424624e-01 3.69616866e-01 -3.04026216e-01
-2.65229512e-02 -3.24441612e-01 6.52082205e-01 -4.71901029e-01
-7.72624135e-01 4.65710461e-01 2.49873281e-01 3.95606816e-01
-3.01171821e-02 -3.74114841e-01 -2.54775047e-01 -5.76772690e-01
1.33958548e-01 7.95778811e-01 1.05607204e-01 -2.33159110e-01
3.64160687e-01 8.25381339e-01 -1.81231165e+00 3.17368537e-01
-1.25825312e-02 7.68468082e-01 2.78297693e-01 1.04256248e+00
4.16055679e-01 5.93231380e-01 -9.29061651e-01 2.15741508e-02
4.89472955e-01 -4.54954833e-01 -6.67984903e-01 3.41090769e-01
3.45451057e-01 -5.64064145e-01 -3.81291032e-01 5.32023728e-01
1.10984337e+00 1.96905404e-01 6.69561327e-01 -1.37795901e+00
3.12734932e-01 -8.61789435e-02 -5.45546174e-01 3.14008802e-01
1.27311051e+00 2.38460787e-02 2.93653667e-01 -8.67275715e-01
6.69003904e-01 1.17533851e+00 6.22836828e-01 4.38507676e-01
-6.17525756e-01 -6.07226133e-01 -5.74971974e-01 5.41271925e-01
-1.35709810e+00 -1.43211737e-01 6.92829847e-01 -7.74624884e-01
1.23796737e+00 2.08765030e-01 1.01294816e+00 8.19702744e-01
1.85851395e-01 6.63291216e-01 1.18475056e+00 -9.23555851e-01
-2.21024379e-01 1.85492665e-01 5.42852640e-01 7.03739941e-01
4.34501380e-01 3.42807323e-01 -5.87175250e-01 3.39261413e-01
6.44316018e-01 1.23885266e-01 -2.21759722e-01 -2.27487832e-01
-9.78660107e-01 4.91760790e-01 2.79071987e-01 7.00332046e-01
-6.07278466e-01 -5.80533966e-02 4.87430036e-01 4.53867093e-02
-9.17407572e-02 2.31637657e-01 -2.56739914e-01 -5.10911345e-01
-7.48815477e-01 4.19686198e-01 8.14114988e-01 6.47687972e-01
2.45661676e-01 -3.36529464e-01 -1.26069844e-01 2.53582448e-01
2.94825226e-01 4.49944496e-01 3.44104230e-01 -6.48060024e-01
3.19893688e-01 1.00274324e+00 2.62342036e-01 -1.24690449e+00
-7.71687746e-01 -1.94683690e-02 -4.21003342e-01 5.43276548e-01
7.78851151e-01 -3.98700893e-01 -9.05716419e-01 1.12627149e+00
5.22854626e-01 -4.15994495e-01 2.20833486e-03 1.20870817e+00
9.16488051e-01 5.12902021e-01 -2.13091910e-01 5.03321178e-02
1.61261928e+00 -6.66719496e-01 -8.82020175e-01 5.95376603e-02
5.10012329e-01 -9.80583787e-01 3.59309345e-01 8.60680103e-01
-8.05615067e-01 -1.03786600e+00 -1.15167356e+00 3.01574707e-01
-4.93389308e-01 7.46830285e-01 4.85976875e-01 8.35600674e-01
-7.66898036e-01 2.41796643e-01 -6.44655824e-01 -1.00158703e+00
-3.01600903e-01 8.48806918e-01 -5.99941194e-01 7.43060857e-02
-8.93723011e-01 1.31853604e+00 5.92483282e-01 4.23410714e-01
-6.38044596e-01 2.17868105e-01 -6.60678804e-01 -3.40429008e-01
1.24118738e-01 -6.91843867e-01 8.91294837e-01 -1.00871599e+00
-1.28844833e+00 1.29583979e+00 -9.13660228e-02 -4.48594123e-01
8.00477147e-01 -5.72245955e-01 -2.48523936e-01 2.62181133e-01
-2.22178400e-01 4.62951511e-01 5.08984685e-01 -9.64329362e-01
-9.26568985e-01 -9.34864640e-01 8.28096569e-02 5.00020742e-01
2.39186380e-02 2.27791116e-01 -5.76293409e-01 -6.19956106e-02
6.58191219e-02 -1.33635712e+00 2.62115002e-01 -5.07380307e-01
-1.97052732e-01 -2.59148836e-01 9.49375868e-01 -1.10443914e+00
9.13290262e-01 -1.82993627e+00 3.79339457e-01 6.67640567e-02
3.33891958e-02 5.47482133e-01 5.24714530e-01 5.80700755e-01
1.61423013e-02 -6.41768217e-01 2.45514274e-01 -1.81550071e-01
-1.57207996e-01 5.71163520e-02 3.93895954e-01 6.79036260e-01
-4.60567385e-01 5.07666826e-01 -3.22279930e-01 -5.01810730e-01
8.20484161e-01 9.06738341e-01 5.00342362e-02 6.06150150e-01
5.50440907e-01 5.90085089e-01 -9.99207497e-02 4.64631498e-01
4.80196595e-01 4.01780725e-01 8.82013738e-02 -5.42320430e-01
-1.36486709e-01 -6.06376946e-01 -1.44748783e+00 1.33355165e+00
-1.29952803e-01 1.01096499e+00 -3.81285399e-02 -9.24031973e-01
1.08354950e+00 6.62585318e-01 6.23886943e-01 -5.11988342e-01
6.34906709e-01 1.47657096e-01 7.26867616e-02 -8.15191090e-01
1.66425094e-01 8.16010013e-02 1.38314232e-01 1.20860912e-01
8.56828466e-02 2.15796486e-01 3.14807296e-01 -2.69050062e-01
8.72396529e-01 5.34833908e-01 5.16316533e-01 1.24378376e-01
9.10180688e-01 1.98936895e-01 -4.49833386e-02 5.15385151e-01
-4.56017524e-01 5.11137724e-01 -1.24451868e-01 -9.09845591e-01
-8.36157382e-01 -5.47902584e-01 2.42781073e-01 5.36242306e-01
8.53013024e-02 4.87172082e-02 -9.15228367e-01 -3.85474861e-01
-1.94537312e-01 8.53719264e-02 -4.68140334e-01 -3.36462259e-02
-5.81347942e-01 -5.06009698e-01 5.39187253e-01 4.78578210e-01
9.80939627e-01 -1.22329903e+00 -1.56390619e+00 -5.59247173e-02
-8.66993591e-02 -1.16503668e+00 1.47198871e-01 -6.16917312e-02
-5.24140537e-01 -1.41871250e+00 -9.48656023e-01 -8.56662273e-01
5.75160325e-01 5.66275008e-02 8.52382481e-01 -1.70433503e-02
-8.52649868e-01 7.77122855e-01 -3.41713876e-01 -5.01794815e-01
-1.00823632e-02 -1.52986363e-01 2.88852137e-02 -5.60299307e-02
9.73296046e-01 2.11513862e-02 -7.04456925e-01 3.01081508e-01
-3.11168104e-01 -1.43775955e-01 5.27888298e-01 4.17642713e-01
-1.35827839e-01 7.20515549e-02 -1.88095450e-01 -3.45973015e-01
5.20230770e-01 1.52731597e-01 -5.15595615e-01 9.28894803e-02
-2.54410475e-01 -1.62069887e-01 5.52583598e-02 -4.72765937e-02
-9.57993388e-01 7.24169016e-01 -1.51126981e-02 -3.15280855e-02
-7.03853905e-01 1.01668529e-01 -1.10978775e-01 -2.62629330e-01
5.85840583e-01 4.36505526e-02 4.79006805e-02 -2.75784910e-01
-2.76505440e-01 8.89069796e-01 4.09755796e-01 -8.40972513e-02
4.41023409e-01 4.80046004e-01 3.07768404e-01 -1.14536560e+00
-4.71687019e-01 -1.10241961e+00 -1.19478953e+00 -9.05136108e-01
1.29398239e+00 -8.05616856e-01 -1.22511530e+00 8.88980687e-01
-1.45227170e+00 3.30880195e-01 4.68699574e-01 9.57416058e-01
-4.56955761e-01 3.80468279e-01 -3.27594340e-01 -1.39193022e+00
-5.41251540e-01 -1.06709659e+00 1.05826378e+00 2.24361226e-01
-5.25375426e-01 -6.23158455e-01 1.36630133e-01 7.25556493e-01
9.82297584e-03 6.48611128e-01 1.35907568e-02 -4.29709852e-01
-3.43921363e-01 -8.56509566e-01 -5.26391761e-03 9.16685313e-02
3.81905586e-02 -3.28727327e-02 -8.19576621e-01 -1.57932103e-01
8.24358389e-02 1.93309896e-02 4.50239450e-01 4.08841640e-01
2.80503452e-01 3.97209704e-01 -3.04338485e-01 2.39352107e-01
1.48603928e+00 7.70428598e-01 7.38772750e-01 6.29752398e-01
9.36515272e-01 7.75532663e-01 8.70998383e-01 5.70582330e-01
5.18238209e-02 1.14453387e+00 1.71535075e-01 6.24923818e-02
-1.10885121e-01 2.24731699e-01 4.67577159e-01 3.29414904e-01
-6.85185611e-01 -1.00017130e-01 -1.23145223e+00 8.57540518e-02
-1.70627534e+00 -1.07620060e+00 -4.83927011e-01 2.14618969e+00
9.24856439e-02 1.89235628e-01 5.47179163e-01 8.62658978e-01
7.54818797e-01 -4.64221388e-01 8.91126543e-02 -4.52641517e-01
1.95566639e-01 8.11066031e-02 5.12501657e-01 5.08111775e-01
-1.10991740e+00 6.11090899e-01 6.12642717e+00 -2.91000977e-02
-1.06390512e+00 -2.41469651e-01 2.89236903e-02 5.53634167e-02
1.02835011e+00 -3.99655491e-01 -1.20586026e+00 3.14177781e-01
7.76669323e-01 2.98815221e-01 6.28594235e-02 5.61047077e-01
3.79285246e-01 -7.47278988e-01 -1.11984158e+00 1.58682144e+00
6.85311973e-01 -4.20669526e-01 -1.93527490e-01 4.66677081e-03
3.79051507e-01 -5.62180161e-01 -3.99797589e-01 9.94494855e-02
-3.84565055e-01 -1.09148157e+00 2.54083931e-01 8.67961586e-01
1.20371647e-01 -9.72852707e-01 1.41632962e+00 5.58384955e-01
-1.24573469e+00 -1.08508252e-01 -4.17110920e-02 -7.12969363e-01
2.69054055e-01 1.45171210e-01 -1.28973055e+00 4.19967562e-01
7.36208320e-01 4.52944458e-01 -6.09005034e-01 1.11736238e+00
8.80779419e-03 1.30481660e-01 -4.65417475e-01 -2.70197153e-01
-5.64072244e-02 -2.70560175e-01 4.92488414e-01 1.26403522e+00
4.21810061e-01 3.02965404e-03 2.27124408e-01 2.18808472e-01
7.60925531e-01 -2.17593368e-02 -8.50826263e-01 1.11923389e-01
-5.71675673e-02 9.91074502e-01 -8.34837317e-01 -2.45131135e-01
-8.04623514e-02 1.22059262e+00 -1.83492005e-01 9.91109908e-02
-6.96926117e-01 -5.08568525e-01 3.27869877e-02 2.67690212e-01
1.48734659e-01 -5.61318159e-01 -1.24452494e-01 -6.30817711e-01
1.34762093e-01 -9.04318035e-01 3.47894907e-01 -9.56460714e-01
-4.41535443e-01 7.26857960e-01 3.39475334e-01 -1.17512226e+00
-5.71796954e-01 -1.12374616e+00 -1.02225274e-01 9.51033652e-01
-3.84146154e-01 -1.34319592e+00 -1.03204954e+00 6.34129465e-01
5.76095283e-01 -4.49007064e-01 7.37464070e-01 3.82147431e-01
-5.04544973e-01 1.28600374e-01 -4.94935840e-01 3.22630465e-01
5.40435731e-01 -1.21911383e+00 -2.78999716e-01 8.28983605e-01
1.44257233e-01 5.28622508e-01 1.08746791e+00 -6.33955717e-01
-1.50194335e+00 -8.61419514e-02 1.00639164e+00 -8.63623261e-01
-1.93061143e-01 -1.91644549e-01 8.31568521e-03 6.18323386e-01
3.61040115e-01 -5.10208905e-01 4.62824315e-01 -2.05514044e-01
3.06685328e-01 -2.32012078e-01 -1.29339552e+00 3.61447990e-01
8.50283802e-01 -3.74926269e-01 -8.59734893e-01 1.39931947e-01
-2.33185381e-01 -4.73223388e-01 -5.31497896e-01 3.27641666e-01
9.38506365e-01 -1.42359221e+00 9.39008951e-01 -5.76065779e-02
-2.84131337e-02 -4.29108649e-01 -4.05644663e-02 -7.81349659e-01
7.74095654e-02 -1.28334746e-01 -1.55908704e-01 8.76937568e-01
-7.78731853e-02 -3.18063498e-01 9.68832195e-01 6.17834210e-01
4.28442270e-01 -3.85444283e-01 -6.52853847e-01 -4.21574503e-01
-8.72379899e-01 -4.58470918e-02 -2.77116392e-02 3.08956653e-01
-1.08905450e-01 5.02039075e-01 -5.46980917e-01 6.40503019e-02
6.71946406e-01 -2.97020555e-01 1.27896047e+00 -1.51272678e+00
-2.48256430e-01 3.52325011e-03 -1.32544053e+00 -6.48932636e-01
-1.76210448e-01 -1.67870596e-01 -8.22766721e-02 -1.75256836e+00
2.49263242e-01 4.47808117e-01 9.77523923e-02 -5.47595471e-02
1.05613604e-01 2.69878089e-01 3.01236749e-01 2.20694263e-02
-3.29418212e-01 -1.12684004e-01 8.74282777e-01 1.88346952e-01
8.73598456e-02 2.47863621e-01 1.70896530e-01 6.32624567e-01
5.67762733e-01 -6.47469908e-02 -3.14033478e-01 2.47026980e-03
5.55829033e-02 5.69233447e-02 5.88637948e-01 -1.87166321e+00
3.13042492e-01 4.52147603e-01 1.01448607e+00 -1.09657502e+00
6.83820665e-01 -1.39442813e+00 5.61110914e-01 1.15820479e+00
8.18400383e-02 3.05860877e-01 1.00942880e-01 2.61709630e-01
-3.05570185e-01 -2.61782676e-01 5.71676075e-01 -4.31931585e-01
-1.14518380e+00 -2.28632018e-01 -3.41485381e-01 -4.81897384e-01
1.58695805e+00 -7.94573665e-01 3.24385375e-01 -4.50499773e-01
-9.01078165e-01 -1.16012245e-01 1.77458301e-01 3.43079925e-01
5.16519785e-01 -1.02208829e+00 -3.73549879e-01 1.11047469e-01
-1.75958037e-01 -3.87021363e-01 9.57229808e-02 8.18346262e-01
-1.15860641e+00 8.11085939e-01 -8.79904330e-01 -7.56374657e-01
-2.23207021e+00 3.75632375e-01 4.65407938e-01 -6.00420088e-02
-3.31974626e-01 6.47586524e-01 -3.99766892e-01 -1.30166009e-01
5.85644364e-01 5.64240590e-02 -8.18935990e-01 1.79565206e-01
4.13956255e-01 8.15522790e-01 -3.06858923e-02 -1.08537173e+00
-5.28348684e-01 1.10561717e+00 3.09495062e-01 -4.85457450e-01
1.02170444e+00 -2.25078672e-01 1.04635984e-01 5.97896278e-01
1.21141946e+00 -2.08097264e-01 -9.49923456e-01 3.47899407e-01
1.95590898e-01 -6.08347416e-01 -1.94788858e-01 -9.40369546e-01
-7.44794130e-01 1.03003001e+00 1.21679866e+00 -5.41903228e-02
1.11659861e+00 -4.86832589e-01 3.94068748e-01 3.76142234e-01
7.22272635e-01 -1.17873478e+00 4.35886346e-02 4.84007061e-01
8.49539340e-01 -1.37126315e+00 2.04663977e-01 -3.52806777e-01
-7.53907442e-01 1.40747738e+00 7.27850437e-01 -2.22618833e-01
5.23465216e-01 2.45782286e-01 8.88176858e-02 -4.61735010e-01
-2.04288680e-02 -4.77655023e-01 4.04955000e-01 7.73548365e-01
8.23463917e-01 5.08988015e-02 -6.74660921e-01 4.31324728e-02
-1.69188783e-01 5.21689594e-01 1.84033602e-01 1.10668659e+00
-4.18340117e-01 -9.15050030e-01 -8.22640479e-01 1.49034455e-01
-4.68325168e-01 5.90408385e-01 -9.00473475e-01 1.04865170e+00
4.60219353e-01 1.07896328e+00 -2.36431345e-01 -6.07253492e-01
6.43335283e-01 1.46087736e-01 8.39886308e-01 -3.91994387e-01
-9.88742769e-01 -2.93599427e-01 1.69139311e-01 -4.47691351e-01
-6.80578768e-01 -6.29708350e-01 -1.08364606e+00 -2.05340117e-01
-7.07900822e-02 4.51595113e-02 9.13219750e-01 1.16351402e+00
2.10226942e-02 3.87525827e-01 2.27789044e-01 -1.19954264e+00
-2.68896788e-01 -1.19290638e+00 -4.92913902e-01 6.55952513e-01
8.84015188e-02 -9.24373567e-01 -2.80194640e-01 1.31358072e-01] | [7.820623874664307, 0.14783835411071777] |
55367960-fd51-4620-a0bd-770b74c42b22 | proactive-image-manipulation-detection | 2203.15880 | null | https://arxiv.org/abs/2203.15880v2 | https://arxiv.org/pdf/2203.15880v2.pdf | Proactive Image Manipulation Detection | Image manipulation detection algorithms are often trained to discriminate between images manipulated with particular Generative Models (GMs) and genuine/real images, yet generalize poorly to images manipulated with GMs unseen in the training. Conventional detection algorithms receive an input image passively. By contrast, we propose a proactive scheme to image manipulation detection. Our key enabling technique is to estimate a set of templates which when added onto the real image would lead to more accurate manipulation detection. That is, a template protected real image, and its manipulated version, is better discriminated compared to the original real image vs. its manipulated one. These templates are estimated using certain constraints based on the desired properties of templates. For image manipulation detection, our proposed approach outperforms the prior work by an average precision of 16% for CycleGAN and 32% for GauGAN. Our approach is generalizable to a variety of GMs showing an improvement over prior work by an average precision of 10% averaged across 12 GMs. Our code is available at https://www.github.com/vishal3477/proactive_IMD. | ['Xiaoming Liu', 'Sijia Liu', 'Tal Hassner', 'Xi Yin', 'Vishal Asnani'] | 2022-03-29 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Asnani_Proactive_Image_Manipulation_Detection_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Asnani_Proactive_Image_Manipulation_Detection_CVPR_2022_paper.pdf | cvpr-2022-1 | ['image-manipulation-detection'] | ['computer-vision'] | [ 1.04614723e+00 1.59886211e-01 -2.29042873e-01 1.86340019e-01
-7.04788506e-01 -8.64771545e-01 8.27665806e-01 -2.36545518e-01
-1.51717186e-01 2.33945340e-01 -2.52366096e-01 -6.12238906e-02
3.07831705e-01 -7.20417023e-01 -9.80511785e-01 -7.74041295e-01
-1.15170581e-02 1.24967165e-01 4.32792842e-01 2.07543912e-04
5.00320733e-01 5.32331288e-01 -1.72306013e+00 1.92991167e-01
3.09271306e-01 9.12425458e-01 4.33168799e-01 1.12682426e+00
4.60778743e-01 5.21646380e-01 -9.74099338e-01 -2.35280737e-01
8.71389806e-01 -4.66459870e-01 -4.57015306e-01 5.02361357e-01
6.94009125e-01 -6.31968975e-01 -2.54936129e-01 1.38639283e+00
4.17014778e-01 -1.34290606e-01 5.94657958e-01 -1.43145263e+00
-6.33201480e-01 3.37479085e-01 -8.31778347e-01 1.36183634e-01
6.01792276e-01 6.01886690e-01 5.01952052e-01 -8.08662117e-01
7.68215835e-01 1.28954530e+00 1.46995425e-01 8.12893629e-01
-1.31117034e+00 -9.57718670e-01 -3.14442441e-02 -1.63464203e-01
-1.32113338e+00 -8.09340119e-01 6.56357586e-01 -4.71254736e-01
4.50366884e-01 4.99879152e-01 3.24385136e-01 1.29494667e+00
1.49876922e-01 7.38107860e-01 1.24341452e+00 -5.71486533e-01
-1.57532007e-01 3.45219493e-01 -1.74718067e-01 8.70334923e-01
6.07160330e-01 4.13091063e-01 -3.97650659e-01 -9.00516659e-02
7.44429290e-01 -3.01998436e-01 -6.30374193e-01 -3.72616827e-01
-1.45629239e+00 4.46685016e-01 -1.87459178e-02 8.10023099e-02
-2.93363363e-01 2.42593095e-01 1.63573265e-01 3.98389220e-01
1.14177227e-01 6.25157773e-01 -8.90560076e-02 3.64463218e-03
-7.90076137e-01 1.09089985e-01 6.39759958e-01 1.25932395e+00
7.01492846e-01 8.18814337e-02 -1.53575778e-01 2.36199424e-01
1.27721414e-01 7.07380056e-01 2.24772573e-01 -1.04021549e+00
3.61726433e-01 4.68670607e-01 2.49550432e-01 -1.09276521e+00
2.77947664e-01 -2.61090457e-01 -5.78069627e-01 5.29524505e-01
5.08552670e-01 2.58096904e-02 -1.03530633e+00 1.48033249e+00
3.91525954e-01 -5.44349179e-02 -1.41594619e-01 7.41838515e-01
4.53249007e-01 3.87948006e-01 -3.36308420e-01 -1.47372156e-01
1.41694045e+00 -6.78484857e-01 -5.69998503e-01 -3.94757062e-01
3.65880787e-01 -1.17299664e+00 9.57880378e-01 5.85027218e-01
-9.01224673e-01 -5.56386054e-01 -1.26635253e+00 2.44076699e-01
-3.58726323e-01 2.91422099e-01 2.51247853e-01 1.25173557e+00
-1.22306681e+00 2.55388945e-01 -3.98824513e-01 -4.34306473e-01
4.12189126e-01 4.07965690e-01 -2.36993119e-01 1.70402035e-01
-7.72401690e-01 6.80855095e-01 6.62604272e-01 -8.21061432e-02
-1.25289440e+00 -4.46733415e-01 -7.22068429e-01 -4.59637344e-01
7.07561076e-01 -3.73266131e-01 1.10483837e+00 -1.09639859e+00
-1.29732168e+00 1.30318916e+00 -6.84026033e-02 -3.98170263e-01
8.30664039e-01 -2.40795743e-02 -3.26246589e-01 4.46734428e-01
4.62615453e-02 8.19495857e-01 1.65424979e+00 -1.74542511e+00
-4.35170531e-01 -9.29578170e-02 3.35233808e-01 -6.14657849e-02
-1.34978235e-01 2.53950149e-01 -5.65147936e-01 -6.07283056e-01
-4.28884476e-02 -8.63838375e-01 1.59797162e-01 2.22086325e-01
-7.23066509e-01 3.60841155e-01 1.16825831e+00 -6.44843519e-01
9.40033793e-01 -2.12650704e+00 -1.17271900e-01 1.97952241e-01
3.31232399e-01 6.48881316e-01 -2.77073801e-01 3.70341152e-01
-9.57791656e-02 3.73479426e-01 6.63090199e-02 -1.85581997e-01
-1.51782827e-02 -2.99274083e-02 -3.74245137e-01 8.29041362e-01
3.34519178e-01 8.38820517e-01 -1.01594806e+00 -4.09314662e-01
4.93456721e-01 2.32361078e-01 -8.71988982e-02 3.72605056e-01
-1.39019713e-01 5.17587423e-01 -1.15457647e-01 1.07829762e+00
8.93967032e-01 -1.00536190e-01 8.92293379e-02 -3.47757190e-01
6.42265426e-03 -9.06564370e-02 -1.18424845e+00 1.00994492e+00
5.92827387e-02 5.99739015e-01 1.44437388e-01 -6.24203146e-01
8.26731682e-01 2.33695716e-01 5.90308420e-02 -4.42688882e-01
2.46197909e-01 1.14637353e-01 1.01864129e-01 -5.07684171e-01
6.09825909e-01 4.22583073e-01 1.11248471e-01 4.77817923e-01
-9.27017108e-02 -2.55722344e-01 3.35929781e-01 2.74758041e-01
1.07418323e+00 2.11120680e-01 5.20239532e-01 -8.36154073e-02
4.93475199e-01 -2.96708167e-01 3.16576272e-01 1.33547735e+00
-5.48084438e-01 6.37815595e-01 5.62629640e-01 1.58568602e-02
-9.95785475e-01 -9.67027485e-01 3.36467288e-02 7.88425863e-01
4.98656213e-01 -2.82804310e-01 -9.11616564e-01 -6.24923706e-01
-9.94786993e-02 4.93457615e-01 -7.33271301e-01 -2.27513164e-01
-6.02692068e-01 -3.47186595e-01 6.76932573e-01 1.36901617e-01
7.55506396e-01 -1.00134075e+00 -8.33625078e-01 -2.83299148e-01
-3.89057808e-02 -1.20804024e+00 -6.17966712e-01 -1.76522732e-01
-5.17275751e-01 -1.32006395e+00 -5.14230967e-01 -5.26689649e-01
9.37713444e-01 6.55899227e-01 8.74531031e-01 3.89494747e-01
-4.23594803e-01 5.69008827e-01 -4.61774260e-01 -4.97444868e-01
-9.13181663e-01 -3.73967886e-01 -5.14419116e-02 1.04214579e-01
1.75156221e-02 -3.11152246e-02 -6.82402313e-01 6.03380919e-01
-1.18200302e+00 1.11177281e-01 6.90891743e-01 6.74690008e-01
2.45327353e-01 7.28474706e-02 -4.40164097e-02 -9.22068000e-01
3.14504474e-01 -2.30547845e-01 -6.87185109e-01 3.56376737e-01
-5.60479701e-01 -1.36104092e-01 2.35140279e-01 -8.12314153e-01
-9.08853769e-01 1.27476215e-01 4.29214925e-01 -5.98259866e-01
-3.24859947e-01 -2.37103343e-01 -2.18032151e-01 -6.55719995e-01
7.36862302e-01 3.74174505e-01 8.74373093e-02 -5.16929217e-02
3.33011925e-01 4.69890326e-01 5.74194968e-01 -4.60476041e-01
1.35034120e+00 5.09888709e-01 -5.19243516e-02 -7.32932866e-01
-2.53297657e-01 -5.09281576e-01 -5.43464124e-01 -6.13978386e-01
5.97584426e-01 -5.62045634e-01 -6.47390604e-01 8.91719818e-01
-1.08713329e+00 -2.49919936e-01 -6.34185746e-02 9.87780467e-02
-4.86731499e-01 7.93788731e-01 -3.07470322e-01 -1.19898593e+00
-2.59425044e-01 -1.33001411e+00 1.54724824e+00 7.65718520e-02
-1.06182292e-01 -5.94880581e-01 -5.98010480e-01 4.34237003e-01
4.23156768e-01 5.63147068e-01 5.34281015e-01 -3.41119945e-01
-9.55886006e-01 -6.06984496e-01 -3.01947683e-01 3.34211409e-01
4.30750966e-01 2.41236180e-01 -1.05472147e+00 -5.30248463e-01
1.97642833e-01 -2.26892516e-01 8.11196983e-01 4.60778996e-02
8.39768767e-01 -3.77690941e-01 -4.46224600e-01 3.68930548e-01
1.40989399e+00 4.03302819e-01 8.51949573e-01 3.57152730e-01
4.40456361e-01 2.81839371e-01 8.87583137e-01 2.38487482e-01
-2.88696498e-01 5.83716810e-01 8.87243867e-01 -2.03796346e-02
-2.63872027e-01 4.55817301e-03 7.44845390e-01 1.93017438e-01
-6.86358362e-02 -6.99464142e-01 -7.48170912e-01 3.90856355e-01
-1.46320701e+00 -1.01830971e+00 -7.77824298e-02 2.49658918e+00
5.70954859e-01 3.36918265e-01 1.46113515e-01 2.25770473e-01
1.20262158e+00 4.07848299e-01 -3.98647696e-01 -8.61537904e-02
-9.57753360e-02 2.13566378e-01 9.27119076e-01 2.50244021e-01
-1.30967116e+00 8.70349884e-01 6.40731621e+00 8.45721185e-01
-1.09482932e+00 1.91490818e-02 4.74033445e-01 1.16740271e-01
1.07029833e-01 3.49979922e-02 -7.60518849e-01 5.32789588e-01
6.19049728e-01 -1.62320688e-01 4.73167121e-01 6.79383814e-01
2.53235489e-01 -5.10129571e-01 -9.69874561e-01 7.58250237e-01
4.38400507e-01 -9.32963550e-01 1.51792258e-01 2.87043899e-01
5.58973312e-01 -4.76066738e-01 3.56302440e-01 3.44925895e-02
1.42448023e-02 -7.21893609e-01 8.38987947e-01 3.93518597e-01
8.02344620e-01 -2.82298595e-01 5.12615919e-01 4.09159869e-01
-1.06057489e+00 1.09018475e-01 -1.95809424e-01 1.62987307e-01
-1.54711023e-01 2.14919731e-01 -1.21015346e+00 4.65352595e-01
2.86882371e-01 3.76381189e-01 -7.54911840e-01 8.41740608e-01
-3.90686899e-01 4.99197036e-01 -8.50626379e-02 1.89851940e-01
-1.48356602e-01 -9.42036808e-02 1.00514650e+00 1.07197177e+00
1.03325948e-01 -3.30881089e-01 1.91914856e-01 8.52143645e-01
-5.78386709e-02 -1.62813574e-01 -8.42175186e-01 -2.78956831e-01
4.28610831e-01 1.21192789e+00 -7.78177023e-01 -3.97602111e-01
-7.68742561e-02 1.30850565e+00 -2.59364218e-01 2.90463358e-01
-1.03371024e+00 -3.20138633e-01 3.38545740e-01 7.12217242e-02
4.00590211e-01 -2.06785187e-01 1.66500807e-01 -1.07681680e+00
2.14715183e-01 -1.33051169e+00 1.69104844e-01 -8.22484195e-01
-7.96407938e-01 4.45362151e-01 1.20590605e-01 -1.49092162e+00
-2.38244027e-01 -7.70078540e-01 -3.57118577e-01 7.98252344e-01
-1.30099058e+00 -1.26885819e+00 -5.89826703e-01 3.18863183e-01
6.54128432e-01 1.91601086e-03 7.49021173e-01 7.34446645e-02
-4.26934063e-01 6.71930075e-01 -2.85840392e-01 3.13402742e-01
8.10491025e-01 -1.10113609e+00 5.12215495e-01 1.43577814e+00
3.80891971e-02 6.24790072e-01 9.47871804e-01 -7.39634037e-01
-1.68695199e+00 -1.15216041e+00 3.24361116e-01 -4.19008821e-01
5.84420085e-01 -5.15204549e-01 -7.35282123e-01 7.45589018e-01
3.50126028e-01 -7.72415102e-02 8.54196921e-02 -6.64319277e-01
-5.64661443e-01 1.62551716e-01 -1.44527423e+00 6.40309453e-01
1.14090109e+00 -4.77237850e-01 -1.91712558e-01 4.71560776e-01
3.53416234e-01 -7.68038332e-01 -6.57946527e-01 5.02706170e-01
7.42451489e-01 -1.03801608e+00 9.67243910e-01 -2.09111542e-01
1.50739744e-01 -6.60655200e-01 -2.27549165e-01 -7.75401592e-01
-8.82866234e-02 -1.05138826e+00 -3.81012708e-01 1.07528532e+00
1.32689595e-01 -8.11048448e-01 6.50921464e-01 2.78071284e-01
3.56956542e-01 -3.49213749e-01 -5.00475466e-01 -1.08352256e+00
-5.25826275e-01 -3.13419908e-01 2.48635978e-01 8.30847204e-01
-4.61768359e-01 -2.97423154e-01 -6.83036029e-01 5.57632804e-01
8.86453509e-01 -2.66488153e-03 1.24492455e+00 -6.41168177e-01
-4.73976195e-01 -3.39174896e-01 -9.54915881e-01 -1.11690116e+00
-1.02278963e-01 -5.53580403e-01 2.54835263e-02 -9.06215131e-01
2.80385762e-01 1.33026496e-03 8.69089887e-02 4.99758333e-01
-2.60890126e-01 6.96693599e-01 2.76678205e-01 2.92832643e-01
-4.01106715e-01 -1.21538699e-01 1.17174363e+00 -2.50347346e-01
4.90232743e-02 -9.18196291e-02 -6.02590799e-01 4.75062668e-01
1.15116000e+00 -4.46584344e-01 -1.50924072e-01 -1.38619646e-01
-1.62350193e-01 -1.51087329e-01 7.09905863e-01 -1.12573874e+00
-5.27938157e-02 -1.62536725e-01 4.64931935e-01 -5.16213536e-01
4.17317331e-01 -5.68461239e-01 3.98936719e-01 7.77103961e-01
-2.40478963e-01 -1.24594823e-01 1.60072356e-01 6.82588339e-01
2.02297598e-01 -4.42483753e-01 8.88742328e-01 -2.98709214e-01
-8.07516992e-01 9.91267636e-02 -5.15985668e-01 -2.80789882e-01
1.25958037e+00 -6.46910846e-01 -7.89116263e-01 -5.28272331e-01
-3.95217478e-01 -2.09325880e-01 8.41333926e-01 4.67959017e-01
6.33145034e-01 -9.24594223e-01 -4.24575388e-01 3.18921149e-01
2.65188128e-01 -4.50926572e-01 -2.27288187e-01 8.96651328e-01
-5.73258460e-01 1.85766950e-01 -9.82988626e-02 -6.99775338e-01
-1.65178537e+00 6.90693378e-01 2.75322139e-01 5.87498471e-02
-3.09102625e-01 6.59183919e-01 1.99302629e-01 1.26948699e-01
8.88951719e-02 -2.59562701e-01 3.56652230e-01 -2.37721741e-01
6.37037873e-01 2.93338209e-01 -5.29823899e-02 -8.62635016e-01
-1.83875769e-01 3.41827452e-01 -3.13169330e-01 8.84037018e-02
6.54572487e-01 -5.66931739e-02 3.12341913e-03 1.82143956e-01
1.12494159e+00 2.72907227e-01 -1.10443175e+00 -1.90815091e-01
2.31905971e-02 -1.16814637e+00 -1.55347392e-01 -6.46241665e-01
-1.03865159e+00 3.77027631e-01 7.82514095e-01 3.53430778e-01
1.35224700e+00 3.98615412e-02 3.21159869e-01 1.66397914e-01
6.07374728e-01 -5.67031682e-01 3.65275770e-01 3.93763520e-02
1.02202845e+00 -1.24087965e+00 1.19964883e-01 -8.35413992e-01
-2.98558354e-01 1.15887809e+00 5.53450882e-01 -2.59936452e-01
7.77429566e-02 3.74750316e-01 2.21358966e-02 -2.38149658e-01
-3.42157751e-01 -3.84766787e-01 1.59634709e-01 8.49370599e-01
8.86165649e-02 3.11091123e-03 -7.42720366e-02 -4.32744175e-01
4.96967584e-02 -2.86837369e-01 6.44553840e-01 1.16263485e+00
-4.48465884e-01 -1.16844666e+00 -7.40754604e-01 4.49500352e-01
-4.18101758e-01 -6.07451536e-02 -7.03359127e-01 1.07559407e+00
1.41722381e-01 1.02037263e+00 -1.61643684e-01 -4.39025134e-01
1.18576556e-01 -1.13340892e-01 7.60551572e-01 -6.97514117e-01
-5.25375843e-01 8.75356942e-02 3.06131560e-02 -7.06487656e-01
-7.81146705e-01 -8.10475707e-01 -5.34203947e-01 -1.05964154e-01
-7.33532667e-01 -4.57426637e-01 5.53057492e-01 5.48255265e-01
2.82097638e-01 1.43708766e-01 7.81057000e-01 -1.05190182e+00
-5.18691003e-01 -7.52994120e-01 -3.52985948e-01 4.19687003e-01
4.40500528e-01 -6.55426800e-01 -6.51304424e-01 3.74773592e-01] | [12.410158157348633, 1.0637781620025635] |
9e999c43-f8a8-45eb-b3cc-59dbf82afdeb | memebot-towards-automatic-image-meme | 2004.14571 | null | https://arxiv.org/abs/2004.14571v1 | https://arxiv.org/pdf/2004.14571v1.pdf | memeBot: Towards Automatic Image Meme Generation | Image memes have become a widespread tool used by people for interacting and exchanging ideas over social media, blogs, and open messengers. This work proposes to treat automatic image meme generation as a translation process, and further present an end to end neural and probabilistic approach to generate an image-based meme for any given sentence using an encoder-decoder architecture. For a given input sentence, an image meme is generated by combining a meme template image and a text caption where the meme template image is selected from a set of popular candidates using a selection module, and the meme caption is generated by an encoder-decoder model. An encoder is used to map the selected meme template and the input sentence into a meme embedding and a decoder is used to decode the meme caption from the meme embedding. The generated natural language meme caption is conditioned on the input sentence and the selected meme template. The model learns the dependencies between the meme captions and the meme template images and generates new memes using the learned dependencies. The quality of the generated captions and the generated memes is evaluated through both automated and human evaluation. An experiment is designed to score how well the generated memes can represent the tweets from Twitter conversations. Experiments on Twitter data show the efficacy of the model in generating memes for sentences in online social interaction. | ['Yezhou Yang', 'Kausic Gunasekar', 'Aadhavan Sadasivam', 'Hasan Davulcu'] | 2020-04-30 | null | null | null | null | ['meme-classification', 'meme-captioning'] | ['natural-language-processing', 'natural-language-processing'] | [ 4.37529325e-01 3.32445204e-01 3.38604718e-01 -4.87146139e-01
-7.74992168e-01 -4.31686491e-01 1.01198304e+00 -1.57651916e-01
-5.32986164e-01 7.82625854e-01 6.22048020e-01 3.05944532e-01
8.59019876e-01 -1.06070983e+00 -1.12886953e+00 -3.53112429e-01
5.12423575e-01 5.75447083e-01 1.86539199e-02 -3.08828115e-01
5.56969464e-01 -2.99193770e-01 -1.43363202e+00 1.08497179e+00
3.99979472e-01 7.50728905e-01 6.71464384e-01 8.11596215e-01
-6.54719889e-01 8.86151850e-01 -7.93831408e-01 -6.14574134e-01
1.23729095e-01 -1.14828551e+00 -8.72228205e-01 3.04877222e-01
4.59433496e-01 -3.35501939e-01 -6.36448085e-01 8.95665407e-01
4.42143500e-01 -2.66495831e-02 7.73983538e-01 -9.89351094e-01
-1.21870923e+00 8.12097311e-01 -2.17889413e-01 9.02761966e-02
6.84526742e-01 2.43649900e-01 6.22578919e-01 -1.29517996e+00
1.16067684e+00 1.29485548e+00 2.29853734e-01 9.17101026e-01
-1.13863337e+00 -5.12869239e-01 -2.33830243e-01 -1.84910044e-01
-1.32450259e+00 -3.66597414e-01 5.36738575e-01 -5.32059431e-01
6.52543902e-01 6.46833470e-03 7.72232831e-01 1.34345829e+00
7.78935790e-01 7.64394701e-01 1.09459901e+00 -3.13793838e-01
1.60815939e-01 7.84097552e-01 -4.28500235e-01 8.24762404e-01
-4.19372231e-01 -4.03288484e-01 -9.48086262e-01 -2.45347455e-01
4.65530068e-01 4.16089185e-02 7.54072815e-02 4.16541815e-01
-1.50793767e+00 1.04116130e+00 6.77390099e-01 2.16912434e-01
-7.14712322e-01 2.71391779e-01 2.00499982e-01 3.26332420e-01
8.91821384e-01 5.10279536e-01 4.82287109e-01 2.29372233e-01
-1.16051376e+00 1.55430421e-01 8.73297036e-01 6.94419980e-01
8.94751430e-01 -4.48874205e-01 -5.19896209e-01 9.52766180e-01
3.65185946e-01 1.05892849e+00 5.53939104e-01 -4.62144196e-01
6.91724479e-01 6.71020806e-01 3.85396272e-01 -1.48920131e+00
1.75824821e-01 2.22819194e-01 -7.14756548e-01 -2.07610145e-01
-2.05505341e-01 -1.54277429e-01 -9.69252825e-01 1.67204022e+00
1.79067329e-02 -1.91835873e-02 2.08456963e-01 8.44628930e-01
1.18243861e+00 1.66422546e+00 6.20339066e-02 6.95830882e-02
1.03029394e+00 -1.09841430e+00 -6.33222163e-01 -6.04001164e-01
1.65398598e-01 -9.88875091e-01 7.92370558e-01 -6.03870630e-01
-1.33620882e+00 -6.26497388e-01 -7.65483141e-01 7.27769826e-03
-3.69689941e-01 8.56554583e-02 -2.93336213e-01 -1.85232740e-02
-1.44197381e+00 1.19565129e-01 -3.30842257e-01 -6.37221515e-01
3.32211316e-01 2.31700346e-01 -3.34630847e-01 -1.75418686e-02
-1.38214839e+00 9.76228237e-01 4.87270147e-01 8.40394851e-03
-1.01957333e+00 7.91668799e-03 -7.58591115e-01 -1.91427723e-01
-4.87317234e-01 -1.17610407e+00 9.43006217e-01 -1.68330145e+00
-1.52401710e+00 1.35339916e+00 -2.92643309e-01 -4.42586720e-01
4.96636659e-01 4.65556622e-01 -2.04883605e-01 4.74464715e-01
4.47233707e-01 1.76082945e+00 9.80054855e-01 -1.35114563e+00
-6.02931321e-01 2.70316541e-01 -6.37003630e-02 5.02196968e-01
-5.74958801e-01 5.81970885e-02 -2.33616561e-01 -4.93603051e-01
-6.75067026e-03 -1.24600339e+00 4.70971689e-02 -7.55332187e-02
-4.38977182e-01 1.67393297e-01 6.30809307e-01 -8.98227990e-01
9.02909636e-01 -2.08521914e+00 2.44200736e-01 1.12621963e-01
2.02347293e-01 -2.41717786e-01 -6.38947606e-01 7.84426630e-01
3.06989610e-01 1.75594479e-01 -4.56513584e-01 -4.58336622e-01
-1.57854378e-01 -1.58274665e-01 -4.82215375e-01 2.46674567e-01
2.52765238e-01 1.19110501e+00 -1.05575550e+00 -6.31969988e-01
-3.09119700e-03 6.35329068e-01 -3.07102084e-01 7.32381880e-01
-4.42329139e-01 4.84391630e-01 -4.57386762e-01 -2.30125323e-01
4.66837704e-01 -4.89506632e-01 -1.11375801e-01 1.56445712e-01
-6.94207400e-02 2.95297533e-01 -5.91473877e-01 1.28975117e+00
-6.17920876e-01 1.03848767e+00 -3.39881957e-01 -2.06749648e-01
1.26992393e+00 3.73414159e-01 1.95384979e-01 -6.81316078e-01
2.38886982e-01 2.36417234e-01 -5.93874872e-01 -7.47663140e-01
8.08865488e-01 -3.73580843e-01 -4.53486711e-01 1.14515579e+00
-9.02384296e-02 -3.77057612e-01 2.73339659e-01 5.31817675e-01
1.13103580e+00 -3.77566256e-02 -1.68479443e-01 2.70247553e-02
5.53856969e-01 3.05159748e-01 -2.50978619e-01 8.18851650e-01
1.33826375e-01 9.08677697e-01 1.13826253e-01 -7.93708205e-01
-1.56811714e+00 -8.81358206e-01 4.29812342e-01 1.18047011e+00
2.27349520e-01 6.40548393e-02 -1.00030792e+00 -5.31327188e-01
-4.01480466e-01 6.85101032e-01 -9.27371085e-01 -2.69903153e-01
-4.98291552e-01 -6.78561091e-01 1.68979600e-01 -1.31782055e-01
8.53263795e-01 -1.93219817e+00 -3.41137111e-01 4.75415647e-01
-1.00981414e+00 -1.01428699e+00 -1.01149094e+00 -6.17743134e-01
-5.56887448e-01 -3.82629216e-01 -1.05309737e+00 -1.43906367e+00
1.18008268e+00 3.49565327e-01 1.31744885e+00 1.22480370e-01
2.84576029e-01 4.17633444e-01 -4.01543349e-01 -1.53311089e-01
-1.41849971e+00 2.56747216e-01 -1.98448539e-01 6.08945668e-01
2.55350232e-01 -1.58063754e-01 -6.59505606e-01 3.24259788e-01
-1.25760710e+00 1.02721179e+00 5.36002934e-01 8.20253134e-01
3.90926361e-01 -5.74167430e-01 6.42017126e-01 -7.90746331e-01
7.16203988e-01 -9.74741638e-01 -1.32506281e-01 2.31639817e-01
8.74440223e-02 1.53653875e-01 4.43183124e-01 -6.63175166e-01
-1.03957987e+00 3.07641745e-01 6.09334409e-02 3.80219594e-02
2.79441833e-01 6.31462097e-01 3.97955894e-01 1.74145028e-01
8.20683897e-01 6.99150205e-01 1.24908842e-01 2.02992782e-01
3.03004324e-01 1.22455049e+00 3.82787347e-01 -1.23322621e-01
7.12732196e-01 5.21212220e-01 -6.54996753e-01 -7.01655746e-01
-7.95877695e-01 -5.63004129e-02 -4.03201163e-01 -5.85803449e-01
1.17746365e+00 -1.02474344e+00 -2.88613677e-01 4.25615311e-01
-1.84734559e+00 -9.72952396e-02 -7.19607696e-02 4.69322838e-02
-6.12913311e-01 -3.57249737e-01 -7.89247632e-01 -6.82734013e-01
-5.77615023e-01 -1.09232354e+00 1.18482411e+00 4.24003929e-01
-3.35576355e-01 -8.91911864e-01 2.97951519e-01 4.64293897e-01
5.01552582e-01 3.54520768e-01 5.87502062e-01 -3.98607224e-01
-8.98024619e-01 -3.48984718e-01 -3.68548095e-01 9.51629877e-02
-2.10505843e-01 -2.13199586e-01 -9.60605264e-01 -1.30880907e-01
-1.93664774e-01 -4.17981386e-01 8.45299125e-01 1.27119213e-01
3.95756781e-01 -7.17370212e-01 -3.50189179e-01 -3.24584916e-02
1.38944030e+00 1.48135439e-01 8.14599276e-01 3.46094698e-01
5.31892419e-01 9.54252720e-01 6.44879416e-03 1.77885145e-01
5.73802710e-01 4.14485604e-01 2.85412133e-01 -1.64080933e-01
-7.71974772e-02 -6.56022608e-01 9.52423871e-01 1.14005935e+00
4.63659167e-01 -7.22553492e-01 -8.02083731e-01 8.03447545e-01
-1.86006165e+00 -1.27567935e+00 3.68055068e-02 1.79631507e+00
8.80865693e-01 -9.58175659e-02 -1.83044728e-02 -6.85820103e-01
1.42818832e+00 2.63071656e-01 -2.94186085e-01 -6.25407636e-01
-3.80238533e-01 -2.72091776e-01 6.72044754e-02 6.23041630e-01
-7.26294458e-01 1.15562904e+00 5.93515348e+00 2.06415713e-01
-1.20063543e+00 4.95226115e-01 9.90261972e-01 -1.58163026e-01
-5.29873550e-01 -2.08805397e-01 -5.71429133e-01 7.37281203e-01
1.28716600e+00 -2.23375171e-01 5.26990592e-01 3.58038008e-01
4.34208155e-01 -1.87787399e-01 -9.47436690e-01 9.99538302e-01
6.95309162e-01 -1.91599739e+00 4.23962861e-01 4.45054285e-02
1.25691700e+00 1.57392561e-01 2.61795253e-01 1.09813631e-01
-1.23784102e-01 -8.27109814e-01 1.22488105e+00 8.64785075e-01
8.36414158e-01 -3.30633819e-01 9.42542553e-01 1.81416512e-01
-8.10184658e-01 1.17176555e-01 -4.56786782e-01 -5.88343330e-02
4.54289377e-01 3.50713849e-01 -1.29484439e+00 -2.69953400e-01
2.85589308e-01 5.15376866e-01 -6.87749445e-01 6.70056581e-01
-2.25510761e-01 4.45257306e-01 -9.00483206e-02 -6.89905226e-01
4.81717080e-01 1.59716792e-02 5.88078499e-01 1.52615750e+00
4.05124962e-01 -1.66112661e-01 1.49584919e-01 1.34001601e+00
-5.90216458e-01 3.64533693e-01 -7.44071484e-01 -5.96213698e-01
4.15368587e-01 1.17900491e+00 -9.46405470e-01 -6.60716832e-01
-7.98011348e-02 1.66243362e+00 1.50621921e-01 2.78965682e-01
-7.06947625e-01 -2.12093815e-01 -8.93599316e-02 3.28392535e-01
-4.90658693e-02 1.09957010e-01 -1.04083814e-01 -9.69201982e-01
1.15653500e-01 -5.71415126e-01 3.17659937e-02 -1.41828561e+00
-1.44541788e+00 1.20309794e+00 -1.26782209e-01 -1.11901200e+00
-3.64338011e-01 -2.17905343e-01 -8.16245198e-01 1.07557607e+00
-8.92473340e-01 -1.27884686e+00 -3.39629799e-01 5.54995909e-02
7.61614442e-01 -4.43765610e-01 8.48524988e-01 8.48456398e-02
-1.50957718e-01 1.15373589e-01 2.28337832e-02 1.98300108e-01
5.68874776e-01 -7.54707754e-01 8.19880664e-01 5.24272501e-01
2.08873630e-01 3.12254101e-01 8.70961368e-01 -1.00675249e+00
-8.21504951e-01 -1.36499155e+00 1.50484610e+00 -8.76316011e-01
4.18951988e-01 -7.30780184e-01 -6.75747931e-01 4.11121994e-01
8.43837023e-01 -2.50935823e-01 5.10054469e-01 -8.69093060e-01
-1.03892319e-01 3.81768346e-01 -1.20883465e+00 7.51024663e-01
7.26163626e-01 -7.36316502e-01 -6.43170893e-01 6.43334448e-01
7.06866384e-01 -3.50926369e-01 -4.15490031e-01 -6.64124310e-01
5.97894311e-01 -6.59904957e-01 5.27968049e-01 -2.35943347e-01
1.32165837e+00 -3.29461366e-01 3.90513055e-02 -1.34681976e+00
-6.41500801e-02 -6.30987704e-01 3.56245846e-01 1.22837448e+00
8.89975011e-01 -2.68806309e-01 6.75257087e-01 4.71019000e-01
1.22298025e-01 -4.21874523e-01 -7.01667309e-01 -9.60933864e-02
-2.18316764e-01 2.28497952e-01 5.14524519e-01 7.28375852e-01
-1.43896431e-01 8.17676723e-01 -6.91734374e-01 -1.50421381e-01
2.12494731e-01 1.12885877e-01 7.75283933e-01 -5.65767527e-01
3.68875675e-02 -1.39303163e-01 -2.38019079e-01 -9.18028057e-01
3.07987809e-01 -1.18860555e+00 7.37651527e-01 -1.99419618e+00
9.86973882e-01 1.79847702e-03 1.38084412e-01 3.30840796e-01
-1.98743001e-01 9.74881530e-01 4.85021681e-01 6.33897543e-01
-7.46004522e-01 5.42694092e-01 1.32084250e+00 -6.17646158e-01
-3.68179440e-01 -3.25725913e-01 -4.49854344e-01 3.71889532e-01
6.56516492e-01 -1.12950206e+00 -3.44329812e-02 -4.31500793e-01
7.52268732e-01 1.40647471e-01 3.68807375e-01 -8.91276896e-01
1.80541739e-01 -1.60850614e-01 4.62332785e-01 -3.61204207e-01
2.89293379e-01 -1.84652656e-01 2.12211534e-01 3.82018834e-01
-8.50636065e-01 3.87194723e-01 -7.55456388e-02 4.57830131e-01
-2.84597099e-01 -2.67118543e-01 9.00547266e-01 -3.91802907e-01
-3.45768839e-01 1.92303613e-01 -7.09550083e-01 -9.87303630e-02
7.58620918e-01 -3.95069867e-01 -1.97172016e-01 -1.05643892e+00
-7.02235460e-01 -5.38591407e-02 5.48853993e-01 8.11983645e-01
9.12501872e-01 -1.72980309e+00 -1.39343691e+00 -6.30156249e-02
3.68731529e-01 -5.41387558e-01 1.73528790e-01 3.03269833e-01
-5.70349813e-01 -3.40049304e-02 -3.31496775e-01 -4.79481369e-01
-8.09228837e-01 1.59429640e-01 2.97482133e-01 -7.49092400e-02
-4.21794206e-01 8.28678191e-01 3.22863311e-01 -4.14779186e-01
-3.63296658e-01 3.28088224e-01 -1.22270145e-01 -1.28479362e-01
8.73149514e-01 -1.09390855e-01 -4.73244280e-01 -1.28393805e+00
1.62316635e-02 2.66798198e-01 5.63219190e-02 -8.01575363e-01
1.45684052e+00 -5.27706087e-01 -3.76061112e-01 5.42148829e-01
1.46418071e+00 -2.77530372e-01 -1.10362637e+00 -8.35877135e-02
-2.01833621e-01 -2.55314887e-01 -2.78777122e-01 -7.08720267e-01
-7.92281806e-01 7.52690136e-01 6.96733236e-01 1.33758113e-01
6.15926385e-01 2.61097997e-01 1.37986529e+00 2.15496749e-01
2.59731114e-01 -1.13517225e+00 6.81999147e-01 6.75138772e-01
1.25401938e+00 -1.34603655e+00 -4.04728115e-01 9.01249573e-02
-8.35276127e-01 1.04969597e+00 4.28720444e-01 -2.15995431e-01
2.03080744e-01 -1.89185023e-01 1.69394642e-01 -2.43361771e-01
-6.45380557e-01 2.61942416e-01 4.10314173e-01 3.13602597e-01
4.32231605e-01 -4.59015630e-02 -2.98411280e-01 1.83407828e-01
-3.34822923e-01 -2.64951829e-02 9.05539513e-01 7.19211459e-01
-7.26337373e-01 -7.51319230e-01 -5.76516211e-01 3.39786083e-01
-1.74350962e-01 -1.79452926e-01 -8.69406044e-01 -2.79530250e-02
2.79614031e-01 9.81898844e-01 5.07275045e-01 -7.15876698e-01
-1.59488887e-01 2.67381698e-01 1.94982812e-01 -1.08111119e+00
-9.84779000e-01 -4.34185952e-01 -1.81354408e-03 -9.46787074e-02
-4.60669577e-01 -2.33035162e-01 -1.22193670e+00 -3.06285828e-01
-2.75570583e-02 2.54899710e-01 1.07181215e+00 9.83913958e-01
6.02686465e-01 8.76978487e-02 9.60981607e-01 -9.68139648e-01
7.76297450e-02 -1.07577527e+00 1.78069863e-02 8.32034111e-01
-9.29394457e-03 8.67477094e-04 -3.08630258e-01 6.99988842e-01] | [11.100171089172363, 0.9403026103973389] |
47744946-d484-424b-a64b-800b74f9b5fa | implicit-diffusion-models-for-continuous | 2303.16491 | null | https://arxiv.org/abs/2303.16491v1 | https://arxiv.org/pdf/2303.16491v1.pdf | Implicit Diffusion Models for Continuous Super-Resolution | Image super-resolution (SR) has attracted increasing attention due to its wide applications. However, current SR methods generally suffer from over-smoothing and artifacts, and most work only with fixed magnifications. This paper introduces an Implicit Diffusion Model (IDM) for high-fidelity continuous image super-resolution. IDM integrates an implicit neural representation and a denoising diffusion model in a unified end-to-end framework, where the implicit neural representation is adopted in the decoding process to learn continuous-resolution representation. Furthermore, we design a scale-controllable conditioning mechanism that consists of a low-resolution (LR) conditioning network and a scaling factor. The scaling factor regulates the resolution and accordingly modulates the proportion of the LR information and generated features in the final output, which enables the model to accommodate the continuous-resolution requirement. Extensive experiments validate the effectiveness of our IDM and demonstrate its superior performance over prior arts. | ['Baochang Zhang', 'XianTong Zhen', 'Jianzhuang Liu', 'Xiaoyan Luo', 'Yanjing Li', 'Sheng Xu', 'Bohan Zeng', 'Xuhui Liu', 'Sicheng Gao'] | 2023-03-29 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Gao_Implicit_Diffusion_Models_for_Continuous_Super-Resolution_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Gao_Implicit_Diffusion_Models_for_Continuous_Super-Resolution_CVPR_2023_paper.pdf | cvpr-2023-1 | ['image-super-resolution'] | ['computer-vision'] | [ 5.61896920e-01 -8.40230733e-02 -1.04386501e-01 -3.31312239e-01
-8.12966406e-01 2.73362640e-02 5.54305971e-01 -6.35638356e-01
-8.22250620e-02 5.59916854e-01 5.91172695e-01 2.71658927e-01
-7.74181336e-02 -7.13331819e-01 -6.36460781e-01 -7.26182938e-01
3.12993944e-01 -3.57327849e-01 2.17046276e-01 -2.53805131e-01
1.12383917e-01 2.50945657e-01 -1.17629409e+00 3.24880600e-01
1.28100657e+00 1.00254333e+00 7.29056597e-01 4.38296467e-01
5.36624976e-02 7.02456772e-01 -2.71914750e-01 1.59552470e-02
3.64500880e-01 -6.37725770e-01 -2.69803613e-01 9.16833654e-02
5.11029959e-01 -5.88488698e-01 -7.50389040e-01 1.19437432e+00
6.75777435e-01 2.16730133e-01 3.85642469e-01 -4.18156445e-01
-1.35696042e+00 4.25687164e-01 -1.11259961e+00 6.29831016e-01
1.99905172e-01 1.18857607e-01 7.17209876e-01 -1.01451218e+00
5.82628429e-01 1.47097027e+00 4.82941598e-01 6.78362608e-01
-1.53880632e+00 -8.51738751e-01 1.14817299e-01 -6.26063021e-03
-1.44609678e+00 -6.14381790e-01 7.74664402e-01 -1.79853797e-01
4.09844875e-01 -1.37687966e-01 3.56277794e-01 1.02578831e+00
2.43655369e-01 4.96024430e-01 1.20155621e+00 -2.36622781e-01
1.49783477e-01 -1.27202794e-01 -2.60718346e-01 3.74927610e-01
1.74898803e-01 3.15259039e-01 -7.33531356e-01 1.80375069e-01
1.88653135e+00 9.63194948e-03 -5.53685844e-01 -8.27092156e-02
-8.95822227e-01 5.80518782e-01 6.80340528e-01 2.89840817e-01
-4.20605212e-01 2.58375674e-01 -6.22935370e-02 5.86503670e-02
6.88009858e-01 1.87612548e-01 -6.73414320e-02 1.62430018e-01
-1.14743781e+00 4.68577594e-02 -1.01894855e-01 8.13432515e-01
3.88492823e-01 3.78443807e-01 -3.97685736e-01 1.12883544e+00
2.09716961e-01 3.12494099e-01 3.59036893e-01 -1.23324490e+00
3.14849228e-01 3.66254956e-01 2.52741635e-01 -8.32391858e-01
-4.30202596e-02 -8.33655596e-01 -1.30451906e+00 2.66317755e-01
-1.21968001e-01 8.01007450e-02 -8.74899626e-01 2.04426122e+00
1.95559576e-01 3.97886276e-01 -1.08919300e-01 1.37354779e+00
6.23081505e-01 6.92025125e-01 2.27822848e-02 -4.31109488e-01
1.15138280e+00 -8.24204087e-01 -1.01618528e+00 -1.46252573e-01
-2.93079853e-01 -5.51759362e-01 1.24513829e+00 2.87031919e-01
-1.51108658e+00 -8.19117367e-01 -1.19175541e+00 -5.10463953e-01
3.56257677e-01 2.14062214e-01 3.13229382e-01 1.68705583e-01
-1.06408858e+00 6.72604024e-01 -8.05673063e-01 1.20461717e-01
5.74756145e-01 2.43772492e-01 -2.61671599e-02 -2.02275142e-01
-1.22494292e+00 6.17186904e-01 -7.77786225e-02 2.28526592e-01
-6.64975584e-01 -7.25864947e-01 -7.31310070e-01 1.22034229e-01
6.19620122e-02 -1.02359855e+00 9.42622423e-01 -1.00490117e+00
-1.77709222e+00 5.17551541e-01 -2.97053486e-01 -3.51236850e-01
5.82103848e-01 -3.51816595e-01 -5.46645463e-01 3.28929693e-01
-1.04629155e-02 4.65028435e-01 1.41480696e+00 -1.39416516e+00
-6.42971635e-01 -3.32545936e-01 -1.50475323e-01 3.42973650e-01
-3.98809463e-01 -1.41555360e-02 -6.08026326e-01 -1.14165747e+00
2.89238185e-01 -2.92878419e-01 -3.50699753e-01 1.74452335e-01
3.88584435e-02 3.64349037e-01 5.48839390e-01 -8.38314950e-01
1.52209878e+00 -2.59354448e+00 4.55418497e-01 -1.46542519e-01
4.71173406e-01 3.76204439e-02 -3.13266188e-01 -1.44072920e-01
-1.55634563e-02 -2.07519829e-02 -2.68259734e-01 -4.86679912e-01
-2.73831546e-01 3.71767618e-02 -4.90844071e-01 4.56071436e-01
4.42759365e-01 7.34286308e-01 -7.43557870e-01 -3.19767386e-01
8.11735690e-02 1.25054193e+00 -4.20199037e-01 2.31255472e-01
8.21566302e-03 8.83120537e-01 -5.06895781e-01 3.36507201e-01
8.99805129e-01 -5.54952800e-01 5.34302853e-02 -5.23981333e-01
-3.66934150e-01 2.41960287e-02 -1.20947754e+00 1.77850878e+00
-6.31606936e-01 4.32033032e-01 5.11600494e-01 -3.18976849e-01
9.61156785e-01 1.24276020e-01 2.54165858e-01 -1.14852810e+00
1.04664098e-02 1.62588969e-01 -2.81309158e-01 -1.64205045e-01
6.96507215e-01 -2.55215794e-01 3.65883887e-01 -7.13288970e-03
-2.94856727e-01 2.26989970e-01 -1.44282654e-01 1.70450404e-01
7.95954525e-01 2.36458197e-01 6.18152060e-02 -1.49954647e-01
5.98338246e-01 -5.61780870e-01 7.00752974e-01 3.56634349e-01
3.54022160e-02 1.00627899e+00 7.24152699e-02 -1.59271970e-01
-1.22939348e+00 -1.13022733e+00 -3.25636655e-01 9.35805023e-01
5.42610288e-01 -2.47566663e-02 -6.66700661e-01 4.73801559e-03
-3.81129116e-01 4.46042120e-01 -6.39326155e-01 -2.27500513e-01
-7.11567879e-01 -4.95879531e-01 1.25372499e-01 5.47855139e-01
9.19264495e-01 -6.93999529e-01 -3.32087338e-01 1.76880836e-01
-2.86552310e-01 -1.21751773e+00 -8.77971053e-01 -3.30535859e-01
-1.09871006e+00 -4.32500780e-01 -8.88215125e-01 -7.83048511e-01
7.54086196e-01 5.17123044e-01 7.47768641e-01 -2.42339760e-01
-1.51948005e-01 -2.84488499e-02 -5.10421768e-02 3.25970739e-01
-2.02230647e-01 -1.55245230e-01 4.69491258e-02 2.62018681e-01
-2.13766500e-01 -9.42214727e-01 -1.03775942e+00 2.50859588e-01
-1.07338190e+00 3.34130973e-01 9.27745879e-01 8.68113220e-01
9.12239432e-01 1.94635943e-01 7.52331197e-01 -5.95825613e-01
6.98198855e-01 -2.41751522e-01 -6.73006833e-01 -9.16531458e-02
-6.48474455e-01 4.26164158e-02 6.06222868e-01 -7.71402478e-01
-1.53477049e+00 -8.86017978e-02 7.64824003e-02 -7.48245001e-01
3.88865560e-01 2.26897776e-01 -4.13789719e-01 -2.19643891e-01
5.28071523e-01 4.15455043e-01 -3.83320525e-02 -6.88785434e-01
5.53545654e-01 4.26396102e-01 8.33581805e-01 -3.27572286e-01
9.35548544e-01 6.15270793e-01 -1.83638215e-01 -6.56131685e-01
-8.84474218e-01 -1.04930878e-01 -3.99031162e-01 -1.38206318e-01
7.40173280e-01 -1.47697341e+00 -3.35381478e-01 3.32037002e-01
-8.14966977e-01 -3.70124787e-01 -3.24641943e-01 4.20494288e-01
-4.81309474e-01 2.43819401e-01 -1.04274142e+00 -8.05618048e-01
-5.10774195e-01 -9.67892766e-01 9.90566075e-01 5.60314238e-01
1.23135969e-01 -5.61461806e-01 -2.47183129e-01 4.13936526e-01
8.83584917e-01 4.33417968e-02 7.88319767e-01 4.28306192e-01
-8.99065375e-01 2.25619838e-01 -8.39291692e-01 4.33084458e-01
8.86781812e-02 -3.85050118e-01 -8.84331942e-01 -4.74312961e-01
1.61857173e-01 -5.57262860e-02 1.09425080e+00 6.95277870e-01
1.13313723e+00 -2.65826613e-01 -3.78466547e-02 9.95503426e-01
1.67985177e+00 -1.59576714e-01 9.61135387e-01 2.49099299e-01
6.72127664e-01 1.20224103e-01 5.14506698e-01 4.57654178e-01
2.17904598e-01 5.94796002e-01 2.54821271e-01 -2.88089812e-01
-6.16186917e-01 -3.36846977e-01 3.84985596e-01 7.93636799e-01
-2.84475327e-01 1.55353025e-01 -2.34591454e-01 3.66081238e-01
-1.61123443e+00 -9.52363491e-01 1.35123566e-01 2.20816779e+00
1.19431365e+00 1.60276100e-01 -1.58667266e-01 -1.46702528e-01
7.88890064e-01 3.45667392e-01 -9.36255217e-01 6.22849390e-02
-3.04485232e-01 2.47199610e-01 3.53532672e-01 5.61774135e-01
-8.05696607e-01 7.31707394e-01 6.30589199e+00 1.03773403e+00
-1.16834998e+00 2.28212699e-01 6.64654076e-01 -2.37183034e-01
-4.11545128e-01 -3.06337923e-01 -7.39675820e-01 4.34196681e-01
5.96808255e-01 -1.06758825e-01 6.45778716e-01 5.58739662e-01
5.92887402e-01 1.30319014e-01 -6.51634991e-01 1.04402673e+00
-1.61066204e-01 -1.19341004e+00 1.58650875e-01 1.30621284e-01
8.62980247e-01 -2.34544203e-01 6.07884705e-01 7.60157928e-02
2.48918906e-02 -1.07375050e+00 6.91029966e-01 7.42190242e-01
1.43162632e+00 -8.22075188e-01 1.66607156e-01 1.58022061e-01
-1.31023324e+00 -3.30903620e-01 -4.23923045e-01 1.40382618e-01
2.60391444e-01 5.78309715e-01 1.41243398e-01 4.71188068e-01
6.64094031e-01 7.38973975e-01 -1.45320281e-01 6.55856729e-01
-1.88542858e-01 4.02432978e-01 7.59262666e-02 7.25133538e-01
-1.81983620e-01 -4.38370138e-01 5.38083196e-01 1.00404274e+00
3.16078216e-01 4.35401052e-01 -5.86927775e-03 1.16782391e+00
-2.43288055e-01 -2.00492591e-01 5.54702953e-02 1.66520566e-01
5.28319478e-01 1.23218679e+00 -2.72776008e-01 -2.70785876e-02
-4.16588128e-01 1.20239186e+00 2.85641611e-01 5.97394288e-01
-7.70497561e-01 -1.16575755e-01 7.69082844e-01 4.36220527e-01
5.25357544e-01 -3.31504673e-01 -5.48067808e-01 -1.17797184e+00
1.24471165e-01 -8.33887815e-01 7.19247237e-02 -9.35461640e-01
-1.30559719e+00 6.52455986e-01 -4.49968934e-01 -1.37654877e+00
1.53804004e-01 -3.06158066e-02 -2.95051962e-01 1.24177659e+00
-1.93860316e+00 -1.19308376e+00 -4.61683929e-01 4.82543945e-01
6.52550519e-01 1.88540250e-01 3.88175935e-01 5.85748315e-01
-7.81309843e-01 5.27792573e-01 2.32745141e-01 -9.20169801e-02
7.61488557e-01 -9.51617301e-01 2.36287951e-01 1.02780664e+00
-5.48092663e-01 7.63660789e-01 7.45567799e-01 -6.83677852e-01
-1.40508294e+00 -1.23547697e+00 3.06919903e-01 -2.32137218e-02
5.52127421e-01 -1.87878788e-01 -1.24816215e+00 3.59372079e-01
6.27072230e-02 2.13827580e-01 2.33628929e-01 -3.48458499e-01
-5.51117420e-01 -3.37886959e-01 -1.08486760e+00 6.15469873e-01
1.14957452e+00 -5.96544683e-01 -3.23237598e-01 -2.19270140e-01
1.03322732e+00 -4.77353752e-01 -1.03422546e+00 3.72263342e-01
5.24102747e-01 -8.15112829e-01 1.18231273e+00 1.04450854e-02
9.68918741e-01 -5.59277236e-01 -2.02584967e-01 -1.09475946e+00
-9.49330509e-01 -6.26498461e-01 -5.74196815e-01 1.29193151e+00
1.19724870e-01 -3.81971508e-01 3.68458480e-01 5.68024993e-01
4.39807847e-02 -7.51948774e-01 -7.68423975e-01 -6.39966130e-01
-1.17198758e-01 1.86805919e-01 4.31720883e-01 8.38938832e-01
-4.61230993e-01 5.42743266e-01 -6.39267981e-01 5.28532505e-01
9.84320283e-01 1.15382358e-01 3.53659540e-01 -9.27816451e-01
-3.55411291e-01 -5.03257036e-01 2.28656605e-02 -1.50703919e+00
-3.32681984e-01 -3.90652150e-01 1.33317960e-02 -1.66378200e+00
3.96050096e-01 -2.44232014e-01 -3.84442687e-01 -1.04240187e-01
-3.35305959e-01 3.34549278e-01 -4.93666269e-02 5.32382607e-01
-4.28248972e-01 8.46864104e-01 1.73900211e+00 2.55420357e-01
-4.83633369e-01 -3.36573869e-01 -9.92353916e-01 5.96135259e-01
5.59311330e-01 -2.73184534e-02 -4.64956075e-01 -7.46712625e-01
1.94218874e-01 2.95080960e-01 2.84017444e-01 -8.13483000e-01
1.78565979e-01 -9.82159153e-02 7.72903502e-01 -3.19274306e-01
5.01907945e-01 -6.16226673e-01 7.45819882e-02 9.09436643e-02
-5.63008487e-01 -3.17731559e-01 -2.15965752e-02 9.51763034e-01
-1.86976567e-01 4.36909080e-01 1.34446537e+00 1.26990139e-01
-5.42388558e-01 6.12065732e-01 8.45128745e-02 -6.18659891e-02
4.96498227e-01 -2.73152828e-01 -2.13916585e-01 -3.59330654e-01
-6.07668281e-01 1.68409675e-01 6.17843449e-01 3.55417311e-01
9.30528462e-01 -1.30784082e+00 -9.74446476e-01 2.20192283e-01
-2.36715078e-01 2.78699309e-01 6.19940341e-01 7.84738839e-01
-1.57191366e-01 -2.28004992e-01 -5.79571836e-02 -2.91801751e-01
-8.95822048e-01 6.29671633e-01 2.95100629e-01 -1.52075127e-01
-1.10366857e+00 6.69857681e-01 5.35052061e-01 2.79002845e-01
1.84700325e-01 7.83633441e-03 -3.06084305e-01 -3.87400389e-01
1.10236871e+00 3.21168274e-01 -5.58893979e-01 -5.44028759e-01
8.84421542e-02 7.55272031e-01 -2.94128656e-01 -1.13428190e-01
1.48293531e+00 -6.89372957e-01 -3.80130894e-02 3.62939775e-01
7.59781480e-01 3.94787453e-02 -1.87270105e+00 -6.70248330e-01
-4.26126033e-01 -7.14234412e-01 5.17534792e-01 -6.47487760e-01
-1.25222814e+00 6.23234093e-01 6.73946679e-01 -2.27743506e-01
1.64885843e+00 -1.72676831e-01 9.93149459e-01 -3.36730242e-01
2.53246486e-01 -9.49776053e-01 2.15663299e-01 3.49886656e-01
1.01989496e+00 -1.04612219e+00 1.72570437e-01 -5.61157227e-01
-5.14354825e-01 8.49797487e-01 5.68313420e-01 -3.11200082e-01
3.75070959e-01 3.42739850e-01 -1.06722347e-01 1.64066926e-01
-7.29216516e-01 2.01645447e-03 2.80762464e-01 4.91351575e-01
4.04800117e-01 -2.09453017e-01 -4.27783579e-01 9.23230767e-01
1.73610523e-01 3.13673168e-01 3.28890324e-01 4.07617122e-01
-5.27179897e-01 -6.06688976e-01 -2.70281374e-01 1.57648161e-01
-4.57889408e-01 -2.76359111e-01 2.10249096e-01 3.19944292e-01
2.99140643e-02 7.77497947e-01 -1.23147992e-02 -2.51350433e-01
2.70950854e-01 -5.46878934e-01 5.14078021e-01 -2.65185118e-01
-2.88765907e-01 4.96020705e-01 -4.26983744e-01 -7.24095941e-01
-4.22951400e-01 -3.11499745e-01 -1.30030191e+00 -2.70596325e-01
-3.59687716e-01 -2.53506929e-01 4.84485686e-01 6.02859020e-01
6.38103783e-01 8.71998429e-01 8.08229566e-01 -8.58405113e-01
-6.04499221e-01 -9.30162966e-01 -8.20096970e-01 2.14245573e-01
5.61620295e-01 -5.10868371e-01 -3.53290051e-01 2.11641565e-01] | [11.26628303527832, -1.9717836380004883] |
83f25537-6cae-4ca9-8abf-3c72b5a1da0c | learning-disentangled-representations-in | null | null | https://openreview.net/forum?id=rm893aOESdu | https://openreview.net/pdf?id=rm893aOESdu | Learning Disentangled Representations in Natural Language Definitions with Semantic Role Labeling Supervision | Disentangling the encodings of neural models is a fundamental aspect for improving interpretability, semantic control and downstream task performance in Natural Language Processing. However, most disentanglement methods are unsupervised or rely on synthetic datasets with known generative factors. We argue that recurrent syntactic and semantic regularities in textual data can be used to provide the models with both structural biases and generative factors. We leverage the semantic structures present in a representative and semantically dense category of sentence types, definitional sentences, for training a Variational Autoencoder to learn disentangled representations. Our experimental results show that the proposed model outperforms unsupervised baselines on several qualitative and quantitative benchmarks for disentanglement, and it also improves the results in the downstream task of definition modeling. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['semantic-role-labeling'] | ['natural-language-processing'] | [ 1.15630299e-01 4.50145334e-01 -4.17322725e-01 -6.56199276e-01
-6.34077191e-01 -8.40687931e-01 1.09298599e+00 5.68668731e-02
-2.83298671e-01 6.03401601e-01 1.04239666e+00 -2.02121139e-01
-1.31579399e-01 -8.14426184e-01 -6.65417373e-01 -6.27159417e-01
4.08456236e-01 7.27125347e-01 -5.63401043e-01 -2.69433409e-01
2.06253724e-03 8.94133672e-02 -1.14639473e+00 5.85549176e-01
9.34452474e-01 5.38181186e-01 9.29368511e-02 4.62772578e-01
-2.68634185e-02 7.74749935e-01 -5.65453708e-01 -8.52716446e-01
1.86483189e-02 -3.16625565e-01 -6.81948900e-01 2.32237242e-02
1.94493100e-01 -2.57988870e-01 -5.95946968e-01 1.20453501e+00
2.22643822e-01 1.68167263e-01 1.17784619e+00 -9.36774611e-01
-9.86360252e-01 1.41756523e+00 -2.73845375e-01 3.60080957e-01
4.86903489e-02 2.27385372e-01 1.89939272e+00 -1.02814710e+00
7.17595041e-01 1.69170666e+00 2.30756462e-01 6.57088816e-01
-1.80507481e+00 -5.78120649e-01 2.05740869e-01 1.55657470e-01
-1.01142704e+00 -5.64654171e-01 1.22225630e+00 -4.76164073e-01
1.03503942e+00 3.35466683e-01 5.84564149e-01 2.02230358e+00
8.85442942e-02 1.16721690e+00 9.70932782e-01 -1.31915793e-01
2.21375659e-01 9.49966908e-02 5.06140888e-01 6.85202539e-01
5.89480400e-01 2.78362632e-01 -5.67395627e-01 -3.04940064e-02
7.27976978e-01 1.08743934e-02 -2.71659464e-01 -2.34978423e-01
-1.29716194e+00 1.27406335e+00 5.00924349e-01 1.58444300e-01
-2.67569214e-01 2.92407304e-01 4.35683489e-01 1.24203973e-01
6.27068698e-01 1.12265873e+00 -5.31008422e-01 -5.79077452e-02
-6.74770594e-01 5.46766460e-01 7.07425237e-01 8.66318345e-01
6.53576076e-01 2.48837829e-01 -4.41900462e-01 7.65378237e-01
6.52063847e-01 5.21905661e-01 5.79701185e-01 -1.01045299e+00
7.99150586e-01 8.11432660e-01 -1.98688164e-01 -1.04002178e+00
-1.36946633e-01 -4.55936372e-01 -7.89829433e-01 -2.86916941e-01
2.88357347e-01 -1.58040598e-01 -7.65450954e-01 1.93640018e+00
-2.45604947e-01 -2.86491543e-01 3.76214385e-01 8.34533811e-01
9.34559286e-01 5.00041246e-01 2.23299250e-01 1.97018251e-01
1.61537695e+00 -7.65055001e-01 -1.04974842e+00 -8.24431717e-01
4.65804458e-01 -2.16824308e-01 1.18299377e+00 -6.97484389e-02
-9.58213031e-01 -3.87173563e-01 -1.29960191e+00 -4.36115474e-01
-1.95370704e-01 1.12657622e-01 8.99106383e-01 3.01203161e-01
-5.11045694e-01 5.47761023e-01 -1.00479817e+00 3.78270656e-01
7.57233799e-01 -8.52427855e-02 -2.41803929e-01 2.59230305e-02
-1.63101840e+00 8.10534656e-01 5.68721056e-01 5.97812375e-03
-1.26514757e+00 -9.36340213e-01 -1.27101898e+00 4.05388594e-01
2.63425231e-01 -1.25389361e+00 1.06968403e+00 -5.95624447e-01
-1.13449109e+00 8.44194591e-01 -1.75460055e-01 -5.85091412e-01
2.12409824e-01 -2.80900866e-01 8.77944008e-02 -1.53387487e-01
6.22442514e-02 4.90904361e-01 7.88863420e-01 -1.35037637e+00
2.64790773e-01 -5.23498774e-01 2.58797288e-01 3.76558095e-01
-5.17323166e-02 -3.36552471e-01 2.35510305e-01 -8.50674689e-01
7.64863268e-02 -5.51986873e-01 -1.99081853e-01 -3.29908133e-01
-1.00366843e+00 -2.64782250e-01 2.59459108e-01 -6.02391601e-01
8.97804856e-01 -1.85328114e+00 8.24782550e-01 -2.93445379e-01
7.80671179e-01 -1.79422840e-01 -9.63949040e-02 3.42544049e-01
-1.32601351e-01 5.79250038e-01 -2.39423603e-01 -8.51987123e-01
3.54286999e-01 4.55871969e-01 -6.52528226e-01 1.27471805e-01
4.41657871e-01 1.33481717e+00 -1.09028924e+00 -2.98887223e-01
8.79927650e-02 2.95488119e-01 -7.76154995e-01 4.52308446e-01
-6.00503623e-01 2.72310257e-01 -4.34939623e-01 1.66194394e-01
2.06018284e-01 -4.63672489e-01 3.52742642e-01 -4.56678689e-01
6.47167146e-01 1.05505526e+00 -5.59538245e-01 1.89268720e+00
-4.67832416e-01 8.51644635e-01 -3.85643393e-01 -1.07903302e+00
6.44567907e-01 4.16967094e-01 -2.79463857e-01 -3.29073459e-01
2.74012387e-01 -1.74366593e-01 2.75935769e-01 -6.84702516e-01
4.43334818e-01 -3.44835997e-01 -3.67758244e-01 6.20923638e-01
4.01232153e-01 -3.36670816e-01 5.53925149e-02 6.12882912e-01
8.27273548e-01 -1.40304402e-01 4.08649623e-01 -3.65808845e-01
-1.48058787e-01 -3.13753128e-01 5.66949546e-01 7.20944941e-01
-5.17225154e-02 5.27720928e-01 1.05404246e+00 -1.29438341e-01
-1.18368375e+00 -1.63136709e+00 -7.70748630e-02 7.62562573e-01
-1.11221164e-01 -5.87808371e-01 -6.95547879e-01 -6.58203065e-01
-7.21220449e-02 1.46332812e+00 -1.00785363e+00 -5.67463815e-01
-2.78225154e-01 -8.90012562e-01 7.46325970e-01 8.07462215e-01
2.08526418e-01 -8.13683510e-01 -2.63280779e-01 -2.14404389e-01
-3.69232267e-01 -1.21382833e+00 -8.28320459e-02 1.85260370e-01
-8.40976655e-01 -1.07641196e+00 -2.16349661e-02 -3.43649864e-01
6.57571971e-01 -1.09639447e-02 1.51916325e+00 -3.73508364e-01
1.32447124e-01 -1.20847866e-01 -2.23769881e-02 -2.90589094e-01
-6.73700809e-01 1.14870727e-01 -5.64906225e-02 -2.12704778e-01
4.77393806e-01 -7.66373813e-01 -5.07778049e-01 -2.76167780e-01
-1.06043994e+00 6.38575792e-01 4.65386331e-01 1.10751438e+00
2.57000774e-01 -4.02061462e-01 4.40487415e-01 -1.40156174e+00
1.12997603e+00 -7.10456431e-01 -2.01529935e-01 1.25161126e-01
-7.19758511e-01 8.49269152e-01 5.15691340e-01 -3.12824398e-01
-1.34731972e+00 -4.95667607e-01 3.72574851e-02 -4.67708081e-01
-2.93758541e-01 6.63793385e-01 -6.06432796e-01 1.09372187e+00
8.27239871e-01 1.49440989e-01 5.65978773e-02 -3.42062503e-01
9.98823881e-01 2.51774430e-01 1.33198008e-01 -7.70219147e-01
7.46993124e-01 3.41352075e-01 -3.63265544e-01 -4.09415007e-01
-1.36812735e+00 2.70249434e-02 -4.41258997e-01 4.13131207e-01
1.02884352e+00 -1.07419550e+00 -3.44340712e-01 1.77447107e-02
-1.65311348e+00 3.62726650e-03 -2.81378955e-01 2.62188673e-01
-5.03065288e-01 2.40431894e-02 -6.96435750e-01 -4.89409506e-01
-1.95451781e-01 -1.36093056e+00 1.26010799e+00 1.28098562e-01
-5.90748131e-01 -1.54786313e+00 2.43488282e-01 6.34355545e-01
1.77010432e-01 3.35025042e-01 1.36538911e+00 -1.10073245e+00
-5.84228337e-01 -6.98940456e-02 -1.26239121e-01 2.70661563e-01
4.67879511e-02 -1.86746478e-01 -1.23744678e+00 1.83326274e-01
2.98242420e-01 -5.55690169e-01 1.26181459e+00 2.18544662e-01
1.08095908e+00 -8.45105648e-01 -2.44173244e-01 4.50972497e-01
1.03512394e+00 -3.66706759e-01 4.91035312e-01 -2.85531431e-01
9.71513271e-01 7.59735525e-01 -7.12164566e-02 1.04282387e-01
4.92959589e-01 4.57695365e-01 3.90322655e-01 2.16763929e-01
2.06417106e-02 -6.81932569e-01 4.37294811e-01 9.95210826e-01
8.23390633e-02 -3.56109530e-01 -8.03950727e-01 4.49194461e-01
-1.86539745e+00 -1.00308526e+00 1.61566541e-01 1.43974972e+00
1.25723600e+00 1.70654312e-01 -5.02325356e-01 -1.19499043e-02
3.56584191e-01 8.10964882e-01 -6.07893407e-01 -2.95986325e-01
-2.81156838e-01 2.37583995e-01 -9.03501350e-04 6.99073434e-01
-8.39109540e-01 9.52283978e-01 6.32381392e+00 6.73173785e-01
-4.80007976e-01 1.71877384e-01 6.29798114e-01 -2.78478563e-01
-1.25593293e+00 -2.64889468e-03 -6.82507455e-01 2.58584261e-01
1.01176167e+00 -2.81302661e-01 4.10981983e-01 6.91982448e-01
3.10626298e-01 3.39640468e-01 -1.77993917e+00 7.96232939e-01
2.29596287e-01 -1.61333561e+00 4.52582926e-01 4.93309041e-03
7.49204814e-01 -9.48406830e-02 2.34623864e-01 3.06654274e-01
7.24318027e-01 -1.44650519e+00 6.62075698e-01 5.36635458e-01
3.98676723e-01 -4.23679054e-01 6.08133137e-01 4.73563641e-01
-5.39050639e-01 3.14779997e-01 -2.53782570e-01 -1.67347804e-01
1.00014150e-01 5.88064194e-01 -9.59881246e-01 3.14850509e-01
-3.34714465e-02 9.33431447e-01 -5.91226220e-01 4.52811606e-02
-1.10805058e+00 7.55155444e-01 2.47163400e-01 -3.62617522e-01
-8.15243460e-03 -1.97778150e-01 8.18679273e-01 1.05675423e+00
-4.85878706e-01 -6.82692006e-02 -4.12643760e-01 2.06155443e+00
-4.28085834e-01 -4.54293460e-01 -8.98142040e-01 -6.71283722e-01
5.87710500e-01 1.04801381e+00 -1.36047184e-01 -4.34306234e-01
-7.27475882e-02 7.70581782e-01 8.19673121e-01 7.22235322e-01
-8.65950942e-01 -1.54032633e-01 1.18357432e+00 -3.25550467e-01
-7.40085021e-02 -4.66657549e-01 -8.11552048e-01 -1.80769145e+00
-1.85393259e-01 -8.44070554e-01 7.12341070e-02 -6.85909271e-01
-1.50240183e+00 4.58718151e-01 2.93585569e-01 -5.55200994e-01
-8.41131806e-01 -7.97386646e-01 -6.04584217e-01 9.23601151e-01
-9.92616177e-01 -1.16832244e+00 8.64964500e-02 1.92513719e-01
1.03637338e+00 -1.73102856e-01 1.09337354e+00 -2.66031921e-01
-8.30230296e-01 6.12634361e-01 -5.67416213e-02 5.03893018e-01
1.44362390e-01 -1.65215945e+00 5.69341362e-01 1.01217484e+00
8.44921529e-01 1.18076956e+00 1.02936029e+00 -3.61571670e-01
-1.40589619e+00 -8.75227094e-01 8.39331150e-01 -9.96900678e-01
7.82596946e-01 -9.71559644e-01 -8.03706467e-01 9.91419315e-01
5.36552131e-01 -3.28648657e-01 1.13420963e+00 5.59665024e-01
-8.88335466e-01 4.06058401e-01 -6.89384878e-01 9.60410357e-01
1.09207559e+00 -1.03776348e+00 -1.36219311e+00 3.10056031e-01
1.33673012e+00 -3.17750722e-01 -7.54018962e-01 1.25287980e-01
4.77353483e-01 -7.40608752e-01 9.86387551e-01 -1.34308791e+00
1.34905159e+00 1.61641166e-01 -2.52647638e-01 -1.85953844e+00
-2.37155586e-01 -2.32555196e-01 -4.50499684e-01 1.18679941e+00
9.18425918e-01 -4.93509889e-01 5.62548876e-01 9.66854811e-01
2.30651349e-02 -7.93024361e-01 -6.66208565e-01 -2.43580163e-01
3.71113628e-01 -6.07454002e-01 4.75998133e-01 1.03901148e+00
-3.05599701e-02 1.16943240e+00 -3.54156569e-02 -1.14298910e-01
6.60908580e-01 1.78095654e-01 4.52088565e-01 -1.04542208e+00
-6.74494326e-01 -5.65898061e-01 -2.56899476e-01 -1.15588605e+00
7.06363261e-01 -1.26624763e+00 -1.77125975e-01 -1.52171230e+00
5.70822418e-01 1.02635279e-01 7.33570382e-02 2.17100114e-01
-4.83880937e-01 -3.08447123e-01 -5.09365238e-02 1.01271376e-01
-3.46789002e-01 1.11395669e+00 1.33715463e+00 -4.38924879e-01
2.66382247e-01 -3.29680771e-01 -1.15511858e+00 7.80337572e-01
4.94725287e-01 -4.61391211e-01 -7.95584142e-01 -9.03578103e-01
6.41538739e-01 -9.37413797e-02 4.67383772e-01 -7.23701864e-02
-2.40935221e-01 -3.48136723e-01 2.67507941e-01 -1.30863056e-01
4.42274153e-01 -3.90642017e-01 -2.60522336e-01 2.00353399e-01
-1.10280263e+00 -7.28776306e-02 -7.10235611e-02 9.22267497e-01
-1.81721941e-01 -1.81529552e-01 4.20458645e-01 -2.42760807e-01
-1.83015227e-01 1.16683461e-01 -2.43163958e-01 6.77102625e-01
4.72376525e-01 2.64284700e-01 -6.87392175e-01 -2.65788287e-01
-8.18573415e-01 2.01912294e-03 2.52952516e-01 7.39342034e-01
8.15593898e-01 -1.44007885e+00 -8.05971265e-01 2.52193451e-01
1.63207337e-01 2.60596514e-01 -1.31186709e-01 4.76493269e-01
-1.92459300e-01 5.54499984e-01 1.45065606e-01 -3.88190448e-01
-6.22283459e-01 3.06985110e-01 4.74390626e-01 -4.74393278e-01
-2.40156174e-01 1.02139056e+00 7.43958414e-01 -4.02409941e-01
-8.34917501e-02 -5.99060655e-01 -1.91775173e-01 1.54314665e-02
3.09320092e-01 3.93498167e-02 -2.48132080e-01 -3.60776901e-01
-1.21758372e-01 -1.93492711e-01 -2.40693241e-01 -3.98390025e-01
1.31111526e+00 -3.54743488e-02 -9.99220908e-02 9.04287934e-01
1.43173635e+00 5.90209523e-03 -1.19205713e+00 -3.07138622e-01
-3.19625884e-02 -3.23568106e-01 5.63321821e-02 -5.71021676e-01
-9.22571301e-01 1.14313340e+00 -1.58036351e-01 1.88412607e-01
3.51121336e-01 4.27574098e-01 3.74421775e-01 3.73803258e-01
-1.04811251e-01 -5.20125568e-01 3.49526346e-01 5.33482790e-01
1.24907422e+00 -1.25989580e+00 -3.70373368e-01 -4.28932101e-01
-8.42356026e-01 9.45015132e-01 6.30394459e-01 -4.57193434e-01
5.66357315e-01 2.94862211e-01 -2.60529369e-01 -5.32754242e-01
-1.25287831e+00 1.62333637e-01 6.47650898e-01 4.12206411e-01
6.85787797e-01 4.43057925e-01 -2.48673260e-01 1.19210052e+00
-5.63868046e-01 -8.71939600e-01 3.11771780e-01 8.22788775e-02
-4.69078831e-02 -9.13557529e-01 2.54727870e-01 6.03871942e-01
-2.59099126e-01 -5.62354267e-01 -6.66509330e-01 6.24653459e-01
-2.11220935e-01 6.82512999e-01 1.62172988e-01 -3.71937603e-01
-6.50461465e-02 2.70090401e-01 5.85214019e-01 -9.07908201e-01
-3.37729752e-01 -3.06694537e-01 4.28678513e-01 -5.42814851e-01
-1.70547336e-01 -5.75785100e-01 -1.14403152e+00 -1.63212921e-02
-1.52238190e-01 1.06393136e-01 4.73555058e-01 1.33145750e+00
4.85216647e-01 8.34274113e-01 3.21524560e-01 -4.57041591e-01
-1.00196946e+00 -1.39447045e+00 -3.79799426e-01 8.20074975e-01
2.67394215e-01 -8.49117100e-01 -7.48268008e-01 6.94026649e-02] | [9.341492652893066, 5.052163124084473] |
c9d571f5-9490-42bb-97eb-1fcb610bfcb0 | sequence-to-sequence-model-with-transformer | 2306.05012 | null | https://arxiv.org/abs/2306.05012v1 | https://arxiv.org/pdf/2306.05012v1.pdf | Sequence-to-Sequence Model with Transformer-based Attention Mechanism and Temporal Pooling for Non-Intrusive Load Monitoring | This paper presents a novel Sequence-to-Sequence (Seq2Seq) model based on a transformer-based attention mechanism and temporal pooling for Non-Intrusive Load Monitoring (NILM) of smart buildings. The paper aims to improve the accuracy of NILM by using a deep learning-based method. The proposed method uses a Seq2Seq model with a transformer-based attention mechanism to capture the long-term dependencies of NILM data. Additionally, temporal pooling is used to improve the model's accuracy by capturing both the steady-state and transient behavior of appliances. The paper evaluates the proposed method on a publicly available dataset and compares the results with other state-of-the-art NILM techniques. The results demonstrate that the proposed method outperforms the existing methods in terms of both accuracy and computational efficiency. | ['Abouzar Estebsari', 'Roozbeh Rajabi', 'Mohammad Irani Azad'] | 2023-06-08 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [-1.28406577e-03 -3.98828954e-01 1.14080094e-01 -4.97204304e-01
-7.43107557e-01 -1.47341415e-01 5.15057862e-01 -1.03108548e-01
-6.43843561e-02 7.03508735e-01 3.35507900e-01 -1.52988583e-01
-1.27483279e-01 -7.03098357e-01 -4.06868905e-01 -7.43977427e-01
-2.38281831e-01 2.13217944e-01 1.69217631e-01 -2.82695651e-01
1.86200351e-01 3.74740750e-01 -1.46493900e+00 3.00652146e-01
6.16099060e-01 1.31072021e+00 2.92445540e-01 6.73205316e-01
5.99561967e-02 1.49523079e+00 -9.46807861e-01 -6.48045167e-02
3.90922651e-02 -1.20832868e-01 -7.86909044e-01 -2.77499706e-01
8.77645463e-02 -4.07139868e-01 -2.72365779e-01 8.00038815e-01
9.93659735e-01 1.57359153e-01 2.97003686e-01 -1.42970407e+00
-4.38742429e-01 5.47216356e-01 -5.37547648e-01 7.55844116e-01
5.52425086e-01 1.75982475e-01 7.30399251e-01 -8.83851275e-02
-1.80399165e-01 1.07787573e+00 9.89799917e-01 5.37517071e-02
-1.07413399e+00 -9.25107479e-01 5.41696474e-02 5.79941273e-01
-1.36111140e+00 -9.41748321e-02 8.20417941e-01 -2.67038465e-01
1.93824339e+00 4.23698366e-01 3.95213723e-01 1.13090122e+00
7.78251350e-01 8.44367027e-01 9.16687548e-01 -1.19531758e-01
3.19830269e-01 -1.25761762e-01 3.94844919e-01 6.06874228e-01
-2.43098333e-01 -3.59825119e-02 -2.75063336e-01 -3.04776311e-01
2.57667482e-01 7.63525143e-02 1.41883761e-01 3.68036807e-01
-7.51780450e-01 7.01134861e-01 3.44528943e-01 6.64295614e-01
-7.46640682e-01 2.50910461e-01 1.00939274e+00 -8.64098296e-02
6.15788877e-01 2.25479469e-01 -6.19855404e-01 -4.58301246e-01
-1.16553104e+00 -1.00036539e-01 8.46984684e-01 9.69693542e-01
2.43431970e-01 3.24211627e-01 -7.32025564e-01 5.88938832e-01
2.36887634e-01 2.74574757e-01 9.37508225e-01 -5.72188199e-01
4.42679256e-01 6.91007018e-01 8.58516246e-03 -6.80917561e-01
-8.35473239e-01 -2.61502177e-01 -1.04845858e+00 -1.11174576e-01
-4.26601768e-01 -5.60169443e-02 -8.79402757e-01 1.85800374e+00
7.77004734e-02 6.20731771e-01 -1.87462211e-01 3.11592042e-01
6.73788071e-01 1.07878566e+00 3.88997823e-01 -4.10129607e-01
1.25371873e+00 -9.63240743e-01 -1.16515505e+00 3.07764143e-01
4.93469805e-01 -5.40519953e-01 9.39330757e-01 2.40116328e-01
-8.83605599e-01 -6.32069826e-01 -9.84225631e-01 1.47311732e-01
-5.53247750e-01 -1.16143443e-01 8.82609040e-02 6.33947730e-01
-9.54636753e-01 5.61950386e-01 -9.78009224e-01 -7.45209038e-01
5.78567982e-01 5.84124207e-01 1.53432220e-01 1.97931230e-01
-1.26314163e+00 9.93174255e-01 4.90687877e-01 6.63547665e-02
-1.08591759e+00 -7.43973017e-01 -7.60696590e-01 3.67074668e-01
3.30208205e-02 -4.05697256e-01 1.51122713e+00 -5.42863905e-01
-1.68490791e+00 2.35131189e-01 -1.35512382e-01 -6.15453660e-01
3.53965014e-01 -3.87244761e-01 -6.60965085e-01 -2.21702546e-01
7.95092713e-03 2.19834819e-02 5.73269725e-01 -5.15624702e-01
-6.58356011e-01 -2.82134145e-01 -2.69483477e-01 -2.48929650e-01
-2.52702475e-01 2.39262849e-01 -3.60171534e-02 -6.10214055e-01
-5.90558529e-01 -6.06238663e-01 -7.45107606e-02 -7.98650265e-01
-5.90181291e-01 -5.36024511e-01 1.51938045e+00 -1.00444543e+00
1.53692734e+00 -2.04110932e+00 -5.05441904e-01 -7.16005117e-02
-2.62606442e-01 7.38065422e-01 -4.43201847e-02 7.10012555e-01
-2.55087554e-01 -1.52334541e-01 -2.23833308e-01 -7.04969883e-01
1.29660457e-01 4.41270322e-01 6.17830977e-02 3.77642304e-01
2.24512041e-01 1.35318196e+00 -7.87279367e-01 -3.15840513e-01
7.47601330e-01 5.70997238e-01 2.76812632e-03 5.62273085e-01
-1.42609596e-01 3.18770200e-01 -2.15315983e-01 5.14972448e-01
5.78594983e-01 -6.22439645e-02 7.28615671e-02 -3.45784098e-01
-3.03750277e-01 4.03285712e-01 -8.50345433e-01 1.51181066e+00
-7.49062955e-01 5.26742756e-01 -3.28466415e-01 -1.30420232e+00
9.30838227e-01 6.98149443e-01 8.02276909e-01 -9.84331250e-01
5.75563908e-01 -2.53037572e-01 -2.80935585e-01 -7.63903201e-01
1.78506285e-01 -6.11132309e-02 -2.12020740e-01 3.44706774e-01
1.82692140e-01 2.55185157e-01 1.27044708e-01 -2.38555327e-01
1.68584156e+00 -5.06150387e-02 4.89566952e-01 -6.04202926e-01
6.67266607e-01 -5.55947304e-01 7.43874669e-01 7.07342744e-01
-5.07423520e-01 -1.75684974e-01 1.45250633e-01 -5.67447543e-01
-9.97824490e-01 -7.74002373e-01 1.44959003e-01 9.71919656e-01
-5.18100917e-01 -2.91035503e-01 -7.24167943e-01 -7.24111021e-01
-6.37602583e-02 1.18800354e+00 -7.41564691e-01 -2.04487175e-01
-5.56835175e-01 -1.15116274e+00 6.39278710e-01 9.14873362e-01
9.52556252e-01 -1.43571055e+00 -1.02171886e+00 8.23396921e-01
-3.39796126e-01 -1.00446391e+00 -3.33021581e-01 6.84671164e-01
-6.20435596e-01 -8.75778198e-01 -2.44191140e-01 -5.93788862e-01
-5.06978109e-02 -2.22087473e-01 1.30002964e+00 -4.38119471e-01
-6.50048137e-01 6.19714521e-02 -2.08212182e-01 -8.68599236e-01
-2.53462642e-01 3.73458773e-01 -2.03975007e-01 -3.67587656e-02
8.99967790e-01 -9.16339099e-01 -3.50692153e-01 2.29607627e-01
-7.42703497e-01 -4.40574884e-01 3.49283397e-01 4.87567961e-01
