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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1794b19a-12e9-4696-adab-e5aca07d6cb5 | feature-encoding-with-autoencoders-for-weakly | 2105.10500 | null | https://arxiv.org/abs/2105.10500v3 | https://arxiv.org/pdf/2105.10500v3.pdf | Feature Encoding with AutoEncoders for Weakly-supervised Anomaly Detection | Weakly-supervised anomaly detection aims at learning an anomaly detector from a limited amount of labeled data and abundant unlabeled data. Recent works build deep neural networks for anomaly detection by discriminatively mapping the normal samples and abnormal samples to different regions in the feature space or fitting different distributions. However, due to the limited number of annotated anomaly samples, directly training networks with the discriminative loss may not be sufficient. To overcome this issue, this paper proposes a novel strategy to transform the input data into a more meaningful representation that could be used for anomaly detection. Specifically, we leverage an autoencoder to encode the input data and utilize three factors, hidden representation, reconstruction residual vector, and reconstruction error, as the new representation for the input data. This representation amounts to encode a test sample with its projection on the training data manifold, its direction to its projection and its distance to its projection. In addition to this encoding, we also propose a novel network architecture to seamlessly incorporate those three factors. From our extensive experiments, the benefits of the proposed strategy are clearly demonstrated by its superior performance over the competitive methods. | ['Lingqiao Liu', 'Ce Zhu', 'Fanxing Liu', 'Yanru Zhang', 'Xucheng Song', 'Yingjie Zhou'] | 2021-05-22 | null | null | null | null | ['supervised-anomaly-detection'] | ['computer-vision'] | [ 4.00327146e-02 -7.73631632e-02 -2.08570510e-02 -6.38135195e-01
-3.69524926e-01 -1.74166605e-01 3.29830408e-01 2.75885165e-02
-3.00678641e-01 4.14535016e-01 1.95616841e-01 -8.45851824e-02
1.39719009e-01 -6.70636415e-01 -3.93335938e-01 -7.60346889e-01
7.54358470e-02 2.31657490e-01 9.64156073e-03 1.15340650e-01
1.79585814e-01 6.58761501e-01 -1.33677256e+00 -5.72615117e-02
1.00167596e+00 1.28664708e+00 -3.90429258e-01 1.14170574e-01
-3.99291515e-01 7.59082496e-01 -4.35950756e-01 -4.61345427e-02
4.24877077e-01 -6.53966188e-01 -4.51818228e-01 3.54995728e-01
2.02595934e-01 -7.22550988e-01 -5.11649132e-01 1.28203130e+00
2.43061662e-01 3.22209597e-01 7.43279397e-01 -1.33949733e+00
-6.67719781e-01 1.93595737e-01 -6.68981671e-01 4.46677029e-01
1.81905627e-01 -6.03388646e-04 9.76234853e-01 -1.06355202e+00
2.67622232e-01 9.08154905e-01 4.41692293e-01 6.62810743e-01
-9.07159030e-01 -4.80865210e-01 4.80337381e-01 4.16226447e-01
-1.30301869e+00 -2.37589717e-01 1.18160748e+00 -3.82273227e-01
4.55643177e-01 2.23991632e-01 6.63626909e-01 1.09953499e+00
-1.12608001e-01 9.93130505e-01 6.51766002e-01 -2.82318056e-01
3.97714525e-01 1.18403904e-01 2.50971884e-01 7.39419281e-01
2.81826556e-01 1.36382561e-02 -1.04536332e-01 -3.07447195e-01
5.35984397e-01 6.86399996e-01 -3.97092611e-01 -6.82599962e-01
-8.31935704e-01 8.08009565e-01 5.31988502e-01 2.60082275e-01
-5.10152221e-01 -2.86595732e-01 4.26304519e-01 2.58570313e-01
4.35348660e-01 2.29978949e-01 -2.49780580e-01 3.60280871e-02
-6.51405454e-01 -9.58449394e-02 5.27771711e-01 5.05184710e-01
7.72585094e-01 5.44189453e-01 -6.97542652e-02 7.83305705e-01
6.99092925e-01 2.32488081e-01 9.60387647e-01 -4.04864848e-01
4.67615902e-01 1.06812108e+00 -1.40415415e-01 -1.23810208e+00
-2.54925221e-01 -5.23379028e-01 -1.00611877e+00 1.53817475e-01
3.77439409e-01 -2.20421746e-01 -9.30633843e-01 1.67504394e+00
4.19041395e-01 5.94121993e-01 2.52775371e-01 1.02622747e+00
4.21122789e-01 6.29579604e-01 -1.71788529e-01 -1.03863545e-01
7.96419263e-01 -8.24744701e-01 -6.93060279e-01 -2.20342115e-01
7.21468747e-01 -3.40820014e-01 9.51171458e-01 3.53806853e-01
-5.58024347e-01 -3.22320133e-01 -1.21087849e+00 2.53511459e-01
-2.68826455e-01 3.41130406e-01 4.28811640e-01 2.81400979e-01
-5.39448977e-01 4.94587004e-01 -1.07584786e+00 -2.51988113e-01
5.42770684e-01 5.76442853e-02 -5.02966523e-01 -1.87275648e-01
-9.46031928e-01 5.92400968e-01 5.24651110e-01 3.89455706e-01
-7.38019884e-01 -2.32448503e-01 -1.10441399e+00 1.58476695e-01
2.32753441e-01 -1.91950396e-01 8.92280459e-01 -1.09105027e+00
-1.26308095e+00 4.88431066e-01 5.42059019e-02 -4.29115206e-01
3.42738479e-01 -2.72156149e-01 -7.39442050e-01 3.59165929e-02
-5.21434359e-02 3.05366218e-02 1.04895270e+00 -9.61373985e-01
-6.51881397e-01 -6.89992309e-01 -4.12743956e-01 1.59970373e-01
-8.26761842e-01 -2.67025590e-01 -2.73435652e-01 -8.73018444e-01
7.85151005e-01 -6.58213794e-01 -3.76057118e-01 -1.29075740e-02
-4.99613404e-01 -9.86900479e-02 1.31300592e+00 -6.95074737e-01
1.13481081e+00 -2.43923068e+00 1.41332060e-01 5.81573844e-01
3.70289475e-01 4.03457999e-01 -4.65819128e-02 8.96406844e-02
-2.17695341e-01 -3.02683741e-01 -5.34106612e-01 -2.57413298e-01
-2.40761321e-02 4.78990346e-01 -4.77668613e-01 6.58742428e-01
4.24296290e-01 6.27889514e-01 -8.56572866e-01 -1.91617891e-01
2.86604136e-01 3.16580892e-01 -5.09740055e-01 5.41745901e-01
7.06050321e-02 5.18670321e-01 -7.73194909e-01 7.06994116e-01
7.34347105e-01 3.47722392e-03 -1.90435514e-01 -3.11121251e-02
3.27832758e-01 -1.05028369e-01 -1.37795627e+00 1.42963445e+00
5.18537313e-02 2.29286090e-01 -7.97369927e-02 -1.49337173e+00
1.37428057e+00 1.44772559e-01 6.00826323e-01 -4.47957397e-01
7.53307268e-02 3.46480638e-01 7.68721849e-02 -5.55022240e-01
1.06719002e-01 -1.14441626e-01 1.50745302e-01 5.65400958e-01
1.97643936e-01 5.48102796e-01 -1.71747655e-01 -4.58626309e-03
1.14606774e+00 -2.81230714e-02 3.04989010e-01 1.00719437e-01
8.97559643e-01 -4.33321178e-01 9.79979694e-01 3.99887145e-01
-4.91370082e-01 5.68016768e-01 4.03004050e-01 -8.30859542e-01
-8.36285830e-01 -1.06438911e+00 -1.48364022e-01 8.06600094e-01
6.56253546e-02 -1.95792601e-01 -6.01931393e-01 -1.20931923e+00
4.44238670e-02 6.27624214e-01 -6.10986590e-01 -7.46546805e-01
-5.46537161e-01 -7.78377116e-01 4.90979642e-01 7.71885872e-01
5.39584517e-01 -9.90545571e-01 -2.05861077e-01 7.57992268e-03
9.65708047e-02 -6.90615177e-01 -4.30430323e-01 2.88424551e-01
-1.03196430e+00 -1.11302030e+00 -4.76254731e-01 -7.69798279e-01
1.17003512e+00 2.65474580e-02 4.41982538e-01 1.54863745e-01
-1.96539685e-01 1.61788449e-01 -4.15721953e-01 -1.66297361e-01
-2.09181115e-01 -2.86325157e-01 3.83399040e-01 6.78902209e-01
7.50123620e-01 -5.66927075e-01 -5.53378522e-01 1.90153614e-01
-1.09995377e+00 -4.19022471e-01 6.85234785e-01 1.10597503e+00
6.92255676e-01 3.73786464e-02 6.11757398e-01 -7.53182769e-01
4.98702615e-01 -8.92688751e-01 -3.36250275e-01 -8.16861913e-02
-6.38559520e-01 2.31891766e-01 1.00307071e+00 -3.01128000e-01
-9.03903365e-01 1.68675229e-01 -3.98367226e-01 -8.07548583e-01
-4.39811647e-01 5.04283965e-01 -4.49018180e-01 6.22524694e-02
5.85820258e-01 5.06689310e-01 3.12216163e-01 -6.53841734e-01
1.25205070e-01 6.02397323e-01 4.96341079e-01 -4.44509029e-01
8.86987984e-01 4.02752191e-01 -1.97125450e-01 -4.66459185e-01
-8.98923516e-01 -5.90459764e-01 -8.44932675e-01 1.30070448e-01
4.72650081e-01 -5.98885536e-01 -8.62438381e-02 4.64028716e-01
-7.41498590e-01 2.42456406e-01 -6.35145485e-01 7.30837047e-01
-2.11930931e-01 7.46395171e-01 -4.85942423e-01 -7.00917244e-01
-3.02591681e-01 -9.68627453e-01 6.23322904e-01 1.45500675e-01
1.19136408e-01 -1.00132120e+00 1.80600598e-01 -1.29308105e-01
3.63941759e-01 3.37813765e-01 8.26478004e-01 -1.45145917e+00
-1.57811254e-01 -8.75691056e-01 -1.11410253e-01 8.59180629e-01
4.05148029e-01 -1.50908917e-01 -7.70436764e-01 -4.70555395e-01
4.46569532e-01 -1.95831716e-01 9.16764140e-01 -4.15642746e-02
1.57855117e+00 -4.20162588e-01 -2.47518733e-01 9.08380151e-01
1.18939722e+00 2.51259714e-01 4.23867047e-01 4.16146994e-01
9.91290092e-01 3.48861933e-01 4.93278593e-01 5.87098360e-01
9.69513133e-02 3.44773799e-01 5.50438464e-01 7.01261982e-02
4.42084581e-01 -3.85248840e-01 3.57627779e-01 1.15785217e+00
1.09561004e-01 4.52741757e-02 -8.79948199e-01 3.71264756e-01
-1.92131567e+00 -8.65445554e-01 2.98147708e-01 2.27315140e+00
3.20831716e-01 1.59292087e-01 -2.92991027e-02 3.90063435e-01
7.85041690e-01 1.38987303e-01 -9.57759798e-01 -2.63272226e-01
-2.47104410e-02 -1.59829035e-01 -1.46044008e-02 4.12036218e-02
-1.20113027e+00 5.72649777e-01 5.85207701e+00 5.29315114e-01
-1.39479971e+00 -2.34348610e-01 5.88465452e-01 1.26816675e-01
-2.44931266e-01 -1.98186427e-01 -5.25397003e-01 7.11071014e-01
7.79350638e-01 3.28696216e-03 2.46917725e-01 1.03449190e+00
-3.89985032e-02 4.22522724e-01 -1.15498424e+00 9.33353782e-01
3.30042958e-01 -7.13085055e-01 4.06580448e-01 6.03507385e-02
4.31055576e-01 -9.00775567e-02 1.03910603e-01 5.24255097e-01
-3.98859531e-02 -9.12998021e-01 2.42576689e-01 5.65854967e-01
3.57322693e-01 -7.86997437e-01 1.04823804e+00 4.94687796e-01
-8.74238014e-01 -2.83444554e-01 -6.15740299e-01 -1.25686824e-01
-2.18999967e-01 5.23826540e-01 -8.28395903e-01 5.26842892e-01
4.54299748e-01 1.05943370e+00 -6.24438405e-01 9.57616150e-01
-1.29883662e-01 8.29640031e-01 -3.08609158e-01 2.99792647e-01
2.90788829e-01 -5.67894042e-01 6.43718421e-01 8.39062274e-01
5.46513736e-01 2.79189590e-02 4.88190383e-01 8.06313038e-01
-3.38544138e-02 4.05922860e-01 -8.98856580e-01 -2.74828300e-02
3.65698725e-01 1.26381505e+00 -4.03884470e-01 -2.13565648e-01
-6.50872946e-01 1.07985163e+00 5.13251245e-01 4.28499460e-01
-5.23075998e-01 -4.72422868e-01 6.46164775e-01 -1.72412112e-01
1.78342164e-01 7.06141889e-02 -4.42979075e-02 -1.48107052e+00
3.03268462e-01 -7.56624699e-01 7.34825552e-01 -2.33203709e-01
-1.62227988e+00 7.16870666e-01 -3.49405736e-01 -1.65839040e+00
-5.05448520e-01 -6.45553172e-01 -9.65619385e-01 8.51962209e-01
-1.20022774e+00 -8.56802106e-01 -3.44127893e-01 7.83835113e-01
2.59917736e-01 -7.79130578e-01 9.10461426e-01 3.15106958e-01
-9.68429446e-01 8.55754077e-01 3.99123877e-01 5.81599891e-01
5.12977779e-01 -1.41049707e+00 1.71858802e-01 1.12878299e+00
3.03585500e-01 5.85499227e-01 2.91025370e-01 -3.94467235e-01
-1.14608955e+00 -1.28996634e+00 2.39587784e-01 -1.84646457e-01
6.49304271e-01 -5.72013631e-02 -1.40338278e+00 9.62059140e-01
-3.23025227e-01 7.28520751e-01 9.52492297e-01 -1.27690509e-02
-4.02777195e-01 -2.60932833e-01 -1.34715807e+00 4.98744041e-01
6.85307741e-01 -4.83828396e-01 -8.62228572e-01 -6.93558902e-02
4.08431739e-01 -4.25626814e-01 -6.16155982e-01 5.24829865e-01
2.58977145e-01 -8.51592839e-01 6.46669447e-01 -8.83467734e-01
2.59337217e-01 -4.94080633e-01 -2.28741854e-01 -1.49173212e+00
-3.05984348e-01 2.86911167e-02 -5.68385482e-01 1.17342567e+00
6.48984388e-02 -8.57201874e-01 1.08207011e+00 5.04831791e-01
-3.92408758e-01 -1.09620392e+00 -1.01035547e+00 -4.98025984e-01
-9.09921378e-02 -4.79530185e-01 6.89990044e-01 1.11711359e+00
-2.68800966e-02 1.45565659e-01 -4.90138143e-01 4.14435446e-01
4.99595582e-01 2.69739926e-02 5.59687376e-01 -1.31831133e+00
-1.20422624e-01 -2.12715149e-01 -1.00941193e+00 -1.00812221e+00
2.97673941e-01 -1.10884750e+00 1.61967482e-02 -1.07330287e+00
5.57159930e-02 -3.20023626e-01 -9.83033895e-01 5.40774822e-01
-3.11975837e-01 9.56659466e-02 -1.92668989e-01 4.06000853e-01
-5.68209171e-01 9.42740262e-01 9.63111699e-01 -1.12734810e-01
-2.03133553e-01 1.73449397e-01 -7.05061495e-01 1.07312834e+00
8.42564404e-01 -2.51010031e-01 -3.68061930e-01 -3.87999892e-01
-3.66924226e-01 -1.86927602e-01 2.88973927e-01 -1.11351800e+00
1.29237324e-01 -1.72170643e-02 8.01904321e-01 -5.89594364e-01
1.16462700e-01 -1.18452394e+00 -3.47987920e-01 3.38464081e-01
-3.46600890e-01 2.64396314e-02 -2.62143603e-03 9.52005029e-01
-5.80603182e-01 -4.17813748e-01 6.59761190e-01 1.04404338e-01
-7.92105496e-01 6.05890751e-01 -1.14052385e-01 1.14323944e-02
1.15502572e+00 -2.75295496e-01 6.20837659e-02 -2.33170420e-01
-9.01457071e-01 3.19687456e-01 4.25844848e-01 4.11258757e-01
1.04283369e+00 -1.78542042e+00 -5.34560382e-01 8.75258505e-01
3.38319987e-01 9.17875320e-02 1.83336183e-01 6.66096210e-01
-3.96770835e-01 1.48067668e-01 -4.38598841e-01 -6.50962114e-01
-7.36845791e-01 6.55177712e-01 3.26875299e-01 -7.93942511e-02
-9.11537468e-01 5.67847908e-01 8.82391483e-02 -6.47508919e-01
3.20922405e-01 -2.49669939e-01 -4.01937664e-01 -2.07105860e-01
6.80351555e-01 2.92809665e-01 1.80204753e-02 -7.07362115e-01
-3.65085632e-01 3.32689464e-01 -3.79362494e-01 2.95522183e-01
1.19690800e+00 -1.19728362e-02 -1.14507556e-01 5.40162861e-01
1.32406247e+00 3.42385359e-02 -1.17717314e+00 -5.94978333e-01
2.14347839e-01 -6.63447976e-01 -1.84667762e-02 -4.05049890e-01
-1.40747225e+00 8.52408469e-01 7.48883188e-01 1.35631740e-01
1.21859765e+00 -1.60952762e-01 8.77143383e-01 6.08665824e-01
-7.57894441e-02 -1.00458944e+00 2.47701749e-01 3.94816458e-01
6.43496692e-01 -1.51250732e+00 -2.97596872e-01 -9.89150442e-03
-5.71386635e-01 1.20729554e+00 9.90926564e-01 -3.57208341e-01
6.64599538e-01 -1.00531422e-01 2.29728252e-01 -1.49116829e-01
-3.92177284e-01 2.65779383e-02 5.00981688e-01 5.29758453e-01
2.02928960e-01 -1.41231477e-01 -1.46860555e-01 8.34934294e-01
1.83757260e-01 -3.44334304e-01 3.46380442e-01 8.13817322e-01
-5.37329018e-01 -8.93215060e-01 -3.14297825e-01 9.09638464e-01
-4.99085873e-01 2.75442511e-01 -1.83391228e-01 4.25773203e-01
3.63048688e-02 5.12187004e-01 3.14778268e-01 -4.77302909e-01
4.55061466e-01 2.86039144e-01 -1.81737348e-01 -7.53302693e-01
1.17353790e-01 -1.66247800e-01 -5.84770203e-01 -5.62038183e-01
6.95951656e-03 -6.19055867e-01 -1.41530406e+00 1.41051576e-01
-1.87521845e-01 3.24881822e-01 2.43689314e-01 1.08640730e+00
2.74005234e-01 3.86140227e-01 1.02424932e+00 -5.88404357e-01
-8.90992999e-01 -1.01295984e+00 -8.46672773e-01 7.69858956e-01
5.19676149e-01 -6.62407398e-01 -6.80606961e-01 -2.37425908e-01] | [7.627014636993408, 2.345529556274414] |
c74e0788-d3ec-4491-bc55-3ffe1298cea5 | exploring-structural-encoding-for-data-to | null | null | https://aclanthology.org/2021.inlg-1.44 | https://aclanthology.org/2021.inlg-1.44.pdf | Exploring Structural Encoding for Data-to-Text Generation | Due to efficient end-to-end training and fluency in generated texts, several encoder-decoder framework-based models are recently proposed for data-to-text generations. Appropriate encoding of input data is a crucial part of such encoder-decoder models. However, only a few research works have concentrated on proper encoding methods. This paper presents a novel encoder-decoder based data-to-text generation model where the proposed encoder carefully encodes input data according to underlying structure of the data. The effectiveness of the proposed encoder is evaluated both extrinsically and intrinsically by shuffling input data without changing meaning of that data. For selecting appropriate content information in encoded data from encoder, the proposed model incorporates attention gates in the decoder. With extensive experiments on WikiBio and E2E dataset, we show that our model outperforms the state-of-the models and several standard baseline systems. Analysis of the model through component ablation tests and human evaluation endorse the proposed model as a well-grounded system. | ['Utpal Garain', 'Joy Mahapatra'] | null | null | null | null | inlg-acl-2021-8 | ['data-to-text-generation'] | ['natural-language-processing'] | [ 5.93582213e-01 6.43176258e-01 3.17695737e-02 -5.08433521e-01
-9.11110818e-01 -3.14002752e-01 8.86497796e-01 -2.55056143e-01
-4.01622206e-01 1.08329129e+00 7.22266555e-01 -8.03412646e-02
2.68625528e-01 -7.46526003e-01 -7.86771774e-01 -2.51068920e-01
5.53353250e-01 7.95992374e-01 -2.03769058e-01 -4.46382135e-01
3.23443472e-01 -2.77660728e-01 -1.47999334e+00 6.31038666e-01
1.26963973e+00 3.44911814e-01 5.48444211e-01 8.49768341e-01
-3.41853142e-01 5.67523837e-01 -8.93826067e-01 -7.34752893e-01
-7.90677741e-02 -1.04772425e+00 -8.74312639e-01 -5.65413311e-02
1.16042055e-01 -4.81451213e-01 -2.30779007e-01 9.02176976e-01
1.04261088e+00 -3.57842594e-01 8.52017522e-01 -9.82854545e-01
-1.35736632e+00 1.31990671e+00 -6.93888515e-02 -7.56605491e-02
3.47895950e-01 1.54336557e-01 8.30320895e-01 -1.02588272e+00
7.86259770e-01 9.70259786e-01 1.57476813e-01 1.07003105e+00
-9.29379582e-01 -4.14320379e-01 -1.36940494e-01 -1.34645909e-01
-1.16092563e+00 -8.06733608e-01 6.32919371e-01 -2.65092045e-01
1.14730847e+00 1.70493737e-01 5.27944446e-01 1.45198393e+00
1.07796177e-01 1.06680977e+00 7.41096139e-01 -8.33873212e-01
-4.06933343e-03 3.93259943e-01 -1.61107123e-01 5.16969681e-01
3.78808856e-01 1.69304475e-01 -8.01565826e-01 3.70172232e-01
5.13692021e-01 -7.86968768e-01 -3.29950154e-01 4.36478518e-02
-1.17707276e+00 6.95292413e-01 1.60287367e-04 1.49607629e-01
-3.52085024e-01 3.71302247e-01 4.95614201e-01 2.08950177e-01
4.73212481e-01 5.45169473e-01 -2.81370699e-01 -6.02041423e-01
-9.84062135e-01 3.40041757e-01 5.77992380e-01 1.59018266e+00
2.28545025e-01 3.82457554e-01 -8.58896971e-01 7.81239092e-01
3.38768244e-01 4.82030600e-01 9.90211964e-01 -2.42599711e-01
1.05958521e+00 5.47311604e-01 6.13193028e-02 -3.70602936e-01
3.27947997e-02 -3.93414944e-01 -9.02297080e-01 -2.12248877e-01
-1.25013307e-01 -5.69792986e-01 -9.84969616e-01 1.82294679e+00
1.13588631e-01 -1.54182315e-01 5.20955980e-01 5.87593555e-01
1.08156610e+00 7.50518620e-01 4.65527587e-02 5.58955893e-02
1.08238828e+00 -1.05342817e+00 -1.10011625e+00 -3.05296868e-01
8.58914614e-01 -7.61614025e-01 1.25761807e+00 1.52596891e-01
-1.34030199e+00 -7.87964940e-01 -1.02319896e+00 -2.41488218e-01
-2.08454043e-01 7.67140210e-01 2.66701669e-01 8.20583045e-01
-1.09361482e+00 3.63494366e-01 -4.45752203e-01 -4.22606915e-01
3.05838227e-01 3.00984532e-01 -2.14755297e-01 1.95474327e-01
-1.60444045e+00 7.92983532e-01 8.26070726e-01 6.08643256e-02
-1.20251024e+00 -4.64804262e-01 -7.48041511e-01 1.71429738e-02
-2.26324305e-01 -1.13564146e+00 1.46573567e+00 -8.81863832e-01
-1.83074307e+00 6.80286527e-01 -4.25319560e-02 -5.79559386e-01
4.18977827e-01 -4.70229805e-01 -5.87500274e-01 -9.97141823e-02
2.15221066e-02 1.03961122e+00 8.27834785e-01 -1.07380676e+00
-5.79866469e-01 1.86086610e-01 -2.09573716e-01 5.55560946e-01
-5.08205414e-01 -2.61355378e-02 -4.15425748e-01 -8.33086967e-01
-5.08603632e-01 -6.72228813e-01 5.61986566e-02 -6.27157211e-01
-7.11146057e-01 -8.66594538e-02 4.72463608e-01 -7.50691473e-01
1.57036138e+00 -1.84232962e+00 4.38175760e-02 -4.06472713e-01
-2.80140936e-01 3.95388424e-01 -2.38649815e-01 8.31817627e-01
-3.19428965e-02 3.75353396e-01 -2.78010488e-01 -4.46985841e-01
2.15415537e-01 -4.68260795e-02 -5.10177970e-01 -7.18113407e-02
5.17591178e-01 1.16482663e+00 -8.95404577e-01 -4.73568261e-01
1.17808685e-01 6.50625348e-01 -7.65640855e-01 6.16673172e-01
-4.12631661e-01 3.83949935e-01 -5.12472510e-01 1.65530145e-01
3.97159994e-01 2.08019651e-02 -1.89009011e-01 -1.19578086e-01
1.39863610e-01 7.18942285e-01 -8.14120054e-01 2.13507462e+00
-5.26703596e-01 6.35767579e-01 -5.32881796e-01 -5.72273433e-01
1.06651664e+00 7.96802163e-01 -2.39908919e-01 -6.66750014e-01
2.22056791e-01 2.44349420e-01 1.59523219e-01 -6.87515497e-01
1.12736952e+00 4.19061445e-02 -2.54622519e-01 4.97156203e-01
3.70715469e-01 -1.87284946e-01 4.30119276e-01 4.20566708e-01
9.32571948e-01 7.29232430e-01 1.37038067e-01 -1.00266665e-01
4.87806857e-01 -3.44518535e-02 9.33040828e-02 6.90788209e-01
2.16984153e-01 9.79230285e-01 2.24731490e-01 5.07153235e-02
-1.34285009e+00 -6.44352496e-01 -6.63931221e-02 8.56740236e-01
-5.00126928e-02 -6.27081037e-01 -1.32978523e+00 -7.31919050e-01
-4.75337356e-01 1.40431666e+00 -7.51811862e-01 -4.73459244e-01
-4.23578322e-01 -7.31594324e-01 1.02591515e+00 4.67713624e-01
5.95471621e-01 -1.35821116e+00 -4.90155250e-01 4.71576780e-01
-4.14899886e-01 -8.94978821e-01 -5.57749212e-01 -4.84354459e-02
-8.32149088e-01 -3.78521740e-01 -6.96393490e-01 -8.15639079e-01
9.06739414e-01 -2.29278252e-01 1.27162886e+00 -1.79639712e-01
-1.82473622e-02 -8.09336156e-02 -8.11117053e-01 -3.34441602e-01
-1.00542855e+00 5.52143633e-01 -3.98492932e-01 -4.85280380e-02
2.30757877e-01 -2.60052472e-01 -5.11760056e-01 -5.20686433e-02
-9.92712677e-01 8.09239209e-01 9.87482786e-01 8.35497618e-01
4.24833506e-01 -2.68363416e-01 9.71424639e-01 -1.34273493e+00
1.24130273e+00 -5.24124920e-01 -2.42441654e-01 4.86375213e-01
-7.33305216e-01 6.34280145e-01 8.00213099e-01 -8.40019360e-02
-1.47978401e+00 -7.45569989e-02 -3.37141424e-01 1.38715446e-01
-1.44116655e-01 5.44408739e-01 -5.56986511e-01 7.26393640e-01
7.29450941e-01 5.29724360e-01 -3.88692200e-01 -4.49523926e-01
7.16075480e-01 1.24944878e+00 5.58326423e-01 -6.97004318e-01
4.65286106e-01 -2.12363586e-01 -5.99220872e-01 -2.79987544e-01
-5.61462343e-01 2.05119073e-01 -5.47662020e-01 -5.29837571e-02
8.78423512e-01 -1.03710151e+00 1.41550124e-01 3.82885218e-01
-1.51545417e+00 -2.50134915e-01 -5.16961873e-01 2.53521472e-01
-6.41329229e-01 -9.95893627e-02 -4.31771487e-01 -7.20540822e-01
-7.98884273e-01 -1.31446540e+00 1.31551194e+00 8.19401667e-02
-4.16489482e-01 -9.04118657e-01 3.34679186e-01 3.45196515e-01
4.94269520e-01 -2.89076827e-02 7.64517248e-01 -7.43225574e-01
-2.54216999e-01 -3.48301798e-01 5.04099391e-02 1.58356190e-01
2.71315187e-01 7.39228502e-02 -1.01740420e+00 -5.79991452e-02
-2.96927154e-01 -5.51534235e-01 7.52383113e-01 4.88993488e-02
1.04721546e+00 -4.34878320e-01 -2.00700715e-01 6.31692111e-01
1.28462029e+00 2.12015182e-01 1.03454876e+00 7.15517625e-02
6.51830673e-01 4.83201444e-01 4.61979419e-01 5.38917363e-01
4.09659177e-01 4.80736971e-01 3.22778106e-01 -6.74295947e-02
-6.06389523e-01 -8.30802739e-01 7.46199846e-01 1.20170045e+00
1.87943891e-01 -1.04052722e+00 -4.90671575e-01 7.76666403e-01
-1.61756873e+00 -9.49200332e-01 -2.16182888e-01 2.04386711e+00
1.32004523e+00 1.55462697e-01 -2.26089060e-01 -7.32272044e-02
9.63209391e-01 -1.14674285e-01 -3.67595464e-01 -6.84751034e-01
-2.58597970e-01 3.56087446e-01 2.30820730e-01 4.76148278e-01
-7.85457492e-01 1.17160451e+00 6.56903744e+00 9.74072397e-01
-7.83340096e-01 4.83993888e-02 6.00801528e-01 -1.35223866e-01
-6.93953753e-01 9.72880051e-02 -1.23939812e+00 6.73904777e-01
1.25735295e+00 -4.13367510e-01 3.00814092e-01 7.74846196e-01
1.70227394e-01 2.49237329e-01 -1.27125609e+00 6.83527291e-01
3.06725532e-01 -1.29619813e+00 6.20355010e-01 -8.77737477e-02
8.02379370e-01 -3.25338662e-01 9.89013538e-03 3.79247248e-01
5.66139519e-01 -1.13329756e+00 1.06024456e+00 4.87961352e-01
1.21382153e+00 -6.93861067e-01 7.17589736e-01 2.17379302e-01
-7.23119617e-01 4.03714448e-01 -3.79572988e-01 9.23522562e-02
4.31679189e-01 5.46457469e-01 -1.20120537e+00 5.91359913e-01
-3.26071423e-03 6.74103856e-01 -7.43402362e-01 5.91815948e-01
-7.40628481e-01 8.20376754e-01 1.94847718e-01 -2.26928860e-01
5.75116985e-02 5.46154454e-02 3.40641886e-01 1.51794493e+00
7.12835133e-01 -2.78075606e-01 -5.54414511e-01 1.12171149e+00
-4.81530547e-01 2.54324824e-01 -7.53606260e-01 -4.73605365e-01
5.07317543e-01 1.17193758e+00 -1.91316128e-01 -5.19113600e-01
-1.78152993e-01 1.36307395e+00 4.63131636e-01 2.18392029e-01
-8.31188858e-01 -7.84829080e-01 1.46497890e-01 6.36992380e-02
3.51813555e-01 7.72929341e-02 -3.47119272e-01 -1.03844929e+00
-6.08003279e-03 -1.04731202e+00 3.02891228e-02 -9.93549645e-01
-9.15043533e-01 1.10887766e+00 -4.48128060e-02 -1.34734857e+00
-7.19371080e-01 -1.27888858e-01 -5.37788332e-01 1.09903908e+00
-1.31176853e+00 -1.15799975e+00 -1.19387701e-01 2.90923685e-01
8.10723841e-01 -6.28169537e-01 9.38822985e-01 2.00937986e-01
-6.54706776e-01 9.20176268e-01 1.96318701e-01 1.87305853e-01
7.66279519e-01 -1.23124337e+00 1.06083477e+00 1.22136688e+00
1.71239734e-01 7.31949687e-01 6.40828907e-01 -9.50278282e-01
-1.21946645e+00 -1.31962895e+00 1.25736272e+00 -3.24432224e-01
1.48553133e-01 -6.21874392e-01 -5.89741707e-01 5.94944775e-01
1.17002022e+00 -4.24547195e-01 7.49881685e-01 -3.65689367e-01
7.26291910e-02 1.05153300e-01 -8.98991585e-01 7.40197718e-01
1.07200897e+00 -2.57971376e-01 -7.01516211e-01 2.19502017e-01
1.00164282e+00 -5.61287880e-01 -5.50769687e-01 2.06888407e-01
3.02707046e-01 -7.62950361e-01 4.23623174e-01 -6.42781317e-01
1.06730354e+00 -2.57770091e-01 8.54528397e-02 -1.76405954e+00
-2.63427824e-01 -1.01611745e+00 -2.11443320e-01 1.73418581e+00
8.74666929e-01 -1.07136838e-01 5.80823183e-01 5.44028342e-01
-6.70941055e-01 -4.60214227e-01 -6.44287407e-01 -4.50887948e-01
6.87940046e-02 -3.85753930e-01 1.01707017e+00 6.20918393e-01
1.20600192e-02 7.75411189e-01 -6.90626740e-01 -2.37727299e-01
2.38667548e-01 -1.74531162e-01 8.41461003e-01 -5.81704855e-01
-4.30290788e-01 -2.74437517e-01 8.47930014e-02 -1.24366426e+00
-7.91511089e-02 -1.24172807e+00 3.00609142e-01 -1.92640197e+00
1.87122092e-01 -3.97958040e-01 1.00816920e-01 3.45153660e-01
-4.40213382e-01 -1.09612681e-01 -1.06609929e-02 -8.26707762e-03
-3.30315053e-01 1.02711129e+00 1.39535844e+00 -7.65748229e-03
-7.17604160e-02 -3.64300191e-01 -1.02542710e+00 6.73356354e-02
1.01316249e+00 -6.52877629e-01 -8.79045725e-01 -1.14574265e+00
4.68647867e-01 -3.64888310e-02 -1.19664885e-01 -1.00120211e+00
1.36272386e-01 -1.35686742e-02 2.64224231e-01 -5.26821911e-01
5.51494658e-02 -3.25958639e-01 1.99142009e-01 2.37554342e-01
-1.02298784e+00 2.07919747e-01 1.74116567e-01 1.87404424e-01
-2.59243518e-01 -7.14279413e-01 4.30092216e-01 -4.67375740e-02
-4.77511227e-01 7.98420608e-02 -2.92705923e-01 5.36564410e-01
7.01439202e-01 -1.73192397e-01 -5.06751657e-01 -5.42322576e-01
-2.20828608e-01 -9.66100618e-02 2.44787514e-01 5.35762846e-01
7.20010102e-01 -1.53134787e+00 -1.31418264e+00 4.17317599e-01
2.90811658e-01 -5.74880242e-02 2.14507058e-02 2.33967170e-01
-5.56432128e-01 6.52095497e-01 -2.21948639e-01 -1.02122657e-01
-8.67741168e-01 1.82098851e-01 1.98948652e-01 -3.46421897e-01
-3.16765398e-01 1.00115108e+00 1.11487634e-01 -3.89293849e-01
5.95891215e-02 -3.05120170e-01 -1.79240778e-01 -2.14196816e-01
5.84049106e-01 5.23690172e-02 1.71667397e-01 -4.90040690e-01
5.69507666e-02 3.76782715e-02 -3.48170072e-01 -5.82804263e-01
1.07200217e+00 -9.63975787e-02 1.62517563e-01 1.35761335e-01
9.27680075e-01 -1.17747955e-01 -1.10749984e+00 1.29343957e-01
-2.85154521e-01 -2.26420432e-01 1.75916180e-02 -1.12710571e+00
-9.14150238e-01 1.16893625e+00 3.60258311e-01 -1.63525105e-01
1.10885048e+00 -4.05847371e-01 9.08807874e-01 1.43387765e-01
7.50657544e-02 -1.30255401e+00 4.88065742e-02 5.95444679e-01
1.11577463e+00 -1.02698755e+00 -2.85562456e-01 -1.45755678e-01
-1.04430878e+00 9.39700723e-01 8.77013862e-01 1.02782033e-01
1.23855680e-01 3.13629389e-01 -1.47462979e-01 1.51534870e-01
-1.07994151e+00 -1.74575493e-01 2.99871922e-01 8.19286644e-01
1.07040787e+00 -2.48213768e-01 -6.42657876e-01 9.04247165e-01
-6.54033720e-01 3.55095476e-01 6.36347115e-01 8.19444060e-01
-2.45591059e-01 -1.58403563e+00 5.01154028e-02 4.17012841e-01
-3.01588804e-01 -7.34655499e-01 -5.49952269e-01 6.00265801e-01
6.12417050e-02 8.88677835e-01 -3.92442793e-02 -4.96099800e-01
2.00878948e-01 2.32005715e-01 4.32048738e-01 -8.20645750e-01
-9.73489523e-01 -9.35176983e-02 4.66816187e-01 8.22832659e-02
-1.90718144e-01 -4.78855550e-01 -1.26620066e+00 1.18677281e-01
-4.70701814e-01 3.60238910e-01 7.37593591e-01 7.22762883e-01
8.62534881e-01 9.94600058e-01 4.88173276e-01 -3.62828165e-01
-5.15112579e-01 -1.40148008e+00 -2.32486859e-01 5.68998694e-01
-1.32925451e-01 -1.37125567e-01 -1.88000090e-02 5.51338494e-01] | [11.84941291809082, 9.094732284545898] |
9e98cd66-619b-43a7-9c04-3bbe223469fe | recist-weakly-supervised-lesion-segmentation | 2303.00205 | null | https://arxiv.org/abs/2303.00205v1 | https://arxiv.org/pdf/2303.00205v1.pdf | RECIST Weakly Supervised Lesion Segmentation via Label-Space Co-Training | As an essential indicator for cancer progression and treatment response, tumor size is often measured following the response evaluation criteria in solid tumors (RECIST) guideline in CT slices. By marking each lesion with its longest axis and the longest perpendicular one, laborious pixel-wise manual annotation can be avoided. However, such a coarse substitute cannot provide a rich and accurate base to allow versatile quantitative analysis of lesions. To this end, we propose a novel weakly supervised framework to exploit the existing rich RECIST annotations for pixel-wise lesion segmentation. Specifically, a pair of under- and over-segmenting masks are constructed for each lesion based on its RECIST annotation and served as the label for co-training a pair of subnets, respectively, along with the proposed label-space perturbation induced consistency loss to bridge the gap between the two subnets and enable effective co-training. Extensive experiments are conducted on a public dataset to demonstrate the superiority of the proposed framework regarding the RECIST-based weakly supervised segmentation task and its universal applicability to various backbone networks. | ['Yefeng Zheng', 'Liansheng Wang', 'Wei Xue', 'Donghuan Lu', 'Dong Wei', 'Lianyu Zhou'] | 2023-03-01 | null | null | null | null | ['weakly-supervised-segmentation', 'lesion-segmentation'] | ['computer-vision', 'medical'] | [ 5.56987822e-01 1.05467595e-01 -8.47545326e-01 -4.19994235e-01
-1.05354416e+00 -4.24630642e-01 3.57525766e-01 3.25360030e-01
-5.42907715e-01 9.19837534e-01 2.74366885e-02 -4.30199534e-01
-5.82371801e-02 -6.67439997e-01 -3.63410980e-01 -1.21654236e+00
3.02295417e-01 3.38428289e-01 4.07157034e-01 3.67478132e-02
-1.57802939e-01 4.82844651e-01 -6.91926479e-01 9.59861204e-02
1.17011011e+00 1.18468881e+00 4.80486661e-01 7.03161210e-02
-1.44060820e-01 7.31812239e-01 -8.41253921e-02 -1.29546374e-01
1.14469379e-01 -3.25426400e-01 -9.15389895e-01 2.63393342e-01
1.91040814e-01 -1.45687953e-01 -2.57441998e-01 1.25268209e+00
4.07532543e-01 -2.41066381e-01 5.52542448e-01 -7.94377327e-01
-1.27756685e-01 7.67373681e-01 -9.45583105e-01 2.72587389e-01
-3.16820711e-01 4.49474275e-01 1.07124281e+00 -5.47293425e-01
5.94025671e-01 5.27302027e-01 6.47185206e-01 4.13270742e-01
-1.08471346e+00 -4.62155074e-01 1.84957176e-01 -1.37251560e-02
-1.39199841e+00 -1.67312860e-01 8.50024998e-01 -1.92638054e-01
2.09688425e-01 3.44408095e-01 5.39077222e-01 7.84837842e-01
1.12737834e-01 8.23050082e-01 1.05723858e+00 -1.75426304e-01
-2.35881470e-02 2.69292712e-01 1.48804531e-01 9.02297258e-01
3.35742123e-02 -2.93117851e-01 1.41625404e-02 2.10133977e-02
8.18366706e-01 1.15126535e-01 -4.62622434e-01 -5.23186445e-01
-1.34109020e+00 6.32016063e-01 1.08392990e+00 4.89775449e-01
-3.22478026e-01 -2.78338939e-02 5.95786810e-01 -2.75049686e-01
5.00069857e-01 -2.16495129e-03 -1.67237744e-02 3.26122671e-01
-1.09494174e+00 -2.93255150e-01 2.05426425e-01 5.34606338e-01
4.27166313e-01 -2.21186504e-01 -5.80554426e-01 7.83768713e-01
1.93519592e-01 5.44080287e-02 7.36689091e-01 -4.76611614e-01
2.56558597e-01 8.58934760e-01 -6.15122840e-02 -8.08857918e-01
-6.78832948e-01 -1.12373662e+00 -1.24568415e+00 -1.13114387e-01
7.75590777e-01 1.74438972e-02 -1.02431738e+00 1.78451395e+00
4.70851064e-01 4.30253565e-01 -1.89316705e-01 1.03849602e+00
7.00634062e-01 8.56544599e-02 3.18927795e-01 -4.52716768e-01
1.43093228e+00 -1.19421709e+00 -4.41954046e-01 -1.40788257e-01
1.13166797e+00 -5.67294180e-01 1.11726987e+00 -7.30126500e-02
-9.95626092e-01 -2.07460895e-01 -9.42851424e-01 2.32942402e-01
1.36604369e-01 5.01947224e-01 6.85010254e-01 4.43568498e-01
-7.48134673e-01 3.66380841e-01 -1.01248598e+00 -2.52356946e-01
8.79468501e-01 2.72625208e-01 -2.05425188e-01 -1.41420439e-01
-8.92471731e-01 5.92827618e-01 4.81469572e-01 3.15764934e-01
-9.42372501e-01 -9.12230551e-01 -4.91371214e-01 -2.34184787e-01
4.39846337e-01 -5.58508098e-01 1.09717989e+00 -9.82575536e-01
-1.24136734e+00 1.07940876e+00 -1.39886409e-01 -6.04701877e-01
8.93780649e-01 5.22051275e-01 -1.49100199e-01 3.87044549e-01
2.06601903e-01 6.05872512e-01 5.47250211e-01 -1.12103438e+00
-7.06684887e-01 -6.71260655e-01 -1.10502407e-01 4.61892545e-01
-4.92475778e-01 -3.57855976e-01 -6.39654696e-01 -7.30850339e-01
5.18095851e-01 -8.94990742e-01 -4.50731635e-01 3.72689515e-01
-8.15112472e-01 -7.54212076e-03 9.35354173e-01 -5.13459146e-01
1.16946518e+00 -2.08862901e+00 -1.71592981e-01 4.65787530e-01
4.57342088e-01 1.43403038e-01 1.03645800e-02 -4.57501888e-01
-2.12521274e-02 1.33743733e-01 -5.49666643e-01 -1.59225494e-01
-3.92967671e-01 1.77296996e-01 1.56258062e-01 7.90201962e-01
3.73416133e-02 1.08445764e+00 -9.73797739e-01 -9.65712607e-01
2.27910504e-01 2.88238734e-01 -1.80520922e-01 7.83774853e-02
-1.57255262e-01 9.55637634e-01 -7.05655336e-01 6.43956184e-01
5.55831969e-01 -4.56787109e-01 1.94512263e-01 -4.71990228e-01
1.42419979e-01 -7.95233250e-02 -6.50147915e-01 1.69087291e+00
-4.56831217e-01 2.28192762e-01 2.84192652e-01 -1.25763512e+00
8.59905124e-01 2.97976404e-01 9.63731170e-01 -6.25257015e-01
2.73949146e-01 4.26427037e-01 1.75537094e-01 -4.82295454e-01
-7.53803179e-02 -2.69297689e-01 1.79153889e-01 3.95455420e-01
-3.79835546e-01 -1.33380592e-01 1.63971111e-01 -9.38965194e-03
9.41217721e-01 -3.13262224e-01 1.92167506e-01 -3.72691453e-01
9.25397336e-01 2.68498566e-02 7.32423306e-01 5.85039556e-01
-5.18603861e-01 6.29039824e-01 4.86965507e-01 -2.95418561e-01
-7.25460529e-01 -8.47410381e-01 -5.53591609e-01 8.16386282e-01
3.44297379e-01 1.28555462e-01 -7.80995548e-01 -1.10391533e+00
-2.61026263e-01 3.47815514e-01 -7.87741840e-01 -1.44671336e-01
-4.99885291e-01 -1.12007153e+00 6.57686710e-01 6.26770735e-01
7.45045543e-01 -6.88469172e-01 -4.53823417e-01 1.73110455e-01
-3.89656842e-01 -1.13664556e+00 -4.51639593e-01 3.94146800e-01
-9.71426308e-01 -1.19444656e+00 -1.13251019e+00 -9.42753017e-01
1.06800997e+00 3.51319999e-01 8.61761630e-01 3.86017561e-01
-1.80029437e-01 -3.39904100e-01 -1.92443624e-01 2.21283827e-02
-4.44368839e-01 2.50657588e-01 -3.94412726e-01 1.99595824e-01
1.35660201e-01 -5.32696187e-01 -8.63779843e-01 5.70987284e-01
-8.63180995e-01 3.26351941e-01 7.06537545e-01 1.08358121e+00
1.03549337e+00 1.07686453e-01 5.29616416e-01 -1.19816339e+00
2.88654983e-01 -3.80251557e-01 -3.56947839e-01 3.35700661e-01
-3.32341194e-01 -2.51087457e-01 5.91938138e-01 -1.69673190e-01
-1.02034271e+00 3.19026291e-01 -2.20363170e-01 3.78301628e-02
-8.45637918e-02 6.84025407e-01 -1.02770217e-01 -4.45389122e-01
5.32155514e-01 2.51189202e-01 -3.21103223e-02 -4.44591604e-03
1.78845063e-01 5.57300150e-01 7.36903965e-01 -5.07771254e-01
6.36021733e-01 8.41963947e-01 1.54955834e-01 -5.04822254e-01
-1.07360470e+00 -5.34004390e-01 -6.98558390e-01 -1.28886774e-01
8.74913871e-01 -5.81772923e-01 -3.90766114e-01 5.24136364e-01
-7.12819695e-01 -3.29643846e-01 -4.77530062e-01 4.85754758e-01
-3.97085637e-01 4.24791366e-01 -6.62106752e-01 -2.15050399e-01
-4.17108506e-01 -1.55857575e+00 9.52766657e-01 3.53858680e-01
1.69675484e-01 -1.04488444e+00 -9.99982357e-02 4.57659185e-01
3.95657480e-01 4.41855371e-01 1.09045851e+00 -7.37643659e-01
-2.70849526e-01 -4.43218857e-01 -7.27152228e-01 2.69959360e-01
2.95003772e-01 -4.61047050e-03 -8.37439299e-01 -2.96870053e-01
-1.91281557e-01 -4.77112472e-01 9.52180922e-01 6.29156470e-01
1.58122861e+00 1.28355682e-01 -6.17809355e-01 8.50642562e-01
1.38618028e+00 -1.07617602e-01 2.74672717e-01 3.35676432e-01
9.24853325e-01 3.85944754e-01 4.35869604e-01 3.25694025e-01
2.19557956e-01 6.46616936e-01 5.80234826e-01 -5.40405154e-01
-2.60491371e-01 -9.59294438e-02 -3.27895939e-01 4.43846226e-01
-1.21380594e-02 -2.29281694e-01 -9.51688588e-01 4.46265906e-01
-1.65794039e+00 -4.58766550e-01 -1.81650296e-01 2.14806032e+00
9.78663921e-01 3.83123338e-01 -5.53027466e-02 8.07426646e-02
8.22692573e-01 2.57097721e-01 -7.59668112e-01 2.31103152e-01
-9.38845500e-02 -1.37742415e-01 5.97304761e-01 1.86763451e-01
-1.16434789e+00 6.72528803e-01 5.38580418e+00 1.20568228e+00
-1.30479419e+00 3.54654223e-01 1.16844594e+00 1.60984561e-01
-2.72523195e-01 -1.74830407e-01 -5.31426549e-01 3.85821581e-01
4.47457910e-01 5.72714061e-02 -1.83514565e-01 5.73267400e-01
5.46282172e-01 -1.86388597e-01 -9.34861243e-01 7.86078691e-01
-3.55744243e-01 -1.33530569e+00 1.43518075e-02 4.90577295e-02
9.36754525e-01 -6.91437870e-02 1.98507115e-01 -2.87100505e-02
9.36874226e-02 -9.97474551e-01 3.66241157e-01 2.11026430e-01
7.73192585e-01 -6.77180171e-01 9.98363256e-01 3.82294804e-01
-1.22338080e+00 6.55320063e-02 -3.14361930e-01 5.93879759e-01
1.47477537e-01 6.64086699e-01 -8.20985794e-01 6.54469490e-01
2.58601785e-01 7.65046775e-01 -6.30907357e-01 1.08929932e+00
-6.76801503e-02 7.41909206e-01 -3.94899756e-01 3.77368331e-01
6.71092570e-01 -7.32528940e-02 5.20959377e-01 1.18451750e+00
4.11273986e-02 -5.75954989e-02 4.51577783e-01 6.44003093e-01
-2.61729926e-01 2.76147455e-01 7.42091015e-02 2.51708150e-01
3.58391911e-01 1.53849602e+00 -1.15816128e+00 -2.11219445e-01
-4.20686841e-01 7.30654895e-01 2.79162556e-01 3.38826567e-01
-9.62106586e-01 7.00245649e-02 3.54215950e-02 2.50417739e-01
1.39799691e-03 2.76505500e-01 -6.44550443e-01 -9.76935089e-01
-4.38653231e-02 -6.54379845e-01 5.63946545e-01 -3.14142376e-01
-1.09110534e+00 7.24484384e-01 -2.75791556e-01 -1.41760230e+00
1.69812709e-01 -1.54513419e-01 -9.88235354e-01 7.76373267e-01
-1.80975759e+00 -1.35885906e+00 -8.40849936e-01 5.66436052e-01
4.46943432e-01 1.20195083e-01 4.85551715e-01 3.33406359e-01
-1.00239122e+00 1.00446308e+00 1.09588057e-02 1.27114847e-01
6.22053742e-01 -1.12349772e+00 -2.65638798e-01 7.57621527e-01
-2.94564515e-01 1.70231774e-01 4.29495305e-01 -4.98131752e-01
-7.91143298e-01 -1.37783575e+00 5.81292771e-02 4.46995087e-02
7.32261121e-01 1.84271187e-01 -1.08877289e+00 4.43972856e-01
-1.44484669e-01 6.38970196e-01 5.46863377e-01 -4.05051470e-01
-1.81417540e-02 -2.47703105e-01 -1.26852596e+00 4.76441592e-01
7.50165820e-01 -2.96551436e-01 8.50478187e-02 7.45198250e-01
5.21055520e-01 -5.67685127e-01 -9.61469591e-01 8.30753088e-01
3.17269176e-01 -7.40217149e-01 8.99541378e-01 -3.28250676e-01
3.62828165e-01 -3.12919557e-01 1.23572320e-01 -1.02021420e+00
-2.74783999e-01 -1.58421457e-01 3.49932104e-01 1.09574294e+00
5.74776709e-01 -3.11296731e-01 1.31751692e+00 4.37211454e-01
-3.09886634e-01 -1.24594867e+00 -8.69461775e-01 -5.06558597e-01
1.00960940e-01 -3.90420377e-01 3.39145452e-01 8.72231781e-01
-2.68690228e-01 -1.58439934e-01 -4.70070578e-02 1.25004113e-01
7.14122355e-01 -9.51699838e-02 4.25500095e-01 -1.02611661e+00
-1.66725844e-01 -7.99787402e-01 -4.27423179e-01 -1.02956688e+00
2.39081215e-02 -1.19320440e+00 1.43357828e-01 -1.33040011e+00
5.49331009e-01 -8.44089568e-01 -7.79364586e-01 4.10760015e-01
-3.31034929e-01 7.23210335e-01 -2.98969716e-01 4.59649086e-01
-5.14494658e-01 4.66298699e-01 1.65724957e+00 -3.74237716e-01
-8.49397555e-02 1.70542940e-01 -6.65275335e-01 1.03545189e+00
7.66702950e-01 -3.12592834e-01 -4.61411685e-01 -2.62164950e-01
-2.64947444e-01 1.58723995e-01 5.37436247e-01 -8.60281944e-01
1.86646968e-01 -2.82358021e-01 3.56033951e-01 -5.20442426e-01
-1.48386151e-01 -8.87290001e-01 -2.41479114e-01 5.31558335e-01
-5.82790136e-01 -4.93145615e-01 8.39850493e-03 6.80985928e-01
-3.32181543e-01 -2.99787790e-01 1.17369151e+00 -1.52147844e-01
-5.23664355e-01 6.24325097e-01 1.93876773e-01 -7.50273466e-02
1.44076753e+00 -3.38114142e-01 -2.53234088e-01 2.89185066e-02
-7.16129899e-01 4.20910418e-01 4.29629445e-01 -1.19457997e-01
4.22895223e-01 -1.15095019e+00 -8.88094902e-01 1.42880846e-02
3.20683360e-01 2.77605683e-01 5.30677855e-01 1.56333792e+00
-5.62301278e-01 3.82944673e-01 -1.16148837e-01 -8.81632626e-01
-1.07685661e+00 2.75230229e-01 7.00697839e-01 -8.25007856e-01
-7.51776338e-01 9.47063625e-01 5.94903946e-01 -2.85171062e-01
4.47766066e-01 -3.36587965e-01 -3.23278576e-01 -3.17604661e-01
4.82440054e-01 -8.32030550e-02 2.08159998e-01 -7.39653289e-01
-3.43911856e-01 4.48767394e-01 -4.22657281e-01 2.43306667e-01
1.07341361e+00 -3.43801945e-01 -1.04612589e-01 1.28011256e-01
1.12819505e+00 -2.79467314e-01 -1.42199445e+00 -6.58812821e-01
-1.68345645e-01 -2.23459333e-01 4.78050679e-01 -6.95764184e-01
-1.52047086e+00 8.21179271e-01 7.99823165e-01 -1.12536468e-01
1.21133661e+00 1.58225928e-04 1.04431939e+00 9.48919654e-02
3.14294919e-02 -7.62382746e-01 -7.62428269e-02 -1.09735109e-01
5.23741782e-01 -1.36309063e+00 -9.26197618e-02 -7.19596326e-01
-5.32590628e-01 1.01015222e+00 7.03926444e-01 3.66956107e-02
3.69157255e-01 9.27468687e-02 5.25553562e-02 -1.27398729e-01
-2.70822495e-01 -1.89262405e-01 2.98764139e-01 4.65924084e-01
5.23521066e-01 2.47313723e-01 -4.85812366e-01 5.17103314e-01
1.85096085e-01 -4.53835679e-03 2.35872537e-01 6.79262280e-01
-5.13229609e-01 -8.70719850e-01 -9.07367170e-02 6.74455166e-01
-5.22601008e-01 1.79537293e-02 -1.36686713e-01 7.37579763e-01
1.05898067e-01 5.23613214e-01 -1.68347985e-01 -1.63369521e-01
1.07053399e-01 -3.47948283e-01 2.68557876e-01 -5.10136187e-01
-4.46063846e-01 2.24612266e-01 -2.78267682e-01 -3.59722525e-01
-4.36930746e-01 -4.43583757e-01 -1.50543725e+00 -1.48231350e-02
-3.01497251e-01 6.14419468e-02 4.04034525e-01 1.09378374e+00
-3.02333564e-01 7.31714904e-01 8.56864631e-01 -4.79998529e-01
-6.75503373e-01 -7.80393779e-01 -5.44853270e-01 4.57936168e-01
1.80948168e-01 -7.17809141e-01 -1.33418620e-01 -8.62745121e-02] | [14.750784873962402, -2.4250223636627197] |
af995005-c7d0-4140-9fa8-49358f1f4836 | the-growing-liberality-observed-in-primary | 2301.02433 | null | https://arxiv.org/abs/2301.02433v1 | https://arxiv.org/pdf/2301.02433v1.pdf | The Growing Liberality Observed in Primary Animal and Plant Cultures is Common to the Social Amoeba | Tissue culture environment liberates cells from ordinary laws of multi-cellular organisms. This liberation enables cells several behaviors, such as proliferation, dedifferentiation, acquisition of pluripotency, immortalization, and reprogramming. Recently, the quantitative value of cellular dedifferentiation and differentiation was defined as liberality, which is measurable as Shannon entropy of numerical transcriptome data and Lempel-Zip complexity of nucleotide sequence transcriptome data. The increasing liberality induced by the culture environment had first been observed in animal cells and had reconfirmed in plant cells. The phenomena may be common across the kingdom, also in a social amoeba. We measured the liberality of the social amoeba which disaggregated from multicellular aggregates and transferred into a liquid medium. | ['Norichika Ogata'] | 2023-01-06 | null | null | null | null | ['culture'] | ['speech'] | [ 2.64011145e-01 5.77774690e-03 4.33885381e-02 1.85303494e-01
-9.81082916e-02 -1.04749620e+00 6.67233646e-01 2.66885728e-01
-5.04315376e-01 1.51741254e+00 1.83419362e-01 -9.49131548e-02
6.72818348e-02 -8.62289906e-01 -2.46911615e-01 -1.20370388e+00
1.66088827e-02 4.47125703e-01 -1.70029640e-01 -5.16064018e-02
8.98880512e-02 6.70912564e-01 -1.28029895e+00 -1.96934119e-01
8.74741077e-01 5.37102878e-01 2.10543931e-01 7.65933573e-01
-1.56146124e-01 6.18080311e-02 -2.48578683e-01 -1.94350600e-01
1.47003293e-01 -6.18636966e-01 -6.22787893e-01 -1.35374725e-01
-5.60348988e-01 1.27891123e-01 8.08858424e-02 8.52372527e-01
2.89693981e-01 -3.76633048e-01 9.22470689e-01 -1.15828633e+00
-7.87465394e-01 5.22070169e-01 -2.86737919e-01 1.37535110e-02
3.88387710e-01 4.15674865e-01 6.83487594e-01 -6.42068923e-01
1.12764859e+00 8.91995907e-01 4.86590624e-01 4.54138368e-01
-1.36007774e+00 -2.00569242e-01 -6.80105805e-01 -3.33722800e-01
-1.30816734e+00 -1.56768203e-01 -2.39709783e-02 -6.72681391e-01
3.46016735e-01 5.66757977e-01 1.29976892e+00 8.55503857e-01
9.88264501e-01 2.93185651e-01 1.62031603e+00 -1.40450358e-01
6.62012458e-01 -8.90996009e-02 -3.00686508e-01 4.47366446e-01
9.23584759e-01 -1.98013797e-01 -4.59349781e-01 -2.86716372e-01
7.45225906e-01 -1.79677933e-01 -4.45770949e-01 1.52068719e-01
-1.45536482e+00 1.51899248e-01 -4.86840934e-01 9.41358864e-01
-3.00213844e-01 1.26954362e-01 3.87184829e-01 2.15997308e-01
-7.43534938e-02 6.79451823e-01 -8.72197390e-01 -6.34687543e-01
-6.55934751e-01 -4.64979678e-01 1.02701402e+00 6.76616013e-01
4.01163250e-01 -2.08311573e-01 2.41150424e-01 3.87361020e-01
-2.56155021e-02 4.33545887e-01 8.76862347e-01 -9.23409879e-01
-5.73835850e-01 8.22617292e-01 -7.80009758e-03 -6.30732894e-01
-4.92190778e-01 -4.21913832e-01 -1.21480525e+00 -3.32780145e-02
7.88839102e-01 -2.15671495e-01 -3.16226214e-01 1.93958676e+00
3.23878914e-01 -1.16691992e-01 2.68244237e-01 5.85064948e-01
3.90832603e-01 6.67433619e-01 3.55772823e-02 -7.57422626e-01
1.46552813e+00 3.65340590e-01 -8.15300167e-01 5.90789616e-01
6.76295817e-01 -5.91940165e-01 9.89406824e-01 6.36761010e-01
-9.72744405e-01 -1.01324417e-01 -9.92763638e-01 2.43250057e-01
-5.78219235e-01 -3.96084577e-01 6.63029790e-01 8.57015789e-01
-9.90255237e-01 8.85043085e-01 -7.34128773e-01 -1.07167017e+00
3.66548628e-01 2.61240780e-01 -8.05017531e-01 4.30386007e-01
-7.72297978e-01 7.73442745e-01 5.29228866e-01 -1.18047543e-01
-6.93836391e-01 -3.95195246e-01 -1.73251301e-01 8.16653483e-03
-2.57966548e-01 -1.14189422e+00 4.56638902e-01 -4.66279268e-01
-1.92785406e+00 1.23167253e+00 1.77188173e-01 -2.42830869e-02
2.37058848e-01 2.63888329e-01 -3.12032253e-01 -1.24471873e-01
-2.88221240e-01 3.52302313e-01 -2.60484576e-01 -1.23145723e+00
-3.08321923e-01 -5.02810776e-01 -4.13090020e-01 -3.43928903e-01
-1.37201592e-01 -3.13584119e-01 3.62537503e-01 -3.89156312e-01
5.38177550e-01 -1.12829244e+00 -2.11720809e-01 -1.72031876e-02
-5.09176791e-01 3.00327867e-01 4.91806328e-01 -3.44077259e-01
6.83768570e-01 -2.10073996e+00 4.15286243e-01 -1.19147561e-02
8.11671615e-02 -1.55713528e-01 1.67336024e-03 8.82427931e-01
2.18887061e-01 9.00246799e-01 -2.49946624e-01 5.19965470e-01
-9.15060863e-02 2.00507492e-01 2.06209749e-01 7.93463826e-01
3.24713886e-02 8.51490796e-01 -1.03272820e+00 -5.75550735e-01
-1.97854042e-01 2.23084986e-01 -3.58021408e-01 -2.02227354e-01
1.79880112e-03 8.70366931e-01 -3.44429165e-01 9.21298683e-01
7.53692150e-01 -2.85894454e-01 4.75298822e-01 9.98734757e-02
-5.03089130e-01 -2.40250483e-01 -4.30112958e-01 1.39402246e+00
2.38452703e-02 5.19920170e-01 1.78785905e-01 -3.09549898e-01
8.38737965e-01 3.59319389e-01 6.58485293e-01 -1.78930551e-01
4.95734721e-01 4.80563998e-01 2.02323511e-01 -3.11566085e-01
5.37483573e-01 -6.35576308e-01 -3.04447263e-01 2.58156478e-01
2.60429233e-02 -7.60759413e-01 3.01559597e-01 1.49356708e-01
1.04003263e+00 3.98087800e-02 8.14652383e-01 -1.02893579e+00
4.63895500e-01 -3.42613578e-01 6.68320239e-01 1.26397997e-01
-3.89955997e-01 4.33919311e-01 8.91473532e-01 7.88456947e-02
-1.12753427e+00 -1.23585856e+00 -4.48458463e-01 4.64552343e-01
1.77259505e-01 -1.51911989e-01 -5.93769550e-01 1.29645333e-01
-9.36644524e-03 4.88185048e-01 -7.62874782e-01 -3.82376611e-01
2.75852662e-02 -1.15648043e+00 8.11475277e-01 -1.51223749e-01
6.45025134e-01 -4.29561853e-01 -4.23719168e-01 1.02045298e-01
-1.09770909e-01 -8.50062966e-01 -3.15520354e-03 4.02440250e-01
-9.90332723e-01 -8.33725750e-01 -5.43216109e-01 -2.45870158e-01
5.02432346e-01 -3.55599254e-01 6.24409795e-01 1.77737698e-01
-3.56081337e-01 -4.03588172e-03 -1.59266621e-01 -2.86029577e-01
-7.00211942e-01 2.96749882e-02 6.83260024e-01 -2.81704813e-01
-9.63485911e-02 -1.13055634e+00 -5.79791129e-01 1.24770820e-01
-1.08114493e+00 -2.90164631e-02 3.14442396e-01 7.75522947e-01
7.13387787e-01 -2.22101852e-01 8.52505744e-01 -6.36626303e-01
3.92874539e-01 -5.19829869e-01 -4.24897343e-01 1.08479224e-02
-4.27684873e-01 1.49910059e-02 9.01599646e-01 -2.36287042e-01
-7.70893335e-01 -1.15527831e-01 -1.93871930e-02 6.49938047e-01
-1.26643062e-01 4.42166507e-01 -4.45623070e-01 6.21126965e-02
3.19290698e-01 6.33892119e-01 -6.11725710e-02 5.91235831e-02
1.85489997e-01 8.21618676e-01 4.23177391e-01 -4.87875253e-01
6.55014575e-01 6.96667135e-01 6.79406643e-01 -1.06617999e+00
-1.59165144e-01 -3.49592492e-02 -8.63043606e-01 -2.67091900e-01
9.38422203e-01 -4.83136743e-01 -1.27027035e+00 5.80776274e-01
-8.05844247e-01 -2.72595048e-01 -4.69884723e-01 5.22489011e-01
-6.47358298e-01 4.70052630e-01 -8.33782732e-01 -7.62653232e-01
-2.53782216e-02 -3.61170381e-01 5.35765409e-01 5.10782480e-01
-4.17426825e-01 -8.30286860e-01 5.29661655e-01 -8.17323849e-02
5.19212723e-01 9.65253234e-01 1.24010980e+00 -4.19609308e-01
-3.65077913e-01 -9.34466347e-02 2.41792187e-01 -1.21710680e-01
5.49548388e-01 6.54223919e-01 -8.36565018e-01 1.71268992e-02
-5.57481274e-02 -1.33095995e-01 4.73958641e-01 1.99040189e-01
5.22448897e-01 -2.62608588e-01 -4.77694839e-01 6.44541740e-01
1.66572642e+00 3.95021379e-01 8.00237894e-01 4.08698350e-01
-8.61707050e-03 4.13417965e-01 3.90954375e-01 6.36819899e-01
-2.28277072e-01 1.89812444e-02 3.76047432e-01 1.91993341e-01
2.73956865e-01 -1.67796221e-02 3.25151116e-01 1.39232314e+00
-4.67415988e-01 -6.18736088e-01 -9.04242992e-01 3.94176513e-01
-1.16913867e+00 -7.30614603e-01 -2.82453835e-01 2.11593223e+00
1.08298767e+00 -1.09772138e-01 5.42717464e-02 -3.67816240e-02
5.99796891e-01 -5.75802445e-01 -6.06545866e-01 -5.46224833e-01
-8.84579301e-01 -1.97296143e-01 6.50287330e-01 1.41356319e-01
-3.20548117e-01 5.98294258e-01 7.64088106e+00 4.39194620e-01
-1.24436569e+00 -9.12848487e-02 7.88281083e-01 -3.74265127e-02
-5.44291258e-01 3.35576117e-01 -2.83827066e-01 6.42810225e-01
7.37068534e-01 -9.33392465e-01 1.10478267e-01 1.45279452e-01
2.69744903e-01 -5.70629716e-01 -1.14328098e+00 4.93752062e-01
-6.14585698e-01 -1.23125958e+00 -1.67396814e-01 5.50404131e-01
7.70634592e-01 -7.82115012e-02 -1.89724699e-01 -7.89870974e-03
3.14836681e-01 -8.79818678e-01 1.82009354e-01 9.09923673e-01
1.13655281e+00 -5.59978426e-01 8.14427972e-01 5.93434870e-01
-7.31683910e-01 2.79433608e-01 -3.68908763e-01 -2.80779421e-01
4.15838242e-01 1.06382966e+00 -1.01873016e+00 5.77632725e-01
1.81814328e-01 1.96886554e-01 -4.55216348e-01 9.18826818e-01
4.01342243e-01 5.84157884e-01 -6.25093460e-01 -4.16798949e-01
-2.27223620e-01 -9.87253010e-01 8.11693251e-01 9.48426664e-01
7.94987321e-01 3.09978396e-01 -6.07021093e-01 7.52296150e-01
-3.84728163e-02 1.50647253e-01 -7.52969742e-01 -7.24717617e-01
5.54452121e-01 1.39494610e+00 -1.32266092e+00 -2.68688440e-01
2.40921572e-01 8.76261473e-01 -2.64091760e-01 1.56890452e-01
-6.33556485e-01 -2.75414914e-01 5.54561734e-01 9.48663503e-02
-1.53354660e-01 -4.36790317e-01 -6.37534618e-01 -1.06507576e+00
-6.53508306e-01 -4.22250092e-01 -5.70476763e-02 -5.42534053e-01
-1.07595837e+00 2.90703416e-01 -4.66337770e-01 -1.15265727e+00
1.15195207e-01 -5.95440209e-01 -2.72258252e-01 2.83541143e-01
-4.47984397e-01 -7.03179538e-01 -1.19560979e-01 -1.77997202e-01
-1.67028263e-01 6.80595115e-02 9.89844799e-01 -4.95775230e-02
-7.02268898e-01 -7.71578476e-02 5.89140773e-01 -4.63242680e-01
4.55762625e-01 -1.21435523e+00 -7.63724074e-02 5.88415742e-01
-4.68259722e-01 7.91612148e-01 1.11747897e+00 -7.31474340e-01
-1.56485820e+00 -8.60359251e-01 7.96754241e-01 -1.32335335e-01
4.67644334e-01 -4.63815033e-01 -4.98493582e-01 3.87779683e-01
3.75700414e-01 -4.33617383e-01 1.11165524e+00 -1.13441490e-01
1.68115824e-01 1.63696781e-01 -1.34261894e+00 1.03515065e+00
1.06058824e+00 -3.40565562e-01 -1.25876844e-01 3.90906334e-01
5.35754502e-01 9.63146761e-02 -1.33374977e+00 3.38543415e-01
8.93130839e-01 -8.30283284e-01 4.22667950e-01 -5.06142914e-01
3.98946017e-01 -4.96767133e-01 -3.78916472e-01 -1.20700455e+00
-5.57560146e-01 -1.00975204e+00 4.28557605e-01 1.44706178e+00
4.65736926e-01 -9.81211662e-01 3.12511027e-01 2.69094288e-01
4.06270148e-03 -6.65617645e-01 -1.34109938e+00 -1.21382487e+00
4.77228045e-01 5.76031506e-01 7.27397859e-01 1.28933942e+00
7.20708132e-01 5.93507737e-02 6.74938038e-02 -5.17762840e-01
1.73596546e-01 -1.78097516e-01 6.22441053e-01 -1.46549249e+00
5.37203997e-02 -3.65037531e-01 -4.05853480e-01 -4.74351823e-01
-2.74387330e-01 -1.16003752e+00 -2.48563830e-02 -1.39591825e+00
3.28085572e-01 -2.56171614e-01 -2.22903684e-01 1.93602107e-02
3.70027661e-01 9.13981125e-02 8.78674090e-02 -3.61275836e-03
-1.77962035e-01 3.70443016e-01 1.43952262e+00 4.17476475e-01
-1.05896816e-01 -5.47121346e-01 -7.87985981e-01 6.39780104e-01
9.53035831e-01 -6.13736212e-01 9.55093428e-02 3.11804175e-01
8.57427657e-01 2.84729779e-01 -6.34412989e-02 -1.01964200e+00
-1.00669101e-01 -5.11450469e-01 2.61102289e-01 -3.65242839e-01
2.58580476e-01 -7.34275520e-01 1.05159235e+00 1.09264886e+00
2.87276283e-02 -2.38093082e-02 -5.57347434e-03 5.50406754e-01
2.02046514e-01 -9.21807513e-02 1.00747824e+00 -1.37580201e-01
1.26347035e-01 -4.67863493e-02 -1.27139139e+00 -1.41857848e-01
1.43907499e+00 -8.46955955e-01 -5.63629150e-01 -1.20688500e-02
-5.51688910e-01 -2.85235882e-01 1.30273807e+00 -2.98756629e-01
1.64537221e-01 -1.26730061e+00 -7.44926929e-01 -1.17795981e-01
-2.42700621e-01 -5.86793602e-01 1.00893550e-01 1.37312603e+00
-9.86151576e-01 2.49877125e-01 -9.28809583e-01 -6.47274017e-01
-8.99331987e-01 4.89164859e-01 6.85303658e-02 -1.31981701e-01
-8.47623646e-02 4.97350156e-01 1.23055279e-01 -1.56050950e-01
-7.93827713e-01 -5.65041661e-01 -5.60656153e-02 2.48491898e-01
-1.81894109e-01 5.34949720e-01 -3.06963891e-01 -4.48828399e-01
-5.92530668e-01 6.06406391e-01 4.22752917e-01 -1.46745846e-01
1.14158916e+00 -2.55560338e-01 -9.49301302e-01 1.12325132e+00
1.04465866e+00 4.04852033e-01 -6.00132525e-01 8.38210046e-01
1.26191512e-01 -3.96238387e-01 -4.88460511e-01 -6.12298965e-01
-4.61117417e-01 1.71650350e-01 1.80555642e-01 7.49939203e-01
1.07500696e+00 -1.38434395e-01 6.62675619e-01 3.72517467e-01
9.10648525e-01 -1.22179472e+00 -2.89701611e-01 7.39779770e-01
7.77274311e-01 -5.30538678e-01 -6.68959692e-03 -3.97163391e-01
-1.80272192e-01 9.04060543e-01 2.74365753e-01 -1.83212310e-02
7.13750482e-01 8.05146635e-01 -3.97669047e-01 1.62111700e-01
-1.26280487e+00 2.08713278e-01 -6.35843158e-01 6.54049933e-01
9.07128215e-01 2.41642192e-01 -1.14948070e+00 8.25375140e-01
-4.52616662e-01 5.17216623e-01 1.12809289e+00 7.03730762e-01
-6.66229069e-01 -9.30725098e-01 -3.15803230e-01 4.01288569e-01
-5.29061079e-01 1.31215170e-01 -8.67050052e-01 8.46769869e-01
3.71735126e-01 5.95433474e-01 3.51566374e-02 -3.56864989e-01
-2.15630233e-01 3.01587880e-01 6.28365099e-01 -4.20424268e-02
-4.73376662e-01 -1.80577904e-01 5.64160869e-02 -1.38604432e-01
-3.57278198e-01 -9.89187121e-01 -1.53886402e+00 -4.98659819e-01
-4.68940347e-01 4.18194443e-01 5.88949561e-01 7.21928418e-01
4.38597023e-01 3.68879169e-01 5.26863933e-01 -2.66530663e-01
1.40959099e-01 -1.07261932e+00 -1.17274320e+00 3.33590694e-02
3.84084582e-02 -1.30128756e-01 -7.56684065e-01 3.65550220e-01] | [5.625861167907715, 4.190752029418945] |
15963369-0900-430c-a4fa-ff2b57aee34f | evaluating-n-best-calibration-of-natural | null | null | https://aclanthology.org/2022.sigdial-1.54 | https://aclanthology.org/2022.sigdial-1.54.pdf | Evaluating N-best Calibration of Natural Language Understanding for Dialogue Systems | A Natural Language Understanding (NLU) component can be used in a dialogue system to perform intent classification, returning an N-best list of hypotheses with corresponding confidence estimates. We perform an in-depth evaluation of 5 NLUs, focusing on confidence estimation. We measure and visualize calibration for the 10 best hypotheses on model level and rank level, and also measure classification performance. The results indicate a trade-off between calibration and performance. In particular, Rasa (with Sklearn classifier) had the best calibration but the lowest performance scores, while Watson Assistant had the best performance but a poor calibration. | ['Staffan Larsson', 'Alexander Berman', 'Ranim Khojah'] | null | null | null | null | sigdial-acl-2022-9 | ['intent-classification'] | ['natural-language-processing'] | [-3.27170819e-01 5.76814175e-01 -4.95403439e-01 -6.26273751e-01
-1.02873039e+00 -7.89585650e-01 8.29754949e-01 5.60510874e-01
-4.59308863e-01 9.13257778e-01 5.74986756e-01 -6.70575857e-01
-2.59589314e-01 -3.39271247e-01 -5.43837771e-02 2.28423346e-02
7.01401010e-02 1.01340425e+00 1.51267415e-02 -1.58263624e-01
5.31485140e-01 1.80290062e-02 -1.16502333e+00 2.91805983e-01
1.19265950e+00 9.68919873e-01 -1.29063785e-01 1.09327126e+00
-1.81575254e-01 9.72744048e-01 -1.00036561e+00 -6.60524309e-01
-2.34428436e-01 -3.61270338e-01 -1.06425929e+00 -7.07077265e-01
3.57710153e-01 -2.14878067e-01 2.64658123e-01 5.99103868e-01
1.30619600e-01 -1.04208939e-01 1.00256681e+00 -1.42994702e+00
-3.65103818e-02 7.36587048e-01 8.21416751e-02 1.17245398e-01
1.18058181e+00 1.68788925e-01 1.34390140e+00 -8.02991748e-01
4.57197607e-01 1.29733753e+00 6.25012517e-01 5.31227469e-01
-1.34652603e+00 -7.31672823e-01 -2.18643174e-01 1.32353539e-02
-1.20589685e+00 -5.36824942e-01 8.26605260e-02 -5.34899354e-01
1.14091218e+00 4.03541327e-01 4.35296863e-01 1.05784774e+00
3.31583589e-01 6.65989101e-01 1.38415253e+00 -6.54192448e-01
6.25294149e-01 7.15716898e-01 9.03459013e-01 8.95120203e-01
1.43795326e-01 -4.95545864e-02 -6.52767658e-01 -5.03410757e-01
2.63251126e-01 -4.55256730e-01 -4.39821929e-01 1.28295748e-02
-1.06271017e+00 1.07870924e+00 3.36435974e-01 1.80109426e-01
-9.52941999e-02 -3.16995382e-01 2.79987335e-01 5.98783433e-01
1.82378024e-01 1.26043200e+00 -8.27687263e-01 -5.57093322e-01
-6.72870815e-01 5.36168739e-02 1.73738968e+00 8.49647164e-01
6.73004031e-01 -3.77079725e-01 -4.98191655e-01 8.90562117e-01
2.77607024e-01 1.69463053e-01 5.44438660e-01 -8.99541497e-01
1.80607721e-01 7.28698969e-01 2.35284105e-01 -5.58556736e-01
-6.65574610e-01 -2.77293026e-01 -4.32273775e-01 1.39777005e-01
4.63940710e-01 -4.10094291e-01 -7.25115895e-01 1.72228026e+00
-2.56288290e-01 -1.51166484e-01 5.40432870e-01 5.00906885e-01
9.49470401e-01 5.70986211e-01 4.34151322e-01 -2.61740297e-01
1.35575712e+00 -8.65369201e-01 -6.79247856e-01 -4.69715118e-01
9.92231309e-01 -6.53835058e-01 1.21427631e+00 3.47763121e-01
-6.57913446e-01 -4.84860212e-01 -1.29986656e+00 2.74428308e-01
-2.99001932e-01 4.59148139e-02 7.22671211e-01 8.75251591e-01
-9.86729145e-01 3.90499353e-01 -4.04806256e-01 -5.79770505e-01
-3.05149972e-01 1.33079633e-01 -4.14179385e-01 1.16201945e-01
-1.29820061e+00 1.50248146e+00 7.63596654e-01 -4.96222943e-01
-5.86807549e-01 -6.85747385e-01 -8.22341561e-01 1.02031223e-01
3.76918674e-01 -5.23349941e-01 1.68854713e+00 -3.99712354e-01
-1.49018312e+00 3.82610023e-01 1.40878394e-01 -7.44102478e-01
5.94104052e-01 -3.89432520e-01 -3.64772737e-01 -2.51843095e-01
7.35714883e-02 7.01623499e-01 1.29511088e-01 -1.19457531e+00
-9.26006436e-01 -6.28540292e-02 2.99777597e-01 3.63624752e-01
-2.13670075e-01 -1.00384980e-01 -1.65518492e-01 -1.35201231e-01
8.90619680e-02 -6.52415752e-01 5.25049679e-02 -6.45413458e-01
-2.59535283e-01 -6.39746130e-01 3.65063161e-01 -6.90110028e-01
1.81132901e+00 -1.56490886e+00 -1.85479634e-02 2.02596739e-01
1.43892407e-01 -1.34982184e-01 3.42971206e-01 3.53774369e-01
3.25149787e-03 3.53968769e-01 1.41026780e-01 -2.71549374e-02
6.58060834e-02 1.45836756e-01 -2.45668009e-01 -1.37469783e-01
-3.32994573e-02 5.61863899e-01 -8.23152900e-01 -6.48137748e-01
1.90450862e-01 -1.97851121e-01 -6.51256979e-01 6.80495858e-01
-6.38614774e-01 -9.87568963e-03 -1.97976738e-01 5.99159718e-01
6.03329353e-02 -2.74360597e-01 2.43787080e-01 -3.14044774e-01
-2.99500916e-02 7.01499462e-01 -8.76340508e-01 1.37812328e+00
-9.21483457e-01 6.40582860e-01 1.10066915e-03 -2.06990913e-01
1.14153230e+00 2.17572421e-01 -1.88397288e-01 -3.76521140e-01
-9.41836685e-02 2.91280717e-01 3.06864679e-01 -2.95364231e-01
6.08019829e-01 -1.23846039e-01 -5.61850190e-01 6.84175551e-01
2.45011687e-01 -2.88278788e-01 -1.05048977e-01 6.12862885e-01
1.22587776e+00 -3.28246236e-01 8.28829825e-01 -6.52001321e-01
4.37590718e-01 3.68794918e-01 -1.06503099e-01 1.19060826e+00
-5.73306158e-02 1.99442998e-01 8.23916078e-01 -2.42755264e-01
-7.35966444e-01 -8.27677488e-01 -1.34698749e-01 1.25702238e+00
-2.39099100e-01 -7.22036004e-01 -7.43925929e-01 -1.07239223e+00
-6.68109953e-02 1.76655221e+00 -5.26113331e-01 -2.07465619e-01
1.38063123e-03 -2.01918736e-01 7.44125605e-01 5.12779832e-01
6.10834897e-01 -9.23787475e-01 -5.84527671e-01 -3.53868008e-02
-3.18552047e-01 -8.42202127e-01 -2.14812741e-01 6.88195109e-01
-7.00816631e-01 -1.01792324e+00 -1.09627627e-01 -4.80802268e-01
2.69209325e-01 -4.03971612e-01 1.46495831e+00 2.21009970e-01
3.00782681e-01 6.62353992e-01 -4.01349097e-01 -5.45263529e-01
-9.45407927e-01 3.46878350e-01 3.07284862e-01 -7.17219353e-01
5.32297492e-01 1.41057238e-01 -5.00235409e-02 5.76555312e-01
-4.39667493e-01 -1.08609879e-02 5.60732603e-01 1.18228507e+00
-1.44567147e-01 -1.95671022e-01 5.85289121e-01 -7.72305965e-01
1.35275412e+00 -5.13777912e-01 -3.88706297e-01 6.59728765e-01
-1.18715334e+00 3.81822795e-01 4.26250994e-01 -5.68637624e-02
-1.04532743e+00 -2.29574889e-01 -2.90764332e-01 -1.19507089e-01
-2.88217276e-01 8.45371664e-01 -6.71050921e-02 3.10312659e-01
1.12540579e+00 -2.01958522e-01 -7.53234923e-02 -2.60873318e-01
2.14346841e-01 1.04284811e+00 5.18341422e-01 -6.63863122e-01
2.21585900e-01 -3.80106807e-01 -6.98189378e-01 -7.60219097e-01
-1.15714324e+00 -3.96512985e-01 -6.25686884e-01 -3.82416189e-01
6.52802885e-01 -6.61983192e-01 -7.09490776e-01 -1.92527529e-02
-1.01415706e+00 -4.24449980e-01 -7.74666741e-02 6.22251809e-01
-5.25517046e-01 7.57914856e-02 -3.03516626e-01 -1.12768209e+00
-4.71050024e-01 -9.45905626e-01 5.40091455e-01 4.82529283e-01
-1.14309919e+00 -8.91959786e-01 2.27039978e-01 3.72877032e-01
3.06015730e-01 -3.44782919e-01 1.30020761e+00 -1.59393716e+00
9.41229612e-02 -3.22924197e-01 -1.51884079e-01 -3.68033834e-02
-5.19372113e-02 -1.40859842e-01 -1.01853180e+00 -2.15225518e-01
1.28574938e-01 -7.13390052e-01 5.67822337e-01 2.21560806e-01
5.53309083e-01 -4.88621056e-01 -4.04105842e-01 5.71909435e-02
1.00191545e+00 3.37308288e-01 4.27257091e-01 4.23246622e-01
1.60065204e-01 7.01148748e-01 1.07138562e+00 2.28979513e-01
2.62802571e-01 6.62607968e-01 -6.00248724e-02 3.56819391e-01
2.91694403e-01 -5.72728217e-01 3.01045686e-01 6.31004333e-01
3.93104523e-01 -3.73389781e-01 -1.42847002e+00 -1.82093885e-02
-1.78010881e+00 -6.18193686e-01 2.97318518e-01 2.31281805e+00
7.47245431e-01 8.50102007e-01 7.46913552e-02 -3.99825387e-02
4.36632186e-01 -1.94694381e-02 -4.18326139e-01 -6.25876129e-01
1.38628438e-01 5.70557788e-02 2.07938686e-01 9.72916543e-01
-8.20068240e-01 9.08723474e-01 8.22198582e+00 7.43404925e-01
-3.80996823e-01 -2.65242457e-01 8.69657695e-01 3.22540015e-01
-3.43244612e-01 2.34312668e-01 -8.61196816e-01 2.87916332e-01
1.45556200e+00 -3.54961634e-01 1.20058935e-02 1.14668798e+00
-3.13527077e-01 -7.11828291e-01 -1.35920548e+00 6.89613342e-01
-1.36197209e-02 -1.25618923e+00 1.95575152e-02 -1.61638007e-01
1.95938289e-01 -2.51798272e-01 -5.38288414e-01 9.48041856e-01
6.98315859e-01 -1.23385191e+00 3.92011642e-01 6.55298650e-01
7.92620897e-01 -8.84858191e-01 1.14044988e+00 7.37463295e-01
-6.44833207e-01 -2.10513353e-01 -6.75745532e-02 -1.44748434e-01
-2.25142375e-01 1.65246248e-01 -1.56103015e+00 2.88470179e-01
5.21666706e-01 2.53829598e-01 -6.97342813e-01 8.09198201e-01
-2.53177792e-01 9.65974927e-01 -4.03995544e-01 -6.62945628e-01
-8.20385665e-03 -8.11211318e-02 4.01490808e-01 1.67821002e+00
-4.04631495e-02 4.80554819e-01 3.68663192e-01 6.03713751e-01
7.46233091e-02 1.69871509e-01 -4.70319331e-01 4.06168811e-02
8.05214882e-01 9.84305620e-01 -5.85202754e-01 -6.52022839e-01
-2.07340121e-01 7.34430850e-01 3.64509106e-01 -1.38112366e-01
-5.47294259e-01 -1.08713925e-01 4.10361886e-01 -1.52336240e-01
-4.42316443e-01 -3.48635539e-02 -5.85416198e-01 -8.53249133e-01
-5.62514663e-01 -1.12230933e+00 9.69558358e-01 -8.92707705e-01
-1.19512618e+00 6.46343529e-01 2.78675377e-01 -9.55571651e-01
-6.78634405e-01 -6.27636492e-01 -6.35730803e-01 8.67989421e-01
-6.35958076e-01 -5.08575857e-01 -5.47715127e-01 5.06209210e-02
7.93971658e-01 -4.67475742e-01 1.37725353e+00 -3.50392729e-01
-4.80574071e-01 7.04657018e-01 -2.68050283e-01 8.34334716e-02
9.03684139e-01 -1.60325968e+00 1.98628217e-01 1.16630800e-01
-4.95646968e-02 9.56637263e-01 1.10474610e+00 -9.02520001e-01
-1.07128680e+00 -4.21845466e-01 8.79565716e-01 -7.91781068e-01
7.48501778e-01 -1.44780546e-01 -9.61255908e-01 6.20236635e-01
4.37574178e-01 -1.04124701e+00 9.55713868e-01 6.40931845e-01
-4.84249562e-01 2.36470237e-01 -1.15429485e+00 5.76189399e-01
4.15186495e-01 -6.05989337e-01 -1.06529069e+00 2.36520201e-01
6.39286876e-01 -5.26355982e-01 -1.10700035e+00 3.58128369e-01
7.76786208e-01 -9.81440067e-01 5.20444334e-01 -6.80129051e-01
4.00951415e-01 -6.76166639e-02 -1.39181212e-01 -1.44540524e+00
-2.60388166e-01 -3.77099842e-01 -1.03530146e-01 1.15298092e+00
1.05585587e+00 -6.11946106e-01 7.07219720e-01 1.33543193e+00
1.03129365e-01 -7.02993751e-01 -5.96907675e-01 -4.98900324e-01
-6.73219785e-02 -7.04266131e-01 2.01465294e-01 8.37935627e-01
6.23840928e-01 1.03580129e+00 -2.37062871e-01 -1.36058718e-01
2.49682754e-01 3.82981636e-02 9.07442272e-01 -1.39221048e+00
-1.13273762e-01 -5.13946891e-01 -2.82637060e-01 -9.52709317e-01
1.96045265e-01 -6.74756467e-01 3.00237805e-01 -1.43657994e+00
3.48644912e-01 -3.55726957e-01 -1.10320218e-01 5.54058850e-01
-3.71288598e-01 -4.29792285e-01 5.01016155e-02 3.86823833e-01
-6.36151075e-01 1.39038637e-01 2.08380401e-01 -2.22573541e-02
-5.12944043e-01 2.84129560e-01 -8.71129692e-01 7.09979653e-01
8.67182136e-01 -3.62964332e-01 -3.51993173e-01 2.23757640e-01
1.04056718e-02 4.36198801e-01 -1.23052597e-01 -1.21647549e+00
5.62497020e-01 -2.78884228e-02 4.98968929e-01 -7.39009917e-01
1.56079173e-01 -5.24564505e-01 4.15027998e-02 6.78031802e-01
-8.82894933e-01 2.24727035e-01 3.06291521e-01 3.88737351e-01
-9.43755209e-02 -7.21555293e-01 7.67336249e-01 -1.12485019e-02
-7.54939735e-01 -4.69169021e-01 -5.40899456e-01 -1.93153881e-02
7.28542745e-01 -1.32484421e-01 -3.62376660e-01 -8.02615166e-01
-6.90977514e-01 5.58561981e-01 4.86915827e-01 3.44374806e-01
5.71307421e-01 -7.87334144e-01 -6.76423132e-01 1.34252265e-01
4.93606269e-01 -5.30380130e-01 -4.73020047e-01 3.89881194e-01
-3.94604802e-01 7.09357083e-01 -9.43061039e-02 -5.91352344e-01
-1.13450480e+00 7.51109421e-02 4.26871240e-01 -5.32163858e-01
-2.25602165e-01 7.70301342e-01 -2.38873705e-01 -7.76161373e-01
5.77620685e-01 -1.92580856e-02 -5.82833052e-01 3.31106573e-01
5.10599077e-01 4.81741756e-01 9.54901986e-03 -1.58488408e-01
-3.38993460e-01 7.82737136e-02 -2.96218723e-01 -4.94028986e-01
6.87207282e-01 1.34416088e-01 1.85458302e-01 6.69623733e-01
8.28059793e-01 -9.81887579e-02 -6.27109826e-01 9.61725321e-03
5.29372156e-01 -2.66570121e-01 3.23910937e-02 -1.57417917e+00
-1.31539613e-01 2.60120869e-01 4.96573478e-01 4.87380952e-01
4.80141193e-01 4.04228866e-02 -2.53013037e-02 8.23587835e-01
5.03224432e-01 -1.17497253e+00 1.18880369e-01 7.94523776e-01
9.43303466e-01 -1.48127854e+00 2.27399111e-01 -8.10500458e-02
-1.15963447e+00 1.25284040e+00 1.12908816e+00 3.79351497e-01
5.97823322e-01 2.60513544e-01 2.14669764e-01 -2.95699775e-01
-1.09584987e+00 1.35104209e-01 3.94729376e-01 2.23354876e-01
7.00071752e-01 1.30391791e-01 -4.71761465e-01 7.07757533e-01
-5.97703338e-01 -3.72634768e-01 3.55182588e-01 6.14945352e-01
-6.06002390e-01 -7.22893536e-01 -2.87735850e-01 1.06861031e+00
1.27718866e-01 -7.48379529e-02 -9.13584411e-01 8.79484415e-01
-6.83508098e-01 1.24914956e+00 -2.90143099e-02 -8.47480774e-01
5.79714179e-01 6.53136432e-01 -1.65958092e-01 -6.57727242e-01
-9.37750459e-01 -3.31815869e-01 7.87882209e-01 -3.70240867e-01
2.66156495e-02 -3.71077806e-01 -1.12538052e+00 -2.09293872e-01
-7.46329308e-01 7.66579807e-01 5.81639051e-01 8.55907798e-01
1.09974286e-02 1.46560594e-01 5.37648201e-01 -2.64390290e-01
-7.64695942e-01 -1.51950860e+00 -6.29678890e-02 8.32779482e-02
-2.92080864e-02 -5.32471836e-01 -6.68534040e-01 -5.25749326e-01] | [12.7022066116333, 8.016667366027832] |
8126f33e-b241-45cb-baf5-e9b79b45d2ab | simultaneous-indoor-and-outdoor-3d | 2207.05344 | null | https://arxiv.org/abs/2207.05344v2 | https://arxiv.org/pdf/2207.05344v2.pdf | Simultaneous Indoor and Outdoor 3D Localization with STAR-RIS-Assisted Millimeter Wave Systems | Simultaneously transmitting (refracting) and reflecting reconfigurable intelligent surfaces (STAR-RISs) have been recently identified to improve the spectrum/energy efficiency and extend the communication range. However, their potential for enhanced concurrent indoor and outdoor localization has not yet been explored. In this paper, we study the fundamental limits, i.e., the Cram\'er Rao lower bounds (CRLBs) via Fisher information analyses, on the three-dimensional (3D) localization performance with a STAR-RIS at millimeter wave frequencies. The effect of the power splitting between refraction and reflection at the STAR-RIS as well as the power allocation between the two mobile stations (MSs) are investigated. By maximizing the principal angle between the two subspaces corresponding to the STAR-RIS reflection and refraction matrices, we are able to find the optimal solutions for these simultaneous operations. We verify that high-accuracy 3D localization can be achieved for both indoor and outdoor MSs when the system parameters are well optimized. | ['George C. Alexandropoulos', 'Aymen Fakhreddine', 'Jiguang He'] | 2022-07-12 | null | null | null | null | ['outdoor-localization'] | ['robots'] | [ 3.34286451e-01 2.92643428e-01 1.80857420e-01 1.23801544e-01
