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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
669b13ee-89ad-476f-95d0-fd189524289a | analyzing-speaker-information-in-self | 2108.00917 | null | https://arxiv.org/abs/2108.00917v1 | https://arxiv.org/pdf/2108.00917v1.pdf | Analyzing Speaker Information in Self-Supervised Models to Improve Zero-Resource Speech Processing | Contrastive predictive coding (CPC) aims to learn representations of speech by distinguishing future observations from a set of negative examples. Previous work has shown that linear classifiers trained on CPC features can accurately predict speaker and phone labels. However, it is unclear how the features actually capture speaker and phonetic information, and whether it is possible to normalize out the irrelevant details (depending on the downstream task). In this paper, we first show that the per-utterance mean of CPC features captures speaker information to a large extent. Concretely, we find that comparing means performs well on a speaker verification task. Next, probing experiments show that standardizing the features effectively removes speaker information. Based on this observation, we propose a speaker normalization step to improve acoustic unit discovery using K-means clustering of CPC features. Finally, we show that a language model trained on the resulting units achieves some of the best results in the ZeroSpeech2021~Challenge. | ['Herman Kamper', 'Matthew Baas', 'Leanne Nortje', 'Benjamin van Niekerk'] | 2021-08-02 | null | null | null | null | ['acoustic-unit-discovery'] | ['speech'] | [ 5.54755926e-01 2.23513424e-01 -1.45521492e-01 -8.70962620e-01
-1.20436037e+00 -6.88694179e-01 7.31823623e-01 4.27172929e-02
-1.98383018e-01 1.23255752e-01 6.06203973e-01 -4.79277462e-01
1.71359643e-01 -6.04467541e-02 -5.72350860e-01 -8.12081516e-01
-3.81519556e-01 2.01817468e-01 -1.35289431e-01 3.33483219e-02
2.88599759e-01 4.88178492e-01 -1.91247940e+00 6.57493174e-01
3.58895540e-01 8.59833300e-01 9.53786448e-02 9.75388527e-01
-1.79770160e-02 6.61661744e-01 -6.02544606e-01 -7.93798640e-02
1.89658571e-02 -5.69361627e-01 -8.79645407e-01 -2.58843601e-02
2.91830540e-01 9.89901125e-02 1.36918247e-01 9.83276784e-01
3.47787112e-01 3.13747972e-01 9.73962009e-01 -1.35969293e+00
-4.98146057e-01 9.24664021e-01 -1.53412625e-01 3.62084389e-01
2.59356171e-01 -2.00333491e-01 1.19196308e+00 -9.90932882e-01
1.63464114e-01 1.42043722e+00 7.34089553e-01 7.64854729e-01
-1.34980583e+00 -8.11116219e-01 4.78719950e-01 1.66413188e-01
-1.37668705e+00 -1.18541491e+00 7.86125660e-01 -3.57894778e-01
1.33379662e+00 5.15064597e-01 1.35413080e-01 1.13330185e+00
-8.51384625e-02 7.34697998e-01 8.68956983e-01 -8.66567791e-01
2.39282086e-01 3.69340420e-01 4.08112615e-01 3.47098559e-01
-4.84252214e-01 1.99372411e-01 -5.66765606e-01 -3.20999205e-01
-4.38465923e-02 -2.78977752e-01 -5.20629585e-01 -2.14898288e-02
-9.52228725e-01 8.61163795e-01 -3.30812633e-02 5.79595029e-01
-1.14791624e-01 1.65622234e-01 4.17112738e-01 4.12391037e-01
4.71326590e-01 2.94965774e-01 -7.59612381e-01 -4.08006757e-01
-8.47774565e-01 -1.60685405e-01 9.38219547e-01 1.01537287e+00
5.70053935e-01 1.72399402e-01 -1.70835126e-02 1.05537951e+00
3.37562710e-01 5.93936980e-01 7.93017626e-01 -1.02918100e+00
3.72251779e-01 -6.77934363e-02 -2.61552423e-01 -6.26897931e-01
-2.18834162e-01 -2.16231287e-01 -5.20579278e-01 -1.14283703e-01
8.36782679e-02 -1.15954816e-01 -8.79656494e-01 2.07802320e+00
-3.20915848e-01 4.06487197e-01 3.19553345e-01 3.63524824e-01
4.46885973e-01 9.18948948e-01 2.23972797e-01 -4.18901920e-01
1.00941455e+00 -6.30885184e-01 -7.75896251e-01 -4.26755548e-01
7.64413536e-01 -8.26144934e-01 8.00126493e-01 3.10667247e-01
-7.22145081e-01 -5.21849096e-01 -9.35230076e-01 1.61733136e-01
-3.15854818e-01 1.99149370e-01 2.92666435e-01 9.08204019e-01
-1.09353387e+00 4.89925295e-01 -9.97546196e-01 -3.82110924e-01
-4.75167185e-02 4.31995153e-01 -2.81827241e-01 1.19194783e-01
-1.06134272e+00 8.01315129e-01 2.48357609e-01 -9.61750746e-02
-6.24843955e-01 -4.38474029e-01 -8.64293873e-01 3.92786384e-01
-2.44569018e-01 1.76607192e-01 1.60888422e+00 -1.08076251e+00
-1.73811054e+00 5.94338357e-01 -7.58166254e-01 -6.89248383e-01
-2.41057113e-01 7.25284219e-02 -7.35195816e-01 -1.20701022e-01
-1.60550535e-01 7.08577812e-01 9.58047211e-01 -1.09310997e+00
-7.94890225e-01 -2.19611764e-01 -4.69796985e-01 1.66906670e-01
-3.89867783e-01 1.75020978e-01 -3.75734419e-01 -4.33483362e-01
3.28785568e-01 -1.07341039e+00 7.55940005e-02 -5.76468289e-01
-4.40940976e-01 -5.38187742e-01 7.44076729e-01 -5.75705171e-01
1.17363870e+00 -2.63686585e+00 -2.52703637e-01 6.46054447e-01
-2.31082872e-01 -1.12643503e-02 -1.17672302e-01 2.41203308e-01
-5.63482046e-01 1.64001629e-01 -1.74451187e-01 -6.57707155e-01
1.06890559e-01 3.02961916e-01 -6.93078220e-01 5.34353793e-01
3.47246289e-01 5.60892999e-01 -3.70870918e-01 -3.55496377e-01
1.41450033e-01 6.34135127e-01 -7.28133500e-01 -5.69301955e-02
1.39930487e-01 2.62208939e-01 5.64303584e-02 2.90555447e-01
4.33183193e-01 5.89116216e-02 3.81442159e-01 1.79349467e-01
-7.74998888e-02 9.26829338e-01 -1.00485468e+00 1.29196227e+00
-4.49084967e-01 9.33009565e-01 2.63313025e-01 -1.38576353e+00
8.30910385e-01 7.36006796e-01 2.89940476e-01 -2.63561726e-01
8.93030763e-02 1.59580186e-01 1.35339931e-01 -9.68737230e-02
2.31763914e-01 -4.74091530e-01 -5.50697930e-02 4.08611149e-01
3.21284264e-01 -3.82265188e-02 -4.03932571e-01 -2.91746613e-02
9.41107452e-01 -5.33507168e-01 2.71131903e-01 -3.14994782e-01
6.64007962e-01 -2.13867426e-01 6.76777780e-01 7.60672748e-01
-5.68300426e-01 6.87944591e-01 2.03016818e-01 2.34524727e-01
-5.57942212e-01 -1.04688108e+00 -2.67759323e-01 1.61984921e+00
-4.54174072e-01 -5.78178406e-01 -7.10313141e-01 -4.05654222e-01
-2.03026786e-01 1.23771667e+00 -4.83401686e-01 -2.70193428e-01
-6.69205427e-01 -5.80851495e-01 6.56335950e-01 6.31710231e-01
-3.73255640e-01 -9.43202674e-01 1.02315426e-01 2.40661725e-01
-2.99375027e-01 -1.04246581e+00 -4.87419218e-01 8.49977195e-01
-6.33262455e-01 -6.32124782e-01 -2.98320055e-01 -1.23040569e+00
3.90018851e-01 3.01559508e-01 9.18923318e-01 -2.58127779e-01
2.22699642e-01 6.17309451e-01 -3.22579235e-01 -5.76905310e-01
-9.77276266e-01 1.08051747e-01 4.46307421e-01 4.93912585e-02
7.68703222e-01 -5.95851660e-01 -3.50085236e-02 -4.92978981e-03
-2.03617766e-01 -2.22077310e-01 4.92221355e-01 7.14196622e-01
4.08437163e-01 2.32068792e-01 6.54724717e-01 -6.95384920e-01
5.46224117e-01 -2.25804567e-01 -1.78353950e-01 2.14105681e-01
-5.51243007e-01 2.55719215e-01 5.62485516e-01 -5.53481758e-01
-1.11403453e+00 4.26973820e-01 -4.15154785e-01 -2.88687795e-01
-3.58994454e-01 2.36191526e-01 -1.21604711e-01 4.19581264e-01
5.58212101e-01 4.85323787e-01 2.75990013e-02 -5.43756962e-01
4.73602384e-01 1.10857189e+00 7.53736913e-01 -3.89278382e-01
5.22421598e-01 1.90760687e-01 -5.58053911e-01 -1.01449025e+00
-7.12316513e-01 -8.25212777e-01 -5.93453526e-01 8.92794058e-02
7.97585309e-01 -1.02255809e+00 -6.79049432e-01 -6.39028326e-02
-1.14354062e+00 -7.60268569e-02 -2.71021485e-01 8.03720653e-01
-7.12147415e-01 1.13342799e-01 -5.34548521e-01 -1.28001750e+00
-1.96644664e-01 -1.05311763e+00 9.06298459e-01 -2.07274795e-01
-5.90945780e-01 -7.70992994e-01 1.51561484e-01 6.74086623e-03
4.12026495e-01 -6.43894613e-01 9.86996174e-01 -1.33143032e+00
-1.68842912e-01 -2.59317368e-01 3.95384312e-01 7.59405792e-01
1.54895380e-01 6.07627220e-02 -1.71209335e+00 -3.18912894e-01
2.39837721e-01 3.94455902e-02 1.02893555e+00 5.20165741e-01
1.15905702e+00 -3.30262333e-01 -4.34334636e-01 3.20583969e-01
8.27997983e-01 3.70334625e-01 3.25877994e-01 -2.51828700e-01
3.54223579e-01 7.25743830e-01 1.17401987e-01 6.69594631e-02
1.74240708e-01 5.28398812e-01 6.58573210e-03 2.99068123e-01
1.30227730e-02 -2.28478730e-01 8.22886884e-01 1.30923259e+00
2.36823395e-01 -1.40806183e-01 -8.40389073e-01 6.78562820e-01
-1.37064457e+00 -1.20675385e+00 1.02808319e-01 2.22040081e+00
8.35448742e-01 2.40573525e-01 -1.72429085e-01 4.85089511e-01
8.83194149e-01 -3.46050598e-02 -1.88313857e-01 -6.29856646e-01
-1.09190680e-01 3.17254037e-01 4.40142065e-01 7.77649641e-01
-1.28429878e+00 7.11873829e-01 7.03078604e+00 6.57191694e-01
-1.22208846e+00 1.68868706e-01 5.77259719e-01 -1.29519284e-01
-1.62597612e-01 -1.38154224e-01 -1.05787158e+00 3.47949266e-01
1.67443073e+00 -2.00459480e-01 4.19733286e-01 1.14324868e+00
4.08058837e-02 4.21080142e-01 -1.61282885e+00 1.04355133e+00
3.61348629e-01 -9.39366519e-01 -2.66682357e-01 3.40919718e-02
4.63887691e-01 3.81028324e-01 2.75453448e-01 7.69706190e-01
4.05764580e-01 -9.89655077e-01 7.60379493e-01 1.52389988e-01
4.37635809e-01 -7.26245403e-01 5.17332733e-01 5.44596672e-01
-1.07230794e+00 -3.99536937e-01 -2.73480147e-01 -3.35574895e-02
-3.60847004e-02 2.30081126e-01 -1.37221146e+00 -8.40302110e-02
6.72753811e-01 5.12541473e-01 -5.31718910e-01 5.85599601e-01
7.54026696e-03 1.34025431e+00 -3.84643525e-01 8.62302408e-02
-2.07889918e-02 2.44115278e-01 3.68173987e-01 1.63781869e+00
4.20990080e-01 -1.47289513e-02 -3.42697203e-02 5.40601671e-01
-4.10509408e-02 1.49829775e-01 -6.09465539e-01 -4.88887019e-02
7.29942620e-01 7.78670251e-01 -3.77882093e-01 -4.85496759e-01
-5.84989607e-01 8.06869924e-01 3.26356977e-01 3.96973640e-01
-5.34038126e-01 -1.97523475e-01 1.03462553e+00 -1.90314844e-01
5.42393744e-01 -2.19366159e-02 -2.34353945e-01 -1.09148669e+00
-1.65542737e-01 -6.68683827e-01 2.15469077e-01 -3.54790598e-01
-1.37870657e+00 5.66551208e-01 -2.65012741e-01 -1.17411375e+00
-9.72796619e-01 -5.92605531e-01 -8.08884919e-01 9.97946560e-01
-1.32654285e+00 -7.00357735e-01 4.49929386e-01 4.43194717e-01
6.55359209e-01 -2.57398218e-01 1.34741819e+00 -4.26472910e-02
-4.53653216e-01 7.05583751e-01 4.56014246e-01 3.26218754e-01
8.18905294e-01 -1.27385855e+00 5.08272111e-01 8.15011322e-01
4.89560843e-01 8.88916254e-01 7.25426853e-01 -1.90519512e-01
-1.02327776e+00 -1.00913215e+00 1.41960621e+00 -5.37024796e-01
4.86804694e-01 -6.40627146e-01 -1.17729104e+00 1.10100877e+00
9.09045115e-02 1.94593275e-03 1.04238033e+00 6.66843891e-01
-5.29088914e-01 -8.04847926e-02 -7.80844390e-01 2.41561368e-01
6.51950777e-01 -1.11656070e+00 -9.26146507e-01 3.04255813e-01
7.90167332e-01 7.03540668e-02 -5.46930969e-01 1.25071630e-01
5.21342456e-01 -7.41006851e-01 1.02704442e+00 -6.15703404e-01
-1.32700369e-01 -1.09599583e-01 -6.96110427e-01 -1.38096464e+00
-4.56558585e-01 -2.88420111e-01 9.73975286e-02 1.55208671e+00
7.38646269e-01 -6.29130602e-01 6.22840822e-01 7.45831013e-01
-3.44136715e-01 -2.06899837e-01 -1.14267409e+00 -9.02883828e-01
3.73793870e-01 -1.05552340e+00 4.55438942e-01 9.63047445e-01
3.64618570e-01 5.34708083e-01 -1.38767913e-01 5.22847056e-01
3.02795798e-01 5.95363490e-02 2.94166923e-01 -1.35849190e+00
-3.74412924e-01 -4.16143179e-01 -5.05181372e-01 -1.23887074e+00
7.44770706e-01 -1.09215009e+00 5.49570739e-01 -8.88385773e-01
-1.12878839e-02 -3.65697771e-01 -6.44672453e-01 4.88316864e-01
-3.06705236e-02 -1.30936988e-02 1.90675020e-01 2.79568166e-01
-3.82629156e-01 3.59011382e-01 3.92440766e-01 -2.96984047e-01
-4.06183988e-01 3.38089138e-01 -7.72935569e-01 7.65872777e-01
1.05944383e+00 -3.77595454e-01 -3.84602696e-01 -2.16712371e-01
-4.70142037e-01 -1.30365059e-01 1.39656551e-02 -1.01683378e+00
3.57121080e-01 7.06903711e-02 2.61807293e-01 -7.26143777e-01
7.03661680e-01 -6.85364425e-01 -2.80597627e-01 2.95684338e-01
-8.18480313e-01 -2.85863072e-01 9.49692726e-02 6.93807065e-01
-4.18111712e-01 -3.04552585e-01 7.26664364e-01 2.27296874e-01
-6.35902047e-01 -3.30130786e-01 -9.44226444e-01 -1.11962497e-01
4.17378664e-01 1.47148862e-01 7.34314546e-02 -7.49710083e-01
-7.46654510e-01 -1.51081368e-01 1.43000139e-02 4.93032932e-01
3.21116120e-01 -1.28819680e+00 -5.94528854e-01 6.30320728e-01
2.91778475e-01 -5.09141982e-01 2.67416444e-02 7.18349516e-01
2.99602211e-01 8.75252008e-01 4.71826315e-01 -7.47323036e-01
-1.61418283e+00 5.84989846e-01 3.28727633e-01 1.53520778e-01
-3.79870027e-01 1.11624539e+00 3.96165848e-01 -7.14716136e-01
5.97234964e-01 -7.58930027e-01 -2.42866620e-01 1.40436506e-02
6.78498745e-01 -5.00480197e-02 3.02349240e-01 -1.08077943e+00
-6.25067949e-01 2.67147005e-01 -1.81326076e-01 -4.14786577e-01
1.39584529e+00 -3.51305604e-01 3.05118531e-01 7.53364086e-01
1.70270348e+00 1.25178635e-01 -1.11521280e+00 -3.40632915e-01
2.31161848e-01 7.30728135e-02 3.86155583e-02 -5.97013831e-01
-6.91803932e-01 1.16217566e+00 7.22426236e-01 3.01103055e-01
1.11489666e+00 2.20813870e-01 3.18627357e-01 5.37068546e-01
4.19260003e-02 -1.11655521e+00 -5.42529881e-01 7.37306356e-01
7.85750389e-01 -1.26750350e+00 -3.42966825e-01 -5.33094764e-01
-6.43926322e-01 1.03091240e+00 4.29777801e-02 2.28248119e-01
1.08883548e+00 4.24719542e-01 1.73815653e-01 1.86687604e-01
-1.11963272e+00 -1.53811440e-01 2.62732774e-01 7.80170262e-01
7.31471837e-01 4.15294528e-01 3.97744030e-01 8.17241669e-01
-7.86338508e-01 -6.01828992e-01 2.98091561e-01 7.60591328e-01
-7.10776746e-01 -1.04508150e+00 -5.12046218e-01 4.58819240e-01
-5.34534156e-01 -3.54037434e-01 -3.63708168e-01 4.13869768e-01
-1.30357206e-01 1.34867227e+00 4.12031800e-01 -5.16818106e-01
1.70489490e-01 8.81746352e-01 9.52995718e-02 -7.09451437e-01
-3.24493200e-01 2.41274074e-01 7.36104175e-02 -3.03534299e-01
-4.49231148e-01 -9.53965008e-01 -1.33719110e+00 -9.23170149e-02
-4.38023835e-01 4.19202119e-01 9.01480854e-01 8.69715035e-01
3.07084888e-01 3.34158212e-01 7.55388916e-01 -6.62069917e-01
-8.53352547e-01 -1.18488503e+00 -6.37889683e-01 3.61612707e-01
5.30477464e-01 -3.75121444e-01 -9.13104773e-01 5.14940679e-01] | [14.42087459564209, 6.373611927032471] |
b1b36b96-8cb4-42e3-93e7-33ef8e4a4a93 | unsupervised-object-segmentation-with | 1911.09228 | null | https://arxiv.org/abs/1911.09228v1 | https://arxiv.org/pdf/1911.09228v1.pdf | Unsupervised Object Segmentation with Explicit Localization Module | In this paper, we propose a novel architecture that iteratively discovers and segments out the objects of a scene based on the image reconstruction quality. Different from other approaches, our model uses an explicit localization module that localizes objects of the scene based on the pixel-level reconstruction qualities at each iteration, where simpler objects tend to be reconstructed better at earlier iterations and thus are segmented out first. We show that our localization module improves the quality of the segmentation, especially on a challenging background. | ['John D. Owens', 'Weitang Liu', 'James Sharpnack', 'Lifeng Wei'] | 2019-11-21 | null | null | null | null | ['unsupervised-object-segmentation'] | ['computer-vision'] | [ 5.88730276e-02 1.91315517e-01 -6.90817833e-02 -2.01882049e-01
-2.89611578e-01 -5.40944338e-01 1.48971826e-01 2.80085653e-01
-4.24560308e-01 4.32313710e-01 -4.27190036e-01 2.69458480e-02
-1.24231167e-02 -8.05940926e-01 -6.25909984e-01 -5.30792534e-01
-2.09842697e-02 5.88112652e-01 1.10820746e+00 9.97756273e-02
3.01687986e-01 7.26337612e-01 -1.39984548e+00 1.26021758e-01
9.36507165e-01 7.55746007e-01 6.42084062e-01 8.23833764e-01
-2.21651465e-01 7.65446126e-01 -3.84670675e-01 -1.73841417e-01
3.33168030e-01 -5.14260054e-01 -9.90511000e-01 5.03363252e-01
4.14665103e-01 -1.49454772e-01 -1.34640560e-01 1.30218518e+00
-5.75170219e-02 6.24393523e-02 4.77492094e-01 -6.52092159e-01
-1.01322467e-02 5.09766221e-01 -7.98834801e-01 2.94574976e-01
5.07579334e-02 7.10306019e-02 8.87339771e-01 -9.67331886e-01
7.35983014e-01 1.08703160e+00 7.20332682e-01 2.35008627e-01
-1.35758555e+00 -1.20615259e-01 5.25873423e-01 2.16339752e-01
-1.49803472e+00 -2.16402128e-01 9.75893617e-01 -4.36684579e-01
2.86215484e-01 1.65096402e-01 6.21169567e-01 2.71660000e-01
7.03633875e-02 1.07271671e+00 9.93330061e-01 -4.50714827e-01
5.26927233e-01 1.84908032e-01 2.41036162e-01 9.16557312e-01
2.82157183e-01 -2.53403842e-01 -1.93517029e-01 2.29948819e-01
1.08887053e+00 2.77643721e-03 -3.54686141e-01 -7.65631378e-01
-1.28033745e+00 6.02303028e-01 6.97500288e-01 6.61754549e-01
-6.66449070e-01 2.14120433e-01 -1.22380219e-01 -1.55363649e-01
4.41718191e-01 4.56826627e-01 -5.59389293e-01 3.41654301e-01
-1.46327758e+00 -1.26626447e-01 5.13489127e-01 6.80274129e-01
1.23755348e+00 -2.40480661e-01 -1.28044188e-01 6.19620442e-01
2.51305640e-01 1.59432396e-01 -1.27931044e-01 -1.30885029e+00
7.76234828e-03 8.01508307e-01 1.86535895e-01 -1.02659321e+00
-3.65639210e-01 -6.72092736e-01 -5.11388361e-01 4.69024420e-01
5.38093150e-01 -6.00046813e-02 -1.14379799e+00 1.25720978e+00
5.34112811e-01 3.97748917e-01 -1.50389448e-01 1.00787258e+00
4.96115267e-01 7.40905821e-01 5.19295037e-02 -1.55571193e-01
1.01266170e+00 -1.27875829e+00 -5.73304296e-01 -3.98521513e-01
2.49582648e-01 -6.23498321e-01 8.20535302e-01 5.69407940e-01
-1.19872439e+00 -8.63958299e-01 -7.19407082e-01 1.34558782e-01
-1.94600355e-02 2.43442744e-01 7.14028180e-01 5.38494408e-01
-1.31449592e+00 7.51015544e-01 -8.69960845e-01 -3.35034579e-01
6.66666210e-01 1.31662071e-01 6.81002513e-02 8.96148682e-02
-3.95995170e-01 6.33366466e-01 6.32284582e-01 6.36158325e-03
-1.07190895e+00 -6.56611323e-01 -5.77705324e-01 -7.15846196e-02
3.62646371e-01 -6.82255328e-01 1.02262044e+00 -1.37120962e+00
-1.45609033e+00 8.52098286e-01 -1.93970427e-01 -3.46506149e-01
5.89190423e-01 -5.10533527e-02 1.74872965e-01 5.26859581e-01
6.93037212e-02 9.91504908e-01 8.69315565e-01 -1.90143955e+00
-1.12274551e+00 -2.52502859e-01 3.17568749e-01 2.66230166e-01
2.37804860e-01 -1.93659917e-01 -1.08671904e+00 -4.37840819e-01
6.15876734e-01 -6.03813767e-01 -8.30760121e-01 9.95865539e-02
-4.36428308e-01 -5.11766225e-03 9.27411616e-01 -4.89649802e-01
9.38043237e-01 -2.26892948e+00 2.77825952e-01 3.82380456e-01
4.61165756e-01 9.28327590e-02 -1.23542108e-01 -1.08866878e-01
1.18174069e-01 3.20939347e-02 -4.99032378e-01 -3.52567255e-01
-4.59443271e-01 2.23764271e-01 -9.19259060e-03 5.77528417e-01
1.15188576e-01 8.05181324e-01 -9.64053512e-01 -8.43853772e-01
4.38811541e-01 3.16576809e-01 -4.59534049e-01 7.66151696e-02
-3.76066834e-01 8.33539665e-01 -6.64734364e-01 6.66969597e-01
8.72221470e-01 -4.86871064e-01 -8.41628462e-02 -2.40153506e-01
-3.55794400e-01 -8.76547471e-02 -1.30269480e+00 1.81157291e+00
-2.46084914e-01 6.16953731e-01 4.05716091e-01 -8.93796265e-01
6.37198627e-01 1.39133530e-02 7.49123573e-01 -4.76749033e-01
1.91500470e-01 4.98633869e-02 7.07077514e-03 -2.36395851e-01
4.59885806e-01 9.88855138e-02 3.50049615e-01 2.08460614e-01
-1.22255892e-01 -2.70241380e-01 4.73370641e-01 2.02893123e-01
9.35948253e-01 2.06259713e-01 1.91072211e-01 -4.12436426e-01
4.68930602e-01 3.17754745e-01 5.44619739e-01 1.07650089e+00
-2.45328337e-01 8.91587555e-01 2.59231329e-01 -3.26486856e-01
-8.61528695e-01 -1.17894828e+00 -5.48091114e-01 8.43631566e-01
9.15419042e-01 -1.68696150e-01 -1.00819445e+00 -9.45378840e-01
-2.31062174e-01 5.30945659e-01 -4.42300588e-01 3.48978162e-01
-6.42231643e-01 -4.70089614e-01 -1.27125755e-02 3.56257915e-01
5.39117396e-01 -9.59409595e-01 -9.76659656e-01 2.22927019e-01
-2.90908664e-01 -1.11698306e+00 -1.89883932e-01 2.39709139e-01
-1.22898519e+00 -9.05465901e-01 -7.93196857e-01 -7.89913595e-01
1.08985651e+00 3.17068517e-01 1.27882242e+00 3.53071421e-01
-2.76007473e-01 3.87272269e-01 -5.24825215e-01 -9.66128558e-02
-3.18671674e-01 -3.99956182e-02 -3.72170359e-01 1.90942064e-01
-2.85599470e-01 -3.75270694e-01 -6.77260339e-01 4.11352038e-01
-8.10542524e-01 2.81315714e-01 7.29051411e-01 2.57050306e-01
1.05672395e+00 6.40945017e-01 -7.34751597e-02 -9.20349598e-01
-4.76329848e-02 -2.26920396e-01 -8.29090357e-01 2.28472769e-01
-2.67333061e-01 4.57430780e-02 4.71673101e-01 -3.31976444e-01
-8.31968904e-01 4.97536868e-01 4.99248505e-04 -4.78855759e-01
-4.82709855e-01 1.69806480e-01 -3.44822109e-02 -2.96284735e-01
4.65939552e-01 1.25437528e-01 -4.78747547e-01 -7.11288214e-01
3.89618307e-01 6.88367262e-02 5.89240849e-01 -4.06853437e-01
7.99310029e-01 8.28575134e-01 -8.17502439e-02 -7.10208476e-01
-8.16304803e-01 -6.09681249e-01 -9.87788916e-01 -5.79280078e-01
9.64413464e-01 -6.25384510e-01 -5.05976737e-01 2.93127090e-01
-1.25168169e+00 -5.87904215e-01 -5.70896089e-01 4.42659020e-01
-3.65427375e-01 4.29829121e-01 -4.18785006e-01 -8.87536943e-01
5.48726693e-02 -1.10106921e+00 1.13571811e+00 4.57860053e-01
3.86596173e-02 -9.79832470e-01 -1.03644058e-01 -9.33248401e-02
5.99800572e-02 3.00492700e-02 5.12712121e-01 -1.27328962e-01
-1.15048432e+00 1.27180681e-01 -3.66389692e-01 1.72213122e-01
-5.87693155e-02 2.24825993e-01 -9.51908350e-01 -4.52352203e-02
1.17947482e-01 2.41533518e-01 1.07394373e+00 8.37414205e-01
1.26318884e+00 -1.13852210e-01 -5.98126352e-01 8.45615983e-01
1.68520343e+00 1.97529837e-01 8.14903677e-01 3.14634621e-01
6.88026726e-01 6.20503068e-01 6.51969135e-01 2.22372681e-01
7.68871903e-02 5.70974052e-01 5.90476513e-01 -6.55714154e-01
-3.88321012e-01 1.19907325e-02 2.73185462e-01 4.55027252e-01
-1.00370254e-02 -2.49461189e-01 -8.29249799e-01 7.97830284e-01
-1.82302010e+00 -6.60041332e-01 -4.63448375e-01 1.94550490e+00
6.28001451e-01 2.36869350e-01 1.20336123e-01 6.62168413e-02
6.45194411e-01 6.43133745e-02 -5.97378612e-01 -7.07133859e-02
-1.80157796e-01 1.23815909e-01 7.30928838e-01 7.13111579e-01
-1.07029581e+00 1.17848396e+00 7.74859238e+00 8.21278036e-01
-1.01276135e+00 1.27279878e-01 9.64834094e-01 2.45884940e-01
-1.48264766e-01 2.26614192e-01 -7.96265364e-01 4.41672653e-01
2.25552499e-01 3.34231585e-01 2.68740565e-01 9.72555578e-01
1.43478215e-01 -8.17880452e-01 -1.08111989e+00 6.69330835e-01
-9.10422578e-02 -1.25599170e+00 -3.92206311e-02 -1.38122797e-01
1.01130390e+00 1.40979394e-01 -1.70359127e-02 -3.13835770e-01
4.90523517e-01 -8.32589269e-01 1.07703435e+00 7.55913317e-01
1.55170456e-01 -5.62775493e-01 6.20954096e-01 6.02260888e-01
-1.16376519e+00 -7.01192915e-02 -2.77842462e-01 1.38500780e-01
2.99502015e-01 1.07437253e+00 -7.48207033e-01 4.16114658e-01
9.54354644e-01 6.54072821e-01 -8.31432104e-01 1.35545754e+00
-3.54572147e-01 6.28917754e-01 -3.86397809e-01 3.30223829e-01
2.39564702e-01 -5.22361934e-01 6.15920305e-01 1.27368951e+00
6.37061000e-02 1.83321387e-01 6.81116581e-01 1.26972342e+00
2.74366081e-01 2.06926484e-02 -3.07530612e-01 4.19089645e-01
-4.65208218e-02 1.44263518e+00 -1.49596047e+00 -4.34790552e-01
-1.02014504e-01 1.17977965e+00 1.81254253e-01 5.10353506e-01
-8.19889009e-01 -1.43925369e-01 2.08990648e-01 3.57486695e-01
6.68689966e-01 -4.19690937e-01 -5.91334581e-01 -9.81819153e-01
-1.44695073e-01 -4.01686430e-01 2.94338554e-01 -8.45567346e-01
-9.76974189e-01 5.93219876e-01 -6.74571395e-02 -1.04609466e+00
2.65836954e-01 -4.65090930e-01 -4.66527581e-01 5.66733778e-01
-1.73246479e+00 -7.16363728e-01 -2.74894208e-01 3.84902090e-01
8.07315767e-01 4.28380430e-01 1.74250662e-01 1.24624312e-01
-4.67569083e-01 4.82481485e-03 2.55845375e-02 1.78002536e-01
-1.19276047e-02 -1.18242908e+00 -1.77926701e-02 1.13384402e+00
4.91522431e-01 4.24089164e-01 6.93373859e-01 -6.99828804e-01
-7.00189531e-01 -1.05015886e+00 4.76107359e-01 -3.81919712e-01
3.39578032e-01 -1.10822223e-01 -8.50105286e-01 3.49520624e-01
1.82962343e-01 7.02428073e-02 4.31095622e-02 -1.07872382e-01
2.86669396e-02 -1.34909898e-01 -1.22227216e+00 4.54514086e-01
9.36755538e-01 -1.70975044e-01 -1.80575103e-01 3.13216388e-01
5.01835048e-01 -3.09765339e-01 -5.17963171e-01 6.09136701e-01
7.06484765e-02 -1.19388163e+00 1.03877938e+00 -9.18536484e-02
1.24955855e-01 -8.32759619e-01 1.32399797e-01 -1.04616785e+00
-5.18249094e-01 -2.68072963e-01 6.40907511e-02 8.59229743e-01
3.94331068e-01 -1.70296341e-01 9.79060590e-01 9.73482728e-02
-6.23236857e-02 -4.84599918e-01 -8.56569588e-01 -7.02580333e-01
-3.57052892e-01 -4.25267369e-01 2.29121402e-01 6.40533447e-01
-6.02888107e-01 2.00859115e-01 3.07755936e-02 5.48857689e-01
8.42019618e-01 5.84280431e-01 5.85964620e-01 -1.36974847e+00
-3.78976822e-01 -6.03950441e-01 -5.10247469e-01 -1.53552437e+00
-9.78423655e-02 -6.11111820e-01 4.47217911e-01 -1.72110021e+00
3.67027044e-01 -6.16147578e-01 -2.62314379e-01 3.50645274e-01
-2.74447650e-01 6.00427985e-01 1.23130232e-01 3.45385194e-01
-1.12308824e+00 1.46289423e-01 1.47107661e+00 -2.06863165e-01
-3.00617576e-01 1.27115056e-01 -3.75400126e-01 1.22399759e+00
5.38688064e-01 -6.20262980e-01 -1.46363303e-01 -5.09160638e-01
-8.50400701e-02 -2.10732013e-01 4.60160822e-01 -1.31823850e+00
1.59881070e-01 -3.32041681e-01 5.77509642e-01 -8.37133884e-01
3.62263739e-01 -1.04858136e+00 2.21253842e-01 5.53643346e-01
-2.34152928e-01 -3.29250455e-01 7.59537444e-02 5.08513093e-01
-2.21034903e-02 -5.96665144e-01 1.22357798e+00 -2.82352060e-01
-9.02510047e-01 8.10268819e-02 -3.19582462e-01 -3.78683358e-01
1.22897243e+00 -6.43043995e-01 3.78563970e-01 -2.65160173e-01
-9.87095714e-01 6.21456839e-02 7.46141374e-01 -9.47662070e-02
5.38136959e-01 -9.19290841e-01 -6.31484926e-01 1.11987531e-01
-1.54138356e-01 3.64674538e-01 -2.23748926e-02 9.38329220e-01
-8.22973907e-01 3.32749151e-02 3.64152230e-02 -1.17916012e+00
-9.59209263e-01 5.66147923e-01 5.27313232e-01 -2.42926016e-01
-7.23768055e-01 9.26868498e-01 6.73848987e-01 -8.65561515e-02
3.11103553e-01 -5.14861107e-01 -2.32663333e-01 -2.56558508e-01
3.45236182e-01 3.88477236e-01 -1.34353057e-01 -7.99951375e-01
-3.05693597e-01 9.46727455e-01 2.58284202e-03 -9.77044776e-02
1.24072123e+00 -4.52305853e-01 -2.67233968e-01 5.19795060e-01
8.33353519e-01 1.86789721e-01 -1.60405767e+00 -2.60287076e-01
1.08098105e-01 -8.65823686e-01 3.24026138e-01 -7.38158643e-01
-1.55625188e+00 5.70525646e-01 6.32253349e-01 2.38772377e-01
1.34079289e+00 3.82246614e-01 6.32349849e-01 3.96002689e-03
4.44382727e-01 -1.06222343e+00 2.03098103e-01 4.53152090e-01
3.84392202e-01 -9.42150354e-01 1.63247809e-01 -7.62817979e-01
-5.31868458e-01 9.66521442e-01 4.17652518e-01 -2.07458347e-01
5.17012358e-01 4.60456699e-01 -3.81057747e-02 -2.71373242e-01
8.09337420e-04 -6.13997400e-01 3.51923108e-01 7.61720181e-01
4.53693904e-02 -1.84503183e-01 -1.60315096e-01 1.92175671e-01
2.45567784e-01 -3.32748026e-01 2.61138618e-01 5.90696990e-01
-9.46102381e-01 -8.85713875e-01 -6.67957544e-01 7.84010664e-02
-2.48756543e-01 1.20266117e-01 -4.37008709e-01 3.91876101e-01
5.86307228e-01 1.01344001e+00 1.88019648e-01 -1.17135849e-02
2.11275995e-01 -4.42494750e-01 5.66597939e-01 -6.63666368e-01
-5.06457806e-01 3.38050455e-01 -3.71874481e-01 -7.95686662e-01
-6.39947593e-01 -6.62280500e-01 -1.65863514e+00 5.62285408e-02
-5.63304603e-01 2.06191421e-01 7.31941223e-01 7.74147809e-01
2.50646174e-01 6.57969594e-01 8.04534316e-01 -1.03703105e+00
2.85039663e-01 -3.59536260e-01 -6.80116296e-01 2.03687862e-01
3.43567878e-01 -5.10386109e-01 -2.44506121e-01 3.17317933e-01] | [9.236400604248047, -0.2590867877006531] |
ce01fae8-aef6-4758-933a-0b2d8dc565b6 | fidelity-weighted-learning | 1711.02799 | null | http://arxiv.org/abs/1711.02799v2 | http://arxiv.org/pdf/1711.02799v2.pdf | Fidelity-Weighted Learning | Training deep neural networks requires many training samples, but in practice
training labels are expensive to obtain and may be of varying quality, as some
may be from trusted expert labelers while others might be from heuristics or
other sources of weak supervision such as crowd-sourcing. This creates a
fundamental quality versus-quantity trade-off in the learning process. Do we
learn from the small amount of high-quality data or the potentially large
amount of weakly-labeled data? We argue that if the learner could somehow know
and take the label-quality into account when learning the data representation,
we could get the best of both worlds. To this end, we propose
"fidelity-weighted learning" (FWL), a semi-supervised student-teacher approach
for training deep neural networks using weakly-labeled data. FWL modulates the
parameter updates to a student network (trained on the task we care about) on a
per-sample basis according to the posterior confidence of its label-quality
estimated by a teacher (who has access to the high-quality labels). Both
student and teacher are learned from the data. We evaluate FWL on two tasks in
information retrieval and natural language processing where we outperform
state-of-the-art alternative semi-supervised methods, indicating that our
approach makes better use of strong and weak labels, and leads to better
task-dependent data representations. | ['Bernhard Schölkopf', 'Stephan Gouws', 'Mostafa Dehghani', 'Jaap Kamps', 'Arash Mehrjou'] | 2017-11-08 | fidelity-weighted-learning-1 | https://openreview.net/forum?id=B1X0mzZCW | https://openreview.net/pdf?id=B1X0mzZCW | iclr-2018-1 | ['ad-hoc-information-retrieval'] | ['natural-language-processing'] | [ 1.01158887e-01 4.56671417e-01 -6.90683603e-01 -9.72558081e-01
-1.15868855e+00 -6.21515334e-01 6.39935493e-01 5.35895765e-01
-1.02942216e+00 9.23381984e-01 9.85804796e-02 -1.05832070e-01
-2.96471208e-01 -8.67592633e-01 -1.01513231e+00 -8.27285230e-01
3.40283036e-01 1.09282708e+00 6.52464852e-02 -5.50147481e-02
3.16671990e-02 2.48884648e-01 -1.52222192e+00 2.66823828e-01
8.56340468e-01 1.31234741e+00 -6.12029573e-04 3.86530489e-01
-1.62383541e-01 1.01695335e+00 -4.89046067e-01 -4.75033313e-01
3.87420386e-01 -9.87191647e-02 -1.02706003e+00 -5.37401475e-02
4.96259153e-01 -2.55534351e-01 2.72452801e-01 1.15759313e+00
3.59887391e-01 1.82427436e-01 8.63170147e-01 -1.14801240e+00
-7.18657255e-01 9.06194508e-01 -3.27453494e-01 9.43449289e-02
-1.10602036e-01 1.06057510e-01 1.24301124e+00 -8.03601682e-01
5.30064225e-01 1.17015910e+00 4.47906554e-01 6.63029075e-01
-1.39113390e+00 -5.33317924e-01 2.43155360e-01 -1.14318080e-01
-9.57826257e-01 -4.95936006e-01 5.13600111e-01 -5.45359612e-01
4.09642875e-01 -2.35988066e-01 3.20273012e-01 1.11194241e+00
-2.33692333e-01 1.02727032e+00 1.31455159e+00 -5.60964704e-01
4.85810697e-01 7.59512305e-01 3.32490742e-01 7.11375237e-01
2.26755053e-01 2.75235683e-01 -7.27395117e-01 -3.77828240e-01
4.04197782e-01 -7.85426330e-03 -3.46533686e-01 -5.04272044e-01
-1.00807381e+00 1.05520058e+00 6.81182683e-01 3.20145637e-01
-5.01415253e-01 1.36061236e-01 3.56576115e-01 6.30703151e-01
1.03513730e+00 6.80338562e-01 -8.90430093e-01 -1.05136679e-03
-1.10620034e+00 1.81035325e-01 8.19003582e-01 6.30496323e-01
1.03946722e+00 -1.72943532e-01 -2.28543833e-01 9.15241778e-01
6.59032047e-01 4.06158954e-01 5.50496757e-01 -9.27768350e-01
2.34814495e-01 4.74428803e-01 2.80523598e-01 -3.73187810e-01
-2.38519803e-01 -2.97657669e-01 -4.75062221e-01 4.59999263e-01
7.61231124e-01 -1.63328335e-01 -1.03838396e+00 1.87288082e+00
5.80511987e-01 -2.09640618e-02 -4.46527302e-02 1.02626073e+00
5.95965087e-01 5.25943637e-01 1.78410336e-01 -1.45633280e-01
1.17942572e+00 -1.02621484e+00 -5.13509750e-01 -2.11973086e-01
8.16492498e-01 -4.13835824e-01 1.03469753e+00 8.33184063e-01
-1.04198134e+00 -5.62052310e-01 -8.87441516e-01 -1.94571912e-01
-4.21276093e-01 2.91706994e-02 4.03023750e-01 6.22973263e-01
-9.96471107e-01 9.34468567e-01 -6.33899093e-01 1.04347602e-01
7.79893517e-01 2.92006344e-01 -3.06929946e-01 -2.35914484e-01
-1.16871083e+00 9.85016763e-01 3.54225039e-01 -1.57800063e-01
-1.39441216e+00 -8.03289294e-01 -5.69657445e-01 1.47394404e-01
5.51494360e-01 -4.98446137e-01 1.60139382e+00 -1.57748044e+00
-1.41484332e+00 1.20021212e+00 1.12998880e-01 -4.92250085e-01
7.39834130e-01 -3.09649050e-01 1.11956008e-01 5.84727302e-02
6.09283447e-02 8.80471110e-01 9.81240749e-01 -1.43264151e+00
-7.82271028e-01 -4.48482692e-01 1.20249569e-01 -7.48994900e-03
-3.10717702e-01 -1.40550390e-01 5.23724779e-02 -2.68894970e-01
-2.58615375e-01 -8.66210401e-01 -5.14576808e-02 3.58485460e-01
-8.40384960e-02 -8.32463980e-01 2.96084642e-01 -1.79350704e-01
7.98755825e-01 -1.87428129e+00 4.08745883e-03 2.44991019e-01
4.20295209e-01 5.71977794e-01 -2.16818988e-01 1.25690624e-01
-3.09581868e-02 1.90617085e-01 -2.25799568e-02 -6.35577798e-01
1.76378474e-01 5.07985830e-01 -2.10803479e-01 5.28356016e-01
4.31800127e-01 7.00158596e-01 -1.30965877e+00 -4.50891674e-01
-1.89119309e-01 4.23512787e-01 -2.94821948e-01 5.67724884e-01
-6.09029591e-01 3.52844685e-01 -5.08695185e-01 3.13601613e-01
3.69022489e-01 -4.30726051e-01 -1.00355238e-01 2.73822211e-02
2.93608725e-01 6.57076538e-01 -1.19794166e+00 1.65606189e+00
-6.01836085e-01 3.46893042e-01 1.40175611e-01 -1.12450480e+00
7.71065533e-01 6.37390971e-01 2.38056436e-01 -6.65088654e-01
2.75945514e-01 3.99172008e-01 -2.27639347e-01 -5.28509915e-01
2.42105901e-01 -5.05691648e-01 2.90894330e-01 9.62298214e-01
5.51867366e-01 -1.82609573e-01 8.92762691e-02 3.69700432e-01
9.51730728e-01 1.53944939e-01 1.99004158e-01 -3.69346112e-01
1.38322741e-01 -1.01539090e-01 6.03722930e-01 8.31826568e-01
-2.73546040e-01 3.37309629e-01 5.83241463e-01 -4.78492230e-01
-9.88212705e-01 -6.86879694e-01 -2.82907277e-01 1.79957187e+00
-3.91739994e-01 -8.29050094e-02 -6.69025123e-01 -1.18560326e+00
3.77892740e-02 6.55900478e-01 -9.11648095e-01 -3.20015490e-01
-1.06663130e-01 -3.38298589e-01 2.58867085e-01 4.95845407e-01
-5.61729595e-02 -1.11669874e+00 -4.78722274e-01 3.28057051e-01
3.20534319e-01 -7.55347788e-01 -2.46157557e-01 8.03434372e-01
-7.17513680e-01 -9.32302296e-01 -5.70709705e-01 -5.04143536e-01
7.58086383e-01 5.07636815e-02 1.56691372e+00 3.18952471e-01
3.24681789e-01 1.77215144e-01 -3.45350891e-01 -5.80880702e-01
-4.61865783e-01 6.41767830e-02 2.92107612e-01 1.31614581e-01
5.95068514e-01 -4.69207883e-01 -3.24255168e-01 1.53934896e-01
-1.09266555e+00 -1.75913915e-01 5.40374279e-01 1.09418917e+00
4.44531649e-01 7.12459087e-02 7.12925732e-01 -1.65500546e+00
6.45292222e-01 -5.71098149e-01 -7.39668667e-01 3.29695165e-01
-9.81297672e-01 5.70782661e-01 6.12328053e-01 -8.50606263e-01
-1.04888940e+00 -3.30576263e-02 -5.85529208e-02 -4.49846298e-01
-3.25407177e-01 5.54656148e-01 -6.60766214e-02 1.06913194e-01
1.09485734e+00 -3.97978902e-01 -1.01923935e-01 -5.33997715e-01
5.36033809e-01 6.50635183e-01 1.49257451e-01 -1.22590196e+00
5.74204922e-01 2.82245249e-01 -3.24951828e-01 -2.09807325e-02
-1.75518417e+00 -3.45499218e-01 -5.86508691e-01 -1.13261528e-01
6.19430900e-01 -9.19491112e-01 -5.66212952e-01 2.95511693e-01
-1.04921412e+00 -7.22611070e-01 -7.82865226e-01 3.56054723e-01
-2.41005331e-01 -3.48145753e-01 -5.01225650e-01 -9.13329780e-01
-8.70141834e-02 -1.21331775e+00 1.10853481e+00 2.19664454e-01
-9.09841992e-03 -1.24415672e+00 2.37769008e-01 4.48662013e-01
5.55188656e-01 -2.04825439e-02 8.71882558e-01 -1.31122696e+00
-3.28486830e-01 -2.65146017e-01 -2.47551486e-01 7.25338459e-01
-5.43179587e-02 -9.34809819e-02 -1.50474179e+00 -2.59805471e-01
9.62012187e-02 -1.41085804e+00 9.82830882e-01 1.28395826e-01
9.74963069e-01 -4.42906141e-01 2.53741462e-02 9.17098820e-02
1.30277193e+00 -5.19175768e-01 -8.28273445e-02 1.38760552e-01
7.18343854e-01 1.10323894e+00 5.64111412e-01 3.58990192e-01
3.69803429e-01 4.39509094e-01 3.44809592e-01 6.76041143e-03
1.08981468e-01 -2.67209470e-01 2.80550927e-01 6.66225135e-01
1.73607633e-01 -1.59919187e-01 -9.09068763e-01 5.15214443e-01
-1.94440210e+00 -5.47857285e-01 5.42690419e-02 2.24962616e+00
1.59171617e+00 4.68503267e-01 2.53956467e-01 1.78057000e-01
3.83985847e-01 3.77035625e-02 -6.82348609e-01 -2.90628552e-01
3.60968828e-01 2.84240574e-01 5.09215772e-01 6.93794608e-01
-8.86301577e-01 7.30880201e-01 5.29315519e+00 9.52450931e-01
-1.22166061e+00 5.03441513e-01 9.45530891e-01 -1.45181373e-01
-4.84798700e-01 -9.74710658e-02 -8.03592145e-01 4.00332421e-01
1.28878045e+00 2.29328960e-01 2.95149773e-01 1.03107619e+00
-2.07693297e-02 -2.76254833e-01 -1.67910945e+00 6.51896358e-01
-9.38121676e-02 -1.10147119e+00 -2.34897390e-01 2.77323931e-01
9.75656509e-01 4.23259348e-01 -7.52379894e-02 4.92230892e-01
9.97785926e-01 -1.10266268e+00 9.97242689e-01 6.89715624e-01
4.45057184e-01 -6.38654828e-01 8.77518594e-01 9.59142387e-01
-5.26519001e-01 -1.53952064e-02 -4.38691139e-01 6.44947812e-02
-1.21603027e-01 1.19447923e+00 -7.83086836e-01 1.18971407e-01
4.80981588e-01 5.57283998e-01 -5.00535548e-01 6.68931603e-01
-5.25157273e-01 9.73037183e-01 -3.79751384e-01 -1.23587817e-01
4.10162151e-01 -4.83111367e-02 -1.22090563e-01 1.07423818e+00
-1.52439237e-01 5.92163764e-03 4.93129075e-01 9.15993094e-01
-5.86406767e-01 -1.00347325e-01 -4.29317683e-01 -4.01576571e-02
4.42475587e-01 1.38765466e+00 -4.19375777e-01 -6.21065497e-01
-2.43614584e-01 2.64636159e-01 8.91280472e-01 1.79655150e-01
-1.86664164e-01 1.66870698e-01 1.65265098e-01 6.78859577e-02
-1.08607905e-02 2.36192018e-01 -1.10355899e-01 -1.00876296e+00
-1.48325577e-01 -1.06962252e+00 3.90190899e-01 -7.41118729e-01
-1.78932011e+00 5.30031741e-01 -6.80017024e-02 -8.31524849e-01
-4.23350602e-01 -8.19092453e-01 -3.34287673e-01 8.20797324e-01
-1.96751785e+00 -8.65341425e-01 8.53924826e-02 4.02778119e-01
3.31789255e-01 -7.03621656e-02 7.59684861e-01 2.37151965e-01
-1.10999525e-01 4.62536067e-01 1.81640889e-02 1.02448665e-01
9.97622728e-01 -1.61398804e+00 -1.38568267e-01 2.24278584e-01
5.46975195e-01 4.30964142e-01 5.53239465e-01 -2.21474081e-01
-9.84631717e-01 -8.75496924e-01 1.00626183e+00 -7.63085604e-01
6.89969718e-01 -3.24513346e-01 -1.35515630e+00 4.64742631e-01
1.94179401e-01 5.67231357e-01 7.82901943e-01 3.79727006e-01
-8.74381781e-01 -1.79325223e-01 -1.50122082e+00 -1.22711539e-01
5.78846812e-01 -7.39522398e-01 -6.95886493e-01 6.82784796e-01
6.74920261e-01 -1.25758514e-01 -8.77723336e-01 1.44490317e-01
2.65149713e-01 -8.75043988e-01 4.96582925e-01 -8.27728391e-01
4.60312545e-01 -7.63498545e-02 2.22494360e-02 -1.77685726e+00
-6.69142157e-02 -1.41396403e-01 -3.71057130e-02 1.38153768e+00
6.35717213e-01 -4.74103749e-01 7.46104062e-01 1.05603850e+00
1.39427707e-01 -9.77325678e-01 -7.43339777e-01 -4.63604301e-01
3.66870672e-01 -4.35767174e-01 5.13986051e-01 1.18716943e+00
-2.32935041e-01 5.53141177e-01 -2.25181684e-01 -7.88212940e-02
5.46676397e-01 -2.34460250e-01 5.10049999e-01 -1.73418403e+00
-3.78138423e-01 -3.12347382e-01 7.33342618e-02 -8.52848709e-01
4.77578312e-01 -8.55188012e-01 4.49163884e-01 -1.16547787e+00
5.38323820e-02 -8.89952302e-01 -5.80822945e-01 7.89155066e-01
-2.73845136e-01 1.42367765e-01 -1.01267301e-01 1.67008519e-01
-9.54727411e-01 4.36854035e-01 1.09902442e+00 -8.42987373e-02
1.59709990e-01 4.57194401e-03 -8.45813930e-01 8.06063235e-01
5.01872063e-01 -1.15893662e+00 -4.35321033e-01 -5.93859017e-01
8.34703982e-01 -9.27877426e-02 1.47172898e-01 -5.96857727e-01
3.46629500e-01 -2.04607412e-01 2.36253932e-01 9.01643485e-02
1.03894375e-01 -1.07628369e+00 -4.94095057e-01 6.15576990e-02
-1.06830204e+00 -4.16992724e-01 -5.23852348e-01 6.20851696e-01
-1.08723067e-01 -7.59690344e-01 9.68021750e-01 -3.60988289e-01
1.40158264e-02 3.67210120e-01 -3.96903567e-02 4.63627636e-01
4.46831316e-01 4.08144802e-01 -4.23384637e-01 -3.31558943e-01
-4.57291156e-01 3.26016188e-01 3.88672650e-01 2.07427487e-01
2.42579922e-01 -1.32530034e+00 -7.98542082e-01 5.55616319e-02
4.20710564e-01 4.75450546e-01 -3.47427309e-01 5.54139435e-01
4.77782078e-02 1.11596070e-01 1.31920546e-01 -5.78204155e-01
-6.70478582e-01 6.24971032e-01 2.43013322e-01 -4.27629054e-01
-8.02523941e-02 1.19135892e+00 -1.45916849e-01 -8.25413823e-01
6.73758686e-01 -4.15723890e-01 -1.66278541e-01 4.20505345e-01
7.19340324e-01 -4.96815778e-02 2.69597322e-01 -5.01997590e-01
1.13141921e-03 1.27702862e-01 -3.02846879e-01 -2.03237176e-01
1.52102137e+00 1.94564521e-01 -1.48566533e-02 8.58965337e-01
1.28017569e+00 -1.59460559e-01 -1.51394320e+00 -1.01183224e+00
3.60508233e-01 -4.63530511e-01 3.56624901e-01 -9.89476860e-01
-1.24624288e+00 1.10319066e+00 5.81096709e-01 4.46303099e-01
6.68691397e-01 1.90605804e-01 5.44437528e-01 6.91594303e-01
4.95194376e-01 -1.37828624e+00 3.73031944e-01 2.06101552e-01
5.09143353e-01 -1.76637208e+00 2.69684512e-02 3.07436347e-01
-7.72916615e-01 7.94598281e-01 5.80650330e-01 -2.89708853e-01
8.34567726e-01 7.21985847e-02 4.32180494e-01 -2.75374413e-01
-1.27821064e+00 -1.38882577e-01 5.78172982e-01 4.32604194e-01
6.21110618e-01 7.50936046e-02 1.17008045e-01 4.67655450e-01
3.51791680e-02 3.63070592e-02 1.34355336e-01 8.68192673e-01
-7.27026701e-01 -1.34326911e+00 -2.36913428e-01 6.79446816e-01
-4.73960310e-01 -2.14075930e-02 -3.28702390e-01 2.96948344e-01
5.26682317e-01 9.40790057e-01 -2.13887155e-01 8.39003325e-02
8.71415585e-02 2.99344182e-01 3.35038424e-01 -1.01135790e+00
-9.02999520e-01 -1.15710422e-01 1.44100741e-01 -5.39978385e-01
-8.08665931e-01 -3.60526443e-01 -1.07487488e+00 -3.12821157e-02
-4.86352772e-01 5.25690556e-01 8.13314617e-01 1.28893006e+00
1.48250591e-02 3.08539540e-01 6.10633969e-01 -9.22067165e-01
-9.72781956e-01 -1.12377262e+00 -6.56608820e-01 6.19607031e-01
6.28717303e-01 -7.32011795e-01 -5.94479322e-01 -5.88539355e-02] | [9.434064865112305, 3.788414478302002] |
f02ebde5-e6da-4692-b9dc-a49f09652784 | learning-towards-selective-data-augmentation | 2303.09719 | null | https://arxiv.org/abs/2303.09719v1 | https://arxiv.org/pdf/2303.09719v1.pdf | Learning towards Selective Data Augmentation for Dialogue Generation | As it is cumbersome and expensive to acquire a huge amount of data for training neural dialog models, data augmentation is proposed to effectively utilize existing training samples. However, current data augmentation techniques on the dialog generation task mostly augment all cases in the training dataset without considering the intrinsic attributes between different cases. We argue that not all cases are beneficial for augmentation task, and the cases suitable for augmentation should obey the following two attributes: (1) low-quality (the dialog model cannot generate a high-quality response for the case), (2) representative (the case should represent the property of the whole dataset). Herein, we explore this idea by proposing a Selective Data Augmentation framework (SDA) for the response generation task. SDA employs a dual adversarial network to select the lowest quality and most representative data points for augmentation in one stage. Extensive experiments conducted on two publicly available datasets, \ie DailyDialog and OpenSubtitles, show that our framework can improve the response generation performance with respect to various metrics. | ['Rui Yan', 'Xiangliang Zhang', 'Xin Gao', 'Jianwei Cui', 'Chen Wei', 'Xiaoqiang Xia', 'Jiayi Zhang', 'Mingzhe Li', 'Xiuying Chen'] | 2023-03-17 | null | null | null | null | ['dialogue-generation', 'response-generation', 'dialogue-generation'] | ['natural-language-processing', 'natural-language-processing', 'speech'] | [ 2.01734021e-01 3.52618933e-01 -2.03710303e-01 -6.48515642e-01
-7.88911343e-01 -5.95158875e-01 7.77016521e-01 -7.01666102e-02
-4.56709862e-01 1.02028656e+00 3.12968314e-01 -2.35453770e-01
2.52855867e-01 -9.00462568e-01 -3.48882347e-01 -5.80600917e-01
4.85637039e-01 8.83843839e-01 6.55299649e-02 -6.45155489e-01
-1.10493489e-01 2.12273672e-01 -1.15366828e+00 2.19758615e-01
1.26843786e+00 1.07633483e+00 -4.01396677e-02 2.20345125e-01
-3.00547957e-01 7.02823400e-01 -8.47865403e-01 -6.84911132e-01
3.59157324e-01 -7.33923197e-01 -1.05718684e+00 3.74490283e-02
5.09730428e-02 -6.34720922e-01 -2.38707438e-01 8.97889912e-01
5.23117065e-01 2.21186712e-01 5.91454268e-01 -1.52954245e+00
-5.82417190e-01 8.56876135e-01 -1.99011594e-01 7.25810155e-02
4.26929444e-01 1.75485834e-01 9.67991710e-01 -9.37946975e-01
5.90714514e-01 1.04033899e+00 2.99338698e-01 1.18294203e+00
-1.17878890e+00 -7.42123961e-01 2.54986078e-01 -2.47357607e-01
-1.06781411e+00 -6.21435225e-01 8.78542244e-01 -2.37693131e-01
2.49449894e-01 3.43948573e-01 2.79053748e-01 1.46085787e+00
-4.98732060e-01 8.83511364e-01 1.25368929e+00 -1.79184377e-01
2.55457848e-01 6.68218136e-01 3.67803067e-01 2.13779598e-01
-2.10858718e-01 -1.29930124e-01 -4.06451970e-01 -3.36446911e-01
6.44799948e-01 -1.91285491e-01 -3.49323362e-01 -6.80175647e-02
-1.08318424e+00 1.06023598e+00 3.22186023e-01 2.21446648e-01
-3.53572637e-01 -6.01491749e-01 3.66111070e-01 6.13808453e-01
4.32685256e-01 5.77777922e-01 -4.42684472e-01 -8.82021710e-02
-4.05061632e-01 5.26811361e-01 8.60225201e-01 1.10362291e+00
7.97145426e-01 9.51303809e-04 -5.20423055e-01 1.06355536e+00
9.53241736e-02 2.54057199e-01 7.89346218e-01 -7.41176069e-01
8.64305139e-01 1.04583585e+00 3.20297956e-01 -5.04235804e-01
-2.30516464e-01 4.61181737e-02 -1.00996196e+00 -1.53501526e-01
6.61771178e-01 -5.57047129e-01 -5.24723709e-01 2.20638824e+00
4.59212005e-01 -3.32645684e-01 3.87205750e-01 9.39898431e-01
1.28265703e+00 5.30137300e-01 1.87707633e-01 -3.21394563e-01
1.28810155e+00 -8.19119155e-01 -9.72992837e-01 -3.54806423e-01
7.75294960e-01 -4.83940125e-01 1.47617209e+00 -2.62311958e-02
-1.14353669e+00 -5.82067251e-01 -1.04549766e+00 5.74939651e-03
-3.79014730e-01 1.60518378e-01 7.05744684e-01 6.93994820e-01
-7.98636913e-01 1.60023034e-01 -7.75572136e-02 -1.69480160e-01
1.03198707e-01 4.32233870e-01 -4.58195955e-01 2.35867083e-01
-1.65145183e+00 7.03411877e-01 3.66087288e-01 -8.19897503e-02
-8.14783573e-01 -5.61599851e-01 -6.49975955e-01 -5.93535183e-03
3.59395087e-01 -4.87850875e-01 1.36864448e+00 -9.82661188e-01
-1.59338486e+00 7.84764171e-01 4.15196130e-03 -4.69148636e-01
8.06424141e-01 -1.68412939e-01 -3.25940907e-01 -1.67260900e-01
-1.53763160e-01 9.21587944e-01 5.66993356e-01 -1.43486512e+00
-3.67526948e-01 -2.65178561e-01 4.89423364e-01 3.43369693e-01
-6.57272935e-01 -1.61116645e-01 -7.27258548e-02 -5.60402930e-01
3.99515741e-02 -1.00958514e+00 -1.73380256e-01 -3.56913745e-01
-6.69307530e-01 -4.71945584e-01 7.35723197e-01 -4.39492673e-01
1.25107145e+00 -1.87639332e+00 -1.68274436e-02 -1.08836308e-01
1.14936560e-01 4.26458150e-01 -1.18459605e-01 3.74230266e-01
-3.59303467e-02 1.76889360e-01 -1.63587227e-01 -5.13256609e-01
-5.55409119e-02 2.54254341e-01 -6.73509538e-01 -1.82383299e-01
3.90942454e-01 7.55351722e-01 -7.11663127e-01 -5.09886444e-01
1.15067614e-02 -1.61818732e-02 -5.61682820e-01 8.01758051e-01
-4.10104901e-01 8.50866139e-01 -7.04443157e-01 3.90542001e-01
6.49914920e-01 -1.94267496e-01 1.34100631e-01 1.53232468e-02
2.08224282e-01 4.82281893e-01 -1.06074595e+00 1.38755548e+00
-4.71508920e-01 6.66145906e-02 4.30371463e-02 -6.69121385e-01
1.26739526e+00 5.29937685e-01 3.63179535e-01 -5.09712994e-01
2.38748178e-01 1.51934639e-01 2.27608696e-01 -4.41352576e-01
8.40484619e-01 -6.88926727e-02 -3.98058623e-01 7.97554195e-01
3.50502543e-02 -3.87126021e-02 1.76231965e-01 4.70506221e-01
7.26805091e-01 -2.31893525e-01 9.00368690e-02 -9.96055752e-02
7.23903954e-01 -6.37483746e-02 7.25359380e-01 7.63201535e-01
-3.27845365e-01 6.34939075e-01 6.62999332e-01 -1.85957044e-01
-9.34102416e-01 -7.68749058e-01 4.49233279e-02 1.32556915e+00
6.37435764e-02 -9.51498449e-02 -9.22446966e-01 -1.08886027e+00
-3.17847222e-01 7.90463805e-01 -5.76667726e-01 -2.71106929e-01
-5.17006695e-01 -8.19703400e-01 6.51615977e-01 6.29033864e-01
7.04746783e-01 -1.24173665e+00 -1.29094958e-01 -1.02885149e-03
-6.45155311e-01 -1.26946902e+00 -5.35260677e-01 8.60359818e-02
-6.11255169e-01 -8.39952230e-01 -5.09707987e-01 -6.39362335e-01
7.38570631e-01 2.52649605e-01 1.02197242e+00 1.37565643e-01
4.89559948e-01 -1.50043651e-01 -5.81706345e-01 -2.95442104e-01
-9.26830888e-01 2.84855425e-01 1.31263733e-01 1.67577401e-01
4.80651706e-01 -4.51209873e-01 -4.18830246e-01 5.03101826e-01
-8.25650513e-01 1.32989660e-01 4.12746668e-01 1.07812560e+00
2.49730721e-01 -4.12596494e-01 1.26677930e+00 -1.16502976e+00
1.17291415e+00 -6.82774246e-01 -2.53603935e-01 3.11431199e-01
-5.67159116e-01 -8.63709152e-02 9.98350322e-01 -7.45614111e-01
-1.24245238e+00 -3.63333113e-02 -2.95502245e-01 -2.78987378e-01
-3.39823574e-01 3.64230812e-01 -5.22752523e-01 3.26274574e-01
7.16913164e-01 2.98425555e-01 2.38468468e-01 -3.89464229e-01
3.54493201e-01 8.44668508e-01 3.29855382e-01 -7.14687169e-01
6.83040440e-01 2.27116495e-01 -3.75649571e-01 -4.13956702e-01
-7.32485771e-01 -1.64258376e-01 -7.04910636e-01 -1.51609644e-01
6.97252333e-01 -6.89516306e-01 -6.12199306e-01 2.43902504e-01
-1.09687257e+00 -3.22574526e-01 -3.42751026e-01 2.59517163e-01
-3.34463596e-01 1.60262004e-01 -5.74962497e-01 -9.29173529e-01
-6.33760691e-01 -1.33929825e+00 6.63029730e-01 3.55455220e-01
-4.06124651e-01 -8.06879640e-01 2.48305639e-03 5.57016075e-01
3.45126867e-01 1.03839245e-02 1.11663127e+00 -1.44097614e+00
-3.22294623e-01 -3.52723897e-01 -8.26017931e-02 3.50811929e-01
1.98251471e-01 -2.51505882e-01 -1.20517504e+00 -9.93495286e-02
2.28589103e-01 -9.70925629e-01 4.65719402e-01 -1.40165582e-01
1.20580399e+00 -5.07679045e-01 -4.11112159e-02 4.60962616e-02
8.18433344e-01 3.72019500e-01 5.30387700e-01 9.08379536e-03
5.25498211e-01 8.99384141e-01 7.52230644e-01 4.21968222e-01
4.91222680e-01 7.82511413e-01 3.04613680e-01 -7.57440254e-02
1.52017459e-01 -3.85513753e-01 2.82003969e-01 7.31115341e-01
2.60396332e-01 -4.49366331e-01 -6.50447309e-01 4.45943356e-01
-1.85063672e+00 -7.63457835e-01 3.02822031e-02 2.25882673e+00
1.24141765e+00 1.34749547e-01 3.98737937e-01 9.69962701e-02
6.90330088e-01 8.74093324e-02 -5.81368029e-01 -3.26852530e-01
-1.64259151e-01 -2.40389928e-02 -1.67868212e-01 3.48625362e-01
-9.89975750e-01 9.14440811e-01 5.98335743e+00 6.38550401e-01
-8.37488651e-01 1.90039769e-01 8.63392949e-01 1.57025039e-01
-4.53614354e-01 -7.33043104e-02 -9.18041348e-01 5.28426886e-01
6.92522526e-01 -2.46304125e-01 1.79193705e-01 9.09662366e-01
3.51430774e-02 2.41890028e-01 -1.21136260e+00 4.18685853e-01
-2.08243921e-01 -9.66459572e-01 3.32880676e-01 -4.11289707e-02
5.70571780e-01 -4.09148008e-01 1.26776770e-01 6.93795264e-01
5.37435114e-01 -1.01772988e+00 2.50475436e-01 1.87948182e-01
6.19400561e-01 -5.30417025e-01 8.55958641e-01 6.43648446e-01
-6.97679281e-01 5.94051480e-02 -2.41576061e-01 1.72840506e-01
1.45457223e-01 1.73308820e-01 -1.12422419e+00 6.33335412e-01
2.92036533e-01 1.02051660e-01 -4.43930775e-01 4.20200318e-01
-1.76804304e-01 3.95141959e-01 -1.24887459e-01 -1.03717603e-01
2.14918926e-01 -4.39559281e-01 5.22484779e-01 7.33703196e-01
1.12759601e-02 2.44446293e-01 3.32947135e-01 9.11188662e-01
-3.64601076e-01 2.68266737e-01 -4.82083619e-01 -1.59956351e-01
8.38113010e-01 1.33403373e+00 -2.83937633e-01 -3.05942863e-01
-4.56101477e-01 6.26987338e-01 3.72411340e-01 1.81410760e-01
-7.14515328e-01 -1.03325956e-01 4.01704222e-01 -7.30198398e-02
-2.83108771e-01 1.65543556e-01 -4.60778356e-01 -1.04569674e+00
8.44993070e-02 -1.17696345e+00 6.13996744e-01 -4.65932816e-01
-1.52111292e+00 1.06559324e+00 -6.92825839e-02 -1.48741901e+00
-6.44896209e-01 -2.69873977e-01 -7.42591262e-01 1.08461571e+00
-1.15499926e+00 -1.20927250e+00 -6.47996008e-01 7.10442841e-01
7.04224944e-01 -3.86826217e-01 9.71454740e-01 3.77525687e-01
-7.39680767e-01 9.82082367e-01 -4.50563341e-01 3.35392445e-01
8.58043373e-01 -1.24076915e+00 2.20310882e-01 5.21750927e-01
-2.48249039e-01 8.50682020e-01 6.55590713e-01 -5.54359794e-01
-1.02037776e+00 -8.54175985e-01 8.18435788e-01 -3.83884043e-01
5.23208678e-01 -4.14092392e-01 -1.35157359e+00 7.16252208e-01
3.14223766e-01 -3.39196324e-01 9.19369042e-01 1.81135684e-01
-2.46883526e-01 1.02504492e-01 -1.31000066e+00 8.18111122e-01
7.20058739e-01 -3.64360452e-01 -5.54366946e-01 4.54626203e-01
8.77736390e-01 -4.31316584e-01 -9.81200218e-01 5.22190094e-01
2.96227843e-01 -7.73574948e-01 7.48883545e-01 -9.39946055e-01
7.04692125e-01 6.20948449e-02 -1.31769806e-01 -1.26119721e+00
2.03614891e-01 -7.00850368e-01 -1.83204994e-01 1.73314559e+00
5.52785218e-01 -5.83974004e-01 8.84710312e-01 1.13233387e+00
-7.24197626e-02 -8.93809319e-01 -7.82396555e-01 -5.60167313e-01
3.40407819e-01 9.07442998e-03 1.11898899e+00 1.29274333e+00
1.38641506e-01 7.95046091e-01 -6.31797671e-01 -1.40863284e-01
1.49075547e-02 2.98146397e-01 1.13738608e+00 -1.28014708e+00
-8.37911367e-02 -2.51814634e-01 2.21316800e-01 -1.39051902e+00
3.89061570e-01 -7.69565284e-01 -3.41870710e-02 -1.26097918e+00
8.85476619e-02 -9.35847580e-01 -4.51384597e-02 5.20078897e-01
-5.45823932e-01 -1.06270693e-01 4.99831587e-02 4.14312273e-01
-1.93717554e-01 8.64581406e-01 1.33134449e+00 4.02355827e-02
-5.33207715e-01 3.89786273e-01 -9.14835274e-01 4.45715070e-01
8.47376049e-01 -1.52522251e-01 -7.28326380e-01 -2.68435508e-01
-1.14817202e-01 3.64358574e-01 1.44957498e-01 -5.71490407e-01
4.03944626e-02 -2.76165396e-01 1.28361121e-01 -5.42213917e-01
6.78368449e-01 -7.87799180e-01 -1.75614491e-01 1.67655915e-01
-8.28286231e-01 -3.24703604e-02 8.35335478e-02 3.09901774e-01
-2.01584697e-01 -4.84879851e-01 7.67898440e-01 -1.36597445e-02
-4.29417104e-01 3.86262774e-01 -3.66101749e-02 2.77662665e-01
8.04958701e-01 9.45033878e-02 -6.43898129e-01 -7.65924871e-01
-6.51378095e-01 4.07312810e-01 3.03300589e-01 5.60464799e-01
4.05502915e-01 -1.39034200e+00 -7.38139153e-01 2.18698785e-01
2.78182775e-01 7.59234205e-02 2.75021464e-01 6.26351118e-01
1.21836610e-01 1.56074494e-01 -2.40629613e-01 -3.25678766e-01
-1.15479541e+00 5.05976617e-01 2.86494464e-01 -4.38128382e-01
-3.35393012e-01 8.05175483e-01 3.26898515e-01 -7.55578876e-01
1.53906122e-01 5.96054196e-02 -5.81739247e-01 2.64119387e-01
4.55268413e-01 8.30265954e-02 -3.71300094e-02 -6.36470377e-01
1.86955594e-02 -2.83919781e-01 -5.22496045e-01 -2.78348029e-01
9.04587567e-01 -1.48602307e-01 1.96386687e-02 3.74875486e-01
8.15394938e-01 -1.15980707e-01 -1.08211589e+00 -5.30884981e-01
-2.52436578e-01 -3.55077356e-01 -3.99322718e-01 -8.06282163e-01
-9.70577478e-01 7.32090294e-01 2.78414100e-01 6.33026540e-01
1.09233236e+00 -1.80065170e-01 8.82042766e-01 5.04929364e-01
1.86894655e-01 -1.12978756e+00 2.65432030e-01 5.14178276e-01
1.07887089e+00 -1.44748902e+00 -2.62782484e-01 -6.48300409e-01
-1.28321362e+00 7.98093796e-01 1.33688819e+00 1.94767311e-01
2.77892441e-01 -6.27554487e-03 2.31098607e-01 -8.53993185e-03
-8.45583677e-01 9.16219279e-02 9.90721583e-02 4.24503148e-01
5.23537695e-01 -9.75492001e-02 -4.10111815e-01 9.44029391e-01
-4.40719396e-01 -4.29649532e-01 6.28893375e-01 6.53394401e-01
-1.10180847e-01 -1.25630736e+00 1.85303367e-03 5.09994030e-01
-2.78976709e-01 2.47840974e-02 -7.76098371e-01 9.25094306e-01
-2.82041192e-01 1.18616962e+00 -2.93401871e-02 -6.11662567e-01
3.99558902e-01 4.83655781e-01 -6.26703128e-02 -5.22186697e-01
-9.35448587e-01 -1.95778400e-01 3.49326611e-01 -7.76726902e-02
-3.66607845e-01 -4.70944285e-01 -1.11176467e+00 -3.48773450e-01
-4.73760873e-01 3.66136700e-01 2.60677367e-01 1.06655145e+00
2.92795658e-01 3.73133183e-01 1.04899573e+00 -2.44460508e-01
-9.20485497e-01 -1.38410676e+00 -3.07362199e-01 7.84411490e-01
7.33525157e-02 -6.98767841e-01 -3.04547429e-01 -7.51789510e-02] | [12.660776138305664, 8.178911209106445] |
3c937030-0195-47a2-a4e0-45276e713fb4 | tart-improved-few-shot-text-classification | 2306.02175 | null | https://arxiv.org/abs/2306.02175v1 | https://arxiv.org/pdf/2306.02175v1.pdf | TART: Improved Few-shot Text Classification Using Task-Adaptive Reference Transformation | Meta-learning has emerged as a trending technique to tackle few-shot text classification and achieve state-of-the-art performance. However, the performance of existing approaches heavily depends on the inter-class variance of the support set. As a result, it can perform well on tasks when the semantics of sampled classes are distinct while failing to differentiate classes with similar semantics. In this paper, we propose a novel Task-Adaptive Reference Transformation (TART) network, aiming to enhance the generalization by transforming the class prototypes to per-class fixed reference points in task-adaptive metric spaces. To further maximize divergence between transformed prototypes in task-adaptive metric spaces, TART introduces a discriminative reference regularization among transformed prototypes. Extensive experiments are conducted on four benchmark datasets and our method demonstrates clear superiority over the state-of-the-art models in all the datasets. In particular, our model surpasses the state-of-the-art method by 7.4% and 5.4% in 1-shot and 5-shot classification on the 20 Newsgroups dataset, respectively. | ['Chang-Tien Lu', 'Fanglan Chen', 'Jianfeng He', 'Xuchao Zhang', 'Shuo Lei'] | 2023-06-03 | null | null | null | null | ['few-shot-text-classification'] | ['natural-language-processing'] | [ 1.33100793e-01 -4.12544608e-01 -3.29697102e-01 -6.50110900e-01
-8.09401989e-01 7.83826858e-02 6.24585271e-01 3.19049180e-01
-6.34637535e-01 4.78910744e-01 1.00504331e-01 2.94061691e-01
-3.62037897e-01 -6.30576611e-01 -1.48531631e-01 -5.64008474e-01
3.60559046e-01 5.02269685e-01 6.76442206e-01 -4.63891655e-01
5.30260086e-01 -2.68917948e-01 -1.93807638e+00 5.99215150e-01
9.43112731e-01 1.13821399e+00 2.36583352e-02 2.63103098e-01
-4.80233878e-01 4.46840823e-01 -5.11189520e-01 -3.39413166e-01
1.62090175e-02 -5.45618534e-01 -5.44978917e-01 6.56903014e-02
2.87073463e-01 1.30993500e-01 -2.54165202e-01 1.20875919e+00
5.20654261e-01 6.29383445e-01 9.11287904e-01 -1.31094229e+00
-9.50587988e-01 6.82285666e-01 -6.89344287e-01 2.67485499e-01
3.59529927e-02 -2.42616281e-01 1.25135601e+00 -1.30380714e+00
4.43563491e-01 1.18877649e+00 6.29010201e-01 5.92410684e-01
-1.13667774e+00 -7.48780668e-01 3.29643011e-01 6.68990016e-01
-1.51681376e+00 -4.62689400e-01 8.12744379e-01 -4.59586084e-01
7.81845391e-01 1.07834317e-01 1.52387589e-01 1.04980326e+00
1.24555863e-01 7.97171891e-01 8.76316845e-01 -4.98943299e-01
6.27276361e-01 2.97983676e-01 7.90082812e-01 3.20368797e-01
7.52427429e-02 -3.71809661e-01 -4.88571584e-01 -2.41864935e-01
-3.80521938e-02 4.51111913e-01 -2.21072406e-01 -5.76865375e-01
-9.65293407e-01 1.09718990e+00 2.76969701e-01 5.78991711e-01
-1.24536984e-01 -4.47816670e-01 7.19561577e-01 3.07416946e-01
8.76124561e-01 1.59113839e-01 -3.29155356e-01 -3.62693109e-02
-9.27525043e-01 8.78233686e-02 5.42383075e-01 8.96276236e-01
4.66219634e-01 2.37416420e-02 -6.48485184e-01 1.37444425e+00
1.25031844e-01 -8.02370235e-02 1.12715900e+00 -3.69024456e-01
5.37505865e-01 8.01449597e-01 -1.94642711e-02 -9.08610404e-01
-3.98702234e-01 -6.23556733e-01 -8.08112741e-01 -1.27654523e-01
2.08530962e-01 6.77056462e-02 -9.92284060e-01 1.52139282e+00
3.80746663e-01 4.71322805e-01 2.06490591e-01 7.65702605e-01
9.04616833e-01 7.75047421e-01 9.19710770e-02 -4.34962451e-01
1.16693449e+00 -1.18680799e+00 -7.55718350e-01 -3.13586593e-01
7.46418893e-01 -5.33426404e-01 1.41693807e+00 1.88212290e-01
-6.38791382e-01 -7.23794997e-01 -1.28451276e+00 2.18696356e-01
-5.01820505e-01 -1.65693894e-01 1.47089258e-01 7.24543333e-01
-3.36220384e-01 6.71964228e-01 -4.36071008e-01 -4.81169760e-01
5.58581054e-01 -7.65387490e-02 -1.08040467e-01 -2.14637995e-01
-1.35882235e+00 8.30370784e-01 6.74916685e-01 -5.01509130e-01
-6.62583649e-01 -9.52306449e-01 -6.86603844e-01 2.32080266e-01
5.01689315e-01 -3.71398568e-01 1.24319243e+00 -5.23256004e-01
-1.49014914e+00 8.17783833e-01 1.16172023e-01 -2.93865949e-01
5.76465011e-01 -2.71438751e-02 -7.57490814e-01 -1.40487164e-01
3.81845161e-02 2.32606024e-01 9.07062948e-01 -9.07675266e-01
-1.02213418e+00 -5.42659461e-01 -3.29385787e-01 2.20330790e-01
-1.01613355e+00 -4.84447666e-02 -1.46665841e-01 -6.07174933e-01
9.85514149e-02 -6.99833989e-01 -3.66649851e-02 -7.58867562e-02
-1.04704663e-01 -6.08765364e-01 1.07930958e+00 -8.60251784e-02
1.50755250e+00 -2.31021118e+00 1.31878763e-01 -3.35288137e-01
9.72138494e-02 5.61185479e-01 -7.52692521e-02 3.41461539e-01
2.43281081e-01 -2.75198013e-01 -2.45976806e-01 -2.82725006e-01
2.03101873e-01 -5.04118428e-02 -2.40369499e-01 5.62372029e-01
-2.65089571e-01 5.75249493e-01 -9.33681905e-01 -4.91943359e-01
2.33726785e-01 3.35885942e-01 -2.72364974e-01 8.13905746e-02
-1.32048488e-01 -1.25712335e-01 -4.53828365e-01 3.43920201e-01
4.84431565e-01 -2.31903180e-01 -6.60848692e-02 -4.47212160e-02
1.47969618e-01 -1.18572757e-01 -1.12859631e+00 1.76467597e+00
-3.27112973e-01 3.67282480e-01 -4.41255450e-01 -1.37407613e+00
1.21367025e+00 2.21327484e-01 2.61149138e-01 -5.59418499e-01
3.33653986e-01 1.84842423e-01 1.38170153e-01 -3.00759554e-01
5.11482358e-01 -3.45405310e-01 -1.73828781e-01 3.65006447e-01
2.23168522e-01 1.31244540e-01 2.57850587e-01 -1.27114981e-01
7.23343253e-01 -2.98477024e-01 5.11054873e-01 -4.62465852e-01
5.84711075e-01 -3.16308916e-01 6.04652286e-01 7.64401436e-01
-4.84748095e-01 6.23447478e-01 2.10208282e-01 -4.87697721e-01
-9.57946301e-01 -9.34518337e-01 -3.60233247e-01 1.68738449e+00
2.63922542e-01 -3.56826931e-01 -6.08839750e-01 -8.33338439e-01
7.71931633e-02 1.26371729e+00 -8.69840145e-01 -5.24502993e-01
-2.29721200e-02 -8.99683535e-01 2.20732674e-01 5.21236897e-01
4.50951874e-01 -7.48070955e-01 -5.76512277e-01 3.01393867e-01
4.51209769e-02 -8.64855230e-01 -6.64287388e-01 1.25340298e-01
-8.00189316e-01 -9.59199786e-01 -1.05032098e+00 -9.22770202e-01
3.09321254e-01 7.35111058e-01 7.38728821e-01 -2.45530292e-01
-1.79551348e-01 -6.12759078e-03 -7.28681982e-01 -4.36324269e-01
-8.18720609e-02 2.17664778e-01 1.69257075e-01 3.68630350e-01
7.95774698e-01 -4.53213543e-01 -3.70305598e-01 7.02521682e-01
-7.52294004e-01 -9.30973366e-02 7.25476071e-02 1.17063999e+00
5.30173063e-01 1.53651252e-01 1.02755415e+00 -1.04026973e+00
7.99470782e-01 -7.80543506e-01 -2.53328890e-01 4.86283183e-01
-9.48020399e-01 -6.99712932e-02 9.24177527e-01 -6.33375764e-01
-1.08744562e+00 -4.08691049e-01 3.75069410e-01 -5.04985929e-01
6.70878589e-02 4.73886013e-01 5.58370836e-02 3.35376889e-01
9.64506030e-01 2.82162040e-01 -1.70815885e-01 -5.23122132e-01
3.29888195e-01 1.14670038e+00 1.77099884e-01 -3.22487235e-01
4.37127262e-01 5.11485696e-01 -3.30833793e-01 -9.06039000e-01
-1.35866547e+00 -9.31795597e-01 -6.47980273e-01 3.20865177e-02
6.63074076e-01 -5.48120499e-01 -4.27220501e-02 5.45726657e-01
-8.07253420e-01 8.23921934e-02 -2.93521345e-01 4.93225038e-01
-4.79725331e-01 2.55180746e-01 -3.11316997e-01 -7.85309792e-01
-6.50049508e-01 -9.38560784e-01 6.97582185e-01 3.04837137e-01
-1.00916080e-01 -8.42522442e-01 1.02961324e-01 1.46789029e-01
4.55243230e-01 -9.18478444e-02 9.44971800e-01 -1.31398451e+00
4.21050012e-01 -4.62717831e-01 -1.76873147e-01 3.12827915e-01
2.37255111e-01 -2.52326578e-01 -9.44433928e-01 -4.23003525e-01
1.93041131e-01 -3.16227287e-01 1.07417715e+00 2.65642762e-01
1.18559074e+00 7.81543776e-02 -3.41463476e-01 4.63149250e-01
1.19036591e+00 4.65933591e-01 3.64240617e-01 4.51757222e-01
5.08536160e-01 5.83649457e-01 9.53321218e-01 7.42326081e-01
1.45433187e-01 7.66266167e-01 6.26560971e-02 5.79900682e-01
1.10403158e-01 -1.50374562e-01 -1.10419160e-02 9.00341570e-01
3.10494542e-01 -1.95620760e-01 -8.88835728e-01 4.63947892e-01
-2.24009228e+00 -1.17329967e+00 5.99804372e-02 2.25820923e+00
6.06180131e-01 4.72327948e-01 2.26209968e-01 3.34472120e-01
1.18910611e+00 4.38849449e-01 -7.32778430e-01 -2.48161942e-01
9.36740264e-02 8.98858253e-03 9.32460278e-03 2.82150637e-02
-1.29339767e+00 8.88367116e-01 5.40929127e+00 1.15570796e+00
-8.95458639e-01 4.57047045e-01 5.78908980e-01 -2.26644218e-01
1.72122326e-02 -3.72027010e-01 -1.01249158e+00 6.68715954e-01
8.83026898e-01 -8.15276027e-01 1.91919394e-02 1.23446715e+00
-1.87323347e-01 3.20423275e-01 -1.06162810e+00 1.09758675e+00
5.74624836e-01 -1.14949107e+00 6.16729669e-02 -3.01124930e-01
9.44055021e-01 -4.63074520e-02 1.45410001e-01 8.98295760e-01
-6.99020028e-02 -6.57893002e-01 6.13786697e-01 3.73273879e-01
5.82435191e-01 -1.04305756e+00 8.71154904e-01 5.89048684e-01
-1.14222074e+00 -4.18950200e-01 -8.53051960e-01 2.96688676e-02
7.51923188e-04 4.76851583e-01 -4.97372329e-01 4.28460807e-01
5.86240113e-01 8.21983933e-01 -5.14518976e-01 1.02852225e+00
2.07380071e-01 4.78634894e-01 5.06076552e-02 -4.86797184e-01
4.20600712e-01 -7.59143233e-02 4.38836336e-01 1.11019075e+00
3.38798463e-01 1.20908014e-01 3.84232968e-01 4.59663510e-01
-5.75300790e-02 5.41159272e-01 -3.61199647e-01 8.99548531e-02
4.64510530e-01 1.08435667e+00 -6.52167618e-01 -5.81730068e-01
-5.61597764e-01 8.13032985e-01 3.68538082e-01 2.89859399e-02
-8.77893150e-01 -1.01879656e+00 4.86423433e-01 -4.24814932e-02
5.30837595e-01 2.85582840e-01 -2.24146798e-01 -1.22036326e+00
1.79665890e-02 -6.06575489e-01 8.24053705e-01 -3.54097068e-01
-1.73451126e+00 6.87305033e-01 1.15395468e-02 -1.49837184e+00
-3.21633406e-02 -4.56461310e-01 -6.85741663e-01 5.04523098e-01
-1.27090430e+00 -8.96934152e-01 -4.27484840e-01 5.31715274e-01
1.17086184e+00 -6.31587148e-01 8.21382761e-01 2.14909151e-01
-5.11614978e-01 8.58629286e-01 7.36561000e-01 -5.46283163e-02
8.58109832e-01 -8.86879444e-01 2.73005038e-01 5.14104545e-01
8.24823752e-02 3.09391558e-01 6.50300562e-01 -4.71957713e-01
-1.08870411e+00 -1.09916008e+00 5.61668754e-01 -2.04606786e-01
8.16023171e-01 -2.69452691e-01 -1.26231992e+00 4.00694191e-01
-4.12046872e-02 3.25007617e-01 8.37311327e-01 2.95812160e-01
-6.47771120e-01 -2.95486391e-01 -1.12524390e+00 5.68267822e-01
9.91943240e-01 -2.68104166e-01 -9.53116059e-01 3.91612709e-01
7.73002923e-01 -4.31082398e-02 -6.09782577e-01 3.07705611e-01
6.61945820e-01 -9.92153168e-01 7.37807035e-01 -9.12678003e-01
3.51708263e-01 -1.01044834e-01 -5.22662044e-01 -1.50435030e+00
-6.46439552e-01 -2.73919970e-01 -2.89454311e-01 1.23880184e+00
2.93262362e-01 -6.24653459e-01 7.25655615e-01 4.01677459e-01
-2.27398887e-01 -8.15113783e-01 -1.14577389e+00 -1.19310582e+00
2.06391782e-01 -4.00785685e-01 4.07156318e-01 1.12740517e+00
4.18899953e-01 6.79732919e-01 -3.94555360e-01 -4.06325221e-01
7.78064311e-01 1.70749605e-01 4.58615094e-01 -1.56067622e+00
-9.60773304e-02 -6.59368694e-01 -6.60621524e-01 -5.83367527e-01
3.84431332e-01 -1.10647488e+00 7.81034529e-02 -1.38860524e+00
4.70188916e-01 -1.58588499e-01 -9.15610790e-01 2.07823411e-01
-4.35387731e-01 4.44057249e-02 2.07462147e-01 2.21203431e-01
-9.71193850e-01 9.22934592e-01 9.09646630e-01 -3.70491624e-01
-3.65724444e-01 1.09027199e-01 -6.66599154e-01 7.32612967e-01
8.98832977e-01 -5.99252939e-01 -7.76064157e-01 -1.21533155e-01
-3.03420395e-01 -2.33131066e-01 -1.92865551e-01 -1.19214559e+00
4.44707036e-01 -2.83453614e-01 2.70648420e-01 -5.86260796e-01
3.33899200e-01 -5.92226088e-01 -1.87515080e-01 4.31018710e-01
-7.61644185e-01 -2.88881481e-01 -1.73938632e-01 9.08222258e-01
-1.51360527e-01 -5.48040926e-01 1.26308870e+00 2.85692215e-02
-8.68584633e-01 3.34851265e-01 3.87868471e-02 6.25264347e-01
1.42586684e+00 -3.24806422e-01 -4.75113541e-01 -3.61967050e-02
-4.22727376e-01 1.81005120e-01 2.11429015e-01 7.90395379e-01
7.39780486e-01 -1.49961793e+00 -9.01789725e-01 1.07154898e-01
7.02204466e-01 -5.06304204e-01 5.33838570e-01 7.14793980e-01
7.92300105e-02 3.24732393e-01 -1.39387771e-01 -6.28204823e-01
-1.18259823e+00 8.94048274e-01 1.02807678e-01 -4.04351503e-02
-7.36778200e-01 7.56452680e-01 1.46385476e-01 -2.86125779e-01
3.23339432e-01 1.24920318e-02 -4.45057482e-01 4.13668901e-01
9.95897710e-01 6.87077165e-01 1.35027975e-01 -5.56025147e-01
-3.49660993e-01 5.20823419e-01 -6.61909461e-01 1.24027185e-01
1.43001628e+00 1.55841224e-02 4.29349899e-01 9.52744007e-01
1.40043116e+00 -6.26466513e-01 -1.10612977e+00 -6.65651977e-01
1.61558911e-01 -7.05228746e-01 1.12702638e-01 -4.49036688e-01
-7.81184912e-01 8.09697390e-01 6.47420526e-01 3.50507170e-01
6.95454478e-01 -1.79183900e-01 8.42830360e-01 4.13209528e-01
4.27554101e-01 -1.58262515e+00 2.72263646e-01 6.99438691e-01
5.79591811e-01 -1.38916564e+00 -8.16096552e-03 -1.03619143e-01
-8.40266109e-01 1.00216699e+00 7.05183685e-01 -1.73925161e-01
8.82661998e-01 -3.00750196e-01 -1.64335072e-01 7.22579733e-02
-8.93543839e-01 -4.07860540e-02 5.26132405e-01 5.33800662e-01
3.94220114e-01 4.69194539e-02 -5.51083803e-01 8.27695727e-01
1.68474868e-01 -2.42520943e-02 2.36063868e-01 9.47954834e-01
-9.47645783e-01 -7.49745786e-01 -7.82021806e-02 7.63123333e-01
-2.55882829e-01 -7.45641440e-02 -8.75369012e-02 6.94134116e-01
-2.32143521e-01 1.05798066e+00 -1.18663823e-02 -5.39177001e-01
5.91980755e-01 4.33271289e-01 1.81924254e-01 -8.76508474e-01
-3.30542505e-01 -1.68292925e-01 -1.99182212e-01 -6.25581071e-02
-1.35163352e-01 -4.81372744e-01 -1.10715985e+00 -3.90579291e-02
-4.94760811e-01 3.72811586e-01 3.46840441e-01 8.81595016e-01
3.34210634e-01 5.07905543e-01 8.73514116e-01 -6.50183916e-01
-1.24539900e+00 -1.11023593e+00 -8.49166989e-01 6.50263667e-01
-7.40757808e-02 -1.02869165e+00 -6.91486835e-01 -3.32477957e-01] | [10.193339347839355, 3.5264813899993896] |
78424d08-2cbf-4030-8935-2b4392b509a1 | convolutional-pose-machines | 1602.00134 | null | http://arxiv.org/abs/1602.00134v4 | http://arxiv.org/pdf/1602.00134v4.pdf | Convolutional Pose Machines | Pose Machines provide a sequential prediction framework for learning rich
implicit spatial models. In this work we show a systematic design for how
convolutional networks can be incorporated into the pose machine framework for
learning image features and image-dependent spatial models for the task of pose
estimation. The contribution of this paper is to implicitly model long-range
dependencies between variables in structured prediction tasks such as
articulated pose estimation. We achieve this by designing a sequential
architecture composed of convolutional networks that directly operate on belief
maps from previous stages, producing increasingly refined estimates for part
locations, without the need for explicit graphical model-style inference. Our
approach addresses the characteristic difficulty of vanishing gradients during
training by providing a natural learning objective function that enforces
intermediate supervision, thereby replenishing back-propagated gradients and
conditioning the learning procedure. We demonstrate state-of-the-art
performance and outperform competing methods on standard benchmarks including
the MPII, LSP, and FLIC datasets. | ['Shih-En Wei', 'Takeo Kanade', 'Varun Ramakrishna', 'Yaser Sheikh'] | 2016-01-30 | convolutional-pose-machines-1 | http://openaccess.thecvf.com/content_cvpr_2016/html/Wei_Convolutional_Pose_Machines_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Wei_Convolutional_Pose_Machines_CVPR_2016_paper.pdf | cvpr-2016-6 | ['car-pose-estimation'] | ['computer-vision'] | [ 1.78254366e-01 5.77249050e-01 -3.12000483e-01 -6.75948918e-01
-7.97327340e-01 -4.06995893e-01 6.05725288e-01 -1.17346719e-01
-4.22165751e-01 6.58959627e-01 2.52566665e-01 -7.78459609e-02
-6.69601411e-02 -5.07580340e-01 -1.33305800e+00 -4.48826283e-01
-3.13271821e-01 8.21165264e-01 2.39293396e-01 3.00494526e-02
1.86577663e-01 7.55107999e-01 -1.12210214e+00 3.10608417e-01
2.13244170e-01 1.00757468e+00 2.77848452e-01 8.16337109e-01
5.22699580e-02 1.05086303e+00 -1.49802864e-01 -4.47061986e-01
1.50316739e-02 -1.22700833e-01 -9.69880939e-01 2.19221562e-01
6.82233930e-01 -5.14383614e-01 -4.76261586e-01 4.32327181e-01
2.52534211e-01 -1.42773418e-02 9.18404818e-01 -8.90350103e-01
-4.42001253e-01 5.15676200e-01 -6.05198026e-01 -1.36508882e-01
2.86748894e-02 3.57879251e-02 1.06601226e+00 -1.10908067e+00
6.98123813e-01 1.29097641e+00 1.11938441e+00 4.97953385e-01
-1.44047153e+00 -2.04490602e-01 5.80956221e-01 3.64478193e-02
-1.24601197e+00 -3.91173035e-01 8.28738153e-01 -5.25576055e-01
9.14790988e-01 1.05879016e-01 6.19603753e-01 1.06063151e+00
1.20118447e-01 1.18672371e+00 6.38686538e-01 -3.89487058e-01
1.23980321e-01 -7.25220069e-02 5.81239872e-02 1.17454994e+00
-2.11325154e-01 4.15926948e-02 -8.68962109e-01 -4.63892817e-02
1.13686502e+00 -1.38309598e-01 -1.57138646e-01 -9.54124749e-01
-9.61132467e-01 7.63347805e-01 9.14036453e-01 -2.68559188e-01
-3.15838069e-01 8.46232176e-01 1.79192871e-01 -2.51956314e-01
6.19220972e-01 3.63844335e-01 -9.61472452e-01 4.15766649e-02
-1.02993011e+00 2.92344093e-01 7.28091180e-01 9.14905548e-01
8.19059610e-01 -2.67840415e-01 -2.72607535e-01 5.59898973e-01
6.99647069e-01 2.69524217e-01 1.60593111e-02 -1.20005167e+00
4.01052028e-01 1.93798617e-01 2.21938238e-01 -7.16723442e-01
-5.52986681e-01 -8.32400262e-01 -4.98409957e-01 2.48334602e-01
5.09674013e-01 -8.49945247e-02 -1.15591252e+00 1.73793423e+00
4.14454401e-01 1.47454932e-01 -4.50320929e-01 7.81063199e-01
4.65282142e-01 4.94761050e-01 1.50135130e-01 2.62684554e-01
8.06851745e-01 -1.40836740e+00 -2.33342141e-01 -4.02244031e-01
5.88589549e-01 -3.42917174e-01 8.51439297e-01 3.40083838e-01
-1.25249994e+00 -5.23136377e-01 -8.86249959e-01 -3.33005130e-01
-1.37738883e-01 5.33382416e-01 7.95435369e-01 1.80547893e-01
-1.23006511e+00 9.45967913e-01 -1.29154134e+00 1.16931602e-01
7.99272358e-01 5.69851875e-01 -3.95782799e-01 1.97241783e-01
-5.49696624e-01 1.01916242e+00 2.88436171e-02 5.20669699e-01
-1.17854285e+00 -8.17016661e-01 -1.16376233e+00 -3.35856564e-02
1.51859671e-01 -8.06609750e-01 1.52981985e+00 -9.10305440e-01
-1.54597545e+00 7.25217938e-01 -2.52788097e-01 -5.52565873e-01
7.52194226e-01 -7.68993497e-01 5.19066572e-01 1.63862668e-02
2.76754219e-02 1.25772178e+00 9.78328586e-01 -1.37251639e+00
-1.79484695e-01 -4.86453503e-01 1.23342752e-01 3.16358596e-01
-1.97668727e-02 -3.38416070e-01 -7.20591128e-01 -4.41310525e-01
2.27147073e-01 -1.00226045e+00 -5.13338149e-01 5.82878292e-01
-5.04949152e-01 -8.85014534e-02 7.15676129e-01 -8.11187088e-01
6.87427819e-01 -1.78406191e+00 7.17194259e-01 1.94052041e-01
1.42697364e-01 -2.00848877e-01 -1.44901171e-01 1.01291247e-01
1.66082025e-01 -1.21538609e-01 -2.33028322e-01 -9.03121889e-01
1.38794258e-01 3.84820461e-01 -3.42056066e-01 6.73774242e-01
5.02247095e-01 1.27480817e+00 -6.93737447e-01 -4.54967111e-01
4.17223841e-01 6.44195080e-01 -8.56739163e-01 2.17285633e-01
-6.50890946e-01 7.33779073e-01 -2.59670287e-01 3.31625730e-01
4.16638941e-01 -5.29420853e-01 2.29458541e-01 -3.46473157e-01
-5.99065721e-02 4.55563277e-01 -8.79899800e-01 2.22207522e+00
-6.02809012e-01 5.51632702e-01 3.55415523e-01 -9.38757241e-01
4.19828176e-01 -1.47521719e-02 3.49563688e-01 -2.06618622e-01
-4.19954509e-02 -1.85201079e-01 -4.67918843e-01 -1.68845579e-01
1.21942259e-01 -2.78235488e-02 -2.61046588e-02 -7.47572333e-02
4.68049139e-01 -1.86091155e-01 -2.90846944e-01 3.16626169e-02
8.69756937e-01 9.14199889e-01 -1.31970853e-01 -2.09528863e-01
3.12382579e-01 -9.63418707e-02 1.75907940e-01 6.44478202e-01
1.53944060e-01 7.24999368e-01 4.75678056e-01 -6.54315174e-01
-9.82677698e-01 -1.24172068e+00 -1.12315618e-01 1.44181859e+00
-7.39523163e-03 -3.91142905e-01 -4.37734604e-01 -9.01873887e-01
1.41781434e-01 3.56598556e-01 -8.60589087e-01 5.38814552e-02
-7.42382824e-01 -3.88458759e-01 3.92075241e-01 1.16483176e+00
2.49510482e-01 -9.83309627e-01 -5.21990120e-01 2.12156758e-01
4.45749164e-02 -7.85547197e-01 -2.75956839e-01 6.21438503e-01
-9.58702803e-01 -8.70757401e-01 -8.05768967e-01 -8.56825531e-01
7.64009714e-01 -3.66335154e-01 1.34533751e+00 -4.13740762e-02
-3.58468622e-01 3.25488985e-01 1.48921221e-01 -2.69588977e-01
1.10420369e-01 4.14556831e-01 -3.99705797e-01 -3.74747127e-01
-2.28117645e-01 -6.48435950e-01 -6.62171543e-01 1.18423976e-01
-4.26898181e-01 3.97305042e-01 8.60733867e-01 1.04106641e+00
8.23696256e-01 -6.50053680e-01 1.37610734e-01 -8.80021393e-01
-5.56418449e-02 -2.20744103e-01 -6.45993531e-01 2.03695476e-01
-1.31826907e-01 6.91576123e-01 3.03736329e-01 -2.65370041e-01
-1.14007771e+00 6.83520496e-01 -3.50349218e-01 -4.81435746e-01
-9.68977436e-02 5.04399717e-01 -1.56605728e-02 -3.25752825e-01
4.33508456e-01 1.68843552e-01 -4.35012914e-02 -7.06011117e-01
7.09589422e-01 3.37132104e-02 6.92889571e-01 -7.28768826e-01
6.97512567e-01 5.00329077e-01 2.10511580e-01 -6.41234338e-01
-1.28411269e+00 -1.99347168e-01 -1.10459530e+00 -1.54868692e-01
9.56801295e-01 -1.09076583e+00 -6.26009464e-01 4.48358655e-01
-1.28195012e+00 -8.34574282e-01 -2.01708525e-01 3.21908563e-01
-7.78622210e-01 6.04795143e-02 -1.02885425e+00 -5.19444883e-01
-5.71056977e-02 -9.49750304e-01 1.61069286e+00 -8.36858526e-02
-3.08797389e-01 -1.06763744e+00 2.83022404e-01 3.08331192e-01
2.61963516e-01 1.69952571e-01 6.91359401e-01 -1.54528499e-01
-8.77770722e-01 6.14460232e-03 -1.36169419e-01 6.68488517e-02
-3.22199672e-01 -1.30415916e-01 -1.05681658e+00 -1.82371959e-01
-3.29833776e-01 -6.92101955e-01 1.03833878e+00 6.07249677e-01
1.63446772e+00 -3.95248741e-01 -4.93697703e-01 9.49333906e-01
1.17292130e+00 -5.44164896e-01 3.86609554e-01 1.41681015e-01
1.08901894e+00 5.01840532e-01 2.92390555e-01 3.16379458e-01
4.92192000e-01 7.62539685e-01 6.98221624e-01 -2.61599988e-01
-1.39092684e-01 -5.22275269e-01 4.05454934e-02 2.22621560e-01
-2.50811223e-02 1.08342379e-01 -8.41094315e-01 4.67232287e-01
-1.98240542e+00 -5.94626844e-01 -1.95394401e-02 1.89672089e+00
9.33582962e-01 1.40304193e-01 -9.43200216e-02 -3.43096226e-01
2.12994173e-01 3.15082908e-01 -5.50400853e-01 -3.42556834e-02
4.12619591e-01 3.65395963e-01 7.09064960e-01 8.87041092e-01
-1.51834750e+00 9.77265477e-01 6.97841644e+00 4.59552318e-01
-1.05358493e+00 7.28500709e-02 8.73188674e-01 -2.89624035e-01
-2.48760328e-01 -1.79681480e-01 -7.70117342e-01 1.97052862e-02
6.83797717e-01 7.28341937e-01 2.29500458e-01 1.05104458e+00
-1.87407695e-02 6.06113523e-02 -1.42953694e+00 7.35929310e-01
-1.15249500e-01 -1.49176633e+00 -1.99743897e-01 4.23625037e-02
6.85555518e-01 3.45050722e-01 1.63567379e-01 2.35974714e-01
4.44545448e-01 -1.31237531e+00 1.13296199e+00 5.33641994e-01
6.36222243e-01 -5.86780012e-01 4.53485221e-01 3.24881762e-01
-1.05083895e+00 7.79831484e-02 -1.89418167e-01 -2.91670114e-01
2.19656616e-01 3.49479645e-01 -7.06334710e-01 2.50351191e-01
5.59907138e-01 7.97386527e-01 -4.90751117e-01 8.51476550e-01
-4.97560948e-01 5.02550542e-01 -4.83452201e-01 1.09065533e-01
3.83124441e-01 1.97700053e-01 2.05299854e-01 1.20453298e+00
-1.15083165e-01 -3.40674251e-01 1.84470043e-01 9.44455206e-01
-1.19063333e-01 -3.24297428e-01 -4.60128576e-01 2.76296407e-01
6.93786442e-02 1.15449595e+00 -6.48937464e-01 -1.39050215e-01
-2.38629103e-01 1.29535782e+00 1.00012505e+00 2.62000352e-01
-1.06203532e+00 1.58019274e-01 5.29339969e-01 2.93306798e-01
7.88083315e-01 -4.93731856e-01 -5.51059127e-01 -9.55333889e-01
1.01216577e-01 -2.67821729e-01 4.24309671e-02 -8.65682423e-01
-9.79985833e-01 2.50274032e-01 -2.82351412e-02 -5.99871874e-01
-4.32198673e-01 -1.02317989e+00 -4.83657002e-01 8.53137136e-01
-1.41890764e+00 -1.53437626e+00 -1.55392125e-01 3.31657588e-01
4.75848436e-01 4.88654464e-01 7.83066452e-01 1.66234281e-02
-3.30823094e-01 4.06152427e-01 -5.46055660e-02 2.38873214e-01
3.49016666e-01 -1.43574810e+00 4.35010910e-01 5.18227398e-01
4.92065638e-01 4.59205717e-01 6.65906250e-01 -4.92850393e-01
-1.31833649e+00 -1.09566844e+00 6.96065307e-01 -8.33490312e-01
5.51817834e-01 -6.14863157e-01 -4.69229788e-01 1.12066770e+00
-7.07219690e-02 2.86589563e-01 2.78964400e-01 4.07514960e-01
-4.00045037e-01 8.71971324e-02 -6.96714997e-01 3.82849932e-01
1.23498631e+00 -6.34558022e-01 -4.08795416e-01 3.93300325e-01
5.65491378e-01 -8.35797131e-01 -6.10358477e-01 3.37814242e-01
7.03675210e-01 -6.84251368e-01 1.29065251e+00 -7.10572600e-01
7.16654658e-01 -1.10758372e-01 9.91372615e-02 -9.66706455e-01
-4.43559587e-01 -3.25242192e-01 -5.55351794e-01 8.00797224e-01
6.82222247e-01 4.65119034e-02 1.26072931e+00 8.01846862e-01
-2.43201390e-01 -1.22836733e+00 -6.85706556e-01 -2.69307762e-01
1.03705063e-01 -3.30834061e-01 3.52293886e-02 3.81239235e-01
-3.42276156e-01 4.58630204e-01 -5.28037965e-01 2.90050149e-01
6.90883875e-01 -2.83289850e-02 7.65834510e-01 -1.02097046e+00
-7.26425886e-01 -1.83834389e-01 -3.72059017e-01 -1.73922157e+00
3.29827130e-01 -6.31877780e-01 4.74118710e-01 -1.52174270e+00
2.46322691e-01 -2.98255384e-01 -1.21222742e-01 7.53134012e-01
8.47395509e-03 4.09216553e-01 5.53831570e-02 6.17345646e-02
-8.84536326e-01 7.69821227e-01 1.24283814e+00 -1.35308176e-01
-1.75845120e-02 1.30990580e-01 -4.44816619e-01 9.26051795e-01
4.96778250e-01 -2.60538071e-01 -5.56136012e-01 -8.84062290e-01
2.51259804e-01 -4.76136170e-02 8.77409637e-01 -9.98288035e-01
2.59391278e-01 1.10459387e-01 9.80055034e-01 -8.49342406e-01
7.09252298e-01 -8.45558643e-01 -9.38974880e-03 4.18696135e-01
-6.35978878e-01 -9.21745822e-02 2.04474866e-01 7.24698961e-01
-1.86085384e-02 1.10233715e-02 5.19972563e-01 -2.51263231e-01
-7.71133423e-01 4.89468127e-01 5.47499470e-02 1.43009750e-02
6.46865666e-01 6.46492541e-02 1.77899629e-01 -3.82598549e-01
-1.13985145e+00 3.20730478e-01 4.88241732e-01 1.28757134e-01
5.05619943e-01 -1.17559505e+00 -4.43954557e-01 1.11826427e-01
-1.60635293e-01 5.11894405e-01 -3.24700177e-02 7.95013249e-01
-8.08582783e-01 3.84089857e-01 1.24073932e-02 -8.39013338e-01
-9.92874861e-01 2.43690446e-01 5.77299714e-01 -5.24460614e-01
-6.51588023e-01 1.48654985e+00 5.02089977e-01 -7.34943926e-01
6.15613639e-01 -4.81541783e-01 2.02854007e-01 -3.32025260e-01
2.67269939e-01 -1.07281096e-02 -3.27367522e-02 -5.25862753e-01
-4.79361743e-01 4.78397965e-01 -1.10061973e-01 -3.97193611e-01
1.58725536e+00 -5.56071103e-02 3.32179554e-02 3.93918067e-01
1.33576560e+00 -3.29807103e-01 -2.07190895e+00 -9.08599868e-02
-1.61462631e-02 -1.24205649e-01 1.32815525e-01 -9.72881675e-01
-8.94610584e-01 9.95657325e-01 3.52743626e-01 -4.79469419e-01
6.15829706e-01 4.44772542e-01 3.49283904e-01 4.52948093e-01
4.00341868e-01 -8.53705883e-01 3.37047756e-01 7.87124693e-01
9.38272655e-01 -1.20627809e+00 -4.13537920e-02 -3.98329198e-01
-3.52334440e-01 1.00133610e+00 7.32360840e-01 -3.42214346e-01
7.27901816e-01 4.71273184e-01 -2.45584562e-01 -3.25449169e-01
-6.02490604e-01 5.16200699e-02 6.11823559e-01 5.07165551e-01
5.93394935e-01 -5.11931144e-02 4.24648495e-03 4.19731647e-01
-1.03249811e-01 -4.35392447e-02 -2.90301651e-01 1.04257250e+00
-4.32472527e-01 -1.04766881e+00 -1.33217484e-01 2.75556505e-01
-3.17122877e-01 -1.57838061e-01 -5.33304870e-01 6.92156434e-01
6.55090436e-02 1.42242700e-01 1.62516624e-01 -1.32571146e-01
-1.44112051e-01 1.95477128e-01 9.18786287e-01 -5.93428969e-01
-2.22435430e-01 -3.36152837e-02 1.25264838e-01 -9.12450671e-01
-2.93408841e-01 -5.94002366e-01 -1.08854365e+00 1.11480318e-01
-2.66665816e-01 -7.08990693e-02 5.85330009e-01 1.13729334e+00
2.86358386e-01 6.20386064e-01 1.29116610e-01 -1.71610773e+00
-6.17810309e-01 -9.47327793e-01 -2.16511294e-01 2.86120474e-01
5.15772820e-01 -7.57228732e-01 -1.62711833e-02 1.06831349e-01] | [7.144000053405762, -1.4093047380447388] |
eb9cb842-0854-4589-a751-cb5442abe2ce | natural-language-sentence-generation-from-api | 2206.06868 | null | https://arxiv.org/abs/2206.06868v1 | https://arxiv.org/pdf/2206.06868v1.pdf | Natural Language Sentence Generation from API Specifications | APIs are everywhere; they provide access to automation solutions that could help businesses automate some of their tasks. Unfortunately, they may not be accessible to the business users who need them but are not equipped with the necessary technical skills to leverage them. Wrapping these APIs with chatbot capabilities is one solution to make these automation solutions interactive. In this work, we propose a system to generate sentences to train intent recognition models, a crucial component within chatbots to understand natural language utterances from users. Evaluation of our approach based on deep learning models showed promising and inspiring results, and the human-in-the-loop interaction will provide further improvement on the system. | ['Yara Rizk', 'Vinod Muthusamy', 'Vatche Isahagian', 'Jayachandu Bandlamudi', 'Kushal Mukherjee', 'Siyu Huo'] | 2022-06-01 | null | null | null | null | ['intent-recognition'] | ['natural-language-processing'] | [-3.40401620e-01 5.20960450e-01 1.46423891e-01 -7.62675107e-01
-2.75333852e-01 -7.39016831e-01 5.86480200e-01 -3.98112118e-01
8.10902193e-03 4.06294018e-01 3.13936472e-01 -8.60293984e-01
2.73952037e-01 -8.80767584e-01 -6.83553591e-02 -1.85137257e-01
4.73329574e-01 4.89779800e-01 1.38781145e-01 -7.45251060e-01
1.79970399e-01 3.56543571e-01 -1.24106467e+00 8.68337154e-01
7.66112149e-01 8.02422702e-01 5.11317015e-01 8.63367260e-01
-8.78112972e-01 1.42341173e+00 -6.31835699e-01 -4.84479547e-01
1.32026300e-01 -3.40062767e-01 -1.16788030e+00 -2.69252330e-01
-2.00879574e-01 -5.07597685e-01 1.37821004e-01 5.47134638e-01
-5.95969148e-03 -7.13244155e-02 1.93459585e-01 -1.26570749e+00
-5.18856466e-01 7.33352721e-01 2.43843302e-01 -3.42870593e-01
7.80808032e-01 6.08509898e-01 9.15408254e-01 -4.18539077e-01
4.61267531e-01 9.07801270e-01 5.07124245e-01 8.04911256e-01
-7.64532149e-01 -3.29394132e-01 -4.23580557e-02 1.31867245e-01
-7.06038952e-01 -5.47252595e-01 7.48874307e-01 -4.63959485e-01
1.31709468e+00 6.37745380e-01 5.39690554e-01 1.30264246e+00
1.09620333e-01 7.92092025e-01 7.60689259e-01 -4.39800084e-01
-4.28585224e-02 7.63395607e-01 5.67857087e-01 7.05591202e-01
5.08884899e-03 -4.68792915e-01 -2.92457402e-01 -1.59055647e-02
6.25934660e-01 3.64567131e-01 5.78601658e-03 8.62576962e-02
-7.47844219e-01 7.85321772e-01 4.49010998e-01 6.52151287e-01
-4.50581551e-01 -1.07311487e-01 3.14636230e-01 4.85000938e-01
1.05672330e-02 9.12780464e-01 -4.15853798e-01 -9.35614765e-01
-1.63296714e-01 -1.91078391e-02 1.58286250e+00 1.33924747e+00
5.98951995e-01 -2.21506059e-01 -1.59293249e-01 4.05681998e-01
4.15678799e-01 9.97994393e-02 3.11101079e-01 -1.06352532e+00
5.22045434e-01 1.10687888e+00 6.43071651e-01 -7.15750515e-01
-3.18139344e-01 2.05989584e-01 -5.29360652e-01 2.14620963e-01
5.30836046e-01 -3.92091811e-01 -3.60150665e-01 8.78521204e-01
7.15945363e-02 -4.45265979e-01 5.81011139e-02 7.50591576e-01
9.49983478e-01 8.61300766e-01 5.59416823e-02 1.61705822e-01
1.53926110e+00 -1.46939182e+00 -9.96668577e-01 -2.77075440e-01
7.95791268e-01 -8.74454200e-01 1.79735053e+00 4.70203608e-01
-9.32676077e-01 -6.07682288e-01 -7.04361200e-01 -1.18583210e-01
-4.67982322e-01 7.28145475e-03 1.11881530e+00 9.09326077e-01
-8.23955178e-01 4.91041780e-01 -8.09981883e-01 -3.83948714e-01
1.24743335e-01 2.22965613e-01 -1.64131701e-01 2.29568884e-01
-1.04426694e+00 1.10803044e+00 9.55935754e-03 2.07389891e-01
-3.70054036e-01 -2.24370003e-01 -8.33156526e-01 4.16461825e-01
2.03907594e-01 -5.72936773e-01 2.09127140e+00 -7.26862669e-01
-2.04850006e+00 3.56580049e-01 -1.22021563e-01 -3.82105768e-01
5.97258151e-01 -5.00329971e-01 -7.15137692e-03 -2.08841711e-01
4.88800369e-02 4.53081876e-01 4.26743388e-01 -8.55283737e-01
-6.91307306e-01 -8.28377753e-02 6.38625503e-01 6.36724532e-02
-5.09921134e-01 2.07516432e-01 -3.19540679e-01 2.80015111e-01
-4.60925996e-01 -7.60019839e-01 -3.91501456e-01 -4.40248907e-01
-4.53995377e-01 -3.56347382e-01 1.14703166e+00 -7.64173567e-01
9.70837951e-01 -1.95867920e+00 -4.93557155e-01 -1.00987151e-01
1.63606566e-03 6.37814939e-01 -3.53081338e-02 9.61179137e-01
5.72175205e-01 3.08936864e-01 5.08004248e-01 -3.45612675e-01
2.98181593e-01 1.67384952e-01 -4.47262138e-01 -3.65056127e-01
1.86105996e-01 9.04330611e-01 -8.16645026e-01 -1.50960803e-01
4.61765200e-01 3.58314961e-01 -4.24903542e-01 9.97548342e-01
-5.16612470e-01 6.61656618e-01 -8.60702515e-01 4.08948749e-01
2.72833407e-01 -4.08489257e-01 2.22671673e-01 2.81829864e-01
-3.08516771e-01 9.12190020e-01 -6.87664330e-01 1.35767055e+00
-1.23097551e+00 5.43814480e-01 3.03724617e-01 -6.63227856e-01
9.44141507e-01 6.48275435e-01 6.30234331e-02 -3.61001194e-01
1.21375494e-01 -5.66662662e-02 6.52890429e-02 -1.14606762e+00
3.88737738e-01 1.88653305e-01 -1.99028984e-01 7.05563664e-01
-1.62533626e-01 -4.21549648e-01 1.06580794e-01 -7.61728873e-03
1.31575942e+00 1.80999950e-01 2.62778550e-01 5.80719531e-01
4.30482715e-01 9.81664956e-02 3.04297619e-02 9.95527565e-01
-1.62961498e-01 2.87290603e-01 5.40512085e-01 -9.45751488e-01
-8.10465932e-01 -5.58707178e-01 3.41284156e-01 1.22579253e+00
-2.87952721e-01 -6.26527607e-01 -8.12122345e-01 -9.98484015e-01
-6.13543391e-01 1.07533395e+00 -2.72169597e-02 1.83857128e-01
-3.22471499e-01 2.10250229e-01 3.30826730e-01 4.39322501e-01
8.27504039e-01 -1.67716730e+00 -9.04411077e-01 4.61077064e-01
-4.77451116e-01 -1.38548875e+00 -2.00914904e-01 1.27517894e-01
-5.32488465e-01 -8.86044323e-01 -1.27461806e-01 -7.04597592e-01
4.07957792e-01 4.37570423e-01 8.31629038e-01 3.37338060e-01
-5.93800992e-02 1.88808829e-01 -7.84476459e-01 -6.44968867e-01
-8.25379789e-01 5.29950917e-01 -4.50496346e-01 -2.92880565e-01
6.27171397e-01 -6.23246729e-01 -3.68678987e-01 3.77076358e-01
-4.79588628e-01 5.04167497e-01 5.29916465e-01 5.71140349e-01
-5.79632998e-01 -2.07775503e-01 3.81465554e-01 -1.14659274e+00
9.29505587e-01 -1.86882004e-01 -4.57184643e-01 1.26963794e-01
-2.87034750e-01 1.26946401e-02 1.02218533e+00 -2.08678797e-01
-1.33816171e+00 2.93224663e-01 -6.95170641e-01 1.85755432e-01
-6.69985056e-01 3.45348477e-01 -3.49406570e-01 1.05863616e-01
4.36007708e-01 -1.37572125e-01 -4.16298844e-02 -7.11511314e-01
5.01770556e-01 1.47784925e+00 1.87084675e-02 -2.72741526e-01
4.85100031e-01 7.88231716e-02 -6.75031960e-01 -6.27778649e-01
-8.94427299e-01 -6.34732842e-01 -3.83657873e-01 -5.67054264e-02
9.50648546e-01 -2.91772693e-01 -1.28761876e+00 1.73184246e-01
-1.75387871e+00 -5.97551525e-01 -8.79787728e-02 1.09403439e-01
-5.30221999e-01 -1.61814746e-02 -6.91778243e-01 -1.06671429e+00
-6.00172102e-01 -1.21896088e+00 7.68895566e-01 5.82342625e-01
-7.76768565e-01 -6.30894780e-01 -1.27314270e-01 1.06706107e+00
9.80539560e-01 -4.27200854e-01 6.94113016e-01 -1.26320970e+00
-7.30040789e-01 -6.94496572e-01 -9.20650363e-02 4.72433597e-01
5.13892889e-01 5.70421852e-02 -1.13818932e+00 5.27440667e-01
2.56774336e-01 -2.45316923e-01 -6.95597939e-03 -1.13317251e-01
1.16561949e+00 -9.24096286e-01 -3.63823801e-01 1.71450973e-01
6.53122067e-01 5.79298139e-01 7.77692080e-01 2.84225225e-01
6.13111377e-01 7.73564219e-01 4.42452073e-01 3.62726301e-01
4.23857033e-01 8.66328359e-01 3.72231334e-01 -1.20496154e-01
2.24026248e-01 -3.65327686e-01 3.94536167e-01 5.97068787e-01
-2.66664773e-01 1.19916135e-02 -9.96641934e-01 2.72186905e-01
-1.98627031e+00 -1.11579645e+00 -5.14079094e-01 1.66526532e+00
7.59517729e-01 2.19822228e-01 1.59660801e-01 -2.41331443e-01
2.86057770e-01 -9.18513760e-02 9.19295177e-02 -1.03415167e+00
9.05549705e-01 2.48840138e-01 -1.78334966e-01 6.10602975e-01
-9.53165770e-01 1.10303581e+00 5.98786688e+00 8.34607929e-02
-1.13205254e+00 2.27868125e-01 5.27320981e-01 4.05190229e-01
-7.56210461e-02 2.04286918e-01 -8.06765854e-01 3.99336547e-01
1.25865817e+00 1.42890424e-01 7.66547084e-01 1.48116159e+00
6.73392236e-01 1.46245793e-01 -1.22657228e+00 5.71453929e-01
-3.55809271e-01 -1.49846160e+00 -4.25478607e-01 -1.23581529e-01
2.02910468e-01 -2.15761065e-01 -2.09914684e-01 6.87492549e-01
4.96766508e-01 -1.06113958e+00 2.20079511e-01 5.29329777e-01
1.21829815e-01 -4.17229444e-01 1.19895411e+00 8.69873106e-01
-7.10506022e-01 -2.20115498e-01 -1.42244533e-01 -8.37025583e-01
3.69861782e-01 1.84276909e-01 -1.63916337e+00 7.63171762e-02
7.01915145e-01 4.42804471e-02 -4.46290910e-01 6.85972452e-01
-6.23806059e-01 6.49576008e-01 -1.88480869e-01 -8.98523271e-01
6.84898719e-02 -2.97518373e-01 1.08701266e-01 1.21315241e+00
2.72095352e-01 1.72367319e-02 9.42834020e-02 9.76163924e-01
5.84837794e-02 6.61327317e-02 -1.00583637e+00 -5.02361834e-01
4.25208211e-01 1.71012938e+00 -5.63393772e-01 -1.70250997e-01
-7.22715795e-01 1.03324866e+00 2.13143900e-01 1.75916418e-01
-6.53956056e-01 -8.87428343e-01 6.70447290e-01 3.00241232e-01
3.69719081e-02 -2.69603431e-01 -5.10631204e-01 -9.09210324e-01
8.83077309e-02 -1.13451195e+00 -1.77541777e-01 -1.00590503e+00
-1.24164569e+00 6.31452084e-01 -3.32743466e-01 -8.34852099e-01
-6.70980573e-01 -7.54195392e-01 -1.22639799e+00 8.01150382e-01
-8.59363616e-01 -1.54325962e+00 -5.65663099e-01 1.51264355e-01
6.82889938e-01 -1.40373379e-01 1.25441897e+00 -1.17025293e-01
-1.08437367e-01 2.04871148e-01 -5.80862403e-01 3.44023079e-01
3.99971575e-01 -1.08632898e+00 5.61936796e-01 4.70944196e-01
4.29278404e-01 1.02869296e+00 7.97220767e-01 -1.82286724e-01
-1.62124431e+00 -8.32978070e-01 1.26442730e+00 -8.39654863e-01
5.69441557e-01 -5.64742625e-01 -6.76838458e-01 1.07514393e+00
6.80653572e-01 -5.59420347e-01 8.03783119e-01 5.33220291e-01
1.77953094e-02 -1.26965687e-01 -1.05934513e+00 7.65128613e-01
6.57606125e-01 -5.04610002e-01 -8.93620133e-01 4.57570851e-01
9.80721116e-01 -4.17784959e-01 -4.48096186e-01 -7.41703287e-02
5.31284094e-01 -1.15820110e+00 5.59860587e-01 -8.11949968e-01
4.71363366e-01 -2.50543002e-02 2.69991279e-01 -1.08587492e+00
7.90152624e-02 -1.10053313e+00 -1.31769609e-02 1.51760888e+00
5.99467933e-01 -7.27055967e-01 9.34734702e-01 1.24600756e+00
-3.47102165e-01 -5.05662322e-01 -3.86328995e-01 -3.76084536e-01
-4.47009891e-01 -6.33845985e-01 1.00673091e+00 6.10573888e-01
7.86832869e-01 7.43596315e-01 -2.05850422e-01 -2.58734152e-02
-3.61806691e-01 1.12284318e-01 1.39865923e+00 -9.15631235e-01
-4.88475740e-01 -2.72304952e-01 -1.11535572e-01 -1.23347676e+00
-2.34492887e-02 -4.37233835e-01 3.06769937e-01 -2.04209566e+00
-9.34765637e-02 -3.60394150e-01 3.69298905e-01 9.11640644e-01
1.45806670e-01 -3.02816540e-01 3.90161872e-01 -9.36519131e-02
-5.54375648e-01 2.45320186e-01 1.03532791e+00 -1.84430137e-01
-6.90935969e-01 7.70558834e-01 -1.11345220e+00 8.63794327e-01
1.04875684e+00 -1.61855802e-01 -5.23162901e-01 -2.83070117e-01
1.59423783e-01 1.18809812e-01 1.72723517e-01 -1.00479829e+00
2.50090599e-01 -4.79802489e-01 -5.51152304e-02 -1.40730530e-01
3.34733814e-01 -1.07982969e+00 3.71130905e-03 3.64668071e-01
-4.31461722e-01 -9.58350524e-02 -1.27278447e-01 -1.46467254e-01
-1.24393009e-01 -7.34380603e-01 1.02893718e-01 -4.09551501e-01
-3.38623255e-01 -2.54920095e-01 -9.26315665e-01 -4.24874991e-01
1.06491721e+00 1.06230296e-01 -4.31442082e-01 -1.10801888e+00
-6.12027347e-01 9.55795869e-02 2.92478830e-01 4.87621486e-01
3.27184469e-01 -8.40973258e-01 -1.21691115e-02 2.94088364e-01
-3.76784615e-02 -9.03532803e-02 -1.98879689e-01 5.24614453e-01
-7.76969671e-01 7.76896238e-01 -2.18832791e-01 -2.32819706e-01
-1.21700776e+00 5.10156572e-01 1.26020893e-01 -6.37206316e-01
-5.31879365e-01 7.45449245e-01 -2.70050436e-01 -7.27921903e-01
5.11593163e-01 -4.39723730e-01 -3.51350516e-01 -2.33636037e-01
8.09527338e-01 -4.81800102e-02 2.43357167e-01 5.80364056e-02
-1.97463542e-01 -2.47690678e-01 -1.23030081e-01 -3.19602966e-01
1.53214192e+00 8.14982578e-02 3.34816836e-02 4.59042639e-01
6.64261639e-01 -1.80622898e-02 -9.66560006e-01 3.45774293e-01
1.54059693e-01 -5.04167914e-01 -5.16313493e-01 -8.98466527e-01
-4.17325705e-01 1.38256133e+00 2.22412884e-01 1.06594610e+00
6.11613095e-01 -4.80393507e-02 1.01081634e+00 1.11078656e+00
6.51353538e-01 -9.66026664e-01 2.28324607e-01 7.32567668e-01
9.34478879e-01 -1.34941125e+00 -6.24917984e-01 -4.01388049e-01
-1.05301774e+00 1.37832415e+00 8.53406429e-01 2.74146169e-01
2.98124641e-01 4.57029581e-01 5.44856191e-01 -2.08775073e-01
-9.52355981e-01 -1.20533533e-01 -1.47649407e-01 8.73013973e-01
1.02740729e+00 -4.24381830e-02 -2.07710221e-01 9.79559839e-01
-3.59913051e-01 2.34670505e-01 5.74668229e-01 1.03251779e+00
-5.19442677e-01 -1.47462356e+00 -1.81090444e-01 3.26205969e-01
-3.07183474e-01 1.06658258e-01 -6.98463678e-01 6.38450980e-01
-2.25280136e-01 1.50550866e+00 -1.37712434e-01 -5.92210352e-01
4.62029040e-01 5.35383284e-01 3.28575820e-02 -1.14885330e+00
-1.02449381e+00 -4.39210057e-01 7.34098315e-01 -5.96150935e-01
1.83003008e-01 -1.58128977e-01 -1.31515431e+00 -3.66943598e-01
-8.95753875e-02 3.98298621e-01 7.52435565e-01 1.05184889e+00
4.40451264e-01 1.72841340e-01 8.47217619e-01 -9.34669077e-01
-5.38205326e-01 -1.31810522e+00 9.74589363e-02 4.50173080e-01
6.60135671e-02 1.05803564e-01 -2.13520415e-02 1.19308785e-01] | [12.73116397857666, 7.8353166580200195] |
62bb2f14-c505-40e3-a373-45ca8a42bb40 | disentangled-ontology-embedding-for-zero-shot | 2206.03739 | null | https://arxiv.org/abs/2206.03739v1 | https://arxiv.org/pdf/2206.03739v1.pdf | Disentangled Ontology Embedding for Zero-shot Learning | Knowledge Graph (KG) and its variant of ontology have been widely used for knowledge representation, and have shown to be quite effective in augmenting Zero-shot Learning (ZSL). However, existing ZSL methods that utilize KGs all neglect the intrinsic complexity of inter-class relationships represented in KGs. One typical feature is that a class is often related to other classes in different semantic aspects. In this paper, we focus on ontologies for augmenting ZSL, and propose to learn disentangled ontology embeddings guided by ontology properties to capture and utilize more fine-grained class relationships in different aspects. We also contribute a new ZSL framework named DOZSL, which contains two new ZSL solutions based on generative models and graph propagation models, respectively, for effectively utilizing the disentangled ontology embeddings. Extensive evaluations have been conducted on five benchmarks across zero-shot image classification (ZS-IMGC) and zero-shot KG completion (ZS-KGC). DOZSL often achieves better performance than the state-of-the-art, and its components have been verified by ablation studies and case studies. Our codes and datasets are available at https://github.com/zjukg/DOZSL. | ['Huajun Chen', 'Feiyu Xiong', 'Yufeng Huang', 'Jeff Z. Pan', 'Zhuo Chen', 'Yajing Xu', 'Wen Zhang', 'Jiaoyan Chen', 'Yuxia Geng'] | 2022-06-08 | null | null | null | null | ['ontology-embedding'] | ['knowledge-base'] | [-1.09383591e-01 3.81391019e-01 -4.57510889e-01 -1.76034853e-01
-3.66519362e-01 -1.94560483e-01 5.99069297e-01 2.95398593e-01
-2.57911510e-03 5.78464448e-01 4.45794076e-01 1.26751721e-01
-6.09064519e-01 -1.20716429e+00 -5.50717711e-01 -6.39507055e-01
-9.58095863e-02 4.29877877e-01 1.38076022e-01 -3.25639874e-01
-4.59615588e-02 -6.75040707e-02 -1.69010425e+00 1.13648988e-01
8.28249454e-01 7.21908450e-01 -1.52506530e-01 1.73133343e-01
-3.53297532e-01 8.56159329e-01 -2.21766323e-01 -8.36037993e-01
1.11200474e-02 -8.08845460e-02 -7.46916294e-01 -1.17380120e-01
3.35497290e-01 -3.73453973e-03 -8.78333211e-01 1.22842801e+00
4.71709847e-01 1.35725349e-01 7.79311597e-01 -1.80500221e+00
-1.42090380e+00 7.96309114e-01 -4.74072844e-01 1.66694410e-02
2.16522157e-01 -7.47663602e-02 1.28931725e+00 -9.53414142e-01
6.55270934e-01 1.49061358e+00 5.77128172e-01 4.02660310e-01
-1.04221821e+00 -8.36214542e-01 2.48212904e-01 8.06178153e-01
-1.71868193e+00 -1.55496225e-01 8.11060667e-01 -5.33080339e-01
8.88634205e-01 -5.94053604e-02 7.46004701e-01 1.20980191e+00
-7.56357014e-02 9.99193311e-01 8.81487846e-01 -4.03766006e-01
2.46884972e-01 5.04156165e-02 6.25415325e-01 1.03632748e+00
7.89619923e-01 1.85654475e-03 -7.68364370e-01 -2.42627472e-01
4.74723339e-01 2.72916824e-01 -5.32783747e-01 -8.31866860e-01
-1.22836375e+00 1.07576716e+00 6.57371700e-01 7.90199935e-02
-1.00254387e-01 1.35014370e-01 4.64242846e-01 7.94965681e-03
5.46586692e-01 3.27797621e-01 -1.88904777e-01 -2.66328864e-02
-2.16608256e-01 1.49436772e-01 7.58768737e-01 1.34949064e+00
9.98398125e-01 -1.12600446e-01 -1.72053203e-01 1.01879549e+00
2.91765392e-01 2.72689521e-01 5.45419753e-01 -6.17278039e-01
2.95752376e-01 1.10525453e+00 -4.54501152e-01 -1.11487830e+00
-2.26431862e-01 -5.33223212e-01 -8.60554338e-01 -1.49951130e-01
-2.85286278e-01 3.25757593e-01 -8.66115868e-01 1.76216841e+00
3.06991577e-01 7.44812727e-01 5.15024364e-01 6.93272889e-01
1.27905333e+00 3.58441114e-01 4.66288552e-02 3.39566857e-01
1.50432014e+00 -1.13444972e+00 -9.19747114e-01 -1.25022121e-02
7.10762143e-01 -2.22244918e-01 1.14896810e+00 1.19751357e-01
-3.69025558e-01 -1.56029627e-01 -1.28253663e+00 -3.06405991e-01
-9.41230714e-01 -9.35145468e-02 1.17578340e+00 5.78798532e-01
-7.24215984e-01 4.82132912e-01 -6.19410336e-01 -4.29789692e-01
8.71308565e-01 4.09261435e-02 -5.91147661e-01 -5.60156047e-01
-1.68120503e+00 6.79839849e-01 6.42001450e-01 -2.53506809e-01
-7.45022118e-01 -9.12813425e-01 -1.20208633e+00 3.53937179e-01
7.68510818e-01 -9.29328024e-01 7.86338866e-01 -1.42643377e-01
-1.11011636e+00 7.61104465e-01 2.92293817e-01 -7.89134353e-02
-1.16792157e-01 -2.04276681e-01 -5.76825917e-01 2.01843023e-01
2.72471845e-01 5.49388766e-01 4.72969621e-01 -1.28121221e+00
-1.95516139e-01 -5.76032102e-01 5.79359174e-01 1.90750301e-01
-7.48125017e-01 -6.01995051e-01 -6.45211041e-01 -7.80906141e-01
9.85967442e-02 -8.10979187e-01 1.02262266e-01 -6.59493497e-03
-5.20368814e-01 -4.57702219e-01 8.01204145e-01 -2.93607086e-01
1.27131927e+00 -2.04368567e+00 3.22842926e-01 -2.25817114e-01
6.07463479e-01 2.21810192e-01 -2.84405440e-01 6.44147635e-01
-1.45264149e-01 2.27014259e-01 -3.41261923e-01 -5.07826269e-01
3.56244296e-01 5.57724893e-01 -2.56552547e-01 2.70525903e-01
9.13348123e-02 1.13473439e+00 -1.32550097e+00 -4.21726346e-01
1.69462517e-01 6.12560153e-01 -5.86215436e-01 -1.97465286e-01
-1.67465936e-02 -1.54580280e-01 -3.72368127e-01 9.21938837e-01
5.30755937e-01 -5.29947281e-01 5.05175292e-02 -6.02174520e-01
4.66761142e-01 -5.75865246e-02 -9.05221939e-01 2.07544541e+00
-4.08510864e-01 3.36703569e-01 -7.46447682e-01 -9.05187309e-01
7.74655461e-01 2.87438840e-01 3.72911215e-01 -4.24490958e-01
1.35156661e-01 -1.49335995e-01 -2.17798978e-01 -5.82109511e-01
3.46527636e-01 -1.26887560e-01 -4.58228886e-02 2.33544230e-01
6.91267848e-01 4.86684367e-02 2.71970183e-01 7.27134466e-01
1.07688034e+00 5.45796491e-02 5.86281419e-01 -2.59146214e-01
3.11751366e-01 -1.69634312e-01 6.24660313e-01 5.37817717e-01
-2.10372105e-01 4.33875591e-01 7.17513800e-01 -1.50583804e-01
-5.16235471e-01 -1.31268740e+00 -9.11607072e-02 7.64080644e-01
5.64745367e-01 -8.73388112e-01 -2.86293060e-01 -7.22929597e-01
1.49407059e-01 9.29459393e-01 -8.07659686e-01 -7.65127540e-01
2.86523938e-01 -8.73964429e-01 6.90605462e-01 5.24564922e-01
6.54839814e-01 -7.29938269e-01 -4.98996563e-02 -2.77513526e-02
-2.01714188e-01 -1.37846124e+00 -1.75442547e-01 -2.68023282e-01
-6.23203576e-01 -1.66311657e+00 -4.33020294e-01 -4.74970341e-01
5.12080848e-01 5.33462167e-01 1.15709853e+00 -1.53658092e-01
-6.26232028e-01 7.42775857e-01 -8.50382328e-01 -5.03595412e-01
1.96371928e-01 6.47167638e-02 9.45846066e-02 1.52133048e-01
7.30742931e-01 -8.32627833e-01 -5.55312932e-01 1.08583160e-01
-1.13638651e+00 2.61275917e-01 4.73001182e-01 9.30528700e-01
5.72955668e-01 2.67016917e-01 7.53890812e-01 -1.22808194e+00
5.39672852e-01 -8.55338573e-01 -2.49325171e-01 4.84925777e-01
-7.74441898e-01 1.77916467e-01 2.60406852e-01 -2.92738348e-01
-8.78928900e-01 -4.96374488e-01 1.59861743e-01 -1.01286352e+00
8.05140883e-02 7.47728765e-01 -4.62698281e-01 -5.61952777e-02
5.33111453e-01 1.37722880e-01 3.99758341e-03 -4.34734285e-01
8.09655368e-01 5.37702978e-01 2.52489835e-01 -6.35192871e-01
7.75633633e-01 5.83658695e-01 -9.90599468e-02 -7.20866680e-01
-1.13654363e+00 -6.15133762e-01 -5.25728106e-01 -4.15160954e-02
7.56631911e-01 -1.09628165e+00 -6.05758369e-01 3.37976396e-01
-8.22500050e-01 4.70156297e-02 -4.60989565e-01 5.72466433e-01
-4.29865599e-01 3.36528480e-01 -4.41630125e-01 -3.78176242e-01
-2.62871206e-01 -9.94565845e-01 1.13039768e+00 1.62407488e-01
1.92110822e-01 -1.18348598e+00 2.50950068e-01 4.64548349e-01
3.17437381e-01 2.57567376e-01 1.32741868e+00 -6.32971227e-01
-7.21421063e-01 -2.40571961e-01 -3.20427328e-01 2.48526707e-01
3.50696236e-01 -3.18285018e-01 -1.08324134e+00 -2.69313753e-01
-4.93422270e-01 -4.10548329e-01 1.14896417e+00 -1.27337664e-01
1.14826453e+00 -2.13504016e-01 -3.74082386e-01 7.50852585e-01
1.69157922e+00 -2.16932774e-01 7.22474217e-01 2.27644101e-01
1.13063657e+00 4.30969775e-01 3.86397362e-01 3.65319163e-01
8.19281399e-01 6.65210783e-01 5.64619839e-01 3.60422730e-01
-4.26922828e-01 -4.16665345e-01 1.62344158e-01 1.23703003e+00
-1.39633447e-01 -1.59740373e-01 -8.64950597e-01 5.51421463e-01
-2.04361916e+00 -8.57456744e-01 -2.12272536e-02 1.89887571e+00
5.11503696e-01 -2.01178536e-01 -3.43695670e-01 3.33505608e-02
7.09580481e-01 4.07194972e-01 -5.70055842e-01 1.62277162e-01
-2.07641676e-01 3.03501576e-01 3.58912349e-01 4.20067817e-01
-1.09622347e+00 1.08682120e+00 5.08406830e+00 1.03224766e+00
-5.37692010e-01 4.61276084e-01 -2.12412365e-02 -1.97617397e-01
-5.77502608e-01 1.36254519e-01 -6.84985161e-01 2.64916509e-01
4.16725844e-01 -5.95217168e-01 3.39756846e-01 8.61743093e-01
-5.35370111e-01 2.16026366e-01 -9.66354549e-01 1.31845224e+00
3.98684710e-01 -1.52003109e+00 5.74308217e-01 -2.14253236e-02
7.72255063e-01 -3.54826450e-02 -7.65239820e-02 9.07822073e-01
3.71941060e-01 -8.96531582e-01 3.31283510e-01 5.88259518e-01
7.54788756e-01 -7.87240267e-01 7.88277984e-01 1.03982516e-01
-1.45721424e+00 7.68548017e-03 -5.98445475e-01 5.68781905e-02
-8.82992893e-02 6.41847610e-01 -4.01742190e-01 1.29353738e+00
6.67828202e-01 1.22150469e+00 -7.78712809e-01 8.33837509e-01
-6.76802933e-01 3.35621774e-01 1.25939131e-01 6.38039187e-02
7.69826397e-02 -2.52375454e-01 5.40791452e-01 8.65388215e-01
1.51116595e-01 3.23201895e-01 1.29714057e-01 1.04140007e+00
-2.22917095e-01 5.74319661e-02 -8.26002657e-01 -4.05227363e-01
5.01597166e-01 1.23354435e+00 -4.79398489e-01 -5.01308739e-01
-7.35077500e-01 9.03916776e-01 5.06443560e-01 5.29653549e-01
-8.30165207e-01 -5.51880598e-01 1.18468952e+00 -2.80837715e-01
7.91904032e-02 -2.53361110e-02 2.68585622e-01 -1.67431474e+00
-1.05697379e-01 -4.55794841e-01 6.58939302e-01 -9.10127163e-01
-1.67188418e+00 4.91956621e-01 1.73072636e-01 -1.15104723e+00
2.97435939e-01 -6.57515764e-01 -2.36086875e-01 5.56393027e-01
-1.71766245e+00 -1.37763321e+00 -6.47707999e-01 7.63385534e-01
3.65459979e-01 -3.60125929e-01 1.18571222e+00 5.73217452e-01
-7.11339056e-01 6.13947213e-01 2.51953839e-03 1.01974152e-01
4.22886848e-01 -1.25401592e+00 1.64737776e-01 5.13561845e-01
3.69635999e-01 6.94533408e-01 1.82913914e-01 -5.65208018e-01
-1.51447761e+00 -1.37055993e+00 4.73363250e-01 -4.67871279e-01
8.93024445e-01 -4.25439417e-01 -9.68791842e-01 8.93784583e-01
8.22048113e-02 2.98257709e-01 1.10618067e+00 4.48737949e-01
-9.22077537e-01 7.04292506e-02 -7.79816747e-01 6.35289609e-01
1.53830087e+00 -9.33289528e-01 -7.38263488e-01 3.66019934e-01
1.22014236e+00 -5.48039153e-02 -9.87759173e-01 5.44698954e-01
3.73639256e-01 -8.02579761e-01 1.07626390e+00 -8.04768443e-01
5.62303782e-01 -3.27517718e-01 -4.23330098e-01 -1.54489148e+00
-4.83134747e-01 -1.08286291e-01 -4.74514514e-01 1.21969056e+00
-6.49454519e-02 -9.42191720e-01 6.84028268e-01 2.13910162e-01
-2.74798185e-01 -1.04383123e+00 -7.81433284e-01 -1.02479541e+00
-2.78843164e-01 -4.61638749e-01 8.29749823e-01 1.49813056e+00
2.13264003e-01 5.86539865e-01 -3.40569109e-01 4.41234231e-01
9.30813432e-01 1.66597202e-01 6.17466927e-01 -1.28974628e+00
-1.95209011e-01 -4.64317530e-01 -1.12834537e+00 -5.08397698e-01
3.55832666e-01 -1.56717634e+00 -6.03680253e-01 -1.84656513e+00
5.62284172e-01 -3.62436801e-01 -5.54242492e-01 7.52911091e-01
-3.94657731e-01 1.99893303e-02 2.73631155e-01 5.34465909e-02
-7.22821772e-01 1.27422321e+00 1.20682132e+00 -4.33499783e-01
1.94149643e-01 -5.27856410e-01 -9.59008813e-01 6.27044082e-01
6.44114316e-01 -3.87290061e-01 -9.67401266e-01 -3.43425691e-01
3.24417502e-01 -2.38657042e-01 6.55251622e-01 -9.86249983e-01
3.67557943e-01 2.03989282e-01 -2.51685947e-01 -3.76508921e-01
5.41441679e-01 -7.21022308e-01 1.84722543e-01 1.33614257e-01
-1.06595322e-01 -4.12981302e-01 1.56765014e-01 1.06812453e+00
-4.86855656e-01 -1.72315136e-01 3.98871571e-01 -7.51461415e-03
-1.21933758e+00 8.82519126e-01 2.04530895e-01 1.14127219e-01
1.08880019e+00 -1.74833685e-02 -6.81853652e-01 -1.48369342e-01
-8.71933520e-01 4.11343306e-01 2.75768965e-01 8.88202786e-01
8.48910153e-01 -1.80931103e+00 -3.86355758e-01 1.42416447e-01
9.94848073e-01 -6.72968552e-02 6.65107846e-01 7.44043708e-01
-3.46080720e-01 4.67788488e-01 -1.02681153e-01 -3.45844179e-01
-9.50513303e-01 8.25872719e-01 1.47087619e-01 -1.32029161e-01
-9.08706427e-01 9.89375055e-01 5.60145319e-01 -7.65682399e-01
2.61967361e-01 -9.90029648e-02 -2.67648876e-01 1.50755227e-01
4.20643717e-01 4.02514905e-01 -9.68855768e-02 -4.70879465e-01
-3.48186314e-01 6.25988126e-01 1.17833531e-02 3.09720010e-01
1.26558113e+00 3.12798694e-02 -1.92129537e-01 6.44307256e-01
1.27229154e+00 -3.25613141e-01 -8.67028534e-01 -6.09762311e-01
-1.28309980e-01 -6.45720303e-01 1.19273417e-01 -5.79607844e-01
-1.28286433e+00 1.00956833e+00 4.68902320e-01 -8.00685585e-02
8.76043260e-01 1.75627753e-01 6.21622026e-01 3.15497398e-01
8.92719686e-01 -4.95508552e-01 2.07591832e-01 4.25093710e-01
7.53601611e-01 -1.21830583e+00 9.16819796e-02 -9.07144129e-01
-6.45857513e-01 9.17126894e-01 6.27687335e-01 -8.29823464e-02
9.99799013e-01 -2.74317116e-01 -2.50972807e-01 -7.07007706e-01
-7.92303622e-01 -5.98168612e-01 2.11883903e-01 5.58308363e-01
1.18750244e-01 3.20550621e-01 -4.25513312e-02 8.77798975e-01
-2.91058905e-02 2.81531066e-02 3.50830734e-01 8.32730830e-01
-1.28821850e-01 -1.09286976e+00 5.89511469e-02 5.07903814e-01
1.24382630e-01 -2.68353701e-01 -2.31029481e-01 9.31020796e-01
2.94148386e-01 7.58035898e-01 -2.19248876e-01 -5.93371689e-01
3.38847637e-01 2.39811227e-01 5.26570737e-01 -8.71999800e-01
1.94966555e-01 -4.40690905e-01 -1.12967566e-02 -7.33090580e-01
-4.07637179e-01 -8.63086879e-02 -1.16690421e+00 -3.02660376e-01
-3.31270397e-01 4.40663062e-02 2.31977195e-01 7.84790277e-01
6.54055417e-01 9.82470632e-01 6.99403062e-02 -5.10952532e-01
-2.64123261e-01 -8.54336560e-01 -1.05169559e+00 6.20015860e-01
-9.29179043e-02 -1.45893705e+00 -3.35063666e-01 -2.76117861e-01] | [8.681775093078613, 7.875489234924316] |
f3b7df1f-b37b-465c-805c-b6994727dc7d | general-purpose-question-answering-with-macaw | 2109.02593 | null | https://arxiv.org/abs/2109.02593v1 | https://arxiv.org/pdf/2109.02593v1.pdf | General-Purpose Question-Answering with Macaw | Despite the successes of pretrained language models, there are still few high-quality, general-purpose QA systems that are freely available. In response, we present Macaw, a versatile, generative question-answering (QA) system that we are making available to the community. Macaw is built on UnifiedQA, itself built on T5, and exhibits strong performance, zero-shot, on a wide variety of topics, including outperforming GPT-3 by over 10% (absolute) on Challenge300, a suite of 300 challenge questions, despite being an order of magnitude smaller (11 billion vs. 175 billion parameters). In addition, Macaw allows different permutations ("angles") of its inputs and outputs to be used, for example Macaw can take a question and produce an answer; or take an answer and produce a question; or take an answer and question, and produce multiple-choice options. We describe the system, and illustrate a variety of question types where it produces surprisingly good answers, well outside the training setup. We also identify question classes where it still appears to struggle, offering insights into the limitations of pretrained language models. Macaw is freely available, and we hope that it proves useful to the community. Macaw is available at https://github.com/allenai/macaw | ['Peter Clark', 'Oyvind Tafjord'] | 2021-09-06 | null | null | null | null | ['generative-question-answering'] | ['natural-language-processing'] | [-3.50567512e-02 2.59760648e-01 2.93530673e-01 -4.92743313e-01
-1.90812755e+00 -1.15438294e+00 6.78909779e-01 -9.90784168e-02
-3.50756526e-01 8.54249835e-01 4.36220556e-01 -8.48020196e-01
-7.52074569e-02 -8.39998245e-01 -6.06063664e-01 -2.53950685e-01
4.07904744e-01 1.08772278e+00 2.47657269e-01 -6.78164661e-01
-5.54060563e-02 -2.20273525e-01 -1.19968736e+00 6.22574389e-01
9.70313907e-01 6.85693622e-01 1.71807751e-01 9.74210739e-01
-3.65666717e-01 9.96858716e-01 -8.91047060e-01 -1.07266796e+00
-8.68748948e-02 -6.53828025e-01 -1.53167033e+00 -3.42927486e-01
6.01816416e-01 -2.18095973e-01 -7.95754790e-02 5.25900960e-01
6.62698507e-01 1.81740776e-01 4.10146415e-01 -1.09280968e+00
-1.25213242e+00 5.50352335e-01 3.08011621e-01 3.85409623e-01
7.88330615e-01 5.28058290e-01 1.54244983e+00 -8.68050694e-01
6.17068887e-01 1.30741155e+00 5.42062640e-01 6.77333057e-01
-1.15058005e+00 -3.02322209e-01 -1.72351629e-01 2.25218818e-01
-1.09641922e+00 -4.44894761e-01 2.23462433e-01 -1.71397462e-01
1.25663626e+00 6.01500690e-01 1.72449559e-01 1.17512012e+00
2.96706468e-01 7.93808401e-01 1.11265516e+00 -3.41569752e-01
2.33751401e-01 -1.28367662e-01 2.37078771e-01 6.04651153e-01
-2.54684001e-01 -5.67738831e-01 -3.22022378e-01 -6.73626721e-01
2.52199829e-01 -4.92129922e-01 -3.46244752e-01 1.97954178e-01
-1.04746902e+00 1.16830897e+00 2.85800964e-01 3.17420810e-01
-2.43637085e-01 5.88606894e-02 -4.88697225e-03 6.44180775e-01
1.79205120e-01 9.04796898e-01 -6.77739441e-01 -5.34865260e-01
-6.20607197e-01 7.99663961e-01 1.16241300e+00 9.79049087e-01
7.34510064e-01 -2.79040784e-01 -5.17639518e-01 8.98621261e-01
8.48249197e-02 6.35056496e-01 4.27886337e-01 -1.28525639e+00
6.91574872e-01 5.30685365e-01 2.60441363e-01 -5.80051184e-01
-2.52194583e-01 -2.38966361e-01 -4.50043112e-01 -3.33927989e-01
6.92816019e-01 -3.28491330e-01 -9.73009467e-01 1.73790646e+00
1.10323153e-01 -3.84857059e-01 2.61475444e-01 7.40146339e-01
1.31835353e+00 7.72250652e-01 2.49843001e-01 3.48620296e-01
1.84554613e+00 -9.96478975e-01 -6.62216663e-01 -6.58967733e-01
7.93969214e-01 -1.14597440e+00 1.70516264e+00 2.70770490e-01
-1.38980079e+00 -3.92345339e-01 -4.92135406e-01 -6.23848796e-01
-4.61217791e-01 -1.46088719e-01 5.94361722e-01 7.51158297e-01
-1.40674198e+00 7.71132186e-02 -4.35005456e-01 -4.42464978e-01
-1.97665603e-03 2.64168419e-02 -1.99467912e-01 -6.11037493e-01
-1.62608719e+00 1.12609994e+00 -5.51950224e-02 -6.01966716e-02
-8.13617170e-01 -6.29575968e-01 -8.60830724e-01 8.80723894e-02
6.33166730e-01 -1.24342406e+00 1.92439306e+00 -4.65297431e-01
-1.46445000e+00 6.61815464e-01 -4.76839006e-01 -3.86860877e-01
1.36510998e-01 -3.26063037e-01 -6.05598629e-01 2.43006662e-01
3.14715713e-01 8.08855593e-01 2.82728136e-01 -7.49299467e-01
-3.30655038e-01 -6.60447702e-02 7.20054984e-01 2.05301732e-01
2.42059842e-01 2.40415245e-01 -4.64733064e-01 -3.14936429e-01
-6.82689436e-03 -6.71838641e-01 -8.49966109e-02 -4.84727025e-01
-3.33433568e-01 -7.55747914e-01 4.36763227e-01 -9.24482584e-01
1.14055634e+00 -1.51634777e+00 -5.54664321e-02 -3.62727016e-01
2.37064604e-02 1.99851245e-01 -3.71156245e-01 9.59573030e-01
1.87513918e-01 3.43655884e-01 -5.60985327e-01 -2.55956441e-01
2.87676305e-01 3.57844770e-01 -5.21469593e-01 -1.91963419e-01
5.40240228e-01 1.46543980e+00 -9.22046065e-01 -3.40773016e-01
-1.17218301e-01 3.21772754e-01 -6.29996955e-01 3.59247118e-01
-8.18669677e-01 1.49016708e-01 -4.61651564e-01 6.11712933e-01
3.95015270e-01 -6.20998144e-01 3.30050439e-02 1.34460360e-01
2.61203438e-01 1.00305164e+00 -7.33747125e-01 1.70441747e+00
-5.08664429e-01 5.67020655e-01 2.84290314e-03 -3.25918019e-01
7.81480074e-01 5.82061112e-01 -2.53806472e-01 -6.68168306e-01
3.36417817e-02 3.82116735e-01 4.28872965e-02 -6.08466566e-01
7.09342301e-01 -2.81260401e-01 -3.30607146e-01 7.18153417e-01
4.29903418e-01 -7.56642759e-01 3.66258562e-01 6.27914906e-01
1.35545266e+00 -6.27882704e-02 -3.03268004e-02 -1.81196272e-01
4.21748847e-01 2.48862192e-01 5.63318729e-02 9.81509924e-01
-2.48969998e-02 7.47176111e-01 5.34779131e-01 -2.02491850e-01
-7.77678668e-01 -1.21544361e+00 4.88558747e-02 1.33058095e+00
-2.78835356e-01 -6.68777764e-01 -6.66969478e-01 -5.89159608e-01
-2.65027374e-01 1.28019309e+00 -5.03459990e-01 1.22508958e-01
-6.31846726e-01 -6.38073504e-01 8.09777081e-01 3.13381761e-01
4.76101786e-01 -1.17516398e+00 -3.16581160e-01 2.06965134e-01
-8.41195762e-01 -9.39491153e-01 -5.14298320e-01 5.24960347e-02
-6.02045178e-01 -1.14883614e+00 -8.21228027e-01 -6.22426093e-01
1.19985111e-01 8.84026140e-02 1.84694993e+00 4.41426132e-03
1.97727457e-01 7.55373597e-01 -3.79289746e-01 -1.79309562e-01
-5.74615955e-01 4.26768601e-01 -6.66548848e-01 -4.30031687e-01
4.35218841e-01 -2.16921523e-01 -6.17704034e-01 1.23970896e-01
-1.06118000e+00 -1.50588900e-01 3.94264013e-01 6.98237181e-01
1.99082106e-01 -5.35277069e-01 9.09010530e-01 -9.59656000e-01
1.23198044e+00 -9.59748864e-01 -3.08259279e-01 5.43182969e-01
-3.47874135e-01 1.15969054e-01 5.86714268e-01 2.09281780e-02
-1.08053291e+00 -6.31713212e-01 -7.77892411e-01 1.32903069e-01
-1.70245245e-01 5.86814404e-01 -5.73881939e-02 3.53113621e-01
8.85770977e-01 2.53388226e-01 -1.40258908e-01 -6.47888362e-01
7.75733590e-01 5.87687969e-01 6.07498586e-01 -6.78808987e-01
6.26891315e-01 -2.51421221e-02 -6.94815457e-01 -4.09868509e-01
-8.66254807e-01 -3.75032157e-01 2.61497926e-02 1.16147913e-01
9.93757188e-01 -8.13787997e-01 -6.12190425e-01 2.23254755e-01
-1.22500956e+00 -5.55317700e-01 -1.58892736e-01 -3.53597887e-02
-4.33017284e-01 2.58732677e-01 -8.33480418e-01 -6.17867768e-01
-5.96103370e-01 -1.09098756e+00 9.00568783e-01 3.68727177e-01
-4.94301647e-01 -1.17974496e+00 4.00625989e-02 9.37303364e-01
7.77050197e-01 -2.36547023e-01 1.31406713e+00 -9.77336943e-01
-6.29345655e-01 -8.37561488e-02 2.18334813e-02 3.06395531e-01
2.40443349e-02 -9.93718579e-02 -7.66390800e-01 -1.08138472e-01
7.83379599e-02 -8.28465641e-01 6.50655508e-01 -6.47244826e-02
8.65205526e-01 -3.81534219e-01 9.52260271e-02 8.10140818e-02
1.09645057e+00 -3.40052769e-02 7.10016847e-01 2.85348922e-01
2.37139419e-01 6.10438108e-01 1.92952171e-01 -1.08281918e-01
1.15322852e+00 5.27498782e-01 3.50496948e-01 3.10072631e-01
-1.88780487e-01 -2.71097749e-01 3.63811493e-01 1.05462301e+00
3.52299452e-01 -6.65087581e-01 -1.12234199e+00 7.31296599e-01
-1.42485011e+00 -7.56686270e-01 -3.78333241e-01 2.02984023e+00
1.10412443e+00 -7.86549300e-02 -5.46361841e-02 -4.71247286e-01
3.05534273e-01 2.38362625e-01 -4.59376186e-01 -6.15008712e-01
-4.43060964e-01 8.14801455e-01 -1.25702813e-01 8.60465229e-01
-5.97479165e-01 9.67478216e-01 7.22842121e+00 8.50123942e-01
-6.27401650e-01 3.81570101e-01 6.79955900e-01 1.32565871e-01
-1.01082671e+00 1.71322644e-01 -7.00568736e-01 3.20276946e-01
1.35185766e+00 -3.82251054e-01 3.67684215e-01 5.48625469e-01
-4.06829715e-01 -3.22350174e-01 -9.65625584e-01 5.79421520e-01
1.89796373e-01 -1.34642804e+00 2.99352318e-01 -4.11971450e-01
4.83695775e-01 1.65001497e-01 -8.87836237e-03 9.31777239e-01
8.43766332e-01 -1.24828279e+00 3.90880138e-01 4.28030431e-01
4.36053187e-01 -5.39254129e-01 7.55476594e-01 5.22668481e-01
-6.70966148e-01 2.66558994e-02 -3.95897597e-01 -1.52501702e-01
4.15081322e-01 2.92810529e-01 -8.26741159e-01 7.29246795e-01
6.80982292e-01 -2.36024708e-01 -8.86115134e-01 8.68488967e-01
-6.09679043e-01 9.72028255e-01 -3.71509969e-01 -2.39165694e-01
4.72453296e-01 -5.59672639e-02 2.37851828e-01 9.65012431e-01
3.64519894e-01 4.34005916e-01 -1.74854487e-01 9.09126818e-01
-2.33432442e-01 1.70143485e-01 -2.12494522e-01 -8.03376213e-02
5.03529847e-01 1.20035315e+00 -1.21401131e-01 -4.49636459e-01
-4.20542836e-01 9.88933861e-01 4.44738209e-01 5.33259034e-01
-7.03943491e-01 -4.16885674e-01 4.29399371e-01 -5.08458801e-02
3.17720436e-02 -1.91035554e-01 1.37113929e-01 -1.20698071e+00
8.42245519e-02 -1.44006348e+00 8.43153477e-01 -1.37123930e+00
-1.47579408e+00 9.30037379e-01 3.59459668e-02 -4.42653745e-01
-7.61142075e-01 -4.57162946e-01 -6.38114810e-01 1.25734663e+00
-1.23941088e+00 -9.29023385e-01 -1.82763293e-01 6.89725459e-01
6.43395185e-01 7.50357807e-02 1.12062848e+00 2.65509695e-01
-1.79343596e-01 5.71195960e-01 -1.35528013e-01 3.05549111e-02
8.27061713e-01 -1.36691034e+00 8.92885089e-01 4.69961882e-01
4.52827066e-01 8.11937034e-01 7.76849568e-01 -3.60735089e-01
-1.44264996e+00 -8.19284737e-01 1.53275514e+00 -1.13794315e+00
7.60706723e-01 -4.72251862e-01 -1.23209906e+00 9.23881173e-01
8.20759952e-01 -5.31984985e-01 8.90737712e-01 1.46606341e-01
-4.35128540e-01 3.03732425e-01 -9.70457613e-01 7.92869747e-01
5.23848712e-01 -6.82462692e-01 -1.12264824e+00 7.50426233e-01
1.16885984e+00 -5.45046270e-01 -7.14462221e-01 2.03016996e-01
2.44885400e-01 -9.60566700e-01 8.17858815e-01 -7.80927002e-01
5.21608591e-01 -2.67929062e-02 -2.57367581e-01 -1.30585611e+00
-3.50316674e-01 -8.11338425e-01 -3.40700448e-02 1.23885179e+00
8.44394743e-01 -1.03862715e+00 4.49163765e-01 9.21265841e-01
-2.96266884e-01 -1.03624737e+00 -1.08491516e+00 -5.47801852e-01
7.33213961e-01 -5.72715819e-01 8.34182322e-01 7.13775277e-01
-2.40047127e-01 7.18585014e-01 -6.38820380e-02 -8.67196843e-02
9.38980356e-02 1.42777294e-01 7.17642784e-01 -6.85004294e-01
-5.82624376e-01 -3.38409364e-01 3.67782593e-01 -1.45727217e+00
-2.97815710e-01 -8.82354498e-01 -1.43029764e-01 -1.98571324e+00
-1.66021734e-01 -3.87038469e-01 1.15460530e-01 5.96772254e-01
-5.95458984e-01 3.42864990e-01 2.54063427e-01 1.49404392e-01
-6.21470690e-01 5.82444549e-01 1.22653675e+00 -2.84286570e-02
2.28464574e-01 3.30941230e-02 -1.11255682e+00 3.70396137e-01
8.81447196e-01 -2.93606162e-01 -4.51680690e-01 -9.70110416e-01
4.97973323e-01 2.23284602e-01 4.17742223e-01 -8.62630427e-01
2.04058766e-01 4.29594517e-02 1.01908617e-01 -5.79834878e-01
6.23112023e-01 -8.31615031e-02 5.80060147e-02 1.19316295e-01
-3.88124526e-01 6.27771199e-01 3.83250684e-01 6.18702956e-02
-3.00005943e-01 -5.47688901e-01 2.82065898e-01 -4.67642874e-01
-4.74970341e-01 5.23590595e-02 -4.78762954e-01 8.05866599e-01
4.68006939e-01 2.21399799e-01 -9.47047412e-01 -9.77363467e-01
-4.82859284e-01 6.31611347e-01 8.37110505e-02 5.98183393e-01
3.13061655e-01 -1.04334199e+00 -9.64585543e-01 -2.02918589e-01
2.60901541e-01 -4.77519780e-02 5.28668761e-01 5.21682680e-01
-3.92932922e-01 8.29499424e-01 3.57161164e-01 -2.95786321e-01
-8.31841230e-01 1.53293222e-01 4.88033414e-01 -5.12641430e-01
-1.40751988e-01 1.16521156e+00 2.08771285e-02 -1.04910135e+00
-2.31361762e-01 -3.08792621e-01 -2.79702041e-02 -6.32594079e-02
5.24331033e-01 1.36913791e-01 2.32537508e-01 -3.50846112e-01
-2.07139745e-01 1.09938286e-01 -2.91328002e-02 -4.05021846e-01
9.57731247e-01 -1.16033122e-01 -2.21501410e-01 4.80273962e-01
9.64600086e-01 -4.96620759e-02 -6.63880169e-01 -3.56762379e-01
-1.71340704e-01 -1.27009556e-01 -4.51047301e-01 -1.49697280e+00
-4.81995583e-01 1.05327380e+00 1.13211393e-01 4.81747925e-01
9.87166405e-01 5.05583465e-01 1.10856628e+00 6.14662051e-01
3.89437228e-01 -5.91842055e-01 2.11449921e-01 8.23646545e-01
1.24897075e+00 -1.07159877e+00 -3.14490676e-01 -1.70730561e-01
-6.75869465e-01 7.88603306e-01 6.46164358e-01 1.43755287e-01
5.02926297e-02 -1.37509912e-01 4.87843603e-01 -4.49314564e-01
-1.33570290e+00 -2.38467485e-01 2.29410350e-01 4.80933994e-01
6.87299073e-01 -4.38523293e-02 -2.81434625e-01 7.85547376e-01
-8.12025666e-01 -2.20767438e-01 4.81628239e-01 7.00032234e-01
-4.64768678e-01 -1.39864254e+00 -4.03894395e-01 4.41380858e-01
-6.87724113e-01 -5.54191172e-01 -4.56063837e-01 7.83687711e-01
-2.10402280e-01 1.51277113e+00 -1.93373397e-01 -6.55140281e-02
4.00960505e-01 6.29422545e-01 3.78045171e-01 -8.16386759e-01
-1.01700675e+00 -4.43913460e-01 4.74441260e-01 -3.22831482e-01
8.19148272e-02 -4.20125395e-01 -9.17721748e-01 -3.86744410e-01
-1.34860665e-01 6.14813745e-01 3.81211460e-01 1.06833768e+00
6.07143402e-01 2.07382202e-01 5.34005910e-02 -2.13209447e-02
-7.97450483e-01 -1.25490701e+00 6.28987923e-02 1.52669445e-01
1.69230819e-01 -2.28500679e-01 -4.27646011e-01 -1.43139720e-01] | [11.383686065673828, 8.142871856689453] |
4604e003-c243-429f-ac0f-64626de55f7c | asymreg-robust-symmetric-image-registration | 2303.10211 | null | https://arxiv.org/abs/2303.10211v2 | https://arxiv.org/pdf/2303.10211v2.pdf | SITReg: Multi-resolution architecture for symmetric, inverse consistent, and topology preserving image registration using deformation inversion layers | Deep learning based deformable medical image registration methods have emerged as a strong alternative for classical iterative registration methods. Since image registration is in general an ill-defined problem, the usefulness of inductive biases of symmetricity, inverse consistency and topology preservation has been widely accepted by the research community. However, while many deep learning registration methods enforce these properties via loss functions, no prior deep learning registration method fulfills all of these properties by construct. Here, we propose a novel multi-resolution registration architecture which is by construct symmetric, inverse consistent, and topology preserving. We also develop an implicit layer for memory efficient inversion of the deformation fields. The proposed method achieves state-of-the-art registration accuracy on two datasets. | ['Pekka Marttinen', 'Joel Honkamaa'] | 2023-03-17 | null | null | null | null | ['deformable-medical-image-registration', 'medical-image-registration'] | ['medical', 'medical'] | [ 1.32318333e-01 -1.73508152e-02 -1.03367195e-01 -6.39616013e-01
-7.72843122e-01 -2.89320797e-01 5.89769363e-01 1.32651493e-01
-6.07987404e-01 6.18290842e-01 4.46972288e-02 1.28745764e-01
-4.81989592e-01 -8.01736593e-01 -4.98658180e-01 -7.70397007e-01
-4.64103296e-02 8.02091718e-01 2.72658944e-01 -3.21237355e-01
1.47513762e-01 5.80440521e-01 -1.01204121e+00 -1.44997463e-01
9.46704745e-01 7.73744941e-01 -7.01095238e-02 5.10682017e-02
2.29705945e-01 2.65506595e-01 1.65582046e-01 -2.95409620e-01
4.10585970e-01 -3.92642230e-01 -1.24657869e+00 -3.22864801e-01
7.35238969e-01 -3.80779237e-01 -4.42005664e-01 1.03942704e+00
7.77090311e-01 2.21417531e-01 5.38237572e-01 -9.62527692e-01
-8.41665924e-01 3.76900077e-01 -5.29727399e-01 1.65570319e-01
1.86070621e-01 -1.71599567e-01 9.44097519e-01 -7.12249398e-01
7.25765824e-01 6.54211164e-01 1.06877279e+00 4.44258749e-01
-1.36914492e+00 -5.56956053e-01 -2.70613015e-01 -9.00609493e-02
-1.37824309e+00 -3.10323924e-01 8.44336092e-01 -4.11373973e-01
5.24427414e-01 2.39710629e-01 5.47189415e-01 8.02602053e-01
4.31883752e-01 2.79825002e-01 1.32619548e+00 -1.65124908e-02
-1.17617913e-01 -3.27396184e-01 3.26662809e-02 7.65711606e-01
4.08430278e-01 1.47804677e-01 -3.01669747e-01 -1.98037654e-01
1.22867560e+00 1.21151514e-01 -2.80583978e-01 -5.43797195e-01
-1.35480332e+00 7.36091614e-01 5.85307777e-01 6.58696294e-01
-4.04351354e-01 2.22943947e-01 2.59700209e-01 1.79222241e-01
6.67278111e-01 3.82955700e-01 -1.37149543e-01 1.06457338e-01
-1.02168107e+00 2.39405215e-01 4.79199737e-01 4.72765297e-01
7.00640023e-01 -1.34249955e-01 5.42968027e-02 6.99290574e-01
5.10828376e-01 3.36759537e-01 5.03418267e-01 -9.39314127e-01
4.66671248e-04 6.81150854e-01 -1.94383323e-01 -1.23746133e+00
-7.04197884e-01 -3.34769487e-01 -1.45682359e+00 1.27056897e-01
3.96054476e-01 3.05830747e-01 -7.40951657e-01 1.87280786e+00
3.59700620e-01 4.86048430e-01 -2.38813400e-01 9.86661911e-01
9.40388441e-01 7.85415806e-03 1.25466049e-01 -9.59916413e-02
1.29434371e+00 -5.21969497e-01 -8.46231759e-01 1.49135649e-01
5.02642810e-01 -8.65111411e-01 7.04979718e-01 -2.29054272e-01
-1.33806551e+00 -3.70123476e-01 -9.71623719e-01 -4.44837719e-01
5.46750948e-02 -4.49716896e-01 8.01477671e-01 5.32455981e-01
-1.33156037e+00 8.68271708e-01 -1.24440682e+00 -3.26035202e-01
6.60944819e-01 7.61941671e-01 -7.99979746e-01 1.42823443e-01
-1.15454292e+00 1.21060419e+00 4.56320326e-04 3.23338300e-01
-4.81977254e-01 -8.84270966e-01 -9.21349764e-01 -4.49603498e-01
-2.85571009e-01 -9.49200749e-01 6.95518911e-01 -7.11343110e-01
-1.41478097e+00 1.62627053e+00 3.30792814e-02 -2.44222775e-01
8.30965698e-01 -2.22685263e-01 -1.92413226e-01 -5.48129436e-03
-7.32681379e-02 4.77095813e-01 4.21310633e-01 -1.16138136e+00
7.52276704e-02 -6.03682518e-01 -1.33522063e-01 1.62816122e-01
-2.18106002e-01 1.30412504e-01 -7.79950544e-02 -8.45392227e-01
6.35200799e-01 -1.01626801e+00 -4.41238761e-01 3.81357372e-01
-2.05692947e-01 -1.00841179e-01 6.31082773e-01 -8.65055144e-01
8.02623332e-01 -1.95027590e+00 5.98844141e-02 4.63467449e-01
3.44068855e-01 1.19817682e-01 -1.85114786e-01 3.13122720e-02
-8.38491768e-02 3.94413397e-02 -6.23905361e-01 -2.90262282e-01
-1.59794703e-01 1.87011600e-01 -1.21173196e-01 1.09890199e+00
-2.56699510e-02 1.19608736e+00 -8.68064344e-01 -6.15190983e-01
1.55365527e-01 1.04063249e+00 -6.90621674e-01 2.76701510e-01
4.42720771e-01 1.16462183e+00 -4.10306722e-01 3.90363812e-01
9.02602077e-01 -4.31885302e-01 2.62046486e-01 -5.81312895e-01
-4.23595943e-02 2.30463952e-01 -1.11363351e+00 2.23166656e+00
-1.20117173e-01 3.15513462e-01 5.74451163e-02 -1.17487943e+00
8.62995744e-01 5.92554092e-01 1.12737060e+00 -6.49201572e-01
3.14326525e-01 4.72836822e-01 2.19179466e-02 -3.54255527e-01
3.82197887e-01 -5.46077967e-01 2.62916148e-01 6.16304517e-01
-1.92983642e-01 -2.13893265e-01 -2.94204295e-01 -2.07940921e-01
7.10749030e-01 4.35334563e-01 2.26496696e-01 -6.00718975e-01
6.11299455e-01 -2.62466013e-01 6.83958709e-01 4.36663032e-01
-1.20802626e-01 1.03598452e+00 -1.17870942e-01 -8.82833064e-01
-9.63602066e-01 -1.20602632e+00 -5.67679107e-01 6.34133399e-01
5.10888457e-01 7.80239031e-02 -6.69705629e-01 -3.97337317e-01
-1.87823847e-01 -1.96173608e-01 -7.02405632e-01 8.04599782e-04
-1.24141455e+00 -8.28675985e-01 7.28138208e-01 4.43550467e-01
8.62018347e-01 -1.04629469e+00 -4.97227669e-01 1.68454856e-01
-5.48630893e-01 -1.19361937e+00 -7.20067382e-01 -3.89769763e-01
-1.24854028e+00 -1.16279602e+00 -7.91748583e-01 -1.07634020e+00
9.94673073e-01 -7.17663988e-02 1.16096449e+00 5.90876579e-01
-2.93920934e-01 2.18486965e-01 8.79954770e-02 2.48834953e-01
-5.13451993e-01 1.90938562e-01 1.43383980e-01 -3.43020111e-02
-1.24891147e-01 -9.84855533e-01 -9.40524101e-01 4.95856881e-01
-1.02011573e+00 -3.20376754e-02 4.93606776e-01 6.48049593e-01
9.71773148e-01 -4.11135525e-01 2.92167932e-01 -8.74567866e-01
3.83070260e-01 5.02736717e-02 -4.55956459e-01 2.64231473e-01
-5.98755479e-01 3.10422748e-01 1.07268803e-01 -3.58640611e-01
-9.53200459e-01 1.70399129e-01 -5.02849996e-01 -1.81870349e-02
-9.93715823e-02 5.41894317e-01 1.39561594e-01 -7.66831279e-01
4.38379645e-01 1.32877693e-01 4.13449258e-01 -3.47689450e-01
1.85412586e-01 -2.15906799e-02 6.50774121e-01 -7.83042789e-01
1.11354697e+00 9.55311060e-01 3.58687997e-01 -5.90118468e-01
-6.30572259e-01 -3.09078544e-01 -1.15191901e+00 1.08576573e-01
1.18100119e+00 -6.94008470e-01 -7.68439591e-01 7.01168358e-01
-1.18570375e+00 -2.80212730e-01 -3.09596896e-01 5.39310753e-01
-6.88462079e-01 7.53962159e-01 -6.50225341e-01 -1.88738272e-01
-7.89183259e-01 -1.30681860e+00 9.78497624e-01 -4.55139279e-02
-3.16429973e-01 -1.34074402e+00 4.29566681e-01 2.26238087e-01
8.50952446e-01 6.93818092e-01 6.15032792e-01 -5.35603523e-01
-5.49697876e-01 -1.64056808e-01 -2.29740784e-01 1.00010045e-01
5.39055586e-01 -5.30947924e-01 -7.58475065e-01 -4.79450554e-01
1.76010817e-01 -1.96611539e-01 6.80781662e-01 5.04417181e-01
1.04987228e+00 -2.21622407e-01 -3.10582966e-01 1.06846893e+00
1.51359117e+00 -2.67481655e-01 9.42712367e-01 3.14422756e-01
9.23124194e-01 5.04312277e-01 9.37880650e-02 1.45712435e-01
6.44594431e-01 9.09452021e-01 3.62822801e-01 -5.98210633e-01
-6.24334365e-02 1.89712703e-01 -2.64868643e-02 1.10723794e+00
-7.27670133e-01 4.80016172e-01 -8.69953156e-01 3.63499522e-01
-1.66509485e+00 -9.99177277e-01 -2.06900954e-01 2.37199306e+00
1.32700646e+00 -2.58577883e-01 -1.40689552e-01 -1.85409069e-01
6.22078359e-01 2.06276000e-01 -1.90226167e-01 1.58499584e-01
-2.49864638e-01 5.22840202e-01 4.56819326e-01 5.55098772e-01
-1.38728058e+00 6.84611201e-01 6.16265249e+00 2.68037647e-01
-1.29362559e+00 3.79823834e-01 5.15013158e-01 5.50119400e-01
-4.22685802e-01 -2.40808427e-02 -3.76671523e-01 -1.49078167e-03
3.19656819e-01 8.19562674e-02 2.74529397e-01 2.09074393e-01
1.00334100e-01 2.78054863e-01 -1.20519149e+00 1.04330301e+00
7.07228482e-02 -1.48359191e+00 5.28083369e-02 3.17973256e-01
8.34356606e-01 8.40035155e-02 1.23211943e-01 -3.65743250e-01
1.67542696e-01 -1.32779837e+00 3.25558335e-01 6.15370035e-01
9.32338595e-01 -3.44518453e-01 7.90168226e-01 -1.06209219e-01
-1.31410432e+00 6.39188766e-01 -8.86005461e-02 2.87378788e-01
4.66220737e-01 4.88600671e-01 -1.33010119e-01 7.90398777e-01
5.72793722e-01 7.61219442e-01 -3.76342565e-01 1.03407300e+00
-5.09368964e-02 2.09583506e-01 -2.53241718e-01 7.39060342e-01
-2.00655013e-02 -5.40561020e-01 4.75368083e-01 9.31230545e-01
2.55759478e-01 3.02012384e-01 2.24301547e-01 9.22724843e-01
-1.80623740e-01 7.47777149e-02 -5.59415042e-01 3.89185727e-01
8.30487236e-02 1.31624842e+00 -8.59289289e-01 1.55356064e-01
-2.80950427e-01 9.17037308e-01 1.50548622e-01 1.11539155e-01
-7.74827123e-01 6.54012188e-02 7.32492626e-01 3.39623928e-01
-8.45058337e-02 -4.89984870e-01 -4.40276772e-01 -1.16251457e+00
1.56751826e-01 -6.37297153e-01 4.16854501e-01 -1.10310502e-01
-1.54574728e+00 6.01811409e-01 -1.53854057e-01 -1.19719124e+00
8.03217366e-02 -3.11563283e-01 -6.03463471e-01 8.10531259e-01
-1.87539923e+00 -1.54123521e+00 -5.95348001e-01 8.20986629e-01
-1.48177937e-01 1.71987027e-01 1.00958741e+00 6.22474790e-01
-6.83256164e-02 6.20421588e-01 -2.47509420e-01 5.63459039e-01
9.79104996e-01 -1.04504538e+00 2.31123894e-01 5.44621050e-01
1.66162401e-02 1.10769749e+00 4.65053886e-01 -6.18321240e-01
-1.39119303e+00 -8.93728673e-01 9.11449671e-01 -4.12453711e-01
4.59471494e-01 -2.75303330e-02 -1.34685516e+00 7.68660247e-01
2.91718006e-01 6.05542600e-01 5.61902821e-01 -1.93451270e-01
-3.21000129e-01 -2.20380664e-01 -1.23173380e+00 2.86789298e-01
1.09820962e+00 -5.32495499e-01 -5.38326502e-01 4.02923316e-01
2.83881426e-01 -7.86780596e-01 -1.42479563e+00 7.70827949e-01
6.45821452e-01 -7.75200427e-01 1.34658754e+00 -3.53873283e-01
1.31880730e-01 -2.42088526e-01 1.23391420e-01 -8.38225186e-01
-3.63518536e-01 -7.37159848e-01 3.27576756e-01 9.91904974e-01
1.13945395e-01 -9.65446591e-01 6.06509805e-01 7.63121843e-01
-2.58373797e-01 -5.96790493e-01 -1.26985729e+00 -6.99716270e-01
4.62516040e-01 -4.05334495e-03 4.87743914e-01 1.50395536e+00
-3.03897738e-01 -1.04398318e-01 -3.13110709e-01 1.83667123e-01
8.93999755e-01 2.06215784e-01 4.75887924e-01 -1.54297221e+00
4.22282703e-02 -6.39328301e-01 -4.49986875e-01 -8.50313067e-01
4.08209562e-01 -1.33720684e+00 2.56597754e-02 -1.72987962e+00
2.67314196e-01 -9.72846568e-01 -4.82883781e-01 6.31655633e-01
4.99907881e-02 8.13947558e-01 -3.52741331e-01 5.31953514e-01
-5.20497620e-01 4.79266912e-01 1.45534313e+00 -4.43399772e-02
-1.11389533e-01 -1.88662902e-01 -4.35448259e-01 7.09592104e-01
7.12208927e-01 -4.06226158e-01 -8.34458172e-02 -5.68386972e-01
2.96125948e-01 -1.88429072e-01 5.09587705e-01 -7.60397077e-01
3.57822239e-01 -7.31509030e-02 1.07642636e-01 -1.58958748e-01
9.20788571e-02 -9.08877790e-01 4.66739237e-01 6.86146200e-01
-2.38322124e-01 3.54443729e-01 2.66765058e-03 1.97947562e-01
-2.24787980e-01 2.66812652e-01 1.17414093e+00 -5.00619560e-02
-3.30197126e-01 8.97522211e-01 3.10873628e-01 4.07558709e-01
6.20987892e-01 -2.75396079e-01 -4.13211174e-02 -6.20511826e-03
-4.87086415e-01 -1.75358310e-01 5.14273524e-01 3.35177958e-01
8.05448771e-01 -1.46863616e+00 -1.13880157e+00 2.72833377e-01
-1.78012803e-01 2.64740944e-01 1.60242423e-01 1.51240110e+00
-7.92599201e-01 -4.66435961e-02 -4.38245445e-01 -7.38397598e-01
-1.12799668e+00 -1.06332600e-02 6.15190387e-01 -5.33055902e-01
-9.02062953e-01 5.75762630e-01 2.49981612e-01 -6.78484678e-01
-2.15087354e-01 -1.07580550e-01 -1.52259380e-01 -2.68583357e-01
1.75147817e-01 2.91754156e-01 2.87280411e-01 -1.07232797e+00
-6.71961129e-01 1.14313686e+00 1.02925226e-01 7.23220780e-02
1.56290865e+00 -1.36274262e-03 -8.28649282e-01 -2.47746799e-02
1.13923538e+00 -1.62644193e-01 -9.90886033e-01 -3.82876933e-01
7.85224214e-02 -2.76058882e-01 -2.24278625e-02 -4.03166741e-01
-1.49388969e+00 6.32748187e-01 7.76909232e-01 -1.18019857e-01
9.70290303e-01 -3.79797630e-02 1.12022638e+00 1.62329644e-01
4.02650803e-01 -7.45719731e-01 3.45689803e-02 5.39742351e-01
1.01351357e+00 -1.41763711e+00 3.86274815e-01 -3.79869610e-01
-1.41877308e-01 1.01217413e+00 3.87963176e-01 -4.59407330e-01
9.44176793e-01 2.48703539e-01 1.44749910e-01 -3.70584935e-01
9.32282582e-02 1.47508621e-01 7.53075480e-01 5.89053035e-01
9.52722311e-01 -1.86721727e-01 -6.76157773e-01 1.22465253e-01
-8.84647667e-02 -8.48144293e-02 -6.07801415e-02 7.34865308e-01
1.10042304e-01 -1.37942386e+00 -1.37254506e-01 1.27054065e-01
-6.52121842e-01 -6.31756112e-02 3.68146896e-02 8.73782814e-01
-1.00971006e-01 3.44345421e-01 1.68186665e-01 3.80920023e-02
2.45459765e-01 -3.53453100e-01 7.60002613e-01 -2.80059099e-01
-8.81600976e-01 -1.81262773e-02 -3.51802617e-01 -6.38618469e-01
-9.82177079e-01 -5.67122161e-01 -1.61119533e+00 -3.14165354e-01
-2.30204508e-01 -1.87714472e-01 5.76549411e-01 1.17392445e+00
3.36035907e-01 1.53713405e-01 3.99008960e-01 -8.27711523e-01
-4.02676165e-01 -5.88499725e-01 -3.15109313e-01 9.78897333e-01
2.46227801e-01 -6.26536071e-01 -9.43040997e-02 2.42286310e-01] | [13.974376678466797, -2.562040328979492] |
512bbaf5-9324-48f3-b8a1-499d4362fa30 | pedagogical-rule-extraction-for-learning | 2112.13285 | null | https://arxiv.org/abs/2112.13285v2 | https://arxiv.org/pdf/2112.13285v2.pdf | Pedagogical Rule Extraction to Learn Interpretable Models - an Empirical Study | Machine-learning models are ubiquitous. In some domains, for instance, in medicine, the models' predictions must be interpretable. Decision trees, classification rules, and subgroup discovery are three broad categories of supervised machine-learning models presenting knowledge in the form of interpretable rules. The accuracy of these models learned from small datasets is usually low. Obtaining larger datasets is often hard to impossible. Pedagogical rule extraction methods could help to learn better rules from small data by augmenting a dataset employing statistical models and using it to learn a rule-based model. However, existing evaluation of these methods is often inconclusive, and they were not compared so far. Our framework PRELIM unifies existing pedagogical rule extraction techniques. In the extensive experiments, we identified promising PRELIM configurations not studied before. | ['Klemens Böhm', 'Benjamin Jochum', 'Vadim Arzamasov'] | 2021-12-25 | null | null | null | null | ['subgroup-discovery'] | ['methodology'] | [ 1.81169033e-01 7.09451616e-01 -9.09184873e-01 -5.17140090e-01
-3.41135293e-01 -4.61582869e-01 4.07686293e-01 4.67056215e-01
2.11153910e-01 1.32434177e+00 7.99095351e-03 -1.05120659e+00
-7.19913602e-01 -7.25374341e-01 -7.59231687e-01 -5.02256751e-01
4.79298132e-03 5.04996896e-01 4.10399705e-01 7.36003891e-02
4.14755404e-01 5.76117992e-01 -1.99784267e+00 7.83651054e-01
1.49747610e+00 9.16105807e-01 -1.21073641e-01 3.43997091e-01
-2.87520230e-01 1.25867593e+00 -3.80995870e-01 -4.29362178e-01
-1.41219785e-02 -3.90940636e-01 -9.03729081e-01 -1.22889228e-01
3.50748688e-01 -1.07617363e-01 1.01347394e-01 7.31036723e-01
-1.91082403e-01 -8.22894722e-02 9.06391799e-01 -1.40720487e+00
-4.31398064e-01 8.37269306e-01 -1.22145981e-01 -3.34589742e-02
5.10999560e-01 -2.88412482e-01 8.93592358e-01 -6.29035354e-01
6.82030678e-01 1.26364112e+00 5.32741904e-01 5.55068254e-01
-1.20593894e+00 -9.59469736e-01 4.05218035e-01 6.42431676e-01
-9.65198696e-01 -2.78126508e-01 7.37757981e-01 -5.48387527e-01
8.21698010e-01 6.50256693e-01 8.43790114e-01 1.00103021e+00
2.23168105e-01 7.93751717e-01 1.48049545e+00 -9.04741228e-01
3.00410450e-01 5.94495237e-01 4.93849128e-01 9.73408997e-01
7.49998927e-01 2.26197079e-01 -6.79663002e-01 -3.31419587e-01
4.88633096e-01 -1.09919377e-01 -3.19304056e-02 -2.15204686e-01
-7.11536825e-01 9.07906473e-01 -1.74160190e-02 4.13713783e-01
-1.08727649e-01 -3.81529659e-01 7.29988664e-02 4.01096791e-01
4.48578328e-01 7.08095014e-01 -9.62031066e-01 -5.56647480e-02
-5.26106596e-01 3.85060430e-01 1.02216804e+00 9.31058466e-01
4.78018254e-01 -2.43098676e-01 1.58471763e-01 6.36316419e-01
4.83699113e-01 2.81407405e-02 4.86607969e-01 -8.81569326e-01
1.69961959e-01 1.01291251e+00 -3.86156552e-02 -8.19612384e-01
-4.04434502e-01 -1.78860977e-01 -6.31739676e-01 1.61443099e-01
5.66821039e-01 -2.67132163e-01 -9.20377314e-01 1.23049045e+00
4.32536095e-01 2.95892119e-01 7.49846026e-02 3.57503444e-01
1.21360099e+00 2.96272904e-01 3.98199290e-01 -7.02212214e-01
1.18375301e+00 -6.24925733e-01 -8.23370099e-01 4.29894663e-02
1.05558479e+00 -5.57062328e-01 7.29347706e-01 1.17099154e+00
-6.80060923e-01 -3.54219139e-01 -7.95887649e-01 2.04758808e-01
-4.19173717e-01 1.72925726e-01 1.26196074e+00 7.18674898e-01
-5.50549507e-01 6.57196879e-01 -5.24993002e-01 -2.11863220e-01
5.59486210e-01 6.43551230e-01 -2.91765869e-01 5.72138578e-02
-1.07183433e+00 9.90877211e-01 6.32712960e-01 -2.91150063e-01
-3.71259511e-01 -9.27126050e-01 -7.16321230e-01 -1.70195363e-02
7.97553837e-01 -6.10512972e-01 1.26224387e+00 -8.30101192e-01
-1.47729957e+00 6.06645644e-01 -2.66964674e-01 -5.28640985e-01
1.28138568e-02 -1.64670736e-01 -5.32726169e-01 -6.81656599e-02
-2.52606958e-01 2.00885534e-01 6.30579114e-01 -1.35270953e+00
-9.48324144e-01 -4.18434739e-01 7.70937279e-02 -1.81159317e-01
-2.09235087e-01 1.07738480e-01 2.57988900e-01 -3.74966204e-01
2.13348538e-01 -6.89784408e-01 -5.01720309e-01 -3.85217965e-01
-4.26723868e-01 -7.85208225e-01 9.18401659e-01 -3.12777013e-01
1.71010995e+00 -1.55921018e+00 -2.26195708e-01 4.60946888e-01
4.82282877e-01 3.20698529e-01 4.10033196e-01 9.33991969e-02
-4.11065221e-01 5.79275250e-01 2.82750398e-01 4.41424727e-01
-3.01242143e-01 4.87307131e-01 -3.38139176e-01 -2.53984094e-01
1.67109996e-01 4.91349548e-01 -8.69917631e-01 -9.62424636e-01
4.01641935e-01 -1.48221284e-01 -6.18600130e-01 1.30423859e-01
-3.85002345e-01 3.11394602e-01 -9.02810335e-01 5.79285324e-01
8.30123574e-02 -2.56401658e-01 6.26014233e-01 8.26973692e-02
-1.21033877e-01 6.57185078e-01 -1.22686279e+00 1.05029321e+00
-2.96799242e-01 5.31110346e-01 -5.17654240e-01 -1.54180872e+00
9.14827764e-01 4.02161777e-01 3.66304189e-01 -1.78013161e-01
-6.49900958e-02 2.71661341e-01 3.91603470e-01 -1.09788215e+00
-2.27731898e-01 -3.22983086e-01 1.61051869e-01 1.20157532e-01
1.10199928e-01 -1.86650306e-01 3.42743248e-01 -6.34814426e-02
9.49826956e-01 2.76451439e-01 9.64083433e-01 -2.39232436e-01
4.33730990e-01 3.07129681e-01 8.52863610e-01 7.63488889e-01
2.33376876e-01 4.12923023e-02 7.09493399e-01 -8.93066406e-01
-4.11868423e-01 -6.78391457e-01 -5.75773895e-01 9.13141191e-01
-4.01836365e-01 -7.93345869e-01 -4.96448994e-01 -1.20737946e+00
1.28298104e-01 8.55677068e-01 -6.27098739e-01 5.98563161e-03
-2.46869117e-01 -7.18930185e-01 2.32014775e-01 2.99995750e-01
9.25792530e-02 -7.81770051e-01 -3.11437845e-01 1.52348414e-01
1.58793926e-01 -1.01082742e+00 4.29533631e-01 4.27331626e-01
-1.40301025e+00 -1.66474605e+00 4.56131756e-01 -7.18439281e-01
7.57185698e-01 -1.81575678e-02 9.88681495e-01 3.96950454e-01
-4.35871519e-02 -3.39800529e-02 -4.92663652e-01 -1.12725043e+00
-5.89610934e-01 1.13967307e-01 4.48447496e-01 -2.49118999e-01
7.87389278e-01 -6.40607893e-01 -1.11096568e-01 7.17279613e-02
-4.97288048e-01 1.52053475e-01 7.94964492e-01 8.45821083e-01
4.49503273e-01 2.17571944e-01 8.38724554e-01 -1.52857828e+00
5.80873132e-01 -4.94826883e-01 -3.41477662e-01 5.47910750e-01
-1.15539789e+00 2.41376728e-01 7.52544701e-01 -6.77527726e-01
-1.11882818e+00 7.45950341e-02 8.80460814e-02 -1.57235637e-01
-9.06713068e-01 6.85387135e-01 -2.22379848e-01 -1.24107957e-01
9.19914246e-01 -1.64028436e-01 -1.80801451e-01 -5.89499533e-01
7.63267130e-02 8.91722977e-01 -2.47745022e-01 -8.46607804e-01
5.81663966e-01 -3.01048964e-01 2.24416882e-01 -8.16194117e-01
-1.10524452e+00 -1.67200804e-01 -7.76359558e-01 -2.08896562e-01
4.24586266e-01 -4.36138153e-01 -7.86899447e-01 -2.94838190e-01
-9.05368149e-01 -3.12189385e-02 -2.52088070e-01 8.25464487e-01
-3.27503413e-01 1.42054871e-01 -2.08865136e-01 -9.50053096e-01
-4.78008203e-02 -8.93191040e-01 3.61494005e-01 4.63494807e-01
-9.28973079e-01 -1.09430575e+00 -3.88960764e-02 6.79965973e-01
-2.75107652e-01 1.59741834e-01 1.45976472e+00 -1.25689292e+00
-2.94873059e-01 -1.94120556e-01 3.87060821e-01 -7.01036006e-02
1.82028040e-01 4.88352954e-01 -1.03806937e+00 4.61171478e-01
-2.93467194e-01 -4.11929697e-01 4.94547784e-01 3.38547349e-01
1.93740761e+00 -7.10154116e-01 -6.47660255e-01 1.37189880e-01
9.13907588e-01 7.74455369e-01 5.06168783e-01 2.37482756e-01
3.88869226e-01 7.59847283e-01 8.39614749e-01 9.02837813e-02
1.64060637e-01 4.63196337e-01 5.57830520e-02 2.17326403e-01
1.10662609e-01 -4.01056141e-01 -3.65855880e-02 7.24785268e-01
-7.08471060e-01 2.75251627e-01 -1.04466248e+00 1.59397155e-01
-1.92929471e+00 -1.06232309e+00 -4.47695702e-01 1.96693873e+00
1.23320365e+00 5.12558818e-01 8.39828625e-02 5.28647959e-01
6.34404793e-02 -5.00045002e-01 -3.32052320e-01 -7.93234587e-01
2.91830480e-01 3.62108380e-01 -1.34684324e-01 4.77818459e-01
-8.57680619e-01 9.13418293e-01 7.39481592e+00 9.65225518e-01
-7.49602199e-01 -2.24113762e-01 8.25282753e-01 3.61913852e-02
-4.34174061e-01 1.78378075e-01 -1.01584375e+00 1.29042178e-01
1.07140553e+00 -1.74529448e-01 -8.32685083e-02 9.37803805e-01
4.31520909e-01 -3.15057069e-01 -1.40869141e+00 6.85846746e-01
-2.36620948e-01 -1.47309637e+00 3.51291358e-01 7.09671155e-02
6.60544693e-01 -8.02882016e-01 -3.03002834e-01 6.61215544e-01
6.93248451e-01 -1.27047944e+00 1.09294124e-01 5.56559622e-01
3.41009289e-01 -6.42497361e-01 5.97854912e-01 7.06242561e-01
-5.37954450e-01 -3.27844232e-01 -1.76648632e-01 -5.99435568e-01
-8.37162673e-01 9.36810732e-01 -1.07109356e+00 5.51444113e-01
6.59010589e-01 7.71355331e-01 -5.91875613e-01 1.03631079e+00
-6.67369723e-01 1.32953131e+00 -2.70337760e-01 -4.67282653e-01
-1.05825327e-01 -3.84850919e-01 2.69768089e-01 8.57566714e-01
3.84270735e-02 6.88138843e-01 1.09173216e-01 6.55150473e-01
1.65049970e-01 3.74698162e-01 -1.08591855e+00 1.26277849e-01
4.89574641e-01 1.14890099e+00 -4.09041375e-01 -4.82019186e-01
-5.01965284e-01 -1.65230427e-02 3.60345304e-01 3.06331128e-01
-4.32631999e-01 -1.54585391e-02 5.59154212e-01 3.48217070e-01
-2.44164541e-01 2.95657665e-01 -7.48683035e-01 -1.22736025e+00
-2.23334551e-01 -1.17665946e+00 6.19451940e-01 -6.25295520e-01
-1.04314876e+00 1.70202270e-01 4.61146921e-01 -1.21203101e+00
-4.82417524e-01 -7.91946530e-01 -5.65531254e-01 3.85217994e-01
-1.02788937e+00 -9.65964794e-01 2.03147307e-01 3.06966990e-01
5.52198946e-01 -3.66731793e-01 1.02574229e+00 -2.10321620e-01
-7.14595973e-01 5.79734623e-01 -9.40353647e-02 -1.20388910e-01
4.44256306e-01 -1.33662176e+00 -7.35854506e-01 4.21425402e-01
2.34181672e-01 7.07954586e-01 8.04922760e-01 -6.69589341e-01
-1.20330143e+00 -6.69863582e-01 1.14683175e+00 -5.77404916e-01
6.51728094e-01 -8.18785727e-02 -1.14384997e+00 6.59768641e-01
-3.20397504e-03 -1.34542048e-01 1.37189913e+00 8.57086778e-01
-1.60618335e-01 -2.55724251e-01 -1.08224809e+00 6.10169530e-01
9.48416591e-01 6.71790168e-02 -1.01600075e+00 3.55180264e-01
4.11428303e-01 -2.84602821e-01 -1.30540705e+00 5.28170824e-01
7.54768193e-01 -7.62691081e-01 8.15445602e-01 -1.67271662e+00
7.52802730e-01 -1.54498694e-02 2.64509171e-01 -1.36506295e+00
-2.83871174e-01 -5.31007648e-01 -7.12013602e-01 9.94116783e-01
9.59152222e-01 -3.71298432e-01 9.88107145e-01 1.09142637e+00
-3.46545540e-02 -1.30264366e+00 -5.48795283e-01 -6.02326930e-01
2.09081843e-01 -6.47424698e-01 7.52084792e-01 1.50637841e+00
7.12351084e-01 6.31166101e-01 -5.50742336e-02 7.16781318e-02
5.89050412e-01 4.12334830e-01 6.71701014e-01 -1.81412446e+00
-1.35188684e-01 -5.02649188e-01 -3.79590482e-01 -4.40841883e-01
2.67307758e-01 -8.18203568e-01 -4.82785046e-01 -1.53391111e+00
2.42164299e-01 -6.87717617e-01 7.28050247e-03 9.32050049e-01
-4.00370985e-01 -5.07926047e-01 -2.50765830e-01 -8.62037539e-02
-4.43387389e-01 1.82113037e-01 1.31282365e+00 4.90134023e-02
-3.99102420e-01 3.49898070e-01 -8.98143888e-01 1.27332509e+00
1.08103895e+00 -5.70297003e-01 -7.60336339e-01 3.64891559e-01
2.82283545e-01 7.13063404e-02 -4.98098470e-02 -6.27460837e-01
1.24694481e-01 -9.83090699e-01 5.43130934e-01 -2.39455149e-01
-2.08596110e-01 -9.73172903e-01 1.38283312e-01 4.95009452e-01
-4.50295866e-01 -5.87510407e-01 2.54147619e-01 3.39204401e-01
-2.00823665e-01 -5.30033290e-01 4.08720434e-01 1.30719347e-02
-6.58313632e-01 -2.46737301e-02 -5.55896044e-01 -1.21552691e-01
1.26043117e+00 -2.99441606e-01 -8.15292224e-02 -2.43645743e-01
-1.01130700e+00 2.09507719e-01 -2.41982579e-01 3.96908939e-01
7.19397962e-01 -1.21309173e+00 -5.04859805e-01 1.62689209e-01
1.17364906e-01 9.10492316e-02 -2.32136160e-01 7.55610526e-01
-2.29047135e-01 7.78642714e-01 3.32093164e-02 -2.73520947e-01
-1.58885550e+00 7.04580843e-01 3.69937837e-01 -5.24282753e-01
-2.76537836e-01 3.59554976e-01 -2.21248399e-02 -5.49824417e-01
3.86213124e-01 -7.79399216e-01 -8.47911179e-01 -2.40515172e-02
6.16692543e-01 3.42113048e-01 8.41902494e-02 1.51916593e-01
-2.76928961e-01 2.93105304e-01 -2.28180960e-01 4.84451294e-01
1.45536876e+00 4.58621204e-01 -7.99744565e-04 6.26595736e-01
5.74952543e-01 -7.78546035e-02 -4.30226207e-01 -8.51609409e-02
3.55440497e-01 -3.22542340e-01 -3.18021956e-03 -9.23623323e-01
-6.96633518e-01 5.46305120e-01 1.01938367e-01 3.64135742e-01
1.17325974e+00 -7.67893670e-03 -9.34036970e-02 7.44355738e-01
4.60984588e-01 -1.13328373e+00 -3.39031607e-01 3.19531322e-01
6.98686600e-01 -1.63287866e+00 3.71969551e-01 -1.02474487e+00
-4.12974030e-01 1.47082126e+00 8.48730206e-01 3.34698379e-01
1.09087098e+00 3.64449829e-01 -2.92362303e-01 -2.85510182e-01
-1.35330904e+00 4.58884006e-03 7.90196836e-01 6.13323629e-01
7.39319801e-01 2.68903792e-01 -7.16459036e-01 1.25314736e+00
-3.67243201e-01 5.43672562e-01 4.00852799e-01 8.28781307e-01
-6.26846492e-01 -1.38757002e+00 -5.53738475e-01 1.16088510e+00
-4.99155134e-01 7.00739026e-02 -8.32967222e-01 1.00161564e+00
4.50487673e-01 1.18690681e+00 -3.07469845e-01 -6.65071547e-01
4.19579923e-01 4.36205834e-01 5.38234293e-01 -7.55836010e-01
-5.29839635e-01 -3.08985829e-01 5.93943596e-01 -3.77282441e-01
-5.98896742e-01 -7.75406003e-01 -1.17143273e+00 -3.36657584e-01
-3.25549394e-01 4.40657884e-01 1.37407094e-01 1.33432686e+00
9.92389396e-02 3.88904631e-01 4.31671381e-01 1.60873234e-01
-2.98657566e-01 -9.96799946e-01 -4.28813100e-01 1.60979912e-01
-8.68704915e-02 -7.98472583e-01 -2.57137507e-01 2.75312096e-01] | [8.903375625610352, 6.76384162902832] |
8b7d488d-b44f-4dc8-9d62-b4cd40f7bdbd | sequence-to-sequence-pre-training-with | 2305.10448 | null | https://arxiv.org/abs/2305.10448v1 | https://arxiv.org/pdf/2305.10448v1.pdf | Sequence-to-Sequence Pre-training with Unified Modality Masking for Visual Document Understanding | This paper presents GenDoc, a general sequence-to-sequence document understanding model pre-trained with unified masking across three modalities: text, image, and layout. The proposed model utilizes an encoder-decoder architecture, which allows for increased adaptability to a wide range of downstream tasks with diverse output formats, in contrast to the encoder-only models commonly employed in document understanding. In addition to the traditional text infilling task used in previous encoder-decoder models, our pre-training extends to include tasks of masked image token prediction and masked layout prediction. We also design modality-specific instruction and adopt both disentangled attention and the mixture-of-modality-experts strategy to effectively capture the information leveraged by each modality. Evaluation of the proposed model through extensive experiments on several downstream tasks in document understanding demonstrates its ability to achieve superior or competitive performance compared to state-of-the-art approaches. Our analysis further suggests that GenDoc is more robust than the encoder-only models in scenarios where the OCR quality is imperfect. | ['Xiaoran Jin', 'Trung Quoc Luong', 'Zhanming Jie', 'Tianyang Zhan', 'Shuwei Feng'] | 2023-05-16 | null | null | null | null | ['optical-character-recognition', 'text-infilling'] | ['computer-vision', 'natural-language-processing'] | [ 9.46695089e-01 1.27790451e-01 -1.13166407e-01 -3.13270926e-01
-1.16653264e+00 -7.16194510e-01 1.02220750e+00 -2.42953468e-02
-1.08159550e-01 4.38385636e-01 4.90958184e-01 -6.48453593e-01
2.14588553e-01 -3.29732776e-01 -9.79298234e-01 -4.62482780e-01
4.60721523e-01 3.07418048e-01 -3.21497582e-02 -1.10091977e-01
8.33079159e-01 5.80606312e-02 -1.60522044e+00 9.96193111e-01
1.16698360e+00 8.01447749e-01 8.77112210e-01 1.13922000e+00
-3.63107353e-01 1.12615347e+00 -6.21802926e-01 -4.26940948e-01
-9.85097215e-02 -4.37139690e-01 -7.03806102e-01 3.61337632e-01
6.60427153e-01 -5.60382366e-01 -5.20835340e-01 6.37262046e-01
4.18375790e-01 -2.16380358e-01 8.12464476e-01 -7.52483428e-01
-1.46147001e+00 5.98963439e-01 -6.42509341e-01 -1.55366231e-02
6.52004600e-01 1.36465803e-01 8.52014542e-01 -1.08249700e+00
3.13959718e-01 1.27621377e+00 3.33933979e-01 4.91114885e-01
-8.98375869e-01 -3.97122085e-01 4.73551810e-01 1.90682247e-01
-1.02578723e+00 -5.09794295e-01 4.21585083e-01 -5.07324815e-01
1.29616678e+00 4.73303050e-02 1.14707295e-02 1.30424321e+00
3.79082054e-01 1.29173434e+00 1.28104436e+00 -8.96604657e-01
-2.77554601e-01 2.53584772e-01 3.28774638e-02 8.41686666e-01
-6.06853329e-02 -8.14104974e-02 -1.02070522e+00 3.45595807e-01
5.58904231e-01 -1.54183105e-01 -3.81066501e-01 -1.10769510e-01
-1.16291678e+00 3.62037629e-01 6.55352250e-02 2.38725008e-03
-1.64282136e-02 2.52827317e-01 4.26529586e-01 2.23152548e-01
4.31340307e-01 2.65058666e-01 -2.67538249e-01 -2.42503509e-01
-1.15476692e+00 -1.66698113e-01 6.89468265e-01 1.38997233e+00
4.07983899e-01 1.18877605e-01 -5.13730645e-01 6.89288437e-01
5.28654099e-01 5.09458065e-01 3.75059277e-01 -6.43846273e-01
1.11521280e+00 4.90922451e-01 1.96118355e-02 -3.37719619e-01
1.24814123e-01 -4.70867217e-01 -6.77052915e-01 3.57965492e-02
1.30344316e-01 3.11934143e-01 -1.47016859e+00 1.53569663e+00
-2.87023395e-01 7.19874501e-02 2.71587342e-01 6.86520875e-01
6.95146203e-01 7.31948555e-01 1.24007940e-01 1.65744916e-01
1.49337149e+00 -1.38296771e+00 -9.76837754e-01 -4.49610472e-01
6.40757442e-01 -1.19843209e+00 1.29582751e+00 5.14521182e-01
-1.06293488e+00 -8.46361101e-01 -1.45331740e+00 -6.10162020e-01
-5.70753992e-01 5.01707673e-01 4.23782200e-01 9.84587908e-01
-1.10983324e+00 1.48198679e-01 -4.88701493e-01 -3.70211393e-01
2.98504859e-01 2.35630244e-01 -2.61066020e-01 -5.31733871e-01
-8.68452072e-01 1.04805422e+00 4.24461156e-01 1.87603459e-01
-1.09591949e+00 -6.07034802e-01 -9.04737830e-01 2.67022371e-01
2.34876513e-01 -8.22726488e-01 1.35902822e+00 -1.12714863e+00
-1.65671384e+00 6.90881491e-01 -5.93386948e-01 -1.05020292e-01
3.50981414e-01 -7.71512985e-01 -3.30811650e-01 2.27347225e-01
-7.36691579e-02 7.92568207e-01 1.11695921e+00 -1.68414068e+00
-4.56942350e-01 -2.04063520e-01 1.01484805e-01 6.87343299e-01
-4.60469872e-01 -1.45997584e-01 -8.18917990e-01 -6.51712418e-01
-1.57726139e-01 -6.38340473e-01 3.20995361e-01 -1.18328847e-01
-7.16888070e-01 2.93976068e-01 1.10444903e+00 -1.01781869e+00
1.30359888e+00 -1.97083223e+00 2.97910511e-01 -1.48526847e-01
-1.01198055e-01 1.68457404e-01 -5.10849357e-01 1.02927244e+00
-2.40889881e-02 1.59093559e-01 -2.48868927e-01 -9.35305774e-01
7.68159479e-02 7.15650544e-02 -4.33355033e-01 1.03267707e-01
3.27419370e-01 8.58020484e-01 -5.10059953e-01 -4.54937220e-01
3.14596176e-01 4.78619993e-01 -3.68622333e-01 4.47453380e-01
-6.11467361e-01 3.84894490e-01 -3.70553404e-01 7.93638349e-01
7.66726017e-01 -4.31318015e-01 2.82999873e-01 -6.82979077e-02
-7.35196993e-02 3.58730823e-01 -8.35444331e-01 2.08826876e+00
-8.37403834e-01 7.95552552e-01 8.82592145e-03 -7.11209357e-01
6.24614477e-01 6.10526025e-01 -3.30277830e-01 -8.31258535e-01
-7.78360441e-02 1.36200726e-01 -9.69924927e-02 -6.00684345e-01
1.00084019e+00 2.07357913e-01 -1.15057431e-01 6.70746386e-01
2.10698411e-01 -1.62743945e-02 6.41493499e-02 5.93564987e-01
9.87032294e-01 6.30845249e-01 1.32421434e-01 -4.12993990e-02
6.57148182e-01 -9.68284607e-02 -1.05901808e-01 1.09466541e+00
1.53717637e-01 7.83721268e-01 4.55265880e-01 3.51347595e-01
-1.28988039e+00 -1.02804589e+00 2.31672272e-01 1.30877507e+00
2.92153388e-01 -2.74163127e-01 -8.39955926e-01 -6.05635107e-01
-1.77825749e-01 8.35633337e-01 -5.61502516e-01 -1.00391014e-02
-3.32325399e-01 -3.68669569e-01 7.34592915e-01 9.01234329e-01
6.30472362e-01 -8.30039144e-01 -8.25582668e-02 -4.09234613e-02
-1.83929592e-01 -1.33702183e+00 -7.12497711e-01 2.07714319e-01
-8.36086929e-01 -9.34219718e-01 -6.38317108e-01 -8.99082363e-01
8.20444226e-01 4.75973159e-01 9.47671354e-01 2.30569005e-01
-1.52434364e-01 7.13698745e-01 -5.28546989e-01 -3.08381408e-01
-6.51012182e-01 1.48203760e-01 -5.33975303e-01 5.25514632e-02
3.38935524e-01 4.31745164e-02 -7.21764505e-01 4.06024083e-02
-1.15007365e+00 5.75851679e-01 1.11241484e+00 1.04805732e+00
1.17360562e-01 -2.87750751e-01 5.41776896e-01 -1.15191114e+00
7.82073021e-01 -3.80476445e-01 -2.07259357e-01 7.58336544e-01
-5.79487741e-01 1.84640780e-01 7.79187858e-01 -2.27293193e-01
-1.76939940e+00 3.44170183e-02 -2.28062551e-02 -8.74212235e-02
-3.37873608e-01 5.39008379e-01 -4.56936449e-01 9.62583572e-02
5.61607182e-02 5.88894546e-01 -1.50057703e-01 -5.82841098e-01
7.14622378e-01 9.74676311e-01 5.83254695e-01 -5.30845046e-01
6.84012294e-01 3.36627156e-01 -3.62069756e-01 -6.52120531e-01
-7.50028133e-01 -4.90964204e-01 -9.43918705e-01 -1.35890827e-01
9.83539820e-01 -1.20182264e+00 -5.30236363e-01 6.72753096e-01
-1.41294622e+00 -2.44468942e-01 2.31812447e-01 1.76005363e-01
-5.29474616e-01 6.10488474e-01 -7.40601599e-01 -9.40927505e-01
-2.66493082e-01 -1.39342403e+00 1.53412747e+00 2.95702457e-01
-1.80174701e-03 -1.03602338e+00 -3.09371769e-01 8.48166347e-01
3.17812115e-01 -2.99621880e-01 1.37522590e+00 -6.04235530e-01
-1.06143868e+00 -1.43105701e-01 -6.60119176e-01 3.40488613e-01
1.61552560e-02 -1.46321608e-02 -1.36649513e+00 -2.58355349e-01
-3.36458027e-01 -5.45654476e-01 1.22171164e+00 1.09591849e-01
1.23126864e+00 6.64653927e-02 -3.75109673e-01 3.44925255e-01
1.63272703e+00 3.17906469e-01 1.02773774e+00 3.34148824e-01
9.44415569e-01 5.01300097e-01 5.44030845e-01 3.01964134e-01
5.80217540e-01 3.66954386e-01 4.15429980e-01 -2.97193974e-01
-3.56357217e-01 -7.00567663e-01 4.45569456e-01 9.88845468e-01
2.04551369e-01 -1.01807320e+00 -6.59420013e-01 5.61925530e-01
-1.78686905e+00 -7.59949923e-01 -8.25612321e-02 1.88604236e+00
5.10046065e-01 9.38757136e-02 -4.95006680e-01 5.11289649e-02
5.54586530e-01 3.78686816e-01 -3.49381834e-01 -9.35648918e-01
-1.95529312e-01 2.44737118e-01 5.25794923e-01 5.49835801e-01
-8.76965582e-01 1.08141863e+00 6.81279564e+00 7.78935075e-01
-7.58183002e-01 1.84575662e-01 5.81412077e-01 2.55696803e-01
-5.35818696e-01 1.90693930e-01 -8.00529718e-01 3.29712540e-01
8.19444120e-01 3.98533195e-01 2.37291008e-01 3.73970717e-01
1.88953966e-01 -3.13092321e-01 -1.31058228e+00 4.58857507e-01
6.35645628e-01 -1.31238532e+00 2.68243611e-01 -1.39072940e-01
8.98187339e-01 -4.92808282e-01 5.03930509e-01 2.46908218e-01
8.21469724e-02 -1.41160131e+00 8.36062312e-01 5.69290161e-01
1.14738619e+00 -4.76129591e-01 5.14972985e-01 3.12863976e-01
-1.14759934e+00 -1.68434903e-01 2.36565955e-02 -2.77834404e-02
3.68926257e-01 7.96874687e-02 -8.40597987e-01 7.54067957e-01
2.24798962e-01 5.48382938e-01 -5.95749855e-01 6.28486812e-01
-3.58226180e-01 5.58565855e-01 2.44178221e-01 4.90510166e-02
3.37618589e-01 1.52673438e-01 1.55424252e-01 1.63730836e+00
3.61100137e-01 -2.77937144e-01 -3.55530046e-02 6.16248131e-01
-1.31955221e-01 -2.76898994e-04 -6.17388844e-01 -3.75368863e-01
5.73969066e-01 8.22889149e-01 -4.17670071e-01 -4.93958980e-01
-8.63397479e-01 1.50196493e+00 1.98801413e-01 6.20726585e-01