1.59535319e-01 6.87591076e-01 8.90559793e-01 -5.75092196e-01
6.96757734e-01 -6.74359679e-01 -8.14990401e-01 1.17119573e-01
-7.64656782e-01 -2.48179697e-02 8.29209745e-01 -5.16999304e-01
-1.20823801e+00 2.18511722e-03 -4.53169703e-01 -3.11019629e-01
-2.18045592e-01 1.16340846e-01 -3.15204442e-01 4.19500768e-01
7.10247010e-02 2.68062949e-01 -3.54254276e-01 -7.22618699e-01
-8.42550173e-02 8.32225978e-01 5.88197112e-01 -1.88014999e-01
2.67487139e-01 2.45861217e-01 -3.11248582e-02 -6.49403214e-01
-8.18054020e-01 -6.70296013e-01 -5.19087076e-01 -1.55805513e-01
1.03573012e+00 -7.62064636e-01 -1.27769649e+00 9.80676174e-01
-1.08763194e+00 -5.39755523e-01 -4.19529110e-01 1.49801016e-01
-7.46012211e-01 7.50700384e-02 -1.07899261e+00 -1.24812412e+00
-8.83801639e-01 -1.07896245e+00 1.23606014e+00 -5.63146882e-02
-4.64163303e-01 -8.31668198e-01 2.68421710e-01 1.61097184e-01
8.99737179e-01 4.15116012e-01 1.07146633e+00 -7.37510383e-01
-3.05351615e-01 -4.25969541e-01 -1.12231374e-01 5.16391456e-01
3.36570293e-01 -2.65125841e-01 -1.20663202e+00 -3.09331447e-01
3.53482485e-01 3.60737555e-02 4.32751894e-01 5.66118836e-01
1.42769575e+00 -4.59336609e-01 -1.94310203e-01 2.61544824e-01
1.82479322e+00 5.57257652e-01 1.04288113e+00 2.59946197e-01
6.62112713e-01 -3.30509059e-02 1.69810042e-01 6.64086759e-01
3.52481574e-01 5.21674812e-01 5.42341232e-01 -7.39954337e-02
1.64478123e-01 -1.91409096e-01 4.44176376e-01 1.08214962e+00
3.86817873e-01 -4.83475000e-01 -7.28863358e-01 6.60329878e-01
-2.12961388e+00 -1.19907165e+00 -1.69183058e-03 1.81953049e+00
4.08914357e-01 1.60481542e-01 2.75882483e-01 6.68880820e-01
6.04432046e-01 1.63672671e-01 -7.10860491e-01 -7.88999617e-01
1.98238924e-01 3.94249558e-01 6.77865505e-01 3.55403841e-01
-1.20350933e+00 2.68503875e-01 7.21940994e+00 6.97762251e-01
-9.84827697e-01 4.89744812e-01 3.94588172e-01 -1.16665117e-01
4.89135653e-01 -8.88736784e-01 -6.56156540e-01 9.11451697e-01
1.67951369e+00 6.27490804e-02 3.54018241e-01 7.77437508e-01
4.75020587e-01 -1.87524498e-01 -1.09740734e+00 9.99884605e-01
1.25001416e-01 -8.16122651e-01 -4.62245673e-01 -2.53870357e-02
7.16913760e-01 4.14590180e-01 -2.36819312e-01 4.84413207e-01
4.78138775e-01 -9.04411018e-01 3.78145099e-01 7.09164798e-01
2.27989063e-01 -9.76870537e-01 1.32228816e+00 3.91857266e-01
-1.66416621e+00 -6.87798858e-01 1.57214195e-01 -3.12031657e-01
3.99196655e-01 6.80511415e-01 -8.38013828e-01 7.18725801e-01
1.13822305e+00 6.86036110e-01 -3.70073497e-01 8.08108985e-01
8.28232542e-02 9.57759619e-01 -5.24440587e-01 -7.61150047e-02
3.22078168e-01 1.31760135e-01 1.69920772e-01 1.46976924e+00
3.00215244e-01 1.28756324e-02 3.20412874e-01 7.48264909e-01
1.00744031e-01 -7.46736452e-02 -5.82557440e-01 2.48713523e-01
2.14063108e-01 1.11867452e+00 -3.41384381e-01 -6.82351410e-01
-5.85647941e-01 1.13731980e+00 -2.49054674e-02 3.80034894e-02
-1.06238353e+00 -3.60111266e-01 5.72722793e-01 -5.08892238e-02
6.94580853e-01 1.26415521e-01 -1.63505316e-01 -6.29280806e-01
-4.12997864e-02 -5.85196018e-01 5.92433095e-01 -8.67670059e-01
-1.63918734e+00 4.87519324e-01 9.65347737e-02 -6.97693527e-01
-2.56937772e-01 -9.65762213e-02 -9.49131250e-01 9.04065847e-01
-1.42622638e+00 -1.20347977e+00 -3.37939680e-01 6.43405378e-01
7.01705813e-01 4.20970656e-02 9.86940622e-01 6.55945659e-01
-8.11142802e-01 4.35292333e-01 2.25446597e-01 -2.82783806e-03
1.47708341e-01 -1.20182264e+00 6.41405761e-01 6.62825525e-01
-5.64301372e-01 2.13880949e-02 5.49314022e-01 -6.37853444e-01
-9.22308326e-01 -1.67792630e+00 8.81837308e-01 -4.72617775e-01
3.65650415e-01 -5.25445461e-01 -8.31824124e-01 9.31970000e-01
6.31400228e-01 -1.66006267e-01 6.91909611e-01 -1.20715402e-01
-6.47972431e-03 -4.38885123e-01 -1.54247081e+00 -1.28883332e-01
5.63438714e-01 -6.13710642e-01 -6.77445054e-01 1.85297668e-01
6.35677636e-01 1.60091445e-01 -1.05117989e+00 4.43694890e-01
4.07882214e-01 -7.81719625e-01 6.28778517e-01 -3.26825321e-01
-1.58336610e-01 -3.81353766e-01 -6.04741514e-01 -1.21172416e+00
-7.49336183e-01 -7.45698392e-01 -4.90568399e-01 1.61791658e+00
-3.43214199e-02 -7.42596507e-01 2.92832583e-01 4.15912420e-01
-1.11024201e-01 -7.58831143e-01 -1.03083265e+00 -1.07569134e+00
-3.54356349e-01 -6.33470953e-01 1.17385578e+00 6.70146406e-01
-4.30398323e-02 4.97671336e-01 -5.54881692e-01 2.60446161e-01
3.60689372e-01 -3.09597284e-01 2.39729121e-01 -1.10622656e+00
3.13693434e-02 -1.08776115e-01 -7.98968971e-01 -3.77404928e-01
1.79692894e-01 -5.56871712e-01 1.69798017e-01 -1.47315931e+00
4.62151825e-01 1.90309823e-01 -1.07816517e+00 5.79112649e-01
4.55915444e-02 2.44148508e-01 1.91824913e-01 -3.40640187e-01
-6.74395800e-01 6.75299525e-01 2.74121195e-01 -2.00892568e-01
-1.55687302e-01 -1.22235663e-01 -1.90800041e-01 4.90095943e-01
1.21070874e+00 -3.52353781e-01 -4.07094687e-01 -1.39949933e-01
-3.68499607e-01 -2.10183859e-01 2.68685699e-01 -1.32570684e+00
-1.28830969e-01 3.42928112e-01 4.40051645e-01 -1.18575001e+00
2.97658920e-01 -1.02419245e+00 2.97410667e-01 7.44353056e-01
-1.88446432e-01 6.24709964e-01 5.49282074e-01 4.45430130e-01
7.21092597e-02 6.03213795e-02 7.83386230e-01 -1.93981707e-01
-5.80922425e-01 2.45918602e-01 -6.21026337e-01 -2.89332807e-01
1.13250625e+00 2.83500195e-01 -1.26058623e-01 -3.48263979e-01
-4.02278006e-01 3.32028091e-01 -1.70965552e-01 4.74471718e-01
5.62317595e-02 -1.62675250e+00 -4.45476919e-01 2.96759218e-01
-6.65780380e-02 -3.56407732e-01 2.01880768e-01 7.60471702e-01
8.04829150e-02 7.26440549e-01 -1.38891026e-01 -5.52988291e-01
-1.27224195e+00 1.00866306e+00 4.83844638e-01 -9.16114390e-01
-5.85255563e-01 4.38906133e-01 -5.88908456e-02 -3.31935406e-01
5.41669309e-01 -7.08985746e-01 -9.78688076e-02 -3.88862789e-02
7.35623240e-01 7.81923413e-01 3.79606158e-01 -6.21617854e-01
-5.86126745e-01 2.99707115e-01 8.09098855e-02 3.33730519e-01
1.45723689e+00 -1.65495455e-01 -1.02685742e-01 9.92543101e-01
1.56500387e+00 -8.06641638e-01 -8.40059757e-01 -8.00117627e-02
1.35251477e-01 -1.06471226e-01 3.12688559e-01 -1.12909019e+00
-1.06786513e+00 9.52303231e-01 1.25171256e+00 3.97204131e-01
1.60614359e+00 -3.07689935e-01 1.29866242e+00 -7.45717390e-03
2.15239093e-01 -1.26389372e+00 -8.14037099e-02 4.52384114e-01
5.67240953e-01 -1.12235343e+00 -3.44941795e-01 1.26329601e-01
-1.07355915e-01 6.72043681e-01 4.19738710e-01 -9.88720171e-03
1.04429686e+00 7.06394434e-01 -6.60682470e-02 -2.11254925e-01
-9.19718981e-01 -4.81851846e-02 -9.22623426e-02 7.18945920e-01
3.83874983e-01 2.81526327e-01 -3.71298529e-02 4.78172362e-01
2.84137260e-02 5.63747346e-01 3.58232222e-02 1.24685991e+00
-4.65031862e-02 -7.78556406e-01 -3.18270177e-01 6.98603451e-01
-7.92006433e-01 2.55075693e-02 1.81219280e-01 5.55374742e-01
2.19600707e-01 1.20223045e+00 2.57844567e-01 -4.34314460e-01
6.91229343e-01 5.25391579e-01 2.12026775e-01 -1.32470414e-01
-8.95087600e-01 -2.37526838e-02 4.92691156e-03 -9.13549066e-01
-4.33282554e-01 -5.58919549e-01 -7.42342293e-01 -5.50975561e-01
-3.36799592e-01 -3.02423149e-01 5.97787738e-01 1.06647193e+00
4.75258380e-01 1.27607310e+00 8.88202488e-01 -8.50834250e-01
-5.38553119e-01 -1.44745302e+00 -6.04721606e-01 6.26105309e-01
4.83136415e-01 -5.03927350e-01 8.34493190e-02 -2.21682444e-01] | [16.060401916503906, 7.577633857727051] |
28aef99b-40c8-42c0-95bf-57fc3742cdd6 | prompt-based-conservation-learning-for-multi | 2209.06923 | null | https://arxiv.org/abs/2209.06923v1 | https://arxiv.org/pdf/2209.06923v1.pdf | Prompt-based Conservation Learning for Multi-hop Question Answering | Multi-hop question answering (QA) requires reasoning over multiple documents to answer a complex question and provide interpretable supporting evidence. However, providing supporting evidence is not enough to demonstrate that a model has performed the desired reasoning to reach the correct answer. Most existing multi-hop QA methods fail to answer a large fraction of sub-questions, even if their parent questions are answered correctly. In this paper, we propose the Prompt-based Conservation Learning (PCL) framework for multi-hop QA, which acquires new knowledge from multi-hop QA tasks while conserving old knowledge learned on single-hop QA tasks, mitigating forgetting. Specifically, we first train a model on existing single-hop QA tasks, and then freeze this model and expand it by allocating additional sub-networks for the multi-hop QA task. Moreover, to condition pre-trained language models to stimulate the kind of reasoning required for specific multi-hop questions, we learn soft prompts for the novel sub-networks to perform type-specific reasoning. Experimental results on the HotpotQA benchmark show that PCL is competitive for multi-hop QA and retains good performance on the corresponding single-hop sub-questions, demonstrating the efficacy of PCL in mitigating knowledge loss by forgetting. | ['Patricia Riddle', 'Michael Witbrock', 'Qianqian Qi', 'Yang Chen', 'Yonghua Zhu', 'Zhenyun Deng'] | 2022-09-14 | null | https://aclanthology.org/2022.coling-1.154 | https://aclanthology.org/2022.coling-1.154.pdf | coling-2022-10 | ['multi-hop-question-answering'] | ['knowledge-base'] | [ 9.46636647e-02 3.74748647e-01 5.97869977e-03 -4.62218553e-01
-1.37856042e+00 -6.68787658e-01 3.17522138e-01 5.42191803e-01
-4.47695374e-01 9.96231854e-01 1.54441923e-01 -6.07480407e-01
-2.72763878e-01 -1.03300488e+00 -9.91634309e-01 -1.87956870e-01
2.68125594e-01 7.69508183e-01 1.03636599e+00 -5.58955610e-01
1.61847785e-01 1.06384143e-01 -1.28023994e+00 8.67078960e-01
1.54578698e+00 7.86592841e-01 2.35942379e-01 9.00837123e-01
-4.20763701e-01 1.41492033e+00 -7.20851541e-01 -7.11029291e-01
-2.53990412e-01 -4.64740068e-01 -1.36812079e+00 -6.27834618e-01
8.85026574e-01 -5.07797599e-01 -1.99499443e-01 6.59838974e-01
2.63678730e-01 2.83080310e-01 2.43195087e-01 -1.06399560e+00
-1.03516400e+00 6.87833130e-01 -1.50465906e-01 5.15051186e-01
6.45973504e-01 2.52874941e-01 1.32910252e+00 -9.21726167e-01
4.01696742e-01 1.44623232e+00 7.36487389e-01 6.00549638e-01
-8.59475195e-01 -3.55858773e-01 3.10794741e-01 6.62839890e-01
-7.89808154e-01 -2.99922973e-01 6.72032118e-01 1.58532768e-01
1.06175196e+00 1.50921687e-01 2.90380269e-01 8.69944096e-01
2.79837787e-01 1.00164962e+00 9.11103249e-01 -5.05097032e-01
2.08323255e-01 7.56160840e-02 5.11350214e-01 1.10254240e+00
1.63313784e-02 -5.21210849e-01 -7.37093806e-01 -3.25694799e-01
-1.11383103e-01 -2.67380536e-01 -1.96111873e-01 -4.60015610e-02
-1.01475823e+00 8.52984250e-01 4.87762421e-01 4.66407359e-01
-5.47323883e-01 2.46343911e-02 2.03055963e-01 7.05821455e-01
1.56050667e-01 7.22145975e-01 -8.14458072e-01 8.71308446e-02
-8.57901514e-01 3.78847152e-01 8.73127580e-01 7.98711002e-01
9.48308289e-01 -3.69245589e-01 -8.22922528e-01 7.34717190e-01
2.99142301e-02 6.09322608e-01 4.23120230e-01 -1.23745573e+00
7.52103806e-01 1.06451797e+00 1.12925917e-01 -7.13516235e-01
-2.56415784e-01 -3.73106122e-01 -4.10833120e-01 -4.14083809e-01
4.68103915e-01 -3.16249877e-02 -8.35696399e-01 1.75594044e+00
3.71083558e-01 -5.49333021e-02 1.73307508e-01 6.27258062e-01
6.66744947e-01 8.70735824e-01 2.03793019e-01 -7.40912706e-02
1.38192821e+00 -1.47051263e+00 -5.94011188e-01 -5.53931952e-01
7.45980978e-01 -4.71695781e-01 1.73198187e+00 4.35790777e-01
-1.22998893e+00 -5.77371359e-01 -8.41455698e-01 -5.57448626e-01
-3.81730378e-01 -1.61890656e-01 3.81956279e-01 1.58378258e-01
-1.07137477e+00 2.69004881e-01 -2.54864812e-01 -8.99785236e-02
3.13829482e-01 2.66907155e-03 -8.08713436e-02 -7.10900962e-01
-1.92482901e+00 1.13214636e+00 2.99436122e-01 -3.12193427e-02
-1.05828846e+00 -1.09274006e+00 -6.48271799e-01 3.67464632e-01
7.08598435e-01 -9.57302332e-01 1.56857634e+00 -3.71477008e-01
-1.27325249e+00 3.50630581e-01 -4.47154701e-01 -4.19348270e-01
3.63931865e-01 -4.59317207e-01 -4.92416769e-01 6.78701162e-01
4.64179724e-01 7.79643953e-01 9.36640859e-01 -1.18697894e+00
-9.57430243e-01 -2.27267742e-01 7.82455325e-01 -8.40527117e-02
-5.04492581e-01 -4.88590539e-01 -3.43727887e-01 -1.38640761e-01
-1.55645415e-01 -4.52492893e-01 1.18312836e-01 1.95748126e-03
1.58477593e-02 -7.50476182e-01 8.57639730e-01 -8.45903754e-01
1.38738704e+00 -1.77222788e+00 2.36264803e-02 -2.59203434e-01
3.57743055e-01 4.98774201e-01 -7.01847255e-01 3.91705513e-01
4.47150946e-01 -1.42315179e-01 -3.91380399e-01 -1.44738659e-01
-6.05380312e-02 5.70645213e-01 -6.26383424e-01 -2.37567529e-01
6.69509768e-01 1.08653450e+00 -1.21113420e+00 -6.41692638e-01
-2.47801840e-01 1.48280591e-01 -7.03584075e-01 3.26267898e-01
-1.01001751e+00 2.10173070e-01 -3.10399741e-01 6.82404876e-01
6.20241284e-01 -5.09227037e-01 -6.85295835e-02 -1.70857951e-01
2.73896188e-01 6.45990789e-01 -7.46873498e-01 1.78019917e+00
-9.33914840e-01 2.40733802e-01 5.72661459e-02 -7.50573099e-01
6.85413897e-01 1.37740955e-01 -6.52088150e-02 -1.24905849e+00
-3.44172031e-01 3.65507245e-01 -5.09238094e-02 -6.97223008e-01
4.95401561e-01 -1.85816482e-01 2.11056732e-02 5.60708284e-01
2.34671891e-01 -2.51092851e-01 5.80840528e-01 7.85884202e-01
1.49183905e+00 -2.04850957e-01 -1.92394674e-01 1.62463650e-01
1.05067205e+00 2.54280835e-01 4.64150399e-01 1.08260751e+00
-2.78542995e-01 1.96028545e-01 4.16433811e-01 -2.76997298e-01
-5.62229335e-01 -1.23976147e+00 4.17701781e-01 1.65694201e+00
1.16857104e-01 -3.52859765e-01 -4.53063905e-01 -1.32228673e+00
7.75977597e-02 1.27944243e+00 -5.71516633e-01 -5.67121744e-01
-8.74369740e-01 -2.59969980e-01 6.91357195e-01 4.74676639e-01
8.60508502e-01 -1.09718835e+00 -6.77732468e-01 4.88530427e-01
-7.17007995e-01 -9.84980404e-01 -3.37685138e-01 1.22828498e-01
-7.47080982e-01 -1.24851215e+00 -6.69633508e-01 -8.03025603e-01
5.27300239e-01 4.40972447e-01 1.39501393e+00 5.77653527e-01
1.24006003e-01 8.44440758e-01 -5.10277808e-01 -2.59010702e-01
-4.66867805e-01 4.95693117e-01 -3.54616374e-01 -1.90283686e-01
3.20936114e-01 -3.39581400e-01 -4.37812656e-01 1.28335759e-01
-1.07488883e+00 -2.53718853e-01 7.78415382e-01 9.78079259e-01
3.14117074e-01 5.33847213e-02 1.04981911e+00 -9.44903493e-01
1.04024982e+00 -5.51375985e-01 -2.16224834e-01 1.04452717e+00
-7.33634651e-01 2.96994746e-01 9.27889585e-01 -4.35574621e-01
-1.47944009e+00 -5.06635249e-01 -2.72349566e-01 5.93390595e-03
2.26097226e-01 7.59107530e-01 7.99768567e-02 1.77421823e-01
6.57121181e-01 3.19941789e-01 -3.38959515e-01 -2.81302005e-01
7.17545867e-01 2.74341762e-01 6.26184285e-01 -9.38480258e-01
8.56179178e-01 3.51136088e-01 -3.33777487e-01 -3.38107407e-01
-1.71327960e+00 -3.93783391e-01 -5.42537630e-01 -1.33454561e-01
6.54038668e-01 -5.80872178e-01 -8.49073410e-01 1.16291933e-01
-1.46436763e+00 -5.18602192e-01 -2.79666215e-01 -6.09310530e-02
-2.69500911e-01 3.68719816e-01 -6.21505737e-01 -6.85518384e-01
-5.38884103e-01 -7.28503406e-01 9.51071560e-01 3.25003028e-01
-1.21261880e-01 -1.08809876e+00 1.89466134e-01 8.52298021e-01
5.33393562e-01 -5.76874793e-01 1.68344533e+00 -6.44380212e-01
-6.16197109e-01 -6.94396272e-02 -2.53486425e-01 4.29240853e-01
-4.09935415e-02 -3.73030365e-01 -8.96809757e-01 -1.46293014e-01
6.73496351e-02 -9.77093816e-01 1.13341856e+00 -2.78698534e-01
1.00793254e+00 -5.47689676e-01 -1.20173462e-01 -2.51747608e-01
9.66992497e-01 -1.61210284e-01 3.97100896e-01 3.45435590e-01
2.70343751e-01 6.91771567e-01 8.12489986e-01 -1.64022017e-02
9.80391443e-01 1.10579245e-01 3.91981959e-01 2.19090700e-01
-2.75618434e-01 -4.66681421e-01 3.22980493e-01 9.49583054e-01
5.92905819e-01 -2.88969159e-01 -1.04051268e+00 9.33941007e-01
-1.80627549e+00 -7.11407304e-01 -2.35967219e-01 1.77040255e+00
1.30670738e+00 2.22447827e-01 -2.33101740e-01 1.22560799e-01
2.62195498e-01 1.47204280e-01 -7.63698876e-01 -3.18064660e-01
-2.79072553e-01 3.46129298e-01 -2.25826636e-01 8.86204302e-01
-5.15014172e-01 1.04942214e+00 5.66390371e+00 6.31457210e-01
-7.29558408e-01 3.82285058e-01 1.32719651e-01 1.36384815e-01
-7.71313131e-01 1.97404459e-01 -8.58288407e-01 1.23851664e-01
1.02443016e+00 -1.68585181e-01 2.90777653e-01 6.22071803e-01
-2.23250195e-01 -5.27393937e-01 -1.05811858e+00 2.04833180e-01
1.60664350e-01 -1.44351685e+00 6.84850395e-01 -8.01018536e-01
5.95518827e-01 -2.29851648e-01 5.41832158e-03 9.73346591e-01
4.29867417e-01 -6.53973937e-01 4.41949785e-01 7.23474681e-01
2.04589069e-01 -5.39925396e-01 7.75297761e-01 8.03629339e-01
-8.15221727e-01 -3.75285953e-01 -3.82709771e-01 -5.89092681e-03
2.88276613e-01 6.28831387e-01 -9.92011845e-01 7.41454244e-01
6.90253437e-01 1.17935054e-01 -1.06426954e+00 7.37902284e-01
-7.38889813e-01 7.33498037e-01 -1.10238627e-01 -1.76355556e-01
5.52338600e-01 3.80191952e-01 2.25705937e-01 9.28352773e-01
1.59623504e-01 1.99214727e-01 -1.33699358e-01 7.54050493e-01
-3.91800702e-01 -1.65624321e-01 -7.62308538e-02 -1.31435990e-01
6.00438178e-01 1.06044853e+00 7.00729107e-03 -5.58171690e-01
-5.53257227e-01 1.03858268e+00 8.79995048e-01 2.30335861e-01
-6.65337920e-01 -5.57874799e-01 5.11133932e-02 1.40650291e-02
2.98673719e-01 -1.17711358e-01 -5.07123843e-02 -1.12568259e+00
2.94200450e-01 -1.05458260e+00 9.65383828e-01 -1.22580731e+00
-1.55475378e+00 4.72764313e-01 -2.70251483e-01 -4.72138613e-01
-2.00320303e-01 -4.09114689e-01 -5.47733486e-01 7.84661591e-01
-2.20883083e+00 -1.23568273e+00 -3.64578187e-01 7.97040224e-01
6.09043658e-01 1.74549162e-01 7.58780539e-01 3.73848498e-01
-1.61186039e-01 6.09748840e-01 -3.18602264e-01 -1.85607448e-01
1.02615368e+00 -1.35530698e+00 -4.82248776e-02 8.21892679e-01
8.95675644e-02 7.44544983e-01 6.14799321e-01 -7.56352544e-01
-1.63136911e+00 -1.00713754e+00 1.48937631e+00 -7.87040293e-01
7.86065400e-01 -8.89103934e-02 -1.62310112e+00 6.26699388e-01
2.03215331e-01 -1.64752573e-01 4.57029760e-01 2.28781149e-01
-7.00372338e-01 -3.32653970e-01 -1.06510472e+00 6.02264106e-01
6.88227594e-01 -6.67144418e-01 -1.40515530e+00 5.62325060e-01
1.23680246e+00 -1.95196256e-01 -6.56383038e-01 4.28692192e-01
1.38946503e-01 -7.14791477e-01 9.15192544e-01 -9.02123809e-01
5.18138230e-01 -3.37242246e-01 1.95373804e-03 -1.24855721e+00
-2.45994627e-01 -3.17791075e-01 -4.96399581e-01 1.24042439e+00
6.67254806e-01 -4.68922883e-01 6.01121008e-01 5.64485788e-01
-2.51082212e-01 -7.06816494e-01 -1.06641197e+00 -6.59803391e-01
3.79854411e-01 -4.49390978e-01 6.22037053e-01 7.95457423e-01
-2.31786251e-01 7.60972559e-01 -1.56794533e-01 4.83933330e-01
4.27646756e-01 1.95038840e-01 6.15349293e-01 -1.06657851e+00
-1.74239993e-01 -6.21468574e-02 5.82258224e-01 -1.30402505e+00
3.42953771e-01 -8.06006610e-01 1.01121716e-01 -1.96748435e+00
3.68609210e-04 -4.11790639e-01 -2.48967439e-01 7.22334027e-01
-6.55338347e-01 -1.78302616e-01 5.82672320e-02 -6.31117672e-02
-1.27355480e+00 7.28067935e-01 1.48898077e+00 -4.92431670e-01
4.48022746e-02 -1.74893633e-01 -8.03523540e-01 4.83291864e-01
5.49288154e-01 -6.29046202e-01 -6.97234333e-01 -8.20819914e-01
6.56804860e-01 1.17333852e-01 6.02610826e-01 -8.88968349e-01
7.55958021e-01 -8.05080459e-02 1.43071428e-01 -7.44376004e-01
2.13596553e-01 -7.06194818e-01 -8.41356993e-01 6.02093339e-01
-6.48759186e-01 7.52904713e-02 2.88940668e-01 6.82886183e-01
-3.60008121e-01 -4.94459957e-01 4.96488363e-01 -2.24127591e-01
-8.00691247e-01 1.13221712e-01 -3.05430681e-01 6.74717665e-01
6.52669907e-01 2.93582529e-01 -9.05400395e-01 -4.06545877e-01
-6.13735139e-01 9.39373791e-01 -1.43187463e-01 4.14567649e-01
7.47750342e-01 -9.96271312e-01 -7.17971563e-01 -1.58003375e-01
3.33705127e-01 9.16221365e-02 4.99696702e-01 6.97833300e-01
-2.43273810e-01 8.04262996e-01 -2.18493361e-02 -3.18727762e-01
-8.67242277e-01 4.98713106e-01 2.81711936e-01 -7.21759915e-01
-3.88383061e-01 9.42554057e-01 -2.22798556e-01 -7.81802654e-01
1.16129830e-01 -4.08011138e-01 -1.94951948e-02 1.54883161e-01
6.88415706e-01 4.09427494e-01 3.55463356e-01 2.13598937e-01
-4.13834840e-01 1.74449921e-01 -4.80427355e-01 -1.21053450e-01
1.22011960e+00 -2.43332624e-01 -3.89469683e-01 4.67036903e-01
8.10477793e-01 5.75424358e-02 -1.05636930e+00 -5.62588573e-01
1.93542555e-01 -1.25439554e-01 -2.36774191e-01 -1.54112113e+00
-5.27187526e-01 1.24130797e+00 8.93914700e-03 1.76538691e-01
1.19353187e+00 2.27802947e-01 1.18037045e+00 1.00573683e+00
4.02129114e-01 -1.05672300e+00 7.74264932e-01 1.01260495e+00
1.19479513e+00 -1.12122667e+00 -3.48522097e-01 -1.11934327e-01
-4.70669091e-01 1.05928993e+00 1.14397538e+00 3.50121379e-01
1.92813948e-01 -1.84248820e-01 1.60050586e-01 -2.66856521e-01
-1.25444710e+00 2.20982227e-02 2.30996728e-01 4.47480023e-01
1.18863724e-01 -4.31388736e-01 -7.88504183e-02 7.54803896e-01
-7.49718696e-02 1.20545477e-01 3.38284105e-01 1.07428360e+00
-7.90464580e-01 -9.18201625e-01 -2.55680352e-01 4.25074100e-01
-6.11181976e-03 -3.76434982e-01 -5.51865280e-01 6.26616120e-01
4.17254157e-02 1.21260703e+00 -2.57787913e-01 -1.15448177e-01
5.53013504e-01 6.06866360e-01 4.53603625e-01 -5.97518027e-01
-8.18057001e-01 -8.34703207e-01 1.68033540e-01 -4.18678105e-01
-1.52931780e-01 -1.56931773e-01 -1.37681162e+00 -2.92388618e-01
-1.71493649e-01 5.10335565e-01 9.74909887e-02 1.19156218e+00
5.03410280e-01 7.64721513e-01 2.11664185e-01 2.72896111e-01
-8.14129889e-01 -9.51796412e-01 2.13373024e-02 1.45797104e-01
5.74566066e-01 -4.52083290e-01 -4.50872272e-01 -1.88591063e-01] | [10.903515815734863, 7.901548385620117] |
7aa4a470-3403-4215-8624-8f1191dd7b4b | cdgnet-class-distribution-guided-network-for | 2111.14173 | null | https://arxiv.org/abs/2111.14173v3 | https://arxiv.org/pdf/2111.14173v3.pdf | CDGNet: Class Distribution Guided Network for Human Parsing | The objective of human parsing is to partition a human in an image into constituent parts. This task involves labeling each pixel of the human image according to the classes. Since the human body comprises hierarchically structured parts, each body part of an image can have its sole position distribution characteristic. Probably, a human head is less likely to be under the feet, and arms are more likely to be near the torso. Inspired by this observation, we make instance class distributions by accumulating the original human parsing label in the horizontal and vertical directions, which can be utilized as supervision signals. Using these horizontal and vertical class distribution labels, the network is guided to exploit the intrinsic position distribution of each class. We combine two guided features to form a spatial guidance map, which is then superimposed onto the baseline network by multiplication and concatenation to distinguish the human parts precisely. We conducted extensive experiments to demonstrate the effectiveness and superiority of our method on three well-known benchmarks: LIP, ATR, and CIHP databases. | ['Wonjun Hwang', 'Jianming Wang', 'Ouk Choi', 'Kunliang Liu'] | 2021-11-28 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Liu_CDGNet_Class_Distribution_Guided_Network_for_Human_Parsing_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Liu_CDGNet_Class_Distribution_Guided_Network_for_Human_Parsing_CVPR_2022_paper.pdf | cvpr-2022-1 | ['human-parsing'] | ['computer-vision'] | [ 2.26173460e-01 4.10314083e-01 -5.19253552e-01 -5.61667800e-01
-2.23743707e-01 -3.44503164e-01 2.55693913e-01 -1.22737035e-01
-2.23981306e-01 4.06886041e-01 3.80496919e-01 4.74220403e-02
3.33266586e-01 -7.88641214e-01 -6.18935823e-01 -7.65562177e-01
3.08736563e-01 4.25440758e-01 3.30475509e-01 1.66870371e-01
-1.00783110e-01 1.24604486e-01 -1.41362178e+00 4.71579790e-01
6.81187332e-01 1.19725180e+00 2.52742082e-01 1.89434230e-01
-9.41064730e-02 7.07429111e-01 -4.95842487e-01 -2.73948878e-01
1.61337748e-01 -6.34877145e-01 -8.18089724e-01 5.58102846e-01
2.53796965e-01 -2.62169391e-01 -2.10442171e-01 1.19999576e+00
3.78437936e-01 5.64053245e-02 7.90096343e-01 -1.02353692e+00
-5.04326046e-01 6.70802891e-01 -1.24424994e+00 -1.52858734e-01
4.09951687e-01 6.83192536e-02 9.76374269e-01 -5.29642761e-01
4.31456327e-01 1.42615592e+00 4.15269256e-01 6.00280344e-01
-8.23028088e-01 -6.34283066e-01 4.21559870e-01 4.68312316e-02
-1.16511357e+00 -1.90351173e-01 1.01746845e+00 -3.53103936e-01
-3.81254926e-02 -7.61412457e-02 6.81467056e-01 8.86894524e-01
2.94963457e-02 1.42108417e+00 9.67912972e-01 -3.27850997e-01
8.10695067e-02 -4.76592267e-03 2.51754582e-01 9.41959560e-01
1.76838860e-01 -2.03302369e-01 -4.06112969e-01 1.92091942e-01
6.52452767e-01 9.85082313e-02 -2.47298509e-01 -7.08927393e-01
-1.17693567e+00 5.43953717e-01 7.61902988e-01 -3.73736061e-02
-3.06775957e-01 -1.69199914e-01 3.48804116e-01 -5.10365784e-01
2.80750580e-02 -3.10760140e-01 -3.63330245e-01 4.03689593e-01
-7.19318807e-01 6.03404492e-02 4.50042367e-01 9.84212220e-01
8.70877802e-01 -5.67298830e-01 -3.77155602e-01 9.97311950e-01
5.39948821e-01 3.92783046e-01 5.55098057e-01 -1.07089496e+00
7.79112697e-01 9.80226159e-01 -5.94693571e-02 -1.05675566e+00
-5.71015835e-01 -3.17459285e-01 -1.09974444e+00 3.60197164e-02
6.65292501e-01 -2.64698774e-01 -1.23534906e+00 1.81519878e+00
5.44230461e-01 -2.92057335e-01 -7.32239857e-02 1.12715316e+00
7.97004282e-01 6.04455352e-01 2.23588377e-01 1.58060491e-01
1.74242234e+00 -1.22215724e+00 -5.58068037e-01 -5.42511582e-01
2.01379061e-01 -3.36838156e-01 1.06815529e+00 2.35485554e-01
-8.36062431e-01 -7.57709920e-01 -9.49237764e-01 -1.43861800e-01
-7.95425549e-02 6.39863968e-01 5.75562119e-01 5.10629952e-01
-4.68094230e-01 3.48750919e-01 -8.83033931e-01 -2.01671049e-01
6.23783648e-01 1.10142780e-02 -4.43564177e-01 -1.87724710e-01
-1.01045191e+00 2.86609977e-01 3.68028194e-01 3.07609528e-01
-5.16606271e-01 -1.72096178e-01 -1.35181189e+00 4.10476001e-03
3.27117264e-01 -7.19202876e-01 9.90263820e-01 -8.90222728e-01
-1.27934325e+00 1.07279611e+00 -2.97951043e-01 -1.02328643e-01
5.93457639e-01 -5.13478853e-02 1.19176945e-02 2.69952178e-01
4.62802529e-01 1.04959786e+00 6.54485822e-01 -1.41755128e+00
-1.05794549e+00 -7.20000148e-01 -1.42689675e-01 4.07662541e-01
6.71733767e-02 -1.53817981e-01 -9.95954335e-01 -7.02115238e-01
6.66997731e-01 -8.69057178e-01 -8.27589259e-02 -1.94011088e-02
-1.09767354e+00 -2.14137971e-01 5.19439876e-01 -8.37517977e-01
1.15398812e+00 -2.25468254e+00 1.94585025e-01 2.56929368e-01
1.60462275e-01 -3.18614393e-01 2.05496579e-01 -1.65356413e-01
-7.53038377e-02 -5.91186294e-03 -4.27650690e-01 -3.14555407e-01
-6.73679635e-02 2.39342898e-01 -2.46844180e-02 5.68776548e-01
2.29129214e-02 6.64686561e-01 -7.40915358e-01 -9.26653981e-01
9.42729190e-02 3.22081298e-01 -4.03130531e-01 2.56070554e-01
5.01219817e-02 4.97708201e-01 -6.66187704e-01 8.40525031e-01
7.81945646e-01 -3.17275017e-01 3.17929417e-01 -3.00349146e-01
2.84849107e-01 1.54353634e-01 -1.16036808e+00 1.53110373e+00
4.69985083e-02 2.73495197e-01 1.15114927e-01 -8.96801591e-01
7.86628008e-01 -2.89024729e-02 3.91567260e-01 -6.67861402e-01
1.94929644e-01 -1.90820128e-01 -7.66016683e-03 -3.66303414e-01
6.91301152e-02 1.06533982e-01 -1.84114739e-01 2.50769317e-01
-2.31946245e-01 2.93835878e-01 1.45731464e-01 1.44252917e-02
5.74397266e-01 1.58489704e-01 2.95649469e-01 -2.99545646e-01
6.72229230e-01 -3.24706346e-01 9.89389062e-01 3.44481856e-01
-5.40625036e-01 9.60751235e-01 8.79551232e-01 -4.67037708e-01
-5.90593874e-01 -1.20057881e+00 -1.20226100e-01 1.30528677e+00
6.05284154e-01 -2.00815678e-01 -1.14538562e+00 -9.61438656e-01
-2.25294188e-01 2.05217987e-01 -9.84124959e-01 -7.65464827e-02
-6.29557610e-01 -5.72924674e-01 2.59112954e-01 7.05968320e-01
8.54574382e-01 -1.39655995e+00 -7.95582831e-01 -1.11364946e-01
-5.07800460e-01 -9.50407684e-01 -6.38967872e-01 1.78079814e-01
-4.04718995e-01 -1.33877027e+00 -1.13176286e+00 -1.19233263e+00
8.36563349e-01 6.51718974e-02 1.03727448e+00 8.85274559e-02
-2.45199233e-01 -2.25125123e-02 -2.75443226e-01 -1.76906914e-01
1.05383463e-01 5.32432459e-02 -1.81653365e-01 3.13578188e-01
1.22190967e-01 -3.19329351e-01 -1.07235408e+00 6.35053098e-01
-5.22416711e-01 2.63124257e-01 6.17645085e-01 8.79580379e-01
8.31310928e-01 3.36807698e-01 -2.77165510e-02 -9.72665787e-01
9.97927319e-03 -4.88950878e-01 -2.07590997e-01 3.79164606e-01
5.88944815e-02 -3.47216539e-02 2.95267850e-01 -2.36434653e-01
-1.07000303e+00 6.16265535e-01 -2.92129535e-02 -1.63849011e-01
-3.82483393e-01 2.17084572e-01 -9.05957162e-01 4.72776294e-01
1.04667895e-01 1.32369190e-01 -3.17444243e-02 -4.53347623e-01
3.85749996e-01 5.88834226e-01 9.44863558e-01 -8.16488266e-01
2.78501630e-01 4.79422420e-01 -1.92269981e-01 -7.31611729e-01
-1.08534169e+00 -4.73647982e-01 -7.82518804e-01 -1.99271843e-01
1.29808104e+00 -7.23658144e-01 -5.35084188e-01 6.27494514e-01
-9.02659059e-01 -3.74015659e-01 3.29868533e-02 8.03452805e-02
-3.89878958e-01 5.39154828e-01 -6.14607513e-01 -6.44705594e-01
-7.89223462e-02 -1.24626386e+00 1.38587224e+00 6.89265430e-01
-1.01752087e-01 -7.56903470e-01 -2.37895980e-01 4.29433316e-01
-3.98005456e-01 2.06539810e-01 9.84735906e-01 -3.91048521e-01
-1.74869120e-01 -1.14256352e-01 -3.74138832e-01 2.66438097e-01
1.77800655e-01 6.01520166e-02 -8.16434026e-01 -4.39387839e-03
-2.17581376e-01 -3.22676986e-01 1.05242431e+00 6.34866178e-01
1.47646463e+00 -5.37792295e-02 -6.85430408e-01 6.95068836e-01
9.48291659e-01 4.32050414e-02 6.45537555e-01 1.96829095e-01
8.33756804e-01 1.10600209e+00 7.02189922e-01 3.21978778e-01
6.05345309e-01 6.36651695e-01 3.71460646e-01 -2.46401727e-01
-2.28763893e-01 -5.95111728e-01 2.02966854e-01 3.34465533e-01
8.05799477e-03 -2.88585156e-01 -9.04980361e-01 3.40537846e-01
-1.64950681e+00 -7.89111435e-01 1.12932637e-01 2.10240769e+00
7.74838924e-01 2.32472897e-01 3.26547623e-01 1.51880085e-01
1.08552325e+00 2.37211138e-01 -5.15544593e-01 3.54720533e-01
-1.26784667e-01 -2.08601907e-01 4.05924350e-01 8.08234885e-02
-1.47916472e+00 6.12683654e-01 6.22667408e+00 7.95910776e-01
-9.23761904e-01 -3.68020624e-01 1.19841683e+00 3.75135869e-01
9.82811302e-03 -2.33861566e-01 -9.24684942e-01 8.09608936e-01
5.87561093e-02 3.22306007e-01 1.14400303e-02 9.42192912e-01
-2.62472451e-01 -2.02419892e-01 -1.02979600e+00 1.11078620e+00
-5.32390662e-02 -6.87029481e-01 -1.08120553e-02 1.56207625e-02
5.29842019e-01 -2.38619089e-01 1.50257587e-01 2.44803533e-01
1.80537045e-01 -9.69352901e-01 1.02296805e+00 2.37842470e-01
6.55210078e-01 -7.73296058e-01 6.53266191e-01 6.31377816e-01
-1.47415817e+00 -1.37572229e-01 -4.70394194e-01 1.58736423e-01
1.26997605e-01 4.21540260e-01 -2.79029757e-01 4.48742896e-01
8.97614658e-01 7.25025833e-01 -4.96979296e-01 7.94129431e-01
-5.61198950e-01 4.32632208e-01 -2.15899825e-01 2.92626053e-01
8.09830427e-02 -3.78788352e-01 9.43358690e-02 1.05860627e+00
9.59748998e-02 2.92098403e-01 6.13007903e-01 7.52595425e-01
-1.30645603e-01 -1.86678804e-02 -2.53518105e-01 3.14555168e-01
3.22401494e-01 1.34406269e+00 -1.01273882e+00 -3.73642117e-01
-4.55971926e-01 1.09820807e+00 2.65502661e-01 3.88127744e-01
-9.01003361e-01 -4.79744673e-01 4.00845677e-01 1.69127002e-01
2.79114932e-01 1.01135634e-01 -4.31599528e-01 -9.63156641e-01
1.38816923e-01 -7.72565246e-01 7.95083344e-01 -7.17774510e-01
-1.16097152e+00 6.88983798e-01 -1.17779844e-01 -1.24103677e+00
-8.57000891e-03 -7.23310828e-01 -5.65583825e-01 8.58966470e-01
-1.12814724e+00 -1.17043138e+00 -5.62853336e-01 5.46371639e-01
5.11946917e-01 2.58879066e-02 3.73729080e-01 2.84497738e-01
-9.16966259e-01 7.74476945e-01 -4.51913357e-01 6.96941435e-01
6.25758231e-01 -1.30507410e+00 5.58159426e-02 6.82433188e-01
-1.42230988e-02 5.63684702e-01 3.49299818e-01 -7.71715581e-01
-7.80026436e-01 -1.09848094e+00 6.34066164e-01 -2.76372313e-01
1.69843823e-01 -3.34605932e-01 -7.56047547e-01 6.52164936e-01
7.86623880e-02 2.57863671e-01 4.61636364e-01 1.00016952e-01
-4.17378336e-01 -1.78055897e-01 -9.92177367e-01 4.22626168e-01
1.01480639e+00 -2.42031157e-01 -7.12781310e-01 1.64312720e-01
3.71736228e-01 -5.97571969e-01 -4.85530406e-01 6.20312929e-01
5.96535563e-01 -1.12673998e+00 1.01580393e+00 -2.50345588e-01
8.56088281e-01 -4.51154858e-01 -7.35503361e-02 -1.05195522e+00
-3.41348529e-01 -1.41087562e-01 1.62716165e-01 1.31017852e+00
3.00296158e-01 -3.79674643e-01 1.14192116e+00 4.55489844e-01
-4.96413223e-02 -8.98113489e-01 -6.78278089e-01 -3.69003713e-01
-2.65271706e-03 -8.69720131e-02 6.31527722e-01 6.30497634e-01
-6.38439730e-02 4.27334249e-01 -3.50275248e-01 2.72492349e-01
7.16074646e-01 3.01836610e-01 6.93485916e-01 -9.37993944e-01
-3.17818850e-01 -5.14956892e-01 -3.96010995e-01 -1.54896045e+00
8.94840583e-02 -5.70446789e-01 3.79999548e-01 -1.56281710e+00
6.64090574e-01 -5.52030325e-01 -4.63084877e-01 7.09511399e-01
-5.44403195e-01 3.40163022e-01 8.24553296e-02 1.42355144e-01
-6.52421355e-01 4.59304988e-01 1.49121773e+00 -3.08993936e-01
-4.09866460e-02 1.85254931e-01 -7.38415420e-01 1.11727941e+00
5.60955822e-01 -1.19152263e-01 -3.65625352e-01 -4.38441128e-01
-1.66662365e-01 1.00580454e-01 2.37940148e-01 -1.02393341e+00
2.01114118e-01 -1.08640634e-01 8.81636977e-01 -8.61574411e-01
1.03841372e-01 -7.55474091e-01 -1.55451417e-01 3.89915735e-01
-3.22411537e-01 -2.48053506e-01 -2.18258321e-01 5.09298563e-01
-3.34441066e-01 1.10662699e-01 9.53762889e-01 -1.46672174e-01
-8.19410026e-01 4.81393486e-01 8.88468102e-02 1.14577770e-01
1.02021420e+00 -2.76351422e-01 -8.08412433e-02 -2.55058587e-01
-7.91893661e-01 5.50871670e-01 5.39419949e-01 2.79831648e-01
4.31919873e-01 -1.28087425e+00 -3.55369031e-01 3.98081899e-01
1.18935511e-01 4.38971817e-01 4.43205595e-01 6.64444566e-01
-5.71886480e-01 1.18103199e-01 -3.84448707e-01 -8.52996647e-01
-1.17979944e+00 6.63519263e-01 5.80359459e-01 -9.18548256e-02
-8.01881731e-01 8.88698161e-01 1.03028512e+00 -4.68787044e-01
5.70967913e-01 -2.88570106e-01 -5.27061284e-01 -1.48869893e-02
5.84031940e-01 1.31864667e-01 -4.61472243e-01 -8.81869495e-01
-5.13348043e-01 8.82528663e-01 1.43064737e-01 -4.37135026e-02
7.87647128e-01 -2.81886041e-01 -1.03690796e-01 1.90834463e-01
1.08650863e+00 3.07883918e-01 -1.45619488e+00 -1.70729026e-01
-1.68366894e-01 -3.75292689e-01 -4.57884699e-01 -7.38898516e-01
-1.48235321e+00 1.13053381e+00 3.73415112e-01 4.48269546e-02
1.31662130e+00 3.59776795e-01 7.79859304e-01 -1.56123325e-01
3.76536697e-01 -1.01986861e+00 -2.34358031e-02 3.01333576e-01
5.78407943e-01 -1.22743750e+00 -4.44813743e-02 -8.64080906e-01
-1.08406365e+00 8.86850953e-01 7.05604851e-01 5.95594049e-02
5.83020687e-01 5.11313453e-02 2.22441956e-01 -6.29604757e-02
-7.06339329e-02 -2.46494636e-01 4.85705584e-01 6.88610077e-01
4.24930364e-01 2.85080135e-01 -2.21526027e-01 1.08251560e+00
-1.98261812e-01 -3.94959986e-01 -9.12313908e-02 7.70829260e-01
-4.73192841e-01 -9.19793367e-01 -3.47678095e-01 2.65762120e-01
-5.07889628e-01 3.72759104e-01 -3.56838077e-01 7.10532963e-01
5.43510497e-01 8.07898521e-01 1.97555557e-01 -3.59554648e-01
3.24504435e-01 -1.40050143e-01 3.12682360e-01 -6.95562005e-01
1.23703852e-02 2.78943241e-01 -1.44655392e-01 -6.12476349e-01
-1.78503335e-01 -6.50769472e-01 -1.75177467e+00 7.87400752e-02
6.26229197e-02 2.36757426e-03 2.68445700e-01 7.76836276e-01
6.64101094e-02 5.97621977e-01 5.41464388e-01 -8.66897941e-01
-3.67232889e-01 -7.76820838e-01 -7.82193124e-01 6.59780204e-01
1.97247908e-01 -9.73662734e-01 -2.08684906e-01 2.59638727e-01] | [8.633820533752441, 0.015389771200716496] |
e2245220-5519-4831-8296-0e44ea701855 | plug-and-play-multilingual-few-shot-spoken | 2305.03058 | null | https://arxiv.org/abs/2305.03058v1 | https://arxiv.org/pdf/2305.03058v1.pdf | Plug-and-Play Multilingual Few-shot Spoken Words Recognition | As technology advances and digital devices become prevalent, seamless human-machine communication is increasingly gaining significance. The growing adoption of mobile, wearable, and other Internet of Things (IoT) devices has changed how we interact with these smart devices, making accurate spoken words recognition a crucial component for effective interaction. However, building robust spoken words detection system that can handle novel keywords remains challenging, especially for low-resource languages with limited training data. Here, we propose PLiX, a multilingual and plug-and-play keyword spotting system that leverages few-shot learning to harness massive real-world data and enable the recognition of unseen spoken words at test-time. Our few-shot deep models are learned with millions of one-second audio clips across 20 languages, achieving state-of-the-art performance while being highly efficient. Extensive evaluations show that PLiX can generalize to novel spoken words given as few as just one support example and performs well on unseen languages out of the box. We release models and inference code to serve as a foundation for future research and voice-enabled user interface development for emerging devices. | ['Vasileios Tsouvalas', 'Aaqib Saeed'] | 2023-05-03 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [-4.79541719e-02 -4.43018138e-01 -2.16855511e-01 -6.17254496e-01
-1.16733730e+00 -5.76626003e-01 4.07482237e-01 -6.11077659e-02
-4.88084584e-01 4.21005577e-01 5.22380650e-01 -2.07529828e-01
1.99012667e-01 -3.99084002e-01 -3.94096971e-01 -1.67842075e-01
4.54677567e-02 5.44699490e-01 2.24098176e-01 -3.48927081e-01
-3.89967799e-01 2.37174720e-01 -1.86890244e+00 3.89545798e-01
6.15124464e-01 9.86788452e-01 2.75850922e-01 1.06459355e+00
-2.85455674e-01 3.16331983e-01 -5.43518484e-01 -3.12744468e-01
-5.01110330e-02 -8.67614374e-02 -3.97598833e-01 -2.04278961e-01
4.39765841e-01 -5.37044227e-01 -4.02713239e-01 6.96101487e-01
1.29161000e+00 2.90797532e-01 5.15459962e-02 -1.05321372e+00
-7.04804480e-01 9.55473423e-01 -5.50830401e-02 2.63489246e-01
8.44036937e-01 2.66818404e-01 1.11403990e+00 -1.37954283e+00
1.36750922e-01 1.20524561e+00 5.67881346e-01 8.25329900e-01
-7.81689167e-01 -8.40275109e-01 1.41184121e-01 1.85809344e-01
-1.53936124e+00 -1.28502309e+00 2.10996538e-01 -7.45966565e-03
1.70442212e+00 4.41902965e-01 5.75489342e-01 1.57264996e+00
-1.21661298e-01 1.13640404e+00 2.72311479e-01 -4.05810326e-01
2.72246450e-01 1.80708826e-01 1.56929955e-01 4.50885534e-01
-1.01350933e-01 -4.52103645e-01 -1.17816138e+00 -1.32944852e-01
3.42567228e-02 1.79494888e-01 -1.95106536e-01 1.64316162e-01
-1.13201630e+00 4.61682707e-01 -1.66076183e-01 5.21218896e-01
-3.32063735e-01 7.25427046e-02 5.34885108e-01 2.92051792e-01
5.72629988e-01 2.11196736e-01 -4.89372939e-01 -1.07935548e+00
-8.82411063e-01 -5.24614416e-02 1.13857651e+00 1.03105271e+00
2.71860600e-01 3.84578645e-01 -1.09258890e-01 1.32320535e+00
-4.13648523e-02 8.43323886e-01 8.11933279e-01 -5.79535425e-01
3.06901366e-01 1.91607758e-01 -1.32566437e-01 -4.69506353e-01
-3.00927430e-01 -3.83225262e-01 -5.33253193e-01 -6.49209440e-01
-8.42373967e-02 -3.39555711e-01 -8.82315457e-01 1.53153872e+00
1.29191950e-01 5.73233545e-01 -8.43255520e-02 7.61423588e-01
9.07298386e-01 8.96253169e-01 -1.69038683e-01 -2.01041341e-01
1.30171680e+00 -8.29549789e-01 -8.31293702e-01 -5.29242754e-01
5.33720434e-01 -7.35260665e-01 1.79988682e+00 7.70527124e-01
-8.62038553e-01 -4.32312787e-01 -9.38739955e-01 -1.83712795e-01
-4.38812464e-01 -3.05214167e-01 5.63409567e-01 7.57998049e-01
-9.51349199e-01 2.03062385e-01 -8.94655943e-01 -8.49157155e-01
1.76047608e-01 3.94336849e-01 -1.45128667e-01 -9.72106904e-02
-1.28503001e+00 4.81647193e-01 -3.97548079e-02 -2.61402071e-01
-6.22510791e-01 -7.47940183e-01 -8.05091381e-01 1.80623069e-01
3.84866625e-01 -2.91805476e-01 1.73069358e+00 -6.15616858e-01
-1.83148682e+00 6.59132719e-01 -3.09395283e-01 -2.99720615e-01
4.42380495e-02 -6.34951174e-01 -1.14807045e+00 -9.40756500e-02
-5.80695532e-02 3.26413542e-01 8.96040380e-01 -6.60272539e-01
-6.94798648e-01 -2.68100649e-01 -4.66643780e-01 5.72922789e-02
-1.05567825e+00 3.75448883e-01 -6.37545168e-01 -3.78245503e-01
-3.39947969e-01 -5.89689553e-01 4.08868313e-01 -5.92295974e-02
-3.19402784e-01 -4.79186147e-01 1.00405848e+00 -6.58131599e-01
1.43905699e+00 -2.41172767e+00 -2.23652557e-01 -9.57335904e-02
8.05097967e-02 5.18065870e-01 -2.06956953e-01 5.26608229e-01
4.64152783e-01 -6.38958216e-02 2.53427595e-01 -7.41534472e-01
3.32920372e-01 2.83821195e-01 -5.61809897e-01 1.06525354e-01
-1.38120607e-01 9.23743069e-01 -1.01739848e+00 -2.81311899e-01
3.69182676e-01 6.53908372e-01 -4.44360912e-01 2.39845246e-01
-1.89473465e-01 7.62285367e-02 -2.33421162e-01 8.51205587e-01
2.91759614e-02 -1.61820710e-01 -2.30650127e-01 1.70267791e-01
-9.83455684e-03 4.43962634e-01 -9.12770092e-01 1.97816253e+00
-8.60533357e-01 5.76980710e-01 1.58208758e-01 -5.04305840e-01
8.46004903e-01 5.33592820e-01 2.43222535e-01 -7.98221409e-01
1.41290545e-01 4.55534607e-01 -2.96034902e-01 -7.94290304e-01
5.18627644e-01 2.48951744e-02 -3.36623341e-01 4.51781750e-01
4.55116957e-01 -1.21872395e-01 -1.30885839e-01 2.39635199e-01
1.22930205e+00 -6.55493140e-01 3.23854327e-01 3.15626234e-01
-2.59400532e-02 -6.41818166e-01 2.96366990e-01 9.38317418e-01
-2.62463093e-01 5.99881411e-01 -4.22614634e-01 -4.01607037e-01
-6.40202641e-01 -1.29961216e+00 9.76647958e-02 1.93555772e+00
-2.61581630e-01 -4.87441599e-01 -4.75282252e-01 -1.15623474e-01
1.89978883e-01 9.56792653e-01 1.72432497e-01 -3.30278993e-01
-1.04642093e-01 -3.90233129e-01 8.67161214e-01 4.58423704e-01
1.49306819e-01 -1.10395300e+00 -4.46203798e-01 4.19556081e-01
-1.71221629e-01 -1.44560313e+00 -7.87623703e-01 1.33728519e-01
-3.60652834e-01 -2.91767418e-01 -8.28655183e-01 -9.49133754e-01
-2.67797768e-01 4.96413529e-01 1.11232531e+00 -3.40705335e-01
-4.91171569e-01 8.35762203e-01 -5.84895849e-01 -4.90666389e-01
-1.52076706e-01 3.28018129e-01 7.99545825e-01 1.81268200e-01
7.89321482e-01 -8.54888499e-01 -4.69245464e-01 -1.34259555e-02
-6.01326466e-01 -3.03604305e-01 3.40184420e-01 6.20067120e-01
1.15624912e-01 -3.62028718e-01 9.13981616e-01 -2.84623802e-01
1.02597833e+00 -5.65965772e-01 -3.47423628e-02 4.32966709e-01
-3.47411335e-01 -2.07040533e-01 5.86962283e-01 -8.22976828e-01
-7.35366940e-01 -1.77362189e-01 -5.42364836e-01 -3.77403051e-01
-2.06507027e-01 3.87466758e-01 -2.37837866e-01 1.98648736e-01
6.62991703e-01 4.50618237e-01 -3.96632582e-01 -6.95685029e-01
6.37286723e-01 1.69502795e+00 7.52378762e-01 -2.69786328e-01
3.43711913e-01 1.86203718e-01 -1.13598108e+00 -1.51807857e+00
-6.67741239e-01 -8.44257295e-01 -2.17000112e-01 -2.01438695e-01
6.59675777e-01 -1.10236299e+00 -8.92997265e-01 4.69315529e-01
-1.03836036e+00 -3.75392586e-01 -3.35050285e-01 5.06333888e-01
-2.43343458e-01 9.16601345e-02 -5.34694374e-01 -1.16712129e+00
-8.09899390e-01 -8.65659237e-01 1.50615525e+00 2.11594284e-01
-7.27993190e-01 -6.06181204e-01 1.30236611e-01 2.07314119e-01
6.96246922e-01 -6.98024929e-01 5.80481887e-01 -1.15396011e+00
-1.75430536e-01 -5.60954213e-01 1.08603776e-01 4.15641308e-01
2.71003306e-01 -1.86785638e-01 -1.30327284e+00 -4.38229561e-01
-1.28679097e-01 -7.05196321e-01 5.57699561e-01 -2.56707817e-02
1.14156759e+00 -4.62756127e-01 -3.13814968e-01 5.30871987e-01
6.55485690e-01 1.16720892e-01 1.43966466e-01 -3.38760555e-01
7.43076682e-01 1.23686828e-01 1.11742884e-01 8.19056153e-01
5.12753487e-01 7.71998227e-01 -8.16580504e-02 3.06866258e-01
-8.24991555e-04 -2.36183956e-01 5.80681026e-01 1.66533494e+00
6.44705176e-01 -5.40776670e-01 -1.13011932e+00 8.16087961e-01
-1.63105166e+00 -7.88704932e-01 5.87089181e-01 2.22417116e+00
1.02829444e+00 2.08414435e-01 1.41999990e-01 1.59902200e-02
4.25912976e-01 2.32211858e-01 -8.48414063e-01 -4.94031042e-01
3.08409147e-03 6.02702022e-01 5.89204545e-04 4.65696096e-01
-7.53399193e-01 1.08090198e+00 6.13552189e+00 8.13634455e-01
-1.57225430e+00 4.43869203e-01 2.64165848e-01 -7.30401576e-01
-3.18469435e-01 -6.56642020e-01 -9.90861595e-01 5.02396643e-01
1.34419799e+00 -3.22774023e-01 8.07153761e-01 9.37247634e-01
2.46600389e-01 2.60264069e-01 -1.14413965e+00 1.63370633e+00
3.97195607e-01 -8.94834936e-01 -1.39897186e-02 -3.11665356e-01
3.62466961e-01 6.66025102e-01 1.65865541e-01 6.87938333e-01
1.86303496e-01 -8.94921064e-01 5.99376798e-01 5.05550265e-01
1.17122006e+00 -6.97951257e-01 2.42955372e-01 4.81261998e-01
-1.03836298e+00 -1.62242964e-01 -6.65726420e-03 -1.44956842e-01
4.01328892e-01 5.94056249e-01 -1.09698045e+00 -3.31204712e-01
8.60731840e-01 5.45475185e-01 -1.99827060e-01 8.58510017e-01
2.13136703e-01 8.51518750e-01 -6.21323645e-01 -5.47809422e-01
-7.77398199e-02 5.64877510e-01 7.06870854e-01 1.37150025e+00
6.67327762e-01 1.42056420e-01 2.67260551e-01 4.00801599e-01
-4.49899524e-01 2.38038525e-01 -7.50418663e-01 -5.68547368e-01
9.48316455e-01 1.00503623e+00 -3.91475081e-01 -4.00001228e-01
-6.42345846e-01 1.42919445e+00 5.61051667e-02 3.70286018e-01
-4.01412249e-01 -6.83822691e-01 1.16418850e+00 -9.16610751e-03
1.51208326e-01 -5.03921032e-01 9.40195695e-02 -1.41324151e+00
2.13682458e-01 -1.16638458e+00 6.83582798e-02 -7.53859520e-01
-1.53867066e+00 5.46527922e-01 -3.03136647e-01 -9.73367989e-01
-5.89128196e-01 -4.06078786e-01 -5.73438108e-01 5.71932912e-01
-8.59175920e-01 -9.79678333e-01 -2.64643252e-01 5.53059220e-01
1.19502151e+00 -4.64695543e-01 1.27655292e+00 5.45002878e-01
-3.16919118e-01 1.05213594e+00 3.07780862e-01 5.50558092e-03
9.63982224e-01 -7.33156264e-01 9.71671581e-01 6.24697387e-01
6.81251049e-01 6.45714700e-01 6.05160475e-01 -5.52202106e-01
-1.89558268e+00 -7.63505459e-01 1.10232341e+00 -4.52461809e-01
9.45179760e-01 -1.18452573e+00 -8.27906370e-01 4.86722022e-01
1.12815864e-01 5.68275377e-02 1.01942027e+00 3.81312013e-01
-4.36736256e-01 -4.78345335e-01 -7.28255033e-01 7.00042307e-01
1.17716753e+00 -1.37898242e+00 -4.41009521e-01 2.84761041e-01
1.37930334e+00 -1.40117183e-01 -5.34999430e-01 2.23633021e-01
9.60879564e-01 -5.40935338e-01 8.51955473e-01 -6.64738059e-01
-4.10591394e-01 3.02203149e-01 -4.94573474e-01 -1.30273247e+00
8.78347307e-02 -1.19456363e+00 -5.33959866e-01 1.23877561e+00
4.59661335e-01 -6.30659640e-01 4.85002160e-01 6.12815499e-01
-1.80193707e-01 -4.55456764e-01 -1.12760150e+00 -8.06704283e-01
-4.92280155e-01 -1.02359831e+00 5.81570387e-01 7.68059373e-01
4.15768892e-01 6.86027110e-01 -6.30676568e-01 -1.65283248e-01
2.36383840e-01 -3.64039212e-01 7.69675672e-01 -1.01008785e+00
-5.64467251e-01 -1.60798132e-01 -5.06500781e-01 -1.38706958e+00
1.12011805e-01 -6.70115113e-01 2.82667667e-01 -1.21857691e+00
-6.99553788e-02 -9.21702459e-02 -3.09875011e-01 5.34451842e-01
3.59237231e-02 1.93143323e-01 1.56902313e-01 -7.93329328e-02
-1.10220969e+00 8.00169885e-01 6.02426589e-01 -3.13027024e-01
-4.35740441e-01 1.40649334e-01 -7.02360392e-01 5.12662351e-01
5.87440610e-01 -9.65300351e-02 -4.07517493e-01 -7.14003444e-01
4.15410936e-01 -2.34889075e-01 -9.92979705e-02 -1.25242519e+00
5.95891774e-01 2.74752408e-01 -8.37842748e-02 -2.92615741e-01
7.01250732e-01 -6.79352760e-01 -4.33187872e-01 -3.46829705e-02
-3.33873153e-01 -8.02436247e-02 2.94391125e-01 4.26748961e-01
1.13284634e-02 2.57821530e-01 2.01408789e-01 1.95325091e-01
-9.58099365e-01 2.46693835e-01 -5.51257730e-01 2.83231199e-01
5.96165895e-01 -1.07097790e-01 -1.95853654e-02 -1.03004801e+00
-8.11658621e-01 1.72868311e-01 3.49039473e-02 1.20286632e+00
7.94032693e-01 -1.29528606e+00 -5.11042595e-01 4.94007558e-01
4.99464929e-01 -2.54751533e-01 1.47687122e-01 3.83494645e-01
-4.72533032e-02 4.13114160e-01 3.68937880e-01 -6.67630553e-01
-1.30872583e+00 2.58502662e-01 1.11721948e-01 5.14157534e-01
-5.15848219e-01 1.17062783e+00 -2.92520046e-01 -5.14251888e-01
9.80601907e-01 -5.06807387e-01 3.41737449e-01 9.79044959e-02
1.01584291e+00 4.03756410e-01 3.17699999e-01 -4.21962559e-01
-5.68048537e-01 2.03046769e-01 -1.18466623e-01 -4.41647261e-01
1.37860405e+00 -3.52952808e-01 1.65182829e-01 1.32061183e+00
1.32266879e+00 7.89058134e-02 -8.38518381e-01 -6.19781494e-01
-1.31754518e-01 -1.45833001e-01 4.25941832e-02 -8.32010567e-01
-5.24715066e-01 1.07424068e+00 8.75077486e-01 6.84725195e-02
8.97614717e-01 2.23559916e-01 1.63490713e+00 9.04038966e-01
7.15669751e-01 -1.23036015e+00 2.56317198e-01 8.77187431e-01
5.65314293e-01 -1.36060834e+00 -5.14150441e-01 1.87792614e-01
-5.62089622e-01 7.94296682e-01 2.92830408e-01 3.94709378e-01
8.92171144e-01 6.19259775e-01 1.96997985e-01 2.70284992e-02
-9.35525358e-01 -1.53684288e-01 3.96738112e-01 4.97366071e-01
4.75447893e-01 1.80581182e-01 3.58859181e-01 1.02090156e+00
-3.29317629e-01 1.01820447e-01 -9.47627351e-02 8.45219970e-01
-6.61689579e-01 -9.24377143e-01 -5.04708998e-02 6.14066422e-01
-4.17976230e-01 -5.50227284e-01 -4.61963177e-01 1.36709005e-01
-2.07497284e-01 1.29202950e+00 2.54760832e-01 -7.75006413e-01
4.18987840e-01 6.31885231e-01 1.61966547e-01 -9.48675394e-01
-5.33506691e-01 1.67726185e-02 1.41303122e-01 -6.31509542e-01
1.88907325e-01 -4.12671387e-01 -1.21382916e+00 -3.47741008e-01
-1.96573392e-01 -3.55363265e-02 8.10450375e-01 1.01755106e+00
8.20737422e-01 3.52325857e-01 5.15452206e-01 -8.66122305e-01
-7.68044412e-01 -1.03752851e+00 -5.34576833e-01 4.92695123e-02
5.30340612e-01 -1.41258970e-01 -1.94024041e-01 -5.59562594e-02] | [14.276729583740234, 6.362946033477783] |
9cdc2413-2391-4c25-89f3-539ec0933e22 | 100-things-you-always-wanted-to-know-about | null | null | https://aclanthology.org/N12-4001 | https://aclanthology.org/N12-4001.pdf | 100 Things You Always Wanted to Know about Linguistics But Were Afraid to Ask* | null | ['Emily M. Bender'] | 2012-06-01 | null | null | null | naacl-2012-6 | ['temporal-action-proposal-generation'] | ['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] | [-7.139074802398682, 3.6218371391296387] |
159aa745-20d7-4e56-8f34-86c7ca873faf | comprehensive-evaluation-of-no-reference | 2010.09414 | null | https://arxiv.org/abs/2010.09414v2 | https://arxiv.org/pdf/2010.09414v2.pdf | Comprehensive evaluation of no-reference image quality assessment algorithms on KADID-10k database | The main goal of objective image quality assessment is to devise computational, mathematical models which are able to predict perceptual image quality consistently with subjective evaluations. The evaluation of objective image quality assessment algorithms is based on experiments conducted on publicly available benchmark databases. In this study, our goal is to give a comprehensive evaluation about no-reference image quality assessment algorithms, whose original source codes are available online, using the recently published KADID-10k database which is one of the largest available benchmark databases. Specifically, average PLCC, SROCC, and KROCC are reported which were measured over 100 random train-test splits. Furthermore, the database was divided into a train (appx. 80\% of images) and a test set (appx. 20% of images) with respect to the reference images. So no semantic content overlap was between these two sets. Our evaluation results may be helpful to obtain a clear understanding about the status of state-of-the-art no-reference image quality assessment methods. | ['Domonkos Varga'] | 2020-10-19 | null | null | null | null | ['no-reference-image-quality-assessment'] | ['computer-vision'] | [ 2.10494354e-01 -3.24972034e-01 -1.12872295e-01 -3.81241143e-01
-1.11691248e+00 -3.16178232e-01 5.48754156e-01 2.15638801e-01
-4.24011946e-01 7.63346493e-01 -4.24455702e-02 -4.85185459e-02
-2.75377363e-01 -5.31160414e-01 -2.57901579e-01 -6.80554271e-01
-2.47550845e-01 3.13515998e-02 3.05624306e-01 -1.75188422e-01
4.85779107e-01 4.05563593e-01 -1.86675096e+00 3.03195864e-01
9.92354453e-01 1.44755459e+00 2.63301700e-01 8.49102378e-01
3.27845633e-01 8.24815571e-01 -9.13086236e-01 -7.26895630e-01
2.36047477e-01 -5.89510918e-01 -8.65315676e-01 3.06858212e-01
5.46375275e-01 -2.36916393e-01 -1.67547151e-01 1.28068113e+00
7.25542724e-01 6.59345910e-02 7.28253961e-01 -1.25401592e+00
-8.79591405e-01 1.41116381e-01 -4.38824087e-01 7.60092795e-01
4.04305667e-01 5.68127513e-01 8.75457287e-01 -6.50706649e-01
5.84654391e-01 1.00852692e+00 3.29485327e-01 2.82167166e-01
-1.18789184e+00 -4.02858496e-01 -3.73733133e-01 6.74906135e-01
-1.48211098e+00 -5.81326127e-01 5.48117697e-01 -4.16673213e-01
5.13362348e-01 4.00598347e-01 2.48355225e-01 7.54443467e-01
4.46294457e-01 3.72199178e-01 1.52777696e+00 -6.82596445e-01
3.38976473e-01 4.36061323e-01 1.10569187e-01 5.06927431e-01
2.58068860e-01 4.60993081e-01 -4.60011572e-01 1.47011817e-01
5.27296185e-01 -7.53379524e-01 -6.16849124e-01 -3.70506585e-01
-1.14224207e+00 6.29425347e-01 4.50094253e-01 5.88907063e-01
-2.73148119e-01 -2.20544964e-01 3.81743073e-01 5.48396766e-01
4.11623687e-01 4.45800245e-01 -1.83632195e-01 -1.09096885e-01
-1.03692341e+00 -5.38959503e-02 5.68868756e-01 6.58261120e-01
6.05832279e-01 1.89685151e-02 -2.92939067e-01 1.12116361e+00
2.31258929e-01 7.13090181e-01 5.59812903e-01 -1.34300983e+00
3.61937046e-01 2.88819253e-01 2.11126894e-01 -1.19638920e+00
1.05186105e-01 -4.68389511e-01 -8.32609296e-01 7.68556476e-01
4.79217798e-01 2.52393544e-01 -8.16149771e-01 1.35392308e+00
-2.09086478e-01 -2.20558539e-01 1.86812386e-01 1.04046798e+00
1.06197488e+00 5.38936555e-01 1.59980088e-01 -3.14839393e-01
1.43409514e+00 -7.57601857e-01 -6.89275563e-01 2.84264356e-01
2.36541599e-01 -1.09320474e+00 1.30323946e+00 6.58117115e-01
-1.32169330e+00 -1.16242623e+00 -1.31946313e+00 1.87314689e-01
-3.97891968e-01 3.57053787e-01 1.45887449e-01 9.87550080e-01
-1.33608305e+00 5.54570973e-01 -3.36361051e-01 -3.42339993e-01
1.90020472e-01 -1.67232919e-02 -4.83262092e-01 -3.24895740e-01
-1.16522610e+00 1.12132370e+00 4.00046766e-01 -1.04109950e-01
-1.01956367e+00 -5.31639934e-01 -6.96539164e-01 -2.52428830e-01
2.86188871e-01 -4.52012748e-01 9.98074353e-01 -1.33318496e+00
-1.36461091e+00 1.26195216e+00 4.98232469e-02 -3.25122714e-01
4.54383373e-01 1.76749960e-01 -9.87734318e-01 6.30081952e-01
3.57956402e-02 5.89057565e-01 7.17742622e-01 -1.72415352e+00
-6.57954037e-01 -1.99627444e-01 4.84590679e-02 1.85341612e-01
9.59543884e-02 2.70338386e-01 -7.64528096e-01 -7.70283401e-01
-1.71837494e-01 -6.17218137e-01 5.26044704e-02 5.96584193e-03
-2.30968058e-01 3.38320173e-02 4.21600014e-01 -7.20117450e-01
1.28351521e+00 -2.25117517e+00 -1.87246814e-01 2.05600455e-01
1.61914766e-01 3.54080886e-01 -4.63407934e-01 1.71269655e-01
-1.88718408e-01 2.32406422e-01 -1.05795428e-01 -1.87678993e-01
-5.08760102e-02 3.49089168e-02 8.07240456e-02 5.43715417e-01
1.92715108e-01 5.53140402e-01 -6.32664263e-01 -8.71100008e-01
5.71920753e-01 2.96330273e-01 -7.73339197e-02 2.29266524e-01
2.52448559e-01 2.34411612e-01 -4.09810580e-02 6.48564637e-01
7.26403952e-01 -1.21416241e-01 -3.07585448e-01 -5.75835824e-01
-7.32946396e-03 -1.39098018e-01 -1.24190593e+00 1.48916495e+00
-3.37720305e-01 8.60554576e-01 -7.05496445e-02 -6.81897938e-01
7.46231973e-01 3.96358252e-01 3.82745117e-01 -1.32477522e+00
1.93724409e-01 1.39004499e-01 1.55560449e-01 -3.98436517e-01
5.73433220e-01 -2.46088319e-02 1.77393228e-01 1.49512708e-01
2.95016855e-01 -2.73096681e-01 8.00849855e-01 5.90104870e-02
7.33817995e-01 -2.72969961e-01 2.73393691e-01 -4.09221560e-01
9.52061832e-01 -9.07705128e-02 2.74539590e-01 7.66480744e-01
-6.61985040e-01 9.66664791e-01 4.08176661e-01 -2.24706382e-01
-1.13767207e+00 -1.34733737e+00 -4.36086416e-01 7.22699940e-01
4.82251942e-01 -2.41176590e-01 -9.06328917e-01 -4.00093615e-01
-4.53094095e-01 5.98204017e-01 -5.88985980e-01 4.87065129e-02
-1.61606416e-01 -9.01928604e-01 4.25064385e-01 1.14125878e-01
7.28708506e-01 -1.29064143e+00 -5.18292844e-01 3.36034410e-02
-2.48763666e-01 -1.13234138e+00 -2.95905620e-01 -1.63385987e-01
-6.99601352e-01 -1.45254338e+00 -9.65379655e-01 -5.74686170e-01
4.56397682e-01 1.61637992e-01 1.71460652e+00 2.92191088e-01
-3.94466996e-01 3.79963964e-01 -6.25243962e-01 -1.17533833e-01
-6.69886410e-01 -2.29001477e-01 -1.98463723e-01 -1.32711947e-01
1.55170724e-01 -1.68805614e-01 -8.33662689e-01 8.32074106e-01
-1.15366554e+00 -3.14495921e-01 7.49299645e-01 6.28291726e-01
8.27780187e-01 4.56795037e-01 1.86688855e-01 -3.45770955e-01
5.96703827e-01 -1.06137596e-01 -6.60879672e-01 4.15761679e-01
-1.00506938e+00 -1.19668029e-01 4.12546426e-01 -1.56067282e-01
-1.10277796e+00 -5.77591240e-01 -1.77596733e-01 -1.37269452e-01
-3.84963840e-01 2.82785177e-01 -3.11550438e-01 -2.30496556e-01
7.81414986e-01 2.39728436e-01 -5.68073206e-02 -3.56500506e-01
2.44143948e-01 6.96156502e-01 8.10004592e-01 -4.66021210e-01
6.66593492e-01 1.62584305e-01 -2.45436475e-01 -8.83788407e-01
-4.36451048e-01 -6.12468302e-01 -4.73128498e-01 -3.91468942e-01
7.84208417e-01 -7.66124725e-01 -4.01919484e-01 6.43881500e-01
-6.68508470e-01 -2.44632870e-01 -3.12722683e-01 5.73313177e-01
-8.26659560e-01 5.99744976e-01 -5.45703948e-01 -4.65251178e-01
-3.31203461e-01 -1.44963038e+00 8.16639781e-01 2.97749758e-01
-5.76104820e-02 -9.76038754e-01 6.61923587e-02 5.97700596e-01
4.66930389e-01 5.22921048e-02 6.93358302e-01 -1.43301278e-01
-4.21752930e-01 -1.20660499e-01 -5.42504013e-01 9.50573027e-01
7.45761395e-02 1.49337023e-01 -8.90739679e-01 -4.21361148e-01
3.08293961e-02 -3.35273415e-01 6.52922332e-01 4.97222304e-01
1.27599490e+00 -3.66465934e-02 8.99318829e-02 4.18130308e-01
1.84559536e+00 3.68166536e-01 1.24498701e+00 4.11317348e-01
2.39884909e-02 4.35075730e-01 8.86675358e-01 1.25887826e-01
9.18181539e-02 9.00366843e-01 3.87200296e-01 -3.46159905e-01
-6.09412670e-01 2.86296993e-01 2.22263470e-01 8.54076743e-01
-2.27388829e-01 -5.03765881e-01 -6.64292634e-01 4.03681129e-01
-1.15553844e+00 -1.05171525e+00 -5.79518415e-02 2.26838684e+00
6.93906724e-01 2.33965710e-01 1.03126183e-01 7.60960877e-01
6.57882214e-01 1.88543543e-01 -2.19198659e-01 -2.26344123e-01
-2.77372360e-01 1.30405858e-01 4.66960341e-01 2.80407786e-01
-1.17896664e+00 2.51839697e-01 7.28110218e+00 1.21319449e+00
-1.06115079e+00 1.27818108e-01 1.14516985e+00 8.11376348e-02
1.75692603e-01 -3.88215482e-01 -3.71713310e-01 7.10118771e-01
1.22976983e+00 -8.60422626e-02 2.27574870e-01 5.56193471e-01
3.76961827e-01 -6.82777107e-01 -8.68028879e-01 1.29509747e+00
3.22850317e-01 -9.08980429e-01 -1.11518338e-01 3.35771311e-03
6.73121572e-01 -8.46110508e-02 3.27740312e-01 -3.99756245e-02
2.25077681e-02 -1.19741833e+00 7.33391464e-01 6.21734798e-01
1.11227846e+00 -6.59090102e-01 1.14461756e+00 -4.06237952e-02
-1.03115487e+00 2.19067365e-01 -3.98316562e-01 3.71607542e-01
3.21688987e-02 7.51052082e-01 -1.61130309e-01 8.99350345e-01
1.01342392e+00 4.08291578e-01 -1.10465968e+00 1.38631630e+00
5.03703952e-02 4.67174113e-01 1.68294787e-01 3.19463909e-01
-5.50695520e-04 -1.86367899e-01 3.64787608e-01 1.17704761e+00
3.64669800e-01 1.78891718e-01 -1.40164971e-01 6.21433020e-01
5.75586669e-02 3.77558082e-01 -1.89219266e-01 1.30314842e-01
4.22456078e-02 1.08088291e+00 -9.16173756e-01 -4.15301412e-01
-4.64420021e-01 9.39498961e-01 -2.28263319e-01 2.80364990e-01
-7.73367763e-01 -5.45360565e-01 4.30453807e-01 -2.34510638e-02
2.74034757e-02 5.77675588e-02 -1.73262611e-01 -1.11265635e+00
-5.28581860e-03 -1.32284129e+00 4.03915405e-01 -1.18418276e+00
-1.41925013e+00 8.80777717e-01 2.72154987e-01 -1.45198274e+00
-1.40129805e-01 -7.13222265e-01 -2.37819403e-01 1.08436584e+00
-1.56832194e+00 -4.77175772e-01 -7.28298545e-01 5.59728682e-01
5.05175889e-01 -2.90849894e-01 6.68486357e-01 4.69778985e-01
-2.91944653e-01 6.87144279e-01 2.70991951e-01 2.02102467e-01
7.35953987e-01 -1.31665587e+00 -5.19798025e-02 8.86851192e-01
1.42687902e-01 8.77797902e-02 9.53369439e-01 -1.22531518e-01
-7.65584528e-01 -7.15567708e-01 4.83574212e-01 -4.77225363e-01
2.66403258e-01 3.99603903e-01 -8.90149832e-01 -1.88231781e-01
5.12352765e-01 1.81907788e-01 6.77527785e-01 -4.44360107e-01
-1.46924973e-01 -4.18040663e-01 -1.36039603e+00 2.99663097e-01
6.65770292e-01 -3.86738241e-01 -4.83740419e-01 -2.27316126e-01
1.94021523e-01 -1.98809043e-01 -1.22394407e+00 5.23164451e-01
4.55968589e-01 -1.58781731e+00 1.13170731e+00 2.27161184e-01
5.49126267e-01 -5.18121779e-01 -3.21549058e-01 -1.35975051e+00
-1.54900745e-01 9.96945128e-02 2.46927455e-01 1.19971704e+00
2.68868446e-01 -4.25942302e-01 5.60131550e-01 1.15839377e-01
8.08282942e-02 -4.93194431e-01 -7.57168651e-01 -9.06225443e-01
-8.00519958e-02 -4.59240139e-01 4.89245921e-01 5.55538297e-01
-4.66512799e-01 5.38942877e-05 -1.86682418e-01 -6.16666935e-02
9.31190491e-01 -7.23267794e-02 5.62499285e-01 -1.03860176e+00
-2.62625664e-01 -6.07518792e-01 -7.80867338e-01 -4.97729093e-01
-2.47792751e-01 -3.87579411e-01 2.02756878e-02 -1.48309565e+00
3.50152045e-01 -3.38560194e-01 -5.73464692e-01 5.91831729e-02
-1.88131332e-01 9.57135677e-01 2.39677504e-01 2.68043369e-01
-6.45413101e-01 4.12922174e-01 1.48335338e+00 -2.74854243e-01
1.44790724e-01 2.94017904e-02 -5.39643526e-01 4.61150438e-01
8.92982185e-01 -2.71012664e-01 -6.74110234e-01 -1.21257685e-01
-2.14050859e-01 1.33409500e-01 3.20080101e-01 -1.53152561e+00
-1.74718037e-01 -6.51113391e-02 5.37507951e-01 -6.17349803e-01
1.54282302e-01 -8.24107289e-01 3.71915191e-01 4.12965477e-01
-4.27506179e-01 1.80295080e-01 5.95074743e-02 3.04660738e-01
-7.46960640e-01 -4.90588903e-01 1.27618515e+00 -1.51900142e-01
-1.10277617e+00 1.25385419e-01 -1.34532616e-01 1.11504346e-01
1.06200624e+00 -5.43488204e-01 -2.91392893e-01 -4.44042891e-01
-7.65910864e-01 -1.91776127e-01 7.61354923e-01 3.41459095e-01
6.73438489e-01 -1.29697323e+00 -8.45387459e-01 -6.89190701e-02
5.26726902e-01 -6.50371194e-01 3.90481114e-01 6.12551868e-01
-8.15618694e-01 4.56478804e-01 -6.07534170e-01 -5.05890429e-01
-1.58607233e+00 6.95984781e-01 5.36127448e-01 -4.06591713e-01
-2.39533469e-01 3.95744324e-01 -1.91579945e-02 8.04807320e-02
2.89576054e-01 -1.00823328e-01 -4.61095244e-01 -2.08428457e-01
7.81612635e-01 5.42508602e-01 2.93947339e-01 -9.85836089e-01
-1.45931542e-01 5.86242795e-01 2.29716361e-01 -2.65050858e-01
8.41325223e-01 -4.62819338e-01 -2.57915836e-02 3.52128595e-01
1.31557894e+00 -1.19136550e-01 -8.66247714e-01 -4.99671586e-02
-3.19473222e-02 -7.36507177e-01 1.65247917e-01 -1.27837706e+00
-1.26987946e+00 7.82574534e-01 1.46290553e+00 4.14184868e-01
1.77898586e+00 -1.26633316e-01 2.32072890e-01 -4.16960344e-02
4.99813348e-01 -9.66548383e-01 2.86686361e-01 5.32990173e-02
9.66186523e-01 -1.52109122e+00 1.03246786e-01 -4.83721614e-01
-6.83307886e-01 9.37449217e-01 4.08012152e-01 1.18161649e-01
3.83912563e-01 -8.05936679e-02 3.97753090e-01 -1.27126247e-01
-4.90696400e-01 -3.60286385e-01 6.79137826e-01 8.73030126e-01
4.59783167e-01 -5.49993031e-02 -4.76698339e-01 2.22492173e-01
-1.13284931e-01 2.09265366e-01 5.07750988e-01 4.31679249e-01
-3.95246327e-01 -1.12526035e+00 -6.24809027e-01 4.50323403e-01
-8.83456945e-01 1.91196240e-02 1.35106072e-01 8.83281767e-01
3.12632740e-01 1.38966978e+00 -9.54686552e-02 -2.43460208e-01
4.96013790e-01 -4.16022718e-01 5.17653346e-01 -1.79278240e-01
-3.81007850e-01 -2.26234511e-01 -2.64428947e-02 -6.64056659e-01
-7.53118098e-01 -3.16428453e-01 -5.79460979e-01 -4.18614805e-01
-1.12255096e-01 2.93381959e-01 7.13089466e-01 5.92541575e-01
-5.97759522e-02 6.12128317e-01 6.44992173e-01 -7.79190004e-01
-3.56153667e-01 -9.66881871e-01 -8.19547474e-01 7.64806867e-01
1.91815421e-01 -6.79782212e-01 -5.37323952e-01 4.86543924e-01] | [11.750015258789062, -1.8983701467514038] |
2543d645-3d56-48bd-8d5c-fc65adcbb691 | person-re-identification-with-correspondence | 1504.06243 | null | http://arxiv.org/abs/1504.06243v1 | http://arxiv.org/pdf/1504.06243v1.pdf | Person Re-identification with Correspondence Structure Learning | This paper addresses the problem of handling spatial misalignments due to
camera-view changes or human-pose variations in person re-identification. We
first introduce a boosting-based approach to learn a correspondence structure
which indicates the patch-wise matching probabilities between images from a
target camera pair. The learned correspondence structure can not only capture
the spatial correspondence pattern between cameras but also handle the
viewpoint or human-pose variation in individual images. We further introduce a
global-based matching process. It integrates a global matching constraint over
the learned correspondence structure to exclude cross-view misalignments during
the image patch matching process, hence achieving a more reliable matching
score between images. Experimental results on various datasets demonstrate the
effectiveness of our approach. | ['Mingliang Xu', 'Yang Shen', 'Junchi Yan', 'Weiyao Lin', 'Jingdong Wang', 'Jianxin Wu'] | 2015-04-23 | person-re-identification-with-correspondence-1 | http://openaccess.thecvf.com/content_iccv_2015/html/Shen_Person_Re-Identification_With_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Shen_Person_Re-Identification_With_ICCV_2015_paper.pdf | iccv-2015-12 | ['patch-matching'] | ['computer-vision'] | [ 1.76602185e-01 -6.08090162e-01 -1.66709006e-01 -6.64989531e-01
-7.35817671e-01 -6.39705300e-01 6.52672410e-01 -9.08614323e-02
-2.82191128e-01 2.38167286e-01 1.22745380e-01 2.57685363e-01
-2.25822791e-03 -5.45990229e-01 -7.84125686e-01 -5.49652219e-01
3.49938989e-01 2.81906635e-01 2.51432240e-01 2.49394551e-01
3.88002515e-01 5.35128713e-01 -1.46530282e+00 1.18479244e-01
6.32981539e-01 7.54772186e-01 1.84348851e-01 4.29324269e-01
4.01194006e-01 1.23575769e-01 -5.65628111e-01 -7.59486079e-01
7.59851813e-01 -2.94127941e-01 -5.53828657e-01 7.18610168e-01
1.20445406e+00 -1.95863754e-01 -2.91999668e-01 1.39103448e+00
4.67880189e-01 1.17020935e-01 2.59000510e-01 -1.09725225e+00
-3.17262828e-01 -2.80042082e-01 -8.77342641e-01 2.97412306e-01
9.21118438e-01 1.29041886e-02 7.68239558e-01 -1.01196516e+00
4.24421966e-01 1.35165536e+00 9.45560992e-01 3.46838892e-01
-1.12105465e+00 -6.11565113e-01 3.00576866e-01 5.23968697e-01
-1.61138761e+00 -3.38825673e-01 8.47739875e-01 -4.87064332e-01
5.78803241e-01 4.14177626e-01 6.90954030e-01 8.11296701e-01
7.02032000e-02 4.62026745e-01 1.14221251e+00 -6.47278666e-01
-1.44150913e-01 1.49438918e-01 1.67687476e-01 5.26444435e-01
3.64995956e-01 3.19064558e-01 -5.39035261e-01 -4.46438879e-01
1.00462973e+00 3.03444117e-01 -3.80484194e-01 -7.43431330e-01
-1.27394736e+00 5.19091904e-01 4.55640137e-01 1.22597754e-01
-1.53101727e-01 -3.64853084e-01 1.76520988e-01 1.65105999e-01
8.88991654e-02 1.81193486e-01 -1.30579084e-01 2.19422758e-01
-6.75111055e-01 2.76965708e-01 3.92237067e-01 1.25296426e+00
8.77501249e-01 -5.20571709e-01 -8.61034170e-02 1.00647366e+00
1.78049013e-01 6.09726250e-01 3.06352317e-01 -6.14766836e-01
8.52842629e-01 6.23405814e-01 2.78018773e-01 -1.41359091e+00
-1.13446347e-01 -2.23211020e-01 -7.94339895e-01 -1.61606371e-01
5.23853302e-01 3.83289665e-01 -5.11755884e-01 1.55622900e+00
5.67408204e-01 3.34333181e-01 -4.05707181e-01 9.74411309e-01
5.48095584e-01 6.99635148e-02 -2.85799921e-01 -3.16944234e-02
1.57900131e+00 -9.52669382e-01 -6.00837231e-01 -4.50490892e-01
9.02398527e-02 -9.54474926e-01 5.40027857e-01 1.55536354e-01
-1.01658213e+00 -1.00009716e+00 -9.97683764e-01 2.19498992e-01
-3.42704281e-02 2.66843259e-01 1.61283687e-02 8.24383795e-01
-9.46918309e-01 2.52398074e-01 -5.12091517e-01 -4.70448911e-01
-7.36773422e-04 5.30845702e-01 -7.75878608e-01 -4.44859058e-01
-9.47626472e-01 9.24750149e-01 3.69654074e-02 1.85945019e-01
-2.74836093e-01 -4.88590628e-01 -1.01840127e+00 -2.17856407e-01
2.44873241e-01 -7.99793780e-01 8.45611036e-01 -9.76455390e-01
-1.08350575e+00 1.14527500e+00 -6.98421538e-01 1.84372999e-02
5.10443807e-01 -1.75389737e-01 -5.23296356e-01 1.09449618e-01
2.54903466e-01 3.00690770e-01 1.00968659e+00 -1.41066933e+00
-6.95072114e-01 -8.37031782e-01 -1.44123271e-01 5.34129083e-01
-1.56240180e-01 1.38372526e-01 -9.98270392e-01 -8.00110161e-01
3.51160556e-01 -1.15593719e+00 -5.25435023e-02 -1.12700924e-01
-3.14554304e-01 4.39196043e-02 5.26553035e-01 -9.10678327e-01
9.72433448e-01 -2.07736754e+00 2.79613793e-01 5.58219194e-01
-8.51962417e-02 1.01189427e-01 -1.30572543e-01 3.06176960e-01
-2.13051572e-01 -3.25826138e-01 5.87834977e-02 -6.08364403e-01
-3.19401741e-01 6.71182647e-02 -8.20117593e-02 8.19881201e-01
-4.17888425e-02 6.43795252e-01 -6.60250902e-01 -4.35336024e-01
5.26872993e-01 5.10824561e-01 -3.56637388e-01 5.18051267e-01
5.94810426e-01 6.94036841e-01 -4.55259308e-02 7.12097883e-01
1.19335437e+00 2.57291887e-02 2.75395036e-01 -5.66426516e-01
1.88884541e-01 -2.89120581e-02 -1.51838839e+00 1.57255459e+00
-3.37682873e-01 2.31374800e-01 8.46182108e-02 -7.24628270e-01
1.08588612e+00 5.35939373e-02 4.59376305e-01 -5.91359258e-01
-4.07039709e-02 -8.48016292e-02 -3.66665453e-01 -2.05793560e-01
4.53617215e-01 1.01507813e-01 -4.29783203e-02 1.16566636e-01
-1.94681689e-01 3.60902190e-01 -2.60740846e-01 -2.03683600e-01
4.82924134e-01 -1.10489734e-01 4.92262721e-01 -2.25253552e-01
1.06934404e+00 -5.05209506e-01 7.69794822e-01 7.55440652e-01
-4.15678710e-01 1.02354097e+00 -2.38889575e-01 -8.47359478e-01
-1.01715064e+00 -1.02304780e+00 -2.31575578e-01 5.63309431e-01
8.56289566e-01 -4.95635122e-01 -8.03047895e-01 -7.48874784e-01
2.92177927e-02 2.49456102e-03 -6.19195938e-01 -7.65449256e-02
-7.75963008e-01 -6.94342315e-01 2.36480609e-01 6.84514642e-01
8.62678289e-01 -3.52237016e-01 -2.14089021e-01 -1.67399332e-01
-5.22605956e-01 -1.46604168e+00 -1.15156925e+00 -5.82558513e-01
-7.73643494e-01 -1.34553373e+00 -7.29671359e-01 -9.09884572e-01
1.19923222e+00 9.78451312e-01 1.01146448e+00 2.24959210e-01
-3.94204050e-01 7.76666224e-01 -1.66787550e-01 1.92171410e-01
-4.12696190e-02 -3.51609498e-01 2.29596764e-01 5.61771512e-01
7.90471196e-01 -2.29848713e-01 -7.41355419e-01 1.06394863e+00
-3.87880296e-01 -2.40913764e-01 2.56807715e-01 1.07832181e+00
7.06223607e-01 2.16922328e-01 -1.85944494e-02 -4.46105838e-01
1.93977162e-01 3.79554480e-02 -6.62249684e-01 6.15193307e-01
-2.79428869e-01 -3.96836728e-01 1.88586473e-01 -4.84228820e-01
-9.13101494e-01 4.26185250e-01 1.68707058e-01 -4.99145836e-01
-2.48570576e-01 -8.32622498e-02 -7.63256073e-01 -5.02750099e-01
2.15962455e-01 3.39352578e-01 1.18154916e-03 -4.00395423e-01
6.45330623e-02 5.54235280e-01 9.12827075e-01 -5.82625806e-01
1.16669559e+00 6.96245968e-01 -1.41083241e-01 -6.54640436e-01
-5.17172217e-01 -1.13132453e+00 -1.23700535e+00 -2.86958694e-01
8.33392560e-01 -1.24009645e+00 -6.74428880e-01 6.49836659e-01
-1.17809570e+00 4.06281352e-01 1.58427790e-01 4.69408810e-01
-3.53571802e-01 9.52643752e-01 -2.65272141e-01 -6.42235041e-01
-9.28190276e-02 -1.30684912e+00 1.32133901e+00 3.45067710e-01
-1.48336217e-01 -1.09676933e+00 1.94132805e-01 6.02878571e-01
-1.35327622e-01 -8.88759941e-02 3.02820712e-01 -3.14575851e-01
-6.76755130e-01 -6.84591413e-01 -1.81224465e-01 1.26833409e-01
4.57123607e-01 -3.80602866e-01 -8.28728080e-01 -7.43978560e-01
8.83947015e-02 2.90671706e-01 4.68925685e-01 2.50312597e-01
8.96066546e-01 -1.05748557e-01 -4.87903833e-01 8.01678896e-01
1.36843789e+00 4.77931574e-02 6.87423527e-01 5.41079462e-01
9.92383957e-01 8.65267694e-01 7.44373143e-01 4.20794606e-01
5.61366439e-01 1.46003234e+00 1.55362815e-01 -2.00704962e-01
-2.55051050e-02 -4.96149987e-01 2.17879161e-01 5.75598776e-01
-9.59832370e-02 1.29795253e-01 -5.29033005e-01 3.34597617e-01
-1.93536699e+00 -1.15339816e+00 -1.20717749e-01 2.89183426e+00
2.96122044e-01 -3.36995423e-01 4.32997376e-01 3.82132679e-02
1.43751776e+00 -1.61102116e-02 -3.42697114e-01 9.82472077e-02
-2.33312652e-01 -3.87127489e-01 6.12660527e-01 7.36007929e-01
-1.25516868e+00 5.46109259e-01 6.75349474e+00 5.46473205e-01
-5.22361279e-01 -5.90385832e-02 3.10790986e-01 2.60789305e-01
-1.75037477e-02 -1.92000773e-02 -1.25959587e+00 5.57647109e-01
7.52368718e-02 2.56171167e-01 3.44869316e-01 6.88801169e-01
-2.26084784e-01 -3.96289863e-03 -1.10924256e+00 1.67588854e+00
6.41485870e-01 -9.41611350e-01 -3.09466440e-02 3.93570065e-01
7.10806966e-01 -5.18985033e-01 8.08252543e-02 -2.81296313e-01
2.67705717e-03 -5.59955180e-01 5.22410333e-01 3.48390073e-01
5.96613228e-01 -7.60940850e-01 7.54897058e-01 2.04635724e-01
-1.49606681e+00 -2.32688412e-02 -6.01145029e-01 5.87040558e-02
5.87308779e-02 4.98875320e-01 -5.29618204e-01 6.94446266e-01
9.03773606e-01 9.76750076e-01 -8.26670408e-01 1.16268718e+00
-4.41114381e-02 -2.33888805e-01 -5.92950955e-02 5.76629400e-01
-4.59064096e-01 -4.50899839e-01 7.16669858e-01 1.11499047e+00
1.80443496e-01 -1.32300079e-01 4.07473445e-01 4.12377059e-01
2.65699655e-01 -1.31396344e-02 -6.56566679e-01 9.30905640e-01
7.36857474e-01 1.15861464e+00 -4.06881034e-01 -1.46653429e-01
-6.83535457e-01 1.51832557e+00 2.25070342e-01 2.40501508e-01
-7.38048136e-01 4.09446880e-02 8.01627100e-01 1.71069205e-01
4.95132864e-01 -1.89066499e-01 -1.70811504e-01 -1.54381561e+00
6.08153045e-01 -1.12507129e+00 6.29068315e-01 -3.96161646e-01
-1.64954352e+00 5.32203734e-01 1.66816428e-01 -1.58442187e+00
-6.84243664e-02 -4.59026039e-01 -6.47435009e-01 1.06846118e+00
-1.24802649e+00 -1.38019073e+00 -6.27558768e-01 9.44160283e-01
4.40947384e-01 -3.60675186e-01 6.75386548e-01 3.86714786e-01
-6.15102351e-01 1.00280893e+00 -2.92612817e-02 2.31396332e-01
1.05974317e+00 -1.12753677e+00 4.89077479e-01 1.16401994e+00
9.79583338e-02 1.08305073e+00 3.94314408e-01 -6.95603967e-01
-1.37973988e+00 -9.84584808e-01 9.58163381e-01 -8.52218211e-01
1.06587328e-01 -4.46227223e-01 -8.01317513e-01 7.65931964e-01
4.97134142e-02 6.11194782e-02 9.37396705e-01 2.36520350e-01
-8.67730975e-01 -4.66337949e-01 -1.18076825e+00 2.46650219e-01
1.07587230e+00 -6.96161985e-01 -6.08374536e-01 1.46481767e-01
-1.46552429e-01 -5.58215737e-01 -8.77468526e-01 4.10414577e-01
6.96816087e-01 -1.09556746e+00 1.53554654e+00 -8.22865516e-02
-3.25497299e-01 -5.51545441e-01 -1.87781096e-01 -1.04123724e+00
-6.75304174e-01 -5.05805850e-01 3.28797787e-01 1.46012592e+00
-2.78640926e-01 -6.12943292e-01 6.91130042e-01 9.34637547e-01
3.29932034e-01 -8.73897374e-02 -1.11964512e+00 -1.01664686e+00