-6.65680528e-01 -5.86127639e-01 1.40834346e-01 -3.75126123e-01
-1.08599916e-01 6.09684706e-01 8.94951459e-04 -6.70139790e-01
-7.14525819e-01 -7.66891181e-01 -5.08872151e-01 -1.13808298e+00
-5.28129339e-01 4.43601608e-02 -2.58734554e-01 -1.24746144e-01
5.28748780e-02 6.06703639e-01 -1.40725851e+00 -5.91221631e-01
1.01600683e+00 1.20626187e+00 2.25945398e-01 6.87723756e-01
3.25889707e-01 -1.13592140e-01 -7.28848636e-01 9.08067673e-02
4.11593348e-01 -9.44321975e-02 5.12733944e-02 -2.18531087e-01
-4.12007086e-02 1.49664690e-03 -2.70756096e-01 9.69462156e-01
6.56459153e-01 -1.89577583e-02 6.37611687e-01 -9.45996463e-01
-9.16398764e-02 3.47993016e-01 -6.21283889e-01 -7.01492503e-02
5.23129344e-01 -5.30839384e-01 4.17404085e-01 -5.08565843e-01
2.35881597e-01 8.62668276e-01 7.17869997e-01 3.71957906e-02
-9.17722702e-01 -7.50839472e-01 -4.37945604e-01 -1.41827583e-01
-1.81835139e+00 -5.65645397e-01 6.45735145e-01 -1.51602566e-01
4.78112459e-01 7.08475292e-01 5.32612622e-01 4.72924918e-01
6.68120861e-01 -1.30234852e-01 8.57745290e-01 -8.55242550e-01
2.71288872e-01 3.27867061e-01 -2.34336317e-01 1.04456365e+00
9.22893882e-01 2.32686877e-01 -4.73134875e-01 -3.03104818e-01
7.55700588e-01 -4.26885009e-01 -4.74922448e-01 -5.54899216e-01
-1.13369119e+00 3.00412387e-01 4.41993177e-01 6.91253603e-01
-2.92722762e-01 2.93262720e-01 -6.33080602e-01 3.33150893e-01
2.25394323e-01 5.32306850e-01 -1.01101343e-02 2.57995516e-01
-5.14398217e-01 -4.57018554e-01 7.90983796e-01 1.14646971e+00
4.65821505e-01 -1.11010894e-01 -1.85121484e-02 3.78142536e-01
8.67133975e-01 1.29450691e+00 -2.24952966e-01 -6.41399384e-01
6.16436839e-01 1.68461040e-01 6.49842203e-01 -1.05837536e+00
-7.54444003e-01 -1.43079090e+00 -7.48046398e-01 -6.32922649e-02
3.99340063e-01 -7.30623007e-01 -4.97997195e-01 1.66320336e+00
3.56461108e-01 1.10400610e-01 5.84433079e-01 7.11476862e-01
4.21299160e-01 5.37079930e-01 -4.80292648e-01 -4.88749236e-01
1.23504281e+00 -1.46300942e-01 -5.57350755e-01 -4.27501440e-01
6.35327101e-01 -7.26808608e-01 4.76086110e-01 2.57304788e-01
-1.04768586e+00 -1.98009628e-04 -1.48356009e+00 5.34374774e-01
1.17249340e-01 3.77631724e-01 3.61269802e-01 1.33175480e+00
-9.63220656e-01 -1.13395400e-01 -1.00595915e+00 -1.22335561e-01
-5.13179973e-02 4.40396070e-01 5.54432496e-02 -1.81765214e-01
-8.92962515e-01 6.09183013e-01 -5.41284084e-01 3.08522463e-01
-1.66094616e-01 -5.30624807e-01 -2.13636562e-01 -3.90522629e-02
1.39150023e-01 -7.14024842e-01 7.10085452e-01 -2.94591784e-01
-1.63572335e+00 3.37240160e-01 -1.93874553e-01 -1.24052040e-01
2.99045090e-02 1.26483634e-01 -6.28049552e-01 7.23964348e-02
-6.13259058e-03 -2.58431911e-01 1.92330569e-01 -1.39861298e+00
-5.69278955e-01 -7.26660013e-01 -1.70486122e-01 3.57304722e-01
-2.53068388e-01 -5.25026023e-01 -1.07649945e-01 -1.97020233e-01
9.37223375e-01 -1.37356985e+00 -3.85672957e-01 -3.27780336e-01
-5.30703068e-01 1.91921756e-01 4.57986295e-01 -2.61014134e-01
7.64363348e-01 -2.48575735e+00 -9.57167894e-02 7.83124924e-01
-2.49096125e-01 -1.80288196e-01 -2.62710918e-02 3.67626816e-01
3.89854342e-01 -1.04592569e-01 3.54713559e-01 -2.09198341e-01
-1.97627366e-01 -2.50017375e-01 8.75717849e-02 1.01969206e+00
-7.59585738e-01 2.67102808e-01 -6.85060143e-01 1.42495364e-01
-4.28447686e-02 5.96042037e-01 -3.88954341e-01 -1.22354463e-01
4.90090996e-01 6.70969307e-01 -1.01642561e+00 5.68427145e-01
8.53486359e-01 -3.82660590e-02 3.52698624e-01 -2.71306723e-01
-5.87494254e-01 8.54057446e-02 -1.22499514e+00 1.42055464e+00
-8.37085843e-01 5.57949603e-01 3.47860545e-01 -5.96236825e-01
1.02974617e+00 2.18562633e-01 6.47743583e-01 -8.71394455e-01
2.71683764e-02 5.24826586e-01 -1.57172427e-01 -2.50433236e-01
3.15311193e-01 -2.85945863e-01 -2.82105714e-01 3.03262025e-01
-2.90283352e-01 -2.00177673e-02 -3.99981856e-01 -1.40294075e-01
1.16003036e+00 -3.07919919e-01 2.02448010e-01 -8.66112053e-01
5.08628368e-01 -3.40943724e-01 3.38348478e-01 7.02876568e-01
3.65877360e-01 -8.85682032e-02 6.55537620e-02 3.26357037e-01
-3.93445283e-01 -1.26109445e+00 -2.64908582e-01 3.38472158e-01
1.10877657e+00 6.18011542e-02 -3.29670966e-01 1.44188613e-01
1.66679919e-01 9.30774033e-01 -3.93512622e-02 -2.54068702e-01
-1.42264113e-01 -8.30342948e-01 4.03837949e-01 -1.31934538e-01
5.51637173e-01 3.59427899e-01 -9.21030939e-01 3.51348370e-02
-2.32024893e-01 -1.12682891e+00 4.70342338e-02 3.13999981e-01
-6.21651590e-01 -8.62205923e-01 -3.08763385e-01 -4.49221194e-01
8.90367210e-01 9.06143129e-01 4.69363719e-01 -3.59211981e-01
-4.92250212e-02 8.88647318e-01 -2.51434892e-01 -3.64093184e-01
1.37134165e-01 -9.96403545e-02 3.88307571e-01 5.44323884e-02
-5.10039628e-01 -6.72998965e-01 -8.21620286e-01 1.01975441e+00
-6.24113232e-02 -5.90941049e-02 5.62833250e-01 1.16592474e-01
4.79976237e-01 4.64158863e-01 4.29463744e-01 -5.44354245e-02
1.24849901e-01 -4.01726574e-01 -1.00863111e+00 2.37068325e-01
-5.14764249e-01 2.91850090e-01 2.63522327e-01 3.61090563e-02
-1.16561830e+00 -7.77597800e-02 1.38135115e-02 4.99348253e-01
1.87793985e-01 3.42613608e-01 -3.69984716e-01 -9.66085792e-01
5.43734133e-01 2.49057431e-02 -4.59195554e-01 -8.59689787e-02
2.03061134e-01 7.21290231e-01 2.01935664e-01 -2.30680227e-01
1.19179749e+00 6.55915558e-01 7.49356985e-01 -1.45251834e+00
-6.27544343e-01 -4.20078129e-01 6.96726069e-02 -3.59806001e-01
6.25965416e-01 -1.06232417e+00 -8.79697740e-01 4.84887091e-03
-7.73163378e-01 -1.03870504e-01 2.50027534e-02 1.13272536e+00
-4.01620835e-01 1.91465825e-01 1.50948331e-01 -1.39369607e+00
-1.06554300e-01 -7.88452923e-01 9.05024230e-01 4.23581541e-01
1.24063730e-01 -7.07093060e-01 -3.67924683e-02 2.23259926e-01
6.61961854e-01 2.95933127e-01 8.42274725e-01 1.85526133e-01
-8.46800745e-01 -4.22638744e-01 1.20881408e-01 -4.80793774e-01
1.03299797e-01 -7.34551668e-01 -5.81942141e-01 -5.94146788e-01
1.75882652e-01 5.87101281e-01 3.35138440e-01 7.41591275e-01
4.93405372e-01 -3.94875228e-01 -9.37956333e-01 9.27077115e-01
1.48772705e+00 4.47205275e-01 6.35801554e-01 9.37310979e-02
-1.17894625e-02 2.86596775e-01 7.67119765e-01 5.19495070e-01
1.40762255e-01 7.61809707e-01 6.15570247e-01 1.48322433e-01
1.13850743e-01 1.76760763e-01 3.25779527e-01 5.18799901e-01
-2.39188135e-01 -8.68119955e-01 -4.76088166e-01 3.36828753e-02
-1.21175492e+00 -4.63503808e-01 -2.80044407e-01 2.61155224e+00
2.27201022e-02 -2.02870682e-01 -4.81997967e-01 -6.94124773e-02
4.79109645e-01 4.98551652e-02 -3.51576775e-01 3.73569131e-01
-1.23951554e-01 -1.56910315e-01 1.28811038e+00 8.17850769e-01
-6.14663899e-01 1.00657634e-01 5.92907667e+00 4.03575182e-01
-8.13958883e-01 6.50569871e-02 9.56446752e-02 -1.62302569e-01
-8.37581158e-01 -1.04827322e-01 -8.76381338e-01 6.35876358e-02
7.46137023e-01 -7.16716349e-02 4.27801520e-01 3.97540003e-01
1.54879466e-01 -4.76866454e-01 -5.31998932e-01 1.02670276e+00
1.35691389e-02 -8.59710872e-01 -5.64602554e-01 6.16915166e-01
5.85264742e-01 -7.74675459e-02 3.03678691e-01 -2.46606782e-01
-8.84759352e-02 -3.23636144e-01 4.88118887e-01 6.70255959e-01
8.35269034e-01 -7.66181529e-01 6.43653512e-01 4.70983237e-01
-1.22831964e+00 -3.22192043e-01 -2.35194951e-01 2.19928682e-01
1.70952722e-01 1.16935802e+00 -7.14481235e-01 8.84161890e-01
3.16842824e-01 -1.34860784e-01 -8.33302811e-02 1.08386934e+00
-1.20627530e-01 2.49153599e-01 -8.93924892e-01 -4.75817263e-01
-7.94649050e-02 -6.30017757e-01 1.00100827e+00 5.56876779e-01
1.29754817e+00 3.93657774e-01 -2.53807366e-01 5.43018997e-01
3.07930529e-01 -2.95185093e-02 -4.20594901e-01 1.42773345e-01
7.20472157e-01 1.02604747e+00 -7.62536645e-01 3.95402312e-01
-2.03500137e-01 4.68087465e-01 -5.67348599e-01 6.11180544e-01
-6.99939847e-01 -6.34324372e-01 9.08778965e-01 2.79635638e-01
9.89848450e-02 -1.05188036e+00 -4.33319777e-01 -8.29171956e-01
-4.07796018e-02 -1.00738384e-01 -7.94344470e-02 -5.23078620e-01
-4.81654882e-01 2.14830980e-01 -2.29927883e-01 -1.16034150e+00
1.01238698e-01 -3.70599210e-01 -9.02395174e-02 6.73020959e-01
-1.37405229e+00 -6.23648822e-01 -4.02176440e-01 4.42507416e-01
-4.02269512e-01 4.47597131e-02 7.55120158e-01 4.87026244e-01
-3.44610155e-01 5.93908846e-01 9.97349203e-01 -4.51489449e-01
1.55145332e-01 -5.70251882e-01 -1.82502568e-01 8.51715147e-01
1.13508694e-01 6.43786848e-01 9.97515321e-01 -5.58016956e-01
-2.32784963e+00 -7.17204213e-01 3.26106697e-01 3.81531939e-02
1.75413370e-01 -4.57516074e-01 2.79376477e-01 3.08908343e-01
-1.05067536e-01 -4.09309655e-01 9.70070362e-01 1.39962122e-01
5.26284501e-02 -3.87942106e-01 -1.24459314e+00 7.74150193e-01
1.28108919e+00 -8.08166787e-02 3.25581342e-01 5.69319665e-01
2.93710142e-01 -4.22776103e-01 -6.61595702e-01 4.77293044e-01
8.46975207e-01 -6.62094355e-01 1.26401567e+00 3.48293811e-01
-6.50373459e-01 -4.49199885e-01 -6.97144032e-01 -1.26982450e+00
-3.74067217e-01 -7.93539762e-01 2.26523206e-01 9.23217475e-01
4.65461314e-01 -1.10631847e+00 8.11197102e-01 1.59475375e-02
-3.18669200e-01 -3.35105747e-01 -1.44224739e+00 -1.08749509e+00
-6.92559123e-01 -2.24110633e-01 4.71057594e-01 3.18343431e-01
1.29202232e-01 4.14583862e-01 -1.22520871e-01 1.23084223e+00
9.81721044e-01 1.89597830e-01 6.69081807e-01 -1.38954437e+00
-2.62405455e-01 -1.90419719e-01 -2.79727370e-01 -1.36892807e+00
-3.10035944e-01 -7.39026845e-01 2.23143578e-01 -1.77284002e+00
-5.02103806e-01 -1.11879277e+00 4.18309588e-04 -2.33755130e-02
5.15187800e-01 1.51904300e-01 -3.20658267e-01 -2.85188779e-02
-2.14180067e-01 6.06911838e-01 9.65692580e-01 1.51103094e-01
-5.00918984e-01 6.76399052e-01 -7.31166661e-01 5.47384501e-01
7.36540675e-01 -3.02021414e-01 -4.30502355e-01 -3.69760841e-01
5.51048338e-01 4.78113770e-01 1.69855401e-01 -1.54657471e+00
1.98910639e-01 -1.00239210e-01 2.63566405e-01 -4.53469038e-01
5.29473007e-01 -1.13004255e+00 8.41322482e-01 7.02213228e-01
6.36911169e-02 -6.54528618e-01 -2.78139506e-02 1.06198335e+00
5.06906509e-01 -4.43937220e-02 8.26062083e-01 2.25653410e-01
7.88104534e-02 -1.21964969e-01 -5.84264576e-01 -6.95151091e-01
1.25584459e+00 -1.85162798e-01 -4.98153001e-01 -7.46583045e-01
-2.17472866e-01 5.68684042e-02 1.79352820e-01 -6.58098534e-02
2.40558743e-01 -1.19456410e+00 -2.44515955e-01 3.16324204e-01
-2.04555616e-02 -5.14707923e-01 3.50545824e-01 1.04504669e+00
-3.38423520e-01 8.22524011e-01 9.05437768e-02 -5.89147031e-01
-1.33656991e+00 1.86844654e-02 5.14541268e-01 5.49322404e-02
-2.46800527e-01 8.48940909e-01 -4.66536172e-02 -1.79550648e-01
1.10085107e-01 -1.99710101e-01 7.86803737e-02 -3.63433599e-01
1.41182899e-01 5.46298087e-01 2.23247379e-01 -4.42039311e-01
-6.19003117e-01 1.05078113e+00 8.02146554e-01 -2.88876295e-01
1.01521230e+00 -6.91820443e-01 9.67764854e-03 -1.72601491e-02
8.62970412e-01 8.64287853e-01 -7.30453432e-01 -3.97420824e-02
-2.83928961e-01 -6.15293026e-01 3.40172172e-01 -5.85086465e-01
-7.82001972e-01 2.98745781e-01 1.06771886e+00 4.61277068e-01
9.76507127e-01 1.84106708e-01 4.15022105e-01 7.39182651e-01
1.21826398e+00 -6.68514609e-01 -1.76328912e-01 1.32064670e-01
4.84279394e-01 -4.44832623e-01 7.28041306e-02 -8.41556549e-01
2.44859248e-01 8.68216038e-01 6.72630891e-02 2.36699190e-02
7.90834904e-01 4.08719152e-01 -2.07247302e-01 -5.14537608e-03
-2.42912699e-03 -1.82983410e-02 2.19009206e-01 6.71718895e-01
3.31328630e-01 3.95585746e-01 -3.77742916e-01 4.75901306e-01
-3.24200124e-01 -6.01357341e-01 5.15164852e-01 7.29975581e-01
-8.02958131e-01 -8.53716493e-01 -7.49008656e-01 3.81413907e-01
-1.10863924e-01 3.41603458e-01 5.71019016e-02 5.50546646e-01
2.49558929e-02 1.41330802e+00 -1.58802643e-01 -4.89065468e-01
5.32053590e-01 -3.85317922e-01 7.36297309e-01 -2.95553982e-01
4.04140711e-01 2.24981874e-01 3.44177693e-01 -4.68337029e-01
-1.81999698e-01 -7.52546608e-01 -1.13534617e+00 7.64072128e-03
-7.95175374e-01 6.20173812e-01 1.15301073e+00 9.51885045e-01
7.30907261e-01 5.87276518e-01 9.26342607e-01 -5.70164263e-01
-4.00328964e-01 -3.79680723e-01 -1.14890397e+00 -8.83241355e-01
2.98972070e-01 -8.02436888e-01 -6.59995854e-01 -8.55027020e-01] | [6.275763511657715, 1.2231241464614868] |
0ae83c79-d4ca-4ec1-9520-973764425ff0 | rtfnet-rgb-thermal-fusion-network-for | null | null | https://ieeexplore.ieee.org/abstract/document/8666745 | https://yuxiangsun.github.io/pub/RAL2019_rtfnet.pdf | RTFNet: RGB-Thermal Fusion Network for Semantic Segmentation of Urban Scenes | Semantic segmentation is a fundamental capability for autonomous vehicles. With the advancements of deep learning technologies, many effective semantic segmentation networks have been proposed in recent years. However, most of them are designed using RGB images from visible cameras. The quality of RGB images is prone to be degraded under unsatisfied lighting conditions, such as darkness and glares of oncoming headlights, which imposes critical challenges for the networks that use only RGB images. Different from visible cameras, thermal imaging cameras generate images using thermal radiations. They are able to see under various lighting conditions. In order to enable robust and accurate semantic segmentation for autonomous vehicles, we take the advantage of thermal images and fuse both the RGB and thermal information in a novel deep neural network. The main innovation of this letter is the architecture of the proposed network. We adopt the encoder–decoder design concept. ResNet is employed for feature extraction and a new decoder is developed to restore the feature map resolution. The experimental results prove that our network outperforms the state of the arts. | ['Ming Liu', 'Weixun Zuo', 'Yuxiang Sun'] | 2019-03-13 | null | null | null | ieee-robotics-and-automation-letters-2019-3 | ['thermal-image-segmentation'] | ['computer-vision'] | [ 4.01093990e-01 -1.56563789e-01 1.10305011e-01 -3.89810532e-01
8.34717900e-02 -2.28036985e-01 3.93603742e-01 -7.43816555e-01
-6.80909812e-01 4.78881001e-01 -2.49286860e-01 -1.13846950e-01
3.31304938e-01 -9.09266233e-01 -6.98878586e-01 -1.08453166e+00
8.80285084e-01 -1.99807167e-01 3.68889064e-01 -3.32971632e-01
5.70003204e-02 3.57903659e-01 -1.81555295e+00 -8.73515382e-02
9.82054472e-01 1.33009315e+00 6.96642578e-01 2.10854277e-01
-5.42694747e-01 5.34743547e-01 -5.49166381e-01 -9.02607515e-02
4.23768133e-01 -5.13698161e-01 -2.22845212e-01 2.99984187e-01
-2.31614779e-03 -6.06097579e-01 -6.35071993e-01 1.32322454e+00
2.46270448e-01 -6.11470044e-02 2.81328470e-01 -1.21001613e+00
-5.51031053e-01 2.06099793e-01 -6.15485489e-01 1.73227608e-01
-2.79677194e-03 3.29890877e-01 2.00098872e-01 -4.79056180e-01
1.78744256e-01 1.07616138e+00 3.64162564e-01 6.57382488e-01
-4.49839503e-01 -6.55034959e-01 1.09400451e-01 6.85632527e-01
-1.08233690e+00 -4.90351468e-01 9.56248581e-01 -3.40144150e-02
5.92869282e-01 8.22346658e-02 7.63580978e-01 9.24318194e-01
3.98537666e-01 5.42316556e-01 1.34912610e+00 -2.14984447e-01
2.34914139e-01 2.58956224e-01 -1.35579824e-01 6.86983705e-01
3.48404408e-01 1.15366302e-01 -1.74965307e-01 5.77192426e-01
6.06628239e-01 5.12168646e-01 -3.61241072e-01 7.42146447e-02
-6.54802918e-01 4.58866298e-01 8.17072392e-01 3.91794384e-01
-1.62139401e-01 1.33773476e-01 1.17284223e-01 4.01777551e-02
1.36976451e-01 -2.32031539e-01 -1.56849131e-01 1.05928838e-01
-7.77866244e-01 -5.68079233e-01 4.35706019e-01 8.92529607e-01
9.41230953e-01 3.79757732e-01 4.69344795e-01 6.60101116e-01
5.46302140e-01 8.83732975e-01 7.52943039e-01 -8.74630988e-01
3.65761608e-01 6.77873015e-01 -2.33705461e-01 -1.05970943e+00
-4.91625220e-01 -3.54247391e-01 -9.39332962e-01 3.36161584e-01
-6.41738325e-02 -1.93100110e-01 -1.29384124e+00 1.17165601e+00
2.21752852e-01 1.56339370e-02 2.48296171e-01 1.27903545e+00
9.93548751e-01 7.99595475e-01 -4.96326908e-02 -2.56920487e-01
1.12324142e+00 -8.94886315e-01 -1.02833688e+00 -6.16757393e-01
7.58180544e-02 -6.26650333e-01 8.23327541e-01 4.39010412e-01
-7.43502915e-01 -8.66935849e-01 -1.53083217e+00 -1.15252435e-01
-6.51072979e-01 2.48911101e-02 5.63786685e-01 9.76558447e-01
-9.66178238e-01 1.78863198e-01 -8.96743834e-01 -3.66288960e-01
4.92360681e-01 2.09286958e-01 -1.64144620e-01 -1.79232955e-01
-1.27462840e+00 9.29726005e-01 5.40028751e-01 7.56768048e-01
-8.10522199e-01 -5.23394533e-02 -5.78292012e-01 -1.40929878e-01
4.19270605e-01 -4.19707000e-01 8.35693061e-01 -1.42834830e+00
-1.78802323e+00 6.10530555e-01 -1.36895806e-01 -2.20614046e-01
5.14869869e-01 -1.93242580e-01 -4.21418071e-01 5.44673264e-01
-2.42358178e-01 2.91936874e-01 8.82468164e-01 -1.51184154e+00
-7.43334949e-01 -5.21737099e-01 -1.10622458e-01 3.49359632e-01
-5.59120297e-01 -1.19374871e-01 -6.68707967e-01 1.36496082e-01
4.11154300e-01 -8.72597098e-01 -3.13265294e-01 -1.94684893e-01
-3.61445725e-01 2.08868042e-01 1.41312993e+00 -7.22992957e-01
6.70869827e-01 -2.05420113e+00 -2.12000728e-01 -6.44779950e-02
1.58494085e-01 5.90413630e-01 3.40396136e-01 -4.40956838e-02
1.66223496e-01 3.13174017e-02 -3.89952332e-01 -6.09371625e-02
-2.18059480e-01 3.92019153e-01 -9.98705402e-02 5.59298456e-01
-7.68938214e-02 6.92529917e-01 -3.66693258e-01 -6.71491265e-01
7.00861812e-01 6.66870534e-01 -9.67276916e-02 1.39935911e-01
-2.51429141e-01 6.11346006e-01 -6.51582122e-01 4.81143683e-01
1.06847310e+00 1.46369770e-01 -1.48973301e-01 -3.46356809e-01
-3.92846435e-01 -2.58769304e-01 -7.87438571e-01 1.62576342e+00
-4.49711293e-01 5.88401139e-01 3.32391918e-01 -1.04547095e+00
1.02969408e+00 9.00722370e-02 4.52610552e-01 -1.34001637e+00
5.97520411e-01 2.89682031e-01 -3.98075461e-01 -7.44826019e-01
4.07776773e-01 -2.20978931e-01 4.26740246e-03 3.60303335e-02
-3.15759569e-01 -1.29287004e-01 -4.19196673e-02 -1.45932004e-01
7.21407294e-01 2.78451681e-01 -1.71055436e-01 3.62070322e-01
7.38535941e-01 -1.11728266e-01 7.92740107e-01 2.07216799e-01
-1.22791842e-01 5.31229973e-01 5.73379584e-02 -4.17761981e-01
-9.87312794e-01 -8.23522806e-01 3.65940034e-02 7.21073329e-01
8.35737765e-01 3.22936058e-01 -8.52717459e-01 -2.57208973e-01
-3.93559992e-01 4.87166852e-01 -1.85964987e-01 -3.78870398e-01
-6.96082413e-01 -8.79578054e-01 2.72182494e-01 3.94325078e-01
1.60593140e+00 -8.55716705e-01 -1.06980491e+00 4.16805185e-02
-1.38082370e-01 -1.47745061e+00 1.32028490e-01 6.80921301e-02
-7.76094615e-01 -1.00141919e+00 -4.09912556e-01 -7.97760427e-01
5.68313420e-01 7.21050680e-01 5.51870823e-01 1.47818953e-01
-4.46738780e-01 1.37685344e-01 -4.01974559e-01 -2.89096951e-01
-3.57768565e-01 1.36384759e-02 -2.11230397e-01 2.28167892e-01
4.26910043e-01 -5.61895251e-01 -8.69483888e-01 2.05438033e-01
-1.24577558e+00 4.12236899e-01 9.03485298e-01 3.63045424e-01
2.83584505e-01 4.22442526e-01 1.68726027e-01 -5.08511722e-01
5.13899997e-02 -2.41925552e-01 -8.80610228e-01 7.58649185e-02
-5.92800200e-01 -9.18825641e-02 8.87390852e-01 2.48801261e-01
-1.62527514e+00 1.66645765e-01 -2.44546518e-01 -2.69382983e-01
-4.78617549e-01 1.19782612e-01 -5.16034722e-01 -2.02377260e-01
1.94513053e-01 5.23612499e-01 1.04059689e-01 -4.15324569e-01
2.46593371e-01 1.02114081e+00 5.15973270e-01 2.96411645e-02
8.26362967e-01 8.18036497e-01 2.65868362e-02 -1.15424168e+00
-4.91562366e-01 -2.69237399e-01 -3.58383000e-01 -7.10164845e-01
1.24538255e+00 -8.72649908e-01 -8.90928388e-01 8.29990327e-01
-1.05916524e+00 5.73649555e-02 3.76376718e-01 3.69807661e-01
-3.07137460e-01 4.12864804e-01 -4.61890697e-01 -7.71876633e-01
-4.03328925e-01 -1.31029665e+00 6.94193065e-01 1.03704309e+00
8.45095694e-01 -6.43564880e-01 -3.70767415e-01 6.84532881e-01
5.75329840e-01 3.03178787e-01 6.50198519e-01 3.54466103e-02
-9.13500309e-01 9.28072631e-02 -5.30178845e-01 6.74126029e-01
1.58553779e-01 -9.55105294e-03 -1.23345160e+00 1.06454298e-01
5.66670358e-01 -1.31250814e-01 1.15228403e+00 2.45595738e-01
1.34093845e+00 -1.55490413e-01 -4.30214107e-01 9.89754200e-01
1.87410510e+00 6.89195275e-01 9.27969992e-01 5.37516892e-01
1.03526199e+00 4.78915185e-01 4.20477629e-01 1.90141961e-01
3.75280738e-01 3.53991151e-01 8.76094639e-01 -5.44033825e-01
-9.28252237e-04 1.18037753e-01 3.09573114e-01 8.44822884e-01
-7.82172102e-03 -1.00548968e-01 -9.14210021e-01 2.84344137e-01
-1.62601495e+00 -7.73152828e-01 -2.65999019e-01 1.97047710e+00
6.46487057e-01 1.63755536e-01 -4.92616087e-01 1.28446102e-01
7.71334350e-01 2.77409822e-01 -7.48842716e-01 -3.02919000e-01
-2.50108570e-01 1.13884687e-01 9.43099618e-01 1.19299948e-01
-9.03375387e-01 8.46511781e-01 5.17429066e+00 5.13112545e-01
-1.49344945e+00 1.80107132e-01 6.03244722e-01 2.31424328e-02
-2.46927157e-01 -1.01992808e-01 -4.74863350e-01 8.14160049e-01
8.36884558e-01 2.61437416e-01 3.53259832e-01 6.99370027e-01
4.43483114e-01 -4.64020044e-01 -3.76558840e-01 1.14391291e+00
2.51380742e-01 -7.41029024e-01 -2.88204730e-01 1.30056040e-02
7.15341628e-01 6.85848743e-02 3.37791502e-01 -1.62021473e-01
6.48749946e-03 -9.11904395e-01 5.25448442e-01 7.00589776e-01
6.26855850e-01 -9.68094170e-01 8.41279268e-01 3.16130251e-01
-1.03768218e+00 -1.96719274e-01 -6.04962766e-01 1.16874151e-01
1.88805997e-01 7.28433073e-01 -4.97060716e-01 6.23138189e-01
8.55563402e-01 7.21685290e-01 -6.27882123e-01 8.32290888e-01
-3.76208901e-01 5.21189272e-01 -3.22924763e-01 -7.20106587e-02
3.37178260e-01 -6.23065770e-01 1.09461337e-01 8.18736792e-01
3.71407598e-01 5.50479330e-02 1.22686513e-02 9.09465790e-01
9.08584967e-02 -2.52821445e-01 -5.03550708e-01 5.87862730e-02
2.23884717e-01 1.34885812e+00 -8.92569840e-01 -3.13053966e-01
-5.63701987e-01 1.18720305e+00 -2.85432279e-01 5.72980046e-01
-1.11226726e+00 -4.28259701e-01 3.24321032e-01 -2.28969485e-01
2.34543577e-01 -3.78875464e-01 -4.95981544e-01 -8.58111680e-01
4.15424258e-02 -4.59324211e-01 -2.43849128e-01 -9.29743230e-01
-6.36439741e-01 5.07381141e-01 -2.98944175e-01 -1.15111136e+00
3.98397356e-01 -8.04059803e-01 -4.25003648e-01 7.55332410e-01
-1.92329299e+00 -9.15393651e-01 -8.84488881e-01 5.87154865e-01
5.26890635e-01 -4.80879843e-03 3.70988488e-01 3.56926292e-01
-8.12224388e-01 1.14845904e-02 4.85390633e-01 2.85998620e-02
3.92328054e-01 -8.12155306e-01 5.20054922e-02 1.08252096e+00
-3.86825681e-01 2.31272340e-01 6.99917436e-01 -3.72860342e-01
-1.69930840e+00 -1.04453743e+00 1.38378426e-01 3.04064423e-01
1.87017471e-01 -2.59857327e-01 -6.39337063e-01 2.88743973e-01
4.91327345e-01 -3.88104133e-02 2.69530356e-01 -6.72462106e-01
-7.80656263e-02 -7.07633913e-01 -1.09924829e+00 3.93238246e-01
7.82784998e-01 -2.65152931e-01 -3.95948648e-01 2.06888065e-01
7.52934158e-01 -3.55605781e-01 -2.86646038e-01 3.43916595e-01
5.68530798e-01 -1.44672859e+00 6.69230282e-01 1.73169285e-01
3.69339734e-01 -4.70223606e-01 9.28119719e-02 -1.08519197e+00
2.43664473e-01 -1.69506416e-01 4.57836330e-01 1.03306592e+00
8.57362822e-02 -1.06604731e+00 8.22183788e-01 5.02506971e-01
-3.95423263e-01 -2.70626962e-01 -8.97284091e-01 -4.01256979e-01
-4.32366788e-01 -1.89905226e-01 4.53670621e-01 4.63652819e-01
-3.55733126e-01 1.54965490e-01 -1.24585256e-01 1.70547634e-01
7.57993042e-01 6.18100911e-02 4.51531887e-01 -1.06969965e+00
1.05382204e-01 -3.61764818e-01 -4.93353069e-01 -8.96241128e-01
4.91976738e-03 -4.04213369e-01 3.32509279e-01 -1.90595281e+00
5.69688082e-02 -2.82464176e-01 -1.02189876e-01 2.07509905e-01
-5.15463278e-02 4.38979477e-01 1.99062660e-01 1.40224919e-01
-6.88709497e-01 8.03027332e-01 1.26860535e+00 -1.69298202e-01
-4.41434197e-02 -2.06110045e-01 -5.06544054e-01 7.82209039e-01
1.18551683e+00 -1.58706725e-01 -4.83078361e-01 -7.31776953e-01
2.13947058e-01 -1.03530698e-01 2.77750015e-01 -1.41013467e+00
3.85311484e-01 -7.37623870e-02 6.30178332e-01 -5.71872830e-01
3.54501635e-01 -1.25660741e+00 1.87498376e-01 5.39652288e-01
1.00593343e-01 -1.61045238e-01 1.44699123e-02 6.62008405e-01
-2.76381195e-01 -9.40932333e-02 8.56346667e-01 -4.55330372e-01
-1.08183956e+00 8.34271461e-02 -3.79027873e-01 -3.33234400e-01
1.18510449e+00 -5.92425406e-01 -5.91738701e-01 -1.87853977e-01
-1.29061505e-01 1.32646024e-01 5.95164299e-01 3.99031550e-01
8.97086859e-01 -1.15379989e+00 -1.03409275e-01 2.97794104e-01
-1.90772370e-01 2.64214069e-01 5.23789883e-01 7.49740243e-01
-9.59880888e-01 4.84843761e-01 -4.12331909e-01 -5.62964261e-01
-1.05054200e+00 4.67607915e-01 4.06119734e-01 4.58058834e-01
-5.77377856e-01 5.28938413e-01 1.84112713e-01 4.43745777e-02
1.31625131e-01 -1.87374130e-01 -4.41179961e-01 -2.00837612e-01
3.57500821e-01 3.74198586e-01 1.07361801e-01 -8.64529371e-01
-2.81288981e-01 7.54366100e-01 1.02583237e-01 -3.46612707e-02
1.25993943e+00 -5.44924915e-01 -2.85127163e-01 3.46531123e-01
1.34126687e+00 -5.49518585e-01 -1.41788924e+00 -9.83999148e-02
-5.04887521e-01 -3.36051464e-01 4.27397281e-01 -5.87030709e-01
-1.66315842e+00 1.13130093e+00 9.07395542e-01 2.56577522e-01
1.49246526e+00 -4.03657913e-01 1.21548104e+00 1.25581563e-01
4.84696001e-01 -1.65718961e+00 -1.12817669e-02 2.19104081e-01
1.85120493e-01 -1.21588135e+00 -8.70791972e-02 -2.96737552e-01
-4.63716835e-01 1.22129178e+00 7.71375000e-01 1.74945310e-01
2.81759590e-01 2.08323419e-01 3.19820821e-01 -4.75531630e-02
-2.44728878e-01 -5.06551266e-01 -2.56365120e-01 5.55764139e-01
1.17936730e-02 -1.77969709e-01 -3.03620189e-01 2.20529020e-01
-9.50816274e-02 -6.00183830e-02 6.36533260e-01 9.30812955e-01
-8.97939503e-01 -5.81926882e-01 -6.72922015e-01 2.84203768e-01
-3.05906862e-01 3.04440856e-01 -1.24529891e-01 5.03953338e-01
4.83874083e-01 1.33687067e+00 -9.09078494e-02 -5.64852953e-01
1.16776213e-01 -4.27910453e-03 3.57484192e-01 -1.16825759e-01
-6.65214285e-02 -1.14291534e-01 -4.02306080e-01 -6.36034966e-01
-7.69369066e-01 -1.95268214e-01 -1.60936880e+00 -2.05685571e-01
-2.44209364e-01 -1.87844127e-01 1.34834504e+00 1.20732009e+00
-4.92133163e-02 8.23219538e-01 9.75981534e-01 -7.76096582e-01
-3.71402465e-02 -7.12517977e-01 -5.68515360e-01 1.60169244e-01
3.73991013e-01 -5.47755480e-01 -5.27016222e-01 1.17612340e-01] | [9.505574226379395, -1.4396198987960815] |
87966630-7aad-47fb-a48d-edc7af2ed163 | join-joint-gans-inversion-for-intrinsic-image | 2305.11321 | null | https://arxiv.org/abs/2305.11321v1 | https://arxiv.org/pdf/2305.11321v1.pdf | JoIN: Joint GANs Inversion for Intrinsic Image Decomposition | In this work, we propose to solve ill-posed inverse imaging problems using a bank of Generative Adversarial Networks (GAN) as a prior and apply our method to the case of Intrinsic Image Decomposition for faces and materials. Our method builds on the demonstrated success of GANs to capture complex image distributions. At the core of our approach is the idea that the latent space of a GAN is a well-suited optimization domain to solve inverse problems. Given an input image, we propose to jointly inverse the latent codes of a set of GANs and combine their outputs to reproduce the input. Contrary to most GAN inversion methods which are limited to inverting only a single GAN, we demonstrate that it is possible to maintain distribution priors while inverting several GANs jointly. We show that our approach is modular, allowing various forward imaging models, that it can successfully decompose both synthetic and real images, and provides additional advantages such as leveraging properties of GAN latent space for image relighting. | ['Julien Philip', 'Svetlana Lazebnik', 'Viraj Shah'] | 2023-05-18 | null | null | null | null | ['intrinsic-image-decomposition', 'image-relighting'] | ['computer-vision', 'computer-vision'] | [ 8.69434118e-01 5.53121567e-01 2.16151476e-01 -1.71331353e-02
-7.59137332e-01 -8.23112607e-01 7.73998439e-01 -1.06333840e+00
1.23052619e-01 7.66425133e-01 2.86180258e-01 -2.35163659e-01
-4.05236967e-02 -9.05656755e-01 -9.59912241e-01 -8.53578985e-01
5.13947368e-01 5.59674263e-01 -4.27256078e-01 -2.43925080e-01
-5.84321804e-02 6.41598642e-01 -1.18917346e+00 2.67186403e-01
7.00644553e-01 7.07906187e-01 -5.31229563e-02 8.02092433e-01
2.83723235e-01 9.77825940e-01 -4.30415899e-01 -4.08583313e-01
3.57894331e-01 -1.01261568e+00 -8.11511993e-01 3.53604704e-01
1.77936107e-01 -5.77418745e-01 -3.87268543e-01 7.32592642e-01
1.08004235e-01 -2.59155273e-01 9.56315935e-01 -1.41198456e+00
-9.27736223e-01 1.64682180e-01 -5.95480382e-01 -3.87581825e-01
3.63530606e-01 2.24558383e-01 9.16957796e-01 -6.64424658e-01
7.28106439e-01 1.03579617e+00 7.11239517e-01 7.64929473e-01
-1.72780550e+00 -4.06530857e-01 -2.77308613e-01 -2.21624658e-01
-9.84497845e-01 -7.52140403e-01 9.40508544e-01 -3.74223351e-01
4.45105433e-01 2.68093854e-01 7.08328485e-01 1.31585228e+00
8.61775279e-02 6.20635509e-01 1.39186394e+00 -6.10129058e-01
2.16859519e-01 -1.22729048e-01 -6.29208148e-01 6.09639287e-01
-6.91803321e-02 5.27402796e-02 -5.19575715e-01 -7.54191726e-02
1.13892674e+00 -2.51135472e-02 -2.95290887e-01 -5.21484971e-01
-1.23933625e+00 9.07615840e-01 3.84473473e-01 1.03794448e-01
-5.48898458e-01 5.69622695e-01 -2.17516437e-01 1.70266315e-01
4.09389675e-01 3.93825382e-01 1.58220172e-01 1.10832304e-01
-1.04425180e+00 2.79020101e-01 8.34676027e-01 7.96098590e-01
8.22756827e-01 2.57895380e-01 6.81457445e-02 5.37730515e-01
4.11987782e-01 6.51588380e-01 7.86750689e-02 -1.44312608e+00
7.02443793e-02 9.05210599e-02 9.58683789e-02 -8.33535075e-01
2.45558962e-01 -3.36291939e-01 -8.23464692e-01 5.13807178e-01
3.74262780e-01 -3.52289043e-02 -1.14959407e+00 2.08158207e+00
6.07856698e-02 1.75882638e-01 8.82395431e-02 5.70737422e-01
2.95630813e-01 6.01742208e-01 -2.53661603e-01 -2.53572799e-02
1.09922516e+00 -7.15555429e-01 -5.83264649e-01 -4.05072898e-01
1.16835743e-01 -7.84612656e-01 9.76002634e-01 3.96012694e-01
-1.51938713e+00 -3.04592699e-01 -1.10596395e+00 -1.85718521e-01
-1.56307276e-02 4.02477123e-02 5.52536011e-01 7.07893670e-01
-1.41396677e+00 4.16905910e-01 -8.91714573e-01 -2.43094534e-01
4.90036666e-01 2.37255380e-01 -3.59601974e-01 -2.57371962e-01
-8.08005571e-01 7.50143468e-01 2.62129726e-03 9.53617543e-02
-1.29049182e+00 -6.55502915e-01 -8.42259288e-01 -5.27092852e-02
-9.02507734e-03 -1.32335222e+00 8.31451893e-01 -1.20263910e+00
-1.78941476e+00 1.02027607e+00 -1.16785720e-01 -1.93539232e-01
5.69364309e-01 1.11459896e-01 1.00608100e-03 3.34640622e-01
2.11067759e-02 8.31589699e-01 1.40119374e+00 -1.64123690e+00
2.19694838e-01 -1.78315222e-01 9.76227522e-02 -1.77059676e-02
-2.10474044e-01 -3.24235857e-01 -1.63398594e-01 -8.27863038e-01
2.43985906e-01 -1.01996899e+00 -1.20550781e-01 -2.29896139e-02
-5.42911530e-01 7.70089149e-01 9.34265912e-01 -8.61592889e-01
5.49537063e-01 -1.99209559e+00 7.55596280e-01 2.36221075e-01
3.17250341e-01 -1.46054104e-01 -3.25149268e-01 6.17573440e-01
-3.05432469e-01 1.12407163e-01 -7.18522727e-01 -7.37323284e-01
1.63374379e-01 5.70448339e-01 -7.12638795e-01 5.31939149e-01
3.37969005e-01 1.30049253e+00 -5.24589181e-01 -5.66742644e-02
2.46552140e-01 9.37497854e-01 -8.37884247e-01 4.20469165e-01
-2.25344047e-01 1.10280561e+00 -2.55734622e-01 3.90427172e-01
7.76277542e-01 -2.28048563e-01 3.04989964e-01 -1.75090089e-01
2.50334889e-01 2.29724705e-01 -8.37296367e-01 1.79752517e+00
-6.75447822e-01 5.01137733e-01 5.45278072e-01 -1.21681976e+00
5.77180445e-01 3.22767437e-01 4.97915328e-01 -3.80483031e-01
-3.59069876e-04 2.51175463e-01 -2.41251171e-01 -2.01277450e-01
1.75951958e-01 -7.41550982e-01 -1.31204790e-02 9.35090065e-01
2.69387484e-01 -7.14488626e-01 -2.43488595e-01 2.29619682e-01
1.09566247e+00 5.78890979e-01 1.68201476e-02 -2.86836438e-02
3.33375454e-01 -3.08707952e-01 -2.53885854e-02 5.54396093e-01
5.14636993e-01 1.12419260e+00 6.07393682e-01 -1.91116333e-01
-1.45941281e+00 -1.35822558e+00 1.62356153e-01 5.75998604e-01
-3.32001776e-01 -4.05864492e-02 -1.13549221e+00 -4.58212256e-01
-3.03821504e-01 7.76542068e-01 -8.28411460e-01 -1.62423477e-01
-5.13087690e-01 -7.00140953e-01 6.69328690e-01 3.05780768e-01
5.27082801e-01 -1.02117765e+00 -5.06595194e-01 -6.78868964e-02
-2.80320376e-01 -1.13229096e+00 -3.90893996e-01 3.14846635e-02
-7.00589716e-01 -8.43874872e-01 -9.43460941e-01 -5.28062761e-01
8.96396875e-01 -1.10128475e-03 1.33827233e+00 5.31493127e-03
-2.52038777e-01 7.93504894e-01 -9.49194375e-03 -3.76673117e-02
-1.01034558e+00 -3.83610502e-02 -2.61490345e-01 3.69972974e-01
-4.82474327e-01 -1.20713568e+00 -6.46423519e-01 1.10542744e-01
-1.40070200e+00 3.87281090e-01 4.81064916e-01 9.15009022e-01
2.46770248e-01 1.93834230e-02 2.78325349e-01 -9.91388619e-01
4.06603694e-01 -4.01460528e-01 -3.46417576e-01 2.50891656e-01
-2.93313146e-01 1.78789511e-01 5.35519242e-01 -1.81163967e-01
-1.18598199e+00 -9.19791637e-04 -3.05836141e-01 -3.90591413e-01
-1.82994395e-01 2.29006499e-01 -3.92875254e-01 -3.18824083e-01
4.10702080e-01 3.95069867e-01 4.36992884e-01 -2.38488793e-01
7.63926685e-01 1.56875983e-01 5.84618330e-01 -7.24335790e-01
1.19876242e+00 9.20773268e-01 2.44327292e-01 -6.59781039e-01
-5.49366593e-01 3.74969870e-01 -5.93950748e-01 -1.18188404e-01
1.05068099e+00 -8.30005288e-01 -7.01846063e-01 5.62530398e-01
-1.12037158e+00 -5.34535587e-01 -5.95696747e-01 -9.61777847e-03
-1.15360773e+00 3.58619064e-01 -5.97108066e-01 -6.25179887e-01
5.96051328e-02 -1.18457723e+00 1.25547671e+00 -1.74334615e-01
-9.63005200e-02 -1.31946480e+00 -2.55574798e-03 4.62085754e-01
6.24466360e-01 6.40499830e-01 8.63219082e-01 2.54135907e-01
-8.94655943e-01 8.39263052e-02 -5.81150595e-03 4.75571454e-01
3.06739122e-01 -1.65808320e-01 -1.14370096e+00 -4.54311460e-01
5.39703429e-01 -4.60387141e-01 9.50695217e-01 3.92279625e-01
1.14377928e+00 -5.95276058e-01 -1.32509604e-01 1.17174840e+00
1.42837155e+00 -9.69892070e-02 1.19252491e+00 6.11054860e-02
7.89963305e-01 5.64788878e-01 -1.77033529e-01 1.16313070e-01
2.37616912e-01 5.78252017e-01 6.32998705e-01 -4.98661220e-01
-2.50286311e-01 -5.30017495e-01 4.00211930e-01 6.12987459e-01
-2.28234693e-01 -3.63486737e-01 -7.35991180e-01 3.81546408e-01
-1.37980318e+00 -8.84870887e-01 2.22752765e-01 1.99429548e+00
7.21751273e-01 -2.17731103e-01 -2.81873401e-02 9.72745866e-02
2.88282901e-01 2.45394304e-01 -5.32414258e-01 -1.36376932e-01
-1.34541869e-01 6.95144296e-01 4.04575914e-01 6.95734918e-01
-7.11378396e-01 6.23850048e-01 7.56717873e+00 5.57633221e-01
-8.83662164e-01 3.33360612e-01 7.87907720e-01 6.62433580e-02
-1.03311253e+00 2.42112905e-01 -9.33432579e-02 2.55646408e-01
6.91618919e-01 2.07067922e-01 1.12313318e+00 2.32547775e-01
-1.34402737e-01 9.93799865e-02 -1.11730111e+00 6.73112094e-01
2.75239319e-01 -1.37556219e+00 2.80623883e-01 4.25143003e-01
8.73006880e-01 -4.39308673e-01 5.59325695e-01 -2.10005820e-01
5.11939466e-01 -1.50008750e+00 7.45822787e-01 5.57598889e-01
1.13335955e+00 -4.68579203e-01 1.01895735e-01 3.57297570e-01
-5.91417789e-01 1.77652389e-01 1.13737270e-01 -5.75571284e-02
4.47422564e-01 4.14870203e-01 -7.24649549e-01 4.50141102e-01
3.25292170e-01 6.64805293e-01 -1.72084138e-01 1.16692908e-01
-4.75717604e-01 5.97826123e-01 -2.44822830e-01 8.19370031e-01
8.52728169e-03 -6.41631365e-01 5.22954464e-01 6.28696263e-01
5.98976314e-01 8.96372497e-02 -2.91925550e-01 1.47188818e+00
-6.77201450e-02 -4.79751438e-01 -9.56047535e-01 -2.25177765e-01
-7.53813013e-02 1.14864028e+00 -7.01116204e-01 -1.06762521e-01
-1.56411722e-01 1.37487137e+00 1.80217355e-01 7.63179958e-01
-9.61500585e-01 1.48099601e-01 4.72251713e-01 2.20509350e-01
5.28623462e-01 -4.09713656e-01 -5.41497767e-01 -1.33393168e+00
-4.47220728e-02 -1.10276604e+00 4.11096553e-05 -1.18053222e+00
-1.11695647e+00 4.76898819e-01 -1.59351707e-01 -1.07533288e+00
-5.31998396e-01 -6.01304650e-01 -5.84950387e-01 1.06877971e+00
-1.44895232e+00 -1.83701277e+00 -3.48166168e-01 6.27933264e-01
7.15887174e-02 -1.03133187e-01 9.36113834e-01 1.21227317e-02
-8.75808373e-02 3.33448350e-01 8.73031691e-02 -6.33497238e-02
3.28510642e-01 -1.17328775e+00 5.04573762e-01 9.76009011e-01
3.00812781e-01 7.72317410e-01 8.23066592e-01 -3.91542643e-01
-1.61036646e+00 -6.98169470e-01 1.69691697e-01 -4.50557083e-01
4.66930032e-01 -5.33138692e-01 -5.51819026e-01 1.19518721e+00
5.85668504e-01 -3.86130698e-02 5.77061296e-01 -3.69364768e-01
-3.92969072e-01 1.04972154e-01 -1.23952043e+00 4.63842571e-01
1.12460494e+00 -8.48053932e-01 -2.45410860e-01 3.05273920e-01
3.78857195e-01 -4.81162548e-01 -7.47036695e-01 2.24139005e-01
5.59027910e-01 -1.23702753e+00 1.38336921e+00 -3.71342689e-01
8.94323349e-01 -3.43372107e-01 -2.30380088e-01 -1.42969513e+00
-1.43506765e-01 -7.16324985e-01 -1.42688707e-01 1.23055542e+00
1.63416028e-01 -8.88411105e-01 7.73228467e-01 4.75218922e-01
4.35503237e-02 -2.47113064e-01 -7.84373701e-01 -4.71692860e-01
2.15996161e-01 -2.22752765e-01 6.36866987e-01 8.07249725e-01
-6.92183673e-01 1.38935328e-01 -7.36861885e-01 6.64164722e-02
9.08810556e-01 8.14273357e-02 8.67638886e-01 -7.14952171e-01
-9.29987252e-01 -1.61699951e-01 -6.11076206e-02 -1.10910857e+00
2.92990625e-01 -8.76619220e-01 -5.60938148e-03 -1.32007110e+00
1.82494536e-01 -4.51814830e-01 2.61013731e-02 4.10911918e-01
2.84496456e-01 9.17049170e-01 2.40012407e-01 3.10890377e-01
1.03402741e-01 7.15009689e-01 1.46877122e+00 -1.48589835e-01
1.67413890e-01 -1.79308295e-01 -9.33474720e-01 5.52728891e-01
3.72838795e-01 -4.12346691e-01 -6.46985888e-01 -7.32918501e-01
3.95741224e-01 1.57896802e-01 8.72221529e-01 -8.15761805e-01
-2.63145193e-02 -3.96771543e-02 5.14374256e-01 9.49300360e-03
6.31188333e-01 -7.53451765e-01 9.12167013e-01 2.95365721e-01
-2.07108751e-01 -1.99849322e-01 5.49581163e-02 4.57465231e-01
-2.03424051e-01 -1.33222952e-01 7.81956255e-01 -4.40298259e-01
-1.16191320e-01 1.29901886e-01 -1.48467228e-01 3.25510092e-02
8.34136665e-01 -1.85145378e-01 -9.74115636e-03 -8.06279778e-01
-8.01916420e-01 -4.47187424e-01 1.11664915e+00 -3.25514562e-02
5.27885616e-01 -1.38195860e+00 -6.04514897e-01 6.75339341e-01
-3.32191497e-01 -3.61201912e-02 1.53938293e-01 7.27143288e-01
-7.38053977e-01 5.84261566e-02 -4.33640569e-01 -6.59518480e-01
-8.28498960e-01 3.47047150e-01 5.58098316e-01 -2.44890526e-01
-4.04707551e-01 7.30547726e-01 6.39414310e-01 -4.14481223e-01
-3.80513221e-01 5.25919162e-02 3.30607384e-01 -2.97935486e-01
1.70405626e-01 -9.16091576e-02 -1.00370862e-01 -6.67355120e-01
-1.30436540e-01 7.28832722e-01 3.58647972e-01 -6.75440133e-01
1.70424807e+00 -1.29172459e-01 -5.38599491e-01 6.14943542e-02
1.12777472e+00 2.83622265e-01 -1.56064320e+00 2.09167421e-01
-6.43518269e-01 -4.95193094e-01 -5.04721142e-02 -5.34346819e-01
-1.23836482e+00 9.28667963e-01 1.08855255e-01 2.98226535e-01
1.48782611e+00 6.05922453e-02 6.62747681e-01 -2.15350538e-01
2.83804923e-01 -5.14997184e-01 3.17077190e-01 1.98979735e-01
1.10183084e+00 -7.29415655e-01 6.26940727e-02 -5.53948104e-01
-3.65236431e-01 1.00806093e+00 -2.17713583e-02 -3.58316153e-01
4.95334417e-01 5.13479650e-01 -1.34970471e-01 -2.29468375e-01
-4.22390521e-01 8.19775835e-02 2.03795925e-01 7.70037174e-01
3.15004587e-01 -1.26254261e-01 1.72598004e-01 -1.68614462e-02
-3.09408754e-01 1.79360688e-01 6.57270551e-01 8.01665962e-01
2.76636064e-01 -1.41424406e+00 -4.83995944e-01 8.67758170e-02
-4.15349692e-01 -7.30554312e-02 -2.97835350e-01 6.13046706e-01
6.67757913e-02 6.01238787e-01 -8.67219120e-02 -1.61836103e-01
-9.59915966e-02 8.08331966e-02 8.80085111e-01 -5.03351212e-01
-1.13129884e-01 1.08926184e-03 -2.51768500e-01 -4.20377612e-01
-6.89573705e-01 -7.01128185e-01 -6.28050745e-01 -2.44153157e-01
1.52255535e-01 -1.94132060e-01 6.11178339e-01 9.65302885e-01
3.74421060e-01 5.67770541e-01 6.03623748e-01 -1.07370448e+00
-3.34205359e-01 -6.60603464e-01 -5.89534283e-01 6.00400746e-01
6.26395166e-01 -5.81914544e-01 -5.72801054e-01 4.71457392e-01] | [11.699898719787598, -0.4202162027359009] |
38bdb963-3d94-4342-8822-7e8df13d4761 | csyngec-incorporating-constituent-based | 2211.08158 | null | https://arxiv.org/abs/2211.08158v1 | https://arxiv.org/pdf/2211.08158v1.pdf | CSynGEC: Incorporating Constituent-based Syntax for Grammatical Error Correction with a Tailored GEC-Oriented Parser | Recently, Zhang et al. (2022) propose a syntax-aware grammatical error correction (GEC) approach, named SynGEC, showing that incorporating tailored dependency-based syntax of the input sentence is quite beneficial to GEC. This work considers another mainstream syntax formalism, i.e., constituent-based syntax. By drawing on the successful experience of SynGEC, we first propose an extended constituent-based syntax scheme to accommodate errors in ungrammatical sentences. Then, we automatically obtain constituency trees of ungrammatical sentences to train a GEC-oriented constituency parser by using parallel GEC data as a pivot. For syntax encoding, we employ the graph convolutional network (GCN). Experimental results show that our method, named CSynGEC, yields substantial improvements over strong baselines. Moreover, we investigate the integration of constituent-based and dependency-based syntax for GEC in two ways: 1) intra-model combination, which means using separate GCNs to encode both kinds of syntax for decoding in a single model; 2)inter-model combination, which means gathering and selecting edits predicted by different models to achieve final corrections. We find that the former method improves recall over using one standalone syntax formalism while the latter improves precision, and both lead to better F0.5 values. | ['Zhenghua Li', 'Yue Zhang'] | 2022-11-15 | null | null | null | null | ['grammatical-error-correction'] | ['natural-language-processing'] | [ 4.30053890e-01 3.45060885e-01 4.35614765e-01 -5.73963165e-01
-7.15172291e-01 -5.09387493e-01 3.01872969e-01 4.87839371e-01
-4.97095942e-01 4.58065689e-01 3.62397730e-01 -6.78465009e-01
2.33742639e-01 -9.35602546e-01 -9.43921626e-01 -2.35875934e-01
3.34013104e-01 2.92702705e-01 2.24132225e-01 -5.94870508e-01
4.15005833e-01 9.60537270e-02 -1.18155968e+00 3.59981120e-01
1.56596458e+00 3.55979383e-01 4.98811603e-01 5.98904788e-01
-4.57837552e-01 5.32141566e-01 -6.28534138e-01 -1.02752733e+00
1.86118372e-02 -6.61300957e-01 -7.18826234e-01 -3.71129453e-01
4.39029694e-01 1.63503498e-01 3.69087309e-01 1.26862371e+00
3.29875886e-01 -2.84076910e-02 2.55848527e-01 -5.49572468e-01
-8.08288395e-01 1.17587745e+00 -7.76054263e-02 -3.27022746e-02
3.48785162e-01 1.93391174e-01 1.29729807e+00 -7.93275416e-01
6.84703887e-01 1.32809484e+00 7.29467869e-01 7.78579593e-01
-1.14616907e+00 -2.87679791e-01 3.56601536e-01 -2.27106661e-02
-1.00049102e+00 -3.82617533e-01 7.28909314e-01 -7.22365230e-02
1.67192078e+00 2.67439783e-01 4.21045482e-01 8.52743566e-01
2.68698424e-01 5.67882478e-01 9.88950610e-01 -9.43326831e-01
7.45259374e-02 -1.93166122e-01 4.68631804e-01 7.81863630e-01
3.80859107e-01 1.86911464e-01 -1.30942121e-01 1.94450229e-01
1.22216947e-01 -5.49503446e-01 -2.79904693e-01 2.08747610e-01
-9.98416483e-01 8.35929751e-01 3.08202237e-01 4.67955321e-01
-2.03933820e-01 3.26210380e-01 6.26586020e-01 4.61672246e-01
3.29846263e-01 6.38646543e-01 -5.95045984e-01 -1.95003390e-01
-7.03374863e-01 3.97451133e-01 7.12538600e-01 1.12996304e+00
5.86799979e-01 6.02659397e-02 -1.47262409e-01 9.38671291e-01
4.11635995e-01 5.35461128e-01 5.24548411e-01 -7.05114424e-01
1.11900342e+00 7.89516449e-01 -3.50988299e-01 -9.22798455e-01
-5.32384157e-01 -3.42304528e-01 -6.40926480e-01 -1.83952138e-01
2.77221799e-01 -9.65128839e-02 -9.03780997e-01 2.08119965e+00
9.41573083e-02 -4.54181343e-01 1.13408469e-01 5.91063976e-01
8.76192987e-01 6.52517736e-01 4.19824451e-01 -8.15480351e-02
1.15609431e+00 -9.17994738e-01 -7.37018824e-01 -3.34640682e-01
1.35906160e+00 -6.62191331e-01 1.25880814e+00 2.63784498e-01
-1.21771860e+00 -5.28779984e-01 -1.07040954e+00 -3.85577649e-01
-4.40011352e-01 2.25255832e-01 4.23471093e-01 7.82137573e-01
-1.24075842e+00 9.46285546e-01 -7.83163786e-01 -4.51007605e-01
-1.29682183e-01 1.38872132e-01 -3.14131767e-01 -3.19023281e-02
-1.27668655e+00 1.16740382e+00 8.70852888e-01 7.25066140e-02
-1.89724132e-01 -5.19214094e-01 -1.15081716e+00 4.13610674e-02
1.88919827e-01 -6.16637170e-01 1.12583280e+00 -9.39360380e-01
-1.23840785e+00 7.62162745e-01 -3.23999316e-01 -4.53712255e-01
5.50611801e-02 -1.60588384e-01 -3.51761371e-01 -3.18458229e-01
-9.24759582e-02 5.29652655e-01 3.13514978e-01 -1.02870691e+00
-6.04117513e-01 -3.21448594e-01 5.30157499e-02 2.63213336e-01
2.07116120e-02 4.18527365e-01 -4.08362806e-01 -8.54399741e-01
2.28163555e-01 -8.33975792e-01 -8.10519531e-02 -6.46097541e-01
-4.51014549e-01 -5.19736946e-01 2.47290716e-01 -1.23907149e+00
1.90317595e+00 -1.82177877e+00 4.15576607e-01 2.94357568e-01
5.37871681e-02 6.80996001e-01 -4.25505340e-01 5.31944931e-01
-1.39695734e-01 5.12156904e-01 -6.45428538e-01 -7.51813054e-01
2.62326095e-02 3.84955943e-01 9.83125716e-02 -1.87252447e-01
5.36446333e-01 1.19030809e+00 -9.58822906e-01 -2.86240488e-01
-2.66220793e-02 6.34784997e-02 -9.13766623e-01 -5.60620204e-02
-2.97082067e-01 2.62529701e-01 7.56757101e-03 5.25129199e-01
7.37565160e-01 1.14719033e-01 5.68331599e-01 7.48125091e-02
-1.03572667e-01 1.10710859e+00 -9.03906465e-01 1.66259658e+00
-6.72930479e-01 1.06028117e-01 -1.84060156e-01 -9.37454045e-01
1.05932319e+00 9.20065418e-02 -4.22839731e-01 -8.53101552e-01
1.17705800e-01 6.37053013e-01 3.49046588e-01 -1.49986997e-01
9.09405708e-01 -4.57055459e-04 -4.57338721e-01 3.46167982e-01
3.39438468e-01 -1.91517547e-01 4.83515888e-01 4.24009651e-01
1.23744047e+00 2.65430689e-01 5.28351605e-01 -2.40099773e-01
6.42748952e-01 -1.95054784e-01 8.62109542e-01 7.91009665e-01
-9.29892138e-02 6.61430120e-01 4.53990579e-01 -2.03908592e-01
-8.62699986e-01 -8.12427282e-01 2.37502173e-01 8.72941017e-01
-2.44721249e-01 -9.72093403e-01 -9.41905677e-01 -1.13550484e+00
-2.06251591e-01 1.22241116e+00 -3.30308646e-01 -2.15511724e-01
-1.27698374e+00 -7.71651208e-01 6.64227009e-01 6.23227775e-01
2.91949004e-01 -1.27571166e+00 -3.62730145e-01 4.69085366e-01
-4.63314652e-01 -8.65354955e-01 -5.07818639e-01 2.72194177e-01
-8.70625794e-01 -9.72642601e-01 -1.10030428e-01 -6.15819514e-01
5.34204304e-01 -1.37460157e-01 1.35128796e+00 7.25170195e-01
2.94520527e-01 -6.47225138e-03 -9.14234161e-01 -2.95955598e-01
-1.01355457e+00 2.02666625e-01 -2.43886828e-01 -5.21191418e-01
3.53025645e-01 -5.93688488e-01 4.23462242e-02 -3.25436145e-01
-6.23307824e-01 2.46055976e-01 4.49689806e-01 6.76073134e-01
4.18287605e-01 -4.50364172e-01 4.31990385e-01 -1.42576993e+00
5.94145238e-01 -1.32652000e-01 -5.87819457e-01 5.98776460e-01
-9.16641295e-01 1.99436650e-01 8.36049616e-01 9.25339535e-02
-1.10110044e+00 -1.90743450e-02 -8.13079238e-01 3.45137626e-01
9.31730145e-04 7.20356166e-01 -5.44669986e-01 4.36532088e-02
4.56446260e-01 1.83959693e-01 -2.57379621e-01 -5.73582292e-01
5.03418148e-01 5.85808814e-01 4.89452749e-01 -7.89466381e-01
5.35008907e-01 -4.76027489e-01 -2.33580485e-01 -3.20300102e-01
-6.68644190e-01 1.80379924e-04 -7.57014632e-01 2.31341794e-02
8.15186799e-01 -5.87553024e-01 -1.24565661e-01 5.00236094e-01
-1.77692759e+00 -2.42596060e-01 -8.05445686e-02 2.38380909e-01
-3.21361691e-01 7.00868666e-01 -7.49357104e-01 -4.68304485e-01
-6.04771554e-01 -1.23962128e+00 1.00998235e+00 -1.79490987e-02
-2.76388973e-01 -1.08670795e+00 9.61309895e-02 1.85603723e-01
4.15179044e-01 1.25456214e-01 1.42449522e+00 -8.88790011e-01
-4.50936019e-01 2.96979491e-02 -2.44348809e-01 6.34205878e-01
-4.17157821e-02 -3.87843139e-02 -4.96463120e-01 -1.78002283e-01
-2.78777510e-01 -2.68475860e-02 9.14875507e-01 2.98092347e-02
8.30723643e-01 -2.91535109e-01 -4.08025086e-02 6.80712938e-01
1.47902715e+00 1.79757923e-01 8.74082267e-01 2.71424860e-01
8.58659685e-01 5.76179504e-01 4.32158828e-01 -8.30566064e-02
8.78038108e-01 6.26150191e-01 3.93543452e-01 2.60404587e-01
-4.30273354e-01 -5.23478329e-01 6.17454469e-01 1.50815761e+00
-5.18376529e-02 -5.60232937e-01 -9.63270903e-01 5.22951126e-01
-1.69974959e+00 -4.63710994e-01 -6.98833942e-01 2.11434436e+00
1.01479757e+00 1.31607592e-01 -3.35756749e-01 1.39543474e-01
7.48440742e-01 -1.30717173e-01 1.34552509e-01 -1.08017969e+00
-3.34331751e-01 6.63309574e-01 4.95410770e-01 5.91238320e-01
-9.93787229e-01 1.40662682e+00 5.58329535e+00 7.22667754e-01
-8.44341457e-01 2.23054349e-01 2.08126411e-01 3.36741656e-01
-7.17054725e-01 4.11949426e-01 -1.13676047e+00 7.34317541e-01
1.26966822e+00 1.40667096e-01 4.13362712e-01 4.28517401e-01
8.04181024e-02 -7.32366666e-02 -9.52685058e-01 6.49567127e-01
1.79954723e-01 -1.12412703e+00 2.73335427e-01 -2.32080013e-01
4.84886378e-01 -7.52610564e-02 -6.30296767e-01 6.83070779e-01
5.79056203e-01 -5.68287969e-01 8.75635326e-01 5.18754244e-01
5.56040764e-01 -6.26361787e-01 9.63311553e-01 3.39082092e-01
-1.13359630e+00 1.16605692e-01 -4.58088577e-01 -1.50758460e-01
4.47610319e-01 6.59883976e-01 -3.05021107e-01 9.79632139e-01
4.70581889e-01 6.57878935e-01 -1.16079378e+00 8.13055992e-01
-8.61718357e-01 1.00005484e+00 -1.57699257e-01 -1.30279809e-01
7.47877583e-02 -2.81134844e-01 7.02529252e-01 1.53191221e+00
4.12847638e-01 4.08574603e-02 -1.30153835e-01 1.01695287e+00
-1.75113127e-01 3.94180626e-01 -4.09908265e-01 1.00195795e-01
7.02483594e-01 1.05437922e+00 -4.54575121e-01 -3.12798321e-01
-3.56378108e-01 1.17929149e+00 9.08157527e-01 -2.34958846e-02
-7.87096083e-01 -6.73956394e-01 1.80500612e-01 -2.48941675e-01
3.53644490e-01 -2.65374839e-01 -5.31467736e-01 -1.22060847e+00
2.96012014e-01 -1.01201999e+00 2.96437621e-01 -5.97357929e-01
-1.03290641e+00 6.04494870e-01 -1.34127602e-01 -7.94923306e-01
-1.38870344e-01 -6.04495168e-01 -8.80032122e-01 1.14224362e+00
-1.55742669e+00 -1.25198340e+00 1.05707474e-01 9.03750509e-02
3.38603318e-01 1.52010247e-01 8.56158018e-01 4.07958299e-01
-6.83071315e-01 1.02616310e+00 -1.91482663e-01 2.32170552e-01
6.14654839e-01 -1.58775723e+00 1.11803973e+00 1.53532743e+00
1.31980374e-01 8.94209743e-01 3.30502212e-01 -1.03982103e+00
-1.07094181e+00 -1.53787148e+00 1.93236995e+00 -5.81193089e-01
3.34545285e-01 -3.64944696e-01 -1.13064468e+00 7.79187858e-01
4.22294177e-02 -3.44974339e-01 3.82248223e-01 2.45754346e-01
-3.72770518e-01 1.65959224e-01 -9.19666946e-01 5.29714525e-01
1.25220740e+00 -2.74273813e-01 -1.01790762e+00 -4.93963324e-02
1.24667668e+00 -5.07447064e-01 -5.84354401e-01 5.00700653e-01
6.31907061e-02 -8.15477073e-01 4.09416795e-01 -6.10565722e-01
6.84133708e-01 -3.41437519e-01 -3.81443888e-01 -1.62769055e+00
-4.18993801e-01 -4.20527428e-01 -6.33567274e-02 1.61577427e+00
8.01931977e-01 -6.35403991e-01 2.10159779e-01 4.69618946e-01
-8.48239422e-01 -5.15813291e-01 -8.41861904e-01 -9.33960140e-01
4.58135694e-01 -7.70660758e-01 8.19363713e-01 7.44644582e-01
4.29316238e-02 3.16020489e-01 -1.16708666e-01 3.80670577e-02
2.08068743e-01 -2.48992667e-01 5.07664979e-01 -1.12630558e+00
-3.46040457e-01 -4.15470868e-01 -1.92171752e-01 -9.02885139e-01
2.95765489e-01 -1.49668789e+00 4.00925785e-01 -1.54096282e+00
-7.59309009e-02 -6.61940098e-01 -8.11569542e-02 6.98939264e-01
-6.70728028e-01 -7.28972554e-02 5.89592278e-01 -1.53203800e-01
-4.61547971e-01 4.47694510e-01 7.56089747e-01 1.63444266e-01
-2.84920096e-01 -2.92488486e-01 -9.12985265e-01 5.75931013e-01
9.23701763e-01 -5.59313238e-01 3.68497744e-02 -8.81060302e-01
5.71424782e-01 -1.26843929e-01 2.03606755e-01 -8.41867030e-01
2.64265034e-02 8.61849040e-02 -2.10893348e-01 -4.04940426e-01
-3.40969563e-01 -3.71286452e-01 -1.55310124e-01 3.33215803e-01
-2.68999219e-01 4.20359939e-01 1.85561657e-01 3.56500804e-01
-8.06832761e-02 -7.66594470e-01 6.32869959e-01 -2.78247416e-01
-5.61273754e-01 -2.84459651e-01 -2.57845879e-01 2.40074039e-01
3.24189544e-01 -2.37567592e-02 -4.18121278e-01 -4.52296510e-02
-5.67825913e-01 1.52886003e-01 4.08176810e-01 3.66630644e-01
4.14051980e-01 -1.14325964e+00 -7.75485873e-01 2.52105743e-01
2.56687582e-01 -1.74181342e-01 -7.82011356e-03 8.46203029e-01
-5.05685508e-01 2.79124200e-01 1.72925845e-01 -2.21533999e-01
-1.25640357e+00 3.39274019e-01 2.13451818e-01 -6.39483213e-01
-4.93675888e-01 1.05433381e+00 -1.44767612e-01 -1.20079327e+00
-1.77053567e-02 -7.96639085e-01 -9.12237540e-02 -3.97541046e-01
4.11637306e-01 1.31006777e-01 7.14617848e-01 -7.05493212e-01
-3.42911303e-01 4.73102808e-01 -2.06920832e-01 1.89047426e-01
1.07668245e+00 -2.62967288e-01 -4.81114060e-01 1.38543099e-01
9.48588550e-01 1.98626325e-01 -6.65232956e-01 -1.55011103e-01
4.89773512e-01 -1.19579360e-01 -8.77901912e-02 -1.32604969e+00
-8.35122764e-01 8.93930376e-01 1.26990646e-01 -1.58629999e-01
1.22565317e+00 -2.00804263e-01 9.88966465e-01 4.09397840e-01
4.28082794e-01 -1.04902971e+00 -5.54263711e-01 1.20914137e+00
8.04229558e-01 -1.12371862e+00 -4.82876033e-01 -6.21820211e-01
-6.70344174e-01 1.18303764e+00 7.05532610e-01 -1.13192335e-01
3.41390789e-01 -1.67515110e-02 -2.20651239e-01 -1.15145221e-01
-8.01207125e-01 -3.25439423e-01 4.01364893e-01 5.28022468e-01
7.67034709e-01 2.56701022e-01 -1.10871398e+00 1.04063737e+00
-5.10768294e-01 -3.08786541e-01 6.28976464e-01 9.08592761e-01
-4.24475133e-01 -1.72150016e+00 -1.82911679e-02 4.98596311e-01
-3.64529103e-01 -7.39362299e-01 -4.28398103e-01 7.72204340e-01
4.09363151e-01 9.86272037e-01 -1.03119701e-01 -4.42775816e-01
6.96445644e-01 3.97908449e-01 5.90091467e-01 -1.08890545e+00
-1.15927541e+00 -2.23926708e-01 4.21284437e-01 -6.47338629e-01
-1.42557546e-01 -6.25702441e-01 -1.46374619e+00 -2.65772581e-01
-4.24196452e-01 5.98305576e-02 7.18493521e-01 1.13356566e+00
4.97452736e-01 5.58236122e-01 3.23916107e-01 -4.65322673e-01
-5.35018981e-01 -1.16901743e+00 -2.69199640e-01 4.56858397e-01
9.81714670e-03 -2.60101259e-01 -3.68618160e-01 -2.51579322e-02] | [11.087369918823242, 10.708354949951172] |
27932171-f025-402b-8ec9-0bdfaea561aa | neonatal-bowel-sound-detection-using | 2108.07467 | null | https://arxiv.org/abs/2108.07467v3 | https://arxiv.org/pdf/2108.07467v3.pdf | Neonatal Bowel Sound Detection Using Convolutional Neural Network and Laplace Hidden Semi-Markov Model | Abdominal auscultation is a convenient, safe and inexpensive method to assess bowel conditions, which is essential in neonatal care. It helps early detection of neonatal bowel dysfunctions and allows timely intervention. This paper presents a neonatal bowel sound detection method to assist the auscultation. Specifically, a Convolutional Neural Network (CNN) is proposed to classify peristalsis and non-peristalsis sounds. The classification is then optimized using a Laplace Hidden Semi-Markov Model (HSMM). The proposed method is validated on abdominal sounds from 49 newborn infants admitted to our tertiary Neonatal Intensive Care Unit (NICU). The results show that the method can effectively detect bowel sounds with accuracy and area under curve (AUC) score being 89.81% and 83.96% respectively, outperforming 13 baseline methods. Furthermore, the proposed Laplace HSMM refinement strategy is proven capable to enhance other bowel sound detection models. The outcomes of this work have the potential to facilitate future telehealth applications for neonatal care. The source code of our work can be found at: https://bitbucket.org/chirudeakin/neonatal-bowel-sound-classification/ | ['Faezeh Marzbanrad', 'Alistair McEwan', 'Anusha Withana', 'Murray Hinder', 'Omid Kavehei', 'Mark Tracy', 'Archana Priyadarshi', 'Jinyuan He', 'Chiranjibi Sitaula'] | 2021-08-17 | null | null | null | null | ['sound-classification'] | ['audio'] | [ 4.43974026e-02 1.90394089e-01 -1.51498362e-01 -1.44537345e-01
-4.72669721e-01 -3.56183469e-01 -2.05406219e-01 6.48970842e-01
-3.79850864e-01 2.56256402e-01 1.79816619e-01 -8.13453615e-01
8.86301622e-02 -6.14713132e-01 -8.47168863e-01 -6.14893019e-01
-5.55117905e-01 1.64036065e-01 2.62362778e-01 3.43238950e-01
-1.02139793e-01 1.82580888e-01 -9.65492189e-01 5.61056018e-01
8.34563076e-01 1.07242811e+00 3.63856256e-01 1.42849851e+00
4.74691354e-02 1.13894558e+00 -2.98969567e-01 1.41147962e-02
4.16917503e-02 -7.88329363e-01 -7.54785061e-01 -1.05472720e+00
2.46715337e-01 -7.44145513e-01 -1.97882608e-01 8.36298287e-01
1.00434148e+00 -9.66800600e-02 3.75205010e-01 -5.80347896e-01
-1.01283811e-01 7.17749357e-01 -2.04291075e-01 7.65647292e-01
4.20179218e-01 -3.83387953e-01 4.00367975e-01 -5.30023158e-01
-4.34038974e-02 5.19731283e-01 1.32991815e+00 6.76328003e-01
-4.29102123e-01 -8.51476252e-01 -5.70294499e-01 8.24346468e-02
-8.72713506e-01 -2.80717164e-01 3.28741193e-01 -4.39269364e-01
9.22949791e-01 3.09319586e-01 1.20089328e+00 7.15602458e-01
3.42165858e-01 7.54300594e-01 5.68114996e-01 -5.75003564e-01
1.45002911e-02 -2.16941923e-01 -1.60214767e-01 1.11112928e+00
4.57455546e-01 -1.24657065e-01 -2.25984558e-01 -1.83544531e-01
8.90136838e-01 4.12486076e-01 -4.32466865e-01 5.58260903e-02
-1.02637625e+00 7.93728232e-01 3.66517752e-01 4.80639815e-01
-7.37674475e-01 3.39447975e-01 4.10932571e-01 8.47926065e-02
4.08366710e-01 4.49191511e-01 -3.37322891e-01 -6.18331611e-01
-8.81053805e-01 -1.31483361e-01 1.15958393e+00 8.32694411e-01
6.13025874e-02 -1.75724044e-01 -1.71360090e-01 7.47847557e-01
5.36262631e-01 4.05366272e-01 4.73697394e-01 -9.07973766e-01
5.00886858e-01 1.71738505e-01 -7.64439255e-02 -7.72908270e-01
-9.01153028e-01 -2.90032595e-01 -1.05012882e+00 -4.43032682e-01
-6.28760681e-02 -7.15667844e-01 -6.40403628e-01 1.31794119e+00
4.65183884e-01 4.88304883e-01 1.13356806e-01 7.09258378e-01
1.68160510e+00 4.01918530e-01 1.77330732e-01 -3.09517056e-01
1.18614125e+00 -9.52316105e-01 -8.63302469e-01 1.71088085e-01
9.33414757e-01 -8.59403431e-01 1.81656137e-01 2.80664474e-01
-1.30679357e+00 -1.62771076e-01 -9.62062776e-01 3.17687064e-01
2.99032360e-01 1.70585752e-01 5.90225577e-01 7.03190148e-01
-1.10354686e+00 7.48005450e-01 -1.64972091e+00 -4.08900112e-01
6.21647954e-01 3.56629640e-01 -7.90165961e-02 1.35293722e-01
-1.21456337e+00 8.06220531e-01 3.27966750e-01 -5.92085011e-02
-3.01041394e-01 -8.39229643e-01 -8.97383511e-01 1.02256067e-01
-2.66421527e-01 -6.91188276e-01 1.78664482e+00 -5.13940036e-01
-1.46059442e+00 5.22098958e-01 7.73396157e-03 -7.04975665e-01
5.32139122e-01 -4.95620459e-01 -1.78201064e-01 8.03067148e-01
-7.84272850e-02 3.27908486e-01 3.94141674e-01 -5.65549374e-01
-8.17051888e-01 2.02339366e-01 -3.06645006e-01 9.33056995e-02
-2.98360467e-01 4.39879179e-01 -2.06767246e-01 -5.62182605e-01
5.98243177e-01 -6.95158601e-01 -8.61472189e-02 7.78421585e-04
-9.34784114e-02 2.44967654e-01 1.12028390e-01 -1.14672995e+00
1.47842157e+00 -1.78744090e+00 -7.23230779e-01 1.02039594e-02
3.85140181e-01 5.81070006e-01 3.20683420e-01 5.23339689e-01
-1.39690578e-01 3.45554985e-02 -2.38300189e-02 -1.46447554e-01
-4.88994539e-01 -1.01000369e-01 5.95665038e-01 6.83295250e-01
-2.56619193e-02 1.02898097e+00 -1.17589736e+00 -4.67380643e-01
5.52545190e-01 6.76175416e-01 -5.92559159e-01 6.62183106e-01
3.99699867e-01 5.31580269e-01 -4.33101386e-01 5.44453621e-01
4.07339424e-01 -3.39311361e-01 -2.73590609e-02 3.78587167e-03
-2.06688896e-01 5.45673907e-01 -8.11420083e-01 1.47498918e+00
-2.64195204e-01 5.49337327e-01 9.51243751e-03 -9.05499578e-01
5.59383988e-01 9.24065053e-01 6.37095451e-01 -5.76953769e-01
5.32441735e-01 4.66462225e-01 2.56584644e-01 -1.41652966e+00
-4.73793268e-01 -1.23888828e-01 8.13760221e-01 4.76526618e-01
9.41188168e-03 7.81776756e-03 9.08725895e-03 -1.91160530e-01
1.24308002e+00 -9.24878046e-02 7.95275807e-01 -3.34037930e-01
1.29501805e-01 -3.46908242e-01 2.89423078e-01 9.43615377e-01
-2.58669794e-01 1.04272640e+00 1.30742088e-01 -6.58559799e-01
-5.21778405e-01 -5.48478425e-01 -2.80034393e-01 8.17307115e-01
-1.90590441e-01 6.16999008e-02 -7.99927115e-01 -3.31734508e-01
-1.05096407e-01 2.67757714e-01 -7.10005343e-01 6.06235005e-02
-7.25823879e-01 -5.20841956e-01 8.09254050e-01 8.99411142e-01
2.46800452e-01 -1.13311410e+00 -1.31758976e+00 5.19600809e-01
-2.35749722e-01 -6.50612772e-01 -2.84339964e-01 2.73572654e-01
-9.18914855e-01 -1.23118305e+00 -1.40108752e+00 -1.10535932e+00
5.06436348e-01 4.42304574e-02 6.54999316e-01 2.85124213e-01
-4.98174816e-01 3.00679088e-01 -7.24353552e-01 -8.17871988e-01
-9.99697268e-01 -1.19198859e-02 -4.56138290e-02 -3.76214534e-01
1.54863030e-01 -5.81403196e-01 -1.48885751e+00 -4.44362730e-01
-5.75882554e-01 2.58541822e-01 6.36314571e-01 6.27303004e-01
2.77958184e-01 -7.31616974e-01 3.77692670e-01 -7.30848014e-01
5.28774500e-01 -9.06930327e-01 -2.53358722e-01 -3.10548432e-02
-6.61190629e-01 -4.96155262e-01 4.54553932e-01 -5.52304566e-01
-4.85711902e-01 -3.65155518e-01 -8.17772150e-01 -3.32213014e-01
-4.37723011e-01 4.83816147e-01 9.51046050e-01 -8.19229111e-02
4.97283459e-01 9.84942019e-02 5.93969114e-02 -7.27195799e-01
-4.53197539e-01 8.59427273e-01 5.45231819e-01 6.62051328e-03
7.28187412e-02 1.00807466e-01 -4.14523259e-02 -5.71794808e-01
-6.07283771e-01 -9.93890762e-01 -2.80567676e-01 -1.27949208e-01
1.16142392e+00 -7.96743035e-01 -7.24308431e-01 4.53170598e-01
-1.21220124e+00 -4.18249369e-01 6.43443167e-02 1.28836524e+00
-3.41034651e-01 4.49316889e-01 -1.10586894e+00 -9.19381440e-01
-1.12400925e+00 -7.33741701e-01 5.31614363e-01 4.43398714e-01
-4.74942356e-01 -1.22388446e+00 3.82443845e-01 1.07325926e-01
8.42097998e-01 6.30771339e-01 9.35413063e-01 -1.05558467e+00
1.05363056e-02 -5.40381789e-01 -1.10832594e-01 4.59186137e-01
3.48165065e-01 5.58947213e-02 -8.91578794e-01 -3.59772950e-01
1.72137678e-01 -7.33364075e-02 6.69336379e-01 9.34630752e-01
8.80595446e-01 -4.23829556e-01 2.42882278e-02 1.05964136e+00
1.25636077e+00 7.38759458e-01 1.30401149e-01 3.59685391e-01
5.24544597e-01 2.92687118e-01 2.81522274e-01 7.61337638e-01
5.08397758e-01 -2.74072051e-01 6.15571737e-01 -5.48090518e-01
-1.87169611e-01 -1.98326945e-01 -3.25348347e-01 1.28873944e+00
-3.70817721e-01 -2.01802030e-01 -1.41584575e+00 7.21555173e-01
-1.45271730e+00 -6.76787496e-01 -3.72617126e-01 1.74331188e+00
8.39783370e-01 -3.65668178e-01 6.60791770e-02 2.24365622e-01
6.20485723e-01 -2.08491459e-01 -1.05034262e-01 -6.01363182e-01
5.47955096e-01 7.50299513e-01 3.87125969e-01 3.32971632e-01
-1.30313909e+00 2.25062206e-01 5.67047405e+00 4.51927334e-02
-1.26356113e+00 2.91136086e-01 4.21864986e-01 1.94711000e-01
1.57590806e-01 -5.76631188e-01 -5.00801690e-02 3.89889628e-01
8.18246067e-01 3.20549160e-01 1.12508543e-01 6.15553081e-01
1.94300398e-01 -2.20930845e-01 -9.86685216e-01 8.48401785e-01
-1.28040165e-01 -1.50227058e+00 -5.71854413e-01 -8.28085959e-01
7.72407055e-01 5.16606629e-01 -1.77638605e-01 -1.17905386e-01
-3.63762051e-01 -9.32647347e-01 4.65645969e-01 5.44892490e-01
1.24658358e+00 -4.85916644e-01 1.12497401e+00 6.40040636e-01
-1.03893507e+00 4.05443870e-02 1.59522280e-01 -1.24582849e-01
-1.88953578e-01 -4.08154130e-02 -1.40360844e+00 1.18405052e-01
1.04503787e+00 3.50103140e-01 2.69280374e-02 1.75374353e+00
-9.30935889e-02 1.13385689e+00 -4.17537510e-01 -4.80936050e-01
9.59933475e-02 3.29946399e-01 3.38987887e-01 1.74254155e+00
7.53917038e-01 4.89296079e-01 -3.99084568e-01 5.67249537e-01
2.21990034e-01 5.33801913e-01 -3.81593406e-01 -3.38004343e-02
3.89987797e-01 8.60782504e-01 -7.65799880e-01 -2.35482901e-01
-6.71340466e-01 5.19669294e-01 -2.52537429e-01 2.74945647e-01
-5.96384346e-01 -5.78748107e-01 -4.01208736e-02 2.03399107e-01
1.76320940e-01 5.86730480e-01 -2.11527541e-01 -7.44842887e-01
-1.94507837e-01 -7.11205721e-01 3.86567146e-01 -4.55701441e-01
-3.19416493e-01 3.56486380e-01 -5.47942258e-02 -1.11275899e+00
-3.34953010e-01 -2.77996123e-01 -9.47860658e-01 7.65755415e-01
-1.71383393e+00 -6.46829665e-01 -5.09445786e-01 -1.18391926e-03
4.31872994e-01 1.92136485e-02 1.12881482e+00 3.88423145e-01
-3.19091231e-01 7.64968157e-01 1.74802512e-01 3.41964394e-01
2.30204090e-01 -1.15122521e+00 3.64720613e-01 8.10676754e-01
-4.71090555e-01 8.48603725e-01 6.10478520e-01 -6.86516643e-01
-1.29028583e+00 -1.23919308e+00 9.69724238e-01 1.14269949e-01
4.07341331e-01 2.36301541e-01 -9.21937108e-01 4.62861985e-01
1.68812722e-01 2.05361217e-01 9.82981265e-01 -7.46988416e-01
3.11560065e-01 2.26060599e-01 -1.12423134e+00 1.24347590e-01
5.28320789e-01 -8.92619416e-03 -4.35787857e-01 2.81482100e-01
6.12776875e-01 -1.03035212e+00 -1.17995405e+00 1.03416348e+00
9.17352080e-01 -1.00121891e+00 6.32303059e-01 -3.36378127e-01
6.41050339e-01 3.95321608e-01 3.41347426e-01 -7.88321316e-01
-3.03662062e-01 -1.14561808e+00 -7.44815767e-01 6.07780695e-01
3.29346746e-01 -9.48587775e-01 6.50999308e-01 3.04393977e-01
-3.05563390e-01 -1.38301480e+00 -6.67253077e-01 -1.15811341e-01
3.68045866e-02 -5.75910866e-01 4.62895989e-01 8.36631715e-01
1.95398510e-01 -4.12983567e-01 1.10447351e-02 3.42391342e-01
1.68771043e-01 -9.43512917e-02 4.07520011e-02 -8.28926563e-01
-4.13195908e-01 -4.18182373e-01 -1.71367645e-01 -8.11314344e-01
-8.02135348e-01 -6.89550936e-01 2.47958153e-01 -2.01881504e+00
2.53433377e-01 -4.40301627e-01 -7.24554777e-01 4.66252804e-01
-3.93399864e-01 1.75407544e-01 -9.94655266e-02 -9.74866375e-02
-5.41776478e-01 -2.00875685e-01 1.28981614e+00 4.96654183e-01
-4.12958771e-01 8.14138651e-01 -3.50108385e-01 9.54831898e-01
1.15165317e+00 -7.47533202e-01 -1.65575787e-01 -4.06485885e-01
1.51089862e-01 8.57293487e-01 4.43485603e-02 -9.75641668e-01
2.65102863e-01 9.61238518e-02 3.96123827e-01 -5.66014886e-01
-2.13007763e-01 -3.99968058e-01 -8.28361437e-02 1.20537615e+00
-3.78809750e-01 9.70527437e-03 2.35551775e-01 1.25478832e-02
-1.35309532e-01 -3.35936964e-01 7.96609998e-01 -1.65563226e-01
-1.50264055e-01 4.98051018e-01 -6.77157879e-01 1.45535141e-01
7.49733329e-01 -9.05358046e-02 2.65241452e-02 -3.43823463e-01
-5.27118206e-01 9.42350551e-02 -9.30294096e-02 2.00180635e-01
9.53077853e-01 -9.84690309e-01 -8.66844296e-01 5.15409827e-01
5.34573644e-02 2.82479465e-01 2.00047806e-01 1.67356479e+00
-1.68575394e+00 5.33677220e-01 1.98955104e-01 -6.50838196e-01
-1.45763421e+00 -4.57170047e-02 4.59711105e-01 2.98823446e-01
-9.87180471e-01 1.56238139e+00 -1.07258417e-01 -1.18797861e-01
8.30093443e-01 -1.17116404e+00 -5.40798068e-01 -4.15080070e-01
5.25396168e-01 6.50120437e-01 6.80918619e-02 -3.24030846e-01
-2.93085366e-01 3.60062987e-01 2.82299072e-01 3.76947761e-01
1.39408541e+00 1.00535870e-01 -1.87059715e-01 1.71339482e-01
1.46589601e+00 -2.37588644e-01 -7.08253741e-01 4.11853381e-02
-2.76022643e-01 -3.08262147e-02 -1.97134122e-01 -4.74827528e-01
-8.15654755e-01 1.15844738e+00 1.01881444e+00 5.08603990e-01
1.14279366e+00 7.14547858e-02 1.01697838e+00 4.06537652e-01
-2.93833837e-02 -4.28347707e-01 -1.39596924e-01 2.34495282e-01
7.43767381e-01 -1.38295984e+00 -3.46038461e-01 -4.54237312e-02
-3.12969536e-01 1.48704040e+00 2.23600462e-01 -2.59192199e-01
7.12354720e-01 6.83079481e-01 6.21930599e-01 -1.04019009e-02
-5.47959208e-01 8.67961720e-02 4.30343628e-01 4.22357529e-01
9.00083721e-01 2.27239907e-01 -3.46209049e-01 6.70522153e-01
-2.90025622e-02 8.20014924e-02 1.53134942e-01 1.14560759e+00
-7.02572286e-01 -3.89978975e-01 -1.15415126e-01 8.51632535e-01
-1.49183261e+00 -5.24962127e-01 1.35081425e-01 5.48834085e-01
2.64042109e-01 7.58470774e-01 -3.20001990e-01 2.80659720e-02
2.03515589e-02 6.30574897e-02 4.76358205e-01 -9.25453484e-01
-7.26347268e-01 4.36836660e-01 1.37663126e-01 -4.63734388e-01
-3.37996155e-01 -7.01449692e-01 -1.67196822e+00 3.47616971e-01
-2.98176557e-01 2.46330380e-01 7.62965381e-01 6.11040592e-01
-7.00136125e-02 6.61843181e-01 3.85341674e-01 -4.46784705e-01
-3.44211698e-01 -1.19237995e+00 -2.93432087e-01 -6.06132634e-02
1.16092014e+00 -1.14032984e-01 -4.18074518e-01 3.65187600e-02] | [14.072134971618652, -2.2439022064208984] |
abd23f5a-eaa6-48aa-a9a2-3bd7f48b4959 | multiobjective-programming-for-type-2 | 1705.05769 | null | http://arxiv.org/abs/1705.05769v1 | http://arxiv.org/pdf/1705.05769v1.pdf | Multiobjective Programming for Type-2 Hierarchical Fuzzy Inference Trees | This paper proposes a design of hierarchical fuzzy inference tree (HFIT). An
HFIT produces an optimum treelike structure, i.e., a natural hierarchical
structure that accommodates simplicity by combining several low-dimensional
fuzzy inference systems (FISs). Such a natural hierarchical structure provides
a high degree of approximation accuracy. The construction of HFIT takes place
in two phases. Firstly, a nondominated sorting based multiobjective genetic
programming (MOGP) is applied to obtain a simple tree structure (a low
complexity model) with a high accuracy. Secondly, the differential evolution
algorithm is applied to optimize the obtained tree's parameters. In the derived
tree, each node acquires a different input's combination, where the
evolutionary process governs the input's combination. Hence, HFIT nodes are
heterogeneous in nature, which leads to a high diversity among the rules
generated by the HFIT. Additionally, the HFIT provides an automatic feature
selection because it uses MOGP for the tree's structural optimization that
accepts inputs only relevant to the knowledge contained in data. The HFIT was
studied in the context of both type-1 and type-2 FISs, and its performance was
evaluated through six application problems. Moreover, the proposed
multiobjective HFIT was compared both theoretically and empirically with
recently proposed FISs methods from the literature, such as McIT2FIS,
TSCIT2FNN, SIT2FNN, RIT2FNS-WB, eT2FIS, MRIT2NFS, IT2FNN-SVR, etc. From the
obtained results, it was found that the HFIT provided less complex and highly
accurate models compared to the models produced by the most of other methods.