-6.88135386e-01 -4.04155701e-01 2.31347948e-01 -1.32556707e-01
6.94449306e-01 -3.36543232e-01 -7.02038109e-01 -1.05327296e+00
3.53385583e-02 -1.22046840e+00 1.57695562e-01 -1.15446174e+00
-9.29014862e-01 5.88200152e-01 -1.93387698e-02 -1.11854303e+00
-1.89019561e-01 -1.02506876e+00 -6.31410837e-01 1.17973959e+00
-1.88440335e+00 -1.82565820e+00 -2.45708078e-01 3.39127988e-01
1.10868621e+00 -2.52188593e-01 8.45850348e-01 5.22025004e-02
-4.11023408e-01 7.28054404e-01 4.27301198e-01 -2.11918920e-01
8.77805829e-01 -1.33394277e+00 4.11889315e-01 1.10670805e+00
6.69631734e-02 7.46758938e-01 6.21161580e-01 -7.30052292e-01
-1.72623289e+00 -7.26015389e-01 7.30912745e-01 -4.60825235e-01
5.25255680e-01 -7.06514657e-01 -8.45420897e-01 8.63021851e-01
1.03068841e+00 -8.29203129e-01 8.09634149e-01 -6.05409406e-02
-6.12006247e-01 2.88244039e-01 -6.90184355e-01 6.94720566e-01
8.74561250e-01 -9.73375976e-01 -5.72804749e-01 2.93425858e-01
6.59895480e-01 -5.67746460e-01 -6.27003491e-01 1.15063258e-01
6.21182919e-01 -8.96576345e-01 7.42494226e-01 -3.95662099e-01
1.02445900e+00 -2.48377115e-01 -2.42500797e-01 -1.09648371e+00
-1.44224152e-01 -5.93467593e-01 -2.97394425e-01 1.30677772e+00
5.85753560e-01 -2.65796095e-01 5.45719981e-01 2.54773557e-01
-5.12335360e-01 -5.59193611e-01 -6.30417883e-01 -4.73225683e-01
1.30696176e-03 -4.34934229e-01 3.62647176e-01 3.55205804e-01
2.11555827e-02 4.51854944e-01 -7.54731476e-01 3.60681802e-01
4.92574513e-01 1.33185655e-01 7.52032876e-01 -5.18167973e-01
-6.17095053e-01 -1.99140742e-01 1.21160559e-01 -1.83460450e+00
1.86103567e-01 -8.01804960e-01 2.43048400e-01 -1.73649180e+00
6.02305412e-01 2.54036635e-02 -8.56024548e-02 2.59286314e-01
-3.79923791e-01 1.62234500e-01 3.67542386e-01 4.87850569e-02
-5.84139705e-01 5.46623290e-01 1.47288048e+00 -2.41979092e-01
1.85182095e-01 -4.63980108e-01 -8.77168834e-01 3.37694287e-01
3.27776551e-01 -5.45073487e-02 -6.74457669e-01 -9.83458579e-01
-1.81663841e-01 9.48751122e-02 1.96265414e-01 -5.99296629e-01
3.09195191e-01 1.41140833e-01 5.59046209e-01 -8.44602585e-01
6.12121820e-01 -6.54556811e-01 -3.10157180e-01 1.68188110e-01
-6.04380429e-01 1.46799833e-01 4.57495123e-01 7.42052913e-01
-3.34224641e-01 -4.59919155e-01 3.55655164e-01 -5.43995947e-02
-1.04429245e+00 -3.80946308e-01 -6.84981942e-01 -2.82104850e-01
7.07523346e-01 -6.29906714e-01 -8.84830952e-01 -6.39594316e-01
-3.75310332e-01 2.42768437e-01 4.48516339e-01 6.01409793e-01
9.16411698e-01 -8.20539773e-01 -5.44365346e-01 2.09352374e-01
3.43170792e-01 -4.35183138e-01 5.16313255e-01 6.87523961e-01
-5.61586797e-01 8.65061879e-01 -9.90189984e-03 -6.42572403e-01
-1.37378407e+00 5.95722198e-01 -2.59747282e-02 -6.17547393e-01
-4.30844665e-01 8.13209593e-01 5.01220286e-01 -2.15605542e-01
3.97989035e-01 -9.73109156e-02 9.77659076e-02 -2.81738281e-01
5.77348351e-01 6.15949463e-03 -4.52833716e-03 -4.84039009e-01
-1.34058237e-01 4.82570469e-01 -5.00828803e-01 -3.33025068e-01
8.68712783e-01 -5.63686609e-01 1.97291940e-01 4.10030782e-01
1.04657280e+00 -1.53927073e-01 -1.39851558e+00 -2.49785185e-02
-5.01863845e-02 -3.88526499e-01 8.64673685e-03 -1.28391755e+00
-5.28711498e-01 1.29674876e+00 5.41043997e-01 -1.56881005e-01
1.39644873e+00 -3.95850718e-01 6.41138077e-01 2.30348602e-01
-2.39238068e-02 -1.20671821e+00 4.63557154e-01 5.96042037e-01
8.06865335e-01 -1.33685005e+00 6.16322979e-02 -4.73124623e-01
-9.74062026e-01 1.07671559e+00 8.46929729e-01 2.10855201e-01
1.36252150e-01 5.73690832e-01 1.54646799e-01 1.08891353e-01
-1.12960172e+00 -1.98080704e-01 4.96009856e-01 6.76449120e-01
8.59588325e-01 -3.08990806e-01 1.15802335e-02 2.42119059e-01
2.63765663e-01 -2.04441044e-02 5.48097789e-01 1.19265389e+00
-3.22016150e-01 -1.24417865e+00 -5.73792696e-01 3.86114568e-01
-3.46512437e-01 -5.08966446e-01 -4.27205861e-01 9.65068460e-01
-7.86552280e-02 1.00539029e+00 1.09549433e-01 -1.89607456e-01
-1.45433880e-02 3.71756852e-01 7.63684273e-01 -6.70028567e-01
-6.55374289e-01 1.46937683e-01 2.54952401e-01 -2.32615873e-01
-3.68297219e-01 -3.13559026e-01 -9.31451619e-01 -3.03987172e-02
-5.65025508e-01 -3.10968846e-01 7.94983208e-01 1.19758022e+00
4.87641484e-01 9.58172143e-01 2.54793346e-01 -7.29422033e-01
-3.77269506e-01 -1.20997298e+00 -3.49133372e-01 1.89966872e-01
3.76602530e-01 -4.59093541e-01 2.10710801e-02 4.59964395e-01] | [11.522623062133789, 2.2159957885742188] |
91fa279d-bf06-468e-909c-212495b796ab | transfer-learning-for-non-intrusive-load | 1902.08835 | null | https://arxiv.org/abs/1902.08835v3 | https://arxiv.org/pdf/1902.08835v3.pdf | Transfer Learning for Non-Intrusive Load Monitoring | Non-intrusive load monitoring (NILM) is a technique to recover source appliances from only the recorded mains in a household. NILM is unidentifiable and thus a challenge problem because the inferred power value of an appliance given only the mains could not be unique. To mitigate the unidentifiable problem, various methods incorporating domain knowledge into NILM have been proposed and shown effective experimentally. Recently, among these methods, deep neural networks are shown performing best. Arguably, the recently proposed sequence-to-point (seq2point) learning is promising for NILM. However, the results were only carried out on the same data domain. It is not clear if the method could be generalised or transferred to different domains, e.g., the test data were drawn from a different country comparing to the training data. We address this issue in the paper, and two transfer learning schemes are proposed, i.e., appliance transfer learning (ATL) and cross-domain transfer learning (CTL). For ATL, our results show that the latent features learnt by a `complex' appliance, e.g., washing machine, can be transferred to a `simple' appliance, e.g., kettle. For CTL, our conclusion is that the seq2point learning is transferable. Precisely, when the training and test data are in a similar domain, seq2point learning can be directly applied to the test data without fine tuning; when the training and test data are in different domains, seq2point learning needs fine tuning before applying to the test data. Interestingly, we show that only the fully connected layers need fine tuning for transfer learning. Source code can be found at https://github.com/MingjunZhong/transferNILM. | ['Michele DIncecco', 'Mingjun Zhong', 'Stefano Squartini'] | 2019-02-23 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [ 1.45508870e-01 1.48064151e-01 -4.11957979e-01 -4.59685534e-01
-7.92647362e-01 -6.66077316e-01 2.68877774e-01 -1.60999656e-01
8.63434300e-02 1.06605196e+00 -2.07322195e-01 -2.86657989e-01
-2.79720873e-01 -8.90820026e-01 -9.95173752e-01 -8.45827818e-01
9.97398719e-02 4.74886417e-01 -1.19535059e-01 -4.04727906e-02
-2.26568475e-01 2.37134933e-01 -1.22361827e+00 2.60920942e-01
8.29835296e-01 1.03287780e+00 3.22296530e-01 4.26048309e-01
2.09763944e-01 7.78746247e-01 -9.29166317e-01 -1.66023329e-01
1.37282386e-01 -5.82843482e-01 -9.23449099e-01 -1.88233536e-02
9.01199207e-02 -5.39031625e-01 9.16745067e-02 1.05569720e+00
3.53015929e-01 -1.91957913e-02 6.77494466e-01 -1.74305224e+00
-6.46003842e-01 6.21009052e-01 -4.60969925e-01 -3.46482068e-01
5.66356361e-01 -3.55316326e-02 6.64782345e-01 -1.46042392e-01
3.92286539e-01 9.56613421e-01 7.89477527e-01 4.27288860e-01
-1.39403927e+00 -1.07326984e+00 -5.04456796e-02 4.03815001e-01
-1.18741179e+00 -1.46546170e-01 1.04927230e+00 -2.52267689e-01
8.53694081e-01 2.56948084e-01 3.01566929e-01 1.62487328e+00
3.28740850e-02 9.94036019e-01 1.26386237e+00 -4.13796455e-01
4.46070403e-01 4.98807788e-01 -2.67878454e-03 2.99420774e-01
3.26956332e-01 -3.25588770e-02 -2.54662395e-01 -2.67544627e-01
4.35418278e-01 -1.27147017e-02 -3.39289337e-01 -2.56561249e-01
-7.86197841e-01 8.78745735e-01 2.91909516e-01 6.16847813e-01
-3.68161500e-01 -1.71013847e-01 5.46889901e-01 5.13263822e-01
3.62444043e-01 2.85993278e-01 -9.74596322e-01 -1.54945731e-01
-8.91661108e-01 -4.96122465e-02 1.07910013e+00 1.03853405e+00
9.75500762e-01 -6.78504705e-02 2.05770642e-01 6.26293242e-01
3.31990011e-02 4.03321892e-01 6.35004580e-01 -8.53363454e-01
5.50110221e-01 5.41813850e-01 1.30902901e-01 -6.83815539e-01
-5.87533653e-01 2.70196348e-01 -9.92787600e-01 1.54468417e-01
5.13998032e-01 -4.63682294e-01 -5.12238562e-01 1.95492327e+00
3.08980793e-01 1.86674044e-01 1.27868414e-01 6.29135251e-01
5.96089423e-01 7.91701376e-01 1.32729173e-01 -1.55524373e-01
1.20451677e+00 -6.57497883e-01 -5.88225842e-01 9.46183652e-02
7.39406526e-01 -4.40921694e-01 1.02261758e+00 5.55289447e-01
-7.23581374e-01 -6.80882990e-01 -1.05227995e+00 6.77102283e-02
-6.32090807e-01 6.59235939e-02 4.64510947e-01 5.33231497e-01
-6.70863330e-01 8.39608073e-01 -8.74373436e-01 -7.92035341e-01
4.29737121e-01 5.70371330e-01 -4.71192181e-01 -1.23203568e-01
-1.30482543e+00 8.37679744e-01 7.60602176e-01 -2.04821616e-01
-6.24854684e-01 -5.98093271e-01 -7.15448976e-01 1.10046580e-01
3.46623302e-01 -3.26587349e-01 1.23668504e+00 -1.08929980e+00
-1.72786689e+00 7.08360791e-01 1.87692940e-01 -3.91373128e-01
6.00153387e-01 -1.44762307e-01 -5.56832910e-01 -4.22477722e-02
1.69175699e-01 3.18410546e-01 7.83411026e-01 -1.23516309e+00
-6.56259477e-01 -2.97519684e-01 7.68163800e-02 -4.09898102e-01
-4.05718952e-01 -2.28067324e-01 -1.30663663e-02 -3.99629682e-01
-4.32329834e-01 -1.05021369e+00 5.87002158e-01 -2.22114876e-01
-5.89180470e-01 -7.03493714e-01 1.04738808e+00 -8.50627959e-01
8.27930868e-01 -2.31895924e+00 -5.50229438e-02 1.50740683e-01
-3.88421118e-01 4.42445576e-01 -1.47983775e-01 4.61337715e-01
-3.59155744e-01 -4.26350720e-02 -2.64330447e-01 -1.47560656e-01
3.69067281e-01 3.05441082e-01 -1.46860197e-01 5.11904716e-01
1.59431890e-01 9.70619619e-01 -9.44527805e-01 -2.86155552e-01
4.59982187e-01 4.80764568e-01 -1.86962098e-01 3.96015197e-01
-3.07731986e-01 6.98148906e-01 -4.20815319e-01 5.36779881e-01
7.57131696e-01 -2.71371156e-01 5.13196766e-01 -4.53725457e-01
3.27009439e-01 3.23341608e-01 -1.03255904e+00 1.73364651e+00
-7.70639837e-01 5.28663635e-01 -1.13063782e-01 -1.25811291e+00
9.16524112e-01 6.16665781e-01 6.50025010e-01 -7.93232381e-01
1.42095163e-01 1.51119798e-01 -1.82447329e-01 -6.11206412e-01
4.34833616e-02 -1.35270804e-01 -2.89343745e-01 4.73934859e-01
1.75100893e-01 3.02557517e-02 -7.41130635e-02 -4.91458356e-01
1.12262762e+00 4.86579955e-01 4.49483812e-01 -1.98722348e-01
3.17013323e-01 -1.61690861e-01 6.79968596e-01 4.36955065e-01
-1.35344565e-01 2.31013522e-01 4.92896855e-01 -2.02439278e-01
-8.35010409e-01 -1.11450565e+00 -1.11356586e-01 1.08718669e+00
-1.02681965e-01 -2.34961346e-01 -8.23275924e-01 -8.75928938e-01
1.49235964e-01 1.05599308e+00 -5.17531335e-01 -3.17664385e-01
-5.13065636e-01 -5.76676846e-01 4.78003472e-01 6.72482431e-01
8.39221895e-01 -1.34832203e+00 -6.95969641e-01 2.53933370e-01
-3.29041779e-01 -9.84923840e-01 -2.58732319e-01 6.49374008e-01
-7.20647573e-01 -1.21249020e+00 -5.27287006e-01 -7.36437678e-01
4.00320768e-01 -5.50974831e-02 1.12785423e+00 -2.00790793e-01
3.44439261e-02 1.81015506e-01 -3.70993882e-01 -4.95622098e-01
-5.38779140e-01 5.31188190e-01 1.30233988e-01 -9.97877792e-02
9.22648430e-01 -9.48434055e-01 -5.23419321e-01 4.07978505e-01
-7.83491433e-01 -2.44202778e-01 3.45349908e-01 6.78091466e-01
2.30895773e-01 4.87462044e-01 9.13479209e-01 -1.10770607e+00
4.56617892e-01 -7.79593110e-01 -6.62145674e-01 2.84347177e-01
-6.39304578e-01 -1.99560404e-01 9.27916646e-01 -7.75943279e-01
-9.23920572e-01 5.46200499e-02 -1.43297702e-01 -5.17607868e-01
-6.26061559e-01 1.21431120e-01 -5.54673433e-01 5.88766336e-01
4.12796557e-01 2.62259971e-02 -2.69295096e-01 -7.16796696e-01
6.92175329e-02 8.93227458e-01 4.52298045e-01 -7.03375340e-01
7.21978724e-01 2.63200462e-01 -2.17149749e-01 -5.62630236e-01
-6.32374406e-01 -3.35793197e-01 -6.94816947e-01 1.40636474e-01
9.87061024e-01 -7.57675052e-01 -1.01744747e+00 5.56361318e-01
-1.03584182e+00 -8.41115236e-01 -4.63226974e-01 3.93048257e-01
-8.06601405e-01 2.35505328e-02 -5.93000114e-01 -7.50329971e-01
-2.23342165e-01 -9.66802061e-01 1.03513825e+00 1.90705791e-01
-7.58039713e-01 -1.25801563e+00 -1.11883648e-01 2.54525185e-01
2.12487221e-01 5.13467789e-01 1.28538001e+00 -9.31250453e-01
-1.23169214e-01 -1.90552294e-01 -9.97593254e-02 6.34900331e-01
8.10352206e-01 -3.85926872e-01 -1.28479445e+00 -6.77536547e-01
3.19026381e-01 -4.78967279e-01 3.04438025e-01 2.10917339e-01
1.24888790e+00 -4.52757210e-01 -3.36964548e-01 4.34049457e-01
1.39643157e+00 4.52063680e-01 7.46992409e-01 4.33199227e-01
6.18086696e-01 5.82236767e-01 4.85147446e-01 2.40376681e-01
3.91111016e-01 7.30524600e-01 1.60272881e-01 -2.13009223e-01
4.83571291e-02 -5.37045956e-01 5.51591814e-01 6.90313518e-01
1.72927588e-01 -3.27553153e-01 -4.35153693e-01 4.38770413e-01
-1.87972319e+00 -7.96163499e-01 1.78619757e-01 2.09284139e+00
8.59369099e-01 4.64864708e-02 3.83030266e-01 5.49037337e-01
6.23784840e-01 -1.96254492e-01 -9.77963150e-01 -4.95370686e-01
2.68631816e-01 3.06721866e-01 3.16884160e-01 2.18583152e-01
-1.03825629e+00 3.75669241e-01 5.01136684e+00 8.58179331e-01
-1.18102467e+00 3.26144487e-01 4.44338262e-01 1.10989764e-01
1.62581861e-01 -3.32863778e-01 -6.07562542e-01 8.92015934e-01
1.22423065e+00 6.72504604e-02 6.39744639e-01 1.01693118e+00
1.25387579e-01 -2.15319917e-01 -1.70687151e+00 1.05825281e+00
-2.13619769e-02 -5.88944793e-01 -5.12278140e-01 7.86421373e-02
5.79449654e-01 -1.34354591e-01 -1.02676928e-01 6.60420656e-01
3.58709127e-01 -7.89953470e-01 3.70239019e-01 3.22724968e-01
7.88616300e-01 -7.69756436e-01 8.03371549e-01 5.64264417e-01
-9.64297414e-01 -8.19871724e-02 -1.90403774e-01 1.41991839e-01
-7.07567707e-02 5.12507737e-01 -8.62724960e-01 7.84834266e-01
9.82780635e-01 5.18329799e-01 -3.04465979e-01 4.17945206e-01
-3.10721427e-01 7.94322848e-01 -3.53891581e-01 2.40905300e-01
-2.04930499e-01 -2.96504974e-01 9.49686170e-02 9.74535286e-01
4.83525515e-01 -2.33196452e-01 9.92619768e-02 9.99937236e-01
-1.39812320e-01 -2.41988450e-01 -7.96514928e-01 1.84494883e-01
2.30210721e-01 1.22287381e+00 -4.31450158e-01 -2.92226642e-01
-5.34854412e-01 1.30812562e+00 1.77172929e-01 3.71669769e-01
-8.90177429e-01 -4.06616300e-01 6.02175772e-01 5.98751456e-02
4.71955597e-01 2.06022471e-01 -1.11641861e-01 -9.81469929e-01
1.07637364e-02 -7.55636752e-01 4.27895725e-01 -7.59229898e-01
-1.81964695e+00 3.29285175e-01 3.69636089e-01 -1.17037606e+00
-6.65429115e-01 -5.81233978e-01 -6.15874231e-01 8.68359745e-01
-1.53827083e+00 -1.15710342e+00 -2.40914196e-01 8.05194139e-01
4.26249236e-01 -1.86850387e-03 1.14484453e+00 5.01909912e-01
-3.91478240e-01 7.60951042e-01 3.53856117e-01 2.93171406e-01
6.77357495e-01 -1.32277381e+00 4.33067605e-02 1.57012478e-01
-7.36368895e-02 2.44707778e-01 5.23399770e-01 -3.85907084e-01
-1.27302718e+00 -1.42035389e+00 5.87193668e-01 -5.85184038e-01
5.57699740e-01 -6.36384606e-01 -1.23614156e+00 9.77342486e-01
2.62124866e-01 -3.31232429e-01 8.09105992e-01 1.79251939e-01
-2.48041376e-01 -4.25419211e-01 -1.61871850e+00 7.14889392e-02
6.50149584e-01 -6.39784276e-01 -6.04782701e-01 4.96378869e-01
4.17052716e-01 -1.76509529e-01 -1.18335378e+00 3.14082831e-01
5.26496947e-01 -8.67927551e-01 6.02186143e-01 -3.99911821e-01
1.82489872e-01 -2.22074136e-01 -4.55615848e-01 -1.40971076e+00
-2.02008799e-01 -4.42533821e-01 -4.85653967e-01 1.79673672e+00
1.19611949e-01 -9.58571374e-01 5.80722332e-01 6.17356241e-01
3.10312569e-01 -4.41501468e-01 -9.25431192e-01 -1.22140872e+00
3.87694389e-01 -4.06546891e-01 1.16176057e+00 1.05264509e+00
-3.78425140e-03 4.51176852e-01 -3.75177234e-01 1.97074413e-01
2.40269154e-01 3.64994466e-01 7.23570645e-01 -1.30715716e+00
-4.68073487e-01 4.55951467e-02 -1.66006163e-01 -7.03683138e-01
5.84587514e-01 -9.55829978e-01 4.24975343e-02 -1.29805481e+00
1.32169560e-01 -4.65238571e-01 -4.45561022e-01 8.51608455e-01
1.77815080e-01 2.26823926e-01 4.05876517e-01 4.37699184e-02
-1.31284013e-01 2.36244589e-01 9.67382431e-01 -3.85359555e-01
-3.80057842e-01 2.87991285e-01 -5.61005354e-01 7.17615128e-01
1.29842913e+00 -5.08198500e-01 -5.69365859e-01 -1.52359903e-01
-1.01781271e-01 3.65997218e-02 3.81181300e-01 -8.98788154e-01
-1.78918585e-01 6.22302480e-03 5.65199852e-01 -4.12662178e-01
3.66756111e-01 -1.30527151e+00 4.13188815e-01 3.90995711e-01
6.10487908e-02 -1.66271091e-01 3.58334184e-01 3.46917421e-01
1.52056873e-01 -2.96946943e-01 5.92166066e-01 -2.42840908e-02
-4.43060786e-01 -4.48532403e-02 -6.95578381e-02 -2.39174411e-01
1.06776965e+00 -1.57988146e-01 -1.53334245e-01 -3.24535996e-01
-6.39511824e-01 6.22539930e-02 4.46475178e-01 5.40987372e-01
-1.66302212e-02 -1.52143860e+00 -5.37133336e-01 2.45030612e-01
-5.68942130e-02 1.51250556e-01 -3.18994895e-02 6.18630409e-01
1.99377567e-01 4.81360257e-01 -8.96845683e-02 -6.70724452e-01
-9.93850529e-01 9.49088395e-01 1.82062492e-01 -3.05600762e-01
-6.38257980e-01 1.17618613e-01 2.54853159e-01 -6.42013431e-01
1.10712685e-01 -6.61632180e-01 1.34738684e-01 1.63336009e-01
3.42111290e-01 4.89879876e-01 2.83737153e-01 -4.25710112e-01
-3.78819734e-01 5.48764467e-01 4.13566232e-02 4.72319633e-01
1.59680259e+00 2.50697043e-02 1.11275651e-02 9.12032366e-01
1.63100100e+00 -3.61618817e-01 -1.22064698e+00 -3.72748300e-02
6.70565814e-02 -1.30923331e-01 -3.45023185e-01 -9.60708916e-01
-1.18880928e+00 7.29383528e-01 8.75850677e-01 5.56837499e-01
1.54359281e+00 6.04438670e-02 9.62265193e-01 3.35716903e-01
6.53897405e-01 -1.07862175e+00 -2.32813641e-01 2.65555531e-02
7.51904488e-01 -1.34116960e+00 -1.74184322e-01 -1.67581216e-01
-3.50044489e-01 9.67962503e-01 5.09254813e-01 -3.21469121e-02
6.72301292e-01 2.76251793e-01 -8.17133859e-02 8.95428658e-02
-4.25260544e-01 1.67854518e-01 -8.80911499e-02 1.02739120e+00
2.56090909e-01 3.62682879e-01 1.59162253e-01 6.54670596e-01
-2.48627096e-01 5.20412266e-01 9.80532542e-02 8.52790296e-01
3.12686235e-01 -1.18686652e+00 -4.71838236e-01 5.76478183e-01
-2.33617947e-01 4.23674494e-01 -1.07236095e-01 9.31947768e-01
4.74977732e-01 1.15123391e+00 5.34268655e-02 -3.29941720e-01
5.27436972e-01 3.34271818e-01 6.45877481e-01 -4.38338935e-01
-4.06862915e-01 -3.28778088e-01 -1.40170857e-01 -3.54066640e-01
-5.80306470e-01 -6.12559915e-01 -1.13496542e+00 -5.74957192e-01
-3.31513554e-01 6.61135539e-02 4.58382666e-01 9.86952543e-01
2.12400034e-01 5.31246722e-01 8.79333556e-01 -8.84858310e-01
-5.58964968e-01 -1.15349793e+00 -8.72245133e-01 7.04758584e-01
4.70774412e-01 -6.27817869e-01 -2.57339686e-01 2.46160582e-01] | [16.053791046142578, 7.571435451507568] |
ac4343bc-fa5b-4713-869f-14c4e42e8fe1 | symantoresearch-at-semeval-2019-task-3 | null | null | https://aclanthology.org/S19-2057 | https://aclanthology.org/S19-2057.pdf | SymantoResearch at SemEval-2019 Task 3: Combined Neural Models for Emotion Classification in Human-Chatbot Conversations | In this paper, we present our participation to the EmoContext shared task on detecting emotions in English textual conversations between a human and a chatbot. We propose four neural systems and combine them to further improve the results. We show that our neural ensemble systems can successfully distinguish three emotions (SAD, HAPPY, and ANGRY) and separate them from the rest (OTHERS) in a highly-imbalanced scenario. Our best system achieved a 0.77 F1-score and was ranked fourth out of 165 submissions. | ['Sanja {\\v{S}}tajner', 'Marc Franco-Salvador', 'Neha Pawar', 'Angelo Basile', 'Mara Chinea Rios', 'Yassine Benajiba'] | 2019-06-01 | null | null | null | semeval-2019-6 | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [-2.41687194e-01 1.71333164e-01 3.62704903e-01 -7.47637212e-01
-4.01429325e-01 -2.03601107e-01 5.00617325e-01 -1.96649164e-01
-5.30016005e-01 9.61167276e-01 1.09670423e-01 1.25812292e-01
4.79866505e-01 -2.55211890e-01 1.72287330e-01 -4.37517107e-01
8.04523230e-02 4.89079297e-01 -3.81373376e-01 -7.17075467e-01
8.85363445e-02 -2.01614797e-02 -1.42120361e+00 8.63340497e-01
5.76056659e-01 1.02135050e+00 -5.92723787e-01 1.08064938e+00
-3.68635394e-02 1.56204581e+00 -1.23008025e+00 -1.02435958e+00
-1.83427781e-01 -6.91967845e-01 -1.25645506e+00 -2.52121508e-01
3.00657321e-02 -1.15864687e-01 5.92746139e-02 5.38787186e-01
7.93795466e-01 3.45465332e-01 5.17246485e-01 -1.75685561e+00
-3.77434224e-01 5.55137038e-01 -2.91904420e-01 1.47075251e-01
5.66198349e-01 2.22343709e-02 1.19338036e+00 -9.37301695e-01
6.89917266e-01 1.14627659e+00 8.73998582e-01 1.01329565e+00
-7.34950066e-01 -6.46845698e-01 -2.72254217e-02 1.76263317e-01
-8.09934318e-01 -8.58362377e-01 8.78063977e-01 -2.19219521e-01
1.63223791e+00 4.78813618e-01 3.82287890e-01 1.52814162e+00
-1.66022908e-02 9.49379861e-01 1.36758435e+00 -1.71983048e-01
1.63615301e-01 5.71828187e-01 3.75691354e-01 3.94567579e-01
-5.42844057e-01 -5.75486898e-01 -7.58817613e-01 -4.07347769e-01
-9.18950289e-02 -4.57732946e-01 -1.10842638e-01 5.59979737e-01
-8.61355960e-01 9.17349517e-01 3.38096283e-02 4.85777944e-01
-4.94152397e-01 -1.86214671e-01 7.37697482e-01 6.35972679e-01
8.90692472e-01 8.44233930e-01 -6.72801256e-01 -7.94757426e-01
-3.56956542e-01 2.02482745e-01 1.48569441e+00 3.30124766e-01
1.93526626e-01 -5.81636243e-02 -2.16450125e-01 1.20981741e+00
-2.54054934e-01 5.72662707e-03 4.24090803e-01 -1.10361922e+00
1.80583045e-01 4.02465969e-01 3.02906126e-01 -1.06751561e+00
-7.48609543e-01 2.60740761e-02 -7.04593778e-01 3.18021094e-03
2.79247403e-01 -9.50097322e-01 -2.31531993e-01 1.62208545e+00
9.85096991e-02 -2.55058885e-01 2.44858846e-01 8.40425313e-01
1.57976544e+00 4.71506864e-01 1.94721475e-01 -2.45336547e-01
1.38410676e+00 -1.17385173e+00 -1.16421008e+00 -4.40868974e-01
6.78975999e-01 -8.79299641e-01 7.89322317e-01 5.47749102e-01
-1.26141608e+00 -3.23702753e-01 -8.37759733e-01 -5.29726557e-02
-5.30379653e-01 2.45824054e-01 7.43559062e-01 7.47063041e-01
-1.14060795e+00 5.74318647e-01 -2.74682105e-01 -2.22256064e-01
9.25544575e-02 4.05654520e-01 -3.74201447e-01 6.92147493e-01
-1.57927334e+00 1.33951557e+00 -2.86375850e-01 2.30619669e-01
3.96559574e-02 -2.06694648e-01 -6.42235041e-01 -1.09254591e-01
3.33030261e-02 -3.22925687e-01 1.47678602e+00 -1.46070468e+00
-2.02376819e+00 1.41689575e+00 -3.33264887e-01 -2.15037629e-01
5.68375826e-01 -1.93511382e-01 -5.43223143e-01 -2.21770987e-01
-2.93168426e-01 5.59932828e-01 4.56929296e-01 -8.38571608e-01
-2.98713416e-01 1.09329103e-02 8.40633363e-02 2.37743929e-01
-3.09701234e-01 8.13876808e-01 2.40742907e-01 -1.25767991e-01
-5.29628217e-01 -9.31008101e-01 -5.72455190e-02 -7.18148649e-01
-4.78293151e-01 -6.58287287e-01 5.82984984e-01 -6.71214223e-01
8.85649025e-01 -1.94101107e+00 7.16620460e-02 -2.12030083e-01
5.00329256e-01 3.13340545e-01 -3.25873867e-02 3.57262760e-01
-3.57272059e-01 2.73234159e-01 1.80455282e-01 -8.81340444e-01
1.98758811e-01 -2.78657917e-02 5.15862256e-02 1.60672829e-01
5.79469025e-01 8.60895157e-01 -6.62413955e-01 -3.61899227e-01
-1.60093990e-03 5.78702867e-01 -2.33469754e-01 7.27206349e-01
2.05699906e-01 2.83485681e-01 7.36165512e-03 3.79651457e-01
5.15984893e-01 -1.20895147e-01 2.53272623e-01 3.29070054e-02
1.32283298e-02 6.49822533e-01 -6.64269149e-01 8.48240852e-01
-6.49597526e-01 1.06312537e+00 4.53896165e-01 -7.51558125e-01
1.36097991e+00 4.98422027e-01 3.74454319e-01 -4.14392203e-01
7.60165572e-01 -1.19140320e-01 3.35535526e-01 -6.24958456e-01
7.98830748e-01 -3.35716903e-01 -5.03994465e-01 8.69845152e-01
1.15895882e-01 -2.66779424e-03 1.36088386e-01 3.42868090e-01
1.07114398e+00 -3.10262561e-01 3.03259999e-01 -1.72300208e-02
5.05996644e-01 -4.75166678e-01 5.78444839e-01 7.07263172e-01
-9.91616547e-01 4.63079602e-01 1.21639359e+00 -5.79179764e-01
-5.05500138e-01 -4.21486676e-01 3.72269392e-01 1.58747649e+00
-2.42890000e-01 -5.32633901e-01 -8.11750889e-01 -7.39926636e-01
-4.75535750e-01 6.38166428e-01 -8.77679288e-01 -1.36807039e-01
-2.07406551e-01 -8.13803077e-01 8.40858221e-01 4.73680407e-01
3.65509868e-01 -1.52178073e+00 -6.18048429e-01 4.71993349e-02
-9.74492252e-01 -1.18935657e+00 7.01440424e-02 5.74659824e-01
-3.93998064e-02 -8.08044314e-01 -2.68693060e-01 -4.42354560e-01
-4.51977588e-02 -2.24778131e-01 1.68330836e+00 3.18361521e-01
-2.31902853e-01 4.43169475e-02 -5.01592875e-01 -8.10087204e-01
-4.25056934e-01 2.74039686e-01 -1.38848573e-02 4.30981256e-02
9.24465358e-01 -2.45913386e-01 -5.19754924e-02 3.17108005e-01
-1.06945291e-01 -2.25128587e-02 -1.42510921e-01 9.39113915e-01
-4.78625715e-01 -4.77813780e-01 9.21346307e-01 -8.08605790e-01
1.16374993e+00 -6.07568622e-01 3.35837811e-01 3.60277630e-02
4.50752378e-02 -4.61600542e-01 5.18710792e-01 -2.00544775e-01
-1.24407911e+00 1.14815282e-02 -5.29954970e-01 9.83331650e-02
-4.88749683e-01 1.33513689e-01 8.15337058e-03 2.04626188e-01
5.42119861e-01 -6.09091401e-01 5.28637553e-04 -1.65920869e-01
5.01742773e-02 1.27468491e+00 2.95345336e-01 -4.30549473e-01
-2.33262658e-01 3.10550127e-02 -6.20246112e-01 -9.02142048e-01
-9.30443704e-01 -3.97238821e-01 -3.23072165e-01 -6.03018820e-01
8.70387912e-01 -9.18981493e-01 -1.34769821e+00 5.99161923e-01
-1.58307469e+00 -4.01087672e-01 9.43636596e-02 2.03927979e-01
-3.35098863e-01 -9.42736343e-02 -1.15475738e+00 -1.30664635e+00
-6.59717858e-01 -8.09461057e-01 7.80578494e-01 3.42507869e-01
-1.15123808e+00 -7.83847749e-01 4.04213101e-01 6.69207633e-01
6.43951833e-01 2.19105348e-01 1.57951683e-01 -1.13299942e+00
7.27564275e-01 -2.36154646e-01 -9.17463750e-02 3.94484162e-01
6.09035138e-03 4.40243781e-01 -1.38222945e+00 2.63798565e-01
1.81373432e-01 -1.10106158e+00 7.24351227e-01 -3.11922487e-02
8.80498707e-01 1.18557168e-02 1.56387582e-01 -2.14358009e-02
4.45339590e-01 1.72529131e-01 7.44093955e-01 1.56805918e-01
1.95095941e-01 1.03658378e+00 3.93593013e-01 6.85938597e-01
4.19011444e-01 4.26721990e-01 2.57759392e-01 -3.26144516e-01
5.44370174e-01 4.57246423e-01 4.43399698e-01 7.57156968e-01
-3.07392120e-01 -6.07576191e-01 -8.00607443e-01 4.70634997e-01
-1.84747887e+00 -1.04646695e+00 -4.25246984e-01 1.20793355e+00
7.67714560e-01 -5.28649651e-02 5.35077989e-01 1.89825982e-01
8.98871601e-01 3.52278948e-01 -2.24558905e-01 -1.62039340e+00
-3.06849599e-01 4.17858124e-01 -3.16451579e-01 5.04510462e-01
-1.35259092e+00 9.44275439e-01 7.02313662e+00 2.32716158e-01
-9.52058494e-01 1.11511633e-01 1.30772185e+00 -4.54804063e-01
2.09906146e-01 -6.67668164e-01 -2.99482942e-01 2.67151326e-01
1.36138797e+00 -5.17352894e-02 3.05709481e-01 9.52847123e-01
-1.37421712e-01 -8.51581395e-02 -9.51230705e-01 9.59128380e-01
3.92075092e-01 -7.45277166e-01 -8.34949672e-01 -3.14837426e-01
4.76832002e-01 1.89920321e-01 -1.68019146e-01 8.14009964e-01
5.32779336e-01 -1.38810718e+00 2.29239643e-01 3.88662755e-01
1.61115214e-01 -9.01914537e-01 1.34256577e+00 2.37937823e-01
-3.38271499e-01 2.44661361e-01 -3.63319330e-02 -9.49367046e-01
6.24738410e-02 4.90960985e-01 -7.76948392e-01 2.36718386e-01
9.83121097e-01 7.35177696e-01 -2.78006643e-01 2.76162505e-01
-2.31335104e-01 5.46129107e-01 -1.71798646e-01 -7.69142210e-01
5.32566383e-02 1.27993733e-01 3.60228151e-01 1.65877807e+00
-3.67678612e-01 4.15756762e-01 1.05862217e-02 6.72049403e-01
-3.91210824e-01 3.23010713e-01 -3.89243573e-01 1.71451773e-02
1.77439284e-02 1.89256752e+00 -6.44604146e-01 -5.07104695e-01
-2.81924587e-02 1.14104009e+00 6.68864965e-01 -1.10355671e-03
-1.00817466e+00 -8.03772867e-01 9.22489047e-01 -9.00679529e-01
3.27702984e-02 4.37504411e-01 -4.93324161e-01 -1.24854958e+00
-2.58917771e-02 -1.06296813e+00 3.51717710e-01 -1.11492717e+00
-1.85448396e+00 1.18845809e+00 -7.33509898e-01 -6.22889698e-01
-2.75597245e-01 -6.93507731e-01 -8.87351930e-01 8.51173997e-01
-7.91031480e-01 -5.96491694e-01 -4.83517498e-01 3.63620818e-01
1.92069277e-01 -5.64019606e-02 1.21486986e+00 3.57715487e-01
-8.34479928e-01 6.69514775e-01 -4.81162161e-01 5.67083299e-01
1.30943680e+00 -1.30974436e+00 3.03279161e-01 1.02407798e-01
-2.34335005e-01 4.16256994e-01 7.86271214e-01 -2.00771466e-01
-6.77851021e-01 -5.61034918e-01 1.47437978e+00 -6.67594254e-01
6.67131841e-01 -5.01239717e-01 -7.99495220e-01 5.57501137e-01
8.38498950e-01 -3.74877036e-01 1.18206596e+00 7.52390742e-01
-4.41487342e-01 3.94634128e-01 -1.36325645e+00 3.75114620e-01
5.95803082e-01 -4.95847672e-01 -6.36648715e-01 1.95572317e-01
2.51361340e-01 -5.05852759e-01 -7.99558938e-01 2.14205310e-01
7.90741384e-01 -1.57781279e+00 5.49036384e-01 -1.29370439e+00
1.02037156e+00 4.03104722e-01 1.04854487e-01 -1.60661721e+00
-5.62429428e-02 -6.98935151e-01 1.30320609e-01 1.20789158e+00
4.11591709e-01 -6.83586538e-01 7.01214612e-01 1.10782135e+00
1.37987211e-01 -7.82910645e-01 -7.67475069e-01 -3.38133603e-01
1.42714068e-01 -3.73520911e-01 4.09118056e-01 1.34974885e+00
8.63430321e-01 9.76938665e-01 -9.01486337e-01 -7.43227005e-01
-1.60742283e-01 6.55111019e-03 8.47016811e-01 -1.23413956e+00
-1.05625868e-01 -7.63165116e-01 -2.15507448e-01 -3.77207160e-01
6.81917310e-01 -4.89477783e-01 4.13426399e-01 -1.13123941e+00
3.15214902e-01 1.28095262e-02 -4.11576003e-01 7.47654676e-01
-4.35674131e-01 6.26712501e-01 2.52812475e-01 -4.75287795e-01
-9.73394752e-01 4.09445703e-01 5.25616884e-01 -6.02966957e-02
-1.61423042e-01 -1.63255557e-01 -9.95099008e-01 8.17036927e-01
1.22306383e+00 -2.81805992e-01 2.78048158e-01 -2.85627414e-02
4.07049745e-01 1.92754380e-02 1.78584322e-01 -4.87603247e-01
-1.89508960e-01 -1.87073778e-02 2.59835780e-01 -3.56950223e-01
6.77765667e-01 -2.60607213e-01 -2.51233280e-01 1.84576467e-01
-7.53268600e-01 5.96713014e-02 2.07661510e-01 -2.58479178e-01
-2.35907018e-01 -7.87190422e-02 6.43418074e-01 -4.17992175e-02
-1.77033663e-01 -5.02495110e-01 -7.95892775e-01 2.75431991e-01
7.57123113e-01 4.12375391e-01 -5.87943316e-01 -1.10355163e+00
-8.91822577e-01 2.22635269e-01 -4.60931771e-02 6.99129224e-01
3.20602566e-01 -9.28168893e-01 -1.05788112e+00 -1.86478019e-01
-9.74312574e-02 -8.88444185e-01 4.10248697e-01 1.05124617e+00
-1.86602369e-01 -4.24952246e-03 -3.98102015e-01 -3.52361709e-01
-1.86752987e+00 -1.52370632e-01 6.79969728e-01 -4.78518546e-01
7.45741948e-02 1.21430123e+00 -3.40025216e-01 -9.33310866e-01
1.85801193e-01 3.86574656e-01 -4.38671589e-01 3.81647170e-01
7.42601156e-01 6.22231781e-01 3.07148248e-01 -8.91173899e-01
-6.74291968e-01 -1.67014092e-01 4.17874195e-02 -8.72290507e-02
1.28238404e+00 9.95091051e-02 -6.14383876e-01 9.13402617e-01
1.33374333e+00 1.32559007e-02 -3.67521137e-01 1.69331506e-01
-1.90010756e-01 -1.19449250e-01 -1.45861849e-01 -1.19586849e+00
-8.97564828e-01 9.19893265e-01 1.42049804e-01 7.55560935e-01
8.30340743e-01 -1.73485935e-01 9.08480108e-01 6.96297646e-01
-2.51118898e-01 -1.47756898e+00 1.13083705e-01 1.08686578e+00
7.09568739e-01 -1.58408391e+00 -1.08025163e-01 -2.75299162e-01
-1.40945613e+00 1.08739614e+00 1.05773580e+00 -5.32044806e-02
2.21143126e-01 4.98437941e-01 7.89640307e-01 -3.83600086e-01
-1.55624068e+00 -8.37706327e-02 9.28688794e-02 3.28716606e-01
1.11341774e+00 1.56782463e-01 -3.30349773e-01 1.01209116e+00
-4.49684620e-01 -2.43467674e-01 7.24037170e-01 6.94440782e-01
3.68234031e-02 -6.67916656e-01 -1.70841038e-01 4.36276525e-01
-8.90442669e-01 1.37249202e-01 -1.52196276e+00 6.33434594e-01
1.35955149e-02 1.56076634e+00 3.01712036e-01 -8.07674170e-01
3.92475963e-01 5.92564106e-01 1.16187394e-01 -2.09815755e-01
-1.53568554e+00 -2.82484204e-01 1.08133340e+00 -5.11072993e-01
-7.75592089e-01 -6.16144657e-01 -9.34314847e-01 -5.54050088e-01
-2.87004322e-01 4.01821494e-01 6.63609445e-01 9.78046238e-01
3.98650974e-01 4.92127687e-01 9.47234988e-01 -8.78316402e-01
-7.33155757e-02 -1.32025814e+00 -2.10754067e-01 6.44631147e-01
1.76491141e-01 -4.63300288e-01 -8.33774686e-01 -2.97362119e-01] | [13.012948989868164, 6.208339214324951] |
353d021a-66d6-4a22-a899-f973de61b5ad | price-graphs-utilizing-the-structural | 2106.02522 | null | https://arxiv.org/abs/2106.02522v5 | https://arxiv.org/pdf/2106.02522v5.pdf | Price graphs: Utilizing the structural information of financial time series for stock prediction | Great research efforts have been devoted to exploiting deep neural networks in stock prediction. While long-range dependencies and chaotic property are still two major issues that lower the performance of state-of-the-art deep learning models in forecasting future price trends. In this study, we propose a novel framework to address both issues. Specifically, in terms of transforming time series into complex networks, we convert market price series into graphs. Then, structural information, referring to associations among temporal points and the node weights, is extracted from the mapped graphs to resolve the problems regarding long-range dependencies and the chaotic property. We take graph embeddings to represent the associations among temporal points as the prediction model inputs. Node weights are used as a priori knowledge to enhance the learning of temporal attention. The effectiveness of our proposed framework is validated using real-world stock data, and our approach obtains the best performance among several state-of-the-art benchmarks. Moreover, in the conducted trading simulations, our framework further obtains the highest cumulative profits. Our results supplement the existing applications of complex network methods in the financial realm and provide insightful implications for investment applications regarding decision support in financial markets. | ['Jichang Zhao', 'Shangzhe Li', 'Xueyuan Chen', 'Ke Xu', 'Junran Wu'] | 2021-06-04 | null | null | null | null | ['stock-prediction'] | ['time-series'] | [-6.2124830e-01 -2.8557798e-01 -2.5230974e-01 -5.4326329e-02
3.3536232e-01 -6.0609013e-01 6.9062155e-01 1.7387876e-01
-3.2286352e-01 5.1402473e-01 6.8223141e-02 -4.8573482e-01
-4.8979118e-01 -1.2405214e+00 -5.0130254e-01 -5.9197438e-01
-7.0663637e-01 1.4062546e-01 2.6493075e-01 -6.2793630e-01
4.0456620e-01 5.4696590e-01 -9.8459888e-01 -2.0439601e-01
7.8298563e-01 1.4516922e+00 -2.5571826e-01 1.6416830e-01
-2.9823026e-01 9.9188143e-01 -5.5219555e-01 -6.4674425e-01
5.3156334e-01 -2.2375455e-01 -1.3568826e-01 -2.1311212e-01
-2.7865863e-01 -2.9292336e-01 -6.0090029e-01 1.2270875e+00
2.6237914e-01 2.7172877e-02 5.2221489e-01 -1.2214911e+00
-1.0872132e+00 9.3445700e-01 -7.4963522e-01 8.7278008e-01
-4.4522586e-01 2.4704374e-01 1.3920261e+00 -6.9421053e-01
2.6473692e-01 9.0710664e-01 7.5722343e-01 -5.3370513e-02
-9.3203992e-01 -9.2852563e-01 5.3905320e-01 4.4514793e-01
-8.4447306e-01 -2.9399380e-02 1.4953082e+00 -4.8490751e-01
1.0535873e+00 -2.1140258e-01 1.1683753e+00 7.0407957e-01
6.5429938e-01 4.0939155e-01 8.8380492e-01 1.0709162e-02
7.8111485e-02 -9.9760391e-02 3.9029515e-01 4.7169489e-01
3.2905746e-01 4.3046352e-01 -4.2215317e-01 7.8511201e-02
1.0585237e+00 3.4152511e-01 -1.5832202e-01 -1.1183648e-02
-1.1646303e+00 1.0821755e+00 9.3266112e-01 6.2066847e-01
-8.1856543e-01 2.7240977e-01 4.3365255e-01 5.1874787e-01
7.4488783e-01 6.0281622e-01 -3.6197889e-01 1.1057947e-01
-8.8653773e-01 1.3997607e-01 6.9479644e-01 5.4373145e-01
4.3303567e-01 6.7002690e-01 1.3465022e-01 3.9404947e-01
3.0206457e-01 3.0136502e-01 7.4811625e-01 -2.6756275e-01
5.7602835e-01 7.9363459e-01 -1.0690098e-01 -1.8165127e+00
-7.1706533e-01 -1.0406605e+00 -1.0939509e+00 6.3990273e-02
2.8613538e-01 -3.0310765e-01 -4.7777045e-01 1.4186311e+00
-1.0259946e-01 7.1085256e-01 2.1977304e-01 7.4594444e-01
4.4919971e-01 8.8992018e-01 -1.9603585e-01 -3.3489335e-01
1.2019192e+00 -9.2850113e-01 -8.1263334e-01 -6.1253086e-02
1.3135408e-01 -5.2307665e-01 7.2759706e-01 6.4572692e-02
-8.7785232e-01 -6.3473511e-01 -1.1831084e+00 5.9446877e-01
-6.3366646e-01 -2.0017740e-01 8.4722441e-01 1.7730941e-01
-1.0650654e+00 9.8745310e-01 -8.2384992e-01 1.7956214e-01
2.2564000e-01 2.9687795e-01 2.6129219e-01 7.4057072e-01
-1.7402065e+00 7.7814281e-01 4.4987839e-01 6.3917589e-01
-3.2712787e-01 -6.7805892e-01 -5.2800661e-01 3.5554910e-01
3.3871233e-01 -3.0604145e-01 8.3242476e-01 -7.8661406e-01
-1.3006701e+00 1.0884242e-01 5.3820747e-01 -1.0025300e+00
4.9761575e-01 9.3870893e-02 -9.1721421e-01 7.2590761e-02
-2.1371339e-01 1.3705036e-01 9.2981714e-01 -7.2497416e-01
-4.9017379e-01 -1.4845714e-01 8.2391605e-02 -5.0514042e-02
-7.8163254e-01 -3.8393366e-01 -1.0257945e-01 -1.3279872e+00
-1.2083107e-02 -7.5806528e-01 -1.4930904e-01 -3.6068770e-01
-2.2893332e-01 -4.1018721e-01 7.0821935e-01 -7.2606379e-01
1.6217893e+00 -1.9631948e+00 3.8378563e-02 4.3891647e-01
4.0269208e-01 2.1986489e-01 -1.0203187e-01 5.5098236e-01
-2.2429255e-01 2.1106969e-01 3.5360411e-02 9.4174124e-02
2.5429070e-01 8.6424630e-03 -7.0187563e-01 2.8933376e-01
5.0634432e-01 1.3954929e+00 -7.7322310e-01 -9.4634429e-02
4.9261857e-02 3.9838001e-01 -1.6555883e-01 -5.3463653e-02
-2.6675031e-01 1.5389290e-01 -6.5736943e-01 3.4898561e-01
6.4494121e-01 -4.7991127e-01 1.3179608e-01 -2.7740321e-01
-8.9214504e-02 2.5042146e-01 -9.1628343e-01 8.6405563e-01
-1.8719345e-01 4.5373860e-01 -5.2672493e-01 -1.2811421e+00
1.2790676e+00 1.1313040e-02 4.5254627e-01 -1.0540835e+00
9.4984941e-02 2.3811550e-01 4.4071946e-01 -9.1473281e-02
5.0143886e-01 -1.7135806e-01 -4.4916157e-02 5.6660509e-01
-1.4095213e-01 1.9063607e-01 1.9211408e-01 -2.6702706e-02
7.3657936e-01 -2.2241862e-01 1.3250969e-01 -3.4340063e-01
4.4456512e-01 -2.1635289e-01 5.8009762e-01 3.2262188e-01
-2.1198174e-01 -5.5036876e-02 1.0080044e+00 -7.8962725e-01
-9.1099530e-01 -9.1624993e-01 1.4828363e-01 6.1729401e-01
2.0091739e-02 -1.1715984e-02 -3.0496007e-01 -6.0633361e-01
3.9581352e-01 5.9602416e-01 -8.9252388e-01 -2.0208190e-01
-6.8366987e-01 -9.7095048e-01 3.6938637e-01 7.1383119e-01
6.2610543e-01 -1.4094611e+00 -4.7349653e-01 6.0461974e-01
3.4954172e-01 -9.2316931e-01 -4.9721289e-01 1.5034945e-01
-1.0075452e+00 -1.1222470e+00 -9.3193388e-01 -7.9628915e-01
3.0633715e-01 -4.6012536e-02 1.1702266e+00 1.5840608e-01
8.6164147e-02 -1.2859828e-03 -3.1815434e-01 -3.0496135e-01
-9.8436005e-02 6.2805645e-02 3.3529688e-02 2.7721286e-01
3.9464191e-01 -8.8623953e-01 -8.5803211e-01 2.6685521e-01
-9.6309018e-01 -3.5361493e-01 6.4318794e-01 9.3914872e-01
4.7431415e-01 5.3932267e-01 1.1055692e+00 -7.1692163e-01
1.2437561e+00 -7.0192420e-01 -1.1996067e+00 1.9734271e-01
-1.0847259e+00 1.3434836e-01 9.5353109e-01 -5.9136677e-01
-5.3076500e-01 -6.0918176e-01 3.0899936e-01 -5.4055458e-01
7.3233950e-01 1.1611807e+00 4.8436475e-01 -4.9475651e-02
-3.8935576e-02 4.9303746e-01 2.7997810e-02 -2.6100987e-01
1.7904499e-01 -9.7153991e-02 2.9946107e-01 -1.2391876e-01
9.6478897e-01 1.3610071e-01 1.9066165e-01 -4.4098765e-01
-5.0032794e-01 8.2150988e-02 -7.6597613e-01 -1.9677025e-01
6.5194279e-01 -6.2563664e-01 -7.7450126e-01 7.1580523e-01
-9.5898503e-01 -7.8743376e-02 -1.5941884e-01 4.6759158e-01
-2.4863048e-01 2.7260444e-01 -8.6337411e-01 -7.5365019e-01
-4.3582141e-01 -8.7735784e-01 2.3893487e-01 2.6394871e-01
2.4748880e-01 -1.6529706e+00 1.1808459e-01 -2.8307220e-01
6.1053032e-01 6.3590336e-01 1.2821598e+00 -9.1448069e-01
-5.6345612e-01 -2.5617620e-01 -3.2270226e-01 3.1635407e-01
1.9406992e-01 -5.5473164e-02 -4.7769094e-01 -2.2073852e-01
1.3973381e-01 1.3619351e-01 1.0445817e+00 3.0947846e-01
8.2294238e-01 -5.7979864e-01 -1.3184581e-02 5.9685594e-01
1.3509036e+00 6.3713628e-01 4.0237492e-01 5.7519734e-01
5.7570618e-01 5.7933623e-01 4.0434960e-01 6.2425417e-01
4.7400805e-01 2.1969526e-01 4.3098381e-01 8.7172948e-02
3.3888993e-01 -3.5685089e-01 3.2276940e-01 1.4718446e+00
-2.6770240e-01 -2.1039499e-01 -1.0085278e+00 4.6315253e-01
-1.9504811e+00 -1.1480619e+00 2.3344378e-01 1.6865921e+00
5.4157674e-01 6.4346957e-01 2.5523955e-01 1.7789631e-01
6.7870808e-01 7.6138180e-01 -1.0595024e+00 -3.5410840e-02
-2.9507998e-01 1.8859166e-01 5.2580208e-01 1.8678571e-01
-1.1775236e+00 7.7278811e-01 5.6355791e+00 4.8777974e-01
-1.4454031e+00 -2.3728953e-01 7.7437377e-01 2.8195547e-02
-4.6884587e-01 -3.1377918e-01 -5.4204208e-01 6.8307465e-01
9.0733433e-01 -6.5432954e-01 5.9753805e-01 5.9890920e-01
1.7106529e-01 6.5036637e-01 -9.5694244e-01 7.9951549e-01
-1.9723682e-01 -1.6830484e+00 1.3788226e-01 1.5095179e-01
6.3052946e-01 -5.8926597e-02 5.8401465e-01 3.7124839e-01
1.9123818e-01 -9.1019744e-01 7.6286072e-01 8.4403896e-01
1.1876178e-01 -1.1069109e+00 9.7345471e-01 8.3550252e-02
-1.6946379e+00 -3.1448320e-01 -5.1982576e-01 -1.8500403e-01
7.6908879e-02 5.6297976e-01 -5.2610958e-01 7.5598371e-01
5.9151512e-01 1.3028420e+00 -6.8536848e-01 7.1866018e-01
-1.5699241e-01 5.5787021e-01 -1.1247970e-01 -4.7424343e-01
7.1567553e-01 -5.9164351e-01 2.7670696e-01 7.8091484e-01
3.9136431e-01 4.2653721e-02 -1.6553387e-02 1.0101588e+00
-1.8928632e-01 6.0905524e-02 -6.7124540e-01 -5.9903759e-01
4.0625641e-01 1.0736617e+00 -9.8106205e-01 -2.6187885e-01
-5.5593896e-01 4.2797020e-01 3.7357005e-01 4.6524885e-01
-8.4756458e-01 -5.7167959e-01 4.7834268e-01 1.4719556e-01
5.7686472e-01 -3.4381109e-01 -2.0508239e-01 -1.2732170e+00
2.4506058e-01 -6.1630708e-01 5.2159125e-01 -4.0328082e-01
-1.6170105e+00 8.8414127e-01 -2.9505137e-01 -1.6054915e+00
-3.9191350e-01 -7.4976450e-01 -9.6863967e-01 8.6709470e-01
-2.0467184e+00 -7.6556087e-01 6.2371336e-02 6.2270474e-01
2.4408439e-01 -6.4490306e-01 4.1529122e-01 2.3884423e-01
-6.1810172e-01 4.9677002e-01 2.6505688e-01 5.8260703e-01
-3.5789680e-02 -1.0732578e+00 8.8228559e-01 9.2020792e-01
2.8500795e-01 5.3861183e-01 2.8919679e-01 -7.9504210e-01
-1.3695251e+00 -1.1027446e+00 4.4541126e-01 1.5992506e-01
1.6872722e+00 -1.9925103e-01 -1.0848664e+00 6.3729584e-01
4.1620106e-01 1.3628228e-01 3.1269079e-01 -2.1468028e-01
-3.5628316e-01 -3.2501024e-01 -7.6609898e-01 5.2517122e-01
6.5114975e-01 -4.6035466e-01 -6.6641760e-01 6.3912965e-02
9.6204501e-01 1.7046042e-01 -1.1561195e+00 2.9203331e-01
3.5047618e-01 -1.0060508e+00 7.8347510e-01 -6.1148256e-01
4.4497040e-01 -8.7150872e-02 1.7584375e-01 -1.5789807e+00
-4.5122814e-01 -6.2757671e-01 -3.3921435e-01 1.0496628e+00
6.4378405e-01 -1.2595525e+00 7.0062882e-01 2.3517856e-01
1.9816655e-01 -8.8453859e-01 -8.7232846e-01 -8.9532930e-01
2.4291556e-01 -1.4625125e-01 9.8595929e-01 1.2623689e+00
-1.2699042e-02 2.1139307e-01 -3.5200733e-01 2.5342654e-02
5.8165038e-01 5.8329248e-01 2.4403162e-01 -1.4740509e+00
-3.0692151e-01 -9.5503592e-01 -5.0171191e-01 -7.5037038e-01
3.2963586e-01 -8.4010249e-01 -6.3059932e-01 -1.1992551e+00
-4.0661913e-01 -3.8600236e-01 -9.8131400e-01 2.4446170e-01
-4.6637837e-02 -9.7570837e-02 4.8465633e-01 3.4319603e-01
-1.8721908e-01 8.4040356e-01 1.3546530e+00 -2.9756716e-01
-1.4011268e-01 1.6321683e-01 -5.7765853e-01 5.5874282e-01
1.0057365e+00 -1.9016413e-01 -5.7314080e-01 -3.3187699e-01
4.4841057e-01 2.6750809e-01 2.5945985e-01 -8.3560467e-01
3.4358910e-01 -5.3053495e-02 4.3023923e-01 -7.4524879e-01
2.1942122e-01 -8.6894208e-01 1.3249667e-02 8.8136953e-01
-2.0176817e-01 9.2028397e-01 2.2788985e-01 8.8403493e-01
-6.6705430e-01 8.8357233e-02 4.7966820e-01 -5.3185921e-02
-8.0925041e-01 7.1930802e-01 4.5480259e-02 1.2060067e-01
1.1463398e+00 8.2798533e-02 -5.4810244e-01 -5.3794378e-01
-6.5500766e-01 5.1398903e-01 -4.9615394e-02 7.0337021e-01
8.1338918e-01 -1.5919248e+00 -5.8697051e-01 3.2888469e-01
-2.1740372e-01 -5.2212387e-01 3.1792548e-02 8.0498952e-01
-6.8085486e-01 5.9439284e-01 -3.5927954e-01 -2.0471810e-01
-5.5985492e-01 9.1042721e-01 4.7404566e-01 -6.1712021e-01
-6.2125969e-01 4.7997925e-01 7.2381303e-02 1.2315251e-03
2.1298568e-01 -7.3654175e-01 -6.6160655e-01 6.1602116e-01
3.3192205e-01 2.8637904e-01 -1.2505385e-01 -5.3097796e-01
-2.9843917e-01 8.7592500e-01 -8.1205599e-02 2.9430591e-02
1.9093323e+00 1.5132819e-01 -2.2968730e-01 6.1685866e-01
1.1427767e+00 -2.7418885e-01 -1.2425981e+00 -3.7101275e-01
3.9776146e-01 -1.0628402e-01 6.9331191e-02 -4.6247524e-01
-1.7666632e+00 1.0164622e+00 3.8957122e-01 1.0096831e+00
1.0755776e+00 -4.0664533e-01 8.9952433e-01 3.6328748e-01
2.2305815e-01 -1.0095141e+00 1.7124589e-01 5.3595769e-01
9.2573839e-01 -1.0229149e+00 -1.0082652e-01 -1.9900089e-02
-6.5941346e-01 1.4518459e+00 3.4168994e-01 -7.3720688e-01
1.3172631e+00 4.7740925e-02 2.1004678e-01 -3.1091452e-01
-8.1266814e-01 6.0122583e-02 4.4332388e-01 2.0544064e-01
2.3645858e-01 1.8704601e-02 -2.8888863e-01 8.3611655e-01
-3.9865613e-01 -2.5223038e-01 4.2880887e-01 6.1708516e-01
-2.0746495e-01 -6.5007573e-01 5.2808441e-02 5.3546095e-01
-5.4722381e-01 -2.8059116e-01 -2.0784125e-01 1.0506945e+00
-4.5382956e-01 6.4743942e-01 4.5359382e-01 -6.0702145e-01
5.0721496e-01 -7.4938662e-02 -1.3862167e-01 -1.6844161e-01
-8.4337515e-01 2.5015491e-01 -2.9053783e-01 -3.4483185e-01
-3.9364749e-01 -4.3737873e-01 -1.1941711e+00 -5.1151514e-01
-2.1764411e-01 1.9801284e-01 3.3991581e-01 6.2148273e-01
4.2117029e-01 9.1863596e-01 1.0396175e+00 -8.7945372e-01
-8.6996680e-01 -8.6130953e-01 -9.7066051e-01 3.0039945e-01
5.7237446e-01 -6.8325984e-01 -4.9934015e-01 -3.6733890e-01] | [4.373880863189697, 4.292166709899902] |
9e7c9aa8-5ecd-48ac-814d-6880ad73c3d2 | offline-primal-dual-reinforcement-learning | 2305.12944 | null | https://arxiv.org/abs/2305.12944v1 | https://arxiv.org/pdf/2305.12944v1.pdf | Offline Primal-Dual Reinforcement Learning for Linear MDPs | Offline Reinforcement Learning (RL) aims to learn a near-optimal policy from a fixed dataset of transitions collected by another policy. This problem has attracted a lot of attention recently, but most existing methods with strong theoretical guarantees are restricted to finite-horizon or tabular settings. In constrast, few algorithms for infinite-horizon settings with function approximation and minimal assumptions on the dataset are both sample and computationally efficient. Another gap in the current literature is the lack of theoretical analysis for the average-reward setting, which is more challenging than the discounted setting. In this paper, we address both of these issues by proposing a primal-dual optimization method based on the linear programming formulation of RL. Our key contribution is a new reparametrization that allows us to derive low-variance gradient estimators that can be used in a stochastic optimization scheme using only samples from the behavior policy. Our method finds an $\varepsilon$-optimal policy with $O(\varepsilon^{-4})$ samples, improving on the previous $O(\varepsilon^{-5})$, while being computationally efficient for infinite-horizon discounted and average-reward MDPs with realizable linear function approximation and partial coverage. Moreover, to the best of our knowledge, this is the first theoretical result for average-reward offline RL. | ['Matteo Papini', 'Nneka Okolo', 'Gergely Neu', 'Germano Gabbianelli'] | 2023-05-22 | null | null | null | null | ['stochastic-optimization', 'offline-rl'] | ['methodology', 'playing-games'] | [-6.39704764e-02 1.69453055e-01 -9.35425699e-01 -8.44898745e-02
-1.12742853e+00 -7.10802615e-01 -6.71787607e-03 3.41196328e-01
-6.91720665e-01 1.27210581e+00 -2.21008882e-01 -6.80546463e-01
-5.16855955e-01 -7.85139561e-01 -9.27233040e-01 -7.61447966e-01
-5.67809343e-01 4.78579998e-01 5.42747080e-02 -1.60313457e-01
1.56141162e-01 2.44851634e-01 -1.36631536e+00 -3.03144634e-01
1.00359583e+00 1.36427701e+00 2.17477962e-01 4.43611264e-01
2.41423443e-01 8.26715648e-01 -5.06076813e-01 -2.36999676e-01
4.96988773e-01 -5.14316380e-01 -6.54173911e-01 2.97227707e-02
-2.51783840e-02 -8.18271160e-01 -2.52833694e-01 1.10562241e+00
4.38513070e-01 4.50860858e-01 1.53447285e-01 -1.30636597e+00
-2.10668102e-01 7.97933698e-01 -5.32426357e-01 1.69354126e-01
3.21735412e-01 4.03726488e-01 1.08568966e+00 5.87722147e-03
3.30312133e-01 1.11088085e+00 2.10752487e-01 5.51833570e-01
-1.18280745e+00 -4.44073260e-01 6.45050466e-01 1.19487539e-01
-7.73855388e-01 -8.09600204e-02 4.37785894e-01 -4.00747918e-02
8.44796658e-01 2.00446203e-01 8.00891161e-01 7.75393426e-01
6.99506374e-03 1.04045868e+00 1.33667052e+00 -3.24578732e-01
5.90475440e-01 6.93668872e-02 -1.79499835e-01 5.94308972e-01
3.01054120e-01 3.17685694e-01 -7.40115196e-02 -2.16101095e-01
7.01657474e-01 2.66337603e-01 -1.28829345e-01 -3.75983417e-01
-7.82626748e-01 1.01231658e+00 2.84405425e-02 -1.91026956e-01
-5.60478687e-01 5.28020382e-01 4.24701601e-01 5.09365380e-01
4.11536902e-01 2.90872604e-01 -4.35261071e-01 -7.72168040e-01
-8.27809215e-01 5.62916338e-01 9.31373596e-01 9.75280762e-01
5.26445150e-01 2.30805844e-01 -4.08420265e-01 4.77256596e-01
-8.73384401e-02 7.96107352e-01 3.30989733e-02 -1.24310231e+00
8.60737622e-01 8.79838318e-02 8.17928731e-01 -4.54725325e-01
-1.77998036e-01 -3.29346716e-01 -1.76737696e-01 2.50902563e-01
7.67568111e-01 -6.54836118e-01 -3.67118835e-01 1.84335387e+00
3.42985153e-01 -3.17834765e-02 -1.00949027e-01 8.48835826e-01
-2.26248443e-01 6.42186940e-01 -3.60209346e-01 -9.26562846e-01
9.63389695e-01 -7.23293662e-01 -5.18213630e-01 -1.48677200e-01
4.63277072e-01 -3.32795709e-01 1.16144931e+00 4.02068049e-01
-1.37227750e+00 1.93163231e-01 -7.57581711e-01 5.18386185e-01
2.02464715e-01 -5.12897260e-02 7.46317685e-01 8.10271382e-01
-7.25712836e-01 6.85479283e-01 -8.80883753e-01 7.03232214e-02
5.13704538e-01 3.52080613e-01 3.35589468e-01 -1.15321487e-01
-9.12419319e-01 8.10346782e-01 3.91335130e-01 -1.95130587e-01
-1.41206729e+00 -5.79133391e-01 -6.04604483e-01 8.42779279e-02
1.53405190e+00 -1.39419034e-01 1.77918732e+00 -7.50860929e-01
-1.87366080e+00 1.31426379e-01 -9.11976025e-02 -8.17237735e-01
7.16005504e-01 1.17284590e-02 8.69826972e-02 2.16763288e-01
-1.04360040e-02 -1.17804734e-02 6.66002035e-01 -9.45433259e-01
-8.83252203e-01 -4.22844112e-01 8.16658974e-01 2.47710377e-01
-2.69811779e-01 -1.51140884e-01 -1.18952729e-01 -3.35366756e-01
-5.05810797e-01 -9.60925162e-01 -6.57882273e-01 -3.84837806e-01
-1.67121217e-01 -2.48778015e-01 2.68540919e-01 -4.75669295e-01
1.45976233e+00 -1.75327337e+00 -3.37653458e-02 1.04904443e-01
-2.84656107e-01 4.81841937e-02 1.00430548e-01 7.55281866e-01
5.40203094e-01 -6.89082444e-02 -2.09194541e-01 -1.65124997e-01
2.05620810e-01 3.63660038e-01 -5.48107684e-01 6.39332116e-01
-2.85114944e-01 6.49299204e-01 -1.19748199e+00 -2.09590629e-01
9.80115458e-02 -3.39986265e-01 -6.60193443e-01 1.27663076e-01
-6.98700488e-01 1.79161429e-01 -6.78773642e-01 6.07045949e-01
5.14729083e-01 -3.91786769e-02 4.89020586e-01 5.02313554e-01
-2.98744470e-01 1.59090981e-01 -1.21897662e+00 1.40137744e+00
-5.87494910e-01 1.22095942e-01 2.67824590e-01 -1.47374058e+00
5.71435392e-01 1.12442620e-01 9.89655793e-01 -8.11391473e-01
2.60796249e-01 2.75566310e-01 -2.78936744e-01 -4.43763494e-01
5.23930252e-01 -3.89221251e-01 -2.92995542e-01 7.46351659e-01
-3.01327556e-01 1.07250713e-01 4.33669716e-01 -7.62734786e-02
1.09961307e+00 1.55682355e-01 3.35420221e-01 -9.37249362e-02
2.48099323e-02 -8.47705379e-02 6.40080333e-01 1.03910506e+00
-4.48088467e-01 -2.18415260e-01 1.05100644e+00 -1.32612929e-01
-8.89115393e-01 -8.31294060e-01 1.50478899e-01 1.03604102e+00
1.72155038e-01 3.94624993e-02 -7.52390802e-01 -8.41556609e-01
3.59861225e-01 8.97882521e-01 -5.63118219e-01 1.53406067e-02
-3.56533349e-01 -5.61012447e-01 2.24697307e-01 3.87618840e-01
2.36462504e-01 -8.28203440e-01 -1.00480998e+00 4.19185638e-01
2.55877580e-02 -1.06646204e+00 -5.86605072e-01 2.02172920e-01
-9.56155717e-01 -9.41356242e-01 -7.22530723e-01 -2.06838623e-01
6.11596465e-01 2.25470468e-01 7.27618575e-01 -5.24207234e-01
5.56556806e-02 7.24791169e-01 -3.20123613e-01 -5.93053520e-01
-9.12428647e-02 -1.16192020e-01 1.20641470e-01 -7.32670873e-02
1.31965019e-02 -2.57436216e-01 -7.28498638e-01 1.74358070e-01
-7.95712590e-01 -3.92182201e-01 3.65688682e-01 7.31092751e-01
8.42753947e-01 1.72160044e-01 8.48422170e-01 -7.36309171e-01
9.02265310e-01 -6.50115252e-01 -1.31047845e+00 3.44815493e-01
-8.39614987e-01 3.99665356e-01 1.19362402e+00 -4.70648497e-01
-8.06659758e-01 -1.01101756e-01 1.15904480e-01 -5.57079911e-01
2.70136148e-01 3.88606161e-01 2.23312005e-02 2.32915983e-01
3.64634573e-01 4.59982753e-01 2.21983850e-01 -3.63305628e-01
2.79548645e-01 4.54799950e-01 1.79410547e-01 -1.12789500e+00
3.51388276e-01 5.05174279e-01 1.89011589e-01 -4.14917201e-01
-8.96254659e-01 -2.01814950e-01 1.36613145e-01 -2.90204942e-01
2.34157115e-01 -6.50393188e-01 -1.52779663e+00 -6.45110989e-03
-4.55629677e-01 -8.52566779e-01 -7.73823082e-01 5.43895185e-01
-1.17600858e+00 5.34540415e-01 -2.76997089e-01 -1.65962493e+00
-1.37319863e-01 -1.11869931e+00 6.00222647e-01 2.36027092e-01
4.26031977e-01 -7.14176714e-01 1.77603900e-01 1.34339720e-01
1.94454968e-01 2.83863425e-01 5.63732624e-01 -1.59509823e-01
-5.89327514e-01 -8.85042250e-02 7.23254234e-02 3.37266147e-01
-1.18470572e-01 -3.91372830e-01 -4.93371546e-01 -8.36104333e-01
-5.88073581e-02 -6.54357314e-01 6.88617527e-01 7.01491892e-01
1.48450541e+00 -8.55841875e-01 -1.01537995e-01 2.26963013e-01
1.67324626e+00 7.20743477e-01 3.24964166e-01 4.40237045e-01
9.04412270e-02 3.04454863e-01 1.25903130e+00 1.31606805e+00
3.72230709e-01 5.98122895e-01 8.02128375e-01 4.93316263e-01
6.65552974e-01 -3.93434614e-01 6.67483926e-01 2.53881123e-02
-1.84330091e-01 -2.54692286e-01 -4.83277649e-01 6.80368006e-01
-2.05032897e+00 -1.21782470e+00 4.87949908e-01 2.92989469e+00
9.77902949e-01 1.75571457e-01 7.30947196e-01 -4.02825661e-02
4.62474436e-01 5.60820103e-02 -1.18159056e+00 -6.92382395e-01
2.48530790e-01 3.60767990e-01 9.92386222e-01 4.37500268e-01
-7.14984655e-01 6.84058309e-01 5.70994854e+00 1.16725326e+00
-8.96906555e-01 1.05608985e-01 6.43765807e-01 -8.00757945e-01
-2.69650012e-01 2.77524471e-01 -8.51504803e-01 8.05331826e-01
1.19494724e+00 -4.72737223e-01 1.13801599e+00 1.08637798e+00
5.76356292e-01 -4.95062709e-01 -1.09842265e+00 8.67406309e-01
-2.85728365e-01 -1.03323221e+00 -5.82670867e-01 4.16230589e-01
9.17171419e-01 -1.78730369e-01 2.16222882e-01 6.18617773e-01
6.69875324e-01 -7.89981723e-01 7.58086979e-01 5.10753632e-01
8.94714355e-01 -1.38672721e+00 3.61945808e-01 6.49900913e-01
-1.03730893e+00 -7.05832481e-01 -3.34497869e-01 -2.27077544e-01
2.37928078e-01 6.58831835e-01 -5.21187901e-01 5.17056406e-01
4.88969058e-01 3.46130818e-01 4.42428410e-01 9.15337384e-01
-5.39285205e-02 6.58482790e-01 -4.25630331e-01 -4.77447838e-01
5.33500016e-01 -2.85776854e-01 4.59051579e-01 7.68118203e-01
3.68766129e-01 4.01066877e-02 8.30093086e-01 5.76522410e-01
-2.86224019e-02 9.31748524e-02 -4.52671468e-01 -3.26459885e-01
5.46899140e-01 8.64465058e-01 -5.13594925e-01 -1.74336672e-01
-2.55961448e-01 6.18134201e-01 4.73183304e-01 2.68341124e-01
-1.14030969e+00 -3.42419654e-01 5.57201982e-01 2.23918892e-02
5.02121150e-01 -3.87536496e-01 3.64031270e-02 -1.00021565e+00
3.77898037e-01 -8.61426055e-01 5.33922851e-01 1.45211115e-01
-9.93771076e-01 -7.27808177e-02 3.19857061e-01 -1.17644596e+00
-5.27743936e-01 -4.03436750e-01 -3.26168209e-01 5.86163461e-01
-1.54149508e+00 -3.45147967e-01 3.58838618e-01 5.96317470e-01
4.52664614e-01 -5.51100969e-02 4.35189575e-01 1.08187839e-01
-6.95410728e-01 7.63121307e-01 8.31382275e-01 -4.21939582e-01
2.02485502e-01 -1.35786200e+00 -3.34010124e-01 6.48817420e-01
-5.48344612e-01 9.52605233e-02 7.59972572e-01 -3.87678951e-01
-1.80273306e+00 -9.28589761e-01 9.53272730e-02 1.22615561e-01
7.06089020e-01 -4.21651006e-02 -3.63977432e-01 5.63466311e-01
-1.19708747e-01 3.08291446e-02 3.36255401e-01 1.35254851e-02
2.16851249e-01 -3.42976421e-01 -1.34514225e+00 7.97284365e-01
8.97808194e-01 -2.63064712e-01 7.80085772e-02 4.03542519e-01
5.95445275e-01 -5.40438771e-01 -9.28900063e-01 3.50461528e-02
5.07963538e-01 -7.29236484e-01 6.06644332e-01 -7.59407580e-01
1.89640343e-01 2.22552940e-01 -1.85440794e-01 -1.28063500e+00
2.35006735e-01 -1.14634597e+00 -6.44241273e-01 7.38886535e-01
2.74211794e-01 -8.91358674e-01 8.27047169e-01 6.19094968e-01
-3.28804851e-02 -1.32774901e+00 -1.32065189e+00 -1.43722463e+00
2.59529561e-01 -5.54829240e-01 5.05607963e-01 4.32107508e-01
3.76228392e-01 -2.74193138e-01 -7.14873075e-01 -1.38885394e-01
8.24464142e-01 4.76550132e-01 5.90147436e-01 -5.36793828e-01
-8.96316171e-01 -3.05911303e-01 1.98886454e-01 -1.45316470e+00
2.37765297e-01 -6.34934008e-01 2.57666349e-01 -1.45024586e+00
3.27329695e-01 -7.46006072e-01 -3.18153590e-01 5.03556967e-01
1.05330274e-01 -7.25439370e-01 3.12811136e-01 -9.10091400e-02
-9.23336923e-01 9.76490796e-01 1.33848739e+00 2.77106860e-03
-3.34109694e-01 4.73530233e-01 -6.58462763e-01 3.47986042e-01
1.02457976e+00 -5.70576370e-01 -7.80508220e-01 -1.28761694e-01
1.80247307e-01 9.28433299e-01 7.48887062e-02 -6.14555776e-01
-1.93591118e-01 -9.23568606e-01 -4.37408715e-01 -5.19878864e-01
2.38585725e-01 -7.04197824e-01 -2.03086466e-01 8.15982580e-01
-4.54619855e-01 -5.92991784e-02 9.45679992e-02 1.03697288e+00
2.76858807e-01 -6.66252196e-01 6.51694536e-01 -3.25902343e-01
-3.01504970e-01 6.19828522e-01 -5.16223550e-01 5.06573558e-01
1.49660838e+00 5.98483197e-02 -2.71282285e-01 -6.97908640e-01
-4.65446740e-01 6.28967762e-01 1.97168559e-01 -3.68930474e-02
3.78677994e-01 -1.05606079e+00 -4.73861903e-01 -5.14520705e-01
-3.78253818e-01 -8.42035711e-02 2.75991559e-01 8.64548266e-01
-7.36378059e-02 6.50744677e-01 -2.80037280e-02 -2.13651255e-01
-6.93569541e-01 8.47380817e-01 3.43321860e-01 -7.96476066e-01
-3.24118912e-01 3.89283717e-01 -4.41537410e-01 -1.64408237e-02
3.61690074e-01 -5.26194453e-01 3.29707474e-01 -9.25126718e-04
4.44894999e-01 8.50566447e-01 -3.12824100e-01 5.30881137e-02
-8.88621286e-02 -3.50251831e-02 -1.17429443e-01 -5.51168203e-01
1.21544933e+00 -9.92888678e-03 3.48788440e-01 3.74023765e-01
1.00378144e+00 -1.80306777e-01 -1.85501075e+00 -3.51775914e-01
-1.74998283e-01 -6.72949791e-01 6.49637356e-02 -5.99900365e-01
-9.95238900e-01 5.99848866e-01 5.65429926e-01 4.23079133e-01
1.07018328e+00 -3.24719101e-01 8.73315930e-01 4.90392089e-01
1.02633584e+00 -1.60656691e+00 6.28678948e-02 4.43055838e-01
5.25054157e-01 -1.13324213e+00 4.04225588e-02 2.16404632e-01
-8.02074254e-01 9.27337706e-01 4.24921572e-01 -2.30921522e-01
3.67459536e-01 1.67984754e-01 -7.78307736e-01 1.60052583e-01
-8.57443988e-01 -5.24959147e-01 -3.87242615e-01 3.09871256e-01
5.16722025e-03 5.09955287e-01 -7.53787994e-01 5.68006873e-01
1.51191115e-01 2.63241172e-01 6.53404713e-01 1.32717133e+00
-6.86300337e-01 -1.21671581e+00 -2.39873573e-01 6.68734610e-01
-6.81762874e-01 2.30951354e-01 1.29142448e-01 5.58974326e-01
-3.02535325e-01 1.11243474e+00 -8.15033168e-02 8.75862166e-02
3.16782027e-01 -2.67301738e-01 7.96265781e-01 -2.73896754e-01
-2.93723732e-01 3.46785709e-02 2.15178058e-01 -7.67810762e-01
-1.59004718e-01 -8.32641006e-01 -1.30353582e+00 -5.82819223e-01
-2.17953995e-01 2.42457315e-01 5.42196274e-01 1.00010109e+00
1.72975883e-01 3.30869168e-01 1.14526963e+00 -5.69989383e-01
-1.46646118e+00 -5.61269701e-01 -8.55724931e-01 -2.29722541e-02
3.04063916e-01 -7.89408863e-01 -2.33839110e-01 -7.52854645e-01] | [4.3239240646362305, 2.752511739730835] |
04ff4fc3-6481-46c9-b9af-ec602f23fcc2 | egyptian-sign-language-recognition-using-cnn | 2107.13647 | null | https://arxiv.org/abs/2107.13647v1 | https://arxiv.org/pdf/2107.13647v1.pdf | Egyptian Sign Language Recognition Using CNN and LSTM | Sign language is a set of gestures that deaf people use to communicate. Unfortunately, normal people don't understand it, which creates a communication gap that needs to be filled. Because of the variations in (Egyptian Sign Language) ESL from one region to another, ESL provides a challenging research problem. In this work, we are providing applied research with its video-based Egyptian sign language recognition system that serves the local community of deaf people in Egypt, with a moderate and reasonable accuracy. We present a computer vision system with two different neural networks architectures. The first is a Convolutional Neural Network (CNN) for extracting spatial features. The CNN model was retrained on the inception mod. The second architecture is a CNN followed by a Long Short-Term Memory (LSTM) for extracting both spatial and temporal features. The two models achieved an accuracy of 90% and 72%, respectively. We examined the power of these two architectures to distinguish between 9 common words (with similar signs) among some deaf people community in Egypt. | ['Rawan Gla', 'Ahmed Elhagry'] | 2021-07-28 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [-1.04718238e-01 -4.90475476e-01 9.24011916e-02 -3.54474843e-01
-3.95116240e-01 -3.68547708e-01 5.41610479e-01 -7.81937122e-01
-7.95434237e-01 6.26933575e-01 6.46621704e-01 -3.20776314e-01
-1.59021512e-01 -5.10842741e-01 -1.36097729e-01 -8.26469660e-01
-1.88804910e-01 3.05371750e-02 1.95357010e-01 -3.40431541e-01
5.90822816e-01 8.34370375e-01 -1.69523299e+00 5.74946404e-01
6.36320710e-01 6.03139162e-01 1.66888177e-01 8.45799088e-01
-4.76831555e-01 9.72425282e-01 -5.05747259e-01 1.63973406e-01
2.08204001e-01 -3.81588638e-01 -8.49078000e-01 -2.82916546e-01
6.77602768e-01 -9.12905157e-01 -8.67905080e-01 8.50522280e-01
1.05050230e+00 -6.54859245e-02 6.13211811e-01 -1.24885046e+00
-9.00146484e-01 4.40217644e-01 -1.75475344e-01 -3.64839695e-02
3.41101110e-01 3.05494845e-01 5.66438496e-01 -8.40117395e-01
5.87324619e-01 1.33539200e+00 7.33725429e-01 7.91207552e-01
-2.43423894e-01 -9.72747862e-01 1.07720383e-01 4.88787264e-01
-1.41040421e+00 -3.63991469e-01 5.54086387e-01 -5.11189759e-01
9.96994376e-01 -1.31441444e-01 9.71635640e-01 7.67965913e-01
-2.11604056e-03 1.04709613e+00 1.35852468e+00 -6.94339275e-01
-1.74688369e-01 -3.71202022e-01 3.69112283e-01 5.11849880e-01
1.69368863e-01 3.42388451e-01 -5.55190325e-01 2.63369292e-01
8.64778340e-01 3.37168276e-02 -4.44510072e-01 3.34047824e-01
-9.56047118e-01 5.37226140e-01 6.37967050e-01 9.44275022e-01
-5.22512436e-01 2.65970320e-01 1.87273294e-01 7.30840027e-01
-4.08380419e-01 -2.75732994e-01 -5.92467964e-01 -3.76907557e-01
-6.84300721e-01 -3.98605578e-02 7.17929959e-01 7.29757369e-01
2.52492689e-02 2.63158143e-01 -1.15520351e-01 1.00229442e+00
9.33933794e-01 7.35642195e-01 8.19996178e-01 -3.82070035e-01
1.55816078e-01 6.91833079e-01 -8.09368119e-02 -6.49811208e-01
-3.24332952e-01 1.73059508e-01 -5.90027153e-01 8.78816664e-01
7.79734313e-01 -4.36716557e-01 -1.75213897e+00 1.52229071e+00
-3.74529153e-01 -6.16957173e-02 1.97238743e-01 9.89639342e-01
1.26157284e+00 4.72748369e-01 1.93316817e-01 3.06095690e-01
1.06262112e+00 -7.31700420e-01 -7.38076329e-01 -4.93197292e-02
6.10648215e-01 -9.81815994e-01 8.53102267e-01 3.08503568e-01
-8.59104872e-01 -3.55815031e-02 -8.04241657e-01 -8.88241529e-02
-8.01551402e-01 3.97320330e-01 6.60014927e-01 6.56555772e-01
-1.49803472e+00 2.76034951e-01 -4.90317076e-01 -1.12399101e+00
3.28408629e-01 5.98769188e-01 -5.12820721e-01 -7.45072141e-02
-8.63642991e-01 1.11863923e+00 4.21056524e-02 6.31410301e-01
-2.07042292e-01 1.78450942e-01 -5.29583812e-01 -5.04096687e-01
-5.73075771e-01 1.33375470e-02 1.27920210e+00 -1.28003263e+00
-1.73180664e+00 1.06051767e+00 -2.08884716e-01 1.39784992e-01
7.37192333e-01 -2.31283501e-01 -6.97250426e-01 -1.61486208e-01
-1.86938494e-01 7.22201228e-01 4.73830640e-01 -1.12162185e+00
-9.91130829e-01 -5.77356279e-01 -2.74005443e-01 -1.28395353e-02
3.70737985e-02 5.62308133e-01 -3.68909895e-01 -6.25857651e-01
7.74150252e-01 -8.95994663e-01 3.06330383e-01 4.96882945e-01
-3.93193960e-02 -2.55558223e-01 1.22835851e+00 -1.18068290e+00
1.05729699e+00 -2.01265025e+00 -3.19626898e-01 3.21642965e-01
-1.90279588e-01 8.06431472e-01 -4.88615274e-01 5.88572741e-01
1.14924200e-02 -3.65781747e-02 -1.72479928e-01 3.06231588e-01
-1.57195434e-01 3.88791859e-01 -7.68833235e-02 4.49318051e-01
-7.67364949e-02 8.99032116e-01 -7.57950127e-01 -3.44352037e-01
3.06094736e-01 8.58774185e-01 -6.06377684e-02 5.55235371e-02
3.06765378e-01 4.45791453e-01 -2.76233733e-01 1.27643907e+00
7.89101362e-01 3.83654475e-01 2.28348821e-02 2.70975032e-03
-5.23538232e-01 -1.59629747e-01 -1.23735166e+00 1.24134517e+00
-2.71408796e-01 1.30845559e+00 1.64678708e-01 -7.58501232e-01
8.44174206e-01 6.25791371e-01 1.60552725e-01 -8.29056144e-01
4.90062326e-01 8.65478218e-01 4.03679073e-01 -1.24001741e+00
1.12051265e-02 2.70149354e-02 5.10262191e-01 3.87399524e-01
-1.97313935e-01 2.24015653e-01 8.32396746e-02 -3.97737771e-01
8.78553569e-01 4.76709865e-02 4.92229238e-02 2.36373339e-02
6.60753727e-01 -2.49250144e-01 3.14106286e-01 5.21459222e-01
-5.65216184e-01 5.57143629e-01 -5.38759530e-02 -5.80258906e-01
-5.80874503e-01 -9.35528874e-01 4.29462790e-02 8.61935854e-01
-1.71381801e-01 5.23716629e-01 -2.16413930e-01 -2.73507655e-01
3.06290593e-02 2.80376494e-01 -3.13792080e-01 2.63431162e-01
-9.29826319e-01 -2.30065674e-01 9.37694252e-01 8.78008485e-01
1.28676713e+00 -1.81366384e+00 -7.90900946e-01 1.54289812e-01
9.13035944e-02 -6.36612594e-01 -3.43764424e-01 -2.14887679e-01
-7.44666219e-01 -1.20379639e+00 -1.44256735e+00 -1.70638156e+00
6.99798703e-01 1.23093069e-01 2.75660068e-01 2.31208444e-01
-2.15323031e-01 2.93393970e-01 -5.45579076e-01 -5.42665184e-01
-6.58204108e-02 -3.68317097e-01 -2.09854811e-01 -1.14497401e-01
1.04178548e+00 -5.21430790e-01 -4.13582325e-01 1.14621833e-01
-7.26755321e-01 -2.50128001e-01 9.42167759e-01 8.01621616e-01
-3.29787850e-01 -5.36317706e-01 3.03447574e-01 8.35578144e-02
8.23873878e-01 1.28895253e-01 -4.61881548e-01 4.60303634e-01
-2.51713723e-01 7.79182911e-02 -2.02416331e-01 -6.13431752e-01
-9.16470230e-01 9.06397402e-02 -2.62279332e-01 3.70416284e-01
-4.06139344e-01 4.74410564e-01 -3.54101919e-02 -5.75059950e-01
1.87547371e-01 5.19871771e-01 6.72014728e-02 -6.31334066e-01
1.23891115e-01 1.54766643e+00 5.84819496e-01 -6.38343543e-02
4.25040334e-01 3.66084635e-01 -4.36995804e-01 -1.16672385e+00
1.58811614e-01 -4.09574598e-01 -8.64996552e-01 -5.57784259e-01
8.04218948e-01 -5.98107517e-01 -9.16630447e-01 1.75618851e+00
-1.36898446e+00 -3.54038894e-01 1.29382938e-01 1.01675320e+00
-1.27687484e-01 3.61687094e-02 -6.57717943e-01 -1.04709804e+00
-2.08461106e-01 -8.35686982e-01 7.38148391e-01 4.26968187e-01
-5.52775338e-02 -6.49406672e-01 9.75684747e-02 -3.03539485e-02
7.85248041e-01 2.41461918e-01 7.48914301e-01 -3.11482906e-01
-5.13659716e-01 -3.91867548e-01 -8.45278502e-01 4.24516261e-01
3.06720614e-01 1.51627794e-01 -1.03355944e+00 -2.16916203e-01
-4.94147718e-01 -2.25042239e-01 1.04794395e+00 7.77838349e-01
3.76594305e-01 -2.44882807e-01 -8.11620429e-02 4.05101568e-01
1.39449620e+00 1.06541789e+00 9.17718410e-01 4.34115320e-01
3.71441871e-01 5.89883387e-01 -5.60253225e-02 1.92149594e-01
3.83504868e-01 3.13831091e-01 -1.03318803e-01 -1.65274724e-01
-5.51471651e-01 -1.87888011e-01 4.94325072e-01 8.49787414e-01
-7.26018667e-01 -4.95985821e-02 -1.12590718e+00 8.83146882e-01
-1.70620263e+00 -1.11816871e+00 -2.49085546e-01 1.90191603e+00
5.92644930e-01 -4.77797568e-01 3.37835923e-02 5.63335538e-01
8.37018430e-01 -2.88098641e-02 -3.06047857e-01 -5.14849663e-01
-4.59813625e-01 4.01937306e-01 5.80984056e-01 6.58882380e-01
-1.02795196e+00 1.20537102e+00 6.54101229e+00 1.78385228e-02
-1.91930163e+00 -2.34582469e-01 -1.29741266e-01 3.77526730e-01
2.67667860e-01 -2.36318186e-01 -4.70023066e-01 3.17532361e-01
4.18816596e-01 2.30827481e-01 2.76035160e-01 5.01626849e-01
3.62833470e-01 -4.07080203e-02 -7.15859354e-01 1.18910980e+00
1.75895855e-01 -8.65950882e-01 2.97715634e-01 6.53269738e-02
4.79060858e-01 3.01017642e-01 -2.24378020e-01 2.54180878e-01
2.93006092e-01 -1.39079797e+00 5.93445480e-01 7.32432127e-01
9.75695908e-01 -2.38749772e-01 1.02566302e+00 -3.15639786e-02
-1.35901546e+00 -3.92019778e-01 1.99191168e-01 -5.19724786e-01
3.15229326e-01 -3.80622745e-01 -4.80429262e-01 -4.15847659e-01
7.35030174e-01 5.62206089e-01 -1.07301727e-01 1.55430079e+00
-4.78024691e-01 4.74121243e-01 -4.12858367e-01 -6.16849780e-01
6.30762100e-01 -7.64178038e-02 3.82591873e-01 1.34203720e+00
6.66237354e-01 3.91108930e-01 -2.93389201e-01 2.50588536e-01
2.51613587e-01 1.88991860e-01 -8.08760345e-01 -2.28560969e-01
2.88817495e-01 3.03761065e-01 -3.25869232e-01 -9.67594683e-02
-7.00510800e-01 1.11680722e+00 -4.52787131e-01 7.82975495e-01
-1.56038612e-01 -9.66509700e-01 5.34662902e-01 -1.52487785e-01
-6.34302944e-02 -3.75058204e-01 -3.52288932e-01 -8.92710268e-01
1.10483497e-01 -7.10915804e-01 2.39246950e-01 -9.50335681e-01
-1.13279426e+00 5.01240969e-01 -3.29457998e-01 -1.27516913e+00
-1.17739588e-01 -1.17674494e+00 -7.92438447e-01 1.00871837e+00
-1.66866815e+00 -1.61211348e+00 -7.60865152e-01 9.13985491e-01
5.93994439e-01 -4.91028190e-01 8.43974054e-01 5.29743433e-01
-2.01392382e-01 5.50470412e-01 1.35884523e-01 9.35285747e-01
5.70606232e-01 -8.02765012e-01 -1.88738704e-02 6.90575182e-01
-2.81418115e-01 5.25120318e-01 2.08784908e-01 -5.68979144e-01
-1.04660690e+00 -5.94817221e-01 1.76165140e+00 2.02835694e-01
4.32247519e-01 4.54986781e-01 -4.21058536e-01 8.72835338e-01
2.54605740e-01 -2.16187388e-01 3.19409192e-01 -4.46733087e-01
-3.81818444e-01 1.04880705e-01 -1.37283576e+00 8.01120102e-01
1.24894261e+00 -6.18194580e-01 -8.94464374e-01 -5.01671918e-02
-1.35425776e-01 -7.63767306e-03 -1.56831980e-01 2.83799559e-01
1.48274934e+00 -7.72561729e-01 6.96876287e-01 -5.68867385e-01
1.17515502e-02 -2.78884739e-01 -5.20260692e-01 -9.43639696e-01
6.43545464e-02 -3.49567860e-01 4.04138058e-01 9.02634621e-01
2.60053307e-01 -1.20517397e+00 5.45687318e-01 6.99810922e-01
1.16298966e-01 -2.56299466e-01 -9.60234463e-01 -8.89519274e-01
1.44891366e-01 -4.10269797e-01 4.14907902e-01 7.14721799e-01
-5.57694100e-02 -2.66757548e-01 -1.97833821e-01 7.08479285e-02
2.49536023e-01 -1.85207099e-01 4.79290903e-01 -1.34461880e+00
4.81385767e-01 -9.13036346e-01 -8.74203801e-01 -1.00728869e+00
-1.05029464e-01 -5.01437843e-01 3.96656655e-02 -1.97748351e+00
-1.63237795e-01 -1.47165135e-01 -1.82755336e-01 8.78720164e-01
6.35975480e-01 1.72484681e-01 1.68463930e-01 2.00439528e-01
4.53314215e-01 2.11775690e-01 1.15925336e+00 -3.88272375e-01
-5.89251995e-01 2.09375307e-01 -2.12605059e-01 1.07298958e+00
8.98932278e-01 1.01100937e-01 3.40559194e-03 -8.65838826e-01
-4.01563317e-01 -2.85391569e-01 4.49965805e-01 -1.04165328e+00
4.97459412e-01 -1.87925473e-01 2.70028025e-01 -8.08917165e-01
2.53528208e-01 -8.07594061e-01 -2.46729016e-01 1.01486051e+00
-1.31713197e-01 -5.11371903e-02 1.40555665e-01 -3.34115148e-01
-5.46057403e-01 -4.64760736e-02 1.02779484e+00 -2.21284971e-01
-1.44448793e+00 1.37909636e-01 -6.27220690e-01 -3.39715242e-01
7.61414826e-01 -7.43719935e-01 -2.77918607e-01 -7.06992507e-01
-6.66450739e-01 4.75653410e-02 -5.47281140e-03 4.95733827e-01
9.75180984e-01 -1.45261157e+00 -8.44478130e-01 6.04058444e-01
1.66407719e-01 -5.87756455e-01 -4.67074551e-02 6.83826029e-01
-1.24379706e+00 6.83709681e-01 -7.31490493e-01 -1.56655952e-01
-1.70360839e+00 -4.57664937e-01 4.68120039e-01 4.74690616e-01
-7.44109392e-01 1.05248165e+00 -6.93309844e-01 -5.74708641e-01
8.12853277e-01 -5.68615079e-01 -5.80035985e-01 1.01278342e-01
6.42162323e-01 4.77867663e-01 -3.14600736e-01 -1.19836950e+00
-6.43079221e-01 1.29248106e+00 2.91163236e-01 -4.94080395e-01
1.36579669e+00 1.80510968e-01 -8.97759870e-02 3.49675953e-01
1.13348711e+00 -3.42212230e-01 -7.59410858e-01 -2.75183052e-01
2.61989117e-01 -5.55061042e-01 -5.76998759e-03 -1.30636203e+00
-1.12525153e+00 1.07342184e+00 1.34984362e+00 -4.50732440e-01
1.25468934e+00 -3.20317417e-01 9.60679591e-01 6.78044021e-01
3.72046888e-01 -1.31771064e+00 -3.95335436e-01 9.95128930e-01
1.39985001e+00 -1.31945729e+00 -6.74095690e-01 2.44732678e-01
-3.71263623e-01 1.43443763e+00 3.85056287e-01 -1.36473060e-01
1.05133939e+00 2.68850297e-01 9.63507354e-01 -6.44293353e-02
-1.25446366e-02 -6.71039522e-01 2.54848570e-01 9.48589742e-01
6.95951879e-01 9.13280621e-02 -9.95238364e-01 2.28064671e-01
-1.09793954e-01 6.44669652e-01 3.56656551e-01 1.41939986e+00
-6.88565731e-01 -9.98898447e-01 -4.47528869e-01 3.20229262e-01
-7.63955563e-02 -1.06366068e-01 -6.50475740e-01 1.06303298e+00
3.07908833e-01 1.04686141e+00 -1.37451619e-01 -4.44877416e-01
4.06407356e-01 3.49869817e-01 4.52671647e-01 -2.59564519e-02
-3.10392112e-01 -1.97253555e-01 -3.39960605e-02 -3.18503082e-01
-5.82407415e-01 -6.34326220e-01 -1.43038106e+00 -3.71940494e-01
-1.00986533e-01 -4.75161225e-01 9.03253436e-01 1.07435548e+00
-2.08574668e-01 -5.74960606e-03 1.45496607e-01 -6.90067828e-01
-3.20020705e-01 -1.40452862e+00 -8.25331390e-01 2.13394493e-01
5.39523900e-01 -4.34447020e-01 -2.83670366e-01 -5.93065731e-02] | [9.077654838562012, -6.385664939880371] |
9bec7d26-1d2f-45ff-9e6d-27dcf4f53b09 | multimodal-sentiment-analysis-with-word-level | 1802.00924 | null | http://arxiv.org/abs/1802.00924v1 | http://arxiv.org/pdf/1802.00924v1.pdf | Multimodal Sentiment Analysis with Word-Level Fusion and Reinforcement Learning | With the increasing popularity of video sharing websites such as YouTube and
Facebook, multimodal sentiment analysis has received increasing attention from
the scientific community. Contrary to previous works in multimodal sentiment
analysis which focus on holistic information in speech segments such as bag of
words representations and average facial expression intensity, we develop a
novel deep architecture for multimodal sentiment analysis that performs
modality fusion at the word level. In this paper, we propose the Gated
Multimodal Embedding LSTM with Temporal Attention (GME-LSTM(A)) model that is
composed of 2 modules. The Gated Multimodal Embedding alleviates the
difficulties of fusion when there are noisy modalities. The LSTM with Temporal
Attention performs word level fusion at a finer fusion resolution between input
modalities and attends to the most important time steps. As a result, the
GME-LSTM(A) is able to better model the multimodal structure of speech through
time and perform better sentiment comprehension. We demonstrate the
effectiveness of this approach on the publicly-available Multimodal Corpus of
Sentiment Intensity and Subjectivity Analysis (CMU-MOSI) dataset by achieving
state-of-the-art sentiment classification and regression results. Qualitative
analysis on our model emphasizes the importance of the Temporal Attention Layer
in sentiment prediction because the additional acoustic and visual modalities
are noisy. We also demonstrate the effectiveness of the Gated Multimodal
Embedding in selectively filtering these noisy modalities out. Our results and
analysis open new areas in the study of sentiment analysis in human
communication and provide new models for multimodal fusion. | ['Louis-Philippe Morency', 'Tadas Baltrušaitis', 'Sen Wang', 'Paul Pu Liang', 'Amir Zadeh', 'Minghai Chen'] | 2018-02-03 | null | null | null | null | ['subjectivity-analysis'] | ['natural-language-processing'] | [ 1.73172250e-01 -3.08263972e-02 -6.68497458e-02 -4.17649835e-01
-9.01095390e-01 -2.99100250e-01 5.85486770e-01 1.74722686e-01
-6.43714190e-01 1.66752577e-01 5.80409706e-01 -9.34095904e-02
8.44359584e-03 -4.11216170e-01 -4.06679034e-01 -8.07811141e-01
7.20261876e-03 -2.86963135e-01 -2.71205366e-01 -6.49157524e-01
-6.59406185e-02 1.02859780e-01 -1.77595866e+00 7.93024659e-01
4.32931602e-01 1.34704936e+00 1.04970045e-01 8.97095144e-01
-2.14607283e-01 9.03890967e-01 -4.82594043e-01 -6.72950804e-01
-4.54705536e-01 -4.12284911e-01 -7.08646834e-01 -1.31899975e-02
1.79421723e-01 -1.39285833e-01 -2.24668860e-01 9.04820204e-01
8.11503768e-01 3.23801011e-01 4.06440824e-01 -1.27872884e+00
-5.49481690e-01 5.40352285e-01 -5.71407497e-01 5.46620525e-02
5.36410451e-01 -1.64024428e-01 1.14055657e+00 -9.52430725e-01
2.93099731e-01 1.37654436e+00 5.13201535e-01 6.18909240e-01
-9.02899444e-01 -3.61351520e-01 3.77432942e-01 5.38942516e-01
-9.13780332e-01 -3.63903552e-01 8.77665341e-01 -2.36876637e-01
1.17228711e+00 3.33450973e-01 4.53651726e-01 1.46556509e+00
1.63832664e-01 1.02389753e+00 8.52528572e-01 -4.74036962e-01
-2.32185293e-02 2.78181851e-01 1.18803076e-01 5.54212749e-01
-7.45466948e-01 -2.01823145e-01 -1.05563307e+00 1.52411401e-01
4.35991511e-02 1.82722718e-01 -4.35953625e-02 1.25579700e-01
-1.26711643e+00 9.09138262e-01 2.97410101e-01 6.21244252e-01
-4.68752384e-01 4.21136469e-01 9.40707207e-01 4.28185612e-01
7.99566925e-01 8.22589695e-02 -6.14291430e-01 -3.29541504e-01
-8.68517280e-01 -1.89513832e-01 4.64035541e-01 3.45892161e-01
5.11003137e-01 2.38969833e-01 -8.60591158e-02 1.25974178e+00
5.62085748e-01 7.54139006e-01 6.90018654e-01 -8.51514220e-01
3.11307251e-01 5.66466510e-01 -3.11881989e-01 -1.14517045e+00
-7.89365113e-01 2.67441361e-03 -6.74231946e-01 6.80809245e-02
3.48933376e-02 -3.40066969e-01 -7.37685680e-01 2.06614757e+00
-7.76948966e-03 -2.40260929e-01 2.97331154e-01 9.09666955e-01
1.27322221e+00 7.91924357e-01 3.86077315e-01 -1.25669882e-01
1.72687113e+00 -8.22745025e-01 -1.15592873e+00 -1.21251747e-01
7.59081066e-01 -7.78527379e-01 9.01214421e-01 2.82445759e-01
-1.20289636e+00 -7.05578148e-01 -9.60805297e-01 -2.24251315e-01
-8.24266255e-01 1.22925833e-01 6.49598837e-01 6.53561175e-01
-1.17804384e+00 2.76381254e-01 -7.80275583e-01 -5.20295143e-01
2.22410411e-01 6.35657549e-01 -5.93740523e-01 3.19110394e-01
-1.51415062e+00 1.01361203e+00 -9.15404111e-02 5.03545225e-01
-4.93624508e-01 -3.68208855e-01 -1.31003308e+00 7.22010657e-02
2.14176308e-02 -4.26697761e-01 1.26107216e+00 -1.64423382e+00
-1.70786023e+00 6.53779984e-01 -5.73267162e-01 -2.00420260e-01
-2.20866054e-01 4.53262888e-02 -5.91810226e-01 5.09393215e-01
-3.04095596e-01 1.04005766e+00 9.67813253e-01 -1.20326567e+00
-5.87386310e-01 -4.36323255e-01 7.15516657e-02 3.48438561e-01
-9.36535478e-01 2.15193525e-01 -2.88877755e-01 -6.26376688e-01
-8.06099176e-02 -7.56489396e-01 4.09720885e-03 -3.90918583e-01
-4.09625396e-02 -1.51785001e-01 1.02334809e+00 -7.56891072e-01
1.30111384e+00 -2.30679393e+00 5.96693218e-01 3.81347053e-02
1.75400048e-01 5.48066013e-03 -4.86513704e-01 6.03065610e-01
-3.44845265e-01 2.22758755e-01 -3.79636213e-02 -1.17311358e+00
1.18815631e-01 1.82725653e-01 -2.47554064e-01 3.53413135e-01
2.88031876e-01 1.03282046e+00 -5.71664989e-01 -4.42417622e-01
4.38868344e-01 9.44963396e-01 -4.82482821e-01 4.85033309e-03
6.51302189e-02 5.00895023e-01 -1.72597274e-01 8.27744126e-01
4.52928334e-01 1.12221790e-02 9.52867419e-02 -6.33431435e-01
-2.02262580e-01 -3.24835815e-02 -6.25914812e-01 1.85142183e+00
-7.21178114e-01 1.00258827e+00 2.08937287e-01 -1.08639443e+00
6.22220755e-01 7.35127687e-01 6.83487296e-01 -1.12735438e+00
7.01911092e-01 2.75656302e-02 -8.73686075e-02 -7.95606494e-01
6.77967370e-01 -4.64129686e-01 -4.15681422e-01 3.24351132e-01
6.22939467e-01 -3.52964662e-02 1.85874239e-01 2.30522677e-01
6.37634456e-01 -5.75811528e-02 -1.42001376e-01 1.90656334e-01
8.07108521e-01 -4.66556042e-01 -1.01685869e-02 3.55599165e-01
-3.43912095e-01 5.28255224e-01 6.61022961e-01 -1.08157337e-01
-6.75596535e-01 -6.56273067e-01 1.40536562e-01 1.79655540e+00
-9.45416689e-02 -4.82625902e-01 -6.51368022e-01 -4.58793133e-01
-4.02039021e-01 4.51982647e-01 -8.76886606e-01 -2.15067431e-01
-2.72453368e-01 -8.17318976e-01 4.40234780e-01 6.09447300e-01
8.86671767e-02 -1.29098737e+00 -3.40898961e-01 -2.17639115e-02
-5.85874796e-01 -1.34533191e+00 -2.02986062e-01 2.91434944e-01
-6.24376535e-01 -6.31091893e-01 -7.56162345e-01 -9.06054318e-01
3.18942726e-01 6.23578951e-03 7.49832273e-01 -1.55856654e-01
5.67236319e-02 9.83806014e-01 -6.59590662e-01 -5.89430630e-01
-2.46505767e-01 -7.05769956e-02 -7.19735324e-02 4.11767751e-01
4.46953654e-01 -3.24400157e-01 -3.87665570e-01 1.67951092e-01
-1.10559273e+00 -7.53166601e-02 1.87891334e-01 8.88032556e-01
2.54968554e-01 -2.49373570e-01 8.46239865e-01 -1.72162130e-01
7.33422339e-01 -6.40255570e-01 -4.30699699e-02 -7.48669207e-02
1.06492892e-01 -2.51208872e-01 3.69190156e-01 -4.44535941e-01
-1.20364642e+00 -1.44096762e-01 -5.27157426e-01 -3.62672061e-01
-6.76296186e-03 8.64391685e-01 1.57472603e-02 -3.19921561e-02
1.45353317e-01 3.64831649e-02 3.11840802e-01 -1.61119938e-01
3.77103597e-01 8.09431672e-01 2.23784655e-01 -1.07755549e-01
5.83645850e-02 6.38643086e-01 -8.49107057e-02 -1.21522868e+00
-7.18764305e-01 -4.41539556e-01 -4.15258706e-01 -6.86832011e-01
1.29794109e+00 -9.79442596e-01 -1.36816525e+00 5.40277302e-01
-1.09820688e+00 -4.55381051e-02 -7.75669366e-02 5.64866722e-01
-5.83182752e-01 5.28248370e-01 -8.10331166e-01 -1.16924930e+00
-3.01811010e-01 -1.43226993e+00 1.47694063e+00 1.89991742e-01
-4.30364639e-01 -1.40574038e+00 3.57936919e-02 6.13174319e-01
3.85659397e-01 7.56601393e-02 7.93207347e-01 -4.28290606e-01
1.36108726e-01 -2.55237371e-01 1.43160205e-03 6.93292439e-01
-2.31717914e-01 -2.00044792e-02 -1.44410586e+00 1.91447139e-02
4.20918427e-02 -5.14055073e-01 1.17538869e+00 6.23256385e-01
9.71561551e-01 9.01624039e-02 1.36429712e-01 3.07134926e-01
9.96224761e-01 2.54366219e-01 5.71897984e-01 1.63149863e-01
6.77668333e-01 1.13524091e+00 4.48444217e-01 4.84482884e-01
5.85300624e-01 4.84326333e-01 7.79211044e-01 -5.08014083e-01
1.38355002e-01 1.84649676e-01 7.34243572e-01 1.19614613e+00
-9.53526422e-03 -4.78221178e-01 -6.06247544e-01 5.57263732e-01
-2.02724171e+00 -9.89247799e-01 -6.11572666e-03 1.78181207e+00
4.54086155e-01 -1.99354887e-01 1.16218641e-01 3.72821689e-01
3.57213557e-01 4.21801001e-01 -4.60997336e-02 -1.09494591e+00
-4.89676237e-01 1.95352703e-01 5.93518578e-02 6.60819829e-01
-1.27840841e+00 8.61890793e-01 5.87073851e+00 8.82671893e-01
-1.26659930e+00 3.35049033e-01 6.28622293e-01 -3.76206726e-01
-3.52095008e-01 -5.12046814e-01 -4.38869894e-01 3.91793787e-01
1.25662708e+00 5.09635746e-01 3.40235233e-01 3.69783193e-01
4.37765181e-01 -4.15203661e-01 -9.73464370e-01 1.06884551e+00
4.02815014e-01 -8.76000047e-01 -1.09134741e-01 -1.31189451e-01
4.54681486e-01 -2.31190473e-02 5.36402524e-01 4.60027754e-01
-5.73270679e-01 -1.12839341e+00 8.15446198e-01 7.33239174e-01
5.85749984e-01 -8.71233344e-01 1.08697736e+00 -5.54367527e-02
-1.12786412e+00 -4.25087273e-01 1.35699555e-01 -2.92530097e-02
5.45830667e-01 2.16951236e-01 -2.28325769e-01 4.64662760e-01
7.81327307e-01 8.28809083e-01 -4.33022052e-01 1.90543801e-01
1.05026156e-01 4.51619744e-01 -1.49738699e-01 -2.22186714e-01
3.30220819e-01 -7.56208450e-02 4.54073370e-01 1.54648566e+00
1.97815940e-01 -8.21477473e-02 -2.72926003e-01 3.49183142e-01
3.87977734e-02 2.13978127e-01 -4.93468612e-01 -5.13627231e-01
-3.26073825e-01 1.44365346e+00 -5.48172891e-01 -1.55881077e-01
-7.86451936e-01 9.23857212e-01 6.93451893e-03 5.87746978e-01
-8.45521748e-01 -1.86165377e-01 8.01024497e-01 -5.46029747e-01
3.28264505e-01 -2.76845872e-01 -3.35290253e-01 -9.88268077e-01
-2.53214359e-01 -8.59095216e-01 3.84582728e-01 -9.59104180e-01
-1.16974938e+00 6.87169731e-01 -2.14761600e-01 -1.03540885e+00
-2.69997627e-01 -9.30462599e-01 -3.44568938e-01 7.79335260e-01
-1.52084637e+00 -1.42237699e+00 -2.39807218e-02 7.17195034e-01
6.01673603e-01 -1.21796250e-01 1.03580689e+00 6.22295380e-01
-5.17825603e-01 5.74594259e-01 -2.69388407e-01 -1.76368192e-01
6.30145907e-01 -9.81390357e-01 -3.25920165e-01 4.32536781e-01
-1.14765331e-01 5.45375645e-01 6.74075544e-01 -1.63752779e-01
-1.44840384e+00 -5.09687364e-01 1.11487079e+00 -4.50223923e-01
9.35550570e-01 -4.00576264e-01 -5.37631035e-01 3.91715646e-01
6.92380965e-01 -5.06391704e-01 1.15928841e+00 1.09351225e-01
-2.17042923e-01 -1.16430618e-01 -8.05023789e-01 6.11396194e-01
2.98846900e-01 -9.42789674e-01 -3.84654671e-01 9.03461203e-02
7.61805713e-01 5.21306088e-03 -9.98631239e-01 6.04211211e-01
7.28651643e-01 -9.94175673e-01 9.14758205e-01 -4.35053140e-01
7.63509035e-01 -8.31709336e-03 -5.62783241e-01 -1.22476232e+00
2.51291007e-01 -2.20028490e-01 -1.30877048e-02 1.09390199e+00
5.63454807e-01 -4.09315377e-01 3.43449622e-01 3.58165741e-01
-1.22945204e-01 -9.04581428e-01 -1.06827521e+00 3.02726924e-02
-2.88520813e-01 -1.04279721e+00 1.46200750e-02 7.08928823e-01
6.07356131e-01 5.22103667e-01 -6.76037252e-01 -6.16437681e-02
9.47771966e-02 -1.83797538e-01 3.43718052e-01 -7.91671574e-01
-1.63184218e-02 -7.02680707e-01 -3.65504265e-01 -7.50736773e-01
5.01270592e-01 -6.53172076e-01 -1.38954878e-01 -1.43850195e+00
6.53484538e-02 5.19371450e-01 -5.50284684e-01 5.71182847e-01
4.07036394e-02 6.64249003e-01 2.69597977e-01 -4.25901145e-01
-8.42552722e-01 9.10212159e-01 1.12540698e+00 -2.06532300e-01
-1.01242483e-01 -3.41514885e-01 -6.76082432e-01 8.37033570e-01
5.40537536e-01 -8.45698640e-02 -1.91470519e-01 -2.50290573e-01
6.45181179e-01 1.99944228e-01 2.29909748e-01 -5.24617374e-01
1.51958391e-01 3.40883821e-01 3.47225666e-01 -7.30123997e-01
1.17780197e+00 -9.09084737e-01 -4.59950894e-01 1.53722644e-01
-3.83891135e-01 2.29284577e-02 4.92998064e-01 2.95139402e-01
-7.14215279e-01 -1.04952499e-01 5.43029606e-01 2.15287969e-01
-7.55100369e-01 -2.76120901e-02 -9.72323120e-01 -5.07981539e-01
5.21955669e-01 -1.34012088e-01 -2.15595797e-01 -8.95559788e-01
-1.16605890e+00 3.66116047e-01 -1.98873654e-01 7.41663873e-01
6.86414361e-01 -1.24296737e+00 -3.21402282e-01 1.78595364e-01
1.91537589e-01 -8.29989076e-01 8.32101643e-01 1.41837049e+00
1.42725125e-01 3.50862861e-01 5.53713664e-02 -6.70797348e-01
-1.50252986e+00 3.44643116e-01 4.10569668e-01 8.54384806e-03
1.48189783e-01 8.96547973e-01 2.70608813e-01 -3.93826157e-01
4.02487218e-01 -2.64794379e-01 -8.93693328e-01 9.21399474e-01
5.22911668e-01 3.74408454e-01 6.36816174e-02 -1.21360135e+00
-4.15276200e-01 7.23430693e-01 2.37504244e-01 -5.15946805e-01
1.34566665e+00 -5.56938052e-01 -3.42523754e-01 9.34134424e-01
1.56154752e+00 -8.31899270e-02 -6.85247600e-01 5.17002195e-02
-2.77253807e-01 5.29211722e-02 2.69223034e-01 -7.82226384e-01
-1.08150077e+00 1.14767635e+00 7.24045753e-01 3.98653746e-01
1.47292030e+00 -6.25676010e-04 8.33355665e-01 2.03749239e-01
-3.72897953e-01 -1.36251366e+00 3.97008598e-01 7.93640196e-01
8.63418460e-01 -1.51809967e+00 -4.33379143e-01 9.68771242e-03
-1.05016208e+00 1.17431629e+00 2.35608205e-01 3.42343420e-01
7.78821170e-01 2.86051989e-01 3.33528668e-01 -3.65634799e-01
-6.70673788e-01 -5.29800117e-01 5.78473866e-01 2.62678951e-01
6.84222043e-01 -5.87241463e-02 -2.82176048e-01 9.53593493e-01
2.33245874e-03 -4.25270706e-01 2.28970349e-01 7.87457943e-01
-2.05704659e-01 -9.47207987e-01 -3.95208359e-01 6.60550520e-02
-7.20875740e-01 -2.66823262e-01 -3.09714228e-01 4.91064012e-01
9.30878147e-02 1.49430811e+00 1.22231007e-01 -6.67608082e-01
4.24025863e-01 3.35549116e-01 3.31080526e-01 -2.50721246e-01
-9.42619026e-01 5.30938327e-01 1.80616036e-01 -7.59877205e-01
-1.01218283e+00 -5.49330831e-01 -1.20730221e+00 -1.24188036e-01
-1.92021802e-01 -1.44764870e-01 1.28309774e+00 1.12663460e+00
2.33704433e-01 9.34732676e-01 6.35641992e-01 -1.31460714e+00
1.40691042e-01 -1.06664157e+00 -4.66811419e-01 4.27010089e-01
7.32665300e-01 -5.20042658e-01 -5.53538561e-01 4.11549322e-02] | [13.242037773132324, 5.221391201019287] |
3841a0fe-500b-451d-80f9-c99981ee0201 | low-light-video-enhancement-with-synthetic | 2208.11014 | null | https://arxiv.org/abs/2208.11014v1 | https://arxiv.org/pdf/2208.11014v1.pdf | Low-Light Video Enhancement with Synthetic Event Guidance | Low-light video enhancement (LLVE) is an important yet challenging task with many applications such as photographing and autonomous driving. Unlike single image low-light enhancement, most LLVE methods utilize temporal information from adjacent frames to restore the color and remove the noise of the target frame. However, these algorithms, based on the framework of multi-frame alignment and enhancement, may produce multi-frame fusion artifacts when encountering extreme low light or fast motion. In this paper, inspired by the low latency and high dynamic range of events, we use synthetic events from multiple frames to guide the enhancement and restoration of low-light videos. Our method contains three stages: 1) event synthesis and enhancement, 2) event and image fusion, and 3) low-light enhancement. In this framework, we design two novel modules (event-image fusion transform and event-guided dual branch) for the second and third stages, respectively. Extensive experiments show that our method outperforms existing low-light video or single image enhancement approaches on both synthetic and real LLVE datasets. | ['Qi Tian', 'Yanfeng Wang', 'Houqiang Li', 'Wengang Zhou', 'Xiangyu Chen', 'Shanxin Yuan', 'Jianzhuang Liu', 'Junfeng An', 'Lin Liu'] | 2022-08-23 | null | null | null | null | ['video-enhancement'] | ['computer-vision'] | [ 7.00846732e-01 -7.25982785e-01 2.59689152e-01 -2.70908117e-01
-3.83447468e-01 -2.31531486e-01 5.57472467e-01 -1.45969599e-01
-5.80463767e-01 8.36599350e-01 4.21868935e-02 -1.20995484e-01
2.54219741e-01 -9.34008360e-01 -6.89912736e-01 -8.22854102e-01
4.03753281e-01 -6.60555840e-01 9.20420170e-01 -1.98079541e-01
2.61094332e-01 2.65046567e-01 -2.01072669e+00 4.07394201e-01
8.51632178e-01 9.16507542e-01 4.33699697e-01 8.54803801e-01
-2.35780999e-01 1.08966029e+00 -3.30001175e-01 -1.50488675e-01
3.50245744e-01 -6.72365725e-01 -1.78039581e-01 3.51965100e-01
3.02635431e-01 -6.88275754e-01 -3.54891181e-01 1.18756831e+00
7.10744917e-01 3.70868176e-01 2.61664726e-02 -1.46954036e+00
-2.70791948e-01 -9.34832320e-02 -1.00737739e+00 5.82207918e-01
4.98891205e-01 5.35550356e-01 1.46440148e-01 -1.17805648e+00
6.47835195e-01 1.07014179e+00 5.61376452e-01 3.29218686e-01
-7.32644379e-01 -8.69545102e-01 6.97344216e-03 6.36427701e-01
-9.93620455e-01 -8.03595126e-01 8.01194370e-01 -6.47324547e-02
6.65218592e-01 -8.54011450e-04 7.13614404e-01 5.67079425e-01
4.89967197e-01 4.62538302e-01 1.38966238e+00 -4.34837848e-01
9.25280601e-02 -2.35295296e-01 -2.42292330e-01 5.10543942e-01
3.62115324e-01 6.21177256e-01 -8.01961124e-01 1.08516060e-01
6.84499681e-01 2.32794538e-01 -4.46391225e-01 1.45108283e-01
-1.35534728e+00 2.23639265e-01 -7.21014217e-02 2.46591885e-02
-6.65871143e-01 2.51403600e-01 2.57468045e-01 1.43300429e-01
3.37028027e-01 -4.98289824e-01 -1.43356636e-01 -1.41131312e-01
-7.89992094e-01 2.17435639e-02 2.96197951e-01 6.30465627e-01
1.04553461e+00 1.62286088e-01 -3.37644577e-01 6.32077515e-01
3.38663369e-01 6.13030076e-01 1.73608288e-01 -9.69933569e-01
3.55115473e-01 1.63231328e-01 4.14514750e-01 -8.69687557e-01
-2.05167934e-01 -2.48253183e-03 -8.98727059e-01 7.35385537e-01
1.16124474e-01 -3.28020692e-01 -8.24202716e-01 1.50279295e+00
6.86769009e-01 7.37725139e-01 2.98975527e-01 9.77365136e-01
9.34453428e-01 1.02817822e+00 2.17994928e-01 -9.85603034e-01
1.30487931e+00 -8.94692600e-01 -1.28376234e+00 -2.33811647e-01
-1.15735248e-01 -1.28773832e+00 7.22977340e-01 3.44226331e-01
-1.71320736e+00 -9.15439367e-01 -1.16946459e+00 -2.98908800e-01
-7.25503191e-02 -7.53150210e-02 5.84363282e-01 4.91050690e-01
-9.87137318e-01 1.43020913e-01 -6.21776044e-01 -1.65778436e-02
9.94560346e-02 1.56452954e-01 -1.88434348e-01 -4.79851484e-01
-1.32595694e+00 6.45267129e-01 3.86206329e-01 2.43892118e-01
-8.67406189e-01 -5.45311093e-01 -8.38634968e-01 -1.74451962e-01
4.65325117e-01 -6.80371881e-01 1.11199713e+00 -8.61835003e-01
-1.48114753e+00 5.47477067e-01 -6.74917459e-01 -1.98842674e-01
3.58912736e-01 -1.15826547e-01 -6.05368674e-01 3.35306495e-01
-5.25891930e-02 4.74214435e-01 1.01120722e+00 -1.28675866e+00
-1.15207505e+00 -1.56372953e-02 1.40216455e-01 3.98982912e-01
6.64045513e-02 2.59694219e-01 -6.34066820e-01 -4.86135215e-01
1.75369158e-01 -4.06636834e-01 -7.67260566e-02 1.24758705e-01
2.30967030e-01 7.65836984e-02 1.25929165e+00 -6.14752233e-01
1.22189283e+00 -2.19550371e+00 -4.45108920e-01 -2.52526373e-01
1.62978053e-01 4.26664531e-01 -1.04946584e-01 6.65800720e-02
-1.11354694e-01 -4.15681779e-01 -9.22541395e-02 -2.01771870e-01
-3.94591868e-01 2.20581979e-01 -1.95759267e-01 4.56754953e-01
2.16071621e-01 5.83212554e-01 -1.15667391e+00 -8.43552351e-01
8.26545537e-01 7.83698261e-01 -4.27568763e-01 2.31582567e-01
1.09520443e-01 4.08926100e-01 -1.32146806e-01 7.76546776e-01
1.12530017e+00 1.37384430e-01 -1.83963314e-01 -7.22909987e-01
-5.38892210e-01 -1.18361764e-01 -1.63550448e+00 1.39280474e+00
-4.96183664e-01 7.18401849e-01 1.46726146e-01 -2.28113875e-01
7.40486801e-01 2.73035556e-01 6.44966304e-01 -9.47434664e-01
3.04902315e-01 2.02623606e-01 -3.00880045e-01 -7.43461490e-01
8.36884320e-01 -1.66530073e-01 6.21185422e-01 6.03078485e-01
-2.67166317e-01 4.56370320e-03 6.04533315e-01 1.13918953e-01
8.71369123e-01 3.09549242e-01 3.26600850e-01 3.40510905e-01
9.47976410e-01 -3.88973802e-01 9.21274483e-01 3.31910461e-01
-5.52831471e-01 5.59502184e-01 -6.35266751e-02 -3.09871048e-01
-9.62449312e-01 -9.79952872e-01 3.89762893e-02 8.34360421e-01
9.73984122e-01 -3.45438570e-01 -4.57887322e-01 -1.63620383e-01
-5.21316648e-01 5.65970778e-01 -4.42733057e-02 -2.19664395e-01
-6.63881600e-01 -9.43275928e-01 -7.80710056e-02 1.86816975e-01
1.13434923e+00 -1.00266755e+00 -9.52167928e-01 3.86862099e-01
-5.97089291e-01 -1.44367874e+00 -6.56659186e-01 -3.75987664e-02
-5.41857660e-01 -1.11267006e+00 -4.61521000e-01 -7.62203753e-01
4.69150335e-01 1.06607831e+00 9.12508845e-01 1.85288012e-01
-2.61128813e-01 3.21671396e-01 -3.01398903e-01 -3.24497879e-01
-3.09888244e-01 -8.57375443e-01 -1.85933307e-01 5.41795492e-01
3.20450395e-01 -6.36698544e-01 -1.00976491e+00 3.90434951e-01
-1.29258454e+00 4.03940260e-01 7.84025729e-01 7.48320341e-01
9.04013753e-01 5.03200948e-01 1.47411913e-01 -3.46433252e-01
4.53018934e-01 -1.46188408e-01 -6.61693573e-01 3.48611236e-01
-4.65051651e-01 -3.87718439e-01 4.41141546e-01 -4.60312068e-01
-1.71035182e+00 -1.87720489e-02 2.20769225e-03 -4.72675472e-01
-1.00563444e-01 -8.13837796e-02 -1.67734891e-01 -3.48695129e-01
4.00568724e-01 4.92610455e-01 -2.37828076e-01 -1.02041602e-01
2.01746985e-01 6.42155170e-01 9.64644730e-01 -2.72031069e-01
6.96957171e-01 8.71794105e-01 2.30667323e-01 -6.06218278e-01
-4.94179517e-01 -3.96339178e-01 -2.94212878e-01 -8.16058874e-01
1.01614249e+00 -1.16979492e+00 -8.24114501e-01 6.89175010e-01
-1.36421263e+00 -1.98175713e-01 -3.02905053e-01 6.96958661e-01
-4.49690372e-01 5.97479284e-01 -5.80753684e-01 -8.69948506e-01
-2.28015348e-01 -1.38586545e+00 1.09539938e+00 7.84397483e-01
5.88391840e-01 -6.31760299e-01 -1.24582157e-01 1.70954894e-02
4.48620051e-01 2.62208253e-01 2.77753323e-01 5.95893681e-01
-9.32054460e-01 1.99457198e-01 -5.48780739e-01 3.17917705e-01
2.44204208e-01 1.08362585e-01 -9.90188003e-01 -8.17632973e-02
1.68553829e-01 -7.92494193e-02 6.97878420e-01 5.78931153e-01
8.82640421e-01 3.19543332e-01 -1.24050498e-01 8.59811008e-01
1.64825344e+00 6.55436039e-01 1.01192200e+00 3.70520711e-01
4.06390190e-01 3.21770221e-01 1.07439041e+00 5.48859954e-01
2.68285960e-01 5.11189580e-01 4.36426878e-01 -5.55645227e-01
-6.03558481e-01 -5.64440303e-02 7.77526736e-01 5.02226293e-01
-3.19497585e-01 -2.71333069e-01 -4.39037979e-01 4.22406107e-01
-1.73695254e+00 -1.32347703e+00 -5.81863940e-01 2.12054276e+00
7.86930561e-01 8.86021256e-02 -3.12743098e-01 2.30299592e-01
1.07329869e+00 3.14052254e-01 -6.20420814e-01 -9.24178138e-02
-5.63971102e-01 1.49299473e-01 5.59289873e-01 4.97943789e-01
-7.94934332e-01 5.86397648e-01 5.88685513e+00 7.92536259e-01
-1.00314593e+00 3.92946124e-01 6.15643859e-01 -1.43819615e-01
-1.58572823e-01 2.15154722e-01 -8.91990423e-01 6.38305962e-01
5.14059067e-01 -2.24139884e-01 3.89889240e-01 2.21855104e-01
7.65993714e-01 -7.02329397e-01 -5.57135046e-01 1.37990057e+00
1.24411330e-01 -1.14187527e+00 -1.27894700e-01 -1.86216503e-01
9.41792786e-01 -2.77320981e-01 -2.18384251e-01 -1.48623124e-01
1.24952875e-01 -2.61300087e-01 7.73546040e-01 5.15270710e-01
9.23706651e-01 -8.15685272e-01 4.48910981e-01 1.66924037e-02
-1.62882543e+00 6.44681081e-02 -2.55884707e-01 8.53077695e-02
8.75326872e-01 8.10447037e-01 5.52391484e-02 6.48168564e-01
9.03592408e-01 1.01534712e+00 -3.65100205e-01 1.05686712e+00
-3.90381336e-01 1.91335648e-01 -1.79547951e-01 4.69641894e-01
-3.74589823e-02 -3.46629411e-01 6.63198829e-01 1.04731154e+00
3.01547855e-01 4.60648477e-01 1.92903981e-01 5.96027076e-01
2.28977010e-01 -2.80339748e-01 -3.69169056e-01 3.62697482e-01
2.39815444e-01 1.37075818e+00 -6.68111444e-01 -4.93614584e-01
-8.36324275e-01 1.10208988e+00 -5.84040225e-01 6.33983195e-01
-1.19571924e+00 -7.40193248e-01 5.68901658e-01 -2.25046314e-02
5.33708669e-02 -3.36840302e-01 -1.35978028e-01 -1.22058189e+00
1.63264632e-01 -6.04579866e-01 2.84362793e-01 -1.50866866e+00
-7.30938494e-01 4.78099912e-01 -1.10775247e-01 -1.48148406e+00
-1.95623543e-02 -2.22036079e-01 -6.72128439e-01 7.98855543e-01
-2.40689754e+00 -9.48989868e-01 -9.09743607e-01 1.18043220e+00
7.85348535e-01 2.48658583e-01 -1.57840967e-01 7.76960611e-01
-4.48124945e-01 -3.98347527e-03 -1.22270007e-02 -3.89455169e-01
1.01673055e+00 -6.66493893e-01 -1.49928659e-01 1.63466823e+00
-4.13285941e-01 6.57006586e-03 5.97258985e-01 -7.14736938e-01
-1.28203511e+00 -1.13079715e+00 8.16291630e-01 2.05726832e-01
2.57253975e-01 1.09798960e-01 -7.01028049e-01 4.39792842e-01
4.57433850e-01 2.81537861e-01 2.41606683e-01 -7.64494002e-01
3.37315768e-01 -4.10783499e-01 -1.13557267e+00 6.50348485e-01
1.09229624e+00 -2.44356945e-01 -3.37293297e-01 2.00593188e-01
7.11703718e-01 -5.62871993e-01 -6.03410721e-01 6.00468218e-01
4.25594777e-01 -1.47476637e+00 1.17284119e+00 2.07425073e-01
4.92576838e-01 -9.90953624e-01 -1.37609616e-01 -7.87793815e-01
-1.14583954e-01 -9.64157820e-01 -1.76756218e-01 1.40591300e+00
-3.09199512e-01 -5.20690680e-01 3.54959995e-01 2.89072931e-01
-2.58439243e-01 -3.23508620e-01 -4.76621181e-01 -4.11603689e-01
-9.67992008e-01 -5.20708203e-01 4.87496525e-01 6.45350516e-01
-5.42483926e-01 1.20261557e-01 -5.71557581e-01 2.97773778e-01
9.68051076e-01 2.29858771e-01 8.39500904e-01 -7.97509193e-01
-2.44830206e-01 -1.98291149e-02 -2.28829488e-01 -9.51451957e-01
-2.80194461e-01 -2.28996485e-01 3.66198123e-01 -1.61343038e+00
3.37370247e-01 -2.79740877e-02 -2.93123096e-01 2.89209038e-01
-4.63171899e-01 5.49492002e-01 6.39313012e-02 7.23794773e-02
-7.52274513e-01 5.71742594e-01 1.34941936e+00 1.41171813e-01
-2.89903194e-01 -3.66679102e-01 -2.51000971e-01 7.86230505e-01
4.14467543e-01 -2.38344625e-01 -3.94786954e-01 -4.76452321e-01
1.94764689e-01 2.52380013e-01 5.29035687e-01 -1.16677713e+00
5.97443104e-01 -4.46804732e-01 5.49583018e-01 -9.21819150e-01
3.57954741e-01 -9.46210265e-01 3.53312939e-01 5.11840761e-01
2.16385409e-01 2.35345647e-01 2.68317968e-01 6.37611508e-01
-5.57913244e-01 -1.35267675e-02 1.17821002e+00 -9.97773185e-02
-1.32273555e+00 4.08836484e-01 -4.83442992e-01 -1.71699166e-01
1.38888836e+00 -5.30950129e-01 -6.54261768e-01 -3.21833581e-01
-4.44889396e-01 1.83629394e-01 4.18588579e-01 2.70466954e-01
9.88264203e-01 -1.34113061e+00 -6.81718647e-01 5.08900583e-01
-1.38732612e-01 -4.58019376e-02 8.11694384e-01 1.21399081e+00
-5.93427896e-01 -1.31315097e-01 -4.44489121e-01 -6.68556511e-01
-1.59160340e+00 4.45609808e-01 3.37173611e-01 -3.33099626e-02
-5.36808848e-01 4.70582455e-01 4.11129296e-01 5.79776406e-01
-8.46587420e-02 -1.74916446e-01 -9.97602642e-02 -1.82239622e-01
9.86495972e-01 6.65202558e-01 7.18502386e-04 -6.67428732e-01
-7.27038607e-02 7.85975039e-01 1.09258048e-01 -1.27370894e-01
1.23385000e+00 -8.80894125e-01 -1.43773749e-01 8.15697908e-02
8.75155032e-01 6.64795414e-02 -1.54441535e+00 -2.20576361e-01
-6.21719718e-01 -8.81456137e-01 4.80217874e-01 -4.82984096e-01
-1.16389048e+00 8.55078459e-01 8.91769350e-01 -1.11633301e-01
2.01447272e+00 -5.76649725e-01 1.19684565e+00 -9.37943533e-02
2.38142878e-01 -9.87348795e-01 1.30540296e-01 2.03959435e-01
3.45029950e-01 -1.26940262e+00 2.99300570e-02 -6.31015539e-01
-2.72502422e-01 1.21772385e+00 7.24949896e-01 2.86510438e-01
5.14309406e-01 5.57877004e-01 1.77134182e-02 7.50118792e-02
-9.62189019e-01 -5.93560815e-01 -1.20758004e-01 6.44753873e-01
1.73322111e-01 -7.35449910e-01 -6.79447532e-01 1.45648599e-01
5.90033591e-01 6.23088956e-01 8.00141692e-01 1.11963713e+00
-6.94528222e-01 -1.01909041e+00 -7.33015597e-01 3.46836005e-03
-4.84386593e-01 -1.41305521e-01 2.19490841e-01 5.35598099e-01
6.32096469e-01 1.36239266e+00 -6.56596944e-02 -3.28417808e-01
3.81479144e-01 -2.83218741e-01 4.15488958e-01 -1.41655020e-02
-3.73572767e-01 3.75776470e-01 -2.07934290e-01 -9.08886671e-01
-1.12551069e+00 -6.60574198e-01 -1.17484832e+00 -4.38238651e-01
-2.37908632e-01 -2.12656245e-01 6.83362663e-01 6.66860402e-01
3.13508928e-01 7.77492225e-01 7.31573343e-01 -1.08184409e+00
2.42357031e-01 -4.64503407e-01 -6.52148008e-01 5.94266832e-01
5.23515880e-01 -7.29638457e-01 -4.87353683e-01 6.71496093e-01] | [10.854111671447754, -2.1480278968811035] |
a19de779-8529-4b05-beef-60b480dfb648 | controlvc-zero-shot-voice-conversion-with | 2209.11866 | null | https://arxiv.org/abs/2209.11866v4 | https://arxiv.org/pdf/2209.11866v4.pdf | ControlVC: Zero-Shot Voice Conversion with Time-Varying Controls on Pitch and Speed | Recent developments in neural speech synthesis and vocoding have sparked a renewed interest in voice conversion (VC). Beyond timbre transfer, achieving controllability on para-linguistic parameters such as pitch and Speed is critical in deploying VC systems in many application scenarios. Existing studies, however, either only provide utterance-level global control or lack interpretability on the controls. In this paper, we propose ControlVC, the first neural voice conversion system that achieves time-varying controls on pitch and speed. ControlVC uses pre-trained encoders to compute pitch and linguistic embeddings from the source utterance and speaker embeddings from the target utterance. These embeddings are then concatenated and converted to speech using a vocoder. It achieves speed control through TD-PSOLA pre-processing on the source utterance, and achieves pitch control by manipulating the pitch contour before feeding it to the pitch encoder. Systematic subjective and objective evaluations are conducted to assess the speech quality and controllability. Results show that, on non-parallel and zero-shot conversion tasks, ControlVC significantly outperforms two other self-constructed baselines on speech quality, and it can successfully achieve time-varying pitch and speed control. | ['Zhiyao Duan', 'Meiying Chen'] | 2022-09-23 | null | null | null | null | ['pitch-control'] | ['audio'] | [-3.66787873e-02 -4.26347554e-02 -3.49148005e-01 -2.99435288e-01
-7.89680183e-01 -8.38322520e-01 5.08350551e-01 4.95358333e-02
-1.46952391e-01 3.61910522e-01 5.91213644e-01 -2.83430099e-01
1.59086928e-01 -7.35164940e-01 -3.78074616e-01 -5.53569019e-01
1.24523342e-01 2.65451875e-02 -1.27915610e-02 -3.88833225e-01
-3.01392078e-01 3.99926960e-01 -1.57209063e+00 1.16484321e-03
7.53910840e-01 1.07710135e+00 2.87286192e-01 1.18629408e+00
-8.14977009e-03 4.43565488e-01 -8.94696891e-01 -2.47872338e-01
2.27767751e-02 -7.02563226e-01 -4.78455365e-01 -9.85976458e-02
1.86093435e-01 -2.27408618e-01 -3.75278354e-01 1.04008651e+00
1.07726693e+00 4.21918124e-01 5.80648839e-01 -9.76615965e-01
-1.13602519e+00 8.13347936e-01 4.18754667e-01 2.72397429e-01
3.11405271e-01 2.62954384e-01 1.27369308e+00 -7.75585115e-01
3.52185875e-01 1.49556625e+00 4.02876794e-01 9.14879620e-01
-1.43994164e+00 -7.13907599e-01 -6.52426705e-02 1.57370821e-01
-1.00230443e+00 -9.40430999e-01 7.70055830e-01 -3.64837259e-01
1.26976025e+00 3.93230617e-01 6.84340835e-01 1.07392108e+00
1.51165247e-01 2.73841918e-01 7.12662220e-01 -6.99818909e-01
4.04266745e-01 1.59497723e-01 -4.04228479e-01 2.30317011e-01
-5.75435936e-01 7.18625247e-01 -6.23274863e-01 2.98199594e-01
8.04661036e-01 -8.05471778e-01 -6.35749400e-01 2.20077410e-01
-1.13088024e+00 8.83037567e-01 3.03203881e-01 2.56106764e-01
-2.77152240e-01 3.05206507e-01 6.23405635e-01 7.21316993e-01
2.38288596e-01 9.37821388e-01 -4.57278252e-01 -6.91659391e-01
-7.16554582e-01 5.90053387e-03 7.42293298e-01 9.93884623e-01
3.73493880e-02 8.90498996e-01 -5.30020773e-01 1.17662084e+00
8.37517306e-02 7.91152477e-01 8.63775015e-01 -1.06864130e+00
4.27920669e-01 -1.93471059e-01 -1.86966676e-02 -8.55202556e-01
-1.76933065e-01 -3.04441929e-01 -4.68772650e-01 1.06455833e-01
-1.28765762e-01 -3.33879888e-01 -8.53661060e-01 1.95811892e+00
1.13502152e-01 -1.37885824e-01 4.06006843e-01 1.02409399e+00
7.23726392e-01 1.11010718e+00 -3.04630399e-02 -4.70006615e-01
1.21786535e+00 -1.19767702e+00 -1.58465755e+00 -2.93212850e-03
-2.17654742e-02 -1.05198920e+00 1.77087629e+00 3.05032849e-01
-1.27302480e+00 -9.86684263e-01 -1.33809662e+00 -1.11370064e-01
-2.19422117e-01 1.47104919e-01 3.69214565e-02 9.33547914e-01
-1.08283675e+00 7.21284389e-01 -7.29950547e-01 1.10827491e-01
-2.33440354e-01 3.56160253e-01 7.34498724e-02 6.22362494e-01
-1.57911694e+00 8.95267546e-01 2.09137574e-01 -2.75085002e-01
-7.70637453e-01 -1.02573085e+00 -7.24929512e-01 2.34516203e-01
9.66750532e-02 -2.86975026e-01 1.86221433e+00 -7.54442513e-01
-2.64937925e+00 9.17443782e-02 1.08233221e-01 -3.98889840e-01
2.78951004e-02 -2.47740462e-01 -1.01804364e+00 2.06811458e-01
-1.56168878e-01 9.03961122e-01 1.24118996e+00 -7.89883375e-01
-6.46460712e-01 3.39386821e-01 -2.63531387e-01 2.88585961e-01
-7.39898682e-01 1.67508256e-02 -3.64249438e-01 -8.83514345e-01
-2.01012716e-01 -8.27555180e-01 1.55049652e-01 -1.18167989e-01
-2.51817644e-01 -1.84579432e-01 8.42664301e-01 -6.98866487e-01
1.56813347e+00 -2.25675392e+00 4.30807859e-01 -3.43296647e-01
-2.32488230e-01 5.72014868e-01 -2.21903667e-01 3.31091374e-01
1.31876260e-01 2.33108953e-01 1.32759526e-01 -3.20307106e-01
3.49897981e-01 1.12788767e-01 -5.68191707e-01 1.02375649e-01
4.64046389e-01 6.56503916e-01 -7.89899290e-01 -2.32029140e-01
4.65695322e-01 8.02064180e-01 -1.13368416e+00 6.49367332e-01
-3.76229733e-01 3.32198501e-01 2.18723670e-01 1.80821523e-01
3.45175564e-02 5.18073440e-01 -8.21086392e-03 -2.99370140e-01
-4.20802414e-01 8.04578185e-01 -8.31616580e-01 1.41579390e+00
-9.80022848e-01 8.36911619e-01 2.94225752e-01 -2.77175814e-01
1.05291653e+00 8.77056539e-01 4.00995277e-02 -8.46745789e-01
3.08039933e-01 1.88777030e-01 3.53484273e-01 -2.74043828e-01
5.67984879e-01 -4.48008299e-01 -6.42627152e-03 -4.33221348e-02
3.60502124e-01 -1.06289828e+00 -7.60859251e-02 -4.36808228e-01
5.43884277e-01 -2.82066315e-01 1.14521585e-01 -1.36210144e-01
4.47021514e-01 -3.26224059e-01 4.09469038e-01 4.35690247e-02
-3.47978204e-01 7.01189995e-01 2.47557059e-01 3.96139562e-01
-1.20279229e+00 -1.42050731e+00 -1.33524284e-01 1.29236293e+00
-2.85249501e-01 -4.77454007e-01 -9.68123794e-01 1.47445172e-01
-1.88360512e-01 1.07354903e+00 -2.61271507e-01 -5.58359742e-01
-5.22079408e-01 1.74740091e-01 7.14995086e-01 4.40994114e-01
1.28965408e-01 -1.06232035e+00 -4.47765380e-01 7.02948868e-01
-1.91902384e-01 -1.17695248e+00 -1.24330699e+00 4.68911141e-01
-4.94380653e-01 -1.91132829e-01 -3.85159373e-01 -8.42144966e-01
-5.14682271e-02 -1.48295373e-01 6.65272415e-01 -4.92202908e-01
-3.24199237e-02 2.65988380e-01 -3.87204200e-01 -5.28904259e-01
-7.80463576e-01 1.59990504e-01 7.08617389e-01 -1.68445185e-01
-2.28873521e-01 -6.89132690e-01 -3.58952820e-01 2.31159344e-01
-6.14361107e-01 -1.53090537e-01 7.96386898e-02 7.41076529e-01
7.02629507e-01 1.90108776e-01 1.07143486e+00 -9.64390561e-02
1.26306725e+00 -2.94742212e-02 -6.97976232e-01 -1.54013291e-01
-6.28453612e-01 3.78890298e-02 1.17991507e+00 -8.63568008e-01
-9.77932274e-01 -3.93631458e-01 -4.92842138e-01 -9.27550375e-01
2.13534251e-01 2.34085277e-01 -5.14806688e-01 4.28057253e-01
7.20084190e-01 -5.05138151e-02 8.71194527e-02 -2.43974313e-01
8.74473691e-01 1.18421268e+00 9.79063809e-01 -4.64952588e-01
7.61968970e-01 -3.54892492e-01 -6.45747304e-01 -1.28633511e+00
-5.39371967e-01 -1.85087696e-02 -4.87756193e-01 -1.96288288e-01
1.00203955e+00 -8.38343203e-01 -5.15368342e-01 1.89928293e-01
-1.14531147e+00 -6.01894677e-01 -5.86258650e-01 7.33910263e-01
-9.70313907e-01 -2.58849442e-01 -7.36937582e-01 -7.74715483e-01
-6.47091091e-01 -1.22779024e+00 7.56925166e-01 2.32491225e-01
-5.33127367e-01 -9.69482124e-01 4.97456565e-02 -5.55286696e-03
8.88335168e-01 -1.25679359e-01 1.05154216e+00 -1.38624221e-01
4.63755503e-02 1.85605183e-01 3.68670404e-01 8.81025016e-01
5.93396842e-01 2.83890367e-01 -1.20859516e+00 -3.28672290e-01
7.63714090e-02 -3.03528041e-01 1.48800403e-01 3.88139576e-01
8.60775113e-01 -5.15729189e-01 3.06661487e-01 7.03849256e-01
8.20606172e-01 7.80875087e-01 2.82481641e-01 -1.96697086e-01
4.19936150e-01 5.04620910e-01 4.76067215e-01 2.95224965e-01
2.92322896e-02 9.65059578e-01 1.69184245e-02 -9.15520713e-02
-6.99157834e-01 -5.58600008e-01 7.15336263e-01 1.90931177e+00
2.70192444e-01 -3.06956828e-01 -3.94235939e-01 4.29288983e-01
-1.03105283e+00 -8.16392958e-01 4.03053701e-01 1.93898225e+00
1.48136580e+00 1.63627133e-01 1.13471961e-02 4.00483727e-01
6.94113314e-01 1.95154101e-01 -6.76313281e-01 -1.13453424e+00
2.13958114e-01 5.63009083e-01 1.23493327e-02 7.73796856e-01
-9.13483083e-01 1.36265159e+00 6.63335371e+00 8.13544393e-01
-1.66587448e+00 1.73272565e-01 1.94181010e-01 -4.37779963e-01
-2.47472003e-01 -6.11662269e-01 -6.56446934e-01 1.80611581e-01
1.62379408e+00 -5.78911722e-01 1.03386688e+00 8.33657086e-01
7.03849852e-01 7.92851746e-01 -1.28891087e+00 8.86255920e-01
-2.65702307e-01 -1.13432360e+00 3.12803239e-02 -1.47753894e-01
7.08213568e-01 -3.35344017e-01 5.33222854e-01 6.64623737e-01
-6.92698918e-03 -1.19881248e+00 1.17787051e+00 1.52603403e-01
1.50453317e+00 -1.06883132e+00 1.55072317e-01 -1.71274040e-02
-1.43220818e+00 -9.74318311e-02 -3.19935501e-01 -4.18677032e-02
2.61844844e-01 1.08170696e-01 -9.71193790e-01 -5.22459708e-02
3.30997378e-01 2.67533213e-01 9.55257192e-02 4.10528421e-01
-4.59074378e-01 1.01211309e+00 6.17835335e-02 -2.50989407e-01
1.47872239e-01 4.66783904e-02 4.57575113e-01 1.21612310e+00
3.37015688e-01 2.52143860e-01 -1.00192711e-01 9.75253463e-01
-2.07673788e-01 2.57540911e-01 -2.01366320e-01 -7.04937339e-01
1.05673575e+00 8.64455998e-01 -6.97432756e-02 -1.26957685e-01
-2.24801332e-01 9.11051631e-01 1.17871433e-01 3.22024256e-01
-1.09090281e+00 -8.99966061e-01 1.37433398e+00 -2.19258055e-01
4.36880291e-01 -4.54387337e-01 -4.40028161e-02 -7.09496498e-01
-3.02789569e-01 -1.04256642e+00 -3.82276416e-01 -6.43654943e-01
-8.84408414e-01 9.54710066e-01 -2.47573674e-01 -1.07931852e+00
-6.17767096e-01 -5.64008236e-01 -5.61449170e-01 1.00772357e+00
-1.37990224e+00 -4.91267681e-01 1.18107602e-01 4.22445416e-01
9.49377775e-01 -3.14545602e-01 1.17511415e+00 2.17619061e-01
-4.91682529e-01 9.53049421e-01 -8.69531557e-02 -2.26029903e-01
6.69455290e-01 -1.24648178e+00 6.54895782e-01 6.35052085e-01
-6.44681156e-02 5.07926583e-01 8.32140326e-01 -1.50585353e-01
-1.54768825e+00 -1.34627581e+00 7.54979849e-01 -1.43547788e-01
6.22288346e-01 -5.60036361e-01 -8.35381687e-01 7.59750828e-02
4.78900284e-01 -6.59358874e-02 7.17774868e-01 -3.04967731e-01
-2.05766439e-01 -2.20636770e-01 -9.29950356e-01 7.31157601e-01
8.08038354e-01 -9.91024911e-01 -7.95668185e-01 -1.39391571e-01
1.81886899e+00 -4.89735395e-01 -1.17867279e+00 5.57162426e-02
3.60772967e-01 -6.99045360e-01 9.89757955e-01 -4.66383785e-01
3.36596489e-01 -1.68636262e-01 -6.49213016e-01 -2.11799955e+00
-2.46492058e-01 -9.36476290e-01 -3.80329341e-02 1.51611328e+00
4.99561250e-01 -3.64788115e-01 6.01449832e-02 1.76316544e-01
-6.59779608e-01 -4.51967061e-01 -9.07774687e-01 -1.06520891e+00
4.70039368e-01 -6.45097554e-01 7.36665130e-01 6.76248312e-01
1.98317379e-01 5.96098304e-01 -2.55860478e-01 3.44441622e-01
4.95931581e-02 -2.99895167e-01 5.28742969e-01 -7.19164133e-01
-3.03262472e-01 -6.17668033e-01 6.10257126e-02 -9.92003858e-01
6.67075887e-02 -7.76006818e-01 4.77544218e-01 -1.21993387e+00
-8.30249786e-01 -9.62647796e-02 -1.64438590e-01 3.13004375e-01
5.43968230e-02 -1.84168935e-01 5.79408228e-01 -1.89055845e-01
2.81718612e-01 1.18820763e+00 1.47649419e+00 -9.35150012e-02
-7.01147437e-01 -7.39771202e-02 -3.45533937e-01 3.21992874e-01
9.46986914e-01 -1.60940975e-01 -8.79274189e-01 -3.37094158e-01
-5.39190233e-01 5.05305171e-01 -2.75716573e-01 -1.30375469e+00
1.21531606e-01 -2.88882256e-01 -1.25588864e-01 -1.68314055e-01
7.69490600e-01 -4.70859200e-01 -1.44128457e-01 3.87340873e-01
-6.50457263e-01 5.69760352e-02 4.73154336e-01 2.51138091e-01
-3.91138345e-01 8.36039931e-02 1.12576997e+00 4.64633554e-01
-2.71664292e-01 1.15925096e-01 -6.51764035e-01 2.40619183e-01
6.04851007e-01 1.79652512e-01 -1.62293501e-02 -5.48889160e-01
-7.66802371e-01 -8.23984519e-02 4.30853106e-03 8.92743945e-01
6.35917366e-01 -1.73447812e+00 -5.10611713e-01 5.05922139e-01
-1.17300071e-01 -3.99094939e-01 -1.63926259e-02 3.28038126e-01
-2.02831864e-01 7.32584655e-01 -1.42091468e-01 -5.24078846e-01
-1.04427814e+00 5.29536545e-01 5.78658164e-01 4.98562187e-01
-4.63570595e-01 8.71023059e-01 -1.61008090e-01 -6.78916633e-01
6.56521440e-01 -8.73915613e-01 -5.88973090e-02 1.99508607e-01
5.04583597e-01 2.25940034e-01 3.60825695e-02 -4.93891150e-01
-1.14196829e-01 4.26977843e-01 4.98897344e-01 -7.31386244e-01
9.85153079e-01 -1.25909105e-01 3.19321781e-01 6.60775185e-01
1.43943179e+00 2.10207835e-01 -1.39684439e+00 4.55801524e-02
-3.12872410e-01 -2.36944228e-01 5.91775596e-01 -8.96849930e-01
-1.08186138e+00 1.08954716e+00 6.71620905e-01 1.77223086e-01
1.19041145e+00 -2.77181804e-01 8.74569654e-01 2.27484390e-01
6.88739046e-02 -1.38648963e+00 4.93498266e-01 8.25171292e-01
1.29226482e+00 -6.27779782e-01 -7.26206005e-01 -2.44211569e-01
-8.40774715e-01 1.12960827e+00 6.35528028e-01 1.17526159e-01
6.14767492e-01 5.98611236e-01 4.41935480e-01 3.94921303e-01
-1.15034747e+00 -1.21224150e-02 3.08885783e-01 7.48097301e-01
7.05523729e-01 4.15024191e-01 3.04631945e-02 7.55785882e-01
-1.21203148e+00 -4.35570568e-01 4.64718431e-01 2.69289583e-01
-5.71553648e-01 -1.02471662e+00 -4.76394624e-01 4.42765728e-02
-3.40158343e-01 -1.58695593e-01 -2.32765168e-01 5.57915866e-01
-5.18223317e-03 1.42599499e+00 3.22728634e-01 -8.29351425e-01
9.19264257e-01 1.85329586e-01 2.82031059e-01 -7.42303193e-01
-7.08015025e-01 4.27869678e-01 2.50099540e-01 -3.17249775e-01
2.15699494e-01 -4.73302066e-01 -1.64466751e+00 -1.76781207e-01
-5.41736901e-01 3.30298245e-01 8.69404376e-01 5.08787811e-01
2.31337234e-01 1.32831037e+00 1.05857503e+00 -8.41985166e-01
-9.57075119e-01 -9.97122705e-01 -4.10723299e-01 2.65965471e-03
7.03627825e-01 -6.27087593e-01 -6.02242649e-01 1.65910095e-01] | [15.037619590759277, 6.518710613250732] |
3d4761be-bdcf-4158-9479-cfed3d14086a | clood-cbr-towards-microservices-oriented-case | null | null | https://rgu-repository.worktribe.com/output/895530/clood-cbr-towards-microservices-oriented-case-based-reasoning | https://rgu-repository.worktribe.com/output/895530/clood-cbr-towards-microservices-oriented-case-based-reasoning | Clood CBR: towards microservices oriented case-based reasoning | CBR applications have been deployed in a wide range of sectors, from pharmaceuticals; to defence and aerospace to IoT and transportation, to poetry and music generation; for example. However, a majority of these have been built using monolithic architectures which impose size and complexity constraints. As such these applications have a barrier to adopting new technologies and remain prohibitively expensive in both time and cost because changes in frameworks or languages affect the application directly. To address this challenge, we introduce a distributed and highly scalable generic CBR system, CLOOD, which is based on a microservices architecture. This splits the application into a set of smaller, interconnected services that scale to meet varying demands. Experimental results show that our CLOOD implementation retrieves cases at a fairly consistent rate as the casebase grows by several orders of magnitude and was over 3,700 times faster than a comparable monolithic CBR system when retrieving from half a million cases. Microservices are cloud-native architectures and with the rapid increase in cloud-computing adoption, it is timely for the CBR community to have access to such a framework. | ['David Corsar', 'Juan A. Recio-García', 'Chamath Palihawadana', 'Nirmalie Wiratunga', 'Ikechukwu Nkisi-Orji'] | 2020-10-03 | null | null | null | international-conference-on-case-based | ['music-generation', 'music-generation'] | ['audio', 'music'] | [-6.83312178e-01 -3.56432855e-01 -7.09870178e-03 -1.89852849e-01
-6.45038366e-01 -1.01884389e+00 7.59830058e-01 -8.06034803e-02
-3.69800702e-02 5.06947696e-01 2.09100142e-01 -5.98462462e-01
-3.17248791e-01 -7.91483164e-01 -1.64289847e-01 -3.79711688e-01
1.60643473e-01 6.72704041e-01 5.22963703e-01 -5.00615180e-01
3.59800309e-01 9.02669549e-01 -2.00248814e+00 4.05042619e-01
6.54925466e-01 1.02358174e+00 4.84021418e-02 3.40005636e-01
-4.25023228e-01 7.62604654e-01 -7.00249612e-01 -2.24057794e-01
2.16725186e-01 8.23644102e-02 -8.94212961e-01 -3.31086934e-01
-5.79769127e-02 -1.55790389e-01 2.35749125e-01 3.60147625e-01
7.16015339e-01 5.87437376e-02 -7.37736773e-05 -1.55335975e+00
-5.50397992e-01 3.90448153e-01 -4.39211130e-01 2.93481410e-01
5.38877785e-01 -1.04955487e-01 5.94026327e-01 -7.43603766e-01
1.12409818e+00 7.62564838e-01 5.94895065e-01 4.46878254e-01
-9.83306766e-01 -3.40975016e-01 2.94536371e-02 1.39506459e-01
-1.45867145e+00 -8.37261319e-01 8.91686529e-02 -2.36782044e-01
1.77993214e+00 8.44798028e-01 7.22760499e-01 7.35529661e-01
2.88555920e-01 -1.33557871e-01 8.56100976e-01 -2.40692943e-01
5.09316802e-01 4.09290701e-01 -2.87547767e-01 -3.48070487e-02
6.75852120e-01 -4.31838214e-01 -2.46478066e-01 -5.12525916e-01
3.57911617e-01 1.90894738e-01 1.61515892e-01 -3.87422800e-01
-9.08141792e-01 2.02263892e-01 1.37412339e-01 6.41424656e-01
-5.71020722e-01 1.22651733e-01 5.30857742e-01 3.61421347e-01
1.22991882e-01 3.63059670e-01 -5.41629016e-01 -6.38453662e-01
-6.63236618e-01 1.02869064e-01 1.41122508e+00 1.31970811e+00
4.65795636e-01 3.83253396e-03 6.34684861e-01 9.41428125e-01
1.63866282e-01 3.06718946e-01 5.82115531e-01 -1.02700317e+00
8.77213851e-03 8.81685138e-01 8.88886005e-02 -8.37645710e-01
-2.41073132e-01 -1.19106777e-01 -3.10915351e-01 2.54797935e-01
-2.49327004e-01 3.32704961e-01 -6.30157113e-01 1.02130651e+00
7.51557112e-01 -3.26188356e-01 1.52352098e-02 9.68342304e-01
7.37740934e-01 5.41423857e-01 -2.11180071e-03 -3.20216119e-02
1.23438251e+00 -7.63511181e-01 -1.35119736e-01 1.19189825e-02
1.85636863e-01 -1.17868626e+00 1.01526761e+00 4.02540714e-01
-1.28513837e+00 1.54737726e-01 -8.71511936e-01 -9.61166713e-03
-6.56485081e-01 -6.51254594e-01 6.56378925e-01 4.48047996e-01
-1.42163324e+00 -2.76692361e-02 -1.02358854e+00 -1.22462082e+00
8.24026880e-04 1.97527841e-01 -2.87959605e-01 -3.37396525e-02
-2.71807432e-01 1.18981886e+00 -2.53809363e-01 -3.23097199e-01
-4.26027030e-01 -5.48145175e-01 -1.10230416e-01 2.27790743e-01
1.12803824e-01 -9.17616308e-01 1.19937813e+00 -6.84440851e-01
-1.40408504e+00 6.57461226e-01 2.49324888e-01 7.32143819e-02
2.56989866e-01 3.42778713e-01 -7.95153320e-01 -1.15276948e-01
3.30291539e-01 4.05400842e-02 2.01260880e-01 -1.01954198e+00
-8.35048437e-01 -4.16066557e-01 3.79579991e-01 2.04862833e-01
-3.82453322e-01 5.63489377e-01 -4.30753171e-01 2.00057819e-01
-1.40593052e-01 -8.95732164e-01 -2.45295614e-01 -1.91627800e-01
2.21302986e-01 1.27661834e-02 8.47833514e-01 -1.64954394e-01
1.03770006e+00 -2.06319547e+00 8.50308388e-02 1.46985888e-01
1.16720289e-01 -1.30521223e-01 -8.50879028e-02 1.25232661e+00
3.55730176e-01 3.96976203e-01 3.46419424e-01 5.19739568e-01
-4.79996577e-02 1.34645358e-01 -2.15916708e-01 1.41751513e-01
-1.14256486e-01 1.88886151e-01 -6.38687193e-01 -2.46515468e-01
1.58355664e-02 7.11181998e-01 -5.25498152e-01 -5.59081174e-02
-1.18806340e-01 1.27944976e-01 -4.10971552e-01 1.02564573e+00
6.58304155e-01 -6.67353094e-01 6.72662735e-01 3.64610165e-01
-4.85786885e-01 5.41668594e-01 -1.11986756e+00 1.54329443e+00
-1.03606629e+00 1.76399648e-01 5.20952702e-01 -7.07307577e-01
6.97642863e-01 4.84824389e-01 8.33561659e-01 -7.52608359e-01
-2.18806788e-01 5.68864942e-01 -1.73138708e-01 -5.39606452e-01
6.18631721e-01 -4.81758602e-02 -1.58792213e-01 6.23494208e-01
-3.89007539e-01 -1.26271367e-01 3.69452327e-01 2.61785150e-01
1.66759193e+00 1.63344845e-01 3.27668279e-01 -2.49446705e-01
1.68932319e-01 4.60694164e-01 6.30453587e-01 1.28718227e-01
1.62958682e-01 2.40280330e-01 1.55611470e-01 -7.38029301e-01
-9.15584743e-01 -1.14678526e+00 -2.71720767e-01 1.15468228e+00
7.83097968e-02 -5.93002737e-01 -4.17018235e-01 -1.99022174e-01
1.35484099e-01 7.47956276e-01 1.20821871e-01 2.22958460e-01
-3.78304750e-01 -4.28714067e-01 1.67953312e-01 3.23300213e-01
3.06415260e-01 -9.81652021e-01 -1.12960267e+00 4.71654326e-01
1.33532763e-01 -1.07407463e+00 -9.38910395e-02 -2.32567117e-01
-7.46318638e-01 -9.47640479e-01 -1.82803601e-01 -4.44316149e-01
3.06873143e-01 8.25755715e-01 1.48659635e+00 3.62007171e-01
-8.34191561e-01 6.96168482e-01 -3.21255267e-01 -4.49319422e-01
-3.92065287e-01 1.33660838e-01 -1.09213017e-01 -5.63219130e-01
4.88460094e-01 -1.10846412e+00 -7.71240354e-01 4.94279891e-01
-8.36151779e-01 -1.52835473e-01 4.49036419e-01 1.65734038e-01
3.97866696e-01 -4.38528270e-01 1.00469625e+00 -6.86382413e-01
7.72643209e-01 -1.10968447e+00 -7.91488171e-01 1.37700945e-01
-1.29375470e+00 -5.68111360e-01 5.96673787e-01 -9.15147588e-02
-9.41230118e-01 -1.46724537e-01 -3.47647853e-02 -1.91970378e-01
-1.70548379e-01 6.34845614e-01 1.40649244e-01 1.31650731e-01
8.42606783e-01 5.41547574e-02 1.55562341e-01 -4.91490543e-01
4.48973089e-01 1.23906255e+00 8.69834051e-02 -4.95537549e-01
2.28265703e-01 6.23982310e-01 -3.92930359e-01 -6.43362105e-01
3.05367947e-01 -6.88415468e-01 2.20808730e-01 -3.06693703e-01
2.53630102e-01 -8.17879260e-01 -7.59450793e-01 -2.18879938e-01
-1.02880549e+00 -7.58079439e-02 -3.50132495e-01 6.49442673e-02
-3.06210876e-01 -2.19214484e-01 -5.66491246e-01 -6.60193920e-01
-6.71196520e-01 -9.99459863e-01 1.13248289e+00 2.08399042e-01
-2.89517760e-01 -5.27051747e-01 3.89142096e-01 3.83851737e-01
1.28532207e+00 3.34116548e-01 8.04980874e-01 -2.71228850e-01
-9.22667980e-01 -4.51780230e-01 -2.28437543e-01 -1.72133461e-01
2.28832260e-01 3.84326488e-01 -7.03869939e-01 -4.98063326e-01
-2.52395660e-01 9.85828713e-02 -2.59871125e-01 -2.99561590e-01
6.04087174e-01 -2.22023517e-01 -6.33788228e-01 3.37125659e-01
1.74312198e+00 3.52517813e-01 8.71214092e-01 8.36397469e-01
2.30639204e-01 4.58669633e-01 4.30573612e-01 5.57328343e-01
5.62524259e-01 7.45292425e-01 4.10607666e-01 2.00676203e-01
-9.52008590e-02 4.04407889e-01 2.45326683e-01 1.37778318e+00
-5.28340757e-01 8.39158744e-02 -1.26446128e+00 6.89696312e-01
-1.91259122e+00 -1.07976758e+00 1.03528751e-02 2.26585197e+00
6.03305459e-01 -2.29192197e-01 3.51960272e-01 -3.59901845e-01
5.72801232e-01 -3.64152670e-01 -7.52385616e-01 -9.49480712e-01
3.70422304e-01 4.76226956e-02 1.09506778e-01 -2.89069504e-01
4.93716970e-02 5.83088577e-01 6.73348188e+00 2.37552196e-01
-1.59370458e+00 3.79680961e-01 1.03987105e-01 -4.36509669e-01
-4.37030256e-01 1.99162498e-01 -3.66674870e-01 5.33885300e-01
1.47252810e+00 -8.32717419e-01 8.71179402e-01 1.37612259e+00
8.28711018e-02 -3.76818366e-02 -7.32224464e-01 9.35157895e-01
-1.39867321e-01 -1.72880292e+00 -2.01526046e-01 4.45940048e-01
5.52737832e-01 8.05844665e-01 -4.43405807e-01 1.50348082e-01
4.65287149e-01 -6.67215705e-01 7.56377041e-01 4.64787632e-01
8.28601599e-01 -7.18652010e-01 3.56662869e-01 1.83670104e-01
-1.23975050e+00 -1.44153818e-01 -3.15835923e-01 -1.15280375e-01
6.23846538e-02 4.66822207e-01 -7.44426966e-01 5.03201783e-01
1.54005241e+00 2.06376128e-02 -1.83390602e-01 1.24687147e+00
6.43129408e-01 1.99553147e-01 -3.58594537e-01 -1.91301078e-01
-2.70789444e-01 -2.84801334e-01 2.88002133e-01 1.11223876e+00
6.87811732e-01 -1.02296583e-01 -1.47774890e-01 4.34184939e-01
-9.36933681e-02 3.98851275e-01 -7.94986665e-01 7.01289773e-02
9.12577033e-01 1.78977954e+00 -9.27694380e-01 -2.23608777e-01
-7.34857202e-01 5.97593546e-01 3.44942689e-01 2.50448138e-01
-7.96114922e-01 -6.74129844e-01 1.01112688e+00 5.00858307e-01
2.71037340e-01 -1.05440132e-01 -1.25424132e-01 -1.05633593e+00
3.32866222e-01 -1.27908421e+00 2.80461371e-01 -8.91115665e-01
-1.51784503e+00 1.17770875e+00 -2.60965556e-01 -1.32149887e+00
-5.02361059e-01 -1.93090364e-01 -2.44499817e-01 5.73553860e-01
-1.16308737e+00 -9.92613912e-01 -4.92939353e-01 3.81383747e-01
2.46478021e-01 -2.39760637e-01 1.12620926e+00 6.55814528e-01
-3.76356423e-01 -4.52367738e-02 4.32862967e-01 -8.64256740e-01
7.37155557e-01 -8.73121500e-01 4.08238053e-01 3.37095022e-01
-8.84083956e-02 1.02196372e+00 3.97864193e-01 -2.70944983e-01
-2.12295818e+00 -7.30482817e-01 7.52808750e-01 -5.95485747e-01
6.97588921e-01 -4.24634516e-01 -4.54026014e-01 6.07425213e-01
4.56446201e-01 2.32014179e-01 7.26101100e-01 3.00315410e-01
-5.00972033e-01 -8.37843359e-01 -1.47228646e+00 5.69491029e-01
1.13522494e+00 -5.24078488e-01 -1.32223815e-01 5.90843916e-01
3.63865137e-01 -2.92184711e-01 -1.16564214e+00 2.77733034e-03
6.90361798e-01 -1.09324205e+00 7.73244798e-01 -4.76806998e-01
2.93319523e-01 -3.92493963e-01 -3.43022794e-01 -1.02492273e+00
-3.87212366e-01 -7.17390180e-01 5.96295521e-02 1.36278653e+00
3.84627402e-01 -1.13865101e+00 3.86179388e-01 1.15521741e+00
-2.00979397e-01 -8.47409904e-01 -1.05865943e+00 -1.07385910e+00
-1.26701444e-01 -1.82051092e-01 1.35446393e+00 1.18242347e+00
4.95468140e-01 4.80579615e-01 3.87478679e-01 -4.24564667e-02
7.91890100e-02 5.31635404e-01 8.92373681e-01 -1.15373158e+00
-4.61260587e-01 -3.16760540e-01 -6.80857182e-01 -4.33768898e-01
-6.93550169e-01 -9.93142962e-01 -2.69428641e-01 -2.11984181e+00
3.16605717e-01 -8.69858027e-01 -2.20962495e-01 4.75967884e-01
6.61168754e-01 2.33190775e-01 7.18769506e-02 6.43637717e-01
-7.17038393e-01 -1.93526298e-01 5.69610476e-01 1.09576933e-01
-1.59120500e-01 -2.27914751e-01 -1.07225311e+00 4.13500935e-01
6.18367076e-01 -3.63294274e-01 -4.65870351e-01 -7.20091105e-01
5.96653759e-01 1.37431249e-01 -5.64296581e-02 -9.79722619e-01
4.73409474e-01 -4.65042174e-01 -3.94368023e-02 2.11535748e-02
2.45042920e-01 -1.16548550e+00 1.10142839e+00 1.05520256e-01
3.15702468e-01 6.46827579e-01 2.09308937e-01 3.03492337e-01
1.06545184e-02 1.83287814e-01 2.55180508e-01 -2.61160526e-02
-4.33815897e-01 -8.84176269e-02 -7.58628666e-01 -6.58422261e-02
1.44923472e+00 -1.09227605e-01 -9.17924285e-01 -2.47581482e-01
-3.42711836e-01 -3.82350422e-02 1.08245981e+00 6.43023133e-01
3.74872804e-01 -1.07926393e+00 -3.10193896e-01 -1.30021095e-01
3.01366806e-01 -2.57932723e-01 -8.63868967e-02 8.77315879e-01
-8.64779651e-01 3.23237270e-01 -3.42887074e-01 -5.48619032e-01
-1.16127849e+00 7.55732298e-01 2.41827369e-01 2.37224847e-01
-7.48394668e-01 1.11908831e-01 -4.90717292e-01 -5.91639042e-01
-1.82414889e-01 -5.33083640e-03 3.74650657e-01 -2.64422566e-01
5.60980558e-01 7.39560604e-01 6.52462184e-01 -4.32402045e-01
-8.36712837e-01 2.83280045e-01 2.13475209e-02 -1.47681504e-01
1.83496428e+00 -1.50094256e-01 -7.20566690e-01 4.46590602e-01
7.84966528e-01 3.47849607e-01 -5.37443638e-01 6.29398704e-01
9.70997885e-02 -5.37821651e-01 -1.84452280e-01 -1.07263446e+00
-1.03048337e+00 2.54820347e-01 4.11495656e-01 8.49326551e-01
1.05439460e+00 3.13091785e-01 6.22132301e-01 2.54801929e-01
1.22475338e+00 -1.19300401e+00 -5.35052776e-01 -1.44787237e-01
1.05826116e+00 -4.16962117e-01 2.01761469e-01 -3.70975405e-01
-2.88064003e-01 1.10008836e+00 3.86697918e-01 -4.77302074e-03
6.03659928e-01 7.15795100e-01 2.81485736e-01 -3.84741664e-01
-1.28653669e+00 1.17284119e-01 -5.87335825e-01 5.40077031e-01
5.75963080e-01 1.21157125e-01 -6.92277670e-01 3.51223886e-01
2.23558508e-02 3.91389251e-01 8.01075697e-01 1.37813258e+00
-9.48344022e-02 -1.48659718e+00 -2.56729722e-01 6.20307386e-01
-5.24858773e-01 6.35836199e-02 -2.64450073e-01 7.20324934e-01
-2.30911121e-01 1.20188069e+00 2.70375252e-01 -1.86497033e-01
6.31317139e-01 -1.34060279e-01 1.77176386e-01 -5.58282614e-01
-9.38396215e-01 -5.05727828e-02 5.90131164e-01 -8.72540295e-01
-1.57296523e-01 -7.42294788e-01 -1.45923972e+00 -7.88677394e-01
-2.35584781e-01 2.09198907e-01 1.34924245e+00 3.97205293e-01
1.46329951e+00 4.62691218e-01 5.13894498e-01 -5.45401812e-01
-6.81874037e-01 -5.00240862e-01 -4.44143802e-01 1.08938716e-01
-5.55610359e-01 -4.17645276e-01 -1.07414588e-01 2.96957530e-02] | [8.829940795898438, 7.371221542358398] |
aafcdd78-7422-4f53-9000-e67a92246166 | improving-the-diagnosis-of-breast-cancer | 2207.06560 | null | https://arxiv.org/abs/2207.06560v1 | https://arxiv.org/pdf/2207.06560v1.pdf | Improving the diagnosis of breast cancer based on biophysical ultrasound features utilizing machine learning | The improved diagnostic accuracy of ultrasound breast examinations remains an important goal. In this study, we propose a biophysical feature based machine learning method for breast cancer detection to improve the performance beyond a benchmark deep learning algorithm and to furthermore provide a color overlay visual map of the probability of malignancy within a lesion. This overall framework is termed disease specific imaging. Previously, 150 breast lesions were segmented and classified utilizing a modified fully convolutional network and a modified GoogLeNet, respectively. In this study multiparametric analysis was performed within the contoured lesions. Features were extracted from ultrasound radiofrequency, envelope, and log compressed data based on biophysical and morphological models. The support vector machine with a Gaussian kernel constructed a nonlinear hyperplane, and we calculated the distance between the hyperplane and data point of each feature in multiparametric space. The distance can quantitatively assess a lesion, and suggest the probability of malignancy that is color coded and overlaid onto B mode images. Training and evaluation were performed on in vivo patient data. The overall accuracy for the most common types and sizes of breast lesions in our study exceeded 98.0% for classification and 0.98 for an area under the receiver operating characteristic curve, which is more precise than the performance of radiologists and a deep learning system. Further, the correlation between the probability and BI RADS enables a quantitative guideline to predict breast cancer. Therefore, we anticipate that the proposed framework can help radiologists achieve more accurate and convenient breast cancer classification and detection. | ['Kevin J. Parker', "Avice M. O'Connell", 'Jihye Baek'] | 2022-07-13 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 1.74251392e-01 7.74064809e-02 -2.19078690e-01 -3.27743471e-01
-7.48896360e-01 -2.84504354e-01 4.13032509e-02 5.02566516e-01
-9.42870453e-02 3.02353799e-01 -2.26876512e-01 -7.37467349e-01
-5.03662527e-01 -7.92769670e-01 -4.42683041e-01 -1.11749315e+00
-5.14762640e-01 2.07797006e-01 1.25139728e-01 3.00333649e-01
1.96392220e-02 8.51199985e-01 -8.58514845e-01 5.96495926e-01
1.06005311e+00 1.51232922e+00 1.80843621e-01 8.56118500e-01
6.60952106e-02 4.73248512e-01 -3.30465376e-01 -1.21875368e-01
-1.14285424e-01 -3.63936543e-01 -2.56023705e-01 -3.46712679e-01
1.19950578e-01 -3.55923623e-01 -4.02408898e-01 9.19105053e-01
6.34489834e-01 -4.54975933e-01 1.23224545e+00 -7.86744118e-01
-7.50252426e-01 3.26096982e-01 -6.54498518e-01 2.83684492e-01
1.85487028e-02 -1.36767611e-01 1.54553249e-01 -5.88055968e-01
2.41368741e-01 7.82269597e-01 9.85261500e-01 3.06079686e-01
-7.59014487e-01 -7.22425818e-01 -7.09157944e-01 1.12129986e-01
-1.27214289e+00 8.09414536e-02 6.34884775e-01 -6.82157934e-01
3.55555892e-01 7.66647220e-01 1.07376385e+00 4.58067000e-01
1.30172324e+00 5.23700953e-01 1.19455695e+00 -4.97074515e-01
1.03371227e-02 2.68766671e-01 7.12084994e-02 1.34978878e+00
5.44415951e-01 2.44750351e-01 1.01915173e-01 -4.61115241e-01
1.02242589e+00 1.79508924e-01 -3.72614115e-01 -6.45240307e-01
-9.40341413e-01 8.00432801e-01 7.45821178e-01 6.31209314e-01
-2.48979449e-01 7.25340992e-02 2.62098819e-01 -4.14003998e-01
6.20544516e-02 2.61747479e-01 -3.73797007e-02 2.01319128e-01
-7.71145225e-01 -3.10064346e-01 7.63217628e-01 2.40879297e-01
2.97484875e-01 -2.22423792e-01 -2.90155143e-01 6.04713678e-01
4.08383608e-01 7.13681161e-01 8.44014466e-01 -6.48521185e-01
-3.13328356e-01 5.94789207e-01 -7.51026645e-02 -1.31726038e+00
-9.74702716e-01 -5.65692246e-01 -1.25988722e+00 8.18877891e-02
4.09211159e-01 -4.09729034e-02 -1.20560849e+00 8.97313297e-01
2.63307452e-01 1.24521613e-01 6.88511133e-02 1.05204046e+00
1.07943642e+00 4.36719000e-01 2.76215672e-01 -3.16849470e-01
1.64668596e+00 -3.84990513e-01 -7.58720994e-01 2.59725183e-01
1.03481221e+00 -3.96507740e-01 6.26137078e-01 1.72055796e-01
-8.03379714e-01 -3.78576189e-01 -1.25068307e+00 4.14126158e-01
-1.09171532e-01 5.42314112e-01 1.10976291e+00 1.00413442e+00
-8.83523643e-01 5.06411552e-01 -1.25543952e+00 -1.99857339e-01
6.60892963e-01 2.37194568e-01 -4.84867871e-01 -1.37509242e-01
-1.23021197e+00 9.47035253e-01 3.33819121e-01 2.30880365e-01
-5.81772685e-01 -9.08886731e-01 -9.80880439e-01 1.33836204e-02
-4.18524295e-01 -3.81567389e-01 8.97554517e-01 -8.12057614e-01
-1.19322348e+00 9.50152159e-01 1.71643645e-01 -5.04221857e-01
4.37165618e-01 1.51567563e-01 -6.66951895e-01 5.89414001e-01
-8.96325801e-03 3.03598970e-01 3.29027116e-01 -1.26845443e+00
-5.38610637e-01 -3.79286796e-01 -5.16143620e-01 -1.48093060e-01
-4.06554908e-01 -2.14272574e-01 -3.08976799e-01 -3.24580699e-01
5.87634146e-01 -7.36845613e-01 -2.51772553e-01 4.00048465e-01
-1.89569473e-01 4.47332002e-02 7.30025828e-01 -9.42256629e-01
1.14759791e+00 -2.19214249e+00 -4.65523511e-01 6.24565721e-01
3.49283457e-01 1.34100556e-01 3.27187002e-01 -3.74862939e-01
-9.98950154e-02 2.66136974e-01 -1.76624060e-01 7.71049559e-01
-4.42505985e-01 -3.72806430e-01 3.98158908e-01 9.36368644e-01
1.30003586e-01 1.20162201e+00 -7.25401938e-01 -9.37340915e-01
2.17592657e-01 6.53369844e-01 -1.09481916e-01 3.19913141e-02
4.03083533e-01 1.78120643e-01 -5.93761563e-01 9.25003767e-01
9.06485915e-01 -5.67084491e-01 3.01051408e-01 -6.88701808e-01
1.06122576e-01 -6.11279905e-01 -6.61545992e-01 1.32126701e+00
-3.79160076e-01 6.14083409e-01 -1.95131943e-01 -9.57384706e-01
1.16573095e+00 2.13099167e-01 9.42346096e-01 -8.22539926e-01
1.55205801e-01 3.98747772e-01 3.57535541e-01 -8.15627098e-01
-1.73775196e-01 -1.78692847e-01 1.98401406e-01 1.70235604e-01
-1.39372051e-01 -9.18838605e-02 -1.93915606e-01 -1.05279274e-01
9.43629503e-01 -1.20877311e-01 4.25223172e-01 -5.29177666e-01
3.85365307e-01 9.97179747e-02 2.00222909e-01 6.92883551e-01
-4.83761698e-01 5.00195146e-01 6.93757832e-01 -6.14151239e-01
-5.42521536e-01 -9.91384149e-01 -1.02072990e+00 5.33391237e-01
5.19531727e-01 3.22312295e-01 -4.79916632e-01 -5.56424320e-01
3.58321577e-01 3.84201735e-01 -9.54448521e-01 -4.25082713e-01
-6.14401340e-01 -1.05444074e+00 5.46025157e-01 6.24724686e-01
4.90449578e-01 -6.92904830e-01 -9.24322724e-01 1.93313867e-01
4.53880802e-02 -5.77640355e-01 1.08269654e-01 3.87704998e-01
-8.99244428e-01 -1.30780518e+00 -1.12801087e+00 -1.04084146e+00
6.60006583e-01 -1.71615422e-01 6.68512881e-01 1.38613172e-02
-9.67631161e-01 2.21942395e-01 -4.23624814e-02 -4.11536396e-01
-7.60621250e-01 -2.69455433e-01 -3.88333708e-01 -4.34637927e-02
3.46312046e-01 3.08071394e-02 -1.10697424e+00 2.28945136e-01
-5.72282672e-01 1.83861569e-01 1.06185853e+00 8.58908772e-01
6.84129834e-01 -1.30390301e-01 3.89385402e-01 -5.22364855e-01
3.24160963e-01 -5.53254664e-01 -1.99171811e-01 4.18646455e-01
-3.67595375e-01 -1.49719819e-01 2.28176832e-01 -3.00792605e-01
-9.63490903e-01 4.55743298e-02 1.58041432e-01 -1.57619566e-01
-1.32373586e-01 7.26472735e-01 3.05160701e-01 -4.80138153e-01
9.56454754e-01 3.00772369e-01 2.31631413e-01 8.72683451e-02
-7.65126720e-02 7.73514807e-01 4.77611661e-01 -1.68841243e-01
1.57961428e-01 5.08085608e-01 4.88450885e-01 -5.18708229e-01
-4.77088571e-01 -4.10505146e-01 -4.62072074e-01 -6.35423422e-01
9.05512094e-01 -5.24431765e-01 -8.75798643e-01 2.63076246e-01
-7.81781852e-01 -2.31460050e-01 2.32344687e-01 8.48743081e-01
-4.71939296e-01 5.64941049e-01 -8.70820045e-01 -7.91676641e-01
-5.90922415e-01 -1.02003992e+00 7.67706215e-01 5.99023342e-01
1.05564035e-01 -1.10017705e+00 -2.16494977e-01 -9.28256586e-02
5.75209200e-01 7.66796768e-01 1.25867569e+00 -6.59139097e-01
-8.94579515e-02 -7.42254555e-01 -6.91278815e-01 -4.20759842e-02
3.78287584e-01 3.24326575e-01 -9.27585781e-01 -2.01251879e-01
3.01721971e-02 -1.02906875e-01 9.97087359e-01 1.12399125e+00
1.88131773e+00 1.23630092e-01 -1.07597888e+00 9.98663962e-01
1.39675593e+00 6.94460928e-01 4.71792489e-01 2.66232222e-01
3.58092725e-01 1.83729604e-01 6.33219004e-01 3.53172779e-01
9.66368914e-02 1.35870770e-01 5.46609759e-01 -6.34962738e-01
-4.40422036e-02 1.40914440e-01 -4.25477862e-01 4.15787011e-01
-1.83101371e-01 7.35392049e-02 -1.38548040e+00 2.73792863e-01
-1.32317781e+00 -5.02836645e-01 -1.55959800e-01 1.86625803e+00
8.08630586e-01 9.52483714e-02 -4.15447652e-01 1.80009119e-02
8.69915903e-01 -3.98337632e-01 -5.47716498e-01 -2.15670347e-01
4.63458933e-02 3.22841913e-01 6.11498475e-01 9.14508924e-02
-1.43976820e+00 1.15697689e-01 6.58136463e+00 8.26063514e-01
-1.66557431e+00 -1.91537514e-01 1.30392814e+00 2.87961513e-01
-1.13150671e-01 -5.58054864e-01 -1.92712560e-01 2.73077369e-01
8.99030387e-01 -8.20293650e-02 -4.31939900e-01 7.72351682e-01
-9.36826393e-02 -4.28406924e-01 -1.01890218e+00 9.11292374e-01
-6.89495057e-02 -1.49757862e+00 -2.50917435e-01 4.19534929e-02
3.79304111e-01 -4.35302585e-01 3.33883971e-01 1.37444854e-01
-3.78417611e-01 -1.37138319e+00 5.40713295e-02 9.88788903e-01
1.46049452e+00 -4.59544808e-01 9.68153358e-01 7.51826093e-02
-9.24797952e-01 -5.39049655e-02 -2.31788695e-01 6.15712404e-01
-5.03350377e-01 5.40799558e-01 -1.10103774e+00 5.18212378e-01
6.54933155e-01 3.41637671e-01 -6.88325644e-01 1.33028293e+00
5.44182479e-01 6.57127619e-01 -2.32377335e-01 -3.50900173e-01
1.61198169e-01 8.06078687e-02 7.46236416e-03 1.72259295e+00
6.94303095e-01 1.88625410e-01 -6.93836138e-02 6.86589420e-01
5.24381340e-01 4.98451561e-01 -2.69463003e-01 1.06899910e-01
2.50161052e-01 1.79055345e+00 -1.07225955e+00 -2.67036706e-01
-2.82072157e-01 5.57742357e-01 -1.61303133e-01 2.59743005e-01
-1.09082794e+00 -4.70557898e-01 -1.21086463e-01 1.65805325e-01
-1.07576437e-01 2.49319240e-01 -5.37638068e-01 -6.69807434e-01
-2.35726610e-01 -1.55751944e-01 4.44116414e-01 -6.52354479e-01
-8.91216636e-01 3.56701881e-01 -2.55098432e-01 -1.18793643e+00
1.22097544e-01 -9.71527636e-01 -5.71256638e-01 9.40354526e-01
-1.51428294e+00 -9.19639349e-01 -7.93867111e-01 1.41232774e-01
-1.73213124e-01 -1.68304265e-01 1.27310288e+00 -5.11966422e-02
-3.12304229e-01 6.41790330e-01 4.53160048e-01 3.07430714e-01
5.67277014e-01 -1.39785898e+00 -4.07556504e-01 9.16943774e-02
-6.45752609e-01 2.42416754e-01 2.45238647e-01 -6.35848582e-01
-1.44425833e+00 -1.06360543e+00 8.71373862e-02 1.50011256e-01
5.58156550e-01 1.43588170e-01 -8.69015396e-01 1.42720997e-01
-1.70982137e-01 5.79876661e-01 1.20136976e+00 -4.81532604e-01
2.40480140e-01 -1.34462968e-01 -1.42484152e+00 2.02532768e-01
3.23390394e-01 -5.68997487e-02 -5.42049445e-02 4.57332104e-01
1.95431173e-01 -8.89265060e-01 -1.39714861e+00 8.50427270e-01
1.08482730e+00 -9.40923870e-01 7.22287059e-01 -4.88678932e-01
5.95042467e-01 1.26153633e-01 1.54961109e-01 -9.85620558e-01
-7.08669066e-01 9.16786194e-02 -2.50026733e-01 3.71306270e-01
3.66850346e-01 -6.48772538e-01 8.81019294e-01 5.30879974e-01
-3.04243565e-01 -1.35741353e+00 -9.56063747e-01 -2.03177303e-01
3.23758990e-01 -1.20572530e-01 3.77876431e-01 8.27040315e-01
1.16841808e-01 -7.91843176e-01 4.69121933e-01 3.57545704e-01
5.85141718e-01 2.63308913e-01 4.52042334e-02 -1.16935980e+00
6.85365777e-03 -6.31563783e-01 -8.40681314e-01 -6.26015842e-01
-2.41425380e-01 -1.14572752e+00 -2.55931258e-01 -1.69487512e+00
5.59964001e-01 -8.91708076e-01 -8.81392717e-01 2.72051692e-01
-2.24971831e-01 1.96566164e-01 -3.68922621e-01 2.41732106e-01
-1.41734317e-01 3.86194773e-02 1.73872972e+00 -4.17020917e-01
-5.77185191e-02 1.05830178e-01 -5.64177096e-01 6.30189955e-01
8.17991257e-01 -1.76912904e-01 2.90555865e-01 1.85564801e-01
-1.85726807e-01 7.24011064e-01 5.84652901e-01 -1.24117136e+00
1.34204581e-01 -2.35353768e-01 1.28931522e+00 -6.48259997e-01
5.66072911e-02 -8.85174036e-01 5.80095872e-02 1.18716037e+00
-1.38362646e-01 -8.70711327e-01 4.45351660e-01 4.23545450e-01
-1.11248724e-01 -1.31344363e-01 9.98355210e-01 -4.99800919e-03
-5.08396149e-01 2.28710622e-01 -2.95693755e-01 -6.66002393e-01
1.41671610e+00 -5.61773777e-01 -2.66369224e-01 -5.91320265e-03
-9.07326102e-01 -1.32843256e-01 -2.47648805e-02 -8.46021622e-02
8.26347113e-01 -1.57343185e+00 -7.51580894e-01 3.90809327e-01
1.66298047e-01 -2.62988895e-01 6.76087260e-01 1.18261158e+00
-1.17200899e+00 6.37098670e-01 -3.47688079e-01 -1.16361153e+00
-1.00502908e+00 1.92562521e-01 1.20115888e+00 -7.48630762e-02
-4.77506340e-01 7.72857964e-01 4.14706022e-01 -1.67561024e-02
6.97863176e-02 -6.71796501e-01 -5.59073091e-01 -3.13878864e-01
3.52609307e-01 2.40250736e-01 1.24074921e-01 -4.12330389e-01
-5.46967983e-01 6.49904668e-01 -5.72610535e-02 4.47822928e-01
9.94099379e-01 3.92028958e-01 -2.31591687e-02 2.24930972e-01
1.17592299e+00 -3.19280863e-01 -1.09063470e+00 8.94474685e-02
-3.19248796e-01 -2.23182857e-01 2.12477967e-01 -1.18437481e+00
-1.13065302e+00 8.81339371e-01 1.43931103e+00 6.48620188e-01
1.13603508e+00 1.41727343e-01 3.80984336e-01 -1.37740246e-03
-1.69834971e-01 -5.20331383e-01 2.09216364e-02 -3.02546591e-01
9.05717313e-01 -1.40966475e+00 8.07010382e-02 -6.59368813e-01
-5.03633797e-01 1.84967947e+00 7.50280738e-01 -2.05773162e-03
9.79431987e-01 4.88779277e-01 2.80526698e-01 -3.22761267e-01
-5.81631884e-02 3.69991213e-01 5.65454423e-01 7.03196228e-01
7.79185355e-01 5.45497835e-01 -3.22982162e-01 1.09131527e+00
1.05491281e-01 1.86673731e-01 1.77007258e-01 7.56954670e-01
-8.10192585e-01 -3.90948914e-02 -5.00977337e-01 9.61515427e-01
-5.86274028e-01 2.42782593e-01 1.45432040e-01 8.33938181e-01
7.33375847e-02 4.41153884e-01 1.58096075e-01 -1.03953607e-01
1.48106709e-01 -3.41386646e-02 5.58418512e-01 -2.07932398e-01
-1.45910740e-01 2.22708255e-01 -3.77350628e-01 -3.50661784e-01
-4.44387496e-02 -2.37854213e-01 -1.55280435e+00 2.72838503e-01
-6.90376222e-01 6.48146670e-04 9.70682800e-01 5.18146634e-01
9.99048650e-02 8.11567128e-01 6.66465282e-01 -5.68873763e-01
-3.80291492e-01 -1.01618552e+00 -8.14329922e-01 9.74386409e-02
1.98267281e-01 -6.51441932e-01 -2.19985589e-01 -8.63865688e-02] | [15.2028169631958, -2.690666913986206] |
575f8729-c74e-4a87-9af2-5855ded6f45f | multi-context-attention-fusion-neural-network | 2104.09225 | null | https://arxiv.org/abs/2104.09225v1 | https://arxiv.org/pdf/2104.09225v1.pdf | Multi-context Attention Fusion Neural Network for Software Vulnerability Identification | Security issues in shipped code can lead to unforeseen device malfunction, system crashes or malicious exploitation by crackers, post-deployment. These vulnerabilities incur a cost of repair and foremost risk the credibility of the company. It is rewarding when these issues are detected and fixed well ahead of time, before release. Common Weakness Estimation (CWE) is a nomenclature describing general vulnerability patterns observed in C code. In this work, we propose a deep learning model that learns to detect some of the common categories of security vulnerabilities in source code efficiently. The AI architecture is an Attention Fusion model, that combines the effectiveness of recurrent, convolutional and self-attention networks towards decoding the vulnerability hotspots in code. Utilizing the code AST structure, our model builds an accurate understanding of code semantics with a lot less learnable parameters. Besides a novel way of efficiently detecting code vulnerability, an additional novelty in this model is to exactly point to the code sections, which were deemed vulnerable by the model. Thus helping a developer to quickly focus on the vulnerable code sections; and this becomes the "explainable" part of the vulnerability detection. The proposed AI achieves 98.40% F1-score on specific CWEs from the benchmarked NIST SARD dataset and compares well with state of the art. | ['Sriram Ravi', 'Sathish Kumar Chandrasekaran', 'Prasanna Ganesan', 'Krishna Sundaresan', 'Hariharan Manikandan', 'Anshul Tanwar'] | 2021-04-19 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [-1.99459475e-02 1.91454232e-01 -1.04688458e-01 -5.83699457e-02
-8.75181735e-01 -8.46243560e-01 -1.13116264e-01 4.37105209e-01
2.39486277e-01 -8.97963941e-02 1.93663299e-01 -7.24921227e-01
-2.03549281e-01 -6.03053033e-01 -7.20541060e-01 -3.12889159e-01
-2.25008309e-01 -3.30437541e-01 7.18127489e-02 -4.07120377e-01
7.23676920e-01 3.10262233e-01 -1.26663172e+00 6.62625492e-01
5.20104527e-01 8.41789961e-01 1.87018380e-01 7.12297618e-01
-1.14643008e-01 1.18527985e+00 -8.58565331e-01 -6.20062590e-01
2.84646656e-02 1.40076131e-01 -9.22859609e-01 -4.28331256e-01
1.23944528e-01 -2.89525181e-01 -6.21324107e-02 1.44469476e+00
1.07685007e-01 -5.63360214e-01 2.68469214e-01 -1.25354505e+00
-7.27944732e-01 1.11485732e+00 -8.67512524e-01 4.57777768e-01
4.49751943e-01 3.49250883e-02 1.11677754e+00 -7.21100092e-01
2.03482747e-01 8.52033854e-01 1.01634800e+00 5.13809085e-01
-5.94412386e-01 -7.30167627e-01 2.13613063e-01 1.62633240e-01
-1.24144077e+00 -1.57564551e-01 9.85524356e-01 -7.33761251e-01
1.34743202e+00 8.73681009e-02 -1.01191539e-03 1.00559926e+00
6.69122159e-01 4.02138978e-01 2.73480892e-01 -3.77359480e-01
1.66259378e-01 1.96040407e-01 6.89550221e-01 6.74982011e-01
2.51587451e-01 5.67903854e-02 -1.15536928e-01 -4.45830762e-01
7.08149448e-02 4.13137734e-01 -3.04495513e-01 1.66199714e-01
-5.56887925e-01 8.44512463e-01 5.35312831e-01 4.78291512e-01
-3.56600344e-01 4.01389271e-01 6.72174990e-01 3.07584673e-01
1.72986254e-01 5.53331375e-01 -7.36003757e-01 -4.54472929e-01
-8.19696724e-01 -1.78800467e-02 5.84606349e-01 7.27738738e-01
7.95651019e-01 1.19721636e-01 2.57183164e-01 4.27799910e-01
5.49497426e-01 1.87492371e-01 3.59774500e-01 -2.59051979e-01
8.16957355e-01 1.19267023e+00 -2.02070385e-01 -1.17006350e+00
-4.03849840e-01 -7.81300306e-01 -3.85632217e-01 4.23955768e-01
-2.77108192e-01 -1.89164639e-01 -8.01750243e-01 1.58106387e+00
-2.42238000e-01 6.71110302e-02 1.84355736e-01 3.66951764e-01
5.12994766e-01 5.83995998e-01 8.40408504e-02 2.84116089e-01
1.44724894e+00 -6.54625893e-01 -5.09803176e-01 -3.88679713e-01
6.57993257e-01 -6.81750715e-01 8.63436759e-01 5.50584555e-01
-6.50255084e-01 -4.93568778e-01 -1.38334024e+00 3.23725700e-01
-6.04872584e-01 1.47584066e-01 4.00021821e-01 1.24684727e+00
-9.32136893e-01 5.57462931e-01 -5.23614407e-01 3.08402833e-02
3.78461629e-01 3.95017296e-01 -1.64499685e-01 1.22542195e-01
-1.16005933e+00 8.71553421e-01 3.99601698e-01 1.52250886e-01
-1.43595243e+00 -9.17249799e-01 -9.43557739e-01 4.20445710e-01
3.91742975e-01 1.51007995e-01 1.08709931e+00 -8.99012864e-01
-8.31331015e-01 5.63232481e-01 1.51876807e-01 -5.21580815e-01
-1.70194805e-01 -5.71827710e-01 -6.98824823e-01 -6.75277114e-02
-9.16286781e-02 9.61724520e-02 1.06249964e+00 -1.21883011e+00
-4.82489020e-01 -7.41666406e-02 5.17150521e-01 -5.22173524e-01
-7.31930435e-01 4.54064250e-01 -1.01066604e-01 -5.77463031e-01
-3.78978312e-01 -8.47141802e-01 -5.18901907e-02 -6.43401504e-01
-5.47598004e-01 -8.63613412e-02 1.13254929e+00 -1.13868833e+00
1.75284982e+00 -2.33118129e+00 1.01307079e-01 3.96950930e-01
4.96653050e-01 4.92754430e-01 -7.92311877e-02 6.07298553e-01
-7.43978083e-01 4.76671904e-01 -5.09116113e-01 4.85043153e-02
-8.45925584e-02 -4.09886867e-01 -8.37143123e-01 3.37653816e-01
4.71599817e-01 8.64426494e-01 -6.64921939e-01 1.83801860e-01
-1.48214102e-01 2.72596866e-01 -5.57038546e-01 1.87788621e-01
-1.64484948e-01 -2.59985536e-01 -5.26689291e-01 9.09438968e-01
7.18671978e-01 -1.49211138e-01 -1.59837365e-01 -4.58739372e-03
-1.41339868e-01 1.17846124e-01 -7.37831533e-01 1.52791607e+00
-6.63962483e-01 6.29593432e-01 -1.41432419e-01 -8.25259149e-01
9.92342174e-01 3.80336106e-01 -1.71377249e-02 -3.79811078e-01
2.55349010e-01 1.00658365e-01 5.08207045e-02 -7.25876451e-01
4.16369557e-01 3.68183166e-01 -5.64591944e-01 7.21648335e-01
-1.78729326e-01 3.10126394e-01 -3.76453876e-01 3.71916950e-01
1.53622079e+00 -2.01598853e-01 2.68543243e-01 -2.93733567e-01
8.36735547e-01 -2.17700571e-01 5.65704882e-01 4.88294601e-01
1.19777368e-02 4.27157193e-01 9.21042800e-01 -5.48330843e-01
-9.50113654e-01 -6.83448195e-01 3.12933236e-01 7.06181824e-01
-8.35297350e-03 -6.99113846e-01 -1.13042533e+00 -1.21522617e+00
-2.18292356e-01 7.65524864e-01 -9.31099176e-01 -7.58792460e-01
-6.23302639e-01 -4.47197855e-01 8.20995092e-01 7.99236715e-01
2.97309816e-01 -1.06137395e+00 -8.79419446e-01 2.32627466e-01
-2.81090960e-02 -7.99281955e-01 -4.38550144e-01 3.84590745e-01
-4.90010828e-01 -1.26463127e+00 -3.06920588e-01 -5.14862061e-01
6.13689482e-01 9.26679000e-02 9.07914400e-01 7.38987088e-01
-5.91135383e-01 1.73407659e-01 -6.59671605e-01 -5.16214550e-01
-7.25936651e-01 5.94551601e-02 -2.86916196e-01 -4.63559739e-02
3.89240146e-01 -2.94075608e-01 -2.92225301e-01 2.00753883e-01
-8.54885638e-01 -6.08571768e-01 6.15352213e-01 4.82453734e-01
-2.94014011e-02 5.98964512e-01 4.43900228e-01 -8.69939983e-01
7.68047035e-01 -9.26674783e-01 -6.89614177e-01 4.36002344e-01
-5.22170663e-01 2.31039539e-01 7.46928513e-01 -2.22590595e-01
-1.07776630e+00 4.70877290e-02 -4.27561611e-01 -4.07810628e-01
4.17539924e-02 7.41152167e-01 -3.56089205e-01 -2.77659148e-01
6.37157857e-01 2.30586566e-02 -3.06059629e-01 -5.33232093e-01
-3.14985514e-02 6.04142547e-01 4.35995489e-01 -5.25665522e-01
1.09496379e+00 7.46461079e-02 -4.18922544e-01 -3.46151143e-01
-5.50236285e-01 -4.04785991e-01 -2.33508661e-01 -2.35218748e-01
7.43559182e-01 -6.67523563e-01 -7.11029112e-01 5.21880627e-01
-1.38402605e+00 1.56103924e-01 1.91634819e-01 -1.05278291e-01
6.98422492e-02 7.22467184e-01 -4.81941581e-01 -9.57054138e-01
-4.99078363e-01 -1.50039887e+00 9.11932945e-01 2.19017386e-01
-2.86891282e-01 -6.65596426e-01 2.42272615e-02 1.22456625e-01
5.40886164e-01 3.32719207e-01 1.35330641e+00 -5.97369790e-01
-5.10543823e-01 -4.01824176e-01 -2.19317675e-01 5.34870923e-01
9.42408293e-02 3.83856446e-01 -1.16027439e+00 -2.94952154e-01
3.12420785e-01 -1.13199852e-01 7.08253264e-01 5.20006455e-02
1.21367252e+00 -1.49080545e-01 -4.78717893e-01 4.58987951e-01
1.66517532e+00 6.35251224e-01 8.34118426e-01 3.08341563e-01
6.46827579e-01 7.98917472e-01 4.63507652e-01 4.29752231e-01
1.47667870e-01 5.99482834e-01 1.15428519e+00 1.31807968e-01
2.74019212e-01 -3.40019539e-02 8.12679529e-01 5.55843532e-01
1.47380784e-01 -1.91809863e-01 -1.44458663e+00 5.96281290e-01
-1.60488510e+00 -9.38422561e-01 -1.63120463e-01 2.05782294e+00
3.40369582e-01 4.33285117e-01 -7.67120197e-02 4.35853243e-01
8.27539861e-01 -1.12820961e-01 -4.64318603e-01 -8.31498146e-01
1.90374583e-01 7.65145496e-02 5.32615900e-01 3.75606537e-01
-1.14409268e+00 5.90871274e-01 5.75034857e+00 7.32006133e-01
-9.71153140e-01 3.28328669e-01 5.49169719e-01 2.39450499e-01
-5.14271736e-01 2.38995939e-01 -8.97937834e-01 4.83750790e-01
1.30403912e+00 -1.42135248e-01 1.93999335e-01 1.22771180e+00
-1.99316233e-01 2.22380504e-01 -1.00806975e+00 6.64689302e-01
2.52751499e-01 -1.28341746e+00 -4.46054608e-01 1.11525595e-01
4.05168742e-01 -1.47782221e-01 4.61078227e-01 3.22675109e-01
2.89646864e-01 -1.01638722e+00 1.07960117e+00 3.19932848e-01
4.76690203e-01 -1.18180704e+00 1.06743395e+00 2.35374764e-01
-1.38993931e+00 -9.36011434e-01 -3.90138537e-01 9.38007832e-02
-1.07747413e-01 4.67818707e-01 -6.68971837e-01 4.65465635e-01
9.70866859e-01 6.72824681e-01 -7.53915727e-01 7.73238599e-01
-4.59032834e-01 5.63404679e-01 1.41143262e-01 1.53382763e-01
4.25920218e-01 6.33205712e-01 3.78035039e-01 1.21950519e+00
4.54321861e-01 -1.29953757e-01 -4.10402149e-01 1.23277843e+00
-2.12520715e-02 -3.37594479e-01 -5.31393766e-01 -1.35775223e-01
3.66007864e-01 1.24049211e+00 -8.13283861e-01 1.41798005e-01
-5.13742626e-01 5.86608887e-01 2.50121683e-01 1.59567609e-01
-1.12411380e+00 -7.76315093e-01 8.09546709e-01 -1.57833129e-01
5.27211607e-01 3.97037715e-02 -2.20501602e-01 -9.24746573e-01
3.44679773e-01 -1.03877532e+00 4.27583337e-01 -4.77822393e-01
-9.94663656e-01 9.63656723e-01 -1.40932977e-01 -1.36607409e+00
-3.30347121e-02 -6.77084327e-01 -1.12478948e+00 8.62153292e-01
-1.49246287e+00 -1.12502158e+00 -9.40169692e-02 3.40080857e-01
5.46174288e-01 -5.40656090e-01 6.47892654e-01 4.32815850e-01
-7.94518948e-01 9.28272069e-01 -3.36466104e-01 3.16967040e-01
8.04926082e-02 -1.09567130e+00 1.15178037e+00 1.34537566e+00
5.04994653e-02 1.20677543e+00 4.89432693e-01 -9.75160897e-01
-1.24083817e+00 -1.07730126e+00 6.83111668e-01 -7.59400725e-01
8.30083966e-01 -5.24607301e-01 -1.16417944e+00 5.58130980e-01
2.19928801e-01 -3.70569736e-01 6.26528561e-01 5.32485917e-02
-8.22776794e-01 1.41828358e-01 -1.10748911e+00 1.54433757e-01
5.47354996e-01 -8.74465585e-01 -7.00215161e-01 1.77599058e-01
1.08828580e+00 -1.63232476e-01 -5.86752176e-01 6.02691248e-02
2.68441349e-01 -1.06537592e+00 9.78050649e-01 -4.24965382e-01
7.74871230e-01 -4.39556450e-01 2.18170304e-02 -8.79051447e-01
-3.08401048e-01 -6.80622160e-01 -1.99071959e-01 1.43598139e+00
8.05783868e-01 -3.54551971e-01 7.81615615e-01 6.39213562e-01
-4.05612022e-01 -5.44651449e-01 -7.52718329e-01 -5.45230627e-01
4.40415144e-02 -8.25238526e-01 1.02634931e+00 8.86577904e-01
2.17661992e-01 -3.09512168e-01 -3.83508652e-01 7.93329239e-01
5.90130508e-01 -7.29363784e-02 2.04137072e-01 -9.72066343e-01
-5.69010437e-01 -5.12948930e-01 -5.34999669e-01 -4.74074781e-01
3.53265166e-01 -5.76487362e-01 -6.09578826e-02 -8.38883162e-01
2.51917034e-01 -1.62282020e-01 -6.37203753e-01 8.25347185e-01
-1.99373037e-01 -9.03225169e-02 1.10628717e-01 9.90918502e-02
-3.13708365e-01 -4.79108132e-02 9.67247188e-02 -5.37179649e-01
5.90467975e-02 2.17627540e-01 -1.05972922e+00 7.24628508e-01
7.65079200e-01 -8.49145293e-01 -4.63514656e-01 -5.43178618e-01
8.14926565e-01 1.44932300e-01 4.20991838e-01 -9.63223755e-01
3.45787942e-01 2.64025986e-01 -1.34289637e-01 -5.91742635e-01
-1.92074552e-01 -1.07489777e+00 1.21945970e-01 7.82090306e-01
-2.85064578e-01 5.03173530e-01 6.10363305e-01 5.59699774e-01
-1.69413328e-01 -9.75998759e-01 5.31376302e-01 3.61802429e-03
-9.89752769e-01 7.85769150e-03 -5.05617321e-01 -2.61183649e-01
1.25132561e+00 -2.11793602e-01 -4.91207600e-01 -5.52049130e-02
-3.77007872e-01 -8.88070371e-03 3.28577429e-01 9.23017502e-01
9.23974037e-01 -8.87620687e-01 -4.79988575e-01 3.10320586e-01
4.42223400e-01 -5.69840789e-01 5.92456877e-01 2.74382591e-01
-5.09012878e-01 3.09127539e-01 -1.39709741e-01 -2.20893100e-01
-1.31337976e+00 9.03497756e-01 3.20694387e-01 -2.79814869e-01
-6.48823798e-01 1.11486065e+00 6.66993931e-02 -1.03970729e-01
1.47742629e-01 -4.35579360e-01 -6.02838516e-01 -2.17185259e-01
9.19803143e-01 1.73584074e-01 2.82887876e-01 -5.76510131e-01
-7.71594584e-01 9.35783744e-01 -5.89208186e-01 4.92816150e-01
1.33468246e+00 1.69852331e-01 -2.88017273e-01 -3.02449185e-02
1.30778325e+00 1.21918738e-01 -1.13284910e+00 9.03272331e-02
4.56503540e-01 -2.72179633e-01 1.37115911e-01 -1.05545998e+00
-1.45335937e+00 1.19375098e+00 6.80284500e-01 4.15967703e-01
1.27612936e+00 -9.50761978e-03 7.45071650e-01 1.54262155e-01
3.80677283e-01 -7.15148330e-01 1.67062715e-01 4.43617135e-01
6.93054914e-01 -1.06761408e+00 -1.85481936e-01 -2.51855373e-01
-5.81595957e-01 1.40636241e+00 5.83818018e-01 3.13359052e-02
6.26036763e-01 8.64282489e-01 3.57154699e-04 -5.80896556e-01
-6.84100866e-01 4.07486230e-01 2.09223136e-01 6.06917262e-01
3.35355014e-01 -2.33222336e-01 2.90432036e-01 9.66986477e-01
1.38335273e-01 -5.54164112e-01 7.94582725e-01 1.05528522e+00
-5.28013766e-01 -1.12904823e+00 -3.85570198e-01 1.71535730e-01
-8.55780661e-01 -3.82056504e-01 -4.98904943e-01 4.54311520e-01
3.12097698e-01 1.12077153e+00 -4.45633143e-01 -8.33291888e-01
3.04936707e-01 -1.32325009e-01 -2.99542427e-01 -7.77951539e-01
-1.23853743e+00 -4.42013979e-01 -1.78597733e-01 -7.44928598e-01
1.59665376e-01 -4.41177040e-01 -1.27430499e+00 -1.07890882e-01
-5.16961694e-01 3.71409893e-01 8.70848417e-01 7.74878263e-01
5.21817148e-01 9.38146174e-01 8.55613887e-01 -4.07438278e-01
-5.58237076e-01 -5.49153388e-01 -2.53365695e-01 3.05896938e-01
5.64977050e-01 -5.90860724e-01 -5.31521976e-01 7.07995221e-02] | [7.054438591003418, 7.771771430969238] |
4fe37ac7-d4e9-4bb9-9f8e-3bfc0d1d0a73 | the-impact-of-cross-lingual-adjustment-of-1 | 2204.06457 | null | https://arxiv.org/abs/2204.06457v1 | https://arxiv.org/pdf/2204.06457v1.pdf | The Impact of Cross-Lingual Adjustment of Contextual Word Representations on Zero-Shot Transfer | Large pre-trained multilingual models such as mBERT and XLM-R enabled effective cross-lingual zero-shot transfer in many NLP tasks. A cross-lingual adjustment of these models using a small parallel corpus can potentially further improve results. This is a more data efficient method compared to training a machine-translation system or a multi-lingual model from scratch using only parallel data. In this study, we experiment with zero-shot transfer of English models to four typologically different languages (Spanish, Russian, Vietnamese, and Hindi) and three NLP tasks (QA, NLI, and NER). We carry out a cross-lingual adjustment of an off-the-shelf mBERT model. We confirm prior finding that this adjustment makes embeddings of semantically similar words from different languages closer to each other, while keeping unrelated words apart. However, from the paired-differences histograms introduced in our work we can see that the adjustment only modestly affects the relative distances between related and unrelated words. In contrast, fine-tuning of mBERT on English data (for a specific task such as NER) draws embeddings of both related and unrelated words closer to each other. The cross-lingual adjustment of mBERT improves NLI in four languages and NER in two languages, while QA performance never improves and sometimes degrades. When we fine-tune a cross-lingual adjusted mBERT for a specific task (e.g., NLI), the cross-lingual adjustment of mBERT may still improve the separation between related and related words, but this works consistently only for the XNLI task. Our study contributes to a better understanding of cross-lingual transfer capabilities of large multilingual language models and of effectiveness of their cross-lingual adjustment in various NLP tasks. | ['Pavel Braslavski', 'Elena Arslanova', 'Leonid Boytsov', 'Pavel Efimov'] | 2022-04-13 | null | null | null | null | ['xlm-r', 'machine-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing'] | [-2.68320650e-01 -1.56889215e-01 -1.82104543e-01 -3.95807683e-01
-1.43470538e+00 -1.01493442e+00 6.37081683e-01 3.66682410e-01
-1.07212901e+00 7.24781215e-01 4.24868822e-01 -6.37556672e-01
3.31667334e-01 -7.59086668e-01 -9.64715183e-01 -4.09524798e-01
3.56780499e-01 7.30117202e-01 -1.48605788e-02 -5.70681453e-01
-3.61033678e-02 3.90205085e-01 -9.18939710e-01 2.29653746e-01
1.09047961e+00 1.03104614e-01 3.88637841e-01 3.06762159e-01
-3.09869766e-01 3.39616612e-02 -3.95264328e-01 -6.08475089e-01
3.60530406e-01 -3.25530261e-01 -7.61989295e-01 -7.38389909e-01
6.24640524e-01 1.23625532e-01 1.67184845e-02 8.76419485e-01
6.78613424e-01 2.87071139e-01 7.24461138e-01 -8.02349508e-01
-1.09535563e+00 7.70599723e-01 -4.99371558e-01 1.49793342e-01