-3.41050357e-01 7.00396001e-02 7.32906342e-01 8.77506077e-01
9.14723873e-02 3.16737033e-02 -7.67502606e-01 5.77237785e-01
9.33761060e-01 1.82052985e-01 1.13250840e+00 -1.14116526e+00
-5.11886358e-01 -1.33732781e-01 -8.44305634e-01 -1.25018954e+00
1.74076021e-01 -4.66414452e-01 -2.63031037e-03 -9.50961471e-01
7.53031433e-01 -2.44240284e-01 -5.44062853e-02 -7.48335710e-03
-6.18305743e-01 4.61820811e-01 2.42285863e-01 6.81332111e-01
-5.24633825e-01 2.65794784e-01 9.18436944e-01 -1.40618354e-01
-6.53136224e-02 2.70622283e-01 -5.62468290e-01 6.32623851e-01
5.41281402e-01 -2.65338302e-01 -2.31655151e-01 -4.87141550e-01
-1.64784625e-01 -1.73062459e-01 6.97338104e-01 -1.03315890e+00
3.07399064e-01 -6.55475184e-02 8.43137622e-01 -6.36236429e-01
4.43119407e-01 -1.21860909e+00 3.65491152e-01 4.57267374e-01
-6.61145300e-02 5.19657791e-01 3.71170719e-03 8.14957082e-01
-3.83335292e-01 -1.14873022e-01 9.84006464e-01 -1.09776169e-01
-5.19889116e-01 4.02111948e-01 8.41595456e-02 -3.49178702e-01
1.06471312e+00 -6.46486640e-01 -6.51157498e-02 -3.44178140e-01
-4.39567447e-01 -3.66633944e-02 8.71964037e-01 5.90217948e-01
6.41135693e-01 -1.75210142e+00 -5.69519699e-01 6.03080153e-01
5.12815356e-01 -2.91414708e-01 3.52755696e-01 6.24851704e-01
-2.05299944e-01 3.39638770e-01 -2.67714173e-01 -8.50431740e-01
-1.90725255e+00 7.14566886e-01 5.37061989e-01 -1.59510896e-01
-5.65055072e-01 8.53952110e-01 5.18037498e-01 -5.39160490e-01
2.60929674e-01 1.67754114e-01 -1.53211579e-01 -3.09726328e-01
8.93652081e-01 3.47681820e-01 2.26656199e-02 -1.26028514e+00
-6.91851854e-01 1.33662236e+00 -2.19972223e-01 -2.69010174e-03
6.52633488e-01 -7.10286736e-01 -1.25505686e-01 1.05347387e-01
1.31658268e+00 9.98444334e-02 -1.26056206e+00 -5.93576729e-01
-1.60890698e-01 -1.35486162e+00 -4.30919558e-01 -4.52818125e-01
-9.58723187e-01 6.67070210e-01 8.00485373e-01 -2.53288239e-01
1.16647971e+00 -1.65090218e-01 6.06399775e-01 1.10868188e-02
6.03299260e-01 -9.99230742e-01 1.58405989e-01 9.07230675e-02
8.17374170e-01 -1.42496443e+00 3.05938303e-01 -7.90969729e-01
-5.77758551e-01 1.04953730e+00 6.38467550e-01 -1.05908386e-01
4.64550942e-01 -2.20245302e-01 7.87696764e-02 1.43431440e-01
-6.70688450e-02 -1.28314178e-02 7.30355084e-01 1.01264679e+00
1.04564697e-01 -4.53234576e-02 2.78976634e-02 2.68016934e-01
3.73697467e-02 -3.80498350e-01 4.89091016e-02 7.19134986e-01
-9.99606997e-02 -1.49394846e+00 -1.00631285e+00 -2.45885521e-01
-1.87813893e-01 1.86778992e-01 -7.41681039e-01 5.70068300e-01
1.52432412e-01 1.00638497e+00 1.47120148e-01 -6.50129497e-01
6.27670825e-01 -3.05830270e-01 8.72055233e-01 -4.33256149e-01
-6.84052885e-01 2.64312953e-01 -2.69187987e-01 -5.48024833e-01
-7.12935984e-01 -1.06298184e+00 -3.70076776e-01 -3.61609876e-01
-5.59964240e-01 -1.19648926e-01 3.70779395e-01 6.45466983e-01
3.81059647e-01 -1.40738815e-01 1.00358999e+00 -9.41717625e-01
-4.79817092e-01 -6.23024583e-01 -4.37991083e-01 9.28953588e-01
3.70547682e-01 -7.69368947e-01 -2.84698963e-01 2.95524925e-01] | [14.747896194458008, 1.0005977153778076] |
aa14b869-a0de-40ee-ab6e-78582152eb71 | can-occupant-behaviors-affect-urban-energy | 2303.03006 | null | https://arxiv.org/abs/2303.03006v1 | https://arxiv.org/pdf/2303.03006v1.pdf | Can occupant behaviors affect urban energy planning? Distributed stochastic optimization for energy communities | To meet carbon emission reduction goals in line with the Paris agreement, planning resilient and sustainable energy systems has never been more important. In the building sector, particularly, strategic urban energy planning engenders large optimization problems across multiple spatiotemporal scales leading to necessary system scope simplifications. This has resulted in disconnected system scales, namely, building occupants and smart-city energy networks. This paper intends on bridging these disjointed scales to secure both resilient and more energy-efficient urban planning thanks to a holistic approach. The intent is to assess the aggregated impact of user behavior stochasticities on optimal urban energy planning. To this end, a stochastic energy community sizing and operation problem is designed, encompassing multi-level utilities founded on energy hub concepts for improved energy and carbon emission efficiencies. To secure the scalability of our approach, an organic spatial problem distribution suitable for field deployment is put forth, validated by a proof of concept. Uncertainty factors affecting urban energy planning are particularly examined through a local sensitivity analysis, namely, economic, climate, and occupant-behavior uncertainties. Founded on historical measurements a typical Dutch energy community composed of 41 residential buildings is designed. Results disclose a fast-converging distributed stochastic problem, where boilers are showcased as the preferred heating utility, and distributed renewable energy and storage systems were identified as unprofitable for the community. Occupant behavior was particularly exposed as the leading uncertainty factor impacting energy community planning. This demonstrates the relevance and value of our approach in connecting occupants to cities for improved, and more resilient, urban energy planning strategies. | ['Wim Zeiler', 'Henrik Madsen', 'Dominik Franjo Dominkovic', 'Daniela Guericke', 'Amos Schledorn', 'Julien Leprince'] | 2023-03-06 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [-3.84978920e-01 3.10675919e-01 1.12043016e-01 3.33676308e-01
-6.31048799e-01 -5.67310750e-01 6.36108458e-01 1.72236532e-01
-8.36375728e-02 1.00435281e+00 1.99117333e-01 -5.26597619e-01
-7.42828488e-01 -1.36134088e+00 -3.19862276e-01 -1.15314353e+00
-6.23433031e-02 4.67632353e-01 -3.63188863e-01 -2.98244685e-01
-2.35879108e-01 8.17688763e-01 -1.51454294e+00 -4.79202271e-01
1.03565145e+00 5.96350253e-01 4.55618501e-01 2.38816902e-01
3.74206841e-01 3.48937325e-02 -2.01350987e-01 1.51451021e-01
2.92529434e-01 -2.41630942e-01 -5.35326779e-01 -2.14976877e-01
-9.24213350e-01 -1.79592669e-01 3.57260346e-01 8.41177225e-01
8.71446908e-01 4.84440386e-01 6.85219586e-01 -1.59405506e+00
-1.22003384e-01 8.06001902e-01 7.58095533e-02 -3.11010331e-01
-8.73476714e-02 4.08008814e-01 9.98493910e-01 -3.93177509e-01
1.29672766e-01 9.17430043e-01 2.88535953e-01 1.61520004e-01
-1.30916989e+00 -3.95156771e-01 1.20545682e-02 2.15004444e-01
-1.73331809e+00 -1.54637396e-01 7.97569633e-01 -2.87785053e-01
1.41748941e+00 9.19740498e-01 1.38212240e+00 7.00460315e-01
1.54551463e-02 1.92703381e-01 1.11294842e+00 -2.85890490e-01
9.64961112e-01 1.22225843e-01 -3.20512325e-01 -5.93364201e-02
6.88904226e-01 3.98832232e-01 4.47764426e-01 4.01367843e-02
8.06295425e-02 -2.11720362e-01 7.72223175e-02 -1.62717506e-01
-7.99507558e-01 4.77809340e-01 6.72865987e-01 8.37390184e-01
-4.99574691e-01 3.58014107e-01 -1.31723538e-01 -3.03237855e-01
3.68662715e-01 2.85251707e-01 -3.18833411e-01 1.03303753e-01
-1.03651655e+00 6.93692937e-02 7.08718777e-01 8.24305117e-01
6.00517154e-01 5.82149252e-02 -3.16467397e-02 2.36595929e-01
6.91187799e-01 1.20105886e+00 -2.21240178e-01 -9.31106746e-01
2.57969499e-01 6.12936795e-01 3.94885421e-01 -7.33784080e-01
-8.57884109e-01 -5.82696736e-01 -1.06330848e+00 2.93698251e-01
-3.26932408e-02 -3.47886711e-01 -4.02100950e-01 1.74400949e+00
5.81273496e-01 -4.34206843e-01 -2.42218450e-02 5.56305230e-01
-1.01517193e-01 8.72971356e-01 4.37800139e-01 -5.27906120e-01
1.29838467e+00 -2.40608647e-01 -4.89558071e-01 2.67969310e-01
2.51898021e-01 -1.39291033e-01 5.67988217e-01 -2.42374569e-01
-1.24351013e+00 6.32609874e-02 -6.96695685e-01 4.73786354e-01
-1.04792690e+00 -3.26460421e-01 7.92693272e-02 1.17212451e+00
-1.25062633e+00 4.11496133e-01 -1.01903450e+00 -6.60691738e-01
2.54936665e-01 1.79584593e-01 4.45463538e-01 3.13831657e-01
-1.41246164e+00 1.28235960e+00 4.87242699e-01 3.17166209e-01
-7.29146779e-01 -8.52459013e-01 -3.77464086e-01 5.55751026e-01
2.70611849e-02 -9.75510895e-01 8.37550104e-01 -1.50131702e-01
-1.33708680e+00 4.39462997e-03 3.36766779e-01 -4.41574007e-01
7.37690687e-01 6.43504620e-01 -5.91134846e-01 -9.93838087e-02
1.27439454e-01 1.38134569e-01 4.81009819e-02 -1.43815947e+00
-6.63179576e-01 -2.34444290e-01 -2.49810278e-01 2.54183412e-01
-3.44190955e-01 -2.43630514e-01 5.24071336e-01 -2.72425503e-01
-2.77614295e-01 -1.04299235e+00 -5.71512997e-01 -8.41255844e-01
-5.57649732e-01 -3.04502428e-01 5.00724971e-01 -6.35906696e-01
1.48021150e+00 -1.68185079e+00 8.90974328e-02 9.01162863e-01
-3.15457493e-01 -3.04674089e-01 4.56440508e-01 9.36589122e-01
-2.26176977e-01 5.76737761e-01 -5.11153281e-01 -2.81645030e-01
8.51342499e-01 3.46573144e-01 1.69023380e-01 4.40787941e-01
-9.93214920e-02 8.72447550e-01 -9.41201389e-01 -2.04924926e-01
7.58041501e-01 4.89163399e-01 -2.87202805e-01 -1.67993307e-01
-2.77602285e-01 3.13320458e-01 -6.35138333e-01 7.33115017e-01
6.17151797e-01 1.25011683e-01 2.40767255e-01 -5.42907007e-02
-7.45882273e-01 -2.61939615e-01 -1.38357174e+00 1.18074131e+00
-1.06015861e+00 9.10282955e-02 3.59827787e-01 -8.55758846e-01
4.77654964e-01 2.79338747e-01 1.06707799e+00 -8.28002572e-01
-3.10648065e-02 6.25643492e-01 -2.76518315e-01 -2.71969348e-01
3.75933498e-01 -3.30380529e-01 -1.35180309e-01 3.85458976e-01
-4.53360438e-01 -5.89364290e-01 1.07259825e-01 -5.53136179e-03
9.22061026e-01 -1.77882418e-01 2.31159300e-01 -1.21124959e+00
5.30669093e-01 -9.38580334e-02 2.36871064e-01 -1.14056759e-01
-2.52773851e-01 -8.78201351e-02 1.64594769e-01 1.37823641e-01
-1.20785296e+00 -1.28810620e+00 -2.98038930e-01 4.94316012e-01
1.51727095e-01 9.41835567e-02 -9.28167999e-01 -7.73380250e-02
9.06076580e-02 1.79201806e+00 -3.68376076e-01 3.91054153e-03
-5.31409740e-01 -1.06909716e+00 2.57648788e-02 -6.64150051e-04
4.88704443e-01 -3.90595436e-01 -1.09319508e+00 2.87812769e-01
-3.66394579e-01 -6.50795877e-01 9.90783097e-04 2.41458073e-01
-4.46520269e-01 -1.06821358e+00 -6.58384442e-01 -7.26446584e-02
7.03399539e-01 5.32649308e-02 1.06701434e+00 -9.72441733e-02
-2.03500330e-01 6.88850403e-01 1.59824878e-01 -4.06807333e-01
-5.62655747e-01 3.27801228e-01 2.62600154e-01 -1.91971436e-01
-3.34157586e-01 -8.64525199e-01 -1.10310280e+00 5.65808833e-01
-7.45877266e-01 6.89536706e-02 2.90841818e-01 1.51716962e-01
5.11824608e-01 9.23661828e-01 9.27057207e-01 -5.92796467e-02
5.76614022e-01 -9.75222230e-01 -8.82211804e-01 6.12793922e-01
-1.30990779e+00 -3.78909819e-02 7.14668751e-01 3.99925530e-01
-1.40540695e+00 6.54727146e-02 -9.89294648e-02 3.56866270e-01
-2.31605574e-01 1.49047315e-01 -7.88122356e-01 2.51884729e-01
1.39827386e-01 1.95596330e-02 -4.93040651e-01 -2.39710554e-01
5.72707236e-01 2.89621085e-01 7.56793981e-03 -8.74643266e-01
1.28792346e+00 4.10179585e-01 5.26595712e-01 -8.00489008e-01
4.21842754e-01 -3.03801119e-01 -3.09225023e-01 -8.19312096e-01
1.03040957e+00 -8.56311142e-01 -9.51765418e-01 2.62044400e-01
-7.76464462e-01 -4.63535935e-01 -8.95406902e-01 2.07562700e-01
-7.75877059e-01 -1.22876056e-01 3.03714573e-01 -1.54990256e+00
-1.63450360e-01 -1.17441273e+00 4.43570435e-01 2.34854519e-01
3.91918570e-02 -1.00724065e+00 3.73691857e-01 1.60690665e-01
9.83335376e-01 9.15459573e-01 9.15476620e-01 -2.78004091e-02
-9.46931660e-01 2.65765399e-01 7.52228424e-02 1.15672342e-01
2.99604982e-01 2.09787905e-01 -9.47374344e-01 -5.49915254e-01
-1.02743104e-01 3.39373976e-01 2.86276311e-01 2.52364904e-01
7.28925467e-01 -6.33549988e-01 -5.52059710e-01 1.45817772e-01
2.13812184e+00 3.26411098e-01 5.00176370e-01 3.70598793e-01
1.87287271e-01 8.70734692e-01 1.91068992e-01 6.77657962e-01
5.91261506e-01 4.78024691e-01 1.00528646e+00 -1.13913067e-01
7.76245296e-02 -7.30117336e-02 1.97034761e-01 6.35408640e-01
-6.35742247e-01 -3.88090163e-01 -1.09339845e+00 9.36139643e-01
-1.53775537e+00 -1.11811757e+00 -1.27299607e-01 2.33418751e+00
3.45642656e-01 -3.00124317e-01 2.35928938e-01 7.34352544e-02
6.03520155e-01 1.51669994e-01 -4.52848315e-01 -3.54209960e-01
-1.64108366e-01 -2.25675683e-02 1.06914914e+00 4.79803979e-01
-3.16630006e-01 2.51506306e-02 4.96709919e+00 8.82621169e-01
-4.62353826e-01 4.90416795e-01 9.16146576e-01 -5.33339083e-01
-1.10675430e+00 1.43067660e-02 -4.75514293e-01 6.57861054e-01
1.64916587e+00 -7.41570890e-01 8.53761137e-01 4.95222151e-01
1.20243657e+00 -4.21498746e-01 -5.58427751e-01 3.33528280e-01
-8.59497666e-01 -1.22598243e+00 -6.70945942e-01 6.79126859e-01
1.15818942e+00 2.79847533e-01 -1.18602730e-01 7.21756294e-02
7.83443570e-01 -7.73256183e-01 9.07786131e-01 7.69317746e-01
5.02826512e-01 -1.34308422e+00 4.37207192e-01 4.26135063e-01
-1.85169733e+00 -5.08127451e-01 2.24089131e-01 3.07123750e-01
7.11614192e-01 8.42358172e-01 -4.84674096e-01 1.10664082e+00
7.86140203e-01 -1.38836682e-01 -1.42346129e-01 6.44357145e-01
-9.48360190e-02 4.06852901e-01 -1.00934827e+00 -4.09281850e-01
3.19233567e-01 -6.22713029e-01 4.35453624e-01 1.02742434e+00
8.98930490e-01 5.02178848e-01 -2.71851212e-01 1.26728082e+00
3.40076804e-01 2.00919494e-01 -6.45839751e-01 2.48803973e-01
7.18857586e-01 1.37455523e+00 -9.52591836e-01 1.87738374e-01
-3.57974917e-02 3.26745778e-01 -4.55329984e-01 5.89760125e-01
-1.21155083e+00 -1.99669451e-02 8.25727165e-01 3.20818752e-01
5.06695509e-02 -2.12982327e-01 -5.76822817e-01 -7.07739532e-01
-2.26542339e-01 9.60454904e-03 1.21986151e-01 -3.42913508e-01
-9.77555454e-01 -5.12696579e-02 5.68298161e-01 -8.24074566e-01
-2.58087605e-01 -1.45583928e-01 -8.75095785e-01 1.12074149e+00
-1.69048142e+00 -1.35031903e+00 -8.32997561e-02 5.98083854e-01
2.36377627e-01 3.29451829e-01 7.05548942e-01 4.31354821e-01
-1.02889884e+00 -7.31400847e-02 8.45349252e-01 -7.66097844e-01
-2.17640400e-01 -1.39689004e+00 2.07412213e-01 1.08490086e+00
-6.36481881e-01 1.49721146e-01 7.66855061e-01 -6.46826744e-01
-1.33801305e+00 -1.20499682e+00 9.09451663e-01 -6.39481023e-02
8.43819618e-01 -3.90395284e-01 -1.08975843e-01 -4.82284464e-02
7.84270585e-01 -5.43100774e-01 3.67418766e-01 -3.75351161e-01
4.86697465e-01 -2.53940344e-01 -1.79045475e+00 7.48720884e-01
1.06083584e+00 -3.99471223e-01 1.49738744e-01 5.65615475e-01
6.25488162e-01 3.17532897e-01 -1.14983118e+00 1.98849678e-01
2.31869727e-01 -9.12507653e-01 9.47426617e-01 -5.72440475e-02
-2.24074200e-01 -2.79490113e-01 -5.00230253e-01 -1.46615326e+00
-4.29559261e-01 -7.50163317e-01 -1.12419777e-01 1.68164027e+00
4.66626227e-01 -1.04841518e+00 4.65092868e-01 1.31905150e+00
-2.37712979e-01 -6.12547338e-01 -1.76690042e+00 -1.13256681e+00
6.55201614e-01 -4.95728552e-01 1.29140925e+00 7.30297565e-01
2.39986042e-03 -3.09950948e-01 3.38867873e-01 5.98158658e-01
9.98223424e-01 -8.18997622e-02 3.05590201e-02 -9.22858417e-01
1.72842488e-01 -1.00574720e+00 3.74757200e-01 1.25613928e-01
-6.75146058e-02 -8.22256923e-01 1.88566484e-02 -2.02065921e+00
-2.34915242e-01 -5.17058253e-01 -4.09226239e-01 2.96790242e-01
4.54928786e-01 -3.39850366e-01 3.73380661e-01 -8.01425725e-02
1.89984426e-01 9.78433430e-01 7.21320152e-01 -3.53455663e-01
-1.38722762e-01 4.83017355e-01 -3.98506939e-01 3.89591157e-01
1.00394118e+00 -2.03012347e-01 -7.02042520e-01 2.29589686e-01
5.25003731e-01 -1.97479706e-02 5.45485020e-01 -1.38607299e+00
3.44828486e-01 -6.39879465e-01 -2.96143860e-01 -7.58883178e-01
-8.78470242e-02 -1.85106015e+00 1.34680176e+00 9.90008414e-01
3.51537138e-01 -8.44241381e-02 4.54775952e-02 4.71542269e-01
3.57510835e-01 2.10649177e-01 8.77518177e-01 -7.45769590e-02
-4.45632070e-01 8.70995820e-02 -5.70213497e-01 -4.01880771e-01
1.81553125e+00 -1.64587721e-01 -3.24860662e-01 4.22031544e-02
-9.23623562e-01 7.69769430e-01 6.06636167e-01 -6.14396706e-02
-6.04258059e-03 -1.38616288e+00 -6.17438197e-01 -2.67449647e-01
-3.67004067e-01 -1.57052770e-01 5.57744324e-01 6.70008302e-01
-2.64232785e-01 6.76990926e-01 1.37454811e-02 -1.85826272e-01
-6.67889833e-01 6.67652190e-01 8.87684286e-01 -3.27950805e-01
-9.99498814e-02 2.79948145e-01 -3.43517005e-01 -3.92365843e-01
-2.45683983e-01 -7.90100217e-01 3.02494287e-01 6.42836750e-01
-1.44927859e-01 1.28142822e+00 2.14805245e-01 -5.23374379e-01
-5.82774639e-01 5.10907590e-01 1.34135437e+00 -2.47977868e-01
1.42941439e+00 -7.61026621e-01 -2.32781947e-01 3.30651164e-01
1.02716160e+00 -8.51896331e-02 -1.08216560e+00 4.33037877e-01
-2.16686763e-02 9.95523632e-02 1.78807303e-01 -1.22941136e+00
-1.04655123e+00 2.68208891e-01 8.66259694e-01 9.81332779e-01
1.54562223e+00 -2.13612527e-01 5.58456838e-01 2.27633655e-01
6.88752532e-01 -1.72100484e+00 -8.34752500e-01 -1.16033919e-01
1.03672969e+00 -9.52059984e-01 -1.34624951e-02 -7.24570220e-03
9.89965573e-02 8.23237181e-01 5.45387380e-02 3.57680172e-01
1.06469083e+00 1.96377978e-01 -7.58418560e-01 -1.51587725e-02
-2.36932129e-01 -6.59472525e-01 -2.98780888e-01 3.40979517e-01
-2.44327694e-01 8.24237168e-01 -2.95652986e-01 4.91128176e-01
2.04678271e-02 -2.97579139e-01 1.27299294e-01 6.00045860e-01
-4.24267977e-01 -9.31199372e-01 -6.39314651e-01 1.20082544e-02
2.30808526e-01 -5.34365885e-02 1.18940160e-01 1.07782316e+00
4.96831447e-01 1.11000454e+00 -5.46237081e-02 -5.73064480e-03
7.50811756e-01 3.58035229e-02 -2.43924201e-01 -2.89429650e-02
-7.32972562e-01 -1.58258826e-01 1.91604659e-01 -5.50773740e-01
-4.88610864e-01 -1.01957893e+00 -1.38769686e+00 -8.50446284e-01
-3.12186122e-01 3.97545516e-01 1.25010717e+00 8.78732443e-01
2.46426448e-01 7.99816549e-01 1.29458582e+00 -9.52549338e-01
-5.90435386e-01 -5.21313310e-01 -9.45900917e-01 -2.99474984e-01
1.04969216e-03 -3.65054250e-01 -6.84671283e-01 -6.35001957e-01] | [5.6946234703063965, 2.444659471511841] |
890cf09b-1e4a-4660-8e3d-627be66e556b | efficient-text-based-reinforcement-learning-1 | null | null | https://openreview.net/forum?id=0McAgVlYSy | https://openreview.net/pdf?id=0McAgVlYSy | Efficient Text-based Reinforcement Learning by Jointly Leveraging State and Commonsense Graph Representations | Text-based games (TBGs) have emerged as useful benchmarks for evaluating progress at the intersection of grounded language understanding and reinforcement learning (RL). Recent work has proposed the use of external knowledge -- like commonsense knowledge -- to improve the efficiency of RL agents for TBGs; and to resemble human-like decision-making skills.
In this paper, we posit that to act efficiently in TBGs, an agent must be able to track the state of the game while retrieving and using relevant commonsense knowledge. Thus, we propose an agent for TBGs that induces a graph representation of the game state and jointly grounds it with a graph of commonsense knowledge from ConceptNet. This combination is achieved through bidirectional knowledge graph attention between the two symbolic representations. We show that agents that incorporate commonsense into the game state graph outperform baseline agents. | ['Anonymous'] | 2021-02-22 | null | null | null | null | ['text-based-games'] | ['playing-games'] | [ 2.39704445e-01 6.49053276e-01 -3.60926762e-02 7.43191019e-02
-4.86617744e-01 -5.41399300e-01 7.82302618e-01 3.96796077e-01
-5.60381055e-01 7.69443452e-01 6.34310782e-01 -4.75369543e-01
-1.48683444e-01 -1.43417037e+00 -6.58548117e-01 -1.18994728e-01
1.63625367e-02 7.95168936e-01 2.52649486e-01 -1.00777233e+00
4.43764746e-01 5.52287586e-02 -1.19576657e+00 1.70231313e-01
1.07835698e+00 3.03517222e-01 2.83429503e-01 6.82034612e-01
-1.56268135e-01 2.05719781e+00 -6.58586025e-01 -5.42868137e-01
1.59427971e-02 -9.19728220e-01 -1.48653126e+00 -3.97492081e-01
1.93259446e-03 -4.32373077e-01 -5.89641750e-01 1.20514464e+00
4.07324098e-02 8.24350297e-01 5.92515349e-01 -1.29153705e+00
-8.84039283e-01 1.26065481e+00 9.14871097e-02 3.05807292e-01
6.61246598e-01 4.95414764e-01 1.41232812e+00 1.12066925e-01
9.11635935e-01 1.42559588e+00 3.71973634e-01 9.60937083e-01
-1.06683636e+00 -5.82977951e-01 2.95083433e-01 4.83269900e-01
-7.88035274e-01 -2.02471346e-01 8.93485963e-01 -4.37120408e-01
1.51819003e+00 -2.69583881e-01 1.21377897e+00 9.15978849e-01
-1.26138246e-02 7.38384783e-01 1.29508257e+00 -6.23831332e-01
5.07545888e-01 -3.80723298e-01 1.25480741e-01 1.24450254e+00
3.07147682e-01 5.32250285e-01 -8.93286288e-01 6.99191063e-04
1.11375070e+00 -3.81246418e-01 3.60452011e-02 -4.59134519e-01
-1.07159579e+00 1.06363451e+00 7.26512253e-01 4.76435423e-01
-5.25236011e-01 9.21855211e-01 5.32455444e-01 4.28929508e-01
9.02303532e-02 1.29461753e+00 1.85866486e-02 -3.65889430e-01
-5.60417354e-01 4.27587092e-01 8.43206763e-01 7.14791536e-01
6.30908430e-01 1.83441043e-01 -3.11288625e-01 3.28217953e-01
2.82940567e-01 4.71188247e-01 6.73880577e-01 -1.12699485e+00
4.86165911e-01 8.06874514e-01 8.54725093e-02 -1.02371228e+00
-5.08371353e-01 -1.56767532e-01 -1.11409843e-01 1.07874408e-01
3.00234288e-01 -2.29016300e-02 -8.42172027e-01 2.21236920e+00
1.17278630e-02 4.83660519e-01 4.61904079e-01 6.92847252e-01
7.77089775e-01 3.72013927e-01 5.39678931e-01 1.20850772e-01
1.11361289e+00 -6.36058331e-01 -5.12981713e-01 -6.70485079e-01
1.03667176e+00 3.51719201e-01 1.20110369e+00 1.53171778e-01
-1.12993860e+00 -1.48240760e-01 -1.10455132e+00 -2.61049628e-01
-4.97340053e-01 -6.37945056e-01 9.84568477e-01 3.88743848e-01
-1.40498042e+00 7.22666621e-01 -6.10738039e-01 -4.69123632e-01
5.98695815e-01 5.35856448e-02 -1.32369727e-01 -2.76973844e-01
-1.73427117e+00 1.55047691e+00 1.14892101e+00 -2.32655436e-01
-1.32521451e+00 -2.24251881e-01 -1.31514251e+00 1.32713001e-02
6.89269602e-01 -9.84941959e-01 1.45228803e+00 -8.04453731e-01
-1.71092319e+00 1.17243230e+00 3.48321259e-01 -6.68456674e-01
1.77531064e-01 -1.44098088e-01 -3.88967656e-02 2.15843976e-01
2.78323829e-01 5.39964736e-01 5.37523150e-01 -1.05728841e+00
-5.03847539e-01 -2.78885186e-01 8.84515584e-01 6.18148446e-01
2.27171645e-01 -1.41397730e-01 -1.38522580e-01 -4.66418803e-01
-7.82595575e-02 -8.77631187e-01 -2.38139704e-01 -6.32173657e-01
-3.94446611e-01 -4.27994400e-01 -3.75399292e-02 -6.00203872e-01
9.09440756e-01 -1.73301601e+00 3.90234202e-01 3.79515439e-01
6.17740750e-01 1.02174759e-01 -3.28251749e-01 3.46373945e-01
8.77255574e-02 5.30735888e-02 5.43234348e-02 1.06114581e-01
3.32164198e-01 3.81845921e-01 -3.74310106e-01 1.07182205e-01
8.77782255e-02 1.62246895e+00 -1.68793333e+00 -2.75449246e-01
2.35923767e-01 -1.76993728e-01 -7.90395379e-01 1.24474630e-01
-8.68424058e-01 -2.84755160e-03 -6.70981765e-01 4.94291913e-03
-1.45546317e-01 -3.03037286e-01 5.22967219e-01 3.20696950e-01
4.44467336e-01 8.07818115e-01 -8.03087473e-01 1.84432471e+00
-4.65439916e-01 5.35274327e-01 -5.66044509e-01 -8.96143258e-01
5.79903245e-01 1.47881910e-01 -1.09691277e-01 -9.19563591e-01
3.40244442e-01 -2.14121595e-01 2.09048629e-01 -2.48537526e-01
5.07062793e-01 -7.94573307e-01 -3.27373832e-01 7.73727953e-01
2.89801478e-01 -8.34461570e-01 2.45722413e-01 8.11434805e-01
1.26600850e+00 4.45399582e-01 6.21116042e-01 -8.93511996e-02
1.02796175e-01 5.16925812e-01 9.96013656e-02 1.11869740e+00
-2.21240178e-01 -4.35353100e-01 7.20120370e-01 -1.06340334e-01
-6.66297317e-01 -1.01434970e+00 9.99108553e-01 1.43192005e+00
1.60422191e-01 -5.92353523e-01 -6.33797109e-01 -5.03793478e-01
-2.22753048e-01 1.21335018e+00 -8.22118580e-01 -8.59305382e-01
-5.40393531e-01 -7.95107987e-03 9.59675372e-01 6.98867619e-01
7.53815114e-01 -1.60126162e+00 -9.75459754e-01 4.53829765e-01
-4.68702525e-01 -1.10806084e+00 -1.68636665e-02 2.68331528e-01
-7.66087115e-01 -1.41833377e+00 3.70987356e-02 -6.38913691e-01
2.05428615e-01 6.50181919e-02 1.59642756e+00 4.82723981e-01
-1.73497591e-02 8.58202815e-01 -5.39921284e-01 -4.02509898e-01
-7.37467587e-01 -2.45805040e-01 -1.25553995e-01 -8.35201144e-01
3.57205808e-01 -5.48843563e-01 -1.26079455e-01 -3.12719345e-01
-8.08014691e-01 4.41788703e-01 7.73035213e-02 7.24405169e-01
1.45842478e-01 1.70423388e-01 4.28398103e-01 -8.67302418e-01
1.28880584e+00 -4.45030242e-01 -6.08474731e-01 2.17229098e-01
-3.33672136e-01 4.59972173e-01 4.59488183e-01 -7.31965527e-02
-1.01931012e+00 -5.30182540e-01 2.73615330e-01 -1.16200075e-01
1.11421920e-01 9.01757658e-01 -2.26937793e-02 7.51081407e-02
1.07600582e+00 5.21914303e-01 -1.15482017e-01 2.26729453e-01
1.04544961e+00 -6.35131970e-02 6.30082309e-01 -1.29588401e+00
7.72455215e-01 1.17527701e-01 -8.55571777e-02 -3.72075886e-01
-1.20913029e+00 -4.27539974e-01 -3.02555531e-01 -1.28638059e-01
9.38280225e-01 -9.30230021e-01 -9.31608260e-01 2.97972828e-01
-1.04337204e+00 -1.14528477e+00 -5.81552148e-01 2.41981953e-01
-1.24360454e+00 2.85756458e-02 -7.33129263e-01 -5.96107900e-01
-2.15188786e-02 -7.41246521e-01 5.85519493e-01 1.86600357e-01
-3.51276219e-01 -1.15607011e+00 3.65433633e-01 3.77206981e-01
3.17441940e-01 2.89953679e-01 1.16504228e+00 -7.41608441e-01
-4.48565125e-01 3.22506636e-01 -2.50278175e-01 -7.30322376e-02
-4.96639963e-03 -4.90063041e-01 -8.79747152e-01 1.01879835e-01
-2.35818297e-01 -9.63984370e-01 9.59352016e-01 2.81762034e-01
6.77051485e-01 -2.72344112e-01 -1.82512298e-01 2.16360435e-01
1.39330590e+00 1.31292999e-01 8.92790675e-01 6.93561435e-01
5.75483739e-01 4.70293283e-01 4.67489243e-01 4.12846625e-01
8.57900858e-01 4.15011317e-01 3.47136527e-01 3.79943550e-01
-1.42104253e-01 -7.89853811e-01 4.92803037e-01 3.34974349e-01
-4.95861560e-01 -1.14119165e-01 -1.19902313e+00 6.21990860e-01
-1.92539001e+00 -1.44926071e+00 4.23968792e-01 1.70757985e+00
1.28982127e+00 2.48378649e-01 1.47488281e-01 -9.65214744e-02
4.68626916e-01 3.07659060e-01 -6.03000820e-01 -4.34102476e-01
-1.56296715e-02 6.20156288e-01 1.17416292e-01 8.87314439e-01
-6.84040546e-01 1.93710101e+00 6.30005598e+00 6.45907223e-01
-5.28692842e-01 7.33904466e-02 3.30989696e-02 1.58827260e-01
-5.13235927e-01 4.55345474e-02 -2.45657310e-01 -1.17605910e-01
8.70112896e-01 -4.33596253e-01 8.88165414e-01 6.06754899e-01
-1.36500731e-01 -1.96135625e-01 -1.14256704e+00 7.99205244e-01
7.62949958e-02 -1.47028565e+00 2.96086043e-01 -1.68595046e-01
6.41419947e-01 1.25210032e-01 -1.96391523e-01 1.00657642e+00
1.62058794e+00 -1.27014387e+00 8.77342343e-01 4.31191027e-01
6.20205879e-01 -6.78872287e-01 4.48521048e-01 3.90437514e-01
-1.01180899e+00 -1.00556158e-01 -1.92463249e-01 -4.65614438e-01
-2.31308341e-01 -5.54676652e-02 -1.04297173e+00 4.25086766e-01
-2.44381670e-02 5.95648348e-01 -5.44868529e-01 5.39824128e-01
-1.02720177e+00 5.68776369e-01 -1.08542452e-02 -4.07947570e-01
5.37946999e-01 -3.64877060e-02 3.42580438e-01 9.22486961e-01
-1.72297090e-01 7.23162115e-01 3.09600621e-01 1.32823956e+00
-2.19041213e-01 -1.70605391e-01 -8.62115681e-01 -6.37158036e-01
4.72625643e-01 5.45095623e-01 -6.48496807e-01 -5.85730016e-01
7.84317181e-02 9.66796100e-01 8.44766915e-01 2.52916604e-01
-6.73216879e-01 -1.67807147e-01 6.51116252e-01 -1.79386228e-01
-5.74313588e-02 -3.92228782e-01 1.17997900e-02 -1.18793917e+00
-5.27184725e-01 -1.01672661e+00 5.63364685e-01 -1.32811987e+00
-8.21394086e-01 3.12795669e-01 3.38189043e-02 -4.05420840e-01
-5.79145908e-01 -7.50879109e-01 -5.98359764e-01 6.14906549e-01
-1.47146094e+00 -1.13862526e+00 -2.69428551e-01 1.01923335e+00
2.01834813e-01 -9.53603387e-02 9.98090148e-01 -8.29938650e-01
-8.95244479e-02 5.51156439e-02 -4.51349348e-01 5.20441234e-01
4.07351442e-02 -1.47670364e+00 7.10554481e-01 7.73519158e-01
4.21680927e-01 8.02812159e-01 7.92549372e-01 -9.69830811e-01
-1.21150291e+00 -8.40101361e-01 4.60347950e-01 -8.26268435e-01
1.00857830e+00 -7.31620658e-03 -6.85508966e-01 1.05613828e+00
9.01458114e-02 -2.68714279e-01 6.39345825e-01 3.95758659e-01
-8.90572071e-01 6.05069280e-01 -9.96360123e-01 8.59785497e-01
1.52279258e+00 -1.07345212e+00 -1.56255102e+00 2.31757253e-01
8.45730722e-01 -4.98638183e-01 -5.58297157e-01 -4.30721968e-01
1.89085666e-03 -6.30084276e-01 8.46737087e-01 -1.33897948e+00
6.75038636e-01 -3.90365392e-01 -1.53400913e-01 -1.95088851e+00
-5.57506442e-01 -5.63948154e-01 -1.40469238e-01 5.29406250e-01
1.22049488e-01 -5.10116875e-01 5.92022836e-01 6.51066005e-01
7.37757310e-02 -4.16035652e-02 -7.75184870e-01 -7.56138384e-01
3.34355801e-01 -7.33554065e-01 5.12454987e-01 1.15486062e+00
8.81984770e-01 6.94189906e-01 1.08907066e-01 -1.12210914e-01
5.13385296e-01 5.79631776e-02 4.97959167e-01 -1.32750106e+00
-4.04117465e-01 -7.23482072e-01 -4.47630137e-01 -7.02799857e-01
7.13536501e-01 -1.32051623e+00 2.02163145e-01 -2.09469867e+00
3.24718773e-01 -2.72067964e-01 -3.62475395e-01 9.30078685e-01
-2.97151029e-01 -2.33432040e-01 5.30085504e-01 -2.67952055e-01
-1.01105547e+00 3.68153244e-01 1.45029676e+00 -2.06303194e-01
-1.43450469e-01 -7.12443173e-01 -1.21677685e+00 7.92684793e-01
9.79993522e-01 -3.78960460e-01 -7.20308065e-01 -1.89658478e-01
1.00678015e+00 1.45505041e-01 7.28182614e-01 -9.04692709e-01
4.28334922e-01 -6.64124668e-01 -3.89105827e-02 1.08486064e-01
2.07124412e-01 -4.00111943e-01 -3.84790480e-01 5.43523550e-01
-6.61387026e-01 -8.00131038e-02 4.39091682e-01 6.72890306e-01
-5.71428845e-03 -2.22949624e-01 5.85185647e-01 -7.59184182e-01
-1.26369798e+00 -7.85965174e-02 -4.31263000e-01 8.26177776e-01
7.42704093e-01 -2.25231320e-01 -6.96623683e-01 -7.08050847e-01
-7.81005919e-01 2.66637176e-01 4.59660083e-01 1.21382251e-01
7.74609149e-01 -1.19094038e+00 -6.49762392e-01 -4.00356531e-01
2.34840721e-01 -1.77928060e-01 -1.52811073e-02 3.15176576e-01
-5.85865378e-01 3.50722790e-01 -4.53851938e-01 3.19787771e-01
-7.59254158e-01 5.48329413e-01 8.30689728e-01 -6.68132067e-01
-6.30545676e-01 1.04274392e+00 1.18135132e-01 -5.86167932e-01
-9.72033963e-02 -4.96050298e-01 -4.04358715e-01 -1.91186428e-01
5.62434375e-01 1.47303358e-01 -1.79482937e-01 -3.54418516e-01
-2.56943077e-01 1.86073110e-01 6.40362278e-02 -5.04813135e-01
1.26255774e+00 2.62879074e-01 -2.17461884e-01 1.49468720e-01
3.77625674e-01 -1.90166295e-01 -9.19362009e-01 -4.54801232e-01
1.91249952e-01 -2.18636990e-01 5.54533936e-02 -1.15157866e+00
-6.10286057e-01 5.73000014e-01 -2.54998356e-01 1.71642676e-01
7.58173883e-01 2.39246383e-01 3.19622815e-01 9.35272038e-01
8.44667554e-01 -1.19913375e+00 5.59787869e-01 1.16755950e+00
9.30998802e-01 -1.02913690e+00 -7.73758665e-02 2.76979394e-02
-8.68528128e-01 8.94082963e-01 7.36834347e-01 -3.83033663e-01
2.12011556e-03 -6.66033551e-02 -3.47824007e-01 -6.98331833e-01
-8.01281691e-01 -8.61847639e-01 9.76450592e-02 1.10621417e+00
1.20256834e-01 4.27776217e-01 3.08831543e-01 7.30868220e-01
-7.04403818e-01 3.11184168e-01 7.26708710e-01 9.35476661e-01
-7.92325199e-01 -6.42410100e-01 -2.31761158e-01 4.24747854e-01
-3.42653617e-02 -5.99249423e-01 -8.20332289e-01 7.70795941e-01
-4.86124493e-02 1.08097029e+00 -2.88841963e-01 -3.55641872e-01
3.18140626e-01 2.77575999e-01 1.12729073e+00 -1.01571667e+00
-6.50587797e-01 -7.13471472e-01 4.90511984e-01 -1.01533914e+00
-3.86611849e-01 -3.27238619e-01 -1.87991953e+00 -5.60091496e-01
-1.51538998e-01 1.83882669e-01 4.91485745e-02 1.35307980e+00
-1.09700009e-01 7.43556976e-01 -2.15713724e-01 -3.64579320e-01
-3.81503910e-01 -7.49002814e-01 -6.26880229e-01 5.88355780e-01
1.09849676e-01 -8.48130226e-01 -7.84936100e-02 -6.43519834e-02] | [3.8686068058013916, 1.2872322797775269] |
08f71be9-a86d-4359-9ef2-811b5080f8b6 | a-generalized-template-based-graph-neural | null | null | https://www.nature.com/articles/s42256-022-00526-z | https://www.nature.com/articles/s42256-022-00526-z | A generalized-template-based graph neural network for accurate organic reactivity prediction | The reliable prediction of chemical reactivity remains in the realm of knowledgeable synthetic chemists. Automating this process by using artificial intelligence could accelerate synthesis design in future digital laboratories. While several machine learning approaches have demonstrated promising results, most current models deviate from how human chemists analyse and predict reactions based on electronic changes. Here, we propose a chemistry-motivated graph neural network called LocalTransform, which learns organic reactivity based on generalized reaction templates to describe the net changes in electron configuration between the reactants and products. The proposed concept dramatically reduces the number of reaction rules and exhibits state-of-the-art product prediction accuracy. In addition to the built-in interpretability of the generalized reaction templates, the high score–accuracy correlation of the model allows users to assess the uncertainty of the machine predictions. | ['Yousung Jung', 'Shuan Chen'] | 2022-09-15 | null | null | null | nature-machine-intelligence-2022-9 | ['chemical-reaction-prediction'] | ['medical'] | [ 4.18777525e-01 4.14932907e-01 -2.88504332e-01 -3.51515740e-01
-1.85083374e-01 -8.45513642e-01 7.32700944e-01 7.35582352e-01
-7.64428303e-02 1.05897820e+00 -1.02977879e-01 -7.76656926e-01
-1.99280959e-02 -1.05415368e+00 -7.28499949e-01 -8.04025233e-01
1.83631182e-01 4.78191137e-01 2.02374071e-01 -4.33999896e-01
4.38806593e-01 9.34588194e-01 -1.17148614e+00 4.01072234e-01
1.12839127e+00 1.04007709e+00 1.55589506e-02 3.93533200e-01
-1.90323010e-01 8.69927704e-01 -2.15769738e-01 -4.92226720e-01
1.38948977e-01 -3.91230136e-01 -6.59771383e-01 -8.06615591e-01
-6.61507100e-02 4.20257747e-01 -1.16563052e-01 9.77585375e-01
3.62988830e-01 3.30866456e-01 1.03359699e+00 -6.69915915e-01
-7.76232421e-01 7.62973607e-01 2.09682658e-01 -1.66879818e-01
5.69757462e-01 5.18711209e-01 1.18923104e+00 -9.40308630e-01
8.92981648e-01 9.01586652e-01 4.89923924e-01 2.73786485e-01
-1.37916088e+00 -5.76178253e-01 1.16999470e-01 5.28109729e-01
-1.10458112e+00 -1.80512071e-01 8.12833786e-01 -5.56030154e-01
1.37954378e+00 1.12794675e-01 8.06049943e-01 9.34083343e-01
5.69014966e-01 -3.71253416e-02 7.38510609e-01 -2.78096199e-01
7.24541426e-01 1.13881968e-01 -1.10735245e-01 8.65876079e-01
6.02929473e-01 3.83956969e-01 -4.94749337e-01 -2.66899709e-02
3.95741314e-01 3.83808985e-02 -1.23094261e-01 -4.67591792e-01
-1.07260299e+00 9.28642154e-01 6.58877313e-01 2.43101478e-01
-5.22926927e-01 -4.98867445e-02 3.00142556e-01 -3.28939408e-02
3.82323325e-01 1.35498476e+00 -8.10480773e-01 2.30314746e-01
-6.73712254e-01 -5.63256294e-02 1.04935455e+00 6.00354314e-01
9.22779799e-01 3.24955225e-01 1.61738753e-01 9.17423069e-02
2.88614571e-01 -1.18184663e-01 1.56326711e-01 -5.04327178e-01
1.26479372e-01 9.67356086e-01 6.46363944e-02 -8.40119183e-01
-7.67533362e-01 -4.30187702e-01 -7.01694727e-01 5.35599232e-01
3.60521048e-01 -2.65043288e-01 -8.81915510e-01 1.35918140e+00
3.79525423e-01 -5.38943231e-01 1.93783283e-01 3.60897243e-01
7.52279937e-01 9.73684430e-01 4.30152893e-01 -4.51349139e-01
6.99897647e-01 -8.32973123e-01 -5.38560331e-01 2.50553608e-01
9.04838264e-01 -5.71701407e-01 4.61618125e-01 7.90530562e-01
-9.32062805e-01 -4.52062488e-01 -1.44715810e+00 -1.83461264e-01
-9.34567511e-01 -2.27467284e-01 1.18618751e+00 7.38193870e-01
-5.68694532e-01 1.49429047e+00 -6.33418620e-01 -5.41085191e-02
1.14998490e-01 8.26543331e-01 -5.27638435e-01 2.08459899e-01
-1.19193828e+00 1.43560529e+00 9.43123877e-01 1.73673213e-01
-7.59846866e-01 -1.04992056e+00 -6.25070333e-01 1.70416459e-01
4.64916855e-01 -6.34499371e-01 1.00631416e+00 -7.46936619e-01
-1.86125124e+00 1.96928784e-01 5.09586111e-02 -3.27958405e-01
3.94204468e-01 4.24473286e-01 -6.29865885e-01 -1.10750072e-01
-3.41857582e-01 6.72270238e-01 3.72084498e-01 -1.07065201e+00
-2.86617279e-01 -2.50381082e-01 -2.95683682e-01 -2.25886077e-01
1.97680384e-01 -4.85808432e-01 3.69357377e-01 -1.70109332e-01
1.83274895e-01 -7.52056539e-01 -6.40783608e-01 4.18365188e-02
-4.45980072e-01 -3.65615577e-01 3.28439176e-01 -2.75807619e-01
9.44038510e-01 -1.52336037e+00 1.97461084e-01 7.29788601e-01
2.09505007e-01 2.52970248e-01 1.23816477e-02 8.79316628e-01
-7.81603992e-01 2.14027360e-01 1.90699546e-04 4.75254685e-01
-1.58831015e-01 -1.94093719e-01 -6.26693889e-02 2.83273727e-01
2.47825712e-01 8.48993838e-01 -9.56955314e-01 2.02138290e-01
3.67430270e-01 5.40330529e-01 -5.87390423e-01 1.51664287e-01
-7.96408176e-01 5.00885189e-01 -6.61315322e-01 5.46581089e-01
3.26121509e-01 -3.08288813e-01 7.81605601e-01 -4.38387185e-01
-4.91130412e-01 6.93981230e-01 -8.05591524e-01 1.51186383e+00
-1.85744852e-01 1.20983422e-02 -7.21233666e-01 -9.69884634e-01
1.28481412e+00 4.53047514e-01 4.19336706e-01 -6.71893001e-01
-4.73074988e-03 2.08330065e-01 3.18712294e-01 -2.88756222e-01
3.72912645e-01 -3.32345128e-01 2.22991139e-01 2.05396771e-01
1.88332915e-01 -4.27539587e-01 3.85962218e-01 9.24607925e-03
6.42825663e-01 4.75416481e-01 8.82532001e-01 -2.70450890e-01
6.75758243e-01 9.76645947e-02 4.25512910e-01 6.02137268e-01
1.91339135e-01 1.95210636e-01 6.94890082e-01 -1.07250547e+00
-1.21986151e+00 -7.02468932e-01 1.56108931e-01 9.78330910e-01
-2.10058019e-01 -6.44125819e-01 -5.67705750e-01 -6.35172784e-01
-1.09961718e-01 9.41388845e-01 -5.85087895e-01 -5.31162798e-01
-1.49861380e-01 -6.91745043e-01 8.85361806e-02 2.07500130e-01
-2.47158334e-01 -1.05856299e+00 2.26043746e-01 6.62028730e-01
3.83860826e-01 -7.03782380e-01 -1.83547437e-01 6.59474432e-01
-6.51723623e-01 -1.19873965e+00 9.87115502e-02 -3.69786888e-01
5.13254881e-01 -4.23679240e-02 7.87009835e-01 -1.96971536e-01
-4.48956043e-01 -3.79587620e-01 -1.34588942e-01 -6.71473086e-01
-1.00309062e+00 1.85295880e-01 2.53304560e-02 -2.97107190e-01
2.03969311e-02 -8.59691381e-01 -7.55281448e-01 -1.09658111e-02
-7.11703658e-01 5.53302057e-02 5.56712747e-01 5.60619891e-01
6.66299880e-01 1.13595068e-01 6.37192070e-01 -9.54912066e-01
4.18457597e-01 -4.51331466e-01 -7.94006169e-01 4.79666770e-01
-1.32647860e+00 8.64885092e-01 1.10678101e+00 -3.35010201e-01
-1.15011120e+00 5.38553476e-01 -1.79772183e-01 3.03733766e-01
-7.52229318e-02 7.49481559e-01 -2.55183995e-01 -3.17124993e-01
8.83099794e-01 -1.80254728e-02 -2.56356120e-01 -2.02576160e-01
5.69461107e-01 3.84940468e-02 2.45721564e-01 -7.42554545e-01
5.53335130e-01 -7.46151805e-02 6.94356024e-01 -5.43903768e-01
-6.14683568e-01 -6.88132867e-02 -5.82164526e-01 -1.12302147e-01
9.10761535e-01 -7.17181981e-01 -1.38304019e+00 -2.23843142e-01
-1.22653806e+00 -2.32957043e-02 -1.93723649e-01 4.01588589e-01
-4.77022856e-01 6.13217115e-01 -2.45360404e-01 -6.38963521e-01
-6.24467254e-01 -1.44749928e+00 3.83199155e-01 7.86294788e-02
-4.26598161e-01 -1.00959980e+00 8.15491676e-02 2.75016397e-01
9.39141363e-02 5.68810642e-01 1.41837704e+00 -8.92107189e-01
-8.29754889e-01 -3.69882375e-01 1.41255513e-01 7.44381873e-03
2.83712000e-01 3.17704171e-01 -7.40017951e-01 6.09220639e-02
-5.32854140e-01 -9.74447373e-03 8.65571976e-01 2.53526896e-01
1.08137012e+00 -2.74628758e-01 -3.43816042e-01 2.01862425e-01
1.26036549e+00 7.96768010e-01 5.46408296e-01 1.73847198e-01
7.77304292e-01 5.84793746e-01 1.27821445e-01 4.89384502e-01
-4.54330631e-02 4.08227861e-01 6.00992203e-01 7.66236782e-02
1.97584182e-01 -6.34154141e-01 3.99627298e-01 4.69173491e-01
-7.19044507e-01 -4.40482140e-01 -7.62085974e-01 -2.83746749e-01
-1.76105380e+00 -9.92071152e-01 -2.83559114e-01 2.20743752e+00
8.31432462e-01 3.12776506e-01 7.23120868e-02 1.58269361e-01
5.68534136e-01 -8.80191382e-03 -6.50621116e-01 -9.04045761e-01
3.11319940e-02 5.70569634e-01 5.89463413e-01 4.96106982e-01
-5.36965370e-01 1.12118220e+00 7.26287174e+00 6.35307610e-01
-1.28109717e+00 -3.55746597e-01 5.78143537e-01 1.82498246e-01
-3.96848321e-01 2.99473912e-01 -6.68872416e-01 1.64200857e-01
1.10502660e+00 -2.65703589e-01 6.60046220e-01 9.18252826e-01
4.47666973e-01 2.82780915e-01 -1.53540659e+00 4.12128508e-01
-4.11109179e-01 -2.07857847e+00 5.40293515e-01 -3.30488905e-02
7.70030260e-01 -3.58819008e-01 -1.18576484e-02 -4.94893380e-02
4.09824520e-01 -1.08812177e+00 7.01191306e-01 6.85322940e-01
5.56774676e-01 -9.83420968e-01 2.27391154e-01 -3.59996082e-03
-8.68729949e-01 -5.61491549e-02 -1.74099162e-01 -1.69981644e-01
-2.22039074e-01 6.42928839e-01 -1.36181355e+00 7.46117949e-01
-9.91582274e-02 6.05058849e-01 -3.81052405e-01 5.80055356e-01
-4.94630218e-01 2.51524478e-01 -1.90467760e-01 -5.38889349e-01
2.07152843e-01 -7.17604637e-01 6.54888377e-02 1.06481826e+00
2.97953099e-01 5.46568930e-02 -7.19111040e-02 1.05868363e+00
-2.12561250e-01 2.32047275e-01 -5.79585671e-01 -5.64341307e-01
3.00043315e-01 1.08788252e+00 -6.63776577e-01 -2.21061781e-01
-2.66463488e-01 5.27153611e-01 3.17616701e-01 3.41186255e-01
-4.77301151e-01 -2.81019598e-01 6.20027065e-01 -6.18169084e-02
3.14039946e-01 -2.77860701e-01 -2.08680794e-01 -6.94107831e-01
-1.27058029e-01 -6.84750855e-01 1.65171474e-01 -7.45352864e-01
-1.00310528e+00 2.81262338e-01 -4.32105601e-01 -7.05640137e-01
-1.40338197e-01 -1.02645779e+00 -6.76985264e-01 5.93388617e-01
-1.31383586e+00 -8.17274868e-01 2.34832570e-01 1.02636494e-01
2.65907168e-01 -1.95462584e-01 1.09648848e+00 -1.99126333e-01
-6.64238214e-01 3.20821494e-01 3.18765908e-01 -5.56196332e-01
6.61970556e-01 -1.28896761e+00 3.96068066e-01 4.57141966e-01
-1.40380830e-01 7.03410745e-01 1.03266382e+00 -6.80539906e-01
-1.56341970e+00 -1.03917646e+00 8.16942751e-01 -2.93043107e-01
7.57415891e-01 -2.58336663e-01 -6.72837973e-01 2.55380839e-01
2.61413082e-02 -6.90077990e-03 7.91754246e-01 1.80725217e-01
-7.61395097e-01 -1.34487569e-01 -1.04361260e+00 6.53356254e-01
9.43846881e-01 -4.06687170e-01 -2.52684474e-01 6.24347866e-01
7.56020725e-01 -5.12799583e-02 -1.33886373e+00 1.71511799e-01
6.07675552e-01 -7.32512653e-01 1.02185547e+00 -1.11056304e+00
5.50218940e-01 -3.98634821e-01 2.00675398e-01 -1.31295705e+00
-5.89694798e-01 -8.69915783e-01 -5.47677092e-02 4.90805298e-01
1.20885849e+00 -6.55513048e-01 8.27024281e-01 1.00211620e+00
-2.90286392e-01 -5.18237591e-01 -6.67287946e-01 -6.51407242e-01
1.09688491e-01 -7.69203380e-02 7.11309671e-01 9.35804605e-01
5.50798953e-01 7.04167306e-01 -1.34787381e-01 9.33193490e-02
1.67130888e-01 -4.68694679e-02 1.92167819e-01 -1.55997717e+00
-2.90146589e-01 -5.17777324e-01 -4.46055740e-01 -4.47084039e-01
1.59686282e-01 -1.24202573e+00 -3.51156503e-01 -1.43943405e+00
-1.38575956e-01 -5.43453917e-02 -4.25337613e-01 5.35226822e-01
3.48825715e-02 -3.33232492e-01 -1.21625327e-01 -3.18914175e-01
-5.47932327e-01 5.40109098e-01 1.28987014e+00 -3.86144131e-01
-4.77086008e-01 -1.35941416e-01 -7.97893941e-01 4.32178587e-01
8.49488914e-01 -6.26349866e-01 -2.57820517e-01 4.66558188e-01
9.54458535e-01 1.44218281e-01 -2.33162612e-01 -9.05886054e-01
3.66216302e-01 -3.82248610e-01 6.13846779e-01 -3.41789573e-01
1.62756503e-01 -6.90057337e-01 7.39005506e-01 6.31633699e-01
-5.99340200e-01 -1.45076990e-01 -2.98543144e-02 7.20581591e-01
1.84160963e-01 -1.25535280e-01 5.92361391e-01 -3.95281225e-01
-1.91966772e-01 5.10113776e-01 -6.98940396e-01 -7.72426188e-01
8.35437179e-01 -3.20132136e-01 -2.94846803e-01 -6.37839437e-02
-8.16837251e-01 -3.26384157e-01 2.66097397e-01 1.98218599e-01
3.22225034e-01 -8.88578713e-01 -1.36237621e-01 -1.58543319e-01
2.62087911e-01 -1.76589951e-01 7.27274418e-02 2.89784700e-01
-6.51651144e-01 7.09365129e-01 -8.39097425e-02 6.73288777e-02
-9.16387081e-01 9.70150769e-01 4.55263019e-01 -3.07702959e-01
-5.75036220e-02 3.10304374e-01 2.84647830e-02 -4.01288718e-01
-2.13294283e-01 -4.68378186e-01 -2.24919710e-02 -1.46403362e-03
1.79842979e-01 3.34533453e-01 5.52578330e-01 -4.38316427e-02
-1.48086950e-01 3.62735391e-02 -4.52883482e-01 4.05162871e-01
1.66883433e+00 6.54374778e-01 -4.71569151e-02 1.14138782e-01
9.15745676e-01 -3.46366137e-01 -1.23573947e+00 9.66475829e-02
1.68904513e-01 2.44681373e-01 2.17784315e-01 -1.11769259e+00
-5.95759273e-01 6.74250185e-01 3.51183116e-01 1.02565788e-01
5.49337387e-01 -3.45404506e-01 3.43907863e-01 1.27368391e+00
8.50857571e-02 -1.25702119e+00 3.00938524e-02 3.92739326e-01
6.89032137e-01 -1.28872979e+00 4.08815920e-01 -6.06584311e-01
-3.33885252e-01 1.64330375e+00 2.17954561e-01 3.47131580e-01
2.89686710e-01 -1.70714065e-01 -4.19730484e-01 -2.57394671e-01
-8.86204839e-01 1.64052576e-01 2.66602546e-01 6.74610496e-01
7.30556846e-01 1.13513291e-01 -4.25935924e-01 4.19050276e-01
-1.36109382e-01 -1.33816376e-01 5.84588230e-01 7.58864939e-01
-5.68058431e-01 -1.55655825e+00 2.23650821e-02 4.00024243e-02
-1.69603094e-01 -2.76446998e-01 -9.60364997e-01 6.32109344e-01
1.97111323e-01 1.07311177e+00 -6.01617992e-01 -2.72082865e-01
4.37737167e-01 3.03596556e-01 6.95470333e-01 -4.72272366e-01
-6.57263875e-01 -5.20184577e-01 4.59469229e-01 -4.36819613e-01
-2.25961804e-01 -2.51893431e-01 -1.39905655e+00 -3.85101765e-01
-5.03876269e-01 5.01994669e-01 1.05843961e+00 1.01378608e+00
6.24776959e-01 4.27687377e-01 7.84248471e-01 -6.99588716e-01
-2.54199386e-01 -5.74276567e-01 -3.99176478e-01 9.13273618e-02
4.97292653e-02 -5.08130908e-01 -8.83927569e-02 -7.06971586e-02] | [4.641829013824463, 5.991127967834473] |
1576210f-9f94-4303-895f-83a027e61b50 | see-through-text-grouping-for-referring-image | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Chen_See-Through-Text_Grouping_for_Referring_Image_Segmentation_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Chen_See-Through-Text_Grouping_for_Referring_Image_Segmentation_ICCV_2019_paper.pdf | See-Through-Text Grouping for Referring Image Segmentation | Motivated by the conventional grouping techniques to image segmentation, we develop their DNN counterpart to tackle the referring variant. The proposed method is driven by a convolutional-recurrent neural network (ConvRNN) that iteratively carries out top-down processing of bottom-up segmentation cues. Given a natural language referring expression, our method learns to predict its relevance to each pixel and derives a See-through-Text Embedding Pixelwise (STEP) heatmap, which reveals segmentation cues of pixel level via the learned visual-textual co-embedding. The ConvRNN performs a top-down approximation by converting the STEP heatmap into a refined one, whereas the improvement is expected from training the network with a classification loss from the ground truth. With the refined heatmap, we update the textual representation of the referring expression by re-evaluating its attention distribution and then compute a new STEP heatmap as the next input to the ConvRNN. Boosting by such collaborative learning, the framework can progressively and simultaneously yield the desired referring segmentation and reasonable attention distribution over the referring sentence. Our method is general and does not rely on, say, the outcomes of object detection from other DNN models, while achieving state-of-the-art performance in all of the four datasets in the experiments.
| [' Tyng-Luh Liu', ' Hwann-Tzong Chen', ' Yi-Chen Lo', ' Songhao Jia', 'Ding-Jie Chen'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['referring-expression-segmentation'] | ['computer-vision'] | [ 5.42988479e-01 5.50387919e-01 -2.19632491e-01 -4.89768177e-01
-1.08384132e+00 -2.83890098e-01 4.96494889e-01 3.66113447e-02
-3.86977017e-01 3.58489037e-01 1.74229816e-01 -2.13025615e-01
-2.47916337e-02 -8.73282313e-01 -9.89901185e-01 -7.42284894e-01
4.81921315e-01 4.23258185e-01 2.32875526e-01 -1.05594516e-01
2.84878671e-01 4.82882112e-01 -1.28292596e+00 3.12331915e-01
7.87077606e-01 1.33610082e+00 2.73534298e-01 6.03558958e-01
-4.35255408e-01 1.10712993e+00 -3.75708550e-01 -5.93040645e-01
9.64744464e-02 -6.57502055e-01 -1.16212738e+00 4.82878774e-01
5.01120687e-01 -4.14921250e-03 -1.01394817e-01 1.21284282e+00
2.35670418e-01 2.28646606e-01 5.16300559e-01 -8.54647338e-01
-1.20945764e+00 8.28831971e-01 -7.42786705e-01 1.82259213e-02
1.92569822e-01 4.99920882e-02 1.31552732e+00 -9.74625528e-01
7.70597696e-01 1.38716972e+00 4.74165678e-01 5.46845019e-01
-1.24256063e+00 -8.98723900e-02 7.37163424e-01 1.15314499e-01
-1.12234056e+00 9.19808522e-02 1.09671879e+00 -2.87500858e-01
6.47453904e-01 1.16240226e-01 6.62090063e-01 7.42238700e-01
-2.18223482e-01 1.37514818e+00 8.43468666e-01 -4.40167427e-01
2.57840693e-01 2.81799603e-02 8.72599706e-02 7.39357293e-01
-3.44378740e-01 -4.15279508e-01 -3.70508313e-01 3.42664421e-01
6.41295314e-01 1.40342012e-01 -2.18352586e-01 -4.21670943e-01
-8.95332098e-01 7.66076446e-01 1.27393675e+00 4.74100292e-01
-7.65003383e-01 4.47868675e-01 3.65726799e-01 -1.57862216e-01
7.38853931e-01 2.69358158e-01 -3.24335009e-01 2.92573631e-01
-1.28066564e+00 -2.46092342e-02 2.93876678e-01 9.16925788e-01
1.13126874e+00 -1.24929301e-01 -5.18360376e-01 7.05140591e-01
2.59001255e-01 1.74615532e-01 2.17669040e-01 -9.87764299e-01
4.71459419e-01 9.14326131e-01 -1.06757663e-01 -8.52729321e-01
-2.76455879e-01 -5.69217443e-01 -7.54670441e-01 7.22208694e-02
5.15905976e-01 -2.61515155e-02 -1.15844703e+00 1.78763807e+00
3.15402508e-01 -2.50182480e-01 -5.47301956e-02 1.00303400e+00
6.67213976e-01 7.61104703e-01 4.09053117e-01 4.73443680e-02
1.47762787e+00 -1.37532103e+00 -6.93995655e-01 -5.29265940e-01
5.11404335e-01 -4.84587312e-01 1.12985170e+00 1.51327267e-01
-1.26059008e+00 -7.17399836e-01 -6.95977271e-01 -6.31772161e-01
-3.65928382e-01 3.18954319e-01 5.31359971e-01 1.75887764e-01
-1.45432317e+00 5.01732886e-01 -6.29835308e-01 -5.95939875e-01
9.11621809e-01 1.80682987e-01 -2.21813028e-03 5.89484628e-03
-9.95866835e-01 8.08476031e-01 2.53905982e-01 5.24336517e-01
-6.64174080e-01 -6.34752572e-01 -8.71894896e-01 8.22196826e-02
4.31144834e-01 -8.51887167e-01 1.01221347e+00 -1.45509577e+00
-1.41187239e+00 1.20635486e+00 -4.58532453e-01 -5.15833139e-01
5.99115491e-01 -3.28034103e-01 2.00161532e-01 3.18895191e-01
3.68617266e-01 1.34589684e+00 9.48490381e-01 -1.57259309e+00
-6.06816292e-01 -3.91470164e-01 2.02462822e-01 1.50275618e-01
-8.00090563e-03 -1.41880177e-02 -9.19353485e-01 -5.17456889e-01
4.91062224e-01 -5.54297447e-01 -2.91613966e-01 2.75527120e-01
-8.86627793e-01 -5.09701848e-01 7.53767788e-01 -5.35319865e-01
9.93468404e-01 -2.17033958e+00 5.99799514e-01 -1.24024525e-02
2.23191872e-01 -6.33355081e-02 -1.37216732e-01 2.52742320e-02
-2.43429586e-01 2.44380549e-01 -5.97392082e-01 -5.77964485e-01
-8.79979879e-03 1.31732285e-01 -3.41129422e-01 4.43488389e-01
6.46572232e-01 1.31236970e+00 -9.39752996e-01 -6.76191032e-01
2.25988075e-01 4.08438891e-01 -5.23534238e-01 2.75873035e-01
-5.55095911e-01 3.59093696e-01 -5.29063940e-01 6.54075921e-01
5.69600642e-01 -2.99397379e-01 -2.33724490e-01 -4.34338689e-01
-2.02485815e-01 -6.23915046e-02 -6.27906203e-01 1.90169346e+00
-5.50854862e-01 7.35732973e-01 9.15660635e-02 -1.33274531e+00
1.11808121e+00 -1.73230037e-01 3.06564867e-01 -6.65664971e-01
4.18747187e-01 4.58500348e-02 -2.54424065e-01 -5.77387333e-01
4.38332468e-01 -8.88438523e-02 -2.27830902e-01 3.99125874e-01
2.53963202e-01 -1.47506744e-01 2.09716391e-02 2.21018955e-01
6.39559031e-01 6.90064311e-01 6.30236119e-02 -2.28275076e-01
7.17978537e-01 2.68111303e-02 2.53607303e-01 8.18083644e-01
-1.40574485e-01 6.11031771e-01 8.84133935e-01 -5.51575959e-01
-9.27790582e-01 -1.02602792e+00 -8.37383643e-02 1.48354495e+00
1.94384903e-01 2.55161494e-01 -1.13473880e+00 -7.73543477e-01
-2.99851477e-01 8.16381335e-01 -1.13457620e+00 3.88632305e-02
-5.20262837e-01 -4.65292811e-01 1.86066270e-01 7.20971227e-01
7.16836631e-01 -1.53995180e+00 -7.00094044e-01 1.80371866e-01
-1.13610789e-01 -9.80772495e-01 -5.54717779e-01 6.65417910e-01
-7.57047832e-01 -6.63969278e-01 -9.51085031e-01 -1.03675854e+00
1.00533831e+00 -1.59529522e-01 1.07087076e+00 1.66759379e-02
-1.39601186e-01 2.79248357e-01 -2.75570750e-01 2.10434403e-02
-4.30725478e-02 1.97334290e-01 -7.51382053e-01 4.26246554e-01
2.81080842e-01 -3.19488466e-01 -7.13750601e-01 -7.06229284e-02
-8.38892817e-01 1.24392256e-01 6.42309189e-01 8.41296196e-01
9.30819273e-01 -4.07550335e-01 4.73580658e-01 -9.88695681e-01
3.98796350e-01 -1.55182451e-01 -5.34716904e-01 3.31501037e-01
-2.36986145e-01 2.93202370e-01 6.25059783e-01 -1.77570581e-01
-1.07455194e+00 3.77575994e-01 -2.73599923e-01 -5.32878935e-01
-1.75989911e-01 3.44298124e-01 -3.18378180e-01 2.76279747e-01
3.82975608e-01 2.77261019e-01 -5.31980753e-01 -1.84127569e-01
1.03196108e+00 3.70634228e-01 8.88006210e-01 -6.51084244e-01
5.90486765e-01 5.49989939e-01 -1.91499576e-01 -4.60979819e-01
-1.50680161e+00 -3.00216794e-01 -1.09898400e+00 -3.29779238e-01
1.53286135e+00 -5.95773280e-01 -6.64255559e-01 2.81337649e-01
-1.47381735e+00 -5.23609579e-01 -6.38258159e-01 -6.53859153e-02
-6.24064505e-01 -1.70745365e-02 -5.97879112e-01 -9.11875129e-01
-3.73624086e-01 -1.23914576e+00 1.49902201e+00 3.76076221e-01
-1.12363711e-01 -9.77781236e-01 -3.77100557e-01 2.90857822e-01
2.73060709e-01 2.30821371e-01 1.14326060e+00 -6.25248551e-01
-8.49722922e-01 -1.90720871e-01 -8.31348717e-01 1.96902737e-01
-5.53044416e-02 1.21689096e-01 -1.43155336e+00 3.49957734e-01
-4.24299203e-02 -3.93794119e-01 1.27693570e+00 6.70407474e-01
1.46288037e+00 -1.85228765e-01 -2.36977816e-01 6.63610756e-01
1.50204659e+00 -8.41600727e-03 6.37546003e-01 4.38834935e-01
9.86480653e-01 8.72944236e-01 5.58056891e-01 1.00202404e-01
4.16972965e-01 5.10008991e-01 7.59976268e-01 -5.15244901e-01
-1.44146994e-01 -3.26662421e-01 6.56800196e-02 2.72854328e-01
1.79285541e-01 -2.47612610e-01 -6.40811265e-01 7.51065075e-01
-1.94358087e+00 -8.84968579e-01 -4.84310538e-02 1.68704021e+00
8.08976352e-01 2.60719210e-01 -2.17521023e-02 -5.93496002e-02
8.83427560e-01 4.28155333e-01 -8.57734621e-01 -7.28532672e-01
-6.96150064e-02 1.99096501e-01 3.06525439e-01 6.02414429e-01
-1.17438495e+00 1.44843817e+00 5.21661425e+00 7.85007954e-01
-1.18619692e+00 1.04944669e-01 1.17363548e+00 2.56242573e-01
-5.79001367e-01 -1.28534093e-01 -6.42192304e-01 -8.05651490e-03
5.08519053e-01 1.52548343e-01 3.07822734e-01 8.40934336e-01
2.45571151e-01 -2.69723982e-01 -1.18299222e+00 6.78940177e-01
2.10940197e-01 -1.28231144e+00 2.73971707e-01 -3.30902547e-01
7.47195899e-01 -1.16117902e-01 2.28463382e-01 3.16836923e-01
2.37335593e-01 -8.37072492e-01 9.47622716e-01 6.96944177e-01
5.93478143e-01 -7.04547703e-01 7.12136805e-01 1.87986612e-01
-1.13477492e+00 -8.55726749e-02 -2.05793589e-01 2.16415703e-01
3.16885740e-01 4.34573889e-01 -4.82637554e-01 2.91469812e-01
5.82233965e-01 8.44426274e-01 -5.97819507e-01 5.90876877e-01
-6.33137167e-01 2.84567326e-01 -1.04076892e-01 -1.29005015e-01
7.68159509e-01 -2.96566904e-01 1.65129215e-01 1.23848522e+00
5.21006621e-03 -6.29355460e-02 4.61551584e-02 1.61611927e+00
-6.26428723e-01 4.04502191e-02 -2.98810124e-01 3.22330415e-01
2.79487640e-01 1.58091414e+00 -1.15899551e+00 -4.87971246e-01
-2.49839470e-01 1.34836602e+00 6.67729616e-01 6.06999993e-01
-8.15450490e-01 -3.41318637e-01 1.60264075e-01 -3.70207131e-02
5.81138074e-01 3.65316778e-01 -5.64110875e-01 -6.41311467e-01
1.31144347e-02 -2.38653973e-01 2.26166934e-01 -1.21496189e+00
-1.21963346e+00 9.31834221e-01 -2.50495672e-01 -9.41984177e-01
-1.45283803e-01 -5.74377656e-01 -8.89554143e-01 1.11931825e+00
-1.69472718e+00 -1.40061009e+00 -2.63510168e-01 4.71701205e-01
7.75265813e-01 3.63001168e-01 5.64048648e-01 -4.70646517e-03
-7.22708285e-01 4.54783589e-01 -4.95865226e-01 2.31221303e-01
3.39212149e-01 -1.48479629e+00 3.06261659e-01 9.54325259e-01
1.20798483e-01 5.23556113e-01 5.42139173e-01 -3.96947592e-01
-9.09557343e-01 -1.48455322e+00 8.90031815e-01 -3.33210438e-01
8.87851119e-01 -3.98597896e-01 -9.72385526e-01 7.05003440e-01
4.47641075e-01 3.15439135e-01 -7.27725681e-03 -6.41636476e-02
-2.27014363e-01 -3.69549617e-02 -1.05302286e+00 6.00743413e-01
8.66634965e-01 -7.87990391e-01 -5.81874967e-01 2.64366478e-01
9.82429802e-01 -3.99438232e-01 -6.06376588e-01 -5.08030169e-02
1.31659493e-01 -8.79117846e-01 8.20958734e-01 -5.83567858e-01
8.01374972e-01 -3.10959041e-01 -1.26104876e-01 -9.55374539e-01
-2.17452228e-01 -3.35544288e-01 4.59442496e-01 1.49317431e+00
6.67279780e-01 -8.80792737e-02 7.98870265e-01 5.84326029e-01
-3.35333437e-01 -1.08635473e+00 -8.43197823e-01 -1.49489850e-01
9.65848044e-02 -6.93281472e-01 1.87743917e-01 5.58584154e-01
-1.89959452e-01 5.10995805e-01 -9.31980908e-02 1.11929365e-01
5.67736208e-01 4.29808378e-01 4.99194384e-01 -8.31984341e-01
5.00400104e-02 -6.06027544e-01 -3.11045021e-01 -1.55098224e+00
4.16440070e-01 -9.48369026e-01 3.68447691e-01 -1.89547908e+00
2.73413479e-01 -6.42659217e-02 -2.60962278e-01 5.63002169e-01
-3.80263805e-01 3.68435562e-01 3.75270456e-01 -2.96558496e-02
-1.02735984e+00 5.45674145e-01 1.54431653e+00 -3.99292052e-01
-2.47893259e-01 -5.32030389e-02 -9.10190403e-01 9.35501695e-01
4.28566664e-01 -3.69720161e-01 -1.29601896e-01 -4.66548771e-01
1.75237253e-01 1.66557208e-02 6.02155805e-01 -5.54917753e-01
3.89567345e-01 2.15710685e-01 4.01865900e-01 -9.37320232e-01
1.94614932e-01 -8.19363177e-01 -5.79781950e-01 1.79280564e-01
-7.39265382e-01 -1.71406463e-01 5.71831949e-02 4.38226372e-01
-2.26974666e-01 -3.71582687e-01 7.99868822e-01 -1.68482050e-01
-6.51140213e-01 1.68958783e-01 -2.33693510e-01 2.69210845e-01
7.51383424e-01 -3.41366708e-01 -1.49145216e-01 -3.36185813e-01
-1.12230587e+00 2.98736006e-01 1.91245362e-01 1.28978148e-01
5.41792929e-01 -1.23148072e+00 -4.30400342e-01 -8.37370008e-02
7.48763159e-02 4.77563173e-01 2.52336830e-01 9.54553366e-01
-2.68720508e-01 3.18676829e-01 3.56385633e-02 -9.41198170e-01
-6.76307559e-01 8.27982247e-01 5.33380628e-01 -2.11127490e-01
-6.23889625e-01 1.18870091e+00 5.59850931e-01 -2.85901815e-01
2.90649712e-01 -6.64338112e-01 -3.37293029e-01 3.58013719e-01
1.75777510e-01 -4.60429043e-02 -1.38682619e-01 -7.61804581e-01
-2.88495362e-01 9.98229027e-01 -1.66113582e-02 1.72960907e-02
1.39476466e+00 -4.57761765e-01 -2.49324515e-01 5.69112360e-01
1.32736254e+00 -2.64673799e-01 -1.62364471e+00 -3.28575522e-01
1.14407457e-01 -2.83771809e-02 1.05309628e-01 -5.29653549e-01
-1.48961055e+00 1.15290916e+00 3.61782014e-01 2.11632550e-01
1.32671285e+00 6.18576348e-01 5.58839560e-01 3.25844258e-01
-1.08416237e-01 -9.25772488e-01 1.55834168e-01 3.97302717e-01
8.78299117e-01 -1.10405219e+00 -9.50135142e-02 -2.59675205e-01
-7.94906676e-01 1.20555353e+00 5.97607195e-01 -3.50488096e-01
3.20664853e-01 8.89651105e-02 1.32088482e-01 -4.92815465e-01
-5.65988719e-01 -5.73664784e-01 3.46711159e-01 4.02391553e-01
4.49981511e-01 -1.67261496e-01 5.56096472e-02 5.16216755e-01
5.00500314e-02 -1.90426946e-01 2.12893024e-01 3.08749199e-01
-4.67970759e-01 -7.21004903e-01 -2.53813267e-01 2.60815144e-01
-2.93354064e-01 -2.28612989e-01 -4.64523971e-01 8.09885025e-01
2.40170285e-01 5.65540135e-01 3.01192701e-01 -7.35582560e-02
5.02024591e-01 7.19267279e-02 2.53119320e-01 -5.26820362e-01
-5.94546974e-01 3.96456271e-01 -2.46184781e-01 -6.83557391e-01
-6.78114355e-01 -7.03824520e-01 -1.59719455e+00 1.27689108e-01
-2.39833340e-01 -6.53992146e-02 2.95956850e-01 1.21429539e+00
-9.12850909e-03 1.05813134e+00 5.85634947e-01 -1.22996092e+00
-1.78028904e-02 -8.34182918e-01 -3.40897858e-01 4.78835344e-01
3.72030079e-01 -3.89891863e-01 -2.69911498e-01 2.46303916e-01] | [10.24178695678711, 1.2149345874786377] |
bc942edf-88ab-4378-8b3f-684012b6ebc9 | unified-language-representation-for-question | 2306.16762 | null | https://arxiv.org/abs/2306.16762v1 | https://arxiv.org/pdf/2306.16762v1.pdf | Unified Language Representation for Question Answering over Text, Tables, and Images | When trying to answer complex questions, people often rely on multiple sources of information, such as visual, textual, and tabular data. Previous approaches to this problem have focused on designing input features or model structure in the multi-modal space, which is inflexible for cross-modal reasoning or data-efficient training. In this paper, we call for an alternative paradigm, which transforms the images and tables into unified language representations, so that we can simplify the task into a simpler textual QA problem that can be solved using three steps: retrieval, ranking, and generation, all within a language space. This idea takes advantage of the power of pre-trained language models and is implemented in a framework called Solar. Our experimental results show that Solar outperforms all existing methods by 10.6-32.3 pts on two datasets, MultimodalQA and MMCoQA, across ten different metrics. Additionally, Solar achieves the best performance on the WebQA leaderboard | ['Yongbin Li', 'Fei Huang', 'Haiyang Yu', 'Cheng Fu', 'Bowen Yu'] | 2023-06-29 | null | null | null | null | ['retrieval', 'question-answering'] | ['methodology', 'natural-language-processing'] | [-1.61688089e-01 -1.41092286e-01 3.23513858e-02 -2.81229347e-01
-1.49054492e+00 -9.69244421e-01 7.82665312e-01 4.14943546e-01
-3.10179561e-01 7.92802513e-01 3.14498693e-01 -5.09312451e-01
-5.75653836e-03 -8.82443666e-01 -5.14547348e-01 -3.63472819e-01
4.79987293e-01 8.95845234e-01 3.21303099e-01 -6.54068291e-01
3.62833232e-01 1.65333882e-01 -1.38720095e+00 7.03144610e-01
1.15268362e+00 1.07505369e+00 2.91858483e-02 5.77759743e-01
-9.62217629e-01 1.00005102e+00 -6.81175649e-01 -9.62127388e-01
-4.84971367e-02 -3.96592140e-01 -1.22694182e+00 -1.64190069e-01
4.56506699e-01 -1.64206192e-01 -5.66341989e-02 8.23155165e-01
4.50536340e-01 1.64617896e-02 8.01543236e-01 -1.21043444e+00
-7.28338242e-01 5.48495770e-01 -3.50554317e-01 -2.33496428e-01
8.84170532e-01 -1.53268680e-01 1.24658239e+00 -9.12594616e-01
5.10352373e-01 1.71302712e+00 2.97591150e-01 4.29999471e-01
-1.05458641e+00 -3.39822769e-01 -4.90289964e-02 3.94785970e-01
-1.24078214e+00 -2.94973642e-01 6.17798209e-01 -4.84579235e-01
7.02685356e-01 3.95002127e-01 1.49814501e-01 8.84612083e-01
-4.07239562e-03 1.03930879e+00 1.33880150e+00 -6.37717605e-01
1.47769764e-01 3.16666633e-01 2.30260924e-01 7.65597284e-01
6.29395097e-02 -5.33749402e-01 -6.17199481e-01 -2.80613810e-01
4.96221304e-01 -2.09758237e-01 -2.86290407e-01 -5.32636583e-01
-1.16594362e+00 1.16550207e+00 5.10443449e-01 2.64514774e-01
-7.94289634e-02 -2.88398445e-01 4.49999094e-01 4.77872759e-01
1.68098047e-01 7.89492130e-01 -3.28946948e-01 -6.48452416e-02
-5.86727321e-01 3.67798120e-01 9.55166757e-01 7.74374068e-01
8.00014198e-01 -5.28460026e-01 -4.22992706e-01 8.91111434e-01
5.97697437e-01 6.75251663e-01 4.42761898e-01 -9.19647038e-01
1.07752430e+00 9.46331322e-01 2.90736824e-01 -1.06482351e+00
-2.59245396e-01 -3.87661792e-02 -5.80953896e-01 -2.98212394e-02
5.82149267e-01 -1.31956702e-02 -9.22644615e-01 1.38325560e+00
2.88079023e-01 -9.04775560e-01 2.82473773e-01 9.40235376e-01
1.12772596e+00 1.00422144e+00 1.21749520e-01 2.71385610e-02
1.51560128e+00 -1.26894534e+00 -1.08603525e+00 -2.29926407e-01
6.11902535e-01 -8.88246059e-01 1.57431543e+00 4.78871584e-01
-1.17896783e+00 -3.71249527e-01 -9.13055718e-01 -6.76001489e-01
-8.40335667e-01 1.77914456e-01 4.89948392e-01 4.75091964e-01
-1.01319599e+00 -1.27001351e-03 -3.51536274e-01 -2.79968500e-01
4.83140647e-02 1.46014569e-02 -2.72398472e-01 -4.48095709e-01
-1.39874744e+00 1.12651765e+00 3.64524871e-01 8.16046894e-02
-7.21588790e-01 -3.01353693e-01 -9.24868762e-01 8.79411101e-02
6.32872880e-01 -9.67455089e-01 1.33052886e+00 -7.42190778e-01
-1.54273546e+00 6.75070345e-01 -2.05497339e-01 -9.83536243e-02
5.27001321e-01 -3.93562198e-01 -2.70472527e-01 3.06166917e-01
1.72711298e-01 5.39339721e-01 6.83049023e-01 -1.48907518e+00
-4.77295727e-01 -4.83871222e-01 6.67871714e-01 4.19738501e-01
-3.30392212e-01 9.11918655e-02 -8.04419756e-01 -3.91705066e-01
3.71112376e-02 -7.61115253e-01 -1.06632233e-01 -7.83671513e-02
-3.26517314e-01 -4.68846232e-01 5.00410020e-01 -9.22117531e-01
1.37840140e+00 -1.80632269e+00 5.66460431e-01 2.51190543e-01
1.74947083e-01 1.47693947e-01 -4.34245653e-02 8.01267147e-01
3.31220806e-01 1.49888694e-01 -1.23320781e-01 -2.69912511e-01
2.91225284e-01 2.08613321e-01 -4.84988779e-01 -1.39738575e-01
9.68963131e-02 9.20029461e-01 -7.88264751e-01 -9.29846287e-01
-7.79836029e-02 2.52289772e-01 -2.92568475e-01 3.57197374e-01
-5.41582644e-01 1.68885544e-01 -6.99618638e-01 7.42858171e-01
4.84312743e-01 -5.18769801e-01 2.11645603e-01 -1.91271782e-01
1.92119423e-02 4.40449625e-01 -1.10574472e+00 1.94397545e+00
-6.34488165e-01 3.69534254e-01 -3.88917193e-04 -7.17457533e-01
8.50468934e-01 3.10269713e-01 1.61442272e-02 -9.94534671e-01
1.54178170e-02 2.50260949e-01 -4.73298281e-01 -5.91745198e-01
4.61720705e-01 1.10568024e-01 -3.14159214e-01 4.39903498e-01
4.81319465e-02 -3.90124887e-01 6.93889380e-01 5.17798007e-01
7.71587193e-01 7.98335746e-02 2.06169039e-02 -5.26937703e-03
7.61088371e-01 3.50104421e-01 -6.67311624e-02 7.15097964e-01
3.46375614e-01 4.87763375e-01 6.21964633e-01 -2.93376297e-01
-7.22160637e-01 -1.15088665e+00 1.72115013e-01 1.21317708e+00
1.07780017e-01 -9.02804077e-01 -6.06786788e-01 -7.94898510e-01
-2.10382566e-01 6.31596327e-01 -4.14006591e-01 7.36627132e-02
-4.26418304e-01 -4.14477617e-01 3.83290559e-01 2.80382872e-01
4.89440799e-01 -8.61314058e-01 -4.94816452e-01 4.59368937e-02
-6.72253847e-01 -8.95696163e-01 -2.05915734e-01 -2.67645195e-02
-7.64364004e-01 -9.38331604e-01 -7.35435903e-01 -5.17564356e-01
4.60581541e-01 2.83232093e-01 1.46327102e+00 8.12329277e-02
-4.03925404e-03 5.73800445e-01 -7.12032497e-01 -4.31672901e-01
-3.11158031e-01 1.94638625e-01 -4.79988486e-01 -7.57534280e-02
1.19903244e-01 -1.07799262e-01 -2.98500746e-01 1.18700482e-01
-1.15256262e+00 2.15764761e-01 8.74345124e-01 9.32925522e-01
4.26582217e-01 -2.49812409e-01 3.25192124e-01 -7.14065790e-01
9.93946612e-01 -3.90769213e-01 -4.80693579e-01 9.50212002e-01
-4.80206013e-01 5.40587127e-01 5.52089036e-01 -1.79853037e-01
-1.12275529e+00 -8.84087309e-02 -2.88227890e-02 -1.82646334e-01
1.01586543e-01 9.73435581e-01 -3.94889325e-01 8.41962323e-02
6.45579755e-01 1.33263722e-01 -3.73518504e-02 -5.35315454e-01
7.41637468e-01 6.23738289e-01 2.58035988e-01 -8.18768919e-01
8.63649368e-01 2.82085449e-01 -1.44834638e-01 -4.52540517e-01
-1.06112266e+00 -4.26411718e-01 -5.32073855e-01 -1.26537129e-01
8.99089038e-01 -8.27427149e-01 -6.06853426e-01 5.76330861e-03
-1.29534996e+00 -1.14176972e-02 4.18503881e-02 1.34729818e-01
-4.67982620e-01 4.36661094e-01 -5.13345659e-01 -7.53774583e-01
-4.04376179e-01 -1.12727654e+00 1.14472497e+00 2.80379295e-01
-8.13992545e-02 -9.85608578e-01 1.47962406e-01 8.78178060e-01
4.69711632e-01 2.39974018e-02 1.36097574e+00 -6.13107860e-01
-6.77247226e-01 -7.27611706e-02 -3.54124576e-01 4.11804765e-02
-3.43086347e-02 -5.48434854e-02 -8.27713668e-01 -1.80976436e-01
-2.65729427e-01 -1.04368722e+00 7.64848888e-01 -2.63147205e-01
1.10844851e+00 -3.44783783e-01 -1.58380702e-01 1.57439113e-01
1.28259957e+00 1.82921037e-01 6.04166329e-01 5.24221122e-01
5.85838318e-01 8.10836375e-01 8.01388562e-01 2.49057159e-01
9.07603741e-01 7.90849984e-01 4.06223148e-01 -3.29436697e-02
1.45807549e-01 -3.31120580e-01 2.31211856e-01 8.79523993e-01
1.92578152e-01 -4.34729964e-01 -1.20215476e+00 5.57612896e-01
-1.97475481e+00 -7.21610546e-01 1.32456971e-02 1.90684426e+00
1.08928978e+00 -1.45219386e-01 1.07086971e-02 -1.03011139e-01
2.74697453e-01 -2.22772639e-02 -1.38691008e-01 -4.38515782e-01
-1.56884968e-01 9.64033902e-02 -1.70816928e-02 3.64783764e-01
-9.48011577e-01 8.03226531e-01 6.46511078e+00 8.90164316e-01
-8.89516830e-01 6.52581453e-03 5.52393377e-01 2.16209576e-01
-6.17471397e-01 -1.08830653e-01 -6.51959717e-01 1.27862856e-01
9.54279125e-01 -1.34603605e-01 3.99162292e-01 6.03367507e-01
-3.29003006e-01 -2.25235224e-01 -1.04676449e+00 1.17985022e+00
3.04165006e-01 -1.28492069e+00 6.36956573e-01 -3.35540771e-01
4.68131572e-01 -4.29541200e-01 1.60041973e-01 5.96720040e-01
4.17756170e-01 -1.31769931e+00 6.73316717e-01 5.57837784e-01
4.91768450e-01 -6.32025421e-01 7.83742964e-01 3.85315418e-01
-9.82177794e-01 2.05572601e-02 -2.10968450e-01 1.39630392e-01
1.47859380e-01 2.24339917e-01 -6.99041545e-01 1.09347105e+00
8.66672039e-01 2.22611174e-01 -9.39599812e-01 1.01217425e+00
-1.53318182e-01 2.22820044e-01 -9.79821086e-02 -2.28762031e-01
3.16092670e-01 -9.85141769e-02 1.14469817e-02 9.42771018e-01
3.53479475e-01 -8.48993286e-03 2.03128681e-01 6.19511425e-01
1.69730093e-02 4.83679622e-01 -4.21121389e-01 -3.07714045e-01
2.15698063e-01 1.36895216e+00 -4.43425924e-01 -3.62046212e-01
-6.36350155e-01 8.14809024e-01 5.07132471e-01 2.96258897e-01
-7.12982178e-01 -4.53847021e-01 1.35873169e-01 -2.38849208e-01
1.15895428e-01 -2.51947850e-01 -1.50911421e-01 -1.35040188e+00
1.71096727e-01 -1.42140675e+00 8.85813475e-01 -1.13822508e+00
-1.39857805e+00 8.11634123e-01 9.26007032e-02 -1.03549242e+00
-5.34756362e-01 -8.02680910e-01 -1.57853171e-01 8.67511570e-01
-1.60070038e+00 -1.34112108e+00 -3.97984236e-01 9.01311696e-01
4.61203456e-01 -2.60195285e-02 9.63595092e-01 4.18657571e-01
-2.41888329e-01 4.08524841e-01 5.05340137e-02 1.19672596e-01
9.56161916e-01 -1.46555316e+00 -8.36879238e-02 4.94557381e-01
4.37420130e-01 6.33400500e-01 7.17241943e-01 -3.11793298e-01
-1.68699682e+00 -5.86630642e-01 1.13388729e+00 -7.22839653e-01
8.43241394e-01 -3.57728630e-01 -1.02514493e+00 4.73468751e-01
6.14639819e-01 -4.11541671e-01 7.17146099e-01 1.51016533e-01
-6.98393166e-01 -8.13652650e-02 -9.05388534e-01 7.11129606e-01
4.88212973e-01 -6.57911956e-01 -9.18106139e-01 5.01037419e-01
7.08261192e-01 -5.83751619e-01 -7.89780140e-01 2.02388644e-01
3.59377116e-01 -6.53639257e-01 9.34685707e-01 -7.80235112e-01
4.88336205e-01 -4.99679327e-01 -2.84733683e-01 -1.33565903e+00
3.90930064e-02 -4.37779188e-01 -4.36449833e-02 1.27767277e+00
7.78786004e-01 -2.14870289e-01 3.17419976e-01 5.63383877e-01
4.77681309e-02 -6.03638947e-01 -8.01464975e-01 -3.03953916e-01
1.10443674e-01 -1.68317482e-01 6.48049533e-01 8.71810436e-01
-4.16074954e-02 7.46041775e-01 -2.88428575e-01 1.91960316e-02
1.00678794e-01 5.13320923e-01 8.64768028e-01 -1.22291219e+00
-9.15325806e-02 -3.42984706e-01 8.36368799e-02 -1.14893293e+00
1.13668978e-01 -6.87059939e-01 8.77920818e-03 -2.07400918e+00
2.02769235e-01 -3.15050811e-01 -1.14349373e-01 5.20181358e-01
-1.81522965e-01 1.19207182e-03 3.66244227e-01 9.02598128e-02
-1.07642543e+00 6.69066370e-01 1.37873363e+00 -4.19367462e-01
8.07844847e-02 -3.84234667e-01 -7.59002447e-01 4.81420547e-01
5.75141549e-01 -1.72910869e-01 -5.31772912e-01 -8.24049950e-01
5.66926658e-01 4.18558031e-01 1.56106532e-01 -7.92419016e-01
3.77479643e-01 -3.24584007e-01 2.85155892e-01 -7.09327936e-01
5.86134911e-01 -8.86192322e-01 -2.45174602e-01 1.05596393e-01
-4.16644663e-01 6.46670282e-01 1.17877275e-01 2.22109541e-01
-6.91542327e-01 -2.33389661e-01 2.45110333e-01 -2.24303201e-01
-6.37181282e-01 -2.40349527e-02 -1.68119058e-01 2.69470960e-01
7.31563747e-01 3.79053473e-01 -7.81997144e-01 -7.37962425e-01
-5.01993597e-01 6.95366621e-01 1.73266008e-01 6.18144095e-01
5.96896470e-01 -1.46561015e+00 -6.18424892e-01 -2.35660002e-01
3.63603741e-01 -1.00069463e-01 2.59550810e-01 6.65431261e-01
-6.93897545e-01 8.89035523e-01 -3.45986366e-01 -5.34218609e-01
-1.13901424e+00 5.97573638e-01 2.03539684e-01 -6.56134188e-01
3.59976478e-02 5.91818333e-01 -1.94902846e-03 -5.97201169e-01
1.69631675e-01 -1.54796734e-01 -5.73348522e-01 4.01684046e-01
5.40213704e-01 7.07462942e-03 1.75501540e-01 -6.73481464e-01
-2.03735098e-01 6.13948047e-01 -1.24974646e-01 -3.16791803e-01
9.57736492e-01 -2.66538769e-01 -4.42847639e-01 6.12368286e-01
9.00279522e-01 1.39400825e-01 -5.55912435e-01 -3.33990574e-01
2.69421101e-01 -3.44047815e-01 -1.34702727e-01 -1.11582315e+00
-7.14268744e-01 1.07320249e+00 3.75298828e-01 4.66129780e-01
1.10984588e+00 2.35700294e-01 6.13202572e-01 8.04504097e-01
3.20731997e-01 -9.56745565e-01 5.18309295e-01 5.69564164e-01
1.20031500e+00 -1.38032711e+00 -1.90197825e-01 -2.46742502e-01
-8.93240988e-01 1.23673439e+00 8.48006427e-01 4.68548447e-01
1.52596906e-01 -2.32857361e-01 4.34406996e-01 -2.81954139e-01
-9.27981317e-01 -2.63045698e-01 7.90038764e-01 3.17137986e-01
5.62027395e-01 -1.27845392e-01 -2.61842340e-01 6.86618507e-01
-2.31593683e-01 -3.24811667e-01 2.29334116e-01 1.10405207e+00
-4.75286216e-01 -1.32624900e+00 -6.44292593e-01 2.18875661e-01
-4.37956899e-01 -5.62545918e-02 -6.83586061e-01 9.15542781e-01
-2.28338763e-01 1.15275753e+00 -1.15184180e-01 -1.42278567e-01
3.91031981e-01 3.87950808e-01 4.03062701e-01 -4.03163522e-01
-5.21107197e-01 -1.54409736e-01 2.71951288e-01 -5.46495855e-01
-4.90214378e-01 -3.54956239e-01 -1.06751406e+00 -1.04678787e-01
-7.37020001e-02 6.11892879e-01 6.79906309e-01 1.04704165e+00
1.87275842e-01 3.02710146e-01 2.99633056e-01 -3.46047372e-01
-5.89498878e-01 -8.62117767e-01 -5.31523377e-02 4.19692874e-01
2.71934152e-01 -6.93654835e-01 -1.67373464e-01 -1.41976953e-01] | [11.039430618286133, 8.03691577911377] |
634f0d94-cc58-4a39-a511-7ab32c5fff94 | techniques-for-effective-and-efficient-fire | 1506.03844 | null | http://arxiv.org/abs/1506.03844v2 | http://arxiv.org/pdf/1506.03844v2.pdf | Techniques for effective and efficient fire detection from social media images | Social media could provide valuable information to support decision making in
crisis management, such as in accidents, explosions and fires. However, much of
the data from social media are images, which are uploaded in a rate that makes
it impossible for human beings to analyze them. Despite the many works on image
analysis, there are no fire detection studies on social media. To fill this
gap, we propose the use and evaluation of a broad set of content-based image
retrieval and classification techniques for fire detection. Our main
contributions are: (i) the development of the Fast-Fire Detection method
(FFDnR), which combines feature extractor and evaluation functions to support
instance-based learning, (ii) the construction of an annotated set of images
with ground-truth depicting fire occurrences -- the FlickrFire dataset, and
(iii) the evaluation of 36 efficient image descriptors for fire detection.