Hence, the proposed HFIT is an efficient and competitive alternative to the
other FISs for function approximation and feature selection. | ['Ajith Abraham', 'Varun Kumar Ojha', 'Vaclav Snasel'] | 2017-05-16 | null | null | null | null | ['time-series-regression'] | ['time-series'] | [-1.76850140e-01 -2.13364020e-01 6.28579110e-02 -1.63946837e-01
1.17301986e-01 -1.47196740e-01 1.46743864e-01 2.52856277e-02
-1.91413999e-01 9.24185216e-01 -4.33677524e-01 -1.63012952e-01
-9.47126925e-01 -1.15620279e+00 -1.62998274e-01 -8.30628455e-01
-5.61439712e-03 7.10260451e-01 3.43180686e-01 -4.86272752e-01
5.29839396e-01 4.76409197e-01 -2.33564734e+00 3.41493785e-02
1.58729827e+00 1.35832405e+00 3.33442330e-01 2.44181499e-01
-2.32526466e-01 6.85646415e-01 -6.02879345e-01 -4.16720241e-01
-2.95161586e-02 -4.10743058e-01 -6.55284226e-01 -1.93444684e-01
-3.43316615e-01 2.04801574e-01 3.40167403e-01 1.04521036e+00
2.14552194e-01 5.04934847e-01 1.12448931e+00 -1.22221947e+00
-4.73195881e-01 5.48481584e-01 -3.19217980e-01 -7.07038026e-03
1.20919771e-01 -2.54573971e-01 7.13122785e-01 -7.93566644e-01
3.86599302e-01 1.31413901e+00 6.40220582e-01 1.29968643e-01
-9.62424397e-01 -3.78668666e-01 -8.62252265e-02 3.47988576e-01
-1.76251400e+00 1.07662961e-01 5.49464166e-01 -5.15463829e-01
7.40902603e-01 6.33192897e-01 8.57433438e-01 2.04802513e-01
5.37749410e-01 5.20845234e-01 1.22140384e+00 -6.29495740e-01
4.35263574e-01 4.89328355e-01 4.54872400e-01 7.58998275e-01
4.62289602e-01 2.38015950e-01 -3.39718573e-02 -1.40999049e-01
5.21693647e-01 -3.13067526e-01 -2.10958585e-01 3.36245596e-02
-5.11969268e-01 8.16002786e-01 4.40392941e-01 8.40967476e-01
-5.13748765e-01 -4.86133456e-01 3.14314842e-01 2.08406821e-01
-6.95509762e-02 5.16364753e-01 -2.47738630e-01 1.28712133e-01
-9.17834878e-01 1.80715159e-01 7.31495380e-01 6.48289263e-01
7.12749600e-01 3.41055810e-01 -1.11521125e-01 8.24966788e-01
3.63120049e-01 3.51977021e-01 6.45250678e-01 -7.65547693e-01
9.73986387e-02 1.03491533e+00 1.36524901e-01 -1.42122340e+00
-4.33320016e-01 -5.53434670e-01 -1.03812778e+00 3.25579494e-01
-1.64636895e-02 -5.00738770e-02 -7.62414098e-01 1.42197049e+00
3.07835013e-01 -3.98033649e-01 2.06257835e-01 5.90635478e-01
8.71902108e-01 7.92881787e-01 -1.35277957e-01 -6.33464098e-01
1.18929267e+00 -6.06155515e-01 -8.58752489e-01 5.50318778e-01
7.54502192e-02 -4.86918926e-01 8.34195733e-01 8.58668327e-01
-1.02803504e+00 -8.82474422e-01 -1.00137293e+00 6.02412343e-01
-4.60014164e-01 3.48078728e-01 5.23187459e-01 7.77332425e-01
-1.07203925e+00 5.06510437e-01 -4.21521425e-01 -1.93182111e-01
-1.06956922e-01 5.49109638e-01 4.42033410e-02 2.63552457e-01
-1.45826340e+00 1.11773300e+00 9.57672954e-01 4.68461365e-01
-2.76459068e-01 -2.71405220e-01 -6.71932220e-01 1.91038787e-01
2.77221411e-01 -6.93548679e-01 6.99849725e-01 -8.18209887e-01
-1.46114707e+00 1.11874491e-01 -1.25834778e-01 -2.78407931e-01
3.87871861e-01 4.14565653e-01 -6.33350909e-01 2.46053338e-01
-3.25900428e-02 2.98655927e-01 6.96649015e-01 -1.31316948e+00
-8.72685313e-01 -2.57564276e-01 -5.76408505e-02 2.24519029e-01
-4.79120225e-01 -1.22735776e-01 -1.35735601e-01 -4.55763817e-01
1.09038599e-01 -4.27451402e-01 -1.09469980e-01 -5.50435066e-01
-3.59537661e-01 -4.31080967e-01 7.14201152e-01 -4.20640320e-01
1.88872445e+00 -1.85704863e+00 4.91491675e-01 8.00053596e-01
-1.54686868e-01 4.08776820e-01 5.75424194e-01 4.19604212e-01
2.16465876e-01 3.48058313e-01 -4.86889243e-01 3.73181134e-01
-1.25538468e-01 2.89644420e-01 3.60712618e-01 -2.77940243e-01
-9.23581868e-02 4.62729275e-01 -6.08372390e-01 -9.17426109e-01
2.84816712e-01 2.92845070e-01 -5.22269607e-01 -1.97786037e-02
-1.28054932e-01 2.11599539e-03 -8.28529537e-01 8.18501890e-01
6.63003504e-01 1.87607378e-01 1.97735325e-01 -4.90644991e-01
-6.30381346e-01 -5.07033110e-01 -1.47182417e+00 6.98228955e-01
-4.05759960e-01 -8.62301439e-02 6.57451302e-02 -1.12511754e+00
1.42407322e+00 4.43441451e-01 5.11356592e-01 -4.56655055e-01
5.68626881e-01 5.72481215e-01 1.12472370e-01 -5.86128175e-01
4.65432823e-01 -1.59852847e-01 -6.56033754e-02 6.69003353e-02
5.21960296e-03 -3.26702267e-01 6.68635666e-01 -3.03529620e-01
4.05367553e-01 6.22172803e-02 4.71685171e-01 -6.82722807e-01
1.24440372e+00 3.52661237e-02 7.94623852e-01 5.33913612e-01
1.06937751e-01 1.86835602e-02 1.39759019e-01 -4.22001272e-01
-5.18184125e-01 -7.15143263e-01 -5.07508993e-01 3.71636629e-01
3.35191160e-01 -1.09175056e-01 -7.31102705e-01 -4.12122846e-01
7.57107213e-02 1.00849187e+00 -4.23325390e-01 -2.81578809e-01
-3.47627610e-01 -9.99808669e-01 4.89739656e-01 1.66239187e-01
9.99090970e-01 -1.16163290e+00 -8.58433425e-01 3.70303124e-01
-1.50518760e-01 -3.47790182e-01 7.52565861e-02 2.92859972e-01
-9.66726542e-01 -1.00696504e+00 -1.56158268e-01 -8.75832379e-01
5.02091467e-01 -2.41259873e-01 9.23979878e-01 4.14345980e-01
-1.15664750e-01 -1.76317319e-01 -5.37418067e-01 -2.27366045e-01
-4.34657454e-01 -2.83537880e-02 1.87334895e-01 2.47740030e-01
5.78235872e-02 -5.79890132e-01 -1.04025476e-01 5.35376251e-01
-8.67164552e-01 -2.77758151e-01 4.83853132e-01 1.12222910e+00
6.89129293e-01 1.28528202e+00 8.67077172e-01 -6.63313448e-01
7.99441814e-01 -3.05628598e-01 -8.04168046e-01 6.07859850e-01
-9.64568317e-01 1.70150533e-01 1.00546062e+00 -1.19720511e-01
-1.65520072e+00 -1.82920590e-01 -2.40411386e-01 -3.09729934e-01
1.03670113e-01 8.37894857e-01 -3.20823699e-01 -3.26162964e-01
5.51911652e-01 3.94396365e-01 3.92896682e-02 -4.53614265e-01
-1.71718463e-01 8.08104992e-01 4.02776867e-01 -8.21543872e-01
6.56577289e-01 -1.59878686e-01 1.84485614e-01 -8.73498976e-01
-2.98431158e-01 2.45408192e-01 -6.13079250e-01 -3.94755572e-01
8.77531886e-01 -1.61250621e-01 -9.17409718e-01 6.20392263e-01
-6.55612230e-01 3.19172084e-01 -1.65990010e-01 5.16726434e-01
-3.59744489e-01 2.05858737e-01 -4.93450820e-01 -1.43408036e+00
-5.05304158e-01 -1.32933521e+00 3.01162183e-01 6.02893293e-01
-1.67852610e-01 -9.15490150e-01 -5.02262890e-01 1.87101170e-01
2.47076839e-01 3.74831319e-01 1.38584113e+00 -5.09929359e-01
-2.45177686e-01 3.50401886e-02 2.94191599e-01 4.48837250e-01
2.00119331e-01 7.66884208e-01 -3.40157896e-01 1.45015232e-02
4.27644998e-01 -1.76962003e-01 6.47322059e-01 5.67995191e-01
8.79264414e-01 -3.97387177e-01 -2.11281136e-01 5.39773107e-01
1.70060742e+00 1.28334069e+00 7.49987841e-01 5.13504207e-01
2.33526409e-01 6.44152761e-01 1.06423044e+00 3.22906375e-01
3.74443769e-01 5.81059456e-01 3.20766240e-01 1.74802095e-01
3.87559146e-01 1.55722871e-01 5.34127429e-02 9.56355453e-01
-4.38965499e-01 -2.76147217e-01 -6.26625717e-01 2.39784881e-01
-1.72483087e+00 -9.75111604e-01 -2.65872627e-01 2.09879541e+00
6.28430128e-01 2.90020347e-01 1.10503569e-01 8.76880050e-01
9.64346349e-01 -4.60472554e-01 -3.44030648e-01 -9.56890702e-01
-2.26351291e-01 2.88392991e-01 -9.83060822e-02 4.91261750e-01
-9.07525599e-01 5.25140166e-01 5.50476503e+00 1.09390152e+00
-1.07096016e+00 -3.99292111e-01 4.96645242e-01 2.23966837e-01
-4.19110984e-01 -1.78243637e-01 -7.25140572e-01 7.38251209e-01
6.81156993e-01 -3.57642621e-01 3.80521119e-01 5.28010309e-01
1.33958146e-01 -4.90599245e-01 -4.23713773e-01 6.40503407e-01
-3.34377229e-01 -1.03667152e+00 3.19470197e-01 -1.49504215e-01
7.62649596e-01 -8.68123829e-01 -2.95447335e-02 2.81019866e-01
-2.10415293e-02 -8.57544005e-01 8.10585439e-01 7.62653887e-01
5.34302950e-01 -1.28598309e+00 1.02969754e+00 4.94208395e-01
-1.41987097e+00 -5.96605718e-01 -3.05803269e-01 1.45924971e-01
7.75539204e-02 7.15704203e-01 -4.21128690e-01 1.47970700e+00
8.18631768e-01 3.30256164e-01 -5.15610337e-01 1.03383327e+00
1.70544580e-01 2.94743329e-01 -4.37613666e-01 -3.71885031e-01
1.34797871e-01 -6.56936526e-01 6.04808688e-01 7.24525571e-01
7.69802570e-01 1.79370910e-01 8.58816877e-02 8.87123227e-01
5.83929598e-01 4.11185354e-01 -2.69380897e-01 1.25898376e-01
9.61049497e-01 1.15511966e+00 -7.73794591e-01 -3.19464743e-01
1.08307093e-01 2.98083723e-01 -2.52087042e-02 -3.10990401e-03
-7.90528655e-01 -7.59200752e-01 9.07108635e-02 -9.33454931e-02
4.14097875e-01 3.46101224e-01 -4.11961019e-01 -8.49955976e-01
-1.07500829e-01 -8.99729550e-01 8.13283324e-01 -7.27463245e-01
-1.17495251e+00 1.18776631e+00 2.43119285e-01 -1.12346184e+00
-4.55368966e-01 -4.49504882e-01 -5.32594621e-01 1.01343930e+00
-1.07334733e+00 -7.92410195e-01 -1.96034014e-01 6.42923772e-01
2.43634090e-01 -3.11932057e-01 7.52842069e-01 2.95221418e-01
-9.23332691e-01 5.91783822e-01 1.49513677e-01 -5.82954407e-01
-8.34906548e-02 -1.11698914e+00 -5.87306917e-01 7.78588891e-01
-6.19673312e-01 5.70768535e-01 5.74434102e-01 -8.14630806e-01
-1.09805834e+00 -8.48763108e-01 9.09359813e-01 3.09928715e-01
1.77460253e-01 2.93288261e-01 -9.64390755e-01 8.97998884e-02
-5.49087636e-02 -5.81370175e-01 1.99300528e-01 -4.56058562e-01
3.64651203e-01 -4.41518426e-01 -1.73162258e+00 3.61166984e-01
6.53212607e-01 1.55507460e-01 -6.99775696e-01 -2.01432943e-01
3.04945558e-01 -6.39499500e-02 -1.44001031e+00 8.31047177e-01
5.94613194e-01 -1.23464644e+00 8.96712542e-01 1.98388882e-02
1.57037824e-01 -6.41938150e-01 -1.20174326e-01 -1.13705575e+00
-7.13929415e-01 -3.00476909e-01 -2.98884548e-02 1.41505849e+00
3.27040434e-01 -9.23203468e-01 1.94949701e-01 3.84796590e-01
-2.84688920e-01 -1.10049391e+00 -8.71497571e-01 -9.13230479e-01
-2.19181567e-01 2.45733321e-01 9.62236226e-01 6.06168807e-01
-1.84048995e-01 1.35396063e-01 -3.12118270e-02 -9.86902639e-02
6.37452066e-01 3.35313022e-01 -2.90034395e-02 -1.77964580e+00
-1.40566245e-01 -6.13706410e-01 -3.26921254e-01 -2.95016080e-01
-4.59599309e-02 -8.05726349e-01 -1.04871660e-01 -1.48397267e+00
-2.02141389e-01 -8.28593671e-01 -3.50078613e-01 3.20406973e-01
-1.38071761e-01 2.15642359e-02 9.35965255e-02 3.20212811e-01
1.20970659e-01 6.45330787e-01 1.38997674e+00 8.38855058e-02
-4.82903093e-01 2.78919935e-01 -6.77330911e-01 7.31378138e-01
7.19360828e-01 -1.82031944e-01 -7.69873857e-01 1.04244679e-01
-8.50116685e-02 3.02725434e-01 -1.55053854e-01 -1.06921732e+00
1.38056606e-01 -2.58874953e-01 4.56434071e-01 -8.28202724e-01
2.46144265e-01 -9.74023998e-01 7.86429644e-01 7.58898973e-01
1.56905577e-01 1.34538457e-01 8.55153576e-02 9.42760110e-02
-4.73377019e-01 -7.70058393e-01 8.54228437e-01 -2.04890758e-01
-6.34021819e-01 -1.25529692e-01 -5.88130713e-01 -2.42344156e-01
1.12503910e+00 -8.17901671e-01 1.43521745e-03 -1.56943873e-03
-6.88686967e-01 3.48626375e-01 2.08501711e-01 -3.53024043e-02
5.39631426e-01 -1.29924202e+00 -4.39008594e-01 2.21132621e-01
-3.70603472e-01 8.50153062e-03 3.49330634e-01 9.19472516e-01
-5.78152716e-01 3.41804117e-01 -5.74621558e-01 -5.12961805e-01
-1.07868600e+00 5.04669309e-01 5.01569390e-01 -3.72845024e-01
-1.91563949e-01 5.39296389e-01 -1.71715721e-01 -5.12784600e-01
-3.72581705e-02 -2.67835021e-01 -1.08440411e+00 4.59473252e-01
4.42615785e-02 7.37455785e-01 3.23707014e-01 -8.02415311e-01
-4.44198400e-01 8.14538121e-01 3.79336983e-01 4.91954498e-02
1.29508460e+00 -7.78068006e-02 -5.61577797e-01 2.59467632e-01
7.60084093e-01 -2.33793378e-01 -5.48558593e-01 3.67842346e-01
5.05958311e-02 -2.70256877e-01 3.14789638e-02 -9.32300866e-01
-1.09098804e+00 3.27403694e-01 3.16223800e-01 4.16199476e-01
1.78269255e+00 -7.05049038e-01 5.24274945e-01 3.45340997e-01
7.50872552e-01 -1.12945592e+00 -4.36433911e-01 4.90429789e-01
9.04958308e-01 -4.34534848e-01 -3.31731886e-02 -5.36170721e-01
-6.64223135e-01 1.38485813e+00 8.15748334e-01 1.24019787e-01
7.07988024e-01 3.24066699e-01 -4.86814857e-01 -1.25314116e-01
-6.23676002e-01 -8.17775056e-02 4.35314387e-01 4.40189719e-01
1.54305175e-01 -2.92366836e-02 -1.02950513e+00 1.05106020e+00
-4.96450067e-01 2.84744263e-01 3.33729863e-01 8.69670331e-01
-6.53631747e-01 -1.04497945e+00 -6.15179062e-01 6.13960564e-01
-1.64262086e-01 2.00227454e-01 -1.23667359e-01 1.04632938e+00
5.69415331e-01 1.20884264e+00 -2.06036910e-01 -8.45197558e-01
5.17371237e-01 1.41170379e-02 4.05997634e-01 -1.98514640e-01
-1.00711083e+00 -5.58231138e-02 1.95812449e-01 -1.08987026e-01
-3.77838314e-01 -5.86925209e-01 -1.49507296e+00 -2.80456036e-01
-6.81403279e-01 1.01675832e+00 4.07157540e-01 8.38803470e-01
6.35169894e-02 7.44632721e-01 9.08363044e-01 -5.56015372e-01
-4.79811341e-01 -7.35913038e-01 -6.78800225e-01 -5.91942519e-02
-2.48126805e-01 -1.24705851e+00 -4.77830261e-01 -3.17293912e-01] | [6.120675086975098, 3.5273244380950928] |
7603c037-3af7-43a9-9e6a-6401311e4aa4 | color-image-restoration-exploiting-inter | null | null | https://ieeexplore.ieee.org/document/9286520 | https://ieeexplore.ieee.org/document/9286520 | Color Image Restoration Exploiting Inter-channel Correlation with a 3-stage CNN | Image restoration is a critical component of image processing pipelines and for low-level computer vision tasks. Conventional image restoration approaches are mostly based on hand-crafted image priors. The inter-channel correlation of color images is not fully exploited. Motivated by the special characteristics of the inter-channel correlation (higher correlation for red/green and green/blue channels than for red/blue) in color images and general characteristics (green channel always shows the best image quality among the three color components) of distorted color images, in this paper, a 3-stage convolutional neural network (CNN) structure is proposed for color image restoration tasks. Since the green channel is found to have the best quality among all three channels, in the first stage, the network is designed to reconstruct the green component. Then, with the guidance of the reconstructed green channel from the first stage, the red and blue channels are reconstructed in the second stage with two parallel networks. Finally, the intermediate reconstructions from the previous stages are concatenated and further refined jointly. We demonstrate the capabilities of the proposed 3-stage structure with three typical color image restoration tasks: color image demosaicking, color compression artifacts reduction, and real-world color image denoising. In addition, we integrate pixel-shuffle convolution into our scheme to improve the efficiency, and also introduce a quality-blind training strategy to simplify the training process for the compression artifacts reduction task. Extensive experimental results and analyses show that the proposed structure successfully exploits the spatial and inter-channel correlation of color images and outperforms the state-of-the-art image reconstruction approaches. | ['Kai Cui; Atanas Boev; Elena Alshina; Eckehard Steinbach'] | 2020-12-08 | null | null | null | null | ['color-image-denoising'] | ['computer-vision'] | [ 4.06238616e-01 -6.51767135e-01 1.96309611e-01 -6.52301311e-02
-5.03818333e-01 -1.71896309e-01 1.94927543e-01 -2.54023373e-01
-5.64825416e-01 4.76006091e-01 1.32570546e-02 -3.69733155e-01
1.83898229e-02 -7.43075490e-01 -6.58507526e-01 -1.06357217e+00
1.06317565e-01 -4.11424965e-01 5.94190471e-02 -2.11198002e-01
2.90169448e-01 5.44296920e-01 -1.50204146e+00 3.11318338e-01
1.14227891e+00 1.18100953e+00 4.10851866e-01 6.77486598e-01
-1.21903822e-01 6.88671827e-01 -2.33240798e-01 -2.57005066e-01
5.15707374e-01 -7.26367831e-01 -3.44263107e-01 4.63432640e-01
2.33879566e-01 -6.67637050e-01 -5.12594402e-01 1.51377881e+00
4.08681214e-01 1.79858491e-01 1.10755637e-01 -9.47845638e-01
-7.84648836e-01 2.52818435e-01 -7.68776476e-01 -8.75412151e-02
-7.82970339e-02 2.94264734e-01 6.18038297e-01 -9.09466445e-01
2.58073568e-01 1.07580590e+00 5.14111817e-01 3.11577380e-01
-1.22079861e+00 -6.83253407e-01 9.97828692e-02 3.48497301e-01
-1.23058414e+00 -3.15241873e-01 1.06149459e+00 -1.14956096e-01
3.45384240e-01 1.17229067e-01 8.05886865e-01 4.53552246e-01
6.30266219e-02 5.98829389e-01 1.31438112e+00 -3.73351127e-01
1.88133672e-01 -1.83789700e-01 -1.95018724e-01 6.67859912e-01
2.42090419e-01 2.36855716e-01 -4.50864971e-01 2.98614323e-01
1.00507712e+00 2.70393133e-01 -6.50222898e-01 -3.12265247e-01
-1.03854680e+00 3.59999239e-01 7.63388634e-01 3.44176233e-01
-5.25150299e-01 1.37200713e-01 4.00766656e-02 1.51522130e-01
1.86721593e-01 -1.51321534e-02 -2.81593800e-01 2.59827375e-01
-1.07381403e+00 -1.00472555e-01 4.03734922e-01 6.77557290e-01
1.03529608e+00 2.14906901e-01 -1.06723443e-01 1.07161140e+00
2.79568940e-01 4.27547753e-01 3.29238981e-01 -1.02975774e+00
3.02566230e-01 4.03949797e-01 2.21224591e-01 -9.35322404e-01
-3.00290376e-01 -4.67402369e-01 -1.46079421e+00 6.43858790e-01
5.27283669e-01 8.64021629e-02 -1.03492475e+00 1.37741911e+00
4.38763611e-02 6.36388510e-02 2.21808627e-01 1.20582700e+00
6.03658974e-01 7.74792314e-01 -1.31894514e-01 -3.25703472e-01
1.30856001e+00 -9.01455700e-01 -6.86058342e-01 -1.53379396e-01
-1.57083482e-01 -9.58144367e-01 9.80689406e-01 5.50940871e-01
-1.26942599e+00 -9.17696357e-01 -1.16481853e+00 -2.83474922e-01
3.44965383e-02 4.72777456e-01 5.24502635e-01 5.09026885e-01
-1.16139817e+00 6.99940026e-01 -7.62923837e-01 4.20558266e-02
3.80638659e-01 -4.85733189e-02 -3.12009811e-01 -8.12006652e-01
-8.21936488e-01 5.09202600e-01 2.34691978e-01 6.83647513e-01
-7.32611954e-01 -4.89920795e-01 -5.68446040e-01 1.99714467e-01
4.47297469e-02 -4.05507505e-01 6.29182696e-01 -1.32171845e+00
-1.60265994e+00 4.86343145e-01 -8.15089792e-02 -9.00322795e-02
5.88432431e-01 -1.73250481e-01 -3.50682557e-01 4.11747545e-01
-1.39015079e-01 4.97742653e-01 1.06807446e+00 -1.64712834e+00
-6.34882808e-01 -2.64076054e-01 -1.66791707e-01 1.24069020e-01
-2.38573223e-01 -7.71979839e-02 -1.12842345e+00 -6.98824584e-01
6.83080137e-01 -6.32641912e-01 -2.08821684e-01 4.04298842e-01
-4.28539991e-01 4.63142574e-01 6.60323501e-01 -1.19761884e+00
9.15676236e-01 -2.44121790e+00 3.41328867e-02 3.77824485e-01
-5.93063468e-03 3.71435583e-01 -3.23247492e-01 1.66058227e-01
-2.76609242e-01 -2.82662690e-01 -6.46384180e-01 -3.13211352e-01
-3.78332883e-01 1.25474662e-01 3.37541066e-02 6.18271768e-01
1.83027342e-01 4.22997326e-01 -7.39913821e-01 -2.63137907e-01
4.06988323e-01 8.26520741e-01 -4.90260690e-01 2.88975924e-01
2.63473094e-01 7.69165337e-01 -9.16103050e-02 6.72015071e-01
1.30990911e+00 -1.64639086e-01 3.40285540e-01 -6.99597001e-01
-2.16595501e-01 -1.03183791e-01 -1.33470297e+00 1.56809926e+00
-5.78959346e-01 5.82194328e-01 5.21532238e-01 -9.99613464e-01
9.36119676e-01 1.68836892e-01 5.48715591e-01 -1.09491098e+00
-2.18727048e-02 3.15727711e-01 2.23462388e-01 -4.12394226e-01
4.25225556e-01 -9.37807113e-02 4.05297369e-01 2.28176355e-01
-8.31174478e-02 -1.25444099e-01 3.24953139e-01 2.03402609e-01
7.20184445e-01 1.60364956e-01 -5.87350130e-02 8.71772412e-03
8.52703691e-01 -3.25037628e-01 7.78235316e-01 6.31863415e-01
-1.27725169e-01 1.07394671e+00 3.61488521e-01 -3.21573555e-01
-1.13391852e+00 -9.00065899e-01 -2.64354851e-02 7.01162875e-01
5.69097340e-01 -9.38506871e-02 -7.10591793e-01 -2.39851996e-01
-2.39946932e-01 3.82402807e-01 -3.70930642e-01 -7.47800712e-03
-6.65306211e-01 -8.05288613e-01 1.05236247e-01 2.95598656e-01
1.12207854e+00 -1.01939797e+00 -3.79722416e-01 2.74446487e-01
-3.09483469e-01 -1.00280511e+00 -3.91343147e-01 1.26881123e-01
-8.74139309e-01 -1.34402919e+00 -8.84459138e-01 -7.92573571e-01
9.99590099e-01 7.37103939e-01 7.84102798e-01 5.93157053e-01
-2.27931499e-01 8.39068145e-02 -4.66472626e-01 2.47868106e-01
-2.88168430e-01 -7.05684006e-01 -3.60245168e-01 5.72918534e-01
-3.01629484e-01 -7.38823593e-01 -1.04603171e+00 1.38892084e-01
-1.27680302e+00 4.27649677e-01 8.64006996e-01 1.11528182e+00
5.31795859e-01 5.87616861e-01 -5.09572849e-02 -6.95764184e-01
3.96371394e-01 8.34311619e-02 -8.59918773e-01 2.45756313e-01
-4.38877881e-01 -2.73087174e-02 9.73565757e-01 -3.04809839e-01
-1.21625316e+00 3.01431715e-01 -2.99199760e-01 -2.51605898e-01
-6.72867149e-02 2.96012580e-01 -3.98141950e-01 -2.36636832e-01
1.90726668e-01 6.22988224e-01 2.54979908e-01 -7.68728912e-01
3.21872145e-01 3.53008717e-01 8.64919543e-01 -2.53993034e-01
9.58631694e-01 6.95611954e-01 -6.81947619e-02 -7.59303451e-01
-3.07148516e-01 -3.52485746e-01 -4.64863300e-01 -2.59525955e-01
7.67546356e-01 -1.04387689e+00 -7.65720844e-01 1.06667101e+00
-1.03972983e+00 -2.80731797e-01 4.27438542e-02 4.06857729e-01
-2.30966717e-01 6.35097802e-01 -8.20441544e-01 -6.06444478e-01
-3.74788046e-01 -1.47346902e+00 7.56671786e-01 4.98341948e-01
7.31720567e-01 -6.24799728e-01 -6.19974971e-01 2.00229526e-01
5.61356664e-01 -1.15725257e-01 9.76781964e-01 2.81772226e-01
-7.52069652e-01 -5.91830239e-02 -7.82708645e-01 8.25493455e-01
1.92513242e-01 -5.92645407e-02 -7.36531734e-01 -3.72687668e-01
-8.33628997e-02 2.45032720e-02 1.22319031e+00 5.51536620e-01
1.39009500e+00 -7.30705336e-02 2.29508117e-01 1.11875415e+00
1.77353251e+00 1.55878872e-01 1.20596278e+00 4.49481398e-01
7.59663761e-01 4.67700899e-01 3.00185770e-01 4.07031953e-01
2.44409770e-01 4.47771788e-01 4.61662531e-01 -7.29405701e-01
-5.44817090e-01 -4.98045720e-02 3.52795899e-01 5.96649766e-01
-1.14590183e-01 6.96822628e-02 -4.08574045e-01 3.50501835e-01
-1.46861660e+00 -6.86881244e-01 -1.47084266e-01 2.19418669e+00
8.70458424e-01 -2.28193849e-01 -3.36142123e-01 2.90420324e-01
6.94150269e-01 1.29943877e-01 -5.34557819e-01 1.27043992e-01
-4.94194120e-01 2.53488719e-01 6.39262855e-01 3.90048414e-01
-1.01331842e+00 6.45135283e-01 5.23620892e+00 7.38565803e-01
-1.13869631e+00 -1.01301290e-01 9.19551492e-01 2.38872290e-01
-1.76269367e-01 1.78270429e-01 -1.23723283e-01 5.12983739e-01
1.80832386e-01 4.35772985e-01 9.98732746e-01 2.94366598e-01
3.21162224e-01 -5.07299304e-01 -7.77858377e-01 1.14667463e+00
-6.15377203e-02 -1.02226913e+00 2.98256730e-03 -2.14619324e-01
7.96601713e-01 -2.46186063e-01 1.80020839e-01 -2.80780494e-01
2.36105248e-01 -8.06435943e-01 8.26425254e-01 4.88725662e-01
8.32111537e-01 -8.32249999e-01 6.73055291e-01 1.57976061e-01
-1.11501050e+00 -3.17615807e-01 -4.96934086e-01 3.01785886e-01
2.18775898e-01 9.44353700e-01 1.16775885e-01 6.95171535e-01
1.14275277e+00 8.81093204e-01 -4.96880352e-01 1.22828054e+00
-5.30897737e-01 4.08183098e-01 -1.23826647e-02 5.96839905e-01
8.46847296e-02 -7.40909934e-01 2.62107134e-01 1.09283853e+00
4.07757998e-01 3.46670628e-01 -6.62416965e-02 9.29453492e-01
-1.53542548e-01 -2.71598041e-01 5.36587536e-02 2.38954619e-01
1.29233152e-01 1.43305993e+00 -7.46614099e-01 -3.99520308e-01
-6.35400355e-01 1.26102221e+00 -1.39853522e-01 8.46721351e-01
-5.50145805e-01 -4.27177727e-01 7.09370017e-01 -1.90991163e-01
5.15366793e-01 -1.49031684e-01 -4.25635576e-01 -1.13546503e+00
8.02750960e-02 -1.04521394e+00 9.88616720e-02 -9.35348690e-01
-1.03615320e+00 5.43117046e-01 -5.49603581e-01 -1.45817101e+00
2.60725200e-01 -5.66141546e-01 -5.92628002e-01 1.15219617e+00
-2.06416273e+00 -1.05682254e+00 -5.35555899e-01 9.48827446e-01
2.40620092e-01 1.07836410e-01 4.11416531e-01 5.20836353e-01
-6.94791079e-01 2.89033353e-01 4.97618496e-01 2.98599601e-01
5.77743530e-01 -9.54615116e-01 1.91483513e-01 1.52012956e+00
-3.26972783e-01 5.80359817e-01 3.98758948e-01 -5.21812320e-01
-1.61917150e+00 -9.24058914e-01 2.62883723e-01 8.17641616e-01
1.84256271e-01 -9.19426791e-03 -1.05148995e+00 4.00987789e-02
1.59972519e-01 3.49062145e-01 2.42150635e-01 -4.19353008e-01
-5.00712752e-01 -5.63992143e-01 -1.03458750e+00 4.14227009e-01
5.74148655e-01 -4.75984484e-01 1.52304605e-01 8.69636312e-02
2.25095153e-01 -2.31241420e-01 -4.78902876e-01 1.80447891e-01
6.46838427e-01 -1.35948682e+00 1.23273706e+00 2.09169909e-01
5.85462689e-01 -7.24744141e-01 -1.59653828e-01 -1.30804014e+00
-2.53102839e-01 -3.62660289e-01 4.07666862e-01 1.06902540e+00
1.45077750e-01 -4.76283222e-01 4.84333605e-01 4.41053480e-01
-1.74183503e-01 -2.56664932e-01 -6.54957891e-01 -3.14100653e-01
-2.36021236e-01 -1.92536786e-01 3.77277851e-01 6.37919664e-01
-5.13405800e-01 -3.32746267e-01 -6.78388298e-01 4.23465073e-01
9.53879952e-01 4.38418210e-01 6.17763042e-01 -8.19162548e-01
-4.14595693e-01 -4.58420753e-01 3.15965107e-03 -1.27138829e+00
-3.32007706e-01 -4.98324007e-01 4.51726735e-01 -1.61963689e+00
2.78745145e-01 -4.49649870e-01 -5.13670564e-01 3.45538437e-01
-5.14221966e-01 6.67954564e-01 4.67855364e-01 2.08659559e-01
-3.50880325e-01 5.22327900e-01 1.55463660e+00 -2.58176625e-01
-1.92181379e-01 -1.40064240e-01 -8.02910209e-01 5.61300278e-01
5.34577787e-01 -9.22124609e-02 -5.25923856e-02 -6.58802211e-01
-1.17617011e-01 2.23551601e-01 5.47528744e-01 -1.12855887e+00
3.67650807e-01 -5.86417839e-02 8.85131717e-01 -4.11886990e-01
3.47744495e-01 -1.09611726e+00 1.96271777e-01 5.64072907e-01
-7.11372942e-02 -2.58280963e-01 3.05139776e-02 4.49525356e-01
-5.63324809e-01 9.96285826e-02 1.30179048e+00 -2.09645659e-01
-7.76642859e-01 2.03812748e-01 -3.34874630e-01 -5.56210995e-01
4.86470461e-01 -3.94068301e-01 -1.73802897e-01 -5.03012955e-01
-4.05853361e-01 -9.01087075e-02 5.72358966e-01 2.58403365e-02
1.04588914e+00 -1.04619491e+00 -7.83042550e-01 6.47338510e-01
-1.73926935e-01 -1.54829413e-01 6.63520873e-01 9.47582424e-01
-8.68159115e-01 -3.14752012e-01 -5.47944903e-01 -4.38870907e-01
-1.04393077e+00 4.09301281e-01 4.43603218e-01 -1.01228856e-01
-8.30157101e-01 6.98616982e-01 1.85234979e-01 -2.15180703e-02
2.63880432e-01 -2.30232909e-01 -1.04306437e-01 -3.13377917e-01
7.32398868e-01 3.72074157e-01 1.50820628e-01 -7.91836619e-01
-8.34074467e-02 6.85668647e-01 4.82834615e-02 -8.49227607e-02
1.54163909e+00 -3.86544645e-01 -6.40911698e-01 -2.05343157e-01
1.27468109e+00 -3.72233316e-02 -1.55719602e+00 -3.27002823e-01
-5.30399442e-01 -8.67140174e-01 4.89108622e-01 -8.61221313e-01
-1.82876909e+00 9.68965650e-01 9.40153480e-01 -8.43742043e-02
2.03829408e+00 -5.68467021e-01 7.80924201e-01 -1.06668569e-01
1.29105374e-01 -1.03700042e+00 8.22588727e-02 1.49918407e-01
7.68681824e-01 -1.10543525e+00 1.92019567e-01 -4.47700322e-01
-4.64509964e-01 1.33495975e+00 4.20383632e-01 -6.72036931e-02
4.53312099e-01 5.97429872e-02 1.23226427e-01 2.07821518e-01
-9.25293341e-02 -3.31140518e-01 2.49910831e-01 5.73496699e-01
3.55497032e-01 -1.10736802e-01 -3.16838592e-01 2.72385985e-01
3.11989307e-01 -1.64208412e-01 5.80922425e-01 5.55825233e-01
-3.07481647e-01 -1.18048930e+00 -6.49819493e-01 1.67746276e-01
-4.45294470e-01 -4.16962534e-01 1.01688072e-01 4.94361311e-01
2.42022321e-01 1.26034760e+00 8.18412676e-02 -3.03430736e-01
2.01723978e-01 -4.58310694e-01 3.37854743e-01 3.32876071e-02
-5.34300804e-01 5.02658248e-01 -3.45181704e-01 -8.66384029e-01
-5.06774187e-01 -3.91033947e-01 -1.10989571e+00 -3.61531883e-01
-1.16919324e-01 -8.04006606e-02 8.99694622e-01 5.30799747e-01
1.09014794e-01 5.79258919e-01 7.85360813e-01 -1.15678346e+00
-5.28538115e-02 -7.82814384e-01 -9.18899894e-01 6.67710185e-01
6.33887589e-01 -2.90421486e-01 -3.19766700e-01 2.69392997e-01] | [10.923559188842773, -2.111231565475464] |
d5da0414-8aa1-406c-bc90-2384342c765c | sidi-kws-a-large-scale-multilingual-dataset | null | null | https://www.isca-speech.org/archive/interspeech_2022/meneses22_interspeech.html | https://www.isca-speech.org/archive/pdfs/interspeech_2022/meneses22_interspeech.pdf | SiDi KWS: A Large-Scale Multilingual Dataset for Keyword Spotting | Keyword spotting (KWS) has become a hot topic in speech processing due to the rise of commercial applications based on voice command detection, such as voice assistants. Like tasks in computer vision, natural language processing, and even speech processing, most current successful approaches for KWS rely on deep learning. However, differently from all those tasks, there is a lack of large-scale datasets designed for training and evaluating deep learning models for KWS. The current work presents SiDi KWS, a public large-scale multilingual dataset currently composed of 24.3 million audio recordings of labeled single-spoken keywords. It intends to boost the development of new KWS systems, especially those based on deep learning. That dataset has been created by applying automatic forced alignment on public datasets of transcribed speech. This work introduces SiDi KWS and KeywordMiner, an open-source framework used to generate that dataset, to benefit the speech processing research community. | ['Gabriela Dantas Rocha', 'Luis Vasconcelos Peres', 'Rafael Bérgamo Holanda', 'Michel Cardoso Meneses'] | 2022-09-22 | null | null | null | interspeech-2022-9 | ['keyword-extraction', 'keyword-spotting'] | ['natural-language-processing', 'speech'] | [ 1.46738723e-01 9.63351876e-03 -1.96561199e-02 -5.68901777e-01
-1.11474311e+00 -4.03833628e-01 6.04125321e-01 1.95958372e-02
-6.08806014e-01 4.14851338e-01 5.11600792e-01 -6.13106489e-01
1.06750773e-02 -2.37972692e-01 -4.99173731e-01 -3.86605501e-01
2.07855105e-01 6.36219203e-01 -1.18629104e-02 2.93314401e-02
1.25548318e-01 4.15306956e-01 -1.32197702e+00 4.22718525e-01
4.35974747e-01 8.51009309e-01 5.18359721e-01 7.76804984e-01
-3.06228787e-01 5.36330998e-01 -7.57267654e-01 -7.73244724e-02
-9.14666057e-02 -2.37983271e-01 -9.22288895e-01 -1.02057658e-01
4.53952879e-01 -2.83319712e-01 -2.18591928e-01 6.61802530e-01
9.90621865e-01 2.29621172e-01 3.84197712e-01 -1.24217594e+00
-4.29348320e-01 1.21242726e+00 5.92777738e-03 3.34987313e-01
3.38490009e-01 7.68346786e-02 1.14153540e+00 -1.13957524e+00
6.38566911e-01 1.38571024e+00 4.76282269e-01 4.67751503e-01
-8.61609280e-01 -6.05973065e-01 -1.55014679e-01 5.34906745e-01
-1.42623413e+00 -9.62651432e-01 7.86508381e-01 -2.02909812e-01
1.49449337e+00 3.59542161e-01 3.86035562e-01 1.35287702e+00
-2.73562640e-01 1.24850881e+00 9.94485617e-01 -7.95447171e-01
1.64758548e-01 2.61491418e-01 1.06971912e-01 3.83575827e-01
-6.13048732e-01 -3.27571601e-01 -1.04819751e+00 -2.11093783e-01
-9.17706117e-02 -7.37088025e-01 -3.40742975e-01 1.23260498e-01
-1.53114855e+00 7.75984287e-01 -3.16795319e-01 6.22638881e-01
-4.54219967e-01 -3.34799975e-01 5.38132608e-01 4.69028711e-01
6.53475344e-01 4.14289474e-01 -8.45904827e-01 -6.39443278e-01
-1.29611301e+00 3.72894168e-01 9.81121719e-01 7.63910294e-01
5.79033732e-01 4.01663572e-01 -1.58036530e-01 1.29185605e+00
2.66563863e-01 5.98088562e-01 9.67720687e-01 -6.30857766e-01
5.11104941e-01 4.77500528e-01 -3.35076421e-01 -5.65508306e-01
-4.48668569e-01 -4.80344184e-02 -5.03039777e-01 -2.71035343e-01
1.95410326e-01 -1.32192582e-01 -8.30710709e-01 1.62260020e+00
2.48276651e-01 -3.85518000e-03 1.23118900e-01 8.24816406e-01
1.15496957e+00 9.93015051e-01 -3.10502142e-01 -2.57451355e-01
1.14131594e+00 -1.03711677e+00 -8.69219720e-01 -6.14571683e-02
6.05080187e-01 -1.15219629e+00 1.44161189e+00 5.59300721e-01
-8.71333599e-01 -3.88951540e-01 -7.13935912e-01 -1.12525754e-01
-8.63942623e-01 2.87564069e-01 1.79374263e-01 5.29802382e-01
-1.27152383e+00 3.78459059e-02 -4.15564686e-01 -6.35304213e-01
3.05053685e-02 2.13783845e-01 -4.32704359e-01 1.07545845e-01
-1.41566753e+00 8.69602740e-01 3.57654572e-01 1.69248655e-01
-1.05988324e+00 -6.54785752e-01 -8.41991425e-01 -8.78561139e-02
4.25447792e-01 -8.54971781e-02 1.82185400e+00 -7.31640637e-01
-1.74988747e+00 1.07698071e+00 -3.15832317e-01 -7.42034614e-01
3.25224251e-01 -3.11121494e-01 -6.37018740e-01 -2.97280043e-01
-8.63765851e-02 6.14063919e-01 1.08150971e+00 -7.26067841e-01
-5.18910170e-01 -4.73858029e-01 -4.77943748e-01 1.71131566e-01
-6.12841666e-01 6.21314168e-01 -3.94170523e-01 -7.10457802e-01
-3.09313029e-01 -8.24772596e-01 1.27298877e-01 -6.35492563e-01
-8.51758480e-01 -7.89621651e-01 1.02946961e+00 -9.99503613e-01
1.26084554e+00 -2.11393571e+00 1.65778637e-01 -1.47456704e-02
-1.47341207e-01 9.72730160e-01 -3.09224337e-01 7.39568710e-01
1.55300811e-01 -5.73940277e-02 -1.63083375e-01 -6.41823709e-01
1.64210066e-01 2.61183470e-01 -5.77249169e-01 2.29167238e-01
1.31326497e-01 6.51638389e-01 -6.13744795e-01 -2.56661892e-01
3.49864095e-01 2.90748626e-01 -5.74866645e-02 4.59722638e-01
-3.37605745e-01 1.51751503e-01 2.09457397e-01 5.51483810e-01
4.06049281e-01 5.04208088e-01 -1.04115583e-01 9.39022526e-02
-4.74830180e-01 8.69858563e-01 -1.01018333e+00 1.81946003e+00
-6.82536900e-01 7.76539743e-01 3.14365089e-01 -1.08518946e+00
8.61161411e-01 6.64400756e-01 2.42635831e-01 -5.27695596e-01
3.66653912e-02 5.37387550e-01 1.66562218e-02 -4.48500156e-01
7.59968460e-01 3.18056077e-01 -1.91438496e-01 5.09237230e-01
4.18282121e-01 -3.54542017e-01 4.44172807e-02 1.47829816e-01
9.36841190e-01 -3.41850996e-01 9.44553614e-02 9.38597415e-03
7.67426789e-01 1.07414819e-01 2.69204229e-01 6.25989497e-01
-1.31154820e-01 6.31796181e-01 3.37088555e-02 -3.57942700e-01
-1.02203405e+00 -6.61781132e-01 -1.10788457e-02 1.43013418e+00
-6.64187908e-01 -4.03320462e-01 -1.06771350e+00 -2.36219093e-01
-9.21789184e-02 6.90384150e-01 1.82398912e-02 2.04042215e-02
-5.93702972e-01 -3.23727041e-01 1.17827570e+00 2.22267155e-02
2.40025237e-01 -1.69707358e+00 -6.33620992e-02 4.33548599e-01
-2.81331569e-01 -1.57231808e+00 -4.92488235e-01 1.88125774e-01
-1.91123486e-01 -8.13981712e-01 -8.61638546e-01 -9.88331854e-01
-1.60590574e-01 2.18113959e-01 9.92881060e-01 -4.21203852e-01
-4.71584231e-01 4.71871525e-01 -5.20285726e-01 -7.02888191e-01
-9.60218430e-01 7.61622846e-01 5.60629904e-01 2.54246015e-02
6.44765794e-01 -2.93782324e-01 -1.36020541e-01 -9.57540274e-02
-6.79658592e-01 -4.25775051e-02 6.30948365e-01 6.17085159e-01
3.69341373e-01 -2.24860623e-01 8.03260267e-01 -5.67775071e-01
1.05383778e+00 -2.89851993e-01 -5.73490322e-01 2.11959437e-01
-5.55699766e-01 -5.60395531e-02 5.39020479e-01 -4.13683087e-01
-9.39339936e-01 -5.34563251e-02 -7.01650620e-01 -3.10143381e-01
-5.93039691e-01 7.43076622e-01 -3.37034553e-01 2.78291494e-01
2.57712632e-01 6.80517018e-01 -6.35929406e-02 -9.90115225e-01
5.55724382e-01 1.59966850e+00 6.44674480e-01 -1.05112500e-01
4.38430935e-01 -2.13813946e-01 -8.59885633e-01 -1.58876514e+00
-6.79705560e-01 -8.42444241e-01 -4.69880342e-01 -2.19040707e-01
6.98734939e-01 -7.69689322e-01 -3.52208704e-01 1.05974901e+00
-1.39255524e+00 -4.81079102e-01 7.72344321e-03 4.69438225e-01
-3.90935391e-01 2.05067351e-01 -5.36944687e-01 -8.52854073e-01
-8.58511388e-01 -1.16704488e+00 1.26656449e+00 -1.08204633e-01
-3.93233865e-01 -6.22996151e-01 4.25902814e-01 6.74435914e-01
6.73189163e-01 -6.49273992e-01 8.13792646e-01 -1.23984635e+00
-2.07378730e-01 4.90594320e-02 2.28679702e-01 8.58797789e-01
1.73586428e-01 1.22046173e-01 -1.31143689e+00 -9.20768902e-02
-3.48371506e-01 -7.36252964e-01 6.39859617e-01 1.37247160e-01
1.08043432e+00 -4.49166626e-01 3.77981737e-02 3.24320525e-01
6.60906613e-01 2.47566864e-01 1.63068861e-01 2.54057676e-01
6.87389672e-01 6.27054870e-01 4.17453140e-01 3.30345601e-01
4.12298441e-01 8.65682781e-01 1.44201294e-01 -2.20498294e-02
-2.53891081e-01 -1.42519683e-01 5.03466070e-01 1.56185305e+00
5.19921839e-01 -2.96293646e-01 -1.21605670e+00 8.60964000e-01
-1.55897713e+00 -6.05219901e-01 1.60064384e-01 1.96957719e+00
1.22110641e+00 -2.14852154e-01 1.62076391e-02 3.97958905e-01
5.22202730e-01 4.13326532e-01 -4.45309848e-01 -8.77012253e-01
-2.27923796e-01 3.68531197e-01 6.44822642e-02 5.66403210e-01
-1.14859307e+00 1.47360075e+00 5.46424484e+00 1.08895910e+00
-1.37187099e+00 1.87192842e-01 4.13145781e-01 5.97047769e-02
-2.38940999e-01 -3.69266421e-01 -1.22711301e+00 3.82835150e-01
1.45317900e+00 -2.13839576e-01 6.43304884e-01 9.93477762e-01
5.59878588e-01 2.36106664e-02 -8.52229178e-01 9.74403679e-01
1.45385996e-01 -1.29982686e+00 4.77562845e-02 -7.49488771e-02
2.26091579e-01 6.75653994e-01 2.07682569e-02 5.66772699e-01
8.60504135e-02 -9.75969732e-01 7.61072457e-01 8.07250366e-02
7.38499165e-01 -9.11568463e-01 7.71337450e-01 6.25476599e-01
-7.52215683e-01 2.75249094e-01 -1.21633291e-01 1.28625646e-01
1.47986397e-01 6.61652267e-01 -1.63741767e+00 1.55184448e-01
7.50551641e-01 1.67470992e-01 -3.95295769e-01 7.64593780e-01
-6.72433600e-02 1.24285281e+00 -4.07159120e-01 -1.82929233e-01
4.65155214e-01 5.60856611e-03 7.41207063e-01 1.72636557e+00
1.84904680e-01 -5.36359787e-01 2.67961472e-01 4.23146963e-01
-3.29177380e-01 6.14931226e-01 -8.60349476e-01 -5.82153082e-01
7.37887561e-01 1.33895051e+00 -2.68776625e-01 -3.72330219e-01
-4.55324501e-01 9.55517590e-01 2.85899997e-01 1.77254304e-01
-3.42500359e-01 -4.47320312e-01 1.06796777e+00 -3.90611589e-02
1.17051803e-01 -5.52160800e-01 2.82225758e-01 -8.04675817e-01
1.78806409e-01 -1.37525928e+00 -3.46814953e-02 -6.51621878e-01
-1.09320033e+00 8.60383511e-01 -6.78928941e-02 -6.83080673e-01
-8.63589585e-01 -7.27039874e-01 -5.79426467e-01 1.01929128e+00
-1.57455969e+00 -1.30620658e+00 3.15727741e-02 6.84825599e-01
1.09721315e+00 -6.50728405e-01 1.14527690e+00 4.16567951e-01
-4.92157370e-01 6.34036481e-01 9.42533240e-02 3.58879179e-01
9.30543423e-01 -1.05381191e+00 8.12651873e-01 7.76809216e-01
7.99160004e-01 3.47675711e-01 5.20281196e-01 -2.03349397e-01
-1.45564806e+00 -9.62784469e-01 1.39747667e+00 -1.09112836e-01
9.75391567e-01 -7.73235619e-01 -9.45763290e-01 3.60002309e-01
7.37440765e-01 -2.42684469e-01 7.14891970e-01 -5.83130633e-03
-1.23791993e-01 -2.13250756e-01 -6.83206856e-01 4.07860845e-01
6.58913791e-01 -9.64807689e-01 -5.76366127e-01 4.94489282e-01
1.02675629e+00 -2.51446784e-01 -6.52385712e-01 2.66436487e-01
2.89710730e-01 -6.85218573e-01 5.66458881e-01 -5.87582052e-01
-1.85092136e-01 -1.21386237e-02 -2.08967805e-01 -1.65535736e+00
2.95330673e-01 -7.80580223e-01 -1.51992207e-02 1.74323428e+00
6.60155714e-01 -4.09250766e-01 5.43089211e-01 7.99805298e-02
-4.46583867e-01 -5.34260213e-01 -1.37072527e+00 -7.14126825e-01
-7.33331442e-02 -9.61848557e-01 5.76989114e-01 9.49242413e-01
-1.93701819e-01 4.73959684e-01 -3.53045195e-01 1.18643763e-02
1.60759524e-01 3.87767851e-02 8.47243965e-01 -9.26550090e-01
-1.68169037e-01 -3.19004685e-01 -4.59219307e-01 -6.83639586e-01
6.06398284e-01 -9.10643101e-01 3.38753939e-01 -1.32911456e+00
-2.23844692e-01 -1.52517304e-01 1.28472848e-02 8.88496220e-01
2.02471718e-01 6.39654249e-02 -5.59416153e-02 3.06311250e-02
-2.69653231e-01 7.01740563e-01 5.60037315e-01 -3.27247083e-01
-2.19025373e-01 9.96572003e-02 -1.15802651e-02 7.36487389e-01
1.12877858e+00 -3.01775962e-01 -2.77168125e-01 -3.62107277e-01
4.74024052e-03 -2.81576723e-01 -4.67623957e-02 -9.63955581e-01
2.43027702e-01 1.38849363e-01 -2.58722544e-01 -8.39386165e-01
3.99300724e-01 -3.62790734e-01 -3.95363450e-01 1.29602349e-03
-4.00988430e-01 1.32593811e-01 2.95549154e-01 -1.53355077e-02
-6.55953705e-01 -2.59638935e-01 5.56221783e-01 -2.01121613e-01
-1.05093277e+00 1.13491237e-01 -1.05895770e+00 1.97383448e-01
5.32587767e-01 3.01609188e-01 -1.10221701e-02 -4.72219229e-01
-5.43722808e-01 -1.08891204e-02 -6.47368357e-02 9.65508759e-01
7.12255776e-01 -9.71407890e-01 -7.72004664e-01 3.88420552e-01
2.51243323e-01 4.33807075e-02 -1.81930616e-01 5.66783011e-01
-2.75302291e-01 7.79965103e-01 2.04299748e-01 -5.23857653e-01
-1.61131227e+00 2.34782174e-01 1.13444291e-01 -9.16522667e-02
-5.16851842e-01 1.00722373e+00 -4.89842892e-01 -1.01396167e+00
7.65347779e-01 -5.67770481e-01 -1.11549027e-01 1.70499399e-01
6.54751360e-01 1.78214863e-01 7.65703380e-01 -8.24063063e-01
-4.61466074e-01 -1.25370845e-01 -3.13806385e-01 -4.96946722e-01
1.27876198e+00 -5.54121546e-02 -2.83495218e-01 6.81066096e-01
1.14292157e+00 7.69765750e-02 -5.71257949e-01 -4.89428401e-01
5.34854949e-01 1.53408349e-01 2.66154766e-01 -7.64198601e-01
-7.46110737e-01 1.02331626e+00 5.51384568e-01 1.42443940e-01
8.42895567e-01 9.27088782e-02 1.13482130e+00 9.36166942e-01
3.10004145e-01 -1.44241393e+00 -2.07525805e-01 9.83860254e-01
1.01386821e+00 -1.26081121e+00 -6.05094969e-01 1.14168398e-01
-5.67727625e-01 8.94260108e-01 3.80459517e-01 4.51805830e-01
5.50335228e-01 5.54036677e-01 5.17825544e-01 8.19057450e-02
-7.00608075e-01 -2.45596811e-01 2.82609999e-01 4.96738106e-01
5.41755915e-01 1.23170972e-01 -1.02371395e-01 5.58393836e-01
-5.89427531e-01 -4.94589098e-02 2.82940358e-01 6.81988716e-01
-3.92999142e-01 -1.38362420e+00 -2.56925374e-01 2.97897696e-01
-5.40596187e-01 -6.35055482e-01 -9.19184446e-01 3.77698630e-01
-1.60878703e-01 1.02664232e+00 -2.83602495e-02 -3.62385988e-01
2.88092881e-01 5.80341280e-01 5.34209497e-02 -7.15920389e-01
-5.24846256e-01 -1.34898126e-01 4.00180012e-01 -4.17614788e-01
-3.07659507e-01 -5.51005304e-01 -9.59779739e-01 -6.71836063e-02
-2.63314694e-01 3.52703333e-01 1.17146981e+00 1.04689658e+00
3.70546013e-01 2.15531722e-01 5.86640179e-01 -8.99400055e-01
-6.82748854e-01 -1.54427338e+00 -5.84323704e-01 -2.56596971e-02
3.13361079e-01 -2.37365603e-01 -4.45940018e-01 3.13361771e-02] | [14.214993476867676, 6.767863750457764] |
0789fd8f-2ac9-4451-a9f8-8ecb8e0a89a9 | zero-day-ddos-attack-detection | 2208.14971 | null | https://arxiv.org/abs/2208.14971v1 | https://arxiv.org/pdf/2208.14971v1.pdf | Zero-day DDoS Attack Detection | The ability to detect zero-day (novel) attacks has become essential in the network security industry. Due to ever evolving attack signatures, existing network intrusion detection systems often fail to detect these threats. This project aims to solve the task of detecting zero-day DDoS (distributed denial-of-service) attacks by utilizing network traffic that is captured before entering a private network. Modern feature extraction techniques are used in conjunction with neural networks to determine if a network packet is either benign or malicious. | ['Troy Januchowski', 'Cameron Boeder'] | 2022-08-31 | null | null | null | null | ['network-intrusion-detection'] | ['miscellaneous'] | [ 0.02742825 -0.6965376 -0.20482226 -0.3060887 0.26681653 -0.9538455
0.63704973 0.05727097 -0.1252752 0.53742677 -0.54692006 -1.1157335
-0.13024506 -1.028467 0.3563434 -0.35056445 -0.5670662 0.50457066
0.6084674 -0.41747504 0.5749666 1.6527933 -0.9889158 0.18209064
-0.07714088 0.9836185 -0.8514949 1.0099425 -0.41733295 0.3778141
-1.1382065 0.05934506 0.7490645 -0.2971212 -0.6363153 -0.06884924
0.12792581 -0.5254995 -0.87716806 1.1636466 0.385359 -0.11665115
0.17156851 -1.7291014 0.0720154 0.17628145 -0.07818723 1.3762648
0.32722625 0.4956305 0.61531216 -0.2962962 0.6995387 1.3361169
0.26444417 0.48155448 -1.2316904 -0.9970499 -0.3146986 0.07007581
-1.0505065 -0.5764436 0.83190066 -0.23050664 1.207943 0.08007148
0.24006443 1.0514495 0.64675766 -0.14995626 0.91466445 -0.30783942
0.48895317 0.17725147 0.44213322 0.3508378 0.5657251 0.5492336
0.26066262 -0.586943 0.56840193 0.4510954 0.09991419 0.01971314
-0.6761505 1.0052834 0.33975914 0.83783185 -0.59718126 -0.18569447
0.8333554 1.129576 0.05181977 0.67233855 -0.7122124 -0.23944466
-0.641277 -0.04381576 1.3868072 0.33281466 0.78609437 0.7102287
0.6301398 0.2636839 0.17963354 0.6270431 0.27583227 -0.552253
-0.18459189 0.7923805 -0.48296127 -1.0945431 -0.37148726 -0.24309205
-0.3625308 0.48266602 0.16522038 -0.4480983 -1.037467 0.9926134
0.2892716 0.2822131 0.21610199 0.35837966 0.11636066 0.5115157
0.06521214 -0.02568254 1.1013168 -0.06406737 -0.41747224 -0.11232857
0.4737463 -0.8382969 0.03349898 0.44770133 -0.2966277 0.05514247
-1.0871806 0.98883414 -0.87419343 -1.008175 0.82032186 1.468563
-0.6853438 0.389167 -0.45035172 -0.69576144 0.42488098 0.6843815
-0.36043563 -0.10474348 -1.2022573 0.7625429 0.60323614 -0.52821136
-1.1884772 -0.47959954 -0.4408294 -0.01066252 0.48500633 0.01985461
0.88220763 -0.9139759 -1.1600918 0.6082424 0.2706968 -0.57973367
-0.09590792 0.46673745 -1.593591 0.66361177 -0.11994741 -0.25795844
1.1196421 -0.75133896 -0.7481381 -0.5715297 0.1109091 -0.9202804
-0.3743579 0.7917702 0.9437619 -0.35786712 0.20807907 -0.9258702
-0.35062644 -0.38196555 -0.3184132 -0.0516055 2.0453374 -0.02736149
1.1018745 -2.00912 -0.9646287 0.84705263 0.34528458 0.9509479
-0.0404559 0.67103654 -0.5216728 0.40558934 0.32997155 0.65140843
-0.32097316 0.44156066 -0.64945865 0.37737793 0.09878998 0.2865419
-0.6465054 0.0129069 0.26426408 0.1644467 -0.16302499 0.07896931
0.03190348 0.3560858 -0.6447209 1.105782 0.6269195 -0.03081163
0.22763033 0.17179932 -0.08948276 0.3605153 -1.1132525 0.43447798
-0.15308653 0.66934854 0.04321476 -1.053682 1.1605586 0.64290357
0.5673731 -0.6001969 0.44876844 0.55353457 0.6081827 -0.31893265
-0.22632855 -0.4228017 -0.00837797 0.7564065 0.15171275 0.42329255
0.42262727 0.20539644 2.0214243 -1.1387622 0.3558503 -0.13122682
0.82850647 0.0476784 0.8301143 0.7684587 -0.779574 -0.3551194
0.6831216 -0.9026053 -0.99212617 -1.3673205 -0.06976127 0.5404598
-0.29959375 0.08009756 -0.326184 -1.1251547 0.14869684 0.61367416
0.12023834 -0.39378667 -0.64893836 -0.51102155 1.0874308 0.05677329
0.5618204 -1.0031246 -0.2892201 0.80049056 0.6502493 -1.3641772
0.12978955 0.35559282 -0.952195 -1.5071439 0.25194895 -0.53145415
0.4187609 0.43344155 0.8773819 0.18542922 -0.9925952 0.49930793
-0.2680043 -0.4333184 -0.7733015 -0.03759379 0.47973418 0.09102648
1.0386928 -1.0091234 -0.27442184 0.4213816 -1.1323824 -1.4743959
0.48193756 0.16492768 -0.4760035 0.7169487 1.0107552 -0.91556007
0.7568722 -0.8346099 -0.62698823 -0.15158252 -0.8184421 -0.31751138
0.9026276 -0.6261367 -0.4517736 -0.3414285 -0.08041175 -0.4931356
-0.87261444 0.1259367 -0.2992063 -0.68902725 0.8884027 -0.04799238
0.13113299 -0.22275208 -0.25044256 0.8948607 0.0221011 -0.32054442
1.197647 0.69948995 0.43042997 -1.2462049 -0.36037433 -0.71869093
-0.43039295 -0.16593856 0.12414444 -0.29266986 -0.7454755 0.40920165
-0.9560263 0.41220626 -0.02479302 0.4990343 0.22363235 0.46486732
-0.80805755 -0.7251541 -0.27939674 -0.59763926 -0.16049439 0.4115015
-0.13632968 -1.086566 0.264669 0.02883006 0.8071839 0.23984891
1.0240476 -1.8270395 -0.42118412 -0.9722013 -0.20682514 0.580785
0.58543235 0.427542 -0.58176047 -0.39235887 0.34057587 0.2816259
0.19588879 -0.11941126 0.34144044 -0.21506184 -0.21434963 0.34904003
1.644019 1.0552098 0.89011455 0.6577611 0.1587103 0.49084038
0.03728775 0.5985123 -0.49086875 0.01359721 0.4864279 0.50597596
0.42840385 0.30526668 0.16418236 -0.08509424 0.572616 -0.35905236
-1.3202149 0.2840045 -1.0406703 -1.3623172 -0.08416489 1.7471521
-0.08437617 0.84792084 0.21612267 0.56474835 0.9708554 0.15142477
-0.4742285 -1.0801718 -0.07497402 0.57127976 0.5962065 0.05374987
-1.1005173 0.938319 6.8936 0.28171867 -1.6688149 -0.1689043
0.26180318 0.47804832 0.34512755 0.50228804 -0.83197916 0.42569768
1.6170065 -0.29046002 0.25895113 0.9503808 0.18575156 0.22349738
-0.62093306 0.57055444 -0.09928328 -1.1589904 0.330214 0.448022
0.385795 -0.00751453 -0.16387406 0.36384666 0.2332634 -0.7574338
-0.5459041 0.15787953 0.23058346 -1.0380177 0.845896 0.07796406
-0.7432922 -0.5593992 -0.07661822 -0.33260712 0.1516099 0.799258
-1.1235855 0.07182253 0.2841785 0.33969298 -0.3367784 1.0486207
-0.02687663 0.85227275 -0.5705491 0.20150162 0.53791106 0.08153996
1.097241 0.89176285 -0.0764814 0.08143811 0.57783735 0.51459146
0.06663915 -0.41678002 -1.0225608 -0.6864259 0.5711325 1.2260659
-1.0930729 -0.21744679 -0.3255461 0.6406931 -0.7769911 0.20931736
-0.11641321 -1.1284698 1.2542233 0.49072605 0.17500466 -0.62851524
-0.03170462 -0.8634002 -0.5276245 -1.0846084 0.58668536 -0.07334975
-1.6210055 0.7969679 -0.35852477 -1.1403807 -0.26484942 -0.75475645
-1.1230181 0.44929245 -1.3055899 -0.9193142 0.23993607 0.840808
0.21800393 -0.9787259 0.9081578 0.60945153 -0.5588705 0.22525619
-0.12625076 0.63790166 0.45684987 -0.6443838 0.40413186 1.0321345
-0.09099156 0.7192748 0.6637183 -0.87475455 -1.2577621 -0.7466234
0.7142287 0.01575359 1.137511 0.08968555 -0.8016241 0.646693
-0.17007762 0.5346745 0.798947 -0.22693127 -0.73151433 -0.3431033
-1.7300446 0.38785103 0.34230638 -0.7182636 -0.47827786 0.09503191
0.47644067 0.47627038 -0.7115068 0.24849385 0.1289903 -1.006152
1.0096176 -1.1461587 -0.50367254 -0.14926651 -0.35633704 -0.6041845
-0.03870776 -0.9054459 -0.17173535 1.1256886 -0.07420668 -1.3545991
0.8580265 0.30061427 0.4135978 -0.054146 -1.034997 -1.2141585
-0.38786754 -0.22708626 0.52443177 1.285917 -0.16844551 0.47807148
0.03605883 0.47711387 0.738956 -0.26547778 0.66299903 -1.796109
0.26638535 -0.49408364 -1.4886649 -0.21121967 0.20875087 -0.38088673
-0.7112706 -0.59131986 -0.6101753 -0.39947724 -0.63768977 0.27630293
0.77742106 0.2619422 0.01020573 0.16967571 -0.21065459 -0.13625236
0.20903362 0.05021013 -0.19501868 0.36956584 -0.2253191 0.7241977
1.3412313 -0.9417792 -0.22548138 0.56377107 -0.07520237 0.1508122
0.47738105 -1.3313047 0.090634 -0.4161348 0.08648297 -0.39454204
-0.06382241 -1.1928992 -0.03504717 0.9853936 0.25731412 0.34104362
0.11444733 0.5788056 -0.29817176 -0.23423222 1.2511455 -0.2246632
-0.9575029 0.56655633 -1.0349435 -0.15737313 1.4676764 -0.33439136
-0.42382294 -0.14384775 -0.86135 -0.13726135 0.38695413 0.42215866
0.7320111 -0.78372467 -0.3805072 0.6432507 -0.08819905 -1.1151386
-0.02689718 0.35912868 -1.0328616 0.57728904 -0.7820282 -0.26644826
-1.4506444 0.775484 0.4963145 -0.1981773 -0.67107713 0.23662788
-0.9050781 -0.12596494 -0.14156356 0.7744087 0.03780185 -0.09357197
0.87044024 0.74152845 0.09628722 -0.41565087 -0.7307273 -0.25045574
-0.54621893 0.04108486 1.1912999 0.18031414 -0.5535294 0.2658498
1.3839347 -0.22926289 -0.18602414 -0.3253434 0.61860543 -0.9552208
0.04770635 -0.634447 -1.103936 0.731906 0.8923569 0.83026063
0.9065224 -0.531743 1.1820524 0.82449937 0.45012102 -0.7919922
0.25854975 0.8565099 -0.19160469 -1.0552797 -0.10162283 -0.2438868
-0.01425369 1.7164619 0.559103 -0.6817083 1.2647954 0.3683996
0.2647655 -0.5673806 -0.7076466 -0.05636806 -0.5416851 0.873275
-0.14958781 -0.10188797 -0.12788008 -0.48532796 0.31171802 -0.33281854
0.9711099 1.3177849 -0.91014284 -1.5005386 -0.5209755 0.7140607
-0.9182721 0.11110353 -0.6471451 0.53260875 -0.30429763 1.2751467
0.0405576 -0.6171363 0.098489 0.22658193 -0.27385634 -0.6755984
-0.8174839 -0.35397798 0.00829975 -0.45664996 0.01762524 -0.2836535
-1.1979718 -1.036976 -0.18032204 0.2608685 1.1122508 0.8044342
0.39309838 0.40381804 1.337604 -0.1375516 -0.5694954 -0.4475596
-0.9166609 0.16164505 0.5266072 -0.50651526 -0.83035445 -0.37252212] | [5.26679801940918, 7.2314629554748535] |
a4fe9297-c052-4801-bbb8-dc5f2668a1df | deepref-a-framework-for-optimized-deep | null | null | https://aclanthology.org/2022.lrec-1.480 | https://aclanthology.org/2022.lrec-1.480.pdf | DeepREF: A Framework for Optimized Deep Learning-based Relation Classification | The Relation Extraction (RE) is an important basic Natural Language Processing (NLP) for many applications, such as search engines, recommender systems, question-answering systems and others. There are many studies in this subarea of NLP that continue to be explored, such as SemEval campaigns (2010 to 2018), or DDI Extraction (2013).For more than ten years, different RE systems using mainly statistical models have been proposed as well as the frameworks to develop them. This paper focuses on frameworks allowing to develop such RE systems using deep learning models. Such frameworks should make it possible to reproduce experiments of various deep learning models and pre-processing techniques proposed in various publications. Currently, there are very few frameworks of this type, and we propose a new open and optimizable framework, called DeepREF, which is inspired by the OpenNRE and REflex existing frameworks. DeepREF allows the employment of various deep learning models, to optimize their use, to identify the best inputs and to get better results with each data set for RE and compare with other experiments, making ablation studies possible. The DeepREF Framework is evaluated on several reference corpora from various application domains. | ['Sébastien Fournier', 'Bernard Espinasse', 'Adrian-Gabriel Chifu', 'Rinaldo Lima', 'Igor Nascimento'] | null | null | null | null | lrec-2022-6 | ['relation-classification'] | ['natural-language-processing'] | [-2.73268014e-01 3.79478514e-01 -4.90186036e-01 -3.02878261e-01
-3.19284022e-01 -4.08579081e-01 9.64584410e-01 3.61352473e-01
-6.04721248e-01 8.18602860e-01 1.53880715e-01 -5.63935101e-01
-3.83120120e-01 -1.14432287e+00 -4.54070628e-01 -2.99648792e-01
-2.18832240e-01 7.41670370e-01 2.39249781e-01 -3.61848027e-01
-3.61747406e-02 7.10975587e-01 -1.58561599e+00 4.77564633e-01
7.40900457e-01 9.60654676e-01 -2.22539213e-02 3.96306068e-01
-6.74308181e-01 8.48861516e-01 -6.50329590e-01 -6.81169152e-01
3.09902988e-02 2.08465457e-02 -1.21591496e+00 -8.17029357e-01
2.04688624e-01 2.04556864e-02 -3.00183445e-01 7.59283125e-01
5.98578990e-01 1.88784733e-01 7.17260540e-01 -1.15013516e+00
-7.51332223e-01 9.98891234e-01 -2.33293697e-01 3.66142124e-01
4.43742156e-01 -3.60123903e-01 1.10376620e+00 -8.18103433e-01
8.42752576e-01 1.28981769e+00 4.56684828e-01 3.40848446e-01
-8.90797138e-01 -8.30989778e-01 -1.40462011e-01 7.68635035e-01
-1.30488443e+00 -2.84246892e-01 6.01620436e-01 -2.17397124e-01
1.55906558e+00 3.00040394e-01 3.05313289e-01 1.33397126e+00
3.82764012e-01 1.01922798e+00 9.18911636e-01 -7.53366768e-01
5.58811799e-03 2.47009128e-01 6.61277354e-01 3.12936157e-01
2.99508125e-01 1.19479455e-01 -4.12874490e-01 -2.53927503e-02
3.20458263e-01 -2.86991805e-01 -9.23803523e-02 -4.49451245e-02
-9.22366023e-01 9.05202508e-01 2.63617277e-01 1.00814486e+00
-3.18071008e-01 -5.22304773e-01 4.58332896e-01 4.29136634e-01
5.62317789e-01 8.55830610e-01 -9.46007967e-01 6.08571544e-02
-8.28739047e-01 6.80825412e-01 1.43652666e+00 1.03401220e+00
5.89124858e-01 -5.10251045e-01 -5.47631741e-01 9.84565556e-01
2.24041566e-01 1.26431808e-01 5.37211299e-01 -4.91375387e-01
8.38603973e-01 9.20128226e-01 -1.95895985e-01 -1.21900833e+00
-8.72702539e-01 -3.86943072e-01 -1.18825459e+00 -2.69666255e-01
7.31241703e-02 -4.70034957e-01 -7.44070709e-01 1.38182926e+00
1.91569164e-01 1.11910760e-01 3.43906522e-01 5.80835700e-01
1.70329368e+00 8.46131146e-01 3.06580395e-01 -1.83566421e-01
1.47793567e+00 -1.00571454e+00 -1.13760853e+00 -2.82834917e-01
8.35062563e-01 -8.73973846e-01 6.57488346e-01 6.21655643e-01
-7.89223850e-01 -4.56361651e-01 -1.08636379e+00 -3.88026536e-01
-1.13755882e+00 3.94334167e-01 8.69305193e-01 4.50457662e-01
-7.07260311e-01 8.10446620e-01 -5.84355295e-01 -6.93471670e-01
3.09075624e-01 4.17719960e-01 -3.97668540e-01 1.98234558e-01
-2.01096630e+00 1.34427691e+00 9.20974553e-01 4.45781857e-01
-4.26048338e-01 -4.71054226e-01 -7.67515361e-01 1.50029063e-01
7.08979487e-01 -8.28114092e-01 1.11223984e+00 -3.63201082e-01
-1.45902765e+00 8.02947462e-01 -9.07377824e-02 -9.61721241e-01
1.42412469e-01 -8.00431967e-01 -6.49997473e-01 -3.03958625e-01
-1.07298262e-01 3.12000781e-01 3.53714406e-01 -8.30390275e-01
-5.73286891e-01 -3.09313953e-01 2.25811809e-01 1.39232036e-02
-4.03144985e-01 5.14176965e-01 -4.06897753e-01 -5.41649103e-01
-3.72689277e-01 -6.93501770e-01 -2.89722562e-01 -8.46264541e-01
-7.46156156e-01 -8.56362760e-01 6.59163117e-01 -5.90614676e-01
1.61947739e+00 -1.71066332e+00 1.94443524e-01 6.87902197e-02
2.38170803e-01 9.36230123e-01 -1.80965766e-01 6.30808115e-01
-2.66851008e-01 3.86178821e-01 5.10145277e-02 -1.50913119e-01
-2.89881211e-02 3.18060726e-01 -2.46955231e-01 -8.58023614e-02
3.33191067e-01 9.97807920e-01 -6.91252828e-01 -5.63225448e-01
2.40795091e-01 4.03679192e-01 -1.69786319e-01 4.37243253e-01
-3.59946251e-01 7.83190504e-02 -5.79107642e-01 4.41364199e-01
7.36958325e-01 -7.44887441e-02 2.79459149e-01 -4.19671893e-01
-1.59626678e-01 6.67735040e-01 -1.34382319e+00 1.37201500e+00
-4.77731436e-01 6.49269164e-01 -2.90494531e-01 -1.19066584e+00
1.22628570e+00 4.39088345e-01 3.96701902e-01 -6.91607237e-01
4.97258872e-01 2.08873197e-01 2.25835830e-01 -8.31061423e-01
6.57520950e-01 1.97798818e-01 1.48928851e-01 1.71401992e-01
5.94881237e-01 7.06462264e-02 6.52749062e-01 5.84053360e-02
1.11804473e+00 1.74892992e-01 6.72817051e-01 -1.35511503e-01
9.13318872e-01 -7.80866593e-02 4.75706011e-01 7.28622973e-01
5.25268055e-02 3.06814790e-01 4.11461234e-01 -7.47037232e-01
-6.53319478e-01 -5.02543509e-01 -3.73554856e-01 9.23981488e-01
-2.91375130e-01 -7.29886055e-01 -5.63518643e-01 -7.16217041e-01
-2.25364134e-01 8.85259330e-01 -4.78843272e-01 -6.98961504e-03
-5.85460663e-01 -1.04412568e+00 6.92579806e-01 3.06631118e-01
7.83833563e-01 -1.77204812e+00 -8.36870223e-02 4.72722352e-01
-2.60642260e-01 -1.45179331e+00 4.20386761e-01 5.30191183e-01
-7.42559671e-01 -1.17745841e+00 -3.81777972e-01 -7.96661615e-01
-1.27689093e-01 -1.85986772e-01 1.45846128e+00 -1.79023936e-01
-1.01737333e-02 -1.18217550e-01 -8.47771227e-01 -7.37088501e-01
-4.10055220e-01 9.84214425e-01 -7.14116544e-02 -3.23516041e-01
1.19128883e+00 -5.59155881e-01 -8.04206133e-02 -6.68890625e-02
-8.76646936e-01 -2.05054179e-01 8.67754936e-01 5.39466321e-01
4.70602840e-01 7.53976852e-02 6.25345647e-01 -1.33800030e+00
1.19513786e+00 -7.37292469e-01 -5.58261871e-01 4.09514904e-01
-8.38549674e-01 2.53305882e-01 7.06309497e-01 -1.77326277e-01
-1.02013183e+00 -3.32647771e-01 -5.45338392e-01 -3.88307683e-02
-5.61754286e-01 9.22152817e-01 -3.81245703e-01 2.65987277e-01
9.15519893e-01 -2.06963196e-01 -5.27971447e-01 -9.27138150e-01
5.26999295e-01 8.25601697e-01 1.39194548e-01 -5.24117529e-01
5.73559642e-01 -1.94773361e-01 -7.55147189e-02 -7.21496344e-01
-1.13543594e+00 -4.87677783e-01 -7.94094861e-01 3.56199980e-01
7.83108473e-01 -5.38746178e-01 -5.41433752e-01 2.39595056e-01
-1.63701105e+00 5.50130149e-03 -9.67766345e-02 4.61721957e-01
-7.37563372e-02 1.44979388e-01 -5.85504591e-01 -4.26963329e-01
-8.62928867e-01 -9.95355546e-01 7.89306760e-01 5.40865719e-01
-3.26070011e-01 -1.09422755e+00 4.42872524e-01 3.00204664e-01
2.95153081e-01 7.53306299e-02 1.00183058e+00 -1.27879834e+00
-1.60289496e-01 -1.93355361e-03 -3.19408834e-01 5.92869103e-01
-1.20646209e-02 7.60810077e-02 -1.02448177e+00 2.15316802e-01
-1.98554993e-01 -2.50094980e-01 8.94188643e-01 1.78122416e-01
1.27180755e+00 -3.44971210e-01 -8.23337436e-01 5.37312031e-01
1.17088115e+00 3.47508371e-01 1.00342298e+00 7.91407704e-01
6.23123765e-01 7.91753650e-01 7.25274682e-01 2.32374808e-03
4.83549863e-01 7.05284536e-01 1.97281271e-01 -7.69542828e-02
-4.12432626e-02 1.76944286e-02 2.14339212e-01 8.56717765e-01
-3.55292380e-01 -4.67357963e-01 -1.14892924e+00 4.84280378e-01
-1.83988273e+00 -8.11921537e-01 -3.15458655e-01 1.77679956e+00
9.08232093e-01 1.64083406e-01 -2.10902885e-01 1.52051672e-01
4.34906095e-01 2.61995912e-01 -1.45824239e-01 -9.37444091e-01
-1.79600164e-01 9.44275677e-01 2.53952414e-01 2.65156418e-01
-1.54708469e+00 1.20731997e+00 5.83870745e+00 1.05042088e+00
-9.13496494e-01 3.68366055e-02 4.89812165e-01 3.18745375e-01
-5.85537404e-02 1.16650343e-01 -1.38006556e+00 -3.87906097e-02
1.19464791e+00 -1.35775164e-01 1.70903832e-01 8.34701777e-01
-5.61297387e-02 1.15479335e-01 -1.20180452e+00 8.38926017e-01
9.93147213e-03 -1.44342172e+00 1.00938834e-01 -7.23317042e-02
3.89434993e-01 2.00809285e-01 -4.64282721e-01 1.00920856e+00
4.03944314e-01 -1.09951556e+00 -9.33030024e-02 5.27861178e-01
2.69703418e-01 -7.83434927e-01 1.40078413e+00 4.59363312e-01
-9.37048793e-01 -2.43619047e-02 -5.12021482e-01 -1.10669509e-01
-1.34496316e-01 9.95283544e-01 -9.98480499e-01 1.24878764e+00
9.37422454e-01 7.29457796e-01 -7.34437525e-01 9.10618603e-01
-5.97871959e-01 6.87393785e-01 -3.20049495e-01 -3.97886187e-01
1.09996311e-01 -1.93517536e-01 5.94099820e-01 1.48703694e+00
1.22735359e-01 -4.50655781e-02 -7.00278059e-02 8.33506525e-01
-2.85626948e-01 7.51703143e-01 -6.64918959e-01 -1.20321035e-01
4.51399326e-01 1.58524823e+00 -3.23985487e-01 -2.76680022e-01
-4.57867771e-01 4.06329185e-01 5.68586409e-01 3.07513863e-01
-6.15591824e-01 -5.88938534e-01 3.32475215e-01 -2.64648926e-02
6.79532737e-02 -1.45687342e-01 -1.52594432e-01 -1.27450991e+00
-1.08511575e-01 -1.07377970e+00 6.04644895e-01 -3.91402394e-01
-1.37621748e+00 9.99512374e-01 3.18449080e-01 -1.03764737e+00
-5.19492090e-01 -6.89094901e-01 -3.29073519e-01 8.57647836e-01
-1.73989165e+00 -1.02390838e+00 -1.57483574e-02 3.96366298e-01
2.48819038e-01 -4.84984696e-01 1.10452771e+00 6.76669478e-01
-8.76231015e-01 5.78599930e-01 -9.41333175e-02 4.83012527e-01
8.96774292e-01 -1.25214410e+00 4.16464090e-01 7.62440383e-01
6.42180085e-01 8.18981051e-01 5.37377357e-01 -4.07228619e-01
-1.05877376e+00 -1.07065046e+00 1.57928979e+00 -4.83210683e-01
7.38242745e-01 -4.14009780e-01 -9.59840655e-01 6.73647702e-01
6.06872022e-01 -8.27036649e-02 7.16563523e-01 8.44290018e-01
-9.30588469e-02 -3.38328302e-01 -9.57538307e-01 5.70245743e-01
7.87294984e-01 -1.02719337e-01 -1.01109147e+00 2.59973556e-01
7.54944742e-01 -4.80559200e-01 -1.10817695e+00 8.85779083e-01
3.41495335e-01 -8.52630854e-01 9.10252571e-01 -7.28160739e-01
5.22472441e-01 -5.39768413e-02 2.62763262e-01 -1.24181199e+00
-3.12359124e-01 -4.80932444e-01 -6.90716028e-01 1.75200725e+00
8.03544581e-01 -6.63989484e-01 5.49431920e-01 2.87322104e-01
1.06216870e-01 -1.07844400e+00 -7.52678454e-01 -6.58579648e-01
3.29178125e-01 -5.80207944e-01 8.14320147e-01 1.00881302e+00
-3.60362768e-01 9.41868901e-01 -1.58076659e-01 5.29993442e-04
-5.94489872e-02 2.49532968e-01 1.01959693e+00 -1.48868275e+00
-1.13450080e-01 -4.36428547e-01 -2.74589628e-01 -9.40563977e-01
3.09924990e-01 -1.06077743e+00 -2.89890707e-01 -1.80488658e+00
-1.25086904e-02 -5.70077717e-01 -4.70841944e-01 8.47582996e-01
-1.11166187e-01 -2.97926515e-01 3.80960219e-02 5.05349673e-02
-4.47268963e-01 5.64762354e-01 1.10361743e+00 -2.84028560e-01
-4.18703437e-01 1.25894487e-01 -7.68319845e-01 8.05923581e-01
8.16450059e-01 -7.46756852e-01 -2.42645130e-01 -2.84339756e-01
5.95029533e-01 -3.10260117e-01 -1.54065173e-02 -9.38506424e-01
2.09257573e-01 7.02556148e-02 2.35949859e-01 -8.11711669e-01
-1.51681840e-01 -8.05601895e-01 -2.27971360e-01 -3.54590192e-02
-2.93406188e-01 -1.87264815e-01 2.98869371e-01 -7.77584985e-02
-6.93720579e-01 -5.02929032e-01 3.06340247e-01 -4.08607945e-02
-7.59112000e-01 3.64366680e-01 -1.17565975e-01 1.69243917e-01
8.60287845e-01 3.20038557e-01 -2.81415284e-01 -2.18230225e-02
-8.15487444e-01 3.11851740e-01 -5.39981067e-01 8.78705382e-01
2.96878874e-01 -1.12023723e+00 -8.29354823e-01 -1.06140442e-01
4.76631150e-02 2.35944271e-01 -9.91803035e-02 4.82193500e-01
-4.47185546e-01 8.67534339e-01 8.82869773e-03 -1.54505700e-01
-1.04449594e+00 7.57813394e-01 2.56374449e-01 -1.28818512e+00
-5.42734444e-01 6.86412454e-01 -2.41346985e-01 -8.17545474e-01
2.23808125e-01 -6.82836413e-01 -1.25695109e+00 3.34178329e-01
7.04321325e-01 2.57050455e-01 5.61892092e-01 -2.32520700e-01
-4.71016675e-01 3.64567101e-01 -4.25938994e-01 2.90578157e-01
1.50142777e+00 2.59136170e-01 -3.33943129e-01 2.50942856e-01
1.02915704e+00 3.78267355e-02 -2.61373192e-01 -3.56987268e-01
4.82379824e-01 1.08541586e-01 5.00794910e-02 -9.20410156e-01
-1.06567454e+00 8.89507174e-01 4.87550855e-01 5.07213414e-01
1.07442951e+00 -6.39021322e-02 8.07881176e-01 7.12640822e-01
3.28194201e-01 -1.01310146e+00 -5.76439261e-01 9.94075358e-01
1.01006234e+00 -1.16443980e+00 2.04891473e-01 -4.78194624e-01
-2.39221409e-01 1.36722553e+00 5.87319374e-01 -1.56698436e-01
1.05802703e+00 5.61236739e-01 -1.26374111e-01 -3.31260562e-01
-7.44440377e-01 -4.96860296e-01 5.91512918e-01 6.33443296e-01
9.11161184e-01 -2.09782342e-03 -9.99859750e-01 1.19727886e+00
-4.13826674e-01 3.99416953e-01 1.61106184e-01 5.40305734e-01
-8.69198442e-02 -1.87731981e+00 -8.77603292e-02 6.80274785e-01
-6.02881312e-01 -3.55914593e-01 -6.47786140e-01 1.01668608e+00
5.24692714e-01 1.10475850e+00 -3.14497560e-01 -5.67018151e-01
6.44295275e-01 1.59033105e-01 3.73961538e-01 -9.24056292e-01
-9.23368931e-01 -5.12467802e-01 6.59339726e-01 -4.27670956e-01
-7.27579951e-01 -3.51515889e-01 -9.84951735e-01 -1.62030458e-01
-4.43995535e-01 2.87229121e-01 4.02438283e-01 1.37277329e+00
3.36804926e-01 5.82324088e-01 2.32375696e-01 -3.60994428e-01
-3.83670293e-02 -1.38645887e+00 -3.14347595e-01 1.41404659e-01
-1.79234952e-01 -8.00143480e-01 -1.35275990e-01 -2.60194898e-01] | [9.358119010925293, 8.781594276428223] |
de7f7084-c770-4110-bd3b-c28f4839e7f7 | sparse-graphical-representation-based | 1607.00137 | null | http://arxiv.org/abs/1607.00137v1 | http://arxiv.org/pdf/1607.00137v1.pdf | Sparse Graphical Representation based Discriminant Analysis for Heterogeneous Face Recognition | Face images captured in heterogeneous environments, e.g., sketches generated
by the artists or composite-generation software, photos taken by common cameras
and infrared images captured by corresponding infrared imaging devices, usually
subject to large texture (i.e., style) differences. This results in heavily
degraded performance of conventional face recognition methods in comparison
with the performance on images captured in homogeneous environments. In this
paper, we propose a novel sparse graphical representation based discriminant
analysis (SGR-DA) approach to address aforementioned face recognition in
heterogeneous scenarios. An adaptive sparse graphical representation scheme is
designed to represent heterogeneous face images, where a Markov networks model
is constructed to generate adaptive sparse vectors. To handle the complex
facial structure and further improve the discriminability, a spatial
partition-based discriminant analysis framework is presented to refine the
adaptive sparse vectors for face matching. We conducted experiments on six
commonly used heterogeneous face datasets and experimental results illustrate
that our proposed SGR-DA approach achieves superior performance in comparison
with state-of-the-art methods. | ['Chunlei Peng', 'Nannan Wang', 'Jie Li', 'Xinbo Gao'] | 2016-07-01 | null | null | null | null | ['heterogeneous-face-recognition'] | ['computer-vision'] | [ 6.57004297e-01 -5.18165052e-01 7.52678700e-03 -3.48710239e-01
-5.54864705e-01 -3.17436904e-01 5.17587543e-01 -6.76364183e-01
2.48302639e-01 4.48900342e-01 -3.21911153e-04 2.93622375e-01
-2.97781110e-01 -5.65428376e-01 -3.22295338e-01 -9.44319010e-01
2.37721741e-01 1.94480985e-01 -1.47041723e-01 8.72909650e-02
4.68056798e-01 8.68321955e-01 -2.12111712e+00 4.24528778e-01
6.19317710e-01 1.02460837e+00 1.43235892e-01 2.86931723e-01
-2.20448658e-01 5.78824341e-01 -4.27837193e-01 -3.92353177e-01
4.94645417e-01 -3.36886555e-01 6.15647361e-02 5.53388894e-01
8.73546481e-01 -3.12036633e-01 -4.57098544e-01 1.33527851e+00
4.85648483e-01 5.89496233e-02 8.09125245e-01 -1.26451075e+00
-9.73143041e-01 8.97460952e-02 -9.40410614e-01 -1.17814243e-01
6.31931901e-01 -1.46728173e-01 2.40361199e-01 -1.30197930e+00
6.12104654e-01 1.75597763e+00 6.46800041e-01 5.58481634e-01
-1.29571462e+00 -9.64084804e-01 1.25797331e-01 2.11415976e-01
-1.85049677e+00 -9.18311298e-01 1.31275475e+00 -3.89686257e-01
2.48076841e-01 3.73365968e-01 3.40877116e-01 1.08711386e+00
1.96804609e-02 4.73245680e-01 1.22089231e+00 -4.74336714e-01
1.74148947e-01 1.26202196e-01 -2.81917542e-01 9.38098431e-01
4.01326805e-01 -7.13182544e-06 -7.37428308e-01 -6.84013903e-01
1.00517702e+00 5.34353256e-01 -2.13475928e-01 -4.83649075e-01
-8.09937358e-01 5.31618476e-01 -1.00422479e-01 3.47107738e-01
-4.45619524e-01 -3.75704348e-01 3.39875370e-02 1.93972126e-01
2.58136600e-01 -3.64825815e-01 2.14178190e-01 3.47161531e-01
-1.24531472e+00 -2.67679971e-02 4.82303232e-01 1.07192183e+00
7.59598792e-01 3.57045352e-01 -2.36715019e-01 1.20233476e+00
7.25536525e-01 7.63905823e-01 4.48636264e-01 -9.26348567e-01
4.93999869e-01 6.71373904e-01 -7.99075663e-02 -1.57070541e+00
1.32746100e-01 -1.54758289e-01 -1.20955467e+00 2.52092212e-01
2.18334183e-01 2.45687649e-01 -7.51394808e-01 1.32469344e+00
3.21682841e-01 5.68038106e-01 1.49364918e-01 8.11403871e-01
7.90356755e-01 6.20638430e-01 -5.01988009e-02 -2.38283217e-01
1.24679554e+00 -6.14783645e-01 -7.26481795e-01 1.76342219e-01
-1.30599305e-01 -1.13540018e+00 6.61196232e-01 4.21556324e-01
-1.00146437e+00 -8.47827911e-01 -8.92874181e-01 4.82965559e-01
8.89763683e-02 6.38292670e-01 2.16044962e-01 8.84723902e-01
-1.01838970e+00 3.31552595e-01 -4.33263928e-01 -3.30118954e-01
4.24835861e-01 3.38008344e-01 -6.67611659e-01 -4.36313540e-01
-7.36084640e-01 2.54193395e-01 -4.36142348e-02 4.55114394e-01
-8.12959552e-01 -6.03511035e-01 -9.57535386e-01 -2.22201389e-03
1.78296752e-02 -4.60501850e-01 5.49157619e-01 -1.16095996e+00
-1.66763198e+00 8.16923082e-01 -3.49563479e-01 2.80023336e-01
3.03811491e-01 2.89089620e-01 -6.41490757e-01 4.52124596e-01
-1.20296493e-01 2.73001283e-01 1.48055136e+00 -1.54206145e+00
-4.10006903e-02 -7.91108727e-01 -4.78932917e-01 1.83791161e-01
-4.85610306e-01 3.47140372e-01 -5.09818017e-01 -9.08447564e-01
3.72764617e-01 -1.01808310e+00 2.27665585e-02 3.48285794e-01
-2.29315460e-01 8.16071704e-02 1.24210775e+00 -7.63048828e-01
1.12679780e+00 -2.22699261e+00 9.45471600e-02 6.68186486e-01
-8.67928192e-02 4.45400745e-01 -3.10973793e-01 4.01090443e-01
-1.18480533e-01 -4.11735982e-01 -2.99320340e-01 -3.74870986e-01
-2.27431372e-01 -1.16494536e-01 4.30096649e-02 7.40232170e-01
1.59061715e-01 1.49722993e-01 -4.98816997e-01 -6.72845840e-01
1.91745892e-01 8.96884561e-01 -4.00705576e-01 3.19461107e-01
3.63308549e-01 4.65618223e-01 -6.97387755e-01 1.29721236e+00
1.20801699e+00 -3.03775501e-02 4.56639469e-01 -5.38815677e-01
7.58215114e-02 -7.93075323e-01 -1.61443353e+00 1.70250428e+00
-3.56418043e-01 3.00465405e-01 5.11476755e-01 -9.10788417e-01
1.36770391e+00 4.51502323e-01 4.94802654e-01 -4.28295940e-01
6.10685814e-03 1.59421548e-01 -2.93984652e-01 -4.19055283e-01
3.23185682e-01 -8.13588277e-02 3.92355591e-01 1.63880393e-01
-1.23891711e-01 1.20997407e-01 -2.51614958e-01 -1.30178481e-01
7.85192251e-01 1.49938449e-01 6.23655170e-02 -3.28565598e-01
1.21620429e+00 -5.21095097e-01 6.36433005e-01 4.23816919e-01
-1.59416303e-01 8.68586421e-01 4.27755117e-02 -4.63165343e-01
-8.57608378e-01 -8.10308039e-01 -2.62945950e-01 6.40434086e-01
2.71895170e-01 -2.46789292e-01 -7.15312421e-01 -2.93081522e-01
1.12481415e-01 -5.49294837e-02 -4.02110606e-01 -5.88685386e-02
-3.79043728e-01 -4.66040522e-01 5.02526760e-01 2.13112101e-01
7.59968340e-01 -8.00653815e-01 -9.22582895e-02 1.86932217e-02
2.99327940e-01 -9.55278099e-01 -4.65645969e-01 -8.58675003e-01
-7.59278953e-01 -1.08143234e+00 -1.17014110e+00 -1.01044428e+00
9.70481098e-01 6.63820565e-01 6.46965444e-01 2.06452653e-01
-5.10075390e-01 6.00008667e-01 -1.48218989e-01 2.12409243e-01
-8.58691484e-02 -4.94019300e-01 2.49937862e-01 9.38626051e-01
2.58398503e-01 -6.23415887e-01 -7.78888345e-01 4.98249948e-01
-1.03565979e+00 -1.32547289e-01 7.39492536e-01 1.02931321e+00
3.92236859e-01 1.85929632e-04 3.97116959e-01 -8.36430490e-01
4.24796283e-01 -5.43173850e-01 -7.04494596e-01 6.07267618e-01
-3.66200984e-01 -5.89090660e-02 6.46366000e-01 -5.42438209e-01
-1.51033998e+00 2.96239376e-01 1.65044665e-01 -8.17049026e-01
-2.77753294e-01 2.96888947e-01 -4.24370944e-01 -7.52788961e-01
2.65874326e-01 4.97779727e-01 2.75251776e-01 -5.16326845e-01
-1.05378985e-01 9.98198867e-01 4.39318597e-01 -8.20050955e-01
9.99330103e-01 5.18771350e-01 1.69564009e-01 -1.16352856e+00
-5.59554994e-02 -3.66599202e-01 -3.70457411e-01 -3.53021801e-01
4.57795441e-01 -1.08894622e+00 -7.09733605e-01 8.12012434e-01
-9.64758217e-01 3.76460224e-01 5.36413789e-01 4.30188298e-01
-3.13126296e-01 7.23179758e-01 -4.76178378e-01 -1.12968373e+00
-2.92624861e-01 -1.19732988e+00 1.39727199e+00 4.35360491e-01
1.77230358e-01 -6.27010703e-01 -5.75265288e-02 5.80667078e-01
4.80708838e-01 1.56888485e-01 7.71148384e-01 -1.81626141e-01
-5.07434785e-01 -9.09050852e-02 -3.44743222e-01 2.29865387e-01
3.99708331e-01 3.38557363e-01 -1.03828263e+00 -4.83610719e-01
-1.82841700e-02 -8.29750448e-02 3.98387969e-01 6.56317621e-02
1.36290550e+00 -3.91361266e-01 -3.85474175e-01 6.31910264e-01
1.47346652e+00 4.99308407e-01 7.40095675e-01 -1.03779733e-01
7.21004486e-01 8.13438535e-01 4.78310853e-01 8.27182055e-01
-3.56208645e-02 8.98861349e-01 -1.23156600e-01 -1.20638348e-02
-7.62355924e-02 -2.31921241e-01 4.45506334e-01 5.91012716e-01
-1.25792846e-01 -1.92651562e-02 -4.81507152e-01 2.16148198e-01
-1.64810228e+00 -1.21994138e+00 1.91358432e-01 2.17891932e+00
5.53000987e-01 -5.30293107e-01 -1.05849117e-01 2.42374033e-01
1.27737069e+00 2.49655455e-01 -3.64745140e-01 -1.33782169e-02
-2.39391536e-01 3.15634310e-01 1.12309463e-01 3.00685197e-01
-7.65566885e-01 6.07114792e-01 5.84766102e+00 9.83499408e-01
-1.12596560e+00 -8.26300755e-02 6.06488228e-01 1.61630377e-01
-4.15843070e-01 -1.65393278e-01 -6.63184404e-01 6.69223130e-01
4.78481948e-01 4.75400463e-02 4.92660582e-01 8.58485699e-01
1.03064179e-01 2.07087919e-01 -7.59055197e-01 1.75381970e+00
6.39796317e-01 -1.10591197e+00 3.30488592e-01 -2.43258085e-02
9.17103648e-01 -7.84086347e-01 4.22632217e-01 -2.28499740e-01
-2.18647614e-01 -9.81164932e-01 5.64882636e-01 7.26183295e-01
1.05144620e+00 -5.90561748e-01 3.75427842e-01 -6.01254925e-02
-1.40662694e+00 1.63424686e-02 -6.35361373e-01 3.40171456e-01
-1.48959011e-01 4.21216398e-01 -3.76956284e-01 6.22189701e-01
5.01794577e-01 7.46608436e-01 -5.65733314e-01 7.84001291e-01
3.96186948e-01 3.22748601e-01 -9.85390544e-02 2.94835418e-01
-1.09470807e-01 -7.06959724e-01 5.07105708e-01 8.50323021e-01
7.36515880e-01 2.67060608e-01 4.39699776e-02 8.90084445e-01
-3.61659825e-02 2.24211350e-01 -1.00998616e+00 1.06573507e-01
5.20756960e-01 1.39836299e+00 -4.86321718e-01 -3.05802852e-01
-7.65810013e-01 9.99693036e-01 -9.03870091e-02 4.88744766e-01
-5.08771360e-01 -1.41323268e-01 5.87446451e-01 5.31876758e-02
3.24693352e-01 -2.18777820e-01 2.24929005e-01 -1.39403653e+00
1.19000115e-01 -1.14824533e+00 2.72657722e-01 -7.51195192e-01
-1.46906149e+00 8.32559109e-01 -8.36523846e-02 -1.42932391e+00
-7.98011944e-02 -5.74716270e-01 -4.72690970e-01 1.04767525e+00
-1.44424486e+00 -1.33708763e+00 -5.78347027e-01 1.08151948e+00
6.30431175e-01 -8.12654197e-01 7.00950265e-01 5.14173567e-01
-7.07788169e-01 7.59513199e-01 3.36542457e-01 -4.32022847e-02
6.78942263e-01 -3.34632516e-01 -2.39158243e-01 8.21121633e-01
6.98000416e-02 7.76461959e-01 2.72928923e-01 -6.64267719e-01
-2.01117277e+00 -9.09720361e-01 3.83459151e-01 9.33377892e-02
1.87843040e-01 -1.88887134e-01 -8.40873599e-01 3.04561436e-01
1.77131146e-02 3.24476689e-01 9.42193747e-01 -3.87271047e-01
-4.99409676e-01 -4.31150287e-01 -1.57810438e+00 5.06592572e-01
1.02837086e+00 -8.02847207e-01 -3.22180420e-01 1.50789060e-02
-3.16512078e-01 -2.62622647e-02 -9.80447292e-01 4.76182431e-01
1.17916191e+00 -1.07855487e+00 1.05699146e+00 -2.73895234e-01
-5.17038777e-02 -6.30857170e-01 -5.39541721e-01 -7.99762666e-01
-4.23377663e-01 -6.54891133e-01 6.22020848e-02 1.48799491e+00
4.61216364e-03 -5.82084894e-01 9.83099282e-01 6.66879237e-01
2.97558367e-01 -3.28675270e-01 -8.12098980e-01 -7.04705417e-01
-6.73426688e-01 2.82778800e-01 7.20961392e-01 9.14445281e-01
-4.39353377e-01 -1.60236672e-01 -5.10346711e-01 3.65831554e-01
1.12226260e+00 3.09700698e-01 5.68258584e-01 -1.36294067e+00
-9.77252275e-02 -3.58734578e-02 -8.44083190e-01 -6.22915447e-01
5.75499773e-01 -6.08228624e-01 -3.20493758e-01 -8.25534523e-01
3.36467803e-01 -6.02229238e-01 -2.37202600e-01 6.30289018e-02
2.06836909e-02 6.42997086e-01 2.37334490e-01 4.40066606e-01
-3.14457983e-01 7.57601142e-01 1.04599190e+00 -4.34016734e-01
-3.75153311e-02 -1.13742232e-01 -4.75485623e-01 5.72339773e-01
4.72481579e-01 -2.13753551e-01 -5.40623844e-01 -3.45747620e-01
-4.14439708e-01 3.18781316e-01 2.73944438e-01 -1.22617257e+00
3.42021972e-01 -2.27329314e-01 7.41395235e-01 -2.86474258e-01
5.58551073e-01 -1.16325164e+00 8.39183688e-01 2.83114314e-01
-1.37791857e-01 1.73866913e-01 5.95484376e-02 6.21807694e-01
-5.70503712e-01 -1.19334713e-01 9.34846580e-01 -6.27246872e-02
-6.09842062e-01 4.96952742e-01 -2.71434575e-01 -4.94433761e-01
1.05308950e+00 -6.35437608e-01 -1.76975504e-01 -2.45989189e-01
-4.52668786e-01 -4.49330121e-01 5.98910153e-01 3.54389936e-01
1.06508648e+00 -1.85988510e+00 -7.69151032e-01 8.75718832e-01
2.23752707e-01 -5.46746671e-01 6.22196078e-01 6.15664959e-01
-7.08698392e-01 2.50227422e-01 -5.05237043e-01 -5.80319643e-01
-1.61994147e+00 4.54357535e-01 9.27163810e-02 2.38572150e-01
-4.69061553e-01 5.18940985e-01 2.23034114e-01 4.53409273e-03
1.45447508e-01 1.17010064e-01 -2.35260010e-01 -3.56202386e-02
6.99442685e-01 3.87902081e-01 1.35383606e-01 -1.05939889e+00
-4.43798572e-01 1.09814775e+00 1.40177175e-01 4.37955260e-02
1.14617848e+00 -1.28547192e-01 -2.57404357e-01 -7.94367790e-02
1.23638833e+00 1.52782097e-01 -1.17524791e+00 -2.58724809e-01
-2.40557015e-01 -1.14702356e+00 -8.63454491e-02 -2.52596140e-01
-1.38807964e+00 5.81569970e-01 9.12118375e-01 -2.01414943e-01
1.26377463e+00 -4.51277643e-01 4.81598884e-01 2.66874462e-01
6.42611802e-01 -7.98453450e-01 -2.38723848e-02 -1.02822043e-01
1.02797198e+00 -1.10533345e+00 -1.25252083e-01 -6.58499539e-01
-4.79918689e-01 1.27192020e+00 5.86624444e-01 -1.31108239e-01
7.27233231e-01 2.00369760e-01 -1.11990690e-01 -8.00910965e-02
-4.26958948e-01 2.51875103e-01 3.75667781e-01 6.70795679e-01
3.60378623e-01 -2.47508287e-01 -7.44881853e-02 3.95251215e-01
4.68419492e-01 -8.94614123e-03 3.09856832e-02 1.05566072e+00
-1.73901990e-01 -1.19532013e+00 -1.07117462e+00 3.21263850e-01
-3.78429532e-01 1.04764163e-01 -2.99213171e-01 2.52107382e-01
-2.14515161e-02 1.01584351e+00 -9.42018777e-02 -4.10585046e-01
1.38529539e-01 8.41543376e-02 6.93869352e-01 -4.49344069e-01
-2.59773254e-01 4.46713120e-02 -2.68682063e-01 -5.43514073e-01
-7.25896060e-01 -6.81672990e-01 -5.11588812e-01 -3.53940248e-01
-1.94082037e-01 1.35504484e-01 5.56763709e-01 4.41143900e-01
6.03655338e-01 -9.69244465e-02 1.03117609e+00 -1.06401050e+00
-4.33893204e-01 -7.00008810e-01 -9.91321504e-01 5.02779603e-01
1.23963930e-01 -9.33524549e-01 -4.08953190e-01 3.00325722e-01] | [12.849709510803223, 0.3983350396156311] |
f8007c49-5a36-4052-99fb-63e56a189f4d | a-comprehensive-review-of-sign-language | 2204.03328 | null | https://arxiv.org/abs/2204.03328v1 | https://arxiv.org/pdf/2204.03328v1.pdf | A Comprehensive Review of Sign Language Recognition: Different Types, Modalities, and Datasets | A machine can understand human activities, and the meaning of signs can help overcome the communication barriers between the inaudible and ordinary people. Sign Language Recognition (SLR) is a fascinating research area and a crucial task concerning computer vision and pattern recognition. Recently, SLR usage has increased in many applications, but the environment, background image resolution, modalities, and datasets affect the performance a lot. Many researchers have been striving to carry out generic real-time SLR models. This review paper facilitates a comprehensive overview of SLR and discusses the needs, challenges, and problems associated with SLR. We study related works about manual and non-manual, various modalities, and datasets. Research progress and existing state-of-the-art SLR models over the past decade have been reviewed. Finally, we find the research gap and limitations in this domain and suggest future directions. This review paper will be helpful for readers and researchers to get complete guidance about SLR and the progressive design of the state-of-the-art SLR model | ['Prof. Partha Pratim Roy', 'Dr. M. Madhiarasan'] | 2022-04-07 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 4.03927773e-01 -4.06255037e-01 -7.60131896e-01 -4.13906664e-01
-4.46455330e-01 -3.67904425e-01 5.80354393e-01 -8.42449129e-01
-5.50497651e-01 3.98190469e-01 3.11928451e-01 -3.28654379e-01
1.57650143e-01 -1.10485166e-01 4.66359966e-02 -5.75862408e-01
1.26237437e-01 -1.93254724e-01 3.55972797e-01 -2.32838374e-02
4.01927382e-01 6.70856535e-01 -1.75321472e+00 3.58311623e-01
7.31633604e-01 6.43214047e-01 -4.80492366e-03 7.80412555e-01
-2.53592253e-01 1.19200492e+00 -7.44353771e-01 -2.76153594e-01
-1.37935597e-02 -5.67095041e-01 -5.14503598e-01 2.94608116e-01
7.39205837e-01 -3.72467548e-01 -8.73212993e-01 8.83297324e-01
7.59282947e-01 -6.17126524e-02 6.17373347e-01 -1.48889792e+00
-1.09507632e+00 6.91566169e-02 -5.55303872e-01 8.82463343e-03
6.22724771e-01 4.16178137e-01 5.94743073e-01 -7.17814445e-01
6.75545692e-01 1.18477750e+00 4.50890005e-01 1.31046093e+00
-4.74827528e-01 -6.55001104e-01 3.57659966e-01 4.48845774e-01
-1.26666343e+00 -4.06644464e-01 6.87291443e-01 -3.15675557e-01
7.44035780e-01 6.41080201e-01 7.62797058e-01 1.20735919e+00
-2.63545603e-01 1.65705442e+00 1.55658102e+00 -6.83176875e-01
-1.37578785e-01 -2.88004521e-02 5.99357843e-01 6.62759840e-01
3.78822982e-01 -5.36284633e-02 -9.06537056e-01 1.75923228e-01
9.06264484e-01 -2.60666497e-02 -4.60476041e-01 -1.77706659e-01
-1.13338137e+00 3.22032064e-01 -1.80697103e-03 6.52758598e-01
-8.96804929e-02 1.81309953e-02 2.51781583e-01 1.58507198e-01
-3.43505830e-01 1.09371372e-01 -1.49462059e-01 -5.04017830e-01
-7.45204210e-01 -2.48741694e-02 6.31080747e-01 7.81077743e-01
-9.21010002e-02 2.02314273e-01 -2.25825191e-01 1.07509530e+00
6.93379521e-01 1.04114485e+00 5.60696065e-01 -9.47985232e-01
2.14397177e-01 5.71755290e-01 1.99095346e-02 -6.94137335e-01
-9.23229679e-02 3.34538937e-01 -6.89076066e-01 4.93581772e-01
5.62691391e-01 1.39892191e-01 -1.52291727e+00 9.96397197e-01
-2.19538480e-01 -5.24080694e-02 1.01994917e-01 1.22971416e+00
1.44822431e+00 3.86205643e-01 3.36136431e-01 -4.69042473e-02
1.49248970e+00 -1.02599490e+00 -1.18692100e+00 -4.78063524e-01
5.69255888e-01 -8.20445776e-01 1.17317760e+00 5.50844610e-01
-8.51198435e-01 -2.26244479e-01 -7.67718315e-01 -4.02394414e-01
-4.79537427e-01 7.59301543e-01 9.11769450e-01 1.01133013e+00
-7.76893914e-01 -1.37704536e-01 -8.22399080e-01 -7.91029751e-01
5.14114857e-01 1.10295556e-01 -5.58290780e-01 -4.29103702e-01
-9.20451105e-01 1.30882967e+00 -1.41221642e-01 4.62773740e-01
-2.80828536e-01 -7.39345420e-03 -7.79887795e-01 -7.58876741e-01
3.41790348e-01 -1.24079488e-01 1.36560702e+00 -6.03725612e-01
-1.61752748e+00 1.35160542e+00 -5.80616534e-01 -3.28715593e-02
6.87793612e-01 -3.53071272e-01 -8.36898685e-01 -5.36652394e-02
-3.53440017e-01 3.38089526e-01 3.68787885e-01 -1.18229163e+00
-9.24437821e-01 -4.22892511e-01 -2.26241902e-01 2.59602129e-01
1.82994843e-01 7.44082928e-01 -8.17784369e-01 -3.38142276e-01
2.49499008e-01 -8.24931324e-01 4.59507629e-02 4.43060726e-01
1.82386190e-02 -4.45808619e-01 1.07197583e+00 -8.96910012e-01
1.27951956e+00 -2.23241997e+00 -3.99099320e-01 4.37995046e-02
1.63120180e-01 8.95353258e-01 -1.88866377e-01 2.48531535e-01
3.31092775e-01 7.35622656e-04 -4.77558747e-02 -3.69867496e-02
-6.65007532e-02 4.41179425e-01 -4.07730728e-01 7.13987887e-01
-2.92248577e-01 1.18107975e+00 -7.82589912e-01 -4.35506344e-01
7.50661433e-01 7.01478779e-01 2.72010177e-01 1.37713462e-01
2.71265775e-01 4.43500012e-01 -5.34629166e-01 1.11179459e+00
4.97052103e-01 -1.81526035e-01 6.50254488e-02 -5.64880110e-02
-2.94183910e-01 8.80312324e-02 -1.22542298e+00 1.26567721e+00
-1.52297899e-01 1.33278704e+00 -5.90255857e-02 -6.89182758e-01
1.07140541e+00 3.02801937e-01 1.76065892e-01 -8.77903640e-01
1.22694716e-01 6.61130667e-01 -4.70721684e-02 -1.01591885e+00
5.64091742e-01 2.19783455e-01 2.28171527e-01 2.57475048e-01
-5.04355729e-01 1.20337561e-01 2.30049461e-01 -1.11722223e-01
8.36227000e-01 1.32215858e-01 6.87714934e-01 5.64286649e-01
8.21725249e-01 -8.31130072e-02 3.06485415e-01 9.25889552e-01
-6.89172685e-01 4.98243481e-01 -4.60651368e-02 -4.98552799e-01
-3.99997145e-01 -8.75673711e-01 6.23260848e-02 9.27589417e-01
4.23652142e-01 -1.00640617e-01 -2.97840506e-01 -7.35267699e-01
-1.53956354e-01 6.08019292e-01 -4.63368833e-01 2.76610166e-01
-8.72078061e-01 -5.41375577e-01 9.64872420e-01 8.72328997e-01
1.07680070e+00 -1.36652648e+00 -7.84971416e-01 -3.38065028e-01
-4.16995198e-01 -1.34142578e+00 -1.71770245e-01 -6.61826968e-01
-6.82448864e-01 -1.30978942e+00 -1.09035230e+00 -1.03682590e+00
8.56333852e-01 7.40139484e-01 4.53899503e-01 3.98875356e-01
-4.98648107e-01 8.94586861e-01 -5.96295416e-01 -4.91910279e-01