1.69232026e-01 1.49104461e-01 1.16850317e+00 -1.05109739e+00
7.26937354e-01 1.47777891e+00 1.01921368e+00 4.42946225e-01
-1.58010650e+00 -7.60471940e-01 -1.27276942e-01 4.15381193e-02
-1.56284833e+00 -5.03767133e-01 1.33703008e-01 -3.17424834e-01
1.40395033e+00 -7.48992488e-02 2.50051111e-01 9.66758847e-01
4.82527196e-01 4.20985162e-01 1.26879084e+00 -8.02492082e-01
-1.20473206e-01 4.79571819e-01 6.00297526e-02 3.84497792e-01
1.09425470e-01 1.18200727e-01 -4.02656555e-01 5.43179139e-02
3.30419689e-01 -3.77702385e-01 -2.42957801e-01 -1.47446930e-01
-1.36927032e+00 1.08830833e+00 4.14463133e-01 6.74298584e-01
-1.77430302e-01 -3.58812921e-02 5.85047424e-01 5.11764348e-01
4.92736936e-01 8.53399873e-01 -9.32599485e-01 -2.01852977e-01
-7.46766746e-01 -1.90310866e-01 6.96998477e-01 7.83884943e-01
1.15592778e+00 -1.05692588e-01 -6.78986311e-02 1.28591692e+00
-2.21308414e-02 6.66887701e-01 8.94512057e-01 -4.86580193e-01
7.29320228e-01 8.60411450e-02 -1.42087817e-01 -4.86715764e-01
-2.80495405e-01 -2.77162552e-01 -2.67476559e-01 7.58403316e-02
6.28233850e-01 -4.31442589e-01 -8.91327143e-01 2.15779233e+00
1.04516648e-01 -4.17277604e-01 4.82429713e-01 5.97352982e-01
3.46194118e-01 1.04250050e+00 4.82795179e-01 1.54255733e-01
1.54134917e+00 -1.02699137e+00 -4.36054498e-01 -5.13833225e-01
1.24724066e+00 -1.17437136e+00 1.77242839e+00 -7.45148361e-02
-9.19282854e-01 -7.62530744e-01 -8.40767682e-01 -5.33804119e-01
-8.05566013e-01 -5.96197462e-03 2.30234325e-01 5.18128276e-01
-1.01967502e+00 5.05057693e-01 -7.52342165e-01 -8.39197040e-01
-5.51863089e-02 2.50958204e-01 -7.01357961e-01 -4.98509675e-01
-1.62725163e+00 1.65943503e+00 3.60228688e-01 -4.59165722e-01
-4.41741765e-01 -8.66531312e-01 -1.15552592e+00 2.43782476e-02
-1.66969538e-01 -4.45938021e-01 1.04420650e+00 -1.04530525e+00
-1.30256999e+00 9.83322918e-01 -2.14659601e-01 -3.41829300e-01
-1.74160134e-02 -2.90025771e-01 -4.31005299e-01 -3.94392490e-01
5.54888189e-01 1.06453705e+00 3.21372569e-01 -8.93487453e-01
-5.32773614e-01 -3.46259564e-01 -1.01294011e-01 5.93259990e-01
-4.83073711e-01 3.11320901e-01 -2.74540454e-01 -6.54451370e-01
-5.18652260e-01 -9.20179844e-01 7.44388700e-02 -4.90095049e-01
2.80676216e-01 -3.05799335e-01 3.00121218e-01 -9.31896091e-01
9.09678221e-01 -2.13721013e+00 1.93620935e-01 -1.30725190e-01
-5.44395983e-01 2.94488460e-01 -6.55870974e-01 7.59119093e-01
-1.34393305e-01 2.82105953e-01 -1.24220677e-01 -3.71189803e-01
5.03611751e-02 4.43757862e-01 -4.25927304e-02 3.93757343e-01
1.54035985e-01 1.13609731e+00 -8.65081370e-01 -3.74039412e-01
1.28376514e-01 5.78990400e-01 -4.98192161e-01 -4.03149519e-03
1.81993619e-01 2.77564734e-01 2.55435705e-01 1.54345751e-01
5.07810593e-01 3.40875804e-01 4.53404747e-02 -2.06277490e-01
-2.00245097e-01 7.82975316e-01 -5.39954662e-01 1.86216128e+00
-1.24722075e+00 6.87965155e-01 -2.72501707e-01 -6.81064129e-01
7.65591979e-01 2.87393838e-01 -2.68724672e-02 -9.84094083e-01
-1.34392783e-01 5.95252514e-01 4.52439934e-01 -1.61719859e-01
5.60414553e-01 -8.61265421e-01 -3.32892090e-01 5.05986035e-01
3.96599680e-01 -3.14820856e-01 1.47260711e-01 -7.73880258e-02
7.72110403e-01 2.32281089e-01 4.88303542e-01 -5.80613613e-01
2.59198010e-01 1.92023963e-01 4.38864291e-01 4.51048374e-01
-6.60254881e-02 4.35364783e-01 -3.74476388e-02 -3.10857110e-02
-1.11743045e+00 -1.26885378e+00 -4.38993454e-01 1.54495227e+00
-1.91750914e-01 -3.23107272e-01 -7.40869582e-01 -8.52598429e-01
1.34346843e-01 1.34642565e+00 -4.65613037e-01 -2.45360896e-01
-7.51308143e-01 -7.03541994e-01 8.32711279e-01 4.84996706e-01
6.06174953e-02 -1.04387558e+00 1.19392522e-01 2.73331404e-01
-3.60635519e-01 -9.56836641e-01 -9.18682218e-01 4.28456783e-01
-6.27831638e-01 -4.33388263e-01 -8.26832175e-01 -1.14360702e+00
4.76594031e-01 1.09719202e-01 1.30501139e+00 -3.69563192e-01
1.55997649e-01 1.84242159e-01 -3.18648458e-01 -3.31322938e-01
-6.61788464e-01 4.75648761e-01 3.05691898e-01 -4.70489413e-01
7.98894167e-01 -3.08966666e-01 -6.55539334e-02 3.36065024e-01
-7.29273319e-01 -3.30260217e-01 5.35390437e-01 8.91194105e-01
5.19873202e-01 -3.11239272e-01 6.82785273e-01 -8.72911036e-01
7.05990791e-01 -4.77486044e-01 -2.71731555e-01 3.68107766e-01
-4.34966862e-01 4.16590750e-01 8.05658698e-01 -6.78123951e-01
-1.10395634e+00 -3.21109504e-01 -3.73402983e-01 -3.28536183e-02
-1.33546874e-01 6.36951447e-01 -1.90389976e-01 5.28283715e-02
9.21941042e-01 -1.18496350e-03 -2.75824219e-01 -5.21597683e-01
7.80318856e-01 7.53310680e-01 2.91747212e-01 -7.82690167e-01
7.46092916e-01 -4.67015617e-02 -6.39795661e-01 -7.36355722e-01
-7.04171419e-01 -3.91289800e-01 -7.47364223e-01 4.20861930e-01
1.01212978e+00 -1.10986960e+00 -1.75267197e-02 1.93469733e-01
-1.23931873e+00 -7.59444118e-01 -4.85299617e-01 8.24599862e-01
-2.41755828e-01 1.17461056e-01 -7.26648271e-01 2.15873532e-02
-2.97802836e-01 -1.15071452e+00 9.74988103e-01 -9.20072272e-02
-6.28357708e-01 -1.48877609e+00 5.99804759e-01 2.30640829e-01
4.19670612e-01 -4.17765260e-01 1.38260305e+00 -9.11566913e-01
2.39036698e-02 -5.55018149e-03 -1.19268656e-01 6.73968911e-01
4.60736871e-01 -3.85484636e-01 -7.38695979e-01 -5.97201645e-01
-1.25904530e-01 -5.23992777e-01 6.65839553e-01 1.41336367e-01
1.70951888e-01 -1.22398458e-01 -2.12535068e-01 5.30582130e-01
1.48748207e+00 7.76827112e-02 5.57821989e-01 4.72684771e-01
7.47134328e-01 6.39711499e-01 7.12995827e-01 -4.76598024e-01
6.07963622e-01 7.56664991e-01 -3.75277787e-01 -2.68499970e-01
-3.39914262e-01 -3.69155675e-01 9.12694633e-01 1.33137965e+00
3.45917016e-01 -3.02006781e-01 -1.08686483e+00 9.28671062e-01
-1.33646142e+00 -4.66101229e-01 1.45391315e-01 2.42621875e+00
1.23129463e+00 -1.35701329e-01 -2.62141168e-01 -6.06533349e-01
6.87922716e-01 -1.74164072e-01 -2.43336365e-01 -9.65164602e-01
-1.70033932e-01 5.45412600e-01 6.80150211e-01 9.32176948e-01
-7.13197947e-01 1.55877769e+00 5.64798737e+00 1.08988583e+00
-1.39170718e+00 5.85395992e-01 3.29222411e-01 6.78518862e-02
-4.85056549e-01 1.90355152e-01 -1.08006012e+00 3.07567149e-01
1.30298007e+00 -2.95925230e-01 6.06276870e-01 5.13840079e-01
-3.45236808e-03 4.70068827e-02 -1.39258122e+00 7.14860082e-01
2.49780610e-01 -9.07819211e-01 1.30243182e-01 7.86985457e-02
8.11506569e-01 7.22497404e-01 -9.05509293e-02 8.40418339e-01
5.14025271e-01 -1.08003867e+00 5.31295002e-01 -4.35023941e-02
1.00404584e+00 -9.20098484e-01 8.60582471e-01 3.73337239e-01
-9.71301615e-01 3.84630382e-01 -5.14716923e-01 9.14963707e-02
2.86014706e-01 9.77557972e-02 -7.68250167e-01 3.90141547e-01
5.17335594e-01 4.38292384e-01 -5.84270775e-01 3.48400623e-01
-4.45747167e-01 6.34126842e-01 -2.49574691e-01 1.86921880e-01
4.50940996e-01 -3.25709283e-01 3.59174550e-01 1.53744745e+00
5.07135093e-01 -2.73286819e-01 -7.08642974e-02 6.13906026e-01
-2.22309068e-01 6.78292513e-01 -8.01716387e-01 -6.88542128e-02
4.93276536e-01 1.24726069e+00 -2.29032338e-01 -4.62242365e-01
-6.60578728e-01 1.19605970e+00 8.71683896e-01 3.70804340e-01
-7.25660563e-01 -6.34768426e-01 8.42453241e-01 6.01692721e-02
3.03966969e-01 -4.32602972e-01 -2.40554437e-01 -1.31647742e+00
-2.15604186e-01 -1.08575368e+00 2.49064684e-01 -7.37007618e-01
-1.51585507e+00 7.55744755e-01 -1.87815636e-01 -8.54780674e-01
-3.22084129e-01 -6.73758686e-01 -5.05586982e-01 1.47601163e+00
-1.47780752e+00 -1.23012102e+00 4.05396104e-01 4.33786452e-01
5.74873209e-01 -1.37133077e-01 1.15262318e+00 3.86801004e-01
-3.12351495e-01 8.86233449e-01 4.52795774e-01 2.27965385e-01
1.57586002e+00 -1.17467642e+00 6.66242719e-01 8.28549087e-01
4.69712377e-01 1.06725073e+00 5.00748098e-01 -6.47303164e-01
-1.05307710e+00 -1.18319488e+00 1.42627358e+00 -5.79958498e-01
8.93591523e-01 -4.59951043e-01 -1.18460417e+00 9.73069131e-01
6.73973262e-01 -2.53036112e-01 8.83932292e-01 5.59615552e-01
-5.17000675e-01 -4.19546179e-02 -1.03149629e+00 7.45439947e-01
6.57946527e-01 -8.41228127e-01 -1.00301981e+00 2.94373691e-01
8.96370947e-01 7.88315013e-03 -1.08414030e+00 1.38150290e-01
4.42305923e-01 -4.21951681e-01 8.44320834e-01 -7.95079112e-01
2.82895476e-01 -1.53040767e-01 -3.84858668e-01 -2.10624957e+00
-3.31156135e-01 -1.89950436e-01 7.24329948e-01 1.48630476e+00
7.34794915e-01 -8.79208922e-01 4.91214395e-02 3.37047398e-01
-3.19449127e-01 -4.52892035e-01 -8.76323283e-01 -1.16072536e+00
1.03749371e+00 -4.00217682e-01 2.72445679e-01 1.19059587e+00
3.68376449e-02 8.71152759e-01 9.63002369e-02 9.21627134e-02
2.86910176e-01 -2.22473815e-01 6.54716432e-01 -8.34269345e-01
-3.08694005e-01 -2.95196652e-01 -1.04828082e-01 -8.54313254e-01
5.21286190e-01 -1.51429605e+00 2.55582809e-01 -1.47715557e+00
2.78615225e-02 -7.10155368e-01 -3.56141716e-01 7.37567484e-01
-3.22121203e-01 3.25177878e-01 3.14462543e-01 1.11199528e-01
-2.46021263e-02 6.01326764e-01 8.96946967e-01 4.30252105e-02
-2.97033370e-01 -5.71537375e-01 -7.32694387e-01 6.22096598e-01
8.37548435e-01 -7.84112155e-01 -1.20637298e-01 -9.56185699e-01
9.54866409e-04 -3.32378119e-01 -1.86337411e-01 -6.02147937e-01
-9.94215980e-02 4.04508114e-02 2.46669665e-01 -4.77606757e-03
2.62054175e-01 -4.83053565e-01 -2.23402902e-01 3.40102524e-01
-1.68630406e-01 3.56539160e-01 7.48269737e-01 2.50931899e-03
-1.89565316e-01 -3.97267669e-01 9.98843849e-01 -1.25604093e-01
-5.85083902e-01 -1.03469118e-01 -4.49537277e-01 5.70241630e-01
6.17941499e-01 3.63706648e-02 -3.04638565e-01 -1.12934716e-01
-4.26100016e-01 6.32254183e-02 7.49918342e-01 6.91393554e-01
-1.08040199e-02 -1.46711361e+00 -1.01998127e+00 2.41531372e-01
2.55818844e-01 -5.82379878e-01 -5.11839725e-02 7.16409802e-01
-2.94345707e-01 5.35848200e-01 -3.97800744e-01 -4.02651519e-01
-1.06261086e+00 6.46726966e-01 3.28564078e-01 -4.25631762e-01
-1.39824465e-01 7.20786154e-01 5.79847813e-01 -1.20141780e+00
-2.57062018e-01 -1.52431861e-01 1.31391853e-01 2.04080328e-01
1.89195812e-01 1.68576404e-01 2.54018158e-01 -9.56090212e-01
-4.64668095e-01 8.42461586e-01 -3.52534235e-01 -4.83104140e-01
1.14925885e+00 -3.39519948e-01 -8.27021226e-02 7.66265869e-01
1.52029955e+00 5.66527724e-01 -6.75162077e-01 -3.24570119e-01
-1.33325607e-01 -1.25649750e-01 -7.44266585e-02 -9.03503060e-01
-5.91352701e-01 1.17172992e+00 3.97988439e-01 -3.86661768e-01
7.79655755e-01 1.74621958e-02 1.00592804e+00 2.61478603e-01
5.09281039e-01 -1.08708537e+00 -2.81749219e-01 8.72232676e-01
7.86366522e-01 -1.22103453e+00 -3.72829795e-01 5.11702038e-02
-8.56085300e-01 8.11538458e-01 5.73036909e-01 -8.47172830e-03
2.56445944e-01 6.78228363e-02 4.18796152e-01 1.56584859e-01
-5.39643824e-01 -1.97400868e-01 3.29005510e-01 5.45753658e-01
8.51035357e-01 2.13803187e-01 -4.42242473e-01 2.40815043e-01
-4.99545395e-01 -4.47430640e-01 2.07859442e-01 5.67450583e-01
-2.22442359e-01 -1.55553782e+00 -3.71313870e-01 5.19116409e-02
-3.69725704e-01 -8.56767356e-01 -1.03089258e-01 1.11163759e+00
2.91439801e-01 7.36243606e-01 2.59637892e-01 -9.58698392e-02
2.81041622e-01 5.44879019e-01 6.10076308e-01 -9.78389323e-01
-7.70280182e-01 -2.34968551e-02 2.73240864e-01 -1.77915603e-01
5.85910194e-02 -7.40175664e-01 -1.27003455e+00 -2.36041933e-01
-2.74660051e-01 2.47915596e-01 8.32490504e-01 9.53963041e-01
2.92450815e-01 2.34575987e-01 2.88846761e-01 -6.03711784e-01
-5.07678568e-01 -1.18459797e+00 -2.81575292e-01 5.03805518e-01
-1.10315336e-02 -4.13214177e-01 -3.39112818e-01 -1.06429577e-01] | [10.983302116394043, 9.952831268310547] |
3c3d2dbe-dc6f-48c8-ab9d-e507e4874979 | the-contribution-of-local-variations-in-hue | 2108.04730 | null | https://arxiv.org/abs/2108.04730v2 | https://arxiv.org/pdf/2108.04730v2.pdf | The contribution of local variations in hue or contrast to symmetry of things in a thing | Symmetry contributes to processes of perceptual organization in biological vision and influences the quality and time of goal directed decision making in animals and humans, as discussed in recent work on the examples of symmetry of things in a thing and bilateral shape symmetry. The present study was designed to show that selective chromatic variations in geometric shape configurations with mirror symmetry can be exploited to highlight functional properties of symmetry of things in a thing in human vision. The experimental procedure uses a psychophysical two alternative forced choice technique, where human observers have to decide as swiftly as possible whether two shapes presented simultaneously on a computer screen are symmetrical or not. The stimuli are computer generated 2D shape configurations consisting of multiple elements, with and without systematic variations in local color, color saturation, or achromatic contrast producing variations in symmetry of things in a thing. All stimulus pairs presented had perfect geometric mirror symmetry. The results show that varying the color of local shape elements selectively in multichromatic and monochromatic shapes significantly slows down perceptual response times, which are a direct measure of stimulus uncertainty. It is concluded that local variations in hue or contrast produce functionally meaningful variations in symmetry of things in thing, revealed here as a relevant perceptual variable in symmetry detection. Disturbance of the latter increases stimulus uncertainty and thereby affects the perceptual salience of mirror symmetry in the time course for goal relevant human decisions. | ['Birgitta Dresp-Langley'] | 2021-08-10 | null | null | null | null | ['symmetry-detection'] | ['computer-vision'] | [ 5.01568377e-01 -2.96777695e-01 2.52637774e-01 -4.37998861e-01
4.09462035e-01 -8.40087414e-01 5.73169589e-01 -6.32698610e-02
-8.01190794e-01 4.52891469e-01 2.62890279e-01 -2.36753300e-01
-4.33896422e-01 -4.55685347e-01 -2.89875299e-01 -6.40699387e-01
2.76526690e-01 1.60687268e-01 3.04042786e-01 -2.46967584e-01
1.05852008e+00 9.83003438e-01 -1.95636415e+00 1.48279488e-01
2.92775124e-01 8.04779828e-01 2.12435439e-01 7.63486981e-01
2.19030783e-01 1.35058865e-01 -4.22776848e-01 -1.75660297e-01
9.33518112e-01 -8.86703074e-01 -5.43074012e-01 4.59832430e-01
5.58834970e-01 2.80362546e-01 3.58266443e-01 1.17733204e+00
7.45517552e-01 2.73295671e-01 7.77454674e-01 -9.23566818e-01
-1.03829670e+00 2.38420144e-01 -5.75095177e-01 3.47091436e-01
4.87560451e-01 6.80675924e-01 6.83683157e-01 -6.36965036e-01
4.54349935e-01 1.36528468e+00 -6.30357191e-02 5.45747697e-01
-1.69230950e+00 -4.07853127e-01 -1.37429133e-01 2.09965169e-01
-1.30746388e+00 -9.36937690e-01 7.27947354e-01 -7.08609104e-01
1.19662786e+00 5.99078655e-01 8.94259512e-01 2.47969985e-01
8.88014674e-01 -5.66660285e-01 1.83711600e+00 -6.27909482e-01
2.48838171e-01 3.50571163e-02 -2.97536403e-01 2.29395926e-01
5.17020583e-01 6.98307216e-01 -5.74170053e-01 1.51270762e-01
1.05293989e+00 -4.34446961e-01 -4.70844179e-01 -2.26919606e-01
-1.49554145e+00 1.93788514e-01 4.59469765e-01 4.32978302e-01
-3.97944570e-01 -2.43179068e-01 -8.48374367e-02 5.44050932e-01
-3.11068207e-01 1.28506386e+00 -8.36252719e-02 1.61777288e-01
-1.90805063e-01 1.09211698e-01 4.38598067e-01 4.66487885e-01
1.42292231e-01 -5.90084232e-02 -4.26245630e-01 6.08158946e-01
1.63432255e-01 6.40489042e-01 5.78052044e-01 -8.44614446e-01
-1.81616917e-01 6.04205191e-01 1.69804931e-01 -9.16696727e-01
-7.49892771e-01 1.35769129e-01 -3.14293355e-01 1.17700601e+00
8.10837746e-01 3.12471479e-01 -7.37826347e-01 1.68930304e+00
2.23874047e-01 -8.74400914e-01 -2.56071597e-01 1.25030720e+00
3.96850288e-01 1.26080424e-01 1.95983976e-01 -5.77932477e-01
1.59152484e+00 3.39007378e-03 -5.30086339e-01 -1.07943706e-01
-2.41617367e-01 -1.26814473e+00 1.36341965e+00 3.36265743e-01
-1.19742525e+00 -7.86029875e-01 -9.12908077e-01 7.44219078e-03
-3.88662189e-01 -3.02467555e-01 2.87389070e-01 9.46220577e-01
-8.43270719e-01 7.00387836e-01 1.01077585e-02 -3.80361646e-01
9.15482268e-02 2.55951703e-01 -3.55675608e-01 2.58728206e-01
-6.28234744e-01 1.10683393e+00 1.85734659e-01 1.44286022e-01
-1.02257073e-01 -2.41956666e-01 -8.12374532e-01 -5.50331399e-02
3.30870338e-02 -8.32152724e-01 1.07184005e+00 -1.06976759e+00
-1.66968822e+00 1.57146823e+00 -6.04467578e-02 7.79914260e-02
4.25448477e-01 5.34332514e-01 -5.22409499e-01 -5.67574129e-02
5.26495762e-02 7.95727491e-01 1.21148932e+00 -9.39032555e-01
4.02276516e-02 -8.64722252e-01 -2.09041148e-01 4.72209215e-01
6.93745077e-01 5.20387650e-01 5.27271807e-01 -2.36378297e-01
5.82688212e-01 -7.59724438e-01 -3.02951764e-02 2.34843165e-01
-2.36416206e-01 -1.47044882e-01 1.16377585e-01 -2.64301449e-02
5.87198794e-01 -2.07204819e+00 -2.82938629e-01 2.73402005e-01
-3.50156128e-02 -1.90244630e-01 -2.55848646e-01 3.59317631e-01
-3.76226276e-01 6.25170022e-02 -4.99770083e-02 6.10192418e-01
2.44578362e-01 -1.78360045e-01 -6.22264966e-02 7.18109250e-01
1.47621602e-01 9.66816366e-01 -5.99628150e-01 -1.21989295e-01
2.01177657e-01 1.85148641e-01 -2.28853196e-01 -9.73625630e-02
4.90487702e-02 5.30018389e-01 2.32357308e-01 4.95714158e-01
8.58270168e-01 2.32914984e-01 -3.26064713e-02 -1.24359928e-01
-5.62204480e-01 3.45858544e-01 -9.69660282e-01 9.38517511e-01
1.90530479e-01 7.00568438e-01 -6.55991673e-01 -3.17826301e-01
1.07207584e+00 -5.20529784e-03 -3.06674182e-01 -1.57711005e+00
2.62638301e-01 1.67803615e-01 9.96413231e-01 -4.35876608e-01
3.11884999e-01 -8.17426801e-01 6.20456599e-02 3.60481352e-01
-6.65537775e-01 -1.08525634e+00 3.17220271e-01 -5.72275400e-01
2.06880271e-01 -5.40659539e-02 8.49368632e-01 -5.69198668e-01
3.68337750e-01 -4.86668706e-01 4.50303584e-01 6.91786528e-01
-4.68169928e-01 4.67620194e-01 6.25133216e-01 -5.11101723e-01
-1.22776961e+00 -1.43758309e+00 -4.46285248e-01 9.12192047e-01
3.78501296e-01 3.98656398e-01 -1.72465712e-01 3.53704691e-01
8.49147961e-02 9.92112875e-01 -6.84395313e-01 -4.06636357e-01
-3.55376691e-01 -4.68919069e-01 -1.39640778e-01 2.44380817e-01
4.40474212e-01 -1.68012393e+00 -1.55168152e+00 -2.36501187e-01
3.95460576e-01 -7.86175549e-01 -5.59519053e-01 -4.20748033e-02
-7.74528444e-01 -1.25808835e+00 -4.42251712e-01 -5.36458433e-01
8.99759710e-01 4.85769510e-01 1.01945996e+00 -3.95439178e-01
-9.21797097e-01 4.89381820e-01 -5.94423637e-02 -7.50598609e-01
-2.00548932e-01 -1.20913744e+00 1.76108450e-01 -1.66652888e-01
4.40103024e-01 -5.10876596e-01 -7.17829645e-01 7.54074275e-01
-6.64161861e-01 -2.32994338e-04 3.68373215e-01 3.64500523e-01
6.11361444e-01 -9.81239229e-02 6.35097921e-02 -1.49268225e-01
8.16993713e-01 3.39275688e-01 -8.33731592e-01 -9.70054418e-02
-3.49347621e-01 1.26603141e-01 2.37539351e-01 -4.38784987e-01
-9.84406531e-01 -3.17433894e-01 6.26561701e-01 2.44392589e-01
-4.61598277e-01 -3.42381336e-02 -1.81519391e-03 -4.23100561e-01
1.17839122e+00 3.15188140e-01 -1.15210548e-01 1.27184793e-01
-1.82037298e-02 1.88837841e-01 3.97972822e-01 -3.18954676e-01
6.37593091e-01 6.43203914e-01 5.35116255e-01 -8.70626986e-01
-2.03705773e-01 -2.33974114e-01 -6.30340695e-01 -3.67239952e-01
9.39280093e-01 -3.23095202e-01 -1.14038610e+00 6.14310265e-01
-6.76971912e-01 -1.92896679e-01 -5.40553212e-01 7.88974166e-01
-8.78430903e-01 1.44426137e-01 2.77203489e-02 -9.29147482e-01
1.44104153e-01 -8.83923173e-01 6.89175308e-01 4.24515635e-01
-2.85792768e-01 -2.46208668e-01 -1.73828136e-02 8.23332884e-05
3.61224264e-01 2.22976476e-01 1.10462654e+00 -2.32304052e-01
-5.88855028e-01 1.75174903e-02 -2.69117624e-01 1.12030422e-02
5.23753226e-01 6.37652129e-02 -8.49194050e-01 -2.72946134e-02
4.16474134e-01 -2.75979251e-01 7.99576759e-01 7.74484277e-01
5.87182701e-01 1.26094803e-01 2.86862671e-01 3.57973844e-01
1.36291611e+00 9.23809588e-01 8.46488535e-01 1.31093904e-01
1.61509477e-02 1.14668536e+00 3.30704808e-01 3.04458767e-01
-3.85052472e-01 4.41403747e-01 3.71299945e-02 1.66166857e-01
-6.25316286e-03 1.85893491e-01 2.83343904e-02 2.20719688e-02
-3.77952278e-01 5.13326935e-02 -6.30536973e-01 2.25662425e-01
-8.88723850e-01 -1.30663431e+00 -5.62645011e-02 2.78492212e+00
4.51729059e-01 2.97279000e-01 8.34719315e-02 3.42349023e-01
1.02274871e+00 -2.75425892e-02 -6.02433324e-01 -9.01444733e-01
-2.84239858e-01 -1.24022504e-02 3.93440455e-01 4.52950567e-01
-5.05045056e-01 5.73126316e-01 7.03394079e+00 1.50625780e-01
-1.21269667e+00 -5.59697390e-01 3.88660312e-01 -2.97049791e-01
-3.84347081e-01 -7.13867918e-02 -3.98182243e-01 2.07828209e-01
2.78884321e-01 -5.96632063e-01 6.46371961e-01 1.57093748e-01
7.05621004e-01 -8.25214565e-01 -1.22495091e+00 1.16699672e+00
2.92448312e-05 -7.25977957e-01 4.40844260e-02 3.14963758e-02
4.27193999e-01 -5.39328754e-01 7.54465759e-01 -3.53167415e-01
-3.12832355e-01 -1.31112814e+00 7.32702911e-01 4.74948972e-01
8.90092492e-01 -5.70345640e-01 2.89937872e-02 5.09774797e-02
-8.18307400e-01 -2.05071613e-01 -7.22241342e-01 -7.94513762e-01
-9.14197341e-02 2.16054246e-01 -7.40491688e-01 -2.62625396e-01
5.66733539e-01 -3.26873451e-01 -4.27077413e-01 1.45462954e+00
-2.43009627e-01 -1.71017423e-01 -2.91310638e-01 -1.96211442e-01
-7.77744874e-02 -2.81356424e-01 8.25588584e-01 6.89495683e-01
-4.44104783e-02 4.27149028e-01 -5.95468879e-01 1.36259305e+00
4.92753327e-01 4.22181576e-01 -6.95248187e-01 -4.86782119e-02
3.27122718e-01 7.52067149e-01 -1.20891345e+00 7.58320019e-02
-1.59969851e-01 8.32466662e-01 -2.20783338e-01 4.97166514e-01
-2.56750911e-01 -2.23470077e-01 5.59133530e-01 1.16826475e-01
7.35371932e-02 -1.51446402e-01 -3.97056013e-01 -6.67963207e-01
1.60884067e-01 -6.03928566e-01 1.79548666e-01 -1.08594966e+00
-1.17225122e+00 2.70281285e-01 1.11329727e-01 -9.51583564e-01
7.91436434e-02 -8.16677928e-01 -7.42006123e-01 1.35610080e+00
-9.09648597e-01 -5.53217649e-01 -1.88146055e-01 7.36397445e-01
3.84973526e-01 -1.42409438e-02 5.25312603e-01 -8.15336883e-01
1.32342353e-01 1.35614619e-01 -2.57839411e-01 -2.17391700e-01
8.58966291e-01 -1.27885675e+00 4.99706268e-01 6.53188944e-01
1.71482578e-01 7.08424747e-01 8.92996073e-01 -5.70538402e-01
-6.68776631e-01 -1.78640753e-01 1.29423428e+00 -3.23190302e-01
3.32868807e-02 -3.10076445e-01 -2.11741984e-01 8.79962593e-02
-3.56970951e-02 -1.96285695e-01 7.60470688e-01 -3.25912088e-01
-5.52810252e-01 1.87504590e-01 -1.45366573e+00 1.08680308e+00
1.19924355e+00 -7.12407410e-01 -9.66073513e-01 1.48818150e-01
4.23733145e-01 7.37992078e-02 -2.42724031e-01 2.84740448e-01
1.01790893e+00 -1.67117512e+00 7.46606112e-01 -6.59746170e-01
-3.66934650e-02 -5.42509198e-01 -1.54751420e-01 -1.23983598e+00
-8.60952973e-01 -5.99587262e-01 1.23013210e+00 3.67195725e-01
1.21385045e-01 -1.16392612e+00 2.60281395e-02 6.69229329e-01
1.00324295e-01 2.55806297e-02 -1.10215449e+00 -9.25582767e-01
-9.32189971e-02 4.84413505e-02 3.60933691e-01 5.75223565e-01
2.02357501e-01 1.19479612e-01 4.00809169e-01 -1.67156488e-01
5.82385123e-01 5.45676589e-01 2.88570404e-01 -1.40633965e+00
-1.57360867e-01 -9.20499623e-01 -9.52401876e-01 -3.15880805e-01
-3.42547894e-01 -4.69140768e-01 -2.20784143e-01 -1.00428557e+00
1.51183113e-01 2.27161780e-01 -3.81440222e-01 -9.76240039e-02
1.13395572e-01 3.00308824e-01 4.59591955e-01 4.40009609e-02
2.11649045e-01 2.56437004e-01 1.84894323e+00 1.34955719e-01
-5.52947998e-01 5.96228950e-02 -8.71739745e-01 6.97753429e-01
7.82265842e-01 -3.22711200e-01 -6.02213502e-01 1.10151023e-01
5.40805221e-01 -1.88150816e-02 7.71169484e-01 -1.02817941e+00
3.39547135e-02 -4.89532351e-01 8.65272760e-01 -1.59527004e-01
2.29191855e-01 -8.36907029e-01 1.07890010e-01 7.65437603e-01
-3.89510840e-01 2.11866364e-01 4.13009763e-01 4.23748016e-01
4.84804958e-01 -1.65182829e-01 1.24315417e+00 -5.04216731e-01
-8.05276453e-01 -2.45825678e-01 -4.29358929e-01 1.09330211e-02
1.20465386e+00 -7.65786409e-01 -3.61900985e-01 -1.14850909e-01
-8.07307124e-01 -5.28609037e-01 8.11547637e-01 3.50687534e-01
5.92543960e-01 -1.09859586e+00 -6.60542130e-01 3.94335061e-01
2.11257353e-01 -8.89084280e-01 3.21448743e-01 1.02431405e+00
-4.83774632e-01 5.95639050e-01 -1.15040672e+00 -4.60306615e-01
-1.39930284e+00 1.05823135e+00 6.45809472e-01 6.74842894e-01
-1.44972652e-01 9.64476287e-01 5.63714921e-01 7.93848112e-02
-2.88656592e-01 -5.45046210e-01 -3.58122319e-01 -2.45891102e-02
4.65656340e-01 3.11820716e-01 -3.00951123e-01 -6.36788964e-01
-3.57918382e-01 1.19470167e+00 2.61343002e-01 -4.33569551e-01
5.14212132e-01 -4.07213151e-01 -2.64829129e-01 9.14763510e-01
5.63248992e-01 3.77669454e-01 -1.10063863e+00 3.77328098e-02
-5.77568293e-01 -1.05535221e+00 -4.32479739e-01 -1.09691036e+00
-1.93818733e-01 6.44309342e-01 7.54206836e-01 2.38703623e-01
1.24342525e+00 -4.31601815e-02 -3.30067694e-01 4.69114482e-01
5.64069450e-01 -1.01543522e+00 8.33747536e-02 -4.62595932e-02
1.39560032e+00 -1.10959387e+00 7.99167752e-02 -3.66007864e-01
-7.82597184e-01 1.10984921e+00 6.00239038e-01 -1.73901275e-01
2.76988178e-01 -1.76085293e-01 1.72668651e-01 -3.40645492e-01
-7.41789043e-01 -6.06795251e-01 6.84195220e-01 1.07979310e+00
4.61718112e-01 2.14206517e-01 -6.71180725e-01 -3.52791369e-01
-4.38573539e-01 -3.73883724e-01 5.28791726e-01 5.61717808e-01
-8.18062723e-01 -5.12823403e-01 -8.36454570e-01 1.90661058e-01
-2.06885133e-02 -2.56569624e-01 -8.05970311e-01 7.56587923e-01
3.97946566e-01 8.78696918e-01 5.27191758e-01 1.14364572e-01
5.30933678e-01 1.23947002e-01 1.28677738e+00 -3.75766516e-01
-3.20905536e-01 5.27497791e-02 -1.48431808e-01 -5.59265375e-01
-4.29622561e-01 -6.40737057e-01 -9.82596338e-01 -9.07067433e-02
6.42879903e-02 -3.05162698e-01 5.63223600e-01 6.88414872e-01
-4.58526537e-02 1.68248057e-01 1.87507257e-01 -9.18799996e-01
-5.34936965e-01 -6.95796013e-01 -1.06695175e+00 4.78517890e-01
1.82611316e-01 -6.41972661e-01 -5.37426889e-01 5.38689978e-02] | [10.114282608032227, 2.0806307792663574] |
2d7aaeab-9673-4cb9-ac5d-a2e2bc13695b | discovering-representation-sprachbund-for | 2109.00271 | null | https://arxiv.org/abs/2109.00271v1 | https://arxiv.org/pdf/2109.00271v1.pdf | Discovering Representation Sprachbund For Multilingual Pre-Training | Multilingual pre-trained models have demonstrated their effectiveness in many multilingual NLP tasks and enabled zero-shot or few-shot transfer from high-resource languages to low resource ones. However, due to significant typological differences and contradictions between some languages, such models usually perform poorly on many languages and cross-lingual settings, which shows the difficulty of learning a single model to handle massive diverse languages well at the same time. To alleviate this issue, we present a new multilingual pre-training pipeline. We propose to generate language representation from multilingual pre-trained models and conduct linguistic analysis to show that language representation similarity reflect linguistic similarity from multiple perspectives, including language family, geographical sprachbund, lexicostatistics and syntax. Then we cluster all the target languages into multiple groups and name each group as a representation sprachbund. Thus, languages in the same representation sprachbund are supposed to boost each other in both pre-training and fine-tuning as they share rich linguistic similarity. We pre-train one multilingual model for each representation sprachbund. Experiments are conducted on cross-lingual benchmarks and significant improvements are achieved compared to strong baselines. | ['Nan Duan', 'Ming Zhou', 'Houqiang Li', 'Hany Hassan', 'Alexandre Muzio', 'Yaobo Liang', 'Yimin Fan'] | 2021-09-01 | null | https://aclanthology.org/2021.findings-emnlp.75 | https://aclanthology.org/2021.findings-emnlp.75.pdf | findings-emnlp-2021-11 | ['multilingual-nlp'] | ['natural-language-processing'] | [-2.65961587e-01 -2.69686401e-01 -5.69749236e-01 -4.67066705e-01
-1.11488152e+00 -7.42531002e-01 7.71839619e-01 1.54207885e-01
-6.39295578e-01 8.73757422e-01 6.40189588e-01 -2.58574367e-01
2.77416825e-01 -7.06526935e-01 -6.84403539e-01 -3.51835370e-01
3.07236731e-01 7.75781214e-01 -8.08654726e-02 -7.03441441e-01
-1.44224152e-01 5.14991730e-02 -1.09230161e+00 5.23713768e-01
1.25067067e+00 1.90888241e-01 3.11520338e-01 2.15557925e-02
-6.57078505e-01 6.26648486e-01 -3.38518411e-01 -6.34470701e-01
2.17346489e-01 -3.71999532e-01 -7.20575333e-01 -2.08226919e-01
4.81874615e-01 2.92483985e-01 -9.08190478e-03 1.21768594e+00
5.79153895e-01 1.09310009e-01 7.63074756e-01 -7.30442941e-01
-9.54489708e-01 1.33675408e+00 -6.51482582e-01 1.43154666e-01
4.26364869e-01 4.21055174e-03 1.26017356e+00 -1.16050315e+00
9.35053468e-01 1.65859449e+00 6.56996369e-01 3.68337452e-01
-1.35996568e+00 -8.01776826e-01 4.01225507e-01 -2.53200289e-02
-1.59491432e+00 -5.39056361e-01 4.87877607e-01 -4.46633160e-01
1.13782847e+00 -1.28518239e-01 1.52312651e-01 1.31077671e+00
-3.66242789e-02 6.82254314e-01 1.14543211e+00 -5.45365930e-01
-3.46872270e-01 4.31941062e-01 8.03546607e-02 5.79057097e-01
1.29879341e-01 -3.09298098e-01 -3.45198482e-01 6.08369410e-02
3.88010561e-01 -2.23242939e-01 -1.88320175e-01 -2.84541219e-01
-1.60850525e+00 9.84745204e-01 5.22521615e-01 8.52100611e-01
-1.68227717e-01 -3.18739176e-01 8.26617956e-01 4.43316281e-01
5.28922677e-01 5.62025964e-01 -5.95865607e-01 3.24248821e-01
-6.06248498e-01 -1.78973023e-02 5.42534828e-01 1.13728321e+00
9.64465082e-01 4.65988107e-02 -2.39644185e-01 1.41937315e+00
5.32441735e-02 5.93398035e-01 8.85698259e-01 -2.31276438e-01
8.98954988e-01 6.25763834e-01 -2.68247873e-01 -5.63839138e-01
-2.93873966e-01 -4.01681662e-01 -9.42207813e-01 -4.11446810e-01
2.81714529e-01 -8.16472471e-02 -7.49836922e-01 1.89491904e+00
8.93208310e-02 -1.14642993e-01 5.93849480e-01 5.91396630e-01
8.65811646e-01 1.03520334e+00 3.21089327e-01 -9.54358056e-02
1.45803058e+00 -1.12941253e+00 -5.17096460e-01 -4.29593444e-01
1.02591944e+00 -1.07448423e+00 1.63930237e+00 1.35856211e-01
-7.79115558e-01 -6.42760873e-01 -7.44518995e-01 -3.67565870e-01
-7.37611353e-01 2.90099323e-01 4.92159098e-01 1.94449529e-01
-7.50405192e-01 1.77486241e-01 -3.30950826e-01 -5.58734000e-01
-5.27044423e-02 -1.70673981e-01 -5.14679492e-01 -4.33862597e-01
-1.69553161e+00 1.11321175e+00 7.90776610e-01 -2.98329711e-01
-6.61739171e-01 -8.52977633e-01 -1.12773013e+00 -1.76441103e-01
1.60628840e-01 -5.57119489e-01 9.19313192e-01 -1.01495636e+00
-1.15784252e+00 1.32309675e+00 -4.62242365e-02 -2.89751112e-01
3.42555255e-01 -5.38375266e-02 -8.27337325e-01 -5.45507371e-01
6.91934168e-01 5.76053500e-01 2.75041908e-01 -1.09439087e+00
-9.30524468e-01 -9.29211825e-02 -4.01041768e-02 4.03264850e-01
-4.09335732e-01 4.87296969e-01 -4.56024975e-01 -8.31530929e-01
-1.66689128e-01 -7.73173809e-01 -1.93271935e-01 -7.04561055e-01
-2.95887411e-01 -5.12297034e-01 1.79702312e-01 -6.57977521e-01
1.26896334e+00 -1.97031462e+00 1.79509133e-01 -6.02374747e-02
-2.58583754e-01 3.27178806e-01 -6.03354812e-01 5.85266352e-01
-8.11206922e-02 1.27982527e-01 -1.37155652e-01 -3.68097991e-01
4.74255495e-02 3.87294799e-01 -3.67250085e-01 2.69072533e-01
1.60634875e-01 8.93530309e-01 -1.30600309e+00 -4.49382752e-01
1.18194900e-01 5.60743928e-01 -3.71165246e-01 4.44386825e-02
-1.31410420e-01 6.48578167e-01 -2.21706793e-01 6.72558546e-01
4.15773481e-01 1.79476991e-01 4.45725828e-01 -3.73082131e-01
-3.14237386e-01 7.99872279e-01 -8.61826241e-01 2.03941941e+00
-1.11606979e+00 2.06894115e-01 -2.08352551e-01 -9.03407812e-01
1.05945361e+00 3.07556033e-01 2.03356758e-01 -8.29018950e-01
-6.42957464e-02 7.64851034e-01 1.88840643e-01 -3.19417089e-01
7.18658209e-01 -3.85458946e-01 -6.74382210e-01 4.31028992e-01
5.02951980e-01 -1.89197045e-02 5.52510858e-01 6.27817288e-02
4.43832546e-01 2.64534593e-01 7.22296298e-01 -5.70123494e-01
7.59121656e-01 -2.89268699e-02 6.69022501e-01 4.48915571e-01
-1.25697076e-01 2.95794457e-01 2.34157145e-02 -3.83845478e-01
-1.04095113e+00 -1.09987903e+00 -1.89360425e-01 1.81875288e+00
-4.08356935e-02 -4.55494255e-01 -3.28746200e-01 -8.06693196e-01
-7.67869353e-02 8.45362842e-01 -3.34941298e-01 -1.06760085e-01
-7.86077797e-01 -8.16664636e-01 7.79888809e-01 2.34279081e-01
1.18323684e-01 -1.26822186e+00 2.15176433e-01 4.07161832e-01
-3.77124429e-01 -1.21789968e+00 -6.15357757e-01 1.93645865e-01
-1.97451711e-01 -5.75901568e-01 -9.10710871e-01 -1.32015049e+00
4.24408793e-01 3.44799608e-01 1.62560725e+00 -3.64695281e-01
1.04074448e-01 -1.53039247e-01 -3.73606384e-01 -1.82163879e-01
-6.68098092e-01 5.05029142e-01 3.38376790e-01 -2.75329296e-02
6.40323341e-01 -4.42709416e-01 -4.62144148e-03 1.10213570e-01
-4.90359873e-01 -4.58787717e-02 4.69049335e-01 6.59156561e-01
8.05139601e-01 -2.56693989e-01 6.64317906e-01 -1.03299499e+00
6.71978354e-01 -7.53794312e-01 -3.69412750e-01 6.72212243e-01
-2.56387860e-01 2.06853241e-01 1.01766837e+00 -5.34438550e-01
-9.77054894e-01 -1.53053746e-01 -7.04834312e-02 -2.30923787e-01
-1.12175964e-01 8.39425564e-01 -4.21309084e-01 2.67997235e-01
7.11721420e-01 2.75922388e-01 -5.48219681e-01 -6.80801630e-01
7.77285218e-01 6.64176464e-01 5.20097017e-01 -9.09836948e-01
6.40758872e-01 3.55800539e-02 -6.47247553e-01 -6.79980636e-01
-1.17559838e+00 -4.64849323e-01 -7.96192825e-01 2.29690686e-01
6.02120936e-01 -1.36530256e+00 3.63096222e-02 6.11912869e-02
-1.21565139e+00 -7.98195079e-02 -2.94892013e-01 6.17538095e-01
-9.16234031e-02 1.32069334e-01 -5.50735176e-01 -3.37754101e-01
-3.59930634e-01 -1.21667421e+00 9.92990196e-01 -1.30458638e-01
-2.43026823e-01 -1.39604712e+00 4.91979063e-01 1.37104765e-01
3.23000163e-01 1.17084779e-01 1.12053227e+00 -9.38872457e-01
-2.02174410e-02 3.54440324e-02 -1.58157095e-01 2.52608597e-01
3.96080136e-01 -1.02552831e-01 -7.09803581e-01 -4.84290808e-01
-4.19587940e-01 -6.19687736e-01 7.94376969e-01 -8.19422025e-03
4.83207136e-01 -2.18964234e-01 -1.96409345e-01 7.35136032e-01
1.43671954e+00 -2.71870971e-01 1.99505299e-01 4.56857383e-01
9.88235950e-01 8.48875642e-01 5.19702733e-01 -1.14565983e-01
7.52699018e-01 7.24688172e-01 -2.29623586e-01 -2.78014749e-01
-3.22125345e-01 -4.96375024e-01 8.47924054e-01 1.66543996e+00
1.70487970e-01 -1.12683788e-01 -1.21377289e+00 8.72837424e-01
-1.69117069e+00 -6.90007985e-01 -1.81026876e-01 2.24530387e+00
1.26684678e+00 2.64511891e-02 2.70390026e-02 -6.23726010e-01
8.27080548e-01 9.31288302e-02 -2.72669941e-01 -5.27956307e-01
-6.84731603e-01 1.11355118e-01 3.32469136e-01 6.84499621e-01
-9.93742764e-01 1.81708086e+00 5.67236328e+00 1.09634268e+00
-1.34716177e+00 2.99561739e-01 3.45300555e-01 1.99957341e-01
-5.54230452e-01 1.05446376e-01 -1.21121275e+00 4.36113000e-01
8.22082341e-01 -7.25840807e-01 4.43136394e-01 7.01456130e-01
-4.96441573e-02 5.15612841e-01 -1.12955403e+00 8.22626710e-01
1.52214304e-01 -8.69476855e-01 6.58682227e-01 -1.79749936e-01
1.03836751e+00 6.12531364e-01 -1.39672905e-01 9.16876078e-01
9.01423752e-01 -1.06942213e+00 6.60012305e-01 1.54585183e-01
9.97484863e-01 -9.25048053e-01 5.55542529e-01 4.05796021e-01
-1.62025690e+00 3.48508239e-01 -7.47394443e-01 2.12031037e-01
3.09760779e-01 4.38872695e-01 -8.09215665e-01 9.73033249e-01
4.08740103e-01 1.02027071e+00 -5.70268035e-01 4.20638472e-01
-3.90465438e-01 3.96375299e-01 -8.20800141e-02 2.80347109e-01
6.69258893e-01 -3.59364986e-01 4.38356787e-01 1.68510091e+00
4.72962528e-01 -5.83643913e-01 7.41235316e-01 7.09919035e-01
-2.37704009e-01 8.71323228e-01 -8.84838760e-01 -1.60900518e-01
5.04331350e-01 1.27646828e+00 -2.57963330e-01 -6.33714080e-01
-8.14377069e-01 8.09964836e-01 8.58432531e-01 2.50423193e-01
-7.08476245e-01 -3.07499081e-01 7.00494349e-01 2.42341515e-02
-4.73261774e-02 -7.83287659e-02 1.12774633e-01 -1.67479360e+00
-2.00748190e-01 -1.09662032e+00 7.34881759e-01 -3.74494880e-01
-1.89078569e+00 1.07476366e+00 -1.53959319e-01 -1.35078573e+00
-4.62091327e-01 -4.55186695e-01 -2.01035410e-01 1.20200455e+00
-1.86654258e+00 -1.67937553e+00 2.15580285e-01 8.56343329e-01
8.19924772e-01 -6.85411572e-01 9.80015993e-01 5.50667703e-01
-5.05283415e-01 8.05154741e-01 2.02463031e-01 2.77879745e-01
1.22664964e+00 -1.14414239e+00 4.34670448e-01 8.41339409e-01
4.78140563e-01 8.25804353e-01 3.67945820e-01 -5.70113480e-01
-9.72798765e-01 -1.41462517e+00 1.63386297e+00 -3.12778711e-01
1.21516359e+00 -4.76239830e-01 -1.26518524e+00 8.20318043e-01
5.82145870e-01 -4.01334018e-02 8.64008129e-01 5.66692710e-01
-7.67213345e-01 1.69079024e-02 -6.84070230e-01 7.78245032e-01
9.92308795e-01 -8.31136346e-01 -8.30828488e-01 5.50938845e-01
7.16813207e-01 -2.57224240e-03 -9.75172520e-01 4.11272317e-01
1.90811113e-01 -5.46870232e-01 8.93024385e-01 -7.86023557e-01
3.25857788e-01 -2.17545897e-01 -3.18235964e-01 -1.83978593e+00
-5.04309654e-01 -4.83346313e-01 5.34912825e-01 1.70319450e+00
7.17936993e-01 -6.40403688e-01 -1.68336213e-01 -2.96324492e-01
-2.59388775e-01 -2.52137631e-01 -7.44643033e-01 -1.07129979e+00
7.65386462e-01 -1.73641786e-01 5.93424618e-01 1.72463334e+00
1.28098100e-01 1.01643872e+00 -4.53698397e-01 -9.68577415e-02
4.19227630e-01 1.70749426e-01 7.18991578e-01 -1.06013668e+00
-8.81372690e-02 -6.91622972e-01 -7.20019937e-02 -7.49956250e-01
7.01421380e-01 -1.78916585e+00 1.50893241e-01 -1.36304951e+00
3.31035733e-01 -5.65732598e-01 -5.67537129e-01 5.74284792e-01
-4.03226793e-01 1.84430897e-01 5.53604364e-02 2.55255818e-01
-6.12562537e-01 4.97413248e-01 1.07397175e+00 -2.00246394e-01
-2.44171947e-01 -4.47600812e-01 -7.57849813e-01 7.99199700e-01
5.89918017e-01 -4.40521210e-01 -1.86027557e-01 -9.19371128e-01
2.41425738e-01 -4.46233094e-01 -2.65042216e-01 -6.95964754e-01
-1.18526988e-01 -1.86150178e-01 -7.09311143e-02 -2.32688785e-01
-7.06553757e-02 -5.69570541e-01 -6.56231567e-02 3.11529785e-01
-4.51222390e-01 1.75720215e-01 1.22621372e-01 9.83298570e-02
-5.20646095e-01 -1.02674529e-01 9.06870484e-01 -3.92387658e-01
-9.61335182e-01 3.06814313e-01 -1.96910352e-01 5.31200945e-01
6.87519848e-01 1.96304128e-01 -1.77310660e-01 -3.81808868e-03
-5.29192686e-01 3.15307766e-01 4.50753361e-01 9.55802858e-01
-5.13083078e-02 -1.72505462e+00 -1.30046046e+00 1.89782724e-01
7.98795938e-01 -2.73975939e-01 2.62165833e-02 5.78911960e-01
-2.74131060e-01 4.55164671e-01 -2.54039288e-01 -4.70658183e-01
-9.49685216e-01 6.80337191e-01 4.02549565e-01 -8.00856888e-01
-5.66064298e-01 7.74633229e-01 5.83615243e-01 -1.24013793e+00
7.34175295e-02 -2.22136110e-01 -3.98355007e-01 3.09973776e-01
3.67375404e-01 -1.25001207e-01 -2.73807114e-03 -1.36203516e+00
-4.51356262e-01 8.02900434e-01 -2.60848403e-01 -7.88884610e-02
1.11130452e+00 -2.29211122e-01 -2.60185212e-01 8.39982212e-01
1.24121368e+00 4.37177986e-01 -6.29464865e-01 -8.74107540e-01
3.11530232e-01 -1.43631339e-01 -3.05020422e-01 -7.51466155e-01
-8.92438293e-01 9.30598795e-01 1.53947413e-01 -1.91068530e-01
6.33602321e-01 1.71624735e-01 7.85260320e-01 3.92364651e-01
5.18389344e-01 -1.23984838e+00 -2.41083324e-01 1.22777653e+00
9.09357965e-01 -1.28491354e+00 -3.23408753e-01 -3.29646081e-01
-9.67935503e-01 8.18787098e-01 6.40434027e-01 -2.22121581e-01
3.83529216e-01 1.00256816e-01 3.32032472e-01 8.36571530e-02
-7.42423236e-01 -5.01925290e-01 5.93053162e-01 4.50354606e-01
1.17338657e+00 4.06940132e-01 -5.78199923e-01 7.10553706e-01
-5.74961066e-01 -3.74399692e-01 9.47999656e-02 5.00437498e-01
-3.34188759e-01 -1.41342270e+00 -1.49858519e-01 -1.38805747e-01
-3.79510373e-01 -5.52692711e-01 -3.14688295e-01 9.09206867e-01
3.58400732e-01 4.82504457e-01 5.75459450e-02 -1.52850330e-01
3.50301474e-01 2.15415582e-01 3.84622127e-01 -1.01280355e+00
-8.97174478e-01 2.42504492e-01 2.45025828e-01 -2.56331414e-01
-1.80078387e-01 -4.98491675e-01 -1.09054244e+00 -1.15260415e-01
1.37929842e-01 1.88208714e-01 4.11654025e-01 9.26648080e-01
9.30979401e-02 4.73250121e-01 5.81329405e-01 -5.45242965e-01
-4.05663073e-01 -1.13244557e+00 -4.36608881e-01 6.89608157e-01
-1.01255886e-01 -4.89783287e-01 -2.34536648e-01 -4.02729996e-02] | [10.9260892868042, 9.945755958557129] |
176874c5-4c3b-417e-ae4b-10cdf2cf9795 | a-fast-and-robust-bert-based-dialogue-state | 2008.12335 | null | https://arxiv.org/abs/2008.12335v1 | https://arxiv.org/pdf/2008.12335v1.pdf | A Fast and Robust BERT-based Dialogue State Tracker for Schema-Guided Dialogue Dataset | Dialog State Tracking (DST) is one of the most crucial modules for goal-oriented dialogue systems. In this paper, we introduce FastSGT (Fast Schema Guided Tracker), a fast and robust BERT-based model for state tracking in goal-oriented dialogue systems. The proposed model is designed for the Schema-Guided Dialogue (SGD) dataset which contains natural language descriptions for all the entities including user intents, services, and slots. The model incorporates two carry-over procedures for handling the extraction of the values not explicitly mentioned in the current user utterance. It also uses multi-head attention projections in some of the decoders to have a better modelling of the encoder outputs. In the conducted experiments we compared FastSGT to the baseline model for the SGD dataset. Our model keeps the efficiency in terms of computational and memory consumption while improving the accuracy significantly. Additionally, we present ablation studies measuring the impact of different parts of the model on its performance. We also show the effectiveness of data augmentation for improving the accuracy without increasing the amount of computational resources. | ['Yang Zhang', 'Evelina Bakhturina', 'Vahid Noroozi', 'Tomasz Kornuta'] | 2020-08-27 | null | null | null | null | ['goal-oriented-dialogue-systems'] | ['natural-language-processing'] | [-7.73087218e-02 8.60572338e-01 1.36314169e-01 -6.24657929e-01
-5.64221323e-01 -5.34561515e-01 1.00162101e+00 1.79955751e-01
-6.52089298e-01 6.39441192e-01 6.96444333e-01 -2.34534577e-01
1.50549918e-01 -4.19818938e-01 -9.35201496e-02 -3.08572233e-01
8.26963410e-02 9.62124884e-01 4.67361391e-01 -8.06940854e-01
3.77484441e-01 1.88321043e-02 -9.77245331e-01 3.84730458e-01
7.57901013e-01 7.08282530e-01 5.12497663e-01 8.92262340e-01
-4.83226240e-01 1.05849326e+00 -7.31695831e-01 -3.38484317e-01
-7.74957612e-02 -5.20897210e-01 -1.13943541e+00 2.27539033e-01
8.01717117e-02 -5.20922840e-01 -3.60640854e-01 8.72241676e-01
5.95391214e-01 3.97662044e-01 4.29715902e-01 -1.20143509e+00
2.99126599e-02 8.17674160e-01 6.55852631e-03 4.24820855e-02
6.55942500e-01 2.74023890e-01 8.54736388e-01 -6.39902890e-01
7.38232911e-01 1.62715662e+00 1.75477132e-01 9.33291495e-01
-7.40376234e-01 -2.06476934e-02 2.50693709e-01 2.31003314e-02
-7.23372757e-01 -9.08115506e-01 4.56082463e-01 -2.67123818e-01
1.31507361e+00 2.43709937e-01 2.63994575e-01 8.83492589e-01
-1.42805979e-01 9.49597836e-01 8.29831004e-01 -6.33317590e-01
2.85630733e-01 5.25284469e-01 6.43609583e-01 8.84782910e-01
-4.24855262e-01 -2.37677544e-01 -6.77365780e-01 -1.80099830e-01
5.59148431e-01 -6.57161057e-01 -1.59179136e-01 -3.67055178e-01
-1.03000653e+00 9.41201806e-01 -2.88025178e-02 3.07885975e-01
-2.98076361e-01 -3.17028075e-01 7.41467535e-01 1.25825524e-01
2.56115496e-01 3.97208035e-01 -5.84895074e-01 -6.77729547e-01
-4.66888070e-01 1.91360936e-01 1.22408426e+00 1.25128603e+00
1.15780555e-01 -3.55914645e-02 -8.97540271e-01 1.01610124e+00
5.10660887e-01 3.18456478e-02 5.13899565e-01 -9.21926022e-01
8.16714227e-01 1.04172063e+00 4.71179128e-01 -2.82775164e-01
-7.62404382e-01 -4.66933213e-02 -4.08781976e-01 -1.61391065e-01
7.27500916e-01 -5.23978472e-01 -7.23925412e-01 1.84593260e+00
4.32212591e-01 -3.82529318e-01 4.75933284e-01 8.73529196e-01
1.09885693e+00 8.39932323e-01 3.02328259e-01 -1.85622349e-01
1.78299010e+00 -1.22251797e+00 -1.15884578e+00 -3.47325534e-01
9.36683238e-01 -5.16299725e-01 1.15088975e+00 -4.79024947e-02
-1.13842344e+00 -3.19814652e-01 -7.74990916e-01 -3.46048295e-01
-3.69663626e-01 3.64844650e-01 4.45487440e-01 6.21033132e-01
-9.50398505e-01 3.67567420e-01 -7.42452681e-01 -7.62180030e-01
-3.15503120e-01 2.62624979e-01 -2.26145208e-01 4.82821256e-01
-1.35616159e+00 1.37223566e+00 7.13617325e-01 -7.85555318e-02
-6.64006412e-01 1.33862123e-02 -1.06111085e+00 3.40443045e-01
5.51767766e-01 -3.45189691e-01 1.72385740e+00 -2.18535140e-01
-1.88130927e+00 6.46932483e-01 -2.96807915e-01 -7.23850548e-01
5.13578534e-01 -3.15150380e-01 -5.02605923e-02 -1.79028258e-01
-1.29063025e-01 7.34621763e-01 3.40922028e-01 -6.66262448e-01
-7.49592721e-01 -4.34092283e-01 3.93675447e-01 7.45319784e-01
-1.22821085e-01 9.12861824e-02 -7.66545773e-01 -8.04673210e-02
-1.82538200e-02 -1.02483726e+00 -2.33246028e-01 -5.35160840e-01
-5.96235037e-01 -5.35530746e-01 7.10251153e-01 -9.31979656e-01
1.37346172e+00 -2.02187991e+00 3.18608791e-01 -4.16426033e-01
-2.71036685e-01 5.36089659e-01 3.30471322e-02 7.40397573e-01
3.59583110e-01 -2.22755998e-01 -2.81242371e-01 -5.93303919e-01
3.32627982e-01 3.41663420e-01 5.32329157e-02 -6.14221580e-02
1.28283948e-01 6.66148305e-01 -6.61316752e-01 -5.26938677e-01
5.13090134e-01 7.28911832e-02 -4.60114479e-01 7.35082328e-01
-5.94073594e-01 5.37213504e-01 -2.61120439e-01 4.72995080e-02
3.05294752e-01 4.63277660e-02 4.00868535e-01 -6.04856350e-02
-2.17399299e-01 1.04081488e+00 -1.13576984e+00 1.94323480e+00
-4.49489921e-01 2.97359526e-01 2.05004692e-01 -3.68916005e-01
7.72886157e-01 6.40853345e-01 -6.64862338e-03 -7.13995457e-01
2.47066081e-01 -1.81690782e-01 2.24192232e-01 -5.36560297e-01
8.95408094e-01 2.55896896e-01 -4.23219323e-01 5.10867536e-01
4.45872426e-01 9.65520740e-02 4.16021496e-01 5.22012293e-01
7.28324115e-01 8.17201436e-02 6.54004335e-01 -2.62959689e-01
9.40570354e-01 1.58508718e-01 3.48836750e-01 5.85136473e-01
-3.36039424e-01 -1.99036244e-02 8.25308502e-01 -7.03243315e-02
-8.80058587e-01 -3.17757159e-01 2.41486713e-01 1.50815153e+00
-1.04235515e-01 -6.29887998e-01 -1.11542439e+00 -8.68608892e-01
-3.76248926e-01 1.30135238e+00 -3.23736459e-01 -1.16627172e-01
-5.64972579e-01 -7.07145154e-01 5.52017629e-01 3.04784805e-01
8.18448126e-01 -1.22087526e+00 -6.70269728e-01 2.97001660e-01
-2.98337877e-01 -1.35897374e+00 -4.07984614e-01 2.18217179e-01
-7.06505954e-01 -9.92331624e-01 -4.74957079e-01 -5.83348513e-01
3.31370234e-01 -1.37541384e-01 8.02732944e-01 -7.77823254e-02
4.31858212e-01 2.01629922e-01 -3.73533487e-01 -3.63526851e-01
-9.84715939e-01 4.46611375e-01 -1.54939264e-01 -1.03389308e-01
3.68302494e-01 1.81042284e-01 -3.18631269e-02 2.47725978e-01
-3.82608086e-01 4.85991597e-01 2.08057195e-01 8.30848753e-01
-2.73898840e-01 -3.72068793e-01 5.10385275e-01 -1.21375978e+00
1.17554414e+00 -2.44315073e-01 -6.75696909e-01 3.69853973e-01
-4.45438385e-01 6.32278144e-01 3.69202584e-01 2.54076459e-02
-1.50706911e+00 5.56494035e-02 -5.61841309e-01 3.50396037e-01
-3.43057483e-01 3.36148858e-01 -2.82710791e-01 3.31478626e-01
4.33082789e-01 3.54803801e-01 4.66063358e-02 -7.83743978e-01
3.72840822e-01 1.05548990e+00 3.41772646e-01 -4.41445231e-01
1.77151471e-01 -2.16072828e-01 -3.80840182e-01 -9.58426714e-01
-7.24459589e-01 -6.73320055e-01 -7.32366443e-01 -1.40433386e-01
7.75770664e-01 -6.79446042e-01 -9.25466776e-01 5.21409810e-01
-1.22020674e+00 -4.93406564e-01 -9.24965069e-02 2.08212882e-01
-6.53264344e-01 5.37156105e-01 -8.17942560e-01 -1.42908704e+00
-7.46858299e-01 -1.15303969e+00 9.35888231e-01 3.32652211e-01
-3.11938167e-01 -1.08584750e+00 1.85001511e-02 3.71018797e-01
4.85166609e-01 -3.55517387e-01 9.92104650e-01 -1.46840811e+00
-1.12104565e-01 1.07655376e-01 -1.28172800e-01 8.14182758e-02
-1.90393813e-02 -5.35450518e-01 -9.74354804e-01 -1.68155804e-01
1.27364531e-01 -2.72627801e-01 4.07632321e-01 7.73310056e-03
1.71215549e-01 -6.22615576e-01 -2.90844608e-02 -1.82953775e-02
9.83350396e-01 7.41700828e-01 3.87832850e-01 3.18415314e-01
4.33226138e-01 9.38955307e-01 1.12209094e+00 4.82880056e-01
6.24832511e-01 1.14204454e+00 3.06102246e-01 1.20447777e-01
3.75142880e-02 -2.30828270e-01 5.75556576e-01 7.02971876e-01
3.61716300e-01 -5.10529280e-01 -8.47842157e-01 5.91873169e-01
-2.11529660e+00 -6.86335504e-01 -1.23099901e-01 2.14645958e+00
8.20336401e-01 1.98208943e-01 3.85728002e-01 -6.51246235e-02
6.31534815e-01 2.92825460e-01 -4.54417616e-01 -9.81943548e-01
3.03232700e-01 -2.86966294e-01 3.01097035e-01 9.57001984e-01
-1.02888596e+00 1.38422775e+00 5.64213848e+00 3.15557837e-01
-6.38833702e-01 2.04364136e-01 2.56440431e-01 2.28333399e-01
3.62704068e-01 -1.63134485e-02 -1.21756041e+00 4.44157422e-01
1.25793326e+00 -1.00448892e-01 2.01344013e-01 9.48798418e-01
3.46050024e-01 -3.69379550e-01 -9.49900866e-01 4.09420013e-01
-1.26131237e-01 -9.69013512e-01 -1.35558233e-01 -1.66984946e-01
3.36418226e-02 -1.79560572e-01 -5.10400891e-01 6.49162292e-01
5.07204950e-01 -3.78639370e-01 5.39667308e-01 2.95407742e-01
4.06380594e-01 -4.98679399e-01 9.85980868e-01 8.20296705e-01
-6.18473291e-01 -9.10663009e-02 -1.85501203e-01 2.23773681e-02
4.39002723e-01 -1.69026241e-01 -1.63382065e+00 3.02299768e-01
9.43519920e-02 1.62854359e-01 -5.09722054e-01 7.33293176e-01
-3.58757287e-01 5.61832368e-01 -3.32087904e-01 -3.82271618e-01
5.14578581e-01 -1.16575412e-01 7.39096761e-01 1.38954818e+00
-5.64063638e-02 1.51800141e-01 1.31384075e-01 6.00828409e-01
1.84219271e-01 2.09031716e-01 -4.52878028e-01 -9.97830853e-02
5.54252267e-01 1.11603570e+00 -2.21133053e-01 -5.70103228e-01
-2.50842690e-01 9.36179459e-01 2.61205316e-01 -7.06683919e-02
-6.54723585e-01 -1.39302120e-01 4.61222440e-01 -1.25746444e-01
5.81620187e-02 -2.29009658e-01 3.86403663e-05 -1.00839293e+00
-3.04933280e-01 -8.98783207e-01 6.07507467e-01 -6.99604869e-01
-5.37058532e-01 7.80691206e-01 7.76447877e-02 -6.46731496e-01
-8.03525150e-01 -3.06343347e-01 -3.97886634e-01 9.88817275e-01
-1.22175884e+00 -9.98617947e-01 -1.92436710e-01 5.13977468e-01
1.09509480e+00 -1.67952076e-01 1.09837210e+00 -1.13075629e-01
-8.94228458e-01 4.37843353e-01 -3.08767885e-01 4.51361895e-01
7.07652152e-01 -1.48860478e+00 8.84953439e-01 8.65567744e-01
-1.22773580e-01 6.00364268e-01 8.67741466e-01 -6.98701203e-01
-1.04377794e+00 -7.55078197e-01 1.22677612e+00 -3.57850581e-01
2.22751036e-01 -5.53396106e-01 -8.93134534e-01 8.25702190e-01
5.62399805e-01 -9.15430307e-01 2.30307043e-01 2.16657564e-01
2.37071320e-01 4.73819733e-01 -1.26872098e+00 5.26558042e-01
6.62568152e-01 -2.85786062e-01 -9.90338564e-01 2.63542384e-01
7.55830407e-01 -9.07228827e-01 -8.94783556e-01 -4.05967236e-02
1.79216966e-01 -9.58481729e-01 5.87882042e-01 -6.82484686e-01
-1.00450836e-01 -6.31526858e-02 1.51279550e-02 -1.23346233e+00
-1.28502473e-01 -7.03691602e-01 -3.96449268e-01 1.41242146e+00
4.42417771e-01 -3.45079005e-01 7.30155706e-01 1.17741299e+00
-2.67116725e-01 -2.76563287e-01 -8.96006882e-01 -3.26180995e-01
-3.83703232e-01 -1.21554723e-02 4.97872710e-01 6.07612908e-01
5.65290391e-01 1.10266352e+00 -6.06594324e-01 2.85112411e-02
2.25291789e-01 -1.82887018e-01 8.95963192e-01 -9.37343895e-01
6.82144938e-03 2.51693167e-02 4.19445261e-02 -1.24995184e+00
-1.54106364e-01 -4.19277728e-01 1.56606540e-01 -1.70817173e+00
-1.34699404e-01 -1.50341392e-01 2.34885499e-01 6.17589593e-01
-2.49305055e-01 -7.47311473e-01 5.06834149e-01 -2.32320540e-02
-7.51417160e-01 6.46741092e-01 9.83594537e-01 1.19493075e-01
-5.92097044e-01 3.03002030e-01 -5.72358012e-01 6.79822028e-01
9.62296963e-01 -3.85710835e-01 -4.88673538e-01 -2.15703666e-01
-3.85360271e-01 6.95430875e-01 -1.74267411e-01 -9.42140698e-01
5.69019556e-01 -4.30670008e-02 -3.32243025e-01 -7.49710619e-01
7.25502312e-01 -6.73585653e-01 -4.11337256e-01 5.07562339e-01
-7.68229246e-01 1.23383757e-02 3.11081648e-01 1.63022891e-01
-2.06514867e-03 -7.25329280e-01 8.57019663e-01 -3.64575922e-01
-9.12783206e-01 -2.65826017e-01 -6.27166986e-01 -1.23777438e-03
8.12426746e-01 8.59351829e-02 -4.26409125e-01 -6.66921973e-01
-7.85848379e-01 5.47388852e-01 1.85318768e-01 5.41176856e-01
2.38629758e-01 -6.58950448e-01 -4.19306487e-01 1.22673973e-01
7.70358518e-02 -1.11357674e-01 1.25407919e-01 6.38508379e-01
-3.85205001e-01 1.14372301e+00 -3.43470722e-01 -2.99543172e-01
-1.52925479e+00 3.62052590e-01 5.53620040e-01 -6.84917867e-01
-3.92640173e-01 6.25857592e-01 -3.17488164e-02 -7.98887312e-01
5.72756112e-01 -1.88635886e-01 -8.60467434e-01 1.00228563e-01
4.44244921e-01 3.50752890e-01 3.72744650e-02 -7.35471010e-01
-2.34192476e-01 -2.58894533e-01 -4.43873763e-01 -5.16440928e-01
1.11002493e+00 -3.63598019e-01 1.50340691e-01 5.48260808e-01
5.49623549e-01 -2.53346592e-01 -1.20939851e+00 -2.89941639e-01
5.02734363e-01 -2.02757083e-02 9.99742933e-03 -1.22618413e+00
-5.27194679e-01 8.52624953e-01 4.54031974e-01 4.29467022e-01
6.99871063e-01 -3.38622779e-01 7.16852963e-01 5.15622258e-01
4.50258255e-01 -1.20188117e+00 -2.80736953e-01 1.06086290e+00
7.95160294e-01 -1.26142013e+00 -2.65392810e-01 -4.19299155e-01
-1.22296202e+00 9.00668979e-01 9.65414703e-01 4.64671463e-01
3.22311968e-02 1.89691335e-02 1.76818579e-01 -3.62589449e-01
-1.14051902e+00 -4.29500192e-01 5.66694699e-02 4.62506294e-01
5.42706430e-01 -2.34517708e-01 -6.14442825e-01 5.05515397e-01
-1.14698045e-01 -2.73734897e-01 6.35825396e-01 9.53137338e-01
-6.07181132e-01 -1.24817324e+00 -9.54604968e-02 1.68099299e-01
-3.75376642e-01 -4.77993712e-02 -8.01536024e-01 8.40384603e-01
-5.03302157e-01 1.26965022e+00 -2.17098117e-01 -2.93170422e-01
7.49698281e-01 7.21662581e-01 8.55390728e-02 -9.04431999e-01
-1.11667180e+00 2.18074676e-03 9.18489456e-01 -4.47581053e-01
-3.03008050e-01 -5.65782845e-01 -1.28847587e+00 -5.89161702e-02
-5.39680898e-01 5.27427435e-01 8.85316372e-01 1.02162683e+00
4.31436479e-01 4.49254066e-01 4.27977353e-01 -3.82511407e-01
-6.79101706e-01 -1.63497055e+00 -2.24776536e-01 3.08524251e-01
-1.66283138e-02 -5.83127618e-01 -1.24407168e-02 -2.79161185e-01] | [12.845308303833008, 7.907772541046143] |
e35ffb4b-c6fc-43b1-97ed-0d83ba55ef12 | maza-at-semeval-2016-task-11-detecting | null | null | https://aclanthology.org/S16-1153 | https://aclanthology.org/S16-1153.pdf | MAZA at SemEval-2016 Task 11: Detecting Lexical Complexity Using a Decision Stump Meta-Classifier | null | ['Marcos Zampieri', 'Shervin Malmasi'] | 2016-06-01 | null | null | null | semeval-2016-6 | ['complex-word-identification'] | ['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] | [-7.251986980438232, 3.6041040420532227] |
5397bd81-7854-48ff-b9c1-4b1a89191847 | locate-then-ask-interpretable-stepwise | 2208.10297 | null | https://arxiv.org/abs/2208.10297v1 | https://arxiv.org/pdf/2208.10297v1.pdf | Locate Then Ask: Interpretable Stepwise Reasoning for Multi-hop Question Answering | Multi-hop reasoning requires aggregating multiple documents to answer a complex question. Existing methods usually decompose the multi-hop question into simpler single-hop questions to solve the problem for illustrating the explainable reasoning process. However, they ignore grounding on the supporting facts of each reasoning step, which tends to generate inaccurate decompositions. In this paper, we propose an interpretable stepwise reasoning framework to incorporate both single-hop supporting sentence identification and single-hop question generation at each intermediate step, and utilize the inference of the current hop for the next until reasoning out the final result. We employ a unified reader model for both intermediate hop reasoning and final hop inference and adopt joint optimization for more accurate and robust multi-hop reasoning. We conduct experiments on two benchmark datasets HotpotQA and 2WikiMultiHopQA. The results show that our method can effectively boost performance and also yields a better interpretable reasoning process without decomposition supervision. | ['Xuanjing Huang', 'Qi Zhang', 'Zhihao Fan', 'Zhongyu Wei', 'Siyuan Wang'] | 2022-08-22 | null | https://aclanthology.org/2022.coling-1.142 | https://aclanthology.org/2022.coling-1.142.pdf | coling-2022-10 | ['multi-hop-question-answering', 'question-generation'] | ['knowledge-base', 'natural-language-processing'] | [-1.13850914e-01 7.58786142e-01 -1.88765541e-01 -6.22335970e-01
-1.41787851e+00 -7.61289775e-01 3.71135354e-01 3.26687783e-01
7.08584264e-02 8.94492924e-01 6.79146647e-01 -8.27511728e-01
-4.43942487e-01 -7.90622652e-01 -6.66063011e-01 -1.29530773e-01
6.34331465e-01 7.18135893e-01 3.76446575e-01 -3.01483363e-01
1.81746662e-01 -1.56472266e-01 -1.05167961e+00 9.44166839e-01
1.44828141e+00 5.51317692e-01 3.77108455e-02 9.84137535e-01
-5.25505006e-01 1.57185340e+00 -7.30469584e-01 -9.94392514e-01
-3.67337428e-02 -6.70499325e-01 -1.59280336e+00 -7.23091699e-03
2.87996352e-01 -6.27159178e-01 -2.33359002e-02 7.98283696e-01
-1.82151534e-02 3.23447771e-03 4.10696864e-01 -1.32342052e+00
-1.07933509e+00 1.26475680e+00 -5.76634645e-01 1.79928228e-01
8.73810470e-01 1.50524691e-01 1.50619626e+00 -6.68167830e-01
5.08644462e-01 1.50822330e+00 5.57233155e-01 3.14323843e-01
-8.02864373e-01 -2.68199801e-01 5.94084144e-01 5.36389172e-01
-8.09411943e-01 -1.79154620e-01 7.86361516e-01 -1.63224731e-02
8.87960851e-01 6.63098395e-01 3.92172068e-01 6.41808927e-01
-2.47507077e-02 1.05582190e+00 9.53287303e-01 -2.75899738e-01
4.16891985e-02 6.16608337e-02 8.40510428e-01 1.10918236e+00
3.04622918e-01 -8.24413717e-01 -4.60398108e-01 -1.67927146e-01
2.48938128e-01 -7.17376769e-02 -1.99018136e-01 3.87671500e-01
-1.14773989e+00 8.75262082e-01 5.17304063e-01 5.33935986e-03
-5.09216726e-01 5.30681461e-02 8.31343681e-02 2.85856366e-01
8.14604685e-02 6.87790811e-01 -4.80879337e-01 -1.16722789e-02
-8.98801863e-01 5.13800204e-01 1.17992282e+00 8.39110374e-01
7.55555928e-01 -6.05716050e-01 -7.39331007e-01 5.01973271e-01
4.43024814e-01 2.89966196e-01 -1.23319067e-02 -1.59641600e+00
1.10098433e+00 1.05194890e+00 2.60794908e-01 -1.04731452e+00
-2.18359783e-01 -2.74928540e-01 -5.26760101e-01 -3.05018187e-01
4.40699041e-01 -2.82372653e-01 -6.28846467e-01 1.36384404e+00
5.97473323e-01 -2.99342245e-01 4.35523838e-01 9.82239425e-01
1.06268311e+00 8.47132623e-01 5.45826219e-02 -3.46865207e-02
1.77838016e+00 -1.78070259e+00 -8.20785284e-01 -2.46196315e-01
5.81307948e-01 -6.06169820e-01 1.34125650e+00 4.15766507e-01
-1.38869441e+00 -3.46873164e-01 -8.87969494e-01 -8.66376758e-01
-6.22228906e-02 2.62271285e-01 7.69259036e-01 2.41322126e-02
-8.73612702e-01 9.76789072e-02 -4.90640968e-01 2.16720365e-02
2.56119698e-01 -1.18207419e-03 8.71945545e-02 -3.53197962e-01
-1.57143533e+00 8.09880733e-01 2.07776040e-01 2.04451978e-01
-5.55867255e-01 -9.35644865e-01 -5.45739174e-01 2.85381824e-01
8.65350723e-01 -1.28143537e+00 1.62238860e+00 -1.52442172e-01
-1.25067461e+00 4.44520593e-01 -7.06053793e-01 -4.28598672e-01
5.92367351e-01 -4.72463727e-01 -1.33771911e-01 4.56721008e-01
5.00060618e-01 6.34291708e-01 4.10145432e-01 -1.39463329e+00
-7.13144898e-01 -2.17963174e-01 1.17812192e+00 4.18711454e-01
-2.92542018e-03 -2.46445358e-01 -7.24832952e-01 -2.95596927e-01
3.52536619e-01 -5.31517863e-01 -1.88188121e-01 -9.86453295e-02
-9.22406554e-01 -4.05473769e-01 5.58814883e-01 -1.07160795e+00
1.48073602e+00 -1.53300142e+00 3.89980167e-01 -2.27703273e-01
6.97647989e-01 -1.59003124e-01 -1.00033231e-01 4.67805505e-01
3.23314190e-01 3.65323633e-01 -2.10579634e-01 -3.02952349e-01
3.10533464e-01 4.18613970e-01 -6.66296244e-01 -4.99261826e-01
3.32238942e-01 1.17799461e+00 -9.79610741e-01 -8.68726194e-01
-2.63034642e-01 5.16819470e-02 -6.82539463e-01 2.55853415e-01
-7.00945675e-01 7.06025064e-02 -8.12810063e-01 6.01519108e-01
5.84096313e-01 -7.75277019e-01 2.36853082e-02 -3.77363354e-01
1.89458802e-01 5.88348210e-01 -1.01481068e+00 1.53831387e+00
-6.60034478e-01 3.67727518e-01 7.61900991e-02 -3.90168995e-01
6.44949675e-01 1.05213024e-01 -1.08141832e-01 -5.04488528e-01
-3.35956931e-01 -5.74339442e-02 -1.32422805e-01 -9.13593948e-01
6.28882349e-01 -1.14839464e-01 -6.25436679e-02 8.66439760e-01
-3.86589170e-01 -1.87160760e-01 6.65508926e-01 8.25625956e-01
1.15245247e+00 -2.59308461e-02 1.41977435e-02 8.45160708e-02
8.86532724e-01 4.13011104e-01 4.31619763e-01 8.27456057e-01
7.83386156e-02 5.31172335e-01 1.00985336e+00 -4.51955467e-01
-6.52081192e-01 -1.03780174e+00 5.07995307e-01 1.06640339e+00
4.68171567e-01 -7.10028648e-01 -8.59846950e-01 -1.12790573e+00
-1.05056912e-03 1.19322693e+00 -4.14479226e-01 1.16272710e-01
-5.75919926e-01 -5.50305545e-01 6.78899944e-01 6.25100434e-01
8.01415443e-01 -7.72660196e-01 -4.15336460e-01 1.21757567e-01
-1.05822194e+00 -1.22216010e+00 -2.75282621e-01 -1.66668802e-01
-7.02896535e-01 -1.20503330e+00 -3.38645518e-01 -4.75935787e-01
8.52286935e-01 3.66333306e-01 1.33827579e+00 6.81189120e-01
1.74801335e-01 3.01530182e-01 -4.81737018e-01 -1.39061347e-01
-4.60576147e-01 1.84278682e-01 -4.99133676e-01 -3.99538100e-01
1.87934473e-01 -1.33792371e-01 -6.29134059e-01 2.09101573e-01
-7.94056833e-01 6.02597654e-01 7.38921821e-01 6.06688976e-01
4.08322245e-01 2.89053351e-01 4.52868730e-01 -1.08154666e+00
1.19144607e+00 -5.77829480e-01 -1.92022026e-01 1.08305061e+00
-6.07973039e-01 5.55530012e-01 1.12575281e+00 1.28074497e-01
-1.59573317e+00 -6.37679160e-01 -9.63317528e-02 1.77128226e-01
1.52832270e-01 7.30506599e-01 -1.57970026e-01 5.22290945e-01
4.73732620e-01 -2.59772222e-02 -3.74701232e-01 -2.37936333e-01
6.05994999e-01 3.35894853e-01 5.72673798e-01 -9.92926359e-01
1.08266532e+00 3.44621450e-01 -2.20157385e-01 -8.97202790e-02
-1.58334339e+00 -1.00011937e-01 -3.38624179e-01 -7.05386475e-02
9.60910559e-01 -7.09415853e-01 -9.49195743e-01 -1.96906343e-01
-1.67976832e+00 -1.45894483e-01 -1.40937194e-01 -6.78734109e-02
-2.22294614e-01 4.85292971e-01 -7.49272287e-01 -7.67487526e-01
-5.42513132e-01 -1.15338337e+00 1.04311681e+00 4.71399963e-01
-5.45185268e-01 -1.06204653e+00 -7.81400353e-02 1.38002777e+00
3.67576480e-02 -5.15188426e-02 1.41679108e+00 -4.48902220e-01
-9.76671696e-01 4.81303735e-03 -5.22003055e-01 6.96787238e-02
-3.87424752e-02 1.20784789e-01 -6.96446240e-01 3.81001592e-01
5.50872367e-03 -5.01889050e-01 7.42391348e-01 -7.55496249e-02
1.15630198e+00 -7.15090096e-01 -8.10278803e-02 2.03069955e-01
1.19010603e+00 -2.43516281e-01 5.08574784e-01 3.03246856e-01
6.57309115e-01 7.13749051e-01 7.68606007e-01 8.22625384e-02
1.27466655e+00 1.34858802e-01 1.73435718e-01 -1.37700632e-01
-1.96362138e-01 -4.96854246e-01 1.90585986e-01 9.80419278e-01
-5.15721403e-02 -4.79183167e-01 -1.08104277e+00 4.76645529e-01
-2.02947092e+00 -9.59985733e-01 -4.86468971e-01 1.35287070e+00
8.56935084e-01 1.20650977e-01 -2.10986868e-01 2.12299302e-01
4.07632440e-01 1.83339059e-01 -5.01701534e-01 -4.78400320e-01
-3.70602496e-02 -1.92568853e-01 -2.72754103e-01 1.05292857e+00
-4.54771549e-01 7.82451272e-01 6.00601482e+00 4.03649479e-01
-3.24133396e-01 -2.90739685e-02 6.54704213e-01 -5.56645691e-02
-1.09787357e+00 3.30999434e-01 -8.49801719e-01 1.43129528e-01
4.80764180e-01 -3.60663444e-01 4.29912299e-01 4.20855999e-01
1.73538998e-01 -3.40748668e-01 -1.14033628e+00 5.83080590e-01
-4.98445444e-02 -1.45827127e+00 5.81841230e-01 -4.91540194e-01
7.20026195e-01 -7.45091975e-01 -7.59532452e-02 4.26255018e-01
6.36279523e-01 -8.92380476e-01 6.35173082e-01 6.09040439e-01
1.63433224e-01 -6.54387772e-01 7.20585763e-01 6.85229599e-01
-1.08961213e+00 -2.65743941e-01 -1.37238398e-01 -1.46905497e-01
3.96893471e-01 5.68400443e-01 -8.26160610e-01 1.02307105e+00
4.51862037e-01 2.97449857e-01 -7.08416939e-01 4.39425886e-01
-9.36291516e-01 4.72127914e-01 -5.54786325e-02 -3.97984833e-01
5.25765181e-01 -1.49377808e-01 3.66624057e-01 9.16350484e-01
1.40283674e-01 5.57008684e-01 6.09732941e-02 1.08467877e+00
-3.13618839e-01 -2.34197438e-01 8.94729048e-02 -1.22278057e-01
5.18571556e-01 1.33356690e+00 -5.59852004e-01 -6.15587056e-01
-3.94578576e-01 9.72753346e-01 7.02508271e-01 5.15582323e-01
-1.15731585e+00 -4.12417978e-01 3.47846538e-01 -4.96739782e-02
1.06777661e-02 -5.30955978e-02 -6.72232270e-01 -1.29270196e+00
4.82523829e-01 -1.00452852e+00 7.97304690e-01 -1.35217655e+00
-1.22609270e+00 5.57508707e-01 1.88216623e-02 -6.22811198e-01
-1.81251422e-01 -3.47404361e-01 -8.90807271e-01 9.83283460e-01
-1.85154378e+00 -1.23486722e+00 -5.28632820e-01 2.63492882e-01
8.15880597e-01 4.95050699e-01 6.58865869e-01 -7.67052919e-02
-6.37682199e-01 3.47850114e-01 -6.54127002e-01 3.92602384e-02
3.64243567e-01 -1.60079837e+00 2.04684868e-01 9.53047216e-01
2.73950007e-02 1.05517006e+00 8.35218370e-01 -6.12522244e-01
-1.43907499e+00 -8.91282976e-01 1.20863032e+00 -7.33148754e-01
7.75756001e-01 9.15051103e-02 -1.10674679e+00 6.99509382e-01
4.73995447e-01 -7.29817271e-01 6.98766470e-01 4.45010334e-01
-4.75739181e-01 -1.79323390e-01 -9.77323711e-01 9.81730878e-01
8.44005167e-01 -5.85190535e-01 -1.30474401e+00 5.44582844e-01
1.10836697e+00 -4.33154464e-01 -6.83898687e-01 1.93848595e-01
3.80559742e-01 -1.01679897e+00 8.64060819e-01 -7.97533333e-01
1.27092481e+00 -6.13332152e-01 9.47500840e-02 -9.28162277e-01
-3.91429812e-01 -5.94368577e-01 -5.70604682e-01 1.22597301e+00
1.08164656e+00 -3.56227189e-01 6.46491528e-01 1.07865429e+00
-8.63237828e-02 -1.11431479e+00 -3.93938124e-01 -2.21532166e-01
-1.23715796e-01 -3.70698154e-01 9.27365899e-01 6.54264212e-01
2.31995910e-01 6.71817720e-01 -1.53972030e-01 5.54252565e-01
6.63820207e-01 6.69843018e-01 8.05472136e-01 -8.57391059e-01
-6.16676867e-01 -2.32229978e-01 4.97267902e-01 -1.43188214e+00
2.24048555e-01 -8.13263535e-01 1.38703227e-01 -2.52630448e+00
4.16277140e-01 6.28097355e-03 -3.15189958e-02 5.47387183e-01
-9.16425884e-01 -2.96681523e-01 1.14430793e-01 9.76173058e-02
-1.16499126e+00 2.73426414e-01 1.78572607e+00 -2.88613856e-01
8.03080350e-02 -1.76898524e-01 -1.48432589e+00 8.01913679e-01
4.72488761e-01 -2.79035628e-01 -9.33201611e-01 -1.11971271e+00
6.63115323e-01 4.18769628e-01 5.27625918e-01 -6.97609723e-01
6.99750364e-01 -2.38715619e-01 2.15101540e-01 -7.65009046e-01
2.07640991e-01 -5.52687883e-01 -2.22748533e-01 2.96979576e-01
-8.06302130e-01 2.29797706e-01 -3.59967463e-02 4.57632512e-01
-3.57950956e-01 -3.67532492e-01 1.80208743e-01 -2.35415980e-01
-4.27633643e-01 1.63746048e-02 -4.85093631e-02 4.44849521e-01
9.16158915e-01 4.11642939e-02 -7.31240928e-01 -5.81396818e-01
-6.60697818e-01 9.49242353e-01 1.68362156e-01 1.91013649e-01
6.03115320e-01 -1.02788627e+00 -8.56863558e-01 -3.71481091e-01
-1.62006557e-01 3.69363397e-01 3.92132044e-01 7.96016097e-01
-6.88671410e-01 6.03680432e-01 2.23612338e-01 -1.80689231e-01
-1.17547297e+00 3.80378693e-01 1.60838068e-01 -8.27297926e-01
-4.03723747e-01 1.00468850e+00 -3.63786034e-02 -4.48591173e-01
1.95604637e-02 -6.20384574e-01 -2.34616160e-01 5.67213595e-02
6.37763500e-01 5.60750484e-01 -1.78025216e-01 1.51733877e-02
-2.20043004e-01 5.15720487e-01 -3.89103115e-01 -1.66482285e-01
1.04539299e+00 -4.20752794e-01 -5.47995687e-01 4.15246904e-01
7.50910878e-01 5.92822768e-02 -9.40839827e-01 -2.31687482e-02
4.18247655e-02 -2.65341967e-01 -2.94136584e-01 -1.18354535e+00
-5.74618101e-01 8.06208789e-01 -7.64769614e-01 5.59072733e-01
1.08905697e+00 1.92713037e-01 1.22030199e+00 8.23783398e-01
9.60053802e-02 -8.47140253e-01 2.04717517e-01 5.89739799e-01
1.14041722e+00 -1.18535054e+00 2.46738285e-01 -7.34744489e-01
-8.66587996e-01 1.18844664e+00 8.79193306e-01 4.16787416e-01
2.51050759e-02 2.54158109e-01 2.77560800e-01 -4.39486146e-01
-1.26938653e+00 2.32320264e-01 3.16084057e-01 1.65026113e-02
3.87544483e-01 -1.72687575e-01 -2.34226167e-01 1.02970338e+00
-3.74417394e-01 -5.13826543e-03 4.68569338e-01 7.90246665e-01
-4.48494971e-01 -8.77495885e-01 -4.25491333e-01 3.62087071e-01
-2.59225577e-01 -1.89409837e-01 -8.53312433e-01 6.34542227e-01
-8.95091370e-02 1.36851740e+00 -3.81597191e-01 -1.60320073e-01
3.17697257e-01 2.89174885e-01 2.99427062e-01 -4.78283942e-01
-7.19429970e-01 -6.28106475e-01 4.64530915e-01 -4.20911729e-01
-2.64488488e-01 -2.45379180e-01 -1.84257662e+00 -5.80120623e-01
-1.94927946e-01 6.13057792e-01 1.33487687e-01 1.36030936e+00
4.65379775e-01 8.83049786e-01 4.02874470e-01 1.43544436e-01
-7.18440771e-01 -5.79184651e-01 -5.57103269e-02 3.49229574e-01
4.07021016e-01 -1.66975573e-01 -4.18825269e-01 1.96633831e-01] | [10.929278373718262, 7.8828606605529785] |
57e78420-ba04-4a68-9b1d-879b4d82c0da | matrix-variate-rbm-and-its-applications | 1601.00722 | null | http://arxiv.org/abs/1601.00722v1 | http://arxiv.org/pdf/1601.00722v1.pdf | Matrix Variate RBM and Its Applications | Restricted Boltzmann Machine (RBM) is an importan- t generative model
modeling vectorial data. While applying an RBM in practice to images, the data
have to be vec- torized. This results in high-dimensional data and valu- able
spatial information has got lost in vectorization. In this paper, a
Matrix-Variate Restricted Boltzmann Machine (MVRBM) model is proposed by
generalizing the classic RBM to explicitly model matrix data. In the new RBM
model, both input and hidden variables are in matrix forms which are connected
by bilinear transforms. The MVRBM has much less model parameters, resulting in
a faster train- ing algorithm while retaining comparable performance as the
classic RBM. The advantages of the MVRBM have been demonstrated on two
real-world applications: Image super- resolution and handwritten digit
recognition. | ['Yongli Hu', 'Jinghua Li', 'Yanfeng Sun', 'Guanglei Qi', 'Junbin Gao'] | 2016-01-05 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [ 1.88046545e-01 -4.94062267e-02 -8.31194445e-02 -3.72768193e-01
-5.74102402e-01 -2.14610234e-01 9.33883071e-01 -5.02592027e-01
-6.70665383e-01 7.62130976e-01 4.98004742e-02 -3.30005020e-01
3.96124981e-02 -9.10804451e-01 -7.03527451e-01 -1.12534630e+00
1.52632460e-01 7.01760828e-01 2.86826175e-02 -1.41927689e-01
2.96171784e-01 7.84160435e-01 -1.13079429e+00 4.39543009e-01
6.61708772e-01 5.77548981e-01 5.36894321e-01 7.86831200e-01
-2.69399434e-01 9.98866200e-01 -3.02320421e-01 -2.10830033e-01
3.03295217e-02 -4.01473522e-01 -7.96994686e-01 -3.52081954e-02
4.18515593e-01 -1.84267640e-01 -6.58043027e-01 9.50402319e-01
1.75419420e-01 4.39251065e-01 1.14528584e+00 -9.17315364e-01
-8.55082631e-01 5.28822958e-01 -5.75999200e-01 1.59620166e-01
-2.39138082e-01 -4.09690410e-01 6.90799057e-01 -1.23470104e+00
4.77607757e-01 1.27604795e+00 4.43928748e-01 7.76875734e-01
-1.94995153e+00 -3.99096638e-01 -2.88341381e-02 4.83934730e-01
-1.56779647e+00 -2.25365981e-02 9.24865961e-01 -6.97087109e-01
1.19085252e+00 7.28555202e-01 6.92239106e-01 9.61339056e-01
4.36196446e-01 7.02777326e-01 1.38248444e+00 -5.33847630e-01
2.50480026e-01 3.16340804e-01 3.36154878e-01 4.77945328e-01
5.81214204e-02 -6.60468489e-02 -5.79504490e-01 -1.64372206e-01
1.27827907e+00 1.39016077e-01 -1.85562581e-01 -8.51033390e-01
-1.12283242e+00 1.17770827e+00 6.60841048e-01 4.62070525e-01
-4.68466640e-01 3.27541023e-01 -5.12327179e-02 -1.12169413e-02
4.20517445e-01 2.72405028e-01 2.05233604e-01 1.12062497e-02
-1.29879332e+00 1.75395787e-01 5.03904581e-01 6.30759716e-01
7.63352513e-01 4.10828441e-01 8.64001885e-02 1.00740659e+00
4.98135597e-01 5.69510818e-01 7.90273309e-01 -6.80632055e-01
5.03467381e-01 1.93734139e-01 -2.91828755e-02 -9.45651531e-01
-1.78742241e-02 -1.14443488e-01 -1.58485937e+00 4.29751098e-01
8.92001912e-02 4.81539249e-01 -1.33202791e+00 1.37564814e+00
-5.64592518e-02 1.79237321e-01 -6.25276417e-02 8.89787793e-01
7.72089422e-01 1.31231391e+00 -1.89770982e-01 7.10373223e-02
9.27466571e-01 -5.63023329e-01 -7.87051857e-01 -1.40131399e-01
2.14836866e-01 -3.05400521e-01 7.95667589e-01 6.08126342e-01
-9.08508837e-01 -6.59318328e-01 -1.14776862e+00 -1.97856352e-01
-4.45611954e-01 -3.38928699e-01 6.14472866e-01 6.09418988e-01
-1.16464567e+00 5.44355810e-01 -1.15971959e+00 -4.37458605e-02
3.66143525e-01 5.45587718e-01 -7.42197454e-01 -2.11000741e-01
-8.48590672e-01 1.12566543e+00 2.05736727e-01 3.95689160e-01
-7.97558665e-01 -2.80740827e-01 -8.88707697e-01 -1.82202905e-01
-3.07422757e-01 -4.69093442e-01 7.57431090e-01 -4.47996557e-01
-1.76987839e+00 7.11159408e-01 -2.64368355e-01 -2.86093563e-01
2.76618570e-01 3.01254299e-02 -3.89546528e-02 -3.21301401e-01
-4.38947082e-01 5.09464800e-01 1.02926242e+00 -1.14059412e+00
1.04132898e-01 -4.71031815e-01 -3.13292384e-01 1.26086235e-01
-4.07632768e-01 -3.34285855e-01 -2.34262630e-01 -7.48004198e-01
5.26164770e-01 -8.09185028e-01 -5.53096890e-01 -5.68678737e-01
-2.72967458e-01 1.53271839e-01 7.90522158e-01 -8.79513800e-01
9.42234278e-01 -1.98912549e+00 1.10358882e+00 4.64112699e-01
1.54746398e-01 1.40619367e-01 -1.96782276e-01 3.89611483e-01
-3.71612281e-01 -1.86828464e-01 -4.21920478e-01 -4.02268171e-01
-2.10932344e-01 7.97740340e-01 -6.37241364e-01 6.73053980e-01
-1.77762672e-01 1.07981265e+00 -3.54559153e-01 -4.67585206e-01
5.02318561e-01 8.23103368e-01 -5.50977588e-01 2.28336558e-01
4.36867148e-01 4.99194682e-01 -1.53291021e-02 1.16290331e-01
7.54483104e-01 -1.84088141e-01 -1.17029287e-01 -2.17170894e-01
-1.12734914e-01 -3.50794755e-02 -1.07460654e+00 1.56171238e+00
-6.55708551e-01 6.80248916e-01 -2.48715430e-02 -1.23453355e+00
9.19906676e-01 7.84096941e-02 3.01183671e-01 -6.59444094e-01
-1.04630090e-01 -9.51575339e-02 -1.32609025e-01 -9.95979756e-02
6.69376433e-01 -3.10462087e-01 4.51599136e-02 1.75182700e-01
3.23239446e-01 -3.10372621e-01 -2.68458694e-01 1.82845488e-01
5.86837530e-01 5.93854114e-02 2.64753759e-01 -3.84894878e-01
3.68772328e-01 -1.39055446e-01 3.36546481e-01 8.26551139e-01
3.96845281e-01 6.26668513e-01 9.90234315e-02 -5.01867294e-01
-1.18637407e+00 -1.26119542e+00 -2.97639966e-01 8.29188526e-01
-4.63961065e-02 -9.53900516e-02 -6.32531047e-01 -1.10800110e-01
-2.28150263e-02 8.99615526e-01 -8.68604422e-01 -1.08122393e-01
-7.05129087e-01 -1.04373503e+00 3.35750908e-01 6.26251280e-01
3.74903649e-01 -9.63695884e-01 -4.71394360e-01 1.42710641e-01
-1.12384766e-01 -6.16186678e-01 -3.19755763e-01 5.04458725e-01
-1.32911491e+00 -3.47382396e-01 -1.07418394e+00 -6.96123123e-01
6.93875253e-01 -5.69395497e-02 7.52412379e-01 -7.55205631e-01
-5.67539096e-01 5.97819500e-02 -5.65508679e-02 -6.27310574e-02
-4.23237234e-01 -1.48482993e-01 5.77844270e-02 2.36124024e-01
1.37989610e-01 -5.31862974e-01 -2.92418182e-01 1.48872271e-01
-9.53752756e-01 5.33555984e-01 8.11430037e-01 1.30486822e+00
7.73808479e-01 -2.44989827e-01 5.18095829e-02 -9.11969602e-01
3.14904839e-01 -1.31472886e-01 -7.19188213e-01 5.15771806e-02
-5.35499334e-01 3.27522397e-01 2.23691687e-01 -8.99678290e-01
-1.09263802e+00 7.83901662e-02 -2.91006058e-01 -5.06060064e-01
6.99086636e-02 5.04374683e-01 -2.35659122e-01 -2.45666519e-01
4.39534307e-01 7.43058562e-01 1.38146132e-01 -8.06518972e-01
5.14628410e-01 5.68575919e-01 4.66690689e-01 -2.55017132e-01
7.27199435e-01 4.33521330e-01 9.66474861e-02 -1.11859465e+00
-5.28055541e-02 -2.53401488e-01 -8.31491828e-01 -5.24950847e-02
9.14575279e-01 -4.79469955e-01 -3.72636259e-01 7.97524154e-01
-1.02520502e+00 -4.86721337e-01 -5.36637127e-01 6.90353155e-01
-6.65293992e-01 5.02436876e-01 -9.07350898e-01 -7.98328221e-01
-4.10608537e-02 -1.17747414e+00 5.77304184e-01 -1.58742115e-01
-7.27095380e-02 -9.95899200e-01 3.05475295e-01 3.09959412e-01
6.86769724e-01 -1.60932943e-01 1.20581627e+00 -3.12512428e-01
-5.23247421e-01 -4.13486689e-01 2.32461970e-02 5.61054409e-01
5.97147346e-02 -3.90446872e-01 -9.89903867e-01 -4.22382981e-01
1.83447286e-01 5.63047118e-02 1.23887002e+00 5.25092602e-01
1.26443028e+00 -4.00645673e-01 -7.38166496e-02 9.24277425e-01
1.44157493e+00 2.06776157e-01 1.00818360e+00 2.77565122e-01
9.61194158e-01 3.36402714e-01 1.75201654e-01 2.21469834e-01
8.30258578e-02 8.15822423e-01 4.12224263e-01 -2.89973289e-01
8.85513723e-02 -1.48305520e-01 3.47070545e-01 1.10794473e+00
-6.99572444e-01 8.20991024e-02 -8.20688426e-01 3.38870853e-01
-1.89988101e+00 -1.14019072e+00 -2.65316218e-01 2.20986533e+00
8.48926961e-01 -3.70230079e-01 -2.31149048e-01 6.27248213e-02
4.87153113e-01 3.24218571e-01 -4.67057675e-01 -5.06918073e-01
-3.21738839e-01 5.33367336e-01 7.14493096e-01 7.37544000e-01
-9.59041893e-01 8.58839095e-01 7.38566113e+00 9.50981796e-01
-1.35990524e+00 4.01536107e-01 4.47714180e-01 -1.89785749e-01
-2.62349486e-01 -2.07601428e-01 -5.53035080e-01 1.80449322e-01
7.45773077e-01 2.42807239e-01 7.27285862e-01 8.06095660e-01
-2.60736287e-01 1.42447231e-02 -1.03041697e+00 1.63028145e+00
2.39789486e-01 -1.46841824e+00 3.23920488e-01 5.14053702e-01
4.54597503e-01 3.27306241e-02 4.70497102e-01 4.28242028e-01
2.53876507e-01 -1.55343521e+00 7.26995647e-01 8.22984278e-01
1.01233029e+00 -8.39675307e-01 7.25176096e-01 4.19395447e-01
-6.84631944e-01 1.64160758e-01 -8.72913241e-01 3.16988789e-02
3.04740481e-02 4.66021508e-01 -6.36381269e-01 1.39768541e-01
6.37569606e-01 3.34190577e-01 -3.37980360e-01 6.61893785e-01
3.59573483e-01 6.63996279e-01 -7.89064318e-02 1.70763806e-02
1.65958211e-01 -6.95676327e-01 4.95261222e-01 1.24789107e+00
3.32202584e-01 -3.43364999e-02 -2.79811263e-01 1.01723599e+00
3.62950027e-01 1.97673604e-01 -6.43208563e-01 1.76353119e-02
-1.37106031e-01 1.19283545e+00 -5.41833401e-01 -2.51647562e-01
-3.04912597e-01 1.56532037e+00 4.41962451e-01 6.91319704e-01
-8.32230389e-01 -9.33197141e-02 5.29036224e-01 1.34778872e-01
4.99628574e-01 -5.41356683e-01 -2.86943823e-01 -1.35167730e+00
-3.53765488e-01 -5.72749019e-01 6.21247478e-02 -6.76892161e-01
-1.27946043e+00 7.97171891e-01 2.98957437e-01 -7.60866880e-01
-4.37157333e-01 -1.01938522e+00 -1.06687941e-01 1.43091750e+00
-1.04534304e+00 -1.27981174e+00 -7.83516988e-02 8.67728889e-01
2.77916759e-01 -3.31972957e-01 1.14092028e+00 2.92450357e-02
-4.45628494e-01 3.96436006e-01 5.62145948e-01 2.36750141e-01
4.30730611e-01 -1.40883565e+00 3.77728492e-01 5.84646881e-01
6.36739373e-01 8.38117659e-01 7.84851968e-01 -4.61159319e-01
-1.51951504e+00 -1.16670573e+00 3.75223190e-01 -5.20712495e-01
2.72094607e-01 -6.50672674e-01 -1.14725614e+00 8.88504267e-01
8.45410377e-02 1.14891484e-01 8.58107984e-01 1.98614329e-01
-5.97979188e-01 -4.43264470e-02 -1.06534374e+00 3.25083435e-01
3.71235758e-01 -7.73014486e-01 -6.60190821e-01 -3.62711996e-02
4.31262180e-02 -3.97358000e-01 -9.18903708e-01 7.88362771e-02
6.81814849e-01 -8.14137995e-01 1.25592554e+00 -7.20519364e-01
-2.29807626e-02 -1.51604235e-01 -6.16893888e-01 -1.40160692e+00
-9.41948533e-01 -5.75184897e-02 -4.39977229e-01 8.05748940e-01
2.81380355e-01 -7.15075195e-01 7.73538589e-01 7.01842189e-01
3.59839678e-01 -8.11200261e-01 -1.34109640e+00 -9.89018142e-01
4.05157447e-01 -4.63324428e-01 4.15128887e-01 9.90992188e-01
-7.88844749e-02 1.84294507e-01 -8.02828133e-01 1.01244621e-01
7.78139174e-01 1.13590352e-01 5.03791213e-01 -1.20032060e+00
-5.46079099e-01 -2.44715869e-01 -8.36364388e-01 -1.24927461e+00
-7.75552494e-03 -9.09893930e-01 1.24414898e-01 -1.54780865e+00
6.62888467e-01 -3.42028052e-01 -5.65352976e-01 4.53764021e-01
1.08785957e-01 6.18457437e-01 1.49402201e-01 3.66833359e-01
-8.01141113e-02 6.38125658e-01 9.46171165e-01 -4.40428138e-01
-1.87625289e-01 -1.35158166e-01 -8.37671757e-02 5.54371953e-01
6.18653595e-01 -2.85633683e-01 -2.32051864e-01 -2.64055699e-01
-8.57929215e-02 1.44628271e-01 6.47899210e-01 -7.32416689e-01
1.61141574e-01 -3.91539395e-01 7.78296351e-01 -8.66284132e-01
1.03412569e+00 -6.43590331e-01 5.34813583e-01 4.36135054e-01
-1.90942928e-01 -3.20399441e-02 1.00480273e-01 3.17477882e-01
-3.22583735e-01 -2.46306717e-01 1.02346027e+00 -5.32167144e-02
-8.09926212e-01 3.36702824e-01 -6.93380713e-01 -6.50720835e-01
9.26633120e-01 -4.94864792e-01 2.73587078e-01 -3.76349211e-01
-9.02812302e-01 -6.18993044e-01 4.03365523e-01 2.73324996e-01
1.05521333e+00 -1.51615345e+00 -6.87192738e-01 5.73516488e-01
-6.89370790e-03 7.11330306e-03 5.31894982e-01 7.51412272e-01
-5.75723350e-01 7.22399831e-01 -4.77637768e-01 -7.64857650e-01
-1.42403853e+00 7.23733962e-01 4.15966094e-01 -2.23044857e-01
-7.72153020e-01 1.04153049e+00 4.92517650e-01 -4.74366546e-01
-4.51851562e-02 -3.85994494e-01 -2.34692127e-01 -1.25197381e-01
4.57328051e-01 3.04975271e-01 -1.50351990e-02 -9.89336252e-01
-1.82314128e-01 3.92846912e-01 -8.10249969e-02 -5.74186265e-01
1.55268252e+00 1.10161975e-01 -4.77652282e-01 8.67104411e-01
1.14724946e+00 -2.28317287e-02 -9.23836708e-01 -3.25424910e-01
-4.21989232e-01 -5.75876653e-01 4.65969115e-01 -3.61480236e-01
-9.14047241e-01 1.16143322e+00 6.61881566e-01 -1.57127678e-01
8.44432592e-01 -6.82530031e-02 1.56964019e-01 4.22405869e-01
5.94606400e-01 -8.61111343e-01 -1.54439449e-01 4.59797889e-01
1.21013927e+00 -9.57642555e-01 -3.27858282e-03 -5.87560572e-02
-6.62257314e-01 1.02365398e+00 3.98890853e-01 -3.17405581e-01
8.09464574e-01 3.58223498e-01 -1.40635580e-01 8.67076293e-02
-2.90353864e-01 4.58468586e-01 5.81381798e-01 7.17265069e-01
3.45872283e-01 1.86311573e-01 1.29921228e-01 4.93081689e-01
-1.64358705e-01 -3.20229143e-01 3.03928882e-01 6.93200231e-01
-1.78325668e-01 -1.12883604e+00 -7.93196023e-01 4.80238974e-01
-2.15192810e-02 -1.88964233e-01 -7.51919746e-02 6.85691953e-01
-2.56314218e-01 3.87440711e-01 2.99361885e-01 -4.42063600e-01
-3.96130607e-02 1.69281363e-01 1.04193223e+00 -4.80328679e-01
5.40853962e-02 -4.93061021e-02 -5.69962084e-01 -6.66833222e-01
-1.40770614e-01 -5.22405982e-01 -1.03332806e+00 -2.99377888e-01
-3.97841603e-01 2.13536322e-01 1.14955485e+00 6.00596189e-01
7.83188790e-02 4.89200383e-01 1.83338955e-01 -1.33152652e+00
-5.90400755e-01 -1.16346371e+00 -1.02285576e+00 1.90417245e-01
3.56503397e-01 -6.80134535e-01 1.61392018e-02 1.00960933e-01] | [9.184768676757812, 2.469943046569824] |
7ee772ba-6c79-48b3-8fbf-804d244b9235 | improved-semantic-representation-for-domain | null | null | https://aclanthology.org/W16-2902 | https://aclanthology.org/W16-2902.pdf | Improved Semantic Representation for Domain-Specific Entities | null | ['Nigel Collier', 'Mohammad Taher Pilehvar'] | 2016-08-01 | null | null | null | ws-2016-8 | ['learning-semantic-representations'] | ['methodology'] | [-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] | [-7.2740349769592285, 3.808521270751953] |
30205812-fd16-4b8f-a7bd-2aead1d06766 | bioflair-pretrained-pooled-contextualized | 1908.05760 | null | https://arxiv.org/abs/1908.05760v1 | https://arxiv.org/pdf/1908.05760v1.pdf | BioFLAIR: Pretrained Pooled Contextualized Embeddings for Biomedical Sequence Labeling Tasks | Biomedical Named Entity Recognition (NER) is a challenging problem in biomedical information processing due to the widespread ambiguity of out of context terms and extensive lexical variations. Performance on bioNER benchmarks continues to improve due to advances like BERT, GPT, and XLNet. FLAIR (1) is an alternative embedding model which is less computationally intensive than the others mentioned. We test FLAIR and its pretrained PubMed embeddings (which we term BioFLAIR) on a variety of bio NER tasks and compare those with results from BERT-type networks. We also investigate the effects of a small amount of additional pretraining on PubMed content, and of combining FLAIR and ELMO models. We find that with the provided embeddings, FLAIR performs on-par with the BERT networks - even establishing a new state of the art on one benchmark. Additional pretraining did not provide a clear benefit, although this might change with even more pretraining being done. Stacking the FLAIR embeddings with others typically does provide a boost in the benchmark results. | ['Ron Daniel Jr', 'Shreyas Sharma'] | 2019-08-13 | null | null | null | null | ['medical-named-entity-recognition'] | ['natural-language-processing'] | [-1.72159802e-02 1.21657558e-01 -9.08002034e-02 -2.49361038e-01
-7.18578756e-01 -4.41287369e-01 6.39906943e-01 7.94278800e-01
-1.13642669e+00 9.98401701e-01 7.52050877e-01 -2.39019558e-01
-1.83029503e-01 -5.24647117e-01 -4.47359025e-01 -5.93126476e-01
-2.85621077e-01 5.25586188e-01 6.32503629e-02 -1.84691250e-01
-1.64771929e-01 5.56309283e-01 -9.03289497e-01 1.04696847e-01
4.75016594e-01 4.90962178e-01 -1.62451476e-01 6.12983644e-01
-3.17341655e-01 4.85053003e-01 -5.20003796e-01 -6.85545862e-01
-1.13840632e-01 -7.12445304e-02 -9.87980664e-01 -7.01615810e-01
3.80954385e-01 1.62908792e-01 -2.89220840e-01 6.70007527e-01
1.00651717e+00 1.43060133e-01 6.30058885e-01 -8.97519588e-01
-7.09927738e-01 7.32065737e-01 -2.86985785e-01 5.00825763e-01
2.84349263e-01 3.20268609e-03 1.09069371e+00 -9.36288059e-01
1.04329431e+00 1.00904071e+00 1.45520878e+00 5.98100603e-01
-1.22975731e+00 -4.43370938e-01 -2.98407942e-01 3.95858362e-02
-1.52713692e+00 -4.81555164e-01 2.03533284e-02 -1.52604133e-01
1.48265064e+00 2.81841069e-01 2.71208465e-01 1.23777652e+00
5.42391241e-01 2.80893713e-01 8.11349392e-01 -2.18814075e-01
1.25757545e-01 8.22218731e-02 5.22688508e-01 7.01784432e-01
6.51858807e-01 -3.36023569e-02 -2.89545596e-01 -5.78376234e-01
2.00528875e-01 9.22207441e-03 -3.88889164e-01 1.03357650e-01
-1.43672597e+00 6.38149798e-01 5.23785651e-01 8.39738667e-01
-4.80816752e-01 1.80676416e-01 6.99535608e-01 2.34198138e-01
5.45919120e-01 1.23745072e+00 -8.44200134e-01 -2.52161533e-01
-1.16848779e+00 -1.66614540e-02 1.03665805e+00 6.21876121e-01
3.10885280e-01 -1.90416709e-01 -3.61312360e-01 8.85861754e-01
-2.03307599e-01 -2.63383627e-01 6.26389563e-01 -4.36453074e-01
1.70075536e-01 4.23613012e-01 -1.73864901e-01 -7.68681228e-01
-9.39245701e-01 -5.26499510e-01 -7.71894276e-01 -2.76120931e-01
6.07659936e-01 -5.85774839e-01 -1.28389359e+00 1.51969445e+00
9.04478505e-02 3.69851500e-01 1.00448571e-01 6.19349003e-01
1.14669633e+00 5.09040713e-01 6.23214006e-01 1.57691777e-01
1.63186729e+00 -8.62693787e-01 -9.48751450e-01 -2.35781223e-01
1.01585805e+00 -6.69541717e-01 5.13540745e-01 1.83270261e-01
-7.06172109e-01 -1.38265833e-01 -1.13248003e+00 -6.09198868e-01
-1.09819043e+00 -1.58555046e-01 9.65474129e-01 7.25701690e-01
-1.15627623e+00 7.85546839e-01 -9.08014715e-01 -6.69196486e-01
5.34202814e-01 4.98097360e-01 -9.76746261e-01 -4.28254128e-01
-1.44950080e+00 1.33771288e+00 5.82948267e-01 -6.32619038e-02
-4.53836501e-01 -1.40190327e+00 -9.82499778e-01 1.92638725e-01
1.60512939e-01 -7.97606766e-01 7.81170011e-01 -1.63601145e-01
-8.89596045e-01 8.65144134e-01 7.09008127e-02 -4.93525445e-01
3.70344877e-01 -1.66010797e-01 -5.81571758e-01 -1.17911495e-01
3.59102823e-02 1.04134047e+00 -8.24769512e-02 -6.59472585e-01
-6.44829497e-02 -2.96904951e-01 -1.79073051e-01 -1.10437728e-01
-5.14233351e-01 1.66552454e-01 -3.09078932e-01 -5.65932870e-01
-4.67864931e-01 -8.09688091e-01 -5.33645093e-01 -3.73263694e-02
-5.25313556e-01 -4.63760883e-01 2.49861479e-01 -6.01545691e-01
1.10761106e+00 -2.01073098e+00 -4.16248702e-02 -1.87970176e-01
3.92423093e-01 2.25541994e-01 -4.35891300e-01 5.43422759e-01
-6.49478436e-01 7.09286869e-01 -2.18991876e-01 -2.43715376e-01
-8.01659301e-02 2.05117360e-01 3.41315210e-01 4.82541680e-01
6.46198630e-01 1.03464782e+00 -9.62384582e-01 -5.11829257e-01
-2.04476178e-01 8.83675933e-01 -6.03785157e-01 -2.39807263e-01
6.63859248e-02 -1.30426571e-01 -2.55736977e-01 5.78005016e-01
4.57042664e-01 -3.58319253e-01 2.15818837e-01 -3.96110237e-01
-2.81618293e-02 2.62486577e-01 -8.13471913e-01 1.84026051e+00
-3.09508681e-01 6.15933776e-01 -1.28400847e-01 -8.24684680e-01
4.22676086e-01 5.62775254e-01 6.03448272e-01 -3.26153368e-01
2.21715763e-01 1.09881133e-01 4.05575216e-01 -4.04968828e-01
5.17672837e-01 -3.24314386e-01 -1.09872289e-01 7.53049925e-02
5.61020792e-01 3.21939886e-01 4.26124372e-02 1.45540923e-01
1.81800723e+00 -2.23042443e-01 5.19022167e-01 -5.66488147e-01
8.35218281e-02 2.02431545e-01 6.22771919e-01 6.56045437e-01
-1.66750386e-01 5.40803373e-01 4.78401989e-01 -4.42736417e-01
-8.12704742e-01 -7.52265871e-01 -6.59260213e-01 1.05212986e+00
-3.51858974e-01 -7.19919503e-01 -4.83763546e-01 -9.75226820e-01
1.37973204e-01 7.27938235e-01 -9.25653338e-01 1.67025011e-02
-2.92049408e-01 -1.41694343e+00 1.13504279e+00 5.37150621e-01
1.19411156e-01 -9.71748412e-01 -4.07371461e-01 3.98158014e-01
3.10909152e-01 -1.07867289e+00 -4.43682998e-01 6.79181814e-01
-7.59512961e-01 -9.61620152e-01 -1.00333881e+00 -6.56350911e-01
4.49739486e-01 -3.74127358e-01 1.25415099e+00 -4.84612286e-02
-6.46471977e-01 1.21504433e-01 -2.28191450e-01 -7.62681842e-01
-1.00315526e-01 4.85325038e-01 -1.72472715e-01 -6.93077624e-01
6.10691905e-01 -3.15837532e-01 -5.22425354e-01 -1.01724923e-01
-1.09594798e+00 -4.87897366e-01 7.84832120e-01 1.18049681e+00
3.49357426e-01 2.54887324e-02 7.08300173e-01 -1.39379966e+00
8.00250828e-01 -7.92791963e-01 7.78294131e-02 9.10301730e-02
-6.83801830e-01 3.58722955e-01 4.57298398e-01 -2.57969230e-01
-6.56668127e-01 -2.49698043e-01 -6.19916499e-01 -1.82629377e-01
-2.15047225e-01 8.94900858e-01 8.00359473e-02 1.01237364e-01
8.20067763e-01 -4.53680247e-01 -2.42035240e-01 -7.22466171e-01
4.20782208e-01 5.39971113e-01 1.99792683e-01 -2.14296654e-01
2.19086379e-01 1.42135933e-01 -1.05420180e-01 -7.62752593e-01
-7.48473763e-01 -5.17575622e-01 -2.44967431e-01 6.01180971e-01
1.30608761e+00 -7.96630442e-01 -4.17866766e-01 -6.19581006e-02
-1.10646892e+00 -3.25964876e-02 -2.62671500e-01 5.98100841e-01
7.86063746e-02 1.04207456e-01 -9.72900093e-01 -1.29549339e-01
-4.22171444e-01 -1.13790560e+00 7.87560701e-01 9.62928832e-02
-5.33357203e-01 -1.32479930e+00 3.15095812e-01 -5.75717771e-04
5.57170630e-01 4.36431110e-01 1.15809202e+00 -1.42611039e+00
2.43927240e-01 -2.71551877e-01 -3.71188700e-01 -5.14523946e-02
2.27711082e-01 -3.01426440e-01 -1.01460278e+00 -9.35375616e-02
-3.92261237e-01 -2.06871778e-01 1.25266194e+00 1.29518971e-01
1.11956608e+00 -1.19846193e-02 -6.82528436e-01 4.54083651e-01
1.55686164e+00 -1.79671205e-03 6.43636763e-01 4.44653422e-01
5.82134366e-01 4.08389568e-01 -3.15589644e-03 3.22774686e-02
2.76404619e-01 3.45709562e-01 1.17191784e-01 -4.33486491e-01
-1.63556740e-01 7.59564936e-02 1.68486074e-01 7.45369852e-01
1.41212896e-01 -4.20619488e-01 -1.34030342e+00 7.94042647e-01
-1.30211043e+00 -7.74342299e-01 4.06080075e-02 1.77483928e+00
1.18290734e+00 7.63388127e-02 -1.46791250e-01 -1.61747217e-01
4.49437052e-01 5.75843379e-02 -4.98433888e-01 -7.88360178e-01
-6.98913336e-02 9.32427287e-01 9.23749030e-01 1.93442538e-01
-1.08838391e+00 6.89909458e-01 6.93686485e+00 7.66853094e-01
-1.07523084e+00 4.08317775e-01 7.84250796e-01 -2.92614829e-02
-1.76874325e-01 -3.36532444e-01 -8.62828791e-01 3.58435512e-01
1.62649071e+00 -2.31276944e-01 -1.00600883e-01 5.02135754e-01
-2.45445773e-01 1.91489697e-01 -1.37094903e+00 8.53199840e-01
1.40884250e-01 -1.71368694e+00 -7.29800910e-02 2.10250929e-01
5.19426763e-01 6.13159716e-01 -3.65328819e-01 6.69292986e-01
4.87449139e-01 -1.46993005e+00 -7.57477582e-02 5.34710467e-01
8.28997433e-01 -6.98247612e-01 1.36132550e+00 -5.99810965e-02
-6.62183881e-01 1.69968188e-01 -4.11817670e-01 4.19399410e-01
3.59982029e-02 8.22011411e-01 -9.83376205e-01 8.39969099e-01
5.91178477e-01 6.42714143e-01 -8.41645420e-01 1.39997697e+00
1.17414437e-01 6.67659223e-01 -3.56228948e-01 -9.84882414e-02
3.20807368e-01 4.22833085e-01 4.54907745e-01 1.91738343e+00
1.93920687e-01 -4.17984761e-02 3.66489589e-02 5.25588989e-01
-6.63082659e-01 2.26959631e-01 -5.58998466e-01 -5.90743184e-01
4.00481969e-01 1.47452092e+00 -6.56127512e-01 -4.69514012e-01
-4.27915454e-01 9.36885953e-01 3.21783632e-01 -1.20902406e-02
-7.46821344e-01 -8.65164876e-01 8.01451445e-01 2.62824595e-02
1.84375077e-01 6.33197650e-02 -1.62159815e-01 -1.01520193e+00
-5.30250311e-01 -7.54692614e-01 6.58674121e-01 -6.17461026e-01
-1.62045646e+00 6.75276399e-01 -2.05339462e-01 -6.30632102e-01
1.41207486e-01 -8.76523495e-01 -4.11982745e-01 8.33924234e-01
-1.62925673e+00 -6.48368299e-01 1.31586373e-01 1.56997845e-01
1.91554204e-01 1.98101223e-01 1.30985773e+00 6.67276800e-01
-7.86233723e-01 9.16721463e-01 1.24331564e-02 3.66397828e-01
1.19468892e+00 -1.44205201e+00 2.32305020e-01 4.08762872e-01
1.72529981e-01 1.08051050e+00 6.19458079e-01 -4.74383086e-01
-1.28071153e+00 -1.27197146e+00 1.30696714e+00 -7.68535852e-01
7.95885980e-01 -2.95513213e-01 -8.88213694e-01 7.30531991e-01
7.41399646e-01 1.38867483e-01 1.25932205e+00 6.07883334e-01
-3.67244244e-01 2.89363146e-01 -1.19592047e+00 4.77598369e-01
1.14328599e+00 -2.75018185e-01 -8.66793096e-01 4.34401721e-01
8.55640590e-01 -2.77055860e-01 -1.51117027e+00 3.67689043e-01
6.05768740e-01 -3.27510387e-01 9.62155819e-01 -1.07832575e+00
5.47849238e-01 6.00136518e-02 4.89933342e-02 -1.60354173e+00
-5.96148968e-01 -2.48582661e-01 3.31961244e-01 1.26375258e+00
9.07409668e-01 -7.18086660e-01 6.94273114e-01 6.78026736e-01
-2.63893694e-01 -8.94197166e-01 -9.14255023e-01 -6.86415315e-01
5.09493828e-01 -2.54862338e-01 5.60148716e-01 1.50015235e+00
2.01011896e-01 5.67963541e-01 3.79170515e-02 -4.15237099e-02
1.20560065e-01 -6.04512751e-01 2.10231677e-01 -1.26191545e+00
-3.07609998e-02 -3.82823378e-01 -7.51369834e-01 -3.06886494e-01
1.43551499e-01 -1.43317068e+00 6.84646750e-03 -1.82355583e+00
2.72870332e-01 -4.05554861e-01 -9.62869048e-01 9.41673577e-01
-5.40323436e-01 3.31242323e-01 1.03186846e-01 -2.08643779e-01
-4.92082626e-01 1.69931054e-01 6.93600118e-01 -2.91069925e-01
-5.97077198e-02 -8.69711101e-01 -1.03677189e+00 3.98052067e-01
6.35661960e-01 -7.78685808e-01 -7.77978227e-02 -6.13228977e-01
2.70147413e-01 -1.40796036e-01 5.13841026e-02 -1.08748007e+00
2.53814369e-01 4.97245342e-01 8.00054073e-01 -8.27499107e-02
1.25027791e-01 -5.10567725e-01 3.27655613e-01 3.16902101e-01
-5.15042126e-01 4.41186786e-01 6.85690224e-01 5.88305295e-01
-6.86955675e-02 -2.14917123e-01 5.05970180e-01 -5.77673018e-02
-5.14928997e-01 2.13957667e-01 -4.32492882e-01 2.03955144e-01
6.40955091e-01 -5.31084649e-02 -3.64408404e-01 2.58264899e-01
-9.68776762e-01 3.42905939e-01 2.72850633e-01 4.77462918e-01
1.44331262e-01 -1.08969164e+00 -6.61973357e-01 -1.98311329e-01
2.73530245e-01 -3.95834863e-01 8.11624229e-02 9.82869565e-01
-4.63689119e-01 6.16994739e-01 -7.51119256e-02 -2.20707983e-01
-1.05233169e+00 4.55748320e-01 3.08330059e-01 -8.06134760e-01
-4.36636716e-01 1.03157818e+00 -4.60805260e-02 -7.31881618e-01
2.07507778e-02 -4.29751843e-01 -3.88479173e-01 4.13922638e-01
4.52591389e-01 3.14038336e-01 5.02126098e-01 -1.70754477e-01
-8.53415549e-01 9.56260227e-03 -3.93613786e-01 1.02467857e-01
1.67472589e+00 5.61992168e-01 -4.78638299e-02 3.90794367e-01
1.50233996e+00 2.00054616e-01 -3.17367554e-01 1.21161841e-01
4.22212929e-01 1.69748142e-01 3.51179004e-01 -1.18133759e+00
-1.02237713e+00 7.89084136e-01 6.29372656e-01 -2.58574337e-01
6.94779038e-01 -2.10789695e-01 8.06629717e-01 4.44553256e-01
1.26058593e-01 -7.14455605e-01 -4.79070485e-01 3.63750398e-01
5.54906845e-01 -1.01633227e+00 2.20115066e-01 -1.59636550e-02
-3.90023470e-01 9.92632389e-01 3.05953562e-01 -6.93917796e-02
8.92798245e-01 5.51799536e-01 -1.54179841e-01 -5.77825963e-01
-8.46216500e-01 -1.55671611e-01 8.87280256e-02 3.50026250e-01
1.02802598e+00 -1.63362488e-01 -7.12835789e-01 6.90340579e-01
-5.45209758e-02 2.85613954e-01 3.92466456e-01 9.26438987e-01
4.61562611e-02 -1.30431354e+00 2.82111485e-02 9.37957168e-01
-1.17296612e+00 -6.24815643e-01 -4.17650700e-01 1.06646359e+00
3.71958405e-01 7.36469030e-01 -2.88949423e-02 -2.27009743e-01
5.74827850e-01 5.12408257e-01 2.30326235e-01 -9.18872118e-01
-1.03107071e+00 -2.04425782e-01 5.38414419e-01 -5.29967308e-01
-2.33833238e-01 -5.37337601e-01 -1.34001088e+00 -2.76471883e-01
-4.49285537e-01 5.12504518e-01 5.64168632e-01 6.70451701e-01
8.33275676e-01 9.19815719e-01 -2.94194371e-01 -4.62826490e-01
-2.22830608e-01 -9.64985073e-01 -3.06508958e-01 2.60360032e-01
1.92334354e-01 -5.94655037e-01 -2.34090969e-01 -3.24278712e-01] | [8.505561828613281, 8.753168106079102] |
4061bbb9-4b68-431f-9caf-f8791d1bbe5b | probabilistic-inference-for-camera | 1910.13740 | null | https://arxiv.org/abs/1910.13740v1 | https://arxiv.org/pdf/1910.13740v1.pdf | Probabilistic Inference for Camera Calibration in Light Microscopy under Circular Motion | Robust and accurate camera calibration is essential for 3D reconstruction in light microscopy under circular motion. Conventional methods require either accurate key point matching or precise segmentation of the axial-view images. Both remain challenging because specimens often exhibit transparency/translucency in a light microscope. To address those issues, we propose a probabilistic inference based method for the camera calibration that does not require sophisticated image pre-processing. Based on 3D projective geometry, our method assigns a probability on each of a range of voxels that cover the whole object. The probability indicates the likelihood of a voxel belonging to the object to be reconstructed. Our method maximizes a joint probability that distinguishes the object from the background. Experimental results show that the proposed method can accurately recover camera configurations in both light microscopy and natural scene imaging. Furthermore, the method can be used to produce high-fidelity 3D reconstructions and accurate 3D measurements. | ['Ge Yang', 'Yuanhao Guo', 'Fons J. Verbeek'] | 2019-10-30 | null | null | null | null | ['key-point-matching'] | ['natural-language-processing'] | [ 3.86069924e-01 -3.03032964e-01 2.74975508e-01 -2.13147894e-01
-6.61676228e-01 -5.30509949e-01 1.90052792e-01 -2.92373523e-02
-5.08732736e-01 4.82724369e-01 -4.64397669e-01 -2.43376151e-01
6.23918474e-02 -5.66910267e-01 -5.94776511e-01 -9.62621093e-01
6.13615751e-01 8.42178941e-01 5.03550351e-01 7.96375573e-01
5.38650811e-01 9.33957040e-01 -1.10354626e+00 -3.87435704e-01
4.44680065e-01 7.47619271e-01 7.56169081e-01 8.39649737e-01
-1.17462695e-01 2.54868537e-01 -3.95501196e-01 -3.43752086e-01
2.81998992e-01 -1.75798148e-01 -7.14479685e-01 5.46866477e-01
3.26221138e-01 -5.74816346e-01 2.16487162e-02 1.28456450e+00
3.15358818e-01 -1.44664682e-02 8.53249848e-01 -7.24056661e-01
-8.16276893e-02 -2.61371117e-02 -8.39078069e-01 -7.21175689e-03
4.26931709e-01 8.36373121e-02 5.17189920e-01 -8.78666222e-01
7.09975243e-01 8.04197192e-01 3.50666910e-01 4.54561502e-01
-1.36482346e+00 -3.62818211e-01 -3.74794215e-01 -3.87979150e-02
-1.43395591e+00 -3.81063282e-01 8.65100980e-01 -7.42948115e-01
3.83601487e-01 2.28587426e-02 6.87497497e-01 4.58853006e-01
5.51927984e-01 1.99985638e-01 1.49325919e+00 -5.49257815e-01
2.72666723e-01 1.72100529e-01 -4.81469557e-02 7.26374447e-01
4.86017823e-01 -2.20703378e-01 -2.75559068e-01 -1.67449996e-01
1.47935438e+00 2.44468391e-01 -4.43717450e-01 -5.25715649e-01
-1.38816071e+00 1.73517972e-01 -1.03213623e-01 6.91954494e-02
-1.51538044e-01 2.62690578e-02 -2.04863325e-01 -4.87830520e-01
-1.87308956e-02 3.12043011e-01 2.13045720e-02 6.62185848e-02
-8.36477757e-01 -1.64312080e-01 5.91872752e-01 9.47179556e-01
7.92698383e-01 -4.05397028e-01 4.25739765e-01 4.50605720e-01
4.56244648e-01 8.50896835e-01 7.09633306e-02 -1.28589153e+00
-2.92053563e-03 5.76650202e-01 2.67757267e-01 -8.42531085e-01
-6.47844672e-02 -2.12140922e-02 -6.09279573e-01 4.52567160e-01
7.55874872e-01 3.95000279e-01 -8.38003218e-01 1.11710215e+00
6.62293494e-01 2.10282624e-01 -1.16443418e-01 9.07529175e-01
6.01091385e-01 4.25386906e-01 -6.55739009e-01 -5.13682961e-01
1.16784739e+00 -2.50525504e-01 -5.87339878e-01 3.71504799e-02
-1.35131270e-01 -1.11099231e+00 8.27110231e-01 4.28804636e-01
-1.16758037e+00 -1.23433761e-01 -8.67688715e-01 -5.60309142e-02
3.88324708e-01 1.29772291e-01 3.29015642e-01 5.13781130e-01
-6.15806043e-01 3.03982794e-01 -1.15555239e+00 -1.90060422e-01
2.10434809e-01 4.12513733e-01 -5.18905103e-01 -2.67199606e-01
-1.00213505e-01 7.58934855e-01 1.23931713e-01 5.32095060e-02
-7.94179142e-01 -3.27567548e-01 -4.99192774e-01 -1.84089825e-01
1.34449854e-01 -8.03438663e-01 9.26175773e-01 -1.06602557e-01
-1.78634202e+00 1.35108924e+00 -3.94716501e-01 2.92225089e-02
4.66796011e-01 1.05214208e-01 2.67851770e-01 6.74049437e-01
-1.30711421e-01 4.20621037e-01 7.42882252e-01 -1.53017938e+00
-3.58707011e-01 -5.90005219e-01 -1.99178025e-01 1.06991336e-01
2.46515989e-01 -3.27271200e-03 -7.42211282e-01 -2.57953852e-02
7.81895459e-01 -8.39973152e-01 -1.42075136e-01 3.71991962e-01
-6.40664935e-01 3.08222681e-01 8.80216122e-01 -3.39109749e-01
5.39090693e-01 -1.96369922e+00 1.56122193e-01 1.27765238e-01
3.06513190e-01 -8.60834420e-02 3.39978278e-01 9.96739715e-02
4.32446688e-01 -1.41626507e-01 -3.20904940e-01 -4.53201622e-01
-3.80204469e-01 2.42091686e-01 -1.19567528e-01 1.13709533e+00
-2.64518619e-01 4.08088982e-01 -6.99521840e-01 -9.31198537e-01
6.37629390e-01 6.80613935e-01 -3.76671553e-01 4.88710791e-01
7.88091645e-02 8.70726705e-01 -4.16871488e-01 6.81252658e-01
9.11118329e-01 -3.93668503e-01 1.40070796e-01 -2.80419558e-01
-3.15947890e-01 -6.39650226e-02 -1.31000495e+00 1.45284629e+00
-1.94226861e-01 5.53096175e-01 3.95352930e-01 -5.77143252e-01
8.36653888e-01 1.70409113e-01 5.88960171e-01 5.49362004e-02
3.42898071e-01 3.52893144e-01 -2.76594043e-01 -4.92332667e-01
3.49843889e-01 -3.97757918e-01 2.85338342e-01 5.70098937e-01
-1.31198749e-01 -1.03326023e+00 -7.78315589e-02 2.03525908e-02
6.45774782e-01 1.72533154e-01 5.24619937e-01 -2.97385573e-01
5.21906674e-01 -2.89708793e-01 8.59736264e-01 4.50700313e-01
-1.62568524e-01 8.86231005e-01 1.59971058e-01 -2.39975184e-01
-1.22717094e+00 -1.14639711e+00 -6.24819636e-01 -2.00998738e-01
6.33564293e-01 1.57742769e-01 -8.29800427e-01 -1.72827125e-01
-2.52347857e-01 1.69348374e-01 -2.26238906e-01 3.63006175e-01
-3.74710590e-01 -6.45730317e-01 -1.13144986e-01 -5.77268610e-03
2.89570779e-01 -4.41907495e-01 -1.02030671e+00 -1.40479859e-02
-2.85732895e-01 -1.38850975e+00 -3.93336833e-01 -5.97771583e-03
-1.06086957e+00 -1.39516437e+00 -4.68825668e-01 -6.40679359e-01
1.19462717e+00 7.50245273e-01 7.99884140e-01 1.49538815e-01
-3.56332928e-01 6.31950140e-01 -1.33308694e-01 -3.43816094e-02
-4.27384496e-01 -4.78534549e-01 1.86583534e-01 -6.96568415e-02
1.99288562e-01 -5.82546890e-01 -6.79065228e-01 7.51766384e-01
-8.16611946e-01 1.88917965e-01 2.78458416e-01 6.93003953e-01
1.21736193e+00 1.85849234e-01 -3.83826643e-01 -6.36963069e-01
-1.49982676e-01 5.94258085e-02 -1.25987506e+00 1.81318104e-01
-2.34296516e-01 -1.40690133e-01 5.18085003e-01 -3.55675280e-01
-9.14779603e-01 3.92070830e-01 2.19120085e-01 -5.14672160e-01
-4.40380275e-01 1.22070394e-01 -3.59056331e-02 -4.38955724e-01
2.70456940e-01 4.65038151e-01 5.22208177e-02 -3.08264196e-01
-1.14560284e-01 5.65269291e-01 8.38498235e-01 -6.35852396e-01
8.35708559e-01 1.25359225e+00 6.11953795e-01 -1.16511476e+00
-4.83553231e-01 -6.94244146e-01 -9.97729778e-01 -4.73417133e-01
1.06288099e+00 -6.32313848e-01 -1.11264920e+00 4.45423275e-01
-1.16076612e+00 -5.71465828e-02 2.29018554e-02 1.02794468e+00
-6.72595918e-01 8.76333714e-01 -6.10093713e-01 -9.34157908e-01
-6.68750182e-02 -1.71465480e+00 1.16572022e+00 4.62594002e-01
9.92157608e-02 -1.03100514e+00 -8.81948322e-03 5.47951996e-01
-4.43941131e-02 1.96156695e-01 7.61480927e-01 2.69061476e-01
-1.24308944e+00 -2.61156917e-01 -7.79022053e-02 8.17938074e-02
2.05321163e-01 6.37871027e-01 -9.62909400e-01 -2.33734891e-01
4.99407798e-01 -1.05027780e-01 4.77619648e-01 7.59172142e-01
1.10601044e+00 1.61374420e-01 -4.52656507e-01 9.77485776e-01
1.67410839e+00 2.51794487e-01 6.03774667e-01 1.63056970e-01
6.60983682e-01 6.67594254e-01 5.25762439e-01 3.77322137e-01
2.20969349e-01 7.81839609e-01 7.35251129e-01 1.18290178e-01
-4.75017391e-02 1.01629414e-01 1.42973363e-01 1.04878855e+00
-1.52809501e-01 -7.79302269e-02 -7.52162755e-01 3.18739831e-01
-1.19829381e+00 -7.96827018e-01 -4.88297731e-01 2.62983680e+00
8.22898984e-01 -1.55300722e-01 -3.09667528e-01 1.15813211e-01
7.71939754e-01 -4.00932133e-01 -5.58382750e-01 1.61719978e-01
-7.41352290e-02 -8.81881043e-02 6.29066110e-01 9.21672344e-01
-8.29247236e-01 6.74776018e-01 6.81147099e+00 4.03076202e-01
-1.07895124e+00 -2.09718794e-01 3.77166122e-01 6.34150729e-02
-3.92019093e-01 1.86124817e-01 -1.12326527e+00 4.57987607e-01
2.79923469e-01 4.44973484e-02 2.66906470e-01 4.39815849e-01
2.68196285e-01 -7.84897208e-01 -9.50749040e-01 1.20921898e+00
5.84877692e-02 -1.07555783e+00 -1.72175728e-02 3.91590953e-01
6.46396935e-01 -1.48364514e-01 -8.35105404e-02 -9.46008027e-01
2.70714819e-01 -5.92438877e-01 6.19937658e-01 6.09903991e-01
9.73280609e-01 -2.95339972e-01 6.10875964e-01 4.80619371e-01
-9.15275931e-01 3.88597935e-01 -6.99382961e-01 3.11081022e-01
3.56217086e-01 1.04468668e+00 -1.23666990e+00 2.08746567e-01
5.04823744e-01 3.55546415e-01 -1.54258549e-01 1.22515559e+00
-2.75137752e-01 3.34491551e-01 -4.95814323e-01 2.00174943e-01
-2.29479119e-01 -7.27047980e-01 9.00413871e-01 8.27605188e-01
4.67001498e-01 2.56702572e-01 1.57605320e-01 9.84913409e-01
1.34327203e-01 -1.86296850e-01 -5.22485971e-01 1.86020628e-01
6.59229815e-01 1.27989590e+00 -1.32370210e+00 -1.25877056e-02
-3.53271008e-01 9.37670469e-01 1.12540707e-01 2.68777937e-01
-5.10717332e-01 -1.09208509e-01 3.04020405e-01 2.87162960e-01
1.54537618e-01 -6.19601309e-01 -2.61635005e-01 -1.33250093e+00
-2.70743296e-02 -2.47364774e-01 -1.80055797e-02 -1.08465278e+00
-9.88276124e-01 3.91125739e-01 1.78092383e-02 -1.22909379e+00
1.29730865e-01 -9.29857671e-01 -3.53696227e-01 7.46256053e-01
-1.43584836e+00 -9.78691161e-01 -3.62476796e-01 4.84590590e-01
2.20299706e-01 2.27769867e-01 7.44961679e-01 -2.24965103e-02
-5.82636297e-01 -1.06240846e-01 2.89731354e-01 -5.87758310e-02
6.70056045e-01 -1.36710393e+00 -1.88007399e-01 1.08698428e+00
-4.91413362e-02 8.70768785e-01 7.30243862e-01 -5.09436369e-01
-1.50325906e+00 -6.31290853e-01 5.55824220e-01 -5.34214556e-01
3.35896164e-01 -1.59981161e-01 -6.92570865e-01 5.82929432e-01
-6.82344437e-02 9.73066837e-02 8.02056849e-01 -5.41712701e-01
-9.77070555e-02 9.55298617e-02 -1.32756412e+00 3.36597621e-01
5.30400813e-01 -5.73388457e-01 -4.43528146e-01 3.22717428e-01
2.42803797e-01 -6.73066914e-01 -9.45072174e-01 4.54724692e-02
7.60454834e-01 -1.07394743e+00 9.43896353e-01 1.26798123e-01
3.07162344e-01 -7.63061702e-01 -3.07194680e-01 -9.17002261e-01
-1.12665787e-01 -5.82639456e-01 1.63591355e-01 8.38826716e-01
-2.29914766e-02 -5.22980571e-01 9.27663445e-01 6.81498110e-01
-1.05139427e-01 -3.16728145e-01 -1.16059852e+00 -6.20416224e-01
-1.83163270e-01 -1.40119165e-01 3.70253146e-01 7.44753718e-01
-1.74641505e-01 8.22546631e-02 -6.05891608e-02 4.53379720e-01
1.34773743e+00 5.33501029e-01 9.04590309e-01 -1.35670400e+00
-2.42701933e-01 -2.48207226e-01 -5.73267341e-01 -1.18333244e+00
-3.76868024e-02 -6.99167848e-01 9.21942145e-02 -1.41918635e+00
8.31727505e-01 -5.13324738e-01 4.58146960e-01 -1.56341061e-01
-2.60032360e-02 3.36169392e-01 -2.02208862e-01 7.04653502e-01
-4.04865295e-01 1.82202682e-01 1.52345967e+00 2.01739132e-01
7.84856675e-04 8.92113075e-02 -2.43539527e-01 9.51152444e-01
4.30268019e-01 -5.11915326e-01 -1.18825935e-01 -5.78683019e-01
2.14166731e-01 2.43326321e-01 4.61537927e-01 -8.65854979e-01
3.31634402e-01 -3.59764516e-01 3.47922564e-01 -8.15503776e-01
5.10029733e-01 -1.37783062e+00 4.33355480e-01 3.97044778e-01
2.37674996e-01 -2.72946626e-01 -2.55826265e-01 7.21201718e-01
6.30376711e-02 -5.87663472e-01 1.34988201e+00 -3.96727413e-01
-7.95981586e-02 2.19601065e-01 -4.00866151e-01 -4.25461411e-01
1.13711810e+00 -5.43879271e-01 -2.14395106e-01 -1.70627579e-01
-5.21816611e-01 -1.54370755e-01 1.25267327e+00 -6.46892607e-01
9.41129863e-01 -1.03418624e+00 -2.82457352e-01 2.86648273e-01
1.44735090e-02 4.69427198e-01 1.25960872e-01 9.56315696e-01
-1.18489408e+00 3.57444942e-01 -6.48558885e-02 -1.17150235e+00
-1.43667412e+00 2.57309288e-01 5.06006658e-01 4.60085161e-02
-6.75960541e-01 7.78027594e-01 2.08698884e-01 -3.89928460e-01
-1.44691944e-01 -4.90913063e-01 6.08290881e-02 -6.48967206e-01
5.76556087e-01 3.64967734e-01 -1.81102410e-01 -8.33223760e-01
-3.27489465e-01 1.16099560e+00 4.26089801e-02 -2.87438244e-01
1.25048995e+00 -4.21384662e-01 -3.23662311e-01 5.05694211e-01
9.81076360e-01 3.02040726e-01 -1.56880689e+00 -2.13233113e-01
-6.41341984e-01 -9.96850371e-01 2.79319137e-01 -2.46255144e-01
-9.68739748e-01 1.11528873e+00 1.90524399e-01 -4.84624803e-02
8.54097843e-01 1.24040164e-01 5.28815389e-01 2.30774760e-01
7.35890329e-01 -7.40909338e-01 -8.36686715e-02 1.56857282e-01
4.09405380e-01 -1.14531589e+00 4.17925715e-01 -7.11224556e-01
-1.61197603e-01 1.32024837e+00 3.39892983e-01 4.12261561e-02
4.71933216e-01 4.16355997e-01 -1.05274317e-03 -3.08533579e-01
-3.35812062e-01 1.57991499e-01 -4.06138115e-02 6.53909326e-01
1.86334237e-01 -7.42686400e-03 -1.89728532e-02 -1.07156724e-01
1.32232875e-01 -2.38091543e-01 9.82055843e-01 7.75849819e-01
-6.15153670e-01 -1.01322687e+00 -7.85741627e-01 1.37522846e-01
-5.02018452e-01 2.69703746e-01 -3.21562797e-01 4.85867620e-01
-2.65402704e-01 7.35273838e-01 2.29190379e-01 -4.07543927e-02
5.95954582e-02 -3.55798334e-01 1.03083193e+00 -7.04833746e-01
1.79960161e-01 6.29311800e-01 -6.10656798e-01 -3.31813753e-01
-7.87135184e-01 -9.10098732e-01 -1.43978238e+00 -2.32066318e-01
-7.22229540e-01 9.96946171e-02 8.56513679e-01 6.87978804e-01
-3.49279009e-02 -1.26987994e-01 7.20461607e-01 -9.12675083e-01
-1.64723247e-01 -6.42012358e-01 -8.93656731e-01 1.68619469e-01
2.58732080e-01 -7.51313627e-01 -7.05064297e-01 3.78564447e-01] | [9.642438888549805, -2.807058572769165] |
86cbcf8a-88cb-480c-bb5a-fb2b86103544 | hub-at-semeval-2021-task-1-fusion-of-sentence | null | null | https://aclanthology.org/2021.semeval-1.75 | https://aclanthology.org/2021.semeval-1.75.pdf | hub at SemEval-2021 Task 1: Fusion of Sentence and Word Frequency to Predict Lexical Complexity | In this paper, we propose a method of fusing sentence information and word frequency information for the SemEval 2021 Task 1-Lexical Complexity Prediction (LCP) shared task. In our system, the sentence information comes from the RoBERTa model, and the word frequency information comes from the Tf-Idf algorithm. Use Inception block as a shared layer to learn sentence and word frequency information We described the implementation of our best system and discussed our methods and experiments in the task. The shared task is divided into two sub-tasks. The goal of the two sub-tasks is to predict the complexity of a predetermined word. The shared task is divided into two subtasks. The goal of the two subtasks is to predict the complexity of a predetermined word. The evaluation index of the task is the Pearson correlation coefficient. Our best performance system has Pearson correlation coefficients of 0.7434 and 0.8000 in the single-token subtask test set and the multi-token subtask test set, respectively. | ['Xiaobing Zhou', 'Yang Bai', 'Bo Huang'] | 2021-08-01 | null | null | null | semeval-2021 | ['lexical-complexity-prediction'] | ['natural-language-processing'] | [-2.48171091e-01 -2.53982395e-01 -1.92718908e-01 -5.43035626e-01
-8.98728669e-01 -3.52650762e-01 4.61897999e-01 1.19236588e-01
-9.57551897e-01 8.85661840e-01 2.82676756e-01 -3.63782197e-01
7.57002532e-02 -6.85677767e-01 -1.78258836e-01 -4.59998339e-01
-1.05350032e-01 1.55256480e-01 3.39455247e-01 -4.32209074e-01
5.78843415e-01 -1.77794933e-01 -1.60873497e+00 8.80841911e-01
7.93285251e-01 1.03442168e+00 8.82335246e-01 9.93694246e-01
-5.64093649e-01 7.59110332e-01 -8.21251273e-01 -2.39653252e-02
9.86098349e-02 -3.62632304e-01 -1.32779253e+00 -5.24053097e-01
6.28721062e-03 1.51930824e-01 8.25950969e-03 8.63818705e-01
6.10964656e-01 4.78766888e-01 7.22962618e-01 -8.24802697e-01
-1.93840593e-01 8.47551942e-01 -2.57903129e-01 5.94070792e-01
7.35241234e-01 -1.83316931e-01 1.34729445e+00 -1.11345494e+00
3.74746859e-01 1.25345528e+00 6.13064110e-01 1.95263714e-01
-7.05283701e-01 -5.42032540e-01 2.35271871e-01 6.14314020e-01
-1.54227149e+00 -2.22309858e-01 4.00798291e-01 -4.68283147e-01
1.81320977e+00 3.36969227e-01 3.69185030e-01 7.36425936e-01
7.81545341e-01 8.10993612e-01 1.12339330e+00 -8.04369986e-01
6.85793674e-03 2.53397673e-01 6.90896928e-01 4.92514074e-01
-1.44734994e-01 -2.52499580e-01 -6.10578299e-01 3.81606072e-03
1.72586650e-01 -1.52047619e-01 -2.02057738e-04 7.02730894e-01
-9.31681871e-01 8.60047281e-01 -9.65620354e-02 8.30332875e-01
-1.33366331e-01 -1.99785009e-01 6.66438341e-01 6.86963141e-01
7.95256674e-01 4.36210215e-01 -1.11555302e+00 -3.62638444e-01
-8.96171868e-01 3.84350270e-01 1.08155620e+00 8.19048464e-01
5.56208372e-01 -3.74916553e-01 -5.12651861e-01 1.17830300e+00
2.84335792e-01 3.29346001e-01 9.99626875e-01 -2.43095979e-01
6.86430097e-01 2.71431714e-01 3.79801430e-02 -6.80179060e-01
-6.84263945e-01 -2.35208288e-01 -5.25527537e-01 -4.17325169e-01
1.96052134e-01 -3.78833026e-01 -6.78963006e-01 1.43422639e+00
-2.25690037e-01 -1.99681252e-01 -1.46491513e-01 4.95228082e-01
1.05268180e+00 8.91049743e-01 2.24392191e-01 -5.88196337e-01
1.80596542e+00 -1.09507036e+00 -9.80410814e-01 -1.78296059e-01
1.24037242e+00 -1.19024825e+00 1.19331932e+00 3.63625169e-01
-1.01991129e+00 -9.00305688e-01 -1.06254268e+00 -1.33656576e-01
-6.80147290e-01 3.26279491e-01 3.20707470e-01 4.87999618e-01
-7.53979266e-01 5.72758794e-01 -2.89665163e-01 -2.87985742e-01
-2.63250411e-01 1.61294952e-01 -1.61377802e-01 2.30403438e-01
-1.86487222e+00 1.24666667e+00 6.36728704e-01 -4.54490364e-01
-4.26235139e-01 -4.91418004e-01 -7.28681266e-01 2.98094422e-01
2.57637948e-01 -4.17128026e-01 1.36007011e+00 -7.05450118e-01
-1.53968632e+00 7.50692964e-01 -3.73084247e-01 -3.36260796e-01
9.67822820e-02 -2.56719559e-01 -6.29974484e-01 -4.26266640e-01
4.00076449e-01 1.34608671e-01 4.55541134e-01 -5.33166230e-01
-1.10780478e+00 -6.79453760e-02 -1.55376911e-01 3.81110638e-01
-4.60313022e-01 2.02597290e-01 -3.14827472e-01 -6.48016870e-01
-3.24907243e-01 -5.28998852e-01 -9.38915089e-02 -1.08029449e+00
-2.99963385e-01 -8.74310732e-01 3.42978448e-01 -7.19771326e-01
2.05965972e+00 -2.13416600e+00 -9.36079547e-02 -7.83432834e-03
-9.92125198e-02 -1.07262190e-02 -2.46054679e-01 7.06976652e-01
-2.53933102e-01 1.47813380e-01 3.26998562e-01 -2.41511211e-01
-4.68457378e-02 -2.31376022e-01 -5.89210540e-02 -1.36695921e-01
-1.05021290e-01 6.27527833e-01 -7.40592301e-01 -5.52955985e-01
-6.42756298e-02 -2.64708757e-01 -3.06702644e-01 2.62488484e-01
-1.71644792e-01 -2.40934297e-01 -3.43379229e-01 1.92758255e-02
3.94756019e-01 -1.87258795e-02 1.78428277e-01 -1.84522852e-01
-3.82908672e-01 8.41830611e-01 -1.09855866e+00 1.67757273e+00
-8.13234627e-01 3.66583824e-01 -3.73922497e-01 -7.11708665e-01
7.90508986e-01 3.87546897e-01 4.44018573e-01 -8.22265029e-01
2.86152989e-01 1.39793336e-01 4.64114100e-02 -8.17545056e-01
7.72879303e-01 -2.49395534e-01 -2.93697774e-01 5.99300444e-01
1.77062288e-01 1.12872850e-02 6.27772033e-01 4.21100073e-02
1.28111529e+00 -1.28475323e-01 7.90805399e-01 -6.99999750e-01
8.80353034e-01 -1.22777425e-01 3.29052120e-01 5.07919610e-01
-7.79386908e-02 1.75477266e-01 3.64450008e-01 -6.02052093e-01
-8.73791814e-01 -6.75269961e-01 -1.69212416e-01 1.62317836e+00
-2.47659236e-01 -9.99138892e-01 -4.73144501e-01 -6.89493597e-01
-1.80535719e-01 8.56482863e-01 -4.58542705e-01 -9.75941122e-02
-2.43401751e-01 -7.19667315e-01 4.27572250e-01 4.89561528e-01
5.52811086e-01 -1.00699997e+00 -1.84507713e-01 2.24808320e-01
-6.57673895e-01 -1.04126036e+00 -9.85206306e-01 3.57858300e-01
-4.20108616e-01 -8.13975275e-01 -4.97072339e-01 -1.22201276e+00
1.36396840e-01 2.56728053e-01 1.20393360e+00 1.02502166e-03
-3.06746691e-01 -1.61708251e-01 -5.64620435e-01 -4.17892247e-01
-6.36002570e-02 4.24908012e-01 1.87339574e-01 -2.92274773e-01
8.49009931e-01 -2.22825930e-02 -3.34168881e-01 3.12529266e-01
-4.09495801e-01 -1.17266236e-03 3.94800127e-01 7.30099380e-01
3.45457256e-01 2.89460838e-01 5.91020823e-01 -8.64885211e-01
1.08333385e+00 -4.73813415e-01 -2.46549994e-01 3.35841596e-01
-5.98415017e-01 5.63800894e-02 6.15660310e-01 -3.32941115e-01
-8.92110169e-01 7.43146762e-02 -4.01635468e-01 2.58102685e-01
1.53411701e-01 1.00678587e+00 -1.64576828e-01 6.46161616e-01
5.89354038e-01 2.20494911e-01 -2.46856079e-01 -6.84207797e-01
-1.21852225e-02 1.09586918e+00 -2.08013728e-02 -4.48780537e-01
1.27246544e-01 -5.50630033e-01 -4.49902743e-01 -1.03740311e+00
-1.12274599e+00 -9.17039752e-01 -5.23420334e-01 -1.49263993e-01
9.12849545e-01 -1.02606940e+00 -6.72318518e-01 4.16011810e-01
-1.25717926e+00 -1.58000708e-01 -1.54781923e-01 6.64382637e-01
-4.32538092e-01 4.39926118e-01 -7.03828514e-01 -7.78952479e-01
-5.22826731e-01 -9.14311111e-01 8.25862586e-01 -1.05867639e-01
-5.50286472e-01 -1.05355322e+00 2.39676848e-01 3.05387050e-01
3.08905452e-01 -4.66792107e-01 1.21102130e+00 -1.20135379e+00
2.68646926e-01 -3.44193101e-01 4.98854928e-02 4.91005629e-01
2.71588653e-01 -1.88222483e-01 -8.26299787e-01 -1.84704855e-01
2.44329914e-01 -2.93564618e-01 8.57476473e-01 4.82865512e-01
1.32989395e+00 -7.85198584e-02 -1.43143803e-01 4.83171344e-02
1.14797640e+00 1.06253214e-01 4.75127161e-01 1.93075255e-01
2.47630984e-01 5.81272840e-01 7.52673447e-01 3.88298392e-01
6.30139053e-01 6.78830087e-01 -2.62879789e-01 3.16520333e-01
-3.91262360e-02 -1.05117010e-02 6.91578507e-01 1.65451562e+00
-7.54770637e-02 -3.07213366e-01 -1.10374689e+00 3.87217849e-01
-1.77446878e+00 -1.05621386e+00 -4.46880847e-01 2.14889693e+00
1.00917828e+00 4.33780491e-01 3.80607545e-01 2.31751427e-01
6.51700079e-01 1.35371834e-01 1.16214760e-01 -8.82181406e-01
7.68297613e-02 4.14974950e-02 1.53067723e-01 9.07505393e-01
-1.08740449e+00 9.41464365e-01 6.89964962e+00 1.29418743e+00
-5.43187082e-01 1.98502049e-01 4.16793317e-01 -2.59113997e-01
-1.31150847e-02 -2.44784459e-01 -1.20913196e+00 7.75187850e-01
1.45687842e+00 -6.27539217e-01 2.17716917e-01 6.77584112e-01
5.71028143e-02 -4.62213099e-01 -1.04483020e+00 8.12549233e-01
2.96150684e-01 -9.61240470e-01 -3.04665882e-02 -1.31039321e-01
3.86937886e-01 -3.42346844e-03 -2.82552361e-01 9.71911192e-01
6.71871975e-02 -1.03319597e+00 4.30087626e-01 6.16221726e-01
8.51523757e-01 -8.98599625e-01 1.19060838e+00 7.28765726e-01
-1.54624093e+00 6.04020543e-02 -3.65363866e-01 -5.62154889e-01
-1.80678397e-01 8.88234377e-01 -5.59421778e-01 5.77597260e-01
5.67505121e-01 6.54124558e-01 -2.51201600e-01 7.33989596e-01
-1.15616597e-01 4.50221747e-01 -8.25279504e-02 -5.16694486e-01
8.21197927e-02 8.64162445e-02 2.10744381e-01 1.67489290e+00
-3.58527340e-02 -2.50336062e-02 5.62208891e-01 1.38356715e-01
-1.37343751e-02 7.12563455e-01 -3.54600489e-01 1.52955800e-01
4.99214768e-01 1.12021482e+00 -4.21641648e-01 -4.99742478e-01
-5.79157352e-01 7.54851699e-01 3.36826056e-01 5.80524430e-02
-7.54978776e-01 -9.95918870e-01 3.35206509e-01 -1.18984148e-01
2.80663490e-01 -1.41937703e-01 -5.16835928e-01 -1.08602881e+00
-6.82636127e-02 -6.07998550e-01 3.65247041e-01 -3.25878501e-01
-1.59130180e+00 7.93694198e-01 1.69608518e-01 -1.10313606e+00
-2.95775026e-01 -7.22653508e-01 -5.99144042e-01 1.54629624e+00
-1.30299020e+00 -5.75799704e-01 -1.60484128e-02 5.97153544e-01
1.09600210e+00 -4.74990338e-01 1.10599244e+00 2.93586075e-01
-4.21030968e-01 6.45396769e-01 7.66638815e-02 1.27315670e-01
8.03850055e-01 -1.22863889e+00 3.26388061e-01 4.28358138e-01
-1.65251657e-01 4.93122637e-01 5.71048439e-01 -7.33983159e-01
-8.64561379e-01 -8.91389966e-01 1.86141706e+00 -3.17230582e-01
7.59916723e-01 -4.04840112e-01 -5.71183145e-01 2.91857839e-01
1.37711257e-01 -3.76806140e-01 1.13980782e+00 8.09888005e-01
-2.53716916e-01 4.82462868e-02 -8.20507646e-01 2.62272179e-01
6.25540078e-01 -7.28483200e-01 -9.09022748e-01 6.70878530e-01
8.22719872e-01 -3.11894208e-01 -1.10358667e+00 9.51781645e-02
5.53724825e-01 -4.14575279e-01 4.95531231e-01 -6.83236778e-01
7.22797394e-01 -4.32525948e-02 -3.40878010e-01 -1.43104815e+00
-6.55650675e-01 -1.01477936e-01 -4.92384881e-02 8.63050163e-01
8.75719428e-01 -4.93391871e-01 4.34506625e-01 2.74174184e-01
-9.51197669e-02 -9.33657289e-01 -9.64836776e-01 -1.03329587e+00
2.39006042e-01 -6.01429224e-01 3.35865706e-01 8.30103338e-01
5.62565744e-01 1.09273326e+00 -2.46635541e-01 -3.91109884e-01
-5.39624225e-03 7.84698650e-02 2.92630732e-01 -1.00401938e+00
-3.46542805e-01 -3.88837218e-01 -4.21137139e-02 -1.01164401e+00
1.46474928e-01 -1.02502322e+00 1.09219953e-01 -1.43260455e+00
5.23909032e-01 6.88666552e-02 -5.57865441e-01 5.31181693e-01
-4.55607623e-01 -2.90711522e-01 2.67204225e-01 9.09949467e-02
-6.77552879e-01 4.04960334e-01 1.02911544e+00 -7.91108161e-02
-3.00254554e-01 1.14250407e-01 -4.59554464e-01 4.74237800e-01
9.94304419e-01 -5.11542916e-01 -4.53337997e-01 -3.47186834e-01
4.88678485e-01 -1.69771456e-03 -4.13905978e-01 -8.04910362e-01
2.98006713e-01 -1.52709424e-01 2.23663554e-01 -7.75346339e-01
7.37784430e-02 -5.35264730e-01 -3.56135339e-01 5.89478552e-01
-4.91761565e-01 3.70281279e-01 1.97266951e-01 1.90948263e-01
-2.70330846e-01 -5.28056860e-01 5.14182031e-01 -2.33610287e-01
-4.49510634e-01 -1.38854040e-02 -8.40917647e-01 1.19807996e-01
9.51091170e-01 2.35529877e-02 -1.82726547e-01 -4.21344489e-02
-9.46703315e-01 4.66131896e-01 -2.43393332e-01 6.94367826e-01
5.36213279e-01 -1.24139130e+00 -1.00395846e+00 2.07658514e-01
9.29386467e-02 -5.15986860e-01 1.19920604e-01 7.18546748e-01
-2.51207232e-01 8.99728060e-01 -5.95500395e-02 -2.17468187e-01
-1.52427876e+00 3.91240716e-01 1.53453723e-01 -7.93185890e-01
-1.65537179e-01 8.69759321e-01 -1.97640806e-01 -2.72836655e-01
2.84654200e-01 -3.66062492e-01 -5.87837815e-01 3.88394207e-01
9.32556510e-01 4.24899846e-01 4.55216110e-01 -2.97371268e-01
-4.43492234e-01 3.78982753e-01 -6.97371721e-01 1.62780359e-02
1.18560922e+00 -3.98927741e-02 -3.20289612e-01 9.74979520e-01
1.40409064e+00 -8.26378688e-02 -1.47973523e-01 -4.22656119e-01
4.12446827e-01 -1.13351934e-01 2.01067060e-01 -1.13133991e+00
-4.15993899e-01 5.27746201e-01 4.78387326e-01 5.26428044e-01
9.33582902e-01 -2.85224706e-01 8.67984474e-01 4.31592166e-01
4.64117020e-01 -1.61339450e+00 1.56317558e-02 1.47178531e+00
8.10276330e-01 -1.08589280e+00 1.58415899e-01 -3.79511237e-01
-7.61324644e-01 1.04641342e+00 6.36718988e-01 2.43002363e-02
9.79649782e-01 3.71685088e-01 -4.84006777e-02 -1.34733655e-02
-1.33188987e+00 -1.35688990e-01 5.87095022e-01 4.18081693e-02
1.07860827e+00 2.88015962e-01 -1.08970499e+00 1.14977980e+00
-4.87937003e-01 -1.75664052e-01 1.60161585e-01 8.40092719e-01
-8.32864761e-01 -1.21458554e+00 -1.36250943e-01 8.66702616e-01
-5.30510545e-01 -5.41035712e-01 -3.03782463e-01 4.59995300e-01
2.07698449e-01 1.20736635e+00 7.66304284e-02 -8.94067347e-01
4.69772339e-01 5.27574778e-01 3.18062276e-01 -1.08353603e+00
-1.05319393e+00 -9.74948928e-02 4.14418578e-01 -6.38227761e-01
-2.73965914e-02 -5.47487676e-01 -1.18846905e+00 -3.87749672e-01
-4.16941017e-01 4.57738817e-01 8.09489489e-01 1.12602699e+00
6.47521168e-02 7.34284818e-01 8.96128654e-01 -4.10073906e-01
-5.96776426e-01 -1.54133904e+00 -6.97417021e-01 1.13494426e-01
-7.06308428e-03 -2.97497928e-01 -3.70676339e-01 -1.60259783e-01] | [10.643777847290039, 10.513359069824219] |
501732c1-2c3e-4231-8474-56e9ae0fafd5 | beyond-white-ground-truth-colors-for-color | null | null | http://openaccess.thecvf.com/content_iccv_2015/html/Cheng_Beyond_White_Ground_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Cheng_Beyond_White_Ground_ICCV_2015_paper.pdf | Beyond White: Ground Truth Colors for Color Constancy Correction | A limitation in color constancy research is the inability to establish ground truth colors for evaluating corrected images. Many existing datasets contain images of scenes with a color chart included; however, only the chart's neutral colors (grayscale patches) are used to provide the ground truth for illumination estimation and correction. This is because the corrected neutral colors are known to lie along the achromatic line in the camera's color space (i.e. R=G=B) ; the correct RGB values of the other color patches are not known. As a result, most methods estimate a 3*3 diagonal matrix that ensures only the neutral colors are correct. In this paper, we describe how to overcome this limitation. Specifically, we show that under certain illuminations, a diagonal 3*3 matrix is capable of correcting not only neutral colors, but all the colors in a scene. This finding allows us to find the ground truth RGB values for the color chart in the camera's color space. We show how to use this information to correct all the images in existing datasets to have correct colors. Working from these new color corrected datasets, we describe how to modify existing color constancy algorithms to perform better image correction. | ['Brian Price', 'Scott Cohen', 'Michael S. Brown', 'Dongliang Cheng'] | 2015-12-01 | null | null | null | iccv-2015-12 | ['color-constancy'] | ['computer-vision'] | [ 1.35013804e-01 -5.58835864e-01 6.04511313e-02 -2.46344894e-01
-4.91074026e-02 -9.73581016e-01 -1.11981221e-01 -2.64445156e-01
-1.81019649e-01 6.25031114e-01 -2.72670984e-01 -2.79990971e-01
2.92638302e-01 -7.59251475e-01 -6.46629155e-01 -8.79900217e-01
3.74184221e-01 -1.38906047e-01 1.96736440e-01 -4.52997863e-01
4.75250483e-01 5.63679278e-01 -1.64626098e+00 1.98568419e-01
8.96782637e-01 7.52720416e-01 -6.30789027e-02 1.01299608e+00
-2.91790634e-01 6.41682684e-01 -7.55017877e-01 -3.26713711e-01
6.95551038e-01 -9.62400675e-01 -4.32300150e-01 2.87116855e-01
6.91652536e-01 -4.38093454e-01 2.31112894e-02 1.63832939e+00
5.90336509e-02 4.23124200e-03 5.23463905e-01 -1.65054476e+00
-1.25539112e+00 -5.94631769e-02 -9.88053739e-01 -4.18812484e-01
3.18656534e-01 -1.93268254e-01 9.34774160e-01 -7.62119830e-01
5.54112732e-01 1.04277992e+00 3.71445835e-01 4.87160176e-01
-1.17745221e+00 -6.04339361e-01 2.72127800e-02 2.91272968e-01
-1.46575844e+00 -7.29632303e-02 8.81850302e-01 -1.21480048e-01
4.24366444e-01 9.33808088e-01 1.12998998e+00 2.41825506e-01
2.40334958e-01 2.69570649e-01 1.65234101e+00 -9.18065190e-01
2.88211316e-01 3.72603774e-01 -1.33351102e-01 7.38773167e-01
7.07196593e-01 -6.43424168e-02 -2.99266785e-01 1.73390493e-01
9.57374752e-01 -9.99281257e-02 -3.46077561e-01 -5.43765008e-01
-1.16117334e+00 5.42584419e-01 7.16139436e-01 2.68782645e-01
-5.93567602e-02 7.56473932e-03 -2.57196665e-01 2.92158544e-01
3.25191952e-02 5.72792530e-01 -2.49116257e-01 1.64330557e-01
-5.99244773e-01 -1.97478995e-01 6.91596746e-01 1.01289940e+00
1.35720730e+00 2.22111106e-01 6.67835623e-02 6.74709260e-01
7.65726194e-02 1.02657199e+00 9.66018587e-02 -1.25351501e+00
1.01990163e-01 8.01338553e-01 4.54807341e-01 -1.43573654e+00
-3.62542212e-01 -8.55351761e-02 -7.80949950e-01 6.08921647e-01
7.92900443e-01 2.54806299e-02 -1.07706559e+00 1.49790084e+00
1.91581994e-01 -9.88577828e-02 4.00662981e-02 1.19395244e+00
4.13041592e-01 6.43540680e-01 -3.53978097e-01 -1.16829030e-01
1.21844637e+00 -6.10337615e-01 -8.95526588e-01 -2.72897810e-01
3.56825590e-01 -1.19243467e+00 1.49074996e+00 5.68714261e-01
-8.02458584e-01 -4.85177606e-01 -1.24584568e+00 -3.15574780e-02
-6.61063969e-01 3.61006558e-01 7.94114470e-01 1.10158396e+00
-1.25009131e+00 1.28267246e-04 -3.28674763e-01 -2.93573648e-01
-3.46185088e-01 -7.30883554e-02 -3.66195500e-01 -2.44958445e-01
-9.21974003e-01 1.06054330e+00 1.51144683e-01 4.47963715e-01
-1.64568558e-01 -2.69423872e-01 -4.83402908e-01 -1.98946297e-01
7.15937614e-02 -1.84459463e-01 7.48898625e-01 -1.70577145e+00
-1.27772093e+00 9.24159706e-01 -1.93446651e-01 3.68717939e-01
6.26377821e-01 2.20316827e-01 -5.73957324e-01 1.62877843e-01
-6.76058829e-02 5.20043612e-01 6.64813578e-01 -2.11466360e+00
-6.65899158e-01 -2.36240000e-01 8.82624388e-02 8.58419389e-02
-2.69195706e-01 -1.34433433e-01 -9.50897992e-01 -3.70492160e-01
7.80193686e-01 -1.27938616e+00 4.11209576e-02 3.08583081e-01
-7.44003057e-01 5.74785590e-01 4.88120407e-01 -7.32059598e-01
1.00336754e+00 -2.20064735e+00 -3.69026005e-01 7.74116874e-01
1.27529979e-01 -1.16681486e-01 -2.08763465e-01 1.26902908e-01
-4.18848395e-01 -3.78773101e-02 -2.03837693e-01 9.96258631e-02
-9.70032662e-02 3.86894226e-01 2.21609417e-02 7.44550228e-01
-1.70977041e-01 5.10483801e-01 -8.55247855e-01 -5.20706117e-01
3.81107897e-01 5.41153431e-01 -6.03986263e-01 2.07417756e-02
2.23855659e-01 2.80278265e-01 3.33177522e-02 7.65617490e-01
1.47933090e+00 2.54293263e-01 2.08253488e-01 -5.51293552e-01
-4.91263807e-01 -5.10375619e-01 -1.59074605e+00 1.12949526e+00
-1.63583159e-01 9.36993837e-01 -1.80597574e-01 -4.56079274e-01
1.00419736e+00 -1.88645273e-01 5.97682357e-01 -1.00328732e+00
1.30565181e-01 2.10949287e-01 1.35646597e-01 -2.37573072e-01
8.63459826e-01 1.90276280e-02 3.04654807e-01 3.84197474e-01
-6.54558599e-01 -5.50975680e-01 3.61826599e-01 2.23177582e-01
3.89216125e-01 -6.32638335e-02 -2.29815263e-02 -1.61201462e-01
4.06761050e-01 3.79103750e-01 7.97630787e-01 6.74227357e-01
-2.62506634e-01 9.46613252e-01 6.55156970e-01 -1.81925043e-01
-1.16014147e+00 -1.17506468e+00 -3.60331647e-02 9.37201321e-01
5.47022223e-01 -1.65210292e-01 -9.49270725e-01 1.71513885e-01
-1.35569051e-01 7.73456752e-01 -9.08303916e-01 -1.92602381e-01
-1.91392645e-01 -8.46152902e-01 7.08295926e-02 3.13539714e-01
7.37734675e-01 -4.81765360e-01 -5.86574733e-01 -3.82602811e-01
-3.58668834e-01 -7.36514986e-01 -4.23796237e-01 6.46033585e-02
-5.35358429e-01 -1.48036730e+00 -5.03235638e-01 -5.92351496e-01
1.53519344e+00 8.78212929e-01 8.71724248e-01 4.57143426e-01
-3.32357854e-01 4.45737183e-01 -5.06837726e-01 -4.07635808e-01
-2.96112746e-01 -7.83761322e-01 -1.99433774e-01 1.86158165e-01
2.54882574e-01 1.34863585e-01 -7.24030554e-01 5.19023180e-01
-1.00394213e+00 3.29402357e-01 4.10147280e-01 4.49569196e-01
7.04583466e-01 4.25513327e-01 -3.60453576e-01 -8.80175233e-01
3.64145249e-01 1.43994778e-01 -8.80037904e-01 7.75116444e-01
-6.64703608e-01 -4.75857444e-02 4.77343142e-01 -2.68216252e-01
-1.11829460e+00 3.52599382e-01 3.69372249e-01 -1.33313268e-01
1.64496914e-01 2.27696411e-02 -2.36953273e-01 -4.57138121e-01
6.72197282e-01 1.55350164e-01 -2.39977613e-01 -2.26833701e-01
6.47998393e-01 3.76157135e-01 7.38737166e-01 -4.31186408e-01
1.07723188e+00 6.46967828e-01 2.02784777e-01 -6.08834684e-01
-5.09719491e-01 -3.58844638e-01 -8.72578502e-01 -6.18635833e-01
7.77659774e-01 -6.31036520e-01 -7.29792535e-01 5.40787339e-01
-9.02249336e-01 -6.36164024e-02 -5.86091727e-02 5.81681907e-01
-3.15538883e-01 4.56175208e-01 -3.33869874e-01 -8.49735796e-01
1.52682751e-01 -1.01440239e+00 5.91793299e-01 4.37842250e-01
2.34913722e-01 -7.92136610e-01 5.49745932e-02 1.91986281e-02
3.24927300e-01 2.90195316e-01 9.87917781e-01 5.76091826e-01
-4.98809427e-01 -2.81604648e-01 -6.59674644e-01 4.92567718e-01
5.17450154e-01 9.21689034e-01 -8.46305847e-01 -1.57923326e-01
-1.44675151e-01 4.02480274e-01 6.76048815e-01 4.09810305e-01
9.85115886e-01 -1.66269884e-01 2.21642002e-01 9.06812847e-01
1.87083876e+00 3.46570760e-01 9.67729270e-01 5.62856913e-01
8.68150532e-01 6.03301108e-01 6.75795197e-01 3.48962337e-01
3.24370831e-01 3.74167144e-01 5.72298169e-01 -7.40962684e-01
-1.40061572e-01 -8.80331248e-02 1.80229470e-01 6.67502224e-01
-4.18497801e-01 9.69066545e-02 -7.80374110e-01 2.44577542e-01
-1.51593304e+00 -6.81910157e-01 -8.96407843e-01 2.48108554e+00
9.91079390e-01 -2.62220830e-01 -1.33338466e-01 3.96551430e-01
9.89078939e-01 -3.22401017e-01 -4.91737545e-01 -5.45581996e-01
-5.94753742e-01 -1.20040677e-01 9.93653357e-01 5.27472436e-01
-7.32777834e-01 7.28644609e-01 6.91513109e+00 4.58115600e-02
-1.28611600e+00 -4.16078597e-01 4.82163727e-01 1.91750348e-01
-5.38005590e-01 3.73429477e-01 -1.18371777e-01 3.02964687e-01
2.37812757e-01 -9.90957320e-02 8.41367245e-01 6.68100297e-01
2.09643066e-01 -7.33717263e-01 -6.72063887e-01 1.31544018e+00
2.85187483e-01 -7.12270439e-01 1.25769079e-01 -1.76822677e-01
1.17610013e+00 -6.57575905e-01 4.00523633e-01 -3.93213063e-01
3.74434531e-01 -6.30011737e-01 9.74317849e-01 6.40468061e-01
9.36039150e-01 -7.62714684e-01 5.21106422e-01 -2.74036974e-01
-9.91295218e-01 1.74262270e-01 -8.62111568e-01 -3.02871987e-02
-3.76805693e-01 5.94288051e-01 -6.48981214e-01 3.06760550e-01
7.57384479e-01 3.96426439e-01 -1.18368280e+00 1.20589852e+00
-4.74836141e-01 1.12364113e-01 -6.56629279e-02 1.00533821e-01
-5.11434190e-02 -7.24142730e-01 -2.81499121e-02 9.04418170e-01
4.61273313e-01 3.25706005e-02 -2.61366844e-01 9.54620898e-01
1.05828546e-01 2.17232749e-01 -2.23545611e-01 2.42815763e-01
3.56188804e-01 1.23922515e+00 -1.14789009e+00 -2.39397272e-01
-5.98508656e-01 1.30622816e+00 -2.46036679e-01 7.96984673e-01
-9.08002913e-01 -5.20118117e-01 6.70248568e-01 -1.75793380e-01
-2.95700192e-01 -3.89397770e-01 -7.61115611e-01 -1.22509968e+00
-1.97829455e-01 -8.40461135e-01 1.78441241e-01 -1.51048803e+00
-9.74307597e-01 3.72399181e-01 -1.47465304e-01 -1.58455420e+00
2.76995748e-01 -9.44554210e-01 -3.88563871e-01 9.26097929e-01
-1.51686776e+00 -1.10260856e+00 -8.65368366e-01 1.11442089e+00
-3.82006556e-01 2.58081824e-01 7.48070300e-01 1.85407594e-01
-5.85602760e-01 4.13034528e-01 5.33531904e-01 3.68324757e-01
1.12249720e+00 -1.54516029e+00 5.28512783e-02 1.28687632e+00
5.63075170e-02 7.43568301e-01 9.73740399e-01 -5.02371550e-01
-1.82714593e+00 -6.90842807e-01 3.28330994e-01 -3.31081390e-01
5.17764747e-01 -1.67386472e-01 -7.39071369e-01 6.13565326e-01
8.02089199e-02 -3.09977643e-02 5.68694651e-01 -1.92935973e-01
-6.49821937e-01 -5.04419804e-01 -1.08885741e+00 8.36588979e-01
5.51560581e-01 -4.42381501e-01 -2.78942008e-02 1.24982461e-01
1.35513365e-01 -4.14742440e-01 -3.42926711e-01 -1.86176702e-01
8.16937268e-01 -1.31953025e+00 8.80415261e-01 -1.56199813e-01
1.65830910e-01 -8.58502746e-01 -3.95963311e-01 -1.36445498e+00
-2.22193226e-01 7.56194536e-03 8.29067647e-01 1.05995703e+00
2.91708052e-01 -8.46464217e-01 5.75487077e-01 1.16853702e+00
1.39133364e-01 1.71748489e-01 -6.24442756e-01 -5.59715867e-01
-1.19626060e-01 -3.04054618e-01 7.12752998e-01 1.11613488e+00
-3.62582415e-01 -4.20533091e-01 -4.79629546e-01 2.40720794e-01
7.27699518e-01 4.05874401e-01 8.49058092e-01 -9.50199425e-01
1.49511039e-01 -3.95927817e-01 -3.15916598e-01 -3.55553567e-01
-2.26861238e-01 -3.44810247e-01 2.01571047e-01 -1.58798885e+00
3.76352727e-01 -6.44506872e-01 -4.08767253e-01 5.68111479e-01
-2.13961661e-01 8.64536345e-01 4.13166046e-01 1.17178269e-01
-2.24543571e-01 -3.59687023e-02 1.61635590e+00 -3.16487044e-01
-1.79796085e-01 -3.47573638e-01 -9.51090455e-01 6.02331221e-01
9.66098726e-01 -2.44529381e-01 -3.39019924e-01 -5.46159208e-01
6.81646645e-01 -4.84962374e-01 3.68521899e-01 -8.70097041e-01
1.90275192e-01 -7.14768648e-01 8.78418326e-01 -4.84762907e-01
2.15511069e-01 -1.18350518e+00 4.25122529e-01 4.41567481e-01
4.02656607e-02 2.33554572e-01 7.10161552e-02 1.78882033e-01
-1.47936121e-01 -1.21226311e-01 1.06045771e+00 -1.65662035e-01
-1.02504969e+00 -3.09883684e-01 -1.72517776e-01 -4.09142077e-01
9.81468678e-01 -7.20427871e-01 -7.29059815e-01 -5.21994531e-01
-2.30766043e-01 -3.57989639e-01 1.19736040e+00 1.58201441e-01
6.44135118e-01 -1.54732323e+00 -4.97266054e-01 4.62767482e-01
2.76119262e-01 -5.19187033e-01 3.84346426e-01 6.07551098e-01
-1.31662166e+00 3.92095670e-02 -8.48737895e-01 -4.52342153e-01
-1.32895267e+00 5.73139012e-01 4.87379283e-01 7.07515538e-01
-1.70947418e-01 6.39282048e-01 2.36235991e-01 -8.53532478e-02
-1.32004768e-01 -6.40109718e-01 -4.78041284e-02 -7.73195475e-02
4.78923321e-01 4.13200945e-01 9.42011774e-02 -8.98084283e-01
-3.86682451e-01 1.09064639e+00 5.94519794e-01 -9.93230939e-02
1.03964853e+00 -3.73742580e-01 -8.22814822e-01 4.82131511e-01
1.17931020e+00 5.80891848e-01 -1.09851527e+00 1.78618416e-01
-5.12810707e-01 -1.33169484e+00 -1.76375173e-02 -8.99798155e-01
-1.47617698e+00 6.53184474e-01 9.04691100e-01 2.45841548e-01
1.72832155e+00 -5.30903459e-01 1.01873197e-01 2.92201757e-01
2.87617654e-01 -1.47203004e+00 -5.86567558e-02 1.84731469e-01
7.70836353e-01 -1.19180322e+00 2.49911398e-01 -5.63647449e-01
-8.96032870e-01 1.38156152e+00 7.33551145e-01 5.51899299e-02
2.82257080e-01 8.73763487e-03 6.30532205e-01 1.02210678e-01
-1.00070551e-01 -3.67829382e-01 2.96456873e-01 8.12988341e-01
5.34074485e-01 3.48371595e-01 -2.21414939e-01 -6.07256778e-02
-3.53886455e-01 -5.24563074e-01 1.09671867e+00 6.84699655e-01
-5.07160008e-01 -1.02078974e+00 -1.16270947e+00 1.38141513e-01
1.15737572e-01 -5.40523455e-02 -7.68259525e-01 7.51569688e-01
2.02274069e-01 1.27531135e+00 1.71365187e-01 -3.45630914e-01
4.88492370e-01 -4.00504231e-01 6.48615479e-01 -9.31783617e-02
-5.66365197e-02 3.48014720e-02 -4.37042922e-01 -6.20562911e-01
-4.84843016e-01 -2.94441581e-01 -1.31421447e+00 -9.00607586e-01
-2.78313369e-01 1.14174537e-01 1.05609965e+00 1.17450573e-01
-2.64866263e-01 3.20457697e-01 8.06344092e-01 -5.09413362e-01
2.36563131e-01 -4.72186595e-01 -1.09373271e+00 7.30671883e-01
3.50408137e-01 -5.30365348e-01 -4.74096209e-01 4.44692135e-01] | [10.461609840393066, -2.5770795345306396] |
78828cc7-c960-416a-8168-e93b64d5dcdb | dialogue-response-ranking-training-with-large | 2009.06978 | null | https://arxiv.org/abs/2009.06978v1 | https://arxiv.org/pdf/2009.06978v1.pdf | Dialogue Response Ranking Training with Large-Scale Human Feedback Data | Existing open-domain dialog models are generally trained to minimize the perplexity of target human responses. However, some human replies are more engaging than others, spawning more followup interactions. Current conversational models are increasingly capable of producing turns that are context-relevant, but in order to produce compelling agents, these models need to be able to predict and optimize for turns that are genuinely engaging. We leverage social media feedback data (number of replies and upvotes) to build a large-scale training dataset for feedback prediction. To alleviate possible distortion between the feedback and engagingness, we convert the ranking problem to a comparison of response pairs which involve few confounding factors. We trained DialogRPT, a set of GPT-2 based models on 133M pairs of human feedback data and the resulting ranker outperformed several baselines. Particularly, our ranker outperforms the conventional dialog perplexity baseline with a large margin on predicting Reddit feedback. We finally combine the feedback prediction models and a human-like scoring model to rank the machine-generated dialog responses. Crowd-sourced human evaluation shows that our ranking method correlates better with real human preferences than baseline models. | ['Xiang Gao', 'Chris Brockett', 'Yizhe Zhang', 'Michel Galley', 'Bill Dolan'] | 2020-09-15 | null | https://aclanthology.org/2020.emnlp-main.28 | https://aclanthology.org/2020.emnlp-main.28.pdf | emnlp-2020-11 | ['conversational-response-selection', 'open-domain-dialog'] | ['natural-language-processing', 'natural-language-processing'] | [-8.03629830e-02 4.91918802e-01 -2.15340763e-01 -9.54292357e-01
-1.00820518e+00 -9.51139450e-01 1.01357424e+00 -2.52250880e-01
-4.09562439e-01 9.89956677e-01 1.10314631e+00 -3.12528461e-01
4.68026578e-01 -4.20275390e-01 -5.44612408e-02 -3.25831361e-02
3.57452780e-02 1.05598879e+00 2.06043646e-01 -8.87891173e-01
2.74630398e-01 -3.01540196e-01 -8.40837777e-01 8.52750957e-01
9.15890872e-01 7.33695686e-01 1.35229036e-01 1.27470231e+00
-1.20889924e-01 1.38032258e+00 -1.00792933e+00 -7.19337285e-01
2.25825608e-01 -6.92859471e-01 -1.19657612e+00 -1.52532086e-01
3.01704407e-01 -7.62500346e-01 -3.52902919e-01 3.24012101e-01
5.41549504e-01 5.55017829e-01 7.08626509e-01 -1.19277608e+00
-6.41158342e-01 9.16374624e-01 -5.85671142e-02 1.15805000e-01
8.25711668e-01 6.28892541e-01 1.51947248e+00 -8.20134997e-01
7.67759919e-01 1.85863841e+00 3.12460542e-01 1.09193909e+00
-1.36781752e+00 -4.20794845e-01 1.58400014e-01 -1.30646870e-01
-2.64265716e-01 -5.27491868e-01 5.27965128e-01 -5.58549464e-01
1.03181660e+00 5.28747261e-01 3.31367463e-01 1.68753111e+00
-2.89675989e-03 9.50585902e-01 9.52234387e-01 -6.62209317e-02
-2.16209926e-02 3.04950744e-01 3.80827665e-01 3.56275111e-01
-7.12804794e-01 1.79054782e-01 -8.10423791e-01 -5.58042109e-01
3.00022036e-01 -2.18781248e-01 -2.42747068e-01 1.58477277e-01
-1.29708898e+00 1.23551548e+00 5.37377477e-01 -3.69338170e-02
-2.11133540e-01 -3.03105831e-01 3.29043031e-01 6.44795358e-01
7.11496651e-01 1.12954569e+00 -6.50129914e-01 -6.24282777e-01
-2.66684771e-01 8.17045569e-01 1.50534463e+00 7.87564337e-01
6.60706997e-01 -3.39799702e-01 -7.54258513e-01 1.37789619e+00
2.96146542e-01 3.08369875e-01 6.15292370e-01 -1.51200676e+00
9.38146412e-01 7.75019050e-01 5.18361449e-01 -6.89668834e-01
-4.38260943e-01 3.87697101e-01 -2.70576596e-01 -1.37799412e-01
7.84497380e-01 -7.70537436e-01 -2.42716029e-01 1.86866283e+00
1.49342239e-01 -7.45539427e-01 2.36307710e-01 1.05899858e+00
9.92185712e-01 8.51764619e-01 3.64582799e-02 -1.17245197e-01
1.05015635e+00 -1.42339849e+00 -6.44470096e-01 -6.35059178e-01
8.09954464e-01 -8.81500781e-01 1.48779869e+00 1.82577759e-01
-1.08871198e+00 -2.47087896e-01 -6.43247724e-01 -1.58006862e-01
2.66393483e-01 -2.44517624e-01 7.19849944e-01 3.53438526e-01
-1.09227002e+00 5.47456682e-01 -1.97338879e-01 -4.36042011e-01
-3.00357640e-01 3.01327437e-01 -7.30032921e-02 3.93714368e-01
-1.32135653e+00 1.27970088e+00 -2.27388740e-01 -2.19958514e-01
-7.43210912e-01 -5.58976948e-01 -6.86597407e-01 -1.32415697e-01
2.09875017e-01 -4.55633461e-01 2.31201601e+00 -9.30738628e-01
-2.13859248e+00 7.17572570e-01 -2.28325590e-01 -5.08323431e-01
7.38634706e-01 -3.40578794e-01 -5.52564263e-02 -1.49624690e-01
6.83569163e-02 1.12658978e+00 1.79905310e-01 -1.02718902e+00
-6.13495231e-01 4.70989980e-02 3.65470409e-01 6.21602297e-01
-2.91488379e-01 1.57902032e-01 9.33547392e-02 -1.30142257e-01
-4.84416366e-01 -1.13196051e+00 -3.97966415e-01 -6.36287868e-01
-4.30646867e-01 -7.66082346e-01 4.94141430e-01 -5.04793406e-01
9.83464718e-01 -1.57069385e+00 4.53305915e-02 -2.28124008e-01
3.81766886e-01 5.32022342e-02 -4.42599624e-01 7.60128081e-01
4.86986309e-01 6.12571165e-02 2.07116127e-01 -5.25134146e-01
2.12487638e-01 3.92319262e-03 -4.91684705e-01 -1.95599824e-01
1.32681832e-01 1.07079470e+00 -1.00613463e+00 -4.72298503e-01
7.58677050e-02 -2.14871004e-01 -1.02203560e+00 1.14515126e+00
-7.49875546e-01 8.00411344e-01 -3.74016851e-01 4.26982343e-02
1.46444747e-02 -2.69117534e-01 3.71602386e-01 5.04809558e-01
-4.63757664e-02 1.07846844e+00 -1.77121475e-01 1.41376221e+00
-6.02624238e-01 7.98413396e-01 1.02213390e-01 -2.65380312e-02
1.26953888e+00 3.32467467e-01 3.64418834e-01 -5.85301220e-01
2.13764921e-01 -4.98111593e-03 4.74015206e-01 -4.13924277e-01
1.03828084e+00 -6.89240918e-02 -4.87182915e-01 9.62937713e-01
8.51792470e-02 -4.94484484e-01 1.35974109e-01 7.38106787e-01
1.20681691e+00 -3.88909638e-01 5.32003492e-02 -1.02586657e-01
2.64334530e-01 1.45330697e-01 5.17106652e-01 7.70053744e-01
-6.14304483e-01 4.65583444e-01 8.02117944e-01 -4.21518922e-01
-8.95422339e-01 -7.86891520e-01 5.45923531e-01 1.86174500e+00
-2.12206505e-04 -3.74202013e-01 -7.79252172e-01 -8.19967926e-01
-1.86776742e-01 8.39580476e-01 -3.53677720e-01 -1.66565157e-03
-7.10652649e-01 -4.38252211e-01 5.31152189e-01 2.43104056e-01
3.12978864e-01 -1.38064814e+00 -1.68928415e-01 3.06051821e-01
-9.10074770e-01 -9.67811108e-01 -1.05565679e+00 5.16096465e-02
-5.00122309e-01 -9.40730274e-01 -4.91293043e-01 -6.54489160e-01
9.24401358e-02 3.11718225e-01 1.68687630e+00 1.53771326e-01
5.17442346e-01 3.57010007e-01 -5.06159663e-01 -3.15046310e-01
-1.00641990e+00 2.47841865e-01 4.23579849e-02 -4.26136076e-01
7.17333734e-01 -3.36059570e-01 -7.40118802e-01 7.56476343e-01
-1.25561520e-01 1.54485986e-01 2.64270395e-01 1.08856344e+00
-7.02658594e-01 -1.35254693e+00 9.46561575e-01 -8.98761332e-01
1.58237612e+00 -5.32798827e-01 -5.40556721e-02 1.21751904e-01
-4.42564458e-01 8.56843069e-02 5.52005351e-01 -6.01810396e-01
-1.48539090e+00 -2.15894058e-01 -6.12334348e-03 2.81138659e-01
6.36900365e-02 -1.75973866e-02 1.83878660e-01 4.21655387e-01
1.23227394e+00 -3.57450217e-01 2.16447562e-01 -1.85209796e-01
6.69866204e-01 1.05082607e+00 3.24586034e-01 -8.18397045e-01
2.26116508e-01 -2.90201366e-01 -8.89488637e-01 -5.67647099e-01
-7.91100979e-01 -5.24608433e-01 -1.56917557e-01 -5.12057662e-01
8.01870883e-01 -8.48919809e-01 -1.21512246e+00 1.66947752e-01
-1.70314360e+00 -9.47731197e-01 1.27638385e-01 3.79915178e-01
-7.00196087e-01 1.38596058e-01 -1.29961252e+00 -1.08677447e+00
-4.90104407e-01 -1.13941371e+00 7.17919707e-01 1.65519550e-01
-1.22306848e+00 -9.17519033e-01 6.65557563e-01 8.55928957e-01
4.05834138e-01 -3.18878204e-01 6.83344781e-01 -1.32363033e+00
-4.48614597e-01 6.19314946e-02 7.12519605e-03 1.82578623e-01
-2.79972907e-02 -1.35665417e-01 -1.02119541e+00 1.13646708e-01
-1.20253139e-03 -1.24123359e+00 5.00541568e-01 6.41137660e-02
2.94850022e-01 -8.19294572e-01 -1.64314061e-02 -2.94279993e-01
3.07889014e-01 2.27540940e-01 2.65606701e-01 -8.65240544e-02
2.41739973e-01 1.16712213e+00 8.24143946e-01 5.22398114e-01
7.70066082e-01 7.27080643e-01 1.83772907e-01 2.02248797e-01
2.44888589e-01 -6.54105604e-01 8.34882975e-01 9.68022168e-01
2.10226759e-01 -5.96169055e-01 -6.74657464e-01 4.18072164e-01
-2.10674024e+00 -1.04383016e+00 -2.11182058e-01 1.79877794e+00
1.21777987e+00 1.48373336e-01 5.92676520e-01 -7.40440845e-01
5.62840819e-01 4.04636145e-01 -5.25064111e-01 -9.38654661e-01
1.12908758e-01 -2.65637666e-01 1.89730711e-02 1.14329231e+00
-5.89571893e-01 1.12866879e+00 7.12895441e+00 2.15634391e-01
-6.51669919e-01 5.20905219e-02 1.00626302e+00 -2.83753991e-01
-4.88921911e-01 -1.47205982e-02 -6.63872600e-01 4.49335068e-01
1.19562995e+00 -1.58256575e-01 6.00731909e-01 9.78676021e-01
5.30868173e-01 -3.29353362e-02 -1.54546344e+00 6.21724844e-01
-1.95453703e-01 -1.00944364e+00 -2.40429431e-01 1.96087384e-03
7.57319689e-01 1.83748379e-02 -2.25260198e-01 9.60083663e-01
1.46536183e+00 -1.09045029e+00 4.26597387e-01 2.28397310e-01
1.71681166e-01 -1.98427364e-01 5.46347082e-01 6.92826390e-01
-4.92860347e-01 -1.41720667e-01 -1.91822797e-01 -6.37645900e-01
3.94755661e-01 -1.53867370e-02 -1.67396998e+00 -2.76172131e-01
2.39821419e-01 3.70497495e-01 -1.84661821e-01 2.34898329e-01
-2.65542388e-01 4.83323365e-01 -1.05063185e-01 -7.78197229e-01
3.13921958e-01 -2.44153559e-01 4.70581442e-01 1.19657302e+00
-1.36085764e-01 3.92152667e-01 4.92220521e-01 8.43594134e-01
-3.91061306e-01 -1.51622556e-02 -5.60721934e-01 -5.22646494e-02
7.81997681e-01 1.34016371e+00 9.82477441e-02 -3.58111233e-01
-2.90921718e-01 8.87865782e-01 7.27078557e-01 2.52602428e-01
-6.24270499e-01 9.96535495e-02 9.83350098e-01 -2.09685981e-01
-4.69436705e-01 -5.97861223e-02 -4.12360638e-01 -1.11034513e+00
-1.40517339e-01 -1.32507324e+00 2.44065806e-01 -5.60294449e-01
-1.63511765e+00 8.83981109e-01 -3.32177758e-01 -9.82323885e-01
-1.21378124e+00 -4.44942534e-01 -1.02453101e+00 9.01834786e-01
-7.57352293e-01 -8.20310354e-01 -9.88004208e-02 3.37757677e-01
1.11836660e+00 -2.33500078e-01 8.96798372e-01 -6.03374764e-02
-1.77941680e-01 5.76407373e-01 -4.48281586e-01 2.03886792e-01
1.37767637e+00 -1.49799526e+00 6.48102462e-01 2.30904460e-01
-2.46819437e-01 7.16132760e-01 1.06584346e+00 -5.82285881e-01
-7.86312401e-01 -6.12802088e-01 1.16012442e+00 -1.17580700e+00
7.67230690e-01 -5.31918168e-01 -5.69642246e-01 8.65519404e-01
6.93154097e-01 -7.71511078e-01 7.97502220e-01 6.55838013e-01
-2.95565158e-01 3.60655725e-01 -8.09174120e-01 8.92237425e-01
9.83318746e-01 -5.10145247e-01 -9.51696277e-01 5.08128345e-01
1.07213712e+00 -4.05869484e-01 -6.99094713e-01 -6.69590458e-02
5.71020544e-01 -1.23212147e+00 6.79822862e-01 -9.22610760e-01
8.26863706e-01 5.41787326e-01 -2.28621010e-02 -1.70498335e+00
-1.00759611e-01 -1.19495177e+00 2.24479940e-02 1.24008203e+00
9.65351641e-01 -3.43715608e-01 9.36967075e-01 1.42941904e+00
-1.95776701e-01 -5.99102080e-01 -4.55216527e-01 -4.34658796e-01
3.23334366e-01 2.43003909e-02 4.42480445e-01 6.85610712e-01
1.05593359e+00 1.24305594e+00 -8.24684441e-01 -6.83290601e-01
4.25141305e-03 -8.89702961e-02 1.42646289e+00 -1.16354311e+00
-4.47258860e-01 -4.85934347e-01 4.35969472e-01 -1.94052649e+00
2.42006764e-01 -5.34712136e-01 6.39628172e-01 -1.25143981e+00
2.32000276e-01 -4.22635257e-01 4.02678281e-01 1.15525201e-01
-4.84796107e-01 -2.48437509e-01 4.33915883e-01 3.55006248e-01
-8.23281109e-01 7.90810406e-01 1.35514927e+00 -1.34846106e-01
-7.97492743e-01 3.20169300e-01 -7.95234859e-01 5.75254261e-01
7.78582215e-01 -1.99208558e-01 -4.93508250e-01 -3.67777199e-01
6.28015101e-02 6.98562860e-01 -2.73661818e-02 -4.16053981e-01
3.02696526e-01 -4.90889549e-01 -1.00383215e-01 -3.86459440e-01
7.41262674e-01 -1.44605607e-01 -5.49977422e-01 7.75895789e-02
-1.28367841e+00 1.17580593e-01 -3.48245353e-01 5.33100605e-01
-1.06824808e-01 -8.97510350e-02 4.84429210e-01 -3.43233496e-01
-3.28410596e-01 1.31035224e-01 -7.47930169e-01 3.93128723e-01
5.91311038e-01 1.04746364e-01 -8.32216144e-01 -1.31542778e+00
-3.54751259e-01 7.09039927e-01 3.14856499e-01 8.02706778e-01
3.18037331e-01 -1.13503277e+00 -8.10840011e-01 -4.54173088e-01
1.78142861e-01 -2.40965262e-01 -1.25522101e-02 3.98984909e-01
-7.06486702e-02 5.87771297e-01 -1.43951857e-02 -3.94327521e-01
-1.28471959e+00 4.63897549e-02 2.90806502e-01 -7.49691010e-01
-1.26581550e-01 1.02980840e+00 3.58711660e-01 -1.18545413e+00
3.42334032e-01 -2.47886643e-01 -2.39575639e-01 -6.10183328e-02
6.31087244e-01 2.75800467e-01 -2.69223571e-01 -2.94974446e-01
-3.69332582e-02 -4.90954667e-01 -2.96024799e-01 -8.07031870e-01
9.73936558e-01 -1.47479564e-01 9.69094634e-02 5.37968457e-01
1.06161177e+00 -4.78362255e-02 -1.64793611e+00 -3.07672083e-01
1.62379920e-01 -3.77576947e-01 -7.90119350e-01 -1.06189513e+00
-2.81241000e-01 7.23746836e-01 -9.34180394e-02 5.08766830e-01
2.17164472e-01 3.56806107e-02 1.01349032e+00 7.58650422e-01
3.65938455e-01 -1.23565698e+00 8.92828524e-01 1.15533161e+00
1.05716097e+00 -1.75191998e+00 -4.70138431e-01 -2.34934047e-01
-1.45828807e+00 8.94289672e-01 1.27717686e+00 -6.77171489e-03
9.03429613e-02 -1.45742178e-01 6.81511581e-01 -4.66545857e-02
-1.62153816e+00 2.09123880e-01 1.37438849e-01 5.43779969e-01
9.97257411e-01 3.20660591e-01 -3.62679839e-01 6.79779351e-01
-7.81377256e-01 -3.90622675e-01 5.21451771e-01 4.81027186e-01
-7.11557448e-01 -1.25999641e+00 -1.62080333e-01 4.49228823e-01
-6.18175566e-02 -5.05497344e-02 -1.32955456e+00 2.30061650e-01
-7.22544372e-01 1.85823548e+00 -7.66261593e-02 -8.60614538e-01
3.54207516e-01 1.56833440e-01 -6.29669726e-02 -7.83072233e-01
-1.30221331e+00 -3.38506341e-01 9.27072883e-01 -5.67969978e-01
-1.54536873e-01 -3.93504828e-01 -8.92617345e-01 -7.53090978e-01
-2.75166035e-01 4.32550430e-01 1.39938712e-01 8.59286368e-01
2.69212008e-01 1.10836141e-01 1.02409244e+00 -9.85508084e-01
-1.12245190e+00 -1.67346501e+00 -1.95083637e-02 7.65289068e-01
1.98546529e-01 -2.86677033e-01 -3.43233168e-01 -1.97838753e-01] | [12.76606273651123, 8.13489818572998] |
a1f9303c-0d23-4350-9ce1-f98b7926ac0d | smoothed-separable-nonnegative-matrix | 2110.05528 | null | https://arxiv.org/abs/2110.05528v2 | https://arxiv.org/pdf/2110.05528v2.pdf | Smoothed Separable Nonnegative Matrix Factorization | Given a set of data points belonging to the convex hull of a set of vertices, a key problem in linear algebra, signal processing, data analysis and machine learning is to estimate these vertices in the presence of noise. Many algorithms have been developed under the assumption that there is at least one nearby data point to each vertex; two of the most widely used ones are vertex component analysis (VCA) and the successive projection algorithm (SPA). This assumption is known as the pure-pixel assumption in blind hyperspectral unmixing, and as the separability assumption in nonnegative matrix factorization. More recently, Bhattacharyya and Kannan (ACM-SIAM Symposium on Discrete Algorithms, 2020) proposed an algorithm for learning a latent simplex (ALLS) that relies on the assumption that there is more than one nearby data point to each vertex. In that scenario, ALLS is probalistically more robust to noise than algorithms based on the separability assumption. In this paper, inspired by ALLS, we propose smoothed VCA (SVCA) and smoothed SPA (SSPA) that generalize VCA and SPA by assuming the presence of several nearby data points to each vertex. We illustrate the effectiveness of SVCA and SSPA over VCA, SPA and ALLS on synthetic data sets, on the unmixing of hyperspectral images, and on feature extraction on facial images data sets. In addition, our study highlights new theoretical results for VCA. | ['Christophe Kervazo', 'Nicolas Gillis', 'Nicolas Nadisic'] | 2021-10-11 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 5.83594978e-01 -2.83567220e-01 8.91582072e-02 5.68487011e-02
-4.07170504e-01 -4.98026907e-01 5.11405110e-01 1.65091828e-01
-5.38642555e-02 5.47135949e-01 2.27548666e-02 -1.67814746e-01
-6.86555207e-01 -6.93857491e-01 -6.57848120e-01 -1.18273604e+00
-2.89994068e-02 4.86670911e-01 -2.37794518e-01 -8.50145295e-02
-1.15383960e-01 8.40917945e-01 -1.58360684e+00 -1.58279374e-01
9.52822208e-01 6.58731461e-01 1.33126721e-01 3.54569227e-01
-2.84225672e-01 2.41614059e-02 -2.12519780e-01 -4.74246703e-02
5.23728549e-01 -3.01029950e-01 -4.42343205e-01 6.12929523e-01
3.08993071e-01 3.47062439e-01 6.48138821e-02 1.29987550e+00
1.74404606e-01 1.84620380e-01 8.60430419e-01 -1.61228859e+00
-2.88987994e-01 4.52587157e-01 -1.25492918e+00 -1.59756213e-01
-1.05883684e-02 -2.66175359e-01 7.46313155e-01 -7.80054569e-01
3.41520935e-01 9.84024704e-01 5.61237216e-01 1.85084492e-01
-1.49620605e+00 -5.02458930e-01 -1.11231795e-02 1.24390543e-01
-1.60247386e+00 -3.87893677e-01 9.52615619e-01 -6.22094631e-01
2.01952130e-01 6.73061430e-01 5.14676392e-01 4.94732767e-01
-3.15653861e-01 5.79135060e-01 1.24203587e+00 -7.57673979e-01
3.94713014e-01 -1.64457068e-01 3.43352914e-01 4.47266251e-01
4.87201691e-01 -8.42682347e-02 -3.20514381e-01 -6.11853898e-01
5.10343254e-01 3.26597765e-02 -5.78682780e-01 -8.04469764e-01
-9.39538658e-01 8.82531941e-01 1.93768725e-01 1.93588108e-01
-5.79809487e-01 -3.86582345e-01 -2.82247504e-03 7.07213804e-02
3.86308819e-01 9.97894779e-02 8.10244903e-02 4.44124997e-01
-1.16950321e+00 -2.18986138e-03 6.35687947e-01 6.29174829e-01
8.95662248e-01 5.71211874e-02 7.53010288e-02 7.32923150e-01
4.51572627e-01 8.07992101e-01 3.24056029e-01 -7.55157590e-01
3.80098492e-01 5.83835602e-01 1.40840709e-01 -1.21090436e+00
-4.75390613e-01 -4.91840065e-01 -1.14470899e+00 3.85668576e-01
4.54669505e-01 1.34800263e-02 -8.24771702e-01 1.69647098e+00
3.55363131e-01 7.30175614e-01 2.62822568e-01 7.82291710e-01
4.42815483e-01 7.58039892e-01 -3.88351798e-01 -9.27511632e-01
8.81936491e-01 -3.64274681e-01 -7.25922942e-01 4.99830991e-02
1.57826826e-01 -7.72234321e-01 7.15225160e-01 7.40393817e-01
-7.15887368e-01 -2.58954972e-01 -9.96555626e-01 5.21017432e-01
-1.44468933e-01 3.46375674e-01 5.06946087e-01 9.22403216e-01
-8.18084300e-01 4.50938910e-01 -8.83187175e-01 -4.10355300e-01
1.66335851e-01 4.11149174e-01 -4.38144296e-01 -1.56194702e-01
-5.23763061e-01 2.48530373e-01 9.29666311e-02 4.01343018e-01
-3.95884275e-01 -4.49542016e-01 -5.91429234e-01 -6.47746325e-02
4.59963322e-01 -3.05449367e-01 4.41146940e-01 -1.25035131e+00
-1.27024996e+00 5.65691233e-01 -4.48069006e-01 -3.29086810e-01
3.35517079e-01 1.95886821e-01 -5.77138066e-01 1.22450151e-01
2.90045068e-02 8.62375554e-03 1.14675391e+00 -1.46811736e+00
-2.12646827e-01 -7.83562005e-01 -3.66767198e-01 8.42237771e-02
-3.15474540e-01 -8.00920278e-02 -2.26781756e-01 -4.01704997e-01
8.67019594e-01 -1.05519032e+00 -1.77912712e-01 -1.42691851e-01
-5.49908459e-01 -8.06114823e-02 8.43172729e-01 -6.93632364e-01
9.71872509e-01 -2.19096017e+00 5.52941620e-01 8.69787991e-01
1.22639515e-01 3.75565618e-01 -2.82696337e-01 4.29718167e-01
-5.11292458e-01 -2.11965457e-01 -7.63566077e-01 -2.50155807e-01
-2.79907733e-01 1.86208054e-01 -3.55839521e-01 8.80917311e-01
-2.10712686e-01 2.13684127e-01 -8.63181233e-01 -4.98116501e-02
3.31614673e-01 3.66356581e-01 -2.00317428e-01 -3.39401335e-01
-6.11580163e-02 2.85576463e-01 3.11084092e-02 5.69585383e-01
1.09792078e+00 -3.40445675e-02 2.88230002e-01 -3.79101157e-01
-3.45160127e-01 -6.42403007e-01 -1.95588028e+00 1.22492027e+00
5.52242137e-02 4.17092443e-01 5.46529472e-01 -1.19821787e+00
9.16162908e-01 4.87148494e-01 8.64884138e-01 -8.19397643e-02
9.46510956e-02 2.05277324e-01 -2.01729655e-01 -3.32887262e-01
2.03769207e-01 -2.64753789e-01 5.44618249e-01 2.84242600e-01
-2.53477722e-01 9.00096297e-02 3.67628932e-01 1.61957309e-01
6.08593047e-01 -2.96744615e-01 2.58432627e-01 -3.40434521e-01
7.56677210e-01 -1.12462230e-01 6.91259563e-01 4.05923218e-01
-6.58424795e-02 7.54044771e-01 4.73683804e-01 1.96088403e-01
-8.44481707e-01 -1.06021547e+00 -3.20604235e-01 3.75501424e-01
8.53535384e-02 -2.37124607e-01 -6.38130069e-01 -1.96006224e-01
-6.84926808e-02 4.86555815e-01 -4.28278565e-01 2.33014822e-01
-5.32528572e-03 -1.18527532e+00 1.05784811e-01 3.54401544e-02
4.16908562e-01 -3.07484597e-01 1.21464565e-01 -4.62617725e-02
-2.24035561e-01 -9.25158918e-01 -1.14105456e-01 2.64138132e-02
-7.84205496e-01 -1.12954843e+00 -6.61298275e-01 -5.62581241e-01
8.73818815e-01 5.49998760e-01 3.87878954e-01 -4.11244363e-01
-4.73588146e-03 4.31138933e-01 -3.06193054e-01 -3.62682283e-01
-3.19451630e-01 -4.72784370e-01 3.87249947e-01 7.48972535e-01
1.66948617e-01 -5.94972670e-01 5.12421466e-02 2.57637620e-01
-1.00550926e+00 7.73231760e-02 3.49450380e-01 4.84858304e-01
7.95904219e-01 5.91353834e-01 2.36158237e-01 -6.97142005e-01
3.84107113e-01 -4.93275166e-01 -9.63697970e-01 4.27358747e-01
-4.05190706e-01 -2.56041326e-02 4.91865605e-01 -2.15224266e-01
-6.47623181e-01 5.38012207e-01 2.13438809e-01 -7.51086533e-01
-6.98005706e-02 7.20916033e-01 -5.38949728e-01 -3.40459168e-01
7.08376944e-01 2.73071349e-01 2.41237834e-01 -5.18929362e-01
2.90749341e-01 6.96597338e-01 3.88722569e-01 -3.98697317e-01
1.04530823e+00 8.94958675e-01 5.54646730e-01 -1.52924538e+00
-3.97061706e-01 -8.99352610e-01 -6.62964582e-01 -2.35662222e-01
6.79405868e-01 -7.24666297e-01 -4.70427573e-01 6.34234428e-01
-7.09218323e-01 1.41592562e-01 -1.91349074e-01 6.13045454e-01
-3.66899788e-01 8.59097838e-01 8.54594484e-02 -1.04914236e+00
-3.20320874e-02 -9.78838801e-01 5.23657262e-01 9.16643068e-02
2.40806431e-01 -9.07626390e-01 1.10175051e-01 4.55266863e-01
-8.85962918e-02 5.12912989e-01 7.84669042e-01 -3.16164941e-01
-2.03009710e-01 -1.99249893e-01 -4.71741632e-02 5.08089483e-01
2.80827016e-01 2.34051257e-01 -8.24889839e-01 -4.62376595e-01
1.41083568e-01 3.39822918e-01 7.62048841e-01 7.47999609e-01
8.32540691e-01 -2.23685935e-01 -3.84106606e-01 6.94870889e-01
1.58636248e+00 1.14490211e-01 4.71871704e-01 -1.62846193e-01
7.47778773e-01 5.29202878e-01 2.36002758e-01 3.07276070e-01
-6.56418875e-02 5.89309096e-01 6.07695341e-01 -1.00880265e-01
2.34523267e-01 2.29227692e-01 2.92056590e-01 6.80851400e-01
-3.83443862e-01 -1.97760299e-01 -8.00467968e-01 4.92925912e-01
-1.81568730e+00 -9.50430155e-01 -7.80527830e-01 2.64616394e+00
2.47968093e-01 -3.08233529e-01 2.32722446e-01 6.89360082e-01
9.63431478e-01 1.15604334e-01 -2.01173484e-01 -8.70314986e-02
-5.41808426e-01 1.20127767e-01 7.47631013e-01 7.33288825e-01
-1.09673429e+00 3.57414693e-01 4.67996454e+00 7.22513080e-01
-1.03087020e+00 -1.23939114e-02 1.20702714e-01 7.09018260e-02
-3.08049828e-01 1.27328381e-01 -5.25402367e-01 2.93745577e-01
4.99091864e-01 -7.77495429e-02 7.50251770e-01 3.59664977e-01
3.47595513e-01 -2.02245027e-01 -5.35546541e-01 1.01310122e+00
3.35152358e-01 -1.00143015e+00 5.72858974e-02 2.80278563e-01
1.11084080e+00 -1.87679276e-01 1.69702426e-01 -3.36770177e-01
-8.68732482e-03 -7.83332825e-01 3.11632723e-01 3.75283569e-01
4.06553566e-01 -7.90788054e-01 4.44491953e-01 4.03641045e-01
-1.04755092e+00 -1.27841517e-01 -4.22589988e-01 3.37051563e-02
2.58523654e-02 9.28828061e-01 -5.28157115e-01 8.92870903e-01
1.53561354e-01 6.04635000e-01 -3.19767684e-01 1.29589415e+00
-1.78574726e-01 7.30569422e-01 -6.67662561e-01 2.87843108e-01
3.21069092e-01 -1.00245392e+00 9.89391267e-01 5.44448018e-01