Using real data from Flickr, our results showed that FFDnR was able to achieve
a precision for fire detection comparable to that of human annotators.
Therefore, our work shall provide a solid basis for further developments on
monitoring images from social media. | ['Mirela Cazzolato', 'Alceu Costa', 'Caetano Traina Jr', 'Willian Oliveira', 'Marcos Bedo', 'Jose Rodrigues', 'Gustavo Blanco', 'Agma Traina'] | 2015-06-11 | null | null | null | null | ['fire-detection'] | ['time-series'] | [ 3.53140771e-01 -3.93387288e-01 3.73341471e-01 -3.21801640e-02
-6.30536854e-01 -7.70990431e-01 7.85318375e-01 7.88818955e-01
-9.19137120e-01 4.92249370e-01 6.39577582e-02 -7.92562664e-02
-3.16846311e-01 -1.28585815e+00 -1.99181914e-01 -8.33448470e-01
-3.64650220e-01 1.73500881e-01 4.77084219e-01 -2.55320877e-01
3.09411168e-01 6.98239028e-01 -2.01209283e+00 6.02850378e-01
1.72095433e-01 1.11552858e+00 4.12329048e-01 7.81369328e-01
4.29401733e-02 1.03433812e+00 -6.04329705e-01 8.72762874e-02
2.13633716e-01 -2.48017877e-01 -9.15409744e-01 1.14279963e-01
1.26675501e-01 -2.87764162e-01 -3.95225316e-01 7.34802783e-01
5.77311277e-01 2.87433445e-01 8.30683351e-01 -1.36496449e+00
-3.42888802e-01 3.02993864e-01 -2.09942222e-01 5.92607856e-01
6.59993768e-01 1.30235866e-01 5.50510347e-01 -6.74347818e-01
8.95037830e-01 8.97679448e-01 7.02921510e-01 8.67970884e-02
-7.43679106e-01 -3.80105793e-01 -2.02347949e-01 4.33889270e-01
-1.88783956e+00 -3.45400989e-01 5.62484086e-01 -7.40762055e-01
1.24546909e+00 5.43570101e-01 9.08465803e-01 7.31987119e-01
-3.12812716e-01 5.45162976e-01 1.24565625e+00 -7.65999913e-01
2.15994671e-01 2.08623763e-02 -9.66259390e-02 6.90088689e-01
3.36556993e-02 8.46464410e-02 -3.16023797e-01 -2.34526530e-01
6.18285060e-01 3.37297946e-01 -2.55092323e-01 7.42294341e-02
-1.24406087e+00 8.58229518e-01 5.00447810e-01 7.24083722e-01
-7.68681586e-01 -6.74465001e-02 4.88946706e-01 3.36230248e-01
9.22691762e-01 2.24938497e-01 -2.67314166e-02 4.62400988e-02
-1.02427173e+00 3.64378542e-01 5.24289548e-01 4.94194239e-01
8.00198495e-01 -3.91237587e-01 -2.27820858e-01 6.84730053e-01
1.84238702e-01 1.01000631e+00 -3.67143960e-03 -7.11537480e-01
3.73106331e-01 6.81895971e-01 2.61853337e-01 -1.77044439e+00
-6.22890472e-01 2.21036181e-01 -8.81140888e-01 5.40767610e-02
2.85137951e-01 5.17125577e-02 -6.10355616e-01 9.47754085e-01
2.69880235e-01 5.81777468e-02 -1.90940544e-01 9.52676952e-01
8.61893892e-01 6.92695618e-01 2.05041185e-01 -3.32245052e-01
1.44725955e+00 -4.67708200e-01 -7.11418748e-01 1.99050769e-01
4.69502002e-01 -5.30354798e-01 7.26439774e-01 3.08337510e-01
-7.13135421e-01 -4.74918097e-01 -5.28886259e-01 4.16290939e-01
-9.20783281e-01 -2.12338939e-01 3.46942812e-01 4.20590937e-01
-8.79582047e-01 4.24014777e-01 -4.87981528e-01 -9.45632756e-01
5.65859437e-01 -6.12252131e-02 -5.61741769e-01 -4.30390565e-03
-1.17720509e+00 9.44522381e-01 6.04645967e-01 3.47762592e-02
-1.06851196e+00 -3.13348353e-01 -6.71316743e-01 -4.51278299e-01
4.43809777e-01 -4.04811352e-01 9.09632087e-01 -7.99009740e-01
-5.66309690e-01 1.24927247e+00 2.50914514e-01 -4.74878520e-01
5.10814130e-01 -9.36790407e-02 -5.93105853e-01 7.55227327e-01
2.20008254e-01 6.20999515e-01 9.45675373e-01 -1.32720208e+00
-8.69441807e-01 -2.41956651e-01 1.83716744e-01 6.79233000e-02
-4.42248255e-01 4.68489319e-01 -3.83189227e-03 -6.60912335e-01
-2.18161955e-01 -7.47683942e-01 -2.93276817e-01 -6.06189445e-02
-2.35738888e-01 -5.82886301e-02 8.43981385e-01 -5.38652897e-01
1.46387661e+00 -2.16509581e+00 -1.20624918e-02 3.20839167e-01
1.49951398e-01 4.67089087e-01 -3.61654833e-02 8.74450445e-01
6.58997521e-02 3.67694587e-01 -4.77295190e-01 1.96779370e-01
-2.57504135e-01 1.98718697e-01 -4.83567595e-01 5.18846035e-01
1.46422952e-01 6.01531744e-01 -1.34557068e+00 -7.56291330e-01
7.76215315e-01 6.01786017e-01 -7.00810403e-02 1.51547372e-01
7.91184679e-02 3.16702396e-01 -3.93814921e-01 5.18138528e-01
3.44808519e-01 2.08490223e-01 -3.79644215e-01 -2.27699056e-01
-4.95580345e-01 -3.43359619e-01 -1.17401242e+00 1.26147723e+00
-1.50118142e-01 6.83629632e-01 -2.27396458e-01 -6.69493437e-01
7.67939866e-01 6.43711269e-01 9.86976027e-01 -5.01804292e-01
2.71512598e-01 -8.97721723e-02 -7.27461696e-01 -9.61966634e-01
5.99077404e-01 -1.11354843e-01 -6.72324374e-02 8.35168958e-01
-1.56782165e-01 -3.65980715e-01 6.15966797e-01 3.49259586e-04
1.30124056e+00 -2.92029023e-01 1.50040850e-01 -1.29073307e-01
5.13577223e-01 5.15251398e-01 -1.01262718e-01 9.04598594e-01
-2.30657443e-01 7.53155291e-01 -1.82306424e-01 -1.08402920e+00
-1.00083959e+00 -6.70954645e-01 -5.56298606e-02 9.73759294e-01
-6.32176772e-02 -7.04255462e-01 -1.01485372e+00 -6.90239608e-01
-1.31505847e-01 3.85508180e-01 -4.25230294e-01 1.97473422e-01
-1.96478650e-01 -9.30413306e-01 7.41666853e-01 2.27421075e-01
7.18360484e-01 -1.63510811e+00 -1.26669097e+00 6.51039183e-02
-6.92833662e-01 -9.22525346e-01 3.36967371e-02 -5.30190133e-02
-4.39170420e-01 -1.58109260e+00 -5.18151700e-01 -4.67852503e-01
5.95361650e-01 7.77214527e-01 9.61402774e-01 4.86402750e-01
-7.48698473e-01 9.46460187e-01 -9.62056756e-01 -6.76554441e-01
-1.70144990e-01 -2.72708721e-02 -3.95913869e-02 1.05753420e-02
5.79557002e-01 -3.15921545e-01 -5.32895863e-01 4.59958851e-01
-1.54038405e+00 -2.34008327e-01 5.20702861e-02 1.09774217e-01
5.54762959e-01 7.83141673e-01 2.30157197e-01 -5.19651949e-01
7.49145925e-01 -4.11749721e-01 -3.58466834e-01 2.57024139e-01
-3.53079826e-01 -5.30326188e-01 2.39768222e-01 -5.98983653e-02
-6.16226792e-01 2.23158449e-01 5.14599048e-02 -1.50578469e-01
-6.22466207e-01 6.83104694e-01 5.12637973e-01 3.70717421e-02
9.99337554e-01 1.19341068e-01 -1.97116524e-01 -3.42515290e-01
8.96580443e-02 1.09113216e+00 4.62494493e-01 -1.12834452e-02
8.43580127e-01 6.93344891e-01 -1.34887367e-01 -1.56123900e+00
-8.81399214e-01 -1.10071433e+00 -9.37871695e-01 -9.05715942e-01
1.07364690e+00 -8.74742568e-01 -3.18323910e-01 7.75402725e-01
-1.03812099e+00 -3.38478029e-01 -3.18292230e-01 3.74153048e-01
-4.29739356e-01 3.47847044e-01 -2.40251735e-01 -1.33679616e+00
-4.17199552e-01 -5.05189717e-01 1.18477213e+00 -2.19257325e-01
-1.08673379e-01 -7.62258053e-01 4.37415302e-01 4.04746234e-01
4.31620359e-01 6.43397391e-01 1.09045528e-01 -7.64794499e-02
-2.13937372e-01 -4.22308087e-01 -2.15464100e-01 3.04484189e-01
2.29561865e-01 3.50701690e-01 -1.06450963e+00 -8.42586458e-02
-1.78524077e-01 -3.84174228e-01 1.15240192e+00 1.55889213e-01
8.77611339e-01 -1.12594619e-01 -3.48399848e-01 1.53674185e-01
1.42226017e+00 1.41670853e-01 1.08346224e+00 8.01705599e-01
5.83331406e-01 6.20269716e-01 8.08719456e-01 7.69538879e-01
2.00615793e-01 6.70823872e-01 5.52819192e-01 -1.84682071e-01
9.76851434e-02 2.00304203e-02 1.35320917e-01 3.27228546e-01
-8.73049557e-01 -4.89237428e-01 -1.10964894e+00 4.83522862e-01
-1.91389990e+00 -1.50165522e+00 -4.55164611e-01 2.03195429e+00
2.59942830e-01 -3.30018938e-01 4.90061700e-01 8.15228820e-01
8.26495349e-01 3.90319735e-01 2.90311694e-01 9.13801044e-02
-1.89197704e-01 1.27010524e-01 5.93012989e-01 2.77189910e-01
-1.79213405e+00 7.28479445e-01 6.66367197e+00 6.02849305e-01
-7.29625583e-01 2.46722490e-01 1.09142832e-01 2.33638093e-01
2.21499503e-01 -2.48308614e-01 -3.97837013e-01 1.87312588e-01
7.20376194e-01 1.89278424e-01 7.00103939e-01 4.67386127e-01
4.57541794e-01 -3.61531824e-01 -2.98438281e-01 9.40815568e-01
2.56964296e-01 -1.03467190e+00 -1.33855224e-01 -1.41839191e-01
3.95144850e-01 1.63486123e-01 -5.60530722e-01 -1.97081104e-01
-3.56156975e-02 -8.45730603e-01 8.83320630e-01 9.28117275e-01
6.45268261e-01 -8.41642022e-01 8.96421373e-01 3.97085726e-01
-1.30164850e+00 -1.53889745e-01 -4.86253411e-01 -2.63623804e-01
2.93177247e-01 6.67942882e-01 -7.22586095e-01 6.25935018e-01
8.61902475e-01 7.50107884e-01 -6.22984767e-01 1.17049515e+00
-4.34760928e-01 4.59949374e-01 -3.95079195e-01 7.38778934e-02
2.24368483e-01 9.74441916e-02 5.51384747e-01 1.52157199e+00
2.15877429e-01 2.42129385e-01 4.40574527e-01 2.98764139e-01
5.66271067e-01 2.37883210e-01 -1.06128061e+00 -3.53733659e-01
2.13654026e-01 1.42740583e+00 -1.26951063e+00 -1.32549256e-01
-9.84389260e-02 8.89240384e-01 5.89437746e-02 5.46925068e-02
-7.64820516e-01 -4.05750841e-01 1.13302462e-01 6.18747950e-01
1.19145019e-02 -4.27834511e-01 4.01553094e-01 -9.09334362e-01
-1.68997824e-01 -5.52192926e-01 7.61439741e-01 -9.73502576e-01
-1.22759771e+00 1.03439796e+00 5.16868114e-01 -1.25100923e+00
-1.52813226e-01 -4.44397151e-01 -2.42030337e-01 2.96709388e-01
-1.24054432e+00 -1.15109611e+00 -7.62361348e-01 8.17999065e-01
2.16327965e-01 -6.25178665e-02 1.15639710e+00 4.65659738e-01
-2.90022880e-01 -2.92072505e-01 -3.70013148e-01 2.23967329e-01
3.09645057e-01 -8.95324945e-01 3.58617716e-02 6.63621426e-01
4.99235690e-01 -7.85217583e-02 4.72870141e-01 -6.66716337e-01
-9.34149623e-01 -1.35600436e+00 1.03566933e+00 -5.10962903e-01
5.60496211e-01 1.13334600e-02 -5.85053265e-01 2.84384400e-01
3.05625815e-02 -2.16922145e-02 7.26624370e-01 -3.31157684e-01
-1.36007354e-01 -4.43686396e-02 -1.37221539e+00 1.96481377e-01
1.10429692e+00 -8.10220659e-01 -5.03345728e-01 7.77907252e-01
-6.88979030e-03 2.19097197e-01 -7.99877584e-01 2.34489381e-01
4.02291834e-01 -9.42356467e-01 1.11162674e+00 -3.02985430e-01
3.75036031e-01 -5.88072836e-01 -2.62930095e-01 -1.16653907e+00
-5.21199107e-01 -1.38608322e-01 2.89672136e-01 1.07994699e+00
2.58786604e-02 -3.11457574e-01 1.02161594e-01 3.22582960e-01
4.07605022e-02 6.54517710e-02 -6.97005332e-01 -6.60025418e-01
-5.76695919e-01 -8.24961185e-01 3.97962034e-01 9.48230863e-01
1.46755487e-01 -9.53713581e-02 -3.51837873e-01 2.45572746e-01
3.81497473e-01 -1.78701267e-01 6.28476322e-01 -1.23949313e+00
2.72190005e-01 -1.99885815e-01 -7.39056170e-01 6.12762310e-02
-1.11215331e-01 -8.13425899e-01 9.33831483e-02 -1.80505192e+00
3.13504517e-01 -3.94573808e-01 -4.01477605e-01 7.97895372e-01
1.82044223e-01 9.83295023e-01 1.66930497e-01 4.65546578e-01
-9.32467341e-01 5.65378629e-02 8.56767595e-01 -2.93116063e-01
-5.37161827e-02 -2.74975210e-01 -9.51387137e-02 8.87575209e-01
9.32802677e-01 -8.30994964e-01 -2.01046616e-01 -3.12843621e-01
4.59883124e-01 -2.55048424e-01 8.37942839e-01 -1.32559335e+00
1.22866929e-01 -3.49426389e-01 5.46603687e-02 -7.18183458e-01
-2.00080313e-02 -1.08726490e+00 4.51717645e-01 4.68228370e-01
-1.95450578e-02 1.76536828e-01 -6.62584277e-03 4.76592094e-01
-2.92666674e-01 -4.28088218e-01 4.11371499e-01 -4.52358693e-01
-8.86186481e-01 1.55411065e-01 -1.00002527e+00 -7.75229484e-02
1.20731938e+00 -1.00696415e-01 -3.39596242e-01 -2.87762552e-01
-6.16832376e-01 -2.00100124e-01 5.57608426e-01 2.94994891e-01
6.15277708e-01 -1.18808901e+00 -8.91865194e-01 1.32568702e-01
3.72017115e-01 6.73488304e-02 3.71352881e-01 7.06945360e-01
-1.01491177e+00 2.20951825e-01 -2.26428986e-01 -5.61875641e-01
-1.36928487e+00 8.47582698e-01 2.24546529e-03 -5.40409014e-02
-6.58458173e-01 4.08223063e-01 -3.79128367e-01 -1.40494198e-01
1.07643619e-01 -8.48392621e-02 -5.65586984e-01 6.50677800e-01
1.10625458e+00 5.57753205e-01 2.41571575e-01 -1.04639900e+00
-5.02853274e-01 5.78446627e-01 5.97941458e-01 -2.21493989e-01
1.68390000e+00 -2.73620273e-04 -3.27978015e-01 4.72695231e-01
5.67443848e-01 -3.34282517e-01 -7.95195639e-01 2.08107457e-01
1.11173801e-02 -7.36573756e-01 3.50632489e-01 -6.93511963e-01
-1.18676257e+00 5.38521647e-01 9.02107298e-01 9.57896113e-01
1.39808559e+00 1.00555886e-02 4.81628627e-01 7.76213884e-01
7.63469160e-01 -1.32038879e+00 -7.89425820e-02 4.61541921e-01
1.15487599e+00 -1.15420079e+00 2.91451346e-02 -2.82344490e-01
-5.28945148e-01 1.06699705e+00 -3.20095122e-01 -7.34190941e-02
8.73143315e-01 2.27104932e-01 5.99220358e-02 -7.57298708e-01
-3.21137518e-01 -7.61540115e-01 1.22166499e-01 7.44356096e-01
3.42677459e-02 1.38912097e-01 -2.40261182e-01 8.22956339e-02
2.65218437e-01 3.17677617e-01 1.92539662e-01 1.42564654e+00
-9.14501071e-01 -8.30158770e-01 -8.00265610e-01 4.30247605e-01
-4.19732004e-01 -8.57371930e-03 -7.22379029e-01 6.89419031e-01
4.95873749e-01 1.46908617e+00 -5.16733229e-02 -6.23531103e-01
5.78650773e-01 -1.03656445e-02 5.38689256e-01 -5.63437164e-01
-8.72104645e-01 -3.38816822e-01 2.65528947e-01 -3.67895871e-01
-1.14987600e+00 -7.47826219e-01 -7.92622387e-01 -2.73085982e-01
-3.89995903e-01 2.39355415e-01 6.11667037e-01 9.04778004e-01
1.94920786e-02 2.25529343e-01 7.12104619e-01 -1.11399448e+00
2.50730187e-01 -1.04354763e+00 -9.42922175e-01 7.27183223e-01
-4.46995161e-02 -7.48743713e-01 -6.04071558e-01 3.95774752e-01] | [9.146439552307129, -1.1490392684936523] |
fc4a9ae4-b634-467a-9d7f-20c37ad984b9 | the-scattering-transform-network-with | 2206.07857 | null | https://arxiv.org/abs/2206.07857v1 | https://arxiv.org/pdf/2206.07857v1.pdf | The Scattering Transform Network with Generalized Morse Wavelets and Its Application to Music Genre Classification | We propose to use the Generalized Morse Wavelets (GMWs) instead of commonly-used Morlet (or Gabor) wavelets in the Scattering Transform Network (STN), which we call the GMW-STN, for signal classification problems. The GMWs form a parameterized family of truly analytic wavelets while the Morlet wavelets are only approximately analytic. The analyticity of underlying wavelet filters in the STN is particularly important for nonstationary oscillatory signals such as music signals because it improves interpretability of the STN representations by providing multiscale amplitude and phase (and consequently frequency) information of input signals. We demonstrate the superiority of the GMW-STN over the conventional STN in music genre classification using the so-called GTZAN database. Moreover, we show the performance improvement of the GMW-STN by increasing its number of layers to three over the typical two-layer STN.} | ['David Weber', 'Naoki Saito', 'Wai Ho Chak'] | 2022-06-16 | null | null | null | null | ['genre-classification'] | ['computer-vision'] | [ 3.81145567e-01 -5.59484959e-01 3.01665753e-01 4.23508920e-02
-5.43163002e-01 -4.45605397e-01 1.82314903e-01 -2.93784559e-01
-1.20716684e-01 6.16221726e-01 3.09865147e-01 -7.95070548e-03
-4.43375945e-01 -6.12651765e-01 -3.02787125e-01 -9.99006629e-01
-4.34223562e-01 -2.31010169e-01 1.47590056e-01 -3.34833533e-01
-1.43172935e-01 6.60418570e-01 -1.70590520e+00 5.04180908e-01
5.11520982e-01 1.37391889e+00 8.25040266e-02 8.97489488e-01
1.32829040e-01 4.45466131e-01 -7.15024889e-01 2.57060900e-02
8.07888061e-02 -4.64427680e-01 -1.65019780e-01 -1.38779804e-01
-5.25593981e-02 4.19905722e-01 -7.19906390e-02 1.10903883e+00
4.62489039e-01 4.30058330e-01 6.58948779e-01 -1.09366620e+00
-5.64159989e-01 6.63935065e-01 -4.23512518e-01 5.86924970e-01
2.61517733e-01 -5.19548357e-01 1.03255832e+00 -1.09077382e+00
2.15424731e-01 1.38785398e+00 1.32550406e+00 2.51929439e-03
-1.42140937e+00 -7.45767117e-01 -4.56717998e-01 3.92810851e-01
-1.36981535e+00 -4.48444426e-01 1.21658659e+00 -2.48996347e-01
5.69462478e-01 3.68730783e-01 6.04208469e-01 1.23135686e+00
2.05265045e-01 6.31784081e-01 1.11049521e+00 -6.79094076e-01
2.58593529e-01 -6.67049348e-01 1.40951067e-01 4.08195287e-01
2.24044546e-01 1.37645945e-01 -9.01186883e-01 -5.70951879e-01
1.16111398e+00 -1.79713815e-02 -4.93421078e-01 4.97881770e-02
-1.30155540e+00 8.15485716e-01 1.60621166e-01 8.45513523e-01
-7.95611143e-01 4.58792001e-01 5.06604433e-01 7.05403149e-01
7.64506400e-01 2.81417280e-01 -2.29779467e-01 -1.05287850e-01
-1.10528302e+00 8.08367729e-02 5.89412808e-01 5.07728517e-01
1.71414927e-01 8.97561252e-01 4.48478349e-02 9.23735321e-01
-2.40947809e-02 8.35412264e-01 4.26064819e-01 -9.28837717e-01
1.86981693e-01 -1.67859420e-01 2.45119214e-01 -1.27006435e+00
-5.44361413e-01 -9.36021149e-01 -1.44720590e+00 1.22825801e-02
3.32381636e-01 2.41900682e-02 -4.67691094e-01 1.42806160e+00
-9.88986939e-02 5.17942131e-01 3.61670554e-01 6.60664558e-01
7.93315411e-01 7.50660837e-01 -4.58916247e-01 -4.89414603e-01
1.35060155e+00 -4.32915151e-01 -1.07094884e+00 1.22281484e-01
-2.91012898e-02 -1.14282656e+00 8.27635407e-01 7.39875913e-01
-9.62628186e-01 -1.00499701e+00 -9.95512128e-01 4.00394469e-01
-7.86143616e-02 1.28209829e-01 4.51937824e-01 5.55940449e-01
-8.37778270e-01 8.89095962e-01 -9.54533100e-01 1.38152242e-02
1.88701570e-01 4.79711257e-02 -1.68416008e-01 1.67169586e-01
-1.16436601e+00 5.90889633e-01 7.66460449e-02 3.66429090e-01
-3.94810855e-01 -6.35519803e-01 -6.99443936e-01 2.88586557e-01
-2.19181970e-01 -1.24925956e-01 1.02759016e+00 -7.76244938e-01
-1.27822888e+00 9.24635381e-02 -4.78312343e-01 -4.34929520e-01
2.84759700e-02 -8.14426914e-02 -9.74424183e-01 4.86280411e-01
-3.25720310e-01 -2.48032480e-01 1.49463248e+00 -9.92180228e-01
-3.97441894e-01 -4.50024428e-03 -4.64935005e-01 -4.60542411e-01
-3.44371527e-01 -2.32921503e-02 7.83186555e-01 -1.51044166e+00
6.74835801e-01 -4.08776641e-01 -2.40591466e-02 -3.67711782e-01
-7.53866732e-02 -1.91023365e-01 1.13818502e+00 -6.32907927e-01
1.26707911e+00 -2.51561666e+00 6.84298947e-02 2.93407053e-01
2.40102801e-02 5.38603589e-02 -3.91201079e-01 8.08398843e-01
-3.69841218e-01 -3.95405054e-01 6.57688593e-03 -8.06833431e-02
-2.94595391e-01 3.00951451e-01 -6.80633724e-01 6.42232776e-01
1.83658659e-01 6.08139753e-01 -7.59861946e-01 2.41778940e-02
-3.52292433e-02 7.60173380e-01 -3.46156508e-01 -3.60405296e-01
1.90065786e-01 6.27691925e-01 -5.42935073e-01 6.91595078e-01
5.13561785e-01 1.80366822e-02 -3.22906971e-01 -4.63510305e-01
-2.29492098e-01 1.08701833e-01 -1.23093355e+00 1.37722945e+00
-5.62284350e-01 7.91680276e-01 2.73015052e-01 -1.31692839e+00
1.34484589e+00 9.34412777e-01 7.01934636e-01 -1.98289782e-01
3.40226181e-02 6.00317240e-01 8.38188455e-02 -4.39085841e-01
-2.66866740e-02 -5.78711867e-01 1.18293919e-01 2.53073812e-01
1.75987795e-01 1.72780007e-01 -5.55156497e-03 -7.30909526e-01
9.78201151e-01 -2.66904503e-01 4.44422752e-01 -8.35325181e-01
6.90444887e-01 -3.39727461e-01 6.06148243e-01 7.81292617e-01
1.29956380e-01 6.41648471e-01 1.18018597e-01 -6.85567737e-01
-9.73319054e-01 -1.12492609e+00 -2.64946312e-01 9.01227951e-01
-4.04903799e-01 -1.52669922e-01 -2.45918438e-01 8.15105885e-02
-4.14860733e-02 6.45720422e-01 -3.32161456e-01 -4.80053015e-02
-8.33974004e-01 -6.89159751e-01 8.45078588e-01 6.42295480e-01
6.44507885e-01 -1.09320045e+00 -7.57057130e-01 5.41120470e-01
-5.62295020e-01 -1.17317986e+00 -3.64497572e-01 2.53214478e-01
-1.07016349e+00 -7.79945314e-01 -8.69354665e-01 -7.39683986e-01
8.58728662e-02 2.45952815e-01 7.90976286e-01 -4.25157905e-01
-8.91098306e-02 3.57477397e-01 -7.51465023e-01 -5.50239682e-01
-4.48824048e-01 -3.77036601e-01 4.32285339e-01 5.03297150e-01
4.54599261e-02 -1.43000555e+00 -4.79117066e-01 4.01030749e-01
-8.45394790e-01 -2.62151778e-01 2.27614179e-01 1.09302640e+00
4.61714089e-01 7.76844442e-01 9.71775115e-01 -5.03484607e-01
7.41026521e-01 -8.36062580e-02 -4.00505602e-01 -1.03913590e-01
-4.72850800e-02 -1.92105368e-01 1.03866649e+00 -9.78278577e-01
-5.42855680e-01 -2.90342838e-01 -2.97009170e-01 -6.31470621e-01
2.79255539e-01 6.87541783e-01 3.66260231e-01 -4.39896941e-01
6.57999873e-01 4.84198093e-01 -1.42724410e-01 -8.06115389e-01
-1.78535283e-01 5.02811253e-01 7.49159217e-01 -5.89569747e-01
1.06966567e+00 6.25477970e-01 4.30041820e-01 -1.28982532e+00
-8.35352898e-01 -6.34841263e-01 -4.56067055e-01 3.99583653e-02
5.28074682e-01 -6.02447331e-01 -9.46538150e-01 2.77325511e-01
-1.33735538e+00 2.63880759e-01 -8.63292634e-01 8.91831994e-01
-8.20424259e-01 2.46827662e-01 -6.09980404e-01 -1.32695520e+00
-3.58794421e-01 -6.71080172e-01 9.12115395e-01 -3.18213463e-01
-2.75256664e-01 -1.24816132e+00 -1.91806257e-01 -2.10984424e-01
8.04628670e-01 6.98853850e-01 1.16625559e+00 -4.97543156e-01
2.43273005e-01 -3.89593095e-01 2.26761356e-01 6.17945850e-01
3.63681883e-01 -4.11447465e-01 -1.13886774e+00 -4.74950880e-01
8.25418830e-01 1.71367675e-01 9.02785778e-01 9.29712057e-01
9.98796523e-01 -3.19771379e-01 2.13735193e-01 5.70548832e-01
1.40784848e+00 3.09025079e-01 4.22345728e-01 2.90638432e-02
1.10256322e-01 2.79341042e-01 1.52598366e-01 4.88622189e-01
-3.63636166e-01 4.26530749e-01 1.40375003e-01 -1.65825546e-01
-1.82841718e-01 6.12367429e-02 5.01838207e-01 1.51618302e+00
-8.03514957e-01 -6.34239540e-02 -5.00235081e-01 2.63162374e-01
-1.70610023e+00 -1.40221107e+00 -2.86375254e-01 2.03191996e+00
4.07258689e-01 7.60541996e-03 7.87938535e-02 1.07330334e+00
7.04765022e-01 4.04582381e-01 -1.44859493e-01 -3.38385850e-01
-5.31488061e-01 8.87603581e-01 1.75703064e-01 3.98055851e-01
-1.05003583e+00 2.31604725e-01 6.91177607e+00 9.73576009e-01
-9.92171168e-01 2.62154669e-01 -5.17709516e-02 2.62664586e-01
1.69789568e-02 -3.90814781e-01 -3.70831311e-01 1.93755403e-01
7.98352718e-01 -3.34720969e-01 5.35453856e-01 3.76496524e-01
4.13985461e-01 4.79074299e-01 -8.72713387e-01 1.21213055e+00
-2.48952821e-01 -1.30712211e+00 5.93890622e-02 -2.28798807e-01
3.84127408e-01 -2.42076173e-01 1.65599883e-01 1.45733386e-01
-1.14764363e-01 -1.20630026e+00 8.64270210e-01 4.99892741e-01
7.24601805e-01 -8.61961722e-01 9.23706293e-01 3.78607184e-01
-1.76670289e+00 -2.19087213e-01 -5.89798808e-01 -1.57672569e-01
9.50674899e-03 8.51608276e-01 -3.78269345e-01 7.99858987e-01
6.69695497e-01 7.55753458e-01 1.66757375e-01 9.45128798e-01
1.39337122e-01 1.05023909e+00 -4.65197355e-01 1.44355670e-01
2.74341404e-01 -1.99056506e-01 1.04472244e+00 1.11618805e+00
8.59384477e-01 3.07015687e-01 3.25603366e-01 5.37558317e-01
2.42512077e-01 -1.73446208e-01 -3.73717159e-01 -1.51487321e-01
1.92784593e-01 1.01741064e+00 -7.29576409e-01 6.91221561e-04
-2.95753360e-01 5.19744813e-01 -7.63947666e-01 8.87226820e-01
-3.34813952e-01 -4.57529277e-01 7.45249569e-01 2.51834989e-01
7.80085862e-01 -5.09859204e-01 -1.14737693e-02 -9.04802978e-01
2.15160362e-02 -7.45051146e-01 3.51358950e-01 -7.54521906e-01
-1.65145433e+00 8.90791416e-01 -4.17051539e-02 -1.82572126e+00
-1.51946917e-02 -6.48851693e-01 -6.28516972e-01 9.23540413e-01
-1.50508869e+00 -9.91537333e-01 -1.42132519e-02 7.93580234e-01
6.17975056e-01 -2.18046471e-01 1.24376786e+00 9.94354486e-02
2.01597020e-01 2.93800533e-01 4.92132992e-01 1.47361904e-01
1.20519280e-01 -9.23063219e-01 1.96494564e-01 5.78438699e-01
2.58053273e-01 3.98878694e-01 1.19301856e+00 -3.53111386e-01
-1.33502662e+00 -9.56516266e-01 6.75871849e-01 1.12299711e-01
8.28543901e-01 -4.49866503e-01 -9.18443203e-01 4.33975309e-01
-3.65461297e-02 3.64201367e-01 8.61859858e-01 8.81536752e-02
-5.71616828e-01 -2.83201814e-01 -1.08509398e+00 2.17312619e-01
8.42997313e-01 -4.31932837e-01 -8.66650045e-01 3.40885133e-01
6.29138350e-01 3.80952507e-02 -1.04972279e+00 5.97127020e-01
8.13326955e-01 -9.44500327e-01 1.42263782e+00 -6.74924612e-01
2.36671895e-01 -3.09044629e-01 -8.05252790e-01 -1.50008309e+00
-7.77636409e-01 -9.96332526e-01 -1.37187228e-01 6.84763670e-01
2.37063151e-02 -1.02906001e+00 4.33435559e-01 -9.10314143e-01
-1.24971732e-01 -5.99040389e-01 -1.60620904e+00 -1.34473383e+00
-2.67583400e-01 -6.23203695e-01 4.82630551e-01 9.20352399e-01
-1.39280006e-01 1.37484252e-01 -3.50123733e-01 5.02613187e-01
8.96839559e-01 3.49466652e-01 5.16268946e-02 -1.78594935e+00
-4.71513629e-01 -3.51450294e-01 -6.95822895e-01 -6.34380937e-01
-2.05377504e-01 -8.25223207e-01 -5.04461229e-02 -1.17115510e+00
-4.74600554e-01 -4.38548364e-02 -6.85152709e-01 2.30586246e-01
4.05198663e-01 6.10657990e-01 1.09105520e-01 3.14345241e-01
1.82507142e-01 5.75478196e-01 1.42936456e+00 9.89215299e-02
-1.44435391e-01 4.87404823e-01 -1.20918870e-01 1.06862640e+00
3.93743455e-01 -4.36502397e-01 -3.08715940e-01 -6.39322260e-03
1.56918198e-01 3.66352588e-01 5.06015539e-01 -1.35613406e+00
1.08179495e-01 1.59127012e-01 6.52740657e-01 -7.53709197e-01
7.42283762e-01 -8.78719509e-01 4.37234461e-01 5.72172642e-01
5.03717130e-03 1.86159536e-01 8.29256177e-02 6.63451314e-01
-6.78326190e-01 8.16804320e-02 9.20759141e-01 1.58515811e-01
-8.90024006e-02 -3.16260234e-02 -5.04471719e-01 -2.65444964e-01
3.98399323e-01 -3.69476229e-01 -1.06291384e-01 -6.83395207e-01
-1.01173413e+00 -4.14405704e-01 -3.22465718e-01 -6.87828362e-02
9.17445183e-01 -1.61987376e+00 -1.00293541e+00 4.94032651e-01
-1.53440714e-01 -4.05283958e-01 1.77077293e-01 1.26456368e+00
-3.92523617e-01 3.05406600e-01 -2.51233250e-01 -5.70879579e-01
-1.22188115e+00 1.63744330e-01 1.65132746e-01 -2.42992952e-01
-1.18209767e+00 8.07311654e-01 2.25394055e-01 1.84854999e-01
2.42258538e-03 -6.85545027e-01 -3.64329070e-01 1.09185375e-01
6.69164777e-01 5.80299437e-01 8.10286924e-02 -6.58711374e-01
-3.23904574e-01 6.96605802e-01 4.96837139e-01 -2.14130044e-01
1.77488387e+00 3.33830476e-01 -3.61829817e-01 8.72753561e-01
9.20001984e-01 2.92371005e-01 -6.92021132e-01 -4.53212053e-01
3.51990312e-02 -2.52731740e-01 3.64697836e-02 -2.48857409e-01
-8.07946384e-01 8.69909227e-01 3.51705819e-01 9.25676405e-01
1.63338888e+00 -2.05652431e-01 1.03753138e+00 4.40641671e-01
5.73476851e-01 -5.31677425e-01 1.17265657e-01 5.60898304e-01
1.25185597e+00 -3.19558352e-01 -6.97615668e-02 -5.08013070e-01
2.14606952e-02 1.69737053e+00 -6.82897806e-01 -5.63755453e-01
9.52429831e-01 4.56644088e-01 3.62391211e-02 -7.09453300e-02
-4.09063965e-01 -6.21369854e-02 5.69405377e-01 5.79060972e-01
3.08173895e-01 6.44921809e-02 -1.49519950e-01 9.42588449e-01
-7.21088707e-01 -1.88217208e-01 4.05667812e-01 7.09531367e-01
-5.72205186e-01 -8.20678949e-01 -8.85698736e-01 4.69966918e-01
-5.38508832e-01 -4.49806117e-02 1.84952006e-01 5.97232997e-01
-3.50040756e-02 1.08858943e+00 -1.39593169e-01 -3.91903192e-01
5.28018773e-01 3.68214875e-01 4.98043209e-01 -4.10813332e-01
-4.40045446e-01 6.89391732e-01 -5.13756461e-02 -1.66273579e-01
-7.08421350e-01 -6.17686987e-01 -9.47166800e-01 -2.31367871e-01
-2.56522357e-01 3.97837788e-01 5.08815765e-01 7.83495843e-01
-1.35706782e-01 7.84314632e-01 8.57381761e-01 -8.87354434e-01
-7.04371929e-01 -1.19155324e+00 -1.08432508e+00 9.56888795e-02
8.87710392e-01 -6.98744655e-01 -5.83973169e-01 1.99323043e-01] | [15.373628616333008, 5.561366081237793] |
409d6de6-e0a0-4401-a49b-d251f9c5b566 | masked-representation-learning-for-domain | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Rao_Masked_Representation_Learning_for_Domain_Generalized_Stereo_Matching_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Rao_Masked_Representation_Learning_for_Domain_Generalized_Stereo_Matching_CVPR_2023_paper.pdf | Masked Representation Learning for Domain Generalized Stereo Matching | Recently, many deep stereo matching methods have begun to focus on cross-domain performance, achieving impressive achievements. However, these methods did not deal with the significant volatility of generalization performance among different training epochs. Inspired by masked representation learning and multi-task learning, this paper designs a simple and effective masked representation for domain generalized stereo matching. First, we feed the masked left and complete right images as input into the models. Then, we add a lightweight and simple decoder following the feature extraction module to recover the original left image. Finally, we train the models with two tasks (stereo matching and image reconstruction) as a pseudo-multi-task learning framework, promoting models to learn structure information and to improve generalization performance. We implement our method on two well-known architectures (CFNet and LacGwcNet) to demonstrate its effectiveness. Experimental results on multi-datasets show that: (1) our method can be easily plugged into the current various stereo matching models to improve generalization performance; (2) our method can reduce the significant volatility of generalization performance among different training epochs; (3) we find that the current methods prefer to choose the best results among different training epochs as generalization performance, but it is impossible to select the best performance by ground truth in practice. | ['Xing Li', 'Zhelun Shen', 'Renjie He', 'Yuchao Dai', 'Mingyi He', 'Bangshu Xiong', 'Zhibo Rao'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['image-reconstruction', 'stereo-matching-1'] | ['computer-vision', 'computer-vision'] | [ 2.11894140e-01 -2.83768535e-01 -2.54739691e-02 -5.02729714e-01
-9.69127834e-01 -2.91729033e-01 4.93427157e-01 -6.11945152e-01
-4.37427133e-01 4.39947158e-01 3.22599262e-01 -4.78037857e-02
-1.41486358e-02 -8.28739524e-01 -7.93746173e-01 -5.95722020e-01
3.01298380e-01 2.43571654e-01 6.01094246e-01 -2.75142014e-01
1.16438955e-01 2.03219205e-01 -1.72803497e+00 7.19317138e-01
1.19858706e+00 1.12682676e+00 5.95914960e-01 1.91955715e-01
-3.51943374e-02 7.51233816e-01 -3.50112140e-01 -4.41297114e-01
6.37122750e-01 -1.97347298e-01 -8.43692541e-01 1.46802053e-01
6.96470141e-01 -4.35175270e-01 -4.29941952e-01 1.08900714e+00
6.80500507e-01 -1.17539756e-01 3.46993446e-01 -1.07045889e+00
-7.81500936e-01 3.45011979e-01 -5.35181165e-01 1.39408084e-02
1.82455122e-01 1.77761510e-01 6.92637384e-01 -9.71683621e-01
4.85058367e-01 1.32776141e+00 7.38088548e-01 7.33747900e-01
-1.16235018e+00 -1.14322174e+00 2.39872739e-01 2.69142061e-01
-1.42106807e+00 -4.89023894e-01 6.68402195e-01 -3.88828784e-01
7.34209538e-01 9.09628440e-03 3.75167280e-01 1.37847662e+00
1.20182008e-01 8.70715439e-01 1.28291070e+00 -2.42638867e-02
-3.14662829e-02 1.17532909e-01 -1.39663249e-01 5.48818052e-01
1.72973182e-02 4.84447300e-01 -5.42329073e-01 -1.48299206e-02
8.58506024e-01 9.17004794e-02 -3.37748170e-01 -4.66793656e-01
-1.34546661e+00 7.29310632e-01 5.85830510e-01 3.35936606e-01
-1.11746125e-01 2.51025353e-02 4.09159124e-01 6.45798743e-01
4.31300610e-01 9.62539837e-02 -4.05806243e-01 3.08615148e-01
-8.59530687e-01 4.42002118e-01 2.97491401e-01 1.27480960e+00
1.11017120e+00 -1.06756724e-01 -2.31471643e-01 1.21836066e+00
5.55630811e-02 3.06069493e-01 9.67882097e-01 -7.19988465e-01
8.15475523e-01 7.42145717e-01 -1.33198932e-01 -8.17233205e-01
-4.53976661e-01 -4.71469015e-01 -1.14325523e+00 2.06667855e-01
2.99582869e-01 6.12722673e-02 -8.65177453e-01 1.89439821e+00
-1.12942502e-01 2.21227989e-01 1.08638547e-01 1.05805087e+00
1.05270398e+00 3.13088298e-01 1.48278624e-02 3.08042228e-01
1.20774376e+00 -1.09638917e+00 -2.53286213e-01 -7.27201700e-01
6.78626955e-01 -7.71129906e-01 1.15743577e+00 1.90058708e-01
-1.02288699e+00 -9.80890274e-01 -1.23402584e+00 -9.66979265e-02
-3.40166003e-01 8.51483271e-02 6.45924866e-01 4.42543507e-01
-1.01026714e+00 7.18552947e-01 -4.67645079e-01 -2.58426219e-01
4.45931345e-01 3.58799130e-01 -5.16806901e-01 -3.46251518e-01
-1.30024469e+00 9.22336698e-01 5.23675203e-01 -1.08424602e-02
-8.59094858e-01 -7.06651926e-01 -8.82104158e-01 -1.84103683e-01
1.97748721e-01 -1.00073230e+00 1.12284458e+00 -1.14545274e+00
-1.34653699e+00 1.22440803e+00 -1.30361199e-01 -2.20098868e-01
7.63882697e-01 -2.47069925e-01 -5.03682077e-01 -2.60986477e-01
3.62924367e-01 9.75781322e-01 8.57207358e-01 -1.16218090e+00
-8.20781887e-01 -4.56904888e-01 4.82638553e-02 2.55087882e-01
-3.45874548e-01 -1.69018343e-01 -6.89600706e-01 -7.80798078e-01
6.65060222e-01 -9.36413109e-01 -2.63010621e-01 -2.25757044e-02
-2.13725597e-01 5.13176173e-02 5.65765142e-01 -4.80655819e-01
1.04929960e+00 -2.26830077e+00 2.20669672e-01 -5.26462793e-02
1.04784146e-01 1.05897285e-01 -3.45463723e-01 2.00311989e-01
-2.67285824e-01 -2.57560104e-01 -3.04144233e-01 -3.59319478e-01
-1.70250744e-01 1.66822165e-01 -5.05699515e-01 4.26202476e-01
6.74002543e-02 9.98107672e-01 -7.27624178e-01 -4.52173680e-01
1.73225373e-01 1.25790074e-01 -7.20284045e-01 3.45669746e-01
2.17270013e-02 5.75869799e-01 -3.47454399e-01 5.94375789e-01
1.04679549e+00 -3.30704242e-01 3.15471366e-02 -4.01127368e-01
-4.63337004e-02 1.37336150e-01 -1.48895276e+00 2.16397524e+00
-6.16146922e-01 3.59520942e-01 -1.30348772e-01 -9.42635417e-01
1.03901005e+00 6.54235855e-02 3.55572641e-01 -1.19994235e+00
-1.69521451e-01 4.47440654e-01 -1.69677824e-01 -4.33275402e-01
5.72037518e-01 -1.17895357e-01 -1.40384138e-01 3.29925597e-01
4.31371219e-02 -8.75983611e-02 -1.61022767e-01 -1.19836137e-01
8.14847350e-01 3.52967381e-01 2.05313474e-01 -4.48115975e-01
7.29414105e-01 -1.60849184e-01 9.04001296e-01 5.64935327e-01
-1.72405601e-01 9.56620574e-01 -6.26150565e-03 -6.24776423e-01
-1.06864238e+00 -1.04896033e+00 -1.30554944e-01 1.31638253e+00
4.67750341e-01 -1.62350461e-01 -4.48249668e-01 -4.61533844e-01
1.18627027e-01 1.24995530e-01 -4.63184655e-01 -2.59854496e-01
-8.19508970e-01 -6.80354416e-01 5.83003283e-01 6.46026373e-01
1.18196476e+00 -9.51945603e-01 -2.53647923e-01 1.60681695e-01
-4.35197562e-01 -1.21970356e+00 -6.15730822e-01 3.98447812e-02
-1.05100131e+00 -9.83035803e-01 -9.19690788e-01 -1.18554378e+00
5.00961602e-01 6.92470074e-01 9.53070164e-01 -5.49923144e-02
1.49475019e-02 -3.82101461e-02 -6.51365817e-02 -6.80144206e-02
-3.14966887e-01 2.88301766e-01 -5.38543686e-02 9.60282236e-02
2.70717859e-01 -7.78766036e-01 -7.58529246e-01 8.55350256e-01
-8.99000347e-01 3.91201168e-01 7.25739956e-01 9.39519465e-01
5.08285880e-01 -3.38877738e-01 4.54103410e-01 -9.06395912e-01
5.17497718e-01 -1.61170825e-01 -6.72360361e-01 2.91087836e-01
-6.67177737e-01 2.53929526e-01 4.92928475e-01 -4.62325990e-01
-1.09864914e+00 1.36754364e-01 -8.27516168e-02 -6.39507949e-01
-2.48902980e-02 1.79530635e-01 -2.00312719e-01 -3.04534525e-01
7.25689530e-01 4.80664045e-01 7.54713081e-03 -7.38054335e-01
3.15709025e-01 7.20215619e-01 7.10758030e-01 -6.53372228e-01
8.93415034e-01 5.87874711e-01 -2.88754523e-01 -2.40711465e-01
-9.66297805e-01 -4.56401736e-01 -6.79559708e-01 -2.07487941e-02
6.59858048e-01 -1.48363841e+00 -5.46839952e-01 6.98985875e-01
-1.09182656e+00 -2.84553558e-01 2.09499955e-01 4.13453102e-01
-7.58134663e-01 6.06851161e-01 -5.32804370e-01 -1.85144067e-01
-3.29889745e-01 -1.31252086e+00 1.29319012e+00 9.23701599e-02
1.04461879e-01 -9.31525230e-01 5.49763115e-03 3.94410282e-01
6.11833155e-01 -7.42804036e-02 8.33777905e-01 -3.08174431e-01
-8.67642105e-01 1.18155755e-01 -6.10791624e-01 3.47122520e-01
7.18061551e-02 -5.05215406e-01 -1.30775368e+00 -5.24283707e-01
1.24282077e-01 -4.91053075e-01 1.09056747e+00 2.34856322e-01
1.55016804e+00 7.55319297e-02 -4.62761730e-01 1.29368496e+00
1.47140563e+00 -2.29843054e-02 9.40694034e-01 7.09189832e-01
6.18281126e-01 5.54175377e-01 4.97841448e-01 5.85609674e-02
5.86685777e-01 9.18024421e-01 3.60370904e-01 -2.44353533e-01
-4.38158691e-01 -4.57852930e-01 3.53261262e-01 7.88716495e-01
1.70825962e-02 3.54724497e-01 -7.84701645e-01 3.84577602e-01
-2.09403586e+00 -1.15317476e+00 1.77886665e-01 2.30980539e+00
6.75185978e-01 2.17248261e-01 -5.91132697e-03 -1.71744570e-01
8.83137643e-01 3.02424133e-01 -5.90728462e-01 9.85908061e-02
-4.73097563e-01 3.89626995e-02 6.14047050e-01 2.92644799e-01
-1.30193543e+00 1.02968967e+00 6.41470003e+00 9.33337271e-01
-1.20898747e+00 1.08509004e-01 4.22233880e-01 -1.43279033e-02
-2.50365704e-01 6.66596144e-02 -8.94499898e-01 4.79646653e-01
2.97721893e-01 -2.29925260e-01 4.00297612e-01 9.61656213e-01
-1.44323409e-01 1.95990235e-01 -1.27739084e+00 1.40527213e+00
-7.86263403e-03 -1.32084966e+00 2.80878544e-01 -7.47277141e-02
7.69210100e-01 1.39707595e-01 7.46140629e-02 5.71567655e-01
1.04982622e-01 -9.44373846e-01 8.05313289e-01 2.84696311e-01
9.61127996e-01 -3.33011270e-01 4.97719437e-01 4.62976187e-01
-1.33777380e+00 -3.38019818e-01 -7.11640120e-01 3.06170210e-02
-5.09855673e-02 3.03687662e-01 -2.62188792e-01 7.74362862e-01
7.78839171e-01 8.69542360e-01 -7.37668157e-01 1.20316684e+00
-2.89243385e-02 -1.14696525e-01 -6.71549421e-03 3.32808465e-01
2.55057126e-01 -3.06290150e-01 2.51869500e-01 1.11533844e+00
3.42970014e-01 -3.15155685e-01 1.44959182e-01 7.73215473e-01
-1.36447269e-02 1.28377974e-02 -7.92039454e-01 6.34496629e-01
3.63628685e-01 9.63829279e-01 -2.72093326e-01 -1.90167949e-01
-8.69344532e-01 1.09813929e+00 5.92216730e-01 4.11939323e-01
-9.00905609e-01 -1.22901276e-01 8.31332922e-01 5.74025773e-02
2.59027630e-01 3.48152826e-04 -5.80130875e-01 -1.43151140e+00
2.05619887e-01 -9.97299135e-01 5.84912896e-01 -7.00880110e-01
-1.48930228e+00 7.44309962e-01 -8.43415037e-02 -1.77750671e+00
-1.70727000e-01 -5.63153923e-01 -4.81520355e-01 1.11354983e+00
-1.71807051e+00 -1.23432016e+00 -6.28832877e-01 9.30447817e-01
6.16143882e-01 -4.88979846e-01 7.12319016e-01 7.68142164e-01
-5.18253803e-01 7.77516425e-01 -3.50611582e-02 1.43121600e-01
1.03439856e+00 -7.73132980e-01 5.80120027e-01 7.28381276e-01
-1.59523576e-01 4.49238658e-01 2.37391546e-01 -4.11590040e-01
-1.14388454e+00 -1.22599840e+00 7.71847010e-01 -3.38616699e-01
4.05315250e-01 -4.33667004e-01 -1.05921066e+00 8.18807662e-01
1.02234026e-02 -2.18221378e-02 4.04186845e-01 1.39866859e-01
-6.79184735e-01 -4.37255204e-01 -8.72200310e-01 7.13269114e-01
1.52888572e+00 -6.88827336e-01 -7.55250216e-01 3.73514026e-01
7.04646468e-01 -7.51716256e-01 -6.35480165e-01 7.67681062e-01
6.19121432e-01 -1.32351720e+00 1.04836857e+00 -5.00274599e-01
4.19908345e-01 -2.03167349e-01 -3.56632888e-01 -1.21492088e+00
-5.49386084e-01 -3.89829546e-01 3.87512296e-01 1.01143610e+00
4.09247488e-01 -8.66850138e-01 8.52699697e-01 5.88911235e-01
-3.60991955e-01 -6.78575933e-01 -1.06985998e+00 -9.75675166e-01
2.03648880e-01 -3.92645568e-01 8.98385823e-01 9.23176944e-01
-3.16233844e-01 2.95336664e-01 -5.85361719e-01 1.26913875e-01
7.25029826e-01 6.26840532e-01 9.94890153e-01 -1.12580824e+00
-3.01421106e-01 -6.03155553e-01 -3.16292644e-01 -1.51614976e+00
9.00685042e-02 -1.04270041e+00 -1.48862496e-01 -1.19861960e+00
5.00465155e-01 -5.73778689e-01 -2.46780664e-01 4.11958456e-01
-2.51124233e-01 8.05497691e-02 1.44443735e-01 5.12567639e-01
-3.96450073e-01 5.43642342e-01 1.47392416e+00 -1.47261709e-01
-8.05466920e-02 2.36788973e-01 -7.92097211e-01 5.66782236e-01
4.97692704e-01 -3.54617715e-01 -3.81116867e-01 -8.74216795e-01
-2.14858185e-02 4.24852967e-02 5.16857982e-01 -1.19462705e+00
3.39024425e-01 -6.40585646e-02 3.87574792e-01 -5.27104855e-01
2.44909465e-01 -8.27436686e-01 2.85423368e-01 3.95073056e-01
-1.42214760e-01 1.07565477e-01 2.48612285e-01 3.10774297e-01
-4.52443570e-01 -8.73988215e-03 9.37682152e-01 -1.72761470e-01
-1.23819482e+00 6.50592983e-01 1.88830987e-01 7.83176646e-02
8.95307004e-01 -5.27551472e-01 -4.16951150e-01 -1.54279038e-01
-6.15684152e-01 4.36985523e-01 5.40040433e-01 8.10056746e-01
6.50328159e-01 -1.64957535e+00 -7.09106565e-01 5.30359328e-01
4.34767932e-01 9.38569307e-02 4.94816124e-01 5.08031368e-01
-3.23248714e-01 4.04274017e-01 -4.75816965e-01 -8.53509426e-01
-1.11036313e+00 7.04614282e-01 6.62218392e-01 -1.44202381e-01
-6.95654273e-01 7.83185780e-01 6.50547504e-01 -6.15429699e-01
3.37973028e-01 -2.37040699e-01 1.56487282e-02 2.13089176e-02
3.47312063e-01 2.04942256e-01 1.09875739e-01 -3.88646811e-01
-3.64716858e-01 1.04136467e+00 -1.24527164e-01 1.39670372e-01
1.23873222e+00 -1.00120276e-01 1.25669986e-01 9.18769389e-02
1.43273878e+00 -3.63822371e-01 -1.31445289e+00 -6.52922273e-01
-4.08216752e-02 -6.70571089e-01 -1.21623695e-01 -4.80576634e-01
-1.15580285e+00 9.14325535e-01 7.92780399e-01 -1.76893771e-01
1.39156079e+00 -5.70276789e-02 8.70695353e-01 2.35210732e-01
5.99666238e-01 -1.10277200e+00 1.46172211e-01 4.28122014e-01
9.45307493e-01 -1.51577771e+00 -1.13852128e-01 -5.41092753e-01
-7.72612035e-01 1.04935002e+00 1.09141469e+00 -1.33598030e-01
6.17897093e-01 4.15527187e-02 2.91911334e-01 -1.39926389e-01
-7.12183118e-01 -4.25986528e-01 3.59480381e-01 5.85600138e-01
3.28727871e-01 -1.72414109e-01 -1.02327198e-01 5.20528853e-01
-1.68475285e-01 1.15756407e-01 -6.66833576e-03 6.28870606e-01
-3.96240234e-01 -1.22108531e+00 -2.18601868e-01 2.29948103e-01
-8.38087350e-02 -1.06882870e-01 -1.16023652e-01 7.78927982e-01
2.26235285e-01 6.30976081e-01 -1.38257081e-02 -7.22234368e-01
7.99527168e-01 -6.96453154e-02 4.50651735e-01 -5.85110724e-01
-5.49617112e-01 -1.71200112e-01 -1.12913273e-01 -8.88477683e-01
-3.92877191e-01 -4.07594085e-01 -7.74882376e-01 -3.48108977e-01
-1.42697860e-02 -2.59281069e-01 2.75870949e-01 8.45378458e-01
4.37311977e-01 3.53725106e-01 6.86083853e-01 -8.10085118e-01
-7.49866962e-01 -9.28173780e-01 -4.26702768e-01 7.02378929e-01
1.63023055e-01 -8.30347598e-01 -2.15992376e-01 -2.02971026e-01] | [8.738137245178223, -2.247398853302002] |
b03d31ce-27ae-4164-abe8-af8ce168f333 | companykg-a-large-scale-heterogeneous-graph | 2306.10649 | null | https://arxiv.org/abs/2306.10649v1 | https://arxiv.org/pdf/2306.10649v1.pdf | CompanyKG: A Large-Scale Heterogeneous Graph for Company Similarity Quantification | In the investment industry, it is often essential to carry out fine-grained company similarity quantification for a range of purposes, including market mapping, competitor analysis, and mergers and acquisitions. We propose and publish a knowledge graph, named CompanyKG, to represent and learn diverse company features and relations. Specifically, 1.17 million companies are represented as nodes enriched with company description embeddings; and 15 different inter-company relations result in 51.06 million weighted edges. To enable a comprehensive assessment of methods for company similarity quantification, we have devised and compiled three evaluation tasks with annotated test sets: similarity prediction, competitor retrieval and similarity ranking. We present extensive benchmarking results for 11 reproducible predictive methods categorized into three groups: node-only, edge-only, and node+edge. To the best of our knowledge, CompanyKG is the first large-scale heterogeneous graph dataset originating from a real-world investment platform, tailored for quantifying inter-company similarity. | ['Dhiana Deva Cavacanti Rocha', 'Armin Catovic', 'Andrew McCornack', 'Richard Anselmo Stahl', 'Mark Granroth-Wilding', 'Vilhelm von Ehrenheim', 'Lele Cao'] | 2023-06-18 | null | null | null | null | ['benchmarking', 'benchmarking'] | ['miscellaneous', 'robots'] | [-2.19351128e-01 3.03567909e-02 -3.89681906e-01 -7.53336996e-02
-5.48661530e-01 -8.94251883e-01 1.00807929e+00 8.29115450e-01
-2.00160444e-01 5.44748962e-01 4.14303243e-01 -8.32166821e-02
-4.33549702e-01 -1.05901337e+00 -2.29713678e-01 1.08848594e-01
-4.44111452e-02 1.00579500e+00 5.83327226e-02 -4.92348969e-01
5.09872377e-01 5.79834163e-01 -1.19498932e+00 4.89055306e-01
5.16455233e-01 1.34330213e+00 -1.35020956e-01 2.23886341e-01
-4.27475125e-01 1.15516353e+00 -5.31195104e-01 -1.33411872e+00
3.35367888e-01 1.39177337e-01 -9.57394123e-01 -2.02814594e-01
2.10105300e-01 5.49448252e-01 -2.80867964e-01 9.90724087e-01
5.27449369e-01 -8.17784443e-02 6.16813838e-01 -1.30451226e+00
-7.60918558e-01 1.19278979e+00 -4.90335464e-01 5.00616968e-01
4.59214002e-01 -3.23663428e-02 1.89156461e+00 -8.90116096e-01
1.26586509e+00 8.02758336e-01 9.59335387e-01 -1.29341632e-01
-1.00345540e+00 -5.81032634e-01 -1.86138511e-01 -1.51428999e-03
-1.35957563e+00 7.11092427e-02 1.04603446e+00 -7.71623075e-01
1.22603440e+00 2.86549777e-01 8.67879748e-01 1.15096498e+00
2.04166636e-01 3.41077268e-01 8.76977324e-01 1.04226414e-02
-1.83778673e-01 4.82353449e-01 1.63181871e-01 4.42844361e-01
7.36988723e-01 -1.48168847e-01 -5.10060430e-01 -2.23652184e-01
6.53969347e-01 -3.14900167e-02 -1.50373742e-01 -5.92879474e-01
-1.50293863e+00 1.02656925e+00 6.85508072e-01 8.03622425e-01
-5.49577296e-01 1.93806410e-01 8.92430246e-01 5.37354052e-01
6.10282421e-01 1.40719509e+00 -3.93696815e-01 -4.10319865e-01
-6.72129631e-01 1.54724389e-01 1.08371675e+00 1.14283693e+00
5.98803222e-01 -1.56406924e-01 4.13759425e-02 6.96585834e-01
-2.08635166e-01 -1.57355934e-01 5.79110861e-01 -7.63188243e-01
8.54873180e-01 8.97966802e-01 -2.69004852e-01 -1.55624211e+00
-3.46482277e-01 -1.15780687e+00 -6.18077219e-01 -4.42149669e-01
-4.07403633e-02 4.43118840e-01 -3.96675579e-02 9.91411567e-01
-2.32085779e-01 2.64551967e-01 4.05718051e-02 5.17012656e-01
1.45896184e+00 -3.36937197e-02 -3.33241284e-01 1.08694285e-01
1.36600435e+00 -1.21027815e+00 -4.32205647e-01 3.60346101e-02
8.26790690e-01 -8.31956863e-01 9.34123516e-01 3.85062173e-02
-1.04884589e+00 -5.27481616e-01 -8.59468281e-01 2.72887170e-01
-8.44014764e-01 -3.89138281e-01 1.05566275e+00 4.54906762e-01
-7.24479496e-01 8.95994484e-01 -1.36850685e-01 -1.66857779e-01
5.84755540e-01 3.05372626e-01 -6.88489497e-01 2.84727011e-02
-1.42207789e+00 1.09191763e+00 3.49221915e-01 -3.57456386e-01
-3.89371812e-01 -9.19787765e-01 -5.10681152e-01 2.48103261e-01
3.48432124e-01 -9.01662886e-01 7.53081262e-01 -6.24848366e-01
-8.24069381e-01 1.22669733e+00 7.66062796e-01 -5.57603061e-01
3.10468942e-01 1.81380421e-01 -9.78022456e-01 -2.01701701e-01
2.91877687e-01 3.62570882e-02 2.19653487e-01 -1.03946006e+00
-5.05390048e-01 -1.86100110e-01 1.01306088e-01 -1.20680608e-01
-5.40154397e-01 9.37343538e-02 -5.49184203e-01 -7.95558274e-01
-2.41521344e-01 -6.03247106e-01 -1.66916177e-01 -1.05274808e+00
-4.81792957e-01 -2.50081897e-01 8.35806131e-02 -5.27279437e-01
1.61153507e+00 -1.85545123e+00 3.22327048e-01 3.89652342e-01
7.82980859e-01 -9.37323123e-02 -4.65426147e-01 8.60309243e-01
-3.07269186e-01 4.08161312e-01 1.83590278e-01 -1.30717471e-01
2.04140842e-01 -1.70038968e-01 -1.89744495e-02 1.52277425e-01
2.43489057e-01 1.53285491e+00 -1.01367152e+00 -4.07097220e-01
2.93980762e-02 2.81567663e-01 -2.50862151e-01 -2.91026384e-01
2.96602011e-01 -5.28247422e-03 -5.15780807e-01 1.06702280e+00
2.07034424e-01 -4.96229529e-01 3.10195446e-01 -5.44813037e-01
2.06310406e-01 2.47242987e-01 -8.03447068e-01 1.52631772e+00
-7.11858153e-01 6.72902524e-01 -3.50965410e-01 -9.33267713e-01
1.20306480e+00 -1.32859722e-01 7.59159565e-01 -6.35506272e-01
4.02472943e-01 6.55094981e-01 1.20955624e-01 5.00873886e-02
9.02334154e-01 2.54857540e-01 -5.88802755e-01 4.64083888e-02
9.09731835e-02 -4.22865331e-01 4.50917989e-01 4.11221892e-01
1.54327524e+00 -5.88456154e-01 5.60030103e-01 -2.01092988e-01
6.54431462e-01 6.16665669e-02 3.26447725e-01 4.47925985e-01
-4.89707245e-03 4.43487525e-01 8.44081700e-01 -6.84625924e-01
-8.82939816e-01 -9.68338966e-01 9.52543989e-02 5.11132121e-01
-1.40949646e-02 -8.24446738e-01 -1.90081626e-01 -8.40090454e-01
6.54477835e-01 4.24063951e-01 -5.31172752e-01 -1.11568980e-01
-3.42357308e-01 -2.21343473e-01 3.51364166e-01 6.01824343e-01
2.83311665e-01 -9.91696179e-01 1.58750758e-01 4.08328861e-01
1.37160212e-01 -1.18106782e+00 -4.66455042e-01 6.23525642e-02
-5.54259121e-01 -1.32625151e+00 -2.40055129e-01 -8.03005099e-01
1.42621771e-01 -1.31356597e-01 2.13466597e+00 -2.23046288e-01
-4.96782124e-01 4.32073623e-01 -3.02148461e-01 7.99774379e-02
-2.13364348e-01 5.34050405e-01 -1.31354302e-01 -2.84705728e-01
3.41415018e-01 -4.35287118e-01 -2.84031332e-01 2.66751081e-01
-3.80515546e-01 -5.99302828e-01 7.35011876e-01 9.44160044e-01
7.45901942e-01 5.13493299e-01 6.42004669e-01 -1.24683642e+00
1.30626225e+00 -6.28173769e-01 -5.06106853e-01 3.33106250e-01
-1.00221753e+00 -1.64497197e-01 4.25391644e-01 -1.59543142e-01
-3.76183867e-01 -3.92564118e-01 1.27268448e-01 -4.25766945e-01
5.16350508e-01 9.69821334e-01 -2.39653643e-02 -3.39510053e-01
5.37128031e-01 -1.33474961e-01 -2.86253750e-01 -3.71954471e-01
7.60740697e-01 6.37616873e-01 4.56306100e-01 -3.69567424e-01
9.23140585e-01 3.94840911e-02 4.38220650e-02 -1.42291188e-01
-6.36245370e-01 -7.59815633e-01 -4.66205686e-01 -3.16564292e-02
5.70913136e-01 -8.88500035e-01 -7.29241967e-01 -9.58007202e-02
-9.05984282e-01 2.86107033e-01 -6.89397514e-01 5.27162373e-01
-3.49892735e-01 1.39098734e-01 -5.87592542e-01 -1.38613671e-01
-4.38918680e-01 -9.78752732e-01 6.78125620e-01 -2.40592033e-01
-5.65746129e-01 -1.31827986e+00 2.34676659e-01 7.05068469e-01
6.38812482e-01 5.13617635e-01 7.16253638e-01 -1.16546452e+00
-5.74221790e-01 -7.31058896e-01 -3.51971596e-01 3.75457972e-01
3.60180646e-01 -8.36615041e-02 -4.17503327e-01 -2.20933691e-01
-6.63377225e-01 8.31908919e-03 1.08201587e+00 8.18117755e-04
9.85484481e-01 1.82927996e-02 -4.31410789e-01 4.37132835e-01
1.38093114e+00 -3.27743702e-02 4.06779975e-01 7.97019422e-01
9.42384720e-01 7.09509254e-01 7.86386549e-01 5.25371552e-01
4.19700891e-01 7.18853295e-01 4.99463707e-01 2.12971255e-01
2.70160157e-02 -1.35616899e-01 -2.52331793e-01 1.24737537e+00
-4.88356978e-01 -3.96396369e-01 -1.05605137e+00 5.32175422e-01
-1.69550204e+00 -1.11061323e+00 -2.62507021e-01 2.15926909e+00
4.20191109e-01 5.46968400e-01 1.29937291e-01 -4.61057909e-02
7.50447989e-01 3.50663185e-01 -3.88999917e-02 -3.23299646e-01
-4.33908194e-01 4.33550209e-01 7.77445197e-01 1.37227729e-01
-1.03425288e+00 9.96606767e-01 6.16374540e+00 1.03274477e+00
-6.49336457e-01 1.82514533e-01 4.61329967e-01 -1.87150940e-01
-9.85157251e-01 1.99310984e-02 -4.92237628e-01 4.74786878e-01
6.94569707e-01 -5.21489799e-01 3.74245048e-01 1.07368970e+00
-7.24908292e-01 7.14999378e-01 -1.03555501e+00 1.12020397e+00
-8.88062716e-02 -2.10503674e+00 -3.38002741e-02 1.67119324e-01
9.57855403e-01 1.29185483e-01 1.63604513e-01 5.62720895e-01
3.67481649e-01 -1.07675743e+00 3.87071311e-01 4.63640928e-01
5.61324894e-01 -1.11990798e+00 1.22733796e+00 -3.41758132e-01
-1.72120225e+00 -1.45120889e-01 -1.79295361e-01 1.74292564e-01
1.40245751e-01 8.57688367e-01 -8.00084114e-01 1.12858272e+00
6.28393352e-01 1.32400787e+00 -7.11339474e-01 6.26512110e-01
1.82476908e-01 5.92314787e-02 2.57713437e-01 -7.35996887e-02
2.56626070e-01 -4.00965989e-01 4.11618948e-01 1.06060910e+00
4.96089995e-01 -7.10647285e-01 -5.63860387e-02 6.69629157e-01
-7.27713406e-01 4.60204244e-01 -7.92846501e-01 -9.71707404e-01
8.71840715e-01 1.64865863e+00 -7.43869007e-01 -1.48874894e-01
-6.92297518e-01 8.25384974e-01 5.36775589e-01 -3.96611951e-02
-9.98367369e-01 -6.71916306e-01 7.15147316e-01 2.11235180e-01
3.18189800e-01 5.93377873e-02 -1.87721118e-01 -9.48484480e-01
-8.27629343e-02 -7.45471001e-01 5.21490872e-01 -3.73215973e-01
-1.96287191e+00 9.53119099e-01 -3.89746189e-01 -1.26726651e+00
-9.44756120e-02 -6.02155447e-01 -4.55683589e-01 7.67079115e-01
-1.32332706e+00 -1.22013128e+00 -4.46170270e-01 1.50162473e-01
-3.47434692e-02 -1.07563210e+00 5.87089181e-01 4.63406533e-01
-5.16810179e-01 5.10190010e-01 1.01892874e-01 -8.18207860e-03
4.40789521e-01 -1.34976184e+00 8.12868714e-01 -2.10808013e-02
5.36124051e-01 7.22763300e-01 3.28108400e-01 -8.12504709e-01
-1.15978265e+00 -1.04317677e+00 1.15987587e+00 -7.92675376e-01
1.66660714e+00 -2.46817976e-01 -5.58797657e-01 5.44894338e-01
2.90902019e-01 1.90102264e-01 8.09635043e-01 3.56894761e-01
-6.28001571e-01 -3.97240400e-01 -9.71968949e-01 1.66314751e-01
1.39060438e+00 -8.62537324e-01 -4.18733537e-01 6.27204359e-01
9.01105404e-01 4.97804917e-02 -1.91908157e+00 3.96832019e-01
3.50712597e-01 -9.21567321e-01 1.38406944e+00 -7.29355991e-01
7.04989374e-01 3.04161757e-02 -3.21488351e-01 -1.36722684e+00
-7.37506986e-01 -4.69203442e-01 -1.46056265e-01 1.49915147e+00
6.67762995e-01 -1.01717150e+00 8.19502413e-01 3.63111123e-02
-2.54219055e-01 -1.04872513e+00 -7.76668191e-01 -1.18684733e+00
-6.98222592e-02 -3.20983753e-02 1.08992767e+00 1.44442177e+00
-9.61841345e-02 3.76494199e-01 -1.45136222e-01 -4.70599860e-01
3.68205130e-01 4.86052096e-01 5.59281707e-01 -1.74705958e+00
-4.86287534e-01 -1.07297850e+00 -1.06808364e+00 -1.82722464e-01
4.02101487e-01 -1.28795981e+00 -8.77467155e-01 -1.73421395e+00
3.03313851e-01 -5.26420355e-01 -7.34553814e-01 -1.46971226e-01
1.73210070e-01 2.12167636e-01 -5.32610863e-02 3.63028467e-01
-5.95024467e-01 1.70344085e-01 1.27555621e+00 -5.08659184e-01
2.72784770e-01 -7.30110779e-02 -9.59014237e-01 3.29990864e-01
7.81042933e-01 -2.47156218e-01 -3.25455964e-01 -3.75236608e-02
7.90033460e-01 -1.81009680e-01 -4.62717488e-02 -6.95080101e-01
9.51833054e-02 8.08234364e-02 -1.24073885e-01 -4.67776239e-01
2.64978230e-01 -6.61910355e-01 7.09724128e-01 4.86543179e-01
-3.20372373e-01 4.25979823e-01 -2.75143862e-01 6.58645689e-01
-8.25412214e-01 -1.87576979e-01 1.68409511e-01 -2.87020057e-01
-9.62200701e-01 4.38818276e-01 2.68841356e-01 4.02057141e-01
1.19610488e+00 -1.16954185e-01 -7.35565901e-01 -1.37528032e-01
-6.11526132e-01 4.42875475e-02 5.38279533e-01 6.40460849e-01
4.16403472e-01 -1.72994113e+00 -8.57375324e-01 -4.65019107e-01
7.75945067e-01 -7.07573831e-01 -6.54450133e-02 9.15198863e-01
-6.44239724e-01 6.44687533e-01 -2.10747689e-01 6.17679171e-02
-1.15998292e+00 7.38224030e-01 -1.09168114e-02 -1.04350328e+00
-2.21284792e-01 1.12087262e+00 -2.08261162e-01 -5.36805034e-01
-2.88161129e-01 -1.44366935e-01 -4.48827296e-01 6.30902588e-01
-1.67993158e-01 6.55054688e-01 2.82926798e-01 -7.62352049e-01
-6.99570119e-01 7.12150872e-01 -7.60310143e-02 3.18392038e-01
1.35822749e+00 3.53752047e-01 -3.78923357e-01 2.80481756e-01
1.36539972e+00 3.67321074e-01 -1.95555925e-01 -1.75596550e-01
4.07088757e-01 -6.97965443e-01 2.64888536e-02 -6.49075866e-01
-1.40314806e+00 3.97604048e-01 3.55531685e-02 6.53636575e-01
7.21615791e-01 3.17549825e-01 6.63565218e-01 3.58072847e-01
7.05205321e-01 -1.17784595e+00 4.48846072e-01 2.63910472e-01
1.04187262e+00 -1.20321107e+00 1.50261760e-01 -7.22928047e-01
-1.03429198e+00 8.75971437e-01 3.00599337e-01 -8.28633606e-02
5.62651932e-01 1.34672180e-01 -3.53811502e-01 -8.51478755e-01
-5.62433958e-01 -3.06661367e-01 7.30270624e-01 4.09412950e-01
6.81851864e-01 3.02288473e-01 -2.77761906e-01 8.77548754e-01
-4.15198207e-01 -4.86459225e-01 2.26625010e-01 5.69811463e-01
-1.00078523e-01 -1.42477083e+00 3.13712269e-01 9.21296656e-01
-2.33024582e-01 -3.76534462e-01 -9.78000164e-01 1.15379322e+00
-7.84596836e-04 6.16836548e-01 1.28362626e-01 -8.60600173e-01
6.85215175e-01 -3.36737096e-01 4.01182890e-01 -6.77791178e-01
-1.21702707e+00 -5.57317913e-01 8.46281350e-01 -6.59303010e-01
-2.70221740e-01 -6.36533976e-01 -8.65112007e-01 -5.94669044e-01
-4.69433755e-01 2.60444731e-01 5.29688001e-01 2.27712795e-01
5.32537639e-01 1.09119904e+00 8.17534626e-01 -3.11552435e-01
-4.92609322e-01 -7.80160248e-01 -1.01482296e+00 6.89122438e-01
-5.97335160e-01 -9.02042866e-01 -5.27065635e-01 -7.51657605e-01] | [8.77233600616455, 7.8751091957092285] |
fe29338e-7bec-4dc7-acca-7f619600901c | shift-reduce-ccg-parsing-using-neural-network | null | null | https://aclanthology.org/N16-1052 | https://aclanthology.org/N16-1052.pdf | Shift-Reduce CCG Parsing using Neural Network Models | null | ['Mark Steedman', 'Bharat Ram Ambati', 'Tejaswini Deoskar'] | 2016-06-01 | null | null | null | naacl-2016-6 | ['ccg-supertagging'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.295064449310303, 3.880946636199951] |
686ccbce-a370-4025-8251-146b3c04ec03 | text-to-kg-alignment-comparing-current | 2306.02871 | null | https://arxiv.org/abs/2306.02871v1 | https://arxiv.org/pdf/2306.02871v1.pdf | Text-To-KG Alignment: Comparing Current Methods on Classification Tasks | In contrast to large text corpora, knowledge graphs (KG) provide dense and structured representations of factual information. This makes them attractive for systems that supplement or ground the knowledge found in pre-trained language models with an external knowledge source. This has especially been the case for classification tasks, where recent work has focused on creating pipeline models that retrieve information from KGs like ConceptNet as additional context. Many of these models consist of multiple components, and although they differ in the number and nature of these parts, they all have in common that for some given text query, they attempt to identify and retrieve a relevant subgraph from the KG. Due to the noise and idiosyncrasies often found in KGs, it is not known how current methods compare to a scenario where the aligned subgraph is completely relevant to the query. In this work, we try to bridge this knowledge gap by reviewing current approaches to text-to-KG alignment and evaluating them on two datasets where manually created graphs are available, providing insights into the effectiveness of current methods. | ['Erik Velldal', 'Lilja Øvrelid', 'Sondre Wold'] | 2023-06-05 | null | null | null | null | ['knowledge-graphs'] | ['knowledge-base'] | [ 1.49028718e-01 5.98532736e-01 -5.47247887e-01 -2.79485375e-01
-5.60820222e-01 -8.32708776e-01 8.91463041e-01 9.09694672e-01
-4.21346843e-01 8.04129660e-01 6.00340605e-01 -3.51578265e-01
-2.50318170e-01 -9.11880732e-01 -4.70321208e-01 -1.43317834e-01
9.25705954e-02 9.49153602e-01 5.95319152e-01 -4.75292951e-01
3.22212070e-01 1.36902243e-01 -1.46314096e+00 3.65516990e-01
9.05553281e-01 5.51842928e-01 1.97806537e-01 1.46280542e-01
-7.56437540e-01 9.95108068e-01 -7.04089046e-01 -7.68965781e-01
-1.33666366e-01 -4.05777514e-01 -1.44457126e+00 -1.49502620e-01
7.88564980e-01 1.74452037e-01 -1.32032812e-01 9.56651568e-01
2.26889029e-01 -5.41481152e-02 6.19072676e-01 -1.17909861e+00
-3.33398521e-01 1.01140857e+00 -2.49538928e-01 2.56428868e-01
6.77921236e-01 -6.09711945e-01 1.32394457e+00 -4.30217713e-01
1.30559933e+00 1.08954895e+00 5.55482209e-01 2.94863939e-01
-1.09159577e+00 -3.01192015e-01 3.95212024e-01 3.61514449e-01
-1.36985195e+00 -3.01314414e-01 6.41762674e-01 -3.05868566e-01
1.46130681e+00 9.68660563e-02 8.15852165e-01 1.02286732e+00
-9.06116888e-02 7.47484922e-01 9.06688094e-01 -9.27239001e-01
-3.69300991e-02 4.54600036e-01 2.86668509e-01 4.58946705e-01
7.19980299e-01 -4.79002833e-01 -7.82424033e-01 -3.55977029e-01
4.37325627e-01 -6.25916123e-01 -4.94347036e-01 -6.08348787e-01
-9.60058331e-01 5.96234918e-01 2.56769359e-01 7.57224083e-01
-2.72849381e-01 -9.77607444e-03 3.12109768e-01 3.57145607e-01
4.40904111e-01 7.25313902e-01 -5.89610219e-01 2.46785283e-02
-1.06303012e+00 4.79563832e-01 1.40075302e+00 1.21795332e+00
9.31802154e-01 -3.75187248e-01 7.49458224e-02 5.27780116e-01
2.70927548e-01 1.21493585e-01 4.24352348e-01 -3.54606271e-01
7.16254771e-01 9.33275223e-01 4.25102152e-02 -1.30574536e+00
-2.26915091e-01 -7.22367346e-01 -4.40829135e-02 -3.60347450e-01
4.93785501e-01 3.67477626e-01 -6.76482141e-01 1.57711732e+00
3.71097207e-01 -1.07869640e-01 2.64478177e-01 6.42725110e-01
1.05262506e+00 6.29697368e-02 5.45603871e-01 -3.26139405e-02
1.29122424e+00 -5.17088950e-01 -6.28863871e-01 -6.68339491e-01
9.39713776e-01 -8.73999298e-01 8.45864654e-01 2.93185376e-02
-5.93219340e-01 -1.22987255e-01 -1.01723588e+00 -2.09935918e-01
-9.18013573e-01 -1.87202245e-01 5.73675513e-01 5.44384420e-01
-1.08567798e+00 6.23430371e-01 -4.57480460e-01 -1.04614902e+00
2.18643293e-01 1.40803065e-02 -3.45394164e-01 -2.97561496e-01
-1.51171196e+00 1.20349741e+00 1.04867494e+00 -3.26327980e-01
-3.66230220e-01 -6.56466126e-01 -8.27190638e-01 5.66239506e-02
9.43861961e-01 -8.43139768e-01 1.27133453e+00 -1.00363588e+00
-8.49597037e-01 1.17251801e+00 -1.29560545e-01 -5.06500781e-01
2.23230571e-01 -2.88525909e-01 -5.66921771e-01 1.71603024e-01
4.04440224e-01 4.67123240e-01 5.27085066e-01 -1.14854288e+00
-5.61759830e-01 -4.53353107e-01 4.94569033e-01 5.01144707e-01
-2.77710408e-01 3.70112062e-01 -6.13375127e-01 -6.65381789e-01
1.54581279e-01 -5.62410235e-01 -1.65583543e-03 -4.69682187e-01
-3.59395981e-01 -3.62797737e-01 7.62934446e-01 -7.44890273e-01
1.46244478e+00 -1.66640723e+00 6.25450462e-02 3.14347923e-01
3.69470388e-01 2.49994352e-01 2.43020132e-01 1.24301219e+00
9.45726335e-02 4.39166844e-01 3.59073766e-02 7.09611028e-02
-7.61258788e-03 4.36555862e-01 -4.97142255e-01 -3.70912999e-02
2.32739344e-01 1.03636634e+00 -1.25037742e+00 -8.28665853e-01
-1.35382116e-01 1.31847233e-01 -1.11072116e-01 -1.11536875e-01
-6.79782987e-01 4.22638208e-02 -6.44878328e-01 4.96907830e-01
1.34209365e-01 -4.16116446e-01 7.37938225e-01 -1.76997632e-01
1.44088820e-01 9.81659055e-01 -1.30180633e+00 1.72210252e+00
-9.25177708e-02 4.87312376e-01 -1.17747381e-01 -1.04228425e+00
6.52465165e-01 4.78424728e-01 2.06529319e-01 -4.64649856e-01
-6.18042015e-02 3.89637142e-01 -2.60915488e-01 -4.32396293e-01
7.89301515e-01 -2.67352760e-01 -1.66961942e-02 4.14795369e-01
3.04497212e-01 -4.69961017e-01 6.12549126e-01 8.55049431e-01
1.09593308e+00 3.66996109e-01 8.05841088e-01 -2.75356948e-01
3.36277306e-01 6.28346980e-01 1.86712101e-01 7.42364883e-01
3.03814381e-01 2.89854944e-01 4.05614227e-01 -1.16186611e-01
-6.81331098e-01 -7.14819372e-01 6.10960685e-02 7.05793083e-01
1.39919937e-01 -1.07312310e+00 -4.80305821e-01 -9.18849409e-01
-3.43817472e-03 9.29731369e-01 -4.20941412e-01 4.93893363e-02
-4.01084393e-01 -2.99697578e-01 5.00942290e-01 5.58312058e-01
1.51923344e-01 -9.47440982e-01 -5.00180483e-01 5.47671199e-01
-3.45032632e-01 -1.44415641e+00 8.38190392e-02 1.21869974e-01
-8.08279097e-01 -1.61670554e+00 -9.21213105e-02 -5.94702005e-01
5.64662099e-01 3.16393584e-01 1.92276239e+00 4.17760313e-01
-4.52705063e-02 8.55399489e-01 -7.55936086e-01 -6.75693274e-01
-4.77984220e-01 4.28299397e-01 -3.63841385e-01 -4.55812395e-01
8.83473635e-01 -4.90490735e-01 -8.64987522e-02 1.27060905e-01
-1.10754609e+00 1.28233701e-01 4.13846999e-01 3.68808240e-01
4.14103508e-01 -5.40660881e-03 5.73576868e-01 -1.11826909e+00
9.80291486e-01 -6.84769213e-01 -4.30636585e-01 6.99547827e-01
-8.07629228e-01 1.45315036e-01 3.83984745e-01 5.14829252e-03
-8.76732707e-01 -2.45408788e-01 1.24179766e-01 -5.93506657e-02
-1.57126218e-01 1.26998639e+00 3.17646340e-02 2.04422455e-02
7.93853641e-01 -2.10260544e-02 -3.40276301e-01 -6.42627358e-01
5.47833383e-01 5.15414536e-01 3.87094855e-01 -8.80667627e-01
7.86424398e-01 2.13939160e-01 -2.39802063e-01 -7.37893939e-01
-1.07696462e+00 -8.40530813e-01 -8.84498537e-01 -2.16992781e-01
5.16714215e-01 -8.03788245e-01 9.68622044e-02 3.31360847e-02
-1.01232481e+00 -3.52113694e-02 -5.86403251e-01 3.27143848e-01
-2.97735333e-01 5.35883665e-01 -2.33228371e-01 -4.96178031e-01
-2.69762933e-01 -6.05126739e-01 8.35672677e-01 8.26628208e-02
-5.50308347e-01 -1.31090188e+00 5.39964996e-02 4.06712830e-01
4.07895207e-01 1.57644570e-01 1.12045705e+00 -1.14137876e+00
-4.99910057e-01 -3.72620851e-01 -7.40812123e-02 -6.06084615e-02
1.79464489e-01 -1.55306831e-01 -7.73429871e-01 -2.50548393e-01
-2.81634212e-01 -5.51153779e-01 5.44163942e-01 -1.95153087e-01
4.91990417e-01 -2.97311813e-01 -7.39857912e-01 -1.78070784e-01
1.59231544e+00 -1.31149590e-01 5.01060843e-01 6.34600878e-01
5.56210637e-01 9.22319531e-01 5.58723271e-01 -1.43217981e-01
5.44025302e-01 8.36140633e-01 3.07207972e-01 1.02939174e-01
-3.19426984e-01 -5.89381456e-01 3.43453921e-02 7.41651952e-01
9.34882462e-02 -3.39392513e-01 -1.16636586e+00 8.62692177e-01
-1.82993460e+00 -9.58901405e-01 -1.93609416e-01 2.21736884e+00
1.06028104e+00 1.21106297e-01 -4.02354337e-02 3.10137868e-03
4.85871136e-01 1.32284775e-01 -1.45956933e-01 6.35311007e-04
-3.64964455e-01 2.46911079e-01 2.62156934e-01 4.00735795e-01
-8.45032275e-01 1.26144147e+00 6.37343884e+00 8.86538088e-01
-7.80438662e-01 -1.17691666e-01 -1.72694307e-02 1.46500617e-01
-4.83252704e-01 5.71069300e-01 -1.03651690e+00 -1.26923164e-02
8.49178255e-01 -5.14931619e-01 1.41566545e-01 7.28088677e-01
-2.26664141e-01 -2.65130579e-01 -1.20108962e+00 4.74688947e-01
4.05940264e-01 -1.36786819e+00 3.35361183e-01 -1.79985594e-02
3.96783322e-01 1.29052371e-01 -5.78359962e-01 3.03150117e-01
6.57380939e-01 -1.03654623e+00 7.33408630e-01 1.37637168e-01
3.08523834e-01 -3.34170491e-01 8.03433478e-01 3.75632167e-01
-1.30027592e+00 3.97173941e-01 -1.41224205e-01 -1.61780547e-02
1.45137766e-02 4.57048357e-01 -1.33232856e+00 1.26232135e+00
6.34017646e-01 6.92180872e-01 -9.23938513e-01 9.86121535e-01
-8.13169122e-01 5.63751638e-01 -4.02098477e-01 -7.64294788e-02
2.80784965e-01 -9.66624543e-03 5.82701087e-01 1.28163588e+00
-4.54000616e-03 -3.91820930e-02 5.34254789e-01 7.73290575e-01
-6.25669956e-03 7.21496880e-01 -8.61650765e-01 -3.94537717e-01
5.21883368e-01 1.17790031e+00 -8.74968231e-01 -5.90904534e-01
-7.05486894e-01 5.40332675e-01 6.91172957e-01 3.77393305e-01
-1.98418960e-01 -4.56437618e-01 2.57394135e-01 4.27275300e-01
2.40402475e-01 -2.31337249e-01 2.21230060e-01 -1.22294188e+00
3.72317255e-01 -9.32476163e-01 9.16079581e-01 -8.89202118e-01
-1.28462195e+00 4.66729939e-01 5.05653858e-01 -8.59065771e-01
-5.83056152e-01 -4.79037493e-01 -1.81443676e-01 8.65873337e-01
-1.59710705e+00 -1.42263603e+00 -1.63588896e-01 5.66614926e-01
2.10851938e-01 2.08968088e-01 8.46753418e-01 1.09404817e-01
-7.43965879e-02 1.59941778e-01 -2.58380830e-01 3.12444091e-01
8.80577743e-01 -1.33117986e+00 3.85692537e-01 9.22086716e-01
8.40568125e-01 1.12908065e+00 8.48812401e-01 -1.18249655e+00
-1.14052022e+00 -7.55843818e-01 1.27747524e+00 -6.97792053e-01
9.37077224e-01 -2.24279657e-01 -1.27397525e+00 9.23262894e-01
3.97084296e-01 -3.44724804e-01 8.15286279e-01 5.40390968e-01
-6.50699317e-01 2.42613569e-01 -8.25498521e-01 4.91398752e-01
1.22732663e+00 -6.35134339e-01 -1.21741784e+00 3.79253507e-01
4.68844414e-01 -4.38733459e-01 -7.96283126e-01 2.42996037e-01
2.42180794e-01 -6.35903656e-01 7.47325778e-01 -9.87712264e-01
2.25311279e-01 -4.45815265e-01 -3.98047343e-02 -1.36560929e+00
6.57881424e-02 -4.40451294e-01 -1.40279174e-01 1.43023372e+00
4.91200715e-01 -6.65013850e-01 7.81840742e-01 6.08703017e-01
-2.14974716e-01 -5.09103477e-01 -8.37089419e-01 -9.17140067e-01
-2.50584586e-03 -5.09394348e-01 5.40091932e-01 1.32610571e+00
5.61329603e-01 6.02706850e-01 1.82572037e-01 1.12152368e-01
3.17003310e-01 2.96738297e-01 7.76805162e-01 -1.61340523e+00
-1.09913364e-01 -3.38637948e-01 -5.91295302e-01 -7.71939337e-01
2.91489482e-01 -1.26256311e+00 -3.36576939e-01 -2.09019947e+00
3.15574348e-01 -4.37285930e-01 -1.26899317e-01 8.29975843e-01
-3.25204372e-01 -2.36725777e-01 -4.25231718e-02 3.55982840e-01
-6.40945733e-01 1.93841383e-01 6.94682062e-01 -1.63194641e-01
-1.69905871e-01 -5.31595945e-01 -1.02501166e+00 5.74329138e-01
8.03122103e-01 -5.56155264e-01 -7.21000254e-01 -4.51939851e-01
6.68271303e-01 -4.05620426e-01 3.06620479e-01 -9.00135040e-01
6.05260730e-01 -1.47981957e-01 1.85145829e-02 -4.59111512e-01
9.37724207e-03 -7.99902380e-01 3.26161653e-01 1.27226144e-01
-2.06102699e-01 -2.55352650e-02 4.66865838e-01 6.33220077e-01
-5.44782698e-01 -6.03593767e-01 1.91879973e-01 -4.79759067e-01
-1.07813692e+00 -9.43916515e-02 -4.51676637e-01 5.62473357e-01
7.45228410e-01 -4.53115731e-01 -6.48118317e-01 -4.91247982e-01
-4.57758367e-01 2.92067409e-01 6.54595077e-01 5.57899714e-01
3.21749508e-01 -9.79832709e-01 -4.84739482e-01 -2.86139250e-01
6.49062693e-01 -4.16325741e-02 -2.71190166e-01 5.73647857e-01
-3.30466628e-01 7.03809142e-01 -7.57655427e-02 -1.41808897e-01
-1.34498096e+00 5.97803056e-01 2.02570692e-01 -6.17425203e-01
-5.54923475e-01 5.62438309e-01 -2.06727281e-01 -2.22095519e-01
6.50995597e-02 -2.61947163e-03 -5.60076654e-01 2.86815912e-01
4.00949031e-01 -1.83022887e-01 4.81461734e-01 -6.77882135e-01
-4.05402869e-01 4.53702927e-01 -4.01542395e-01 -2.64722295e-02
1.31404793e+00 -1.68937352e-02 -4.08292264e-01 3.88676107e-01
5.91479063e-01 3.13047856e-01 -3.83745134e-01 -7.33733237e-01
6.99815035e-01 -4.99089122e-01 -8.67393538e-02 -9.71305490e-01
-8.04723740e-01 4.90943134e-01 -1.44672289e-01 4.05195028e-01
9.18085158e-01 4.53038156e-01 2.22428322e-01 5.46630561e-01
7.33357191e-01 -1.05131197e+00 -5.30523881e-02 5.36593199e-01
9.46300387e-01 -9.18273628e-01 3.80275100e-01 -9.05357301e-01
-4.18754458e-01 1.14430141e+00 6.02506161e-01 3.69224519e-01
3.78207892e-01 2.83285417e-02 1.13684855e-01 -8.72225106e-01
-6.14552915e-01 -5.66356242e-01 3.47435057e-01 8.55327725e-01
6.95335388e-01 -3.18806380e-01 -5.33540964e-01 2.93210357e-01
-3.76586109e-01 3.76674868e-02 4.05723542e-01 1.31254649e+00
-3.45787734e-01 -1.56576157e+00 -7.17651546e-02 5.44010401e-01
-5.30727625e-01 -4.36573386e-01 -1.09513640e+00 1.14689183e+00
-2.09408909e-01 1.05169857e+00 -3.90386432e-01 -5.28981723e-02
3.68065536e-01 4.94301766e-01 4.91539538e-01 -1.20269787e+00
-8.71381044e-01 -2.20440060e-01 8.16505551e-01 -4.26309288e-01
-7.10733354e-01 -5.49049973e-01 -9.33827460e-01 4.46689576e-02
-4.81685460e-01 5.50518155e-01 4.73007411e-01 1.15577757e+00
3.36485714e-01 1.74169466e-01 -3.44722003e-01 -2.58492380e-01
-3.36107910e-01 -9.25814331e-01 -5.67676544e-01 6.27089620e-01
-3.82187366e-01 -6.45823240e-01 -9.11951959e-02 1.26678407e-01] | [9.45775032043457, 8.195188522338867] |
57c02420-5e8a-4ddb-abc8-556de3752179 | a-radiomics-approach-to-traumatic-brain | 1811.05699 | null | http://arxiv.org/abs/1811.05699v1 | http://arxiv.org/pdf/1811.05699v1.pdf | A Radiomics Approach to Traumatic Brain Injury Prediction in CT Scans | Computer Tomography (CT) is the gold standard technique for brain damage
evaluation after acute Traumatic Brain Injury (TBI). It allows identification
of most lesion types and determines the need of surgical or alternative
therapeutic procedures. However, the traditional approach for lesion
classification is restricted to visual image inspection. In this work, we
characterize and predict TBI lesions by using CT-derived radiomics descriptors.