-2.55281150e-01 -3.36801499e-01 -9.14985761e-02 -6.74173236e-02
1.10835612e+00 1.60274133e-02 -4.32241619e-01 6.93808675e-01
-6.46829426e-01 -5.57599440e-02 7.30528355e-01 5.85578442e-01
2.96380609e-01 -4.96838331e-01 7.43468329e-02 -3.75357300e-01
6.75136209e-01 3.83086741e-01 -4.82781321e-01 7.53026545e-01
-4.30960953e-01 -1.69612259e-01 -1.29619673e-01 -7.00899065e-01
-1.12288523e+00 -1.58574745e-01 2.88279783e-02 1.58140346e-01
-5.97969532e-01 8.58784616e-02 -2.27743700e-01 -4.86723810e-01
5.07338762e-01 6.94806576e-01 1.48825377e-01 -6.92067802e-01
5.24544537e-01 1.50212264e+00 6.25182152e-01 -1.85726926e-01
3.73512417e-01 6.45914316e-01 -1.07682891e-01 -1.41053963e+00
-5.18832147e-01 -7.68897891e-01 -4.37895358e-01 -7.14368165e-01
4.87677276e-01 -5.59590161e-01 -7.63691425e-01 1.27455735e+00
-1.06267107e+00 -3.80146325e-01 -2.98198640e-01 7.47168720e-01
-5.29235542e-01 6.13135934e-01 -5.62710822e-01 -1.24679017e+00
-1.50206462e-01 -1.06800425e+00 9.41070676e-01 6.44520879e-01
-2.48232827e-01 -6.10270858e-01 -6.28401414e-02 8.73262584e-01
4.48468983e-01 -4.17120941e-02 1.15038946e-01 -2.66969115e-01
-7.46584535e-01 -6.01600826e-01 -6.17863953e-01 4.10857290e-01
3.98209095e-01 4.13604488e-04 -1.02331889e+00 2.21290737e-02
-3.94816279e-01 -3.49474281e-01 8.53937924e-01 5.26908100e-01
8.97858977e-01 4.49769124e-02 -5.40239453e-01 2.10623860e-01
9.39535201e-01 4.61889386e-01 9.97607231e-01 3.05774868e-01
6.35626137e-01 4.99735415e-01 8.03650379e-01 -2.18559980e-01
4.07524765e-01 5.34134448e-01 -2.99781263e-01 -5.80496714e-02
-7.39651084e-01 -3.53983134e-01 4.50293481e-01 7.96872139e-01
-9.04378295e-01 -1.02809981e-01 -8.69046628e-01 3.81087035e-01
-1.85733676e+00 -1.09129393e+00 -3.16963851e-01 1.94748330e+00
5.46627343e-01 -5.35984278e-01 -6.94347429e-04 3.51849079e-01
5.98178267e-01 3.77626896e-01 -3.24371278e-01 -1.34981260e-01
-5.13077259e-01 -3.73624146e-01 5.72882593e-01 5.66948354e-01
-1.05828929e+00 1.21613181e+00 7.61002922e+00 5.89583993e-01
-1.55227494e+00 -2.11933166e-01 -3.05437855e-02 2.86238819e-01
3.74324292e-01 -3.53432029e-01 -9.98216867e-01 1.98691815e-01
3.48756462e-01 1.13306813e-01 3.88225794e-01 8.72084558e-01
4.39539075e-01 -4.58170176e-01 -8.37173462e-01 1.61274552e+00
6.52550876e-01 -1.16599023e+00 6.52674288e-02 2.19052985e-01
4.22990978e-01 6.94786906e-02 4.84800488e-02 8.31269696e-02
2.06067324e-01 -1.04532504e+00 4.64029789e-01 5.37511885e-01
1.17842829e+00 1.50450870e-01 7.74152339e-01 2.46394038e-01
-1.31058979e+00 -6.35336014e-03 7.57392049e-02 -1.58643529e-01
4.01455909e-01 -9.34205130e-02 -2.29993418e-01 1.45665884e-01
5.31839252e-01 9.20330882e-01 -5.80697715e-01 1.25686073e+00
-7.99672663e-01 6.46854162e-01 -2.38657102e-01 -4.83350128e-01
-8.01472738e-02 -1.59048066e-02 4.93758261e-01 1.45374513e+00
7.03279153e-02 3.61323029e-01 -6.50921017e-02 2.69479901e-01
2.85209805e-01 1.35623485e-01 -6.68564260e-01 -3.13218206e-01
1.99727908e-01 7.49669194e-01 -4.92299736e-01 -2.20449194e-01
-8.55680406e-01 1.05441010e+00 -2.72808790e-01 6.94644153e-01
-3.62472951e-01 -4.32122707e-01 5.70449412e-01 1.28682241e-01
-2.00546503e-01 -5.37785351e-01 -4.86240894e-01 -1.34964705e+00
2.97925562e-01 -1.13828897e+00 4.48722988e-01 -9.98295546e-01
-1.24859357e+00 4.54578325e-02 -7.09120631e-02 -1.49798608e+00
-1.19623840e-01 -1.12555563e+00 -1.27532795e-01 6.35618150e-01
-1.77186358e+00 -1.59573627e+00 -5.01275957e-01 4.92620975e-01
6.71913266e-01 -3.69428575e-01 8.23652744e-01 1.35241389e-01
-1.91982552e-01 7.36471951e-01 -2.25627646e-01 8.04829419e-01
6.51084602e-01 -6.55698717e-01 2.18297750e-01 1.07382917e+00
1.23464338e-01 7.41587222e-01 4.03814554e-01 -6.62177026e-01
-1.42999208e+00 -4.95811671e-01 1.14493430e+00 -5.93435049e-01
4.09395933e-01 1.31301343e-01 -5.89494944e-01 7.55532444e-01
-1.34060740e-01 1.53639868e-01 6.97155058e-01 -1.72121912e-01
-5.16754210e-01 7.18251839e-02 -1.24181414e+00 1.12328267e+00
1.41829002e+00 -6.15349770e-01 -8.04936945e-01 -1.70407400e-01
-2.21321180e-01 -4.31300998e-01 -2.91666895e-01 2.92687446e-01
1.29308939e+00 -5.73932767e-01 7.35710025e-01 -5.38388431e-01
-1.04889765e-01 -5.74513555e-01 -1.45710245e-01 -6.11018181e-01
-2.17557158e-02 -4.98510182e-01 -4.16938484e-01 8.08294237e-01
-1.18374759e-02 -8.76585960e-01 7.39117622e-01 1.13404608e+00
2.80723333e-01 -3.88082623e-01 -8.23738575e-01 -9.71436977e-01
-4.31988180e-01 -8.43705833e-01 1.70619205e-01 6.18752658e-01
1.64904252e-01 2.79952977e-02 -6.87782228e-01 2.31126342e-02
7.23224878e-01 -1.21674106e-01 1.08781838e+00 -1.06954217e+00
2.52254516e-01 -5.72143018e-01 -8.17785561e-01 -1.69140458e+00
-1.96773559e-01 -4.20024306e-01 -1.52850430e-02 -2.04174542e+00
1.93452150e-01 4.92668636e-02 -8.72534066e-02 7.94704080e-01
3.01683154e-02 3.36519778e-01 6.02919102e-01 3.97685528e-01
-6.12570405e-01 7.16849118e-02 1.56159365e+00 -2.19190568e-01
-2.63820440e-01 8.63780603e-02 -6.51800752e-01 1.07204103e+00
8.33008587e-01 1.22691952e-02 -7.93193877e-02 -3.32646132e-01
-3.20138365e-01 -4.84563023e-01 4.29490924e-01 -6.71598732e-01
4.85025793e-01 -4.83065039e-01 1.27981991e-01 -7.52414048e-01
4.50462461e-01 -8.20320129e-01 -3.17067623e-01 3.91587228e-01
-3.41572225e-01 -4.95316565e-01 -1.93231523e-01 2.40856186e-01
-2.09027410e-01 -9.60266069e-02 8.83396924e-01 -1.04267120e-01
-1.46134055e+00 -1.12386711e-01 -8.20579946e-01 -1.43316209e-01
8.72882366e-01 -7.97686577e-01 -5.29592335e-01 -5.61960042e-01
-4.55252320e-01 2.12185666e-01 2.27450460e-01 9.20716941e-01
1.00380147e+00 -1.19828260e+00 -4.47948098e-01 4.06235397e-01
3.39287460e-01 -4.68146175e-01 2.15144500e-01 8.01388144e-01
-4.65551794e-01 7.54702449e-01 -7.09118471e-02 -3.62349093e-01
-2.05812383e+00 -5.81478626e-02 3.12537253e-01 -2.14365199e-02
-8.23139668e-01 4.09324169e-01 -2.15215743e-01 -1.81485772e-01
6.40992939e-01 -3.06525975e-01 -5.42219043e-01 -4.47292030e-01
1.22506511e+00 7.85388470e-01 -5.75370312e-01 -9.47682023e-01
-5.60008466e-01 1.04611051e+00 -9.63046215e-03 -2.44905889e-01
8.44305098e-01 -2.41715580e-01 9.35723074e-03 5.06843865e-01
6.22835457e-01 -1.10489525e-01 -7.99885392e-01 -3.10777903e-01
-1.28783494e-01 -7.51473486e-01 -7.61863068e-02 -1.22545338e+00
-9.10411894e-01 9.16191101e-01 8.60112667e-01 -3.24745595e-01
1.00670266e+00 2.06194758e-01 6.53422236e-01 4.60256517e-01
8.26825440e-01 -1.34281230e+00 -5.63168451e-02 5.29538214e-01
1.06463706e+00 -1.36815977e+00 5.25181442e-02 -5.36687553e-01
-8.84256899e-01 1.09490621e+00 4.76463020e-01 1.27911493e-01
5.83688080e-01 3.04081172e-01 7.83086658e-01 6.23140708e-02
1.19334921e-01 -4.35266882e-01 4.44643855e-01 1.09904301e+00
4.26630646e-01 2.09906593e-01 -8.58743846e-01 4.09613132e-01
2.05836236e-01 8.61830175e-01 3.87677163e-01 1.26330400e+00
-4.34635967e-01 -1.26274753e+00 -5.47614813e-01 3.57038558e-01
-2.07591608e-01 3.99262547e-01 -5.86405873e-01 9.46563005e-01
-2.41497555e-03 1.02438021e+00 -4.04911608e-01 -3.39335591e-01
6.22226655e-01 1.48950592e-01 6.73102021e-01 -1.72277868e-01
1.08055517e-01 -1.98238745e-01 2.61843681e-01 -5.76760352e-01
-8.71582031e-01 -7.31238484e-01 -1.18505526e+00 -1.60302147e-01
-1.66406900e-01 -4.65112358e-01 6.65451527e-01 1.01190722e+00
4.41232175e-02 -9.83404666e-02 -8.19012597e-02 -5.04011989e-01
-2.26399258e-01 -9.16137695e-01 -7.91130483e-01 2.32851088e-01
3.64336222e-01 -5.61723351e-01 -4.00755465e-01 3.16084951e-01] | [9.119780540466309, -6.429074764251709] |
fac9c796-85e2-44da-9618-61447b6095ca | let-the-flows-tell-solving-graph | 2305.17010 | null | https://arxiv.org/abs/2305.17010v1 | https://arxiv.org/pdf/2305.17010v1.pdf | Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNets | Combinatorial optimization (CO) problems are often NP-hard and thus out of reach for exact algorithms, making them a tempting domain to apply machine learning methods. The highly structured constraints in these problems can hinder either optimization or sampling directly in the solution space. On the other hand, GFlowNets have recently emerged as a powerful machinery to efficiently sample from composite unnormalized densities sequentially and have the potential to amortize such solution-searching processes in CO, as well as generate diverse solution candidates. In this paper, we design Markov decision processes (MDPs) for different combinatorial problems and propose to train conditional GFlowNets to sample from the solution space. Efficient training techniques are also developed to benefit long-range credit assignment. Through extensive experiments on a variety of different CO tasks with synthetic and realistic data, we demonstrate that GFlowNet policies can efficiently find high-quality solutions. | ['Ling Pan', 'Yoshua Bengio', 'Aaron Courville', 'Nikolay Malkin', 'Hanjun Dai', 'Dinghuai Zhang'] | 2023-05-26 | null | null | null | null | ['combinatorial-optimization'] | ['methodology'] | [ 1.76069841e-01 -2.89012585e-02 -3.11308622e-01 -2.55946070e-01
-1.26977515e+00 -4.44486290e-01 5.74720740e-01 8.30021277e-02
-6.87812865e-01 1.29221940e+00 -1.47122517e-01 -3.61902207e-01
-3.38026792e-01 -8.21944475e-01 -7.32376277e-01 -7.64824927e-01
-5.63187227e-02 1.28614056e+00 -1.26183331e-01 3.36889595e-01
3.84853095e-01 4.33096796e-01 -1.30229342e+00 -4.14818572e-03
1.28053367e+00 9.02056932e-01 5.00047326e-01 7.42872179e-01
-3.52161735e-01 4.01510805e-01 -5.74937105e-01 -4.39612836e-01
1.96699381e-01 -3.28107208e-01 -7.22030699e-01 1.77289248e-01
1.88397393e-01 5.80554269e-02 4.48599160e-02 1.12921309e+00
5.28957248e-01 3.70219111e-01 9.94844735e-01 -1.14332342e+00
-3.81531149e-01 6.23084009e-01 -5.14909267e-01 2.90739715e-01
1.59714073e-01 3.44450206e-01 1.32859206e+00 -8.05844247e-01
4.84549105e-01 1.44879973e+00 3.55166197e-01 6.48613393e-01
-1.59695482e+00 -5.60714662e-01 1.32535070e-01 1.33268535e-01
-1.26373827e+00 -3.43076348e-01 3.25703740e-01 -1.14836469e-02
9.11377311e-01 2.22597107e-01 5.55731177e-01 1.15560591e+00
1.87973469e-01 1.11961925e+00 1.19546485e+00 -5.14793158e-01
7.42840171e-01 -5.31002693e-02 -1.15812540e-01 6.53178155e-01
3.59783649e-01 -5.42768724e-02 -5.65558672e-01 -5.01665711e-01
7.54270136e-01 -4.20071855e-02 -1.15707934e-01 -3.38709831e-01
-7.81354606e-01 1.00661671e+00 1.68477893e-01 -6.50211796e-02
-3.04982781e-01 4.01282579e-01 -6.29990697e-02 1.50784507e-01
3.61793041e-01 7.32496083e-01 -4.45218205e-01 -4.03370023e-01
-8.83629799e-01 9.47444260e-01 1.10496140e+00 1.02401388e+00
9.15310800e-01 -1.11178577e-01 -4.82180297e-01 8.11965883e-01
2.64156967e-01 5.30401528e-01 1.20669492e-01 -1.07998979e+00
8.92722130e-01 1.28014535e-01 5.53082585e-01 -6.73414648e-01
-1.01266690e-01 -3.67697120e-01 -6.81916177e-01 4.90730256e-02
9.02670383e-01 -3.45023721e-01 -1.12430859e+00 1.55841756e+00
2.36030728e-01 3.40788923e-02 -1.33304149e-01 7.34489977e-01
-3.00962210e-01 1.00893569e+00 1.34373978e-01 -2.97947317e-01
9.33055520e-01 -9.56122577e-01 -3.47544879e-01 -4.36118901e-01
4.89761859e-01 -5.56281567e-01 1.23033071e+00 5.63117743e-01
-1.42988241e+00 -1.10640474e-01 -6.13902867e-01 2.59360701e-01
8.15046057e-02 -2.24801481e-01 9.67370689e-01 6.66772723e-01
-8.74534845e-01 9.31831717e-01 -1.10273695e+00 -8.51493329e-02
8.88964534e-01 4.05160844e-01 2.98866659e-01 -6.12834871e-01
-8.51249754e-01 8.15154016e-01 3.95949036e-01 3.11708055e-03
-1.08130753e+00 -6.31912470e-01 -5.67270458e-01 5.26285350e-01
7.28266180e-01 -7.80827284e-01 1.40117037e+00 -1.00321186e+00
-1.49197114e+00 3.17588300e-01 -4.03654277e-01 -7.54557490e-01
5.63089609e-01 -1.47219403e-02 9.80726331e-02 8.04308802e-03
9.86143425e-02 7.60138154e-01 9.78645742e-01 -9.73467708e-01
-8.81848812e-01 -5.52614033e-02 -2.07834139e-01 3.32846999e-01
-2.21410424e-01 3.98876742e-02 -3.07730287e-01 -4.33164537e-01
-3.27534713e-02 -8.28920007e-01 -9.30468082e-01 -2.40120456e-01
-6.04935408e-01 -4.98954415e-01 1.11964747e-01 -1.98970050e-01
1.17823327e+00 -1.69481754e+00 2.78195500e-01 6.18886471e-01
9.35733244e-02 7.48524815e-02 -8.46067518e-02 2.14424908e-01
5.21260381e-01 3.95259559e-01 -4.04690832e-01 -6.26509905e-01
4.74080086e-01 7.12639093e-01 -1.73661023e-01 2.92948335e-01
2.12269410e-01 1.01374221e+00 -9.77122426e-01 -5.33250213e-01
-2.48713478e-01 -9.91225913e-02 -1.00834584e+00 1.27011999e-01
-9.57198679e-01 1.52813703e-01 -6.01915359e-01 6.33425653e-01
5.55659235e-01 -4.76040870e-01 3.00739676e-01 6.37481689e-01
2.82623798e-01 2.55015343e-01 -1.23794281e+00 1.35497463e+00
-6.10142291e-01 2.01663449e-01 1.16346896e-01 -1.03517282e+00
6.01366580e-01 -1.01081334e-01 1.76660374e-01 -5.97235024e-01
6.50017038e-02 2.80360281e-01 1.30094171e-01 -1.36087582e-01
5.07792294e-01 -4.28253949e-01 -1.00410827e-01 4.81989473e-01
4.87020053e-02 -7.22117797e-02 5.58157682e-01 -9.14396904e-03
1.09747171e+00 -7.44060427e-02 -6.71433732e-02 -3.10384572e-01
1.73780993e-01 -3.04434896e-02 8.77606332e-01 1.19988143e+00
-4.09161905e-03 4.34878916e-01 7.46507287e-01 -3.10098231e-01
-1.08999598e+00 -1.41800940e+00 6.90773427e-02 8.76285017e-01
-7.01148361e-02 -1.62339747e-01 -7.08916128e-01 -7.14608371e-01
-2.35539041e-02 8.55315447e-01 -1.96127087e-01 1.44975275e-01
-5.59635460e-01 -1.04546034e+00 1.76957279e-01 4.51887310e-01
1.57437488e-01 -1.08071089e+00 -2.49355078e-01 5.82431018e-01
-1.26742348e-01 -1.03429127e+00 -4.61186618e-01 5.14025271e-01
-1.02297592e+00 -6.73635840e-01 -9.29601431e-01 -5.61467528e-01
8.16697538e-01 -8.05909261e-02 1.36357462e+00 -1.12546019e-01
-3.50032061e-01 1.52872708e-02 4.02562022e-02 -4.64122057e-01
-4.35885906e-01 6.05775833e-01 2.25799326e-02 -1.02421224e-01
3.64023149e-01 -5.44973969e-01 -3.31638932e-01 1.32178217e-01
-6.53246403e-01 -6.41452521e-02 5.65274060e-01 9.40983891e-01
7.86117971e-01 1.33621514e-01 5.64393699e-01 -8.62933397e-01
8.76139164e-01 -5.99964857e-01 -1.08230209e+00 2.45859385e-01
-6.75009727e-01 5.27326167e-01 7.87011504e-01 -4.86328423e-01
-1.07712102e+00 -1.66629523e-01 -1.73774082e-02 -5.27485490e-01
-2.00203583e-01 4.84344512e-01 -1.63808167e-01 2.81157106e-01
5.96129835e-01 1.27340317e-01 -8.91058818e-02 -4.60110277e-01
1.75168023e-01 1.76857322e-01 8.02057907e-02 -1.23964012e+00
7.54382849e-01 2.78749615e-01 5.43430410e-02 -6.76875532e-01
-1.22966504e+00 -2.92703599e-01 -1.60132736e-01 4.65249158e-02
8.66430283e-01 -5.92198312e-01 -4.97193277e-01 7.51522332e-02
-1.06879723e+00 -6.88961267e-01 -3.49940479e-01 2.84016460e-01
-7.22251058e-01 7.10954368e-02 -5.22761941e-01 -1.05920362e+00
-8.54311511e-02 -1.16845417e+00 7.18438089e-01 5.25359869e-01
-1.70610636e-01 -1.00573838e+00 1.50709748e-01 2.79339015e-01
2.63603777e-01 -9.30908620e-02 1.03050065e+00 -4.32508916e-01
-1.19310868e+00 -1.40822846e-02 -1.82163864e-01 1.67358771e-01
-2.32273251e-01 6.20994018e-04 -6.87793493e-01 -3.07017922e-01
-2.53444016e-01 -4.80773717e-01 9.34873700e-01 6.51749849e-01
1.50810134e+00 -4.27571744e-01 -4.64066654e-01 6.60313964e-01
1.36807096e+00 -3.69971097e-02 6.42007232e-01 2.34807566e-01
4.15329099e-01 3.14847171e-01 6.30989790e-01 6.32534087e-01
1.81442171e-01 4.37938988e-01 1.56795368e-01 1.72803134e-01
3.27447474e-01 -4.35898930e-01 2.93717474e-01 4.78633672e-01
9.73593667e-02 -5.25110245e-01 -9.89702284e-01 4.09486234e-01
-1.90609586e+00 -1.06733859e+00 1.63236678e-01 2.06879425e+00
9.93419349e-01 4.48047757e-01 1.37972623e-01 -2.14797676e-01
6.76751614e-01 -1.12853482e-01 -5.34222901e-01 -4.00257975e-01
-9.66211373e-04 6.93268418e-01 7.07779646e-01 6.50906742e-01
-8.80726993e-01 1.14654648e+00 6.43026018e+00 1.34472275e+00
-5.90499878e-01 7.37848971e-03 1.07640636e+00 -3.58418941e-01
-4.56167579e-01 9.37995985e-02 -1.23035538e+00 5.03094673e-01
1.07995772e+00 -1.88105524e-01 7.02568829e-01 9.40358639e-01
9.62207019e-02 -3.74892831e-01 -1.05932963e+00 9.97958899e-01
-3.85994911e-01 -1.57821023e+00 -3.04702073e-02 2.77316302e-01
1.15489066e+00 -1.06989399e-01 1.91321686e-01 3.77797931e-01
9.62764502e-01 -1.38111508e+00 6.09829068e-01 4.21687484e-01
5.79887331e-01 -1.16406691e+00 3.89323294e-01 6.10827684e-01
-7.92519987e-01 -2.98724890e-01 -8.35850775e-01 -7.36236349e-02
4.84768867e-01 7.13657975e-01 -9.15163875e-01 2.20008373e-01
6.02074206e-01 3.36496443e-01 -2.40744039e-01 1.38485980e+00
-2.29330942e-01 8.09910178e-01 -6.99328244e-01 -5.33414423e-01
6.72200561e-01 -4.37744796e-01 3.91994089e-01 9.17893589e-01
3.99839520e-01 -2.01867998e-01 3.67273688e-01 1.30782187e+00
-1.18275844e-01 -6.38264418e-02 -2.47447714e-01 -3.52483034e-01
4.18617159e-01 1.05753779e+00 -9.46051419e-01 -1.51816726e-01
-1.25227958e-01 7.04366863e-01 6.66638434e-01 6.40087485e-01
-9.64983642e-01 -9.98314545e-02 9.59216237e-01 -5.71343489e-02
4.29902226e-01 -4.94024396e-01 -4.16292936e-01 -1.25025225e+00
-1.46463588e-01 -7.83180177e-01 3.34799290e-01 -2.08840057e-01
-1.70242751e+00 1.57284111e-01 -4.49604401e-03 -6.90406561e-01
-1.83534071e-01 -4.49868530e-01 -6.43010616e-01 1.05934763e+00
-1.49177158e+00 -4.40179944e-01 1.72703326e-01 3.85494888e-01
6.49116874e-01 -7.85012692e-02 4.24061596e-01 1.84015274e-01
-6.65415704e-01 5.20577669e-01 4.12326008e-01 -3.67028326e-01
3.21897864e-01 -1.40315008e+00 3.71015400e-01 7.16613591e-01
4.19118732e-01 3.69121760e-01 7.16416359e-01 -4.78486985e-01
-1.22202408e+00 -8.50360930e-01 6.72799408e-01 -3.51558417e-01
6.68143094e-01 -6.08626604e-01 -8.10807705e-01 2.89652348e-01
-1.10818543e-01 -2.75000155e-01 5.12750924e-01 2.33672321e-01
6.94380105e-02 3.36258598e-02 -1.01318276e+00 9.07988191e-01
1.00138044e+00 -1.21770069e-01 -1.10270448e-01 5.94995916e-01
5.09662569e-01 -2.95227110e-01 -4.18067753e-01 -3.94934937e-02
4.69095483e-02 -7.96384513e-01 1.03428543e+00 -6.61392033e-01
5.18498540e-01 8.70074183e-02 -9.65707451e-02 -1.30513000e+00
-1.94570839e-01 -1.02543056e+00 -3.83549541e-01 1.12031209e+00
7.61833906e-01 -6.28863454e-01 1.27848709e+00 7.51895368e-01
7.17934519e-02 -1.12396514e+00 -9.59608078e-01 -8.65974188e-01
4.12307382e-01 -5.95054805e-01 5.04531801e-01 3.46930772e-01
-3.39572489e-01 3.02631378e-01 -1.57475963e-01 -2.22323854e-02
1.03783464e+00 1.34360522e-01 6.96333528e-01 -1.11711109e+00
-6.46236658e-01 -5.67745030e-01 2.39751026e-01 -1.44828510e+00
4.04471487e-01 -9.41283584e-01 2.74285555e-01 -1.56240487e+00
1.46712080e-01 -1.17963517e+00 4.29688320e-02 7.00565651e-02
-2.57957637e-01 -3.59408617e-01 2.49648914e-01 1.03296064e-01
-7.57957280e-01 8.13205242e-01 1.32790101e+00 4.81058359e-02
-8.83218572e-02 5.01286805e-01 -6.68774903e-01 6.58850193e-01
9.58178818e-01 -7.32530773e-01 -4.88434106e-01 -2.22890615e-01
3.70312035e-01 2.45142266e-01 3.09733246e-02 -9.25795138e-01
5.91367483e-02 -7.96197891e-01 4.41388160e-01 -5.41024268e-01
2.62799233e-01 -4.74177659e-01 -2.46088162e-01 2.69872934e-01
-3.90841633e-01 6.95909187e-02 -8.34799632e-02 9.31475699e-01
1.01460978e-01 -6.60961390e-01 9.48566556e-01 -4.91359830e-01
-3.00941259e-01 6.71599269e-01 -6.93772078e-01 5.93914330e-01
9.43016887e-01 -4.71543781e-02 1.23834297e-01 -3.30401242e-01
-6.85566485e-01 6.22547507e-01 4.45636630e-01 -1.69940174e-01
5.45214593e-01 -9.46077645e-01 -3.46119642e-01 -1.31192192e-01
-4.11047846e-01 5.36414325e-01 1.01505883e-01 6.04171574e-01
-6.27802193e-01 3.98164302e-01 2.56539080e-02 -5.71159899e-01
-7.70812809e-01 4.41506773e-01 2.96965629e-01 -8.94768894e-01
-4.50644553e-01 1.09327316e+00 -3.50976326e-02 -4.13869232e-01
4.43962008e-01 -3.66796345e-01 4.25927162e-01 -1.16075546e-01
4.33648795e-01 4.13960755e-01 -3.11193436e-01 2.81883031e-01
6.97863623e-02 -9.69134346e-02 -1.81992441e-01 -5.31189322e-01
1.33009195e+00 -5.51322401e-02 6.79593254e-03 1.72665566e-01
8.61666024e-01 -3.93005341e-01 -1.52395642e+00 -1.80040419e-01
4.75447506e-01 -6.87312007e-01 -2.60615498e-01 -5.05098999e-01
-6.85756326e-01 9.92751539e-01 7.09238350e-02 2.58275181e-01
6.46607816e-01 6.67085813e-04 9.48112547e-01 6.53047860e-01
5.66088021e-01 -1.36428010e+00 3.67603391e-01 7.10606515e-01
3.17601293e-01 -9.58068669e-01 -1.57620490e-01 -1.93735570e-01
-5.41996539e-01 1.03411424e+00 7.26045787e-01 -2.52377838e-01
4.64797974e-01 3.62743676e-01 -3.38008732e-01 9.85913128e-02
-1.14602935e+00 -2.05135167e-01 -1.95876256e-01 5.66077948e-01
3.12467031e-02 2.19706651e-02 -3.13829780e-02 4.02848363e-01
-4.82130945e-02 -2.16878057e-01 5.90176940e-01 8.37861478e-01
-6.26714230e-01 -1.52336586e+00 -3.89643043e-01 8.07597578e-01
-5.96176386e-01 -1.28051847e-01 1.20221063e-01 3.42526823e-01
-2.73729682e-01 6.00916147e-01 9.76554602e-02 2.49031752e-01
-1.12390079e-01 5.26655763e-02 6.86641455e-01 -7.79160500e-01
-3.90190989e-01 1.52267247e-01 1.58545673e-01 -5.22503257e-01
2.91908272e-02 -8.76601934e-01 -1.06458759e+00 -3.72824818e-01
-3.38307887e-01 2.39187106e-01 3.62217903e-01 1.12384427e+00
2.00499937e-01 2.59386241e-01 3.12932640e-01 -1.10644138e+00
-1.18515491e+00 -6.60555840e-01 -6.13672912e-01 -6.88187927e-02
-1.54246703e-01 -7.21228361e-01 -2.96996504e-01 -2.89488107e-01] | [6.822367191314697, 3.9349334239959717] |
c03a072e-f99f-4491-9fe9-5fc950acd9ab | sc6d-symmetry-agnostic-and-correspondence | 2208.02129 | null | https://arxiv.org/abs/2208.02129v3 | https://arxiv.org/pdf/2208.02129v3.pdf | SC6D: Symmetry-agnostic and Correspondence-free 6D Object Pose Estimation | This paper presents an efficient symmetry-agnostic and correspondence-free framework, referred to as SC6D, for 6D object pose estimation from a single monocular RGB image. SC6D requires neither the 3D CAD model of the object nor any prior knowledge of the symmetries. The pose estimation is decomposed into three sub-tasks: a) object 3D rotation representation learning and matching; b) estimation of the 2D location of the object center; and c) scale-invariant distance estimation (the translation along the z-axis) via classification. SC6D is evaluated on three benchmark datasets, T-LESS, YCB-V, and ITODD, and results in state-of-the-art performance on the T-LESS dataset. Moreover, SC6D is computationally much more efficient than the previous state-of-the-art method SurfEmb. The implementation and pre-trained models are publicly available at https://github.com/dingdingcai/SC6D-pose. | ['Esa Rahtu', 'Janne Heikkilä', 'Dingding Cai'] | 2022-08-03 | null | null | null | null | ['6d-pose-estimation'] | ['computer-vision'] | [-2.26628140e-01 -3.04033309e-01 -3.13902825e-01 -2.99452811e-01
-7.73854733e-01 -5.94663501e-01 6.15941107e-01 -3.19780171e-01
-1.62328050e-01 -6.33314066e-03 -1.02533519e-01 -4.08492275e-02
1.40457034e-01 -3.27270865e-01 -8.18732560e-01 -4.66030329e-01
1.42664328e-01 9.26259100e-01 3.27552140e-01 2.54318982e-01
5.19200623e-01 1.13542342e+00 -1.38022280e+00 -1.39000535e-01
1.53940842e-01 1.54431415e+00 1.90431491e-01 5.71455121e-01
2.92617798e-01 3.07476074e-01 -2.56259561e-01 -1.12225935e-01
6.65349543e-01 -4.38407771e-02 -6.93713129e-01 4.43443298e-01
7.54248440e-01 -6.20369613e-01 -5.10175109e-01 8.06712866e-01
4.87703323e-01 -1.77866265e-01 7.18582034e-01 -1.29623687e+00
-3.61570984e-01 -2.66999960e-01 -9.26596761e-01 -2.91699886e-01
6.60700500e-01 1.34554103e-01 6.74876690e-01 -1.50708294e+00
9.73395050e-01 1.24835110e+00 5.58804035e-01 2.70684987e-01
-9.99348342e-01 -5.71698964e-01 -1.12875521e-01 2.45373085e-01
-1.57519901e+00 -1.99646786e-01 9.09243166e-01 -4.98626918e-01
9.78038013e-01 3.72123793e-02 6.66682720e-01 8.07983398e-01
-2.20324076e-03 7.46590734e-01 1.07610106e+00 -5.37981689e-01
2.61302125e-02 -3.22622329e-01 -5.11517599e-02 5.80550313e-01
3.56821805e-01 2.62075346e-02 -7.10904777e-01 -2.72668041e-02
1.27703881e+00 1.16516888e-01 -2.15575248e-01 -1.27295530e+00
-1.37936461e+00 4.54495996e-01 6.93813443e-01 -2.57964641e-01
-2.61008978e-01 2.47296825e-01 1.68286994e-01 2.23833788e-02
3.34286988e-01 1.98339537e-01 -5.46264291e-01 -1.00193493e-01
-3.60270292e-01 3.05097789e-01 4.97790366e-01 1.54444599e+00
7.26343572e-01 -3.13030124e-01 3.98814499e-01 4.37757164e-01
4.95561957e-01 8.27438653e-01 1.00756779e-01 -1.09109807e+00
6.07871890e-01 6.58153892e-01 4.99364525e-01 -7.89352775e-01
-5.50483346e-01 -2.13676795e-01 -4.22889411e-01 2.83164859e-01
4.91691351e-01 3.33377272e-01 -9.26047444e-01 1.16728961e+00
7.91985393e-01 -3.47299725e-02 -3.23673815e-01 1.25132501e+00
9.65597212e-01 4.02929634e-01 -7.20541775e-01 2.13296026e-01
1.28890216e+00 -1.11780953e+00 -2.85543561e-01 -2.49177501e-01
3.64377797e-01 -1.22291911e+00 7.86574066e-01 3.29405010e-01
-1.08418989e+00 -5.18008351e-01 -1.04972136e+00 -4.02141869e-01
-2.62507558e-01 5.50645173e-01 7.21313834e-01 3.36197555e-01
-6.32259309e-01 2.20688611e-01 -1.04624605e+00 -3.93238395e-01
4.24006194e-01 3.79179180e-01 -7.73918092e-01 -2.73214817e-01
-3.94692332e-01 1.12518835e+00 2.22664982e-01 3.44073102e-02
-6.60594165e-01 -7.20962644e-01 -9.83462989e-01 -4.69373554e-01
6.03951812e-01 -6.42003596e-01 1.51040971e+00 -2.81683326e-01
-1.67978299e+00 1.39994037e+00 -2.15724438e-01 3.01879328e-02
8.04305792e-01 -5.69575369e-01 2.33161494e-01 3.17065388e-01
1.76061373e-02 5.90611637e-01 8.97298872e-01 -1.26558411e+00
-2.54980505e-01 -9.58595395e-01 -2.09142655e-01 5.05497754e-01
2.47270867e-01 -1.96105242e-03 -1.11135268e+00 -5.84265888e-01
9.07393336e-01 -1.23029375e+00 1.36033803e-01 8.22138250e-01
-4.83092964e-01 -3.43869865e-01 1.12518191e+00 -5.02794623e-01
4.39285487e-01 -2.19280005e+00 1.80880666e-01 8.11010003e-02
-4.42711599e-02 2.33820245e-01 -1.14827953e-01 2.33153522e-01
-1.90195113e-01 -4.14040625e-01 3.13730776e-01 -5.03738999e-01
1.02543615e-01 -2.57756889e-01 -1.61173075e-01 1.12390530e+00
5.41127585e-02 9.54641521e-01 -4.96967167e-01 -3.00968975e-01
5.82882822e-01 5.73077559e-01 -2.71014303e-01 2.18896374e-01
-9.70158502e-02 3.54216605e-01 -4.21579480e-01 1.10301042e+00
1.15247047e+00 -3.69979888e-01 -8.13379735e-02 -5.72798789e-01
-9.16158780e-02 4.04195309e-01 -1.42955136e+00 2.09432554e+00
-1.83328554e-01 4.07269031e-01 -1.30610451e-01 -5.37858427e-01
1.09891963e+00 -5.75883985e-02 5.61668694e-01 -5.02445459e-01
2.78458983e-01 2.55797178e-01 -4.89158273e-01 -1.33330926e-01
2.74340361e-01 5.42782009e-01 9.00764465e-02 5.53529501e-01
9.11194757e-02 -6.44962966e-01 -1.35253832e-01 -1.12500317e-01
6.24175727e-01 7.09163725e-01 5.04848421e-01 -7.19183981e-02
3.05510253e-01 -7.26921335e-02 2.89933205e-01 3.27642679e-01
-3.86920236e-02 1.07955182e+00 1.85971528e-01 -3.86879623e-01
-1.10459232e+00 -1.16787970e+00 -3.55710417e-01 2.92539179e-01
6.95218086e-01 -3.58879238e-01 -4.23572749e-01 -4.47748691e-01
6.46420181e-01 2.67764568e-01 -4.96223509e-01 1.87259123e-01
-6.67421639e-01 -1.43170714e-01 -1.35545447e-01 7.25484610e-01
6.14204288e-01 -5.82443833e-01 -8.92526507e-01 -2.22142667e-01
1.61944162e-02 -1.36447060e+00 -6.41948342e-01 1.73064038e-01
-1.05249918e+00 -1.23238623e+00 -7.73549795e-01 -6.20910347e-01
7.67875075e-01 6.63058281e-01 9.02591109e-01 -2.58337229e-01
-5.83626509e-01 4.87679809e-01 -3.14504504e-01 -5.06360710e-01
3.93836319e-01 9.65370331e-03 7.62948245e-02 -2.19094872e-01
5.13338864e-01 -2.71313697e-01 -7.88768470e-01 1.00967181e+00
-2.40799785e-01 9.91856903e-02 5.19889712e-01 4.01285112e-01
9.45339859e-01 -4.94315267e-01 -2.48789981e-01 -1.50301009e-01
-5.34509659e-01 9.78752375e-02 -1.10058618e+00 4.58026258e-03
-3.42217654e-01 -2.83548892e-01 -1.16454683e-01 -5.01309037e-01
-8.51653636e-01 5.91088116e-01 2.74655491e-01 -8.10881853e-01
-4.74801324e-02 -5.20608649e-02 -3.43043476e-01 -4.09155011e-01
4.62069452e-01 -8.45413208e-02 1.06629275e-01 -8.50449920e-01
3.54795277e-01 6.45426631e-01 6.59745395e-01 -4.02768731e-01
1.12195075e+00 7.23403931e-01 3.53146791e-01 -8.24871302e-01
-7.82473683e-01 -8.68941307e-01 -1.18363094e+00 -2.62590736e-01
7.02848017e-01 -1.02882028e+00 -8.70235860e-01 9.35044050e-01
-1.32641554e+00 -2.46654510e-01 -1.30040661e-01 7.88183391e-01
-9.25240576e-01 1.55314729e-01 -4.10969287e-01 -5.48325300e-01
-3.02586824e-01 -1.23040068e+00 1.66240191e+00 6.32613897e-02
-1.01570420e-01 -2.69550234e-01 -2.55027384e-01 5.53132713e-01
-9.43803936e-02 4.53662544e-01 5.50048351e-01 -1.56830773e-01
-1.17310596e+00 -5.38535357e-01 -4.76139933e-01 2.75033638e-02
6.51406646e-02 -2.78149303e-02 -8.33990335e-01 -2.92153895e-01
-1.24108251e-02 -6.13704920e-01 4.82531130e-01 3.71417284e-01
1.09220028e+00 7.97442347e-02 -3.85062009e-01 1.00178444e+00
1.48112333e+00 -2.57788766e-02 5.22631228e-01 5.66900611e-01
8.41287017e-01 3.43266189e-01 1.17532611e+00 4.73666906e-01
3.31946582e-01 1.31791699e+00 8.42986226e-01 9.67027098e-02
-3.38846087e-01 -2.68440068e-01 1.54127955e-01 4.86790717e-01
-2.36120328e-01 1.49465457e-01 -1.01406884e+00 1.29550070e-01
-1.61473405e+00 -4.08393472e-01 -2.47028708e-01 2.19897604e+00
4.60348964e-01 -2.69269869e-02 1.52519256e-01 1.38553038e-01
5.88534832e-01 1.03052676e-01 -8.98766577e-01 1.85551614e-01
-1.77352829e-03 -1.31367566e-02 5.48060298e-01 3.30087245e-01
-1.14949238e+00 9.67605293e-01 5.37411213e+00 5.60944915e-01
-1.19899750e+00 -5.74931577e-02 1.84583843e-01 -5.77428862e-02
4.36225504e-01 1.06334183e-02 -1.08830047e+00 1.84858423e-02
6.61720484e-02 2.51053840e-01 1.67719468e-01 1.23983252e+00
-2.93114096e-01 -3.39106709e-01 -1.28883278e+00 1.52320814e+00
4.14102405e-01 -1.20799923e+00 -2.62122095e-01 4.03360948e-02
7.02166021e-01 2.71633208e-01 -1.98055767e-02 -2.64187157e-01
-2.23041251e-01 -6.71544790e-01 1.13633060e+00 2.82982886e-01
1.14419353e+00 -5.47966421e-01 5.67968309e-01 1.84895426e-01
-1.42824543e+00 2.59807527e-01 -4.89088237e-01 2.80279130e-01
-6.85674176e-02 3.49207044e-01 -7.56061614e-01 7.31455684e-01
1.01717365e+00 9.84989524e-01 -7.66043544e-01 1.06389999e+00
-5.33611596e-01 -3.65684070e-02 -4.50761408e-01 1.28538474e-01
-1.42793074e-01 -9.69748944e-02 6.01677656e-01 6.88135743e-01
3.09805393e-01 2.24463299e-01 4.20626737e-02 5.37505627e-01
-1.59339719e-02 -2.43817389e-01 -3.58319879e-01 2.56120980e-01
4.10946906e-01 1.12869704e+00 -8.52673948e-01 1.36545688e-01
-2.89957851e-01 9.66349483e-01 4.36619185e-02 2.42273331e-01
-6.74846292e-01 -2.92947650e-01 5.56377649e-01 3.32835168e-01
8.33707154e-01 -4.58627701e-01 -2.85546571e-01 -1.23799253e+00
4.00125653e-01 -7.39740193e-01 1.33034542e-01 -1.37312818e+00
-1.02806306e+00 3.80862981e-01 1.32613435e-01 -1.64408529e+00
-1.78147197e-01 -1.16452587e+00 -1.29660442e-01 8.53016853e-01
-1.35028422e+00 -1.37130249e+00 -7.92125702e-01 5.60980260e-01
4.71281469e-01 2.08949260e-02 6.45273447e-01 -3.55402343e-02
-2.62493752e-02 4.55170602e-01 -1.06307879e-01 1.84623688e-01
8.31252635e-01 -9.58366394e-01 5.37368834e-01 5.12645125e-01
-4.12719287e-02 3.45163971e-01 3.75758648e-01 -4.34143871e-01
-2.18964839e+00 -9.55685616e-01 7.40878344e-01 -9.19672847e-01
3.39785814e-01 -6.97026968e-01 -5.06142497e-01 7.94105768e-01
-4.24288601e-01 4.97507930e-01 3.55943036e-03 -2.87896216e-01
-7.03242004e-01 -2.43703738e-01 -1.07439363e+00 3.21376503e-01
1.32532227e+00 -4.92760986e-01 -5.82873583e-01 4.30255353e-01
5.94810843e-01 -1.16595137e+00 -8.61999094e-01 5.04225910e-01
8.48967493e-01 -8.68978202e-01 1.28589642e+00 -1.43871114e-01
1.62607193e-01 -6.13942564e-01 -3.77864540e-01 -7.04471171e-01
-8.74439105e-02 -5.40692925e-01 -3.45237434e-01 8.59179080e-01
-7.53708184e-02 -5.31016588e-01 9.90102589e-01 2.43216082e-01
3.56124230e-02 -9.09081578e-01 -9.98661876e-01 -1.00871468e+00
-4.05597538e-01 -4.76863205e-01 5.46940923e-01 6.41179800e-01
-4.84552175e-01 1.58276290e-01 -1.57791153e-02 3.85283589e-01
8.86618197e-01 6.33572280e-01 1.53618109e+00 -1.14895105e+00
8.48645195e-02 -2.52010643e-01 -9.49258506e-01 -1.57573926e+00
-8.16027448e-02 -6.72412634e-01 -1.65448308e-01 -1.24814999e+00
1.50279298e-01 -2.22370610e-01 3.33903849e-01 4.56422061e-01
2.87247092e-01 4.93070632e-01 2.73308843e-01 4.09375608e-01
-4.27974701e-01 5.25958061e-01 1.34951210e+00 1.02886513e-01
-8.22633412e-03 1.02018863e-01 -1.20359369e-01 7.88508594e-01
4.58804667e-01 -4.15554285e-01 -7.85260797e-02 -3.07918042e-01
-1.27741210e-02 1.06610037e-01 5.94997525e-01 -7.92170227e-01
1.53808132e-01 -1.62531734e-01 8.29813302e-01 -1.48970246e+00
7.63498962e-01 -9.57828045e-01 1.73231766e-01 5.16037464e-01
1.55745000e-01 1.22386210e-01 5.57309799e-02 2.43605450e-01
6.44292030e-03 -1.10217489e-01 9.20114398e-01 1.65386423e-02
-6.75554872e-01 6.36544704e-01 3.98926198e-01 -9.54930633e-02
1.19474947e+00 -4.81497496e-01 -4.05176699e-01 -3.50111693e-01
-3.33771646e-01 -8.56201816e-03 8.90784502e-01 5.54420054e-01
7.74604738e-01 -1.50060213e+00 -5.18133104e-01 3.58428806e-01
4.85763252e-01 5.18303156e-01 -8.90541598e-02 8.34194183e-01
-8.40759873e-01 6.53504491e-01 -1.30740970e-01 -1.32217324e+00
-1.44501972e+00 5.30957878e-01 2.44585574e-01 1.85959578e-01
-6.55763149e-01 7.77255237e-01 1.35471240e-01 -6.91875398e-01
5.04583299e-01 -3.69120985e-01 4.22764182e-01 -9.41402689e-02
4.13444102e-01 6.32226527e-01 3.39337617e-01 -9.35786963e-01
-6.17792904e-01 1.36650002e+00 6.82238257e-03 -8.00403729e-02
1.43937886e+00 1.60112958e-02 -4.24338318e-02 2.20879719e-01
1.37691498e+00 -3.60324025e-01 -1.54470086e+00 -4.53349560e-01
-8.36243182e-02 -9.63846982e-01 -6.17340542e-02 -6.11210346e-01
-9.55257475e-01 8.76744568e-01 6.86557412e-01 -4.69792128e-01
9.89525437e-01 3.47706854e-01 3.95785928e-01 5.52082539e-01
7.32552171e-01 -7.70027578e-01 5.35422027e-01 5.89802027e-01
1.43615723e+00 -1.18614173e+00 5.71701527e-01 -7.45113730e-01
-3.41164440e-01 1.29340231e+00 7.43552148e-01 -2.63188303e-01
7.53684282e-01 6.65565059e-02 5.86572997e-02 -1.33415997e-01
-1.67673618e-01 5.88831585e-03 7.33751953e-01 5.43064892e-01
-2.67776474e-02 -1.48262516e-01 2.76501805e-01 -7.68988281e-02
-1.36596262e-01 -2.05385208e-01 -1.61037594e-02 1.24507689e+00
-1.23770915e-01 -8.17806482e-01 -7.50179052e-01 -8.29104241e-03
-8.89475644e-02 4.92862850e-01 -6.03280604e-01 1.16259289e+00
-2.06321076e-01 4.36502993e-01 1.88623101e-01 -1.64570287e-01
4.99290973e-01 -1.69851109e-01 1.17088377e+00 -5.75414956e-01
5.59451282e-02 2.62096882e-01 -2.20887631e-01 -1.13198006e+00
-4.86387640e-01 -8.68828595e-01 -1.11864638e+00 -1.63732886e-01
-6.81714356e-01 -4.34761196e-01 1.17902982e+00 5.20874083e-01
3.62938404e-01 -2.90750146e-01 7.85243750e-01 -1.52016234e+00
-7.99989760e-01 -8.17834675e-01 -5.27631819e-01 3.44104260e-01
3.65369409e-01 -9.78561282e-01 -3.64795029e-01 -9.45900828e-02] | [7.404983997344971, -2.6434547901153564] |
318c22fb-2ce0-4dab-a311-fc4e434c0dc3 | acquiring-annotated-data-with-cross-lingual | 1808.10290 | null | http://arxiv.org/abs/1808.10290v2 | http://arxiv.org/pdf/1808.10290v2.pdf | Acquiring Annotated Data with Cross-lingual Explicitation for Implicit Discourse Relation Classification | Implicit discourse relation classification is one of the most challenging and
important tasks in discourse parsing, due to the lack of connective as strong
linguistic cues. A principle bottleneck to further improvement is the shortage
of training data (ca.~16k instances in the PDTB). Shi et al. (2017) proposed to
acquire additional data by exploiting connectives in translation: human
translators mark discourse relations which are implicit in the source language
explicitly in the translation. Using back-translations of such explicitated
connectives improves discourse relation parsing performance. This paper
addresses the open question of whether the choice of the translation language
matters, and whether multiple translations into different languages can be
effectively used to improve the quality of the additional data. | ['Frances Yung', 'Wei Shi', 'Vera Demberg'] | 2018-08-30 | acquiring-annotated-data-with-cross-lingual-1 | https://aclanthology.org/W19-2703 | https://aclanthology.org/W19-2703.pdf | ws-2019-6 | ['implicit-discourse-relation-classification'] | ['natural-language-processing'] | [ 4.32208896e-01 8.68218482e-01 -6.34210825e-01 -2.80944705e-01
-1.01158655e+00 -1.04729462e+00 7.00370073e-01 3.79984260e-01
-3.82186085e-01 1.17167509e+00 5.99706888e-01 -9.33081865e-01
7.57181272e-02 -5.81664383e-01 -6.99318707e-01 -2.41726086e-01
2.86550105e-01 6.67293608e-01 1.88139305e-01 -6.25255704e-01
6.40886053e-02 -1.13204755e-01 -8.67521524e-01 7.44577527e-01
1.25740480e+00 3.68968189e-01 3.58732134e-01 4.85955030e-01
-4.72434103e-01 1.05356145e+00 -8.14178526e-01 -6.22825325e-01
-2.03436278e-02 -8.13850999e-01 -1.41989458e+00 -8.46618563e-02
1.81757331e-01 -1.48199484e-01 8.06482062e-02 8.88415635e-01
6.11649677e-02 -2.26132125e-01 3.34879875e-01 -6.47379935e-01
-5.22935033e-01 1.27440155e+00 -1.94101766e-01 4.43144530e-01
4.92398888e-01 -1.43301517e-01 1.37522590e+00 -5.46556354e-01
1.10495150e+00 1.21254277e+00 3.27563643e-01 6.77070618e-01
-1.27371669e+00 -2.35882521e-01 1.41381264e-01 5.16913891e-01
-6.79239929e-01 -7.07176328e-01 7.75434971e-01 -3.92163634e-01
1.18813932e+00 6.39603376e-01 3.98298681e-01 1.30572295e+00
-3.91377330e-01 6.84299946e-01 1.34237242e+00 -1.03172171e+00
-1.03729084e-01 1.72932461e-01 4.10736680e-01 5.02559662e-01
7.37610981e-02 -1.48305401e-01 -6.89220190e-01 6.78236187e-02
4.29727882e-01 -9.11555946e-01 -4.11086410e-01 3.04364234e-01
-1.27760589e+00 9.65581596e-01 3.25073451e-01 6.06830716e-01
1.33452863e-01 -2.93677598e-01 7.50473261e-01 6.37562633e-01
5.61830819e-01 1.06020677e+00 -7.27731287e-01 -6.37454689e-01
-3.82824928e-01 2.99005248e-02 9.34584320e-01 9.69395876e-01
5.20735979e-01 -3.52169991e-01 1.03590963e-02 7.84362495e-01
-5.12348861e-02 2.71354914e-01 2.83329278e-01 -1.23128021e+00
1.16658378e+00 7.21036911e-01 -3.28601673e-02 -6.40707135e-01
-8.04991275e-02 -8.14516768e-02 -2.55127072e-01 -2.62801051e-01
9.77305174e-01 -3.83062214e-01 -3.22167337e-01 1.58057261e+00
3.34628224e-01 -4.49465871e-01 2.82034755e-01 8.24923456e-01
8.50153387e-01 5.69080114e-01 1.04513496e-01 -4.57659841e-01
1.47373629e+00 -1.05882347e+00 -9.64473307e-01 -5.15214622e-01
1.17010748e+00 -9.53729212e-01 1.21218777e+00 -1.17705621e-01
-1.00917733e+00 -1.72673628e-01 -8.38003278e-01 -5.13881564e-01
-5.77911502e-03 1.58744395e-01 7.71046877e-01 2.56221026e-01
-5.90374947e-01 4.42016125e-01 -7.66216040e-01 -1.58144549e-01
3.41532916e-01 1.44990072e-01 -3.65109593e-01 -1.14488870e-01
-1.50024068e+00 1.29615903e+00 4.56213206e-01 1.28502846e-01
7.74813220e-02 -4.30773675e-01 -9.54996109e-01 -1.64579988e-01
7.46478677e-01 -3.72737527e-01 1.42143011e+00 -1.25287521e+00
-1.63610160e+00 1.18578160e+00 -3.16997826e-01 -3.89648080e-01
5.65126300e-01 -4.89503264e-01 1.57438546e-01 1.06807753e-01
1.52127907e-01 3.45014304e-01 3.77505720e-01 -9.40035582e-01
-6.76610947e-01 -2.37389088e-01 4.86378938e-01 3.91495883e-01
-6.32238388e-02 4.24988121e-01 6.21231534e-02 -4.61148381e-01
2.02144593e-01 -8.99783850e-01 7.43335113e-02 -5.08734941e-01
-4.15528297e-01 -6.64659739e-01 5.56428194e-01 -1.13121855e+00
1.14079809e+00 -1.88101006e+00 5.59234142e-01 -3.57116431e-01
-9.97616909e-03 2.67599255e-01 5.80952950e-02 3.69059145e-01
1.99896753e-01 4.07688141e-01 -1.38985619e-01 -1.40885293e-01
-1.29043654e-01 7.75795102e-01 -4.38965827e-01 1.39884487e-01
7.10851908e-01 1.10927558e+00 -9.66619074e-01 -5.00114977e-01
-1.75052032e-01 -2.93106697e-02 -1.57400519e-01 2.01974183e-01
-6.60993576e-01 9.62603271e-01 -4.33377117e-01 4.25893724e-01
1.14605494e-01 -6.90549612e-02 7.75687099e-01 4.19175252e-02
-1.42165840e-01 1.69741392e+00 -6.37638569e-01 1.50574577e+00
-3.73533398e-01 1.03102255e+00 1.65711582e-01 -1.08295584e+00
6.00890279e-01 5.91467619e-01 -1.69389620e-02 -7.62497902e-01
1.72243759e-01 5.06091058e-01 6.45526171e-01 -6.88152909e-01
5.05011559e-01 -3.02377164e-01 -2.42553502e-01 4.39031154e-01
-1.39232844e-01 -1.59800336e-01 3.82842213e-01 -5.00705764e-02
9.90909159e-01 1.71149433e-01 4.00531322e-01 -3.45022947e-01
5.00212371e-01 6.93927467e-01 7.94529021e-01 3.30650955e-01
2.25488041e-02 3.13941687e-01 9.78385091e-01 -1.40259087e-01
-1.12121069e+00 -5.15700340e-01 -2.78878301e-01 1.15631187e+00
-2.22519100e-01 -5.17932415e-01 -7.51337767e-01 -8.60594153e-01
-3.97916347e-01 7.50900865e-01 -5.82865059e-01 3.35036218e-01
-1.63685942e+00 -5.16199470e-01 7.17202425e-01 5.11905968e-01
2.02513695e-01 -1.08335531e+00 -4.78355199e-01 2.37967491e-01
-8.44269097e-01 -1.43263006e+00 7.95954913e-02 4.14641589e-01
-1.02433693e+00 -1.20561349e+00 -1.64702281e-01 -8.83923650e-01
6.30689859e-01 -2.16715604e-01 1.18259132e+00 3.04743558e-01
4.38362420e-01 -3.23799938e-01 -6.83991432e-01 -1.35865420e-01
-1.11551309e+00 6.30932570e-01 -4.58429396e-01 -7.66663790e-01
3.27069432e-01 -2.16366947e-01 1.48789704e-01 -1.11618601e-01
-2.66222328e-01 3.72716457e-01 3.24045181e-01 1.11573541e+00
2.90403992e-01 -2.48662204e-01 4.30318415e-01 -1.59450340e+00
7.54580319e-01 -1.45432204e-01 -4.66236383e-01 3.09262604e-01
-1.87793821e-01 3.19711089e-01 5.29336154e-01 -5.00539899e-01
-1.10268760e+00 -3.56507510e-01 -1.15828648e-01 4.24392253e-01
2.69403607e-02 8.10002983e-01 -2.68210649e-01 4.04386848e-01
6.41000628e-01 -4.51716840e-01 -1.22854393e-02 -5.22327006e-01
4.93694931e-01 5.21394730e-01 2.21276298e-01 -1.03555000e+00
5.88078558e-01 -2.00553149e-01 -1.18449897e-01 -7.89404213e-01
-1.15102923e+00 -2.39875346e-01 -9.86714482e-01 2.61578441e-01
9.45583999e-01 -7.41433978e-01 -3.41192663e-01 -7.44069666e-02
-1.57914054e+00 -7.03499079e-01 -3.13915610e-01 4.05757308e-01
-3.50026816e-01 3.07931274e-01 -1.12886190e+00 -5.68916619e-01
-1.79815635e-01 -1.06014037e+00 7.11207867e-01 -1.84584081e-01
-8.18231761e-01 -1.03405905e+00 -3.93253975e-02 8.63411069e-01
1.24851972e-01 1.59327343e-01 1.59072340e+00 -6.13396108e-01
-5.80721617e-01 2.86563247e-01 -1.35018617e-01 3.06100339e-01
4.23018336e-01 -6.82332963e-02 -7.76300192e-01 1.06073923e-01
2.65220702e-01 -4.96614873e-01 4.42192674e-01 2.63470076e-02
4.28707093e-01 -7.59436190e-01 -1.19857281e-01 1.55826673e-01
8.92321289e-01 -1.04210572e-02 5.20796955e-01 6.20219767e-01
6.51786983e-01 1.11756432e+00 7.32793093e-01 -4.23229575e-01
5.33301651e-01 6.50382578e-01 -6.29108548e-02 9.02362019e-02
-4.16489840e-01 1.07430294e-02 4.89234775e-01 1.58843064e+00
-2.09991992e-01 -2.89425030e-02 -1.04700136e+00 4.78728652e-01
-1.81360078e+00 -7.04224885e-01 -6.51788592e-01 1.79390097e+00
1.53408897e+00 3.25167149e-01 -1.10257000e-01 3.18572521e-01
4.56596583e-01 6.11339957e-02 -7.55257457e-02 -5.49022794e-01
-3.84632468e-01 3.91710162e-01 1.77923888e-01 1.00884092e+00
-9.77437675e-01 1.14564872e+00 5.88264704e+00 5.10404885e-01
-1.07061458e+00 2.61928946e-01 5.07962108e-01 3.24368209e-01
-3.80707383e-01 3.89331609e-01 -7.18783617e-01 3.25484455e-01
9.62231755e-01 4.46170755e-02 3.92812192e-01 4.77225423e-01
-6.35156259e-02 -2.79175490e-01 -1.46698868e+00 2.59392530e-01
-1.26249507e-01 -1.46067166e+00 -3.14553827e-01 -2.77530532e-02
4.99765456e-01 -2.52749026e-02 -2.41006687e-01 2.92255759e-01
2.51056463e-01 -1.05594075e+00 6.75221980e-01 -1.34325579e-01
8.19226503e-01 -2.61494100e-01 8.84870231e-01 4.99610007e-01
-7.42601216e-01 2.38059208e-01 -1.42333001e-01 -6.17654622e-01
2.26335630e-01 1.77115977e-01 -1.17511094e+00 4.62535173e-01
2.65163094e-01 4.58512038e-01 -5.34754217e-01 2.31445998e-01
-9.82909739e-01 1.23235404e+00 -4.03418124e-01 -2.55178988e-01
1.57485038e-01 -2.75843740e-01 8.18257451e-01 1.12682068e+00
-1.50432408e-01 1.96623966e-01 1.77154034e-01 5.81817627e-01
-1.63752899e-01 1.43874601e-01 -3.18506062e-01 -2.52885610e-01
4.87534404e-01 8.95662844e-01 -5.85207701e-01 -2.59203702e-01
-4.34865147e-01 7.30582595e-01 8.38127375e-01 2.54522204e-01
-3.78266692e-01 1.72726259e-01 4.20245647e-01 1.94210291e-01
1.94878682e-01 -4.37773079e-01 -6.31198823e-01 -1.05631113e+00
3.93312246e-01 -1.19231677e+00 3.10166895e-01 -2.95637965e-01
-1.32228196e+00 6.32704437e-01 -5.03486581e-03 -7.53509820e-01
-4.54463959e-01 -8.50224674e-01 -3.56932074e-01 1.07193935e+00
-1.56086147e+00 -1.12912071e+00 1.16166919e-01 -5.76521605e-02
7.41922379e-01 1.11900538e-01 1.13065398e+00 2.09333822e-01
-4.37471420e-01 4.24394816e-01 -1.68151796e-01 6.06959045e-01
7.21285105e-01 -1.41858506e+00 3.00450206e-01 7.61688352e-01
2.32726336e-01 7.38935947e-01 8.00407887e-01 -5.69724500e-01
-1.20052075e+00 -5.58574975e-01 1.41045427e+00 -8.10932994e-01
9.61274266e-01 -4.92469907e-01 -1.29896080e+00 8.80278468e-01
6.87831938e-01 -3.49911153e-01 8.35638583e-01 7.09424317e-01
-2.01094344e-01 4.29257482e-01 -6.30716741e-01 5.82767248e-01
9.21829224e-01 -8.26729357e-01 -1.18593478e+00 3.44242156e-01
1.15899575e+00 -9.04625595e-01 -8.48252594e-01 3.54348540e-01
-1.86075363e-02 -3.69779170e-01 4.97189522e-01 -8.29213142e-01
9.35918808e-01 -1.62520446e-02 -1.57723874e-01 -1.21920145e+00
1.31671965e-01 -6.01888120e-01 -2.62403250e-01 1.55945492e+00
8.49598229e-01 -5.60598791e-01 6.42212927e-01 6.56634569e-01
-1.77217454e-01 -8.39573026e-01 -1.11478913e+00 -5.38934350e-01
6.32503390e-01 -7.41811991e-02 2.86387563e-01 1.24133253e+00
6.53132677e-01 1.27668500e+00 8.68338048e-02 -2.91003674e-01
3.07330023e-02 3.33610594e-01 6.44680023e-01 -1.13499939e+00
-5.48285425e-01 -2.93928146e-01 1.39145672e-01 -1.09795010e+00
4.31372583e-01 -1.08788753e+00 3.20687555e-02 -1.64627898e+00
-9.22687054e-02 -9.95292664e-01 3.51647019e-01 5.15047729e-01
-4.60829377e-01 -1.51828155e-01 3.46888512e-01 3.06525409e-01
-3.16513002e-01 4.01840091e-01 1.46479213e+00 -1.98838534e-03
-3.40530932e-01 -5.99177554e-02 -6.53136075e-01 7.20495582e-01
9.24978673e-01 -5.48382401e-01 -2.29739189e-01 -9.91261184e-01
3.67996961e-01 2.90105283e-01 -1.46197423e-01 -1.92469075e-01
2.47771200e-02 -3.76816541e-01 -1.55008137e-01 -3.07986885e-01
3.00458729e-01 -4.99119341e-01 -2.86796778e-01 2.33388796e-01
-4.70918447e-01 1.63058206e-01 2.23044768e-01 4.51484658e-02
-4.02421147e-01 -6.22182012e-01 1.81995794e-01 -3.42156470e-01
-3.89955938e-01 -6.19279981e-01 -4.08801019e-01 6.55513406e-01
6.12698317e-01 -1.16701527e-02 -8.02906334e-01 4.62987348e-02
-4.92821217e-01 4.28237244e-02 3.76759261e-01 2.23687917e-01
1.34050697e-02 -9.84144211e-01 -9.01905715e-01 -3.27534735e-01
-1.60899058e-01 4.54076022e-01 -4.93604511e-01 9.21344340e-01
-5.54602921e-01 4.81168628e-01 6.22093268e-02 -4.51639175e-01
-1.61334550e+00 1.43933222e-01 4.88001034e-02 -6.20653629e-01
-7.05828309e-01 8.27371061e-01 -2.52287865e-01 -5.59143722e-01
-3.22634429e-02 -5.31483829e-01 -2.55108416e-01 8.96073729e-02
2.02361882e-01 2.15206757e-01 1.44629657e-01 -7.86501169e-01
-3.22070032e-01 2.55343784e-02 -1.87676266e-01 -2.81881124e-01
1.45659399e+00 -2.52653271e-01 -5.09331405e-01 6.34906292e-01
8.72757316e-01 5.13332367e-01 -8.62178624e-01 -3.51973653e-01
7.05804884e-01 -2.69178599e-01 -3.79056096e-01 -1.01785064e+00
-2.85746366e-01 8.48060846e-01 -2.96655625e-01 2.46604711e-01
6.51105225e-01 3.46831948e-01 8.29889357e-01 5.49933672e-01
2.35870302e-01 -1.20709014e+00 -4.00545187e-02 1.14223564e+00
8.50830376e-01 -1.46747231e+00 3.55077684e-02 -1.05428779e+00
-6.08348846e-01 1.14748287e+00 5.91949463e-01 6.33752346e-02
9.12645012e-02 3.59519601e-01 3.89907449e-01 -2.06985012e-01
-7.71409988e-01 -1.33583516e-01 2.19669744e-01 4.75647032e-01
9.49402392e-01 2.58790582e-01 -8.38133156e-01 5.45198679e-01
-6.37606680e-01 -5.46952486e-01 4.92586881e-01 1.00553071e+00
-8.57784450e-02 -1.75673592e+00 -3.23516488e-01 3.00372005e-01
-5.64563215e-01 -2.33317062e-01 -7.53922880e-01 7.31730759e-01
2.08606899e-01 1.12096262e+00 -1.14631020e-02 8.41988847e-02
1.51224643e-01 3.03613842e-01 8.43309104e-01 -1.05157661e+00
-7.06767499e-01 -9.56575051e-02 9.63454962e-01 -1.08039737e-01
-7.29722142e-01 -8.71074915e-01 -1.47031760e+00 -2.39663228e-01
-6.18836880e-01 4.35943246e-01 3.17198843e-01 1.33456147e+00
-5.84974289e-02 6.37910604e-01 1.01842664e-01 -1.71226129e-01
-5.39875388e-01 -1.27245629e+00 2.75942862e-01 4.40919846e-01
3.51361454e-01 -5.61659813e-01 -2.78774530e-01 2.64530599e-01] | [10.740315437316895, 9.265174865722656] |
46fa9854-bc08-4249-b802-8d5a7bcdc04b | ambipun-generating-humorous-puns-with | null | null | https://openreview.net/forum?id=RhhZo7CpksR | https://openreview.net/pdf?id=RhhZo7CpksR | $AmbiPun$ : Generating Humorous Puns with Ambiguous Context | Computational humor has garnered interest of the natural language processing community due to its wide applications to real world scenarios. One way to express humor is via the use of puns. A homographic pun plays on words that are spelled the same way but have different meanings. In this paper, we propose a simple yet effective way to generate pun sentences that does not require any pun sentences to train on. Our approach is inspired by humor theories that ambiguity comes from the context rather than the pun word itself. Given a pair of definitions of a pun word, our model first produces a list of related concepts through a reverse dictionary. We then utilize one-shot GPT3 to generate context words, and then generate punning sentences that incorporate context words from both worlds.
We also investigate how the position of a pun word appearing in the sentence will influence the generated results. We compare our proposed $\textsc{AmbiPun}$ with well crafted baselines. Human evaluation shows that our method successfully generates pun 52% of the time, outperforming the the state-of-the-art model by a large margin. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['reverse-dictionary'] | ['natural-language-processing'] | [ 2.07358271e-01 -2.43819907e-01 3.09922099e-01 -5.31884562e-03
-6.20010555e-01 -7.00810373e-01 6.36853814e-01 1.09368026e-01
-4.58334982e-01 8.59514832e-01 7.71762013e-01 -3.17911282e-02
4.62330997e-01 -1.04786944e+00 -4.28575218e-01 -4.78253484e-01
7.60873079e-01 4.56309915e-01 5.17691299e-03 -9.07311022e-01
6.87585771e-01 1.48770045e-02 -1.12131739e+00 4.69307303e-01
6.98697448e-01 1.54307589e-01 2.52156615e-01 3.91061068e-01
-3.73855531e-01 1.02924216e+00 -1.15390527e+00 -8.68448019e-01
3.18065107e-01 -1.03771043e+00 -8.52989256e-01 -4.96730246e-02
2.95444161e-01 -5.95669746e-02 2.92899739e-02 1.14264154e+00
5.38785636e-01 2.52455652e-01 4.38400626e-01 -7.56942749e-01
-1.01446331e+00 1.12597871e+00 -6.44316196e-01 6.10244758e-02
5.08899808e-01 2.15744600e-01 1.34908247e+00 -8.64553809e-01
5.89879155e-01 1.39986336e+00 5.07973969e-01 7.43120372e-01
-1.29290426e+00 -5.90625644e-01 -4.79106486e-01 5.00645280e-01
-1.25597489e+00 1.40651539e-01 1.06085587e+00 -2.52666175e-01
9.77962255e-01 4.27347869e-01 5.48344195e-01 1.33405828e+00
2.42462113e-01 6.30382955e-01 1.30962920e+00 -7.59508669e-01
2.54691362e-01 4.74500395e-02 6.27175793e-02 3.51152539e-01
5.55244684e-02 -1.29399626e-02 -7.40301788e-01 -4.59397882e-01
3.98051679e-01 -3.30170423e-01 -1.85074046e-01 4.23098803e-01
-9.57676649e-01 1.23082185e+00 2.67843395e-01 6.07617438e-01
-4.34226483e-01 3.15265298e-01 5.08665860e-01 1.02451622e-01
-6.14573760e-03 1.02645719e+00 5.57790771e-02 -3.64668936e-01
-9.43771958e-01 6.96660519e-01 9.31043684e-01 6.77566528e-01
4.96367544e-01 -4.62561958e-02 -3.62570971e-01 9.66845036e-01
-1.23550326e-01 5.45604527e-01 5.79408467e-01 -8.48817647e-01
2.30477929e-01 4.02596503e-01 3.50144476e-01 -1.28710437e+00
-7.20884725e-02 -2.95850545e-01 -5.16707540e-01 4.81405593e-02
4.10284638e-01 6.83739781e-02 -5.17494917e-01 1.86121023e+00
-1.34824529e-01 5.40541997e-03 8.23479984e-03 1.11489522e+00
7.05413163e-01 7.47924924e-01 3.44697782e-03 -2.15034217e-01
1.40974045e+00 -1.10469210e+00 -5.47854483e-01 -4.72589999e-01
5.52549601e-01 -1.02290761e+00 1.63228083e+00 5.48694611e-01
-1.00526798e+00 -2.22639740e-01 -1.06489551e+00 -2.98197120e-01
-1.92516685e-01 -3.79517913e-01 3.23106676e-01 6.32450879e-01
-6.51857853e-01 7.04986453e-01 3.01614571e-02 -2.38222122e-01
8.82489234e-02 -3.09608191e-01 2.65846923e-02 1.01027101e-01
-1.53617573e+00 1.31766272e+00 4.70770210e-01 -2.68028021e-01
-6.49964452e-01 -3.81719381e-01 -7.53175914e-01 2.09366277e-01
3.26813400e-01 -8.52985919e-01 1.32930088e+00 -9.89456058e-01
-1.26005590e+00 8.84812951e-01 -2.50430584e-01 -6.84444368e-01
3.37695211e-01 -4.02481079e-01 -7.43734390e-02 -4.15188372e-02
2.85636336e-01 4.04439867e-01 6.94275558e-01 -1.22950745e+00
-1.25621542e-01 1.63935035e-01 1.21804647e-01 1.60228252e-01
-4.68116045e-01 3.85642260e-01 1.39704704e-01 -8.19098711e-01
-3.36055487e-01 -7.98790514e-01 -2.83184290e-01 -6.05141222e-01
-3.10825408e-01 -3.26888174e-01 3.30194294e-01 -6.45317554e-01
1.58373737e+00 -1.85567760e+00 1.03790760e-01 6.52973652e-02
1.06667750e-01 2.22276896e-01 -3.22956055e-01 9.57059264e-01
1.82098359e-01 3.23307484e-01 -5.01170993e-01 -2.47590825e-01
2.47974619e-01 4.46742445e-01 -7.69405305e-01 -6.99141249e-02
-5.23038805e-02 1.03462517e+00 -1.27046657e+00 -4.72188324e-01
1.57855786e-02 2.60415345e-01 -5.80379069e-01 1.97750002e-01
-4.68228400e-01 1.49399281e-01 -7.53675625e-02 1.44945666e-01
3.52993786e-01 -1.52831137e-01 2.47602597e-01 2.31723502e-01
2.28717670e-01 6.96541071e-01 -7.87678599e-01 1.51890409e+00
-4.61255729e-01 3.51628929e-01 -7.38430917e-01 -2.25801572e-01
1.16745496e+00 1.71876281e-01 -2.30565831e-01 -6.63853049e-01
3.66710305e-01 2.92901486e-01 5.01079410e-02 -3.91502142e-01
1.01419866e+00 -9.93256807e-01 -5.58138072e-01 7.12032318e-01
-3.35546374e-01 -6.10796988e-01 5.98936021e-01 3.54292542e-01
1.24897420e+00 -9.24291313e-02 8.23942661e-01 -2.70474762e-01
7.21401989e-01 6.81882072e-03 5.93590021e-01 8.96858692e-01
-1.49796575e-01 6.42818570e-01 5.98922968e-01 -4.17091727e-01
-1.20265853e+00 -1.07122135e+00 3.11600417e-01 9.29659724e-01
2.45574694e-02 -7.01372981e-01 -7.40821064e-01 -4.56969261e-01
-2.06172675e-01 1.58158565e+00 -8.05414021e-01 -9.71020237e-02
-9.36303914e-01 -4.50344533e-01 7.10946441e-01 3.48023176e-01
3.16364676e-01 -1.68466735e+00 -9.55503225e-01 5.11579037e-01
-5.65994799e-01 -9.14984584e-01 -6.47913456e-01 -1.48372903e-01
-4.78197604e-01 -8.47671568e-01 -3.25984180e-01 -5.90995252e-01
3.75345200e-01 2.94724107e-01 1.50293982e+00 3.33848655e-01
2.13931054e-02 -5.86537570e-02 -6.55470610e-01 -5.19331098e-01
-7.95406103e-01 -3.78898889e-01 -9.15538967e-02 -4.73621190e-01
9.24405217e-01 -8.27650309e-01 -2.16180533e-01 -4.87439148e-02
-1.13035166e+00 2.09856465e-01 3.46678555e-01 1.12149072e+00
2.81669348e-01 -7.05501586e-02 5.30655921e-01 -1.03979802e+00
1.17997098e+00 -2.95677722e-01 1.88802462e-03 -4.05494608e-02
-2.52669036e-01 1.44482881e-01 1.08031642e+00 -3.19808692e-01
-6.99876666e-01 -3.56471032e-01 3.01391445e-03 -2.16766626e-01
8.10384192e-03 5.04534125e-01 -8.17820523e-03 3.95436645e-01
1.08784354e+00 2.67547339e-01 -3.37084234e-01 -2.19637781e-01
7.40949273e-01 4.32286918e-01 6.41064525e-01 -9.01881754e-01
9.91339803e-01 3.11193258e-01 -1.18078709e-01 -6.06312931e-01
-1.11826086e+00 -3.38550776e-01 -3.06908369e-01 -5.18369302e-02
6.66286886e-01 -5.00923038e-01 -3.46284151e-01 1.97315782e-01
-1.85293067e+00 -2.68308669e-02 -3.14808428e-01 -7.23131597e-02
-5.54161847e-01 6.58702374e-01 -7.38494277e-01 -7.98031628e-01
-7.88341343e-01 -6.97581172e-01 6.19391680e-01 3.48539144e-01
-9.69953060e-01 -7.34380543e-01 2.54979104e-01 5.85071325e-01
2.75877655e-01 2.69616753e-01 1.11075962e+00 -6.66757345e-01
-8.64384025e-02 -4.79087755e-02 -9.66079980e-02 5.45924544e-01
3.49948034e-02 -5.67038178e-01 -7.03529418e-01 1.83484122e-01
2.64905393e-01 -5.84400594e-01 7.65288055e-01 -2.81026512e-01
7.27919996e-01 -7.22254515e-01 4.09083873e-01 -1.31015390e-01
1.48810577e+00 -5.28466627e-02 9.74334955e-01 5.08159041e-01
4.84767854e-01 4.33398247e-01 3.86703104e-01 6.71112716e-01
1.18382275e-01 4.40678775e-01 3.47054780e-01 3.20913255e-01
-1.30174672e-02 -5.46889186e-01 4.29718345e-01 8.28380823e-01
8.30610394e-02 -2.24586353e-01 -7.91978776e-01 8.60960603e-01
-1.74672019e+00 -1.45131230e+00 -3.10812443e-01 2.05217004e+00
1.37135446e+00 2.69361347e-01 -2.95677837e-02 9.93317068e-02
4.86254394e-01 4.03432786e-01 1.07872680e-01 -1.12149286e+00
-4.32655871e-01 6.88630283e-01 -1.50672914e-02 7.86433756e-01
-5.43216169e-01 1.40225172e+00 5.47688150e+00 8.95730138e-01
-7.11375356e-01 4.99351695e-02 3.05671453e-01 -2.13035062e-01
-6.79356694e-01 2.74121553e-01 -3.31495255e-01 3.92521620e-01
5.86175203e-01 -6.35026455e-01 6.60618722e-01 7.62953579e-01
6.53021395e-01 -1.07173428e-01 -1.07295692e+00 9.84450698e-01
7.02482045e-01 -1.14773428e+00 4.09177899e-01 -3.00018936e-01
8.45358253e-01 -5.42216480e-01 -1.95962936e-01 3.42355013e-01
2.35329032e-01 -1.30555212e+00 9.36810791e-01 3.31893057e-01
1.11351587e-01 -8.02533329e-01 8.74323606e-01 5.50781846e-01
-6.40175402e-01 7.66062438e-02 -6.05308712e-01 -7.89629459e-01
4.55986530e-01 5.84086716e-01 -1.07939577e+00 1.16468564e-01
2.30449140e-01 1.94696903e-01 -5.34684539e-01 8.32062185e-01
-1.17596304e+00 8.04832399e-01 -1.26913699e-04 -6.27308965e-01
2.30976269e-01 -1.51252389e-01 8.27054620e-01 1.50646627e+00
2.89645553e-01 4.48603541e-01 1.87517166e-01 1.46618104e+00
-3.41026127e-01 3.83083135e-01 -4.06310499e-01 -1.40762433e-01
5.13683200e-01 1.26072979e+00 -3.43362898e-01 -4.14379805e-01
-9.05148461e-02 1.39717257e+00 4.23173547e-01 -1.31230392e-02
-1.00770104e+00 -2.77471602e-01 4.02026743e-01 1.25150189e-01
1.86072648e-01 -2.72850990e-01 -8.65943491e-01 -8.48691046e-01
1.03739671e-01 -1.03728867e+00 1.95633560e-01 -1.11294925e+00
-1.77391684e+00 5.07001042e-01 -2.86563486e-01 -8.90964270e-01
-2.18538746e-01 -3.51129025e-01 -1.06694245e+00 1.05677652e+00
-1.15139091e+00 -1.11741006e+00 3.62249501e-02 1.46354541e-01
6.55339479e-01 1.84183598e-01 1.04663444e+00 -2.31619671e-01
1.75836571e-02 5.06802380e-01 -4.77878630e-01 2.52699435e-01
8.43502820e-01 -1.30768418e+00 4.66379762e-01 1.21929383e+00
4.08961356e-01 9.92430508e-01 1.50206280e+00 -7.20628560e-01
-8.69178116e-01 -7.30930090e-01 1.73174238e+00 -6.62321925e-01
9.76194859e-01 -1.87824845e-01 -9.38633382e-01 1.91869363e-01
7.35224843e-01 -7.03920305e-01 8.83634686e-01 3.29205766e-02
-9.54953134e-01 4.28350061e-01 -9.61456299e-01 9.89266098e-01
7.69892871e-01 -4.34128582e-01 -1.34565377e+00 1.50353611e-01
7.59808123e-01 -2.43961930e-01 -9.83988494e-02 -1.29983574e-01
2.77337909e-01 -1.03929663e+00 4.72995102e-01 -8.74857962e-01
1.37714231e+00 -3.60326320e-01 -3.33688021e-01 -1.46598077e+00
-5.86808801e-01 -7.86584139e-01 1.33213103e-01 1.03572035e+00
4.06580061e-01 -2.02785388e-01 6.33821011e-01 5.49327493e-01
-5.11143245e-02 -6.19241238e-01 -7.99487531e-01 -8.70544374e-01
5.71292102e-01 -4.83965904e-01 5.67411780e-01 8.38384151e-01
6.06227636e-01 8.90286982e-01 -6.90742910e-01 -2.98034608e-01
5.20259917e-01 3.90841573e-01 7.55364001e-01 -6.01404965e-01
-7.82955706e-01 -5.75116456e-01 -2.73435414e-01 -8.76251459e-01
2.18679950e-01 -1.03534615e+00 2.95454800e-01 -1.42546546e+00
7.70534992e-01 1.25899151e-01 -2.77213991e-01 7.38678753e-01
-5.63963950e-01 5.37706196e-01 7.16227829e-01 -3.06019466e-02
-4.28431213e-01 6.68003976e-01 1.11782813e+00 -6.00151755e-02
-2.79164702e-01 -6.61154389e-01 -1.04444253e+00 7.63248384e-01
1.06154144e+00 -6.30181789e-01 -1.48995101e-01 -3.83591980e-01
5.73212862e-01 -1.40531033e-01 5.55362582e-01 -8.62071455e-01
-3.28879873e-03 -6.05492294e-01 -6.18439987e-02 -2.72617698e-01
1.93546191e-01 -3.78521740e-01 -7.15330914e-02 3.82306218e-01
-3.07862669e-01 1.49112865e-01 1.99891612e-01 2.20024273e-01
3.20513286e-02 -6.60675228e-01 9.61154938e-01 -5.67788482e-01
-4.33383554e-01 -5.49391985e-01 -4.22829568e-01 2.95670271e-01
7.32798636e-01 -8.78465623e-02 -5.44965148e-01 -5.16970217e-01
-9.72711816e-02 -1.31088689e-01 7.65773594e-01 5.67124307e-01
7.14321911e-01 -1.37254882e+00 -1.13925123e+00 -3.87595415e-01
1.50262341e-01 -4.81091082e-01 5.60504235e-02 3.80360425e-01
-4.42731231e-01 1.94890395e-01 -1.72228113e-01 1.07938595e-01
-1.38624930e+00 6.59428120e-01 6.82286993e-02 -5.13786912e-01
-5.45881391e-01 8.89890611e-01 7.45405331e-02 -1.83800787e-01
-3.16353500e-01 2.41275191e-01 -7.94052184e-02 -2.21372023e-01
8.25158000e-01 2.28274614e-01 -2.25274250e-01 -7.41611958e-01
-1.90349236e-01 4.92606275e-02 -1.77303642e-01 -4.71081316e-01
1.10246003e+00 1.93288639e-01 -6.40425622e-01 5.73815286e-01
8.01741481e-01 5.21364152e-01 -3.73991609e-01 -1.74849942e-01
1.48035556e-01 -7.93519616e-01 -5.39873838e-01 -1.18617201e+00
-3.71945322e-01 6.53646231e-01 -2.08796501e-01 -9.03248638e-02
9.74683404e-01 -9.77950767e-02 1.43171692e+00 3.02641869e-01
7.00979948e-01 -1.16609585e+00 5.95619440e-01 1.01617789e+00
1.25461543e+00 -7.56079853e-01 -7.59793669e-02 -3.09902489e-01
-9.56794083e-01 1.17311013e+00 6.52429223e-01 -4.13526595e-01
-2.27317750e-01 -9.85568166e-02 2.37417251e-01 -1.07494347e-01
-7.04707384e-01 5.57326190e-02 6.64379969e-02 2.86334753e-01
7.06720948e-01 3.29139173e-01 -1.46321893e+00 1.19937801e+00
-9.26732838e-01 -2.28035655e-02 1.19619417e+00 5.82921326e-01
-8.36473823e-01 -1.44215298e+00 -5.36710262e-01 2.38279238e-01
-3.17199409e-01 -7.60395110e-01 -9.89881039e-01 1.86170474e-01
2.07668930e-01 1.12388253e+00 -6.44318163e-01 -5.60741663e-01
1.66528672e-01 2.14093372e-01 4.88717973e-01 -7.89330959e-01
-1.04662061e+00 -2.40289867e-01 2.93679863e-01 -1.91569716e-01
8.37010890e-02 -5.77916920e-01 -1.50794375e+00 -7.01447070e-01
-7.57199600e-02 3.50454986e-01 1.22420982e-01 1.12602580e+00
-1.41448798e-02 7.62467384e-02 5.42611718e-01 -4.34876412e-01
-9.26006317e-01 -1.01633120e+00 -5.32904863e-01 8.81183088e-01
-4.63264704e-01 -1.94830462e-01 -3.57993305e-01 1.16124917e-02] | [11.337682723999023, 9.07463264465332] |
2d5ee83e-8b03-4d9d-9d06-9c42983e6325 | span-based-joint-entity-and-relation-2 | 2210.12720 | null | https://arxiv.org/abs/2210.12720v1 | https://arxiv.org/pdf/2210.12720v1.pdf | Span-based joint entity and relation extraction augmented with sequence tagging mechanism | Span-based joint extraction simultaneously conducts named entity recognition (NER) and relation extraction (RE) in text span form. However, since previous span-based models rely on span-level classifications, they cannot benefit from token-level label information, which has been proven advantageous for the task. In this paper, we propose a Sequence Tagging augmented Span-based Network (STSN), a span-based joint model that can make use of token-level label information. In STSN, we construct a core neural architecture by deep stacking multiple attention layers, each of which consists of three basic attention units. On the one hand, the core architecture enables our model to learn token-level label information via the sequence tagging mechanism and then uses the information in the span-based joint extraction; on the other hand, it establishes a bi-directional information interaction between NER and RE. Experimental results on three benchmark datasets show that STSN consistently outperforms the strongest baselines in terms of F1, creating new state-of-the-art results. | ['Jing Yang', 'Huijun Liu', 'Jun Ma', 'Jie Yu', 'Hao Xu', 'Shasha Li', 'Bin Ji'] | 2022-10-23 | null | null | null | null | ['joint-entity-and-relation-extraction'] | ['natural-language-processing'] | [-6.13655262e-02 3.81476171e-02 -4.14637476e-01 -2.15466410e-01
-8.98535728e-01 -4.37448084e-01 3.46003562e-01 -4.45676642e-03
-6.51687384e-01 9.18754399e-01 4.44409966e-01 -4.29076821e-01
1.88895300e-01 -7.76461780e-01 -6.22492015e-01 -5.20900488e-01
1.74241066e-02 1.38969451e-01 5.20053618e-02 -6.33380562e-02
-7.66592175e-02 5.58064163e-01 -7.75133848e-01 2.46454611e-01
9.61405993e-01 1.20095873e+00 1.10050803e-02 3.14081848e-01
-4.42312747e-01 1.07708120e+00 -5.69795787e-01 -5.43771625e-01
-5.95437922e-02 -3.14421743e-01 -1.17523539e+00 -4.53663737e-01
1.01738259e-01 -2.27175742e-01 -6.11819923e-01 7.97055662e-01
5.84622741e-01 2.16324687e-01 4.90447968e-01 -1.00770700e+00
-8.16307008e-01 9.73355830e-01 -5.39590657e-01 2.71299426e-02
1.10873811e-01 5.53529784e-02 1.43017149e+00 -9.01607692e-01
5.51715374e-01 9.99405801e-01 8.97869468e-01 3.50847989e-01
-9.61252928e-01 -7.52977550e-01 3.10924053e-01 2.41270214e-01
-1.36755896e+00 -2.04157755e-01 7.47023880e-01 -1.54464066e-01
1.36473620e+00 -8.58323500e-02 5.89610457e-01 1.25980580e+00
2.45238140e-01 1.17211187e+00 9.14775312e-01 -3.30369532e-01
-1.87194750e-01 -5.21439910e-01 3.83886039e-01 6.59889400e-01
1.59029618e-01 4.10141386e-02 -4.75724071e-01 1.37917534e-01
5.59243143e-01 3.25042047e-02 -2.87717521e-01 3.00144851e-01
-1.34688234e+00 5.52669406e-01 7.70598948e-01 6.90211713e-01
-4.86876130e-01 1.90805390e-01 6.58814788e-01 -1.06645919e-01
5.83520412e-01 5.52607536e-01 -8.51351261e-01 -6.71683103e-02
-9.80602026e-01 -2.26563737e-01 7.99935579e-01 1.04354024e+00
5.69713414e-01 1.85078327e-02 -7.33378351e-01 7.43399024e-01
2.20255554e-01 1.76337972e-01 4.83402163e-01 -3.94436091e-01
6.65371478e-01 7.33956218e-01 -1.86084975e-02 -5.05125940e-01
-6.09745681e-01 -7.38652468e-01 -9.46955025e-01 -4.60792869e-01
1.43999070e-01 -5.02355695e-01 -9.54403758e-01 2.03711438e+00
2.46084347e-01 3.86977136e-01 3.32331918e-02 5.77068150e-01
1.05943310e+00 5.05450547e-01 6.06365919e-01 -2.28011742e-01
1.66593432e+00 -1.33300936e+00 -9.80796218e-01 -9.93473381e-02
9.85479832e-01 -4.72225755e-01 7.08207250e-01 -2.83796005e-02
-7.35792756e-01 -6.45507395e-01 -1.00187576e+00 -4.94148046e-01
-6.54448330e-01 4.27542508e-01 6.22028649e-01 2.85341084e-01
-7.60526061e-01 5.15128851e-01 -8.13062966e-01 -2.30116427e-01
3.82054716e-01 1.72112450e-01 -4.54710186e-01 7.96315894e-02
-1.79980123e+00 9.38216269e-01 7.86784649e-01 4.36370343e-01
-6.50487125e-01 -8.05070400e-01 -1.12613821e+00 6.02228701e-01
4.31261420e-01 -6.78897083e-01 1.20424318e+00 -1.53093576e-01
-1.32118034e+00 6.77801609e-01 -2.55405933e-01 -3.59295845e-01
1.70234665e-01 -6.47711456e-01 -6.14945531e-01 2.49360204e-02
1.63257241e-01 6.05216682e-01 1.19705550e-01 -8.96227956e-01
-5.68313003e-01 -1.38858289e-01 -2.55602133e-02 -1.14205807e-01
-5.33560455e-01 1.13365099e-01 -2.72293955e-01 -7.16140270e-01
-2.67181069e-01 -5.82969010e-01 -1.38639629e-01 -5.17158747e-01
-8.99230361e-01 -8.08483839e-01 6.57982707e-01 -9.54175055e-01
1.83493292e+00 -1.98184860e+00 -1.32097110e-01 -2.75615573e-01
2.81684071e-01 5.06166458e-01 -3.45079035e-01 8.56902063e-01
-3.30522597e-01 4.68402743e-01 -2.70381957e-01 -5.36009729e-01
1.84047788e-01 -2.37185329e-01 -3.20991635e-01 1.22994497e-01
6.33125365e-01 1.33091879e+00 -1.06848514e+00 -8.63585114e-01
-2.37277836e-01 3.18957627e-01 1.98161099e-02 4.58035141e-01
-1.94742978e-01 3.13094795e-01 -4.60383981e-01 5.09915709e-01
6.13881171e-01 -2.91307986e-01 3.15892041e-01 -4.31473315e-01
-1.36202216e-01 7.76839852e-01 -6.15621150e-01 2.00080609e+00
-6.39244914e-01 3.98162693e-01 -3.99179459e-01 -8.95775318e-01
9.21257973e-01 7.48061061e-01 4.88329440e-01 -4.97008920e-01
5.72931357e-02 1.23930134e-01 -1.00367509e-01 -3.52089554e-01
4.22385633e-01 -1.83458537e-01 -4.89255279e-01 4.53155458e-01
4.40590233e-01 6.07417464e-01 3.78552407e-01 2.08909705e-01
1.45124316e+00 3.47988933e-01 5.64036071e-01 6.72431216e-02
5.06273985e-01 -3.32366467e-01 1.11534488e+00 4.29769546e-01
-2.24850237e-01 2.64724582e-01 6.84560120e-01 -3.91243726e-01
-8.71042669e-01 -8.77052486e-01 -1.22647263e-01 1.09012735e+00
-9.72741768e-02 -5.79520226e-01 -6.72383189e-01 -1.42979801e+00
-3.82037014e-01 6.96262479e-01 -7.02026427e-01 -2.07854688e-01
-8.58624697e-01 -5.10075510e-01 1.16616905e+00 1.11362934e+00
9.02707338e-01 -1.52720582e+00 -5.96641377e-02 5.31388700e-01
-5.98715484e-01 -1.37751651e+00 -7.91478574e-01 2.77612150e-01
-7.11401045e-01 -9.27631080e-01 -7.58707047e-01 -8.35437059e-01
2.84423590e-01 -1.51336268e-01 1.26332355e+00 7.22313821e-02
-4.14205827e-02 -1.48812011e-01 -5.53193927e-01 -3.67220700e-01
2.27908567e-01 8.66879702e-01 -2.78882027e-01 -1.35642380e-01
5.93055129e-01 -6.82021201e-01 -4.46725935e-01 9.60337580e-04
-7.79415905e-01 -7.83079788e-02 1.09227872e+00 1.09603179e+00
3.30551118e-01 -2.55489290e-01 9.94664371e-01 -9.35702622e-01
4.59082514e-01 -5.30398428e-01 -1.15105473e-01 6.38863325e-01
-4.22812343e-01 2.18865052e-01 7.48289049e-01 -2.49992698e-01
-1.26146972e+00 -4.24129479e-02 -5.39941847e-01 -2.25257367e-01
-2.49671906e-01 7.49298573e-01 -7.11575210e-01 4.19717610e-01
2.13103339e-01 3.42998058e-01 -6.63347125e-01 -7.91151702e-01
4.27092552e-01 6.93914235e-01 4.53399360e-01 -5.99661827e-01
3.99506241e-01 3.51867937e-02 1.14137484e-02 -4.63219076e-01
-1.56291139e+00 -4.00836855e-01 -8.69058609e-01 2.24154681e-01
1.18083012e+00 -7.95235574e-01 -7.77665138e-01 6.29938841e-01
-1.54734993e+00 -2.31342390e-01 -1.99695826e-01 5.32862782e-01
-3.28366935e-01 3.81636590e-01 -1.17389095e+00 -8.55743885e-01
-6.83747053e-01 -6.94387555e-01 1.08340299e+00 3.62460732e-01
-9.50383842e-02 -9.83235717e-01 1.20881557e-01 8.13089535e-02
6.70183375e-02 2.89491206e-01 9.28799868e-01 -1.02812397e+00
-3.63808811e-01 -1.92566782e-01 -6.10312343e-01 1.58699393e-01
-5.30586438e-03 -3.75999480e-01 -1.05789208e+00 -3.93796489e-02
-3.40116680e-01 -4.57411885e-01 1.15481269e+00 -7.57945925e-02
1.24175489e+00 -2.32795656e-01 -5.18371642e-01 5.56518495e-01
1.44798183e+00 9.62012783e-02 7.26263225e-01 2.99900025e-01
1.03657269e+00 3.54386449e-01 4.61348504e-01 1.52568236e-01
5.88009894e-01 3.80711704e-01 1.96566001e-01 -2.79871404e-01
-2.81418890e-01 -5.62623322e-01 2.57477432e-01 9.58127618e-01
-7.90770128e-02 -3.83461148e-01 -8.47592533e-01 6.67713165e-01
-1.85228026e+00 -9.83813763e-01 8.59433934e-02 1.78912985e+00
1.10075331e+00 1.65362209e-01 -2.14256905e-02 -2.73117982e-02
7.09800661e-01 3.75403374e-01 -5.89225471e-01 -2.47683018e-01
5.73636554e-02 3.41522843e-01 3.62610012e-01 -2.53152810e-02
-1.35194349e+00 1.21161473e+00 5.40856886e+00 1.04223239e+00
-8.83385003e-01 2.02839017e-01 4.56250101e-01 3.86133820e-01
-1.40963778e-01 -3.32261138e-02 -1.10968578e+00 5.08595347e-01
1.00137377e+00 9.16063860e-02 -7.95006081e-02 7.32050896e-01
-2.20520556e-01 3.23553771e-01 -1.13708794e+00 5.97898424e-01
-2.24070132e-01 -1.00707686e+00 -6.12623841e-02 1.77965149e-01
6.50168896e-01 2.79324707e-02 -4.12844181e-01 8.36902201e-01
6.74622357e-01 -1.04276383e+00 4.88265365e-01 8.01163375e-01
1.06043720e+00 -8.88191521e-01 1.17757237e+00 4.23443615e-01
-1.77203929e+00 2.38956008e-02 -1.76710799e-01 2.66011119e-01
4.93349403e-01 8.95730615e-01 -5.06031275e-01 1.06584823e+00
4.03994292e-01 7.85752833e-01 -1.99784264e-01 1.04237306e+00
-7.60151029e-01 9.00618851e-01 -1.84677303e-01 -5.42581826e-03
3.07782441e-01 7.90451393e-02 2.69182295e-01 1.76230311e+00
2.54047006e-01 1.63294315e-01 3.14623564e-01 7.54525483e-01
-6.55185640e-01 1.38440832e-01 -3.87337416e-01 -3.83497775e-01
7.91759014e-01 1.46204138e+00 -6.48212135e-01 -3.80693048e-01
-4.94096339e-01 9.95209038e-01 1.08460236e+00 3.80785346e-01
-8.20450604e-01 -1.02438831e+00 4.82559413e-01 -4.72081900e-01
6.79775953e-01 -3.86284739e-01 -4.95394021e-01 -1.28228939e+00
7.77538046e-02 -4.60911959e-01 5.31781554e-01 -4.34588164e-01
-1.50387239e+00 8.46993625e-01 -4.62989926e-01 -1.02511740e+00
2.71471981e-02 -4.81451392e-01 -8.22793961e-01 1.01498711e+00
-1.63094091e+00 -1.67610025e+00 -2.84950677e-02 8.39672014e-02
2.66831160e-01 2.94596732e-01 9.76128936e-01 2.80593008e-01
-9.54400659e-01 1.02923834e+00 -6.66887239e-02 1.07106471e+00
7.41577268e-01 -1.36511993e+00 5.93595684e-01 9.13096189e-01
2.34196916e-01 8.84225726e-01 -1.16358615e-01 -7.59612441e-01
-8.19298208e-01 -1.27892399e+00 1.50606740e+00 -2.88616329e-01
6.02169156e-01 -4.51100558e-01 -8.53904545e-01 9.54154789e-01
3.38154882e-01 1.57593727e-01 9.49331582e-01 4.85958576e-01
-6.80526674e-01 -9.49200615e-02 -7.72568405e-01 3.70047182e-01
1.40887594e+00 -6.12391889e-01 -7.28502154e-01 5.25319688e-02
1.20293808e+00 -9.53365862e-02 -1.24807584e+00 5.41969597e-01
4.82154965e-01 -6.31461203e-01 7.62024641e-01 -9.65140402e-01
5.88206232e-01 -1.52764231e-01 1.73179805e-01 -1.26026058e+00
-5.53694606e-01 -3.46013010e-01 -5.01126051e-01 1.87126327e+00
6.19717300e-01 -5.69726527e-01 3.66695821e-01 2.48380467e-01
-4.17071015e-01 -1.27065885e+00 -8.26463938e-01 -1.06550574e+00
2.30141804e-01 -3.84559482e-01 1.00112498e+00 9.95422363e-01
1.91198587e-01 8.89173090e-01 -5.10230660e-01 -2.04049200e-02
2.94410527e-01 2.38304287e-01 1.89445093e-01 -1.15528965e+00
3.25486213e-02 -4.90755558e-01 -5.16095087e-02 -1.30521500e+00
5.39927423e-01 -8.82424593e-01 2.31829777e-01 -1.80361462e+00
2.54946947e-01 -4.58916605e-01 -6.87683403e-01 9.04816270e-01
-7.27845013e-01 1.76507737e-02 3.12173158e-01 5.12920953e-02
-9.83444214e-01 9.08037663e-01 1.16076648e+00 -2.81143058e-02
1.42183840e-01 -3.28191429e-01 -7.87570119e-01 4.80963618e-01
7.81759977e-01 -6.04505837e-01 9.32071954e-02 -4.29228693e-01
2.47050896e-01 3.35228369e-02 -8.43664259e-02 -9.18561876e-01
2.31802613e-01 1.30566219e-02 5.55010617e-01 -7.24692225e-01
6.84098601e-02 -5.88564813e-01 -4.87219751e-01 4.00954694e-01
-4.46660191e-01 -1.87023863e-01 -2.68956274e-02 6.21186912e-01
-2.70390958e-01 -2.43276387e-01 5.03125668e-01 -8.95039067e-02
-6.78316832e-01 7.14945257e-01 -1.59524530e-01 2.88141996e-01
7.89001882e-01 3.39366853e-01 -2.29040891e-01 9.16565955e-02
-5.40165305e-01 3.90828282e-01 -1.02146909e-01 2.77213514e-01
1.99964032e-01 -1.60880423e+00 -7.14345038e-01 -1.94340274e-01
1.14516251e-01 1.88127890e-01 3.52393419e-01 8.08058202e-01
-9.11051482e-02 6.29044652e-01 1.48173440e-02 2.64365263e-02
-7.06631124e-01 9.38726783e-01 8.37009698e-02 -1.31526566e+00
-4.40018564e-01 9.46147323e-01 2.44917914e-01 -5.51431000e-01
-8.41542240e-03 -2.59898126e-01 -4.63429093e-01 7.93306082e-02
5.10330737e-01 3.72674257e-01 7.61161298e-02 -4.42170918e-01
-3.80457520e-01 3.50627333e-01 -2.26233065e-01 7.67206447e-03
1.20052099e+00 -3.33844312e-02 -1.27752066e-01 4.32597458e-01
1.25971043e+00 -4.17843238e-02 -1.01389635e+00 -5.66061854e-01
4.41124141e-01 1.59611061e-01 -1.42191336e-01 -9.06532884e-01
-9.93108213e-01 1.17683637e+00 -1.36164188e-01 3.89995761e-02
1.17488170e+00 -1.14776306e-01 1.57500792e+00 3.57273847e-01
1.76588088e-01 -8.89353752e-01 -5.14802150e-02 9.06486630e-01
5.99498808e-01 -1.01263309e+00 -3.86920184e-01 -4.69876260e-01
-3.99980962e-01 1.04922664e+00 9.16722357e-01 3.92147563e-02
6.75044417e-01 3.38387847e-01 -7.80084282e-02 2.75725871e-02
-8.02939773e-01 -5.62251031e-01 3.61334711e-01 4.98630047e-01
8.39611948e-01 -3.93416546e-03 -7.11462736e-01 1.28030002e+00
-7.67198217e-04 4.52481024e-02 -2.22742438e-01 7.91018665e-01
-2.55980521e-01 -1.17055583e+00 1.06716976e-01 3.70124340e-01
-7.63297915e-01 -5.84664941e-01 -2.89770961e-01 6.46569729e-01
3.03065538e-01 8.48764718e-01 -1.51638061e-01 -5.00886202e-01
2.28454024e-01 4.72916603e-01 1.16070971e-01 -6.46406412e-01
-8.08047235e-01 -2.05966026e-01 4.17294323e-01 -4.12376195e-01
-1.81191862e-01 -3.63589585e-01 -1.39693809e+00 -5.63181266e-02
-5.63057899e-01 5.20897448e-01 3.86918008e-01 1.30159688e+00
4.73891705e-01 8.15786660e-01 5.39744973e-01 -4.62048352e-01
-5.79238296e-01 -1.52061486e+00 -5.60139596e-01 2.00342909e-01
6.66752234e-02 -6.06898129e-01 1.49026543e-01 -1.02050439e-01] | [9.500022888183594, 9.38221549987793] |
08a28078-7396-4246-8f6e-ab40672eefd8 | gecko-a-grammatical-and-discourse-error | null | null | https://aclanthology.org/2021.jeptalnrecital-demo.3 | https://aclanthology.org/2021.jeptalnrecital-demo.3.pdf | GECko+: a Grammatical and Discourse Error Correction Tool | GECko+ : a Grammatical and Discourse Error Correction Tool We introduce GECko+, a web-based writing assistance tool for English that corrects errors both at the sentence and at the discourse level. It is based on two state-of-the-art models for grammar error correction and sentence ordering. GECko+ is available online as a web application that implements a pipeline combining the two models. | ['Ajinkya Kulkarni', 'Miguel Couceiro', 'Maxime Amblard', 'Thibo Rosemplatt', 'Léo Jacqmin', 'Eduardo Calò'] | null | null | null | null | jep-taln-recital-2021-6 | ['sentence-ordering'] | ['natural-language-processing'] | [ 2.14961059e-02 8.42210114e-01 1.99474782e-01 -3.35742503e-01
-8.33688974e-01 -2.58062065e-01 2.47284696e-01 8.85261059e-01
-1.53590098e-01 7.98955381e-01 6.49226129e-01 -4.60459888e-01
-1.98075414e-01 -6.07917666e-01 -4.79132712e-01 5.84254563e-01
5.02840579e-01 7.12824404e-01 4.21471804e-01 -8.11422706e-01
1.02590346e+00 -4.24330264e-01 -1.16639471e+00 8.10732782e-01
1.28308249e+00 1.35877222e-01 3.70691985e-01 1.23276114e+00
-3.83874059e-01 1.24252808e+00 -8.03383887e-01 -6.26734495e-01
-6.93065703e-01 -4.38692212e-01 -1.37818420e+00 -5.30993998e-01
4.56072956e-01 -2.36861818e-02 -2.46840529e-02 9.06597376e-01
6.69718206e-01 -2.79535681e-01 4.04750817e-02 -5.07217586e-01
-9.08174574e-01 8.67646515e-01 5.36757767e-01 4.73113418e-01
1.04298627e+00 1.75083056e-02 7.84328580e-01 -1.28489923e+00
1.42906642e+00 1.24352181e+00 1.12522793e+00 8.96696925e-01
-9.71524060e-01 -7.91191682e-02 -1.72473833e-01 4.68681097e-01
-8.26913655e-01 -6.50881052e-01 5.34571826e-01 -6.53757811e-01
2.20018625e+00 4.12719637e-01 5.95497727e-01 9.82964814e-01
3.11603695e-01 5.24210155e-01 7.50772834e-01 -1.12938786e+00
-1.05171032e-01 -7.79890493e-02 5.73750556e-01 1.04201233e+00
1.27713412e-01 -2.39369497e-01 -1.26583612e+00 4.63932455e-02
-4.80134860e-02 -9.54127610e-01 -1.98077098e-01 6.74954951e-01
-7.87319958e-01 4.77458149e-01 -2.70154119e-01 3.99071366e-01
-6.94376007e-02 2.06910253e-01 8.78979683e-01 5.37788272e-01
8.52406383e-01 7.74125218e-01 -3.53473425e-01 -9.81971443e-01
-7.85582244e-01 8.52930427e-01 9.49655235e-01 1.10903656e+00
1.32144898e-01 -2.18854949e-01 -4.51186866e-01 8.86939168e-01
7.72313356e-01 4.14257236e-02 4.32050288e-01 -7.45703995e-01
1.17820752e+00 9.44988489e-01 1.67424679e-01 -1.08306062e+00
-4.09866005e-01 -2.93002445e-02 1.48421273e-01 -1.75130032e-02
5.77084541e-01 1.96525395e-01 -2.89216429e-01 1.32803476e+00
3.80118527e-02 -4.09251571e-01 1.03672244e-01 4.80686873e-01
1.59769821e+00 3.07273239e-01 2.39889082e-02 -7.15446994e-02
9.69458878e-01 -1.08314764e+00 -1.35673177e+00 -6.01041794e-01
1.64280391e+00 -1.10698903e+00 1.16236806e+00 1.67295799e-01
-1.49575388e+00 -1.93397492e-01 -1.07304740e+00 -7.17724085e-01
-5.12339175e-01 2.93344647e-01 9.41693932e-02 4.32258517e-01
-1.11292279e+00 1.02484405e+00 -9.25460458e-01 -7.54410625e-01
6.74695447e-02 -1.69967726e-01 -2.20209911e-01 1.50001168e-01
-1.18338907e+00 1.50451314e+00 5.59560537e-01 -9.27026942e-02
7.62680769e-02 -6.45805538e-01 -1.19567966e+00 -2.10795283e-01
3.69474828e-01 -5.40217698e-01 1.74601507e+00 -3.00203562e-01
-1.56854534e+00 1.22123384e+00 -5.49755752e-01 -2.49603540e-01
4.83931065e-01 -4.97841686e-01 -4.92669851e-01 -3.29406828e-01
2.58969337e-01 -2.56859541e-01 1.37291968e-01 -5.68016946e-01
-5.86111367e-01 -4.49025244e-01 -3.78904998e-01 2.64980167e-01
2.95429319e-01 8.28741431e-01 1.66402000e-03 -8.42379332e-01
3.83020639e-02 -2.34876931e-01 4.20967460e-01 -3.53066981e-01
-3.78963113e-01 -6.93202317e-01 5.28499007e-01 -1.86860371e+00
2.59011889e+00 -1.67787802e+00 3.39971721e-01 -1.33464187e-01
7.89699256e-02 7.07410276e-01 8.32671523e-02 9.52051044e-01
1.71894878e-01 5.76697767e-01 -3.57816517e-01 -9.78762627e-01
2.66476005e-01 1.18875660e-01 2.12171957e-01 -1.81828529e-01
3.12845469e-01 9.15087461e-01 -1.50486505e+00 -3.80975425e-01
2.98604280e-01 -9.81313959e-02 -5.09909511e-01 1.20053396e-01
-3.84339601e-01 1.83870882e-01 2.20467851e-01 4.50241864e-01
2.22369567e-01 4.93391268e-02 4.32869107e-01 7.12458968e-01
-3.78585100e-01 1.48945165e+00 -9.82114077e-01 1.89828527e+00
-1.56331658e-01 7.41630256e-01 -5.33318184e-02 -2.41187662e-01
8.65951180e-01 4.51922417e-01 -6.45858049e-01 -5.07103443e-01
3.61268818e-02 9.11567569e-01 -3.56516719e-01 -1.02890944e+00
1.20874274e+00 4.59689766e-01 -1.27648279e-01 4.76888359e-01
3.90848637e-01 -3.01008016e-01 7.51114547e-01 5.16447544e-01
1.36181211e+00 3.57782573e-01 8.59143794e-01 -1.72931358e-01
5.69345653e-01 3.31046820e-01 6.48355484e-01 8.83236885e-01
8.92971456e-02 5.15614510e-01 3.72980356e-01 -3.37903082e-01
-1.24977374e+00 -4.54035580e-01 6.70815632e-02 8.28989387e-01
-3.27950418e-01 -1.35867906e+00 -1.10057044e+00 -6.46394193e-01
-3.63241020e-03 1.33646405e+00 -5.59323311e-01 -4.34773266e-02
-6.94407582e-01 -3.01284432e-01 7.79229820e-01 4.12679344e-01
6.72024339e-02 -1.52444220e+00 -3.53962779e-01 6.00426078e-01
-4.91401404e-01 -7.35309482e-01 -4.02783126e-01 -3.36374253e-01
-3.24200004e-01 -1.41999185e+00 4.08801705e-01 -6.48287475e-01
3.37572336e-01 -4.82575864e-01 1.54440165e+00 1.09303415e+00
9.93618891e-02 2.37710074e-01 -9.09402132e-01 -7.61129737e-01
-1.52513623e+00 2.54122764e-01 -2.57343113e-01 -9.64545369e-01
6.23255551e-01 -2.37124786e-01 2.96577096e-01 -5.10626376e-01
-3.80629241e-01 4.34715867e-01 -4.97162551e-01 9.10638332e-01
2.87593082e-02 -6.12644970e-01 6.69692338e-01 -1.32667065e+00
1.10393858e+00 -3.00917685e-01 -3.51446748e-01 6.28904641e-01
-8.86029780e-01 -2.03192428e-01 3.31473976e-01 6.17165327e-01
-9.27977800e-01 -5.84284902e-01 -8.70268345e-01 8.18288624e-01
1.54219255e-01 9.65816081e-01 -6.94701001e-02 4.24850844e-02
5.01111984e-01 -1.98425725e-01 -1.43000007e-01 -7.27319598e-01
5.27760834e-02 1.06834328e+00 7.09384620e-01 -2.17283413e-01
2.68367268e-02 -6.60402477e-01 -4.38319564e-01 -5.80306292e-01
-7.33003676e-01 -1.77463576e-01 -8.17382157e-01 -5.05456924e-01
5.53637087e-01 -3.86824250e-01 -3.60364586e-01 7.08841562e-01
-1.74388373e+00 -6.69517875e-01 -3.70573431e-01 -1.47240832e-01
-4.64055866e-01 5.76077640e-01 -7.46576250e-01 -7.80699670e-01
-6.16213262e-01 -6.81574345e-01 9.29747880e-01 1.02451526e-01
-1.05077457e+00 -1.27048087e+00 2.61282265e-01 3.47118556e-01
1.75655127e-01 1.79472849e-01 9.73509490e-01 -6.54113054e-01
-4.83766943e-02 7.44739845e-02 6.49211630e-02 2.12274101e-02
-2.40064889e-01 1.80364773e-01 -2.76642531e-01 6.50944114e-02
-5.52976251e-01 -1.26662597e-01 4.62789208e-01 -8.63294452e-02
6.00263894e-01 -5.93468368e-01 -8.59379619e-02 4.58733559e-01
1.12535024e+00 7.11498410e-02 6.83221340e-01 1.00060546e+00
7.27471590e-01 5.90102255e-01 8.30195189e-01 5.30208588e-01
9.12200391e-01 9.22134578e-01 3.36147994e-01 6.65523589e-01
-4.28528786e-01 -6.19632483e-01 1.93159178e-01 1.34984314e+00
3.08757365e-01 -5.81652224e-01 -1.30458069e+00 8.30053270e-01
-2.40699315e+00 -9.05306637e-01 -9.18644607e-01 1.84549153e+00
1.12517238e+00 5.02073914e-02 -2.61436969e-01 4.85806376e-01
5.01285255e-01 -7.01298118e-02 2.07658768e-01 -1.30447245e+00
-3.51024181e-01 4.40372735e-01 9.64326486e-02 1.32426429e+00
-8.01519215e-01 1.17679429e+00 7.20125532e+00 5.21287560e-01
-6.65916741e-01 4.80293989e-01 -2.40265414e-01 1.76519826e-01
-2.85034150e-01 2.34813944e-01 -1.11868393e+00 9.86209810e-01
9.98943031e-01 -3.55782360e-01 4.23715740e-01 5.58597147e-01
5.24034917e-01 -3.55144083e-01 -9.86075997e-01 4.64250714e-01
2.79097885e-01 -1.98048162e+00 -1.79449737e-01 -5.70631742e-01
2.63143033e-01 -8.75559300e-02 -7.50303090e-01 1.99678302e-01
3.78023088e-01 -9.36115682e-01 1.36295009e+00 9.46129322e-01
6.12195134e-01 -3.61974150e-01 1.07446814e+00 4.55330133e-01
-4.68496561e-01 5.59012704e-02 9.02277529e-02 -5.04078150e-01
7.18979597e-01 3.99322510e-01 -9.69455302e-01 5.39674878e-01
7.18703210e-01 8.79235029e-01 -1.21796763e+00 7.12299228e-01
-8.16723883e-01 7.53020883e-01 9.48464572e-02 -1.42755806e-01
-4.43328530e-01 2.01796100e-01 1.21094990e+00 1.75042689e+00
3.65874201e-01 2.83633411e-01 -7.06431344e-02 6.68457508e-01
1.32021859e-01 2.08604470e-01 -3.13739806e-01 -1.20617718e-01
1.17965734e+00 5.91489792e-01 -9.10277441e-02 -6.21115386e-01