5.50115228e-01 2.91008890e-01 2.32850924e-01 6.91606343e-01
1.60702407e-01 3.72241586e-01 -3.18783730e-01 -5.94731495e-02
2.56107867e-01 1.27509928e+00 -9.46302831e-01 -4.62835729e-02
-5.19648790e-01 8.62342119e-01 -1.84522435e-01 6.70767486e-01
-5.10471880e-01 1.00328075e-02 6.36049509e-01 3.08453619e-01
1.72273248e-01 -3.27511102e-01 -3.72295767e-01 -9.55909431e-01
-2.98349969e-02 -7.85004675e-01 4.60555673e-01 -6.67286873e-01
-1.14253259e+00 2.12143779e-01 6.55634748e-03 -1.27529311e+00
2.98855335e-01 -7.28749394e-01 -5.28939247e-01 1.08656561e+00
-1.34424186e+00 -1.09572303e+00 -3.41117978e-01 8.40668321e-01
-1.05011021e-03 -2.30320901e-01 6.60753131e-01 5.13317697e-02
-7.30629325e-01 -8.66709203e-02 7.80346572e-01 -3.69154140e-02
3.22171599e-01 -9.57106173e-01 -2.95634747e-01 1.22387505e+00
5.34167290e-01 4.94708657e-01 7.70859301e-01 -6.70179784e-01
-1.31782532e+00 -9.71903980e-01 5.53955078e-01 -2.65673045e-02
5.87642133e-01 -1.60703078e-01 -8.69227827e-01 6.98569775e-01
-7.47626498e-02 -2.56013330e-02 7.49862254e-01 1.38592094e-01
-3.24048512e-02 -2.51500934e-01 -1.05791020e+00 4.32315797e-01
5.16170919e-01 -3.76226693e-01 -2.53759146e-01 7.84570813e-01
1.36191901e-02 -1.23820342e-01 -5.68089783e-01 3.72762948e-01
1.83181286e-01 -9.78190482e-01 1.01703167e+00 -2.98404098e-01
-7.40541741e-02 -8.04225385e-01 -3.23687255e-01 -1.25887787e+00
-2.94647574e-01 -5.52282035e-01 1.97878808e-01 1.08213723e+00
1.70483664e-01 -6.21864855e-01 7.33269751e-01 3.28360051e-01
6.91322014e-02 -1.94007695e-01 -1.07162809e+00 -9.85202789e-01
-1.26341939e-01 -4.43950891e-01 4.77884263e-01 1.19573736e+00
-1.52210951e-01 1.45763963e-01 -4.29574639e-01 8.88679087e-01
1.05193949e+00 2.76649684e-01 5.28496206e-01 -1.69313753e+00
-3.94098550e-01 -2.82958269e-01 -4.13491756e-01 -3.90478045e-01
2.03052402e-01 -9.73036528e-01 -1.92353562e-01 -1.61179304e+00
2.72352453e-02 -4.34154540e-01 -1.41187221e-01 2.59567469e-01
6.43288568e-02 3.55861574e-01 1.97593346e-01 2.31082201e-01
1.04869194e-01 2.72047490e-01 8.51844668e-01 -1.06892407e-01
-5.20290434e-01 3.74492139e-01 -3.59280050e-01 8.81129801e-01
5.16410708e-01 -2.57367998e-01 -4.36261922e-01 -2.17494383e-01
2.76945978e-01 1.21255316e-01 4.24296230e-01 -1.12461245e+00
3.15786541e-01 -2.38991901e-01 1.24953359e-01 -5.36997497e-01
5.67552626e-01 -1.06349182e+00 6.61637843e-01 3.08765948e-01
1.63672939e-01 -4.40165699e-01 7.57293925e-02 7.71536529e-01
-7.12773725e-02 -3.91421884e-01 8.26765180e-01 8.45989212e-02
-6.30973101e-01 2.43782461e-01 -2.73684680e-01 -4.47166622e-01
1.23026800e+00 -3.03146571e-01 -1.14855982e-01 -3.69751215e-01
-1.04658806e+00 5.69213592e-02 3.33883047e-01 -2.82311849e-02
5.30651867e-01 -1.09089029e+00 -8.83245349e-01 6.20042503e-01
2.21731383e-02 -1.67122230e-01 3.15801680e-01 1.11696076e+00
-4.21235472e-01 1.99004501e-01 2.75772344e-02 -7.06924677e-01
-1.71526587e+00 6.62289143e-01 1.72715396e-01 3.09560120e-01
-1.86528683e-01 7.52196133e-01 -9.72430557e-02 -1.56262681e-01
2.85036899e-02 -1.21003553e-01 -2.61073589e-01 3.54237765e-01
3.94932836e-01 7.35923231e-01 9.24180523e-02 -1.03085208e+00
-1.37330741e-01 7.46274710e-01 4.69320923e-01 -2.20888928e-01
1.25742543e+00 -4.20398787e-02 -5.69186747e-01 4.34639871e-01
1.01315796e+00 4.49434280e-01 -8.66006076e-01 -4.46178883e-01
-2.27119088e-01 -4.28516656e-01 3.96294802e-01 -4.31006968e-01
-1.02338612e+00 7.64142513e-01 6.71514750e-01 4.08537924e-01
1.34835100e+00 -3.65631312e-01 7.05197155e-02 1.60298124e-01
1.57083675e-01 -8.92930150e-01 -4.87003863e-01 1.09373860e-01
8.20385337e-01 -9.65366364e-01 1.49853036e-01 -8.06312382e-01
-4.34856772e-01 1.16922033e+00 -1.47012044e-02 3.19916345e-02
8.87652457e-01 5.46504781e-02 -9.87016186e-02 -1.84001215e-02
-2.08488647e-02 -4.12275761e-01 2.28391692e-01 4.89177108e-01
-6.02102317e-02 3.28437418e-01 -1.62970781e-01 2.92688400e-01
1.04814835e-01 -1.04510456e-01 5.40804923e-01 6.19018734e-01
-4.44569618e-01 -9.35627759e-01 -1.03136504e+00 4.92657274e-01
1.36894539e-01 -6.01985492e-02 -2.73904949e-01 4.28407729e-01
2.83703625e-01 1.08691204e+00 -7.06097856e-02 -8.59870389e-02
1.19997963e-01 1.55597880e-01 5.25893986e-01 -4.88501281e-01
4.50431183e-02 3.21077347e-01 -2.46444494e-01 -1.67599142e-01
-7.90168524e-01 -1.05415761e+00 -1.11529171e+00 -3.58328193e-01
-5.97858429e-01 4.25454676e-01 8.62938702e-01 9.70851481e-01
3.61547656e-02 7.63987526e-02 7.39171624e-01 -5.02703249e-01
-3.55802387e-01 -6.68382943e-01 -1.23074865e+00 1.77215472e-01
2.99186110e-01 -5.08236408e-01 -6.71127975e-01 -2.82462295e-02] | [10.074056625366211, -1.9937150478363037] |
998ac095-3cb3-431b-b241-1e2d0010acec | enhanced-word-representations-for-bridging | 1803.04790 | null | http://arxiv.org/abs/1803.04790v2 | http://arxiv.org/pdf/1803.04790v2.pdf | Enhanced Word Representations for Bridging Anaphora Resolution | Most current models of word representations(e.g.,GloVe) have successfully
captured fine-grained semantics. However, semantic similarity exhibited in
these word embeddings is not suitable for resolving bridging anaphora, which
requires the knowledge of associative similarity (i.e., relatedness) instead of
semantic similarity information between synonyms or hypernyms. We create word
embeddings (embeddings_PP) to capture such relatedness by exploring the
syntactic structure of noun phrases. We demonstrate that using embeddings_PP
alone achieves around 30% of accuracy for bridging anaphora resolution on the
ISNotes corpus. Furthermore, we achieve a substantial gain over the
state-of-the-art system (Hou et al., 2013) for bridging antecedent selection. | ['Yufang Hou'] | 2018-03-13 | enhanced-word-representations-for-bridging-1 | https://aclanthology.org/N18-2001 | https://aclanthology.org/N18-2001.pdf | naacl-2018-6 | ['bridging-anaphora-resolution'] | ['natural-language-processing'] | [-2.26862758e-01 3.32234800e-01 -5.85192740e-01 -2.52511829e-01
-6.51012719e-01 -4.66843873e-01 5.56720197e-01 6.09783888e-01
-7.01819777e-01 6.65957451e-01 9.60005581e-01 7.09828921e-03
-3.71835172e-01 -9.10824776e-01 -2.50235587e-01 -1.45611599e-01
6.35765716e-02 7.85784543e-01 1.60250083e-01 -7.14730501e-01
2.51095712e-01 5.28702676e-01 -1.44877148e+00 2.14964136e-01
6.80056274e-01 4.45146233e-01 3.37640867e-02 -2.28721723e-02
-5.79595983e-01 4.92615163e-01 -7.09008634e-01 -6.34156108e-01
-2.61868834e-01 -5.51565662e-02 -1.21395195e+00 -7.19119072e-01
3.93049747e-01 -1.40405700e-01 -8.51112843e-01 1.04561806e+00
3.46988380e-01 3.97203326e-01 7.01222658e-01 -1.09709620e+00
-1.24372423e+00 1.00934708e+00 -2.57306665e-01 8.34188640e-01
6.59713984e-01 -5.12744486e-01 1.75656021e+00 -1.22988868e+00
9.77890730e-01 1.51009858e+00 6.84502840e-01 6.93502307e-01
-1.08446789e+00 -6.32592440e-01 -2.94334814e-02 7.49606371e-01
-1.44308782e+00 -2.12450236e-01 6.58342898e-01 -2.75014192e-02
1.68967056e+00 -1.24649152e-01 3.88595998e-01 1.43148935e+00
5.55225797e-02 6.03671193e-01 5.18428445e-01 -5.38557351e-01
-3.38792913e-02 -3.57283980e-01 9.07791555e-01 2.02934757e-01
6.50429010e-01 2.50669885e-02 -6.46122396e-01 -4.88511473e-01
5.26596069e-01 -4.23850901e-02 -3.99158388e-01 -4.06031817e-01
-9.42632556e-01 1.26198030e+00 6.79824650e-01 6.40826941e-01
-3.85245383e-01 2.52063006e-01 5.94382465e-01 1.49993047e-01
6.30244389e-02 1.00044835e+00 -4.23358649e-01 -3.08480382e-01
-2.13546753e-01 5.94069958e-01 7.48723626e-01 9.69782770e-01
5.59948742e-01 -2.74852574e-01 -2.71764472e-02 1.03037333e+00
1.98672205e-01 1.13090642e-01 8.39729130e-01 -9.18820977e-01
5.97377062e-01 7.00507343e-01 2.08474889e-01 -1.04566681e+00
-5.64925611e-01 1.55449614e-01 -3.69603513e-03 -3.71528387e-01
2.80152231e-01 1.75981849e-01 -3.89443278e-01 2.22167516e+00
1.41148046e-01 1.64940014e-01 5.19787550e-01 1.07607210e+00
1.17650390e+00 4.32916641e-01 4.05633509e-01 -1.24737397e-01
1.86932313e+00 -6.39995575e-01 -1.25743186e+00 -6.07546389e-01
7.63666332e-01 -4.75580931e-01 1.08755291e+00 -3.29369366e-01
-1.05914414e+00 -1.53449684e-01 -1.22094417e+00 -4.98167366e-01
-6.06495678e-01 -6.78348958e-01 9.40713882e-01 1.45111397e-01
-4.62512940e-01 6.31371379e-01 -6.51353061e-01 -8.68692875e-01
2.62985528e-01 2.45936096e-01 -7.31284320e-01 -1.92883372e-01
-2.07477784e+00 1.65187764e+00 6.40762746e-01 -4.78693366e-01
-3.47753435e-01 -8.15738082e-01 -1.34374118e+00 3.76221627e-01
1.74640805e-01 -4.33100909e-01 9.99563456e-01 8.18875730e-02
-7.50070751e-01 1.16538107e+00 -2.23949566e-01 -4.85913366e-01
-4.39686298e-01 -7.34081984e-01 -6.26607299e-01 2.19362125e-01
3.43993008e-01 6.93111837e-01 2.60843098e-01 -9.81717587e-01
-3.15178782e-01 -4.27139163e-01 3.81636739e-01 3.57699245e-01
-8.50707412e-01 4.53957677e-01 -7.00570131e-03 -8.05682242e-01
2.10242480e-01 -6.03385150e-01 1.37965307e-01 -1.02016069e-01
7.10038915e-02 -9.95309711e-01 7.69903898e-01 -4.52285558e-01
1.54670835e+00 -2.26414657e+00 4.67780858e-01 -3.38760227e-01
1.89210236e-01 3.44605416e-01 -4.20507133e-01 7.76993096e-01
-3.08425218e-01 2.21263394e-01 -2.11600170e-01 7.60031864e-02
3.29697222e-01 6.57944262e-01 -6.13802433e-01 2.96248734e-01
3.86186957e-01 1.03310895e+00 -1.11298919e+00 -5.48447669e-01
2.24597640e-02 4.10319239e-01 -6.95925176e-01 2.93303221e-01
8.71775821e-02 -4.29306716e-01 -4.47041959e-01 5.45713365e-01
3.23781729e-01 5.53465225e-02 5.65930068e-01 -4.86152440e-01
3.24696600e-01 9.75129604e-01 -8.87982666e-01 2.05942464e+00
-5.26811123e-01 5.38695455e-01 -3.53018731e-01 -1.03712010e+00
1.09469581e+00 6.65238678e-01 1.51305616e-01 -5.58157682e-01
3.81229639e-01 2.85501152e-01 1.95630118e-01 -2.96368659e-01
6.84024096e-01 -3.26145411e-01 -7.82552958e-01 7.39612281e-01
3.86686027e-01 -7.39579946e-02 1.82693258e-01 3.96897376e-01
1.28545511e+00 3.10025439e-02 6.12295806e-01 -4.19580132e-01
2.27368683e-01 2.33137116e-01 7.60989606e-01 3.77637148e-01
-3.70148718e-01 4.16412294e-01 3.86625826e-01 -4.13102925e-01
-8.50764632e-01 -1.57505929e+00 -5.23616374e-01 1.11927032e+00
5.18951893e-01 -8.45711648e-01 -6.37673914e-01 -5.23913682e-01
3.51933867e-01 9.87719536e-01 -6.96926236e-01 -6.26316547e-01
-1.00389969e+00 -6.88267350e-01 8.59010935e-01 1.15217698e+00
-1.19568653e-01 -1.38269067e+00 -6.69298708e-01 4.79081631e-01
-2.27478430e-01 -1.23253763e+00 -2.16490731e-01 4.68784422e-01
-7.24234343e-01 -1.17925322e+00 1.46359161e-01 -9.51400042e-01
1.11301370e-01 8.70674402e-02 1.33617020e+00 1.35275751e-01
-4.31921571e-01 1.61714897e-01 -4.45779324e-01 -6.53324798e-02
9.72172618e-02 -6.97753802e-02 5.02626240e-01 -8.02385032e-01
1.38063633e+00 -8.19231451e-01 -9.69177037e-02 1.96824577e-02
-8.12147796e-01 -9.14615929e-01 1.63935333e-01 1.11790371e+00
4.09859300e-01 -4.60984498e-01 8.79368126e-01 -8.14193249e-01
1.13570082e+00 -6.62267685e-01 1.25020579e-01 1.31043047e-01
-5.55671275e-01 1.36866912e-01 2.80427873e-01 -6.17905676e-01
-9.90862548e-01 -6.96899533e-01 7.01603591e-02 -5.91587484e-01
-8.21972042e-02 5.41562498e-01 -2.06882000e-01 4.23519850e-01
7.38291204e-01 -4.61082190e-01 -3.87333870e-01 -6.51454926e-01
9.10311937e-01 6.90305293e-01 1.00381887e+00 -1.01208222e+00
5.38130403e-01 1.23052597e-01 -4.37079012e-01 -3.59647065e-01
-1.23703742e+00 -5.24860978e-01 -7.32508838e-01 6.46536648e-01
1.14414263e+00 -6.83124721e-01 -2.62852877e-01 -3.91155064e-01
-1.77052772e+00 4.45617765e-01 -4.10872340e-01 5.87034523e-01
-5.01010060e-01 2.61913627e-01 -1.09293401e+00 -4.87992130e-02
-2.23677441e-01 -6.47353113e-01 6.48724139e-01 1.35393366e-01
-1.16212761e+00 -1.15872884e+00 2.29221672e-01 3.31003100e-01
3.69648874e-01 7.01103359e-02 1.49141920e+00 -1.23397422e+00
6.16956539e-02 -3.06323767e-01 -4.52993810e-01 -2.91770786e-01
3.32659304e-01 -5.66268086e-01 -9.51940060e-01 -1.46287195e-02
-8.47913474e-02 -3.98254842e-01 7.24469364e-01 -9.52616856e-02
6.43346012e-01 -5.59793003e-02 -6.26883805e-01 3.98399979e-01
1.10429168e+00 -9.22298525e-03 5.95695138e-01 5.78618169e-01
6.00490093e-01 5.11936247e-01 8.87728214e-01 3.48924071e-01
3.03761929e-01 8.74066830e-01 3.15415233e-01 5.91123998e-01
-3.73830795e-01 -2.93863297e-01 9.20213014e-02 7.06002235e-01
1.84603527e-01 -1.08508475e-01 -9.63566303e-01 9.97562587e-01
-1.86023974e+00 -9.70109344e-01 -2.38047112e-02 1.61120057e+00
1.20595694e+00 3.62136140e-02 -5.59274554e-01 2.04700723e-01
8.87982249e-01 3.01668495e-01 -2.25021258e-01 -9.77842093e-01
-2.00218379e-01 8.02100062e-01 4.61560761e-04 6.13125980e-01
-1.01755190e+00 1.52347863e+00 6.25567102e+00 5.76936781e-01
-2.42941484e-01 4.33076024e-01 -6.35312736e-01 1.12319604e-01
-5.72428882e-01 1.82981998e-01 -8.38735104e-01 8.99947882e-02
7.65782535e-01 -3.26007158e-01 2.48905778e-01 6.29523635e-01
-6.63488805e-01 4.23758090e-01 -1.38154852e+00 7.61588395e-01
4.32343543e-01 -1.37671888e+00 2.95998633e-01 -3.04898411e-01
3.69159102e-01 -2.28307277e-01 -4.36328292e-01 5.11403203e-01
4.53548819e-01 -1.13718700e+00 2.43230045e-01 9.38166827e-02
5.91196954e-01 -7.65637636e-01 1.07181752e+00 -2.04424143e-01
-1.04972422e+00 -1.52269378e-01 -7.40732610e-01 -1.95961252e-01
4.23102915e-01 1.15564100e-01 -4.50247437e-01 3.28154355e-01
6.31752253e-01 7.98002779e-01 -2.79043913e-01 3.21978986e-01
-9.81053591e-01 1.89204186e-01 -1.16643727e-01 7.99068138e-02
2.41606995e-01 1.88809291e-01 5.62683284e-01 1.29112184e+00
3.20361219e-02 3.86393905e-01 -9.23784450e-03 9.57748592e-01
-2.10557312e-01 -1.08868532e-01 -6.02206469e-01 -2.69394726e-01
1.45815146e+00 9.67527330e-01 -7.17450753e-02 -2.81376392e-01
-5.11103272e-01 8.62177193e-01 9.03623939e-01 9.55624133e-03
-7.16069221e-01 -7.09079623e-01 1.45026362e+00 -3.08142155e-01
3.38317156e-01 -2.25239769e-02 -2.37381443e-01 -1.00434041e+00
-2.62449265e-01 -6.13296449e-01 1.00996077e+00 -7.79989541e-01
-1.81870043e+00 4.62274998e-01 1.73916131e-01 -7.33274341e-01
-3.36569875e-01 -7.54352629e-01 -7.17668355e-01 9.06590402e-01
-1.30301023e+00 -1.03919995e+00 -5.55303954e-02 4.71088499e-01
3.46246839e-01 -6.23046681e-02 1.62415493e+00 1.94492459e-01
-3.31871271e-01 6.61416709e-01 -5.56500733e-01 3.03361893e-01
9.42504883e-01 -1.15009308e+00 5.13735652e-01 4.01664227e-01
3.17890912e-01 1.22893798e+00 7.20822752e-01 -6.09494865e-01
-1.46145797e+00 -5.96559227e-01 1.71885443e+00 -8.25860858e-01
1.21305120e+00 -1.60259102e-02 -1.36546111e+00 9.15415823e-01
4.06781495e-01 2.17452794e-01 9.00526702e-01 6.40475869e-01
-1.11615646e+00 3.03037912e-01 -1.18253720e+00 6.93853855e-01
1.48732460e+00 -8.92584264e-01 -2.08505344e+00 7.29155540e-03
1.14416027e+00 -3.15905422e-01 -1.30221903e+00 5.83594918e-01
1.96429506e-01 -2.99207032e-01 1.35941839e+00 -1.49016428e+00
4.71819669e-01 -2.54188441e-02 -7.60116875e-01 -1.25602913e+00
-5.54821849e-01 -1.20083980e-01 -5.66861153e-01 1.28937840e+00
3.15105468e-01 -3.97202909e-01 3.21145326e-01 5.58005273e-01
-3.05976629e-01 -5.66324174e-01 -1.07784295e+00 -8.91761601e-01
4.51293498e-01 -3.50464970e-01 8.77736330e-01 1.50871789e+00
9.82742131e-01 5.91008902e-01 2.69766420e-01 3.26521724e-01
4.54753786e-01 4.34551477e-01 9.61909443e-02 -1.53881514e+00
1.99351579e-01 -4.36561674e-01 -7.59288788e-01 -5.71446180e-01
1.07309079e+00 -1.13560939e+00 -3.06951374e-01 -1.67759836e+00
2.02670798e-01 -2.77153283e-01 -7.49959469e-01 5.93739748e-01
-3.88811350e-01 2.99999654e-01 1.14680171e-01 1.13218009e-01
-3.14043850e-01 6.23178184e-01 7.60406494e-01 -7.87273645e-02
1.81391329e-01 -1.08324623e+00 -8.60424459e-01 7.62619555e-01
7.98339128e-01 -7.03625262e-01 -1.14425585e-01 -8.01421940e-01
7.84248635e-02 -2.76244599e-02 2.25937322e-01 -4.41375434e-01
1.43659994e-01 -2.38321081e-01 -3.49967517e-02 -6.98903427e-02
6.34593248e-01 -5.75847983e-01 -4.31980520e-01 2.34375402e-01
-5.09943426e-01 3.63113344e-01 1.98529750e-01 4.62557912e-01
-4.58085030e-01 -6.51115596e-01 3.93411517e-01 1.44294411e-01
-1.11942971e+00 -7.54855275e-02 -1.85214907e-01 7.05214918e-01
6.76132202e-01 7.97095895e-02 -7.18066752e-01 7.53686801e-02
-6.93608642e-01 1.13836482e-01 4.88808960e-01 9.11563158e-01
8.34194362e-01 -1.82794273e+00 -5.88272810e-01 -1.51728705e-01
5.73936403e-01 -3.52707714e-01 -2.58570611e-01 4.36848491e-01
1.27499914e-02 3.27572137e-01 -3.97973180e-01 5.33825457e-02
-1.09198964e+00 7.75269449e-01 -4.97708544e-02 -7.08460808e-02
-8.37007165e-01 8.52741539e-01 -4.67516445e-02 -3.82771641e-01
1.92909315e-01 -1.20082743e-01 -5.18278182e-01 3.77879977e-01
5.14265716e-01 1.41628444e-01 -6.15974143e-02 -5.70724368e-01
-9.06812608e-01 5.26441395e-01 -3.24161440e-01 -1.84800997e-02
1.45070887e+00 1.56842977e-01 -3.18233967e-01 2.40289211e-01
1.15361202e+00 1.70517731e-02 -4.54977959e-01 -4.08719778e-01
6.52546942e-01 -4.64496166e-01 -3.97985429e-02 -4.10178274e-01
-6.28114223e-01 9.53412652e-01 1.36503920e-01 -1.98456123e-01
4.45444345e-01 6.59148753e-01 1.15012765e+00 6.88798845e-01
4.48154420e-01 -9.87686276e-01 6.32982180e-02 7.59242475e-01
8.44666839e-01 -7.78331876e-01 -2.08198968e-02 -5.51981926e-01
-5.86130738e-01 1.09362221e+00 8.23594451e-01 -5.72561264e-01
4.75058317e-01 4.65851039e-01 3.18215862e-02 -6.27902031e-01
-8.24748576e-01 -2.98406422e-01 3.29086818e-02 8.01360250e-01
8.61348510e-01 9.25761387e-02 -8.85282397e-01 1.26003242e+00
-2.82330394e-01 -6.02842927e-01 4.27705526e-01 8.52797031e-01
-5.39256275e-01 -1.12883890e+00 -1.71882033e-01 1.34087384e-01
-4.54050243e-01 -3.41567069e-01 -5.71178675e-01 9.07181799e-01
-1.18041046e-01 9.17066932e-01 5.07056534e-01 -1.36502668e-01
5.97679317e-01 3.35377723e-01 6.37431502e-01 -9.86601055e-01
-4.39721793e-01 -6.46806538e-01 5.28272331e-01 -7.23403037e-01
-3.28558505e-01 -5.54177523e-01 -1.85864937e+00 -2.13809729e-01
-2.96461135e-01 4.01700497e-01 1.12141907e-01 1.07548702e+00
1.57678336e-01 3.29111099e-01 1.47539973e-02 -3.81844640e-01
-7.70859480e-01 -1.11733234e+00 -5.57830989e-01 1.02407789e+00
-2.16577619e-01 -1.29041314e+00 -4.71965164e-01 -4.64435309e-01] | [9.95705509185791, 9.016329765319824] |
88fca47d-76fd-481f-be83-1f48d2e7d750 | stack-operation-of-tensor-networks | 2203.16338 | null | https://arxiv.org/abs/2203.16338v2 | https://arxiv.org/pdf/2203.16338v2.pdf | Stack operation of tensor networks | The tensor network, as a facterization of tensors, aims at performing the operations that are common for normal tensors, such as addition, contraction and stacking. However, due to its non-unique network structure, only the tensor network contraction is so far well defined. In this paper, we propose a mathematically rigorous definition for the tensor network stack approach, that compress a large amount of tensor networks into a single one without changing their structures and configurations. We illustrate the main ideas with the matrix product states based machine learning as an example. Our results are compared with the for loop and the efficient coding method on both CPU and GPU. | ['Tianqi Chen', 'Erping Li', 'Bo Yang', 'L. K. Ang', 'Tianning Zhang'] | 2022-03-28 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [ 1.40155673e-01 -1.02136001e-01 8.67732912e-02 -1.79616079e-01
2.71870315e-01 -8.43255341e-01 4.75835502e-01 6.79378584e-02
-1.90486088e-01 3.22209388e-01 2.08596677e-01 -6.66999757e-01
-4.21122432e-01 -5.45998812e-01 -6.11103415e-01 -7.29512334e-01
-6.34020865e-01 5.32295942e-01 2.56411970e-01 -3.30442488e-01
3.28810096e-01 8.49577010e-01 -1.50394154e+00 4.00790542e-01
4.38912034e-01 9.25221205e-01 -2.25509182e-01 8.77743602e-01
-3.39752316e-01 9.59858298e-01 -2.00272635e-01 -7.70853341e-01
3.64611089e-01 1.10732615e-02 -1.20346427e+00 9.17261839e-02
5.41854560e-01 -2.52088726e-01 -4.56744254e-01 9.95083690e-01
-1.33281842e-01 2.67698951e-02 3.19993556e-01 -1.25966406e+00
-8.10586661e-02 1.11573410e+00 -2.69919157e-01 -2.34703310e-02
9.26003307e-02 -1.26992404e-01 1.32074714e+00 -6.88039124e-01
4.72816825e-01 1.00146365e+00 6.34729862e-01 2.44686350e-01
-1.43413365e+00 -2.19191551e-01 -2.73163438e-01 1.50001451e-01
-1.11886740e+00 -2.39964381e-01 8.49801958e-01 -6.28874898e-01
1.07480729e+00 8.25106144e-01 9.03856099e-01 3.74614596e-01
1.74970776e-01 5.67731261e-01 8.54063869e-01 -3.19278449e-01
1.42192006e-01 -3.43458086e-01 6.53006434e-01 8.24778199e-01
4.49357778e-01 -1.66304559e-01 -3.47123265e-01 -3.87503326e-01
8.64793360e-01 2.12026775e-01 -4.29686159e-04 -7.59973645e-01
-1.80518675e+00 6.17576122e-01 3.78170431e-01 5.46992064e-01
-1.64323241e-01 4.96271908e-01 8.31059277e-01 4.22184706e-01
2.35489368e-01 6.34662926e-01 -5.81643224e-01 -2.74755001e-01
-7.92909086e-01 2.67722368e-01 1.30805993e+00 7.09764719e-01
9.60698545e-01 1.44409299e-01 1.86128318e-01 3.94038618e-01
3.13808508e-02 6.04232661e-02 1.74383640e-01 -9.19873774e-01
4.10512269e-01 7.29693353e-01 -2.91991472e-01 -1.07119262e+00
-5.83851695e-01 -6.01954877e-01 -1.16696405e+00 -1.90984920e-01
3.19777906e-01 6.35019168e-02 -4.78436679e-01 1.23863041e+00
2.75364578e-01 1.79835930e-01 -1.59009248e-01 6.53490901e-01
4.92345423e-01 5.09572268e-01 -4.42113966e-01 -5.35663776e-02
1.24191070e+00 -1.02517891e+00 -6.23258829e-01 3.17216039e-01
1.30628157e+00 -1.00414884e+00 5.70180535e-01 6.69816852e-01
-1.09301233e+00 -1.65528297e-01 -8.54441404e-01 -3.35069060e-01
-3.60174149e-01 -5.87409250e-02 1.49382687e+00 6.34654105e-01
-1.26302648e+00 1.01885319e+00 -1.01114798e+00 -2.33948082e-01
1.68345449e-03 6.35417938e-01 -6.93533301e-01 8.17715153e-02
-6.64093077e-01 6.96289361e-01 7.50589907e-01 4.63669538e-01
-4.26919639e-01 -7.43659854e-01 -5.05878329e-01 2.77325034e-01
4.03689325e-01 -7.14133620e-01 1.19627750e+00 -3.59501809e-01
-1.50356436e+00 5.82131982e-01 5.91858178e-02 -3.45750064e-01
1.31292775e-01 9.07735080e-02 5.43330237e-02 -6.15106849e-03
-5.37064552e-01 2.40031064e-01 6.86523497e-01 -9.03151870e-01
-1.43751070e-01 -3.34881097e-01 5.30142128e-01 -1.70436874e-01
-3.22694957e-01 2.44551867e-01 -1.41692162e-01 -5.31442881e-01
7.07332075e-01 -1.13001919e+00 -4.64267522e-01 -4.07938898e-01
-8.59047830e-01 -7.43494034e-02 6.27641022e-01 -4.61593151e-01
1.41803777e+00 -2.01509023e+00 7.16279864e-01 7.24489093e-01
9.20021772e-01 -2.06217635e-02 8.25411379e-02 7.58623421e-01
-6.95592284e-01 2.59354770e-01 -2.22849980e-01 -3.42314810e-01
-1.18504144e-01 5.48686564e-01 -2.70193368e-01 4.55877602e-01
-1.52566314e-01 5.76897562e-01 -9.27842975e-01 -3.99760425e-01
7.32637942e-02 2.79666066e-01 -9.18310523e-01 4.20129970e-02
-1.00463800e-01 1.90072954e-01 -2.57302552e-01 4.69507068e-01
8.52502406e-01 -2.15139180e-01 6.70249104e-01 -8.72248828e-01
-3.94625753e-01 3.93593490e-01 -1.31803870e+00 1.50471902e+00
-7.86044076e-02 3.77062976e-01 1.50166482e-01 -1.30930936e+00
7.18410611e-01 2.25485459e-01 8.30007315e-01 7.83833191e-02
9.44683254e-02 4.24942225e-01 4.50199842e-01 -3.99774820e-01
7.30998874e-01 -5.33693247e-02 4.37855199e-02 6.78804159e-01
1.85167700e-01 -4.21668530e-01 7.79043555e-01 5.43664932e-01
1.33944869e+00 -2.56147593e-01 -9.28306952e-04 -4.55277026e-01
5.48118532e-01 -5.96879534e-02 5.90779632e-02 2.75850415e-01
3.51412773e-01 2.67268896e-01 1.28412199e+00 -6.55295849e-01
-1.43737197e+00 -8.97009373e-01 7.75807649e-02 8.59597683e-01
-5.96814692e-01 -1.06597316e+00 -7.84281433e-01 -4.25066561e-01
5.46572730e-02 2.40270853e-01 -3.29179049e-01 8.08762014e-02
-7.59632468e-01 -6.26215160e-01 4.91181314e-01 1.70704290e-01
1.85370266e-01 -3.28419000e-01 -1.70927286e-01 -9.69898999e-02
1.09621890e-01 -1.04775989e+00 -4.66603845e-01 1.48795068e-01
-1.41607475e+00 -1.04320681e+00 -1.95473060e-01 -4.87313181e-01
7.10566938e-01 3.01311255e-01 9.39204037e-01 3.27881753e-01
9.33705568e-02 1.11050799e-01 -1.45564988e-01 2.67010510e-01
-4.26810890e-01 2.06528619e-01 1.79266065e-01 2.42910743e-01
-1.36124492e-01 -8.79877865e-01 -1.50162548e-01 5.57842962e-02
-1.19920838e+00 3.83425891e-01 7.44209647e-01 7.12590039e-01
2.39818513e-01 2.59534180e-01 -6.41108632e-01 -9.09806550e-01
6.23070836e-01 -2.13731647e-01 -5.61387599e-01 9.61457640e-02
-3.45508456e-01 5.51830053e-01 5.59328616e-01 -2.72095799e-01
-4.60547984e-01 1.10383958e-01 2.50054985e-01 -4.80454654e-01
4.78561819e-01 9.03071821e-01 1.36250660e-01 -4.44219708e-01
4.16933000e-01 -1.11464895e-02 2.24323466e-01 -9.39435840e-01
3.96410614e-01 1.47285908e-01 3.46852839e-01 -8.89522731e-01
8.91181707e-01 4.05927658e-01 7.88790762e-01 -9.02045012e-01
-5.58410466e-01 -3.74399930e-01 -8.96419048e-01 -2.84173608e-01
4.88480657e-01 -2.28719980e-01 -1.25843203e+00 4.70637172e-01
-1.55180287e+00 1.40897542e-01 -1.63293719e-01 7.22991526e-01
-4.35713589e-01 6.71422660e-01 -9.16081965e-01 -5.03650069e-01
-1.53357893e-01 -1.21411490e+00 7.09157348e-01 -5.65481544e-01
-3.50132920e-02 -1.05653286e+00 2.32012451e-01 2.11208671e-01
4.49061871e-01 1.97670951e-01 1.19686270e+00 -5.84485054e-01
-9.68234658e-01 -1.15474299e-01 -2.98176169e-01 7.16795802e-01
-2.73894519e-01 5.77319145e-01 -4.06454712e-01 -2.26772666e-01
2.85278652e-02 2.84890711e-01 7.14188933e-01 -1.43274546e-01
1.24523115e+00 -5.25557280e-01 -8.65391344e-02 7.13549495e-01
1.52878237e+00 -2.25213289e-01 6.27702534e-01 -1.02141783e-01
1.23391747e+00 6.46969318e-01 -1.48015186e-01 3.48747253e-01
1.56045601e-01 5.98302066e-01 4.99819428e-01 1.48075551e-01
1.63602322e-01 4.54450250e-02 2.52744377e-01 1.80844760e+00
-8.58601213e-01 4.03810620e-01 -1.09890437e+00 2.42945880e-01
-1.79097795e+00 -8.78819883e-01 -7.56957591e-01 2.31947899e+00
6.57052636e-01 1.68522984e-01 4.41622883e-02 4.52000618e-01
5.21499634e-01 2.28491485e-01 1.22452557e-01 -8.56475830e-01
-1.15235284e-01 4.79729593e-01 9.77895439e-01 5.31175911e-01
-8.47790837e-01 6.29315257e-01 7.52928305e+00 4.63666290e-01
-1.12415481e+00 2.74003118e-01 -8.04384202e-02 -1.21072687e-01
-4.13243949e-01 3.55864882e-01 -2.25009978e-01 1.02663942e-01
9.16929126e-01 -1.37962967e-01 9.20905828e-01 7.10543871e-01
2.42022518e-02 1.52821198e-01 -1.43084574e+00 1.18025577e+00
-3.15949231e-01 -1.65610933e+00 3.01844299e-01 1.40527949e-01
5.40340364e-01 7.26340860e-02 -2.24027812e-01 -2.68070232e-02
7.48592019e-02 -7.57119179e-01 7.27976322e-01 6.35207772e-01
2.67210752e-01 -6.07721269e-01 6.43294573e-01 -1.50039062e-01
-1.17839456e+00 -9.31948237e-03 -1.17637679e-01 -5.33205509e-01
8.87021143e-03 1.12412143e+00 -5.81325471e-01 7.87015378e-01
4.45470750e-01 7.19750583e-01 -4.67506170e-01 1.01637793e+00
7.85984173e-02 6.23757005e-01 -6.05952680e-01 -1.33916020e-01
4.18524474e-01 -7.71997929e-01 5.98784208e-01 1.23780298e+00
2.68080175e-01 -1.40851274e-01 1.34776048e-02 6.41571105e-01
3.83945629e-02 6.58691600e-02 -5.28340101e-01 -4.68475312e-01
1.00263193e-01 1.31037903e+00 -5.65626919e-01 -3.77454877e-01
-3.58605981e-01 6.07141972e-01 2.27150768e-01 2.86944568e-01
-4.87550735e-01 -4.54221100e-01 9.46605563e-01 7.64783844e-02
2.18583003e-01 -9.06380177e-01 -5.03763437e-01 -1.41425061e+00
3.27729821e-01 -7.30307817e-01 3.21691204e-03 -6.43113017e-01
-9.86278415e-01 4.52044487e-01 3.46123755e-01 -1.17925775e+00
-1.43859208e-01 -1.07527018e+00 -2.28332609e-01 7.50125468e-01
-1.04955328e+00 -1.01077425e+00 1.01867348e-01 5.45012593e-01
-4.52376366e-01 1.55319571e-01 7.03719735e-01 5.57408273e-01
-7.26502597e-01 1.62300467e-01 1.89864397e-01 5.09176850e-02
6.65759295e-02 -1.29714370e+00 2.86654025e-01 7.92062819e-01
-2.65164912e-01 1.05802464e+00 8.99761856e-01 -4.12926257e-01
-2.24587107e+00 -6.82506084e-01 1.13835764e+00 -5.53418696e-02
1.36893439e+00 -5.27754903e-01 -9.41652179e-01 1.05997753e+00
-5.96036352e-02 3.09811514e-02 4.17606860e-01 3.63781363e-01
-6.22680783e-01 -3.74054551e-01 -8.29473495e-01 6.95584416e-01
1.31853855e+00 -6.71202421e-01 -3.31833959e-01 7.41704345e-01
8.33082557e-01 -5.98093808e-01 -1.42987692e+00 9.71773714e-02
5.44925451e-01 -1.07972658e+00 1.02367496e+00 -8.18060279e-01
4.25225824e-01 -4.20783162e-01 -2.56425828e-01 -1.09268522e+00
-4.11468625e-01 -1.20644164e+00 -3.11552405e-01 7.66754985e-01
3.01801443e-01 -9.23233032e-01 8.14836204e-01 4.66543585e-01
-2.84176975e-01 -6.65612519e-01 -9.32893753e-01 -9.79263067e-01
-3.53015482e-01 -6.88816786e-01 8.44685614e-01 1.09986031e+00
3.22437644e-01 3.91265392e-01 -3.38070422e-01 1.31235104e-02
6.83917880e-01 -1.55879036e-01 7.89269567e-01 -1.26930571e+00
-3.97908658e-01 -6.39898777e-01 -8.47073197e-01 -9.13315296e-01
-1.25537992e-01 -1.23912263e+00 -6.68506622e-01 -1.12053049e+00
1.61422253e-01 -6.63396597e-01 -9.47498083e-02 4.48918343e-01
6.21716857e-01 1.14661865e-02 3.92417967e-01 3.73008877e-01
-2.60504782e-01 1.80195555e-01 1.33389759e+00 -9.40283239e-02
1.27706215e-01 -2.44476721e-01 -3.52339417e-01 5.17030060e-01
5.90395510e-01 -3.89663279e-01 -1.90153912e-01 -6.42017007e-01
8.76827717e-01 8.93563963e-03 2.07852766e-01 -8.63009870e-01
2.70739496e-01 -1.12741828e-01 -4.94561583e-01 -7.10795343e-01
5.09893537e-01 -8.29564750e-01 6.50633931e-01 6.69230103e-01
-7.23876432e-02 5.29278100e-01 -1.12416372e-01 -4.41328995e-02
-2.51455307e-01 -3.95264566e-01 2.55188376e-01 -1.22186154e-01
-3.62171024e-01 4.02309000e-01 -2.07440108e-02 -4.49313492e-01
6.13858819e-01 -1.64603040e-01 -3.52226198e-01 -3.80077958e-03
-9.28725123e-01 -2.74459273e-01 3.54447007e-01 -1.62821990e-02
4.09009278e-01 -1.40492868e+00 -6.48649454e-01 1.45642474e-01
-1.78096473e-01 -1.92324631e-02 2.80371785e-01 1.33738852e+00
-1.09425282e+00 7.12343216e-01 -2.22612888e-01 -7.04630315e-01
-1.28041780e+00 8.77342045e-01 4.07030255e-01 -6.23134255e-01
-3.23911428e-01 4.11511749e-01 -1.59285605e-01 -7.39719927e-01
-8.40208530e-02 -1.10867405e+00 1.89892977e-01 -1.48290336e-01
3.42288375e-01 6.62271976e-01 4.98505771e-01 -4.57853317e-01
-2.49418199e-01 5.97230792e-01 1.40871465e-01 7.54452422e-02
1.27599907e+00 3.96551937e-01 -1.42079198e+00 5.26160121e-01
1.50487351e+00 -7.37983063e-02 -2.70326793e-01 -1.44857228e-01
3.07173058e-02 -3.83150548e-01 6.88196793e-02 -9.94819254e-02
-1.26259506e+00 7.28554010e-01 3.85441855e-02 7.85690665e-01
8.27266335e-01 -1.58624724e-02 6.31278217e-01 8.95645618e-01
2.70599991e-01 -6.92645431e-01 -2.57335961e-01 9.29677129e-01
9.34520125e-01 -4.57083285e-01 2.16351971e-01 -8.76024127e-01
4.74699438e-02 1.55874801e+00 3.70353125e-02 -1.93822563e-01
9.81586933e-01 3.84336054e-01 -6.94922984e-01 -4.50799882e-01
-7.81455159e-01 -3.09463218e-02 1.31395116e-01 1.63462669e-01
6.00528777e-01 5.22226572e-01 -5.11763990e-01 -1.56688318e-01
-7.25687325e-01 7.39162788e-02 8.09017181e-01 9.04430032e-01
-1.28594846e-01 -1.26801372e+00 -5.78989327e-01 6.05642796e-01
-2.26947069e-01 -2.52658665e-01 -1.45985454e-01 6.53609216e-01
6.02867529e-02 4.72073495e-01 3.18659008e-01 -8.50650907e-01
3.29526067e-01 9.98032093e-03 8.85433555e-01 -4.92751032e-01
-7.45063841e-01 -3.61591369e-01 5.05834818e-01 -6.39129460e-01
-4.67563838e-01 -6.49646103e-01 -1.10023534e+00 -1.12709498e+00
-1.63934767e-01 2.39322677e-01 1.19285917e+00 8.55258822e-01
3.34559411e-01 4.73995477e-01 6.73991323e-01 -9.68611121e-01
-6.53572977e-01 -9.69217539e-01 -7.63419747e-01 3.60408217e-01
2.56624669e-01 -5.62667370e-01 -4.03035820e-01 -7.99218565e-02] | [6.207220554351807, 5.006288528442383] |
71551d16-f1da-423a-a023-4f64edb6088e | pypots-a-python-toolbox-for-data-mining-on | 2305.18811 | null | https://arxiv.org/abs/2305.18811v1 | https://arxiv.org/pdf/2305.18811v1.pdf | PyPOTS: A Python Toolbox for Data Mining on Partially-Observed Time Series | PyPOTS is an open-source Python library dedicated to data mining and analysis on multivariate partially-observed time series, i.e. incomplete time series with missing values, A.K.A. irregularlysampled time series. Particularly, it provides easy access to diverse algorithms categorized into four tasks: imputation, classification, clustering, and forecasting. The included models contain probabilistic approaches as well as neural-network methods, with a well-designed and fully-documented programming interface for both academic researchers and industrial professionals to use. With robustness and scalability in its design philosophy, best practices of software construction, for example, unit testing, continuous integration (CI) and continuous delivery (CD), code coverage, maintainability evaluation, interactive tutorials, and parallelization, are carried out as principles during the development of PyPOTS. The toolkit is available on both Python Package Index (PyPI) and Anaconda. PyPOTS is open-source and publicly available on GitHub https://github.com/WenjieDu/PyPOTS. | ['Wenjie Du'] | 2023-05-30 | null | null | null | null | ['imputation', 'imputation', 'philosophy', 'multivariate-time-series-imputation', 'time-series-clustering', 'irregular-time-series', 'imputation', 'classification-on-time-series-with-missing', 'traffic-data-imputation'] | ['computer-vision', 'miscellaneous', 'miscellaneous', 'time-series', 'time-series', 'time-series', 'time-series', 'time-series', 'time-series'] | [-3.05515766e-01 -5.02323091e-01 -1.86480522e-01 -3.45153600e-01
-7.42106020e-01 -6.73083246e-01 1.65432274e-01 8.75482485e-02
5.26842624e-02 7.08584607e-01 -6.06534258e-02 -5.51090956e-01
-3.71799767e-01 -6.71713412e-01 -5.85334539e-01 -8.58012319e-01
-4.12831843e-01 5.42092383e-01 -3.87533039e-01 1.98579177e-01
2.45148093e-01 3.37901264e-01 -1.84667110e+00 9.41798091e-02
8.48521769e-01 9.04011846e-01 9.58428234e-02 6.34860933e-01
1.49458438e-01 8.74809086e-01 -3.85756254e-01 1.15604103e-02
2.02076919e-02 3.22759300e-02 -3.74351114e-01 -3.59580219e-01
-6.62874937e-01 -2.46315926e-01 6.39677420e-02 5.59604466e-01
5.07790685e-01 -3.05372536e-01 4.69351560e-01 -1.99951994e+00
-4.68366325e-01 6.96534812e-01 -4.14904326e-01 8.84705037e-03
3.43493670e-01 2.84874916e-01 3.83937746e-01 -9.13599432e-01
1.95843667e-01 7.03097939e-01 9.67122674e-01 -1.55453742e-01
-1.39007378e+00 -8.98862660e-01 -4.39895988e-01 3.51522118e-01
-1.47303617e+00 -3.58587891e-01 7.26081789e-01 -7.56265938e-01
9.14595246e-01 7.47130513e-01 3.60332966e-01 1.34308410e+00
2.22355783e-01 7.34020233e-01 1.09851730e+00 -2.11239591e-01
5.44364870e-01 8.37826654e-02 5.25696278e-01 -1.53040186e-01
1.20911412e-01 3.28298450e-01 -2.81906605e-01 -7.71735907e-01
4.77172792e-01 4.53869760e-01 -1.32442787e-01 4.43677939e-02
-1.36756134e+00 4.75582600e-01 -3.85419428e-01 1.18317202e-01
-8.00422013e-01 -3.73011202e-01 6.10283434e-01 5.96577704e-01
5.96309423e-01 -6.22353368e-02 -1.21788514e+00 -7.26316690e-01
-9.63171959e-01 4.72542048e-01 1.09105182e+00 1.08646750e+00
6.78976536e-01 1.83157936e-01 -2.18731090e-02 7.82046914e-01
-2.84175146e-02 4.38695550e-01 5.95793843e-01 -1.08890367e+00
2.28903517e-01 5.45159280e-01 2.26601809e-01 -7.61582136e-01
-6.85419381e-01 -2.95782417e-01 -1.21245670e+00 9.61604118e-02
2.97651380e-01 -5.00001609e-01 -4.27171022e-01 1.54462004e+00
3.02215934e-01 3.05466503e-01 -1.08608365e-01 4.70001489e-01
6.88428760e-01 6.50310874e-01 -5.10533117e-02 -6.43058598e-01
1.02361906e+00 -6.34121418e-01 -6.77169561e-01 2.30041206e-01
3.91310662e-01 -9.14218545e-01 7.86293983e-01 6.42222703e-01
-9.98605311e-01 -3.91000748e-01 -4.87079352e-01 3.63247663e-01
-5.20258009e-01 1.88541397e-01 5.56136787e-01 4.52048808e-01
-8.05502534e-01 8.36144149e-01 -1.36352015e+00 -1.38265520e-01
1.80653840e-01 2.37000316e-01 -3.17226529e-01 1.62412524e-01
-9.29210424e-01 4.87331927e-01 3.29635203e-01 5.15661389e-02
-5.97551942e-01 -1.19741976e+00 -8.30376387e-01 -8.83456692e-02
1.99298963e-01 -5.78018486e-01 1.13973200e+00 -6.59273207e-01
-1.37112558e+00 4.88790393e-01 -2.60571599e-01 -1.43032968e-01
4.98844594e-01 -2.21817493e-01 -6.91811740e-01 -4.64596540e-01
2.38280252e-01 -1.77320451e-01 5.46488106e-01 -6.60759211e-01
-3.43818426e-01 -5.35404503e-01 -8.80046666e-01 -3.51201624e-01
-9.53481868e-02 4.22222614e-01 -1.65076345e-01 -9.21729684e-01
-2.00763583e-01 -7.55052686e-01 -2.19937012e-01 -3.48534495e-01
-5.36079168e-01 -2.58148640e-01 8.09793055e-01 -1.31206954e+00
1.47544801e+00 -2.18165660e+00 -7.83925876e-02 2.91112453e-01
-2.97222465e-01 -1.83583871e-01 1.20747201e-01 1.21218526e+00
-6.08868480e-01 -1.38893425e-01 -5.97891808e-01 -3.06708068e-01
2.82497585e-01 2.13811174e-03 -3.40016693e-01 6.10649228e-01
7.70065263e-02 4.51942205e-01 -5.73344171e-01 8.47235136e-03
5.85438371e-01 3.65045309e-01 -1.40835911e-01 4.63115335e-01
-1.39246598e-01 8.98665726e-01 -1.69591963e-01 1.13946307e+00
1.04214644e+00 -4.00925577e-01 2.16592059e-01 3.54015201e-01
-7.49031007e-01 5.89770973e-02 -1.35095882e+00 1.49748600e+00
-2.81967342e-01 2.07841113e-01 -8.49249661e-02 -8.59185576e-01
1.08152330e+00 5.21788359e-01 7.38840759e-01 -3.47188443e-01
-5.46938218e-02 2.92355776e-01 -4.31144476e-01 -7.09150314e-01
2.66145080e-01 6.89017951e-01 -7.64301568e-02 8.76040399e-01
-1.52919933e-01 4.49524671e-01 2.81562716e-01 -2.48857677e-01
1.13609231e+00 5.93523264e-01 5.37284076e-01 -1.85301587e-01
2.64057755e-01 2.01930299e-01 7.93676436e-01 5.00592291e-01
1.40507549e-01 5.22576928e-01 5.97570002e-01 -3.66283387e-01
-1.38164163e+00 -9.42143142e-01 -5.82233012e-01 1.12231600e+00
-5.93579233e-01 -4.25540447e-01 -4.29857999e-01 2.81793118e-01
1.27474979e-01 9.03814614e-01 -4.06993508e-01 3.61656398e-01
-2.89346874e-01 -9.90126491e-01 3.71439070e-01 3.98881912e-01
1.95813812e-02 -1.47510624e+00 -6.10544980e-01 4.04008597e-01
-3.43225688e-01 -6.87071145e-01 -5.56528335e-03 3.15957308e-01
-8.66929591e-01 -1.12347865e+00 -5.45651257e-01 -5.35715222e-01
3.22447091e-01 -2.18238309e-01 1.30889630e+00 -2.26235017e-01
-4.52472091e-01 2.55770177e-01 -4.18348700e-01 -7.15684295e-01
-1.37224987e-01 -1.58780321e-01 -5.71463909e-03 -2.53761500e-01
5.72475672e-01 -1.23846662e+00 -4.60249692e-01 5.09989619e-01
-7.75057793e-01 2.00389046e-02 1.78468153e-01 8.00907731e-01
6.71625674e-01 1.61305219e-01 6.37186110e-01 -3.58787864e-01
5.43512642e-01 -1.24988055e+00 -1.02985024e+00 3.05351645e-01
-6.83700502e-01 -5.08703947e-01 7.14246094e-01 -2.55344659e-01
-6.63109064e-01 1.20693231e-02 -3.73975158e-01 -4.63892877e-01
-7.83148646e-01 9.49527264e-01 -1.48937497e-02 5.27446806e-01
3.60415936e-01 7.14561999e-01 2.43144289e-01 -8.97982121e-01
-1.49156928e-01 1.06638348e+00 4.22953606e-01 -5.73166251e-01
3.72381955e-01 2.91381955e-01 -4.28864509e-01 -7.32267201e-01
-4.45828885e-02 -4.56237137e-01 -5.08546293e-01 -6.12375699e-02
2.00361758e-01 -1.07352841e+00 -9.47231710e-01 1.07658613e+00
-8.56864035e-01 -3.82930636e-01 1.16087593e-01 4.57401663e-01
-5.75885177e-01 1.23109616e-01 -5.14686346e-01 -1.13464415e+00
-6.10968888e-01 -8.15895379e-01 8.04675817e-01 2.31987201e-02
-5.36649644e-01 -9.40375924e-01 3.62429559e-01 -1.32109886e-02
4.23669189e-01 9.00014818e-01 6.97219729e-01 -7.83614635e-01
-2.14901134e-01 -4.61349815e-01 1.49446934e-01 4.04447526e-01
3.45469126e-03 6.21878028e-01 -9.35350120e-01 -2.70803124e-01
5.48874103e-02 -1.94672242e-01 1.22456485e-02 7.15362489e-01
1.78247046e+00 -5.08107424e-01 -1.76114246e-01 6.14775360e-01
1.25520921e+00 2.61012495e-01 7.27307141e-01 5.24422407e-01
1.43434480e-01 7.74375856e-01 6.37903869e-01 1.24834692e+00
6.04985118e-01 4.54294145e-01 4.38237578e-01 7.67874271e-02
8.03705871e-01 2.15464205e-01 3.92647535e-01 8.91603351e-01
-7.03151599e-02 2.16490448e-01 -1.20806646e+00 6.46551311e-01
-2.09327888e+00 -1.20282304e+00 -8.85357797e-01 2.46930742e+00
8.93576026e-01 -2.96095908e-01 4.44560647e-01 4.54168797e-01
6.99299097e-01 -2.33688802e-01 -5.61774194e-01 -1.99263051e-01
-1.89625338e-01 6.66613132e-02 2.97731549e-01 -2.39179991e-02
-1.06684399e+00 5.79265952e-02 5.85700655e+00 8.14089298e-01
-8.89165938e-01 8.68626758e-02 6.50864005e-01 1.01504318e-01
-4.43754420e-02 -2.70970576e-02 -3.32200199e-01 1.03553855e+00
1.35544395e+00 -5.81067681e-01 5.73743522e-01 1.09739721e+00
7.05956697e-01 2.94702090e-02 -9.14884269e-01 8.64513755e-01
-5.49595475e-01 -1.32454479e+00 -8.50334525e-01 -5.75158298e-02
4.91649508e-01 4.76369083e-01 -7.13406578e-02 3.79171073e-01
2.46623725e-01 -7.89005756e-01 7.27917492e-01 6.69610977e-01
5.89265168e-01 -8.26857448e-01 9.16384041e-01 6.37589037e-01
-9.57170904e-01 -1.50441870e-01 -2.23343391e-02 -4.49842274e-01
2.09753841e-01 9.89201248e-01 -2.12109819e-01 1.04720402e+00
1.23782599e+00 8.45902681e-01 -2.83848375e-01 1.31648791e+00
1.01312824e-01 8.45849574e-01 -4.25289631e-01 4.38454270e-01
-4.29694891e-01 -3.59690070e-01 4.66209054e-01 9.98003423e-01
5.37388504e-01 -1.26427561e-01 2.87650615e-01 8.18492889e-01
6.92695260e-01 -9.21114162e-02 -3.02916437e-01 2.05756053e-02
7.86160469e-01 1.33111072e+00 -3.42233986e-01 -1.11253023e-01
-3.99715483e-01 3.88818145e-01 -1.21291749e-01 4.93847817e-01
-9.69299853e-01 -4.84311819e-01 8.09814155e-01 6.42827973e-02
1.00582212e-01 -2.83457369e-01 -6.95018888e-01 -9.46808696e-01
1.88618600e-01 -1.00448632e+00 7.18279958e-01 -8.81379485e-01
-1.69682050e+00 4.51003641e-01 2.16351658e-01 -1.58659434e+00
-5.36257207e-01 -3.98378730e-01 -9.59623456e-01 1.11080194e+00
-1.21269464e+00 -9.15875733e-01 -4.73352194e-01 7.94518232e-01
1.96582556e-01 -1.20913655e-01 1.24608397e+00 6.93419576e-01
-1.04127300e+00 1.59219563e-01 7.47516155e-01 -4.01699767e-02
4.60717112e-01 -1.08568966e+00 4.61102754e-01 6.73637331e-01
-6.63482368e-01 8.14400852e-01 7.35654294e-01 -6.65111959e-01
-1.50894034e+00 -1.30494666e+00 8.05984557e-01 -3.29175919e-01
1.00498843e+00 -2.94017285e-01 -1.03621757e+00 8.31674814e-01
1.88538834e-01 -3.31196487e-01 1.13179147e+00 1.68370530e-01
-1.22809716e-01 -8.44324455e-02 -9.61943448e-01 2.83937216e-01
3.16350967e-01 -1.38030916e-01 -3.27761739e-01 5.69639146e-01
4.27267969e-01 -3.98339152e-01 -1.31713307e+00 3.51798892e-01
6.42251790e-01 -1.09102786e+00 8.06623697e-01 -2.55574226e-01
3.79156440e-01 -4.14553136e-01 -4.04460239e-04 -8.43605757e-01
-2.47787610e-01 -8.15485299e-01 -2.94909745e-01 1.38623524e+00
3.54997754e-01 -9.22973335e-01 2.54265785e-01 6.88354552e-01
-2.82277882e-01 -4.25161690e-01 -9.03364480e-01 -9.40195858e-01
-1.67882577e-01 -9.40331280e-01 1.03745854e+00 1.32959139e+00
7.93932900e-02 -5.36218762e-01 -6.66535735e-01 3.67671192e-01
9.74716246e-01 2.66997904e-01 1.03600526e+00 -1.32483268e+00
-4.05496329e-01 -2.63073891e-01 -6.00577611e-03 -2.71591723e-01
-1.27313420e-01 -5.03239274e-01 -3.72101933e-01 -1.17739296e+00
2.18682755e-02 -4.85829502e-01 -3.46884251e-01 8.29104125e-01
1.76742375e-01 -2.54049208e-02 -4.20668036e-01 6.15064502e-01
-1.41148776e-01 3.91913742e-01 5.02800763e-01 2.43428588e-01
-3.66225392e-01 5.57614565e-01 -3.14808041e-01 4.48624641e-01
1.21369708e+00 -8.34287345e-01 9.76942033e-02 -2.26937398e-01
-1.21175475e-01 5.19687593e-01 8.51682603e-01 -7.03138351e-01
1.42959535e-01 -3.52050781e-01 3.51347893e-01 -1.19677734e+00
3.04900557e-02 -9.13268685e-01 1.26137316e+00 4.75132346e-01
9.07515883e-02 5.09735048e-01 3.76406729e-01 2.02416897e-01
-2.47941419e-01 1.10820882e-01 2.40605921e-01 4.30246219e-02
-4.52018142e-01 3.40673834e-01 -2.32007831e-01 -4.15973783e-01
1.21773827e+00 -1.74423933e-01 -3.46977830e-01 -3.29293787e-01
-9.55673158e-01 5.04679024e-01 4.26003307e-01 4.97040451e-01
2.71863908e-01 -1.26117861e+00 -1.01864624e+00 3.95867348e-01
2.06239268e-01 -2.47937933e-01 6.64600670e-01 1.36172986e+00
-4.89091158e-01 1.99708536e-01 -3.51276487e-01 -7.63008416e-01
-9.91615355e-01 6.70804381e-01 -1.30258501e-01 -1.14064723e-01
-7.30812788e-01 2.10931271e-01 -4.65114415e-01 -9.91201341e-01
3.46785933e-01 -1.65886864e-01 -6.45755380e-02 -1.04710981e-01
7.64250398e-01 9.09042835e-01 1.80812418e-01 -1.37778834e-01
-3.59732896e-01 2.48697877e-01 4.37719405e-01 2.09988266e-01
1.96614754e+00 -1.24824516e-01 -7.68726528e-01 9.78764653e-01
8.71119022e-01 -4.45399523e-01 -1.05797565e+00 9.29177329e-02
2.28553951e-01 -3.76955807e-01 -3.33714098e-01 -9.49487984e-01
-5.88786304e-01 3.42566133e-01 5.01347303e-01 6.42113328e-01
1.34443700e+00 -2.65406907e-01 3.14566135e-01 -9.23235789e-02
4.84860271e-01 -7.79949605e-01 -7.83511341e-01 3.44033360e-01
1.07622802e+00 -1.05276847e+00 -9.79663059e-02 -7.56902620e-02
-5.89365780e-01 1.07143819e+00 2.03509033e-01 -9.59309563e-03
1.04887009e+00 5.36204636e-01 2.34758612e-02 9.44102407e-02
-1.01329565e+00 2.91660607e-01 -8.79295543e-02 6.28128707e-01
5.87531090e-01 2.93944746e-01 -2.95575678e-01 1.17102802e+00
-3.11666697e-01 5.37260652e-01 4.84564096e-01 1.02071917e+00
2.57062286e-01 -1.07159448e+00 -8.55594039e-01 5.21709621e-01
-5.08394957e-01 -2.15390891e-01 2.31512308e-01 6.93004310e-01
-1.33551657e-01 1.07167232e+00 1.67248905e-01 -2.19561383e-01
2.30432838e-01 7.93339387e-02 -3.34143430e-01 -8.38900208e-02
-7.77835250e-01 2.36629397e-01 -2.92089619e-02 -4.99435782e-01
-1.45530507e-01 -1.05409062e+00 -7.99122930e-01 -7.49966681e-01
8.94368291e-02 3.15706819e-01 9.51896787e-01 5.95133245e-01
6.96457565e-01 3.32249820e-01 8.88784468e-01 -8.20718586e-01
-3.76897246e-01 -1.10034347e+00 -6.93342805e-01 3.48573066e-02
1.70704991e-01 -3.52477342e-01 -4.34073091e-01 6.72075525e-02] | [7.209350109100342, 3.5607569217681885] |
4a02546e-3858-4a70-895c-0fc0fcb8c9c6 | dependency-language-models-for-transition | 1607.04982 | null | http://arxiv.org/abs/1607.04982v2 | http://arxiv.org/pdf/1607.04982v2.pdf | Dependency Language Models for Transition-based Dependency Parsing | In this paper, we present an approach to improve the accuracy of a strong
transition-based dependency parser by exploiting dependency language models
that are extracted from a large parsed corpus. We integrated a small number of
features based on the dependency language models into the parser. To
demonstrate the effectiveness of the proposed approach, we evaluate our parser
on standard English and Chinese data where the base parser could achieve
competitive accuracy scores. Our enhanced parser achieved state-of-the-art
accuracy on Chinese data and competitive results on English data. We gained a
large absolute improvement of one point (UAS) on Chinese and 0.5 points for
English. | ['Juntao Yu', 'Bernd Bohnet'] | 2016-07-18 | dependency-language-models-for-transition-1 | https://aclanthology.org/W17-6302 | https://aclanthology.org/W17-6302.pdf | ws-2017-9 | ['transition-based-dependency-parsing'] | ['natural-language-processing'] | [-3.61776084e-01 3.39781702e-01 -2.28175521e-01 -7.25169599e-01
-1.39077687e+00 -6.12998188e-01 3.01441818e-01 1.85002878e-01
-7.07067907e-01 7.86468148e-01 3.07484120e-01 -5.79966009e-01
3.73959064e-01 -6.83879495e-01 -4.53955531e-01 -2.38426074e-01
-1.70904920e-01 2.41647616e-01 6.44785285e-01 -3.91755551e-01
2.15825930e-01 3.94548118e-01 -9.33147371e-01 5.23323953e-01
9.77349222e-01 4.93157744e-01 3.53211224e-01 9.62256491e-01
-4.67578143e-01 7.84728169e-01 -6.09023571e-01 -7.52413750e-01
-6.78365454e-02 -1.77469701e-01 -1.05038464e+00 -5.17355084e-01
3.47880856e-03 8.74229819e-02 -7.56743401e-02 8.86466205e-01
2.37152740e-01 -4.00596410e-01 1.95440948e-02 -6.54572427e-01
-6.48012638e-01 1.19650412e+00 -2.86223024e-01 4.24833387e-01
5.29530942e-01 -3.59276056e-01 1.31425703e+00 -7.93637574e-01
6.99700296e-01 1.13204515e+00 5.88308632e-01 4.95063841e-01
-8.52766573e-01 -6.59521639e-01 3.69121045e-01 -1.89374387e-01
-9.96175468e-01 -3.84674907e-01 5.57127893e-01 3.48320119e-02
1.63217151e+00 -2.01824710e-01 2.21654937e-01 8.41996789e-01
6.82064831e-01 8.02231908e-01 1.32024288e+00 -8.57613504e-01
-6.03784509e-02 1.03444960e-02 8.88372242e-01 8.01268458e-01
1.50464758e-01 3.17445360e-02 -5.13528109e-01 3.41784433e-02
2.97908485e-01 -6.75701559e-01 3.55473191e-01 3.98469746e-01
-7.96483338e-01 1.05958951e+00 3.57133269e-01 6.20637417e-01
-1.44703656e-01 -1.30261242e-01 4.55810040e-01 2.87688971e-01
5.71164370e-01 2.96979696e-01 -1.04484367e+00 -4.32654917e-01
-6.23712838e-01 -1.32190749e-01 1.09536135e+00 1.20680535e+00
4.72854793e-01 -3.84858131e-01 5.30823544e-02 7.24136949e-01
1.78955570e-01 6.09340250e-01 3.38557750e-01 -5.90602338e-01
1.05879211e+00 5.24989069e-01 -3.55004258e-02 -3.65913272e-01
-7.14357018e-01 -1.65860638e-01 -3.78576159e-01 -1.80693284e-01
4.38960224e-01 -4.38533634e-01 -9.93110478e-01 1.57688451e+00
8.98157582e-02 -3.98401350e-01 5.39623320e-01 3.75053942e-01
6.69456065e-01 6.06297433e-01 5.40636241e-01 -3.18613619e-01
1.39859951e+00 -1.20792627e+00 -7.92066813e-01 -4.12020266e-01
1.26050925e+00 -8.74666929e-01 8.86816978e-01 2.08355054e-01
-1.07857478e+00 -6.21611595e-01 -8.60638857e-01 -1.15611441e-01
-2.93517321e-01 2.52694845e-01 8.27256680e-01 9.27976668e-01
-9.87426877e-01 6.29551828e-01 -1.20405543e+00 -4.10487264e-01
-1.86663121e-02 5.59210241e-01 -7.93934643e-01 -9.35719982e-02
-1.20432436e+00 1.01517653e+00 5.37542105e-01 -6.82537109e-02
-1.26937419e-01 -3.07479054e-01 -9.83167231e-01 1.05744861e-01
1.61889613e-01 8.07580724e-02 1.40492690e+00 -3.87737840e-01
-1.60347390e+00 7.13034451e-01 -2.73059696e-01 -4.47659075e-01
-7.15973154e-02 -6.97346270e-01 -7.68220603e-01 7.14598224e-02
3.05781156e-01 3.94796222e-01 8.20493624e-02 -7.10576236e-01
-1.16654289e+00 -3.39993298e-01 1.27682701e-01 -1.57392815e-01
-2.29731873e-01 7.05477297e-01 -7.60900199e-01 -3.98660302e-01
1.51365623e-01 -1.18525445e+00 -3.27609032e-01 -9.68640804e-01
-2.32972518e-01 -4.78151202e-01 6.02347910e-01 -1.03768575e+00
1.55161047e+00 -1.94859076e+00 -2.07777321e-01 -7.08829090e-02
-2.91293353e-01 4.72035319e-01 -4.11443830e-01 5.29208541e-01
-6.25427288e-04 4.94603872e-01 -2.58038223e-01 -2.52531528e-01
-3.44132841e-01 4.97199118e-01 1.22956879e-01 -3.90205462e-03
8.06667686e-01 8.96423697e-01 -8.83925080e-01 -7.41306067e-01
-1.93250105e-01 2.58799255e-01 -5.74945748e-01 4.49026704e-01
1.08155482e-01 3.26644421e-01 -7.04807520e-01 7.76528418e-01
7.80132651e-01 2.02075690e-01 6.89242721e-01 1.22851841e-01
-3.91072094e-01 7.14383006e-01 -7.82971978e-01 1.87691450e+00
-6.38578653e-01 2.02852979e-01 -1.30216964e-02 -6.48382723e-01
1.23835111e+00 3.35787505e-01 1.12997539e-01 -7.95355856e-01
4.38533239e-02 2.92253822e-01 2.48624906e-01 -3.96243572e-01
4.61844951e-01 -3.80148403e-02 -7.98869550e-01 2.59995043e-01
4.38050449e-01 -2.08864100e-02 5.60554147e-01 2.25866720e-01
1.44261515e+00 2.68879950e-01 5.02743065e-01 -6.25425518e-01
7.88482308e-01 2.41161436e-01 7.36962795e-01 5.63737035e-01
-1.64032206e-01 3.34833473e-01 8.08981240e-01 -4.14301366e-01
-7.41635203e-01 -8.24882984e-01 -1.70341372e-01 1.27988005e+00
-4.40737486e-01 -1.00986993e+00 -8.06403041e-01 -1.36516392e+00
-4.17410672e-01 7.19083428e-01 -5.56384206e-01 2.95640260e-01
-1.25564861e+00 -9.56850827e-01 8.43711317e-01 1.11944985e+00
4.44410890e-01 -1.11409616e+00 -3.51572245e-01 5.66150725e-01
-9.76059735e-02 -1.80048454e+00 -9.57465321e-02 7.20353484e-01
-9.81769443e-01 -9.91095245e-01 -4.13125083e-02 -1.12157941e+00
3.96093547e-01 -2.58281827e-01 1.28065252e+00 1.05411574e-01
2.32071951e-01 -3.44446123e-01 -1.05977368e+00 -3.71893764e-01
-8.18318069e-01 6.03111506e-01 -1.32892638e-01 -7.96273649e-01
4.66530949e-01 -2.02309132e-01 9.40967798e-02 3.22147720e-02
-4.92926806e-01 -2.85654366e-01 8.95698130e-01 8.40296745e-01
1.34524062e-01 -3.17496657e-01 5.32439411e-01 -1.37333560e+00
6.80424750e-01 -2.84996986e-01 -6.58565104e-01 2.83678979e-01
-8.35523009e-01 6.10509872e-01 8.63827407e-01 8.42434317e-02
-1.53351390e+00 4.37375456e-01 -9.10372615e-01 4.68441874e-01
-1.61705583e-01 7.77399600e-01 -7.32214525e-02 1.17536843e-01
2.16926634e-01 -3.54102224e-01 -6.42091870e-01 -8.66094470e-01
4.49634463e-01 8.19515705e-01 6.24455333e-01 -8.85249615e-01
4.17023957e-01 -1.01307191e-01 -1.25462994e-01 -4.12951112e-01
-9.19714868e-01 -4.22369123e-01 -1.46097815e+00 4.85035598e-01
9.93149400e-01 -9.47535694e-01 -9.56052467e-02 3.35349262e-01
-1.30028725e+00 -1.66132659e-01 2.17976674e-01 5.87828755e-01
-4.02333401e-02 4.55749243e-01 -1.22207737e+00 -4.98322457e-01
-5.72782695e-01 -9.23343897e-01 9.00754213e-01 1.12154394e-01
9.88665819e-02 -1.01534891e+00 4.79416579e-01 7.55166262e-02
8.44863579e-02 -3.70755382e-02 9.24288094e-01 -9.38884914e-01
1.23809457e-01 -3.81091803e-01 -1.85693577e-01 3.89026374e-01
1.29181743e-01 3.28811407e-01 -7.59888113e-01 -1.45810202e-01
-3.52329642e-01 -2.88986981e-01 7.91229188e-01 4.49932367e-03
5.13717890e-01 3.08365285e-01 -4.01287228e-01 3.50555331e-01
1.43988109e+00 1.65401220e-01 4.32657391e-01 3.40894997e-01
5.03205299e-01 5.07301569e-01 1.09249949e+00 1.64000750e-01
4.50191140e-01 4.25785691e-01 5.10179214e-02 2.17949655e-02
-1.04648070e-02 -3.01913708e-01 5.51834404e-01 1.34498060e+00
-1.65188015e-01 -1.69243827e-01 -1.16613817e+00 5.47468066e-01
-1.65173268e+00 -2.95097142e-01 -5.92755616e-01 1.70459473e+00
7.91159332e-01 7.89124608e-01 -4.33771685e-02 -5.37213385e-02
6.92241728e-01 -1.45419657e-01 1.62364691e-01 -1.16736650e+00
-5.63476421e-03 8.86357963e-01 5.79988420e-01 5.73822677e-01
-1.18228042e+00 1.69968379e+00 7.50319481e+00 2.67454177e-01
-1.02186608e+00 1.99637011e-01 2.23970756e-01 4.77435440e-01
2.34006435e-01 2.92630881e-01 -1.39961326e+00 9.96711180e-02
1.80375099e+00 -1.94395632e-02 -3.02246511e-01 9.48377490e-01
-1.51305377e-01 -1.13498874e-01 -9.42748427e-01 1.62676558e-01
-3.06131065e-01 -1.02208650e+00 -3.42515171e-01 2.72708945e-03
4.43517715e-01 4.91662085e-01 -5.19166946e-01 6.82198942e-01
7.39335001e-01 -7.15484500e-01 4.47868824e-01 -1.69861734e-01
7.33904719e-01 -8.66961837e-01 1.26044619e+00 4.16290164e-01
-1.38294101e+00 -1.14396170e-01 -5.82309425e-01 -4.75666910e-01
3.59410435e-01 4.43842858e-01 -7.28527725e-01 8.02183151e-01
7.82962084e-01 4.82410222e-01 -7.87048817e-01 4.32027370e-01
-7.86225855e-01 9.46279347e-01 -3.19101632e-01 -9.38924029e-02
4.52828377e-01 1.37194112e-01 -9.36522260e-02 1.90897131e+00
1.43369839e-01 3.92597795e-01 6.81366995e-02 5.58984019e-02
-1.22518077e-01 3.11186373e-01 -3.59654099e-01 7.82907903e-02
5.35955131e-01 1.48536599e+00 -7.33434737e-01 -3.60832751e-01
-9.24923122e-01 7.91148067e-01 9.06824529e-01 -2.87118435e-01
-8.19364965e-01 -4.98839200e-01 2.38871917e-01 -4.69387680e-01
6.63650632e-01 -4.50780183e-01 -1.96259409e-01 -1.19173348e+00
3.39969136e-02 -5.88103950e-01 6.87779665e-01 -2.44675219e-01
-1.18352044e+00 1.10810828e+00 5.39878085e-02 -7.96482325e-01
-4.09480304e-01 -1.00967300e+00 -6.87783062e-01 8.80647123e-01
-1.63084924e+00 -1.23688459e+00 1.40084118e-01 4.35915440e-01
5.51168144e-01 7.29594380e-02 1.28302193e+00 1.83005169e-01
-7.74890959e-01 7.79690504e-01 -1.00248441e-01 7.21127868e-01
8.00526917e-01 -1.31047702e+00 1.20494819e+00 1.26101041e+00
2.46648461e-01 6.41878784e-01 3.13691884e-01 -5.97413719e-01
-1.32422376e+00 -9.28705871e-01 1.60997808e+00 -5.69936097e-01
9.36520338e-01 -4.13973391e-01 -7.83287704e-01 9.09017503e-01
3.86205137e-01 1.79635331e-01 6.99503481e-01 6.10633433e-01
-4.53038871e-01 1.54869646e-01 -1.09441233e+00 -1.21717630e-02
9.91393089e-01 -3.43386859e-01 -1.05746961e+00 -9.20277834e-02
8.20183516e-01 -4.74118263e-01 -1.29017961e+00 4.76794511e-01
6.12055004e-01 -8.67537379e-01 4.12899345e-01 -9.14264500e-01
5.44776559e-01 1.73105180e-01 -2.53230184e-01 -1.32241654e+00
-6.14174306e-01 -2.87740469e-01 2.71350350e-02 1.38020301e+00
9.52945471e-01 -4.59755778e-01 6.50620461e-01 6.92778885e-01
-3.74497056e-01 -5.60442626e-01 -8.67297411e-01 -7.39725292e-01
6.96371496e-01 -5.32799482e-01 3.72350514e-01 6.27062201e-01
5.37691474e-01 7.12207496e-01 -1.89414788e-02 3.54621947e-01
2.65175998e-01 8.37620348e-02 4.46032524e-01 -1.19669855e+00
-1.21916912e-01 2.03217998e-01 -1.13648437e-01 -8.17246258e-01
6.16740704e-01 -7.98831999e-01 1.95855275e-01 -1.47983873e+00
1.55283213e-01 -5.32995284e-01 -6.14815235e-01 5.96475899e-01
-5.97955406e-01 1.26593605e-01 5.06078720e-01 -8.37686360e-02
-5.70370615e-01 5.37290704e-03 8.35822046e-01 2.80115634e-01
-2.23730132e-01 -9.68890935e-02 -9.13457155e-01 7.32798517e-01
1.03327870e+00 -9.38958108e-01 1.82557479e-01 -8.74160588e-01
4.92333956e-02 3.47563386e-01 -6.63394451e-01 -9.31139112e-01
8.34346786e-02 9.24400389e-02 1.75019488e-01 -4.91674602e-01
-1.23910151e-01 -4.30564284e-01 -5.61858535e-01 6.22249424e-01
-1.96282625e-01 5.94918787e-01 3.41034889e-01 2.34532878e-01
-4.26145047e-01 -3.31780493e-01 5.66025972e-01 -1.25767604e-01
-9.15110171e-01 -1.57774508e-01 -2.62982100e-01 1.01030678e-01
7.60034502e-01 4.21399236e-01 -4.18719262e-01 2.96348631e-01
-7.53690898e-01 1.37707308e-01 1.83227390e-01 6.48413837e-01
2.71306753e-01 -8.17100823e-01 -9.50043797e-01 4.07091707e-01
1.10955469e-01 -3.75847429e-01 -3.54276747e-01 4.84664142e-01
-6.66597843e-01 8.80008936e-01 -3.25155139e-01 -3.05688024e-01
-1.45592105e+00 4.63055223e-01 -2.39111573e-01 -9.63726342e-01
-5.19341409e-01 8.45108509e-01 -3.28716069e-01 -6.88549697e-01
-2.49151006e-01 -7.50734508e-01 -3.18208158e-01 -3.13176811e-01
3.90498966e-01 -4.01138775e-02 4.54551816e-01 -7.01523840e-01
-8.19278598e-01 4.94753927e-01 -3.45334172e-01 -1.53710216e-01
1.47633755e+00 8.57425034e-02 -1.57683671e-01 3.41728181e-02
1.12367702e+00 5.37067056e-01 -9.25379455e-01 1.82906985e-02
6.71787143e-01 -9.87176374e-02 -4.01362479e-02 -9.47264373e-01
-8.09481263e-01 7.96504617e-01 2.26613313e-01 1.41204566e-01
1.03955090e+00 3.27239454e-01 9.04942334e-01 7.11567283e-01
6.02928460e-01 -1.11790764e+00 -4.53950554e-01 1.00182986e+00
2.04852059e-01 -1.28180826e+00 -1.56157732e-01 -8.11148107e-01
-7.64837623e-01 1.36175025e+00 5.84756792e-01 -3.85836333e-01
8.28676105e-01 9.56545591e-01 3.98857772e-01 9.47479159e-02
-9.22104597e-01 -3.91510904e-01 -1.25600621e-01 4.96941447e-01
1.04965627e+00 2.51468360e-01 -1.03250432e+00 1.07790470e+00
-2.15400472e-01 -7.25853518e-02 5.86813033e-01 1.33775604e+00
-4.84691083e-01 -1.95768225e+00 -1.32758021e-01 1.86275765e-02
-9.77750123e-01 -4.45945054e-01 -3.18276584e-01 1.12613368e+00
-2.52955616e-01 1.41967225e+00 -9.15567353e-02 -5.63132882e-01
5.85213721e-01 3.03166419e-01 5.43503940e-01 -8.37363780e-01
-8.81663799e-01 1.15114942e-01 6.94686413e-01 -6.99722588e-01
-6.12744212e-01 -5.24752378e-01 -1.64467764e+00 -6.24513328e-02
-4.29529756e-01 5.46432376e-01 6.96729481e-01 9.90630031e-01
1.85163274e-01 3.82679731e-01 7.09520638e-01 -3.80043358e-01
-4.31253165e-01 -1.14532268e+00 -4.34427619e-01 1.84455007e-01
-1.86101392e-01 -2.41659209e-01 -8.50475952e-02 6.67567626e-02] | [10.343497276306152, 9.821457862854004] |
a2efee67-47bb-4013-8742-c76f60339b20 | jdcfc-a-japanese-dialogue-corpus-with-feature | null | null | https://aclanthology.org/L18-1461 | https://aclanthology.org/L18-1461.pdf | JDCFC: A Japanese Dialogue Corpus with Feature Changes | null | ['Daisuke Kawahara', 'Tetsuaki Nakamura'] | 2018-05-01 | jdcfc-a-japanese-dialogue-corpus-with-feature-1 | https://aclanthology.org/L18-1461 | https://aclanthology.org/L18-1461.pdf | lrec-2018-5 | ['dialogue-understanding'] | ['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] | [-7.270104885101318, 3.7413828372955322] |
90434c04-4a8b-4a5f-be31-842193a8cd7f | few-shot-text-classification-with-dual | 2209.15069 | null | https://arxiv.org/abs/2209.15069v1 | https://arxiv.org/pdf/2209.15069v1.pdf | Few-shot Text Classification with Dual Contrastive Consistency | In this paper, we explore how to utilize pre-trained language model to perform few-shot text classification where only a few annotated examples are given for each class. Since using traditional cross-entropy loss to fine-tune language model under this scenario causes serious overfitting and leads to sub-optimal generalization of model, we adopt supervised contrastive learning on few labeled data and consistency-regularization on vast unlabeled data. Moreover, we propose a novel contrastive consistency to further boost model performance and refine sentence representation. After conducting extensive experiments on four datasets, we demonstrate that our model (FTCC) can outperform state-of-the-art methods and has better robustness. | ['Jiawei Han', 'Liwen Sun'] | 2022-09-29 | null | null | null | null | ['few-shot-text-classification'] | ['natural-language-processing'] | [ 1.41880304e-01 -1.86380893e-01 -4.62336361e-01 -6.42496586e-01
-1.02330887e+00 -2.53863633e-01 6.03578269e-01 3.11890453e-01
-7.23631561e-01 9.50463533e-01 1.75239727e-01 -1.39137402e-01
2.04176113e-01 -4.75775212e-01 -1.66283637e-01 -3.36732835e-01
3.51261586e-01 2.19559640e-01 2.04670966e-01 -2.33474359e-01
2.81183153e-01 -1.92838967e-01 -1.04108393e+00 1.74926594e-01
1.06160367e+00 9.94253516e-01 1.09419554e-01 4.45009053e-01
-5.53574264e-01 1.05188954e+00 -6.49595141e-01 -5.54976285e-01
6.39128238e-02 -5.21263182e-01 -8.27533007e-01 1.77603498e-01
2.02494815e-01 -1.96998939e-01 -3.72852743e-01 1.15546989e+00
5.01264215e-01 6.41840339e-01 8.43484521e-01 -8.35473657e-01
-1.03942502e+00 7.52510488e-01 -7.40965903e-01 4.10136372e-01
2.35566005e-01 7.92847499e-02 1.16727948e+00 -1.16581416e+00
3.83525580e-01 1.13038051e+00 6.25777185e-01 7.82590926e-01
-1.07251823e+00 -6.59484506e-01 5.83078325e-01 1.19271189e-01
-1.44772327e+00 -4.94278520e-01 8.26183259e-01 -1.75938502e-01
9.90943730e-01 1.42230950e-02 1.05886601e-01 1.14594650e+00
1.93803608e-01 1.00294721e+00 9.28991914e-01 -7.04119086e-01
4.86820072e-01 2.46156648e-01 7.33974516e-01 7.06773579e-01
1.71661317e-01 -3.43926102e-01 -2.49438360e-01 -2.03327194e-01
1.06662340e-01 2.29833633e-01 -2.48070240e-01 -8.44164044e-02
-7.54676878e-01 9.77477133e-01 1.73946947e-01 4.12165254e-01
-6.52102157e-02 -1.03878617e-01 7.76094675e-01 2.80543506e-01
8.20010245e-01 2.70991236e-01 -6.49086177e-01 -1.32856622e-01
-1.03546250e+00 -1.59856781e-01 6.89212561e-01 9.02160466e-01
5.53934515e-01 3.19311731e-02 -5.93047619e-01 1.24712586e+00
1.33460477e-01 5.35395294e-02 1.02799249e+00 -4.60837573e-01
7.14962423e-01 6.38168216e-01 -1.32847115e-01 -6.26211226e-01
-2.12554514e-01 -5.25553644e-01 -9.98088598e-01 -4.79304165e-01
7.52926059e-03 -2.11479902e-01 -1.08756173e+00 1.47891438e+00
-6.63936883e-02 2.72979677e-01 1.85412839e-01 6.90691471e-01
7.19321787e-01 7.89252162e-01 4.16082889e-01 -5.85330904e-01
1.17946196e+00 -1.36805809e+00 -9.61500943e-01 -5.56450665e-01
1.05353284e+00 -4.40083206e-01 1.55043685e+00 2.19225451e-01
-8.12139034e-01 -5.26826680e-01 -1.09692585e+00 -8.43888745e-02
-3.17603260e-01 4.95563298e-02 4.01899487e-01 5.15982747e-01
-4.25548911e-01 4.17269289e-01 -5.88152468e-01 -3.05371493e-01
4.67221111e-01 -2.66233906e-02 -1.73832878e-01 -2.80043930e-01
-1.60666072e+00 9.47881460e-01 6.80995345e-01 -1.43769294e-01
-6.59478068e-01 -4.36485499e-01 -1.02656996e+00 1.27567202e-01
5.67231655e-01 -4.80882823e-01 1.25109541e+00 -5.26427686e-01
-1.51093745e+00 7.47109771e-01 -3.18023115e-01 -4.40242738e-01
4.58200365e-01 -3.19533557e-01 -4.73726749e-01 -5.88338077e-02
-1.98709834e-02 3.83223355e-01 5.88275909e-01 -1.09478688e+00
-3.73266160e-01 -1.09334789e-01 -1.77769050e-01 1.73461854e-01
-9.12856519e-01 4.66505624e-02 -5.26871860e-01 -8.33057106e-01
-2.80380100e-02 -5.01074016e-01 -4.64158267e-01 -1.27754956e-01
-3.90616685e-01 -4.04763728e-01 7.52475560e-01 -4.85536039e-01
1.82666898e+00 -1.81043041e+00 -1.64697602e-01 -2.05729559e-01
-3.99457067e-02 4.99957681e-01 -1.39252618e-01 3.81503135e-01
2.13399723e-01 2.65689254e-01 -4.70060796e-01 -5.77301085e-01
-1.50058710e-03 2.33545661e-01 -3.50078285e-01 3.13140184e-01
1.29894570e-01 7.87749410e-01 -9.39002097e-01 -7.21817017e-01
2.08503887e-01 2.66887218e-01 -3.55724871e-01 2.34459877e-01
-2.18894780e-01 3.66970487e-02 -6.94094837e-01 4.59336460e-01
6.51080012e-01 -5.38141787e-01 -9.94457025e-03 4.53049876e-02
3.68775070e-01 2.64828950e-01 -1.00065148e+00 1.84797859e+00
-7.01499760e-01 3.07908893e-01 -3.93645346e-01 -1.27046537e+00
1.07377958e+00 1.41602576e-01 2.81667598e-02 -4.38269824e-01
5.07440984e-01 -1.09984614e-01 -2.79879928e-01 -6.13772154e-01
4.49599206e-01 -4.12411541e-01 -1.65477440e-01 5.27731001e-01
2.89552122e-01 2.17328146e-02 3.31447452e-01 2.51792848e-01
9.64002907e-01 -7.82757625e-02 6.80858016e-01 -2.08539754e-01
8.49507868e-01 -2.62348354e-01 5.90188146e-01 8.74665439e-01
-6.14536524e-01 5.33424020e-01 2.16484144e-01 -1.93338215e-01
-8.87709796e-01 -6.47382259e-01 -2.67470121e-01 1.38884604e+00
1.13273352e-01 -4.52212065e-01 -5.57129681e-01 -1.18311369e+00
-3.21538925e-01 1.18768191e+00 -5.51610112e-01 -4.76756632e-01
-2.22024679e-01 -1.10963178e+00 4.17235970e-01 6.17375493e-01
6.77932858e-01 -6.77300572e-01 -5.02823442e-02 1.59643188e-01
-1.77619934e-01 -1.01994896e+00 -9.07878399e-01 3.30194443e-01
-8.77989709e-01 -7.50859439e-01 -6.43253803e-01 -1.05699217e+00
6.14459753e-01 4.04827386e-01 1.05066061e+00 1.01693459e-01
-2.01941162e-01 1.00211926e-01 -6.52109385e-01 -1.76471904e-01
-2.74250239e-01 1.96798250e-01 6.68289587e-02 -2.17667624e-01
8.48216414e-01 -1.48492441e-01 -3.55986297e-01 -3.24819703e-04
-9.02875304e-01 -1.47700921e-01 2.64582545e-01 1.31703150e+00
3.39396954e-01 8.97257775e-03 8.38504136e-01 -1.08361137e+00
1.01960409e+00 -5.29683113e-01 -2.47358814e-01 7.17339396e-01
-1.00948644e+00 1.08591229e-01 1.04786694e+00 -6.39171600e-01
-1.23218536e+00 -1.33537501e-01 1.21836020e-02 -4.00148630e-01
9.62820575e-02 7.25182414e-01 6.95076585e-02 2.83097684e-01
8.19282830e-01 3.56553614e-01 -3.25612158e-01 -4.39344585e-01
3.83815140e-01 1.02912104e+00 2.01488867e-01 -3.91763926e-01
5.30158699e-01 1.94055155e-01 -4.50578272e-01 -6.99718118e-01
-1.64293563e+00 -6.27901077e-01 -7.21311331e-01 1.52083814e-01
5.28260291e-01 -1.05602801e+00 -7.80231655e-02 1.28377676e-01
-1.02528834e+00 -1.72983408e-01 -5.51463552e-02 5.44230163e-01
-2.07342848e-01 6.76357508e-01 -7.75122285e-01 -1.13751411e+00
-7.33196616e-01 -7.17429280e-01 8.43856096e-01 1.59888685e-01
-9.63263512e-02 -1.31271291e+00 3.01651120e-01 2.68820018e-01
1.24986440e-01 -4.07691509e-01 6.02516532e-01 -1.22365141e+00
2.24278733e-01 -5.69464684e-01 -1.23588912e-01 6.51666105e-01
1.71481401e-01 -2.16808155e-01 -1.07864237e+00 -3.77732962e-01
1.96927786e-01 -8.38265002e-01 1.24770987e+00 1.41174182e-01
1.38831723e+00 -1.55782998e-01 -2.53372043e-01 2.92565465e-01
1.30769491e+00 1.16360537e-03 3.98602813e-01 1.79754347e-01
6.45424306e-01 5.08507729e-01 9.01886463e-01 6.20715916e-01
3.90114874e-01 3.75196964e-01 -3.81580621e-01 2.51363665e-01
2.05621332e-01 -3.41717929e-01 3.38594079e-01 1.24844289e+00
4.40935761e-01 -4.23679382e-01 -8.19119513e-01 2.50828654e-01
-1.98820388e+00 -9.51301873e-01 2.65948385e-01 2.02948141e+00
1.19329262e+00 4.69346970e-01 -1.71015188e-01 3.78666096e-03
9.62844193e-01 3.16945523e-01 -5.16800642e-01 -2.80288011e-01
-1.42428368e-01 1.76737800e-01 2.25358516e-01 6.70448601e-01
-1.25347519e+00 1.26453543e+00 7.11822367e+00 1.24716866e+00
-9.66079473e-01 2.67015934e-01 9.87847686e-01 -2.50859022e-01
-2.47090459e-01 2.49902625e-02 -9.33641136e-01 5.63535690e-01
9.92110968e-01 -4.43916172e-01 1.81915864e-01 9.40779507e-01
1.52449489e-01 2.13052370e-02 -8.84822488e-01 1.00669396e+00
4.78101194e-01 -1.07950580e+00 1.38953999e-01 -5.68886220e-01
8.76904130e-01 -1.41656682e-01 -3.22615087e-01 9.54469919e-01
3.89641196e-01 -8.15612972e-01 2.01196238e-01 4.45365727e-01
5.70209742e-01 -6.42152727e-01 8.13807964e-01 7.47543514e-01
-1.06919348e+00 -1.74186572e-01 -8.60155702e-01 -2.19773039e-01
1.60765827e-01 5.62827289e-01 -5.49060404e-01 3.03820491e-01
2.48957068e-01 6.62194848e-01 -5.98850191e-01 8.45338523e-01
-1.84237078e-01 6.77339256e-01 1.14645511e-01 -5.53525090e-01
3.08110893e-01 -7.99629092e-02 1.25007913e-01 1.41665018e+00
1.30257383e-01 3.36526155e-01 5.27932167e-01 5.97102225e-01
-3.27832431e-01 5.21573544e-01 -3.86068314e-01 7.42946863e-02
5.21872759e-01 1.23898101e+00 -4.94578719e-01 -7.40553677e-01
-6.93404377e-01 1.07639229e+00 8.00755680e-01 4.37433153e-01
-6.02993131e-01 -6.30673647e-01 1.36127055e-01 -3.95300746e-01
1.16393983e-01 4.17494681e-03 -4.07934517e-01 -1.76538813e+00
3.86352977e-03 -4.00731236e-01 6.67350650e-01 -5.79177856e-01
-1.77973807e+00 6.05760574e-01 -1.78898543e-01 -1.30184054e+00
-2.30767518e-01 -5.45087099e-01 -7.92558253e-01 7.68588722e-01
-1.57577264e+00 -9.19554949e-01 -8.36482048e-02 3.41124922e-01
1.02627301e+00 -3.67715240e-01 8.59247863e-01 1.11373693e-01
-8.41687918e-01 8.60545874e-01 3.45158666e-01 2.09490851e-01
8.94694507e-01 -1.08806610e+00 2.83375621e-01 7.71972716e-01
2.25244425e-02 6.76775515e-01 6.51610732e-01 -6.46359980e-01
-9.62071955e-01 -1.16853333e+00 8.62105966e-01 -2.75994271e-01
8.10789347e-01 -2.75512189e-01 -1.23471570e+00 5.96286714e-01
2.47728243e-01 3.43766600e-01 1.05283916e+00 1.00192837e-01
-4.27329332e-01 -3.36554907e-02 -1.08789635e+00 6.08573675e-01
8.33917141e-01 -6.38801754e-01 -9.89946723e-01 3.70597959e-01
8.20665121e-01 8.46267678e-03 -6.46597266e-01 3.66259038e-01
2.56441772e-01 -4.57435936e-01 5.05024731e-01 -9.96934772e-01
4.17724878e-01 1.66473463e-02 -1.62693888e-01 -1.32985723e+00
-2.60722846e-01 -4.41850275e-01 -3.61986071e-01 1.32584989e+00
5.55083334e-01 -3.91245633e-01 7.25516081e-01 7.79868484e-01
-3.73583026e-02 -8.68500292e-01 -7.52629399e-01 -9.66456234e-01
3.62799644e-01 -4.83738422e-01 8.80705267e-02 1.28144717e+00
6.19245052e-01 8.65102947e-01 -7.03224242e-01 -3.23554516e-01
5.07753313e-01 -7.91053325e-02 4.53547955e-01 -1.10333073e+00
-2.98336416e-01 -3.49391788e-01 -1.27891943e-01 -1.25523591e+00
6.06455982e-01 -9.30424869e-01 2.20759943e-01 -1.43082047e+00
7.64877975e-01 -5.48883043e-02 -7.04876065e-01 4.55046564e-01
-8.30924332e-01 8.82608220e-02 -4.30029333e-02 1.34926096e-01
-1.12297428e+00 1.03656983e+00 1.13600242e+00 -3.22514713e-01
-1.85720518e-01 -1.21590637e-01 -7.29437232e-01 5.84060371e-01
7.19642043e-01 -4.69091564e-01 -4.69402850e-01 -2.15281963e-01
-9.00081322e-02 -4.26522307e-02 -2.08955199e-01 -6.86290741e-01
2.62349963e-01 -2.30176032e-01 3.25334013e-01 -4.11388546e-01
1.36842623e-01 -5.66225648e-01 -8.11073959e-01 4.24171001e-01
-9.63808894e-01 -3.46311331e-01 6.28073215e-02 7.44516909e-01
-2.42661506e-01 -8.62237632e-01 8.65421951e-01 -1.67373821e-01
-6.15577996e-01 3.76215100e-01 -1.68765336e-01 4.54935700e-01
9.91594076e-01 1.50818855e-01 -3.86753500e-01 -4.35470223e-01
-5.51970720e-01 5.72805703e-01 2.09286675e-01 4.23545003e-01
5.29053390e-01 -1.41136658e+00 -6.50311291e-01 -5.42898588e-02
4.37610716e-01 -2.96930254e-01 4.14807409e-01 4.74013507e-01
-1.28979743e-01 4.19772506e-01 3.40624899e-01 -3.13595533e-01
-1.01463568e+00 8.83305311e-01 2.06207559e-01 -4.28672850e-01
-4.91263509e-01 8.12465250e-01 5.09470329e-02 -5.59258461e-01
4.40121114e-01 -9.28909183e-02 -3.04407179e-01 -3.89012098e-02
8.51544499e-01 2.07105175e-01 -3.99084091e-02 -3.39236110e-01
-3.06704760e-01 3.70340943e-01 -6.46450222e-01 4.27082926e-02
9.56579566e-01 -3.94108742e-01 1.86878607e-01 8.53448033e-01
1.44762802e+00 -3.40969503e-01 -1.28093064e+00 -5.59003174e-01
1.18395016e-01 -3.73248726e-01 2.81630635e-01 -7.18448877e-01
-6.59431636e-01 1.03658223e+00 4.02331084e-01 1.97034404e-01
9.43964958e-01 -2.26616740e-01 8.03733587e-01 9.10288095e-01
9.84504595e-02 -1.60237467e+00 3.19558382e-01 8.43701184e-01
4.13460940e-01 -1.76729512e+00 2.21163005e-01 -2.82736897e-01
-1.00450706e+00 1.11670768e+00 9.17178214e-01 -7.16980174e-02
7.75240541e-01 5.87376840e-02 7.70001188e-02 1.86029449e-01
-1.15178204e+00 -1.15573630e-01 3.60722512e-01 1.61698967e-01
7.77541041e-01 -2.16707379e-01 -6.42195225e-01 9.18572187e-01
1.43118694e-01 1.53736517e-01 3.95233274e-01 1.07151127e+00
-7.80276477e-01 -9.58859682e-01 1.08292699e-02 7.36804724e-01
-5.38693309e-01 -3.64342421e-01 -1.74531013e-01 4.98178035e-01
-4.06169474e-01 1.07732677e+00 -9.67806056e-02 -2.98756033e-01
1.72079608e-01 4.60435480e-01 2.72621483e-01 -9.47961748e-01
-3.26166660e-01 1.25213504e-01 5.79659306e-02 1.22121826e-01
-2.45292544e-01 -2.58402288e-01 -1.27024698e+00 -5.59445992e-02
-7.11834490e-01 2.46829107e-01 1.45040601e-01 1.24816358e+00
1.18871644e-01 4.94575679e-01 8.03707719e-01 -4.39910412e-01
-1.19201958e+00 -1.44122887e+00 -7.21122682e-01 4.04864371e-01
1.27528831e-01 -5.05249560e-01 -6.25965059e-01 -1.44187966e-03] | [10.721917152404785, 7.461348056793213] |
19cc9dc8-0fe7-4e25-a9fa-4ee10ccd94dd | road-extraction-by-deep-residual-u-net | 1711.10684 | null | http://arxiv.org/abs/1711.10684v1 | http://arxiv.org/pdf/1711.10684v1.pdf | Road Extraction by Deep Residual U-Net | Road extraction from aerial images has been a hot research topic in the field
of remote sensing image analysis. In this letter, a semantic segmentation
neural network which combines the strengths of residual learning and U-Net is
proposed for road area extraction. The network is built with residual units and
has similar architecture to that of U-Net. The benefits of this model is
two-fold: first, residual units ease training of deep networks. Second, the
rich skip connections within the network could facilitate information
propagation, allowing us to design networks with fewer parameters however
better performance. We test our network on a public road dataset and compare it
with U-Net and other two state of the art deep learning based road extraction
methods. The proposed approach outperforms all the comparing methods, which
demonstrates its superiority over recently developed state of the arts. | ['Qingjie Liu', 'Zhengxin Zhang', 'Yunhong Wang'] | 2017-11-29 | null | null | null | null | ['skin-cancer-segmentation', 'lung-nodule-segmentation'] | ['medical', 'medical'] | [ 4.09851968e-01 2.14931831e-01 -1.79947823e-01 -3.95159602e-01
-7.98821449e-02 -2.42841169e-01 3.68393660e-01 -4.58540976e-01
-5.00942707e-01 8.54532778e-01 1.07303159e-02 -4.80601013e-01
-8.52882117e-02 -1.61349893e+00 -6.12827957e-01 -5.07489562e-01
-1.19629819e-02 -1.55571491e-01 5.86266756e-01 -1.71537191e-01
5.49714379e-02 7.62644768e-01 -1.36275673e+00 -2.46035740e-01
1.21309125e+00 1.15733862e+00 3.12786072e-01 3.84651899e-01
-1.49721622e-01 8.70444298e-01 -1.51065573e-01 6.02739267e-02
4.64659423e-01 -1.88293159e-01 -1.03273594e+00 -2.06598610e-01
2.77413726e-01 -7.11327851e-01 -6.70729280e-01 9.48312700e-01
5.33087373e-01 2.41496369e-01 7.04200029e-01 -7.58229434e-01
-5.24693847e-01 7.41445303e-01 -6.66419029e-01 1.76788506e-03
-3.53830397e-01 -2.20450819e-01 9.66493905e-01 -5.34104407e-01
3.51844281e-01 1.00119257e+00 9.03008461e-01 1.92974746e-01
-8.42104018e-01 -7.23792255e-01 1.98336557e-01 -2.02305857e-02
-1.52689505e+00 -4.16671075e-02 7.42280602e-01 -4.71267700e-01
7.09060848e-01 -1.20858707e-01 5.26798487e-01 3.32112253e-01
-5.74174635e-02 1.10813379e+00 9.76568341e-01 -8.10037330e-02
-3.77189256e-02 -6.82258457e-02 3.29359591e-01 8.39994490e-01
2.36302957e-01 1.04908403e-02 3.13108116e-01 3.43914181e-01
1.01971626e+00 2.69026637e-01 -2.88215995e-01 -3.55753452e-02
-8.18149269e-01 8.39184821e-01 1.27235162e+00 3.31194729e-01
-5.24267912e-01 2.64712960e-01 1.72232285e-01 -7.87274987e-02
6.47806346e-01 3.28334391e-01 -5.08921266e-01 5.28524935e-01
-1.23072517e+00 6.82418048e-02 3.93760026e-01 7.47603416e-01
9.85497832e-01 3.64474833e-01 -8.88713747e-02 8.37334394e-01
3.07910770e-01 5.19032896e-01 1.63759097e-01 -6.79050982e-01
4.18186456e-01 9.06754673e-01 -1.63069427e-01 -9.74151254e-01
-6.76030815e-01 -4.12268490e-01 -1.11046422e+00 3.75017911e-01
1.63808092e-01 -6.91813529e-01 -1.25869250e+00 1.28767347e+00
-3.80908810e-02 2.35340580e-01 1.69055775e-01 8.58997464e-01
1.38707793e+00 8.28317225e-01 9.40366462e-03 5.89298666e-01
1.02864289e+00 -1.08705080e+00 -4.38221246e-01 -3.71632397e-01
3.29373479e-01 -3.56714875e-01 5.75776756e-01 -8.65913928e-02
-5.21373093e-01 -8.07470560e-01 -1.24391067e+00 5.28456531e-02
-6.92243516e-01 6.25582993e-01 1.00087821e+00 4.73711282e-01
-1.08191323e+00 8.94888341e-01 -5.37824690e-01 -4.78382945e-01
1.01060009e+00 2.99429864e-01 -1.20957419e-01 1.09549321e-01
-1.37971854e+00 6.13069832e-01 7.51948714e-01 6.86787844e-01
-6.76807404e-01 -2.67884701e-01 -9.58535552e-01 1.91553935e-01
3.75007451e-01 -5.06233513e-01 8.84104669e-01 -7.37182319e-01
-1.52842033e+00 5.80558479e-01 7.93792829e-02 -4.86192882e-01
3.97879124e-01 -4.95555311e-01 -2.29349628e-01 1.82399258e-01
5.37567884e-02 1.14921045e+00 5.08314371e-01 -1.09178221e+00
-8.38810921e-01 -2.56181121e-01 1.58034086e-01 8.77764747e-02
-8.35628435e-02 -4.42973405e-01 -2.74293333e-01 -6.20643616e-01
3.74651402e-01 -7.76197314e-01 -5.71821213e-01 -7.77232796e-02
-6.33849323e-01 -1.03864796e-01 1.14282250e+00 -6.87500298e-01
1.22349656e+00 -1.76744044e+00 -2.55994618e-01 3.74665767e-01
1.89124882e-01 7.55045474e-01 -1.84109345e-01 1.34963527e-01
7.10858172e-03 4.10748184e-01 -6.82224989e-01 4.17235762e-01
-3.78380507e-01 1.49350036e-02 -2.13108152e-01 2.37979308e-01
4.43414837e-01 1.16941345e+00 -6.75539911e-01 -3.93554211e-01
4.15473104e-01 5.04117429e-01 -1.83355257e-01 5.19656800e-02
-3.77881248e-03 2.77149290e-01 -7.97952056e-01 7.62096286e-01
9.98260140e-01 3.11843567e-02 -7.43128136e-02 -3.70031059e-01
-4.70150352e-01 1.13543019e-01 -1.11641002e+00 1.12099421e+00
-3.78786057e-01 9.46742296e-01 -2.40062717e-02 -1.10282362e+00
1.28533268e+00 1.40435517e-01 3.70326459e-01 -6.08963728e-01
3.26175421e-01 7.54672140e-02 -1.18920818e-01 -4.35532719e-01
5.36444426e-01 3.01217139e-01 3.20274532e-01 9.24682617e-02
-1.57339051e-01 -1.69949055e-01 -3.42479721e-03 -1.55448169e-01
6.44048274e-01 5.37228107e-01 9.37947556e-02 -3.05873364e-01
5.95679164e-01 -2.42794864e-02 4.87563670e-01 6.51349247e-01
-1.88715383e-01 2.95087904e-01 3.22796762e-01 -6.29325569e-01
-6.88880265e-01 -7.27423787e-01 -4.36033666e-01 6.96360886e-01
4.64027554e-01 9.44038630e-02 -8.41555417e-01 -7.09813356e-01
-2.15386525e-02 1.69694319e-01 -5.49873650e-01 3.19275051e-01
-6.08052254e-01 -9.47022021e-01 8.35170746e-01 9.56233978e-01
1.54257059e+00 -1.24609375e+00 -6.67920530e-01 2.17727378e-01
-1.98826820e-01 -1.21727479e+00 1.64105237e-01 5.07138716e-03
-1.25519884e+00 -1.12647498e+00 -8.08893919e-01 -8.73086393e-01
6.89166903e-01 6.66540682e-01 8.96661818e-01 -2.83721313e-02
-2.42753565e-01 -2.64910668e-01 -3.29224437e-01 -4.78745908e-01
9.41607654e-02 6.60074890e-01 -6.04866207e-01 -1.85866147e-01
3.56438786e-01 -4.61265296e-01 -7.78499365e-01 3.71383131e-01
-9.21660662e-01 8.88831764e-02 9.91663516e-01 7.59029031e-01
5.75404346e-01 5.64232409e-01 6.19817853e-01 -1.08605719e+00
3.58540297e-01 -4.44456756e-01 -7.81707942e-01 -3.62095907e-02
-6.31928563e-01 -8.55201632e-02 4.13252950e-01 2.11757049e-01
-1.13098001e+00 3.31070155e-01 -4.07245725e-01 2.56576359e-01
-3.65171283e-01 6.08959198e-01 -1.63230479e-01 -2.89690524e-01
3.75398606e-01 -5.08887582e-02 -1.43539295e-01 -4.41330940e-01
4.63104904e-01 9.31710005e-01 3.99558425e-01 -7.44390115e-02
7.53326416e-01 4.74573940e-01 1.14626743e-01 -1.14550614e+00
-9.61810410e-01 -5.20549357e-01 -9.57483232e-01 -1.84682846e-01
1.08676970e+00 -1.06643403e+00 -5.32006681e-01 8.68594766e-01
-9.41056788e-01 -5.45069456e-01 -4.06368934e-02 4.69483733e-01
-2.71146595e-01 2.11589813e-01 -4.58061963e-01 -7.73075819e-01
-6.71709597e-01 -9.74085152e-01 8.23746860e-01 8.16102087e-01
3.25455397e-01 -9.40488696e-01 -2.88723260e-01 2.51895219e-01
6.06664300e-01 5.27515411e-01 4.94067788e-01 -1.15280241e-01
-7.76094794e-01 -1.07137427e-01 -1.01009071e+00 6.56037867e-01
2.33339399e-01 2.30351940e-01 -1.15076470e+00 1.84067681e-01
-5.37504494e-01 -1.35284722e-01 1.64396226e+00 9.14260685e-01
1.29366076e+00 -1.29218772e-01 -5.77386916e-01 8.61026585e-01
1.74942183e+00 1.92334995e-01 1.09026694e+00 6.14421725e-01
1.00811934e+00 7.23089576e-01 5.26962161e-01 8.32618997e-02
4.50263143e-01 2.02121526e-01 7.52183199e-01 -9.25335228e-01
8.35306495e-02 -7.86360204e-02 -7.62077654e-03 2.00074479e-01
-5.02084196e-01 -6.79268464e-02 -9.88142788e-01 9.13676739e-01
-1.95180106e+00 -1.02473104e+00 -4.10230786e-01 1.76123190e+00
3.98036689e-01 1.74761161e-01 1.22911386e-01 1.39842480e-01
6.64500117e-01 4.25382793e-01 -5.46214879e-01 -1.65394336e-01
-2.26310015e-01 4.49608862e-01 1.12999499e+00 3.31373692e-01
-1.82105553e+00 1.45532334e+00 6.24186802e+00 6.45368040e-01
-1.21470773e+00 -2.30964452e-01 7.03843474e-01 6.09081626e-01
8.82168487e-02 -1.00477763e-01 -8.58664453e-01 5.33065051e-02
5.36616564e-01 4.89211768e-01 2.80868318e-02 7.97802925e-01
2.33485878e-01 -3.12005669e-01 -3.84080470e-01 4.14228678e-01
-3.02485824e-01 -1.36462331e+00 -5.48912250e-02 -4.20842506e-02
9.02974367e-01 6.46799803e-01 -2.48490453e-01 2.78947204e-01
5.27277529e-01 -1.21664023e+00 3.21422845e-01 3.74784857e-01
8.40766788e-01 -9.44837809e-01 1.30503619e+00 1.21816285e-01
-1.68476462e+00 -1.49001375e-01 -6.10157311e-01 -2.23183930e-01
1.53960258e-01 6.69588327e-01 -5.93681455e-01 7.85867095e-01
7.54749179e-01 1.11238062e+00 -4.56202984e-01 1.06776690e+00
-5.29024959e-01 8.52575004e-01 -2.28761122e-01 2.11175218e-01
8.21060538e-01 -6.37243867e-01 2.15829596e-01 1.32057714e+00
1.65990382e-01 1.62614778e-01 2.97189534e-01 9.62578833e-01
-1.74162373e-01 7.42450804e-02 -1.03148615e+00 -3.39287962e-03
3.81577700e-01 1.46961510e+00 -9.20923769e-01 -2.58391857e-01
-3.96815777e-01 6.73368573e-01 7.06630945e-02 3.80920112e-01
-6.95830166e-01 -1.09762764e+00 5.65667093e-01 -4.27668430e-02
4.52544063e-01 -2.52807260e-01 -4.18200135e-01 -6.96551800e-01
-2.39580363e-01 -1.56837940e-01 1.04613386e-01 -6.16579413e-01
-8.49586666e-01 8.33822131e-01 -1.91302836e-01 -1.20889139e+00
3.16576600e-01 -8.16874743e-01 -9.23913419e-01 9.15448606e-01
-2.42712355e+00 -1.45929885e+00 -7.44424284e-01 2.35504016e-01
4.11765963e-01 -1.26453534e-01 5.08707523e-01 2.27409571e-01
-8.34565401e-01 8.51309001e-02 7.21061900e-02 7.47391939e-01
1.83141679e-01 -1.16604745e+00 7.64938235e-01 1.08149362e+00
-2.27996960e-01 3.67619365e-01 -1.17725782e-01 -6.18193388e-01
-5.62806547e-01 -1.73458469e+00 5.68256915e-01 1.56212732e-01
5.17105222e-01 1.00959688e-01 -8.79944861e-01 4.74913478e-01
8.08301196e-02 -1.66079596e-01 4.31386948e-01 -1.13137759e-01
1.76558383e-02 -3.32225740e-01 -1.00193918e+00 4.21279222e-01
9.31413710e-01 -2.70553738e-01 -4.01074558e-01 4.99802381e-02
6.32843852e-01 -1.47616386e-01 -7.91385949e-01 7.60895610e-01
6.76308513e-01 -8.65491927e-01 9.58624840e-01 -3.93926539e-02
6.47825360e-01 -4.31961864e-01 -4.18186858e-02 -1.30917597e+00
-5.07072031e-01 -1.06467605e-01 2.72033691e-01 1.24099684e+00
5.50204098e-01 -8.03394675e-01 8.23807299e-01 2.36063041e-02
-8.45734924e-02 -8.08276772e-01 -2.12609470e-01 -6.19161487e-01
1.47089735e-01 -2.31980402e-02 7.25266814e-01 8.05685461e-01
-6.72574759e-01 5.18070519e-01 -3.62866342e-01 2.39137411e-01
4.77407187e-01 1.21595316e-01 5.64910829e-01 -1.72717285e+00
3.76854837e-01 -5.99612236e-01 -2.30954140e-01 -1.35962307e+00
2.84024775e-01 -8.01490963e-01 5.49155809e-02 -2.22689748e+00
-1.66383535e-01 -6.81324184e-01 -2.15824738e-01 8.45735312e-01
-2.59249419e-01 4.79678363e-01 -2.93451548e-02 1.38428107e-01
3.29072657e-03 7.25809872e-01 1.38762426e+00 -3.38897258e-01
-5.92412829e-01 3.54182303e-01 -6.49796069e-01 9.74681735e-01
1.26520252e+00 -1.23001494e-01 -4.16418970e-01 -5.63745618e-01
2.31269244e-02 -3.42141002e-01 2.52745301e-01 -1.13424158e+00
3.05520203e-02 5.21679558e-02 4.03103441e-01 -1.18681276e+00
-3.38969767e-01 -8.90822053e-01 -3.14412236e-01 4.09134567e-01
1.70743633e-02 -3.82535517e-01 3.55759650e-01 3.63047630e-01
-4.70075548e-01 -1.76726997e-01 8.56376112e-01 -2.70979583e-01
-1.17361629e+00 4.74445432e-01 -4.86381114e-01 -2.55351335e-01
9.19718683e-01 -5.88391960e-01 -3.93365592e-01 -1.16522036e-01
-3.68057728e-01 5.15527666e-01 1.55842826e-02 2.88140416e-01
7.19642520e-01 -1.04477322e+00 -7.57541955e-01 7.68982619e-02
-7.49666095e-02 5.66447675e-01 1.55459270e-01 6.84068978e-01
-1.04227686e+00 5.75146854e-01 -4.59500015e-01 -4.93694395e-01
-9.25277412e-01 2.37934813e-02 4.37385976e-01 -1.82009816e-01
-8.64319384e-01 5.77488959e-01 1.86990604e-01 -7.88036704e-01
1.06260069e-02 -3.62011462e-01 -8.39816391e-01 1.32872120e-01
2.90497929e-01 7.31588781e-01 -1.97902799e-01 -6.03988290e-01
-3.31375450e-01 8.73824120e-01 2.32158244e-01 3.45031977e-01
1.60395384e+00 -8.17873552e-02 -2.15841621e-01 6.60452875e-04
8.78956676e-01 -3.48330140e-01 -1.28730500e+00 -3.79341930e-01
-1.54229626e-01 -1.31488219e-01 6.72088563e-01 -6.35554135e-01
-1.58263075e+00 9.76118863e-01 6.54901445e-01 1.56950485e-02
1.20885360e+00 -3.41315717e-01 1.00765824e+00 6.69929147e-01
1.04229197e-01 -1.23737788e+00 -3.12495977e-01 7.11690307e-01
4.26112950e-01 -1.51910865e+00 9.50408801e-02 -5.51100314e-01
-3.82620960e-01 1.18403065e+00 7.02601850e-01 -2.45178506e-01
8.61848116e-01 1.18371420e-01 2.11034864e-01 -3.68618816e-01
1.70287862e-01 -1.01661491e+00 3.60761046e-01 5.19316494e-01
5.33695221e-01 3.10440660e-01 -1.81925535e-01 2.59325683e-01
7.67112225e-02 2.27855265e-01 4.62336838e-01 7.89456427e-01
-7.71260381e-01 -8.01546514e-01 -4.12605815e-02 6.08710527e-01
-3.83412808e-01 -2.81900018e-01 -4.92216289e-01 9.53983784e-01
1.57212257e-01 8.77672791e-01 -8.42258241e-03 -4.49111015e-01
3.99330974e-01 -3.47658932e-01 6.26699552e-02 -4.28546995e-01
-3.26925427e-01 -1.78039908e-01 8.01726505e-02 -4.89823788e-01
-8.81652474e-01 -1.60710007e-01 -1.27760196e+00 -2.19095930e-01
-6.60139143e-01 -1.25057429e-01 6.93025708e-01 9.02058125e-01
2.35888883e-01 8.41606140e-01 7.78897464e-01 -9.41917062e-01
1.27362043e-01 -9.90467548e-01 -6.62542403e-01 -1.79696798e-01
2.51670480e-01 -6.53263211e-01 2.99046710e-02 -9.63258073e-02] | [9.22656536102295, -1.3573840856552124] |
0e39e903-7c2f-4ece-bb93-26415b6eecb4 | conflict-based-search-for-explainable-multi | 2202.09930 | null | https://arxiv.org/abs/2202.09930v2 | https://arxiv.org/pdf/2202.09930v2.pdf | Conflict-Based Search for Explainable Multi-Agent Path Finding | In the Multi-Agent Path Finding (MAPF) problem, the goal is to find non-colliding paths for agents in an environment, such that each agent reaches its goal from its initial location. In safety-critical applications, a human supervisor may want to verify that the plan is indeed collision-free. To this end, a recent work introduces a notion of explainability for MAPF based on a visualization of the plan as a short sequence of images representing time segments, where in each time segment the trajectories of the agents are disjoint. Then, the explainable MAPF problem asks for a set of non-colliding paths that admits a short-enough explanation. Explainable MAPF adds a new difficulty to MAPF, in that it is NP-hard with respect to the size of the environment, and not just the number of agents. Thus, traditional MAPF algorithms are not equipped to directly handle explainable-MAPF. In this work, we adapt Conflict Based Search (CBS), a well-studied algorithm for MAPF, to handle explainable MAPF. We show how to add explainability constraints on top of the standard CBS tree and its underlying A* search. We examine the usefulness of this approach and, in particular, the tradeoff between planning time and explainability. | ['Morteza Lahijanian', 'Shaull Almagor', 'Justin Kottinger'] | 2022-02-20 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [ 1.32823616e-01 7.44432211e-01 -5.62465154e-02 -9.94984582e-02
-3.53207499e-01 -9.12411034e-01 4.50025439e-01 5.90285838e-01
-1.32959159e-02 9.54011798e-01 -2.38042861e-01 -5.60330927e-01
-8.56241643e-01 -9.59187746e-01 -7.24347711e-01 -4.78411287e-01
-5.38285196e-01 1.03998411e+00 5.92049837e-01 -3.32616299e-01
2.65021712e-01 7.24132955e-01 -1.20639610e+00 -8.50922614e-02
7.65416622e-01 2.99016774e-01 3.65678787e-01 7.44140148e-01
5.53369932e-02 4.36059475e-01 -5.47792792e-01 3.26607414e-02
2.05633715e-01 -5.73837101e-01 -1.31017196e+00 2.13241264e-01
-2.26668999e-01 -8.90916735e-02 -2.13258430e-01 9.04920697e-01
-3.67188394e-01 1.24178752e-01 4.82153028e-01 -2.41130614e+00
-1.57116517e-01 7.08785832e-01 -3.53203773e-01 -4.85188812e-02
4.78672355e-01 1.76775858e-01 7.53874481e-01 -2.41543934e-01
7.74349213e-01 1.00249767e+00 3.57529223e-01 5.10682821e-01
-1.24712265e+00 -7.17714727e-02 5.22160411e-01 5.79214513e-01
-1.20244992e+00 1.02213912e-01 1.78461224e-01 -2.96885878e-01
1.25028837e+00 6.51553333e-01 7.56512761e-01 4.68910992e-01
5.34660697e-01 2.11021513e-01 1.02519131e+00 -4.57829535e-01
5.64204693e-01 -1.00950420e-01 3.22177380e-01 7.89459765e-01
4.34604615e-01 4.69266325e-01 -3.17992091e-01 -2.29758054e-01
7.34789073e-01 1.10819377e-01 -2.89003730e-01 -6.60101891e-01
-1.41604626e+00 8.88886273e-01 2.49372214e-01 3.33317161e-01
-1.67823061e-01 3.53426099e-01 2.02629752e-02 3.00274819e-01
-2.00785771e-01 8.11712146e-01 -2.35677436e-01 -5.49749210e-02
-3.66197586e-01 5.85602522e-01 9.78410363e-01 1.03586733e+00
1.01529264e+00 -3.88177246e-01 9.36728865e-02 -4.90530312e-01
1.25174463e-01 4.46918666e-01 -3.49047363e-01 -1.32397985e+00
2.30620936e-01 6.46607280e-01 3.94254327e-01 -1.03083062e+00
-7.20790744e-01 -6.66248724e-02 -4.41113025e-01 8.29893887e-01
4.91213828e-01 3.11167598e-01 -5.29243171e-01 1.91719830e+00
5.39642632e-01 3.42827380e-01 1.80183202e-01 1.09563911e+00
2.69926600e-02 8.53626192e-01 -3.58482867e-01 -4.74187255e-01
1.24057317e+00 -1.36643684e+00 -4.95175540e-01 -3.26521486e-01
5.21791279e-01 -2.08723083e-01 8.15501571e-01 3.76729995e-01
-1.13411117e+00 1.08253453e-02 -1.07398272e+00 3.96198750e-01
-1.98739514e-01 -6.69846654e-01 6.73256874e-01 2.22642615e-01
-1.15812218e+00 4.21743304e-01 -7.92371929e-01 -3.81273866e-01
-2.06771269e-01 5.29640317e-01 -7.40016699e-01 -3.28077823e-01
-7.72882521e-01 9.24966693e-01 5.52346349e-01 -1.27044871e-01
-1.06960976e+00 -1.28900558e-01 -8.07228804e-01 2.60023057e-01
1.35980201e+00 -7.63309479e-01 1.36526489e+00 -4.61087018e-01
-9.83432293e-01 2.90221810e-01 -4.53125745e-01 -5.49686313e-01
3.78946394e-01 2.85641968e-01 -1.09086566e-01 9.68994498e-02
4.62608457e-01 1.82718202e-01 4.44692254e-01 -1.29662347e+00
-7.79449940e-01 -4.24967498e-01 7.19196379e-01 7.56093860e-02
4.73523915e-01 -3.02846193e-01 -2.33700708e-01 1.03708439e-01
1.82668641e-01 -1.35791230e+00 -8.20846498e-01 -1.70079187e-01
-7.17467785e-01 -3.56382310e-01 6.32842779e-01 -6.81229010e-02
1.02416766e+00 -1.96446228e+00 3.66829753e-01 6.40881121e-01
5.58897913e-01 -2.74799407e-01 -2.70804822e-01 9.32492673e-01
-2.05027983e-02 3.69552970e-01 -3.78930748e-01 7.93087669e-03
2.85552204e-01 7.47118950e-01 -3.01684827e-01 5.21093786e-01
-2.50811204e-02 6.02429450e-01 -1.05400252e+00 -2.97239989e-01
1.30459100e-01 -3.42221737e-01 -5.09274662e-01 1.57633126e-02
-5.09061217e-01 4.37899113e-01 -5.69063604e-01 2.79018916e-02
6.26249433e-01 -1.18244551e-01 2.05939174e-01 6.93677664e-01
-6.95670187e-01 8.35589692e-02 -1.43969738e+00 1.61138690e+00
-8.19257647e-02 4.40875620e-01 5.40608093e-02 -7.21112669e-01
5.64873278e-01 1.62713349e-01 4.63340640e-01 -3.95269454e-01
-1.76237181e-01 3.27937841e-01 -3.39912921e-02 -2.63692558e-01
5.02869725e-01 -6.15235083e-02 -4.12105769e-01 9.74093854e-01
-7.11428463e-01 -7.98337832e-02 3.93040001e-01 3.83113950e-01
1.55689263e+00 -3.99462551e-01 5.54833114e-01 -3.14475298e-01
5.84798276e-01 6.95009410e-01 5.83841085e-01 9.64080334e-01
-1.56356260e-01 1.80175871e-01 7.30710089e-01 -6.59269571e-01
-8.26063633e-01 -1.05928433e+00 5.05272090e-01 4.97752130e-01
8.52738500e-01 -5.91822743e-01 -1.05713701e+00 -7.19810009e-01
-1.57119244e-01 8.58037472e-01 -5.81275642e-01 -1.33979008e-01
-8.05404007e-01 2.64129519e-01 -9.98440012e-03 5.49677834e-02
1.77129418e-01 -9.93899167e-01 -1.32505512e+00 3.99997979e-01
-3.54114205e-01 -9.15370822e-01 -6.25908911e-01 3.07918370e-01
-5.63408136e-01 -1.48385692e+00 5.54431491e-02 -6.22760057e-01
9.60874081e-01 7.80355692e-01 8.75420034e-01 3.31157237e-01
5.58854342e-02 5.06720066e-01 -4.29920226e-01 -3.97312343e-01
-4.62900311e-01 1.41927391e-01 -7.81270042e-02 -5.06097734e-01
-1.13626488e-01 -2.78468370e-01 -1.59021184e-01 7.06324458e-01
-8.24022174e-01 2.97948420e-01 2.68218368e-01 4.09224749e-01
1.00093853e+00 8.83817792e-01 2.35655054e-01 -4.94029760e-01
6.30795956e-01 -3.48310292e-01 -7.30412602e-01 4.91700381e-01
-7.17943192e-01 2.45825708e-01 5.92738330e-01 -6.48321956e-02
-4.50647563e-01 1.10584415e-01 5.64207017e-01 -7.17235953e-02
-2.68397182e-01 7.00722575e-01 -1.99172631e-01 1.06682956e-01
3.83052021e-01 4.31357846e-02 3.83362472e-02 1.03437640e-01
3.67317557e-01 -1.06900195e-02 8.25272799e-01 -4.63810802e-01
1.01048720e+00 4.70474988e-01 5.93640924e-01 -2.36334413e-01
-2.13016123e-01 -1.84319496e-01 -4.28242356e-01 -3.77047211e-01
7.89322317e-01 -4.72485498e-02 -1.21393728e+00 -2.20771849e-01
-1.52793634e+00 -4.73692864e-01 -4.86973405e-01 2.56079704e-01
-1.06451631e+00 1.82495356e-01 -6.84400797e-02 -1.10679340e+00
3.11487228e-01 -1.36960852e+00 7.53504634e-01 1.91439167e-02
-4.62647468e-01 -5.43809593e-01 2.96228379e-01 3.30648832e-02
1.11389525e-01 7.72559762e-01 1.06787407e+00 -7.81557381e-01
-1.08395493e+00 9.18289870e-02 3.86865661e-02 -8.12369585e-01
-6.58462718e-02 -2.27561980e-01 -1.60208866e-01 -2.86317974e-01
1.69100892e-02 3.83281976e-01 2.40099624e-01 3.73197913e-01
6.92768276e-01 -6.96047306e-01 -6.93852425e-01 -2.03950956e-01
1.45068634e+00 6.37089133e-01 3.72734815e-01 7.65679538e-01
1.13218956e-01 1.02002954e+00 9.85186994e-01 3.93530428e-01
5.91566086e-01 9.55379725e-01 9.98855412e-01 -1.28793688e-02
3.95584583e-01 -4.69170548e-02 9.58674029e-02 8.60716179e-02
-2.15064690e-01 -4.55494881e-01 -1.07935500e+00 4.24384266e-01
-2.38036442e+00 -1.16475260e+00 -6.66421115e-01 2.17629051e+00
1.22096814e-01 1.30121127e-01 2.81439424e-01 4.73784864e-01
7.48992085e-01 -3.21045667e-01 -6.90855563e-01 -8.13494205e-01
1.08539857e-01 -2.62806445e-01 4.50562984e-01 1.10152423e+00
-7.52943873e-01 5.22789180e-01 5.57828283e+00 2.07779810e-01
-5.04969478e-01 1.53134078e-01 1.03536412e-01 9.85722095e-02
-4.69202965e-01 4.59926784e-01 -3.00967067e-01 2.50355750e-02
8.87008309e-01 -6.58106029e-01 9.53142941e-01 8.97181809e-01
3.94668162e-01 -3.59807283e-01 -1.39551735e+00 3.71935070e-01
-2.73449987e-01 -1.17053330e+00 -4.95464593e-01 3.58575642e-01
4.92704391e-01 -5.07341743e-01 -1.26540795e-01 -3.27850059e-02
2.65779227e-01 -1.20912671e+00 1.04300332e+00 2.27031335e-01
2.58221477e-01 -1.09420764e+00 5.07698298e-01 7.35296905e-01
-1.36707938e+00 -2.43704036e-01 -4.72928472e-02 -3.60423982e-01
5.59744775e-01 1.64182648e-01 -1.10488439e+00 8.71790588e-01
4.83736843e-01 1.70846894e-01 1.95561782e-01 1.13864911e+00
-4.21083510e-01 -2.43465587e-01 -3.36991787e-01 3.52795720e-02
5.55687845e-01 -1.56695366e-01 8.56080890e-01 7.17705846e-01
5.39542198e-01 4.98790324e-01 5.55433154e-01 9.74453866e-01
7.20597327e-01 -5.09748399e-01 -8.30452919e-01 1.85899213e-01
6.04244351e-01 8.41078162e-01 -1.08095908e+00 -4.93575111e-02
-1.38113886e-01 8.03580284e-01 9.25256535e-02 3.12216312e-01
-1.03383172e+00 -1.75233424e-01 7.03735709e-01 3.13832432e-01
2.13252604e-02 -5.49263954e-01 -2.87492692e-01 -4.11176056e-01
5.73665909e-02 -7.98014462e-01 4.20305312e-01 -9.29478526e-01
-5.25836349e-01 9.24450099e-01 3.57448667e-01 -1.03956449e+00
-5.25487423e-01 -1.49806499e-01 -6.57806218e-01 7.61763096e-01
-1.45137334e+00 -8.38809788e-01 -4.02703911e-01 7.96906650e-01
6.50417686e-01 2.92035192e-01 9.00630593e-01 -4.35196280e-01
-4.07236367e-01 -2.35749885e-01 -4.95114982e-01 -7.77316570e-01
-2.34835059e-03 -1.35867167e+00 5.43841660e-01 1.05621016e+00
8.88588801e-02 4.67675358e-01 1.16166151e+00 -6.89147651e-01
-1.57501149e+00 -9.31377947e-01 1.14342391e+00 -2.91482151e-01
4.94629264e-01 -1.88749526e-02 -7.83616662e-01 9.80089545e-01
2.88741499e-01 -3.53264332e-01 2.11379364e-01 -1.26448512e-01
7.72553449e-03 2.13942841e-01 -1.04565775e+00 7.00186849e-01
1.26247883e+00 -9.76212397e-02 -4.40378934e-01 2.51590103e-01
9.53383625e-01 -4.23607558e-01 -4.61142361e-01 1.70671299e-01
7.14697242e-02 -1.22782123e+00 6.25245214e-01 -6.38193548e-01
1.24662623e-01 -1.09053838e+00 1.21120043e-01 -1.48705494e+00
-5.54219723e-01 -8.69055688e-01 1.25274211e-01 6.78407431e-01
5.05894661e-01 -7.68430531e-01 7.82935560e-01 7.93783188e-01
-5.80654383e-01 -8.33349288e-01 -1.28693640e+00 -1.19537568e+00
-4.24569607e-01 -4.92879182e-01 1.03508437e+00 5.84521413e-01
6.15163684e-01 1.59608707e-01 -2.71414399e-01 7.93135524e-01
5.03004253e-01 5.51810980e-01 8.57111990e-01 -1.30287051e+00
-4.61752206e-01 -3.05692732e-01 -1.40393704e-01 -7.39453554e-01
1.62411079e-01 -6.96381569e-01 3.99951369e-01 -2.00889349e+00
2.66308337e-01 -5.73940337e-01 2.89851665e-01 8.36657286e-01
3.63044649e-01 -5.85280716e-01 6.37810886e-01 4.08498764e-01
-7.76652813e-01 7.74655193e-02 1.40045512e+00 -2.17803359e-01
-3.45505178e-01 1.21148609e-01 -4.32045400e-01 4.56589133e-01
8.38162005e-01 -7.65240729e-01 -7.39648163e-01 -4.00027633e-01
3.09000552e-01 8.77749562e-01 4.89736915e-01 -8.03713739e-01
5.04630625e-01 -8.59611154e-01 -7.25364208e-01 -4.73139703e-01
2.42114305e-01 -1.30781841e+00 9.74819303e-01 1.02790856e+00
-2.02685535e-01 6.77785695e-01 1.20921597e-01 7.34255612e-01
-2.35608190e-01 -5.16402602e-01 1.27115473e-01 -6.70722872e-02
-7.90184915e-01 2.51167685e-01 -7.72575378e-01 -3.85189861e-01
1.76366472e+00 -4.52077657e-01 -6.12396598e-01 -5.21304309e-01
-8.33301842e-01 5.08148670e-01 6.37104809e-01 9.82186943e-03
7.46272683e-01 -1.10743380e+00 -3.64980102e-01 -5.81530221e-02
5.06574474e-02 2.65508652e-01 3.06879371e-01 9.32013392e-01
-3.58134121e-01 5.78166127e-01 -2.78840452e-01 -3.81594747e-01
-1.33035052e+00 1.09837973e+00 2.14609325e-01 -4.49474066e-01
-6.65666580e-01 2.46439889e-01 4.03669626e-01 -1.83957309e-01
7.06898198e-02 -4.82519656e-01 -2.39828415e-02 -7.13879347e-01
5.78121603e-01 6.17156327e-01 -2.63372689e-01 -3.20948541e-01
-6.32046103e-01 5.31405509e-01 3.35471511e-01 -4.11278307e-01
1.18300402e+00 -3.01993310e-01 -3.09250891e-01 6.71498552e-02
3.95595878e-01 -4.60707918e-02 -1.10748744e+00 2.38091573e-01
2.08913773e-01 -3.17310959e-01 -4.94319856e-01 -7.37212062e-01
-6.75332367e-01 4.23532367e-01 -2.81189624e-02 9.34803963e-01
1.07904267e+00 3.70052487e-01 6.04341090e-01 4.65977043e-01
1.05463529e+00 -5.71067810e-01 -1.12733077e-02 6.36133552e-01
9.40093696e-01 -6.42667472e-01 -3.03500324e-01 -9.00748909e-01
-5.90687931e-01 1.19889390e+00 4.76076245e-01 1.97640091e-01
3.89855057e-02 4.06006545e-01 -3.14104855e-01 -3.38069439e-01
-7.04560578e-01 -1.83796778e-01 -3.53516549e-01 7.34603107e-01
-5.75604260e-01 2.11497352e-01 -3.96461576e-01 5.92309833e-01
-4.44785416e-01 -1.33739725e-01 9.25319493e-01 1.05143774e+00
-8.36890459e-01 -1.22682059e+00 -6.48575246e-01 -1.79445416e-01
3.08709860e-01 5.14114141e-01 -6.56628966e-01 1.03353035e+00
2.07537487e-01 1.23408043e+00 -5.51595725e-02 -3.11483592e-01
6.80739045e-01 -4.58020747e-01 3.97403568e-01 -5.10086238e-01
-2.03860775e-01 -2.41530940e-01 2.06751972e-01 -8.32776904e-01
-4.14877206e-01 -6.07321799e-01 -1.87036383e+00 -6.73903763e-01
-1.22837022e-01 6.74915969e-01 4.88081217e-01 1.14697289e+00
1.74012274e-01 4.79199558e-01 6.72695220e-01 -6.06626153e-01
-1.25501409e-01 -4.30866443e-02 -6.27677858e-01 -1.28372163e-01
3.19785595e-01 -5.82747340e-01 -2.39755407e-01 -2.76797980e-01] | [4.892002582550049, 1.7628222703933716] |
366ef9f3-c20e-4cfa-aa4a-a6daf2606940 | chinese-named-entity-recognition-with-graph | null | null | https://aclanthology.org/W15-3103 | https://aclanthology.org/W15-3103.pdf | Chinese Named Entity Recognition with Graph-based Semi-supervised Learning Model | null | ['Aaron Li-Feng Han', 'Lidia S. Chao', 'Xiaodong Zeng', 'Derek F. Wong'] | 2015-07-01 | null | null | null | ws-2015-7 | ['chinese-named-entity-recognition'] | ['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] | [-7.4049072265625, 3.7947235107421875] |
f15423be-9033-469f-b775-cb8ee7d7d96b | chinese-zero-pronoun-resolution-with-deep-1 | null | null | https://aclanthology.org/P16-1074 | https://aclanthology.org/P16-1074.pdf | Chinese Zero Pronoun Resolution with Deep Neural Networks | null | ['Vincent Ng', 'Chen Chen'] | 2016-08-01 | null | null | null | acl-2016-8 | ['chinese-zero-pronoun-resolution'] | ['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] | [-7.365434169769287, 3.696352958679199] |
cb04c335-8eb4-4e4a-ab56-a086d3fe7a6a | breast-cancer-detection-using-convolutional | 2003.07911 | null | https://arxiv.org/abs/2003.07911v3 | https://arxiv.org/pdf/2003.07911v3.pdf | Breast Cancer Detection Using Convolutional Neural Networks | Breast cancer is prevalent in Ethiopia that accounts 34% among women cancer patients. The diagnosis technique in Ethiopia is manual which was proven to be tedious, subjective, and challenging. Deep learning techniques are revolutionizing the field of medical image analysis and hence in this study, we proposed Convolutional Neural Networks (CNNs) for breast mass detection so as to minimize the overheads of manual analysis. CNN architecture is designed for the feature extraction stage and adapted both the Region Proposal Network (RPN) and Region of Interest (ROI) portion of the faster R-CNN for the automated breast mass abnormality detection. Our model detects mass region and classifies them into benign or malignant abnormality in mammogram(MG) images at once. For the proposed model, MG images were collected from different hospitals, locally.The images were passed through different preprocessing stages such as gaussian filter, median filter, bilateral filters and extracted the region of the breast from the background of the MG image. The performance of the model on test dataset is found to be: detection accuracy 91.86%, sensitivity of 94.67% and AUC-ROC of 92.2%. | ['Yaecob Girmay', 'Simon Hadush', 'Gebrekirstos Hagos', 'Abiot Sinamo'] | 2020-03-17 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 1.76707402e-01 4.60121781e-01 -1.93271950e-01 -3.73333365e-01
-3.19797963e-01 -1.49965107e-01 4.14946824e-02 5.78550637e-01
-4.58107620e-01 3.18670154e-01 -3.97608317e-02 -7.49469101e-01
-6.55942261e-02 -1.11772442e+00 -3.67289752e-01 -8.21724057e-01
-2.22863659e-01 3.12550664e-01 2.04019979e-01 1.77086070e-01
2.10021824e-01 9.95526969e-01 -8.49218369e-01 5.74761689e-01
3.21882546e-01 1.13996470e+00 1.65883154e-01 1.19328928e+00
1.01085715e-01 8.50428462e-01 -2.47575119e-01 -2.82907218e-01
4.94672596e-01 -2.67854691e-01 -7.14122534e-01 1.39591902e-01
1.01005040e-01 -7.00303495e-01 -4.11359131e-01 8.25080633e-01
6.83596671e-01 -2.03917220e-01 9.53345656e-01 -7.93821394e-01
-4.50943053e-01 4.29165900e-01 -1.16929984e+00 7.24553943e-01
-6.42225444e-01 -1.97242931e-01 7.11934343e-02 -7.64434576e-01
4.16511714e-01 1.03405404e+00 1.04259002e+00 4.88204598e-01
-5.98701715e-01 -9.57028091e-01 -7.56738603e-01 -1.47305150e-02
-1.24007702e+00 -3.56641471e-01 2.20842272e-01 -2.38673478e-01
8.11716795e-01 2.17626914e-01 7.00627625e-01 2.20782116e-01
1.05663788e+00 4.63351339e-01 8.09128940e-01 -5.22058427e-01
-1.32584363e-01 2.94692606e-01 -3.80777754e-02 8.82670522e-01
6.21777654e-01 5.56503870e-02 8.65106359e-02 -1.81126297e-01
9.98806238e-01 3.09952706e-01 3.79748225e-01 -6.70087039e-02
-7.57074475e-01 1.09407461e+00 6.53506875e-01 4.93802935e-01
-6.57205582e-01 6.27668798e-02 6.33837819e-01 1.59496628e-02
2.73972601e-01 -2.30755433e-01 -4.08328980e-01 4.85982031e-01
-1.04042745e+00 4.95661236e-02 4.09790933e-01 4.73451227e-01
9.88928080e-02 -3.06805447e-02 -1.90443024e-01 6.55876219e-01
5.74684799e-01 3.98296595e-01 6.18689001e-01 -3.37263048e-01
1.51550323e-01 9.43061113e-01 -2.28880376e-01 -1.48860240e+00
-7.80273974e-01 -7.28515446e-01 -1.25571120e+00 -2.42280923e-02
1.97417513e-01 -3.96480381e-01 -1.52669728e+00 9.60436702e-01
5.94677508e-01 -4.04183120e-01 1.36050835e-01 5.89599192e-01
1.03024268e+00 5.92958212e-01 4.41190958e-01 1.94842771e-01
1.61612499e+00 -4.60929245e-01 -5.61013341e-01 5.35069443e-02
7.47981429e-01 -7.82075822e-01 8.38467106e-02 -2.23442372e-02
-1.19305015e+00 -4.94735032e-01 -1.11908555e+00 5.57568632e-02
-5.31676769e-01 7.13917553e-01 7.14105844e-01 6.35395288e-01
-1.02190065e+00 1.83651850e-01 -1.19931281e+00 -8.05413485e-01
1.08704770e+00 7.91030526e-01 -5.56398571e-01 -5.87387867e-02
-5.34326792e-01 9.36599791e-01 5.59214711e-01 3.29348981e-01
-1.01882720e+00 -4.44979161e-01 -6.50750041e-01 -2.08448589e-01
-1.28921673e-01 -3.88986826e-01 1.13584912e+00 -1.22324920e+00
-7.27919936e-01 1.20944965e+00 5.84631339e-02 -7.11418092e-01
3.82128149e-01 2.72557437e-01 -3.12552840e-01 4.08622712e-01
9.90708470e-02 9.35189068e-01 4.70496923e-01 -9.07937109e-01
-1.21964550e+00 -5.86213887e-01 -4.64289427e-01 -6.36651218e-02
9.35946628e-02 2.73733646e-01 2.82745156e-02 -6.76094294e-01
7.23186135e-01 -7.04895496e-01 -3.66952300e-01 1.62798673e-01
-1.24176733e-01 1.29326865e-01 1.15399468e+00 -1.23931003e+00
1.03565657e+00 -2.17408943e+00 -6.64081872e-01 4.59180593e-01
3.08912307e-01 3.38079453e-01 3.37601304e-01 -7.31597096e-03
-2.63060361e-01 1.22126259e-01 -4.19636741e-02 3.33164662e-01
-5.38515806e-01 -5.83124869e-02 6.57355011e-01 9.02950644e-01
4.86997813e-01 8.82093310e-01 -5.20984650e-01 -8.38500202e-01
2.08437443e-01 5.83708644e-01 -1.05865508e-01 9.48715732e-02
3.82489413e-01 1.31707177e-01 -2.79877186e-01 1.23940825e+00
1.03494406e+00 -4.95026782e-02 -4.20611911e-02 -4.82568562e-01
1.81115925e-01 -4.61790651e-01 -9.14778769e-01 9.46701467e-01
-7.94071555e-02 7.65799880e-01 3.82998884e-01 -1.17787087e+00
9.18719709e-01 4.67528850e-01 5.02160132e-01 -5.03419578e-01
8.47573876e-01 2.37652436e-01 6.73673391e-01 -1.02687621e+00
2.44944409e-01 -1.87710598e-01 3.92350852e-01 2.65121430e-01
3.71384202e-03 3.19364488e-01 -1.28160372e-01 8.13305005e-02
1.10876334e+00 -3.27926278e-01 7.38034308e-01 -3.28321636e-01
2.37668902e-01 3.13897312e-01 2.89903998e-01 6.42493248e-01
-7.00353205e-01 2.94359803e-01 2.32219845e-01 -6.63611114e-01
-1.11410427e+00 -9.09113705e-01 -3.72125030e-01 8.86775911e-01
-2.68676937e-01 6.04552567e-01 -6.04573667e-01 -5.77416778e-01
1.39418378e-01 1.60366029e-01 -1.05054772e+00 -2.34416112e-01
-6.07011378e-01 -9.94993925e-01 5.97403705e-01 7.63038218e-01
9.94823813e-01 -1.12149143e+00 -9.79279220e-01 2.22393259e-01
2.46938109e-01 -5.79002678e-01 3.31141800e-02 4.70574200e-01
-1.29306555e+00 -1.28770912e+00 -8.61031532e-01 -1.19794142e+00
1.15193653e+00 1.50703371e-01 7.68503249e-01 1.48735911e-01
-1.11959112e+00 -2.16797188e-01 -1.85936600e-01 -1.05095720e+00
-5.98647714e-01 -4.86467294e-02 -5.84342837e-01 -2.49156862e-01
8.44355166e-01 -4.36184891e-02 -8.94053936e-01 -6.99318945e-02
-9.33869064e-01 -3.72562860e-03 1.22921574e+00 6.16087794e-01
4.22326744e-01 3.08819443e-01 6.49813592e-01 -9.28583324e-01
3.22071463e-01 -9.26512003e-01 -1.73520878e-01 -6.21889494e-02
2.52670585e-03 -8.31149817e-01 -6.06468618e-02 -3.28911185e-01
-7.96318531e-01 2.07287252e-01 1.38674140e-01 1.34095222e-01
-3.34080458e-01 3.86610478e-01 3.80630165e-01 -2.19426736e-01
6.31637335e-01 -5.50427735e-02 2.18664959e-01 -6.42834604e-02
-3.92716020e-01 1.04195595e+00 6.47464871e-01 3.07931423e-01
4.55721527e-01 5.93849659e-01 3.32027793e-01 -6.00090623e-01
-4.27542627e-01 -6.40276372e-01 -6.66481793e-01 -3.09365749e-01
1.10396183e+00 -8.14888775e-01 -3.38101596e-01 3.74022782e-01
-7.44569659e-01 1.22001797e-01 3.11681211e-01 5.98805904e-01
1.69483021e-01 -1.34507298e-01 -7.93277681e-01 -1.08658135e+00
-9.28259909e-01 -8.32888246e-01 6.95581555e-01 6.59179270e-01
-1.48559615e-01 -8.26864421e-01 -5.92656136e-01 2.74306417e-01
9.61260140e-01 7.31185496e-01 1.10248983e+00 -7.81990528e-01
3.47538330e-02 -9.55755651e-01 -9.27209437e-01 1.56934008e-01
4.65980083e-01 1.35401532e-01 -8.01558912e-01 -1.36994168e-01
5.89844445e-03 3.66215035e-02 7.64885664e-01 1.16389811e+00
1.21596611e+00 -2.41772518e-01 -6.72151804e-01 4.15437520e-01
1.64596736e+00 8.19072366e-01 4.65716511e-01 3.62314552e-01
2.81956911e-01 4.11925524e-01 4.77699578e-01 3.63103777e-01
-4.16533537e-02 -2.94002056e-01 5.88244200e-01 -5.37318468e-01
-1.00520521e-01 2.59531260e-01 -1.52919650e-01 2.45985895e-01
1.13148075e-02 -1.31589277e-02 -1.05525303e+00 8.14945281e-01
-1.36148429e+00 -5.88251829e-01 -2.94807144e-02 1.61175883e+00
5.04424274e-01 -4.54551764e-02 1.88774288e-01 3.15978527e-01
8.74616325e-01 -6.87776923e-01 -2.96763361e-01 -6.91513062e-01
2.36826256e-01 5.75352311e-01 9.64123607e-01 -1.26674189e-03
-1.53169882e+00 4.69598621e-01 5.90487385e+00 4.65599358e-01
-1.30590236e+00 2.04999194e-01 1.32115769e+00 -2.29509603e-02
7.44113743e-01 -5.25022030e-01 -5.99002481e-01 -1.80586204e-02
9.14182842e-01 3.55118275e-01 -2.68162757e-01 7.38672972e-01
3.28342140e-01 -7.31896281e-01 -5.56579053e-01 5.79438984e-01
-1.19540719e-02 -1.23609948e+00 4.89915572e-02 4.76566367e-02
6.72979653e-01 -2.59686373e-02 -1.36804551e-01 9.50619727e-02
-4.84464988e-02 -1.51949370e+00 1.28736556e-01 3.83505702e-01
4.46508229e-01 -1.01926088e+00 1.43447578e+00 2.30413839e-01
-9.96389091e-01 -2.99545079e-01 -4.14273202e-01 2.32387558e-01
-5.61256111e-01 2.91766018e-01 -1.44507527e+00 1.05254926e-01
9.01623666e-01 8.72228965e-02 -7.52634645e-01 1.20252478e+00
9.82706547e-01 6.89139664e-01 -2.91457593e-01 -1.80158928e-01
4.87740755e-01 2.98578411e-01 -7.22008571e-02 1.51646137e+00
4.55998093e-01 7.18060881e-02 -1.19800486e-01 6.16212964e-01
2.18513794e-02 4.20027196e-01 -5.78788280e-01 -1.90319970e-01
1.56150311e-01 1.41356301e+00 -1.43971324e+00 -1.70427844e-01
-1.70144081e-01 3.05179387e-01 -2.85505533e-01 -1.58479035e-01
-6.67259634e-01 -6.61911368e-01 -3.46750855e-01 3.52783442e-01
2.19714403e-01 3.62048984e-01 -3.24915618e-01 -2.05257591e-02
-3.17555189e-01 -7.87185848e-01 5.36997318e-01 -1.62015691e-01
-7.69171178e-01 1.52260929e-01 7.63505697e-02 -6.11177146e-01
3.45098704e-01 -7.58683324e-01 -7.34291255e-01 8.71611297e-01
-1.12261403e+00 -1.46817601e+00 -6.62707210e-01 3.37741166e-01
6.19621456e-01 -3.43928188e-01 6.86409354e-01 3.50978673e-01
-6.08416021e-01 5.69261134e-01 -8.78934339e-02 6.16465569e-01
2.33085126e-01 -9.67896998e-01 -2.36269087e-01 6.45771980e-01
-8.20831299e-01 3.25740933e-01 2.50443459e-01 -8.59197676e-01
-1.16275632e+00 -1.58060873e+00 6.48943365e-01 2.60174751e-01
1.91777393e-01 1.29343212e-01 -4.34672028e-01 6.05677426e-01
1.15205295e-01 2.80371249e-01 7.15038240e-01 -7.68799484e-01
4.11992699e-01 -1.70366578e-02 -1.82870102e+00 1.72342271e-01
6.21140227e-02 1.92516908e-01 5.25494367e-02 2.49945968e-01
-3.92997786e-02 -5.63685298e-01 -7.95844972e-01 8.00014794e-01
8.69293392e-01 -8.14000010e-01 7.73770928e-01 -4.33790773e-01
5.43629944e-01 1.19293369e-02 -2.34051168e-01 -4.99735415e-01
-4.46662217e-01 7.52520338e-02 -5.92600480e-02 8.85933578e-01
5.47007799e-01 -3.83031428e-01 9.10743535e-01 1.72177970e-01
-2.88146976e-02 -1.28203797e+00 -7.54586160e-01 9.15977433e-02
5.80573305e-02 -9.27507505e-02 2.06005409e-01 5.76002777e-01
-6.23750806e-01 -4.26087290e-01 2.95308411e-01 3.44222814e-01
3.60271394e-01 -4.98093963e-01 4.83052939e-01 -9.86662507e-01
8.66546631e-02 -3.38374138e-01 -7.07460344e-01 5.52082881e-02
-7.04984546e-01 -6.78383350e-01 -4.08412516e-02 -1.66756070e+00
7.24876583e-01 -3.88910532e-01 -2.59610683e-01 5.52180469e-01
3.86361033e-02 4.09600466e-01 -3.79768491e-01 -1.59190997e-01
8.62232074e-02 -4.56968486e-01 1.38900781e+00 -2.08475500e-01
-9.26033854e-02 2.44889900e-01 -6.49113357e-01 7.60218441e-01
1.23744905e+00 -5.28171778e-01 -1.76344752e-01 -8.91398713e-02
-1.42798021e-01 1.63655639e-01 4.54069167e-01 -1.11265373e+00
9.50324908e-02 -1.84595302e-01 1.48954034e+00 -1.22258747e+00
-7.89754093e-03 -8.50396156e-01 1.92135423e-01 1.02794504e+00
-1.90181181e-01 -5.49873849e-03 5.17857254e-01 2.86082983e-01
1.42353207e-01 -4.60687697e-01 1.27768540e+00 -5.61485469e-01
-2.29433686e-01 2.68485099e-01 -4.68772918e-01 -7.21597850e-01
1.49181271e+00 -5.87784290e-01 7.10854158e-02 1.17367199e-02
-6.97791457e-01 -9.95535403e-02 -8.20296481e-02 -3.04483436e-02
6.60083115e-01 -1.13142037e+00 -9.07935500e-01 1.71311960e-01
-6.91531375e-02 2.89437652e-01 1.27858669e-01 1.19823873e+00
-1.42288971e+00 4.84745979e-01 -4.82075453e-01 -4.98634428e-01
-1.62946248e+00 8.56223330e-02 7.44008958e-01 -1.33253321e-01
-6.79523110e-01 1.03966749e+00 -1.55485690e-01 5.29852398e-02
5.33185482e-01 -5.59498429e-01 -5.76125920e-01 -1.64738357e-01
5.65338910e-01 5.06372750e-01 8.82977545e-02 -7.85575628e-01
-2.18299270e-01 3.94028813e-01 -4.77177531e-01 3.37983817e-01
1.30035746e+00 7.68631995e-02 -3.10126901e-01 1.14016436e-01
1.22132099e+00 -4.69879031e-01 -4.06441689e-01 1.17429666e-01
-1.20248728e-01 -2.22581148e-01 5.02443314e-01 -9.27798212e-01
-1.38116431e+00 6.00313306e-01 1.40571475e+00 1.57682672e-01
1.24976528e+00 -1.01149835e-01 6.56481624e-01 2.34058678e-01
-3.20233196e-01 -1.04821062e+00 -2.42338926e-01 -1.07876316e-01
6.59425974e-01 -1.61253798e+00 1.75395802e-01 -7.57823810e-02
-2.18497291e-01 1.44013155e+00 8.20540369e-01 -5.29599011e-01
1.10386813e+00 3.51776630e-01 1.50828406e-01 -5.34283638e-01
-4.25931484e-01 2.39791811e-01 4.67258096e-02 7.96009302e-01
8.00583661e-01 2.94424772e-01 -1.70508340e-01 5.54702997e-01
9.10967705e-04 2.73320109e-01 2.99488455e-01 1.39054120e+00
-8.67478371e-01 -4.08452265e-02 -6.06261969e-01 1.31223965e+00
-1.27993178e+00 2.19904840e-01 -3.84244949e-01 1.12163103e+00
6.65804029e-01 8.92243505e-01 2.54272461e-01 4.88714688e-02
-2.47763395e-01 -3.36073130e-01 3.88050795e-01 -5.28407753e-01
-7.34310389e-01 2.00793564e-01 -2.52596498e-01 -6.82827143e-04
-3.48077804e-01 -3.05396438e-01 -1.40217102e+00 -2.38866016e-01
-5.87185979e-01 -1.72713503e-01 1.11167026e+00 7.26675510e-01
-4.64511365e-02 7.52342820e-01 3.64944726e-01 -6.47843719e-01
-3.43172550e-01 -1.53000259e+00 -6.26503110e-01 -1.16138861e-01
5.27810633e-01 -3.37937325e-01 2.79349945e-02 2.01843947e-01] | [15.28353500366211, -2.5559744834899902] |
7033488d-90f0-45db-a4e3-fd287580ff2b | lifelong-3d-object-recognition-and-grasp | 2109.11544 | null | https://arxiv.org/abs/2109.11544v2 | https://arxiv.org/pdf/2109.11544v2.pdf | Lifelong 3D Object Recognition and Grasp Synthesis Using Dual Memory Recurrent Self-Organization Networks | Humans learn to recognize and manipulate new objects in lifelong settings without forgetting the previously gained knowledge under non-stationary and sequential conditions. In autonomous systems, the agents also need to mitigate similar behavior to continually learn the new object categories and adapt to new environments. In most conventional deep neural networks, this is not possible due to the problem of catastrophic forgetting, where the newly gained knowledge overwrites existing representations. Furthermore, most state-of-the-art models excel either in recognizing the objects or in grasp prediction, while both tasks use visual input. The combined architecture to tackle both tasks is very limited. In this paper, we proposed a hybrid model architecture consists of a dynamically growing dual-memory recurrent neural network (GDM) and an autoencoder to tackle object recognition and grasping simultaneously. The autoencoder network is responsible to extract a compact representation for a given object, which serves as input for the GDM learning, and is responsible to predict pixel-wise antipodal grasp configurations. The GDM part is designed to recognize the object in both instances and categories levels. We address the problem of catastrophic forgetting using the intrinsic memory replay, where the episodic memory periodically replays the neural activation trajectories in the absence of external sensory information. To extensively evaluate the proposed model in a lifelong setting, we generate a synthetic dataset due to lack of sequential 3D objects dataset. Experiment results demonstrated that the proposed model can learn both object representation and grasping simultaneously in continual learning scenarios. | ['Hamidreza Kasaei', 'Krishnakumar Santhakumar'] | 2021-09-23 | null | null | null | null | ['3d-object-recognition'] | ['computer-vision'] | [ 2.76669741e-01 -9.37396213e-02 2.35604912e-01 4.78381030e-02
2.74656326e-01 -3.33992839e-01 4.50177491e-01 2.02721823e-02
-3.67436230e-01 8.00923765e-01 -4.21514630e-01 2.78822273e-01
-1.42879859e-01 -1.03730905e+00 -1.23353708e+00 -1.11063218e+00
-2.43797585e-01 6.44219100e-01 4.57663924e-01 -1.26367986e-01
2.18958691e-01 6.48328006e-01 -2.29805446e+00 3.04507405e-01
6.72921538e-01 1.03300285e+00 9.50723588e-01 5.71128964e-01
-1.30117729e-01 6.62975311e-01 -3.85637879e-01 3.13039958e-01
1.78112879e-01 -1.53578594e-01 -4.68104601e-01 1.20663807e-01
1.00488644e-02 -3.76854211e-01 -5.40856779e-01 6.74437761e-01
3.75883043e-01 4.30958867e-01 5.03520131e-01 -1.00791669e+00
-9.28109527e-01 4.16751146e-01 -2.02487886e-01 2.58585155e-01
8.47278610e-02 4.31648791e-01 1.08077899e-01 -9.66037035e-01
4.77018505e-01 1.11432993e+00 4.90836710e-01 7.56145239e-01
-8.02295446e-01 -4.19297159e-01 3.81835312e-01 5.21822870e-01
-1.01434827e+00 -8.86142254e-02 8.37334871e-01 -2.04042643e-01
1.25978374e+00 -2.34408662e-01 1.15780854e+00 1.44798732e+00
5.76840818e-01 8.61458480e-01 1.12635970e+00 -2.42482245e-01
4.88546848e-01 5.24605773e-02 1.28457874e-01 5.66691220e-01
4.08578008e-01 4.20811266e-01 -6.51857197e-01 1.90949872e-01
9.03513968e-01 5.89639306e-01 -1.58274591e-01 -4.77809489e-01
-1.08869839e+00 2.08376795e-01 4.84910101e-01 4.87008035e-01
-9.04298902e-01 9.19980332e-02 5.03459275e-02 5.34654021e-01
-1.56302494e-03 1.95053086e-01 -5.45135498e-01 -1.09671414e-01
-5.62838733e-01 2.72436172e-01 7.64104605e-01 9.21062708e-01
7.11601853e-01 3.62797767e-01 1.43959269e-01 5.42648971e-01
-9.62086171e-02 5.72840750e-01 1.01937473e+00 -5.05040348e-01
3.93626653e-02 7.60304332e-01 1.92063630e-01 -9.86314535e-01
-5.07578254e-01 -6.26485884e-01 -8.75095785e-01 4.91418928e-01
2.03618273e-01 2.53809839e-01 -1.39541471e+00 1.84571767e+00
3.00026983e-01 1.51085272e-01 2.74936199e-01 9.16840017e-01
4.56348777e-01 7.33984709e-01 3.23393904e-02 -3.49497586e-01
8.84908319e-01 -8.76908600e-01 -6.29153669e-01 -3.11582208e-01
5.56274019e-02 -2.48806059e-01 1.04686940e+00 4.31885362e-01
-1.21595097e+00 -8.25247288e-01 -1.23648083e+00 2.56450862e-01
-7.78887212e-01 2.95842672e-03 3.84748906e-01 -3.35252136e-02
-7.55247891e-01 9.16908979e-01 -1.14977479e+00 -5.09467781e-01
2.75491357e-01 4.45329428e-01 -2.56649047e-01 -2.55127437e-02
-9.99462485e-01 1.02595544e+00 8.71891856e-01 3.75995547e-01
-1.46971071e+00 -3.29260051e-01 -5.09390831e-01 1.71972275e-01
4.40159649e-01 -6.03947401e-01 1.00513840e+00 -1.05685043e+00
-1.58761966e+00 5.04479110e-01 1.59661680e-01 -5.81280470e-01
3.71509910e-01 -3.55498224e-01 -2.85563856e-01 -8.70539621e-02
-3.17802161e-01 5.66156209e-01 1.38889229e+00 -1.43875372e+00
-4.61215585e-01 -6.19763792e-01 -1.23935826e-01 2.72923648e-01
-4.60903823e-01 -1.03039992e+00 2.86622308e-02 -7.33302057e-01
4.30167079e-01 -9.78919506e-01 1.98472247e-01 -2.09524602e-01
2.84461200e-01 3.02269179e-02 1.42560911e+00 -4.87248033e-01
7.53254414e-01 -2.07026243e+00 6.28467441e-01 -2.62466431e-01
-1.95857242e-01 3.45499516e-01 -2.30064020e-01 4.22101557e-01
-3.14633846e-02 -5.01898289e-01 -2.79653668e-01 -1.04111776e-01
-3.74382168e-01 5.93903244e-01 -6.02825224e-01 1.52678341e-01
2.02864811e-01 8.72475803e-01 -8.74373317e-01 3.02416161e-02
9.18228179e-02 5.22119939e-01 -2.66220778e-01 4.51059371e-01
-5.58957934e-01 4.56162363e-01 -2.47214749e-01 7.50611424e-01
7.16668963e-01 -2.04363123e-01 1.59528449e-01 -4.74029519e-02
-1.30319104e-01 -3.91372919e-01 -1.14326680e+00 1.80676675e+00
-5.00509858e-01 9.84290615e-02 -1.13202654e-01 -1.20895028e+00
1.14120090e+00 1.24310270e-01 3.26588809e-01 -8.12469959e-01
1.29547775e-01 2.05904782e-01 -3.33803073e-02 -6.85381591e-01
3.14343154e-01 -2.62944996e-02 1.98579133e-01 5.12879014e-01
4.18227136e-01 9.84672606e-02 -1.16518609e-01 -2.88169146e-01
1.16832292e+00 3.30447167e-01 6.31805286e-02 5.09609422e-03
3.49937767e-01 -7.56671652e-03 4.85511214e-01 9.44208026e-01
7.74407089e-02 3.34240645e-01 -2.89365292e-01 -1.00518513e+00
-1.01579809e+00 -1.29721105e+00 2.69398063e-01 9.28283811e-01
4.45755869e-01 3.33848476e-01 -4.16259348e-01 -4.16613698e-01
1.45954683e-01 7.79743254e-01 -8.17102134e-01 -8.63305807e-01
-7.36276746e-01 -6.74346983e-01 1.62131165e-03 5.39229751e-01
8.16149116e-01 -1.77227175e+00 -1.64136493e+00 5.44914246e-01
2.53764600e-01 -5.51587999e-01 5.68067357e-02 4.45197284e-01
-1.14507234e+00 -1.09891427e+00 -7.68685639e-01 -1.11065471e+00
7.28372872e-01 3.13943446e-01 6.72009706e-01 1.36284530e-01
-5.23922980e-01 6.42476857e-01 -3.88372779e-01 -3.64384532e-01
-3.13118309e-01 1.71748996e-01 4.87152606e-01 5.18700890e-02
2.92903900e-01 -9.96575773e-01 -4.92111355e-01 1.96352899e-01
-1.21660757e+00 7.81522542e-02 9.82096970e-01 1.24306488e+00
6.31956220e-01 3.19666982e-01 8.95114481e-01 -3.78027171e-01
5.59107602e-01 -5.65800428e-01 -3.91792119e-01 5.03976941e-01
-5.24619043e-01 2.84148157e-01 6.50054812e-01 -1.07665884e+00
-1.18998289e+00 1.06931239e-01 2.58271486e-01 -8.04210901e-01
-2.22406060e-01 4.24319118e-01 -3.39934370e-03 1.12175643e-02
4.03586447e-01 8.42545331e-01 9.88519937e-02 -4.96330738e-01
5.23396581e-02 3.69452089e-01 5.64817011e-01 -6.58125162e-01
3.93835366e-01 2.39832789e-01 -1.09964855e-01 -7.05787659e-01
-4.95357901e-01 1.57615587e-01 -7.67572403e-01 -4.57679361e-01
4.86151576e-01 -5.35484016e-01 -9.63908970e-01 9.94354844e-01
-1.12492597e+00 -4.48693216e-01 -6.57327473e-01 3.73560995e-01
-8.92071187e-01 7.31380074e-04 -6.04398847e-01 -1.06122315e+00
-2.25331873e-01 -8.82209182e-01 6.89413309e-01 5.13558686e-01
3.21583480e-01 -4.67741251e-01 8.60894751e-03 -2.34855354e-01
6.16905868e-01 3.12940001e-01 1.12703788e+00 -3.53355139e-01
-7.33623624e-01 -1.77763656e-01 2.73270875e-01 1.93615004e-01
3.37744474e-01 -5.45970142e-01 -7.13070512e-01 -6.99979782e-01
4.52651739e-01 -5.43300927e-01 1.15237153e+00 1.16330139e-01
1.10635912e+00 -5.49628854e-01 -3.35670143e-01 1.29048541e-01
1.37325776e+00 6.10367298e-01 5.12808681e-01 2.70892441e-01
2.85834998e-01 5.18014133e-01 4.92853552e-01 5.69393694e-01
9.86864567e-02 2.64446855e-01 7.50700772e-01 6.41572833e-01
-7.53802583e-02 -4.43434417e-01 3.67685318e-01 1.01357412e+00
-1.66560382e-01 -1.92781717e-01 -8.58323634e-01 7.64244080e-01
-2.05349851e+00 -9.82788265e-01 7.83227384e-01 2.17707920e+00
5.93568683e-01 2.43724987e-01 -2.42759958e-01 9.39138308e-02
6.45879686e-01 -1.51775479e-01 -1.29289246e+00 -9.11886841e-02
-3.40486765e-01 1.26150131e-01 1.18337549e-01 1.18582822e-01
-6.84734166e-01 9.50609863e-01 5.43092966e+00 4.15955484e-01
-1.31424725e+00 5.82778864e-02 1.14928894e-01 -2.13506624e-01
1.97892010e-01 -1.81246743e-01 -5.13265669e-01 4.74630713e-01
6.98697388e-01 5.56602590e-02 7.42468178e-01 9.57731664e-01
-3.06124210e-01 -2.89867997e-01 -9.53448832e-01 1.00325537e+00
1.93508238e-01 -1.06398416e+00 3.83875638e-01 -3.27595741e-01
4.60109085e-01 -2.78775662e-01 4.43251908e-01 5.46311498e-01
-8.82255360e-02 -9.06240523e-01 8.24843705e-01 1.30956626e+00
3.59890819e-01 -4.74139690e-01 5.75855911e-01 7.17041731e-01
-9.16828394e-01 -7.41843462e-01 -4.46452618e-01 -2.71868616e-01
-8.36183503e-03 1.64284393e-01 -6.44159853e-01 2.15006709e-01
1.11607838e+00 6.75492704e-01 -3.75659436e-01 9.63339210e-01
9.87079665e-02 2.05621719e-01 -4.80466992e-01 -1.81511506e-01
1.14716649e-01 1.73604041e-01 8.51645887e-01 5.53267479e-01
6.35782599e-01 3.64500552e-01 5.20208627e-02 9.68255460e-01
1.65591732e-01 -3.69922698e-01 -7.45374620e-01 -1.57719165e-01
4.00215477e-01 7.51166999e-01 -8.40277135e-01 -3.15275699e-01
1.16644718e-01 1.16599226e+00 5.32956898e-01 4.89347160e-01
-5.30847430e-01 -2.34056234e-01 1.51684001e-01 6.67808354e-02
6.25935376e-01 -6.15804017e-01 4.30223951e-03 -1.15268254e+00
2.59922236e-01 -6.31723344e-01 2.26676613e-01 -7.72035062e-01
-1.22490346e+00 7.56882906e-01 -1.20531037e-01 -1.08470607e+00
-2.89317429e-01 -5.40168941e-01 -4.41316813e-01 3.66886139e-01
-1.29901612e+00 -1.09516799e+00 -6.83757842e-01 6.91545665e-01
8.23504031e-01 -3.83105248e-01 8.39145243e-01 -5.73648931e-03
-3.52537453e-01 2.50000358e-01 9.27985385e-02 -5.50265968e-01
3.75155747e-01 -7.65580297e-01 -1.61883861e-01 4.45631564e-01
-1.25891536e-01 5.97015560e-01 8.55173945e-01 -1.08131969e+00
-1.75944519e+00 -1.06200707e+00 1.55368596e-01 -1.18345141e-01
1.39504254e-01 -3.27152044e-01 -1.39336371e+00 5.39113224e-01
-1.08614177e-01 -8.00133348e-02 8.05538669e-02 -3.56352538e-01
-4.92780544e-02 -1.62606671e-01 -1.19400942e+00 3.75349611e-01
1.20916867e+00 -1.20458215e-01 -8.78877938e-01 1.10989273e-01
7.72449851e-01 -3.42675030e-01 -5.90240300e-01 6.63465142e-01
8.65388393e-01 -8.56758535e-01 6.69090450e-01 -7.56091952e-01
6.31428212e-02 -3.47323895e-01 -1.99226260e-01 -1.28849113e+00
-3.14578116e-01 -1.23127334e-01 -7.98220396e-01 7.48884737e-01
1.84823421e-03 -6.16046190e-01 7.65873849e-01 2.87759304e-01
-1.09799907e-01 -8.10539901e-01 -1.06764448e+00 -9.46932256e-01
-1.36019692e-01 4.68834117e-02 7.32815087e-01 4.50952470e-01
-3.06475699e-01 1.50778759e-02 -5.14535546e-01 3.81019622e-01
5.85988641e-01 4.85776991e-01 3.96960080e-01 -1.16041648e+00
-3.05240244e-01 -9.16513354e-02 -4.79869872e-01 -9.11938369e-01
1.59474105e-01 -5.47755420e-01 9.17425826e-02 -1.13637829e+00
5.83274066e-02 -3.82575184e-01 -6.16637886e-01 5.33720076e-01
2.23909780e-01 -2.28880823e-01 2.75689721e-01 4.33927178e-01
-4.55439091e-01 1.13050580e+00 1.14821029e+00 -4.79288995e-01
-6.06546462e-01 -6.63313363e-03 -7.13241026e-02 5.11391819e-01
8.11403632e-01 -3.58614624e-01 -5.58860123e-01 -6.73125863e-01
1.00988224e-01 9.50952768e-02 6.59964085e-01 -1.56594670e+00
5.94119608e-01 -1.27576113e-01 8.00062120e-01 -7.98108757e-01
5.38807273e-01 -1.13992274e+00 3.87551963e-01 8.15422595e-01
-1.11420304e-01 2.20600411e-01 3.73027653e-01 9.74104762e-01
-8.77336413e-02 -2.00335667e-01 6.47545874e-01 -5.45369864e-01
-9.15314019e-01 3.44538331e-01 -3.64814281e-01 -5.64932525e-01
1.31708550e+00 -3.96483600e-01 -1.19548127e-01 -1.29155055e-01
-1.26368999e+00 1.39225528e-01 4.36926454e-01 8.46394479e-01
1.08791602e+00 -1.29266322e+00 -1.82392061e-01 5.47423840e-01
-1.13566734e-01 5.94606362e-02 7.02942252e-01 3.89314264e-01
-1.99319214e-01 2.27551654e-01 -9.51935112e-01 -5.73384881e-01
-5.65550566e-01 1.11703515e+00 3.63267690e-01 8.70696753e-02
-7.61693239e-01 5.19333065e-01 -7.51570892e-03 -2.20996320e-01
4.76682544e-01 -2.06198663e-01 -1.44985318e-01 -7.76738822e-02
4.33392256e-01 3.40207845e-01 1.01238117e-02 -2.82434851e-01
-1.14158317e-01 5.38464963e-01 -3.58436197e-01 6.98017851e-02
1.69924045e+00 -9.66547132e-02 -4.79720905e-02 9.55451012e-01
6.73718870e-01 -9.48710918e-01 -1.49867153e+00 -2.48068660e-01
-1.79247811e-01 -2.21493661e-01 -2.58035213e-01 -9.61778402e-01
-9.48708892e-01 9.68992651e-01 1.09488463e+00 -9.80971530e-02
1.31286919e+00 -2.43443608e-01 9.18389261e-01 8.93777549e-01
8.20476592e-01 -1.37936759e+00 6.52572870e-01 7.21515834e-01
1.28184402e+00 -9.43855286e-01 -3.86360675e-01 3.55610967e-01
-3.84159863e-01 1.37819684e+00 1.04025817e+00 -4.21664596e-01
7.87284553e-01 1.46331832e-01 -4.95933205e-01 -1.95796102e-01
-9.77262378e-01 1.06771372e-01 -2.26926237e-01 5.42649686e-01
-4.33244467e-01 -2.09060624e-01 9.03670713e-02 8.01322222e-01
1.40323984e-02 2.60989010e-01 3.55960608e-01 1.43787301e+00
-7.98127711e-01 -6.87773466e-01 -1.63185135e-01 4.23482269e-01
2.58738071e-01 4.11718279e-01 -6.44313321e-02 6.25084817e-01
3.98552716e-01 4.23825502e-01 1.05437994e-01 -4.65645671e-01
3.82735044e-01 2.86355138e-01 7.15384483e-01 -3.13601255e-01
-3.06760013e-01 -4.54178214e-01 -8.02958250e-01 -4.12032455e-01
-3.18366468e-01 -6.03993058e-01 -1.29567075e+00 6.28823861e-02
-1.92876697e-01 2.56626606e-02 5.95847011e-01 6.94149137e-01
5.62512577e-01 6.34932458e-01 3.12136084e-01 -1.08220220e+00
-7.04376936e-01 -9.59404886e-01 -5.40833235e-01 4.12682593e-01
4.59317207e-01 -1.28600299e+00 -1.93700209e-01 7.75510669e-02] | [9.826327323913574, 3.4024224281311035] |
3968273a-cb38-4574-8948-c18927d97b01 | few-shot-speaker-identification-using-1 | 2305.19541 | null | https://arxiv.org/abs/2305.19541v1 | https://arxiv.org/pdf/2305.19541v1.pdf | Few-Shot Speaker Identification Using Lightweight Prototypical Network with Feature Grouping and Interaction | Existing methods for few-shot speaker identification (FSSI) obtain high accuracy, but their computational complexities and model sizes need to be reduced for lightweight applications. In this work, we propose a FSSI method using a lightweight prototypical network with the final goal to implement the FSSI on intelligent terminals with limited resources, such as smart watches and smart speakers. In the proposed prototypical network, an embedding module is designed to perform feature grouping for reducing the memory requirement and computational complexity, and feature interaction for enhancing the representational ability of the learned speaker embedding. In the proposed embedding module, audio feature of each speech sample is split into several low-dimensional feature subsets that are transformed by a recurrent convolutional block in parallel. Then, the operations of averaging, addition, concatenation, element-wise summation and statistics pooling are sequentially executed to learn a speaker embedding for each speech sample. The recurrent convolutional block consists of a block of bidirectional long short-term memory, and a block of de-redundancy convolution in which feature grouping and interaction are conducted too. Our method is compared to baseline methods on three datasets that are selected from three public speech corpora (VoxCeleb1, VoxCeleb2, and LibriSpeech). The results show that our method obtains higher accuracy under several conditions, and has advantages over all baseline methods in computational complexity and model size. | ['Qianhua He', 'Qisheng Huang', 'Wenchang Cao', 'Hao Chen', 'Yanxiong Li'] | 2023-05-31 | null | null | null | null | ['speaker-identification'] | ['speech'] | [ 6.72330335e-02 -2.17806384e-01 1.65329620e-01 -6.70175433e-01
-7.37465143e-01 -1.09883212e-01 3.59014302e-01 -2.52930254e-01
-5.23259163e-01 2.31579185e-01 2.93200642e-01 -1.66588724e-01
5.60478233e-02 -5.17997324e-01 -2.28054106e-01 -9.19349313e-01
-1.04387969e-01 -1.64880291e-01 3.26846927e-01 -1.63691148e-01
4.09157649e-02 3.76814127e-01 -1.77698863e+00 2.83219486e-01
6.95728183e-01 1.23542607e+00 2.24284902e-01 7.13599801e-01
-3.84314448e-01 7.16195762e-01 -8.18070710e-01 -1.56292453e-01
-1.10996746e-01 -1.56367630e-01 -4.69574958e-01 -9.39392745e-02
1.46859899e-01 -4.54373062e-01 -6.56635880e-01 9.69374955e-01
1.18254888e+00 5.28722882e-01 3.12340558e-01 -1.24634540e+00
-2.84805685e-01 9.51028824e-01 -2.04297632e-01 3.60621065e-01
1.64857537e-01 1.21242084e-01 9.87479270e-01 -9.22294080e-01
8.91036987e-02 1.41600287e+00 6.85021460e-01 6.25593662e-01
-9.00420547e-01 -9.39985216e-01 2.46132091e-01 5.31292021e-01
-1.62966347e+00 -1.11359131e+00 8.27248514e-01 -4.53613289e-02
1.28153419e+00 5.14383376e-01 3.41548771e-01 9.22929049e-01
-2.74520457e-01 9.66638088e-01 4.68401909e-01 -3.78191680e-01
3.04486573e-01 4.21205729e-01 6.32129431e-01 5.67524552e-01
-3.48756045e-01 9.81778000e-03 -8.39090705e-01 -1.62389979e-01
3.08585912e-01 3.20211090e-02 -1.41057089e-01 2.69860178e-01
-1.11663663e+00 7.51452923e-01 2.18222663e-01 4.44219559e-01
-2.86961555e-01 -8.53442624e-02 6.90157592e-01 4.81040001e-01
4.46530342e-01 -3.13877106e-01 -2.73531139e-01 -4.21518594e-01
-1.02448201e+00 -8.58043656e-02 9.58483040e-01 9.94913101e-01
5.49528062e-01 3.70149046e-01 -2.31337950e-01 1.24506485e+00
4.28975075e-01 3.17106873e-01 1.07355702e+00 -3.04906547e-01
7.35806286e-01 4.64764416e-01 -1.69236034e-01 -8.34931195e-01
-3.22020352e-01 -3.22780401e-01 -8.61404359e-01 -7.93248713e-02
-2.48412803e-01 -2.49691948e-01 -6.73423886e-01 1.62118757e+00
4.01847273e-01 4.77638304e-01 3.74007404e-01 7.84290314e-01
1.29285491e+00 1.17641103e+00 4.52647358e-02 -2.08463997e-01
1.55300844e+00 -1.10558319e+00 -9.24001336e-01 -9.42920521e-02
6.48318946e-01 -6.97075665e-01 1.25096512e+00 5.50218038e-02
-9.08032060e-01 -8.23206902e-01 -1.25197530e+00 -1.44272208e-01
-4.64539737e-01 5.30185401e-01 2.81363219e-01 9.51837063e-01
-1.08695161e+00 3.57514769e-01 -7.02487528e-01 -1.20882750e-01
2.38444343e-01 6.44086003e-01 -2.48015523e-01 3.79204243e-01
-1.38811469e+00 5.46292603e-01 3.55761051e-01 3.26563239e-01
-6.50796711e-01 -6.15876257e-01 -1.01657307e+00 6.38302088e-01
7.99559057e-02 -7.04498589e-03 1.20001328e+00 -7.36773312e-01
-2.00675702e+00 3.67886961e-01 -4.31364655e-01 -5.60277879e-01
-6.93419855e-03 7.40975738e-02 -8.36327493e-01 -1.30601197e-01
-3.45291138e-01 4.95688736e-01 1.06607759e+00 -5.16526699e-01
-7.23100662e-01 -3.94147336e-01 -1.33441035e-02 2.16385618e-01
-1.07666194e+00 4.87352014e-01 -3.31513673e-01 -4.98761058e-01
1.48732752e-01 -4.92198348e-01 4.17498918e-03 -3.23130578e-01
-3.84062678e-01 -5.33683300e-01 1.17089379e+00 -1.00313246e+00
1.78370500e+00 -2.78655839e+00 1.30466335e-02 8.17610472e-02
2.69007869e-02 5.89698374e-01 -1.45541236e-01 3.08517903e-01
5.58584668e-02 -9.76293832e-02 -7.50121176e-02 -6.85883343e-01
2.57275403e-01 -1.00583583e-01 -2.66374111e-01 3.57735246e-01
7.59726539e-02 5.11312485e-01 -4.54532802e-01 -6.26694083e-01
3.49718958e-01 7.18162417e-01 -3.85422319e-01 3.86361450e-01
2.21311957e-01 -9.21150818e-02 -2.50923783e-01 4.28952456e-01
7.16422915e-01 2.75934428e-01 -1.98115692e-01 -1.74460754e-01
-3.30798537e-01 7.82620192e-01 -1.55626857e+00 1.51648176e+00
-7.71501720e-01 6.02842808e-01 2.20560223e-01 -8.88227701e-01
1.05431902e+00 6.76010728e-01 1.80741489e-01 -3.40750933e-01
3.31717223e-01 5.58543690e-02 9.11717340e-02 -5.00607610e-01
3.55776012e-01 6.62444681e-02 -3.80341709e-02 6.47495568e-01
3.25079978e-01 2.24863186e-01 -8.91989022e-02 1.09307431e-01
9.27471757e-01 -6.09791756e-01 5.78675345e-02 -2.57837661e-02
1.05430067e+00 -6.27089679e-01 5.96727610e-01 3.58214676e-01
-3.36505026e-01 1.34071216e-01 1.85378060e-01 -3.10654104e-01
-8.11542332e-01 -9.15893793e-01 -1.54765276e-02 1.40125620e+00
-1.38213709e-01 -4.15533751e-01 -9.60629463e-01 -4.56400156e-01
-1.56398684e-01 6.43476367e-01 -2.29259863e-01 -3.93541455e-01
-5.92166305e-01 -4.52952862e-01 8.33397329e-01 5.59843838e-01
7.28102624e-01 -1.36716545e+00 -4.03466225e-01 4.11652774e-01
2.81439479e-02 -9.37659562e-01 -9.17046487e-01 1.43684313e-01
-6.58330858e-01 -3.67201865e-01 -5.07168174e-01 -1.16389644e+00
3.55243862e-01 3.55358452e-01 4.66980636e-01 -1.11727148e-01
-2.56085157e-01 -5.84896319e-02 -2.64880031e-01 -4.42004263e-01
-2.52688855e-01 3.97780806e-01 3.76235783e-01 3.64329189e-01
6.73177183e-01 -4.70651478e-01 -3.46673340e-01 2.85456955e-01
-7.44488120e-01 -1.95767760e-01 4.08257723e-01 1.00956547e+00
9.26942378e-02 3.59766595e-02 8.45750749e-01 -4.33076590e-01
6.69732988e-01 -2.83787787e-01 -5.75983465e-01 2.62290865e-01
-1.34611428e-01 -1.08188428e-01 7.15173364e-01 -5.91492593e-01
-1.29558349e+00 -2.08505336e-03 -5.73273957e-01 -2.42163673e-01
-1.80109534e-02 3.99366915e-01 -5.78321457e-01 -2.43427921e-02
3.04491609e-01 6.22682750e-01 2.79183775e-01 -7.54947841e-01
2.08769709e-01 1.64705896e+00 9.85460877e-02 3.59431617e-02
5.19081593e-01 1.50606081e-01 -7.84292877e-01 -1.38095355e+00
-3.37360710e-01 -5.23669720e-01 -5.36849439e-01 1.49737233e-02
5.55140674e-01 -1.00607455e+00 -9.59856033e-01 6.97962940e-01
-1.18017709e+00 6.18448518e-02 -1.94587439e-01 7.95720518e-01
-2.08669633e-01 2.32673004e-01 -7.29993820e-01 -1.09366488e+00
-8.77278030e-01 -1.22698593e+00 8.79163325e-01 4.00862724e-01
-3.01982947e-02 -6.98941708e-01 -1.82883114e-01 1.57952935e-01
7.55617678e-01 -7.79364109e-01 6.50239587e-01 -9.27207172e-01
-2.53388911e-01 -3.34509462e-01 -7.68492892e-02 8.37030590e-01
2.79003024e-01 -8.60655010e-02 -1.48291039e+00 -4.82991815e-01
2.28102282e-01 -7.73366466e-02 7.39829838e-01 2.66176134e-01
1.31009245e+00 -5.21166503e-01 -1.67094693e-01 7.06757367e-01
9.07554090e-01 5.10491788e-01 4.32304114e-01 -2.53494754e-02
6.17362499e-01 5.27815640e-01 2.79875368e-01 4.27799702e-01
2.79395878e-01 7.08249748e-01 -8.49913284e-02 9.86037999e-02
-2.18604833e-01 5.36312535e-02 8.54827106e-01 1.63206387e+00
2.83293217e-01 -1.19000124e-02 -4.55195725e-01 4.67982292e-01
-1.49541688e+00 -1.17954779e+00 4.58366066e-01 2.29174924e+00
7.85725594e-01 -9.07803550e-02 2.71552712e-01 4.27630484e-01
8.72826099e-01 4.67507958e-01 -5.34350216e-01 -5.12964070e-01
2.91974023e-02 2.46360138e-01 -3.07649318e-02 4.67955172e-01
-1.08645749e+00 1.09093344e+00 5.78873205e+00 1.06623101e+00
-1.39555097e+00 3.57850641e-01 3.70292336e-01 -4.17229354e-01
1.17137201e-01 -4.56468016e-01 -1.28923810e+00 6.22784913e-01
1.47073126e+00 -1.69336379e-01 5.40062726e-01 9.58823800e-01
1.65534586e-01 5.06026089e-01 -9.27271724e-01 1.29050112e+00
2.83762157e-01 -1.11135221e+00 -2.15214327e-01 -3.13434958e-01
1.39476284e-01 -4.07951102e-02 9.17299762e-02 5.93492270e-01
-1.33722514e-01 -7.83907712e-01 5.96507549e-01 1.46690100e-01
7.84243047e-01 -9.60572183e-01 7.52110898e-01 1.18821308e-01
-1.50983989e+00 -4.59458739e-01 -4.94710058e-01 7.64677003e-02
1.46470204e-01 3.20212364e-01 -8.25480223e-01 3.00792247e-01
7.48432636e-01 4.72193182e-01 -2.46844366e-01 8.60578239e-01
5.77329397e-02 8.31608653e-01 -4.24562395e-01 -4.89244640e-01
2.65373498e-01 -1.66494993e-03 4.47335780e-01 1.39648175e+00
3.21715713e-01 -4.02386067e-03 -2.29441792e-01 4.77676719e-01
-6.41222596e-02 2.85998642e-01 -4.26445216e-01 -1.54776618e-01
8.76808047e-01 1.34084690e+00 -1.68635935e-01 -6.23860240e-01
-5.89699507e-01 9.25083995e-01 2.48103991e-01 2.91294664e-01
-7.92470276e-01 -1.13160241e+00 9.78126943e-01 -3.27967942e-01
4.30694699e-01 -3.47975008e-02 -2.09383592e-01 -1.05564082e+00
2.27338657e-01 -8.55221272e-01 2.95637339e-01 -3.27097088e-01
-1.01692069e+00 9.60553288e-01 -3.37292165e-01 -1.20628464e+00
-3.51774395e-01 -4.23688143e-01 -9.84320819e-01 1.15569735e+00
-1.25912738e+00 -1.05636883e+00 -2.46373713e-01 7.03084171e-01
8.10738027e-01 -7.01860189e-01 9.56404626e-01 6.47810221e-01
-9.89115596e-01 1.06558228e+00 1.12930834e-01 1.67193264e-01
5.11476815e-01 -6.72894239e-01 5.48977077e-01 6.64337933e-01
1.69913486e-01 7.62333751e-01 3.19293916e-01 -1.41659975e-01
-1.36483836e+00 -1.05475771e+00 1.10897410e+00 3.06300312e-01
5.05289733e-01 -7.83949256e-01 -1.03236878e+00 5.68684101e-01
1.38245210e-01 -9.91194881e-03 8.37578535e-01 1.26029059e-01
-3.20289373e-01 -5.81422448e-01 -1.24560010e+00 5.45745254e-01
8.41932297e-01 -8.54131401e-01 -5.47582865e-01 1.42587334e-01
1.08842206e+00 -1.38164297e-01 -7.43328691e-01 -1.44158959e-01
6.89768612e-01 -5.84416032e-01 1.07302713e+00 -5.49403429e-01
-2.39168331e-01 -2.79147208e-01 -2.18078867e-01 -1.30954075e+00
-1.37684897e-01 -5.83707631e-01 -1.19919844e-01 1.56205595e+00
4.57197726e-01 -8.92815173e-01 5.25834203e-01 4.57779765e-01
-3.81120950e-01 -6.09133124e-01 -1.38175976e+00 -9.09539580e-01
-4.38374519e-01 -3.97449762e-01 1.08669436e+00 6.53218448e-01
2.39416569e-01 4.36696827e-01 -3.26603055e-01 3.43387842e-01
4.46043730e-01 -2.44898319e-01 6.88629329e-01 -1.06852841e+00
-1.61933422e-01 -3.20058286e-01 -6.58121526e-01 -9.58172202e-01
2.83414811e-01 -7.48826265e-01 2.50108372e-02 -1.16327488e+00
-3.69730145e-02 -4.36497778e-01 -4.84515190e-01 4.57042187e-01
-1.03068531e-01 -2.54896402e-01 6.61099032e-02 -5.00240177e-02
-3.73647571e-01 7.91963398e-01 5.00404060e-01 -3.47464949e-01
-6.28955305e-01 2.27665246e-01 -3.18202108e-01 5.36996424e-01
8.64864767e-01 -2.19728291e-01 -4.21316266e-01 -3.98303717e-01
-5.34059823e-01 -6.82339519e-02 1.60953358e-01 -1.22268295e+00
6.21018589e-01 3.70601714e-01 7.20987245e-02 -7.26076543e-01
8.18128765e-01 -6.51807487e-01 -1.19229041e-01 4.72305089e-01
-4.31563854e-01 -1.62819654e-01 2.03245595e-01 1.05084598e-01
-4.87613499e-01 -4.42317605e-01 5.49795628e-01 2.88451970e-01
-6.93598986e-01 4.23550934e-01 -3.84009182e-01 -4.80270952e-01
7.90618479e-01 -8.90493914e-02 -2.92270295e-02 -2.61033148e-01
-7.24090338e-01 2.12885235e-02 -2.76804894e-01 6.22392476e-01
8.27586532e-01 -1.56548238e+00 -6.03262722e-01 7.81357169e-01
1.43613024e-02 -2.15588495e-01 7.49382436e-01 6.42667592e-01
-1.61684930e-01 4.63841766e-01 3.02785784e-02 -3.97022724e-01
-1.68714190e+00 3.58260632e-01 1.87269717e-01 5.27255572e-02
-6.11636698e-01 1.13693464e+00 8.89384001e-03 -6.38880610e-01
7.93168783e-01 -3.64836693e-01 -4.27418083e-01 2.45057255e-01
1.11521077e+00 5.05768061e-01 1.38985112e-01 -8.17515492e-01
-4.24325675e-01 2.74808854e-01 -2.73082256e-01 -2.91921526e-01
1.40145278e+00 -2.48432219e-01 -6.67947158e-02 6.46839857e-01
1.47962332e+00 -1.75561383e-01 -9.24924970e-01 -5.71762502e-01
-1.75368145e-01 -2.46226713e-01 2.27517322e-01 -3.05266738e-01
-1.09332812e+00 1.32208991e+00 8.92384589e-01 1.63527116e-01
1.08049202e+00 -9.03512165e-02 1.08614182e+00 4.55544472e-01
1.08304657e-01 -1.24631751e+00 -1.93278432e-01 5.55248439e-01
8.27529907e-01 -1.02606320e+00 -5.08906960e-01 -2.73204088e-01
-7.52732515e-01 1.01894104e+00 4.95499641e-01 5.20160273e-02
9.19060528e-01 3.36237878e-01 5.07371798e-02 7.37690628e-02
-7.48033702e-01 -1.71932075e-02 1.57632411e-01 4.28327322e-01
4.12785947e-01 1.77811712e-01 -4.74209636e-02 1.20113003e+00
-4.27829802e-01 -2.71570534e-01 -3.31811011e-02 7.24175096e-01
-6.15510046e-01 -8.90912712e-01 -3.30028027e-01 4.20555681e-01
-2.05715671e-01 -3.03464472e-01 -2.87848320e-02 2.32005492e-01
-5.63095920e-02 1.14660335e+00 3.64822179e-01 -8.41565371e-01
3.76009226e-01 4.21842635e-01 -1.46715224e-01 -5.95043778e-01
-9.07440841e-01 -1.04676383e-02 2.65584793e-02 -3.39743346e-01
6.65444881e-02 -5.59359550e-01 -1.19755995e+00 -2.27382526e-01
-5.33066034e-01 2.67022103e-01 1.05227590e+00 8.42998862e-01
5.77481449e-01 6.85999691e-01 9.79972959e-01 -8.93528938e-01
-9.66837108e-01 -1.47628248e+00 -6.91220105e-01 7.32074901e-02
4.35641438e-01 -4.78312075e-01 -5.90294063e-01 -5.21404371e-02] | [14.358156204223633, 5.985490798950195] |
1dbe0431-138b-4ec1-9dc4-d52a62ec1cf0 | efficient-and-interpretable-neural-models-for | 2208.14252 | null | https://arxiv.org/abs/2208.14252v1 | https://arxiv.org/pdf/2208.14252v1.pdf | Efficient and Interpretable Neural Models for Entity Tracking | What would it take for a natural language model to understand a novel, such as The Lord of the Rings? Among other things, such a model must be able to: (a) identify and record new characters (entities) and their attributes as they are introduced in the text, and (b) identify subsequent references to the characters previously introduced and update their attributes. This problem of entity tracking is essential for language understanding, and thus, useful for a wide array of downstream applications in NLP such as question-answering, summarization. In this thesis, we focus on two key problems in relation to facilitating the use of entity tracking models: (i) scaling entity tracking models to long documents, such as a novel, and (ii) integrating entity tracking into language models. Applying language technologies to long documents has garnered interest recently, but computational constraints are a significant bottleneck in scaling up current methods. In this thesis, we argue that computationally efficient entity tracking models can be developed by representing entities with rich, fixed-dimensional vector representations derived from pretrained language models, and by exploiting the ephemeral nature of entities. We also argue for the integration of entity tracking into language models as it will allow for: (i) wider application given the current ubiquitous use of pretrained language models in NLP applications, and (ii) easier adoption since it is much easier to swap in a new pretrained language model than to integrate a separate standalone entity tracking model. | ['Shubham Toshniwal'] | 2022-08-30 | null | null | null | null | ['coreference-resolution'] | ['natural-language-processing'] | [ 1.58060759e-01 2.24527985e-01 -2.24726677e-01 -1.93640113e-01
-6.81035042e-01 -9.74134266e-01 8.39895785e-01 9.74061370e-01
-7.12298453e-01 8.86064053e-01 4.84220773e-01 -4.91656423e-01
-3.47804099e-01 -7.43989289e-01 -6.58444047e-01 -7.47608766e-02
-2.44120136e-01 7.86061108e-01 2.53379613e-01 -8.55041817e-02
1.60263658e-01 7.29412198e-01 -1.48157525e+00 1.50095165e-01
6.49344265e-01 4.27073300e-01 1.52893201e-01 6.96114898e-01
-8.27517092e-01 7.74402082e-01 -7.88596392e-01 -6.10311329e-01
-2.34644964e-01 -1.07876016e-02 -1.19813097e+00 -2.39004761e-01
3.52699846e-01 5.95404021e-02 -2.60535210e-01 6.21591985e-01
3.36194903e-01 5.26426911e-01 7.62737989e-01 -1.09377801e+00
-6.03368700e-01 8.78393233e-01 -3.62695336e-01 3.23107958e-01
6.68068886e-01 -4.92701083e-01 1.12688112e+00 -5.77160478e-01
1.11197662e+00 1.06349838e+00 8.21323454e-01 5.04100680e-01
-1.17251611e+00 -4.81541187e-01 4.29912359e-01 -6.65042782e-03
-1.31646979e+00 -5.78401804e-01 3.05762917e-01 -5.53487420e-01
1.34295857e+00 3.12730551e-01 3.34662050e-01 9.66647804e-01
-1.46159694e-01 9.15498078e-01 1.93915576e-01 -6.30775630e-01
8.30994397e-02 5.38316905e-01 5.90125203e-01 4.82300371e-01
5.82434237e-01 -5.51277220e-01 -3.74717981e-01 -4.03116137e-01
4.57649231e-01 -2.72655696e-01 -5.44493981e-02 -4.02414948e-01
-1.17347038e+00 8.03590417e-01 1.61028922e-01 7.67094493e-01
-4.35593277e-01 1.46529049e-01 6.26362681e-01 2.12924071e-02
2.46992826e-01 9.36686218e-01 -7.84059167e-01 -3.29713017e-01
-1.12070906e+00 4.04656529e-01 1.04578733e+00 1.06775522e+00
6.49528086e-01 -2.05855340e-01 2.25138021e-04 7.76812494e-01
2.36898959e-01 1.50837243e-01 7.04812646e-01 -7.01364815e-01
5.72535813e-01 8.27447712e-01 2.82414794e-01 -9.91454065e-01
-5.24148464e-01 -4.11139101e-01 -5.52162111e-01 -4.61956382e-01
2.88205028e-01 -1.95887163e-01 -7.22426116e-01 1.84648502e+00
3.13055694e-01 3.58130373e-02 3.47736657e-01 -1.72776170e-02
9.85840678e-01 9.62101579e-01 4.14897084e-01 -4.13807988e-01
1.43034327e+00 -4.89432991e-01 -6.32602692e-01 -3.00408721e-01
1.15481448e+00 -7.07117736e-01 4.63443696e-01 -7.22163469e-02
-9.54540968e-01 -4.53746170e-01 -7.66875923e-01 -4.04072851e-01
-9.90444005e-01 1.03294216e-01 9.31385636e-01 5.21595836e-01
-8.15140724e-01 4.24043059e-01 -7.51931548e-01 -8.29494178e-01
3.66559476e-02 4.65755612e-01 -5.36334813e-01 2.86065191e-01
-1.41561437e+00 1.11243868e+00 7.93795645e-01 -1.95155684e-02
6.61498681e-02 -7.46202290e-01 -1.20828187e+00 1.80154979e-01
4.07004803e-01 -8.44185710e-01 1.15444577e+00 -6.50082052e-01
-7.57669806e-01 7.49002695e-01 -5.56332231e-01 -4.93245244e-01
1.65893987e-01 -3.67410392e-01 -6.56100094e-01 -1.66302800e-01
2.89370060e-01 4.67766315e-01 3.02189440e-01 -1.19110501e+00
-8.75383437e-01 -3.51080984e-01 1.21360235e-01 3.26894671e-01
-5.50776005e-01 1.48958877e-01 -7.96458066e-01 -5.15882432e-01
-7.69743770e-02 -7.08219349e-01 -8.36709738e-02 -3.44641387e-01
-2.84753203e-01 -4.50664014e-01 6.85010970e-01 -7.33530283e-01
1.80751860e+00 -1.90119028e+00 8.59407261e-02 2.17150450e-01
1.37172818e-01 5.57691514e-01 5.69521338e-02 8.06903124e-01
-6.52651787e-02 4.20490384e-01 -1.98712945e-02 -2.37736002e-01
1.41207263e-01 1.77796394e-01 -5.75561345e-01 3.24743167e-02
8.92000447e-04 8.82372558e-01 -7.99675107e-01 -5.39561450e-01
5.12722358e-02 4.93912876e-01 -3.44902396e-01 -2.80822605e-01
-1.66832358e-01 -2.26192430e-01 -5.12533903e-01 4.04716402e-01
2.17272043e-01 -1.81224406e-01 2.94117272e-01 -1.64831981e-01
-2.44760662e-01 3.44601750e-01 -1.59166431e+00 1.34307408e+00
-4.30668443e-01 8.91151845e-01 -5.16170375e-02 -7.79635072e-01
8.13321412e-01 4.24755037e-01 5.05771577e-01 -3.59214187e-01
-8.71486142e-02 1.41205087e-01 -3.40596229e-01 -6.06385291e-01
1.25513041e+00 -9.23491940e-02 -4.24528092e-01 5.25631785e-01
1.01055168e-01 2.06882372e-01 7.52752900e-01 6.53576374e-01
9.15303528e-01 -4.88457317e-03 5.77855825e-01 8.45328420e-02
7.93151736e-01 1.53106079e-01 4.45161253e-01 8.70407820e-01
2.40280777e-01 1.07635416e-01 2.68154621e-01 -3.73989046e-01
-9.50116158e-01 -8.19268465e-01 -3.32088202e-01 1.25802076e+00
-2.24221557e-01 -6.73966646e-01 -3.24585170e-01 -6.45678341e-01
4.18201029e-01 1.05925691e+00 -5.50884068e-01 3.70562859e-02
-7.63558269e-01 -5.41557729e-01 8.51521134e-01 7.41548061e-01
-9.04232338e-02 -9.75573480e-01 -4.67109591e-01 4.67911392e-01
-1.69848800e-01 -9.70450342e-01 -1.80736240e-02 3.77512604e-01
-6.99688554e-01 -8.57072532e-01 -6.84885621e-01 -8.76217604e-01
5.36506414e-01 -1.29282519e-01 1.04304707e+00 1.40263513e-01
-6.05338439e-02 7.99767375e-01 -3.68274629e-01 -5.82359850e-01
-4.35101688e-01 5.71480751e-01 1.51021406e-01 -2.83665836e-01
4.22737420e-01 -1.51499793e-01 8.96924268e-03 -8.36983249e-02
-1.18635046e+00 -3.00572932e-01 3.62614185e-01 6.40213609e-01
2.97543228e-01 5.62708266e-02 6.46899760e-01 -1.37239051e+00
8.50645304e-01 -6.11874104e-01 -3.71792406e-01 7.69447148e-01
-6.55644178e-01 1.40885517e-01 4.23931211e-01 -3.89238983e-01
-1.07685018e+00 -6.37110919e-02 -2.99211025e-01 1.62292093e-01
-6.85709268e-02 9.63163435e-01 3.40805985e-02 3.66143256e-01
5.84247410e-01 2.05637336e-01 -3.68672609e-01 -5.29188991e-01
7.32832193e-01 6.68300211e-01 5.77467144e-01 -7.23455191e-01
8.80863905e-01 1.92173675e-01 -1.92680031e-01 -1.18459666e+00
-5.78763843e-01 -9.01017666e-01 -9.91563141e-01 1.46725193e-01
5.87290168e-01 -9.24495578e-01 -5.34189403e-01 1.03771642e-01
-1.29156923e+00 5.83177917e-02 -5.71718812e-01 4.08845007e-01
-2.91276693e-01 5.59512556e-01 -3.89167070e-01 -7.01017201e-01
-3.05796176e-01 -5.17195463e-01 8.90281856e-01 5.26003242e-01
-7.76998758e-01 -1.41379130e+00 2.11454317e-01 2.65087664e-01
2.47423172e-01 6.16784990e-02 1.35402203e+00 -1.34489357e+00
-2.34977514e-01 -5.93713522e-01 -5.18931225e-02 -6.98593855e-02
1.79535553e-01 2.71805108e-01 -6.56363189e-01 -3.05983245e-01
-4.50645238e-01 -3.98272872e-02 5.65594137e-01 1.47219017e-01
6.23519480e-01 -2.36460716e-01 -8.23992729e-01 3.91793936e-01
1.12312984e+00 2.81786442e-01 4.53878820e-01 6.18261814e-01
5.55494428e-01 7.04932451e-01 3.15309018e-01 2.90739536e-01
5.91259837e-01 6.06776714e-01 -3.10905784e-01 1.07395805e-01
3.14458609e-02 -5.07910430e-01 8.80825147e-02 8.58368337e-01
3.22201818e-01 -5.29168665e-01 -1.09322071e+00 8.73399496e-01
-1.87288916e+00 -1.18300772e+00 -1.84032485e-01 2.20150685e+00
9.01937723e-01 5.95385395e-02 -9.10901278e-02 -1.26724735e-01
6.69703364e-01 1.76676467e-01 -5.82138598e-01 -5.90502381e-01
-1.31252214e-01 1.46513835e-01 2.64996439e-01 4.43901271e-01
-1.10904217e+00 1.07618558e+00 6.63154364e+00 5.46854556e-01
-9.55862105e-01 -1.68348938e-01 1.57201409e-01 1.11671396e-01
-3.73423040e-01 1.70737907e-01 -1.38834333e+00 2.46872827e-01
1.06979561e+00 -9.43275452e-01 -4.76850942e-02 6.94892764e-01
-1.17765866e-01 -1.52244881e-01 -1.27009642e+00 7.67376959e-01
3.10063064e-01 -1.52631795e+00 3.97857934e-01 -9.38106515e-03
5.35259008e-01 -3.43205035e-01 -3.64887595e-01 7.23081470e-01
4.10205275e-01 -8.87543559e-01 4.24861431e-01 5.66595376e-01
4.51256305e-01 -5.63937306e-01 6.45548522e-01 4.00618017e-01
-1.27180731e+00 -4.57290486e-02 -3.55778784e-01 2.14665413e-01
4.43457276e-01 2.21978083e-01 -9.80773926e-01 8.03010106e-01
3.74001443e-01 5.89295626e-01 -7.05535769e-01 1.12232661e+00
-8.10866356e-02 3.12674433e-01 -4.05714720e-01 -1.93135098e-01
2.49496460e-01 1.50198445e-01 5.65130115e-01 1.49870169e+00
2.54214197e-01 -7.57473111e-02 -1.35173053e-01 4.90445733e-01
-3.26192290e-01 3.35736722e-01 -5.81125617e-01 -4.52414215e-01
7.11021602e-01 1.07444513e+00 -7.59016514e-01 -5.92142463e-01
-4.14149314e-01 7.14664221e-01 3.09226513e-01 3.43218893e-01
-6.10442162e-01 -5.94426334e-01 5.21556199e-01 1.22925758e-01
4.35737610e-01 -3.16666603e-01 -6.11816347e-02 -1.22396076e+00
-4.11549248e-02 -7.43379831e-01 7.66232848e-01 -7.40287900e-01
-1.07927024e+00 4.80154514e-01 1.91235512e-01 -8.05635691e-01
-5.61151206e-01 -3.75240505e-01 -3.56606483e-01 9.11272645e-01
-1.22227669e+00 -1.10091341e+00 1.73133805e-01 4.00575131e-01
4.93489653e-01 2.03432404e-02 1.00477099e+00 3.56433332e-01
-6.35043800e-01 6.36689842e-01 4.22516525e-01 3.82059276e-01
6.88252568e-01 -1.22885442e+00 5.37193716e-01 9.40123081e-01
6.56799734e-01 1.35150433e+00 5.94188690e-01 -7.97953367e-01
-1.24555159e+00 -1.07652760e+00 1.63148904e+00 -8.41185391e-01
8.26502919e-01 -2.24414572e-01 -1.25415421e+00 1.04298735e+00
-1.01441950e-01 -4.74311680e-01 8.98082852e-01 5.64352632e-01
-2.77450651e-01 9.07333344e-02 -8.75784159e-01 4.69952673e-01
6.20346963e-01 -6.11396492e-01 -1.13846910e+00 3.94973338e-01
5.58291972e-01 -3.31699252e-01 -8.92004251e-01 1.26763493e-01
7.15532482e-01 -2.08937317e-01 1.02992415e+00 -1.08005524e+00
-7.32698515e-02 -3.75139892e-01 -2.55541392e-02 -9.88348484e-01
-4.14568216e-01 -5.10510147e-01 -3.98431510e-01 1.85060740e+00
7.73548484e-01 -5.92556477e-01 6.98970616e-01 1.12897003e+00
2.56602079e-01 -5.40530562e-01 -7.91241229e-01 -8.74072433e-01
3.10173303e-01 -3.43620777e-01 5.44962704e-01 1.16183734e+00
5.62799424e-02 5.50253391e-01 -2.59427056e-02 2.33617634e-01
3.07435207e-02 1.40437623e-03 7.98939466e-01 -1.56278753e+00
4.03755382e-02 -4.64391351e-01 -5.05994916e-01 -1.16548407e+00
1.58629090e-01 -9.86846924e-01 -1.97342351e-01 -2.08051324e+00
8.58606100e-02 -6.57289743e-01 -1.45434201e-01 5.47672689e-01
-1.33943453e-01 -3.49812061e-01 2.03122497e-01 3.87585938e-01
-5.68466723e-01 2.29915366e-01 2.79523462e-01 -8.33638534e-02
-4.29446369e-01 -8.82894616e-04 -1.00767708e+00 7.38107026e-01
4.44940627e-01 -5.74870884e-01 -3.22334349e-01 -5.85173249e-01
6.15066707e-01 2.47372836e-02 -1.36627838e-01 -8.80297780e-01
8.12607884e-01 4.63007800e-02 5.18251181e-01 -6.46102726e-01
3.37434113e-01 -7.79269576e-01 2.53754884e-01 1.88214853e-01
-4.86680418e-01 3.27212989e-01 5.70859432e-01 6.20280325e-01
-2.43554860e-01 -6.85723662e-01 2.06207335e-01 -2.27417096e-01
-1.13833869e+00 -4.40360010e-02 -5.29329479e-01 2.10726872e-01
1.06126559e+00 -4.14992750e-01 -3.85700226e-01 -1.66806549e-01
-7.16395557e-01 3.49483430e-01 5.22363961e-01 6.67850018e-01
1.73431963e-01 -9.87581849e-01 -3.67285222e-01 -8.05971399e-02
1.78931504e-01 -2.26785354e-02 4.17743884e-02 3.46235365e-01
-4.15838152e-01 6.60946310e-01 1.24506310e-01 -1.96332425e-01
-1.44858658e+00 4.39089328e-01 -4.99028713e-02 -5.05066514e-01
-5.32482803e-01 8.72661471e-01 -8.95439610e-02 -5.39587855e-01
4.01077926e-01 -1.80941314e-01 -6.25773907e-01 7.60307431e-01
5.91172278e-01 2.19603226e-01 2.04993457e-01 -8.92624915e-01
-6.32343888e-01 5.31087935e-01 -5.45606017e-01 -1.12931188e-02
1.53402376e+00 -2.77785271e-01 -1.04794994e-01 5.98726451e-01
9.54977930e-01 4.00802761e-01 -4.81639862e-01 -3.70287627e-01
6.30942464e-01 -3.21816318e-02 -1.44952416e-01 -7.80990660e-01
-3.36426705e-01 6.06324732e-01 8.91820788e-02 3.24309319e-01
6.67335272e-01 1.51977047e-01 7.77815938e-01 8.67103159e-01
2.23075941e-01 -1.01886356e+00 -5.11933208e-01 6.73828721e-01
5.88901222e-01 -7.75652111e-01 4.87950802e-01 -8.45921338e-02
-6.78497136e-01 1.22150385e+00 2.34577104e-01 5.22946119e-01
3.92472357e-01 1.03763342e-01 -4.96104881e-02 -3.83564979e-01
-7.26839900e-01 -2.50279605e-01 3.00277799e-01 5.93948781e-01
6.08081400e-01 -1.57196745e-01 -2.42044702e-01 5.77087581e-01
-2.07807183e-01 -5.36369048e-02 6.89807475e-01 1.11977911e+00
-4.41855133e-01 -1.28294861e+00 -1.97796121e-01 6.25223935e-01
-4.81656879e-01 -1.53000519e-01 -5.75262427e-01 1.07665217e+00
2.14427188e-01 8.31689894e-01 1.73985690e-01 7.15676546e-02
5.28472781e-01 6.00580871e-01 9.02200267e-02 -1.01320791e+00
-7.34775126e-01 -3.97239566e-01 3.56458485e-01 8.85484666e-02
-3.26321781e-01 -1.06415844e+00 -1.24241161e+00 -1.81093812e-01
-4.48929131e-01 4.91101563e-01 8.64486873e-01 9.54615116e-01
5.46689987e-01 3.56330782e-01 1.56662520e-02 -2.87846953e-01
-2.74325043e-01 -7.36514032e-01 -4.07700151e-01 3.86691749e-01
7.22837597e-02 -5.16131103e-01 -1.75334960e-01 1.82137847e-01] | [9.490523338317871, 9.045074462890625] |
d2927675-2600-40d9-aaff-0b68c2e7ceed | retrieval-enhanced-visual-prompt-learning-for | 2306.02243 | null | https://arxiv.org/abs/2306.02243v1 | https://arxiv.org/pdf/2306.02243v1.pdf | Retrieval-Enhanced Visual Prompt Learning for Few-shot Classification | Prompt learning has become a popular approach for adapting large vision-language models, such as CLIP, to downstream tasks. Typically, prompt learning relies on a fixed prompt token or an input-conditional token to fit a small amount of data under full supervision. While this paradigm can generalize to a certain range of unseen classes, it may struggle when domain gap increases, such as in fine-grained classification and satellite image segmentation. To address this limitation, we propose Retrieval-enhanced Prompt learning (RePrompt), which introduces retrieval mechanisms to cache the knowledge representations from downstream tasks. we first construct a retrieval database from training examples, or from external examples when available. We then integrate this retrieval-enhanced mechanism into various stages of a simple prompt learning baseline. By referencing similar samples in the training set, the enhanced model is better able to adapt to new tasks with few samples. Our extensive experiments over 15 vision datasets, including 11 downstream tasks with few-shot setting and 4 domain generalization benchmarks, demonstrate that RePrompt achieves considerably improved performance. Our proposed approach provides a promising solution to the challenges faced by prompt learning when domain gap increases. The code and models will be available. | ['Yifan Liu', 'Xinyi Yu', 'Linlin Ou', 'Tianxiao Chen', 'Hao Chen', 'Jintao Rong'] | 2023-06-04 | null | null | null | null | ['domain-generalization'] | ['methodology'] | [ 2.27958843e-01 -3.72074842e-01 -5.57484627e-01 -6.35408401e-01
-1.03812122e+00 -6.46593690e-01 7.64784217e-01 1.15113877e-01
-6.41468704e-01 6.37622237e-01 6.32622689e-02 -3.71036679e-02
-1.69558004e-02 -7.05660343e-01 -7.54259467e-01 -5.93344808e-01
1.68366343e-01 4.67225730e-01 8.20446372e-01 -1.42326683e-01
2.76741803e-01 4.27157462e-01 -1.68403351e+00 3.58123928e-01
9.15639281e-01 1.04335153e+00 6.88087404e-01 4.79898602e-01
-4.39104676e-01 6.42043352e-01 -5.39622128e-01 -2.43671928e-02
4.27188009e-01 3.10444515e-02 -8.85934114e-01 3.70308943e-02
5.76145828e-01 -6.07441247e-01 -2.45383739e-01 8.77335072e-01
5.29853463e-01 5.65666258e-01 7.92256951e-01 -1.27800190e+00
-1.06189513e+00 4.87806499e-01 -5.07517755e-01 4.16468203e-01
9.18365642e-02 4.46288347e-01 6.08071744e-01 -1.24497688e+00
6.85343146e-01 1.09952092e+00 5.18603921e-01 9.30069685e-01
-1.00212359e+00 -7.91548252e-01 5.78250170e-01 3.62004519e-01
-1.32315016e+00 -5.78527391e-01 4.91131246e-01 -3.50400269e-01
9.68757749e-01 -2.50163496e-01 2.54426837e-01 1.22003698e+00
-2.82253027e-01 1.00758648e+00 9.59498405e-01 -2.92623371e-01
4.69303042e-01 1.83896124e-01 5.39878249e-01 1.97643667e-01
-4.59094085e-02 2.28689849e-01 -6.34899080e-01 -1.59771368e-01