Relevant shape, intensity and texture biomarkers characterizing the different
lesions are isolated and a lesion predictive model is built by using Partial
Least Squares. On a dataset containing 155 scans (105 train, 50 test) the
methodology achieved 89.7 % accuracy over the unseen data. When a model was
build using only texture features, a 88.2 % accuracy was obtained. Our results
suggest that selected radiomics descriptors could play a key role in brain
injury prediction. Besides, the proposed methodology is close to reproduce
radiologists decision making. These results open new possibilities for
radiomics-inspired brain lesion detection, segmentation and prediction. | ['Thijs Vande Vyvere', 'Bjoern Menze', 'Jan S. Kirschke', 'Ezequiel de la Rosa', 'Diana M. Sima'] | 2018-11-14 | null | null | null | null | ['injury-prediction'] | ['playing-games'] | [ 3.37789506e-01 -2.49324307e-01 -1.45164788e-01 -1.94088116e-01
-9.22331691e-01 -1.51161283e-01 4.95788962e-01 6.07713163e-01
-7.44551718e-01 6.27587378e-01 -2.05767658e-02 -1.06793515e-01
-2.04231843e-01 -8.06109548e-01 -2.50848591e-01 -1.02167869e+00
-2.11676940e-01 1.10260856e+00 6.24551177e-01 1.81680530e-01
4.01994973e-01 1.12869608e+00 -1.48698080e+00 4.90287960e-01
6.64965093e-01 1.10184574e+00 6.64886773e-01 3.37759167e-01
-2.18436897e-01 8.27686608e-01 -4.01871562e-01 1.97988063e-01
4.40670736e-02 -2.58279681e-01 -8.87939751e-01 -2.98555288e-02
4.80876081e-02 -2.62174577e-01 -8.46512094e-02 6.67016685e-01
6.81508183e-01 8.04783553e-02 1.39064062e+00 -6.78835571e-01
-9.42548290e-02 2.45374769e-01 -5.69704950e-01 6.85539603e-01
3.47354949e-01 -4.37786104e-03 2.95737684e-01 -1.03379846e+00
4.60786700e-01 7.49873400e-01 2.78774381e-01 5.71502566e-01
-1.19288707e+00 -5.30743897e-01 -3.84234369e-01 9.71205354e-01
-1.22884238e+00 -8.91892090e-02 4.14726645e-01 -1.07221723e+00
5.80371141e-01 5.27967632e-01 8.92092109e-01 7.17875004e-01
5.27642190e-01 6.83345437e-01 1.79139507e+00 -4.77168083e-01
3.15307885e-01 -1.75262451e-01 4.27232206e-01 8.73449147e-01
3.47108752e-01 1.54918134e-01 -2.20906943e-01 -2.27504984e-01
5.02579212e-01 2.27475449e-01 -4.26085681e-01 -3.01021039e-01
-1.02079701e+00 6.78144038e-01 5.12760818e-01 6.97718680e-01
-5.73531449e-01 -1.79470345e-01 4.69245881e-01 -7.53218308e-02
3.09067369e-01 4.83957753e-02 2.26912089e-02 3.24907094e-01
-1.10695958e+00 -3.00575972e-01 1.50883302e-01 2.29037762e-01
5.09687006e-01 -1.36779502e-01 -3.29541415e-01 1.18050086e+00
2.54106317e-02 5.67761958e-01 9.99534667e-01 -4.13590461e-01
-1.09960638e-01 5.56783795e-01 -4.80066866e-01 -5.69463611e-01
-7.48183966e-01 -7.58313909e-02 -6.99269474e-01 5.03929794e-01
4.99783516e-01 1.00869328e-01 -1.19152272e+00 7.95538425e-01
1.32467002e-01 -1.16671190e-01 -5.67132652e-01 1.13148236e+00
8.02135587e-01 1.16410859e-01 3.23198438e-01 -2.18611896e-01
1.60998833e+00 -4.93758619e-01 -3.02820355e-01 -1.39791176e-01
6.42715931e-01 -6.63263559e-01 7.44349301e-01 5.59405565e-01
-7.07675517e-01 4.76657376e-02 -7.54699886e-01 3.77052039e-01
-1.93058059e-01 3.23476523e-01 2.23432139e-01 6.07506752e-01
-8.25838625e-01 3.86043191e-01 -8.87591898e-01 -6.49235845e-01
5.81034362e-01 5.12323737e-01 -7.94084668e-01 -2.54272580e-01
-4.48654473e-01 1.45548046e+00 4.77291316e-01 -1.85084552e-01
-7.65682697e-01 -4.27290112e-01 -2.80255049e-01 -3.92620951e-01
-1.88111942e-02 -5.63785851e-01 7.48675883e-01 -3.88345927e-01
-1.24701297e+00 1.32902980e+00 -3.08203131e-01 -5.01919985e-01
6.13607705e-01 1.28533289e-01 -1.60143644e-01 8.61103296e-01
8.22732970e-02 4.35925990e-01 6.69056594e-01 -1.42812669e+00
-4.20131743e-01 -7.21817911e-01 -6.28974557e-01 -1.37822226e-01
-7.77854174e-02 4.71079409e-01 8.01844820e-02 -5.65228581e-01
4.85462308e-01 -7.96835303e-01 -9.01167914e-02 -1.37812391e-01
-2.06316441e-01 -9.91989449e-02 6.34545684e-01 -9.52496886e-01
7.93685973e-01 -1.55513453e+00 -3.64230983e-02 5.69526792e-01
4.93502557e-01 1.51895583e-01 2.90070236e-01 1.42857447e-01
-3.45079899e-01 -3.00506931e-02 -3.46859604e-01 9.45733115e-02
-4.10065114e-01 -1.23335406e-01 2.47555241e-01 8.42985988e-01
-2.50419319e-01 7.56635547e-01 -3.77485603e-01 -8.82921517e-01
5.48059702e-01 3.41270864e-01 -1.82241186e-01 4.11666371e-02
3.19444954e-01 8.09652448e-01 -6.73637211e-01 6.99776411e-01
5.34393132e-01 1.63645297e-01 -1.60120577e-01 -1.23740599e-01
-3.66111733e-02 -4.11554188e-01 -4.01418209e-01 1.08523095e+00
-3.46996933e-01 6.57244802e-01 -2.65431672e-01 -1.33990633e+00
1.11699092e+00 3.63456160e-01 1.02252710e+00 -9.29960310e-01
6.39480770e-01 4.31582332e-01 -3.64600979e-02 -9.03237998e-01
-3.70608330e-01 -6.29525244e-01 4.74784344e-01 4.71447080e-01
-1.98721752e-01 -2.05594838e-01 9.62757971e-04 -2.40122765e-01
1.10642684e+00 -4.43013519e-01 1.54250398e-01 -4.27084059e-01
5.85705042e-01 3.30784917e-01 -1.24893673e-01 5.49909472e-01
-3.96068692e-01 9.52909172e-01 1.89221099e-01 -5.78840733e-01
-7.61740923e-01 -9.96812046e-01 -6.73652709e-01 6.51189089e-01
-5.98285310e-02 3.69337827e-01 -9.70178843e-01 -3.88957679e-01
-2.58811295e-01 4.65670049e-01 -8.61344397e-01 -1.24215387e-01
-6.45628393e-01 -1.08458424e+00 3.05434614e-01 2.85406470e-01
2.67993748e-01 -8.47061634e-01 -7.82613575e-01 1.46085441e-01
-3.79312396e-01 -7.77108252e-01 1.09717526e-01 2.77493328e-01
-1.12782538e+00 -1.44604588e+00 -1.11209345e+00 -9.07253444e-01
6.95502341e-01 3.47619593e-01 5.38602054e-01 4.23098296e-01
-8.13097596e-01 4.35448498e-01 -4.93296146e-01 -2.25480348e-01
-2.85612881e-01 -2.90505469e-01 -1.00400589e-01 6.05033860e-02
2.26563245e-01 -4.52030480e-01 -9.89421487e-01 3.24199021e-01
-7.20376253e-01 -2.39177227e-01 7.78623164e-01 6.73050463e-01
8.32712173e-01 -2.57610857e-01 3.48842412e-01 -6.48695707e-01
4.43975717e-01 -4.36713159e-01 5.22529893e-02 4.02201802e-01
-5.56295276e-01 -1.79502532e-01 3.88783127e-01 -2.64123708e-01
-9.14633274e-01 -2.40081456e-03 -9.24849883e-02 -2.08686784e-01
-8.04260015e-01 3.93106252e-01 3.17624480e-01 -4.92237329e-01
9.24208999e-01 3.33562225e-01 8.46261904e-02 -6.61991715e-01
-3.60599607e-01 5.85372865e-01 4.38694507e-01 -7.35331118e-01
2.93771446e-01 7.19732046e-01 2.08085552e-01 -9.84997571e-01
-4.26442891e-01 -9.51974392e-01 -1.06226397e+00 -7.27404237e-01
1.05151415e+00 -4.23573554e-01 -3.65342110e-01 3.63084733e-01
-1.01444697e+00 -2.84648508e-01 -1.97163727e-02 8.85954618e-01
-5.41090548e-01 5.58254242e-01 -4.65518713e-01 -7.06836462e-01
-5.41685879e-01 -1.45862877e+00 9.36901927e-01 -1.76004440e-01
-7.30998367e-02 -8.15126956e-01 1.47463799e-01 7.68986344e-01
2.19081208e-01 4.74968225e-01 1.40963972e+00 -6.56055272e-01
-1.16598375e-01 -5.48696637e-01 -3.03342879e-01 2.56482828e-02
-1.59292053e-02 -1.76422074e-01 -9.04478252e-01 -1.65108144e-01
1.31708518e-01 -1.73774719e-01 1.07496250e+00 5.26789963e-01
1.24848819e+00 2.80112535e-01 -4.02242124e-01 2.48258337e-01
1.52989686e+00 5.01002729e-01 5.54975867e-01 7.78060615e-01
4.00181204e-01 7.83704102e-01 4.00281012e-01 3.74755591e-01
5.97126149e-02 6.31407619e-01 3.89340639e-01 -7.67445425e-03
-5.31922936e-01 2.79783219e-01 1.00004911e-01 6.67863667e-01
-5.10944188e-01 1.16890565e-01 -1.35055053e+00 3.56421411e-01
-1.19397545e+00 -9.83885109e-01 -7.89378703e-01 2.13957310e+00
3.51580352e-01 -1.33148462e-01 3.15659791e-01 5.57715595e-01
9.27437544e-01 -5.05927265e-01 -2.00973108e-01 -1.88026935e-01
-3.38033766e-01 6.06778741e-01 5.42703390e-01 1.26736298e-01
-9.23429072e-01 5.74388742e-01 6.77505159e+00 1.03264236e+00
-1.48130620e+00 3.87163192e-01 6.10507011e-01 -3.34892161e-02
9.39901397e-02 -2.14456618e-01 -6.94352016e-02 3.93192828e-01
6.18746042e-01 8.17300528e-02 1.58501461e-01 2.86614805e-01
4.96927381e-01 -8.61129165e-01 -5.97898960e-01 8.25590849e-01
1.68520197e-01 -8.96947742e-01 1.38392761e-01 7.42765293e-02
3.89033467e-01 2.22254157e-01 -1.28879145e-01 -2.74426222e-01
-2.96893805e-01 -1.09383500e+00 8.49893987e-01 9.76708412e-01
9.47128117e-01 -5.57214499e-01 7.88915515e-01 3.32247108e-01
-8.65993202e-01 -1.55456066e-01 -2.23687336e-01 2.50656933e-01
7.75312856e-02 4.98111010e-01 -1.18154132e+00 2.69530267e-01
6.95463955e-01 8.90505984e-02 -6.93896711e-01 1.72584999e+00
8.12485144e-02 7.67662525e-01 -2.67731994e-01 1.04771584e-01
-1.36743495e-02 -3.05364430e-01 5.29919446e-01 1.31955218e+00
2.68872231e-01 2.41027683e-01 1.50886402e-01 3.89161736e-01
6.36086226e-01 7.69807100e-01 -3.66972208e-01 6.81617677e-01
1.09958567e-01 9.72751081e-01 -1.55476999e+00 -2.18553275e-01
-2.48799682e-01 7.42330372e-01 2.59329617e-01 1.18214130e-01
-5.07919729e-01 7.65077472e-02 -3.92334983e-02 4.32632685e-01
-2.44035915e-01 -1.39474288e-01 -6.78470433e-01 -9.61640179e-01
-2.39309058e-01 -4.68782961e-01 5.28364122e-01 -7.52049863e-01
-1.00631702e+00 6.21647120e-01 2.95206219e-01 -1.11837316e+00
2.31386125e-01 -8.55979323e-01 -7.69587755e-01 6.34488225e-01
-1.12459874e+00 -9.28667188e-01 -4.19181794e-01 6.14093125e-01
3.68185550e-01 -7.89412409e-02 7.90673256e-01 1.57322645e-01
-7.26064503e-01 -1.77113153e-02 2.61828750e-01 7.83273112e-03
5.32542765e-01 -1.14142632e+00 -7.41665065e-01 5.62271237e-01
-2.41081357e-01 6.44397140e-02 7.81532228e-01 -5.49418747e-01
-8.76318157e-01 -7.57237613e-01 5.64135790e-01 -1.75046176e-01
6.33842528e-01 3.02469820e-01 -8.35797668e-01 1.97933868e-01
-1.59757078e-01 -8.43722373e-02 8.68332922e-01 -6.96314275e-01
3.98164056e-02 2.73400307e-01 -1.37610519e+00 2.24872023e-01
5.45266628e-01 -3.85926574e-01 -6.62408471e-01 5.19164264e-01
-2.68650234e-01 -1.67643849e-03 -1.03135216e+00 5.10296941e-01
6.07547760e-01 -1.15474224e+00 1.05847371e+00 -5.25807559e-01
1.29001319e-01 3.70423980e-02 -1.38558792e-02 -1.10992324e+00
-3.90387654e-01 4.82166886e-01 4.37098116e-01 4.52814519e-01
1.58504322e-01 -5.88201523e-01 8.40729415e-01 6.94806278e-01
-2.86020577e-01 -7.87903428e-01 -1.12789679e+00 -6.97117150e-01
4.69639868e-01 -5.82562625e-01 2.74816841e-01 6.99392736e-01
1.86209261e-01 -4.11744267e-01 2.42567644e-01 -5.67427911e-02
7.44403005e-01 2.80428175e-02 4.31411229e-02 -1.55511320e+00
2.12780535e-01 -8.62478614e-01 -7.91298330e-01 -2.95238674e-01
2.50817537e-01 -1.25697315e+00 -1.29309714e-01 -1.83465946e+00
6.17232919e-01 -6.31147683e-01 -4.82867301e-01 2.89279819e-01
1.96355775e-01 6.54123783e-01 5.16126007e-02 5.97479820e-01
4.70160246e-02 2.48592153e-01 1.44245327e+00 -1.79319426e-01
2.39721194e-01 6.46493062e-02 -2.22142804e-02 7.48846948e-01
1.14896226e+00 -5.94196558e-01 -3.24714072e-02 -2.17043862e-01
-6.41164899e-01 1.99810788e-01 7.40749657e-01 -1.25687981e+00
3.11477691e-01 -2.91577280e-01 4.86356586e-01 -3.85360330e-01
2.59522140e-01 -8.77318144e-01 1.61636651e-01 8.53961706e-01
-1.00996532e-01 -2.46097311e-01 -3.15940194e-03 3.34152967e-01
-1.75393209e-01 -6.91552460e-01 1.14722407e+00 -1.43593267e-01
-6.05803847e-01 3.46105933e-01 -1.15717089e+00 -2.09684342e-01
1.28130507e+00 -6.33605778e-01 -1.37547284e-01 -1.51423435e-03
-1.17435932e+00 -4.36520845e-01 3.13374937e-01 -4.52692173e-02
9.03870523e-01 -1.05526567e+00 -6.98493898e-01 -7.05866218e-02
2.85652637e-01 -3.38602751e-01 4.72501487e-01 1.54232550e+00
-1.00632656e+00 5.85856736e-01 -5.83391845e-01 -7.63863504e-01
-1.47193110e+00 2.75405079e-01 5.09007037e-01 3.64004914e-03
-7.84598112e-01 6.05481327e-01 1.76020741e-01 1.06422298e-01
-7.71326274e-02 -1.64103091e-01 -7.48070717e-01 2.17367366e-01
3.19637746e-01 7.72671223e-01 6.37208700e-01 -1.24940181e+00
-3.90384912e-01 8.76297534e-01 1.84220493e-01 -1.08162269e-01
1.36407375e+00 -1.23837769e-01 -3.56045634e-01 3.24295908e-01
1.14355230e+00 -2.91516960e-01 -4.90725338e-01 -8.30853209e-02
2.15903446e-01 -2.50138611e-01 4.73422050e-01 -8.86088431e-01
-1.25839245e+00 1.07083702e+00 1.18509746e+00 -5.76230988e-04
9.90320921e-01 2.53841877e-01 6.46027982e-01 2.37179890e-01
8.06776822e-01 -7.90134788e-01 -1.41366139e-01 1.46452114e-01
9.90822256e-01 -1.19092405e+00 -1.29730359e-01 -4.73387241e-01
-5.61087310e-01 1.53095114e+00 3.98632288e-01 -2.01018184e-01
6.99326396e-01 9.70607102e-02 5.25355376e-02 -5.22499919e-01
-2.73458540e-01 -3.79047304e-01 5.23611128e-01 8.11302662e-01
5.63929379e-01 4.47215825e-01 -7.01924205e-01 4.22518611e-01
-1.35324880e-01 -1.41098976e-01 3.12407494e-01 8.75557065e-01
-1.12709999e+00 -1.07678640e+00 -8.20593059e-01 1.06768835e+00
-5.29410660e-01 8.97115991e-02 -4.03361738e-01 6.85548604e-01
1.71108305e-01 8.76376033e-01 -3.01745772e-01 -2.79826462e-01
2.81415910e-01 1.76683933e-01 8.13037694e-01 -5.47274172e-01
-4.06309605e-01 6.74377382e-02 -1.66864768e-01 -3.14635038e-01
-4.11747843e-01 -9.99134481e-01 -1.51586485e+00 -1.72301829e-01
-2.94819415e-01 1.00462474e-01 7.02058554e-01 1.26514041e+00
-2.20337316e-01 3.02234024e-01 4.47687924e-01 -7.44340420e-01
1.06983885e-01 -8.86553109e-01 -8.59013677e-01 2.36799136e-01
2.36479333e-03 -1.14738762e+00 -1.55518800e-01 1.78860217e-01] | [14.364632606506348, -2.1605327129364014] |
931c68bf-dfd3-446f-be24-b96a383691ea | bringing-structure-into-summaries-a-faceted | 2106.00130 | null | https://arxiv.org/abs/2106.00130v2 | https://arxiv.org/pdf/2106.00130v2.pdf | Bringing Structure into Summaries: a Faceted Summarization Dataset for Long Scientific Documents | Faceted summarization provides briefings of a document from different perspectives. Readers can quickly comprehend the main points of a long document with the help of a structured outline. However, little research has been conducted on this subject, partially due to the lack of large-scale faceted summarization datasets. In this study, we present FacetSum, a faceted summarization benchmark built on Emerald journal articles, covering a diverse range of domains. Different from traditional document-summary pairs, FacetSum provides multiple summaries, each targeted at specific sections of a long document, including the purpose, method, findings, and value. Analyses and empirical results on our dataset reveal the importance of bringing structure into summaries. We believe FacetSum will spur further advances in summarization research and foster the development of NLP systems that can leverage the structured information in both long texts and summaries. | ['Daqing He', 'Tong Wang', 'Xingdi Yuan', 'Yue Dong', 'Lei Zhang', 'Khushboo Thaker', 'Rui Meng'] | 2021-05-31 | null | https://aclanthology.org/2021.acl-short.137 | https://aclanthology.org/2021.acl-short.137.pdf | acl-2021-5 | ['unsupervised-extractive-summarization'] | ['natural-language-processing'] | [ 2.08184153e-01 4.21318144e-01 -7.15530157e-01 -1.39744177e-01
-1.20240355e+00 -9.44209099e-01 7.77241766e-01 6.48594499e-01
1.28662691e-01 1.00320458e+00 1.62846863e+00 -1.92873567e-01
-2.08955333e-01 -4.00919229e-01 -3.58970374e-01 -5.46784960e-02
3.23534876e-01 3.97547036e-01 -1.67805195e-01 -2.13049024e-01
8.42680693e-01 8.89762491e-02 -1.03244472e+00 5.40221572e-01
1.41923249e+00 5.16401350e-01 3.06606516e-02 5.33241034e-01
-5.64718366e-01 6.24778271e-01 -1.04373825e+00 -5.38082063e-01
-2.52730012e-01 -7.42563665e-01 -8.92290294e-01 3.49641591e-01
7.42613018e-01 -2.57468343e-01 -9.57556739e-02 7.94028461e-01
4.61867362e-01 3.07607874e-02 7.90445507e-01 -7.93940127e-01
-7.35906959e-01 1.03799772e+00 -5.61761022e-01 2.64140576e-01
8.65332603e-01 2.97866366e-03 1.50941014e+00 -4.40565079e-01
9.35309470e-01 9.33635354e-01 4.82753277e-01 3.04508954e-01
-9.11228597e-01 -1.68466315e-01 8.07240084e-02 -9.31114629e-02
-5.51955640e-01 -5.74969530e-01 5.87024629e-01 -3.02872479e-01
1.09653139e+00 3.53708625e-01 9.08687234e-01 1.14902127e+00
3.58941823e-01 9.01191950e-01 7.41299391e-01 -1.47599712e-01
-3.83631401e-02 -4.10433039e-02 6.94869041e-01 2.94425964e-01
1.05565560e+00 -7.90403306e-01 -8.34869742e-01 -2.24307969e-01
-2.25258227e-02 -1.17791489e-01 -4.91159886e-01 5.12395017e-02
-1.37138212e+00 6.55983508e-01 -9.16995332e-02 2.64186084e-01
-7.75388002e-01 -2.74526566e-01 8.03551137e-01 1.61434829e-01
6.41593516e-01 1.00989866e+00 -2.16917336e-01 -6.24728739e-01
-1.49965048e+00 5.87878168e-01 1.46607924e+00 1.17624414e+00
4.82289404e-01 -3.92065085e-02 -8.15859437e-01 5.68481445e-01
-2.61028975e-01 5.17805636e-01 3.42952430e-01 -1.05396318e+00
8.31114709e-01 1.05591309e+00 1.44547522e-01 -8.77599299e-01
-2.58722991e-01 -6.09595239e-01 -5.49340010e-01 -6.90099895e-01
-1.46385625e-01 -4.27889168e-01 -4.36131090e-01 9.92292762e-01
-1.64369524e-01 -5.34541667e-01 2.53224999e-01 5.11462629e-01
1.62622559e+00 7.93402731e-01 -4.09714401e-01 -4.69837159e-01
1.47196686e+00 -1.03863633e+00 -1.04439592e+00 -2.20381796e-01
5.47914982e-01 -8.63439620e-01 1.22274971e+00 2.85470217e-01
-1.35520279e+00 2.15754732e-02 -1.12731624e+00 -4.00268644e-01
-1.21336348e-01 9.70976278e-02 3.58399391e-01 3.76661509e-01
-7.86651433e-01 3.43995392e-01 -5.04018843e-01 -5.83002687e-01
5.19928992e-01 -4.84239757e-01 -2.34986261e-01 -2.04264939e-01
-6.68416560e-01 8.15228820e-01 2.07411110e-01 -4.42698836e-01
-3.44928026e-01 -9.94722366e-01 -7.09098101e-01 3.31039459e-01
7.61843860e-01 -1.14048970e+00 1.47955430e+00 -1.85636818e-01
-1.16377985e+00 3.69178861e-01 -5.90095460e-01 -4.95310187e-01
2.00246468e-01 -6.65313125e-01 -3.99303101e-02 3.83130461e-01
5.86350858e-01 5.88487349e-02 4.72655803e-01 -9.64129984e-01
-4.34731364e-01 -3.74502689e-01 -1.80080950e-01 3.67685199e-01
-3.22729409e-01 2.48450022e-02 -5.07049263e-01 -6.67550743e-01
-3.20738494e-01 -4.73795861e-01 2.97715981e-02 -8.19844484e-01
-1.14017510e+00 -2.64633685e-01 6.33229852e-01 -8.89889181e-01
1.77047765e+00 -1.64802551e+00 1.27862692e-01 -3.05434734e-01
4.40386176e-01 5.65753430e-02 -4.33138460e-02 1.34609699e+00
5.90325534e-01 6.13309324e-01 -2.01065347e-01 -7.75996149e-02
9.51516926e-02 -2.26633310e-01 -7.31950402e-01 6.48964546e-04
-7.38372654e-02 1.28779304e+00 -9.05282974e-01 -6.25696898e-01
-2.83877462e-01 -4.89878878e-02 -4.01140571e-01 -6.48855604e-03
-5.26372612e-01 -6.48695529e-02 -7.09470809e-01 5.68048060e-01
3.34377855e-01 -3.35401535e-01 -1.30493179e-01 -1.44716263e-01
-3.79640609e-01 8.78406703e-01 -4.38917935e-01 1.62850201e+00
-1.34489447e-01 9.35941875e-01 -2.29100093e-01 -6.15181863e-01
6.55487359e-01 3.04968625e-01 3.41762513e-01 -4.51715022e-01
-8.26254562e-02 4.50487831e-04 -3.54612410e-01 -5.04516363e-01
1.24696207e+00 1.47876784e-01 -5.09959280e-01 1.03569043e+00
-1.08286530e-01 -6.68474019e-01 8.85880888e-01 8.62963438e-01
1.23825347e+00 -2.02369824e-01 7.59624660e-01 -1.76262513e-01
-5.60610406e-02 6.49904847e-01 2.57854909e-01 1.00192273e+00
2.46484101e-01 7.35457838e-01 1.03205037e+00 4.73274738e-02
-9.86159861e-01 -7.68299699e-01 -2.75043324e-02 5.94971180e-01
-2.16737494e-01 -1.14604676e+00 -6.88650548e-01 -5.50750375e-01
1.51254281e-01 1.17611313e+00 -4.93548900e-01 9.35371891e-02
-3.18432033e-01 -2.67595589e-01 6.33392870e-01 2.54482001e-01
4.83886361e-01 -1.02514768e+00 -7.69093573e-01 1.25386655e-01
-5.61389983e-01 -1.12695968e+00 -8.30773413e-01 -1.61365390e-01
-1.04035330e+00 -1.17562413e+00 -7.05841780e-01 -4.77526218e-01
2.10149810e-01 5.51865160e-01 1.37679970e+00 -2.49962762e-01
2.50597447e-01 6.40490115e-01 -6.13007605e-01 -6.93911254e-01
-4.96758103e-01 7.20730364e-01 -3.82684201e-01 -6.39815569e-01
1.93232059e-01 -2.59395123e-01 -2.79234052e-01 -2.17357337e-01
-9.50580955e-01 1.95523322e-01 7.20703721e-01 5.39228022e-01
3.77192885e-01 -2.91080505e-01 9.77051318e-01 -1.27162063e+00
1.74954593e+00 -7.00729907e-01 -1.27752610e-02 5.38118243e-01
-7.15288401e-01 1.76635846e-01 4.92502123e-01 -4.53874050e-03
-1.20082331e+00 -9.07681108e-01 2.37789065e-01 4.67947692e-01
1.73878446e-01 1.24519897e+00 1.38253152e-01 6.37051821e-01
6.33298635e-01 5.25214911e-01 7.07520097e-02 -4.65146422e-01
5.08726656e-01 6.55877054e-01 5.71674287e-01 -4.40872967e-01
4.98933524e-01 1.64548829e-01 -3.18289280e-01 -1.23575747e+00
-1.33043611e+00 -6.35313988e-01 -1.44440353e-01 -1.13787062e-01
4.26047057e-01 -7.49955118e-01 -2.15396792e-01 4.03684378e-03
-1.03123271e+00 7.12296516e-02 -6.71107054e-01 2.27608100e-01
-2.99743295e-01 7.41318107e-01 -4.01384443e-01 -2.13198334e-01
-1.02336919e+00 -6.03307486e-01 8.58090043e-01 5.42921007e-01
-8.98434401e-01 -8.14777136e-01 2.72280514e-01 6.64845824e-01
4.58501637e-01 3.49328429e-01 1.07186043e+00 -9.72790539e-01
-3.84626865e-01 -4.52265471e-01 -1.36898786e-01 1.88112020e-01
2.73932099e-01 2.46660531e-01 -2.16258660e-01 7.33224973e-02
6.39610216e-02 -4.15549308e-01 1.08292532e+00 6.27983987e-01
6.30258560e-01 -9.16998446e-01 -3.01637977e-01 1.52528137e-01
9.05498147e-01 -1.65263891e-01 4.58883375e-01 4.95415181e-01
4.42722142e-01 5.44794500e-01 7.09176481e-01 7.56474435e-01
6.80566788e-01 9.97295305e-02 -5.57809928e-03 4.61082578e-01
-3.64012718e-01 -4.05535996e-01 2.39369780e-01 1.11275148e+00
2.03893617e-01 -5.44122934e-01 -7.29781091e-01 7.43420780e-01
-1.70688009e+00 -1.29351389e+00 -1.81766927e-01 1.59389269e+00
8.50801110e-01 1.32261977e-01 4.48454134e-02 -1.77957863e-01
3.44465762e-01 7.02215612e-01 -6.05543256e-01 -7.07233012e-01
-3.68999749e-01 -9.80508998e-02 9.17038992e-02 1.43729553e-01
-4.39585865e-01 8.17547381e-01 6.51674795e+00 5.90176642e-01
-7.91105092e-01 -5.25221288e-01 1.58598632e-01 -2.45650947e-01
-8.23878467e-01 1.70163348e-01 -9.46024477e-01 2.51642257e-01
7.76161909e-01 -1.08632982e+00 3.65634561e-02 8.42047751e-01
4.99796122e-01 -4.58835781e-01 -1.06378055e+00 4.84237760e-01
4.32954609e-01 -1.84953082e+00 5.43163955e-01 -3.35266977e-03
1.03385174e+00 -7.99881443e-02 -1.18609577e-01 3.31361145e-01
8.88296217e-02 -8.48248720e-01 6.15889490e-01 6.22079253e-01
5.74371099e-01 -6.28953516e-01 7.28798091e-01 4.36026782e-01
-7.52911806e-01 2.40889937e-01 -2.34940514e-01 -1.30279198e-01
3.66249084e-01 6.82172358e-01 -9.51668561e-01 7.93762267e-01
3.36368769e-01 1.18267214e+00 -6.38963699e-01 1.11593437e+00
-4.20416474e-01 9.25071537e-01 1.54717937e-01 -5.02568305e-01
3.50728214e-01 -3.64445120e-01 9.91730511e-01 1.40511096e+00
3.39693904e-01 1.33112907e-01 5.53796031e-02 7.99074769e-01
-4.14942384e-01 2.49687180e-01 -7.26421118e-01 -7.14224100e-01
7.32607782e-01 1.19536245e+00 -4.57594126e-01 -5.48268259e-01
-4.34934229e-01 5.05885005e-01 1.14156730e-01 3.58892709e-01
-1.83051452e-01 -6.06225789e-01 2.74043441e-01 2.14567721e-01
2.46212468e-01 -3.11889678e-01 -8.87482405e-01 -1.34500897e+00
1.69283345e-01 -1.24743986e+00 3.03482711e-01 -7.53988981e-01
-1.00551462e+00 3.54037732e-01 3.44494820e-01 -7.75087893e-01
-2.75793523e-01 2.21936271e-01 -1.04453397e+00 6.10323131e-01
-1.32260358e+00 -7.21793890e-01 -3.16393524e-01 -5.39571680e-02
1.02442777e+00 -4.33452092e-02 5.69630384e-01 -4.51562792e-01
-6.22398376e-01 1.37361273e-01 2.43473187e-01 -1.29991189e-01
9.27909136e-01 -1.16276515e+00 6.35616899e-01 7.90967762e-01
-1.55221662e-02 9.13509548e-01 9.30865526e-01 -1.06239474e+00
-1.56432235e+00 -7.72110164e-01 1.28965795e+00 -5.47772169e-01
6.29129708e-01 2.86582649e-01 -9.23529804e-01 6.27303183e-01
9.68491316e-01 -1.16301930e+00 9.77120817e-01 4.47438806e-01
-1.04087561e-01 1.33525297e-01 -7.82162428e-01 7.14017868e-01
7.71945715e-01 -3.11924845e-01 -1.39815784e+00 1.70708194e-01
8.72840703e-01 -3.08959991e-01 -8.22293103e-01 4.35146615e-02
5.37194073e-01 -8.10136437e-01 4.93845582e-01 -4.59650040e-01
1.10346270e+00 1.06640533e-01 2.14571968e-01 -1.71906424e+00
-1.18082933e-01 -7.62747169e-01 -3.89390290e-01 1.31474698e+00
4.86092389e-01 -5.17867327e-01 6.05950177e-01 5.07940233e-01
-6.83243155e-01 -8.62865806e-01 -3.38584453e-01 -4.86952960e-01
4.53575701e-02 -1.47200331e-01 6.93088591e-01 4.66364682e-01
5.91693342e-01 8.57591808e-01 -1.45383403e-01 -5.09881616e-01
5.05167902e-01 6.52859747e-01 1.14066577e+00 -1.15904510e+00
2.50229329e-01 -7.10886717e-01 3.02513212e-01 -1.16278088e+00
6.56329617e-02 -8.70458186e-01 -3.30640227e-01 -2.53298855e+00
8.71650279e-01 3.90801460e-01 6.43333972e-01 2.86517590e-01
-4.28968698e-01 -3.03624034e-01 1.80157930e-01 3.11868072e-01
-9.60038960e-01 6.40342712e-01 1.52027619e+00 -1.75140351e-01
-3.81193072e-01 -5.42828105e-02 -1.69853520e+00 5.06031394e-01
6.93052709e-01 -3.23738784e-01 -4.52564031e-01 -1.76223055e-01
4.71357137e-01 4.22797978e-01 -1.83402240e-01 -7.45945215e-01
3.95226479e-01 -3.08522344e-01 1.65355355e-01 -1.25894892e+00
-1.20681850e-02 -6.13646507e-02 -2.15550005e-01 2.20620677e-01
-6.29438818e-01 3.29635382e-01 1.32569849e-01 5.36054611e-01
-3.00413966e-01 -2.08641961e-01 9.29079130e-02 -2.37772688e-01
-2.69536197e-01 1.56473462e-02 -5.49411535e-01 8.69399011e-01
5.75121641e-01 -3.41381192e-01 -9.74130988e-01 -7.62335718e-01
2.24690512e-02 4.98585492e-01 6.62585616e-01 3.36988091e-01
6.39592052e-01 -7.83423364e-01 -1.05138552e+00 -3.07395130e-01
1.65767476e-01 5.34657501e-02 3.86310741e-02 7.25297451e-01
-3.41307908e-01 9.86166358e-01 -2.38243669e-01 -1.75635964e-01
-1.09120393e+00 1.98522046e-01 -5.20890653e-01 -5.38792372e-01
-9.74691153e-01 9.29464623e-02 -1.44264519e-01 -3.81192118e-02
1.01870537e-01 -4.78319138e-01 -5.47648251e-01 7.13926077e-01
7.69852698e-01 7.29052484e-01 -1.31448535e-02 -4.39451814e-01
-3.98421884e-02 2.85663128e-01 -4.20550168e-01 -1.50320917e-01
1.48566246e+00 -2.94273376e-01 -2.07573667e-01 5.02885401e-01
8.84630919e-01 5.02433538e-01 -8.56245220e-01 -1.82075754e-01
3.34287196e-01 -1.59953207e-01 -1.67444006e-01 -9.42616642e-01
-3.85265261e-01 4.46744621e-01 -8.13481867e-01 4.54506010e-01
7.65094399e-01 1.72105774e-01 1.09761512e+00 6.70095265e-01
-2.65605479e-01 -1.03915417e+00 4.14914578e-01 8.91638815e-01
1.24105239e+00 -8.59114408e-01 7.06529200e-01 2.54110992e-02
-1.17660403e+00 1.12110794e+00 2.25940570e-01 4.33168709e-02
1.22222759e-01 2.97980066e-02 -1.87586501e-01 -6.93594337e-01
-9.75511134e-01 1.55501068e-03 5.89147806e-01 3.35605323e-01
5.81375182e-01 3.12073994e-02 -7.53597617e-01 9.75799501e-01
-6.65939271e-01 -1.63937435e-01 1.22121847e+00 9.37932789e-01
-8.55655134e-01 -7.88274229e-01 -2.24988416e-01 1.17255139e+00
-7.40452707e-01 -2.37274736e-01 -9.30129647e-01 6.36360407e-01
-9.95713651e-01 1.20300293e+00 -3.49515975e-01 1.24284357e-01
4.50208545e-01 1.09030008e-01 3.10619235e-01 -1.10099173e+00
-8.71268868e-01 -1.30657405e-01 6.06641352e-01 -1.62391603e-01
-2.65916169e-01 -8.82396758e-01 -1.16189492e+00 -7.17135847e-01
-8.27226192e-02 6.19878709e-01 6.46538198e-01 8.31935108e-01
7.29997277e-01 6.25048280e-01 2.08909690e-01 -6.58538878e-01
-6.53090537e-01 -1.12389827e+00 -4.02688533e-01 5.91782965e-02
3.59156489e-01 -8.55393931e-02 -1.84623674e-01 -1.26361519e-01] | [12.506623268127441, 9.5230131149292] |
7d80ab4d-5864-423e-ac6f-d9ddebfa594c | multi-label-image-recognition-by-recurrently | 1711.02816 | null | http://arxiv.org/abs/1711.02816v1 | http://arxiv.org/pdf/1711.02816v1.pdf | Multi-label Image Recognition by Recurrently Discovering Attentional Regions | This paper proposes a novel deep architecture to address multi-label image
recognition, a fundamental and practical task towards general visual
understanding. Current solutions for this task usually rely on an extra step of
extracting hypothesis regions (i.e., region proposals), resulting in redundant
computation and sub-optimal performance. In this work, we achieve the
interpretable and contextualized multi-label image classification by developing
a recurrent memorized-attention module. This module consists of two alternately
performed components: i) a spatial transformer layer to locate attentional
regions from the convolutional feature maps in a region-proposal-free way and
ii) an LSTM (Long-Short Term Memory) sub-network to sequentially predict
semantic labeling scores on the located regions while capturing the global
dependencies of these regions. The LSTM also output the parameters for
computing the spatial transformer. On large-scale benchmarks of multi-label
image classification (e.g., MS-COCO and PASCAL VOC 07), our approach
demonstrates superior performances over other existing state-of-the-arts in
both accuracy and efficiency. | ['Zhouxia Wang', 'Liang Lin', 'Tianshui Chen', 'Ruijia Xu', 'Guanbin Li'] | 2017-11-08 | multi-label-image-recognition-by-recurrently-1 | http://openaccess.thecvf.com/content_iccv_2017/html/Wang_Multi-Label_Image_Recognition_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Wang_Multi-Label_Image_Recognition_ICCV_2017_paper.pdf | iccv-2017-10 | ['multi-label-image-classification'] | ['computer-vision'] | [ 5.29272974e-01 -1.85453296e-02 -2.28356063e-01 -6.02937698e-01
-1.15515041e+00 -4.45881039e-01 4.29240614e-01 3.16982538e-01
-5.29809058e-01 3.44888598e-01 -1.03648476e-01 -3.60048622e-01
3.08762848e-01 -4.20567006e-01 -8.90495002e-01 -8.47894371e-01
4.31128234e-01 2.69149214e-01 1.38446108e-01 2.85793692e-01
5.28219998e-01 5.15832901e-01 -1.74687123e+00 9.06549156e-01
4.75350648e-01 1.30480051e+00 4.06958044e-01 5.84089220e-01
-1.03926882e-01 1.12605047e+00 -2.87424654e-01 3.71108912e-02
-2.39374593e-01 -4.04258966e-01 -9.42969978e-01 1.46773905e-01
5.61245501e-01 -1.06698200e-01 1.42773986e-01 1.03211629e+00
2.68833905e-01 1.07459515e-01 7.86282480e-01 -8.27804804e-01
-5.79599380e-01 2.83891946e-01 -8.43253732e-01 1.48530409e-01
-2.65982598e-02 -1.41398981e-01 1.13531625e+00 -1.19354844e+00
4.45061624e-01 1.15799499e+00 4.70627606e-01 4.89775777e-01
-1.16561544e+00 -6.14278078e-01 3.77577126e-01 3.53500098e-01
-1.50700116e+00 -3.63663465e-01 5.42511702e-01 -3.40480566e-01
1.01386642e+00 3.06128711e-01 1.94772914e-01 1.00122213e+00
4.02688861e-01 8.90316188e-01 1.16538334e+00 -5.28624237e-01
1.03408672e-01 2.65382379e-01 2.13546097e-01 9.66364682e-01
-3.24543446e-01 -5.31093851e-02 -3.91200572e-01 -2.64478419e-02
6.48751855e-01 2.85321862e-01 -1.04072494e-02 -3.14342886e-01
-1.08523357e+00 8.93405139e-01 7.28725374e-01 3.59005988e-01
-4.20939386e-01 3.43575358e-01 4.97280836e-01 8.07656795e-02
6.36218250e-01 1.87893853e-01 -5.25382876e-01 5.59997082e-01
-1.07088554e+00 3.18109430e-02 1.44792169e-01 7.90699720e-01
1.06468153e+00 -3.02693546e-01 -5.07522583e-01 1.00461531e+00
4.90778685e-01 -2.47658808e-02 5.50015688e-01 -4.80253190e-01
5.37369430e-01 8.45989227e-01 -5.86814284e-02 -7.15799391e-01
-6.85476780e-01 -7.29420424e-01 -8.30253124e-01 1.60983548e-01
2.45937452e-01 2.42357269e-01 -1.21844828e+00 1.57620847e+00
3.60430479e-01 2.56426752e-01 -2.13513166e-01 9.55620766e-01
8.18634093e-01 6.25561833e-01 5.86613178e-01 9.55793336e-02
1.67795479e+00 -1.48562944e+00 -3.80644649e-01 -5.25556326e-01
8.84617925e-01 -6.64917886e-01 9.12708759e-01 1.55573329e-02
-9.75636780e-01 -5.14555156e-01 -9.89177287e-01 -3.91350180e-01
-6.27325118e-01 5.85934877e-01 2.90203363e-01 6.30066618e-02
-1.09584761e+00 5.04609346e-01 -3.91411930e-01 -2.41601363e-01
6.95542693e-01 3.33853930e-01 -3.55296195e-01 -1.08354457e-01
-8.70574176e-01 8.53532016e-01 5.49773395e-01 1.84520215e-01
-1.10593724e+00 -6.26774192e-01 -9.99589324e-01 3.15328032e-01
1.81371197e-01 -4.30172473e-01 1.14921331e+00 -1.35450923e+00
-1.15490830e+00 1.28574049e+00 -3.84692401e-01 -3.73076558e-01
2.41826981e-01 1.50643447e-02 -9.73401368e-02 2.79433429e-01
3.11986715e-01 1.05151105e+00 9.21926856e-01 -1.40459704e+00
-9.61262584e-01 -4.90088612e-01 -1.69415131e-01 2.46485159e-01
-1.77745581e-01 2.70776987e-01 -4.34725434e-01 -7.33798087e-01
2.08374545e-01 -1.03096366e+00 -3.37598771e-01 -1.40878499e-01
-5.62398076e-01 -4.24374342e-01 6.41093731e-01 -6.62214994e-01
9.07046378e-01 -2.06560826e+00 3.29531692e-02 1.19884141e-01
1.80255800e-01 1.77764878e-01 -2.13713631e-01 1.86876524e-02
-4.26468372e-01 7.00330501e-03 5.18512949e-02 -6.00748420e-01
-2.27651924e-01 -1.82050586e-01 -3.40797484e-01 7.07021832e-01
4.02235121e-01 1.24339283e+00 -6.65652335e-01 -7.54321516e-01
3.74599814e-01 4.20695186e-01 -2.57489979e-01 2.61428446e-01
-3.79447281e-01 5.00550866e-01 -3.34441900e-01 8.08761537e-01
4.34223771e-01 -7.68934429e-01 1.84000567e-01 -3.63199800e-01
-2.52063990e-01 -3.62769552e-02 -7.30728984e-01 1.82085943e+00
-7.24268317e-01 4.21193242e-01 -5.55696711e-02 -1.05761027e+00
6.70302927e-01 3.26724321e-01 2.97463626e-01 -8.72698069e-01
3.73987228e-01 1.36101380e-01 -4.91334856e-01 -2.73303926e-01
5.14407337e-01 1.02136515e-01 -4.87951562e-02 5.58584213e-01
1.46543309e-01 7.26751268e-01 -1.66438475e-01 -1.96471453e-01
9.18490052e-01 1.59032851e-01 4.62963402e-01 -3.98165643e-01
6.10455751e-01 -2.34657928e-01 4.51229155e-01 5.41189969e-01
-2.33038291e-01 6.04691863e-01 3.90482455e-01 -6.77760482e-01
-9.58425581e-01 -7.07303882e-01 2.28832606e-02 1.66268897e+00
2.39885420e-01 -1.41434968e-01 -5.92284024e-01 -1.01117623e+00
-9.56701785e-02 6.31419659e-01 -1.11442876e+00 7.25618601e-02
-6.34187877e-01 -5.61829448e-01 3.35387558e-01 7.29828119e-01
4.01026160e-01 -1.21279275e+00 -8.15167069e-01 1.24118567e-01
-5.68457842e-02 -1.02751780e+00 -5.57665050e-01 6.88109875e-01
-5.89756787e-01 -1.01943302e+00 -7.33644783e-01 -1.13925767e+00
8.75766814e-01 4.39815074e-01 9.72663105e-01 1.06650159e-01
-4.00227576e-01 -2.18338877e-01 -3.57411467e-02 -7.86783546e-02
-1.27425611e-01 2.00211883e-01 -8.08376372e-01 3.62126470e-01
2.59095103e-01 -1.69686571e-01 -8.78339231e-01 5.82816958e-01
-6.86590731e-01 4.62001771e-01 1.02124321e+00 1.07958853e+00
1.11086273e+00 -4.55521673e-01 6.86316431e-01 -9.85920548e-01
5.33265471e-02 -5.11381388e-01 -5.71810484e-01 7.31729150e-01
-5.13801396e-01 2.36699894e-01 7.34162748e-01 -2.45178252e-01
-9.98325408e-01 4.49198186e-01 -2.08383091e-02 -5.93009651e-01
-3.78154516e-01 1.88889340e-01 -9.18062702e-02 -1.93996266e-01
3.12923968e-01 3.71100545e-01 -2.47546613e-01 -5.50041318e-01
4.52265501e-01 6.93707883e-01 2.70671904e-01 -2.80679852e-01
2.23962098e-01 4.39628839e-01 -1.04121147e-02 -4.59712833e-01
-1.37321234e+00 -7.88176835e-01 -6.81831002e-01 -3.59765530e-01
1.27903736e+00 -9.93832409e-01 -8.74769747e-01 3.10318470e-01
-1.27322555e+00 -2.68408626e-01 1.73839167e-01 5.97498938e-03
-4.54382628e-01 -1.69452205e-02 -8.59152853e-01 -4.83976424e-01
-5.64544261e-01 -1.39801800e+00 1.60848784e+00 3.29733789e-01
9.50078107e-03 -8.68702233e-01 -7.25473389e-02 5.83825469e-01
2.53395140e-01 1.87544614e-01 1.07843649e+00 -6.57361448e-01
-6.72218382e-01 4.92857620e-02 -8.15820575e-01 3.63218896e-02
-3.03180248e-01 -4.05338645e-01 -1.44219863e+00 -4.57298428e-01
-2.40420774e-01 -6.77239954e-01 1.19313180e+00 2.62434930e-01
1.60833848e+00 -2.80092180e-01 -7.18678951e-01 5.96987069e-01
1.60452259e+00 1.36962980e-01 4.48293090e-01 2.71957666e-01
8.71572196e-01 6.38758242e-01 8.36693585e-01 1.35052741e-01
4.89702523e-01 8.18727553e-01 6.84488773e-01 -4.97210085e-01
-1.00992754e-01 -5.96149042e-02 -1.02628507e-01 5.19885063e-01
4.43971157e-01 -1.93631947e-01 -7.63927877e-01 5.44620454e-01
-2.16318226e+00 -7.88435102e-01 1.81696951e-01 1.87128687e+00
6.36547685e-01 -1.61582157e-01 -1.77076623e-01 -1.96840748e-01
8.90817702e-01 4.20891345e-01 -6.88893259e-01 -5.30672371e-01
1.37100950e-01 1.53442517e-01 6.53030276e-01 2.74906427e-01
-1.59652376e+00 1.13291502e+00 5.39381790e+00 9.43503439e-01
-1.18644345e+00 3.54432017e-01 1.08809292e+00 -1.33352324e-01
-6.96364418e-02 -2.20970988e-01 -1.01108229e+00 2.28252843e-01
9.50275660e-01 6.25700772e-01 2.18020245e-01 9.25953805e-01
-1.17996968e-01 -1.26610801e-01 -1.04873788e+00 9.57609415e-01
3.60618770e-01 -1.34952223e+00 1.90233767e-01 8.65957513e-02
6.08874023e-01 5.22427447e-02 2.45287836e-01 1.62500247e-01
3.04005500e-02 -1.29262877e+00 9.87948060e-01 4.61767524e-01
1.02858257e+00 -8.57077003e-01 6.28003955e-01 2.73748308e-01
-1.43733525e+00 -3.36518556e-01 -3.25446695e-01 4.57033753e-01
-1.85671389e-01 3.77777308e-01 -7.71702051e-01 4.54975814e-01
4.55629140e-01 7.25204527e-01 -7.65584469e-01 7.00743675e-01
-8.73571336e-02 3.09336960e-01 1.72961503e-01 1.09982371e-01
6.81591988e-01 1.53767869e-01 -1.37006998e-01 1.30531585e+00
6.63305297e-02 -1.90645024e-01 2.01479241e-01 8.46435130e-01
-2.22678617e-01 2.81737924e-01 -3.68259728e-01 2.90276587e-01
2.45332122e-01 1.63980818e+00 -1.03081036e+00 -3.08836222e-01
-5.78359067e-01 1.02595031e+00 1.03735030e+00 3.61678183e-01
-1.14963615e+00 -2.25537866e-01 5.02172649e-01 -1.62332490e-01
5.95045030e-01 1.94541022e-01 -3.96835744e-01 -8.59347105e-01
-2.49988839e-01 -7.35691488e-01 6.06158257e-01 -5.91315091e-01
-7.90998042e-01 6.65400088e-01 -5.15022099e-01 -9.69733953e-01
-1.47258788e-01 -5.75006068e-01 -4.43057001e-01 9.76584435e-01
-1.71347177e+00 -1.74311674e+00 -3.18155169e-01 3.99899840e-01
7.48412967e-01 8.02773684e-02 1.06737769e+00 3.79825979e-01
-8.56698871e-01 8.15621138e-01 9.57317725e-02 -1.30573995e-02
6.96066439e-01 -1.20216811e+00 2.73448497e-01 6.18377268e-01
4.01139557e-01 2.95339912e-01 1.78057358e-01 -3.17195743e-01
-8.71392488e-01 -1.62711167e+00 8.63328218e-01 -2.78080732e-01
2.82264739e-01 -5.38603067e-01 -8.30843985e-01 8.28666985e-01
2.51915783e-01 2.26153359e-01 7.34748363e-01 -7.54045025e-02
-5.03962874e-01 -9.28597003e-02 -9.58015203e-01 4.85280842e-01
7.02407777e-01 -6.22636914e-01 -2.31505707e-01 4.04597163e-01
5.51160395e-01 -2.65900612e-01 -8.02468479e-01 3.22940350e-01
6.45626009e-01 -8.44053805e-01 1.04260147e+00 -4.46586788e-01
4.17495906e-01 -3.27641875e-01 -5.46327084e-02 -1.08544970e+00
-5.62044144e-01 5.87955713e-02 1.11486860e-01 9.65947270e-01
5.87582409e-01 -4.10301834e-01 7.11388826e-01 1.62602216e-01
-2.80071169e-01 -1.32461166e+00 -8.99653733e-01 -1.40523165e-01
-1.93595126e-01 -7.96941668e-02 3.79518420e-01 8.84354115e-01
-3.28939348e-01 5.43348610e-01 -2.42196694e-01 1.85588822e-01
6.84696972e-01 5.16485095e-01 2.08925918e-01 -8.53306293e-01
-8.46750438e-02 -4.75968868e-01 -1.31390825e-01 -1.07471621e+00
5.55876970e-01 -1.05609226e+00 1.88565359e-01 -1.50552833e+00
5.73556304e-01 -4.63018477e-01 -9.02478099e-01 7.76236892e-01
-2.41791084e-01 3.95133227e-01 -3.02435947e-03 2.76834697e-01
-1.18301392e+00 2.47217149e-01 1.09521580e+00 -2.58755088e-01
1.44636884e-01 -2.07852811e-01 -5.55218697e-01 6.63281500e-01
7.57817149e-01 -5.61068714e-01 -2.69732654e-01 -4.12793666e-01
-4.57574055e-02 1.38080627e-01 5.35722733e-01 -8.82373810e-01
3.20260465e-01 -7.38958716e-02 6.88334703e-01 -8.77409756e-01
3.49336147e-01 -6.47629559e-01 -9.08218473e-02 3.80801976e-01
-7.57090032e-01 1.98309720e-01 1.79978788e-01 6.71361446e-01
-1.85554400e-01 -2.20748484e-01 1.09158337e+00 -2.65596002e-01
-8.32722962e-01 2.14229286e-01 -2.13579431e-01 -2.35862970e-01
1.32773304e+00 1.56238712e-02 -5.18782854e-01 1.82954818e-01
-5.90147138e-01 2.51398623e-01 4.44814056e-01 4.54315931e-01
6.86709404e-01 -1.23060167e+00 -4.21984255e-01 2.89767861e-01
4.04355437e-01 -9.27888602e-02 4.89125073e-01 7.38936186e-01
-4.29372042e-01 7.88446009e-01 -1.22916974e-01 -7.49207258e-01
-1.42008829e+00 7.44511127e-01 4.94974762e-01 -7.57177472e-01
-5.34496725e-01 1.02815306e+00 6.52306974e-01 -1.60992607e-01
2.68270403e-01 -3.04550648e-01 -4.11756068e-01 2.11173534e-01
6.89571321e-01 -9.09429230e-03 1.87196463e-01 -8.16057861e-01
-5.63184440e-01 7.61224687e-01 -4.82939839e-01 1.84267536e-01
1.11254835e+00 -1.23396054e-01 -1.76672876e-01 5.17557859e-01
1.59500420e+00 -6.06873035e-01 -1.14170384e+00 -2.12826908e-01
3.50769222e-01 -1.90873131e-01 1.71531647e-01 -8.89365613e-01
-1.17097282e+00 9.15482879e-01 7.34395921e-01 -1.63187072e-01
1.05146432e+00 1.30700931e-01 7.33923674e-01 6.61930591e-02
2.63001978e-01 -1.10224962e+00 1.58603847e-01 3.03809851e-01
6.72750533e-01 -1.38661337e+00 -2.65648335e-01 -1.13536365e-01
-5.80813468e-01 9.55979764e-01 7.52040625e-01 5.20025790e-02
5.14229715e-01 -7.87971392e-02 1.78353176e-01 -3.73594910e-01
-1.04184389e+00 -1.33218259e-01 5.70491016e-01 -1.21404231e-02
5.60133100e-01 5.27423695e-02 -1.60102919e-02 3.07842374e-01
5.61507106e-01 -4.18191493e-01 -1.10548638e-01 6.90388560e-01
-6.51722968e-01 -6.67246401e-01 -1.88490286e-01 6.11418962e-01
-6.37552142e-01 -1.07953817e-01 -1.06553346e-01 5.69820166e-01
1.46335125e-01 7.89767325e-01 1.74764365e-01 -3.02036881e-01
7.10938349e-02 2.62084007e-01 3.37759614e-01 -6.58139467e-01
-9.43397164e-01 5.56706786e-02 -5.28197885e-02 -8.71436775e-01
-2.68577307e-01 -4.16670769e-01 -1.14724278e+00 1.52594551e-01
-4.49996889e-01 -2.99942065e-02 7.96493828e-01 9.42731202e-01
4.33347970e-01 7.68963039e-01 6.09576225e-01 -8.88423562e-01
-4.58050996e-01 -1.02834165e+00 -3.89887005e-01 4.02080178e-01
2.86838561e-01 -5.97521245e-01 -4.65798490e-02 -2.14111611e-01] | [9.82436466217041, 3.98407244682312] |
fadd0668-d18a-44d5-9f78-458f5d3ea468 | generative-imagination-elevates-machine | 2009.09654 | null | https://arxiv.org/abs/2009.09654v2 | https://arxiv.org/pdf/2009.09654v2.pdf | Generative Imagination Elevates Machine Translation | There are common semantics shared across text and images. Given a sentence in a source language, whether depicting the visual scene helps translation into a target language? Existing multimodal neural machine translation methods (MNMT) require triplets of bilingual sentence - image for training and tuples of source sentence - image for inference. In this paper, we propose ImagiT, a novel machine translation method via visual imagination. ImagiT first learns to generate visual representation from the source sentence, and then utilizes both source sentence and the "imagined representation" to produce a target translation. Unlike previous methods, it only needs the source sentence at the inference time. Experiments demonstrate that ImagiT benefits from visual imagination and significantly outperforms the text-only neural machine translation baselines. Further analysis reveals that the imagination process in ImagiT helps fill in missing information when performing the degradation strategy. | ['Lei LI', 'Quanyu Long', 'Mingxuan Wang'] | 2020-09-21 | null | https://aclanthology.org/2021.naacl-main.457 | https://aclanthology.org/2021.naacl-main.457.pdf | naacl-2021-4 | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 6.23640060e-01 4.50089365e-01 -3.59254599e-01 -3.60569715e-01
-7.62695372e-01 -8.16647768e-01 1.01641762e+00 -5.05773544e-01
-9.61505622e-02 8.05879593e-01 3.42583805e-01 -5.25049388e-01
9.63833690e-01 -5.59945107e-01 -1.35561359e+00 -3.35498035e-01
8.37428987e-01 5.15451252e-01 -4.57227260e-01 -1.34970814e-01
1.47109434e-01 3.13360244e-02 -8.29866946e-01 9.72505391e-01
8.43750715e-01 6.84685409e-01 6.96950674e-01 4.89368826e-01
-3.28877419e-01 1.03762114e+00 -5.90247035e-01 -7.52888918e-01
2.33359486e-01 -1.23722363e+00 -7.86657691e-01 3.69831741e-01
9.44523633e-01 -6.67605579e-01 -3.79765481e-01 1.07028508e+00
3.08162630e-01 -2.53040433e-01 7.61518300e-01 -1.42496586e+00
-1.58175170e+00 4.98333335e-01 -6.65257692e-01 5.52951992e-02
6.35679483e-01 6.16936386e-01 6.46849573e-01 -1.41664112e+00
1.07706511e+00 1.55975795e+00 2.87643354e-02 8.47871602e-01
-1.53229392e+00 -3.72730821e-01 -1.61185861e-01 8.23870227e-02
-9.54119444e-01 -6.94028854e-01 7.41608441e-01 -2.74733394e-01
1.07757390e+00 2.92964727e-01 9.64380920e-01 1.61474419e+00
3.99397969e-01 9.86372411e-01 1.55154264e+00 -5.24855971e-01
-2.33484924e-01 4.00174558e-01 -4.91288066e-01 1.05436397e+00
-1.02483287e-01 2.88520485e-01 -1.09531665e+00 3.99457633e-01
1.19546986e+00 -2.52967387e-01 -3.70006680e-01 -2.13014364e-01
-1.87285066e+00 6.79455221e-01 4.85246688e-01 8.61822218e-02
-3.96692723e-01 4.92363989e-01 -7.02156648e-02 7.81006932e-01
4.85136420e-01 5.36275923e-01 3.80298942e-02 3.37498970e-02
-9.87052381e-01 -2.63793588e-01 3.03339958e-01 1.05561757e+00
7.51122296e-01 3.30089867e-01 -5.42442977e-01 5.50702989e-01
2.38116518e-01 1.18274021e+00 3.82049173e-01 -1.13402033e+00
9.05538082e-01 5.70469320e-01 1.38928667e-01 -8.03635001e-01
2.44180769e-01 -1.15690857e-01 -5.60574830e-01 1.17856406e-01
4.18936938e-01 -1.59614801e-01 -9.26074922e-01 1.91975355e+00
8.03519133e-03 -7.60973692e-02 4.80137080e-01 1.22581172e+00
9.00719821e-01 8.85266721e-01 -1.55376181e-01 -2.94268638e-01
1.35225463e+00 -1.37502348e+00 -8.06440711e-01 -7.55541027e-01
3.20355594e-01 -9.15629566e-01 1.69356108e+00 -2.13036895e-01
-1.70533228e+00 -6.01600707e-01 -9.93219674e-01 -6.39707506e-01
-5.77971041e-02 4.53769118e-01 3.48420829e-01 1.39649123e-01
-1.49390495e+00 1.60468191e-01 -3.95596743e-01 -4.44419414e-01
4.06463742e-01 -8.03426728e-02 -5.14055252e-01 -1.83967695e-01
-1.07341301e+00 1.34277689e+00 -2.87194327e-02 5.01088537e-02
-1.27891827e+00 -3.79619569e-01 -1.02361667e+00 -2.08062857e-01
2.55349800e-02 -1.52576625e+00 1.39376640e+00 -1.94979441e+00
-1.52117074e+00 1.28818309e+00 -6.73927844e-01 -3.85555655e-01
5.24193585e-01 -1.02507293e-01 1.17724210e-01 6.61755681e-01
2.74327695e-01 1.61733973e+00 1.26310396e+00 -1.52949536e+00
-1.31229954e-02 -2.97216535e-01 1.29124716e-01 6.53357565e-01
-4.95238565e-02 -4.05594222e-02 -5.17275870e-01 -5.81282794e-01
2.67580412e-02 -8.32108617e-01 3.02539170e-01 3.25288236e-01
-5.50361991e-01 1.84081674e-01 4.42463666e-01 -8.10878813e-01
4.21146691e-01 -1.95773351e+00 6.79169953e-01 -4.66977626e-01
1.33018285e-01 -3.36221099e-01 -6.20377302e-01 5.31773686e-01
-5.54072671e-02 4.05160105e-03 9.28240828e-03 -4.73222613e-01
4.79164347e-02 1.78887770e-01 -8.69872630e-01 3.19746882e-01
4.05545563e-01 1.89342523e+00 -9.18910086e-01 -8.25896502e-01
2.03739271e-01 4.36763555e-01 1.08534507e-02 3.24755013e-01
-4.78499800e-01 8.08393657e-01 -6.06595241e-02 4.33875769e-01
5.08316040e-01 -3.75992119e-01 6.46936521e-02 -3.98875564e-01
1.06512561e-01 1.12797931e-01 2.69535594e-02 2.18439889e+00
-5.62617958e-01 1.10452175e+00 -2.22057909e-01 -7.22381353e-01
6.46889269e-01 3.67356330e-01 -1.16543189e-01 -1.21094275e+00
5.97457290e-02 1.15197234e-01 -1.94937170e-01 -7.01260746e-01
2.55712211e-01 -4.97730345e-01 4.31034230e-02 7.50370443e-01
-1.83957610e-02 -6.29976451e-01 -5.80832176e-02 4.48449403e-01
4.92795825e-01 5.59537172e-01 -9.79617760e-02 3.40138264e-02
9.73627791e-02 5.27252734e-01 -5.40427491e-02 4.79358763e-01
-5.55661395e-02 7.05777705e-01 4.82293844e-01 -2.84035891e-01
-1.27597725e+00 -1.59414363e+00 5.41106582e-01 8.35239172e-01
3.74832422e-01 2.23666668e-01 -9.98965204e-01 -6.27194464e-01
-3.89677912e-01 1.35769916e+00 -6.72424734e-01 -2.75306523e-01
-5.09220004e-01 -1.53863057e-01 5.43914676e-01 4.22320545e-01
7.62696564e-01 -1.26480603e+00 -7.13300884e-01 -2.19700038e-01
-8.80837619e-01 -1.24507344e+00 -1.03407836e+00 -5.89284062e-01
-9.78386998e-01 -5.66030860e-01 -1.03529477e+00 -1.13231707e+00
1.23071337e+00 6.99150324e-01 1.38263655e+00 -3.49602327e-02
6.60543218e-02 5.71765542e-01 3.98966298e-02 -1.48309305e-01
-7.77879894e-01 -4.93574262e-01 -2.97086328e-01 2.33562645e-02
1.28555939e-01 -3.42433184e-01 -5.85488915e-01 7.71566927e-02
-6.72612011e-01 1.34325826e+00 9.18103456e-01 8.47259164e-01
3.65799785e-01 -8.90376449e-01 2.94854164e-01 -3.32534701e-01
7.13612318e-01 -2.36778826e-01 -1.50040746e-01 5.39055347e-01
-2.22310707e-01 4.61427197e-02 3.17245185e-01 -7.40239918e-01
-1.23408878e+00 7.56301126e-03 5.16649485e-01 -6.44942641e-01
1.36123374e-01 3.61782610e-01 -4.72053401e-02 2.97590762e-01
6.33873165e-01 7.74235606e-01 3.93328905e-01 4.47949693e-02
6.78128481e-01 3.47776115e-01 6.35246217e-01 -5.15458763e-01
7.79638290e-01 6.02931738e-01 -2.18168780e-01 -4.46685255e-01
-8.44058692e-01 4.18290406e-01 -5.86069942e-01 -4.19410795e-01
1.15238070e+00 -1.18624270e+00 -3.96546453e-01 1.73186094e-01
-1.82534039e+00 -5.49073160e-01 -3.43261421e-01 2.93653041e-01
-8.84697020e-01 1.99085712e-01 -5.86504996e-01 -5.67891836e-01
-3.38509858e-01 -1.21568286e+00 1.43983090e+00 -2.02591158e-02
-1.89149469e-01 -9.12067831e-01 -2.61636585e-01 8.72159660e-01
2.83703774e-01 2.16117591e-01 9.33736324e-01 1.16150118e-01
-1.01523125e+00 3.47777992e-01 -7.20500052e-01 9.65815783e-02
-5.13562188e-02 -4.58228290e-01 -6.82435155e-01 -1.94090024e-01
2.32374296e-01 -7.33355582e-01 8.77001524e-01 1.87909052e-01
4.64956880e-01 -6.43741846e-01 -9.99316499e-02 4.56774473e-01
1.23645830e+00 4.08919640e-02 8.23680401e-01 -1.27040550e-01
7.89551854e-01 7.98514187e-01 5.07994890e-01 -4.11517978e-01
6.98463321e-01 5.77436447e-01 2.47443348e-01 -6.64028585e-01
-6.57582879e-01 -7.32215762e-01 1.21564198e+00 9.51875687e-01
1.88790530e-01 -2.27578387e-01 -7.23059118e-01 6.45534456e-01
-1.83421874e+00 -1.01529896e+00 -8.46309215e-02 1.81969321e+00
1.06152785e+00 -3.56129348e-01 -4.01796281e-01 -7.18516111e-01
5.47188818e-01 8.18235651e-02 -7.31506169e-01 -5.61945140e-01
-5.94075441e-01 -1.44371495e-01 8.22183788e-02 6.98648930e-01
-4.17337865e-01 1.54160106e+00 6.35745049e+00 5.15280545e-01
-1.26962638e+00 5.15438855e-01 7.76676059e-01 -3.46619695e-01
-6.79321289e-01 1.19248338e-01 -4.79570031e-02 1.26664296e-01
7.47697949e-01 -7.17436671e-02 9.05880809e-01 1.58992156e-01
5.10235071e-01 -2.70478815e-01 -1.36188698e+00 1.10614693e+00
7.05284774e-01 -1.56019104e+00 6.40038311e-01 -4.26080637e-02
8.37414980e-01 -1.91988841e-01 5.61102867e-01 7.50702471e-02
1.66520938e-01 -1.26958358e+00 1.10849833e+00 6.91370487e-01
1.28202605e+00 -2.00274587e-01 2.40077734e-01 3.78686577e-01
-5.83009839e-01 2.71398842e-01 -1.98967248e-01 -7.34871626e-02
3.92255396e-01 5.72787412e-02 -9.73312199e-01 2.88037091e-01
7.85698444e-02 5.60198486e-01 -7.11588085e-01 -6.17606230e-02
-6.08790159e-01 4.07112300e-01 2.94060796e-01 -5.85022420e-02
2.18298599e-01 -2.68537134e-01 5.15719712e-01 7.98072100e-01
4.19849694e-01 -3.42629641e-01 1.87970269e-02 1.45516407e+00
-2.76495695e-01 2.14253917e-01 -1.05997944e+00 -3.99347335e-01
2.14323044e-01 9.07630861e-01 -3.78515393e-01 -7.44547367e-01
-4.24639940e-01 2.12239385e+00 4.79248047e-01 9.20072973e-01
-9.77192342e-01 2.74763793e-01 1.22484423e-01 5.80373593e-02
-8.10152441e-02 -3.61906052e-01 -6.04017615e-01 -1.50740659e+00
1.58762172e-01 -9.68930662e-01 -3.54742616e-01 -1.96896839e+00
-9.29407060e-01 5.19083202e-01 2.04306599e-02 -1.11876762e+00
-3.55174243e-01 -4.74743634e-01 -5.11250913e-01 1.04092729e+00
-1.15187418e+00 -1.96180534e+00 -1.31749913e-01 5.50857067e-01
1.09525394e+00 -1.51355594e-01 7.23298550e-01 -3.36515874e-01
-1.54452816e-01 4.42655712e-01 -3.91451389e-01 5.91485463e-02
8.83474410e-01 -1.16239154e+00 5.44079363e-01 8.54302526e-01
5.72916687e-01 6.67460978e-01 7.22858846e-01 -7.98300564e-01
-1.84613776e+00 -9.81404126e-01 1.07793820e+00 -7.80646324e-01
4.46019143e-01 -4.15985376e-01 -3.38749319e-01 9.05496538e-01
1.28066933e+00 -5.89772165e-01 4.09174860e-01 -7.00942218e-01
-4.86148059e-01 2.17619091e-01 -9.88545895e-01 1.21365416e+00
1.11758161e+00 -8.88671100e-01 -9.14346635e-01 5.58993459e-01
9.99111950e-01 -2.69910276e-01 -3.80069256e-01 -7.13395700e-02
4.85102475e-01 -6.59898877e-01 1.03526366e+00 -6.68489158e-01
1.14311695e+00 -4.80924487e-01 -2.33192459e-01 -1.56607437e+00
1.58111587e-01 -6.13446176e-01 -2.43506450e-02 7.40583241e-01
7.93133259e-01 -3.47416431e-01 3.40002656e-01 2.63906687e-01
-3.44489068e-02 -4.30266023e-01 -9.79929268e-01 -5.50559759e-01
2.49783218e-01 -1.90489978e-01 3.70115787e-01 1.02568543e+00
1.88364074e-01 1.10706413e+00 -6.53006434e-01 -1.92401245e-01
5.76857388e-01 5.51770687e-01 9.70625103e-01 -3.89989555e-01
-3.50900441e-01 -2.44209722e-01 1.12613104e-01 -1.32454526e+00
3.34263146e-01 -1.24755728e+00 -1.97190903e-02 -1.98088300e+00
7.24007905e-01 6.33002996e-01 2.80183583e-01 6.89463139e-01
-8.07077438e-02 7.25865960e-01 5.21332324e-01 4.43226576e-01
-4.42972511e-01 6.26569450e-01 2.21130586e+00 -4.63973910e-01
2.14794099e-01 -7.29071438e-01 -7.80639470e-01 3.91018301e-01
6.07260108e-01 -4.27559800e-02 -6.61279619e-01 -1.13980746e+00
4.20266330e-01 5.35509765e-01 9.32404995e-01 -3.29777002e-01
-7.40160868e-02 -4.23960805e-01 7.77501941e-01 -5.10888636e-01
6.20534062e-01 -4.94158179e-01 1.65907294e-01 4.85908955e-01
-5.19380748e-01 6.08044803e-01 9.22170654e-02 4.15867865e-01
-3.61885764e-02 4.27781381e-02 6.76892340e-01 -3.60182613e-01
-3.73221815e-01 -1.91386670e-01 -5.11197031e-01 -2.64417450e-03
7.90151954e-01 -3.97606939e-01 -6.20426774e-01 -7.35036016e-01
-7.99882293e-01 3.82008515e-02 9.49276268e-01 5.62692702e-01
1.09518397e+00 -1.66796494e+00 -9.46783066e-01 -8.66244063e-02
5.17279617e-02 -4.67342585e-01 6.95250034e-02 1.13668048e+00
-4.86467868e-01 3.11401755e-01 -5.05435228e-01 -6.54578984e-01
-1.17578053e+00 7.45442867e-01 3.48836690e-01 1.66600585e-01
-3.92137021e-01 7.39160120e-01 6.84714556e-01 -3.17351162e-01
-3.14188778e-01 -1.25478700e-01 3.87962878e-01 -2.97517121e-01
3.69662404e-01 -2.71616846e-01 -7.44219184e-01 -9.45938110e-01
9.14103016e-02 4.91467655e-01 1.63759366e-01 -1.14395547e+00
7.99438298e-01 -5.38501680e-01 -3.55067879e-01 7.60992050e-01
1.05166376e+00 -1.13787577e-01 -1.30619192e+00 -1.37611538e-01
-4.30469811e-01 -4.04198050e-01 -2.62770772e-01 -1.28099525e+00
-9.10458982e-01 1.36774278e+00 4.02273327e-01 -5.19687414e-01
1.07072663e+00 2.16625586e-01 9.15006459e-01 3.13998520e-01
3.51592153e-01 -9.44565535e-01 7.86286950e-01 3.42892140e-01
1.58514929e+00 -1.32676446e+00 -2.64735550e-01 -3.01301360e-01
-1.22657263e+00 8.55026960e-01 8.43237877e-01 1.09809205e-01
-2.52793312e-01 -1.91791356e-01 3.35820466e-01 -1.95808232e-01
-9.88784432e-01 1.88056398e-02 5.18135250e-01 5.73334157e-01
3.54602575e-01 9.29178670e-02 -6.71999231e-02 1.84156373e-01
-2.99261868e-01 -2.37179592e-01 3.93314570e-01 5.73695123e-01
-3.12269866e-01 -8.33592534e-01 -5.27957141e-01 3.46485451e-02
9.77555960e-02 -5.81820190e-01 -1.22094929e+00 4.70030546e-01
1.47980660e-01 9.58082676e-01 1.95884094e-01 -2.45625764e-01
-4.28195521e-02 2.76800752e-01 1.17234910e+00 -5.92281818e-01
-4.72386926e-01 1.08870760e-01 1.32548615e-01 -6.07533872e-01
-4.21089560e-01 -6.52607977e-01 -1.28231084e+00 -1.40548691e-01
7.51845613e-02 -2.87400603e-01 7.86879241e-01 1.02747226e+00
5.24108469e-01 2.31392175e-01 3.84228706e-01 -7.04934120e-01
-1.96101829e-01 -1.10179257e+00 2.36933723e-01 5.05351663e-01
3.17431033e-01 -2.89385080e-01 -1.04999609e-01 7.47079372e-01] | [11.359959602355957, 1.423008918762207] |
bac610d0-8b3b-4df1-a4f6-dd8c513821c3 | 3d-object-class-detection-in-the-wild | 1503.05038 | null | http://arxiv.org/abs/1503.05038v1 | http://arxiv.org/pdf/1503.05038v1.pdf | 3D Object Class Detection in the Wild | Object class detection has been a synonym for 2D bounding box localization
for the longest time, fueled by the success of powerful statistical learning
techniques, combined with robust image representations. Only recently, there
has been a growing interest in revisiting the promise of computer vision from
the early days: to precisely delineate the contents of a visual scene, object
by object, in 3D. In this paper, we draw from recent advances in object
detection and 2D-3D object lifting in order to design an object class detector
that is particularly tailored towards 3D object class detection. Our 3D object
class detection method consists of several stages gradually enriching the
object detection output with object viewpoint, keypoints and 3D shape
estimates. Following careful design, in each stage it constantly improves the
performance and achieves state-ofthe-art performance in simultaneous 2D
bounding box and viewpoint estimation on the challenging Pascal3D+ dataset. | ['Peter Gehler', 'Michael Stark', 'Tobias Ritschel', 'Bojan Pepik', 'Bernt Schiele'] | 2015-03-17 | null | null | null | null | ['viewpoint-estimation'] | ['computer-vision'] | [ 1.24118350e-01 -2.85590559e-01 -6.41676709e-02 -3.53280932e-01
-5.70129633e-01 -8.23194265e-01 8.90927553e-01 3.80485445e-01
-4.37148809e-01 -4.67429459e-02 -2.09014192e-01 -1.32110015e-01
1.02821812e-01 -1.92851603e-01 -4.89662826e-01 -3.77155215e-01
-1.84033468e-01 6.86552882e-01 8.53836238e-01 9.43926945e-02
6.82505608e-01 1.10005176e+00 -1.82568586e+00 1.07849114e-01
1.67733148e-01 1.14086294e+00 2.77620405e-01 7.78043747e-01
-3.20109576e-01 4.58269209e-01 -5.94315171e-01 -1.37948379e-01
4.23697442e-01 -2.43709926e-02 -4.65346962e-01 4.69315976e-01
7.94013560e-01 -4.08372462e-01 1.00604273e-01 7.70375252e-01
4.41236109e-01 -1.55219689e-01 9.09469187e-01 -1.24987984e+00
1.44784087e-02 -1.18380375e-01 -8.84121656e-01 2.81664252e-01
4.31725711e-01 -5.72607480e-02 8.43611419e-01 -1.19178700e+00
6.37732863e-01 1.39216280e+00 6.63204551e-01 4.58119094e-01
-1.33148670e+00 -4.08509791e-01 3.56921583e-01 -4.67214882e-02
-1.47262537e+00 -2.05499932e-01 7.96189010e-01 -8.98413956e-01
1.19416451e+00 9.61443707e-02 7.41795361e-01 6.22069240e-01
-1.02673076e-01 1.06992900e+00 1.07480025e+00 -6.74467862e-01
9.58755016e-02 2.01412156e-01 3.04349482e-01 6.55927420e-01
3.97570074e-01 2.96309471e-01 -3.35529089e-01 -1.48731112e-01
8.37561190e-01 -9.96578708e-02 1.11027420e-01 -1.21013784e+00
-1.05062139e+00 6.49844944e-01 2.71005332e-01 -2.27187220e-02
-7.47565106e-02 9.51325595e-02 4.80798781e-01 -1.71406627e-01
6.26289129e-01 2.15460464e-01 -3.76922637e-01 -1.32132173e-01
-9.01217043e-01 7.02666402e-01 6.27765775e-01 1.15526390e+00
3.65208447e-01 -3.52136433e-01 9.36199650e-02 5.68621993e-01
4.54073578e-01 5.65128684e-01 -1.63460389e-01 -8.07980299e-01
2.22964793e-01 8.83086979e-01 3.55349541e-01 -6.75492048e-01
-5.50003409e-01 -2.56771207e-01 -2.48762563e-01 8.44315588e-01
6.04904354e-01 3.86991441e-01 -1.08664346e+00 1.13379395e+00
8.41676772e-01 -2.84702957e-01 -3.85593832e-01 8.59098196e-01
7.78713405e-01 3.35996866e-01 3.22941877e-02 2.53817976e-01
1.40783942e+00 -4.25757140e-01 -1.40826732e-01 -2.11149111e-01
4.74833548e-01 -8.33094776e-01 3.94006163e-01 3.50899845e-01
-1.04725945e+00 -6.92667961e-01 -1.01677728e+00 -1.52378812e-01
-4.93710518e-01 6.97300285e-02 5.67843795e-01 7.47971237e-01
-5.22865534e-01 1.99926659e-01 -8.29771221e-01 -4.56995368e-01
7.56033957e-01 3.07293952e-01 -5.02386928e-01 9.62994695e-02
-3.26267928e-01 1.34433579e+00 7.74421632e-01 -1.57876357e-01
-9.06498492e-01 -6.29019916e-01 -8.11652541e-01 -3.73577237e-01
5.76857924e-01 -5.57511091e-01 1.12898910e+00 -2.91917056e-01
-1.08612728e+00 1.59818017e+00 -5.93904522e-04 -3.51771772e-01
7.49991119e-01 -3.63267183e-01 1.66073009e-01 -1.04753435e-01
-4.76983860e-02 8.39094102e-01 1.16430759e+00 -1.30797994e+00
-1.20764410e+00 -7.81968653e-01 -2.06476539e-01 2.99660683e-01
2.99012989e-01 3.62187535e-01 -5.41950107e-01 -2.36476988e-01
5.23131073e-01 -8.88261199e-01 -1.75079983e-02 3.98085475e-01
-1.97064593e-01 -7.00676978e-01 1.00777674e+00 -1.62906229e-01
7.01007068e-01 -2.32227755e+00 4.70965207e-02 -1.75038725e-02
3.94454300e-01 3.32497388e-01 1.65865287e-01 1.92910641e-01
-3.66194509e-02 -1.44475356e-01 -8.90145898e-02 -4.57072109e-01
7.15702623e-02 3.85552086e-02 -3.50326836e-01 8.63873482e-01
5.49230337e-01 7.58820415e-01 -8.53371382e-01 -6.99971199e-01
6.25783801e-01 4.39985782e-01 -4.74776685e-01 2.52253324e-01
-2.47176677e-01 1.63869187e-01 -5.36858439e-01 8.35904658e-01
8.94007504e-01 -8.05146545e-02 -2.86819369e-01 -7.50494003e-02
-4.32272106e-01 2.94471048e-02 -1.39874756e+00 1.44707501e+00
1.33254722e-01 7.35059202e-01 1.37564316e-01 -8.33844066e-01
1.11566758e+00 7.69755319e-02 4.90790486e-01 -2.64660329e-01
2.29150131e-01 4.39515978e-01 -2.21435457e-01 -3.18908513e-01
4.68968004e-01 3.46919224e-02 2.12420542e-02 1.09700277e-01
6.75246492e-02 -8.65399420e-01 2.18590438e-01 -1.35428533e-01
6.11575723e-01 5.30323088e-01 8.48948002e-01 -2.02471942e-01
4.14455175e-01 3.27233434e-01 -2.75713224e-02 8.28982770e-01
-5.52718580e-01 6.91950619e-01 4.24515009e-01 -4.57968503e-01
-1.23556316e+00 -1.24045265e+00 -6.71263099e-01 8.87454152e-01
2.20044777e-01 -4.35286202e-02 -4.22512680e-01 -9.18787301e-01
5.27650237e-01 2.30334893e-01 -6.16302133e-01 2.52600223e-01
-7.53728390e-01 -4.64432746e-01 1.87265888e-01 6.13211870e-01
2.55229920e-01 -7.50874758e-01 -1.11634314e+00 5.30637838e-02
4.46212411e-01 -1.25475645e+00 -2.27835566e-01 6.27893209e-01
-9.93860304e-01 -1.32305169e+00 -8.39034557e-01 -8.28549564e-01
5.88972032e-01 5.22423804e-01 1.13011861e+00 -1.11857362e-01
-8.60713482e-01 4.76977319e-01 -3.31088066e-01 -8.69202793e-01
-3.54478657e-01 -1.31272793e-01 9.04990956e-02 -3.11668366e-01
6.32349253e-01 -1.56217262e-01 -3.40747207e-01 4.41087306e-01
-5.83896935e-01 -6.04460314e-02 5.34293354e-01 5.40891230e-01
6.52845263e-01 -2.10766941e-01 5.65762166e-04 -4.83587533e-01
-7.82322809e-02 -6.05405271e-02 -1.11390436e+00 -1.00966893e-01
-2.21856460e-01 -2.52241701e-01 -1.02262966e-01 -3.72047871e-01
-6.58994138e-01 4.73573297e-01 1.56543367e-02 -2.75414675e-01
-6.82466090e-01 -2.00088128e-01 -7.77077004e-02 -2.42475525e-01
8.41311872e-01 1.73495799e-01 6.99713407e-03 -7.10432589e-01
6.26862228e-01 7.49748468e-01 4.00014073e-01 -1.76640168e-01
9.24454391e-01 7.96067595e-01 3.69333893e-01 -9.66156363e-01
-1.03907633e+00 -1.02578390e+00 -1.24890971e+00 -3.13284397e-01
9.97180104e-01 -9.36506867e-01 -6.90376997e-01 4.10079598e-01
-1.28425980e+00 3.50124910e-02 -2.63074160e-01 4.72222686e-01
-6.98115349e-01 2.20627710e-01 1.08840853e-01 -1.04780281e+00
-3.67102586e-02 -9.09100533e-01 1.53295469e+00 6.71994910e-02
-1.50995538e-01 -6.11762285e-01 3.14401314e-02 2.55944163e-01
-2.69858707e-02 4.86635685e-01 7.97823489e-01 -7.06647396e-01
-8.17061782e-01 -6.76041663e-01 -5.52288175e-01 3.07871819e-01
-3.01416051e-02 -2.85884794e-02 -1.04779446e+00 -1.36715710e-01
-1.39923409e-01 -1.33871883e-01 8.15439761e-01 4.39244509e-01
8.25887144e-01 5.49950778e-01 -5.89125216e-01 5.39608121e-01
1.35932672e+00 1.53110132e-01 -3.93930748e-02 2.73065269e-01
4.30849582e-01 6.42375588e-01 7.01022625e-01 4.85913992e-01
4.56531793e-02 8.50931823e-01 6.74800813e-01 -2.35886220e-02
-3.82760435e-01 -1.13634847e-01 -1.82052702e-01 -9.36577842e-02
-9.25860833e-03 1.01378910e-01 -1.06420255e+00 5.40611029e-01
-1.54805732e+00 -7.12833524e-01 -1.93426222e-01 2.23950219e+00
4.95729834e-01 6.12818837e-01 5.25578141e-01 2.49935031e-01
6.52079523e-01 5.03327921e-02 -4.70845163e-01 7.92279840e-02
-3.37090045e-02 -1.25980794e-01 5.21133840e-01 2.38987550e-01
-1.44546163e+00 8.07775855e-01 6.68641710e+00 7.44834483e-01
-6.72022879e-01 -3.55851084e-01 1.21517502e-01 4.24345471e-02
3.63020539e-01 -8.07755068e-02 -1.48759115e+00 3.92563455e-02
-2.47929450e-02 1.49861157e-01 -1.17428452e-01 1.24820828e+00
-6.10956065e-02 -3.85705054e-01 -1.47082818e+00 1.06627119e+00
2.93274254e-01 -1.11280394e+00 -1.31745368e-01 -6.83214664e-02
5.19961774e-01 7.80668333e-02 -8.84215459e-02 1.45889774e-01
-1.41112760e-01 -7.31881738e-01 9.83997941e-01 1.66641381e-02
5.12255192e-01 -4.94472057e-01 3.96427780e-01 6.04902148e-01
-1.24968004e+00 -2.19814232e-04 -3.51216555e-01 1.47551363e-02
2.71474775e-02 3.10685366e-01 -1.25258172e+00 5.12811802e-02
7.19838321e-01 6.40281737e-01 -6.47049785e-01 1.61470962e+00
-6.01226501e-02 1.60961851e-01 -5.28089821e-01 -3.30406070e-01
2.52490312e-01 1.15841344e-01 9.38460946e-01 1.32962835e+00
-1.92861870e-01 9.64491516e-02 3.95259559e-01 8.19887877e-01
2.14546576e-01 -5.49586117e-02 -6.97126508e-01 1.46835148e-01
3.57335985e-01 1.20622420e+00 -1.09575033e+00 -2.67892450e-01
-3.17782402e-01 4.72078592e-01 1.90017477e-01 -8.26751888e-02
-7.10788190e-01 -3.62915695e-01 5.66889226e-01 2.69630432e-01
6.95423603e-01 -7.29632497e-01 -2.28487015e-01 -6.28599167e-01
7.02038705e-02 -3.76012743e-01 1.78826049e-01 -6.54110134e-01
-9.94834661e-01 3.59875083e-01 3.69750232e-01 -1.09186971e+00
-9.93313864e-02 -1.04970515e+00 -6.41180575e-02 6.90114081e-01
-1.53816319e+00 -1.25969422e+00 -3.92046124e-01 2.59903818e-01
6.59556687e-01 4.01477963e-02 6.84342682e-01 1.58290416e-01
-6.29338473e-02 1.26851857e-01 -2.13031955e-02 2.09268659e-01
4.96248275e-01 -1.43227458e+00 5.45492828e-01 6.37576103e-01
4.25965905e-01 2.14843154e-01 6.49280667e-01 -5.65543771e-01
-1.26630414e+00 -8.76043975e-01 6.59940004e-01 -1.19115090e+00
4.82190162e-01 -8.56978238e-01 -6.78951740e-01 4.63096082e-01
-4.94003117e-01 2.59672344e-01 2.27208272e-01 2.35211756e-03
-5.39902866e-01 1.13286681e-01 -9.95116115e-01 4.25709218e-01
1.04759598e+00 -3.82147998e-01 -8.51160586e-01 3.59319508e-01
4.17780429e-01 -7.47806609e-01 -5.76648355e-01 5.97511828e-01
5.74776053e-01 -9.35074985e-01 1.32489002e+00 -5.24745822e-01
-1.90214664e-01 -6.60748005e-01 -2.80624956e-01 -5.32851219e-01
-2.31317043e-01 -4.35362101e-01 -2.34094799e-01 7.86028326e-01
5.02776727e-02 -1.68220207e-01 1.13542533e+00 2.37965956e-01
-1.53793678e-01 -4.97592300e-01 -1.14214003e+00 -8.92037034e-01
-2.31189817e-01 -5.36003053e-01 1.58448488e-01 2.06397027e-01
-4.62640911e-01 2.40652591e-01 -2.53223199e-02 1.05217069e-01
9.65638518e-01 5.47246933e-01 1.09403145e+00 -1.59474671e+00
4.91932593e-02 -8.76797259e-01 -9.44659173e-01 -1.44602609e+00
-2.22326815e-01 -7.49946654e-01 2.03700751e-01 -1.20412326e+00
3.98408532e-01 -3.85373563e-01 1.24148494e-02 2.65190989e-01
-2.45979100e-01 5.35251141e-01 3.44941914e-01 1.74758300e-01
-6.58150792e-01 1.48906425e-01 1.15294600e+00 9.17845890e-02
-1.81290746e-01 2.90682822e-01 -1.60018250e-01 8.42429221e-01
3.22793871e-01 -4.75848407e-01 4.43313979e-02 -3.06100190e-01
-4.22670729e-02 -2.72144496e-01 6.32540762e-01 -9.55761313e-01
-3.86375450e-02 1.45683780e-01 6.52313292e-01 -1.34545696e+00
6.44306123e-01 -1.02122247e+00 -3.56961906e-01 4.59059596e-01
-1.74642503e-01 -3.93366426e-01 2.59068638e-01 5.73460877e-01
2.21098028e-02 -2.82655418e-01 1.14549589e+00 -2.14283139e-01
-1.12249660e+00 2.78442889e-01 -1.84941143e-01 9.31759328e-02
1.28766227e+00 -7.86060691e-01 1.34005636e-01 3.55672210e-01
-7.38443196e-01 1.37071759e-01 4.06620890e-01 5.12426555e-01
6.69328094e-01 -9.78141725e-01 -7.64602840e-01 3.40785682e-01
4.00120556e-01 2.21121550e-01 -7.91069642e-02 7.11920440e-01
-5.09006679e-01 5.50223291e-01 -5.26759075e-03 -1.26724625e+00
-1.68143046e+00 7.70202160e-01 3.17633688e-01 2.34393790e-01
-7.61904001e-01 1.14963341e+00 1.45540312e-01 -1.17552057e-01
6.00108266e-01 -3.99427444e-01 -1.95771620e-01 2.20195934e-01
4.73610014e-01 4.21469808e-01 4.55526635e-03 -5.44375658e-01
-6.04037762e-01 1.04022682e+00 -1.06964745e-01 1.58413664e-01
1.11336517e+00 -5.71795776e-02 2.31046334e-01 5.50792158e-01
1.10994124e+00 -2.16688216e-01 -1.45542133e+00 -2.41491795e-01
4.06074017e-01 -7.21238852e-01 3.50616723e-02 -6.57184303e-01
-3.29417974e-01 9.51306403e-01 7.98884094e-01 3.76755059e-01
6.96212530e-01 5.02853751e-01 1.42623261e-01 3.16347241e-01
4.63832229e-01 -8.75510335e-01 1.30249590e-01 6.69332683e-01
7.89848089e-01 -1.59775209e+00 5.02170265e-01 -4.73309129e-01
-1.54955238e-01 1.30607116e+00 3.59798789e-01 -3.84370178e-01
7.99470663e-01 2.41543785e-01 -5.00431359e-02 -3.93520653e-01
-2.91011184e-01 -5.14591634e-01 6.70023382e-01 8.58787477e-01
3.11646253e-01 -4.29733656e-02 1.80721655e-01 -1.72581542e-02
1.65458396e-01 -4.46035236e-01 1.54049397e-01 1.11102724e+00
-1.04529548e+00 -9.19195592e-01 -6.31687760e-01 1.76429376e-01
-2.71507800e-01 2.23751768e-01 -4.72280860e-01 1.32125378e+00
3.24013531e-01 5.42571962e-01 2.80375034e-01 1.37237564e-01
5.51577926e-01 1.61188524e-02 8.86952043e-01 -9.16276813e-01
-3.05259734e-01 2.30114311e-01 -1.29440024e-01 -5.91069102e-01
-4.65106100e-01 -8.30252051e-01 -1.06976807e+00 3.34020585e-01
-7.20485091e-01 -1.34403870e-01 1.19494081e+00 8.46217990e-01
2.12483499e-02 7.46550485e-02 4.02205437e-01 -1.38978589e+00
-6.75762057e-01 -7.06471741e-01 -5.66379309e-01 3.17946523e-02
4.42221433e-01 -9.19465899e-01 -2.58675784e-01 -5.28111085e-02] | [7.63656759262085, -2.675401210784912] |
4287dcf6-9967-4914-8b29-8ae61d9005d1 | a-conceptual-framework-for-implicit | 2104.03940 | null | https://arxiv.org/abs/2104.03940v1 | https://arxiv.org/pdf/2104.03940v1.pdf | A Conceptual Framework for Implicit Evaluation of Conversational Search Interfaces | Conversational search (CS) has recently become a significant focus of the information retrieval (IR) research community. Multiple studies have been conducted which explore the concept of conversational search. Understanding and advancing research in CS requires careful and detailed evaluation. Existing CS studies have been limited to evaluation based on simple user feedback on task completion. We propose a CS evaluation framework which includes multiple dimensions: search experience, knowledge gain, software usability, cognitive load and user experience, based on studies of conversational systems and IR. We introduce these evaluation criteria and propose their use in a framework for the evaluation of CS systems. | ['Gareth J. F. Jones', 'Abhishek Kaushik'] | 2021-04-08 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [ 3.05730179e-02 -3.93022783e-02 -1.66127637e-01 -2.00509980e-01