-1.44399911e-01 1.09042871e+00 8.39869440e-01 -3.84786010e-01
-1.98166251e-01 -7.93439627e-01 2.76894391e-01 3.51724863e-01
-2.00999361e-02 -2.89060861e-01 -8.36568594e-01 -1.11928117e+00
3.36655736e-01 -9.75580931e-01 3.40314478e-01 -1.07940650e+00
-8.72417331e-01 2.90349543e-01 -3.76297504e-01 -6.55135036e-01
-5.93656063e-01 -5.00011563e-01 -5.75766861e-01 9.64675903e-01
-1.05567789e+00 -1.22919106e+00 -4.57240373e-01 -1.31610528e-01
6.47758126e-01 -1.08668722e-01 1.09967041e+00 2.91342646e-01
-6.71176076e-01 6.17425203e-01 9.76144001e-02 6.45536035e-02
7.53042579e-01 -1.66297877e+00 1.14995778e+00 1.24293685e+00
-4.49735224e-01 7.39053905e-01 9.02908862e-01 -1.20748556e+00
-5.48673093e-01 -1.13795781e+00 2.55809426e+00 -1.03552735e+00
5.95340252e-01 -2.22167090e-01 -1.20927179e+00 9.48910177e-01
3.59789401e-01 -4.47865337e-01 4.51922923e-01 8.67422521e-02
1.83923408e-01 5.52311599e-01 -1.10311139e+00 2.59383857e-01
1.32512534e+00 -7.30366886e-01 -1.34055674e+00 5.08006632e-01
6.10178351e-01 -1.39334548e+00 -1.02572250e+00 1.15219332e-01
2.09258094e-01 -8.44183683e-01 1.85629427e-01 -6.12673879e-01
9.67796445e-01 -3.97278875e-01 4.74515408e-02 -1.64836943e+00
-2.23025069e-01 -9.86377478e-01 -5.96985340e-01 1.66875792e+00
6.73790455e-01 -5.63994229e-01 1.08572409e-01 6.00023270e-01
-9.46925640e-01 -4.71264303e-01 -1.24671960e+00 -5.98240912e-01
-2.09695786e-01 -6.64327621e-01 7.31987298e-01 8.45199704e-01
1.00015616e+00 3.74539047e-01 -1.59551755e-01 -8.53929892e-02
-6.02823161e-02 -7.65652835e-01 6.26057029e-01 -1.12364686e+00
-2.36490294e-01 -4.37547565e-01 -1.78978056e-01 -3.94341379e-01
-2.18108972e-03 -1.18622124e+00 7.88279623e-02 -1.83484519e+00
-3.02216381e-01 -1.25034124e-01 6.58199787e-01 5.88133752e-01
-4.44402814e-01 5.04323468e-03 2.88100541e-01 1.18381187e-01
-6.17803514e-01 2.35484451e-01 4.20219481e-01 1.20298989e-01
-4.61165488e-01 -5.28007030e-01 -6.71389043e-01 7.25954831e-01
6.80896640e-01 -7.17230022e-01 4.11646128e-01 -9.56130087e-01
9.35601354e-01 -1.85724899e-01 1.55151144e-01 -8.72298539e-01
2.89171636e-01 9.78864208e-02 -3.87622327e-01 -7.02175975e-01
-2.72124797e-01 2.01025352e-01 1.72920704e-01 3.82609874e-01
-5.11394680e-01 6.33935988e-01 8.72122347e-02 -4.10568295e-03
-1.53855264e-01 -9.25686657e-01 5.79681695e-01 -3.25728238e-01
-7.94672549e-01 -7.21360743e-01 -7.96344161e-01 1.81078926e-01
5.34671664e-01 -2.87202239e-01 -1.14961123e+00 -1.66587055e-01
-7.39110768e-01 3.60975474e-01 5.76701522e-01 5.48286796e-01
5.47309875e-01 -1.14013851e+00 -7.42312014e-01 1.41174495e-01
5.05693972e-01 -9.82862338e-02 -2.37973586e-01 7.68128335e-01
-1.24415565e+00 7.02707469e-01 9.68810171e-02 -9.00351927e-02
-1.82950008e+00 -1.23615041e-01 3.38073313e-01 -7.08159566e-01
-6.60620213e-01 8.72590721e-01 -1.14002955e+00 -9.91156340e-01
-7.94369206e-02 -2.68062651e-01 -4.18403000e-01 -2.81488657e-01
8.97611558e-01 1.00691783e+00 8.60166848e-01 -6.88968897e-01
-5.95821679e-01 5.07667139e-02 -2.44623292e-02 -9.73504037e-02
1.23061442e+00 -7.34017670e-01 -5.53154588e-01 5.77872217e-01
3.17246586e-01 2.41015181e-01 -2.44172707e-01 -4.28228714e-02
8.71655643e-01 -5.47615707e-01 -1.27698094e-01 -1.50831032e+00
2.71118224e-01 5.07002175e-01 1.11721888e-01 6.03845865e-02
4.60937440e-01 -1.82108134e-01 8.77130687e-01 2.29240716e-01
1.15731023e-01 -1.83684409e+00 -3.22813332e-01 1.42574060e+00
1.30155635e+00 -1.06718612e+00 -7.43213221e-02 -8.82651448e-01
-2.31955722e-01 1.23285949e+00 7.40420401e-01 8.76705945e-02
1.81354180e-01 -4.64028530e-02 1.72026545e-01 -4.01638627e-01
-9.07876968e-01 1.15182385e-01 2.12414622e-01 6.88176930e-01
1.08640313e+00 -3.30590010e-02 -1.29025984e+00 1.08548415e+00
-6.91645384e-01 3.66205543e-01 9.57691431e-01 1.14737177e+00
-2.64721662e-01 -1.51723957e+00 -2.84524590e-01 3.86594802e-01
-4.35364485e-01 -4.63678479e-01 -8.68074238e-01 5.47485828e-01
2.17943892e-01 1.23513973e+00 -2.94477846e-02 -4.27433074e-01
6.84277296e-01 3.11745882e-01 5.24305820e-01 -1.22286391e+00
-1.36607337e+00 -7.77434587e-01 1.12153566e+00 -7.42909372e-01
1.06220515e-02 -1.07301533e+00 -1.37564397e+00 -6.51508808e-01
-2.46400401e-01 7.55520450e-05 6.10281885e-01 1.23083282e+00
7.49752343e-01 7.00331628e-01 -2.35429496e-01 -5.70220649e-01
-1.90531388e-01 -1.50498617e+00 -1.77426666e-01 2.45202929e-01
1.39857963e-01 -2.96574324e-01 6.50198059e-03 -1.26003236e-01] | [11.084371566772461, 10.705199241638184] |
ed305093-db09-413d-bec3-4be2e8b043fd | a-coupling-enhancement-algorithm-for-zro2 | 2205.11145 | null | https://arxiv.org/abs/2205.11145v2 | https://arxiv.org/pdf/2205.11145v2.pdf | A Coupling Enhancement Algorithm for ZrO2 Ceramic Bearing Ball Surface Defect Detection Based on Cartoon-texture Decomposition Model and Multi-Scale Filtering Method | This study aimed to improve the surface defect detection accuracy of ZrO2 ceramic bearing balls. Combined with the noise damage of the image samples, a surface defect detection method for ZrO2 ceramic bearing balls based on cartoon-texture decomposition model was proposed. Building a ZrO2 ceramic bearing ball surface defect detection system. The ZrO2 ceramic bearing ball surface defect image was decomposed by using the Gaussian curvature model and the decomposed image layer was filtered by using Winner filter and wavelet value domain filter. Then they were fused into a clear and undamaged ZrO2 ceramic bearing ball surface defect image and detected. The experimental results show that the image denoising method of ZrO2 ceramic bearing ball surface defect based on cartoon-texture decomposition model can denoise while retaining the image details. The PSNR of image is 34.1 dB, the SSIM is 0.9476, the detection accuracy is 95.8%, and the detection speed of a single defect image is 191ms / img. This method can effectively improve the efficiency and accuracy of ZrO2 ceramic bearing ball surface defect detection. | ['Zhenhong Li', 'Wenjie Li', 'Xianqi Liao', 'Jiaqi Yi', 'Xin Zhang', 'Wei Wang'] | 2022-05-23 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [-6.28623515e-02 -3.42850626e-01 5.83143651e-01 5.81111491e-01
-3.05829167e-01 5.19254446e-01 -2.90977895e-01 -1.21263433e-02
-1.02683745e-01 1.59807310e-01 -2.11024791e-01 1.07604556e-01
2.21437626e-02 -1.21994257e+00 1.06808748e-02 -1.03519392e+00
-1.98490284e-02 1.53583258e-01 8.67517292e-01 -6.73194468e-01
5.04783511e-01 4.92809743e-01 -1.58627331e+00 5.05681217e-01
7.13665068e-01 1.31146812e+00 8.41939390e-01 4.51830536e-01
3.14875335e-01 5.40944099e-01 -9.53699827e-01 3.04796010e-01
1.32949442e-01 5.30765168e-02 -7.95981884e-02 5.54756284e-01
-7.00677693e-01 -3.64779472e-01 -2.93353021e-01 1.20878613e+00
5.88084519e-01 -2.92387605e-01 6.98649168e-01 -7.03913569e-01
-7.27417052e-01 -4.87504214e-01 -1.00240111e+00 5.27942419e-01
4.70330626e-01 8.83381441e-03 3.81386042e-01 -1.42133176e+00
6.90997183e-01 1.59775126e+00 4.82234657e-01 4.34241116e-01
-1.09064007e+00 -7.34968364e-01 -4.71795499e-01 7.73862958e-01
-1.37692380e+00 -2.63724141e-02 6.00605190e-01 -3.99970293e-01
4.98155206e-01 5.28244495e-01 8.01889718e-01 5.12313433e-02
8.62789392e-01 7.16194808e-01 9.63712931e-01 -6.34527206e-01
-3.60927843e-02 -3.48960698e-01 -7.41907656e-02 4.96558160e-01
5.55979431e-01 2.57871687e-01 -2.62502879e-01 2.71350235e-01
7.64264762e-01 3.50093901e-01 -4.19988543e-01 3.89779449e-01
-9.08557117e-01 5.96636593e-01 3.27473998e-01 3.91534925e-01
-1.41690537e-01 -2.86508888e-01 6.06360078e-01 6.83069587e-01
9.47499275e-01 -2.41304442e-01 1.16648808e-01 2.12375015e-01
-4.12425548e-01 1.31261617e-01 3.87121290e-01 6.86734676e-01
4.54640746e-01 2.50649363e-01 2.28101656e-01 1.28731549e+00
8.26397181e-01 8.54408443e-01 5.67201555e-01 -4.01165098e-01
-9.89258289e-02 5.24505615e-01 -1.13570824e-01 -1.22518945e+00
-3.19517441e-02 3.77003759e-01 -8.32239270e-01 9.98768628e-01
-2.08937190e-02 3.80309850e-01 -8.62483263e-01 2.02900082e-01
5.66604435e-01 -1.75671324e-01 -6.73228651e-02 9.96977389e-01
1.07067859e+00 7.55443513e-01 -6.35670483e-01 -1.73061252e-01
1.99897635e+00 -5.83994508e-01 -1.02156973e+00 2.04252064e-01
3.30525428e-01 -1.31888187e+00 1.00909615e+00 7.88563848e-01
-9.99876201e-01 -5.45316160e-01 -1.71751881e+00 1.22817375e-01
3.37898999e-01 5.88043749e-01 9.65498164e-02 6.82699442e-01
-6.77951813e-01 3.85518700e-01 -9.64680612e-01 -7.52841681e-02
3.81410331e-01 2.11506560e-01 -3.88136715e-01 -7.23111570e-01
-6.88521981e-01 7.39269197e-01 4.36639786e-02 2.73835599e-01
-6.86899245e-01 -5.83555341e-01 -6.17145240e-01 -2.52894014e-01
1.25207797e-01 -2.16850311e-01 1.06787491e+00 -3.31335574e-01
-1.63460183e+00 1.04456234e+00 2.92933509e-02 6.67731687e-02
3.18320751e-01 -6.70977309e-02 -7.38985419e-01 4.51122254e-01
6.29715562e-01 -3.30741018e-01 7.84291565e-01 -1.16251528e+00
-9.54915047e-01 -7.74192393e-01 -8.08761179e-01 3.27027813e-02
2.02252895e-01 4.15323585e-01 -2.66966552e-01 -8.78173411e-01
1.01575649e+00 -4.02815074e-01 -1.10265911e-01 1.12057760e-01
-2.68163145e-01 -4.77011770e-01 1.19703317e+00 -7.24126518e-01
9.58916485e-01 -2.68307328e+00 -4.82946515e-01 2.16407761e-01
2.98103243e-01 -1.71925709e-01 2.60348737e-01 4.70212042e-01
-1.38755113e-01 -4.29619700e-02 2.25282133e-01 -4.06003818e-02
-4.68630075e-01 -1.26530588e-01 7.92142004e-02 1.09437108e+00
-1.73086390e-01 1.46675959e-01 -4.90655363e-01 -3.43299657e-01
1.36878371e-01 4.22113597e-01 -2.11569697e-01 -5.34145758e-02
4.78678882e-01 1.35654867e-01 -4.55853581e-01 1.24816239e+00
1.28456819e+00 4.92342681e-01 -7.11505771e-01 -3.10924947e-01
-1.46143407e-01 -1.52391002e-01 -1.31291139e+00 1.00183094e+00
-1.74112469e-01 6.12288117e-01 6.64689004e-01 -8.91533792e-01
1.68688226e+00 4.09020424e-01 4.43971843e-01 -1.05187309e+00
8.70063677e-02 6.53662324e-01 -3.52419049e-01 -9.10675704e-01
1.14392638e-01 -7.62507141e-01 5.22401333e-01 -1.69530213e-01
-2.98040032e-01 -7.50558615e-01 1.25879839e-01 3.40279974e-02
7.75444567e-01 -4.73854512e-01 -2.97180504e-01 -5.79857528e-01
6.09069705e-01 1.85960278e-01 6.08734667e-01 2.18841553e-01
6.27430007e-02 1.04860556e+00 2.19903648e-01 -5.93449771e-01
-1.15371573e+00 -9.80556250e-01 -7.17545867e-01 3.17032337e-01
6.67441785e-01 -4.68364768e-02 -6.19543076e-01 3.38985592e-01
6.13914207e-02 3.53790373e-02 -3.90533507e-01 -4.16553319e-01
-3.06259483e-01 -8.80111814e-01 -5.50324610e-03 1.50206253e-01
7.32039928e-01 -1.02657413e+00 -1.08222388e-01 5.12239933e-01
5.59909754e-02 -6.82421625e-01 -1.93019107e-01 -6.71823800e-01
-1.23093843e+00 -1.46276045e+00 -6.72795057e-01 -1.30035651e+00
5.64355612e-01 5.85749626e-01 7.31841505e-01 7.36088276e-01
-1.03575516e+00 4.09114063e-01 -5.72221100e-01 -6.10478997e-01
-4.74199712e-01 -9.80996966e-01 1.13204680e-01 6.73694536e-02
4.32149887e-01 -1.95961550e-01 -8.05775046e-01 7.60422289e-01
-1.09491169e+00 -1.18643381e-01 2.66953975e-01 1.06284487e+00
6.25840902e-01 8.06609333e-01 5.18707633e-01 -1.93837702e-01
8.84695470e-01 3.74963582e-02 -5.49874842e-01 -5.26804268e-01
-6.87260628e-01 -8.32628250e-01 -2.59091258e-02 -2.30271205e-01
-9.76690173e-01 -4.80086803e-01 -1.71796873e-01 -3.12209455e-03
5.35495020e-02 5.22841752e-01 -1.15475588e-01 -3.80353443e-02
6.61620975e-01 2.55955935e-01 7.86401927e-01 -7.97086596e-01
-4.16177720e-01 1.06578851e+00 7.31516600e-01 -3.31460923e-01
5.44155359e-01 9.41366076e-01 -2.28663683e-01 -1.18988383e+00
3.30751238e-04 -8.43625784e-01 -1.87974870e-02 -6.99340522e-01
9.22383606e-01 -1.21787906e+00 -1.33379006e+00 1.12941885e+00
-1.02500403e+00 2.75957078e-01 -2.79204965e-01 8.17026794e-01
-2.43578479e-01 7.66131759e-01 -1.34587586e+00 -6.75717473e-01
-5.31594336e-01 -1.25560331e+00 7.68149912e-01 1.37640536e-02
5.96991479e-01 -4.74410832e-01 -3.30895066e-01 4.97950763e-01
2.47784778e-01 1.98944703e-01 6.78163409e-01 1.03179604e-01
-3.49586457e-01 -9.15584385e-01 -2.43590310e-01 6.18871033e-01
3.43387306e-01 -1.18574895e-01 -4.65276927e-01 -5.19221783e-01
1.20370150e+00 2.85176575e-01 8.36353958e-01 4.37642694e-01
4.50286865e-01 2.50550896e-01 -4.12728667e-01 1.24016523e-01
1.49812329e+00 6.09723449e-01 1.31434059e+00 1.03792489e+00
2.56446779e-01 2.78069437e-01 1.14472437e+00 2.81232119e-01
3.13879028e-02 6.06054008e-01 6.12946212e-01 -2.19875246e-01
-2.85012275e-01 2.85580873e-01 4.50342774e-01 9.97728407e-01
-3.06895226e-01 2.03041196e-01 -8.10450673e-01 6.91556513e-01
-1.09053671e+00 -7.31669605e-01 -9.42538023e-01 2.03996038e+00
4.56279188e-01 2.32873499e-01 -2.62233526e-01 9.19565201e-01
1.08843124e+00 -6.06293678e-01 -1.37026399e-01 -3.39408256e-02
-2.86313921e-01 3.08067501e-01 5.02963901e-01 3.81167620e-01
-5.93377054e-01 4.23284292e-01 5.42590857e+00 1.48344171e+00
-1.11554861e+00 -1.82927190e-03 2.85234272e-01 1.71440855e-01
1.92609467e-02 -2.04510793e-01 -5.89779794e-01 7.32390523e-01
3.86453718e-02 -2.67959177e-01 -1.99156180e-01 5.94799221e-01
5.61812699e-01 -7.47033298e-01 -3.83334339e-01 1.30182004e+00
-1.66922361e-01 -1.03056872e+00 -2.44143948e-01 8.22367966e-02
9.29563046e-02 -1.96220055e-01 -7.10214628e-03 -3.17817748e-01
-1.34845167e-01 -9.94895816e-01 8.39376390e-01 3.74072880e-01
9.14238214e-01 -7.06312120e-01 1.05108249e+00 6.38033897e-02
-1.19310844e+00 -1.07366316e-01 -9.06058550e-01 -2.96061963e-01
1.89881742e-01 1.20813560e+00 -7.10793376e-01 5.24738610e-01
1.50758851e+00 8.24750066e-01 -1.97161138e-01 1.47971117e+00
-2.13792279e-01 5.44457555e-01 -4.95452881e-01 1.95416808e-01
-2.77286828e-01 -6.36794567e-01 7.35520601e-01 5.20723939e-01
6.20951533e-01 4.86796439e-01 1.71778217e-01 4.58874136e-01
4.20432270e-01 4.74454671e-01 -2.68464029e-01 8.28808904e-01
4.44990069e-01 7.86757171e-01 -9.22745228e-01 -2.22048298e-01
-5.17141700e-01 7.23277271e-01 -7.91701615e-01 7.01916590e-02
-2.92887568e-01 -7.70692408e-01 4.45877045e-01 6.80707037e-01
3.92893970e-01 -2.44939521e-01 -7.33128548e-01 -8.49766016e-01
2.18359888e-01 -7.83205271e-01 2.28272140e-01 -9.29321587e-01
-1.13094628e+00 2.07450718e-01 -4.54200029e-01 -1.49540019e+00
9.79144394e-01 -8.11446130e-01 -8.00145566e-01 7.25961447e-01
-1.19106221e+00 -7.99753189e-01 -2.83755064e-01 6.61386549e-01
6.29442811e-01 -2.87976533e-01 3.95853013e-01 3.79575729e-01
-5.63037157e-01 -9.72157642e-02 4.68361765e-01 -1.91145036e-02
5.24721444e-01 -9.22002256e-01 4.33874242e-02 6.86192870e-01
-6.17672026e-01 -2.48576924e-02 8.45907569e-01 -7.34893262e-01
-1.37476158e+00 -2.80376613e-01 2.66798407e-01 1.80151671e-01
3.22079778e-01 -2.18038172e-01 -1.05592752e+00 1.32328823e-01
2.32784101e-03 1.91114932e-01 2.44402483e-01 -6.41662419e-01
2.75167257e-01 -2.27184653e-01 -1.51470757e+00 3.40192765e-01
1.19057506e-01 -2.47175485e-01 -5.73149681e-01 5.35726249e-01
4.04437631e-01 -6.18388057e-01 -1.29183233e+00 4.68921483e-01
2.54599512e-01 -9.58197474e-01 1.28311861e+00 2.16293469e-01
8.35603625e-02 -8.52244377e-01 -5.52609526e-02 -7.92071581e-01
-5.11200607e-01 5.41239753e-02 8.40002894e-01 8.14564884e-01
-5.35401329e-02 -7.59652615e-01 7.99665511e-01 -4.80901822e-02
-6.42689109e-01 -6.47575498e-01 -1.43994009e+00 -7.49762297e-01
-1.63464621e-01 -3.07511210e-01 1.89760044e-01 3.29410046e-01
1.57872289e-01 -2.56128376e-03 -7.67053887e-02 4.33010608e-01
4.92987484e-01 -3.14701587e-01 4.42254096e-01 -1.60985672e+00
4.74122167e-01 -6.32557869e-02 -1.15870094e+00 -1.07574332e+00
-6.51312292e-01 -4.85526592e-01 -7.32899681e-02 -1.69279563e+00
-2.89485633e-01 -5.21282792e-01 -8.56735483e-02 5.50004728e-02
-1.02340624e-01 5.53243637e-01 -6.86009079e-02 7.26027548e-01
2.63025016e-01 5.33316791e-01 1.62590563e+00 -4.29497033e-01
-7.76092708e-02 7.94455856e-02 -1.55528799e-01 3.68800849e-01
3.82227391e-01 -3.84067267e-01 1.12674989e-01 -1.61206633e-01
-1.24470063e-01 2.45652407e-01 4.21758682e-01 -1.19575083e+00
4.07870919e-01 2.45696187e-01 2.70916015e-01 -6.98142290e-01
2.79137433e-01 -9.37680364e-01 1.61707923e-01 1.10758674e+00
9.85946059e-01 -4.66155797e-01 2.23699465e-01 6.76069319e-01
-8.24492455e-01 -2.38675818e-01 1.17841184e+00 -7.92407766e-02
-7.19179928e-01 7.13072196e-02 -9.64744866e-01 -5.08024275e-01
1.31220686e+00 -9.85656440e-01 5.18380925e-02 1.79657146e-01
-1.04092467e+00 -1.86601534e-01 6.34054482e-01 2.22029448e-01
1.33141017e+00 -1.50424302e+00 -1.03449237e+00 7.60591507e-01
1.48857972e-02 2.38569111e-01 6.46941543e-01 9.49035108e-01
-1.46653116e+00 -3.02948922e-01 -2.47782990e-01 -9.93682265e-01
-1.56706202e+00 3.19952011e-01 5.86766005e-01 1.06671579e-01
-1.34979784e+00 1.01880455e+00 6.34785816e-02 1.34208187e-01
-3.80026311e-01 -1.72929496e-01 -3.38453501e-01 -4.06010039e-02
6.98527575e-01 7.85561383e-01 5.11586130e-01 -6.14785075e-01
-2.47093275e-01 8.56407881e-01 -1.74137920e-01 1.50093079e-01
1.57172370e+00 -2.62123704e-01 -7.00746357e-01 1.29655868e-01
9.97329772e-01 2.84199268e-01 -7.41750956e-01 1.70961484e-01
-6.16133548e-02 -9.38308418e-01 3.54754299e-01 -2.67319620e-01
-1.39394391e+00 6.60053313e-01 1.04270482e+00 4.62407202e-01
1.34769917e+00 -9.87938493e-02 8.43488634e-01 -1.27512366e-01
4.13466096e-01 -1.06935430e+00 2.90090084e-01 9.89808366e-02
1.17753887e+00 -8.90926600e-01 2.88674533e-01 -8.37867796e-01
-9.06977803e-02 1.61822891e+00 3.39858085e-01 -7.96489358e-01
1.00840211e+00 3.13740343e-01 3.21895033e-01 -7.37225056e-01
-3.50172520e-01 2.98318028e-01 -1.98811203e-01 7.23719001e-01
1.01388261e-01 -2.22431961e-02 -6.31386817e-01 4.89206702e-01
-6.46097511e-02 4.21512593e-03 8.38575065e-01 8.96878481e-01
-1.02956760e+00 -8.51752460e-01 -1.30563271e+00 4.93973255e-01
-7.92285621e-01 2.09679335e-01 6.82201326e-01 6.33849561e-01
-1.21459134e-01 1.48836088e+00 5.36465012e-02 -2.89518625e-01
9.68038499e-01 -6.92089975e-01 1.58994883e-01 -5.57354927e-01
1.59402907e-01 5.19631684e-01 -3.33884209e-01 -4.92542088e-01
-1.39103353e-01 -4.71159130e-01 -1.26417327e+00 -3.17878991e-01
-8.40991914e-01 2.53949314e-01 6.70156956e-01 5.63076317e-01
-4.12937487e-03 5.30055344e-01 7.43181467e-01 -7.02580214e-01
-1.29580796e-01 -1.14739859e+00 -1.74083006e+00 4.10295755e-01
3.77272546e-01 -9.31035221e-01 -7.75503099e-01 -4.96944264e-02] | [7.491899490356445, 1.6142016649246216] |
a92990ed-e9bf-4eec-8eb9-7c08fa3605f4 | sam-da-uav-tracks-anything-at-night-with-sam | 2307.01024 | null | https://arxiv.org/abs/2307.01024v1 | https://arxiv.org/pdf/2307.01024v1.pdf | SAM-DA: UAV Tracks Anything at Night with SAM-Powered Domain Adaptation | Domain adaptation (DA) has demonstrated significant promise for real-time nighttime unmanned aerial vehicle (UAV) tracking. However, the state-of-the-art (SOTA) DA still lacks the potential object with accurate pixel-level location and boundary to generate the high-quality target domain training sample. This key issue constrains the transfer learning of the real-time daytime SOTA trackers for challenging nighttime UAV tracking. Recently, the notable Segment Anything Model (SAM) has achieved remarkable zero-shot generalization ability to discover abundant potential objects due to its huge data-driven training approach. To solve the aforementioned issue, this work proposes a novel SAM-powered DA framework for real-time nighttime UAV tracking, i.e., SAM-DA. Specifically, an innovative SAM-powered target domain training sample swelling is designed to determine enormous high-quality target domain training samples from every single raw nighttime image. This novel one-to-many method significantly expands the high-quality target domain training sample for DA. Comprehensive experiments on extensive nighttime UAV videos prove the robustness and domain adaptability of SAM-DA for nighttime UAV tracking. Especially, compared to the SOTA DA, SAM-DA can achieve better performance with fewer raw nighttime images, i.e., the fewer-better training. This economized training approach facilitates the quick validation and deployment of algorithms for UAVs. The code is available at https://github.com/vision4robotics/SAM-DA. | ['Jia Pan', 'Changhong Fu', 'Guangze Zheng', 'Haobo Zuo', 'Liangliang Yao'] | 2023-07-03 | null | null | null | null | ['transfer-learning'] | ['miscellaneous'] | [-1.54378965e-01 -6.55308902e-01 -1.78534776e-01 -2.31721420e-02
-4.17939216e-01 -7.55749285e-01 3.92879426e-01 -6.31162345e-01
-2.44824901e-01 8.25768471e-01 -6.38947248e-01 -2.01386586e-01
-3.10465842e-01 -6.16848528e-01 -4.89190429e-01 -1.13363171e+00
-1.57019254e-02 2.89210916e-01 3.36703658e-01 -2.58593172e-01
-5.87540686e-01 5.16647279e-01 -1.72242880e+00 -3.71892065e-01
1.06168926e+00 1.07241797e+00 6.58313453e-01 5.71099341e-01
4.75646377e-01 3.14387716e-02 -5.86859345e-01 1.95562631e-01
7.80003428e-01 -1.56703651e-01 6.70051053e-02 2.90033698e-01
5.91818929e-01 -6.39124453e-01 -2.58892953e-01 1.22656178e+00
3.63809586e-01 2.64532328e-01 2.76939094e-01 -1.70029891e+00
-4.70371366e-01 -1.56109080e-01 -5.79448283e-01 3.52850467e-01
-2.67977327e-01 6.75701320e-01 4.80483621e-01 -8.30548406e-01
3.21012735e-01 7.47046947e-01 7.20465064e-01 6.91024959e-01
-7.70536661e-01 -1.15687430e+00 2.85106093e-01 9.23922434e-02
-1.44769895e+00 -1.39810652e-01 6.42909408e-01 -4.39789355e-01
3.78038734e-01 1.38932303e-01 1.06235397e+00 9.88595486e-01
-1.97209362e-02 6.94824338e-01 8.57730210e-01 8.12583566e-02
7.18121827e-02 -1.60562899e-02 -2.58316696e-01 7.89933085e-01
6.73123240e-01 6.78442359e-01 -1.95363145e-02 6.08278699e-02
8.24267983e-01 2.95170069e-01 -3.84662777e-01 -4.60614711e-01
-1.39264679e+00 5.40281534e-01 5.82909048e-01 -5.65334857e-02
-2.37375081e-01 -2.11584672e-01 3.32817733e-01 2.80134469e-01
4.05186385e-01 3.77334237e-01 -8.14252615e-01 4.98627387e-02
-1.09808719e+00 2.10800424e-01 3.63668144e-01 1.35838068e+00
6.98048472e-01 6.93739772e-01 -5.47909401e-02 4.35428739e-01
2.49836192e-01 1.26733935e+00 2.64607221e-01 -5.74824095e-01
1.39137119e-01 6.06515944e-01 5.15949190e-01 -7.67746985e-01
-4.97081727e-01 -9.45827365e-01 -6.54309571e-01 3.92078280e-01
3.04532796e-01 -6.23058379e-01 -9.63270605e-01 1.59447360e+00
7.22916663e-01 2.98536271e-01 1.97090492e-01 1.38125062e+00
7.09306002e-01 8.57756376e-01 8.53875950e-02 -4.62329090e-01
1.07920039e+00 -9.74379778e-01 -4.22806472e-01 -4.81331497e-01
3.56894851e-01 -5.02544403e-01 8.70408177e-01 2.20299587e-01
-2.92312026e-01 -8.99700940e-01 -1.18817234e+00 6.04964137e-01
-3.55249375e-01 8.82864356e-01 6.87474608e-01 4.09848511e-01
-5.24844050e-01 1.04833074e-01 -6.53353035e-01 -4.90486383e-01
5.90318501e-01 3.13749790e-01 -3.40780646e-01 -9.29586068e-02
-1.00716496e+00 5.03624678e-01 6.39625549e-01 1.80822492e-01
-1.56867480e+00 -9.82167482e-01 -6.28884733e-01 -3.12091023e-01
6.99381113e-01 -6.08062267e-01 1.31862128e+00 -1.08250368e+00
-1.32404709e+00 7.01606691e-01 2.69349277e-01 -5.60276449e-01
2.36350566e-01 -3.08010697e-01 -7.11775959e-01 2.44880598e-02
3.18239689e-01 8.74035060e-01 1.27463126e+00 -1.10970974e+00
-1.17302346e+00 -4.05240804e-01 1.17808981e-02 1.16441332e-01
-3.91062051e-01 -2.98311025e-01 -5.94004281e-02 -8.51239800e-01
-8.62653702e-02 -1.19171226e+00 2.48422585e-02 4.41562593e-01
9.73666981e-02 -2.26181298e-01 1.45640159e+00 -3.37172389e-01
1.03797865e+00 -2.11840582e+00 7.13433279e-03 -4.80781198e-01
6.60328642e-02 7.82550037e-01 -1.46908164e-01 9.04336348e-02
7.29795769e-02 -5.49863577e-01 -1.84970677e-01 1.12610616e-01
-1.99859813e-01 8.28322023e-02 -5.35220146e-01 6.33903384e-01
2.47258663e-01 7.28547156e-01 -1.15823030e+00 -2.41738990e-01
4.66335475e-01 1.71253666e-01 -1.01105452e-01 3.06059301e-01
-5.51987469e-01 7.11747289e-01 -5.80986142e-01 1.29578698e+00
9.81930614e-01 -1.09156147e-01 -1.54091403e-01 -4.42270488e-01
-5.92301309e-01 -5.67865610e-01 -6.80741906e-01 1.54821515e+00
-7.70342648e-02 8.38561833e-01 2.62775004e-01 -5.47537386e-01
1.17239201e+00 -1.34861551e-03 5.76269925e-01 -4.85045493e-01
3.28977287e-01 2.14421526e-01 -2.48163864e-02 -2.86061972e-01
4.94288534e-01 -2.97479760e-02 -1.24006554e-01 -2.57462323e-01
-5.75211197e-02 -3.65589224e-02 1.13515839e-01 -1.87413424e-01
6.85330987e-01 4.75859612e-01 2.22628832e-01 -2.61811823e-01
5.44482470e-01 7.71701336e-01 1.05442178e+00 3.67828667e-01
-7.17883587e-01 1.63386881e-01 -4.78832960e-01 -5.50798774e-01
-1.01376045e+00 -1.00633776e+00 -1.72058880e-01 8.96151543e-01
5.91767073e-01 4.09451760e-02 -5.70707738e-01 -6.65186405e-01
1.63339674e-01 4.05691683e-01 -4.19334948e-01 -3.76564801e-01
-2.19185367e-01 -7.07034767e-01 5.78465998e-01 4.47905719e-01
9.07711983e-01 -7.32178032e-01 -1.04676247e+00 9.13545191e-02
9.31067299e-03 -1.09861243e+00 -2.96390951e-01 -7.38168806e-02
-7.51887321e-01 -1.25457740e+00 -9.01517034e-01 -7.33002424e-01
5.65730810e-01 1.04307175e+00 5.17666578e-01 -2.98211485e-01
-5.07374048e-01 6.29205406e-01 -4.23732132e-01 -8.06777477e-01
1.40463272e-02 -1.68875679e-01 6.06880307e-01 -8.30510333e-02
5.85071743e-01 -2.09263936e-01 -5.77921987e-01 5.58198571e-01
-6.32938743e-01 -5.28283268e-02 7.55468786e-01 9.23991859e-01
7.85763860e-01 4.17295158e-01 5.62581778e-01 -1.29856557e-01
1.45181194e-01 -5.60132384e-01 -1.45892096e+00 2.07663998e-01
-5.93488991e-01 -5.78147769e-01 9.52278376e-01 -8.03941727e-01
-7.84329593e-01 3.55795115e-01 5.04363298e-01 -1.28482938e+00
-2.80278563e-01 2.44806677e-01 -1.71212673e-01 -3.94040763e-01
6.94527626e-01 4.94577229e-01 2.20694214e-01 -1.83749095e-01
6.32349774e-03 5.67667425e-01 7.54456222e-01 -3.56968105e-01
1.47267532e+00 5.72031140e-01 1.26083747e-01 -9.12802637e-01
-1.03767931e+00 -8.43045712e-01 -4.35683995e-01 -4.99382526e-01
7.62994468e-01 -1.51076782e+00 -4.20209825e-01 6.44047320e-01
-8.14446509e-01 -5.38985908e-01 -3.58412415e-01 6.61923766e-01
-3.98943901e-01 1.13697566e-01 7.48279318e-02 -8.56435657e-01
-4.88518596e-01 -8.31116199e-01 9.95730281e-01 8.65499973e-01
5.49931705e-01 -7.65107632e-01 -9.62047055e-02 -4.55098115e-02
2.14440182e-01 4.07444358e-01 4.32196930e-02 -2.03471825e-01
-8.16510797e-01 -3.25150579e-01 -3.89591724e-01 4.17349547e-01
3.83725137e-01 7.45300204e-02 -9.01492417e-01 -9.26423907e-01
-1.04297414e-01 -6.99905157e-02 6.57849014e-01 5.66375732e-01
7.88947701e-01 -3.81076038e-02 -6.89947426e-01 8.79603207e-01
1.34772277e+00 4.82026130e-01 6.28538951e-02 5.54454803e-01
7.48941243e-01 1.04996592e-01 1.88786757e+00 7.06022918e-01
2.19694957e-01 5.84683120e-01 8.18338037e-01 -1.76031321e-01
-3.66930850e-02 -2.25978807e-01 6.31164908e-01 2.01557428e-01
6.74944371e-02 -1.82787195e-01 -5.79252005e-01 6.88078463e-01
-1.80158138e+00 -9.43206787e-01 -1.37577532e-02 2.29028034e+00
3.35265607e-01 -1.59609511e-01 3.15918565e-01 -2.36430034e-01
7.45911002e-01 1.48049951e-01 -1.01726210e+00 5.47920644e-01
-1.64190039e-01 -3.81768405e-01 8.97954941e-01 -1.82925224e-01
-1.62418318e+00 1.07705724e+00 4.50569630e+00 8.91099215e-01
-1.32097125e+00 8.53386745e-02 -2.37066165e-01 2.11203787e-02
4.42258775e-01 -5.23997657e-02 -1.35016966e+00 6.48628175e-01
7.74488986e-01 -4.92634386e-01 4.10998970e-01 1.41076303e+00
3.10901731e-01 8.66362453e-02 -4.92916077e-01 1.10740995e+00
3.25766094e-02 -1.11951971e+00 1.42070940e-02 1.54892504e-01
7.03898966e-01 1.67193636e-01 2.10144579e-01 3.82312745e-01
1.02475367e-01 -2.66076565e-01 6.20019019e-01 4.32078660e-01
1.13747346e+00 -6.48696721e-01 5.35940468e-01 5.06147385e-01
-1.62228930e+00 -7.17130005e-01 -9.23950076e-01 1.06285922e-01
-1.15846723e-01 4.01957124e-01 -9.82441068e-01 7.23875284e-01
9.49585140e-01 1.18520761e+00 -6.04138255e-01 1.38958764e+00
-4.96012457e-02 5.95631778e-01 -3.29825968e-01 9.58999917e-02
5.21076262e-01 -1.97421685e-01 1.17061257e+00 7.01925933e-01
6.99922562e-01 2.85486132e-01 7.19287097e-01 6.25429809e-01
2.98737466e-01 -4.44313675e-01 -7.95585394e-01 -2.51702756e-01
6.66640639e-01 1.62260914e+00 -3.96542668e-01 -1.94353983e-01
-2.10440829e-01 4.69683528e-01 -2.96140701e-01 1.38966992e-01
-1.27305484e+00 -1.94022611e-01 1.07744873e+00 6.86343536e-02
4.28453177e-01 -5.76188624e-01 3.91377032e-01 -1.05430007e+00
-2.45035380e-01 -8.04337144e-01 5.32723188e-01 -1.03377593e+00
-1.31056356e+00 7.84278572e-01 5.97715937e-02 -2.36897397e+00
1.22635111e-01 -9.83142376e-01 -4.44658279e-01 5.51211298e-01
-1.70436525e+00 -1.73997605e+00 -1.06908619e+00 6.57132268e-01
9.53442752e-01 -6.66402102e-01 5.44580281e-01 3.05363625e-01
-7.31669664e-01 3.94695103e-01 9.54222158e-02 2.41925754e-03
8.56520891e-01 -7.24135280e-01 1.09514713e-01 1.28315961e+00
-1.57113478e-01 2.72818178e-01 7.08076715e-01 -8.96458268e-01
-1.65404284e+00 -1.76709616e+00 -3.08216035e-01 -2.42186755e-01
7.43591428e-01 7.95940310e-02 -6.10466897e-01 5.42025685e-01
-1.89253837e-01 3.22054327e-01 3.71481508e-01 -4.67516214e-01
-1.66924205e-02 -6.50740802e-01 -1.08682346e+00 3.41597050e-01
8.82500112e-01 -5.35536073e-02 -3.60053778e-01 5.27376533e-01
9.22428846e-01 -3.76700610e-01 -7.91778624e-01 7.10733891e-01
5.53969085e-01 -6.86096907e-01 8.73799980e-01 -3.41461509e-01
-2.16194898e-01 -1.12209320e+00 -1.28751501e-01 -1.38625944e+00
-5.78341782e-01 -5.32756984e-01 -1.21757016e-01 1.02903044e+00
-1.02808692e-01 -7.86138535e-01 7.00041234e-01 -3.95162813e-02
-6.40148461e-01 -2.65876770e-01 -7.39216089e-01 -1.46257520e+00
-2.91451842e-01 1.10640027e-01 6.31507933e-01 7.75919318e-01
-7.23798811e-01 -8.56172144e-02 -5.09912014e-01 8.18420827e-01
9.14084375e-01 3.39997560e-01 1.01599205e+00 -1.69754779e+00
4.57884893e-02 1.40991732e-01 -3.09772700e-01 -9.99879718e-01
5.34539744e-02 -5.59127748e-01 2.41384029e-01 -1.25346994e+00
-3.83373588e-01 -6.22947097e-01 -1.36422247e-01 4.28834081e-01
6.61913976e-02 1.99748412e-01 1.66457519e-01 3.55596870e-01
-7.06701875e-01 6.57444775e-01 1.46302426e+00 -2.55657107e-01
-2.21602783e-01 2.92949975e-01 -2.74965554e-01 4.56235737e-01
9.65641916e-01 -3.67365867e-01 -6.31745815e-01 -3.25639158e-01
-4.08256978e-01 -6.00317046e-02 8.16632271e-01 -1.68990326e+00
3.47925544e-01 -3.83005887e-01 6.52912080e-01 -8.65099132e-01
4.55333143e-01 -1.15662229e+00 2.97745287e-01 6.73384488e-01
6.05192840e-01 -1.09566003e-01 6.97713614e-01 7.69067049e-01
-1.64214298e-01 9.71251801e-02 9.34856117e-01 -1.62462741e-01
-1.52387571e+00 8.42614532e-01 1.51092988e-02 -1.41642541e-01
1.58130157e+00 -5.29894948e-01 -6.55509710e-01 1.42122060e-01
-3.16403627e-01 3.75393063e-01 6.57695889e-01 5.92472970e-01
5.67016661e-01 -1.21353865e+00 -5.46615124e-01 5.39881229e-01
4.05914068e-01 1.39371157e-01 4.75180835e-01 1.03165174e+00
-2.75462180e-01 5.15560031e-01 -5.99791288e-01 -8.50717247e-01
-1.33663857e+00 7.60558307e-01 3.64909798e-01 3.19159657e-01
-5.19579411e-01 7.66205788e-01 2.78739303e-01 -2.11431265e-01
-2.11085752e-01 -1.56403571e-01 -5.02199195e-02 -1.72747183e-03
6.35907888e-01 3.80976290e-01 -2.85665751e-01 -6.97627842e-01
-5.11591911e-01 6.51591957e-01 1.75026760e-01 5.25622427e-01
1.12501395e+00 -1.31107852e-01 1.78933412e-01 8.01156834e-02
4.97785360e-01 -2.74149686e-01 -2.05455995e+00 1.33982629e-01
-2.77524829e-01 -6.33245289e-01 1.41836733e-01 -7.41218805e-01
-9.65548813e-01 6.83251500e-01 1.03060949e+00 6.55980632e-02
1.39654446e+00 -4.71695930e-01 8.82078230e-01 5.05165756e-01
8.31944048e-01 -9.26332831e-01 -5.35926148e-02 4.02392477e-01
6.14751577e-01 -1.47903085e+00 1.05249815e-01 -1.50736034e-01
-6.28151655e-01 1.14420307e+00 1.05444348e+00 -1.59828275e-01
3.04372609e-01 7.15080127e-02 1.42052934e-01 -4.17048447e-02
-6.49251640e-01 -5.92500627e-01 3.41977805e-01 1.08959651e+00
-3.71180981e-01 2.10171133e-01 2.31641755e-01 4.60168153e-01
4.25272658e-02 1.52111381e-01 2.69426942e-01 7.06309795e-01
-7.47730017e-01 -5.91213524e-01 -4.34198350e-01 2.67305374e-01
1.39262468e-01 1.56759828e-01 -2.93498561e-02 9.93670285e-01
2.89194524e-01 9.37298596e-01 -8.39501917e-02 -6.47777140e-01
2.74143428e-01 -4.58337098e-01 4.13307130e-01 -3.85143757e-01
-8.88512209e-02 -5.39283222e-03 -2.07248122e-01 -3.08276325e-01
-4.09434915e-01 -5.85822821e-01 -1.07572782e+00 -1.12935930e-01
-5.27416706e-01 1.86926320e-01 6.06672049e-01 6.28631175e-01
5.45463979e-01 3.80269229e-01 8.34321260e-01 -9.39437449e-01
-5.38661361e-01 -6.38765335e-01 -5.63106120e-01 -2.86208540e-01
5.24571598e-01 -1.19958186e+00 -3.98709416e-01 2.81585548e-02] | [6.921438694000244, -1.861168622970581] |
d9e71850-0f15-4720-87c6-18797b17f4a1 | modularized-textual-grounding-for | 1904.03589 | null | https://arxiv.org/abs/1904.03589v2 | https://arxiv.org/pdf/1904.03589v2.pdf | Modularized Textual Grounding for Counterfactual Resilience | Computer Vision applications often require a textual grounding module with precision, interpretability, and resilience to counterfactual inputs/queries. To achieve high grounding precision, current textual grounding methods heavily rely on large-scale training data with manual annotations at the pixel level. Such annotations are expensive to obtain and thus severely narrow the model's scope of real-world applications. Moreover, most of these methods sacrifice interpretability, generalizability, and they neglect the importance of being resilient to counterfactual inputs. To address these issues, we propose a visual grounding system which is 1) end-to-end trainable in a weakly supervised fashion with only image-level annotations, and 2) counterfactually resilient owing to the modular design. Specifically, we decompose textual descriptions into three levels: entity, semantic attribute, color information, and perform compositional grounding progressively. We validate our model through a series of experiments and demonstrate its improvement over the state-of-the-art methods. In particular, our model's performance not only surpasses other weakly/un-supervised methods and even approaches the strongly supervised ones, but also is interpretable for decision making and performs much better in face of counterfactual classes than all the others. | ['Shu Kong', 'Charless Fowlkes', 'Zhiyuan Fang', 'Yezhou Yang'] | 2019-04-07 | modularized-textual-grounding-for-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Fang_Modularized_Textual_Grounding_for_Counterfactual_Resilience_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Fang_Modularized_Textual_Grounding_for_Counterfactual_Resilience_CVPR_2019_paper.pdf | cvpr-2019-6 | ['phrase-grounding', 'natural-language-visual-grounding'] | ['natural-language-processing', 'reasoning'] | [ 4.08350736e-01 4.06108797e-01 -4.76947546e-01 -3.15981269e-01
-6.76736116e-01 -6.20611787e-01 8.90995741e-01 1.88707963e-01
-1.93749711e-01 8.61339688e-01 1.32539093e-01 -5.70229709e-01
3.59419361e-02 -7.54289806e-01 -9.26553607e-01 -4.21209514e-01
7.66354725e-02 3.21732908e-01 2.16198012e-01 -2.40752131e-01
-4.80552465e-02 7.37162679e-02 -1.61973524e+00 6.03931069e-01
1.10866809e+00 1.28039110e+00 -2.21239224e-01 2.20138520e-01
-8.30779523e-02 1.16124654e+00 -4.88333732e-01 -1.00906897e+00
2.99066275e-01 -2.71599650e-01 -6.65058255e-01 2.53056109e-01
8.28156888e-01 -2.68425167e-01 -2.10701287e-01 1.34476328e+00
1.99847013e-01 -3.02965581e-01 5.11819124e-01 -1.75593770e+00
-1.21979368e+00 6.92655802e-01 -6.28448367e-01 -2.82690227e-01
1.93352878e-01 6.27248764e-01 1.37829030e+00 -8.56764317e-01
6.74409270e-01 1.45375967e+00 6.47570014e-01 6.04749918e-01
-1.43944430e+00 -6.05816603e-01 5.94634950e-01 1.90734610e-01
-1.16637063e+00 -3.74799371e-01 9.15351748e-01 -5.12018323e-01
6.77528262e-01 3.92113328e-01 5.69541574e-01 1.36438942e+00
-8.79587382e-02 1.05073285e+00 1.50936961e+00 -2.52095819e-01
4.76053387e-01 1.15179881e-01 -1.46288455e-01 1.08405912e+00
4.66476947e-01 2.50635982e-01 -6.68901324e-01 -3.56016792e-02
4.60932881e-01 -3.94620523e-02 -4.08233881e-01 -6.68822289e-01
-1.62189543e+00 6.81659162e-01 8.83304596e-01 -1.25119135e-01
-1.68400824e-01 3.90357286e-01 2.87216425e-01 1.85083181e-01
7.16259718e-01 5.15929461e-01 -1.88349187e-01 4.92298484e-01
-9.53738153e-01 3.23613852e-01 5.05710363e-01 9.83611345e-01
7.49322653e-01 -2.09914483e-02 -2.86271960e-01 4.94007200e-01
4.24569488e-01 5.03855288e-01 2.56121933e-01 -7.83299685e-01
7.75922894e-01 8.99887562e-01 1.66942030e-01 -1.08407128e+00
-3.61785799e-01 -5.37974596e-01 -9.93268251e-01 5.88709295e-01
4.06082630e-01 3.84712741e-02 -1.02375150e+00 1.81874228e+00
2.93706775e-01 8.04953203e-02 3.34731638e-02 1.03454077e+00
8.48919153e-01 3.98947686e-01 3.86614889e-01 -1.15016513e-01
1.39046776e+00 -1.01159811e+00 -8.39059949e-01 -4.80272830e-01
6.36979461e-01 -4.43555534e-01 1.70054817e+00 1.74369082e-01
-7.54096389e-01 -4.41347033e-01 -1.37967396e+00 -1.03564158e-01
-4.74117219e-01 1.40306562e-01 8.76366377e-01 7.07159042e-01
-8.68258595e-01 4.97337699e-01 -4.50968981e-01 -6.55444786e-02
8.20589364e-01 7.52096400e-02 -3.78319740e-01 7.92504475e-02
-1.20656538e+00 8.12626481e-01 5.83359480e-01 -2.50579938e-02
-8.95205021e-01 -7.04066336e-01 -1.01851773e+00 -6.33302480e-02
6.48140907e-01 -1.10841990e+00 1.18633389e+00 -1.26021516e+00
-1.31653321e+00 1.04109299e+00 2.54893064e-01 -5.94278812e-01
1.05276477e+00 -1.78221822e-01 -3.03255618e-01 7.64744580e-02
2.86374867e-01 8.25755179e-01 8.24037910e-01 -1.59346175e+00
-6.86361074e-01 -3.78493696e-01 6.13254189e-01 2.01916203e-01
-2.88796306e-01 -3.27076972e-01 -3.15694809e-01 -8.78285289e-01
-7.19580948e-02 -9.04743314e-01 -1.66327208e-01 5.69387436e-01
-6.50917888e-01 -2.02383790e-02 8.66504967e-01 -3.13813686e-01
1.07398236e+00 -2.25717592e+00 -1.95119351e-01 -1.42103106e-01
4.37578738e-01 4.96121421e-02 3.25058401e-03 3.91892381e-02
-7.59530813e-04 3.97080332e-01 -4.56874430e-01 -2.88417280e-01
3.56964588e-01 1.64113253e-01 -6.61572814e-01 4.28240776e-01
4.51682448e-01 1.12194014e+00 -1.27786565e+00 -8.05176377e-01
3.00718099e-01 -6.57594725e-02 -5.15032113e-01 1.48863450e-01
-6.26808405e-01 1.72004327e-02 -2.66784072e-01 7.83303738e-01
3.97184670e-01 -5.13432980e-01 2.51295835e-01 -4.00848508e-01
5.10930363e-03 2.40306370e-02 -1.18483448e+00 1.61875856e+00
-4.05265361e-01 4.70646918e-01 -1.82522178e-01 -8.01706195e-01
5.92872322e-01 1.87368453e-01 1.04769208e-01 -6.93735838e-01
-7.93031156e-02 2.37785414e-01 -1.89985111e-01 -5.48772573e-01
4.28175807e-01 -4.29594606e-01 -1.28279611e-01 3.51209223e-01
-3.98937240e-02 -1.84118927e-01 7.86526203e-02 3.38563621e-01
6.67990804e-01 4.83714730e-01 4.42512631e-01 -1.74982727e-01
1.92362249e-01 2.68734097e-01 5.75291753e-01 8.32871318e-01
-2.72960186e-01 7.29879737e-01 6.25480831e-01 -5.31805336e-01
-1.00526130e+00 -1.20327628e+00 1.91257477e-01 1.13171697e+00
4.85219300e-01 -3.19079399e-01 -7.42271304e-01 -1.10248697e+00
1.34511292e-01 8.11221421e-01 -7.84389436e-01 -3.11465979e-01
-1.34901732e-01 -7.63572931e-01 7.00751603e-01 6.70997381e-01
8.54367971e-01 -8.07899594e-01 -6.18864715e-01 -7.49473646e-02
-3.59812498e-01 -1.20835018e+00 -3.56953770e-01 -1.52459800e-01
-8.05055857e-01 -1.17000985e+00 -4.23169255e-01 -1.82361692e-01
6.64848089e-01 2.81379431e-01 1.23524022e+00 1.22723266e-01
3.81758660e-02 6.39165640e-02 -1.67213127e-01 -7.95928717e-01
-2.67380595e-01 -2.82546520e-01 1.14321457e-02 3.12136769e-01
-7.28937462e-02 -2.60676861e-01 -6.61433458e-01 2.44635150e-01
-1.04008222e+00 5.74669719e-01 7.27193177e-01 9.20061350e-01
6.90773785e-01 -8.99021626e-02 4.84640270e-01 -1.22236824e+00
5.24222076e-01 -2.52496809e-01 -6.01924539e-01 4.82727200e-01
-9.32063758e-01 1.82318330e-01 6.89299464e-01 -4.05971259e-01
-1.28340745e+00 1.25021979e-01 2.91537017e-01 -4.29755211e-01
-7.42180869e-02 4.75340873e-01 -4.96714741e-01 1.71198264e-01
1.04890084e+00 -1.26215219e-01 -1.96076751e-01 -1.25746906e-01
8.70227993e-01 4.79176283e-01 8.49477649e-01 -6.19184077e-01
9.71861839e-01 9.46043253e-01 -7.19083101e-03 -3.88942450e-01
-1.19275773e+00 -5.66950813e-02 -3.90928924e-01 -2.85695195e-01
7.32626736e-01 -8.73897851e-01 -6.98624969e-01 1.91183269e-01
-1.11371279e+00 -3.53959382e-01 -4.20251340e-01 2.12891296e-01
-5.91683805e-01 4.34072495e-01 -2.05891043e-01 -8.77570033e-01
-1.98969245e-01 -9.09120440e-01 1.14035439e+00 -1.40931994e-01
-1.99379921e-01 -8.15546870e-01 -3.60116243e-01 5.59991241e-01
1.42662331e-01 6.52283728e-01 9.08095837e-01 -1.79802030e-01
-6.84296489e-01 1.29978498e-03 -6.08004451e-01 1.95374936e-01
-3.62001769e-02 1.00252405e-01 -1.34635019e+00 -1.65403374e-02
-3.11595917e-01 -5.12676954e-01 1.06319368e+00 1.45111829e-01
1.23807490e+00 -8.16905260e-01 -1.94722429e-01 5.28711677e-01
1.29966521e+00 -2.24633545e-01 5.18651485e-01 4.83950794e-01
9.13715005e-01 8.07335973e-01 7.59572625e-01 1.95248678e-01
5.97306967e-01 6.30491138e-01 1.04544532e+00 -6.63582206e-01
-1.83799654e-01 -6.10074759e-01 3.67043227e-01 -7.16903061e-02
-2.53639698e-01 -2.17798546e-01 -8.58708680e-01 6.23594165e-01
-2.33037114e+00 -1.15116417e+00 -1.32401511e-01 1.96077013e+00
1.03203201e+00 3.83589357e-01 1.51626617e-01 2.50678360e-01
5.30192077e-01 3.79744887e-01 -5.84635735e-01 -1.82087049e-01
-3.53239119e-01 -3.14924717e-01 4.92830366e-01 2.87695467e-01
-1.36518252e+00 9.92847443e-01 6.33275414e+00 7.43885577e-01
-1.02963090e+00 7.73018003e-02 9.64580655e-01 -1.25853404e-01
-7.08039343e-01 4.25095409e-02 -1.46005809e-01 4.22851741e-01
3.98330957e-01 -8.38868544e-02 2.00469688e-01 1.17116642e+00
1.12555429e-01 9.61065292e-02 -1.52413964e+00 1.00056994e+00
-2.19907045e-01 -1.75887609e+00 3.89515460e-01 -7.08319023e-02
8.08027983e-01 -1.61304146e-01 2.89634377e-01 1.26479954e-01
6.01139367e-01 -1.23706269e+00 1.57692337e+00 1.74977392e-01
9.79902446e-01 -3.37104857e-01 7.03093767e-01 3.59914660e-01
-9.76935744e-01 -1.01895556e-01 -1.52263165e-01 -3.21321130e-01
-5.49923107e-02 7.14943707e-01 -6.66116774e-01 6.68598890e-01
6.20264471e-01 5.22195339e-01 -5.99132240e-01 5.96799910e-01
-6.91799760e-01 5.20589411e-01 -1.66597843e-01 1.01054590e-02
4.04173553e-01 1.51087761e-01 2.72229433e-01 1.34492850e+00
-2.35725604e-02 -1.12821274e-02 2.31092662e-01 1.15533686e+00
-3.46005350e-01 -6.13012910e-02 -8.41739416e-01 6.31418005e-02
2.82985836e-01 1.04905081e+00 -6.73243344e-01 -6.03968859e-01
-4.43158865e-01 8.28759909e-01 3.48692119e-01 5.70734024e-01
-8.96867752e-01 -1.28368974e-01 5.59489727e-01 1.37842476e-01
-8.21316838e-02 8.53299052e-02 -7.26238012e-01 -1.42001510e+00
3.28297973e-01 -1.02074909e+00 6.55292749e-01 -9.15446043e-01
-1.40417719e+00 4.95426923e-01 -1.63921099e-02 -1.31660032e+00
-1.34409204e-01 -8.94145846e-01 -3.29695046e-01 5.17453074e-01
-1.75207078e+00 -1.68464255e+00 -4.40979362e-01 5.24562478e-01
3.64960492e-01 1.29958913e-01 6.50527120e-01 5.35291433e-02
-4.12865132e-01 5.31176507e-01 -4.58958417e-01 1.27839088e-01
6.37312293e-01 -1.53175712e+00 3.46912205e-01 1.03256345e+00
5.04132032e-01 7.09046781e-01 7.90511012e-01 -4.80117142e-01
-1.16885936e+00 -1.28169632e+00 7.17229009e-01 -6.19461238e-01
7.24955797e-01 -4.20597792e-01 -6.58669829e-01 7.20252454e-01
8.03840011e-02 3.10709476e-01 6.29571259e-01 3.21257025e-01
-9.31392431e-01 -2.30317503e-01 -1.16245747e+00 1.03705883e+00
1.33206058e+00 -6.86978579e-01 -7.44585812e-01 2.67913193e-01
8.28695297e-01 -4.23168510e-01 -4.02778238e-01 4.36001539e-01
6.12676144e-01 -1.18749046e+00 1.00916004e+00 -6.92164600e-01
7.83117831e-01 -7.47481048e-01 -1.25265256e-01 -1.12071455e+00
-3.53720933e-01 -5.57452679e-01 -1.73413754e-01 1.11001337e+00
6.35938287e-01 -5.16845107e-01 7.28140354e-01 7.21812665e-01
1.56775583e-03 -8.22195411e-01 -8.03110063e-01 -8.81443024e-01
-3.51249158e-01 -7.96583414e-01 8.73943448e-01 1.17824864e+00
1.75024569e-01 4.43397075e-01 -4.27891254e-01 2.46088520e-01
8.38752151e-01 5.18185019e-01 7.05183387e-01 -1.04195309e+00
-2.59368181e-01 -5.29286146e-01 -3.75799388e-01 -7.64645875e-01
1.35699794e-01 -8.04355025e-01 -5.22570573e-02 -1.60788751e+00
2.52286047e-01 -3.62995982e-01 -3.52773398e-01 9.64269876e-01
-3.80415708e-01 6.11863554e-01 3.23781073e-01 2.66329050e-01
-7.49424458e-01 5.04746735e-01 1.13881803e+00 -6.04244649e-01
1.22425064e-01 -3.37312222e-01 -9.68511343e-01 1.06771159e+00
3.86781424e-01 -3.22087646e-01 -6.31969810e-01 -5.86072803e-01
4.53759313e-01 -3.58567595e-01 9.64676976e-01 -5.99604607e-01
-1.39794312e-03 -6.40191913e-01 2.53313035e-01 -1.77172914e-01
-7.72086233e-02 -8.02814841e-01 1.30079180e-01 4.16264236e-01
-4.20343101e-01 -2.57225007e-01 8.83374363e-02 7.32675374e-01
-2.06610233e-01 2.00794369e-01 6.68540478e-01 -1.63949654e-01
-8.74198079e-01 3.46784234e-01 1.50643826e-01 3.03219073e-03
1.02496052e+00 -2.12889925e-01 -5.70017934e-01 -2.67719686e-01
-2.93150008e-01 2.66332686e-01 6.95725322e-01 3.62525225e-01
5.18856466e-01 -1.46326363e+00 -7.27347732e-01 -2.05476865e-01
6.93580627e-01 4.58604880e-02 -2.02884525e-02 6.99561000e-01
-4.27607358e-01 1.72995836e-01 -1.62521347e-01 -5.89190960e-01
-1.03044963e+00 8.46998930e-01 2.24529982e-01 -3.01536202e-01
-5.48653126e-01 7.05416322e-01 7.84684837e-01 -1.00864455e-01
1.80855870e-01 -5.98775327e-01 7.32205808e-02 -9.71942674e-03
4.04032767e-01 1.31513387e-01 -1.45180643e-01 -3.38308275e-01
-4.21024799e-01 3.18899035e-01 1.81742907e-01 -1.53517410e-01
1.11065304e+00 -4.52914685e-02 7.17712343e-02 3.87936413e-01
7.13960886e-01 -2.02135101e-01 -1.52063692e+00 -3.16518217e-01
1.57784209e-01 -6.14231110e-01 -4.08634096e-02 -1.04192138e+00
-9.26202297e-01 8.42657208e-01 3.13371181e-01 3.23508561e-01
1.07641435e+00 1.53997004e-01 3.59412968e-01 4.07750428e-01
3.31866980e-01 -1.10636568e+00 1.98045582e-01 -6.37913868e-02
1.10400677e+00 -1.56070220e+00 2.14041859e-01 -5.64573765e-01
-8.48540306e-01 7.88513541e-01 5.61390102e-01 2.27331132e-01
1.04703037e-02 1.68711334e-01 2.33231694e-01 -2.19891548e-01
-7.51853347e-01 -3.59143764e-01 6.19411111e-01 6.98989451e-01
3.66556704e-01 1.77043483e-01 -7.29783177e-02 5.30673444e-01
-2.24072635e-01 3.12143005e-02 7.12001175e-02 4.28984880e-01
-1.71623722e-01 -5.81074297e-01 -3.16620529e-01 1.57776952e-01
-3.50707591e-01 -1.40511408e-01 -7.12197185e-01 9.86979067e-01
3.29052806e-01 9.37768877e-01 -3.62827897e-01 -2.80744463e-01
4.50322449e-01 -1.98952541e-01 3.02068025e-01 -5.68395793e-01
-3.53231043e-01 -2.43514538e-01 3.54196131e-01 -7.15190709e-01
-5.69965839e-01 -2.86218882e-01 -1.25481534e+00 -1.97369158e-01
-2.50024229e-01 -1.91806912e-01 4.11559284e-01 1.11343312e+00
2.75529087e-01 4.59372342e-01 3.18208039e-01 -5.33185124e-01
-7.66727567e-01 -6.64322019e-01 -2.99461573e-01 9.32374001e-01
4.12423939e-01 -7.66985059e-01 -1.70979559e-01 4.21457767e-01] | [10.355557441711426, 1.7494498491287231] |
2b820469-52cb-4cc2-9214-daf9a4be7bec | ecological-semantics-programming-environments | 2003.04567 | null | https://arxiv.org/abs/2003.04567v2 | https://arxiv.org/pdf/2003.04567v2.pdf | Ecological Semantics: Programming Environments for Situated Language Understanding | Large-scale natural language understanding (NLU) systems have made impressive progress: they can be applied flexibly across a variety of tasks, and employ minimal structural assumptions. However, extensive empirical research has shown this to be a double-edged sword, coming at the cost of shallow understanding: inferior generalization, grounding and explainability. Grounded language learning approaches offer the promise of deeper understanding by situating learning in richer, more structured training environments, but are limited in scale to relatively narrow, predefined domains. How might we enjoy the best of both worlds: grounded, general NLU? Following extensive contemporary cognitive science, we propose treating environments as "first-class citizens" in semantic representations, worthy of research and development in their own right. Importantly, models should also be partners in the creation and configuration of environments, rather than just actors within them, as in existing approaches. To do so, we argue that models must begin to understand and program in the language of affordances (which define possible actions in a given situation) both for online, situated discourse comprehension, as well as large-scale, offline common-sense knowledge mining. To this end we propose an environment-oriented ecological semantics, outlining theoretical and practical approaches towards implementation. We further provide actual demonstrations building upon interactive fiction programming languages. | ['Gabriel Stanovsky', 'Reut Tsarfaty', 'Ronen Tamari', 'Dafna Shahaf'] | 2020-03-10 | null | null | null | null | ['grounded-language-learning'] | ['natural-language-processing'] | [ 3.94204915e-01 4.91650909e-01 -2.06232950e-01 -3.58391613e-01
-2.69446254e-01 -8.41851652e-01 8.00461590e-01 4.63448286e-01
-2.01788962e-01 4.28773940e-01 9.33074236e-01 -6.50513470e-01
-5.97086251e-01 -9.97836351e-01 -8.28802347e-01 -2.24721462e-01
-1.97442189e-01 6.27491772e-01 -1.56573094e-02 -7.16600597e-01
4.09389764e-01 1.93343624e-01 -1.93076408e+00 4.02536094e-01
1.01111460e+00 3.18962544e-01 6.42484307e-01 3.46378893e-01
-1.75153106e-01 1.17935574e+00 -1.09703735e-01 -8.30025002e-02
-3.36464606e-02 -4.33743745e-01 -1.20971453e+00 -1.44417897e-01
2.03615770e-01 -4.20564055e-01 -1.20285757e-01 6.79199636e-01
2.18266487e-01 4.47341889e-01 2.58600503e-01 -9.33155358e-01
-1.13167870e+00 9.89892721e-01 2.21719384e-01 5.30632287e-02
8.32915545e-01 3.95182878e-01 1.20484948e+00 -4.23433989e-01
8.34488630e-01 1.44726741e+00 6.00193441e-01 5.51089823e-01
-1.34616363e+00 -2.34806120e-01 6.05774403e-01 1.83770061e-01
-9.78614867e-01 -5.18495858e-01 6.11676872e-01 -6.17260575e-01
1.38270855e+00 3.06529850e-01 1.08012545e+00 1.19262350e+00
-2.58601308e-01 7.82487988e-01 1.23755062e+00 -6.56787872e-01
2.58830011e-01 9.22285840e-02 1.53478041e-01 6.45300150e-01
3.41963619e-01 1.93918705e-01 -8.66199911e-01 3.36102694e-02
8.49109411e-01 1.13408998e-01 -2.50219822e-01 -5.62711000e-01
-1.47165251e+00 7.49571681e-01 6.71970963e-01 5.31891048e-01
-3.91368359e-01 1.72520056e-01 1.85411990e-01 1.98142231e-01
3.13931942e-01 9.05434608e-01 -3.43205422e-01 -3.13641638e-01
-4.54055667e-01 6.24633729e-01 9.03945327e-01 7.51937091e-01
8.40386868e-01 -3.38547140e-01 2.13818312e-01 5.75682759e-01
3.22621346e-01 3.02139819e-01 3.64663243e-01 -1.22845578e+00
3.15705419e-01 7.83167481e-01 3.12644094e-01 -1.12589014e+00
-5.78702450e-01 -1.18980654e-01 -8.64430070e-02 -4.80245873e-02
3.46932441e-01 -7.39107803e-02 -6.45558953e-01 2.05723810e+00
3.69295537e-01 1.59839928e-01 7.42482468e-02 1.00069737e+00
4.44750339e-01 5.82432210e-01 5.73823273e-01 2.45694816e-01
1.32739735e+00 -6.33240461e-01 -3.40503216e-01 -7.39412785e-01
8.52978706e-01 -6.41098991e-02 1.65237248e+00 1.44753188e-01
-9.42679405e-01 -2.47630984e-01 -9.57157552e-01 -4.68092114e-01
-6.51333928e-01 -4.94597018e-01 1.32915115e+00 7.09720910e-01
-9.79152620e-01 4.92068917e-01 -8.30211997e-01 -9.10137177e-01
4.66686636e-01 -1.02468863e-01 -1.77064136e-01 -4.40503024e-02
-1.26942456e+00 1.17342114e+00 5.84004343e-01 6.51847571e-02
-9.51465666e-01 -6.48436725e-01 -1.03945827e+00 1.40985042e-01
6.52996063e-01 -9.40638244e-01 1.12175965e+00 -1.21250284e+00
-1.05189991e+00 1.05338490e+00 1.29914172e-02 -3.71635735e-01
1.35798171e-01 -4.97421950e-01 -1.01303719e-01 -1.01855844e-01
3.27098459e-01 5.97177386e-01 -1.24819085e-01 -1.47218084e+00
-5.00874579e-01 -5.60781598e-01 8.64501715e-01 6.93496644e-01
-3.07556897e-01 3.12468093e-02 2.63401926e-01 -2.45821849e-01
3.59103888e-01 -6.36143506e-01 -3.01416874e-01 1.98218152e-02
6.00909665e-02 -2.12234497e-01 2.18975097e-01 -5.43393016e-01
1.13490283e+00 -2.06785846e+00 1.33001089e-01 8.80961493e-03
1.85415626e-01 -3.73732559e-02 -7.30095059e-02 8.67458701e-01
2.37654358e-01 3.84843439e-01 4.66199964e-02 6.09476417e-02
4.21526164e-01 2.99909890e-01 -4.62614119e-01 1.28239900e-01
-1.16136715e-01 9.23248768e-01 -1.42775905e+00 -2.56066859e-01
3.98954600e-01 7.30437785e-02 -8.77225459e-01 1.76836234e-02
-5.39622247e-01 2.62499005e-01 -6.95140243e-01 2.74235427e-01
-4.43237126e-02 -1.52773231e-01 6.38152182e-01 2.84296334e-01
-3.99312258e-01 6.51098788e-01 -1.19599950e+00 2.09321666e+00
-6.88263357e-01 7.25474477e-01 -1.59058347e-02 -1.03741300e+00
5.99081397e-01 1.68687806e-01 1.47327498e-01 -7.21476138e-01
-1.12330116e-01 2.13627532e-01 2.38434836e-01 -1.05085790e+00
4.91772979e-01 -4.54149187e-01 1.86996907e-02 6.82297230e-01
-1.76574618e-01 -1.42970428e-01 -4.73586693e-02 2.34624788e-01
8.74044061e-01 7.16620982e-01 4.59768295e-01 -4.37316775e-01
-1.10997930e-01 4.01914150e-01 2.70918399e-01 8.79550934e-01
-1.99879825e-01 1.27909347e-01 3.11219692e-01 -5.70934951e-01
-9.79777157e-01 -9.78702843e-01 -9.40729901e-02 1.59585774e+00
4.64580864e-01 -4.72877920e-01 -7.98233211e-01 -3.77930142e-02
-1.89407229e-01 1.16386235e+00 -5.96261322e-01 -9.23708975e-02
-5.92448235e-01 -3.39006782e-01 6.43361986e-01 6.29335225e-01
3.51481676e-01 -1.35037661e+00 -1.45843554e+00 3.45248848e-01
-3.51614594e-01 -7.44051218e-01 3.13881278e-01 3.16814035e-01
-7.78440237e-01 -1.06105804e+00 7.95396268e-02 -8.74238491e-01
3.90584439e-01 5.53397000e-01 1.24273980e+00 3.89632523e-01
-9.55549628e-02 6.93757534e-01 -4.83280271e-01 -6.97762609e-01
-2.79960334e-01 -1.30647078e-01 -1.21708162e-01 -5.56975365e-01
6.18029237e-01 -8.72682452e-01 -5.51179528e-01 1.35773093e-01
-9.71054852e-01 3.98077190e-01 2.95167238e-01 7.00711250e-01
2.64540970e-01 -2.05833986e-01 7.01030016e-01 -9.00948584e-01
8.32098126e-01 -9.33944046e-01 -6.60885721e-02 4.55952585e-01
-5.57123542e-01 -8.44193250e-02 3.93874973e-01 -1.72501162e-01
-1.46636248e+00 -3.38571995e-01 -5.43031190e-03 1.34619713e-01
-6.15456402e-01 9.14369524e-01 -3.09627235e-01 3.75104368e-01
1.22816825e+00 6.62275776e-02 -1.57053526e-02 -3.41728032e-01
9.52868879e-01 5.38297296e-01 3.25195432e-01 -1.26961434e+00
5.59569061e-01 5.60666323e-01 -6.13364637e-01 -9.54080999e-01
-9.61629748e-01 -2.84680665e-01 -7.06177056e-01 -2.00945437e-01
8.87789130e-01 -9.70795155e-01 -6.51661217e-01 -5.01254983e-02
-8.07828546e-01 -8.24090958e-01 -4.37859178e-01 1.98605344e-01
-7.80626476e-01 4.96480279e-02 -1.28386185e-01 -7.85672307e-01
2.54033208e-01 -7.96873093e-01 9.29483533e-01 2.37780243e-01
-7.14969635e-01 -1.34271717e+00 1.14999786e-01 5.48109710e-01
4.82768118e-01 3.64738703e-01 1.19624615e+00 -9.23376203e-01
-6.19092226e-01 1.06144026e-01 1.14895487e-02 -6.01101667e-02
1.51898608e-01 -3.10267061e-01 -9.43953276e-01 1.51576623e-01
9.43076760e-02 -6.30742371e-01 3.86373550e-01 9.26386118e-02
1.02203310e+00 -4.00153011e-01 -4.55662996e-01 3.90641540e-01
1.42870462e+00 2.75507987e-01 5.93595207e-01 7.40507305e-01
4.92715240e-01 1.03748226e+00 6.22403383e-01 2.65951812e-01
9.78387117e-01 3.19049686e-01 2.26370454e-01 3.55195850e-02
1.15085080e-01 -8.09452534e-01 2.11482421e-01 2.26260424e-01
-2.41868526e-01 -6.92307353e-02 -1.34304094e+00 7.75101602e-01
-2.14561391e+00 -1.56745946e+00 1.29224673e-01 1.92249978e+00
7.01252520e-01 -3.30793038e-02 -3.08347642e-02 -1.71599090e-01
2.24490747e-01 4.64998335e-01 -5.36791742e-01 -3.97543937e-01
-1.57249480e-01 3.23192813e-02 -2.00727165e-01 6.44816279e-01
-7.46380568e-01 1.27766252e+00 6.47931862e+00 2.33509436e-01
-8.91956329e-01 2.07102031e-01 5.06770074e-01 -5.67360735e-03
-9.12932575e-01 4.10402924e-01 -7.18487725e-02 -5.39866351e-02
8.32285523e-01 -1.69020165e-02 9.00119424e-01 8.11721981e-01
4.88767296e-01 -3.38647723e-01 -1.40445912e+00 5.54698527e-01
4.94519882e-02 -1.53531754e+00 -1.41186835e-02 -8.38578194e-02
5.19664645e-01 3.83332148e-02 -2.90544659e-01 2.54121870e-01
5.87691188e-01 -1.28818774e+00 1.18146312e+00 3.14152241e-01
3.32618833e-01 -3.14218551e-01 1.77087426e-01 7.52835631e-01
-9.24989581e-01 -3.84568393e-01 -2.13720411e-01 -8.93491924e-01
1.10630035e-01 -2.02055335e-01 -3.98940116e-01 3.35030109e-01
6.07859552e-01 6.02814972e-01 -1.75723374e-01 5.65597296e-01
-2.73807883e-01 4.28553939e-01 -3.50453496e-01 -3.49963605e-01
2.68529564e-01 -1.93777621e-01 4.02817696e-01 9.74777102e-01
4.14578989e-02 5.88654459e-01 3.90925050e-01 1.17886877e+00
3.69384438e-01 4.79217321e-02 -1.00737977e+00 -2.11209103e-01
9.36380386e-01 8.11831951e-01 -7.31876254e-01 -2.75282621e-01
-4.25398320e-01 5.07236481e-01 4.32148755e-01 4.98915583e-01
-5.54806471e-01 2.95032468e-03 8.04748833e-01 3.65697086e-01
-3.23266268e-01 -2.99563229e-01 -5.83238661e-01 -1.17685497e+00
-6.61780536e-02 -1.06256974e+00 1.72941133e-01 -1.00971532e+00
-1.00685060e+00 1.04751429e-02 3.58721167e-01 -5.34575522e-01
-2.18787476e-01 -7.04194367e-01 -5.50736070e-01 6.17601693e-01
-1.18589890e+00 -1.55757058e+00 -4.24191803e-01 4.68073666e-01
6.01345420e-01 3.08053881e-01 9.75738525e-01 -1.90852627e-01
-1.99701577e-01 -1.99353039e-01 -1.11182019e-01 -1.04664125e-01
2.18313783e-01 -1.10297418e+00 4.19013232e-01 6.67086422e-01
3.18438649e-01 1.23546648e+00 8.20875823e-01 -6.15397811e-01
-1.54704070e+00 -4.88244414e-01 8.51790130e-01 -9.51013207e-01
8.37040424e-01 -5.64810038e-01 -9.12075460e-01 1.20822608e+00
2.32276857e-01 -6.15710497e-01 9.66904700e-01 7.28352487e-01
-3.73351932e-01 2.95687169e-01 -9.98475254e-01 9.35816526e-01
1.66761327e+00 -9.86309171e-01 -1.11139357e+00 4.82836336e-01
8.25306714e-01 -3.66508126e-01 -7.22570896e-01 -1.10186227e-01
7.62129366e-01 -1.14294362e+00 1.11702824e+00 -9.37364876e-01
5.67924440e-01 -3.17751288e-01 -4.82349932e-01 -1.13628566e+00
-4.17382956e-01 -5.67622960e-01 3.59986097e-01 1.13083577e+00
3.63400966e-01 -6.56517267e-01 4.46714550e-01 9.80615377e-01
-3.57638419e-01 -6.08170092e-01 -4.86536235e-01 -4.08109635e-01
2.99175888e-01 -8.86930466e-01 9.21449602e-01 1.03553939e+00
6.82154834e-01 3.99335980e-01 1.04013443e-01 1.65245265e-01
3.40170085e-01 9.19297412e-02 7.26363540e-01 -1.37210023e+00
-3.47453922e-01 -4.27762181e-01 -4.48363610e-02 -9.92592037e-01
2.30449617e-01 -8.48295152e-01 1.52928784e-01 -1.97491086e+00
3.44800800e-01 -7.42921650e-01 -6.77486509e-02 6.01448834e-01
-1.69823572e-01 -4.10519898e-01 2.18747243e-01 2.85282880e-01
-5.12781084e-01 2.68854022e-01 1.02428830e+00 9.93089899e-02
-3.45696837e-01 -5.82511723e-01 -1.30740297e+00 9.62621152e-01
6.94234967e-01 7.68051133e-04 -9.25246418e-01 -8.20399284e-01
6.05560660e-01 -2.73614854e-01 7.49189317e-01 -8.06236982e-01
1.53561443e-01 -8.13138604e-01 -1.85359828e-03 1.83665678e-01
9.02353004e-02 -7.79915631e-01 2.64801413e-01 2.33237162e-01
-6.37020588e-01 -1.28964931e-01 1.67343333e-01 5.04185736e-01
9.36567336e-02 -1.75878718e-01 2.36646712e-01 -5.71091413e-01
-1.28059673e+00 -2.82101899e-01 -2.96921074e-01 1.64473459e-01
9.75180924e-01 -6.56080723e-01 -4.65231806e-01 -3.61471653e-01
-7.34748423e-01 2.97699839e-01 9.29253280e-01 5.41607440e-01
3.49546105e-01 -8.35245371e-01 -3.05935651e-01 -8.57198760e-02
3.26639384e-01 2.30194286e-01 3.53835970e-01 3.65765333e-01
-5.55791438e-01 2.60893822e-01 -1.84507415e-01 -3.13732296e-01
-4.95493025e-01 5.47456443e-01 5.59876144e-01 2.49762118e-01
-6.63479328e-01 7.08500326e-01 4.63802934e-01 -7.01153994e-01
3.59510407e-02 -3.50979805e-01 -2.70435750e-01 1.36808977e-01
3.41545582e-01 2.01182112e-01 -3.62913489e-01 -4.20457155e-01
-2.72316039e-01 3.03607047e-01 5.97854614e-01 -1.41140655e-01
1.54091263e+00 -4.03244823e-01 -1.66279852e-01 6.73234642e-01
4.57905471e-01 -1.21400170e-01 -1.23921716e+00 -6.89802542e-02
2.43057251e-01 -3.53171378e-01 -1.95900336e-01 -1.05397248e+00
-1.55049101e-01 8.57177556e-01 3.05904239e-01 3.99779171e-01
7.85005629e-01 4.25049841e-01 3.37790549e-01 6.23839557e-01
6.48098946e-01 -1.14099729e+00 1.52662843e-02 2.00052783e-01
8.14689577e-01 -1.06572258e+00 -5.59815988e-02 -1.42890913e-02
-7.35366344e-01 8.75231504e-01 6.72700346e-01 -8.20926279e-02
3.47259074e-01 -6.92835823e-02 2.06100941e-03 -5.02392292e-01
-8.83491814e-01 -3.65875483e-01 -1.48539484e-01 1.08230317e+00
7.00062633e-01 3.32084954e-01 -5.43582998e-02 4.89934146e-01
-4.69870508e-01 -1.03653111e-01 2.98781157e-01 9.55257416e-01
-8.35355043e-01 -7.34397948e-01 -3.08235943e-01 2.14388311e-01
1.05837032e-01 -1.10264182e-01 -4.42800492e-01 1.09011936e+00
4.01485264e-01 1.05278909e+00 1.02706067e-01 -4.98574413e-02
4.39013660e-01 1.51386142e-01 5.32291353e-01 -8.53593647e-01
-6.70165777e-01 -3.58953834e-01 4.38759714e-01 -8.41050327e-01
-6.59253776e-01 -8.70112836e-01 -1.49313033e+00 -3.44405085e-01
-4.46645059e-02 -9.78766754e-02 4.40968692e-01 1.19996929e+00
3.93448800e-01 3.09327900e-01 9.25883353e-02 -8.96141708e-01
-4.14685518e-01 -6.76570415e-01 -1.47334620e-01 5.71722150e-01
1.95360050e-01 -6.21473134e-01 -1.47355407e-01 4.69805412e-02] | [9.31600284576416, 6.813937187194824] |
4631c50f-cb27-4b83-9b58-940c2cdc6b60 | language-models-as-knowledge-embeddings | 2206.12617 | null | https://arxiv.org/abs/2206.12617v3 | https://arxiv.org/pdf/2206.12617v3.pdf | Language Models as Knowledge Embeddings | Knowledge embeddings (KE) represent a knowledge graph (KG) by embedding entities and relations into continuous vector spaces. Existing methods are mainly structure-based or description-based. Structure-based methods learn representations that preserve the inherent structure of KGs. They cannot well represent abundant long-tail entities in real-world KGs with limited structural information. Description-based methods leverage textual information and language models. Prior approaches in this direction barely outperform structure-based ones, and suffer from problems like expensive negative sampling and restrictive description demand. In this paper, we propose LMKE, which adopts Language Models to derive Knowledge Embeddings, aiming at both enriching representations of long-tail entities and solving problems of prior description-based methods. We formulate description-based KE learning with a contrastive learning framework to improve efficiency in training and evaluation. Experimental results show that LMKE achieves state-of-the-art performance on KE benchmarks of link prediction and triple classification, especially for long-tail entities. | ['Yanghua Xiao', 'Jiaqing Liang', 'Qianyu He', 'Xintao Wang'] | 2022-06-25 | null | null | null | null | ['triple-classification'] | ['graphs'] | [-5.49167395e-01 2.87793130e-01 -9.93561268e-01 -1.47397012e-01
-3.48492026e-01 -4.64612097e-01 6.20951653e-01 5.79985023e-01
-3.17427546e-01 8.31113875e-01 4.65217710e-01 -2.22640947e-01
-3.84129047e-01 -1.25540173e+00 -6.52011573e-01 -2.92407811e-01
-4.21965361e-01 7.39410639e-01 3.26150715e-01 -3.93590927e-01
-3.14414173e-01 2.24380106e-01 -1.17991686e+00 1.25772342e-01
9.59456623e-01 9.34511602e-01 -1.59720868e-01 2.60098189e-01
-6.11761451e-01 1.09394634e+00 -1.14095598e-01 -1.02350545e+00
-2.65621036e-01 2.20625490e-01 -1.00096798e+00 -4.24870253e-01
2.61599660e-01 -4.22522286e-03 -9.74566638e-01 8.29256594e-01
4.50187951e-01 1.92393195e-02 9.48196471e-01 -1.50897765e+00
-1.53307855e+00 1.04283619e+00 -3.63952816e-01 4.86783162e-02
3.24109942e-01 -5.78399062e-01 1.70015037e+00 -1.33196473e+00
8.45304430e-01 1.07972538e+00 9.84434783e-01 3.55318040e-01
-1.26858568e+00 -5.37497580e-01 2.29515389e-01 7.69317091e-01
-1.79030192e+00 -1.40362233e-01 7.76522219e-01 -4.77131218e-01
1.25065470e+00 -1.16049945e-02 8.28862488e-01 1.03489935e+00
-3.66163313e-01 1.14331496e+00 3.82727087e-01 -2.77716219e-01
-1.52783439e-01 4.33545977e-01 4.28784966e-01 9.52403843e-01
9.70104218e-01 -2.53834754e-01 -5.25593996e-01 -3.43154997e-01
5.15550852e-01 -9.44802687e-02 -4.88816768e-01 -1.05507958e+00
-1.28614986e+00 1.05002737e+00 6.63894594e-01 2.98395962e-01
-3.61391544e-01 2.15587825e-01 6.96327567e-01 2.80201077e-01
5.11920154e-01 4.38770443e-01 -7.72788644e-01 3.24001759e-01
-5.74412227e-01 2.44928718e-01 9.11869049e-01 1.42304754e+00
1.00081265e+00 2.26709396e-02 -2.12676287e-01 7.79765666e-01
4.06427056e-01 2.82433987e-01 5.36278605e-01 -1.00823715e-01
6.03261769e-01 9.58723545e-01 2.64074858e-02 -1.20552230e+00
-4.67586964e-01 -6.91465616e-01 -8.78754914e-01 -7.40092754e-01
-8.68671909e-02 9.07498449e-02 -6.95166528e-01 1.51531076e+00
3.55146289e-01 3.83240014e-01 4.28677827e-01 5.85493445e-01
1.40242684e+00 5.96756220e-01 2.75906831e-01 -1.70046017e-02
1.14569533e+00 -1.25823891e+00 -8.99605393e-01 7.52027109e-02
1.31486714e+00 -1.87138081e-01 7.67793119e-01 -2.33940661e-01
-7.79933155e-01 -2.49264896e-01 -1.06036317e+00 -1.48307666e-01
-9.98476982e-01 3.99157345e-01 1.14194787e+00 5.12225449e-01
-8.96573603e-01 4.83143866e-01 -5.64479053e-01 -2.52530187e-01
4.22632486e-01 2.53417462e-01 -8.11641216e-01 -2.16942027e-01
-1.83720362e+00 9.19884562e-01 1.04511797e+00 -1.10567465e-01
-4.88616258e-01 -1.08289933e+00 -1.45475948e+00 4.42241877e-01
6.02827609e-01 -9.78217125e-01 6.79194450e-01 -1.16450906e-01
-9.17062759e-01 6.99272275e-01 5.83584383e-02 -3.77967477e-01
-4.40959446e-02 -2.14555442e-01 -9.18346286e-01 -1.90239936e-01
-4.68076300e-03 1.13482766e-01 5.43855548e-01 -1.33343685e+00
-4.01948929e-01 -1.21878177e-01 1.57096952e-01 -1.27100116e-02
-1.01909804e+00 -4.64358777e-01 -6.01837635e-01 -5.85465431e-01
-3.46523561e-02 -5.30923128e-01 1.17160290e-01 -2.38559783e-01
-7.15108693e-01 -7.12949991e-01 7.27910936e-01 -5.35070002e-01
1.88961792e+00 -1.69504070e+00 2.48750612e-01 2.22857684e-01
5.43304503e-01 7.47127354e-01 -1.54465482e-01 9.58633423e-01
-1.89932212e-01 3.68574560e-01 3.31670493e-01 -3.38186204e-01
4.28402573e-01 3.82630825e-01 -3.26214790e-01 2.37754405e-01
2.47236595e-01 1.51678157e+00 -1.24598193e+00 -7.00028777e-01
3.86733515e-03 5.70289612e-01 -5.28668046e-01 2.94504985e-02
-1.86924726e-01 -4.57342476e-01 -6.32877469e-01 8.52980554e-01
5.88855982e-01 -6.83350861e-01 6.86785519e-01 -7.54899859e-01
3.27458829e-01 3.77926379e-01 -1.09079003e+00 1.57965839e+00
-4.88028765e-01 2.03903779e-01 -4.66451257e-01 -1.27608359e+00
8.03948641e-01 1.91774189e-01 3.05172265e-01 -2.91731983e-01
-2.18524933e-01 3.15871835e-01 -4.21046972e-01 -4.34633344e-01
7.98453808e-01 6.89785331e-02 -5.11843860e-02 2.02722233e-02
5.52190244e-01 3.56419921e-01 2.67858773e-01 7.20555007e-01
1.03627753e+00 -7.80806839e-02 5.80851018e-01 -6.32018149e-02
4.61170405e-01 -1.53794408e-01 6.29546165e-01 4.25683051e-01
1.51772603e-01 -9.77737010e-02 4.91007388e-01 -3.74586850e-01
-8.20267558e-01 -1.19974399e+00 -2.01170463e-02 1.04931486e+00
2.07571253e-01 -1.17848575e+00 8.02935585e-02 -1.15334427e+00
6.13961637e-01 6.38390183e-01 -7.16144025e-01 -4.87969637e-01
-1.98708996e-01 -6.59581959e-01 6.19753182e-01 8.10998678e-01
1.12502515e-01 -5.30884624e-01 4.80108202e-01 3.95519733e-01
1.42082376e-02 -9.89682376e-01 -2.45735705e-01 2.09896564e-01
-6.02943480e-01 -1.23460877e+00 -7.10067213e-01 -1.05438304e+00
6.99747503e-01 1.31262869e-01 1.46089208e+00 -1.34137467e-01
-1.47216529e-01 5.11878133e-01 -6.74501419e-01 -4.58208397e-02
7.19016343e-02 3.35276157e-01 2.40748599e-01 -1.74679548e-01
5.60841680e-01 -6.05913281e-01 -3.50405753e-01 5.60685433e-02
-7.39019632e-01 -2.99257874e-01 8.37510407e-01 1.27461767e+00
6.83131635e-01 1.88059822e-01 8.96327317e-01 -1.08398867e+00
6.73935115e-01 -7.92441845e-01 -2.26789922e-01 8.06747973e-01
-1.10482216e+00 3.07848901e-01 5.23993850e-01 -3.93520206e-01
-6.84350491e-01 -3.32350701e-01 1.57263830e-01 -6.19971097e-01
5.68890691e-01 1.28054392e+00 -2.40000203e-01 -1.47916943e-01
4.72066879e-01 3.32077652e-01 -5.46853364e-01 -6.17722690e-01
7.95799911e-01 4.10258144e-01 3.21125329e-01 -7.45539546e-01
1.05378520e+00 1.63133129e-01 -4.78051640e-02 -5.44356585e-01
-1.12141824e+00 -8.49788368e-01 -5.95679522e-01 2.33698130e-01
3.05327296e-01 -1.24929059e+00 -4.57444280e-01 -6.34398833e-02
-8.90883327e-01 1.82748944e-01 -4.36976910e-01 6.02750182e-01
-2.59547204e-01 4.65390891e-01 -5.45518398e-01 -4.21797276e-01
-3.99287939e-01 -2.80831814e-01 7.54088879e-01 1.86275467e-02
2.03539446e-01 -1.55373335e+00 3.52647930e-01 1.57729939e-01
2.38194406e-01 9.39935446e-02 1.22078729e+00 -9.38359559e-01
-5.14889479e-01 -4.87593293e-01 -4.81718659e-01 3.30252409e-01
1.39337957e-01 -2.52536088e-01 -6.03112817e-01 -2.79270738e-01
-1.17099011e+00 -5.56149781e-01 1.23282683e+00 -1.10241935e-01
1.01251769e+00 -5.15860379e-01 -7.74287343e-01 6.76654279e-01
1.74219859e+00 -5.25194228e-01 5.16342819e-01 3.67807925e-01
1.12836909e+00 3.60103965e-01 6.26091659e-01 6.00061834e-01
1.03304100e+00 5.85401893e-01 3.21389318e-01 1.14608988e-01
-1.67943299e-01 -8.72080922e-01 9.67329219e-02 1.14262211e+00
-8.39826688e-02 -4.33482349e-01 -9.18638289e-01 1.19089460e+00
-2.04409862e+00 -1.02023220e+00 -2.40560517e-01 1.80925024e+00
1.08654845e+00 -9.95383710e-02 1.93275400e-02 9.52131748e-02
5.56368947e-01 2.43424386e-01 -3.60474080e-01 -3.92749980e-02
-3.12279552e-01 2.26570778e-02 6.20621741e-01 1.26235768e-01
-1.26097131e+00 1.18012536e+00 5.58137417e+00 1.07988846e+00
-6.35397613e-01 1.18116774e-01 -2.46794835e-01 3.07625771e-01
-8.81776214e-01 -1.39588878e-01 -1.09351981e+00 1.20309919e-01
7.66272604e-01 -7.81870723e-01 -5.75871691e-02 1.13785017e+00
-5.73674083e-01 5.90597570e-01 -1.23615479e+00 1.17634273e+00
2.12902878e-03 -1.76650214e+00 3.88401061e-01 -6.79142475e-02
9.07553375e-01 6.00962527e-02 1.60673354e-02 1.04792643e+00
5.08008718e-01 -1.00957727e+00 3.12510312e-01 6.27259433e-01
9.07168388e-01 -7.43256986e-01 1.03423309e+00 4.24630679e-02
-1.66627085e+00 1.21063828e-01 -6.88010156e-01 2.74742484e-01
1.18310347e-01 8.18401694e-01 -8.56329679e-01 1.27660906e+00
4.32155132e-01 1.25158322e+00 -5.88263571e-01 8.23953629e-01
-4.67708170e-01 5.54485977e-01 -1.50349252e-02 -1.16166376e-01
1.93471462e-01 8.97013992e-02 2.00596675e-01 1.42853701e+00
2.83262104e-01 -2.30815977e-01 2.78997570e-01 7.23390460e-01
-6.00663662e-01 3.98374557e-01 -8.24787021e-01 -7.57277608e-01
7.34839976e-01 1.32172644e+00 -7.01054633e-02 -4.22560304e-01
-7.96788096e-01 7.88959622e-01 9.19703662e-01 4.05284643e-01
-7.10428834e-01 -5.34804046e-01 7.27376938e-01 1.56921372e-01
8.03702056e-01 -2.63496310e-01 3.54727119e-01 -1.59601164e+00
9.65015292e-02 -1.35880366e-01 7.63629377e-01 -4.11466360e-01
-1.79568040e+00 2.87821352e-01 3.82025652e-02 -1.11876512e+00
-2.43653625e-01 -7.21573651e-01 -9.28151309e-02 6.95976615e-01
-2.23192382e+00 -1.76426077e+00 5.37431724e-02 6.09067202e-01
-1.53786093e-01 -2.92960972e-01 1.08521473e+00 6.34757400e-01
-5.47335088e-01 1.00291193e+00 6.12208307e-01 4.82085556e-01
6.00415289e-01 -1.48691702e+00 1.16102390e-01 1.24720015e-01
3.80322278e-01 9.00983751e-01 1.93483397e-01 -7.00666130e-01
-1.66077518e+00 -1.34712017e+00 1.42418540e+00 -4.79364991e-01
1.24348736e+00 -1.99142233e-01 -9.65147614e-01 1.02831757e+00
-2.46148080e-01 7.44486272e-01 1.19499278e+00 9.20809925e-01
-9.97685254e-01 7.81232165e-03 -8.87502313e-01 3.69325846e-01
1.31458807e+00 -9.13626134e-01 -7.77323663e-01 3.48106742e-01
7.81991601e-01 -2.66120657e-02 -1.52933764e+00 6.52470648e-01
5.08076549e-01 -3.46679986e-01 1.33680272e+00 -1.19237626e+00
2.77638584e-01 -3.66539121e-01 -1.16285220e-01 -1.61520374e+00
-6.72614455e-01 -1.19499318e-01 -1.13693678e+00 1.43400562e+00
5.34632146e-01 -5.44825315e-01 8.77371848e-01 4.89100367e-02
-1.09524675e-01 -1.15751207e+00 -6.88905120e-01 -1.37314522e+00
1.09152645e-01 -1.15049675e-01 6.22317076e-01 1.56846607e+00
4.87371236e-01 5.67690849e-01 -3.74134958e-01 3.39499474e-01
5.50010562e-01 3.60673070e-01 5.82560182e-01 -1.44187367e+00
-9.10130888e-02 -1.28000304e-01 -1.06220353e+00 -8.98506224e-01
6.91865385e-01 -1.44714844e+00 -6.61073983e-01 -2.07939911e+00
4.02855277e-01 -5.84308803e-01 -5.54254293e-01 7.09634662e-01
-2.42434025e-01 -1.09015286e-01 -2.44995162e-01 1.13331802e-01
-1.04208624e+00 1.09979713e+00 9.61578548e-01 -5.79297185e-01
1.35620907e-01 -5.58892846e-01 -6.08260751e-01 5.91728926e-01
4.53541577e-01 -2.34482169e-01 -7.99712241e-01 -3.97768497e-01
7.07087398e-01 -3.27134401e-01 2.91710317e-01 -6.10482693e-01
3.12551409e-01 7.79534690e-03 2.00547218e-01 -5.71751773e-01
2.34353364e-01 -8.08643639e-01 9.93192643e-02 4.93198857e-02
-2.64488220e-01 -3.89610738e-01 -7.26858974e-02 1.13856685e+00
-5.65369785e-01 -1.47731096e-01 2.19701126e-01 1.53500006e-01
-1.34817553e+00 8.04657161e-01 2.92652696e-01 3.88859302e-01
9.09537792e-01 1.32928699e-01 -5.35178721e-01 -2.38453582e-01
-1.01955032e+00 6.02068186e-01 1.83821410e-01 4.42940593e-01
7.53671229e-01 -1.94760704e+00 -6.47741854e-01 -1.14153489e-01
9.47618544e-01 -3.76391970e-02 2.38347188e-01 6.83559239e-01
-1.85862139e-01 6.28347754e-01 2.05344632e-01 2.15185750e-02
-1.10037887e+00 9.16543782e-01 2.64152996e-02 -8.10272157e-01
-5.78688741e-01 1.19325423e+00 6.21420480e-02 -6.29021823e-01
2.20115736e-01 -2.23741114e-01 -4.92558151e-01 4.02439445e-01
4.41360593e-01 3.10881227e-01 2.01484207e-02 -5.43392777e-01
-5.15704632e-01 3.03991884e-01 -3.41378510e-01 5.96121728e-01
1.42241180e+00 -1.34195490e-02 -4.00286689e-02 3.52230966e-01
1.37761712e+00 5.47175333e-02 -5.09383678e-01 -8.64955425e-01
3.58313948e-01 -3.61291409e-01 2.40592092e-01 -6.29071057e-01
-1.00842428e+00 6.81812286e-01 -7.47647807e-02 1.32711068e-01
6.75940275e-01 3.85746777e-01 9.03152466e-01 8.99387479e-01
6.44094050e-01 -1.10790634e+00 -5.97787164e-02 4.77278888e-01
7.22493529e-01 -1.09201872e+00 2.23454118e-01 -7.72607684e-01
-4.38002914e-01 9.87759292e-01 5.70158780e-01 1.25193223e-01
1.04421425e+00 -1.43860817e-01 -5.81106305e-01 -4.81819034e-01
-9.16190267e-01 -6.73169851e-01 6.21133506e-01 8.99305105e-01
3.49054575e-01 2.26711676e-01 -3.61972898e-01 1.13004518e+00
1.89594343e-01 -3.93987373e-02 1.83213294e-01 7.26808786e-01
-2.86700189e-01 -1.19557309e+00 3.65701348e-01 6.47176325e-01
-9.25229266e-02 -3.90699059e-01 -3.61956120e-01 1.01505578e+00
1.44894466e-01 5.42584240e-01 -2.95101464e-01 -6.96124256e-01
3.83646160e-01 3.05295169e-01 4.37026590e-01 -8.10063183e-01
2.65968405e-02 -6.32119238e-01 6.63043857e-01 -4.30917799e-01
-3.95880401e-01 -3.32351953e-01 -1.13435757e+00 -5.28486788e-01
-6.83454573e-01 4.57036763e-01 3.22504163e-01 5.77700198e-01
4.83964086e-01 5.18758297e-01 3.98957521e-01 -2.26502687e-01
-6.75253391e-01 -8.24519157e-01 -1.20357323e+00 4.11638349e-01
3.74465361e-02 -1.02033377e+00 -2.27429762e-01 -1.93386748e-01] | [8.784525871276855, 7.944107532501221] |
b4881bae-394c-4f50-a2f1-ee847609be3e | 190807836 | 1908.07836 | null | https://arxiv.org/abs/1908.07836v1 | https://arxiv.org/pdf/1908.07836v1.pdf | PubLayNet: largest dataset ever for document layout analysis | Recognizing the layout of unstructured digital documents is an important step when parsing the documents into structured machine-readable format for downstream applications. Deep neural networks that are developed for computer vision have been proven to be an effective method to analyze layout of document images. However, document layout datasets that are currently publicly available are several magnitudes smaller than established computing vision datasets. Models have to be trained by transfer learning from a base model that is pre-trained on a traditional computer vision dataset. In this paper, we develop the PubLayNet dataset for document layout analysis by automatically matching the XML representations and the content of over 1 million PDF articles that are publicly available on PubMed Central. The size of the dataset is comparable to established computer vision datasets, containing over 360 thousand document images, where typical document layout elements are annotated. The experiments demonstrate that deep neural networks trained on PubLayNet accurately recognize the layout of scientific articles. The pre-trained models are also a more effective base mode for transfer learning on a different document domain. We release the dataset (https://github.com/ibm-aur-nlp/PubLayNet) to support development and evaluation of more advanced models for document layout analysis. | ['Antonio Jimeno Yepes', 'Xu Zhong', 'Jianbin Tang'] | 2019-08-16 | null | null | null | null | ['document-layout-analysis'] | ['computer-vision'] | [ 1.97128132e-01 -1.24843322e-01 -2.23599404e-01 -3.88696790e-01
-7.86843121e-01 -1.10122335e+00 4.29195881e-01 5.52772164e-01
-4.79653180e-01 3.59068006e-01 3.14836770e-01 -6.63761318e-01
-1.87596492e-02 -5.83028436e-01 -1.06191707e+00 -4.05683428e-01
2.53140748e-01 6.94760144e-01 -1.54396966e-01 4.35872972e-01
6.80208921e-01 6.16955638e-01 -9.49363291e-01 1.00061727e+00
8.13147008e-01 8.39063346e-01 6.55109644e-01 1.11242080e+00
-6.77449524e-01 5.46305716e-01 -8.33675921e-01 -1.51737273e-01
-1.00651488e-01 -2.31346205e-01 -7.70083427e-01 -2.96905071e-01
1.00609744e+00 -4.02480930e-01 -3.36161911e-01 9.52879906e-01
5.98322630e-01 -4.08377200e-01 7.47073770e-01 -9.08281386e-01
-1.23049843e+00 6.67074084e-01 -5.27082920e-01 4.49429095e-01
1.46276027e-01 -3.42561528e-02 1.02602291e+00 -7.45242536e-01
1.03954101e+00 1.07446706e+00 7.76029408e-01 2.94621021e-01
-8.19291234e-01 -3.14580411e-01 -1.34523302e-01 3.35387319e-01
-8.40949535e-01 -2.58010983e-01 6.31249070e-01 -7.54540443e-01
1.07990241e+00 1.40627399e-01 5.25655746e-01 1.05570900e+00
6.37239814e-01 7.92795718e-01 7.48249173e-01 -6.05332851e-01
1.27205253e-01 -2.86373049e-01 4.71978903e-01 8.83945525e-01
5.38519323e-01 -6.38492048e-01 -1.55451268e-01 1.44737840e-01
6.33381665e-01 4.23309058e-02 -1.74052671e-01 -3.66842449e-01
-1.20163119e+00 7.29679048e-01 5.49255252e-01 4.99078006e-01
-1.36852227e-02 -2.16443036e-02 8.06586623e-01 2.89850291e-02
1.86699882e-01 9.01818037e-01 -3.12128514e-01 -1.67495698e-01
-1.06077099e+00 1.48733348e-01 5.99812925e-01 1.06110728e+00
1.80430442e-01 -3.67858380e-01 -2.32785225e-01 7.93352723e-01
2.38369077e-01 2.33440459e-01 4.70401257e-01 -8.08346868e-01
7.51256764e-01 7.56641269e-01 -1.87175885e-01 -1.04193664e+00
-6.66971087e-01 -1.55413136e-01 -9.09343719e-01 -1.65720358e-01
6.88635170e-01 8.60248134e-02 -1.05112433e+00 1.01802671e+00
-1.70081899e-01 -5.17855942e-01 -1.12076588e-01 4.68723506e-01
1.29992402e+00 7.71558464e-01 -7.54574463e-02 3.69684458e-01
1.50989854e+00 -1.07963288e+00 -8.54018509e-01 1.01623721e-01
8.34738910e-01 -9.85912383e-01 1.09259677e+00 7.30563343e-01
-9.81248140e-01 -6.00735784e-01 -1.32752895e+00 -8.54457915e-01
-9.39019501e-01 5.97131371e-01 6.09127343e-01 4.07214046e-01
-1.01842642e+00 5.21141589e-01 -5.85551858e-01 -5.90849578e-01
1.11986530e+00 7.11564794e-02 -3.53214383e-01 -3.31111848e-01
-5.71880817e-01 6.80714905e-01 3.70208502e-01 2.95300186e-01
-6.01881862e-01 -8.22049856e-01 -6.01358414e-01 1.71265289e-01
-7.27090389e-02 -5.86680830e-01 1.32077992e+00 -6.92569971e-01
-8.58306348e-01 1.10787928e+00 6.07963428e-02 -3.69083554e-01
4.39566463e-01 -1.47885025e-01 -2.19999149e-01 4.89236116e-01
8.62223804e-02 7.93115616e-01 3.32981020e-01 -9.40707862e-01
-3.17163587e-01 -4.94057238e-01 -2.01104850e-01 -9.52666774e-02
-5.01310408e-01 1.42921463e-01 -7.96082795e-01 -5.24821579e-01
-3.51571500e-01 -6.20442986e-01 1.11743644e-01 4.44054067e-01
-5.89277625e-01 -2.00681061e-01 9.00451779e-01 -1.06607127e+00
9.73438382e-01 -1.75577569e+00 -1.11762017e-01 6.43256530e-02
4.35351670e-01 4.87112403e-01 -5.06248176e-01 5.79083860e-01
-3.05771809e-02 3.71245831e-01 3.79394218e-02 1.69097930e-01
-1.26944229e-01 -1.92566857e-01 -2.26133585e-01 4.52720791e-01
-6.92623258e-02 1.39828718e+00 -6.82335138e-01 -3.60082358e-01
-9.61061753e-03 3.24448884e-01 -3.37084144e-01 1.77540660e-01
-5.81232369e-01 1.82721898e-01 -3.93259019e-01 5.05912960e-01
7.29667604e-01 -5.99318922e-01 5.16628683e-01 -5.30712306e-01
-1.08349413e-01 5.79887927e-02 -5.10461152e-01 2.02376103e+00
-6.90563396e-02 1.38442028e+00 -1.01030096e-01 -1.23922443e+00
8.06314111e-01 -5.81412157e-03 4.18932825e-01 -9.29099321e-01
2.97522724e-01 -1.94181219e-01 2.52442390e-01 -7.73601651e-01
3.72677416e-01 7.55860507e-01 -3.22664939e-02 5.20830691e-01
1.26676157e-01 -1.62857398e-01 6.73087358e-01 3.48112702e-01
1.20529413e+00 2.83585731e-02 -4.00501862e-02 -2.52033889e-01
1.97675422e-01 2.94954538e-01 6.70290440e-02 8.72982681e-01
5.37927635e-02 7.52345324e-01 9.90054548e-01 -6.77826762e-01
-1.69493413e+00 -8.86956692e-01 -3.19285244e-01 9.46410000e-01
-4.46219355e-01 -6.10522211e-01 -1.11455882e+00 -4.39122647e-01
4.67377715e-02 3.92165601e-01 -6.51565015e-01 2.51117975e-01
-4.61530209e-01 -5.29551923e-01 1.03465986e+00 8.38654220e-01
3.70006204e-01 -1.21598887e+00 -4.90123421e-01 -2.58100568e-03
9.34929550e-02 -1.12966192e+00 -3.67716342e-01 1.97678581e-01
-6.40434802e-01 -1.34489155e+00 -1.02694154e+00 -1.01013207e+00
1.06784463e+00 5.16404510e-02 8.53651106e-01 2.24909753e-01
-8.75727415e-01 3.86448175e-01 -9.91342068e-02 -9.65135872e-01
-4.31968212e-01 2.05111772e-01 -4.12040859e-01 -7.26939082e-01
3.83541226e-01 4.09263298e-02 -6.70790434e-01 -3.84027988e-01
-1.09852397e+00 1.87405303e-01 6.73429906e-01 7.66222239e-01
6.27627671e-01 1.36995345e-01 2.19587073e-01 -1.31682420e+00
1.02442145e+00 -2.12380230e-01 -8.02469015e-01 4.32215303e-01
-5.61609745e-01 -2.34655645e-02 8.68304193e-01 -1.11920483e-01
-9.60005105e-01 -1.21739574e-01 -2.58755773e-01 -1.18405454e-01
-3.63272041e-01 9.24148977e-01 -1.15440853e-01 2.67390132e-01
5.49933374e-01 5.06540015e-02 2.03589406e-02 -7.23141730e-01
5.65733969e-01 9.07703221e-01 8.43112111e-01 -4.90338922e-01
3.23245287e-01 2.45240360e-01 -3.39126997e-02 -9.34984326e-01
-7.95493066e-01 -4.31112409e-01 -1.04045904e+00 7.99130127e-02
1.00348508e+00 -5.20284355e-01 -6.90364122e-01 5.95931172e-01
-1.30937099e+00 -4.85787541e-01 2.99766153e-01 -1.02251982e-02
-3.45766097e-01 5.45607090e-01 -8.29750776e-01 3.02415658e-02
-7.17719913e-01 -1.12310743e+00 9.69817996e-01 2.15272918e-01
-2.86032349e-01 -1.04311752e+00 2.66437471e-01 4.30929095e-01
1.12607419e-01 2.25375578e-01 1.66815746e+00 -6.09882772e-01
-4.42241251e-01 -3.14538449e-01 -7.19275415e-01 1.54386878e-01
1.43900901e-01 7.89332747e-01 -8.21799695e-01 -1.35071173e-01
-7.82391191e-01 -2.26055890e-01 1.10257840e+00 6.77379787e-01
1.91183031e+00 -3.38916481e-01 -5.15446186e-01 6.56576455e-01
1.36466169e+00 4.90116835e-01 8.66074562e-01 7.06525266e-01
9.31093156e-01 4.98170078e-01 1.23710804e-01 1.19845577e-01
2.08912954e-01 1.74338341e-01 5.27769625e-02 -4.32876080e-01
-1.88144326e-01 -2.39188746e-01 -1.81037605e-01 6.92269981e-01
3.13835859e-01 -5.87516010e-01 -1.42586064e+00 2.61472136e-01
-1.68350863e+00 -8.67680550e-01 -1.50241911e-01 1.63359439e+00
7.23212600e-01 2.41438940e-01 -4.09233347e-02 -1.51794001e-01
5.45812666e-01 -5.68023175e-02 -5.80296516e-01 -7.77548611e-01
-1.37748331e-01 1.96000084e-01 4.04987097e-01 -1.10342633e-02
-1.27338278e+00 8.05300713e-01 6.76361036e+00 4.84184265e-01
-1.25506032e+00 -2.37739593e-01 7.47247040e-01 -4.67175618e-03
-2.64577866e-01 -3.94317687e-01 -7.44217992e-01 5.58324218e-01
1.18273199e+00 -1.78435799e-02 1.52343884e-01 8.05676103e-01
1.87317938e-01 -1.02325454e-01 -1.39181578e+00 9.47013557e-01
-1.08505562e-02 -2.27990127e+00 2.28894934e-01 1.69323578e-01
5.95430970e-01 1.28729790e-01 3.08214903e-01 -4.18772288e-02
1.65387973e-01 -1.28315234e+00 5.21102726e-01 7.63670743e-01
9.98342633e-01 -6.29492342e-01 5.86633980e-01 -4.92263474e-02
-6.52112842e-01 -7.47852307e-03 -8.05387676e-01 2.95346946e-01
-4.07795727e-01 7.22766876e-01 -1.06628954e+00 1.45543694e-01
8.95075023e-01 9.99424994e-01 -1.08418763e+00 1.24544752e+00
3.28723192e-02 2.86930472e-01 1.68517351e-01 -3.55618000e-01
3.30551237e-01 -2.61798769e-01 -3.04694381e-02 1.57522118e+00
2.51715422e-01 -4.51939464e-01 -1.59842804e-01 9.32379246e-01
-5.03851235e-01 2.18991667e-01 -8.25142086e-01 -1.02112103e+00
2.25262761e-01 1.52043509e+00 -1.07497299e+00 -3.69131237e-01
-5.43655694e-01 7.39956677e-01 5.87081015e-01 4.56437498e-01
-5.20063341e-01 -6.92953467e-01 7.73603469e-02 -6.78967461e-02