5.50225019e-01 2.87348568e-01 -2.94269413e-01 -2.51637965e-01
-1.01717162e+00 6.63477361e-01 5.72727382e-01 -2.11447343e-01
-3.19586337e-01 8.65814183e-03 5.36676049e-01 5.06851971e-01
6.04937971e-01 1.51980042e-01 -7.11854994e-01 8.63101855e-02
-8.98733914e-01 3.14644516e-01 5.88810623e-01 1.52821159e+00
1.01221383e+00 -8.01603198e-02 -5.06626844e-01 1.00131583e+00
2.40397722e-01 5.74859858e-01 6.79883420e-01 -8.19893956e-01
4.58660543e-01 3.72463137e-01 1.16433255e-01 -2.30873168e-01
3.29528786e-02 -2.93799311e-01 -6.18381560e-01 -4.11459571e-03
2.04455480e-01 -8.37840587e-02 -1.43642807e+00 1.73114896e+00
4.63700503e-01 6.13382041e-01 3.67377281e-01 9.17139530e-01
1.09651685e+00 9.25743222e-01 5.20734429e-01 -1.04267910e-01
1.08843577e+00 -1.38552964e+00 -2.44450256e-01 -5.66926241e-01
4.72852111e-01 -5.82035780e-01 1.36585653e+00 1.03899315e-01
-6.67279899e-01 -7.70424247e-01 -8.16797256e-01 -1.05821319e-01
-5.27236402e-01 -1.83677927e-01 4.23546404e-01 -5.72799239e-03
-9.71959710e-01 4.83345896e-01 -7.43937969e-01 -6.74251974e-01
5.55409431e-01 1.16615266e-01 -1.99181914e-01 -5.41418016e-01
-1.10602868e+00 5.96291363e-01 4.81443107e-01 -2.18741477e-01
-1.26383162e+00 -9.72061396e-01 -7.80129492e-01 4.27815318e-02
5.03627539e-01 -5.40084720e-01 1.71871150e+00 -9.36676204e-01
-1.33932137e+00 7.41063356e-01 -4.80307229e-02 -5.98776937e-01
2.85570085e-01 -4.24791932e-01 -2.69861281e-01 1.29001871e-01
2.77459413e-01 1.04182601e+00 1.27080500e+00 -1.07966399e+00
-6.45408332e-01 -2.49394447e-01 9.37384814e-02 3.04486185e-01
-4.00197834e-01 -1.17606178e-01 -6.85392380e-01 -5.23076653e-01
-1.85576051e-01 -7.81550825e-01 -3.96243364e-01 2.32417583e-01
6.51090145e-02 -6.23518109e-01 1.00287747e+00 -9.81060416e-02
8.39706957e-01 -2.24349952e+00 -1.94185421e-01 -2.34491274e-01
-1.38492525e-01 5.50892055e-01 -5.83497465e-01 3.79730105e-01
1.19031169e-01 -1.34861514e-01 -1.56959206e-01 -2.34781444e-01
-1.23044826e-01 4.85714078e-01 -6.77204430e-01 3.73175293e-02
5.57313323e-01 8.15139711e-01 -1.12861824e+00 -4.39808488e-01
-2.52489522e-02 8.05863440e-02 -5.20112216e-01 5.03196120e-01
-6.66856587e-01 2.36755654e-01 -6.87721312e-01 7.89444327e-01
6.98005617e-01 -4.60913748e-01 -3.21437836e-01 5.81161045e-02
4.42864336e-02 -3.84734501e-03 -9.91500497e-01 2.07428360e+00
-3.64667982e-01 3.40755552e-01 -1.42764464e-01 -1.17475784e+00
1.19368660e+00 6.28749356e-02 8.45967457e-02 -7.43583202e-01
-2.63727695e-01 4.41437624e-02 -4.12282228e-01 -6.69229150e-01
6.55202508e-01 2.32009795e-02 -1.48327067e-01 4.52697575e-01
2.74388462e-01 -8.29136521e-02 1.16175301e-01 2.89515167e-01
1.06640160e+00 4.82605547e-01 2.89446980e-01 -7.07613006e-02
3.50610048e-01 4.90178436e-01 7.14925587e-01 1.17120600e+00
-4.99037594e-01 5.66720426e-01 -1.46918222e-01 -4.53573883e-01
-7.53677726e-01 -1.13709188e+00 -5.69173582e-02 1.76068473e+00
2.52090186e-01 -2.47867256e-01 -3.39004755e-01 -8.68155718e-01
1.46909922e-01 5.08200169e-01 -3.93894166e-01 -4.85377818e-01
-3.10665280e-01 -4.46889549e-01 2.63533503e-01 8.85576189e-01
6.99485958e-01 -1.28130436e+00 -5.00405073e-01 1.98887706e-01
7.37445876e-02 -1.12014556e+00 -5.07136106e-01 4.79300499e-01
-1.02035367e+00 -8.25779438e-01 -8.27397764e-01 -1.00835192e+00
5.85051179e-01 7.56200612e-01 1.13295984e+00 -3.11972946e-02
-2.35865295e-01 5.97859025e-01 -6.93029344e-01 -4.24083948e-01
-1.40038431e-01 2.58510202e-01 1.40601903e-01 -2.87501127e-01
8.56534541e-01 -3.48005265e-01 -5.98631859e-01 1.68055803e-01
-1.02867842e+00 -2.07459152e-01 6.30742610e-01 1.11199129e+00
8.72589469e-01 -4.39344287e-01 8.96830320e-01 -9.49722409e-01
5.69348514e-01 -7.77285278e-01 -5.90333521e-01 4.13665950e-01
-6.66247725e-01 9.16896760e-02 8.55336249e-01 -8.67391109e-01
-1.20467603e+00 2.75681257e-01 6.34645075e-02 -8.36920738e-01
-3.93980294e-01 5.74411392e-01 4.30268571e-02 1.59709215e-01
1.03274179e+00 3.68978560e-01 -1.92795634e-01 -6.59071565e-01
6.83025956e-01 9.87746656e-01 8.31351757e-01 -9.80027735e-01
8.22236657e-01 3.33791226e-01 -5.68318069e-01 -6.46213591e-01
-1.19883716e+00 -7.89147198e-01 -6.18652225e-01 4.28585000e-02
5.88661373e-01 -1.20399344e+00 -8.91632438e-02 4.64959979e-01
-9.98707831e-01 -6.99226975e-01 -5.43062091e-01 2.63877690e-01
-4.70968097e-01 1.34708971e-01 -3.38957757e-01 -5.27628362e-01
-5.16050637e-01 -9.44588244e-01 1.18856788e+00 6.90143347e-01
1.55372262e-01 -8.82134557e-01 9.18555781e-02 1.26639217e-01
5.47542751e-01 -2.19384998e-01 7.37929523e-01 -1.00077856e+00
-6.41134799e-01 -1.99490059e-02 -3.96639466e-01 4.50848997e-01
1.12111546e-01 -1.90316528e-01 -1.10771883e+00 -4.18601871e-01
-2.26423651e-01 -1.09007633e+00 1.20394945e+00 1.16489194e-01
1.35147846e+00 -4.59371693e-02 -5.06712973e-01 6.85623109e-01
1.53807950e+00 7.80149698e-02 2.00667515e-01 4.57079321e-01
3.48135233e-01 4.84435201e-01 1.24074125e+00 3.97327840e-01
4.08288658e-01 4.12029952e-01 2.06544191e-01 1.22562103e-01
-1.97352454e-01 -3.81370664e-01 3.60967785e-01 5.34402132e-01
3.11350793e-01 -5.68717495e-02 -1.01246810e+00 7.65079379e-01
-1.97239077e+00 -7.75732100e-01 3.22622746e-01 1.96068513e+00
9.56763327e-01 7.10167512e-02 -1.01018921e-01 -4.74304706e-01
6.03208423e-01 8.19938350e-03 -1.24319100e+00 -1.06907524e-01
1.88008323e-01 2.96554655e-01 3.27501833e-01 1.94827393e-01
-1.06206191e+00 1.42782557e+00 6.11865759e+00 8.55916619e-01
-1.19422400e+00 1.84816003e-01 2.68280566e-01 -5.28812483e-02
-1.72653958e-01 5.86168058e-02 -1.16933405e+00 3.11246842e-01
8.43288839e-01 -3.91039521e-01 1.72759041e-01 1.25209522e+00
8.66903067e-02 -6.67718202e-02 -1.34183407e+00 9.89527583e-01
5.44202700e-02 -1.36817658e+00 3.92887861e-01 -5.32293558e-01
7.41774857e-01 5.01744032e-01 3.56762297e-02 9.63859677e-01
4.44998682e-01 -7.40709364e-01 3.71860206e-01 4.77203667e-01
8.93503010e-01 -4.22769696e-01 4.20548946e-01 6.13536775e-01
-1.08298433e+00 -2.09504589e-01 -8.69609952e-01 -1.72477812e-02
-7.48679191e-02 1.20055802e-01 -9.59912598e-01 3.95433068e-01
9.60069180e-01 9.45783854e-01 -6.72031224e-01 1.16790104e+00
-1.60971761e-01 5.45237303e-01 -1.69194832e-01 9.54777077e-02
4.24705625e-01 6.66902214e-02 2.95382857e-01 1.26764071e+00
2.23634779e-01 3.36468935e-01 6.27546608e-01 5.69874465e-01
-8.95978808e-02 1.04646022e-02 -7.65612245e-01 6.13770112e-02
9.11190569e-01 1.12342393e+00 -3.82354468e-01 -6.07307673e-01
-6.27480626e-01 1.06386817e+00 6.28745198e-01 6.52622521e-01
-4.96212304e-01 -3.27058643e-01 7.15860248e-01 -1.81156993e-01
3.86397213e-01 -3.47345113e-03 2.19413280e-01 -1.17817998e+00
-6.24128245e-02 -7.26887941e-01 6.43563807e-01 -9.31568384e-01
-1.70655537e+00 5.16370058e-01 2.60931015e-01 -1.32078731e+00
-2.74717510e-01 -5.58444023e-01 -7.00224161e-01 7.45024562e-01
-2.12045431e+00 -1.23000872e+00 -6.45203412e-01 9.50075448e-01
1.21965992e+00 -3.56463075e-01 7.83922732e-01 6.69087321e-02
-3.83648813e-01 5.04984260e-01 1.91328302e-01 2.81794611e-02
1.21607244e+00 -1.19033682e+00 3.78997982e-01 7.17799425e-01
-3.31421643e-02 6.20526850e-01 4.21063721e-01 -5.47709167e-01
-1.19076276e+00 -1.63114202e+00 5.21750748e-01 -3.09055924e-01
5.89918137e-01 -2.02354908e-01 -1.39252520e+00 8.35847080e-01
1.10099383e-01 5.50071239e-01 6.44569099e-01 4.32213433e-02
-5.42146802e-01 -3.49026144e-01 -9.70634758e-01 4.34899777e-01
1.26264644e+00 -5.05629420e-01 -9.92346108e-01 5.63720882e-01
1.15993571e+00 -3.88477921e-01 -7.31767952e-01 3.25995833e-01
2.09598526e-01 -5.02475858e-01 9.22386765e-01 -9.58792806e-01
3.16954792e-01 -2.35833943e-01 -2.02235371e-01 -1.37287104e+00
-3.75507116e-01 -4.01590437e-01 -1.62262365e-01 1.41320908e+00
2.88178772e-01 -3.81748855e-01 7.03971386e-01 5.93181789e-01
-3.51970315e-01 -5.57851970e-01 -6.34524047e-01 -1.18898952e+00
2.54669189e-01 -4.15855676e-01 6.13874078e-01 7.73124516e-01
-3.74141812e-01 5.21839559e-01 -2.09727407e-01 2.69851506e-01
6.17181242e-01 3.45172286e-01 9.92168188e-01 -1.29838419e+00
-3.14306796e-01 4.18183319e-02 4.30314913e-02 -1.56480408e+00
2.76530087e-01 -1.05941355e+00 3.78618300e-01 -1.48860335e+00
4.13426727e-01 -8.00346494e-01 -5.52094102e-01 7.61962891e-01
-3.07006121e-01 2.35367026e-02 1.43525109e-01 5.38528323e-01
-9.41111088e-01 7.15478837e-01 1.12919581e+00 -4.02936220e-01
-3.89304399e-01 7.10408539e-02 -7.41732597e-01 6.28766119e-01
6.96833134e-01 -5.81292570e-01 -8.07008266e-01 -7.57163763e-01
-2.09673300e-01 -1.59329236e-01 2.97105223e-01 -8.56852889e-01
5.72815299e-01 -4.47378486e-01 3.25963706e-01 -6.04155838e-01
1.65685356e-01 -7.80162632e-01 -4.05914545e-01 1.28835097e-01
-4.23949599e-01 -1.65999666e-01 3.62474769e-01 8.94342363e-01
-1.84447214e-01 -4.53355163e-01 7.73802698e-01 -3.05791765e-01
-1.56971753e+00 6.99221849e-01 -1.45356759e-01 3.23977888e-01
1.08178830e+00 -2.27113858e-01 -6.16768003e-01 -2.60998178e-02
-8.08361650e-01 7.44006336e-01 4.80636567e-01 7.61712253e-01
8.40784371e-01 -1.30019009e+00 -7.13214457e-01 1.62322313e-01
7.87524521e-01 5.30939519e-01 1.96500316e-01 3.95551085e-01
-9.48198363e-02 1.47598207e-01 -1.92557871e-01 -9.78430212e-01
-1.00823057e+00 8.01827431e-01 2.98409443e-02 -2.08867025e-02
-8.15803885e-01 9.78787839e-01 2.90868372e-01 -5.06868660e-01
4.62052047e-01 -3.54784310e-01 -2.15527490e-01 3.78875174e-02
7.78085530e-01 2.77697900e-03 -1.11396402e-01 -3.52983698e-02
-3.41682762e-01 5.00807762e-01 -5.38766623e-01 9.59307626e-02
1.50537848e+00 -2.50370860e-01 3.55232507e-01 5.25575161e-01
8.76814246e-01 -5.77292025e-01 -1.84166276e+00 -8.66725683e-01
2.30961055e-01 -3.72685671e-01 -2.77292524e-02 -7.65226960e-01
-7.71526277e-01 8.82276714e-01 7.30857551e-01 -2.29924977e-01
1.20390773e+00 1.82627916e-01 7.74533868e-01 9.12352204e-01
4.93895143e-01 -1.25546491e+00 4.02479470e-01 7.14318156e-01
6.97153628e-01 -1.68904841e+00 -2.08180040e-01 -2.54575331e-02
-7.41862476e-01 9.93532479e-01 1.18028450e+00 -2.01009497e-01
6.33869231e-01 1.49600908e-01 2.06506446e-01 7.07495376e-04
-1.22852457e+00 -4.18402314e-01 7.28229508e-02 1.08250105e+00
9.86488163e-02 -2.49531046e-01 -3.86810936e-02 5.76356113e-01
2.81881064e-01 2.73049802e-01 2.98704952e-01 1.13205063e+00
-8.92105758e-01 -1.08598495e+00 -1.83247387e-01 4.11135644e-01
6.35317713e-02 -1.96324781e-01 -3.02117944e-01 6.89682245e-01
7.05952197e-02 7.94994414e-01 1.71303913e-01 -1.17446482e-01
3.95022541e-01 2.65618622e-01 2.20766634e-01 -1.27041650e+00
-3.49572659e-01 -8.82254988e-02 -2.09867284e-01 -5.57016730e-01
-4.51017290e-01 -5.41042089e-01 -1.21383822e+00 2.15843424e-01
-1.76009238e-01 1.60782561e-01 2.86221147e-01 1.02505028e+00
4.48651403e-01 3.25999558e-01 3.74690682e-01 -6.51585579e-01
-1.06950855e+00 -1.00928807e+00 -4.13511783e-01 4.00505006e-01
4.69900370e-01 -5.94963551e-01 -1.68456584e-01 3.15873384e-01] | [10.121963500976562, 2.4556355476379395] |
4bb2298b-165e-48b0-b3bb-3a9c8e702c24 | recent-advance-in-content-based-image | 1706.06064 | null | https://arxiv.org/abs/1706.06064v2 | https://arxiv.org/pdf/1706.06064v2.pdf | Recent Advance in Content-based Image Retrieval: A Literature Survey | The explosive increase and ubiquitous accessibility of visual data on the Web have led to the prosperity of research activity in image search or retrieval. With the ignorance of visual content as a ranking clue, methods with text search techniques for visual retrieval may suffer inconsistency between the text words and visual content. Content-based image retrieval (CBIR), which makes use of the representation of visual content to identify relevant images, has attracted sustained attention in recent two decades. Such a problem is challenging due to the intention gap and the semantic gap problems. Numerous techniques have been developed for content-based image retrieval in the last decade. The purpose of this paper is to categorize and evaluate those algorithms proposed during the period of 2003 to 2016. We conclude with several promising directions for future research. | ['Qi Tian', 'Houqiang Li', 'Wengang Zhou'] | 2017-06-19 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [ 1.77791193e-01 -4.93073374e-01 -4.53904539e-01 -9.32910815e-02
-7.33583689e-01 -6.02841496e-01 8.23351085e-01 5.85986078e-01
-5.00011921e-01 3.37633789e-01 3.53830487e-01 -2.57181495e-01
-3.07453334e-01 -6.26522899e-01 -9.52121839e-02 -3.96657377e-01
2.58875132e-01 -1.18794337e-01 5.04323781e-01 -2.77835459e-01
8.25645089e-01 5.63329518e-01 -2.03397894e+00 5.80498815e-01
6.29326224e-01 1.27391219e+00 5.26829600e-01 3.65452051e-01
-5.62712073e-01 7.13921905e-01 -5.68599880e-01 -2.54406631e-01
3.82464193e-02 -4.00979698e-01 -1.02313840e+00 2.59151936e-01
5.59591770e-01 -2.70410687e-01 -5.17639756e-01 1.25834012e+00
6.57653332e-01 -9.52390656e-02 6.22555375e-01 -1.31015670e+00
-1.20704436e+00 -1.69230431e-01 -7.50948787e-01 4.25514698e-01
4.76733148e-01 -3.35744351e-01 9.93587852e-01 -1.10143507e+00
8.25036347e-01 1.21405149e+00 -6.33754507e-02 3.06512415e-01
-4.64504302e-01 -2.44713351e-01 9.23788399e-02 8.98245037e-01
-1.77293169e+00 -1.97806302e-02 8.38441014e-01 -6.23181999e-01
6.12978756e-01 6.19845450e-01 9.42333221e-01 6.65219486e-01
1.03122965e-01 7.81138897e-01 1.08903849e+00 -8.84347081e-01
-1.35109499e-02 4.06226248e-01 1.27812596e-02 5.49777627e-01
3.35498273e-01 -2.02909142e-01 -5.28902769e-01 -8.19798335e-02
6.20324433e-01 3.67227465e-01 -1.65663838e-01 -3.07056397e-01
-6.62839055e-01 8.59670937e-01 6.91786408e-01 5.25032341e-01
-2.67239213e-01 -1.80936500e-01 6.10749900e-01 1.40877694e-01
5.78635573e-01 1.86076835e-01 3.03046405e-01 3.46214354e-01
-1.10483861e+00 3.39416116e-01 3.41427594e-01 8.32169890e-01
5.43312013e-01 -1.74417764e-01 -2.34283879e-01 1.08471584e+00
5.68639159e-01 7.59548724e-01 6.21707439e-01 -7.32940495e-01
1.65126652e-01 7.40034878e-01 -4.29363325e-02 -1.64086723e+00
1.66387826e-01 -9.35546309e-02 -7.12779164e-01 1.92828223e-01
1.41125947e-01 7.37118304e-01 -1.09012282e+00 8.89730811e-01
2.17829853e-01 -7.07770467e-01 -3.06797802e-01 1.17265224e+00
1.20252800e+00 7.56083965e-01 3.09681475e-01 -3.10778439e-01
1.63305283e+00 -8.22274864e-01 -8.73772323e-01 -1.32215053e-01
-5.71929943e-03 -1.19849288e+00 1.13329649e+00 1.78552538e-01
-1.11777878e+00 -3.79330724e-01 -9.78304505e-01 -2.88623840e-01
-8.66774201e-01 2.55837627e-02 2.03202009e-01 3.42785835e-01
-1.22086394e+00 -6.68924823e-02 -1.06505848e-01 -7.71864653e-01
3.36315423e-01 -1.95467740e-01 -2.44923979e-01 -2.77701914e-01
-1.01376212e+00 9.21792805e-01 4.25412267e-01 7.89511576e-02
-4.55812752e-01 -3.06627333e-01 -5.00071645e-01 -7.73436576e-02
4.32266533e-01 -4.26610172e-01 7.80804157e-01 -1.13934863e+00
-7.84039736e-01 1.37267423e+00 -2.06386998e-01 -1.17882036e-01
3.47494781e-01 -2.06914339e-02 -5.91525733e-01 6.94142640e-01
2.20420390e-01 5.32551885e-01 8.45298946e-01 -1.48899209e+00
-8.06733310e-01 -5.25758266e-01 -1.48529023e-01 3.72710168e-01
-6.08366609e-01 4.65295702e-01 -9.52406108e-01 -7.74047375e-01
1.60035521e-01 -6.80944264e-01 5.03543392e-02 4.24043119e-01
-1.35698155e-01 -4.36884582e-01 1.04480934e+00 -6.87194765e-01
1.51735127e+00 -2.00986171e+00 -2.72976458e-01 3.83467615e-01
2.85894454e-01 2.91949391e-01 -1.19353339e-01 9.07922268e-01
7.38135949e-02 5.17456770e-01 2.31136993e-01 3.32166582e-01
-4.66915131e-01 -6.88397908e-04 -4.83649403e-01 4.28907007e-01
-3.43990952e-01 9.70736980e-01 -7.63725281e-01 -1.19677508e+00
3.59231085e-01 4.28838193e-01 -3.73470113e-02 4.71094763e-03
7.57343322e-02 -1.53722554e-01 -6.65836692e-01 1.09146225e+00
6.21732712e-01 -7.09741890e-01 6.11609817e-02 -3.73236328e-01
-4.33614075e-01 -1.95007816e-01 -6.79018021e-01 1.32603574e+00
-1.06594048e-01 9.85680759e-01 -2.36479357e-01 -8.36417198e-01
6.02768123e-01 1.97825015e-01 6.71664536e-01 -1.48484993e+00
-4.28988449e-02 1.89407438e-01 -3.28403801e-01 -8.81349087e-01
7.13447690e-01 5.80973253e-02 1.42201737e-01 8.15090761e-02
-4.21812773e-01 -7.97986239e-02 3.39320362e-01 4.94298041e-01
4.84677881e-01 -1.88581318e-01 4.49465752e-01 -1.75488159e-01
5.78297734e-01 4.96280104e-01 -1.41571879e-01 6.12183690e-01
-2.13241786e-01 7.45305538e-01 -2.28304654e-01 -5.96957326e-01
-1.19475615e+00 -9.16024566e-01 -1.40831187e-01 8.16815853e-01
6.93567097e-01 -4.43925470e-01 -3.45892727e-01 -4.00697559e-01
-4.62574176e-02 8.35213214e-02 -5.78736544e-01 1.07498787e-01
-2.87055194e-01 -3.35922182e-01 6.60777912e-02 6.36278018e-02
6.44963026e-01 -1.29505956e+00 -6.39033020e-01 -1.83956102e-01
-4.18662459e-01 -8.89108539e-01 -3.59044701e-01 -5.30014396e-01
-7.00658679e-01 -1.12540817e+00 -1.29169762e+00 -9.69737053e-01
8.59564960e-01 1.02558899e+00 1.07043576e+00 5.45558929e-01
-7.92883575e-01 7.76230991e-01 -7.90513694e-01 -4.04598564e-01
-9.02881473e-02 -1.47189915e-01 -4.30415332e-01 -1.55234709e-01
4.12577063e-01 1.08622491e-01 -1.16643548e+00 2.27950901e-01
-1.36833954e+00 -4.13605431e-03 4.64831263e-01 5.58794677e-01
6.88336968e-01 4.03701589e-02 3.01286250e-01 -4.35239881e-01
7.52833426e-01 -5.91767430e-01 -6.02896333e-01 7.53094792e-01
-1.07588840e+00 -1.61597863e-01 -7.89384842e-02 -1.82400495e-01
-8.87838721e-01 -4.13739026e-01 2.52039135e-01 -2.82572657e-01
1.25168562e-01 8.69498134e-01 2.26297706e-01 -1.58581942e-01
6.07224882e-01 3.64860117e-01 1.41142353e-01 -5.01485467e-01
2.28440762e-01 9.81811166e-01 2.00909778e-01 -2.42582187e-02
5.40278554e-01 6.14487767e-01 -1.85154621e-02 -1.19717300e+00
-7.55061209e-01 -1.05719316e+00 -2.64998265e-02 -9.18467224e-01
8.70068908e-01 -8.81210625e-01 -5.75428486e-01 2.18688041e-01
-1.15568471e+00 4.47349459e-01 4.03721109e-02 1.99425146e-01
1.73480287e-02 9.37262356e-01 -3.00092280e-01 -8.49583030e-01
-5.76962054e-01 -9.60397720e-01 9.25387561e-01 2.91648716e-01
-7.53199086e-02 -7.99313903e-01 -9.05594900e-02 5.94315410e-01
5.04315376e-01 -1.59157827e-01 8.82092297e-01 -2.03977183e-01
-5.63315213e-01 -3.90599668e-01 -8.26416373e-01 1.12129375e-01
1.64118573e-01 5.83462939e-02 -7.18458354e-01 -2.16823190e-01
-1.77074790e-01 -1.66830570e-01 7.92897344e-01 4.22525495e-01
1.35349023e+00 -5.02364099e-01 -4.28918689e-01 -6.21151403e-02
2.02547264e+00 5.22287667e-01 9.55217957e-01 8.28478217e-01
5.02515376e-01 8.81254375e-01 7.77608752e-01 4.10735548e-01
2.16118842e-01 7.52962887e-01 5.84768474e-01 -4.24688980e-02
-3.68119448e-01 -3.12663078e-01 -3.45314085e-01 8.64501595e-01
-1.57059118e-01 -4.93619591e-01 -1.01568818e+00 8.62861991e-01
-1.79998207e+00 -1.20736921e+00 1.26588289e-02 2.17012525e+00
6.49873555e-01 -5.13838470e-01 1.08250931e-01 1.20956227e-01
8.34014118e-01 1.47451147e-01 -2.42862627e-01 -1.35783240e-01
-1.94037944e-01 -2.37474605e-01 5.34770310e-01 1.81793451e-01
-8.77651393e-01 7.13747978e-01 6.95513439e+00 1.33970189e+00
-1.31617904e+00 -6.19713552e-02 7.15323329e-01 2.39310071e-01
-3.52024138e-01 -1.09248841e-02 -4.85072374e-01 5.52648544e-01
1.45163089e-01 -3.94806325e-01 1.84030578e-01 8.60210598e-01
2.59435803e-01 -4.61693615e-01 -3.32794011e-01 1.40661168e+00
5.31636477e-01 -1.36238003e+00 7.61885822e-01 6.63123280e-02
6.65082932e-01 -2.45015264e-01 4.33014095e-01 -3.44514757e-01
-3.35701168e-01 -9.10089314e-01 8.88471007e-01 4.05629009e-01
8.49985123e-01 -5.69683433e-01 5.48936963e-01 7.79341832e-02
-1.23133230e+00 2.36904677e-02 -4.30168152e-01 1.69698089e-01
1.28081394e-02 3.26928079e-01 -4.30992216e-01 5.38277149e-01
1.00277102e+00 7.24087477e-01 -8.75113010e-01 1.61944127e+00
1.46602079e-01 2.19403595e-01 1.54829532e-01 -2.74632454e-01
1.42587349e-01 -1.92461252e-01 4.56554234e-01 9.53480244e-01
4.43174750e-01 -1.00866124e-01 5.85897341e-02 4.16445345e-01
1.12765804e-01 6.11858606e-01 -8.18542182e-01 -2.91727751e-01
4.11929131e-01 1.18586349e+00 -1.04173899e+00 -3.75163138e-01
-6.28935933e-01 1.03289914e+00 1.03390182e-03 4.59942520e-01
-3.32109869e-01 -3.16913128e-01 3.87089066e-02 5.22079349e-01
7.20188022e-02 -7.33720064e-02 -1.56565949e-01 -8.39919448e-01
1.89678505e-01 -8.71081173e-01 6.00967050e-01 -1.16538274e+00
-1.36609530e+00 6.25399947e-01 2.94102490e-01 -1.61416113e+00
4.22280421e-03 -4.16164726e-01 4.33976054e-02 7.18003094e-01
-1.64387250e+00 -1.06879485e+00 -3.17295372e-01 5.33827901e-01
7.88877964e-01 -1.11227617e-01 5.38271189e-01 3.35681051e-01
1.09012499e-01 1.16653241e-01 1.80836052e-01 5.88909984e-02
7.33016610e-01 -6.45228267e-01 -3.87487262e-01 5.83129287e-01
1.60165310e-01 5.05728900e-01 7.16452241e-01 -7.37015367e-01
-1.36261559e+00 -6.13526464e-01 1.13659608e+00 -1.64110258e-01
5.63561797e-01 1.97011471e-01 -6.29533887e-01 -4.18911502e-02
5.68455994e-01 -1.13700315e-01 5.21338105e-01 -4.48696792e-01
-4.46298450e-01 -1.50438696e-01 -9.09193993e-01 8.34530830e-01
6.19406044e-01 -6.93673968e-01 -3.84639800e-01 1.45235822e-01
3.20838869e-01 2.39992328e-02 -4.40767467e-01 1.70708582e-01
7.96811938e-01 -6.26907527e-01 1.37682438e+00 -1.09369755e-01
4.75377858e-01 -1.79957107e-01 -2.02466697e-01 -7.05408275e-01
-1.48348615e-01 6.49412200e-02 5.31704545e-01 9.36393738e-01
1.17644414e-01 -1.47862345e-01 6.09234929e-01 5.67120433e-01
3.38488191e-01 -3.92756552e-01 -7.14912653e-01 -6.15601897e-01
-2.08742648e-01 -7.57518644e-03 4.50424738e-02 6.52327657e-01
1.77769661e-01 2.75762469e-01 -4.34486479e-01 -4.31063265e-01
5.93808949e-01 2.90357649e-01 2.49094531e-01 -1.35805261e+00
3.45234036e-01 -7.19093978e-01 -5.87146521e-01 -8.61859679e-01
-3.16340208e-01 -8.44863236e-01 -2.44467646e-01 -1.94721389e+00
8.67955565e-01 -6.21590242e-02 -2.53057897e-01 2.86566079e-01
-1.38527557e-01 7.57411599e-01 4.93952751e-01 7.34286666e-01
-9.46311593e-01 1.78867966e-01 1.43091261e+00 -5.06286323e-01
1.82346404e-01 -3.79432440e-01 -6.51752114e-01 5.37219882e-01
6.12015069e-01 -3.53461504e-01 -5.35915077e-01 -2.66061306e-01
7.31772304e-01 -1.48369759e-01 5.35519719e-01 -5.33952355e-01
3.29943657e-01 -2.99968809e-01 3.62203658e-01 -8.49636376e-01
2.68240511e-01 -1.07023311e+00 1.06623963e-01 2.61942327e-01
-4.38008368e-01 4.88358796e-01 3.02316491e-02 6.54495895e-01
-8.23390186e-01 -3.08619231e-01 4.88847166e-01 -3.47529560e-01
-1.18989146e+00 2.49659777e-01 -7.56025434e-01 -1.16856163e-02
9.11109328e-01 -5.79040349e-01 -2.70143688e-01 -5.32216370e-01
-4.82000738e-01 9.44732726e-02 4.20944840e-01 6.87316716e-01
1.17763638e+00 -1.42154849e+00 -5.20773590e-01 -2.77467608e-01
5.47967911e-01 -6.24186814e-01 5.74367583e-01 3.67960602e-01
-8.22724700e-01 6.70848072e-01 -3.07786554e-01 -3.84613901e-01
-1.88823390e+00 9.51684594e-01 -1.62521154e-01 -1.67356163e-01
-5.13360322e-01 1.12451456e-01 2.13293478e-01 5.92086613e-01
3.13219011e-01 3.07525218e-01 -6.96413159e-01 3.24913740e-01
8.50710869e-01 5.30387580e-01 -6.53909147e-02 -8.50433350e-01
-4.37005460e-01 7.98117638e-01 -7.57247880e-02 -2.03399226e-01
9.15874302e-01 -7.01403618e-01 -3.45432460e-01 3.44418108e-01
1.43534815e+00 -1.74759090e-01 -2.41100311e-01 -1.35972098e-01
1.91520646e-01 -8.69761646e-01 2.90816605e-01 -8.72414231e-01
-1.07457232e+00 7.85797715e-01 8.79016221e-01 6.40609145e-01
1.11100888e+00 1.78324193e-01 3.66576672e-01 9.83335152e-02
3.06784302e-01 -1.35993278e+00 3.49862009e-01 9.57432538e-02
1.05599904e+00 -1.59634042e+00 3.84655654e-01 -4.49260145e-01
-5.57726860e-01 1.04979253e+00 2.27915764e-01 1.94161147e-01
7.94895351e-01 -2.93265522e-01 2.92175770e-01 -5.42607427e-01
-2.94935882e-01 -5.14674664e-01 7.96434224e-01 5.02122164e-01
5.14635324e-01 -3.70661229e-01 -9.63713527e-01 -1.08607374e-01
4.22754824e-01 -4.59998101e-02 -7.17404708e-02 1.07504964e+00
-6.93865836e-01 -1.19864988e+00 -5.43707550e-01 2.85703838e-01
-8.70716870e-01 -2.72116512e-01 -5.69993317e-01 8.21437836e-01
-4.57602262e-01 1.23341846e+00 6.30198643e-02 7.33739287e-02
7.56873712e-02 -3.34209085e-01 3.58743817e-01 -1.77015036e-01
-2.96923816e-01 2.43832305e-01 -3.24657559e-01 -4.29115832e-01
-7.07069218e-01 -3.27550828e-01 -8.45694840e-01 -1.10088035e-01
-2.57292569e-01 2.53059089e-01 9.54434097e-01 6.58257127e-01
3.39591563e-01 5.15941530e-04 3.96275997e-01 -4.95718628e-01
-1.93788081e-01 -6.34783506e-01 -6.71823800e-01 5.36955953e-01
1.60585552e-01 -4.52184498e-01 -2.34994262e-01 1.41241804e-01] | [10.868904113769531, 0.31133797764778137] |
ae35f0be-5b85-4233-94e0-8ec068ccdbc8 | evaluating-dense-passage-retrieval-using | 2208.06959 | null | https://arxiv.org/abs/2208.06959v1 | https://arxiv.org/pdf/2208.06959v1.pdf | Evaluating Dense Passage Retrieval using Transformers | Although representational retrieval models based on Transformers have been able to make major advances in the past few years, and despite the widely accepted conventions and best-practices for testing such models, a $\textit{standardized}$ evaluation framework for testing them has not been developed. In this work, we formalize the best practices and conventions followed by researchers in the literature, paving the path for more standardized evaluations - and therefore more fair comparisons between the models. Our framework (1) embeds the documents and queries; (2) for each query-document pair, computes the relevance score based on the dot product of the document and query embedding; (3) uses the $\texttt{dev}$ set of the MSMARCO dataset to evaluate the models; (4) uses the $\texttt{trec_eval}$ script to calculate MRR@100, which is the primary metric used to evaluate the models. Most importantly, we showcase the use of this framework by experimenting on some of the most well-known dense retrieval models. | ['Nima Sadri'] | 2022-08-15 | null | null | null | null | ['passage-retrieval'] | ['natural-language-processing'] | [ 1.00153960e-01 -4.25392181e-01 -8.80024489e-03 -3.89853686e-01
-1.06539309e+00 -6.67033851e-01 1.05633318e+00 2.75082916e-01
-7.46795416e-01 3.27444404e-01 8.05988163e-02 -4.08069015e-01
-5.59382498e-01 -7.26693392e-01 -4.21924829e-01 -4.27376539e-01
-2.40367770e-01 5.36392093e-01 4.06518310e-01 -3.73916805e-01
5.03860176e-01 3.02587599e-01 -1.84314406e+00 2.42157191e-01
5.34543872e-01 1.24540007e+00 -4.21824232e-02 4.28308219e-01
-7.12057203e-02 4.72467333e-01 -4.92608964e-01 -7.95042574e-01
2.23011032e-01 -3.47879797e-01 -8.94374192e-01 -5.41067898e-01
3.77180964e-01 -1.25226215e-01 -3.23492229e-01 8.94512236e-01
4.46746618e-01 2.53593892e-01 9.81337309e-01 -8.12625289e-01
-8.47371697e-01 4.51106936e-01 -2.16531992e-01 4.20733869e-01
4.12207544e-01 -2.83090081e-02 1.45963681e+00 -1.15568542e+00
7.73205638e-01 1.12337506e+00 3.51230502e-01 1.61314666e-01
-1.04820883e+00 -5.19759893e-01 -4.08964604e-02 2.72900611e-01
-1.58363271e+00 -2.27380261e-01 4.05823380e-01 -1.91462010e-01
1.10869801e+00 5.45492709e-01 5.03000140e-01 9.90421772e-01
1.27766794e-02 6.73556268e-01 1.13759017e+00 -7.45537341e-01
1.05571173e-01 4.43798184e-01 5.81761181e-01 4.61852759e-01
1.81963220e-01 2.50768334e-01 -3.37286919e-01 -2.38315001e-01
4.02582496e-01 -1.76576018e-01 -1.71461090e-01 -2.70408750e-01
-9.33450103e-01 9.25217748e-01 3.25553864e-01 8.00976574e-01
-2.53183693e-01 1.32153600e-01 4.31914866e-01 5.69055676e-01
3.00223351e-01 5.71779728e-01 -4.28025499e-02 -2.38062397e-01
-1.15902877e+00 4.38128620e-01 7.31915772e-01 6.53639495e-01
6.84144855e-01 -3.11729193e-01 -3.25268656e-01 9.76172745e-01
4.23067331e-01 2.97035456e-01 5.95631421e-01 -6.91505194e-01
3.83608311e-01 6.00586474e-01 1.58958122e-01 -1.15334535e+00
5.47441207e-02 -4.48485345e-01 -3.83362025e-01 -1.12994224e-01
1.31137967e-01 2.36124173e-01 -7.21384704e-01 1.47526836e+00
-1.48260847e-01 -3.16956252e-01 4.70613455e-03 8.16191137e-01
6.58132434e-01 5.64728856e-01 1.40940502e-01 5.42003587e-02
1.28559768e+00 -7.95218945e-01 -4.86095279e-01 -1.12257069e-02
8.44146848e-01 -9.65102255e-01 1.35244036e+00 2.39290252e-01
-1.14936590e+00 -4.42832679e-01 -1.12826157e+00 -1.26872420e-01
-8.07877898e-01 -9.99524519e-02 5.55862844e-01 8.12658787e-01
-1.29886162e+00 4.19912457e-01 -6.34428620e-01 -5.26322961e-01
-1.50988951e-01 2.99491554e-01 -3.52795035e-01 -3.81573230e-01
-1.45821655e+00 1.13836622e+00 1.86138660e-01 1.75713286e-01
-8.58108997e-01 -3.34673941e-01 -6.85356498e-01 1.54520541e-01
1.88681096e-01 -5.37244976e-01 9.73658442e-01 -5.43963432e-01
-1.18769109e+00 8.80778968e-01 7.77118281e-02 -2.20495179e-01
2.93175936e-01 -2.11602330e-01 -5.64659119e-01 5.32030426e-02
-7.24969581e-02 7.12699175e-01 4.83227879e-01 -9.88323212e-01
-4.09810424e-01 -4.00063992e-01 2.29100928e-01 1.66982874e-01
-5.51049471e-01 4.13115978e-01 -9.11753356e-01 -6.70434356e-01
6.84875846e-02 -9.55139279e-01 3.42753045e-02 -3.43385249e-01
-3.15103292e-01 -3.36827308e-01 5.03769696e-01 -4.24651951e-01
1.65435517e+00 -2.27347445e+00 1.14637293e-01 5.46055257e-01
-1.76541973e-02 2.82967210e-01 -3.69396329e-01 1.01175213e+00
-4.03009132e-02 2.76060730e-01 -4.36369814e-02 -5.08158803e-01
4.22685593e-01 -6.30546063e-02 -5.43132484e-01 2.64557123e-01
-8.02295133e-02 8.16043615e-01 -5.76333761e-01 -5.17501116e-01
1.02004357e-01 7.23976016e-01 -4.68891263e-01 -2.45616939e-02
-5.75511716e-02 -3.74725908e-01 -5.41435182e-01 3.96435738e-01
2.83744991e-01 -2.65314490e-01 2.07565799e-01 -9.13610309e-02
-1.36330351e-01 5.55407226e-01 -1.17418528e+00 1.47264063e+00
-1.93538904e-01 6.04662895e-01 -3.36741120e-01 -8.34892154e-01
7.90260077e-01 2.37272948e-01 3.22617352e-01 -9.76238489e-01
1.08645923e-01 3.12415898e-01 -1.93159446e-01 -2.08311170e-01
8.30440283e-01 2.81917062e-02 -1.88405946e-01 6.90640271e-01
5.33589236e-02 -2.56987721e-01 5.10608196e-01 3.37645113e-01
1.27624202e+00 -5.51011786e-02 -1.34690672e-01 -1.50668204e-01
5.45305550e-01 -1.61329880e-01 -1.99871987e-01 8.86969388e-01
3.01483590e-02 6.28914356e-01 4.61013436e-01 -1.69871151e-01
-8.48717391e-01 -1.06269038e+00 -2.90815830e-01 1.22353303e+00
-8.42003524e-02 -7.25741029e-01 -5.94170153e-01 -5.36973596e-01
-3.54630314e-02 7.87088752e-01 -8.20424139e-01 -1.29088759e-01
-2.05944747e-01 -7.06510305e-01 7.79408574e-01 2.67516345e-01
2.99898833e-01 -1.12788093e+00 -5.45932472e-01 -4.71498296e-02
-6.03651889e-02 -6.76483691e-01 -2.65805274e-01 1.84913408e-02
-7.03334033e-01 -8.95283759e-01 -6.83886826e-01 -5.92107296e-01
3.10998738e-01 2.10324794e-01 1.17036700e+00 3.31317842e-01
-1.52241960e-01 6.89701259e-01 -5.65210998e-01 -2.24269241e-01
-1.28649637e-01 1.69633940e-01 -2.12158293e-01 -2.75564194e-01
5.70687473e-01 -2.46241361e-01 -7.03823447e-01 3.68040383e-01
-1.49925911e+00 -3.78589451e-01 4.94379818e-01 5.87404251e-01
7.30786979e-01 -2.03620747e-01 4.46673244e-01 -7.58327246e-01
1.14156640e+00 -3.50635827e-01 -6.95155203e-01 5.19879639e-01
-1.20907581e+00 1.40593231e-01 1.41613483e-01 -1.71046574e-02
-4.45598662e-01 -6.28015995e-01 -2.52985030e-01 -1.41001984e-01
1.94859341e-01 1.08193624e+00 2.46932179e-01 4.64050025e-02
6.73745096e-01 3.23266178e-01 -1.61713764e-01 -5.90668023e-01
5.01875043e-01 7.29680598e-01 2.46530697e-01 -6.22427344e-01
7.37159848e-01 2.23019123e-01 -3.32255304e-01 -6.53626561e-01
-4.66456085e-01 -5.01239300e-01 -2.86669970e-01 -8.77378136e-02
5.61931968e-01 -6.04316413e-01 -3.48797828e-01 8.07401836e-02
-8.05304885e-01 -4.38129082e-02 -2.66126722e-01 6.15731061e-01
-3.33667099e-01 5.02596378e-01 -5.62696636e-01 -7.05752730e-01
-4.26981121e-01 -1.31896031e+00 9.16252911e-01 -6.04648180e-02
-3.78751725e-01 -1.00850177e+00 3.98491144e-01 2.31860191e-01
8.07891488e-01 -1.34532735e-01 1.36574066e+00 -7.90651798e-01
-4.14158225e-01 -7.72449076e-01 -3.71418059e-01 5.09134173e-01
-2.10512295e-01 1.86316893e-01 -1.01545286e+00 -4.37460691e-01
-1.45089030e-01 -4.75460857e-01 8.79503012e-01 5.39576896e-02
1.05944777e+00 1.01806551e-01 -2.98216909e-01 1.12184741e-01
1.41323805e+00 1.02783749e-02 1.02384675e+00 4.70634878e-01
-1.75519194e-02 6.78195357e-01 4.71973568e-01 5.70751764e-02
4.03607130e-01 9.85951483e-01 2.21647948e-01 1.95657998e-01
-1.74666066e-02 -1.40579864e-01 4.43689853e-01 9.55559015e-01
-7.81074092e-02 -2.75149941e-01 -8.59728515e-01 5.09254575e-01
-1.54755306e+00 -8.80256295e-01 2.55386144e-01 2.38442230e+00
7.92524993e-01 1.61972255e-01 -1.07432812e-01 1.13238662e-01
7.90814683e-02 3.43302876e-01 2.30528824e-02 -3.34179401e-01
-6.45749271e-02 7.01974392e-01 -3.97088118e-02 4.39286232e-01
-7.85849094e-01 8.69815648e-01 6.74984741e+00 7.63084590e-01
-1.28815019e+00 -8.54175761e-02 3.50948066e-01 -2.24859014e-01
-5.32641768e-01 1.34327903e-01 -6.53977394e-01 3.51260632e-01
1.18239665e+00 -2.53506482e-01 4.03919458e-01 6.34178460e-01
-2.40829989e-01 -1.59023061e-01 -1.39183390e+00 7.36777306e-01
1.61171183e-01 -1.16158509e+00 1.82375595e-01 7.37813860e-02
3.23793173e-01 1.48329884e-01 4.61333156e-01 8.42891455e-01
2.72422403e-01 -1.13525581e+00 6.71711147e-01 5.56392610e-01
8.27105224e-01 -3.69142652e-01 8.15673530e-01 6.60454035e-02
-8.98312807e-01 1.21138498e-01 -3.53950053e-01 2.23147973e-01
-2.27545261e-01 3.67605746e-01 -5.80039799e-01 7.62780666e-01
8.64409626e-01 3.39171261e-01 -9.17319655e-01 1.00926232e+00
1.15046902e-02 6.49895251e-01 -2.85291135e-01 -2.02182427e-01
3.89709473e-01 -2.37392694e-01 3.09356838e-01 1.38889050e+00
4.08189476e-01 -3.57424468e-01 -2.03021467e-01 7.07903862e-01
-1.06748700e-01 5.15715182e-01 -5.90954840e-01 -2.63076514e-01
5.31282485e-01 1.04854715e+00 -5.47471404e-01 -1.35230288e-01
-3.57336253e-01 7.24367738e-01 4.22094762e-01 5.14174759e-01
-6.90135300e-01 -6.53168499e-01 3.98236334e-01 1.81887105e-01
3.35955679e-01 -1.51787609e-01 1.13350615e-01 -9.88395929e-01
2.35511914e-01 -1.02657640e+00 5.93507230e-01 -8.54375541e-01
-1.35602808e+00 7.83734024e-01 5.23664594e-01 -1.01471543e+00
-4.18000668e-01 -5.46119809e-01 -2.04499394e-01 1.16093397e+00
-1.45603240e+00 -7.37968504e-01 -3.86800915e-02 5.06400168e-01
3.51919010e-02 -1.56345367e-01 1.26601934e+00 5.50311387e-01
-4.96939391e-01 7.51517117e-01 2.28545412e-01 1.94865257e-01
7.29949653e-01 -9.61371779e-01 2.30918229e-01 4.19827521e-01
5.01729608e-01 1.31111050e+00 6.14522696e-01 -2.51032978e-01
-1.43599880e+00 -6.67110622e-01 1.12614155e+00 -5.82745552e-01
7.60410130e-01 -2.81949788e-01 -8.38329136e-01 6.08034194e-01
2.50257879e-01 -2.56176829e-01 8.82142484e-01 2.60181129e-01
-6.77516878e-01 -2.93280572e-01 -9.06232238e-01 5.92086554e-01
6.04079962e-01 -7.76583850e-01 -6.96453691e-01 1.43007830e-01
4.36789423e-01 -7.79862925e-02 -9.79328156e-01 4.58238840e-01
8.40352714e-01 -1.15800548e+00 9.58203733e-01 -5.00952661e-01
3.18976820e-01 -1.31753772e-01 -5.68262339e-01 -9.81684983e-01
-2.00738057e-01 -9.49747786e-02 7.26981461e-02 1.06563628e+00
7.60726094e-01 -6.50655389e-01 3.88254791e-01 7.03434646e-01
6.11285158e-02 -9.79284406e-01 -9.12225723e-01 -6.12075150e-01
2.64551133e-01 -7.14967012e-01 5.27492881e-01 7.29896486e-01
-6.42198101e-02 1.54006824e-01 4.94988114e-02 -1.90358326e-01
1.27602354e-01 -4.07560170e-02 5.88002086e-01 -1.17509258e+00
-3.41159433e-01 -6.73386455e-01 -3.29786748e-01 -8.57466400e-01
-1.36955395e-01 -9.19229090e-01 -2.45461211e-01 -1.55833316e+00
3.32910061e-01 -5.76741040e-01 -7.54986644e-01 3.96560788e-01
9.15670022e-03 4.10762042e-01 2.05474675e-01 3.50631684e-01
-7.31577575e-01 4.45572436e-01 7.75218666e-01 -1.00275107e-01
1.61047325e-01 -3.39415669e-01 -8.43804479e-01 1.63278222e-01
3.68907452e-01 -4.65423852e-01 -5.99015832e-01 -5.07737219e-01
5.99854410e-01 -1.22370034e-01 3.99408340e-01 -9.79130387e-01
1.16840050e-01 2.14866757e-01 1.33405387e-01 -4.81990546e-01
5.12926161e-01 -7.61126518e-01 1.21669389e-01 2.04382598e-01
-6.10596418e-01 5.11832952e-01 2.47093275e-01 4.30446386e-01
-5.64068377e-01 -5.09085357e-01 4.35202301e-01 6.40483499e-02
-3.71180147e-01 7.28012174e-02 -1.89390481e-01 -2.71497183e-02
7.66585827e-01 -1.41493276e-01 -3.73257756e-01 -4.76375490e-01
-3.77475888e-01 6.27396032e-02 5.34998715e-01 5.61150372e-01
5.37877321e-01 -1.13411868e+00 -5.81261158e-01 1.48366705e-01
4.39628363e-01 -5.78101933e-01 9.74777639e-02 7.00245619e-01
-4.40148503e-01 9.23279941e-01 1.62929237e-01 -3.31795752e-01
-1.15219343e+00 3.76133412e-01 3.58475089e-01 -7.07545161e-01
-2.52829701e-01 6.48579061e-01 -2.00600520e-01 -3.29144299e-01
3.28491747e-01 -2.23171502e-01 -3.03456426e-01 6.26231655e-02
5.83436847e-01 1.48163065e-01 4.84289020e-01 -5.14244914e-01
-3.41615289e-01 3.76923591e-01 -4.36095357e-01 -6.56843722e-01
1.38358474e+00 1.86767861e-01 -2.75596023e-01 3.66676360e-01
1.37563825e+00 -1.00952245e-01 -4.01820153e-01 -2.33578265e-01
2.63625294e-01 -3.96617591e-01 5.10790646e-02 -9.43397284e-01
-7.69331574e-01 8.85664284e-01 8.09894383e-01 5.37863731e-01
8.28711271e-01 -1.93320364e-01 3.01251441e-01 4.98826236e-01
5.38255453e-01 -9.49244678e-01 7.47596612e-03 6.42183959e-01
8.40827763e-01 -7.92528987e-01 -8.72463640e-03 -4.40191031e-02
-3.88751388e-01 7.79960036e-01 1.00003421e-01 -2.09105108e-02
7.12823033e-01 -1.12692118e-01 1.09545127e-01 -5.91306925e-01
-6.75057471e-01 4.53568101e-02 6.15034521e-01 2.74529278e-01
8.81983995e-01 -1.85086086e-01 -6.05563641e-01 4.30600882e-01
-2.41306290e-01 1.21947303e-01 4.19088081e-02 9.49403107e-01
-2.11934358e-01 -1.48877704e+00 -1.37942046e-01 4.95593369e-01
-5.15892982e-01 -3.42431813e-01 -5.20645022e-01 1.02385616e+00
-3.59513581e-01 9.72225785e-01 -2.10098386e-01 -4.93765950e-01
5.60008764e-01 3.78689170e-01 4.66182113e-01 -6.35013402e-01
-8.55349541e-01 -2.72606581e-01 3.75648253e-02 -3.83428693e-01
-1.71732277e-01 -6.27490759e-01 -8.57372642e-01 -2.53104001e-01
-2.11835533e-01 5.31901598e-01 6.85828209e-01 8.42398226e-01
4.11145270e-01 2.13914126e-01 3.77431184e-01 -5.38727164e-01
-7.56596804e-01 -1.23013711e+00 -5.59363842e-01 6.46437407e-01
-2.41564438e-01 -7.12373078e-01 -4.37569857e-01 -3.32138598e-01] | [11.414566993713379, 7.609040260314941] |
4b742f1f-6dfd-4221-8155-46aa093bf267 | pointclip-point-cloud-understanding-by-clip | 2112.02413 | null | https://arxiv.org/abs/2112.02413v1 | https://arxiv.org/pdf/2112.02413v1.pdf | PointCLIP: Point Cloud Understanding by CLIP | Recently, zero-shot and few-shot learning via Contrastive Vision-Language Pre-training (CLIP) have shown inspirational performance on 2D visual recognition, which learns to match images with their corresponding texts in open-vocabulary settings. However, it remains under explored that whether CLIP, pre-trained by large-scale image-text pairs in 2D, can be generalized to 3D recognition. In this paper, we identify such a setting is feasible by proposing PointCLIP, which conducts alignment between CLIP-encoded point cloud and 3D category texts. Specifically, we encode a point cloud by projecting it into multi-view depth maps without rendering, and aggregate the view-wise zero-shot prediction to achieve knowledge transfer from 2D to 3D. On top of that, we design an inter-view adapter to better extract the global feature and adaptively fuse the few-shot knowledge learned from 3D into CLIP pre-trained in 2D. By just fine-tuning the lightweight adapter in the few-shot settings, the performance of PointCLIP could be largely improved. In addition, we observe the complementary property between PointCLIP and classical 3D-supervised networks. By simple ensembling, PointCLIP boosts baseline's performance and even surpasses state-of-the-art models. Therefore, PointCLIP is a promising alternative for effective 3D point cloud understanding via CLIP under low resource cost and data regime. We conduct thorough experiments on widely-adopted ModelNet10, ModelNet40 and the challenging ScanObjectNN to demonstrate the effectiveness of PointCLIP. The code is released at https://github.com/ZrrSkywalker/PointCLIP. | ['Hongsheng Li', 'Peng Gao', 'Yu Qiao', 'Bin Cui', 'Xupeng Miao', 'Kunchang Li', 'Wei zhang', 'Ziyu Guo', 'Renrui Zhang'] | 2021-12-04 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Zhang_PointCLIP_Point_Cloud_Understanding_by_CLIP_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Zhang_PointCLIP_Point_Cloud_Understanding_by_CLIP_CVPR_2022_paper.pdf | cvpr-2022-1 | ['training-free-3d-point-cloud-classification', 'zero-shot-transfer-3d-point-cloud', 'training-free-3d-part-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-5.06288484e-02 -1.72371134e-01 -3.14037919e-01 -4.32237536e-01
-9.18990135e-01 -5.42413950e-01 8.07158053e-01 -2.57360131e-01
-3.21240164e-02 -1.47807285e-01 1.95040718e-01 -1.41704828e-01
1.27760038e-01 -7.08550215e-01 -1.13304424e+00 -4.74814445e-01
4.29338366e-01 5.61602175e-01 1.89120203e-01 -5.69458976e-02
1.30522892e-01 4.69789684e-01 -1.77045465e+00 1.99244976e-01
5.27404010e-01 1.11512125e+00 5.22237480e-01 4.46610332e-01
-6.04856789e-01 5.12200415e-01 1.12655170e-01 -3.51534426e-01
5.65206707e-01 1.04489908e-01 -5.62326074e-01 3.57151806e-01
1.01148295e+00 -7.50280082e-01 -4.94328320e-01 9.42875564e-01
5.17354369e-01 4.73324209e-02 6.71102703e-01 -1.27801478e+00
-9.24585462e-01 1.14737570e-01 -6.91548824e-01 -1.03694700e-01
3.48209292e-01 4.82788771e-01 1.02692771e+00 -1.54855680e+00
6.38596952e-01 1.14159977e+00 6.93575621e-01 5.29903591e-01
-9.68327820e-01 -8.09143841e-01 1.78982347e-01 2.95883566e-01
-1.43685365e+00 -4.65471983e-01 8.84228408e-01 -6.02892995e-01
1.30578470e+00 -2.70447843e-02 8.03694725e-01 1.38874006e+00
-1.07971609e-01 8.67378712e-01 6.92261875e-01 -3.88079494e-01
8.85757655e-02 1.77285057e-02 3.41081992e-02 6.78524315e-01
-1.35316357e-01 2.29454786e-01 -8.62593234e-01 4.41919453e-02
9.61481690e-01 4.29721147e-01 -3.22625548e-01 -9.82653975e-01
-1.37329054e+00 6.05479479e-01 5.63440382e-01 2.31907442e-02
-2.05035791e-01 2.14566559e-01 2.85073668e-01 2.03404054e-01
6.58750117e-01 1.02213070e-01 -3.60477388e-01 -1.03092872e-01
-7.87122607e-01 -8.41045752e-02 5.71945012e-01 1.52892125e+00
1.02058983e+00 -1.49648175e-01 -8.07441212e-03 8.29893470e-01
3.40575576e-01 9.51501131e-01 9.32308957e-02 -9.57406700e-01
6.61761940e-01 5.22930145e-01 -8.33812281e-02 -9.08503413e-01
-9.13171247e-02 -1.77374050e-01 -9.24702048e-01 1.58256426e-01
3.33177857e-02 4.06985253e-01 -1.07889485e+00 1.34237492e+00
4.07001913e-01 6.11141264e-01 1.49864759e-02 1.21734130e+00
1.03278744e+00 7.38077402e-01 -2.21408516e-01 1.85050130e-01
1.23437381e+00 -1.24928343e+00 -2.03704655e-01 -3.62746030e-01
4.49441016e-01 -6.79027200e-01 1.42476988e+00 2.87587568e-02
-9.35353279e-01 -7.54939079e-01 -1.01043177e+00 -4.59980637e-01
-2.40095869e-01 -9.80448052e-02 5.00602126e-01 1.61700591e-01
-8.47848237e-01 3.45948040e-01 -9.43692803e-01 -6.04476750e-01
5.78520536e-01 2.84921774e-03 -6.27061248e-01 -5.33291638e-01
-9.10845399e-01 7.83345759e-01 1.32763296e-01 -1.89663038e-01
-1.12156582e+00 -1.19550729e+00 -1.10022700e+00 -6.79547638e-02
4.96191829e-01 -1.08634806e+00 1.21568441e+00 -4.98021632e-01
-1.34135580e+00 1.31812763e+00 -1.67878851e-01 -1.99293151e-01
4.93073821e-01 -2.90268809e-01 -3.72993574e-02 3.49355698e-01
2.35002175e-01 8.56458724e-01 1.07774663e+00 -1.33293128e+00
-5.02452135e-01 -5.61077893e-01 2.21095994e-01 4.21939790e-01
-2.07401127e-01 -2.65294850e-01 -9.97607946e-01 -2.93660253e-01
2.73868531e-01 -7.05242395e-01 8.62722322e-02 7.01632142e-01
-2.48177573e-01 -2.08980933e-01 6.10178947e-01 -2.43126184e-01
4.18189883e-01 -2.28958821e+00 1.48104221e-01 -1.35342062e-01
3.04694384e-01 1.29231408e-01 -3.43331337e-01 5.31686902e-01
1.41736483e-02 -1.64320290e-01 -9.87772718e-02 -6.28476501e-01
1.83165267e-01 2.84062237e-01 -5.55944026e-01 6.14029706e-01
2.30785668e-01 1.26994908e+00 -8.62831593e-01 -4.27107185e-01
8.84103656e-01 5.10635376e-01 -5.98287463e-01 3.65058601e-01
-3.50243002e-01 2.54186571e-01 -3.24523330e-01 8.30343306e-01
8.29182386e-01 -6.41366482e-01 -3.99401426e-01 -5.16272068e-01
-1.45336628e-01 -5.54030463e-02 -8.46044660e-01 2.54484367e+00
-6.73091173e-01 4.89936471e-01 -2.78522044e-01 -1.04382181e+00
9.33237851e-01 6.94037899e-02 4.69099790e-01 -6.93779886e-01
2.56156977e-02 -2.48153284e-02 -6.97708488e-01 -7.11068630e-01
2.13092864e-01 -1.28612414e-01 2.92969346e-02 3.04504961e-01
4.27294135e-01 -5.47064900e-01 -5.24118900e-01 2.00389728e-01
8.48883033e-01 4.03387100e-01 3.03429961e-01 1.60468772e-01
3.03358078e-01 6.09635599e-02 8.54354650e-02 8.75535786e-01
-8.14633965e-02 1.10500085e+00 1.49845695e-02 -3.21801811e-01
-1.29916847e+00 -1.25948000e+00 -3.10398247e-02 9.01476860e-01
6.01252556e-01 -2.76724786e-01 -2.90773928e-01 -5.90250075e-01
1.19872220e-01 7.27069378e-01 -5.42169213e-01 -6.29338101e-02
-2.05505714e-01 -2.88480371e-02 3.69820118e-01 6.33650601e-01
6.14820361e-01 -5.30526817e-01 -4.10841286e-01 -1.95627660e-01
-8.91368613e-02 -1.45704186e+00 -6.57295942e-01 1.26968846e-01
-7.03119934e-01 -9.07707274e-01 -8.79079401e-01 -8.52486253e-01
5.12440920e-01 1.07043374e+00 1.11070275e+00 -2.14054715e-02
-9.67768729e-02 8.90494406e-01 -5.62669396e-01 -3.32338959e-01
1.61580574e-02 5.03974780e-03 3.94994952e-02 -1.62702501e-01
6.94173157e-01 -8.92305017e-01 -6.94303155e-01 4.02511150e-01
-6.03947341e-01 4.81942326e-01 5.65303802e-01 6.92178130e-01
8.73010278e-01 -5.91160119e-01 5.10014780e-02 -4.54797804e-01
-2.15080604e-01 -5.11635721e-01 -4.60363507e-01 2.85122514e-01
-2.85948545e-01 -1.51289567e-01 4.12246138e-01 -4.38822687e-01
-8.64102006e-01 9.94159579e-02 -1.96692541e-01 -1.44871700e+00
-4.06843096e-01 2.48190522e-01 -2.51320302e-01 -2.74604023e-01
5.15834451e-01 5.49119890e-01 1.61088645e-01 -5.79453707e-01
7.42206693e-01 8.29823494e-01 6.51010752e-01 -5.32545388e-01
1.15431190e+00 9.10657465e-01 -2.52795875e-01 -9.09216106e-01
-1.19685709e+00 -9.18928027e-01 -8.02434266e-01 -1.99387163e-01
9.85480666e-01 -1.37467277e+00 -4.04017061e-01 3.83255690e-01
-1.37462914e+00 -3.21811467e-01 -1.81810021e-01 5.03929853e-01
-8.27972412e-01 4.41426605e-01 -2.99234003e-01 -4.63170469e-01
-3.93841594e-01 -8.79703999e-01 1.65027833e+00 -8.94751623e-02
2.11695209e-01 -7.72720754e-01 8.47789124e-02 5.18045187e-01
4.38912064e-02 -1.42211035e-01 8.26027513e-01 -5.26597798e-01
-8.05934906e-01 -2.64535006e-02 -5.94310284e-01 3.60636145e-01
-1.01631254e-01 -2.67731071e-01 -1.10192657e+00 -2.60254204e-01
5.10708950e-02 -5.87373197e-01 6.95496678e-01 2.44765818e-01
1.14737737e+00 1.63185164e-01 -2.60655046e-01 1.17324078e+00
1.61102402e+00 -2.63547093e-01 5.18579602e-01 2.25326717e-01
1.08286226e+00 3.79438907e-01 6.16586566e-01 3.86728436e-01
6.15398645e-01 8.34873915e-01 6.34675264e-01 3.54648642e-02
-2.92536706e-01 -6.75293028e-01 1.78459361e-01 1.03741610e+00
-7.64793158e-02 -8.62924382e-02 -1.16701055e+00 5.19133925e-01
-1.87695348e+00 -9.54027355e-01 1.68333963e-01 1.95216417e+00
4.83436555e-01 3.83170545e-02 -2.16138333e-01 -4.51756626e-01
5.73453963e-01 4.46084589e-01 -7.59612501e-01 2.58743256e-01
-1.00743540e-01 6.67451546e-02 5.06615877e-01 3.39476407e-01
-9.37718093e-01 1.06372690e+00 4.74513102e+00 9.06937182e-01
-1.16426647e+00 4.08134848e-01 2.32874259e-01 -3.07760060e-01
-3.59953493e-01 6.48897216e-02 -8.96288872e-01 2.20173299e-01
4.33701962e-01 -1.98116884e-01 3.57258141e-01 1.10180092e+00
7.13494420e-02 4.39261824e-01 -1.41636109e+00 1.54855585e+00
5.52529573e-01 -1.62450683e+00 2.58942664e-01 -3.53464251e-03
7.05669224e-01 6.61609113e-01 -1.43264547e-01 4.91647691e-01
-5.60298972e-02 -8.29704583e-01 8.51971447e-01 5.38444281e-01
1.09565556e+00 -3.24729055e-01 4.63181853e-01 5.94220817e-01
-1.16534185e+00 2.02618688e-01 -7.57756710e-01 9.04925726e-03
3.78464043e-01 3.29462379e-01 -5.62232912e-01 8.04439366e-01
8.62073481e-01 1.22387958e+00 -3.53434980e-01 9.41056073e-01
-6.49234578e-02 2.02209517e-01 -3.69137526e-01 1.84228167e-01
4.12592500e-01 -6.40279576e-02 6.67993188e-01 8.31027031e-01
4.93188649e-01 2.05151156e-01 2.61597693e-01 1.08852136e+00
-3.86482850e-03 2.37533469e-02 -8.86254430e-01 1.59748584e-01
5.15338898e-01 9.88238454e-01 -2.02819690e-01 -3.98392677e-01
-9.32096481e-01 1.08543706e+00 4.28683728e-01 3.35644662e-01
-8.82809520e-01 -1.18316717e-01 8.60905051e-01 9.79300812e-02
7.07878768e-01 -3.52723658e-01 -1.93839073e-01 -1.55544436e+00
1.71649992e-01 -5.15174448e-01 8.26845691e-03 -1.24951386e+00
-1.67115796e+00 4.09797370e-01 1.35791197e-01 -1.79447138e+00
1.21092878e-01 -7.39849389e-01 -5.89987338e-01 6.16887629e-01
-1.64228320e+00 -1.61434293e+00 -7.66605079e-01 8.76433969e-01
1.03630567e+00 -8.87911469e-02 6.88279748e-01 2.22327530e-01
-1.87949464e-01 4.88923997e-01 -2.69258171e-02 -4.65984233e-02
7.43100226e-01 -9.71755564e-01 7.67425120e-01 5.16899049e-01
4.31793511e-01 3.68091613e-01 3.98773938e-01 -5.38697422e-01
-1.79065669e+00 -1.16368484e+00 4.89321589e-01 -7.00124264e-01
8.33996594e-01 -6.14642501e-01 -1.12732875e+00 6.14174366e-01
-1.19161662e-02 3.09498847e-01 4.34891880e-01 -1.69094756e-01
-7.74424076e-01 -6.32854700e-02 -7.07249522e-01 5.65081775e-01
1.62054586e+00 -9.44318414e-01 -8.33190858e-01 4.35616553e-01
1.18478501e+00 -7.04371274e-01 -8.98061514e-01 3.21262062e-01
4.87460226e-01 -9.06527281e-01 1.35350609e+00 -4.89170969e-01
7.73844719e-01 -1.07919648e-01 -6.26163304e-01 -1.02891648e+00
-2.40047351e-01 -2.75757372e-01 -2.72959024e-01 1.02326596e+00
8.41918141e-02 -3.56009007e-01 8.03506732e-01 3.74679744e-01
-4.13693964e-01 -7.67391503e-01 -9.44224358e-01 -8.72173607e-01
1.90090463e-01 -8.78013730e-01 5.49823463e-01 9.14431930e-01
-2.73695976e-01 4.17592078e-01 -5.19012988e-01 4.20286596e-01
8.63355517e-01 3.73507172e-01 1.24638081e+00 -1.02192378e+00
-2.99862444e-01 -2.33977139e-01 -6.04029119e-01 -1.64848912e+00
3.90728772e-01 -1.12554705e+00 3.92758735e-02 -1.45995986e+00
3.12320262e-01 -4.60824192e-01 -1.06602393e-01 4.19452786e-01
1.13048293e-01 3.24685007e-01 4.51954037e-01 5.21404862e-01
-6.31462216e-01 8.65634739e-01 1.36090541e+00 -4.59473103e-01
3.84737663e-02 -2.49562547e-01 -3.84441882e-01 6.09066784e-01
5.43865681e-01 -2.61174738e-01 -4.82513696e-01 -8.39829445e-01
6.85813352e-02 4.77765761e-02 8.24396253e-01 -8.77493799e-01
5.33682227e-01 -2.69530594e-01 1.24549411e-01 -9.64609921e-01
6.32529676e-01 -1.00078928e+00 8.69294554e-02 -4.98804599e-02
-1.42998934e-01 -3.50401759e-01 1.17184883e-02 9.04278219e-01
-7.03917295e-02 -1.29523844e-01 4.37848330e-01 -2.45926902e-01
-1.12378216e+00 8.71592164e-01 3.89522046e-01 -2.89555285e-02
9.71492290e-01 -3.94768447e-01 -3.77387464e-01 -2.35647038e-01
-4.95371759e-01 3.36684793e-01 8.65566850e-01 7.97097325e-01
9.42434609e-01 -1.40498638e+00 -5.35651684e-01 5.13200223e-01
8.06538343e-01 4.39100832e-01 7.37717867e-01 9.40490782e-01
-5.01098692e-01 4.27842319e-01 -1.55432433e-01 -1.24341619e+00
-9.79527831e-01 7.71398723e-01 2.87211418e-01 3.13529611e-01
-1.31659281e+00 7.70897925e-01 5.66787302e-01 -6.39361501e-01
4.56777602e-01 -2.12843716e-01 2.69004047e-01 -2.36107171e-01
5.79410672e-01 -8.36304054e-02 -9.02745221e-03 -5.72470188e-01
-4.47513223e-01 1.20172656e+00 -1.17617838e-01 9.28563476e-02
1.45959139e+00 -3.77124369e-01 2.06233218e-01 7.37764120e-01
1.54263985e+00 -4.77667361e-01 -1.60473418e+00 -5.22330821e-01
-5.33638477e-01 -7.40155756e-01 2.30975360e-01 -3.05533320e-01
-9.89159465e-01 1.18449283e+00 5.76857328e-01 -2.91804641e-01
7.76936412e-01 4.49952215e-01 8.39439094e-01 5.03808379e-01
7.32274950e-01 -6.49578333e-01 2.95967400e-01 5.12819886e-01
8.93612564e-01 -1.65799248e+00 -1.39846891e-01 -4.34554011e-01
-7.19607472e-01 1.07777691e+00 7.73107886e-01 -2.31902853e-01
7.13802993e-01 -6.29671514e-02 4.34875637e-02 -4.68349427e-01
-9.46036100e-01 -2.88147271e-01 3.51033062e-01 8.07081103e-01
-2.83593476e-01 -1.73740998e-01 5.14346063e-01 4.63167638e-01
-5.32480851e-02 1.21493071e-01 1.41624928e-01 7.00829387e-01
-4.36813891e-01 -5.53293645e-01 -1.56094596e-01 2.65757799e-01
2.89927959e-01 -2.21034020e-01 -1.59222722e-01 7.77930558e-01
4.49638516e-02 4.41549242e-01 3.23501617e-01 -5.45630336e-01
4.42745328e-01 -6.25427365e-02 5.99381804e-01 -7.40127981e-01
-1.65340230e-02 -1.05928630e-01 -3.10312808e-01 -8.14633071e-01
-5.48038244e-01 -4.13576365e-01 -9.89545047e-01 -2.69775122e-01
-3.47293735e-01 -3.73841107e-01 5.92794716e-01 1.02837706e+00
5.08749723e-01 9.42973197e-02 5.76793373e-01 -1.23706424e+00
-5.63138187e-01 -7.16803074e-01 -4.09100592e-01 3.95505995e-01
3.33789855e-01 -9.31857049e-01 -5.71888864e-01 -1.11248940e-02] | [8.132299423217773, -3.31589674949646] |
5f908701-b8ba-4215-aba1-b0cb7bb83331 | energy-dissipative-evolutionary-deep-operator | 2306.06281 | null | https://arxiv.org/abs/2306.06281v1 | https://arxiv.org/pdf/2306.06281v1.pdf | Energy-Dissipative Evolutionary Deep Operator Neural Networks | Energy-Dissipative Evolutionary Deep Operator Neural Network is an operator learning neural network. It is designed to seed numerical solutions for a class of partial differential equations instead of a single partial differential equation, such as partial differential equations with different parameters or different initial conditions. The network consists of two sub-networks, the Branch net and the Trunk net. For an objective operator G, the Branch net encodes different input functions u at the same number of sensors, and the Trunk net evaluates the output function at any location. By minimizing the error between the evaluated output q and the expected output G(u)(y), DeepONet generates a good approximation of the operator G. In order to preserve essential physical properties of PDEs, such as the Energy Dissipation Law, we adopt a scalar auxiliary variable approach to generate the minimization problem. It introduces a modified energy and enables unconditional energy dissipation law at the discrete level. By taking the parameter as a function of time t, this network can predict the accurate solution at any further time with feeding data only at the initial state. The data needed can be generated by the initial conditions, which are readily available. In order to validate the accuracy and efficiency of our neural networks, we provide numerical simulations of several partial differential equations, including heat equations, parametric heat equations and Allen-Cahn equations. | ['Guang Lin', 'Jie Shen', 'Shiheng Zhang', 'Jiahao Zhang'] | 2023-06-09 | null | null | null | null | ['operator-learning'] | ['miscellaneous'] | [-1.51457610e-02 1.57242343e-01 4.22635585e-01 1.89388260e-01
-9.02721360e-02 -4.14544374e-01 -5.82715683e-02 -1.35551644e-02
-4.72929716e-01 1.04011774e+00 -6.00236595e-01 -1.87309444e-01
-7.94347823e-02 -1.11404300e+00 -7.48830974e-01 -1.04269767e+00
-1.76800266e-01 1.50366826e-02 -4.43015136e-02 -3.16774875e-01
-3.44738849e-02 6.49179339e-01 -1.36449695e+00 -2.79907137e-01
1.11108470e+00 1.48716271e+00 -1.05261028e-01 7.29706407e-01
2.58020669e-01 6.49760246e-01 -5.37723482e-01 1.57550916e-01
5.28814852e-01 -8.77367318e-01 -5.78750134e-01 -3.24928671e-01
-5.03128886e-01 -3.31045419e-01 -3.23698699e-01 1.27892292e+00
7.39720643e-01 6.96750998e-01 7.13649154e-01 -1.27862108e+00
-6.74908817e-01 3.46324056e-01 1.19481876e-01 -1.31412059e-01
-3.23146209e-02 5.59182286e-01 5.92594981e-01 -7.99938023e-01
5.98145247e-01 8.23999643e-01 1.10794663e+00 5.11523902e-01
-1.36139476e+00 -1.76790237e-01 -5.29695868e-01 -1.55064806e-01
-1.57672787e+00 7.87846819e-02 9.87311065e-01 -3.62346411e-01
7.53298938e-01 4.05844331e-01 9.93522525e-01 5.77676594e-01
8.24279308e-01 3.59109551e-01 8.79055619e-01 -2.86456674e-01
7.63725698e-01 2.06689328e-01 -2.70880044e-01 1.06116009e+00
-7.21523389e-02 3.53269815e-01 -5.08451415e-03 -3.92153978e-01
9.75171745e-01 -8.18487182e-02 -5.87652266e-01 -3.71804535e-01
-7.34557569e-01 9.40761626e-01 6.94656074e-01 3.08915913e-01
-5.77709079e-01 3.09400588e-01 8.95136744e-02 4.15159494e-01
3.66435587e-01 9.31359828e-01 -5.24921298e-01 -2.23104246e-02
-7.00895548e-01 2.89057583e-01 1.00172317e+00 4.36994225e-01
1.01045287e+00 3.78883034e-01 -1.27956972e-01 3.48262370e-01
2.17614882e-02 4.68307436e-01 6.09425068e-01 -1.38739693e+00
-2.72384554e-01 7.02850878e-01 2.53640205e-01 -9.05095279e-01
-3.29702765e-01 -3.77023190e-01 -1.26513815e+00 6.08066618e-01
2.37418130e-01 -9.22617793e-01 -7.29719996e-01 1.72410500e+00
3.96028161e-01 8.92981887e-02 7.08455890e-02 1.05866277e+00
4.30494606e-01 1.07549524e+00 -3.55825245e-01 -6.58373058e-01
6.66492105e-01 -7.57119715e-01 -7.24489629e-01 4.64992225e-01
7.75557399e-01 -1.87623009e-01 8.94258082e-01 -7.03730760e-03
-1.50594127e+00 -4.96134102e-01 -9.27841365e-01 3.95510765e-03
-5.42057455e-01 -2.67930645e-02 2.41778433e-01 1.96260631e-01
-1.37853909e+00 1.43237662e+00 -8.98609221e-01 2.11230963e-01
4.42897156e-02 2.19589323e-01 9.52220038e-02 8.46910417e-01
-1.55476844e+00 8.14458847e-01 2.76991755e-01 4.18989986e-01
-6.93317473e-01 -1.28518939e+00 -7.23099411e-01 3.03869754e-01
-3.38557027e-02 -7.18886793e-01 1.03367102e+00 -1.15895164e+00
-2.13520598e+00 3.63932729e-01 -2.12403722e-02 -3.17485571e-01
7.99329758e-01 3.05726856e-01 -5.69584370e-02 1.12446435e-01
-1.37791559e-01 4.86068219e-01 6.14006102e-01 -1.03678823e+00
-7.81817287e-02 1.85348213e-01 -1.84974179e-01 4.90075089e-02
-2.17178196e-01 -3.40135872e-01 2.71665841e-01 -5.19079864e-01
2.39027794e-02 -9.39747989e-01 -4.41414893e-01 5.18129587e-01
-3.73361468e-01 4.49496321e-03 7.13325858e-01 -7.17685580e-01
8.97951305e-01 -2.16328979e+00 7.23573148e-01 3.11994582e-01
8.01525824e-03 -7.39909848e-03 2.92378724e-01 2.53185183e-01
-1.93846852e-01 2.71368593e-01 -9.49268460e-01 -1.62310660e-01
-7.09750876e-03 2.77343005e-01 -2.86980718e-01 4.58704263e-01
4.47156399e-01 1.01384246e+00 -9.94256079e-01 -3.74131471e-01
-9.12019331e-03 6.89351201e-01 -6.61433160e-01 3.79703075e-01
-4.49896991e-01 6.17755711e-01 -5.21914661e-01 1.92650333e-01
5.34893394e-01 -6.14211634e-02 -1.64529353e-01 -1.57032877e-01
-5.99560022e-01 -2.42089778e-01 -1.34255040e+00 1.39200091e+00
-5.49985528e-01 5.04036069e-01 5.93363047e-01 -1.40396607e+00
9.22823727e-01 4.96607393e-01 9.40693021e-01 -5.18027484e-01
6.68647230e-01 4.51746166e-01 9.60953813e-03 -4.53914583e-01
2.54921824e-01 -3.63546938e-01 -4.97520417e-02 5.48239827e-01
-4.78764214e-02 -5.67017078e-01 -3.94525155e-02 -1.99478477e-01
1.00402164e+00 2.37349998e-02 -7.24661052e-02 -6.12528741e-01
6.22225642e-01 -1.41921183e-02 5.96123755e-01 4.55555379e-01
-3.87638398e-02 5.02981722e-01 8.19811165e-01 -4.96923625e-01
-1.15449214e+00 -8.96131754e-01 -3.56910735e-01 5.91389716e-01
2.54806101e-01 2.09443197e-01 -7.81749129e-01 -6.64327070e-02
8.98526907e-02 6.55861378e-01 -8.79324615e-01 -5.52974582e-01
-8.25258136e-01 -6.88585401e-01 3.54073375e-01 3.36916059e-01
9.11572397e-01 -1.35094237e+00 -1.14896691e+00 2.27609441e-01
5.95648363e-02 -5.00220835e-01 -3.35579276e-01 6.31162941e-01
-7.49823868e-01 -9.46973681e-01 -7.77372599e-01 -7.63478577e-01
8.74229968e-01 -8.29389036e-01 7.41380215e-01 3.06706101e-01
-3.65227252e-01 1.31587744e-01 2.05328345e-01 -8.32477063e-02
-5.94853461e-01 -3.18150893e-02 9.92057994e-02 2.20303535e-01
-6.99065506e-01 -7.62552798e-01 -4.83378679e-01 2.14265347e-01
-8.87929618e-01 -1.53896660e-01 5.49761914e-02 8.71227920e-01
8.07440400e-01 4.07170117e-01 5.37377968e-02 -1.11043766e-01
9.35148180e-01 -4.00560498e-01 -8.03166091e-01 1.16167158e-01
-4.30707514e-01 4.73104686e-01 1.15495992e+00 -7.42418170e-01
-9.57616448e-01 7.44307637e-02 -3.41737904e-02 -5.22694945e-01
4.29433942e-01 4.18836772e-01 5.44434562e-02 -3.16846460e-01
7.08478153e-01 3.07015985e-01 1.76970437e-01 -3.21868330e-01
2.38236599e-02 1.59117728e-01 6.51930571e-01 -9.16726887e-01
5.80474615e-01 7.37610087e-02 6.92609549e-01 -7.11008668e-01
-2.29351029e-01 4.38625544e-01 -5.41977108e-01 -3.64639133e-01
7.66628146e-01 -5.37889339e-02 -1.04665923e+00 8.02738249e-01
-1.21107030e+00 -8.66574228e-01 -1.07719636e+00 8.34161118e-02
-4.47976500e-01 -1.18066683e-01 -6.93372130e-01 -1.01823819e+00
-4.30416971e-01 -1.07716906e+00 6.66304886e-01 4.07714218e-01
1.19685167e-02 -1.33835053e+00 1.50288731e-01 -9.04376864e-01
5.98169446e-01 9.34951305e-01 9.15155530e-01 2.62537636e-02
-3.20172429e-01 -7.33272871e-04 3.49602848e-01 5.86003184e-01
-8.44458640e-02 3.45446318e-01 -5.68823636e-01 -2.13761464e-01
5.37161052e-01 -2.01524049e-01 6.42991185e-01 4.87811148e-01
1.27424145e+00 -4.45640028e-01 -2.91682661e-01 9.70462620e-01
1.53648221e+00 3.83854777e-01 3.59458029e-01 -3.23081017e-01
4.13306296e-01 2.22340614e-01 -2.26703897e-01 4.62718368e-01
-2.55822569e-01 2.78136641e-01 3.70527357e-01 -1.08176827e-01
3.26596260e-01 3.33495326e-02 3.42890561e-01 5.96091390e-01
-2.26817310e-01 -7.04875663e-02 -7.02639222e-01 3.64119560e-01
-1.67669821e+00 -7.53440857e-01 -4.37598713e-02 2.21131587e+00
1.04980421e+00 -1.46687850e-01 -1.24925792e-01 2.76162684e-01
5.82834899e-01 -1.76259905e-01 -1.02551091e+00 -9.67546463e-01
-9.09923092e-02 6.14345610e-01 4.33216333e-01 7.92139411e-01
-5.41849971e-01 2.65187562e-01 6.31461239e+00 2.35812962e-01
-1.61920404e+00 -8.22429135e-02 5.74103177e-01 1.16398007e-01
-3.81087452e-01 -1.88208431e-01 -6.47031814e-02 8.02837789e-01
8.39418352e-01 -4.79196638e-01 8.16782415e-01 7.53111899e-01
1.67360619e-01 -1.86462492e-01 -9.91513133e-01 5.39883077e-01
-6.03395045e-01 -1.31173289e+00 -4.70791787e-01 -6.99305609e-02
1.00685883e+00 -2.15951577e-01 1.85018498e-02 1.14457376e-01
2.95995533e-01 -9.43027318e-01 6.67137802e-01 9.84103441e-01
6.82533622e-01 -7.08915472e-01 6.74283266e-01 6.27700388e-01
-1.05416787e+00 -4.51714322e-02 -2.91970193e-01 -4.27371651e-01
3.35284084e-01 7.47767925e-01 -2.20522761e-01 3.27019274e-01
5.60030878e-01 3.74457747e-01 -6.75102789e-03 8.18358600e-01
-1.11355364e-01 2.75128812e-01 -6.75070524e-01 -3.45534921e-01
1.00478753e-01 -7.50257194e-01 5.08672833e-01 6.55702293e-01
6.74704015e-01 3.84395689e-01 4.69375178e-02 1.50188911e+00
-1.57547936e-01 -8.33370723e-04 -3.51048321e-01 1.60321489e-01
2.88208395e-01 1.10179317e+00 -5.83424687e-01 -2.06355527e-01
2.64545679e-01 1.02500665e+00 1.01000555e-01 6.94789231e-01
-8.92430961e-01 -6.75355971e-01 7.02420771e-01 -1.67072773e-01
3.51054929e-02 -2.97905385e-01 -4.14008200e-01 -8.80710304e-01
9.26880911e-02 3.38764489e-02 1.67946711e-01 -9.21323597e-01
-7.89985001e-01 4.58082289e-01 -8.91874433e-02 -8.75994325e-01
-4.03483480e-01 -7.15118885e-01 -1.04933262e+00 1.24111331e+00
-1.01434314e+00 -1.88035131e-01 -1.58835024e-01 5.27801514e-01
-2.04774782e-01 1.51372612e-01 8.69373918e-01 -7.93129951e-03
-8.44214261e-01 2.40831673e-01 3.02554369e-01 7.17346072e-02
-1.16559550e-01 -1.29344189e+00 -6.00745901e-02 7.03513145e-01
-8.12581241e-01 2.54096329e-01 8.64581525e-01 -5.10203302e-01
-1.35112250e+00 -9.41491306e-01 6.30941451e-01 -6.38780147e-02
6.48052216e-01 -2.03038961e-01 -1.23012555e+00 3.97504330e-01
2.42608458e-01 5.71109891e-01 -1.20699525e-01 -7.77023315e-01
6.11619353e-01 -9.82318521e-02 -1.56052518e+00 4.18769866e-01
7.69699693e-01 -3.02351445e-01 -9.42102224e-02 2.36373827e-01
8.06248605e-01 -8.16645682e-01 -1.00107455e+00 5.20389795e-01
3.48772049e-01 -8.27751040e-01 7.15207040e-01 -5.68400621e-01
8.00074220e-01 -2.49719486e-01 1.64963752e-01 -1.72553241e+00
-1.00510076e-01 -1.03843963e+00 -4.90175515e-01 7.58136392e-01
3.73530984e-01 -1.04198551e+00 4.47562546e-01 8.81483316e-01
-1.99207887e-01 -1.42071557e+00 -1.03723383e+00 -8.30213130e-01
5.67415178e-01 -1.66769847e-01 6.09385073e-01 8.68266463e-01
-7.46650100e-02 -2.33051136e-01 -8.99789259e-02 2.67004877e-01
3.25178683e-01 1.00562833e-01 7.78789297e-02 -1.10975397e+00
-2.80320793e-01 -6.55909181e-01 6.25344217e-02 -6.80362403e-01
3.98613364e-01 -8.47635686e-01 2.52262115e-01 -1.33878601e+00
-6.02028370e-01 -4.20197666e-01 -1.60155371e-01 2.59110630e-01
3.53450887e-02 -4.72025052e-02 -1.68917939e-01 -1.00443594e-01
2.86213785e-01 9.26080585e-01 1.55208659e+00 1.08112693e-02
-7.31345654e-01 -6.74663559e-02 -5.53626008e-02 6.67263329e-01
9.28916693e-01 -3.22727889e-01 -3.60258281e-01 -2.19015449e-01
1.75327316e-01 2.38174513e-01 5.68785310e-01 -1.12565446e+00
3.57102573e-01 -2.85023868e-01 2.77007073e-01 2.46725585e-02
2.07778752e-01 -6.56240880e-01 4.38518614e-01 9.38565373e-01
-3.72972757e-01 2.16703966e-01 2.34917793e-02 -5.62251657e-02
-7.78409764e-02 -3.14400733e-01 1.04372609e+00 -1.12312734e-01
-2.10069388e-01 2.80101806e-01 -4.46697056e-01 1.42382145e-01
9.79055822e-01 -3.86473835e-02 2.61590406e-02 -3.04835051e-01
-6.43602550e-01 2.94385314e-01 5.20651281e-01 -1.83666125e-01
3.64713997e-01 -1.55572951e+00 -4.33253616e-01 5.92983902e-01
-6.17808282e-01 4.07474875e-01 8.42386037e-02 7.67281711e-01
-7.02044070e-01 -2.64609665e-01 -1.99752316e-01 -4.22310919e-01
-2.33546227e-01 4.81803000e-01 1.25868559e+00 -4.28088456e-02
-5.05693257e-01 7.25962162e-01 -3.09859067e-01 -5.77782691e-01
-6.22679889e-02 -7.40134358e-01 3.09045494e-01 -7.68199414e-02
7.83424228e-02 5.98764002e-01 -1.03446290e-01 -3.51978302e-01
-9.92885977e-02 5.72435915e-01 1.06157351e+00 -2.59036124e-01
1.25250649e+00 3.82083684e-01 -5.05708218e-01 5.01849115e-01
1.54479718e+00 -3.90997678e-01 -1.47924232e+00 5.44965081e-02
-6.05771780e-01 2.49657393e-01 1.83075845e-01 -6.36565685e-01
-1.48681831e+00 7.69578457e-01 3.63262206e-01 7.91386962e-01
1.30167258e+00 -4.42904770e-01 1.04146945e+00 2.71545470e-01
1.28350750e-01 -1.35451019e+00 -9.77302417e-02 6.82163358e-01
9.13744092e-01 -5.61343670e-01 -1.69766158e-01 -8.61469284e-02
-7.59982318e-02 1.31450307e+00 5.53698599e-01 -3.53138715e-01
8.88732135e-01 6.10443294e-01 -1.40654147e-01 1.30415067e-01
-4.11431760e-01 2.72216111e-01 1.89566702e-01 1.10544324e-01
2.18219563e-01 -1.26673087e-01 -4.41486955e-01 4.49766517e-01
-2.10360751e-01 6.09787926e-02 5.26052713e-01 9.13541317e-01
-3.31165016e-01 -7.19976425e-01 -2.81064749e-01 1.25927180e-01
-1.87558420e-02 2.78257400e-01 -2.59635091e-01 7.62765825e-01
3.03918898e-01 3.17555130e-01 3.59602690e-01 -3.09187779e-03
3.93428385e-01 4.17087555e-01 2.59358793e-01 -1.22723468e-01
-5.06286442e-01 -4.86999065e-01 -3.53444546e-01 -5.89394629e-01
-7.74666891e-02 -4.57003713e-01 -1.82523048e+00 -4.09604877e-01
1.55243528e-04 5.52055240e-01 5.16108036e-01 8.71568084e-01
3.25861484e-01 8.22991848e-01 8.87522697e-01 -1.04687190e+00
-5.76643765e-01 -6.55549943e-01 -6.95457220e-01 1.73597589e-01
6.72344625e-01 -4.86303955e-01 -6.93412840e-01 -3.05407513e-02] | [6.504624843597412, 3.431114912033081] |
073d4e5a-7969-408a-bc6e-f49e4e61d82a | paxqa-generating-cross-lingual-question | 2304.12206 | null | https://arxiv.org/abs/2304.12206v1 | https://arxiv.org/pdf/2304.12206v1.pdf | PAXQA: Generating Cross-lingual Question Answering Examples at Training Scale | Existing question answering (QA) systems owe much of their success to large, high-quality training data. Such annotation efforts are costly, and the difficulty compounds in the cross-lingual setting. Therefore, prior cross-lingual QA work has focused on releasing evaluation datasets, and then applying zero-shot methods as baselines. In this work, we propose a synthetic data generation method for cross-lingual QA which leverages indirect supervision from existing parallel corpora. Our method termed PAXQA ({P}rojecting {a}nnotations for cross-lingual ({x}) QA) decomposes cross-lingual QA into two stages. In the first stage, we apply a question generation (QG) model to the English side. In the second stage, we apply annotation projection to translate both the questions and answers. To better translate questions, we propose a novel use of lexically-constrained machine translation, in which constrained entities are extracted from the parallel bitexts. We release cross-lingual QA datasets across 4 languages, totaling 662K QA examples. We then show that extractive QA models fine-tuned on these datasets outperform both zero-shot and prior synthetic data generation models, showing the sufficient quality of our generations. We find that the largest performance gains are for cross-lingual directions with non-English questions and English contexts. Ablation studies show that our dataset generation method is relatively robust to noise from automatic word alignments. | ['Chris Callison-Burch', 'Bryan Li'] | 2023-04-24 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation', 'cross-lingual-question-answering', 'question-generation'] | ['medical', 'miscellaneous', 'natural-language-processing', 'natural-language-processing'] | [ 1.49595216e-01 2.67595619e-01 8.07575230e-03 -4.85656947e-01
-1.94313478e+00 -9.43952203e-01 6.76908851e-01 -3.04612964e-01
-3.73550415e-01 9.58755672e-01 5.58234811e-01 -6.28183007e-01
2.68826663e-01 -8.28970969e-01 -8.47641408e-01 -2.07521185e-01
6.72746599e-01 9.51121807e-01 1.38108894e-01 -7.37734914e-01
-2.30182245e-01 -3.97105038e-01 -9.65906382e-01 5.91425598e-01
1.41631961e+00 5.15667856e-01 9.32260603e-03 3.97325635e-01
-4.61653233e-01 6.97393239e-01 -7.25063443e-01 -1.30366790e+00
3.18299294e-01 -8.17679644e-01 -1.23534489e+00 -1.36200860e-01
6.66456461e-01 -2.20400944e-01 1.08084805e-01 6.89104438e-01
5.80172658e-01 -8.73817280e-02 4.65026259e-01 -1.12770116e+00
-1.10119259e+00 8.83803129e-01 -1.29979938e-01 -2.50811130e-01
5.72970331e-01 3.48924041e-01 1.67436945e+00 -1.11094475e+00
8.88684094e-01 1.13638067e+00 6.04879320e-01 9.13460195e-01
-1.29113829e+00 -3.01832169e-01 -3.22651476e-01 1.93973899e-01
-1.18246746e+00 -6.09862089e-01 5.63802302e-01 -9.32644531e-02
1.07866073e+00 2.01224282e-01 9.61683318e-02 1.27749479e+00
-6.17824793e-02 9.17026997e-01 1.15525651e+00 -8.64313066e-01
1.42623439e-01 9.18107182e-02 -5.22274338e-02 5.19453645e-01
-3.05306586e-03 -3.11439276e-01 -4.51469213e-01 -2.18258470e-01
9.62276533e-02 -7.41057873e-01 -3.04280311e-01 1.36673097e-02
-1.27743423e+00 1.09755290e+00 1.01152971e-01 1.16952658e-01
-2.08244324e-01 -6.55990243e-02 4.21411633e-01 5.86541355e-01
3.86650890e-01 9.40894485e-01 -9.09432471e-01 -1.97462112e-01
-6.97546244e-01 4.31161851e-01 9.62567747e-01 1.15037847e+00
7.00881779e-01 -2.84446865e-01 -5.48850894e-01 1.15701771e+00
1.70788467e-01 7.79531181e-01 7.01984465e-01 -1.09966516e+00
1.11767423e+00 6.91495359e-01 2.11493269e-01 -2.53808051e-01
1.72998086e-02 2.96460986e-02 -2.19197214e-01 -4.03437972e-01
6.11324131e-01 -5.07770717e-01 -9.44984972e-01 1.83469570e+00
4.20038670e-01 -7.63323843e-01 6.63543344e-01 5.23432136e-01
8.01701963e-01 7.89322197e-01 1.70038179e-01 6.86659589e-02
1.79728091e+00 -1.36759770e+00 -8.13179493e-01 -3.86873782e-01
9.57647800e-01 -9.71027434e-01 1.84550655e+00 -9.64098144e-03
-1.18367445e+00 -4.10345435e-01 -7.72072613e-01 -6.10882342e-01
-3.78261834e-01 3.73117328e-01 3.36503536e-01 7.93909609e-01
-8.51468503e-01 1.73720419e-02 -6.57986343e-01 -3.31594467e-01
6.92582056e-02 -1.56206623e-01 -2.77589858e-01 -5.69156945e-01
-1.64108360e+00 1.09265196e+00 1.88717797e-01 -1.10032998e-01
-7.15174735e-01 -5.76802552e-01 -1.16740632e+00 -1.77173004e-01
4.19589996e-01 -9.68332946e-01 1.55554628e+00 -7.04831243e-01
-1.60168028e+00 7.94281721e-01 -4.22111601e-01 -4.06490326e-01
2.45073751e-01 -4.73338515e-01 -4.68365133e-01 -3.09255924e-02
5.99578679e-01 8.42365205e-01 5.65773606e-01 -1.04822612e+00
-5.29934466e-01 -2.04212770e-01 2.56794542e-01 3.76080811e-01
-7.68207982e-02 1.88390389e-01 -5.27595758e-01 -5.10876477e-01
-1.68115467e-01 -9.08746839e-01 -5.84359244e-02 -5.40601492e-01
-3.56514901e-01 -4.97605294e-01 3.47467482e-01 -9.34408188e-01
1.15230370e+00 -1.60684800e+00 1.12219655e-03 -2.05871791e-01
-5.05675077e-01 1.81976557e-01 -3.85966033e-01 7.91644990e-01
2.68839896e-01 1.60921708e-01 -5.19433200e-01 -5.35567224e-01
3.40719759e-01 5.92249036e-01 -7.18589544e-01 -3.30722243e-01
6.96014702e-01 1.42380810e+00 -8.30805182e-01 -5.70703149e-01
-3.72250199e-01 5.55424765e-02 -7.26711214e-01 4.37406003e-01
-5.98594666e-01 2.92466938e-01 -2.02082574e-01 7.39473760e-01
2.27430016e-01 -1.54508740e-01 3.13912004e-01 -2.46601611e-01
1.93710983e-01 1.14618576e+00 -6.04737878e-01 2.12019205e+00
-8.82238686e-01 1.77949607e-01 -3.13892215e-01 -3.92094880e-01
6.96507096e-01 6.55705154e-01 -6.16379678e-02 -9.01606858e-01
-7.79866502e-02 6.44427061e-01 5.08101918e-02 -5.82073748e-01
6.98764265e-01 -3.23895782e-01 -4.92446989e-01 7.19472408e-01
4.29861248e-01 -7.62728810e-01 6.08708680e-01 3.95023614e-01
1.10487747e+00 5.63357532e-01 9.05903652e-02 2.14789994e-02
4.62524503e-01 6.60560489e-01 3.91186774e-01 4.25283253e-01
6.22618496e-02 6.74172103e-01 4.03281510e-01 -3.28516513e-02
-1.09394860e+00 -9.51087654e-01 -5.12857698e-02 1.10266531e+00
-2.60883600e-01 -6.91160500e-01 -1.15458190e+00 -1.14148486e+00
-3.43303502e-01 1.22635603e+00 -5.08659780e-01 -5.45124747e-02
-9.70654905e-01 -7.86253393e-01 1.13489807e+00 3.00615430e-01
4.10419732e-01 -1.13125241e+00 -2.77422786e-01 3.65281671e-01
-9.38205361e-01 -1.31210089e+00 -5.98606527e-01 -5.09210452e-02
-5.17899096e-01 -8.86315942e-01 -6.42699122e-01 -8.08146238e-01
4.26006585e-01 -1.35988355e-01 1.66466725e+00 -3.01936954e-01
1.98668793e-01 3.44547927e-01 -6.02348626e-01 -2.54219025e-01
-6.57838464e-01 4.03004855e-01 -1.89202964e-01 -1.36435196e-01
6.24426186e-01 -1.50163680e-01 -3.14289540e-01 2.07198188e-01
-8.25045168e-01 2.68189586e-03 5.05869508e-01 1.00556111e+00
7.18156993e-01 -7.01203287e-01 8.01110029e-01 -1.09787822e+00
1.01195931e+00 -5.03179133e-01 -4.47569668e-01 7.54398406e-01
-4.13107693e-01 2.97108740e-01 8.09768677e-01 -1.36686325e-01
-1.41562462e+00 -1.25445992e-01 -5.05478561e-01 1.92084014e-01
-8.03474977e-04 5.70368826e-01 -5.97998023e-01 4.72564250e-01
9.30692971e-01 1.50939077e-01 -3.26907754e-01 -5.06602466e-01
1.12618184e+00 7.06430912e-01 4.69558269e-01 -1.06431091e+00
7.54850030e-01 1.21458717e-01 -6.63694739e-01 -3.19396466e-01
-1.20697677e+00 -1.46524400e-01 -5.51656604e-01 2.77149558e-01
1.02656102e+00 -9.29059744e-01 -1.38894901e-01 1.36275381e-01
-1.39604962e+00 -4.81500536e-01 -6.56443357e-01 2.84099013e-01
-5.33090591e-01 2.04713345e-01 -8.26090693e-01 -3.68931711e-01
-7.20066190e-01 -1.14598429e+00 1.26387870e+00 2.66553722e-02
-3.84022534e-01 -9.77021873e-01 5.77580810e-01 1.00852406e+00
3.55638593e-01 -2.71711648e-01 1.12146354e+00 -7.53635466e-01
-4.94757801e-01 -4.26741131e-03 -1.95065141e-02 5.11482954e-01
2.56735742e-01 -2.78513968e-01 -7.94689357e-01 -4.52479050e-02
-3.98102542e-03 -1.10836351e+00 3.58664900e-01 -3.00114095e-01
4.60018754e-01 -4.67365831e-01 1.59544244e-01 2.64642090e-01
1.17696619e+00 -1.30438998e-01 5.59794009e-01 2.86051273e-01
6.19199753e-01 8.56160104e-01 7.09138989e-01 -3.23638707e-01
1.10035253e+00 7.94170380e-01 -5.22844866e-03 1.93843134e-02
-4.66655314e-01 -5.29541016e-01 7.11131334e-01 1.37225175e+00
3.58511806e-01 -4.03742164e-01 -1.08134806e+00 1.07806623e+00
-1.54699588e+00 -6.37995720e-01 -2.12694362e-01 1.98904383e+00
1.59205878e+00 -1.81910902e-01 -2.43662357e-01 -4.18649167e-01
2.70255446e-01 -1.80327117e-01 -2.63291657e-01 -4.00312245e-01
-4.21955287e-01 7.73691714e-01 5.06527871e-02 6.82612002e-01
-7.19593167e-01 1.38036704e+00 5.62796688e+00 8.56444478e-01
-5.88043988e-01 5.66096187e-01 3.88496101e-01 9.26530808e-02
-8.85641992e-01 4.00678277e-01 -1.04232228e+00 3.79723519e-01
1.15601528e+00 -8.90356600e-02 2.77346075e-01 6.44570768e-01
-3.16268325e-01 6.82411194e-02 -1.11093056e+00 5.33046246e-01
3.59559983e-01 -1.15422511e+00 3.41012210e-01 -1.66692629e-01
9.89701152e-01 2.70308316e-01 -1.99140295e-01 7.81554163e-01
9.54269588e-01 -9.63549614e-01 6.47885323e-01 1.59576703e-02
1.08675575e+00 -5.20547509e-01 8.21242392e-01 3.34081382e-01
-9.30062890e-01 2.92743862e-01 -3.50485623e-01 2.88326055e-01
8.41461241e-01 3.00588071e-01 -8.71608913e-01 8.51828814e-01
5.06677032e-01 3.24985944e-02 -6.04363978e-01 4.34829146e-01
-9.13937807e-01 9.95509028e-01 -2.60529637e-01 1.20377094e-01
3.60546649e-01 -2.73963124e-01 2.11648464e-01 1.08337235e+00
3.42162997e-01 6.60027098e-03 -5.80691546e-02 8.90934944e-01
-5.08239210e-01 4.59099233e-01 -4.21616912e-01 4.67583016e-02
4.82312411e-01 1.32808709e+00 -4.55857143e-02 -5.87135255e-01
-6.64928913e-01 1.25337422e+00 6.56541407e-01 2.75868475e-01
-5.51692784e-01 -6.15139067e-01 4.27088231e-01 -2.92845696e-01
2.05554739e-01 -1.75963596e-01 -2.11517274e-01 -1.55691636e+00
3.20184797e-01 -1.47806442e+00 6.32284224e-01 -8.26606631e-01
-1.64136684e+00 9.06989932e-01 -2.86162555e-01 -9.84909475e-01
-6.92681968e-01 -4.62059468e-01 -3.47972780e-01 1.14545822e+00
-1.65356445e+00 -1.66726744e+00 9.19272006e-02 6.86671734e-01
8.11669171e-01 6.65821042e-03 1.14970458e+00 3.99391681e-01
-3.27849388e-01 8.03615332e-01 -1.06970549e-01 3.85008216e-01
1.16728115e+00 -1.42906570e+00 9.85103130e-01 1.00892472e+00
6.02980018e-01 9.13493752e-01 3.49583447e-01 -6.72115147e-01
-1.29070246e+00 -1.15053201e+00 1.56550682e+00 -1.19811630e+00
9.27828252e-01 -5.85476100e-01 -1.01815474e+00 8.52293611e-01
7.41946936e-01 -2.17951074e-01 8.90452087e-01 1.68252140e-01
-5.01154900e-01 5.89011274e-02 -9.16641772e-01 7.81654000e-01
6.46525443e-01 -7.57591546e-01 -1.22516990e+00 3.73508364e-01
1.16181231e+00 -3.72146249e-01 -9.13819253e-01 3.31852525e-01
3.16729873e-01 -3.44906271e-01 5.56355059e-01 -8.51169169e-01
6.78707242e-01 -3.93786371e-01 -3.52384239e-01 -1.50767374e+00
2.86981314e-01 -6.17104888e-01 1.30721971e-01 1.58841407e+00
1.05915594e+00 -4.49462682e-01 3.98350537e-01 5.85961938e-01
-3.79805654e-01 -5.76326787e-01 -1.05601943e+00 -7.51396000e-01
6.02581799e-01 -5.26915967e-01 8.00173998e-01 1.06168306e+00
6.37658238e-02 1.11104095e+00 -2.60697126e-01 -6.22459874e-02
2.86108464e-01 2.59398550e-01 9.46297407e-01 -5.11369228e-01
-5.65142155e-01 4.46329936e-02 4.48501796e-01 -1.27101660e+00
6.84730038e-02 -1.00547850e+00 2.56959468e-01 -1.67599154e+00
-6.35810867e-02 -6.41876340e-01 2.93255448e-01 6.62107289e-01
-5.28077066e-01 3.57013673e-01 -9.82407387e-03 1.97682187e-01
-4.62825269e-01 9.47443366e-01 1.13049841e+00 6.03501052e-02
-5.92877995e-03 -3.62236977e-01 -7.00708628e-01 3.76881450e-01
6.75239146e-01 -5.25565803e-01 -4.25710857e-01 -1.15926671e+00
4.50288802e-01 -6.15186766e-02 -5.60287684e-02 -6.29251182e-01
1.25976428e-01 1.31406942e-02 -2.20847249e-01 -3.24024647e-01
4.21624064e-01 -4.25610036e-01 -3.49664927e-01 6.67482093e-02
-3.40681463e-01 2.95428902e-01 8.40768144e-02 3.07205897e-02
-4.71957803e-01 -4.26775426e-01 3.41829211e-01 -3.05377007e-01
-2.20573828e-01 2.59749097e-04 -9.29193646e-02 8.65290105e-01
5.12369931e-01 2.55505741e-01 -8.00817132e-01 -3.87351722e-01
-3.52272421e-01 4.27612394e-01 4.08948720e-01 5.07209897e-01
1.15023628e-01 -1.45076013e+00 -1.04529035e+00 1.29474290e-02
5.66874981e-01 1.31551176e-02 2.83283368e-02 6.43570781e-01
-3.49646747e-01 6.32163167e-01 3.72704342e-02 -1.96560234e-01
-8.09553146e-01 2.26873189e-01 2.97291249e-01 -5.64015806e-01
-8.16063676e-03 9.54388857e-01 -1.26354769e-01 -1.22380090e+00
-2.84770370e-01 -5.01178689e-02 -2.04095920e-03 4.14273329e-02
3.70257765e-01 5.36347441e-02 4.10136163e-01 -6.87560856e-01
-1.76182970e-01 3.58565688e-01 -2.05939204e-01 -6.36667192e-01
1.05159366e+00 -2.10781306e-01 -2.04474568e-01 2.36781821e-01
8.43224466e-01 6.34161472e-01 -8.64445984e-01 -4.60051894e-01
3.94544423e-01 -9.95795056e-02 -5.97920835e-01 -1.27175212e+00
-5.63091815e-01 1.07219911e+00 2.44279616e-02 -4.79612239e-02
9.16420698e-01 3.54494750e-01 1.41414309e+00 4.35965568e-01
4.93602663e-01 -1.12083054e+00 1.01787381e-01 8.63230586e-01
9.29679573e-01 -1.34689653e+00 -5.22826433e-01 -2.72550970e-01
-9.93594348e-01 6.80833101e-01 8.31250131e-01 2.81798810e-01
-1.91385355e-02 -1.77931190e-01 6.26236916e-01 -2.99970299e-01
-1.00885713e+00 -4.15932298e-01 3.40459079e-01 5.04012942e-01
6.38449192e-01 3.07424255e-02 -5.18261552e-01 9.13626194e-01
-8.20904970e-01 -2.48707578e-01 3.49485457e-01 8.61652613e-01
-1.62968156e-03 -1.81546223e+00 -1.22980744e-01 1.39455944e-01
-7.21082032e-01 -7.98322141e-01 -5.05232632e-01 7.67281592e-01
1.78230464e-01 1.18944252e+00 -2.09663823e-01 -1.30683362e-01
6.22339547e-01 5.57510972e-01 3.93373162e-01 -9.28744793e-01
-7.10541189e-01 -2.58886665e-01 5.89009523e-01 -3.92057419e-01
-2.04800114e-01 -6.46557629e-01 -9.71221387e-01 1.41724512e-01
-4.46773976e-01 5.38459718e-01 6.21146739e-01 1.13874817e+00
5.77743351e-01 1.77442774e-01 2.93831527e-01 2.78827418e-02
-7.70281315e-01 -1.20019591e+00 1.10648729e-01 4.75764066e-01
-1.15731709e-01 -8.25109035e-02 -2.05294505e-01 2.37080768e-01] | [11.350125312805176, 8.375720024108887] |
4abf1100-169b-4bbb-a7a7-60d3748488f4 | explainability-of-text-processing-and | 2212.07126 | null | https://arxiv.org/abs/2212.07126v1 | https://arxiv.org/pdf/2212.07126v1.pdf | Explainability of Text Processing and Retrieval Methods: A Critical Survey | Deep Learning and Machine Learning based models have become extremely popular in text processing and information retrieval. However, the non-linear structures present inside the networks make these models largely inscrutable. A significant body of research has focused on increasing the transparency of these models. This article provides a broad overview of research on the explainability and interpretability of natural language processing and information retrieval methods. More specifically, we survey approaches that have been applied to explain word embeddings, sequence modeling, attention modules, transformers, BERT, and document ranking. The concluding section suggests some possible directions for future research on this topic. | ['Mandar Mitra', 'Debapriyo Majumdar', 'Sourav Saha'] | 2022-12-14 | null | null | null | null | ['document-ranking'] | ['natural-language-processing'] | [ 7.66739622e-03 4.29444313e-01 -5.25593162e-01 -5.53918123e-01
-1.82384238e-01 -5.22843599e-01 1.01877892e+00 5.82957923e-01
-3.27109903e-01 2.08686680e-01 6.97269320e-01 -7.48086214e-01
-2.45707959e-01 -4.51211035e-01 -2.67079741e-01 -9.00139436e-02
-1.36967555e-01 4.81141150e-01 -3.79109591e-01 -1.40202060e-01
6.56489611e-01 4.51187521e-01 -1.34437346e+00 7.02364087e-01
5.79470932e-01 6.60209358e-01 -1.64102716e-03 7.84412980e-01
-6.31685495e-01 1.01917768e+00 -5.29884279e-01 -7.96258569e-01
-2.97137499e-01 -1.96612403e-01 -1.20747817e+00 -1.95055529e-01
4.53294069e-01 -4.43301678e-01 -7.35660613e-01 9.40957308e-01
1.86161965e-01 1.34965971e-01 1.06642497e+00 -1.20184481e+00
-1.61287439e+00 8.67237329e-01 -2.58645386e-01 3.83860081e-01
1.92564175e-01 -1.28134578e-01 1.58032930e+00 -1.13623273e+00
3.82227749e-01 1.57544243e+00 4.84823585e-01 8.11225593e-01
-8.17262650e-01 -2.27467760e-01 4.65464383e-01 6.71030223e-01
-1.12028992e+00 -1.93605945e-02 4.80646431e-01 -3.71356934e-01
1.39637995e+00 2.84286529e-01 5.53740144e-01 1.11055124e+00
7.43433237e-01 1.03848183e+00 6.04001880e-01 -6.54259622e-01
-4.74495143e-02 3.87230098e-01 9.80545044e-01 6.38577938e-01
5.74657738e-01 1.72163233e-01 -4.77321595e-01 -3.42506431e-02
5.85078478e-01 3.34380865e-01 -1.69543192e-01 -3.17017198e-01
-8.91460896e-01 1.24005544e+00 6.48615003e-01 5.23948610e-01
-3.39658976e-01 4.28174168e-01 5.34426987e-01 2.15899318e-01
6.30302727e-01 8.39554548e-01 -6.02973580e-01 1.31321978e-02
-7.16058671e-01 1.49266332e-01 7.14890659e-01 8.04938972e-01
4.15825814e-01 2.56651402e-01 -9.28352103e-02 5.49472153e-01
7.39998043e-01 1.63434207e-01 6.30847335e-01 -5.08174241e-01
1.89636230e-01 4.14605737e-01 8.60595331e-03 -1.21896839e+00
-6.40999079e-01 -3.13355595e-01 -8.03106308e-01 -2.20540687e-01
-1.48844510e-01 1.65722921e-01 -9.34434354e-01 1.34099555e+00
-3.57166052e-01 -2.51064479e-01 1.63933337e-01 1.02267969e+00
1.02303874e+00 9.99526203e-01 3.42314601e-01 1.58367887e-01
1.48699856e+00 -1.25620639e+00 -1.05991697e+00 -5.98835945e-01
7.22384393e-01 -6.26098871e-01 1.03958488e+00 1.35008216e-01
-1.06766987e+00 -5.09384990e-01 -8.89950991e-01 -7.07656860e-01
-6.38183594e-01 -6.42298386e-02 1.08165324e+00 4.95784342e-01
-1.10963488e+00 4.32590455e-01 -1.00597632e+00 -5.05603313e-01
3.59326571e-01 3.37550253e-01 -3.07944231e-02 -2.03237340e-01
-1.43956077e+00 1.28684628e+00 3.89048845e-01 1.74320862e-01
-6.29216611e-01 -5.93099535e-01 -9.70571458e-01 4.54269588e-01
-2.37680465e-01 -8.14490855e-01 1.42386889e+00 -9.02260125e-01
-1.10144556e+00 7.55613565e-01 -3.71444523e-01 -6.85263991e-01
-1.52177453e-01 -5.46642423e-01 -5.33463776e-01 -3.44793573e-02
-3.67862821e-01 7.08901346e-01 4.61311847e-01 -1.06710792e+00
-2.60621637e-01 -5.87999821e-01 1.16864495e-01 1.68485537e-01
-7.25218713e-01 3.00919652e-01 -2.13377744e-01 -6.04395092e-01
-1.89292133e-01 -6.92895174e-01 -1.13491170e-01 4.47759964e-03
-4.42589939e-01 -6.59655988e-01 3.79444033e-01 -7.26361334e-01
1.45223856e+00 -2.00636339e+00 2.00256035e-01 -1.12129971e-01
5.31313896e-01 3.61425489e-01 -4.18945909e-01 9.92414474e-01
-2.32880086e-01 8.67603779e-01 1.40387669e-01 -1.90663517e-01
2.60297030e-01 2.36075446e-01 -7.86810100e-01 3.37352097e-01
2.95772016e-01 1.30491757e+00 -7.86474705e-01 -1.11039989e-01
2.87554771e-01 8.32752824e-01 -5.04651666e-01 2.06614792e-01
-2.13187858e-01 -2.16383249e-01 -5.30261219e-01 2.31303960e-01
3.10376883e-01 -4.66929615e-01 2.26889417e-01 5.03179133e-02
-6.88433722e-02 9.22997475e-01 -6.09926581e-01 1.46747279e+00
-5.09152770e-01 1.38144445e+00 -3.89207393e-01 -1.00173092e+00
5.45762181e-01 3.76506835e-01 -5.53598255e-02 -6.36778772e-01
1.02488413e-01 -1.43262610e-01 1.90976173e-01 -6.18041635e-01
1.06274819e+00 -1.33807302e-01 1.87820598e-01 7.64278054e-01
-1.12552449e-01 7.10950717e-02 9.94063821e-03 5.62517285e-01
8.16240549e-01 -4.03173715e-01 3.50560337e-01 -2.03412518e-01
3.31642509e-01 -4.25381251e-02 -1.19257227e-01 9.83878911e-01
-8.85930434e-02 4.35749084e-01 2.72840649e-01 -9.35611844e-01
-1.28511393e+00 -8.00154865e-01 -1.13771647e-01 1.27127492e+00
1.49773747e-01 -7.52321482e-01 -5.87831259e-01 -4.57744896e-01
7.14757890e-02 1.02043784e+00 -8.93334627e-01 -4.70815331e-01
-3.52645248e-01 -4.34632391e-01 4.75998610e-01 8.68451059e-01
1.23113941e-03 -1.17409408e+00 -1.88983470e-01 -6.96452111e-02
-7.41157383e-02 -7.37118542e-01 -2.81257212e-01 2.33444534e-02
-1.16518557e+00 -8.09769154e-01 -3.24558914e-01 -9.04769003e-01
6.44109488e-01 6.50378287e-01 1.30341327e+00 5.28497338e-01
-2.99822599e-01 6.79107904e-01 -5.40218472e-01 -6.35817468e-01
-2.57871568e-01 1.20777406e-01 7.00737759e-02 -5.10684013e-01
9.76312935e-01 1.16261035e-01 -5.27267218e-01 -1.48910239e-01
-1.43606114e+00 -4.54397164e-02 5.26810706e-01 1.06706452e+00
1.62550956e-01 -1.66590348e-01 3.21182638e-01 -1.10052288e+00
1.20168102e+00 -5.67190528e-01 -1.28278211e-01 3.52587402e-01
-1.05052865e+00 4.51713771e-01 3.85925919e-01 -3.72135520e-01
-8.63767266e-01 -6.59868300e-01 -1.43979788e-01 6.26348183e-02
1.56682022e-02 1.03636897e+00 2.34255001e-01 2.68858910e-01
7.02255011e-01 1.14206150e-01 -3.37479003e-02 -3.74283075e-01
8.48870158e-01 7.82325327e-01 -1.23560339e-01 -4.67972718e-02
6.54258430e-01 2.61755586e-01 -5.61814606e-01 -9.46031451e-01
-1.00031543e+00 -3.16957831e-01 -3.75593215e-01 2.39911541e-01
7.30372250e-01 -6.43002450e-01 -3.66249412e-01 2.25009173e-02
-1.78112531e+00 6.54235780e-02 -1.11506991e-01 5.44703305e-01
-1.18074380e-01 4.74141449e-01 -9.38601494e-01 -7.91547894e-01
-4.97153610e-01 -8.26577008e-01 8.24465811e-01 1.78799778e-01
-7.85900474e-01 -1.58816731e+00 6.37413189e-02 1.66564822e-01
6.01959944e-01 -5.07074714e-01 1.30422306e+00 -1.08904982e+00
-3.30608010e-01 -3.60448927e-01 -3.50931138e-01 1.85130477e-01
-1.02559477e-01 1.54789016e-01 -9.55212295e-01 -1.18698746e-01
-4.54324372e-02 -3.62482190e-01 1.04882133e+00 5.83513618e-01
1.36223078e+00 -5.96769214e-01 -4.39654708e-01 2.25677326e-01
1.21146965e+00 1.77365616e-01 6.75474346e-01 3.78223449e-01
5.01879930e-01 7.73965120e-01 2.37838462e-01 2.90344357e-01
5.11875749e-01 1.93434775e-01 4.36211765e-01 -1.14341497e-01
-1.27177134e-01 -4.44262177e-01 2.58908838e-01 1.12708020e+00
2.69537479e-01 -9.50253844e-01 -1.01655304e+00 4.97400731e-01
-1.74463201e+00 -1.07820046e+00 -1.89484090e-01 1.66246831e+00
4.38157320e-01 -1.75812557e-01 -5.13064981e-01 -2.07040124e-02
7.41748750e-01 3.02051187e-01 -5.04981995e-01 -1.06951773e+00
-1.32708132e-01 1.37243256e-01 7.45703503e-02 7.42834270e-01
-8.56461883e-01 1.10832024e+00 7.99801254e+00 2.65742391e-01
-9.50073957e-01 -1.72659993e-01 6.55279994e-01 1.49081022e-01
-8.21243405e-01 -1.13366693e-01 -5.31266689e-01 1.14129540e-02
1.21281493e+00 -3.45212072e-01 3.69271457e-01 8.15785289e-01
2.48240858e-01 4.14647758e-01 -1.44643331e+00 8.76942158e-01
2.44473577e-01 -1.56053305e+00 6.31095707e-01 -6.38533011e-02
5.21318793e-01 1.65475935e-01 4.16069806e-01 3.58638436e-01
2.80608743e-01 -1.22587371e+00 5.33907175e-01 3.63842994e-01
2.92837024e-01 -5.48321664e-01 9.16373610e-01 1.12210847e-01
-6.22319877e-01 -2.12112427e-01 -6.83077753e-01 -5.92751980e-01
6.86398670e-02 3.35401356e-01 -6.59602880e-01 1.35002792e-01
5.01555860e-01 9.52711225e-01 -4.56119388e-01 8.26858878e-01
-5.62243223e-01 6.66990757e-01 3.01524818e-01 -6.39087617e-01
3.45733672e-01 1.18010126e-01 2.64188200e-01 1.41654181e+00
-1.82049386e-02 7.59092867e-02 -1.77257285e-01 8.82785320e-01
-1.83314577e-01 4.45982516e-02 -7.39382684e-01 -6.85861409e-01
4.75715071e-01 9.90273237e-01 -4.57011968e-01 -4.19302911e-01
-6.54513776e-01 9.65860844e-01 4.72793967e-01 5.93087018e-01
-6.68167889e-01 -3.99616152e-01 9.27131772e-01 -1.43051520e-02
8.59035626e-02 -5.44866979e-01 -5.35700619e-01 -1.38208997e+00
-1.08244017e-01 -8.89306784e-01 4.76418436e-01 -1.03380871e+00
-1.47953331e+00 7.96121895e-01 -1.26163378e-01 -6.88271701e-01
-4.47170675e-01 -9.98448372e-01 -4.09265131e-01 7.33896315e-01
-1.61368585e+00 -7.39835680e-01 4.32405770e-02 3.68469581e-02
7.99831748e-01 -3.34679782e-02 1.14211309e+00 -1.33539783e-02
-5.31312168e-01 3.93649459e-01 3.67867351e-01 8.19574669e-02
4.26519573e-01 -1.04063880e+00 8.15417469e-01 5.22254169e-01
6.65013194e-01 1.29411232e+00 7.19073296e-01 -2.05478355e-01
-1.73538208e+00 -7.98541248e-01 1.48613930e+00 -7.14130759e-01
8.60664070e-01 -3.20100754e-01 -1.06618333e+00 1.00440383e+00
7.60725021e-01 -5.43268383e-01 9.47026730e-01 4.51803029e-01
-6.06504440e-01 1.84104085e-01 -6.38344824e-01 7.75050521e-01
5.11494219e-01 -8.35487962e-01 -8.07378650e-01 4.92714852e-01
8.71603608e-01 6.46248041e-03 -5.19925296e-01 7.97799416e-03
6.27583265e-01 -7.37913549e-01 8.15351248e-01 -1.29599440e+00
8.09077621e-01 2.02627480e-01 2.63655603e-01 -1.40495646e+00
-6.94944620e-01 -4.01497990e-01 -3.38703930e-01 8.47010851e-01
6.10785246e-01 -5.92883289e-01 6.54080987e-01 1.19800448e+00
-2.55790532e-01 -7.07892299e-01 -4.38392073e-01 -5.12875617e-01
6.17237329e-01 -6.77100420e-01 2.75502980e-01 8.91959548e-01
5.68631351e-01 7.04036534e-01 -2.70320594e-01 -4.60678805e-03
4.07841392e-02 -1.84468795e-02 4.12456155e-01 -1.28559315e+00
7.18417019e-02 -6.75193965e-01 -4.78919744e-01 -1.49717832e+00
4.13387060e-01 -1.07309222e+00 -1.23580068e-01 -2.14255953e+00
5.45467496e-01 2.53388315e-01 -3.85085523e-01 3.84658873e-01
-6.68834746e-02 -5.69771603e-02 5.97003959e-02 2.81328082e-01
-5.95580578e-01 5.64166605e-01 9.39432561e-01 -4.40236747e-01
1.67432606e-01 -4.20878738e-01 -1.17146385e+00 5.92219949e-01
1.14122105e+00 -3.92479867e-01 -4.82548803e-01 -1.46023405e+00
6.77305579e-01 -2.79694855e-01 1.67247996e-01 -1.42641813e-01
3.12763244e-01 -8.78201351e-02 3.20976704e-01 -3.99149597e-01
2.30152383e-01 -8.23438525e-01 -4.54405159e-01 5.96754730e-01
-1.16308570e+00 5.83454370e-01 3.76260459e-01 6.60131395e-01
-4.24334377e-01 -3.15218180e-01 2.50700742e-01 -6.86664656e-02
-6.97225094e-01 3.37494254e-01 -9.95299578e-01 -1.10844739e-01
4.39078778e-01 1.54963419e-01 -4.35114026e-01 -7.04670429e-01
-4.91209060e-01 2.28659168e-01 2.89488044e-02 9.86046791e-01
9.15795326e-01 -1.16985559e+00 -5.81551552e-01 -8.75193402e-02
1.60749808e-01 -6.23844981e-01 1.58013832e-02 1.73886150e-01
-3.93010467e-01 1.29756355e+00 8.12403709e-02 -1.26585171e-01
-9.94042993e-01 8.91206920e-01 1.14295028e-01 -2.50910640e-01
-4.45761561e-01 7.85136700e-01 2.97685921e-01 -3.33359063e-01
3.91435653e-01 -4.62328434e-01 -4.49541986e-01 -3.08632791e-01
8.85428965e-01 3.04489821e-01 -1.49497181e-01 -3.23373675e-01
-4.77617234e-01 2.23348275e-01 -5.79424977e-01 -7.45453313e-02
1.23721337e+00 -2.69005328e-01 -4.32842433e-01 4.48916912e-01
1.37796772e+00 -4.79780316e-01 -3.62842411e-01 -1.57173976e-01
3.65480185e-01 -1.45598263e-01 2.47061983e-01 -7.29783356e-01
-7.85172760e-01 1.67355907e+00 3.50265622e-01 7.02568233e-01
7.40995169e-01 1.41344909e-02 5.48834920e-01 6.64941847e-01
-1.78401440e-01 -9.13002431e-01 6.10452779e-02 9.62501705e-01
9.89571452e-01 -1.26771891e+00 -6.19251914e-02 -1.60175830e-01
-5.45318782e-01 1.49512446e+00 5.58325171e-01 -1.68120101e-01
8.03825557e-01 -4.52625193e-02 2.69708317e-02 -4.40501124e-01
-1.08671832e+00 1.71419948e-01 4.70390022e-01 5.62624633e-01
1.21819735e+00 -1.66519478e-01 -6.59196198e-01 6.61974967e-01
-1.11083575e-01 -3.39846641e-01 4.71358120e-01 5.80311477e-01
-5.16478062e-01 -1.09840381e+00 -8.38894770e-02 3.00985157e-01
-5.48122764e-01 -8.56598914e-01 -9.02965605e-01 7.48974562e-01
-6.61184013e-01 1.12218022e+00 4.01707262e-01 -4.18211341e-01
-1.43190771e-01 1.77065223e-01 1.51260629e-01 -7.32165277e-01
-6.00720882e-01 -4.35999393e-01 4.11930606e-02 -3.44907194e-01
-2.23053187e-01 -2.04719171e-01 -1.26651251e+00 -4.66219902e-01
-3.89857024e-01 4.19942886e-01 9.01488602e-01 9.08455610e-01
7.59391665e-01 5.65971255e-01 2.48273581e-01 -4.47737247e-01
-5.92202425e-01 -1.17763245e+00 -3.67741466e-01 2.26862669e-01
3.03019375e-01 -2.47763574e-01 -4.00361240e-01 3.82518582e-02] | [10.858099937438965, 8.109405517578125] |
ad691664-2799-43ea-98f5-abce0aad7ca1 | corri2p-deep-image-to-point-cloud | 2207.05483 | null | https://arxiv.org/abs/2207.05483v3 | https://arxiv.org/pdf/2207.05483v3.pdf | CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondence | Motivated by the intuition that the critical step of localizing a 2D image in the corresponding 3D point cloud is establishing 2D-3D correspondence between them, we propose the first feature-based dense correspondence framework for addressing the image-to-point cloud registration problem, dubbed CorrI2P, which consists of three modules, i.e., feature embedding, symmetric overlapping region detection, and pose estimation through the established correspondence. Specifically, given a pair of a 2D image and a 3D point cloud, we first transform them into high-dimensional feature space and feed the resulting features into a symmetric overlapping region detector to determine the region where the image and point cloud overlap each other. Then we use the features of the overlapping regions to establish the 2D-3D correspondence before running EPnP within RANSAC to estimate the camera's pose. Experimental results on KITTI and NuScenes datasets show that our CorrI2P outperforms state-of-the-art image-to-point cloud registration methods significantly. We will make the code publicly available. | ['Xiaodong Chen', 'Junhui Hou', 'Yiming Zeng', 'Siyu Ren'] | 2022-07-12 | null | null | null | null | ['point-cloud-registration', 'image-to-point-cloud-registration'] | ['computer-vision', 'computer-vision'] | [-3.75505909e-02 -4.23001796e-01 9.57500339e-02 -2.63509154e-01
-8.68884087e-01 -7.98415661e-01 6.82238102e-01 -1.28107131e-01
-1.75026432e-01 -1.64077953e-01 -1.61617905e-01 3.78994793e-02
9.75196362e-02 -6.61787868e-01 -7.43140399e-01 -5.26796758e-01
1.50830418e-01 7.19489694e-01 4.32211429e-01 4.28217053e-02
5.99642396e-01 9.66411352e-01 -1.31112647e+00 -3.12186360e-01
5.63945472e-01 1.14546204e+00 9.71703902e-02 2.31876180e-01
-7.45886937e-02 -1.38293803e-01 -4.08074446e-02 1.04571935e-02
7.50842094e-01 -3.34488861e-02 -6.37493193e-01 4.70930338e-01
4.82404113e-01 -3.90561253e-01 -2.65935063e-01 1.05486965e+00
1.32614374e-01 6.60327375e-02 6.91116035e-01 -1.36538792e+00
-3.68880183e-01 -3.84108305e-01 -1.07753634e+00 -1.34697706e-01
7.74294198e-01 -3.97039093e-02 6.33325100e-01 -1.53144884e+00
6.75621569e-01 1.30493546e+00 8.15995514e-01 -7.04600289e-03
-1.10435581e+00 -7.19187438e-01 -1.80768818e-01 -2.60116696e-01
-1.87674129e+00 -2.80115336e-01 1.00203288e+00 -7.24792182e-01
6.28898025e-01 1.13379754e-01 6.88248754e-01 2.59068191e-01
-2.69242320e-02 3.25731963e-01 9.94352043e-01 -4.95348811e-01
-1.32356316e-01 -8.39560330e-02 6.03447706e-02 7.20278203e-01
-1.17676005e-01 3.01509142e-01 -3.54714930e-01 -4.95486647e-01
1.22271812e+00 3.02311689e-01 -2.32174948e-01 -9.08411145e-01
-1.39203918e+00 7.53016055e-01 6.60716414e-01 1.16726801e-01
-2.47527137e-01 -3.94438133e-02 -8.39550272e-02 -2.09273547e-02
5.33618033e-01 2.34399438e-01 -6.63995296e-02 9.20765176e-02
-7.80764759e-01 2.33938798e-01 3.65540087e-01 1.24529302e+00
1.32714820e+00 -6.51162922e-01 5.46084344e-01 4.63868707e-01
5.79761684e-01 6.35647595e-01 1.55675545e-01 -8.99389625e-01
5.58906257e-01 9.53207791e-01 2.83378422e-01 -1.54009748e+00
-1.99298143e-01 1.04374744e-01 -5.91993928e-01 1.00825936e-01
2.89163571e-02 2.71546930e-01 -5.77064395e-01 9.17048216e-01
8.95653307e-01 5.40230095e-01 -5.84933050e-02 1.14657688e+00
7.09735215e-01 6.30944669e-01 -7.29053080e-01 1.76811948e-01
1.35146487e+00 -6.43504798e-01 -1.30203784e-01 -1.78415045e-01
3.58216286e-01 -1.06318343e+00 5.63359559e-01 -2.95457155e-01
-9.52064395e-01 -5.20737529e-01 -1.03393686e+00 -2.45208964e-01
-1.30533114e-01 2.93423444e-01 1.65302604e-01 1.27010234e-02
-7.33296514e-01 3.81631851e-01 -9.55435157e-01 -3.87796372e-01
2.16153845e-01 3.82109493e-01 -8.02834749e-01 -1.12995550e-01
-6.27101898e-01 7.30275869e-01 2.72101164e-01 1.30516157e-01
-3.15490127e-01 -6.90603137e-01 -1.14923894e+00 -2.67646849e-01
3.28776658e-01 -6.56348944e-01 8.74973953e-01 -5.15288353e-01
-1.08960462e+00 1.29519761e+00 -3.50706995e-01 1.12159833e-01
4.00263667e-01 -1.49097115e-01 -1.72259882e-01 1.88403606e-01
4.12524164e-01 5.29564619e-01 8.73009264e-01 -1.61424410e+00
-6.09370589e-01 -8.14510763e-01 -3.19182515e-01 4.38713133e-01
2.98969656e-01 2.09875792e-01 -1.03631520e+00 -1.78575978e-01
1.02114046e+00 -1.14112067e+00 -1.05194144e-01 3.50287378e-01
-4.78995949e-01 -1.44850031e-01 1.22058189e+00 -3.85001600e-01
6.73579276e-01 -2.44218278e+00 3.14486958e-02 5.86277723e-01
3.40747237e-01 -2.42114048e-02 -2.25676134e-01 2.82280982e-01
-2.06029445e-01 -1.85296327e-01 -1.59835681e-01 -5.18234968e-01
-2.62985766e-01 -3.69301401e-02 -1.56142250e-01 1.14667428e+00
3.03032309e-01 8.55832577e-01 -9.41235065e-01 -5.07448733e-01
6.10941231e-01 4.85275418e-01 -2.00776055e-01 3.64903063e-01
4.30897981e-01 5.39725363e-01 -7.14169145e-01 6.44746244e-01
1.30219495e+00 -4.66172807e-02 -4.60146874e-01 -5.61094522e-01
-5.73137283e-01 5.32982014e-02 -1.48803282e+00 1.75663555e+00
-6.39957190e-02 4.09726173e-01 5.31234685e-03 -5.83633959e-01
1.22719657e+00 -3.28526809e-03 8.30020964e-01 -2.79760569e-01
2.15701014e-01 2.28121698e-01 -4.48599964e-01 -1.69284552e-01
5.79049051e-01 2.00450420e-01 -2.84500979e-02 2.09158093e-01
-1.29508257e-01 -7.02960789e-01 -4.87579495e-01 4.77577001e-02
6.50799215e-01 1.36921301e-01 3.29114914e-01 -1.38624489e-01
7.41206765e-01 1.84499040e-01 5.07027447e-01 2.22836018e-01
8.86110589e-03 1.12915504e+00 1.32147014e-01 -3.95079315e-01
-1.33837533e+00 -1.05102289e+00 -3.00748825e-01 -7.55459629e-03
8.33033502e-01 -4.63364065e-01 -5.99421263e-01 -6.05898201e-01
3.44024241e-01 -4.50946242e-02 -5.18247604e-01 1.84575632e-01
-5.44566274e-01 2.49785814e-03 -6.90551028e-02 4.20527488e-01
5.27991295e-01 -3.19930881e-01 -5.51178575e-01 -1.69772312e-01
-9.22392607e-02 -1.30907798e+00 -1.05795288e+00 -2.37234727e-01
-8.79874289e-01 -1.29211950e+00 -5.39951563e-01 -9.54008639e-01
1.02404368e+00 9.78152037e-01 8.15589726e-01 1.18247181e-01
-1.25550553e-01 4.87126738e-01 -2.76434213e-01 -8.05629343e-02
-1.01794444e-01 -2.81320155e-01 2.12679043e-01 1.57673329e-01
4.67539400e-01 -4.49286759e-01 -6.68736577e-01 9.48494196e-01
-6.29301846e-01 1.37260348e-01 3.24032485e-01 5.57349801e-01
1.07169271e+00 -9.99220163e-02 -4.30438042e-01 -3.33788157e-01
5.12014069e-02 -2.05570295e-01 -1.05429173e+00 5.71859032e-02
-1.27664700e-01 -1.41460016e-01 -2.13167258e-02 -2.00392857e-01
-4.09206361e-01 7.80315936e-01 9.73404869e-02 -1.05119467e+00
-4.88996282e-02 4.00498122e-01 -2.07705215e-01 -5.90556383e-01
3.33975136e-01 3.62802833e-01 2.35066980e-01 -4.33759362e-01
4.22951341e-01 8.35043669e-01 7.88597047e-01 -2.92348653e-01
1.58144879e+00 8.74706388e-01 8.00333321e-02 -7.61928201e-01
-6.65022612e-01 -1.26106274e+00 -1.40404356e+00 -1.08504027e-01
1.02188694e+00 -1.07741702e+00 -6.19647324e-01 4.44129676e-01
-1.48964250e+00 4.62804019e-01 -9.18069109e-02 6.47048175e-01
-6.76050484e-01 6.82867825e-01 -1.03318952e-01 -4.75335568e-01
-2.97613591e-01 -1.33152235e+00 1.67208993e+00 1.67383164e-01
1.10181555e-01 -6.19710386e-01 4.39477175e-01 3.51156920e-01
-1.20277047e-01 3.59281451e-01 4.85524893e-01 -1.99416414e-01
-7.44496047e-01 -7.48984754e-01 -4.52270687e-01 1.30145505e-01
2.70798087e-01 1.67157754e-01 -8.24798942e-01 -3.57093662e-01
2.17707694e-01 1.27869651e-01 1.67749107e-01 9.34946612e-02
7.86731064e-01 1.39544994e-01 -6.08809173e-01 9.61537957e-01
1.51247990e+00 -3.78948860e-02 6.16418779e-01 4.32960272e-01
8.73374045e-01 3.69216084e-01 9.82790172e-01 2.83695787e-01
4.23179150e-01 1.14478695e+00 6.69170499e-01 -2.45138302e-01
1.25486106e-01 -5.80370784e-01 9.00533125e-02 8.83099556e-01
-7.55869895e-02 5.58453321e-01 -9.35875654e-01 3.99857134e-01
-1.81162488e+00 -5.66878557e-01 -2.97265619e-01 2.50222349e+00
4.45623815e-01 -2.76313931e-01 4.24853824e-02 -2.29605764e-01
1.02634132e+00 -1.00829788e-02 -4.45336282e-01 2.54157484e-01
1.87140793e-01 -1.25609308e-01 6.37740791e-01 4.95469213e-01
-1.24255347e+00 1.01874852e+00 5.29910851e+00 5.69949150e-01
-1.09389293e+00 -1.16759643e-01 7.04756826e-02 3.56278569e-01
1.25190439e-02 4.40464675e-01 -8.85198295e-01 3.60302448e-01
3.19754183e-01 -1.01002976e-01 2.38168642e-01 9.49886262e-01
6.58370480e-02 -9.34884325e-02 -1.23841333e+00 1.50817120e+00
2.20820874e-01 -1.45124698e+00 -7.81298950e-02 2.48783901e-01
6.99727714e-01 3.42550337e-01 -2.30348155e-01 -2.89877564e-01
-9.38025191e-02 -7.39420414e-01 8.00466895e-01 2.84312487e-01
9.28156316e-01 -6.33483469e-01 5.73595166e-01 4.81185347e-01
-1.63855028e+00 3.88438791e-01 -6.69540048e-01 3.01213980e-01
1.86480656e-01 4.71158028e-01 -6.84724212e-01 8.12600315e-01
6.51962519e-01 1.03252912e+00 -4.00890946e-01 1.24623621e+00
-1.00285960e-02 -1.87939808e-01 -5.51697254e-01 5.72417259e-01
1.92228615e-01 -7.17959166e-01 7.37569094e-01 6.32529497e-01
6.06957853e-01 2.90545285e-01 3.75934124e-01 1.12640333e+00
4.61866334e-02 -4.03406247e-02 -8.86189997e-01 4.56204653e-01
8.18298340e-01 1.36273921e+00 -6.81451023e-01 -3.22754830e-02
-5.49618542e-01 1.17782772e+00 8.36754888e-02 -3.10870092e-02
-7.46749640e-01 -3.27148706e-01 8.15487504e-01 3.49776447e-01
3.66991431e-01 -6.43747985e-01 -1.75475813e-02 -1.29392159e+00
2.51427859e-01 -4.45339054e-01 -7.91122019e-02 -1.09091544e+00
-1.17940760e+00 5.60919225e-01 1.47390217e-02 -1.96265745e+00
7.37291798e-02 -4.63460416e-01 -7.91895211e-01 1.23076987e+00
-1.33078265e+00 -1.27789474e+00 -7.57343709e-01 8.24057996e-01
1.98507652e-01 2.46037945e-01 6.62725151e-01 1.61969773e-02
-2.33375996e-01 2.23128170e-01 3.50264192e-01 5.44027507e-01
4.59341735e-01 -7.81614244e-01 7.46276855e-01 7.31441259e-01
2.66263127e-01 6.41865551e-01 9.71058384e-02 -6.54621959e-01
-1.82232642e+00 -1.04170299e+00 6.15270078e-01 -8.58985960e-01
5.64145327e-01 -6.11402392e-01 -8.25026155e-01 7.38244176e-01
-2.69412965e-01 4.01764035e-01 1.05655164e-01 -3.28697324e-01
-3.67537290e-01 7.39623420e-03 -1.18762147e+00 3.14564407e-01
9.46871698e-01 -7.17001259e-01 -7.81136274e-01 3.79466146e-01
5.73173225e-01 -1.06138706e+00 -9.88075614e-01 2.48266548e-01
3.20038259e-01 -6.57675147e-01 1.33344233e+00 -6.99076205e-02
1.43508792e-01 -7.26431191e-01 -3.47134858e-01 -1.21901166e+00
-2.72827536e-01 -5.51480532e-01 4.22690958e-01 1.12346125e+00
-4.43758890e-02 -6.14170134e-01 8.30157876e-01 5.49043238e-01
-1.87203974e-01 -5.20049334e-01 -1.26246512e+00 -8.55912387e-01
-1.05925031e-01 -3.07023555e-01 8.00045252e-01 9.59379137e-01
-1.51374891e-01 1.73454151e-01 9.73555818e-03 7.78175414e-01
8.04845095e-01 5.43552160e-01 1.34690142e+00 -1.29770386e+00
3.00493419e-01 -3.77144143e-02 -9.78747129e-01 -1.44531167e+00
1.53708160e-01 -7.78655112e-01 1.67826131e-01 -1.07634032e+00
2.69862890e-01 -7.84518659e-01 3.22297603e-01 2.35572010e-01
2.85066478e-02 3.67488414e-01 2.51258165e-01 8.85504484e-01
-3.23758096e-01 6.62283123e-01 1.04926145e+00 9.57286209e-02
-2.47508183e-01 -3.32264937e-02 -3.19947243e-01 7.84422755e-01
2.26667941e-01 -4.96020675e-01 1.06776461e-01 -4.77942884e-01
-3.39354217e-01 4.81978729e-02 5.36414742e-01 -1.10475647e+00
3.25855404e-01 -2.52784580e-01 4.58229512e-01 -1.22726905e+00
5.86250365e-01 -1.31099367e+00 3.36383581e-01 6.94596991e-02
1.66887194e-01 3.42082262e-01 9.47564319e-02 4.89341110e-01
-3.25137168e-01 -2.32230321e-01 7.73564875e-01 7.40774907e-03
-6.01235509e-01 7.81823575e-01 3.89124542e-01 -1.84645981e-01
1.41429722e+00 -6.49068713e-01 -7.76901543e-02 -1.48932919e-01
-2.78026700e-01 2.95578152e-01 1.19272697e+00 4.05254930e-01
1.04591477e+00 -1.65759254e+00 -7.08292842e-01 7.03205407e-01
5.42724550e-01 6.46551311e-01 3.16047133e-03 9.15266454e-01
-7.59962738e-01 4.65313703e-01 -1.23706758e-02 -1.33046877e+00
-1.60846603e+00 4.96424824e-01 4.89151150e-01 2.92378813e-01
-9.52196777e-01 6.71584368e-01 3.31080079e-01 -7.51290560e-01
-1.17751807e-01 -4.54881698e-01 1.83108255e-01 -3.03742558e-01
4.11248058e-01 1.33802325e-01 1.22407496e-01 -1.31748760e+00
-7.12027907e-01 1.53918099e+00 -3.40004116e-02 -4.25786972e-02
1.28245318e+00 -1.46085054e-01 -3.50488752e-01 9.10388678e-02
1.74347496e+00 8.83073639e-03 -1.45015442e+00 -4.95168805e-01
-2.36868843e-01 -1.18392372e+00 1.67481765e-01 -2.32664752e-03
-1.15869832e+00 6.94488645e-01 6.70242667e-01 -1.32213295e-01
8.15643072e-01 4.52081680e-01 7.40448952e-01 3.45802605e-02
6.21225178e-01 -6.23492956e-01 -9.96269658e-02 5.84487855e-01
8.78157496e-01 -1.23843205e+00 3.05715442e-01 -7.90781677e-01
-4.08683538e-01 1.25710404e+00 3.81977439e-01 -3.97454470e-01
7.93349862e-01 -5.98240830e-02 5.14742583e-02 -5.35971463e-01
5.17146178e-02 4.04914618e-02 5.58752418e-01 6.84461772e-01
-1.25327095e-01 -1.24413311e-01 2.08744869e-01 -4.76603061e-02
-1.80384740e-01 -2.13940330e-02 2.80684322e-01 1.00035882e+00
-2.72484213e-01 -1.02616978e+00 -8.91984761e-01 9.91285667e-02
3.17275017e-01 1.08757593e-01 -4.71494853e-01 8.65379035e-01
-7.86703303e-02 6.34053409e-01 5.92827618e-01 -6.05958641e-01
5.48017085e-01 -2.62475103e-01 4.37393457e-01 -7.43517578e-01
-3.63597631e-01 1.58412188e-01 -4.40093517e-01 -8.11658859e-01
-4.98849779e-01 -9.01182413e-01 -1.33624828e+00 -3.90338719e-01
-6.96749210e-01 1.57867938e-01 9.48300660e-01 7.84348607e-01
5.80566049e-01 -5.33297539e-01 1.20395446e+00 -1.49245286e+00
-5.31191647e-01 -5.45965552e-01 -6.27636433e-01 5.48818946e-01
3.74467134e-01 -8.59051108e-01 -4.43732977e-01 -1.51130676e-01] | [7.65981388092041, -2.840498447418213] |
11cc611f-6ef0-4634-834b-9fe38004b0df | combining-verbal-and-nonverbal-features-to | null | null | https://aclanthology.org/W12-1634 | https://aclanthology.org/W12-1634.pdf | Combining Verbal and Nonverbal Features to Overcome the ``Information Gap'' in Task-Oriented Dialogue | null | ['Christopher Mitchell', 'Kristy Elizabeth Boyer', 'Joseph F. Grafsgaard', 'James C. Lester', 'Eun Young Ha'] | 2012-07-01 | null | null | null | ws-2012-7 | ['dialogue-act-classification'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.288228511810303, 3.7071192264556885] |
14e0359b-e145-483c-a71d-f858b2562f26 | a-corpus-for-detecting-high-context-medical | 2003.03044 | null | https://arxiv.org/abs/2003.03044v1 | https://arxiv.org/pdf/2003.03044v1.pdf | A Corpus for Detecting High-Context Medical Conditions in Intensive Care Patient Notes Focusing on Frequently Readmitted Patients | A crucial step within secondary analysis of electronic health records (EHRs) is to identify the patient cohort under investigation. While EHRs contain medical billing codes that aim to represent the conditions and treatments patients may have, much of the information is only present in the patient notes. Therefore, it is critical to develop robust algorithms to infer patients' conditions and treatments from their written notes. In this paper, we introduce a dataset for patient phenotyping, a task that is defined as the identification of whether a patient has a given medical condition (also referred to as clinical indication or phenotype) based on their patient note. Nursing Progress Notes and Discharge Summaries from the Intensive Care Unit of a large tertiary care hospital were manually annotated for the presence of several high-context phenotypes relevant to treatment and risk of re-hospitalization. This dataset contains 1102 Discharge Summaries and 1000 Nursing Progress Notes. Each Discharge Summary and Progress Note has been annotated by at least two expert human annotators (one clinical researcher and one resident physician). Annotated phenotypes include treatment non-adherence, chronic pain, advanced/metastatic cancer, as well as 10 other phenotypes. This dataset can be utilized for academic and industrial research in medicine and computer science, particularly within the field of medical natural language processing. | ['Jonathan Welt', 'Joy T. Wu', 'Franck Dernoncourt', 'David W. Grant', 'Leo Anthony Celi', 'John Foote', 'Edward T. Moseley', 'Eric T. Carlson', 'Sebastian Gehrmann', 'Patrick D. Tyler'] | 2020-03-06 | a-corpus-for-detecting-high-context-medical-1 | https://aclanthology.org/2020.lrec-1.170 | https://aclanthology.org/2020.lrec-1.170.pdf | lrec-2020-5 | ['patient-phenotyping'] | ['medical'] | [ 4.98536974e-01 3.06105852e-01 -6.07210040e-01 -4.74233061e-01
-9.31021929e-01 -5.24098456e-01 -1.79297268e-01 1.38622153e+00
-2.07028180e-01 9.22056794e-01 8.60930324e-01 -4.63497579e-01
-4.00244385e-01 -5.88163733e-01 -1.28527477e-01 -4.28012818e-01
4.21996787e-02 8.74940813e-01 -8.10128093e-01 5.94518006e-01
-5.74980266e-02 2.80581504e-01 -1.04017472e+00 5.15870929e-01
6.65981293e-01 8.77269149e-01 8.30061641e-03 6.38989925e-01
2.33704850e-01 9.28067982e-01 -4.33207959e-01 -2.80277580e-01
7.43339304e-03 -5.77229202e-01 -7.39463806e-01 -1.52799208e-02
-3.36982612e-03 -1.74542442e-01 4.93052043e-02 8.69202852e-01
6.82538807e-01 -3.83151531e-01 6.22303128e-01 -9.03028607e-01
-3.99819881e-01 4.81106788e-01 1.77562580e-01 3.98907028e-02
8.35971415e-01 -7.23051652e-02 9.59707081e-01 -2.56395876e-01
7.78210163e-01 5.19854903e-01 8.25017750e-01 5.93421161e-01
-1.10835755e+00 -2.92293668e-01 -3.82843643e-01 -3.12639207e-01
-1.35662627e+00 -3.66413176e-01 6.46437258e-02 -8.71364176e-01
6.32336795e-01 5.90455890e-01 7.43932009e-01 8.78086686e-01
3.83044094e-01 3.22082460e-01 7.00528204e-01 -2.26955071e-01
5.22125363e-01 9.24094394e-02 3.40473294e-01 6.12376750e-01
5.65593183e-01 -2.42828920e-01 -6.85609505e-02 -1.20659065e+00
4.67383683e-01 7.37203538e-01 -4.93033111e-01 5.68836741e-02
-1.51523852e+00 5.25597095e-01 -2.26799682e-01 -5.88998795e-02
-8.39357734e-01 -3.73100638e-01 5.30751705e-01 9.00913775e-02
7.26267248e-02 4.88742560e-01 -1.02398980e+00 -2.76548833e-01
-7.62855291e-01 1.17204078e-01 1.09071279e+00 1.00958848e+00
3.14560831e-01 -6.96365595e-01 -6.40626073e-01 7.25593328e-01
3.30123186e-01 2.41998389e-01 8.17103326e-01 -7.88951099e-01
3.17986637e-01 1.21793473e+00 4.21155274e-01 -6.03767276e-01
-8.46565008e-01 3.58159244e-02 -9.92733955e-01 -6.27960920e-01
4.11410779e-02 -4.15370941e-01 -5.38255334e-01 1.34387958e+00
3.95330757e-01 -2.04300154e-02 3.74317616e-01 5.41625857e-01
9.73552525e-01 1.05555780e-01 5.88261604e-01 -6.99924052e-01
1.99353004e+00 -2.40751266e-01 -9.89316106e-01 1.48293570e-01
1.11387122e+00 -5.42215109e-01 6.45856082e-01 2.26280972e-01
-9.56923008e-01 2.44191419e-02 -3.72013211e-01 1.93295717e-01
-3.10456365e-01 3.18967164e-01 2.38898367e-01 4.46945310e-01
-3.79837364e-01 5.24338186e-01 -8.56101871e-01 -6.24939442e-01
7.24509060e-01 2.55629092e-01 -4.52964783e-01 -9.13386568e-02
-9.72267866e-01 6.56795263e-01 3.23207915e-01 -2.69239515e-01
-2.67515421e-01 -1.10682940e+00 -1.06699002e+00 1.37264103e-01
2.19138175e-01 -1.23621774e+00 1.22773480e+00 -2.95429289e-01
-6.51465416e-01 1.18002176e+00 -3.63578498e-01 -3.26291978e-01
1.95482433e-01 -1.09559871e-01 -7.66261041e-01 5.30244373e-02
2.64625728e-01 4.86733904e-03 1.35377392e-01 -6.19323194e-01
-6.59251928e-01 -7.69915521e-01 -3.77829939e-01 1.93672001e-01
-2.36974910e-01 2.04512194e-01 -2.36774117e-01 -7.93539882e-01
-1.69155449e-01 -8.59385729e-01 -4.81554657e-01 -2.53110733e-02
-4.76743788e-01 3.80918048e-02 3.36773157e-01 -9.66840148e-01
1.83963060e+00 -2.15726376e+00 -2.34692276e-01 -1.32664070e-01
4.14907098e-01 1.16860904e-01 5.14465213e-01 5.77493489e-01
-2.36863330e-01 4.22822356e-01 -5.09631217e-01 -7.68265268e-03
-2.03916162e-01 2.35799909e-01 6.62139952e-02 5.20814776e-01
8.95034969e-02 1.04030788e+00 -8.62760901e-01 -6.23850644e-01
1.13601305e-01 3.38556796e-01 -6.43108666e-01 4.13674802e-01
-1.39921695e-01 7.24039376e-01 -6.47330761e-01 9.06672657e-01
1.02792710e-01 -7.46902704e-01 4.62364018e-01 -6.18023947e-02
8.27981904e-02 4.48147804e-01 -8.38278890e-01 1.42360842e+00
-3.21567841e-02 -8.02985281e-02 1.96598724e-01 -6.03952169e-01
5.38148761e-01 8.78713965e-01 1.11889040e+00 -6.07672147e-02
1.24837182e-01 -1.05641969e-01 -1.17480882e-01 -1.14262867e+00
2.47042235e-02 -2.89886981e-01 -2.81848192e-01 7.29039252e-01
-7.39796638e-01 2.44788274e-01 7.30208084e-02 -4.71744239e-02
1.66126168e+00 -3.45135361e-01 1.15678489e+00 -2.24801496e-01
4.36687440e-01 3.71220231e-01 1.00455713e+00 3.79742295e-01
-1.61838651e-01 5.64055502e-01 3.23054850e-01 -5.85549116e-01
-8.04797232e-01 -7.33120322e-01 -4.92106050e-01 4.21806097e-01
-3.85659158e-01 -6.81349158e-01 -3.37072611e-01 -3.20919037e-01
1.38026148e-01 4.88622844e-01 -5.98657191e-01 -2.40090996e-01
-2.41000801e-01 -9.85222876e-01 5.89360416e-01 5.74707031e-01
-9.54482108e-02 -1.28371441e+00 -1.26521909e+00 5.65677702e-01
-3.60055447e-01 -9.30280268e-01 -5.59191048e-01 1.49094000e-01
-1.02775645e+00 -1.69578767e+00 -5.22353053e-01 -6.93832874e-01
7.76274323e-01 -6.94985330e-01 1.19353843e+00 1.42269790e-01
-7.65603602e-01 5.46487570e-01 -4.09702837e-01 -4.70881909e-01
-6.63271129e-01 -1.66955248e-01 5.40630147e-02 -9.17926058e-02
7.47618854e-01 -5.07378317e-02 -6.90809369e-01 -7.76464343e-02
-1.06946552e+00 -6.71522319e-02 4.32117999e-01 7.69778311e-01
8.06861639e-01 -5.86812152e-03 4.62195307e-01 -1.37890768e+00
8.63376558e-01 -7.48614848e-01 -6.72443062e-02 5.07066011e-01
-7.26920605e-01 -1.99612781e-01 6.65371597e-01 -2.19254103e-02
-6.13359213e-01 4.48700726e-01 -7.65288174e-02 2.84390170e-02
-9.48285401e-01 8.76099408e-01 -2.54404694e-01 8.72068584e-01
3.04965615e-01 -3.78271118e-02 -1.90085873e-01 -4.82782692e-01
-5.04922450e-01 1.23564982e+00 3.89718235e-01 -3.42209339e-01
3.40129845e-02 1.80402309e-01 -2.74019614e-02 -5.39666176e-01
-1.00754786e+00 -8.67609441e-01 -5.78643322e-01 2.81326950e-01
1.20951235e+00 -9.15021837e-01 -1.11430633e+00 -3.63055915e-02
-6.61705077e-01 -2.80482769e-01 -4.98250425e-01 7.17872679e-01
-2.48559549e-01 2.55993217e-01 -5.85738778e-01 -5.72711170e-01
-6.03009760e-01 -9.41384733e-01 1.06144345e+00 -6.64139865e-03
-1.04121566e+00 -1.12755680e+00 2.11905584e-01 2.75026351e-01
1.36812413e-02 7.28850365e-01 1.68359447e+00 -7.80632257e-01
3.02436888e-01 -4.51059133e-01 8.47068653e-02 -2.43098676e-01
7.25367248e-01 -4.60982472e-02 -5.08361161e-01 -1.45056680e-01
3.40229928e-01 3.05897295e-02 2.08188280e-01 5.75307488e-01
1.30968237e+00 -6.89468741e-01 -4.34460670e-01 4.60684568e-01
1.33100867e+00 6.14018023e-01 4.21962440e-01 -2.54112426e-02
5.96317947e-01 5.24166405e-01 4.91120905e-01 1.02228248e+00
5.29196322e-01 2.95215726e-01 7.49531835e-02 -8.85206237e-02
5.53542078e-01 9.31196436e-02 -6.08104318e-02 5.07121146e-01
1.42355844e-01 -3.26197594e-01 -1.38913023e+00 5.40217578e-01
-1.74045074e+00 -6.52830601e-01 -5.11046171e-01 2.28685951e+00
1.04297149e+00 -4.29340541e-01 -2.41529569e-02 1.63091481e-01
7.77724385e-01 -6.17439389e-01 -6.97238147e-01 -2.79853940e-01
2.06957504e-01 1.06020860e-01 5.58047652e-01 1.83808233e-03
-8.66567969e-01 8.22120830e-02 6.46877623e+00 -2.55134940e-01
-5.79689026e-01 -1.52730495e-01 8.74577403e-01 -1.38097629e-02
-1.54467328e-02 -3.24052006e-01 -5.87125719e-01 6.38548613e-01
1.12943327e+00 1.19885979e-02 1.92827582e-01 6.57974839e-01
6.13484085e-01 -2.64933705e-01 -1.56320608e+00 1.00303006e+00
-1.43607318e-01 -1.20710397e+00 -2.76520312e-01 2.92102218e-01
6.72157168e-01 -2.92151839e-01 -3.23512197e-01 -1.36998892e-01
2.78766543e-01 -1.03889418e+00 1.72436938e-01 8.30189168e-01
1.30549133e+00 -2.00173452e-01 8.57444644e-01 4.87432659e-01
-9.79606152e-01 -2.57054836e-01 1.53357312e-01 -6.23565093e-02
2.37068564e-01 6.45094872e-01 -1.12338364e+00 5.60609877e-01
6.79044306e-01 9.24862981e-01 -2.64443904e-01 1.01097071e+00
1.89373270e-01 7.69727230e-01 5.13122641e-02 4.24655914e-01
-2.84193575e-01 -2.43794426e-01 -4.80570197e-02 1.03524578e+00
3.30536664e-01 9.00494039e-01 3.44385475e-01 4.42438662e-01
-1.51321962e-01 2.52961457e-01 -5.51319063e-01 -5.60762286e-01
3.85138988e-01 1.11536777e+00 -5.43842375e-01 -7.89412260e-01
-6.05835557e-01 6.30923033e-01 -9.11532640e-02 1.52608722e-01
-5.09223938e-01 1.09879725e-01 8.27603102e-01 5.41306496e-01
-1.09613329e-01 4.22310859e-01 -5.25070071e-01 -1.01764882e+00
-2.62395054e-01 -1.03214180e+00 1.03353024e+00 -5.11727631e-01
-1.38235056e+00 3.70357931e-01 -3.94084036e-01 -1.25104201e+00
-3.85536641e-01 -3.45285088e-01 -1.79229960e-01 8.18996131e-01
-1.04593384e+00 -4.27912623e-01 -5.27866960e-01 5.83854735e-01
1.83550566e-01 -3.52870673e-02 1.59239805e+00 5.16368806e-01
-8.57340991e-01 1.61410406e-01 -3.20471860e-02 3.38382989e-01
8.71955395e-01 -1.05781138e+00 -2.95756739e-02 -2.20141048e-03
-6.00357652e-01 8.49252284e-01 3.26853842e-01 -1.10405648e+00
-1.26568317e+00 -1.44351602e+00 1.41441739e+00 -5.99337518e-01
2.04236344e-01 1.02941982e-01 -8.63670707e-01 7.95906544e-01
-2.38181919e-01 -2.73852088e-02 1.65491676e+00 -2.64774501e-01
2.66727746e-01 3.96012425e-01 -1.49401593e+00 3.14472675e-01
7.10351050e-01 -4.96578902e-01 -7.40685642e-01 4.53936011e-01
4.18813318e-01 -4.36665356e-01 -1.51159596e+00 5.79690337e-01
4.18336451e-01 -2.94105977e-01 6.62162781e-01 -1.13451469e+00
5.48073888e-01 -1.26156211e-01 -1.83875933e-02 -8.07311416e-01
-2.88250566e-01 -5.15418053e-01 -3.00282508e-01 8.77963483e-01
2.95251667e-01 -4.99782026e-01 6.48689926e-01 1.52448750e+00
4.29493785e-02 -8.19931746e-01 -4.47867215e-01 -1.32353947e-01
-5.67206919e-01 -1.72173455e-01 7.55717099e-01 1.34231102e+00
3.54016036e-01 2.65431285e-01 1.78351235e-02 2.22203538e-01
1.68646082e-01 2.82893449e-01 3.49370450e-01 -1.83997262e+00
-1.27343386e-02 -1.28263026e-01 -1.28423482e-01 -3.70003045e-01
-2.66944110e-01 -7.74423122e-01 -8.97375792e-02 -2.18740368e+00
4.71712738e-01 -5.90046465e-01 -3.63212943e-01 7.57095993e-01
-3.15146357e-01 -9.42340344e-02 -5.72071493e-01 3.59327048e-01
-3.68260115e-01 1.71205886e-02 7.58807898e-01 1.49591953e-01
-3.45756561e-01 2.42890328e-01 -9.05385554e-01 8.10230911e-01
8.21012855e-01 -9.44488645e-01 3.00422683e-02 -1.11320890e-01
3.87207866e-01 7.05175638e-01 3.13712597e-01 -5.96443713e-01
7.77291059e-02 -3.67339492e-01 3.03752720e-01 -5.40786266e-01
-4.96568829e-02 -1.15894186e+00 6.24899328e-01 9.46533680e-01
-5.94781339e-01 3.76936436e-01 9.85548049e-02 5.22904277e-01
-1.62799865e-01 -2.88239807e-01 3.39347959e-01 -3.82519901e-01
-1.64970294e-01 3.92866373e-01 -7.38377392e-01 1.95458189e-01
1.14507782e+00 -2.20389739e-02 -1.01394594e-01 -3.25051919e-02
-1.16066122e+00 2.39558816e-01 6.86174631e-01 1.19700432e-01
4.78755623e-01 -1.18503344e+00 -7.63426781e-01 1.29900739e-01
4.81329888e-01 1.17177159e-01 2.14563116e-01 1.12800276e+00
-6.30262494e-01 4.22470748e-01 1.22578815e-01 -1.77133009e-01
-1.40496922e+00 7.48174131e-01 -8.24686065e-02 -3.07964355e-01
-8.10083330e-01 1.39018878e-01 -6.88377842e-02 -2.82657087e-01
2.99055964e-01 -7.05168009e-01 -4.01381671e-01 1.38116136e-01
6.92707539e-01 2.31051132e-01 2.31196836e-01 -6.82441354e-01
-5.16513109e-01 1.66870534e-01 1.50002584e-01 3.48979294e-01
1.52883494e+00 -8.10389444e-02 -4.23127502e-01 5.77525437e-01
1.21334529e+00 -5.83556779e-02 -4.33664262e-01 -3.13558877e-02
1.61786646e-01 -5.08003905e-02 -3.84922296e-01 -9.02345300e-01
-6.13583565e-01 2.74826527e-01 3.63093019e-01 1.56646028e-01
1.18884063e+00 -3.70276906e-02 6.56716108e-01 2.78197259e-01
-5.01148887e-02 -8.17581177e-01 -7.11027741e-01 2.63614237e-01
3.58092785e-01 -1.15506005e+00 2.24481188e-02 -2.81907737e-01
-8.40155244e-01 8.45208287e-01 1.82370260e-01 4.08771038e-01
8.42947245e-01 4.30510640e-01 1.81871369e-01 -4.76883650e-01
-9.15230751e-01 2.24401802e-01 1.29007474e-01 1.75949186e-01
7.88329363e-01 3.84574950e-01 -4.64511335e-01 9.42223310e-01
-6.15363121e-02 3.02948147e-01 5.16478539e-01 1.24423575e+00
-8.37115478e-03 -1.01942539e+00 -5.16493797e-01 1.31117761e+00
-9.46043968e-01 -4.35864329e-01 -4.46722746e-01 1.26953408e-01
3.02271783e-01 7.88297296e-01 4.34167124e-02 1.35371253e-01
3.51355791e-01 4.54834670e-01 -1.69345569e-02 -1.22105145e+00
-1.03489125e+00 -1.46267796e-02 1.46279395e-01 -3.09586346e-01
-5.41047454e-01 -9.23057675e-01 -1.57615376e+00 3.48697603e-01
6.09045625e-02 3.36922795e-01 4.14044082e-01 8.85620892e-01
5.40196598e-01 7.51121402e-01 1.69848651e-02 2.25683630e-01
-2.41781771e-01 -6.80446565e-01 -5.71010351e-01 6.20153368e-01
4.72458482e-01 -4.47234176e-02 2.91137826e-02 7.22518802e-01] | [8.411185264587402, 8.512163162231445] |
b20a985c-432d-473a-af6b-971228cc9f9f | a-two-stage-approach-towards-generalization | 2111.05825 | null | https://arxiv.org/abs/2111.05825v2 | https://arxiv.org/pdf/2111.05825v2.pdf | A Two-Stage Approach towards Generalization in Knowledge Base Question Answering | Most existing approaches for Knowledge Base Question Answering (KBQA) focus on a specific underlying knowledge base either because of inherent assumptions in the approach, or because evaluating it on a different knowledge base requires non-trivial changes. However, many popular knowledge bases share similarities in their underlying schemas that can be leveraged to facilitate generalization across knowledge bases. To achieve this generalization, we introduce a KBQA framework based on a 2-stage architecture that explicitly separates semantic parsing from the knowledge base interaction, facilitating transfer learning across datasets and knowledge graphs. We show that pretraining on datasets with a different underlying knowledge base can nevertheless provide significant performance gains and reduce sample complexity. Our approach achieves comparable or state-of-the-art performance for LC-QuAD (DBpedia), WebQSP (Freebase), SimpleQuestions (Wikidata) and MetaQA (Wikimovies-KG). | ['Gaetano Rossiello', 'Achille Fokoue', 'Pavan Kapanipathi', 'Tahira Naseem', 'Nandana Mihidukulasooriya', 'Ibrahim Abdelaziz', 'June Thai', 'Srinivas Ravishankar'] | 2021-11-10 | null | null | null | null | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-2.70498872e-01 6.08254433e-01 -3.61436546e-01 -6.06407225e-01
-1.12438428e+00 -9.66079652e-01 3.44427407e-01 4.92351055e-01
-2.70093679e-01 1.04920340e+00 1.99740693e-01 -4.56490964e-01
-4.94899422e-01 -1.43541920e+00 -1.27951181e+00 1.44340396e-01
5.34887984e-02 9.85641062e-01 8.97365868e-01 -7.89671183e-01
-3.75588238e-02 -1.61505360e-02 -1.48377275e+00 8.39130044e-01
1.16446984e+00 1.01050401e+00 -7.83963427e-02 5.18269241e-01
-9.13894236e-01 1.21411669e+00 -5.45703828e-01 -1.18669677e+00
1.47097465e-02 -1.80194110e-01 -1.71430171e+00 -7.53952980e-01
8.21469665e-01 -1.07655391e-01 -9.35112834e-02 1.00297642e+00
1.82577223e-01 3.26668397e-02 4.91165072e-01 -1.26347625e+00
-1.00728762e+00 9.68164086e-01 5.42635098e-02 1.19622417e-01
6.31507099e-01 -1.83828548e-01 1.36243284e+00 -3.08074802e-01
7.14916348e-01 1.23520613e+00 8.70404661e-01 5.14708877e-01
-1.08909559e+00 -4.59809601e-01 2.28249729e-02 6.76571012e-01
-1.14938760e+00 -1.75569713e-01 3.06223154e-01 -1.42198861e-01
1.16390991e+00 1.97697341e-01 3.83778781e-01 6.57552540e-01
-3.62539381e-01 5.26711881e-01 9.56385672e-01 -5.03227174e-01
2.31508702e-01 2.97184139e-01 6.38108134e-01 8.11170518e-01
6.76321566e-01 -4.60684836e-01 -5.43660820e-01 -1.91012055e-01
3.75379026e-01 -5.97241044e-01 -1.29093677e-01 -8.13684642e-01
-8.15150976e-01 7.18253136e-01 7.15213180e-01 1.34066464e-02
-2.77123034e-01 1.85581565e-01 5.50765932e-01 6.67893589e-01
-1.75033808e-01 8.37338805e-01 -9.03466344e-01 -8.48072618e-02
-3.66812915e-01 6.37002110e-01 1.35035348e+00 1.25177372e+00
1.34135401e+00 -4.67602968e-01 8.08013882e-03 5.72604775e-01
1.86044380e-01 5.68502545e-01 2.40763918e-01 -9.64630604e-01
8.24154198e-01 1.00216961e+00 2.55247474e-01 -6.16894305e-01
-3.89501125e-01 -7.53192976e-02 2.30650660e-02 -2.15453640e-01
5.49574733e-01 -9.32053011e-03 -9.91357803e-01 1.62977290e+00
4.49749112e-01 2.93366481e-02 6.50305688e-01 5.98434389e-01
1.28914595e+00 3.47054362e-01 5.00378430e-01 2.86079317e-01
1.57328820e+00 -8.10202599e-01 -5.37261069e-01 -2.65451908e-01
9.47705507e-01 -7.47022703e-02 1.31207895e+00 3.37780043e-02
-1.06355703e+00 -2.94369936e-01 -9.97911155e-01 -5.93650281e-01
-1.05322719e+00 -5.41895926e-01 8.15182924e-01 8.65510821e-01
-1.01975417e+00 1.89658046e-01 -5.02650499e-01 -5.54506063e-01
3.23660284e-01 2.79572368e-01 -3.83134544e-01 -4.05680269e-01
-1.58448637e+00 1.12505829e+00 1.20365608e+00 -4.04764593e-01
-6.81796908e-01 -1.21678531e+00 -8.47456098e-01 1.16739936e-01
8.84219110e-01 -1.25872886e+00 1.38320315e+00 -7.37741053e-01
-1.28876066e+00 7.61155605e-01 1.79207519e-01 -6.94078326e-01
1.22581400e-01 -4.08833802e-01 -6.13119960e-01 1.09096989e-01
1.37011379e-01 4.90405381e-01 3.03096294e-01 -1.28834581e+00
-6.75917149e-01 -5.23954570e-01 1.03541708e+00 3.15931797e-01
-1.70044675e-01 -2.72923291e-01 -7.08693027e-01 -6.78574294e-02
-1.94844455e-01 -5.76433003e-01 1.93818569e-01 -5.20569324e-01
-3.33677262e-01 -2.31035352e-01 4.75192755e-01 -8.29331100e-01
1.17578113e+00 -1.56401312e+00 1.36492521e-01 2.87873358e-01
-5.65451905e-02 3.54736418e-01 -1.48581386e-01 6.79010630e-01
2.53353298e-01 1.97005168e-01 -1.92537278e-01 5.40044606e-01
1.45832971e-01 6.03876412e-01 -4.11304951e-01 -4.48089778e-01
2.52132654e-01 1.26075673e+00 -1.16331041e+00 -5.35618961e-01
-2.66076028e-01 -3.46132405e-02 -7.64203727e-01 1.37160376e-01
-9.22065496e-01 -9.44761187e-02 -6.13661587e-01 5.68250299e-01
5.24205863e-01 -4.87955421e-01 5.68144381e-01 -5.56493402e-01
5.21773636e-01 7.23966122e-01 -1.25518286e+00 2.10997868e+00
-6.11322463e-01 -8.32131729e-02 -1.91431597e-01 -9.45298493e-01
6.44300938e-01 2.57696748e-01 2.41087768e-02 -8.68345797e-01
-3.83511424e-01 3.76862407e-01 -1.69972613e-01 -6.39951229e-01
6.98686481e-01 -1.43879026e-01 -1.69310033e-01 3.76060307e-01
5.16192794e-01 -2.67091185e-01 5.13654113e-01 3.19185257e-01
1.13621056e+00 2.17924595e-01 3.31014454e-01 -2.33845890e-01
4.61608052e-01 5.48108995e-01 2.15723783e-01 9.24865305e-01
2.62704343e-01 3.76899876e-02 4.09605205e-01 -4.99171555e-01
-7.42478848e-01 -1.43400216e+00 5.54594509e-02 1.42496753e+00
2.32340649e-01 -6.27156198e-01 -6.94913328e-01 -1.22084749e+00
3.25826705e-01 9.22002792e-01 -5.56370676e-01 -2.05827236e-01
-5.47157228e-01 -3.91923428e-01 8.67867470e-01 8.92071724e-01
6.28609478e-01 -1.09331632e+00 -3.69080067e-01 1.58027768e-01
-2.97268063e-01 -1.24370956e+00 4.41305131e-01 1.31519511e-01
-9.23931062e-01 -1.40717578e+00 -9.29624364e-02 -5.17685294e-01
2.96660602e-01 -2.77501717e-03 1.95293546e+00 2.55684219e-02
1.58475488e-01 8.02906513e-01 -5.41092873e-01 -6.50815248e-01
-4.96649295e-01 5.57645738e-01 -3.73326749e-01 -6.93535089e-01
6.64885223e-01 -4.06187057e-01 -2.11209089e-01 -1.13613987e-02
-1.04910111e+00 -1.64857283e-01 4.52691436e-01 6.05177462e-01
4.18335438e-01 1.22738779e-02 1.00449181e+00 -1.48237562e+00
6.86603963e-01 -6.78639352e-01 -6.08214021e-01 9.45043921e-01
-5.75101912e-01 3.76415372e-01 5.01128733e-01 2.09532052e-01
-1.35218942e+00 -3.88851434e-01 -1.48880363e-01 -3.79872974e-03
-9.24608782e-02 1.02467394e+00 -3.74192595e-01 -1.02374582e-02
1.02951086e+00 -9.72531587e-02 -2.73236662e-01 -4.85889673e-01
1.11808431e+00 3.17518234e-01 7.23850667e-01 -1.17996025e+00
7.96351373e-01 2.93942124e-01 -1.47219017e-01 -3.19800198e-01
-1.31602991e+00 -3.45822215e-01 -6.28132582e-01 3.50025028e-01
7.60416269e-01 -1.01818001e+00 -7.64302611e-01 -3.32214721e-02
-9.03563201e-01 -4.83603716e-01 -5.53021371e-01 1.34627568e-02
-6.63268447e-01 1.12993568e-01 -4.45765406e-01 -2.91190505e-01
-5.04472017e-01 -4.98691887e-01 8.00956786e-01 2.61416763e-01
-9.07694772e-02 -1.22210777e+00 3.14314187e-01 8.37366581e-01
4.28886920e-01 1.55596377e-03 1.63082612e+00 -9.81531322e-01
-8.34870577e-01 7.21457452e-02 -4.41215426e-01 3.73454064e-01
1.81766316e-01 -5.08549631e-01 -8.52332830e-01 -5.23713268e-02
-7.94552863e-01 -9.03285563e-01 5.65796971e-01 -2.07352892e-01
9.57821190e-01 -3.08917105e-01 -8.88394862e-02 4.39298183e-01
1.65315390e+00 -2.27372572e-02 8.12875807e-01 7.67980278e-01
7.45076656e-01 5.57398260e-01 4.98827040e-01 -1.25844821e-01
1.02398300e+00 5.23679256e-01 4.09532160e-01 3.10228854e-01
-2.29620650e-01 -3.14210832e-01 6.77661747e-02 6.11699462e-01
-9.34014842e-02 3.72976810e-02 -1.26285899e+00 8.03607106e-01
-1.87257767e+00 -8.22797656e-01 9.03791860e-02 2.04914904e+00
1.25970721e+00 8.91878363e-03 2.24147335e-01 -1.73157260e-01
2.53910720e-01 -1.84762418e-01 -6.95512831e-01 -4.07820672e-01
-1.11695498e-01 5.33129513e-01 4.23981547e-01 3.59124780e-01
-7.24191725e-01 1.19710779e+00 6.82483435e+00 5.05291283e-01
-6.63297057e-01 5.62401451e-02 -5.40773235e-02 2.45856732e-01
-5.30438244e-01 5.23377731e-02 -9.58860219e-01 8.52875710e-02
1.05828416e+00 -3.88921499e-01 5.09662330e-01 9.19260442e-01
-9.24759328e-01 1.26879826e-01 -1.28447425e+00 6.11951947e-01
-1.71897903e-01 -1.81113553e+00 6.49662018e-01 -4.48911041e-01
5.52144527e-01 1.57086357e-01 -2.52871424e-01 9.40093338e-01
8.54967177e-01 -8.55815887e-01 4.04326379e-01 5.71853817e-01
6.55207574e-01 -7.55909681e-01 8.08074117e-01 -3.11510991e-02
-9.74550009e-01 -1.25134289e-01 -5.33221841e-01 1.44303292e-01
-1.13904670e-01 1.51254326e-01 -1.09766901e+00 1.32543659e+00
9.81655717e-01 2.57887512e-01 -8.61426234e-01 7.61571229e-01
-4.69857752e-01 5.39723217e-01 -1.50986895e-01 1.07539944e-01
4.13777530e-02 5.18271141e-03 1.21854041e-02 1.11883461e+00
3.67970690e-02 1.74800098e-01 5.48040271e-02 7.33229518e-01
-4.23782021e-01 2.26188049e-01 -4.01849806e-01 -1.26619875e-01
5.86057782e-01 9.90674675e-01 -1.34338215e-01 -6.89728975e-01
-8.05220306e-01 5.19635320e-01 8.74118030e-01 4.29819256e-01
-7.07715273e-01 -4.76474881e-01 7.29978383e-01 1.22882381e-01
5.54879248e-01 2.67094642e-01 4.66602072e-02 -1.25900710e+00
2.96058148e-01 -1.02841020e+00 1.25814402e+00 -9.17553365e-01
-1.45633090e+00 3.54202569e-01 3.40548992e-01 -4.38026041e-01
-4.97097790e-01 -6.88201725e-01 -1.23853512e-01 7.58073986e-01
-2.14333677e+00 -1.51474416e+00 -5.74835598e-01 9.31009769e-01
-1.83362618e-01 6.62961528e-02 1.06061602e+00 2.55010635e-01
-6.69508055e-02 5.42958498e-01 -1.52966529e-01 3.05506915e-01
7.37473309e-01 -1.71198976e+00 4.96519148e-01 4.14969355e-01
1.47177011e-01 7.61992395e-01 4.79443789e-01 -6.34120464e-01
-1.78252435e+00 -1.24654329e+00 5.24461865e-01 -9.86850023e-01
8.57030571e-01 -2.56553348e-02 -1.39076877e+00 1.13142002e+00
2.18021452e-01 2.19527800e-02 9.03115630e-01 8.25963795e-01
-1.10385346e+00 -3.29678059e-01 -9.93629515e-01 4.05863643e-01
1.22855544e+00 -8.19520533e-01 -1.26596951e+00 2.07259893e-01
8.78124595e-01 -4.95035470e-01 -1.39257371e+00 4.94530916e-01
4.55674827e-01 -7.99527586e-01 1.15813148e+00 -1.48904788e+00
2.95702428e-01 -4.06225622e-01 -3.78384501e-01 -1.35053861e+00
-2.66132593e-01 -1.31684572e-01 -6.62204206e-01 1.19383705e+00
6.78773165e-01 -5.04596710e-01 9.30426300e-01 8.03628147e-01
-1.46011472e-01 -3.81498218e-01 -7.14806855e-01 -1.05415857e+00
2.96783417e-01 -3.89442682e-01 1.10882294e+00 1.10436940e+00
2.36402571e-01 6.19265020e-01 2.41287351e-01 6.28649473e-01
3.90479654e-01 4.07430440e-01 1.07643688e+00 -1.29042971e+00
-3.27880621e-01 -6.10377491e-02 -4.20029819e-01 -7.76112080e-01
1.48905069e-01 -1.08102548e+00 -2.92853236e-01 -2.06798983e+00
8.51566568e-02 -7.68830359e-01 -5.10289252e-01 8.41263294e-01
-3.93870831e-01 -4.70157750e-02 -1.85129680e-02 -1.44908130e-01
-1.04304242e+00 1.41381577e-01 1.03222871e+00 -6.18156977e-02
-2.41955672e-03 -5.61967134e-01 -1.05372870e+00 3.76269668e-01
5.38415432e-01 -2.71313488e-01 -9.04685736e-01 -8.09255600e-01
1.00104296e+00 1.92458276e-02 2.93591172e-01 -1.00590646e+00
3.40737224e-01 -1.24327727e-01 -7.00166496e-03 -2.33286798e-01
2.60041833e-01 -7.09951639e-01 8.25203434e-02 1.00899935e-01
-1.86771244e-01 4.09081252e-03 5.61842680e-01 5.04850686e-01
-4.66185510e-01 -4.00951356e-01 3.53587985e-01 -2.77447194e-01
-1.11439061e+00 3.43996435e-02 3.10458511e-01 8.38354707e-01
7.17526317e-01 -4.75574881e-02 -9.61883545e-01 -2.01428026e-01
-4.90392864e-01 3.83016378e-01 3.68742257e-01 4.37243879e-01
1.56351462e-01 -1.13092542e+00 -5.33793151e-01 -2.26426631e-01
6.66093946e-01 2.27960646e-01 2.34717399e-01 3.71966243e-01
-6.44816697e-01 6.19985580e-01 -4.98116165e-01 -3.30520153e-01
-8.15323651e-01 7.23734796e-01 3.65947187e-01 -5.27959466e-01
-2.54287750e-01 7.63961852e-01 1.03117935e-02 -9.37895894e-01
8.65228754e-03 -5.22092044e-01 -1.97016001e-01 8.93163756e-02
5.75040936e-01 3.58831495e-01 5.47894239e-01 -1.37493014e-01
-3.68689001e-01 3.27481657e-01 -3.60520035e-01 2.78228134e-01
1.19709110e+00 1.17890880e-01 -1.24388166e-01 2.32769623e-01
7.24102736e-01 -7.41949901e-02 -5.99435449e-01 -5.28401017e-01
5.42488754e-01 -1.30848020e-01 -2.76875734e-01 -1.34584582e+00
-7.53465414e-01 4.42966521e-01 1.64437562e-01 2.56527007e-01
9.32209611e-01 3.51215899e-01 7.95995533e-01 1.23737836e+00
7.62066901e-01 -9.28918719e-01 -1.33970752e-01 6.33420587e-01
7.58557498e-01 -1.23362863e+00 -1.36456275e-02 -6.73362136e-01
-4.89171565e-01 1.03459179e+00 7.98549771e-01 1.94446757e-01
4.33759987e-01 -1.28055111e-01 2.31608644e-01 -6.00327730e-01
-7.36879647e-01 -4.64775711e-01 4.40173864e-01 8.66662800e-01
2.60354996e-01 -1.10309646e-01 1.50715262e-01 8.53101850e-01
-2.97116995e-01 2.75323570e-01 2.86317527e-01 1.18379796e+00
-4.35031831e-01 -1.18696606e+00 2.00110953e-02 5.09573042e-01
-3.52371633e-01 -3.40652674e-01 -3.09680700e-01 1.21319413e+00
-7.21099153e-02 6.35858119e-01 -5.11410385e-02 -6.08631149e-02
7.62395561e-01 6.36714518e-01 8.34016323e-01 -8.79506171e-01
-7.00767219e-01 -9.98650253e-01 7.26082325e-01 -7.03814089e-01
-4.81897920e-01 -1.83459818e-01 -1.41886234e+00 -1.47582427e-01
-2.76955128e-01 5.29963970e-01 4.28947538e-01 9.40483630e-01
7.05842435e-01 3.63364160e-01 -3.55815321e-01 3.90571803e-01
-4.01173115e-01 -6.89625800e-01 -3.71854782e-01 7.16512084e-01
-1.08156212e-01 -6.02197409e-01 2.64805377e-01 2.66706496e-01] | [10.191878318786621, 7.872884750366211] |
ea94c80c-fe66-4700-9eb0-b3886072d0d6 | identifying-electrocardiogram-abnormalities | 2206.10592 | null | https://arxiv.org/abs/2206.10592v1 | https://arxiv.org/pdf/2206.10592v1.pdf | Identifying Electrocardiogram Abnormalities Using a Handcrafted-Rule-Enhanced Neural Network | A large number of people suffer from life-threatening cardiac abnormalities, and electrocardiogram (ECG) analysis is beneficial to determining whether an individual is at risk of such abnormalities. Automatic ECG classification methods, especially the deep learning based ones, have been proposed to detect cardiac abnormalities using ECG records, showing good potential to improve clinical diagnosis and help early prevention of cardiovascular diseases. However, the predictions of the known neural networks still do not satisfactorily meet the needs of clinicians, and this phenomenon suggests that some information used in clinical diagnosis may not be well captured and utilized by these methods. In this paper, we introduce some rules into convolutional neural networks, which help present clinical knowledge to deep learning based ECG analysis, in order to improve automated ECG diagnosis performance. Specifically, we propose a Handcrafted-Rule-enhanced Neural Network (called HRNN) for ECG classification with standard 12-lead ECG input, which consists of a rule inference module and a deep learning module. Experiments on two large-scale public ECG datasets show that our new approach considerably outperforms existing state-of-the-art methods. Further, our proposed approach not only can improve the diagnosis performance, but also can assist in detecting mislabelled ECG samples. Our codes are available at https://github.com/alwaysbyx/ecg_processing. | ['Jian Wu', 'Danny Z. Chen', 'Xiaoxian Yang', 'Xiaojun Chen', 'Jintai Chen', 'Yuexin Bian'] | 2022-06-16 | null | null | null | null | ['ecg-classification', 'clinical-knowledge'] | ['medical', 'miscellaneous'] | [ 2.24401817e-01 -2.11744353e-01 5.77359693e-03 -5.91813505e-01
-5.60408592e-01 -2.00317904e-01 -3.82548898e-01 4.48722750e-01
-2.29416534e-01 7.45911181e-01 -2.22120568e-01 -6.13136709e-01
-2.73141742e-01 -9.57136869e-01 -2.44442910e-01 -5.63301563e-01
-2.52444237e-01 4.75936979e-01 -1.56107873e-01 7.69741684e-02
-1.24791719e-01 4.10584539e-01 -1.04228258e+00 3.83737862e-01
9.28277194e-01 1.28356874e+00 -3.07375163e-01 7.76346624e-01
3.38580519e-01 9.28130388e-01 -6.38758361e-01 -3.92606348e-01
-3.45688313e-02 -8.85255754e-01 -5.86717784e-01 -1.82844281e-01
-1.70978680e-01 -4.47477371e-01 -3.23018044e-01 8.64882708e-01
1.03893352e+00 -3.55202317e-01 3.44655812e-01 -6.78003311e-01
-2.42387295e-01 6.44080043e-01 -2.78734416e-01 4.73629445e-01
-8.88670906e-02 2.71938771e-01 6.47024989e-01 -4.82159913e-01
1.01802066e-01 5.14172852e-01 9.84514236e-01 4.77421045e-01
-7.49357760e-01 -7.04095244e-01 -3.53559703e-01 4.94142979e-01
-1.41817343e+00 -2.96796709e-01 9.35761392e-01 -3.59787375e-01
4.55857843e-01 3.77930582e-01 9.66078997e-01 8.36027443e-01
2.92395711e-01 5.40368021e-01 6.99642360e-01 -2.15682462e-01
5.21077514e-02 -2.93394566e-01 1.19604178e-01 7.70416856e-01
3.70416552e-01 -3.51732317e-03 -3.71445119e-02 -4.17918116e-01
8.31209362e-01 3.92467499e-01 -4.55262750e-01 1.39025569e-01
-1.20126653e+00 5.25210500e-01 3.93887967e-01 4.87450302e-01
-6.58943713e-01 -1.75851628e-01 7.39550292e-01 2.43558183e-01
2.19516188e-01 3.06803226e-01 -6.98218703e-01 -2.00146720e-01
-7.90113091e-01 1.24631017e-01 5.22008061e-01 2.70692050e-01
3.72861683e-01 7.37767741e-02 -2.58371443e-01 8.65320861e-01
1.02472343e-01 4.07803714e-01 6.52741969e-01 -8.15578282e-01
2.67554492e-01 8.12736869e-01 -1.32435739e-01 -1.18299305e+00
-6.44383550e-01 -1.03203750e+00 -1.59753847e+00 -1.58573821e-01
3.20825636e-01 -4.45525736e-01 -5.76403439e-01 1.26132095e+00
6.60305023e-02 3.50046903e-01 -2.16116384e-02 9.86940324e-01
8.90235543e-01 3.17470700e-01 6.04666658e-02 -4.36049610e-01
1.33682716e+00 -2.50991762e-01 -6.85034811e-01 3.19468677e-02
7.29341447e-01 -3.28845203e-01 5.65054536e-01 7.97199726e-01
-1.03160751e+00 -7.63637841e-01 -1.00667179e+00 3.26236516e-01
6.24938421e-02 4.84350771e-01 4.68071043e-01 6.36755407e-01
-6.24849916e-01 9.52158988e-01 -8.04856777e-01 -1.44492090e-02
8.65266979e-01 3.24021757e-01 4.76644514e-03 -4.81040962e-02
-1.61836231e+00 6.29705966e-01 5.47973335e-01 6.57726705e-01
-6.86209142e-01 -4.77577001e-01 -7.52527058e-01 1.07764497e-01
3.52604717e-01 -7.82052934e-01 1.10152578e+00 -8.49078238e-01
-1.15058863e+00 8.01796436e-01 -8.47847259e-04 -7.03483701e-01
5.55084467e-01 -1.31509185e-01 -6.37960911e-01 3.73897374e-01
-1.33491293e-01 7.58668706e-02 7.73872733e-01 -7.26300061e-01
-6.18995547e-01 -6.09533310e-01 -1.94285437e-01 -8.65523815e-02
-3.31597984e-01 4.87251244e-02 -2.27638423e-01 -6.82419956e-01
1.14309567e-03 -5.88997304e-01 -4.19672757e-01 -1.08371735e-01
-4.19004977e-01 -1.49215147e-01 4.04271007e-01 -9.90816593e-01
1.57754493e+00 -2.02501893e+00 -2.31846988e-01 4.04469699e-01
6.90630496e-01 8.59484553e-01 3.53096277e-01 7.84842074e-02
-8.63204449e-02 2.67099380e-01 -4.86611247e-01 1.37531728e-01
-4.36972082e-01 3.09861779e-01 2.15060711e-02 3.97812486e-01
2.56834269e-01 1.06525874e+00 -7.07810760e-01 -4.76508081e-01
3.16095233e-01 5.64893067e-01 -2.41780147e-01 3.73757750e-01
2.00288683e-01 8.95035863e-01 -5.13482451e-01 5.82000971e-01
4.74298835e-01 -5.41434646e-01 5.04475892e-01 -1.29479066e-01
4.04836714e-01 1.56822622e-01 -1.07525313e+00 1.36889994e+00
4.72692447e-03 1.58760071e-01 -4.25275981e-01 -1.44430673e+00
1.08628273e+00 6.42169595e-01 6.20502591e-01 -5.43019235e-01
3.73796582e-01 3.08281660e-01 5.86252153e-01 -7.33153880e-01
-5.12027204e-01 -1.01695843e-01 1.91094324e-01 2.72278011e-01
-3.66758436e-01 5.00500977e-01 1.71479702e-01 -2.20377117e-01
1.10679376e+00 -2.02017218e-01 6.63470328e-01 1.20744184e-01
8.45903814e-01 -2.88723290e-01 1.24869967e+00 7.56489575e-01
-3.11860979e-01 6.65037930e-01 4.94535238e-01 -1.04246044e+00
-6.46183312e-01 -7.01347172e-01 -2.98480541e-01 3.66782427e-01
-1.65937021e-01 -4.02503908e-01 -7.36584961e-01 -7.07136869e-01
-1.41909197e-01 1.18233711e-01 -4.31089818e-01 -3.00742626e-01
-7.61449695e-01 -1.15457213e+00 1.01750863e+00 9.17486548e-01
6.91662610e-01 -1.21557891e+00 -1.07176292e+00 6.22831047e-01
-3.82444829e-01 -8.41059804e-01 -4.30801436e-02 8.68406370e-02
-1.18275142e+00 -1.46613634e+00 -9.21684742e-01 -6.20372534e-01
4.70927745e-01 -3.21378380e-01 1.12633204e+00 6.18521214e-01
-4.90181923e-01 -1.91176534e-01 -4.87534076e-01 -6.79855287e-01
-4.34651911e-01 3.83265354e-02 -2.34013796e-02 3.08774978e-01
4.24517870e-01 -6.19944930e-01 -1.08566689e+00 7.71799535e-02
-5.87786376e-01 -4.65883873e-02 7.10089505e-01 9.41309094e-01
6.37842953e-01 1.23600818e-01 9.90901411e-01 -1.21778035e+00
6.38504624e-01 -2.72834301e-01 -7.30950013e-02 1.61910430e-01
-8.63464832e-01 -3.86901557e-01 7.91607320e-01 9.12561640e-03
-6.12289369e-01 2.14120839e-02 -7.23204076e-01 -4.41186041e-01
-6.34967923e-01 7.83158720e-01 -1.42795578e-01 3.06124330e-01
6.32228792e-01 2.76238114e-01 -5.77626415e-02 -4.64207113e-01
-3.79549563e-01 8.01960886e-01 5.94829082e-01 -3.00936103e-01
3.69515628e-01 1.93584561e-01 7.34391659e-02 -3.90017509e-01
-7.92863965e-01 -3.73797357e-01 -7.37744808e-01 -1.52447611e-01
8.25195312e-01 -7.77226329e-01 -7.52121508e-01 5.78110397e-01
-1.06264234e+00 -1.13727786e-01 -2.37461090e-01 4.23808873e-01
-2.89540678e-01 5.55234075e-01 -8.63651097e-01 -7.57151067e-01
-8.93487751e-01 -8.01945388e-01 7.13888824e-01 1.65859699e-01
-2.68130243e-01 -9.18275058e-01 -7.29562342e-02 2.22989187e-01
3.46851557e-01 6.02847874e-01 1.15069437e+00 -9.02531505e-01
-2.48993188e-02 -3.14069778e-01 -1.14001840e-01 7.48324096e-01
2.15923324e-01 -3.00843358e-01 -8.97745073e-01 -1.66318491e-01
7.78374076e-02 -1.03730775e-01 7.58603752e-01 5.29943049e-01
1.68195009e+00 -6.07910454e-02 -1.92555740e-01 7.15219438e-01
1.03485429e+00 5.29265165e-01 8.59718859e-01 -3.15055102e-02
7.29310572e-01 8.85068327e-02 2.28205532e-01 6.63792551e-01
3.91111046e-01 2.84801632e-01 3.04190159e-01 -6.11406982e-01
2.47857884e-01 2.55376905e-01 -1.99564189e-01 9.66977894e-01
-6.53463304e-01 3.29933725e-02 -1.28475237e+00 3.69921237e-01
-1.96101522e+00 -7.99281657e-01 -3.96094441e-01 2.04310083e+00
1.01051354e+00 2.12100312e-01 2.41306826e-01 9.27259445e-01
6.83657885e-01 -3.43281835e-01 -7.75004983e-01 -3.03935498e-01
1.04484865e-02 6.05979025e-01 -7.45537579e-02 -1.71537846e-01
-1.21272838e+00 2.01857269e-01 5.42920494e+00 3.32259893e-01
-1.30177963e+00 1.16213329e-01 1.06633222e+00 4.12132204e-01
3.35535645e-01 -4.11379009e-01 -3.18834215e-01 4.76783067e-01
9.24877405e-01 -2.07050610e-02 -4.91548479e-02 6.41584933e-01
4.19788212e-01 4.74306256e-01 -1.00272799e+00 1.30861580e+00
1.20350709e-02 -1.17576253e+00 -1.69744581e-01 -7.92970881e-02
1.90079361e-01 -3.00695926e-01 -2.42094174e-01 2.24567056e-01
-3.90329450e-01 -9.29409325e-01 1.22085944e-01 6.85154498e-01
1.05384076e+00 -9.09620106e-01 1.38046443e+00 3.86370480e-01
-1.03188729e+00 -1.96708873e-01 -2.79532462e-01 -3.29010725e-01
-6.68559000e-02 1.03366125e+00 -6.45823300e-01 7.78449357e-01
7.44210362e-01 1.11889005e+00 -6.34522617e-01 1.18085587e+00
-4.50268120e-01 1.04285920e+00 1.13628894e-01 3.28262568e-01
-2.59496748e-01 1.77491322e-01 3.36054146e-01 1.03101015e+00
2.40224838e-01 5.06279945e-01 3.48118722e-01 7.42925048e-01
-1.57257169e-01 3.48417401e-01 -3.85688752e-01 1.59535438e-01
3.40108834e-02 1.22384858e+00 -7.33421445e-01 -6.62525654e-01
-2.56385595e-01 6.94117486e-01 -4.14769836e-02 1.83411837e-01
-8.53930414e-01 -6.87272549e-01 2.96051830e-01 2.62243032e-01
-1.02717578e-02 3.32148522e-01 -5.92917383e-01 -1.35075581e+00
1.37299687e-01 -1.17397738e+00 6.85783386e-01 -4.36106026e-01
-1.31581211e+00 7.15580285e-01 -4.97804910e-01 -1.41889596e+00
-3.17523748e-01 -4.38395560e-01 -7.66191900e-01 8.41802597e-01
-1.53506684e+00 -6.92348242e-01 -3.46255004e-01 6.15892887e-01
2.81624198e-01 -9.59494114e-02 1.14058101e+00 7.79556215e-01
-6.97334766e-01 6.91613913e-01 -2.21098989e-01 7.75854349e-01
6.54827893e-01 -1.16555691e+00 1.81227610e-01 7.41963327e-01
6.62663020e-03 6.12900317e-01 1.08483486e-01 -5.28352916e-01
-1.00781476e+00 -1.41970038e+00 9.44952011e-01 -2.01852441e-01
1.13513339e-02 2.21668005e-01 -1.16341209e+00 3.42715442e-01
-1.24019109e-01 3.18366289e-01 1.03213644e+00 1.20426551e-01
1.20481253e-01 -4.25367475e-01 -9.80978966e-01 1.93854958e-01
9.01274383e-01 -2.70187140e-01 -6.43537402e-01 1.01748377e-01
-5.00388183e-02 -4.93034452e-01 -1.15730083e+00 1.04404843e+00
7.60633469e-01 -8.36822391e-01 8.67788732e-01 -6.92431152e-01
4.59123403e-01 -1.95162505e-01 3.98732990e-01 -1.19647479e+00
-4.62043226e-01 -3.72714400e-01 -4.03043926e-01 8.58404398e-01
2.86445588e-01 -7.84688413e-01 6.61576629e-01 3.33941013e-01
-6.28393739e-02 -1.18443024e+00 -5.09653151e-01 -3.51778030e-01
-1.22328013e-01 -5.63006103e-01 5.87125063e-01 1.16111779e+00
-3.54256555e-02 2.96661943e-01 -4.15929258e-01 7.83260316e-02
5.73970735e-01 2.16724738e-01 3.91562134e-01 -1.59338808e+00
-4.85197246e-01 -4.16040212e-01 -6.49535179e-01 -5.03919840e-01
-3.75026017e-01 -9.18481767e-01 -2.78085113e-01 -1.62071109e+00
2.35525668e-01 -4.61792111e-01 -9.42570984e-01 8.14075589e-01
-5.66201627e-01 5.88684022e-01 -8.36049467e-02 1.43539146e-01
-4.71100360e-01 8.17010552e-02 1.13447165e+00 -2.19110519e-01
-2.92158633e-01 3.21303397e-01 -8.60581398e-01 1.06932116e+00
1.14195144e+00 -5.38425088e-01 -1.99267089e-01 -2.82051980e-01
3.27338696e-01 3.88936460e-01 4.52388734e-01 -1.34628296e+00
-2.84937061e-02 2.34841526e-01 9.47987139e-01 -4.28286195e-01
-2.68298954e-01 -6.06212914e-01 6.52256384e-02 9.02774334e-01
-4.34635371e-01 -1.02147087e-02 2.54551042e-02 3.06397438e-01
-3.80383998e-01 -1.74638480e-02 5.90188444e-01 -2.42528707e-01
-3.17339450e-01 5.29208243e-01 -4.38922435e-01 1.47578567e-01
7.49659717e-01 6.43725842e-02 1.34381875e-01 -1.97677180e-01
-1.15498042e+00 2.31577247e-01 -2.25508779e-01 9.97551158e-02
9.10239458e-01 -1.20810294e+00 -1.04715455e+00 3.52869660e-01
1.59608707e-01 1.49231970e-01 4.70233023e-01 1.14410865e+00
-7.67932951e-01 1.21423468e-01 -1.42899051e-01 -7.50514209e-01
-1.31298971e+00 4.15171593e-01 6.02225244e-01 -4.02379036e-01
-9.42933738e-01 4.89464819e-01 -2.36440286e-01 -1.15316302e-01
4.34405714e-01 -5.41897476e-01 -4.58450884e-01 -1.43252194e-01
7.86556423e-01 3.42062235e-01 4.37819242e-01 -2.12121204e-01
-4.09207672e-01 5.51565945e-01 -4.71943989e-02 6.16492271e-01
1.29420114e+00 1.29192308e-01 -1.03413031e-01 3.72143418e-01
6.84550464e-01 -4.94168639e-01 -6.20936215e-01 -2.51189560e-01
-2.81963408e-01 -1.22175351e-01 -2.41946895e-02 -1.00084484e+00
-1.36070657e+00 1.49621534e+00 9.19597387e-01 2.84300625e-01
1.58994246e+00 -4.24653351e-01 1.21954620e+00 5.81442773e-01
2.36907989e-01 -7.59533644e-01 -4.19049025e-01 1.17803261e-01
4.46946055e-01 -1.21941984e+00 -1.07736036e-01 -2.61451513e-01
-6.93949938e-01 1.40618646e+00 2.99637437e-01 -6.53051063e-02
8.47838879e-01 1.27053827e-01 3.95870268e-01 -1.36737749e-01
-5.47322810e-01 -8.79652873e-02 1.29805997e-01 4.65711325e-01
4.66672570e-01 2.73688853e-01 -5.55690646e-01 1.13644600e+00
2.98361152e-01 5.78371763e-01 1.66884243e-01 8.65989566e-01
-2.05069661e-01 -1.20883608e+00 -2.15151727e-01 9.05788481e-01
-1.01358962e+00 -8.40091333e-02 -2.95789897e-01 2.31877103e-01
6.22049153e-01 8.94123733e-01 -3.30801576e-01 -5.65460145e-01
3.70856017e-01 4.03966337e-01 3.03280056e-01 -5.17811239e-01
-7.98237860e-01 1.33383617e-01 -1.67685058e-02 -2.66747534e-01
-4.23392087e-01 -4.43411469e-01 -1.25516582e+00 1.32150978e-01
-1.61556769e-02 1.75503939e-01 1.17762692e-01 1.01636040e+00
5.31176746e-01 9.81404483e-01 5.51228940e-01 -2.91988969e-01
-4.68034118e-01 -9.81187940e-01 -5.96796751e-01 4.93430316e-01
2.98148334e-01 -2.24246919e-01 1.09018743e-01 4.48291957e-01] | [14.285202980041504, 3.249361276626587] |
4f845ab5-bab3-49d7-b94d-ef8c488056bf | multilevel-profiling-of-situation-and | 2109.06488 | null | https://arxiv.org/abs/2109.06488v1 | https://arxiv.org/pdf/2109.06488v1.pdf | Multilevel profiling of situation and dialogue-based deep networks for movie genre classification using movie trailers | Automated movie genre classification has emerged as an active and essential area of research and exploration. Short duration movie trailers provide useful insights about the movie as video content consists of the cognitive and the affective level features. Previous approaches were focused upon either cognitive or affective content analysis. In this paper, we propose a novel multi-modality: situation, dialogue, and metadata-based movie genre classification framework that takes both cognition and affect-based features into consideration. A pre-features fusion-based framework that takes into account: situation-based features from a regular snapshot of a trailer that includes nouns and verbs providing the useful affect-based mapping with the corresponding genres, dialogue (speech) based feature from audio, metadata which together provides the relevant information for cognitive and affect based video analysis. We also develop the English movie trailer dataset (EMTD), which contains 2000 Hollywood movie trailers belonging to five popular genres: Action, Romance, Comedy, Horror, and Science Fiction, and perform cross-validation on the standard LMTD-9 dataset for validating the proposed framework. The results demonstrate that the proposed methodology for movie genre classification has performed excellently as depicted by the F1 scores, precision, recall, and area under the precision-recall curves. | ['Aditya Sharma', 'Ayush Mittal', 'Mayank Jindal', 'Dinesh Kumar Vishwakarma'] | 2021-09-14 | null | null | null | null | ['genre-classification'] | ['computer-vision'] | [-1.56163583e-02 -4.67114478e-01 -1.35156035e-01 -3.76384646e-01
-6.85908973e-01 -7.13756859e-01 8.72759759e-01 7.13256896e-01
-3.09348106e-01 3.32214296e-01 7.62387037e-01 4.30957586e-01
-2.99813598e-01 -4.64878500e-01 8.29721838e-02 -6.15758121e-01
-1.74649999e-01 -2.94547677e-01 6.29647374e-02 -4.21371043e-01
6.37053788e-01 3.69272642e-02 -1.91382241e+00 6.40907288e-01
3.25370222e-01 1.53889525e+00 -9.07216668e-02 7.02037990e-01
6.39544576e-02 1.35259843e+00 -5.88809907e-01 -3.56540650e-01
-5.23874238e-02 -5.07681191e-01 -6.52921200e-01 1.28583342e-01
-1.46251485e-01 -7.99436718e-02 -2.83168089e-02 6.34464502e-01
3.43054235e-01 5.56766808e-01 6.62144840e-01 -1.17167938e+00
-2.68368989e-01 3.59189361e-01 -2.90406764e-01 4.94909227e-01
8.98582876e-01 -1.23395719e-01 8.02669466e-01 -5.92819929e-01
8.92598808e-01 1.15929282e+00 5.13098061e-01 6.05710968e-02
-4.68111604e-01 -3.81525636e-01 -1.25196837e-02 6.76630914e-01
-1.07419991e+00 -3.67031723e-01 9.95443106e-01 -7.71092713e-01
8.73013556e-01 3.49002391e-01 9.99834359e-01 1.15480721e+00
5.08879781e-01 5.23263276e-01 1.29497790e+00 -4.06272262e-01
2.48891324e-01 4.73801136e-01 3.82909596e-01 1.81940556e-01
-5.76590419e-01 -3.25584203e-01 -8.64070773e-01 1.86586175e-02
2.44388297e-01 -6.05448261e-02 -9.04138945e-03 2.00999647e-01
-1.04973471e+00 7.58869350e-01 -2.78676003e-01 5.79879940e-01
-7.23122001e-01 -3.94858241e-01 1.04458070e+00 6.19304061e-01
7.32661307e-01 2.08500251e-01 -1.10184349e-01 -9.88234878e-01
-7.91881859e-01 3.33019286e-01 6.36095464e-01 5.40310085e-01
3.42410535e-01 -5.97012565e-02 -2.87940562e-01 9.59795952e-01
1.98952049e-01 2.13318050e-01 6.34204030e-01 -8.95740032e-01
1.54855520e-01 7.27955103e-01 -4.45383787e-02 -1.57244134e+00
-5.17992198e-01 -4.08163033e-02 -4.50929731e-01 -1.21102795e-01
-1.75936386e-01 -3.91199514e-02 -2.11529315e-01 1.55298746e+00
4.38484281e-01 -1.80838734e-01 1.73946366e-01 9.04513299e-01
1.44750297e+00 8.33406985e-01 1.64109126e-01 -9.25247848e-01
1.59082758e+00 -6.07766449e-01 -1.16352224e+00 3.20084691e-01
2.56143510e-01 -9.84097779e-01 1.06417084e+00 7.93158591e-01
-1.05101788e+00 -7.38692582e-01 -1.16733444e+00 2.64119923e-01
-4.53105032e-01 1.89211309e-01 3.52141231e-01 4.60709900e-01
-6.67102814e-01 5.03734469e-01 -2.41495654e-01 -6.96040273e-01
-3.39408033e-02 1.59221534e-02 -6.44610286e-01 2.82777876e-01
-1.15100467e+00 8.22977364e-01 3.38288069e-01 -3.53421658e-01
-8.07422638e-01 -3.06472927e-01 -6.36566699e-01 -3.06590676e-01
1.80601969e-01 2.35502962e-02 1.02497327e+00 -1.46458268e+00
-1.65415525e+00 7.82373726e-01 1.26350313e-01 -1.02284424e-01
-1.85400471e-01 -7.74344057e-02 -9.15412366e-01 6.99835718e-01
8.13708454e-02 2.06328198e-01 7.29428887e-01 -7.72828698e-01
-6.20998919e-01 -5.48256278e-01 4.28873241e-01 6.59829736e-01
-8.26226890e-01 6.63184822e-01 -3.32607657e-01 -8.49735558e-01
-9.75365415e-02 -6.27336323e-01 4.33355033e-01 -5.53296387e-01
1.28922909e-01 -2.51355737e-01 8.63029242e-01 -9.38873112e-01
1.71574306e+00 -2.43101668e+00 3.51758868e-01 2.16491856e-02
2.55590808e-02 -2.08929524e-01 2.94958681e-01 9.30100501e-01
1.03059486e-01 -3.62113833e-01 2.84904093e-01 -1.49086148e-01
-2.35480934e-01 -1.44121855e-01 1.74096525e-01 4.12918538e-01
-2.86608845e-01 1.72480091e-01 -8.23445678e-01 -7.25333571e-01
2.08251268e-01 3.08957666e-01 -2.09416598e-01 2.27108702e-01
8.34721923e-02 4.18248296e-01 -4.52055216e-01 7.06894338e-01
-2.07112692e-02 3.44240159e-01 -1.88158005e-02 -4.99954760e-01
-4.91712630e-01 -3.77061102e-03 -9.46321249e-01 1.68997920e+00
-4.26967978e-01 9.06190217e-01 1.05313905e-01 -8.44760835e-01
1.06454027e+00 7.47473300e-01 6.72335804e-01 -6.80224955e-01
6.00249231e-01 -2.22298518e-01 -2.96898484e-01 -1.15349984e+00
6.83341384e-01 -2.67885357e-01 -3.48392218e-01 3.80237192e-01
3.24604630e-01 8.91051143e-02 3.50397885e-01 1.76810369e-01
9.41883862e-01 2.09664941e-01 5.95532298e-01 -8.43658298e-02
7.12299168e-01 1.33804888e-01 3.81708622e-01 1.21842166e-02
-2.80108958e-01 3.49857062e-01 5.10709882e-01 -2.43200287e-01
-7.05688715e-01 -4.90944594e-01 4.02005911e-02 1.22486842e+00
1.23020470e-01 -8.14211011e-01 -7.27622807e-01 -4.07046646e-01
-4.55450237e-01 5.34672201e-01 -8.18981528e-01 -1.90802783e-01
-7.82268643e-02 -5.97152650e-01 3.84597182e-01 2.19272226e-01
6.27591491e-01 -1.16521180e+00 -6.98181450e-01 1.58188134e-01
-3.82353991e-01 -9.76346731e-01 -6.90843153e-05 -8.73429105e-02
-5.31202972e-01 -1.15555620e+00 -2.11361974e-01 -6.25280380e-01
-9.01949480e-02 -2.71375217e-02 7.58386075e-01 -5.81882179e-01
-6.84015602e-02 7.15014100e-01 -1.00389338e+00 -1.93390980e-01
-4.89818156e-01 -4.15823340e-01 1.84497178e-01 4.41264421e-01
4.17097539e-01 -7.14014828e-01 -3.36837262e-01 1.92573085e-01
-6.35435820e-01 5.89865260e-03 2.08935708e-01 3.58068019e-01
3.96313846e-01 1.73807770e-01 7.73754597e-01 -4.17231530e-01
8.77838373e-01 -9.14859474e-01 2.93581069e-01 -6.35637939e-02
-3.78512591e-01 -6.85697198e-01 6.98335707e-01 -4.26014811e-01
-1.25661230e+00 -3.29065681e-01 -1.39566898e-01 -4.17094350e-01
-3.33287477e-01 7.10405469e-01 -1.15302600e-01 1.54721841e-01
5.66125572e-01 1.80431500e-01 5.26178069e-02 -5.42028129e-01
6.43186793e-02 1.29064453e+00 5.86910069e-01 -4.34591919e-01
-1.15251150e-02 4.35271233e-01 -1.36664644e-01 -1.01484358e+00
-6.84456348e-01 -8.30654681e-01 -4.73604888e-01 -1.28442550e+00
1.02514791e+00 -9.22004998e-01 -9.52911735e-01 3.86827379e-01
-8.19578171e-01 2.86660016e-01 -1.45955086e-01 6.98369920e-01
-6.87757790e-01 4.66060877e-01 -6.93915248e-01 -9.54660237e-01
-4.58030671e-01 -8.53799403e-01 7.15564489e-01 1.90565720e-01
-5.97131371e-01 -8.02253067e-01 1.73833489e-01 5.02645016e-01
1.16603673e-01 9.21044290e-01 8.02491426e-01 -6.05459213e-01
4.33852851e-01 -3.15090626e-01 1.87920749e-01 2.82920569e-01
1.33693114e-01 -2.44690031e-02 -7.55302906e-01 3.54318358e-02
2.80643582e-01 -5.95862031e-01 4.96777624e-01 2.03413591e-01
6.48107231e-01 -3.50242913e-01 1.51764214e-01 6.57560006e-02
1.34858346e+00 6.45024359e-01 3.47658873e-01 4.31370288e-01
3.78263593e-01 6.88825071e-01 1.25258470e+00 1.04335165e+00
3.95529151e-01 9.03220773e-01 3.18332791e-01 5.19062161e-01
8.45394060e-02 -1.74901057e-02 7.00638354e-01 1.17308688e+00
-4.59438741e-01 -1.98732316e-01 -5.14073431e-01 3.67298663e-01
-1.97866547e+00 -1.41211247e+00 -2.12645948e-01 1.96437252e+00
6.20293796e-01 3.92684713e-02 5.71738660e-01 7.14707553e-01
3.79548997e-01 2.49191821e-01 -9.01662279e-03 -8.28592598e-01
-1.76623285e-01 -1.62962601e-01 -2.10334539e-01 1.50674656e-01
-1.31068099e+00 5.52613199e-01 5.25828934e+00 1.10708225e+00
-9.44501996e-01 3.94147307e-01 3.51527125e-01 -4.50779974e-01
-1.80388596e-02 -3.79062831e-01 -2.68507779e-01 5.27784586e-01
1.12321627e+00 -1.30417258e-01 4.06192064e-01 7.68906772e-01
8.26783061e-01 -4.92406726e-01 -7.90996730e-01 1.09782851e+00
7.21110761e-01 -1.03052640e+00 -1.56150401e-01 -3.26167345e-01
3.93099576e-01 -6.10144615e-01 -1.68585166e-01 3.02189827e-01
-5.96796393e-01 -4.56613183e-01 9.94351745e-01 6.76094055e-01
7.89182425e-01 -8.70950401e-01 7.48503923e-01 1.90689236e-01
-1.18603003e+00 -2.06943586e-01 1.13101892e-01 -5.00299990e-01
1.48608565e-01 1.08058609e-01 -2.41765797e-01 5.98325074e-01
8.31561804e-01 1.26136029e+00 -4.74895626e-01 5.40482223e-01
3.62119883e-01 6.88467741e-01 6.02946943e-03 -2.83538043e-01
1.65401921e-01 -1.93531007e-01 7.44213462e-01 1.29823589e+00
2.35996306e-01 4.10910070e-01 4.31797057e-02 1.65183753e-01
4.06116217e-01 8.10999513e-01 -5.24610698e-01 -1.81794226e-01
3.16812307e-01 1.57975173e+00 -9.98746276e-01 -3.16094548e-01
-3.89628887e-01 6.56123936e-01 -1.24333635e-01 -1.20213687e-01