-7.52474010e-01 -9.39930916e-01 1.11565197e+00 3.73935223e-01
-5.62890589e-01 1.97141871e-01 2.26486355e-01 -6.67081356e-01
-5.59543848e-01 1.19168513e-01 8.85591358e-02 4.34236452e-02
1.28136069e-01 2.37100080e-01 3.15317184e-01 -4.79278594e-01
1.20984101e+00 2.82715738e-01 -1.85500598e+00 2.55806059e-01
1.04649246e+00 8.31809163e-01 7.48458564e-01 9.19208109e-01
-3.62396538e-01 8.52222860e-01 -8.70570302e-01 -1.76240385e-01
-2.86384970e-01 -2.45240286e-01 -1.46367300e+00 -4.15097535e-01
6.64421394e-02 2.71353014e-02 -9.79394317e-02 5.78230321e-01
8.92702460e-01 4.79809880e-01 2.81415045e-01 -1.41403747e+00
-7.89789140e-01 2.94520438e-01 2.22685367e-01 5.97720802e-01
1.19899464e+00 -3.51224810e-01 8.06878865e-01 -8.29181135e-01
7.23202944e-01 1.13247168e+00 3.31601381e-01 3.81000161e-01
-9.39201355e-01 -5.78773856e-01 -1.45826951e-01 5.28555214e-01
-1.38981056e+00 -5.31357169e-01 3.82156223e-01 -5.13321459e-01
1.45361650e+00 7.13758767e-01 6.16714239e-01 6.64724767e-01
1.21710487e-01 8.25260341e-01 1.07040727e+00 -1.12994421e+00
3.22112530e-01 8.70834589e-01 6.01795137e-01 -2.96135750e-02
2.95265131e-02 -1.97907552e-01 -6.25303328e-01 -5.13460755e-01
4.50012326e-01 -2.34997988e-01 -2.86346674e-01 -2.77264714e-01
-6.37036324e-01 5.61477184e-01 7.41844177e-02 7.15471208e-01
-3.72341543e-01 -4.65599410e-02 2.92161465e-01 4.69510108e-01
3.72627735e-01 9.97038782e-01 -1.86242476e-01 -8.93834352e-01
-4.54514772e-01 3.07048887e-01 1.44651711e+00 1.17340040e+00
3.85273248e-01 -8.19847226e-01 -4.17139679e-01 1.21613276e+00
3.81483167e-01 3.43952298e-01 5.48905969e-01 -1.02075839e+00
-1.02730222e-01 6.82642102e-01 3.51512522e-01 -1.10420227e+00
-4.27656323e-02 -1.75274536e-01 2.68086225e-01 -2.92977303e-01
-4.89607006e-01 2.09471826e-02 -3.39711726e-01 1.33470380e+00
-1.63579285e-01 -5.18737376e-01 9.27775353e-02 7.03219593e-01
1.11344731e+00 3.89175981e-01 3.08556348e-01 -3.14168364e-01
1.21814883e+00 -9.99072075e-01 -9.19081628e-01 -2.66189873e-01
6.85615838e-01 -1.21886134e+00 1.34789538e+00 1.96248576e-01
-1.37951243e+00 -2.63184518e-01 -9.01818216e-01 -6.43221363e-02
-6.13672853e-01 -3.40235978e-02 6.44428492e-01 9.82374609e-01
-1.63729596e+00 8.06480348e-02 -5.14100194e-01 -8.73280108e-01
-5.49746037e-01 1.29200906e-01 2.80735642e-01 -1.33650720e-01
-1.14489508e+00 1.40020776e+00 -1.30042076e-01 -1.04891799e-01
-1.67148545e-01 -2.61845738e-01 -3.95190507e-01 2.49199107e-01
3.31583381e-01 -7.18177915e-01 2.02599263e+00 -4.72871393e-01
-1.38683486e+00 6.33523703e-01 -1.49470121e-01 3.45665216e-01
-1.06503047e-01 -3.14658999e-01 -5.24037123e-01 1.77830026e-01
3.01399678e-01 1.45230219e-01 -9.73260552e-02 -1.43726659e+00
-8.24878752e-01 -2.76342213e-01 4.14004207e-01 9.73719358e-01
-3.67280662e-01 7.80174017e-01 -8.98178697e-01 -1.46230474e-01
-2.28268430e-01 -1.10835433e+00 2.12667719e-03 -5.77852011e-01
1.62824228e-01 -7.58134365e-01 5.05600214e-01 -3.92708898e-01
1.99168599e+00 -1.95073116e+00 -1.03889041e-01 2.08660066e-01
1.40385047e-01 4.11588848e-01 6.61137179e-02 1.25666356e+00
2.42234677e-01 5.95385194e-01 5.58721840e-01 -1.34790853e-01
1.57045081e-01 -5.56454778e-01 2.51756340e-01 -1.77271321e-01
-6.20589435e-01 6.70062721e-01 -1.00240982e+00 -5.56842387e-01
-3.39655429e-02 4.09586370e-01 -2.34443605e-01 3.66231263e-01
1.90275133e-01 -1.73121691e-01 -5.94572604e-01 6.00717008e-01
2.42473751e-01 -4.94697392e-01 1.58222511e-01 5.09711385e-01
-6.25024736e-01 5.84611833e-01 -6.97579265e-01 1.76396799e+00
-7.60697126e-01 8.94815803e-01 7.19461367e-02 -2.66358674e-01
5.79972923e-01 6.61873877e-01 3.73994291e-01 -1.18247795e+00
2.46117804e-02 3.06953210e-02 -1.23780638e-01 -8.50157320e-01
7.47193336e-01 5.09682119e-01 1.35397941e-01 9.49997425e-01
-3.42906386e-01 -2.45015383e-01 2.22571477e-01 5.86789012e-01
1.30975521e+00 -3.40924382e-01 1.83928877e-01 -4.42130059e-01
4.99014258e-01 3.35842192e-01 -3.60002577e-01 1.18591928e+00
-2.63661742e-01 9.39435661e-02 -3.60135660e-02 2.16567189e-01
-4.72697705e-01 -4.66003865e-01 -1.16302922e-01 1.46472621e+00
5.92038512e-01 -8.79447043e-01 -1.05003703e+00 -1.80096015e-01
-4.07130808e-01 7.70752728e-01 -1.61654368e-01 -2.01821059e-01
5.84989451e-02 -5.90738617e-02 3.42552572e-01 1.50064483e-01
4.38208520e-01 -1.51215947e+00 -9.16637182e-01 -6.06684899e-03
-6.10008121e-01 -9.37510729e-01 -6.15099549e-01 -2.93749779e-01
-5.87010026e-01 -1.05059588e+00 -4.70145494e-01 -1.08697498e+00
2.64245093e-01 1.10552573e+00 1.29258287e+00 8.03372324e-01
-6.02496147e-01 1.62402439e+00 -8.07564080e-01 -5.06365120e-01
-6.89630881e-02 6.63098646e-03 -2.68709093e-01 -9.83118474e-01
7.22043455e-01 -1.08749874e-01 -9.93396580e-01 6.44727886e-01
-8.35835934e-01 -3.90396833e-01 7.14938045e-01 4.99805242e-01
-2.87553370e-01 7.16946796e-02 3.50927353e-01 -6.16477787e-01
2.06145716e+00 -5.79129994e-01 -1.00169387e-02 8.62071812e-01
-1.36795819e+00 -2.35975936e-01 -5.88055372e-01 -2.95917481e-01
-1.37970817e+00 -7.30948508e-01 3.19591135e-01 2.48928964e-01
-1.14032745e-01 1.09781718e+00 4.34457093e-01 -5.77214420e-01
1.04985189e+00 1.21738285e-01 1.54358491e-01 -2.47664154e-01
6.70019630e-03 1.31104672e+00 -1.22184874e-02 -6.77022994e-01
5.17236851e-02 -3.52077127e-01 -7.71675467e-01 -1.16456878e+00
-1.35018528e-01 -1.23876274e+00 -8.55173320e-02 -6.97825849e-01
4.14596319e-01 -6.85071290e-01 -1.20504868e+00 -3.58129069e-02
-1.09532535e+00 -2.11541981e-01 5.23422003e-01 6.07004523e-01
-3.00347179e-01 4.90288347e-01 -5.57649910e-01 -1.38799477e+00
-7.13909984e-01 -1.23313022e+00 7.42515326e-01 4.21632528e-01
-7.22952902e-01 -9.70811963e-01 4.50700909e-01 6.64478302e-01
9.74171758e-01 -5.35693228e-01 8.62201393e-01 -5.28989196e-01
-5.41727424e-01 -4.91890818e-01 -2.36371651e-01 -7.84719810e-02
-9.08174366e-02 -4.98444960e-02 -6.56996071e-01 -2.81829536e-01
2.94601340e-02 -3.58152121e-01 -1.26086995e-01 2.98763156e-01
6.73833609e-01 -2.58785129e-01 -6.94980204e-01 -3.65597069e-01
1.48472548e+00 9.39818442e-01 5.61152339e-01 7.65010774e-01
-1.81410119e-01 7.89150774e-01 9.48831201e-01 1.58106729e-01
3.68343771e-01 8.83029878e-01 -2.56233722e-01 3.18972677e-01
1.81642503e-01 1.02540463e-01 6.42318577e-02 6.22304797e-01
-2.33777612e-01 -4.36031073e-01 -1.29931498e+00 4.40122455e-01
-1.78571963e+00 -7.88760424e-01 3.53268892e-01 2.08237576e+00
5.40066004e-01 -5.32628968e-02 -2.40529906e-02 -1.97873294e-01
5.82895994e-01 -4.10153925e-01 -1.49083152e-01 -6.63969576e-01
6.43741310e-01 1.89291716e-01 3.68781127e-02 6.08390868e-01
-2.85662472e-01 9.21429992e-01 7.84999132e+00 4.67501909e-01
-7.55992353e-01 -2.18644924e-02 1.22939095e-01 2.73231000e-01
-4.25513148e-01 3.66408139e-01 -5.42194426e-01 1.91062510e-01
8.02036047e-01 -7.83838809e-01 6.43902242e-01 1.14098525e+00
6.47332221e-02 -5.68254948e-01 -1.04982722e+00 1.21805167e+00
1.94868520e-01 -1.08325672e+00 -3.15456897e-01 -5.56740761e-02
2.36892432e-01 -1.96716815e-01 8.68270472e-02 5.88380456e-01
3.28948081e-01 -8.15788865e-01 1.69240296e-01 4.32904124e-01
5.16535640e-01 -4.11263347e-01 6.77038431e-01 5.08101940e-01
-8.59524190e-01 2.39739828e-02 2.66379416e-02 -2.44671479e-01
-1.38794690e-01 -2.41227910e-01 -1.11319458e+00 2.20190674e-01
1.08530736e+00 4.62213568e-02 -6.82053089e-01 1.23223805e+00
3.95269871e-01 9.99844000e-02 -6.72968701e-02 -8.35361898e-01
7.16385096e-02 -2.30526656e-01 3.50683510e-01 1.37986743e+00
1.13357589e-01 7.06761479e-01 2.09901154e-01 5.15104413e-01
4.78992373e-01 4.45065409e-01 -6.46920145e-01 -2.00560674e-01
1.18998516e+00 1.15990567e+00 -7.40340948e-01 -2.01776102e-01
-3.94774526e-01 7.57794976e-01 5.24959452e-02 5.41110933e-01
-7.69389123e-02 -6.64975047e-01 5.15588999e-01 -6.08023852e-02
-2.74027616e-01 -1.54686213e-01 -1.77365124e-01 -5.44217050e-01
2.24365428e-01 -9.02297616e-01 1.90949932e-01 -1.04679453e+00
-7.90661216e-01 9.54794407e-01 4.19383585e-01 -8.57164741e-01
-5.35350025e-01 4.45922539e-02 -5.07066131e-01 1.01589942e+00
-1.10484135e+00 -4.99357283e-01 -7.54014790e-01 3.62331361e-01
5.62903225e-01 -9.64235067e-02 1.00631285e+00 3.61430675e-01
-4.11258936e-01 4.88253176e-01 -5.67102619e-02 -7.17639983e-01
8.77283394e-01 -1.14769220e+00 1.17933109e-01 2.85925895e-01
-4.31841969e-01 1.51650846e+00 7.29947150e-01 -6.67407393e-01
-1.62848997e+00 1.67713389e-01 1.17672503e+00 -7.25326061e-01
2.82712579e-01 -2.41249993e-01 -4.95828658e-01 3.05811524e-01
7.70090163e-01 -9.84412014e-01 9.85449910e-01 6.78450048e-01
1.78404793e-01 2.74743557e-01 -8.93256545e-01 7.54220843e-01
9.28284049e-01 -8.31106246e-01 -3.69042128e-01 4.00478393e-01
6.52023554e-01 -2.86858618e-01 -9.04648185e-01 1.98564947e-01
7.58189380e-01 -8.91544521e-01 1.11148345e+00 -1.02062792e-01
1.55086089e-02 2.05397308e-01 2.97500491e-01 -1.21115017e+00
-4.51233208e-01 -7.70680428e-01 4.12832230e-01 1.05031872e+00
5.97199857e-01 -5.01697004e-01 4.57423925e-01 1.63247597e+00
-3.16530108e-01 -4.13021296e-01 -2.32516140e-01 -5.56867480e-01
-4.38165575e-01 -2.16007292e-01 2.38860264e-01 9.58844423e-01
1.20604599e+00 8.09106588e-01 1.67048603e-01 -2.20564976e-01
1.12229232e-02 -2.45681405e-01 7.19381988e-01 -1.48503065e+00
1.24756008e-01 -8.72285426e-01 -5.69863096e-02 -1.25955451e+00
-3.08326900e-01 -4.80340213e-01 2.72053182e-01 -1.77692950e+00
6.09359741e-01 -4.04592216e-01 -6.98528364e-02 -8.29873886e-03
-2.08287939e-01 -4.18076009e-01 1.38800800e-01 5.88374734e-01
-9.26487923e-01 1.25864536e-01 8.84059608e-01 4.15354609e-01
-5.33937037e-01 7.87096098e-02 -1.22927773e+00 7.94408098e-02
6.72577739e-01 -1.01901412e-01 -9.04259205e-01 -2.31394127e-01
4.73014534e-01 5.15673697e-01 -2.80302644e-01 -7.75555730e-01
1.10804677e+00 -2.33874425e-01 -2.84974337e-01 -4.60879028e-01
3.36932838e-01 -9.19812918e-01 1.80912212e-01 2.02627584e-01
-8.47117543e-01 6.29469931e-01 1.75923973e-01 4.32652354e-01
-4.93950009e-01 -5.00034451e-01 -1.11692637e-01 -1.41278014e-01
-6.42802715e-01 -1.31194875e-01 -8.11611950e-01 -2.37079501e-01
9.29674566e-01 -4.04126436e-01 -2.82814413e-01 -9.53245819e-01
-3.93524289e-01 5.64182699e-01 2.99907267e-01 6.54221356e-01
5.66759467e-01 -9.38153088e-01 -1.04359552e-01 -3.04889232e-01
5.72578132e-01 -4.93806869e-01 7.91679099e-02 4.86074656e-01
-5.97781837e-01 1.22332048e+00 -1.63355179e-03 -3.96615833e-01
-1.67672884e+00 4.28179324e-01 -2.07879841e-01 2.10248102e-02
-1.02822356e-01 6.03270113e-01 -1.48522586e-01 -2.32214093e-01
8.55094016e-01 1.92335948e-01 -7.30035722e-01 -2.39209697e-01
6.35904431e-01 4.83402550e-01 2.79121399e-01 1.79175623e-02
-3.36480856e-01 2.73548990e-01 -3.44086587e-01 -8.17258477e-01
8.20813000e-01 -6.04527295e-01 -2.01696426e-01 3.02472085e-01
9.81614709e-01 -3.36748660e-01 -8.67650937e-03 -3.69829327e-01
6.41202152e-01 -5.96327364e-01 3.36274296e-01 -1.27745700e+00
-1.78428277e-01 4.20737833e-01 6.91702366e-01 7.45936036e-01
9.00904894e-01 5.21456972e-02 1.99851006e-01 8.53911579e-01
1.89670429e-01 -1.59904730e+00 3.77920866e-01 4.82811272e-01
1.10803044e+00 -1.21343338e+00 -1.96541190e-01 -5.86557984e-01
-8.43597710e-01 9.29332674e-01 7.50849605e-01 6.66139901e-01
1.08031881e+00 -1.20876610e-01 2.29517803e-01 -7.00365424e-01
-8.96256506e-01 -1.58012822e-01 3.99977088e-01 2.29873419e-01
1.17601478e+00 -2.91777641e-01 -1.11834908e+00 3.94055903e-01
-8.11639875e-02 2.28078008e-01 2.91638494e-01 1.75110924e+00
-5.07028699e-01 -1.13640857e+00 -1.30264208e-01 1.85070664e-01
-3.85575235e-01 -4.05115783e-01 -1.08785760e+00 7.58912921e-01
-1.00851047e+00 1.69201589e+00 -2.18118817e-01 -6.12319469e-01
4.47388321e-01 2.12858528e-01 8.72413218e-02 -8.39073956e-01
-9.20086980e-01 -2.71924995e-02 5.64565361e-01 -5.90944648e-01
-5.49525797e-01 -7.82889426e-01 -6.35776043e-01 -4.46631312e-01
-8.20903242e-01 1.07893121e+00 1.40785921e+00 7.69097149e-01
8.05296481e-01 7.16901198e-02 5.60388684e-01 -3.85905057e-01
-3.27887356e-01 -1.49755740e+00 -2.79674292e-01 1.77839249e-01
-9.19000730e-02 -6.28615558e-01 -3.78987730e-01 -2.27725714e-01] | [12.22993278503418, 7.724972248077393] |
27f46ab7-e8bc-4d0c-bd88-5031efc58f3e | dependency-based-decipherment-for-resource | null | null | https://aclanthology.org/D13-1173 | https://aclanthology.org/D13-1173.pdf | Dependency-Based Decipherment for Resource-Limited Machine Translation | null | ['Qing Dou', 'Kevin Knight'] | 2013-10-01 | null | null | null | emnlp-2013-10 | ['decipherment'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.212493419647217, 3.53434419631958] |
eed12dce-49b0-4bf3-ae56-73e959061959 | transformer-based-neural-marked-spatio | 2302.09276 | null | https://arxiv.org/abs/2302.09276v1 | https://arxiv.org/pdf/2302.09276v1.pdf | Transformer-Based Neural Marked Spatio Temporal Point Process Model for Football Match Events Analysis | With recently available football match event data that record the details of football matches, analysts and researchers have a great opportunity to develop new performance metrics, gain insight, and evaluate key performance. However, most sports sequential events modeling methods and performance metrics approaches could be incomprehensive in dealing with such large-scale spatiotemporal data (in particular, temporal process), thereby necessitating a more comprehensive spatiotemporal model and a holistic performance metric. To this end, we proposed the Transformer-Based Neural Marked Spatio Temporal Point Process (NMSTPP) model for football event data based on the neural temporal point processes (NTPP) framework. In the experiments, our model outperformed the prediction performance of the baseline models. Furthermore, we proposed the holistic possession utilization score (HPUS) metric for a more comprehensive football possession analysis. For verification, we examined the relationship with football teams' final ranking, average goal score, and average xG over a season. It was observed that the average HPUS showed significant correlations regardless of not using goal and details of shot information. Furthermore, we show HPUS examples in analyzing possessions, matches, and between matches. | ['Keisuke Fujii', 'Tony Sit', 'Calvin C. K. Yeung'] | 2023-02-18 | null | null | null | null | ['point-processes'] | ['methodology'] | [-2.66025811e-01 -7.23224103e-01 -3.19956541e-01 -1.67364761e-01
-4.69735891e-01 -2.92072147e-01 5.42892933e-01 5.46613455e-01
-4.97023821e-01 4.74585712e-01 4.78800118e-01 8.29647928e-02
-6.17815077e-01 -1.19640148e+00 -6.45755231e-01 -3.46357167e-01
-3.54714692e-01 1.71347395e-01 5.52952707e-01 -4.23373580e-01
5.98921418e-01 4.03519630e-01 -1.70104408e+00 2.90405869e-01
5.67749143e-01 9.79140997e-01 7.46584982e-02 3.28386605e-01
1.91028994e-02 1.07433867e+00 -6.48412466e-01 -6.49840236e-01
4.17953283e-01 -3.46223742e-01 -3.90920073e-01 -3.20188135e-01
9.24474522e-02 -1.03709243e-01 -6.15728140e-01 5.76945066e-01
3.08655322e-01 7.05287576e-01 3.55487525e-01 -1.36883771e+00
-5.16961217e-01 7.40194082e-01 -5.69751441e-01 7.43929625e-01
4.18881178e-01 2.09356129e-01 1.00256586e+00 -7.97087252e-01
3.64151478e-01 8.97431970e-01 7.94370353e-01 -2.67150819e-01
-7.48075008e-01 -8.50199103e-01 3.91229421e-01 6.68789208e-01
-1.37559950e+00 1.04316257e-01 6.32237792e-01 -5.95957041e-01
7.13981092e-01 9.72820595e-02 1.12896359e+00 9.51261580e-01
5.24688482e-01 8.92648697e-01 1.04706371e+00 -1.03679210e-01
-2.26056315e-02 -5.45977294e-01 4.08683449e-01 2.66108990e-01
-3.94603275e-02 4.55675542e-01 -1.15713131e+00 4.05790210e-02
7.94612348e-01 4.69251961e-01 2.05532625e-01 9.46065560e-02
-1.36332929e+00 4.53429729e-01 2.60944337e-01 1.90557599e-01
-6.84618413e-01 1.34458184e-01 5.09685040e-01 1.24298468e-01
4.27735865e-01 3.48677665e-01 -3.51097733e-02 -8.34974527e-01
-1.26449978e+00 6.60052776e-01 5.25921464e-01 5.76873779e-01
4.80187446e-01 2.66207987e-03 -8.09501708e-01 5.89625418e-01
-3.74966145e-01 1.14478968e-01 2.83604711e-01 -8.77423465e-01
6.36947215e-01 8.32089663e-01 -8.76166206e-03 -1.45547724e+00
-4.27456260e-01 -6.33162081e-01 -6.81132615e-01 -2.26450339e-01
5.08773685e-01 1.88017398e-01 -6.03008032e-01 1.53102684e+00
-2.34192293e-02 5.73934317e-01 -3.36779028e-01 1.03743005e+00
5.61805427e-01 6.78026795e-01 2.84781247e-01 -2.02297628e-01
1.37263036e+00 -6.05443239e-01 -7.65493155e-01 1.02434292e-01
1.47659987e-01 -5.14947593e-01 1.28172004e+00 3.66552681e-01
-1.15717471e+00 -9.77468133e-01 -9.15197909e-01 1.49963871e-01
-9.07558277e-02 2.87945122e-02 6.75724924e-01 2.14104235e-01
-5.45986652e-01 8.87244999e-01 -1.10646152e+00 -1.07681245e-01
1.77954897e-01 3.37594189e-03 -9.51768756e-02 5.79595491e-02
-1.35451007e+00 8.14141214e-01 5.92160523e-01 4.25954983e-02
-9.79010940e-01 -8.57513547e-01 -5.90270698e-01 2.84189522e-01
6.64648831e-01 -2.10686907e-01 1.02627647e+00 -1.38673991e-01
-1.08990729e+00 4.19521630e-01 -9.37944371e-03 -6.78623438e-01
5.40996969e-01 -4.09793794e-01 -5.91045916e-01 -5.81740178e-02
3.08326662e-01 -7.66604021e-02 4.55222502e-02 -6.85964465e-01
-9.77845669e-01 -4.14785147e-01 3.23258117e-02 3.07536185e-01
-2.77855992e-01 2.57432401e-01 -5.81478953e-01 -7.48748243e-01
2.40504801e-01 -7.68566787e-01 -1.75733268e-01 -5.03171504e-01
-2.15108499e-01 -3.42352748e-01 1.66190937e-01 -8.82642031e-01
1.89932537e+00 -2.32946038e+00 8.48416314e-02 2.24533886e-01
-3.26587334e-02 -7.89367184e-02 1.47988245e-01 8.36069345e-01
1.84444129e-01 -2.87262909e-03 1.97267160e-01 -2.71499958e-02
4.49875817e-02 1.26661554e-01 -3.12308460e-01 1.61231086e-01
-7.48993084e-02 8.49041402e-01 -8.71206224e-01 -6.52082980e-01
1.24514356e-01 -4.60256971e-02 -2.83469141e-01 -7.25210411e-03
1.50148377e-01 3.04175347e-01 -4.46682453e-01 7.90539384e-01
1.77179411e-01 1.00326359e-01 -1.11174740e-01 -1.51902065e-01
-5.35830796e-01 8.03147182e-02 -1.19888067e+00 1.61921000e+00
-1.07560508e-01 5.32316566e-01 -6.47689819e-01 -8.94130230e-01
1.16561496e+00 2.58275151e-01 8.71647120e-01 -1.11315262e+00
-7.98269585e-02 -1.16173849e-01 -2.49328651e-02 -3.89692545e-01
1.05028248e+00 -2.74661869e-01 -5.66221833e-01 3.17419589e-01
-7.36324117e-02 4.90457505e-01 6.67469501e-01 5.14037944e-02
1.23185337e+00 4.07602787e-01 2.07474083e-01 4.67122532e-03
-2.00526062e-02 2.58752167e-01 9.13184643e-01 7.90695250e-01
-3.61966699e-01 5.82760036e-01 5.89013100e-01 -7.35151708e-01
-9.72284317e-01 -1.30986142e+00 1.19058512e-01 1.25227606e+00
3.17882627e-01 -6.83487117e-01 -6.71343058e-02 -8.97717774e-02
-2.47285981e-02 4.88713920e-01 -4.99625564e-01 -3.14689696e-01
-7.93353379e-01 -7.78748453e-01 8.57213736e-01 9.40838516e-01
6.67293012e-01 -9.38758790e-01 -5.13339460e-01 5.87849200e-01
-6.04799688e-01 -9.94873047e-01 -3.95057738e-01 -1.06774665e-01
-6.86089098e-01 -9.99836445e-01 -5.47793329e-01 -4.29400355e-01
-2.36266509e-01 2.35058740e-02 1.04095316e+00 -2.28955343e-01
2.36615688e-02 2.04505101e-02 -4.35156107e-01 -6.16664052e-01
2.27196082e-01 1.33040920e-01 2.94686675e-01 4.77564596e-02
5.86481094e-01 -7.27877021e-01 -6.14101052e-01 4.56969798e-01
-6.41106606e-01 1.07138544e-01 5.69787443e-01 4.82099712e-01
8.33170414e-01 1.35354355e-01 4.80020285e-01 -7.45532662e-02
9.02273834e-01 -5.54086983e-01 -2.34862134e-01 3.35103959e-01
-5.06080329e-01 -3.14187080e-01 3.65787417e-01 -5.85904896e-01
-1.02698839e+00 -4.26678628e-01 4.26650979e-02 -4.94263291e-01
1.10056981e-01 8.54916453e-01 2.17246264e-01 6.23504937e-01
7.15203285e-01 5.52552462e-01 -3.01347762e-01 -4.82382566e-01
-5.46796657e-02 1.79983899e-01 7.90156305e-01 -8.27056408e-01
6.21001005e-01 4.34210122e-01 1.17096268e-01 -3.44428420e-01
-9.14570987e-01 -6.90044701e-01 -6.48688495e-01 -6.61663115e-01
8.59637380e-01 -1.14471054e+00 -1.05022407e+00 4.27437544e-01
-7.17145860e-01 2.81611662e-02 -5.05066395e-01 9.78856385e-01
-4.88102973e-01 3.09418052e-01 -9.88327563e-01 -8.59557211e-01
1.38115093e-01 -7.40434289e-01 7.53207207e-01 2.48883694e-01
-2.76670486e-01 -6.07616186e-01 3.22280198e-01 3.60666454e-01
8.08862597e-02 5.65466464e-01 5.69122314e-01 -5.75226128e-01
-5.93582571e-01 -4.51795280e-01 -9.85809416e-02 2.70025488e-02
-7.74538741e-02 -1.41749397e-01 -3.07560652e-01 1.24990635e-01
-1.43025234e-01 1.65098026e-01 7.20710933e-01 6.02839530e-01
1.09822679e+00 1.17714480e-01 -1.29342958e-01 3.71911824e-01
1.27106345e+00 2.59492606e-01 6.16163731e-01 7.54231989e-01
5.61125994e-01 4.94768262e-01 1.15911233e+00 6.64933205e-01
4.62450475e-01 8.44310164e-01 3.76083218e-02 1.01383552e-01
6.58853650e-02 -7.85036147e-01 2.82608151e-01 8.94010901e-01
-8.64727855e-01 -1.38836220e-01 -1.02132607e+00 7.59692729e-01
-2.23988748e+00 -1.53996575e+00 -3.87690008e-01 2.02206588e+00
4.61451203e-01 4.07523215e-01 4.40639436e-01 3.11655223e-01
7.99406111e-01 1.73926890e-01 -2.99973547e-01 -2.36279801e-01
-1.52199075e-01 1.24957584e-01 5.50891221e-01 -2.22402170e-01
-9.48085308e-01 7.31793284e-01 6.37306452e+00 1.05368614e+00
-5.80327511e-01 1.70762777e-01 3.10675591e-01 -7.14318693e-01
1.72104672e-01 7.83929899e-02 -8.70743871e-01 6.73034549e-01
8.57891500e-01 -5.20610154e-01 -6.48562098e-03 5.59522867e-01
5.47398090e-01 -3.09964735e-02 -8.40086997e-01 8.51926625e-01
-1.52791813e-01 -1.22430718e+00 1.40549820e-02 1.98568285e-01
5.00448048e-01 -2.24330842e-01 -1.55913401e-02 8.13555002e-01
4.47281450e-01 -7.90722549e-01 8.72198761e-01 1.11350870e+00
3.63837212e-01 -8.88422251e-01 6.20930493e-01 5.70370376e-01
-1.47066426e+00 -4.47725624e-01 -2.34297767e-01 -7.03466177e-01
7.03496277e-01 4.05188829e-01 -2.65758812e-01 9.53680336e-01
1.02994156e+00 8.63582730e-01 -3.78796905e-01 1.28502560e+00
6.51645362e-02 7.79542327e-01 -2.92697757e-01 1.22629844e-01
2.64910042e-01 -1.95325702e-01 5.01307845e-01 9.96457040e-01
7.10509598e-01 1.21632375e-01 5.12866676e-01 6.31395400e-01
2.64020324e-01 2.08243147e-01 -2.58925915e-01 7.12378416e-03
4.38239992e-01 8.43196094e-01 -7.72663713e-01 -2.77817160e-01
-3.22624654e-01 4.65892047e-01 1.90659508e-01 2.03350395e-01
-1.11345208e+00 4.80404589e-03 7.69475698e-01 4.90791976e-01
-8.86290371e-02 -3.69021177e-01 -3.63800257e-01 -1.09851444e+00
2.91328758e-01 -5.43674707e-01 6.01935625e-01 -6.66812599e-01
-1.33610499e+00 1.93961248e-01 4.75743026e-01 -1.59181309e+00
-2.69665550e-02 -6.60813972e-02 -7.61288226e-01 7.46639550e-01
-1.13452518e+00 -1.15352952e+00 -3.51325929e-01 5.80565751e-01
5.42308927e-01 -3.40699732e-01 2.31951103e-01 4.87706155e-01
-6.50328934e-01 2.71477520e-01 -2.74852186e-01 2.37542301e-01
5.31949043e-01 -8.43878925e-01 2.90780932e-01 9.92020607e-01
2.22165421e-01 4.79161710e-01 7.32073545e-01 -9.25039709e-01
-9.19499576e-01 -8.77670109e-01 6.74586117e-01 -3.93194824e-01
7.71113217e-01 1.00779265e-01 -8.22344005e-01 6.94530904e-01
-2.31874928e-01 -5.63393593e-01 8.92641962e-01 5.56141853e-01
2.56351948e-01 -1.52683496e-01 -4.03854012e-01 7.42155671e-01
1.37022352e+00 -4.04756933e-01 -9.06218708e-01 1.17889993e-01
6.21590555e-01 -3.47678661e-01 -1.20825779e+00 6.04268193e-01
7.19470143e-01 -7.92678952e-01 1.18005264e+00 -5.55054784e-01
6.69269919e-01 -3.93411458e-01 -1.54004574e-01 -1.00091958e+00
-7.37125754e-01 -1.26367779e-02 -1.46114871e-01 1.16660595e+00
7.68447444e-02 -2.88264025e-02 6.72439754e-01 5.39579630e-01
-4.98762280e-01 -7.81623065e-01 -9.49377537e-01 -1.01470804e+00
-1.45786598e-01 -9.34415400e-01 9.80626225e-01 7.07572460e-01
2.38388002e-01 -9.76707041e-02 -8.68657827e-01 2.24401001e-02
4.48062956e-01 3.10756326e-01 6.75437093e-01 -1.22007275e+00
-4.25243795e-01 -5.02201974e-01 -7.61727810e-01 -7.77422249e-01
-2.79706985e-01 -6.76483870e-01 -2.29313299e-01 -1.65878201e+00
3.62025321e-01 -1.91672176e-01 -7.64070570e-01 3.54225993e-01
-2.71126419e-01 1.92826703e-01 4.31883603e-01 6.13007367e-01
-6.74951494e-01 3.94535810e-01 1.27816629e+00 1.21432737e-01
-3.54242742e-01 3.63467902e-01 -4.54284936e-01 5.40588439e-01
7.54341364e-01 -4.97506887e-01 -2.61788666e-01 -2.42267311e-01
5.19528091e-01 5.10034740e-01 5.27196646e-01 -1.36514878e+00
5.71924567e-01 -4.90759313e-01 3.52601737e-01 -9.50812578e-01
5.05266488e-01 -2.54534155e-01 4.18836027e-01 2.99954981e-01
-3.62718880e-01 3.56240243e-01 1.18669188e-02 6.54739261e-01
-7.82665670e-01 2.35872194e-01 -1.84862763e-02 -1.33061722e-01
-1.12083757e+00 5.50888002e-01 -4.16194767e-01 -2.33803526e-01
1.29776275e+00 -6.43544137e-01 -8.97110850e-02 -1.99572504e-01
-1.13127196e+00 3.51962298e-01 2.53935248e-01 6.27400339e-01
6.11187518e-01 -1.47296011e+00 -8.35935175e-01 -1.23696096e-01
2.47085184e-01 -3.91158789e-01 7.12809265e-01 9.91830945e-01
-4.22683984e-01 1.79492995e-01 -5.71063817e-01 -5.78805625e-01
-9.79918957e-01 5.55163980e-01 5.89101985e-02 -6.19508862e-01
-7.71343291e-01 3.74275476e-01 1.44919092e-02 -2.23934591e-01
3.26015390e-02 -4.04715896e-01 -4.25449342e-01 2.60756671e-01
5.27029395e-01 6.91123486e-01 -1.01790607e-01 -5.13472795e-01
-6.16230480e-02 3.48641872e-01 3.46865147e-01 -1.38099998e-01
1.29662538e+00 -2.06883978e-02 3.24624747e-01 8.81833911e-01
4.49418843e-01 -1.92173034e-01 -1.29767299e+00 -2.92778403e-01
1.21266678e-01 -5.53264558e-01 -2.23574683e-01 -5.01827121e-01
-8.50857377e-01 7.14711785e-01 4.03990924e-01 1.31864190e-01
1.12239718e+00 -2.13917986e-01 9.90226805e-01 5.26255704e-02
6.76280737e-01 -1.20994210e+00 1.26145169e-01 4.04515624e-01
4.48191136e-01 -7.75240660e-01 2.07600789e-03 -2.23871469e-01
-1.00329125e+00 8.01283360e-01 5.65274239e-01 -1.91325590e-01
6.96646154e-01 1.11197695e-01 -3.74349684e-01 -3.68052125e-01
-6.75583422e-01 -2.43605986e-01 3.99554849e-01 2.32665524e-01
2.41277874e-01 3.54326040e-01 -5.92119515e-01 1.06994903e+00
-5.83829284e-01 2.81748921e-01 3.46152544e-01 9.90865052e-01
-4.70857352e-01 -8.10271144e-01 -3.10009241e-01 6.03640437e-01
-6.22862101e-01 6.65137321e-02 1.84764355e-01 8.88410330e-01
3.41943979e-01 8.18343282e-01 2.97546923e-01 -7.46950805e-01
8.73666763e-01 7.08391517e-02 2.60394752e-01 -5.42145610e-01
-9.46043372e-01 -2.26431102e-01 -2.76877582e-02 -7.04577565e-01
-4.21324939e-01 -9.03654873e-01 -9.43064570e-01 -6.92507565e-01
-1.90893859e-02 1.58428058e-01 3.54715317e-01 1.00052667e+00
1.20912269e-01 7.23449588e-01 3.01779389e-01 -5.95528543e-01
-2.36186922e-01 -1.10901463e+00 -8.63406479e-01 6.30264938e-01
-4.21534181e-01 -8.97053540e-01 1.86451226e-01 -5.16578555e-02] | [6.788328170776367, 0.35883572697639465] |
ba16d0e4-5861-4e92-8721-64cdce0220db | snel-a-structured-neuro-symbolic-language-for | 2306.06036 | null | https://arxiv.org/abs/2306.06036v1 | https://arxiv.org/pdf/2306.06036v1.pdf | SNeL: A Structured Neuro-Symbolic Language for Entity-Based Multimodal Scene Understanding | In the evolving landscape of artificial intelligence, multimodal and Neuro-Symbolic paradigms stand at the forefront, with a particular emphasis on the identification and interaction with entities and their relations across diverse modalities. Addressing the need for complex querying and interaction in this context, we introduce SNeL (Structured Neuro-symbolic Language), a versatile query language designed to facilitate nuanced interactions with neural networks processing multimodal data. SNeL's expressive interface enables the construction of intricate queries, supporting logical and arithmetic operators, comparators, nesting, and more. This allows users to target specific entities, specify their properties, and limit results, thereby efficiently extracting information from a scene. By aligning high-level symbolic reasoning with low-level neural processing, SNeL effectively bridges the Neuro-Symbolic divide. The language's versatility extends to a variety of data types, including images, audio, and text, making it a powerful tool for multimodal scene understanding. Our evaluations demonstrate SNeL's potential to reshape the way we interact with complex neural networks, underscoring its efficacy in driving targeted information extraction and facilitating a deeper understanding of the rich semantics encapsulated in multimodal AI models. | ['Ivanovitch Silva', 'Allan Martins', 'Silvan Ferreira'] | 2023-06-09 | null | null | null | null | ['scene-understanding'] | ['computer-vision'] | [ 4.07053083e-01 1.84680149e-01 -2.72335917e-01 -4.13311005e-01
-3.77145499e-01 -9.74705517e-01 7.73445129e-01 6.67948723e-01
-5.55752575e-01 3.03698719e-01 4.43741888e-01 -2.32566610e-01
-3.70399714e-01 -8.08929443e-01 -1.33080304e-01 -4.42385264e-02
-2.53630191e-01 2.83458740e-01 8.94803032e-02 -5.87334931e-01
4.47251908e-02 7.02770293e-01 -1.99511564e+00 7.12633073e-01
4.09250498e-01 1.17378235e+00 -1.24998160e-01 4.86801028e-01
-5.02141416e-01 1.16752958e+00 -5.18648624e-01 -4.35844004e-01
-4.41977590e-01 -7.68102333e-02 -9.07081485e-01 -3.48507047e-01
4.03785378e-01 -4.40392531e-02 -7.31061026e-02 8.91939580e-01
2.81997889e-01 -2.76232045e-02 3.12102944e-01 -1.39040399e+00
-2.25800157e-01 7.55682468e-01 7.40123019e-02 6.94095194e-02
8.82542849e-01 1.42273143e-01 1.14522707e+00 -8.22879851e-01
8.12619627e-01 1.48109639e+00 6.14020109e-01 3.99948329e-01
-1.43916500e+00 -3.36079121e-01 9.09310505e-02 -4.71503846e-03
-1.32803512e+00 -8.04070175e-01 4.44016755e-01 -5.23136854e-01
1.07782137e+00 6.52047873e-01 8.01108539e-01 7.97864795e-01
-4.58025068e-01 9.65012312e-01 5.15636027e-01 -4.13356245e-01
2.35775799e-01 1.48311213e-01 2.10505784e-01 7.66854644e-01
-1.41384423e-01 -3.15557718e-02 -1.24484491e+00 -2.16257781e-01
6.13214970e-01 -2.92368591e-01 -1.57023251e-01 -2.27786601e-01
-1.56798220e+00 5.14319062e-01 3.54727030e-01 3.19090128e-01
-2.76948124e-01 2.10161522e-01 7.43217230e-01 2.86971122e-01
-2.09756628e-01 8.93445551e-01 -2.35851228e-01 -3.56427401e-01
-5.77114820e-01 3.19537848e-01 9.23391283e-01 9.69346404e-01
5.51778734e-01 -9.58887190e-02 -2.08497167e-01 1.02751625e+00
3.55117202e-01 4.14024889e-01 1.86670497e-01 -1.33536339e+00
4.19190615e-01 1.03613961e+00 -7.94734806e-02 -1.03038824e+00
-5.57598352e-01 -2.59170830e-01 -5.12376606e-01 2.29690224e-01
4.46459740e-01 -5.14010489e-02 -4.59361911e-01 2.11405301e+00
2.44588554e-01 -6.58640742e-01 1.91547722e-01 6.98360264e-01
9.81620193e-01 2.29974821e-01 7.10489810e-01 2.81152576e-01
1.61424446e+00 -1.17066137e-01 -6.24341428e-01 -3.98402959e-01
7.84122169e-01 -2.56893009e-01 1.15924954e+00 3.72076571e-01
-1.29098761e+00 -2.17500910e-01 -8.99660408e-01 -5.08689940e-01
-1.11310625e+00 -8.47406685e-02 9.26513851e-01 3.83801371e-01
-1.16159606e+00 2.90452093e-01 -6.42543852e-01 -4.40959930e-01
5.47740102e-01 5.45923650e-01 -6.21391952e-01 7.15192705e-02
-1.32958472e+00 9.47366834e-01 7.42947996e-01 1.14781298e-01
-5.23955338e-02 -6.66584790e-01 -1.27712965e+00 2.12268859e-01
4.96142358e-01 -9.68469143e-01 1.28554559e+00 -7.36291230e-01
-1.08499813e+00 9.41080630e-01 -1.10995382e-01 -2.54503727e-01
2.42666751e-02 9.00190789e-04 -6.44777477e-01 1.64164126e-01
-1.13843322e-01 1.13712192e+00 2.26161629e-01 -1.05348957e+00
-5.32697976e-01 -6.50185823e-01 2.26357728e-01 5.14311910e-01
-2.31728390e-01 2.72646964e-01 -5.18022358e-01 -4.08269465e-01
2.07291067e-01 -2.43016422e-01 1.27826348e-01 4.12914068e-01
-4.27393883e-01 -1.95966750e-01 8.19204271e-01 -1.76689953e-01
1.34625864e+00 -2.37471247e+00 4.41981941e-01 5.19447148e-01
6.38541281e-01 -5.43203205e-02 2.71798708e-02 7.82133758e-01
-6.92604342e-03 1.23212911e-01 -1.31243199e-01 -3.45919132e-01
4.83649313e-01 2.89731503e-01 -3.51927042e-01 -2.57471561e-01
4.20120507e-01 1.23549247e+00 -8.96713316e-01 -6.83394969e-01
1.62679479e-01 5.45792103e-01 -4.72402096e-01 -1.31493315e-01
-5.39631963e-01 1.62638482e-02 -2.96211898e-01 1.04545915e+00
-9.52809379e-02 -2.29777291e-01 1.94733307e-01 -4.43463147e-01
-2.90156215e-01 5.95476814e-02 -1.05508721e+00 1.82694077e+00
-2.54301488e-01 8.07339311e-01 4.42352891e-01 -3.77546042e-01
5.95815301e-01 1.38431057e-01 2.21200302e-01 -7.02893257e-01
2.71957785e-01 1.88132040e-02 -2.31549695e-01 -8.66987109e-01
6.30039513e-01 -8.84333625e-02 -3.07249904e-01 3.36621791e-01
1.38259470e-01 -2.92858720e-01 4.23848689e-01 4.18124110e-01
8.62448394e-01 1.57418460e-01 2.50605673e-01 5.94463944e-02
3.75357181e-01 2.12508649e-01 -9.19429809e-02 6.88182354e-01
1.26599580e-01 -3.33381630e-02 6.38495982e-01 -1.86398938e-01
-5.50494611e-01 -1.21995056e+00 -1.04932621e-01 1.50898409e+00
-1.62612330e-02 -5.17777026e-01 -4.94448900e-01 1.53484717e-02
7.63711408e-02 6.07491732e-01 -4.31346297e-01 -9.14803073e-02
-1.03843413e-01 -2.23306984e-01 9.87936735e-01 4.00543302e-01
4.00273800e-01 -1.25690734e+00 -1.07624459e+00 -5.51191112e-03
-2.96060711e-01 -1.21197021e+00 2.09950190e-02 2.16791511e-01
-7.71486700e-01 -9.32026923e-01 -1.21801198e-01 -5.27171910e-01
4.03542936e-01 -3.49597245e-01 1.35515332e+00 -2.46329933e-01
-3.70059520e-01 7.93419957e-01 3.89963086e-03 -5.43684900e-01
-3.87846291e-01 7.96229765e-02 -2.74146676e-01 -2.51321495e-01
3.38507622e-01 -5.63130677e-01 -1.86791584e-01 5.96971363e-02
-1.39287663e+00 3.78312349e-01 4.52608258e-01 4.53846306e-01
2.81595767e-01 -3.11334461e-01 5.19269288e-01 -7.42751956e-01
9.06322062e-01 -6.00628555e-01 -4.56012100e-01 4.42200363e-01
-1.30165547e-01 1.97123110e-01 2.69900948e-01 -4.64944482e-01
-7.65186310e-01 -1.65260248e-02 1.12532146e-01 -8.74653384e-02
-4.98371601e-01 1.06668389e+00 -3.32517207e-01 5.20153418e-02
6.67832434e-01 -4.24422808e-02 3.83738995e-01 -1.92670360e-01
7.55764484e-01 4.10107762e-01 1.17319345e+00 -7.47136056e-01
3.90528560e-01 4.20489192e-01 1.70624584e-01 -8.66965413e-01
-3.79700154e-01 -1.40080482e-01 -5.26191056e-01 -3.61935824e-01
8.55484128e-01 -6.43148363e-01 -1.35160518e+00 2.96751678e-01
-1.11156952e+00 -1.76304087e-01 -5.16081572e-01 3.75928253e-01
-4.08686489e-01 1.71070099e-01 -5.01730621e-01 -8.32010567e-01
-1.04387231e-01 -9.08300221e-01 1.06940031e+00 2.31750421e-02
-8.50929141e-01 -9.43884254e-01 -2.35199586e-01 2.28071943e-01
4.05551523e-01 4.53194380e-01 1.39632761e+00 -8.42360079e-01
-5.33078492e-01 -3.53629947e-01 -4.72691238e-01 -2.62453496e-01
-2.36252576e-01 -1.65973991e-01 -1.08885670e+00 4.29396659e-01
-4.76129234e-01 -5.83712459e-01 3.70882660e-01 -3.10205311e-01
9.24290359e-01 -1.04714118e-01 -2.37584263e-01 4.03408706e-01
1.10321486e+00 1.87320352e-01 2.60903388e-01 1.71818674e-01
5.31074405e-01 1.02225256e+00 1.69379085e-01 3.47153097e-01
6.45761073e-01 4.94799167e-01 7.17058361e-01 -8.35282952e-02
1.82921648e-01 -2.67544568e-01 -9.43884999e-03 3.06212872e-01
2.97019124e-01 -7.53998831e-02 -1.23020649e+00 1.43820345e-01
-1.76883113e+00 -1.03219843e+00 1.43761948e-01 2.02368569e+00
1.00845480e+00 -2.48433817e-02 9.00274888e-02 3.17773148e-02
2.98197269e-01 9.09694135e-02 -6.44995391e-01 -3.89304340e-01
-3.46314907e-01 -6.78917319e-02 5.88363372e-02 4.82118696e-01
-8.77988994e-01 9.20440912e-01 7.18526745e+00 5.88427663e-01
-1.10348558e+00 -4.92945522e-01 2.41054401e-01 -2.14657515e-01
-5.40221930e-01 -3.90902787e-01 -3.82187605e-01 -4.45432626e-02
7.68895149e-01 1.15350835e-01 8.92750025e-01 3.56978685e-01
2.24128626e-02 -3.81015569e-01 -1.64055049e+00 1.21333897e+00
1.82763543e-02 -1.45879018e+00 2.35264525e-01 -1.60497561e-01
-1.88264489e-01 3.79199088e-02 -1.26198158e-01 2.47426629e-01
1.73875645e-01 -1.44208515e+00 1.01357162e+00 8.44264269e-01
9.64727640e-01 -4.40498352e-01 2.98260957e-01 2.82738775e-01
-1.04999053e+00 -2.24245250e-01 4.66469526e-01 -3.13582569e-02
1.95554480e-01 8.63869041e-02 -5.73097408e-01 3.83875370e-01
3.97129118e-01 4.14144278e-01 -6.07630491e-01 7.93040395e-01
2.31482968e-01 -1.95075408e-01 -6.98679924e-01 -1.21107809e-01
-9.92820188e-02 8.94612968e-02 5.66230118e-01 1.49263966e+00
9.61847417e-03 1.79655090e-01 -1.39057592e-01 1.06349826e+00
7.80679807e-02 -5.48509834e-03 -9.25827265e-01 -4.46318418e-01
8.27617943e-01 9.10193384e-01 -6.12096131e-01 -3.54359657e-01
-2.78096288e-01 5.29707491e-01 2.92260259e-01 5.48747838e-01
-5.10246456e-01 -6.31000280e-01 5.70654035e-01 -1.51838347e-01
-2.37722278e-01 -5.05142152e-01 -5.82617283e-01 -9.26456630e-01
-1.45353809e-01 -1.11482728e+00 5.38170159e-01 -1.21498060e+00
-8.27214122e-01 6.12382591e-01 4.64694381e-01 -7.85615087e-01
-4.65294272e-01 -6.74246192e-01 -1.91643864e-01 7.63444006e-01
-8.20606351e-01 -1.29126942e+00 -2.91121632e-01 6.93782091e-01
6.19408004e-02 -2.53499262e-02 1.16639268e+00 2.18020707e-01
-3.90765071e-01 2.98821539e-01 -5.95031142e-01 1.32263526e-01
2.10589945e-01 -8.50541711e-01 1.47710815e-02 2.72538930e-01
1.48593798e-01 8.94693017e-01 5.28808951e-01 -2.92216212e-01
-1.78403330e+00 -3.11889261e-01 7.10256815e-01 -7.05793858e-01
7.94324815e-01 -4.52181727e-01 -7.91576445e-01 5.37008107e-01
1.04682483e-01 -3.72034490e-01 8.65205050e-01 2.75418907e-01
-8.58611822e-01 1.23846931e-02 -1.04426610e+00 9.76510644e-01
1.07198584e+00 -1.17410672e+00 -5.57608128e-01 -2.24968970e-01
6.43896043e-01 -4.40879375e-01 -9.88384306e-01 5.17979801e-01
8.52854729e-01 -9.03733730e-01 1.15236568e+00 -6.74169660e-01
2.36848146e-01 -2.46886894e-01 -3.97893876e-01 -6.07289672e-01
1.58624813e-01 -5.42212844e-01 -1.14190631e-01 1.17287600e+00
6.40176952e-01 -5.12676716e-01 4.35696095e-01 1.19516349e+00
1.09124973e-01 -7.85535276e-01 -6.70705199e-01 -2.06128452e-02
-5.50173759e-01 -1.12120759e+00 7.39520073e-01 9.62866545e-01
7.01434314e-01 3.64931136e-01 3.41778070e-01 8.50227028e-02
2.74212480e-01 -2.56477714e-01 4.69442695e-01 -1.37463975e+00
-1.81477398e-01 -9.86993432e-01 -5.88683069e-01 -8.00397456e-01
-2.58719400e-02 -9.59847391e-01 -3.57416034e-01 -1.35032415e+00
-1.74783185e-01 -2.67347217e-01 -1.25947660e-02 9.39153016e-01
6.00548029e-01 1.69998169e-01 3.66427779e-01 9.56767276e-02
-6.71843290e-01 2.76197672e-01 8.84034276e-01 -1.52149320e-01
-2.65270025e-01 -4.50589299e-01 -8.31063032e-01 9.07854617e-01
5.44061720e-01 1.21905275e-01 -5.52562594e-01 -6.28143072e-01
7.57725239e-01 1.93549860e-02 7.58902848e-01 -1.01265073e+00
6.87755823e-01 -1.34936348e-01 2.74081707e-01 -4.03502196e-01
8.45993280e-01 -1.00283229e+00 2.20115677e-01 4.30667587e-02
-8.66684496e-01 9.90768299e-02 7.34328032e-01 2.37972587e-01
-4.36248839e-01 2.80143358e-02 2.32538372e-01 -4.69530560e-02
-9.14156914e-01 -1.24535084e-01 -4.72629160e-01 1.82114750e-01
6.74519181e-01 -4.86599624e-01 -5.18471062e-01 -4.30996537e-01
-1.25006962e+00 4.96647030e-01 2.76575595e-01 5.29825866e-01
6.90146387e-01 -1.17548180e+00 7.86420926e-02 5.21145821e-01
4.72777337e-01 1.88582409e-02 3.41030490e-03 6.92709088e-01
-3.18325371e-01 3.14827412e-01 -1.15281083e-01 -5.26399910e-01
-1.33309162e+00 1.89773366e-01 4.23405349e-01 2.70153791e-01
-1.48232535e-01 6.62191808e-01 -1.41017973e-01 -5.60015976e-01
7.53816545e-01 -5.42276382e-01 -3.27194244e-01 4.31711107e-01
4.68512744e-01 1.76170006e-01 -2.13429958e-01 -5.42905509e-01
-3.75596911e-01 2.83964574e-01 3.46411467e-01 -5.90087771e-01
1.02843988e+00 -1.05588265e-01 -7.22302079e-01 8.84702921e-01
9.08139229e-01 -2.58208495e-02 -6.66427612e-01 -2.31206864e-01
3.90093774e-01 1.31368846e-01 -1.64044738e-01 -1.30245185e+00
-3.37024301e-01 8.32600594e-01 1.37678370e-01 4.54643309e-01
1.12648404e+00 4.06066179e-01 3.90234023e-01 7.59091437e-01
3.69702131e-01 -8.34723592e-01 -2.46707782e-01 7.61320055e-01
9.56034362e-01 -9.07354712e-01 -2.45357126e-01 -4.02912855e-01
-6.18536413e-01 1.34785819e+00 4.94571686e-01 7.69219637e-01
3.74710947e-01 5.66700459e-01 1.53412819e-01 -6.98188663e-01
-8.59607756e-01 -2.92914838e-01 3.43770742e-01 5.68068922e-01
4.45654958e-01 -2.16640308e-01 4.77531701e-01 3.58966738e-01
-2.33097389e-01 1.74219370e-01 -1.40735209e-01 1.06659222e+00
-4.54735130e-01 -1.07817960e+00 -4.48090494e-01 3.41181278e-01
-2.62462348e-01 -2.77867019e-01 -7.17376053e-01 9.00968432e-01
2.74146974e-01 7.73414373e-01 9.96285155e-02 -2.61854649e-01
4.10250872e-01 4.32133555e-01 3.77729923e-01 -5.70573211e-01
-7.71930933e-01 -2.26485372e-01 4.56682622e-01 -7.36199617e-01
-3.36062163e-01 -5.33736527e-01 -1.49971712e+00 -2.77180206e-02
2.63505638e-01 -8.30449015e-02 8.20297360e-01 9.19604838e-01
6.35019183e-01 3.94193113e-01 -2.82553017e-01 -7.50319958e-01
-2.42827430e-01 -4.75654840e-01 -2.13025078e-01 1.94290400e-01
5.05682468e-01 -3.30559045e-01 4.89973649e-02 5.81028499e-02] | [10.690797805786133, 1.845304012298584] |
5cda6aa9-c192-4f99-afe2-44464cc3a8a7 | a-prototype-malayalam-to-sign-language | 1412.7415 | null | http://arxiv.org/abs/1412.7415v2 | http://arxiv.org/pdf/1412.7415v2.pdf | A prototype Malayalam to Sign Language Automatic Translator | Sign language, which is a medium of communication for deaf people, uses
manual communication and body language to convey meaning, as opposed to using
sound. This paper presents a prototype Malayalam text to sign language
translation system. The proposed system takes Malayalam text as input and
generates corresponding Sign Language. Output animation is rendered using a
computer generated model. This system will help to disseminate information to
the deaf people in public utility places like railways, banks, hospitals etc.
This will also act as an educational tool in learning Sign Language. | ['Kannan Balakrishnan', 'Jestin Joy'] | 2014-12-23 | null | null | null | null | ['sign-language-translation'] | ['computer-vision'] | [ 1.47079498e-01 -1.44103661e-01 8.18605274e-02 -1.75360799e-01
-6.97577372e-02 -6.56796217e-01 7.49267936e-01 -2.59485453e-01
-6.90541327e-01 1.06984103e+00 7.91308165e-01 -7.04858899e-01
1.75462484e-01 -8.17498446e-01 1.20498680e-01 -5.26393652e-01
2.19024733e-01 4.00554687e-01 4.11726385e-01 -5.74767530e-01
6.68104291e-01 5.61500490e-01 -1.26193595e+00 1.73420280e-01
5.21505475e-01 -5.77786863e-02 2.96204388e-01 1.12461579e+00
-6.29926801e-01 9.29375231e-01 -7.51397848e-01 -1.01469040e-01
7.29837939e-02 -8.99876416e-01 -8.01533520e-01 -1.92139581e-01
9.00972039e-02 -7.69956112e-01 -3.69712152e-02 1.00195825e+00
6.45230711e-01 -1.44820005e-01 9.91667151e-01 -9.77348864e-01
-5.89026690e-01 4.68050480e-01 -3.09587210e-01 -4.71178353e-01
8.35722744e-01 -9.23936665e-02 9.09171030e-02 -5.47070026e-01
9.28766727e-01 1.39042175e+00 2.64693528e-01 7.52797782e-01
-6.18203163e-01 -8.82425547e-01 -2.88776636e-01 1.62654370e-01
-1.06780541e+00 -8.38402137e-02 6.51663005e-01 -3.00013602e-01
6.85488462e-01 2.20182449e-01 1.14688683e+00 2.14988261e-01
3.40438783e-01 7.19473362e-01 1.53732491e+00 -1.13859892e+00
8.60557035e-02 7.85677806e-02 1.39133364e-01 6.58891737e-01
2.26397008e-01 -2.54467726e-01 -4.50856358e-01 -1.19108267e-01
8.97201538e-01 -2.64719605e-01 -4.41615671e-01 3.86486530e-01
-9.38329339e-01 6.43776655e-01 -7.12685511e-02 5.55797040e-01
-3.50697488e-01 2.59167433e-01 1.80699304e-01 5.48818767e-01
-3.64404082e-01 -5.03611624e-01 -1.76214665e-01 -8.24599802e-01
-7.12114513e-01 1.98756546e-01 1.08690619e+00 7.12786019e-01
-1.27021983e-01 4.27610189e-01 3.76703739e-01 6.37106419e-01
1.21622980e+00 9.80766416e-01 5.90590715e-01 -4.62939978e-01
2.19122842e-01 5.21673918e-01 1.46631047e-01 -3.87774020e-01
-2.66437680e-01 5.15361667e-01 -3.09190542e-01 1.23685730e+00
5.38727283e-01 -4.48860943e-01 -1.48351085e+00 1.01646662e+00
3.63787413e-01 -2.99736857e-01 3.43713790e-01 8.02830935e-01
1.20666087e+00 5.76183021e-01 1.12292588e-01 1.22738615e-01
1.52856410e+00 -3.77338111e-01 -8.42226624e-01 6.04489684e-01
5.42716682e-01 -1.59222770e+00 1.00932717e+00 4.65106964e-01
-9.98766840e-01 1.33303881e-01 -6.91268861e-01 -1.46458456e-02
-5.46449721e-01 3.19719873e-02 2.69225806e-01 8.73627186e-01
-1.03378618e+00 1.25089839e-01 -6.18604124e-01 -9.32011664e-01
2.23820835e-01 4.48465317e-01 -2.34904587e-01 3.87537256e-02
-8.80920112e-01 1.31486535e+00 2.54608154e-01 1.61484838e-01
7.34160841e-02 7.85198957e-02 -8.63150299e-01 -6.74746037e-01
-5.80282629e-01 -4.23957765e-01 1.37512219e+00 -7.16208696e-01
-2.05180812e+00 9.23697293e-01 -4.30024594e-01 -9.03475732e-02
1.03857100e+00 -6.18036017e-02 -5.12641072e-01 1.60318822e-01
-2.93038636e-01 5.32175720e-01 4.60886508e-01 -1.11398053e+00
-9.71646130e-01 -6.33425340e-02 -1.75280899e-01 1.45185441e-01
5.73253334e-01 7.01371253e-01 -9.02101323e-02 -6.62719369e-01
3.79122496e-01 -7.26554573e-01 1.38056248e-01 2.78381795e-01
-1.18686043e-01 -1.11754142e-01 1.54659557e+00 -1.14513195e+00
8.98457348e-01 -1.57450008e+00 -2.90364683e-01 6.22898877e-01
-3.37929785e-01 5.05529523e-01 1.96789950e-01 9.55559194e-01
4.51052725e-01 -5.69735095e-02 -2.07307309e-01 3.79633039e-01
3.76018905e-03 5.84892690e-01 8.89927596e-02 3.26433688e-01
-2.87186444e-01 5.16642809e-01 -6.43674552e-01 -7.65201986e-01
6.19304478e-01 8.71526003e-01 2.60480754e-02 -3.69910628e-01
4.68253255e-01 4.99391615e-01 -3.30436558e-01 8.83286417e-01
6.19500101e-01 6.09431446e-01 -3.55809182e-02 2.82056540e-01
-5.56300044e-01 1.22882701e-01 -1.37376010e+00 1.08044159e+00
-5.07864058e-01 1.02857828e+00 1.91950485e-01 -5.70484281e-01
1.01489496e+00 8.02698851e-01 1.19568072e-01 -4.85741913e-01
4.70157713e-01 5.87488890e-01 1.29291520e-01 -8.37651193e-01
1.49543002e-01 -4.31851327e-01 3.80211025e-01 8.01600277e-01
-5.46432853e-01 -5.30515671e-01 4.46460158e-01 1.65831804e-01
3.47971350e-01 4.40578401e-01 3.34140211e-01 4.44150306e-02
6.75987363e-01 -6.41939649e-03 -1.32504597e-01 1.20407201e-01
-8.06423202e-02 5.51115423e-02 -1.24999760e-02 -1.46851897e-01
-7.85145283e-01 -8.91734064e-01 -1.67456716e-01 4.31954086e-01
-6.95351437e-02 1.53791115e-01 -4.58499759e-01 -1.10958703e-01
-2.00158894e-01 6.20979905e-01 -1.96288284e-02 3.82970184e-01
-8.86893272e-01 9.51862559e-02 4.34495121e-01 3.77125710e-01
8.10370028e-01 -1.55813205e+00 -1.23085415e+00 5.45272648e-01
1.65576980e-01 -6.55330718e-01 -2.67065108e-01 -3.92795622e-01
-9.58541989e-01 -9.64838743e-01 -1.48326683e+00 -1.67528045e+00
1.06395268e+00 -1.23002380e-01 3.10788870e-01 4.30698037e-01
-2.97585338e-01 4.06096578e-01 -7.60738432e-01 -1.05402005e+00
-1.00439703e+00 -5.92649579e-01 -2.42509753e-01 -6.14059925e-01
6.26066566e-01 -6.72315121e-01 -3.90079856e-01 -1.00089490e-01
-8.79510403e-01 4.47080016e-01 4.51539755e-01 5.51940382e-01
-3.27068597e-01 -1.59980536e-01 2.41392434e-01 -5.20324230e-01
1.12420690e+00 4.41085130e-01 -6.99032724e-01 1.95441946e-01
-2.90805727e-01 1.17199853e-01 3.48662853e-01 -3.43955010e-01
-9.08736885e-01 9.36589837e-02 -2.52377361e-01 1.01826048e+00
-3.01846147e-01 3.95489246e-01 4.21224348e-02 -6.04339421e-01
2.66023874e-01 6.78509533e-01 1.29208982e-01 -5.66748857e-01
1.33990526e-01 1.35323834e+00 6.65228963e-01 -2.06291303e-01
9.51980591e-01 3.52705270e-01 1.92427099e-01 -1.22172201e+00
8.38474631e-01 -4.28757548e-01 -7.60527849e-01 -7.38224745e-01
6.14837766e-01 -2.74954319e-01 -8.26943815e-01 7.44012475e-01
-1.30900741e+00 -1.68745428e-01 2.05631554e-01 1.02705407e+00
-3.23905379e-01 3.28778952e-01 -5.26000857e-01 -1.36435139e+00
-4.07789737e-01 -7.70910203e-01 6.82614923e-01 5.75142145e-01
-5.50737977e-01 -9.25609112e-01 -8.77398252e-02 -5.83288900e-04
4.93424237e-01 4.07695502e-01 9.19598818e-01 9.43812132e-02
-1.89427078e-01 -4.44072247e-01 -2.07620963e-01 4.63137962e-02
5.29236019e-01 2.98818141e-01 -5.07309794e-01 1.20942742e-01
-7.53231525e-01 -5.55378497e-02 5.81430733e-01 4.70635533e-01
-2.24286586e-01 -4.90157515e-01 -6.33729771e-02 -6.27088221e-03
1.54724729e+00 1.08521450e+00 8.86137903e-01 3.86568040e-01
1.73045933e-01 5.37923515e-01 2.58761823e-01 2.81695336e-01
4.20303911e-01 2.88146913e-01 -3.47104937e-01 -2.45212793e-01
-4.75834668e-01 -3.50879461e-01 4.00623918e-01 8.88996601e-01
-8.48265648e-01 1.29075438e-01 -1.04719222e+00 4.03555363e-01
-1.51454759e+00 -1.08000958e+00 -6.78243101e-01 2.07759333e+00
8.31438124e-01 -1.80825323e-01 2.57607222e-01 8.46679986e-01
2.81211972e-01 -6.13412619e-01 1.87787965e-01 -8.41992319e-01
8.08656886e-02 7.57696509e-01 6.33684218e-01 1.13989663e+00
-6.24038994e-01 1.02048433e+00 6.04901886e+00 -9.52063352e-02
-1.76155484e+00 -1.16844267e-01 -6.70063913e-01 5.68749070e-01
-1.48354843e-01 1.62768617e-01 -5.36222219e-01 4.32144642e-01
3.53041738e-01 -6.26735806e-01 2.91050345e-01 2.32676864e-01
1.01508152e+00 -6.57278597e-01 -4.35324371e-01 8.50225091e-01
-2.23631710e-01 -1.13963425e+00 2.49158546e-01 6.77096471e-02
5.83106041e-01 -3.71578991e-01 -5.29213727e-01 -1.47045523e-01
5.10242581e-01 -8.93504262e-01 7.46607900e-01 4.61906075e-01
7.01741695e-01 -5.62742591e-01 8.75564635e-01 9.83826444e-02
-1.27322292e+00 4.45789605e-01 1.38534427e-01 -4.99582350e-01
8.25214446e-01 -4.69566971e-01 -1.18302405e+00 -9.36697200e-02
2.23454893e-01 -1.08702160e-01 -2.55278163e-02 1.58922744e+00
-5.60926974e-01 7.75123894e-01 -6.48414433e-01 -7.00536847e-01
3.51139009e-01 -3.28747928e-01 4.23746794e-01 1.59852242e+00
5.20584941e-01 3.09821248e-01 -1.65264234e-01 7.10555613e-02
4.70320761e-01 8.33095670e-01 -7.04167008e-01 -6.17148429e-02
1.10370651e-01 2.78949559e-01 -9.50297713e-01 -5.19588888e-01
-2.70131767e-01 1.22069037e+00 -9.64297414e-01 6.52147472e-01
-2.35810816e-01 -1.08895910e+00 1.63939983e-01 3.28056604e-01
-2.89452851e-01 -6.55167580e-01 -4.50399444e-02 -5.91501296e-01
-1.74697801e-01 -6.37385964e-01 5.86253963e-02 -7.88196504e-01
-1.81818441e-01 7.68250376e-02 2.36616172e-02 -1.50917697e+00
-5.09495199e-01 -8.25821698e-01 -1.01349378e+00 1.15914738e+00
-1.28425372e+00 -1.67015684e+00 -2.36569434e-01 6.21837318e-01
5.41764200e-01 -2.32340813e-01 1.04162586e+00 4.02756095e-01
4.13867533e-01 2.60012895e-01 1.09342821e-01 5.83199918e-01
4.75135744e-01 -1.08663750e+00 -4.20374535e-02 7.60430276e-01
1.05222780e-02 5.42497516e-01 8.79960835e-01 -7.85676837e-01
-9.17245984e-01 -1.15297578e-01 1.40580976e+00 1.19906470e-01
3.09336752e-01 3.22667569e-01 -3.00398737e-01 3.29829484e-01
4.92183596e-01 -6.73381448e-01 5.59472024e-01 -7.59418666e-01
2.42027670e-01 3.35992634e-01 -1.44598913e+00 9.88013804e-01
7.41975844e-01 -3.58164430e-01 -6.48600101e-01 1.69690773e-01
-6.60535395e-02 -3.54055673e-01 -2.92466193e-01 -3.60438168e-01
1.46035886e+00 -3.97039324e-01 4.50868428e-01 -6.52905524e-01
9.35337618e-02 -5.36073327e-01 4.44641635e-02 -1.13606250e+00
2.92783797e-01 -8.49885464e-01 5.61302960e-01 1.13085830e+00
5.79281807e-01 -1.02496409e+00 8.50757658e-01 7.78219104e-01
2.81352401e-01 -7.33029097e-02 -8.35890889e-01 -4.49912190e-01
3.06492914e-02 -4.59650129e-01 2.98340887e-01 4.89165694e-01
2.72976905e-01 -1.41699761e-01 -4.57818180e-01 -8.49827528e-02
5.36908329e-01 -3.65355611e-01 8.56089354e-01 -1.26184881e+00
2.27525324e-01 -5.60117126e-01 -9.24094856e-01 -6.93883598e-01
-4.55959946e-01 -6.64429009e-01 -1.90847725e-01 -2.56928873e+00
-4.88067418e-01 4.55378518e-02 3.85817468e-01 5.50472200e-01
5.92587292e-01 5.14534354e-01 5.26029944e-01 4.58468348e-02
6.48855984e-01 -1.89335287e-01 1.63128018e+00 4.46985587e-02
-6.79238319e-01 2.46176183e-01 -2.58316487e-01 9.96653736e-01
1.07439137e+00 -3.38451445e-01 -1.85712412e-01 -1.34307563e-01
-1.30025223e-01 -3.13432030e-02 3.33922714e-01 -6.60958469e-01
1.80571735e-01 -5.87070167e-01 2.39949122e-01 -6.53751969e-01
2.36671399e-02 -1.26248276e+00 1.14018790e-01 1.23908055e+00
7.67795369e-02 1.10443039e-02 3.01913265e-02 -2.14258417e-01
-1.30396158e-01 -4.42983717e-01 7.57419527e-01 -1.51455149e-01
-9.32831287e-01 -5.71987331e-01 -1.13381112e+00 -5.12246907e-01
8.99571002e-01 -1.05526376e+00 -1.70782328e-01 -1.18645298e+00
-6.40969157e-01 1.65452659e-01 2.78970420e-01 1.18594840e-01
8.30776274e-01 -1.33997715e+00 -9.62750852e-01 3.64558458e-01
-1.25309229e-01 -4.06797498e-01 -3.21476251e-01 7.48425722e-01
-1.76156271e+00 3.48755151e-01 -6.83788359e-01 2.58844197e-02
-2.23005605e+00 -7.17844725e-01 2.19061598e-02 3.97817701e-01
-9.47548807e-01 5.98138988e-01 -8.66080701e-01 -8.66656527e-02
2.94633478e-01 -5.56414425e-01 -4.82262343e-01 -2.50369251e-01
7.24660635e-01 4.34501290e-01 -6.07663214e-01 -1.06504762e+00
-4.37930644e-01 1.33998513e+00 4.93482769e-01 -1.06452453e+00
9.28993344e-01 8.56042188e-03 1.38704041e-02 2.78395742e-01
8.52798462e-01 4.91465688e-01 -3.25336903e-01 1.30255580e-01
5.56191541e-02 -6.18356645e-01 -1.50190368e-01 -1.28306389e+00
-3.69768649e-01 8.02035153e-01 1.02278233e+00 -2.97453225e-01
1.05915082e+00 -3.18110108e-01 7.98505902e-01 5.37417591e-01
5.77341557e-01 -1.46898282e+00 -8.02326262e-01 4.11338389e-01
1.20180357e+00 -1.02051449e+00 -4.37514763e-03 2.97137629e-02
-6.91989720e-01 1.62509644e+00 9.94030535e-02 2.57279038e-01
9.77106214e-01 7.12876081e-01 1.10334599e+00 1.48412362e-01
-7.95710310e-02 -4.03245777e-01 1.50277153e-01 1.24249184e+00
9.91142750e-01 2.41078779e-01 -1.62011647e+00 -1.39743760e-01
-6.54061496e-01 7.31399119e-01 8.57844889e-01 1.76549459e+00
-8.81520748e-01 -1.47047257e+00 -1.10948265e+00 3.15977365e-01
-3.86007488e-01 8.56534094e-02 -6.75572276e-01 9.35777426e-01
1.28904507e-01 1.01562572e+00 -2.99358726e-01 -1.99775957e-02
3.56388181e-01 3.86695653e-01 5.66606343e-01 5.40740862e-02
-2.35200062e-01 1.08342871e-01 3.64080936e-01 4.05319780e-01
-5.63061893e-01 -2.23143905e-01 -2.05536699e+00 -4.68392462e-01
8.02388415e-02 1.31967798e-01 1.28420150e+00 9.14866507e-01
-5.64072788e-01 1.11883536e-01 -3.84183563e-02 -5.88498890e-01
2.35712398e-02 -1.15349650e+00 -5.78073800e-01 2.20055625e-01
2.25031078e-01 -1.99998260e-01 1.20022893e-01 5.24462223e-01] | [9.077159881591797, -6.383749961853027] |
5e17d329-da5d-4ffb-b097-6ca04dea2272 | robust-high-resolution-video-matting-with | 2108.11515 | null | https://arxiv.org/abs/2108.11515v1 | https://arxiv.org/pdf/2108.11515v1.pdf | Robust High-Resolution Video Matting with Temporal Guidance | We introduce a robust, real-time, high-resolution human video matting method that achieves new state-of-the-art performance. Our method is much lighter than previous approaches and can process 4K at 76 FPS and HD at 104 FPS on an Nvidia GTX 1080Ti GPU. Unlike most existing methods that perform video matting frame-by-frame as independent images, our method uses a recurrent architecture to exploit temporal information in videos and achieves significant improvements in temporal coherence and matting quality. Furthermore, we propose a novel training strategy that enforces our network on both matting and segmentation objectives. This significantly improves our model's robustness. Our method does not require any auxiliary inputs such as a trimap or a pre-captured background image, so it can be widely applied to existing human matting applications. | ['Soumyadip Sengupta', 'Imran Saleemi', 'Linjie Yang', 'Shanchuan Lin'] | 2021-08-25 | null | null | null | null | ['image-matting', 'video-matting'] | ['computer-vision', 'computer-vision'] | [ 1.95514947e-01 -3.16348642e-01 -1.33468330e-01 -2.55034983e-01
-5.81772864e-01 -1.95361644e-01 3.02502185e-01 -4.00085807e-01
-4.87773836e-01 5.03365576e-01 -3.23542207e-02 -3.53513092e-01
5.05557060e-01 -6.83914840e-01 -9.26153600e-01 -5.44069529e-01
2.43875206e-01 5.09209156e-01 5.41618049e-01 -5.52438572e-02
-1.30685702e-01 9.09694061e-02 -1.22836483e+00 4.37333822e-01
9.17479575e-01 8.60137761e-01 2.29557201e-01 1.06183648e+00
2.49121666e-01 1.18155921e+00 -5.38441241e-01 -2.74966210e-01
5.04461467e-01 -4.65102434e-01 -8.82826746e-01 5.14566660e-01
8.32104206e-01 -7.87790060e-01 -8.66295099e-01 6.92976534e-01
3.02449763e-01 1.57047451e-01 5.40723167e-02 -9.31962073e-01
-5.19075453e-01 3.05489868e-01 -9.57546413e-01 4.30571169e-01
2.49112174e-01 5.65888941e-01 4.82204556e-01 -7.03409791e-01
4.34946805e-01 1.21091878e+00 5.50031900e-01 5.76796591e-01
-1.24846709e+00 -5.21773636e-01 3.19988489e-01 2.10198417e-01
-1.04898751e+00 -2.99002707e-01 4.17689115e-01 -1.98691368e-01
7.95361042e-01 2.49891356e-01 1.14752305e+00 1.04669189e+00
2.81051576e-01 1.11337650e+00 1.15411592e+00 -2.66754776e-01
-2.00773831e-02 -6.14786923e-01 -9.66776046e-04 8.26415837e-01
7.71839917e-02 -1.46279648e-01 -8.17338586e-01 -4.23647510e-03
1.44050336e+00 1.31811291e-01 -2.60653883e-01 -7.25516081e-02
-1.57929564e+00 4.16799784e-01 2.73314178e-01 -1.62796192e-02
-3.66566986e-01 7.32384920e-01 3.86943400e-01 1.86245620e-01
7.05966175e-01 -6.33726195e-02 -3.19196284e-03 -4.38886821e-01
-1.55817986e+00 1.66320771e-01 4.35973138e-01 9.87509072e-01
5.28581321e-01 4.54030871e-01 -1.74183711e-01 4.06506836e-01
1.35123417e-01 7.46424139e-01 3.84239793e-01 -1.39180708e+00
5.60707152e-01 2.52100945e-01 1.15069926e-01 -9.41773117e-01
1.40568689e-02 -5.28212041e-02 -9.12247896e-01 1.74140409e-01
3.48042101e-01 -6.67261779e-02 -1.39772928e+00 1.25695872e+00
4.10586923e-01 7.80444801e-01 -2.13419512e-01 1.24160600e+00
6.88923597e-01 8.75057399e-01 -4.63122576e-02 -1.03505716e-01
1.08103943e+00 -1.74037611e+00 -8.24197173e-01 -4.61927652e-01
1.34848088e-01 -9.55155790e-01 1.11523414e+00 4.17559981e-01
-1.57683098e+00 -6.13443255e-01 -9.66660798e-01 -2.98545033e-01
2.85899132e-01 1.48317918e-01 8.70876133e-01 4.65160221e-01
-1.35403943e+00 6.63490593e-01 -1.47578275e+00 -8.10156688e-02
4.32359099e-01 4.26404923e-01 -1.05952494e-01 -9.95564014e-02
-7.42922366e-01 6.30868793e-01 2.72698104e-01 3.12794298e-01
-1.16081429e+00 -5.99933505e-01 -8.15602303e-01 -2.42113560e-01
5.19782186e-01 -1.01572680e+00 1.46609473e+00 -1.29107690e+00
-1.94298232e+00 8.56075466e-01 -5.03106534e-01 -6.41221762e-01
8.02345335e-01 -8.24403644e-01 -1.36597201e-01 4.50000286e-01
-7.88989738e-02 6.57188535e-01 1.28400826e+00 -9.98642623e-01
-4.44840014e-01 -1.67435762e-02 -1.12454772e-01 3.94783020e-01
-2.90148348e-01 2.89272610e-02 -1.05071497e+00 -9.52578187e-01
5.93731068e-02 -1.06041420e+00 -4.06555951e-01 2.63753366e-02
-4.35664386e-01 4.99237150e-01 1.08925378e+00 -8.13255429e-01
1.13691187e+00 -1.91455138e+00 4.60257053e-01 -1.27483355e-02
5.05801916e-01 6.11978769e-01 -9.65225026e-02 1.08549118e-01
1.64117187e-01 -1.48360848e-01 -1.55424893e-01 -6.72844112e-01
-3.73739809e-01 2.76097775e-01 -3.58314782e-01 5.04298270e-01
-1.06982678e-01 1.12785470e+00 -7.48564124e-01 -5.05788088e-01
5.31086683e-01 6.28523946e-01 -4.22151685e-01 3.70771229e-01
-1.58056557e-01 5.80663383e-01 -1.92440689e-01 5.71174383e-01
6.87793195e-01 -4.01938975e-01 1.74007863e-01 -9.92149413e-02
5.61058894e-02 2.17807367e-01 -8.83987427e-01 2.02442980e+00
-2.68963605e-01 7.80617595e-01 2.00428754e-01 -6.88245237e-01
5.46664894e-01 2.86939174e-01 5.39732337e-01 -5.02154768e-01
1.75945789e-01 -4.18375246e-02 -3.88089240e-01 -1.91985026e-01
8.09225380e-01 2.48368531e-01 3.59376848e-01 6.69181049e-01
-2.15453282e-01 -2.19317358e-02 1.63198128e-01 4.81040686e-01
1.12870383e+00 2.87313819e-01 -1.00534633e-01 -1.43224031e-01
1.19019710e-01 6.03783354e-02 6.32388413e-01 6.23003960e-01
-1.20577298e-01 9.02512252e-01 1.82473838e-01 -6.93293810e-01
-1.33791614e+00 -1.21432471e+00 3.79887730e-01 1.03261578e+00
4.73258644e-01 -5.73839843e-01 -9.41233456e-01 -3.83547992e-01
-3.31984937e-01 1.43689945e-01 -5.22347093e-01 1.38196364e-01
-1.11369383e+00 -6.50615990e-01 5.66839755e-01 8.95619214e-01
9.83088374e-01 -9.52736914e-01 -7.76356697e-01 2.45963439e-01
-4.00895596e-01 -1.40912998e+00 -9.30296063e-01 -2.21963719e-01
-1.28940141e+00 -9.07693744e-01 -8.32335353e-01 -7.40068436e-01
6.63069665e-01 7.22326159e-01 1.43687415e+00 3.04046810e-01
-3.29047263e-01 4.09570217e-01 -1.81066528e-01 7.47325048e-02
-1.23403609e-01 -7.57542253e-02 -2.35061105e-02 -1.04436390e-01
6.24229200e-02 -3.92244607e-01 -7.24559247e-01 3.69877547e-01
-9.58360374e-01 6.64315104e-01 3.45407128e-01 7.89981186e-01
8.07697833e-01 -8.75702575e-02 -8.77544358e-02 -7.90516675e-01
2.15695396e-01 -1.54249426e-02 -5.62384427e-01 7.14179799e-02
-2.60627151e-01 -3.64366800e-01 4.44871485e-01 -5.73947549e-01
-1.07451403e+00 2.21334562e-01 1.14601925e-01 -9.95655775e-01
1.16121180e-01 1.32670760e-01 1.55421764e-01 -1.86457917e-01
2.97170341e-01 1.96721047e-01 2.22171560e-01 -2.35734254e-01
3.20147127e-01 1.51065379e-01 9.89388525e-01 -4.96521324e-01
1.04363358e+00 7.86027431e-01 -2.22043172e-01 -6.80205882e-01
-7.75236785e-01 -4.40990120e-01 -7.86401153e-01 -2.12952197e-01
1.03055394e+00 -1.24289250e+00 -8.57398748e-01 8.72226417e-01
-1.03890717e+00 -9.97605085e-01 -5.63889295e-02 3.15673590e-01
-6.28180444e-01 6.70050979e-01 -1.40136540e+00 -5.24626553e-01
-7.57305801e-01 -1.19506586e+00 1.21648550e+00 2.71729678e-01
-9.35054868e-02 -9.88045931e-01 -1.69432625e-01 7.26594567e-01
4.95410532e-01 2.16640830e-01 1.73273653e-01 2.92989641e-01
-1.06512356e+00 6.51917309e-02 -2.29856297e-01 1.95714325e-01
2.14978799e-01 2.15675443e-01 -6.14582241e-01 -5.54504812e-01
3.25297117e-02 -3.31944615e-01 1.10954094e+00 5.82514048e-01
1.22879803e+00 -4.62721050e-01 -2.80391663e-01 8.45116854e-01
1.21247554e+00 -5.36176749e-02 9.88891423e-01 3.99673462e-01
1.32469809e+00 3.36180404e-02 7.56686449e-01 3.54841262e-01
3.90235484e-01 7.81640410e-01 3.00499469e-01 -6.02791131e-01
-1.83096871e-01 -3.95385846e-02 5.03581464e-01 8.34624052e-01
-4.22097504e-01 -2.58810252e-01 -7.19879150e-01 3.52032214e-01
-2.36485362e+00 -1.14961922e+00 -2.03624800e-01 2.12091923e+00
8.07601750e-01 6.19337782e-02 4.54396367e-01 -2.87160315e-02
7.35891640e-01 3.74952048e-01 -6.15284681e-01 -1.77747324e-01
-1.93212349e-02 4.08133864e-01 4.74914104e-01 5.24170518e-01
-1.26330590e+00 1.45950937e+00 7.43829870e+00 7.23766983e-01
-1.18851459e+00 5.88941984e-02 8.87666047e-01 -5.16980767e-01
-6.34437278e-02 -2.48805791e-01 -5.57800591e-01 4.59894836e-01
8.58345151e-01 9.56207588e-02 6.74789310e-01 7.68553793e-01
4.91960615e-01 -2.47159109e-01 -9.63604271e-01 1.17966115e+00
2.36096516e-01 -1.54587066e+00 5.33300564e-02 2.25303858e-03
6.77198946e-01 -1.85741007e-01 2.87910312e-01 -6.93738982e-02
4.48060721e-01 -1.38814211e+00 9.01502430e-01 2.57379204e-01
7.47659922e-01 -7.62427688e-01 4.99799639e-01 3.58817652e-02
-1.34187257e+00 4.69905913e-01 -4.00220335e-01 -2.11873412e-01
6.26546681e-01 5.67469418e-01 -5.47473073e-01 4.23273653e-01
8.07261169e-01 9.59244430e-01 -5.32754242e-01 7.09055007e-01
-2.21760854e-01 8.63934040e-01 -3.25913161e-01 5.46916246e-01
3.93425763e-01 -2.13635236e-01 2.92044908e-01 1.18723953e+00
7.64939934e-02 2.64122993e-01 5.85986555e-01 5.84351182e-01
4.06868495e-02 -2.46050134e-01 -2.56826818e-01 -2.60536168e-02
2.54085749e-01 1.18918467e+00 -1.04043818e+00 -7.67574430e-01
-3.35656017e-01 1.52876019e+00 1.37616932e-01 3.85060519e-01
-1.37669778e+00 1.76427856e-01 6.12278759e-01 2.29860425e-01
4.72289592e-01 -7.26209164e-01 -4.70685959e-01 -1.58970034e+00
1.08482227e-01 -1.14194822e+00 2.73874491e-01 -7.15715766e-01
-9.69526649e-01 7.22453654e-01 -1.75700188e-01 -1.15384114e+00
-1.74397141e-01 -3.41932029e-01 -7.21886396e-01 5.55006683e-01
-1.27002549e+00 -1.32118654e+00 -7.90562034e-01 8.68588686e-01
7.49312818e-01 9.41538066e-02 3.85426164e-01 3.12350750e-01
-6.51094854e-01 4.80825424e-01 -1.95888266e-01 2.36233473e-01
7.78148592e-01 -1.13723743e+00 9.22642648e-01 1.31086707e+00
1.61641389e-01 5.14881849e-01 7.27588296e-01 -9.17826653e-01
-1.53569818e+00 -1.29968047e+00 2.26326257e-01 -3.52207810e-01
4.22943622e-01 -2.33119711e-01 -9.54323053e-01 1.06799817e+00
6.55155301e-01 1.80347577e-01 3.51166427e-01 -2.06292465e-01
-2.85840839e-01 -2.08158046e-03 -8.08970094e-01 7.63913333e-01
1.12180519e+00 -2.89701730e-01 -2.72850454e-01 4.27052617e-01
7.64342606e-01 -1.15439010e+00 -6.35642111e-01 3.17854702e-01
4.19464737e-01 -9.99728620e-01 1.04405308e+00 -3.64745140e-01
5.70544899e-01 -4.53830838e-01 1.72731027e-01 -1.00208855e+00
-4.43131387e-01 -9.75554883e-01 -4.73121017e-01 7.73411512e-01
-1.06227957e-02 -3.85895222e-01 1.11801541e+00 6.44319415e-01
-2.70808011e-01 -8.70421886e-01 -7.11377740e-01 -7.53158033e-01
-3.13083053e-01 -2.42730767e-01 2.75069028e-01 7.45543242e-01
-2.58780956e-01 6.70541525e-02 -1.08033597e+00 1.67681709e-01
8.19310844e-01 2.61718422e-01 9.81772006e-01 -4.96764779e-01
-5.79949856e-01 -5.21611869e-02 -3.21853399e-01 -1.42234755e+00
-1.19439382e-02 -3.63283575e-01 1.45999789e-01 -1.36734867e+00
3.06498110e-01 -1.69420213e-01 1.94101453e-01 6.66757405e-01
-3.79718810e-01 8.04157257e-01 2.96018481e-01 4.25286233e-01
-8.78546178e-01 5.31203389e-01 1.41305006e+00 -1.41630545e-01
-1.73697025e-01 -2.57413805e-01 -1.60734549e-01 7.96438515e-01
6.96099937e-01 -1.12790100e-01 -2.91606128e-01 -8.79232943e-01
-3.07555348e-01 1.05288893e-01 4.23008382e-01 -1.15429056e+00
2.43267715e-01 -3.64858717e-01 6.06738389e-01 -5.93082786e-01
7.32253909e-01 -3.72927696e-01 3.38023841e-01 5.63205421e-01
-2.63131615e-02 5.22935212e-01 3.33642989e-01 5.25034547e-01
-1.53606296e-01 1.51283965e-01 9.87486720e-01 -3.25095981e-01
-6.74478292e-01 5.42530060e-01 -5.50519645e-01 1.39590465e-02
1.19930458e+00 -1.94342032e-01 -3.62594247e-01 -5.09392977e-01
-5.44748545e-01 2.01195851e-01 9.28142011e-01 2.47516513e-01
7.54524648e-01 -1.31188607e+00 -6.01551414e-01 -4.96205688e-03
-6.28077149e-01 2.30706260e-01 3.52769375e-01 7.58718789e-01
-1.15695131e+00 1.28982812e-01 -1.91176623e-01 -9.39018548e-01
-1.56780243e+00 4.42880213e-01 1.61167264e-01 -2.50280470e-01
-1.18230319e+00 7.75654078e-01 2.38512933e-01 2.65323073e-01
1.26593724e-01 -1.30322397e-01 3.12478572e-01 -4.92250204e-01
7.89521694e-01 3.91140282e-01 -2.09437191e-01 -5.85806668e-01
-1.86196283e-01 4.89957809e-01 -3.98113459e-01 -2.11210310e-01
1.18251789e+00 -2.23405644e-01 -1.37598634e-01 3.76526117e-01
5.87994754e-01 -2.01592267e-01 -1.58640325e+00 -6.19038083e-02
-5.37276566e-01 -9.10446048e-01 -3.93083170e-02 -2.79612154e-01
-1.52323306e+00 6.46331429e-01 3.34084183e-01 -2.00343162e-01
1.15419614e+00 -3.06378812e-01 1.58202326e+00 3.79850827e-02
4.13379461e-01 -1.11362016e+00 7.24344552e-01 4.27142084e-01
4.58769679e-01 -1.13814449e+00 2.21016899e-01 -7.15109825e-01
-7.54079521e-01 1.15034389e+00 9.63339150e-01 -4.30685014e-01
3.04587390e-02 5.13321757e-01 3.71363968e-01 -7.88927451e-02
-6.87930882e-01 7.77993873e-02 2.91405529e-01 3.71209532e-01
4.89532918e-01 -1.15798758e-02 -9.54404995e-02 -1.91371471e-01
-2.07051784e-01 2.34574564e-02 6.77171588e-01 1.06890547e+00
-3.40816289e-01 -1.02998149e+00 -3.35294157e-01 3.40553284e-01
-5.15845478e-01 -2.45540455e-01 -1.94732565e-02 5.10138094e-01
-3.26476455e-01 7.77644157e-01 1.56236023e-01 -3.97677034e-01
-1.08487651e-01 -2.58249462e-01 6.28740549e-01 -5.16141176e-01
-6.92695141e-01 3.66320580e-01 -1.07235499e-01 -9.87161338e-01
-5.34821272e-01 -4.48399752e-01 -1.15354192e+00 -1.03380859e+00
-1.47464216e-01 -1.53668121e-01 2.00564027e-01 8.80471110e-01
5.50392866e-01 6.54398561e-01 3.12631458e-01 -1.28620005e+00
1.79880902e-01 -6.48295760e-01 -2.81610399e-01 3.72481316e-01
2.86873102e-01 -3.03734720e-01 1.02730412e-02 4.55953032e-01] | [10.628860473632812, -0.888198733329773] |
2a738d5c-9724-4531-8695-f987039b3158 | mcmlsd-a-dynamic-programming-approach-to-line | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Almazan_MCMLSD_A_Dynamic_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Almazan_MCMLSD_A_Dynamic_CVPR_2017_paper.pdf | MCMLSD: A Dynamic Programming Approach to Line Segment Detection | Prior approaches to line segment detection typically involve perceptual grouping in the image domain or global accumulation in the Hough domain. Here we propose a probabilistic algorithm that merges the advantages of both approaches. In a first stage lines are detected using a global probabilistic Hough approach. In the second stage each detected line is analyzed in the image domain to localize the line segments that generated the peak in the Hough map. By limiting search to a line, the distribution of segments over the sequence of points on the line can be modeled as a Markov chain, and a probabilistically optimal labelling can be computed exactly using a standard dynamic programming algorithm, in linear time. The Markov assumption also leads to an intuitive ranking method that uses the local marginal posterior probabilities to estimate the expected number of correctly labelled points on a segment. To assess the resulting Markov Chain Marginal Line Segment Detector (MCMLSD) we develop and apply a novel quantitative evaluation methodology that controls for under- and over-segmentation. Evaluation on the YorkUrbanDB dataset shows that the proposed MCMLSD method outperforms the state-of-the-art by a substantial margin.