5.15159786e-01 -3.22871476e-01 -5.63914716e-01 -9.19392884e-01
3.64778899e-02 -1.11970115e+00 4.23754454e-01 -1.28782523e+00
-1.14135802e+00 2.59099096e-01 -4.27028298e-01 -9.07840848e-01
2.92553097e-01 -1.30931532e+00 -7.11704254e-01 6.41892850e-01
-1.19411850e+00 -1.20880699e+00 -2.64344484e-01 3.84385884e-02
5.06151319e-01 -3.82487923e-01 9.89572346e-01 8.78584310e-02
-9.08588171e-01 4.13758814e-01 8.61445904e-01 6.38288140e-01
7.97100782e-01 -1.48904669e+00 5.75518847e-01 4.74190474e-01
2.55391151e-01 1.08600461e+00 3.51394892e-01 -6.38487279e-01
-1.70426357e+00 -1.11817217e+00 4.98801827e-01 -6.68692291e-01
5.48896968e-01 -6.90580785e-01 -9.61325347e-01 8.30975771e-01
7.33628988e-01 -4.51013505e-01 1.07994270e+00 1.62622988e-01
-5.25376320e-01 3.42307277e-02 -7.93743253e-01 6.50539279e-01
6.54700637e-01 -4.80953634e-01 -7.00819254e-01 6.67387784e-01
3.02591115e-01 -5.80359161e-01 -9.41428244e-01 -6.12100028e-02
8.60904038e-01 -4.35661405e-01 8.35619628e-01 -8.05104733e-01
1.13740849e+00 -1.62226781e-01 3.07446629e-01 -1.02459240e+00
-4.08634216e-01 -1.00452945e-01 -4.11896892e-02 1.02072036e+00
5.01286626e-01 -2.46453688e-01 7.44468272e-01 2.80285805e-01
-3.16372156e-01 -6.80293977e-01 -4.95856076e-01 -5.25863409e-01
4.49071765e-01 -7.82951489e-02 4.01161134e-01 8.16850424e-01
1.59377456e-01 4.90632117e-01 3.57219934e-01 -2.98666179e-01
3.55811119e-01 1.70010284e-01 6.86177015e-01 -1.25886798e+00
1.08041584e-01 -8.75544786e-01 -2.02907562e-01 -7.79957056e-01
1.92471981e-01 -1.35001206e+00 -1.26946881e-01 -2.23350596e+00
5.51441610e-01 -3.42941172e-02 -3.90509427e-01 6.34694219e-01
2.07924008e-01 8.69796351e-02 -7.31221884e-02 2.28514746e-01
-7.21774399e-01 -6.47494420e-02 1.41722834e+00 -7.51862705e-01
1.21065401e-01 -5.95467269e-01 -9.72191870e-01 5.57501793e-01
8.03142905e-01 -1.29747912e-01 -2.83598661e-01 -7.14657903e-01
3.96886319e-01 -4.59665477e-01 3.29030398e-03 -7.43681908e-01
3.23810130e-01 1.52759060e-01 1.40625572e+00 -1.20763814e+00
-2.04227716e-01 -3.54167879e-01 -2.93735892e-01 3.27552557e-01
-8.40059042e-01 1.92857236e-01 6.10422254e-01 2.94426650e-01
1.62105426e-01 -4.75375295e-01 4.16100383e-01 -3.79746050e-01
-7.92309642e-01 1.18598543e-01 -7.96410978e-01 -1.32931814e-01
5.99563062e-01 -2.16899291e-01 -1.05493760e+00 -8.28710645e-02
-3.65967214e-01 7.73595795e-02 5.31691730e-01 4.87659603e-01
7.33146906e-01 -9.90067065e-01 -3.32078546e-01 -4.32843417e-02
3.36141944e-01 -4.80039008e-02 1.82543427e-01 4.87525165e-01
-1.38860226e+00 1.25315261e+00 -7.17103779e-01 -5.39674461e-01
-1.29137254e+00 9.46893096e-01 -3.12270690e-02 -2.23785892e-01
-8.21711183e-01 6.16579831e-01 2.33150452e-01 -4.88717735e-01
3.30553502e-01 -6.57048583e-01 -3.33239168e-01 5.74073233e-02
7.26503789e-01 7.12711737e-02 3.57213348e-01 -1.25255197e-01
-8.03279802e-02 4.26983237e-01 -5.87082744e-01 2.62098342e-01
1.59465015e+00 2.41950870e-01 -4.20609653e-01 2.20629871e-01
1.51537657e+00 -9.85522047e-02 -8.19571614e-01 1.95610672e-01
2.15310648e-01 -2.65614390e-02 8.08610991e-02 -8.39630485e-01
-7.77625799e-01 1.11597085e+00 5.25930703e-01 -4.57807109e-02
8.31466913e-01 -1.76945239e-01 5.85648119e-01 7.74004877e-01
-2.78706431e-01 -1.29907906e+00 3.05118054e-01 4.58171278e-01
1.14903402e+00 -1.05401528e+00 3.43025416e-01 -8.22535083e-02
-1.67595193e-01 1.69549513e+00 6.01627290e-01 1.02189697e-01
3.44428033e-01 4.81339186e-01 3.43400121e-01 -4.24441546e-01
-4.88197237e-01 2.80966669e-01 6.29279673e-01 7.26990759e-01
8.56934726e-01 -1.25161946e-01 -2.69433320e-01 5.62040508e-01
-2.27556542e-01 1.21327594e-01 6.78201556e-01 1.01309037e+00
-2.75411099e-01 -1.15174305e+00 -3.87381732e-01 9.33965147e-01
-5.02399862e-01 -2.08489284e-01 -7.44962692e-01 4.67749596e-01
-2.38576099e-01 5.48716545e-01 4.26793247e-01 2.59787381e-01
6.84187561e-02 7.73143768e-02 8.49131942e-01 -6.58825636e-01
-2.17084721e-01 2.48723775e-01 7.16163814e-02 -2.79634744e-01
-3.74144688e-02 -4.29113090e-01 -1.27777874e+00 -1.36093795e-01
2.78202206e-01 -2.58106142e-01 8.47999215e-01 6.07068956e-01
5.96815050e-01 7.90417552e-01 -1.06012650e-01 -6.55654490e-01
3.31791081e-02 -8.09458554e-01 -3.71301055e-01 2.38435715e-01
1.07413702e-01 -1.98798671e-01 3.43353838e-01 5.22648335e-01] | [11.631739616394043, 2.756769895553589] |
0d907530-4b7b-48df-a888-ed9292bce7d5 | utilizing-multimodal-feature-consistency-to | null | null | https://aclanthology.org/2020.clinicalnlp-1.29 | https://aclanthology.org/2020.clinicalnlp-1.29.pdf | Utilizing Multimodal Feature Consistency to Detect Adversarial Examples on Clinical Summaries | Recent studies have shown that adversarial examples can be generated by applying small perturbations to the inputs such that the well- trained deep learning models will misclassify. With the increasing number of safety and security-sensitive applications of deep learn- ing models, the robustness of deep learning models has become a crucial topic. The robustness of deep learning models for health- care applications is especially critical because the unique characteristics and the high financial interests of the medical domain make it more sensitive to adversarial attacks. Among the modalities of medical data, the clinical summaries have higher risks to be attacked because they are generated by third-party companies. As few works studied adversarial threats on clinical summaries, in this work we first apply adversarial attack to clinical summaries of electronic health records (EHR) to show the text-based deep learning systems are vulnerable to adversarial examples. Secondly, benefiting from the multi-modality of the EHR dataset, we propose a novel defense method, MATCH (Multimodal feATure Consistency cHeck), which leverages the consistency between multiple modalities in the data to defend against adversarial examples on a single modality. Our experiments demonstrate the effectiveness of MATCH on a hospital readmission prediction task comparing with baseline methods. | ['Li Xiong', 'Pengfei Tang', 'Ian Molloy', 'Taesung Lee', 'Youngja Park', 'Wenjie Wang'] | null | null | null | null | emnlp-clinicalnlp-2020-11 | ['readmission-prediction'] | ['medical'] | [ 1.60755783e-01 3.07428330e-01 1.20652348e-01 -3.33798677e-01
-1.02199221e+00 -9.67976093e-01 5.23376763e-01 5.23974717e-01
-2.32315689e-01 5.56755722e-01 4.39588606e-01 -6.09414458e-01
-1.83669284e-01 -8.41530144e-01 -9.57359612e-01 -6.57711685e-01
-2.73943901e-01 1.66075751e-01 -3.42502803e-01 -1.59965247e-01
-2.55592763e-01 4.25964922e-01 -5.88135779e-01 8.65722537e-01
6.70180500e-01 9.88009751e-01 -7.91300774e-01 7.40639329e-01
4.73035306e-01 9.46349919e-01 -7.38425970e-01 -9.67304051e-01
6.72978044e-01 -3.26044947e-01 -6.99699402e-01 -6.22050405e-01
4.50539619e-01 -6.01990581e-01 -8.00011814e-01 1.34650707e+00
9.30136204e-01 -1.53088510e-01 6.57754242e-01 -1.46423852e+00
-9.08940196e-01 8.85285497e-01 -6.19828552e-02 6.40614182e-02
2.74966985e-01 4.16951150e-01 6.29968047e-01 -3.68905187e-01
3.81420076e-01 1.16681600e+00 7.87259579e-01 9.93532360e-01
-8.89620185e-01 -8.90022695e-01 -1.65195707e-02 -1.15792394e-01
-9.81140375e-01 -3.08428258e-01 6.25101507e-01 -4.91828173e-01
5.54533362e-01 4.44597304e-01 1.04728490e-02 1.92565167e+00
9.22942340e-01 6.42218947e-01 8.90896082e-01 9.75387990e-02
2.65250266e-01 1.61219090e-01 6.29261583e-02 4.90858227e-01
4.54271615e-01 4.63857859e-01 -1.84984222e-01 -9.27876115e-01
2.35880107e-01 6.30009472e-01 -4.89084542e-01 -1.08620703e-01
-1.31176543e+00 1.02201033e+00 5.58506846e-01 4.66045700e-02
-3.12983364e-01 1.34208098e-01 1.03893137e+00 3.92055809e-01
7.04394281e-02 6.50874913e-01 -5.23162842e-01 2.46509269e-01
-2.63995498e-01 2.14754567e-01 6.88151300e-01 7.75815487e-01
-3.16440731e-01 -8.08858126e-03 -4.66972768e-01 2.46707425e-01
2.27142498e-01 6.53787136e-01 5.42381525e-01 -4.26099062e-01
7.86528170e-01 5.31466424e-01 -2.00774670e-01 -1.31007957e+00
-3.27521622e-01 -3.21453869e-01 -1.51684093e+00 4.46193926e-02
3.39861274e-01 -4.75893825e-01 -7.78406262e-01 1.91311824e+00
1.99109644e-01 1.30567268e-01 6.94663107e-01 6.20217204e-01
9.31363821e-01 2.92627931e-01 3.40016752e-01 1.86741084e-01
1.41423750e+00 -2.44021386e-01 -8.19752038e-01 7.65071586e-02
7.02367961e-01 -5.04784346e-01 7.74351239e-01 3.56215686e-01
-9.01847422e-01 -2.31269673e-01 -1.00977671e+00 3.42459440e-01
-6.54959321e-01 -6.96054399e-01 3.95820737e-01 8.37059498e-01
-4.87558156e-01 5.62695026e-01 -8.78729224e-01 5.34338169e-02
8.40862870e-01 2.70002335e-01 -6.36398017e-01 -6.28425926e-02
-1.96822989e+00 8.11632156e-01 2.40412056e-01 -5.72653897e-02
-1.12432766e+00 -9.44097161e-01 -1.03004062e+00 1.65894419e-01
-5.57891354e-02 -8.27387512e-01 1.05955660e+00 -8.66040051e-01
-7.42581606e-01 8.09921205e-01 7.43196487e-01 -7.70365834e-01
1.04829800e+00 -3.87714505e-01 -7.14317143e-01 2.70607293e-01
-2.96875685e-01 -5.50009906e-02 9.23768461e-01 -1.06928861e+00
-2.06707343e-01 -4.60325122e-01 8.92657414e-02 -3.37771446e-01
-6.73471332e-01 2.53221276e-03 3.67090970e-01 -1.05347109e+00
-5.72106719e-01 -1.07601070e+00 -4.33394939e-01 -1.65708587e-01
-9.47830081e-01 2.36450300e-01 7.94788957e-01 -5.33951283e-01
1.27185881e+00 -2.33016610e+00 -3.05136532e-01 2.12435856e-01
4.34738725e-01 5.45412302e-01 -4.54331040e-02 4.82543766e-01
-4.64036047e-01 5.24000406e-01 -2.59188503e-01 2.27444842e-02
3.03587597e-02 4.42181081e-02 -9.35152292e-01 6.18331194e-01
3.31677586e-01 1.11358261e+00 -9.10457551e-01 -2.99663246e-01
-8.36929455e-02 5.33994436e-01 -5.44777095e-01 5.04221201e-01
1.35926127e-01 4.03932959e-01 -5.84204078e-01 7.27870226e-01
6.94529653e-01 -2.07760602e-01 7.98423029e-03 -3.18255216e-01
8.52668941e-01 -1.01020597e-01 -7.70713508e-01 1.09078825e+00
-1.06830627e-01 2.63428599e-01 -3.23635072e-01 -8.53930116e-01
5.25829315e-01 7.03094423e-01 4.58321512e-01 -3.62095356e-01
3.55205059e-01 -1.14492983e-01 1.39925927e-01 -7.04602718e-01
9.64106247e-02 -2.65560359e-01 -6.44538760e-01 6.19295120e-01
-3.53741616e-01 1.47764251e-01 -7.03241110e-01 6.43156230e-01
1.43983972e+00 -5.14907718e-01 2.37596661e-01 5.58475330e-02
2.67455369e-01 -1.66867211e-01 6.12047434e-01 1.10409927e+00
-5.42568624e-01 8.39207828e-01 4.35621291e-01 -7.60807753e-01
-8.07068944e-01 -1.27367556e+00 -4.13525045e-01 4.59295541e-01
-2.10041419e-01 -6.77286983e-02 -7.87418485e-01 -1.36668622e+00
3.70561391e-01 4.31613684e-01 -9.79782462e-01 -1.04921436e+00
-4.74925280e-01 -8.26437891e-01 1.23132372e+00 7.78014779e-01
3.32139552e-01 -9.01266038e-01 -5.74896455e-01 1.29198674e-02
-4.75865267e-02 -1.00362492e+00 -6.06803298e-01 4.78502996e-02
-5.35617948e-01 -1.52255380e+00 -5.04786134e-01 -3.70326221e-01
6.26026273e-01 -4.20775801e-01 1.04419088e+00 7.95049593e-02
-4.04853761e-01 6.54316485e-01 -2.32432947e-01 -8.95195901e-01
-1.17565334e+00 -1.13752626e-01 4.22287375e-01 1.68482438e-01
3.21358621e-01 -9.34837312e-02 -6.74045146e-01 -2.47644037e-02
-1.48276281e+00 -5.09861767e-01 2.81937897e-01 9.96212423e-01
2.62760460e-01 -9.97162461e-02 1.05283201e+00 -1.38659024e+00
9.58359897e-01 -9.08844471e-01 -1.76013798e-01 3.39837819e-01
-3.85124564e-01 -1.07914239e-01 1.06015909e+00 -7.09461629e-01
-6.52250826e-01 1.41580496e-02 -5.35943061e-02 -5.97070515e-01
-2.12424904e-01 5.04349053e-01 -3.64404857e-01 1.89625382e-01
8.77186418e-01 -1.05581127e-01 -1.83305312e-02 -1.43410578e-01
8.39384720e-02 7.61844337e-01 5.88481367e-01 -2.75331795e-01
9.57789600e-01 4.55451339e-01 4.24866490e-02 -1.75169826e-01
-7.60593176e-01 1.84832186e-01 -1.11709245e-01 3.06255594e-02
9.76211905e-01 -7.73620963e-01 -1.08012652e+00 5.85320652e-01
-1.15565121e+00 6.84176683e-02 -6.38573021e-02 4.74189043e-01
-3.01156789e-01 5.62451303e-01 -8.19820583e-01 -6.33822143e-01
-7.81084538e-01 -1.07771671e+00 7.26784587e-01 -2.03452617e-01
-3.38511318e-01 -1.10142100e+00 1.57886192e-01 2.26628199e-01
4.08571512e-01 1.09500277e+00 1.29291999e+00 -1.53177726e+00
-3.69547129e-01 -7.58668125e-01 2.59496421e-01 5.92166483e-01
2.88917363e-01 -3.63007467e-03 -1.18315780e+00 -5.60336769e-01
2.36193359e-01 -4.65244055e-01 5.72198927e-01 1.62746087e-01
1.27240217e+00 -8.04965496e-01 -2.77247161e-01 7.79984176e-01
1.13156128e+00 1.80599779e-01 5.87908030e-01 3.06582123e-01
7.40537286e-01 6.43002510e-01 4.91801679e-01 4.57142860e-01
1.35434389e-01 8.31407383e-02 9.58813727e-01 -3.10023069e-01
5.76245904e-01 -2.26691291e-01 3.20584267e-01 4.24581230e-01
4.78793412e-01 -3.99806887e-01 -1.08030391e+00 4.10097718e-01
-1.87021279e+00 -1.07374036e+00 3.11670359e-02 2.17654085e+00
9.50901091e-01 7.18596727e-02 -1.24571823e-01 5.78621253e-02
6.44367456e-01 -8.38183984e-02 -8.50866079e-01 -4.64499980e-01
-2.40957245e-01 -1.09878637e-01 4.16563243e-01 1.70394704e-02
-1.51431406e+00 3.55850726e-01 6.42894602e+00 2.78026193e-01
-1.01148868e+00 -5.81921674e-02 1.03013361e+00 -2.06448674e-01
-2.22303540e-01 -8.07910383e-01 -4.08828288e-01 7.92882979e-01
1.10691667e+00 -2.62493998e-01 -1.99630260e-01 9.42016006e-01
-3.25674899e-02 6.87120676e-01 -1.58922637e+00 7.54763365e-01
-3.83036844e-02 -1.40383613e+00 3.73385727e-01 8.22227746e-02
7.66677618e-01 -5.14800139e-02 5.74836850e-01 2.89022833e-01
7.44241536e-01 -1.37968373e+00 3.45210075e-01 5.42268872e-01
7.10775554e-01 -1.09441578e+00 1.21564150e+00 2.51556009e-01
-5.40246129e-01 -2.33002409e-01 -1.73143625e-01 5.93254030e-01
-2.37694774e-02 3.55529428e-01 -7.47464895e-01 6.08568549e-01
9.23762798e-01 4.85513657e-01 -4.78289515e-01 4.65281457e-01
1.98518991e-01 5.48892558e-01 9.92926732e-02 4.50814247e-01
7.71557540e-02 4.23214853e-01 4.53686118e-01 1.30474007e+00
1.08784050e-01 9.50489342e-02 4.18614522e-02 6.55465603e-01
-5.14837444e-01 -1.94219857e-01 -1.38172829e+00 -2.90301889e-01
2.87002802e-01 9.86161292e-01 3.02579701e-02 -2.81496853e-01
-1.71792805e-01 7.34319150e-01 -5.22767752e-02 2.26282358e-01
-9.12151992e-01 -4.15520728e-01 9.63102460e-01 -1.53231725e-01
-2.38469705e-01 3.61774474e-01 -2.95075685e-01 -1.18577814e+00
-6.85441345e-02 -1.59062016e+00 1.01629341e+00 -4.00193512e-01
-1.96385503e+00 9.29616153e-01 -3.03380311e-01 -1.55066800e+00
-3.61603796e-01 -4.96237427e-01 -4.70360368e-01 8.75715673e-01
-1.23547029e+00 -1.13493347e+00 7.97932297e-02 1.05648959e+00
-3.08396034e-02 -6.37100875e-01 1.20193708e+00 1.31118163e-01
-5.40711582e-01 1.29821312e+00 1.94528863e-01 7.53853261e-01
9.09100056e-01 -1.10892332e+00 5.23019314e-01 9.77691174e-01
-3.03842038e-01 1.00333714e+00 4.03259993e-01 -7.49009252e-01
-1.41711962e+00 -1.43195415e+00 5.73099613e-01 -1.11703873e+00
6.75453484e-01 -3.77644360e-01 -1.24073648e+00 9.41022158e-01
2.05466613e-01 2.64155060e-01 1.29930127e+00 -5.49061418e-01
-7.24994779e-01 -1.23145264e-02 -1.64508104e+00 7.48342037e-01
5.57626188e-01 -7.56631136e-01 -7.24985778e-01 4.51246411e-01
9.59791124e-01 -4.51051772e-01 -1.30829716e+00 6.46382749e-01
5.99852145e-01 -6.57228053e-01 1.21729016e+00 -1.67150164e+00
8.38754773e-01 1.60242505e-02 -1.78118721e-01 -1.33584166e+00
-1.02211736e-01 -7.79598355e-01 -3.20969582e-01 1.11216855e+00
3.19609910e-01 -9.08801973e-01 4.03535336e-01 1.06940103e+00
1.48146808e-01 -6.48431420e-01 -8.54683638e-01 -5.98479450e-01
3.61408412e-01 -2.00683951e-01 9.18438733e-01 1.62316644e+00
1.29445702e-01 -2.33845696e-01 -5.64137280e-01 7.27004945e-01
4.89584476e-01 -1.59865484e-01 6.73924088e-01 -9.34196532e-01
-3.56947839e-01 7.13289389e-03 -5.27198911e-01 -6.39359057e-02
-1.13405082e-02 -9.76469159e-01 -2.16635615e-01 -9.75013077e-01
1.77460551e-01 -2.84190089e-01 -8.34116638e-01 6.97273493e-01
-6.32604957e-01 1.05772682e-01 2.83667505e-01 1.77121744e-01
-2.45357573e-01 2.44359076e-01 8.18677425e-01 -4.37206745e-01
2.18721017e-01 2.09245503e-01 -1.01023912e+00 7.18217969e-01
6.92369580e-01 -9.25238431e-01 -2.25088581e-01 -3.29098105e-01
2.93131173e-01 4.12717238e-02 4.07027572e-01 -5.27829468e-01
1.53952971e-01 -1.40397679e-02 4.59277689e-01 -3.25713098e-01
-1.71030551e-01 -1.33867764e+00 1.40971795e-01 9.85146999e-01
-7.39876628e-01 4.14513946e-01 4.26031828e-01 8.42507005e-01
-1.50293857e-01 1.38181105e-01 7.59515285e-01 -1.46651909e-01
1.57022431e-01 5.51203787e-01 -3.92067701e-01 2.58116007e-01
1.16579998e+00 1.59029275e-01 -6.71249330e-01 -3.38847667e-01
-7.07772493e-01 2.46756315e-01 2.13599920e-01 6.35875404e-01
7.87878513e-01 -1.41424692e+00 -9.28967237e-01 2.37257272e-01
3.79883170e-01 -3.21784206e-02 4.50052887e-01 4.79454786e-01
-2.37219423e-01 1.21510802e-02 -1.26756653e-01 -4.75510567e-01
-1.42963004e+00 1.19323266e+00 5.74921250e-01 -3.66208166e-01
-6.35650754e-01 5.83604157e-01 5.86034477e-01 -4.66008514e-01
4.82246071e-01 -1.91847026e-01 9.23160836e-02 -2.48482525e-01
7.67390609e-01 1.28321424e-01 1.46664202e-01 -3.31587434e-01
-6.69686437e-01 -3.06707695e-02 -4.42963809e-01 4.14860934e-01
1.33502281e+00 4.41771060e-01 1.72713816e-01 6.17464148e-02
1.29329967e+00 -6.39874069e-03 -8.77444506e-01 -1.70684651e-01
-1.25865087e-01 -2.34584630e-01 -2.76275307e-01 -1.09257150e+00
-1.22937012e+00 9.72592950e-01 7.48343110e-01 5.78988135e-01
9.58482563e-01 -3.64162415e-01 9.68775868e-01 4.32811201e-01
-1.40537575e-01 -5.40406823e-01 2.37201020e-01 2.17424437e-01
8.63436997e-01 -1.51830924e+00 -2.70151615e-01 8.87073725e-02
-9.99782324e-01 9.47812319e-01 4.37919766e-01 6.18853793e-02
6.35862887e-01 5.18376052e-01 5.84724486e-01 -1.69810176e-01
-4.87196535e-01 8.49422932e-01 2.48452485e-01 8.24844837e-01
1.70905322e-01 1.08181328e-01 2.78364062e-01 1.06536257e+00
2.05969159e-03 -3.08503479e-01 5.53227961e-01 9.51532543e-01
3.52018058e-01 -9.32327390e-01 -5.70319831e-01 5.13851404e-01
-1.24694073e+00 -3.42525274e-01 -5.99463940e-01 5.45462966e-01
-1.98246196e-01 9.30884123e-01 -3.44992369e-01 -5.84778965e-01
5.79265237e-01 9.29971784e-02 -1.67389378e-01 -4.50437635e-01
-1.38176799e+00 -4.73515838e-01 -9.55778584e-02 -5.10578871e-01
6.34458512e-02 -4.21041876e-01 -1.06644094e+00 -4.62192059e-01
3.49750929e-02 -6.38553351e-02 3.50759774e-01 6.41733646e-01
7.86645889e-01 7.28718936e-01 9.17345047e-01 -1.18310414e-01
-1.29354787e+00 -6.52908683e-01 -3.60637724e-01 1.06071115e+00
8.43587756e-01 -1.28879070e-01 -5.25767505e-01 8.02213475e-02] | [5.771636009216309, 7.690955638885498] |
d1f7f469-d000-4640-9fdb-78c44ed12955 | summarizing-opinions-aspect-extraction-meets | 1808.08858 | null | http://arxiv.org/abs/1808.08858v1 | http://arxiv.org/pdf/1808.08858v1.pdf | Summarizing Opinions: Aspect Extraction Meets Sentiment Prediction and They Are Both Weakly Supervised | We present a neural framework for opinion summarization from online product
reviews which is knowledge-lean and only requires light supervision (e.g., in
the form of product domain labels and user-provided ratings). Our method
combines two weakly supervised components to identify salient opinions and form
extractive summaries from multiple reviews: an aspect extractor trained under a
multi-task objective, and a sentiment predictor based on multiple instance
learning. We introduce an opinion summarization dataset that includes a
training set of product reviews from six diverse domains and human-annotated
development and test sets with gold standard aspect annotations, salience
labels, and opinion summaries. Automatic evaluation shows significant
improvements over baselines, and a large-scale study indicates that our opinion
summaries are preferred by human judges according to multiple criteria. | ['Stefanos Angelidis', 'Mirella Lapata'] | 2018-08-27 | summarizing-opinions-aspect-extraction-meets-1 | https://aclanthology.org/D18-1403 | https://aclanthology.org/D18-1403.pdf | emnlp-2018-10 | ['aspect-extraction'] | ['natural-language-processing'] | [ 5.99686325e-01 6.17715299e-01 -9.32442844e-01 -7.40548372e-01
-1.31771791e+00 -7.53586769e-01 5.46978891e-01 8.04256201e-01
-2.38418967e-01 6.89652801e-01 9.09321189e-01 -4.86234799e-02
2.28993088e-01 -4.91733998e-01 -5.05803883e-01 -1.79763243e-01
4.14105952e-01 5.14040411e-01 -2.08176434e-01 -3.74506652e-01
6.34032488e-01 -3.80371243e-01 -1.39478350e+00 7.78013289e-01
1.23263943e+00 1.11222982e+00 -4.08453375e-01 7.32789636e-01
-6.74990490e-02 7.74892509e-01 -1.01390719e+00 -1.16881359e+00
2.30124947e-02 -4.13995594e-01 -5.79650104e-01 4.35206652e-01
9.10221159e-01 -2.69961953e-01 6.98984504e-01 8.65579426e-01
5.68731487e-01 2.44269356e-01 9.88467395e-01 -7.72968471e-01
-1.37862027e+00 6.75207615e-01 -8.59884977e-01 -1.83322392e-02
4.20300663e-01 -2.45532505e-02 1.91336119e+00 -9.39474523e-01
6.25417888e-01 1.06279528e+00 6.41682208e-01 2.53940552e-01
-1.32823431e+00 -1.58040687e-01 6.62243605e-01 -3.22411776e-01
-3.93281490e-01 -6.08269155e-01 7.21117377e-01 -2.76667535e-01
1.57341623e+00 -1.41575202e-01 5.69533885e-01 1.02911234e+00
3.90552074e-01 1.09232795e+00 5.16613841e-01 -3.79894465e-01
5.16151965e-01 5.86476684e-01 6.73669159e-01 2.76225626e-01
8.49627852e-01 -6.44290924e-01 -7.33381867e-01 -4.39348072e-01
6.05843775e-02 -3.38969767e-01 -3.67419571e-02 -2.40908340e-01
-7.49164164e-01 9.87897336e-01 4.86154994e-03 -1.10240109e-01
-8.60402465e-01 -2.94506580e-01 5.86653888e-01 3.17862898e-01
1.33988512e+00 1.03899908e+00 -1.05164611e+00 1.10167548e-01
-8.98365200e-01 3.64503264e-01 1.30215240e+00 8.40330839e-01
7.06376731e-01 2.99606144e-01 -2.95352608e-01 9.95428979e-01
2.17874080e-01 8.92718077e-01 5.19857645e-01 -9.21753049e-01
5.73304653e-01 8.29417050e-01 3.16158056e-01 -1.11408830e+00
-1.97245747e-01 -5.52462518e-01 -4.21081990e-01 9.44999456e-02
-4.49372321e-01 -7.16631114e-01 -8.90269578e-01 1.31555831e+00
1.73454415e-02 -6.17502451e-01 4.55921113e-01 4.81028557e-01
1.33960640e+00 5.94425082e-01 1.57317489e-01 -4.95237619e-01
1.31324136e+00 -1.30100036e+00 -8.65615666e-01 -6.34668350e-01
5.95358849e-01 -7.06071377e-01 9.13387418e-01 7.51577973e-01
-1.38584077e+00 -4.57334280e-01 -1.32601404e+00 -7.33954906e-02
-4.55026567e-01 5.21927118e-01 6.33454859e-01 3.88411164e-01
-1.05272031e+00 3.83107841e-01 -3.53337467e-01 -1.05910636e-01
5.89379251e-01 2.94108748e-01 -3.33846301e-01 1.68018237e-01
-8.30281913e-01 1.00375080e+00 6.66882172e-02 -1.90990791e-01
-5.03826559e-01 -5.51165521e-01 -1.36530185e+00 2.60530040e-02
4.02306259e-01 -9.88067925e-01 1.75049138e+00 -1.68218708e+00
-1.46724355e+00 7.11120427e-01 -4.24371064e-01 -3.27018529e-01
-3.01735580e-01 -7.51801789e-01 -4.70935404e-01 2.62013137e-01
6.30292714e-01 5.31481147e-01 9.23236549e-01 -1.20714283e+00
-8.84012043e-01 -3.49023968e-01 2.24879369e-01 5.69850862e-01
-5.49302161e-01 7.53653944e-02 -1.06625043e-01 -5.03926992e-01
-5.04836559e-01 -6.15104795e-01 -5.12857199e-01 -7.81811833e-01
-5.93414724e-01 -4.91344661e-01 3.09846193e-01 -7.57776737e-01
1.05496180e+00 -1.69622684e+00 5.47961146e-02 -4.59690206e-02
2.50795782e-01 1.89598322e-01 -5.43787241e-01 1.85741231e-01
-8.16558152e-02 2.65781045e-01 -6.77167848e-02 -7.23432899e-01
1.50384039e-01 -4.49769586e-01 -4.00530577e-01 -1.19468249e-01
5.92720449e-01 1.03681219e+00 -1.04657698e+00 -3.14348847e-01
-3.04610878e-01 1.47917882e-01 -5.25100946e-01 -1.82812121e-02
-6.38212562e-01 -2.54764259e-01 -4.56077576e-01 8.03987026e-01
2.63450176e-01 -4.21375006e-01 -8.65515545e-02 -4.19095129e-01
1.61957696e-01 6.50154173e-01 -7.09302127e-01 1.36911345e+00
-5.24474680e-01 4.14398819e-01 -1.57776371e-01 -5.94273686e-01
1.03835237e+00 4.27638054e-01 2.43241206e-01 -5.39890766e-01
1.76517278e-01 1.81210458e-01 -4.72680271e-01 -3.58527809e-01
1.10530210e+00 4.37204493e-03 -3.38246077e-01 9.40839350e-01
4.42479521e-01 -5.61729431e-01 6.85796082e-01 4.89259273e-01
1.06947458e+00 2.80424953e-02 7.10115194e-01 -8.91195387e-02
3.12005132e-01 4.18419510e-01 6.50460720e-01 5.70834398e-01
1.40157202e-02 7.21792161e-01 7.69844174e-01 -4.55192804e-01
-9.18019295e-01 -5.99799752e-01 3.24940354e-01 1.45289874e+00
-1.31251514e-01 -7.87990212e-01 -3.99069011e-01 -1.21203208e+00
3.96643579e-02 1.12846696e+00 -6.86887681e-01 -2.38470867e-01
-1.99569210e-01 -6.30309701e-01 -1.72061384e-01 6.70915723e-01
3.03176627e-03 -1.25049865e+00 -1.48520932e-01 1.29325897e-01
3.00863869e-02 -9.96576607e-01 -6.25807166e-01 3.05752933e-01
-7.47018337e-01 -9.74669456e-01 -6.77412450e-01 -9.81272340e-01
7.41775990e-01 2.60906875e-01 1.90311170e+00 -4.74683434e-01
4.46016163e-01 4.88417357e-01 -4.51691598e-01 -1.04070222e+00
-3.43076825e-01 3.49823147e-01 1.95294879e-02 -1.88867286e-01
9.34742332e-01 -2.73956150e-01 -4.95779872e-01 -1.29488111e-01
-6.55595541e-01 -1.89614087e-01 9.71483052e-01 6.23970032e-01
5.73565364e-01 -1.50494382e-01 1.16501200e+00 -1.19780529e+00
1.37318552e+00 -4.17861611e-01 -1.72343895e-01 2.57330865e-01
-8.79445255e-01 -1.47805154e-01 4.97860610e-01 -2.74065107e-01
-1.44992232e+00 -1.76867753e-01 2.12340578e-01 2.67301232e-01
-3.34104784e-02 9.26988125e-01 -9.37298834e-02 6.90222383e-01
1.04838490e+00 -2.24116698e-01 -1.83006465e-01 -9.57377627e-02
7.81290114e-01 6.62105799e-01 3.14877659e-01 -1.94602817e-01
3.46694559e-01 1.17178023e-01 -6.70825839e-01 -6.56708837e-01
-1.61926365e+00 -4.75435823e-01 -5.48328459e-01 1.29265174e-01
6.59225941e-01 -1.24013937e+00 -1.46820378e-02 7.14506134e-02
-1.36903858e+00 5.79771660e-02 -7.48293459e-01 3.84845376e-01
-2.11121291e-01 -8.38133693e-03 -6.25367701e-01 -6.81204677e-01
-1.19494045e+00 -6.61373079e-01 1.31128287e+00 5.69415689e-01
-9.20530856e-01 -1.11468792e+00 3.80751848e-01 5.91404915e-01
3.82434338e-01 2.47747138e-01 8.10828447e-01 -1.16461146e+00
1.58791393e-02 -6.41723335e-01 1.56771895e-02 8.88227820e-01
4.07204360e-01 2.17855170e-01 -6.87392831e-01 -2.13884767e-02
-5.37446979e-03 -8.32832396e-01 1.23164546e+00 8.56846035e-01
3.04954231e-01 -5.94646215e-01 -1.71882197e-01 -2.71368414e-01
1.04721963e+00 -7.95258135e-02 1.98759395e-03 1.04586512e-01
7.20676899e-01 9.23996270e-01 6.51394963e-01 3.21508378e-01
6.72586977e-01 5.62432036e-02 -2.71215253e-02 -1.21403798e-01
1.55685484e-01 -9.02184919e-02 6.26059115e-01 1.01501644e+00
7.22601116e-02 -4.58915949e-01 -3.58506858e-01 1.04394186e+00
-1.96916986e+00 -7.67788529e-01 1.10348865e-01 1.71761310e+00
1.07208776e+00 5.45610547e-01 1.65820658e-01 -3.71397495e-01
5.14729321e-01 6.22296631e-01 -9.29162502e-01 -1.02679074e+00
-3.00458938e-01 2.09745318e-01 1.69994012e-01 4.48844701e-01
-1.15977037e+00 7.99146533e-01 6.74051762e+00 4.49704498e-01
-5.61800659e-01 -4.85004224e-02 8.88269722e-01 -5.10412574e-01
-9.81234789e-01 -2.14156494e-01 -8.48531067e-01 -8.91278535e-02
9.51045632e-01 -6.01996660e-01 -3.48167956e-01 1.21963274e+00
9.69531760e-02 -1.92848310e-01 -1.06804347e+00 4.80239481e-01
8.35300207e-01 -1.27067769e+00 3.71871561e-01 -1.57901004e-01
1.53621948e+00 4.41556387e-02 1.44162729e-01 4.86287147e-01
9.30022597e-01 -7.11601138e-01 3.76173675e-01 2.15617239e-01
4.69768465e-01 -8.92009079e-01 1.18484259e+00 -3.18160094e-02
-8.34400177e-01 1.63505420e-01 -2.88052768e-01 -1.83659628e-01
4.15445209e-01 9.94598389e-01 -5.85382342e-01 5.14602780e-01
3.50446403e-01 1.55454719e+00 -6.64149106e-01 4.81712937e-01
-7.68480718e-01 5.56030989e-01 2.28059411e-01 -2.64423907e-01
1.52902439e-01 -1.94867730e-01 5.24345994e-01 1.19170988e+00
-1.15875982e-01 -1.54781610e-01 2.50313282e-01 6.21265829e-01
-3.49919617e-01 4.80158895e-01 -7.36870587e-01 -2.82799155e-01
3.00272629e-02 1.72114265e+00 -4.52882081e-01 -6.45885944e-01
-6.45984352e-01 7.04196215e-01 1.27757758e-01 5.08429646e-01
-1.29256666e-01 -5.92599154e-01 4.53953236e-01 -3.76446635e-01
6.89934909e-01 3.01200151e-01 -8.19443047e-01 -1.24479246e+00
1.29642054e-01 -1.09601283e+00 1.73507378e-01 -7.65788317e-01
-1.45542681e+00 7.83421040e-01 -2.32018054e-01 -1.09769690e+00
-4.76165444e-01 -3.94859403e-01 -9.72183108e-01 6.43339097e-01
-1.57232583e+00 -1.03401732e+00 1.73421949e-01 -1.92957491e-01
9.63184595e-01 -4.71470743e-01 8.03975284e-01 -2.08517939e-01
-6.30796373e-01 4.28651452e-01 -3.07668924e-01 -1.38315305e-01
1.00015223e+00 -1.74149454e+00 7.58530200e-01 6.80483699e-01
1.41541347e-01 7.34163105e-01 8.10967982e-01 -9.32750523e-01
-8.59955430e-01 -1.00462234e+00 1.26680195e+00 -8.78978372e-01
4.88927573e-01 1.14697464e-01 -6.84722841e-01 6.89037442e-01
1.05007672e+00 -7.45836198e-01 1.43242145e+00 7.21684039e-01
-3.71839553e-01 -1.97438374e-01 -8.96166682e-01 4.67037320e-01
4.49462444e-01 -3.57765973e-01 -1.09952366e+00 4.48472917e-01
1.04316354e+00 -1.20543890e-01 -7.60475278e-01 4.34144706e-01
5.61227977e-01 -6.07149720e-01 4.84887093e-01 -7.01338947e-01
1.10593271e+00 -1.28996477e-01 9.70515087e-02 -1.90983593e+00
-2.10172623e-01 -3.77214432e-01 -3.68319273e-01 1.37597263e+00
1.26385427e+00 -3.58260244e-01 6.48469269e-01 8.50381434e-01
-3.09500515e-01 -9.96043026e-01 8.71036295e-03 -4.67341989e-02
-1.97340280e-01 -7.59857967e-02 2.52024621e-01 5.50034702e-01
3.41865838e-01 1.54949605e+00 -1.50617510e-01 -1.67275935e-01
4.54961747e-01 4.26130831e-01 7.73437560e-01 -1.28276753e+00
-1.81676313e-01 -5.96966565e-01 1.33938417e-01 -9.51211989e-01
4.32901084e-01 -5.36521971e-01 1.85916081e-01 -2.15495086e+00
3.97448123e-01 5.07967174e-01 -3.85866255e-01 5.41139364e-01
-4.61131632e-01 6.54844120e-02 -2.99780846e-01 6.61644638e-02
-1.46325338e+00 6.63338900e-01 1.07207382e+00 -4.94417071e-01
-4.61809397e-01 2.62469471e-01 -1.80367231e+00 9.50024426e-01
7.52831817e-01 -3.49659413e-01 -5.75400770e-01 -4.49173570e-01
9.11306500e-01 -4.44973320e-01 -4.06714797e-01 -3.99838865e-01
2.01695681e-01 5.32886237e-02 4.32751507e-01 -9.01518047e-01
1.04220048e-01 -2.43493274e-01 -7.79822588e-01 -1.93810329e-01
-9.06856656e-01 2.68644392e-01 1.76339597e-01 6.93796217e-01
-3.87109399e-01 -1.68333352e-01 2.10281864e-01 -3.33389997e-01
-4.32214677e-01 -2.04752143e-02 -1.94911599e-01 2.08631620e-01
4.55761373e-01 -8.32895190e-02 -7.12408185e-01 -7.70702481e-01
-3.86936545e-01 5.20265818e-01 2.95775235e-01 5.82145870e-01
5.25653839e-01 -1.14180768e+00 -1.06209898e+00 -1.70758605e-01
4.23249543e-01 7.89978355e-02 -8.25301036e-02 4.35098767e-01
2.96718240e-01 4.70008850e-01 2.10770428e-01 -2.86361992e-01
-1.08388388e+00 2.08068684e-01 -1.09415285e-01 -5.68564534e-01
-6.07012473e-02 8.12813282e-01 4.55211438e-02 -5.07066190e-01
9.76135954e-02 -5.07597506e-01 -8.05708647e-01 6.03393674e-01
9.04916942e-01 1.76212981e-01 2.23470852e-01 -5.64788997e-01
-7.90177509e-02 5.60976207e-01 -7.21868157e-01 -2.52388984e-01
1.45000613e+00 -1.67680055e-01 -1.23665825e-01 6.91562653e-01
1.01083934e+00 1.97024539e-01 -9.69174743e-01 -4.33308095e-01
1.46880880e-01 8.39480609e-02 5.28983176e-02 -1.10684478e+00
-7.53930688e-01 1.61377043e-01 -2.40365535e-01 4.71290499e-01
9.26393628e-01 1.12622134e-01 8.60853255e-01 6.85771167e-01
-2.46959850e-01 -1.47043169e+00 5.80945849e-01 5.14779747e-01
1.03264832e+00 -1.73914182e+00 4.33585525e-01 -1.56549633e-01
-1.32284737e+00 7.21376240e-01 7.86478877e-01 -3.39121729e-01
5.58883488e-01 -1.47878304e-01 3.53117824e-01 -5.45212269e-01
-1.36967576e+00 -1.38154045e-01 8.40044558e-01 5.87761641e-01
7.11142421e-01 -4.89918701e-02 -4.47347075e-01 1.16560066e+00
-2.08420873e-01 -2.36263007e-01 5.37670016e-01 8.17182600e-01
-6.19480371e-01 -7.87003994e-01 2.68298924e-01 1.03959453e+00
-6.49660945e-01 -2.42792562e-01 -9.30231929e-01 1.24590553e-01
-2.28172839e-01 1.47986758e+00 -1.86169162e-01 -1.85734421e-01
7.35931098e-01 -1.60541683e-01 -5.67650832e-02 -1.32418108e+00
-9.12121356e-01 1.00698620e-01 7.96728313e-01 -3.08473259e-01
-7.71385133e-01 -6.17508709e-01 -7.58865654e-01 2.24677503e-01
-6.04775071e-01 4.45744991e-01 6.42876387e-01 1.05624580e+00
7.86213875e-01 6.52175903e-01 6.65275216e-01 -8.86536241e-01
-5.31965315e-01 -1.23029482e+00 -4.03775901e-01 5.26760519e-01
5.26917636e-01 -6.69645816e-02 -5.16756594e-01 1.92410201e-01] | [11.43282413482666, 6.756292343139648] |
76b2d38e-bddd-4e0a-b826-63d77c03e9c0 | forecasting-algorithms-for-causal-inference | 2208.03489 | null | https://arxiv.org/abs/2208.03489v1 | https://arxiv.org/pdf/2208.03489v1.pdf | Forecasting Algorithms for Causal Inference with Panel Data | Conducting causal inference with panel data is a core challenge in social science research. Advances in forecasting methods can facilitate this task by more accurately predicting the counterfactual evolution of a treated unit had treatment not occurred. In this paper, we draw on a newly developed deep neural architecture for time series forecasting (the N-BEATS algorithm). We adapt this method from conventional time series applications by incorporating leading values of control units to predict a "synthetic" untreated version of the treated unit in the post-treatment period. We refer to the estimator derived from this method as SyNBEATS, and find that it significantly outperforms traditional two-way fixed effects and synthetic control methods across a range of settings. We also find that SyNBEATS attains comparable or more accurate performance relative to more recent panel estimation methods such as matrix completion and synthetic difference in differences. Our results highlight how advances in the forecasting literature can be harnessed to improve causal inference in panel settings. | ['Justin Young', 'Julian Nyarko', 'Jacob Goldin'] | 2022-08-06 | null | null | null | null | ['matrix-completion'] | ['methodology'] | [ 3.20763111e-01 1.93326265e-01 -5.45409620e-01 -4.98606980e-01
-5.09765685e-01 -6.18384719e-01 1.10297751e+00 -5.60446270e-02
-2.66811214e-02 1.31942689e+00 1.21184027e+00 -7.27405429e-01
-1.90601975e-01 -8.48257840e-01 -9.73389208e-01 -6.33843005e-01
-2.71830291e-01 1.90604970e-01 -8.78207386e-01 2.52521373e-02
2.60278076e-01 8.28085616e-02 -1.19193172e+00 3.89227062e-01
8.33178937e-01 4.76899862e-01 -6.08733594e-01 3.49921167e-01
5.24324834e-01 9.69008446e-01 -4.05161023e-01 -5.58563411e-01
3.71270090e-01 -4.87011641e-01 -4.79316205e-01 -4.16154683e-01
5.17591238e-01 -5.56206286e-01 -7.36589849e-01 4.26916033e-01
7.88675010e-01 2.10538134e-01 8.47473681e-01 -1.16084933e+00
-7.61373699e-01 1.11772752e+00 -6.45519853e-01 3.79443258e-01
4.20683742e-01 1.84755936e-01 9.84307885e-01 -4.07481998e-01
5.62584937e-01 1.66307604e+00 1.17484164e+00 1.13273449e-01
-1.92770612e+00 -9.38037157e-01 1.72367796e-01 -9.16811973e-02
-7.79891014e-01 -7.42815375e-01 6.42830133e-01 -9.24418092e-01
5.55176914e-01 1.99641466e-01 6.52498841e-01 1.77490580e+00
6.16317034e-01 6.31621242e-01 1.31084836e+00 -2.28336275e-01
3.07991177e-01 -6.79573178e-01 1.31986141e-02 3.59500526e-03
3.27070057e-01 8.53361130e-01 -5.42755783e-01 -6.40787363e-01
7.66622961e-01 7.37540424e-02 -2.83284634e-01 2.17201151e-02
-1.64130664e+00 1.22419667e+00 2.89129972e-01 5.41991591e-02
-9.30289567e-01 7.03035772e-01 5.82478642e-01 3.62100184e-01
9.95690286e-01 3.42706650e-01 -5.31581223e-01 -7.37447590e-02
-1.18371677e+00 7.58636475e-01 1.00741446e+00 2.53309786e-01
1.18450820e-01 3.17513108e-01 -9.52681303e-01 3.52502167e-01
7.54222460e-03 6.19991183e-01 3.59069675e-01 -1.44549155e+00
2.01673344e-01 1.26737192e-01 6.12312555e-01 -1.01183188e+00
-5.72830915e-01 -4.42579389e-01 -1.08138204e+00 -8.29326883e-02
6.86014533e-01 -1.02214813e+00 -7.13956654e-01 2.15948272e+00
4.42894101e-01 6.05370641e-01 -2.66727060e-01 4.40423071e-01
3.89609337e-01 6.94820225e-01 2.38538176e-01 -6.46895826e-01
9.43054140e-01 -3.12007934e-01 -8.56021166e-01 -3.51363011e-02
5.54519355e-01 -3.80921841e-01 3.83612096e-01 2.29700282e-02
-9.89124656e-01 -2.40209684e-01 -5.87036312e-01 2.30173677e-01
-8.60177800e-02 -9.42936763e-02 1.15064144e+00 7.51076877e-01
-1.19319677e+00 1.04503703e+00 -6.48219407e-01 -1.13642707e-01
4.90791053e-01 2.93454409e-01 -6.48641139e-02 2.79183567e-01
-1.52392054e+00 6.72569871e-01 -4.04342450e-02 2.01865770e-02
-9.75775898e-01 -1.66836452e+00 -6.77516341e-01 3.10324013e-01
4.63038422e-02 -1.11669159e+00 1.62115705e+00 -1.03501117e+00
-1.35061598e+00 3.16171736e-01 -3.14592123e-01 -8.02096725e-01
6.35391653e-01 4.09079045e-02 -3.90613765e-01 -6.13995552e-01
3.60996753e-01 4.45138514e-01 7.48876274e-01 -8.44629347e-01
-5.94121635e-01 -3.32493216e-01 -1.98649526e-01 -1.70013726e-01
-5.15968874e-02 -9.77671370e-02 7.68513918e-01 -9.24010932e-01
-6.33844495e-01 -9.88808692e-01 -2.31846079e-01 -5.09296060e-01
-5.54689109e-01 -3.84736508e-01 1.67793438e-01 -8.26073289e-01
1.35221314e+00 -1.65513694e+00 -1.22382410e-01 -3.81097160e-02
4.52722162e-01 -1.06234483e-01 6.99147284e-02 9.13000941e-01
-6.36730194e-01 1.86884135e-01 -1.14730619e-01 -2.77008235e-01
2.51601577e-01 -2.49345824e-01 -8.35545301e-01 6.78841233e-01
-8.24988782e-02 1.09870374e+00 -9.28765714e-01 6.42380118e-02
1.43795118e-01 2.73268014e-01 -6.21699214e-01 -1.19210780e-01
-8.32378119e-02 5.66943705e-01 6.20810427e-02 5.52915409e-02
4.76953447e-01 -3.03060114e-01 5.91135979e-01 2.02756912e-01
-4.66103047e-01 6.00463569e-01 -8.10369551e-01 1.16032994e+00
-3.20441484e-01 9.44502294e-01 -2.91102856e-01 -1.22508073e+00
5.40958881e-01 8.52445006e-01 7.12149680e-01 -5.60866475e-01
2.56153028e-02 -2.22271532e-02 3.33713770e-01 -3.25961739e-01
2.11087584e-01 -4.17793006e-01 1.43947387e-02 7.54480839e-01
-2.96377659e-01 4.80737984e-01 1.81986138e-01 1.63095251e-01
1.16619420e+00 8.16351622e-02 4.03451055e-01 -7.15822220e-01
-2.27343768e-01 -1.72963962e-01 7.24448323e-01 1.13754988e+00
3.32516991e-02 9.31816697e-02 9.02842641e-01 -7.39123046e-01
-1.22620010e+00 -9.36518013e-01 -2.22907707e-01 1.08951426e+00
-9.19238508e-01 3.61832120e-02 -5.01463950e-01 -3.49847645e-01
7.06484675e-01 1.13470352e+00 -1.35585618e+00 5.97096272e-02
-5.36980033e-01 -1.21720827e+00 4.23562974e-01 5.87082565e-01
6.40107617e-02 -9.10050690e-01 -3.11957926e-01 4.38674092e-01
-1.97763607e-01 -2.82930702e-01 -4.32522625e-01 -2.90736347e-01
-8.23972762e-01 -8.27621281e-01 -8.23539853e-01 -6.50017336e-02
-6.09584674e-02 8.16007853e-02 1.38913012e+00 -5.69962800e-01
3.03875864e-01 6.91985488e-02 8.03763419e-02 -7.70005167e-01
-5.41835010e-01 -2.81143542e-02 2.33521581e-01 2.26490766e-01
3.19126099e-01 -9.99684572e-01 -9.41243410e-01 -2.43165135e-01
-6.52289450e-01 1.13909185e-01 1.71254620e-01 1.05817544e+00
-3.21147054e-01 -4.56795096e-01 1.26444042e+00 -1.13069475e+00
9.01563644e-01 -1.07850826e+00 -6.50842607e-01 1.13737293e-01
-8.59280348e-01 1.12365022e-01 4.90181655e-01 -5.64674079e-01
-1.39239872e+00 -3.80944669e-01 4.17973787e-01 -1.76417202e-01
1.32769004e-01 9.46874797e-01 5.97511411e-01 6.52749717e-01
8.07307720e-01 -3.75334412e-01 7.97002241e-02 -2.18714654e-01
4.89381969e-01 2.65614778e-01 4.43553239e-01 -4.76745933e-01
3.19722891e-01 7.12741375e-01 6.59143254e-02 -4.72081564e-02
-8.10082078e-01 1.13239907e-01 -2.57136106e-01 -2.14987457e-01
6.96396172e-01 -1.28608906e+00 -1.34876859e+00 3.76606971e-01
-1.08336055e+00 -8.36957693e-01 -2.39042073e-01 7.58725643e-01
-4.06362236e-01 -1.73203394e-01 -5.74123740e-01 -9.96601820e-01
-2.36225262e-01 -5.74086845e-01 1.06553245e+00 8.71227607e-02
-5.88598311e-01 -1.49119437e+00 7.24949539e-01 -1.80414304e-01
3.41061383e-01 8.31724405e-01 7.75375843e-01 -4.05511349e-01
-5.20991273e-02 -2.40880191e-01 -7.49912634e-02 -3.15620482e-01
9.35015455e-02 3.75600696e-01 -9.78049159e-01 -3.17863345e-01
-1.20378576e-01 -5.44562994e-04 1.03973639e+00 1.41475224e+00
9.39288139e-01 -5.89857340e-01 -6.11962736e-01 4.47245568e-01
1.01616335e+00 2.82151043e-01 6.00817800e-01 -1.23774335e-01
5.03416598e-01 7.36240387e-01 1.14668764e-01 8.85253251e-01
7.21150637e-01 4.93946999e-01 -2.41324082e-02 -2.33434349e-01
2.87014633e-01 -5.41663706e-01 4.09068316e-01 3.12466800e-01
-1.52596831e-01 -4.66042429e-01 -8.25549364e-01 7.69811511e-01
-2.18485546e+00 -1.68296373e+00 -5.01625896e-01 2.23816824e+00
9.02531624e-01 -3.11155468e-01 3.82154167e-01 -2.26175919e-01
7.85657048e-01 2.78237939e-01 -6.65523112e-01 -3.54664922e-01
-1.69436991e-01 2.19078064e-01 6.79179966e-01 4.95433241e-01
-1.09599411e+00 2.81679392e-01 7.60788012e+00 5.67053735e-01
-8.50912213e-01 9.17403027e-02 1.16244400e+00 -2.33703732e-01
-7.26809859e-01 2.50806093e-01 -4.93032038e-01 6.50998950e-01
1.57567084e+00 -6.57378972e-01 4.60498512e-01 1.72709122e-01
8.82431030e-01 -6.92838430e-02 -1.59621286e+00 5.31907678e-01
-4.15688068e-01 -1.76634192e+00 -3.13444972e-01 4.12632316e-01
1.45017731e+00 -8.26826915e-02 2.25971147e-01 4.05822545e-01
1.28060782e+00 -1.15041780e+00 5.80897450e-01 8.65892649e-01
7.75579631e-01 -4.81133908e-01 5.37357748e-01 1.30768523e-01
-6.18688285e-01 -5.07794082e-01 -2.00359851e-01 -8.37374210e-01
3.73734325e-01 1.00828874e+00 -9.23610032e-01 4.99806195e-01
3.59042734e-01 9.11353290e-01 -1.21265547e-02 7.37965703e-01
2.54051425e-02 1.21453607e+00 -1.73079073e-01 1.87180877e-01
-9.91761237e-02 -2.08821476e-01 2.84096628e-01 8.17903697e-01
5.46536922e-01 2.42948920e-01 -3.39758515e-01 9.90865886e-01
-2.81505138e-01 -3.01595569e-01 -1.05130959e+00 -5.28807230e-02
8.11055124e-01 7.12468147e-01 -1.66106924e-01 -6.65686190e-01
-4.28478390e-01 2.80842513e-01 1.42388910e-01 5.58396876e-01
-8.41168761e-01 1.75677568e-01 6.55671120e-01 -9.89166498e-02
4.39179912e-02 1.60075784e-01 -6.51865304e-01 -1.27736330e+00
-4.85386789e-01 -9.54486012e-01 6.63959146e-01 -8.51354480e-01
-1.58386099e+00 -3.79879057e-01 2.39412844e-01 -7.17861950e-01
-7.59422243e-01 -1.44389421e-01 -8.56681049e-01 1.05944622e+00
-7.52309978e-01 -7.99447954e-01 2.97876686e-01 3.57977659e-01
2.06059873e-01 1.78105175e-01 7.27070093e-01 7.62150213e-02
-7.69536674e-01 4.38312978e-01 7.62563646e-01 -4.35651205e-02
8.46587360e-01 -1.36521244e+00 8.97700429e-01 8.17983389e-01
-1.61548272e-01 6.97182715e-01 9.20511961e-01 -9.21326220e-01
-1.11817586e+00 -9.12326753e-01 1.14122105e+00 -6.90795243e-01
8.43428433e-01 -3.85863662e-01 -3.67794901e-01 1.23514140e+00
5.76255977e-01 -5.61920702e-01 7.75196671e-01 8.06450307e-01
-1.96529508e-01 -1.10894188e-01 -8.18477929e-01 9.81348813e-01
9.63789523e-01 -3.58135074e-01 -3.58739257e-01 5.71731091e-01
8.54923010e-01 5.32981418e-02 -1.13086271e+00 3.43089879e-01
9.60051239e-01 -1.02242267e+00 8.89001966e-01 -9.89595354e-01
1.04069483e+00 3.39567840e-01 7.83159360e-02 -1.87455034e+00
-6.98378861e-01 -1.10903275e+00 -8.79580807e-03 1.16182327e+00
3.58349830e-01 -1.00645661e+00 5.03569782e-01 9.67170894e-01
2.81815156e-02 -2.93638945e-01 -8.38908970e-01 -2.94568330e-01
5.45568109e-01 1.81297958e-02 9.32895124e-01 1.52226031e+00
3.62709127e-02 3.03913981e-01 -7.03302503e-01 -1.06091365e-01
7.72258103e-01 4.21983093e-01 8.64443064e-01 -1.60623252e+00
-2.05781475e-01 -4.25703913e-01 3.19693625e-01 -5.25937855e-01
5.55141926e-01 -6.99011922e-01 -4.41703439e-01 -1.22677314e+00
3.24002832e-01 -1.11238033e-01 -1.61802217e-01 3.35721165e-01
-5.09990990e-01 7.49270953e-07 1.69575270e-02 1.14155142e-02
3.12758476e-01 6.53134227e-01 1.04719353e+00 -1.14460848e-02
-1.49721831e-01 3.31671312e-02 -8.84023190e-01 3.30658287e-01
7.62077749e-01 -5.11836171e-01 -2.86376446e-01 -2.12010890e-01
4.62437570e-01 9.38102543e-01 6.41544342e-01 -3.63105446e-01
-1.00714482e-01 -6.11383438e-01 7.18042850e-01 -2.65710503e-01
-1.00825414e-01 -1.52985469e-01 8.76555920e-01 6.30739033e-01
-8.89610887e-01 4.34851229e-01 1.67381868e-01 6.99605465e-01
2.35786274e-01 3.08329225e-01 1.24186233e-01 -1.00383442e-02
1.25140980e-01 1.19917788e-01 -7.03768015e-01 -1.45827224e-02
6.71477973e-01 2.27037445e-01 -6.94578171e-01 -9.36582565e-01
-5.62250257e-01 2.97375292e-01 1.17782511e-01 1.52582750e-01
-1.74130142e-01 -1.64805543e+00 -1.28412509e+00 -2.20672742e-01
-4.72395569e-01 -7.52073586e-01 5.46113968e-01 1.22446954e+00
5.29836193e-02 6.87447011e-01 -6.93098828e-02 -8.12756494e-02
-5.75038910e-01 7.19903648e-01 2.72656143e-01 -3.48713845e-01
-2.70886481e-01 4.23619777e-01 7.38468051e-01 -4.20469999e-01
-3.87562871e-01 -1.97643518e-01 9.04097632e-02 3.92272234e-01
5.38043022e-01 8.02530646e-01 -3.17544401e-01 -1.74535185e-01
7.01690316e-02 -3.77859116e-01 2.47389182e-01 -3.81370783e-01
1.56583798e+00 -1.60561636e-01 -2.51903206e-01 9.28112864e-01
9.07300830e-01 4.15273793e-02 -1.59828126e+00 3.15177441e-02
-1.78812712e-01 -3.10062557e-01 6.70734718e-02 -9.09165919e-01
-7.66872525e-01 5.25293052e-01 3.60586882e-01 6.36676192e-01
7.65226364e-01 -5.12523055e-01 2.79109687e-01 -1.74425356e-02
1.47742014e-02 -7.54659414e-01 -5.90834856e-01 8.50616768e-02
9.45435941e-01 -1.14874828e+00 1.65430456e-01 1.86759233e-01
-2.10892156e-01 7.67991006e-01 -3.54479477e-02 -4.38211739e-01
6.84273541e-01 -3.30561120e-03 -1.59265086e-01 -1.96729451e-01
-1.36910498e+00 2.86136657e-01 2.34177992e-01 3.78159970e-01
1.03357995e+00 6.46638095e-01 -6.15639329e-01 2.18261406e-01
-4.10346478e-01 1.52047977e-01 7.89152563e-01 3.65708470e-01
2.56110609e-01 -8.29799771e-01 -5.34709752e-01 1.06907082e+00
-7.09206522e-01 -4.16219175e-01 -1.81729019e-01 9.29455101e-01
-3.15040201e-01 1.14842403e+00 3.99448335e-01 -1.08385898e-01
3.39332246e-03 3.01830739e-01 2.24699289e-01 -1.76221222e-01
-7.26714790e-01 7.20520988e-02 3.97987634e-01 -5.18292189e-01
-6.98045194e-01 -1.29781663e+00 -2.86575109e-01 -1.14976668e+00
-1.30099446e-01 1.09849628e-02 2.32177094e-01 7.77091861e-01
6.14935100e-01 5.67710757e-01 8.55583310e-01 -1.24272752e+00
-8.95819366e-01 -1.43686819e+00 -4.15839136e-01 3.13886046e-01
5.35571933e-01 -7.91517675e-01 -4.11290556e-01 1.47195011e-01] | [8.01328182220459, 5.390271186828613] |
2e087ca9-3ea1-44cf-8586-027eb657ae93 | category-query-learning-for-human-object | 2303.14005 | null | https://arxiv.org/abs/2303.14005v1 | https://arxiv.org/pdf/2303.14005v1.pdf | Category Query Learning for Human-Object Interaction Classification | Unlike most previous HOI methods that focus on learning better human-object features, we propose a novel and complementary approach called category query learning. Such queries are explicitly associated to interaction categories, converted to image specific category representation via a transformer decoder, and learnt via an auxiliary image-level classification task. This idea is motivated by an earlier multi-label image classification method, but is for the first time applied for the challenging human-object interaction classification task. Our method is simple, general and effective. It is validated on three representative HOI baselines and achieves new state-of-the-art results on two benchmarks. | ['Yichen Wei', 'Shuang Liang', 'Yue Hu', 'Fangao Zeng', 'Chi Xie'] | 2023-03-24 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Xie_Category_Query_Learning_for_Human-Object_Interaction_Classification_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Xie_Category_Query_Learning_for_Human-Object_Interaction_Classification_CVPR_2023_paper.pdf | cvpr-2023-1 | ['human-object-interaction-detection', 'multi-label-image-classification'] | ['computer-vision', 'computer-vision'] | [ 8.56310606e-01 2.90139616e-01 -5.61151326e-01 -6.62483513e-01
-9.47627187e-01 -3.16386133e-01 7.68016100e-01 2.75517434e-01
-3.11099768e-01 5.79961121e-01 2.77044773e-01 -4.14618477e-02
-1.16531104e-01 -4.57063705e-01 -7.88210809e-01 -4.76373136e-01
2.97210902e-01 7.51168966e-01 3.63838494e-01 4.70527709e-02
1.94339886e-01 -2.97917239e-02 -1.86545098e+00 6.80389524e-01
4.49398935e-01 1.12556124e+00 -1.70884177e-01 4.91531342e-01
4.13775712e-01 1.15184879e+00 -4.61356103e-01 -2.93688059e-01
-1.79408759e-01 -5.79362333e-01 -1.38645518e+00 2.60957211e-01
7.21556485e-01 -1.17004707e-01 -1.59600779e-01 5.77625751e-01
5.40702879e-01 3.13705057e-01 1.09473062e+00 -1.63064802e+00
-8.59758914e-01 4.51087058e-01 -3.98121953e-01 -2.34950289e-01
6.69372082e-01 2.23436691e-02 1.32463610e+00 -8.82667780e-01
9.10970747e-01 1.62361789e+00 4.79922533e-01 6.72960520e-01
-1.40918088e+00 -4.94141281e-01 7.01708719e-02 4.34956253e-01
-1.30293906e+00 -7.54795074e-02 5.21582901e-01 -6.13861144e-01
9.74561870e-01 8.53540301e-02 2.55957842e-01 1.50948405e+00
8.04936141e-02 1.32630229e+00 1.28816664e+00 -4.91733938e-01
-1.17833272e-01 2.51616895e-01 5.09093523e-01 7.07634747e-01
-1.74543902e-01 8.00492316e-02 -6.10647142e-01 4.83928099e-02
1.89891234e-01 1.31584890e-02 -3.28402787e-01 -9.47974265e-01
-1.57631993e+00 1.10312152e+00 5.82909107e-01 3.76447916e-01
5.64315282e-02 2.04209581e-01 7.20859408e-01 2.19394758e-01
3.48887652e-01 1.70343235e-01 -3.95324051e-01 3.93898748e-02
-4.56649661e-01 4.73739952e-01 8.80511940e-01 1.30458796e+00
8.01447213e-01 -6.39613211e-01 -8.71426702e-01 8.81176949e-01
2.66748250e-01 3.91315013e-01 2.92880416e-01 -9.55793262e-01
1.94751963e-01 6.14676893e-01 -3.12881887e-01 -4.79185820e-01
-5.43139279e-01 -3.70653987e-01 -7.03740060e-01 1.17540374e-01
1.63572997e-01 3.54921430e-01 -8.90803397e-01 1.52369368e+00
1.80605069e-01 6.01458885e-02 8.98459777e-02 7.26788402e-01
1.14013541e+00 2.91463643e-01 4.46948051e-01 1.74180791e-01
1.41282439e+00 -1.63487768e+00 -6.80805624e-01 -1.57292426e-01
7.05262661e-01 -5.13628721e-01 1.40564191e+00 4.24465388e-01
-7.81279862e-01 -8.31794143e-01 -9.09403443e-01 -6.24183953e-01
-7.33088672e-01 1.93755507e-01 8.76879930e-01 4.93065447e-01
-1.03686881e+00 2.19804019e-01 -1.33190751e-01 -7.30678737e-01
6.72895074e-01 4.53200102e-01 -3.20088536e-01 -2.56782889e-01
-1.00524020e+00 8.00600529e-01 6.27116263e-01 -3.09954464e-01
-1.22125888e+00 -6.27036214e-01 -1.07500970e+00 -1.12174839e-01
5.44883430e-01 -7.66130745e-01 1.56830144e+00 -8.59792531e-01
-1.63608658e+00 1.22648263e+00 -1.87382251e-01 -2.93473482e-01
2.83281088e-01 -1.56678304e-01 -1.27696514e-01 9.74629298e-02
1.65981382e-01 1.31128168e+00 8.12142551e-01 -1.68301368e+00
-9.21164811e-01 -2.26477489e-01 2.76297837e-01 2.64206350e-01
-2.25516945e-01 4.79935892e-02 -3.54814172e-01 -4.51255441e-01
-3.38597327e-01 -1.25822699e+00 1.47663966e-01 -4.46119905e-02
-4.68485713e-01 -6.42222583e-01 1.01142669e+00 -4.66364533e-01
9.07652736e-01 -2.22317338e+00 4.99294847e-01 -2.69865133e-02
2.34207168e-01 -1.30558619e-02 -2.67755389e-01 1.79435849e-01
7.04284525e-03 3.61159481e-02 -3.48138452e-01 -4.70927656e-01
2.57620633e-01 1.04655333e-01 -8.83109793e-02 3.65724534e-01
-2.42596641e-02 1.33001554e+00 -1.07241571e+00 -6.29201353e-01
4.36331689e-01 5.45595646e-01 -6.10168695e-01 4.44801092e-01
-2.23972857e-01 6.65392160e-01 -3.13204713e-02 7.16440856e-01
2.09556580e-01 -3.28159273e-01 6.21373877e-02 -5.52055299e-01
1.25509545e-01 1.79737687e-01 -5.11497140e-01 1.86154485e+00
-5.98664641e-01 6.63930953e-01 -6.60437226e-01 -1.27909625e+00
5.04544318e-01 5.41798234e-01 5.06995261e-01 -6.88596547e-01
1.95444837e-01 5.05411811e-02 -1.79184929e-01 -3.27670783e-01
2.63880759e-01 1.71387687e-01 -4.99294281e-01 8.37902501e-02
5.67882657e-01 -2.88061261e-01 1.07403569e-01 4.37019877e-02
9.15253043e-01 4.78756964e-01 6.32628679e-01 -2.10443571e-01
6.22277915e-01 -1.22374982e-01 1.33333161e-01 9.59453821e-01
-4.88370925e-01 7.18405008e-01 3.36103797e-01 -2.57822305e-01
-7.93429494e-01 -9.68907416e-01 -2.69850254e-01 1.63512468e+00
3.80469382e-01 -3.00343841e-01 -6.17835939e-01 -1.25086510e+00
-3.87236550e-02 6.25930667e-01 -8.96383345e-01 -3.05949330e-01
-1.95136726e-01 -1.97118133e-01 1.47498727e-01 6.73315227e-01
6.40508592e-01 -1.51873398e+00 -5.76628685e-01 1.01345569e-01
-3.10200751e-01 -1.34376228e+00 -5.47606230e-01 3.75391036e-01
-4.54068035e-01 -1.11551094e+00 -7.02104509e-01 -1.05796635e+00
2.99136341e-01 3.33137155e-01 1.31636345e+00 -4.64788936e-02
-3.57549965e-01 7.20892310e-01 -6.15307927e-01 -4.90137935e-01
-1.10557899e-01 4.65220511e-01 -4.73992825e-01 5.62620871e-02
4.51252133e-01 -6.15656562e-02 -4.50388461e-01 4.89414334e-01
-7.38901854e-01 1.75442189e-01 7.06034422e-01 1.16695523e+00
5.99024057e-01 -1.86086059e-01 3.46734047e-01 -1.18038630e+00
1.32970363e-01 -3.07450056e-01 -1.04481600e-01 5.43281376e-01
-4.36856836e-01 3.02206874e-02 2.06442192e-01 -5.99164486e-01
-1.15370154e+00 3.61091554e-01 6.87080175e-02 -1.99554071e-01
-6.52617872e-01 2.72358745e-01 -3.82528633e-01 -3.70513022e-01
5.02993703e-01 -1.96851790e-01 -3.83045584e-01 -3.31043392e-01
7.41443694e-01 7.67871499e-01 6.76824450e-01 -3.27828079e-01
5.26797593e-01 4.26584393e-01 2.66016480e-02 -6.36555731e-01
-1.38020408e+00 -9.58814502e-01 -1.03174996e+00 -2.53838599e-01
1.31044316e+00 -8.38166475e-01 -1.07341659e+00 5.56700647e-01
-1.16431403e+00 -7.16663778e-01 -1.30347386e-01 2.38871992e-01
-9.28458571e-01 2.21922308e-01 -5.75165510e-01 -4.52053756e-01
-1.78791866e-01 -1.28872252e+00 1.62481260e+00 -1.46469086e-01
-1.11989662e-01 -8.99492562e-01 -1.15934284e-02 6.96295083e-01
1.58639684e-01 4.28496934e-02 1.02163529e+00 -4.77365464e-01
-5.34402311e-01 -9.57621485e-02 -7.32657373e-01 2.91870058e-01
-9.53305587e-02 -4.97221410e-01 -1.20547485e+00 -4.59503800e-01
-3.33159983e-01 -9.32016313e-01 9.93035257e-01 1.08638540e-01
1.46089375e+00 -5.97847514e-02 -7.00287104e-01 3.76422435e-01
1.44217217e+00 2.21341059e-01 6.80952191e-01 1.02089532e-01
1.19094908e+00 6.72954500e-01 7.55499780e-01 7.28871375e-02
7.92982101e-01 1.32342923e+00 3.03933442e-01 -2.82037675e-01
-5.80684841e-01 -2.71164943e-02 1.46690607e-01 4.46688145e-01
1.13027468e-01 -5.15106320e-01 -7.89564967e-01 5.71020067e-01
-2.23729324e+00 -8.05060804e-01 -6.67422488e-02 1.99250579e+00
6.40466869e-01 1.16564156e-02 4.34848189e-01 1.00655504e-01
5.53891540e-01 6.63240477e-02 -4.33681846e-01 -2.14955553e-01
1.04767650e-01 3.61047477e-01 5.25345623e-01 3.68185759e-01
-1.63903654e+00 1.24986804e+00 6.96644449e+00 6.97212994e-01
-6.06140733e-01 7.14063942e-01 4.49104369e-01 2.33269975e-01
5.82585298e-02 2.87011378e-02 -8.10281694e-01 1.62741262e-02
8.81139576e-01 2.37705186e-01 2.54168421e-01 9.76300180e-01
-4.04341191e-01 -1.62093192e-01 -1.57793248e+00 1.24308836e+00
6.10175192e-01 -8.70776474e-01 1.31699234e-01 2.11861134e-01
6.14747226e-01 -2.14143068e-01 3.17097716e-02 7.71275520e-01
1.57902345e-01 -8.75163317e-01 8.40897381e-01 4.26738113e-01
8.30139697e-01 -5.30323505e-01 8.54969203e-01 -2.90980279e-01
-1.36763310e+00 -2.66981721e-01 -1.23826070e-02 6.34897053e-02
1.02719419e-01 8.64422694e-02 -4.56524640e-01 4.29858863e-01
6.80166364e-01 1.02836764e+00 -9.35837030e-01 9.55035925e-01
-2.78097600e-01 5.87700665e-01 1.48738801e-01 1.10757872e-01
2.28929862e-01 2.54632980e-01 -5.87907471e-02 1.35854959e+00
-1.05314925e-01 -5.35678566e-02 7.49868929e-01 5.12862921e-01
-1.59106523e-01 2.09500894e-01 -9.10140872e-01 1.62138566e-01
7.29970112e-02 1.20105410e+00 -7.08950162e-01 -5.18474162e-01
-8.90537918e-01 1.48709929e+00 4.98704225e-01 3.37056041e-01
-9.25521374e-01 -5.66278577e-01 1.70391560e-01 -4.23071623e-01
2.64265865e-01 7.24787787e-02 -1.47116691e-01 -9.58367467e-01
-2.42923573e-01 -8.87060225e-01 6.29871428e-01 -5.08439958e-01
-1.21474242e+00 3.92464280e-01 3.28004360e-01 -1.05611825e+00
-3.75001788e-01 -9.27960038e-01 -1.84263930e-01 1.55180797e-01
-1.66096926e+00 -1.88617563e+00 -7.17561901e-01 7.31280446e-01
7.75668025e-01 -3.24996375e-02 1.09016955e+00 5.03688097e-01
-4.02356833e-01 9.53965425e-01 -1.98331133e-01 -1.39337197e-01
8.43801320e-01 -1.52847958e+00 8.39237794e-02 2.56857842e-01
1.51603743e-01 1.13337800e-01 4.43921894e-01 -3.49851668e-01
-9.77831721e-01 -1.09165168e+00 1.05322099e+00 -8.99440706e-01
3.94526899e-01 -7.60834455e-01 -6.41712487e-01 9.59119618e-01
4.30452138e-01 5.30837961e-02 8.24774981e-01 2.80159414e-01
-5.45255065e-01 7.02279881e-02 -9.72421408e-01 2.70674527e-01
1.51592803e+00 -8.19787204e-01 -5.54395258e-01 7.98066616e-01
7.60653734e-01 -2.42399797e-01 -9.52759027e-01 6.75429881e-01
6.61592007e-01 -7.14499056e-01 1.10107565e+00 -4.05533820e-01
2.93763161e-01 -3.69006842e-02 -2.70453423e-01 -1.13361573e+00
-4.88246530e-01 -3.94570112e-01 -1.34476766e-01 1.07234752e+00
3.74022007e-01 -2.34059796e-01 5.13717830e-01 8.56056660e-02
-2.70722210e-01 -6.36299849e-01 -5.99696577e-01 -8.50992382e-01
-7.93949068e-02 -2.65685380e-01 2.33629927e-01 6.13637149e-01
9.12437588e-02 9.65541482e-01 -6.63344800e-01 -2.62269676e-01
9.69916046e-01 1.79686204e-01 8.05052280e-01 -1.35973763e+00
-3.61635476e-01 -4.01865393e-01 -6.59351707e-01 -9.40769851e-01
7.27734387e-01 -1.12052476e+00 4.70138311e-01 -1.54172647e+00
7.78551400e-01 -1.03139900e-01 -4.01801229e-01 8.79397511e-01
-1.52974099e-01 8.39805543e-01 2.62341112e-01 1.54561341e-01
-1.23002541e+00 6.34802341e-01 9.59896684e-01 -5.73609889e-01
6.07673004e-02 -6.26871064e-02 -5.02102733e-01 4.44698989e-01
5.07089734e-01 -1.91937625e-01 -5.85120857e-01 -2.24341869e-01
-2.95519203e-01 -2.71156758e-01 6.62476957e-01 -1.31682456e+00
-9.72600952e-02 1.63842693e-01 5.24197519e-03 -6.86800480e-01
3.50268006e-01 -7.88181961e-01 -1.50197461e-01 2.85232127e-01
-7.94266284e-01 -5.89320898e-01 -1.99522749e-01 4.96077657e-01
-2.56447822e-01 -1.39619321e-01 8.85324419e-01 7.66263530e-02
-1.17738807e+00 3.05685282e-01 -3.15505803e-01 -7.01642185e-02
1.10046053e+00 -9.68312994e-02 -2.67156273e-01 -3.54321867e-01
-7.36690223e-01 1.83650479e-01 2.36235648e-01 7.81779826e-01
4.98123646e-01 -1.42418587e+00 -4.00232524e-01 -1.42080396e-01
9.62213337e-01 -4.64253545e-01 -5.87201305e-03 7.93048382e-01
1.41676188e-01 8.85009289e-01 -2.26713762e-01 -7.73638964e-01
-1.30910337e+00 1.04364920e+00 1.68862715e-01 -2.16835275e-01
-6.41023159e-01 6.19842231e-01 6.31544769e-01 -5.98333001e-01
6.43967867e-01 -2.86842376e-01 -3.51276189e-01 -5.59766591e-02
3.50783825e-01 2.04612315e-01 -1.25541821e-01 -1.05983019e+00
-3.33733082e-01 7.60186255e-01 1.53812114e-02 -8.71779099e-02
8.46539259e-01 -2.55455732e-01 -1.09311445e-02 8.39348376e-01
1.73763645e+00 -7.39918768e-01 -9.43706810e-01 -3.18218738e-01
2.11290553e-01 -2.33690754e-01 1.52010284e-02 -1.00928748e+00
-8.05913508e-01 9.37362790e-01 1.09079158e+00 -1.15410566e-01
9.98136401e-01 4.80648011e-01 6.73976839e-01 5.89066386e-01
6.44904554e-01 -8.65952671e-01 5.31368971e-01 5.17923117e-01
9.46640015e-01 -1.73629045e+00 -2.99592018e-01 -8.05262029e-01
-7.82890260e-01 6.13991916e-01 8.47776055e-01 3.20591211e-01
6.44821048e-01 -2.32481286e-01 3.18966992e-02 -4.38154459e-01
-3.80862832e-01 -7.06703842e-01 5.78566968e-01 7.89710760e-01
7.00738847e-01 4.80602160e-02 -4.28463876e-01 1.86760172e-01
3.47161651e-01 2.15953782e-01 -3.31875496e-02 9.86353099e-01
-1.75769359e-01 -1.22247136e+00 2.76761442e-01 4.06883568e-01
-6.69781715e-02 5.64077683e-02 -2.95912862e-01 7.82088220e-01
3.76465678e-01 1.00629032e+00 -1.46461278e-01 -3.94577891e-01
4.86216605e-01 2.61428982e-01 6.73103154e-01 -1.02264094e+00
-6.06707633e-01 -1.53106228e-01 8.57002009e-03 -9.06391084e-01
-8.09039593e-01 -6.12940550e-01 -9.09385026e-01 4.06772465e-01
-3.32976431e-01 -1.26423612e-01 4.87397134e-01 1.00493777e+00
2.24219769e-01 5.68114221e-01 4.48802024e-01 -1.28704000e+00
-1.68447033e-01 -1.04029059e+00 -4.07629848e-01 1.08763683e+00
1.01005793e-01 -1.22636199e+00 -1.60244465e-01 2.21786633e-01] | [10.120262145996094, 1.989902138710022] |
c75e4e54-e707-4196-9e5e-3955a4b85ae3 | h-denseformer-an-efficient-hybrid-densely | 2307.01486 | null | https://arxiv.org/abs/2307.01486v1 | https://arxiv.org/pdf/2307.01486v1.pdf | H-DenseFormer: An Efficient Hybrid Densely Connected Transformer for Multimodal Tumor Segmentation | Recently, deep learning methods have been widely used for tumor segmentation of multimodal medical images with promising results. However, most existing methods are limited by insufficient representational ability, specific modality number and high computational complexity. In this paper, we propose a hybrid densely connected network for tumor segmentation, named H-DenseFormer, which combines the representational power of the Convolutional Neural Network (CNN) and the Transformer structures. Specifically, H-DenseFormer integrates a Transformer-based Multi-path Parallel Embedding (MPE) module that can take an arbitrary number of modalities as input to extract the fusion features from different modalities. Then, the multimodal fusion features are delivered to different levels of the encoder to enhance multimodal learning representation. Besides, we design a lightweight Densely Connected Transformer (DCT) block to replace the standard Transformer block, thus significantly reducing computational complexity. We conduct extensive experiments on two public multimodal datasets, HECKTOR21 and PI-CAI22. The experimental results show that our proposed method outperforms the existing state-of-the-art methods while having lower computational complexity. The source code is available at https://github.com/shijun18/H-DenseFormer. | ['Xudong Xue', 'Hong An', 'Zhaohui Wang', 'Liang Qiao', 'Minfan Zhao', 'Ziqi Zhu', 'Shulan Ruan', 'Hongyu Kan', 'Jun Shi'] | 2023-07-04 | null | null | null | null | ['tumor-segmentation'] | ['computer-vision'] | [ 1.30647197e-01 -1.06081508e-01 -4.28398639e-01 -3.02423537e-01
-1.13206828e+00 -1.84805036e-01 3.03561270e-01 -8.82597193e-02
-4.05549377e-01 4.89310890e-01 4.32492852e-01 -2.09933460e-01
1.80149805e-02 -8.51069152e-01 -3.24474841e-01 -1.02904606e+00
3.23395342e-01 -6.87793270e-02 2.26837561e-01 -1.29959330e-01
-2.95849860e-01 1.38294354e-01 -9.28618491e-01 4.54589546e-01
9.44024742e-01 1.16837394e+00 3.16702843e-01 3.01053852e-01
-9.47717018e-03 9.27721798e-01 1.12949066e-01 -4.13181067e-01
-1.84441283e-01 -4.46705848e-01 -7.00399935e-01 1.03482477e-01
-3.24001722e-02 -5.27826250e-01 -8.17455649e-01 1.09461033e+00
6.92333698e-01 -1.89617500e-01 3.87451977e-01 -1.19565666e+00
-3.06465149e-01 6.65177941e-01 -6.62026286e-01 9.22786817e-03
3.74307111e-02 1.73413023e-01 8.31978440e-01 -9.10411119e-01
5.54342210e-01 9.14316654e-01 5.37702262e-01 4.86869663e-01
-8.08800101e-01 -8.55474770e-01 -5.23389615e-02 4.19896960e-01
-1.52218807e+00 -2.73915261e-01 8.62678409e-01 -1.37405172e-01
5.32423794e-01 3.59835148e-01 6.42500401e-01 1.01942730e+00
2.02290177e-01 1.23985434e+00 9.56423104e-01 4.66340333e-02
-1.94768608e-01 -8.59918967e-02 1.71381123e-02 1.13459122e+00
-1.23207793e-01 -1.36787891e-01 -2.44913906e-01 -3.51312049e-02
7.46621907e-01 5.09612143e-01 -6.15254641e-01 -1.33957177e-01
-1.44628084e+00 9.43346679e-01 1.08068407e+00 5.61572671e-01
-4.10115451e-01 2.02390492e-01 5.62150538e-01 -1.97863672e-02
2.06636310e-01 -3.60408813e-01 -1.53102845e-01 5.20783886e-02
-7.03871012e-01 -2.67647833e-01 3.15337807e-01 7.09084749e-01
5.60910761e-01 -2.63006836e-01 -3.46065819e-01 8.21024895e-01
6.67284310e-01 3.17363143e-01 7.84204066e-01 -5.75947821e-01
5.02442002e-01 8.66318822e-01 -6.43478155e-01 -8.53569865e-01
-5.51074505e-01 -5.46748579e-01 -1.32063949e+00 -4.34080362e-01
2.38245241e-02 -1.92576498e-01 -8.71150196e-01 1.61643088e+00
3.62792820e-01 3.90985250e-01 1.35392487e-01 1.00256598e+00
1.45068431e+00 5.62861621e-01 1.43483967e-01 3.17381807e-02
1.52329183e+00 -1.25166464e+00 -8.21329474e-01 1.80121347e-01
7.70180881e-01 -9.04071331e-01 7.47112989e-01 7.89593533e-02
-1.11625302e+00 -3.44159037e-01 -8.12974811e-01 -3.70400220e-01
-2.62015402e-01 4.92528439e-01 8.07034552e-01 3.44891399e-01
-9.23712969e-01 1.03637725e-01 -1.08368456e+00 -2.21605495e-01
8.51247072e-01 4.94905293e-01 -5.05807042e-01 -4.18109417e-01
-1.14646232e+00 6.05734706e-01 4.10133004e-01 2.18857646e-01
-1.10959637e+00 -5.55833757e-01 -1.08566976e+00 1.40290126e-01
2.29194775e-01 -8.41055453e-01 1.21870232e+00 -7.60605574e-01
-1.43213224e+00 4.41787392e-01 -1.42222807e-01 6.14896752e-02
3.02136332e-01 2.14363158e-01 -2.54156679e-01 5.44673443e-01
-2.25069076e-01 9.84602451e-01 4.31779951e-01 -9.39888477e-01
-5.80949008e-01 -2.87966132e-01 -2.66377348e-02 3.88341188e-01
-7.61484861e-01 -3.16184312e-01 -7.25821912e-01 -6.11243486e-01
1.76793262e-01 -7.24802494e-01 -3.19138288e-01 8.11364949e-02
-6.56970501e-01 -7.64578357e-02 7.50639260e-01 -8.03988159e-01
1.26416457e+00 -2.29896188e+00 4.67245817e-01 2.43299782e-01
4.89036858e-01 4.02615853e-02 -3.18136156e-01 3.13077658e-01
-5.06543107e-02 -1.04817428e-01 -4.20370668e-01 -4.97531652e-01
-8.45354646e-02 2.49908581e-01 3.01320165e-01 5.56539953e-01
-3.50040831e-02 1.22932494e+00 -7.46286571e-01 -8.20778012e-01
3.63453388e-01 8.03321362e-01 -5.09719729e-01 1.51260138e-01
2.55190343e-01 4.91497636e-01 -5.95000684e-01 1.02217603e+00
7.53248990e-01 -4.24368143e-01 7.17113391e-02 -7.11681306e-01
7.48534948e-02 3.08406986e-02 -5.84094703e-01 2.09138036e+00
-3.65216672e-01 3.97923380e-01 8.03636014e-02 -9.92374182e-01
4.71135318e-01 5.59078455e-01 7.70647943e-01 -6.85127854e-01
7.46275306e-01 3.22066694e-01 7.76572153e-02 -6.54131413e-01
1.09587900e-01 -1.75857514e-01 8.92118663e-02 1.67397223e-02
2.00292856e-01 1.96833745e-01 3.13054584e-02 2.45231599e-01
1.12510955e+00 -3.76636624e-01 1.19168237e-01 1.55056864e-01
8.62844586e-01 -1.85625121e-01 5.24176180e-01 1.72556154e-02
-1.32423818e-01 5.66590667e-01 4.87180978e-01 2.83655114e-02
-5.54639041e-01 -9.47822213e-01 -1.00194402e-01 6.57970369e-01
3.32267374e-01 -3.57543379e-01 -6.69675529e-01 -8.66363168e-01
-2.81276584e-01 1.68063253e-01 -7.22572923e-01 -3.08526218e-01
-4.41932231e-01 -9.49885070e-01 6.67547524e-01 6.49852812e-01
8.42053413e-01 -7.16495991e-01 -3.27490807e-01 -1.04917735e-02
-6.23348951e-01 -1.04953432e+00 -5.83529830e-01 -2.21311133e-02
-9.59194779e-01 -1.13157296e+00 -1.06261504e+00 -1.09913397e+00
9.66732204e-01 3.19681257e-01 5.70723951e-01 1.92727029e-01
-2.41734102e-01 2.85985172e-01 -4.05150354e-01 1.23141594e-02
6.49638921e-02 3.27493995e-01 -7.33386874e-01 2.29768619e-01
1.46232188e-01 -2.74902582e-01 -9.05499041e-01 2.36658469e-01
-1.13629043e+00 4.98422563e-01 1.17967284e+00 1.22510266e+00
7.25971282e-01 5.22285551e-02 3.01725179e-01 -5.66688836e-01
4.20685291e-01 -7.00140059e-01 -5.96184433e-02 1.19810335e-01
5.65503445e-03 -3.13857138e-01 4.48939443e-01 -2.84022689e-01
-9.79180038e-01 1.77985117e-01 -5.43176115e-01 -3.74793023e-01
4.73690107e-02 9.51031387e-01 -3.58991086e-01 -2.71823734e-01
-3.33422944e-02 4.37355906e-01 1.48599967e-01 -3.64175051e-01
2.02354819e-01 6.19706988e-01 4.09838110e-01 -1.61319762e-01
6.54832780e-01 4.54032242e-01 4.44617160e-02 -3.82279903e-01
-5.80276251e-01 -2.26954222e-01 -2.24469900e-01 -2.32460156e-01
1.01141918e+00 -1.14554620e+00 -7.61704028e-01 6.02937102e-01
-9.51058447e-01 -1.45480052e-01 -2.99416762e-03 7.80609906e-01
-2.25171253e-01 3.27754170e-01 -1.06359553e+00 -7.57620037e-02
-5.43138504e-01 -1.63971472e+00 1.12608218e+00 6.42138124e-01
2.98918247e-01 -1.05276000e+00 -8.53300318e-02 5.73738039e-01
4.50837463e-01 3.45081985e-01 9.04654503e-01 -3.65585357e-01
-5.69099188e-01 -2.28563100e-01 -4.95966166e-01 2.12443203e-01
2.94542521e-01 -3.21872473e-01 -8.93305719e-01 -3.84469539e-01
-2.84970671e-01 -4.11788374e-01 1.17421710e+00 3.91687989e-01
1.48756683e+00 -1.88498259e-01 -5.94575524e-01 7.58099735e-01
1.51255071e+00 1.01448976e-01 7.36542284e-01 1.57697469e-01
9.37618732e-01 1.71036869e-01 3.10574561e-01 3.00683469e-01
7.34135389e-01 3.29127878e-01 6.58577144e-01 -7.18490481e-01
-2.14293465e-01 -1.56526148e-01 2.34950468e-01 1.31240916e+00
1.06613375e-01 -2.46098429e-01 -8.26360524e-01 5.74030221e-01
-1.92613566e+00 -7.34591842e-01 7.78666660e-02 1.71639502e+00
8.67244840e-01 -1.74057782e-01 -9.38074738e-02 2.16429681e-01
5.26979744e-01 9.92050301e-03 -4.24084961e-01 1.32755965e-01
-1.10427793e-02 1.21362805e-01 2.90207475e-01 3.21734339e-01
-1.07632375e+00 5.44950783e-01 4.87095642e+00 1.19860017e+00
-1.21647680e+00 5.43065429e-01 7.61066198e-01 -8.97033960e-02
-5.29272258e-01 -2.85929769e-01 -3.03483218e-01 5.07649362e-01
5.58497846e-01 5.75314611e-02 7.78002217e-02 3.51358891e-01
-9.74294078e-03 2.59523224e-02 -8.12539101e-01 1.25157392e+00
8.58096704e-02 -1.32289660e+00 1.17815308e-01 1.60270423e-01
5.33443511e-01 5.59220985e-02 3.02376539e-01 2.35929981e-01
-6.00722991e-02 -1.09990656e+00 1.18907496e-01 7.03052342e-01
7.75549233e-01 -9.54493821e-01 1.30094409e+00 1.76349118e-01
-1.43851709e+00 -5.92563264e-02 -1.68334484e-01 5.20238221e-01
1.00729905e-01 5.80620170e-01 -4.09219176e-01 9.16386962e-01
4.36125994e-01 1.08997381e+00 -5.71554124e-01 1.05867743e+00
-8.70417878e-02 5.43070495e-01 -2.33207837e-01 1.09244376e-01
5.79044521e-01 1.25828892e-01 1.57568157e-01 1.13903761e+00
5.19880354e-01 1.59383789e-01 1.69540495e-01 4.82329845e-01
-2.92124391e-01 1.06683619e-01 -3.01677465e-01 -1.59181073e-01
2.62139350e-01 1.77352238e+00 -4.92214173e-01 -3.20801884e-01
-7.36073494e-01 7.82764971e-01 8.22353065e-02 2.16165543e-01
-1.09751904e+00 -4.93820131e-01 2.56521195e-01 -1.77182034e-01
3.26255769e-01 1.21464893e-01 -1.18244708e-01 -1.27352035e+00
-2.28285328e-01 -9.59067941e-01 6.87470496e-01 -6.11337602e-01
-1.22938299e+00 6.77492917e-01 -2.40135416e-01 -1.56029320e+00
4.04706657e-01 -4.17100966e-01 -6.18419230e-01 7.56264210e-01
-1.77756321e+00 -1.67447114e+00 -7.28228033e-01 1.06777549e+00
4.18789715e-01 -2.69774646e-01 7.22242057e-01 7.30716825e-01
-8.69622707e-01 8.05780947e-01 -4.50009406e-02 3.87868375e-01
5.44728756e-01 -8.79906595e-01 -6.16147101e-01 5.41749597e-01
-4.57679242e-01 3.68965805e-01 -2.09996790e-01 -3.59044462e-01
-1.55027115e+00 -1.08093750e+00 6.15793347e-01 2.85982907e-01
5.61201811e-01 -4.63070683e-02 -6.70763910e-01 4.81830955e-01
6.53439939e-01 1.49436817e-01 1.20599234e+00 -4.79560316e-01
3.42875496e-02 -2.89941847e-01 -1.11654997e+00 5.80715597e-01
6.04609191e-01 -3.83281142e-01 -1.73413828e-01 2.73206234e-01
6.20909810e-01 -6.30960345e-01 -1.28914917e+00 7.21367478e-01
4.59786564e-01 -6.29446805e-01 8.28946173e-01 -9.09422934e-02
8.14305007e-01 -3.50144088e-01 -2.59260684e-01 -1.17870259e+00
-3.60640973e-01 -3.00179161e-02 2.77715847e-02 1.07512927e+00
2.94777572e-01 -6.32123828e-01 5.53331256e-01 2.02302694e-01
-4.49691266e-01 -1.42291260e+00 -1.01514411e+00 -1.56178609e-01
8.63835216e-03 -1.97182730e-01 5.38269877e-01 9.31495070e-01
2.06192434e-01 2.73868203e-01 -1.68887660e-01 6.25394583e-02
5.15555263e-01 -7.18617998e-03 3.01684588e-01 -6.25694573e-01
-1.18823498e-01 -5.47477543e-01 -5.02040386e-01 -1.00506103e+00
-1.39334157e-01 -1.34726667e+00 -2.09678173e-01 -1.67955649e+00
6.44947529e-01 -2.71017969e-01 -6.89532459e-01 9.13456202e-01
-1.52580455e-01 4.65298921e-01 1.20906062e-01 1.09094866e-01
-6.57102823e-01 9.23405111e-01 1.83453202e+00 -5.64024210e-01
1.59018010e-01 -4.56139624e-01 -7.56162226e-01 7.37279832e-01
8.60607326e-01 -2.58994639e-01 -4.43708241e-01 -6.17564380e-01
-9.92548391e-02 2.02933058e-01 4.67040360e-01 -9.96055901e-01
4.29134279e-01 7.25493953e-02 7.24571705e-01 -7.75427163e-01
3.29727739e-01 -1.04121804e+00 9.64968354e-02 7.39330232e-01
-3.77737492e-01 -8.95774961e-02 2.56011993e-01 1.43384919e-01
-5.94786465e-01 -4.47745547e-02 7.23591030e-01 -3.09320502e-02
-4.37474877e-01 7.95895100e-01 -2.52242684e-01 -3.64675313e-01
1.04364514e+00 2.39068136e-01 -5.17001808e-01 -5.91784157e-02
-6.63363278e-01 5.33703923e-01 7.59633780e-02 1.62362978e-01
1.06227946e+00 -1.75463963e+00 -6.51326418e-01 9.04678181e-02
5.68312295e-02 8.80143493e-02 6.34154260e-01 1.55223608e+00
-5.49400270e-01 2.13412121e-01 -1.70322075e-01 -6.11931741e-01
-1.35498917e+00 3.07806373e-01 4.34610158e-01 -4.80535179e-01
-5.66996336e-01 8.03908944e-01 3.36720973e-01 -3.82627845e-01
2.77051419e-01 -2.54422277e-01 -3.27514678e-01 6.06956379e-03
4.04342979e-01 1.79987296e-01 -2.12667003e-01 -7.29393661e-01
-4.82955098e-01 4.89230871e-01 -2.69883364e-01 -2.01715250e-03
1.15520847e+00 1.00340322e-02 -3.43624175e-01 -3.40961479e-02
1.62653542e+00 -2.59783775e-01 -9.31504369e-01 -4.23847616e-01
-6.76870286e-01 -2.82782048e-01 4.37891126e-01 -7.59998739e-01
-1.84316242e+00 1.14033580e+00 9.34292316e-01 -3.69051874e-01
1.71586609e+00 5.76696321e-02 1.41647530e+00 8.74830484e-02
-7.93233290e-02 -7.05342889e-01 2.04990536e-01 2.12599978e-01
7.14570403e-01 -1.20794261e+00 -8.56136382e-02 -5.54463327e-01
-6.05980635e-01 1.17397809e+00 6.26397252e-01 1.44982249e-01
8.04496467e-01 3.16419572e-01 8.99282694e-02 -2.70393044e-01
-5.45353889e-01 -3.25353712e-01 3.25462073e-01 1.79323241e-01
4.74388957e-01 1.62058860e-01 -4.30090845e-01 8.61873567e-01
1.02146655e-01 2.29446366e-01 1.14524968e-01 9.03022826e-01
-1.26583904e-01 -1.07055724e+00 -1.15712591e-01 5.37503242e-01
-3.05994958e-01 -1.72676504e-01 -1.68193489e-01 6.53496742e-01
4.65654969e-01 8.35749567e-01 -1.63640931e-01 -7.18166828e-01
3.21573503e-02 -3.05295587e-01 4.80657429e-01 -2.22811252e-01
-5.19998491e-01 5.37952483e-01 -1.45997882e-01 -4.86620486e-01
-6.27208948e-01 -5.63505054e-01 -1.45898128e+00 -1.93531483e-01
-2.37682372e-01 -3.72320451e-02 3.17369789e-01 8.24650407e-01
1.79756388e-01 9.19789433e-01 7.34927177e-01 -6.07769907e-01
-1.14833690e-01 -9.26406205e-01 -2.63925493e-01 1.76413372e-01
3.21241856e-01 -6.47132277e-01 4.90000993e-02 -2.30517592e-02] | [14.564372062683105, -2.414703607559204] |
c5994f6f-2cae-4481-8703-eedf3fd00992 | a-unified-model-for-reverse-dictionary-and | 2205.04602 | null | https://arxiv.org/abs/2205.04602v2 | https://arxiv.org/pdf/2205.04602v2.pdf | A Unified Model for Reverse Dictionary and Definition Modelling | We build a dual-way neural dictionary to retrieve words given definitions, and produce definitions for queried words. The model learns the two tasks simultaneously and handles unknown words via embeddings. It casts a word or a definition to the same representation space through a shared layer, then generates the other form in a multi-task fashion. Our method achieves promising automatic scores on previous benchmarks without extra resources. Human annotators prefer the model's outputs in both reference-less and reference-based evaluation, indicating its practicality. Analysis suggests that multiple objectives benefit learning. | ['Zheng Zhao', 'Pinzhen Chen'] | 2022-05-09 | null | null | null | null | ['definition-modelling', 'reverse-dictionary'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.33459151e-01 -1.70537755e-02 -6.86980128e-01 -5.31394422e-01
-1.44257033e+00 -1.03085470e+00 8.49272609e-01 6.31734580e-02
-9.08328354e-01 8.25089216e-01 4.90208030e-01 -3.53565335e-01
-4.24396209e-02 -8.35881174e-01 -4.40393329e-01 -5.38238525e-01
4.00538325e-01 7.40421534e-01 -2.93219864e-01 -4.33611006e-01
-1.21421441e-02 1.24839559e-01 -1.17457330e+00 1.78497314e-01
5.31208277e-01 8.88947487e-01 2.28988916e-01 4.29685891e-01
-3.13424826e-01 3.89135748e-01 -8.33211184e-01 -9.98219252e-01
4.35331851e-01 5.74278198e-02 -9.72254574e-01 -5.00603676e-01
6.89661145e-01 -1.10732093e-01 -3.77737403e-01 8.83619368e-01
5.92526734e-01 6.77766681e-01 6.25987589e-01 -9.36272919e-01
-1.66922784e+00 7.24626482e-01 -2.49783233e-01 2.51478553e-01
5.72170436e-01 2.09426314e-01 1.77681887e+00 -1.46707475e+00
7.26698160e-01 1.27852118e+00 3.61646891e-01 6.82024837e-01
-1.41205990e+00 -8.96952033e-01 2.40956455e-01 -1.25745520e-01
-1.50421226e+00 -4.74887967e-01 1.58857048e-01 -9.49191228e-02
1.58808124e+00 1.34105310e-01 2.29671240e-01 1.46973264e+00
1.88992754e-01 5.22109151e-01 6.49492383e-01 -2.86306441e-01
8.18152651e-02 4.53841081e-03 4.35988307e-01 5.21226287e-01
3.94118190e-01 5.11662997e-02 -4.67278391e-01 -3.73137414e-01
4.38391119e-01 -2.71822885e-02 -3.15791637e-01 7.79176950e-02
-1.28158331e+00 9.44283605e-01 3.69032145e-01 3.37229162e-01
-3.61403495e-01 4.88698363e-01 3.22286695e-01 6.50899768e-01
6.07313931e-01 1.29621363e+00 -7.36335158e-01 -1.10453509e-01
-6.25197768e-01 3.54124367e-01 7.28854537e-01 1.12491834e+00
8.74371052e-01 -9.35266092e-02 -4.04280275e-01 1.06580174e+00
2.24496409e-01 8.47474337e-01 7.94439375e-01 -6.19289339e-01
3.99692208e-01 4.51245725e-01 9.00724977e-02 -8.02081704e-01
-1.63208842e-01 -5.26833355e-01 -4.75305915e-01 -3.06291521e-01
-1.28608495e-01 -2.98846215e-01 -1.20141852e+00 2.09359574e+00
-9.82033461e-02 1.47663942e-02 1.79744437e-01 1.00747454e+00
1.19903970e+00 6.17280662e-01 2.67401725e-01 2.60808319e-01
1.37933600e+00 -1.06023037e+00 -7.95446515e-01 -7.37685144e-01
6.06495023e-01 -5.95708132e-01 1.33087444e+00 2.11919054e-01
-9.40205872e-01 -5.06089032e-01 -1.01394939e+00 -6.76261246e-01
-7.25205123e-01 1.56637147e-01 5.80528080e-01 2.93730766e-01
-1.20976901e+00 5.89734241e-02 -4.31833774e-01 -5.11189848e-02
9.60484892e-02 2.64691234e-01 -5.31555295e-01 -2.74027437e-01
-1.81755137e+00 1.39994597e+00 2.74308801e-01 -3.33994627e-01
-1.11100411e+00 -7.23016977e-01 -1.19686699e+00 1.08667903e-01
1.40936092e-01 -9.61376011e-01 1.34617066e+00 -6.17042601e-01
-9.24123526e-01 1.07908726e+00 -3.42643619e-01 -3.58209640e-01
-2.56575763e-01 -3.95115495e-01 -6.34491920e-01 -3.23894024e-01
6.26357675e-01 7.85689235e-01 7.54826665e-01 -1.16518819e+00
-4.38361019e-01 -1.02924734e-01 6.30131900e-01 1.86662778e-01
-3.37424219e-01 -2.35655650e-01 -6.16962969e-01 -8.67746770e-01
-2.06509843e-01 -5.61096311e-01 -2.67637074e-01 -1.45667288e-02
-4.62334901e-02 -8.72273088e-01 4.04771529e-02 -3.69314790e-01
1.52703094e+00 -1.92835438e+00 2.53596663e-01 4.39127907e-02
4.20340300e-01 1.15613669e-01 -7.76875615e-01 4.27841663e-01
-3.63370359e-01 3.83772522e-01 -1.95782214e-01 -5.70638537e-01
2.71115601e-01 5.07583678e-01 -7.89277613e-01 2.65027255e-01
2.59576350e-01 1.13031018e+00 -1.27532804e+00 -8.45535025e-02
-9.38339457e-02 3.29651415e-01 -4.13635761e-01 3.46940905e-01
-3.34378630e-01 -6.24743223e-01 -4.24227208e-01 4.74032551e-01
4.70358282e-01 -3.32981646e-01 8.17419291e-02 -1.00646585e-01
2.96661049e-01 7.72783518e-01 -9.21273232e-01 2.19875026e+00
-1.13669586e+00 5.43547869e-01 -5.92197366e-02 -7.00206161e-01
1.14972174e+00 3.83484095e-01 -8.47765431e-02 -7.68234134e-01
-1.81264207e-01 2.67706513e-01 -3.42537850e-01 -4.21310961e-01
8.41925800e-01 -2.29702011e-01 -3.82377148e-01 8.32807302e-01
5.03237307e-01 -3.44328821e-01 2.29937676e-03 3.61803025e-01
1.32147253e+00 2.96483655e-02 3.21779042e-01 -2.62912244e-01
7.58336037e-02 -1.70598581e-01 4.20350313e-01 9.23442066e-01
3.30417693e-01 5.49978673e-01 1.82878941e-01 -7.88817644e-01
-5.93553603e-01 -1.42444003e+00 4.08868631e-03 1.63792372e+00
3.03259969e-01 -6.97656751e-01 4.63332981e-02 -8.46313953e-01
4.04797465e-01 1.15504038e+00 -7.91572630e-01 -5.02531767e-01
-3.20297360e-01 -3.17283064e-01 7.52961814e-01 5.29058814e-01
-1.38687789e-01 -1.07169914e+00 -4.17291373e-01 2.12944329e-01
-8.04756433e-02 -7.25287259e-01 -9.25915778e-01 6.54991031e-01
-1.58551395e-01 -9.02264893e-01 -6.08548462e-01 -1.06431317e+00
4.80039239e-01 3.37939203e-01 1.85283411e+00 1.23737887e-01
-9.06588137e-02 4.02126878e-01 -2.65009761e-01 -1.70422301e-01
2.92793494e-02 4.81873482e-01 2.97267169e-01 -2.90923536e-01
9.40764308e-01 -3.45286399e-01 -3.36448133e-01 1.27704784e-01
-1.08385587e+00 -5.04349947e-01 4.76980567e-01 1.12650847e+00
5.12475610e-01 -5.97607553e-01 9.70823526e-01 -8.74150634e-01
1.31130636e+00 -9.48131204e-01 -3.94750297e-01 4.55883503e-01
-9.61405516e-01 5.22494733e-01 4.66262221e-01 -5.90342402e-01
-7.47506142e-01 -1.75371364e-01 -1.46498725e-01 -6.17914140e-01
2.26836085e-01 7.47665584e-01 -1.82551835e-02 5.93247414e-01
1.07548833e+00 9.89197940e-02 -2.02267677e-01 -4.77460414e-01
9.01009858e-01 6.46991134e-01 5.28110981e-01 -8.80664289e-01
9.72704589e-01 -5.76224998e-02 -8.75022531e-01 -2.60269254e-01
-1.42801166e+00 -5.77319503e-01 -2.71764189e-01 4.44878072e-01
8.27006400e-01 -1.28724587e+00 -9.37607139e-02 -3.04042935e-01
-1.65142894e+00 -1.30003422e-01 -5.97054601e-01 2.14597166e-01
3.50064114e-02 -2.31596798e-01 -9.33118314e-02 -4.86166030e-01
-6.55606031e-01 -1.07119143e+00 1.38708103e+00 6.11328073e-02
-5.87263942e-01 -1.22882271e+00 4.36710298e-01 1.07804433e-01
6.54666185e-01 -3.19753289e-01 9.11220074e-01 -1.13062096e+00
-1.90676227e-01 -1.73551232e-01 -3.06066960e-01 3.51259857e-01
3.52954865e-01 -4.63136047e-01 -8.79257083e-01 -3.38654548e-01
-2.25838169e-01 -7.38588154e-01 1.17943668e+00 -1.94684297e-01
1.11922812e+00 -3.88279170e-01 -2.39729658e-01 6.20904863e-01
1.42201471e+00 -1.45310871e-02 2.68657506e-01 2.09318355e-01
3.57947022e-01 1.33768365e-01 4.15108442e-01 2.41358951e-01
4.56213444e-01 5.17002463e-01 2.77726382e-01 -2.48159409e-01
-2.65518665e-01 -3.14135045e-01 3.16755921e-01 7.35770583e-01
6.27757549e-01 -5.68933964e-01 -1.02862489e+00 1.25119674e+00
-1.50102794e+00 -7.50323653e-01 4.64991421e-01 1.84470606e+00
1.40592849e+00 3.22231427e-02 -4.82161194e-01 -6.84689462e-01
5.40171564e-01 6.28013372e-01 -9.05045271e-01 -6.78348720e-01
-3.78701180e-01 1.19749975e+00 3.85437667e-01 6.99202418e-01
-7.79634953e-01 1.11056745e+00 7.37282944e+00 8.85224164e-01
-9.07155097e-01 4.26360965e-01 4.11356330e-01 -4.14007127e-01
-1.29465890e+00 -3.74023896e-03 -6.81114733e-01 1.64662912e-01
7.24957645e-01 -6.22323692e-01 4.92526740e-01 8.31422210e-01
-5.66139579e-01 4.20417815e-01 -1.29228103e+00 1.22847271e+00
3.95842493e-01 -1.37258005e+00 6.17469549e-01 -1.66085601e-01
7.51081586e-01 4.62270498e-01 2.21650138e-01 7.88379908e-01
7.69873559e-01 -1.47592509e+00 4.01108950e-01 5.65264881e-01
1.34519231e+00 -5.54989100e-01 9.05223727e-01 -4.81805652e-02
-1.05816972e+00 2.06192151e-01 -5.45374393e-01 -3.15222800e-01
4.69332896e-02 5.06816268e-01 -7.81617641e-01 2.24709898e-01
8.07556435e-02 5.70613265e-01 -6.29093528e-01 4.39064950e-01
-7.12552369e-01 4.64857280e-01 -2.04551637e-01 -2.45714888e-01
3.76162708e-01 -1.98828261e-02 3.59732032e-01 1.33945644e+00
1.49256200e-01 7.13154674e-02 2.94938296e-01 1.26607203e+00
-7.30919540e-01 -2.07137503e-02 -9.48178530e-01 -1.69105008e-01
9.86116111e-01 1.38221335e+00 2.67153233e-01 -4.21507448e-01
-3.48007977e-01 8.24136972e-01 7.02125549e-01 5.24777591e-01
-7.35920429e-01 -7.31639922e-01 1.24458265e+00 -2.95059323e-01
1.79785714e-02 -2.76922017e-01 -3.67754906e-01 -1.45469320e+00
7.86942169e-02 -9.21354651e-01 5.62301755e-01 -8.32739294e-01
-1.69986367e+00 7.85956264e-01 -1.69431716e-01 -8.02735090e-01
-4.94469881e-01 -6.48202956e-01 -5.20387888e-01 1.24678516e+00
-1.55425072e+00 -1.06090391e+00 1.31139252e-02 5.16531825e-01
7.59117126e-01 -4.17323768e-01 1.44960904e+00 3.85890841e-01
-3.07482809e-01 1.08494830e+00 -2.29166090e-01 2.92721838e-01
1.15550983e+00 -1.40452683e+00 8.69599640e-01 7.16575384e-01
6.05719328e-01 1.19170141e+00 3.99397731e-01 -3.58817577e-01
-1.48923945e+00 -1.04265547e+00 1.26397598e+00 -8.67030680e-01
7.99258292e-01 -3.64084244e-01 -8.05153847e-01 6.88539326e-01
6.02810681e-01 1.50582463e-01 9.51373816e-01 5.91999948e-01
-1.03726614e+00 2.84359250e-02 -8.57110381e-01 4.90748763e-01
1.16136348e+00 -1.12647653e+00 -1.37855983e+00 3.80880445e-01
1.35016370e+00 -4.87708449e-01 -8.35837185e-01 9.37296450e-03
6.92145109e-01 -6.32262379e-02 8.52021694e-01 -1.11675227e+00
5.41850269e-01 -1.37744382e-01 -5.24192333e-01 -1.63758421e+00
-3.25991780e-01 -5.77872455e-01 -1.63611189e-01 8.70205939e-01
1.06145561e+00 -7.51945376e-01 1.78942904e-01 8.60278964e-01
-6.11786284e-02 -1.02688098e+00 -1.14773047e+00 -5.87060630e-01
4.08393115e-01 -5.76874733e-01 9.10170436e-01 1.04282498e+00
-2.67133325e-01 9.21223998e-01 -1.06507251e-02 -2.72063306e-03
7.55459219e-02 2.23958656e-01 5.58842957e-01 -8.80207956e-01
-1.64319798e-01 -3.88121814e-01 -8.23297799e-02 -1.51372325e+00
8.26071203e-01 -1.38265479e+00 -2.39133500e-02 -1.67141831e+00
-5.08011952e-02 -6.09265089e-01 -7.77085960e-01 1.06683171e+00
-4.79040653e-01 2.95678258e-01 -1.18112929e-01 -6.96518272e-02
-7.85419643e-01 5.84114611e-01 8.75731885e-01 -5.34039259e-01
1.25944272e-01 -4.74711776e-01 -1.27478945e+00 1.12110928e-01
5.05441189e-01 -5.56044281e-01 -5.01655281e-01 -1.53974295e+00
6.03406072e-01 -3.43945205e-01 9.03108045e-02 -4.04233068e-01