-7.98330903e-01 -2.41674423e-01 3.51589352e-01 1.16850749e-01
-4.15248200e-02 8.87178034e-02 5.98616153e-02 -8.93568695e-01
-9.65970457e-02 -1.00946593e+00 5.30642271e-01 -7.72589028e-01
-9.24745679e-01 8.51031542e-01 3.21146786e-01 -1.60307777e+00
-2.33261138e-01 -2.95401454e-01 -4.88615811e-01 8.97050053e-02
-5.21799684e-01 -1.17729104e+00 -5.57945430e-01 7.05291688e-01
9.47336853e-01 -4.35952395e-01 8.52385283e-01 5.22417247e-01
-4.70027626e-01 2.84521103e-01 -6.52067177e-03 -1.73541591e-01
6.67329550e-01 -8.94655406e-01 -7.38833249e-01 4.18891042e-01
-7.07581118e-02 1.95141122e-01 8.33784819e-01 -4.63703126e-01
-1.44478512e+00 -6.66434586e-01 7.15063572e-01 -2.28805199e-01
5.56923270e-01 2.30091829e-02 -2.03374609e-01 1.86611041e-01
3.72524768e-01 -6.91138983e-01 1.07862961e+00 5.14674634e-02
8.30793288e-03 -1.71578571e-01 -1.16661453e+00 1.81770667e-01
5.77296793e-01 -4.40331489e-01 -6.02276444e-01 1.96375296e-01
3.43846470e-01 -8.67169127e-02 -1.29045236e+00 1.92808598e-01
9.20141518e-01 -1.17173576e+00 6.07349455e-01 -4.21796173e-01
7.67765224e-01 -2.23695666e-01 -5.28295040e-01 -1.03123689e+00
-1.72482222e-01 -3.54462773e-01 -8.75942316e-03 1.68408215e+00
-6.03726925e-03 1.90890908e-01 1.84924468e-01 9.27312374e-02
-2.54372448e-01 -7.64702559e-01 -8.28682959e-01 -2.62063295e-01
-4.90348130e-01 -7.75571823e-01 1.00964844e-01 1.10173929e+00
7.72930264e-01 6.76422119e-01 -6.07657552e-01 -5.00819147e-01
1.46866828e-01 1.32944286e-01 3.88449043e-01 -1.18123698e+00
-2.29187235e-01 -3.69172543e-01 -8.31530869e-01 -2.30339125e-01
-4.64594103e-02 -5.54838181e-01 -3.00448805e-01 -1.30933273e+00
5.04478276e-01 1.27942026e-01 -2.34333947e-01 5.16948923e-02
2.23209843e-01 6.09890521e-01 2.89196402e-01 1.91144392e-01
-9.52369690e-01 4.14913893e-01 8.91994536e-01 9.31176096e-02
-2.42650896e-01 7.49129849e-03 -3.05304199e-01 9.27313983e-01
5.92455208e-01 -9.22133848e-02 -5.15543103e-01 3.02186400e-01
4.19791788e-01 5.41337907e-01 6.89112842e-02 -1.16598606e+00
-1.99142724e-01 -9.92971584e-02 1.31934106e-01 -5.50675929e-01
7.64996827e-01 -6.24079645e-01 3.71175826e-01 3.81802991e-02
-4.67399895e-01 1.87303811e-01 1.00465454e-01 4.93233442e-01
-6.02925301e-01 -2.07110215e-02 5.83082378e-01 -4.20144983e-02
-1.01016831e+00 -1.60563648e-01 -8.63470554e-01 -9.94318202e-02
1.24389791e+00 -5.04248679e-01 -7.51993358e-02 -8.03881466e-01
-1.02629316e+00 -2.47368887e-01 2.87145674e-01 7.41894186e-01
5.16985893e-01 -1.35187352e+00 -6.10627532e-01 -1.78917691e-01
3.02348137e-01 -8.96639884e-01 6.23162329e-01 1.06942844e+00
-3.16096365e-01 7.52904788e-02 -7.09155321e-01 -2.94368744e-01
-1.76857781e+00 5.40991724e-01 -1.97505787e-01 5.96736185e-02
-1.91369250e-01 4.63093758e-01 -1.64884120e-01 2.39673093e-01
1.26814187e-01 -1.00400764e-02 -1.28650308e+00 9.21577275e-01
4.94113237e-01 7.65558481e-01 1.13504035e-02 -1.26617873e+00
-3.71720076e-01 6.34605825e-01 3.72314245e-01 -4.09323871e-01
1.21727133e+00 -4.69724506e-01 -1.88318342e-01 1.07126319e+00
1.16219366e+00 -9.37625170e-02 -5.39011598e-01 1.10125333e-01
-1.69840097e-01 -2.82202631e-01 3.49550277e-01 -9.12850201e-01
-7.08793521e-01 7.21670628e-01 5.76316714e-01 5.35312831e-01
1.54039776e+00 -1.11899711e-01 5.98682761e-01 -9.92480516e-02
2.71010518e-01 -1.45703208e+00 2.23586306e-01 2.38383621e-01
9.83324230e-01 -9.57402229e-01 1.06411733e-01 -2.46461749e-01
-9.75347400e-01 1.26327932e+00 5.04026771e-01 9.66834128e-02
8.69876266e-01 -3.16195674e-02 -5.18586859e-02 -3.58815312e-01
-9.15307522e-01 -7.59944841e-02 4.06294495e-01 3.04778636e-01
6.45381808e-01 7.28077814e-02 -8.99578035e-01 1.44577205e+00
-2.76097178e-01 2.65742224e-02 4.82127011e-01 8.42290401e-01
-5.34600317e-01 -7.49220490e-01 -3.40011567e-01 4.37645912e-01
-7.81634629e-01 3.22979003e-01 -5.30533969e-01 5.62251985e-01
4.69725877e-01 1.29017007e+00 1.12090423e-03 -9.92312372e-01
2.05046922e-01 -6.46743625e-02 5.02353787e-01 -3.77010733e-01
-1.16781211e+00 1.60170391e-01 6.23531580e-01 -5.32077730e-01
-8.82403374e-01 -8.86208892e-01 -8.51203740e-01 -3.29091936e-01
1.01414643e-01 3.23646158e-01 6.62891090e-01 1.05280793e+00
2.05537438e-01 4.49836880e-01 9.49528813e-01 -9.16092038e-01
5.73759973e-02 -1.35753310e+00 -6.14366114e-01 5.54493546e-01
9.96529823e-04 -7.92330503e-01 -2.67795563e-01 3.27274740e-01] | [15.231179237365723, 4.871703624725342] |
bc7df4d6-33c0-4303-9b6e-f82698b56faa | human-like-controllable-image-captioning-with | 2103.12204 | null | https://arxiv.org/abs/2103.12204v1 | https://arxiv.org/pdf/2103.12204v1.pdf | Human-like Controllable Image Captioning with Verb-specific Semantic Roles | Controllable Image Captioning (CIC) -- generating image descriptions following designated control signals -- has received unprecedented attention over the last few years. To emulate the human ability in controlling caption generation, current CIC studies focus exclusively on control signals concerning objective properties, such as contents of interest or descriptive patterns. However, we argue that almost all existing objective control signals have overlooked two indispensable characteristics of an ideal control signal: 1) Event-compatible: all visual contents referred to in a single sentence should be compatible with the described activity. 2) Sample-suitable: the control signals should be suitable for a specific image sample. To this end, we propose a new control signal for CIC: Verb-specific Semantic Roles (VSR). VSR consists of a verb and some semantic roles, which represents a targeted activity and the roles of entities involved in this activity. Given a designated VSR, we first train a grounded semantic role labeling (GSRL) model to identify and ground all entities for each role. Then, we propose a semantic structure planner (SSP) to learn human-like descriptive semantic structures. Lastly, we use a role-shift captioning model to generate the captions. Extensive experiments and ablations demonstrate that our framework can achieve better controllability than several strong baselines on two challenging CIC benchmarks. Besides, we can generate multi-level diverse captions easily. The code is available at: https://github.com/mad-red/VSR-guided-CIC. | ['Wei Liu', 'Jun Xiao', 'Zhihong Jiang', 'Long Chen'] | 2021-03-22 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Chen_Human-Like_Controllable_Image_Captioning_With_Verb-Specific_Semantic_Roles_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Chen_Human-Like_Controllable_Image_Captioning_With_Verb-Specific_Semantic_Roles_CVPR_2021_paper.pdf | cvpr-2021-1 | ['controllable-image-captioning'] | ['computer-vision'] | [ 6.47399724e-01 4.38242912e-01 -3.22164059e-01 -5.64101577e-01
-9.04765129e-01 -7.67388105e-01 9.09313560e-01 -1.39459074e-01
-2.01293543e-01 8.31860423e-01 5.96087098e-01 4.06142846e-02
2.32915014e-01 -6.81959987e-01 -1.12993073e+00 -6.50672793e-01
4.12706316e-01 4.43234235e-01 1.97603256e-01 -3.12905937e-01
4.15544286e-02 1.73653573e-01 -1.43762183e+00 6.08982205e-01
9.50305700e-01 7.70592630e-01 6.03071332e-01 1.96217448e-01
3.66874672e-02 9.81373966e-01 -7.30153203e-01 -1.53347135e-01
-5.21475822e-02 -7.73215234e-01 -8.29072356e-01 4.12078619e-01
1.28165424e-01 -4.44604680e-02 -2.90193349e-01 1.07140398e+00
4.27581191e-01 2.45365292e-01 5.97898602e-01 -1.67848587e+00
-1.10228062e+00 7.30857372e-01 -3.11569273e-01 -1.20054595e-01
7.71437705e-01 6.05697453e-01 1.20471907e+00 -6.78026915e-01
1.03676331e+00 1.43245113e+00 -1.65191457e-01 1.03845680e+00
-1.37630200e+00 -5.06566346e-01 5.76843977e-01 5.25896102e-02
-1.01559520e+00 -4.29435641e-01 7.95923829e-01 -2.65071452e-01
5.01712739e-01 3.52070212e-01 6.96241081e-01 1.64834464e+00
-4.42530990e-01 1.31910920e+00 9.14636612e-01 -2.79905647e-01
2.78030545e-01 8.58614668e-02 -1.74660414e-01 4.55624729e-01
7.07681626e-02 -2.43967876e-01 -4.78440136e-01 3.07263315e-01
7.91138887e-01 -4.03686434e-01 -6.56171203e-01 -4.57278550e-01
-1.75778854e+00 7.46405184e-01 4.62060988e-01 3.16049069e-01
-3.37740481e-01 5.24721503e-01 2.71841019e-01 -8.45675766e-02
9.05903876e-02 9.70842123e-01 -3.00649136e-01 5.39377518e-02
-4.18878138e-01 3.54487121e-01 4.05220479e-01 1.47101390e+00
6.04416728e-01 -7.68189318e-04 -1.07053459e+00 6.79564953e-01
4.96596657e-02 6.32467091e-01 2.29880095e-01 -1.19669962e+00
6.10392392e-01 6.89569235e-01 4.23241645e-01 -7.90970504e-01
-1.47751644e-02 -9.21486095e-02 -7.53327429e-01 -3.68879348e-01
1.58741146e-01 4.86160070e-02 -1.05037129e+00 2.23114657e+00
1.59702778e-01 1.44392416e-01 2.39354715e-01 1.20502210e+00
9.78742898e-01 1.05515623e+00 4.52608913e-01 -6.94841519e-02
1.68383849e+00 -1.13051808e+00 -8.54835391e-01 -4.90773290e-01
3.03969234e-01 -2.68234968e-01 1.74421036e+00 -2.94199605e-02
-1.08992743e+00 -4.38457042e-01 -8.12198222e-01 -8.75112340e-02
-1.26579985e-01 2.44831711e-01 5.99614501e-01 6.72153607e-02
-1.00529635e+00 1.23136071e-03 -2.82646835e-01 -2.65689492e-01
4.86356020e-01 3.69599909e-02 -4.18729842e-01 -4.95269783e-02
-1.48281181e+00 4.84576494e-01 5.21385789e-01 -3.89328972e-02
-1.38149881e+00 -6.95414543e-01 -1.24220240e+00 1.04968354e-01
6.17821932e-01 -1.01645613e+00 1.42272329e+00 -1.34428525e+00
-1.15390420e+00 1.06323421e+00 -3.30299675e-01 -2.43272603e-01
3.84983838e-01 -9.05283242e-02 -1.91948339e-01 5.20347476e-01
6.18433714e-01 1.39006925e+00 8.22783530e-01 -1.80483162e+00
-5.01851439e-01 1.40068680e-01 5.79914331e-01 4.94629711e-01
-9.42413956e-02 1.48587883e-01 -9.14439797e-01 -8.65946174e-01
-2.28699848e-01 -8.94554853e-01 -2.12034151e-01 -7.16042146e-02
-8.12035322e-01 -3.12258869e-01 3.20740342e-01 -2.90417969e-01
1.09715354e+00 -2.18227029e+00 4.27106977e-01 -1.90954164e-01
-1.03944410e-02 2.54878365e-02 -5.78908741e-01 3.15736145e-01
-2.38292500e-01 4.52775329e-01 -4.69521105e-01 -3.52430552e-01
1.32100731e-01 2.56714761e-01 -4.92077112e-01 5.19065931e-02
5.52305400e-01 1.23959124e+00 -1.34494603e+00 -5.63475192e-01
9.64617878e-02 2.97796071e-01 -4.60878789e-01 4.53712970e-01
-9.12299931e-01 6.89832389e-01 -7.19480336e-01 7.01162338e-01
3.54503483e-01 -5.19354463e-01 1.83560967e-01 -4.39866543e-01
2.29417980e-01 1.19912013e-01 -8.98714125e-01 1.71172023e+00
-4.44238931e-01 3.60425830e-01 -3.88555638e-02 -8.10975492e-01
6.75028980e-01 2.44747862e-01 2.90547699e-01 -8.01656604e-01
-8.00005272e-02 -2.82277688e-02 -2.35844359e-01 -4.88365471e-01
4.19979542e-01 1.31160274e-01 -3.81985813e-01 1.97214156e-01
-1.24793686e-01 -3.85210127e-01 4.76723313e-01 4.83862579e-01
8.29229891e-01 5.05448699e-01 2.25094393e-01 -1.02585040e-01
6.75944030e-01 1.88167527e-01 6.76810145e-01 7.12444901e-01
-2.67667890e-01 1.07381475e+00 9.61505234e-01 -9.44617987e-02
-9.23895299e-01 -1.07750309e+00 3.63288850e-01 9.68123019e-01
6.83728576e-01 -3.04812968e-01 -8.01114857e-01 -7.36852884e-01
-2.85180539e-01 1.12562513e+00 -7.76154220e-01 -1.04753360e-01
-6.79311156e-01 -2.71977246e-01 4.05408233e-01 5.31444311e-01
6.17679536e-01 -1.80770087e+00 -6.20699704e-01 7.41446391e-02
-6.41187787e-01 -1.49438059e+00 -9.54132378e-01 -1.13780737e-01
-2.61024982e-01 -1.03738129e+00 -7.03364909e-01 -1.12263155e+00
1.11357415e+00 2.97612667e-01 1.43506789e+00 -1.21302474e-02
1.37675464e-01 6.18461967e-01 -5.40012419e-01 -3.23585659e-01
-4.29538876e-01 -1.37379125e-01 -3.01220447e-01 3.23754072e-01
-4.99581844e-02 -1.40451729e-01 -7.11942196e-01 4.19231057e-01
-1.07102036e+00 6.58782482e-01 5.85700095e-01 6.47935331e-01
9.09665883e-01 -3.59878987e-01 6.27644897e-01 -8.46875489e-01
7.61139512e-01 -3.96803856e-01 -3.64291817e-01 5.49581826e-01
-3.68466936e-02 2.33063638e-01 7.58914590e-01 -5.25785446e-01
-1.13647103e+00 2.66136706e-01 1.95142820e-01 -5.20423472e-01
-3.52982640e-01 1.87907711e-01 -7.54502952e-01 7.06266999e-01
5.63282490e-01 6.57715440e-01 -5.40220588e-02 2.19251662e-01
7.36423910e-01 3.02568585e-01 6.58597946e-01 -1.06960428e+00
7.77798712e-01 4.97321695e-01 -2.23427907e-01 -4.25719231e-01
-1.08392417e+00 -3.18798125e-01 -1.42796189e-01 -3.12723786e-01
1.33573210e+00 -1.11811745e+00 -6.01726234e-01 3.02295804e-01
-1.43258560e+00 -6.18595541e-01 -2.79438555e-01 6.06670678e-02
-9.22701478e-01 5.78207150e-02 -2.86762595e-01 -5.55694580e-01
-3.34157884e-01 -1.44166899e+00 1.53288925e+00 3.15409899e-01
-2.13099241e-01 -7.18813539e-01 -4.39153880e-01 4.34615791e-01
9.51345637e-02 5.06775022e-01 7.52920508e-01 -3.57505798e-01
-8.38945091e-01 2.18599990e-01 -3.46777469e-01 2.30398372e-01
8.88102502e-02 -1.45950869e-01 -6.50166690e-01 -1.37563189e-02
-3.93133968e-01 -5.62003016e-01 6.49733245e-01 4.06978518e-01
1.39951527e+00 -4.27466154e-01 -3.91507417e-01 4.59888011e-01
1.24344409e+00 4.16533232e-01 9.47000086e-01 4.04913753e-01
7.36497223e-01 7.53740489e-01 9.58851278e-01 8.42388347e-02
5.39463460e-01 7.63561606e-01 5.66529512e-01 -2.59266257e-01
-2.70621300e-01 -7.58810639e-01 4.50940877e-01 6.42392263e-02
5.44965789e-02 -5.98929107e-01 -6.88872397e-01 7.05871403e-01
-2.10573959e+00 -1.14840889e+00 -4.85962816e-02 1.74422705e+00
1.07213509e+00 4.93001118e-02 -2.15918608e-02 -3.29719901e-01
1.10955191e+00 3.72339875e-01 -6.63725317e-01 8.79030973e-02
-3.12552571e-01 -1.78965062e-01 1.36904880e-01 2.89515764e-01
-1.09754324e+00 1.26348317e+00 4.80865097e+00 8.06212068e-01
-7.94435501e-01 5.00847474e-02 8.56939554e-01 5.84440604e-02
-8.94496858e-01 -4.69142746e-04 -7.25580990e-01 6.63637280e-01
3.16338390e-01 -2.83829391e-01 3.94222647e-01 8.57363403e-01
5.98850906e-01 -1.01785585e-02 -1.20038223e+00 1.01696658e+00
1.52306274e-01 -1.42342234e+00 6.45281315e-01 -2.66452938e-01
6.33861780e-01 -5.71686864e-01 1.11897394e-01 3.29615653e-01
2.34964252e-01 -9.85011637e-01 1.31940806e+00 2.97593206e-01
1.19229043e+00 -3.66878211e-01 2.84803897e-01 9.73672941e-02
-1.20079660e+00 -6.66979998e-02 -6.92972764e-02 2.88219184e-01
4.30798709e-01 2.55940467e-01 -5.46766579e-01 5.15018284e-01
4.08972293e-01 7.80269623e-01 -4.26920742e-01 5.73406637e-01
-1.05354643e+00 5.52812874e-01 1.75138727e-01 -2.31244028e-01
4.06796336e-01 -1.87553465e-01 6.46221399e-01 9.72117424e-01
9.93430763e-02 1.75407350e-01 3.07623111e-02 1.25271952e+00
-2.52347738e-01 -5.59472991e-03 -5.98985672e-01 -1.47860214e-01
6.79555058e-01 1.16285014e+00 -7.36590445e-01 -3.84014875e-01
-2.48099014e-01 1.20171273e+00 1.39693022e-01 4.97211695e-01
-1.24496329e+00 -2.33547062e-01 6.67484462e-01 2.43228048e-01
9.96218622e-02 1.08563855e-01 -9.08183455e-02 -1.32470000e+00
2.04527915e-01 -1.13986957e+00 3.24301481e-01 -1.49840486e+00
-9.65018332e-01 6.53342128e-01 3.20083648e-01 -1.31330168e+00
-6.42663315e-02 -3.52481842e-01 -5.97525775e-01 6.43712699e-01
-1.53829658e+00 -1.39608490e+00 -5.06193817e-01 5.92122734e-01
9.81233001e-01 8.14569741e-02 5.46333730e-01 2.80522779e-02
-5.41946769e-01 3.57280761e-01 -9.00498807e-01 1.16535857e-01
5.89128852e-01 -1.41185737e+00 5.19461095e-01 9.69144940e-01
9.64852516e-03 6.01446748e-01 7.06798971e-01 -6.10169828e-01
-1.19751179e+00 -1.34613955e+00 8.12319994e-01 -6.34809554e-01
4.34873998e-01 -6.70262814e-01 -7.03487277e-01 6.13653302e-01
3.43582630e-01 -1.71578780e-03 1.43595368e-01 -5.32877266e-01
-3.04586887e-01 -5.52971326e-02 -8.96581471e-01 9.84828591e-01
1.50637615e+00 -3.10273170e-01 -6.44672275e-01 4.95385796e-01
1.22401214e+00 -3.60573977e-01 -2.80833393e-01 3.67740244e-01
1.43386349e-01 -7.99958467e-01 9.98900950e-01 -6.26805365e-01
8.45424533e-01 -7.89467752e-01 -6.10913299e-02 -1.34429121e+00
-2.90811986e-01 -7.87092447e-01 1.12199128e-01 1.33047569e+00
5.65039694e-01 -2.39529058e-01 5.64605176e-01 6.71892941e-01
-4.62171495e-01 -6.13451838e-01 -4.99347985e-01 -7.77658701e-01
-2.96640605e-01 -2.16080561e-01 7.14687347e-01 8.97127450e-01
-2.61908799e-01 6.42396212e-01 -3.36070806e-01 1.86880857e-01
3.43818933e-01 2.42898464e-01 7.53666162e-01 -7.03317106e-01
-2.36046538e-01 -2.67170906e-01 -4.34486903e-02 -1.11234140e+00
5.11356771e-01 -9.05924201e-01 1.95128590e-01 -2.12142086e+00
3.51864249e-01 -3.17384899e-01 -3.66401370e-03 9.07559097e-01
-5.39239943e-01 4.59004603e-02 4.91625696e-01 2.24510193e-01
-9.49403048e-01 7.79255271e-01 1.89199328e+00 -3.64419430e-01
-2.33838364e-01 -3.45855474e-01 -1.17815149e+00 3.60711396e-01
8.27805281e-01 -2.34865710e-01 -8.94818068e-01 -4.78158444e-01
3.60617340e-01 1.30967572e-01 4.63418692e-01 -7.44148612e-01
-1.14793830e-01 -5.53672373e-01 4.52948771e-02 -9.93878245e-02
2.13161916e-01 -5.60069978e-01 1.09228380e-01 2.07507744e-01
-6.66002512e-01 -7.92290792e-02 -3.24343517e-02 5.12659669e-01
-4.14947957e-01 -4.26985547e-02 7.06186354e-01 -3.18686604e-01
-1.06604898e+00 3.09756130e-01 -1.65886566e-01 3.21382105e-01
1.40269506e+00 -1.01726307e-02 -4.20353591e-01 -6.53243899e-01
-5.67472935e-01 6.99369311e-01 4.98317808e-01 7.66773105e-01
7.06486404e-01 -1.37491679e+00 -8.35019588e-01 -1.75110295e-01
6.84321523e-01 3.37401837e-01 9.17332470e-02 3.41676354e-01
-3.69804889e-01 2.37042412e-01 -4.17963266e-02 -4.96534765e-01
-8.39179516e-01 7.69131422e-01 2.87226975e-01 -6.50915280e-02
-4.59580272e-01 7.62078464e-01 8.03002596e-01 -2.37625375e-01
2.74533909e-02 -2.82466084e-01 -3.26874673e-01 7.52274171e-02
4.41784948e-01 -3.18915367e-01 -5.02883554e-01 -6.41282141e-01
-2.95889676e-01 3.33294094e-01 1.70254529e-01 -2.34110206e-01
1.13241482e+00 -6.55974448e-02 -1.00911327e-01 5.20644449e-02
7.74727643e-01 -1.47252813e-01 -1.55408978e+00 2.14643598e-01
-1.79209575e-01 -2.74172455e-01 -4.09089088e-01 -9.51924503e-01
-9.63575900e-01 3.75009626e-01 -6.83033764e-02 -1.42524941e-02
1.26911950e+00 4.27387089e-01 4.94432330e-01 2.00911924e-01
3.68294984e-01 -8.62821698e-01 5.54216743e-01 2.18947425e-01
1.58071280e+00 -1.02265036e+00 -4.97207344e-01 -7.51834333e-01
-1.35282469e+00 6.21755242e-01 1.09213150e+00 1.74167410e-01
-9.65446606e-02 -1.97399065e-01 -5.34085324e-03 -1.98284775e-01
-8.49399805e-01 -5.58420181e-01 1.90056026e-01 8.32645535e-01
2.27231503e-01 6.23001941e-02 -2.62860924e-01 6.07396007e-01
9.30182114e-02 -6.78914413e-02 6.56102538e-01 6.70075059e-01
-1.09370105e-01 -9.71078038e-01 -1.73283756e-01 1.49386108e-01
-7.33372495e-02 -2.62604445e-01 -3.95655632e-01 7.81901717e-01
7.36885667e-02 1.03011608e+00 3.26438993e-03 -6.05294965e-02
6.26996934e-01 -6.48128316e-02 3.69891167e-01 -1.01984596e+00
-2.97669888e-01 -1.02997802e-01 2.47335508e-01 -7.52646685e-01
-7.15686083e-01 -5.49651861e-01 -1.54356229e+00 4.81346339e-01
1.23930633e-01 2.12081864e-01 3.54581892e-01 4.94320363e-01
4.82249558e-01 6.72544718e-01 5.24648786e-01 -6.89954460e-01
-2.52077073e-01 -7.84192443e-01 -3.04319888e-01 9.77589309e-01
4.61964235e-02 -6.32537425e-01 -3.88062090e-01 5.37755013e-01] | [10.843063354492188, 0.9699905514717102] |
79d8014d-9032-4ddb-bf4e-02e0b3f48d38 | vimi-vehicle-infrastructure-multi-view | 2303.10975 | null | https://arxiv.org/abs/2303.10975v1 | https://arxiv.org/pdf/2303.10975v1.pdf | VIMI: Vehicle-Infrastructure Multi-view Intermediate Fusion for Camera-based 3D Object Detection | In autonomous driving, Vehicle-Infrastructure Cooperative 3D Object Detection (VIC3D) makes use of multi-view cameras from both vehicles and traffic infrastructure, providing a global vantage point with rich semantic context of road conditions beyond a single vehicle viewpoint. Two major challenges prevail in VIC3D: 1) inherent calibration noise when fusing multi-view images, caused by time asynchrony across cameras; 2) information loss when projecting 2D features into 3D space. To address these issues, We propose a novel 3D object detection framework, Vehicles-Infrastructure Multi-view Intermediate fusion (VIMI). First, to fully exploit the holistic perspectives from both vehicles and infrastructure, we propose a Multi-scale Cross Attention (MCA) module that fuses infrastructure and vehicle features on selective multi-scales to correct the calibration noise introduced by camera asynchrony. Then, we design a Camera-aware Channel Masking (CCM) module that uses camera parameters as priors to augment the fused features. We further introduce a Feature Compression (FC) module with channel and spatial compression blocks to reduce the size of transmitted features for enhanced efficiency. Experiments show that VIMI achieves 15.61% overall AP_3D and 21.44% AP_BEV on the new VIC3D dataset, DAIR-V2X-C, significantly outperforming state-of-the-art early fusion and late fusion methods with comparable transmission cost. | ['Ya-Qin Zhang', 'Yilun Chen', 'Jingjing Liu', 'Yan Wang', 'Tongda Xu', 'Xiaoliang Huo', 'Siqi Fan', 'Zhe Wang'] | 2023-03-20 | null | null | null | null | ['feature-compression'] | ['computer-vision'] | [ 1.82574242e-02 -4.29965019e-01 -1.86856121e-01 -2.71882594e-01
-9.50929761e-01 -6.50397599e-01 6.55171037e-01 -3.64688188e-01
-3.64793688e-01 2.26137370e-01 -5.61949983e-03 -3.59690905e-01
3.23882788e-01 -7.28378356e-01 -9.77818370e-01 -6.67228222e-01
5.78782037e-02 -1.79794729e-01 5.96100807e-01 -1.27186045e-01
1.12256035e-02 6.33156121e-01 -1.73446012e+00 1.26226187e-01
6.86729193e-01 1.27896869e+00 5.50375879e-01 7.33036160e-01
1.29244342e-01 5.17285526e-01 -3.03457618e-01 -3.69445294e-01
6.34537458e-01 1.15722775e-01 3.85921039e-02 3.27134848e-01
5.03131628e-01 -8.61801386e-01 -7.08182395e-01 1.10981917e+00
3.45475972e-01 -2.46367007e-01 3.35789502e-01 -1.87231028e+00
-2.59303480e-01 -1.06992029e-01 -1.10792387e+00 4.39257830e-01
2.26580247e-01 3.01010936e-01 4.60126936e-01 -1.11469519e+00
4.64507699e-01 1.49208462e+00 5.85925460e-01 2.87176847e-01
-8.66558015e-01 -1.05956352e+00 2.47676671e-01 6.50944650e-01
-1.61749315e+00 -6.70681477e-01 6.89719558e-01 -2.13417456e-01
7.62400806e-01 1.98723689e-01 4.69489008e-01 7.23306417e-01
3.58635128e-01 7.75625348e-01 9.22841847e-01 2.12519392e-02
-1.80333391e-01 2.40421832e-01 -8.43532607e-02 4.10107255e-01
4.54895914e-01 6.08381569e-01 -6.31715894e-01 8.22394118e-02
3.71610492e-01 3.27760518e-01 -2.56739527e-01 -1.93119064e-01
-1.19206870e+00 6.68543220e-01 3.07731301e-01 -2.76407331e-01
-4.24659759e-01 3.30506533e-01 2.59433478e-01 3.19798678e-01
4.09783661e-01 -6.37874484e-01 -3.30342174e-01 7.85326958e-02
-6.30001366e-01 6.91070259e-02 1.30198807e-01 1.65654039e+00
1.08451617e+00 -1.38506377e-02 -6.11900128e-02 4.55550879e-01
6.02629662e-01 1.34600484e+00 -2.78140485e-01 -1.24072802e+00
9.14300323e-01 3.49622518e-01 -1.00906692e-01 -9.46227372e-01
-1.43134356e-01 -1.98398411e-01 -7.41693139e-01 4.07888353e-01
-2.51992613e-01 -2.49719873e-01 -7.32166767e-01 1.60511458e+00
6.45126045e-01 4.53849703e-01 4.46730256e-01 1.09882832e+00
8.75988901e-01 7.55964458e-01 -3.39234352e-01 -2.72213668e-01
1.41246343e+00 -6.76616192e-01 -6.81743979e-01 -2.60113031e-01
4.39454436e-01 -8.11244428e-01 6.66772872e-02 4.30615433e-02
-1.08204019e+00 -7.55873263e-01 -1.32470357e+00 1.10732566e-03
-3.37321997e-01 2.53092535e-02 7.01973587e-02 8.08999121e-01
-1.02340555e+00 -4.88103956e-01 -6.40968382e-01 -1.06239110e-01
3.81691486e-01 2.31441006e-01 -4.68555689e-01 -6.20297074e-01
-1.20168412e+00 9.43846822e-01 -7.24227130e-02 3.96546870e-02
-1.15050280e+00 -8.11995029e-01 -9.02639329e-01 -1.21659264e-01
6.37252867e-01 -5.26004672e-01 8.25522423e-01 -4.82183963e-01
-1.10341299e+00 5.04485607e-01 -5.50332844e-01 -5.83886504e-01
4.89068568e-01 -2.20512226e-01 -6.95852339e-01 3.69133592e-01
2.28743985e-01 7.83608079e-01 8.12675714e-01 -1.63861227e+00
-1.17109883e+00 -5.30343056e-01 -1.60152882e-01 4.49671537e-01
2.35798791e-01 -6.26138300e-02 -1.10504711e+00 -2.23606020e-01
1.73390061e-01 -1.07306588e+00 -7.96951652e-02 1.31560951e-01
-2.27058098e-01 1.67311430e-01 1.43173063e+00 -4.22768086e-01
7.75261819e-01 -2.47728348e+00 -1.59669071e-01 1.58315688e-01
5.10519683e-01 2.04512149e-01 -2.36271739e-01 1.46514177e-01
3.69771779e-01 -3.07815194e-01 7.74058402e-02 -5.35592973e-01
-2.18798101e-01 3.63179684e-01 -3.33515406e-01 7.36517966e-01
6.18367828e-02 8.09827447e-01 -7.53428519e-01 -4.61614013e-01
7.49943078e-01 7.77470767e-01 -4.85731930e-01 4.05764915e-02
4.01587635e-01 2.44840205e-01 -3.78981709e-01 9.05008614e-01
1.60678244e+00 6.30816445e-02 -2.81966329e-01 -4.69184101e-01
-3.30724716e-01 -1.72426060e-01 -1.25645518e+00 1.36635160e+00
-4.85713720e-01 8.45353901e-01 4.75050718e-01 -6.15403771e-01
6.88928723e-01 2.70200729e-01 4.37112421e-01 -8.79466414e-01
2.01205000e-01 2.57169724e-01 -4.72819090e-01 -3.53693157e-01
7.02148199e-01 3.88659388e-01 -2.56314278e-01 1.21129286e-02
-1.49474917e-02 1.37247927e-02 -1.66226894e-01 5.42401016e-01
1.10172617e+00 -4.31558460e-01 -8.94422531e-02 1.65199220e-01
8.39707851e-01 -2.62162685e-01 7.02243567e-01 5.38063765e-01
-3.68560284e-01 4.60297465e-01 1.34197459e-01 -6.40421808e-02
-8.96362960e-01 -1.22185767e+00 -7.73343146e-02 4.02444273e-01
9.89214540e-01 -2.27348223e-01 -4.03666973e-01 -6.35536194e-01
3.72234583e-01 5.78608215e-01 -4.05752420e-01 -2.75138557e-01
-4.17252779e-01 -2.34617159e-01 3.58759999e-01 3.80662650e-01
8.60249400e-01 3.02180219e-02 -7.70204127e-01 1.80546165e-01
-4.19498324e-01 -1.90715945e+00 -7.79660523e-01 -1.19141959e-01
-4.59735751e-01 -1.12638116e+00 -4.61221129e-01 -1.32175475e-01
4.07261878e-01 1.53779435e+00 6.41290665e-01 -1.21835224e-01
-8.13399181e-02 7.45777071e-01 -4.43377852e-01 -4.41193312e-01
-4.98589963e-01 -5.48791170e-01 1.64960802e-01 4.23667580e-01
5.31458974e-01 -3.04361999e-01 -6.35793805e-01 6.07677400e-01
-6.54454947e-01 1.84722766e-01 7.82003462e-01 4.69211668e-01
6.42675042e-01 2.38538887e-02 2.31377333e-01 -1.19394377e-01
-2.19258025e-01 -7.20572710e-01 -9.60860014e-01 -1.16286054e-01
-5.26633680e-01 -5.25029063e-01 1.90144360e-01 -9.69373211e-02
-1.08001268e+00 2.20934153e-01 1.17254920e-01 -9.84817147e-01
-3.02708149e-02 -1.27559736e-01 -5.08575141e-01 -3.76072913e-01
3.82365137e-02 4.32954609e-01 2.35396340e-01 -1.03040926e-01
5.45775831e-01 9.84714091e-01 6.81347370e-01 8.17561001e-02
1.10039866e+00 1.06858897e+00 1.62019446e-01 -9.48558569e-01
-2.32737452e-01 -7.18944013e-01 -2.38629773e-01 -6.38917387e-01
9.67120528e-01 -1.87951434e+00 -7.61887133e-01 5.42582393e-01
-1.34307635e+00 1.94434613e-01 1.38590321e-01 7.57446110e-01
-3.84789854e-01 5.13170063e-01 -1.75386235e-01 -6.90589428e-01
-7.62614086e-02 -1.56379211e+00 1.38213348e+00 -7.95850754e-02
6.27943337e-01 -3.07551235e-01 -3.27719688e-01 5.40467680e-01
3.15875322e-01 -5.87415881e-03 1.84268430e-01 5.87801412e-02
-1.28626728e+00 -1.62529305e-01 -7.68507183e-01 3.66741955e-01
-5.76765649e-02 -3.66723150e-01 -1.11276472e+00 -3.62050831e-01
-3.91916707e-02 2.73716003e-01 1.03058410e+00 4.19282466e-01
6.56119347e-01 8.73557106e-02 -7.29837060e-01 8.66562128e-01
1.52300847e+00 2.24603891e-01 5.19545615e-01 -5.53908832e-02
8.25918734e-01 2.55670458e-01 8.42709959e-01 4.84520465e-01
9.77913380e-01 9.84086096e-01 8.18004906e-01 -6.66180626e-02
-4.29605514e-01 -4.08125371e-02 6.47218406e-01 5.39555669e-01
1.85518980e-01 -4.42073107e-01 -4.32817996e-01 6.43643677e-01
-1.70293331e+00 -9.73402619e-01 -3.73398721e-01 2.10950494e+00
-1.14264700e-03 1.31215289e-01 -7.50547349e-02 5.49557991e-02
9.44856942e-01 2.67688245e-01 -5.98489165e-01 4.95089553e-02
-3.65986407e-01 -6.99437201e-01 1.29897320e+00 5.28990626e-01
-9.60128188e-01 4.77699041e-01 4.91907310e+00 9.57162380e-01
-9.18910027e-01 5.54105639e-01 2.24709168e-01 -2.38638774e-01
-3.06607962e-01 -6.23459853e-02 -1.15010500e+00 7.54753947e-01
9.57713485e-01 1.42806973e-02 2.79086947e-01 5.63565612e-01
3.10357362e-01 -2.82390118e-01 -8.41729105e-01 1.25590289e+00
2.89060473e-01 -1.51924598e+00 -1.01381615e-01 6.72360897e-01
6.29675627e-01 7.72935510e-01 2.82665081e-02 -9.28868055e-02
1.50612310e-01 -3.16747427e-01 1.15039897e+00 2.04858854e-01
9.77819741e-01 -8.98974538e-01 5.79054773e-01 2.33095244e-01
-1.78684914e+00 -3.30774248e-01 -2.03255937e-01 4.76543278e-01
7.55072653e-01 6.90159321e-01 -3.00290346e-01 7.84862638e-01
9.74709392e-01 8.39493275e-01 -4.62129802e-01 7.80907333e-01
1.48351654e-01 1.64430007e-01 -6.21658206e-01 4.84708995e-01
2.79454112e-01 1.24376595e-01 1.06876314e+00 9.95185375e-01
6.16192102e-01 2.93099225e-01 6.72202930e-02 5.99874794e-01
3.24662365e-02 -5.95254779e-01 -1.02023923e+00 8.53705227e-01
1.03654611e+00 1.09552789e+00 -2.60146499e-01 -3.82845074e-01
-1.11646736e+00 7.82265067e-01 -4.18855608e-01 5.34709513e-01
-1.11382639e+00 -6.32928237e-02 1.08724523e+00 -3.59610803e-02
8.13065767e-01 -3.72431040e-01 -2.07240451e-02 -1.08137131e+00
2.87141740e-01 -3.32674950e-01 6.33639991e-02 -8.44610870e-01
-7.39871562e-01 3.51148218e-01 1.76978409e-01 -1.77296782e+00
2.29375497e-01 -1.22203626e-01 -2.85052240e-01 5.71151972e-01
-2.17218304e+00 -1.21929908e+00 -5.89629412e-01 9.24261272e-01
5.05071402e-01 -2.45975956e-01 -4.99958917e-02 8.77927899e-01
-2.92005956e-01 5.73581040e-01 1.96841896e-01 -2.79578596e-01
5.64082444e-01 -5.81184685e-01 4.92337435e-01 1.14276075e+00
-2.22575605e-01 -2.62492239e-01 3.52657080e-01 -4.46001530e-01
-2.04635119e+00 -1.45847881e+00 7.02266753e-01 -3.45322609e-01
6.96006864e-02 -4.40281868e-01 -6.08166575e-01 5.64540327e-01
2.26110756e-01 4.42034006e-01 2.64618039e-01 -7.90138245e-01
-4.14867371e-01 -4.85570461e-01 -1.15468884e+00 2.49631882e-01
1.03472936e+00 -5.16193151e-01 -8.58914629e-02 -8.30660090e-02
9.91941750e-01 -5.46125710e-01 -6.32245243e-01 5.24471521e-01
4.37620968e-01 -1.01393867e+00 1.20092416e+00 5.02025247e-01
-2.87037015e-01 -8.81455719e-01 -7.68066823e-01 -8.19546103e-01
1.52327493e-02 -6.63464904e-01 -3.07762235e-01 1.19079268e+00
3.43714654e-02 -8.75000060e-01 4.41602975e-01 2.38984525e-01
-4.53954250e-01 -1.67222396e-01 -1.54393160e+00 -7.37732887e-01
-6.10815048e-01 -9.97769773e-01 8.30641031e-01 5.40187359e-01
-5.28716147e-01 1.53559297e-01 -5.09555578e-01 8.24967146e-01
1.12764120e+00 -1.76255126e-02 1.22455931e+00 -9.59512770e-01
2.51716912e-01 -8.80346596e-02 -7.85926700e-01 -1.44844747e+00
-5.32130525e-02 -5.90524137e-01 7.61713088e-02 -1.03315413e+00
1.68464720e-01 -3.12415689e-01 1.52183287e-02 -6.80213347e-02
3.21973301e-02 4.10774946e-01 4.51003969e-01 2.31684864e-01
-7.89787591e-01 7.30116606e-01 9.16634738e-01 -1.54070973e-01
1.22976555e-02 -2.11757533e-02 -5.49045622e-01 4.43715870e-01
2.41194278e-01 -4.07614738e-01 -3.77449393e-01 -7.52608180e-01
-1.04831442e-01 4.14779603e-01 7.84563243e-01 -1.13472116e+00
5.26042223e-01 3.25767957e-02 2.41401911e-01 -1.47370410e+00
6.75998628e-01 -1.48429596e+00 3.26499820e-01 3.28748703e-01
3.44049901e-01 5.10022752e-02 4.33870703e-01 1.11037326e+00
-2.85666198e-01 5.54135144e-01 7.71881938e-01 3.13392699e-01
-1.12701905e+00 5.72158694e-01 -6.06721401e-01 -2.47951686e-01
1.50129139e+00 -5.24618447e-01 -5.61661363e-01 -2.63409495e-01
1.19962404e-02 7.94032037e-01 4.83353019e-01 7.68266857e-01
9.30685937e-01 -1.60562396e+00 -9.19223666e-01 6.18778765e-01
4.81549531e-01 -7.22441599e-02 8.74981105e-01 1.00799084e+00
-2.38290220e-01 5.31997681e-01 3.42216492e-02 -1.17405057e+00
-1.38797319e+00 5.92403114e-01 3.33361551e-02 4.16198730e-01
-6.84626043e-01 5.69277465e-01 4.92693931e-01 -1.30215660e-01
2.04295479e-02 -2.00626761e-01 8.61788169e-02 -9.53563079e-02
6.58283830e-01 5.82411647e-01 2.64443662e-02 -1.45074213e+00
-5.70075035e-01 8.94344985e-01 7.95211084e-03 1.82499771e-03
9.99167025e-01 -1.17351007e+00 4.04350519e-01 -1.76449344e-01
1.52234006e+00 1.55524844e-02 -1.77663624e+00 -5.59557259e-01
-7.31105983e-01 -8.06323409e-01 4.77440178e-01 -3.13401371e-01
-1.38718951e+00 8.85816097e-01 1.10258722e+00 -4.56931815e-02
1.20765126e+00 1.61854431e-01 1.08883619e+00 4.12770770e-02
4.89195973e-01 -7.94975519e-01 -2.60430098e-01 2.92589545e-01
5.46224177e-01 -1.38019907e+00 8.48362520e-02 -5.34600794e-01
-6.20549262e-01 8.27473521e-01 5.20552278e-01 2.05365792e-01
8.04069757e-01 3.32447916e-01 8.47282931e-02 -2.41127178e-01
-9.30470705e-01 -4.07311738e-01 5.02429940e-02 8.10348332e-01
-6.92868173e-01 -1.26624126e-02 3.91876847e-01 2.46423841e-01
4.68467325e-01 -2.35698104e-01 5.49474001e-01 7.73790002e-01
-3.87848735e-01 -6.27053499e-01 -6.25904739e-01 1.47983566e-01
-1.88312531e-01 9.34501663e-02 2.37706244e-01 7.65850425e-01
4.32431191e-01 1.35935497e+00 1.62968084e-01 -8.24751914e-01
4.38817322e-01 -6.40204191e-01 1.92982897e-01 -1.11144818e-02
8.13794285e-02 3.17120761e-01 -3.38770859e-02 -9.27309036e-01
-5.90259671e-01 -8.09773326e-01 -9.51237321e-01 -5.72158873e-01
-5.43402970e-01 -1.85458630e-01 9.36327994e-01 8.76837969e-01
9.21077728e-01 4.17241216e-01 1.20083570e+00 -1.23384833e+00
-5.44926189e-02 -3.29450250e-01 -5.50536394e-01 -1.80991620e-01
1.02309239e+00 -8.78508031e-01 -6.30646229e-01 3.19104008e-02] | [7.9682416915893555, -1.7866265773773193] |
cb4b894f-993d-4d5a-b424-bafef8c18822 | fast-template-matching-and-update-for-video | 2004.07538 | null | https://arxiv.org/abs/2004.07538v1 | https://arxiv.org/pdf/2004.07538v1.pdf | Fast Template Matching and Update for Video Object Tracking and Segmentation | In this paper, the main task we aim to tackle is the multi-instance semi-supervised video object segmentation across a sequence of frames where only the first-frame box-level ground-truth is provided. Detection-based algorithms are widely adopted to handle this task, and the challenges lie in the selection of the matching method to predict the result as well as to decide whether to update the target template using the newly predicted result. The existing methods, however, make these selections in a rough and inflexible way, compromising their performance. To overcome this limitation, we propose a novel approach which utilizes reinforcement learning to make these two decisions at the same time. Specifically, the reinforcement learning agent learns to decide whether to update the target template according to the quality of the predicted result. The choice of the matching method will be determined at the same time, based on the action history of the reinforcement learning agent. Experiments show that our method is almost 10 times faster than the previous state-of-the-art method with even higher accuracy (region similarity of 69.1% on DAVIS 2017 dataset). | ['Jimin Xiao', 'Mingjie Sun', 'Yao Zhao', 'Eng Gee Lim', 'Bingfeng Zhang'] | 2020-04-16 | fast-template-matching-and-update-for-video-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Sun_Fast_Template_Matching_and_Update_for_Video_Object_Tracking_and_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Sun_Fast_Template_Matching_and_Update_for_Video_Object_Tracking_and_CVPR_2020_paper.pdf | cvpr-2020-6 | ['template-matching', 'video-object-tracking'] | ['computer-vision', 'computer-vision'] | [ 3.92251164e-01 -2.06894338e-01 -2.15608507e-01 -2.90952176e-01
-7.58022070e-01 -5.51343918e-01 4.48990703e-01 2.97143787e-01
-8.50083888e-01 6.98698163e-01 -4.15741324e-01 -4.91330214e-02
1.64336920e-01 -8.80394459e-01 -7.05092072e-01 -9.11259294e-01
2.75806278e-01 7.63310432e-01 9.54665959e-01 5.10914139e-02
4.94014025e-01 3.86248678e-01 -1.63850009e+00 1.81850076e-01
8.71860802e-01 1.27912533e+00 4.28490788e-01 5.91850340e-01
-5.28634489e-02 7.83088148e-01 -3.75663817e-01 -2.92486250e-01
2.55439848e-01 -4.66365606e-01 -8.95703852e-01 6.24089003e-01
2.61195123e-01 -4.76189405e-01 1.21195547e-01 1.08855116e+00
3.39313060e-01 2.66334325e-01 5.41325331e-01 -9.83802319e-01
2.90854365e-01 3.96072865e-01 -6.00987673e-01 2.07971707e-01
2.73623854e-01 2.04575926e-01 8.25484693e-01 -6.79166436e-01
8.82288337e-01 9.04245436e-01 1.64308235e-01 5.14419138e-01
-9.26052749e-01 -3.51832062e-01 5.33491075e-01 5.66921592e-01
-1.29074252e+00 -3.88968676e-01 7.11623073e-01 -4.87003088e-01
3.80341053e-01 -8.33350196e-02 7.68723905e-01 5.85545540e-01
6.47064745e-02 9.02468801e-01 1.17345059e+00 -4.58432227e-01
6.16044939e-01 -3.28404605e-02 -3.24116379e-01 8.84849191e-01
-7.49415457e-02 1.27083454e-02 -3.22802663e-01 8.41307119e-02
6.66586578e-01 -2.49569397e-02 -2.63644189e-01 -6.29985750e-01
-1.11419678e+00 5.35673738e-01 3.65493923e-01 4.21291590e-01
-6.69374347e-01 -3.71120945e-02 3.83730084e-01 -6.34922311e-02
2.01670364e-01 2.22682506e-01 -4.11595017e-01 -1.41102538e-01
-1.02715373e+00 3.42292964e-01 5.00642121e-01 6.13728404e-01
7.78996527e-01 -2.39862800e-01 -2.73201138e-01 6.01077020e-01
2.74316519e-01 9.86115187e-02 2.15536028e-01 -9.49153066e-01
3.47295910e-01 5.60525894e-01 4.26160097e-01 -7.35156357e-01
-2.32808992e-01 -2.62047410e-01 -4.29819465e-01 4.80808258e-01
8.43748629e-01 -2.17157900e-01 -1.01986098e+00 1.30141687e+00
8.29073787e-01 2.93359905e-01 -1.61928087e-01 1.23932445e+00
3.42371076e-01 7.51363039e-01 2.22534046e-01 -3.25864494e-01
1.22888684e+00 -1.03056562e+00 -5.95738769e-01 -3.68867278e-01
4.88651037e-01 -7.24308848e-01 5.41155219e-01 4.45260227e-01
-1.02859282e+00 -6.70614541e-01 -1.05958581e+00 5.98710775e-01
-1.19465418e-01 3.70922327e-01 2.41335481e-01 2.78743267e-01
-7.01035261e-01 8.69410515e-01 -8.70243907e-01 -2.21033737e-01
4.14760292e-01 2.68336773e-01 -1.67503834e-01 -1.67898655e-01
-9.84141648e-01 8.25850904e-01 7.12319195e-01 2.05404624e-01
-8.75242472e-01 -2.70184815e-01 -6.02885187e-01 -1.19628757e-01
1.01155365e+00 -4.24359024e-01 1.30059075e+00 -1.30384016e+00
-1.67009532e+00 6.81272686e-01 6.58769459e-02 -6.52658701e-01
9.62538898e-01 -1.08722867e-02 -4.40811738e-02 4.23291385e-01
2.02667147e-01 8.55781615e-01 1.01558566e+00 -1.14392638e+00
-1.37086833e+00 -4.15035754e-01 3.23274225e-01 2.93175668e-01
1.11686997e-01 -9.38105509e-02 -1.03042638e+00 -3.97654533e-01
5.05110174e-02 -9.83855903e-01 -4.42975253e-01 1.70477331e-01
-1.20739266e-01 -4.00647014e-01 7.22053707e-01 -5.67396462e-01
9.36327696e-01 -1.78208578e+00 1.56367630e-01 -5.77257015e-02
-9.05567780e-02 5.14539838e-01 9.47090834e-02 2.17940356e-03
4.32056874e-01 -2.10964411e-01 -4.02868092e-01 -2.01885968e-01
-3.88913959e-01 -7.51370490e-02 1.06481276e-01 5.95824122e-01
3.24060917e-01 4.26132411e-01 -9.62716579e-01 -8.66824985e-01
6.21123970e-01 1.30393907e-01 -4.62677807e-01 5.08557081e-01
-4.55962479e-01 6.27871156e-01 -7.38135040e-01 5.33226132e-01
5.40545821e-01 -1.41602337e-01 2.00892165e-01 -1.85668305e-01
-2.83363104e-01 5.78663535e-02 -1.54829252e+00 1.67782581e+00
-1.98641598e-01 2.52528727e-01 -1.22825362e-01 -1.13832986e+00
8.16547453e-01 1.75749838e-01 6.22971892e-01 -5.82792699e-01
3.36720318e-01 2.75001019e-01 2.55184710e-01 -4.54083294e-01
2.59941280e-01 3.79261672e-02 1.91245586e-01 2.80123293e-01
-2.12986365e-01 -5.49067259e-02 6.03486776e-01 -1.45907387e-01
8.93669188e-01 7.94580281e-01 3.42738688e-01 9.18315351e-02
1.00317562e+00 1.83063880e-01 1.00514495e+00 7.86856830e-01
-5.12343943e-01 7.71134436e-01 4.70716745e-01 -6.49192452e-01
-8.57049048e-01 -4.58042651e-01 1.26141950e-01 7.26635516e-01
6.55839086e-01 -6.27680719e-02 -1.04661393e+00 -1.11955154e+00
-3.49478751e-01 6.21387243e-01 -5.42704046e-01 -5.52483834e-02
-7.06711531e-01 -3.58128965e-01 -1.27284899e-01 5.25409698e-01
8.25170696e-01 -1.24882877e+00 -1.21910930e+00 5.55831194e-01
-1.30291075e-01 -1.43226719e+00 -4.41439182e-01 3.64230052e-02
-7.88771391e-01 -1.30113852e+00 -6.73014402e-01 -4.65305477e-01
7.67904341e-01 1.23919444e-02 7.89065480e-01 3.39936882e-01
-1.13043964e-01 -3.71659026e-02 -4.44463134e-01 -7.84777943e-03
-5.89397371e-01 -1.05199581e-02 -2.79854983e-01 4.81814146e-01
-9.44695622e-02 6.41230866e-02 -7.15015948e-01 6.45582616e-01
-8.96509171e-01 2.47217953e-01 6.65672898e-01 6.21071160e-01
9.69867110e-01 2.77728647e-01 4.75589782e-01 -8.30650508e-01
-2.62413546e-02 2.52942704e-02 -1.00956190e+00 5.35367727e-01
-4.67076153e-01 1.37104183e-01 6.86626375e-01 -6.15219831e-01
-1.05724585e+00 6.35317683e-01 -1.40365854e-01 -4.68013734e-01
-2.67744362e-01 1.62862524e-01 -1.79219320e-01 -5.47619201e-02
2.36385897e-01 1.83400780e-01 -1.51957750e-01 -1.97870135e-01
3.19362469e-02 3.95380527e-01 4.09077048e-01 -4.01940823e-01
5.63607275e-01 3.57970089e-01 -1.66240141e-01 -1.34255961e-01
-7.57545054e-01 -4.84109670e-01 -1.04591763e+00 -6.81170344e-01
9.90427434e-01 -6.69898808e-01 -6.22043371e-01 5.43192267e-01
-1.14312863e+00 -3.61834943e-01 3.17640342e-02 4.10638213e-01
-7.88899601e-01 1.66399091e-01 -2.33134925e-01 -8.48422766e-01
-2.01464117e-01 -1.62456453e+00 1.05198622e+00 3.71859699e-01
1.63773377e-03 -4.83462453e-01 -2.98528105e-01 6.13474786e-01
9.07491148e-03 2.91276336e-01 5.02297640e-01 -5.27135253e-01
-8.53827953e-01 -1.60008356e-01 -1.08794212e-01 1.20230049e-01
7.08152130e-02 2.22338185e-01 -5.64566076e-01 4.37923387e-04
-2.16251895e-01 -3.48287448e-02 7.39334226e-01 4.83187705e-01
1.05731797e+00 1.51732162e-01 -2.67549038e-01 2.28803590e-01
1.35890841e+00 5.76969266e-01 5.67219973e-01 6.03423357e-01
4.95208472e-01 5.74325442e-01 1.36158323e+00 4.12413806e-01
2.88474232e-01 1.09280884e+00 6.93139732e-01 1.70108169e-01
-1.42584100e-01 -5.80011979e-02 2.00364709e-01 6.98104734e-03
-1.12523526e-01 -8.91368613e-02 -7.40292788e-01 3.82826447e-01
-2.12287951e+00 -9.13422883e-01 2.47933567e-01 2.40501451e+00
8.14103663e-01 4.82457489e-01 3.32087398e-01 2.80845702e-01
9.47168827e-01 1.51070997e-01 -7.47805059e-01 1.43903997e-02
2.75826484e-01 -4.11792845e-01 4.44707394e-01 3.20649832e-01
-1.36733079e+00 1.21849954e+00 4.27454805e+00 8.64732981e-01
-1.27935219e+00 -1.22035146e-01 8.17071319e-01 1.36971056e-01
2.89332151e-01 1.64986104e-01 -7.95051932e-01 5.23866892e-01
4.69018310e-01 3.32492590e-01 3.92720968e-01 6.62716746e-01
2.52340764e-01 -6.81962848e-01 -1.07837069e+00 8.70342374e-01
-2.52057444e-02 -1.29752052e+00 -1.70470878e-01 -3.38944733e-01
6.24912441e-01 -3.74497712e-01 -4.57992345e-01 9.49421078e-02
-1.55024290e-01 -3.75057310e-01 1.06756949e+00 5.40380061e-01
2.03291357e-01 -1.01117992e+00 7.28614390e-01 5.88896453e-01
-1.29279399e+00 -4.70144525e-02 -1.84637055e-01 1.75937921e-01
2.81934112e-01 4.10768867e-01 -9.13065374e-01 5.06919146e-01
6.41221166e-01 5.15981078e-01 -3.78719121e-01 1.33127296e+00
-5.28232813e-01 3.41180712e-01 -1.47026405e-01 -4.06918637e-02
3.62680018e-01 -2.17860803e-01 5.36845624e-01 9.70169067e-01
1.19538933e-01 1.51662692e-01 6.60089076e-01 7.15281248e-01
1.44759089e-01 1.70824513e-01 -2.44214851e-02 1.16319999e-01
2.61899590e-01 1.32736564e+00 -1.43867862e+00 -4.92244005e-01
-7.60940686e-02 9.77399111e-01 4.19457018e-01 9.46764424e-02
-9.53861117e-01 -5.41393906e-02 3.34589601e-01 3.35694551e-01
6.40025914e-01 -1.25573009e-01 -3.45617421e-02 -8.90304685e-01
6.13314584e-02 -7.84092963e-01 4.79771286e-01 -6.65289879e-01
-7.36006021e-01 7.71210194e-01 -1.02258302e-01 -1.33283496e+00
-5.45718789e-01 -3.19743395e-01 -4.73341256e-01 3.62193555e-01
-1.59715414e+00 -6.46655381e-01 -1.75805405e-01 2.93074697e-01
9.63950038e-01 1.31939292e-01 3.10717940e-01 2.11154312e-01
-7.54693687e-01 1.53245658e-01 -2.57659703e-01 1.69913113e-01
5.41498482e-01 -1.14862466e+00 7.12746233e-02 9.73967791e-01
1.41134217e-01 -1.10816680e-01 7.49599218e-01 -7.11199403e-01
-1.19453359e+00 -9.67006266e-01 4.90472674e-01 1.45083845e-01
3.70467871e-01 -4.96907495e-02 -9.48622823e-01 1.56438619e-01
-1.37771875e-01 3.78358454e-01 -7.40126567e-03 -4.52986985e-01
1.53459191e-01 -2.81809121e-01 -1.19188845e+00 6.71035230e-01
7.22636819e-01 4.47105803e-02 -3.24165314e-01 1.32530779e-01
3.93702120e-01 -7.94627309e-01 -5.28589725e-01 6.99307263e-01
4.60507780e-01 -1.04124427e+00 6.84971273e-01 -2.40530536e-01
3.81307989e-01 -7.34059036e-01 1.69363514e-01 -1.03289485e+00
-3.03477682e-02 -4.21176463e-01 -8.07874743e-03 1.11071754e+00
2.93188423e-01 -3.61523479e-01 8.34071696e-01 7.22358108e-01
2.43242383e-01 -1.24123275e+00 -1.00143361e+00 -4.51476216e-01
-6.03693843e-01 -2.64400005e-01 4.43250775e-01 4.61404622e-01
-4.75993127e-01 1.14460081e-01 -3.24803829e-01 3.16552311e-01
5.10053396e-01 6.31817102e-01 7.49651253e-01 -1.05316496e+00
-1.39797777e-01 -3.47088993e-01 -5.14464796e-01 -1.06716812e+00
1.09656580e-01 -3.36941570e-01 4.70073789e-01 -1.47681272e+00
5.62980138e-02 -6.64994240e-01 -2.79041797e-01 3.95067245e-01
-4.39356208e-01 1.40951246e-01 5.35626948e-01 1.05773419e-01
-9.49074507e-01 4.09813195e-01 1.36393499e+00 -1.71959609e-01
-3.91493946e-01 5.48026443e-01 -2.67048180e-01 8.60960543e-01
7.35404372e-01 -5.09046853e-01 -2.45167121e-01 -2.36997038e-01
-3.37673396e-01 4.16057587e-01 3.07615638e-01 -1.09801054e+00
4.77199733e-01 -4.18434083e-01 4.98941749e-01 -7.00579941e-01
1.93827763e-01 -9.02324736e-01 -3.94056179e-02 6.41306818e-01
-2.89654821e-01 4.66091409e-02 2.60958727e-02 6.58272147e-01
-6.55920580e-02 -5.52221775e-01 1.07422447e+00 -3.46214920e-02
-1.06886232e+00 5.59589684e-01 -3.64020348e-01 -1.81675032e-01
1.36663580e+00 -2.61626393e-01 2.01764956e-01 -2.59821475e-01
-6.55488849e-01 2.95750141e-01 3.10804963e-01 4.20596898e-01
5.77739000e-01 -9.17902946e-01 -6.62515402e-01 2.30513816e-03
-7.90305156e-03 1.99220181e-01 1.70033202e-01 7.68345416e-01
-3.99941444e-01 -1.43023981e-02 -1.57933891e-01 -8.11390042e-01
-1.22794664e+00 6.17256761e-01 5.55861413e-01 -3.86252940e-01
-5.48576236e-01 4.11981374e-01 -8.28196555e-02 1.28358930e-01
3.52182120e-01 -1.28486067e-01 -5.37632287e-01 2.02794611e-01
5.38696885e-01 2.49233618e-01 1.97813615e-01 -6.93405569e-01
-3.31540227e-01 6.06405497e-01 -1.63359150e-01 -1.09935373e-01
1.23560631e+00 -1.90579280e-01 1.54818565e-01 2.50625163e-01
8.48815382e-01 -3.79342109e-01 -1.71166098e+00 -3.45489442e-01
6.98970770e-03 -6.39339685e-01 6.26823083e-02 -8.48442078e-01
-1.30458438e+00 5.29749751e-01 6.98463798e-01 -1.91682186e-02
1.17746389e+00 -1.82956070e-01 6.00818574e-01 1.39932439e-01
5.70409596e-01 -1.40768635e+00 -1.70558095e-02 1.77618489e-01
6.57898962e-01 -1.40687978e+00 1.78036392e-01 -4.23061460e-01
-8.68392229e-01 1.27918196e+00 9.63085055e-01 -7.93946609e-02
2.96864480e-01 1.00458249e-01 5.46586588e-02 1.00902386e-01
-5.81806064e-01 -3.92845213e-01 3.55015665e-01 1.94135398e-01
4.81480621e-02 -2.34071851e-01 -4.62832749e-01 -3.58172576e-03
3.34865183e-01 4.01607007e-02 2.73190200e-01 9.17661846e-01
-5.45666456e-01 -1.14152360e+00 -3.89633179e-01 3.09146583e-01
-4.02147800e-01 4.63483453e-01 -4.77376617e-02 5.47211945e-01
2.46873483e-01 9.86740410e-01 1.19439736e-02 -1.82363153e-01
3.01405251e-01 -1.00478910e-01 5.19547939e-01 -4.68261212e-01
-4.22420204e-01 2.69857228e-01 -6.10649362e-02 -6.83649361e-01
-7.71299541e-01 -8.28224242e-01 -1.67976046e+00 2.07704529e-01
-5.44178545e-01 4.92492802e-02 6.63288713e-01 1.34107244e+00
1.46562219e-01 5.36430478e-01 7.44583726e-01 -1.16228592e+00
-4.72552210e-01 -4.60593045e-01 -1.59035355e-01 3.24276626e-01
7.48510957e-02 -7.97093868e-01 1.12300031e-01 1.18484534e-01] | [9.01196575164795, -0.17457488179206848] |
cca8910d-2fa9-4feb-933c-c2a84da0098a | cgodial-a-large-scale-benchmark-for-chinese | 2211.11617 | null | https://arxiv.org/abs/2211.11617v1 | https://arxiv.org/pdf/2211.11617v1.pdf | CGoDial: A Large-Scale Benchmark for Chinese Goal-oriented Dialog Evaluation | Practical dialog systems need to deal with various knowledge sources, noisy user expressions, and the shortage of annotated data. To better solve the above problems, we propose CGoDial, new challenging and comprehensive Chinese benchmark for multi-domain Goal-oriented Dialog evaluation. It contains 96,763 dialog sessions and 574,949 dialog turns totally, covering three datasets with different knowledge sources: 1) a slot-based dialog (SBD) dataset with table-formed knowledge, 2) a flow-based dialog (FBD) dataset with tree-formed knowledge, and a retrieval-based dialog (RBD) dataset with candidate-formed knowledge. To bridge the gap between academic benchmarks and spoken dialog scenarios, we either collect data from real conversations or add spoken features to existing datasets via crowd-sourcing. The proposed experimental settings include the combinations of training with either the entire training set or a few-shot training set, and testing with either the standard test set or a hard test subset, which can assess model capabilities in terms of general prediction, fast adaptability and reliable robustness. | ['Yongbin Li', 'Jian Sun', 'Zhongqi An', 'Zheng Cao', 'Yuchuan Wu', 'Bowen Li', 'Wanwei He', 'Yinpei Dai'] | 2022-11-21 | null | null | null | null | ['goal-oriented-dialog'] | ['natural-language-processing'] | [-4.08608347e-01 2.39400379e-02 3.24710719e-02 -5.71525633e-01
-6.60996199e-01 -7.64031172e-01 6.37932301e-01 -1.90634459e-01
-3.24478596e-01 1.24732924e+00 6.30333066e-01 -2.67425865e-01
-8.10413214e-04 -5.74491918e-01 2.96499282e-01 -4.47816938e-01
3.14083129e-01 1.15147042e+00 5.12243986e-01 -9.66065586e-01
7.66144395e-02 -1.19764388e-01 -8.87633145e-01 3.50402772e-01
1.23293126e+00 1.17058921e+00 2.46216729e-01 4.71232444e-01
-5.65473795e-01 1.17176199e+00 -9.96769488e-01 -6.42254353e-01
-7.13411942e-02 -5.88405490e-01 -1.16591787e+00 2.48243272e-01
-3.23048025e-01 -5.00590861e-01 -5.10973752e-01 6.32465601e-01
8.91530454e-01 6.28943920e-01 3.99582922e-01 -1.27194035e+00
-8.05247307e-01 6.38406157e-01 2.08927155e-01 9.14668962e-02
6.43934131e-01 3.36912096e-01 8.44687521e-01 -8.50955009e-01
8.18637550e-01 1.76870191e+00 1.35058165e-01 1.02936018e+00
-7.69868791e-01 -4.22643751e-01 -4.36725747e-03 1.34477049e-01
-9.45331395e-01 -6.82148516e-01 5.63496530e-01 -3.40133220e-01
7.78837025e-01 1.27667010e-01 -4.18594554e-02 1.41116273e+00
-4.62519854e-01 8.49304616e-01 1.12576807e+00 -2.34926388e-01
3.05672556e-01 7.38224149e-01 6.78759098e-01 3.90119672e-01
-6.44138217e-01 -1.12014525e-01 -6.29177153e-01 -3.59364480e-01
2.30432808e-01 -3.38610947e-01 -5.28993309e-01 1.13171846e-01
-1.19599080e+00 9.28963423e-01 5.17108776e-02 1.79803967e-01
4.30639684e-02 -1.00746846e+00 5.58200359e-01 6.62474692e-01
2.08982974e-01 5.08423984e-01 -8.22368503e-01 -4.16553736e-01
-1.05776437e-01 3.08828712e-01 1.63427711e+00 1.31669164e+00
5.21629393e-01 -4.89089526e-02 -6.55073225e-01 1.38852966e+00
1.11124411e-01 4.48469609e-01 9.80390966e-01 -8.91276240e-01
9.50233400e-01 9.07911837e-01 4.32695091e-01 -5.02743602e-01
-4.66938615e-01 5.68497717e-01 -7.46927619e-01 -5.08585453e-01
8.50109398e-01 -7.00589895e-01 -2.82620341e-01 1.88146043e+00
4.12847668e-01 -2.09414870e-01 8.77364814e-01 9.31182027e-01
1.67695940e+00 6.95992887e-01 -2.82019041e-02 -4.52670664e-01
1.37515450e+00 -1.40193737e+00 -9.79766309e-01 -8.50355551e-02
5.50929785e-01 -6.84910834e-01 1.44512272e+00 2.22756907e-01
-8.53835642e-01 -5.46563745e-01 -6.49696589e-01 -9.75812525e-02
-5.94065368e-01 9.71244350e-02 3.26127768e-01 6.64736390e-01
-8.17211986e-01 -6.20469190e-02 3.24574299e-03 -3.91576231e-01
-1.31832331e-01 -1.22016435e-02 -1.89693525e-01 -1.32438943e-01
-1.70767665e+00 9.93684232e-01 4.19766963e-01 -1.03672519e-01
-8.67569685e-01 -4.71979499e-01 -6.52924061e-01 1.46194220e-01
6.66365325e-01 -1.43264085e-01 1.57431841e+00 -4.36731845e-01
-1.97096241e+00 6.42112792e-01 1.01241164e-01 -1.53366655e-01
7.06039906e-01 -1.00828581e-01 -3.78981620e-01 -1.92002326e-01
-1.31420702e-01 4.63305771e-01 8.33354890e-02 -1.04784286e+00
-3.61397386e-01 -3.09648871e-01 2.65297979e-01 4.43663120e-01
-4.59936589e-01 7.58765778e-03 -4.90319133e-01 -2.76457310e-01
-2.62896448e-01 -9.65279698e-01 -1.36954151e-02 -5.30044973e-01
-7.43366718e-01 -6.46950066e-01 1.00164199e+00 -7.86188960e-01
1.15081763e+00 -2.12583017e+00 1.24785557e-01 -3.64632010e-01
5.58116883e-02 4.76989716e-01 -1.72298655e-01 6.35045648e-01
5.19931972e-01 -1.50900155e-01 8.07123333e-02 -3.86611819e-01
1.85128272e-01 3.72848988e-01 -3.70206773e-01 -4.56852704e-01
2.62884479e-02 7.10424840e-01 -9.09373879e-01 -6.90596819e-01
2.32193559e-01 -3.36042970e-01 -2.60574073e-01 8.82478058e-01
-5.93042493e-01 6.94686115e-01 -8.41634691e-01 7.03302383e-01
3.99154693e-01 -1.54153988e-01 4.27098513e-01 1.14775375e-01
9.76864845e-02 2.66751200e-01 -1.14730251e+00 1.70277166e+00
-4.06854182e-01 3.92292142e-01 3.19036603e-01 -6.06251597e-01
1.51464117e+00 6.05686843e-01 1.83883503e-01 -8.73646319e-01
2.58397311e-01 -8.13329592e-02 1.41017571e-01 -7.70346105e-01
6.21724308e-01 8.27391744e-02 -5.20328999e-01 4.50764030e-01
4.07265931e-01 -1.43841892e-01 2.21481800e-01 3.38031441e-01
1.12557840e+00 -5.77084243e-01 1.75861433e-01 -1.79028824e-01
8.81675184e-01 1.57682896e-01 6.53019428e-01 6.92982435e-01
-7.28376448e-01 2.29823396e-01 7.34948099e-01 -3.89612257e-01
-4.53191668e-01 -9.79590178e-01 5.79809323e-02 1.56939399e+00
3.39141190e-01 -2.69584000e-01 -5.34662187e-01 -7.86727071e-01
-1.71983153e-01 7.97660172e-01 -1.46669999e-01 -2.35622615e-01
-3.70145023e-01 -6.58616066e-01 8.38166833e-01 2.96853572e-01
1.26810396e+00 -1.40660357e+00 3.46942656e-02 2.32674658e-01
-6.24087632e-01 -1.41049445e+00 -5.05474627e-01 7.82091692e-02
-2.41255537e-01 -1.25861824e+00 -6.45611286e-01 -9.29258585e-01
-2.40424357e-04 1.25208080e-01 1.39129543e+00 -3.09208240e-02
1.01407103e-01 2.75751203e-01 -7.50365257e-01 -1.83205649e-01
-6.99119925e-01 1.47529468e-01 4.45600860e-02 -3.42123270e-01
6.26077056e-01 -1.53358668e-01 -7.70587027e-02 1.05440211e+00
-2.75037408e-01 -5.70387766e-02 1.90662280e-01 1.39041138e+00
-2.68059045e-01 -2.22893134e-01 1.22242248e+00 -5.65622449e-01
1.40452635e+00 -5.92841864e-01 -3.64889234e-01 7.31813252e-01
-2.52412677e-01 -1.88015655e-01 6.73466384e-01 -5.31112373e-01
-1.69645238e+00 -1.60219893e-01 -7.59415030e-02 -2.47365639e-01
-3.18543613e-01 4.08440232e-01 -5.73579609e-01 2.89260596e-01
7.53934264e-01 3.78765851e-01 1.90815896e-01 -4.56188172e-01
4.79071975e-01 1.31530654e+00 4.72396225e-01 -1.01789188e+00
1.48202956e-01 -3.28372627e-01 -8.78440320e-01 -9.39830422e-01
-7.14608848e-01 -4.41318631e-01 -6.39572203e-01 -2.66978383e-01
9.11955059e-01 -8.37431610e-01 -1.08357847e+00 6.44235790e-01
-1.29464996e+00 -7.65772820e-01 1.65199898e-02 2.10661262e-01
-2.77488738e-01 2.00438082e-01 -9.04990256e-01 -9.77871299e-01
-5.32797158e-01 -1.36358774e+00 7.67724037e-01 5.40354133e-01
-1.16593026e-01 -9.24426436e-01 7.42097944e-02 8.35085273e-01
4.09779638e-01 -2.57700205e-01 1.01647592e+00 -1.65755892e+00
-2.18332067e-01 4.82505606e-03 -1.82874858e-01 4.39357013e-01
2.70698875e-01 -3.51754576e-01 -9.46750164e-01 -3.72610800e-02
-1.07508600e-02 -1.34380484e+00 2.30999827e-01 -2.97170818e-01
6.26903236e-01 -2.70280480e-01 -7.95271434e-03 -3.43734980e-01
7.09916294e-01 7.74782121e-01 3.45819533e-01 -7.91412145e-02
1.29975438e-01 8.62127483e-01 9.15693343e-01 6.68439090e-01
7.41816878e-01 8.17765117e-01 -8.96125585e-02 5.67728281e-01
1.48808599e-01 -1.28045216e-01 1.38540551e-01 1.06162000e+00
3.61463130e-01 -6.91890895e-01 -1.10067213e+00 3.51905674e-01
-2.00755954e+00 -6.89733922e-01 2.71592319e-01 1.85834193e+00
1.20894849e+00 3.51358131e-02 3.79732996e-01 -2.58398175e-01
7.69028842e-01 1.90391302e-01 -7.47506499e-01 -1.16388217e-01
-2.47326925e-01 -3.76167774e-01 -2.73466468e-01 5.85688889e-01
-9.79172468e-01 1.27150142e+00 5.36596584e+00 9.14004147e-01
-8.01363766e-01 1.00270502e-01 8.24120045e-01 2.00502679e-01
4.33939137e-02 -2.29386181e-01 -9.79554951e-01 6.80770576e-01
8.29275727e-01 -2.81667978e-01 4.32781965e-01 1.05481207e+00
7.28214458e-02 -5.59273958e-02 -8.85859311e-01 8.92386734e-01
-1.02637641e-01 -1.06457770e+00 -1.57511622e-01 -3.99838984e-01
4.13949817e-01 -1.41614005e-01 -4.07718033e-01 1.11007941e+00
9.24873412e-01 -9.01245415e-01 3.53418104e-02 3.99516404e-01
3.92209679e-01 -1.65550366e-01 9.19711590e-01 9.23250318e-01
-9.15232241e-01 -6.83274642e-02 -4.19931233e-01 9.17237834e-04
2.19666228e-01 -2.82518584e-02 -1.15361297e+00 6.66148305e-01
5.36310196e-01 3.93235505e-01 -3.23367149e-01 5.47166288e-01
4.83109988e-02 4.14162606e-01 -2.00207949e-01 -5.93617558e-01
2.88609326e-01 -4.18864965e-01 3.27891499e-01 1.06791985e+00
-8.17262828e-02 5.08961022e-01 6.51880264e-01 6.59195244e-01
-9.21869949e-02 1.83597326e-01 -4.39665973e-01 -2.46058293e-02
1.11590850e+00 1.20674646e+00 -2.13214070e-01 -5.01982152e-01
-5.11252880e-01 6.42393172e-01 3.54423255e-01 3.99610847e-01
-5.21067023e-01 -4.03349251e-01 5.76974154e-01 -7.30026960e-01
-1.37897298e-01 -4.45084609e-02 6.96128458e-02 -1.22356200e+00
-2.07726642e-01 -1.12170756e+00 5.84996223e-01 -6.97189450e-01
-1.67135537e+00 9.36942577e-01 -1.02523272e-03 -1.00020635e+00
-4.95859832e-01 -5.34463286e-01 -8.89062881e-01 9.18428481e-01
-1.04333138e+00 -8.80123734e-01 -9.11429942e-01 9.26769018e-01
1.19754803e+00 -7.76674449e-01 1.01214445e+00 4.13045436e-01
-8.40874434e-01 6.56121612e-01 1.28427625e-01 5.83804071e-01
1.08726227e+00 -1.14999819e+00 -7.55466819e-02 -4.64449376e-02
-4.74898607e-01 2.36979023e-01 5.22544205e-01 -4.04989123e-01
-1.24683845e+00 -8.92969728e-01 7.46690929e-01 -6.42287970e-01
6.68477833e-01 -4.08447474e-01 -1.13838899e+00 5.63269556e-01
2.21935838e-01 -4.28977549e-01 7.39057004e-01 1.08161807e-01
2.34639384e-02 6.31778091e-02 -1.24502087e+00 4.78611797e-01
8.90610874e-01 -3.60511571e-01 -9.48614120e-01 6.07195258e-01
1.07219231e+00 -8.23939383e-01 -1.05661118e+00 2.43429556e-01
2.60254532e-01 -9.13827062e-01 7.41982579e-01 -9.92817223e-01
3.56964111e-01 3.42976481e-01 -4.41589117e-01 -1.42654812e+00
3.54891777e-01 -5.91442823e-01 9.10397619e-02 1.83468199e+00
4.20241326e-01 -4.46691006e-01 7.87152767e-01 1.07964599e+00
-2.88035393e-01 -8.16623807e-01 -1.01005089e+00 -6.69730425e-01
2.13605419e-01 1.64283007e-01 7.66782820e-01 1.15093601e+00
5.71189225e-01 8.99582624e-01 -7.68288314e-01 -4.63087380e-01
-1.39156073e-01 1.99328914e-01 1.19368756e+00 -1.07154465e+00
-2.85466556e-02 -2.15059415e-01 1.16265804e-01 -1.43883300e+00
4.24535036e-01 -2.75402814e-01 2.30389938e-01 -1.07393944e+00
-1.00462377e-01 -7.49589682e-01 3.20615917e-01 3.38440478e-01
-3.34270477e-01 -7.02245712e-01 1.46649778e-01 2.65658677e-01
-8.63343120e-01 1.16185224e+00 1.42381692e+00 -1.34722024e-01
-6.28579736e-01 2.99195170e-01 -4.29544091e-01 4.58069682e-01
6.13800466e-01 1.80001676e-01 -7.62021244e-01 -1.37422517e-01
-6.12987280e-01 9.92244244e-01 -3.73931006e-02 -6.03487492e-01
5.01776040e-01 -3.01617950e-01 -2.70809799e-01 -4.33716178e-01
7.81503677e-01 -4.78485465e-01 -5.92137039e-01 2.20217183e-01
-7.66281366e-01 -2.77158543e-02 1.50217175e-01 4.80171204e-01
-4.27646518e-01 -2.66783416e-01 6.32160485e-01 -3.28728646e-01
-1.06774473e+00 1.32475331e-01 -2.27858722e-01 6.43227816e-01
9.34551537e-01 3.18770677e-01 -1.15001285e+00 -8.87058020e-01
-8.08892608e-01 9.43731725e-01 -5.38054481e-02 8.23543847e-01
4.45399910e-01 -1.17527878e+00 -5.82390904e-01 -1.04484916e-01
2.26145580e-01 -8.78456160e-02 4.24049765e-01 4.89925057e-01
-1.20527707e-01 6.05651557e-01 -3.65086883e-01 -3.23714525e-01
-1.09591472e+00 3.77567410e-01 3.49121571e-01 -4.08640385e-01
-1.76676586e-01 8.89070928e-01 -9.52351652e-03 -1.22145176e+00
6.64123297e-01 9.27483365e-02 -4.68760699e-01 1.56279474e-01
4.74792659e-01 3.97313833e-01 -2.17619210e-01 -4.78710413e-01
-2.72125661e-01 -2.99263388e-01 -6.60838410e-02 -1.54147923e-01
6.11141443e-01 -3.49113494e-01 1.93760589e-01 5.49793243e-01
8.29700232e-01 -3.75792503e-01 -9.35200751e-01 -8.19042861e-01
6.80960491e-02 -2.08027124e-01 -6.02054358e-01 -1.22832203e+00
-7.54400134e-01 8.01402569e-01 3.74673128e-01 4.30570632e-01
6.77035153e-01 1.41992047e-01 8.47188950e-01 1.11321139e+00
3.63640904e-01 -1.19326544e+00 6.36045456e-01 1.12272871e+00
1.05762494e+00 -1.82192814e+00 -5.89998841e-01 -3.73029798e-01
-1.47021735e+00 9.26706553e-01 1.41985023e+00 4.45995480e-01
7.64623642e-01 -1.04627244e-01 4.49112564e-01 3.92359905e-02
-1.22294772e+00 -2.13946193e-01 -7.17567280e-02 5.35217345e-01
2.20419303e-01 -1.51397943e-01 -1.35622263e-01 1.27039158e+00
-3.69530767e-01 -1.62843317e-01 1.80152982e-01 7.32046425e-01
-5.34395516e-01 -9.83420253e-01 -2.87298113e-02 4.14546698e-01
3.08845341e-01 -5.67805097e-02 -8.66088033e-01 9.89687562e-01
-3.86314452e-01 1.61742461e+00 -1.36221647e-01 -7.94070601e-01
5.53051293e-01 5.37422419e-01 -3.51824760e-01 -5.32031059e-01
-6.86994374e-01 -2.45616287e-01 7.72100866e-01 -9.38009396e-02
-1.22461163e-01 -2.60999799e-01 -9.04145360e-01 -4.47461307e-01
-5.77652693e-01 6.73676431e-01 9.08605754e-02 1.22693932e+00
2.99174905e-01 2.21417248e-01 6.71880245e-01 -3.39948237e-01
-1.04857135e+00 -1.59675729e+00 -4.75710452e-01 5.31365037e-01
-8.43102187e-02 -8.44898462e-01 -2.92954385e-01 -1.97547823e-01] | [12.810440063476562, 7.965880870819092] |
3d5421de-11c1-456f-bb78-d4752ed003db | high-quality-rgb-d-reconstruction-via-multi | 2210.12202 | null | https://arxiv.org/abs/2210.12202v1 | https://arxiv.org/pdf/2210.12202v1.pdf | High-Quality RGB-D Reconstruction via Multi-View Uncalibrated Photometric Stereo and Gradient-SDF | Fine-detailed reconstructions are in high demand in many applications. However, most of the existing RGB-D reconstruction methods rely on pre-calculated accurate camera poses to recover the detailed surface geometry, where the representation of a surface needs to be adapted when optimizing different quantities. In this paper, we present a novel multi-view RGB-D based reconstruction method that tackles camera pose, lighting, albedo, and surface normal estimation via the utilization of a gradient signed distance field (gradient-SDF). The proposed method formulates the image rendering process using specific physically-based model(s) and optimizes the surface's quantities on the actual surface using its volumetric representation, as opposed to other works which estimate surface quantities only near the actual surface. To validate our method, we investigate two physically-based image formation models for natural light and point light source applications. The experimental results on synthetic and real-world datasets demonstrate that the proposed method can recover high-quality geometry of the surface more faithfully than the state-of-the-art and further improves the accuracy of estimated camera poses. | ['Daniel Cremers', 'Xingxing Zuo', 'Bjoern Haefner', 'Lu Sang'] | 2022-10-21 | null | null | null | null | ['rgb-d-reconstruction'] | ['computer-vision'] | [ 3.51188719e-01 -2.95264840e-01 4.97024506e-01 -3.75926167e-01
-5.39920509e-01 -3.09101611e-01 3.91571462e-01 -8.55762139e-03
-1.13597028e-01 4.33448255e-01 -1.53051674e-01 2.23487407e-01
9.52308178e-02 -9.97881711e-01 -6.72394216e-01 -5.44840753e-01
5.24469376e-01 5.73694289e-01 4.64348316e-01 -3.12648863e-01
5.21781743e-01 8.34555268e-01 -1.77865779e+00 -1.79809198e-01
7.42497087e-01 1.05808783e+00 2.94943601e-01 4.33351219e-01
-1.97421148e-01 2.48012900e-01 -1.96170226e-01 -1.90840736e-01
4.89253730e-01 -2.48410314e-01 -1.78309277e-01 4.75748897e-01
4.68262225e-01 -4.07877028e-01 1.69360906e-01 9.86061513e-01
3.36829245e-01 -2.26314384e-02 6.39340401e-01 -7.25897968e-01
-1.50585100e-01 -8.41808856e-01 -8.84749472e-01 -5.03346205e-01
8.09192538e-01 2.73629446e-02 4.45699066e-01 -1.11987591e+00
6.28215492e-01 1.10205293e+00 6.76474273e-01 2.43274570e-01
-1.04390252e+00 -4.45064366e-01 6.03180863e-02 -7.44603649e-02
-1.61230969e+00 -2.78505057e-01 1.33902454e+00 -4.31258887e-01
4.89069581e-01 2.87853688e-01 9.03679073e-01 4.90655959e-01
1.32032216e-01 9.53349471e-02 1.46125185e+00 -5.25466144e-01
3.24943602e-01 5.73153719e-02 -2.57426649e-01 8.30469370e-01
2.56885290e-01 6.93203211e-02 -6.83361709e-01 -2.43304461e-01
1.21693599e+00 1.42882392e-01 -6.78400815e-01 -8.27531993e-01
-1.10849559e+00 4.01620686e-01 2.14271680e-01 -5.85434921e-02
-4.48641181e-01 -6.47111908e-02 -2.15359896e-01 -3.04998964e-01
6.79765940e-01 -1.22624904e-01 -2.62873113e-01 -1.00447632e-01
-7.61177480e-01 2.41238065e-02 7.28155196e-01 8.11388314e-01
1.41696250e+00 -1.18067048e-01 5.25611401e-01 5.62949836e-01
8.53368282e-01 8.61669719e-01 5.74168488e-02 -1.07617760e+00
3.11062306e-01 7.99934030e-01 4.51129884e-01 -1.20844948e+00
-5.43016791e-02 -8.98117349e-02 -7.12581992e-01 5.08538008e-01
1.23062819e-01 3.56981844e-01 -6.94255590e-01 9.76077259e-01
8.49279165e-01 3.75435770e-01 2.82417424e-02 9.88168001e-01
6.95876122e-01 5.86016297e-01 -7.65003085e-01 -4.05944914e-01
1.00593483e+00 -4.78987455e-01 -6.46966457e-01 -2.39686206e-01
-7.86353499e-02 -9.43601310e-01 1.09937656e+00 6.16113901e-01
-1.03912270e+00 -3.47458303e-01 -9.14408982e-01 -1.64629862e-01
1.66968524e-01 1.85806692e-01 2.85361320e-01 5.10716081e-01
-9.19778228e-01 5.12537062e-01 -9.10073519e-01 -2.68421978e-01
-7.54543990e-02 7.71345347e-02 -3.85118574e-01 -3.01518977e-01
-6.77506745e-01 7.78503001e-01 -1.50975689e-01 3.02725822e-01
-6.65399432e-01 -6.90399289e-01 -7.23845363e-01 -2.44706705e-01
2.21722692e-01 -8.68623137e-01 7.62305498e-01 -7.51697600e-01
-2.02890229e+00 8.63509119e-01 -3.39899421e-01 5.08939028e-01
4.69313949e-01 -1.39105260e-01 -5.11497678e-03 3.68185014e-01
-1.64242893e-01 -1.16146378e-01 8.02999914e-01 -1.96725619e+00
-1.26604646e-01 -6.38919592e-01 2.67743468e-02 4.85002935e-01
6.10011853e-02 -5.08936107e-01 -6.15346253e-01 -8.76023769e-02
6.11617923e-01 -5.85798264e-01 -1.60336018e-01 6.38885140e-01
-2.46227071e-01 3.21238548e-01 8.11538756e-01 -5.38673222e-01
7.72104740e-01 -2.01765347e+00 2.58249134e-01 2.82020390e-01
-1.21497773e-01 -2.47819535e-02 2.50826031e-01 5.19941449e-01
3.80660564e-01 -2.41429031e-01 -5.67380130e-01 -6.65544212e-01
-4.20162886e-01 3.60780895e-01 5.70295267e-02 9.11526322e-01
-2.92217076e-01 1.38189659e-01 -7.26112187e-01 -5.46698749e-01
6.98822320e-01 1.06979012e+00 -4.34470296e-01 4.75644022e-01
-2.16919303e-01 7.46144295e-01 -6.39759004e-01 6.89941764e-01
1.05239809e+00 -2.05341354e-02 2.16401294e-02 -5.75430810e-01
-4.42789674e-01 -2.34422043e-01 -1.63578916e+00 2.11524177e+00
-7.77929962e-01 1.71456441e-01 4.46495086e-01 -3.76436830e-01
1.16233611e+00 -3.34399119e-02 5.16437352e-01 -6.02461278e-01
7.91148618e-02 3.56912971e-01 -7.40921974e-01 -4.43762094e-01
2.74811745e-01 -4.01527166e-01 4.64275867e-01 1.33864045e-01
-7.09898055e-01 -8.29571009e-01 -4.59302723e-01 -1.00380257e-01
5.65105617e-01 7.51175225e-01 3.07350993e-01 -1.16054520e-01
9.36553001e-01 -9.14616436e-02 6.18554473e-01 2.12617815e-02
3.88462245e-01 1.02517080e+00 -1.35765865e-01 -3.42164487e-01
-1.08649123e+00 -1.04029882e+00 -2.83394933e-01 5.22349030e-02
7.33938932e-01 -2.00172633e-01 -7.57126272e-01 -6.41522259e-02
2.97730360e-02 6.57694936e-01 -3.29118669e-01 3.45641673e-01
-5.39800823e-01 -5.55677652e-01 -2.66477764e-01 -7.17517138e-02
6.57890379e-01 -3.91634941e-01 -9.51683939e-01 -1.50309995e-01
-1.52184054e-01 -1.07725394e+00 -1.59115896e-01 -5.79679906e-01
-1.07984269e+00 -1.29765773e+00 -6.85522616e-01 -2.30182469e-01
9.87049699e-01 4.93327856e-01 1.09891427e+00 1.65110335e-01
-1.61651507e-01 8.41159582e-01 -3.36604476e-01 -1.52041838e-01
-1.74615070e-01 -5.58048427e-01 -1.90007389e-01 5.22329271e-01
-9.45997983e-02 -7.04322994e-01 -9.20540750e-01 6.37171030e-01
-9.49912310e-01 4.49710608e-01 3.49588305e-01 2.55916864e-01
1.08986175e+00 -1.68793693e-01 -2.41462633e-01 -6.47825241e-01
-6.19086390e-03 -2.28226021e-01 -9.75976944e-01 2.00217217e-01
-6.72343791e-01 -3.26545760e-02 3.62563014e-01 -4.57737362e-03
-1.37168133e+00 3.72229725e-01 -2.84359902e-02 -6.11322880e-01
-1.19671360e-01 9.33378264e-02 -3.44584018e-01 -3.81770879e-01
3.54829252e-01 4.57457542e-01 3.68587039e-02 -8.79860520e-01
2.69912947e-02 6.14484370e-01 2.32497826e-01 -5.92793643e-01
9.52606559e-01 1.22422969e+00 5.66053331e-01 -1.11114836e+00
-6.69067800e-01 -6.15037799e-01 -7.58923173e-01 -5.84601343e-01
6.85560644e-01 -9.21085536e-01 -8.64180446e-01 5.46077192e-01
-1.38817441e+00 -1.80275321e-01 -1.23910956e-01 5.07921159e-01
-5.76182306e-01 7.77323604e-01 -1.54060155e-01 -9.53521192e-01
-1.41206428e-01 -1.33402205e+00 1.80074930e+00 5.74078299e-02
4.19575095e-01 -9.99865234e-01 4.01118815e-01 4.05397266e-01
2.26292863e-01 6.97324753e-01 4.42147046e-01 9.45058465e-01
-1.12944865e+00 3.62752117e-02 -1.10610329e-01 2.66088873e-01
1.54349431e-01 2.04050213e-01 -1.03001273e+00 -1.49607852e-01
4.28898782e-01 1.41692445e-01 2.38575682e-01 1.26004249e-01
8.67933512e-01 7.70962089e-02 -1.06069960e-01 1.11556542e+00
2.28919911e+00 -1.62982121e-01 7.38604844e-01 3.88909280e-01
9.10284221e-01 5.22816479e-01 8.10792804e-01 7.36855745e-01
6.52371228e-01 1.03255415e+00 1.11159539e+00 4.81703365e-03
-1.21657945e-01 -1.81040376e-01 3.61319810e-01 1.03064227e+00
-5.46895027e-01 -1.72745124e-01 -8.20409536e-01 1.04111783e-01
-1.53890300e+00 -4.22641128e-01 -7.85044789e-01 2.64637542e+00
5.80583751e-01 -3.41233164e-01 -4.15365338e-01 1.50722057e-01
5.24979949e-01 8.34910497e-02 -4.88820285e-01 -1.68294191e-01
-1.76727446e-03 1.30440891e-01 6.20596230e-01 9.70713556e-01
-2.86433131e-01 6.30814314e-01 5.52679157e+00 5.19748092e-01
-1.18103909e+00 1.87360853e-01 1.10593230e-01 7.64673129e-02
-7.64914095e-01 2.58541137e-01 -7.20443666e-01 3.08621496e-01
4.95270371e-01 1.63377509e-01 4.40259844e-01 6.31950796e-01
4.37458277e-01 -7.19741642e-01 -7.64276206e-01 1.31179178e+00
3.24083418e-01 -9.79869127e-01 4.20626551e-02 9.66956019e-02
7.21889615e-01 -1.24974335e-02 -3.25874269e-01 -7.13505089e-01
-2.52608806e-01 -5.06914020e-01 7.51685679e-01 1.06085563e+00
1.06934166e+00 -4.57745463e-01 5.29860735e-01 6.38203919e-01
-1.19060123e+00 5.25021374e-01 -4.29153800e-01 -3.06712510e-03
5.19326746e-01 1.06276929e+00 -5.17977774e-01 8.37014616e-01
5.80741107e-01 8.01454365e-01 -4.37540621e-01 7.41659880e-01
-4.03970808e-01 1.34107739e-01 -5.19776702e-01 2.24041089e-01
-2.25783944e-01 -8.63256395e-01 4.93006170e-01 5.88194311e-01
5.12910187e-01 3.70746613e-01 7.85580203e-02 8.56696308e-01
3.41985404e-01 3.40511829e-01 -5.23917079e-01 6.50310099e-01
1.29161566e-01 1.30557311e+00 -5.95052481e-01 2.40829471e-03
-5.13690710e-01 1.11214817e+00 1.23044051e-01 3.26596797e-01
-6.37489915e-01 1.07993841e-01 3.84372294e-01 7.74833798e-01
-7.02597052e-02 -6.19578898e-01 -2.70173460e-01 -1.35075808e+00
2.48921752e-01 -2.34254748e-01 -3.38345200e-01 -1.32732821e+00
-9.27474678e-01 4.48299348e-01 4.37424332e-02 -1.50442123e+00
1.14251360e-01 -5.81502140e-01 -3.03596050e-01 1.11046374e+00
-1.95224071e+00 -1.15404391e+00 -8.57472062e-01 6.83726966e-01
3.97344917e-01 4.30009365e-01 7.43844807e-01 2.10648239e-01
-3.02265406e-01 -2.93340862e-01 2.84844041e-01 -4.15035546e-01
4.65506136e-01 -8.52080464e-01 -4.94939908e-02 7.45908558e-01
-2.58109629e-01 4.87673730e-01 5.44720829e-01 -7.22347319e-01
-1.95835590e+00 -6.81886256e-01 3.58603597e-01 -2.66735673e-01
5.23036756e-02 -2.20111638e-01 -7.03697681e-01 4.42638129e-01
-1.50328036e-02 3.75732422e-01 3.03816617e-01 -4.75217074e-01
1.76172983e-02 -3.95563990e-01 -1.44494462e+00 1.23575993e-01
9.57794070e-01 -4.88082498e-01 -2.89582253e-01 2.12844595e-01
1.39439434e-01 -7.81346083e-01 -1.01257300e+00 4.38747317e-01
6.53500617e-01 -1.61245120e+00 1.12272370e+00 4.62097585e-01
2.39909679e-01 -8.35085869e-01 -3.50606829e-01 -1.20262694e+00
1.48757264e-01 -4.32274252e-01 -1.28886467e-02 1.09915662e+00
-2.35360295e-01 -7.11266577e-01 7.92237759e-01 6.15545690e-01
-1.61135361e-01 -8.11716318e-01 -9.13080871e-01 -4.02328789e-01
-6.01246178e-01 -3.38868290e-01 6.14967823e-01 7.99113810e-01
-7.52554655e-01 2.25299336e-02 -7.10326284e-02 6.95706606e-01
1.09333098e+00 4.63449508e-01 1.09820473e+00 -1.34950435e+00
-1.55970275e-01 1.65759042e-01 -4.84506220e-01 -1.12462282e+00
-9.91811976e-02 -4.61537689e-01 4.31402326e-02 -1.78413856e+00
1.30320983e-02 -7.31026709e-01 4.53059614e-01 -1.79388613e-01
9.78895873e-02 2.34229863e-01 -9.85257104e-02 4.42081273e-01
-1.10809326e-01 7.18786538e-01 1.44146538e+00 3.49822909e-01
-5.67918718e-02 -1.06148437e-01 -1.40131071e-01 1.04815745e+00
3.60494107e-01 -2.69984275e-01 -4.08458173e-01 -7.92413235e-01
5.83565831e-01 1.73880368e-01 4.21136469e-01 -9.66115713e-01
5.99571690e-02 -4.67476636e-01 1.82860017e-01 -8.00234139e-01
8.12923551e-01 -1.36544859e+00 7.23071516e-01 1.83281511e-01
3.57499093e-01 -4.00208235e-02 -1.75976798e-01 7.42295027e-01
-4.39607650e-02 -3.00883412e-01 9.84910429e-01 -4.76257056e-01
-4.68188316e-01 3.75978768e-01 2.92468220e-01 -1.66713610e-01
1.08883667e+00 -5.62972963e-01 1.02495037e-01 -3.28797042e-01
-2.42690623e-01 -3.13642085e-01 1.37122262e+00 -2.35916629e-01
1.14002526e+00 -1.21196055e+00 -5.49440682e-01 4.82967079e-01
1.47256836e-01 7.41673589e-01 1.31470934e-01 5.95823765e-01
-1.32779193e+00 -1.29918233e-01 1.26009732e-01 -1.00547647e+00
-1.10225296e+00 2.19859079e-01 5.95939219e-01 1.55423924e-01
-7.30864942e-01 4.86476034e-01 1.90073475e-01 -4.36080247e-01
-2.50641167e-01 -3.28741610e-01 8.49731565e-02 -3.46214235e-01
1.63850978e-01 6.34880781e-01 1.58308551e-01 -1.04979575e+00
-3.66662681e-01 1.68266201e+00 6.86891317e-01 -2.22368017e-01
1.41070640e+00 -4.47908342e-01 -2.78281718e-01 5.65697908e-01
1.11679399e+00 4.01108414e-01 -1.41487873e+00 -3.01271044e-02
-5.39080501e-01 -1.10354865e+00 4.78641599e-01 -2.95232654e-01
-1.18286681e+00 1.01878381e+00 5.22428751e-01 -2.52305269e-01
1.17633867e+00 -2.47275665e-01 7.34900534e-01 4.33313362e-02
1.06100607e+00 -8.37147415e-01 -2.13600561e-01 -2.17764289e-03
9.62908268e-01 -1.04060686e+00 6.42963767e-01 -9.38152552e-01
-2.66150355e-01 1.32723725e+00 4.38837618e-01 -1.22232549e-01
8.27988207e-01 9.48199555e-02 9.79436375e-03 -3.62648487e-01
-4.20688689e-02 -2.04558987e-02 8.74625370e-02 4.99636710e-01
2.16910496e-01 -1.82571515e-01 -1.18932843e-01 -2.34470889e-01
2.19433278e-01 6.52647614e-02 6.93248034e-01 9.13958728e-01
-3.97033632e-01 -1.15450335e+00 -8.37541759e-01 -2.06055176e-02
5.27674519e-02 1.34967178e-01 -1.46444201e-01 7.01038897e-01
1.72748163e-01 7.36467779e-01 1.00474678e-01 4.16021347e-02
6.24301851e-01 -3.10297579e-01 7.63858438e-01 -7.91435301e-01
-1.14460871e-01 1.74508825e-01 -1.40167102e-01 -8.38911057e-01
-1.02802432e+00 -7.64354825e-01 -1.26010847e+00 -1.34024974e-02
-4.77697909e-01 -1.12337075e-01 1.24289465e+00 5.94266713e-01
3.56653869e-01 -4.44782786e-02 9.72692132e-01 -1.24611187e+00
-1.09272793e-01 -6.26376450e-01 -7.41003335e-01 5.06254017e-01
3.60535026e-01 -9.92535949e-01 -6.97490275e-01 4.76000551e-03] | [9.319488525390625, -2.918010711669922] |
800f50c8-2106-4e11-8910-59f8af7d0670 | towards-multimodal-multitask-scene | 2209.13156 | null | https://arxiv.org/abs/2209.13156v1 | https://arxiv.org/pdf/2209.13156v1.pdf | Towards Multimodal Multitask Scene Understanding Models for Indoor Mobile Agents | The perception system in personalized mobile agents requires developing indoor scene understanding models, which can understand 3D geometries, capture objectiveness, analyze human behaviors, etc. Nonetheless, this direction has not been well-explored in comparison with models for outdoor environments (e.g., the autonomous driving system that includes pedestrian prediction, car detection, traffic sign recognition, etc.). In this paper, we first discuss the main challenge: insufficient, or even no, labeled data for real-world indoor environments, and other challenges such as fusion between heterogeneous sources of information (e.g., RGB images and Lidar point clouds), modeling relationships between a diverse set of outputs (e.g., 3D object locations, depth estimation, and human poses), and computational efficiency. Then, we describe MMISM (Multi-modality input Multi-task output Indoor Scene understanding Model) to tackle the above challenges. MMISM considers RGB images as well as sparse Lidar points as inputs and 3D object detection, depth completion, human pose estimation, and semantic segmentation as output tasks. We show that MMISM performs on par or even better than single-task models; e.g., we improve the baseline 3D object detection results by 11.7% on the benchmark ARKitScenes dataset. | ['Jian Zhang', 'Ali Farhadi', 'Hanlin Goh', 'Yao-Hung Hubert Tsai'] | 2022-09-27 | null | null | null | null | ['traffic-sign-recognition', 'depth-completion'] | ['computer-vision', 'computer-vision'] | [-4.37438153e-02 -1.53890759e-01 1.01512820e-01 -5.51280618e-01
-8.02179217e-01 -5.64552844e-01 4.49727714e-01 -4.33137976e-02
-5.24246573e-01 4.20436293e-01 -1.32667109e-01 -3.79675299e-01
2.63235658e-01 -7.79453993e-01 -1.00717592e+00 -6.69236958e-01
3.39229703e-01 9.22343016e-01 5.34124792e-01 -1.50853753e-01
1.65299594e-01 4.86344248e-01 -1.99272656e+00 9.50702354e-02
1.01447117e+00 1.26415408e+00 5.93408525e-01 9.41031158e-01
-3.63860764e-02 4.63262558e-01 -3.10770512e-01 -2.58687407e-01
1.97984233e-01 3.44044119e-01 -3.75689298e-01 3.51932675e-01
6.45622551e-01 -7.59969532e-01 -2.38890171e-01 1.02946675e+00
4.15477931e-01 1.53301999e-01 6.04506612e-01 -1.70566058e+00
-4.09311533e-01 -1.67706579e-01 -4.82768476e-01 -7.42885321e-02
5.55317581e-01 5.05004764e-01 4.69720274e-01 -1.24773633e+00
4.92866457e-01 1.56161749e+00 4.49816406e-01 4.66104418e-01
-7.98884213e-01 -4.40857768e-01 5.76488197e-01 6.80585563e-01
-1.24568737e+00 -4.09537226e-01 4.30835098e-01 -4.83549923e-01
9.80851471e-01 2.92236030e-01 7.68140018e-01 1.41352594e+00
-1.16699340e-03 1.16386032e+00 1.15020096e+00 4.52906154e-02
3.29401374e-01 3.25096756e-01 3.68412495e-01 7.78865576e-01
3.53122920e-01 2.26445511e-01 -4.75633055e-01 3.56759250e-01
6.05776131e-01 2.21767083e-01 1.32883996e-01 -5.56435943e-01
-1.21721911e+00 4.22083259e-01 5.81510246e-01 -4.42631781e-01
-3.50350231e-01 1.40485734e-01 1.42664984e-01 -1.34155691e-01
2.81982958e-01 -2.48438388e-01 -6.96643829e-01 -1.57427222e-01
-4.07837838e-01 5.86799443e-01 5.37116230e-01 1.57048452e+00
9.78823841e-01 -2.53765851e-01 1.44113988e-01 5.19825697e-01
6.39082909e-01 1.23754895e+00 -2.71368716e-02 -9.81396139e-01
8.31065655e-01 6.03994191e-01 3.27360719e-01 -7.30095506e-01
-7.24408090e-01 -1.77285761e-01 -5.87376833e-01 2.86003351e-01
6.97317600e-01 -4.88462821e-02 -1.25379324e+00 1.44648755e+00
6.71006680e-01 1.48911893e-01 2.18021814e-02 1.25300431e+00
1.27781641e+00 6.64905012e-01 3.50467376e-02 2.68817067e-01
1.55699658e+00 -1.12143624e+00 -3.86421114e-01 -8.28524768e-01
3.78752351e-01 -4.81855243e-01 1.05839109e+00 3.98143381e-01
-9.32357907e-01 -7.13310301e-01 -6.74033463e-01 -5.28557062e-01
-8.56336772e-01 2.69485533e-01 7.65434682e-01 8.15944910e-01
-8.59548330e-01 -2.88441122e-01 -1.07491207e+00 -6.73283577e-01
4.39843744e-01 3.01710337e-01 -3.24816823e-01 -5.76550186e-01
-6.95697010e-01 9.76426244e-01 1.61713347e-01 1.20489351e-01
-1.14047575e+00 -6.80124760e-01 -1.00210035e+00 -2.65786737e-01
8.22462976e-01 -1.07047975e+00 1.30172563e+00 -8.90094414e-02
-1.30615962e+00 8.69374990e-01 -4.20306206e-01 -3.22411686e-01
6.91502988e-01 -4.71874565e-01 2.66668200e-02 -2.64626592e-01
2.48426110e-01 9.99858379e-01 3.53763372e-01 -1.50805712e+00
-1.10616863e+00 -1.07537305e+00 1.36270925e-01 5.14403045e-01
3.10499638e-01 -3.86956066e-01 -9.87831175e-01 7.89717808e-02
4.73764449e-01 -1.09798110e+00 -5.61674416e-01 1.95080936e-01
-6.53808951e-01 -1.62053615e-01 8.80395472e-01 -6.97903931e-01
3.08651179e-01 -1.95522976e+00 4.86220531e-02 1.24106660e-01
1.90658450e-01 -1.83363795e-01 5.22983633e-02 -1.25611633e-01
5.52286983e-01 -2.38072559e-01 -1.23189591e-01 -9.15385962e-01
3.57917547e-01 4.54498976e-01 -1.93991996e-02 4.17995900e-01
-1.30433753e-01 9.75540280e-01 -8.36479247e-01 -5.21108091e-01
7.79891014e-01 5.33290684e-01 -6.09190881e-01 1.04025509e-02
-3.44427884e-01 6.01790547e-01 -6.21223152e-01 1.18461740e+00
9.84801054e-01 -1.58945382e-01 -2.35968530e-01 -1.70028597e-01
-2.63720512e-01 1.42884791e-01 -1.42376280e+00 1.89467824e+00
-5.61258376e-01 5.38006723e-01 -4.64646742e-02 -6.26760066e-01
7.26043522e-01 -1.06723502e-01 3.64117503e-01 -8.24941039e-01
2.24650174e-01 3.08358401e-01 -4.39125866e-01 -5.59427142e-01
6.84152603e-01 3.47398490e-01 -3.18179637e-01 2.18045842e-02
-2.26527199e-01 -3.61163080e-01 -5.61728887e-02 -5.37566543e-02
7.39226520e-01 2.90050387e-01 1.91399485e-01 2.10709959e-01
3.19629073e-01 4.72148806e-01 4.40839112e-01 7.89375126e-01
-5.46527088e-01 4.98707324e-01 5.97269237e-02 -4.29815322e-01
-9.72799897e-01 -1.23841274e+00 -3.80690321e-02 9.44325209e-01
9.00927842e-01 -6.84824884e-02 -8.05514276e-01 -5.69357932e-01
2.98761964e-01 6.45103693e-01 -4.37471539e-01 6.90036491e-02
-5.16548038e-01 -6.98708296e-01 1.07681520e-01 7.72649646e-01
7.15342879e-01 -7.39067376e-01 -9.94839251e-01 -5.86609654e-02
-3.21722895e-01 -1.88379097e+00 -1.81883320e-01 1.74566582e-01
-6.41330600e-01 -1.11631978e+00 -4.75020319e-01 -5.63548088e-01
4.89646852e-01 7.08099723e-01 1.05742955e+00 -3.81869048e-01
-3.07535082e-01 6.21129155e-01 -9.85217094e-02 -7.05811799e-01
1.56350642e-01 -3.81589420e-02 2.16799974e-01 -1.11988172e-01
6.19127572e-01 -2.15338066e-01 -7.38802612e-01 5.47384918e-01
-3.99508774e-01 3.42111468e-01 5.62673628e-01 3.00848991e-01
8.69495392e-01 -2.70286083e-01 5.46588823e-02 -4.72046077e-01
-6.52740523e-02 -3.27873349e-01 -7.49523342e-01 1.85100362e-01
-2.27283746e-01 -3.93416792e-01 2.42211342e-01 -2.17065349e-01
-1.12427807e+00 3.89660865e-01 -2.89968133e-01 -3.95438284e-01
-8.03407609e-01 -6.98504820e-02 -7.17260718e-01 7.07823783e-02
4.86912727e-01 2.60829777e-01 -1.67571276e-01 -4.16646898e-01
6.72415137e-01 6.87583625e-01 7.16765285e-01 -3.97648424e-01
7.33530998e-01 7.84079194e-01 -3.58255841e-02 -9.14816678e-01
-6.75813735e-01 -7.10776627e-01 -7.80549049e-01 -3.49422127e-01
1.11794186e+00 -1.20499468e+00 -1.09229755e+00 6.34565711e-01
-1.41056716e+00 -3.67212355e-01 -1.24840744e-01 5.68134308e-01
-8.61373663e-01 3.02669946e-02 -3.02668393e-01 -1.06031656e+00
1.21977299e-01 -1.54703367e+00 1.56672001e+00 3.27108890e-01
1.73287511e-01 -5.53953409e-01 -3.83755058e-01 8.61629009e-01
-3.75252552e-02 2.16653049e-01 5.43931723e-01 -2.60835797e-01
-1.40284920e+00 -3.70624699e-02 -4.32717055e-01 -1.94067359e-01
-1.63636133e-01 -3.91272604e-01 -1.07579124e+00 1.64300159e-01
-2.48330966e-01 -2.09683359e-01 8.97624433e-01 6.64059460e-01
1.07601559e+00 4.95349728e-02 -6.13128841e-01 7.04173446e-01
9.94074941e-01 3.36778939e-01 5.39208710e-01 3.94323498e-01
9.11855578e-01 7.70500481e-01 9.65727687e-01 4.85484868e-01
1.26355791e+00 9.63620663e-01 8.65561903e-01 -2.13536561e-01
-1.14492834e-01 -2.12320015e-01 2.60820299e-01 3.63420039e-01
-1.48151428e-01 -2.41110966e-01 -1.03142238e+00 2.58487523e-01
-2.12841010e+00 -6.39816880e-01 -4.16757673e-01 1.92894101e+00
3.20824347e-02 1.77154332e-01 3.86750013e-01 5.20583950e-02
4.84471440e-01 -1.78783819e-01 -1.08863723e+00 7.11228624e-02
-1.26086488e-01 -4.26883608e-01 7.85483003e-01 5.97317517e-01
-1.21481061e+00 1.14362323e+00 4.93730879e+00 4.22122180e-01
-6.36492670e-01 3.88456345e-01 5.50855577e-01 -1.05243303e-01
-9.14650261e-02 -3.23458165e-01 -1.07138944e+00 3.53771508e-01
4.38870102e-01 4.19610709e-01 3.98178458e-01 1.19331586e+00
2.49628231e-01 -2.79237449e-01 -1.14479315e+00 1.43034053e+00
1.86873972e-01 -8.98416638e-01 -1.60587713e-01 2.22518265e-01
5.44019699e-01 2.34482095e-01 2.98118442e-01 5.47346413e-01
3.14002693e-01 -7.54782379e-01 1.24505079e+00 5.37642181e-01
5.05725205e-01 -6.53501987e-01 6.29241824e-01 7.88268149e-01
-1.37761176e+00 -2.47339025e-01 -2.70371288e-01 7.44088441e-02
4.27303821e-01 4.58233625e-01 -9.02450323e-01 4.98029709e-01
9.91778910e-01 7.34563112e-01 -6.01359487e-01 1.21098769e+00
-2.31141180e-01 -1.95703834e-01 -5.99477828e-01 -3.19383055e-01
2.56584048e-01 -2.34438851e-01 5.34539819e-01 1.01786613e+00
3.78676921e-01 2.24852711e-01 5.37346303e-01 8.70774806e-01
4.10712808e-01 -2.28295743e-01 -5.92634559e-01 7.20357597e-01
3.61708492e-01 9.72848296e-01 -7.75788724e-01 -4.33502525e-01
-3.92974734e-01 9.58609998e-01 1.54527638e-03 6.51945651e-01
-1.07899129e+00 1.93381608e-01 1.20348072e+00 1.42298803e-01
1.22691028e-01 -6.74532950e-01 -6.64634049e-01 -1.09139252e+00
1.33715183e-01 -5.08202016e-01 9.62261334e-02 -9.84591424e-01
-8.80587280e-01 3.19469452e-01 5.50755002e-02 -1.02914667e+00
-1.23523466e-01 -1.08929956e+00 -9.70854610e-02 6.69723868e-01
-1.72788918e+00 -1.29977965e+00 -9.56075847e-01 7.73860931e-01
9.23618793e-01 2.13258311e-01 3.49537671e-01 5.10321438e-01
-6.00516200e-01 2.90379524e-01 -1.97814569e-01 -2.55432308e-01
2.85844415e-01 -1.30908990e+00 6.47626638e-01 6.57960653e-01
3.59264240e-02 1.74762264e-01 5.39957821e-01 -6.03325069e-01
-1.87311101e+00 -1.26925135e+00 5.96513748e-01 -9.71751630e-01
1.16625182e-01 -6.98698699e-01 -3.38818014e-01 7.13037491e-01
-5.36614716e-01 5.95015138e-02 2.70296693e-01 3.27209160e-02
-1.90253019e-01 -1.39745638e-01 -1.20238328e+00 7.76441991e-01
1.63728130e+00 -2.98806399e-01 -1.08851254e-01 5.04163682e-01
7.94221580e-01 -1.06408906e+00 -2.47192323e-01 3.82906675e-01
6.22750282e-01 -1.04252887e+00 1.56543779e+00 -4.35952216e-01
1.29902586e-01 -5.29606581e-01 -7.23719239e-01 -8.82549942e-01
6.29565567e-02 4.72931042e-02 -3.06412816e-01 6.70900285e-01
4.04822677e-01 -5.19499838e-01 1.08746707e+00 6.60137713e-01
-4.84063715e-01 -6.45935774e-01 -1.14116752e+00 -6.54152930e-01
-4.12004888e-01 -1.12314320e+00 7.66881227e-01 2.96279550e-01
-5.75198591e-01 2.20598549e-01 -2.61321485e-01 7.90674329e-01
8.81918848e-01 2.62989312e-01 1.37283385e+00 -1.15062749e+00
6.44968897e-02 -4.54821706e-01 -5.56307018e-01 -1.77779472e+00
-8.47111344e-02 -6.18485332e-01 3.23014110e-01 -2.04313731e+00
1.26526460e-01 -6.04251921e-01 2.01802149e-01 2.08980128e-01
-1.68933928e-01 6.81113750e-02 3.16170007e-01 2.88671330e-02
-8.92937243e-01 6.05714262e-01 1.27668440e+00 -2.84447521e-01
-2.80452847e-01 3.51942599e-01 -4.44635898e-01 8.95989597e-01
5.65657794e-01 -4.22427729e-02 -3.31228942e-01 -6.47638023e-01
-2.06090942e-01 1.20711148e-01 1.06185365e+00 -1.09904742e+00
4.86426234e-01 -3.51087332e-01 5.39997578e-01 -1.29890168e+00
1.03104782e+00 -1.04903424e+00 -1.41194269e-01 4.97751951e-01
3.83428603e-01 -8.46486837e-02 9.45749953e-02 6.83152437e-01
1.70567676e-01 1.86333895e-01 3.68752956e-01 -3.49026918e-01
-1.22700739e+00 6.06720090e-01 -1.91878691e-01 -2.96669543e-01
1.15449393e+00 -7.75673330e-01 -4.28260595e-01 -1.64450243e-01
-8.11605930e-01 6.25960112e-01 3.16297501e-01 6.89997792e-01
9.19306874e-01 -1.22062278e+00 -5.72382152e-01 2.52237529e-01
3.17945153e-01 6.49803698e-01 5.00447810e-01 8.47530127e-01
-2.81981885e-01 6.23753190e-01 -5.27561177e-03 -1.24145508e+00
-1.35087430e+00 5.23621857e-01 2.66257137e-01 1.46911561e-01
-5.49025655e-01 7.44800031e-01 5.08773327e-01 -8.42838228e-01
5.50418437e-01 -8.59379411e-01 -8.62245187e-02 -1.81804597e-01
3.16404313e-01 5.96072376e-01 2.63544656e-02 -6.04982436e-01
-6.94001853e-01 8.95297468e-01 3.04524302e-01 1.66210029e-02
1.12734449e+00 -5.05561173e-01 5.08537650e-01 4.71421182e-01
9.13139224e-01 -4.83400553e-01 -1.53493488e+00 -3.05438995e-01
-2.48801485e-01 -5.79634845e-01 -1.88700780e-01 -7.25313663e-01
-7.58856297e-01 1.07971179e+00 8.66131186e-01 -1.23648636e-01
8.55842292e-01 4.11099136e-01 8.50048840e-01 5.73201299e-01
9.06339586e-01 -1.12034202e+00 1.75898895e-01 7.16310441e-01
7.32191503e-01 -1.76385176e+00 -2.14547664e-01 -7.49269187e-01
-6.08076334e-01 7.24591792e-01 1.02470338e+00 4.20605659e-01
5.04731953e-01 1.58915624e-01 2.65780836e-03 -1.47304848e-01
-3.85627687e-01 -6.97934926e-01 2.27779880e-01 8.83506775e-01
-3.58684033e-01 4.00833458e-01 5.78445673e-01 6.62251294e-01
-5.01883566e-01 -2.71140426e-01 1.27983794e-01 7.78454602e-01
-6.42322361e-01 -4.84122097e-01 -7.92146564e-01 2.36227542e-01
4.28484648e-01 1.83726311e-01 -1.99845150e-01 8.25468719e-01
6.49388075e-01 1.06156754e+00 1.36959344e-01 -4.61989373e-01
8.75625968e-01 -2.17885002e-01 5.94574571e-01 -4.89354491e-01
-8.36382136e-02 -1.48578495e-01 -4.20030989e-02 -7.74122417e-01
-1.68663025e-01 -9.47014630e-01 -1.17983091e+00 -3.12135458e-01
-1.41473144e-01 -6.09498918e-01 1.19409311e+00 1.08812845e+00
1.51412234e-01 5.11113167e-01 2.60492861e-01 -1.31997931e+00
-1.72327489e-01 -6.61981463e-01 -2.40217894e-01 2.01352090e-01
4.54609007e-01 -9.75730717e-01 -2.98769530e-02 -1.15605839e-01] | [8.016133308410645, -2.331118583679199] |
36c64352-c149-4de9-a61e-f59585c70eb4 | deep-learning-for-short-latency-epileptic | 2301.03465 | null | https://arxiv.org/abs/2301.03465v2 | https://arxiv.org/pdf/2301.03465v2.pdf | Shorter Latency of Real-time Epileptic Seizure Detection via Probabilistic Prediction | Although recent studies have proposed seizure detection algorithms with good sensitivity performance, there is a remained challenge that they were hard to achieve significantly short detection latency in real-time scenarios. In this manuscript, we propose a novel deep learning framework intended for shortening epileptic seizure detection latency via probabilistic prediction. We are the first to convert the seizure detection task from traditional binary classification to probabilistic prediction by introducing a crossing period from seizure-oriented EEG recording and proposing a labeling rule using soft-label for crossing period samples. And, a novel multiscale STFT-based feature extraction method combined with 3D-CNN architecture is proposed to accurately capture predictive probabilities of samples. Furthermore, we also propose rectified weighting strategy to enhance predictive probabilities, and accumulative decision-making rule to achieve significantly shorter detection latency. We implement the proposed framework on two prevalent datasets -- CHB-MIT scalp EEG dataset and SWEC-ETHZ intracranial EEG dataset in patient-specific leave-one-seizure-out cross-validation scheme. Eventually, the proposed algorithm successfully detected 94 out of 99 seizures during crossing period and 100% seizures detected after EEG onset, averaged 14.84% rectified predictive ictal probability (RPIP) errors of crossing samples, 2.3 s detection latency, 0.08/h false detection rate (FDR) on CHB-MIT dataset. Meanwhile, 84 out of 89 detected seizures during crossing period, 100% detected seizures after EEG onset, 16.17% RPIP errors, 4.7 s detection latency, and 0.08/h FDR are achieved on SWEC-ETHZ dataset. The obtained detection latencies are at least 50% shorter than state-of-the-art results reported in previous studies. | ['Mohamad Sawan', 'Shuang Wang', 'Wenjie Ming', 'Jie Yang', 'Yankun Xu'] | 2023-01-04 | null | null | null | null | ['seizure-detection'] | ['medical'] | [ 2.65807807e-01 -2.00027466e-01 2.97224253e-01 -3.19075525e-01
-1.13294411e+00 -3.18194002e-01 2.58507282e-01 3.53863120e-01
-6.70830667e-01 9.64326024e-01 -2.55741030e-01 -1.08556166e-01
-6.62615776e-01 -4.19099897e-01 -4.63105232e-01 -7.20340014e-01
-6.40721083e-01 3.74159366e-02 3.13052088e-01 2.86236674e-01
2.23153636e-01 4.26883608e-01 -1.26114571e+00 3.08378458e-01
8.12017143e-01 1.34381318e+00 2.72667825e-01 4.01297003e-01
6.10252559e-01 3.06018561e-01 -8.26783836e-01 1.35449573e-01
-2.93561239e-02 -1.42316699e-01 -2.72135496e-01 -1.73283845e-01
-4.29250598e-01 -1.36848360e-01 -2.52164125e-01 1.00734711e+00
9.09454525e-01 -3.57410610e-01 1.04708874e+00 -1.25789607e+00
-1.73401654e-01 4.88846272e-01 -6.78510606e-01 5.56106925e-01
2.48266041e-01 1.26152098e-01 4.01290983e-01 -7.47881174e-01
1.90780491e-01 3.60565990e-01 6.87015533e-01 2.86215872e-01
-1.05913329e+00 -1.22458923e+00 -1.97135508e-01 4.05859292e-01
-1.98794389e+00 -6.76428825e-02 5.78631580e-01 -8.37725461e-01
1.22828674e+00 2.51917243e-01 8.80922019e-01 1.26077664e+00
8.37916613e-01 4.52191561e-01 1.33642876e+00 -4.60626148e-02
3.75073433e-01 -4.53439355e-02 3.17909539e-01 -1.43737709e-02
2.96193808e-01 2.60031968e-01 -5.87415814e-01 -1.93918243e-01
5.50362408e-01 1.74156144e-01 -5.23785293e-01 3.59634995e-01
-1.10360479e+00 4.69424695e-01 1.02873012e-01 3.86082560e-01
-6.58509910e-01 -1.69736072e-01 3.67247045e-01 -3.08210105e-02
3.18415016e-01 2.63291866e-01 -6.84328318e-01 -3.33885521e-01
-1.27002931e+00 2.36265957e-01 4.67190981e-01 8.91978383e-01
3.55157793e-01 1.10064439e-01 -5.05926669e-01 5.69236398e-01
4.69095744e-02 5.78948677e-01 6.86707437e-01 -5.70293032e-02
2.03868613e-01 5.32792687e-01 -1.09486002e-02 -7.53241777e-01
-8.76321018e-01 -1.01182008e+00 -9.58567858e-01 -7.31131658e-02
-1.06877223e-01 -2.81176418e-01 -7.43439257e-01 1.52910936e+00
-4.52298194e-01 2.52430797e-01 4.01784554e-02 7.78986454e-01
3.80662203e-01 3.51651251e-01 1.25802815e-01 -6.55968308e-01
1.67032766e+00 3.11314315e-02 -6.89167738e-01 1.00437306e-01
4.44089681e-01 -6.99959934e-01 5.66488087e-01 9.97244954e-01
-6.76698029e-01 -2.12731108e-01 -1.07735062e+00 7.19307542e-01
1.64684802e-01 5.22607207e-01 2.64934123e-01 5.05144536e-01
-9.15316463e-01 3.20678234e-01 -1.06370926e+00 -1.96188882e-01
6.63082123e-01 6.41489685e-01 -3.48161429e-01 3.37151259e-01
-1.17651153e+00 7.34677851e-01 4.01238382e-01 -6.79098740e-02
-1.20795703e+00 -7.53638744e-01 -3.88460100e-01 1.05380356e-01
-1.49271697e-01 -1.80524155e-01 9.52380836e-01 -3.51122051e-01
-1.02159774e+00 5.09371459e-01 -2.86802173e-01 -7.60010898e-01
5.57565391e-01 -2.33396277e-01 -7.55903780e-01 -9.71649028e-03
1.65495306e-01 4.06423032e-01 6.17249846e-01 -6.12454891e-01
-7.16335356e-01 -6.34546876e-01 -5.19813955e-01 -3.59148644e-02
-1.48951367e-01 1.50372908e-01 2.67308988e-02 -8.12485397e-01
1.57783464e-01 -6.54552937e-01 -5.27612492e-02 -6.48839772e-01
-4.35979128e-01 -3.54290664e-01 5.66597879e-01 -6.75703645e-01
1.33982253e+00 -2.11141181e+00 -4.12521660e-01 1.19818270e-01
3.42255473e-01 1.83591843e-01 6.71626851e-02 1.69583142e-01
-3.23490769e-01 -1.38196051e-01 -4.88553971e-01 -5.58925420e-02
-1.37868330e-01 -5.83793283e-01 -4.21601236e-01 5.88908792e-01
4.42102939e-01 5.38351655e-01 -6.85335755e-01 -2.35074386e-03
3.28831226e-01 8.10576618e-01 -1.96597382e-01 1.75796524e-01
5.18289447e-01 4.46204633e-01 -2.38208830e-01 6.10312402e-01
8.22559655e-01 1.79386973e-01 -1.20596997e-01 -3.28846186e-01
-2.96100706e-01 1.82771042e-01 -1.11914682e+00 1.38743007e+00
-1.52515396e-01 7.20803618e-01 -6.43868625e-01 -1.03946090e+00
1.20081818e+00 5.35896301e-01 3.92749459e-01 -6.07454956e-01
4.07468230e-01 3.95379454e-01 2.18990311e-01 -5.21747530e-01
-9.92954746e-02 -7.65512437e-02 -5.56031801e-03 5.29722422e-02
6.56203926e-02 1.03504546e-01 -1.43485963e-01 -2.21499324e-01
1.52551126e+00 -4.78591435e-02 4.99653429e-01 -6.04122996e-01
3.70124608e-01 -1.88822523e-01 8.32976520e-01 6.74138606e-01
-3.73884857e-01 5.63016295e-01 5.12518048e-01 -4.35132056e-01
-3.92654300e-01 -1.09080780e+00 -8.38853061e-01 2.52858818e-01
2.25826520e-02 -3.40191394e-01 -8.33419502e-01 -3.44510466e-01
-4.22745854e-01 7.12895513e-01 -3.19653273e-01 -2.27153391e-01
-1.99256718e-01 -1.39002752e+00 7.55607247e-01 5.87493420e-01
4.30490196e-01 -1.18372905e+00 -1.13523793e+00 5.76458395e-01
2.17205808e-02 -1.12471783e+00 5.31340309e-04 6.08318150e-01
-6.52776957e-01 -1.22823536e+00 -7.18893945e-01 -6.18423581e-01
4.10865486e-01 -3.81581485e-01 5.88335991e-01 -5.19941866e-01
-6.06643617e-01 -3.03754091e-01 -4.44658041e-01 -5.05034566e-01
3.13499123e-01 -1.91468656e-01 4.20270711e-01 -1.35569861e-02
8.18953454e-01 -8.54876161e-01 -9.01974022e-01 2.04798311e-01
-5.16432762e-01 -8.27714354e-02 8.50690484e-01 7.88687706e-01
7.91983545e-01 1.91485777e-01 1.10987425e+00 -2.59444058e-01
6.37766719e-01 -5.19637108e-01 -6.05605006e-01 8.08428004e-02
-6.17339909e-01 -3.71314883e-01 5.93935668e-01 -6.37719035e-01
-3.81510943e-01 5.90549409e-02 -2.20035851e-01 -3.44411135e-01
-3.92840385e-01 2.36959442e-01 3.81377563e-02 1.97813258e-01
6.67005897e-01 7.33712435e-01 -7.25195110e-01 -1.66648671e-01
-6.15787089e-01 9.43898618e-01 3.90705526e-01 -4.56589945e-02
1.59957618e-01 6.35438412e-02 -1.58898592e-01 -5.42002559e-01
-1.02205932e-01 -2.40921602e-01 -2.91967809e-01 1.04703024e-01
8.07534218e-01 -1.24360013e+00 -8.16112459e-01 6.61964655e-01
-1.20442784e+00 -1.14535436e-01 2.95073867e-01 1.03581822e+00
-3.69028300e-01 6.05345294e-02 -3.30981940e-01 -1.22315323e+00
-1.01049459e+00 -1.53759992e+00 1.03403747e+00 -4.15004194e-02
-5.38052499e-01 -2.14324683e-01 -2.60918915e-01 -2.69048244e-01
3.40414554e-01 4.76573348e-01 7.09150434e-01 -1.06037080e+00
-2.75183260e-01 -5.91282547e-01 -2.68433869e-01 2.29774371e-01
2.49476269e-01 -3.01287353e-01 -1.08828533e+00 -3.87401015e-01
6.25731796e-02 1.64070502e-01 5.86404979e-01 6.86106086e-01
1.34702981e+00 4.27985415e-02 -5.71115434e-01 5.93831241e-01
1.37282038e+00 8.98847520e-01 5.51243901e-01 5.04175760e-02
4.66475114e-02 -1.91924325e-03 3.10090661e-01 8.75928104e-01
2.59713084e-01 5.49868345e-01 1.88782886e-01 2.43624568e-01
-7.99373351e-03 2.63169080e-01 3.36478621e-01 8.08182955e-01
4.37362306e-02 -1.63830906e-01 -1.12188971e+00 8.11463833e-01
-1.51606941e+00 -4.84740585e-01 -3.52769017e-01 2.37871909e+00
7.15057969e-01 4.27973300e-01 -1.25152305e-01 6.76320910e-01
7.79443920e-01 -5.23275912e-01 -5.28925598e-01 1.21433347e-01
-1.32027596e-01 6.96520567e-01 3.54614288e-01 2.01628774e-01
-1.03760314e+00 5.23440480e-01 5.10204172e+00 8.92129660e-01
-1.48789203e+00 3.54788154e-01 6.31510258e-01 -3.27500463e-01
2.33587742e-01 -2.31245577e-01 -8.39470983e-01 7.69862175e-01
1.05442095e+00 -8.58017728e-02 2.00273424e-01 5.43454885e-01
4.42191035e-01 -2.62381166e-01 -1.03860497e+00 1.55195558e+00
-1.07109822e-01 -1.01540339e+00 -2.37896398e-01 -9.66832936e-02
4.13100272e-01 1.55209169e-01 -1.31182224e-01 2.11884767e-01
-4.67220485e-01 -1.03052258e+00 7.93195307e-01 8.22308540e-01
1.25115871e+00 -9.18689847e-01 1.06840861e+00 4.13043171e-01
-1.09858310e+00 -6.38390258e-02 -1.79069012e-01 -2.39350423e-01
-3.91930789e-02 1.03504992e+00 -9.39760864e-01 5.74258804e-01
1.10024059e+00 7.04249382e-01 -4.16602582e-01 1.39558947e+00
-3.29028368e-01 8.42071176e-01 -3.18717688e-01 -2.14012250e-01
-1.19309612e-01 2.99774468e-01 6.76961899e-01 1.24570024e+00
7.90753365e-01 2.16457888e-01 -2.61126399e-01 7.03932345e-01
1.73167363e-01 -4.25091423e-02 -3.05808097e-01 3.31383139e-01
8.02196145e-01 1.26316071e+00 -7.67484367e-01 -1.00610554e-01
-5.24118654e-02 6.39746010e-01 3.61207873e-02 3.33839357e-01
-1.01520181e+00 -9.41688597e-01 2.08174333e-01 1.72879156e-02
-5.98525219e-02 1.52633429e-01 -4.54969406e-01 -1.03197265e+00
2.79459715e-01 -5.85727096e-01 -3.85805555e-02 -6.58347070e-01
-1.04038382e+00 1.18838418e+00 -7.86631182e-02 -1.31708574e+00
-7.83353820e-02 -5.09104669e-01 -7.24846721e-01 9.48483169e-01
-1.37127769e+00 -6.83530927e-01 -2.22572967e-01 7.80132949e-01
5.38815200e-01 -2.62335390e-01 1.00629127e+00 3.93829614e-01
-7.34239340e-01 7.70038128e-01 -1.38698295e-01 -1.00306392e-01
6.14851296e-01 -7.78957844e-01 -1.02757543e-01 8.71702135e-01
-2.96927631e-01 4.79356319e-01 4.18116510e-01 -5.52852929e-01
-9.35716093e-01 -1.23018706e+00 7.81509697e-01 -2.17031129e-02
5.94578862e-01 -5.05314112e-01 -8.13942671e-01 3.70283633e-01
1.14609487e-02 1.58513989e-02 6.20181561e-01 -1.95240051e-01
1.76421091e-01 -3.86154294e-01 -1.26871359e+00 2.12622210e-01
7.02364445e-01 -3.67332548e-01 -8.91505122e-01 2.74632037e-01
2.56536543e-01 -1.56806692e-01 -1.01226306e+00 1.04024339e+00
6.50414228e-01 -7.98327625e-01 4.23602790e-01 1.97883099e-01
-7.14184940e-02 -3.16868424e-01 -1.28693208e-01 -1.07862401e+00
-2.63587981e-01 -6.82268739e-01 3.23109291e-02 9.70382810e-01
3.74576032e-01 -8.09678674e-01 3.67867082e-01 2.05619127e-01
-5.00758767e-01 -1.27800739e+00 -1.22635961e+00 -6.66819453e-01
-1.26521736e-01 -9.44959104e-01 5.04575133e-01 4.94350344e-01
2.08396003e-01 -2.13972121e-01 -1.00072011e-01 5.10455608e-01
5.77663362e-01 -2.87847131e-01 -2.41101533e-02 -1.10647643e+00
2.67293137e-02 -2.73359954e-01 -7.80120909e-01 -2.73974270e-01
-2.10983977e-01 -8.66832078e-01 6.21605031e-02 -1.23274636e+00
4.17967916e-01 -2.59663135e-01 -9.87827301e-01 7.40854084e-01
-1.10849708e-01 3.69704723e-01 -3.55502814e-01 1.13568135e-01
-2.40396887e-01 3.74767542e-01 5.41788578e-01 6.83176331e-03
-3.42297018e-01 -2.42436514e-03 -3.00798088e-01 7.25213647e-01
9.51070249e-01 -7.47965872e-01 -4.10025328e-01 -2.12193981e-01
-1.61084563e-01 2.28380114e-01 3.75846535e-01 -1.54123127e+00
3.37045103e-01 2.66633421e-01 8.88214707e-01 -8.78649533e-01
2.93611109e-01 -4.39818084e-01 3.14954042e-01 7.16756344e-01
5.95092662e-02 -8.15170724e-03 5.76166332e-01 3.55551690e-01
-1.21968739e-01 2.05900893e-02 5.87948322e-01 4.91655558e-01
-5.31584799e-01 2.97250271e-01 -7.71910012e-01 -9.97024402e-02
1.40796041e+00 -2.98039407e-01 4.54606814e-03 2.35353425e-01
-9.56043363e-01 -3.13968919e-02 -2.56483644e-01 2.79414386e-01
9.71438229e-01 -9.91575837e-01 -8.25280964e-01 6.03438914e-01
3.08844417e-01 -3.30360919e-01 5.27018249e-01 1.31400728e+00
-2.55471647e-01 5.81391156e-01 -3.75667542e-01 -7.77953565e-01
-9.52123642e-01 -6.31627487e-03 2.21025378e-01 2.98435092e-02
-7.31906772e-01 1.07069623e+00 2.39154235e-01 2.51716614e-01
2.19629943e-01 -5.86136162e-01 -4.76922132e-02 1.90326106e-02
6.67992294e-01 2.37798095e-01 5.82756579e-01 -2.57177591e-01
-9.73142564e-01 5.51487148e-01 1.22520165e-03 -1.35961836e-02
1.45901215e+00 3.50167185e-01 4.63427380e-02 2.41221547e-01
8.01356733e-01 -3.18769097e-01 -1.26234782e+00 3.51762593e-01
-1.10191982e-02 -1.24853160e-02 3.16675097e-01 -1.33902800e+00
-7.92836428e-01 1.06179023e+00 1.32502687e+00 -1.39791789e-02
1.55773365e+00 -1.90383509e-01 6.48213148e-01 -2.16460638e-02
8.00198257e-01 -5.64812720e-01 -5.53030491e-01 2.36519545e-01
8.90541792e-01 -8.73873949e-01 -1.30049393e-01 1.50493681e-02
-5.64706624e-01 1.06201863e+00 3.79121125e-01 -3.99145007e-01
7.55638242e-01 4.77897704e-01 -4.27754849e-01 -1.40222341e-01
-8.42775583e-01 1.70498520e-01 3.34986269e-01 6.39340580e-01
4.26306993e-01 5.20270586e-01 -6.59446537e-01 1.58666873e+00
-2.48264030e-01 3.62697005e-01 3.77993286e-01 6.75352156e-01
-2.33995408e-01 -5.68219066e-01 -1.48823142e-01 8.62077355e-01
-6.56753898e-01 -4.13258255e-01 1.64018363e-01 7.32054234e-01
3.71918797e-01 1.10029185e+00 1.32572874e-01 -8.16188276e-01
3.10884923e-01 2.52809703e-01 4.59359527e-01 -4.90535885e-01
-5.52243590e-01 4.70796764e-01 -3.53056908e-01 -3.80577475e-01
-6.44052625e-02 -6.83504581e-01 -1.36690974e+00 3.87913883e-01
-4.50478643e-01 2.76024431e-01 5.61152101e-01 1.03921413e+00
7.33652055e-01 8.35230708e-01 4.26096499e-01 -4.76479560e-01
-2.16537878e-01 -1.45869350e+00 -9.35622931e-01 -2.16852888e-01
3.43755037e-01 -8.42664838e-01 -4.54580188e-01 -6.11827150e-03] | [13.224255561828613, 3.5161561965942383] |
26765b21-7a4e-4918-83a0-1883555fd740 | graph-constrained-reinforcement-learning-for-1 | 2001.08837 | null | https://arxiv.org/abs/2001.08837v1 | https://arxiv.org/pdf/2001.08837v1.pdf | Graph Constrained Reinforcement Learning for Natural Language Action Spaces | Interactive Fiction games are text-based simulations in which an agent interacts with the world purely through natural language. They are ideal environments for studying how to extend reinforcement learning agents to meet the challenges of natural language understanding, partial observability, and action generation in combinatorially-large text-based action spaces. We present KG-A2C, an agent that builds a dynamic knowledge graph while exploring and generates actions using a template-based action space. We contend that the dual uses of the knowledge graph to reason about game state and to constrain natural language generation are the keys to scalable exploration of combinatorially large natural language actions. Results across a wide variety of IF games show that KG-A2C outperforms current IF agents despite the exponential increase in action space size. | ['Prithviraj Ammanabrolu', 'Matthew Hausknecht'] | 2020-01-23 | null | https://openreview.net/forum?id=B1x6w0EtwH | https://openreview.net/pdf?id=B1x6w0EtwH | iclr-2020-1 | ['action-generation'] | ['computer-vision'] | [-7.89433438e-03 6.23965263e-01 -2.16353297e-01 3.43643576e-01
-4.09608632e-01 -1.05025303e+00 1.03829372e+00 -3.95887703e-01
-5.08282661e-01 1.09401000e+00 2.67430484e-01 -7.61635125e-01
-1.64447621e-01 -1.19794047e+00 -2.91197419e-01 -6.55922741e-02
-4.79069442e-01 7.93073475e-01 4.78180140e-01 -8.26615930e-01
1.38018563e-01 1.53729945e-01 -1.43442512e+00 -9.52215046e-02
8.19362223e-01 5.86282611e-02 1.79815665e-01 1.01636362e+00
-9.54284519e-02 1.37031949e+00 -3.96446824e-01 8.47075954e-02
6.37752712e-01 -7.61570752e-01 -1.10672021e+00 1.57880057e-02
-4.18160647e-01 -5.64308643e-01 -7.13329911e-01 8.15249503e-01
9.90924314e-02 3.31421822e-01 1.92295954e-01 -1.43014848e+00
-8.44570771e-02 1.00600648e+00 4.70435694e-02 -1.61080971e-01
8.25647235e-01 1.09442663e+00 8.63174140e-01 5.53848930e-02
9.79004085e-01 1.58312964e+00 9.48839933e-02 9.91609216e-01
-1.13264585e+00 -4.48044300e-01 3.61316741e-01 3.98551859e-02
-1.36818647e+00 -1.10556722e-01 2.25545928e-01 -2.52848208e-01
1.64072669e+00 2.59178787e-01 1.31582129e+00 9.55755889e-01
5.78194499e-01 6.49483204e-01 1.20460033e+00 -4.35919642e-01
7.24328339e-01 -3.38640004e-01 -5.26010275e-01 1.09397221e+00
1.07632041e-01 7.38642275e-01 -6.89443052e-01 -4.34491813e-01
1.10873544e+00 -7.21936464e-01 2.15488046e-01 -5.10446429e-01
-1.08557761e+00 1.01394904e+00 -2.39963651e-01 -1.46215603e-01
-3.81603807e-01 7.40971625e-01 2.92507768e-01 6.40975535e-01
-3.31245720e-01 1.44497097e+00 -4.92794663e-01 -9.84487116e-01
-2.74042517e-01 8.53952825e-01 1.32188022e+00 1.13963866e+00
5.99340618e-01 4.88691479e-01 -9.76383016e-02 -4.70540896e-02
4.21103358e-01 7.88140297e-01 4.99710977e-01 -1.40550685e+00
4.73472983e-01 8.00566196e-01 5.66132128e-01 -4.40357417e-01
-5.11477888e-01 3.76295559e-02 1.28063574e-01 7.17276871e-01
4.09633249e-01 -9.28265274e-01 -9.51329827e-01 1.93970895e+00
2.66936362e-01 -1.62546620e-01 5.20737708e-01 5.20385087e-01
2.10394695e-01 6.30817294e-01 2.47887686e-01 -9.14093181e-02
1.10875893e+00 -6.27365291e-01 -4.58266675e-01 -7.33363867e-01
9.20522630e-01 5.39285205e-02 1.10748053e+00 3.45231831e-01
-1.24015939e+00 -1.27689643e-02 -1.07110906e+00 3.87727916e-01
-4.68289882e-01 -7.11173892e-01 1.23830223e+00 8.36269855e-01
-1.32605076e+00 3.15274775e-01 -9.19744611e-01 -3.40833426e-01
1.19493254e-01 5.40351510e-01 2.40706862e-03 2.92745262e-01
-1.62529945e+00 1.14264846e+00 9.66567338e-01 -3.52418810e-01
-1.39074528e+00 -1.65767699e-01 -1.00299752e+00 6.79091662e-02
1.16756475e+00 -5.28688967e-01 1.64083278e+00 -4.01846379e-01
-2.12569761e+00 3.45554978e-01 5.58717489e-01 -7.09933281e-01
7.45781600e-01 1.48500100e-01 -7.50764981e-02 -9.04130116e-02
9.57362577e-02 9.04964030e-01 1.96156621e-01 -7.76580632e-01
-7.51117706e-01 5.01883924e-02 8.60147595e-01 8.41495991e-01
1.97180063e-01 -3.78101975e-01 -2.67615616e-01 -9.70286578e-02
-2.75550157e-01 -1.30589092e+00 -8.70456040e-01 -4.49388176e-01
-3.11935574e-01 -1.76997408e-01 5.03966272e-01 5.30464463e-02
9.71289039e-01 -1.45377064e+00 2.03997001e-01 3.44709188e-01
2.34315142e-01 6.24800697e-02 -5.03881097e-01 1.02376688e+00
4.41030830e-01 3.13270241e-01 2.97620118e-01 6.13826454e-01
4.69425261e-01 3.97558510e-01 -2.36479297e-01 -7.70512521e-02
-5.40456697e-02 1.45211017e+00 -1.28708780e+00 -3.63262594e-01
2.50722229e-01 -4.54150766e-01 -7.74715006e-01 7.72106797e-02
-1.10448730e+00 3.20348978e-01 -1.16987002e+00 8.15781131e-02
2.55815927e-02 -9.03260242e-03 9.23068523e-01 9.33484972e-01
-1.47825077e-01 3.35179001e-01 -1.27430511e+00 1.54782510e+00
-4.26829368e-01 3.93655002e-01 -4.47468311e-02 -3.01042855e-01
6.45111203e-01 2.58487552e-01 2.38853499e-01 -8.66231978e-01
1.37775004e-01 -3.34902154e-03 5.22193730e-01 -4.04929042e-01
5.92202723e-01 -2.54258066e-01 -6.33750379e-01 1.00758755e+00
-1.71310678e-01 -1.00032091e+00 6.94221973e-01 5.39569497e-01
1.36920357e+00 3.98352206e-01 5.93716860e-01 -4.85509396e-01
1.77265242e-01 5.64320982e-01 3.54568541e-01 1.57875431e+00
-1.37882784e-01 -4.68820632e-01 8.55101585e-01 -6.41262829e-01
-1.00187516e+00 -1.04967284e+00 6.93526924e-01 1.18265772e+00
3.50423276e-01 -8.42645407e-01 -7.28604376e-01 -4.05742079e-01
-1.57111987e-01 1.25012779e+00 -4.66821998e-01 -2.89527982e-01
-6.28169060e-01 -1.19788758e-01 8.96022320e-01 3.79811764e-01
5.88883698e-01 -1.46111166e+00 -1.25715935e+00 5.17176926e-01
-9.30367038e-02 -8.25058103e-01 -3.61819327e-01 2.73972839e-01
-4.88010794e-01 -1.27238214e+00 1.12340681e-01 -5.68830550e-01
9.23374221e-02 -1.47952542e-01 1.03128934e+00 1.17879920e-01
-2.10855410e-01 8.81519914e-01 -2.40994304e-01 -4.26227659e-01
-9.65319633e-01 4.72542271e-02 1.02986522e-01 -9.57702518e-01
2.44408920e-01 -5.90743065e-01 -1.86217383e-01 1.97126359e-01
-8.72672081e-01 6.41714036e-01 2.12641746e-01 7.20442951e-01
-5.57332709e-02 2.76154131e-01 4.08310950e-01 -6.86002970e-01
1.29646027e+00 -2.29893416e-01 -1.11117923e+00 3.10119510e-01
-5.19353867e-01 4.76586550e-01 7.19844639e-01 -4.37066197e-01
-1.21421635e+00 1.46002188e-01 4.27714080e-01 4.24535006e-01
-1.86899286e-02 5.89144468e-01 -1.71596482e-01 3.95948626e-02
9.95126486e-01 5.93470216e-01 4.40450087e-02 4.33764726e-01
9.90793288e-01 1.50233924e-01 1.71317711e-01 -1.17601073e+00
1.00490236e+00 6.77865744e-02 6.30180538e-02 -5.07747531e-01
2.19741520e-02 1.32743835e-01 -2.37340540e-01 -2.85548151e-01
7.59861231e-01 -8.17264855e-01 -1.30416131e+00 2.98275143e-01
-7.96200156e-01 -1.31449497e+00 -6.62082016e-01 2.33353198e-01
-1.34566224e+00 9.63517949e-02 -4.67246622e-01 -9.97372329e-01
-1.30185718e-02 -1.29217017e+00 6.14569008e-01 3.59201372e-01
-4.54528719e-01 -8.44948411e-01 4.70007390e-01 1.00500293e-01
3.67552578e-01 1.00475892e-01 8.49525571e-01 -7.29770362e-01
-9.33390617e-01 1.38302088e-01 3.14150304e-01 -6.12011135e-01
2.31342658e-01 -3.49891722e-01 -2.88911521e-01 -2.40529820e-01
-4.03274000e-01 -8.18322241e-01 2.48753335e-02 1.77233778e-02
5.93153656e-01 -5.15184224e-01 -2.99819171e-01 6.13239780e-02
1.17695427e+00 8.38047504e-01 7.12304175e-01 6.03670239e-01
2.11077314e-02 1.77898377e-01 5.93397796e-01 8.48850548e-01
4.54423070e-01 5.11871517e-01 3.17036480e-01 5.96432984e-01
2.14442760e-01 -8.09468567e-01 4.25892174e-01 1.56189293e-01
-1.16613610e-02 -5.22683978e-01 -1.21647096e+00 5.26235700e-01
-2.04352474e+00 -1.05223966e+00 4.73564833e-01 1.74519408e+00
1.27009320e+00 5.34192920e-01 1.87031642e-01 -5.42678237e-01
3.87172908e-01 1.60668880e-01 -9.75550950e-01 -6.47065461e-01
9.74942222e-02 3.90333712e-01 5.82304716e-01 1.05465293e+00
-6.50072455e-01 1.71985102e+00 7.33967638e+00 9.33340430e-01
-6.14374876e-01 -4.81359243e-01 2.10302040e-01 -1.32815555e-01
-3.43765646e-01 2.79817969e-01 -4.92191762e-01 -9.99807101e-03
1.02301443e+00 -8.94204855e-01 1.00474477e+00 7.99975157e-01
4.41602558e-01 -6.38416111e-01 -1.09042442e+00 3.46484691e-01
-5.30656576e-01 -1.63257480e+00 1.10929221e-01 2.58677959e-01
8.29161227e-01 -1.77448243e-02 -7.92908743e-02 7.16920316e-01
1.81734002e+00 -1.25824821e+00 7.01009393e-01 1.50987610e-01
7.91466713e-01 -7.80045927e-01 1.50916263e-01 6.92439079e-01
-1.23147678e+00 -3.46015841e-01 4.33742069e-02 -6.98213160e-01
1.67569771e-01 -6.97084308e-01 -1.23033726e+00 1.27158195e-01
-9.79340356e-03 -6.59530684e-02 -3.34899276e-01 6.89804733e-01
-6.12055540e-01 4.55377787e-01 -3.83340597e-01 -6.47174537e-01
7.82006800e-01 -1.24568209e-01 7.31338978e-01 6.78954005e-01
2.28775013e-02 6.76500738e-01 6.44283712e-01 1.03201854e+00
4.07245308e-01 -3.65886748e-01 -9.00597990e-01 -5.93051791e-01
5.64340711e-01 6.45065427e-01 -7.16615558e-01 -4.88611698e-01
-2.03706175e-02 6.95406497e-01 1.39690787e-01 4.00102824e-01
-7.53770292e-01 -2.41565406e-01 7.04701126e-01 -1.93235159e-01
-1.64549146e-02 -6.96102023e-01 -6.08372018e-02 -9.90883350e-01
-5.65197408e-01 -1.44942009e+00 1.51336446e-01 -8.86351585e-01
-5.03042817e-01 4.11428303e-01 2.23448873e-01 -7.44132102e-01
-1.05444038e+00 -4.92912829e-01 -5.88613868e-01 7.00624585e-01
-7.19251931e-01 -8.18261981e-01 -1.93933453e-02 6.36179626e-01
7.66347587e-01 -4.82821584e-01 1.09129632e+00 -9.64324296e-01
-8.64781514e-02 2.55482465e-01 -2.18355078e-02 -2.29995069e-03
-1.66925192e-01 -1.37267435e+00 9.96695817e-01 6.17378652e-01
5.70925139e-02 5.41550457e-01 9.25572097e-01 -1.14569724e+00
-1.98654485e+00 -7.29036689e-01 -8.48947465e-03 -5.18835187e-01
1.05492520e+00 -4.33070749e-01 -2.03222990e-01 8.11222315e-01
3.25639933e-01 -5.73919415e-01 2.60639399e-01 -1.09779991e-01
-1.90368503e-01 5.28428733e-01 -9.63592112e-01 1.51692164e+00
1.32366025e+00 -5.14562488e-01 -6.57442331e-01 2.51112431e-01
8.39611351e-01 -6.50381207e-01 -5.04221976e-01 -2.15415508e-01
4.56830233e-01 -5.82199872e-01 6.19805455e-01 -1.10511172e+00
2.23515436e-01 -4.20913935e-01 4.99438345e-02 -1.55254459e+00
-3.17035913e-01 -1.41155601e+00 1.01554908e-01 2.35891953e-01
3.94723147e-01 -6.68971896e-01 1.25132203e+00 1.02157438e+00
3.45172971e-01 -3.62173915e-01 -7.65520632e-01 -1.01728177e+00
4.14787591e-01 -4.93448108e-01 5.84704101e-01 5.23799479e-01
9.47615087e-01 3.02325964e-01 -2.83053935e-01 -2.01747969e-01
3.41622829e-01 -6.37784600e-02 8.73646080e-01 -7.74827838e-01
-6.35566890e-01 -4.34712827e-01 -1.09372161e-01 -1.15983272e+00
1.42314941e-01 -5.18738508e-01 3.99675369e-01 -1.44099736e+00
1.19887792e-01 -3.37973446e-01 3.60479265e-01 7.01192021e-01
1.14881910e-01 -4.74498004e-01 5.75225174e-01 -1.61915794e-01
-1.03084457e+00 7.10079551e-01 1.66459227e+00 -1.66202471e-01
-6.11111939e-01 -3.14331293e-01 -7.34320223e-01 7.67823815e-01
9.45516944e-01 9.61930901e-02 -1.08136165e+00 -1.37599126e-01
6.33735240e-01 6.56311393e-01 -7.64516667e-02 -9.48582649e-01
4.68637943e-01 -1.31807709e+00 -2.27210209e-01 -2.32631736e-03
3.26678216e-01 -3.70494723e-01 1.47640944e-01 9.56236839e-01
-6.79259300e-01 1.46969303e-01 5.64740598e-01 6.43026888e-01
3.59065235e-01 5.19226380e-02 4.14437413e-01 -7.09818482e-01
-1.12078488e+00 6.61813542e-02 -1.56265235e+00 5.56848526e-01
1.32206690e+00 -3.23326409e-01 -3.13000739e-01 -1.10480464e+00
-7.12598741e-01 7.48923361e-01 4.24900144e-01 3.09186608e-01
5.42310059e-01 -9.72602844e-01 -5.18442750e-01 3.14228013e-02
5.07259741e-02 -1.56409845e-01 -1.13120101e-01 -1.48793563e-01
-7.52405882e-01 5.65243840e-01 -4.98511374e-01 1.94322452e-01
-9.17017519e-01 2.90845782e-01 7.40802646e-01 -6.66372716e-01
-6.53439045e-01 5.23671269e-01 3.26107234e-01 -6.57075405e-01
-8.46305043e-02 -1.43480361e-01 2.36232132e-02 -7.59876907e-01
5.20584226e-01 1.44295469e-01 -5.76489270e-01 6.64877668e-02
-2.26389915e-01 -2.74194051e-02 -9.44308762e-04 -8.09089363e-01
1.06529772e+00 -9.98608172e-02 2.47055888e-01 -1.09395675e-01
-7.48402402e-02 -1.81283966e-01 -1.48471832e+00 -1.48124233e-01
4.53023389e-02 -2.38437012e-01 -2.55150616e-01 -1.17183352e+00
-3.95800799e-01 1.30060777e-01 7.20098540e-02 3.45064372e-01
5.63651621e-01 -9.89163592e-02 3.13219011e-01 1.19169533e+00
1.07015169e+00 -1.46379554e+00 4.33602452e-01 1.04937577e+00
7.79653430e-01 -6.55169368e-01 1.08333826e-02 -1.24880888e-01
-9.19556558e-01 1.04124641e+00 1.17682564e+00 3.25596817e-02
6.85678273e-02 6.71464801e-01 -6.75338879e-02 -2.96299219e-01
-1.24562645e+00 -3.94873112e-01 -5.26634037e-01 1.05193281e+00
-2.58515298e-01 2.12817416e-01 -1.57486811e-01 1.66636571e-01
-6.58344984e-01 1.02127697e-02 1.15177023e+00 1.19456947e+00
-7.34984398e-01 -1.35494912e+00 -2.61134416e-01 2.85662174e-01
1.78144813e-01 -4.00794894e-02 -7.33306706e-01 1.13610780e+00
-2.07610890e-01 1.31832397e+00 -2.64076322e-01 -3.63242775e-01
-6.92913756e-02 3.41736525e-02 7.22935081e-01 -8.80649745e-01
-4.13170546e-01 -3.32544655e-01 6.04869425e-01 -9.67380404e-01
2.41263971e-01 -5.62491119e-01 -1.70627654e+00 -4.58028018e-01
-3.92481238e-02 3.86430591e-01 3.98463421e-02 1.03129065e+00
1.43074229e-01 3.51538152e-01 3.24921757e-01 -3.51192981e-01
-8.01487327e-01 -5.79468548e-01 -5.44770062e-01 4.86415364e-02
-1.91584732e-02 -5.89369357e-01 4.77513857e-02 -3.28612626e-01] | [3.7996842861175537, 1.4434049129486084] |
cac5fa75-a3a8-4b43-b620-5442475597c5 | identifying-individual-differences-in-gender | null | null | https://aclanthology.org/W16-0806 | https://aclanthology.org/W16-0806.pdf | Identifying Individual Differences in Gender, Ethnicity, and Personality from Dialogue for Deception Detection | null | ['Yocheved Levitan', 'Michelle Levine', 'Sarah Ita Levitan', 'Rivka Levitan', 'Andrew Rosenberg', 'Guozhen An', 'Julia Hirschberg'] | 2016-06-01 | null | null | null | ws-2016-6 | ['deception-detection'] | ['miscellaneous'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.23374605178833, 3.792767286300659] |
8531a677-35bc-4b53-975b-96fda1c21b82 | learning-ordinal-relationships-for-mid-level | null | null | http://openaccess.thecvf.com/content_iccv_2015/html/Zoran_Learning_Ordinal_Relationships_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Zoran_Learning_Ordinal_Relationships_ICCV_2015_paper.pdf | Learning Ordinal Relationships for Mid-Level Vision | We propose a framework that infers mid-level visual properties of an image by learning about ordinal relation- ships. Instead of estimating metric quantities directly, the system proposes pairwise relationship estimates for points in the input image. These sparse probabilistic ordinal mea- surements are globalized to create a dense output map of continuous metric measurements. Estimating order rela- tionships between pairs of points has several advantages over metric estimation: it solves a simpler problem than metric regression; humans are better at relative judgements, so data collection is easier; ordinal relationships are invari- ant to monotonic transformations of the data, thereby in- creasing the robustness of the system and providing qualitatively different information. We demonstrate that this frame- work works well on two important mid-level vision tasks: intrinsic image decomposition and depth from an RGB im- age. We train two systems with the same architecture on data from these two modalities. We provide an analysis of the resulting models, showing that they learn a number of simple rules to make ordinal decisions. We apply our algo-rithm to depth estimation, with good results, and intrinsic image decomposition, with state-of-the-art results. | ['Daniel Zoran', 'William T. Freeman', 'Phillip Isola', 'Dilip Krishnan'] | 2015-12-01 | null | null | null | iccv-2015-12 | ['intrinsic-image-decomposition'] | ['computer-vision'] | [ 1.64397553e-01 2.82192491e-02 -2.18252599e-01 -8.61282527e-01
-8.38734031e-01 -5.21473646e-01 5.70354998e-01 1.68475419e-01
-7.07316995e-01 4.96255994e-01 3.43969494e-01 1.92702189e-01
-1.82886884e-01 -7.01839507e-01 -6.71569884e-01 -6.46976769e-01
-2.07819715e-01 7.51983404e-01 1.03921227e-01 4.97672744e-02
6.50669277e-01 5.70465624e-01 -1.91082275e+00 6.93267286e-01
4.19092864e-01 1.32704449e+00 -3.69362235e-02 9.25582767e-01
2.27863520e-01 9.49721575e-01 -3.76173228e-01 -1.84438691e-01
3.39341164e-01 -3.30538273e-01 -8.22493613e-01 3.12352687e-01
8.10250461e-01 -4.80717659e-01 -2.65875161e-01 8.85607243e-01
4.32474837e-02 -1.22635297e-01 1.18342519e+00 -1.23579443e+00
-6.07999921e-01 3.77404302e-01 -7.41966009e-01 -1.23268351e-01
6.78748846e-01 -1.49895936e-01 1.47789860e+00 -9.49961841e-01
4.74001795e-01 1.50174582e+00 6.07142568e-01 3.67316395e-01
-1.56574655e+00 -1.53732732e-01 4.35287058e-02 4.23135161e-01
-1.37083650e+00 -4.51063484e-01 6.36165261e-01 -5.35875142e-01
9.05294538e-01 3.23834628e-01 4.37292159e-01 7.44719625e-01
3.26944292e-02 6.78883374e-01 1.41706288e+00 -4.23125386e-01
2.75081515e-01 -1.50904372e-01 -1.51805595e-01 8.74668539e-01
-2.07416397e-02 4.68597002e-02 -9.39391851e-01 1.52505293e-01
9.94907618e-01 -1.29511412e-02 -7.18541518e-02 -6.02747858e-01
-1.69611287e+00 4.79023337e-01 7.74128914e-01 1.21153310e-01
-1.44026190e-01 5.08035302e-01 -3.10930051e-02 4.13179398e-01
1.18192405e-01 6.01474822e-01 -2.33059123e-01 -1.05697937e-01
-8.89302969e-01 5.26638255e-02 7.07934022e-01 7.35129714e-01
1.07815397e+00 -5.72320282e-01 3.37858558e-01 5.08865833e-01
4.30647820e-01 3.45274836e-01 3.53322089e-01 -1.69302809e+00
4.15566802e-01 4.38799888e-01 9.10430327e-02 -1.17456591e+00
-5.35911918e-01 2.22157016e-01 -7.90056467e-01 9.32633340e-01
7.75337815e-01 3.33688468e-01 -9.21475530e-01 1.86097348e+00
-1.53351411e-01 -2.24357828e-01 -2.86551237e-01 7.57785738e-01
5.61805010e-01 3.58077884e-01 -3.12165022e-01 8.54915977e-02
1.27070069e+00 -3.69416833e-01 -3.92731428e-01 -1.64362743e-01
4.12995040e-01 -4.63002920e-01 1.22516477e+00 7.91023970e-01
-1.19182658e+00 -4.75025922e-01 -1.23526895e+00 -5.88864505e-01
-4.80165064e-01 -8.77761021e-02 6.52446628e-01 2.34105274e-01
-1.43837166e+00 9.39561725e-01 -9.65504706e-01 -4.09609586e-01
4.16764021e-01 4.93740946e-01 -7.06511557e-01 1.46877840e-02
-4.83788669e-01 1.02662838e+00 1.93393260e-01 1.38900742e-01
-7.01687992e-01 -4.08295542e-01 -1.09773970e+00 -2.83281416e-01
-1.05715260e-01 -6.89303517e-01 1.13397753e+00 -6.91851795e-01
-1.23812342e+00 1.29152679e+00 -3.88615280e-01 -2.69052107e-02
5.51054239e-01 1.22296154e-01 2.42035165e-01 1.26734719e-01
1.40180603e-01 1.26494873e+00 7.19124317e-01 -1.34125626e+00
-6.98053777e-01 -6.71696246e-01 3.91894281e-01 4.86936480e-01
-2.13486388e-01 -1.71797380e-01 -1.01214141e-01 -3.40604901e-01
8.42548311e-01 -7.14339614e-01 -2.30445135e-02 7.44616926e-01
-2.87270844e-01 -1.44420162e-01 4.28327203e-01 -2.53553063e-01
8.45008194e-01 -2.00679946e+00 6.24452114e-01 1.95700347e-01
4.47729826e-01 -9.01142478e-01 -6.99623525e-02 1.17952645e-01
-2.96581443e-03 5.99238239e-02 -4.59101945e-01 -7.54751325e-01
2.03412756e-01 5.63580215e-01 -1.84576064e-02 7.04059839e-01
1.45103067e-01 6.91506803e-01 -9.81196165e-01 -6.83949649e-01
4.73997742e-01 1.83084577e-01 -4.79197234e-01 3.89288157e-01
-1.11325055e-01 2.39568621e-01 1.80820838e-01 9.00446177e-01
4.07296538e-01 -8.68099406e-02 -7.29532689e-02 -5.17997503e-01
-1.45456241e-02 2.65567631e-01 -1.30404091e+00 2.04718804e+00
-5.54622233e-01 9.02541041e-01 -1.45712063e-01 -8.11194479e-01
7.11725652e-01 -5.51609024e-02 5.43845832e-01 -5.85297883e-01
3.40635441e-02 1.23338580e-01 -1.33260995e-01 -3.96086097e-01
3.65778655e-01 -1.25043079e-01 -1.53482944e-01 5.47437191e-01
6.40553907e-02 -9.01209891e-01 2.94457376e-01 1.13001294e-01
8.60857368e-01 4.94840235e-01 4.28644180e-01 -2.03663215e-01
2.44597524e-01 -2.70976782e-01 3.21580619e-01 6.56656682e-01
-2.25861847e-01 1.00060403e+00 7.48570979e-01 -6.36097312e-01
-1.05121839e+00 -1.66773462e+00 -2.69500732e-01 1.10645628e+00
2.68251628e-01 -4.96169716e-01 -5.67207932e-01 -4.33465153e-01
1.30359247e-01 3.11462075e-01 -1.17414975e+00 1.47609338e-01
-3.72724980e-01 -3.60063702e-01 2.51787841e-01 8.38042200e-01
3.93842965e-01 -1.00914252e+00 -8.27883720e-01 -2.68971473e-01
-3.17931741e-01 -1.12014019e+00 -2.05539942e-01 4.46621269e-01
-1.01080894e+00 -1.07717264e+00 -3.31269294e-01 -7.22385406e-01
9.40583110e-01 1.65060371e-01 1.38409579e+00 -1.70942649e-01
-3.55336815e-01 2.86597073e-01 -1.18286349e-01 -2.06029922e-01
-1.15968771e-01 -2.78270155e-01 3.19739372e-01 -1.74004272e-01
4.34731871e-01 -8.12808335e-01 -6.57241762e-01 6.07131660e-01
-6.68921471e-01 1.02895677e-01 5.96114755e-01 7.45263219e-01
7.45382488e-01 -2.79644996e-01 -8.08480680e-02 -3.91019285e-01
1.05902754e-01 2.46644747e-02 -4.14458543e-01 6.46808371e-02
-2.44171590e-01 4.88677084e-01 4.33769494e-01 -1.94487080e-01
-5.57462811e-01 2.42324799e-01 1.82186961e-01 -1.31726071e-01
-2.57196277e-01 2.36696675e-01 -2.27809936e-01 -1.00854561e-01
1.03767514e+00 -2.86547720e-01 -1.24563135e-01 -3.04543614e-01
8.31114173e-01 5.16006708e-01 9.39149320e-01 -6.79929137e-01
6.41972125e-01 1.06518960e+00 3.76662165e-01 -5.10720968e-01
-1.04463017e+00 -4.55807656e-01 -1.16892350e+00 -1.15537651e-01
9.38705504e-01 -7.69480526e-01 -9.95989501e-01 5.45673549e-01
-1.06340694e+00 -4.83927846e-01 -4.30368632e-01 4.74472940e-01
-1.15929019e+00 3.33271533e-01 -5.87659001e-01 -5.30034423e-01
4.63233650e-01 -1.18926096e+00 1.38477957e+00 -6.08794950e-02
-4.61375207e-01 -9.79567587e-01 3.90619598e-02 1.02272652e-01
-1.11776628e-01 6.59339726e-01 7.50557840e-01 1.93679631e-01
-6.50686979e-01 1.06017683e-02 -5.44651687e-01 2.81906754e-01
3.33020538e-01 5.94066903e-02 -1.26914704e+00 -9.82248634e-02
4.58319634e-02 -7.16012537e-01 1.07569230e+00 4.28445607e-01
1.42702508e+00 -1.96890593e-01 2.65065990e-02 9.90822256e-01
1.20890975e+00 -4.37795758e-01 7.44372785e-01 4.23299611e-01
8.82309794e-01 9.45914328e-01 7.06712306e-01 4.03183043e-01
6.82532847e-01 6.29374266e-01 8.23811948e-01 -9.36428308e-02
-1.59557328e-01 -9.70373824e-02 3.02632451e-01 5.97746789e-01
-3.99099797e-01 2.75954396e-01 -7.66215146e-01 4.20442402e-01
-1.93635869e+00 -8.60009789e-01 1.36237234e-01 2.22029614e+00
9.86122966e-01 2.08927095e-01 2.29564473e-01 2.58138210e-01
2.36884907e-01 2.89839894e-01 -5.95415354e-01 -4.39221650e-01
-2.47561872e-01 -5.26887290e-02 4.88091826e-01 8.25934172e-01
-1.07971537e+00 5.79906464e-01 7.23351192e+00 2.93673068e-01
-7.59583175e-01 -3.67078692e-01 8.10425341e-01 -2.21525133e-01
-2.76818246e-01 -8.55202228e-03 -3.37969333e-01 1.44620165e-01
5.74386060e-01 1.96994379e-01 6.96204126e-01 6.86953366e-01
4.85011451e-02 -4.34395194e-01 -2.06996107e+00 1.49085712e+00
2.36297354e-01 -1.14305031e+00 -1.81469861e-02 1.25248104e-01
5.29018819e-01 1.56892702e-01 1.97238520e-01 -2.30978042e-01
4.44827139e-01 -1.37239170e+00 1.03219938e+00 8.94767284e-01
1.17458630e+00 -6.32261932e-01 4.95385438e-01 3.01750869e-01
-1.03489971e+00 -8.30724612e-02 -6.68428719e-01 -4.48853850e-01
-9.97145176e-02 5.20326197e-01 -6.04932904e-01 1.58079844e-02
9.04073715e-01 1.05704582e+00 -7.52239108e-01 7.42968440e-01
-5.16153812e-01 -1.80660561e-01 -5.69424212e-01 2.44223833e-01
-1.25561103e-01 -2.77470917e-01 1.47354066e-01 1.13262999e+00
1.62827477e-01 4.56868000e-02 -1.77762806e-01 8.27723503e-01
-7.34875426e-02 -3.56889546e-01 -7.43830442e-01 3.73647958e-01
3.37882698e-01 1.28333545e+00 -6.39818430e-01 -1.64808929e-01
-2.55871296e-01 9.23341513e-01 6.66336596e-01 2.41360888e-01
-5.10504961e-01 -3.04706424e-01 8.00811112e-01 1.09955743e-01
5.03673479e-02 -7.22239554e-01 -9.75521564e-01 -1.20535564e+00
1.70085713e-01 -8.74723911e-01 4.60213840e-01 -1.15131187e+00
-1.12573588e+00 3.66415799e-01 1.46973595e-01 -1.35270560e+00
-6.86024904e-01 -1.11244512e+00 -2.52570987e-01 8.03135633e-01
-1.25526440e+00 -1.04797995e+00 -7.17054307e-01 4.15159792e-01
2.48030007e-01 3.67374659e-01 9.89179015e-01 -3.55542868e-01
8.73005241e-02 4.20949459e-01 -1.62016854e-01 2.19323233e-01
8.04178894e-01 -1.84540510e+00 4.96099621e-01 7.54310489e-01
6.99107170e-01 5.64168155e-01 6.29305363e-01 -2.04256102e-01
-1.26169121e+00 -4.20941830e-01 7.28380263e-01 -1.09143412e+00
6.82629049e-01 -7.50481486e-01 -5.72549045e-01 5.34480274e-01
-1.16796568e-01 1.41923234e-01 5.87354064e-01 3.59918296e-01
-9.84183729e-01 -2.92773217e-01 -1.23299980e+00 5.76124370e-01
1.16470850e+00 -1.00347042e+00 -6.64845765e-01 1.98348939e-01
4.32453454e-01 -4.67646837e-01 -1.03555906e+00 2.75208741e-01
1.00378585e+00 -1.68023908e+00 1.17387807e+00 -3.34866524e-01
5.95970809e-01 -5.59294283e-01 -6.78181589e-01 -1.14892614e+00
-3.43769938e-01 -2.82479078e-01 -3.09553117e-01 6.20890200e-01
3.42262119e-01 -3.29599589e-01 8.62623751e-01 7.02334523e-01
2.59493440e-01 -8.02957773e-01 -9.56573606e-01 -6.31739438e-01
-4.80469503e-02 -5.80628753e-01 3.30566078e-01 7.69776285e-01
1.71885282e-01 1.34332329e-01 2.92257201e-02 8.47789273e-02
9.59439933e-01 3.74942236e-02 6.72306597e-01 -1.17570508e+00
-3.89288396e-01 -6.84635103e-01 -9.91802096e-01 -1.09516978e+00
-1.00664183e-01 -4.80716288e-01 2.99796343e-01 -1.56994581e+00
1.03066005e-01 -1.73558325e-01 -4.17371243e-01 5.96141815e-01
8.17354955e-03 8.56050909e-01 7.85265565e-02 3.12662303e-01
-6.53904438e-01 1.93879202e-01 1.02661109e+00 -9.23122838e-02
1.95183908e-03 -3.89024585e-01 -6.46109819e-01 1.14429152e+00
4.47506309e-01 -3.11083317e-01 -3.70877415e-01 -7.21008182e-01
5.07582366e-01 -1.86446816e-01 4.41973597e-01 -1.06535351e+00
1.73554197e-01 -3.21936578e-01 7.68432140e-01 -6.40263140e-01
5.95370770e-01 -7.18647242e-01 -3.44976604e-01 1.27358109e-01
-4.72551912e-01 2.91183591e-01 -2.60403246e-01 3.42478126e-01
-2.82059550e-01 -8.00454896e-03 8.68427753e-01 -2.19955206e-01
-8.86825681e-01 3.23639959e-01 1.28765196e-01 -1.23693477e-02
6.58961952e-01 -6.32385790e-01 -1.14772312e-01 -5.67234278e-01
-7.86271393e-01 1.76548716e-02 9.90139127e-01 1.42015025e-01
7.14064002e-01 -1.36294198e+00 -7.87987649e-01 1.30827457e-01
4.71577168e-01 2.55160540e-01 -3.07110697e-01 7.43400216e-01
-7.26219475e-01 -1.88767202e-02 -4.78473783e-01 -1.10078681e+00
-1.01965749e+00 4.37136292e-01 2.77607471e-01 1.51347905e-01
-2.32546955e-01 1.00678241e+00 4.94228929e-01 -4.16748881e-01
3.85551423e-01 -9.21122193e-01 3.38130561e-03 1.99167341e-01
5.34458876e-01 3.54787767e-01 -7.12713376e-02 -5.35990894e-01
-4.89569426e-01 1.06264949e+00 1.99810237e-01 -6.27428412e-01
1.28679419e+00 -3.73400867e-01 -5.48077583e-01 1.03374469e+00
1.32685459e+00 -2.93539196e-01 -1.70055330e+00 -2.24974230e-01
3.95157076e-02 -7.82267809e-01 -9.93755013e-02 -6.15441442e-01
-4.27559018e-01 1.09803259e+00 2.67100364e-01 1.72816023e-01
1.20653296e+00 2.22710133e-01 5.09528704e-02 6.60182118e-01
6.27939343e-01 -9.94897664e-01 3.83909464e-01 4.07052159e-01
1.10766530e+00 -1.62202060e+00 3.59080821e-01 -7.28835687e-02
-3.91708046e-01 1.24799967e+00 4.36450183e-01 -3.36225450e-01
4.69914109e-01 3.66678625e-01 2.27137387e-01 -1.05474286e-01
-5.41960657e-01 -2.76660830e-01 5.66812277e-01 8.40288639e-01
4.44559187e-01 1.40183166e-01 2.55793482e-01 -3.03332329e-01
-5.30668080e-01 -4.57299083e-01 6.02534056e-01 6.28182173e-01
-5.44396281e-01 -9.23251152e-01 -4.26390439e-01 3.30714434e-01
2.84306277e-02 -6.03111386e-02 -5.79074264e-01 7.70170629e-01
6.69749305e-02 8.35447609e-01 4.34534520e-01 -4.24269706e-01
9.40625966e-02 -3.51184815e-01 1.11744428e+00 -6.23289406e-01
2.35022828e-01 -2.81637013e-01 1.28066046e-02 -1.07053125e+00
-7.37910151e-01 -9.14698184e-01 -1.17329633e+00 -2.56968975e-01
1.28059447e-01 -2.83285350e-01 9.27154779e-01 9.15242910e-01
-1.48776725e-01 1.45353526e-01 7.99199343e-01 -1.18605626e+00
-3.56445283e-01 -8.67645979e-01 -5.92540085e-01 6.49478316e-01
6.84365273e-01 -6.21676266e-01 -6.23710811e-01 3.02307636e-01] | [8.42219066619873, -2.4956612586975098] |
8b31a9b9-1bb1-4e43-a908-cef2f41216fb | universal-denoising-networks-a-novel-cnn | 1711.07807 | null | http://arxiv.org/abs/1711.07807v2 | http://arxiv.org/pdf/1711.07807v2.pdf | Universal Denoising Networks : A Novel CNN Architecture for Image Denoising | We design a novel network architecture for learning discriminative image
models that are employed to efficiently tackle the problem of grayscale and
color image denoising. Based on the proposed architecture, we introduce two
different variants. The first network involves convolutional layers as a core
component, while the second one relies instead on non-local filtering layers
and thus it is able to exploit the inherent non-local self-similarity property
of natural images. As opposed to most of the existing deep network approaches,
which require the training of a specific model for each considered noise level,
the proposed models are able to handle a wide range of noise levels using a
single set of learned parameters, while they are very robust when the noise
degrading the latent image does not match the statistics of the noise used
during training. The latter argument is supported by results that we report on
publicly available images corrupted by unknown noise and which we compare
against solutions obtained by competing methods. At the same time the
introduced networks achieve excellent results under additive white Gaussian
noise (AWGN), which are comparable to those of the current state-of-the-art
network, while they depend on a more shallow architecture with the number of
trained parameters being one order of magnitude smaller. These properties make
the proposed networks ideal candidates to serve as sub-solvers on restoration
methods that deal with general inverse imaging problems such as deblurring,
demosaicking, superresolution, etc. | ['Stamatios Lefkimmiatis'] | 2017-11-21 | universal-denoising-networks-a-novel-cnn-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Lefkimmiatis_Universal_Denoising_Networks_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Lefkimmiatis_Universal_Denoising_Networks_CVPR_2018_paper.pdf | cvpr-2018-6 | ['color-image-denoising'] | ['computer-vision'] | [ 3.31750363e-01 -2.79688597e-01 3.39698702e-01 -2.13433966e-01
-5.95396638e-01 -2.14842975e-01 7.19911635e-01 -3.44196521e-02
-7.29844987e-01 6.07072175e-01 5.34806512e-02 8.31560120e-02
-4.56325769e-01 -7.16099977e-01 -6.46987200e-01 -1.13467181e+00
-4.07310165e-02 3.73849332e-01 2.08256885e-01 -5.19195318e-01
1.79057419e-01 7.43535042e-01 -1.74505293e+00 1.66525185e-01
8.98516357e-01 1.15344381e+00 2.61400223e-01 5.41827559e-01
6.28247485e-02 6.22912109e-01 -3.92515540e-01 -1.42241307e-02
2.72399187e-01 -4.40253049e-01 -4.39850211e-01 8.20787400e-02
6.01150393e-01 -1.82741672e-01 -3.25549185e-01 1.24462199e+00
6.12274289e-01 2.43517995e-01 5.86603999e-01 -4.36905354e-01
-6.19425058e-01 2.41777927e-01 -4.09633815e-01 4.17860985e-01
1.00512810e-01 -1.40119111e-03 5.93582392e-01 -8.16740215e-01
6.52091324e-01 1.06389129e+00 9.33280408e-01 5.03056705e-01
-1.45445073e+00 -1.71753421e-01 -5.64622954e-02 4.60723877e-01
-1.09162533e+00 -6.52476907e-01 8.35010648e-01 -2.36796021e-01
5.54334998e-01 2.33308926e-01 4.38605428e-01 1.27036321e+00
9.61673781e-02 3.03415895e-01 1.54663467e+00 -4.96083230e-01
3.28665674e-01 4.91035096e-02 -7.00463261e-03 4.21039790e-01
1.65810257e-01 1.18342645e-01 -2.43346050e-01 -4.73657958e-02
7.85833478e-01 -1.17987104e-01 -6.62279129e-01 -5.12709022e-01
-1.08988070e+00 7.29550183e-01 6.26801491e-01 9.60828662e-01
-5.77517927e-01 6.85809478e-02 4.04063165e-01 3.99203330e-01
5.80858827e-01 1.44523248e-01 -2.14322716e-01 3.47150117e-01
-1.22882211e+00 2.63352841e-01 7.36863077e-01 2.12598652e-01
8.11619580e-01 4.37745333e-01 4.66499515e-02 8.42491150e-01
1.52574703e-01 3.45220864e-01 6.61326468e-01 -8.05047274e-01
1.53592169e-01 1.05806023e-01 2.37629145e-01 -1.25538039e+00
-5.02761781e-01 -8.90943110e-01 -1.40954912e+00 6.29786134e-01
5.26991427e-01 1.30671009e-01 -1.02734458e+00 1.69496715e+00
1.02191478e-01 3.14020425e-01 1.10552467e-01 1.06548727e+00
6.43283308e-01 5.42923868e-01 -1.83253512e-01 -4.12031293e-01
1.16076732e+00 -5.73838830e-01 -9.20743704e-01 -2.34354481e-01
-3.55746644e-03 -9.13888097e-01 7.79608846e-01 8.09207559e-01
-1.09570229e+00 -7.67434955e-01 -1.07823992e+00 4.72488254e-02
-3.83060902e-01 2.10502550e-01 4.29096758e-01 7.61835217e-01
-1.48743045e+00 1.07939732e+00 -6.43927395e-01 -2.96798617e-01
3.38189125e-01 1.46139607e-01 -4.73437756e-01 -2.23263443e-01
-1.00308537e+00 1.07050312e+00 1.29945040e-01 7.74941802e-01
-9.18913603e-01 -2.82594621e-01 -5.80498815e-01 1.71238557e-01
1.73598602e-01 -5.63148677e-01 4.97647226e-01 -1.37444484e+00
-1.59984553e+00 9.08835769e-01 -3.60330753e-02 -5.30429959e-01
8.13196421e-01 -2.86671042e-01 -4.77501482e-01 3.12983066e-01
-2.66331851e-01 4.30138074e-02 1.37817121e+00 -1.58213651e+00
8.90424177e-02 -3.48457903e-01 -8.12041014e-02 -1.56896994e-01
-3.13451260e-01 -2.80837603e-02 -2.31903404e-01 -9.80325818e-01
3.74842077e-01 -6.01402879e-01 -5.01320064e-01 -6.98925555e-02
-1.61554307e-01 3.05643916e-01 7.70601869e-01 -9.01165307e-01
7.59026468e-01 -2.10667348e+00 5.32351136e-01 2.49994189e-01
-4.02645990e-02 5.34049034e-01 -3.26986611e-01 4.11794424e-01
-4.43973660e-01 -1.77911505e-01 -5.69582820e-01 -4.87714440e-01
-2.14718074e-01 6.65016994e-02 -1.78651750e-01 8.47276628e-01
-7.80588388e-03 3.97856206e-01 -5.55408359e-01 -8.15426186e-02
4.41478133e-01 8.55713785e-01 -3.29210818e-01 1.29403830e-01
-1.54424310e-02 6.20960414e-01 -1.47597209e-01 1.58618450e-01
9.92418528e-01 1.09780848e-01 1.14967145e-01 -5.22481501e-01
-1.48796856e-01 -1.97021887e-01 -1.58622575e+00 1.62044251e+00
-5.91728806e-01 4.68939722e-01 6.42982543e-01 -1.58139384e+00
1.00481963e+00 4.12111431e-01 4.73540664e-01 -6.28336191e-01
1.56045064e-01 5.09416223e-01 -1.30810305e-01 -4.89475936e-01
9.05214325e-02 -4.19867247e-01 5.66772342e-01 2.24487260e-01
2.46402293e-01 5.09060882e-02 2.12866440e-01 -1.73416555e-01
1.03248274e+00 1.44211262e-01 1.29710153e-01 -5.29240608e-01
1.02347922e+00 -4.20989215e-01 4.38201725e-01 9.04739797e-01
1.90610699e-02 8.96052241e-01 3.08158249e-01 -7.62164772e-01
-1.05793262e+00 -8.68973970e-01 -2.27368668e-01 6.61303997e-01
7.76785612e-02 5.35141602e-02 -8.46137464e-01 -3.84198993e-01
-4.93842453e-01 3.40344131e-01 -6.33333921e-01 1.52186632e-01
-8.94646943e-01 -1.20224547e+00 2.42220685e-01 1.00553825e-01
7.46946692e-01 -1.06201911e+00 -4.31962609e-01 2.66483754e-01
-1.46528691e-01 -1.26640105e+00 1.48328140e-01 3.42302978e-01
-1.05689323e+00 -1.09470510e+00 -1.03921449e+00 -7.90192425e-01
4.43195373e-01 1.27631739e-01 1.17885303e+00 2.55029976e-01
-3.64564955e-01 3.71270895e-01 -3.50231618e-01 1.57355398e-01
-4.93489921e-01 -2.69608527e-01 -1.21305257e-01 4.57043409e-01
-1.99791431e-01 -1.08999479e+00 -7.33008206e-01 7.02132061e-02
-1.28766263e+00 -2.26854101e-01 7.29437590e-01 1.07225084e+00
3.82261813e-01 3.34013969e-01 3.76162887e-01 -8.05523396e-01
4.56977934e-01 -3.18186522e-01 -5.85831583e-01 2.33356934e-02
-4.21391606e-01 1.38485029e-01 1.02428305e+00 -3.91620934e-01
-1.22117555e+00 1.37485906e-01 -5.31628191e-01 -1.68329343e-01
-4.66242999e-01 2.82012612e-01 -2.47244537e-02 -4.97196883e-01
6.96270943e-01 4.82048392e-01 4.31202240e-02 -9.56810236e-01
1.58736199e-01 1.36905223e-01 6.19177222e-01 -2.68763423e-01
1.09594703e+00 7.96649754e-01 2.76979059e-01 -1.03955066e+00
-4.88398314e-01 -3.65801334e-01 -5.74654281e-01 -6.00814410e-02
8.91072214e-01 -7.63726413e-01 -5.61072588e-01 8.17653239e-01
-1.16593206e+00 -1.49880484e-01 -2.33460799e-01 3.38015199e-01
-5.64630747e-01 7.96344817e-01 -7.57998407e-01 -7.41862833e-01
-2.41408214e-01 -1.09985197e+00 7.24106252e-01 3.08148284e-02
4.64821488e-01 -1.11244380e+00 7.32020885e-02 2.21554711e-01
9.83111084e-01 3.05680305e-01 9.78183925e-01 -3.45315903e-01
-4.35407549e-01 -3.37365150e-01 -1.83234468e-01 9.95804191e-01
-7.66183287e-02 -3.84450197e-01 -1.04176962e+00 -5.42360485e-01
6.67970836e-01 -2.15886101e-01 1.23586702e+00 5.62904716e-01
1.11761522e+00 -1.14455454e-01 1.38887316e-01 7.75829136e-01
1.85089612e+00 -3.68009061e-01 9.81416881e-01 5.37161589e-01
3.74549121e-01 5.60746729e-01 4.15290222e-02 2.42831036e-01
-2.07743347e-01 9.46570158e-01 8.51106465e-01 -5.26903808e-01
-3.29709411e-01 3.81076247e-01 1.87380210e-01 5.27228713e-01
-4.87603605e-01 -1.26574561e-01 -6.00551248e-01 5.17134547e-01
-1.73547697e+00 -1.00800407e+00 -3.57344866e-01 2.27571058e+00
4.94349360e-01 7.30357468e-02 -1.66655570e-01 4.56714779e-01
5.29858112e-01 5.08898616e-01 -1.30630359e-01 -2.26322725e-01
-5.59410989e-01 6.63493335e-01 4.42095280e-01 5.64307511e-01
-1.24616051e+00 6.33766770e-01 6.31368160e+00 9.17511940e-01
-1.09784627e+00 2.62842894e-01 4.81913567e-01 3.47602189e-01
-1.68535978e-01 -1.53892905e-01 -1.15559861e-01 3.36447001e-01
7.94505537e-01 5.19417226e-01 5.63310206e-01 4.71797198e-01
4.10306185e-01 -1.83352888e-01 -6.44762158e-01 9.88779068e-01
2.46806100e-01 -1.17435551e+00 -7.41409836e-03 -1.97459787e-01
7.06365466e-01 -6.35152683e-03 1.23934790e-01 -1.39451191e-01
-1.58393607e-01 -1.19618309e+00 4.74501699e-01 7.97348499e-01
2.43747070e-01 -5.11490345e-01 1.06441903e+00 3.44893306e-01
-7.12669909e-01 -2.59172142e-01 -5.09576738e-01 2.61267498e-02
2.16576353e-01 1.04837418e+00 9.69502628e-02 8.28286469e-01
9.42244649e-01 5.24710238e-01 -6.19843543e-01 1.04977369e+00
-1.68117300e-01 5.08783102e-01 -2.02738971e-01 5.23635626e-01
3.11805874e-01 -4.74229813e-01 7.44546354e-01 1.16981018e+00
4.85333532e-01 -2.33631581e-01 -1.88934624e-01 8.10512483e-01
1.64591447e-01 2.05210611e-01 -3.36976945e-01 4.17654783e-01
-2.72135437e-01 1.34192085e+00 -7.37990379e-01 -2.50669718e-01
-3.12574416e-01 1.18087971e+00 4.65568565e-02 6.64487600e-01
-4.30924594e-01 -1.40334651e-01 5.00214815e-01 9.73337963e-02
6.38896465e-01 -3.27807784e-01 -1.19487152e-01 -1.30960763e+00
1.18057750e-01 -1.12190092e+00 8.09782892e-02 -6.37895107e-01
-1.30435085e+00 9.91463304e-01 -2.31236279e-01 -1.11573195e+00
-2.82278359e-01 -9.36990440e-01 -6.48843765e-01 1.08647954e+00
-1.87072968e+00 -1.05638635e+00 -4.17967707e-01 6.85371995e-01
2.91124880e-01 -1.99322239e-01 9.95133698e-01 4.06964093e-01
-3.21752459e-01 -6.29452616e-03 4.96420205e-01 -1.43573076e-01
6.30605757e-01 -1.12244809e+00 1.18156530e-01 1.17393303e+00
1.32016599e-01 5.58774352e-01 1.14028764e+00 -2.96335518e-01
-1.20898259e+00 -5.76091945e-01 5.77929854e-01 1.31419390e-01
5.05367398e-01 -2.41797403e-01 -1.12718189e+00 1.89586937e-01
3.26651096e-01 4.78015840e-01 1.76106200e-01 -1.47074401e-01
-3.92096370e-01 -3.56292874e-01 -1.17326224e+00 1.18221648e-01
7.21274257e-01 -2.72055507e-01 -3.90102506e-01 3.14758748e-01
-2.05965643e-03 -1.95090249e-01 -5.19915819e-01 4.51034814e-01
3.00942302e-01 -1.71314633e+00 1.23824298e+00 -3.50421786e-01
3.55942816e-01 -3.23239088e-01 -3.00311763e-02 -1.30524981e+00
-3.67396086e-01 -4.73295599e-01 1.07361861e-01 8.92988563e-01
3.62085849e-02 -6.65672243e-01 5.07008791e-01 -1.00813940e-01
-6.45815060e-02 -3.98887932e-01 -1.14347649e+00 -7.26828456e-01
-6.79941848e-02 -2.44128510e-01 5.80311418e-02 8.02140474e-01
-8.05785596e-01 -1.14632495e-01 -7.07039773e-01 2.23024070e-01
1.03696561e+00 4.31802459e-02 3.92098308e-01 -1.28416133e+00
-4.82299209e-01 -4.78401005e-01 -5.46105742e-01 -7.48772800e-01
1.93022951e-01 -4.75630194e-01 1.25390083e-01 -1.39363337e+00
-5.76539896e-02 -3.21792215e-01 -3.53043854e-01 2.00304940e-01
1.27212191e-02 7.83374488e-01 5.12749888e-02 1.53373361e-01
-9.64937210e-02 5.21651328e-01 9.36349869e-01 -8.10301974e-02
8.99736434e-02 2.34630704e-01 -4.66098338e-01 9.83075857e-01
5.72320282e-01 -3.37915987e-01 -3.83029766e-02 -4.93519187e-01
1.73642054e-01 -1.91002220e-01 7.42771626e-01 -1.25725865e+00
1.66066810e-01 2.49336973e-01 5.69281220e-01 -8.41966718e-02
4.18765843e-01 -1.06788111e+00 2.33416244e-01 4.92351830e-01
-1.22149594e-01 -1.98527083e-01 9.10965204e-02 6.76786482e-01
-4.50461179e-01 -5.33142805e-01 1.16166449e+00 -4.28606719e-01
-6.53286874e-01 -8.00002962e-02 -4.97933388e-01 -3.86932462e-01
5.42709172e-01 -1.54160082e-01 -9.51885507e-02 -5.67503333e-01
-9.70039964e-01 -5.55900812e-01 4.42985475e-01 1.16367370e-01
5.34540355e-01 -1.02753425e+00 -8.54261160e-01 2.28830740e-01
-3.78255814e-01 -4.59825456e-01 7.23081529e-01 1.12118387e+00
-6.60346091e-01 -5.63010052e-02 -4.66909289e-01 -4.61434245e-01
-1.02742171e+00 5.92848182e-01 5.80007851e-01 -4.31248933e-01
-8.69610548e-01 5.32818973e-01 1.59758963e-02 -1.77031189e-01
1.22424878e-01 -4.07612137e-02 -4.06610787e-01 -1.08594541e-02
6.29133105e-01 4.32364076e-01 4.81903613e-01 -8.19351614e-01
-2.49674454e-01 8.30784380e-01 2.63568133e-01 6.79406226e-02
1.90497386e+00 -1.27119064e-01 -4.98679042e-01 1.56441227e-01
1.07494879e+00 7.70758763e-02 -1.13451719e+00 -2.74256766e-01
3.42603438e-02 -4.49298650e-01 4.77016300e-01 -7.02091098e-01
-1.36958957e+00 8.65192771e-01 1.11662936e+00 3.10661167e-01
1.59411597e+00 -4.35931623e-01 3.90972853e-01 2.59290695e-01
1.18919738e-01 -9.24529910e-01 1.36885315e-01 3.27168912e-01
1.15589142e+00 -1.17320192e+00 1.64226323e-01 -3.19066614e-01
-1.12996727e-01 1.44439089e+00 4.22700457e-02 -5.87864697e-01
5.65707684e-01 2.97597796e-01 1.93895936e-01 -6.32210895e-02
-1.28774941e-01 -4.77545798e-01 3.50879669e-01 6.74126208e-01
3.01763445e-01 -4.43567157e-01 -5.31900585e-01 1.82587385e-01
3.54307503e-01 -2.19269600e-02 5.39478481e-01 5.63433111e-01
-3.44642490e-01 -1.25802720e+00 -7.38661706e-01 -4.22021262e-02
-6.20879829e-01 -1.14822008e-01 5.39195491e-03 7.54994631e-01
2.40836099e-01 8.64221513e-01 -4.09749359e-01 1.04064852e-01
3.62951636e-01 -3.55337374e-02 5.96060991e-01 -2.00292572e-01
-6.60439074e-01 2.78071135e-01 -1.75248116e-01 -7.18079150e-01
-9.22808468e-01 -4.42807227e-01 -5.39286673e-01 -5.89149594e-02
-1.82436883e-01 -1.05669849e-01 6.99797988e-01 1.04668343e+00
-1.12077417e-02 4.85156417e-01 5.16875267e-01 -1.45654237e+00
-7.06519604e-01 -1.05054319e+00 -7.52164841e-01 6.54405773e-01
6.81482613e-01 -5.58381200e-01 -6.51668847e-01 1.09051250e-01] | [11.494464874267578, -2.3639068603515625] |
5efb4ef2-4c62-4f7a-8763-00338ec37fd6 | anise-assembly-based-neural-implicit-surface | 2205.13682 | null | https://arxiv.org/abs/2205.13682v2 | https://arxiv.org/pdf/2205.13682v2.pdf | ANISE: Assembly-based Neural Implicit Surface rEconstruction | We present ANISE, a method that reconstructs a 3D~shape from partial observations (images or sparse point clouds) using a part-aware neural implicit shape representation. The shape is formulated as an assembly of neural implicit functions, each representing a different part instance. In contrast to previous approaches, the prediction of this representation proceeds in a coarse-to-fine manner. Our model first reconstructs a structural arrangement of the shape in the form of geometric transformations of its part instances. Conditioned on them, the model predicts part latent codes encoding their surface geometry. Reconstructions can be obtained in two ways: (i) by directly decoding the part latent codes to part implicit functions, then combining them into the final shape; or (ii) by using part latents to retrieve similar part instances in a part database and assembling them in a single shape. We demonstrate that, when performing reconstruction by decoding part representations into implicit functions, our method achieves state-of-the-art part-aware reconstruction results from both images and sparse point clouds.When reconstructing shapes by assembling parts retrieved from a dataset, our approach significantly outperforms traditional shape retrieval methods even when significantly restricting the database size. We present our results in well-known sparse point cloud reconstruction and single-view reconstruction benchmarks. | ['Evangelos Kalogerakis', 'Radomir Mech', 'Matheus Gadelha', 'Dmitry Petrov'] | 2022-05-27 | null | null | null | null | ['point-cloud-reconstruction'] | ['computer-vision'] | [ 3.32217485e-01 1.94024965e-01 5.46981115e-03 -3.35210830e-01
-1.03453267e+00 -8.79243135e-01 7.05081105e-01 -3.25201303e-02
4.38983291e-01 1.92485258e-01 3.87600332e-01 1.19725451e-01
-1.18400333e-02 -1.18867052e+00 -1.52134907e+00 -5.30514061e-01
3.13615799e-01 1.43380618e+00 1.50417805e-01 -3.01016085e-02
3.68807763e-01 1.30216873e+00 -1.80286634e+00 6.07494295e-01
1.85489938e-01 1.16085017e+00 3.00039768e-01 4.90873128e-01
-7.00335383e-01 2.70912468e-01 -3.03807054e-02 -2.95645386e-01
5.61101377e-01 1.95229977e-01 -8.34434986e-01 5.51479876e-01
7.73069024e-01 -4.02767628e-01 -3.08278322e-01 8.65554452e-01
4.07234989e-02 -1.41174555e-01 9.75623667e-01 -8.10063064e-01
-7.52037108e-01 3.33895832e-02 -3.82743239e-01 -6.78162575e-01
4.26357269e-01 -1.25999168e-01 1.07299256e+00 -1.44704008e+00
9.26039577e-01 1.42156959e+00 8.24350119e-01 1.62627116e-01
-1.92302895e+00 -3.76086503e-01 -1.85345914e-02 -4.83110189e-01
-1.63929939e+00 -7.99717247e-01 9.24537480e-01 -5.24043739e-01
1.34097075e+00 2.75257766e-01 7.92518675e-01 2.75979042e-01
-2.42937710e-02 7.06032932e-01 5.76971889e-01 -1.04624897e-01
1.71667889e-01 -2.50805736e-01 -2.47617811e-01 8.37871432e-01
1.44283324e-01 -8.55268165e-02 -4.45392847e-01 -7.57735789e-01
1.36196184e+00 4.40407634e-01 -2.77986974e-01 -9.24630344e-01
-1.24687195e+00 5.99367380e-01 6.38797641e-01 -9.15427133e-02
-6.84888363e-01 5.78096509e-01 -6.76266616e-03 2.92990446e-01
7.30531335e-01 4.48660292e-02 -6.50771976e-01 1.08459577e-01
-9.12409246e-01 4.78829920e-01 9.64405715e-01 1.35171497e+00
1.26752687e+00 -6.60628378e-02 2.90988773e-01 9.21688437e-01
4.57057923e-01 9.35561121e-01 -6.37406632e-02 -1.26639640e+00
4.51190799e-01 7.69943595e-01 1.15765095e-01 -8.61098826e-01
1.87615409e-01 -3.49932201e-02 -6.59891725e-01 1.53571561e-01
-2.61127502e-01 6.89111412e-01 -1.10258591e+00 1.35980582e+00
2.83539295e-01 2.65388876e-01 -2.80453980e-01 6.08234763e-01
8.94740283e-01 8.27430069e-01 -3.59871238e-01 1.18221626e-01
1.18398499e+00 -5.61092257e-01 -1.57490626e-01 1.43144622e-01
1.90036714e-01 -9.25054193e-01 3.39607894e-01 4.92285430e-01
-1.44306624e+00 -5.44463098e-01 -7.54172862e-01 -3.73497844e-01
-8.12418386e-02 -3.58678214e-02 4.14359689e-01 3.97980660e-02
-1.49414337e+00 8.32333565e-01 -9.38377321e-01 -2.17892043e-03
5.04950106e-01 5.89323640e-01 -7.47613490e-01 -4.14532125e-01
4.54840623e-02 3.19473326e-01 7.10089412e-03 -3.07794750e-01
-7.63162911e-01 -7.34806895e-01 -9.65550423e-01 1.39637798e-01
4.91358619e-03 -1.19659686e+00 1.14860702e+00 -8.01297545e-01
-1.22329831e+00 1.02390289e+00 -6.79149032e-01 -9.08123180e-02
-1.38698384e-01 -8.83431509e-02 1.32785067e-01 3.06418598e-01
1.08937234e-01 9.36528683e-01 1.13209844e+00 -1.91043139e+00
-8.57484937e-02 -4.82973069e-01 -1.46399260e-01 2.47857675e-01
4.73301083e-01 -4.25390095e-01 -8.13693762e-01 -5.39298773e-01
1.03640103e+00 -8.49685311e-01 -1.36562914e-01 7.66885877e-01
-3.37360352e-01 -1.16268825e-02 9.25140917e-01 -5.40144920e-01
4.54940677e-01 -2.08408093e+00 4.77030694e-01 4.48252350e-01
4.19004440e-01 -1.54580370e-01 -3.87907028e-01 5.53243697e-01
-1.98639840e-01 -2.85353437e-02 -5.39794981e-01 -1.00169885e+00
7.49634728e-02 6.71772063e-01 -8.22939157e-01 5.57444036e-01
2.62576759e-01 1.04128492e+00 -5.74302256e-01 -1.22117668e-01
2.23085836e-01 7.03546405e-01 -9.58265066e-01 3.19496304e-01
-5.07592261e-01 2.41780147e-01 -5.26756585e-01 9.84931886e-01
9.25068200e-01 -3.77480030e-01 -4.67946492e-02 -4.58147854e-01
3.80796241e-03 3.57824504e-01 -1.05100310e+00 1.98137808e+00
-3.58022809e-01 1.62229761e-01 4.44881052e-01 -1.14462781e+00
1.04312396e+00 4.84652638e-01 8.83392334e-01 -2.12390989e-01
-3.04849654e-01 3.97241235e-01 -7.40716994e-01 1.07293678e-02
5.82959890e-01 -5.36797941e-01 1.72434926e-01 6.34038448e-01
1.14334390e-01 -8.54689240e-01 -5.14687598e-01 1.47199422e-01
1.06020451e+00 5.79437494e-01 1.50743946e-01 8.72688368e-03
1.55383185e-01 1.23089217e-01 2.31514946e-02 3.73395354e-01
7.17210352e-01 1.18685043e+00 4.86397184e-02 -8.20736766e-01
-1.38802230e+00 -1.58249664e+00 -2.72742361e-01 5.13580739e-01
-9.42319706e-02 -4.40878212e-01 -2.83832014e-01 -2.71461785e-01
5.24666846e-01 5.09709060e-01 -5.85458696e-01 1.01734675e-01
-6.81765020e-01 -3.57964006e-03 1.24956690e-01 4.99874860e-01
-9.39866528e-02 -1.14668429e+00 -1.44032061e-01 2.34303012e-01
-3.44183445e-02 -9.04845715e-01 -4.98088449e-01 2.85430551e-01
-1.53567231e+00 -9.08852577e-01 -5.48849583e-01 -8.18950355e-01
1.26379466e+00 6.79416239e-01 1.61074173e+00 4.06570345e-01
-4.41169627e-02 8.33642721e-01 -1.71505749e-01 4.52335039e-03
-4.61368471e-01 -4.32069480e-01 1.83977839e-02 1.44837126e-02
-6.18678145e-02 -1.23963916e+00 -4.77424532e-01 1.53719544e-01
-1.10862434e+00 1.97096422e-01 7.92751789e-01 5.75842500e-01
1.57457757e+00 -3.85928303e-01 -1.09414622e-01 -9.04836893e-01
-1.28901722e-02 -5.76609552e-01 -5.62013626e-01 8.25641677e-02
-8.65221322e-02 4.49671626e-01 5.14156163e-01 -2.13376120e-01
-6.28601313e-01 5.90345621e-01 -3.57699156e-01 -1.22752988e+00
-3.69584858e-01 3.49978834e-01 -7.20309392e-02 -1.74217850e-01
3.90212506e-01 7.77793109e-01 3.66312146e-01 -1.04304612e+00
5.80778241e-01 2.12654963e-01 5.81983447e-01 -1.08383560e+00
1.06811941e+00 7.66670406e-01 1.86956123e-01 -9.65960622e-01
-4.88654494e-01 -6.62086427e-01 -9.72799361e-01 2.47396722e-01
5.25378108e-01 -1.02239227e+00 -4.04619694e-01 1.30307272e-01
-1.42563689e+00 -1.00926541e-01 -6.80457294e-01 -1.83851585e-01
-1.07095218e+00 4.87595558e-01 -5.90752065e-01 -4.26561892e-01
-4.28418368e-01 -1.10278654e+00 2.03245926e+00 -6.03109658e-01
-4.02629524e-02 -5.83007753e-01 2.95406908e-01 7.54130483e-02
1.22528121e-01 1.66722387e-01 1.04057324e+00 -2.02159628e-01
-1.21360648e+00 -2.05000848e-01 -2.47198448e-01 1.55509472e-01
1.20243736e-01 -2.34404072e-01 -6.45499408e-01 -3.58009398e-01
-5.17585129e-03 -2.44092926e-01 5.96066952e-01 4.48239207e-01
1.37586701e+00 -5.05992711e-01 -5.26929021e-01 9.99967456e-01
1.77201641e+00 -2.07071379e-01 5.57800889e-01 -2.16740981e-01
8.22886169e-01 2.18817398e-01 -1.24958260e-02 4.19994324e-01
3.58950853e-01 7.49467671e-01 7.37026513e-01 1.95874676e-01
-3.33605379e-01 -6.06926382e-01 1.70562312e-01 1.14903104e+00
-2.50431448e-01 2.10242137e-01 -6.74792051e-01 5.80106258e-01
-1.64117634e+00 -7.71274328e-01 -1.57401130e-01 2.24842501e+00
5.48354745e-01 -2.93897271e-01 -1.41578436e-01 -1.33934572e-01
2.72578031e-01 1.14624895e-01 -5.63199341e-01 -2.55485773e-01
-1.35419872e-02 6.03489578e-01 3.77359241e-01 3.95784944e-01
-6.40977025e-01 7.09608436e-01 6.77054644e+00 5.58522403e-01
-8.40641201e-01 3.57973129e-02 2.88545012e-01 -1.00539590e-03
-1.05080950e+00 1.93680644e-01 -5.39469421e-01 -1.61731541e-01
6.48052156e-01 2.72979468e-01 8.21625292e-01 8.48823965e-01
-3.99234802e-01 3.08442354e-01 -1.25147331e+00 1.21855319e+00
1.64355233e-01 -1.84929264e+00 7.08142817e-01 4.87166077e-01
6.45640433e-01 5.15174389e-01 -2.94399887e-01 2.74648629e-02
3.14529181e-01 -1.02579641e+00 1.29197991e+00 9.29664314e-01
1.19006228e+00 -4.31037128e-01 1.12053595e-01 5.80678403e-01
-1.46383142e+00 3.91243339e-01 -7.95584977e-01 1.60656363e-01
1.73062816e-01 3.75519633e-01 -7.48742223e-01 6.50019109e-01
4.91442204e-01 9.49019551e-01 -9.84153673e-02 9.45361137e-01
2.92516887e-01 3.60581517e-01 -6.99777901e-01 4.59062904e-01
-1.53020754e-01 -5.32333016e-01 6.39035583e-01 7.55737424e-01
6.44919753e-01 5.04062831e-01 3.30881506e-01 1.44488549e+00
-2.12473914e-01 -2.13855922e-01 -1.01651311e+00 -7.92756826e-02
4.37447429e-01 9.85379696e-01 -6.05199039e-01 -3.87453586e-01
-4.26262766e-01 1.01051688e+00 6.45483792e-01 2.95978695e-01
-2.12879375e-01 2.90374011e-01 7.51361907e-01 5.31219184e-01
8.76024246e-01 -4.29895490e-01 -4.37923104e-01 -1.04061246e+00
-1.59649216e-02 -5.07355511e-01 -2.84274906e-01 -1.17600596e+00
-1.38110745e+00 6.02023840e-01 1.48571879e-01 -1.52509034e+00
-1.34333447e-01 -6.73842549e-01 -2.48141915e-01 1.06488979e+00
-1.16671288e+00 -1.25841403e+00 1.43670300e-02 6.42253101e-01
6.67321026e-01 -1.41008407e-01 1.23581016e+00 -2.23202817e-02
3.62386078e-01 -1.11351766e-01 8.51779953e-02 -1.16721988e-01
5.80765605e-02 -1.07098603e+00 8.33278954e-01 2.24448860e-01
6.48922026e-01 8.09751928e-01 3.26386333e-01 -7.30902076e-01
-2.01049709e+00 -8.88222516e-01 4.80814844e-01 -6.36293292e-01
1.46147072e-01 -3.47589374e-01 -9.81227219e-01 9.96546805e-01
-3.61995190e-01 5.08256197e-01 3.60253900e-01 -8.71905684e-02
-6.42322302e-01 1.98106349e-01 -1.11602879e+00 2.27525905e-01
1.26194561e+00 -8.89805198e-01 -7.97666550e-01 2.69045025e-01
8.50375533e-01 -5.72383404e-01 -1.03704739e+00 4.10086572e-01
6.33412838e-01 -8.65372121e-01 1.58135390e+00 -3.77811879e-01
5.64316273e-01 -3.86096507e-01 -8.58139455e-01 -1.12434745e+00
-6.65988266e-01 -1.11812979e-01 -4.64325935e-01 5.26983380e-01
6.06833920e-02 -4.77839440e-01 9.22417939e-01 4.35942173e-01
-5.43998182e-01 -1.03949499e+00 -9.11900282e-01 -3.93587589e-01
-2.55200900e-02 -5.09809077e-01 9.84246731e-01 6.66189551e-01
-8.52787018e-01 7.65942112e-02 -1.49744907e-02 3.20079744e-01
6.62649632e-01 8.45507920e-01 9.54979777e-01 -1.52507150e+00
-4.48578477e-01 -3.97625178e-01 -3.57807606e-01 -1.68635142e+00
1.93558320e-01 -1.13484490e+00 2.27476239e-01 -1.75230587e+00
2.07574978e-01 -7.36769080e-01 1.53116778e-01 7.42011786e-01
5.23371756e-01 5.57074726e-01 7.94065148e-02 7.32546508e-01
-3.28926086e-01 7.13155329e-01 1.25768638e+00 -2.74846107e-01
2.09752604e-01 1.40551636e-02 -4.87885773e-01 6.40433550e-01
1.34307891e-01 -5.75769603e-01 -2.32601121e-01 -6.19535565e-01
2.92756557e-01 6.26411736e-01 4.45408911e-01 -8.78950119e-01
8.04237574e-02 8.91507603e-03 5.17934620e-01 -1.20305479e+00
1.08743441e+00 -1.14536715e+00 9.25778508e-01 2.56693751e-01
1.98244050e-01 1.78048447e-01 2.80967951e-01 8.18337739e-01
-2.03893736e-01 -1.52913913e-01 3.12682211e-01 -5.93986332e-01
-2.00267404e-01 7.79944062e-01 7.34030902e-02 -4.48626488e-01
1.88258141e-01 -6.13491058e-01 1.80738136e-01 -3.21707696e-01
-7.71570742e-01 -3.41713309e-01 9.99670208e-01 2.83568919e-01
1.14622772e+00 -1.71835279e+00 -6.11442566e-01 7.81051278e-01
1.21804982e-01 6.25339806e-01 7.23608062e-02 4.16573316e-01
-7.71504581e-01 4.89010245e-01 -6.75071329e-02 -9.76292253e-01
-1.03060114e+00 4.67666686e-01 1.66300371e-01 5.43175340e-02
-1.07768595e+00 7.19810784e-01 6.86608374e-01 -9.48737800e-01
-1.14357583e-01 -4.29042131e-01 4.08513963e-01 -4.90282744e-01
9.14580598e-02 -2.44311854e-01 2.55129337e-01 -1.23436594e+00
-1.14953548e-01 1.09640622e+00 3.37600201e-01 -7.65760541e-02
1.84427500e+00 3.54896367e-01 -6.73194945e-01 3.92654300e-01
1.43957400e+00 3.08658779e-01 -1.27397990e+00 -4.87149835e-01
-5.65537870e-01 -5.55697918e-01 -1.31247174e-02 -4.94734645e-01
-1.08374500e+00 5.51269948e-01 -3.48357260e-02 -2.53940485e-02
8.23957324e-01 4.97129798e-01 9.08779740e-01 5.54170609e-01
9.63128328e-01 -2.90501446e-01 -6.64578527e-02 6.06864095e-01
1.46178150e+00 -7.07229733e-01 3.39033097e-01 -6.72192454e-01
1.45486807e-02 1.18680823e+00 -1.28859565e-01 -6.44427717e-01
1.08714342e+00 1.91837311e-01 -4.85778064e-01 -8.21658611e-01
-8.28982234e-01 1.07351132e-01 8.17418575e-01 4.64604080e-01
2.01641142e-01 3.53112161e-01 4.65308845e-01 1.95286945e-01
-2.15984792e-01 -1.52774274e-01 -1.40498513e-02 8.78143907e-01
-5.35435796e-01 -1.29257405e+00 -7.02410579e-01 8.20594192e-01
1.04092665e-01 -1.99182574e-02 -4.66072023e-01 3.63087326e-01
-1.72925647e-03 5.56747392e-02 3.64854693e-01 -2.37185925e-01
4.69629765e-01 1.29761249e-01 6.79671168e-01 -1.03387225e+00
-1.52376786e-01 1.91902012e-01 -7.01822713e-02 -9.09165025e-01
-3.11301917e-01 -9.47701275e-01 -1.22673476e+00 -1.84695035e-01
-2.96007842e-03 -8.86394307e-02 7.47440338e-01 8.05726767e-01
6.41306520e-01 5.45656569e-02 4.21053767e-01 -1.74470532e+00
-4.27171379e-01 -6.28251255e-01 -6.31154597e-01 5.58259487e-01
3.36716652e-01 -8.74489725e-01 -1.60626635e-01 1.04560032e-01] | [8.65503978729248, -3.6074912548065186] |
c24b42f1-f6c5-442f-a49d-004ba6af6771 | predicting-soil-properties-from-hyperspectral | null | null | https://ieeexplore.ieee.org/abstract/document/9897254 | https://github.com/ridvansalihkuzu/hyperview_eagleeyes/blob/master/challenge_submission_eagleeyes/hyperview_for_ICIP_camera_ready_eagleeyes.pdf | Predicting Soil Properties from Hyperspectral Satellite Images | The AI4EO HYPERVIEW challenge seeks machine learning
methods that predict agriculturally relevant soil parameters
(K, Mg, P2O5, pH) from airborne hyperspectral images.