| ['Yiming Qian', 'Ron Tal', 'James H. Elder', 'Emilio J. Almazan'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['line-segment-detection'] | ['computer-vision'] | [ 2.45141387e-01 2.33164430e-01 -2.23082200e-01 -2.20149964e-01
-1.02978969e+00 -6.02109373e-01 5.78231394e-01 6.47069931e-01
-5.48746884e-01 4.30116594e-01 -3.17132205e-01 -1.68242618e-01
2.09400188e-02 -6.74813330e-01 -7.07641423e-01 -7.08541989e-01
6.46223128e-03 9.40950990e-01 1.09745777e+00 3.11391622e-01
5.08713901e-01 7.68704295e-01 -1.32784152e+00 1.91185046e-02
7.48103142e-01 8.59770954e-01 4.69397977e-02 8.53321075e-01
-2.94007450e-01 5.12251318e-01 -5.83944619e-01 -1.57745913e-01
2.49323338e-01 -6.04587197e-01 -7.45804787e-01 3.45016539e-01
3.80180836e-01 -1.68181539e-01 1.33962527e-01 1.03346813e+00
3.55993956e-01 7.26909265e-02 9.97124374e-01 -1.16851890e+00
3.83207649e-01 5.71874619e-01 -1.09993804e+00 1.74044333e-02
5.28191388e-01 -1.61060721e-01 1.11993134e+00 -8.32803905e-01
7.07525134e-01 1.06648684e+00 6.70003355e-01 -2.10078910e-01
-1.52966607e+00 -1.64670631e-01 -1.36939198e-01 -3.68566811e-02
-1.66367316e+00 1.39104202e-02 5.99594116e-01 -7.19469249e-01
4.72276300e-01 1.01482935e-01 7.16061234e-01 6.49969131e-02
9.01727304e-02 9.89891589e-01 1.16374421e+00 -9.87621009e-01
3.95083755e-01 7.48449489e-02 2.56749541e-01 8.69880795e-01
4.02171612e-02 1.15853228e-01 -4.40097332e-01 -7.96094686e-02
7.63689876e-01 -5.46606064e-01 3.77868824e-02 -9.36955154e-01
-1.14207447e+00 7.28221595e-01 3.31024468e-01 1.66998595e-01
-4.70758379e-01 5.94447516e-02 2.27064729e-01 -3.53439301e-01
2.19893724e-01 1.08120777e-01 8.40942860e-02 5.99754713e-02
-1.68173003e+00 4.05516565e-01 9.50552821e-01 7.98075259e-01
9.39650595e-01 -5.66672504e-01 -3.71731430e-01 5.97424924e-01
6.55700803e-01 2.93646067e-01 -3.58900102e-03 -8.78981292e-01
1.61943361e-01 5.45107007e-01 2.21659452e-01 -8.42758715e-01
-3.51252377e-01 -3.43252599e-01 -2.37777546e-01 7.21079290e-01
9.59840894e-01 -1.61645319e-02 -9.54643309e-01 1.16840529e+00
5.41132569e-01 -8.58358964e-02 -3.23747158e-01 5.80884159e-01
4.39336076e-02 6.58526421e-01 4.67502326e-03 -2.26151198e-01
1.51767290e+00 -8.58370841e-01 -4.38085347e-01 -8.91071465e-03
4.40599591e-01 -1.06990790e+00 7.29367256e-01 5.02519906e-01
-1.05518031e+00 -4.70911652e-01 -1.25186694e+00 1.20742552e-01
-9.78457406e-02 2.54802734e-01 1.68865725e-01 7.73708165e-01
-8.42904389e-01 5.17596841e-01 -9.90640223e-01 -4.09808904e-01
2.79923201e-01 8.28176141e-02 6.03923872e-02 6.64760694e-02
-5.79556465e-01 7.54606187e-01 9.00683761e-01 -2.23307102e-03
-6.51510656e-01 -2.62969196e-01 -7.61065662e-01 -2.11434774e-02
3.52745861e-01 -4.54937458e-01 1.19075036e+00 -7.25477517e-01
-1.35378003e+00 1.03066325e+00 -2.40203455e-01 -7.53858447e-01
9.79617894e-01 -3.03477049e-01 7.57242367e-02 5.21474242e-01
2.76180178e-01 9.22010481e-01 7.39897609e-01 -1.56629205e+00
-1.16241860e+00 -1.17010055e-02 -3.31506670e-01 2.69253969e-01
5.63513100e-01 5.55415563e-02 -9.96546268e-01 -3.19897324e-01
4.77496088e-01 -1.13873267e+00 -2.29202241e-01 -5.36409877e-02
-7.40915954e-01 -3.65090132e-01 7.34077871e-01 -5.51298916e-01
1.23547554e+00 -2.00139999e+00 -2.92689383e-01 7.59842455e-01
5.06708659e-02 -1.00900792e-01 4.78644788e-01 4.34278697e-01
8.55772495e-02 -1.48976341e-01 -4.85579550e-01 -2.82908827e-01
-9.57112163e-02 -7.79088214e-02 -1.01631820e-01 7.90248573e-01
-9.92264599e-02 3.96061033e-01 -8.10320735e-01 -1.06668794e+00
5.17597795e-01 1.88539162e-01 -2.27980778e-01 1.04622863e-01
-2.33743459e-01 6.11333847e-02 -2.53898412e-01 3.70878905e-01
8.63042951e-01 2.79531758e-02 -7.80769587e-02 -2.11313277e-01
-6.09655380e-01 -1.63357764e-01 -1.61092436e+00 1.46712601e+00
3.18164863e-02 8.51732373e-01 -5.50238788e-01 -6.38397336e-01
1.08781517e+00 8.79740939e-02 5.31423092e-01 -1.49477333e-01
-5.65629527e-02 2.35534325e-01 -1.43155023e-01 -6.90581724e-02
7.10922658e-01 3.80074158e-02 -2.09500417e-01 3.48021656e-01
-3.62713672e-02 -3.47570956e-01 5.41363597e-01 2.63493150e-01
8.22235703e-01 5.72455406e-01 6.29966855e-01 -2.30379596e-01
4.33007658e-01 4.69618261e-01 2.67182589e-01 1.14745867e+00
-3.54797363e-01 7.54908979e-01 7.78573513e-01 2.18473617e-02
-1.10420740e+00 -1.23711944e+00 -1.68548331e-01 8.53350222e-01
4.48649824e-01 -1.03875175e-01 -1.14565492e+00 -5.97471178e-01
-2.62567669e-01 1.06084561e+00 -4.85136449e-01 3.26501817e-01
-4.80757713e-01 -5.95254123e-01 4.17419821e-01 2.29982480e-01
5.19475222e-01 -9.09288406e-01 -1.08092260e+00 3.45489055e-01
-3.99511233e-02 -9.76281583e-01 -3.90992910e-01 3.51074457e-01
-7.36338079e-01 -9.13342118e-01 -9.01929855e-01 -7.94571579e-01
6.74331784e-01 8.62145647e-02 9.69762504e-01 -2.75355816e-01
-2.82691568e-01 1.20020308e-01 -3.77602637e-01 -1.56933188e-01
-6.21457756e-01 -1.83729343e-02 -6.06380939e-01 -5.62580563e-02
2.75568634e-01 1.39021069e-01 -6.09414279e-01 5.09651184e-01
-4.90728408e-01 6.00799657e-02 7.49729276e-01 5.57326138e-01
9.76140857e-01 3.15059721e-01 -1.84970856e-01 -8.30502093e-01
3.89230251e-01 -2.09683981e-02 -1.00097334e+00 2.87224084e-01
-5.30291021e-01 2.06730992e-01 -4.10901904e-02 -1.65007859e-01
-1.07321370e+00 6.57524586e-01 -4.95644584e-02 -1.65989161e-01
-5.81694901e-01 4.56653744e-01 -4.79601622e-02 7.22131040e-03
7.26071894e-01 7.99605623e-02 -3.09811652e-01 -2.21608475e-01
6.58752024e-01 3.24563712e-01 8.36121440e-01 -3.11042517e-01
7.31865466e-01 6.24997079e-01 2.15889215e-01 -8.69815707e-01
-6.10637069e-01 -9.76748049e-01 -1.01933563e+00 -6.36515081e-01
1.09376812e+00 -4.62659150e-01 -4.13056850e-01 4.97812122e-01
-1.16314638e+00 -2.28055328e-01 -5.20199120e-01 4.92599636e-01
-8.39959145e-01 6.47960722e-01 -4.09540415e-01 -1.06475234e+00
1.22202210e-01 -1.09027064e+00 1.04104841e+00 3.74182075e-01
-3.65598261e-01 -8.07011068e-01 3.04002941e-01 5.35880364e-02
-3.31204087e-01 5.15936732e-01 7.59251952e-01 -6.19911134e-01
-6.63999259e-01 -6.83332801e-01 -3.80479723e-01 1.77569032e-01
-3.93944889e-01 4.82627124e-01 -8.07027936e-01 1.88049763e-01
-3.47569138e-01 1.71283916e-01 7.75826156e-01 8.10765564e-01
6.09271348e-01 1.98670134e-01 -6.69002533e-01 1.38575450e-01
1.66508424e+00 3.07491004e-01 7.31763840e-01 4.44993079e-01
5.41266501e-01 7.52531528e-01 1.07192600e+00 3.72935981e-01
1.31806940e-01 8.52092445e-01 1.64412111e-01 -3.36490095e-01
-2.44046777e-01 -5.40949523e-01 1.87940165e-01 1.91585094e-01
8.98252353e-02 -4.85048801e-01 -1.15978515e+00 5.98572314e-01
-1.83039677e+00 -9.38135266e-01 -6.18204117e-01 2.51246381e+00
4.90339696e-01 7.93614507e-01 5.47909319e-01 5.22405803e-01
1.09323108e+00 -1.93779990e-01 -1.82121858e-01 -3.18180501e-01
5.09023927e-02 -1.66056603e-01 1.07757962e+00 7.54287183e-01
-1.33883262e+00 9.46622074e-01 6.55778933e+00 1.24712181e+00
-6.27141893e-01 -1.96592018e-01 5.40273726e-01 6.91779077e-01
1.51727647e-01 5.01507878e-01 -1.00661647e+00 3.08194607e-01
5.75848699e-01 4.64641340e-02 -1.59651592e-01 8.05082321e-01
3.89912069e-01 -9.96105373e-01 -9.81417358e-01 6.81192815e-01
9.88441426e-03 -9.87263262e-01 -2.49523744e-01 2.01812372e-01
6.85437977e-01 -7.09510893e-02 -1.87481135e-01 -1.96540371e-01
4.33016002e-01 -5.31293690e-01 1.14585912e+00 7.26566613e-01
3.78626823e-01 -8.38766396e-01 6.40718281e-01 5.00469387e-01
-1.21821344e+00 3.58968049e-01 -1.73754022e-02 3.98377448e-01
6.08675301e-01 7.61581719e-01 -1.58045304e+00 4.78601724e-01
3.97245079e-01 1.28544778e-01 -6.61513925e-01 1.74589658e+00
-5.28782785e-01 8.56735647e-01 -6.70609653e-01 5.31145968e-02
3.18293869e-01 -3.53085786e-01 6.65413499e-01 1.53146064e+00
1.55673966e-01 -4.07855511e-01 3.80543947e-01 7.72149503e-01
4.70883042e-01 2.60122985e-01 -1.55276805e-01 3.37996811e-01
4.52829897e-01 1.17317116e+00 -1.73082340e+00 -3.81367892e-01
-1.94305807e-01 1.07504010e+00 3.07938419e-02 2.11815834e-01
-6.64906979e-01 -5.48166573e-01 -4.38136905e-01 2.63578385e-01
4.14977878e-01 -3.50660861e-01 -4.20147955e-01 -3.22306782e-01
-2.76855946e-01 -2.94059753e-01 3.90183717e-01 -7.12830186e-01
-7.55995333e-01 4.47365820e-01 3.60667020e-01 -1.34657586e+00
-5.48325658e-01 -2.75098473e-01 -5.98722100e-01 8.14751446e-01
-1.19074178e+00 -1.01861405e+00 -1.99740440e-01 1.43460721e-01
5.74365318e-01 4.08224642e-01 3.88609082e-01 -6.80475160e-02
-3.73627514e-01 3.54182243e-01 8.08004588e-02 1.48305863e-01
4.74101156e-01 -1.74309540e+00 3.29865783e-01 1.22732568e+00
2.59470254e-01 1.99813768e-01 1.16442513e+00 -7.05273211e-01
-1.72131777e-01 -6.13428056e-01 8.91595840e-01 -1.67481661e-01
4.31330681e-01 -2.23542422e-01 -8.26002657e-01 3.63543034e-01
1.00752495e-01 -3.16307962e-01 4.46175516e-01 -2.28648975e-01
1.29749805e-01 2.64608592e-01 -1.12056589e+00 4.78217155e-01
3.31745952e-01 -1.73940167e-01 -5.81933320e-01 1.43763989e-01
1.89441472e-01 -3.54646623e-01 -6.99134946e-01 1.73542425e-01
6.79811597e-01 -1.37297928e+00 7.34321535e-01 4.03446615e-01
5.97446561e-02 -8.70495379e-01 2.45242015e-01 -9.87083852e-01
-1.21830329e-01 -5.68167746e-01 2.93084890e-01 1.34226382e+00
4.48205262e-01 -2.62346894e-01 9.45743918e-01 2.48552948e-01
-7.90494122e-03 -5.05822361e-01 -9.70933259e-01 -5.94066083e-01
-3.80492151e-01 -5.06706595e-01 1.12978570e-01 2.69486845e-01
-1.78361490e-01 8.39902684e-02 -9.01008695e-02 3.01639229e-01
8.96801114e-01 1.09801598e-01 8.41978192e-01 -1.09778488e+00
-3.73424053e-01 -6.31350100e-01 -6.61599815e-01 -1.40959191e+00
-2.81233847e-01 -5.26546717e-01 6.81602597e-01 -1.71164155e+00
5.27222753e-02 -4.94165212e-01 9.30953026e-02 5.23532294e-02
-1.37488946e-01 9.08003300e-02 1.62756175e-01 4.37171608e-01
-6.09626889e-01 -7.10384697e-02 7.96652913e-01 2.34056368e-01
-3.05845052e-01 3.18178326e-01 8.50696415e-02 1.09003305e+00
4.99010831e-01 -6.65179133e-01 -1.01926148e-01 2.71683395e-01
3.42466012e-02 6.53006509e-02 3.23241919e-01 -1.23271072e+00
5.59859276e-01 2.45394260e-01 3.97465080e-01 -1.34710693e+00
1.54134065e-01 -7.71508515e-01 -9.56661329e-02 4.88927424e-01
-4.22166526e-01 -1.67961225e-01 -1.16956886e-02 7.23269880e-01
-1.52984753e-01 -8.56449127e-01 1.06020701e+00 1.03232972e-01
-7.79864669e-01 -2.06638366e-01 -6.99649155e-01 -2.77024299e-01
1.44177186e+00 -6.88050866e-01 2.62516677e-01 -2.95270443e-01
-7.53269970e-01 1.45593420e-01 6.36700213e-01 -1.92328304e-01
4.20454770e-01 -9.21095192e-01 -5.63628852e-01 1.76719539e-02
1.97513342e-01 5.70139550e-02 4.13397290e-02 8.23404610e-01
-1.15791881e+00 1.22981504e-01 1.53712988e-01 -1.05488050e+00
-1.28411651e+00 4.58899319e-01 3.62232000e-01 -4.29698408e-01
-4.41455334e-01 7.28941381e-01 -2.49367747e-02 9.13487598e-02
1.46163717e-01 -3.21596563e-01 -2.62110800e-01 2.53466874e-01
1.27273500e-02 7.15810239e-01 -1.32942408e-01 -9.50472951e-01
-2.11692974e-01 7.96302140e-01 -6.19669519e-02 -7.49758303e-01
6.41852975e-01 -3.46563637e-01 2.00260788e-01 7.08106339e-01
8.56253922e-01 2.73721457e-01 -1.49952412e+00 2.78109666e-02
4.96916771e-01 -4.60096687e-01 1.42338693e-01 -6.80886686e-01
-4.17900592e-01 6.63368225e-01 8.60082686e-01 5.29053092e-01
8.58176410e-01 1.94844961e-01 3.19014281e-01 -2.32633010e-01
1.95481300e-01 -1.13979232e+00 -1.91049173e-01 7.96751156e-02
4.06200826e-01 -8.69507849e-01 2.42442861e-01 -9.40555632e-01
-7.23581433e-01 1.33495331e+00 3.25675867e-02 -3.51287425e-01
5.47593176e-01 2.90047705e-01 8.20078142e-03 -1.91819772e-01
-2.43005250e-02 -3.89882624e-01 3.54619026e-01 3.92256260e-01
2.82982647e-01 9.48296636e-02 -4.29880351e-01 -2.16727227e-01
-1.43625304e-01 -1.00359321e-04 5.44278383e-01 9.50811028e-01
-8.56553197e-01 -9.96482372e-01 -8.04338872e-01 1.59188434e-01
-2.66162157e-01 1.49461165e-01 -1.41080454e-01 9.17380214e-01
2.98908412e-01 8.82888258e-01 3.39891583e-01 6.97674081e-02
3.53097290e-01 -3.87106165e-02 4.29123074e-01 -4.27948475e-01
-7.54564852e-02 7.50026584e-01 -3.31917256e-02 -2.93178886e-01
-4.78926867e-01 -1.12637043e+00 -1.45169473e+00 3.49910468e-01
-7.91726530e-01 1.52730554e-01 7.91736305e-01 7.80278742e-01
-2.42927700e-01 2.92676300e-01 4.03065115e-01 -1.02575696e+00
-1.07684545e-01 -7.26759434e-01 -7.28944898e-01 1.03913769e-01
6.08979166e-02 -5.27883410e-01 -3.86528760e-01 3.43491733e-01] | [8.29963207244873, -1.3834388256072998] |
993f796a-e517-439c-9e6b-ee6557268a08 | last-layer-state-space-model-for | 2307.01566 | null | https://arxiv.org/abs/2307.01566v1 | https://arxiv.org/pdf/2307.01566v1.pdf | Last layer state space model for representation learning and uncertainty quantification | As sequential neural architectures become deeper and more complex, uncertainty estimation is more and more challenging. Efforts in quantifying uncertainty often rely on specific training procedures, and bear additional computational costs due to the dimensionality of such models. In this paper, we propose to decompose a classification or regression task in two steps: a representation learning stage to learn low-dimensional states, and a state space model for uncertainty estimation. This approach allows to separate representation learning and design of generative models. We demonstrate how predictive distributions can be estimated on top of an existing and trained neural network, by adding a state space-based last layer whose parameters are estimated with Sequential Monte Carlo methods. We apply our proposed methodology to the hourly estimation of Electricity Transformer Oil temperature, a publicly benchmarked dataset. Our model accounts for the noisy data structure, due to unknown or unavailable variables, and is able to provide confidence intervals on predictions. | ['Sylvain Le Corff', 'Maurice Charbit', 'Max Cohen'] | 2023-07-04 | null | null | null | null | ['representation-learning'] | ['methodology'] | [ 1.04021199e-01 1.69491902e-01 -9.27107856e-02 -7.02581584e-01
-1.13801956e+00 -5.63520968e-01 8.74781251e-01 1.62532330e-01
-1.27289146e-01 1.19324756e+00 1.07743531e-01 -5.91616273e-01
-3.12020689e-01 -9.99996722e-01 -8.08342814e-01 -5.79408824e-01
-1.88825816e-01 7.40753531e-01 -3.45849961e-01 5.16042709e-01
-2.29012724e-02 5.44684947e-01 -1.36160290e+00 -1.37044087e-01
8.51991653e-01 1.44886220e+00 -2.04145536e-02 5.32029808e-01
-8.78617167e-02 7.79655457e-01 -7.76689947e-01 -1.22097850e-01
-2.67225429e-02 -2.44869664e-01 -4.61405128e-01 -1.93902820e-01
-1.91673413e-01 -5.33735037e-01 -1.76941887e-01 9.49584126e-01
3.32023025e-01 3.38883549e-01 1.12554085e+00 -1.10866833e+00
-2.77209520e-01 1.11626065e+00 -1.08056992e-01 -8.00138190e-02
-1.36229828e-01 1.37214242e-02 7.00863540e-01 -5.72865129e-01
2.18076855e-02 1.05079055e+00 6.67526245e-01 4.17937875e-01
-1.55689299e+00 -6.07243359e-01 7.02682585e-02 -1.04613721e-01
-1.42780697e+00 -4.13202524e-01 7.80491412e-01 -7.03444064e-01
9.71326649e-01 -9.11831260e-02 3.94348800e-01 1.47456551e+00
2.22422734e-01 5.23085713e-01 9.95201886e-01 -6.05968535e-02
9.81706917e-01 1.67112544e-01 1.47563368e-01 2.16823786e-01
4.75085497e-01 5.68503380e-01 -6.72218725e-02 -3.33450675e-01
6.32611275e-01 1.74081683e-01 -6.45698458e-02 -3.73930007e-01
-4.91417527e-01 9.80131328e-01 2.29925498e-01 2.32600179e-02
-5.35016000e-01 6.42514408e-01 2.06169605e-01 7.50642717e-02
6.72395110e-01 4.81101573e-01 -6.53885841e-01 -3.56352597e-01
-1.34499097e+00 6.87525272e-02 1.06324124e+00 6.69584095e-01
7.31440187e-01 4.70309734e-01 -1.82113126e-01 5.84815919e-01
4.94567961e-01 5.15673399e-01 4.51399714e-01 -7.29018390e-01
4.07188624e-01 2.25661367e-01 3.11476886e-01 -5.36193788e-01
-5.08859813e-01 -7.01270461e-01 -9.98716176e-01 2.05168009e-01
1.64174408e-01 -8.07565093e-01 -1.12262440e+00 1.79151917e+00
6.78721024e-03 3.81863415e-01 2.96445340e-01 4.89694744e-01
3.48348975e-01 8.33738029e-01 -3.33380559e-03 -1.56800315e-01
9.29807425e-01 -4.37403381e-01 -5.86110711e-01 -3.35583925e-01
2.81047076e-01 -1.68785051e-01 4.47258502e-01 4.09089595e-01
-1.02332950e+00 -3.76031250e-01 -1.18510628e+00 4.44784522e-01
-4.03064996e-01 3.67678225e-01 6.42065644e-01 7.42067516e-01
-6.88857734e-01 1.06720483e+00 -1.32815325e+00 1.88877970e-01
5.07208169e-01 2.32188597e-01 1.51444614e-01 2.26873845e-01
-1.39958537e+00 1.12869906e+00 7.94199049e-01 5.21811485e-01
-1.21671510e+00 -6.48967743e-01 -1.17179179e+00 5.75227976e-01
3.80307615e-01 -5.50491929e-01 1.29479635e+00 -5.21479726e-01
-1.86803114e+00 -1.57962129e-01 2.20646933e-01 -8.13284039e-01
4.41296875e-01 -2.40109354e-01 -4.30958211e-01 -3.16992760e-01
-3.96287650e-01 3.80833372e-02 1.16063833e+00 -1.24996448e+00
-2.41732240e-01 -2.57899791e-01 -2.10810393e-01 -3.42090309e-01
4.11023647e-02 -4.98995751e-01 7.86589086e-02 -6.55655742e-01
5.38129061e-02 -7.93472648e-01 -4.58342999e-01 -5.46664059e-01
-3.51257086e-01 -3.14952284e-02 4.55356687e-01 -7.03015387e-01
1.40269208e+00 -1.79627597e+00 3.19450498e-01 7.01862395e-01
-1.54199943e-01 -5.55647835e-02 2.12658271e-01 4.19922441e-01
-6.09468073e-02 1.13578796e-01 -5.77569723e-01 -6.46395504e-01
4.47057277e-01 4.79107678e-01 -6.27709925e-01 2.25122601e-01
5.77218771e-01 7.44829953e-01 -5.99896550e-01 -5.07763810e-02
3.00766289e-01 4.26390827e-01 -2.47872680e-01 2.75611997e-01
-6.77870035e-01 3.83978456e-01 -4.93814707e-01 4.51211989e-01
5.67535996e-01 -4.63811427e-01 3.80883723e-01 -4.28256765e-02
-5.12594320e-02 4.28576827e-01 -1.44782627e+00 1.71798360e+00
-8.00908029e-01 2.93818146e-01 -1.74418956e-01 -1.09978592e+00
8.67574811e-01 3.03881347e-01 1.27655894e-01 -1.16402939e-01
4.77875113e-01 1.53762937e-01 -2.47017846e-01 5.59787266e-02
4.01371807e-01 -8.55312794e-02 -4.88290757e-01 5.83231509e-01
3.03790480e-01 -5.25010467e-01 2.25210041e-01 -6.57345876e-02
8.45374107e-01 2.78891146e-01 2.85514086e-01 -1.09073818e-01
2.50608712e-01 -4.17982370e-01 5.00437260e-01 8.55795383e-01
3.03164154e-01 3.40731174e-01 7.41736948e-01 -4.54616427e-01
-8.33875477e-01 -1.15756083e+00 -4.24988061e-01 4.71323162e-01
-5.39023340e-01 -1.49803326e-01 -5.85186660e-01 -4.71073717e-01
3.66043270e-01 1.19995773e+00 -7.33269513e-01 -4.38316524e-01
-1.17780045e-01 -9.41536009e-01 3.70788008e-01 8.35071325e-01
1.96547341e-02 -6.86692476e-01 -5.44265568e-01 5.22695363e-01
4.54841703e-01 -7.69932449e-01 2.74811804e-01 8.19749832e-01
-9.75673616e-01 -6.54627860e-01 -5.42913795e-01 1.85197473e-01
3.83381099e-01 -9.60685492e-01 1.13346624e+00 -5.89486539e-01
-1.14569314e-01 2.48799950e-01 1.00746103e-01 -6.33108139e-01
-5.73322594e-01 2.00159773e-01 9.86746848e-02 -1.21768162e-01
-1.46104703e-02 -7.98770607e-01 -3.32524389e-01 -5.32649979e-02
-9.33426142e-01 -2.10816860e-01 7.23119497e-01 9.68746483e-01
4.94709820e-01 2.43152037e-01 7.46255517e-01 -6.66230679e-01
7.20771849e-01 -8.38286340e-01 -1.28046167e+00 3.01620871e-01
-9.98200595e-01 7.21298933e-01 5.59742630e-01 -1.30455464e-01
-1.27325130e+00 1.81847900e-01 9.95556489e-02 -6.68359816e-01
-1.51556879e-01 9.62449133e-01 6.64889440e-02 3.99556458e-01
4.66342896e-01 4.57985587e-02 -1.45625532e-01 -4.72134292e-01
4.28802907e-01 5.19004405e-01 4.35067326e-01 -9.11222160e-01
8.27969670e-01 -2.34983265e-02 1.93266541e-01 -2.88587451e-01
-9.36387658e-01 1.73679143e-01 -3.92271996e-01 5.93990423e-02
5.08886695e-01 -9.12551105e-01 -5.55283964e-01 2.82724738e-01
-1.01641643e+00 -3.38448554e-01 -5.42535841e-01 7.80181170e-01
-5.89076459e-01 3.05567030e-03 -5.42285442e-01 -1.28416181e+00
-3.29593778e-01 -1.01055276e+00 9.19328213e-01 2.70112455e-01
-9.97717977e-02 -1.06778002e+00 3.48642528e-01 -1.47301212e-01
6.76005185e-01 3.83954346e-01 9.56630051e-01 -7.70720601e-01
-6.23082519e-01 -3.52221012e-01 -1.59003660e-01 7.37678111e-01
-4.09411415e-02 1.05759591e-01 -1.20322049e+00 -2.16829360e-01
5.15695438e-02 -4.38288540e-01 9.99756217e-01 6.24075115e-01
1.43344116e+00 -3.20514768e-01 -1.56948194e-01 6.12284243e-01
1.45130718e+00 2.71216452e-01 3.10294539e-01 -1.86623707e-01
2.59641707e-01 3.54913831e-01 2.43365794e-01 8.83975923e-01
1.85667738e-01 9.61953253e-02 4.94748801e-01 3.81152928e-01
5.27091801e-01 -3.00447285e-01 2.33395606e-01 5.22165596e-01
-1.11234024e-01 -3.14684480e-01 -1.12637603e+00 3.38770092e-01
-1.91540968e+00 -9.19934154e-01 5.37464797e-01 2.20483971e+00
8.72511864e-01 2.99343616e-01 -3.09349179e-01 1.51137680e-01
3.18011373e-01 1.53293774e-01 -8.53140175e-01 -3.99507850e-01
3.13533634e-01 4.75222826e-01 4.31726575e-01 4.77056533e-01
-1.09532547e+00 4.79728281e-01 6.71838093e+00 7.24397123e-01
-9.67253447e-01 -2.32296422e-01 8.16109121e-01 -2.48591855e-01
-4.24829811e-01 -5.85161597e-02 -7.07328260e-01 6.37237370e-01
1.58755016e+00 -1.86683476e-01 5.13983369e-01 9.57696557e-01
-3.22614796e-02 -2.09080428e-01 -1.50970113e+00 8.27198505e-01
-2.71532476e-01 -1.43423867e+00 -1.25048280e-01 -8.25357139e-02
8.59964013e-01 9.90254506e-02 2.08842710e-01 6.99431181e-01
8.41630399e-01 -1.38458264e+00 6.41091347e-01 1.00995064e+00
7.05027997e-01 -1.10510612e+00 9.64432359e-01 4.09018904e-01
-8.94948542e-01 -3.04416955e-01 -2.18760520e-01 -1.49274930e-01
1.00842752e-01 9.25083101e-01 -7.56975353e-01 6.54281378e-01
4.71209317e-01 5.07541418e-01 -1.50984198e-01 7.25908458e-01
-3.49796116e-01 6.87536418e-01 -5.68158805e-01 -1.66494042e-01
1.49108559e-01 -2.80908167e-01 1.45834520e-01 8.90143216e-01
5.93123138e-01 -2.19580486e-01 5.61877787e-02 1.46441209e+00
-3.98705043e-02 -6.37563884e-01 -4.30351168e-01 -3.33014876e-01
5.54505825e-01 1.09424663e+00 -2.99754322e-01 -3.60249728e-01
-1.97131276e-01 6.18838370e-01 3.29851210e-01 5.84872961e-01
-9.37690556e-01 -2.64161915e-01 5.44287503e-01 -4.80220407e-01
4.65574533e-01 -2.82219499e-01 -2.57783741e-01 -1.30898535e+00
-2.16392264e-01 -6.00104630e-01 2.64394999e-01 -5.43918669e-01
-1.38655984e+00 5.48183560e-01 2.99548686e-01 -8.06621611e-01
-1.22073102e+00 -6.12985790e-01 -5.49234569e-01 1.18158579e+00
-1.56160128e+00 -7.89504945e-01 9.42863300e-02 2.60077447e-01
1.65446594e-01 -1.63037255e-01 9.37853158e-01 -3.05885337e-02
-9.61034894e-01 2.03782201e-01 6.20349407e-01 9.39997435e-02
1.57624811e-01 -1.30737078e+00 2.49244347e-01 8.16212356e-01
-3.86834592e-02 5.07702053e-01 6.69044793e-01 -4.75558639e-01
-1.16710842e+00 -1.15339696e+00 3.96433502e-01 -4.43159521e-01
9.85872507e-01 -4.57089543e-01 -7.78434813e-01 8.09402406e-01
-1.33631647e-01 1.03902109e-01 6.32853746e-01 2.72847503e-01
-1.42797455e-01 -3.26391645e-02 -1.24759078e+00 5.52957952e-02
3.71707201e-01 -6.34860694e-01 -6.50874078e-01 -9.26055238e-02
4.63938117e-01 -4.81942415e-01 -1.15603793e+00 3.16657424e-01
5.40476918e-01 -5.81553400e-01 7.18079329e-01 -5.46769500e-01
4.01572466e-01 -1.73962831e-01 -2.37045273e-01 -1.62438416e+00
2.24941280e-02 -5.37272394e-01 -6.27215266e-01 1.21404672e+00
6.42975330e-01 -7.27469683e-01 7.26295233e-01 1.06020677e+00
4.27033752e-02 -6.57397509e-01 -1.18567765e+00 -6.84093952e-01
2.03916952e-01 -7.92185068e-01 1.04290807e+00 6.52448475e-01
-1.94699422e-01 2.46957392e-01 -3.63774031e-01 4.81111944e-01
7.65947342e-01 3.49940956e-01 3.33816648e-01 -1.57280052e+00
-6.86229408e-01 -3.74400318e-01 -2.63328820e-01 -6.98533535e-01
3.21168184e-01 -5.10303319e-01 1.71289086e-01 -1.21458578e+00
-1.03579618e-01 -4.71119314e-01 -5.38042247e-01 3.98816735e-01
1.54996276e-01 -4.13033217e-01 -1.56009227e-01 -7.64999613e-02
-1.27785355e-01 9.57268476e-01 4.57596421e-01 -1.05187707e-01
-1.37809953e-02 3.56380969e-01 -3.01968813e-01 7.66037524e-01
8.78238916e-01 -3.86478215e-01 -6.91303313e-01 -2.60736823e-01
3.87008071e-01 4.67221707e-01 4.28309143e-01 -1.15805829e+00
1.32674873e-01 -1.49293840e-01 7.60261476e-01 -7.89519548e-01
4.67227340e-01 -1.06062782e+00 3.96789819e-01 4.91754413e-01
-3.28893840e-01 -1.65711343e-01 3.65908027e-01 8.43424022e-01
-2.92251647e-01 -4.42547828e-01 5.37689447e-01 6.63244054e-02
-3.28084826e-01 3.12689364e-01 -3.47954363e-01 -3.31993103e-01
8.38285267e-01 3.54770035e-01 -1.02215596e-01 -3.62208605e-01
-9.54657972e-01 3.48281026e-01 -3.07264347e-02 1.35563493e-01
4.57467675e-01 -1.18387127e+00 -4.65635657e-01 3.23536545e-01
-2.33554095e-01 1.99264616e-01 2.01767206e-01 2.28185549e-01
-3.50623094e-02 5.14054477e-01 1.22024089e-01 -5.17576516e-01
-3.14428091e-01 3.56980175e-01 5.94404340e-01 -6.33314490e-01
-1.00221358e-01 6.61675036e-01 -1.92781582e-01 -4.22177941e-01
3.79132569e-01 -9.22668993e-01 -1.03603929e-01 1.83913633e-01
2.90948123e-01 4.64509606e-01 1.89059395e-02 5.42833097e-02
-1.41266912e-01 1.06090888e-01 2.12915570e-01 -3.00144315e-01
1.42421818e+00 2.95177530e-02 1.65803224e-01 8.19330633e-01
9.15428579e-01 -5.74956656e-01 -1.47366691e+00 -2.50017732e-01
1.23107746e-01 9.16280821e-02 3.83075505e-01 -1.08233452e+00
-9.41279829e-01 9.86998558e-01 5.42631090e-01 2.54950583e-01
8.91160488e-01 -1.99795783e-01 3.35819215e-01 5.83113313e-01
3.73001456e-01 -1.12519217e+00 -4.15167004e-01 5.88137269e-01
7.31696665e-01 -1.26146710e+00 5.94802983e-02 2.90364206e-01
-3.02954137e-01 1.02595913e+00 1.92371190e-01 -5.13115451e-02
9.26961064e-01 5.88433623e-01 -3.55667830e-01 6.04830533e-02
-1.06366205e+00 9.15818363e-02 5.23077607e-01 4.00525719e-01
2.04511553e-01 1.88511446e-01 1.78910151e-01 1.17705607e+00
-1.59357190e-01 1.74809888e-01 2.65013844e-01 8.37991238e-01
-3.17408353e-01 -9.63541269e-01 -2.54956484e-01 7.13676929e-01
-3.42779070e-01 -1.70802474e-01 3.16495061e-01 5.30939937e-01
-2.60952264e-01 6.92789257e-01 3.23791862e-01 -1.12465367e-01
1.01281583e-01 4.11514372e-01 2.91748554e-01 -5.49562335e-01
-1.88037664e-01 -2.32218131e-01 2.44835496e-01 -5.91362536e-01
1.29912971e-02 -7.06039786e-01 -1.04228830e+00 -3.96301523e-02
-3.74369472e-01 3.91946435e-01 1.09809029e+00 9.03439283e-01
3.54648888e-01 8.49457562e-01 5.97963572e-01 -1.00653422e+00
-1.39355397e+00 -1.12392104e+00 -8.07022035e-01 -2.68177748e-01
4.12793696e-01 -7.24848628e-01 -6.78960919e-01 -3.05996060e-01] | [7.2464280128479, 3.7531585693359375] |
8af3973d-f42e-4e91-82ca-b9abd70fbe82 | contrastive-learning-for-low-light-raw | 2305.03352 | null | https://arxiv.org/abs/2305.03352v1 | https://arxiv.org/pdf/2305.03352v1.pdf | Contrastive Learning for Low-light Raw Denoising | Image/video denoising in low-light scenes is an extremely challenging problem due to limited photon count and high noise. In this paper, we propose a novel approach with contrastive learning to address this issue. Inspired by the success of contrastive learning used in some high-level computer vision tasks, we bring in this idea to the low-level denoising task. In order to achieve this goal, we introduce a new denoising contrastive regularization (DCR) to exploit the information of noisy images and clean images. In the feature space, DCR makes the denoised image closer to the clean image and far away from the noisy image. In addition, we build a new feature embedding network called Wnet, which is more effective to extract high-frequency information. We conduct the experiments on a real low-light dataset that captures still images taken on a moonless clear night in 0.6 millilux and videos under starlight (no moon present, <0.001 lux). The results show that our method can achieve a higher PSNR and better visual quality compared with existing methods | ['Yuhan Dong', 'Taoyong Cui'] | 2023-05-05 | null | null | null | null | ['video-denoising'] | ['computer-vision'] | [ 3.30654204e-01 -5.53235888e-01 5.35195649e-01 -2.36812770e-01
-4.61848438e-01 -2.71696988e-02 4.31411535e-01 -4.00731623e-01
-6.03004336e-01 8.99246812e-01 3.65486205e-01 2.27595285e-01
-5.64175248e-02 -7.43863285e-01 -7.35472620e-01 -1.18719733e+00
7.63147101e-02 -5.29763877e-01 2.23013833e-01 -3.52223724e-01
1.57550171e-01 2.23744094e-01 -1.64702129e+00 4.34417911e-02
8.28407466e-01 8.42220664e-01 5.93090653e-01 5.50135612e-01
6.15730993e-02 8.31077158e-01 -4.32627827e-01 -9.55464467e-02
5.76329172e-01 -6.30401492e-01 -1.92145884e-01 1.99418068e-01
3.33760649e-01 -3.39816064e-01 -5.38538992e-01 1.42887223e+00
6.87554538e-01 2.35535011e-01 3.02584976e-01 -7.42156565e-01
-5.80534279e-01 1.04916689e-03 -7.11225331e-01 5.54984748e-01
8.13998803e-02 4.64242131e-01 6.37160718e-01 -7.46683896e-01
3.72423977e-01 1.15033770e+00 4.47177112e-01 5.11996388e-01
-1.02513933e+00 -5.81246018e-01 -6.75683767e-02 5.51641285e-01
-1.05683041e+00 -6.67428613e-01 9.62225080e-01 3.80242942e-04
4.44564402e-01 2.67133355e-01 7.45356560e-01 1.04364991e+00
5.74210048e-01 2.96909064e-01 1.59127998e+00 -4.79727358e-01
5.66810295e-02 -1.57311529e-01 -8.80786628e-02 4.76178259e-01
3.99355590e-01 2.93013096e-01 -4.84801203e-01 2.08243877e-01
5.87089241e-01 3.08975726e-01 -7.90889442e-01 1.63580254e-01
-9.46445942e-01 6.82659209e-01 6.76771820e-01 3.76421481e-01
-4.03852552e-01 3.29331487e-01 -2.26941369e-02 3.38438898e-01
5.97608626e-01 1.88699231e-01 -5.65855764e-02 1.83414012e-01
-5.73557794e-01 -3.37850563e-02 5.76695681e-01 3.12723249e-01
8.68284047e-01 3.53847802e-01 -1.46033047e-02 9.75316107e-01
4.19567555e-01 6.06375813e-01 4.11468148e-01 -1.01267004e+00
7.38710389e-02 8.90388116e-02 5.83783351e-02 -1.02843106e+00
-2.11582258e-01 -6.57509983e-01 -1.12791359e+00 6.68206930e-01
-5.80231398e-02 3.60654667e-02 -1.00044310e+00 1.53229463e+00
1.69399619e-01 5.60920477e-01 2.98398614e-01 1.31679356e+00
9.39599395e-01 1.04700291e+00 -2.67856061e-01 -7.21027017e-01
1.22621906e+00 -8.80425870e-01 -1.11707151e+00 -3.19205761e-01
-1.87704220e-01 -8.88779402e-01 1.08900392e+00 7.30999529e-01
-9.58597243e-01 -7.03478038e-01 -1.17482686e+00 -1.53990984e-01
-1.71520356e-02 -1.51368350e-01 4.72308278e-01 4.88022953e-01
-7.51685321e-01 5.40489137e-01 -6.11202359e-01 -2.74134576e-01
2.83691496e-01 -8.08577314e-02 -3.65077138e-01 -5.88412762e-01
-1.14818919e+00 8.14916551e-01 1.63597807e-01 5.09049654e-01
-1.04746282e+00 -1.85791045e-01 -7.70564914e-01 1.74403433e-02
4.83506709e-01 -6.75185740e-01 7.36471951e-01 -9.97730732e-01
-1.52229297e+00 4.65443850e-01 -1.56956658e-01 -3.21332425e-01
2.59781510e-01 -2.74532229e-01 -5.04138529e-01 3.37825209e-01
4.02496047e-02 1.58533007e-01 1.16915667e+00 -1.57733381e+00
-4.11539674e-01 -3.13566297e-01 6.08157590e-02 4.13312733e-01
-2.85116434e-01 -8.68539140e-02 -3.62200290e-01 -5.83174527e-01
1.89958796e-01 -5.60315609e-01 -2.10613683e-01 -8.89886264e-03
4.96750250e-02 9.09415185e-02 7.45472133e-01 -6.43326640e-01
7.98681974e-01 -2.25863004e+00 2.59967119e-01 -2.29563728e-01
2.85116106e-01 3.68714571e-01 -1.75084680e-01 3.20504218e-01
7.87751451e-02 -2.04802498e-01 -5.21904707e-01 -3.22566003e-01
-4.48342264e-01 5.37244558e-01 4.02304530e-02 7.15430081e-01
1.15451746e-01 5.35315454e-01 -8.51756275e-01 -4.33862090e-01
3.99514288e-01 6.82744384e-01 -2.86246210e-01 3.55193466e-01
1.44220114e-01 6.40860260e-01 -2.07849324e-01 6.15354896e-01
1.13643503e+00 1.94000632e-01 -2.74021119e-01 -5.29143512e-01
-4.42348093e-01 -2.84044832e-01 -1.25123513e+00 1.52529502e+00
-4.82703537e-01 6.77012205e-01 4.68474478e-01 -1.02865708e+00
9.33289528e-01 -8.01921338e-02 2.09226161e-01 -1.15513134e+00
1.75771788e-01 1.69996828e-01 -2.49713510e-02 -1.00977933e+00
3.44798118e-01 -6.19695783e-01 5.08162081e-01 -9.91780460e-02
-7.27916881e-02 -3.46352607e-01 9.12963524e-02 -4.41029519e-02
1.00961673e+00 5.96169978e-02 1.97515205e-01 -2.01289847e-01
7.20132470e-01 -5.55833936e-01 9.28385317e-01 6.82811260e-01
-2.57122368e-01 7.18298018e-01 6.37140404e-03 -5.09262264e-01
-8.38885903e-01 -9.11447167e-01 -2.23820224e-01 6.41694248e-01
5.70711255e-01 -2.65747309e-01 -4.50994760e-01 -1.50324568e-01
-4.17682916e-01 3.28269124e-01 -3.42591792e-01 -3.40648025e-01
-5.69910407e-01 -1.13463855e+00 -5.19934706e-02 -3.70114599e-03
1.12318754e+00 -1.00018060e+00 -3.13883960e-01 3.86472195e-02
-3.55736196e-01 -1.00918436e+00 -1.20229878e-01 2.08947092e-01
-6.44630015e-01 -1.03799760e+00 -6.04555607e-01 -7.54728019e-01
4.53203797e-01 6.55108809e-01 9.57746208e-01 1.30706623e-01
-4.39223468e-01 2.05736190e-01 -6.10414028e-01 -4.71287578e-01
-8.31833556e-02 -5.64784527e-01 2.83173304e-02 3.72471750e-01
1.59859776e-01 -5.69158971e-01 -8.74886751e-01 -1.11827189e-02
-1.34927440e+00 -1.26607940e-01 6.35157347e-01 1.06530774e+00
6.18167460e-01 6.09873712e-01 2.46718898e-01 -4.44490403e-01
5.99668145e-01 -3.44707191e-01 -7.01121509e-01 -1.40872851e-01
-5.27239382e-01 -1.27605721e-01 7.55433381e-01 -2.24948540e-01
-1.23368037e+00 -2.75915507e-02 -2.76786149e-01 -3.17639381e-01
3.03815883e-02 2.85465688e-01 -1.14893362e-01 -4.50749069e-01
5.89622021e-01 5.28608561e-01 -1.01907831e-02 -5.74150801e-01
8.35811123e-02 6.93807483e-01 6.31941497e-01 -6.33897334e-02
1.04999506e+00 7.64478385e-01 2.81293690e-01 -1.20753253e+00
-8.42731118e-01 -3.94570172e-01 -1.24659345e-01 -3.95221144e-01
9.66385305e-01 -1.10483372e+00 -8.76261771e-01 5.18423975e-01
-9.13410246e-01 -1.36992544e-01 -1.72702447e-01 8.24699402e-01
-3.16029936e-01 6.92306817e-01 -5.53969800e-01 -9.00437534e-01
-4.02381212e-01 -1.11506200e+00 7.14175284e-01 5.65594077e-01
8.46638918e-01 -7.74824798e-01 6.84404820e-02 3.35372657e-01
5.99259973e-01 2.77386576e-01 4.21791077e-01 4.13689494e-01
-8.56906533e-01 1.50718287e-01 -3.64232302e-01 1.00510061e+00
2.33308405e-01 -2.70250112e-01 -1.09869480e+00 -3.86554986e-01
1.01110017e+00 -1.93791106e-01 1.43480098e+00 5.24230897e-01
1.09009182e+00 -2.08931975e-02 1.48054764e-01 1.17256451e+00
1.89436841e+00 4.91223074e-02 1.11699414e+00 5.89702666e-01
6.54914439e-01 4.33155537e-01 6.20284438e-01 3.29620987e-01
1.65444389e-02 4.18085605e-01 9.28250372e-01 -4.52985168e-01
-4.05235648e-01 2.00635359e-01 3.84880841e-01 8.50799561e-01
-3.28044116e-01 -3.89341623e-01 -2.84424841e-01 4.17175680e-01
-1.53432918e+00 -1.07214296e+00 -4.83338237e-01 2.10236454e+00
8.12976062e-01 5.49131408e-02 -4.71647084e-01 2.39958450e-01
5.57672560e-01 4.34809208e-01 -2.62649179e-01 -1.35326624e-01
-5.73328793e-01 2.40316823e-01 5.28432190e-01 6.42921090e-01
-9.58652794e-01 6.23452902e-01 5.85861063e+00 7.01910794e-01
-1.18218589e+00 1.65731758e-01 4.09108400e-01 -4.42780331e-02
-2.64886081e-01 1.25643745e-01 -4.38087195e-01 8.51133704e-01
6.36741996e-01 1.65165186e-01 7.57925570e-01 2.97781438e-01
5.38886070e-01 -5.48103869e-01 -4.27046508e-01 1.41153097e+00
4.45891351e-01 -9.82451916e-01 -1.72352836e-01 -2.96267848e-02
6.30897224e-01 4.13155295e-02 -6.60409927e-02 1.03782691e-01
-2.28583485e-01 -8.95019174e-01 2.38535538e-01 1.06216931e+00
5.61384976e-01 -7.81314254e-01 1.01347077e+00 4.11946803e-01
-8.74511600e-01 -2.58141696e-01 -8.85713339e-01 -3.38713884e-01
2.36870989e-01 9.97801960e-01 -1.59816399e-01 6.53109074e-01
1.19998896e+00 9.61182594e-01 -4.80053514e-01 1.15798008e+00
-4.49737191e-01 5.54022133e-01 -1.15022860e-01 2.28484884e-01
6.64121434e-02 -6.73235416e-01 7.02071726e-01 1.01980114e+00
2.58839518e-01 4.24612641e-01 -4.53573838e-02 6.10338688e-01
-2.36108527e-01 -1.23463176e-01 -7.01298952e-01 3.79565418e-01
-1.78326696e-01 1.45208168e+00 -4.42038596e-01 -1.46621421e-01
-5.25916040e-01 1.16874719e+00 -1.55063391e-01 4.25353110e-01
-6.92238629e-01 -5.60204029e-01 4.96762574e-01 6.47319423e-04
1.24586180e-01 -4.26202178e-01 8.00584108e-02 -1.36704361e+00
1.86363816e-01 -7.71555066e-01 -1.14528937e-02 -1.17155099e+00
-1.38483596e+00 6.35289431e-01 -2.58672297e-01 -1.32327628e+00
5.34142137e-01 -5.24669290e-01 -7.04741180e-01 9.19655323e-01
-2.26382256e+00 -8.88147414e-01 -9.38171208e-01 6.94846153e-01
6.59916818e-01 1.90624148e-02 4.08012331e-01 4.68758255e-01
-4.31444407e-01 -1.35827348e-01 5.31788170e-01 -3.07487905e-01
9.24720705e-01 -1.08211422e+00 -1.12303786e-01 1.20538473e+00
-1.54727519e-01 3.28801870e-01 8.84868562e-01 -4.53026175e-01
-1.56327844e+00 -9.59421515e-01 5.00486910e-01 -3.57007086e-02
3.32642585e-01 -3.08125615e-01 -9.93565798e-01 3.16610247e-01
5.83384037e-01 3.48126262e-01 3.16829860e-01 -4.49213624e-01
1.08531965e-02 -7.57054806e-01 -1.28273129e+00 3.97528946e-01
9.54791725e-01 -3.06863129e-01 -7.22941518e-01 4.65519100e-01
7.01295912e-01 -8.46630707e-02 -5.16639173e-01 4.89884317e-01
2.38070458e-01 -1.22239161e+00 9.46252882e-01 9.63788107e-02
2.98388332e-01 -6.42485976e-01 -3.06254983e-01 -1.47640812e+00
-4.38940883e-01 -6.44752204e-01 1.35540143e-01 1.13278973e+00
-3.74929935e-01 -6.84812248e-01 3.06969076e-01 -2.24837527e-01
-3.73558104e-01 -2.94600636e-01 -9.05387878e-01 -7.41150618e-01
-3.11071575e-01 -2.18765989e-01 6.66131228e-02 6.92153931e-01
-6.89033687e-01 4.51874286e-01 -9.53720391e-01 3.38088632e-01
1.12153518e+00 2.24005133e-02 6.47853673e-01 -1.27459323e+00
-1.58276096e-01 1.83052585e-01 -4.65955555e-01 -8.60538661e-01
3.14092822e-02 -5.47084033e-01 2.95144588e-01 -1.73339593e+00
4.24753696e-01 2.63678338e-02 -3.47775340e-01 1.12295374e-01
-2.83218741e-01 7.29027390e-01 1.57596871e-01 2.64994800e-01
-4.52663332e-01 9.05749917e-01 1.39925027e+00 -2.05222845e-01
-1.04277544e-02 -2.09869727e-01 -6.29464924e-01 6.76429212e-01
6.00164235e-01 -5.62354743e-01 -1.73715934e-01 -7.35063314e-01
1.81087092e-01 -1.50457203e-01 5.02547979e-01 -1.24917102e+00
1.46764889e-01 -2.41401389e-01 5.51220000e-01 -2.87354857e-01
6.26327097e-01 -9.08962071e-01 1.05237737e-02 5.32377601e-01
1.73118487e-01 -4.30763453e-01 -2.12053433e-02 7.97416389e-01
-5.08425593e-01 -3.77400964e-01 1.19360948e+00 -4.85489100e-01
-7.60786593e-01 5.18658496e-02 -1.87723130e-01 -3.43008846e-01
7.67487705e-01 -2.60620505e-01 -5.34076154e-01 -4.31084991e-01
-5.33449411e-01 1.61282927e-01 4.47647691e-01 1.13376625e-01
9.46405053e-01 -1.10360205e+00 -8.53281617e-01 3.29508334e-01
1.15988255e-02 -1.46559671e-01 4.24794018e-01 9.17367220e-01
-7.39799857e-01 -1.85659960e-01 -3.72542560e-01 -6.31822646e-01
-1.29641497e+00 4.46055681e-01 3.86202157e-01 2.84341544e-01
-7.57371247e-01 6.60821736e-01 2.23895684e-01 1.65135157e-03
-1.24584220e-01 -2.19815612e-01 -3.81332129e-01 -2.01302066e-01
7.07545757e-01 5.42416930e-01 6.36171773e-02 -5.37957489e-01
-7.56826624e-02 9.00163710e-01 2.35849202e-01 2.33308151e-01
1.74057996e+00 -4.69880849e-01 -4.96148556e-01 3.77055913e-01
1.29845715e+00 4.01835948e-01 -1.38686109e+00 -1.20427236e-01
-4.98630732e-01 -9.02394891e-01 5.09459853e-01 -5.63729048e-01
-1.14352822e+00 7.85075366e-01 1.08606815e+00 3.64860684e-01
1.66025376e+00 -2.44028509e-01 7.54574835e-01 4.76901233e-01
1.06362417e-01 -1.00099015e+00 3.23080033e-01 3.82591635e-01
9.04475033e-01 -1.64110684e+00 2.58486867e-01 -2.07812101e-01
-2.72199243e-01 1.23822820e+00 3.84710252e-01 -2.92017430e-01
4.05512363e-01 1.56002436e-02 1.42488614e-01 -1.49958432e-01
-4.24972266e-01 -6.28152549e-01 -1.74664795e-01 5.92545927e-01
1.02352068e-01 -3.85819376e-01 -6.62135661e-01 1.87068388e-01
2.20170692e-01 4.49961331e-03 9.55349505e-01 8.17542434e-01
-9.57864463e-01 -9.14567828e-01 -6.25386000e-01 2.45560631e-01
-7.01285481e-01 -1.33701861e-01 6.01160564e-02 4.65222806e-01
4.23804969e-01 1.34539557e+00 -9.57073569e-02 -2.90493459e-01
3.58852595e-01 -3.22668731e-01 3.88591319e-01 -3.95479798e-01
-1.88191935e-01 4.66211021e-01 -2.29302481e-01 -5.99339008e-01
-8.85720432e-01 -2.09836230e-01 -9.52097178e-01 -7.16196820e-02
-3.38125139e-01 9.78612751e-02 8.42236817e-01 5.95266879e-01
-6.30383641e-02 6.89375103e-01 1.04739881e+00 -9.54882562e-01
-2.14929193e-01 -1.04178548e+00 -9.08309758e-01 5.10663092e-01
7.40983605e-01 -5.45107484e-01 -1.00864434e+00 8.93357173e-02] | [10.926766395568848, -2.5599827766418457] |
6c00d2e9-65e7-4f9e-8732-109657ec53ac | freedom-target-label-source-data-domain | 2307.02493 | null | https://arxiv.org/abs/2307.02493v1 | https://arxiv.org/pdf/2307.02493v1.pdf | FREEDOM: Target Label & Source Data & Domain Information-Free Multi-Source Domain Adaptation for Unsupervised Personalization | From a service perspective, Multi-Source Domain Adaptation (MSDA) is a promising scenario to adapt a deployed model to a client's dataset. It can provide adaptation without a target label and support the case where a source dataset is constructed from multiple domains. However, it is impractical, wherein its training heavily relies on prior domain information of the multi-source dataset -- how many domains exist and the domain label of each data sample. Moreover, MSDA requires both source and target datasets simultaneously (physically), causing storage limitations on the client device or data privacy issues by transferring client data to a server. For a more practical scenario of model adaptation from a service provider's point of view, we relax these constraints and present a novel problem scenario of Three-Free Domain Adaptation, namely TFDA, where 1) target labels, 2) source dataset, and mostly 3) source domain information (domain labels + the number of domains) are unavailable. Under the problem scenario, we propose a practical adaptation framework called FREEDOM. It leverages the power of the generative model, disentangling data into class and style aspects, where the style is defined as the class-independent information from the source data and designed with a nonparametric Bayesian approach. In the adaptation stage, FREEDOM aims to match the source class distribution with the target's under the philosophy that class distribution is consistent even if the style is different; after then, only part of the classification model is deployed as a personalized network. As a result, FREEDOM achieves state-of-the-art or comparable performance even without domain information, with reduced final model size on the target side, independent of the number of source domains. | ['Chan-Hyun Youn', 'Gyusang Cho', 'Eunju Yang'] | 2023-07-04 | null | null | null | null | ['philosophy'] | ['miscellaneous'] | [ 1.68879181e-01 1.03139249e-03 -5.10416210e-01 -5.65433025e-01
-6.33633792e-01 -8.39264750e-01 5.83688855e-01 -2.09827349e-01
-1.52272880e-01 8.22267234e-01 -1.20831020e-01 3.60100269e-02
-2.07128882e-01 -8.91507566e-01 -6.12917900e-01 -8.96685302e-01
4.27558839e-01 1.09827793e+00 3.20388168e-01 -3.29196602e-02
-2.85866082e-01 5.52772164e-01 -1.38446105e+00 2.23966867e-01
8.44919741e-01 1.26973867e+00 1.75897330e-01 1.63817182e-01
-4.10007894e-01 3.06471616e-01 -7.07943618e-01 -6.55983090e-01
5.21604478e-01 -3.00406903e-01 -6.80287123e-01 3.67014170e-01
1.08262636e-01 -3.10159862e-01 -1.54157713e-01 1.08100784e+00
4.29908335e-01 -1.22889653e-02 6.70252502e-01 -1.71493912e+00
-6.32310748e-01 3.51350933e-01 -1.39430135e-01 -2.15769783e-01
5.52145913e-02 -1.45784602e-01 7.82037139e-01 -6.83513105e-01
7.32224822e-01 8.69215012e-01 3.53449017e-01 6.58479631e-01
-1.41262841e+00 -6.56017005e-01 4.73946363e-01 2.16899812e-01
-1.41254401e+00 -4.96232212e-01 8.28851998e-01 -3.64039600e-01
3.90214682e-01 2.03988060e-01 2.63665110e-01 1.56124473e+00
-2.99464643e-01 6.19309306e-01 8.68142903e-01 -2.60287195e-01
8.07665646e-01 8.43550146e-01 6.25520758e-03 -8.22856799e-02
1.86418295e-01 -5.29224388e-02 -5.03832042e-01 -6.09882474e-01
5.30906379e-01 1.53806970e-01 -1.64891005e-01 -1.08777285e+00
-7.49300063e-01 6.04543030e-01 -9.80941951e-02 1.65900499e-01
-4.38244492e-01 -6.74685836e-01 2.73150474e-01 4.98474300e-01
4.01925325e-01 -1.00984015e-01 -1.03598523e+00 5.83735667e-02
-8.52502048e-01 4.83640190e-03 1.25345588e+00 1.61734772e+00
9.07334924e-01 -3.70605327e-02 1.38180226e-01 9.27657306e-01
3.27436924e-01 7.74120390e-01 5.25542200e-01 -6.51615202e-01
4.50818270e-01 6.81456268e-01 2.95333683e-01 -5.25632799e-01
-8.32016617e-02 -5.75211525e-01 -7.63645768e-01 1.03101186e-01
5.66924751e-01 -2.03951579e-02 -7.71632254e-01 2.17942548e+00
6.66242421e-01 1.64501116e-01 2.72342891e-01 7.67167866e-01
4.58305240e-01 3.35891336e-01 -6.46358207e-02 -4.78811800e-01
1.19216645e+00 -7.56760418e-01 -5.87372541e-01 -3.60298544e-01
2.28259534e-01 -5.17584860e-01 1.08195472e+00 4.52243418e-01
-8.50090086e-01 -3.55187207e-01 -7.69496858e-01 3.83034408e-01
-4.85198706e-01 -2.00945176e-02 1.51874363e-01 8.35371971e-01
-8.17542434e-01 7.85386637e-02 -4.32739705e-01 -7.10442662e-01
4.30594563e-01 4.99206275e-01 -5.58634281e-01 -4.03306484e-01
-1.13143635e+00 6.94127202e-01 3.98140877e-01 -4.53072011e-01
-8.88307095e-01 -6.50636494e-01 -3.83054197e-01 2.62674212e-01
6.52704418e-01 -1.00010264e+00 1.14051521e+00 -1.31302524e+00
-1.70659912e+00 6.38974309e-01 -2.59145021e-01 -8.14067423e-02
6.75467074e-01 1.22482568e-01 -9.46589351e-01 -8.37576240e-02
-6.02133758e-02 -5.85398152e-02 1.16338468e+00 -1.48026550e+00
-8.64071131e-01 -5.58333457e-01 -2.54140869e-02 6.31720871e-02
-6.04912758e-01 -1.91830769e-01 -6.39544189e-01 -3.92735749e-01
2.48316377e-02 -9.27433312e-01 4.22565304e-02 1.82343662e-01
-3.11895758e-01 -8.99129882e-02 1.02558219e+00 -4.69249159e-01
1.13297927e+00 -2.34349465e+00 2.14481249e-01 5.83944738e-01
1.28144979e-01 3.96994591e-01 -2.44364426e-01 5.10016501e-01
1.79913899e-04 5.30391745e-02 -2.39031211e-01 -4.28783774e-01
2.15018302e-01 6.02832556e-01 -2.76101887e-01 2.31486902e-01
-3.67431417e-02 3.55033398e-01 -7.87240088e-01 -2.24360839e-01
-1.37578622e-01 4.09788251e-01 -5.60518682e-01 6.13561094e-01
-2.70302057e-01 6.70225441e-01 -5.92618585e-01 7.55655289e-01
1.10149348e+00 -3.87812078e-01 6.70869350e-01 -2.45551974e-01
4.62717772e-01 7.56358653e-02 -1.44606638e+00 1.79038298e+00
-4.93106127e-01 -8.74367058e-02 3.39587748e-01 -9.80239272e-01
1.09636366e+00 5.11711538e-01 7.13214636e-01 -3.98639292e-01
-2.03450248e-01 4.13964361e-01 -1.75125197e-01 -3.57152134e-01
-2.97568813e-02 -8.72375667e-02 -1.45606235e-01 5.13434768e-01
3.41917276e-01 3.93891543e-01 -8.42459425e-02 -2.17930451e-02
1.07284498e+00 7.92659745e-02 3.95184577e-01 2.07168004e-03
4.03712273e-01 -1.93866849e-01 8.17860901e-01 7.64444888e-01
-2.22569808e-01 4.59632188e-01 3.96816462e-01 -3.23899239e-01
-1.01124024e+00 -1.38513184e+00 -1.02596134e-01 1.30377293e+00
6.74292669e-02 -4.18208092e-02 -6.43841505e-01 -1.24220538e+00
1.88416302e-01 7.80635774e-01 -4.32631761e-01 -2.63507694e-01
-3.21210742e-01 -5.14279783e-01 3.88984799e-01 2.15620548e-01
3.57780963e-01 -5.83657444e-01 5.11374474e-02 2.76456535e-01
-2.62431204e-01 -1.16877389e+00 -4.87998307e-01 2.41078734e-01
-7.00393260e-01 -1.01179683e+00 -5.66295862e-01 -4.91355777e-01
6.40468776e-01 -1.07896058e-02 1.18173528e+00 -4.50756550e-01
5.28183103e-01 4.83387440e-01 -4.22186196e-01 -1.39638737e-01
-5.15495658e-01 3.88853997e-01 3.15571338e-01 5.55162489e-01
6.39909863e-01 -1.10601640e+00 -3.37939113e-01 7.44214356e-01
-1.04858971e+00 -4.03108120e-01 5.67739844e-01 7.92398751e-01
5.25313795e-01 1.35475680e-01 9.34727311e-01 -1.19212472e+00
3.30141038e-01 -1.00109541e+00 -4.22249496e-01 5.24526954e-01
-9.78620708e-01 -3.10973167e-01 8.49038303e-01 -9.37292039e-01
-1.14162636e+00 -2.00735014e-02 7.21729323e-02 -5.48812389e-01
-6.13017619e-01 3.11679542e-01 -1.00616360e+00 1.66206151e-01
6.36021972e-01 3.71382743e-01 5.94397895e-02 -9.42108452e-01
3.48105401e-01 1.02698672e+00 4.49082822e-01 -6.91271961e-01
9.09996808e-01 4.57055390e-01 -2.43418843e-01 -3.23195696e-01
-7.47089863e-01 -4.82354254e-01 -8.87419641e-01 1.50351122e-01
2.98701286e-01 -9.07631457e-01 -3.86123985e-01 3.56994003e-01
-9.36974168e-01 -1.11867324e-01 -5.31058073e-01 3.00087124e-01
-5.74273050e-01 2.47156277e-01 2.49199569e-02 -6.13546729e-01
-6.89888671e-02 -1.02950227e+00 6.36767626e-01 7.22607225e-02
-3.33446562e-02 -1.23792779e+00 -1.74863577e-01 1.06955275e-01
6.45440102e-01 -2.88053360e-02 1.10649407e+00 -1.56896150e+00
-4.43832397e-01 -3.84487182e-01 -6.82756901e-02 4.57606047e-01
4.52567846e-01 -4.17879820e-01 -1.01710212e+00 -5.21911740e-01
1.91535071e-01 9.34945643e-02 -6.65515102e-03 1.36008812e-02
1.04011428e+00 -6.06083274e-01 -2.64537334e-01 6.43132031e-01
1.35423708e+00 2.56820232e-01 2.97792047e-01 1.61561146e-01
6.00131929e-01 5.74282110e-01 5.20682156e-01 8.63256574e-01
5.57653904e-01 1.08903909e+00 4.22385931e-01 5.28998598e-02
-1.34155467e-01 -4.09176260e-01 3.98741603e-01 5.43742418e-01
4.24674332e-01 -5.59391916e-01 -7.45826542e-01 5.18420577e-01
-1.92468512e+00 -7.42598593e-01 3.30009043e-01 2.66649508e+00
8.55549395e-01 -1.83209315e-01 4.81066138e-01 -1.85298666e-01
8.21585417e-01 -3.23127300e-01 -1.16422367e+00 -2.14213073e-01
-9.09771994e-02 -1.42598420e-01 4.51570660e-01 1.60761148e-01
-7.19456434e-01 6.23940825e-01 6.05775023e+00 9.19444025e-01
-9.29197431e-01 3.87023211e-01 2.54625469e-01 -1.59672976e-01
-4.16982651e-01 7.88884014e-02 -9.88422155e-01 8.47656786e-01
1.03206885e+00 -2.97167748e-01 7.41351843e-01 1.13503408e+00
-1.06038950e-01 2.79605120e-01 -1.50558531e+00 7.44934082e-01
-1.01154312e-01 -9.08869922e-01 5.84142692e-02 3.55476052e-01
4.85369563e-01 7.53465593e-02 7.37309307e-02 3.73066604e-01
4.49062645e-01 -3.61428261e-01 6.29203796e-01 4.70248729e-01
1.13580084e+00 -6.15056694e-01 6.70535386e-01 8.24032784e-01
-9.15081561e-01 -3.58162791e-01 -1.81366861e-01 3.99901986e-01
1.83868065e-01 6.59538448e-01 -8.45436215e-01 7.50606358e-01
7.85319388e-01 4.56144363e-01 -2.76672304e-01 7.93392956e-01
1.52274996e-01 6.08123779e-01 -4.69993651e-01 6.25894129e-01
-4.08902764e-01 -2.53340393e-01 6.26568079e-01 8.37361753e-01
5.53650439e-01 -1.09396197e-01 4.28920507e-01 7.00483918e-01
-8.68618116e-02 5.03851585e-02 -5.70263863e-01 1.77719116e-01
1.07471371e+00 1.04896772e+00 -1.32386237e-01 -3.92266661e-01
-5.98854065e-01 1.05304456e+00 2.67283469e-01 7.41386831e-01
-6.48145258e-01 5.99828623e-02 9.93779302e-01 2.81619728e-01
3.31824094e-01 1.56235501e-01 -2.43417859e-01 -1.36035621e+00
1.59744218e-01 -9.48928356e-01 6.65741682e-01 -4.97166812e-01
-1.89236498e+00 5.96566439e-01 1.16006127e-02 -1.43808961e+00
-3.50325733e-01 -2.90917844e-01 -2.71079510e-01 1.05917406e+00
-1.57516670e+00 -1.51795805e+00 -1.36277318e-01 1.17622864e+00
2.62512445e-01 -5.80815613e-01 1.16505396e+00 6.33105814e-01
-5.07091880e-01 1.17922497e+00 4.98172730e-01 -1.71878442e-01
1.07070506e+00 -9.92435396e-01 -2.69629676e-02 6.69392169e-01
-6.21887818e-02 5.80262363e-01 4.90261972e-01 -5.42961001e-01
-1.19192827e+00 -1.20423627e+00 8.51227164e-01 -5.17025113e-01
4.63530868e-01 -4.42094028e-01 -1.14053500e+00 8.04138362e-01
-1.83677182e-01 3.06588143e-01 1.19698799e+00 1.11172654e-01
-6.01848781e-01 -4.67576832e-01 -1.77230179e+00 2.05282673e-01
1.06888402e+00 -5.17058194e-01 -1.30915672e-01 3.00730377e-01
6.63869023e-01 -2.01241419e-01 -1.25284982e+00 1.20711021e-01
4.27966803e-01 -8.36171925e-01 9.21502292e-01 -7.91260600e-01
-1.87246367e-01 -4.28333461e-01 -6.40021145e-01 -1.23731470e+00
-4.51641649e-01 -4.96999651e-01 -5.51140428e-01 1.75215948e+00
3.24047744e-01 -9.48359847e-01 6.85685456e-01 1.05112231e+00
6.19285088e-03 -3.70603800e-01 -1.28152668e+00 -1.15757799e+00
-1.04091801e-02 -1.66969389e-01 1.30853033e+00 1.05243099e+00
-2.85699725e-01 1.78900078e-01 -5.29784620e-01 5.76001465e-01
4.62990195e-01 9.74646360e-02 1.00032926e+00 -1.53948867e+00
-4.65111613e-01 -2.78964117e-02 -7.65527338e-02 -1.04641247e+00
2.78574377e-01 -7.75670588e-01 -3.02271485e-01 -1.15116644e+00
9.13225934e-02 -8.26098621e-01 -5.28572798e-01 6.28846824e-01
2.80132055e-01 -2.13510692e-01 1.81668058e-01 6.82635367e-01
-5.31451762e-01 4.29532111e-01 8.67410064e-01 1.19489960e-01
-3.07173640e-01 6.15109444e-01 -9.09424901e-01 4.90444541e-01
7.00852931e-01 -7.14739263e-01 -6.95465386e-01 -3.94544125e-01
-1.28634751e-01 1.22960255e-01 1.63720131e-01 -6.89741671e-01
2.09258556e-01 -5.80335796e-01 1.19498476e-01 -1.27595007e-01
3.66867393e-01 -1.62104261e+00 6.46603227e-01 -8.87506679e-02
-1.69835523e-01 -2.72837728e-01 -1.93478972e-01 8.07215869e-01
-7.69433230e-02 -1.41751096e-01 7.98515677e-01 1.41421616e-01
-6.76902056e-01 7.55557954e-01 -6.16283193e-02 -7.22921342e-02
1.08721352e+00 -2.76260376e-01 -3.02916765e-01 -3.42404842e-01
-1.04134345e+00 -1.62720587e-02 8.74743640e-01 4.67681617e-01
1.70851961e-01 -1.52625954e+00 -4.48298454e-01 5.30994177e-01
3.94510448e-01 2.22291015e-02 2.58235425e-01 5.85973740e-01
4.40274030e-01 -8.51564109e-02 -1.60697848e-01 -6.10641360e-01
-9.35924947e-01 8.69214833e-01 1.20953999e-01 -2.68839300e-01
-3.07657540e-01 4.68812734e-01 2.76609480e-01 -7.71268308e-01
2.21374899e-01 2.47391880e-01 -2.07431614e-02 1.51334420e-01
3.80613625e-01 3.39246601e-01 1.42630950e-01 -6.22540474e-01
-3.73659074e-01 2.75642782e-01 -2.15472355e-01 -2.86890101e-02
1.20689285e+00 -6.89753354e-01 4.84229587e-02 3.67670357e-01
1.08748388e+00 4.50190566e-02 -1.49250865e+00 -9.95605171e-01
-1.74226966e-02 -6.78948402e-01 -3.68932277e-01 -1.26677692e+00
-1.08689535e+00 6.13045871e-01 6.99910820e-01 2.15849116e-01
1.47830987e+00 1.04077958e-01 6.31917596e-01 7.29748011e-02
5.49161494e-01 -1.13296533e+00 -1.44728929e-01 2.86094517e-01
5.72681904e-01 -1.24230623e+00 -2.65861750e-01 -3.16150993e-01
-8.16340804e-01 9.22821701e-01 6.21116161e-01 3.19023013e-01
8.07265759e-01 7.69918412e-02 1.10808648e-01 2.41560951e-01
-7.53436685e-01 -2.48579569e-02 2.00741410e-01 1.31039572e+00
-2.27368876e-01 1.61153451e-01 1.15410142e-01 1.14764333e+00
1.07763670e-01 7.77599812e-02 3.11751783e-01 6.18469357e-01
-3.39937881e-02 -1.74956000e+00 -2.09750012e-01 2.85848647e-01
-2.27702796e-01 1.29935995e-01 -2.05534905e-01 5.22273064e-01
4.24932301e-01 9.78108287e-01 -9.43063572e-02 -3.11393410e-01
4.72404182e-01 4.07671720e-01 -1.05050916e-03 -7.86057889e-01
-3.66503000e-01 1.10051155e-01 -5.53866029e-02 -3.94757122e-01
-1.96529195e-01 -6.71555758e-01 -7.82330453e-01 -4.66681927e-01
-3.23980242e-01 1.75905585e-01 6.71077490e-01 9.53068495e-01
8.60443115e-01 7.46990293e-02 9.61292386e-01 -4.79075253e-01
-9.54203308e-01 -7.32205927e-01 -9.19310153e-01 4.76902425e-01
2.23706111e-01 -7.20341444e-01 -2.80551553e-01 2.51252025e-01] | [10.378844261169434, 3.1971089839935303] |
aab964a8-7fcd-45d4-a7fc-464a4b1b7b43 | practical-lessons-on-optimizing-sponsored | 2304.09107 | null | https://arxiv.org/abs/2304.09107v1 | https://arxiv.org/pdf/2304.09107v1.pdf | Practical Lessons on Optimizing Sponsored Products in eCommerce | In this paper, we study multiple problems from sponsored product optimization in ad system, including position-based de-biasing, click-conversion multi-task learning, and calibration on predicted click-through-rate (pCTR). We propose a practical machine learning framework that provides the solutions to such problems without structural change to existing machine learning models, thus can be combined with most machine learning models including shallow models (e.g. gradient boosting decision trees, support vector machines). In this paper, we first propose data and feature engineering techniques to handle the aforementioned problems in ad system; after that, we evaluate the benefit of our practical framework on real-world data sets from our traffic logs from online shopping site. We show that our proposed practical framework with data and feature engineering can also handle the perennial problems in ad systems and bring increments to multiple evaluation metrics. | ['Musen Men', 'Jayanth Korlimarla', 'Weizhi Du', 'Bo Liu', 'Yanbing Xue'] | 2023-04-05 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [ 6.28914461e-02 -4.06347275e-01 -7.81059682e-01 -7.94232011e-01
-9.21102047e-01 -3.82017404e-01 1.94993749e-01 1.05190471e-01
-2.16368228e-01 6.89156234e-01 -7.92286396e-02 -7.58881807e-01
-3.36745083e-01 -7.08916545e-01 -6.23261273e-01 -2.09074736e-01
-9.97125078e-03 4.51393872e-01 2.34004721e-01 -5.25512934e-01
4.63664085e-01 1.76553920e-01 -1.31679833e+00 4.36872184e-01
8.37504506e-01 1.46961796e+00 1.13278516e-02 5.89868784e-01
-1.86032057e-01 4.86814946e-01 -3.95413429e-01 -8.26700747e-01
5.15591323e-01 2.85026968e-01 -5.09494305e-01 -6.21649846e-02
6.59314036e-01 -5.33308625e-01 -7.62323141e-02 6.53807998e-01
5.38777649e-01 -1.46100596e-01 6.47544742e-01 -1.53239429e+00
-5.78059673e-01 4.60129082e-01 -9.99307036e-01 2.86867172e-01
2.62926370e-02 -2.22141638e-01 1.49318266e+00 -9.49159086e-01
6.02952130e-02 1.29581237e+00 7.12287486e-01 9.52996388e-02
-1.19083774e+00 -6.51765108e-01 3.91097724e-01 3.03952068e-01
-9.11371291e-01 -1.65327281e-01 7.17635512e-01 -4.95914906e-01
6.45444274e-01 3.10365975e-01 2.60487258e-01 1.20392513e+00
4.09181297e-01 1.25744522e+00 1.12766206e+00 -4.34703767e-01
1.12889305e-01 7.09466338e-01 8.51768434e-01 4.83785927e-01
3.15944344e-01 1.53394595e-01 -4.97449189e-01 -4.68443781e-01
6.18182659e-01 3.73491883e-01 3.90312195e-01 -4.08534408e-01
-7.19820440e-01 1.30601871e+00 2.75965452e-01 -9.38810557e-02
-2.46777251e-01 2.38389462e-01 2.13353306e-01 4.49392796e-01
4.84454095e-01 1.49119884e-01 -1.12600541e+00 9.76427048e-02
-8.59971225e-01 6.21536493e-01 1.05587471e+00 1.33805931e+00
7.79439032e-01 -2.22112432e-01 -4.75776374e-01 1.16788220e+00
6.85813189e-01 3.25258642e-01 5.46337485e-01 -6.27401352e-01
7.08716512e-01 3.59236985e-01 3.90972197e-01 -8.03886354e-01
-3.29371899e-01 -6.99180663e-01 -3.00395250e-01 -1.31276786e-01
3.58002186e-01 -3.30386400e-01 -7.08516359e-01 1.31888127e+00
2.97094584e-01 -1.83751941e-01 -4.76541132e-01 6.16447866e-01
2.89566934e-01 2.84996331e-01 2.52985626e-01 -4.13991399e-02
1.46923661e+00 -1.36181343e+00 -7.64629960e-01 -3.83446693e-01
8.69721651e-01 -8.92651379e-01 1.25376129e+00 8.70094776e-01
-8.70121658e-01 -6.54196143e-01 -1.27016306e+00 -4.35648002e-02
-5.83326101e-01 2.32958749e-01 1.40046573e+00 1.11275673e+00
-4.08717871e-01 4.97236878e-01 -3.56766880e-01 -1.01278357e-01
4.91781950e-01 4.13086891e-01 5.07917523e-01 -1.65515691e-01
-1.21195495e+00 7.84283042e-01 -9.03518591e-03 -6.83587417e-02
-9.51902032e-01 -8.68688226e-01 -3.66832942e-01 1.31151512e-01
6.41075552e-01 -5.76651037e-01 1.75942481e+00 -7.28678286e-01
-1.39503133e+00 1.69844568e-01 -1.23021483e-01 -5.18023670e-01
5.13423622e-01 -6.39369309e-01 -6.92880511e-01 -7.38944948e-01
-3.20411138e-02 1.94732070e-01 8.78432989e-01 -1.00273705e+00
-1.07458651e+00 -6.43843293e-01 1.21278934e-01 -1.33829251e-01
-6.76638842e-01 -2.17554774e-02 -5.56135811e-02 -5.77741981e-01
-1.56268016e-01 -8.94799948e-01 -4.91820514e-01 -1.14036880e-01
-3.10312927e-01 -3.53369713e-01 7.47491360e-01 -5.79162776e-01
1.58646333e+00 -1.90624022e+00 -3.79879206e-01 3.22662473e-01
2.05900311e-01 6.24980070e-02 -1.38204262e-01 1.97002947e-01
1.87337160e-01 7.94664472e-02 4.61078495e-01 -1.74320757e-03
1.64019570e-01 4.77301404e-02 -2.33673617e-01 -9.52247251e-03
4.87047397e-02 8.04659724e-01 -4.10038918e-01 -5.17668843e-01
-1.23061940e-01 -3.22796330e-02 -8.15436065e-01 9.26373452e-02
-4.55330640e-01 -2.59854048e-01 -7.77578771e-01 1.06102633e+00
8.73264134e-01 -6.64498687e-01 1.83530942e-01 -3.21764141e-01
1.80864751e-01 3.10624540e-01 -1.43139184e+00 1.45341897e+00
-9.73525226e-01 -8.86382461e-02 1.99702922e-02 -9.49099422e-01
8.65389526e-01 -1.57928795e-01 5.10717869e-01 -8.32825065e-01
-5.46698906e-02 2.91967064e-01 -2.63171028e-02 -4.83363062e-01
5.60439527e-01 2.73629755e-01 -3.76963280e-02 3.08715910e-01
7.85274953e-02 4.33624864e-01 1.20496847e-01 1.27264872e-01
8.50774646e-01 -4.60220277e-02 2.08221912e-01 -1.08301915e-01
4.36433434e-01 9.13631171e-02 3.67534220e-01 8.70046675e-01
-2.16500834e-01 1.90549955e-01 2.54958510e-01 -4.43107635e-01
-1.04173017e+00 -9.39717710e-01 -4.30832595e-01 1.87305069e+00
-4.73450013e-02 -5.29045343e-01 -2.70772219e-01 -1.00398171e+00
7.29267597e-01 7.05019712e-01 -2.57998347e-01 4.45185266e-02
-4.17947620e-01 -1.11447251e+00 6.03655577e-02 4.71549958e-01
5.30771017e-01 -1.83301941e-01 2.75514573e-01 4.35792059e-01
1.84287697e-01 -9.65260506e-01 -5.66002429e-01 4.01702106e-01
-1.09626639e+00 -7.62426496e-01 -6.52959466e-01 -6.31069601e-01
-3.59212980e-02 5.57351589e-01 1.06316030e+00 -2.76613772e-01
-4.23354924e-01 7.62055963e-02 -1.70005217e-01 -6.60511613e-01
-2.75520887e-03 4.85754371e-01 -1.12352230e-01 4.60696742e-02
8.77210617e-01 -4.03032929e-01 -8.01984191e-01 9.23381746e-01
-6.41520917e-01 -2.77186215e-01 9.38188195e-01 9.35692847e-01
3.94351691e-01 -1.08436026e-01 8.48549426e-01 -1.19309390e+00
1.00238335e+00 -7.37018883e-01 -7.51480699e-01 5.17512381e-01
-1.50981200e+00 1.72559753e-01 1.86469600e-01 -6.86712086e-01
-9.40893769e-01 5.06493077e-02 -4.54243161e-02 2.11368188e-01
1.97868600e-01 5.97252190e-01 -3.73469740e-02 -1.83714554e-01
7.80918896e-01 -1.13336444e-01 -1.56215876e-01 -9.76439238e-01
5.61113536e-01 1.24113572e+00 -2.96875119e-01 -5.09591341e-01
4.92939740e-01 8.40175748e-02 -7.08158463e-02 -3.06061834e-01
-1.19131315e+00 -7.11511970e-01 -4.97521251e-01 2.37594500e-01
3.52735668e-01 -9.74280477e-01 -7.91613400e-01 1.01402916e-01
-8.34035516e-01 1.20565750e-01 2.56227553e-01 5.04138470e-01
-5.29172063e-01 2.42379397e-01 -6.22213781e-01 -8.75931025e-01
-4.14166033e-01 -1.01573658e+00 9.51088428e-01 1.28645077e-01
-6.23908713e-02 -1.03756046e+00 1.07722566e-01 8.41973066e-01
7.06203878e-01 -3.57630312e-01 1.04612684e+00 -9.96029079e-01
-6.45316005e-01 -3.69544476e-01 -3.33527774e-01 4.29006666e-01
3.52792158e-05 -2.55065501e-01 -8.88912797e-01 -3.07929039e-01
-1.17347628e-01 -3.74432057e-01 9.51207042e-01 6.09302342e-01
1.50364590e+00 -3.41839194e-01 -5.77671647e-01 3.58407855e-01
1.43408346e+00 1.75001159e-01 3.02277386e-01 5.54139316e-01
3.42055738e-01 3.96096796e-01 1.11029851e+00 6.46166444e-01
5.32343626e-01 9.48395789e-01 3.66439223e-01 -1.11914404e-01
9.45139974e-02 -4.78997052e-01 2.59500682e-01 3.88847828e-01
1.02314085e-01 -1.79518551e-01 -3.60628098e-01 1.88231900e-01
-2.05105352e+00 -4.56760287e-01 -2.49259040e-01 2.09619975e+00
6.42498910e-01 3.60699236e-01 5.01511455e-01 -1.33300483e-01
7.28023052e-01 2.15704143e-02 -7.14326441e-01 -4.69861656e-01
1.36760116e-01 1.98509604e-01 1.10609996e+00 3.12353075e-01
-1.31368780e+00 7.75245130e-01 7.15262318e+00 1.10380995e+00
-5.85011482e-01 5.79151750e-01 6.71306372e-01 -2.03630269e-01
-2.45489776e-01 -2.45540831e-02 -1.62656462e+00 4.90071952e-01
1.12019861e+00 -1.92349061e-01 4.91463512e-01 1.58104920e+00
1.86910123e-01 1.96340740e-01 -1.38223672e+00 1.08714783e+00
-1.43388048e-01 -1.20840096e+00 -8.46243836e-03 5.01440585e-01
5.81539154e-01 -1.06480643e-02 3.67602885e-01 8.44850779e-01
7.45209157e-01 -8.32552433e-01 3.04908931e-01 -7.03120157e-02
3.15195680e-01 -4.11554158e-01 6.71803892e-01 7.56290853e-02
-7.23818839e-01 -6.07862115e-01 -4.00654048e-01 4.45362329e-02
3.36438864e-01 1.02318478e+00 -7.53186405e-01 4.01894867e-01
5.70045710e-01 5.97595930e-01 -7.37083137e-01 1.25174248e+00
3.08469087e-01 5.11013329e-01 -2.52000272e-01 -5.57330370e-01
2.57512271e-01 -1.00955993e-01 -5.26175648e-02 1.01049793e+00
4.05584872e-01 -4.28546935e-01 1.51345402e-01 6.59516037e-01
3.23994905e-02 3.96711439e-01 -2.70395607e-01 -1.09802396e-03
1.34194821e-01 1.62358737e+00 5.71310632e-02 -2.03357548e-01
-8.60108972e-01 6.43314481e-01 1.38577402e-01 4.55384195e-01
-1.13096905e+00 -3.38532329e-01 6.47064984e-01 5.16698956e-01
4.02385384e-01 -2.49512821e-01 -4.66459155e-01 -1.15937722e+00
2.03254908e-01 -8.70103240e-01 6.25149727e-01 -3.82392079e-01
-1.63277280e+00 1.75741434e-01 -1.01741895e-01 -1.10664630e+00
-3.70201506e-02 -1.08988309e+00 -4.46346998e-01 6.17280662e-01
-1.93907344e+00 -1.17950928e+00 -1.41773045e-01 5.98557889e-01
7.36365736e-01 -5.45736253e-01 3.60645384e-01 9.67696011e-01
-5.53840518e-01 9.47577953e-01 2.23096058e-01 -2.34182283e-01
9.63077724e-01 -1.23417687e+00 3.44795734e-01 1.00478008e-02
3.53039280e-02 6.92066669e-01 5.30221641e-01 -5.26448548e-01
-1.49189770e+00 -9.75936472e-01 6.27897441e-01 -3.83606702e-01
9.84014571e-01 -5.16502440e-01 -2.92310685e-01 8.68464947e-01
-1.07015014e-01 -2.54642993e-01 1.04515064e+00 1.00855196e+00
-4.08243299e-01 -5.40079832e-01 -1.12604070e+00 2.65959591e-01
9.42283869e-01 -1.54996132e-02 -1.53553486e-01 9.88258660e-01
6.52090728e-01 -9.47738662e-02 -1.07864296e+00 2.76532799e-01
8.53594661e-01 -5.56724072e-01 1.15583360e+00 -1.32237697e+00
3.35034698e-01 3.09879005e-01 -4.76056278e-01 -1.20403814e+00
-5.64487696e-01 -5.28209627e-01 -2.94233918e-01 1.12763715e+00
1.17455173e+00 -7.71906972e-01 1.05351460e+00 8.38867366e-01
5.73645607e-02 -8.35845888e-01 -7.22234786e-01 -8.38084102e-01
3.99647877e-02 -4.23653841e-01 6.79167092e-01 6.27902627e-01
-2.66355276e-01 6.83939099e-01 -7.80505359e-01 3.63187417e-02
7.07539380e-01 1.72029108e-01 8.75154734e-01 -1.41028214e+00
-8.75531554e-01 -1.00161351e-01 3.72274313e-03 -1.63741875e+00
-3.58886182e-01 -7.74183214e-01 -6.20193124e-01 -1.09770703e+00
3.42513293e-01 -8.37488234e-01 -5.20762861e-01 2.10980177e-01
3.02008931e-02 -3.51205766e-01 -2.28956854e-03 1.53230727e-01
-6.74558640e-01 3.55783165e-01 1.08300889e+00 -3.67421210e-01
-2.09538892e-01 7.44665623e-01 -1.05194592e+00 3.75772566e-01
6.51914358e-01 -6.02972507e-01 -2.43071049e-01 -3.15774202e-01
3.37783962e-01 -1.30595282e-01 1.42751008e-01 -4.49288785e-01
-1.43310418e-02 -3.40673923e-01 4.62975889e-01 -6.93620682e-01
3.33424032e-01 -9.27959979e-01 -2.58313268e-01 3.96013856e-01
-5.67532837e-01 5.87508045e-02 -1.67305246e-01 8.44583809e-01
-2.21635979e-02 -3.34076226e-01 4.58630592e-01 -2.12615520e-01
-5.06097138e-01 4.30794030e-01 1.31096765e-01 -2.36026630e-01
9.63017166e-01 1.10967152e-01 -6.87207162e-01 -2.66124785e-01
-7.63506711e-01 6.33206666e-01 -3.81788343e-01 6.96907640e-01
2.05395415e-01 -1.56333864e+00 -5.58420122e-01 1.31634086e-01
3.64198387e-01 -7.44565785e-01 -7.55172446e-02 6.59830153e-01
-1.31464139e-01 8.35648596e-01 3.84387076e-02 -4.88276929e-01
-1.06771564e+00 7.37005472e-01 4.73460443e-02 -5.61017454e-01
1.44780815e-01 5.92999518e-01 -1.72206447e-01 -3.68059307e-01
4.53615814e-01 -3.62751514e-01 -3.48051190e-02 8.49478543e-02
3.91163886e-01 5.09916484e-01 3.54908764e-01 3.02709877e-01
-1.22582756e-01 3.15905005e-01 -8.56358230e-01 -4.20238040e-02
1.34221542e+00 -2.46611476e-01 4.63591754e-01 3.18469524e-01
1.25865686e+00 -2.46209741e-01 -9.14097905e-01 -3.60949844e-01
3.85048270e-01 -8.23879182e-01 4.15084660e-01 -1.12914884e+00
-9.55891728e-01 8.58234823e-01 1.03424191e+00 3.50030988e-01
6.97647154e-01 -6.25534132e-02 1.11531544e+00 5.61133742e-01
4.39110368e-01 -1.65546775e+00 2.75186300e-01 -1.01448670e-01
5.46636403e-01 -1.78314900e+00 1.65439203e-01 -5.41136742e-01
-6.96022689e-01 9.83251214e-01 7.72609234e-01 -3.62424068e-02
1.11462510e+00 9.80721265e-02 -2.48840123e-01 -1.61461264e-01
-9.51442540e-01 9.96404979e-03 1.08047873e-01 4.67828870e-01
5.66540241e-01 -3.85436304e-02 -7.12807775e-01 8.46012533e-01
7.88190067e-02 2.98087329e-01 2.38492802e-01 8.40184689e-01
-6.19168878e-01 -1.49564064e+00 -6.59225136e-02 9.29083765e-01
-6.14653528e-01 -2.73692369e-01 -1.71629284e-02 6.99858665e-01
-1.95445433e-01 9.02398050e-01 -2.51343131e-01 -4.88486171e-01
6.17884159e-01 7.68625215e-02 3.06153834e-01 -4.95812982e-01
-5.01774311e-01 3.69727969e-01 4.73296225e-01 -4.22797859e-01
1.79052888e-03 -6.12348676e-01 -3.72652590e-01 -2.28658512e-01
-8.56489301e-01 9.20264274e-02 1.15118718e+00 5.93757927e-01
4.75938261e-01 4.30043012e-01 1.07369328e+00 -3.67435515e-01
-1.52558458e+00 -1.17797923e+00 -9.43049908e-01 1.42885029e-01
1.52088687e-01 -9.70448136e-01 -5.16083419e-01 -3.64514589e-01] | [9.856561660766602, 5.470157623291016] |