2.09559292e-01 -1.96765900e-01 2.04718813e-01 -1.35713574e-02
4.28417146e-01 -5.39733827e-01 -2.46670440e-01 -1.14777340e-02
-9.41684663e-01 6.92252934e-01 6.26034960e-02 3.35880727e-01
-2.60346502e-01 -1.88760594e-01 4.65170473e-01 -4.78319377e-02
-5.88159680e-01 3.65007311e-01 1.04889520e-01 5.91726124e-01
4.56461132e-01 2.33995870e-01 -3.22741359e-01 -4.73394334e-01
-5.54828048e-01 5.46395242e-01 2.80983329e-01 1.01009417e+00
7.52122641e-01 -1.55270493e+00 -9.12682116e-01 2.61496484e-01
5.38075745e-01 8.76053199e-02 -4.72474337e-01 -4.92553674e-02
2.00378951e-02 3.49317819e-01 2.33961776e-01 -2.16487795e-01
-5.97433150e-01 4.61455375e-01 4.53054875e-01 -3.20061505e-01
-1.96184888e-01 1.24390829e+00 1.46601841e-01 -8.10730517e-01
7.94834271e-02 -2.23557562e-01 -7.14371502e-02 3.05899531e-01
7.40003109e-01 6.81699216e-02 1.07512273e-01 -4.70211297e-01
-3.90502304e-01 2.69757926e-01 -2.67656833e-01 -2.29103893e-01
1.30456793e+00 6.68942407e-02 -1.75856620e-01 4.75591838e-01
1.67372441e+00 -2.07786765e-02 -4.37425584e-01 -6.19105816e-01
6.60229474e-02 -4.84508902e-01 2.48388290e-01 -1.15613055e+00
-9.81251717e-01 5.17819703e-01 3.45553279e-01 -3.45880091e-02
8.95260096e-01 1.93067163e-01 1.02725530e+00 8.41585636e-01
3.87613237e-01 -8.79553974e-01 1.10933185e-01 9.29176331e-01
1.23980749e+00 -1.17299461e+00 -7.28059933e-02 3.84151846e-01
-4.82395828e-01 9.63006556e-01 8.37539077e-01 -2.50468791e-01
4.06021833e-01 2.12915316e-01 2.43345797e-01 -4.92167026e-01
-1.45205438e+00 -2.26477876e-01 4.95705843e-01 6.26469254e-01
6.67238712e-01 6.71642274e-02 -3.31741512e-01 9.21058893e-01
-2.62647241e-01 -4.07515407e-01 1.30267933e-01 4.27448660e-01
-1.29214451e-01 -1.29959583e+00 2.58013904e-01 7.85673261e-01
-2.87292242e-01 -6.66496515e-01 -5.13747692e-01 5.33686042e-01
1.62845284e-01 9.30648565e-01 1.05960414e-01 -4.93521243e-01
4.48691398e-01 2.91305155e-01 -4.09983620e-02 -1.29390538e+00
-8.59741092e-01 -5.78591168e-01 2.52340227e-01 -6.34133458e-01
-1.39483884e-02 -9.03583020e-02 -1.14906156e+00 1.18104398e-01
-5.42980552e-01 4.19819146e-01 5.36336780e-01 8.62323225e-01
7.74083853e-01 3.02184999e-01 6.86112642e-01 -2.48389006e-01
-1.00545299e+00 -1.00790620e+00 -2.04671726e-01 6.63843334e-01
5.17968237e-01 -6.77512527e-01 -5.59651136e-01 -3.52257788e-01] | [10.57789421081543, 8.823745727539062] |
dad80328-3124-43ce-98c8-07b195846c6b | rank-aware-negative-training-for-semi | 2306.07621 | null | https://arxiv.org/abs/2306.07621v1 | https://arxiv.org/pdf/2306.07621v1.pdf | Rank-Aware Negative Training for Semi-Supervised Text Classification | Semi-supervised text classification-based paradigms (SSTC) typically employ the spirit of self-training. The key idea is to train a deep classifier on limited labeled texts and then iteratively predict the unlabeled texts as their pseudo-labels for further training. However, the performance is largely affected by the accuracy of pseudo-labels, which may not be significant in real-world scenarios. This paper presents a Rank-aware Negative Training (RNT) framework to address SSTC in learning with noisy label manner. To alleviate the noisy information, we adapt a reasoning with uncertainty-based approach to rank the unlabeled texts based on the evidential support received from the labeled texts. Moreover, we propose the use of negative training to train RNT based on the concept that ``the input instance does not belong to the complementary label''. A complementary label is randomly selected from all labels except the label on-target. Intuitively, the probability of a true label serving as a complementary label is low and thus provides less noisy information during the training, resulting in better performance on the test data. Finally, we evaluate the proposed solution on various text classification benchmark datasets. Our extensive experiments show that it consistently overcomes the state-of-the-art alternatives in most scenarios and achieves competitive performance in the others. The code of RNT is publicly available at:https://github.com/amurtadha/RNT. | ['Yunfeng Liu', 'Wenze Zhang', 'Xinxin Cao', 'Jianlin Su', 'Wen Bo', 'Shengfeng Pan', 'Ahmed Murtadha'] | 2023-06-13 | null | null | null | null | ['text-classification', 'semi-supervised-text-classification-1'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.72269189e-01 1.97017133e-01 -3.99356276e-01 -7.07518995e-01
-1.00506830e+00 -5.31810939e-01 5.78454196e-01 4.17778045e-01
-4.90936399e-01 8.19095910e-01 -7.01425225e-02 -3.28476071e-01
-2.95216832e-02 -6.29265845e-01 -6.52926445e-01 -9.35780823e-01
4.05007184e-01 8.62677932e-01 -1.07037704e-02 -1.28672391e-01
1.14574619e-01 -1.41586050e-01 -1.50429869e+00 5.98725855e-01
9.74924624e-01 1.25378191e+00 -3.37846763e-02 1.41437396e-01
-3.73176962e-01 9.95416462e-01 -5.97872078e-01 -5.58522820e-01
1.71807155e-01 -3.68328631e-01 -8.00011456e-01 2.54563600e-01
6.27122670e-02 -1.65751830e-01 8.99375752e-02 1.17535865e+00
3.10455352e-01 1.72562018e-01 7.78785646e-01 -1.27264106e+00
-2.97082633e-01 1.06602430e+00 -6.35815442e-01 -1.32184312e-01
1.23883188e-01 -1.47923574e-01 1.30128491e+00 -1.06589293e+00
3.79619271e-01 9.72641230e-01 5.79973817e-01 5.60005784e-01
-9.99246716e-01 -7.05643296e-01 4.32554096e-01 2.03159615e-01
-1.26582766e+00 -3.45280856e-01 7.89360642e-01 -1.39994785e-01
3.26301396e-01 1.94318131e-01 1.65637463e-01 1.22805285e+00
-2.57366508e-01 1.25258243e+00 1.29412448e+00 -6.43861055e-01
4.91442233e-01 4.56268877e-01 5.34210563e-01 5.21713138e-01
1.02204315e-01 -1.12785496e-01 -5.64974487e-01 -2.17310518e-01
-1.53104499e-01 3.77149992e-02 -3.11666340e-01 -1.28468484e-01
-8.82391512e-01 7.78112531e-01 3.92907530e-01 3.67460012e-01
-3.30332398e-01 -1.84183612e-01 5.30597031e-01 2.38186091e-01
8.14312637e-01 6.54292107e-02 -7.60672629e-01 1.99095622e-01
-9.16643858e-01 -9.28779468e-02 7.72416413e-01 8.10532451e-01
6.14391804e-01 -2.55387247e-01 -3.10823709e-01 8.99056494e-01
4.75723535e-01 3.24200898e-01 5.95278084e-01 -7.17583120e-01
7.09664643e-01 7.93400586e-01 1.40689954e-01 -5.42891324e-01
-4.41573054e-01 -8.10708523e-01 -8.35372925e-01 -3.57118361e-02
4.48950291e-01 -2.51807809e-01 -8.62734437e-01 1.60904443e+00
4.45866972e-01 1.53986081e-01 3.15945506e-01 9.61353719e-01
8.55078340e-01 6.14745438e-01 6.85301647e-02 -3.99314344e-01
1.12411380e+00 -9.44395900e-01 -7.83698797e-01 -2.81517357e-01
1.08399010e+00 -5.96333325e-01 9.65978920e-01 5.50709248e-01
-5.19172430e-01 -3.44775558e-01 -1.06348622e+00 2.87458718e-01
-3.51125538e-01 4.43505019e-01 2.55374402e-01 5.92160165e-01
-7.58046210e-01 6.15155220e-01 -5.06490648e-01 -1.64488796e-02
4.56124127e-01 2.82484829e-01 -2.51735181e-01 -1.96920723e-01
-1.37309504e+00 6.76838398e-01 7.17031240e-01 1.41925529e-01
-7.82503426e-01 -1.70503139e-01 -6.83121681e-01 6.86033443e-02
6.90807521e-01 -1.29758239e-01 1.38512123e+00 -1.15884674e+00
-1.26057315e+00 8.54415894e-01 -2.61866540e-01 -4.48719472e-01
9.28943396e-01 -1.85702130e-01 -2.36303240e-01 -7.83257484e-02
4.30150688e-01 5.17080367e-01 7.59516716e-01 -1.66023135e+00
-8.33293021e-01 -3.82453591e-01 -8.58500898e-02 3.90569508e-01
-5.61606705e-01 -3.02889526e-01 -3.42891097e-01 -5.41772723e-01
4.07185465e-01 -9.18182254e-01 -1.12590477e-01 -1.95988715e-01
-7.39211142e-01 -6.24840260e-01 8.37059557e-01 -1.00872509e-01
1.03378296e+00 -2.11735487e+00 -2.58670717e-01 2.31647521e-01
6.26819804e-02 3.01447600e-01 1.42112970e-01 2.47749493e-01
-1.45289496e-01 1.44427642e-01 -2.74242550e-01 -6.82235003e-01
1.71073340e-02 1.88110739e-01 -4.31476384e-01 4.70808774e-01
8.72677341e-02 5.82957268e-01 -1.05920637e+00 -6.87673926e-01
9.75635499e-02 1.31400168e-01 4.36867923e-02 1.17338575e-01
-5.10097921e-01 3.71776342e-01 -6.54040337e-01 5.08778870e-01
7.21439004e-01 -5.94884336e-01 3.42697501e-01 6.99363425e-02
1.89839274e-01 4.23702329e-01 -1.34476590e+00 1.19767272e+00
-3.24044377e-01 3.52974713e-01 -6.68135509e-02 -1.26436794e+00
9.12747085e-01 5.12639999e-01 2.48475805e-01 -4.35900241e-01
4.60499108e-01 3.48928392e-01 -2.22858191e-01 -2.96236873e-01
3.48168194e-01 -2.80044436e-01 6.05502166e-02 8.39712620e-01
-2.24500205e-02 1.77494317e-01 2.40168810e-01 3.94735307e-01
8.09955597e-01 1.65735289e-01 2.59273648e-01 -2.02966884e-01
7.97341764e-01 -8.45957175e-03 6.39464200e-01 8.12846005e-01
-2.76817381e-01 4.53210771e-01 5.40256321e-01 -1.23264946e-01
-5.47656655e-01 -7.15316415e-01 -3.08734506e-01 1.45620584e+00
1.22509412e-01 -2.96470195e-01 -6.37574434e-01 -1.30864120e+00
-1.91595227e-01 1.09594345e+00 -8.14933777e-01 -8.66862829e-04
-3.80460471e-02 -7.66645849e-01 2.86165595e-01 3.95045638e-01
4.66528624e-01 -9.58021522e-01 -1.27178356e-01 9.52374041e-02
-4.18808103e-01 -9.49327588e-01 -2.69799054e-01 8.66995156e-01
-7.71531284e-01 -1.03671920e+00 -4.50579315e-01 -8.37055385e-01
8.39661896e-01 4.39261556e-01 9.46861625e-01 2.17695773e-01
5.38527191e-01 2.78317519e-02 -7.67216980e-01 -3.30389351e-01
-6.11164093e-01 7.57720321e-02 5.81628457e-02 2.20191017e-01
4.62089807e-01 -1.38837904e-01 -3.98133606e-01 4.59753215e-01
-7.63984621e-01 6.46291152e-02 2.50671268e-01 1.14075792e+00
6.33112073e-01 4.79726970e-01 9.12646174e-01 -1.42075706e+00
4.33996499e-01 -7.09179461e-01 -3.95761013e-01 3.22692007e-01
-8.19213688e-01 1.67287201e-01 1.07176375e+00 -4.76161569e-01
-1.24713421e+00 1.25033066e-01 -7.32061416e-02 -1.76118910e-01
-1.34715527e-01 6.39736176e-01 -9.35509130e-02 5.59198499e-01
7.96508372e-01 1.66503429e-01 -3.06287408e-01 -2.34980032e-01
2.10475430e-01 9.89515305e-01 1.21407382e-01 -6.51329279e-01
5.74678004e-01 4.97722447e-01 -1.74825653e-01 -2.47719422e-01
-1.66149211e+00 -6.29552782e-01 -4.88971055e-01 -3.86469007e-01
2.86611706e-01 -9.45584834e-01 -4.82892156e-01 2.37443760e-01
-8.97914052e-01 -3.31286430e-01 -1.07672893e-01 4.43395108e-01
-1.54997125e-01 4.24266219e-01 -5.07957637e-01 -1.12488198e+00
-4.84146655e-01 -1.15126824e+00 1.00898564e+00 1.98185518e-01
-1.01487920e-01 -9.58416760e-01 -3.60709131e-01 5.83525538e-01
-6.31024316e-02 -7.63960704e-02 8.03624809e-01 -1.42132270e+00
-1.47809386e-01 -5.39458990e-01 -5.75958900e-02 5.50694048e-01
-6.71453923e-02 -2.08987400e-01 -1.23894382e+00 -2.18684420e-01
1.39816090e-01 -7.78834581e-01 8.88769090e-01 1.60770774e-01
9.36754346e-01 -1.26770973e-01 -3.03877681e-01 -4.15437520e-02
1.18288994e+00 4.00958918e-02 7.38510117e-02 4.12008941e-01
5.29983759e-01 8.86155725e-01 1.05648637e+00 5.76027811e-01
3.13659489e-01 2.94962764e-01 4.97650892e-01 8.81530046e-02
3.25092733e-01 -2.63437152e-01 4.08908725e-01 6.12146616e-01
5.51549554e-01 -7.49621391e-01 -8.91728163e-01 2.99177289e-01
-2.11894345e+00 -5.98072112e-01 -5.31990707e-01 2.14875174e+00
1.10071099e+00 5.24850667e-01 -2.96370029e-01 5.79717517e-01
1.07075727e+00 -1.29931584e-01 -6.08627796e-01 1.26055062e-01
-2.98791409e-01 -1.87434163e-03 2.62869954e-01 3.41824591e-01
-1.37716627e+00 8.77385974e-01 4.83807755e+00 9.62297320e-01
-8.92659605e-01 2.45794460e-01 1.08459532e+00 1.78204924e-02
-2.15653151e-01 -4.82011996e-02 -8.36520612e-01 4.34024811e-01
8.36985171e-01 -2.88950168e-02 5.50737232e-02 9.54295814e-01
2.51149207e-01 -2.48249814e-01 -1.07998109e+00 5.61891079e-01
8.40355009e-02 -8.92235160e-01 -5.48344031e-02 -1.38452545e-01
9.15674865e-01 6.04759865e-02 7.29331523e-02 5.72619736e-01
3.53387564e-01 -6.04157329e-01 9.63063002e-01 4.63373400e-02
6.74675584e-01 -7.58104742e-01 1.19172192e+00 9.06118691e-01
-7.63230085e-01 -1.35106504e-01 -2.54888088e-01 6.02374785e-02
-2.52616704e-01 8.73327434e-01 -9.21615899e-01 5.54414928e-01
5.79525352e-01 7.52679765e-01 -6.01561368e-01 5.46229899e-01
-6.14614964e-01 9.95143056e-01 -3.48672539e-01 -2.32941732e-01
2.93326437e-01 -6.08616546e-02 2.03861147e-01 9.98667955e-01
3.56025286e-02 1.27512828e-01 3.81908745e-01 5.41306615e-01
-4.77931052e-01 5.03440022e-01 -3.57956827e-01 2.73719162e-01
4.90307719e-01 1.32824290e+00 -1.00369287e+00 -6.96475983e-01
-3.28165412e-01 7.03863442e-01 4.81178761e-01 1.53070509e-01
-6.59711421e-01 9.89682134e-03 -3.16226631e-01 -1.92103997e-01
4.15063612e-02 5.51775753e-01 -5.37202835e-01 -1.06389344e+00
2.00516224e-01 -8.17948520e-01 6.36207759e-01 -7.09069848e-01
-1.50421834e+00 6.83186948e-01 -2.93798149e-01 -1.42238414e+00
-2.62215197e-01 -5.87546051e-01 -4.83710289e-01 6.44131541e-01
-1.31476772e+00 -8.05048347e-01 -2.47953296e-01 3.90001833e-01
5.73162973e-01 -8.89547765e-02 7.09818244e-01 1.14809312e-01
-6.81611598e-01 7.59908974e-01 5.15848696e-01 2.37073168e-01
9.10917699e-01 -1.43973577e+00 -3.04110758e-02 7.21109450e-01
2.26236507e-02 4.22454417e-01 7.57155418e-01 -6.42289579e-01
-8.53860438e-01 -1.27421343e+00 8.26095700e-01 -4.64859843e-01
6.14785373e-01 -2.28157312e-01 -1.06669438e+00 5.17247498e-01
2.85156220e-02 4.64565754e-02 8.71647716e-01 1.19161762e-01
-3.85491312e-01 -1.28442226e-02 -1.16046798e+00 4.37911958e-01
6.03848279e-01 -3.46430659e-01 -5.05410016e-01 7.80240834e-01
5.38797319e-01 -1.96553558e-01 -4.29974973e-01 3.99982810e-01
2.08127812e-01 -9.60137069e-01 3.89628798e-01 -3.71720463e-01
4.95271564e-01 -3.91556859e-01 -1.51490554e-01 -1.25124133e+00
8.80234018e-02 -9.35939550e-02 -4.79739271e-02 1.41009796e+00
7.06042051e-01 -6.20396733e-01 9.29114938e-01 4.66342688e-01
-1.19512782e-01 -8.52901459e-01 -9.08298671e-01 -5.56581259e-01
3.19284089e-02 -5.84887683e-01 2.51034021e-01 1.25915611e+00
1.47983521e-01 7.13210344e-01 -1.93905637e-01 1.00369751e-01
6.71122551e-01 1.61836714e-01 4.17963117e-01 -1.48866725e+00
-1.29913047e-01 -2.06005722e-01 1.28462791e-01 -9.74420786e-01
5.91661811e-01 -1.20237505e+00 5.35838246e-01 -1.34903085e+00
3.88613611e-01 -6.25149667e-01 -4.64471966e-01 7.04343498e-01
-4.60451812e-01 1.22015193e-01 -8.37440640e-02 3.51310402e-01
-1.03121448e+00 6.41622841e-01 9.91675436e-01 -2.48898208e-01
-1.35064229e-01 3.32210511e-01 -7.99774587e-01 7.18073666e-01
1.05545998e+00 -8.35022926e-01 -4.51037496e-01 -9.61738601e-02
2.70077139e-01 5.36203347e-02 1.64656888e-03 -6.83383822e-01
3.27551335e-01 -1.03772078e-02 3.94854456e-01 -7.50326216e-01
6.23019896e-02 -9.38930750e-01 -3.07661831e-01 3.73141378e-01
-9.14417982e-01 -3.71992320e-01 -1.33727118e-01 7.66598940e-01
-3.87541726e-02 -7.86777496e-01 8.04523468e-01 6.61092112e-03
-2.77918100e-01 -8.85903463e-02 -2.66841024e-01 1.72158793e-01
8.40158999e-01 1.21966355e-01 -4.31901336e-01 -4.24327731e-01
-5.60942590e-01 6.02386236e-01 3.02552134e-01 1.42210335e-01
3.60644758e-01 -1.10899401e+00 -8.17122161e-01 -5.71246706e-02
4.56060320e-01 1.67078629e-01 1.27706334e-01 6.86820745e-01
-8.64578560e-02 4.03836519e-01 5.83832741e-01 -7.54895091e-01
-1.01657188e+00 5.41317105e-01 2.59296507e-01 -4.58778322e-01
-3.35730046e-01 7.73226917e-01 1.67165741e-01 -6.21660352e-01
5.69895804e-01 7.33191818e-02 -3.79185677e-01 8.83313492e-02
4.40144151e-01 1.81058154e-01 2.86997080e-01 -6.50861204e-01
-2.63814270e-01 2.71281414e-03 -4.35356200e-01 -1.84376717e-01
1.08695996e+00 -1.33159876e-01 2.42009684e-02 7.79576957e-01
1.10422206e+00 -2.06881866e-01 -1.09307492e+00 -7.17968643e-01
2.12750807e-01 -1.81627423e-01 3.14779133e-01 -1.16689622e+00
-9.44420040e-01 7.15976894e-01 5.15983224e-01 2.94361234e-01
9.60105121e-01 -3.46172601e-03 3.94847959e-01 6.19890630e-01
3.00287962e-01 -1.33859932e+00 1.21320724e-01 5.71584642e-01
4.83730078e-01 -1.62110496e+00 -5.23864925e-02 -3.91763359e-01
-1.00225794e+00 9.73257959e-01 7.19545543e-01 3.46463948e-01
5.87889433e-01 8.27406347e-02 4.05298114e-01 -7.25079998e-02
-9.96946156e-01 -2.38106981e-01 4.02855165e-02 1.74776390e-01
7.08741665e-01 1.87159657e-01 -3.11358303e-01 7.51579881e-01
2.53547821e-02 -2.30613291e-01 4.59611237e-01 8.82259905e-01
-4.40261096e-01 -9.95655537e-01 -4.66955036e-01 6.98902011e-01
-5.08151889e-01 -1.66266441e-01 -4.82852697e-01 4.59025741e-01
1.43613517e-01 1.41477191e+00 -3.33750337e-01 -4.06071365e-01
1.46044508e-01 4.08958405e-01 -7.96333700e-02 -8.22232783e-01
-6.39289558e-01 1.76050588e-01 7.11145103e-02 1.44619226e-01
-5.06179154e-01 -6.62363946e-01 -1.60253131e+00 -1.21039845e-01
-8.88332605e-01 5.84251761e-01 4.54994768e-01 1.13340127e+00
9.15969610e-02 3.20925862e-01 9.41792965e-01 -4.61271077e-01
-8.52241874e-01 -1.19913900e+00 -5.40943861e-01 4.78731245e-01
1.77705780e-01 -8.07648480e-01 -7.59696126e-01 -1.13532640e-01] | [9.50970458984375, 3.9967451095581055] |
fdbbceed-c111-457c-b605-67ae442fa9e1 | handwritten-digit-recognition-using-machine | 2106.12614 | null | https://arxiv.org/abs/2106.12614v1 | https://arxiv.org/pdf/2106.12614v1.pdf | Handwritten Digit Recognition using Machine and Deep Learning Algorithms | The reliance of humans over machines has never been so high such that from object classification in photographs to adding sound to silent movies everything can be performed with the help of deep learning and machine learning algorithms. Likewise, Handwritten text recognition is one of the significant areas of research and development with a streaming number of possibilities that could be attained. Handwriting recognition (HWR), also known as Handwritten Text Recognition (HTR), is the ability of a computer to receive and interpret intelligible handwritten input from sources such as paper documents, photographs, touch-screens and other devices [1]. Apparently, in this paper, we have performed handwritten digit recognition with the help of MNIST datasets using Support Vector Machines (SVM), Multi-Layer Perceptron (MLP) and Convolution Neural Network (CNN) models. Our main objective is to compare the accuracy of the models stated above along with their execution time to get the best possible model for digit recognition. | ['Rishika Kushwah', 'Ritik Dixit', 'Samay Pashine'] | 2021-06-23 | null | null | null | null | ['handwriting-recognition', 'handwritten-digit-recognition'] | ['computer-vision', 'computer-vision'] | [ 4.14936870e-01 -6.84541687e-02 6.27031550e-02 -3.54871780e-01
9.35405940e-02 -6.83215499e-01 8.76365125e-01 -1.21995896e-01
-3.89936954e-01 6.36280417e-01 -1.00142375e-01 -6.93123043e-01
-6.89495057e-02 -5.39058805e-01 -3.57853383e-01 -3.98815751e-01
3.50578517e-01 3.26101810e-01 1.05350576e-01 -1.20824419e-01
5.69727004e-01 1.32709205e+00 -1.41221356e+00 6.29565775e-01
2.02347010e-01 1.15576804e+00 1.39634132e-01 1.17887568e+00
-1.53246328e-01 1.47656882e+00 -7.71923542e-01 -4.87066031e-01
2.00037882e-02 -2.18967736e-01 -8.89337540e-01 3.72203499e-01
1.41725212e-01 -5.64106584e-01 -7.18055010e-01 7.06370652e-01
3.41653377e-01 4.99498621e-02 6.72914684e-01 -6.56805873e-01
-1.13351727e+00 2.98093110e-01 -2.16093332e-01 1.11713253e-01
3.83080542e-01 1.51278809e-01 2.75340438e-01 -8.45837951e-01
4.20923233e-01 1.13132131e+00 5.12458146e-01 4.01733994e-01
-8.60215783e-01 -6.95410594e-02 -4.64645863e-01 2.03730732e-01
-9.61306512e-01 -2.60771334e-01 7.23337710e-01 -4.71444666e-01
1.35550368e+00 4.95464772e-01 2.47421771e-01 1.12805831e+00
2.72361100e-01 7.35475838e-01 1.12870622e+00 -9.09959376e-01
2.78213948e-01 4.70778048e-01 2.30997384e-01 6.46937311e-01
4.42238636e-02 -5.70396259e-02 -2.95786262e-01 3.34283412e-01
9.81910646e-01 1.00719169e-01 -1.79971427e-01 5.89896083e-01
-1.21601880e+00 4.72748458e-01 2.21706718e-01 7.31814206e-01
-3.97986323e-01 -6.16462603e-02 3.32861871e-01 4.56893444e-01
-7.38743842e-02 4.65278387e-01 -2.02495575e-01 -3.01607102e-01
-1.00591230e+00 -2.88476914e-01 1.00653207e+00 6.56471372e-01
1.76395893e-01 4.46604967e-01 2.29270961e-02 8.50041330e-01
1.78471565e-01 5.50099671e-01 9.81787086e-01 -5.77079475e-01
4.34167862e-01 6.15664065e-01 -4.33389917e-02 -1.18820286e+00
-8.94597694e-02 -1.53040541e-02 -1.14742374e+00 8.18014920e-01
5.95186055e-01 -1.24734066e-01 -1.04387081e+00 5.39607286e-01
-5.00930011e-01 -4.85240787e-01 2.96646118e-01 8.30095172e-01
7.86858559e-01 9.58591640e-01 -3.02917123e-01 1.98063791e-01
1.14680445e+00 -9.66240346e-01 -8.05087447e-01 -3.65835696e-01
2.15262413e-01 -9.22729313e-01 8.24400783e-01 9.15725946e-01
-8.86403501e-01 -7.93752670e-01 -1.34487128e+00 -3.10657501e-01
-6.57076716e-01 9.17266607e-01 4.80684578e-01 7.18184114e-01
-7.61538625e-01 1.01429939e+00 -9.13746119e-01 -3.42199355e-01
2.81424820e-01 4.69807774e-01 -5.62376499e-01 5.99093474e-02
-4.65694726e-01 1.30951464e+00 3.64958942e-01 4.64282721e-01
-4.04065847e-01 1.31404877e-01 -3.33397388e-01 1.06433049e-01
-7.48296976e-02 1.58865899e-01 1.05347645e+00 -1.51982260e+00
-1.70216751e+00 9.09574389e-01 5.03376834e-02 -5.48099756e-01
7.18065441e-01 1.54538581e-03 -6.79927588e-01 2.44117558e-01
-9.15817440e-01 3.90834123e-01 9.18796301e-01 -8.02415431e-01
-2.96622127e-01 -5.56873083e-01 -2.53306001e-01 -2.64383554e-01
-2.17088461e-01 2.53188133e-01 1.40232101e-01 -5.06892741e-01
1.89516231e-01 -5.16621828e-01 2.91828811e-01 2.26482749e-01
-4.66328263e-01 -1.74264297e-01 1.22927642e+00 -1.25275576e+00
8.61384630e-01 -2.27840662e+00 -1.89762481e-03 -2.95393285e-03
-2.52214279e-02 6.48009837e-01 1.07985049e-01 6.12435222e-01
2.47029215e-02 -1.72740221e-02 5.53589761e-02 -1.42953396e-01
5.35972323e-03 4.05191749e-01 -6.22678876e-01 3.03524166e-01
5.08681536e-01 6.62701428e-01 -2.42456391e-01 -2.16836095e-01
6.56083286e-01 6.03745997e-01 3.14118087e-01 -1.72997080e-02
-7.14340946e-03 -3.96370776e-02 -3.63781959e-01 8.45650256e-01
2.82702118e-01 -2.31991887e-01 1.52644500e-01 -2.25088932e-02
-3.20325911e-01 -2.22655982e-01 -9.66693282e-01 9.70333934e-01
-4.21243995e-01 1.47881997e+00 -4.25528586e-01 -1.00195241e+00
1.39890540e+00 5.92609167e-01 -3.80254179e-01 -6.00250900e-01
2.78193742e-01 5.24630487e-01 -7.14096949e-02 -8.64063203e-01
4.30557400e-01 2.17484199e-02 4.82182652e-01 3.37309569e-01
8.36025551e-03 8.81357640e-02 -1.99438110e-01 -3.68947655e-01
9.15988028e-01 -1.62323117e-01 2.31843352e-01 1.94876626e-01
5.81736624e-01 -2.71767229e-02 -4.37301308e-01 7.48626232e-01
9.40802507e-03 7.16243088e-01 2.77614772e-01 -9.47619021e-01
-1.33482778e+00 -6.66087627e-01 2.92662401e-02 4.74365950e-01
-4.10172999e-01 6.20485485e-01 -5.49727917e-01 -1.03342034e-01
-1.47567555e-01 7.62003958e-01 -4.01370913e-01 2.05576494e-01
-5.18841267e-01 -3.67526829e-01 8.57797027e-01 6.68594897e-01
8.68415236e-01 -1.55854154e+00 -8.80633056e-01 2.07962900e-01
4.34417099e-01 -1.02995861e+00 2.76757628e-01 5.24004579e-01
-1.04116237e+00 -8.15979481e-01 -9.02757525e-01 -9.13182795e-01
6.56157911e-01 -1.52600333e-01 4.64032441e-01 -9.04138293e-03
-6.11676753e-01 2.59869784e-01 -4.56438065e-01 -3.20191085e-01
-7.77776182e-01 -1.46478832e-01 -2.71774884e-02 5.76948114e-02
4.77516562e-01 -3.60991448e-01 -7.96771497e-02 -1.01262610e-02
-1.10652792e+00 6.82524964e-02 1.05999494e+00 6.95862293e-01
5.36000282e-02 1.18084945e-01 3.61653805e-01 -4.80216235e-01
7.61915624e-01 4.90082763e-02 -4.78277981e-01 5.82015812e-01
-2.19833955e-01 -2.27336492e-02 9.19108212e-01 -5.90120494e-01
-8.76917779e-01 2.28268191e-01 -1.45257473e-01 -2.85466224e-01
-7.36426234e-01 3.75269741e-01 3.33419591e-02 -3.10574323e-01
9.31983531e-01 7.43263900e-01 2.33416092e-02 -5.80379844e-01
-1.78822912e-02 1.46995413e+00 8.80385041e-01 1.15146562e-02
2.44814590e-01 2.91440845e-01 -7.05294907e-02 -1.24770570e+00
-2.37945721e-01 3.86179127e-02 -8.70252192e-01 -4.93182540e-01
8.01328301e-01 -2.00021952e-01 -8.09042096e-01 1.08659446e+00
-1.41969240e+00 -1.76670760e-01 3.13573442e-02 5.29817879e-01
-2.52230316e-01 4.32431668e-01 -8.74601781e-01 -1.31693816e+00
-1.82632089e-01 -9.29750502e-01 7.00203896e-01 4.33148503e-01
-2.76681155e-01 -1.12436104e+00 -3.33093405e-01 4.20396388e-01
4.92222548e-01 2.34997123e-01 9.41932499e-01 -9.19499815e-01
-5.12679100e-01 -9.15929258e-01 -3.26002866e-01 9.39089119e-01
3.27313870e-01 3.90990049e-01 -1.25881350e+00 3.83079201e-02
5.63977957e-02 -4.90928560e-01 7.16265380e-01 9.11020339e-02
1.37580490e+00 -5.40458381e-01 1.60987213e-01 1.88322976e-01
1.49299908e+00 8.05921495e-01 1.09474123e+00 4.00422186e-01
2.62276083e-01 2.85035640e-01 -2.55027980e-01 4.39767778e-01
-4.25923705e-01 3.11909407e-01 6.32797629e-02 1.47201046e-02
-7.32238144e-02 -5.67959696e-02 1.98940098e-01 5.94758093e-01
-5.25842309e-01 -4.32294965e-01 -1.11166954e+00 3.07028703e-02
-1.27651870e+00 -1.06415236e+00 -8.63219202e-02 1.94065630e+00
4.82091904e-01 2.54697293e-01 -2.77641296e-01 9.06057179e-01
7.07782149e-01 -1.68577418e-01 -3.76306653e-01 -9.69349921e-01
-3.73855442e-01 2.11330965e-01 4.21123415e-01 2.80533552e-01
-9.88138378e-01 4.58015203e-01 6.55395317e+00 6.07646704e-01
-1.76411104e+00 -6.02145314e-01 7.83115625e-01 3.38471860e-01
3.49453181e-01 -6.12692177e-01 -5.10597408e-01 4.79389280e-01
8.70705366e-01 1.68481335e-01 7.97118306e-01 8.88963759e-01
1.16150178e-01 -2.34350547e-01 -1.20471179e+00 1.18423545e+00
2.49697313e-01 -1.38977766e+00 2.13162661e-01 -1.43528774e-01
4.41646665e-01 -3.97272915e-01 2.04230472e-01 3.03729735e-02
-2.33803689e-01 -1.47546041e+00 6.56058311e-01 7.43362188e-01
5.33395231e-01 -3.95691156e-01 8.80957842e-01 5.72425365e-01
-4.02550578e-01 -3.26650620e-01 -2.68606931e-01 -4.07721758e-01
-2.79074878e-01 3.10553938e-01 -7.14450061e-01 1.52787820e-01
2.34728158e-01 3.22545677e-01 -4.64773923e-01 8.83504450e-01
1.18115842e-02 6.02029204e-01 -1.75329447e-01 -6.99119747e-01
4.50432152e-01 1.36563048e-01 4.95116785e-02 1.36155415e+00
3.30876708e-01 2.79980153e-01 -3.98818016e-01 7.64093935e-01
-1.17100915e-02 -1.64889053e-01 -4.93543416e-01 -7.64732301e-01
1.27246380e-01 1.04577303e+00 -9.73396182e-01 -3.34931672e-01
-3.99764001e-01 1.41621840e+00 -1.13270797e-01 2.37727001e-01
-3.16065460e-01 -9.49856579e-01 -1.00015320e-01 3.73848453e-02
2.06379265e-01 -4.06303138e-01 -6.00323200e-01 -8.60288441e-01
2.44849339e-01 -8.50848556e-01 -2.29028180e-01 -1.16731095e+00
-1.04632401e+00 7.82730520e-01 -6.32519543e-01 -7.98663020e-01
-2.12158710e-01 -1.44243670e+00 -5.55227220e-01 8.93582880e-01
-1.01490033e+00 -8.89595985e-01 -2.96906769e-01 2.52267212e-01
7.36555517e-01 -6.83832884e-01 9.43509698e-01 1.95435673e-01
-3.72131437e-01 2.96011239e-01 3.79359484e-01 6.19877338e-01
2.59676930e-02 -1.02529919e+00 2.36687079e-01 5.66334426e-01
2.24858642e-01 7.88562894e-02 5.72032332e-01 -3.60633731e-01
-1.50679946e+00 -6.94626510e-01 1.09557903e+00 -2.79318094e-01
4.46293771e-01 -5.98295368e-02 -9.40139949e-01 7.25592971e-01
1.56801835e-01 -2.18819141e-01 3.88441354e-01 -5.47273397e-01
-3.01792085e-01 -7.88533781e-03 -1.16946065e+00 3.04743409e-01
1.06223643e-01 -6.15289569e-01 -7.69166052e-01 3.28577280e-01
1.42792761e-02 -3.13102692e-01 -4.51071233e-01 -2.56862491e-01
7.93459296e-01 -1.04448581e+00 7.34675109e-01 -7.45413303e-01
1.01853800e+00 1.08148053e-01 -2.95236707e-01 -8.80496979e-01
-4.37577255e-02 -1.78435847e-01 -4.15609702e-02 9.53930378e-01
4.94484454e-01 -6.23849332e-01 6.25817239e-01 8.78935039e-01
-2.13424787e-02 -5.28532743e-01 -6.71431005e-01 -8.71361971e-01
-1.30929857e-01 -3.35752606e-01 2.54778713e-01 1.00969696e+00
1.34939656e-01 -9.49004367e-02 -5.73179781e-01 6.77274242e-02
2.80413181e-01 -1.28761709e-01 2.32429057e-01 -1.18373227e+00
-4.65636730e-01 -5.39316237e-01 -9.46915686e-01 -8.84506285e-01
-2.25744799e-01 -6.89566731e-01 -1.99467286e-01 -1.63337839e+00
-2.36677393e-01 1.03490457e-01 1.05896212e-01 6.20659709e-01
6.72645926e-01 8.49935263e-02 4.05044168e-01 2.62799084e-01
-6.55644685e-02 -3.37340347e-02 1.16803956e+00 -2.95374364e-01
7.79115856e-02 1.49738461e-01 -1.11838490e-01 7.32768893e-01
7.06206501e-01 -8.93175229e-03 8.11521187e-02 -7.50146568e-01
9.04080551e-03 3.65946054e-01 6.74522817e-01 -1.16508090e+00
5.96455097e-01 1.51634857e-01 1.12105358e+00 -4.19121921e-01
3.23273003e-01 -8.62979531e-01 2.08057418e-01 5.10314405e-01
-6.48914456e-01 -9.69949961e-02 2.09443703e-01 1.24969974e-01
-4.05724555e-01 -6.37341022e-01 7.57685184e-01 -3.42188835e-01
-7.07029521e-01 -3.97707373e-01 -9.94997382e-01 -7.02270031e-01
9.09407854e-01 -6.76060319e-01 -3.52239937e-01 -3.42347354e-01
-8.46954644e-01 -3.59122962e-01 -2.36766152e-02 3.44331950e-01
9.35524344e-01 -9.25393581e-01 -4.55712885e-01 3.10011059e-01
-4.20664310e-01 -3.96356553e-01 1.25791684e-01 2.54033536e-01
-1.27039981e+00 9.94438291e-01 -6.32035911e-01 -8.64543542e-02
-1.32770777e+00 3.83776158e-01 5.67282736e-01 1.45309567e-01
-6.60886586e-01 4.46622729e-01 -8.18486750e-01 -4.41965349e-02
6.31843507e-01 -4.44783837e-01 -1.97310254e-01 -4.04596537e-01
9.70021546e-01 5.71221113e-01 1.66486457e-01 -1.61669880e-01
-1.36992872e-01 3.46155524e-01 -3.95932160e-02 -1.41615912e-01
1.37224185e+00 5.05422175e-01 -5.81627488e-02 6.24720097e-01
1.03107798e+00 -5.29031277e-01 -8.23394001e-01 2.75231183e-01
7.56268203e-02 -4.40580279e-01 1.85492679e-01 -1.30164719e+00
-8.35565865e-01 1.34214342e+00 6.23674572e-01 4.87191856e-01
1.04436719e+00 -3.59578907e-01 3.84269953e-01 1.27224255e+00
1.21141210e-01 -1.42138004e+00 1.19679496e-01 3.27434331e-01
1.09920526e+00 -1.27077854e+00 -2.13354871e-01 1.56927228e-01
-6.22503400e-01 2.14219642e+00 1.69859335e-01 -2.12529182e-01
5.55078506e-01 5.43507278e-01 4.18277532e-02 1.15304887e-01
-6.07721627e-01 3.49991858e-01 1.73247457e-01 4.82083678e-01
4.93144810e-01 -6.85450956e-02 6.62127733e-02 5.47895372e-01
1.03860511e-03 5.91027975e-01 6.74418807e-01 1.01153910e+00
-7.21583903e-01 -6.63295031e-01 -4.77811217e-01 5.68392992e-01
-4.59042072e-01 1.83868840e-01 -7.85113037e-01 5.55989504e-01
-2.16892540e-01 8.13371420e-01 1.63547829e-01 -3.27705353e-01
2.15033293e-01 2.67540812e-01 6.44796193e-01 -9.31081548e-02
-5.63946009e-01 -5.64206302e-01 -1.20540045e-01 1.35649487e-01
-1.90948054e-01 -3.33265483e-01 -1.09337687e+00 -3.95168751e-01
-7.40143284e-02 -4.03802961e-01 1.33545077e+00 1.16528058e+00
-1.03527971e-01 4.03074205e-01 4.50694323e-01 -8.02174568e-01
-8.81349206e-01 -1.00007463e+00 -8.71955156e-01 -2.86641140e-02
2.80731797e-01 5.62205389e-02 -4.14249510e-01 5.49000800e-01] | [11.831443786621094, 2.648480176925659] |
afb78020-764e-4338-adf4-c0a15cfd9603 | a-new-neuromorphic-computing-approach-for | 2102.12773 | null | https://arxiv.org/abs/2102.12773v1 | https://arxiv.org/pdf/2102.12773v1.pdf | A New Neuromorphic Computing Approach for Epileptic Seizure Prediction | Several high specificity and sensitivity seizure prediction methods with convolutional neural networks (CNNs) are reported. However, CNNs are computationally expensive and power hungry. These inconveniences make CNN-based methods hard to be implemented on wearable devices. Motivated by the energy-efficient spiking neural networks (SNNs), a neuromorphic computing approach for seizure prediction is proposed in this work. This approach uses a designed gaussian random discrete encoder to generate spike sequences from the EEG samples and make predictions in a spiking convolutional neural network (Spiking-CNN) which combines the advantages of CNNs and SNNs. The experimental results show that the sensitivity, specificity and AUC can remain 95.1%, 99.2% and 0.912 respectively while the computation complexity is reduced by 98.58% compared to CNN, indicating that the proposed Spiking-CNN is hardware friendly and of high precision. | ['Mohamad Sawan', 'Shiqi Zhao', 'Jie Yang', 'Fengshi Tian'] | 2021-02-25 | null | null | null | null | ['seizure-prediction'] | ['medical'] | [ 3.35069597e-01 -2.18299851e-01 2.63382137e-01 -1.48351148e-01
-5.28989658e-02 -2.11173460e-01 9.75782946e-02 1.08557511e-02
-4.02880698e-01 1.21998370e+00 -2.79354364e-01 1.31943524e-01
5.18999137e-02 -7.40209162e-01 -5.05765915e-01 -7.20966101e-01
-1.85003936e-01 -3.81804705e-01 4.14347291e-01 1.08027428e-01
1.90630436e-01 5.61025500e-01 -1.57815564e+00 1.87117189e-01
8.99542928e-01 1.44572115e+00 2.59098768e-01 3.38859409e-01
4.41621244e-01 5.76400161e-01 -6.22431815e-01 -1.51041999e-01
1.02153262e-02 -5.43849409e-01 1.09151535e-01 -4.89002109e-01
-5.90577006e-01 -1.42352805e-02 -5.30275643e-01 1.16399038e+00
8.03526878e-01 -2.70829171e-01 5.45744598e-01 -1.32934618e+00
-3.90866786e-01 4.67411697e-01 -2.07923740e-01 2.50823200e-01
-6.98560998e-02 2.09934101e-01 -4.69259694e-02 -5.96652210e-01
1.89017743e-01 3.85644674e-01 1.21685994e+00 7.71822989e-01
-9.18591917e-01 -1.22811580e+00 -6.55390918e-01 -1.13757640e-01
-1.66234839e+00 -3.37106228e-01 5.34483731e-01 -2.18555987e-01
1.40246212e+00 1.94448754e-01 1.24116921e+00 1.16091323e+00
1.03274250e+00 1.70961797e-01 1.28187811e+00 -9.55164954e-02
6.59779787e-01 -5.00050224e-02 -4.16182950e-02 5.09307921e-01
9.81228292e-01 2.31224000e-01 -7.18373120e-01 -1.14853889e-01
9.20509636e-01 4.19222832e-01 -5.39816804e-02 5.44638813e-01
-9.67006087e-01 4.06210035e-01 6.01516664e-01 5.36948323e-01
-6.51557088e-01 5.47757089e-01 5.22032678e-01 -2.50052840e-01
4.73123677e-02 2.78667599e-01 -1.63649037e-01 -3.97346109e-01
-1.16804767e+00 3.05585504e-01 5.91030478e-01 9.25806284e-01
2.31249854e-01 5.80980361e-01 -1.77796930e-01 4.78526235e-01
-2.03407882e-03 5.68021953e-01 8.32226992e-01 -3.31897497e-01
-6.00739121e-02 8.12357306e-01 -1.87791452e-01 -8.21091950e-01
-7.77623892e-01 -5.00501812e-01 -1.37673366e+00 2.08291858e-02
-1.26566678e-01 -4.31497276e-01 -9.33785558e-01 1.43513739e+00
-4.37887698e-01 5.34767032e-01 3.25892925e-01 7.34988332e-01
9.44860816e-01 4.77155715e-01 3.74932498e-01 -7.96188191e-02
1.35792553e+00 -4.67351139e-01 -8.25511396e-01 -1.80597067e-01
2.46753201e-01 -3.89340788e-01 4.85866368e-01 2.91802049e-01
-1.10180604e+00 -4.17167664e-01 -1.46478403e+00 2.69201458e-01
-2.72857249e-01 4.67073143e-01 6.67600095e-01 7.82840669e-01
-1.07558537e+00 5.92545211e-01 -1.19798541e+00 -3.67301732e-01
8.15257251e-01 8.76422584e-01 2.26107463e-02 4.47356313e-01
-8.69558215e-01 8.21763694e-01 5.48592806e-01 -3.51445377e-02
-8.57336879e-01 -4.03161913e-01 -2.87748635e-01 3.02513242e-01
-7.31472075e-01 -4.82027531e-01 1.07033122e+00 -7.76336432e-01
-1.62749004e+00 2.63867170e-01 -1.25201821e-01 -8.25620711e-01
1.02955021e-01 4.21843976e-01 -6.90846443e-01 -3.47743817e-02
-1.61594987e-01 7.12694407e-01 2.21809775e-01 -3.68805319e-01
-4.13392425e-01 -3.62702876e-01 -5.87325633e-01 -1.39521450e-01
-4.06352967e-01 3.56478058e-02 2.31081963e-01 -6.78313196e-01
2.42822185e-01 -7.99131095e-01 -2.84705758e-01 1.47573918e-01
-3.32998574e-01 -4.26304713e-02 7.10026801e-01 -5.07761717e-01
1.03287745e+00 -1.89555740e+00 -5.26909411e-01 1.44478112e-01
1.70287520e-01 4.30888176e-01 3.17982316e-01 1.84280679e-01
4.47537340e-02 -1.81025505e-01 -3.25599194e-01 1.95778742e-01
-4.04278368e-01 -7.32567906e-02 1.05787054e-01 3.48339796e-01
4.63711202e-01 1.25671196e+00 -6.69590473e-01 -2.32184883e-02
1.22150339e-01 9.71982896e-01 -3.14146101e-01 2.25301519e-01
2.69153446e-01 3.50587994e-01 -2.94176459e-01 9.09709930e-01
6.57900095e-01 -5.35494797e-02 1.04721695e-01 -2.13941008e-01
-1.95893660e-01 1.94912344e-01 -8.27711821e-01 1.58387017e+00
-6.46292120e-02 6.30803943e-01 -6.74890101e-01 -9.41018045e-01
1.60280335e+00 4.65030849e-01 2.79129893e-01 -1.08872879e+00
5.93613386e-01 6.82951272e-01 1.54914200e-01 -5.64865351e-01
-1.18302628e-01 -3.77663732e-01 -3.74438353e-02 1.41313687e-01
4.13148433e-01 2.60530710e-01 -3.59815598e-01 -4.90112126e-01
1.47050166e+00 2.39088889e-02 3.22981507e-01 -5.69548428e-01
1.79355443e-02 -1.23371489e-01 6.93040788e-01 4.06468272e-01
-2.15731874e-01 3.27557147e-01 8.76377821e-02 -7.91180134e-01
-1.14642739e+00 -9.82622802e-01 -2.62541533e-01 2.15226337e-02
1.29538894e-01 -1.12968840e-01 -1.03238285e+00 2.35136405e-01
-4.37292427e-01 1.08601302e-01 -2.26378158e-01 -3.36052477e-01
-3.77237320e-01 -9.32851970e-01 1.16337490e+00 8.71662259e-01
9.55099285e-01 -1.28330910e+00 -1.50283229e+00 7.31185138e-01
1.58682153e-01 -1.14886701e+00 2.67513812e-01 7.44437456e-01
-1.23933411e+00 -6.48431659e-01 -5.85491121e-01 -9.89848197e-01
5.67619205e-01 -4.66707677e-01 6.11074030e-01 -7.23884925e-02
-6.90486729e-01 -6.87458277e-01 -9.57349017e-02 -8.82289350e-01
4.22848582e-01 -4.25714515e-02 3.21399212e-01 -2.24712953e-01
7.81769574e-01 -1.10196567e+00 -8.39253247e-01 -3.70077819e-01
-5.07887185e-01 2.49076456e-01 7.88518488e-01 7.82395184e-01
8.08770895e-01 3.43659930e-02 9.94342029e-01 -3.55628669e-01
6.69588387e-01 -5.05097985e-01 -6.48184597e-01 -5.50038069e-02
-5.21933675e-01 -1.06198102e-01 9.28783238e-01 -6.31924272e-01
-4.52580273e-01 6.12626374e-01 -8.46569836e-02 -1.68319345e-01
-1.31890237e-01 1.90783918e-01 2.56152511e-01 -4.18754399e-01
6.49462938e-01 6.30628526e-01 -6.85925707e-02 -1.31337389e-01
-7.93661118e-01 7.24690199e-01 6.41243875e-01 2.21680179e-02
1.48986608e-01 1.46188229e-01 2.38866627e-01 -5.86850405e-01
5.60594648e-02 2.05380082e-01 -7.20801353e-02 -2.96390563e-01
8.10334861e-01 -1.10394382e+00 -1.03313506e+00 6.15201890e-01
-1.28611529e+00 -1.48718953e-02 6.77332059e-02 6.75671399e-01
-4.70786422e-01 -2.36781776e-01 -7.43934512e-01 -1.10004520e+00
-1.08678627e+00 -1.07374823e+00 5.07039547e-01 7.13178337e-01
-3.28028709e-01 -4.32695955e-01 -1.70991838e-01 -5.19260406e-01
8.91564071e-01 7.88344204e-01 5.22948980e-01 -6.22526646e-01
-3.98296684e-01 -4.11809653e-01 -3.80709678e-01 1.13593116e-02
5.34625091e-02 -3.05270731e-01 -1.17648005e+00 -7.93142989e-02
7.90108815e-02 -2.28426799e-01 4.11894262e-01 4.61817443e-01
1.52427566e+00 -2.97545969e-01 -5.51113009e-01 7.13530719e-01
2.06169820e+00 8.32058847e-01 9.75098491e-01 6.94744885e-02
3.93527262e-02 -9.81794149e-02 -2.62141913e-01 6.08812511e-01
-2.35650800e-02 1.50606319e-01 3.37111264e-01 9.62284859e-03
-5.70570864e-02 -3.19214582e-01 1.88399628e-01 1.05240452e+00
-2.60935456e-01 -2.22886667e-01 -9.81128097e-01 5.76013207e-01
-1.54811442e+00 -8.01266789e-01 -2.07778141e-01 1.85753000e+00
7.83717513e-01 4.26012874e-02 -1.43138021e-01 4.79709566e-01
7.40137339e-01 -5.43786645e-01 -7.55919814e-01 -6.04907393e-01
-1.86760589e-01 1.20890319e+00 6.84676588e-01 -4.08721417e-01
-6.70316637e-01 4.64085996e-01 6.08231020e+00 4.55453038e-01
-1.38874948e+00 2.85264671e-01 5.30965388e-01 -2.96917737e-01
1.29664421e-01 -4.54153687e-01 -6.57193899e-01 1.06425047e+00
1.61854506e+00 -1.02380447e-01 5.10487199e-01 6.21689260e-01
1.30090401e-01 -2.29611564e-02 -7.05889881e-01 1.43649995e+00
-3.39455009e-01 -1.55001128e+00 -1.75542116e-01 -1.85550541e-01
6.97712779e-01 1.26048788e-01 -1.79031223e-01 1.22570964e-02
-2.03512624e-01 -1.40016520e+00 4.74680722e-01 7.62684464e-01
1.09130561e+00 -1.13248754e+00 1.16020179e+00 3.57157171e-01
-1.20969760e+00 -1.73185647e-01 -6.98849618e-01 -4.58243012e-01
-1.38546154e-01 7.73574889e-01 -5.98071754e-01 -4.25541699e-02
1.17702103e+00 5.55073082e-01 -2.64701128e-01 1.41716015e+00
2.24972814e-01 6.19265378e-01 -4.75304276e-01 -9.34773266e-01
1.61007881e-01 1.67861342e-01 -1.44159853e-01 1.32391036e+00
9.53313589e-01 5.17304063e-01 -4.10499096e-01 1.15479052e+00
-2.60287672e-01 -1.60935491e-01 -6.26739204e-01 -2.24933565e-01
8.07146609e-01 1.24333775e+00 -9.49404597e-01 -1.73566982e-01
-5.43116666e-02 7.96393514e-01 3.69643769e-03 -1.23282835e-01
-9.98812854e-01 -1.08366418e+00 3.95066172e-01 9.75067392e-02
3.44818123e-02 5.92040494e-02 -1.04125607e+00 -8.46763074e-01
1.72010899e-01 -2.53935277e-01 -4.35233146e-01 -8.32209170e-01
-8.14994097e-01 1.10562992e+00 -6.38251901e-01 -1.47594464e+00
7.73075595e-02 -6.30757868e-01 -7.14958072e-01 8.34076703e-01
-1.25883293e+00 -8.61094356e-01 -6.33095205e-01 5.03733099e-01
2.26703659e-01 -1.96114689e-01 1.13678384e+00 2.17298388e-01
-5.72423518e-01 6.90402091e-01 2.54528262e-02 8.83359462e-02
-5.95253147e-02 -4.80368495e-01 3.59805226e-01 7.67129719e-01
-4.19659674e-01 4.16953623e-01 2.87396818e-01 -5.53173065e-01
-1.23411298e+00 -1.38244438e+00 1.28497314e+00 3.19474757e-01
3.14051777e-01 -6.56480432e-01 -7.12770283e-01 2.59253591e-01
3.29846442e-01 2.03691602e-01 6.69989109e-01 -8.49661529e-01
9.66694355e-02 -2.47424796e-01 -1.57264483e+00 4.64895874e-01
1.00954795e+00 -2.74901420e-01 -2.65566885e-01 1.76312998e-01
1.85498282e-01 -2.34540880e-01 -8.70167673e-01 4.86689150e-01
7.83432662e-01 -7.74703026e-01 5.12143254e-01 -8.77896547e-02
3.14101696e-01 -2.20966995e-01 -1.06405623e-01 -9.74512458e-01
-2.32359812e-01 -3.45681161e-01 -1.52880818e-01 8.02338004e-01
4.30621922e-01 -8.29765439e-01 8.11288774e-01 5.10800898e-01
-2.19017774e-01 -1.14009857e+00 -1.22987258e+00 -8.69443655e-01
-3.11219752e-01 -1.49447143e-01 8.36109281e-01 4.98699516e-01
4.14143354e-01 -1.30660772e-01 -2.02623457e-01 1.02259079e-02
4.70382512e-01 -3.90094787e-01 -2.66470999e-01 -1.48060548e+00
2.55664319e-01 -1.69483479e-02 -1.10943985e+00 -1.79453418e-01
-3.10990900e-01 -7.47913539e-01 2.29268670e-01 -1.33398092e+00
2.67536730e-01 -4.84603286e-01 -8.41457963e-01 7.68214583e-01
4.54521388e-01 7.96605527e-01 -2.62142032e-01 7.82174617e-02
-1.96860656e-01 3.62629801e-01 5.85233510e-01 2.75942199e-02
-9.86567661e-02 -2.99531609e-01 -3.77286643e-01 5.11814713e-01
1.37428045e+00 -8.12158406e-01 -3.03834468e-01 -2.61606753e-01
-1.18154921e-01 1.77673593e-01 3.79369467e-01 -2.00534344e+00
7.35560894e-01 2.86789566e-01 1.14423430e+00 -3.47802877e-01
3.64278615e-01 -6.77058160e-01 7.56027997e-01 1.24059701e+00
-5.24151896e-04 3.03334445e-01 4.60686326e-01 3.74149948e-01
-6.91950470e-02 -4.14555334e-02 8.57438385e-01 7.60020362e-03
-4.23082411e-01 3.46987933e-01 -7.04062581e-01 -4.01356876e-01
1.30073452e+00 -7.96622336e-01 -1.60306200e-01 2.88026482e-01
-5.73486686e-01 -6.08671725e-01 -4.08619717e-02 1.03847630e-01
1.04475963e+00 -1.52523577e+00 -2.98542082e-01 6.09724104e-01
7.76858777e-02 -4.14530486e-01 2.53612399e-01 8.22603106e-01
-7.50993252e-01 7.34505594e-01 -1.09525394e+00 -4.42723185e-01
-8.21885586e-01 -4.00274480e-03 4.74363774e-01 1.55242249e-01
-6.18259132e-01 8.26769412e-01 -5.27383745e-01 3.07050049e-02
1.55353144e-01 -5.33772469e-01 -1.91540718e-01 -5.83632588e-01
5.60010493e-01 3.47600043e-01 3.67067248e-01 -8.90211090e-02
-6.84255540e-01 3.85273844e-01 5.16660154e-01 8.90253037e-02
1.72135377e+00 6.05523586e-01 -2.10818797e-01 1.43150151e-01
1.04609776e+00 -8.34977806e-01 -1.09731591e+00 2.83962250e-01
-3.86229418e-02 1.85761571e-01 4.46066745e-02 -1.07309711e+00
-1.13784027e+00 9.81464386e-01 1.19643283e+00 -2.18770891e-01
1.54788840e+00 -5.60849726e-01 1.18800294e+00 2.40243375e-01
7.19722569e-01 -1.00256896e+00 -1.60022601e-01 1.51966333e-01
5.29512823e-01 -6.25575423e-01 -4.20421064e-01 -6.88654333e-02
-2.81511396e-01 1.48743260e+00 8.09761584e-01 -7.48618603e-01
7.50496745e-01 9.56830919e-01 -5.53065717e-01 -1.58207864e-01
-8.66456926e-01 2.44343206e-01 -1.83840707e-01 8.26665044e-01
3.26558530e-01 3.78852278e-01 -7.19689369e-01 1.35947824e+00
-2.59363502e-01 8.80744457e-01 4.87305373e-01 1.10819900e+00
-4.84542727e-01 -5.44634879e-01 1.62948012e-01 9.58541632e-01
-6.77961469e-01 -2.43708730e-01 -1.41639411e-01 2.94372559e-01
4.93499160e-01 7.66385734e-01 2.83095270e-01 -9.51978266e-01
2.20810577e-01 1.11313559e-01 3.77018154e-01 -2.73991555e-01
-1.00348425e+00 -1.78574160e-01 -3.26789320e-01 -4.16331530e-01
-2.19165415e-01 -2.02644303e-01 -1.51213300e+00 -2.77793765e-01
-3.07057321e-01 -2.19773501e-02 1.10562503e+00 7.55578518e-01
7.81352103e-01 7.42646277e-01 3.16258401e-01 -5.53886652e-01
-9.15243030e-02 -9.82586920e-01 -7.07464218e-01 -3.11599463e-01
4.32547182e-02 -5.64514577e-01 -2.06765663e-02 7.18469545e-02] | [8.320710182189941, 2.5108392238616943] |
da4d2220-1112-4667-8c0e-604ee912159d | markerless-outdoor-human-motion-capture-using | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Saini_Markerless_Outdoor_Human_Motion_Capture_Using_Multiple_Autonomous_Micro_Aerial_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Saini_Markerless_Outdoor_Human_Motion_Capture_Using_Multiple_Autonomous_Micro_Aerial_ICCV_2019_paper.pdf | Markerless Outdoor Human Motion Capture Using Multiple Autonomous Micro Aerial Vehicles | Capturing human motion in natural scenarios means moving motion capture out of the lab and into the wild. Typical approaches rely on fixed, calibrated, cameras and reflective markers on the body, significantly limiting the motions that can be captured. To make motion capture truly unconstrained, we describe the first fully autonomous outdoor capture system based on flying vehicles. We use multiple micro-aerial-vehicles(MAVs), each equipped with a monocular RGB camera, an IMU, and a GPS receiver module. These detect the person, optimize their position, and localize themselves approximately. We then develop a markerless motion capture method that is suitable for this challenging scenario with a distant subject, viewed from above, with approximately calibrated and moving cameras. We combine multiple state-of-the-art 2D joint detectors with a 3D human body model and a powerful prior on human pose. We jointly optimize for 3D body pose and camera pose to robustly fit the 2D measurements. To our knowledge, this is the first successful demonstration of outdoor, full-body, markerless motion capture from autonomous flying vehicles.
| [' Michael J. Black', ' Aamir Ahmad', ' Igor Martinovic', ' Roman Ludwig', ' Raffi Enficiaud', ' Rahul Tallamraju', ' Eric Price', 'Nitin Saini'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['markerless-motion-capture'] | ['computer-vision'] | [-5.58433346e-02 -1.58723205e-01 1.37598449e-02 6.41288385e-02
-3.88294071e-01 -7.43108928e-01 2.62574464e-01 -6.74230039e-01
-7.09333241e-01 3.39594692e-01 -1.95888102e-01 3.46589953e-01
5.45181751e-01 -2.17570767e-01 -7.80654609e-01 -1.77630231e-01
5.60133979e-02 6.58712506e-01 6.10240996e-01 -1.15082517e-01
-3.17272246e-01 4.86825287e-01 -1.39734328e+00 -6.15913332e-01
2.44679242e-01 7.52409041e-01 2.83536673e-01 1.20092499e+00
1.03475416e+00 4.59080368e-01 -4.87904817e-01 -3.38938534e-01
6.09828353e-01 -1.73379898e-01 -8.60939920e-02 3.88500571e-01
9.82763112e-01 -8.47365737e-01 -6.43707275e-01 4.77892339e-01
5.77554345e-01 1.26197278e-01 1.26978055e-01 -1.38785124e+00
1.93475392e-02 -1.30069301e-01 -5.67774892e-01 -1.28193125e-01
1.24212003e+00 4.33030844e-01 2.84959137e-01 -6.88783407e-01
8.54299247e-01 9.27281380e-01 1.08705533e+00 9.76587474e-01
-7.42994785e-01 -2.90373206e-01 -6.22082725e-02 -3.86401981e-01
-1.58989096e+00 -5.63825905e-01 6.34828389e-01 -6.40716076e-01
7.94851005e-01 1.47721171e-01 1.28210092e+00 1.18764889e+00
1.89563423e-01 3.96363378e-01 4.62352157e-01 -3.57263803e-01
4.85964268e-02 -1.65737998e-02 -2.79218316e-01 1.09911919e+00
6.63885891e-01 -1.19440548e-01 -6.66354656e-01 -7.11069852e-02
1.09657729e+00 4.82184887e-01 -6.66457593e-01 -1.16171241e+00
-1.40005052e+00 1.54283211e-01 3.74669433e-01 -2.15600640e-01
-2.75495380e-01 5.83856344e-01 -3.42129588e-01 -2.48454809e-01
7.47076124e-02 6.04585484e-02 -2.89935917e-01 -3.31041425e-01
-1.16952968e+00 3.37425083e-01 6.91687226e-01 1.47914517e+00
4.19875741e-01 -6.94756806e-02 1.89962238e-01 -1.01304911e-01
7.32087433e-01 1.23897243e+00 3.27839404e-01 -1.21124685e+00
5.01970828e-01 5.72441280e-01 7.47735918e-01 -9.46089327e-01
-4.71777618e-01 4.78185117e-02 -8.88825133e-02 2.40088329e-01
3.63960952e-01 -5.19057691e-01 -7.26712644e-01 1.57717752e+00
7.25954533e-01 1.20578185e-01 -2.33986601e-01 1.48824573e+00
3.66626114e-01 1.43277109e-01 -3.04756969e-01 1.28136992e-01
1.37354958e+00 -9.18010175e-01 -4.58884209e-01 -9.69340086e-01
5.18391132e-01 -6.39576316e-01 8.04739535e-01 3.16800982e-01
-9.94310677e-01 -5.06361783e-01 -1.20771646e+00 -1.72384024e-01
6.16325848e-02 5.57489216e-01 3.07961762e-01 7.23520100e-01
-9.60936546e-01 2.72526830e-01 -1.49637640e+00 -8.63321841e-01
-3.80190343e-01 6.54200315e-01 -7.26556897e-01 -1.90445051e-01
-8.67510319e-01 1.08630145e+00 -1.22877277e-01 3.38167757e-01
-9.51643348e-01 -1.86635554e-01 -9.54190254e-01 -5.76712608e-01
4.56251085e-01 -1.53882456e+00 1.29883587e+00 -3.54317814e-01
-1.69306028e+00 1.18585980e+00 2.05833875e-02 -4.26642746e-01
7.20057845e-01 -1.18034649e+00 -7.39510655e-02 4.36643779e-01
-1.01030335e-01 6.07246101e-01 8.93416405e-01 -1.00441718e+00
-1.76831365e-01 -7.54327536e-01 -5.63317649e-02 4.56232905e-01
-3.13453786e-02 -2.53562629e-01 -9.90201175e-01 -1.31565422e-01
1.42475829e-01 -1.62737334e+00 -1.52793437e-01 4.59359407e-01
-2.17465326e-01 7.10480034e-01 8.87927949e-01 -5.10709703e-01
8.93386424e-01 -1.89888990e+00 3.53750318e-01 -1.21343158e-01
1.75275922e-01 2.89714307e-01 5.65151095e-01 1.80979431e-01
6.80767894e-01 -5.49314201e-01 -7.50474539e-03 -8.17870796e-01
-1.06248327e-01 1.22552849e-01 9.91369039e-02 8.71954739e-01
-4.94033128e-01 7.41783202e-01 -7.81264842e-01 -5.05502403e-01
4.55576241e-01 7.22391427e-01 -4.08903629e-01 4.92433786e-01
1.86652854e-01 5.18897235e-01 -3.72715145e-01 9.56605852e-01
4.70405966e-01 1.55116124e-02 1.35472536e-01 -1.93687797e-01
-1.48125648e-01 -3.17296475e-01 -1.44132853e+00 2.09733057e+00
-2.14523554e-01 5.40105700e-01 7.41996050e-01 -1.46834612e-01
6.13665760e-01 2.98042715e-01 4.84154135e-01 2.16837168e-01
2.29675949e-01 9.74238515e-02 -6.50011539e-01 -4.83144492e-01
7.09597766e-01 -1.63773522e-01 -2.72264391e-01 -9.14599672e-02
9.20622945e-02 -3.31888914e-01 -2.07066491e-01 2.13048290e-02
1.20515168e+00 7.53216565e-01 3.53763282e-01 2.26999491e-01
1.52662173e-01 2.29246750e-01 4.33407396e-01 4.72567469e-01
-4.30110127e-01 1.00256550e+00 -5.20844638e-01 -3.39914709e-01
-7.52977252e-01 -1.10854125e+00 4.53129590e-01 6.85382664e-01
5.25784135e-01 -5.15414655e-01 -9.11364794e-01 -5.80025502e-02
3.79816234e-01 -1.55161604e-01 -3.31604570e-01 -1.43389732e-01
-7.51477182e-01 -1.20045841e-01 5.71485102e-01 6.81767285e-01
4.76399720e-01 -1.15418501e-01 -1.62110293e+00 1.12743966e-01
-1.87696218e-01 -1.54542685e+00 -5.50339282e-01 -2.55873322e-01
-7.44881928e-01 -1.18668020e+00 -9.83699918e-01 -3.58203381e-01
4.88376826e-01 7.12775648e-01 9.84575748e-01 -2.28240207e-01
-3.30242008e-01 1.24243557e+00 -7.26844147e-02 -2.80133754e-01
4.22054023e-01 -4.12136793e-01 6.44750834e-01 -1.41211152e-01
2.74536721e-02 -2.71386504e-01 -8.49339426e-01 4.97359961e-01
-1.39536306e-01 -9.52388048e-02 3.40461314e-01 1.51420176e-01
6.02265954e-01 -6.43179953e-01 -8.64409149e-01 -1.92809641e-01
-2.24973232e-01 -1.77238569e-01 -6.76886976e-01 -2.15232698e-03
5.15287183e-02 -6.14572585e-01 2.27918595e-01 -4.88406718e-01
-6.44827425e-01 8.28557491e-01 2.98733592e-01 -8.81648123e-01
-2.49949455e-01 -1.89725876e-01 -2.82335728e-01 -4.85273629e-01
6.50486529e-01 -2.29577258e-01 1.14192963e-01 -2.37267703e-01
4.61705744e-01 4.80964899e-01 1.18906367e+00 -1.46818474e-01
8.49604547e-01 9.02474642e-01 2.21778765e-01 -9.98759210e-01
-4.21683282e-01 -8.51560950e-01 -1.35863316e+00 -6.38254225e-01
1.05261803e+00 -1.49150968e+00 -9.20274615e-01 6.90293074e-01
-9.42745507e-01 -4.84961838e-01 -1.34599283e-02 1.02843118e+00
-7.54980683e-01 3.59627247e-01 -4.85510051e-01 -9.50328350e-01
-3.52925330e-01 -8.85720968e-01 1.84171557e+00 3.18226933e-01
-5.01560032e-01 -7.76421309e-01 3.87156188e-01 5.26323259e-01
2.83653699e-02 7.30449617e-01 -6.81304812e-01 3.97380471e-01
-7.10083842e-01 -8.13874006e-01 7.65935659e-01 -2.73888320e-01
-1.48930505e-01 9.39432159e-02 -7.63296902e-01 -5.01143634e-01
-6.32190332e-02 -2.37606600e-01 4.93983239e-01 6.64429247e-01
-8.87367353e-02 -2.87951995e-02 -1.04038334e+00 9.88613069e-01
1.15879560e+00 -2.51314133e-01 2.96908557e-01 4.44374949e-01
1.12131321e+00 2.32615992e-01 8.18760812e-01 5.61364174e-01
6.11455142e-01 1.25642812e+00 4.82172728e-01 2.12825686e-02
-2.88097691e-02 -4.97068405e-01 8.89488935e-01 5.61594129e-01
-3.61034244e-01 -3.30276579e-01 -1.01313770e+00 2.92475969e-01
-1.72444046e+00 -7.24000752e-01 -2.74749398e-01 2.51517272e+00
1.94698200e-01 -9.36074406e-02 3.96669894e-01 -2.96923190e-01
5.65370619e-01 -4.46181744e-02 -4.59046811e-01 2.92059332e-01
2.87866890e-01 -3.17558378e-01 8.96520019e-01 5.70688426e-01
-1.36261296e+00 9.87968445e-01 6.17786646e+00 -4.53140259e-01
-9.85672951e-01 5.51053360e-02 -5.80419242e-01 -6.34153843e-01
4.46314156e-01 1.28368765e-01 -1.14755058e+00 3.91033024e-01
8.51074755e-01 3.31599146e-01 -7.00813765e-03 1.23670924e+00
-1.05527945e-01 -3.01655918e-01 -1.03215623e+00 1.22999907e+00
4.67752457e-01 -4.72754002e-01 -5.67619145e-01 2.21259654e-01
4.64246452e-01 1.12980545e-01 -1.75554827e-01 -2.22908750e-01
1.49068177e-01 -4.61690247e-01 1.04473865e+00 1.03163159e+00
6.19013369e-01 -2.77304351e-01 4.74386245e-01 7.98317969e-01
-1.42186975e+00 8.09135064e-02 -3.95798087e-01 -3.02345961e-01
6.00181937e-01 3.56787801e-01 -5.54226935e-01 2.63439924e-01
7.10656047e-01 5.75163782e-01 -6.11705959e-01 9.98409092e-01
-3.43422711e-01 1.23979047e-01 -7.06602335e-01 1.93960950e-01
-1.87710509e-01 -9.59687158e-02 7.86995888e-01 8.45384419e-01
7.27635801e-01 3.32291842e-01 5.57128370e-01 3.69750381e-01
1.25671938e-01 -5.52604079e-01 -1.04741943e+00 4.70357060e-01
1.68420762e-01 1.33950210e+00 -5.38540423e-01 -2.37314865e-01
-2.95103073e-01 1.43613017e+00 -5.21519743e-02 -8.96359459e-02
-8.54342997e-01 -7.64706209e-02 7.69567013e-01 5.05676091e-01
2.26240292e-01 -1.03726733e+00 4.38246399e-01 -1.75310719e+00
3.72691602e-01 -3.87517631e-01 1.37812672e-02 -1.18266726e+00
-4.53479379e-01 3.79528970e-01 8.59166496e-03 -1.77381396e+00
-4.76084203e-01 -5.56968808e-01 -1.74035221e-01 5.34318864e-01
-7.28213370e-01 -1.45936358e+00 -8.25051546e-01 8.64217699e-01
2.05272824e-01 2.72130072e-01 7.27020800e-01 1.89588755e-01
-4.76559550e-01 3.85963202e-01 -3.27841401e-01 5.90879470e-02
1.03777671e+00 -1.10106039e+00 4.46104079e-01 1.21538734e+00
6.67894483e-02 9.30357754e-01 6.98727548e-01 -9.55876946e-01
-2.23265290e+00 -7.75215089e-01 4.17325199e-01 -1.40876579e+00
1.28258288e-01 -7.53720164e-01 -3.01570058e-01 1.12181544e+00
-2.23214135e-01 3.10857952e-01 5.01739323e-01 -3.90751511e-01
2.76763469e-01 -1.08058721e-01 -1.02193952e+00 5.03917217e-01
1.23317993e+00 -3.58032912e-01 -6.72520161e-01 2.41793349e-01
4.32691753e-01 -1.12493992e+00 -7.02350140e-01 1.98554963e-01
1.23419952e+00 -7.09130049e-01 1.27070117e+00 1.24303168e-02
-3.53645265e-01 -8.27821910e-01 -2.06496641e-01 -9.12903845e-01
-3.83281440e-01 -8.36055279e-01 -5.29068053e-01 9.39227462e-01
-2.72575319e-01 -1.01134650e-01 1.12557459e+00 1.12138724e+00
-3.91358994e-02 -9.73443463e-02 -8.33347380e-01 -8.28974724e-01
-7.91285276e-01 -1.22226298e-01 5.04591018e-02 3.76370937e-01
-3.14153805e-02 2.49046549e-01 -9.97686088e-01 6.15001738e-01
8.65343690e-01 -9.76161752e-03 1.54485238e+00 -1.21991611e+00
-4.08998042e-01 3.33216697e-01 -9.35924053e-01 -1.42745066e+00
-2.45347172e-01 -6.13552704e-02 2.49693692e-01 -1.26850116e+00
-1.37276202e-01 4.28239226e-01 7.06332862e-01 1.20836243e-01
-2.02958256e-01 3.46492290e-01 2.90585220e-01 5.19393206e-01
-8.17810118e-01 2.35437512e-01 6.62837565e-01 4.20651674e-01
-5.36571592e-02 1.49342865e-01 -1.39485732e-01 1.11278951e+00
2.90162802e-01 -4.55138236e-01 -3.11664641e-01 -6.24351442e-01
-2.87203752e-02 3.84370565e-01 8.75197887e-01 -1.77324605e+00
7.58325458e-01 7.90616199e-02 8.90524864e-01 -4.77563590e-01
8.85969102e-01 -1.03115165e+00 7.50614583e-01 6.88982904e-01
2.84395695e-01 4.16586578e-01 1.51309654e-01 6.06194258e-01
1.29273966e-01 1.81329057e-01 5.38108051e-01 -6.06693745e-01
-7.36395538e-01 3.05592626e-01 -3.36943746e-01 -9.54974294e-02
1.22001898e+00 -6.27479553e-01 -8.93995091e-02 -4.39236522e-01
-7.22529054e-01 1.76551640e-01 1.29394162e+00 1.97403908e-01
6.54107273e-01 -1.21888065e+00 -1.54542819e-01 4.33826715e-01
-2.67001446e-02 1.02769680e-01 2.15247199e-01 8.52862954e-01
-9.64710593e-01 5.09029031e-01 -2.64171183e-01 -9.42111671e-01
-1.53627956e+00 5.45283675e-01 4.43342507e-01 1.70292765e-01
-7.10427642e-01 6.92718089e-01 -2.24306345e-01 -5.01728117e-01
-9.56644043e-02 -3.49604368e-01 2.99078614e-01 -4.27278012e-01
4.56056833e-01 5.70779026e-01 -6.08476698e-02 -1.10130811e+00
-8.36116433e-01 1.46271861e+00 6.39634490e-01 -5.70676565e-01
8.21436703e-01 -6.19329333e-01 3.76979679e-01 3.32328290e-01
8.11993182e-01 4.04029161e-01 -1.66145766e+00 8.67461786e-02
-5.37699640e-01 -6.88515246e-01 -2.50382155e-01 -1.91101268e-01
-6.56995118e-01 8.40499461e-01 7.27571189e-01 -3.85524690e-01
8.30794930e-01 -1.16483591e-01 7.93739438e-01 4.62037027e-01
9.99225140e-01 -1.06042314e+00 -4.49881367e-02 4.22632515e-01
5.62304258e-01 -1.00369275e+00 3.01307052e-01 -4.24007803e-01
-6.44416690e-01 9.60826933e-01 7.85854757e-01 -2.49756962e-01
2.71207720e-01 4.23440933e-01 1.05758123e-01 -1.76095307e-01
-3.49770039e-01 -2.80070484e-01 2.39698902e-01 8.51035595e-01
2.53413558e-01 1.69833690e-01 3.47190708e-01 2.92265177e-01
-4.22025919e-01 2.04493284e-01 4.33843523e-01 1.42157781e+00
-4.31594819e-01 -6.16873145e-01 -8.33303154e-01 -3.44872832e-01
-3.38070124e-01 4.02398676e-01 -7.64549136e-01 1.08865190e+00
1.72881857e-01 8.99524391e-01 -3.76349017e-02 -5.65664887e-01
8.65090370e-01 1.21791013e-01 8.50109696e-01 -5.75421453e-01
-2.77346015e-01 1.83736980e-01 -9.93231982e-02 -1.02410853e+00
-4.60581213e-01 -8.10916722e-01 -1.00190425e+00 -1.66935563e-01
-3.50147814e-01 -2.59001195e-01 5.70114195e-01 5.94057679e-01
5.13659537e-01 -1.82941393e-03 2.42911190e-01 -1.53223658e+00
-2.60127097e-01 -6.27463579e-01 -7.15168476e-01 2.63081282e-01
5.54232538e-01 -9.20662880e-01 -1.34651884e-01 3.14748168e-01] | [7.258089065551758, -1.3411588668823242] |
3904dab4-85cf-4bb4-84a6-3482a4ad2eda | tell-me-what-you-see-a-zero-shot-action | 2112.09976 | null | https://arxiv.org/abs/2112.09976v1 | https://arxiv.org/pdf/2112.09976v1.pdf | Tell me what you see: A zero-shot action recognition method based on natural language descriptions | Recently, several approaches have explored the detection and classification of objects in videos to perform Zero-Shot Action Recognition with remarkable results. In these methods, class-object relationships are used to associate visual patterns with the semantic side information because these relationships also tend to appear in texts. Therefore, word vector methods would reflect them in their latent representations. Inspired by these methods and by video captioning's ability to describe events not only with a set of objects but with contextual information, we propose a method in which video captioning models, called observers, provide different and complementary descriptive sentences. We demonstrate that representing videos with descriptive sentences instead of deep features, in ZSAR, is viable and naturally alleviates the domain adaptation problem, as we reached state-of-the-art (SOTA) performance on the UCF101 dataset and competitive performance on HMDB51 without their training sets. We also demonstrate that word vectors are unsuitable for building the semantic embedding space of our descriptions. Thus, we propose to represent the classes with sentences extracted from documents acquired with search engines on the Internet, without any human evaluation on the quality of descriptions. Lastly, we build a shared semantic space employing BERT-based embedders pre-trained in the paraphrasing task on multiple text datasets. We show that this pre-training is essential for bridging the semantic gap. The projection onto this space is straightforward for both types of information, visual and semantic, because they are sentences, enabling the classification with nearest neighbour rule in this shared space. Our code is available at https://github.com/valterlej/zsarcap. | ['Helio Pedrini', 'David Menotti', 'Rayson Laroca', 'Valter Estevam'] | 2021-12-18 | null | null | null | null | ['zero-shot-action-recognition'] | ['computer-vision'] | [ 1.73865139e-01 -5.78927882e-02 -2.88985521e-01 -3.44545871e-01
-5.68557441e-01 -4.66784030e-01 8.50851536e-01 -7.56716961e-03
-3.60139936e-01 4.97016579e-01 5.22255838e-01 2.81010747e-01
-1.26607820e-01 -6.96442842e-01 -7.26948798e-01 -5.98150849e-01
1.22091651e-01 2.27856025e-01 4.26937878e-01 -2.13115528e-01
1.51003018e-01 1.62993297e-01 -2.05567551e+00 7.70994425e-01
4.53088194e-01 1.14773130e+00 4.14719373e-01 5.59985101e-01
-2.75478542e-01 8.72123480e-01 -6.75576806e-01 -6.52381241e-01
4.28614952e-02 -4.64303255e-01 -6.43305480e-01 3.15207958e-01
7.07044005e-01 -2.88522661e-01 -8.58721912e-01 9.87411201e-01
2.09246591e-01 3.60883832e-01 7.85683751e-01 -1.52679908e+00
-9.46544290e-01 3.50362659e-01 -1.01437181e-01 8.59658420e-02
7.91145205e-01 6.94847330e-02 1.23901451e+00 -9.69150424e-01
8.33760500e-01 1.25588381e+00 4.51189339e-01 6.50669515e-01
-9.95005488e-01 -4.19766963e-01 5.65356575e-02 9.86214399e-01
-1.51759541e+00 -4.36287522e-01 8.17076683e-01 -4.37757194e-01
1.00008643e+00 5.31230152e-01 8.54726076e-01 1.83894932e+00
-2.14397296e-01 9.88079846e-01 6.02896392e-01 -4.76699501e-01
1.37186527e-01 5.02591193e-01 1.43139690e-01 5.11165082e-01
-6.94035040e-03 -2.90166229e-01 -9.11661506e-01 8.13796744e-02
5.05049348e-01 3.10747325e-01 -6.59179151e-01 -6.80218101e-01
-1.22629559e+00 8.80343139e-01 3.62044364e-01 6.11870706e-01
-1.83695644e-01 1.63380310e-01 6.04713142e-01 2.05738068e-01
3.72075200e-01 3.93733919e-01 -2.25334704e-01 -1.44856006e-01
-7.37144887e-01 1.74498074e-02 4.13432002e-01 1.12652612e+00
6.04170024e-01 -2.52209902e-01 -3.64334881e-01 8.52167130e-01
1.34374797e-01 4.49493498e-01 9.43752766e-01 -8.90655696e-01
4.64493066e-01 5.80490649e-01 1.43265463e-02 -1.27225375e+00
-8.73129666e-02 -1.86112374e-01 -6.70153379e-01 -2.45632455e-01
2.81975120e-01 5.59684575e-01 -7.87588775e-01 1.90621519e+00
7.68280169e-03 3.15719843e-01 3.34130168e-01 9.23516929e-01
9.25143540e-01 7.91597664e-01 1.12454817e-01 -1.25895485e-01
1.56038499e+00 -1.13121128e+00 -1.00097370e+00 -1.29116029e-01
6.50066793e-01 -4.16952103e-01 1.32259274e+00 2.96894342e-01
-8.44598293e-01 -7.14252889e-01 -1.04940891e+00 -2.27120444e-01
-7.96860218e-01 2.68045932e-01 3.62262785e-01 3.10151339e-01
-1.08874786e+00 7.69229889e-01 -6.34764969e-01 -8.28443706e-01
4.53228652e-01 -2.27066457e-01 -7.79121518e-01 3.95934954e-02
-1.42065966e+00 9.74624336e-01 4.03006643e-01 -3.01454842e-01
-7.79411912e-01 -3.77461374e-01 -1.02973354e+00 2.07424447e-01
4.43747789e-01 -5.69872200e-01 9.10789430e-01 -1.23031080e+00
-1.23683357e+00 1.04485357e+00 -1.74429506e-01 -4.62912172e-01
4.59547549e-01 -2.59304285e-01 -5.22503197e-01 7.69891262e-01
1.87594369e-01 7.98707426e-01 1.06498611e+00 -1.11970484e+00
-3.39075863e-01 -3.05589378e-01 2.54924476e-01 2.08673000e-01
-9.34960842e-01 -6.97795078e-02 -5.81691325e-01 -7.18554556e-01
6.09376170e-02 -7.10394979e-01 2.08958521e-01 3.32817942e-01
-1.80480704e-01 -4.28379178e-01 8.62402737e-01 -7.55381227e-01
1.06566536e+00 -2.46527553e+00 4.05314028e-01 -1.20365672e-01
9.64192972e-02 1.97168678e-01 -2.40764871e-01 6.63634777e-01
-2.07891524e-01 -2.98546236e-02 -1.45743072e-01 -5.08097172e-01
2.41571963e-01 2.91855425e-01 -5.15096605e-01 4.38979536e-01
1.19255364e-01 8.83455396e-01 -1.04123724e+00 -5.55118918e-01
3.56502295e-01 5.75752974e-01 -4.14407909e-01 3.22033137e-01
-3.08251753e-02 1.80955708e-01 -2.74897724e-01 3.73118758e-01
2.81325459e-01 -3.40776533e-01 1.63533628e-01 -5.31069696e-01
1.84818178e-01 1.14094496e-01 -1.00429618e+00 2.20042491e+00
-3.07108343e-01 7.95233607e-01 -4.66720641e-01 -1.41133928e+00
7.68093944e-01 5.75045228e-01 3.63943845e-01 -7.01064944e-01
1.11112893e-02 -2.38446109e-02 -5.50484300e-01 -8.39089751e-01
4.56148654e-01 -5.16519621e-02 -6.92632645e-02 3.87619361e-02
5.56910694e-01 -5.57221763e-04 3.02015722e-01 3.22668076e-01
1.18057144e+00 2.79906988e-01 4.45159048e-01 9.95516255e-02
5.66080809e-01 -3.49916033e-02 2.88513750e-01 6.19588971e-01
-3.18368554e-01 8.49213123e-01 5.02963424e-01 -3.33230704e-01
-9.94751334e-01 -1.18406653e+00 -2.50938684e-01 1.02774274e+00
2.91844815e-01 -8.22362721e-01 -5.83998621e-01 -6.30096138e-01
-2.03107178e-01 8.50060642e-01 -7.12834656e-01 -4.30889904e-01
-2.12145373e-01 -8.31798092e-02 4.63096350e-01 5.39651275e-01
3.76937777e-01 -1.01495469e+00 -7.23094046e-01 -1.11995302e-01
-4.25277144e-01 -1.28788269e+00 -2.22709477e-01 1.67644694e-02
-5.84502339e-01 -1.24215436e+00 -9.52588081e-01 -6.52265131e-01
5.03661394e-01 4.20828491e-01 9.93299007e-01 -2.81318482e-02
-3.21451098e-01 8.38088691e-01 -9.20348585e-01 1.76966693e-02
-3.80580753e-01 -4.53199089e-01 2.41361946e-01 2.55380511e-01
5.51084101e-01 -5.81663311e-01 -5.91385841e-01 2.94755548e-01
-1.06078029e+00 2.58732289e-01 3.73889506e-01 7.56949663e-01
3.45575184e-01 -3.27314615e-01 4.20220762e-01 -3.19709092e-01
1.99045330e-01 -4.43574935e-01 -3.85711566e-02 4.19943511e-01
-1.47532001e-01 8.15160573e-03 5.88161051e-01 -5.32827497e-01
-7.25553572e-01 -7.82730803e-03 1.95794110e-03 -1.10475099e+00
-3.75914365e-01 2.23757908e-01 -2.63311356e-01 2.68774450e-01
6.56522095e-01 5.09136200e-01 5.55335172e-02 -5.56493342e-01
5.91541111e-01 7.63861001e-01 4.25394595e-01 -1.93897605e-01
6.68007851e-01 5.48544943e-01 -1.69644281e-01 -8.04746151e-01
-7.43557930e-01 -7.63881862e-01 -6.60604000e-01 -2.27393150e-01
1.18214178e+00 -9.17802036e-01 -2.66137034e-01 5.76929599e-02
-1.37113833e+00 1.62731245e-01 -5.17308652e-01 5.87977231e-01
-9.81973171e-01 7.15772986e-01 -4.34308976e-01 -5.51871300e-01
6.00315034e-02 -9.89421844e-01 1.09224200e+00 -1.70508370e-01
-3.19900393e-01 -8.34204376e-01 1.75185874e-02 4.70981807e-01
1.32542580e-01 7.24507943e-02 7.38669097e-01 -9.60447490e-01
-3.70423228e-01 -1.45264626e-01 -1.51911676e-01 6.12066209e-01
2.04511419e-01 -1.64056540e-01 -1.10066152e+00 -2.00633422e-01
1.48626417e-01 -4.11106259e-01 9.43500698e-01 2.16513705e-02
1.35282838e+00 -3.74123812e-01 -3.70048583e-01 4.82262164e-01
1.33194053e+00 1.01608753e-01 7.77208447e-01 3.13960880e-01
6.60611391e-01 8.03306341e-01 5.57855487e-01 4.70715344e-01
1.13449328e-01 1.08956265e+00 4.12356675e-01 1.82576284e-01
-3.20263952e-01 -3.95333946e-01 5.96100986e-01 6.22361422e-01
-7.79315159e-02 -3.96653324e-01 -7.08013058e-01 5.34342647e-01
-2.00591755e+00 -1.36794508e+00 1.26469254e-01 2.01731396e+00
6.78854764e-01 -1.97747320e-01 -7.34108090e-02 1.70757368e-01
7.88940787e-01 3.15990716e-01 -2.65199751e-01 -1.79719441e-02
-3.32265288e-01 -4.93153594e-02 1.61750481e-01 5.67519441e-02
-1.19501436e+00 9.53411281e-01 5.42014694e+00 1.13153553e+00
-8.35221529e-01 3.76086414e-01 2.59254575e-01 -1.80816859e-01
-1.26640141e-01 -2.41616592e-01 -4.26200181e-01 5.45077503e-01
9.65547860e-01 -2.00729206e-01 2.26751551e-01 9.09843862e-01
7.61523247e-02 6.97438642e-02 -1.42971194e+00 1.42027032e+00
6.92695677e-01 -1.30558968e+00 4.91627812e-01 -3.27590019e-01
3.73692185e-01 -3.31577450e-01 3.32439579e-02 3.45763385e-01
-4.68630433e-01 -7.99004555e-01 1.03102493e+00 6.50403678e-01
7.95748711e-01 -3.46738070e-01 6.17868662e-01 1.25502080e-01
-1.14002860e+00 -1.54164225e-01 -6.04715228e-01 1.11553613e-02
1.39658660e-01 -2.06163339e-02 -4.53929007e-01 6.39962077e-01
8.23622465e-01 1.26265311e+00 -7.67636776e-01 8.92129898e-01
-3.02749664e-01 1.76771939e-01 7.40997046e-02 1.05855586e-02
1.85626000e-01 -2.20524073e-01 6.72659099e-01 1.15072584e+00
4.95333970e-01 -1.11725898e-02 -7.04990402e-02 8.53117049e-01
1.32019281e-01 1.89590111e-01 -9.61891770e-01 -7.84216225e-02
1.20083801e-01 9.79455769e-01 -5.85224152e-01 -6.27209604e-01
-6.22140050e-01 1.30548894e+00 2.14134425e-01 4.85463828e-01
-1.11557567e+00 -3.73797387e-01 7.23155379e-01 1.52365342e-01
3.52406740e-01 -3.43969576e-02 3.39650482e-01 -1.52812147e+00
2.07462430e-01 -6.95148468e-01 3.34301174e-01 -1.17907286e+00
-1.35125220e+00 7.19129324e-01 3.92394841e-01 -1.73217094e+00
-2.28892565e-01 -8.75709772e-01 -2.05146208e-01 3.54939938e-01
-1.28745890e+00 -1.05364513e+00 -4.36613888e-01 8.59320641e-01
8.52653563e-01 -3.08009237e-01 8.99293959e-01 3.53309542e-01
-3.20376277e-01 4.64417696e-01 2.21981645e-01 1.61388695e-01
8.22477221e-01 -1.01756632e+00 -6.74416423e-02 6.45697832e-01
7.16920793e-01 4.30798233e-01 8.68335605e-01 -3.25380534e-01
-1.26735878e+00 -8.74670029e-01 7.78642237e-01 -6.97252929e-01
9.17694390e-01 -5.50591409e-01 -9.39618826e-01 7.01713443e-01
2.53670067e-01 -2.96750609e-02 7.63027966e-01 -1.37557358e-01
-5.57042778e-01 8.87894258e-02 -8.91619325e-01 5.94546735e-01
1.43721998e+00 -8.00712466e-01 -1.00266111e+00 7.04520047e-01
9.04785156e-01 -1.11760786e-02 -4.52730030e-01 1.43614918e-01
3.49032879e-01 -1.03066909e+00 1.18334913e+00 -8.08855295e-01
6.29809916e-01 -1.79280207e-01 -5.74016273e-01 -1.27086556e+00
-1.77767649e-01 -6.92912005e-03 -1.44121349e-01 1.17505407e+00
6.58005774e-02 -3.31293046e-01 5.76857328e-01 3.07195246e-01
-2.36163169e-01 -4.72089171e-01 -1.15791392e+00 -1.14482665e+00
-3.34421694e-01 -4.53156263e-01 2.67125785e-01 9.96173441e-01
2.38405868e-01 2.95390308e-01 -4.75830019e-01 -1.12162158e-01
4.34283257e-01 -3.86904962e-02 5.65108955e-01 -9.94366825e-01
-2.66291797e-01 -3.15879464e-01 -1.08022916e+00 -1.03452885e+00
5.81694722e-01 -1.01768041e+00 -1.36531338e-01 -1.40866947e+00
4.66042340e-01 1.21968657e-01 -3.26728314e-01 4.54473406e-01
1.23218745e-01 3.71152937e-01 4.57305133e-01 4.39954340e-01
-9.38708305e-01 9.86161947e-01 1.02195430e+00 -4.57931459e-01
1.99977845e-01 -4.76781666e-01 -2.69120038e-01 7.62181580e-01
4.78364885e-01 -3.26990396e-01 -5.32273710e-01 -3.70259881e-01
1.08406521e-01 3.61361951e-02 7.45091319e-01 -1.28031886e+00
3.21007408e-02 -5.69506455e-03 2.52908796e-01 -3.90804231e-01
9.34808373e-01 -1.05243862e+00 2.07231328e-01 2.43750989e-01
-5.61302900e-01 -1.35906547e-01 -4.48533855e-02 8.42332482e-01
-5.11408746e-01 -5.17068684e-01 5.70921361e-01 -1.45843014e-01
-1.07027352e+00 1.49157166e-01 -2.49541104e-01 -2.10147738e-01
1.15069473e+00 -4.68526661e-01 -4.16402280e-01 -6.28187776e-01
-1.04154623e+00 -6.20152317e-02 5.60359478e-01 6.39768362e-01
8.27401340e-01 -1.60611498e+00 -3.98971140e-01 6.61384389e-02
6.32737756e-01 -5.94368279e-01 4.19447929e-01 6.87323868e-01
-4.07877594e-01 5.41662157e-01 -4.87018734e-01 -6.85987532e-01
-1.32578683e+00 9.40995812e-01 1.12682484e-01 1.22309126e-01
-7.96295643e-01 7.36189246e-01 4.93055850e-01 7.54457191e-02
3.45210522e-01 -2.35177293e-01 -5.81812978e-01 4.13585007e-01
7.11450934e-01 9.25821587e-02 -1.75400943e-01 -9.39095795e-01
-5.35227478e-01 7.32402086e-01 2.15836585e-01 -9.92930681e-02
1.23349476e+00 -1.75191909e-01 1.74641162e-01 6.73727393e-01
1.60691547e+00 -2.92300701e-01 -1.08322108e+00 -3.11474890e-01
-2.18961779e-02 -7.34359980e-01 -5.76387011e-02 -4.16977763e-01
-7.75706470e-01 9.70274568e-01 6.53245091e-01 2.52215564e-01
9.82306480e-01 3.92377913e-01 5.81516504e-01 5.87237477e-01
4.23621982e-01 -9.58186507e-01 5.75425982e-01 2.09047556e-01
1.25088847e+00 -1.34290469e+00 -1.90116271e-01 -3.46932977e-01
-9.43976939e-01 1.22221696e+00 4.28979129e-01 -1.34410411e-02
3.37205321e-01 -5.13249695e-01 -1.24995030e-01 -1.72101721e-01
-8.03577244e-01 -3.98784786e-01 3.10678005e-01 6.68373525e-01
8.08482170e-02 -1.13852052e-02 -3.27957720e-01 7.17913747e-01
-1.44628556e-02 -2.33768344e-01 4.91114676e-01 7.34457791e-01
-4.29919243e-01 -9.68273342e-01 -1.63085192e-01 1.25960708e-01
-1.00161836e-01 -8.60307291e-02 -3.38770002e-01 7.53377676e-01
1.33713894e-02 8.85879695e-01 3.84572685e-01 -4.43137646e-01
3.52184266e-01 3.30375373e-01 4.96311963e-01 -6.54334486e-01
-4.67499299e-03 -2.94521034e-01 -3.95153463e-02 -8.89834344e-01
-6.06655538e-01 -5.80992997e-01 -9.44658399e-01 3.97883728e-02
2.16906033e-02 2.47821555e-01 5.77075481e-01 8.43126535e-01
2.60657728e-01 3.92129421e-01 4.62884694e-01 -8.27074647e-01
-5.48764110e-01 -9.08455312e-01 -6.26439512e-01 1.02798939e+00
1.01377428e-01 -1.08620059e+00 -7.50899911e-01 3.15787107e-01] | [8.8910493850708, 0.9901512861251831] |
1db40ddd-6480-4338-a8a3-c3fb8401e8dc | on-device-detection-of-sentence-completion | null | null | https://aclanthology.org/2020.icon-main.53 | https://aclanthology.org/2020.icon-main.53.pdf | On-Device detection of sentence completion for voice assistants with low-memory footprint | Sentence completion detection (SCD) is an important task for various downstream Natural Language Processing (NLP) based applications. For NLP based applications, which use the Automatic Speech Recognition (ASR) from third parties as a service, SCD is essential to prevent unnecessary processing. Conventional approaches for SCD operate within the confines of sentence boundary detection using language models or sentence end detection using speech and text features. These have limitations in terms of relevant available data for training, performance within the memory and latency constraints, and the generalizability across voice assistant domains. In this paper, we propose a novel sentence completion detection method with low memory footprint for On-Device applications. We explore various sequence-level and sentence-level experiments using state-of-the-art Bi-LSTM and BERT based models for English language. | ['Ranjan Samal', 'Anmol Bhasin', 'Priyadarshini Pai', 'Godawari Sudhakar Rao', 'Chandan Pandey', 'Vijeta Gour', 'Rahul Kumar'] | null | null | null | null | icon-2020-12 | ['boundary-detection'] | ['computer-vision'] | [ 5.91802180e-01 -1.21118240e-01 -4.20009084e-02 -4.30852950e-01
-1.18614078e+00 -4.95865613e-01 1.87451884e-01 5.09882510e-01
-5.89692533e-01 4.99723792e-01 4.87658709e-01 -9.71004963e-01
3.07233870e-01 -3.59554350e-01 -2.04527780e-01 -1.75379619e-01
9.33792666e-02 1.80070624e-02 3.28603178e-01 -1.94218546e-01
2.25588635e-01 6.72232509e-01 -9.20217097e-01 8.61881137e-01
7.83302426e-01 8.46970320e-01 7.13012516e-01 1.16810656e+00
-4.96845424e-01 7.85493076e-01 -8.85133505e-01 -1.15170870e-02
-1.57951564e-01 -8.25464576e-02 -8.82781506e-01 -1.05542742e-01
6.65965676e-02 -2.16045380e-01 -5.82831502e-01 8.47735643e-01
7.88188756e-01 2.79842317e-01 9.00575221e-02 -7.93051839e-01
-4.66149777e-01 4.65881705e-01 -3.57990980e-01 6.85069561e-01
7.74843812e-01 5.45072556e-02 5.31249166e-01 -1.10947692e+00
3.21089804e-01 1.42068553e+00 4.92122471e-01 8.93900514e-01
-1.00358045e+00 -2.45013773e-01 2.35133290e-01 -6.61543980e-02
-1.25730574e+00 -1.23935056e+00 5.97923756e-01 -1.28893942e-01
1.95314252e+00 4.55746710e-01 -2.34561145e-01 1.05538619e+00
3.50044906e-01 9.97945487e-01 7.17985451e-01 -8.27892601e-01
2.35317945e-01 -3.06358859e-02 4.34593439e-01 5.56859016e-01
-3.98441315e-01 -2.88620502e-01 -8.39234114e-01 -2.08277777e-01
4.16000307e-01 -6.16847835e-02 -1.86088309e-01 4.75024521e-01
-8.93708408e-01 6.15704536e-01 -2.80943811e-01 4.09335107e-01
-4.07209814e-01 -3.36368650e-01 9.99011040e-01 4.54579830e-01
5.97157061e-01 -8.98503885e-02 -6.57453001e-01 -5.31839252e-01
-1.08974481e+00 -4.30325538e-01 8.94486666e-01 1.27118278e+00
8.91977772e-02 3.01813930e-01 -4.51482296e-01 1.16897011e+00
4.57027346e-01 2.64211416e-01 8.81608009e-01 -5.00311673e-01
1.05157661e+00 3.52041751e-01 4.03471524e-03 -5.59768140e-01
-3.27306241e-01 1.07866563e-01 -8.12678277e-01 -2.63921976e-01
-1.89560559e-02 -3.96327883e-01 -9.35693502e-01 1.22202575e+00
3.42269018e-02 -1.72060877e-02 3.36234987e-01 6.16719842e-01
9.98669147e-01 1.00594985e+00 1.42573074e-01 -5.28313160e-01
1.36810422e+00 -9.86388028e-01 -9.34902549e-01 -5.72578728e-01
1.00232112e+00 -9.79553401e-01 1.36624014e+00 3.05391699e-01
-1.05845618e+00 -4.31576610e-01 -9.34783638e-01 -3.55069578e-01
-2.59496421e-01 3.76614511e-01 2.39217803e-01 9.87983048e-01
-1.27434695e+00 2.67918468e-01 -9.35971320e-01 -6.69356704e-01
5.52354380e-02 4.66824591e-01 -1.65020049e-01 1.43084243e-01
-1.20330238e+00 7.27390766e-01 1.23631783e-01 4.16117311e-01
-4.24649358e-01 -9.50659215e-02 -9.97866631e-01 2.08808765e-01
-9.07982513e-03 -2.37848699e-01 1.70010102e+00 -5.55672407e-01
-1.81860280e+00 8.55478406e-01 -8.25038433e-01 -6.65097952e-01
1.03016131e-01 -3.28501105e-01 -9.07383621e-01 7.47877583e-02
-1.09469533e-01 3.51384640e-01 8.25648308e-01 -5.24112940e-01
-6.94042265e-01 -4.00750637e-01 -2.62236089e-01 2.16810271e-01
-3.53537202e-01 8.97221267e-01 -1.79933637e-01 -5.46415329e-01
2.22050622e-02 -5.08034706e-01 -6.53337240e-02 -4.05507445e-01
-4.20067757e-01 -4.86819059e-01 1.11015940e+00 -1.17349577e+00
1.67615342e+00 -2.32696629e+00 -6.74741805e-01 -3.32554519e-01
-2.54499286e-01 7.11609423e-01 -2.92091966e-01 7.20212698e-01
1.35035470e-01 3.34541827e-01 4.32571620e-02 -5.64507484e-01
-1.31826147e-01 -2.50983308e-03 -5.96669853e-01 1.81989849e-01
2.09679529e-01 6.93500578e-01 -5.45177877e-01 -5.52900910e-01
3.24076444e-01 2.58172274e-01 -9.92582813e-02 1.89291164e-01
5.74762896e-02 1.35118114e-02 -2.51648188e-01 4.89941955e-01
3.76008391e-01 3.45117152e-01 5.71267195e-02 2.97368497e-01
-2.44136736e-01 1.20942461e+00 -8.17799270e-01 1.84562349e+00
-7.71060944e-01 9.10730064e-01 6.44808054e-01 -8.63720477e-01
1.00152791e+00 8.28000128e-01 -9.74190906e-02 -7.08918810e-01
-5.93746118e-02 1.92260608e-01 -1.62155665e-02 -6.01282239e-01
7.36835778e-01 5.40242381e-02 1.74177773e-02 2.30963126e-01
-1.64511010e-01 1.60365611e-01 -4.57616262e-02 2.17506081e-01
1.42792487e+00 -3.73066008e-01 2.96685040e-01 -9.79109481e-02
6.25528038e-01 -1.72786653e-01 5.29825985e-01 6.84217036e-01
-6.25344992e-01 5.67453027e-01 1.63318425e-01 -4.82414588e-02
-8.12792420e-01 -9.34244692e-01 1.71838045e-01 1.23812401e+00
-4.62499887e-01 -3.47848296e-01 -8.64075422e-01 -4.99306381e-01
-3.53119582e-01 1.05938184e+00 3.63145202e-01 -3.53907198e-02
-8.00759494e-01 -1.35107741e-01 8.36452603e-01 5.67878008e-01
2.65925884e-01 -1.43648899e+00 -2.60734379e-01 5.59387505e-01
-3.09999079e-01 -1.70385337e+00 -9.23303008e-01 1.48348466e-01
-1.08333981e+00 -3.61701012e-01 -3.49696338e-01 -1.30026042e+00
2.71470755e-01 5.21238625e-01 6.39247417e-01 -2.42985800e-01
-1.81458339e-01 2.41823748e-01 -3.17309022e-01 -3.45339715e-01
-7.94111490e-01 7.89681002e-02 3.08783412e-01 -1.56577528e-01
8.06678772e-01 -3.64178061e-01 -2.39958942e-01 4.86361906e-02
-5.70771337e-01 -1.38795063e-01 4.89886671e-01 7.05530882e-01
1.32845044e-01 -7.40531608e-02 1.01230180e+00 -5.56924045e-01
1.33266187e+00 -3.42149734e-02 -2.13695765e-01 3.96182269e-01
-2.18734905e-01 -2.23338664e-01 8.17767799e-01 -3.24811578e-01
-1.41380823e+00 1.38232529e-01 -6.84391499e-01 -1.18381664e-01
-5.54839015e-01 5.18718302e-01 -4.57155824e-01 4.16600376e-01
4.07441527e-01 5.33401370e-01 -1.90602720e-01 -6.81081712e-01
3.99895459e-02 1.77698243e+00 3.87869477e-01 -1.92227975e-01
-1.42053915e-02 -7.17105716e-02 -7.05903471e-01 -1.55512333e+00
-4.25801814e-01 -8.34260046e-01 -6.34183109e-01 1.31318435e-01
6.40930295e-01 -8.74701619e-01 -8.46557915e-01 3.03779364e-01
-1.78851211e+00 -1.66680127e-01 8.35331157e-02 4.85155344e-01
-2.89703310e-01 7.38964677e-01 -1.14505303e+00 -1.38155901e+00
-1.06423044e+00 -9.91131783e-01 1.19539106e+00 -4.57300805e-02
-5.43325961e-01 -7.69302607e-01 -4.77417260e-01 5.85901737e-01
2.65424997e-01 -5.68220675e-01 8.92664850e-01 -8.48366261e-01
-1.21774025e-01 -3.98376882e-01 -1.41086295e-01 6.12678707e-01
2.85725266e-01 -2.53088653e-01 -1.12673318e+00 -2.91397542e-01
4.54543710e-01 2.92571466e-02 6.15119755e-01 5.22420347e-01
8.98856103e-01 -5.78220248e-01 -4.13652003e-01 -7.30591342e-02
9.18833971e-01 5.43603897e-01 4.85772073e-01 1.77584193e-03
4.01656002e-01 6.30022109e-01 5.80825806e-01 3.47380489e-01
-7.04450458e-02 4.22636360e-01 -2.70878762e-01 1.74327821e-01
-9.69793722e-02 -3.71792406e-01 9.39297199e-01 1.28212082e+00
8.20268929e-01 -7.23154187e-01 -1.06357753e+00 6.54234231e-01
-1.82875848e+00 -9.11982298e-01 -2.76025862e-01 2.25591636e+00
8.20429385e-01 4.79142904e-01 -2.20675301e-02 2.09465533e-01
1.24056613e+00 2.04664737e-01 -2.55003661e-01 -1.28782713e+00
2.64398843e-01 8.58919844e-02 3.27435881e-01 7.89419115e-01
-9.65346932e-01 1.15719199e+00 6.50353193e+00 9.89755571e-01
-1.20257640e+00 4.99960512e-01 6.04425669e-01 -6.88479468e-02
7.65577629e-02 -2.92980462e-01 -1.12560833e+00 4.82841671e-01
1.38774669e+00 4.88678403e-02 3.42326611e-01 7.83043265e-01
1.06827462e+00 -1.67844951e-01 -1.12521720e+00 1.11515665e+00
-8.10368732e-02 -1.10192180e+00 -1.82321802e-01 -4.36431140e-01
-1.41244121e-02 2.64163226e-01 -3.08748990e-01 3.58200014e-01
-1.87629208e-01 -7.85392225e-01 4.52705950e-01 -4.56152596e-02
9.83297467e-01 -5.37989914e-01 4.21085626e-01 7.08914220e-01
-1.16748190e+00 1.12672113e-02 -2.38700211e-01 -4.25662518e-01
5.15045285e-01 5.93692958e-01 -1.14667618e+00 6.88226819e-02
4.36360061e-01 1.88754737e-01 -1.42305344e-01 5.68513274e-01
-4.03167717e-02 1.02604115e+00 -3.84843856e-01 -4.18878973e-01
1.63092762e-01 9.70319659e-02 5.92445910e-01 1.71160448e+00
2.61122346e-01 1.38337150e-01 1.39296949e-01 4.11321819e-01
-8.10039192e-02 1.67188704e-01 -6.93067908e-01 -3.47959876e-01
7.20597625e-01 1.04681754e+00 -6.17472112e-01 -2.86874235e-01
-6.55532718e-01 1.33065271e+00 2.36161768e-01 4.27685618e-01
-3.98014158e-01 -7.69547403e-01 5.31897962e-01 7.74431881e-03
-4.42207716e-02 -7.27352440e-01 -5.56807637e-01 -1.01914155e+00
6.59342229e-01 -7.32715070e-01 1.19571596e-01 -6.19393826e-01
-1.00911975e+00 7.18667924e-01 -6.43696666e-01 -1.05714846e+00
-1.83153659e-01 -5.97215891e-01 -8.98388743e-01 1.01450646e+00
-1.30508482e+00 -8.38848233e-01 3.16670060e-01 2.39406839e-01
1.32668531e+00 -2.35674366e-01 9.88491178e-01 3.76614064e-01
-6.95250630e-01 5.59416533e-01 1.55805334e-01 3.42576504e-01
3.87373120e-01 -8.05674493e-01 1.00348067e+00 1.15217054e+00
-5.90231735e-03 7.95811355e-01 5.19919813e-01 -6.28248751e-01
-1.53852808e+00 -1.03573465e+00 1.79511762e+00 -1.55302122e-01
5.48513830e-01 -8.28017890e-01 -8.26481998e-01 6.32803679e-01
3.56524587e-01 -2.88073212e-01 7.27829516e-01 1.89598426e-01
1.81533784e-01 -1.68484077e-01 -1.12173736e+00 7.20908165e-01
1.03420997e+00 -1.14113986e+00 -7.23465085e-01 5.63549399e-01
1.16096604e+00 -3.32960546e-01 -6.09182894e-01 1.86843634e-01
2.68362641e-01 -4.87733841e-01 4.80069607e-01 -5.30008852e-01
1.79546028e-02 -3.87563482e-02 -2.12423369e-01 -9.63182211e-01
2.53969394e-02 -1.03726339e+00 -1.88839987e-01 1.74532843e+00
7.00177073e-01 -5.85121334e-01 7.75991380e-01 1.03162551e+00
-5.11440516e-01 -5.91782629e-01 -1.14619911e+00 -9.08953786e-01
-3.15748811e-01 -8.93239439e-01 3.57413381e-01 4.22891021e-01
5.96395433e-01 9.14678812e-01 -9.75080431e-02 2.74005055e-01
-1.44516706e-01 -2.55912304e-01 2.15457305e-01 -5.56598842e-01
-1.33683654e-02 -1.94966719e-01 -9.12515447e-02 -1.60092020e+00
2.75758564e-01 -6.49788499e-01 3.78752351e-01 -1.69803488e+00
-2.18822315e-01 -3.31472605e-02 -5.84967434e-02 3.06930900e-01
1.89139068e-01 -6.36695683e-01 1.53690681e-01 1.41318282e-02
-5.05681634e-01 5.62168777e-01 5.27220070e-01 -1.63449809e-01
-6.71986759e-01 2.05364063e-01 -2.97126293e-01 5.77423573e-01
1.15845954e+00 -4.86722261e-01 -3.65983903e-01 -8.17284346e-01
-2.54764467e-01 5.16874850e-01 -8.05735737e-02 -8.40210915e-01
5.55804074e-01 3.16959880e-02 2.69004051e-03 -8.54634702e-01
3.86091232e-01 -5.97700596e-01 -6.61366045e-01 3.28277677e-01
-6.21205866e-01 3.17473471e-01 2.75567651e-01 4.13331836e-01
-3.32347810e-01 -4.94481713e-01 5.06791353e-01 -4.20960896e-02
-7.27216423e-01 -6.71345294e-02 -1.07471657e+00 -6.06241450e-02
6.49694443e-01 -3.40943187e-01 -2.27346241e-01 -5.02833128e-01
-6.23251498e-01 1.19113170e-01 -9.74915922e-02 7.36576259e-01
1.08026397e+00 -8.17630053e-01 -4.50589389e-01 3.72237325e-01
-2.81355418e-02 -1.66231513e-01 1.04865342e-01 7.14503050e-01
-4.20317501e-01 1.03703082e+00 4.99072611e-01 -4.41585600e-01
-1.62351096e+00 5.87729037e-01 1.32562757e-01 -1.35104686e-01