We present a hybrid model fusing Random Forest and K-
nearest neighbor regressors that exploit the average spectral
reflectance, as well as derived features such as gradients,
wavelet coefficients, and Fourier transforms. The solution is
computationally lightweight and improves upon the challenge
baseline by 21.9%, with the first place on the public leader-
board. In addition, we discuss neural network architectures
and potential future improvements. | ['Roshni Kamath', 'Caroline Arnold', 'Frauke Albrecht', 'Rıdvan Salih Kuzu'] | 2022-10-18 | null | null | null | conference-2022-10 | ['seeing-beyond-the-visible'] | ['computer-vision'] | [ 6.94632113e-01 -9.70689654e-02 -4.45391834e-01 -4.33307052e-01
-5.21906376e-01 -6.91619039e-01 3.44461709e-01 1.09155692e-01
-9.89821106e-02 1.11122370e+00 1.48634464e-01 -7.23207593e-01
-5.75142086e-01 -1.24434757e+00 -6.65182173e-01 -8.46758425e-01
-3.93659055e-01 -3.21853943e-02 -1.31460249e-01 -5.11363685e-01
5.47593720e-02 6.03455544e-01 -1.68928730e+00 5.23528576e-01
1.00389850e+00 1.28135216e+00 2.33367532e-01 7.54855990e-01
4.45870548e-01 5.14711201e-01 1.24033153e-01 3.31023276e-01
5.78223288e-01 2.74198145e-01 -7.80628979e-01 -2.87569314e-01
6.94087029e-01 -6.48119748e-01 -5.52257895e-02 9.72108066e-01
4.11917627e-01 -2.68670693e-02 8.31325889e-01 -1.06837487e+00
-4.11261052e-01 6.14353955e-01 -9.00721550e-01 -3.77520137e-02
-2.53578126e-01 3.06519389e-01 9.80270624e-01 -6.75513268e-01
4.13308620e-01 7.24088848e-01 9.40189719e-01 -2.02484369e-01
-1.53314304e+00 -8.20665658e-01 3.67529571e-01 6.40787303e-01
-1.39671850e+00 -4.12722409e-01 1.86062083e-01 -5.18158317e-01
1.35260725e+00 4.77058709e-01 9.16917443e-01 7.43720293e-01
3.38292360e-01 6.87420547e-01 1.35912108e+00 -2.01788709e-01
3.90120447e-02 -1.74956381e-01 2.50825137e-01 4.93893474e-01
4.67855126e-01 2.51350373e-01 -5.49504936e-01 -4.28455621e-01
3.94538492e-01 -1.20242059e-01 -5.03257871e-01 -3.64938557e-01
-1.05367255e+00 7.18931079e-01 8.60425830e-01 -4.85170007e-01
-9.60379183e-01 -5.26871607e-02 1.28122969e-02 2.88348854e-01
8.75289142e-01 6.65841520e-01 -1.14351439e+00 5.89723229e-01
-9.81140375e-01 3.85515451e-01 6.77181602e-01 4.71713334e-01
1.17563868e+00 1.60050660e-01 1.48085311e-01 8.49851131e-01
3.58721703e-01 1.15328336e+00 -1.38413031e-02 -1.17004979e+00
2.91168131e-02 1.78705677e-01 5.21460295e-01 -9.91845667e-01
-4.83588487e-01 -3.61031950e-01 -6.50429606e-01 2.22192109e-01
-1.31954700e-01 -4.03972924e-01 -9.87547517e-01 1.09274137e+00
1.60724804e-01 9.10811201e-02 2.33574077e-01 7.64846921e-01
8.39754939e-01 9.62171853e-01 1.30216897e-01 -1.40577510e-01
7.85494149e-01 -8.51037145e-01 -4.74577427e-01 -5.93139052e-01
2.69603223e-01 -4.39038843e-01 6.13276005e-01 5.96483588e-01
-6.81095421e-01 -7.31889457e-02 -1.16554201e+00 3.68586063e-01
-9.50956047e-01 1.40816435e-01 1.03296983e+00 4.45465863e-01
-1.02041066e+00 9.60239589e-01 -1.04384959e+00 -2.38600284e-01
3.90971720e-01 2.35690787e-01 -1.94395021e-01 6.75635263e-02
-1.16798234e+00 8.25751662e-01 6.58111691e-01 5.58397710e-01
-6.65327251e-01 -1.02267253e+00 -6.94705009e-01 -7.77469352e-02
3.14805090e-01 -2.72333771e-01 1.03985190e+00 -8.07453513e-01
-1.68106258e+00 7.10203946e-01 -2.01821223e-01 -4.03863877e-01
2.19686717e-01 -5.53437829e-01 -6.06453776e-01 6.88966289e-02
1.92166269e-01 8.73355627e-01 6.40556216e-01 -1.04627347e+00
-8.69891942e-01 -6.59677565e-01 -1.74848840e-01 4.25045431e-01
-4.87278283e-01 -4.28156763e-01 5.75833499e-01 -3.86548750e-02
6.67056680e-01 -1.04699492e+00 -4.42098558e-01 1.89884409e-01
-4.19524789e-01 2.34564453e-01 6.84535921e-01 -9.76800203e-01
4.78002638e-01 -2.05733442e+00 -1.61011785e-01 5.55942297e-01
-4.44002002e-02 1.96564198e-01 -3.54929745e-01 1.52570710e-01
-4.30558711e-01 -2.55774930e-02 -4.06377465e-01 8.80283773e-01
-1.07708007e-01 7.51968101e-02 -4.98726010e-01 5.00593126e-01
4.14320678e-01 6.76794231e-01 -9.34986770e-01 3.19055051e-01
2.28007451e-01 2.43811965e-01 -1.12270661e-01 -5.06767221e-02
-4.55592752e-01 -2.11323559e-01 -2.42128044e-01 8.20422828e-01
1.20371866e+00 -3.25075358e-01 3.29610556e-01 -4.01417941e-01
-5.51149428e-01 3.47001135e-01 -1.14387989e+00 1.06648004e+00
-7.15175271e-02 6.72344565e-01 5.56909323e-01 -9.95497048e-01
9.92116153e-01 3.26545574e-02 4.02727693e-01 -3.95169109e-01
-6.72297001e-01 3.84143859e-01 -2.50966668e-01 -4.73401695e-01
4.73837346e-01 3.57102722e-01 9.11227524e-01 2.40479916e-01
-2.52038866e-01 -4.23822165e-01 -1.54649451e-01 -2.55543917e-01
7.68981814e-01 6.20965481e-01 3.90492618e-01 -7.28158593e-01
2.29013816e-01 2.89118022e-01 1.74555331e-01 6.59577310e-01
-9.10541490e-02 1.52093291e-01 3.38601708e-01 -6.31723702e-01
-8.36092412e-01 -1.05262554e+00 -5.54552555e-01 1.47619295e+00
-1.32169768e-01 -1.33022919e-01 -7.32088163e-02 -2.53964573e-01
4.86638069e-01 5.54985702e-01 -6.84732497e-01 6.65299222e-03
-1.51766231e-02 -1.37972796e+00 2.25310504e-01 6.01554096e-01
7.04391360e-01 -9.06169116e-01 -3.92986625e-01 2.15506285e-01
-2.53327310e-01 -9.38361943e-01 6.03450894e-01 1.01530278e+00
-9.97866690e-01 -1.02093565e+00 -3.85967851e-01 -3.62604201e-01
1.24252383e-02 6.64017081e-01 1.10462391e+00 -5.03356218e-01
-4.39641625e-01 -1.33782983e-01 -2.93668836e-01 -8.99674833e-01
1.77423686e-01 4.66861963e-01 -3.07757229e-01 -2.57905036e-01
6.01830661e-01 -5.72871983e-01 -8.07174921e-01 4.72336709e-02
-6.15583658e-01 1.29460081e-01 5.83034396e-01 7.62061119e-01
7.72823274e-01 -6.82857707e-02 2.61766493e-01 -8.66447389e-01
1.48445219e-01 -7.36499131e-01 -8.16054761e-01 2.80390650e-01
-8.98533404e-01 -2.99714506e-01 1.91043079e-01 1.31321624e-01
-1.11259508e+00 3.45938772e-01 2.51377702e-01 3.25643510e-01
-3.23365211e-01 1.21955776e+00 1.53071553e-01 -2.12761208e-01
1.19429386e+00 1.77089944e-01 -3.41169126e-02 -1.58447176e-01
3.13476861e-01 8.68748248e-01 6.95426643e-01 -2.99827516e-01
7.25317299e-01 4.98669147e-01 8.72142538e-02 -1.34270346e+00
-1.20875776e+00 -3.64847779e-01 -6.28492713e-01 2.94559542e-03
3.36870402e-01 -1.46610320e+00 -3.90059263e-01 7.04573452e-01
-6.99792325e-01 -7.43821979e-01 6.57207072e-02 5.96989632e-01
-3.85208786e-01 -7.17635965e-03 -4.20973331e-01 -6.99327111e-01
-8.57818067e-01 -5.96466243e-01 9.30542231e-01 1.16565369e-01
5.89925982e-02 -4.38623309e-01 -2.06991181e-01 3.49235803e-01
6.42073750e-01 4.87933248e-01 7.85326540e-01 -8.41514170e-02
-5.16645849e-01 1.13702729e-01 -7.30737388e-01 3.49207193e-01
4.23930377e-01 4.33951467e-01 -1.41262591e+00 -2.97894657e-01
-4.00721371e-01 -5.99940598e-01 1.32306397e+00 1.12108183e+00
1.30855691e+00 -5.18151522e-02 -4.24023420e-01 1.14219701e+00
1.40077853e+00 -2.06852898e-01 4.99783576e-01 7.35710442e-01
3.32131773e-01 6.99204147e-01 5.82201898e-01 5.35102010e-01
2.01767772e-01 -1.80405140e-01 9.98926699e-01 -4.02438134e-01
3.60954314e-01 2.12244496e-01 2.02221572e-01 2.36244589e-01
-5.49933195e-01 9.43588465e-02 -1.12354696e+00 4.40261573e-01
-1.77095127e+00 -9.98335361e-01 -4.36971039e-01 1.88424778e+00
7.72469163e-01 -2.25818604e-01 -2.52132088e-01 3.05928588e-02
5.44003606e-01 7.20212281e-01 -1.00347233e+00 6.53847009e-02
-6.87745810e-01 5.47718465e-01 1.28760540e+00 4.33294028e-01
-1.54540777e+00 1.06465816e+00 8.07075119e+00 1.24795087e-01
-1.35988760e+00 -4.48636651e-01 5.23387969e-01 -1.48209482e-01
-2.66033351e-01 -1.25563145e-01 -5.43650270e-01 -1.59650594e-01
9.58911002e-01 1.29400760e-01 8.43929589e-01 9.11484897e-01
2.26245105e-01 -2.95528769e-01 -4.54798847e-01 3.99596214e-01
-3.00568789e-01 -1.34270620e+00 -3.31229270e-01 5.43734804e-03
8.90618742e-01 9.70757425e-01 2.22243398e-01 -2.73199845e-02
7.81678379e-01 -1.01160383e+00 3.50160152e-01 6.31460905e-01
9.15865541e-01 -6.18192554e-01 7.00834632e-01 -1.66186392e-01
-1.13087595e+00 -2.03882903e-01 -5.57585955e-01 -3.39150935e-01
-7.48587370e-01 7.25467086e-01 -7.46425986e-01 7.24317431e-01
1.26378548e+00 1.24835050e+00 -5.80992460e-01 7.99867392e-01
-3.32498699e-01 8.90360236e-01 -6.62525952e-01 1.32337257e-01
3.67153764e-01 -2.78856039e-01 4.19569053e-02 1.10471869e+00
2.38983661e-01 3.75652134e-01 1.11823425e-01 7.03540981e-01
3.59695762e-01 1.50494903e-01 -6.68035686e-01 -6.74112886e-02
5.22080004e-01 1.29563725e+00 -4.61085767e-01 -5.20047033e-04
-2.79330492e-01 6.21865511e-01 -4.81772535e-02 7.65877604e-01
-2.68024772e-01 -4.44124609e-01 1.01808715e+00 -1.06419630e-01
5.17482221e-01 -1.75558969e-01 -2.09118962e-01 -9.07850742e-01
-5.89292832e-02 -1.03410327e+00 7.27994815e-02 -1.25740504e+00
-1.42965472e+00 3.00234929e-02 -1.68037936e-01 -7.58378446e-01
-3.12122051e-02 -1.21156931e+00 -2.65059322e-01 1.25299382e+00
-2.22889161e+00 -1.08538508e+00 -8.23747098e-01 3.16837251e-01
-8.01781490e-02 -2.48016164e-01 1.33739758e+00 -2.30373263e-01
-4.66598064e-01 -1.37196764e-01 8.08039963e-01 -3.15739870e-01
6.50603116e-01 -1.29747367e+00 5.32465816e-01 4.21407998e-01
-5.26090026e-01 1.00870371e-01 4.87443984e-01 -6.59581184e-01
-1.36383784e+00 -1.36494112e+00 5.60894489e-01 1.68705270e-01
9.70753908e-01 3.72027606e-01 -7.14667439e-01 7.54739583e-01
1.22246742e-02 3.20249707e-01 7.32986212e-01 4.75536585e-01
-4.22991276e-01 -4.86288249e-01 -9.59865630e-01 3.35078090e-01
7.08471775e-01 -6.88902974e-01 2.35503122e-01 7.60632217e-01
3.80087435e-01 -3.76101524e-01 -9.92034435e-01 8.69246900e-01
9.83744323e-01 -7.86283910e-01 1.03602529e+00 -7.80625045e-01
6.21797502e-01 -2.71467775e-01 -6.92677855e-01 -1.40554583e+00
-7.59237826e-01 -2.07833588e-01 4.23266217e-02 3.41951907e-01
6.86460555e-01 -7.12540448e-01 1.04715919e+00 2.11075470e-01
6.41996935e-02 -4.31089193e-01 -2.85508931e-01 -6.11590743e-01
1.83096856e-01 -1.21861756e-01 7.84286261e-01 1.06554258e+00
-3.38928163e-01 3.59681621e-02 -2.47382417e-01 7.94182003e-01
5.99673986e-01 3.36375654e-01 5.25789917e-01 -1.76837158e+00
-8.34637228e-03 -3.26738685e-01 -7.13209808e-02 -5.73439717e-01
2.14959636e-01 -7.72196293e-01 6.20605722e-02 -1.37476039e+00
9.34485495e-02 -2.25154519e-01 -5.44726849e-01 1.00770116e+00
-4.76532519e-01 6.81415796e-02 -2.95228988e-01 6.66833296e-02
1.43689707e-01 3.82692963e-01 8.06040168e-01 -7.39390194e-01
-3.07820737e-01 5.02335466e-03 -5.58736503e-01 6.74888670e-01
1.26775944e+00 -1.76346496e-01 -1.75143600e-01 -6.55392885e-01
4.67537105e-01 -9.10747424e-02 5.24161100e-01 -8.25523376e-01
-2.31348723e-01 -8.51049840e-01 9.30407107e-01 -8.13911378e-01
2.22105354e-01 -5.85640669e-01 1.41461566e-01 5.88531613e-01
-3.56361419e-01 -2.92322427e-01 4.42397654e-01 3.09075236e-01
1.79056495e-01 -1.41685680e-01 4.18516278e-01 -1.26670077e-01
-7.96755731e-01 2.11867884e-01 -6.16730928e-01 -7.51777828e-01
5.26243985e-01 -1.01685055e-01 -7.70655632e-01 -1.19159251e-01
-6.82321966e-01 4.06392664e-01 7.67828748e-02 5.17583340e-02
2.49061778e-01 -8.51647913e-01 -1.16702807e+00 3.33055258e-01
2.21775472e-01 -1.14392892e-01 -2.17590407e-02 3.60884666e-01
-8.87504399e-01 4.32576537e-01 -6.53226376e-01 -6.35633290e-01
-9.92352128e-01 -2.19604298e-01 7.48773754e-01 2.10118853e-02
-1.04939200e-01 8.20136964e-01 -3.72719169e-01 -8.64864945e-01
-1.25803232e-01 -1.60712972e-01 -1.68146804e-01 9.79712009e-02
3.96693170e-01 6.33282840e-01 4.00915951e-01 -6.64256222e-04
-3.58511716e-01 9.10650268e-02 6.00876808e-02 1.41208738e-01
2.05771303e+00 1.01922475e-01 -5.17622054e-01 4.62580115e-01
7.26254225e-01 -6.23291135e-01 -1.33438301e+00 -5.11159003e-01
5.79655841e-02 -2.95582861e-01 7.64316499e-01 -1.21809101e+00
-1.05593097e+00 5.04135191e-01 1.10386443e+00 2.14894116e-01
1.18784058e+00 -6.96827710e-01 2.80243248e-01 1.36896729e+00
-1.45265600e-02 -1.27772486e+00 -4.24376637e-01 7.31321812e-01
8.81120443e-01 -1.56852531e+00 7.74671972e-01 -5.18035412e-01
-8.20389241e-02 1.38103580e+00 8.27548325e-01 1.27843171e-01
1.04797006e+00 3.35881144e-01 3.62233192e-01 -1.64852083e-01
-6.98336959e-01 -2.40726203e-01 3.18131633e-02 9.51043665e-01
7.69342124e-01 7.34392464e-01 2.55114079e-01 3.48950587e-02
-3.37365955e-01 1.66146845e-01 3.03108931e-01 1.01377189e+00
-8.32119763e-01 -4.98420149e-01 -3.95273507e-01 1.12458515e+00
7.00826719e-02 -5.24376988e-01 -3.27835321e-01 2.54482359e-01
-1.40422583e-01 8.57672513e-01 -5.71702458e-02 -2.00903922e-01
2.47455373e-01 5.76474555e-02 1.21492818e-01 -4.57948178e-01
-2.84373581e-01 3.83485816e-02 1.60941735e-01 -5.80129325e-01
-7.12288082e-01 -7.81181395e-01 -5.14031410e-01 -4.89145309e-01
-5.79397023e-01 -2.00182945e-01 9.57116365e-01 6.32493079e-01
2.82688051e-01 2.15812743e-01 7.72052526e-01 -1.17327142e+00
-6.72723472e-01 -1.16351295e+00 -1.10058832e+00 -3.80121440e-01
6.32615089e-01 -3.50744307e-01 -1.72990277e-01 1.17969103e-01] | [9.465338706970215, -1.545013666152954] |
2b810d7e-c8ae-49f7-8c99-491dd2a81e13 | cqare-contrastive-question-answering-for-few | null | null | https://openreview.net/forum?id=FEg_0BrW4Ks | https://openreview.net/pdf?id=FEg_0BrW4Ks | CQARE: Contrastive Question-Answering for Few-shot Relation Extraction with Prompt Tuning | Prompt tuning with pre-trained language models (PLM) has exhibited outstanding performance by closing the gap between pre-training tasks and various downstream applications, without the need for uninitialized parameters to be introduced. However, prompt tuning requires vast amounts of prompt engineering and predefined label word mapping, which obstructs its implements in practice. Besides, the ample label space makes prompt tuning more arduous and challenging when it comes to relation extraction (RE). To tackle these issues, we propose a Contrastive Question-Answering method with prompt tuning for few-shot RE (CQARE). CQARE carries out a RE task-specific pre-training with four entity-relation-aware pre-training objects, including a prompt pre-training to automatically generate continuous prompts. The proposed pre-training can provide more robust initialization with prompt tuning while maintaining semantic consistency with the proposed PLM. Furthermore, CQARE can effectively avoid label words mapping by reformulating RE as contrastive question answering. The results indicate CQARE raising averaged accuracy of 5.11\% on a cross-domain few-shot dataset, demonstrating that robust initialization is crucial for prompt tuning and effective contrastive question answering. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['cross-domain-few-shot'] | ['computer-vision'] | [ 2.84525692e-01 1.98480785e-01 -6.81572454e-03 -4.71560836e-01
-9.75563586e-01 -4.40273046e-01 6.95895493e-01 3.32773119e-01
-7.91415691e-01 4.93776053e-01 1.40471414e-01 -2.67074168e-01
-3.60190928e-01 -8.54794323e-01 -3.56337011e-01 -3.49417478e-01
4.52212602e-01 6.56592906e-01 5.08639574e-01 -7.18497336e-01
1.04875021e-01 -2.75304262e-03 -1.58938277e+00 1.37911767e-01
1.24840975e+00 9.89076674e-01 6.31066799e-01 6.31986260e-01
-6.41044915e-01 8.57974112e-01 -8.99383783e-01 -7.16263473e-01
-2.19109595e-01 -3.48983198e-01 -1.14004612e+00 -1.54382691e-01
2.65059650e-01 -9.97684374e-02 -5.70421070e-02 7.39543259e-01
7.11534083e-01 7.40130782e-01 6.00971282e-01 -1.15423119e+00
-9.23613667e-01 6.25581145e-01 -1.58039391e-01 3.82395327e-01
2.25648105e-01 2.03606278e-01 1.27837241e+00 -1.01846993e+00
4.41002727e-01 1.05781269e+00 4.38314736e-01 7.86738515e-01
-9.98739004e-01 -6.40862405e-01 1.58953369e-02 5.32404244e-01
-1.44208980e+00 -5.24913490e-01 5.93668342e-01 -2.28827909e-01
1.33747101e+00 4.39579278e-01 7.73883983e-02 9.54150081e-01
-4.44258571e-01 4.99717742e-01 7.23858476e-01 -6.67871356e-01
2.68871337e-01 2.03740969e-01 4.86703545e-01 4.64181244e-01
-1.36759773e-01 -2.79520184e-01 -3.88395220e-01 -2.65358184e-02
3.93753737e-01 -1.50809765e-01 -2.12449223e-01 1.96195528e-01
-1.07272553e+00 7.68589735e-01 3.23116899e-01 4.22702581e-01
-2.76915997e-01 -1.43468067e-01 5.37397206e-01 2.39724323e-01
1.74216673e-01 8.19633901e-01 -6.19693756e-01 -9.46831927e-02
-7.47692168e-01 1.07357666e-01 6.00469947e-01 1.19140041e+00
7.39718378e-01 -2.31641725e-01 -9.43843961e-01 1.31292021e+00
-3.93988937e-02 3.26562524e-01 5.38687289e-01 -9.07339871e-01
7.16886938e-01 7.35427916e-01 1.21986516e-01 -8.23590457e-01
-3.95382911e-01 -4.56886232e-01 -7.56933510e-01 -6.07502103e-01
2.42631778e-01 -3.17816325e-02 -7.47414768e-01 1.94030976e+00
5.06590068e-01 2.62103170e-01 2.84282893e-01 6.95946157e-01
1.31489611e+00 8.45665276e-01 6.90203667e-01 -3.30545127e-01
1.88611090e+00 -1.22641027e+00 -1.03976429e+00 -2.78110892e-01
8.49873424e-01 -7.33230770e-01 1.94635534e+00 -3.93722951e-02
-6.87840462e-01 -7.78408706e-01 -8.45235169e-01 -4.12539989e-01
-4.27050054e-01 3.59916836e-02 5.19702613e-01 4.20705169e-01
-5.34855425e-01 1.16673760e-01 -2.99448133e-01 -6.20471060e-01
1.13141090e-01 1.33970499e-01 -6.50318265e-02 -1.25796616e-01
-1.68652141e+00 9.35965419e-01 5.87180018e-01 -2.23178118e-02
-5.71145475e-01 -1.01258552e+00 -7.94869840e-01 2.49061823e-01
9.16560233e-01 -8.08535874e-01 1.54231727e+00 -1.53554216e-01
-1.54321194e+00 8.58676851e-01 -2.27852285e-01 -1.33825108e-01
2.11561933e-01 -3.74384522e-01 -7.06922472e-01 1.53480887e-01
2.31544748e-01 6.52148366e-01 7.44943857e-01 -9.05353844e-01
-5.48755705e-01 1.28384739e-01 3.97606403e-01 3.58010262e-01
-6.52356923e-01 2.59980857e-01 -6.00496233e-01 -6.01676106e-01
-3.32877457e-01 -3.68827760e-01 -1.42117843e-01 -5.25537372e-01
-2.15575665e-01 -8.12124848e-01 6.11845851e-01 -5.10808587e-01
1.78325379e+00 -1.97375703e+00 -4.13440824e-01 -3.40345263e-01
1.17995232e-01 6.94652140e-01 -5.27917743e-01 4.71613467e-01
-9.61007625e-02 3.46168801e-02 -2.93430001e-01 -1.47231370e-01
1.72475621e-01 3.67150187e-01 -3.87752175e-01 -2.75753975e-01
5.08929431e-01 1.33506262e+00 -1.16890848e+00 -7.78378069e-01
1.38741255e-01 1.04309425e-01 -4.95907068e-01 8.03725123e-01
-5.82744300e-01 2.81495869e-01 -3.34906816e-01 4.39401537e-01
3.02837044e-01 -5.64576864e-01 -2.24674135e-01 -3.17250043e-01
2.66175479e-01 4.08428639e-01 -1.09938300e+00 1.73119903e+00
-7.89605975e-01 1.56322271e-01 -2.82188952e-01 -1.03061914e+00
1.16795278e+00 4.20226306e-01 1.93281874e-01 -1.13200057e+00
1.52072713e-01 8.53160471e-02 -1.46257877e-01 -1.03456414e+00
7.44569361e-01 -2.44080126e-01 -2.11017966e-01 4.20352548e-01
2.99514025e-01 -2.20042437e-01 3.48337322e-01 3.92161518e-01
1.21074474e+00 -1.77538004e-02 2.92688966e-01 -1.62401214e-01
7.77246118e-01 9.35953408e-02 5.26457131e-01 6.21197343e-01
-2.07859889e-01 5.40404797e-01 6.44555455e-03 5.26752360e-02
-6.87979162e-01 -8.39981019e-01 9.44261812e-03 1.49983513e+00
2.87526876e-01 -6.66780233e-01 -8.72099340e-01 -8.60569477e-01
-4.37002391e-01 1.12499213e+00 -2.98077315e-01 -3.76432568e-01
-6.15375459e-01 -7.26840258e-01 5.33037663e-01 3.96966010e-01
7.15460956e-01 -1.29259217e+00 -4.73552793e-01 5.78255057e-01
-7.07275510e-01 -1.32861507e+00 -6.57685459e-01 3.43877286e-01
-5.78460693e-01 -1.03804386e+00 -6.54694855e-01 -9.10751402e-01
5.80526054e-01 3.61984193e-01 1.47289312e+00 7.14351311e-02
-2.93296576e-01 2.98506945e-01 -6.23880208e-01 -1.35427773e-01
-2.10511357e-01 3.13214809e-01 -1.52199104e-01 -2.34302521e-01
7.92347670e-01 -5.08762836e-01 -6.77732885e-01 4.95045513e-01
-1.00171864e+00 1.04442134e-01 3.29140931e-01 9.66085732e-01
4.60359633e-01 -1.31697193e-01 1.13351965e+00 -1.08964980e+00
9.24182415e-01 -4.31120187e-01 -2.91872382e-01 8.81630540e-01
-6.72885358e-01 9.09273624e-02 7.78473556e-01 -5.69408715e-01
-1.45808649e+00 -4.52717900e-01 -4.18821901e-01 1.39677208e-02
-1.77006021e-01 2.37292439e-01 -3.65196228e-01 5.14448583e-01
1.14053380e+00 -8.52807239e-02 -4.59166795e-01 -5.20520926e-01
8.22410226e-01 7.87869871e-01 7.62671530e-01 -9.48188722e-01
7.88532555e-01 -4.91221398e-02 -5.87865055e-01 -7.29654312e-01
-1.41813266e+00 -6.98136568e-01 -4.54167157e-01 2.04281099e-02
1.06632864e+00 -6.45846546e-01 -6.32525563e-01 1.94102332e-01
-1.21330488e+00 -4.16954964e-01 -6.25279486e-01 1.94614872e-01
-4.02711630e-01 2.64364362e-01 -5.58073461e-01 -6.19458437e-01
-7.31019020e-01 -6.44704759e-01 8.45885277e-01 3.97282511e-01
-5.48259437e-01 -9.85055268e-01 2.50594374e-02 7.64192402e-01
5.60738027e-01 -2.99624950e-01 1.17639315e+00 -9.40982521e-01
-2.86887199e-01 1.32104278e-01 -4.06293452e-01 2.77091563e-01
2.50767529e-01 -3.20362896e-01 -1.13219404e+00 1.65407717e-01
6.56078905e-02 -6.12040043e-01 4.30742204e-01 -1.28424108e-01
1.06459367e+00 -2.78890997e-01 -6.60217032e-02 4.56937581e-01
1.15800846e+00 1.42347157e-01 3.94661486e-01 4.04283702e-01
5.25751591e-01 7.81102598e-01 1.27085245e+00 2.41617411e-01
7.61566162e-01 6.31540895e-01 6.06664680e-02 -2.18677055e-02
-3.14607114e-01 -4.37994063e-01 -8.92715901e-02 1.16773474e+00
2.50634849e-01 -2.38745227e-01 -8.50474656e-01 6.06943369e-01
-1.87558937e+00 -7.04837084e-01 -1.06029563e-01 1.88288617e+00
1.38431728e+00 -5.10560535e-02 -2.12906301e-01 4.61635739e-02
7.76543617e-01 8.12211037e-02 -5.56226492e-01 -2.38972068e-01
9.20724049e-02 5.78615427e-01 -1.01613775e-01 4.95462388e-01
-8.64580572e-01 1.38559020e+00 4.88238573e+00 1.24957764e+00
-7.36735106e-01 4.45145905e-01 4.13247377e-01 2.91230142e-01
-4.87218052e-01 9.56782326e-03 -1.13733137e+00 2.32567713e-01
1.04821742e+00 -3.55505645e-01 1.80575743e-01 6.89871848e-01
1.10398397e-01 -8.46752077e-02 -1.01013947e+00 1.15086246e+00
-3.44507396e-02 -1.15800297e+00 -6.65828809e-02 -5.98829567e-01
5.07015467e-01 -3.81953567e-01 -2.68713534e-01 1.11683619e+00
3.16558123e-01 -8.65165472e-01 3.41957211e-01 3.93165261e-01
9.46993589e-01 -5.88393569e-01 6.35426700e-01 6.54009819e-01
-1.18301630e+00 -8.38966854e-03 -4.68104094e-01 -6.64959773e-02
3.41312498e-01 5.51019371e-01 -1.06069303e+00 6.54540420e-01
5.44138193e-01 1.06240854e-01 -6.82751358e-01 7.26453960e-01
-5.61639547e-01 6.62593961e-01 -1.56887218e-01 -1.25333637e-01
4.89327125e-02 -3.84714045e-02 1.40367314e-01 1.33890414e+00
2.04220548e-01 6.23073101e-01 7.81384930e-02 6.58873796e-01
-1.48215532e-01 4.39459026e-01 1.01538017e-01 -1.29420832e-01
8.44945610e-01 1.50766754e+00 -6.32266164e-01 -6.34522259e-01
-4.04684603e-01 9.62249994e-01 5.63399851e-01 4.70959544e-01
-9.90256786e-01 -7.33065009e-01 4.09301877e-01 1.19574703e-02
4.34528250e-04 -6.02556691e-02 -1.11472428e-01 -1.22783732e+00
3.95003818e-02 -7.31829107e-01 6.28833413e-01 -6.18975222e-01
-1.48754382e+00 6.96583688e-01 -4.98725362e-02 -8.80982339e-01
-2.76698172e-01 -1.98960021e-01 -5.93797565e-01 8.68156612e-01
-1.76530898e+00 -1.04595220e+00 -5.70336938e-01 6.02197528e-01
7.92965829e-01 6.74854666e-02 1.01532185e+00 6.57505274e-01
-7.95658529e-01 8.87156010e-01 -5.91480911e-01 -3.65627408e-02
1.08835161e+00 -1.25158787e+00 4.16552395e-01 7.98999131e-01
1.53746247e-01 7.19394147e-01 6.02116346e-01 -4.04199749e-01
-9.97827709e-01 -1.34354913e+00 1.30386901e+00 -5.16872466e-01
6.94092095e-01 -3.44755888e-01 -1.29170680e+00 2.53695399e-01
1.40973374e-01 4.85153459e-02 7.84923673e-01 2.11420447e-01
-3.69735390e-01 -2.09110692e-01 -9.28244770e-01 5.88499308e-01
1.25157964e+00 -7.40541577e-01 -1.23193526e+00 4.29459184e-01
1.27028751e+00 -1.78963661e-01 -8.65227282e-01 6.08025134e-01
-1.48696810e-01 -4.57242131e-01 9.42803741e-01 -5.52536428e-01
2.21489400e-01 -2.73057669e-01 -1.17949493e-01 -9.81361151e-01
-3.19239885e-01 -6.81473732e-01 -4.11067039e-01 1.98874247e+00
3.83483559e-01 -3.86142582e-01 4.36666518e-01 8.35124791e-01
-1.82912260e-01 -7.63704181e-01 -5.33529937e-01 -1.02242887e+00
-1.92339823e-01 -6.23552501e-01 6.59047008e-01 1.13116074e+00
-5.50618246e-02 1.12433898e+00 -7.25298077e-02 3.49318311e-02
2.53935099e-01 9.19027347e-03 7.80389845e-01 -1.07774365e+00
-3.56856912e-01 -2.09116414e-01 3.80534887e-01 -1.33800185e+00
1.48824513e-01 -1.05946124e+00 2.68329650e-01 -1.65118420e+00
1.70406133e-01 -9.20722246e-01 -2.36462638e-01 5.97756386e-01
-8.59254599e-01 3.13705988e-02 -8.99349451e-02 9.54691768e-02
-8.99769664e-01 9.43042994e-01 1.29266965e+00 3.52610312e-02
-2.48017073e-01 -1.52379841e-01 -8.11728716e-01 4.38243240e-01
7.07026362e-01 -4.79876310e-01 -8.85065734e-01 -3.50104302e-01
2.40618542e-01 5.43146394e-02 5.26111089e-02 -6.72020733e-01
4.04006332e-01 -2.52734154e-01 -2.36135840e-01 -3.94384027e-01
2.76803672e-01 -6.14689767e-01 -2.87040502e-01 5.18898070e-02
-4.39932168e-01 -1.02112509e-01 2.12507565e-02 2.80321598e-01
-3.39741915e-01 -6.25240505e-01 7.62989640e-01 1.83466950e-03
-1.20413113e+00 3.36273432e-01 1.77898481e-01 8.51754069e-01
9.12327290e-01 6.44354802e-03 -4.67268825e-01 -1.78637952e-01
-6.42814100e-01 5.69260418e-01 1.09051652e-02 6.26917958e-01
3.63022536e-01 -1.23574567e+00 -5.32424629e-01 -9.20866579e-02
6.24148726e-01 2.99505264e-01 5.72183669e-01 5.50609946e-01
-3.62468436e-02 3.33395243e-01 3.65410566e-01 -4.50387508e-01
-1.13200665e+00 7.49296367e-01 2.25360729e-02 -5.17548859e-01
-6.91611409e-01 1.20965135e+00 1.38026416e-01 -7.84811080e-01
3.79064113e-01 -1.96328476e-01 -4.63074923e-01 3.71591300e-01
6.67791069e-01 2.44901136e-01 2.93099195e-01 -1.24507509e-01
-1.88657165e-01 4.21656370e-01 -1.39471352e-01 -1.11802034e-01
1.13756967e+00 -2.99769312e-01 3.09851840e-02 3.24926466e-01
1.12070882e+00 -1.13369428e-01 -1.05134165e+00 -6.91827357e-01
5.37533581e-01 -1.83508217e-01 -2.35662028e-01 -9.91549730e-01
-5.32152474e-01 9.44284976e-01 2.29502201e-01 7.57483616e-02
1.14652967e+00 1.30629361e-01 1.36110520e+00 6.91329896e-01
2.62856662e-01 -1.29954481e+00 4.29252446e-01 7.87441015e-01
7.21981466e-01 -1.34260249e+00 -4.05597508e-01 -5.86595595e-01
-6.56101108e-01 7.67428875e-01 9.94812131e-01 2.94844002e-01
3.94961983e-01 4.14005779e-02 2.84851283e-01 -2.75054246e-01
-8.82905006e-01 -5.62017679e-01 4.30714965e-01 6.02540970e-01
5.40274799e-01 -1.79605335e-01 -5.16566873e-01 1.07144034e+00
-3.65186006e-01 -1.30307987e-01 1.12569937e-02 7.30649650e-01
-5.26697040e-01 -1.14958930e+00 -4.46537361e-02 2.41119727e-01
-1.97337270e-01 -4.13419932e-01 9.58038419e-02 5.66491842e-01
9.88679603e-02 1.26537812e+00 -5.69571927e-02 -2.11375117e-01
6.31421983e-01 4.00105923e-01 2.94185519e-01 -1.09954047e+00
-7.27378905e-01 -3.15817982e-01 3.00816774e-01 -3.23112428e-01
-1.75993428e-01 -1.16421200e-01 -1.41855407e+00 8.20930451e-02
-6.74019098e-01 6.11576617e-01 2.74307132e-01 1.16371024e+00
4.57028925e-01 8.30969334e-01 4.56053764e-01 -3.41415778e-02
-6.64855719e-01 -1.21084678e+00 -3.46783064e-02 5.70756197e-01
-9.12031382e-02 -7.15873837e-01 -2.71821141e-01 1.99155882e-02] | [10.580045700073242, 8.128897666931152] |
1149e2c0-40ff-49a7-8ea3-4f6ce457fc2d | efficient-model-monitoring-for-quality | 2104.05533 | null | https://arxiv.org/abs/2104.05533v1 | https://arxiv.org/pdf/2104.05533v1.pdf | Efficient Model Monitoring for Quality Control in Cardiac Image Segmentation | Deep learning methods have reached state-of-the-art performance in cardiac image segmentation. Currently, the main bottleneck towards their effective translation into clinics requires assuring continuous high model performance and segmentation results. In this work, we present a novel learning framework to monitor the performance of heart segmentation models in the absence of ground truth. Formulated as an anomaly detection problem, the monitoring framework allows deriving surrogate quality measures for a segmentation and allows flagging suspicious results. We propose two different types of quality measures, a global score and a pixel-wise map. We demonstrate their use by reproducing the final rankings of a cardiac segmentation challenge in the absence of ground truth. Results show that our framework is accurate, fast, and scalable, confirming it is a viable option for quality control monitoring in clinical practice and large population studies. | ['Maria A. Zuluaga', 'Francesco Galati'] | 2021-04-12 | null | null | null | null | ['cardiac-segmentation'] | ['medical'] | [ 2.77038634e-01 9.28631797e-02 -3.42809260e-01 -4.60557610e-01
-1.27228510e+00 -5.39335608e-01 2.54609138e-01 8.01442802e-01
-6.24992549e-01 6.52188897e-01 -7.19351545e-02 -3.11996400e-01
-1.96385950e-01 -6.44529104e-01 -4.29289937e-01 -6.05143428e-01
-9.95755121e-02 8.28178346e-01 2.87799239e-01 3.13784778e-01
1.14286244e-01 6.34644389e-01 -9.26111341e-01 3.38385403e-01
8.94347966e-01 1.00923920e+00 -1.76748097e-01 8.17203283e-01
1.05259597e-01 5.94235599e-01 -5.54292738e-01 -3.77862781e-01
3.63962948e-01 -4.76798594e-01 -8.61230016e-01 1.42200768e-01
6.30798876e-01 -2.65331298e-01 2.04196557e-01 7.52449155e-01
8.19585681e-01 -4.50488120e-01 4.41016734e-01 -9.76291656e-01
1.00472584e-01 3.57074559e-01 -2.13314414e-01 4.66915667e-01
-1.80324525e-01 5.65110564e-01 8.54484856e-01 -2.94364363e-01
6.18966579e-01 5.86093605e-01 9.33133721e-01 3.51694494e-01
-1.60585856e+00 -2.18356788e-01 -2.69829392e-01 -1.52358606e-01
-1.08162773e+00 -2.31162235e-01 3.23812842e-01 -8.11116993e-01
4.74647284e-01 3.81191641e-01 9.60242689e-01 6.97474301e-01
4.18523312e-01 5.54863453e-01 1.34862590e+00 -1.91180617e-01
2.87599891e-01 -2.69794296e-02 1.73491031e-01 8.40910017e-01
3.41099709e-01 -1.83181595e-02 -2.18615577e-01 -2.00604185e-01
9.26727653e-01 -1.59870163e-01 -8.91722739e-02 -5.55401266e-01
-1.39877415e+00 7.89415538e-01 4.57529217e-01 5.32024086e-01
-6.30157769e-01 1.75373778e-01 5.26625574e-01 -1.39583386e-02
4.73001033e-01 6.35315835e-01 -4.40158457e-01 -2.95405626e-01
-1.68363011e+00 3.10403287e-01 5.18353701e-01 1.43211439e-01
1.69414401e-01 -2.46137843e-01 -7.14440107e-01 5.95405519e-01
2.24050283e-01 3.97171110e-01 4.57234293e-01 -1.26986921e+00
-2.02074815e-02 7.81390071e-01 2.35922411e-02 -9.12136614e-01
-8.15501273e-01 -7.96545088e-01 -7.21619725e-01 3.45453590e-01
7.98868835e-01 -1.71233594e-01 -1.00651467e+00 1.54667437e+00
3.15180153e-01 1.98388219e-01 -4.93939072e-01 8.84449959e-01
6.52242720e-01 2.35129874e-02 2.60573536e-01 -3.16331595e-01
1.18839180e+00 -6.81694627e-01 -4.19882268e-01 8.06229413e-02
8.30128849e-01 -5.23030221e-01 9.50088680e-01 5.13857067e-01
-1.28575587e+00 -5.20647764e-01 -8.57697725e-01 2.81913519e-01
-1.71380430e-01 2.18356833e-01 3.79183114e-01 8.46504450e-01
-1.30945933e+00 1.03737879e+00 -1.20077682e+00 -1.58103526e-01
8.84081900e-01 4.61405277e-01 -2.02667475e-01 1.50597349e-01
-7.70792425e-01 8.70179892e-01 1.73674598e-01 1.38644233e-01
-8.75713706e-01 -9.66382384e-01 -5.17355442e-01 -4.80017178e-02
5.63801080e-02 -7.54657269e-01 1.24734855e+00 -6.44795001e-01
-1.21558928e+00 1.48908126e+00 1.60297304e-01 -6.98128998e-01
1.23220193e+00 -7.55894631e-02 -9.71888471e-03 3.27319950e-01
2.50569969e-01 6.91800654e-01 6.01891935e-01 -9.75813627e-01
-5.80198586e-01 -4.65124577e-01 -4.37126219e-01 -2.47560441e-01
-6.23841435e-02 -5.26599772e-02 -3.93591106e-01 -3.52791727e-01
1.69859067e-01 -8.62961173e-01 -6.44056976e-01 1.68298095e-01
-4.61130291e-01 1.76037282e-01 1.93965405e-01 -8.13742220e-01
1.35305607e+00 -1.76666713e+00 -1.69732451e-01 4.87097979e-01
6.65396869e-01 6.23566806e-01 1.46235690e-01 -1.75890148e-01
9.54199210e-02 4.92706358e-01 -5.94281197e-01 -1.70907110e-01
-3.21927547e-01 6.43666908e-02 3.23361307e-01 5.38967192e-01
4.47051018e-01 1.18015122e+00 -1.05039692e+00 -8.50684226e-01
4.50014353e-01 5.20615816e-01 -5.12296617e-01 8.83551985e-02
1.17268162e-02 1.00755239e+00 -4.18593049e-01 4.71850663e-01
3.51008892e-01 -5.85264504e-01 3.40483457e-01 -6.72018602e-02
-1.50095522e-02 7.54256025e-02 -1.08285296e+00 1.77722013e+00
-1.52966559e-01 3.73499036e-01 3.62508669e-02 -9.52870965e-01
7.84674227e-01 3.29461277e-01 1.07936132e+00 -7.01595366e-01
2.99266696e-01 4.35932219e-01 1.97286218e-01 -3.47735822e-01
-1.44324645e-01 -3.88535053e-01 1.26365468e-01 2.73991406e-01
-5.87778725e-02 -2.18152195e-01 3.00105810e-01 2.75634807e-02
1.24382460e+00 -1.54472245e-02 1.65583462e-01 -5.38355708e-01
5.37209511e-01 2.06309378e-01 5.71628869e-01 1.01443326e+00
-7.04450905e-01 1.05068278e+00 8.48814070e-01 -7.38286376e-01
-9.35176253e-01 -1.17126179e+00 -3.74042958e-01 4.21753019e-01
-3.32362473e-01 -2.89682746e-01 -1.01529884e+00 -9.43088472e-01
3.03100348e-02 1.57876834e-01 -8.34849536e-01 -1.81346852e-02
-6.82457685e-01 -1.09814656e+00 6.58037543e-01 5.28803587e-01
3.01081032e-01 -1.01369774e+00 -1.16257834e+00 5.38137913e-01
-2.12336257e-01 -1.03256345e+00 -1.28036574e-01 -7.41700977e-02
-1.31170678e+00 -1.33579385e+00 -1.04271412e+00 -4.92424101e-01
4.38397259e-01 -4.82692033e-01 1.62171984e+00 3.13168734e-01
-7.57646918e-01 2.05203846e-01 5.46001233e-02 -4.06019986e-01
-6.87003672e-01 1.30629972e-01 -2.45605379e-01 -1.11243486e-01
-6.63247555e-02 -2.57114559e-01 -1.11132932e+00 2.07714483e-01
-7.42262483e-01 -1.55523777e-01 5.46824872e-01 6.84554815e-01
9.92121756e-01 -4.09837872e-01 4.63797927e-01 -1.12573111e+00
5.67541420e-01 -8.78106952e-02 -6.73126578e-01 2.89427668e-01
-1.10793340e+00 -1.87039152e-01 1.05859280e-01 1.89985335e-01
-5.52637935e-01 1.24651484e-01 -2.79821932e-01 -2.84233034e-01
-3.56904149e-01 5.20371914e-01 4.81277555e-01 5.22837155e-02
8.94651473e-01 -1.84902593e-01 3.88262570e-01 -4.46110576e-01
1.76674128e-02 1.60276070e-01 5.32220364e-01 -4.86444205e-01
4.04216737e-01 5.77348530e-01 4.55619663e-01 -4.31916684e-01
-8.81213427e-01 -6.52006209e-01 -9.87034738e-01 -4.09744263e-01
1.07283723e+00 -6.01722002e-01 -4.16377425e-01 4.28995788e-01
-7.99081624e-01 -5.78445137e-01 -5.27455330e-01 2.22844079e-01
-6.06799304e-01 3.74420166e-01 -6.20085478e-01 -4.79783148e-01
-5.94593704e-01 -1.48930335e+00 1.19165409e+00 8.59631002e-02
-2.56069720e-01 -1.25942016e+00 3.65945965e-01 4.53007609e-01
6.13235652e-01 9.56822574e-01 7.41043031e-01 -7.02695310e-01
-5.20035744e-01 -4.21967119e-01 -2.12602019e-01 4.96031791e-01
1.27192259e-01 1.59030259e-01 -9.40865934e-01 -2.26843074e-01
-1.35046691e-01 -2.31742710e-02 7.71664858e-01 1.03056872e+00
1.17975271e+00 1.53041467e-01 -2.27941364e-01 5.86489975e-01
1.36311615e+00 -1.25335798e-01 4.62059587e-01 4.18812662e-01
5.82878709e-01 3.87486756e-01 3.42828482e-01 2.35273227e-01
1.76876783e-01 5.68057358e-01 3.55206132e-01 -6.54071152e-01
-1.50328740e-01 6.80875704e-02 -3.53367895e-01 4.92474228e-01
-1.28596678e-01 1.98935941e-01 -1.49361384e+00 4.95787740e-01
-1.90506399e+00 -5.92646241e-01 -4.50758487e-01 2.48046803e+00
7.06082344e-01 5.69921494e-01 4.74759609e-01 9.59997550e-02
5.38142562e-01 -2.75443941e-01 -4.27427292e-01 -3.49748611e-01
1.19938456e-01 4.57517147e-01 5.31432807e-01 2.37500340e-01
-1.26130819e+00 6.27319276e-01 7.19533491e+00 4.14007276e-01
-1.21794677e+00 2.40966052e-01 1.36401117e+00 5.04088439e-02
1.19144730e-01 -2.86210328e-01 -2.93983072e-01 4.71814871e-01
1.04160047e+00 2.60386050e-01 -1.69881299e-01 5.30203700e-01
6.03702486e-01 -3.67804579e-02 -9.44526792e-01 6.98814273e-01
-2.85232455e-01 -1.71004879e+00 -2.26931497e-01 1.40040234e-01
6.45329893e-01 2.66591460e-01 2.27406491e-02 8.58881325e-03
2.01072022e-02 -1.13726032e+00 3.17475021e-01 5.26246250e-01
9.74181354e-01 -3.24729204e-01 8.81217837e-01 4.91982028e-02
-6.90910459e-01 2.03349993e-01 2.14066729e-01 3.45153093e-01
2.84429401e-01 8.57575595e-01 -1.03921092e+00 4.17431802e-01
5.95101178e-01 5.92905521e-01 -7.69424975e-01 1.45823824e+00
-7.13488506e-03 8.04390669e-01 -1.71006426e-01 4.59638804e-01
3.93229663e-01 -1.59253925e-01 5.19687116e-01 1.43050396e+00
-1.82080630e-03 -3.09714228e-01 3.71334702e-01 9.20219541e-01
-2.11546794e-02 3.62834483e-01 -1.86226830e-01 7.02088550e-02
-1.31315425e-01 1.45166636e+00 -1.36519480e+00 -4.73974973e-01
-1.33922890e-01 5.74174523e-01 1.24194343e-02 -8.15701261e-02
-8.38360071e-01 1.83679029e-01 2.78485596e-01 5.17896473e-01
-9.57331061e-02 8.86178166e-02 -8.75628889e-01 -8.22129309e-01
-4.39363681e-02 -8.64467978e-01 6.92335427e-01 -2.62862206e-01
-1.10908496e+00 4.45038915e-01 -2.17061281e-01 -1.14574504e+00
-1.08814664e-01 -7.00478137e-01 -5.70488274e-01 8.53124976e-01
-1.72102118e+00 -9.15537238e-01 -4.33786303e-01 2.06413925e-01
2.18485877e-01 1.17202654e-01 8.38446319e-01 3.74990672e-01
-6.35263562e-01 5.07449031e-01 -6.63627908e-02 2.87587345e-01
6.36499524e-01 -1.57185602e+00 2.10846841e-01 7.31435120e-01
9.72087085e-02 3.69029939e-01 5.46666086e-01 -5.54054379e-01
-6.70784533e-01 -9.39824760e-01 6.36067152e-01 -6.80016756e-01
4.14437741e-01 2.57379025e-01 -8.75980020e-01 3.06646794e-01
-3.08718830e-01 3.04403096e-01 7.66952157e-01 5.53072467e-02
5.83919324e-02 -1.83270156e-01 -1.42223239e+00 7.57314637e-02
5.82714677e-01 -5.18296175e-02 -2.95970291e-01 4.19712543e-01
1.64556623e-01 -6.83912218e-01 -1.15545762e+00 6.85499907e-01
6.45262420e-01 -1.30507565e+00 9.26961720e-01 -8.21857095e-01
4.98731285e-01 -1.01430401e-01 2.73491830e-01 -9.94615257e-01
-2.69913226e-01 -4.73190308e-01 4.46075499e-02 7.68876314e-01
6.87758207e-01 -4.20754194e-01 1.07643473e+00 8.12482238e-01
-1.60065755e-01 -1.00357449e+00 -1.05833399e+00 -4.73817617e-01
3.60069543e-01 -3.11124861e-01 2.24050418e-01 9.57335114e-01
-3.54733855e-01 -4.65556644e-02 7.08886385e-02 -1.08045287e-01
7.32227445e-01 1.63618512e-02 5.34185886e-01 -1.59147823e+00
-2.48024270e-01 -8.84307504e-01 -7.38591313e-01 -1.64360076e-01
-1.97334155e-01 -9.26636875e-01 -7.17019662e-02 -1.89689326e+00
3.51875216e-01 -4.38149095e-01 -8.44298661e-01 3.71409237e-01
-3.71667027e-01 5.61720252e-01 1.03411213e-01 1.34939134e-01
-5.90584874e-01 -1.81898564e-01 1.29703736e+00 -6.55978173e-02
-2.58401662e-01 2.09355339e-01 -4.91559923e-01 6.19475842e-01
1.03020287e+00 -4.92370337e-01 -9.46699604e-02 -2.06818163e-01
9.04152263e-03 -7.45602185e-04 5.51301241e-01 -1.27961922e+00
-1.81681976e-01 1.77020952e-01 5.20753801e-01 -3.13516825e-01
-1.22420132e-01 -6.47577941e-01 -4.87324744e-02 9.88059402e-01
-4.29286242e-01 1.41419441e-01 2.82483906e-01 2.80400574e-01
-1.49918154e-01 7.90459812e-02 1.23471475e+00 -4.07448888e-01
-3.98482710e-01 4.17167813e-01 -4.20006439e-02 4.12759095e-01
9.36195791e-01 -2.60579437e-01 -5.20680025e-02 -4.76348288e-02
-1.16084373e+00 3.74480218e-01 2.82307267e-01 1.14106745e-01
3.94699156e-01 -9.07788217e-01 -1.10348272e+00 1.25578148e-02
-5.94253764e-02 -1.16698435e-02 3.47264618e-01 1.32268655e+00
-9.08358693e-01 3.47638696e-01 -2.59806812e-01 -1.25393188e+00
-1.22589505e+00 1.90567642e-01 8.12358439e-01 -8.32745671e-01
-6.66496158e-01 6.90037370e-01 -1.79220036e-01 -3.88329357e-01
1.37417197e-01 -4.94930416e-01 -7.09501803e-02 2.34933142e-02
3.58032197e-01 3.79073679e-01 4.81402248e-01 -5.05448341e-01
-4.16194856e-01 4.73218948e-01 1.51555300e-01 2.59615630e-02
1.23739028e+00 1.56197056e-01 -1.24313235e-01 4.26046550e-01
8.51926565e-01 -3.95830959e-01 -1.16044819e+00 2.93831415e-02
2.85240471e-01 -3.76580417e-01 3.76753688e-01 -1.28277326e+00
-1.22989345e+00 1.00720036e+00 1.24812138e+00 2.96591729e-01
9.40783083e-01 -1.33644834e-01 7.29182422e-01 -1.76162452e-01
3.31900679e-02 -1.07579088e+00 -8.74355733e-02 4.46650684e-02
3.83777320e-01 -1.56592870e+00 -2.05111150e-02 -2.89601564e-01
-5.40549517e-01 1.04689717e+00 3.13701838e-01 -1.65318325e-02
5.89032173e-01 1.82745203e-01 6.63718581e-01 -4.52293068e-01
-3.22009981e-01 -2.04324573e-01 5.13156831e-01 5.22863269e-01
7.56199956e-01 3.16302061e-01 -3.10792089e-01 2.91723698e-01
9.89791751e-02 1.56924635e-01 3.06547612e-01 7.77640700e-01
-3.62116903e-01 -1.14785576e+00 -8.36086571e-02 8.64293754e-01
-9.79911566e-01 5.36195301e-02 -2.13885963e-01 7.80432582e-01
2.07162544e-01 7.87505746e-01 -1.73942536e-01 6.46219868e-03
6.93783641e-01 3.52058917e-01 3.22315902e-01 -5.93180656e-01
-9.39679921e-01 1.76039696e-01 -2.01304659e-01 -8.01701486e-01
-4.39261258e-01 -7.08728969e-01 -1.20096540e+00 1.25925452e-01
-4.00284566e-02 6.91289231e-02 7.53224194e-01 7.65996575e-01
4.33445245e-01 9.48260427e-01 4.94312584e-01 -5.23479939e-01
-4.83716339e-01 -6.93719685e-01 -3.54639113e-01 7.14483559e-01
4.08713073e-01 -4.75987524e-01 -8.35970696e-03 1.88675486e-02] | [14.302199363708496, -2.4328978061676025] |
fdd8054d-968f-442e-8476-6326deb5f4cc | manydg-many-domain-generalization-for | 2301.08834 | null | https://arxiv.org/abs/2301.08834v2 | https://arxiv.org/pdf/2301.08834v2.pdf | ManyDG: Many-domain Generalization for Healthcare Applications | The vast amount of health data has been continuously collected for each patient, providing opportunities to support diverse healthcare predictive tasks such as seizure detection and hospitalization prediction. Existing models are mostly trained on other patients data and evaluated on new patients. Many of them might suffer from poor generalizability. One key reason can be overfitting due to the unique information related to patient identities and their data collection environments, referred to as patient covariates in the paper. These patient covariates usually do not contribute to predicting the targets but are often difficult to remove. As a result, they can bias the model training process and impede generalization. In healthcare applications, most existing domain generalization methods assume a small number of domains. In this paper, considering the diversity of patient covariates, we propose a new setting by treating each patient as a separate domain (leading to many domains). We develop a new domain generalization method ManyDG, that can scale to such many-domain problems. Our method identifies the patient domain covariates by mutual reconstruction and removes them via an orthogonal projection step. Extensive experiments show that ManyDG can boost the generalization performance on multiple real-world healthcare tasks (e.g., 3.7% Jaccard improvements on MIMIC drug recommendation) and support realistic but challenging settings such as insufficient data and continuous learning. | ['M. Brandon Westover', 'Jimeng Sun', 'Chaoqi Yang'] | 2023-01-21 | null | null | null | null | ['seizure-detection'] | ['medical'] | [ 3.31795633e-01 -4.24788967e-02 -5.10732949e-01 -5.37985682e-01
-5.70736110e-01 -3.62590939e-01 2.72449523e-01 2.33951569e-01
-1.68035269e-01 9.68642712e-01 3.96929055e-01 -1.48674533e-01
-3.86194944e-01 -4.84451473e-01 -6.04689658e-01 -8.55865955e-01
-3.16721834e-02 8.11259687e-01 -1.58542544e-01 -9.67234671e-02
-2.91956097e-01 7.54815117e-02 -9.59439933e-01 2.62833595e-01
1.20155394e+00 6.83732152e-01 1.42274037e-01 -1.61835849e-01
2.80444562e-01 3.48522276e-01 -6.81946158e-01 -1.28803372e-01
3.43768388e-01 -3.67095530e-01 -3.38698387e-01 3.07751242e-02
-1.31933853e-01 -4.66996312e-01 -3.59222054e-01 9.03280199e-01
7.27057636e-01 -1.54974937e-01 8.10964048e-01 -1.35457289e+00
-5.91611505e-01 7.03897119e-01 -6.03075266e-01 -9.31943208e-02
2.46389657e-01 -1.36080742e-01 6.87063396e-01 -6.47144496e-01
4.31502402e-01 9.64881003e-01 9.05248582e-01 9.22101974e-01
-1.35509634e+00 -1.09643507e+00 2.19575301e-01 -3.94910015e-02
-1.36702323e+00 -6.57743439e-02 7.95705259e-01 -6.67414963e-01
3.42985094e-01 3.28944713e-01 3.06235433e-01 1.54649913e+00
3.04814093e-02 6.59428418e-01 1.02111125e+00 2.26001009e-01
4.36901569e-01 2.32398316e-01 2.76313365e-01 -1.16629258e-01
5.06419778e-01 8.93712491e-02 -2.68406838e-01 -8.23835850e-01
6.61358118e-01 8.42966616e-01 -5.64342856e-01 -5.00586092e-01
-1.36667204e+00 8.42263401e-01 3.16162765e-01 -1.56657100e-02
-5.59812129e-01 -6.50531650e-01 4.37882274e-01 3.61538857e-01
3.60955417e-01 3.70573997e-01 -7.15478003e-01 2.24169701e-01
-5.92111707e-01 3.97377312e-01 7.89733410e-01 1.22751963e+00
4.63411480e-01 -3.41690272e-01 -8.46226811e-02 9.27241027e-01
-2.13705882e-01 5.05670369e-01 8.34668458e-01 -2.72989661e-01
7.59293497e-01 8.26734900e-01 1.85882196e-01 -8.73266757e-01
-7.47210860e-01 -6.33201420e-01 -1.46251094e+00 -5.34305573e-01
3.75906885e-01 -3.20239812e-01 -8.72895658e-01 1.96848500e+00
3.92448664e-01 4.60604727e-01 2.10048527e-01 8.06960225e-01
7.41208196e-01 1.22970253e-01 1.35563374e-01 -2.26022467e-01
1.23683333e+00 -6.03010774e-01 -5.56838334e-01 -3.33179176e-01
9.51511800e-01 -3.23652416e-01 9.58032429e-01 4.81626451e-01
-5.27977765e-01 -1.49235010e-01 -7.72717535e-01 4.78683025e-01
-5.49512021e-02 3.09183728e-02 5.80754220e-01 4.27479506e-01
-3.76619786e-01 6.06989443e-01 -7.67414808e-01 -3.45794618e-01
6.91759586e-01 5.97771764e-01 -4.18603599e-01 -5.86995721e-01
-1.28627956e+00 4.42539185e-01 2.16381729e-01 -3.27725053e-01
-6.51418567e-01 -1.09148860e+00 -5.28756201e-01 -8.09576884e-02
4.46399599e-01 -1.00979972e+00 9.16248381e-01 -6.80404782e-01
-9.15996671e-01 5.91070354e-01 -6.77256137e-02 -4.09747124e-01
6.11325264e-01 -2.74782628e-01 -6.08762145e-01 -3.23818952e-01
1.52955353e-01 1.74429849e-01 5.50292969e-01 -1.02665985e+00
-5.67570508e-01 -7.67889202e-01 -3.19520682e-01 2.22876415e-01
-6.79419994e-01 -2.04262987e-01 -1.72404796e-01 -1.04597294e+00
2.00504169e-01 -1.10621107e+00 -5.37166655e-01 -4.46513593e-01
-3.59646529e-01 -2.92626750e-02 5.55282950e-01 -7.91924059e-01
1.41680658e+00 -2.19965839e+00 1.88824609e-01 1.00620419e-01
6.21696174e-01 1.93366736e-01 -1.09366059e-01 3.16959381e-01
-3.99357885e-01 2.18335874e-02 -4.96007770e-01 -1.92760944e-01
-3.85847211e-01 3.44682217e-01 -3.98810327e-01 4.59519148e-01
1.04190959e-02 6.86706126e-01 -8.33016872e-01 -8.57533887e-02
-1.05416246e-01 4.21434730e-01 -7.06333339e-01 2.62881786e-01
2.51249932e-02 1.00494933e+00 -8.48329961e-01 5.13329387e-01
9.38409269e-01 -6.00694478e-01 2.13569269e-01 -2.08760314e-02
5.94854355e-01 2.25703120e-01 -1.06921434e+00 1.60206878e+00
-7.88445845e-02 -6.94489479e-02 -1.82004943e-01 -1.37773454e+00
9.34450686e-01 4.14051443e-01 9.84121799e-01 -4.61184502e-01
4.51988652e-02 2.32390404e-01 1.81421354e-01 -3.92572671e-01
-1.58882976e-01 -2.68168151e-01 -1.96845040e-01 2.90890008e-01
-4.45317388e-01 4.45098698e-01 -5.08113861e-01 8.65589920e-03
1.23911870e+00 -4.69511330e-01 6.84288025e-01 -2.22409129e-01
2.28899285e-01 -2.36305986e-02 1.22912562e+00 6.47802591e-01
-2.12591633e-01 8.43943238e-01 3.80760908e-01 -5.19557953e-01
-9.32681203e-01 -9.59229529e-01 -6.10994220e-01 7.59268641e-01
1.56075299e-01 -3.47764850e-01 -4.94857013e-01 -7.46660233e-01
3.15786660e-01 3.83190542e-01 -7.29260683e-01 -6.04246378e-01
-4.21116233e-01 -1.36604321e+00 3.45682234e-01 5.86726725e-01
-8.40189010e-02 -6.78637743e-01 -2.51742333e-01 4.71733421e-01
-1.79691002e-01 -1.11733496e+00 -5.34329474e-01 2.45025605e-01
-1.09038532e+00 -1.15668952e+00 -1.00179327e+00 -6.23554707e-01
6.86782897e-01 2.70805418e-01 1.14201796e+00 -1.27639517e-01
-1.75676540e-01 -4.79227342e-02 -3.54824394e-01 -5.40247142e-01
-3.77576143e-01 8.75624269e-02 5.97761035e-01 5.40483035e-02
7.56049991e-01 -7.78958797e-01 -8.97821486e-01 6.42626882e-01
-8.99953187e-01 -9.90933403e-02 6.44684315e-01 1.31581295e+00
6.57139361e-01 1.03574164e-01 1.03187597e+00 -1.45872378e+00
6.58528686e-01 -1.11568570e+00 -1.75600484e-01 1.71275929e-01
-8.49697113e-01 -1.05002098e-01 9.15804744e-01 -9.57745314e-01
-5.31843185e-01 8.74904245e-02 3.45303774e-01 -6.06209695e-01
-2.68897325e-01 4.95429307e-01 -4.88479853e-01 3.28657866e-01
6.75930142e-01 1.87088951e-01 1.04034245e-01 -8.15085709e-01
-1.97628573e-01 9.02379811e-01 1.81552887e-01 -4.10787344e-01
6.31659031e-01 4.19206142e-01 -2.46824939e-02 -4.37404603e-01
-8.14678907e-01 -6.44432962e-01 -4.08831984e-01 6.11444056e-01
4.15389627e-01 -1.38395035e+00 -4.48642850e-01 4.81017768e-01
-6.72924161e-01 -9.60028172e-02 -1.67149529e-02 7.53213823e-01
-2.66091228e-01 3.88174295e-01 -2.90378898e-01 -4.82879251e-01
-3.60711336e-01 -1.12852824e+00 9.00378704e-01 -8.01943690e-02
-3.21600914e-01 -9.31469440e-01 -5.61186159e-03 2.44998962e-01
1.93200484e-01 2.92541742e-01 1.02379751e+00 -1.14851642e+00
7.61717465e-03 -1.33220017e-01 -1.12293154e-01 2.80749679e-01
6.26225233e-01 -7.63559818e-01 -7.83581853e-01 -6.01158321e-01
2.83979058e-01 -6.35172985e-03 7.16507316e-01 4.93950993e-01
1.44031358e+00 -3.83130819e-01 -8.55760157e-01 8.33551466e-01
9.83670235e-01 3.32502753e-01 2.83155143e-01 7.74644017e-02
9.01307464e-01 5.53797662e-01 6.76160574e-01 8.79129827e-01
5.68366230e-01 7.50975370e-01 1.94322348e-01 -3.24778676e-01
3.35561872e-01 -2.23007947e-01 -4.64466251e-02 7.27077723e-01
2.53121823e-01 1.04198918e-01 -1.12880325e+00 5.12663305e-01
-2.09931087e+00 -5.95752478e-01 3.68136801e-02 2.46697044e+00
1.04135978e+00 -1.60542071e-01 4.13632423e-01 2.36069504e-02
8.58070672e-01 -4.99742091e-01 -1.24608672e+00 3.28074008e-01
-2.73240596e-01 -1.26577727e-02 6.77587807e-01 -1.04123712e-01
-1.03692913e+00 3.91348183e-01 5.52564240e+00 5.56565523e-01
-1.07597268e+00 1.49017692e-01 8.18385541e-01 -2.13178203e-01
-2.31996581e-01 -2.85231799e-01 -7.13477910e-01 7.84104526e-01
6.55353725e-01 -2.72196174e-01 3.91937315e-01 8.90025496e-01
2.41609678e-01 3.77817094e-01 -1.47379255e+00 1.33041012e+00
-1.90314520e-02 -8.26343715e-01 6.69262931e-02 2.94180542e-01
8.29022169e-01 9.21409950e-02 1.23230815e-01 2.87883967e-01
2.13401377e-01 -1.17338347e+00 -1.04744022e-03 4.02614504e-01
8.62664282e-01 -5.47988892e-01 6.41454995e-01 6.85860753e-01
-6.79175317e-01 -3.89249980e-01 -5.19999564e-01 3.05501875e-02
-1.20032199e-01 8.26484442e-01 -1.01301229e+00 6.13669932e-01
7.91848481e-01 1.07039607e+00 -2.70774305e-01 1.12071431e+00
3.93916249e-01 7.45463431e-01 -4.08419520e-01 3.88221323e-01
-2.61741370e-01 -2.52967745e-01 3.51422012e-01 6.64136052e-01
5.95882535e-01 3.62540394e-01 4.42024350e-01 5.26862800e-01
-1.24229930e-01 2.91716158e-01 -6.79601133e-01 1.15840323e-01
6.97409689e-01 7.31634319e-01 -9.52679291e-02 -2.78349072e-01
-6.31975889e-01 8.16782176e-01 1.96154624e-01 4.64899570e-01
-7.71948099e-01 1.05432412e-02 1.01489890e+00 3.58912289e-01
5.17903641e-02 3.18398476e-01 -4.57618952e-01 -1.51269317e+00
2.30829120e-01 -1.36253846e+00 7.51745522e-01 -5.13368957e-02
-1.80002701e+00 4.88323092e-01 -9.90912393e-02 -1.60544598e+00
-1.37668252e-01 -3.73773575e-01 -2.08829984e-01 9.93391514e-01
-1.53856337e+00 -7.60121346e-01 -3.09598327e-01 9.55840230e-01
3.10082510e-02 -4.59530622e-01 9.84827638e-01 7.31341362e-01
-7.54116595e-01 8.75857949e-01 5.75600684e-01 2.49608583e-03
1.24832380e+00 -9.96489763e-01 1.21368162e-01 2.67604560e-01
-2.90294558e-01 8.85004461e-01 4.77627665e-01 -7.06273258e-01
-1.17158949e+00 -1.33667743e+00 5.71851134e-01 -5.78267276e-01
4.27516043e-01 -4.28224444e-01 -1.26243806e+00 7.15754271e-01
-6.37198031e-01 1.89649742e-02 1.08047116e+00 3.41367036e-01
-4.28739637e-01 -3.39989096e-01 -1.26209402e+00 5.10809481e-01
1.26960313e+00 -2.79893726e-01 -5.48070908e-01 5.48095226e-01
5.84448993e-01 -3.92674834e-01 -1.16859829e+00 7.71184325e-01
3.16996515e-01 -6.12530947e-01 1.13396394e+00 -1.12167394e+00
1.92625627e-01 -4.92093451e-02 -4.08216827e-02 -1.39957333e+00
-4.31249052e-01 -6.53352678e-01 -1.28550097e-01 9.89390969e-01
3.70381862e-01 -9.60890889e-01 8.61111343e-01 9.96987522e-01
1.92927137e-01 -8.77781451e-01 -8.42675209e-01 -7.89976299e-01
3.76119405e-01 -9.54514146e-02 1.25970316e+00 1.35444844e+00
1.94228247e-01 4.57553238e-01 -8.53852749e-01 3.92417163e-01
6.41585588e-01 2.01610237e-01 8.13271523e-01 -1.79203057e+00
-4.27049518e-01 -2.32678935e-01 -4.16459084e-01 -9.74458635e-01
-6.71948045e-02 -7.75566161e-01 -3.91072065e-01 -1.16466761e+00
6.26670420e-01 -9.43055987e-01 -6.61608696e-01 6.07473195e-01
-6.29057705e-01 -2.70410478e-01 -2.55288064e-01 6.00035548e-01
-1.66159496e-01 4.48118865e-01 1.20944166e+00 -6.35880530e-02
-3.91595185e-01 5.48142910e-01 -1.17067325e+00 5.88873744e-01
8.95083547e-01 -7.35209525e-01 -6.43549323e-01 -3.01321328e-01
-1.58996359e-01 1.76876843e-01 2.16011882e-01 -9.16249573e-01
2.22207140e-03 -2.80583113e-01 3.59518111e-01 -3.32205653e-01
1.82057142e-01 -1.05851698e+00 3.99014205e-01 5.30988932e-01
-1.15508810e-01 5.61911464e-02 -3.71632054e-02 9.51120675e-01
-1.32612005e-01 2.65214533e-01 5.64399064e-01 1.09317265e-01
-3.24691743e-01 6.31460845e-01 3.19656521e-01 2.72091031e-01
1.20789790e+00 -4.50099446e-02 -2.09136039e-01 -2.52215177e-01
-9.33949888e-01 4.36242759e-01 5.36014199e-01 5.37518919e-01
4.58431393e-01 -1.32380259e+00 -8.50366712e-01 3.25577378e-01
5.51710010e-01 2.34455198e-01 3.28703642e-01 9.81379867e-01
3.58316958e-01 3.72392595e-01 -2.49638855e-02 -6.50275290e-01
-1.23627532e+00 9.40933764e-01 2.40412667e-01 -2.79729128e-01
-9.39752340e-01 4.90862042e-01 9.37848568e-01 -4.98810887e-01
3.72836679e-01 -4.13901150e-01 -1.80590332e-01 -6.06891140e-02
5.82556725e-01 4.91681285e-02 2.44878139e-02 -4.65910286e-01
-5.63377976e-01 4.19063121e-01 -3.55515122e-01 6.06985509e-01
1.56232512e+00 -6.78860992e-02 2.50978321e-01 2.36207068e-01
1.18793571e+00 -2.40018651e-01 -1.16957366e+00 -7.31084168e-01
-8.30467790e-02 -3.28881145e-01 -3.49258751e-01 -9.03085113e-01
-1.08199739e+00 7.03988016e-01 5.72887301e-01 -1.14497840e-01
1.31497610e+00 3.40366140e-02 8.44252646e-01 3.18962008e-01
3.82944763e-01 -7.87707567e-01 -2.61118799e-01 1.06893271e-01
5.74452400e-01 -1.56525481e+00 -5.89785986e-02 -4.73190844e-01
-9.21361744e-01 5.15790343e-01 4.33701754e-01 1.28494591e-01
7.24862397e-01 1.54881671e-01 4.79064658e-02 9.48343351e-02
-5.72257102e-01 3.11166108e-01 2.70125568e-01 7.36678541e-01
2.67294109e-01 5.10926604e-01 -3.86498898e-01 1.51888001e+00
-4.47664708e-02 2.00273708e-01 4.28047001e-01 4.96845454e-01
7.21231773e-02 -1.40234733e+00 -2.71844029e-01 9.26632047e-01
-4.39822674e-01 -2.26953045e-01 -1.68612927e-01 5.40055394e-01
4.95938510e-02 7.97476113e-01 -3.12269509e-01 -4.59806442e-01
6.13492846e-01 -1.46442205e-02 8.27984437e-02 -7.73598611e-01
-4.25931901e-01 -1.10075675e-01 -2.93252826e-01 -3.68428171e-01
-7.25522712e-02 -8.57481003e-01 -1.12270617e+00 -2.12238520e-01
-6.57707304e-02 1.83402732e-01 1.85461327e-01 7.33694494e-01
7.53767490e-01 4.65576589e-01 7.75505245e-01 -2.18702540e-01
-1.10622931e+00 -8.48181903e-01 -7.62319267e-01 1.05969036e+00
4.91919100e-01 -8.31123173e-01 -3.02656859e-01 -1.57541841e-01] | [10.346600532531738, 3.276689291000366] |
b4a1e374-3ce4-463e-86d4-d307d901e222 | multi-task-deep-cnn-model-for-no-reference | 2008.11961 | null | https://arxiv.org/abs/2008.11961v1 | https://arxiv.org/pdf/2008.11961v1.pdf | Multi-task deep CNN model for no-reference image quality assessment on smartphone camera photos | Smartphone is the most successful consumer electronic product in today's mobile social network era. The smartphone camera quality and its image post-processing capability is the dominant factor that impacts consumer's buying decision. However, the quality evaluation of photos taken from smartphones remains a labor-intensive work and relies on professional photographers and experts. As an extension of the prior CNN-based NR-IQA approach, we propose a multi-task deep CNN model with scene type detection as an auxiliary task. With the shared model parameters in the convolution layer, the learned feature maps could become more scene-relevant and enhance the performance. The evaluation result shows improved SROCC performance compared to traditional NR-IQA methods and single task CNN-based models. | ['Ja-Ling Wu', 'Chen-Hsiu Huang'] | 2020-08-27 | null | null | null | null | ['no-reference-image-quality-assessment'] | ['computer-vision'] | [ 2.28959456e-01 -6.59597397e-01 -1.32757425e-02 -5.36160290e-01
-8.21648896e-01 -3.67402107e-01 8.00851285e-02 -3.85148555e-01
-5.46048760e-01 3.06746423e-01 5.41457394e-03 -4.16332185e-01
1.54938325e-01 -8.35810304e-01 -6.71726704e-01 -7.12867737e-01
5.32128572e-01 -2.91377813e-01 1.77974403e-01 -2.79402226e-01
3.24607015e-01 2.44845867e-01 -1.29885745e+00 4.23027456e-01
9.87712383e-01 1.56922197e+00 4.25857216e-01 8.17672670e-01
1.34951428e-01 6.97556078e-01 -5.76781690e-01 -9.33387160e-01
5.04813612e-01 -6.31210729e-02 -2.14306235e-01 2.32102156e-01
3.80764097e-01 -6.91937149e-01 -3.85517538e-01 1.54239893e+00
6.16715610e-01 3.21836281e-03 2.74968892e-01 -1.08419859e+00
-1.07561529e+00 1.53334066e-01 -7.64106870e-01 6.35027230e-01
1.53674901e-01 4.44090784e-01 6.97850287e-01 -8.88502002e-01
2.08539143e-02 1.20022273e+00 7.93710530e-01 -2.02409878e-01
-7.85671711e-01 -9.95732725e-01 -4.39647883e-02 8.98669243e-01
-1.45435572e+00 -4.00001973e-01 9.05122697e-01 -1.71902046e-01
7.13610888e-01 -6.77930862e-02 6.00149393e-01 8.32867384e-01
2.38740951e-01 7.84051299e-01 1.21638441e+00 -9.93534997e-02
-2.12929845e-02 3.38757366e-01 -1.93400159e-01 4.24888402e-01
3.06109667e-01 -5.57525009e-02 -1.75938234e-01 4.79368895e-01
9.34567153e-01 2.68461943e-01 1.77589823e-02 2.46881709e-01
-9.07162666e-01 8.20031941e-01 5.39323151e-01 2.87987530e-01
-3.20858449e-01 6.70785233e-02 2.93658733e-01 1.64809033e-01
7.37717688e-01 1.83339149e-01 -4.29507434e-01 -1.78543016e-01
-1.06751108e+00 -1.39290318e-01 2.97169000e-01 1.00813580e+00
6.09352231e-01 5.98719791e-02 -6.87516555e-02 9.57717061e-01
2.88962662e-01 8.38193357e-01 3.02428752e-01 -8.43615234e-01
4.51467782e-01 7.62578905e-01 4.02805284e-02 -1.35607004e+00
-1.81147560e-01 -9.73786592e-01 -1.02121425e+00 2.58309999e-03
2.03083858e-01 2.49776454e-03 -6.25405908e-01 1.05799627e+00
1.88591868e-01 1.35353133e-01 -1.57357350e-01 1.10045516e+00
7.33537078e-01 8.40602279e-01 -8.79920945e-02 -1.38996348e-01
1.54296148e+00 -1.18738520e+00 -8.89141977e-01 6.14382420e-03
2.69520879e-01 -1.19754779e+00 1.09322679e+00 8.69756162e-01
-1.02100003e+00 -1.08669591e+00 -1.32983029e+00 -4.31768507e-01
-4.26340103e-01 6.93622828e-01 8.33238959e-01 1.02575302e+00
-9.65382278e-01 3.10994893e-01 -1.81821167e-01 -5.69784306e-02
9.19797242e-01 2.04640344e-01 1.26931285e-02 -5.12192786e-01
-1.21120965e+00 7.10301876e-01 1.98641330e-01 5.66257656e-01
-8.77793372e-01 -4.04267222e-01 -5.95574200e-01 2.21293494e-01
5.00895917e-01 -4.67432708e-01 1.43248940e+00 -1.49483597e+00
-1.77933693e+00 5.70181608e-01 -3.19638886e-02 -2.72562802e-01
4.66857105e-01 -2.61306852e-01 -8.49819958e-01 2.58267581e-01
3.50663960e-02 3.16650808e-01 1.05539775e+00 -9.56636369e-01
-9.86930311e-01 -1.77331552e-01 3.51009637e-01 4.94200736e-01
-1.48713470e-01 3.53914171e-01 -9.19677615e-01 -5.87439597e-01
-1.85976803e-01 -6.74764633e-01 -1.40772164e-01 -2.45362818e-02
-3.34763974e-01 -7.31186196e-02 9.41158652e-01 -1.08203638e+00
9.84293640e-01 -2.29786754e+00 -5.28841257e-01 -7.17222989e-02
8.76226723e-02 5.95896125e-01 -1.31647334e-01 -2.17692852e-01
9.65755060e-02 -2.01809276e-02 1.83348119e-01 -3.66834849e-01
-7.80647397e-02 -3.69689137e-01 6.39391989e-02 5.29530227e-01
1.00881338e-01 9.38426912e-01 -7.05794156e-01 -5.16296864e-01
7.43139148e-01 5.47850966e-01 -3.00240517e-01 4.01437990e-02
1.30295455e-01 3.31957966e-01 -2.19912484e-01 8.63348722e-01
1.29814875e+00 -6.09492600e-01 -2.18636468e-01 -8.07319880e-01
-9.17077735e-02 7.15475678e-02 -1.16051173e+00 1.65305841e+00
-8.27545464e-01 7.09200263e-01 6.47476614e-02 -8.23826194e-01
6.47128403e-01 -5.90077154e-02 2.07178548e-01 -1.29131794e+00
4.64308470e-01 5.18505052e-02 1.69226691e-01 -5.77561140e-01
7.02970505e-01 1.31139427e-01 1.04512669e-01 6.83389157e-02
-1.32406697e-01 7.66094998e-02 -1.58392146e-01 2.84769759e-02
6.54554188e-01 -7.95376003e-02 2.92098552e-01 -1.08400382e-01
7.04075634e-01 -2.35976964e-01 6.89935207e-01 4.39374089e-01
-4.42439944e-01 7.42690146e-01 4.50701788e-02 -4.99733478e-01
-1.28208733e+00 -8.65800381e-01 -1.63574785e-01 8.51488531e-01
5.78248262e-01 2.46701501e-02 -7.95919597e-01 -5.26537597e-01
-4.32138532e-01 5.05952775e-01 -2.52838045e-01 -1.47064015e-01
-4.15207505e-01 -8.97307158e-01 3.70293379e-01 3.54715079e-01
1.52892959e+00 -7.69357204e-01 -2.56972700e-01 5.83027005e-02
-3.56498778e-01 -1.57051909e+00 -7.39970505e-01 -3.47852439e-01
-3.81222576e-01 -1.04805541e+00 -7.67163098e-01 -6.41259789e-01
3.79752398e-01 8.13588023e-01 8.06334913e-01 -1.62362665e-01
-2.53956884e-01 -6.62015304e-02 -4.25073236e-01 -4.69967186e-01
-1.13807134e-02 -1.06606618e-01 -2.94121444e-01 5.11460781e-01
4.71319586e-01 -4.14611369e-01 -1.37855291e+00 3.81738633e-01
-7.95180738e-01 2.53509790e-01 9.40376997e-01 5.42740643e-01
4.88247067e-01 6.98071599e-01 4.68123257e-01 -6.19665444e-01
5.68346679e-01 -3.76290828e-01 -8.28317285e-01 8.69937539e-02
-6.91305339e-01 -7.66288042e-01 7.32390821e-01 -3.89642268e-01
-1.47283483e+00 3.67698744e-02 -1.94549337e-01 -4.05126303e-01
-1.42889425e-01 1.32639930e-01 -6.46987379e-01 -1.81529775e-01
1.72285736e-01 3.60571861e-01 -3.41404110e-01 -4.30398166e-01
3.43301952e-01 1.01630986e+00 3.83411974e-01 3.47442478e-01
6.18999302e-01 5.17748773e-01 -2.08522320e-01 -9.36304927e-01
-7.60588944e-01 -6.75997794e-01 -3.54157925e-01 -4.13403749e-01
9.67106044e-01 -1.34914589e+00 -8.98492217e-01 8.59711111e-01
-1.39308965e+00 -1.40502200e-01 1.38573453e-01 5.66899955e-01
-2.09328115e-01 4.67157483e-01 -6.32680595e-01 -7.20137119e-01
-5.31198740e-01 -1.45148361e+00 1.22855127e+00 4.66029167e-01
5.53929031e-01 -5.97543418e-01 -3.84239912e-01 7.77684450e-01
6.29181325e-01 -3.88137221e-01 5.38815558e-01 -3.22837800e-01
-9.57241237e-01 -2.77091920e-01 -1.04130483e+00 9.23313618e-01
-7.39519577e-03 -2.64932454e-01 -1.32453060e+00 -4.97628450e-02
3.16724569e-01 8.54476020e-02 5.26212871e-01 7.50545084e-01
1.48621309e+00 -3.35893810e-01 5.29969409e-02 9.18851495e-01
1.66735065e+00 5.90727448e-01 1.01794147e+00 3.38458061e-01
9.07883227e-01 2.55377412e-01 7.53603101e-01 2.94968426e-01
4.59025770e-01 7.35612214e-01 4.99808848e-01 -6.11092806e-01
-2.65418470e-01 -1.06512159e-01 3.11488539e-01 7.90313065e-01
4.11555618e-02 -4.30040479e-01 -2.92668402e-01 3.96259487e-01
-1.61273682e+00 -8.15803707e-01 -2.08265021e-01 1.81613219e+00
3.46080691e-01 -1.53113455e-01 1.40149659e-02 1.57621235e-01
1.02727270e+00 2.58912537e-02 -8.05620193e-01 -3.46585959e-01
-4.48301047e-01 7.80215114e-02 8.58104885e-01 6.76754862e-02
-1.24884856e+00 8.12891841e-01 6.33916807e+00 1.23952878e+00
-1.11898494e+00 6.83227420e-01 1.07678640e+00 -9.41630080e-02
-3.89189906e-02 -3.47380280e-01 -6.38550997e-01 7.90157378e-01
7.53439963e-01 2.83503711e-01 5.21251619e-01 9.35886025e-01
5.54887056e-01 -4.34413105e-01 -6.85220957e-01 1.74931717e+00
3.73655289e-01 -1.16511083e+00 -1.62534416e-01 4.22212094e-01
7.99015343e-01 -2.32467204e-02 4.76253718e-01 1.95096463e-01
2.09424749e-01 -9.94244814e-01 7.84600317e-01 3.69842589e-01
9.32654142e-01 -8.66540849e-01 1.12212765e+00 -1.28355429e-01
-1.19603825e+00 -2.85801709e-01 -4.54179943e-01 -1.27099633e-01
3.53034705e-01 9.04000878e-01 -4.90388989e-01 4.63898361e-01
1.07743371e+00 7.21264839e-01 -8.14810157e-01 1.05286932e+00
-5.25044948e-02 6.26864731e-01 -3.44313122e-02 -9.52806976e-03
1.03601687e-01 -4.57416683e-01 3.14685524e-01 1.25295877e+00
5.24601758e-01 6.75627068e-02 1.02887156e-04 7.89631844e-01
-2.80577421e-01 2.74908096e-01 -4.04335171e-01 2.16989920e-01
5.16567864e-02 1.69699609e+00 -7.91170120e-01 -4.00502086e-01
-8.14605594e-01 1.39779270e+00 -3.15901548e-01 3.13711941e-01
-1.04708469e+00 -5.49263179e-01 6.18820786e-01 2.01397650e-02
4.86317456e-01 -9.40716118e-02 -3.83800507e-01 -1.13261211e+00
1.76818803e-01 -9.61687803e-01 -1.41618494e-02 -1.20868981e+00
-1.23287582e+00 3.37440223e-01 -4.04985458e-01 -1.39610398e+00
3.04480433e-01 -8.16739380e-01 -5.60445189e-01 8.78630161e-01
-1.80483425e+00 -1.32065761e+00 -4.63233739e-01 7.92552531e-01
7.76553452e-01 -1.28071845e-01 1.83426887e-01 8.04829419e-01
-7.40812302e-01 7.73960650e-01 2.03512460e-01 2.39912108e-01
4.38873649e-01 -9.26295340e-01 2.68329412e-01 1.34597945e+00
-2.06919894e-01 3.51628780e-01 4.23363328e-01 -4.95262414e-01
-1.32380569e+00 -1.23857641e+00 3.52678567e-01 -8.90437588e-02
3.96724463e-01 -2.61072904e-01 -5.21583974e-01 2.88463384e-01
5.04118025e-01 -1.05575018e-01 6.19064033e-01 -2.91670449e-02
-2.14373037e-01 -6.63980544e-01 -1.06895041e+00 4.59597617e-01
9.45414007e-01 -6.94534123e-01 1.23083703e-02 2.62172073e-01
8.57503474e-01 -9.47707370e-02 -7.32772291e-01 2.46168062e-01
5.47225773e-01 -1.09508038e+00 9.84113336e-01 2.45125726e-01
2.60692000e-01 -4.48186338e-01 -2.96195537e-01 -1.12973332e+00
-2.63964266e-01 -4.63324279e-01 3.46502662e-01 1.14390600e+00
2.95368373e-01 -5.38108528e-01 6.12651587e-01 2.51001358e-01
-1.29073888e-01 -3.68394315e-01 -7.03975439e-01 -5.69932044e-01
-5.20669281e-01 -7.81113148e-01 7.52523303e-01 5.04398882e-01
-6.12211823e-01 4.07688200e-01 -6.57084644e-01 3.53362262e-01
5.42914152e-01 2.58474182e-02 8.54643226e-01 -8.27887893e-01
-4.07214254e-01 -3.11200976e-01 -4.37380314e-01 -1.14744270e+00
-4.03713584e-01 -3.71791214e-01 -6.50055408e-02 -1.43486452e+00
4.53239501e-01 -1.64648697e-01 -4.06869441e-01 -1.12647869e-01
-2.88072843e-02 5.37990093e-01 1.66283414e-01 1.84122086e-01
-1.12062752e+00 4.41107988e-01 1.45714045e+00 -2.90677726e-01
5.14110774e-02 6.27872050e-02 -9.38403547e-01 7.77236402e-01
7.55679905e-01 -7.17095584e-02 -5.01269400e-01 -4.24119502e-01
6.38818562e-01 -2.17840374e-01 4.95384157e-01 -9.49390173e-01
2.78687477e-01 8.64914060e-02 6.65120363e-01 -7.92195857e-01
4.75650340e-01 -8.92359912e-01 8.00929889e-02 3.61377507e-01
-5.30574471e-02 1.33814082e-01 1.02700859e-01 6.65319920e-01
-3.74672025e-01 9.82088596e-02 7.81129003e-01 -2.21185580e-01
-8.49768996e-01 4.67408299e-01 -2.28452548e-01 -4.49113816e-01
7.26867557e-01 -5.50446510e-01 -3.76639426e-01 -6.18825197e-01
-1.06573552e-01 -1.18482791e-01 2.55628705e-01 3.89651150e-01
8.28447342e-01 -1.23515034e+00 -4.61401969e-01 1.43836783e-02
-4.27873153e-03 -3.25730368e-02 8.76758397e-01 7.96600759e-01
-5.81578493e-01 4.02337074e-01 -2.09266469e-01 -5.89171231e-01
-1.08887124e+00 7.14252830e-01 2.58375257e-01 -1.38664350e-01
-2.39668325e-01 5.61910689e-01 4.75745291e-01 1.02436943e-02
-6.05269894e-02 -2.98013955e-01 -4.60413426e-01 -2.05977157e-01
7.65731633e-01 6.60964310e-01 2.96810627e-01 -8.36131811e-01
-2.57303193e-02 5.35990119e-01 -1.29328668e-01 -4.53917608e-02
1.14597905e+00 -7.48890102e-01 1.58556268e-01 1.57336771e-01
1.47146702e+00 -1.71375051e-01 -1.31031990e+00 -2.19410554e-01
-5.43237865e-01 -6.79261267e-01 6.25211179e-01 -9.33480620e-01
-1.33387518e+00 1.12201083e+00 1.20468736e+00 -2.44444050e-02
1.58150268e+00 -3.77773136e-01 1.10715806e+00 2.14481860e-01
2.52806455e-01 -1.42285430e+00 3.39002997e-01 5.48302829e-02
7.17575014e-01 -1.61778128e+00 4.46981099e-03 -6.09179556e-01
-5.77321827e-01 9.02616262e-01 5.88778377e-01 -5.58454022e-02
6.46899939e-01 -1.10986479e-01 9.12472755e-02 -7.82344192e-02
-8.91342834e-02 -3.23666751e-01 4.47912306e-01 6.44457281e-01
9.89602786e-03 2.29140986e-02 1.31711051e-01 7.69974768e-01
-1.21529996e-01 1.54762268e-01 3.50191206e-01 2.90159345e-01
-2.63530076e-01 -5.42750418e-01 -2.52162665e-01 4.59917694e-01
-7.46935666e-01 -4.53948885e-01 1.87471464e-01 5.30503511e-01
5.19309282e-01 1.41449583e+00 8.52341205e-02 -7.24587202e-01
2.75842994e-01 -4.97243971e-01 2.61875570e-01 -3.63255262e-01
-4.71851617e-01 2.85183012e-01 -1.38833271e-02 -7.89665401e-01
-5.59531748e-01 -5.06302238e-01 -5.17717779e-01 -5.84932089e-01
-5.69265604e-01 -4.51360792e-01 9.96672571e-01 1.00673711e+00
4.18157279e-01 6.12310171e-01 9.45597827e-01 -7.55963802e-01
-4.18509431e-02 -1.08406973e+00 -5.86251378e-01 3.54049087e-01
1.30075306e-01 -4.62612808e-01 -1.28713071e-01 2.68722661e-02] | [11.642634391784668, -1.9256056547164917] |
c81a4d6f-0fc0-403d-87c9-e238bef66ec4 | foit-fast-online-instance-transfer-for | null | null | https://www.researchgate.net/profile/Jinpeng-Li-3/publication/348932155_FOIT_Fast_Online_Instance_Transfer_for_Improved_EEG_Emotion_Recognition/links/6017f2a692851c2d4d0b0b69/FOIT-Fast-Online-Instance-Transfer-for-Improved-EEG-Emotion-Recognition.pdf | https://www.researchgate.net/profile/Jinpeng-Li-3/publication/348932155_FOIT_Fast_Online_Instance_Transfer_for_Improved_EEG_Emotion_Recognition/links/6017f2a692851c2d4d0b0b69/FOIT-Fast-Online-Instance-Transfer-for-Improved-EEG-Emotion-Recognition.pdf | FOIT: Fast Online Instance Transfer for Improved EEG Emotion Recognition | The Electroencephalogram (EEG)-based emotion recognition is promising yet limited by the requirement of a large number of training data. Collecting substantial labeled samples in the training trails is the key to the generalization on the test trails. This process is time-consuming and laborious. In recent years, several studies have proposed various semi-supervised learning (e.g., active learning) and transfer learning (e.g., domain adaptation, style transfer mapping) methods to alleviate the requirement on training data. However, most of them are iterative methods, which need considerable training time and are unfeasible in practice. To tackle this problem, we present the Fast Online Instance Transfer (FOIT) for improved affective Brain-computer Interface (aBCI). FOIT selects auxiliary data from historical sessions and (or) other subjects heuristically, which are then combined with the training data for supervised training. The predictions on the test trails are made by the multiclassifier ensemble. As a one-shot algorithm, FOIT avoids time-consuming iterations. Experimental results show that FOIT brings significant improvement in accuracy for the three category classification (1%-8%) on the SEED dataset and four-category classification (1%-14%) on the SEED-IV dataset in the cross-subject, cross-session, and cross-all scenarios. The time cost over the baselines is moderate (~35s in average for our machine). To achieve the comparative accuracies, the iterative methods require much more time (~45s-~900s). FOIT provides a simple, fast, and practically feasible solution to improve the generalization of aBCIs and allows various choices of classifiers without constrains. Our codes are available online. | ['Ting Cai', 'Hao Chen', 'Jinpeng Li'] | 2021-02-01 | null | null | null | 2020-ieee-international-conference-on-3 | ['eeg-emotion-recognition'] | ['miscellaneous'] | [ 3.13779145e-01 -1.57999724e-01 -1.54251426e-01 -7.55104899e-01
-8.29792261e-01 -2.84704328e-01 3.25458422e-02 3.32139693e-02
-6.55189931e-01 1.06006718e+00 -4.63710368e-01 -1.10124074e-01
-2.47215673e-01 -4.50005323e-01 -4.00221616e-01 -8.43307674e-01
-2.91373134e-01 2.90263712e-01 1.62065879e-01 -1.26525104e-01
1.68721989e-01 3.52027893e-01 -1.50457680e+00 3.83075178e-01
1.46917069e+00 1.29594874e+00 2.24867046e-01 4.94601205e-02
-1.09604508e-01 3.36931556e-01 -6.58774257e-01 -3.47768664e-01
4.75938767e-02 -5.37451744e-01 -6.24736965e-01 2.15926743e-03
-3.24847817e-01 -8.19573775e-02 2.82150745e-01 8.27111363e-01
7.43172586e-01 3.31028283e-01 8.10818613e-01 -1.73898852e+00
-2.98377484e-01 2.69135118e-01 -6.33842945e-01 7.50073791e-02
2.26745158e-01 -1.06705077e-01 3.42989981e-01 -1.21580064e+00
1.00766607e-01 7.14226365e-01 6.80580437e-01 7.73252666e-01
-1.14584351e+00 -1.40747559e+00 1.98076218e-01 5.81007540e-01
-1.47511077e+00 -5.94490290e-01 8.16858411e-01 -3.71693403e-01
7.95404136e-01 2.62793273e-01 7.95855701e-01 1.38657308e+00
1.54214635e-01 9.10861254e-01 1.46391129e+00 -3.44646960e-01
7.18270183e-01 6.31677806e-01 5.49444437e-01 2.29633316e-01
2.35279538e-02 1.07919832e-03 -7.49054968e-01 -1.82581365e-01
4.60841238e-01 1.18816523e-02 -3.14572066e-01 -1.40179604e-01
-6.31601393e-01 6.87096477e-01 2.96313584e-01 1.67071581e-01
-4.80590791e-01 -6.31067991e-01 5.95633149e-01 5.59428394e-01
7.31353402e-01 4.85393375e-01 -6.69692934e-01 -3.71990949e-01
-9.41506445e-01 -2.50644654e-01 6.85042322e-01 1.01616788e+00
6.76741421e-01 3.43239270e-02 6.68123066e-02 1.14838243e+00
-8.10714439e-02 3.28053743e-01 5.98606825e-01 -2.59967506e-01
4.84442234e-01 6.77523971e-01 7.32933506e-02 -7.90693581e-01
-5.50021529e-01 -3.91122192e-01 -9.57777202e-01 9.12562013e-02
-2.58389059e-02 -4.43935066e-01 -9.26676929e-01 1.54786885e+00
9.56592485e-02 4.43644196e-01 1.08591430e-02 7.67820179e-01
6.60927534e-01 6.32837057e-01 3.98235202e-01 -9.19871032e-01
8.55475664e-01 -7.82276332e-01 -7.09155083e-01 -3.26612175e-01
6.38841033e-01 -3.62585217e-01 1.15269601e+00 6.61814094e-01
-7.85201252e-01 -5.09564042e-01 -1.09576619e+00 6.70385659e-01
-3.17443937e-01 5.93821965e-02 9.10306096e-01 7.75099218e-01
-8.50075245e-01 4.32423919e-01 -8.33132565e-01 -3.83680612e-01
6.78783715e-01 7.80909061e-01 -3.23005944e-01 3.11394650e-02
-1.24827874e+00 8.00471425e-01 3.83816123e-01 1.24524929e-01
-7.46584594e-01 -8.09233069e-01 -5.16237080e-01 1.42875865e-01
2.06578255e-01 -1.98075306e-02 8.09206367e-01 -1.46936548e+00
-1.64658475e+00 4.59130526e-01 -9.99132767e-02 -2.22679436e-01
2.21184790e-01 -6.45950735e-02 -6.19722188e-01 4.53561805e-02
-1.96366042e-01 6.27177119e-01 7.37134159e-01 -1.06016147e+00
-4.53595340e-01 -3.43501329e-01 -3.41543049e-01 3.08951735e-01
-8.01838934e-01 2.16970235e-01 -2.45672405e-01 -4.44384217e-01
-3.72403637e-02 -1.15862489e+00 1.50980055e-01 -1.79856986e-01
1.10306926e-01 -4.18658078e-01 7.49596357e-01 -6.34808183e-01
1.13041735e+00 -2.39739656e+00 -7.11207464e-02 2.59543628e-01
-2.05842480e-01 4.72461432e-01 -8.11539590e-02 1.06395014e-01
-4.47388649e-01 -6.85989633e-02 -3.85459423e-01 -1.33618608e-01
-3.22532535e-01 -7.88546447e-03 -1.30501196e-01 1.96585268e-01
2.07226172e-01 7.38067687e-01 -7.58755088e-01 -3.68956000e-01
-1.27596064e-02 2.00357512e-01 -4.24027532e-01 4.79931980e-01
4.99580026e-01 7.14342713e-01 -1.67833924e-01 5.64602852e-01
5.59579194e-01 -1.88575789e-01 1.36686638e-02 -1.16037749e-01
1.09379448e-01 1.33914426e-01 -1.07229161e+00 1.68951309e+00
-4.20212746e-01 4.24982369e-01 -2.18498483e-01 -1.42853582e+00
1.06646466e+00 5.76847374e-01 6.66512072e-01 -8.52206290e-01
1.66918829e-01 3.19907248e-01 2.09496573e-01 -4.96990323e-01
-6.01549745e-02 -2.28790388e-01 -6.78834915e-02 3.86639178e-01
3.41101199e-01 3.21904987e-01 -1.21425003e-01 -2.41266619e-02
8.68828297e-01 8.48453492e-02 3.15366536e-01 -4.01573837e-01
2.01108322e-01 -2.20650017e-01 9.92457569e-01 4.26910043e-01
-3.70466292e-01 1.24168180e-01 1.25914142e-01 -2.14151248e-01
-6.20961845e-01 -8.84369731e-01 -4.30546224e-01 8.87649894e-01
1.54322520e-01 -1.53811306e-01 -7.57697344e-01 -7.05897272e-01
-3.00896585e-01 8.91080856e-01 -5.78774989e-01 -6.54728711e-01
-2.43711490e-02 -1.15731490e+00 7.56417355e-03 8.46223831e-01
7.81770170e-01 -1.08587062e+00 -5.18911779e-01 2.28824794e-01
-2.40961730e-01 -7.25884974e-01 -3.51631641e-02 5.96818030e-01
-1.05197537e+00 -6.38032913e-01 -6.42888308e-01 -7.21946955e-01
8.27451527e-01 1.44482762e-01 6.34081066e-01 -8.36070627e-02
-1.51527941e-01 2.34535992e-01 -5.81125081e-01 -7.50479221e-01
3.39561045e-01 -4.46362868e-02 5.68820357e-01 2.90949970e-01
5.84522367e-01 -7.55976975e-01 -6.06930554e-01 6.04446888e-01
-3.05755734e-01 1.05968066e-01 6.73690021e-01 1.09669292e+00
4.28370833e-01 9.34037045e-02 1.17115116e+00 -7.90000975e-01
5.56393027e-01 -6.31652415e-01 -3.30518156e-01 4.75829899e-01
-9.02858913e-01 -4.27957058e-01 6.06139898e-01 -1.07184303e+00
-1.33528125e+00 4.37318496e-02 1.52037472e-01 -3.65845025e-01
-1.60143659e-01 4.75390345e-01 -1.93314329e-01 -3.34456593e-01
6.94525838e-01 1.28179446e-01 -2.04816297e-01 -2.83452272e-01
-2.97720581e-01 1.17596126e+00 5.04093990e-02 -4.19042885e-01
4.73227084e-01 -1.15687653e-01 -4.91769999e-01 -6.45957530e-01
-6.96889579e-01 -3.23498100e-01 -6.91717565e-01 -5.12134850e-01
5.71202576e-01 -9.87790287e-01 -5.34730792e-01 5.01813352e-01
-7.97857106e-01 -5.47736108e-01 1.11079149e-01 8.42808068e-01
-4.11711007e-01 -7.48804510e-02 -3.85955632e-01 -1.11664307e+00
-6.66463733e-01 -1.01438415e+00 5.57943821e-01 4.53603446e-01
-5.23723006e-01 -6.86685622e-01 -2.87064791e-01 2.32366964e-01
3.58046770e-01 1.55295087e-02 9.70426559e-01 -8.52590799e-01
-1.01218745e-01 -1.71711504e-01 -4.26498875e-02 3.00694406e-01
1.75274625e-01 -4.20476884e-01 -1.23081481e+00 -4.86986846e-01
2.90245235e-01 -6.28793120e-01 3.21982920e-01 1.69236332e-01
1.39250696e+00 -1.26572736e-02 -3.73023957e-01 4.12784785e-01
1.20531666e+00 1.12575996e+00 8.19709897e-01 1.67533144e-01
2.46279866e-01 4.94438976e-01 8.65687728e-01 3.60743344e-01
4.51148711e-02 5.32934546e-01 -2.26080462e-01 -2.37752706e-01
3.34853381e-01 2.51325607e-01 4.78780538e-01 1.22035289e+00
-1.04765423e-01 9.87028033e-02 -9.30538177e-01 4.62103158e-01
-1.69440663e+00 -8.08302522e-01 1.03075199e-01 2.46276903e+00
9.75636780e-01 1.54105589e-01 1.03242755e-01 4.97668415e-01
6.07801974e-01 -6.41678452e-01 -8.47670734e-01 -4.48296279e-01
3.01106095e-01 4.95555758e-01 -2.93527208e-02 2.92807519e-02
-9.18770730e-01 6.08526945e-01 5.72731161e+00 8.66092980e-01
-1.38423991e+00 2.73020655e-01 7.67524660e-01 -3.30068648e-01
3.15922022e-01 -1.90617800e-01 -5.89044750e-01 8.22530568e-01
1.21832585e+00 -2.88915724e-01 6.14280283e-01 9.04352069e-01
8.51029158e-02 -4.12867188e-01 -1.14738548e+00 1.38605690e+00
7.68278986e-02 -7.75609076e-01 -4.21036184e-01 -2.73240387e-01
6.28783822e-01 -2.79302746e-01 -1.49820372e-01 8.19839239e-01
-3.26616198e-01 -6.98926091e-01 4.09819186e-01 5.29442847e-01
9.40385818e-01 -9.29854929e-01 7.46309757e-01 4.93831307e-01
-8.19209754e-01 -3.44826311e-01 -3.36120397e-01 7.13891760e-02
-2.06916511e-01 4.72862035e-01 -3.82854730e-01 4.98194665e-01
1.05550456e+00 6.35658920e-01 -4.36981767e-01 1.14800906e+00
-1.75054055e-02 9.98730302e-01 -1.64587498e-01 -3.44038308e-01
-1.75867215e-01 -3.55752707e-01 1.89120620e-01 8.83640707e-01
3.59419048e-01 5.88923812e-01 1.87999353e-01 5.43390274e-01
8.31275657e-02 2.95745164e-01 -3.27517182e-01 7.43962675e-02
4.45378870e-01 1.31202245e+00 -7.44827032e-01 -3.48638624e-01
-4.11689073e-01 1.17749918e+00 4.05738950e-01 4.98053432e-01
-1.04024148e+00 -8.63899171e-01 1.20170057e-01 -1.92815751e-01
-2.07394913e-01 1.73875332e-01 -4.84155774e-01 -1.18289840e+00
-9.74097941e-03 -8.08952212e-01 2.71274686e-01 -9.38695014e-01
-1.32003200e+00 7.51086712e-01 1.04857504e-01 -1.46042943e+00
-1.16777427e-01 -4.30282295e-01 -6.49964869e-01 7.82568812e-01
-1.00767958e+00 -4.29186791e-01 -4.62992102e-01 8.62260699e-01
5.14589548e-01 -3.09893101e-01 1.23772562e+00 6.50847197e-01
-8.80590260e-01 8.99091840e-01 1.91566423e-01 1.53344050e-01
8.76233995e-01 -7.40077555e-01 -2.45929569e-01 5.03034234e-01
-1.55630216e-01 7.17876852e-01 2.68783152e-01 -4.18127239e-01
-1.08667970e+00 -9.30989563e-01 5.27172506e-01 -8.17302167e-02
5.07775009e-01 -6.61553860e-01 -1.22453785e+00 4.69885230e-01
1.30853146e-01 -2.05182970e-01 1.15567791e+00 5.19789875e-01
8.94863009e-02 -6.12335801e-01 -1.17337978e+00 4.83093768e-01
9.34559643e-01 -4.65273887e-01 -5.06719649e-01 2.06541076e-01
2.25790143e-01 -1.22753970e-01 -9.62085962e-01 4.57895577e-01
4.63991076e-01 -6.35567367e-01 4.00769800e-01 -5.25096476e-01
3.39753143e-02 1.84566349e-01 1.86784819e-01 -1.58120573e+00
-4.31218952e-01 -5.83288968e-01 1.59877047e-01 1.34805620e+00
7.16735363e-01 -7.88554311e-01 5.43959796e-01 1.10934615e+00
-2.38909766e-01 -1.05581641e+00 -9.28719759e-01 -9.30090427e-01
-1.89550146e-01 -5.29737234e-01 3.97558004e-01 1.17579937e+00
6.27870619e-01 4.82852906e-01 -3.18664312e-01 -1.90743268e-01
5.04228234e-01 6.87011867e-04 3.51593286e-01 -1.46215892e+00
1.10597081e-01 9.46101174e-02 -4.18770820e-01 -4.57122922e-01
1.15943156e-01 -9.50898051e-01 1.05130136e-01 -1.19326127e+00
3.98357242e-01 -6.81549430e-01 -7.60166466e-01 9.07793880e-01
-2.68144161e-01 -1.75823807e-03 -8.51946175e-02 2.01923564e-01
-5.51347435e-01 7.62894094e-01 8.00140202e-01 -5.22137759e-03
-4.52145249e-01 -2.22291872e-02 -5.77164412e-01 5.75068176e-01
1.03249395e+00 -6.99170947e-01 -8.55328083e-01 -1.03792995e-01
-1.02998398e-01 1.97120070e-01 3.63079309e-02 -1.21471441e+00
1.76054209e-01 -1.61881134e-01 7.52983749e-01 -3.84979904e-01
4.75381702e-01 -8.45389009e-01 1.29564971e-01 1.47607729e-01
-2.81544536e-01 -2.08849475e-01 4.77014959e-01 5.15067041e-01
-9.42431577e-03 -2.09573388e-01 9.62219417e-01 2.06006020e-01
-7.90025413e-01 4.93641824e-01 -4.29867774e-01 -6.21313490e-02
1.32987773e+00 -3.43435913e-01 -1.98552348e-02 -1.53088152e-01
-8.90943289e-01 1.95442379e-01 -7.79784992e-02 4.21892375e-01
6.42938197e-01 -1.32278371e+00 -3.27153713e-01 3.39604467e-01
2.24627420e-01 -3.75463724e-01 3.95920396e-01 1.24456823e+00
2.71384537e-01 2.25178301e-01 -5.53180575e-01 -6.28895819e-01
-1.23113012e+00 4.73679572e-01 2.72750288e-01 7.71636739e-02
-4.03144866e-01 8.21076989e-01 4.50407118e-02 -1.84705555e-01
3.46274555e-01 1.00308307e-01 -1.96667045e-01 3.56434375e-01
5.60111701e-01 3.01844269e-01 2.99744427e-01 -2.76510343e-02
-5.41940749e-01 1.59233302e-01 -2.59448975e-01 2.88810711e-02
1.60633910e+00 1.32401481e-01 5.78843877e-02 8.05329382e-01
1.33322728e+00 -5.47814190e-01 -1.08415401e+00 -9.20119882e-02
3.65213305e-02 -2.60245860e-01 4.88100760e-02 -1.12500393e+00
-1.00895190e+00 9.56137002e-01 1.03441322e+00 -9.06160027e-02
1.60506201e+00 -3.94718587e-01 4.97465670e-01 4.63744074e-01
9.23224509e-01 -1.36046994e+00 9.64101870e-03 1.23587012e-01
9.28154290e-01 -1.19249141e+00 -2.08695550e-02 -2.92763114e-01
-1.02270305e+00 8.62575054e-01 9.47326303e-01 4.52870727e-02
7.07502544e-01 2.02190205e-01 -2.71829162e-02 6.19197525e-02
-9.42966640e-01 3.16043347e-01 3.12130600e-01 6.12867177e-01
2.82789320e-01 1.16185293e-01 -4.87221569e-01 1.12913847e+00
2.16264501e-01 2.32714206e-01 1.10072143e-01 8.06557417e-01
-1.86892390e-01 -9.14999664e-01 -3.88532430e-02 7.67535567e-01
-3.94714437e-02 -9.11838114e-02 -2.42199570e-01 7.31888831e-01
1.12906590e-01 1.21872556e+00 2.46469360e-02 -7.22654462e-01
3.50362569e-01 5.65683544e-01 4.31337386e-01 -5.63495874e-01
-6.69313312e-01 4.88901958e-02 -2.03711241e-02 -4.86430317e-01
-5.08978605e-01 -6.71671748e-01 -1.25156987e+00 3.28936763e-02
-6.85276926e-01 4.16359365e-01 5.40346622e-01 9.10922647e-01
5.51413953e-01 3.69248807e-01 9.50510204e-01 -6.44158721e-01
-3.03453684e-01 -1.19900739e+00 -7.26844728e-01 4.49688405e-01
-3.16352159e-01 -1.06741691e+00 -5.19809723e-01 -3.46300714e-02] | [13.128241539001465, 3.4134652614593506] |
ea9b4201-76a9-4c34-aa38-2d9ce88fb608 | care-coherent-actionable-recourse-based-on | 2108.08197 | null | https://arxiv.org/abs/2108.08197v1 | https://arxiv.org/pdf/2108.08197v1.pdf | CARE: Coherent Actionable Recourse based on Sound Counterfactual Explanations | Counterfactual explanation methods interpret the outputs of a machine learning model in the form of "what-if scenarios" without compromising the fidelity-interpretability trade-off. They explain how to obtain a desired prediction from the model by recommending small changes to the input features, aka recourse. We believe an actionable recourse should be created based on sound counterfactual explanations originating from the distribution of the ground-truth data and linked to the domain knowledge. Moreover, it needs to preserve the coherency between changed/unchanged features while satisfying user/domain-specified constraints. This paper introduces CARE, a modular explanation framework that addresses the model- and user-level desiderata in a consecutive and structured manner. We tackle the existing requirements by proposing novel and efficient solutions that are formulated in a multi-objective optimization framework. The designed framework enables including arbitrary requirements and generating counterfactual explanations and actionable recourse by choice. As a model-agnostic approach, CARE generates multiple, diverse explanations for any black-box model in tabular classification and regression settings. Several experiments on standard data sets and black-box models demonstrate the effectiveness of our modular framework and its superior performance compared to the baselines. | ['Ingrid Chieh Yu', 'Peyman Rasouli'] | 2021-08-18 | null | null | null | null | ['counterfactual-explanation'] | ['miscellaneous'] | [ 3.85206848e-01 7.03447521e-01 -6.19045615e-01 -7.11725771e-01
-6.24464989e-01 -5.37961185e-01 6.84936166e-01 2.43123192e-02
-1.61047339e-01 1.24136484e+00 4.27207768e-01 -6.13858163e-01
-8.38185728e-01 -6.96312129e-01 -7.23312199e-01 -6.38955712e-01
1.13345027e-01 6.19254053e-01 -4.97288644e-01 -7.49539807e-02
6.07212245e-01 2.97357440e-01 -1.87831581e+00 5.20336568e-01
1.21309638e+00 8.35368395e-01 7.82823712e-02 3.89617354e-01
3.04632075e-02 5.76405227e-01 -2.48547837e-01 -6.48845136e-01
4.27406728e-01 -5.07598698e-01 -7.30527103e-01 2.56489277e-01
2.25053027e-01 -9.84605178e-02 1.35503829e-01 7.80879438e-01
1.52632952e-01 1.70791760e-01 8.78446877e-01 -1.68478739e+00
-8.31557989e-01 7.68498123e-01 -2.68302917e-01 -1.65465266e-01
3.02529424e-01 3.49358112e-01 1.29190922e+00 -5.71961045e-01
3.70185196e-01 1.34362793e+00 3.69078636e-01 7.19359219e-01
-1.58503389e+00 -3.94767404e-01 6.22069001e-01 3.29763204e-01
-6.78525090e-01 -2.95642942e-01 7.96962321e-01 -3.34525436e-01
8.34193885e-01 9.12850916e-01 4.41141605e-01 1.16288972e+00
4.01461184e-01 4.51999933e-01 1.14173400e+00 -4.58473325e-01
5.72004914e-01 5.10876536e-01 -2.17581564e-03 5.16290188e-01
6.73883617e-01 4.93555516e-01 -6.69650435e-01 -3.79910737e-01
2.46077597e-01 1.73592895e-01 -3.92288566e-01 -7.83799112e-01
-1.29138803e+00 1.07692790e+00 2.88587749e-01 -4.14608382e-02
-5.86007476e-01 1.35577887e-01 3.20785701e-01 5.01946568e-01
4.41384643e-01 8.79440427e-01 -1.08141840e+00 3.49004060e-01
-5.90174913e-01 3.67471874e-01 6.99624836e-01 7.51301825e-01
6.08228981e-01 1.22676238e-01 -3.01580548e-01 2.78094798e-01
4.83545601e-01 4.07782406e-01 7.28791535e-01 -9.63778675e-01
3.95895302e-01 7.78135121e-01 5.37190080e-01 -1.00516534e+00
-3.31455380e-01 -6.89361453e-01 -6.51978552e-01 3.81034791e-01
2.88893193e-01 -2.24822134e-01 -7.09420264e-01 1.96842420e+00
5.12219489e-01 -5.98578453e-02 2.03480974e-01 9.90971863e-01
3.37320477e-01 4.68320161e-01 7.58172572e-02 -6.39003396e-01
1.12000465e+00 -7.42758811e-01 -8.03531110e-01 -2.71279305e-01
6.98258162e-01 -3.61319840e-01 1.32022476e+00 3.10952038e-01
-7.36808300e-01 -3.27950776e-01 -1.03300726e+00 2.32395440e-01
-3.60269994e-01 6.57785013e-02 9.03334618e-01 8.83293688e-01
-6.75305247e-01 7.14861989e-01 -2.69019425e-01 3.50377969e-02
1.30449027e-01 5.21056235e-01 -3.25364769e-01 3.97885710e-01
-1.20133841e+00 9.15819466e-01 6.41099095e-01 1.69178128e-01
-7.97989249e-01 -9.27430153e-01 -8.66332471e-01 4.26870584e-01
7.08051085e-01 -1.24057412e+00 1.14164305e+00 -1.39158142e+00
-1.49894452e+00 5.10798395e-01 -3.86970714e-02 -7.67317355e-01
9.42966640e-01 -2.69869417e-02 -4.19726163e-01 -3.76405448e-01
1.67480186e-01 4.08222973e-01 8.12770724e-01 -1.37919426e+00
-7.40225196e-01 -3.99473369e-01 2.76860923e-01 2.28520587e-01
-2.65511312e-02 -4.84294206e-01 3.80163372e-01 -6.38445795e-01
4.33435999e-02 -8.14618051e-01 -5.14025211e-01 -2.06894934e-01
-6.62966669e-01 8.36841613e-02 5.69524586e-01 -2.48379484e-01
1.20770633e+00 -1.76617026e+00 1.71472747e-02 2.66710073e-01
-3.14791836e-02 -2.52597868e-01 -4.81266230e-02 2.54562259e-01
-4.66562748e-01 3.59776706e-01 -4.62295473e-01 -1.50956810e-01
3.72017890e-01 1.50982231e-01 -6.26347005e-01 5.12143970e-01
5.33916280e-02 6.75886869e-01 -8.32697928e-01 -3.20306629e-01
3.06702167e-01 -1.12632252e-01 -8.79531145e-01 2.51634568e-01
-5.72348297e-01 2.08568245e-01 -4.85396653e-01 3.29609811e-01
4.33429569e-01 -1.27623588e-01 4.21855271e-01 -2.09271163e-02
-8.72813538e-02 3.35288912e-01 -1.40845644e+00 1.37260795e+00
-6.68770909e-01 2.43111566e-01 -4.20635492e-01 -1.14223969e+00
8.38695467e-01 3.68602902e-01 1.09503739e-01 -3.71073395e-01
3.86724025e-02 4.01584744e-01 1.08403474e-01 -5.23927689e-01
2.26845130e-01 -4.77078885e-01 -2.42151052e-01 5.31964540e-01
-9.94107053e-02 5.65427653e-02 -1.31833017e-01 -1.16658926e-01
5.40103197e-01 2.09672332e-01 9.78507042e-01 -6.68438554e-01
6.00013316e-01 1.79681018e-01 7.27486789e-01 1.01973951e+00
2.81680137e-01 5.83807945e-01 6.97327971e-01 -7.66047120e-01
-7.42558837e-01 -8.59963059e-01 2.46988479e-02 7.61698306e-01
-1.29709408e-01 6.70921504e-02 -3.96336019e-01 -1.19487786e+00
2.83570319e-01 1.50444865e+00 -1.06672132e+00 -3.43588352e-01
-1.71594501e-01 -9.36716735e-01 -1.67951897e-01 1.99423805e-01
3.42458516e-01 -8.79519403e-01 -9.69016492e-01 5.15743643e-02
3.44707668e-02 -3.64408463e-01 -3.56300354e-01 2.63002723e-01
-9.15164828e-01 -1.23345292e+00 -1.08102687e-01 -4.13776413e-02
7.69812524e-01 8.34009498e-02 1.00250709e+00 1.09679913e-02
1.95992798e-01 3.99309322e-02 -6.02256507e-02 -5.45517504e-01
-2.53539532e-01 -2.72824049e-01 1.28509328e-01 3.95740420e-01
4.89696823e-02 -5.66819727e-01 -6.36729717e-01 2.24338919e-01
-8.16376984e-01 3.75896037e-01 5.71676493e-01 1.23706627e+00
4.88834202e-01 -9.43891257e-02 9.84635532e-01 -1.30186081e+00
7.33528674e-01 -7.70546615e-01 -7.16507137e-01 6.18445694e-01
-1.35167551e+00 7.49889851e-01 1.06759655e+00 -3.47457200e-01
-1.45193362e+00 2.22895481e-02 3.92172307e-01 -1.27862841e-01
-2.17776641e-01 5.22962987e-01 -6.63295567e-01 5.31576753e-01
7.31380582e-01 6.14590384e-02 -2.84074515e-01 -4.50669169e-01
8.38060915e-01 5.18283188e-01 4.03328627e-01 -7.54355490e-01
6.61543131e-01 5.07404983e-01 8.58912244e-02 -2.17186920e-02
-8.82894695e-01 7.03334063e-02 -4.70574468e-01 -3.99087183e-02
3.60291213e-01 -3.93176347e-01 -8.33449185e-01 -3.52544159e-01
-1.08430815e+00 2.46778000e-02 -6.89554155e-01 5.12414932e-01
-9.73493278e-01 2.78051253e-02 5.67049742e-01 -1.08125842e+00
-2.15900108e-01 -1.13402057e+00 6.81301594e-01 3.68143693e-02
-5.42438030e-01 -1.07678127e+00 1.53680602e-02 3.14076304e-01
2.04972550e-01 6.11390531e-01 1.20164621e+00 -1.04086292e+00
-5.81355274e-01 1.67538561e-02 -9.29804146e-03 -1.00539796e-01
2.68232673e-01 -9.03020874e-02 -1.00472426e+00 -1.05042616e-02
1.54535264e-01 5.33219166e-02 8.31443131e-01 4.53965843e-01
1.21788454e+00 -1.23892176e+00 -2.52592504e-01 4.70787942e-01
1.49713755e+00 3.67886811e-01 9.39688906e-02 5.31262279e-01
8.46175104e-02 9.14042771e-01 7.76182771e-01 6.72574818e-01
2.50733584e-01 7.04233170e-01 8.45963418e-01 4.27414030e-02
4.96209532e-01 -4.57539797e-01 1.73084036e-01 -1.76569119e-01
1.76736981e-01 -2.99553543e-01 -6.17797375e-01 7.48389602e-01
-2.26573396e+00 -1.20768607e+00 -2.96013236e-01 2.38196039e+00
6.09796286e-01 1.18564539e-01 -5.41896932e-02 1.55172423e-01
4.97565597e-01 2.14708410e-02 -9.14652526e-01 -9.63273227e-01
-4.02114503e-02 -3.12151074e-01 3.61875623e-01 6.76219583e-01
-8.01878393e-01 3.73530746e-01 5.80470514e+00 6.03612483e-01
-9.58808720e-01 5.62668219e-02 8.47051084e-01 -4.40753043e-01
-1.12304556e+00 3.35056186e-01 -3.76914382e-01 5.02960026e-01
1.07466257e+00 -8.64567339e-01 3.75339180e-01 1.03079236e+00
7.71242857e-01 2.72805002e-02 -1.49140573e+00 5.34354031e-01
-1.92394346e-01 -1.57314062e+00 3.88701826e-01 2.14861669e-02
8.65919054e-01 -5.43988943e-01 2.40011558e-01 2.20598459e-01
4.64190751e-01 -1.03759968e+00 1.15587163e+00 6.32444203e-01
5.45466721e-01 -7.47460008e-01 8.42190981e-01 6.23286664e-01
-5.37013710e-01 -4.17298287e-01 -1.47651017e-01 -3.57666582e-01
-1.46244824e-01 5.30676365e-01 -8.64660740e-01 1.03213644e+00
2.64995664e-01 3.90680432e-01 -2.64032185e-01 7.32952237e-01
-3.79677504e-01 3.62873584e-01 -2.37253346e-02 -8.90918076e-02
2.85710990e-01 -1.63940206e-01 5.12038410e-01 9.99671698e-01
3.39223474e-01 5.45848571e-02 -2.42936954e-01 1.14286995e+00
1.85291320e-01 1.91838011e-01 -7.81709611e-01 3.93161744e-01
3.82075518e-01 8.67791831e-01 -4.27283049e-01 -3.29659253e-01
-1.19016014e-01 5.95089853e-01 1.29438743e-01 4.23362970e-01
-8.36922109e-01 4.64006551e-02 6.86744153e-01 -3.67915034e-02
-1.14802592e-01 6.09471500e-01 -9.03519809e-01 -1.25666022e+00
1.32633552e-01 -1.09054184e+00 8.08344901e-01 -6.92319691e-01
-1.20963967e+00 4.77219373e-01 3.37121099e-01 -1.27651906e+00
-6.16380155e-01 -4.84330982e-01 -6.97387576e-01 8.15941393e-01
-1.53282809e+00 -9.31362927e-01 1.11862481e-01 2.94448286e-01
6.11629844e-01 -1.34920999e-01 8.66668701e-01 -2.89237320e-01
-5.18495977e-01 4.43951190e-01 1.85760871e-01 -7.35157907e-01
3.24575216e-01 -1.51232219e+00 -7.71326646e-02 8.61831784e-01
1.55114830e-01 7.82907665e-01 1.27521789e+00 -4.57554340e-01
-1.05246747e+00 -1.07966208e+00 1.25970495e+00 -3.05337161e-01
5.21045089e-01 -9.52245891e-02 -6.56464696e-01 7.61062384e-01
2.30138242e-01 -4.16748971e-01 6.93941712e-01 2.73026079e-01
-2.71074444e-01 -1.88487157e-01 -1.50548351e+00 7.91826844e-01
8.98081839e-01 -7.01215640e-02 -1.06356561e+00 2.47227237e-01
7.98671305e-01 -5.01436815e-02 -2.53598362e-01 3.47841471e-01
6.89320147e-01 -1.16753662e+00 7.52637804e-01 -1.31862593e+00
5.82183123e-01 -3.13782930e-01 -2.94005603e-01 -1.58836162e+00
-1.89630538e-01 -7.27881730e-01 -1.07149575e-02 1.05791640e+00
8.52740169e-01 -8.71405661e-01 6.16657555e-01 1.17316663e+00
4.07881849e-02 -8.52207065e-01 -1.16747689e+00 -7.09774137e-01
-1.94571584e-01 -4.87041563e-01 1.23611999e+00 1.02521002e+00
1.83388382e-01 1.36270866e-01 -6.25692070e-01 2.70655930e-01
5.99717319e-01 6.77526951e-01 7.51375139e-01 -1.00094700e+00
-3.96813095e-01 -4.27091718e-01 -1.60933211e-02 -4.08443421e-01
3.78909737e-01 -8.49279523e-01 -2.86111087e-01 -1.35325789e+00
2.77066439e-01 -1.62811860e-01 -4.53508586e-01 5.94660401e-01
-3.05546403e-01 -5.40263236e-01 3.35202605e-01 -7.73257464e-02
-3.16462398e-01 7.85348654e-01 8.82097423e-01 -7.04688802e-02
-2.35949412e-01 1.83264077e-01 -1.23397326e+00 8.41102064e-01
8.64227772e-01 -7.74410367e-01 -7.75134444e-01 -2.04909161e-01
3.70339870e-01 3.42216551e-01 6.37755692e-01 -3.48290980e-01
-1.09866433e-01 -9.03104722e-01 7.57277384e-02 -2.49642685e-01
7.64193684e-02 -1.14820349e+00 7.13745177e-01 5.39306521e-01
-8.68076205e-01 5.29172011e-02 7.80784572e-03 8.04001749e-01
-4.17711511e-02 -3.82178277e-01 5.49181819e-01 -9.61312503e-02
-4.44790214e-01 -5.30245565e-02 -5.68714999e-02 -2.12458909e-01
1.19237840e+00 -1.98212489e-01 -3.00506741e-01 -3.42430919e-01
-7.49431670e-01 3.27872515e-01 3.33574444e-01 4.86663520e-01
4.98286992e-01 -1.21751165e+00 -8.25114667e-01 1.12882286e-01
2.68654108e-01 -4.54953730e-01 2.76993841e-01 4.94682014e-01
8.47939923e-02 8.67344797e-01 -1.99617341e-01 -5.19999824e-02
-9.01971519e-01 7.21861303e-01 5.60824752e-01 -5.65278709e-01
-2.07187563e-01 3.14601690e-01 4.21210647e-01 -6.61791325e-01
-1.82126269e-01 -3.90912712e-01 -2.24434808e-01 5.77683523e-02
1.72593802e-01 3.71280491e-01 -1.93975978e-02 -6.96621984e-02
-3.03992063e-01 -8.57389197e-02 2.33064905e-01 -2.10343212e-01
1.42052984e+00 -2.30317086e-01 2.55822062e-01 4.20279890e-01
7.31702507e-01 -6.97578490e-02 -1.28667748e+00 6.80097863e-02
3.23649079e-01 -7.81723857e-01 -5.30358702e-02 -1.24820113e+00
-8.44453275e-01 5.50296366e-01 2.66591936e-01 3.21332335e-01
1.17548895e+00 -2.40257084e-01 -1.52644709e-01 3.55342686e-01
2.28016302e-01 -8.04815829e-01 -4.55114275e-01 -2.86438227e-01
1.38705945e+00 -1.27984285e+00 -4.19868715e-02 -1.05285212e-01
-8.53849709e-01 9.88446712e-01 5.94370544e-01 7.04969242e-02
2.62814403e-01 -1.58361748e-01 -1.61211953e-01 -6.83451071e-02
-1.35830760e+00 5.45411222e-02 5.45358419e-01 5.11941075e-01
2.92508453e-01 3.32095832e-01 -8.54628801e-01 1.13726461e+00
-3.84142458e-01 -8.55426192e-02 5.26496530e-01 2.74870664e-01
-1.89254865e-01 -9.51019168e-01 -4.89914387e-01 6.71342134e-01
-2.95085639e-01 -1.77978501e-02 -4.23709780e-01 1.13338530e+00
1.61204919e-01 1.09960234e+00 -2.24723876e-01 -1.37202397e-01
5.23446500e-01 1.72777176e-01 -7.18785748e-02 -6.24566436e-01
-5.87431848e-01 -2.75063932e-01 3.79730523e-01 -6.19265437e-01
-3.36207509e-01 -7.78821886e-01 -1.04826355e+00 -2.24603578e-01
-3.82983685e-01 4.80133176e-01 4.77327615e-01 1.25137591e+00
5.77198803e-01 5.29956162e-01 8.32314849e-01 -3.56494486e-01
-1.09213340e+00 -7.04560518e-01 -4.36906517e-01 4.37106878e-01
5.12008309e-01 -7.90910423e-01 -6.91674888e-01 4.99244183e-02] | [8.729100227355957, 5.640194892883301] |
3ad5bca9-9349-40ea-9d88-24905becd68b | constrained-sampling-for-class-agnostic | 2209.09195 | null | https://arxiv.org/abs/2209.09195v1 | https://arxiv.org/pdf/2209.09195v1.pdf | Constrained Sampling for Class-Agnostic Weakly Supervised Object Localization | Self-supervised vision transformers can generate accurate localization maps of the objects in an image. However, since they decompose the scene into multiple maps containing various objects, and they do not rely on any explicit supervisory signal, they cannot distinguish between the object of interest from other objects, as required in weakly-supervised object localization (WSOL). To address this issue, we propose leveraging the multiple maps generated by the different transformer heads to acquire pseudo-labels for training a WSOL model. In particular, a new discriminative proposals sampling method is introduced that relies on a pretrained CNN classifier to identify discriminative regions. Then, foreground and background pixels are sampled from these regions in order to train a WSOL model for generating activation maps that can accurately localize objects belonging to a specific class. Empirical results on the challenging CUB benchmark dataset indicate that our proposed approach can outperform state-of-art methods over a wide range of threshold values. Our method provides class activation maps with a better coverage of foreground object regions w.r.t. the background. | ['Eric Granger', 'Aydin Sarraf', 'Marco Pedersoli', 'Soufiane Belharbi', 'Shakeeb Murtaza'] | 2022-09-09 | null | null | null | null | ['weakly-supervised-object-localization'] | ['computer-vision'] | [ 5.42026043e-01 1.50869176e-01 -2.59312183e-01 -4.77972120e-01
-9.54180419e-01 -5.55531919e-01 6.64271355e-01 1.10288054e-01
-2.54724860e-01 6.32667184e-01 -3.36863339e-01 2.37024620e-01
2.23087549e-01 -8.88322949e-01 -1.00695622e+00 -1.02907670e+00
3.60629737e-01 6.39864683e-01 9.24714029e-01 3.22824687e-01
2.52492398e-01 5.32879353e-01 -1.52334917e+00 4.40113395e-01
8.95780683e-01 1.40721905e+00 5.75685799e-01 9.21024382e-02
-2.79541820e-01 1.03363276e+00 -8.36822629e-01 8.17511678e-02
2.04913259e-01 -4.03779536e-01 -6.35122418e-01 3.68047655e-01
7.24130094e-01 -1.70321077e-01 1.86090529e-01 1.18902791e+00
4.88102920e-02 -1.26031026e-01 7.79247761e-01 -1.19680393e+00
-3.58644396e-01 6.03641450e-01 -4.96883273e-01 2.99176723e-01
-1.20631959e-02 1.24113321e-01 9.51733470e-01 -1.11729217e+00
5.16812146e-01 1.03734148e+00 3.32422346e-01 3.89508396e-01
-1.64980590e+00 -6.73431098e-01 4.62843001e-01 1.95150763e-01
-1.45405567e+00 -4.16535020e-01 1.12364769e+00 -4.35044497e-01
2.44582072e-01 1.90863654e-01 6.45925164e-01 1.11986017e+00
-1.36894241e-01 1.08870459e+00 1.37106323e+00 -2.97853470e-01
5.45473814e-01 5.48918188e-01 2.19007745e-01 7.72728920e-01
2.50172436e-01 -2.29848698e-01 -6.92497313e-01 -8.47815350e-02
7.53975213e-01 1.10767111e-01 -2.40287170e-01 -8.24533641e-01
-1.33692479e+00 6.18674934e-01 8.10279965e-01 2.92507440e-01
-4.28633839e-01 1.31132692e-01 7.14158118e-02 -3.50030154e-01
4.82241064e-01 3.27871203e-01 -2.31203333e-01 5.91817141e-01
-1.07456934e+00 -1.18184045e-01 4.83026475e-01 9.49577689e-01
1.17926979e+00 1.24162261e-03 -4.92047071e-01 7.57494509e-01
3.62853050e-01 5.31749010e-01 2.38571689e-01 -6.10734761e-01
3.78378451e-01 9.43155766e-01 1.52652219e-01 -8.87863874e-01
5.31593859e-02 -6.99789584e-01 -6.01875544e-01 1.94952235e-01
5.42095542e-01 3.24059248e-01 -1.01609075e+00 1.58635938e+00
5.88638842e-01 4.16325420e-01 3.39313969e-02 1.00549245e+00
7.12236762e-01 6.94972456e-01 1.03217244e-01 1.43422723e-01
1.31807363e+00 -1.09056926e+00 -3.62854153e-01 -6.55338109e-01
9.60229486e-02 -4.05559689e-01 1.06980431e+00 3.02574277e-01
-8.66432488e-01 -7.59991109e-01 -9.67872977e-01 2.37641960e-01
-3.11783373e-01 7.31987000e-01 5.23496270e-01 3.61682355e-01
-8.53308678e-01 1.64160937e-01 -8.81195724e-01 -1.81752428e-01
9.89402115e-01 1.58791199e-01 -1.45774394e-01 3.47889327e-02
-7.57545412e-01 7.16481209e-01 6.62168443e-01 2.18239084e-01
-1.82383418e+00 -4.30278510e-01 -7.13616669e-01 4.07596752e-02
5.41425049e-01 -8.92646313e-02 8.40248466e-01 -1.47256994e+00
-1.30052114e+00 9.64291692e-01 -2.19553292e-01 -5.27140856e-01
5.56397974e-01 -5.73049784e-02 -3.16786394e-02 3.32838267e-01
5.08789420e-01 7.68793464e-01 1.30168772e+00 -1.62739742e+00
-9.18408811e-01 -3.15014929e-01 1.62666321e-01 -8.58130679e-03
-2.05700025e-01 -1.33585423e-01 -5.51336050e-01 -4.88385081e-01
3.29392165e-01 -7.76003957e-01 -5.97781651e-02 3.67563032e-02
-6.96962416e-01 -3.89031798e-01 9.80292618e-01 -2.89577216e-01
6.05328798e-01 -2.03010583e+00 1.42168313e-01 2.05211729e-01
3.22478294e-01 1.82881132e-01 -1.21482452e-02 -1.93855733e-01
2.29638904e-01 -3.95547599e-01 -3.08845460e-01 -3.16628516e-01
-2.39936322e-01 1.58209786e-01 -4.90704268e-01 6.32774889e-01
6.39987171e-01 7.28125453e-01 -9.31300879e-01 -7.07264841e-01
3.58412594e-01 3.80382329e-01 -1.48583442e-01 3.79305929e-01
-5.29143155e-01 6.25210226e-01 -5.57284772e-01 9.15416002e-01
6.51980221e-01 -4.90652770e-01 -1.81982536e-02 -3.47741932e-01
1.00646533e-01 1.59290656e-01 -1.00354886e+00 1.48513973e+00
-3.14237684e-01 6.18323207e-01 -4.28370982e-02 -1.29283929e+00
1.19426513e+00 -9.02153626e-02 2.17298567e-01 -3.46373469e-01
6.47706538e-02 1.17121048e-01 -1.43167898e-01 -9.92122889e-02
-4.10761870e-02 9.81455296e-02 8.09642151e-02 2.81022608e-01
4.07541573e-01 1.33291721e-01 1.08875714e-01 4.31209579e-02
9.12024915e-01 2.73442686e-01 8.37378427e-02 -4.87408817e-01
7.57530808e-01 9.24226567e-02 7.18039572e-01 8.08675706e-01
-1.90410033e-01 5.32481730e-01 3.26258779e-01 -4.55537915e-01
-6.20956540e-01 -1.30178976e+00 -2.23447278e-01 1.19141150e+00
6.18016839e-01 6.86513409e-02 -1.00646400e+00 -1.10368192e+00
-1.83163449e-01 5.82052350e-01 -7.74271071e-01 -1.29342601e-01
-4.91382897e-01 -6.66928589e-01 3.96180093e-01 5.61873794e-01
6.68306053e-01 -1.18105721e+00 -1.05008841e+00 1.51182324e-01
-3.29194218e-01 -1.27571642e+00 -2.03370154e-01 6.28895938e-01
-7.21176088e-01 -1.24268782e+00 -7.47906804e-01 -9.53374386e-01
1.23196661e+00 2.61431992e-01 1.01966941e+00 -2.36799344e-01
-2.23344818e-01 2.19363689e-01 -1.57010868e-01 -4.07065302e-01
-2.73728132e-01 7.56796896e-02 -2.07402885e-01 7.03018785e-01
3.97608280e-01 -1.52122647e-01 -5.34312427e-01 6.53026640e-01
-6.57316506e-01 1.63623229e-01 8.41011822e-01 7.58296132e-01
8.88891578e-01 -1.62687264e-02 5.39355338e-01 -9.14591134e-01
-3.68990265e-02 -4.38118428e-01 -9.75100338e-01 4.69075292e-01
-2.65789717e-01 1.72685295e-01 5.98149061e-01 -6.91061258e-01
-1.11358070e+00 5.45625329e-01 4.30635750e-01 -3.77049625e-01
-4.94180650e-01 -8.07276964e-02 -3.42904329e-01 -9.05212238e-02
6.57461345e-01 6.63680613e-01 -3.41337651e-01 -3.40421110e-01
3.29332441e-01 4.25414234e-01 6.07087553e-01 -5.34963906e-01
8.99530888e-01 7.50407934e-01 -2.02603623e-01 -4.48196948e-01
-1.15211844e+00 -4.31663573e-01 -7.90459931e-01 -4.33169425e-01
9.77602124e-01 -9.40948069e-01 -2.24348709e-01 2.56396204e-01
-1.19104254e+00 -3.72733772e-01 -3.06177258e-01 2.91315138e-01
-4.08223242e-01 -9.44096223e-02 -2.08976179e-01 -8.38609397e-01
-1.03984371e-01 -1.35347438e+00 1.31513405e+00 2.92777419e-01
4.45426404e-02 -6.85987651e-01 -3.77655327e-01 3.64705622e-01
3.10733289e-01 2.78854638e-01 6.30004764e-01 -8.12140882e-01
-1.14321506e+00 -3.15823138e-01 -4.01147395e-01 4.86933470e-01
2.60336548e-01 -1.70984790e-01 -1.21263111e+00 -1.82746127e-01
-3.99720967e-02 -4.30296749e-01 1.00497031e+00 2.65107453e-01
1.29442811e+00 -2.69299001e-01 -6.49675488e-01 4.62638050e-01
1.31914401e+00 1.82601705e-01 2.31645733e-01 2.16271907e-01
7.30017662e-01 5.98284066e-01 8.51328671e-01 1.56848684e-01
1.33745268e-01 5.91413081e-01 7.59314239e-01 -3.27909082e-01
-1.49857089e-01 -2.38417014e-01 5.10557353e-01 1.16857111e-01
1.37731194e-01 3.02897058e-02 -7.77831495e-01 7.42065132e-01
-1.69244587e+00 -6.12629712e-01 1.68275699e-01 2.13184309e+00
8.82725179e-01 3.72982502e-01 1.16602303e-02 2.03605015e-02
8.75853419e-01 2.20750943e-01 -7.66158640e-01 3.66817266e-01
-7.78781921e-02 1.86834320e-01 6.68391645e-01 2.38322958e-01
-1.32955337e+00 1.19099510e+00 5.52980900e+00 8.22415173e-01
-1.31397736e+00 1.82997152e-01 7.88507104e-01 1.51979476e-01
1.73400939e-02 9.16544795e-02 -1.16383624e+00 5.55658281e-01
3.63064438e-01 2.20507190e-01 2.16035098e-01 1.36043298e+00
-1.21295109e-01 -2.87488610e-01 -1.28180635e+00 7.65131176e-01
2.78290063e-01 -1.31038845e+00 4.32515629e-02 -2.46253565e-01
6.49579167e-01 -6.73846826e-02 -3.33636720e-03 1.10062957e-01
2.82313973e-01 -8.78027797e-01 1.14856839e+00 4.58563119e-01
4.79105949e-01 -3.95713806e-01 6.29652262e-01 5.34453869e-01
-1.15888464e+00 1.82575993e-02 -5.62645972e-01 3.58310878e-01
-2.58844107e-01 7.34557867e-01 -1.16914380e+00 2.19581351e-01
7.22995520e-01 6.35792553e-01 -8.47713530e-01 9.49959159e-01
-5.96071005e-01 9.03125703e-01 -2.63781309e-01 -5.43698929e-02
2.06596389e-01 8.15222599e-03 3.26199383e-01 1.12367427e+00
9.84337777e-02 -3.20402354e-01 4.55383748e-01 1.45562530e+00
6.41315058e-02 -7.95854107e-02 -3.96713376e-01 2.61512458e-01
5.02818823e-01 1.43585646e+00 -1.19569743e+00 -5.40317714e-01
-1.51705608e-01 9.76976514e-01 3.78791422e-01 4.73730057e-01
-1.04229164e+00 -2.74938315e-01 1.77726641e-01 2.27775648e-01
4.75226551e-01 7.27973580e-02 -8.16973820e-02 -1.07667089e+00
4.68925312e-02 -6.05904698e-01 1.28732085e-01 -6.33410871e-01
-1.02041113e+00 6.50453866e-01 2.23054155e-03 -1.23795831e+00
7.71726370e-02 -5.62516451e-01 -5.57939291e-01 7.28105962e-01
-1.60802245e+00 -1.34831548e+00 -7.63117313e-01 5.56386650e-01
6.56838000e-01 -2.88374424e-01 5.76573610e-01 2.79604066e-02
-5.35088480e-01 2.08060384e-01 -2.32541904e-01 3.44003886e-01
5.36144853e-01 -1.31801784e+00 -5.28935753e-02 1.03284490e+00
5.19140482e-01 4.89817619e-01 4.00470197e-01 -4.11807954e-01
-1.12080181e+00 -1.51714540e+00 3.43493283e-01 -5.26681840e-01
3.42699468e-01 -7.49094367e-01 -8.37426603e-01 6.54568851e-01
-9.58980173e-02 6.49271667e-01 2.58034587e-01 -2.83667177e-01
-3.86604130e-01 -5.52323163e-01 -1.11972964e+00 2.80895025e-01
7.74602413e-01 -5.55352867e-01 -5.06946623e-01 2.88438708e-01
4.99687046e-01 -2.22482175e-01 -3.40411872e-01 3.82958114e-01
8.53107125e-02 -8.80495191e-01 1.07362866e+00 -1.66935906e-01
2.62960911e-01 -7.80455351e-01 -1.52991176e-01 -1.07203197e+00
-2.29305789e-01 -4.57192361e-02 -1.07317314e-01 1.38485694e+00
2.53175527e-01 -5.63251913e-01 8.61686587e-01 1.06318884e-01
2.09457837e-02 -5.16872108e-01 -8.92260611e-01 -7.96238899e-01
-5.60277104e-01 -3.18236947e-01 5.11404693e-01 7.90711224e-01
-5.68652153e-01 3.72917980e-01 -2.31345184e-02 4.45128113e-01
1.04338813e+00 3.94036114e-01 8.36376369e-01 -1.33446872e+00
-1.64836153e-01 -2.46755183e-01 -4.81775820e-01 -1.10488641e+00
3.33084375e-01 -1.01634562e+00 5.15052974e-01 -1.41303205e+00
3.54765654e-01 -8.48748863e-01 -6.48394048e-01 7.62385845e-01
-1.85784787e-01 5.12026608e-01 -1.69913576e-03 3.26067865e-01
-8.87816966e-01 5.47002733e-01 9.63827670e-01 -5.58615863e-01
7.20104426e-02 1.20390825e-01 -6.06610715e-01 8.08103859e-01
7.49653041e-01 -6.28154516e-01 -5.00080407e-01 -2.30978683e-01
-4.57290351e-01 -2.73120135e-01 8.80669296e-01 -1.22079849e+00
1.36979535e-01 -2.36278713e-01 6.55408323e-01 -8.08343709e-01
2.77853787e-01 -9.62444186e-01 -1.68077335e-01 3.87657940e-01
-4.52103615e-01 -5.89866400e-01 6.18736371e-02 6.33873343e-01
-2.90973723e-01 -2.10278228e-01 1.09015393e+00 -1.15615770e-01
-8.64878893e-01 9.96605307e-02 -3.05642992e-01 -1.19855039e-01
1.43422461e+00 -1.13868624e-01 -1.99247554e-01 7.98421651e-02
-5.47205389e-01 7.12201698e-04 3.89203966e-01 2.40308568e-01
6.87569201e-01 -1.24441338e+00 -3.94801080e-01 4.31443214e-01
4.66876358e-01 4.73835111e-01 -2.69078404e-01 6.53051496e-01
-3.07208955e-01 3.09045047e-01 -2.49380559e-01 -1.23523426e+00
-1.17782056e+00 5.24864554e-01 3.98904502e-01 -4.90861945e-02
-5.82451403e-01 9.80414569e-01 7.21984327e-01 -2.02115610e-01
4.57203180e-01 -7.35608399e-01 -8.77187178e-02 3.05133276e-02
4.39922720e-01 3.14986035e-02 -1.34745613e-01 -7.27893233e-01
-5.77670932e-01 4.07951713e-01 -7.28052435e-03 1.07695043e-01
1.13922763e+00 1.46876678e-01 -1.92566797e-01 6.11970365e-01
9.56730723e-01 -7.22377077e-02 -1.60999882e+00 -6.14524126e-01
3.09957355e-01 -5.08751512e-01 9.81972218e-02 -7.73876786e-01
-1.22450531e+00 8.68062794e-01 8.36396813e-01 -3.04310670e-04
1.10078120e+00 3.97438556e-01 3.02530497e-01 1.92366287e-01
6.43239915e-01 -1.00402129e+00 5.05048931e-01 1.38943806e-01
6.89942122e-01 -1.29132760e+00 -2.75248319e-01 -4.68078166e-01
-6.44447386e-01 8.87097001e-01 9.41193640e-01 -3.55834812e-01
2.71168202e-01 2.71542549e-01 1.29438758e-01 -1.25877365e-01
-4.17632550e-01 -3.44567955e-01 3.52468848e-01 6.81717932e-01
6.88465834e-02 -2.17658766e-02 3.04466963e-01 4.21677083e-01
3.05137515e-01 -1.13967143e-01 1.07069999e-01 6.97150230e-01
-6.53660893e-01 -7.70022869e-01 -6.28996611e-01 3.42755258e-01
-2.85973877e-01 1.00098953e-01 -4.98437911e-01 4.64737952e-01
3.18281144e-01 7.90237427e-01 6.22196943e-02 -9.46154073e-02
-2.12805513e-02 -3.37304175e-02 4.39904392e-01 -7.88597047e-01
-3.43960851e-01 1.91124767e-01 -2.87962019e-01 -6.65497839e-01
-6.86994672e-01 -4.99746412e-01 -1.20867646e+00 5.89175284e-01
-5.40924668e-01 5.12589514e-02 4.63104576e-01 8.29918861e-01
2.42971350e-02 6.41349673e-01 6.19867504e-01 -1.14453483e+00
-3.42729539e-01 -8.95928383e-01 -4.27516252e-01 3.78381699e-01
3.98732066e-01 -9.34720278e-01 -2.19353199e-01 1.34732381e-01] | [9.496408462524414, 0.8478989005088806] |
1d72f21a-1c7c-425c-bacb-05ae00a065a5 | approximate-adapt-anonymize-3a-a-framework | 2307.01875 | null | https://arxiv.org/abs/2307.01875v1 | https://arxiv.org/pdf/2307.01875v1.pdf | Approximate, Adapt, Anonymize (3A): a Framework for Privacy Preserving Training Data Release for Machine Learning | The availability of large amounts of informative data is crucial for successful machine learning. However, in domains with sensitive information, the release of high-utility data which protects the privacy of individuals has proven challenging. Despite progress in differential privacy and generative modeling for privacy-preserving data release in the literature, only a few approaches optimize for machine learning utility: most approaches only take into account statistical metrics on the data itself and fail to explicitly preserve the loss metrics of machine learning models that are to be subsequently trained on the generated data. In this paper, we introduce a data release framework, 3A (Approximate, Adapt, Anonymize), to maximize data utility for machine learning, while preserving differential privacy. We also describe a specific implementation of this framework that leverages mixture models to approximate, kernel-inducing points to adapt, and Gaussian differential privacy to anonymize a dataset, in order to ensure that the resulting data is both privacy-preserving and high utility. We present experimental evidence showing minimal discrepancy between performance metrics of models trained on real versus privatized datasets, when evaluated on held-out real data. We also compare our results with several privacy-preserving synthetic data generation models (such as differentially private generative adversarial networks), and report significant increases in classification performance metrics compared to state-of-the-art models. These favorable comparisons show that the presented framework is a promising direction of research, increasing the utility of low-risk synthetic data release for machine learning. | ['Matthew Howard', 'Olivia Choudhury', 'Weijie Xu', 'Tamas Madl'] | 2023-07-04 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation'] | ['medical', 'miscellaneous'] | [ 2.25865752e-01 4.24388051e-01 -2.31128529e-01 -6.05649531e-01
-1.01034999e+00 -1.04380810e+00 6.44567609e-01 3.14655751e-01
-4.80838925e-01 1.10141206e+00 1.65687680e-01 -1.23327419e-01
-1.35749847e-01 -1.04784107e+00 -9.06689942e-01 -7.89608777e-01
-1.21540383e-01 4.67955351e-01 -3.77162725e-01 9.74171758e-02
-2.81801492e-01 8.20972204e-01 -1.26034880e+00 4.30782914e-01
8.80706251e-01 8.38486612e-01 -9.22909677e-01 4.90362436e-01
4.04753417e-01 3.24545950e-01 -6.11118436e-01 -1.33681965e+00
9.36151385e-01 -3.73711258e-01 -6.57058239e-01 -3.63735378e-01
4.31482702e-01 -5.08419216e-01 -3.70279819e-01 9.48246539e-01
6.94646299e-01 -1.40997648e-01 6.82963669e-01 -1.72124732e+00
-8.49356115e-01 5.02480626e-01 4.90896264e-03 -4.02100593e-01
2.23617405e-01 2.66526371e-01 7.10468769e-01 -1.25522569e-01
7.50510752e-01 8.73818278e-01 9.36758220e-01 8.27100515e-01
-1.74210656e+00 -7.35556126e-01 -5.67551255e-01 -4.00560230e-01
-1.45060313e+00 -5.26499271e-01 5.17131090e-01 -3.14401478e-01
3.97454053e-01 8.97871315e-01 6.25046372e-01 1.38896179e+00
2.04310447e-01 7.35599101e-01 1.24838722e+00 -5.26677594e-02
5.49300969e-01 9.29015100e-01 -1.61656156e-01 2.22424090e-01
7.61572838e-01 3.90065640e-01 -4.13884550e-01 -8.83037448e-01
3.14734757e-01 2.05962166e-01 -2.04586908e-01 -1.05252671e+00
-4.94172812e-01 1.01187575e+00 2.33173460e-01 -1.98966730e-02
-3.01805317e-01 -8.45280513e-02 3.20094049e-01 4.48494047e-01
6.58159912e-01 4.55720663e-01 -2.17453986e-01 1.57164574e-01
-1.17050803e+00 8.88990402e-01 1.03389394e+00 1.10471177e+00
6.83483303e-01 -2.40301996e-01 -4.68250692e-01 3.66678566e-01
-2.23105773e-02 2.79079020e-01 3.81046176e-01 -7.87490249e-01
5.59550166e-01 6.12186551e-01 1.55747190e-01 -1.01245368e+00
1.95410833e-01 -3.47989708e-01 -8.95583630e-01 3.18911523e-01
3.55136663e-01 -3.84066463e-01 -3.61349463e-01 1.96164727e+00
4.41250235e-01 -2.94049352e-01 5.49470723e-01 3.04151893e-01
2.09741279e-01 3.38595122e-01 2.34848157e-01 2.43247747e-02
9.25079525e-01 -2.52039641e-01 -5.44385076e-01 2.89185494e-01
7.69393444e-01 -2.26962149e-01 8.59178007e-01 1.36645243e-01
-1.18082035e+00 -5.59910797e-02 -9.93691206e-01 -1.92604750e-01
-6.05538666e-01 -3.04563761e-01 7.77782202e-01 1.55014634e+00
-1.20476341e+00 7.68752456e-01 -6.79303765e-01 -3.73479962e-01
1.14176714e+00 6.89851940e-01 -9.09740686e-01 -4.43043374e-02
-1.29656267e+00 3.78204197e-01 3.76606017e-01 -4.63422716e-01
-8.47866774e-01 -1.16649437e+00 -7.44979143e-01 9.40907449e-02
-2.51049966e-01 -1.03198361e+00 7.64987648e-01 -8.98481369e-01
-9.46480632e-01 1.21531940e+00 3.65441233e-01 -1.21235859e+00
1.30935085e+00 -3.13364826e-02 -2.89440036e-01 -1.11502640e-01
-1.90822482e-01 6.22872233e-01 6.40668213e-01 -1.28678656e+00
-6.03237212e-01 -6.80833161e-01 -1.18544243e-01 -3.54043879e-02
-6.57311916e-01 -9.26195905e-02 2.69844588e-02 -7.12226450e-01
-5.78168631e-01 -8.45127225e-01 -3.12942237e-01 3.78140599e-01
-5.85148752e-01 4.01470512e-01 9.40042555e-01 -8.81894350e-01
9.94773149e-01 -2.10687327e+00 -3.10659468e-01 3.53353411e-01
5.73403686e-02 3.38107854e-01 2.16030359e-01 4.57268089e-01
-4.52227481e-02 5.38140535e-01 -7.28786767e-01 -7.99658477e-01
2.30291501e-01 1.83123022e-01 -5.01197636e-01 7.86264658e-01
5.66610508e-02 1.03834784e+00 -4.02199715e-01 -1.77255630e-01
5.05415443e-03 6.64744437e-01 -9.23129141e-01 2.94069260e-01
-2.74690036e-02 3.74117494e-01 -4.46203858e-01 5.27984381e-01
1.10379899e+00 3.97969693e-01 2.66278461e-02 3.42348337e-01
2.57404149e-01 -1.55175552e-01 -1.06135237e+00 1.48527777e+00
-6.64517283e-02 2.80817837e-01 1.31889641e-01 -6.11463964e-01
7.74182856e-01 2.91007102e-01 6.77496672e-01 -3.28380436e-01
-3.93023379e-02 -4.82202359e-02 -4.66093093e-01 -1.72748104e-01
4.65370476e-01 -1.10742077e-01 -3.83731723e-01 6.24850214e-01
-8.80109519e-02 7.48862922e-02 -4.31034595e-01 1.78771108e-01
9.08868670e-01 -1.19011700e-02 3.36516470e-01 -2.68824041e-01
2.37170249e-01 -2.06084058e-01 4.90159482e-01 7.23343790e-01
-2.80535817e-01 7.91000903e-01 6.52247012e-01 -3.82211268e-01
-1.33713639e+00 -1.05630970e+00 -2.65743732e-01 6.38431787e-01
-3.40189278e-01 -1.48162872e-01 -1.31149530e+00 -9.35671568e-01
6.21892452e-01 1.11409903e+00 -8.48735511e-01 -4.48931545e-01
-1.41964555e-01 -1.04879379e+00 1.27055776e+00 7.92733505e-02
4.30042475e-01 -6.20313287e-01 -2.99643159e-01 -8.62184763e-02
1.89822093e-01 -6.85864627e-01 -4.65028524e-01 -1.62686288e-01
-8.50482285e-01 -9.21057463e-01 -4.93380189e-01 -1.60982445e-01
8.05772424e-01 -2.87027299e-01 6.98070943e-01 -3.10985863e-01
-2.83828825e-01 5.08328915e-01 -2.68329717e-02 -5.46228945e-01
-7.39160657e-01 1.97516114e-01 5.68149537e-02 4.93195981e-01
3.79162967e-01 -8.15410733e-01 -5.29049933e-01 -2.89382972e-02
-1.42474198e+00 -3.44766915e-01 3.89045000e-01 5.50191522e-01
6.59946680e-01 1.80553775e-02 2.69058734e-01 -1.41338515e+00
7.28962898e-01 -6.85311496e-01 -4.86197710e-01 2.65236259e-01
-9.84920561e-01 2.07437083e-01 7.88568437e-01 -3.68604213e-01
-1.01196063e+00 2.02493623e-01 -1.03787266e-01 -4.51643378e-01
-2.79173106e-01 -2.25468889e-01 -7.30453134e-01 -3.60674381e-01
7.62221813e-01 3.90163839e-01 3.82811040e-01 -5.43328941e-01
6.31179750e-01 7.67434955e-01 5.83412886e-01 -6.09959483e-01
8.75713885e-01 7.94825554e-01 2.19778880e-01 -3.65814775e-01
-1.71911076e-01 -2.38468442e-02 -5.10576189e-01 4.13185358e-01
5.26094854e-01 -7.81984568e-01 -5.87824464e-01 5.15815020e-01
-6.72081828e-01 9.84891877e-02 -9.40195441e-01 1.73028618e-01
-7.80233681e-01 3.06446850e-01 -2.47452408e-01 -9.35897052e-01
-6.98445201e-01 -8.47935975e-01 8.94010186e-01 1.22165903e-01
-2.33243421e-01 -1.13431489e+00 1.97512209e-01 5.57992339e-01
5.53263009e-01 1.00163090e+00 6.39774382e-01 -1.33337450e+00
-6.50840342e-01 -8.93988311e-01 2.91675091e-01 5.93759596e-01
6.59507960e-02 -2.09792495e-01 -1.26313317e+00 -5.38740635e-01
2.02689499e-01 -2.76200384e-01 5.46915114e-01 1.41346753e-01
1.24526191e+00 -1.14582551e+00 -3.62824351e-01 1.16034710e+00
1.41441977e+00 -1.39687121e-01 9.51672971e-01 1.05395563e-01
4.83812571e-01 8.35661411e-01 2.14594573e-01 6.61827505e-01
3.23434055e-01 6.09831154e-01 4.10396725e-01 -1.59912467e-01
3.17581236e-01 -8.63700449e-01 2.95922607e-01 -2.74926215e-01
2.68021613e-01 -3.41851503e-01 -3.62204075e-01 5.79267740e-01
-1.68931746e+00 -1.14297056e+00 1.26358137e-01 2.76592326e+00
1.01143479e+00 -2.60425836e-01 3.86288911e-01 -1.71428412e-01
4.80220139e-01 -2.64303572e-02 -5.57108104e-01 -6.48537338e-01
-3.70764196e-01 3.10383707e-01 9.83538032e-01 3.62377763e-01
-1.17369926e+00 6.08234644e-01 6.54560661e+00 7.75490820e-01
-8.34424496e-01 1.79356575e-01 1.05376911e+00 -5.19629121e-01
-7.13021517e-01 1.08138546e-01 -7.72305369e-01 6.66740716e-01
1.22888517e+00 -7.08553493e-01 4.45604801e-01 1.13551629e+00
-3.89361084e-02 4.64386195e-01 -1.38389432e+00 6.81702971e-01
9.40108448e-02 -1.26587582e+00 2.26657912e-01 6.78931773e-01
7.77370989e-01 -4.57715541e-01 4.93134439e-01 8.46688300e-02
6.84426069e-01 -1.16771019e+00 4.69236612e-01 8.49867702e-01
7.96903610e-01 -1.18048656e+00 6.37017250e-01 4.70910072e-01
-4.35212582e-01 1.12966388e-01 -4.39632803e-01 4.41251278e-01
-1.29435286e-01 6.28195643e-01 -9.74518955e-01 7.54353881e-01
5.96299887e-01 3.12877774e-01 -7.00849831e-01 6.98511958e-01
1.05473541e-01 4.81655508e-01 -5.12308061e-01 1.43936709e-01
-1.75281674e-01 -2.55853951e-01 5.91335058e-01 1.11031151e+00
3.85797769e-01 -2.32485935e-01 -3.28416586e-01 1.25610864e+00
-5.04432857e-01 2.02839166e-01 -1.09513152e+00 -6.72033653e-02
4.70466584e-01 9.86820757e-01 1.54329717e-01 9.50520113e-02
-6.67695329e-02 1.25203753e+00 2.64867127e-01 1.30398184e-01
-6.53449893e-01 -1.78659096e-01 1.25033033e+00 3.99875849e-01
7.80593511e-03 2.96996981e-01 -5.15926838e-01 -9.88576710e-01
3.18354428e-01 -8.95131707e-01 7.66811788e-01 -1.79624408e-01
-1.47755623e+00 3.22687089e-01 -6.07537031e-02 -1.18809247e+00
-5.81368566e-01 -1.11362353e-01 -3.09787124e-01 1.16938818e+00
-1.08673215e+00 -1.54852915e+00 3.28902841e-01 6.56624615e-01
-3.81657481e-01 -3.48205328e-01 1.09812939e+00 9.51699317e-02
-3.86517644e-01 1.39762473e+00 6.46053016e-01 2.72071436e-02
6.08056724e-01 -1.14286411e+00 6.47087514e-01 9.33820367e-01
-1.38112092e-02 9.39667583e-01 5.71060061e-01 -6.50848985e-01
-1.38702679e+00 -1.49658144e+00 8.82629395e-01 -9.06023383e-01
-5.34300543e-02 -7.94397473e-01 -9.07578290e-01 9.51500714e-01
-2.48034000e-02 2.54380018e-01 1.16763544e+00 -3.16285372e-01
-4.34037536e-01 -4.11679506e-01 -2.30300641e+00 3.48991305e-01
8.42726886e-01 -5.59529424e-01 -2.80938353e-02 2.27663577e-01
7.94497013e-01 -2.36852333e-01 -1.08262479e+00 5.41459187e-04
6.37347460e-01 -1.22617364e+00 1.04931498e+00 -1.11942327e+00
4.17213812e-02 2.92356964e-02 -2.88741857e-01 -1.05586553e+00
4.78494763e-02 -1.08664227e+00 -1.67202979e-01 1.82505155e+00
4.43116635e-01 -9.96649027e-01 1.29997635e+00 1.63345838e+00
5.91084957e-01 -4.75093901e-01 -1.06467617e+00 -7.43959248e-01
4.25179482e-01 -1.08049028e-01 1.15623856e+00 1.18998384e+00
-3.48959923e-01 -5.67330837e-01 -7.45835423e-01 2.29072645e-01
9.94342566e-01 -1.90108910e-01 1.15287197e+00 -1.16351080e+00
-2.18241766e-01 -1.43548831e-01 -7.15514183e-01 -1.67768687e-01
2.14658514e-01 -1.15985310e+00 -7.21301198e-01 -8.90875161e-01
2.68868208e-01 -5.69253683e-01 -3.41206565e-02 5.62883258e-01
9.56214778e-03 8.55163112e-02 -4.08957303e-02 1.26032203e-01
7.65282810e-02 6.29037201e-01 5.91551542e-01 2.03376301e-02
-2.26261050e-01 4.96999472e-01 -1.31455100e+00 7.28141516e-02
8.14607561e-01 -8.13594759e-01 -5.15606642e-01 1.32378072e-01
-1.77333713e-01 -2.61178523e-01 5.88501930e-01 -1.02003145e+00
2.31383815e-01 -1.22235760e-01 4.94904816e-01 -4.30580368e-03
2.00179428e-01 -1.15902340e+00 8.98275316e-01 4.92885083e-01
-6.42259836e-01 -1.77696049e-01 1.08187445e-01 5.68582177e-01
7.70062059e-02 3.80645208e-02 8.88569415e-01 9.51905772e-02
3.25638358e-03 8.53473783e-01 2.95946687e-01 4.65249680e-02
1.50000441e+00 -2.14158729e-01 -1.68103173e-01 -4.26984251e-01
-4.93396848e-01 -4.54151854e-02 1.08098197e+00 2.02581286e-01
2.75626332e-01 -1.32945526e+00 -9.77035761e-01 5.48991680e-01
1.94560364e-01 -1.06651537e-01 2.89736062e-01 1.15687869e-01
-4.47217345e-01 2.70536989e-01 -3.04297447e-01 -1.27780035e-01
-1.33887911e+00 8.17078412e-01 3.89322191e-01 -3.54133457e-01
-4.43164617e-01 5.82807541e-01 -7.95063153e-02 -7.21862793e-01
1.46151304e-01 9.55360979e-02 3.97772789e-01 -1.10692851e-01
4.98088539e-01 4.37623620e-01 1.10223129e-01 -5.80103934e-01
-1.35843694e-01 -1.61705688e-01 -1.38470665e-01 -2.47510135e-01
1.28687942e+00 -2.77189370e-02 -1.24801509e-01 -3.09502929e-01
1.53892136e+00 3.02475363e-01 -1.38532078e+00 -3.95393297e-02
-3.41838807e-01 -7.62007654e-01 -3.24308634e-01 -6.33391440e-01
-1.01176786e+00 5.54765701e-01 7.25912809e-01 2.21884146e-01
1.17329717e+00 -3.15187901e-01 5.54195225e-01 2.07475387e-02
5.00406444e-01 -7.79688478e-01 -8.18384588e-01 -4.72705841e-01
7.75287092e-01 -9.24740613e-01 1.68515623e-01 -3.10298204e-01
-7.63982654e-01 4.55173939e-01 1.16205707e-01 -1.09526208e-02
6.04045331e-01 3.55207890e-01 -1.76751837e-01 3.26434791e-01
-5.67077935e-01 4.64333385e-01 7.49494284e-02 1.25030673e+00
-9.88316908e-02 2.66130209e-01 -1.86335936e-01 9.31483030e-01
-6.08803451e-01 8.31517503e-02 3.44155282e-01 8.92742276e-01
2.75200278e-01 -1.57638431e+00 -2.75097281e-01 4.56871361e-01
-8.18016231e-01 4.95285727e-02 -8.75025630e-01 9.02199626e-01
1.43628702e-01 5.32650888e-01 -1.39628902e-01 -2.75603801e-01
3.93989474e-01 2.46091932e-01 5.04538789e-02 -1.73520222e-01
-1.08356118e+00 -7.53904462e-01 -5.70691861e-02 -6.06572866e-01
1.08544454e-02 -1.10723603e+00 -6.81043923e-01 -8.66890371e-01
3.32553804e-01 1.38352871e-01 6.73987091e-01 3.68040681e-01
8.51347208e-01 -1.61134526e-01 8.71164322e-01 -6.90917224e-02
-9.30882812e-01 -3.53225440e-01 -9.13336098e-01 8.91857147e-01
2.79442221e-01 1.44951224e-01 -3.26379627e-01 1.18027903e-01] | [6.02170991897583, 6.908913612365723] |
1f38fa04-dd54-4a08-aebc-051bc660821e | a-machine-learning-pressure-emulator-for | 2306.13116 | null | https://arxiv.org/abs/2306.13116v1 | https://arxiv.org/pdf/2306.13116v1.pdf | A Machine Learning Pressure Emulator for Hydrogen Embrittlement | A recent alternative for hydrogen transportation as a mixture with natural gas is blending it into natural gas pipelines. However, hydrogen embrittlement of material is a major concern for scientists and gas installation designers to avoid process failures. In this paper, we propose a physics-informed machine learning model to predict the gas pressure on the pipes' inner wall. Despite its high-fidelity results, the current PDE-based simulators are time- and computationally-demanding. Using simulation data, we train an ML model to predict the pressure on the pipelines' inner walls, which is a first step for pipeline system surveillance. We found that the physics-based method outperformed the purely data-driven method and satisfy the physical constraints of the gas flow system. | ['Alberto Costa Nogueira Junior', 'Elie Alhajjar', 'João Lucas de Sousa Almeida', 'Minh Triet Chau'] | 2023-06-22 | null | null | null | null | ['physics-informed-machine-learning'] | ['graphs'] | [-4.00249988e-01 1.43206030e-01 2.00220987e-01 6.08806349e-02
-1.40280724e-01 -4.40923929e-01 5.34845412e-01 1.64626881e-01
1.74057763e-02 6.17350459e-01 -3.68449599e-01 -7.59232342e-01
4.12336551e-02 -1.20521438e+00 -8.26800287e-01 -1.02553093e+00
-2.17894554e-01 4.45525378e-01 5.29943287e-01 -1.43598437e-01
2.22105071e-01 9.69916523e-01 -1.00745118e+00 -1.92858398e-01
5.92973590e-01 8.05320561e-01 5.20219617e-02 8.07756901e-01
1.94114268e-01 8.97207856e-01 -2.57814556e-01 1.90227985e-01
3.78914326e-01 -3.22730869e-01 -8.28897357e-01 -2.89679259e-01
-5.61144948e-01 -4.79629010e-01 -1.12918901e+00 5.59052646e-01
1.01008803e-01 3.60853463e-01 8.62974286e-01 -1.47050798e+00
1.88846543e-01 3.82732838e-01 -3.59475434e-01 -1.11172929e-01
-2.70261522e-02 8.43664467e-01 4.25467342e-01 -4.37482119e-01
-6.11612992e-03 9.25380588e-01 8.57275784e-01 3.91488522e-01
-7.76684940e-01 -1.91848800e-01 -3.33223939e-01 -1.75819531e-01
-1.34779644e+00 6.39100969e-02 7.48687625e-01 -7.26038277e-01
1.18201470e+00 5.79059124e-01 1.11452675e+00 2.98795819e-01
6.47105575e-01 5.49767494e-01 8.47022831e-01 -6.92383274e-02
6.94721937e-01 2.18148887e-01 -2.75739253e-01 5.83532155e-01
1.58885702e-01 2.56063521e-01 -1.36065930e-01 -8.55490118e-02
1.09495199e+00 -1.99997470e-01 -1.81792557e-01 1.70398112e-02
-4.60817724e-01 6.59394264e-01 4.42627966e-01 1.12084337e-01
-3.92212719e-01 1.80062875e-01 2.78292418e-01 -3.87608975e-01
2.63247788e-01 6.78828180e-01 -4.98536617e-01 -1.87109381e-01
-9.56791520e-01 7.02575386e-01 1.34437716e+00 9.64700997e-01
6.32627189e-01 1.79993004e-01 -1.11779161e-01 -1.04499318e-01
7.99768686e-01 5.37399411e-01 -1.50378630e-01 -1.08001947e+00
3.06810644e-02 3.53911489e-01 6.04387760e-01 -6.24111176e-01
-5.63157201e-01 1.32207021e-01 -1.03541100e+00 3.53150934e-01
4.99983341e-01 -7.73912311e-01 -7.23297596e-01 7.46627331e-01
4.76440936e-01 2.59244680e-01 -5.61070330e-02 8.31350744e-01
6.80409610e-01 1.18249893e+00 2.41367593e-01 -2.47328177e-01
8.96032035e-01 -1.06092680e+00 -7.51006722e-01 1.42217919e-01
7.12332904e-01 -3.37493390e-01 4.18199867e-01 2.55309016e-01
-1.25928342e+00 -3.88060510e-01 -1.02282417e+00 1.75169110e-01
-3.63084793e-01 -4.55248743e-01 7.13552773e-01 6.66353226e-01
-7.81191289e-01 1.24284458e+00 -1.19679201e+00 6.28782436e-03
2.93935031e-01 1.27589583e-01 2.00567439e-01 4.66847122e-01
-1.20048678e+00 1.15639555e+00 8.41448456e-02 4.43070978e-01
-1.12640738e+00 -1.23679638e+00 -8.66085231e-01 6.46763295e-02
-5.24792187e-02 -4.21844870e-01 1.68006575e+00 7.46157587e-01
-1.73601973e+00 5.21701649e-02 -7.78107718e-02 -3.65833074e-01
8.64157021e-01 -2.27834448e-01 -4.46204981e-03 1.46832034e-01
-5.32373786e-01 5.77980578e-02 2.00448394e-01 -1.29363441e+00
-5.60177326e-01 3.42706084e-01 -1.06598593e-01 -5.02987988e-02
4.32943180e-02 -1.49151400e-01 -7.94796925e-03 2.15736791e-01
-1.52616858e-01 -9.58930969e-01 -7.59170055e-01 6.58687577e-02
-8.37460041e-01 -4.87731844e-01 1.03730774e+00 -8.30026150e-01
7.56791890e-01 -1.69934177e+00 -1.90296039e-01 1.57434359e-01
-2.40265295e-01 4.48891073e-01 8.85668933e-01 9.68104005e-01
4.91497219e-02 2.41190836e-01 -4.90345806e-01 -2.89284617e-01
9.84216761e-03 -2.09266152e-02 -2.05511466e-01 7.87385046e-01
5.06671309e-01 5.78277826e-01 -1.07687795e+00 -2.43894771e-01
6.57462478e-01 4.49992031e-01 -2.79419273e-01 7.99167216e-01
-2.13096321e-01 1.03645086e+00 -6.84805632e-01 3.34933519e-01
8.94139826e-01 9.69551727e-02 -1.78792760e-01 -3.97740722e-01
-7.41341949e-01 9.94677767e-02 -8.42499077e-01 1.21554923e+00
-4.18255925e-01 2.71178633e-01 1.95590124e-01 -9.20713007e-01
1.17059124e+00 5.22559643e-01 1.14436877e+00 -4.41136599e-01
1.29014671e-01 2.04310212e-02 -3.32353383e-01 -1.06711030e+00
4.52160388e-01 -4.56394792e-01 8.84281695e-02 -4.34834920e-02
-5.43461084e-01 -1.22275686e+00 -9.60711017e-02 -3.28393914e-02
1.24580181e+00 3.08795631e-01 -3.10758591e-01 -6.51170790e-01
3.36007625e-01 5.55303156e-01 5.21192372e-01 2.86739856e-01
-5.08303903e-02 3.88673306e-01 4.06177551e-01 -5.07209599e-01
-1.42136681e+00 -8.98207605e-01 -2.36607999e-01 -3.81131284e-02
5.83175182e-01 -2.31612161e-01 -8.75019670e-01 -3.20659280e-01
2.51774073e-01 9.83265340e-01 3.42954509e-02 -3.07232499e-01
-8.55259359e-01 -9.58213210e-01 3.07925582e-01 7.20964372e-01
5.44468105e-01 -6.54178679e-01 -7.64226437e-01 3.52944613e-01
4.18013930e-01 -9.94288087e-01 8.68961588e-02 8.15973654e-02
-5.43721080e-01 -1.35367167e+00 -3.87459308e-01 -2.04896808e-01
4.73194838e-01 -1.27384543e-01 8.64352047e-01 2.35518157e-01
-4.84903127e-01 1.67048097e-01 -1.01343445e-01 -7.47748435e-01
-9.08929348e-01 -2.20916852e-01 -1.02606341e-02 -6.05229855e-01
2.55551771e-03 -1.15802363e-01 -9.76805449e-01 4.01083499e-01
-5.45864582e-01 3.17324460e-01 5.34162074e-02 2.40017638e-01
2.52307296e-01 1.10324574e+00 2.89588481e-01 -4.61156458e-01
1.53107360e-01 -7.28057861e-01 -8.83487105e-01 -1.03890426e-01
-4.70906854e-01 -2.48464480e-01 7.07620919e-01 -1.73503682e-01
-1.31928885e+00 4.88600165e-01 -4.86220270e-01 -3.12240005e-01
-2.86554128e-01 3.60966116e-01 -1.90266058e-01 -5.99503033e-02
1.10185146e-01 -1.86911359e-01 -1.50561377e-01 -6.63475513e-01
-1.07644998e-01 4.07179207e-01 4.98183489e-01 -5.53893864e-01
1.35787940e+00 4.25226003e-01 6.68231070e-01 -1.30581760e+00
-3.29565436e-01 -4.61523056e-01 -6.21188939e-01 -3.81563783e-01
1.03746891e+00 -7.11748004e-01 -1.24957228e+00 8.51991534e-01
-1.18110430e+00 -7.66306400e-01 -4.59910631e-01 4.01033103e-01
-3.52879286e-01 2.08318532e-01 -9.08982098e-01 -1.66385269e+00
-2.34817788e-01 -1.21514857e+00 8.51365328e-01 6.85005963e-01
5.34778088e-02 -1.11323833e+00 1.38420299e-01 -1.10001944e-01
6.66867673e-01 5.95185935e-01 6.87348783e-01 -1.73674509e-01
-8.40861380e-01 -2.30783269e-01 -1.72959954e-01 -4.48840708e-02
1.64208978e-01 6.17319643e-01 -1.17956638e+00 -3.73801082e-01
3.66891146e-01 1.49171323e-01 5.60097039e-01 4.13469434e-01
1.35991883e+00 -2.48046950e-01 -6.93999827e-01 5.27973354e-01
1.44169796e+00 5.61436236e-01 4.74122912e-01 2.39300400e-01
8.11882198e-01 7.93508112e-01 7.15531707e-01 7.67723024e-01
1.38685584e-01 1.66514367e-01 6.34228885e-01 -4.63321328e-01
4.02260423e-01 -3.40241671e-01 2.97460314e-02 5.67546964e-01
-1.26464650e-01 -3.71897608e-01 -1.10599935e+00 4.10678506e-01
-1.58902621e+00 -6.88751042e-01 -8.93779159e-01 2.11581635e+00
6.41608179e-01 1.79336593e-01 1.30862638e-01 2.75060665e-02
4.15054560e-01 -4.05592769e-01 -6.46630704e-01 -4.56875235e-01
4.56891119e-01 1.81253739e-02 1.01474440e+00 1.91883922e-01
-1.24017096e+00 2.82186270e-01 7.62661743e+00 3.59883100e-01
-9.98242617e-01 -1.40091136e-01 6.16362214e-01 4.08552766e-01
-1.72224775e-01 2.68602520e-01 -9.55712795e-01 4.07587230e-01
1.38420832e+00 -3.15022826e-01 5.72301447e-02 4.99022245e-01
7.98261404e-01 -4.08029497e-01 -1.18086851e+00 1.80973738e-01
-8.75307083e-01 -1.50816524e+00 -7.68447161e-01 -9.74537432e-02
4.08373266e-01 1.63718201e-02 -5.65970063e-01 1.72584683e-01
3.41044426e-01 -8.41759920e-01 6.44487560e-01 7.41487324e-01
4.43282485e-01 -4.40558076e-01 7.27123320e-01 5.28645575e-01
-1.30312359e+00 -1.90163311e-02 -8.98935795e-02 -2.41168782e-01
8.26536417e-01 7.20330596e-01 -1.01701391e+00 6.60442472e-01
8.92144382e-01 2.39612579e-01 -8.84592235e-02 1.40510702e+00
9.21032801e-02 7.25058734e-01 -6.90037251e-01 -1.28164381e-01
1.34834096e-01 -3.20120126e-01 2.70623118e-01 1.14463246e+00
3.03213060e-01 -5.59446327e-02 3.12756598e-01 1.31391203e+00
5.59657276e-01 -4.69916284e-01 -7.40011871e-01 -1.08882852e-01
2.76640266e-01 1.21839094e+00 -5.08455873e-01 1.47619054e-01
-4.48810935e-01 4.53045487e-01 -3.79019648e-01 4.48619634e-01
-1.30565000e+00 -4.17215496e-01 6.24428511e-01 4.49442476e-01
-2.07109660e-01 -6.17588818e-01 -1.86491609e-01 -3.70387226e-01
-3.94546449e-01 3.17099869e-01 -8.90108570e-02 -5.42787731e-01
-1.25770319e+00 9.78603736e-02 5.10481298e-01 -8.98160756e-01
-5.12721501e-02 -8.19167733e-01 -1.26437616e+00 1.02048337e+00
-1.73332858e+00 -9.86758173e-01 -3.80971819e-01 -6.17945194e-02
4.53033745e-01 1.66242421e-01 6.65954471e-01 2.07753435e-01
-1.07787669e+00 -5.94832599e-02 2.47321323e-01 6.95511401e-02
6.53865933e-02 -1.33519685e+00 7.97117710e-01 6.08314335e-01
-1.01109660e+00 2.88438261e-01 1.29314339e+00 -1.06931603e+00
-2.26547647e+00 -1.22791767e+00 3.04373622e-01 -8.90598148e-02
8.51994991e-01 -2.02117592e-01 -1.22692823e+00 3.88373733e-01
3.53855997e-01 1.00690745e-01 1.70799181e-01 -5.19015193e-01
6.97057843e-01 2.07399815e-01 -1.45960259e+00 1.84618592e-01
3.89123201e-01 -3.05204362e-01 -7.00592846e-02 8.15042973e-01
7.42338717e-01 -7.16945827e-01 -1.28662527e+00 6.58506513e-01
1.88641980e-01 -4.26930517e-01 8.16964805e-01 -4.90934163e-01
2.83915430e-01 -2.58000463e-01 4.51805025e-01 -1.05062366e+00
-1.44972563e-01 -9.56417561e-01 -4.51849639e-01 1.34333825e+00
4.88042533e-01 -3.57460320e-01 1.03702164e+00 1.63124263e+00
-4.02601004e-01 -8.21367443e-01 -8.85034680e-01 -8.17141294e-01
5.80759943e-01 -1.95038766e-01 5.94790161e-01 7.19893992e-01
2.13863119e-01 -2.65522271e-01 -1.94287285e-01 7.69144237e-01
6.01319492e-01 -2.73029715e-01 5.65162480e-01 -8.44933033e-01
-2.92602628e-01 -1.03769317e-01 1.05938233e-01 -7.30070591e-01
9.41576734e-02 -2.92374223e-01 6.53947651e-01 -1.73269618e+00
-1.18670367e-01 -9.04404104e-01 1.18144915e-01 1.23106577e-01
4.98754680e-02 -4.16651011e-01 -1.95274651e-01 -9.33354348e-02
2.81233341e-01 7.53853798e-01 1.18744302e+00 -3.35636497e-01
-3.50064784e-01 4.17736262e-01 3.00053984e-01 7.65236020e-01
1.12226593e+00 -3.71456683e-01 -1.12679958e-01 -1.51388109e-01
2.56060511e-02 2.23374367e-01 4.20320451e-01 -1.04094374e+00
6.22764707e-01 -5.58793426e-01 1.84693590e-01 -9.07862723e-01
4.39108908e-01 -1.03669393e+00 5.16779363e-01 9.51836228e-01
1.33828774e-01 -1.23628914e-01 5.33023655e-01 3.94138187e-01
2.15367019e-01 -5.16862333e-01 7.82843590e-01 -4.01098192e-01
-3.79924417e-01 4.95871782e-01 -8.09583187e-01 -3.45345348e-01
1.27547991e+00 9.12332907e-02 -5.18831909e-01 1.33006513e-01
-2.13767424e-01 5.12723148e-01 4.37634110e-01 -6.61615580e-02
3.92597675e-01 -6.64039910e-01 -6.65484667e-01 2.34280944e-01
-6.82515919e-01 8.20501149e-01 3.61850739e-01 5.24064183e-01
-1.26389468e+00 3.10452312e-01 2.43736073e-01 -6.03913605e-01
-4.83395576e-01 6.36540234e-01 7.94417918e-01 -1.01179838e-01
-9.56460238e-01 8.47974777e-01 4.05551344e-02 -3.69628817e-01
-8.31654668e-02 -3.48858714e-01 1.28440127e-01 -6.46033108e-01
3.48266542e-01 7.58916080e-01 4.57114121e-03 -2.75895536e-01
-6.58098683e-02 3.39069963e-01 6.43337071e-01 2.31151208e-01
1.51783121e+00 2.69614726e-01 -2.60417107e-02 5.28963506e-01
9.26142514e-01 -5.51158369e-01 -1.70430636e+00 3.11677694e-01
4.97209914e-02 -4.21929002e-01 3.61663461e-01 -2.35005811e-01
-1.21604764e+00 6.35639489e-01 6.82028234e-02 7.89985895e-01
7.98602939e-01 -1.09677978e-01 9.78915453e-01 2.74867505e-01
2.43215963e-01 -9.35441256e-01 -3.65093976e-01 2.29163930e-01
6.65580451e-01 -9.62717950e-01 1.55919105e-01 -8.75474632e-01
-2.71030277e-01 9.98267949e-01 9.56382930e-01 -1.22383997e-01
1.10542822e+00 9.31219220e-01 -3.65112722e-01 -1.93794474e-01
-4.46651608e-01 5.27528584e-01 -3.46363425e-01 5.59921026e-01
2.20557764e-01 -5.54352440e-02 3.13675374e-01 1.54015824e-01
-3.84694641e-03 -1.87415510e-01 4.95991260e-01 1.35167110e+00
-3.67929608e-01 -8.21862102e-01 -3.80659401e-01 2.17896163e-01
-8.22722539e-02 4.07804430e-01 2.10702315e-01 8.18010032e-01
-5.59592880e-02 1.22659492e+00 1.24804735e-01 -1.21427760e-01
7.58779049e-01 -1.73253015e-01 1.16323456e-01 -2.29197904e-01
-6.37524903e-01 -1.60076648e-01 1.55723616e-01 -4.63073939e-01
-2.52538174e-01 -6.20281398e-01 -1.83707809e+00 -4.57521886e-01
-4.58482742e-01 4.56427336e-01 1.13112783e+00 1.10641563e+00
-1.63770854e-01 7.62743771e-01 1.05188668e+00 -1.14956009e+00
-5.12264967e-01 -7.56771207e-01 -1.12353861e+00 2.83652306e-01
3.35506082e-01 -6.41732037e-01 -5.87228954e-01 -1.16807714e-01] | [6.369542121887207, 3.32501482963562] |
c061215e-757e-4f47-a527-a3c158f0b458 | multiple-reflection-symmetry-detection-via | 1704.06392 | null | http://arxiv.org/abs/1704.06392v1 | http://arxiv.org/pdf/1704.06392v1.pdf | Multiple Reflection Symmetry Detection via Linear-Directional Kernel Density Estimation | Symmetry is an important composition feature by investigating similar sides
inside an image plane. It has a crucial effect to recognize man-made or nature
objects within the universe. Recent symmetry detection approaches used a
smoothing kernel over different voting maps in the polar coordinate system to
detect symmetry peaks, which split the regions of symmetry axis candidates in
inefficient way. We propose a reliable voting representation based on weighted
linear-directional kernel density estimation, to detect multiple symmetries
over challenging real-world and synthetic images. Experimental evaluation on
two public datasets demonstrates the superior performance of the proposed
algorithm to detect global symmetry axes respect to the major image shapes. | ['Christophe Ducottet', 'Philippe Colantoni', 'Olivier Alata', 'Cecile Barat', 'Mohamed Elawady'] | 2017-04-21 | null | null | null | null | ['symmetry-detection'] | ['computer-vision'] | [ 7.51319081e-02 -9.55526307e-02 -1.31277218e-01 -3.95066082e-01
-3.95782471e-01 -6.93494558e-01 1.13437188e+00 -4.49012190e-01
-9.38863978e-02 2.87679523e-01 3.70778114e-01 -6.10284023e-02
-4.50472891e-01 -9.23127234e-01 -2.14385018e-01 -7.67240942e-01
-6.76082820e-02 7.85975158e-01 7.17275798e-01 2.57806629e-02
1.04535651e+00 8.69728565e-01 -1.43442798e+00 3.14019978e-01
5.85061193e-01 9.85837996e-01 -2.95720786e-01 4.77782130e-01
4.43903282e-02 3.49914908e-01 -2.92757869e-01 -5.47042131e-01
5.34580529e-01 -2.25271508e-01 -7.54399776e-01 2.25775480e-01
4.13164884e-01 2.03027785e-01 3.80096980e-03 1.30730677e+00
1.16516367e-01 4.91564721e-03 1.39923120e+00 -9.63807166e-01
-5.17505884e-01 6.21286891e-02 -9.87820506e-01 2.15867117e-01
1.63935184e-01 -4.78877544e-01 9.68513489e-01 -1.02118981e+00
9.88672435e-01 8.82877111e-01 6.41542137e-01 -5.95227554e-02
-7.32706070e-01 -5.96713901e-01 -4.62124646e-01 4.74055588e-01
-1.46415949e+00 -2.33852670e-01 1.02827132e+00 -6.64644659e-01
8.28125477e-01 4.48636323e-01 3.76335710e-01 3.92735451e-01
5.17994046e-01 2.47119412e-01 1.14132130e+00 -4.74284649e-01
9.80067402e-02 -1.88718751e-01 -1.61970034e-01 7.83827841e-01
2.67709255e-01 -1.93730101e-01 -3.91691953e-01 -5.81681669e-01
1.03658342e+00 -2.06313834e-01 -2.69529849e-01 -9.52922165e-01
-1.43023360e+00 8.23794961e-01 2.59229928e-01 1.79938138e-01
-2.05319211e-01 -6.93774581e-01 1.78473771e-01 3.58927157e-03
4.38820839e-01 6.51405573e-01 -1.95991695e-01 -1.13089427e-01
-8.93174291e-01 2.39270538e-01 5.00179231e-01 6.18525088e-01
5.27106941e-01 -3.29039425e-01 4.39667702e-03 8.27149332e-01
1.96288466e-01 4.06419218e-01 3.96220326e-01 -6.88998818e-01
3.76975209e-01 1.02750194e+00 -4.01480347e-02 -1.16162634e+00
-6.24993026e-01 -1.81440964e-01 -1.18099535e+00 4.06790167e-01
5.02300739e-01 4.63651061e-01 -7.26915121e-01 1.07661176e+00
5.25162816e-01 -1.52693093e-01 -1.51885077e-01 9.09925759e-01
7.89423108e-01 5.90911269e-01 -7.13720560e-01 1.70732543e-01
1.71020436e+00 -6.45921767e-01 -6.66246116e-02 2.36672446e-01
1.14566155e-01 -1.24396622e+00 5.53157151e-01 6.10258877e-01
-7.62501419e-01 -2.14619935e-01 -9.59954619e-01 5.99750765e-02
-2.22612232e-01 5.13865113e-01 7.24311531e-01 7.46973515e-01
-5.59409082e-01 3.75474602e-01 -5.91451764e-01 -1.49720386e-01
1.88593507e-01 1.97819900e-02 -7.20088542e-01 3.16634297e-01
-6.11764193e-01 6.61268413e-01 1.54362500e-01 -1.22797661e-01
1.16571531e-01 -4.79684383e-01 -9.70830500e-01 -4.01356407e-02
1.04795508e-01 -4.58683491e-01 9.13491011e-01 -6.03787303e-01
-1.39251852e+00 1.35818756e+00 -1.92478597e-01 -9.15071592e-02
8.07505906e-01 2.80386984e-01 -5.37173390e-01 2.73632675e-01
-6.46717325e-02 2.63475865e-01 1.12733781e+00 -9.16683018e-01
-5.02711356e-01 -5.88805139e-01 -5.52662790e-01 6.46097139e-02
2.85464913e-01 3.28361183e-01 -2.91974425e-01 -6.53991222e-01
1.16142237e+00 -9.01349783e-01 1.69879600e-01 -4.35118154e-02
-5.25533676e-01 -5.07896304e-01 1.07529998e+00 -7.56867468e-01
8.04392993e-01 -1.77913237e+00 -6.97629973e-02 8.15754712e-01
1.11404017e-01 -1.17066368e-01 3.41063887e-01 2.26328507e-01
-3.30728501e-01 -1.58893943e-01 -2.74044067e-01 3.96402657e-01
-1.18746288e-01 -4.19559270e-01 -4.99816149e-01 1.09181595e+00
1.10220894e-01 5.41983008e-01 -3.14527303e-01 -2.91205108e-01
2.33421102e-01 3.85340273e-01 -3.49766731e-01 1.51657201e-02
4.65476811e-01 2.57332981e-01 -1.18497185e-01 6.30773008e-01
1.20910943e+00 -1.50636539e-01 -7.95874447e-02 -4.23301250e-01
-5.07370591e-01 2.43002594e-01 -1.42768073e+00 9.80751991e-01
1.02229610e-01 5.24926245e-01 -2.85965323e-01 -9.58472967e-01
1.30497515e+00 -1.01128057e-01 2.58479387e-01 -7.77798653e-01
5.37073761e-02 2.14049723e-02 1.58796445e-01 -1.36650771e-01
5.81672490e-01 -2.11133555e-01 -1.08396403e-01 2.94349164e-01
3.48564014e-02 -5.40095687e-01 9.57012028e-02 -3.07714880e-01
5.40366471e-01 -7.65705928e-02 7.98214495e-01 -6.79506361e-01
9.33269680e-01 -3.50454926e-01 8.00762594e-01 4.55052137e-01
-1.63376823e-01 1.21351314e+00 7.70208061e-01 -1.08266795e+00
-1.51044583e+00 -1.20744181e+00 -6.42447710e-01 5.17228961e-01
2.83596069e-01 -3.60997111e-01 -5.86028576e-01 -4.53592867e-01
-1.17858797e-01 3.81458282e-01 -6.33230925e-01 -2.40134578e-02
-6.58707082e-01 -9.79191244e-01 2.71849036e-01 2.15137899e-01
8.37808788e-01 -8.05755377e-01 -8.08094680e-01 -6.24537766e-01
-7.76749253e-02 -8.05570960e-01 -3.90741915e-01 -2.22160935e-01
-5.37205160e-01 -1.49095368e+00 -1.04049075e+00 -7.65046358e-01
9.16706383e-01 4.23435092e-01 8.55571210e-01 -4.34243888e-01
-6.40743971e-01 4.26549651e-02 -2.20697299e-02 -8.15368593e-02
-9.00236517e-02 -3.44569564e-01 1.35968821e-02 1.71019584e-01
2.48455375e-01 -3.54314148e-01 -5.68119645e-01 8.65887523e-01
-4.21291620e-01 2.19236985e-01 3.38710636e-01 7.08917856e-01
4.83583093e-01 2.41929665e-01 -1.22589327e-01 -4.82274830e-01
3.36286992e-01 -2.21855298e-01 -1.07453418e+00 5.14544129e-01
-1.47945866e-01 3.52633417e-01 2.18723118e-01 -2.84438524e-02
-1.28041518e+00 8.64496641e-03 3.14368516e-01 -4.92268205e-02
4.88300957e-02 -1.30831050e-02 1.16333283e-01 -3.12668711e-01
4.29397285e-01 6.35985911e-01 -3.03034902e-01 -1.35084167e-01
1.20629802e-01 6.30018532e-01 6.75523818e-01 -6.60460889e-01
9.52400863e-01 9.96993005e-01 3.81702274e-01 -1.22506654e+00
-1.74546868e-01 -9.74724293e-01 -8.37605715e-01 -1.63090467e-01
7.90893435e-01 -6.21900380e-01 -7.02413678e-01 7.55176306e-01
-1.24156678e+00 6.44213438e-01 3.04125119e-02 6.84010267e-01
-7.16511965e-01 8.19919884e-01 -1.44771323e-01 -8.22553098e-01
-3.58266145e-01 -8.31702173e-01 1.41625261e+00 5.35732865e-01
-1.53055876e-01 -5.44010341e-01 5.42790711e-01 2.98467010e-01
1.98994204e-01 -6.87570870e-02 1.07928061e+00 -5.22322476e-01
-5.02886772e-01 -1.88386559e-01 -5.93314648e-01 -4.03689221e-02
9.76294950e-02 5.63662171e-01 -7.52995789e-01 8.07187706e-02
1.94379196e-01 -5.79001531e-02 9.85263228e-01 3.25904012e-01
1.08485293e+00 2.08472937e-01 -7.37751601e-03 1.00948882e+00
9.58684981e-01 1.26638874e-01 8.63456905e-01 4.33938116e-01
1.88926592e-01 6.70259297e-01 3.64819050e-01 5.78963757e-01
7.24957585e-02 6.43464506e-01 -1.31400274e-02 -3.05804349e-02
1.38193801e-01 -4.91630919e-02 4.89199199e-02 5.71331263e-01
-5.85212767e-01 1.96333259e-01 -1.06344426e+00 2.59939462e-01
-1.64860511e+00 -1.26666903e+00 -1.99429274e-01 2.48382330e+00
1.69342279e-01 2.95669436e-01 2.68478859e-02 2.75094032e-01
1.14557803e+00 2.89375991e-01 -1.68177009e-01 -4.35909122e-01
-5.01990736e-01 6.50592893e-02 7.02845275e-01 1.01755254e-01
-1.38613796e+00 3.98685426e-01 7.00517178e+00 9.63440895e-01
-1.31095386e+00 -2.94430614e-01 4.05855060e-01 5.60988784e-01
-2.01981202e-01 8.82585570e-02 -6.41877234e-01 1.72505334e-01
-7.09913075e-02 -3.06693524e-01 -1.09727375e-01 1.14349091e+00
-2.68427253e-01 -3.10641378e-01 -3.91461313e-01 1.29400742e+00
4.03329492e-01 -1.46317887e+00 1.00970790e-01 -6.40223846e-02
9.55993533e-01 -7.02958927e-02 1.26508757e-01 -2.63720423e-01
-7.71382591e-03 -8.10616612e-01 6.77883029e-01 4.93800849e-01
6.24635160e-01 -7.89669693e-01 5.66613674e-01 3.17594767e-01
-1.42530632e+00 1.90717101e-01 -6.61699355e-01 -1.44689977e-01
4.14870083e-02 5.64585507e-01 -1.32267773e+00 7.69477427e-01
5.93160391e-01 2.96002477e-01 -6.27246559e-01 1.38883984e+00
-5.08475065e-01 2.58287132e-01 -4.90596741e-01 1.14877917e-01
2.10258476e-02 -8.20773244e-01 5.79025865e-01 1.10398304e+00
5.49179196e-01 -1.64779887e-01 -2.62083352e-01 8.55759919e-01
2.73530096e-01 2.30360851e-01 -6.86251760e-01 4.17406291e-01
9.60330665e-03 1.32350397e+00 -1.60001945e+00 -4.15518433e-02
-4.66352433e-01 9.18108284e-01 1.06780507e-01 -1.66539401e-02
-7.22974598e-01 -5.94306409e-01 4.48215216e-01 2.50536650e-01
4.92700279e-01 -5.40839672e-01 -3.06380928e-01 -1.44156861e+00
2.53615648e-01 -3.86283785e-01 5.37830710e-01 -7.30598032e-01
-1.19943583e+00 3.53619069e-01 7.07044676e-02 -1.45250034e+00
1.21481098e-01 -1.04429710e+00 -1.19904006e+00 7.18941987e-01
-1.07375300e+00 -1.24501920e+00 -1.14960350e-01 5.05653083e-01
4.08333123e-01 -4.23372537e-01 6.42824888e-01 -2.79468745e-01
-1.51370734e-01 1.97853088e-01 3.49373311e-01 2.19461128e-01
6.78823590e-01 -1.22719896e+00 6.41116023e-01 9.47036505e-01
3.51314694e-01 5.78939319e-01 5.45752168e-01 -6.88162148e-01
-1.00214589e+00 -5.03144026e-01 1.14476418e+00 -2.74857789e-01
3.24434102e-01 -5.55461228e-01 -5.80990255e-01 3.03671598e-01
7.36205429e-02 3.79364900e-02 5.41058540e-01 8.99154618e-02
-8.67867947e-01 2.55344123e-01 -1.26872385e+00 3.80020142e-01
9.57616210e-01 -5.03525972e-01 -1.02994287e+00 2.14820892e-01
-2.76717037e-01 -2.28504092e-01 -4.76706952e-01 9.52156961e-01
8.21078360e-01 -1.48370540e+00 1.00180340e+00 -3.89479399e-01
2.29119033e-01 -7.26556242e-01 1.27164841e-01 -9.10673201e-01
-5.40958822e-01 -5.54309309e-01 4.75329012e-01 8.23178589e-01
5.64404465e-02 -6.36650145e-01 1.06897485e+00 2.20775589e-01
2.45228872e-01 -3.12799335e-01 -1.49257779e+00 -7.61472106e-01
-1.05073631e-01 -8.22119266e-02 5.26591837e-01 8.84447157e-01
1.01437375e-01 -8.43969658e-02 -1.91306606e-01 3.24059308e-01
7.16213107e-01 5.79363585e-01 9.42169189e-01 -1.46242559e+00
1.06418423e-01 -8.12750518e-01 -1.26632762e+00 -6.66922987e-01
9.96954367e-02 -6.88164473e-01 -3.91276777e-01 -9.51121628e-01
4.90390122e-01 -5.58323674e-02 -5.69606759e-02 -2.04274401e-01
3.02220970e-01 5.08701921e-01 -2.09989280e-01 2.78881907e-01
-5.53004086e-01 7.27954209e-01 1.06623089e+00 -1.20856375e-01
3.95689040e-01 2.90384948e-01 -1.21121861e-01 1.24548900e+00
7.21043050e-01 -1.81381688e-01 -1.63414001e-01 2.44500220e-01
6.01176739e-01 -4.08639222e-01 2.13772312e-01 -1.11620140e+00
3.45386565e-01 -2.57695198e-01 7.88811147e-01 -1.01482189e+00
1.63566284e-02 -5.16240239e-01 2.02186674e-01 4.28874135e-01
2.76914567e-01 1.88942373e-01 -2.45167658e-01 2.19852045e-01
-4.08369541e-01 -3.50127697e-01 1.11864126e+00 1.16612054e-01
-7.66916752e-01 -1.11762723e-02 -1.19330011e-01 -1.28143191e-01
1.28555524e+00 -4.25250590e-01 -3.02503556e-01 -1.98179483e-01
-2.23558873e-01 -2.25016147e-01 8.46997619e-01 4.57024425e-01
7.41852999e-01 -1.29091966e+00 -6.92288041e-01 8.24890018e-01
4.59319055e-01 -1.02916345e-01 3.18541616e-01 7.04603791e-01
-1.05895901e+00 4.86342877e-01 -5.94467103e-01 -7.38914192e-01
-1.69072497e+00 2.41405025e-01 4.38992798e-01 -1.85299158e-01
-5.89895010e-01 8.61376882e-01 5.15370250e-01 -7.37427354e-01
-2.36393273e-01 -6.44298911e-01 -4.86553192e-01 -2.55296752e-02
6.08413339e-01 5.97919762e-01 2.70831585e-01 -1.07799554e+00
-5.17852843e-01 1.30884457e+00 3.39553244e-02 -1.01166949e-01
1.12398899e+00 2.42276698e-01 -2.01150417e-01 9.48097631e-02
9.85037863e-01 5.92301250e-01 -8.92261386e-01 -1.69739887e-01
1.22330263e-01 -8.09788644e-01 -4.75769341e-01 -2.60881186e-01
-6.77559078e-01 5.01477301e-01 3.10599416e-01 2.37217695e-01
7.80249298e-01 3.02163363e-01 1.58390224e-01 3.79715592e-01
3.41688693e-01 -1.09557748e+00 -3.02264631e-01 5.94930649e-01
1.18540275e+00 -1.17923093e+00 3.61951083e-01 -7.78981388e-01
-6.13138258e-01 1.81521523e+00 3.82143140e-01 -4.15758848e-01
8.22160661e-01 -1.71930697e-02 -1.65635590e-02 -3.52262795e-01
-2.49718919e-01 1.47786632e-01 7.80429125e-01 5.52117348e-01
3.91553253e-01 9.00488570e-02 -5.30163646e-01 3.98682058e-01
-5.71892142e-01 -6.14172697e-01 3.98476005e-01 6.70049131e-01
-6.81274176e-01 -1.02580595e+00 -7.61084616e-01 2.00095594e-01
-2.61210442e-01 3.19167167e-01 -7.19657600e-01 5.68296969e-01
2.18950480e-01 2.89556175e-01 4.92142320e-01 -7.74576962e-02
2.38501862e-01 7.68574476e-02 7.07408130e-01 -4.20132130e-02
-7.30245635e-02 -1.88769996e-01 -1.45587206e-01 -4.12296891e-01
-4.26289856e-01 -8.24199080e-01 -1.09984660e+00 -6.68127462e-02
-1.93397135e-01 1.59005180e-01 7.42331266e-01 6.30169511e-01
2.31807888e-01 -2.05987692e-01 8.51051986e-01 -9.00465071e-01
-5.12297392e-01 -7.59585083e-01 -8.97616148e-01 6.13616288e-01
-3.87893356e-02 -1.01726806e+00 -5.71354687e-01 -1.54816568e-01] | [8.934125900268555, -2.0412662029266357] |
59f59af3-c756-4984-897e-7a35fd7837f3 | an-ai-ready-multiplex-staining-dataset-for | 2305.16465 | null | https://arxiv.org/abs/2305.16465v1 | https://arxiv.org/pdf/2305.16465v1.pdf | An AI-Ready Multiplex Staining Dataset for Reproducible and Accurate Characterization of Tumor Immune Microenvironment | We introduce a new AI-ready computational pathology dataset containing restained and co-registered digitized images from eight head-and-neck squamous cell carcinoma patients. Specifically, the same tumor sections were stained with the expensive multiplex immunofluorescence (mIF) assay first and then restained with cheaper multiplex immunohistochemistry (mIHC). This is a first public dataset that demonstrates the equivalence of these two staining methods which in turn allows several use cases; due to the equivalence, our cheaper mIHC staining protocol can offset the need for expensive mIF staining/scanning which requires highly-skilled lab technicians. As opposed to subjective and error-prone immune cell annotations from individual pathologists (disagreement > 50%) to drive SOTA deep learning approaches, this dataset provides objective immune and tumor cell annotations via mIF/mIHC restaining for more reproducible and accurate characterization of tumor immune microenvironment (e.g. for immunotherapy). We demonstrate the effectiveness of this dataset in three use cases: (1) IHC quantification of CD3/CD8 tumor-infiltrating lymphocytes via style transfer, (2) virtual translation of cheap mIHC stains to more expensive mIF stains, and (3) virtual tumor/immune cellular phenotyping on standard hematoxylin images. The dataset is available at \url{https://github.com/nadeemlab/DeepLIIF}. | ['Saad Nadeem', 'Christine H. Chung', 'Robbert JC Slebos', 'Janis V. de la Iglesia', 'Juan Hernandez-Prera', 'Joseph Marino', 'Parmida Ghahremani'] | 2023-05-25 | null | null | null | null | ['style-transfer'] | ['computer-vision'] | [-5.41018695e-02 2.14752229e-03 -4.81210113e-01 -2.13572998e-02
-1.28398144e+00 -7.30694830e-01 2.18899637e-01 3.18810284e-01
-6.64508879e-01 9.68710482e-01 1.35892332e-02 -6.76281631e-01
3.52425545e-01 -6.46271646e-01 -1.84848323e-01 -1.19379020e+00
2.29478016e-01 1.03784263e+00 7.93037564e-02 -1.21008284e-01
-1.75166413e-01 8.68173718e-01 -1.01955116e+00 2.34789476e-01
2.99592793e-01 8.68851781e-01 -1.21933986e-02 1.20959461e+00
-2.43751168e-01 7.85708666e-01 -2.00877234e-01 -1.26931190e-01
1.84537843e-01 -1.07773334e-01 -8.60892773e-01 -1.85252994e-01
4.08788115e-01 -4.11250383e-01 1.75016791e-01 7.54015446e-01
5.57354927e-01 -4.74552900e-01 7.16795385e-01 -1.37873220e+00
-2.76283413e-01 -2.93074325e-02 -7.13833869e-01 3.06934863e-01
-7.15509951e-02 5.07834494e-01 7.48069823e-01 -7.44607508e-01
1.15640128e+00 6.09812200e-01 7.43553340e-01 6.28820300e-01
-1.47468865e+00 -7.90347040e-01 -6.85513198e-01 -1.63182206e-02
-1.49684691e+00 -1.24782361e-01 9.42230597e-02 -7.14229286e-01
8.48953664e-01 3.79767358e-01 8.92299712e-01 8.30253005e-01
6.01625383e-01 6.89627171e-01 1.49624586e+00 -1.81681022e-01
3.59005094e-01 2.48574197e-01 3.11587244e-01 9.42523479e-01
1.67990386e-01 4.49482724e-02 2.18329467e-02 -2.49700755e-01
6.71906531e-01 4.85470057e-01 -2.42975757e-01 -6.86212033e-02
-1.08257699e+00 7.55653143e-01 3.45438719e-01 4.30172205e-01
-6.38651475e-02 -4.43000458e-02 5.42456925e-01 2.35591218e-01
2.68817216e-01 -4.54669595e-02 -3.23427349e-01 1.72965929e-01
-9.75554109e-01 -3.19496423e-01 6.44226849e-01 4.28962469e-01
9.32978392e-01 -5.21762550e-01 2.05020189e-01 5.74068487e-01
9.04046670e-02 5.57627559e-01 6.74134612e-01 -7.74381578e-01
-5.81157088e-01 6.31623387e-01 -8.03403929e-02 -3.38063449e-01
-9.10451710e-01 -5.30733824e-01 -1.10651207e+00 6.97417021e-01
7.76536226e-01 2.23782822e-01 -8.14860106e-01 1.42012119e+00
5.12275517e-01 -4.73895930e-02 -3.00045639e-01 9.13264155e-01
5.17181098e-01 -1.26629800e-01 2.47385234e-01 -2.38711745e-01
1.63540590e+00 -6.74327135e-01 -5.17761767e-01 3.50800186e-01
1.37665009e+00 -7.08749652e-01 1.22647047e+00 2.17364892e-01
-6.85736060e-01 1.93610698e-01 -8.14156473e-01 -2.76984990e-01
-6.97587669e-01 7.20928749e-03 9.55063641e-01 5.74260890e-01
-1.54375613e+00 1.16770737e-01 -1.04307675e+00 -6.50682151e-01
8.29529703e-01 7.01138854e-01 -9.19422567e-01 6.91810483e-03
-5.61656177e-01 8.30161989e-01 -1.86281279e-01 -1.03829771e-01
-6.28918827e-01 -1.18819737e+00 -4.67482448e-01 -5.22988617e-01
-2.70897359e-01 -9.08779740e-01 1.20755458e+00 -1.09649765e+00
-1.51055956e+00 1.62762511e+00 -2.73935020e-01 -2.33334213e-01
5.17576277e-01 5.99041224e-01 -1.07850999e-01 3.47318649e-01
3.11550852e-02 8.36218774e-01 -5.43128140e-02 -8.68195474e-01
-7.37245440e-01 -6.10383570e-01 -4.90014583e-01 -2.02410817e-01
-1.97091654e-01 -2.17758551e-01 -2.57503420e-01 -3.05493176e-01
-3.48993987e-01 -1.06322265e+00 -3.63530777e-02 5.90031326e-01
-2.92134225e-01 9.34290886e-02 7.67777503e-01 -6.25057578e-01
6.65871799e-01 -2.07639647e+00 -1.83878839e-01 3.96527350e-01
5.10034800e-01 1.47426635e-01 -1.61189958e-01 -1.33823538e-02
-1.92920249e-02 4.55318272e-01 7.71720633e-02 -4.00109679e-01
5.78846084e-03 1.41675353e-01 1.61002889e-01 1.06246746e+00
-1.37811512e-01 1.20707572e+00 -9.47692335e-01 -9.83686090e-01
1.30471528e-01 6.16935015e-01 -3.07098985e-01 -5.90203926e-02
1.88111052e-01 5.30106366e-01 -2.27025580e-02 1.31966758e+00
6.34678066e-01 -4.57094610e-01 3.48099738e-01 -3.72308135e-01
7.43628442e-02 -2.42007717e-01 -4.93873388e-01 1.32011640e+00
-4.22877967e-01 6.74690247e-01 7.15989947e-01 -1.73582062e-01
3.29937011e-01 1.63394123e-01 6.60918474e-01 -6.18730605e-01
5.25277197e-01 4.74668145e-01 -2.35559613e-01 -1.43943593e-01
-9.60034579e-02 -8.08427215e-01 2.65439302e-01 6.73892617e-01
-5.31693883e-02 -3.50462735e-01 9.79239270e-02 7.13410378e-02
1.39511681e+00 -4.71362323e-01 3.96612853e-01 -4.84738946e-01
5.12688279e-01 6.09517336e-01 4.78761613e-01 1.17888540e-01
-8.34199309e-01 4.79073972e-01 4.93555605e-01 -5.73759615e-01
-9.47395921e-01 -1.15408516e+00 -3.71888667e-01 7.97825992e-01
-2.65401900e-01 2.49989808e-01 -4.52070355e-01 -6.29548013e-01
2.06078917e-01 6.62416145e-02 -1.11238933e+00 4.15443540e-01
-3.38752002e-01 -9.12494183e-01 8.21851492e-01 5.67404985e-01
-8.89014229e-02 -5.29817522e-01 -4.57315207e-01 -6.85488433e-02
1.05104566e-01 -6.62197769e-01 -4.18314755e-01 3.77487034e-01
-7.69271970e-01 -1.41602051e+00 -7.63214290e-01 -9.29534793e-01
1.03724778e+00 1.41276807e-01 9.96362984e-01 6.37268722e-01
-8.39959979e-01 2.54198372e-01 -5.87502010e-02 -4.76451635e-01
-5.33586562e-01 1.48861319e-01 -7.68029019e-02 -3.21462989e-01
4.81989861e-01 -5.11614323e-01 -9.02220905e-01 2.10049346e-01
-1.04233778e+00 2.54209220e-01 9.05873656e-01 1.19629061e+00
1.35723031e+00 -4.12907928e-01 1.65924132e-01 -1.18685138e+00
1.76158294e-01 -3.49843115e-01 -5.54932177e-01 1.47607282e-01
-3.15613538e-01 -3.70152801e-01 6.39448881e-01 -1.75153479e-01
-5.90887010e-01 1.61575869e-01 -3.22078079e-01 -2.84581929e-01
-2.17640415e-01 3.00649792e-01 5.39473653e-01 -3.59368950e-01
5.76178610e-01 -1.80847809e-01 7.84508705e-01 1.65309638e-01
-1.47950441e-01 5.42900443e-01 6.43178403e-01 -4.03698198e-02
7.73779571e-01 1.19473290e+00 3.65556568e-01 -5.06430566e-01
-4.77647394e-01 -6.59633398e-01 -4.95171309e-01 -3.09233330e-02
8.54113221e-01 -8.20946515e-01 -1.00913346e+00 5.17910182e-01
-5.23115993e-01 -9.87625360e-01 -2.48229548e-01 4.98880595e-01
-3.69073033e-01 9.11714137e-02 -1.27617717e+00 -2.60380119e-01
-8.19281042e-01 -1.07936752e+00 1.26907670e+00 1.67832673e-01
-6.56223178e-01 -1.29376781e+00 5.64592957e-01 3.72454762e-01
7.10769236e-01 6.66556418e-01 1.04135776e+00 -5.60593426e-01
-1.93279564e-01 -5.60570478e-01 -3.26556027e-01 -2.32870758e-01
2.27395996e-01 6.09533310e-01 -1.07574856e+00 -3.28353107e-01
-4.27310765e-01 -4.90601242e-01 4.95667160e-01 3.14493775e-01
8.34694684e-01 8.20422359e-03 -8.01005065e-01 9.27601695e-01
1.61061990e+00 -6.54260665e-02 7.35542297e-01 4.26797181e-01
5.09881377e-01 5.39546549e-01 3.69866312e-01 1.48987785e-01
4.55438018e-01 2.98483759e-01 3.29718560e-01 -8.42991292e-01
-3.13410074e-01 9.37560350e-02 3.97546683e-03 4.86323267e-01
2.58884847e-01 -4.56437543e-02 -1.09825039e+00 6.11916184e-01
-1.09684455e+00 -6.50560141e-01 -1.26484677e-01 1.73430407e+00
1.02405751e+00 -1.87561572e-01 1.34835020e-01 1.57474959e-03
4.39676285e-01 -5.50844312e-01 -5.21662176e-01 -2.35051900e-01
-3.55429828e-01 5.82528532e-01 6.59158587e-01 7.23638654e-01
-6.89266622e-01 4.38255519e-01 6.34479570e+00 8.50390434e-01
-1.70619035e+00 3.69023532e-01 1.17848611e+00 -3.96136820e-01
-3.85749578e-01 -1.37755215e-01 -5.82739294e-01 1.84738621e-01
6.49840772e-01 1.00287059e-02 8.39668736e-02 5.75757504e-01
-3.39134559e-02 -4.33338612e-01 -1.00674939e+00 8.25979769e-01
-2.63540208e-01 -1.79978478e+00 -2.95125365e-01 6.11457646e-01
4.08706188e-01 3.44853759e-01 9.15692598e-02 1.81814209e-01
5.17004848e-01 -9.89546239e-01 7.82897994e-02 4.39867049e-01
1.38576829e+00 -4.38388586e-01 1.37193525e+00 5.20243198e-02
-8.85940313e-01 3.50514740e-01 9.73736942e-02 4.13752645e-01
-2.09565148e-01 7.10199237e-01 -1.43897891e+00 5.46452962e-02
4.88678306e-01 2.93932110e-01 -6.68753684e-01 3.92806858e-01
1.67514265e-01 2.57198155e-01 -5.15219033e-01 -1.63166970e-01
-2.08626781e-02 1.69015944e-01 -6.98961094e-02 1.40013528e+00
1.34123102e-01 3.13306367e-03 -2.04605430e-01 4.96400535e-01
1.04934402e-01 1.62298083e-01 -1.20270617e-01 3.09177693e-02
2.18573511e-01 2.08203745e+00 -1.16757846e+00 -2.10746825e-01
-2.55688071e-01 7.32973516e-01 4.24228311e-01 1.21420071e-01
-7.29807019e-01 -7.39953294e-02 6.36895716e-01 4.56150562e-01
-1.27402544e-01 2.47738183e-01 -4.44348663e-01 -7.07861722e-01
-6.62728488e-01 -6.00300550e-01 7.65118539e-01 -5.93875885e-01
-1.51188350e+00 4.02920544e-01 -6.61503136e-01 -1.07037103e+00
-4.70069274e-02 -9.75888312e-01 -7.83652246e-01 8.25974822e-01
-1.79856670e+00 -1.38070011e+00 -5.85768342e-01 4.85958993e-01
-6.80998489e-02 2.26761296e-01 1.12046754e+00 1.67814940e-01
-8.28175426e-01 8.98724914e-01 -1.77446961e-01 1.20632835e-01
1.03829324e+00 -1.36069047e+00 -6.21867895e-01 1.87207073e-01
-8.42988133e-01 5.91642976e-01 4.00202960e-01 -4.71568972e-01
-1.73791325e+00 -1.19514120e+00 6.10866368e-01 -6.84862971e-01
8.80524337e-01 -1.30995572e-01 -5.47452390e-01 7.31855392e-01
2.05970079e-01 5.22394717e-01 1.59281480e+00 -3.03798288e-01
-1.94170311e-01 -2.19929457e-01 -1.61371458e+00 7.29411542e-01
3.83808374e-01 -6.01985395e-01 2.92605609e-01 5.39539695e-01
-8.55456293e-02 -5.57663798e-01 -1.29060817e+00 2.72649884e-01
8.87767792e-01 -8.40826571e-01 4.69333291e-01 -4.05556038e-02
2.42059883e-02 -6.99849963e-01 -7.22463205e-02 -8.62290263e-01
-2.91477114e-01 -2.60840923e-01 3.50082248e-01 7.31343091e-01
5.96077800e-01 -1.00452745e+00 1.03483176e+00 5.04755735e-01
-1.94399014e-01 -1.09820616e+00 -1.03975165e+00 -3.10127050e-01
4.11019832e-01 1.74014822e-01 4.17831868e-01 9.31278229e-01
3.47731829e-01 -2.65970558e-01 7.34848797e-01 -1.09358899e-01
4.94887769e-01 -9.41804647e-02 9.90868211e-01 -8.47161472e-01
-3.11818600e-01 -9.46624756e-01 -5.21028042e-01 5.55388890e-02
1.93865046e-01 -1.21553922e+00 -2.97854930e-01 -1.30921412e+00
5.38945735e-01 -4.62553948e-01 -3.89630347e-01 8.35264802e-01