6d3dfded-bff8-4d76-82d0-8f915f75c202 | context-free-textspotter-for-real-time-and | 2106.05611 | null | https://arxiv.org/abs/2106.05611v1 | https://arxiv.org/pdf/2106.05611v1.pdf | Context-Free TextSpotter for Real-Time and Mobile End-to-End Text Detection and Recognition | In the deployment of scene-text spotting systems on mobile platforms, lightweight models with low computation are preferable. In concept, end-to-end (E2E) text spotting is suitable for such purposes because it performs text detection and recognition in a single model. However, current state-of-the-art E2E methods rely on heavy feature extractors, recurrent sequence modellings, and complex shape aligners to pursue accuracy, which means their computations are still heavy. We explore the opposite direction: How far can we go without bells and whistles in E2E text spotting? To this end, we propose a text-spotting method that consists of simple convolutions and a few post-processes, named Context-Free TextSpotter. Experiments using standard benchmarks show that Context-Free TextSpotter achieves real-time text spotting on a GPU with only three million parameters, which is the smallest and fastest among existing deep text spotters, with an acceptable transcription quality degradation compared to heavier ones. Further, we demonstrate that our text spotter can run on a smartphone with affordable latency, which is valuable for building stand-alone OCR applications. | ['Naoaki Yamashita', 'Takumi Fujino', 'Kenji Doi', 'Tomohiro Tanaka', 'Ryota Yoshihashi'] | 2021-06-10 | null | null | null | null | ['text-spotting'] | ['computer-vision'] | [ 4.49042469e-01 -4.92168814e-01 1.67221785e-01 -1.33232057e-01
-7.67818451e-01 -3.33925426e-01 6.08528018e-01 -2.07267031e-01
-6.07464433e-01 6.22937679e-02 8.43405277e-02 -7.65354753e-01
3.56294245e-01 -3.07319999e-01 -4.50497061e-01 -4.84368503e-01
5.07980049e-01 5.21395266e-01 3.55136722e-01 -3.15718591e-01
4.92190093e-01 1.01126172e-01 -1.49585903e+00 4.00765836e-01
9.13877487e-01 9.82834697e-01 5.06063282e-01 1.16094184e+00
-4.52448636e-01 5.02751410e-01 -5.85677326e-01 -6.34475231e-01
5.94257973e-02 -8.70700553e-02 -2.56500483e-01 -1.55533433e-01
6.02386355e-01 -5.51862895e-01 -4.05239165e-01 6.86355114e-01
8.82618189e-01 -1.93923220e-01 2.93016762e-01 -8.81862164e-01
-3.34139824e-01 5.45476139e-01 -6.94266140e-01 -7.11674318e-02
3.04702282e-01 4.11447793e-01 8.93750787e-01 -1.48280597e+00
2.86626279e-01 8.73929083e-01 9.89592314e-01 4.04362857e-01
-9.42760944e-01 -5.98743439e-01 -8.81437734e-02 -5.76292835e-02
-1.53699684e+00 -9.01758492e-01 3.54716092e-01 -1.86707333e-01
1.38204205e+00 7.63493598e-01 5.31389415e-01 1.11678231e+00
2.22159490e-01 1.18711483e+00 8.86303306e-01 -4.43509489e-01
1.66962966e-01 -1.44654676e-01 -3.89610268e-02 6.77913964e-01
6.30872026e-02 -4.22614038e-01 -1.00244284e+00 7.59520456e-02
4.19720381e-01 -5.34673929e-02 -2.41722450e-01 2.75042742e-01
-1.51458561e+00 3.86384517e-01 -2.07097858e-01 1.95417881e-01
-5.64642996e-02 3.27715605e-01 5.45170307e-01 4.26559746e-01
6.23396337e-01 5.97747676e-02 -1.96322173e-01 -8.70386183e-01
-1.68908679e+00 -2.38639805e-02 7.90006518e-01 1.00937259e+00
5.14132082e-01 -1.84649289e-01 -2.34224811e-01 9.61352885e-01
1.22699134e-01 9.48606670e-01 7.04764783e-01 -1.61242574e-01
9.95470226e-01 2.99396694e-01 -1.64796654e-02 -8.70145202e-01
-2.99267083e-01 -1.07873440e-01 -7.39807785e-01 -5.47655039e-02
2.48902068e-01 -2.09525347e-01 -9.77831602e-01 8.86577547e-01
5.48211336e-02 3.07363600e-01 -4.29802179e-01 9.13902819e-01
5.18782139e-01 7.30504811e-01 -4.61682111e-01 2.16449961e-01
1.45228803e+00 -1.31410611e+00 -7.50043035e-01 -6.06824338e-01
1.07108605e+00 -1.33409739e+00 1.59885871e+00 4.83611465e-01
-8.41936409e-01 -2.62565970e-01 -1.03989100e+00 -4.84505683e-01
-3.02230597e-01 6.39351130e-01 4.97269273e-01 1.10523295e+00
-1.12072706e+00 4.80605036e-01 -1.06205809e+00 -5.30912220e-01
2.28642598e-01 3.16328466e-01 -4.34908457e-02 2.36426964e-01
-6.49331987e-01 6.39331222e-01 -2.28968356e-02 3.34245324e-01
-2.41846815e-01 -4.52772379e-01 -4.20658469e-01 2.77761430e-01
5.22022009e-01 -4.18933421e-01 1.20570052e+00 -5.80136776e-01
-2.06429791e+00 7.00476050e-01 -4.28586215e-01 -3.36783558e-01
8.54401529e-01 -6.33751512e-01 -4.60342258e-01 5.39785512e-02
-2.69652426e-01 2.98787236e-01 1.09258080e+00 -5.90283811e-01
-6.42190576e-01 -1.89973667e-01 -5.78628063e-01 3.53224576e-01
-7.00416863e-01 4.55180228e-01 -1.06411803e+00 -8.53551328e-01
3.44039425e-02 -9.82507885e-01 1.60792798e-01 2.73276746e-01
-5.83512604e-01 -3.17164771e-02 1.25190520e+00 -9.08572495e-01
1.65525651e+00 -2.30393577e+00 -3.80325377e-01 -2.63872109e-02
2.84118414e-01 5.57990670e-01 -4.12662067e-02 5.77599883e-01
2.79042214e-01 6.59062713e-02 -4.94458005e-02 -9.90076065e-01
3.67885709e-01 -2.28359297e-01 -5.86191237e-01 5.36588848e-01
-4.94319946e-02 1.15810287e+00 -4.98716176e-01 -5.95375299e-01
5.08737981e-01 3.67103428e-01 -3.98645431e-01 -3.72580923e-02
-2.37890720e-01 -2.92354256e-01 -1.84975207e-01 6.79373443e-01
6.35102689e-01 -1.98824361e-01 2.67327931e-02 2.50233591e-01
-4.53808397e-01 5.80448210e-01 -1.08009970e+00 1.82795584e+00
-6.95106208e-01 1.29538035e+00 1.64932877e-01 -4.84943628e-01
9.58939016e-01 1.36002377e-01 9.33261588e-02 -7.24340081e-01
2.45960414e-01 5.67591012e-01 -4.89238262e-01 -3.64659607e-01
1.14932942e+00 4.93433475e-01 1.20287709e-01 8.53326619e-01
-4.10261214e-01 -3.57569307e-01 3.25896218e-02 1.23137839e-01
1.18970180e+00 1.43042356e-01 -2.24360362e-01 -1.02864727e-01
-1.13044493e-02 -3.70920390e-01 1.03309110e-01 9.53621745e-01
5.01070619e-02 9.68418777e-01 1.93917960e-01 -4.23277438e-01
-1.19869530e+00 -4.99667585e-01 6.40402511e-02 1.34990168e+00
1.52284011e-01 -9.13827360e-01 -1.01155043e+00 -2.94277638e-01
-5.01538992e-01 5.64506590e-01 -1.53047368e-01 3.16435248e-01
-7.19011843e-01 -7.78893709e-01 9.88632798e-01 4.32769805e-01
5.02166688e-01 -7.44578004e-01 -9.07077670e-01 2.61706918e-01
-2.84554154e-01 -1.21890068e+00 -1.03868997e+00 1.69562086e-01
-8.31926882e-01 -5.12340188e-01 -1.19892442e+00 -6.82461262e-01
3.21512967e-01 6.27898514e-01 8.42976570e-01 2.54902780e-01
-2.64660150e-01 -9.49589908e-02 -4.67670172e-01 -2.67403245e-01
-1.15448885e-01 3.99344862e-01 -1.67894170e-01 1.65790051e-01
4.99116391e-01 -4.79029834e-01 -8.10149014e-01 5.95697582e-01
-8.55002522e-01 6.43972874e-01 6.73759758e-01 6.96607411e-01
1.86289787e-01 -2.09132686e-01 -1.78998917e-01 -4.38084930e-01
5.87076902e-01 9.45969447e-02 -5.91686606e-01 5.39575100e-01
-6.56285763e-01 -1.36128470e-01 7.33479023e-01 -5.73421478e-01
-8.12554181e-01 8.05107784e-03 -3.47397864e-01 -2.49515042e-01
1.66001469e-01 3.86401355e-01 2.01188475e-01 -7.01557919e-02
5.48334837e-01 7.18558550e-01 -1.99275821e-01 -5.91087043e-01
2.93855667e-01 1.25735128e+00 5.08756101e-01 -4.01878536e-01
6.39001727e-01 4.25982475e-01 -4.90316480e-01 -1.28734481e+00
-2.45275557e-01 -6.76695526e-01 -4.40739244e-01 -1.50760382e-01
5.87171674e-01 -7.78077662e-01 -8.07068884e-01 9.46734011e-01
-1.19921231e+00 -7.82101750e-01 1.89192042e-01 2.72232592e-01
-5.08177102e-01 6.19470000e-01 -5.66796422e-01 -9.08408642e-01
-8.29689741e-01 -1.04969025e+00 1.96790898e+00 2.12568846e-02
-1.54869601e-01 -5.78647852e-01 -6.59366548e-02 2.88404524e-01
7.31046915e-01 -5.23389518e-01 4.04467136e-01 -4.06477630e-01
-4.75157201e-01 -1.92737982e-01 -4.89747912e-01 -2.28898868e-01
-1.74952015e-01 1.15677126e-01 -1.07885969e+00 -3.01009029e-01
-1.33214414e-01 -5.10330498e-02 8.92505527e-01 7.40363682e-03
1.23856771e+00 -3.16276193e-01 -4.16860729e-01 9.86107051e-01
1.17917359e+00 3.45266238e-02 9.18925941e-01 3.09454381e-01
9.29027617e-01 2.16438323e-01 3.98895830e-01 5.63150764e-01
2.68116027e-01 1.07853532e+00 -9.33312476e-02 -3.60201955e-01
-2.06746310e-01 -2.76416451e-01 5.34962893e-01 1.24551249e+00
3.83454353e-01 -6.24119639e-01 -1.23699820e+00 3.92293125e-01
-2.00518155e+00 -6.92400932e-01 -2.54448026e-01 2.10906720e+00
8.34120870e-01 2.08766907e-01 6.47358000e-02 2.92227775e-01
8.79698932e-01 3.75161201e-01 -4.55632329e-01 -4.05621022e-01
-1.24018997e-01 3.70252371e-01 6.10649645e-01 3.38726729e-01
-1.01017141e+00 1.25601792e+00 6.05096769e+00 1.42376614e+00
-1.58123624e+00 -1.98739395e-03 6.24511719e-01 -3.32235157e-01
-9.71729085e-02 2.71897204e-02 -9.46899414e-01 8.20791841e-01
1.12284803e+00 1.10439762e-01 5.22314131e-01 7.11289823e-01
3.89359474e-01 -7.34945089e-02 -8.66444170e-01 1.63293481e+00
1.47402629e-01 -1.41883612e+00 -2.48294070e-01 -4.87529896e-02
3.30808908e-01 3.33657980e-01 4.53915782e-02 7.29011558e-03
-2.94228108e-03 -1.04367769e+00 1.00758660e+00 3.65932345e-01
1.28574002e+00 -4.42956626e-01 4.93565440e-01 3.22526664e-01
-1.30394852e+00 2.20248759e-01 -1.91702470e-01 1.18249983e-01
3.78503948e-01 8.82363319e-01 -1.00004458e+00 2.59144515e-01
5.44301450e-01 3.95002306e-01 -7.65886009e-01 1.01925671e+00
8.42489377e-02 8.64491820e-01 -8.32342803e-01 -5.47767401e-01
2.92312413e-01 -5.17661460e-02 4.46567833e-01 1.80655658e+00
7.23811567e-01 -1.88334003e-01 -8.16007704e-02 5.27095318e-01
-1.07348830e-01 2.65817612e-01 -2.72729576e-01 -1.54097408e-01
3.96099746e-01 1.20742249e+00 -1.25017893e+00 -4.04010892e-01
-3.69105428e-01 1.63754499e+00 -4.11264971e-02 1.86668754e-01
-9.53660905e-01 -1.02168572e+00 3.47056985e-01 -2.14816555e-02
4.74086523e-01 -6.21092260e-01 -7.30990946e-01 -1.46527267e+00
4.78517205e-01 -1.12578511e+00 -3.01767498e-01 -8.02780867e-01
-7.04402447e-01 5.66780448e-01 -8.20981383e-01 -1.18230665e+00
4.43945602e-02 -7.57701039e-01 -6.41922832e-01 6.32315993e-01
-1.18606687e+00 -1.10095561e+00 -4.24693793e-01 5.63908041e-01
9.07644868e-01 -1.45976022e-02 5.86692989e-01 4.97886300e-01
-7.27727413e-01 1.10268354e+00 6.26798570e-01 3.04555416e-01
7.74585187e-01 -1.06746912e+00 1.57136548e+00 9.90536630e-01
6.21411860e-01 5.23077846e-01 6.24258757e-01 -7.23557711e-01
-1.93568611e+00 -7.77013123e-01 1.14131176e+00 -3.62025768e-01
8.13469231e-01 -1.12604225e+00 -6.78292572e-01 3.37955713e-01
9.62916613e-02 -1.23691678e-01 4.00440305e-01 -2.53541791e-03
-1.96073174e-01 1.29023641e-01 -4.32709634e-01 1.21397662e+00
1.08265448e+00 -8.78266633e-01 -2.21742555e-01 4.45478946e-01
6.94010019e-01 -6.55964851e-01 -1.86142132e-01 -1.49569750e-01
9.45728242e-01 -1.09121764e+00 6.44159317e-01 2.25637063e-01
2.30666354e-01 -2.36526549e-01 -6.62816912e-02 -7.96767116e-01
2.77649164e-01 -1.54145837e+00 -1.04514964e-01 1.12335241e+00
4.30824369e-01 -5.90716720e-01 9.90339220e-01 5.02975643e-01
-2.28003874e-01 -7.28518546e-01 -1.07283247e+00 -8.81695747e-01
-3.59050721e-01 -8.05707037e-01 6.81181490e-01 6.22809112e-01
1.91613343e-02 2.22763136e-01 -7.80824125e-01 -2.59367913e-01
2.71731108e-01 1.16525091e-01 1.01086688e+00 -7.62348831e-01
-4.32671756e-01 -6.65447295e-01 -1.20709270e-01 -1.91884422e+00
-2.38563538e-01 -5.47555923e-01 4.57713991e-01 -1.10039508e+00
5.59709407e-02 -5.40957928e-01 3.71087939e-01 5.01853168e-01
-3.90568405e-01 4.08000708e-01 4.24876392e-01 3.74888897e-01
-7.52034187e-01 6.87367141e-01 7.20131576e-01 -5.08382432e-02
-2.57210791e-01 -1.65443733e-01 -2.86431849e-01 5.88066638e-01
4.95468110e-01 -3.29086334e-01 -4.78045270e-02 -9.05665517e-01
4.30317789e-01 -5.14608659e-02 8.31102505e-02 -1.07720649e+00
7.78051734e-01 2.19584480e-01 -7.97424018e-02 -8.84740472e-01
3.72213632e-01 -5.60633779e-01 -1.08423308e-01 1.64513588e-01
-2.54301757e-01 1.68341354e-01 1.17069304e-01 2.05770969e-01
4.47058231e-02 -2.82973558e-01 3.38328481e-01 4.46727306e-01
-4.36340272e-01 1.94809139e-01 -5.17603040e-01 -5.97014949e-02
4.64336723e-01 -4.40113276e-01 -4.90356117e-01 -4.34549153e-01
-1.70663632e-02 -4.82969768e-02 6.27662778e-01 3.61190349e-01
6.34843528e-01 -1.04039741e+00 -6.77301586e-01 2.09557891e-01
4.67240177e-02 -1.29450500e-01 -4.86602075e-02 8.91213953e-01
-9.68239963e-01 5.43899655e-01 3.17562461e-01 -7.38140643e-01
-1.41941929e+00 3.10118109e-01 1.80746503e-02 -2.77409494e-01
-1.10661304e+00 8.93289864e-01 5.43072112e-02 -2.31085941e-01
3.59554470e-01 -3.91620487e-01 4.56693053e-01 -9.75300372e-02
9.47056532e-01 3.65386814e-01 6.06560051e-01 -2.61848599e-01
-3.27242613e-01 8.08265030e-01 -2.04816237e-01 -4.72931355e-01
1.05274987e+00 -3.32487494e-01 1.59275964e-01 3.16839576e-01
1.01676929e+00 1.40165612e-01 -1.22199345e+00 -1.06624268e-01
1.96338832e-01 -6.57129109e-01 2.79728413e-01 -6.92951858e-01
-6.96893692e-01 1.11403811e+00 5.76228738e-01 8.05173814e-03
1.16247928e+00 -5.42097867e-01 1.42895758e+00 7.57307351e-01
3.16410005e-01 -1.41691363e+00 -1.95234944e-03 6.71289861e-01
6.55551016e-01 -1.00808299e+00 -1.06347851e-01 -2.88683236e-01
-4.49662089e-01 1.23906767e+00 1.35640606e-01 1.80947393e-01
2.40777776e-01 8.57262313e-01 8.23130310e-02 -3.52435932e-02
-8.44842255e-01 -6.80957809e-02 2.89618343e-01 2.35185713e-01
3.66542459e-01 -3.33805978e-02 -9.23158228e-02 1.88568488e-01
-4.72113341e-01 -5.49061187e-02 3.62577856e-01 8.05933595e-01
-4.19098228e-01 -9.88873899e-01 -4.15236741e-01 3.83370608e-01
-5.28150260e-01 -8.30798030e-01 -5.09733915e-01 4.27316338e-01
-5.59069812e-01 1.07275212e+00 6.58199787e-02 -7.02720940e-01
1.88867922e-03 5.31190634e-02 7.86917359e-02 -3.28031480e-01
-7.09095895e-01 3.44428182e-01 9.43961069e-02 -7.00261176e-01
2.44364068e-01 -2.50628531e-01 -8.16794991e-01 -7.71157265e-01
-6.63441002e-01 -3.31159562e-01 1.10109365e+00 1.03709936e+00
7.17429519e-01 4.42503504e-02 5.03152251e-01 -1.08196890e+00
-3.52424711e-01 -9.17104185e-01 -2.87006795e-01 -7.92783424e-02
3.35098922e-01 -5.30263744e-02 -2.45713294e-01 2.72499453e-02] | [11.974152565002441, 2.2451441287994385] |
6e7ddb20-a161-4583-b9fd-94e5a73139e7 | multi-exposure-image-fusion-based-on-exposure | 1806.09607 | null | http://arxiv.org/abs/1806.09607v1 | http://arxiv.org/pdf/1806.09607v1.pdf | Multi-Exposure Image Fusion Based on Exposure Compensation | This paper proposes a novel multi-exposure image fusion method based on
exposure compensation. Multi-exposure image fusion is a method to produce
images without color saturation regions, by using photos with different
exposures. However, in conventional works, it is unclear how to determine
appropriate exposure values, and moreover, it is difficult to set appropriate
exposure values at the time of photographing due to time constraints. In the
proposed method, the luminance of the input multi-exposure images is adjusted
on the basis of the relationship between exposure values and pixel values,
where the relationship is obtained by assuming that a digital camera has a
linear response function. The use of a local contrast enhancement method is
also considered to improve input multi-exposure images. The compensated images
are finally combined by one of existing multi-exposure image fusion methods. In
some experiments, the effectiveness of the proposed method are evaluated in
terms of the tone mapped image quality index, statistical naturalness, and
discrete entropy, by comparing the proposed one with conventional ones. | ['Sayaka Shiota', 'Hitoshi Kiya', 'Yuma Kinoshita', 'Taichi Yoshida'] | 2018-06-23 | null | null | null | null | ['multi-exposure-image-fusion'] | ['computer-vision'] | [ 8.7475991e-01 -6.6748369e-01 3.4477600e-01 -2.1295339e-01
-3.0202448e-01 -3.8296840e-01 2.5073913e-01 1.6298585e-02
-5.9660548e-01 7.7134860e-01 -9.9835172e-03 -4.0327787e-02
-4.2335767e-01 -9.2056155e-01 -3.1864068e-01 -7.9511076e-01
5.0272536e-01 -4.1100556e-01 2.0535339e-01 -3.0212882e-01
4.9891284e-01 2.8714254e-01 -1.6368148e+00 -2.2253567e-02
1.0994332e+00 1.0764362e+00 3.9735892e-01 6.5917265e-01
2.1190243e-02 4.7106406e-01 -7.6705962e-01 -2.2286770e-01
4.0399429e-01 -1.0603328e+00 -4.9094790e-01 5.1389843e-01
3.2447442e-01 -4.2293844e-01 -6.2773690e-02 1.5172993e+00
2.8737369e-01 3.2630548e-01 6.4774013e-01 -9.3713397e-01
-8.8394928e-01 2.0227617e-01 -7.3574221e-01 3.5779217e-01
2.9893756e-01 9.4337603e-03 2.4867290e-01 -4.1528374e-01
2.2143078e-01 1.0379845e+00 2.6752463e-01 -5.3417936e-02
-1.2550501e+00 -4.2892441e-01 -2.3082660e-01 3.1641579e-01
-1.5169847e+00 -2.5194600e-01 8.9875335e-01 -2.5838935e-01
1.8153420e-01 4.4663513e-01 5.7143939e-01 1.2681440e-01
7.1356523e-01 -2.1712919e-01 1.9832617e+00 -9.7200614e-01
3.7444334e-02 4.7988603e-01 9.8244729e-04 3.5863581e-01
1.4123063e-01 1.5227567e-01 -3.5039052e-02 1.0473064e-01
5.9129041e-01 1.1649050e-01 -7.2438520e-01 1.2909245e-01
-8.7779617e-01 5.0782591e-01 3.3268899e-01 7.9496735e-01
-4.2832690e-01 -3.4126258e-01 -5.2123375e-02 3.2878396e-01
2.6555169e-01 1.7891262e-01 2.1105038e-01 1.6717337e-01
-1.0634656e+00 -2.4476956e-01 4.5788601e-01 4.7215909e-01
9.0865374e-01 -1.8413141e-02 5.4433118e-03 6.2204063e-01
2.4729142e-01 6.5090716e-01 5.8675563e-01 -8.5822052e-01
1.6115788e-01 3.1642008e-01 1.5956967e-01 -1.1011255e+00
6.8886429e-02 2.5450572e-02 -7.9162657e-01 7.3629898e-01
3.1076190e-01 -2.2392380e-01 -8.3111387e-01 1.5382798e+00
1.1907095e-01 -1.6964760e-02 5.1623899e-01 1.0257862e+00
4.7850144e-01 9.8008287e-01 -3.3663176e-02 -9.4852179e-01
1.2792519e+00 -6.5727204e-01 -1.3602839e+00 1.2966943e-01
-3.2377380e-01 -1.2277814e+00 9.9095076e-01 7.0075393e-01
-1.1552949e+00 -9.0608060e-01 -1.4952593e+00 1.9601232e-01
-3.7850034e-01 9.1914654e-02 -1.1694694e-01 9.0935248e-01
-9.9669999e-01 4.4498336e-01 -1.8709129e-01 -2.7843228e-01
-2.6305404e-01 1.7736222e-01 -1.7753433e-01 -1.1979202e-01
-1.3472873e+00 1.2054843e+00 4.2987102e-01 1.6512416e-01
-2.0776784e-01 -3.4403840e-01 -6.0901731e-01 3.9814647e-02
6.5033056e-02 -2.3328453e-01 8.4111083e-01 -1.4361870e+00
-1.6984680e+00 4.1161132e-01 -8.5363985e-04 -1.6450584e-01
4.2212912e-01 8.4494412e-02 -8.3289748e-01 5.2602541e-01
-3.4760720e-01 2.1631488e-01 8.6472648e-01 -1.3851447e+00
-5.8497185e-01 -8.4108941e-02 2.2196408e-01 5.0717515e-01
-1.8885595e-01 9.3329727e-04 -3.8069490e-01 -2.4465713e-01
3.8563687e-02 -6.1692601e-01 -8.3765090e-02 -4.6920730e-03
-9.6054353e-02 4.5465431e-01 1.1161836e+00 -9.6921080e-01
1.2058120e+00 -2.4069972e+00 -2.3031789e-01 5.0540739e-01
-3.6046699e-01 3.0435205e-01 2.2475016e-01 2.6471218e-01
-4.5658618e-02 -3.7374131e-02 -5.4983538e-01 1.6040573e-01
-3.2719451e-01 -2.6166993e-01 4.4233717e-02 5.4105699e-01
-1.5896754e-01 5.3541530e-02 -4.6448100e-01 -8.9706337e-01
8.5876703e-01 7.9351503e-01 2.4238034e-01 1.5013048e-01
2.4687365e-01 3.8133720e-01 -1.2080144e-03 3.3551261e-01
1.0913517e+00 1.9700080e-01 2.6195934e-01 -5.6258607e-01
-4.4523880e-01 -5.2947611e-01 -1.4582067e+00 1.0754547e+00
-6.1061746e-01 7.2015429e-01 -1.0175034e-01 -5.4946148e-01
1.1405324e+00 5.0587416e-01 4.0593472e-01 -8.7209147e-01
4.9040633e-01 1.6945221e-01 1.1826838e-02 -4.8880410e-01
7.1537536e-01 -2.0036243e-01 3.1965673e-01 3.0735168e-01
-4.6569446e-01 -2.9681835e-01 5.2428728e-01 -1.7229110e-01
2.5905868e-01 3.2392181e-02 5.3698963e-01 -1.5815549e-01
9.1772544e-01 -3.1443283e-01 2.9883787e-01 2.7986035e-01
-1.2501869e-01 5.5923897e-01 5.9514489e-02 -3.1663235e-02
-1.1012417e+00 -8.1580108e-01 -4.6523154e-01 2.8038800e-01
8.8331294e-01 2.0754021e-01 -9.5819592e-01 7.8805275e-02
-4.7051895e-01 6.5438390e-01 -2.3379889e-01 -2.0260307e-01
-3.8257954e-01 -7.9036790e-01 7.3457532e-02 -1.8556030e-01
1.1827984e+00 -1.0219779e+00 -8.6861575e-01 1.2258234e-01
-3.6700997e-01 -9.3102920e-01 -4.8699042e-01 -2.1671240e-01
-6.3378507e-01 -9.4247484e-01 -7.3051435e-01 -8.0036372e-01
8.1074506e-01 5.4670483e-01 4.4080240e-01 4.5595512e-01
-9.9029176e-02 1.1575354e-01 -4.7389778e-01 -1.2548022e-01
-6.5878195e-01 -4.7126213e-01 -2.4323009e-01 3.7792489e-01
1.3211146e-01 -3.6996064e-01 -7.6796561e-01 2.6710343e-01
-1.3706775e+00 2.5678702e-02 7.8968638e-01 4.4604981e-01
6.4218605e-01 6.5558195e-01 3.6405104e-01 -5.0701845e-01
7.4609613e-01 3.0411370e-02 -8.7692487e-01 7.1924531e-01
-8.3851534e-01 -1.9821915e-01 7.4662995e-01 -4.5159397e-01
-1.5143660e+00 -1.4371701e-01 1.8579525e-01 -1.0027038e-01
-1.7944691e-01 1.9529828e-01 -3.2035109e-01 -5.7123482e-01
3.6321038e-01 3.9424774e-01 2.1182209e-01 -1.3235562e-01
2.6795161e-01 9.4017339e-01 6.3146609e-01 7.6907769e-02
6.5871930e-01 2.4744597e-01 1.4086436e-01 -8.7736022e-01
-1.4434692e-01 2.0219095e-02 -3.8248768e-01 -6.1889464e-01
1.1275303e+00 -6.2217343e-01 -7.0091051e-01 6.7166120e-01
-8.9164686e-01 1.6339332e-01 8.6719923e-02 8.4454608e-01
-3.5972211e-01 7.7705353e-01 -3.9892724e-01 -9.5789528e-01
-3.9237297e-01 -1.3107648e+00 3.9961448e-01 8.2032853e-01
1.6996935e-01 -1.0257148e+00 -2.3304726e-01 1.3618222e-01
6.1620581e-01 3.4691140e-01 7.9693741e-01 2.1705432e-01
-5.5402493e-01 9.8513536e-02 -1.3718303e-01 6.7785460e-01
5.6716251e-01 2.4945694e-01 -7.2732240e-01 -2.1731785e-01
3.3446181e-01 -8.9375880e-03 4.9039933e-01 6.0567331e-01
1.0600905e+00 -3.6587320e-02 1.8862835e-03 5.9330523e-01
2.2474492e+00 8.1567568e-01 1.2533822e+00 4.0552726e-01
-1.3205810e-01 3.3072224e-01 8.6566877e-01 1.8160373e-01
-2.6674785e-02 5.2399588e-01 1.8022351e-01 -5.5551237e-01
-2.3127133e-01 1.9563965e-01 1.9105652e-01 5.4388189e-01
-6.5998663e-03 -6.0480106e-01 -1.1420028e-01 3.2572165e-01
-1.2709627e+00 -1.1252925e+00 -2.8658813e-01 2.3526797e+00
7.8375411e-01 1.4185901e-01 -3.5198462e-01 6.5959847e-01
1.2257406e+00 -9.1056200e-03 -1.7975172e-01 -8.0104125e-01
-3.9778736e-01 2.2170582e-01 6.4186484e-01 9.5028448e-01
-8.7002856e-01 1.8520533e-01 6.3904200e+00 7.0737100e-01
-1.5317664e+00 -1.7735999e-02 5.6612444e-01 3.5076818e-01
-2.9737857e-01 1.8965413e-01 -4.3192890e-01 9.0232271e-01
5.2743334e-01 -3.7488526e-01 4.9609429e-01 9.6417025e-02
3.8068700e-01 -8.8740492e-01 -3.5299659e-01 9.0813398e-01
3.8959670e-01 -6.6435903e-01 -1.3920844e-01 -4.4766273e-02
7.8780848e-01 -1.0565225e+00 3.5589257e-01 -4.4771156e-01
-3.9332113e-01 -5.0597644e-01 5.5917925e-01 9.6607554e-01
6.4860773e-01 -7.6059949e-01 8.6360133e-01 -9.1274688e-03
-1.1723186e+00 5.4925336e-03 -2.0256436e-01 3.9578494e-02
4.9533784e-01 4.7081161e-01 -3.0638176e-01 8.2035506e-01
4.5352340e-01 -5.9003636e-02 -6.3990098e-01 1.2513683e+00
-1.0860998e-01 2.9743293e-01 -2.4906369e-01 9.8199368e-02
-2.1119406e-02 -7.9761237e-01 2.0509239e-01 9.2658782e-01
6.9897616e-01 3.7063903e-01 -1.8496895e-01 9.1980672e-01
2.3826553e-01 4.1162536e-01 -3.5494366e-01 2.5760239e-01
4.8036063e-01 1.2524059e+00 -8.3224779e-01 -5.0414735e-01
-4.3701005e-01 9.4659013e-01 -6.4772260e-01 4.7874555e-01
-9.3183374e-01 -8.1797063e-01 -1.0047592e-01 -1.9659743e-01
1.6559571e-01 -4.6801709e-02 -2.4670599e-01 -5.5973202e-01
-7.0683183e-03 -7.7602416e-01 1.4773962e-01 -1.1976392e+00
-7.6390463e-01 6.5560251e-01 2.7900651e-01 -1.4136183e+00
7.4635752e-02 -2.0236081e-01 -6.4051592e-01 1.3317554e+00
-1.5142385e+00 -7.7241379e-01 -4.9118271e-01 5.9112710e-01
3.2256329e-01 1.3293625e-01 4.4678247e-01 5.1385105e-01
-2.5810611e-01 4.0705201e-01 3.6045364e-01 -4.3612382e-01
7.8879911e-01 -1.0297973e+00 -5.0572532e-01 1.3350289e+00
-2.8893155e-01 2.3966169e-01 9.0976632e-01 -6.7386991e-01
-7.1068650e-01 -5.6251222e-01 6.3774401e-01 4.3969291e-01
-1.4221835e-01 2.4195901e-01 -1.0144657e+00 1.0888229e-01
9.4377321e-01 -3.5731149e-01 5.9867316e-01 -8.2511961e-01
3.9261261e-01 -5.1950204e-01 -1.5988088e+00 3.5095963e-01
1.3247402e-01 -3.1062844e-01 -5.4155636e-01 -1.0165598e-01
5.8141738e-01 -1.5955365e-01 -1.0843527e+00 3.0052775e-01
5.8411467e-01 -1.0538678e+00 7.0690382e-01 5.9816188e-01
2.6120582e-01 -8.6179048e-01 -2.1343681e-01 -1.2839748e+00
-2.9083833e-01 -2.9087192e-01 7.2817147e-01 1.3757789e+00
3.5461116e-01 -5.7691449e-01 2.9582668e-02 4.4852364e-01
1.6120872e-01 -2.2426865e-01 -4.4277558e-01 -4.8880580e-01
-4.9673423e-01 4.5612511e-01 7.3662400e-01 6.7746824e-01
-1.1201897e-01 5.0053369e-02 -5.5986488e-01 4.5607537e-01
7.5231445e-01 1.9509873e-01 4.4595483e-01 -7.4944037e-01
-1.7183951e-01 -1.9729555e-01 -2.9557428e-01 -4.9935776e-01
-3.7640190e-01 -1.2138419e-01 6.0566846e-02 -1.5786200e+00
1.4253460e-01 -6.7658000e-02 -4.2896616e-01 -1.2786938e-01
-3.6603868e-01 6.0828453e-01 4.4865090e-01 1.8020324e-01
-6.3293047e-02 2.3679660e-01 1.4881901e+00 -7.5183935e-02
-2.3142609e-01 -1.8543877e-01 -5.2500665e-01 4.0523934e-01
8.4006226e-01 -8.5247666e-02 -6.3610953e-01 -3.0850127e-01
-1.7047530e-01 2.5229561e-01 2.5132611e-01 -1.3332486e+00
2.2965252e-01 -4.0473565e-01 5.4453814e-01 -5.9415835e-01
4.2503217e-01 -1.1802400e+00 6.9798380e-01 5.0542140e-01
-2.4079254e-01 2.3003735e-01 -4.6250809e-02 2.8202507e-01
-3.6923227e-01 -5.6676918e-01 1.0922407e+00 -5.6528334e-02
-8.7816346e-01 -2.5855634e-01 -4.0376109e-01 -7.0951515e-01
1.3340416e+00 -7.2802615e-01 -3.2030430e-01 -4.4071129e-01
-4.2583901e-01 -3.5322326e-01 7.6517302e-01 1.5352488e-02
6.7927605e-01 -1.2736347e+00 -3.5022828e-01 2.4227858e-01
-2.1365577e-01 -6.6786712e-01 6.2499183e-01 9.0459603e-01
-8.7246168e-01 -2.0248578e-01 -6.7242795e-01 -2.5441131e-01
-1.5405607e+00 7.6597166e-01 3.7284088e-01 1.4776780e-01
-2.7258953e-01 2.7200673e-02 -4.1925091e-02 5.4699606e-01
-1.6039707e-01 -2.1578979e-01 -5.3180867e-01 -3.3041984e-01
6.9154084e-01 4.7957397e-01 -6.2419888e-02 -7.5346816e-01
9.8307338e-03 1.0649796e+00 3.2028869e-01 -5.4881603e-01
7.0669144e-01 -6.9249070e-01 -3.1536829e-01 3.8326046e-01
1.1125020e+00 1.4146486e-01 -9.9082810e-01 6.0901251e-02
-5.8749253e-01 -9.6169591e-01 1.7475529e-01 -8.8415122e-01
-8.9151186e-01 7.2718537e-01 1.3141489e+00 3.4104428e-01
1.8651080e+00 -7.0635092e-01 6.9288427e-01 -6.4932592e-02
8.2987249e-02 -1.3427918e+00 -1.3375124e-01 -2.8394845e-01
6.1364228e-01 -1.1999568e+00 3.0386224e-01 -5.0618154e-01
-4.1856804e-01 1.2952725e+00 5.7181853e-01 8.5201740e-02
5.6260890e-01 1.0774532e-01 2.6855400e-01 2.3975229e-01
-2.1157700e-01 -1.7453066e-01 1.2534340e-01 4.6157947e-01
3.2394126e-01 -1.4106975e-01 -1.1796303e+00 -9.2429660e-02
7.0698131e-03 1.9503905e-01 9.1746157e-01 7.9416049e-01
-7.9971403e-01 -1.0091941e+00 -9.9005264e-01 1.0373602e-01
-6.6815859e-01 1.0457053e-01 2.2098845e-02 7.9979628e-01
6.0854894e-01 1.4247017e+00 3.0107738e-02 -2.5148401e-01
2.7130568e-01 -2.7917814e-01 5.9029758e-01 1.7730564e-01
-2.9641265e-01 8.1919849e-02 -3.0418763e-01 2.0000930e-01
-9.9295270e-01 -3.1760627e-01 -1.0183249e+00 -3.9287055e-01
-6.4975983e-01 4.0684655e-01 9.2089760e-01 8.5879654e-01
-3.0404636e-01 5.1951528e-01 1.0429307e+00 -3.3320180e-01
-2.3917931e-01 -9.3698794e-01 -7.0354986e-01 5.2572477e-01
3.2968968e-01 -3.8311970e-01 -5.6596112e-01 4.6773279e-01] | [10.936078071594238, -2.454396963119507] |
52cd1f36-ec19-45f2-ab86-5f868c1f67c6 | a-divide-and-conquer-approach-for-multi-label | null | null | https://aclanthology.org/2021.findings-emnlp.412 | https://aclanthology.org/2021.findings-emnlp.412.pdf | A Divide-And-Conquer Approach for Multi-label Multi-hop Relation Detection in Knowledge Base Question Answering | Relation detection in knowledge base question answering, aims to identify the path(s) of relations starting from the topic entity node that is linked to the answer node in knowledge graph. Such path might consist of multiple relations, which we call multi-hop. Moreover, for a single question, there may exist multiple relation paths to the correct answer, which we call multi-label. However, most of existing approaches only detect one single path to obtain the answer without considering other correct paths, which might affect the final performance. Therefore, in this paper, we propose a novel divide-and-conquer approach for multi-label multi-hop relation detection (DC-MLMH) by decomposing it into head relation detection and conditional relation path generation. In specific, a novel path sampling mechanism is proposed to generate diverse relation paths for the inference stage. A majority-vote policy is employed to detect final KB answer. Comprehensive experiments were conducted on the FreebaseQA benchmark dataset. Experimental results show that the proposed approach not only outperforms other competitive multi-label baselines, but also has superiority over some state-of-art KBQA methods. | ['Yunbo Cao', 'Qian-Wen Zhang', 'Chenchen Ye', 'Linhai Zhang', 'Yanzheng Xiang', 'Deyu Zhou'] | null | null | null | null | findings-emnlp-2021-11 | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-7.31411725e-02 3.93594414e-01 -2.88107872e-01 -3.26373369e-01
-1.06109071e+00 -5.08478105e-01 3.84107292e-01 4.54151869e-01
-1.14012614e-01 1.06644678e+00 1.12625711e-01 -4.70783263e-01
-2.76464075e-01 -1.17490530e+00 -6.16367280e-01 -5.49609601e-01
3.90817314e-01 8.82822096e-01 1.03168976e+00 -2.82663912e-01
-1.21713929e-01 1.77088246e-01 -9.41917479e-01 4.76676971e-01
1.03026295e+00 7.33479083e-01 -1.15970016e-01 3.13830942e-01
-4.20631945e-01 1.12458515e+00 -6.39681339e-01 -8.54584634e-01
-3.32153112e-01 -6.16844475e-01 -1.38432395e+00 -2.38070488e-01
2.35225394e-01 -2.53433943e-01 -1.29487991e-01 1.09859228e+00
4.10965711e-01 1.34276226e-01 6.10370815e-01 -1.22060299e+00
-3.69785786e-01 8.62325728e-01 -7.10922658e-01 3.38879108e-01
6.71484649e-01 -4.29916710e-01 1.54033518e+00 -5.37643552e-01
7.76311219e-01 1.64690113e+00 2.39425629e-01 1.68232784e-01
-8.11299801e-01 -6.98899686e-01 2.52391964e-01 6.46504045e-01
-1.56197739e+00 -1.45708010e-01 5.12479365e-01 -1.26598388e-01
7.73134589e-01 3.06373268e-01 2.86166698e-01 4.40108448e-01
3.53223383e-02 7.60175467e-01 9.46755350e-01 -4.34921026e-01
-6.79116994e-02 1.64952978e-01 4.82876360e-01 9.08410907e-01
2.58691669e-01 -6.14345789e-01 -2.81320661e-01 -5.64464629e-01
4.03785884e-01 -6.21293485e-01 -5.13891459e-01 -9.25183445e-02
-1.08107781e+00 8.65252018e-01 6.02597713e-01 1.36295021e-01
-3.05508852e-01 -1.04677856e-01 1.90480754e-01 2.10112408e-01
1.60403296e-01 1.27780020e-01 -5.74271500e-01 3.70372593e-01
-2.68097639e-01 5.46802998e-01 1.02342939e+00 1.04181862e+00
1.01089096e+00 -1.04835355e+00 -4.82410818e-01 6.92802846e-01
5.75462759e-01 1.40900254e-01 9.01457891e-02 -7.92493403e-01
9.10476744e-01 1.09228694e+00 1.74641371e-01 -1.19609797e+00
-3.87225419e-01 -3.07842195e-01 -6.57850862e-01 -7.18907893e-01
3.56950641e-01 -2.05472514e-01 -6.06052041e-01 1.40295088e+00
1.22399378e+00 2.02131584e-01 2.85962999e-01 1.00588989e+00
1.47149432e+00 8.77789497e-01 2.03487948e-01 -4.62559670e-01
1.94570267e+00 -1.21896887e+00 -8.99819076e-01 -1.06681913e-01
8.10189843e-01 -7.44303405e-01 6.31474733e-01 3.26840393e-02
-7.27502823e-01 -2.10616037e-01 -8.22789907e-01 -1.91195890e-01
-1.18483566e-01 2.00076699e-02 5.74008048e-01 3.87451440e-01
-5.57989955e-01 -2.53136605e-02 -3.87013346e-01 -3.30311239e-01
1.06031202e-01 6.95942864e-02 -2.77904570e-01 -3.98365527e-01
-1.80992532e+00 9.32008445e-01 8.64630818e-01 1.58232942e-01
-5.82496226e-01 -2.71769315e-01 -5.21414638e-01 1.37099940e-02
1.14123487e+00 -9.13298488e-01 1.36644030e+00 -1.77792147e-01
-1.06051648e+00 5.79492748e-01 -6.10753357e-01 -1.14834890e-01
1.70496449e-01 -5.46176434e-02 -6.46259785e-01 4.44141239e-01
2.75880277e-01 4.03628320e-01 2.74277925e-01 -1.35806906e+00
-1.02778018e+00 -2.29707807e-01 5.96592665e-01 5.23017824e-01
1.56359479e-01 2.41911307e-01 -8.58375072e-01 -1.94103748e-03
4.87584203e-01 -7.42527306e-01 -7.90563524e-02 -7.20869958e-01
-9.10546720e-01 -8.58274341e-01 5.82819760e-01 -7.06905365e-01
1.59075165e+00 -1.55600405e+00 1.68961659e-02 3.42689961e-01
2.50829250e-01 1.77142948e-01 2.50791043e-01 3.96506101e-01
1.69652745e-01 1.28199294e-01 -1.03147216e-01 3.10067862e-01
-3.19734305e-01 2.96561152e-01 -3.04027677e-01 7.34294429e-02
1.08656183e-01 9.90805089e-01 -1.01142859e+00 -1.22907138e+00
-4.41295743e-01 1.38577940e-02 -2.13265404e-01 1.05504416e-01
-6.88405693e-01 4.69320893e-01 -9.60302234e-01 7.42839098e-01
7.25313008e-01 -5.87111533e-01 3.57025713e-01 -4.38179016e-01
2.71050692e-01 6.21008396e-01 -1.30495131e+00 1.23434818e+00
-2.05947280e-01 -1.63875908e-01 -8.89284909e-02 -6.46921337e-01
8.99958611e-01 5.58855116e-01 -1.18117249e-02 -4.68296707e-01
4.47290950e-04 4.22741920e-01 2.18887269e-01 -7.13606000e-01
4.54538345e-01 -8.63609016e-02 7.40258843e-02 3.48605901e-01
-7.90078118e-02 2.47992441e-01 4.78898853e-01 4.92840469e-01
1.14513600e+00 5.08314855e-02 4.42057252e-01 1.54870916e-02
1.04644811e+00 2.64617413e-01 9.63405311e-01 4.86785680e-01
-5.18275946e-02 1.54659152e-01 1.08006597e+00 7.69814476e-02
-9.59491059e-02 -8.56956005e-01 7.48925358e-02 9.09577489e-01
6.66914821e-01 -6.04026675e-01 -5.44322908e-01 -1.25924683e+00
-1.48625299e-01 6.29549384e-01 -3.31892848e-01 3.51444893e-02
-8.67677391e-01 -7.43288159e-01 4.59187418e-01 2.36046854e-02
7.34348297e-01 -1.20618367e+00 -1.02017067e-01 4.53953028e-01
-9.45244610e-01 -1.33151817e+00 -2.33539596e-01 -2.36467257e-01
-7.24373341e-01 -1.49862754e+00 -3.56504619e-01 -9.05889332e-01
5.00434101e-01 3.48942995e-01 1.20630836e+00 2.42614493e-01
2.20733404e-01 -8.09582025e-02 -6.21746719e-01 -5.67665175e-02
-2.80482590e-01 5.05966008e-01 -5.78648150e-01 4.45650667e-02
5.66425025e-01 -3.45168531e-01 -6.91194534e-01 5.23553789e-01
-6.30639195e-01 -1.51133016e-01 7.82066643e-01 5.11962771e-01
8.55290651e-01 5.20592391e-01 8.89118791e-01 -1.23778522e+00
8.34095120e-01 -7.63316751e-01 -2.91575581e-01 9.65833843e-01
-5.57678401e-01 1.04457103e-01 4.11500424e-01 -1.23947524e-01
-1.44056058e+00 -2.65318513e-01 -1.44394219e-01 2.24094108e-01
-4.91711535e-02 8.64021659e-01 -6.83972716e-01 2.16496393e-01
3.71432811e-01 -6.56741410e-02 -5.56105256e-01 -4.00482923e-01
6.27409875e-01 5.84000945e-01 2.19358459e-01 -7.23237753e-01
5.78887522e-01 1.97105035e-01 1.06020607e-01 -3.24337751e-01
-1.17965901e+00 -6.30549073e-01 -4.87835944e-01 -1.34265646e-01
8.84185374e-01 -8.12514484e-01 -7.61583209e-01 1.03352256e-01
-1.46598041e+00 2.10796386e-01 3.04246604e-01 4.02035087e-01
7.13246688e-02 2.91538864e-01 -7.65375257e-01 -6.07649624e-01
-5.08914292e-01 -1.01103306e+00 8.07205677e-01 6.73896372e-01
-1.64706185e-01 -7.71966219e-01 1.15542807e-01 8.20569456e-01
-2.40000978e-01 9.21147503e-03 1.43571830e+00 -8.97405088e-01
-9.68430161e-01 -3.42630185e-02 -5.39372921e-01 -1.98904276e-01
-9.73573793e-03 -1.78420559e-01 -5.93221366e-01 2.05912545e-01
-3.65502656e-01 -3.95594478e-01 8.23769867e-01 -7.63617754e-02
5.03047228e-01 -3.73612195e-01 -7.83045650e-01 -7.01344535e-02
1.13828170e+00 2.55517513e-01 5.40042996e-01 3.77483696e-01
9.18897927e-01 9.25992608e-01 1.14816785e+00 -2.18793243e-01
1.04484427e+00 6.62102461e-01 2.19926387e-01 1.67651132e-01
-1.13773897e-01 -3.22108626e-01 -9.97670963e-02 9.71752942e-01
1.13978021e-01 -3.65689874e-01 -8.08680236e-01 5.21238267e-01
-2.00682640e+00 -5.99854290e-01 -7.59578109e-01 1.79314053e+00
1.09870160e+00 1.62864149e-01 -2.80021634e-02 8.19729939e-02
9.49108541e-01 4.80853692e-02 -4.83257592e-01 -1.75429750e-02
-1.35387219e-02 1.13017067e-01 5.27022630e-02 6.68046594e-01
-9.00476038e-01 1.05171013e+00 4.41205835e+00 1.14969146e+00
-4.10270870e-01 3.92602801e-01 4.69432324e-01 5.35817385e-01
-5.57104826e-01 4.00313139e-01 -1.29044104e+00 3.57115492e-02
5.18501163e-01 -1.99075177e-01 -7.46068880e-02 5.31585872e-01
-2.17853546e-01 -4.22126681e-01 -6.99328005e-01 3.98341149e-01
-1.74027503e-01 -8.95077765e-01 1.94912374e-01 -1.67692944e-01
6.33390963e-01 -4.96659368e-01 -6.84760213e-01 4.10148412e-01
4.05212522e-01 -6.47166014e-01 3.03945571e-01 3.93235117e-01
3.52772206e-01 -9.29304242e-01 9.16238427e-01 5.70445120e-01
-1.66209662e+00 2.18851253e-01 -3.50626379e-01 4.25394446e-01
3.74155402e-01 7.94577241e-01 -8.90519679e-01 1.26710856e+00
5.15629113e-01 1.97546259e-01 -4.62862104e-01 1.01713753e+00
-8.89826238e-01 7.15290010e-01 -1.43911585e-01 -3.16278368e-01
1.24675110e-01 -2.27975130e-01 3.35839003e-01 9.57372248e-01
1.68941930e-01 5.42793453e-01 2.07676888e-01 5.36460519e-01
-3.25487226e-01 5.71880162e-01 -1.18860099e-02 3.40460420e-01
8.95407200e-01 1.45330143e+00 -8.01320851e-01 -3.75071555e-01
-4.94458348e-01 7.02063441e-01 6.02240264e-01 3.96019608e-01
-7.76938081e-01 -5.61151505e-01 1.28320139e-02 -9.22163799e-02
2.58323431e-01 2.93026060e-01 2.12530583e-01 -9.30640697e-01
2.25835547e-01 -8.69292080e-01 1.00455928e+00 -6.56739354e-01
-1.11543190e+00 4.89288658e-01 1.24664821e-01 -7.12310910e-01
5.38667850e-03 1.41855583e-01 -5.92661560e-01 9.89067793e-01
-1.76701844e+00 -1.24228692e+00 -3.93916488e-01 5.09474516e-01
1.02411099e-01 2.49639988e-01 7.43216991e-01 4.11026865e-01
-7.73424327e-01 5.44388473e-01 -3.82738262e-01 4.85777482e-02
7.17627168e-01 -1.07386220e+00 5.68507835e-02 7.37464845e-01
1.28852144e-01 6.79745138e-01 3.87133271e-01 -8.95218313e-01
-9.39873517e-01 -1.08097124e+00 1.33197892e+00 -1.55311391e-01
5.60821295e-01 1.42464235e-01 -1.27011836e+00 7.91139841e-01
6.75688013e-02 -2.12717146e-01 6.60853088e-01 3.08590561e-01
-4.00014281e-01 -2.27561012e-01 -1.05562496e+00 5.50830185e-01
7.74627149e-01 -2.61916101e-01 -7.82536805e-01 5.14064431e-01
1.05590570e+00 -4.66301262e-01 -8.72133434e-01 8.08869064e-01
3.38789284e-01 -8.38476062e-01 8.46657693e-01 -3.72262627e-01
5.17981648e-01 -7.53857493e-01 1.67500511e-01 -9.80672777e-01
-2.43387103e-01 -2.59749144e-01 -7.96065152e-01 1.81395078e+00
8.45812380e-01 -7.23655224e-01 6.76566064e-01 2.39502624e-01
1.01803482e-01 -1.17795599e+00 -9.50357735e-01 -3.65120828e-01
-1.70051709e-01 8.07214603e-02 8.77777338e-01 9.27410960e-01
-8.24422762e-02 9.83147383e-01 1.91144105e-02 6.23541176e-01
4.48468119e-01 5.83810449e-01 6.16071165e-01 -1.19821441e+00
-3.93376440e-01 -2.19715968e-01 8.71089473e-02 -1.38953733e+00
6.58205450e-02 -7.94429600e-01 1.03831694e-01 -2.08476210e+00
3.41464341e-01 -8.29269588e-01 -1.14811100e-01 2.69039422e-01
-9.46687639e-01 -9.01548192e-02 -3.39176625e-01 3.83296579e-01
-9.68177617e-01 4.03444320e-01 1.54524410e+00 -3.56598827e-03
-1.62135556e-01 1.54491380e-01 -8.45509231e-01 5.89666903e-01
6.87677085e-01 -6.81982875e-01 -6.54014707e-01 -1.00517742e-01
5.23365140e-01 5.27576745e-01 -7.05430284e-02 -6.13158941e-01
3.94232601e-01 -1.66408718e-01 -2.14759737e-01 -8.51200998e-01
1.93482742e-01 -5.24501979e-01 -1.86081789e-02 4.23367321e-01
-2.35170767e-01 -2.93502003e-01 -2.33629614e-01 8.01885486e-01
-3.41614366e-01 -5.85577309e-01 2.81251520e-01 -1.69538334e-01
-5.91331303e-01 2.63753891e-01 8.43830034e-02 3.83773148e-01
1.12354124e+00 3.95174444e-01 -8.02415729e-01 -2.58632600e-01
-4.19626415e-01 8.08820605e-01 -2.00239033e-01 3.16090614e-01
3.76766711e-01 -1.34575295e+00 -8.44108403e-01 -6.22079849e-01
3.99560332e-01 5.93798578e-01 2.40834162e-01 8.20251048e-01
-5.81754088e-01 5.54138482e-01 6.13481343e-01 -1.63128197e-01
-1.45591760e+00 3.91317189e-01 2.33236581e-01 -8.52404594e-01
-3.15733165e-01 1.20749962e+00 4.87326719e-02 -5.03730655e-01
3.82435098e-02 -5.89169376e-02 -8.35502386e-01 4.12428379e-01
4.19505209e-01 3.38002801e-01 4.08291295e-02 -7.37285495e-01
-4.69472826e-01 5.37753224e-01 -3.65349829e-01 -2.08938401e-02
5.01406074e-01 -2.27341890e-01 -7.51861989e-01 4.12820429e-01
8.52061510e-01 1.29687786e-01 -4.90972191e-01 -6.70632184e-01
2.42125705e-01 -1.55846149e-01 -2.75895387e-01 -7.80398428e-01
-8.91125381e-01 5.80162764e-01 -7.03064501e-02 3.67117614e-01
9.26928461e-01 5.10073066e-01 1.20881188e+00 5.28684735e-01
5.57710230e-01 -6.30545855e-01 -9.24925804e-02 4.75583613e-01
5.99714637e-01 -1.07360864e+00 1.46210179e-01 -1.23984981e+00
-5.26895821e-01 8.29317212e-01 1.08585143e+00 4.86022502e-01
5.23531199e-01 -2.63131440e-01 9.89325494e-02 -5.88200808e-01
-8.16741347e-01 -4.14438963e-01 2.80207813e-01 1.72551185e-01
3.94898206e-01 1.10535815e-01 -8.52155387e-01 5.93663216e-01
-1.15156628e-01 -2.75966346e-01 2.27746204e-01 9.45352972e-01
-7.68011034e-01 -1.33676171e+00 -3.51524591e-01 5.20655274e-01
-5.63309252e-01 -1.30736649e-01 -4.93115932e-01 6.91933393e-01
4.15230066e-01 1.46458685e+00 -3.48402858e-01 -4.04488236e-01
4.50461894e-01 2.86311120e-01 3.90818477e-01 -6.81203008e-01
-5.91999829e-01 -1.60234302e-01 6.27017498e-01 -1.46030903e-01
-5.66936255e-01 -4.15961117e-01 -1.68043172e+00 -1.76501706e-01
-9.34660554e-01 6.89107239e-01 2.42358856e-02 1.23662674e+00
1.39837712e-01 7.11235762e-01 3.02005470e-01 4.27128792e-01
-7.66408443e-02 -1.12570739e+00 -3.27592880e-01 4.80346829e-02
-5.72358035e-02 -6.62604213e-01 -9.53388512e-02 -1.89369708e-01] | [9.501321792602539, 8.522119522094727] |
35a55641-2a0b-4c5e-86b5-edaf22dda58e | neural-ranking-models-with-weak-supervision | 1704.08803 | null | http://arxiv.org/abs/1704.08803v2 | http://arxiv.org/pdf/1704.08803v2.pdf | Neural Ranking Models with Weak Supervision | Despite the impressive improvements achieved by unsupervised deep neural
networks in computer vision and NLP tasks, such improvements have not yet been
observed in ranking for information retrieval. The reason may be the complexity
of the ranking problem, as it is not obvious how to learn from queries and
documents when no supervised signal is available. Hence, in this paper, we
propose to train a neural ranking model using weak supervision, where labels
are obtained automatically without human annotators or any external resources
(e.g., click data). To this aim, we use the output of an unsupervised ranking
model, such as BM25, as a weak supervision signal. We further train a set of
simple yet effective ranking models based on feed-forward neural networks. We
study their effectiveness under various learning scenarios (point-wise and
pair-wise models) and using different input representations (i.e., from
encoding query-document pairs into dense/sparse vectors to using word embedding
representation). We train our networks using tens of millions of training
instances and evaluate it on two standard collections: a homogeneous news
collection(Robust) and a heterogeneous large-scale web collection (ClueWeb).
Our experiments indicate that employing proper objective functions and letting
the networks to learn the input representation based on weakly supervised data
leads to impressive performance, with over 13% and 35% MAP improvements over
the BM25 model on the Robust and the ClueWeb collections. Our findings also
suggest that supervised neural ranking models can greatly benefit from
pre-training on large amounts of weakly labeled data that can be easily
obtained from unsupervised IR models. | ['Mostafa Dehghani', 'W. Bruce Croft', 'Hamed Zamani', 'Jaap Kamps', 'Aliaksei Severyn'] | 2017-04-28 | null | null | null | null | ['ad-hoc-information-retrieval'] | ['natural-language-processing'] | [ 3.3391601e-01 5.6785047e-02 -3.7462640e-01 -5.2863508e-01
-1.1517755e+00 -6.4548057e-01 9.8446226e-01 2.7606368e-01
-8.8052177e-01 6.4087796e-01 4.3994758e-01 -2.0556679e-01
-4.1159177e-01 -8.1067419e-01 -8.7548673e-01 -3.8344893e-01
-1.8894070e-01 9.1042143e-01 2.4521385e-01 -5.0466108e-01
7.2333559e-02 1.3481396e-01 -1.6295350e+00 4.1105914e-01
7.6175845e-01 1.0851490e+00 1.8306710e-01 5.8589774e-01
-6.8928249e-02 7.8274745e-01 -5.6479228e-01 -3.5184664e-01
3.7573043e-01 2.4819670e-02 -9.3068779e-01 -3.1474414e-01
5.8121198e-01 -5.8931088e-01 -4.1680878e-01 8.3840704e-01
6.4112562e-01 2.4824327e-01 7.2652185e-01 -6.7299181e-01
-8.6286110e-01 7.6036066e-01 -2.0332178e-01 6.4772274e-04
2.1665883e-01 -3.3141309e-01 1.6551124e+00 -8.9053917e-01
8.2738245e-01 1.0376548e+00 3.2018417e-01 4.3224552e-01
-1.3499585e+00 -3.3787602e-01 -1.9234473e-01 6.4970873e-02
-1.1025734e+00 -3.6924613e-01 7.6588970e-01 -2.4414870e-01
7.3497224e-01 1.7255187e-01 -8.1989489e-02 1.2144699e+00
-2.2932604e-01 8.9983892e-01 9.5993751e-01 -5.6380332e-01
1.3837114e-01 4.7736764e-01 4.7659197e-01 5.1170802e-01
2.5058293e-01 1.1000006e-01 -3.9606664e-01 -4.1970351e-01
4.1229922e-01 9.0449922e-02 -2.0199490e-01 -5.2189231e-01
-1.2103711e+00 1.0514976e+00 8.7435901e-01 3.5426375e-01
-2.3836450e-01 2.2078277e-01 4.6528706e-01 6.4107376e-01
6.0412502e-01 8.5091501e-01 -6.4893061e-01 2.7351770e-01
-8.9192355e-01 -3.0540006e-02 7.0540506e-01 6.1406308e-01
8.5110754e-01 -3.0490547e-01 -2.3678878e-01 1.3170975e+00
2.5754493e-01 4.4820738e-01 6.3438600e-01 -6.6443074e-01
5.2950025e-01 4.2529401e-01 2.9214558e-01 -9.7488642e-01
-3.7900516e-01 -5.0574160e-01 -6.5204453e-01 -9.8711848e-02
3.9963749e-01 7.4338927e-03 -1.0348639e+00 1.7490319e+00
-1.9382752e-01 -1.0206061e-01 1.9553134e-01 1.0380088e+00
9.6557695e-01 6.6959375e-01 -1.4369845e-01 6.3703306e-02
1.0489620e+00 -1.0758660e+00 -4.3776467e-01 -2.4956678e-01
8.5885984e-01 -6.1606550e-01 1.2991918e+00 2.9289299e-01
-8.6739892e-01 -5.1589966e-01 -1.0591969e+00 -2.6388738e-01
-8.2822222e-01 4.1310778e-01 7.2521210e-01 7.4707985e-02
-1.1436032e+00 5.4460627e-01 -5.2412355e-01 -4.4330981e-01
2.1708031e-01 7.0121980e-01 -4.2128211e-01 -3.5934269e-01
-1.6350304e+00 8.7478697e-01 3.0713624e-01 1.6833445e-01
-9.7360009e-01 -9.5507212e-02 -6.5445828e-01 1.7039968e-01
4.0457532e-01 -5.6215227e-01 1.1942190e+00 -1.1716336e+00
-1.3556358e+00 7.8475845e-01 1.2028353e-01 -4.6315604e-01
3.4164831e-01 -4.6313873e-01 -2.1986935e-01 1.7089176e-01
2.2708749e-02 7.4565428e-01 5.0863707e-01 -1.4087126e+00
-3.6141700e-01 -3.4227282e-01 4.1805810e-01 1.3605116e-01
-8.1826639e-01 1.0898272e-01 -5.9331632e-01 -4.4264388e-01
-1.1585806e-01 -1.0353551e+00 -1.8084882e-01 -2.7834189e-01
-3.9226207e-01 -4.7376359e-01 3.7133956e-01 -6.0341990e-01
9.2863297e-01 -1.8790578e+00 6.0082372e-02 3.6248732e-01
7.2085172e-02 3.5363388e-01 -6.5565926e-01 4.3449330e-01
7.1815297e-02 1.4400648e-01 3.8807422e-02 -8.8147730e-02
9.6513309e-02 2.4969958e-01 -4.4239381e-01 3.4531295e-02
2.7329469e-01 1.0116169e+00 -1.1594285e+00 -2.4615182e-01
-2.2120337e-01 5.2476126e-01 -5.6765264e-01 2.9817787e-01
-3.4176800e-01 -2.8080476e-02 -6.2898076e-01 4.6288306e-01
2.3323092e-01 -5.6956661e-01 2.9390952e-01 -2.4909550e-01
3.1599540e-01 6.7447472e-01 -8.3583325e-01 1.7235050e+00
-7.3955220e-01 6.6826028e-01 -2.8127593e-01 -1.2798902e+00
8.4336579e-01 2.7406147e-01 3.3459201e-01 -9.2917377e-01
-2.6333872e-02 3.6468956e-01 -5.0038975e-02 -2.7577868e-01
6.5658110e-01 8.7781318e-02 -1.7105150e-01 6.9412696e-01
3.7278694e-01 2.3534568e-01 4.2919904e-01 3.9517599e-01
1.3950183e+00 -2.1597141e-01 -2.4449292e-01 -4.7063366e-02
2.8637445e-01 -4.1761160e-02 1.6438320e-01 1.0242949e+00
2.9086855e-01 7.4594378e-01 3.9449733e-01 -4.0711138e-01
-9.0122271e-01 -1.0420337e+00 -1.9961540e-01 1.7058886e+00
-2.2693165e-02 -3.0785674e-01 -2.5114292e-01 -8.9303750e-01
-6.5606311e-02 3.6118528e-01 -6.0794920e-01 -2.6504621e-01
-4.1722596e-01 -8.1940484e-01 2.7399898e-01 4.8234487e-01
3.1806952e-01 -1.2071712e+00 -5.1911980e-02 2.3313126e-01
5.0439505e-04 -9.8225707e-01 -1.0810388e-01 6.3458347e-01
-9.1067648e-01 -9.0471685e-01 -8.5354644e-01 -8.9240754e-01
7.3749018e-01 2.6311493e-01 1.4458480e+00 2.3357582e-01
3.6848433e-02 4.3162942e-01 -3.5836738e-01 -9.9229462e-02
-1.6292438e-02 3.1790927e-01 1.9564442e-01 -1.1907411e-01
4.5682657e-01 -2.7845079e-01 -5.4946423e-01 3.7506247e-01
-1.1355374e+00 -2.8853834e-01 9.0599382e-01 1.3263819e+00
4.5453167e-01 -1.5520781e-01 7.8099322e-01 -1.2031734e+00
9.4910628e-01 -3.8627133e-01 -5.2722698e-01 4.4331193e-01
-7.2396421e-01 4.9613115e-01 5.8066106e-01 -4.7543752e-01
-7.8498012e-01 -1.3893577e-01 2.6551232e-02 -2.0563601e-01
1.8230970e-01 9.0072423e-01 1.5696894e-01 2.2262861e-01
9.4576770e-01 2.2461757e-02 -3.2916531e-01 -5.5366033e-01
5.5080169e-01 9.0259188e-01 2.1333940e-01 -6.6154099e-01
8.4834170e-01 2.3710917e-01 -4.3229344e-01 -4.4258878e-01
-1.2771720e+00 -6.8136185e-01 -5.4237735e-01 1.5342815e-01
6.7130452e-01 -8.3706063e-01 -1.9432139e-01 -1.8337210e-01
-1.0364928e+00 -3.6781833e-01 -6.5847777e-02 5.9285563e-01
-2.8592008e-01 -7.1699060e-03 -7.9422426e-01 -4.7648677e-01
-3.8630074e-01 -1.0131561e+00 1.0896088e+00 -5.0218035e-03
7.3950209e-02 -9.6511996e-01 3.6536801e-01 4.7920281e-01
6.0721403e-01 -1.8686154e-01 9.1862696e-01 -1.2091467e+00
-5.9528774e-01 -5.9912622e-01 -5.3030044e-01 5.6728446e-01
8.6176239e-02 -2.5006020e-01 -1.1460190e+00 -2.8339115e-01
-5.8656615e-01 -1.0399283e+00 1.3047268e+00 7.3241696e-02
1.1569959e+00 -3.6183915e-01 -1.9696444e-01 1.5917724e-01
1.3434659e+00 -2.4288094e-01 5.1151973e-01 4.9654153e-01
6.0521269e-01 7.7855271e-01 4.7000495e-01 -2.0301754e-03
2.0748529e-01 7.7134335e-01 2.6888135e-01 -1.3768341e-01
3.4246542e-02 -2.9496640e-01 4.2568946e-01 1.0241181e+00
-2.4574210e-01 -2.2592053e-01 -8.3488405e-01 4.8494169e-01
-1.8325284e+00 -7.9551768e-01 4.0588617e-01 2.5599217e+00
1.1410093e+00 3.6015424e-01 -1.9949482e-01 -1.6238602e-01
5.3232104e-01 2.7135861e-01 -4.8845747e-01 -2.0470345e-01
-4.3416210e-02 2.2177203e-01 4.4060978e-01 4.0038389e-01
-1.1973161e+00 8.6975157e-01 5.7893357e+00 6.0837328e-01
-1.2061090e+00 3.6115307e-02 7.7615732e-01 -1.7892599e-01
-4.2785406e-01 -2.4628948e-01 -6.7040771e-01 2.6621398e-01
1.1676450e+00 8.8293776e-02 5.5404842e-01 8.5165107e-01
-1.3190408e-01 3.3423051e-01 -1.2903607e+00 7.9920399e-01
1.7241684e-01 -1.1097542e+00 1.8476984e-01 2.0652160e-02
8.4291905e-01 6.7087746e-01 1.0668311e-01 7.7672797e-01
7.0895737e-01 -1.0183141e+00 1.7604271e-01 4.3643188e-01
6.2118977e-01 -4.7211191e-01 1.2039278e+00 2.9277077e-01
-6.9430465e-01 -1.1297317e-02 -8.6468697e-01 2.5649786e-01
-2.5765792e-01 6.3660222e-01 -8.7708360e-01 3.8992158e-01
6.4636242e-01 7.4941814e-01 -8.4251016e-01 8.7987590e-01
-2.8905410e-01 7.2591656e-01 -4.0818676e-01 -2.0091864e-01
4.1465381e-01 -7.4462570e-02 9.5585287e-02 1.1829438e+00
1.2269250e-01 -2.2315429e-01 1.6806515e-01 5.0705987e-01
-6.7012644e-01 2.1285971e-01 -7.5552845e-01 -2.4260406e-01
1.8312281e-01 1.4315792e+00 -4.3420449e-01 -4.4614464e-01
-5.0062078e-01 7.1580786e-01 7.4239671e-01 7.6218295e-01
-4.5505148e-01 -4.7481409e-01 1.5435843e-01 4.5533579e-02
2.2379081e-01 -1.3147749e-02 1.3949087e-01 -1.2649798e+00
-9.0186717e-03 -7.3371214e-01 5.0561631e-01 -8.0278152e-01
-1.4919009e+00 8.0530798e-01 -1.3005416e-01 -1.1488501e+00
-5.5154747e-01 -8.0480480e-01 -4.0303409e-01 5.3123242e-01
-1.9659072e+00 -9.4191903e-01 -4.9725279e-02 5.0343817e-01
2.5381732e-01 -2.6525789e-01 8.1827176e-01 4.5407292e-01
-2.8852430e-01 6.8265736e-01 6.4893842e-01 4.1977870e-01
1.0898068e+00 -1.4465986e+00 -2.3428455e-01 1.8678272e-01
6.5494382e-01 9.6769106e-01 4.2834333e-01 -1.3783249e-01
-1.4487039e+00 -9.8734057e-01 1.0648572e+00 -6.0507309e-01
8.0201465e-01 -4.8681283e-01 -8.2114583e-01 5.3633922e-01
1.8335102e-01 2.8799930e-01 7.1287280e-01 6.9577074e-01
-6.4008498e-01 -3.5231611e-01 -7.9370326e-01 5.3484792e-01
8.2906485e-01 -8.3996850e-01 -8.5066724e-01 6.7005360e-01
7.7276576e-01 1.5660341e-01 -7.9309922e-01 4.6989378e-01
6.3763803e-01 -5.0668687e-01 1.2511399e+00 -7.8833091e-01
6.9665992e-01 -1.4066255e-01 -3.7140337e-01 -1.3418232e+00
-1.1295864e-01 8.9066010e-03 3.2659311e-02 1.0027866e+00
8.5442257e-01 -5.3960752e-01 6.8480486e-01 6.0655475e-01
1.2080283e-01 -6.6571128e-01 -5.3674960e-01 -6.3554734e-01
2.0887662e-02 -2.7549958e-01 1.7231496e-01 1.0074456e+00
-3.6759991e-02 8.4515804e-01 -3.9520213e-01 -5.9930991e-02
3.5787779e-01 2.3161849e-01 7.8456241e-01 -1.5380116e+00
-3.4379649e-01 -2.9079092e-01 -1.5683395e-01 -1.1553620e+00
3.2799083e-01 -1.2820679e+00 1.8903093e-01 -1.6906712e+00
3.5115331e-01 -5.5295020e-01 -1.0600228e+00 6.7084759e-01
-1.2838282e-01 4.6877220e-01 -2.6185717e-02 3.7503389e-01
-1.0916107e+00 6.3100582e-01 9.3379635e-01 -5.9228581e-01
-5.8757842e-02 -1.5967570e-01 -6.5353578e-01 5.8844167e-01
6.2085533e-01 -6.1186999e-01 -4.6818671e-01 -8.0983233e-01
6.6421139e-01 -6.9422491e-02 1.4027795e-01 -6.6008955e-01
2.5654769e-01 2.7367997e-01 3.8997778e-01 -2.7736738e-01
2.9335174e-01 -7.6264423e-01 -6.3085097e-01 1.0805047e-01
-1.0963444e+00 -1.6994734e-01 -1.7142686e-01 7.1771508e-01
-5.9209812e-01 -4.2224982e-01 2.6708555e-01 -1.7392900e-02
-6.1387235e-01 2.0398217e-01 5.4059748e-02 -5.3742565e-03
2.2417592e-01 3.5078451e-01 -5.4542345e-01 -6.0160512e-01
-6.8050504e-01 2.9608935e-01 2.4174649e-01 5.5378383e-01
4.9802947e-01 -1.4847547e+00 -7.3018622e-01 -7.3618919e-02
4.9552751e-01 -1.6605392e-01 -3.5626531e-01 4.0293506e-01
-2.4335089e-01 7.9471320e-01 1.2488321e-01 -4.8162392e-01
-8.7112898e-01 5.1585615e-01 -9.3576737e-02 -8.4797412e-01
-1.9070695e-01 6.4867282e-01 1.9462878e-01 -9.0312040e-01
3.9197958e-01 -1.9177230e-01 -5.2119559e-01 1.5374759e-01
5.2288628e-01 -7.3062584e-02 1.7536123e-01 -3.1246918e-01
-1.8443392e-01 4.8672515e-01 -4.2537677e-01 -1.0596231e-01
1.6031810e+00 2.3292442e-01 2.6093073e-02 3.9223528e-01
1.5870947e+00 3.4157243e-02 -7.9862022e-01 -6.2123674e-01
4.6676654e-01 -2.1642239e-01 3.4353083e-01 -9.8795849e-01
-1.0474700e+00 9.1771525e-01 6.2278801e-01 2.8905702e-01
8.2650030e-01 2.9899752e-01 6.7263246e-01 1.2105452e+00
3.9687112e-01 -1.2114843e+00 3.3921239e-01 6.1230093e-01
9.1230875e-01 -1.6787868e+00 -2.2260245e-02 2.7988973e-01
-4.2747357e-01 9.0316111e-01 3.9971104e-01 -3.5847354e-01
5.1695287e-01 -3.3875740e-01 2.0529155e-01 -3.7972873e-01
-1.0134059e+00 -4.6562266e-01 7.4178582e-01 2.1268147e-01
7.4452770e-01 -9.0627894e-02 -3.9849234e-01 4.7617131e-01
6.3026115e-02 -3.4842133e-03 1.6029519e-01 6.7785484e-01
-4.6796206e-01 -1.2606027e+00 1.3760053e-02 8.8322932e-01
-5.3175533e-01 -4.0350881e-01 -4.8508790e-01 5.8307302e-01
-3.8597646e-01 9.2408645e-01 -5.7172842e-02 -4.4260037e-01
1.4270584e-01 1.1107278e-01 6.7783226e-03 -8.0902386e-01
-4.2469066e-01 -6.0880326e-02 4.2571232e-01 -4.9107251e-01
-5.1937056e-01 -1.9131300e-01 -9.2587411e-01 3.3959258e-01
-5.5776012e-01 3.9310002e-01 7.2269517e-01 8.1707799e-01
3.0818692e-01 3.1335571e-01 7.5492674e-01 -8.2269669e-01
-8.7949985e-01 -1.1732509e+00 -4.3932173e-01 6.4265096e-01
1.9137499e-01 -6.2138170e-01 -6.6540575e-01 -2.0342812e-01] | [11.416539192199707, 7.563056468963623] |
1495faf6-5bdf-426d-a271-35a741751be4 | learning-through-transcription | null | null | https://aclanthology.org/2022.computel-1.11 | https://aclanthology.org/2022.computel-1.11.pdf | Learning Through Transcription | Transcribing speech for primarily oral, local languages is often a joint effort involving speakers and outsiders. It is commonly motivated by externally-defined scientific goals, alongside local motivations such as language acquisition and access to heritage materials. We explore the task of ‘learning through transcription’ through the design of a system for collaborative speech annotation. We have developed a prototype to support local and remote learner-speaker interactions in remote Aboriginal communities in northern Australia. We show that situated systems design for inclusive non-expert practice is a promising new direction for working with speakers of local languages. | ['Steven Bird', 'Mat Bettinson'] | null | null | null | null | computel-acl-2022-5 | ['language-acquisition'] | ['natural-language-processing'] | [ 1.06556788e-01 6.93333030e-01 3.29520181e-02 -7.25143969e-01
-1.59044349e+00 -7.64923215e-01 5.03183901e-01 2.17062682e-01
-4.72611278e-01 6.72814965e-01 1.32917333e+00 -6.24221146e-01
2.92935856e-02 -8.41117129e-02 -3.35167259e-01 -4.06835347e-01
1.88642323e-01 3.59862804e-01 1.13152321e-02 -5.58318734e-01
2.94058949e-01 3.90528083e-01 -1.43484700e+00 5.07105649e-01
6.45817876e-01 -2.31349945e-01 4.12064314e-01 8.63221407e-01
-5.78147054e-01 1.43526316e+00 -7.67688930e-01 7.96822086e-02
-2.62881756e-01 -6.16530836e-01 -1.50272453e+00 4.20638174e-02
4.75316465e-01 -1.09079666e-01 1.73049435e-01 6.42530024e-01
1.03187811e+00 5.52042902e-01 1.00222409e-01 -4.37518030e-01
-6.47168696e-01 1.06414211e+00 2.04531267e-01 3.12522128e-02
9.41044986e-01 -6.90287054e-02 7.37611294e-01 -8.75758827e-01
7.46462226e-01 1.19612122e+00 9.94425118e-01 6.93384588e-01
-9.78330553e-01 -4.25003022e-01 8.37674513e-02 -2.96800166e-01
-1.60630286e+00 -1.32328594e+00 3.80328029e-01 -5.61307371e-01
9.40027535e-01 7.07276702e-01 6.49205208e-01 1.05374527e+00
-6.06007755e-01 4.90282595e-01 9.46372390e-01 -8.52623701e-01
2.76218325e-01 7.59717643e-01 -4.25233454e-01 2.74998873e-01
-6.57544672e-01 -3.65384430e-01 -1.29502332e+00 -1.37591302e-01
5.67923367e-01 -4.32051420e-01 -5.00226378e-01 5.16482830e-01
-1.26803958e+00 6.82960868e-01 8.13508928e-02 7.28930175e-01
2.39151213e-02 6.70432970e-02 4.32314634e-01 4.30401444e-01
6.41899288e-01 3.34086329e-01 -2.09355175e-01 -8.22964489e-01
-9.31604147e-01 -1.49213642e-01 1.32502866e+00 9.98838723e-01
3.08541864e-01 -2.46660486e-01 1.48969308e-01 1.31248033e+00
1.14445055e+00 1.51596099e-01 3.06675017e-01 -1.06630814e+00
1.74809039e-01 2.61444479e-01 8.61722827e-02 -2.08031401e-01
5.82389086e-02 1.61669493e-01 9.10333619e-02 1.44181788e-01
4.78972584e-01 -3.09005558e-01 -5.41082978e-01 1.33707404e+00
6.12676144e-01 -5.21009326e-01 5.65842867e-01 7.99313009e-01
1.31588161e+00 7.49654949e-01 5.54465592e-01 -3.17427516e-01
1.16276062e+00 -7.75562465e-01 -9.31059837e-01 -6.11176044e-02
8.76507223e-01 -1.25511682e+00 1.30870438e+00 1.64042860e-01
-1.14759815e+00 1.05854213e-01 -3.04563731e-01 -6.79276168e-01
-3.42556477e-01 4.23832648e-02 4.62278843e-01 1.10198784e+00
-1.67002988e+00 1.95366040e-01 -7.77559519e-01 -1.09097075e+00
2.47169659e-01 8.20174143e-02 -5.32204032e-01 2.49655947e-01
-8.85792196e-01 8.89186919e-01 -2.30889484e-01 4.01970953e-01
-8.48577738e-01 -4.29623812e-01 -6.32524848e-01 -3.59413356e-01
6.03851676e-02 2.49815196e-01 1.79269695e+00 -1.02917564e+00
-1.92505050e+00 1.31587946e+00 4.55952100e-02 2.61356086e-01
5.65367341e-01 -4.51384515e-01 -5.83690941e-01 -1.19288340e-01
2.61637747e-01 3.64520609e-01 -2.61718705e-02 -1.21597958e+00
-6.75987184e-01 -7.84785077e-02 -2.35822067e-01 7.28287935e-01
-2.01260239e-01 8.55411768e-01 1.59350500e-01 -4.33489472e-01
1.02147479e-02 -5.82547784e-01 -1.01289719e-01 8.69096145e-02
1.61967903e-01 -2.74750680e-01 7.16301799e-01 -9.97913718e-01
1.14626801e+00 -2.30301428e+00 -2.56477743e-01 1.98247150e-01
-2.62767281e-02 8.95048007e-02 4.44574416e-01 1.27994728e+00
4.21913683e-01 4.81350839e-01 3.34221311e-02 -1.09988585e-01
-3.44259404e-02 3.17190081e-01 -3.31057981e-02 5.10663211e-01
-2.98377126e-01 9.69300345e-02 -1.34054887e+00 -6.47369027e-01
-2.77163267e-01 8.50794017e-01 -5.78209646e-02 5.34663424e-02
3.44552279e-01 5.04436553e-01 -3.43147278e-01 6.84451282e-01
1.15304157e-01 4.31349427e-01 4.63188559e-01 7.40621209e-01
-1.03520095e+00 9.12103415e-01 -1.02082598e+00 2.10718489e+00
-7.43171513e-01 1.10268295e+00 9.91898358e-01 -3.37277889e-01
1.09044242e+00 1.16146708e+00 -6.10806085e-02 -1.61191031e-01
-1.45901695e-01 8.48057210e-01 -8.41057077e-02 -9.92956817e-01
4.23686117e-01 -4.67333913e-01 1.37625232e-01 9.11128700e-01
1.16878092e-01 -2.82291770e-01 -6.70561314e-01 1.79051369e-01
8.19176793e-01 6.22090638e-01 3.87076527e-01 -8.18032444e-01
3.25617790e-01 1.38263091e-01 2.04974040e-01 5.56923568e-01
-5.64208865e-01 2.35842884e-01 -1.38698265e-01 -4.15728420e-01
-8.49332571e-01 -7.78625071e-01 -2.03133523e-01 1.98220193e+00
-4.21105713e-01 -4.21075165e-01 -7.85168707e-01 -3.88751090e-01
-7.95595229e-01 8.94159973e-01 -1.60553977e-01 2.70278573e-01
-3.62769157e-01 2.69214064e-01 9.01641607e-01 3.51503819e-01
1.25615835e-01 -1.35109067e+00 -7.41180003e-01 5.67464352e-01
-3.00540686e-01 -6.39615238e-01 -7.83124030e-01 4.22185361e-01
-3.63917440e-01 -2.54573256e-01 -8.37681592e-01 -1.42630196e+00
4.85282958e-01 2.08146408e-01 8.24872613e-01 2.82952160e-01
7.59143308e-02 7.13188112e-01 -6.39273345e-01 -6.96138859e-01
-9.00805354e-01 1.10514998e-01 2.70547904e-02 -2.44480774e-01
3.33488464e-01 -6.19122982e-01 -1.51544929e-01 3.99137944e-01
-5.77118695e-01 1.05727263e-01 1.90537721e-01 2.61306107e-01
-2.47293469e-02 -8.02163541e-01 7.56240606e-01 -9.15689766e-01
8.49346399e-01 -3.40589374e-01 1.31946176e-01 5.59740841e-01
3.93064857e-01 -4.18846935e-01 2.20446482e-01 -5.21948814e-01
-1.51887441e+00 2.42349058e-01 -2.91716933e-01 5.74319124e-01
-5.12649059e-01 6.49645746e-01 -2.53527164e-01 -3.91764402e-01
1.11831617e+00 -2.17586577e-01 9.57626849e-02 -3.93288642e-01
1.69221044e-01 1.52341664e+00 4.71915096e-01 -9.39874947e-01
3.41088623e-01 8.76538381e-02 -8.55115414e-01 -1.71004927e+00
-4.66088474e-01 -8.32815886e-01 -7.26942778e-01 -9.09777701e-01
8.34187031e-01 -1.36688769e+00 -6.95967436e-01 -1.01538874e-01
-1.02715850e+00 -9.54861462e-01 -5.54726601e-01 8.33566189e-01
-2.53802657e-01 -2.35616088e-01 -1.90414265e-01 -1.45713294e+00
-1.23031951e-01 -7.81687856e-01 1.13764453e+00 3.68267149e-01
-8.64774942e-01 -1.39122939e+00 3.33447903e-01 6.05204880e-01
6.02701843e-01 9.39916521e-02 2.42818236e-01 -9.43729043e-01
-1.59034699e-01 2.63581038e-01 3.76089752e-01 5.59418797e-02
3.43547255e-01 1.73144504e-01 -1.31380284e+00 1.26098037e-01
1.25643909e-01 -7.02112734e-01 -1.52449280e-01 -3.69943976e-01
1.83688194e-01 -7.53056228e-01 -1.63111580e-03 -1.10652320e-01
9.21655834e-01 2.76023448e-01 2.72495210e-01 3.25832009e-01
6.20227695e-01 1.17699325e+00 3.32120866e-01 1.00177996e-01
4.55693543e-01 3.90257329e-01 -6.66955709e-01 6.44072099e-03
4.72454578e-02 -3.18849832e-01 7.38394916e-01 1.39545143e+00
-3.55513573e-01 1.78968925e-02 -1.70200300e+00 1.10142720e+00
-1.77026713e+00 -1.16965079e+00 -3.61345559e-01 2.13459349e+00
1.36702502e+00 -5.82530141e-01 1.65208653e-01 -1.84356928e-01
5.46504617e-01 -1.91246033e-01 3.18097830e-01 -9.57263649e-01
3.17809343e-01 4.83630635e-02 3.50383162e-01 1.09222507e+00
-4.77237403e-01 1.08125973e+00 6.75024986e+00 3.17762166e-01
-8.63467455e-01 4.41103011e-01 3.39759231e-01 -1.07828170e-01
-7.16487706e-01 2.14161381e-01 -7.14929581e-01 -2.22020492e-01
1.42343414e+00 -1.70218974e-01 5.28510630e-01 5.42903185e-01
7.07301974e-01 -5.89273274e-02 -1.02071667e+00 3.44408721e-01
1.01077765e-01 -1.20216846e+00 -4.21262711e-01 1.59139574e-01
5.25145233e-01 1.28532778e-02 -6.06303871e-01 3.98991406e-02
8.64122033e-01 -1.34785295e+00 1.29206979e+00 4.29783642e-01
7.47620285e-01 -7.52996445e-01 2.52020389e-01 5.78684866e-01
-1.10217512e+00 2.34860137e-01 3.61243874e-01 -2.36246169e-01
2.63550222e-01 -2.97829747e-01 -1.21932852e+00 -1.95114329e-01
1.09646130e+00 8.43220726e-02 -1.26777858e-01 7.90300965e-01
-2.50642359e-01 1.07074201e+00 -6.85847938e-01 -2.68893987e-01
5.78386903e-01 -6.57576174e-02 5.33980250e-01 1.52098203e+00
2.21842900e-01 5.49955606e-01 2.93497115e-01 3.67164880e-01
3.36990893e-01 5.54553986e-01 -8.24453354e-01 -2.15710610e-01
7.97066867e-01 1.02267814e+00 -8.80302787e-01 7.40243541e-03
-3.66238594e-01 5.37239850e-01 2.07412884e-01 3.72101516e-01
-1.40783429e-01 -3.70779008e-01 3.47185880e-01 3.80198538e-01
-1.94734290e-01 -2.98642755e-01 -9.52344835e-02 -4.91376251e-01
-1.53446555e-01 -1.32924688e+00 1.33921355e-02 -4.23634082e-01
-8.14166903e-01 2.84329236e-01 -1.95240319e-01 -8.29582930e-01
-1.59261078e-01 -2.22777389e-03 -5.86137831e-01 1.19983041e+00
-7.50531733e-01 -1.48277724e+00 -3.10985982e-01 3.61316979e-01
6.59794569e-01 2.43391767e-02 1.31546116e+00 2.49263436e-01
-1.72270656e-01 2.20579445e-01 2.11839631e-01 -2.79869325e-02
8.04272950e-01 -1.18576276e+00 1.81380540e-01 5.29122829e-01
3.05110753e-01 9.03532624e-01 6.97969317e-01 -6.90054715e-01
-1.21029592e+00 -5.22788763e-01 1.52027929e+00 -4.57358330e-01
7.39848018e-01 -8.57541144e-01 -1.05510390e+00 9.23257113e-01
6.86725914e-01 -5.60001612e-01 1.65110147e+00 9.80143994e-02
1.63357452e-01 2.71355838e-01 -1.17011130e+00 8.28166902e-01
1.30060506e+00 -1.22615516e+00 -5.38860977e-01 8.11444402e-01
7.58832693e-01 -4.26449060e-01 -1.19582427e+00 -5.92699051e-01
6.35820210e-01 -4.06117648e-01 6.75565839e-01 -3.06196868e-01
-3.37813720e-02 -3.22354198e-01 -1.19204842e-01 -9.97885883e-01
1.94156647e-01 -1.72447741e+00 8.37985754e-01 2.09660888e+00
5.70195198e-01 -6.20183885e-01 3.46899778e-01 1.05784738e+00
-7.17983723e-01 1.80216432e-01 -1.15167654e+00 -2.98438549e-01
1.03364296e-01 -3.80136371e-01 2.76669860e-01 1.17581511e+00
5.62658548e-01 4.48452890e-01 -6.31826669e-02 1.29273400e-01
1.03638656e-01 -8.76091182e-01 6.26656771e-01 -1.05906689e+00
3.79839428e-02 -2.82915264e-01 -1.18079938e-01 -4.27509397e-01
-1.02158748e-01 -8.13874245e-01 9.19649184e-01 -1.84984100e+00
-5.53136230e-01 -3.56750667e-01 5.68886936e-01 6.65238619e-01
3.63414794e-01 -1.56754732e-01 -2.74361879e-01 3.63696814e-01
-3.58472914e-01 2.24098444e-01 9.54492569e-01 3.93219173e-01
-9.59796250e-01 -1.94306940e-01 -1.16792953e+00 6.08605266e-01
6.89519465e-01 -6.31233335e-01 -5.45145869e-01 -5.39222002e-01
4.09178227e-01 1.99113060e-02 8.64790678e-02 -8.87580037e-01
6.93443418e-01 -5.61018348e-01 -2.92621344e-01 -4.11886685e-02
-2.47910082e-01 -7.36080825e-01 5.11764050e-01 -6.27056882e-02
-8.15053403e-01 -4.43019181e-01 1.26932591e-01 9.54701602e-02
-2.27696508e-01 -5.25292218e-01 4.52258199e-01 -4.30651248e-01
-4.90733683e-01 -6.17328346e-01 -1.04175973e+00 1.45108730e-01
1.03061724e+00 -4.65693891e-01 -2.47335956e-01 -6.79311752e-01
-8.75773132e-01 2.84720719e-01 4.16548908e-01 4.06272970e-02
3.12577337e-01 -1.03928351e+00 -7.80696571e-01 -1.05979890e-01
7.22956732e-02 3.56183082e-01 -1.20791800e-01 8.43651652e-01
-1.06659210e+00 3.86287607e-02 1.59989178e-01 -2.84415275e-01
-1.37938631e+00 -6.40244007e-01 4.38019335e-01 6.41861498e-01
-8.86738241e-01 1.16437042e+00 -3.64667922e-01 -9.75037456e-01
5.04111409e-01 -5.87965921e-02 -3.28525901e-01 3.70373994e-01
7.15072870e-01 4.25506383e-01 4.73964810e-02 -1.14660215e+00
-4.28702384e-01 -8.43798593e-02 1.71636060e-01 -1.02059591e+00
1.56627941e+00 -3.96388978e-01 6.61889017e-02 1.12796569e+00
9.48815942e-01 5.65913737e-01 -9.32917058e-01 -5.75755611e-02
2.51296729e-01 -4.63031590e-01 2.04558447e-01 -8.97987425e-01
-1.71374246e-01 6.28049433e-01 5.25560081e-01 2.44583905e-01
4.93700266e-01 3.61190468e-01 2.36823946e-01 5.78338742e-01
1.22331686e-01 -1.60965383e+00 -2.79328197e-01 4.06679779e-01
1.34369349e+00 -9.60901022e-01 -1.02514789e-01 -1.73658133e-01
-7.88380623e-01 1.16214144e+00 3.11880767e-01 8.11488032e-01
6.47486389e-01 3.02243769e-01 1.06077564e+00 -4.25701231e-01
-7.37236202e-01 -7.38293305e-02 -3.64628993e-02 9.28567588e-01
1.48735774e+00 1.57145143e-01 -3.01181823e-01 1.76531225e-01
-7.44335890e-01 -1.22972026e-01 6.35096550e-01 1.50673664e+00
-5.52184939e-01 -1.10861194e+00 -7.47932911e-01 -2.01769620e-01
-4.82880443e-01 -2.29056701e-01 -1.08452308e+00 6.44031465e-01
2.38703359e-02 1.22664452e+00 1.96905583e-02 1.80671960e-01
2.44008005e-01 4.13551539e-01 -8.58849436e-02 -1.14780152e+00
-1.45284414e+00 5.60112178e-01 8.91339123e-01 -6.01154417e-02
-8.70628238e-01 -1.27956617e+00 -1.30804074e+00 -2.64317483e-01
-3.33971381e-01 6.70222700e-01 1.13352251e+00 8.70150506e-01
-1.83491334e-02 2.17627212e-01 3.42102259e-01 -8.16652775e-01
-7.24839121e-02 -1.03095102e+00 -5.04209757e-01 -2.49745801e-01
2.85439163e-01 2.23138854e-01 -3.96910280e-01 3.29967082e-01] | [14.073956489562988, 7.042368412017822] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.