8.25257301e-02 9.12355483e-01 -2.64476150e-01 4.92888540e-01
-3.88423920e-01 -2.23958626e-01 1.27866602e+00 -3.50917459e-01
6.70071170e-02 -6.35250330e-01 -7.74895847e-01 4.46631461e-01
8.62866998e-01 -3.60423237e-01 1.18721403e-01 1.50448140e-02
2.08015695e-01 -1.78806037e-02 5.81830263e-01 -7.97249377e-01
3.51184696e-01 -2.33544484e-01 7.76421189e-01 -6.07514024e-01
1.23838857e-02 -6.60099864e-01 5.92782617e-01 9.12479043e-01
5.43630831e-02 7.51997381e-02 4.67786312e-01 7.53246471e-02
-7.01544061e-02 4.48926330e-01 1.02838910e+00 -2.20818892e-01
-3.18613708e-01 3.76788944e-01 -6.95335090e-01 -4.23706383e-01
1.25239122e+00 -5.59413910e-01 -9.52401042e-01 2.43697464e-02
-5.67030966e-01 3.52924109e-01 1.07121658e+00 -4.16770875e-01
1.91558599e-01 -1.01063740e+00 -6.49504542e-01 4.39248420e-02
2.63414323e-01 -4.99632955e-02 5.83521068e-01 1.60548460e+00
-1.09891677e+00 1.72052696e-01 -3.75491112e-01 -8.50721300e-01
-1.24887133e+00 4.27473158e-01 7.04648256e-01 -8.40574324e-01
-2.97130525e-01 1.01309454e+00 3.68515521e-01 -5.06378710e-01
-1.53387159e-01 -8.59243795e-02 2.37966359e-01 -2.55544651e-02
6.56618416e-01 2.34016493e-01 3.17131490e-01 -5.46083689e-01
-5.26782632e-01 3.73964339e-01 -4.56758797e-01 -3.02498345e-03
1.05502963e+00 8.91093835e-02 -4.82098192e-01 2.41227597e-01
1.53739512e+00 3.59153487e-02 -9.13384378e-01 2.42733672e-01
-3.58846784e-01 -1.28114074e-02 2.65493810e-01 -9.41611409e-01
-1.25450039e+00 5.17564893e-01 8.64385545e-01 -3.20179015e-01
1.28956175e+00 -3.26609868e-03 8.70946109e-01 -1.14790760e-01
3.88420850e-01 -9.28644896e-01 -2.32110158e-01 1.68970227e-01
5.94196081e-01 -1.22721326e+00 4.22258452e-02 -3.58952850e-01
-3.20139199e-01 1.15081179e+00 5.22960603e-01 1.07455775e-02
6.35149121e-01 1.09171140e+00 7.45678127e-01 -3.12673867e-01
-9.24609482e-01 -5.24345189e-02 -3.69729996e-01 6.12061620e-01
6.71973526e-01 3.87781829e-01 -1.27237871e-01 4.16333139e-01
-4.17268455e-01 3.79665822e-01 4.23224598e-01 1.06893694e+00
3.89663316e-02 -9.67117190e-01 -2.43420810e-01 5.05235493e-01
-4.56896901e-01 1.73882365e-01 -5.75258255e-01 1.12608278e+00
4.97917943e-02 3.84968817e-01 2.33455896e-01 -1.51894018e-01
-2.73289438e-02 -9.38067064e-02 4.35648382e-01 -4.49058175e-01
-8.12927425e-01 2.71836728e-01 -2.15412602e-01 -4.62341189e-01
-2.38479152e-01 -6.20758295e-01 -1.55157757e+00 -5.69263339e-01
-1.48710936e-01 -1.98058635e-01 5.81172228e-01 6.86784983e-01
2.75550097e-01 4.39682454e-01 4.66328770e-01 -7.06052125e-01
4.66773845e-02 -7.88940489e-01 -8.53233933e-01 2.92168111e-01
4.99409467e-01 -4.76141304e-01 -7.68927813e-01 1.25407144e-01] | [15.059233665466309, -3.0642971992492676] |
6f587b6e-2310-403b-89c1-d9fde22a5f3d | dgpose-disentangled-semi-supervised-deep | 1804.06364 | null | https://arxiv.org/abs/1804.06364v2 | https://arxiv.org/pdf/1804.06364v2.pdf | DGPose: Deep Generative Models for Human Body Analysis | Deep generative modelling for human body analysis is an emerging problem with many interesting applications. However, the latent space learned by such approaches is typically not interpretable, resulting in less flexibility. In this work, we present deep generative models for human body analysis in which the body pose and the visual appearance are disentangled. Such a disentanglement allows independent manipulation of pose and appearance, and hence enables applications such as pose-transfer without specific training for such a task. Our proposed models, the Conditional-DGPose and the Semi-DGPose, have different characteristics. In the first, body pose labels are taken as conditioners, from a fully-supervised training set. In the second, our structured semi-supervised approach allows for pose estimation to be performed by the model itself and relaxes the need for labelled data. Therefore, the Semi-DGPose aims for the joint understanding and generation of people in images. It is not only capable of mapping images to interpretable latent representations but also able to map these representations back to the image space. We compare our models with relevant baselines, the ClothNet-Body and the Pose Guided Person Generation networks, demonstrating their merits on the Human3.6M, ChictopiaPlus and DeepFashion benchmarks. | ['Adnane Boukhayma', 'N. Siddharth', 'Arnab Ghosh', 'Thalaiyasingam Ajanthan', 'Rodrigo de Bem', 'Philip Torr', 'Ondrej Miksik'] | 2018-04-17 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 2.94283628e-01 4.88696784e-01 8.60919729e-02 -3.62672389e-01
-1.62980929e-01 -5.20242155e-01 7.49988556e-01 -3.90077800e-01
-2.65184879e-01 6.00686610e-01 3.16536486e-01 3.38043213e-01
2.32062079e-02 -7.09419191e-01 -8.49873543e-01 -9.44414735e-01
1.78024054e-01 9.33977723e-01 3.45398411e-02 -2.24169135e-01
-6.08796656e-01 1.20589450e-01 -1.58367264e+00 -7.52431378e-02
7.62019873e-01 8.56744051e-01 8.54650512e-02 6.97549462e-01
4.90363896e-01 3.57379109e-01 -4.50394511e-01 -6.30312085e-01
3.73989254e-01 -6.56064868e-01 -5.17803848e-01 5.69347441e-01
6.44767702e-01 -3.68297875e-01 -1.79684177e-01 8.46732914e-01
5.41465998e-01 -3.12808827e-02 8.56983244e-01 -1.31706452e+00
-3.43203902e-01 4.27703112e-01 -4.80487496e-01 -5.66499054e-01
4.87646520e-01 1.18299738e-01 8.17925811e-01 -5.08534968e-01
7.07497239e-01 1.48870158e+00 4.56191748e-01 8.48165631e-01
-1.54246950e+00 -3.58225346e-01 1.92807257e-01 -1.33282840e-01
-1.22925234e+00 -2.99317837e-01 8.38996649e-01 -8.53509486e-01
1.50089383e-01 3.57908010e-01 1.11204505e+00 1.48250675e+00
2.61192676e-02 7.97309518e-01 1.21501458e+00 -5.65680385e-01
1.04037926e-01 5.78561649e-02 -3.06302607e-01 8.25867712e-01
2.40603775e-01 5.00019938e-02 -5.06774008e-01 3.53253111e-02
1.12267256e+00 1.28750443e-01 -3.44456136e-01 -1.12207878e+00
-1.32003558e+00 7.67208815e-01 4.62647468e-01 -1.87855393e-01
-4.14506555e-01 3.16776305e-01 2.20171437e-01 -6.67830836e-03
4.50429380e-01 1.13871418e-01 -8.40195715e-02 2.08341196e-01
-9.49311435e-01 5.10286331e-01 8.02839935e-01 7.94084907e-01
4.02996421e-01 8.15922245e-02 -3.55564862e-01 4.81694311e-01
7.55924344e-01 7.72294641e-01 2.41503075e-01 -7.63257146e-01
4.79584217e-01 5.57034612e-01 1.40856624e-01 -1.03783298e+00
-3.94438416e-01 -5.69404483e-01 -1.05213666e+00 4.21807766e-01
6.84512675e-01 -1.79499045e-01 -1.29433143e+00 2.12394857e+00
5.82341552e-01 -4.31831360e-01 -1.69373482e-01 1.24693680e+00
6.70429468e-01 3.07129741e-01 1.92795545e-01 1.14465654e-01
1.58515644e+00 -8.99778962e-01 -6.09355688e-01 -4.53364581e-01
3.59650180e-02 -5.37167192e-01 8.67935717e-01 5.23626745e-01
-1.03894687e+00 -7.27578282e-01 -9.57973659e-01 -2.96030380e-02
-2.60237232e-02 4.69313055e-01 5.31434119e-01 6.00424767e-01
-8.60873759e-01 5.43981910e-01 -1.15956688e+00 -4.79353130e-01
2.19415158e-01 5.20692825e-01 -5.18842518e-01 1.63031489e-01
-1.11207080e+00 6.75524473e-01 3.10684592e-01 4.51524794e-01
-9.60243702e-01 -7.48285800e-02 -1.17368388e+00 -5.49330786e-02
3.36532265e-01 -1.27688992e+00 8.85527611e-01 -9.88861561e-01
-1.71282053e+00 1.20888019e+00 1.60192490e-01 -3.62090617e-01
1.17542219e+00 -5.25129139e-01 7.02705085e-02 9.39460397e-02
-1.38712227e-01 8.22226763e-01 1.39554918e+00 -1.29810524e+00
-6.50488585e-02 -6.04132533e-01 8.05021152e-02 3.43512118e-01
-5.06701916e-02 -2.92989492e-01 -5.87083578e-01 -8.42042208e-01
2.37203553e-01 -1.43322182e+00 -2.61915267e-01 2.70482302e-01
-7.59211957e-01 1.63570940e-01 5.37347734e-01 -9.37239945e-01
7.91376293e-01 -1.83716631e+00 1.09693837e+00 2.90825367e-01
5.15779614e-01 6.19840771e-02 2.37920791e-01 2.70360768e-01
-8.61568749e-02 -2.34391570e-01 -1.51306242e-01 -8.78593564e-01
3.38500619e-01 4.24365819e-01 7.71861244e-03 6.92249298e-01
-1.56621123e-03 1.00402892e+00 -7.30377078e-01 -6.12030447e-01
4.37951922e-01 6.71131968e-01 -6.07181013e-01 3.75016600e-01
-2.69801795e-01 1.08364511e+00 -2.30679393e-01 2.17593506e-01
3.71792018e-01 -2.48641759e-01 3.86502922e-01 -4.28789258e-01
2.77933002e-01 -1.04328975e-01 -1.14016020e+00 1.92543960e+00
-3.97947490e-01 1.59697190e-01 1.92229643e-01 -6.52045608e-01
7.02717662e-01 5.57422340e-01 5.73729277e-01 -2.72659004e-01
1.85976028e-01 2.11168081e-02 1.61141098e-01 -3.05673480e-01
7.27446824e-02 -2.78063536e-01 -1.83250457e-01 3.03448200e-01
2.34874427e-01 1.22326063e-02 1.90717891e-01 -3.12898308e-04
5.00706434e-01 7.58986950e-01 2.15690672e-01 -1.90576762e-01
3.51152599e-01 -5.08737326e-01 4.58763987e-01 3.04169714e-01
1.56100869e-01 9.16532159e-01 3.13386768e-01 -3.99241894e-01
-8.83125782e-01 -1.44198322e+00 2.30173886e-01 9.54383790e-01
-2.22997647e-02 -4.00604963e-01 -9.17997956e-01 -7.52834201e-01
8.51873010e-02 2.56312788e-01 -9.14297223e-01 -1.83294892e-01
-5.91307580e-01 -5.30068159e-01 2.24623233e-01 6.47213221e-01
2.47813970e-01 -1.01978970e+00 -8.66678059e-01 1.99514553e-02
-4.25571084e-01 -1.02562451e+00 -4.82950598e-01 -2.74115335e-02
-7.59752393e-01 -8.21530581e-01 -1.18862081e+00 -4.16758835e-01
9.48788643e-01 -5.27184010e-01 1.11952817e+00 -2.95539498e-01
-3.15688699e-01 4.99822259e-01 -2.41707653e-01 -3.88788819e-01
-4.38453078e-01 4.22569439e-02 2.65877813e-01 4.20451641e-01
-2.12140858e-01 -8.68339479e-01 -9.15296793e-01 4.53338683e-01
-6.97993577e-01 5.69006085e-01 6.94989800e-01 7.43897080e-01
6.75187707e-01 -4.99278128e-01 1.16238981e-01 -7.58267403e-01
1.24199644e-01 -7.99550861e-02 -3.58958632e-01 2.21232697e-01
-4.63172555e-01 1.92654222e-01 2.67972052e-01 -6.07492983e-01
-9.71351147e-01 5.64217269e-01 -1.95568293e-01 -5.55050254e-01
-2.34355643e-01 8.57194066e-02 -6.80714428e-01 1.51269138e-01
6.51253998e-01 8.44690204e-02 2.67401397e-01 -7.62810707e-01
5.32434344e-01 8.67026597e-02 6.27438784e-01 -6.72817588e-01
1.04950213e+00 6.26468658e-01 2.03163922e-01 -6.77617848e-01
-6.89466238e-01 -1.64383933e-01 -1.06265330e+00 -3.35723788e-01
1.27312040e+00 -8.22547674e-01 -5.02189875e-01 5.00156343e-01
-1.06839967e+00 -3.13261241e-01 -5.01496673e-01 3.77091944e-01
-9.12692785e-01 4.33026671e-01 -5.66323519e-01 -7.72537112e-01
-3.95799875e-01 -1.04385328e+00 1.55016983e+00 -3.14037502e-03
-5.69788873e-01 -1.13917053e+00 -2.27774996e-02 4.57001030e-01
4.76588644e-02 9.21656966e-01 6.12490296e-01 -3.03340375e-01
-3.80484283e-01 -2.51150012e-01 1.91566527e-01 3.92204881e-01
2.34377772e-01 -3.85398865e-01 -1.04345286e+00 -5.23503363e-01
-1.41692936e-01 -4.09497410e-01 6.25761330e-01 5.28282404e-01
8.35609198e-01 -4.81210500e-01 -2.63985664e-01 7.10809469e-01
9.16364551e-01 -4.66180354e-01 6.26570225e-01 -1.47855580e-01
1.17009497e+00 8.76159310e-01 3.95683438e-01 4.14824486e-01
3.57702911e-01 1.15699625e+00 5.68485618e-01 -4.69084173e-01
-2.38154769e-01 -6.08313680e-01 3.78053069e-01 6.73354268e-01
-7.53437996e-01 -3.21220577e-01 -6.72952116e-01 1.94201902e-01
-1.96173406e+00 -6.66965485e-01 -1.97948247e-01 2.25879836e+00
7.63606310e-01 -4.84059304e-02 4.62598503e-01 1.82779863e-01
5.35421073e-01 1.98394246e-02 -4.10216093e-01 5.83158731e-02
1.69132680e-01 2.04783022e-01 2.63091624e-01 2.98949212e-01
-1.20616794e+00 5.76918662e-01 5.12870073e+00 4.03841913e-01
-9.45905209e-01 1.12799078e-01 2.94112891e-01 -1.30517036e-01
-4.88767698e-02 -1.48310110e-01 -6.39637768e-01 3.68570983e-01
4.69035387e-01 2.70600796e-01 1.56264737e-01 6.92554951e-01
1.09922662e-01 1.34644270e-01 -1.54760528e+00 1.12569404e+00
1.75180286e-01 -5.74565947e-01 2.47963831e-01 4.48906898e-01
5.18115342e-01 -5.06824493e-01 1.91137269e-01 5.50574586e-02
5.86874560e-02 -1.00086057e+00 1.26984584e+00 7.74278522e-01
8.49257827e-01 -4.27487284e-01 4.62781847e-01 4.69919771e-01
-9.76806581e-01 2.70287901e-01 1.05523303e-01 9.64381024e-02
5.42177200e-01 3.91825974e-01 -5.35023093e-01 7.50476122e-01
4.81423408e-01 4.69929695e-01 -5.43918490e-01 6.64954245e-01
-5.72748125e-01 3.90025973e-01 -3.53968054e-01 3.96050572e-01
-3.64252210e-01 -2.38409966e-01 7.39191234e-01 9.79211330e-01
2.12339312e-01 -4.24337715e-01 3.88680071e-01 1.09877431e+00
1.27019748e-01 -5.13321310e-02 -2.68096775e-01 5.73600903e-02
-3.45859796e-01 1.19950163e+00 -7.04088092e-01 -1.67592496e-01
1.04978666e-01 1.33406162e+00 6.60814047e-02 2.71182895e-01
-8.37196171e-01 1.71748832e-01 4.48062241e-01 6.02982700e-01
2.38833755e-01 -2.99670517e-01 9.09388363e-02 -1.31359148e+00
7.26317018e-02 -8.61593962e-01 2.68870533e-01 -6.29104018e-01
-1.20976901e+00 4.50847685e-01 4.72652137e-01 -1.15747929e+00
-8.39154720e-01 -6.51436985e-01 -2.97075778e-01 9.01594937e-01
-8.98506343e-01 -1.73168445e+00 -5.28376758e-01 5.51612258e-01
2.85533369e-01 3.25024784e-01 8.72263789e-01 2.83342123e-01
-3.98001671e-01 5.40220559e-01 -4.22580391e-01 2.53210872e-01
7.21349537e-01 -1.59018278e+00 3.93893957e-01 6.84734881e-01
3.41722697e-01 7.34074593e-01 1.08864391e+00 -6.77590668e-01
-1.20985794e+00 -9.79861498e-01 3.79138529e-01 -6.98809445e-01
8.43422115e-02 -8.21014762e-01 -5.44794023e-01 7.15512872e-01
-4.58098985e-02 1.08137533e-01 5.70379913e-01 1.16228610e-01
-1.15264669e-01 -9.71283466e-02 -8.63223493e-01 4.79217708e-01
1.27028728e+00 -2.37750381e-01 -3.90492201e-01 3.06260407e-01
3.13967347e-01 -7.12616205e-01 -8.35949659e-01 2.65650421e-01
9.96008515e-01 -9.08970594e-01 1.12682855e+00 -4.76647258e-01
5.08754492e-01 -3.15743476e-01 8.53855014e-02 -1.25006199e+00
-3.00197899e-01 -5.72590888e-01 -5.59246778e-01 1.22847462e+00
9.92667377e-02 -4.22652066e-01 8.78588557e-01 5.27228951e-01
1.50765106e-01 -6.00068092e-01 -7.30105639e-01 -7.68265367e-01
-1.32008776e-01 -8.59968737e-03 3.36289287e-01 6.14658594e-01
-4.39116389e-01 4.34470832e-01 -9.53917980e-01 1.46302164e-01
9.30902839e-01 2.84529388e-01 1.14107394e+00 -1.46161652e+00
-9.47016716e-01 -6.22050092e-02 -6.64211571e-01 -1.25808549e+00
-2.53188275e-02 -6.28852367e-01 1.45915359e-01 -1.58644438e+00
1.82422981e-01 -1.33676112e-01 8.20210725e-02 5.80757558e-01
-1.91219915e-02 3.57623547e-01 3.34608555e-01 1.50730848e-01
-3.49181533e-01 7.33775616e-01 1.50363564e+00 -6.00103661e-02
-1.70071766e-01 2.57622719e-01 -2.58571744e-01 9.95245814e-01
4.46930200e-01 -3.01892638e-01 -5.89681447e-01 -3.19688797e-01
3.54447335e-01 7.47493282e-02 9.55804706e-01 -1.03892994e+00
-1.08263776e-01 1.58523589e-01 6.11191273e-01 -3.07025135e-01
7.42689133e-01 -8.46851349e-01 6.05073094e-01 5.87134540e-01
-1.51933923e-01 -2.59018064e-01 -9.16537270e-02 7.33584404e-01
3.87328379e-02 1.32422701e-01 6.96638346e-01 -2.27913499e-01
-2.72380203e-01 3.50535810e-01 1.12390652e-01 -4.97728121e-03
8.32580924e-01 -3.39843035e-01 2.27038085e-01 -5.92427552e-01
-1.32694781e+00 1.11487433e-02 6.22757673e-01 4.73245800e-01
4.58256245e-01 -1.37576723e+00 -7.53959358e-01 2.15165600e-01
1.89244375e-01 2.59370506e-01 3.75606447e-01 1.04568958e+00
-3.32980275e-01 6.16490878e-02 -3.13777000e-01 -9.22797799e-01
-1.48817062e+00 5.26420474e-01 4.75347459e-01 -4.07207161e-01
-7.98566341e-01 5.92193007e-01 8.19611788e-01 -3.75939012e-01
8.36424753e-02 -3.69016200e-01 -9.20256823e-02 -1.37519419e-01
7.94150233e-02 1.22269556e-01 -2.49127448e-01 -1.05144691e+00
-1.76587567e-01 6.17961228e-01 2.52282411e-01 -2.48976991e-01
1.23142707e+00 -1.64115816e-01 8.80149454e-02 4.37645257e-01
8.16167593e-01 1.22169189e-01 -1.52811992e+00 1.70918237e-02
-5.26286483e-01 -2.42845818e-01 -3.31064939e-01 -7.69226134e-01
-1.08068848e+00 1.06745934e+00 7.01268673e-01 -3.54140848e-02
1.03578603e+00 1.62196904e-01 4.78383541e-01 -1.67507604e-01
5.26012301e-01 -8.30559373e-01 1.63844913e-01 4.75208052e-02
1.22815704e+00 -9.89099085e-01 1.37896687e-01 -6.46124721e-01
-5.38226783e-01 8.98942590e-01 5.98384619e-01 2.11227220e-03
4.02018517e-01 -6.40723780e-02 -8.23774040e-02 -2.81713665e-01
-1.86226845e-01 -3.47403437e-01 9.63590741e-01 7.24837184e-01
3.20071101e-01 2.58744359e-01 -1.12823419e-01 6.68514967e-01
-4.55925703e-01 -1.87105596e-01 -1.16009362e-01 7.21831799e-01
1.09040797e-01 -1.35565293e+00 -5.39462805e-01 5.60921356e-02
-2.34296337e-01 4.19499159e-01 -5.76089919e-01 8.37565601e-01
4.86532509e-01 4.51422989e-01 -1.75232381e-01 -3.04694921e-01
4.94313240e-01 1.21986985e-01 9.63512957e-01 -7.95281887e-01
-1.70066893e-01 3.06500286e-01 -1.98874399e-02 -4.24462885e-01
-5.40659547e-01 -6.51286602e-01 -8.08019757e-01 1.12069525e-01
-2.72825569e-01 -6.90651238e-02 6.05294466e-01 8.59165609e-01
1.30493373e-01 6.72368169e-01 9.21809897e-02 -1.34269845e+00
-6.00724876e-01 -1.07464504e+00 -6.40721083e-01 8.09959054e-01
2.49678969e-01 -1.07167137e+00 -1.06950663e-01 5.20171106e-01] | [7.231204986572266, -0.9237357378005981] |
ce2e65d6-edb8-434a-8620-226b7bc62c59 | fine-tuning-of-explainable-cnns-for-skin | 2304.01399 | null | https://arxiv.org/abs/2304.01399v1 | https://arxiv.org/pdf/2304.01399v1.pdf | Fine-tuning of explainable CNNs for skin lesion classification based on dermatologists' feedback towards increasing trust | In this paper, we propose a CNN fine-tuning method which enables users to give simultaneous feedback on two outputs: the classification itself and the visual explanation for the classification. We present the effect of this feedback strategy in a skin lesion classification task and measure how CNNs react to the two types of user feedback. To implement this approach, we propose a novel CNN architecture that integrates the Grad-CAM technique for explaining the model's decision in the training loop. Using simulated user feedback, we found that fine-tuning our model on both classification and explanation improves visual explanation while preserving classification accuracy, thus potentially increasing the trust of users in using CNN-based skin lesion classifiers. | ['Daniel Sonntag', 'Fabrizio Nunnari', 'Md Abdul Kadir'] | 2023-04-03 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 4.16604914e-02 7.27345943e-01 -9.37067866e-02 -5.02134323e-01
-6.51184190e-03 -5.32193124e-01 2.09019884e-01 3.00785780e-01
-4.18503374e-01 3.39129567e-01 1.76975220e-01 -5.27122974e-01
2.45988786e-01 -6.04871392e-01 -4.25551504e-01 -2.67184138e-01
4.23452407e-01 -8.52149725e-02 2.26491496e-01 -2.62851864e-02
5.37470341e-01 4.14507002e-01 -1.39291990e+00 1.04260159e+00
8.21080923e-01 8.59959543e-01 -6.52180016e-02 1.09437549e+00
2.09797144e-01 7.45302200e-01 -8.10627937e-01 -3.95343453e-01
2.85322778e-02 -3.07801515e-01 -8.93118322e-01 1.50205255e-01
7.19463646e-01 -4.32076573e-01 2.48968359e-02 7.31413782e-01
4.28171039e-01 -1.20903715e-01 4.58986521e-01 -1.18240714e+00
-8.42873096e-01 1.66752622e-01 -1.52689770e-01 3.23614240e-01
3.51293027e-01 5.54339051e-01 6.22562051e-01 -4.95861173e-01
7.86130130e-01 1.09546769e+00 8.94161880e-01 1.06885767e+00
-1.26048553e+00 -5.59949756e-01 1.21567585e-01 1.82340056e-01
-1.08706069e+00 -1.56038567e-01 4.51785594e-01 -6.74855769e-01
9.21225488e-01 5.13095617e-01 1.02649844e+00 1.04309380e+00
5.60191631e-01 4.52204943e-01 1.31230640e+00 -4.59939718e-01
4.11386400e-01 8.80597591e-01 2.10634366e-01 7.99276531e-01
1.66912094e-01 4.10346061e-01 -4.28646058e-01 -1.32199064e-01
1.20380688e+00 -3.34654301e-02 -3.28371912e-01 -1.35240331e-01
-5.86096585e-01 7.58309066e-01 1.03276896e+00 1.65800035e-01
-3.35952997e-01 3.24376583e-01 1.78129286e-01 3.53998810e-01
4.59099770e-01 7.72973895e-01 -4.82697546e-01 1.37097910e-01
-6.01525009e-01 -7.85216540e-02 7.24250257e-01 2.75540978e-01
4.50069934e-01 -1.50149748e-01 -5.57659328e-01 5.66830695e-01
2.53094763e-01 -2.99938489e-02 5.45056820e-01 -8.04799676e-01
-2.85507262e-01 9.46990728e-01 -1.09404087e-01 -1.09300363e+00
-5.45137942e-01 -6.52578831e-01 -5.45283675e-01 9.56098080e-01
4.03086990e-01 -1.84664622e-01 -1.30831540e+00 1.54663646e+00
1.32176846e-01 -1.76472977e-01 -2.86125988e-01 1.04786766e+00
9.04490054e-01 -8.77778158e-02 4.70658690e-01 3.60306799e-01
1.33471346e+00 -9.52728629e-01 -5.53779662e-01 -1.46577358e-01
7.73070395e-01 -4.13338095e-01 1.33503699e+00 3.87460440e-01
-9.70528126e-01 -7.26500988e-01 -9.40439165e-01 -2.00924166e-02
-3.98621798e-01 3.60170037e-01 5.42370319e-01 6.33626282e-01
-1.52303600e+00 1.01443362e+00 -6.06193364e-01 -7.74551749e-01
6.54839933e-01 4.78609473e-01 -4.17280406e-01 1.27654612e-01
-6.98740900e-01 1.14956188e+00 -1.58793256e-02 1.62272248e-03
-6.43640757e-01 -6.73984885e-01 -4.78775471e-01 2.70676404e-01
-2.59166002e-01 -9.73329425e-01 1.64563596e+00 -1.53513432e+00
-1.24922347e+00 9.74262893e-01 -1.14263281e-01 -1.48526967e-01
6.97026610e-01 7.68179297e-02 -1.20874420e-01 -1.27361983e-03
-3.09898317e-01 1.08386779e+00 8.91972601e-01 -1.35938597e+00
-3.18950921e-01 -3.77709299e-01 4.48470622e-01 5.29564209e-02
-4.73302275e-01 -4.50782001e-01 -1.22748755e-01 -4.44794536e-01
-2.59168118e-01 -7.97985077e-01 -5.18335700e-01 6.56210542e-01
-3.01881433e-01 1.35064974e-01 8.97030532e-01 -5.18729627e-01
1.01830375e+00 -2.24485874e+00 -2.51732111e-01 4.96115059e-01
7.98561752e-01 4.33594555e-01 -1.44456327e-01 3.27864915e-01
-4.06860918e-01 4.41078305e-01 2.79255897e-01 -3.83066952e-01
-3.61946493e-01 2.45246604e-01 1.84768170e-01 1.09753720e-01
3.36994499e-01 1.08625150e+00 -7.75041044e-01 -2.76839584e-01
6.48789406e-02 7.65245914e-01 -6.88231885e-01 3.15041304e-01
6.91073984e-02 3.49361181e-01 -9.82724689e-03 3.67444277e-01
5.48057973e-01 -6.19496405e-01 2.67969251e-01 -2.70501196e-01
1.46226093e-01 -2.21946239e-02 -6.94700241e-01 1.15615571e+00
-4.90055323e-01 1.07650089e+00 3.61559093e-02 -1.98548630e-01
5.54099739e-01 3.79325986e-01 -3.12316447e-01 -8.27623606e-01
3.12313914e-01 -2.32268393e-01 2.40904585e-01 -6.73042655e-01
1.93224847e-01 -1.47356898e-01 5.95023453e-01 6.54987872e-01
3.11803725e-03 6.07593060e-02 -2.09037319e-01 3.88023645e-01
1.09115398e+00 -2.02142939e-01 5.19703448e-01 -1.33063123e-01
8.21508765e-02 9.09769908e-02 -8.55463892e-02 8.04039717e-01
-4.79925483e-01 5.84802508e-01 7.51622379e-01 -7.08275199e-01
-8.84069562e-01 -6.28606081e-01 3.00299644e-01 1.11058772e+00
-3.39045286e-01 -2.84949630e-01 -1.14318216e+00 -1.06788266e+00
1.11823171e-01 6.72235131e-01 -1.56263340e+00 -4.37312126e-01
1.30051807e-01 -5.43439090e-02 3.04941416e-01 8.80901098e-01
1.68747336e-01 -1.28808177e+00 -1.06065321e+00 -1.80328652e-01
2.67677426e-01 -4.17494446e-01 -3.66279274e-01 3.56838673e-01
-1.05317247e+00 -1.13940537e+00 -4.65966135e-01 -4.74454492e-01
1.30231297e+00 2.37138405e-01 9.75990295e-01 9.55789447e-01
-5.64240992e-01 3.54686618e-01 -3.14159632e-01 -4.48796570e-01
-5.28755188e-01 1.23942964e-01 -6.00945294e-01 -2.03153357e-01
1.88050661e-02 -2.98783958e-01 -9.69840467e-01 1.58556625e-01
-1.03573549e+00 4.94338095e-01 8.31077695e-01 7.24388301e-01
-3.21141421e-03 -4.97648299e-01 9.89011154e-02 -1.31855464e+00
9.91974354e-01 -1.74692765e-01 -4.41186093e-02 1.45145953e-01
-8.89376998e-01 3.36946994e-02 4.05577481e-01 -5.95608532e-01
-7.32643425e-01 2.22797826e-01 -3.02345399e-02 -3.73414606e-01
-2.66751856e-01 4.01894987e-01 5.95791638e-01 -7.00936317e-01
1.20440781e+00 -2.96595961e-01 8.68744180e-02 -2.07342833e-01
2.73257315e-01 5.40667355e-01 4.59399194e-01 2.57152975e-01
6.59840286e-01 4.19034690e-01 -3.96257222e-01 -3.15826416e-01
-8.18057239e-01 -2.75260299e-01 -6.22859657e-01 -6.75081968e-01
8.52437973e-01 -4.30954039e-01 -1.08537138e+00 1.64665282e-01
-1.41941035e+00 -6.45396948e-01 -2.23961905e-01 -1.00001715e-01
-6.70047104e-02 -1.71149179e-01 -4.95884955e-01 -5.75219929e-01
-2.97131538e-01 -1.00737739e+00 8.54164004e-01 5.20784616e-01
-7.84653723e-01 -1.38477600e+00 1.98486462e-01 9.27965194e-02
8.88171494e-01 1.57735229e-01 9.83035803e-01 -5.61688364e-01
-3.11853170e-01 -3.74251664e-01 -6.60839081e-01 2.13790387e-01
4.46060039e-02 1.77683830e-01 -1.51212215e+00 -2.84042805e-01
-4.04633194e-01 -1.40711471e-01 8.52867007e-01 2.80818701e-01
1.36417925e+00 -3.27022940e-01 -3.79969418e-01 4.56694394e-01
1.31035948e+00 -2.75386292e-02 8.85293484e-01 7.81414062e-02
5.57090819e-01 6.25057697e-01 1.95016861e-01 1.93057284e-01
1.50104269e-01 4.62312043e-01 5.94745517e-01 -8.76423657e-01
-3.82726938e-01 -2.66801029e-01 -7.49022514e-02 1.11318171e-01
-2.11463034e-01 -3.93119715e-02 -8.24942708e-01 3.44893843e-01
-1.83328354e+00 -5.86430132e-01 -1.59475669e-01 1.83214831e+00
5.66035390e-01 1.23620495e-01 -1.51225850e-01 5.24613354e-03
4.35240746e-01 -3.77863288e-01 -4.73251790e-01 -8.93614292e-01
5.21518946e-01 4.74814802e-01 4.25134271e-01 6.98013246e-01
-7.28590131e-01 8.20127547e-01 7.76778746e+00 2.96162099e-01
-1.54405236e+00 7.64698982e-02 6.26640856e-01 8.42125490e-02
-3.98520827e-01 -1.74365029e-01 -1.34650305e-01 3.01920660e-02
6.88149929e-01 2.11355388e-01 2.65055925e-01 8.93354654e-01
3.78119946e-01 -3.51878494e-01 -1.09084380e+00 6.05159998e-01
5.22360317e-02 -1.60527492e+00 2.09125832e-01 2.47514825e-02
5.28339207e-01 -4.97533023e-01 3.56854498e-01 -4.24888395e-02
2.86309749e-01 -1.28566003e+00 5.40531516e-01 6.01589203e-01
8.04380953e-01 -4.93812114e-01 8.38457465e-01 1.02615178e-01
-5.74063838e-01 -1.45603091e-01 1.87105179e-01 -4.57510889e-01
-5.68573833e-01 1.61373511e-01 -1.65183640e+00 -1.14854008e-01
6.31710052e-01 1.71171784e-01 -1.13649678e+00 1.13253403e+00
-5.28630078e-01 4.92901981e-01 2.94148684e-01 -1.76430330e-01
-2.81304494e-02 6.39096558e-01 2.92885229e-02 1.38794267e+00
-2.17339054e-01 3.47474702e-02 -1.86487526e-01 8.67236972e-01
2.24110246e-01 -6.98868632e-02 -4.04800147e-01 4.06226292e-02
1.09898308e-02 1.51844168e+00 -8.21684420e-01 -3.11408430e-01
1.77578479e-01 1.26085079e+00 5.83191335e-01 3.85825276e-01
-4.86618936e-01 -3.38289022e-01 6.24175787e-01 4.22621101e-01
9.83434990e-02 1.01147234e-01 -9.60900903e-01 -7.06368208e-01
-2.81147331e-01 -6.77225709e-01 2.42219567e-01 -1.47431898e+00
-8.45225692e-01 7.91987836e-01 -6.04034960e-01 -7.88829327e-01
-1.77050903e-01 -8.19561362e-01 -9.12088096e-01 8.64328146e-01
-1.38974285e+00 -1.11177516e+00 -1.07361317e+00 4.10591781e-01
1.25110626e-01 3.24133337e-01 1.08114874e+00 -8.40207562e-02
-4.10183787e-01 7.74027646e-01 -7.14395642e-01 2.93827113e-02
8.70474577e-01 -1.48324788e+00 3.79974633e-01 4.11445796e-01
-1.28000200e-01 6.78678036e-01 8.18597078e-01 -6.68597937e-01
-5.80588222e-01 -9.32519019e-01 7.70663857e-01 -5.91539443e-01
2.67583191e-01 -2.89760470e-01 -1.01422596e+00 5.55613041e-01
5.10698676e-01 -5.69804646e-02 9.71490622e-01 1.78918049e-01
-5.86609542e-01 2.36879691e-01 -1.45312977e+00 8.10672224e-01
6.17494822e-01 -6.33074939e-01 -2.79920012e-01 1.70527071e-01
5.26410580e-01 -4.55137104e-01 -4.38394845e-01 3.94447483e-02
9.21674073e-01 -1.09156120e+00 4.55439508e-01 -1.04508710e+00
7.69124150e-01 -9.48438328e-03 3.51828992e-01 -1.76945400e+00
-6.29996538e-01 -1.97950438e-01 2.54567683e-01 4.31267351e-01
6.13800585e-01 -6.41548574e-01 1.01399124e+00 6.57926559e-01
-1.73032749e-02 -8.64911139e-01 -5.04981458e-01 3.47645991e-02
-2.15693429e-01 -2.22368151e-01 4.87858057e-02 8.53619754e-01
2.95110255e-01 1.65952817e-01 -3.29854749e-02 1.52781397e-01
7.33938217e-02 -4.46447283e-01 7.15571702e-01 -1.16490984e+00
-4.65030015e-01 -4.36342746e-01 -5.43505073e-01 -3.79462481e-01
-4.63264287e-01 -7.82916427e-01 -1.19944038e-02 -1.68480504e+00
4.24228847e-01 5.58031984e-02 -3.42000782e-01 9.63535964e-01
-2.13779807e-01 5.41464388e-01 2.40386620e-01 5.02580889e-02
-3.89574051e-01 -2.28413656e-01 1.45320511e+00 1.62263751e-01
-1.71428211e-02 -2.26429880e-01 -1.12490451e+00 7.01114118e-01
6.44345760e-01 -4.86042649e-01 -1.59297824e-01 -4.75884169e-01
2.91644961e-01 -1.58807725e-01 1.05572605e+00 -1.03236818e+00
4.51413274e-01 -4.62882780e-02 1.00165451e+00 -1.98022388e-02
1.03343263e-01 -7.57262647e-01 2.52191647e-04 1.01091290e+00
-7.61970878e-01 -2.26069521e-02 5.72978735e-01 4.02790397e-01
1.01605013e-01 4.75730635e-02 1.01718009e+00 -1.39087319e-01
-5.29357851e-01 -7.62593448e-02 -6.49760127e-01 -6.80822194e-01
8.67884576e-01 -4.05332118e-01 -4.32244807e-01 -6.33174598e-01
-1.23026371e+00 4.54816036e-02 6.20568871e-01 2.07239017e-01
7.41664827e-01 -1.20480371e+00 -4.17857945e-01 4.59066480e-01
2.65529484e-01 -7.77799487e-01 2.61784852e-01 8.37575436e-01
-4.90357578e-01 2.21917257e-01 -6.33402348e-01 -6.01702988e-01
-1.65221250e+00 1.54653400e-01 6.63885891e-01 -2.40882114e-01
-1.68154046e-01 9.68929768e-01 1.94794565e-01 -3.70702922e-01
3.40696454e-01 -4.16952699e-01 -3.14674675e-01 -2.17873797e-01
5.78433871e-01 2.25372631e-02 2.62608469e-01 2.82319099e-01
-1.38825208e-01 1.83743626e-01 -3.45617205e-01 -2.09631488e-01
9.46798205e-01 1.47535995e-01 2.05411538e-01 3.24022591e-01
9.54286039e-01 -1.84317082e-01 -1.17464495e+00 2.66636461e-01
-4.20080364e-01 -6.98548317e-01 5.27359284e-02 -1.61262119e+00
-1.15038013e+00 1.22960246e+00 9.41133618e-01 5.20531237e-01
1.10013676e+00 -1.37437314e-01 2.70398140e-01 1.25860676e-01
-1.10707454e-01 -8.78401458e-01 3.97498250e-01 1.12519160e-01
1.11362958e+00 -1.23498893e+00 -1.22208275e-01 -5.90110302e-01
-8.13790679e-01 1.30266976e+00 1.05100274e+00 -2.82411188e-01
4.95249033e-01 2.51738936e-01 6.83113515e-01 -5.95566094e-01
-1.13999200e+00 -6.22235946e-02 5.87150753e-01 8.41996491e-01
6.99161172e-01 1.37302145e-01 -3.92035097e-02 3.34175557e-01
-7.81460404e-02 4.17938024e-01 6.18379056e-01 7.60972440e-01
-4.74916011e-01 -6.94211364e-01 -1.72558427e-01 5.18201351e-01
-7.05721527e-02 -5.80365472e-02 -1.20631278e+00 8.10359418e-01
3.05741191e-01 6.56911731e-01 1.88493937e-01 -6.61509514e-01
5.75405002e-01 -1.13915160e-01 4.94420737e-01 -8.76493931e-01
-1.30262148e+00 -3.81268412e-01 7.80332685e-02 -8.91120017e-01
1.81451440e-01 -1.84733674e-01 -1.01454568e+00 -1.89563468e-01
-3.51230055e-01 -1.17016494e-01 7.91989803e-01 8.09459269e-01
4.89637643e-01 9.38057780e-01 4.40297633e-01 -8.00310075e-01
-3.94642591e-01 -9.66875613e-01 -1.79163367e-01 3.77332449e-01
3.87013376e-01 -4.04745281e-01 -3.27065647e-01 -3.75766046e-02] | [8.905305862426758, 5.579439640045166] |
2a9b99bc-792f-43e8-9dc8-3f06354da354 | discovering-customer-service-dialog-system | 2212.12363 | null | https://arxiv.org/abs/2212.12363v1 | https://arxiv.org/pdf/2212.12363v1.pdf | Discovering Customer-Service Dialog System with Semi-Supervised Learning and Coarse-to-Fine Intent Detection | Task-oriented dialog(TOD) aims to assist users in achieving specific goals through multi-turn conversation. Recently, good results have been obtained based on large pre-trained models. However, the labeled-data scarcity hinders the efficient development of TOD systems at scale. In this work, we constructed a weakly supervised dataset based on a teacher/student paradigm that leverages a large collection of unlabelled dialogues. Furthermore, we built a modular dialogue system and integrated coarse-to-fine grained classification for user intent detection. Experiments show that our method can reach the dialog goal with a higher success rate and generate more coherent responses. | ['Zheyu Zhang', 'Anqi Liu', 'Xing Ma', 'Zhitong Yang'] | 2022-12-23 | null | null | null | null | ['intent-detection'] | ['natural-language-processing'] | [-1.87599093e-01 5.33794403e-01 -8.08965489e-02 -7.97891438e-01
-8.62420261e-01 -5.11272490e-01 8.10866475e-01 -1.65422007e-01
-1.58313364e-01 1.08705854e+00 7.32361913e-01 -2.87406802e-01
2.91858107e-01 -4.90953207e-01 2.82962799e-01 -2.60198146e-01
4.73101825e-01 8.16921353e-01 1.11914262e-01 -8.14044237e-01
1.97588310e-01 -6.90687895e-02 -1.09175968e+00 6.18217885e-01
1.10342896e+00 6.62555218e-01 1.89330280e-01 8.68554235e-01
-4.44417417e-01 1.32679260e+00 -8.17184925e-01 -5.08132875e-01
-1.66730687e-01 -7.17031598e-01 -1.40341365e+00 1.85075119e-01
7.88435787e-02 -6.24134481e-01 -1.24356803e-02 5.00826776e-01
6.36280358e-01 2.94222385e-01 5.03557324e-01 -1.18275201e+00
-5.22661209e-01 6.59155786e-01 1.58177450e-01 -2.05827609e-01
7.79243767e-01 2.75971889e-01 1.29226255e+00 -8.55407178e-01
4.00095850e-01 1.42324471e+00 3.94674271e-01 1.09384203e+00
-1.14071631e+00 -2.39916995e-01 -2.92620459e-03 -1.29409492e-01
-6.09163880e-01 -6.28695607e-01 7.92475045e-01 -4.10581827e-01
9.68229890e-01 3.25648814e-01 3.97561908e-01 1.34573817e+00
-3.25760961e-01 1.04895115e+00 1.37540340e+00 -6.02760792e-01
1.06428586e-01 4.89671886e-01 5.07410884e-01 6.90582812e-01
-6.42427802e-01 -4.23152119e-01 -5.55448771e-01 -1.87367439e-01
3.77289623e-01 -1.66792855e-01 -1.37736291e-01 5.29072694e-02
-8.30972314e-01 1.25145936e+00 1.51053905e-01 5.31441629e-01
-1.47951141e-01 -7.23332345e-01 4.88522440e-01 5.71421444e-01
7.33881533e-01 7.76022673e-01 -5.75121045e-01 -6.27964497e-01
-2.99550503e-01 2.01739728e-01 1.51621985e+00 9.39285755e-01
4.82148975e-01 -1.60330713e-01 -5.86603105e-01 1.24373627e+00
2.28274465e-01 3.03252518e-01 5.29567063e-01 -1.00111711e+00
5.69430232e-01 1.03170478e+00 3.15426379e-01 -5.05203903e-01
-5.98279715e-01 1.95049718e-01 -7.32958853e-01 -2.52398431e-01
7.25687683e-01 -5.93562782e-01 -2.94970632e-01 1.74240124e+00
3.78985852e-01 -5.27088046e-01 4.58633333e-01 8.51573050e-01
1.23318636e+00 4.12404239e-01 2.60389566e-01 -2.10786536e-01
1.42856479e+00 -1.34705341e+00 -9.86884296e-01 -6.53487965e-02
9.02660370e-01 -7.10016072e-01 1.51588762e+00 3.64027023e-01
-1.05415165e+00 -6.07309639e-01 -4.09981877e-01 -1.19070634e-01
-1.04514763e-01 8.77940655e-02 7.64752150e-01 8.33246946e-01
-1.09038556e+00 1.23828255e-01 -1.83554277e-01 -5.17001629e-01
2.22742651e-03 2.51434743e-01 -1.96593300e-01 1.59234479e-01
-1.17811441e+00 1.00359786e+00 7.09182322e-02 -2.02452108e-01
-6.54424369e-01 -3.12323540e-01 -8.07419658e-01 -5.95282726e-02
3.41487020e-01 -5.53903401e-01 2.01179028e+00 -4.47914749e-01
-2.23862767e+00 8.27740729e-01 -7.71605894e-02 -3.77059639e-01
3.83520871e-01 -3.47496629e-01 3.95710804e-02 4.79581170e-02
3.24931182e-02 6.92501426e-01 3.67555499e-01 -9.44136977e-01
-7.96494961e-01 -3.34931970e-01 4.48120624e-01 6.48189664e-01
-7.51168907e-01 2.19731435e-01 -1.05570769e-02 -1.16428122e-01
-4.03429538e-01 -9.54058290e-01 -2.66358435e-01 -8.24410379e-01
-3.84655029e-01 -1.02634752e+00 7.71453142e-01 -4.96944696e-01
1.16790771e+00 -1.75297773e+00 1.58350542e-01 -6.08189881e-01
6.45746291e-01 4.81850505e-01 4.41972762e-02 9.27557766e-01
5.23720682e-01 -1.20771825e-02 1.97318316e-01 -5.78258514e-01
1.35311007e-01 3.63008603e-02 -3.71922940e-01 -2.77971774e-01
3.45966802e-03 9.30594027e-01 -1.07045972e+00 -4.89773244e-01
3.46663952e-01 -1.24153800e-01 -5.99182725e-01 1.27096486e+00
-5.14131546e-01 7.78520286e-01 -6.95529044e-01 4.98266399e-01
1.25458211e-01 -3.70256484e-01 2.48120442e-01 2.16912940e-01
-1.19297065e-01 7.93380916e-01 -2.95721412e-01 1.60052669e+00
-7.66624629e-01 4.28603709e-01 3.48025858e-01 -7.67076373e-01
1.21581757e+00 6.90757453e-01 2.99319029e-01 -5.11470199e-01
1.27922654e-01 3.26099284e-02 3.83224227e-02 -6.73123360e-01
7.18322396e-01 -1.88794374e-01 -4.69280183e-01 8.11343789e-01
2.79517770e-01 -5.03227890e-01 5.63194118e-02 4.01841760e-01
9.11280692e-01 -2.31978595e-01 4.30867523e-01 -1.01187743e-01
7.25181282e-01 2.59456724e-01 1.54982924e-01 7.78334618e-01
-4.68453348e-01 2.96272844e-01 4.30709213e-01 -3.49034816e-01
-6.08612299e-01 -4.87938315e-01 1.30341768e-01 1.81918526e+00
-2.06056103e-01 -4.53296661e-01 -9.78070796e-01 -1.09392619e+00
-3.47401708e-01 6.92856431e-01 -2.01444894e-01 3.40495072e-02
-4.94376987e-01 -5.01569033e-01 7.41425157e-01 2.91695416e-01
9.28161979e-01 -1.26956034e+00 -9.44399387e-02 3.13931674e-01
-5.10966718e-01 -1.22690237e+00 -4.54471648e-01 6.79826364e-02
-6.03003025e-01 -1.03005910e+00 -5.67182422e-01 -1.09945488e+00
3.97206128e-01 4.35903668e-01 1.17244768e+00 -2.57201791e-02
2.06050739e-01 4.87383455e-01 -6.93114281e-01 -3.28478396e-01
-7.94796586e-01 3.56555581e-01 1.57066941e-01 -1.66644737e-01
6.14021003e-01 -2.71681607e-01 -3.03520620e-01 6.86654150e-01
-2.79825538e-01 4.28501815e-01 2.90834904e-01 1.12825525e+00
-5.55252552e-01 -3.69700223e-01 1.18965983e+00 -9.87371147e-01
1.60392034e+00 -2.75181413e-01 -1.20817527e-01 2.38250017e-01
-5.84506035e-01 5.75047033e-03 9.05481398e-01 -3.58947039e-01
-1.63890922e+00 -1.41881883e-01 -5.02109051e-01 3.12935591e-01
-5.72020531e-01 1.70678154e-01 -1.09079070e-01 1.36476874e-01
7.26807296e-01 1.19409747e-01 1.44887313e-01 -6.47523344e-01
5.61228335e-01 1.55530369e+00 1.44377425e-01 -9.24661517e-01
3.14668179e-01 -2.42769718e-01 -7.48680413e-01 -9.65352654e-01
-1.23139620e+00 -6.99937224e-01 -6.45505667e-01 -4.52565342e-01
8.31991017e-01 -9.22359228e-01 -9.78093863e-01 5.71522892e-01
-1.19589186e+00 -7.75550842e-01 -1.10826805e-01 2.50683039e-01
-4.85561699e-01 1.58323273e-01 -1.02109802e+00 -1.12759054e+00
-6.85146928e-01 -1.04371107e+00 8.50134790e-01 4.81843293e-01
-6.56970143e-01 -1.24431038e+00 3.05614263e-01 1.04559588e+00
4.97346163e-01 -5.05477428e-01 6.71653688e-01 -1.46574962e+00
-1.12757482e-01 1.15229525e-02 -1.23293623e-01 3.50061238e-01
3.54145378e-01 -4.86997604e-01 -1.11262858e+00 -1.33804947e-01
3.96860838e-01 -1.37743378e+00 1.78444594e-01 -1.70716852e-01
5.48110723e-01 -6.50226116e-01 2.29007583e-02 -2.74986088e-01
4.07290369e-01 1.40354618e-01 8.12963769e-02 6.85330248e-03
5.45879602e-01 1.11764359e+00 8.78489912e-01 5.15875697e-01
8.52491319e-01 7.49160290e-01 -2.59930175e-02 3.22930589e-02
-7.87530467e-02 -3.84247422e-01 3.06268692e-01 9.56040502e-01
7.70380348e-02 -2.58208722e-01 -8.39032531e-01 5.47372282e-01
-2.01978016e+00 -6.86287344e-01 -2.55453169e-01 1.70021713e+00
1.41715121e+00 8.81034359e-02 4.83061612e-01 -2.88850486e-01
5.31616032e-01 2.00232625e-01 -2.27200523e-01 -5.33915520e-01
2.28154823e-01 -8.94370601e-02 -4.26652998e-01 7.50527203e-01
-9.12858367e-01 1.35963631e+00 6.53388691e+00 6.83247983e-01
-7.63116777e-01 2.07812145e-01 7.06894696e-01 3.11337203e-01
-6.20421767e-02 -1.65524766e-01 -1.21382987e+00 1.86777011e-01
1.13115466e+00 -1.87388346e-01 1.19421452e-01 1.09331095e+00
2.91680455e-01 -1.54424896e-02 -9.55912948e-01 5.25350928e-01
-6.43289089e-02 -9.49345052e-01 -2.11933732e-01 6.27687871e-02
6.03664994e-01 -3.11466813e-01 -2.96410680e-01 9.74124908e-01
1.02095914e+00 -8.88739169e-01 4.50550206e-03 2.15343907e-01
3.48611653e-01 -3.41464728e-01 7.10783482e-01 1.07526374e+00
-7.87413478e-01 -1.12020440e-01 -1.53477520e-01 -5.20888805e-01
1.04074672e-01 1.87234521e-01 -1.72749543e+00 2.38629699e-01
2.76521683e-01 3.63436222e-01 -2.52991796e-01 3.31388593e-01
-4.90059227e-01 7.40786195e-01 6.43553399e-03 -6.52667642e-01
3.32997710e-01 -3.04826200e-01 2.89775848e-01 1.24836016e+00
-2.87040532e-01 4.50647980e-01 7.99684405e-01 3.92914653e-01
-2.76565880e-01 4.49346215e-01 -7.31177986e-01 2.70408187e-02
4.66239065e-01 1.62313831e+00 -1.43504381e-01 -3.16937387e-01
-4.87770885e-01 7.58515358e-01 7.85659075e-01 2.00088490e-02
-2.66881138e-01 -3.89004983e-02 4.17012304e-01 -2.74856031e-01
-3.27116907e-01 -1.59229666e-01 -3.28176528e-01 -1.16260087e+00
-3.38310242e-01 -1.21946728e+00 4.52205062e-01 -5.63166022e-01
-1.52357328e+00 8.37755322e-01 -1.30702332e-01 -8.49026501e-01
-8.65157366e-01 -4.02942538e-01 -7.82684505e-01 9.45560336e-01
-1.13839102e+00 -1.21267152e+00 -3.41053128e-01 6.28844380e-01
1.26827145e+00 -4.74263877e-01 1.29447186e+00 -1.62498415e-01
-6.99810147e-01 4.70726311e-01 -3.78357410e-01 3.39557618e-01
1.11392188e+00 -1.52744329e+00 1.00241400e-01 2.25916490e-01
-1.77863836e-01 6.62066102e-01 6.05704725e-01 -5.93717217e-01
-9.80929494e-01 -8.34815919e-01 1.23273361e+00 -6.80095136e-01
6.83798015e-01 -4.35707271e-01 -8.35436702e-01 5.44738531e-01
7.94829011e-01 -6.44031286e-01 9.65492308e-01 5.69939077e-01
-1.04649149e-01 2.92408913e-01 -1.23801649e+00 7.24132538e-01
7.78272927e-01 -6.03898227e-01 -9.74934995e-01 5.74158490e-01
7.79210031e-01 -4.17175561e-01 -1.04589868e+00 2.30332270e-01
2.27177545e-01 -1.03845024e+00 6.72668695e-01 -9.59042311e-01
4.12480235e-01 4.96032476e-01 -2.56665275e-02 -1.47845697e+00
-3.45769785e-02 -9.47757721e-01 -1.81145072e-01 1.59859371e+00
3.20135087e-01 -3.60346973e-01 7.12655246e-01 1.21875656e+00
-3.29739630e-01 -6.42899334e-01 -5.35627246e-01 -4.22727078e-01
9.74883214e-02 6.19088039e-02 3.41255486e-01 8.65808368e-01
8.68301511e-01 1.34464037e+00 -7.47018635e-01 -3.84129405e-01
1.98232889e-01 2.95460969e-01 1.23157609e+00 -1.41848147e+00
-3.10728531e-02 -3.16055149e-01 4.32557076e-01 -1.66118717e+00
4.58559752e-01 -5.87257445e-01 3.24201256e-01 -1.38311183e+00
1.39826357e-01 -5.22803247e-01 3.82542998e-01 6.23938262e-01
-6.26424551e-01 -3.32412869e-02 -6.40166476e-02 3.97712111e-01
-1.03399396e+00 8.37418258e-01 1.21856773e+00 -5.34121180e-03
-6.99374318e-01 4.34819043e-01 -9.52841103e-01 8.12390327e-01
1.15474343e+00 -1.57792196e-01 -4.15090591e-01 5.46166524e-02
-4.93928373e-01 5.49152374e-01 -1.83667243e-01 -5.39373338e-01
2.80154347e-01 -3.17133039e-01 -2.67155051e-01 -4.22154039e-01
5.07418513e-01 -2.89469689e-01 -7.35681236e-01 1.82679787e-01
-1.00649762e+00 -3.93516362e-01 -1.25246629e-01 3.07645231e-01
-2.03162670e-01 -3.84149939e-01 6.67240739e-01 -3.58930200e-01
-4.72629577e-01 -1.00072801e-01 -6.32082582e-01 3.41969639e-01
8.56432796e-01 3.43238115e-01 -4.10329193e-01 -9.14984167e-01
-5.62028348e-01 4.96651173e-01 1.52933940e-01 5.93456268e-01
3.97779375e-01 -1.04602313e+00 -7.84410417e-01 -1.20134957e-01
2.80699134e-01 -1.01953343e-01 -6.15637418e-06 5.88827014e-01
4.88912389e-02 9.05849040e-01 -9.95105654e-02 -5.53531766e-01
-1.62971938e+00 2.68173683e-02 2.09674060e-01 -8.58970344e-01
-4.93849516e-01 8.17425013e-01 5.68578057e-02 -1.13785636e+00
4.26680624e-01 1.74978465e-01 -7.32379556e-01 2.05573931e-01
6.45023942e-01 1.88142896e-01 -6.41724467e-02 -4.32023019e-01
2.12634832e-01 -2.30546236e-01 -2.93502033e-01 -3.82922947e-01
1.16336727e+00 -4.23352510e-01 1.16812624e-02 4.68958825e-01
7.54240036e-01 -6.78391149e-03 -1.20771563e+00 -5.41891932e-01
1.25432625e-01 -3.39904368e-01 -2.64470339e-01 -9.23168182e-01
-3.33552480e-01 7.83554912e-01 -4.10028882e-02 9.16700304e-01
6.43816233e-01 1.38104679e-02 9.67196643e-01 8.99990320e-01
5.11387765e-01 -1.04178154e+00 6.67445421e-01 1.01556349e+00
7.92744815e-01 -1.78998291e+00 -4.95935142e-01 -3.58027488e-01
-1.23461401e+00 1.05228913e+00 1.20032704e+00 4.63060826e-01
1.92592084e-01 9.11502838e-02 5.28671861e-01 -1.50964484e-01
-1.07431650e+00 -3.91356528e-01 1.56860054e-01 5.85495651e-01
8.32729697e-01 -7.11935712e-03 -3.46384227e-01 8.81369412e-01
-3.65262538e-01 -1.24993816e-01 4.96545553e-01 8.88422310e-01
-7.38294542e-01 -1.49601448e+00 -2.08381072e-01 2.70012796e-01
-2.29771525e-01 -2.01850623e-01 -1.07022214e+00 4.64887530e-01
-6.91255212e-01 1.81759214e+00 -5.07566512e-01 -5.95329821e-01
4.49668586e-01 5.54591000e-01 7.96449631e-02 -1.11997795e+00
-1.11451995e+00 -1.11500166e-01 8.52685034e-01 -2.46296287e-01
-3.76895368e-01 -2.38824069e-01 -1.07593465e+00 -2.07814530e-01
-3.61053258e-01 7.41288662e-01 2.85615087e-01 1.08453691e+00
2.88253129e-01 2.33823344e-01 1.19477880e+00 -5.76103866e-01
-9.81202006e-01 -1.61042821e+00 -2.54742742e-01 4.78954285e-01
1.08958617e-01 -2.82711774e-01 -3.08975160e-01 -1.29490629e-01] | [12.834452629089355, 7.983392238616943] |
91349a52-2459-4c3d-8ca8-75321523f52b | can-bert-eat-rucola-topological-data-analysis | 2304.01680 | null | https://arxiv.org/abs/2304.01680v1 | https://arxiv.org/pdf/2304.01680v1.pdf | Can BERT eat RuCoLA? Topological Data Analysis to Explain | This paper investigates how Transformer language models (LMs) fine-tuned for acceptability classification capture linguistic features. Our approach uses the best practices of topological data analysis (TDA) in NLP: we construct directed attention graphs from attention matrices, derive topological features from them, and feed them to linear classifiers. We introduce two novel features, chordality, and the matching number, and show that TDA-based classifiers outperform fine-tuning baselines. We experiment with two datasets, CoLA and RuCoLA in English and Russian, typologically different languages. On top of that, we propose several black-box introspection techniques aimed at detecting changes in the attention mode of the LMs during fine-tuning, defining the LM's prediction confidences, and associating individual heads with fine-grained grammar phenomena. Our results contribute to understanding the behavior of monolingual LMs in the acceptability classification task, provide insights into the functional roles of attention heads, and highlight the advantages of TDA-based approaches for analyzing LMs. We release the code and the experimental results for further uptake. | ['Ekaterina Artemova', 'Irina Piontkovskaya', 'Irina Proskurina'] | 2023-04-04 | null | null | null | null | ['topological-data-analysis', 'linguistic-acceptability'] | ['graphs', 'natural-language-processing'] | [-1.78257167e-01 2.92454541e-01 -1.96849838e-01 -4.40942228e-01
-8.66248965e-01 -1.01310146e+00 5.84120512e-01 5.88576436e-01
-2.35681295e-01 2.83580720e-01 5.73539615e-01 -6.72552228e-01
-4.45375413e-01 -7.27016926e-01 -7.98173487e-01 -3.93971473e-01
-3.60388011e-01 7.53556848e-01 1.93514097e-02 -4.86848533e-01
7.63623536e-01 3.34399939e-01 -1.41655231e+00 6.27210915e-01
1.14442563e+00 8.47842932e-01 9.30274725e-02 7.19226897e-01
-3.28076184e-01 6.02538407e-01 -4.95593220e-01 -6.04170561e-01
-2.05102459e-01 -2.00146034e-01 -1.20688164e+00 -6.44543648e-01
9.14883971e-01 3.51292998e-01 -1.21566191e-01 8.05742145e-01
1.58158600e-01 1.48429543e-01 1.14344954e+00 -7.61103392e-01
-1.22696853e+00 1.28102160e+00 -3.97834510e-01 6.61474943e-01
4.26621586e-01 -2.37508453e-02 1.59776330e+00 -1.00195444e+00
7.63252318e-01 1.70686460e+00 9.44483161e-01 2.43187144e-01
-1.28377259e+00 -1.18456259e-01 6.30775094e-01 3.27940375e-01
-1.25375235e+00 -5.33171415e-01 6.48471892e-01 -7.35295475e-01
1.19384789e+00 3.63631368e-01 6.11679554e-01 9.30664062e-01
2.64124215e-01 8.05736244e-01 1.16964161e+00 -7.25240707e-01
5.97834308e-03 -1.30424909e-02 6.59146249e-01 9.10323620e-01
8.98713395e-02 2.91043743e-02 -7.78246462e-01 2.98029836e-02
3.60565245e-01 -8.40724170e-01 -1.83484539e-01 -4.21618611e-01
-1.23296225e+00 9.43439126e-01 5.48059881e-01 5.66210330e-01
3.15897353e-02 -1.27291918e-01 5.22911191e-01 4.53900129e-01
5.46810687e-01 7.24671900e-01 -6.36814296e-01 -1.65697783e-01
-5.44454038e-01 -8.73495918e-03 4.96190041e-01 8.34021926e-01
4.40730274e-01 -2.32805192e-01 -2.46368110e-01 8.95639420e-01
2.51641363e-01 3.17711323e-01 6.39756143e-01 -5.45171797e-01
7.03159273e-01 6.25116825e-01 -5.18625379e-01 -1.06858623e+00
-7.45952249e-01 -3.98679137e-01 -2.34107971e-01 -1.13144778e-01
7.84249663e-01 2.63865203e-01 -4.95429844e-01 1.97079706e+00
-5.55024967e-02 -5.86584926e-01 -3.42398703e-01 6.23449922e-01
6.06386662e-01 4.68816161e-01 9.14523154e-02 2.08114423e-02
1.23883367e+00 -7.09087491e-01 -2.99302667e-01 -1.43050522e-01
1.15745640e+00 -4.74734396e-01 1.99748182e+00 3.60004455e-01
-1.15423882e+00 -7.97863781e-01 -8.37674618e-01 -5.16196012e-01
-7.71390438e-01 1.77191183e-01 5.56432843e-01 6.04601979e-01
-1.17103422e+00 7.75588930e-01 -6.50423944e-01 -4.84228194e-01
2.73870915e-01 4.45626199e-01 -1.85021520e-01 6.40934050e-01
-1.13914335e+00 1.08272696e+00 1.24747001e-01 5.67427874e-02
-2.53361136e-01 -6.35027587e-01 -9.94926810e-01 1.22986645e-01
-1.01051621e-01 -3.06362748e-01 1.00156450e+00 -7.99368203e-01
-1.32037699e+00 1.18115389e+00 -7.70272762e-02 -2.14234158e-01
1.57301798e-01 -1.38041049e-01 -1.39798746e-01 -3.34101290e-01
1.93587869e-01 5.12700140e-01 5.20046830e-01 -7.74979174e-01
-4.51151490e-01 -5.48907697e-01 1.02956429e-01 1.36014357e-01
-4.36439842e-01 -3.27796079e-02 -1.37117386e-01 -7.27306426e-01
-1.88317094e-02 -8.79116118e-01 2.82294750e-01 -5.68348169e-01
-5.91735899e-01 -5.47106564e-01 3.30860615e-01 -7.24074960e-01
1.65312541e+00 -2.23671460e+00 4.86727983e-01 4.77639258e-01
3.86257112e-01 -1.15616523e-01 -2.71573305e-01 2.69073844e-01
3.71344984e-02 5.06554425e-01 -1.16030492e-01 -1.77225575e-01
2.96213359e-01 -2.95897007e-01 -3.15782934e-01 3.96201462e-01
1.34689763e-01 1.26059425e+00 -8.80902350e-01 -5.10608852e-01
2.28820398e-01 2.56066948e-01 -6.64023757e-01 -3.04590791e-01
-2.96438694e-01 4.03532058e-01 -2.46530786e-01 5.22280455e-01
1.54443353e-01 -8.89532343e-02 5.23178816e-01 -2.25088686e-01
-2.63119608e-01 9.30376709e-01 -6.31025910e-01 1.36821175e+00
-5.91622829e-01 1.10701013e+00 -1.66859701e-01 -7.78649509e-01
8.64782751e-01 -4.45302606e-01 -3.56147379e-01 -9.48816180e-01
2.10812837e-01 1.54594973e-01 3.33586395e-01 -3.44609231e-01
5.31856716e-01 4.70124185e-01 -6.28475666e-01 3.77296925e-01
2.94678044e-02 2.25324072e-02 3.19813102e-01 1.05214216e-01
7.82151043e-01 2.45183930e-01 3.79299551e-01 -9.27236497e-01
6.48167491e-01 -1.11058950e-01 2.77708083e-01 7.62592733e-01
-1.28368467e-01 2.45038599e-01 9.59971189e-01 -5.48362970e-01
-9.84849632e-01 -1.01159012e+00 -3.40295941e-01 2.05896115e+00
-2.11344153e-01 -7.40852654e-01 -9.20742214e-01 -7.10489690e-01
1.05709508e-01 6.62580192e-01 -1.19074094e+00 -1.66525498e-01
-9.98216331e-01 -4.57765818e-01 7.07930803e-01 7.45123208e-01
-1.06362537e-01 -1.35888755e+00 -3.14312875e-01 -1.45309240e-01
-1.76030517e-01 -7.57503152e-01 -5.63328564e-01 2.12104768e-01
-6.63939595e-01 -1.04170203e+00 -2.28737921e-01 -1.03331923e+00
4.92081881e-01 -3.80758494e-01 1.43598485e+00 1.87471539e-01
-6.37709796e-02 1.36677427e-02 -1.08274348e-01 -1.51545227e-01
-4.72347707e-01 8.74072611e-01 2.62242466e-01 -3.11072290e-01
2.90420681e-01 -3.53734046e-01 -1.72216251e-01 1.58248376e-02
-1.40491605e-01 -2.07193077e-01 4.32758629e-01 6.02504671e-01
4.11377698e-01 -5.89811444e-01 4.04194325e-01 -1.17773747e+00
1.11300731e+00 -2.38445401e-01 -6.99846148e-01 4.68711197e-01
-5.79759896e-01 5.31200886e-01 8.29372048e-01 -1.36310667e-01
-8.13862681e-01 -1.43094242e-01 -4.36623357e-02 3.64512354e-02
1.17615936e-02 6.94488466e-01 -2.43247956e-01 -2.36272842e-01
9.08770442e-01 -2.01402113e-01 -3.61450732e-01 -6.66049361e-01
7.15618193e-01 6.09066427e-01 5.69359243e-01 -9.12457049e-01
2.37406582e-01 -4.39763181e-02 -2.14868933e-01 -9.58727598e-01
-7.15206087e-01 1.56270154e-02 -1.09358740e+00 4.47369181e-02
7.13438213e-01 -2.85715252e-01 -9.24426377e-01 1.72470883e-01
-8.53181601e-01 -6.18719161e-01 -1.18477598e-01 2.11399376e-01
-6.57952130e-01 2.84499645e-01 -8.72555852e-01 -4.95555997e-01
-1.73944011e-01 -9.96254265e-01 1.21644914e+00 -1.09471254e-01
-7.40031302e-01 -1.37913203e+00 3.42818230e-01 1.29255235e-01
3.04761261e-01 -5.87664470e-02 1.76369560e+00 -6.56150341e-01
-3.18293929e-01 4.18027133e-01 -6.96807206e-02 -3.78296018e-01
-3.42529237e-01 2.91973770e-01 -9.88829315e-01 -5.75028695e-02
-6.42760098e-01 -3.22080374e-01 1.03441882e+00 6.98070884e-01
1.26586807e+00 -2.47621283e-01 -4.49448735e-01 8.11640322e-01
8.19897950e-01 -1.26221240e-01 1.26986384e-01 5.27529597e-01
8.35500240e-01 8.59183133e-01 4.52220559e-01 -7.62906000e-02
6.79814935e-01 8.25265050e-01 3.36575583e-02 2.63412088e-01
-1.44312352e-01 -3.50027829e-01 4.88254637e-01 1.21905291e+00
-8.99560973e-02 -1.78736925e-01 -1.19918084e+00 6.44382954e-01
-1.59707046e+00 -5.66746414e-01 -2.43783370e-01 2.07227540e+00
6.84117794e-01 3.63883823e-01 3.71511638e-01 2.30904311e-01
6.16559982e-01 1.52626112e-01 -3.07200938e-01 -1.05691254e+00
-2.71945685e-01 4.03926492e-01 2.69814730e-01 9.94129777e-01
-1.07302022e+00 1.50330722e+00 6.71333647e+00 5.71001291e-01
-1.19331813e+00 -2.36653551e-01 5.47634780e-01 5.39452545e-02
-4.16231692e-01 -1.49955571e-01 -7.97099054e-01 1.67732328e-01
1.19510210e+00 1.15105771e-01 6.17900968e-01 5.96412301e-01
-3.83839197e-02 3.31684619e-01 -1.39832580e+00 5.98450124e-01
3.57167460e-02 -1.29972637e+00 1.07252948e-01 7.82655850e-02
3.80204022e-01 2.49361604e-01 3.96444827e-01 5.18377125e-01
1.87162593e-01 -1.13613772e+00 1.05760348e+00 6.26536846e-01
9.85088468e-01 -6.22546673e-01 2.96070814e-01 6.24791440e-03
-1.34485281e+00 -2.76586890e-01 -2.82828957e-01 -2.22486615e-01
-1.95838749e-01 1.10329956e-01 -6.31365001e-01 1.20965548e-01
7.48362243e-01 6.00825787e-01 -1.21952701e+00 5.94844878e-01
-2.51490861e-01 7.80150294e-01 -1.09225772e-01 -3.85992855e-01
8.02343562e-02 6.92469883e-04 5.27971804e-01 1.49774063e+00
-3.88049930e-02 -4.25894529e-01 -1.71019256e-01 9.38399911e-01
1.41331460e-02 5.96792698e-01 -4.78944480e-01 -2.04025581e-01
6.17141962e-01 1.05398989e+00 -8.56409013e-01 -1.57585010e-01
-1.24964431e-01 6.82362020e-01 1.12326515e+00 1.25690475e-01
-6.20197475e-01 -5.63348055e-01 5.45336843e-01 1.23797953e-01
3.42382610e-01 -2.23123819e-01 -6.53562069e-01 -1.13928854e+00
2.06018426e-02 -8.54287326e-01 4.99576181e-01 -6.90770984e-01
-1.29899979e+00 8.43614817e-01 -1.72459811e-01 -6.19413376e-01
-3.38491738e-01 -9.39931631e-01 -5.66657662e-01 7.14328170e-01
-8.88720930e-01 -1.23587537e+00 -4.91050743e-02 2.50837356e-01
5.86956561e-01 -1.02949306e-01 8.43914986e-01 -1.16822928e-01
-5.98037362e-01 9.18035150e-01 1.09536864e-01 3.19184035e-01
5.36491036e-01 -1.65716577e+00 1.22009301e+00 4.31430638e-01
6.51203930e-01 9.22327757e-01 3.83740455e-01 -5.67669988e-01
-9.23222959e-01 -7.54720628e-01 1.26346624e+00 -9.33887720e-01
8.76918852e-01 -9.61784482e-01 -9.80367243e-01 8.73879373e-01
1.44332228e-02 -2.79377967e-01 3.94252807e-01 9.87496555e-01
-4.99922216e-01 1.95117861e-01 -5.43390095e-01 5.17228007e-01
1.48736489e+00 -9.03270006e-01 -9.36636150e-01 1.32292032e-01
6.20276749e-01 -3.49880815e-01 -8.27194333e-01 1.49548173e-01
6.33508146e-01 -1.04370451e+00 7.87285328e-01 -8.95618916e-01
2.21588865e-01 8.80141854e-02 5.45834750e-02 -1.63508737e+00
-8.54853392e-01 -4.22497153e-01 -1.23127550e-01 1.03277838e+00
8.98818314e-01 -5.76455832e-01 4.95795757e-01 2.28514038e-02
-4.16853249e-01 -7.75989771e-01 -8.51212084e-01 -5.71407676e-01
7.67157137e-01 -4.35872227e-01 6.06270969e-01 9.59962249e-01
5.29043198e-01 8.00990403e-01 2.89827764e-01 -9.14482921e-02
3.17833751e-01 1.90459415e-01 2.58049756e-01 -1.60587561e+00
-2.01430336e-01 -1.00474584e+00 -3.16939116e-01 -6.64372742e-01
6.09033048e-01 -1.30900097e+00 -2.27731228e-01 -1.19977522e+00
-1.30444258e-01 -5.00196636e-01 -2.43931398e-01 2.89757997e-01
-1.00673206e-01 4.81672883e-02 2.71408826e-01 1.50040999e-01
-7.70396590e-01 1.17576115e-01 8.33723664e-01 -1.81735665e-01
-3.89517695e-01 -3.73857886e-01 -8.57222676e-01 8.97862732e-01
8.76293421e-01 -7.94016644e-02 -2.70203382e-01 -6.43382370e-01
8.10691476e-01 -3.46613139e-01 4.78077121e-02 -8.72740149e-01
6.72625378e-02 3.61265391e-02 4.09809470e-01 -4.32517767e-01
-9.36609283e-02 -8.73481706e-02 -6.55814946e-01 2.09169120e-01
-8.18609357e-01 4.33225930e-01 3.61899137e-01 2.00968072e-01
7.94489086e-02 9.84384194e-02 6.35525405e-01 3.83505225e-02
-6.32575870e-01 -1.16823189e-01 -4.56866950e-01 7.59063005e-01
5.43928325e-01 9.88358818e-03 -4.56347734e-01 -1.62094831e-01
-1.08521152e+00 -8.96608923e-03 4.73349094e-01 7.67592669e-01
1.34238064e-01 -1.14080215e+00 -6.45408690e-01 4.98827726e-01
3.87127161e-01 -4.99516845e-01 -5.01940288e-02 8.20552647e-01
-5.32504678e-01 8.61982763e-01 -2.07051292e-01 -4.78789866e-01
-1.11573863e+00 6.53852284e-01 4.19754654e-01 -1.63154975e-01
-4.25535440e-01 1.12740290e+00 5.02299428e-01 -8.95103693e-01
2.17867583e-01 -1.01828265e+00 -4.21876043e-01 1.56068519e-01
2.46883929e-01 4.41781044e-01 3.70273322e-01 -8.70218158e-01
-5.31583071e-01 9.71859157e-01 -1.50586754e-01 -1.40615076e-01
1.01662886e+00 -4.67873216e-01 -1.13430932e-01 8.85399461e-01
9.33332801e-01 6.26868248e-01 -7.58132279e-01 -4.09651082e-03
5.49604893e-01 6.37481641e-03 -1.80219382e-01 -9.09747839e-01
-6.19260490e-01 1.08559418e+00 1.92962319e-01 4.04794067e-01
6.54574454e-01 3.16122353e-01 3.16151589e-01 3.43649179e-01
9.54290256e-02 -1.26985645e+00 -3.38180393e-01 1.00474155e+00
1.03336489e+00 -8.99949849e-01 -4.70454901e-01 -3.00075173e-01
-5.09902775e-01 1.17090642e+00 6.42009258e-01 -2.30983496e-02
3.83830100e-01 1.26562551e-01 1.76859237e-02 -3.85951698e-01
-9.13826048e-01 -1.61988005e-01 7.64538646e-01 7.32220769e-01
9.01790440e-01 2.60856390e-01 -2.81078637e-01 6.78118825e-01
-1.02058518e+00 -6.21996760e-01 3.01962525e-01 3.90511006e-01
-4.41655070e-01 -8.34254146e-01 -3.03517789e-01 5.53802907e-01
-3.19283485e-01 -6.22490287e-01 -9.30453420e-01 9.46480691e-01
1.12168379e-01 6.09974086e-01 3.49482596e-01 -5.52119911e-01
3.20498407e-01 3.17054629e-01 6.46244824e-01 -5.36433458e-01
-7.50753164e-01 -2.77919948e-01 3.14108580e-01 -5.72040915e-01
4.47475582e-01 -9.15727437e-01 -1.22795236e+00 -3.77421558e-01
-1.73100993e-01 1.39028877e-01 4.35947478e-01 8.95746350e-01
4.09356654e-01 4.82211411e-01 1.83004960e-01 -5.88680327e-01
-3.11247975e-01 -1.12712765e+00 -3.11596930e-01 3.62218350e-01
3.87032658e-01 -6.89275682e-01 -4.10966426e-01 -2.56126434e-01] | [10.812097549438477, 9.584216117858887] |
266e4510-ad9b-4096-b797-7907c1aed85b | h2o-two-hands-manipulating-objects-for-first | 2104.11181 | null | https://arxiv.org/abs/2104.11181v2 | https://arxiv.org/pdf/2104.11181v2.pdf | H2O: Two Hands Manipulating Objects for First Person Interaction Recognition | We present a comprehensive framework for egocentric interaction recognition using markerless 3D annotations of two hands manipulating objects. To this end, we propose a method to create a unified dataset for egocentric 3D interaction recognition. Our method produces annotations of the 3D pose of two hands and the 6D pose of the manipulated objects, along with their interaction labels for each frame. Our dataset, called H2O (2 Hands and Objects), provides synchronized multi-view RGB-D images, interaction labels, object classes, ground-truth 3D poses for left & right hands, 6D object poses, ground-truth camera poses, object meshes and scene point clouds. To the best of our knowledge, this is the first benchmark that enables the study of first-person actions with the use of the pose of both left and right hands manipulating objects and presents an unprecedented level of detail for egocentric 3D interaction recognition. We further propose the method to predict interaction classes by estimating the 3D pose of two hands and the 6D pose of the manipulated objects, jointly from RGB images. Our method models both inter- and intra-dependencies between both hands and objects by learning the topology of a graph convolutional network that predicts interactions. We show that our method facilitated by this dataset establishes a strong baseline for joint hand-object pose estimation and achieves state-of-the-art accuracy for first person interaction recognition. | ['Marc Pollefeys', 'Federica Bogo', 'Jan Stuhmer', 'Bugra Tekin', 'Taein Kwon'] | 2021-04-22 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Kwon_H2O_Two_Hands_Manipulating_Objects_for_First_Person_Interaction_Recognition_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Kwon_H2O_Two_Hands_Manipulating_Objects_for_First_Person_Interaction_Recognition_ICCV_2021_paper.pdf | iccv-2021-1 | ['hand-object-pose'] | ['computer-vision'] | [-7.75463656e-02 -4.96133715e-02 6.01276346e-02 -3.21093261e-01
-2.71149695e-01 -7.13063896e-01 7.73374498e-01 -4.67632502e-01
-1.67758986e-01 1.90470349e-02 4.30872679e-01 3.24972004e-01
-1.13547429e-01 -1.36395782e-01 -7.09813714e-01 -2.86001861e-01
-1.00509338e-01 1.22378552e+00 1.35371417e-01 -2.06010640e-01
1.34659573e-01 9.83650744e-01 -1.64257765e+00 3.09951216e-01
-6.17055595e-02 9.63046551e-01 -1.13022011e-02 1.12839425e+00
5.48287094e-01 6.86153889e-01 -5.75933695e-01 -2.63286680e-01
3.53955716e-01 -7.44274259e-02 -8.89355600e-01 4.40603673e-01
8.03944886e-01 -7.64246881e-01 -7.49883294e-01 3.07236195e-01
6.97229505e-01 3.63872409e-01 7.19396770e-01 -1.56377375e+00
-1.86200872e-01 8.29490051e-02 -5.34724593e-01 -2.00122088e-01
1.32379341e+00 3.44199121e-01 8.09696078e-01 -9.19309139e-01
1.15233445e+00 1.41450012e+00 4.69286531e-01 6.83129966e-01
-1.11186337e+00 -1.88380674e-01 3.70838732e-01 1.54786734e-02
-1.49974275e+00 -4.83612597e-01 9.17296290e-01 -8.78038108e-01
1.42901683e+00 3.05380106e-01 1.01270604e+00 1.59673798e+00
-5.39762042e-02 8.02819669e-01 6.58741593e-01 -6.18621171e-01
2.51189023e-02 -1.15097545e-01 7.26850331e-02 7.62858570e-01
-2.40827098e-01 -2.55777538e-01 -8.17021370e-01 5.83137944e-02
1.07567692e+00 3.01346958e-01 -2.37232700e-01 -1.30765641e+00
-1.46261990e+00 -2.64233183e-02 4.06831712e-01 -9.43862949e-04
-2.52574593e-01 4.48262572e-01 2.45564848e-01 -2.90483892e-01
3.62734228e-01 -5.47022186e-02 -7.63154507e-01 -4.99532104e-01
8.57843645e-03 5.61877489e-01 8.14374149e-01 1.48401928e+00
2.21945882e-01 -5.74059129e-01 -3.46929878e-01 1.96595296e-01
2.74448425e-01 4.62528288e-01 -2.39373520e-01 -1.09995222e+00
6.65281236e-01 9.18110251e-01 5.05439878e-01 -8.79254758e-01
-6.71383500e-01 1.67323843e-01 -1.78936854e-01 2.75802374e-01
6.96001112e-01 1.84289798e-01 -7.60686278e-01 1.67391145e+00
6.22914135e-01 -1.46685645e-01 -2.77152717e-01 1.09054375e+00
8.20794106e-01 -2.21484080e-01 -2.50554532e-01 3.10406864e-01
1.46135139e+00 -7.79047370e-01 -6.28262997e-01 -2.46482180e-03
6.73135936e-01 -4.36209142e-01 1.02693486e+00 2.53970474e-01
-1.17563260e+00 -6.65451944e-01 -6.37149870e-01 -3.67267340e-01
-5.13231456e-01 4.10224497e-01 7.56627679e-01 4.43158507e-01
-6.89770103e-01 3.74228865e-01 -1.04589713e+00 -7.18044639e-01
2.51495242e-01 6.65098071e-01 -8.60915422e-01 2.41678268e-01
-4.34271783e-01 1.09927917e+00 1.29813731e-01 2.97957987e-01
-9.86466169e-01 -5.96573889e-01 -9.81027782e-01 -3.62737209e-01
5.35397291e-01 -9.21789467e-01 1.11271620e+00 -2.81874657e-01
-1.52814436e+00 1.39485109e+00 -6.04080632e-02 2.91201383e-01
8.09408903e-01 -7.61981189e-01 6.42014891e-02 -4.75836322e-02
-1.42830372e-01 6.23160303e-01 5.86193621e-01 -1.53371084e+00
-2.28163302e-01 -1.12745523e+00 4.91369933e-01 5.29180825e-01
1.57869726e-01 -1.61340222e-01 -8.42214406e-01 -2.75884509e-01
1.96319789e-01 -1.29295909e+00 2.15285331e-01 2.35644639e-01
-6.96062207e-01 -3.71432930e-01 9.05213714e-01 -6.05061233e-01
4.53414977e-01 -2.15928245e+00 9.35723960e-01 1.21083938e-01
3.76450896e-01 -9.17233676e-02 7.04769343e-02 2.47572988e-01
-2.08706394e-01 -3.32330495e-01 3.81140083e-01 -8.84416819e-01
4.48868603e-01 1.30751461e-01 1.76384449e-02 7.49350905e-01
-1.83824793e-01 1.17314053e+00 -8.27788115e-01 -2.03006864e-01
6.87032580e-01 1.04558861e+00 -4.83454198e-01 7.58561075e-01
-5.99814095e-02 9.36659098e-01 -2.54982710e-01 6.30055845e-01
4.12934005e-01 -1.46853164e-01 1.25783354e-01 -5.08478999e-01
2.68193126e-01 1.40719637e-01 -1.44371951e+00 2.29187918e+00
-3.15516859e-01 3.51183087e-01 -4.87585738e-02 -3.44328135e-01
5.10166109e-01 6.01111114e-01 8.50283325e-01 -1.15210973e-01
4.91667300e-01 -1.52952328e-01 -5.03946900e-01 -6.54930651e-01
1.42253965e-01 3.98198605e-01 -2.72027373e-01 4.36820567e-01
3.44955862e-01 -4.98206317e-01 -2.56600648e-01 4.52043973e-02
9.39458728e-01 9.11781788e-01 1.57887280e-01 2.54741609e-01
4.31333095e-01 -3.79396588e-01 -1.68221548e-01 4.18091685e-01
-1.42330900e-01 6.82862520e-01 3.46520722e-01 -6.62908733e-01
-1.03848386e+00 -1.44165754e+00 2.78148741e-01 1.16561913e+00
2.09488481e-01 -4.68844622e-01 -7.90029347e-01 -8.37595046e-01
3.35936785e-01 3.30099732e-01 -7.87543535e-01 -7.42398053e-02
-6.55421793e-01 2.30258361e-01 2.67318308e-01 8.75284016e-01
2.62700379e-01 -9.25882161e-01 -8.82052004e-01 -5.97358823e-01
-1.54946029e-01 -1.56685746e+00 -6.63069963e-01 -3.46803814e-02
-5.65854669e-01 -1.46547592e+00 -5.79402983e-01 -6.66940391e-01
9.22823310e-01 7.85121396e-02 1.10046053e+00 -3.73994768e-01
-4.92533237e-01 1.36988080e+00 -2.68427938e-01 -1.96530074e-01
2.57865220e-01 -1.55854017e-01 4.99673784e-01 6.23444282e-02
3.16296190e-01 -6.71877563e-01 -6.11519158e-01 5.14479876e-01
-3.19976620e-02 7.16035515e-02 -1.21440142e-01 2.75739014e-01
4.62785363e-01 -5.81866443e-01 -4.74631369e-01 -4.98276412e-01
9.13754925e-02 2.69003838e-01 -2.86378145e-01 2.42190793e-01
3.98420304e-01 -4.07116205e-01 7.29997382e-02 -4.41007227e-01
-9.41079497e-01 7.55145967e-01 3.64956170e-01 -7.63872862e-01
-7.85678506e-01 -3.37997496e-01 -5.62461078e-01 -9.82802659e-02
7.13236153e-01 -3.18624973e-01 -2.50968903e-01 -6.07664049e-01
8.60117793e-01 5.78212678e-01 7.86680520e-01 -6.46377087e-01
6.28214240e-01 8.53881240e-01 3.22889328e-01 -5.80119729e-01
-7.58618355e-01 -6.93965435e-01 -1.88846457e+00 -5.72425663e-01
1.19352567e+00 -8.87579858e-01 -1.64027953e+00 6.65495098e-01
-1.44980419e+00 -4.36898172e-01 -3.39011192e-01 6.10633850e-01
-1.27208054e+00 1.10957079e-01 -2.18467221e-01 -1.03756380e+00
3.93582061e-02 -1.19579673e+00 1.84734440e+00 -4.34379131e-01
-6.55179620e-01 -7.13737905e-01 -1.20827690e-01 6.14067614e-01
-3.32920313e-01 8.00103068e-01 3.23302090e-01 -2.70465732e-01
-5.30592799e-01 -5.53428471e-01 2.19136029e-01 -1.69886395e-01
2.06121191e-01 -2.02396035e-01 -1.05695426e+00 -2.42579237e-01
-2.67199099e-01 -3.38331610e-01 2.39324346e-01 3.67367536e-01
9.66732800e-01 9.39218253e-02 -4.92202371e-01 7.41635919e-01
6.87859118e-01 1.72499362e-02 3.64252597e-01 -1.05150118e-01
1.42512822e+00 8.65220428e-01 4.78921950e-01 6.87738121e-01
4.76646721e-01 1.43589306e+00 5.10788798e-01 2.50012845e-01
-1.35355949e-01 -2.34216973e-01 2.16302857e-01 3.46335649e-01
-8.06142449e-01 -2.40101218e-01 -1.02213299e+00 2.21405268e-01
-1.84124815e+00 -6.29642129e-01 -1.42575055e-01 2.27673459e+00
1.81162193e-01 1.77163631e-02 5.30114889e-01 1.47760719e-01
4.52282697e-01 -5.05417250e-02 -6.96153522e-01 4.20775078e-02
3.75048846e-01 1.14065848e-01 1.86602235e-01 4.81700778e-01
-1.20757973e+00 1.02117443e+00 5.93991852e+00 -1.81334019e-01
-4.48675990e-01 -1.05757922e-01 -1.51933625e-01 -5.64627290e-01
5.70097029e-01 -9.72182527e-02 -7.21119225e-01 -1.23673081e-01
4.63420868e-01 6.06112659e-01 5.93858242e-01 8.27757299e-01
-2.02427823e-02 3.84229384e-02 -1.89345467e+00 1.43735933e+00
3.96584779e-01 -8.71851504e-01 -1.16565086e-01 9.32915732e-02
5.57719529e-01 -2.24076480e-01 -4.46550548e-02 -8.34666714e-02
1.37906238e-01 -7.81520545e-01 9.94388759e-01 9.58456933e-01
8.55762720e-01 -5.42150259e-01 4.17610377e-01 4.18018430e-01
-1.44963980e+00 -3.72654162e-02 4.13231492e-01 -3.65565419e-01
3.98738623e-01 -1.82009131e-01 -6.46641076e-01 3.97107154e-01
8.42288852e-01 8.78016412e-01 -5.12333632e-01 2.91806668e-01
-3.21623951e-01 -4.69346642e-01 -3.47286850e-01 3.07699561e-01
-3.75613689e-01 -2.10169354e-04 6.43977046e-01 7.66321540e-01
-1.54712856e-01 2.78555602e-01 3.21087629e-01 6.99089527e-01
8.89676586e-02 -3.42198551e-01 -8.36447597e-01 1.09802909e-01
6.19439557e-02 9.20374453e-01 -6.97126210e-01 -2.60500640e-01
-2.54160278e-02 1.40951216e+00 4.07540113e-01 4.43347305e-01
-8.67498696e-01 -1.25002965e-01 8.38255644e-01 1.53773054e-01
1.60605490e-01 -8.15644264e-01 -5.38797565e-02 -1.36210990e+00
4.08432245e-01 -6.53270900e-01 1.89412817e-01 -1.15624976e+00
-9.99451101e-01 2.62545228e-01 3.93045455e-01 -1.09184515e+00
-3.23995918e-01 -1.05841517e+00 -6.28700480e-02 7.53834546e-01
-3.45370114e-01 -1.69387031e+00 -1.04283595e+00 8.76089454e-01
4.25112963e-01 -1.14022888e-01 9.97624993e-01 -1.35490343e-01
-3.01138550e-01 4.23285455e-01 -5.71069181e-01 2.91300356e-01
5.97138882e-01 -1.37623143e+00 6.15592182e-01 3.19532529e-02
5.23672223e-01 7.56490111e-01 4.43052649e-01 -5.21010876e-01
-2.04221940e+00 -4.54635024e-01 6.42412484e-01 -1.83618462e+00
1.78601533e-01 -1.25464487e+00 -4.71438020e-02 1.39939141e+00
-4.98474002e-01 2.34847695e-01 3.67615342e-01 4.22118157e-01
-3.89098823e-01 5.35830557e-02 -9.66337681e-01 7.24614978e-01
1.93158674e+00 -9.33676243e-01 -5.53353250e-01 8.66008818e-01
4.23643380e-01 -1.22503948e+00 -9.36678290e-01 2.63599664e-01
1.04341221e+00 -9.44737077e-01 1.42697465e+00 -7.24133790e-01
4.03397009e-02 -3.86681497e-01 -4.17066097e-01 -8.12507987e-01
-3.55094858e-02 -5.81372559e-01 -6.73361659e-01 8.74617100e-01
-3.89833421e-01 -7.68290609e-02 9.30193603e-01 8.79541278e-01
1.24231681e-01 -5.92128873e-01 -9.02510405e-01 -6.36141717e-01
-5.88067770e-01 -6.15936160e-01 5.14078736e-01 5.04894376e-01
2.50677109e-01 2.14827269e-01 -3.22494417e-01 1.52836636e-01
5.63555181e-01 -5.86258918e-02 1.71476460e+00 -1.28562939e+00
-1.46606401e-01 -3.58958066e-01 -1.02593732e+00 -1.24031031e+00
5.31885505e-01 -6.75689220e-01 -5.83046414e-02 -1.35279846e+00
2.22927839e-01 3.55180465e-02 2.41986409e-01 3.08016568e-01
2.48147786e-01 4.92630929e-01 2.45402798e-01 8.17215815e-02
-7.57609665e-01 4.02750909e-01 1.29492700e+00 6.95338771e-02
-3.49946290e-01 -7.37269521e-02 1.62124023e-01 9.47485805e-01
2.17780828e-01 -5.87211037e-03 -2.44408280e-01 -4.47906941e-01
3.81230749e-02 -8.75740200e-02 9.84738648e-01 -1.04729211e+00
1.87778190e-01 -1.05886772e-01 7.46143699e-01 -7.86360681e-01
1.09503222e+00 -1.10403109e+00 5.06394804e-01 3.03793550e-01
-3.40586871e-01 -6.76215291e-02 6.20192811e-02 5.36772490e-01
4.20681983e-01 5.04232883e-01 1.14023201e-01 -4.06295031e-01
-5.02437174e-01 4.36870664e-01 1.25549689e-01 -1.65830836e-01
1.31141829e+00 -4.89674747e-01 4.64783795e-02 -4.95319277e-01
-1.34958851e+00 8.18288475e-02 6.38888478e-01 7.72064626e-01
5.85701525e-01 -1.33125436e+00 -2.56330132e-01 4.75951284e-01
2.83334196e-01 2.59380817e-01 1.82652101e-01 8.71283948e-01
-3.31080288e-01 5.69742858e-01 -2.99300134e-01 -9.85750854e-01
-1.61478853e+00 6.06250823e-01 5.24915695e-01 2.76523858e-01
-7.13297486e-01 9.68729734e-01 3.33593994e-01 -8.95704150e-01
5.78359365e-01 -3.95350456e-01 3.14106084e-02 -1.06426939e-01
2.69522846e-01 6.47185624e-01 -9.15260687e-02 -1.29107237e+00
-6.42497301e-01 9.81876850e-01 2.66902536e-01 -1.37669250e-01
1.15666199e+00 -5.12949079e-02 -1.55155346e-01 8.37687910e-01
1.20471251e+00 -2.32236445e-01 -1.55018008e+00 -5.14594931e-03
-3.81027043e-01 -9.95571077e-01 -3.79370064e-01 -7.66880095e-01
-7.04822183e-01 9.21459198e-01 5.56497216e-01 -3.88127148e-01
6.52801514e-01 5.81593990e-01 2.59551823e-01 6.08130693e-01
7.80071199e-01 -9.38743114e-01 6.55351102e-01 6.39756620e-01
1.34471464e+00 -1.06210279e+00 1.45414263e-01 -6.16897404e-01
-5.32017708e-01 9.17259276e-01 8.99810016e-01 -1.60647780e-01
5.85285604e-01 6.32328019e-02 -1.81154996e-01 -6.74247026e-01
-1.10635966e-01 -1.16440602e-01 8.00535381e-01 8.80472541e-01
3.24037164e-01 2.30521426e-01 5.66835284e-01 2.97996670e-01
-3.26648951e-01 -2.09909692e-01 -2.22088560e-01 1.16297889e+00
3.47010404e-01 -8.96544695e-01 -3.40808600e-01 3.42909358e-02
5.31468950e-02 6.67464077e-01 -1.02317131e+00 9.17568803e-01
3.76727343e-01 6.51982486e-01 2.68811494e-01 -5.15720010e-01
1.10589600e+00 2.66034544e-01 1.40800142e+00 -7.60673404e-01
-5.34301341e-01 -1.78613946e-01 -1.13096513e-01 -1.02685153e+00
-5.98925710e-01 -6.28612280e-01 -1.07422304e+00 -3.03317070e-01
-2.40702540e-01 -5.70007741e-01 6.22482598e-01 1.13984799e+00
4.14071679e-01 4.13678110e-01 2.81245589e-01 -2.00927377e+00
-2.45390654e-01 -8.93023789e-01 -7.36308575e-01 9.96740460e-01
3.76785070e-01 -1.28530014e+00 -2.07465887e-01 1.63806140e-01] | [6.618928909301758, -0.9094635248184204] |
bf2636f5-a2a5-4a37-906a-eaa1d335ee48 | knowledge-base-question-answering-by-case | 2202.10610 | null | https://arxiv.org/abs/2202.10610v2 | https://arxiv.org/pdf/2202.10610v2.pdf | Knowledge Base Question Answering by Case-based Reasoning over Subgraphs | Question answering (QA) over knowledge bases (KBs) is challenging because of the diverse, essentially unbounded, types of reasoning patterns needed. However, we hypothesize in a large KB, reasoning patterns required to answer a query type reoccur for various entities in their respective subgraph neighborhoods. Leveraging this structural similarity between local neighborhoods of different subgraphs, we introduce a semiparametric model (CBR-SUBG) with (i) a nonparametric component that for each query, dynamically retrieves other similar $k$-nearest neighbor (KNN) training queries along with query-specific subgraphs and (ii) a parametric component that is trained to identify the (latent) reasoning patterns from the subgraphs of KNN queries and then apply them to the subgraph of the target query. We also propose an adaptive subgraph collection strategy to select a query-specific compact subgraph, allowing us to scale to full Freebase KB containing billions of facts. We show that CBR-SUBG can answer queries requiring subgraph reasoning patterns and performs competitively with the best models on several KBQA benchmarks. Our subgraph collection strategy also produces more compact subgraphs (e.g. 55\% reduction in size for WebQSP while increasing answer recall by 4.85\%)\footnote{Code, model, and subgraphs are available at \url{https://github.com/rajarshd/CBR-SUBG}}. | ['Andrew McCallum', 'Hannaneh Hajishirzi', 'Manzil Zaheer', 'Robin Jia', 'Elliot Tower', 'Ankita Naik', 'Ameya Godbole', 'Rajarshi Das'] | 2022-02-22 | null | null | null | null | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-4.82351780e-01 6.15615070e-01 -4.29831326e-01 -4.32073414e-01
-1.38764644e+00 -9.33948457e-01 8.90274271e-02 3.92327964e-01
3.12052201e-02 1.00930977e+00 4.70457494e-01 -5.01801074e-01
-4.70413625e-01 -1.36672652e+00 -1.13999176e+00 -2.01546803e-01
-6.88463524e-02 1.01399279e+00 9.19357717e-01 -4.25106674e-01
-8.73302370e-02 3.15632522e-01 -1.29275584e+00 7.06516623e-01
1.14166987e+00 8.35869551e-01 -7.53299817e-02 6.41614437e-01
-3.68502021e-01 1.11653805e+00 -3.83333325e-01 -6.51203215e-01
9.73742455e-02 -1.78602591e-01 -1.56504560e+00 -6.10418022e-01
7.35825956e-01 -1.62663937e-01 -7.78851151e-01 9.65107977e-01
2.08399504e-01 3.35280061e-01 6.02127314e-01 -9.66310322e-01
-7.22311020e-01 7.10025549e-01 -1.95797160e-01 7.09529936e-01
9.39679325e-01 -3.32065403e-01 1.50073516e+00 -6.36935294e-01
9.20512915e-01 9.07702863e-01 3.72109801e-01 3.25022519e-01
-1.02944493e+00 -5.34077168e-01 7.12002218e-02 4.07712787e-01
-1.65309596e+00 -3.73888433e-01 3.48038107e-01 3.26085873e-02
1.43345010e+00 6.76769257e-01 5.28434575e-01 4.34732944e-01
-6.92748791e-03 6.30166948e-01 5.76727569e-01 -2.10572302e-01
3.20771754e-01 6.73381835e-02 6.00390792e-01 1.06903911e+00
4.41747576e-01 -7.22541511e-01 -5.34966111e-01 -8.34267437e-01
2.81456381e-01 -3.24266076e-01 -4.18059707e-01 -4.69526172e-01
-7.30255365e-01 9.27434564e-01 7.63072789e-01 1.64758667e-01
-3.50491405e-01 3.82600427e-01 -2.94134095e-02 5.33915818e-01
1.45538807e-01 5.32631755e-01 -9.46685493e-01 1.43622383e-01
-5.81462145e-01 6.22496068e-01 1.45887911e+00 1.46330619e+00
1.38698971e+00 -7.66521335e-01 -1.95808217e-01 7.45739877e-01
2.15676278e-01 5.74424922e-01 2.57405728e-01 -9.87194777e-01
8.89502227e-01 1.18131208e+00 2.32768804e-01 -1.05687344e+00
-3.44631970e-01 -1.61090210e-01 -3.65532815e-01 -9.92498636e-01
3.45038950e-01 6.52284473e-02 -7.66493499e-01 1.81012797e+00
7.56625056e-01 -1.20472357e-01 2.24823371e-01 5.41387737e-01
1.16924298e+00 6.73773289e-01 1.39802158e-01 1.58905908e-01
1.75224149e+00 -6.83417499e-01 -1.57977492e-01 -2.58884817e-01
1.20365763e+00 -1.39799133e-01 1.07094264e+00 -1.31983757e-01
-9.01244342e-01 7.51573709e-04 -5.70836306e-01 -4.23463821e-01
-6.75641119e-01 -3.52189869e-01 6.23078525e-01 5.13700128e-01
-1.44208837e+00 -9.04748142e-02 -4.55906808e-01 -7.04748094e-01
2.82858700e-01 3.76411438e-01 -3.19695622e-01 -5.77948272e-01
-1.70643413e+00 6.60391092e-01 5.29832184e-01 -3.41303736e-01
-6.11404657e-01 -9.14377391e-01 -8.86521757e-01 2.37759203e-01
9.70360160e-01 -1.15989411e+00 1.11647713e+00 -1.80847049e-01
-7.63197064e-01 6.42649353e-01 -6.28578961e-01 -3.82608593e-01
-2.19928071e-01 -8.46413150e-02 -5.06921649e-01 4.52425778e-01
5.25661230e-01 3.78980130e-01 1.98101893e-01 -9.63249564e-01
-6.62320197e-01 -4.93776560e-01 6.26127064e-01 3.54086488e-01
-8.84504765e-02 -3.17154050e-01 -9.06830072e-01 -1.99669868e-01
3.57107401e-01 -7.46970534e-01 -6.78333715e-02 -7.09597766e-01
-5.33884346e-01 -5.97494006e-01 3.73312473e-01 -7.44871855e-01
1.59940255e+00 -1.72306788e+00 9.19016823e-02 6.92498684e-01
5.05317271e-01 -2.01602951e-01 -1.62937841e-03 7.00727522e-01
2.85042524e-01 1.68633506e-01 1.93129797e-02 5.65332294e-01
1.82555672e-02 4.07618821e-01 -6.52720213e-01 6.83515891e-02
-1.57216474e-01 1.40962934e+00 -9.72896338e-01 -6.05606794e-01
-5.56989849e-01 -1.86743468e-01 -8.68875802e-01 -9.59154293e-02
-8.43044281e-01 -2.05419794e-01 -1.02894819e+00 8.76479983e-01
3.23420554e-01 -1.00997472e+00 4.32001561e-01 -4.19913977e-01
6.22954249e-01 5.80639243e-01 -1.14665926e+00 1.47224307e+00
-3.14976787e-03 -4.36365567e-02 -1.57132760e-01 -8.16302359e-01
5.14114916e-01 2.21297398e-01 2.71389991e-01 -6.33952022e-01
-4.71623451e-01 4.00431991e-01 -4.21608835e-01 -6.53187215e-01
5.36219835e-01 1.61100671e-01 -4.76764768e-01 5.34396946e-01
2.18788430e-01 -2.24473804e-01 4.92576003e-01 9.33190703e-01
1.85231161e+00 -4.84120727e-01 3.34446460e-01 -4.24167395e-01
4.59833533e-01 5.54670572e-01 4.09057915e-01 9.81481791e-01
1.40841797e-01 -2.36341245e-02 6.60888791e-01 -4.36361760e-01
-6.64841712e-01 -1.30954540e+00 1.98552996e-01 1.22661638e+00
3.48756015e-01 -6.25306010e-01 -5.11266351e-01 -1.09571385e+00
4.94750708e-01 8.60945821e-01 -5.47276080e-01 -1.58530623e-01
-6.47473335e-01 -3.73125136e-01 8.20443869e-01 4.57402110e-01
4.12455678e-01 -8.96153390e-01 -3.76419583e-03 2.35676542e-02
-6.91664815e-01 -9.85195577e-01 -5.79730511e-01 1.50664467e-02
-8.54630768e-01 -1.52748370e+00 -2.82634288e-01 -7.53832340e-01
6.50046408e-01 3.16819131e-01 1.50487280e+00 1.86801985e-01
-4.99724559e-02 9.85394478e-01 -3.60201389e-01 4.38177250e-02
-1.14295639e-01 5.01803279e-01 -2.65447885e-01 -4.51873899e-01
6.39602184e-01 -4.36914921e-01 -7.98886478e-01 2.94159681e-01
-9.54696953e-01 -4.53324854e-01 5.22290885e-01 4.39341396e-01
7.55204737e-01 1.28042534e-01 7.60769308e-01 -1.37363100e+00
6.11145794e-01 -1.15441382e+00 -6.05062544e-01 9.32779908e-01
-6.07797801e-01 2.30938405e-01 4.52515721e-01 1.25441700e-02
-9.77073431e-01 -3.05111140e-01 7.93787017e-02 -4.52224091e-02
-3.57898846e-02 9.34137225e-01 -1.72682613e-01 4.11101803e-02
9.92563605e-01 2.58732319e-01 -4.06970322e-01 -2.23035321e-01
7.39786327e-01 2.47837469e-01 1.90266296e-01 -9.59699929e-01
8.30982745e-01 4.78012711e-01 -3.37272286e-01 -6.67520046e-01
-1.11965585e+00 -9.76162016e-01 -1.75733939e-01 1.17017731e-01
6.48445606e-01 -1.06179047e+00 -6.19234443e-01 -1.90982312e-01
-8.84561837e-01 -4.33219910e-01 -4.86416042e-01 1.34496108e-01
-4.49445069e-01 4.41734582e-01 -7.87417352e-01 -2.82100976e-01
-2.88358718e-01 -6.00717902e-01 9.58745360e-01 3.02841038e-01
-3.22168648e-01 -9.18110013e-01 4.48992312e-01 9.17943358e-01
1.75030023e-01 -2.01182544e-01 1.58628619e+00 -1.12967670e+00
-1.05365670e+00 -1.68525890e-01 -4.12148952e-01 -2.51980752e-01
1.68551728e-01 -6.35994971e-01 -5.63538849e-01 -2.02803582e-01
-2.93418199e-01 -5.87922812e-01 6.38818622e-01 1.34598717e-01
9.11710501e-01 -7.38067210e-01 -7.61890888e-01 3.34971517e-01
1.47632241e+00 -7.65212104e-02 5.94748735e-01 1.28469139e-01
5.21364748e-01 4.35086638e-01 4.55926001e-01 3.02029084e-02
9.73326802e-01 3.37908149e-01 8.01319554e-02 3.40147942e-01
-5.50153374e-04 -3.91418070e-01 5.86914532e-02 6.30938888e-01
3.22370343e-02 -3.71667773e-01 -1.08131301e+00 8.20330739e-01
-1.85619354e+00 -9.25332785e-01 -3.90619524e-02 1.90957892e+00
1.04222000e+00 -1.76966071e-01 6.24621585e-02 -5.02405882e-01
4.35328543e-01 3.58440503e-02 -8.31983984e-01 -4.81142849e-02
-2.36591265e-01 4.85470325e-01 5.55508077e-01 6.11081719e-01
-6.59310699e-01 1.13879395e+00 6.05377150e+00 9.10081565e-01
-2.59147376e-01 2.07554847e-02 3.50319535e-01 -5.04147597e-02
-9.60072577e-01 2.18024537e-01 -1.05062056e+00 7.80720860e-02
1.04206288e+00 -3.63131046e-01 5.64216137e-01 7.57367373e-01
-4.59303379e-01 -3.39741677e-01 -9.22498167e-01 3.93650413e-01
-4.21214662e-02 -1.60043895e+00 6.06606781e-01 -3.30508593e-03
7.55275071e-01 1.67216793e-01 -3.24036688e-01 7.26670265e-01
8.62325728e-01 -8.04917574e-01 2.40811855e-01 7.07790613e-01
5.19455791e-01 -5.65745950e-01 6.01936698e-01 3.29759330e-01
-1.52757025e+00 -1.86183944e-01 -4.14879113e-01 4.99427646e-01
-1.96389362e-01 5.68214774e-01 -9.84739482e-01 1.15595615e+00
1.06251740e+00 9.85475630e-02 -6.63807392e-01 6.63360178e-01
-1.94310144e-01 6.96464717e-01 -5.43969095e-01 -3.00674826e-01
1.30242348e-01 -7.56325945e-02 3.58963668e-01 1.03382838e+00
2.28663146e-01 7.26915240e-01 -1.26017779e-01 8.41573656e-01
-4.83364880e-01 3.10770988e-01 -5.38159370e-01 -5.07315099e-02
8.85659933e-01 8.15805852e-01 -5.34108162e-01 -5.64826727e-01
-5.50252736e-01 8.12255859e-01 1.01478684e+00 8.11570942e-01
-8.06129694e-01 -4.71765012e-01 5.23276031e-01 2.81961054e-01
5.24977565e-01 1.52038202e-01 3.46080422e-01 -1.29387772e+00
3.42349917e-01 -8.88139427e-01 1.35173178e+00 -8.31987560e-01
-1.38012445e+00 4.29497808e-01 2.91079640e-01 -3.14352989e-01
-2.62479305e-01 -2.19984174e-01 -1.35569707e-01 8.14093769e-01
-1.38802755e+00 -1.04486692e+00 -2.63548166e-01 1.14379930e+00
9.37868934e-03 1.82636023e-01 8.35813224e-01 2.47193694e-01
-2.68551469e-01 6.00830317e-01 -3.71765234e-02 2.10196182e-01
5.69093406e-01 -1.37464988e+00 1.41320273e-01 3.65543514e-01
3.88942063e-01 1.02557731e+00 3.42476249e-01 -9.50442553e-01
-1.65830195e+00 -1.13646460e+00 1.20613253e+00 -9.26406085e-01
8.47148776e-01 -1.42722055e-01 -1.03773272e+00 1.18504369e+00
-1.86799556e-01 1.16701387e-02 8.91536236e-01 4.95783359e-01
-7.77890265e-01 -3.04393888e-01 -1.22423089e+00 5.84580839e-01
1.08711886e+00 -7.76776075e-01 -8.92545819e-01 5.27521074e-01
1.06298578e+00 -4.48223740e-01 -1.09083879e+00 5.41864097e-01
3.08326095e-01 -7.81712115e-01 1.03247535e+00 -9.99983132e-01
-9.58959833e-02 -4.31436956e-01 -3.95327419e-01 -7.00822294e-01
-3.80734414e-01 -4.02972728e-01 -7.52436161e-01 7.30823159e-01
9.81709599e-01 -1.00835156e+00 1.05274510e+00 1.06064820e+00
7.76727870e-02 -1.14405203e+00 -1.07227981e+00 -5.43489695e-01
-4.09898683e-02 -1.54857874e-01 8.35503936e-01 8.78624141e-01
2.71416843e-01 3.22687805e-01 3.93753827e-01 7.92616308e-01
2.89089620e-01 5.18825829e-01 7.50322819e-01 -8.12515557e-01
-3.27455312e-01 1.36032552e-02 -1.82008654e-01 -1.54014325e+00
-5.78313321e-02 -1.08382010e+00 -2.42670685e-01 -1.85000360e+00
3.92607272e-01 -6.34401798e-01 -8.53798091e-02 6.12144530e-01
-3.21104884e-01 -1.39531106e-01 -4.43650752e-01 3.12126905e-01
-1.13215470e+00 3.20051521e-01 9.13772464e-01 -1.40794620e-01
-1.66809529e-01 -6.91666156e-02 -9.51652467e-01 3.54438990e-01
6.22528315e-01 -4.21120852e-01 -8.02464187e-01 -2.83831209e-01
8.78093421e-01 1.98713765e-01 4.31199253e-01 -6.32624984e-01
7.24045157e-01 -1.04718812e-01 6.59461394e-02 -5.95489740e-01
2.38566205e-01 -4.28857327e-01 2.88919955e-01 3.58701050e-01
-2.22436592e-01 -6.90050274e-02 2.03328788e-01 1.03526127e+00
-2.51395226e-01 -1.16342850e-01 1.99195251e-01 -3.47652733e-01
-7.09907293e-01 4.48281676e-01 -1.88388735e-01 7.89077222e-01
8.53050470e-01 -1.20160095e-01 -6.87253177e-01 -5.58895111e-01
-8.06640029e-01 5.97723365e-01 2.28091046e-01 -6.88057542e-02
5.89491546e-01 -1.06052423e+00 -4.25834268e-01 -2.79142529e-01
4.81563866e-01 4.31728929e-01 4.56024498e-01 6.70411348e-01
-5.25740147e-01 8.69307697e-01 6.39642537e-01 -1.66997924e-01
-8.37375641e-01 5.09864092e-01 3.99559557e-01 -6.41132832e-01
-3.45818907e-01 1.07606030e+00 2.32525006e-01 -7.58628666e-01
-1.63832270e-02 -5.36503255e-01 3.06138899e-02 -1.81605443e-01
1.15067892e-01 4.68244314e-01 2.35843211e-01 -2.60463357e-01
-4.39945489e-01 2.99129128e-01 -3.19813967e-01 2.54362196e-01
1.02483201e+00 -1.20801136e-01 -4.31662500e-01 1.50642797e-01
9.90386546e-01 3.84050488e-01 -3.92518908e-01 -5.22221804e-01
2.75860101e-01 -1.85128808e-01 -5.60524285e-01 -8.45719635e-01
-8.44807982e-01 -1.99637488e-01 -1.84475496e-01 5.18443346e-01
9.25870895e-01 9.90179420e-01 9.53556836e-01 1.00915146e+00
8.51593673e-01 -7.37890840e-01 -2.06387565e-02 6.02092206e-01
8.04442883e-01 -8.30596089e-01 -4.54804301e-02 -6.78015947e-01
-2.53587693e-01 6.72213793e-01 6.38763607e-01 4.90079150e-02
9.42584455e-01 -1.52475610e-01 -1.80418268e-01 -1.07983112e+00
-9.10060108e-01 -1.39383465e-01 4.34834659e-01 3.07825476e-01
8.73508980e-04 -3.78692821e-02 -9.19014588e-02 7.12128043e-01
-3.30669612e-01 -1.21291980e-01 -1.10291233e-02 9.49084282e-01
-6.10230148e-01 -8.43095958e-01 -1.14758395e-01 1.05678952e+00
-1.97059184e-01 -3.80043387e-01 -5.00390172e-01 9.53006327e-01
-1.83063447e-01 1.00884962e+00 -2.53275037e-01 -2.69060522e-01
4.88314807e-01 3.53790790e-01 4.39733773e-01 -9.45997119e-01
-4.53864992e-01 -6.33243382e-01 4.62992728e-01 -8.40819061e-01
-1.24649525e-01 -4.77682889e-01 -1.46998525e+00 -3.69780570e-01
-4.11595583e-01 8.19959223e-01 -2.56398082e-01 8.15672636e-01
8.33930075e-01 -3.12150177e-02 9.81064662e-02 4.78000551e-01
-5.22217453e-01 -7.79814601e-01 -7.46989131e-01 2.90170670e-01
-8.72542709e-02 -3.87395680e-01 -4.00397629e-01 -1.53264046e-01] | [10.415923118591309, 7.84948205947876] |
3435118f-9ad4-4fb0-b049-02de90fdaa37 | collision-free-motion-planning-for-mobile | 2306.17445 | null | https://arxiv.org/abs/2306.17445v1 | https://arxiv.org/pdf/2306.17445v1.pdf | Collision-free Motion Planning for Mobile Robots by Zero-order Robust Optimization-based MPC | This paper presents an implementation of robust model predictive control (MPC) for collision-free reference trajectory tracking for mobile robots. The presented approach considers the robot motion to be subject to process noise bounded by ellipsoidal sets. In order to efficiently handle the evolution of the disturbance ellipsoids within the MPC, the zero-order robust optimization (zoRO) scheme is applied. The idea is to fix the disturbance ellipsoids within one optimization iteration and solve the problem repeatedly with updated disturbance ellipsoid trajectories. The zero-order approach is suboptimal in general. However, we show that it does not impair convergence to the reference trajectory in the absence of obstacles. The experiments on an industrial mobile robot prototype demonstrate the performance of the controller. | ['Moritz Diehl', 'Niels van Duijkeren', 'Jonathan Frey', 'Florian Messerer', 'Yunfan Gao'] | 2023-06-30 | null | null | null | null | ['motion-planning'] | ['robots'] | [ 9.56282914e-02 5.27209997e-01 -2.07420856e-01 5.05081296e-01
-1.08569421e-01 -4.01887357e-01 5.88579118e-01 -1.40371382e-01
-4.56340760e-01 9.82737780e-01 -6.51277184e-01 -4.48193252e-01
-4.58288103e-01 -4.19892132e-01 -5.68597317e-01 -1.00841784e+00
-6.57313913e-02 6.65445864e-01 3.29297096e-01 -4.07348365e-01
3.49931806e-01 8.57807457e-01 -1.16944683e+00 -7.96786189e-01
9.81771946e-01 5.91200590e-01 5.68688154e-01 8.42573941e-01
5.58126688e-01 2.83081263e-01 -1.99608803e-01 2.62629956e-01
4.13230002e-01 1.38916150e-01 -5.54247200e-01 6.30243838e-01
-3.95794660e-01 1.50428871e-02 -1.44497871e-01 1.32352591e+00
1.78219050e-01 8.16538334e-01 8.27951968e-01 -1.34994853e+00
-4.13025618e-02 -2.15408772e-01 -2.04663843e-01 -1.21395007e-01
-8.48970562e-02 7.80377388e-02 1.92050666e-01 -8.87457788e-01
8.50373268e-01 1.30375397e+00 4.73742664e-01 2.43940890e-01
-1.03706670e+00 5.52020818e-02 3.10582258e-02 -4.09142300e-02
-1.58386564e+00 -3.09499800e-01 1.20688908e-01 -6.00729287e-01
8.20388913e-01 2.84176797e-01 1.75569922e-01 2.36612722e-01
9.14803207e-01 1.05869107e-01 5.50421476e-01 -2.72097230e-01
3.71156693e-01 1.94433898e-01 -1.33336186e-01 6.44725859e-01
8.08691025e-01 3.40447307e-01 8.14642966e-01 -6.76537827e-02
8.60911608e-01 -1.06314935e-01 -2.90862411e-01 -8.41640115e-01
-1.04190719e+00 8.94485593e-01 2.23297045e-01 1.22588128e-01
-6.35011554e-01 3.11352797e-02 1.64365724e-01 2.52370417e-01
1.10608742e-01 5.82957387e-01 -1.26849324e-01 6.46299347e-02
-1.27111092e-01 6.52099371e-01 1.11139739e+00 1.44389677e+00
1.92868695e-01 5.16564250e-01 2.58456260e-01 5.38459182e-01
6.18535876e-01 7.19100296e-01 1.03010625e-01 -1.33469975e+00
5.33711612e-01 2.65532583e-01 9.83553529e-01 -1.16900659e+00
-4.17973787e-01 -1.06032483e-01 -6.63855374e-01 7.29085386e-01
4.04920816e-01 -4.99431223e-01 -6.87789917e-01 1.02045536e+00
5.24925947e-01 -2.67436415e-01 2.82313019e-01 9.97426510e-01
-2.81294048e-01 9.92838025e-01 -2.73192346e-01 -8.60992432e-01
6.60351932e-01 -8.39607537e-01 -1.13268971e+00 -1.33275926e-01
4.31756437e-01 -7.21503139e-01 3.06294054e-01 4.85145688e-01
-1.02984262e+00 -4.52246100e-01 -1.28792179e+00 5.30488908e-01
-1.21238329e-01 2.19464317e-01 -3.69173557e-01 3.40080798e-01
-7.36474454e-01 7.20862269e-01 -7.63688326e-01 -3.22398722e-01
-4.90068257e-01 6.52955115e-01 -2.06604466e-01 3.48831952e-01
-8.87124121e-01 1.46837842e+00 8.00448895e-01 3.81231308e-01
-6.89148188e-01 -1.81943551e-01 -7.87353754e-01 -3.52192789e-01
7.18075454e-01 -2.80072033e-01 1.14621627e+00 -5.49131036e-01
-2.16968584e+00 2.41719589e-01 -1.39412433e-01 -2.89417624e-01
7.19324589e-01 -2.15903953e-01 -9.08729061e-02 1.06533960e-01
-1.71176851e-01 -4.63242903e-02 1.00718999e+00 -1.48326898e+00
-7.30479240e-01 6.49451315e-02 -2.84054577e-01 3.79437685e-01
2.38769323e-01 -1.90246478e-01 -4.17282939e-01 -2.28381485e-01
1.98426649e-01 -1.55716217e+00 -1.00418770e+00 -2.51222610e-01
-3.28923196e-01 6.03317842e-02 9.36697721e-01 -2.54431367e-01
1.13839388e+00 -1.92564380e+00 6.22350574e-01 5.39465904e-01
-3.33995789e-01 3.06454837e-01 1.95750698e-01 5.84072888e-01
1.54450506e-01 -1.20503604e-01 -1.49386033e-01 6.03631791e-03
-1.04481496e-01 4.76887465e-01 -3.05242181e-01 9.42658544e-01
1.52952090e-01 2.93455809e-01 -8.91517878e-01 -1.65940985e-01
3.49328995e-01 1.12345241e-01 -3.27646911e-01 -1.00522526e-02
-8.08057860e-02 2.81567723e-01 -6.81525230e-01 2.12813795e-01
7.45958507e-01 4.23229367e-01 1.41717374e-01 2.10304201e-01
-9.81024921e-01 -3.98330122e-01 -1.72975004e+00 5.09578705e-01
-2.84718007e-01 4.27254319e-01 1.04145467e+00 -9.79986787e-01
9.85501409e-01 3.47162992e-01 5.46164870e-01 -4.21013460e-02
5.89194357e-01 4.39972520e-01 -5.50817279e-03 -4.27663118e-01
7.42177725e-01 -2.39889607e-01 2.15230078e-01 -2.91003525e-01
-4.23366219e-01 -6.75911963e-01 1.42195582e-01 -3.87654126e-01
5.51986039e-01 -2.28896916e-01 6.52251780e-01 -7.16688752e-01
1.07553649e+00 4.71520483e-01 6.62851214e-01 3.60882193e-01
-3.30249608e-01 1.60644248e-01 1.70051292e-01 1.67818442e-02
-1.34665036e+00 -6.38415813e-01 -1.49219096e-01 4.74287361e-01
7.23707438e-01 2.31860742e-01 -4.49568391e-01 7.44019598e-02
1.29007339e-01 8.05187166e-01 -1.48790240e-01 -1.82722494e-01
-8.45426559e-01 -4.96134490e-01 -1.26087025e-01 1.03521079e-01
-2.26683673e-02 -4.25293684e-01 -6.77332103e-01 7.35913217e-01
2.44440958e-01 -9.05073524e-01 -3.74547124e-01 3.99956293e-02
-8.51792932e-01 -1.17061639e+00 -5.70808172e-01 -9.52003002e-01
9.70512450e-01 2.57977307e-01 6.99467212e-02 -8.30826387e-02
9.86004993e-02 5.32786608e-01 -8.74859691e-02 -5.49728692e-01
-7.63134480e-01 -3.16268772e-01 5.23466468e-01 -4.27766144e-02
-4.91623193e-01 2.00244695e-01 1.78450178e-02 9.45103526e-01
-4.33337092e-01 -4.75870818e-01 1.88496366e-01 7.21945107e-01
8.79242897e-01 6.30111039e-01 4.48424399e-01 -2.15317637e-01
9.51696277e-01 -4.83633786e-01 -1.31823349e+00 -1.66855633e-01
-5.24554193e-01 -1.98439375e-01 8.60103011e-01 -7.19314814e-01
-1.03050697e+00 4.36498135e-01 2.45515123e-01 -6.05024636e-01
1.28335342e-01 3.92243624e-01 -1.89683050e-01 -4.30672854e-01
3.06861430e-01 -3.19427461e-03 6.15566134e-01 -2.42288947e-01
3.90735775e-01 5.86209238e-01 6.47417843e-01 -3.30876440e-01
1.05686188e+00 4.83158350e-01 6.33854449e-01 -1.13650739e+00
1.86742887e-01 -8.15421402e-01 -7.71442890e-01 -4.55632061e-01
7.51871347e-01 -5.67663968e-01 -9.75326538e-01 9.65725482e-02
-1.09182727e+00 -3.84398341e-01 -2.08305731e-01 6.20273530e-01
-1.21451902e+00 3.34116697e-01 -3.73442262e-01 -1.44671035e+00
-1.01181962e-01 -1.36675060e+00 2.83777744e-01 1.89209640e-01
-2.81335432e-02 -1.06452465e+00 1.09407544e-01 -5.12300611e-01
3.33429992e-01 5.52431881e-01 4.05663639e-01 -3.27547014e-01
-3.81190836e-01 -5.54510176e-01 4.80978608e-01 1.01087958e-01
-8.82693976e-02 2.53292590e-01 -1.12101480e-01 -6.06387615e-01
5.39279163e-01 3.19116592e-01 5.69973094e-03 4.59478915e-01
-2.38339156e-02 -6.79711938e-01 -8.67034912e-01 1.81582883e-01
1.74612200e+00 9.97157156e-01 4.47079092e-01 8.34998786e-01
5.31594515e-01 7.42411017e-01 1.46210110e+00 2.42835343e-01
3.27323899e-02 7.54440427e-01 7.09433854e-01 3.61348957e-01
7.34832764e-01 3.17665964e-01 4.60282326e-01 4.43536699e-01
-7.35464394e-02 -1.76836416e-01 -8.62856030e-01 8.07049334e-01
-1.98981178e+00 -8.59739840e-01 -8.26348662e-01 2.24642324e+00
2.87457108e-01 -8.65865126e-03 -9.62924659e-02 8.11530352e-02
1.05288208e+00 -4.73139226e-01 -2.73581088e-01 -1.10225236e+00
2.95949250e-01 -7.47190893e-01 1.19582188e+00 1.27064741e+00
-1.13500178e+00 5.06326318e-01 6.18498707e+00 5.73960900e-01
-1.10008717e+00 -1.85479060e-01 -1.72471866e-01 4.32965636e-01
3.64922017e-01 -1.03923306e-01 -9.81482804e-01 3.74618649e-01
9.73191261e-01 -6.99829400e-01 3.97175193e-01 1.21532321e+00
8.04551423e-01 -3.41690332e-01 -4.86495823e-01 4.72244263e-01
-2.63094366e-01 -8.79734516e-01 -4.09348100e-01 1.82971045e-01
9.31077421e-01 -3.43768507e-01 1.18430987e-01 1.53897718e-01
2.19144911e-01 -5.38413167e-01 7.37622082e-01 6.04581058e-01
2.90156662e-01 -1.16070545e+00 7.94888616e-01 8.46841633e-01
-1.16590095e+00 -5.83319902e-01 -7.76217520e-01 -1.82438567e-01
7.66261339e-01 3.19354534e-02 -1.23335242e+00 6.93536401e-01
6.94267824e-03 2.77045637e-01 1.23123974e-01 1.47400391e+00
1.84749216e-01 -4.51081432e-02 -4.21677858e-01 -3.17730039e-01
5.08988082e-01 -7.68358052e-01 1.45901978e+00 9.58539665e-01
4.57944870e-01 2.18032047e-01 4.96818215e-01 5.82525313e-01
8.23937654e-01 5.68299927e-02 -9.39492881e-01 3.09449703e-01
2.60409564e-01 7.25934982e-01 -4.86805290e-01 -1.53828964e-01
6.08548261e-02 6.83149874e-01 -1.44699365e-01 3.80439520e-01
-6.38866842e-01 -8.23916376e-01 7.57138193e-01 -5.87832220e-02
4.32371140e-01 -7.65076101e-01 -3.23025256e-01 -4.05168712e-01
-2.16265649e-01 -3.26878697e-01 -1.11493446e-01 -3.20783943e-01
-8.29884231e-01 3.74232650e-01 2.32799456e-01 -1.72290504e+00
-6.39785051e-01 -7.05235004e-01 -5.34387469e-01 9.25300777e-01
-1.09212875e+00 -3.47211182e-01 3.69315028e-01 2.23164901e-01
7.05948651e-01 -2.70906519e-02 4.14962828e-01 -1.08833000e-01
-6.30204558e-01 -2.38483459e-01 8.67489100e-01 -5.22215426e-01
3.86097401e-01 -1.15039062e+00 -2.21302301e-01 1.00801051e+00
-1.24758863e+00 8.24525714e-01 1.46327221e+00 -9.30254698e-01
-1.61728954e+00 -1.44762468e+00 6.42544985e-01 -2.07870379e-01
9.39482272e-01 2.52868444e-01 -8.85783553e-01 7.05692112e-01
-1.14050023e-02 -2.88353860e-01 -3.62081081e-01 -8.32281351e-01
8.51580679e-01 1.24906540e-01 -1.16326165e+00 8.51034522e-01
2.75547117e-01 2.26700202e-01 -6.24886453e-01 2.35006943e-01
7.28594184e-01 -6.37709260e-01 -1.02878642e+00 6.32842660e-01
9.79066938e-02 6.43043369e-02 7.89609551e-01 -3.03109318e-01
-4.62523997e-01 -8.81766200e-01 9.98583883e-02 -1.52109110e+00
-4.15622562e-01 -1.22826445e+00 -5.95382899e-02 7.62278140e-01
2.65528351e-01 -7.24586010e-01 4.10532653e-01 5.46494067e-01
-4.10062224e-01 -7.75519669e-01 -1.20305133e+00 -1.24190354e+00
1.58795834e-01 -6.68832064e-02 -2.44977906e-01 4.46419656e-01
4.58298415e-01 5.00110611e-02 -3.36898953e-01 9.64654565e-01
5.61131239e-01 -3.44153494e-01 7.76623130e-01 -1.05933607e+00
2.64766872e-01 -3.13317209e-01 -2.05408037e-01 -8.73927236e-01
2.51861930e-01 -3.12219381e-01 7.98373163e-01 -1.67748702e+00
-6.27895474e-01 -3.70899528e-01 4.18990105e-01 -3.33020598e-01
-1.65963411e-01 -3.83085936e-01 3.02641511e-01 1.97948352e-01
-3.75925213e-01 5.22129953e-01 1.29206693e+00 -2.25400217e-02
-7.28842974e-01 6.30915165e-01 1.48714483e-01 8.42736185e-01
9.45074677e-01 -3.47731769e-01 -6.32832229e-01 1.02006318e-03
-3.39722723e-01 3.72736454e-01 -4.48798165e-02 -1.00712705e+00
4.54688281e-01 -7.37941980e-01 -3.55823040e-01 -8.10532153e-01
3.35494727e-01 -1.33841717e+00 5.37900209e-01 1.09783626e+00
1.12642050e-01 1.90323189e-01 1.85987383e-01 9.05438244e-01
-1.19523816e-01 -6.91745698e-01 1.27402008e+00 2.38674551e-01
-5.24650216e-01 5.42914942e-02 -1.15883386e+00 -4.89443541e-01
1.53635597e+00 -4.55179989e-01 6.80074394e-02 -2.01429307e-01
-9.02747214e-01 5.74238420e-01 6.25725985e-01 3.40031445e-01
5.01577914e-01 -9.84355450e-01 -2.47507349e-01 -8.13547224e-02
-4.06673372e-01 -1.14195332e-01 -6.18947521e-02 9.95914221e-01
-7.36897767e-01 9.43845749e-01 -1.68474391e-01 -5.37634492e-01
-1.08077502e+00 1.07348406e+00 7.13560164e-01 -5.79088032e-02
-2.77596682e-01 3.81420910e-01 -2.21447542e-01 -3.89055252e-01
2.77077239e-02 -4.12562877e-01 -4.05149966e-01 -9.09079835e-02
3.54428321e-01 9.66301620e-01 -1.57031059e-01 -8.32994163e-01
-1.07055038e-01 7.33183980e-01 3.33180159e-01 -3.58297080e-01
9.84570205e-01 -7.20662236e-01 -6.05013669e-02 4.54526395e-01
1.00292563e+00 -8.74347761e-02 -1.52135646e+00 4.13159102e-01
3.55833054e-01 -2.60454118e-01 -2.87207037e-01 -5.54358512e-02
-3.69440526e-01 6.19162805e-02 3.61825138e-01 1.82639420e-01
4.80376452e-01 -7.85521626e-01 1.55716777e-01 6.54963255e-01
5.89237690e-01 -1.41594422e+00 -2.70762682e-01 1.17027974e+00
1.22956192e+00 -6.74581885e-01 3.21742222e-02 -8.88630509e-01
-8.22379470e-01 1.25771511e+00 5.54588556e-01 -6.26869798e-01
6.57181740e-01 2.79291004e-01 -1.95224151e-01 4.79671806e-01
-6.92269266e-01 -2.55455375e-01 6.83408529e-02 6.78083777e-01
-3.14908028e-01 -3.32188904e-02 -9.58095372e-01 1.84941255e-02
3.22207868e-01 -7.22289681e-02 8.35321546e-01 1.01558208e+00
-1.04323518e+00 -6.87267840e-01 -1.06903684e+00 -1.83383882e-01
-2.49408171e-01 6.64632022e-01 2.17699304e-01 1.27936614e+00
-9.91804525e-02 1.29971874e+00 -1.28574483e-02 9.97030213e-02
9.55487072e-01 -2.08033144e-01 5.64827882e-02 -4.48554724e-01
-9.35887471e-02 3.12693983e-01 2.99896806e-01 -4.99795318e-01
8.42621103e-02 -5.96924067e-01 -1.63727212e+00 5.54523207e-02
-4.46506739e-01 3.26610297e-01 6.59300208e-01 6.61041379e-01
7.08213747e-02 2.02348962e-01 7.74761200e-01 -1.14164329e+00
-9.77567911e-01 -6.69183254e-01 -4.96874094e-01 -2.39488035e-01
4.60714608e-01 -9.98998165e-01 -4.92663115e-01 5.25835082e-02] | [5.214152812957764, 2.2320449352264404] |
c4d29f79-85c4-45a6-a6d0-45c7f739bd11 | arabic-dialect-identification-using-bert-fine | null | null | https://aclanthology.org/2020.wanlp-1.33 | https://aclanthology.org/2020.wanlp-1.33.pdf | Arabic Dialect Identification Using BERT Fine-Tuning | In the last few years, deep learning has proved to be a very effective paradigm to discover patterns in large data sets. Unfortunately, deep learning training on small data sets is not the best option because most of the time traditional machine learning algorithms could get better scores. Now, we can train the neural network on a large data set then fine-tune on a smaller data set using the transfer learning technique. In this paper, we present our system for NADI shared Task: Country-level Dialect Identification, Our system is based on fine-tuning of BERT and it achieves 22.85 F1-score on Test Set and our rank is 5th out of 18 teams. | ['Marwan Torki', 'Zeyad Ezzat', 'Moustafa Tohamy', 'Moataz Mansour'] | null | null | null | null | coling-wanlp-2020-12 | ['dialect-identification'] | ['natural-language-processing'] | [-5.2365750e-01 -3.2605022e-01 -1.1074528e-01 -6.6835368e-01
-7.6705950e-01 -6.5797526e-01 4.8013061e-01 -7.2920278e-02
-5.4811561e-01 9.5031160e-01 1.8545045e-01 -3.5385126e-01
-3.8740760e-01 -1.0054462e+00 -4.8345914e-01 -4.4856757e-01
-1.3263641e-01 1.1565299e+00 8.7449603e-02 -7.3947006e-01
2.2048756e-01 4.1435322e-01 -1.1364238e+00 4.3443793e-01
1.0321143e+00 6.4144629e-01 1.4674883e-01 5.0900936e-01
-1.8433785e-01 7.2517836e-01 -8.1774634e-01 -6.1306328e-01
4.0347496e-01 -1.4964303e-01 -9.6457511e-01 -5.8631384e-01
6.2058878e-01 -2.1602342e-01 -1.7933394e-01 7.4397480e-01
6.9611412e-01 3.0166285e-02 5.0270259e-01 -7.1275520e-01
-6.4423198e-01 8.4883755e-01 -4.8166162e-01 2.9142517e-01
-5.6692682e-02 -2.3100142e-01 8.3678573e-01 -7.4843478e-01
4.9171376e-01 1.0696170e+00 9.1731709e-01 3.7728414e-01
-1.1501095e+00 -9.4485474e-01 -1.4578460e-01 3.7555757e-01
-1.4458590e+00 -2.0830360e-01 6.2912709e-01 -5.4015052e-01
9.2232549e-01 -7.9439752e-02 2.4387056e-01 8.8952535e-01
-1.0597532e-01 6.0769045e-01 1.3555946e+00 -3.8936487e-01
-1.4520791e-01 4.6799919e-01 1.2759557e-01 6.6259706e-01
-1.9412497e-01 1.6961268e-01 -3.5473672e-01 6.5838084e-02
8.2414824e-01 -1.7992394e-01 -6.8940438e-02 9.9329986e-02
-1.0814794e+00 1.2184328e+00 6.4503056e-01 5.5230618e-01
-2.1370293e-01 -4.9617678e-01 4.0081382e-01 9.0720397e-01
5.3582829e-01 6.0852641e-01 -9.3572062e-01 -3.2796779e-01
-6.0126013e-01 1.7627527e-01 6.3518733e-01 2.6356193e-01
6.1923027e-01 -6.0701314e-02 1.4086087e-01 1.4262673e+00
-2.9834726e-01 2.2059007e-01 8.6644608e-01 -4.1181502e-01
3.8635105e-01 6.5397161e-01 -1.2901935e-01 -6.0420901e-01
-6.4359385e-01 -7.0017034e-01 -9.7494107e-01 3.9088619e-01
7.8147870e-01 -7.2736585e-01 -7.5774258e-01 1.8060871e+00
1.7197112e-02 -7.4253663e-02 -8.5881270e-02 8.7453431e-01
6.3904405e-01 6.7602473e-01 -3.6686403e-01 1.3550249e-01
8.2842237e-01 -9.4111454e-01 -2.3396312e-01 -2.2330377e-02
6.8991977e-01 -7.8355628e-01 9.5044947e-01 7.1040589e-01
-7.7743220e-01 -7.8130496e-01 -8.1518567e-01 3.1591532e-01
-6.6836357e-01 6.6004582e-02 7.2098798e-01 8.2539624e-01
-1.1454561e+00 4.0319881e-01 -1.5016675e-01 -6.3969880e-01
2.0084901e-01 8.9499658e-01 -5.0953341e-01 -4.5352943e-02
-1.2539139e+00 7.0981270e-01 4.5430419e-01 -1.1298231e-01
-6.4454687e-01 -6.1977088e-01 -2.7485490e-01 4.2805310e-02
2.7261136e-02 -3.7818378e-01 9.9397999e-01 -1.1052724e+00
-1.7050209e+00 1.1916025e+00 3.3765700e-01 -5.6423026e-01
5.3389454e-01 -1.0171944e-01 -6.7153853e-01 -5.1303059e-01
-6.3676042e-03 4.2582586e-01 3.1086531e-01 -7.3621970e-01
-9.0877336e-01 -4.4475532e-01 1.2021281e-01 -1.5888510e-03
-4.9931869e-01 2.6795939e-01 -8.3210431e-02 -4.1745627e-01
-2.4692228e-01 -9.2943561e-01 1.1964116e-02 -9.5808154e-01
-2.4239912e-03 -4.6763757e-01 3.7771571e-01 -7.5177640e-01
1.0774037e+00 -2.1874092e+00 1.2077284e-01 3.8438696e-01
2.6551394e-02 4.8913395e-01 -8.4622346e-02 4.5337823e-01
-1.8234089e-01 -4.8613522e-02 2.7157018e-01 2.0843931e-01
-2.2104941e-01 -1.8144688e-01 -1.2504560e-01 1.5232804e-01
-7.5228058e-02 5.1970589e-01 -7.1269250e-01 -2.4862966e-02
4.2993810e-02 1.0232518e-02 -6.2155253e-01 2.6019591e-01
8.9750767e-02 5.2345866e-01 -6.5139462e-03 4.2778772e-01
7.5427902e-01 -8.2070522e-02 2.6264992e-01 2.6699221e-01
-1.5583624e-01 1.9613974e-01 -9.9759066e-01 1.5688415e+00
-5.9345645e-01 7.4243039e-01 -7.2466269e-02 -1.0699108e+00
1.2368600e+00 1.6425674e-01 2.7275085e-01 -9.9704647e-01
8.9739729e-03 4.2105198e-01 4.9100912e-01 -2.8953624e-01
9.8707989e-02 -3.4863853e-01 -3.4245184e-01 4.6561822e-01
9.8989487e-02 3.0883551e-01 -1.0165744e-01 -1.0767889e-01
9.4807470e-01 -2.6288342e-01 9.6288890e-02 -2.7929097e-01
6.4555281e-01 1.8314910e-01 5.0358623e-01 7.9196000e-01
-1.9601935e-01 6.0557956e-01 3.6245516e-01 -1.0209469e+00
-1.0798143e+00 -8.9644092e-01 -3.2640010e-01 1.7145052e+00
-5.5062222e-01 -4.5541268e-02 -6.7698377e-01 -8.2894117e-01
5.8033157e-02 2.5012571e-01 -7.5673270e-01 -1.5442479e-02
-6.2660241e-01 -9.4470358e-01 7.4562520e-01 5.1574117e-01
8.4647667e-01 -1.0450326e+00 3.0495194e-01 2.1582524e-01
-6.7686006e-02 -7.0525700e-01 -1.8711995e-01 2.0828019e-01
-6.2564063e-01 -8.4451514e-01 -7.6148474e-01 -1.4516484e+00
1.7367032e-01 2.1898682e-01 1.1949786e+00 2.8306052e-02
2.9865389e-03 -5.3054851e-01 -2.8615963e-01 -3.9451978e-01
-3.7044591e-01 6.5431660e-01 3.6665103e-01 9.6152350e-02
6.5595043e-01 -4.9940023e-01 -3.5068622e-01 5.0406051e-01
-2.9848430e-01 -8.7731093e-02 5.1039302e-01 1.0747803e+00
4.2852473e-02 3.8125989e-01 1.0818377e+00 -1.1387355e+00
7.5762773e-01 -4.0706614e-01 -6.9665450e-01 1.4710633e-01
-4.1000712e-01 -1.5891813e-01 7.5966942e-01 -2.9898748e-01
-1.0933940e+00 -1.5400615e-01 -5.8476937e-01 4.6247859e-02
-3.0525354e-01 7.6411927e-01 8.0687068e-02 -4.8398435e-02
8.7203354e-01 -5.2915223e-02 -1.3817257e-01 -8.1426311e-01
-2.0996338e-02 1.1852273e+00 3.8970825e-01 -3.3322233e-01
5.7745785e-01 1.9045857e-01 -4.7115031e-01 -4.9630281e-01
-9.7479886e-01 -4.9612716e-01 -8.9781326e-01 1.3008506e-02
7.7109146e-01 -1.0471189e+00 -6.8886578e-01 6.9334668e-01
-5.8985859e-01 -7.1065950e-01 2.4509609e-01 5.2132660e-01
-1.8590917e-01 -3.4890899e-01 -5.8148068e-01 -3.2539365e-01
-3.0153519e-01 -8.8138622e-01 5.2828813e-01 1.5805878e-01
-1.7228454e-01 -1.1156313e+00 6.1894786e-01 4.8901168e-01
5.7069051e-01 1.4967942e-01 1.1191797e+00 -1.0272651e+00
-1.5926108e-01 -2.8029817e-01 -2.4007553e-01 3.5758728e-01
3.2906491e-01 -2.9330984e-01 -1.1163197e+00 -4.5802605e-01
-4.8319823e-01 -6.8468320e-01 7.8939986e-01 4.6669653e-01
1.2903850e+00 1.2810606e-01 -5.0230272e-04 7.7627480e-01
1.4739659e+00 2.5751582e-01 4.4054767e-01 7.4004489e-01
7.2738242e-01 5.7131952e-01 4.6257937e-01 2.1910585e-01
4.8778823e-01 8.1614560e-01 6.1814595e-02 -2.2867967e-01
3.8878191e-03 -1.8993489e-01 2.4550824e-01 8.4438956e-01
-3.4697649e-01 1.6026783e-01 -1.2393562e+00 4.7889203e-01
-1.7944844e+00 -9.5896870e-01 -1.3270578e-01 2.4087677e+00
6.9303626e-01 3.7291715e-01 7.2505069e-01 -5.8245964e-02
5.9134471e-01 -4.1687685e-01 -3.5635397e-01 -5.7799399e-01
-3.8691223e-01 2.9826179e-01 4.3147635e-01 6.0581058e-01
-1.0029424e+00 1.2011099e+00 6.8712025e+00 6.4572090e-01
-1.3574440e+00 1.7712943e-01 6.4612913e-01 -1.2711403e-01
1.6510427e-01 -4.4828445e-01 -8.8306803e-01 4.7757882e-01
1.1831996e+00 -4.5726471e-02 6.4552683e-01 6.4566392e-01
-2.1995399e-02 1.5064934e-01 -8.6581886e-01 1.0814701e+00
-1.4988941e-02 -1.2320074e+00 -1.6431107e-01 4.1524327e-01
9.6991348e-01 5.0044918e-01 2.0465073e-01 9.3930900e-01
6.2135726e-01 -1.3192959e+00 2.9016805e-01 2.0473765e-01
6.2708324e-01 -1.2646732e+00 1.2441170e+00 5.8859789e-01
-6.0009098e-01 -3.6545318e-01 -5.9754115e-01 -3.4275705e-01
-2.7097973e-01 3.8457283e-01 -1.2723601e+00 3.4685519e-01
9.8891199e-01 4.3838757e-01 -6.6501015e-01 1.2344002e+00
1.5873839e-01 8.9512217e-01 -2.1735007e-01 -1.6813887e-03
4.7481844e-01 -3.5221133e-01 3.2753229e-02 1.1526866e+00
2.6199615e-01 -2.5160259e-01 1.8780650e-01 1.1983000e-01
-2.5442132e-01 2.9983187e-01 -7.5791526e-01 2.3013540e-01
8.8521361e-02 1.1630949e+00 -3.9383209e-01 -2.3606032e-01
-4.8979694e-01 8.6789089e-01 8.8822204e-01 1.5050046e-02
-6.8581694e-01 -4.3542451e-01 6.2289488e-01 -2.2719122e-02
3.0716756e-01 -1.0171027e-01 -3.3842024e-01 -1.0644583e+00
-5.1853538e-01 -1.2426536e+00 8.0800509e-01 -4.4711578e-01
-1.4350691e+00 7.3041838e-01 -4.7769821e-01 -9.9260026e-01
-4.4613954e-01 -8.4630090e-01 -5.8804142e-01 1.0879004e+00
-1.2090392e+00 -1.0086677e+00 -2.5420049e-01 7.6762593e-01
5.4790968e-01 -9.1261429e-01 1.0318824e+00 6.6426158e-01
-5.0729311e-01 8.0187052e-01 7.9068595e-01 5.7766855e-01
1.2483503e+00 -1.4005675e+00 2.9749593e-01 3.7112877e-01
2.5280493e-01 5.9748703e-01 2.9924497e-01 -3.1935832e-01
-9.7852832e-01 -7.3039103e-01 8.9036286e-01 -4.7146472e-01
7.1660227e-01 -6.0548753e-01 -8.9177370e-01 7.6026726e-01
3.2937062e-01 -6.6363204e-01 1.0879152e+00 1.0583210e+00
-3.2400951e-01 -4.9633989e-01 -1.0005718e+00 2.8365025e-01
7.0039326e-01 -5.3982353e-01 -4.0893590e-01 5.5887961e-01
4.0955961e-01 -2.2146225e-01 -1.0907393e+00 2.7319336e-01
6.2669724e-01 -1.0833715e+00 8.0208492e-01 -8.4037322e-01
1.7650409e-01 1.8727100e-02 -8.8919505e-02 -1.8824506e+00
-7.4568105e-01 -3.8074109e-01 6.5925723e-01 1.3436823e+00
7.1768719e-01 -7.2333103e-01 8.7575078e-01 7.8681216e-02
4.9706127e-02 -4.6536300e-01 -6.4365685e-01 -7.9322702e-01
5.0313509e-01 -1.2288997e-01 9.4114298e-01 1.4012882e+00
-6.5044545e-02 6.5302593e-01 -5.3815746e-01 -6.8893651e-03
3.3853939e-01 4.2345390e-01 1.0248964e+00 -1.5555646e+00
-5.9185666e-01 -6.6367596e-01 -2.6089245e-01 -5.3054440e-01
1.6615944e-01 -9.9476123e-01 -2.6870823e-01 -1.2210324e+00
2.8852269e-01 -7.4414343e-01 -5.6927049e-01 4.0384752e-01
1.9106455e-02 4.6378610e-01 -1.2228812e-01 9.4871767e-02
-3.2145497e-01 4.6748642e-02 9.9319202e-01 -2.5946578e-01
-3.7349871e-01 3.3176389e-01 -8.7778354e-01 5.3953403e-01
1.1740158e+00 -1.5188643e-01 -1.8107589e-01 -6.4777845e-01
3.4251618e-01 -1.5060726e-01 -3.2524416e-01 -1.0287380e+00
-4.1114040e-02 -6.5570623e-02 6.9087249e-01 -4.8508093e-01
9.9652410e-02 -4.5825914e-01 -3.5051212e-02 3.9815077e-01
-1.7827864e-01 6.7603476e-02 2.5256193e-01 -1.2138972e-02
-3.7245625e-01 -1.2020985e-01 7.8403777e-01 -1.2798433e-01
-8.7634623e-01 2.5021571e-01 -2.8862634e-01 -1.0945342e-01
6.9514167e-01 5.2021760e-02 -2.0597827e-01 -3.9247873e-01
-7.1713519e-01 1.9254504e-01 4.0083435e-01 6.5366769e-01
1.7083676e-01 -1.2522634e+00 -1.1167632e+00 3.7254462e-01
1.1335985e-01 -3.5188827e-01 1.9642670e-01 4.9492010e-01
-7.1095151e-01 5.6649995e-01 -7.2351432e-01 -4.1130841e-01
-1.2654840e+00 3.9911050e-01 4.6715975e-01 -5.1146352e-01
-2.8862128e-01 1.0985422e+00 2.7521509e-01 -1.1218920e+00
6.7986259e-03 1.8429969e-01 -5.7131737e-01 5.8433179e-02
6.9618624e-01 2.4377166e-01 3.7717804e-01 -5.4703057e-01
-3.3321232e-01 5.7608885e-01 -5.1171988e-01 -1.4061876e-01
1.6915532e+00 1.2569652e-01 -2.4132602e-02 4.1929409e-01
1.2655426e+00 2.1540292e-01 -7.2303671e-01 -2.4888627e-01
-7.3428646e-02 -3.7833616e-01 1.8850298e-01 -1.2560941e+00
-9.1654325e-01 7.8162509e-01 9.4006729e-01 3.9021492e-01
1.0429046e+00 -1.6893440e-01 6.4038098e-01 5.1514560e-01
4.4944268e-01 -1.3083035e+00 -1.0251549e-01 9.2998064e-01
6.3104326e-01 -1.6386324e+00 -4.4901472e-01 1.2054136e-01
-5.2639234e-01 9.8584610e-01 7.6553923e-01 -2.3329255e-01
6.4170504e-01 -3.1828526e-02 3.9646485e-01 -6.4798303e-02
-5.7481873e-01 -1.8480250e-01 1.8189874e-01 6.1573231e-01
6.4482015e-01 4.5518124e-01 3.0135684e-02 3.4927568e-01
-7.0489025e-01 -1.0471404e-01 2.7129045e-01 3.4626102e-01
-4.9000749e-01 -1.2880698e+00 -4.8803422e-01 5.5301863e-01
-6.3034993e-01 -8.2921498e-02 -5.7045573e-01 8.0643582e-01
2.5740123e-01 9.1602790e-01 1.6915576e-01 -7.3026484e-01
2.7729785e-01 1.4010382e-01 5.2919400e-01 -5.2020156e-01
-1.0152752e+00 -1.4981280e-01 2.6376471e-01 -1.5409684e-01
-9.0097480e-02 -6.2624007e-01 -8.8113111e-01 -8.3139646e-01
2.9373450e-02 1.4735967e-01 6.9174069e-01 7.5851053e-01
2.6106557e-01 2.0728877e-01 1.0240902e+00 -3.8636166e-01
-4.6467200e-01 -1.2803676e+00 -6.5924805e-01 5.5043358e-01
1.9352964e-01 -4.3084428e-01 -3.2036148e-03 -3.0823112e-01] | [10.178744316101074, 10.687644004821777] |
25a27777-0fc1-4940-8041-7dfc25d4cda9 | neural-transition-system-for-end-to-end | 2110.02001 | null | https://arxiv.org/abs/2110.02001v2 | https://arxiv.org/pdf/2110.02001v2.pdf | Mastering the Explicit Opinion-role Interaction: Syntax-aided Neural Transition System for Unified Opinion Role Labeling | Unified opinion role labeling (ORL) aims to detect all possible opinion structures of 'opinion-holder-target' in one shot, given a text. The existing transition-based unified method, unfortunately, is subject to longer opinion terms and fails to solve the term overlap issue. Current top performance has been achieved by employing the span-based graph model, which however still suffers from both high model complexity and insufficient interaction among opinions and roles. In this work, we investigate a novel solution by revisiting the transition architecture, and augmenting it with a pointer network (PointNet). The framework parses out all opinion structures in linear-time complexity, meanwhile breaks through the limitation of any length of terms with PointNet. To achieve the explicit opinion-role interactions, we further propose a unified dependency-opinion graph (UDOG), co-modeling the syntactic dependency structure and the partial opinion-role structure. We then devise a relation-centered graph aggregator (RCGA) to encode the multi-relational UDOG, where the resulting high-order representations are used to promote the predictions in the vanilla transition system. Our model achieves new state-of-the-art results on the MPQA benchmark. Analyses further demonstrate the superiority of our methods on both efficacy and efficiency. | ['Chong Teng', 'Yijiang Liu', 'Meishan Zhang', 'Donghong Ji', 'Fei Li', 'Hao Fei', 'Shengqiong Wu'] | 2021-10-05 | null | null | null | null | ['fine-grained-opinion-analysis'] | ['natural-language-processing'] | [ 2.56919354e-01 3.23741019e-01 -3.32464129e-01 -4.69774336e-01
-7.97068715e-01 -6.95847154e-01 4.90470797e-01 4.67509478e-01
5.38446978e-02 4.63037014e-01 5.80017984e-01 -7.55964875e-01
-2.11814299e-01 -9.26920235e-01 -3.79148096e-01 -5.72248280e-01
1.18984058e-01 5.41245878e-01 4.27674741e-01 -8.59426975e-01
2.48148158e-01 -1.06480261e-02 -1.56513667e+00 5.20436049e-01
1.41218495e+00 8.92273724e-01 -3.52977604e-01 4.04465824e-01
-5.24404645e-01 1.12984300e+00 -6.66156769e-01 -1.15457666e+00
4.20351960e-02 -4.84308064e-01 -1.02477753e+00 -1.60876587e-01
3.62864077e-01 5.09475805e-02 -3.95147353e-02 1.04825759e+00
4.38691050e-01 3.20026159e-01 5.16873300e-01 -1.21682918e+00
-1.10325837e+00 9.49541569e-01 -6.01614535e-01 1.54740170e-01
4.76480693e-01 -4.39316154e-01 1.79132855e+00 -7.43911266e-01
5.83182931e-01 1.43021739e+00 5.19972205e-01 3.74947876e-01
-8.20612192e-01 -2.64510602e-01 7.11919129e-01 3.01412821e-01
-1.09469092e+00 -2.79694468e-01 7.47799277e-01 7.62825683e-02
1.40329182e+00 3.68674815e-01 6.58558071e-01 7.98694015e-01
5.19653320e-01 7.70357072e-01 9.91926551e-01 -3.97249788e-01
1.25852600e-01 -2.42039040e-01 8.22484076e-01 9.62105930e-01
3.17838639e-01 -4.49114263e-01 -7.98176944e-01 -3.61651123e-01
1.95702478e-01 -1.79417446e-01 -2.80098081e-01 -2.46375009e-01
-5.60641587e-01 8.57163489e-01 4.77065533e-01 4.58405018e-01
-2.38968566e-01 -1.84122905e-01 3.07177484e-01 5.00620544e-01
1.06204414e+00 4.02088553e-01 -8.13720465e-01 1.04828246e-01
-6.53337657e-01 1.32781804e-01 1.11622894e+00 7.82575607e-01
6.87930286e-01 -3.08458775e-01 -3.72912735e-01 7.76163816e-01
6.78525627e-01 3.27739984e-01 2.14645311e-01 -6.92451417e-01
4.13340122e-01 1.36207128e+00 -1.95319012e-01 -1.11380136e+00
-3.96307081e-01 -8.08211625e-01 -7.17443824e-01 -3.12778503e-01
-1.25588581e-01 3.35629424e-03 -1.15719402e+00 1.56835020e+00
5.65165281e-01 -1.04288511e-01 3.58835429e-01 6.06108904e-01
9.26610112e-01 7.51257420e-01 8.98994952e-02 -4.86487210e-01
1.62397790e+00 -1.31760430e+00 -8.91915143e-01 -2.96296805e-01
9.77537751e-01 -5.71711242e-01 8.96371722e-01 2.18830332e-01
-9.74781275e-01 -2.01463431e-01 -8.90170634e-01 -3.50032121e-01
-5.40544271e-01 -3.49586070e-01 8.72462392e-01 7.78767526e-01
-1.37920594e+00 2.62789100e-01 -4.77487266e-01 -1.40587702e-01
9.25987884e-02 5.32851815e-01 -2.59114176e-01 -2.97294468e-01
-1.74587870e+00 1.10336661e+00 1.21378750e-01 3.96420211e-01
-2.31319770e-01 -7.11038768e-01 -1.14847004e+00 2.17832282e-01
8.33851576e-01 -1.11158884e+00 1.12346387e+00 -8.32915187e-01
-1.35919702e+00 5.61801374e-01 -7.71174967e-01 -3.02131623e-01
-1.89913064e-01 -2.35889748e-01 -6.95151508e-01 6.36635795e-02
9.16303545e-02 3.14722091e-01 5.84609449e-01 -1.31930161e+00
-8.29507411e-01 -5.89877605e-01 1.01154780e+00 7.00097322e-01
-2.25649819e-01 2.36231342e-01 -7.35908151e-01 -4.66894597e-01
3.10623586e-01 -7.70177305e-01 -4.25751597e-01 -5.02622306e-01
-2.30414763e-01 -7.54457355e-01 3.85328919e-01 -5.29171765e-01
1.77717090e+00 -1.66462243e+00 4.68498230e-01 2.66464740e-01
4.35789734e-01 1.99596241e-01 -1.29923418e-01 5.03653347e-01
3.67973484e-02 3.98576826e-01 -4.66664106e-01 -5.62750399e-01
1.09693997e-01 5.96930981e-01 -2.62852699e-01 6.73536286e-02
-3.86564396e-02 1.16617286e+00 -1.05987835e+00 -5.70708275e-01
-3.21839899e-01 4.55145419e-01 -7.81465650e-01 2.20798515e-02
-3.73356611e-01 -7.77643695e-02 -4.28579181e-01 8.99187982e-01
6.51736677e-01 -3.57079685e-01 6.78782165e-01 -4.51291651e-01
6.27162829e-02 6.14684820e-01 -9.56269741e-01 1.62331164e+00
-4.33753431e-01 -5.35670966e-02 1.01642974e-01 -5.79166472e-01
7.57725835e-01 1.68314278e-01 3.63536805e-01 -6.64944768e-01
1.18776642e-01 2.51051515e-01 1.75267875e-01 -3.19809228e-01
8.54078531e-01 -7.92527646e-02 -2.10036322e-01 3.10480505e-01
1.19887330e-01 4.99176532e-02 5.15848994e-01 6.11953080e-01
1.13038588e+00 1.05567873e-01 4.53574121e-01 -4.28828359e-01
8.56994569e-01 -6.92674592e-02 4.98098671e-01 5.32424212e-01
2.84117740e-02 2.25957349e-01 1.01450372e+00 -2.62002766e-01
-2.69939542e-01 -6.65427506e-01 2.90779322e-01 1.56861591e+00
3.72952580e-01 -1.03888869e+00 -3.42822909e-01 -1.24447691e+00
-1.57223091e-01 8.83346379e-01 -6.73519254e-01 -1.88590094e-01
-6.20295823e-01 -8.49558532e-01 2.53337860e-01 4.09592718e-01
2.42683381e-01 -8.00321877e-01 -1.19998746e-01 1.36439621e-01
-3.31277460e-01 -9.31378901e-01 -4.81530964e-01 2.43865624e-01
-5.47809720e-01 -1.06646287e+00 -2.38080427e-01 -5.74002683e-01
7.73153841e-01 3.16802561e-01 1.40830517e+00 2.37661511e-01
5.18913746e-01 2.04071537e-01 -8.53762925e-01 -1.77577153e-01
-5.25816716e-02 2.97749549e-01 -1.56133220e-01 4.61225323e-02
2.49407649e-01 -5.58223009e-01 -7.09314823e-01 1.98827118e-01
-9.81607139e-01 -2.10426793e-01 6.13537431e-01 9.05453622e-01
4.59748715e-01 8.19807202e-02 4.19819921e-01 -1.36451805e+00
9.46786702e-01 -5.27603447e-01 -2.77785122e-01 6.89756036e-01
-1.00640583e+00 2.58126736e-01 5.10676086e-01 1.71119168e-01
-1.31347406e+00 -4.99860615e-01 -3.92353445e-01 -3.69124152e-02
4.76690650e-01 1.10401094e+00 -2.94965893e-01 1.01118781e-01
3.77823740e-01 -3.39576229e-02 -3.86779159e-01 -2.39124715e-01
7.33114421e-01 4.28939015e-01 1.54599160e-01 -6.66558504e-01
3.71824235e-01 1.96471870e-01 -3.07375807e-02 -3.14874977e-01
-1.42647481e+00 -6.25813544e-01 -4.29402083e-01 4.94228676e-02
6.83879614e-01 -8.16075683e-01 -4.26941097e-01 3.51217091e-01
-1.39818203e+00 2.43330449e-01 -6.65527955e-02 -1.21477760e-01
2.14020416e-01 6.41149342e-01 -6.16960406e-01 -9.80122507e-01
-8.11550200e-01 -9.85000551e-01 1.14237595e+00 3.44444104e-02
-3.74537632e-02 -1.11294758e+00 2.40423918e-01 5.23407161e-01
4.42841440e-01 -6.58165812e-02 1.33096015e+00 -8.70896757e-01
-3.37488383e-01 -3.88175133e-03 -1.77415162e-01 3.11979085e-01
-1.67445485e-02 -3.91610116e-02 -6.58885181e-01 -1.41602919e-01
-3.75840701e-02 -1.32971197e-01 9.63173449e-01 7.43019506e-02
6.20520890e-01 -3.40884387e-01 -2.00977758e-01 2.06153989e-01
1.23121643e+00 1.26344442e-01 3.91600281e-01 -4.12755795e-02
9.59944427e-01 9.64553237e-01 6.44241989e-01 -1.05526578e-02
1.02211738e+00 4.96313214e-01 5.33482969e-01 6.37085512e-02
1.17314234e-01 -1.18021585e-01 6.02827847e-01 1.35693741e+00
-4.08081830e-01 -9.18395221e-01 -8.21479321e-01 5.14904857e-01
-2.18398452e+00 -5.97115517e-01 -5.80937803e-01 1.62634778e+00
6.02031231e-01 2.80144811e-01 -3.01317126e-01 5.11537753e-02
5.22465646e-01 6.58605933e-01 -3.80742073e-01 -6.10127330e-01
-2.21424997e-01 5.62036395e-01 3.31654251e-01 8.52240801e-01
-7.62307942e-01 1.12724292e+00 6.16651917e+00 7.74943769e-01
-5.48138142e-01 3.68168950e-01 4.14051324e-01 2.23601505e-01
-9.22691941e-01 3.24886352e-01 -9.18056190e-01 -5.24328984e-02
1.09180832e+00 -2.82698423e-01 1.63020894e-01 5.86531878e-01
-1.93406031e-01 -3.99549343e-02 -7.80080378e-01 5.92070103e-01
4.49793041e-01 -1.15486217e+00 5.45924187e-01 -8.91699642e-02
6.35545850e-01 -1.63218275e-01 -1.93930641e-01 6.69776201e-01
3.68111253e-01 -7.23758519e-01 5.66381037e-01 2.88385332e-01
4.35320854e-01 -8.41453135e-01 1.08716989e+00 1.94837570e-01
-1.64124835e+00 -9.97678265e-02 -2.89747208e-01 -7.76867047e-02
3.36351842e-01 6.89848423e-01 -1.96218640e-01 1.47173202e+00
3.26754540e-01 7.45689094e-01 -5.80773115e-01 5.12521565e-01
-6.82081103e-01 5.40629387e-01 -1.24933861e-01 -1.31744385e-01
3.36674869e-01 -2.72159457e-01 3.88142437e-01 9.36796665e-01
2.22609445e-01 3.29283088e-01 1.90914989e-01 2.54964709e-01
-2.42262840e-01 3.76868516e-01 -4.38030213e-01 6.18429668e-02
3.40697646e-01 1.38165987e+00 -6.82891905e-01 -4.16791379e-01
-5.71279168e-01 1.02665758e+00 5.15189230e-01 2.31915727e-01
-8.16479325e-01 -1.21366434e-01 5.78370631e-01 -1.97269246e-01
2.38990724e-01 1.51395071e-02 1.65853985e-02 -1.39829552e+00
2.19682828e-01 -1.06175518e+00 7.31859565e-01 -7.44886577e-01
-1.37222886e+00 9.64654565e-01 9.02387649e-02 -7.46366501e-01
-2.09876150e-01 -6.71828449e-01 -4.37974662e-01 8.42152894e-01
-1.58995175e+00 -1.56982493e+00 2.00587004e-01 6.10946059e-01
5.75241268e-01 3.36107433e-01 9.67673123e-01 2.63152242e-01
-6.13527715e-01 6.21142328e-01 -5.49540222e-01 -4.20960039e-01
5.03020644e-01 -1.49514079e+00 3.66076767e-01 9.23581064e-01
1.54460728e-01 1.02132022e+00 6.07683778e-01 -7.23849535e-01
-1.46304500e+00 -8.78685236e-01 1.49423814e+00 -6.57632113e-01
8.92888427e-01 -2.77765721e-01 -1.02481472e+00 5.86872578e-01
5.98272085e-01 -7.43817985e-02 8.52274477e-01 6.02313280e-01
-5.97919106e-01 -1.32286862e-01 -7.31420636e-01 6.19660676e-01
1.42885709e+00 -6.82848811e-01 -1.04643869e+00 1.57536179e-01
1.34515631e+00 -4.18654263e-01 -8.16846848e-01 6.35225832e-01
5.17114639e-01 -1.05668557e+00 5.95204651e-01 -6.73962951e-01
4.84540135e-01 -4.25028771e-01 -1.50985330e-01 -1.38412392e+00
-3.03688079e-01 -6.74613237e-01 -6.95168495e-01 1.37596607e+00
9.49949145e-01 -8.53920937e-01 5.83658278e-01 5.44200659e-01
-4.62130994e-01 -1.32981062e+00 -8.92835379e-01 -1.84937611e-01
-2.00086772e-01 -4.80037510e-01 6.85102820e-01 8.31776679e-01
1.78832188e-01 1.05450487e+00 -3.10582727e-01 3.58973056e-01
2.43272126e-01 2.82747865e-01 1.35969758e-01 -1.17117810e+00
-3.56086522e-01 -3.27274472e-01 4.01668847e-02 -1.25928724e+00
3.25316399e-01 -9.57768261e-01 -1.44079402e-01 -2.12247825e+00
3.00751142e-02 -1.93586856e-01 -7.23562241e-01 5.36293447e-01
-5.83600938e-01 -5.26079945e-02 6.66478649e-02 1.94108307e-01
-1.09650433e+00 6.00015998e-01 1.27517641e+00 -1.55071348e-01
-8.01487491e-02 -6.67696074e-02 -1.38181448e+00 8.30460489e-01
5.10806501e-01 -4.34983432e-01 -8.49557281e-01 -6.64500475e-01
9.81852233e-01 8.12632143e-02 -2.58449674e-01 -3.71192783e-01
5.40385664e-01 1.40361890e-01 -4.09835398e-01 -7.77641296e-01
2.16044977e-01 -6.79287493e-01 -1.60671681e-01 6.53196722e-02
-2.69958019e-01 5.27791440e-01 -2.22199678e-01 7.26427317e-01
-4.19079691e-01 -1.34909794e-01 4.20488678e-02 -9.29004513e-03
-7.03783333e-01 2.96904653e-01 -1.24519609e-01 -5.93507998e-02
6.65331006e-01 1.92332849e-01 -9.23940778e-01 -1.91054806e-01
-5.18231630e-01 5.33977151e-01 2.45347708e-01 4.29469466e-01
5.68508565e-01 -1.08685052e+00 -5.57798982e-01 -8.28253031e-02
2.49970123e-01 1.63315997e-01 3.28410208e-01 9.50330138e-01
-1.93159923e-01 2.90178001e-01 3.43418509e-01 -9.96432453e-02
-1.33381367e+00 4.99181509e-01 3.20310056e-01 -1.26895225e+00
-8.85622427e-02 1.10377598e+00 4.97523323e-02 -5.46331942e-01
-1.77961722e-01 -4.77289528e-01 -7.80980408e-01 4.05686826e-01
1.97907746e-01 1.76622212e-01 3.53704035e-01 -9.03034389e-01
-5.14428735e-01 5.80155790e-01 -2.02094629e-01 3.11361291e-02
9.72673118e-01 -1.55780628e-01 -9.12914991e-01 2.74301112e-01
1.00245249e+00 3.69744420e-01 -3.87146711e-01 -1.63118154e-01
1.45983264e-01 -5.79842888e-02 7.95814581e-03 -8.54413867e-01
-9.93765175e-01 5.06018162e-01 2.60763206e-02 7.00669110e-01
1.18050110e+00 6.88643456e-02 8.19144905e-01 4.18687373e-01
4.84425098e-01 -6.21379912e-01 -1.68516189e-01 1.04415274e+00
7.37276852e-01 -8.26049447e-01 6.59094751e-02 -8.66598427e-01
-7.53033519e-01 7.45217383e-01 7.36509979e-01 1.88241571e-01
7.02135026e-01 -4.01290208e-02 1.11678995e-01 -6.70834422e-01
-1.29672909e+00 -6.28631830e-01 5.65369129e-01 2.75378436e-01
6.78804696e-01 -6.44649146e-04 -8.97443354e-01 8.49152982e-01
-9.11610201e-02 -3.76448482e-01 1.06593639e-01 1.09607518e+00
-4.09672230e-01 -1.48683345e+00 -1.20343855e-02 4.77253228e-01
-6.51056111e-01 -6.95683181e-01 -6.03353441e-01 4.30983603e-01
4.62873966e-01 1.37132764e+00 -1.19890623e-01 -4.77771372e-01
6.88938200e-01 2.16743365e-01 1.78825229e-01 -8.26997161e-01
-1.01871026e+00 -6.14249408e-02 6.05159461e-01 -6.25529826e-01
-8.94151807e-01 -8.50441754e-02 -1.13604999e+00 -1.98232427e-01
-5.42515337e-01 5.07343411e-01 2.64212340e-01 1.09416914e+00
6.85711265e-01 8.25506806e-01 4.60959703e-01 -3.10154915e-01
-4.46580589e-01 -8.61386478e-01 -4.65056777e-01 2.23187804e-01
8.33021551e-02 -5.51549911e-01 -3.97939295e-01 -4.41927761e-01] | [11.457380294799805, 6.7039690017700195] |
ebcdec32-8ee8-4043-ab45-312f6c4f0476 | zhixiaobao-at-semeval-2022-task-10 | null | null | https://aclanthology.org/2022.semeval-1.187 | https://aclanthology.org/2022.semeval-1.187.pdf | ZHIXIAOBAO at SemEval-2022 Task 10: Apporoaching Structured Sentiment with Graph Parsing | This paper presents our submission to task 10, Structured Sentiment Analysis of the SemEval 2022 competition. The task aims to extract all elements of the fine-grained sentiment in a text. We cast structured sentiment analysis to the prediction of the sentiment graphs following (Barnes et al., 2021), where nodes are spans of sentiment holders, targets and expressions, and directed edges denote the relation types between them. Our approach closely follows that of semantic dependency parsing (Dozat and Manning, 2018). The difference is that we use pre-trained language models (e.g., BERT and RoBERTa) as text encoder to solve the problem of limited annotated data. Additionally, we make improvements on the computation of cross attention and present the suffix masking technique to make further performance improvement. Substantially, our model achieved the Top-1 average Sentiment Graph F1 score on seven datasets in five different languages in the monolingual subtask. | ['Yongliang Wang', 'Chong Yang', 'Jing Xu', 'Chen Liang', 'Yangkun Lin'] | null | null | null | null | semeval-naacl-2022-7 | ['semantic-dependency-parsing'] | ['natural-language-processing'] | [ 0.37650552 0.52611005 -0.27483326 -0.8275967 -0.7983042 -1.1254382
0.7286422 0.42257422 -0.50991863 0.7517616 0.66821676 -0.4715861
0.4507364 -0.48582068 -0.8972672 -0.20391372 -0.10518327 0.19505891
0.12565264 -0.67680985 0.3497858 -0.08656322 -0.8291558 0.7168327
0.5278644 1.05102 -0.03783771 0.7566005 -0.36421213 1.1707959
-0.5831623 -1.2873958 -0.0591805 -0.35073668 -1.0752741 -0.13573146
0.45880118 0.3575871 0.07218222 1.1228626 0.10550494 -0.21548098
0.7046863 -1.0793817 -0.6424729 1.1393049 -0.81600857 0.17561251
0.25940487 -0.41732037 1.591104 -0.96090627 0.7815906 1.0343516
0.7684076 0.633754 -0.8941127 -0.35212645 0.63634557 -0.14827022
-0.76846427 -0.5825888 0.5144253 -0.38398182 1.4410081 0.0982994
0.3996393 1.0797826 0.23293717 0.7844575 1.2035408 -0.5354154
0.05881073 0.21864817 0.49210408 0.77601796 0.2188255 -0.5171447
-0.7583447 0.05164456 -0.04147417 -0.60306096 0.13128778 -0.1961781
-0.9782468 0.83700323 0.38542998 0.18161756 -0.08372458 0.05800498
0.8157986 0.4539543 0.9005617 0.42499554 -1.240301 -0.02577419
-0.72974986 -0.14779213 1.1331737 1.0857291 0.49521932 -0.14955768
-0.03124527 0.67303437 0.27992335 0.33109075 0.37046054 -0.5001056
0.9032842 0.41040018 -0.17468594 -0.8411753 -0.42916387 -0.36601314
-0.41846964 -0.3167786 0.39606744 -0.64256585 -0.8562799 1.7603279
0.04659779 -0.3133275 0.36699703 0.59316164 0.8581862 0.40022582
0.34590316 0.11506473 1.5855058 -1.4546092 -0.58107775 -0.9006021
1.1552831 -0.85284406 1.1132523 0.18316685 -0.92701447 -0.15373866
-1.0183406 -0.23994017 -0.655168 -0.06963781 0.9829745 0.74427485
-1.2261884 0.42205232 -0.77440315 -0.25879455 0.47977403 0.39904302
-0.42161828 0.15270871 -1.2549078 0.96486247 0.02578031 0.08789116
-0.25513127 -0.50641495 -1.213591 -0.08428595 0.3214656 -0.42164475
1.342181 -1.5464239 -1.3551816 1.467213 -0.4505292 -0.5689242
0.0357141 -0.4338788 -0.2717415 -0.11566482 0.35761952 0.51335496
0.5203217 -0.88337934 -0.5723805 -0.49368536 0.32722571 0.2834039
-0.36614558 0.62690836 -0.5689367 -0.6352725 -0.2827232 -1.0496063
-0.2585666 -0.6965821 -0.65837795 -0.20157845 0.4160969 -0.79520243
0.9385977 -2.1095903 0.17692664 0.21574588 0.11273196 -0.12175972
-0.24919897 0.46013314 -0.3251885 0.37704265 -0.47272623 -0.757812
0.13669285 0.07282924 -0.32019335 0.35237122 0.34207067 1.223341
-0.7192122 -0.33631375 -0.36739746 0.37318736 -0.4168765 -0.09649035
-0.46727252 0.40696728 -0.54114354 0.50600135 0.37344223 -0.3323238
0.5846047 -0.11697048 0.02050357 0.97072273 -0.61000395 1.7218153
-0.74514914 0.58769464 0.3710905 -0.9989795 0.72089934 0.07744943
0.11635661 -0.70054305 0.17105225 0.325717 -0.05378165 -0.11387355
0.5734864 -0.2542083 -0.6923548 0.36662003 0.05328647 -0.36252955
0.45623526 0.55165565 1.0232873 0.3805489 0.37128967 -0.6117859
0.65552634 0.337637 0.2912098 0.3558253 0.02232998 0.40956867
1.1956887 -0.18172115 -0.8432774 -0.56949353 0.11380843 1.5087194
-0.18276246 -0.74955875 -0.89790547 -1.3218747 -0.19530843 0.81914556
-0.83655286 0.19914156 -0.70542395 -0.7268321 0.50775564 0.5962503
0.07929919 -1.1972128 -0.06329476 -0.06847234 -0.17054154 -1.44413
-0.59707785 0.5686726 -0.6344608 -0.9772549 -0.34215307 -1.2532166
0.62489504 -0.28330192 1.5935261 -0.01052918 0.30644676 0.11348167
-0.5353394 -0.5864329 -0.16852471 0.45129713 -0.4958896 -0.1852819
0.59486514 -0.1030167 -0.21036023 -0.2257947 -0.4241137 0.09420586
0.25306377 0.761825 0.68300885 -0.29930055 0.54242843 -1.8050252
0.6309212 -0.51360303 -0.5513284 0.13109215 -0.5827771 0.03390577
0.8915412 0.26309147 -0.961747 0.03192752 -0.45889258 0.5193132
0.08742883 0.74524724 -0.16185892 0.16467519 0.30433407 -0.22464728
-0.45748433 -0.27468058 0.52762955 0.640904 0.2797409 -0.6200537
0.35556224 0.31517503 -0.00608743 -0.44952777 -1.698356 -0.26972657
-0.77874106 0.1782218 1.0256487 -1.2052495 -0.31432196 0.46039602
-1.1437557 -0.6725495 -0.28079757 0.10730068 -0.3363804 0.20127663
-0.96822804 -0.49565485 -0.5577751 -0.8840087 1.3323787 -0.11346504
-0.5258627 -1.3766611 0.09815768 0.47905412 0.26920265 0.2153962
0.93553036 -0.9704456 0.02894215 -0.02984204 -0.22054228 0.50290155
-0.05924408 -0.208377 -0.9020783 -0.1396143 -0.01880202 -0.7044406
1.0235646 0.22530812 0.98645335 -0.29750812 -0.127392 0.54507095
1.3188839 -0.18475388 0.35158038 0.51506996 0.9666443 1.2297838
0.60425705 -0.10215502 0.8451191 0.25390944 0.31485176 -0.0870548
-0.1132062 -0.3681402 0.7283836 1.1743194 0.2259944 -0.50290096
-0.7475649 0.74862313 -1.6592363 -0.5077192 -0.5610962 1.6920599
0.8678964 0.585067 -0.05715576 -0.21835355 0.5457493 0.4464219
-0.29148504 -0.9381673 -0.46174008 0.63786167 0.76578283 0.89252627
-1.1241153 1.590412 5.9615097 0.45262593 -0.8994971 0.188379
0.86063206 -0.05462224 -0.6183834 0.09840565 -0.92958295 0.27369702
1.2403467 0.21722537 0.31295457 0.5355624 -0.25662816 0.01658776
-0.9064287 0.3730981 0.28272074 -0.99934995 -0.32531637 -0.34395632
0.8927871 0.6169055 0.02549586 0.35725552 0.6488924 -1.1239524
0.9080731 0.02204835 0.81239873 -0.7444023 0.84722507 0.02284357
-1.0872375 0.28687176 -0.0660173 -0.24087505 0.30960044 0.68423986
-0.4255142 0.42506772 0.6877794 1.1441461 -0.6491799 -0.01839523
-0.98288393 0.9461154 -0.09392103 -0.44560286 0.33527055 -0.29658765
0.1843459 1.4922034 -0.1308939 -0.1770587 -0.16175655 0.01775288
-0.5223048 0.4420138 -0.43833163 -0.37993166 -0.04033514 1.5422179
-0.97729933 -0.45076907 -0.8760342 1.0740092 0.71438116 0.2981419
-0.48230788 -0.448466 0.48697847 -0.13428734 0.6156093 -0.08356141
-0.70806384 -1.2909216 0.13355054 -0.99062085 0.5135921 -0.7481973
-1.4136938 0.8767696 -0.5156322 -0.26803914 -0.11912008 -0.97774494
-0.53274846 1.0169089 -1.7500405 -1.388237 0.31015655 0.44671354
0.44277614 -0.01228412 0.8602186 0.1236965 -0.40528467 0.5837514
-0.11393356 0.3671806 0.82058305 -1.5301567 1.0901004 0.8901862
0.2551703 0.6191296 0.60789883 -0.84243476 -1.2739196 -1.0717735
1.7412789 -0.84381485 1.3267164 -0.89636284 -0.46775135 1.257784
0.6421425 -0.21874286 0.80252945 0.5350115 -0.6026003 0.2557135
-0.76440626 0.5681084 1.1829225 -0.7226754 -0.537999 0.41978076
0.9602026 -0.4342587 -0.69892293 0.28970212 0.32923374 -0.80718607
0.3620438 -0.8746551 0.9410546 0.07435015 -0.34181333 -1.46289
0.1245384 -0.50285417 0.10153797 1.1829365 1.1954477 -0.65701425
0.8760672 0.33971906 -0.36771792 -0.9236378 -0.39767867 -0.19068363
0.34737542 -0.49320644 0.39730084 1.0226926 0.25738013 0.97511727
-0.08637928 -0.05742458 0.48384118 0.2380883 0.48566788 -0.96471375
-0.5159416 -0.27175575 0.04027615 -1.0257578 0.75358605 -1.1835704
-0.03542386 -1.7925605 0.0364321 -0.24605602 -0.34279218 0.6711927
-0.25367734 0.48389557 0.07440475 -0.14117922 -0.8186086 0.19370598
1.0599092 -0.06798662 0.23509452 -0.15720871 -1.4067658 0.8475861
0.9001399 -0.66079026 -0.3551375 -0.7268939 1.0370384 -0.20143946
-0.11027406 -0.10632404 0.04104061 0.16308573 0.11348221 -0.34105727
0.14586397 -0.5247774 -0.5638955 0.27332765 -0.5488628 0.28995946
0.3629307 0.18787071 -0.43574005 -0.24727678 0.41399428 -0.09979629
-0.6502966 0.12465318 -0.10590167 0.5830748 0.7912575 0.27221933
-0.4939524 -0.30330333 -0.7118082 0.32827738 0.486822 0.505295
0.04051809 -0.7056991 -0.6819962 -0.04448352 0.11088776 -0.06674661
-0.1299022 0.8697501 -0.38171953 0.46896964 0.15066935 0.1077837
-1.2061121 0.24953714 0.09318897 -0.5322159 -0.21329173 1.3146237
0.18287231 -0.69521314 -0.22455336 -0.08080916 -0.55683213 0.1647417
0.33439738 -0.12588695 0.39399648 -0.82262963 -0.74933064 0.5793953
-0.2177998 -0.23556468 1.5167803 -0.21326047 -0.5825328 0.4400755
1.3983469 0.7697359 -0.8801573 -0.07156681 0.44732967 0.22309443
-0.12367854 -1.0541611 -1.0884069 0.7810155 -0.41215846 0.34129396
0.9489909 0.33441767 0.81090796 0.20211406 0.16659562 -1.0550637
-0.37348905 1.1323155 0.46915084 -1.2571449 -0.06617366 -0.60684526
-1.1529201 0.8481566 0.48769763 -0.26988515 0.63739455 0.66717356
0.3369063 -0.56396586 -0.81805176 -0.14859834 0.29351118 0.36058164
0.99538344 0.12494747 -0.6040848 0.9200319 -0.7092462 -0.5451494
0.41321087 0.771915 -0.08812556 -1.1806839 0.25490633 0.58185923
-1.1248237 -0.8061196 -0.8402039 0.5981245 -0.25745842 1.1280429
0.0449655 -0.28113773 0.51111007 0.14663225 0.45590279 -1.0327666
-0.8986489 -0.10588132 0.82341415 -0.475907 -0.6168429 -0.6679021
-1.448282 -0.052441 -0.11751058 0.4751244 1.0323582 1.1352066
0.2878687 0.6021546 0.47106707 -0.2857799 -0.2911389 -1.035161
-0.63143706 0.3086873 0.33034435 -0.02327031 -0.42590633 0.2318135 ] | [11.377893447875977, 6.849254131317139] |
1a20ba2c-df4b-4d7b-894c-b6b54ca33a0d | eyebag-accurate-control-of-eye-blink-and-gaze | 2306.17391 | null | https://arxiv.org/abs/2306.17391v1 | https://arxiv.org/pdf/2306.17391v1.pdf | EyeBAG: Accurate Control of Eye Blink and Gaze Based on Data Augmentation Leveraging Style Mixing | Recent developments in generative models have enabled the generation of photo-realistic human face images, and downstream tasks utilizing face generation technology have advanced accordingly. However, models for downstream tasks are yet substandard at eye control (e.g. eye blink, gaze redirection). To overcome such eye control problems, we introduce a novel framework consisting of two distinct modules: a blink control module and a gaze redirection module. We also propose a novel data augmentation method to train each module, leveraging style mixing to obtain images with desired features. We show that our framework produces eye-controlled images of high quality, and demonstrate how it can be used to improve the performance of downstream tasks. | ['Wonjong Ryu', 'Jeong Young Jeong', 'Bryan S. Kim'] | 2023-06-30 | null | null | null | null | ['gaze-redirection', 'face-generation'] | ['computer-vision', 'computer-vision'] | [ 4.44455534e-01 2.34911174e-01 4.02020723e-01 -4.19703156e-01
-2.74624169e-01 -4.38544422e-01 6.83267534e-01 -7.76786029e-01
9.46466997e-02 7.04383969e-01 9.72889960e-02 -2.34531105e-01
2.05228344e-01 -4.64476138e-01 -6.46835625e-01 -7.10099757e-01
4.84066159e-01 -5.08174486e-02 -2.67974198e-01 -2.36726046e-01
2.77962625e-01 5.62461019e-01 -1.96007180e+00 1.52684838e-01
8.34656417e-01 8.09903979e-01 -1.71461925e-01 7.83263087e-01
2.69311368e-01 6.29758358e-01 -6.77817106e-01 -3.01143169e-01
2.40109578e-01 -8.00519228e-01 -4.42653805e-01 6.17125928e-01
8.03576648e-01 -4.37416822e-01 -1.01358168e-01 6.75624669e-01
7.36976147e-01 1.08694263e-01 5.83149552e-01 -1.77277648e+00
-1.17386389e+00 -2.13239297e-01 -8.24293435e-01 -3.93837504e-02
3.23015511e-01 4.12950933e-01 4.85693634e-01 -1.01062894e+00
4.87496674e-01 1.34003830e+00 3.04470479e-01 1.34197247e+00
-1.47532916e+00 -9.23554540e-01 6.81779832e-02 -2.02983501e-03
-1.13334906e+00 -1.15238833e+00 6.62074029e-01 -2.89357245e-01
5.80897689e-01 9.54720080e-02 5.63899815e-01 1.43561375e+00
-1.65821309e-03 6.18685365e-01 1.34568095e+00 -5.39291680e-01
-5.31950593e-02 1.27105057e-01 -4.94786590e-01 7.14792073e-01
1.04250133e-01 3.64622682e-01 -8.46113801e-01 1.31365359e-01
1.10301006e+00 -2.92159110e-01 -5.36340296e-01 -1.92548245e-01
-9.03580010e-01 6.22740924e-01 4.32329774e-01 -2.60667115e-01
-1.67510450e-01 8.94889459e-02 -3.79607499e-01 1.45895228e-01
7.29270160e-01 3.33508044e-01 -1.43217131e-01 1.75827697e-01
-9.20214057e-01 3.46727550e-01 5.85888505e-01 1.31431186e+00
6.52396441e-01 1.43237948e-01 -5.01210630e-01 7.42333353e-01
6.37199163e-01 6.42683446e-01 1.57947168e-01 -1.04998016e+00
1.29144236e-01 3.90272796e-01 2.54738569e-01 -5.26251256e-01
-2.15740472e-01 -9.48483050e-02 -6.60179496e-01 7.74707198e-01
4.04998690e-01 -2.78673083e-01 -1.39138472e+00 2.07805395e+00
5.49940169e-01 3.05596203e-01 -1.43926963e-01 8.05649757e-01
7.18859136e-01 4.59914833e-01 1.94733098e-01 -1.60345882e-01
1.21815193e+00 -1.06427848e+00 -8.56233716e-01 -2.72397965e-01
2.02415034e-01 -8.76918733e-01 1.06865978e+00 2.87520558e-01
-1.38660288e+00 -6.25258744e-01 -8.50115895e-01 -4.41905707e-01
7.81602487e-02 2.04863220e-01 5.44406652e-01 9.26429570e-01
-1.61394179e+00 2.55043149e-01 -6.55905664e-01 -1.93454936e-01
7.71741390e-01 3.01958442e-01 -4.89924341e-01 1.92985654e-01
-5.71750700e-01 6.74472988e-01 -2.37158373e-01 1.69005617e-01
-9.44405437e-01 -9.53237534e-01 -1.10543513e+00 4.55989689e-02
1.09987445e-01 -1.19007123e+00 1.35792971e+00 -1.09937775e+00
-1.89308870e+00 1.08185077e+00 -6.39236629e-01 1.74435694e-03
4.24965143e-01 -1.70545861e-01 -4.10226524e-01 -4.19437140e-02
-1.11861877e-01 1.16947353e+00 1.64613605e+00 -1.46077371e+00
-6.62014902e-01 -4.40764844e-01 -1.01693332e-01 1.00254171e-01
-3.47712934e-01 2.31476128e-01 -4.92417574e-01 -6.03582621e-01
-3.37398887e-01 -1.06384420e+00 1.04492761e-01 3.75920177e-01
-4.60944533e-01 -5.18092848e-02 1.00781345e+00 -3.90275836e-01
1.04288208e+00 -2.00423050e+00 8.20642784e-02 -1.08534761e-01
5.71870208e-01 3.48788619e-01 -4.44551587e-01 9.03725699e-02
-3.57805133e-01 1.60216480e-01 -1.19526871e-01 -8.85520756e-01
-2.45531037e-01 -1.37853459e-01 -2.41112590e-01 1.71601683e-01
7.65388131e-01 1.02952409e+00 -6.36826754e-01 -3.60643923e-01
1.09097604e-02 8.09282422e-01 -7.57506251e-01 3.75222087e-01
-1.71765327e-01 8.60023797e-01 8.52890022e-04 6.35529995e-01
8.39067042e-01 -2.93628693e-01 -2.31567025e-01 -2.07827576e-02
-7.33185634e-02 -7.39884377e-02 -6.50589168e-01 1.43574822e+00
-5.78259826e-01 7.70507753e-01 1.20468870e-01 3.97619121e-02
8.30079973e-01 2.24713817e-01 8.78254473e-02 -4.86296415e-01
3.15509319e-01 -3.08564812e-01 3.12730558e-02 -5.47034323e-01
3.10313553e-01 -1.93853363e-01 5.77889025e-01 7.04895258e-01
1.08718507e-01 -2.22169429e-01 1.13053575e-01 3.98052484e-02
6.87842607e-01 5.55368006e-01 -1.51261196e-01 -7.23065063e-03
5.29122293e-01 -2.61284113e-01 3.27731580e-01 2.93681711e-01
-2.41965950e-01 1.10137272e+00 3.78100574e-01 6.26536971e-03
-1.06536365e+00 -1.02111173e+00 1.05207292e-02 1.08830595e+00
-1.34726197e-01 -2.42027581e-01 -1.14906478e+00 -5.48017025e-01
-4.40601930e-02 7.21793234e-01 -8.08251262e-01 -4.29138273e-01
-4.03450459e-01 -6.79138362e-01 4.34916884e-01 6.00258350e-01
4.13616270e-01 -1.17882264e+00 -4.49164212e-01 -3.04897815e-01
9.30675715e-02 -1.03997159e+00 -7.44755507e-01 -6.21327639e-01
-5.79338908e-01 -1.05997229e+00 -9.11424875e-01 -5.78672349e-01
1.15805805e+00 5.25418937e-01 1.16272736e+00 2.50056624e-01
-3.56941223e-01 4.21038091e-01 -5.07284589e-02 -6.38954103e-01
-6.00190639e-01 -1.71477571e-01 1.55269392e-02 4.58374888e-01
2.93866217e-01 -4.45044190e-01 -9.58869100e-01 2.36329585e-01
-9.27376688e-01 3.94577473e-01 3.66812199e-01 7.65825272e-01
2.41021723e-01 -5.64571321e-01 6.16080582e-01 -9.43033338e-01
8.67034554e-01 -2.02863336e-01 -7.54001141e-01 1.57624960e-01
-8.06477547e-01 5.68457246e-02 2.70279378e-01 -4.19950902e-01
-1.52388382e+00 -1.79934427e-02 -8.57753158e-02 -5.95698118e-01
-3.59976321e-01 -2.29392678e-01 -3.18787217e-01 -2.71159381e-01
8.03723574e-01 8.55487138e-02 4.26604033e-01 -4.01584655e-01
4.10435885e-01 7.23371685e-01 4.47061211e-01 -3.13958824e-01
1.01392579e+00 6.15461051e-01 1.45676881e-01 -7.49800682e-01
-1.01338792e+00 1.57334879e-01 -4.60666835e-01 -2.62809962e-01
7.59032011e-01 -1.03806496e+00 -1.04743016e+00 8.20508480e-01
-9.66549814e-01 -4.11217928e-01 -1.55598238e-01 6.02682345e-02
-7.55572796e-01 -1.61172748e-01 -3.32332760e-01 -8.56365919e-01
-3.39923561e-01 -9.92623627e-01 1.45341313e+00 7.83532858e-01
-5.73802963e-02 -7.40085602e-01 6.68614358e-02 5.07169306e-01
5.18247187e-01 2.10939944e-02 7.12184489e-01 -1.86070763e-02
-5.86558819e-01 4.61797304e-02 -3.27647150e-01 3.64043742e-01
3.29377025e-01 4.01650667e-01 -1.46163738e+00 -4.67543364e-01
-8.64954591e-02 -3.42009515e-01 6.50360763e-01 3.76997948e-01
1.22073138e+00 2.53335387e-02 -4.72406685e-01 8.67987990e-01
8.67720962e-01 9.16111693e-02 9.02272105e-01 2.73233242e-02
7.95192957e-01 6.68485343e-01 3.66104782e-01 2.72391617e-01
3.86896014e-01 5.66622794e-01 1.87197939e-01 -4.04304028e-01
-7.34694004e-01 -4.28640157e-01 2.08416536e-01 -5.87049499e-02
-2.24508733e-01 -4.31810588e-01 -6.95256114e-01 4.39648449e-01
-1.54533112e+00 -8.34199369e-01 -6.34263977e-02 2.13692999e+00
9.64285254e-01 -2.43230566e-01 1.67406335e-01 -1.44212186e-01
7.70066679e-01 -1.59835652e-01 -7.25219429e-01 -2.16017410e-01
1.45374700e-01 4.18106496e-01 -5.09943552e-02 3.53783399e-01
-8.66507947e-01 1.14279103e+00 7.39271307e+00 1.14765249e-01
-1.11978579e+00 -2.00620499e-02 7.78358221e-01 -3.46354872e-01
-4.28464741e-01 -1.03182442e-01 -1.00227654e+00 4.46922541e-01
8.00802648e-01 -1.13325134e-01 4.17325646e-01 5.17806053e-01
5.15534222e-01 -2.19082087e-02 -1.29314971e+00 1.14948452e+00
4.04411048e-01 -1.24704158e+00 2.82276869e-01 2.90083528e-01
8.12770247e-01 -5.27118444e-01 4.57246929e-01 2.21679006e-02
2.20835671e-01 -1.17307580e+00 5.16357422e-01 7.18702912e-01
1.21216023e+00 -7.73369968e-01 -1.35980368e-01 6.00968078e-02
-6.99256718e-01 9.06120762e-02 -5.64004481e-02 1.85521290e-01
1.52476206e-01 2.41244048e-01 -8.23252082e-01 1.58548310e-01
5.03858864e-01 4.82191086e-01 -7.60295570e-01 9.65884268e-01
-5.80185711e-01 2.07763672e-01 2.58335117e-02 3.33122462e-01
-2.66719848e-01 -1.71380416e-01 4.71332103e-01 5.78262091e-01
4.27558511e-01 -2.30617020e-02 -4.75541294e-01 1.37282586e+00
-2.71807671e-01 -3.50239635e-01 -7.35833645e-01 7.70169646e-02
4.90158200e-01 1.46427727e+00 -1.98100388e-01 -4.37988788e-02
-3.33905220e-01 1.22307682e+00 3.21058661e-01 6.34704649e-01
-8.24497759e-01 -2.21928984e-01 1.21912837e+00 2.61794358e-01
5.32259755e-02 -1.70899540e-01 -2.24179104e-01 -1.21652603e+00
-1.34228945e-01 -9.31823313e-01 -1.39617175e-02 -1.17088020e+00
-1.10964918e+00 7.34879971e-01 -2.16546208e-01 -1.06486475e+00
-4.21153337e-01 -6.97926819e-01 -6.84805691e-01 1.22296393e+00
-1.79753602e+00 -1.49126768e+00 -7.28662312e-01 9.64260459e-01
4.52857673e-01 -2.05639303e-01 6.92337811e-01 1.13959022e-01
-7.74485171e-01 7.91227758e-01 -4.19659883e-01 -9.59684476e-02
1.02714682e+00 -1.25003731e+00 6.72349751e-01 9.81581867e-01
1.91658363e-02 6.31867945e-01 5.01919210e-01 -4.25067455e-01
-1.26968396e+00 -1.07630432e+00 6.13333166e-01 -6.58844829e-01
1.48957789e-01 -5.03779769e-01 -5.97324848e-01 8.99285197e-01
4.43686128e-01 7.88044631e-02 7.64156580e-01 -4.65224124e-03
-3.18357170e-01 -1.09502845e-01 -1.15911305e+00 9.94376004e-01
1.28938437e+00 -3.79837066e-01 -2.24171653e-01 1.54086009e-01
3.91258687e-01 -4.76462573e-01 -3.77622604e-01 2.85468847e-01
4.75611597e-01 -1.10591221e+00 8.18675935e-01 -8.87211442e-01
5.38874328e-01 -3.20402652e-01 5.88436007e-01 -1.50205445e+00
-2.72580892e-01 -1.24781311e+00 -3.47058177e-01 1.45112729e+00
3.37089986e-01 -6.19659424e-01 7.76257455e-01 9.26445484e-01
5.70463808e-03 -2.55321175e-01 -4.09960449e-01 -4.90570337e-01
-2.19213963e-02 -1.07450215e-02 6.82547271e-01 4.33171928e-01
-3.21478605e-01 6.18193626e-01 -4.74766344e-01 9.53711346e-02
8.12754214e-01 2.44785041e-01 1.18783319e+00 -1.18177938e+00
2.13935953e-02 -3.44430923e-01 -9.07150581e-02 -1.04676998e+00
2.03888535e-01 -5.16449630e-01 -7.98226707e-03 -1.16324866e+00
1.44919410e-01 -1.81510881e-01 1.21073782e-01 5.00153661e-01
-5.35420060e-01 6.27874911e-01 2.27314070e-01 7.07371384e-02
5.83474301e-02 5.55967093e-01 1.74046111e+00 3.53572637e-01
-2.69909918e-01 1.57495722e-01 -1.19885147e+00 5.67892373e-01
6.59207463e-01 -1.86587662e-01 -7.77572632e-01 -5.68307757e-01
1.57526031e-01 -2.10566178e-01 6.03266239e-01 -8.72175932e-01
1.46772891e-01 -2.82340087e-02 8.69127452e-01 -1.68080971e-01
4.57332760e-01 -5.24892092e-01 -4.93072346e-02 2.31230985e-02
-1.80221632e-01 -5.90592436e-02 3.84292990e-01 5.72945774e-01
-4.50607669e-03 7.98118412e-02 1.02143407e+00 1.47369981e-01
-2.88883418e-01 6.13848329e-01 -2.63763815e-01 -1.21298335e-01
1.23809767e+00 -2.87789464e-01 -4.70151424e-01 -5.39133251e-01
-8.17182660e-01 2.01459736e-01 6.34250402e-01 7.32802808e-01
7.57189929e-01 -1.31116569e+00 -8.81609321e-01 8.59106183e-01
2.57751137e-01 -8.70490819e-02 1.96523085e-01 7.37237215e-01
-2.84176111e-01 6.96722940e-02 -3.72167259e-01 -6.27478421e-01
-1.47811306e+00 4.89506811e-01 3.66913974e-01 4.15966928e-01
-3.95803779e-01 1.03475177e+00 5.98271728e-01 6.49396852e-02
-1.27995098e-02 9.15347133e-03 -2.57177830e-01 -2.25987304e-02
9.18960035e-01 1.58759907e-01 6.61628693e-02 -6.51682258e-01
-7.60531723e-02 4.18798953e-01 -9.49866176e-02 -1.57638133e-01
1.29856300e+00 -5.30240834e-01 7.36684874e-02 -3.87482531e-02
7.10921645e-01 1.17726564e-01 -1.83469820e+00 1.56430289e-01
-4.66290206e-01 -7.97252357e-01 4.05498929e-02 -8.88713419e-01
-1.14576888e+00 8.50896716e-01 5.89869618e-01 1.18799426e-01
1.45763636e+00 -1.83227614e-01 6.96589172e-01 -1.43311396e-01
1.04243569e-01 -7.88236022e-01 1.14742249e-01 2.06508860e-01
1.07243729e+00 -1.47866750e+00 -4.46103454e-01 -5.23745358e-01
-5.89220166e-01 1.00855184e+00 9.25605059e-01 1.17358133e-01
6.97540164e-01 4.03402001e-01 3.77674282e-01 -3.38764757e-01
-1.02689850e+00 -5.41983962e-01 5.26929498e-01 8.60166192e-01
5.44254005e-01 -3.87076348e-01 6.42490387e-02 -7.17786420e-03
-1.75799459e-01 2.68096238e-01 5.21438837e-01 7.79909849e-01
-1.14198484e-01 -1.14119852e+00 -5.30166745e-01 1.89465120e-01
-3.51136714e-01 -1.08233899e-01 -5.31179965e-01 7.75926232e-01
-8.55375752e-02 1.22997177e+00 3.51259053e-01 -1.65590480e-01
2.14673176e-01 2.13362888e-01 9.36076820e-01 -8.64112496e-01
-2.44985566e-01 1.43745914e-01 -1.29034296e-01 -5.87886810e-01
-3.97915632e-01 -4.46271956e-01 -6.01130188e-01 -5.11480689e-01
-3.90317619e-01 -3.94041777e-01 5.16676366e-01 7.10897863e-01
8.23140681e-01 4.66779888e-01 7.67007291e-01 -1.07253098e+00
-2.92724937e-01 -1.03981483e+00 -3.91556442e-01 5.38763762e-01
4.91353095e-01 -7.10834444e-01 -3.16532046e-01 5.29989123e-01] | [13.958250045776367, -0.009716873057186604] |
ee6bf209-43f0-41a8-871a-eed9d3710edf | source-aware-embedding-training-on | 2307.04336 | null | https://arxiv.org/abs/2307.04336v1 | https://arxiv.org/pdf/2307.04336v1.pdf | Source-Aware Embedding Training on Heterogeneous Information Networks | Heterogeneous information networks (HINs) have been extensively applied to real-world tasks, such as recommendation systems, social networks, and citation networks. While existing HIN representation learning methods can effectively learn the semantic and structural features in the network, little awareness was given to the distribution discrepancy of subgraphs within a single HIN. However, we find that ignoring such distribution discrepancy among subgraphs from multiple sources would hinder the effectiveness of graph embedding learning algorithms. This motivates us to propose SUMSHINE (Scalable Unsupervised Multi-Source Heterogeneous Information Network Embedding) -- a scalable unsupervised framework to align the embedding distributions among multiple sources of an HIN. Experimental results on real-world datasets in a variety of downstream tasks validate the performance of our method over the state-of-the-art heterogeneous information network embedding algorithms. | ['Guosheng Yin', 'Jiajun Shen', 'Chi Ho Wong', 'Tsai Hor Chan'] | 2023-07-10 | null | null | null | null | ['graph-embedding', 'representation-learning', 'network-embedding', 'recommendation-systems'] | ['graphs', 'methodology', 'methodology', 'miscellaneous'] | [-5.16922362e-02 6.02542698e-01 -7.98826516e-01 -1.19113527e-01
-1.31090760e-01 -5.90617359e-01 6.86978400e-01 5.42794228e-01
2.36855745e-01 4.56006825e-01 7.14922667e-01 -3.61485958e-01
-6.29617572e-01 -1.26872480e+00 -2.39906356e-01 -4.47845221e-01
-2.40299612e-01 5.76941013e-01 3.06687564e-01 -3.04673463e-01
3.26176360e-03 1.98240757e-01 -7.99181819e-01 2.40062073e-01
4.67347682e-01 3.69056851e-01 -1.37190506e-01 3.76087993e-01
-6.55090928e-01 9.63992894e-01 -2.95962811e-01 -6.71011567e-01
3.21002394e-01 -4.29029167e-01 -8.39318097e-01 -4.45956707e-01
3.16557318e-01 1.44020945e-01 -1.28516507e+00 1.24541712e+00
3.97615254e-01 -1.60399467e-01 7.64642596e-01 -1.73451829e+00
-9.67501044e-01 1.28377891e+00 -5.52492201e-01 4.39137936e-01
1.49059489e-01 -3.41868192e-01 1.68723226e+00 -4.11383629e-01
1.04881549e+00 1.34185696e+00 7.58352935e-01 9.64084640e-03
-1.23658574e+00 -4.68289196e-01 2.83959001e-01 1.78666383e-01
-1.22419298e+00 1.24381401e-01 1.08558977e+00 -4.47409570e-01
8.16947758e-01 2.39637762e-01 8.09816182e-01 1.09952521e+00
2.17077062e-01 6.32686853e-01 6.59476280e-01 -1.28597617e-01
-5.41716814e-04 1.14689805e-01 5.19690394e-01 8.51239085e-01
8.95300865e-01 -2.79981852e-01 -5.96723378e-01 -4.16982949e-01
6.03476882e-01 4.01695073e-01 -6.50574937e-02 -6.52927041e-01
-1.29006636e+00 1.05179942e+00 9.92110848e-01 8.11491609e-01
-1.59483999e-01 2.57701963e-01 6.74171448e-01 5.03399491e-01
7.89267719e-01 2.93933928e-01 -2.50405073e-01 2.47995973e-01
-4.32916701e-01 -2.17937306e-01 1.02370870e+00 1.18903470e+00
9.61104393e-01 -1.20475173e-01 5.64622357e-02 5.89656234e-01
4.59924996e-01 3.58169083e-03 2.01602370e-01 -5.48371911e-01
8.23151231e-01 1.07325327e+00 -6.98665023e-01 -1.93042028e+00
-3.07932079e-01 -3.26199234e-01 -1.07919240e+00 -5.08093894e-01
1.70386702e-01 1.11127138e-01 -3.09676260e-01 1.47960353e+00
4.11080927e-01 2.95062393e-01 -1.43833041e-01 5.71125865e-01
1.00775433e+00 4.65101242e-01 4.86248769e-02 2.93659538e-01
1.06892872e+00 -1.18036985e+00 -6.78378284e-01 -7.04551488e-02
7.26970434e-01 -3.06853205e-01 5.15015006e-01 -5.91949344e-01
-6.15624249e-01 -2.66791910e-01 -8.93612087e-01 -1.36513086e-02
-7.89808452e-01 -6.73520982e-01 1.07222116e+00 3.40467185e-01
-1.04839432e+00 6.78817987e-01 -4.41442162e-01 -7.90645778e-01
4.96310860e-01 1.55933812e-01 -6.81214213e-01 -3.52441609e-01
-1.26742625e+00 2.94322342e-01 3.91583681e-01 -1.80260509e-01
-4.53313291e-01 -1.14787793e+00 -9.62857366e-01 4.22746480e-01
3.03869337e-01 -1.00973105e+00 3.61747533e-01 -7.34285653e-01
-9.20481682e-01 6.02067411e-01 -3.96491624e-02 -7.94808269e-02
-2.89950073e-02 3.25906366e-01 -7.20620692e-01 2.62218535e-01
1.96675047e-01 2.73434460e-01 5.41563809e-01 -1.11698079e+00
-2.24798203e-01 -4.08652365e-01 1.84835345e-01 3.72903608e-02
-8.92394245e-01 -1.20300397e-01 -3.67045492e-01 -6.37856305e-01
1.71636865e-01 -9.55881953e-01 -1.36625260e-01 7.54603297e-02
-8.23477089e-01 -4.15319830e-01 9.70571041e-01 -4.03652489e-01
1.43949628e+00 -2.10377169e+00 3.86570305e-01 6.10793710e-01
1.02679467e+00 -7.15156123e-02 -4.98939335e-01 9.16051507e-01
-3.14294994e-01 5.42498469e-01 1.65600821e-01 9.82593074e-02
1.89333051e-01 1.13603689e-01 1.45280119e-02 5.47554910e-01
-8.16781893e-02 1.12928724e+00 -1.29885781e+00 -5.74590087e-01
3.73289958e-02 4.63237375e-01 -5.04826903e-01 3.37408245e-01
1.06062151e-01 8.77076238e-02 -6.28961265e-01 5.62475741e-01
3.21129382e-01 -8.44370246e-01 8.52861226e-01 -5.99705160e-01
3.67507219e-01 2.88492888e-01 -9.39429641e-01 1.57592010e+00
-6.83569536e-02 9.16799843e-01 -2.63039947e-01 -1.21794093e+00
6.57508731e-01 1.52154833e-01 9.81212616e-01 -4.50672626e-01
1.88154134e-03 -1.62028477e-01 3.01390737e-01 -3.41853440e-01
3.07010889e-01 1.42194271e-01 -8.81711394e-02 5.57494938e-01
3.98587584e-01 4.29526567e-01 2.84677655e-01 1.00459647e+00
1.66234684e+00 -6.44061208e-01 2.42864802e-01 -5.68734646e-01
-2.05085799e-03 -2.63377279e-01 3.07341427e-01 6.47366703e-01
-2.77538270e-01 3.07443291e-01 8.71788502e-01 -3.93161029e-01
-1.20165706e+00 -1.15519297e+00 9.26473588e-02 1.01122475e+00
2.90956259e-01 -9.98760045e-01 -3.99031699e-01 -1.15343022e+00
3.95943105e-01 -9.46701169e-02 -7.69526124e-01 -3.00865114e-01
-3.07125956e-01 -6.64166212e-01 4.63959843e-01 4.10363495e-01
1.21500291e-01 -6.90889359e-01 7.90698290e-01 8.10315460e-02
-1.05903819e-01 -1.08308589e+00 -6.25520289e-01 -1.98970735e-01
-8.27459335e-01 -1.63194108e+00 -2.96770692e-01 -9.36991513e-01
8.32908988e-01 6.27254844e-01 1.36970878e+00 2.47752696e-01
-2.88951039e-01 8.22153926e-01 -1.77796990e-01 3.06781381e-01
-2.65653908e-01 5.17876685e-01 -9.34280977e-02 -9.55485553e-02
4.31367725e-01 -7.01968193e-01 -5.85284948e-01 1.70983672e-01
-1.08326542e+00 -2.06375048e-01 4.95520204e-01 7.12704003e-01
2.27502540e-01 2.64317840e-01 6.42941356e-01 -1.59870577e+00
7.74955451e-01 -1.23655617e+00 -1.59375429e-01 4.13336664e-01
-7.38964081e-01 1.74714878e-01 5.65777004e-01 -4.07216161e-01
-6.11154497e-01 -7.93879032e-01 3.73476207e-01 -2.01338619e-01
1.28453761e-01 8.62056911e-01 -2.93813616e-01 -3.54970872e-01
3.59304637e-01 -2.50873089e-01 -1.51089594e-01 -3.06169480e-01
6.06459975e-01 4.95234102e-01 -5.76736331e-02 -4.43383127e-01
9.14167225e-01 3.54416937e-01 2.98467398e-01 -9.34658527e-01
-8.71761322e-01 -8.43586147e-01 -5.85978150e-01 -6.50784224e-02
6.07439876e-01 -9.69382465e-01 -4.15261179e-01 -3.61624360e-02
-8.56199622e-01 2.40923576e-02 -7.61429369e-02 2.82095015e-01
-1.51877493e-01 5.80649495e-01 -8.97731423e-01 -1.35002527e-02
-2.05155939e-01 -6.10718608e-01 5.79436779e-01 3.13459113e-02
-2.18092203e-01 -1.81556606e+00 5.78159928e-01 4.49549377e-01
5.62143624e-01 1.40169382e-01 1.36197782e+00 -8.26963067e-01
-7.17356563e-01 -1.23542145e-01 -5.22923589e-01 -1.72598690e-01
3.70087296e-01 -9.08503123e-03 -5.75852513e-01 -3.73479009e-01
-8.88212860e-01 2.35170610e-02 7.46345699e-01 -5.15940562e-02
1.20206130e+00 -6.60222530e-01 -8.33497047e-01 6.93690419e-01
1.70329356e+00 -4.94893700e-01 6.37739480e-01 2.48304695e-01
1.29372525e+00 6.27272248e-01 -1.40763432e-01 1.96292728e-01
5.71606517e-01 4.41465437e-01 3.79358470e-01 -2.37714857e-01
-4.03675169e-01 -8.25407803e-01 2.33652622e-01 1.45866084e+00
4.15606722e-02 -5.15254855e-01 -6.94822192e-01 7.49545872e-01
-1.92219007e+00 -1.06326270e+00 -2.53287494e-01 1.70592153e+00
6.17348552e-01 -1.09741256e-01 1.27848998e-01 -2.64355958e-01
1.05452085e+00 6.66180432e-01 -5.09994507e-01 -4.31616828e-02
-2.09344119e-01 -1.68615088e-01 4.91044819e-01 3.54701996e-01
-9.49477792e-01 8.38759065e-01 6.55016947e+00 6.07982934e-01
-5.77651143e-01 1.50443152e-01 3.59450579e-01 2.13201404e-01
-1.05648494e+00 2.19908476e-01 -5.82623780e-01 5.67460239e-01
1.08882356e+00 -4.44030762e-01 4.83947605e-01 7.73785710e-01
-4.33302015e-01 6.73484921e-01 -1.10131299e+00 8.62113774e-01
1.88274719e-02 -1.63795698e+00 3.59520853e-01 1.97459251e-01
1.05446041e+00 4.83507097e-01 -1.46758080e-01 3.76589000e-01
5.84482193e-01 -7.47613370e-01 9.75448489e-02 3.73400509e-01
4.99098718e-01 -5.31103134e-01 7.82237232e-01 -6.98814243e-02
-1.74015367e+00 1.68166250e-01 -6.80370927e-01 3.33310217e-01
8.04014951e-02 9.81841445e-01 -7.66962409e-01 9.19749081e-01
4.61096048e-01 1.40122437e+00 -6.73006117e-01 8.15986276e-01
-1.14557587e-01 5.79694927e-01 -3.60607617e-02 1.72658470e-02
1.64824873e-01 -5.07789016e-01 5.84572136e-01 1.18227792e+00
1.25754088e-01 -1.57066420e-01 1.44389942e-01 8.29406142e-01
-6.24348640e-01 1.00839615e-01 -1.29402530e+00 -8.67979527e-01
6.20576739e-01 1.45827508e+00 -6.13079369e-01 -8.69833976e-02
-8.40606987e-01 7.83445477e-01 8.68995845e-01 5.79976916e-01
-5.63796639e-01 -5.64218163e-01 8.70587707e-01 3.06256592e-01
1.06787398e-01 -2.38743469e-01 5.08351251e-02 -1.50266480e+00
-2.38016918e-01 -4.98367965e-01 8.26235414e-01 -4.15531695e-01
-2.08184123e+00 6.02333307e-01 -8.94231573e-02 -1.03666258e+00
1.03125811e-01 -5.18226504e-01 -7.19519675e-01 4.53269869e-01
-1.70904410e+00 -1.30221355e+00 -2.54668087e-01 7.71351874e-01
-6.08060136e-02 -4.48815823e-01 7.26779103e-01 5.28320611e-01
-6.94304585e-01 6.26287282e-01 3.18575203e-01 5.17302215e-01
4.94741112e-01 -1.24193347e+00 3.82641524e-01 4.60864305e-01
5.13406098e-01 8.41291726e-01 2.81229913e-01 -8.22889507e-01
-1.68921161e+00 -1.27706122e+00 9.92202401e-01 -2.54239649e-01
1.40006101e+00 -3.33773792e-01 -7.75854230e-01 9.63131428e-01
5.02148628e-01 4.40838099e-01 1.19208169e+00 6.00467682e-01
-8.65989327e-01 -1.65325806e-01 -1.01270306e+00 6.49906814e-01
1.54164004e+00 -9.06868100e-01 -2.35606954e-01 5.08344471e-01
9.67442334e-01 3.55418742e-01 -1.54902256e+00 2.37974286e-01
3.21081400e-01 -7.96922505e-01 1.24477839e+00 -1.25532317e+00
4.10721689e-01 -3.74010466e-02 -1.75746113e-01 -1.55638850e+00
-8.18252563e-01 -4.93984461e-01 -5.26660919e-01 1.26078475e+00
3.25213492e-01 -9.09323692e-01 8.87606442e-01 5.07897556e-01
9.49113742e-02 -4.63587046e-01 -5.78792572e-01 -6.22260451e-01
-1.15021348e-01 1.38490155e-01 6.89219296e-01 1.58124185e+00
5.02513111e-01 4.72614676e-01 -6.70451969e-02 3.04305166e-01
9.14841294e-01 1.61932796e-01 6.53896630e-01 -1.48182106e+00
-2.51471430e-01 -4.79489654e-01 -1.01428735e+00 -6.99166417e-01
6.91417515e-01 -1.50657487e+00 -7.94986010e-01 -1.82744765e+00
7.36246228e-01 -6.88327014e-01 -5.97648203e-01 4.86583084e-01
-1.72279164e-01 -1.65925914e-04 2.12512657e-01 2.82144934e-01
-9.57618117e-01 6.94134593e-01 1.08415449e+00 -6.07393622e-01
3.47318411e-01 -4.25139576e-01 -1.00609899e+00 4.09225523e-01
8.28479171e-01 -6.52091622e-01 -7.35108376e-01 -4.23775375e-01
6.15427494e-01 -2.07163006e-01 1.09083079e-01 -5.38713932e-01
3.21717173e-01 -9.50867087e-02 -6.27350947e-03 -1.21052906e-01
-1.05425097e-01 -9.45029855e-01 3.67275774e-01 2.10404783e-01
-4.67487335e-01 6.71231970e-02 -1.67633072e-01 1.15111434e+00
-3.33587468e-01 7.04053715e-02 3.40892553e-01 -9.42560583e-02
-6.84346080e-01 8.07622790e-01 3.47246155e-02 4.49686438e-01
9.86604691e-01 2.98487812e-01 -1.03058863e+00 -2.76669472e-01
-5.23943782e-01 -4.25625034e-02 4.33084309e-01 6.77399516e-01
6.78656220e-01 -1.62560999e+00 -4.85558510e-01 1.40069366e-01
3.49868178e-01 -3.88787657e-01 8.21298584e-02 6.42118871e-01
-3.19763422e-01 4.05073702e-01 3.32242623e-02 -2.62971252e-01
-1.04709256e+00 6.63402736e-01 1.55250311e-01 -4.78647709e-01
-7.84329057e-01 4.47700411e-01 1.76206291e-01 -7.35062122e-01
-2.85629984e-02 8.61869305e-02 -8.36484730e-02 3.77690271e-02
2.22823232e-01 5.36224663e-01 -1.78536087e-01 -4.74651426e-01
-2.43072227e-01 3.91946584e-01 -3.31045501e-02 5.08852839e-01
1.41400242e+00 -1.22248590e-01 -7.36940980e-01 5.04463792e-01
1.80757999e+00 -9.44712013e-03 -5.72953105e-01 -4.08076137e-01
1.46470055e-01 -6.87428772e-01 5.70571646e-02 -1.84172630e-01
-1.43616080e+00 5.33406079e-01 -1.96876526e-02 7.40395665e-01
4.31328118e-01 4.53078181e-01 7.94439971e-01 5.72656631e-01
5.30290306e-01 -6.79355681e-01 2.21951768e-01 3.05657357e-01
4.15063143e-01 -1.25590730e+00 -5.37396362e-03 -7.68076539e-01
-4.71244335e-01 1.01061118e+00 6.07996166e-01 -2.38626331e-01
1.33233786e+00 5.14906235e-02 -3.23552072e-01 -6.36280477e-01
-6.59259915e-01 -4.85457592e-02 5.08283257e-01 6.13813639e-01
4.20788586e-01 1.13030866e-01 -3.46331820e-02 2.50915825e-01
3.37894499e-01 -5.50358176e-01 3.38473797e-01 6.57530725e-01
-1.85215279e-01 -1.26496279e+00 3.78942013e-01 7.19502747e-01
-2.42143944e-01 -2.66401857e-01 -5.69644451e-01 7.10311472e-01
-3.68216157e-01 8.77411962e-01 3.76046821e-02 -5.08762181e-01
-6.50636107e-02 -2.33448222e-01 2.58440852e-01 -6.94260836e-01
-5.16757727e-01 -3.67193550e-01 2.10920334e-01 -7.07547843e-01
-4.49171066e-01 -2.52823114e-01 -1.04085541e+00 -7.69586802e-01
-1.81297973e-01 2.27973297e-01 3.86747539e-01 6.67439818e-01
8.21535885e-01 5.73800445e-01 7.77174234e-01 -4.67216790e-01
-2.06607148e-01 -7.81545758e-01 -1.13081300e+00 8.57046485e-01
3.24398316e-02 -5.64225614e-01 -7.47143745e-01 -3.47770780e-01] | [7.183073043823242, 6.239923477172852] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.