-5.40882528e-01 7.37756431e-01 -1.29701674e-01 -6.13675654e-01
5.05497694e-01 -5.49832046e-01 3.86708714e-02 -4.33971047e-01
6.93513274e-01 3.41760606e-01 6.43654168e-01 -4.04090464e-01
-6.47196054e-01 -2.81190276e-01 -2.95791537e-01 -3.49987030e-01
8.09437275e-01 -6.14890754e-01 7.60131553e-02 4.16258365e-01
1.15860915e+00 -1.15330324e-01 -6.07208192e-01 -2.31478855e-01
5.92492282e-01 -6.96078315e-02 2.40074947e-01 -6.45179570e-01
-2.59795725e-01 1.29739141e+00 4.39166546e-01 2.23280117e-01
9.30049777e-01 -2.99846381e-01 1.32544601e+00 6.18402779e-01
2.49921560e-01 -1.70908630e+00 -4.54652280e-01 8.63569558e-01
9.27241981e-01 -1.22971272e+00 -6.37977302e-01 -4.20552880e-01
-5.32485306e-01 1.07135248e+00 5.37871659e-01 3.25836599e-01
5.93810380e-01 5.39808273e-01 1.98187664e-01 2.70841122e-01
-9.24821973e-01 -2.27706265e-02 -1.16622970e-01 5.53682745e-01
9.90665317e-01 9.09314975e-02 -5.57046354e-01 7.06431329e-01
1.05546579e-01 5.96180465e-03 3.85475963e-01 1.15071452e+00
-6.92869067e-01 -1.03161180e+00 -2.38557681e-01 5.65750480e-01
-7.47743845e-01 -6.72686815e-01 -5.64667940e-01 4.84305173e-02
-5.02673268e-01 1.83351135e+00 1.09757669e-02 -3.11448812e-01
3.10813755e-01 4.13558394e-01 6.16242141e-02 -1.04792464e+00
-6.27477050e-01 5.60264401e-02 6.38369262e-01 -2.81359524e-01
2.20634323e-02 -5.83020985e-01 -1.55624521e+00 -2.09374592e-01
-5.07797301e-01 7.97922760e-02 7.89034963e-01 1.12434316e+00
9.38930690e-01 3.61403942e-01 6.39065802e-01 -4.00986373e-01
-6.66926861e-01 -1.34731948e+00 -4.78868246e-01 -2.16877982e-02
4.27650362e-01 2.60506898e-01 -1.63327926e-03 1.03620134e-01] | [14.4154052734375, 6.902316093444824] |
13e72faf-d5ed-4a02-81e0-4bd8a616b379 | beyond-tabula-rasa-reincarnating | 2206.01626 | null | https://arxiv.org/abs/2206.01626v2 | https://arxiv.org/pdf/2206.01626v2.pdf | Reincarnating Reinforcement Learning: Reusing Prior Computation to Accelerate Progress | Learning tabula rasa, that is without any prior knowledge, is the prevalent workflow in reinforcement learning (RL) research. However, RL systems, when applied to large-scale settings, rarely operate tabula rasa. Such large-scale systems undergo multiple design or algorithmic changes during their development cycle and use ad hoc approaches for incorporating these changes without re-training from scratch, which would have been prohibitively expensive. Additionally, the inefficiency of deep RL typically excludes researchers without access to industrial-scale resources from tackling computationally-demanding problems. To address these issues, we present reincarnating RL as an alternative workflow or class of problem settings, where prior computational work (e.g., learned policies) is reused or transferred between design iterations of an RL agent, or from one RL agent to another. As a step towards enabling reincarnating RL from any agent to any other agent, we focus on the specific setting of efficiently transferring an existing sub-optimal policy to a standalone value-based RL agent. We find that existing approaches fail in this setting and propose a simple algorithm to address their limitations. Equipped with this algorithm, we demonstrate reincarnating RL's gains over tabula rasa RL on Atari 2600 games, a challenging locomotion task, and the real-world problem of navigating stratospheric balloons. Overall, this work argues for an alternative approach to RL research, which we believe could significantly improve real-world RL adoption and help democratize it further. Open-sourced code and trained agents at https://agarwl.github.io/reincarnating_rl. | ['Marc G. Bellemare', 'Aaron Courville', 'Pablo Samuel Castro', 'Max Schwarzer', 'Rishabh Agarwal'] | 2022-06-03 | null | null | null | null | ['humanoid-control'] | ['robots'] | [-1.18195839e-01 2.02671885e-02 -3.42759460e-01 5.43850735e-02
-6.84850216e-01 -1.08095467e+00 5.57434738e-01 -3.02635580e-01
-7.05159187e-01 1.20815694e+00 -1.59192130e-01 -9.90621507e-01
-2.37768129e-01 -7.66221225e-01 -9.46337402e-01 -7.60727167e-01
-2.77546048e-01 7.97366917e-01 -2.74304077e-02 -6.85450554e-01
2.30703562e-01 5.38671374e-01 -1.49326229e+00 -1.36061892e-01
6.93748474e-01 3.63385051e-01 2.03338176e-01 7.68189371e-01
1.28675103e-01 9.92270470e-01 -6.44201875e-01 7.47872964e-02
6.61234677e-01 -5.05573511e-01 -8.61326158e-01 -1.47310540e-01
1.51387721e-01 -3.00762832e-01 -1.60262838e-01 5.34358919e-01
5.63560307e-01 3.65962148e-01 1.22897346e-02 -1.52009177e+00
1.57609865e-01 6.71116114e-01 -5.44384360e-01 7.63973445e-02
9.69667733e-02 8.83310735e-01 9.40484762e-01 -1.94932017e-02
6.45013094e-01 1.12486255e+00 6.58625841e-01 6.79501235e-01
-1.00565040e+00 -7.91850924e-01 2.83249855e-01 -9.56195667e-02
-9.16281044e-01 -3.79947066e-01 4.58231926e-01 -1.96584284e-01
1.06094301e+00 2.61892259e-01 9.66650248e-01 1.17473996e+00
9.11247432e-02 8.42450500e-01 1.14173102e+00 -2.26059258e-01
5.08775234e-01 -1.28395915e-01 -5.59019268e-01 8.73110294e-01
4.86543387e-01 3.84531528e-01 -3.37823302e-01 -6.72940239e-02
7.74445415e-01 -5.10801196e-01 5.97768612e-02 -6.84294105e-01
-1.14463770e+00 9.72862363e-01 4.29939896e-01 1.71186090e-01
-3.47932339e-01 7.20632374e-01 3.82184118e-01 5.30989349e-01
-5.16608506e-02 1.07513154e+00 -7.26202548e-01 -5.65763414e-01
-9.29240882e-01 6.29406154e-01 8.95986676e-01 7.51709819e-01
8.04782331e-01 4.35550123e-01 3.02622765e-01 4.77128923e-01
2.53580749e-01 4.85544086e-01 3.78319710e-01 -1.58444381e+00
3.91181618e-01 3.35042834e-01 4.20983732e-01 -5.91429293e-01
-7.62805164e-01 -6.22547925e-01 -3.41167837e-01 6.71931207e-01
5.39987266e-01 -8.01796734e-01 -4.90456879e-01 1.82450163e+00
5.46871126e-01 2.77700484e-01 4.24362630e-01 1.13250184e+00
2.11903721e-01 6.64320290e-01 -1.47670329e-01 -1.10741377e-01
1.04517639e+00 -1.25257647e+00 -1.47363603e-01 -3.83850724e-01
8.70143473e-01 -5.06435096e-01 1.25212240e+00 6.41994357e-01
-1.24106085e+00 -1.69701800e-01 -1.18541014e+00 3.56284231e-01
-4.66982484e-01 -1.62152156e-01 1.15602922e+00 7.23847866e-01
-1.19489825e+00 6.13941848e-01 -8.73246372e-01 -4.47058618e-01
3.33352506e-01 5.78493834e-01 -2.87676565e-02 1.86506689e-01
-1.13705528e+00 9.82398689e-01 2.98761338e-01 3.92013155e-02
-1.12131190e+00 -6.40820622e-01 -5.54080009e-01 -8.70243013e-02
9.20521200e-01 -9.11600709e-01 1.71442866e+00 -1.18188298e+00
-2.05508542e+00 3.73643547e-01 4.64256406e-01 -7.05616295e-01
7.04499125e-01 -8.17928761e-02 9.19415355e-02 -2.04503596e-01
-8.05901662e-02 7.35483050e-01 7.89709985e-01 -1.45323980e+00
-9.42790508e-01 1.59938887e-01 6.90721035e-01 6.87547028e-01
6.51812255e-02 -2.54193366e-01 -2.12105408e-01 -1.81102857e-01
-4.83364880e-01 -1.17218232e+00 -4.67590332e-01 -5.13918102e-01
1.61640227e-01 -1.41339317e-01 7.55658507e-01 -1.24165863e-01
9.02576029e-01 -1.72256601e+00 2.58703887e-01 7.99464583e-02
2.47311834e-02 3.84221286e-01 -3.72339934e-01 5.94421566e-01
2.76795894e-01 1.72850430e-01 -1.85245827e-01 -1.28499851e-01
3.24344575e-01 5.47116220e-01 -3.20731640e-01 3.76155108e-01
-1.89757288e-01 8.89681756e-01 -1.39218044e+00 -2.16934249e-01
3.01590949e-01 1.31390523e-02 -8.90008152e-01 3.86135764e-02
-6.79374695e-01 5.94017148e-01 -4.82009202e-01 4.73137230e-01
3.27173084e-01 -1.18571892e-01 6.69140756e-01 3.05239350e-01
-4.92998987e-01 3.27700198e-01 -1.09165549e+00 1.96244025e+00
-9.30215359e-01 4.93296176e-01 2.74581313e-01 -1.13205433e+00
5.76235533e-01 3.51932310e-02 8.65254462e-01 -9.33506787e-01
1.74551383e-01 2.53123492e-01 2.39940658e-01 -3.56656134e-01
5.75323284e-01 -1.46399841e-01 -2.55079925e-01 7.76428223e-01
-8.86213183e-02 -3.78332764e-01 4.88563746e-01 4.06038240e-02
1.38144600e+00 6.86428010e-01 3.00228179e-01 -1.49421558e-01
2.15814158e-01 5.14003813e-01 5.15622139e-01 1.13481319e+00
-3.09056669e-01 5.44328727e-02 5.13457060e-01 -7.12276936e-01
-1.06379211e+00 -7.73722410e-01 3.09579283e-01 1.42784250e+00
-7.09962449e-04 -4.75665629e-01 -5.15259326e-01 -9.15289342e-01
1.90940544e-01 7.65756488e-01 -4.59588796e-01 -7.10030869e-02
-8.73360574e-01 -8.85387123e-01 9.07301724e-01 6.27558529e-02
5.80091357e-01 -1.39437091e+00 -1.40277207e+00 4.72004592e-01
-9.08065140e-02 -6.64294064e-01 -8.36777389e-02 3.19908619e-01
-6.02030277e-01 -1.09145296e+00 -2.79900521e-01 -3.91024351e-01
2.45586365e-01 3.13717782e-01 1.28800440e+00 3.61282170e-01
-1.41255468e-01 4.73441660e-01 -3.65115076e-01 -2.79589444e-01
-4.12863970e-01 3.99461925e-01 9.94320661e-02 -6.56182885e-01
-1.97123557e-01 -4.78000224e-01 -6.11532867e-01 3.32094282e-01
-7.27161288e-01 1.88187972e-01 6.92451656e-01 8.02907884e-01
1.19755968e-01 2.05659434e-01 7.49759614e-01 -7.67799020e-01
7.65971482e-01 -6.26388252e-01 -1.09801209e+00 4.61348221e-02
-6.52140439e-01 1.80326000e-01 8.91779959e-01 -3.52945656e-01
-8.69694650e-01 1.66975078e-04 -2.74896137e-02 -2.86317825e-01
3.99926193e-02 5.90500295e-01 2.10075513e-01 -1.39778838e-01
7.01121986e-01 9.07078981e-02 1.77338660e-01 -5.10043018e-02
4.55415279e-01 3.18945616e-01 1.33275777e-01 -1.03093362e+00
8.87069941e-01 2.79049069e-01 2.42311191e-02 -4.91125375e-01
-5.42006612e-01 -1.78996831e-01 -5.33849299e-02 -3.05735111e-01
4.35930341e-01 -7.06008852e-01 -1.16618860e+00 1.07714973e-01
-7.06820548e-01 -1.36388636e+00 -5.35924792e-01 3.17980945e-01
-9.47826326e-01 -1.12480730e-01 -2.83668101e-01 -5.71015418e-01
-1.56204581e-01 -1.31247354e+00 6.26145065e-01 4.87734646e-01
-4.47490886e-02 -1.01799536e+00 5.27203918e-01 6.46571100e-01
6.71408117e-01 8.73830914e-02 6.13042831e-01 -2.45043591e-01
-5.03834605e-01 3.24293405e-01 8.02427530e-02 -1.14562400e-01
-1.17765898e-02 3.82663757e-02 -6.48864925e-01 -7.01142132e-01
-4.96829122e-01 -7.79829085e-01 4.52751279e-01 1.73834354e-01
9.35281813e-01 -4.07600462e-01 -9.04498771e-02 5.78688741e-01
1.18333852e+00 3.94664496e-01 3.94584030e-01 1.18181491e+00
4.25145715e-01 3.39831561e-01 8.59249651e-01 6.03565812e-01
5.77409089e-01 5.94908834e-01 6.54747665e-01 -2.04827935e-01
3.41781676e-02 -3.26523781e-01 7.08593726e-01 5.81805170e-01
-1.97541192e-01 -4.17534173e-01 -1.03050327e+00 2.27511942e-01
-2.08795786e+00 -9.90851045e-01 5.80036819e-01 2.02817655e+00
7.52939701e-01 1.16745517e-01 3.20432723e-01 -3.11602324e-01
1.12193346e-01 1.69106036e-01 -1.00060344e+00 -6.72458291e-01
2.09552050e-01 2.34825745e-01 8.23924601e-01 4.79438573e-01
-8.82123530e-01 1.55529535e+00 5.79980993e+00 7.61205852e-01
-1.35517132e+00 9.55055642e-04 2.18250588e-01 -3.46309662e-01
-9.89757478e-02 2.37964794e-01 -4.37696338e-01 2.22322688e-01
1.24643731e+00 -2.98818499e-01 1.31241500e+00 8.21894348e-01
6.62046254e-01 -2.34190807e-01 -7.98119068e-01 6.20200038e-01
-2.86754429e-01 -1.42107081e+00 -1.94194376e-01 2.08574042e-01
6.32989645e-01 3.12375993e-01 9.76244211e-02 1.06167269e+00
1.17308688e+00 -1.09781396e+00 7.16133118e-01 1.52120784e-01
4.40571129e-01 -1.06215954e+00 6.31709576e-01 5.24868906e-01
-9.94872868e-01 -9.90118161e-02 -3.64059091e-01 -3.74416769e-01
-1.44736901e-01 -3.11335206e-01 -1.12937081e+00 7.52753437e-01
6.73435450e-01 5.75682402e-01 -2.81429201e-01 8.84809792e-01
-1.73552185e-01 7.88675070e-01 -4.41372782e-01 -1.60239115e-01
8.81028712e-01 -3.20480675e-01 4.05422688e-01 9.27913845e-01
1.35841459e-01 -5.96412197e-02 3.86630088e-01 5.89954257e-01
-2.03372762e-01 -8.47926959e-02 -7.79796958e-01 -1.08612828e-01
3.91287774e-01 1.45307720e+00 -8.10103118e-01 -1.05209328e-01
-2.00609341e-01 4.33489203e-01 4.51332718e-01 3.15807909e-01
-1.18982041e+00 -1.25814170e-01 8.22582126e-01 -1.04717270e-01
1.57876879e-01 -3.47472996e-01 -4.39169183e-02 -1.09160304e+00
-4.38975990e-01 -1.53362334e+00 3.72206330e-01 -7.19789743e-01
-8.19100678e-01 2.53440291e-01 -1.22897141e-02 -1.26234496e+00
-4.94149148e-01 -2.90259629e-01 -4.46484983e-01 3.23359668e-01
-1.70149136e+00 -1.01195061e+00 -7.95303285e-02 4.30890828e-01
7.02438235e-01 -3.90677601e-01 5.86253464e-01 5.54740652e-02
-6.54104650e-01 4.13516402e-01 1.81498706e-01 -3.83445501e-01
6.13374293e-01 -1.26776505e+00 2.00869545e-01 7.04498827e-01
-4.67035733e-03 5.22226036e-01 8.22376311e-01 -4.26933140e-01
-1.93783367e+00 -8.40227783e-01 -3.48949358e-02 -3.23870599e-01
9.46431756e-01 -3.49204242e-01 -3.40645760e-01 7.65797079e-01
4.82305497e-01 -1.93303168e-01 3.00080329e-01 3.00326794e-01
2.23234817e-01 -8.87358412e-02 -9.90374029e-01 9.68595684e-01
1.04166174e+00 1.41404858e-02 -3.21813792e-01 2.81453401e-01
5.09257913e-01 -6.65944517e-01 -6.23769701e-01 2.79264987e-01
6.07288003e-01 -9.16471958e-01 9.98643398e-01 -8.05489361e-01
1.44407302e-01 -5.43638051e-01 1.24326482e-01 -1.66155589e+00
5.73227182e-02 -1.12436795e+00 4.90263291e-02 7.23006129e-01
4.29346323e-01 -8.57328773e-01 9.45719659e-01 2.41530016e-01
-1.94142506e-01 -7.17833400e-01 -8.33354473e-01 -8.48159373e-01
6.54438496e-01 -4.77937430e-01 6.25056624e-01 1.05726552e+00
6.54247729e-03 2.19996646e-01 -5.07391036e-01 2.36058041e-01
3.91114175e-01 2.12320641e-01 1.35183156e+00 -6.92806661e-01
-5.97608685e-01 -5.07347405e-01 1.92957759e-01 -9.32760000e-01
3.20425302e-01 -8.39878857e-01 2.82850236e-01 -1.51967311e+00
-2.74028420e-01 -1.07645607e+00 -2.13655278e-01 9.09740388e-01
3.10794234e-01 -5.88658378e-02 5.83153546e-01 2.44958803e-01
-9.56395209e-01 5.55099428e-01 1.44287670e+00 -1.58572197e-02
-5.00952601e-01 -8.05703700e-02 -7.73475289e-01 7.73051202e-01
1.38492227e+00 -4.91268396e-01 -8.25423717e-01 -3.85236204e-01
4.76470798e-01 2.49684572e-01 2.74886042e-01 -1.19555306e+00
2.57851779e-01 -6.56224847e-01 -1.05356395e-01 -4.03319448e-02
8.36892724e-02 -7.77113080e-01 5.93715049e-02 6.52390838e-01
-2.07872197e-01 2.69507766e-01 5.44231117e-01 1.67675346e-01
2.27608815e-01 -2.70422697e-01 6.84669316e-01 -4.29805219e-01
-6.92998827e-01 -1.32614356e-02 -8.41400325e-01 3.25433105e-01
1.26516104e+00 6.22624345e-02 -5.74664474e-01 -3.07936311e-01
-4.19815660e-01 7.57869899e-01 6.61853552e-01 3.50081682e-01
1.69737265e-01 -6.09601676e-01 -4.57844913e-01 2.33242474e-02
-3.16302657e-01 3.71292531e-02 -2.73550805e-02 8.29123855e-01
-8.63289475e-01 3.42831314e-01 -5.65781057e-01 -2.08699122e-01
-8.94892216e-01 3.84628534e-01 8.09156358e-01 -6.55921757e-01
-6.30514145e-01 4.18557823e-01 -1.85694441e-01 -9.36891079e-01
7.00799450e-02 -3.15907508e-01 1.01244599e-02 -7.41114765e-02
8.38365406e-02 2.22570539e-01 -2.61489861e-02 -1.27191201e-01
-2.68175840e-01 2.97812045e-01 -4.27400768e-02 -4.70292449e-01
1.42391825e+00 1.69552974e-02 3.04888636e-01 3.00333679e-01
4.15552437e-01 3.95602779e-03 -1.80012846e+00 2.14453086e-01
-5.12281246e-02 -2.34995380e-01 2.04807028e-01 -1.17893195e+00
-1.16255569e+00 5.22218883e-01 2.84986079e-01 1.52878305e-02
9.48544085e-01 -3.72153729e-01 5.22949398e-01 9.01183963e-01
9.22195554e-01 -1.29132962e+00 8.41743946e-02 6.42389476e-01
6.90737844e-01 -1.18467224e+00 2.01613471e-01 5.26285946e-01
-7.43563771e-01 1.10105228e+00 8.51577520e-01 -1.90532327e-01
1.45054022e-02 3.89499843e-01 1.90635160e-01 -1.80788934e-01
-1.20041466e+00 -4.55280691e-01 -5.25818288e-01 4.36641335e-01
2.22968068e-02 9.57465619e-02 -1.87997729e-01 2.20745444e-01
-5.53043008e-01 1.80432662e-01 7.75457025e-01 1.47162485e+00
-3.81702632e-01 -1.24230897e+00 -3.29177439e-01 1.40215963e-01
-3.91444191e-02 1.38802186e-01 -1.22991376e-01 1.36403644e+00
2.37598211e-01 8.87223125e-01 -1.19452909e-01 -2.32173562e-01
2.58048117e-01 -2.76689529e-01 5.45692146e-01 -3.09162050e-01
-9.44891572e-01 -5.15272878e-02 3.03077072e-01 -8.13932419e-01
-3.22981894e-01 -8.28310907e-01 -1.58776784e+00 -7.70954788e-01
2.60238320e-01 3.50685984e-01 6.73961341e-01 8.70090842e-01
3.03466856e-01 6.57114446e-01 4.71083194e-01 -1.07921529e+00
-4.12478268e-01 -4.70960677e-01 -1.14170298e-01 -1.85025170e-01
2.60729045e-01 -9.46145415e-01 -3.64867657e-01 -3.36437732e-01] | [3.9503321647644043, 1.6331194639205933] |
e9227383-d97f-4934-904d-31456ecbf530 | grounding-hindsight-instructions-in-multi | 2204.04308 | null | https://arxiv.org/abs/2204.04308v2 | https://arxiv.org/pdf/2204.04308v2.pdf | Grounding Hindsight Instructions in Multi-Goal Reinforcement Learning for Robotics | This paper focuses on robotic reinforcement learning with sparse rewards for natural language goal representations. An open problem is the sample-inefficiency that stems from the compositionality of natural language, and from the grounding of language in sensory data and actions. We address these issues with three contributions. We first present a mechanism for hindsight instruction replay utilizing expert feedback. Second, we propose a seq2seq model to generate linguistic hindsight instructions. Finally, we present a novel class of language-focused learning tasks. We show that hindsight instructions improve the learning performance, as expected. In addition, we also provide an unexpected result: We show that the learning performance of our agent can be improved by one third if, in a sense, the agent learns to talk to itself in a self-supervised manner. We achieve this by learning to generate linguistic instructions that would have been appropriate as a natural language goal for an originally unintended behavior. Our results indicate that the performance gain increases with the task-complexity. | ['Stefan Wermter', 'Manfred Eppe', 'Frank Röder'] | 2022-04-08 | null | null | null | null | ['multi-goal-reinforcement-learning'] | ['methodology'] | [ 2.34184787e-01 6.94221973e-01 -7.29179978e-02 -3.74034792e-01
-9.34429765e-01 -6.98987365e-01 5.81295133e-01 -1.15861654e-01
-6.77350402e-01 1.08087015e+00 6.93173289e-01 -3.22944969e-01
-5.03525436e-02 -6.15722001e-01 -1.01140702e+00 -6.36856019e-01
-1.74491927e-01 3.26527625e-01 -1.17105350e-01 -5.76760709e-01
4.20225084e-01 3.99618208e-01 -1.51502776e+00 2.81322241e-01
6.34846509e-01 4.32288080e-01 4.86495286e-01 9.24973667e-01
6.46175304e-03 1.62858033e+00 -4.81308579e-01 1.65287450e-01
3.82736653e-01 -1.03848231e+00 -1.08191681e+00 1.81684598e-01
1.34116650e-01 -7.11793721e-01 -3.39171052e-01 9.05007422e-01
3.00890058e-01 3.87215227e-01 6.85000360e-01 -1.33640051e+00
-3.89975637e-01 1.08581710e+00 -2.40708455e-01 -2.96915025e-01
4.07113552e-01 6.24546468e-01 1.05483198e+00 -2.41190910e-01
6.86605811e-01 1.43736768e+00 1.75613642e-01 1.07575214e+00
-1.19748521e+00 -3.61525744e-01 3.77341956e-01 -1.76729500e-01
-6.30982816e-01 -4.19727236e-01 6.52819276e-01 -4.59389836e-01
1.05194247e+00 -1.42496049e-01 3.24773014e-01 1.09309971e+00
3.26997727e-01 9.76858675e-01 1.17244208e+00 -6.24813080e-01
5.06810129e-01 -5.04973456e-02 -1.24227092e-01 1.05209255e+00
-1.81047954e-02 7.21394181e-01 -4.86923307e-01 1.26224965e-01
7.20685005e-01 -3.12015325e-01 -9.12391022e-02 -7.87560523e-01
-1.25984836e+00 1.11233342e+00 5.12884259e-01 2.33764857e-01
-5.64024627e-01 6.77250981e-01 4.84049141e-01 7.65787184e-01
-2.38124147e-01 1.04012692e+00 -6.02384686e-01 -2.12933436e-01
-1.80520236e-01 3.99450213e-01 8.93035293e-01 8.47815752e-01
9.23257530e-01 3.87379766e-01 -5.50653189e-02 5.14960349e-01
1.16378725e-01 5.18892944e-01 7.49715269e-01 -1.76069677e+00
4.48317379e-01 2.37825364e-01 3.85557353e-01 -4.43933576e-01
-5.58577895e-01 -6.70251027e-02 -1.98287800e-01 8.43526840e-01
7.01340020e-01 -7.12944686e-01 -5.46414137e-01 2.28098726e+00
-7.15550035e-02 -5.40756881e-01 6.90538704e-01 8.78993094e-01
2.24245802e-01 6.00557208e-01 9.27735120e-02 -2.35196084e-01
8.69697511e-01 -1.17716181e+00 -4.85217541e-01 -6.91826761e-01
1.14691067e+00 -2.93968201e-01 1.47119892e+00 4.98925567e-01
-9.09444511e-01 -3.45803171e-01 -8.13978553e-01 1.17406890e-01
9.83537510e-02 -6.99695498e-02 8.06142867e-01 7.13154227e-02
-1.08038294e+00 5.49077988e-01 -7.67015100e-01 -5.79337001e-01
-1.51661104e-02 3.38285357e-01 -3.44309539e-01 1.36111915e-01
-8.02038789e-01 1.00745130e+00 4.09645081e-01 -5.58884501e-01
-1.33076692e+00 -1.33619592e-01 -1.07841766e+00 1.16748922e-01
6.02252126e-01 -5.06659687e-01 2.11968851e+00 -1.32601368e+00
-1.92024326e+00 6.04387343e-01 7.76235983e-02 -7.57754266e-01
2.85451889e-01 -1.36253491e-01 2.74374813e-01 1.62629604e-01
4.00629342e-01 9.01557386e-01 4.45387959e-01 -1.28695059e+00
-6.28762841e-01 -4.79940146e-01 3.85772407e-01 4.79999572e-01
9.75789130e-02 -2.52087682e-01 4.47271049e-01 -4.93864417e-01
-4.63737041e-01 -1.05763328e+00 -6.42156720e-01 -1.74177408e-01
-7.93086812e-02 -3.00743133e-01 1.12071313e-01 -6.81532472e-02
5.02550602e-01 -2.17411423e+00 2.20121711e-01 -1.61814079e-01
5.34472689e-02 -1.51233554e-01 -5.90804040e-01 6.72425210e-01
4.34541367e-02 -2.64309883e-01 -2.57270575e-01 2.61714868e-02
1.82265595e-01 5.14870763e-01 -6.84553802e-01 3.24900389e-01
1.35412514e-01 9.94908512e-01 -1.17809784e+00 -2.80065536e-01
-5.73875271e-02 -1.34161308e-01 -9.60926056e-01 5.36347926e-01
-6.81804359e-01 6.03387833e-01 -6.63827002e-01 5.58043458e-03
-8.42172578e-02 3.08709200e-02 2.90896177e-01 5.01577199e-01
-2.23708093e-01 5.56035757e-01 -7.88659096e-01 1.84923387e+00
-6.43968999e-01 3.71767312e-01 1.65654868e-01 -1.15767908e+00
7.20418215e-01 1.01855859e-01 3.26758713e-01 -7.45001733e-01
1.54744267e-01 3.86943281e-01 2.34470144e-01 -6.49285018e-01
3.32495034e-01 -4.52299386e-01 -4.45136249e-01 8.37172031e-01
1.05708815e-01 -4.80111867e-01 1.05044298e-01 2.28042558e-01
1.13646233e+00 6.21778965e-01 6.78730190e-01 -2.60703564e-01
1.55114755e-01 5.30745268e-01 3.23442519e-01 1.14319146e+00
-2.52000183e-01 -5.76153472e-02 7.75660932e-01 -4.48958516e-01
-8.26285660e-01 -8.87013257e-01 6.51116610e-01 1.57472777e+00
-1.41194180e-01 -4.56540473e-02 -6.88180685e-01 -7.54550576e-01
-8.41133893e-02 1.27449858e+00 -6.09228909e-01 -4.11351383e-01
-8.35191190e-01 -3.61440927e-02 4.60790277e-01 5.98308444e-01
1.65766761e-01 -1.69747877e+00 -1.34309375e+00 2.11263463e-01
1.08678313e-02 -8.47537458e-01 -5.83021283e-01 8.25716674e-01
-8.60508502e-01 -1.00958812e+00 -5.07167399e-01 -9.77581620e-01
6.46297634e-01 1.80942237e-01 9.83134508e-01 -7.00742677e-02
9.83300880e-02 8.21424067e-01 -3.20065439e-01 -3.08725923e-01
-8.07577968e-01 -1.41163081e-01 3.14580470e-01 -5.29838324e-01
3.02608907e-01 -4.79168326e-01 -2.35220805e-01 -1.63492516e-01
-6.35332465e-01 6.25737570e-03 7.06020355e-01 9.65470016e-01
1.56747267e-01 -3.23844731e-01 7.90792227e-01 -7.87925541e-01
9.34699118e-01 -3.22762251e-01 -7.98225224e-01 4.36310796e-03
-5.51887572e-01 9.79958951e-01 1.00548291e+00 -4.26838845e-01
-1.07132173e+00 6.51592612e-01 7.52724409e-02 1.41365603e-01
-3.56927961e-01 2.52180696e-01 1.11653909e-01 1.69970810e-01
1.18739450e+00 4.76873040e-01 4.41852033e-01 -1.43774673e-01
6.73985660e-01 5.70543051e-01 3.86603415e-01 -9.66292381e-01
3.93458068e-01 1.61386862e-01 -4.52789105e-02 -6.76110327e-01
-8.74917388e-01 -9.53000188e-02 -2.28408337e-01 2.61913948e-02
8.29269171e-01 -8.67765248e-01 -1.22428310e+00 -1.91833694e-02
-1.17469656e+00 -1.12092173e+00 -6.04900539e-01 7.06106246e-01
-1.60464644e+00 2.40230441e-01 -5.62440336e-01 -1.00624573e+00
-5.53510599e-02 -1.21768224e+00 8.19393158e-01 9.94803533e-02
-3.41000080e-01 -5.37851214e-01 1.05701424e-01 -1.26422539e-01
2.59474218e-01 -7.16998354e-02 7.84583509e-01 -7.63098121e-01
-4.10732806e-01 4.56346750e-01 8.17074105e-02 1.40389115e-01
2.04543203e-01 -6.34755552e-01 -6.14165664e-01 -2.45081410e-01
1.45035371e-01 -1.11605775e+00 6.48584843e-01 3.04952562e-01
7.65425205e-01 -7.47069359e-01 1.34007558e-02 1.83664024e-01
1.27682686e+00 5.16980290e-01 1.94964185e-01 3.52387071e-01
1.84333071e-01 9.76192474e-01 6.39064670e-01 4.71490562e-01
3.35721016e-01 4.07130301e-01 5.38227916e-01 3.05495530e-01
-3.57569046e-02 -7.37706065e-01 7.91735411e-01 2.57838577e-01
2.84291357e-01 4.68146615e-02 -7.07938790e-01 5.89419901e-01
-2.08124137e+00 -1.02903223e+00 2.41985530e-01 2.06865740e+00
9.55308974e-01 -8.79809782e-02 4.84400094e-01 -2.85608649e-01
1.75000519e-01 -1.47371277e-01 -8.18282187e-01 -7.70032167e-01
1.54059798e-01 -1.70392394e-02 6.43082201e-01 1.03848720e+00
-6.04753792e-01 1.28539252e+00 6.88677025e+00 1.65717885e-01
-9.91453290e-01 -2.26631030e-01 2.60332406e-01 -9.88525972e-02
-2.54506826e-01 -8.74312520e-02 -5.81689119e-01 2.90006157e-02
8.90366435e-01 -4.36706662e-01 9.57491577e-01 1.09910274e+00
3.52738678e-01 -1.86241269e-01 -1.49907541e+00 6.42451227e-01
-7.68668298e-03 -9.76684749e-01 -1.19956583e-01 1.71534009e-02
5.67122221e-01 1.54593363e-01 -2.23043621e-01 7.14575708e-01
1.21321821e+00 -9.23527002e-01 6.91773951e-01 1.59882084e-01
4.99369740e-01 -7.89401472e-01 3.15520853e-01 9.46364224e-01
-5.36387265e-01 -6.01442099e-01 -3.10486972e-01 -6.73121870e-01
-5.27395606e-02 -1.88838303e-01 -1.06843758e+00 1.00208856e-02
1.31058753e-01 2.68698514e-01 -4.36303429e-02 6.05157495e-01
-6.67751968e-01 3.93296868e-01 -1.67647731e-02 -5.57719290e-01
7.56232262e-01 -8.69766548e-02 3.00008059e-01 9.44316864e-01
1.92101747e-01 1.81076840e-01 5.14498532e-01 8.29839230e-01
1.17492065e-01 8.59420374e-03 -1.30804050e+00 -2.24240944e-01
9.40582380e-02 7.88617671e-01 -2.76450515e-01 -1.77869380e-01
-2.23100364e-01 1.01153803e+00 7.38233149e-01 3.46649408e-01
-3.98351759e-01 -3.47540259e-01 5.40792108e-01 -2.24132657e-01
1.64971635e-01 -2.05783755e-01 -2.32950673e-02 -9.78218257e-01
-2.10874289e-01 -1.10175204e+00 1.49890065e-01 -8.27393949e-01
-1.02390933e+00 4.69346821e-01 -1.34147510e-01 -9.76885438e-01
-1.16439462e+00 -6.53050303e-01 -1.93196833e-01 5.12326241e-01
-1.34029686e+00 -5.98818719e-01 1.88356847e-01 4.93862957e-01
8.12551260e-01 -3.47196460e-01 1.03196216e+00 -3.17277193e-01
2.08017305e-01 3.17302018e-01 -1.30845487e-01 9.08857286e-02
6.07611179e-01 -1.38125265e+00 4.38862219e-02 5.70147932e-01
6.52602762e-02 5.62590957e-01 8.91261160e-01 -3.70271116e-01
-1.46808541e+00 -8.79248679e-01 8.29935014e-01 -2.76177406e-01
7.02518463e-01 -2.71335006e-01 -3.35326374e-01 1.11547279e+00
4.15969998e-01 -3.19896609e-01 4.52914447e-01 -8.44430029e-02
-4.35157448e-01 8.73439014e-02 -1.13696432e+00 9.65708673e-01
1.05678391e+00 -3.94702464e-01 -1.03504145e+00 4.95330513e-01
1.18460226e+00 -1.78312406e-01 -8.01441669e-02 -1.40126094e-01
4.32829231e-01 -7.96486139e-01 5.13451099e-01 -1.19276357e+00
7.73752153e-01 -2.03515962e-01 -3.88378948e-01 -1.69582784e+00
-2.73102850e-01 -8.09885859e-01 2.28714973e-01 6.29591644e-01
3.51273060e-01 -6.06158853e-01 7.48561740e-01 3.81622195e-01
-2.04868764e-01 -3.26110154e-01 -6.28076017e-01 -9.44253206e-01
5.58538675e-01 -1.61942244e-01 2.82375306e-01 5.50462961e-01
7.11851537e-01 8.64423394e-01 -5.27463496e-01 -2.63218760e-01
4.68790025e-01 2.27664441e-01 8.85039687e-01 -7.97618747e-01
-5.93954504e-01 -3.96906972e-01 1.26645207e-01 -1.58811474e+00
5.30119479e-01 -9.35391486e-01 7.76643574e-01 -1.35043800e+00
3.55058536e-02 -1.38196275e-01 1.81136448e-02 6.70858443e-01
2.35970452e-01 -3.97612810e-01 4.78119612e-01 -1.02430135e-01
-6.92816079e-01 5.80150187e-01 1.43598962e+00 -4.48288433e-02
-4.23861206e-01 2.32332125e-02 -1.20538294e+00 7.58014560e-01
1.20468020e+00 -4.71314192e-01 -6.62344098e-01 -6.64029598e-01
2.24217892e-01 4.72779095e-01 2.40355417e-01 -6.18251026e-01
2.35540479e-01 -6.04466319e-01 -1.98434949e-01 -9.30846110e-02
6.38210699e-02 -8.01616967e-01 -6.39221132e-01 1.02839923e+00
-1.07422149e+00 1.60337687e-01 1.22315861e-01 5.60525298e-01
1.17530689e-01 -4.13826615e-01 8.47092628e-01 -6.53739870e-01
-7.18554020e-01 -3.04898769e-01 -1.14085376e+00 2.08896875e-01
1.01590228e+00 1.84557423e-01 -1.87783718e-01 -8.19372892e-01
-5.09047389e-01 4.60435420e-01 3.58587772e-01 3.39479715e-01
4.76558596e-01 -1.11012578e+00 -7.49083877e-01 5.01530841e-02
2.52155036e-01 -2.17731416e-01 -2.67305166e-01 3.62984598e-01
-2.43923500e-01 5.51684260e-01 -4.69997764e-01 -2.05807388e-01
-7.85508335e-01 8.37266266e-01 4.48075324e-01 -2.11881980e-01
-5.71288526e-01 6.24046266e-01 5.12189507e-01 -8.64604115e-01
4.61758971e-01 -4.06899065e-01 -5.09429723e-02 -4.26960588e-01
5.52675664e-01 -9.18143690e-02 -4.51557100e-01 -1.92150503e-01
-1.62161723e-01 2.33754918e-01 7.06292018e-02 -5.12385607e-01
1.45900178e+00 2.18641702e-02 1.84149593e-01 5.54825604e-01
9.10053134e-01 1.34782912e-02 -1.44900644e+00 -2.34496042e-01
1.66029438e-01 -2.03448888e-02 -4.37008113e-01 -1.04436612e+00
-5.98638773e-01 7.43519545e-01 1.83190778e-01 -5.39798439e-02
8.39989901e-01 1.52045667e-01 3.87284368e-01 1.20497811e+00
7.17163742e-01 -1.11304665e+00 4.57571834e-01 8.34911466e-01
1.09243476e+00 -1.28008008e+00 -4.19864535e-01 2.92138875e-01
-1.05193651e+00 1.00342298e+00 8.23799551e-01 -2.94393361e-01
-7.91734904e-02 3.05202872e-01 1.45593271e-01 -1.85234204e-01
-9.54275072e-01 -2.98604846e-01 -4.59893435e-01 8.53219032e-01
5.30337751e-01 1.03353962e-01 -3.25001299e-01 4.25526202e-01
-5.40394068e-01 1.41698435e-01 8.08522642e-01 9.18018222e-01
-9.32337582e-01 -1.02675116e+00 -2.75724471e-01 1.29629225e-01
-2.29351908e-01 7.28475377e-02 -6.91469252e-01 6.49217248e-01
-3.48537236e-01 1.10399950e+00 -1.00754172e-01 -1.94450170e-01
2.37336352e-01 1.09011166e-01 7.07831621e-01 -9.34369445e-01
-4.08250779e-01 -2.46436507e-01 1.15678579e-01 -7.91137040e-01
-1.23558082e-01 -4.80660945e-01 -1.81525469e+00 9.40732881e-02
2.64002383e-01 2.72895724e-01 4.05653805e-01 9.14942145e-01
2.19783857e-01 4.79817957e-01 6.88849211e-01 -5.90648115e-01
-1.36588407e+00 -7.24330008e-01 -4.16887552e-01 1.97751507e-01
7.10791767e-01 -3.83752972e-01 -4.95838374e-01 -4.85331826e-02] | [4.095640182495117, 1.5034393072128296] |
4f3377bc-4499-4d22-9e9a-fd5c0b55ccbe | joint-neural-architecture-and-hyperparameter | 2211.16126 | null | https://arxiv.org/abs/2211.16126v2 | https://arxiv.org/pdf/2211.16126v2.pdf | Joint Neural Architecture and Hyperparameter Search for Correlated Time Series Forecasting | Sensors in cyber-physical systems often capture interconnected processes and thus emit correlated time series (CTS), the forecasting of which enables important applications. The key to successful CTS forecasting is to uncover the temporal dynamics of time series and the spatial correlations among time series. Deep learning-based solutions exhibit impressive performance at discerning these aspects. In particular, automated CTS forecasting, where the design of an optimal deep learning architecture is automated, enables forecasting accuracy that surpasses what has been achieved by manual approaches. However, automated CTS solutions remain in their infancy and are only able to find optimal architectures for predefined hyperparameters and scale poorly to large-scale CTS. To overcome these limitations, we propose SEARCH, a joint, scalable framework, to automatically devise effective CTS forecasting models. Specifically, we encode each candidate architecture and accompanying hyperparameters into a joint graph representation. We introduce an efficient Architecture-Hyperparameter Comparator (AHC) to rank all architecture-hyperparameter pairs, and we then further evaluate the top-ranked pairs to select a final result. Extensive experiments on six benchmark datasets demonstrate that SEARCH not only eliminates manual efforts but also is capable of better performance than manually designed and existing automatically designed CTS models. In addition, it shows excellent scalability to large CTS. | ['Christian S. Jensen', 'Bin Yang', 'Chenjuan Guo', 'Miao Zhang', 'Dalin Zhang', 'Xinle Wu'] | 2022-11-29 | null | null | null | null | ['correlated-time-series-forecasting'] | ['time-series'] | [-1.43650351e-02 -3.68232548e-01 -8.75373855e-02 -1.54929325e-01
-6.77591085e-01 -8.29138696e-01 6.42395377e-01 1.49726138e-01
2.52814233e-01 3.53479832e-01 7.10425004e-02 -6.47328913e-01
-6.74199104e-01 -7.56972134e-01 -3.64501774e-01 -7.09056199e-01
-6.79824412e-01 3.81106913e-01 1.70401827e-01 -1.23837993e-01
2.19115287e-01 5.44734120e-01 -1.20461047e+00 -1.27815783e-01
8.82322252e-01 1.31517172e+00 4.02967148e-02 6.90970480e-01
-1.93656206e-01 6.76230013e-01 -5.90991139e-01 -6.24947995e-02
2.61748999e-01 -3.20148170e-01 -2.47609511e-01 -1.25294983e-01
-4.46051545e-02 9.73025337e-02 -3.55304837e-01 6.08811438e-01
8.01743343e-02 -9.32919607e-02 5.64604521e-01 -1.44539678e+00
-3.45737368e-01 6.68844044e-01 -5.06723464e-01 3.17370683e-01
2.20204797e-02 4.57134366e-01 1.13250959e+00 -4.18396175e-01
1.55605465e-01 1.01585758e+00 8.70676577e-01 -8.42017774e-03
-1.30235422e+00 -5.47807872e-01 3.59030604e-01 -8.88036489e-02
-1.30161273e+00 -2.36538261e-01 1.02533305e+00 -4.79995161e-01
1.20565939e+00 2.26888344e-01 9.37402070e-01 1.03908813e+00
7.27436721e-01 4.85146612e-01 7.85181463e-01 8.97910148e-02
4.62761760e-01 -5.99824846e-01 9.86609831e-02 6.16458774e-01
2.55791813e-01 6.91817626e-02 -4.13819075e-01 -2.87427396e-01
6.88402474e-01 2.24980101e-01 -1.91697404e-01 -1.56006977e-01
-1.40377498e+00 5.66192389e-01 5.23843169e-01 4.59141612e-01
-6.45136476e-01 4.81186390e-01 3.06473166e-01 4.44007993e-01
2.79008359e-01 9.20286477e-01 -7.61514068e-01 -1.98692203e-01
-6.96116447e-01 7.30062351e-02 8.58143449e-01 7.48204410e-01
5.50891817e-01 4.52037215e-01 1.05589842e-02 4.23270345e-01
4.52982336e-02 5.02688348e-01 3.74071211e-01 -7.60023952e-01
4.70899463e-01 9.90485668e-01 -2.02392135e-02 -1.31966603e+00
-8.79715562e-01 -8.15617085e-01 -1.22469366e+00 -4.50528085e-01
9.56690460e-02 -2.60955751e-01 -8.26410949e-01 1.63551819e+00
2.59448349e-01 4.21201378e-01 -1.28188748e-02 6.67954683e-01
3.10637951e-01 9.50032949e-01 1.04665376e-01 -2.40533516e-01
1.12044978e+00 -5.81433713e-01 -5.66774666e-01 -2.94901103e-01
5.05800426e-01 -4.90927100e-01 1.06514895e+00 3.33234847e-01
-6.24772847e-01 -2.80152619e-01 -1.28448665e+00 4.87670839e-01
-3.97450835e-01 2.39094403e-02 8.78501892e-01 3.44674915e-01
-1.06325483e+00 8.07794154e-01 -1.28166997e+00 -1.63057119e-01
2.48307675e-01 4.43303794e-01 1.86988652e-01 2.34589726e-01
-1.21681261e+00 6.37711406e-01 2.57472515e-01 2.49535933e-01
-8.02414536e-01 -9.02311683e-01 -5.14654100e-01 2.64166206e-01
4.50362295e-01 -6.79624021e-01 1.13148403e+00 -6.09554648e-01
-1.42352152e+00 -1.88800722e-01 7.44394585e-02 -6.43372059e-01
1.80572554e-01 1.20344378e-01 -8.56420457e-01 4.32377234e-02
-2.78769135e-01 2.12610960e-01 9.06569004e-01 -9.50917304e-01
-4.51770931e-01 -2.31787637e-01 -2.09423050e-01 -1.78345010e-01
-6.99303031e-01 -3.78653198e-01 -5.19863188e-01 -6.47517741e-01
2.19124839e-01 -1.06999564e+00 -5.69732845e-01 -3.80472451e-01
-4.97909635e-01 -2.89751679e-01 8.94399285e-01 -4.81363088e-01
1.59039354e+00 -1.92771018e+00 4.37177792e-02 5.38068712e-01
6.42012775e-01 1.06615722e-01 -3.68346572e-01 9.18173492e-01
1.54952794e-01 -2.97303237e-02 -9.17716101e-02 -2.55525619e-01
5.67920245e-02 1.50033891e-01 -6.01495028e-01 3.76005322e-01
3.01085144e-01 1.07994366e+00 -8.58154356e-01 -1.23283692e-01
2.38904521e-01 3.04945499e-01 -4.92775477e-02 4.53026593e-02
-4.90987301e-01 3.07118714e-01 -6.72115803e-01 7.26955891e-01
2.16084465e-01 -9.58218277e-01 5.38667738e-01 -4.14973885e-01
-6.85163289e-02 2.53457695e-01 -9.11867857e-01 9.69003439e-01
-4.85300869e-01 7.05315590e-01 -2.49801651e-01 -9.15914178e-01
1.25877178e+00 1.12141170e-01 9.89485919e-01 -7.48641729e-01
2.20226645e-01 2.28348166e-01 3.26052271e-02 -2.11655930e-01
3.79876018e-01 3.73216838e-01 -2.59583414e-01 4.82473105e-01
-3.41578722e-01 -7.48162717e-02 6.05529435e-02 1.67237464e-02
1.67163408e+00 -4.27887022e-01 2.79812396e-01 -1.77791432e-01
2.05810159e-01 1.48571402e-01 5.07752419e-01 5.04828215e-01
-2.09644791e-02 2.65682817e-01 7.55982876e-01 -8.75504911e-01
-1.03496110e+00 -8.59095335e-01 3.28973889e-01 7.83144295e-01
5.15127033e-02 -6.59074605e-01 -3.15197468e-01 -5.89400530e-01
2.15233773e-01 5.80787778e-01 -5.60486794e-01 -2.68372923e-01
-6.03030682e-01 -7.75226057e-01 3.65716726e-01 7.30176389e-01
2.79229492e-01 -7.99904764e-01 -6.45078480e-01 5.09045720e-01
1.91897079e-02 -1.20236301e+00 -2.90063113e-01 3.09101462e-01
-9.72226918e-01 -1.13984561e+00 -2.25246355e-01 -3.29264700e-01
3.89050007e-01 3.12010735e-01 1.10253918e+00 -4.20155982e-03
-1.09131500e-01 1.95466563e-01 -1.53926000e-01 -4.25813258e-01
-2.26187691e-01 2.61346310e-01 -3.14481300e-03 -2.03891601e-02
3.05212647e-01 -7.32362151e-01 -7.44735599e-01 4.03757066e-01
-8.71160984e-01 -1.96777195e-01 7.39128351e-01 5.80994308e-01
5.53490341e-01 3.52956504e-01 7.52099812e-01 -2.84397721e-01
9.16030347e-01 -6.09974027e-01 -9.98733044e-01 4.20269340e-01
-9.98825550e-01 1.87722042e-01 9.96904612e-01 -6.12782776e-01
-3.09654117e-01 1.77512303e-01 2.76956379e-01 -7.63548851e-01
9.21031609e-02 1.06321037e+00 2.03433558e-01 3.31799909e-02
4.98243600e-01 1.45183668e-01 -4.48630713e-02 -3.12252522e-01
1.41951218e-01 2.05216631e-01 3.22844297e-01 -4.28578943e-01
7.23277271e-01 3.12421530e-01 2.13318899e-01 -7.29203880e-01
-6.51016355e-01 -3.88427109e-01 -4.05663997e-01 -3.74026686e-01
5.20514131e-01 -8.06996405e-01 -8.07605386e-01 2.43334696e-01
-9.93755996e-01 -4.02176827e-01 -3.22688483e-02 3.79707694e-01
-2.44327843e-01 7.61469314e-03 -4.48503971e-01 -7.46316314e-01
-3.63639176e-01 -9.73509073e-01 1.26246750e+00 -1.03445038e-01
-4.61340636e-01 -1.18157542e+00 2.26047710e-01 -1.51868373e-01
5.50916493e-01 4.97384399e-01 1.09870911e+00 -6.74452305e-01
-7.68869936e-01 -4.31461394e-01 -1.04294717e-01 -1.37418315e-01
1.93063676e-01 3.84178519e-01 -6.04322553e-01 -3.70705932e-01
-2.52616554e-01 1.97850332e-01 7.04126954e-01 5.06547153e-01
1.17919028e+00 -3.08782011e-01 -5.73630929e-01 4.90281343e-01
1.37860382e+00 4.11175162e-01 2.85356760e-01 1.60516292e-01
8.06111991e-01 4.99547213e-01 3.16123188e-01 5.74788332e-01
5.74856102e-01 5.75819612e-01 5.09170532e-01 -5.46532199e-02
3.18345994e-01 -2.14061454e-01 2.87467241e-01 1.26067293e+00
6.96952641e-02 -4.45530981e-01 -1.44935465e+00 5.36064506e-01
-1.94291437e+00 -6.82560861e-01 -1.52179420e-01 1.94857669e+00
3.29759032e-01 3.92710835e-01 1.59955084e-01 2.76239336e-01
3.98157746e-01 2.82432616e-01 -1.03659832e+00 -6.57801703e-02
4.98310057e-03 -2.70526186e-02 6.71415567e-01 8.19368064e-02
-1.14068890e+00 7.42228806e-01 6.62414980e+00 4.44322228e-01
-1.38505507e+00 -1.96939260e-01 4.83118057e-01 -1.11074694e-01
-4.18485254e-01 -4.04778495e-02 -4.66131121e-01 4.02309567e-01
1.44683552e+00 -2.77898997e-01 5.34949064e-01 7.10877836e-01
4.43683982e-01 4.68465954e-01 -1.03768992e+00 9.70026195e-01
-4.73560691e-01 -1.58892798e+00 -8.53699744e-02 2.93504864e-01
7.67087162e-01 1.99167669e-01 1.39665246e-01 2.13520721e-01
6.47627711e-01 -9.27167833e-01 4.28710371e-01 5.11849701e-01
3.94744933e-01 -6.45226717e-01 5.16446710e-01 8.31534490e-02
-1.72839785e+00 -6.04491711e-01 -1.07292816e-01 5.65794222e-02
9.23954770e-02 8.78432631e-01 -9.56034660e-01 5.92358470e-01
6.17795646e-01 8.68172348e-01 -4.46289718e-01 9.90168273e-01
7.92453960e-02 8.20095897e-01 -4.29751486e-01 -4.67459977e-01
3.91737074e-01 -2.06922734e-04 4.77239817e-01 9.24434364e-01
6.95815861e-01 2.80097891e-02 3.82185608e-01 6.19537950e-01
2.66177386e-01 -8.34955350e-02 -7.93165863e-01 -6.59307599e-01
9.42092061e-01 1.08276176e+00 -9.38419044e-01 -2.43748456e-01
-3.16027313e-01 3.38525712e-01 2.41598323e-01 4.92238134e-01
-7.71634281e-01 -1.03003502e-01 8.34738135e-01 4.32534469e-03
3.85185450e-01 -7.93802559e-01 -4.81106460e-01 -9.05118287e-01
-3.24123390e-02 -9.43813562e-01 5.76805234e-01 -4.72831815e-01
-1.37577856e+00 7.33559489e-01 -2.40296900e-01 -1.64588892e+00
-5.01357138e-01 -3.32447648e-01 -7.10076094e-01 3.98601145e-01
-1.08949959e+00 -1.06150091e+00 -4.70586926e-01 5.07859170e-01
4.78024691e-01 -1.61334544e-01 6.40278220e-01 -5.70752309e-04
-9.08220589e-01 2.71388113e-01 2.36838177e-01 -2.30241150e-01
2.15136632e-01 -1.16137981e+00 7.58304417e-01 8.07323456e-01
1.16899870e-01 4.90285426e-01 8.32968295e-01 -7.96437681e-01
-2.10677552e+00 -1.34605634e+00 6.42559290e-01 -2.91229606e-01
1.28551543e+00 -3.93777102e-01 -9.52524841e-01 4.44643557e-01
7.38783032e-02 -3.11905682e-01 5.38154602e-01 2.92568982e-01
-3.79337072e-01 -5.13469994e-01 -5.14988720e-01 7.37885237e-01
8.13559711e-01 -3.48531216e-01 -2.24860653e-01 4.10168409e-01
9.25544739e-01 -2.43978295e-02 -1.12725091e+00 4.86942708e-01
4.59995419e-01 -5.64154148e-01 8.69014084e-01 -4.63783443e-01
3.13294768e-01 -2.64477074e-01 -1.05653115e-01 -1.48654735e+00
-4.13883984e-01 -9.44360971e-01 -4.56814170e-01 9.22077358e-01
5.78661859e-01 -8.13473761e-01 5.30294478e-01 5.79495668e-01
-2.58111745e-01 -9.73668635e-01 -5.91925740e-01 -9.99177337e-01
-3.54733258e-01 -6.68450236e-01 1.27302194e+00 1.13687325e+00
-9.22690108e-02 2.75368214e-01 -2.06101224e-01 5.40260434e-01
7.07382441e-01 5.43790758e-01 7.71468043e-01 -1.50494409e+00
-1.52972907e-01 -8.69292498e-01 -3.81619096e-01 -7.25981593e-01
-5.25555909e-02 -3.29348803e-01 1.70083274e-03 -1.48351514e+00
-4.25519437e-01 -7.17690885e-01 -5.93994260e-01 4.80983794e-01
-2.47323867e-02 -2.90043354e-01 -5.42858429e-02 5.41360915e-01
-5.52279413e-01 6.36428833e-01 1.05860627e+00 -1.68087646e-01
-4.48317260e-01 4.39515449e-02 -4.91808712e-01 4.37965572e-01
1.05121696e+00 -2.24088311e-01 -7.56065369e-01 -2.90090948e-01
2.62869984e-01 3.69164139e-01 3.91722977e-01 -1.09338260e+00
4.26509619e-01 -4.38280374e-01 2.67593473e-01 -8.01707745e-01
2.71656692e-01 -9.21380937e-01 5.20963013e-01 4.35708791e-01
-3.24034542e-01 6.46900833e-01 1.54654652e-01 8.92014086e-01
-2.54945070e-01 4.73488957e-01 2.94374764e-01 2.13073671e-01
-7.10828900e-01 6.48234248e-01 -3.39994758e-01 -2.91668981e-01
9.43239391e-01 4.90967929e-02 -5.73286295e-01 -5.70064425e-01
-3.61987472e-01 4.19780850e-01 1.56679422e-01 6.74101651e-01
5.67642868e-01 -1.26880455e+00 -2.72290379e-01 1.37870207e-01
1.37772113e-01 -1.93162948e-01 1.21333711e-01 9.24435139e-01
-2.56294131e-01 6.08564079e-01 7.75204897e-02 -7.98588514e-01
-8.26190531e-01 6.87362671e-01 2.83295155e-01 -4.69447374e-01
-7.15729833e-01 2.97962189e-01 -9.02768224e-02 -1.80429861e-01
2.07455799e-01 -7.43240595e-01 -1.30917639e-01 -1.07287042e-01
3.23776782e-01 3.55848879e-01 2.72782475e-01 -1.56891316e-01
-3.57751399e-01 5.79129040e-01 2.41993725e-01 1.67565688e-01
1.61366975e+00 -6.78260028e-02 8.29725266e-02 4.03655499e-01
1.22604370e+00 -3.78334135e-01 -1.36866391e+00 -1.43981397e-01
3.51307154e-01 -2.18643859e-01 1.76500440e-01 -5.33002734e-01
-1.28927243e+00 6.23743832e-01 2.71620482e-01 9.81528759e-01
1.31867588e+00 -1.10953197e-01 9.88947034e-01 5.93089640e-01
3.42446774e-01 -8.79710674e-01 2.60491580e-01 5.35460114e-01
8.32920492e-01 -9.46637213e-01 -7.01672137e-02 -2.29264826e-01
-5.12843192e-01 1.20027471e+00 5.68245649e-01 -7.31878877e-02
9.01059866e-01 5.20665348e-01 -4.21609357e-02 -6.62578583e-01
-1.39958835e+00 -6.22626394e-02 5.28235435e-01 3.42559457e-01
2.08307728e-01 2.81753957e-01 4.90918867e-02 3.54232639e-01
-3.31657708e-01 -2.65310735e-01 1.56436563e-01 8.99285853e-01
-2.85172909e-01 -7.80892015e-01 -2.35590339e-01 7.94176400e-01
4.53574321e-04 2.50929624e-01 -4.48289603e-01 6.93934321e-01
-5.65491080e-01 1.06098664e+00 2.35331163e-01 -9.74249125e-01
4.30047840e-01 -2.39353240e-01 -4.49770354e-02 3.77945006e-02
-7.05528319e-01 3.49950612e-01 1.63268402e-01 -8.51206601e-01
-2.47308388e-02 -6.04870021e-01 -1.00675011e+00 -4.21632141e-01
-1.39352292e-01 -2.01700907e-02 7.95239151e-01 9.55107987e-01
8.25235248e-01 8.45313072e-01 1.03312910e+00 -7.30279326e-01
-6.46237135e-01 -6.39826179e-01 -2.44155854e-01 -2.67441780e-03
4.57573652e-01 -7.12613404e-01 -2.94761032e-01 -3.04351687e-01] | [6.989395618438721, 2.947340726852417] |
d96a36ea-6e22-4d27-8ac4-53e04fdb6199 | unified-extraction-of-health-condition | null | null | https://aclanthology.org/N12-2005 | https://aclanthology.org/N12-2005.pdf | Unified Extraction of Health Condition Descriptions | null | ['Ivelina Nikolova'] | 2012-06-01 | null | null | null | naacl-2012-6 | ['negation-detection'] | ['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] | [-7.265807628631592, 3.6932621002197266] |
0bef1567-5318-4c13-8c31-035e906f35f9 | localized-vision-language-matching-for-open | 2205.06160 | null | https://arxiv.org/abs/2205.06160v2 | https://arxiv.org/pdf/2205.06160v2.pdf | Localized Vision-Language Matching for Open-vocabulary Object Detection | In this work, we propose an open-vocabulary object detection method that, based on image-caption pairs, learns to detect novel object classes along with a given set of known classes. It is a two-stage training approach that first uses a location-guided image-caption matching technique to learn class labels for both novel and known classes in a weakly-supervised manner and second specializes the model for the object detection task using known class annotations. We show that a simple language model fits better than a large contextualized language model for detecting novel objects. Moreover, we introduce a consistency-regularization technique to better exploit image-caption pair information. Our method compares favorably to existing open-vocabulary detection approaches while being data-efficient. Source code is available at https://github.com/lmb-freiburg/locov . | ['Thomas Brox', 'Sudhanshu Mittal', 'Maria A. Bravo'] | 2022-05-12 | null | null | null | null | ['open-vocabulary-object-detection', 'open-vocabulary-attribute-detection', 'open-world-object-detection'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 2.66366839e-01 2.02783018e-01 -4.55329627e-01 -4.26600724e-01
-1.46940005e+00 -8.55232179e-01 6.92652166e-01 4.18699354e-01
-5.38550913e-01 4.57260489e-01 3.38888355e-02 -1.58829749e-01
2.52164215e-01 -4.90595043e-01 -1.08145130e+00 -4.56548095e-01
1.37006059e-01 7.25388169e-01 5.00945628e-01 2.03781545e-01
3.03009450e-01 3.81924510e-01 -1.57921076e+00 4.93286222e-01
2.24415526e-01 7.30795860e-01 5.99764824e-01 7.62628913e-01
-8.04525539e-02 6.13241255e-01 4.91206208e-03 -2.93603897e-01
4.24562812e-01 -2.00202033e-01 -9.98312175e-01 2.12454706e-01
1.07412994e+00 -1.67228580e-01 -3.02772462e-01 1.02493143e+00
2.91381001e-01 -2.73572654e-03 6.17334008e-01 -1.02666783e+00
-7.17986763e-01 1.53949544e-01 -5.82271159e-01 4.29767787e-01
3.81454974e-01 1.02716021e-01 1.07761717e+00 -1.30659175e+00
9.05157447e-01 1.02479315e+00 6.07385278e-01 7.33243585e-01
-1.33311069e+00 -6.53817952e-01 2.47786865e-01 2.18801111e-01
-1.72122920e+00 -3.83103877e-01 6.03729367e-01 -6.68273926e-01
6.69285715e-01 2.48012155e-01 3.98542494e-01 8.27099085e-01
-2.75265902e-01 1.06126320e+00 8.31254244e-01 -7.76255667e-01
-2.17355378e-02 6.91993773e-01 3.16675633e-01 8.66171658e-01
3.77058595e-01 5.54405674e-02 -2.52170682e-01 -2.97686070e-01
6.08250499e-01 7.62725919e-02 -8.90633166e-02 -9.75042284e-01
-1.22952545e+00 8.43705416e-01 6.25416040e-01 4.12039161e-01
-3.37691866e-02 2.93042302e-01 3.88179809e-01 -8.39246288e-02
7.57517815e-01 2.29048640e-01 -4.16523069e-01 3.81678075e-01
-9.63406503e-01 8.72250795e-02 3.94125164e-01 1.13848066e+00
1.16024220e+00 -4.91429538e-01 -2.75338084e-01 7.49700189e-01
4.16717559e-01 7.69153357e-01 3.35506260e-01 -7.41905749e-01
2.55368680e-01 3.99023920e-01 7.59725124e-02 -8.11116993e-01
-1.63650662e-01 -2.11734295e-01 -1.49041265e-01 -3.38367186e-02
4.53000575e-01 3.84301543e-01 -9.65061367e-01 1.67333221e+00
5.65183640e-01 3.93986583e-01 -2.00852472e-03 8.35244596e-01
9.01689529e-01 4.26729649e-01 1.04975574e-01 -8.55989009e-03
1.58946645e+00 -1.23043191e+00 -3.74449372e-01 -4.53491658e-01
7.34338820e-01 -7.95016706e-01 9.68979657e-01 -1.19224256e-02
-8.75657260e-01 -4.86678720e-01 -6.44680262e-01 -2.53591090e-01
-6.78100765e-01 2.71560639e-01 4.13196176e-01 5.84700465e-01
-1.08605874e+00 1.02605494e-02 -5.83073556e-01 -8.05212557e-01
7.44129419e-01 1.69639632e-01 -4.92981225e-01 -2.74982661e-01
-6.38449252e-01 8.21707308e-01 6.05573237e-01 -3.88215333e-01
-1.15416312e+00 -5.20513773e-01 -1.07297182e+00 -3.37780267e-01
4.84262496e-01 -5.89243114e-01 1.41454077e+00 -1.22265136e+00
-6.77990854e-01 1.62751734e+00 -4.23073500e-01 -4.50800985e-01
3.51758003e-01 -7.79996589e-02 -1.79967418e-01 5.06118894e-01
4.33297068e-01 1.16327357e+00 1.02291799e+00 -1.50129521e+00
-7.67460763e-01 -3.55918109e-01 3.21347490e-02 8.34788778e-04
-2.93829620e-01 2.00343475e-01 -6.85125709e-01 -5.88266313e-01
5.52183725e-02 -9.99554038e-01 -1.37335002e-01 3.67104173e-01
-1.78748086e-01 -3.89735430e-01 7.29242861e-01 -4.71024066e-01
8.17683160e-01 -2.10083413e+00 -2.53715575e-01 6.74977107e-03
2.73519725e-01 3.52969736e-01 -4.77301687e-01 2.96903372e-01
-1.58684164e-01 -3.74942087e-02 -2.76130468e-01 -4.78343934e-01
-2.63615131e-01 1.92883313e-01 -4.93925869e-01 6.62252963e-01
3.30999762e-01 1.18133688e+00 -1.17754209e+00 -7.13439167e-01
1.85610950e-01 3.06967586e-01 -3.39403212e-01 9.25288647e-02
-5.25023997e-01 3.90122384e-01 -3.14647704e-01 7.50027061e-01
7.55956888e-01 -5.36525309e-01 -2.60655768e-02 -6.59093112e-02
-7.17205480e-02 -1.03595346e-01 -1.00471950e+00 1.80905330e+00
-4.06227767e-01 8.36512208e-01 -2.07565762e-02 -9.87069726e-01
6.56349361e-01 1.79854855e-01 2.39093527e-01 -3.49874586e-01
1.25296980e-01 2.92495072e-01 -6.26407564e-01 -6.45746827e-01
2.42578372e-01 -5.60118183e-02 -1.89895332e-02 4.10553426e-01
3.48841459e-01 -6.20303042e-02 2.83090144e-01 5.53390384e-01
9.08126831e-01 1.74006760e-01 4.48621064e-01 -1.85331911e-01
5.26123106e-01 3.29500049e-01 3.01905513e-01 1.09400666e+00
-3.27300161e-01 9.14579749e-01 -7.30092824e-02 -6.18801832e-01
-1.18309391e+00 -1.21941316e+00 -4.48901445e-01 1.19334269e+00
3.01132202e-01 -3.16621602e-01 -6.46000624e-01 -9.05628562e-01
8.57385695e-02 5.01948297e-01 -6.96435571e-01 -3.46093276e-03
-3.55572999e-01 -2.72168845e-01 4.04047698e-01 5.17763913e-01
1.39474526e-01 -9.61186349e-01 -4.24147695e-01 -1.70069441e-01
-3.37852806e-01 -1.15020454e+00 -6.50766611e-01 7.77968392e-02
-6.20727122e-01 -1.30350304e+00 -8.87948036e-01 -1.38082385e+00
1.03505123e+00 5.81140578e-01 1.12868547e+00 2.28490368e-01
-7.04424322e-01 1.01759529e+00 -4.11975026e-01 -5.75299203e-01
-3.97514641e-01 -1.28119230e-01 8.93561020e-02 1.91207796e-01
7.48884380e-01 -2.85468344e-02 -5.30488551e-01 2.36414135e-01
-8.82825315e-01 1.01961203e-01 6.36674345e-01 6.85714185e-01
7.80419946e-01 -5.82362354e-01 2.51629353e-01 -7.15216160e-01
-8.75801221e-03 -4.62289810e-01 -7.19086707e-01 5.64949095e-01
-3.95423502e-01 -1.99266970e-01 -2.76077073e-02 -6.67565465e-01
-8.01449537e-01 6.36744738e-01 2.02459753e-01 -5.41564226e-01
-4.51527894e-01 2.24899594e-02 2.30958521e-01 -3.81302327e-01
7.31596053e-01 4.28067118e-01 -1.10813016e-02 -4.46526051e-01
7.43974864e-01 8.08473766e-01 5.50939381e-01 -3.39059711e-01
1.08336961e+00 9.40795302e-01 -4.79586095e-01 -7.29832470e-01
-1.14235067e+00 -1.22406733e+00 -9.96455252e-01 -1.81369483e-01
8.39688718e-01 -1.29087150e+00 -7.87622109e-02 2.25860491e-01
-1.32836080e+00 -3.24428856e-01 -4.45946485e-01 4.83983964e-01
-6.34910822e-01 4.98927951e-01 -3.54819894e-01 -8.28011513e-01
-2.40261301e-01 -7.62121201e-01 1.63623428e+00 6.70695603e-02
5.77024417e-03 -8.93758237e-01 2.20723689e-01 5.98077357e-01
3.28686573e-02 -9.44784135e-02 2.68065244e-01 -9.56648231e-01
-1.00572443e+00 -5.63047588e-01 -5.59643209e-01 2.19888017e-01
-1.21781759e-01 -4.68035340e-01 -1.05483508e+00 -4.85674590e-01
-2.44574502e-01 -5.45980752e-01 1.12733090e+00 1.48507535e-01
8.93536568e-01 -2.29449138e-01 -8.04507196e-01 3.30182225e-01
1.43768990e+00 -3.94532770e-01 4.01539415e-01 4.44160640e-01
5.74490488e-01 4.79653448e-01 7.39751458e-01 2.07026854e-01
4.53685582e-01 6.74773455e-01 1.86024889e-01 -1.96273789e-01
-3.00444692e-01 -3.03330511e-01 9.54287499e-02 1.24889828e-01
4.67934787e-01 -1.50427505e-01 -1.20415831e+00 1.14420664e+00
-2.02412200e+00 -9.36603010e-01 -1.92810312e-01 2.18039799e+00
7.69787788e-01 -4.86673824e-02 1.45596370e-01 -4.20197129e-01
1.01970112e+00 -8.31010118e-02 -3.70939910e-01 1.12785883e-01
-1.07697360e-01 5.02665108e-03 5.82378924e-01 5.11886179e-01
-1.32098341e+00 1.10284901e+00 6.12390614e+00 9.74523187e-01
-7.16739416e-01 6.89043820e-01 3.98080558e-01 -1.11675404e-01
-5.92767037e-02 8.76491293e-02 -1.07174230e+00 1.27007440e-01
6.28633261e-01 -8.92069042e-02 3.06393616e-02 9.38018441e-01
-3.99127752e-02 -3.00357640e-01 -1.14759111e+00 1.20770562e+00
7.64473319e-01 -1.51727295e+00 1.46189734e-01 -2.14793548e-01
6.82485998e-01 4.71477777e-01 -4.38550375e-02 1.23740412e-01
4.42107990e-02 -5.28231382e-01 8.28204870e-01 4.05886650e-01
6.99937701e-01 -5.62892817e-02 3.99933130e-01 3.89136761e-01
-1.25937104e+00 -1.63143292e-01 -3.48807991e-01 -1.01221381e-02
-1.28112733e-02 3.36988956e-01 -9.46254730e-01 7.02190176e-02
8.90621126e-01 6.53438449e-01 -9.00540411e-01 1.35600889e+00
-4.33421046e-01 4.88634706e-01 -3.91248614e-01 1.83062941e-01
2.39910319e-01 1.88196659e-01 6.66067779e-01 1.28622341e+00
-1.60913616e-01 -6.33959025e-02 5.22659481e-01 7.89080381e-01
-1.03830904e-01 3.36205959e-01 -7.79717445e-01 2.53377467e-01
1.74058422e-01 1.37281322e+00 -9.13720191e-01 -5.38190603e-01
-5.99731445e-01 1.13072848e+00 5.81846297e-01 2.29674593e-01
-7.32898355e-01 -1.60359845e-01 2.79777408e-01 3.54743093e-01
6.30052030e-01 -1.56264216e-01 6.78805560e-02 -1.33734596e+00
2.80247033e-01 -3.29323083e-01 7.24281073e-01 -1.00458837e+00
-1.24434090e+00 4.66695189e-01 6.41762987e-02 -1.20079839e+00
8.74675289e-02 -7.09235251e-01 -4.94485408e-01 4.41978276e-01
-1.76136017e+00 -1.72357178e+00 -3.00743818e-01 5.63320279e-01
7.19771385e-01 -2.51677353e-02 6.62539482e-01 3.39699358e-01
-3.32357198e-01 5.11356890e-01 3.29946399e-01 3.84353429e-01
8.08333755e-01 -1.03941035e+00 3.13529015e-01 8.75160575e-01
6.60113573e-01 4.91016030e-01 5.13228357e-01 -6.68502033e-01
-9.23696518e-01 -1.46722841e+00 1.08143926e+00 -1.07014275e+00
7.20826864e-01 -8.76046002e-01 -8.79272699e-01 9.81651723e-01
1.20132469e-01 6.52767062e-01 6.58393860e-01 -9.71381441e-02
-7.49956131e-01 7.10770786e-02 -1.02774799e+00 2.74660856e-01
1.01553309e+00 -7.33204067e-01 -7.34338164e-01 1.07513380e+00
7.02397466e-01 -3.44749093e-01 -2.95830518e-01 2.95309514e-01
3.72858137e-01 -3.31556201e-01 1.18943012e+00 -5.89311361e-01
-8.17921385e-02 -5.19005239e-01 -1.71098217e-01 -6.12554371e-01
-2.95041353e-01 -3.52256328e-01 7.53167197e-02 1.03796208e+00
5.32307565e-01 -5.67345560e-01 6.41559184e-01 4.29444671e-01
2.49574795e-01 -3.35025311e-01 -9.28509474e-01 -1.16945851e+00
1.87599789e-02 -4.52712655e-01 -7.24357888e-02 7.61717200e-01
-5.20093217e-02 2.88842410e-01 -2.43926093e-01 4.39125270e-01
8.49618018e-01 3.80015731e-01 7.51559854e-01 -1.00766706e+00
-1.31623730e-01 -9.94443223e-02 -6.64225578e-01 -9.82735813e-01
2.58328944e-01 -1.18449140e+00 1.93832546e-01 -1.39257765e+00
7.36483634e-01 -5.27188957e-01 -3.80629629e-01 7.98678041e-01
-1.33485058e-02 1.03103745e+00 1.06884830e-01 4.87985611e-01
-1.34869814e+00 1.84652805e-01 5.96455872e-01 -3.19710702e-01
-3.51732261e-02 -1.22387521e-01 -5.25770962e-01 6.63520753e-01
7.74132967e-01 -8.94600928e-01 -3.78782414e-02 -4.52563643e-01
9.23911780e-02 -3.61993492e-01 8.48150432e-01 -9.19466257e-01
3.73804390e-01 8.91441777e-02 1.95582211e-01 -6.17109716e-01
1.85351238e-01 -7.33214438e-01 -2.41574049e-01 4.38377589e-01
-5.00486732e-01 -4.87333775e-01 3.95704627e-01 1.04912841e+00
-2.52149832e-02 -4.46150869e-01 8.93052101e-01 -2.42635548e-01
-1.07281578e+00 4.05263394e-01 -4.18500721e-01 4.98714894e-02
1.44602311e+00 -2.55441785e-01 -1.80905655e-01 -1.38566017e-01
-8.63686025e-01 3.16852212e-01 5.36242604e-01 7.77971268e-01
5.49465895e-01 -1.25260210e+00 -8.16201448e-01 1.34533361e-01
9.85407591e-01 -9.62872282e-02 9.22124460e-02 7.98689067e-01
-3.52106273e-01 5.67444265e-01 2.91789174e-01 -9.52276170e-01
-1.58350825e+00 1.01986265e+00 3.74618024e-01 5.51962517e-02
-5.56687295e-01 1.02731252e+00 5.94568014e-01 -6.61662579e-01
2.80943036e-01 7.79428855e-02 5.61643578e-02 -1.14292987e-01
7.79000223e-01 -1.20304234e-01 -1.51289597e-01 -9.99314845e-01
-6.18176877e-01 7.49499500e-01 -2.10473493e-01 9.99245867e-02
1.11757421e+00 -4.75869447e-01 -1.44344911e-01 3.78838360e-01
1.32773125e+00 -1.90933838e-01 -1.01965559e+00 -8.81905496e-01
2.21521199e-01 -6.31995261e-01 1.52182356e-01 -7.16894209e-01
-6.25516355e-01 6.60735369e-01 1.02285266e+00 -1.88801721e-01
8.46397221e-01 7.85440087e-01 5.10823667e-01 5.33936739e-01
3.71860683e-01 -1.01387882e+00 3.44324201e-01 2.97440469e-01
8.04407120e-01 -1.86839795e+00 -1.26844943e-01 -6.21856570e-01
-3.03192526e-01 9.68798459e-01 5.31058490e-01 -5.34810685e-02
6.50319099e-01 -1.13281034e-01 1.47685722e-01 -3.18827152e-01
-5.27679265e-01 -7.49135792e-01 4.41064149e-01 7.03120172e-01
7.25418553e-02 -1.40032962e-01 -1.07630193e-02 2.73456752e-01
4.42113847e-01 -3.59449983e-02 3.78304482e-01 9.94014680e-01
-6.81451738e-01 -1.02327430e+00 -5.92092693e-01 2.78953582e-01
-1.85601056e-01 -3.57234210e-01 -2.98010916e-01 6.56278670e-01
2.53620654e-01 8.77776384e-01 2.83941925e-01 1.79504707e-01
7.77425840e-02 1.45048812e-01 5.60593605e-01 -1.13181388e+00
-1.08881027e-01 -5.07444777e-02 -2.19730154e-01 -4.77085978e-01
-7.42871225e-01 -8.51757765e-01 -9.01385665e-01 4.25732166e-01
-7.26787806e-01 1.38214588e-01 6.36449575e-01 8.88502657e-01
4.91594106e-01 -1.53307065e-01 4.48823482e-01 -7.79388607e-01
-2.00133622e-01 -6.48854017e-01 -4.67497438e-01 5.03198922e-01
5.90053499e-01 -5.86515188e-01 -3.59994292e-01 3.04858953e-01] | [9.750372886657715, 1.4800716638565063] |
b5d65aa0-0356-4f4d-8f72-a6ac5b65fb6a | xy-network-for-nuclear-segmentation-in-multi | 1812.06499 | null | https://arxiv.org/abs/1812.06499v5 | https://arxiv.org/pdf/1812.06499v5.pdf | HoVer-Net: Simultaneous Segmentation and Classification of Nuclei in Multi-Tissue Histology Images | Nuclear segmentation and classification within Haematoxylin & Eosin stained histology images is a fundamental prerequisite in the digital pathology work-flow. The development of automated methods for nuclear segmentation and classification enables the quantitative analysis of tens of thousands of nuclei within a whole-slide pathology image, opening up possibilities of further analysis of large-scale nuclear morphometry. However, automated nuclear segmentation and classification is faced with a major challenge in that there are several different types of nuclei, some of them exhibiting large intra-class variability such as the tumour cells. Additionally, some of the nuclei are often clustered together. To address these challenges, we present a novel convolutional neural network for simultaneous nuclear segmentation and classification that leverages the instance-rich information encoded within the vertical and horizontal distances of nuclear pixels to their centres of mass. These distances are then utilised to separate clustered nuclei, resulting in an accurate segmentation, particularly in areas with overlapping instances. Then for each segmented instance, the network predicts the type of nucleus via a devoted up-sampling branch. We demonstrate state-of-the-art performance compared to other methods on multiple independent multi-tissue histology image datasets. As part of this work, we introduce a new dataset of Haematoxylin & Eosin stained colorectal adenocarcinoma image tiles, containing 24,319 exhaustively annotated nuclei with associated class labels. | ['Nasir Rajpoot', 'Ayesha Azam', 'Simon Graham', 'Yee Wah Tsang', 'Quoc Dang Vu', 'Shan E Ahmed Raza', 'Jin Tae Kwak'] | 2018-12-16 | null | null | null | null | ['multi-tissue-nucleus-segmentation', 'nuclear-segmentation'] | ['medical', 'medical'] | [ 4.65814471e-01 4.77395877e-02 -7.64134079e-02 -1.40385062e-01
-1.08613718e+00 -7.50886858e-01 4.86432433e-01 9.37692046e-01
-9.11638975e-01 5.20927489e-01 -1.39310556e-02 -4.43856120e-02
-1.13273777e-01 -8.83409441e-01 -2.81112790e-01 -1.45048964e+00
-1.75034791e-01 8.05322111e-01 3.71559024e-01 7.72499815e-02
1.35901734e-01 8.67095232e-01 -1.15089869e+00 4.49095696e-01
3.61656785e-01 8.50204229e-01 1.10350281e-01 9.77506697e-01
-1.38377711e-01 4.27071095e-01 -3.14189583e-01 3.88653576e-03
-5.98512702e-02 -2.98865348e-01 -9.74821687e-01 1.01417981e-01
3.92169058e-01 -1.35706300e-02 2.34961975e-02 1.12327826e+00
5.46344340e-01 -2.30260342e-01 9.63711202e-01 -6.06390476e-01
1.56254664e-01 4.40286368e-01 -6.78728640e-01 6.19382203e-01
-3.25818747e-01 1.40681550e-01 1.01505160e+00 -5.24330199e-01
8.86163533e-01 5.76146960e-01 9.02974725e-01 2.51879841e-01
-1.56465149e+00 -3.62310588e-01 -6.07034922e-01 -1.20573454e-01
-1.64387655e+00 -1.26705408e-01 3.18767786e-01 -6.90189958e-01
7.82045305e-01 3.32263678e-01 9.91264999e-01 2.78253347e-01
2.95622498e-01 5.75832903e-01 1.21526206e+00 -3.32534492e-01
3.08033198e-01 1.16227688e-02 1.00409269e-01 6.04645729e-01
1.03636548e-01 -3.74547660e-01 8.49534795e-02 4.31169644e-02
6.18374288e-01 2.08497345e-01 -1.42111212e-01 -2.67337948e-01
-1.49739110e+00 7.33250737e-01 6.55689478e-01 7.51203120e-01
-1.50733680e-01 1.71283260e-01 6.56934321e-01 -1.48384780e-01
4.02802885e-01 2.32998237e-01 -2.57482082e-01 4.48541529e-02
-1.31376517e+00 1.22903194e-02 5.54128349e-01 1.53018251e-01
8.73773217e-01 -6.23381197e-01 -3.79287712e-02 6.48350239e-01
2.85799980e-01 -6.57849014e-03 6.63065910e-01 -6.05341911e-01
-1.52458236e-01 1.22477376e+00 -4.13101703e-01 -8.82886589e-01
-9.70738173e-01 -6.49834931e-01 -1.32457793e+00 4.17247742e-01
9.18502450e-01 2.44861141e-01 -9.73267555e-01 1.18189991e+00
7.93409169e-01 -2.02918574e-01 -2.62794018e-01 7.07743108e-01
6.66124463e-01 2.80335754e-01 1.30500138e-01 -9.39974785e-02
1.70638132e+00 -5.03781855e-01 -5.99959433e-01 1.17905773e-01
1.20186138e+00 -5.65631688e-01 6.13910317e-01 -2.93601397e-02
-8.62450600e-01 -4.38537151e-02 -7.69294441e-01 -1.37794390e-01
-7.33757675e-01 1.12025082e-01 5.40920615e-01 4.42197382e-01
-1.22881305e+00 5.39898098e-01 -1.04094195e+00 -5.23547828e-01
9.21022475e-01 7.98527837e-01 -6.61643744e-01 2.03161165e-01
-5.72280765e-01 5.30760646e-01 6.71452880e-01 1.27761692e-01
-3.98864269e-01 -9.14296031e-01 -6.75436437e-01 -6.40315339e-02
-3.20474766e-02 -4.54397708e-01 9.01888788e-01 -7.96110868e-01
-9.22537208e-01 1.41438794e+00 1.02322496e-01 -3.12348425e-01
5.04864156e-01 8.26499760e-01 7.22984001e-02 3.75932664e-01
1.02515012e-01 1.03189135e+00 2.42967859e-01 -1.06114054e+00
-9.36935484e-01 -5.77831924e-01 -4.00980890e-01 1.36395305e-01
-8.07371289e-02 -1.93148017e-01 -3.19501400e-01 -5.36032796e-01
1.27473205e-01 -7.83634067e-01 -4.53715384e-01 1.69619903e-01
-5.10890901e-01 -1.85130835e-02 7.26140618e-01 -4.96047676e-01
1.11590791e+00 -2.27383447e+00 2.98304558e-01 5.70143700e-01
7.13698626e-01 -2.85224449e-02 3.06951612e-01 9.92054567e-02
2.74618957e-02 2.82768875e-01 -2.43946135e-01 -2.89439708e-01
-1.54927984e-01 9.34653655e-02 4.49128181e-01 1.06768382e+00
6.55530617e-02 9.85667467e-01 -8.78162324e-01 -9.26993906e-01
2.01189846e-01 6.56342328e-01 -3.38196903e-01 -8.88819695e-02
3.83954942e-02 5.52827716e-01 -8.55901316e-02 9.29908216e-01
4.12358522e-01 -4.75428790e-01 3.28528047e-01 -2.96061903e-01
-2.11597830e-02 -2.71846116e-01 -8.63159657e-01 1.28616428e+00
-2.58645508e-02 7.50025392e-01 3.66735667e-01 -7.78955281e-01
4.44917858e-01 2.73757011e-01 8.82113397e-01 -3.33075166e-01
5.40668070e-01 5.02927959e-01 2.97724783e-01 -1.40898257e-01
2.96462268e-01 -5.94198883e-01 -2.37213727e-02 4.76053774e-01
1.94120079e-01 -2.25966141e-01 5.90212762e-01 4.25149483e-05
1.29795456e+00 -5.59609175e-01 5.75132012e-01 -3.54230851e-01
4.84431267e-01 1.70463189e-01 3.66397411e-01 2.56788373e-01
-4.54242080e-01 8.97783339e-01 7.77306259e-01 -6.20024681e-01
-1.14882743e+00 -6.62764788e-01 -4.61935222e-01 7.21393228e-01
-2.11190075e-01 -7.78084695e-02 -7.74641395e-01 -6.89362049e-01
-8.64080340e-02 -2.39051446e-01 -1.25477648e+00 3.04089904e-01
-4.81366038e-01 -1.29831052e+00 7.36796021e-01 4.15489107e-01
1.32896125e-01 -1.18847048e+00 -7.21594751e-01 2.60318786e-01
-5.24780415e-02 -7.64270008e-01 -1.81667835e-01 7.83369601e-01
-8.08778703e-01 -1.29538453e+00 -7.14800417e-01 -9.68557000e-01
1.21970081e+00 3.85330878e-02 1.16795707e+00 5.97175539e-01
-9.16592896e-01 -2.97084320e-02 -1.08417057e-01 -1.25752375e-01
-5.78934312e-01 3.38925183e-01 -3.95483613e-01 -7.11201057e-02
3.78177434e-01 -3.85543078e-01 -7.69423068e-01 1.48276716e-01
-1.13695669e+00 -3.48863490e-02 8.62593353e-01 1.05860579e+00
1.30754542e+00 3.69100720e-01 1.12557903e-01 -1.02855206e+00
6.65449053e-02 -5.27205646e-01 -4.85243589e-01 2.36081015e-02
-1.33127975e-03 -3.72269541e-01 5.67442358e-01 -1.37435660e-01
-3.35890591e-01 3.00800681e-01 -1.50973111e-01 2.04671174e-01
-4.95035529e-01 6.26032650e-01 2.29764670e-01 -4.09592181e-01
5.70861220e-01 1.20415598e-01 3.36692214e-01 5.23850173e-02
7.90832862e-02 5.51152349e-01 4.13618416e-01 8.63082632e-02
5.56530237e-01 9.81726289e-01 3.71866286e-01 -8.31766963e-01
-5.16586781e-01 -9.78889883e-01 -1.00439763e+00 -1.84746832e-01
1.15097070e+00 -5.31462014e-01 -7.04034150e-01 6.11689568e-01
-5.22714436e-01 -7.05736995e-01 -5.58791518e-01 2.44915277e-01
-4.16561157e-01 2.13356689e-01 -1.02534795e+00 -2.92026281e-01
-4.18641835e-01 -1.29404616e+00 1.26107156e+00 3.40009570e-01
-4.41082269e-01 -1.39952266e+00 5.00240088e-01 4.17955250e-01
2.06049100e-01 5.32855868e-01 1.16627932e+00 -7.54034996e-01
-2.87188321e-01 -6.28959954e-01 -2.40077093e-01 -8.35466757e-02
2.34626129e-01 4.06000644e-01 -7.29554296e-01 -4.44972783e-01
-6.36969030e-01 -2.78413773e-01 9.86971438e-01 6.31701708e-01
1.09267759e+00 -1.46586195e-01 -7.97332346e-01 9.00144756e-01
1.64396620e+00 -1.94470793e-01 5.04069746e-01 5.89630127e-01
5.35357475e-01 7.60693073e-01 2.74432182e-01 1.28688261e-01
9.95895490e-02 4.01388198e-01 4.85395700e-01 -7.16242313e-01
4.90536029e-03 4.41930354e-01 -3.65440130e-01 4.88499939e-01
-5.36685549e-02 -1.15749866e-01 -1.26201713e+00 9.04447496e-01
-1.27522242e+00 -8.38530481e-01 -3.06892663e-01 1.85168695e+00
1.08550858e+00 -1.90526526e-02 2.39600226e-01 4.57923263e-01
7.10813046e-01 -1.39510527e-01 -3.90451968e-01 1.71883740e-02
-1.02078177e-01 2.47419164e-01 6.55876458e-01 2.08799526e-01
-1.41567874e+00 3.90476733e-01 6.27091217e+00 1.12505567e+00
-1.04819226e+00 -2.08803490e-01 1.27285516e+00 -1.06412128e-01
1.41564831e-01 -4.42947567e-01 -7.53304660e-01 3.51160794e-01
6.79192007e-01 1.65508315e-01 1.27049625e-01 2.83680677e-01
8.57685730e-02 -4.71546680e-01 -9.33004558e-01 5.11681855e-01
-2.66499996e-01 -1.52129972e+00 -1.99733168e-01 6.28877759e-01
7.67970800e-01 1.89444482e-01 -3.82882133e-02 -1.20838188e-01
1.34446368e-01 -1.36433065e+00 2.66883075e-01 2.17021510e-01
1.00456202e+00 -8.21291447e-01 1.33912814e+00 2.52281100e-01
-1.17311954e+00 4.86136936e-02 -2.48582974e-01 5.07924378e-01
-2.98611581e-01 5.94543338e-01 -1.18735909e+00 2.27865234e-01
5.45752227e-01 4.84481484e-01 -8.45739484e-01 1.25202215e+00
3.94921988e-01 5.12750745e-01 -4.15400624e-01 -4.22438001e-03
3.15911442e-01 -1.04821809e-01 7.86606148e-02 1.59658802e+00
2.59732038e-01 1.01938630e-02 5.42510748e-02 6.12074792e-01
-9.23917517e-02 3.03094536e-01 -9.38158780e-02 -2.26245150e-01
1.10540479e-01 2.10572243e+00 -1.86213601e+00 -1.70621559e-01
-1.23846926e-01 4.54599679e-01 4.79350924e-01 -3.35484557e-02
-5.78801811e-01 -2.54815906e-01 4.07268107e-01 4.40028965e-01
1.19878478e-01 4.43031304e-02 -3.03160518e-01 -4.41937208e-01
-4.82824087e-01 -6.65266693e-01 6.05377674e-01 -1.28917903e-01
-1.34767151e+00 3.96486849e-01 -5.08264899e-01 -1.12070262e+00
1.19232930e-01 -6.02856994e-01 -5.18050909e-01 5.02919912e-01
-1.46659088e+00 -1.37531650e+00 -4.23114002e-01 9.68036950e-02
1.04465976e-01 1.08924918e-01 1.10095274e+00 2.33856529e-01
-4.24058467e-01 3.41821343e-01 4.11404282e-01 5.30770600e-01
4.59555835e-01 -1.89511836e+00 -1.09736480e-01 2.21880719e-01
-8.09263289e-02 3.55198622e-01 3.50180656e-01 -4.58678901e-01
-9.67147410e-01 -1.25645578e+00 6.10068500e-01 -2.51836032e-01
8.38673651e-01 -2.90859878e-01 -9.25256133e-01 4.07269657e-01
9.00446475e-02 6.14229918e-01 1.40569830e+00 -3.54639173e-01
2.04074204e-01 7.17018247e-02 -1.37619770e+00 4.84835267e-01
2.08765179e-01 -3.69193584e-01 -9.17187601e-04 4.63294655e-01
3.55737843e-02 -5.55000305e-01 -1.14212012e+00 1.20860897e-01
4.44642693e-01 -1.04562509e+00 7.77077854e-01 -2.09092684e-02
3.84732485e-01 -4.68633831e-01 2.40907803e-01 -1.17203438e+00
-4.61568445e-01 -7.35986084e-02 3.85483414e-01 8.77801359e-01
4.24221694e-01 -3.74792546e-01 1.20955968e+00 6.87188730e-02
-1.23296551e-01 -1.13738728e+00 -1.21582246e+00 -9.73195136e-02
3.51898521e-01 1.47257805e-01 1.96357757e-01 7.80147672e-01
5.49792312e-02 -2.70780157e-02 5.07877290e-01 -1.31681845e-01
5.59402823e-01 -1.70996949e-01 4.16102916e-01 -1.20861447e+00
2.15367042e-02 -8.93011808e-01 -1.07958770e+00 -1.80963308e-01
3.21365311e-03 -1.20789433e+00 -5.80696762e-03 -1.60643339e+00
5.42909086e-01 -3.86083633e-01 -3.21689993e-01 5.29319108e-01
-1.71233743e-01 1.10918617e+00 -3.78820300e-01 2.85928637e-01
-7.96812117e-01 -6.67769685e-02 1.28773141e+00 -4.48303401e-01
-2.96877529e-02 -2.17976630e-01 -3.26906502e-01 6.72445536e-01
7.17068613e-01 -4.49623227e-01 2.29241207e-01 2.21378788e-01
3.14532906e-01 -2.04249457e-01 5.27459800e-01 -1.23798871e+00
3.81290108e-01 8.14950168e-02 7.64193475e-01 -8.64464283e-01
4.72419858e-02 -9.45856750e-01 2.61379212e-01 5.61875701e-01
-3.05840135e-01 -3.20081711e-01 1.62173674e-01 5.77904999e-01
-3.35645884e-01 -1.47709167e-02 1.14871514e+00 -3.57353032e-01
-1.45070717e-01 2.88485855e-01 -7.87231088e-01 -2.04369992e-01
1.27836180e+00 -6.89736724e-01 -2.98739195e-01 2.31105745e-01
-7.60627627e-01 1.44103333e-01 7.27530003e-01 -5.41882455e-01
2.15919733e-01 -1.19332433e+00 -6.55168712e-01 5.54798823e-03
2.53087223e-01 6.09769881e-01 5.48341572e-01 1.40680945e+00
-1.07788777e+00 2.66201586e-01 -1.53662130e-01 -9.30689573e-01
-1.57609808e+00 1.25446662e-01 8.17235529e-01 -9.00572121e-01
-5.93778789e-01 9.86397326e-01 2.12300941e-01 -4.54338759e-01
-4.40429524e-02 -5.89950204e-01 -5.93216300e-01 3.95380348e-01
4.97424871e-01 1.78822368e-01 2.54641235e-01 -8.91103268e-01
-1.81710571e-01 5.48730969e-01 -1.30922273e-01 2.18675807e-01
1.39119947e+00 7.63453171e-02 -7.57038951e-01 5.55573225e-01
1.25213468e+00 -6.58757836e-02 -1.21024454e+00 -2.86171585e-02
1.67022422e-01 -2.60563999e-01 1.06994137e-01 -7.16934144e-01
-1.17276037e+00 7.43923247e-01 6.77302182e-01 3.90769750e-01
1.05501521e+00 2.18801126e-02 5.61873674e-01 1.69678926e-02
5.20620197e-02 -1.02658868e+00 -1.44043043e-01 2.60412484e-01
1.88759014e-01 -1.32284355e+00 1.52434140e-01 -3.76863182e-01
-1.28280208e-01 1.43009150e+00 1.46382332e-01 -2.14728117e-02
4.73178148e-01 6.74998403e-01 2.51460761e-01 -6.31474078e-01
-5.25713503e-01 -1.47748813e-01 1.65444851e-01 4.28126782e-01
6.60222292e-01 2.16950655e-01 -5.86709306e-02 4.45751250e-01
-1.19881146e-01 -1.64810240e-01 4.64973211e-01 9.30479109e-01
-4.73226130e-01 -8.76264513e-01 -3.36035222e-01 7.88906574e-01
-9.15852904e-01 1.54303955e-02 -5.09872437e-01 1.00231886e+00
2.58653075e-01 3.04514766e-01 4.17530775e-01 1.22355381e-02
-1.59516826e-01 -2.46823579e-01 3.75021607e-01 -7.05034316e-01
-1.06795108e+00 3.66603583e-01 -3.24800551e-01 -1.53149972e-02
-4.57786977e-01 -6.45562172e-01 -1.79409540e+00 -2.24993318e-01
-3.26582283e-01 -2.21028719e-02 5.13040364e-01 8.68658721e-01
-1.70174748e-01 7.26120949e-01 3.92016739e-01 -1.00257766e+00
-1.04294211e-01 -9.02168453e-01 -1.07545018e+00 3.67706299e-01
5.28675377e-01 -2.57961899e-01 -6.11593962e-01 2.66419441e-01] | [15.0261812210083, -3.117263078689575] |
212eed41-58f1-4f7f-87ce-ac70755bf7d2 | comparison-of-single-and-multi-objective | 2306.17038 | null | https://arxiv.org/abs/2306.17038v1 | https://arxiv.org/pdf/2306.17038v1.pdf | Comparison of Single- and Multi- Objective Optimization Quality for Evolutionary Equation Discovery | Evolutionary differential equation discovery proved to be a tool to obtain equations with less a priori assumptions than conventional approaches, such as sparse symbolic regression over the complete possible terms library. The equation discovery field contains two independent directions. The first one is purely mathematical and concerns differentiation, the object of optimization and its relation to the functional spaces and others. The second one is dedicated purely to the optimizational problem statement. Both topics are worth investigating to improve the algorithm's ability to handle experimental data a more artificial intelligence way, without significant pre-processing and a priori knowledge of their nature. In the paper, we consider the prevalence of either single-objective optimization, which considers only the discrepancy between selected terms in the equation, or multi-objective optimization, which additionally takes into account the complexity of the obtained equation. The proposed comparison approach is shown on classical model examples -- Burgers equation, wave equation, and Korteweg - de Vries equation. | ['Alexander Hvatov', 'Mikhail Maslyaev'] | 2023-06-29 | null | null | null | null | ['symbolic-regression'] | ['knowledge-base'] | [ 8.00772682e-02 7.86121935e-02 1.52554289e-01 -1.49112567e-01
2.12263986e-02 -3.15172404e-01 3.66810292e-01 2.56117642e-01
-6.90088928e-01 1.13421667e+00 -3.51593107e-01 -2.46897891e-01
-7.39253104e-01 -6.12627864e-01 -1.48275018e-01 -8.33089828e-01
-2.10159384e-02 7.55798638e-01 -1.24089278e-01 -6.13680124e-01
5.14681935e-01 7.79111862e-01 -1.74389744e+00 -4.57823545e-01
1.06244159e+00 9.57389355e-01 5.87633476e-02 5.40885746e-01
-3.57696414e-01 5.51254928e-01 -5.59984088e-01 -1.62125289e-01
1.35838091e-02 -6.09701693e-01 -6.18373692e-01 -8.73909369e-02
-5.16143799e-01 3.24578047e-01 5.10523260e-01 1.10082257e+00
2.84720242e-01 2.65579432e-01 1.08154666e+00 -1.10719883e+00
-2.38819927e-01 3.44726175e-01 -3.96759547e-02 2.39499420e-01
5.27023971e-01 7.31833652e-02 5.51146030e-01 -7.57613957e-01
6.69075727e-01 9.15736139e-01 5.44216216e-01 1.29415378e-01
-1.22841656e+00 -8.01127777e-02 -1.64400205e-01 2.83212990e-01
-1.62686837e+00 -1.20002456e-01 8.94731998e-01 -9.15556550e-01
8.27271581e-01 4.45308566e-01 8.14415634e-01 4.87823337e-01
2.12085918e-01 1.53693035e-01 1.33000958e+00 -5.45560002e-01
5.10771632e-01 7.47075796e-01 5.38811386e-01 5.71069598e-01
6.30822897e-01 8.72173086e-02 -1.27957940e-01 -7.92511627e-02
3.67380798e-01 -5.97380698e-01 -4.13183600e-01 -4.80918795e-01
-5.58080792e-01 9.64479446e-01 -6.60717636e-02 1.10008967e+00
-5.82963109e-01 -2.63866454e-01 1.19156055e-01 2.06650957e-01
4.35641199e-01 1.05222201e+00 -4.57524002e-01 -2.46574506e-01
-1.09957790e+00 4.07553107e-01 1.19495630e+00 5.27436793e-01
8.84278297e-01 4.66104060e-01 3.75929892e-01 1.63068712e-01
3.92728627e-01 3.21705937e-01 5.33582747e-01 -2.57769525e-01
1.13198534e-01 8.11542690e-01 6.95417672e-02 -1.26255643e+00
-5.45573771e-01 -7.31857777e-01 -5.65907717e-01 5.62718093e-01
5.09260237e-01 -5.06703019e-01 -3.81511450e-01 1.35046768e+00
5.57371616e-01 -1.86868146e-01 9.54211056e-02 9.08904374e-01
7.18035102e-01 4.45941567e-01 -2.71326482e-01 -8.90361249e-01
9.43988979e-01 -4.47635084e-01 -9.61085618e-01 5.95529258e-01
6.31679773e-01 -7.88720369e-01 4.57382053e-01 5.08952320e-01
-1.02692866e+00 -1.58070579e-01 -1.02523255e+00 3.52375716e-01
-8.41126144e-01 2.73431271e-01 5.05954683e-01 6.32964969e-01
-9.04964089e-01 9.34341729e-01 -5.31764388e-01 -1.99259624e-01
-1.68833509e-01 4.62701827e-01 -1.42210037e-01 5.66670597e-01
-9.74276960e-01 1.46064448e+00 4.11572427e-01 3.80863994e-01
-1.29643127e-01 -6.50354743e-01 -6.02914572e-01 5.07316776e-02
4.42477256e-01 -5.59099317e-01 6.48766339e-01 -1.04968250e+00
-1.90592861e+00 7.33459949e-01 -2.65482843e-01 -2.81756818e-01
8.14159691e-01 1.91945702e-01 -3.59614551e-01 -6.10130094e-03
-1.27672061e-01 -1.55071825e-01 7.87925541e-01 -1.25724506e+00
9.82963368e-02 -1.24952339e-01 -2.96372473e-01 -2.74964403e-02
2.32743216e-03 -4.86039296e-02 2.36564994e-01 -3.73538762e-01
1.71350747e-01 -5.61150372e-01 -3.14177930e-01 -6.08753301e-02
-3.72049838e-01 -5.72066456e-02 4.47785437e-01 -5.20685852e-01
1.16330695e+00 -1.66733050e+00 1.03610468e+00 5.46139717e-01
2.43107043e-02 3.92508626e-01 5.67168415e-01 5.93550265e-01
-2.96506643e-01 1.04919240e-01 -5.00053704e-01 -2.38761678e-01
-3.23861390e-02 6.33371323e-02 1.23026453e-01 7.12674916e-01
2.42923573e-01 2.16097742e-01 -4.72274542e-01 -7.19259501e-01
3.08130115e-01 5.40606558e-01 -5.09369612e-01 4.38878350e-02
-1.97167203e-01 6.43253207e-01 -6.65512621e-01 3.27484697e-01
6.20510936e-01 2.20627010e-01 -7.87147880e-02 -4.25908774e-01
-8.44496369e-01 -2.03647569e-01 -1.58483362e+00 1.19596553e+00
-3.56924206e-01 4.07417446e-01 2.51205295e-01 -1.46226037e+00
1.14801526e+00 3.03599209e-01 9.03878629e-01 -3.20306033e-01
7.65386462e-01 6.18504405e-01 1.97551310e-01 -6.89699829e-01
4.85283852e-01 -2.15845689e-01 3.77896160e-01 2.23358676e-01
8.71045142e-02 -2.84847945e-01 6.01539552e-01 -4.65843290e-01
4.54356253e-01 3.14677268e-01 6.81834638e-01 -8.20131004e-01
1.14815390e+00 2.23415330e-01 1.25141829e-01 1.56948850e-01
1.70579776e-01 5.49420007e-02 5.91616988e-01 -2.92293668e-01
-7.60922909e-01 -3.96690726e-01 -5.15941978e-01 1.21988498e-01
3.11900169e-01 8.13900828e-02 -6.14468873e-01 -2.62922738e-02
3.15709710e-02 8.27755511e-01 -5.99696815e-01 -5.85176572e-02
-4.98017073e-01 -8.90759289e-01 2.57287085e-01 -4.55269575e-01
4.84241515e-01 -7.30074942e-01 -9.65855300e-01 3.24503869e-01
2.33381733e-01 -6.63706481e-01 5.40541887e-01 2.12376401e-01
-9.70056295e-01 -1.11084247e+00 -8.59322131e-01 -3.01270813e-01
5.35285056e-01 -6.89354122e-01 9.30975199e-01 1.97146043e-01
-7.83201575e-01 3.98579478e-01 -2.82608300e-01 -4.95739818e-01
-7.93234110e-01 5.82395159e-02 2.85493527e-02 1.39745131e-01
6.35630786e-02 -8.30290616e-01 8.29784665e-03 2.28196472e-01
-6.19063914e-01 -4.22943890e-01 4.42493051e-01 5.48703313e-01
3.42320859e-01 6.57028034e-02 3.40092629e-01 -3.75079304e-01
7.02661157e-01 -3.97302687e-01 -8.01728964e-01 1.93591043e-01
-8.54532838e-01 5.52923441e-01 6.29336298e-01 -3.95669669e-01
-1.05073214e+00 -1.02824252e-02 -3.06087732e-01 -1.99325457e-01
-1.81802109e-01 5.56281626e-01 -1.39462024e-01 -5.71501315e-01
5.15421987e-01 4.62843806e-01 9.36822072e-02 -7.63205588e-01
5.18264621e-02 2.82424003e-01 2.20108837e-01 -6.10445142e-01
8.31496119e-01 8.10530931e-02 6.22313857e-01 -1.24132919e+00
1.00456752e-01 -4.50381786e-01 -5.79038203e-01 -4.64384824e-01
9.06094193e-01 -5.01412638e-02 -9.53856766e-01 3.74766946e-01
-1.40889215e+00 1.86736211e-01 -5.15745699e-01 7.55142033e-01
-4.74378079e-01 2.80250877e-01 -6.53524920e-02 -1.38081574e+00
-1.23312376e-01 -1.34094095e+00 5.63423097e-01 2.16422990e-01
-2.11769119e-01 -1.19914460e+00 4.02101457e-01 -1.19477585e-01
4.25803483e-01 5.02084553e-01 8.69475365e-01 -5.04206419e-01
-2.47282162e-01 -1.38464704e-01 3.03155810e-01 1.01535857e-01
-7.45686144e-02 2.71731466e-01 -6.35042131e-01 7.93301016e-02
6.52607679e-01 1.81709945e-01 5.05699337e-01 2.08775967e-01
4.16056871e-01 -1.31833375e-01 -2.67206639e-01 6.50629759e-01
1.91068721e+00 3.83558095e-01 4.19902533e-01 2.19419584e-01
1.94409385e-01 5.66993177e-01 5.40987432e-01 5.27782619e-01
-1.78449929e-01 1.13346922e+00 4.85987335e-01 -7.90575370e-02
2.10020691e-01 4.47142750e-01 8.03684816e-02 7.33910561e-01
-7.20223188e-01 -1.41827941e-01 -7.45174050e-01 2.66916931e-01
-1.52020419e+00 -9.73045111e-01 -4.89347309e-01 2.08868098e+00
4.71296102e-01 4.78273481e-02 1.11707307e-01 6.04456902e-01
6.10797703e-01 -2.11180478e-01 -2.34171301e-01 -8.29892099e-01
-2.31104806e-01 3.11395228e-01 4.00027901e-01 9.99811590e-01
-6.87098444e-01 2.12749958e-01 5.33193684e+00 5.09483516e-01
-1.40911901e+00 -2.18768138e-02 -3.26002836e-02 1.73176691e-01
-4.22135353e-01 5.39172329e-02 -6.76055610e-01 6.88630700e-01
6.25612140e-01 -3.47483814e-01 4.25765425e-01 6.15328133e-01
2.44447231e-01 -6.08607292e-01 -7.61672318e-01 8.15890670e-01
1.67624578e-01 -1.06032753e+00 -2.06290931e-01 1.47234410e-01
5.45396566e-01 -5.90553343e-01 -2.41173208e-01 -3.25493142e-02
-4.36067045e-01 -9.80749011e-01 7.71931767e-01 1.06735587e+00
2.39280701e-01 -5.06561399e-01 9.17796552e-01 4.36133325e-01
-9.96049821e-01 2.19380006e-01 -3.46102528e-02 -3.04843187e-01
5.01064718e-01 6.11507773e-01 -5.44169307e-01 1.05712235e+00
1.16663001e-01 3.96388561e-01 -2.98708618e-01 1.23314655e+00
2.35311449e-01 3.47697943e-01 -5.75166702e-01 -7.48643756e-01
1.55796185e-01 -9.18866396e-01 1.15194547e+00 9.98702586e-01
5.17575562e-01 1.73532963e-01 -2.47758880e-01 1.38571072e+00
8.68105590e-01 6.76045716e-01 -6.20046020e-01 -6.84751123e-02
-8.15525204e-02 1.04388058e+00 -9.58235264e-01 -8.81100297e-02
1.14696205e-01 6.70353949e-01 -9.32044610e-02 2.94463098e-01
-7.83862233e-01 -3.58743012e-01 4.71306324e-01 2.67690811e-02
1.77089617e-01 -3.89219016e-01 -3.78595680e-01 -1.20236063e+00
-1.99321704e-03 -5.71838260e-01 -7.14452565e-02 -4.60652411e-01
-5.89423954e-01 6.93443358e-01 4.01084214e-01 -8.27794790e-01
-5.69875240e-01 -9.83157039e-01 -3.87295842e-01 1.11423278e+00
-1.23090303e+00 -7.46213973e-01 4.65944596e-02 5.42645752e-01
2.62837201e-01 -1.64519310e-01 7.93026388e-01 4.42519188e-01
-7.01447845e-01 1.36248097e-01 3.11010152e-01 -6.05418324e-01
4.68893871e-02 -1.08400118e+00 -7.04970181e-01 7.61716366e-01
-3.88061672e-01 3.36744666e-01 1.40419662e+00 -5.13345242e-01
-1.45604372e+00 -1.71705946e-01 1.06994402e+00 7.37849995e-03
6.33571327e-01 5.67119801e-03 -9.33671653e-01 5.38594387e-02
7.62478411e-02 -3.19535434e-01 2.70459622e-01 -1.72149256e-01
4.37938899e-01 1.22164614e-01 -1.19044900e+00 3.42258245e-01
5.49276590e-01 1.79504417e-02 -5.71403563e-01 1.16520129e-01
2.36238942e-01 -2.94732451e-01 -1.05386639e+00 3.67664933e-01
3.27580541e-01 -1.04881573e+00 9.07247663e-01 -5.02924979e-01
2.75122464e-01 -2.45378390e-01 1.82571083e-01 -1.09510934e+00
2.00831100e-01 -7.76585162e-01 1.72029540e-01 1.23314083e+00
4.61687922e-01 -9.41342056e-01 4.65334028e-01 4.88447517e-01
-1.68308988e-01 -1.07317114e+00 -1.10373020e+00 -8.30594182e-01
1.58340912e-02 -2.07685471e-01 4.36182529e-01 1.09078538e+00
3.79481688e-02 2.10182726e-01 -9.34477076e-02 1.04923934e-01
4.00417984e-01 1.81351632e-01 4.78221774e-01 -1.40517092e+00
-5.47062457e-01 -9.02235150e-01 -6.78419888e-01 -3.19774359e-01
5.23315370e-02 -7.07382023e-01 -1.57276332e-01 -1.15245807e+00
-6.84548199e-01 -3.92100930e-01 2.71540761e-01 -3.33822191e-01
2.61524588e-01 -1.87447160e-01 6.80291206e-02 5.89833967e-02
-1.46552511e-02 7.21544921e-01 1.08636296e+00 1.84352994e-01
-3.79295111e-01 1.96131036e-01 -2.36714616e-01 9.29353416e-01
6.17696047e-01 -4.73034710e-01 -3.99688005e-01 5.61918467e-02
3.01541746e-01 1.00582443e-01 1.44907176e-01 -1.08813190e+00
2.69458175e-01 -2.07745418e-01 -9.88738388e-02 -2.70049214e-01
4.20212626e-01 -9.27221656e-01 6.22348964e-01 6.33307040e-01
-4.14209180e-02 1.52223974e-01 1.67214319e-01 7.83821791e-02
-3.47659856e-01 -8.98856401e-01 7.11017668e-01 -1.82556972e-01
-5.19604921e-01 -3.04799557e-01 -5.63813031e-01 1.60845146e-02
1.23258746e+00 -5.24882257e-01 2.19450518e-01 -2.12102816e-01
-1.03306055e+00 -1.11295074e-01 3.51795465e-01 -4.57551807e-01
1.51482895e-01 -7.82746792e-01 -4.55106765e-01 -1.34016136e-02
-4.84825611e-01 -2.15014562e-01 1.49521694e-01 1.22341466e+00
-8.94402504e-01 3.74229759e-01 -2.94558704e-01 -5.15962303e-01
-1.29446471e+00 6.96469128e-01 8.18557739e-01 -3.78987074e-01
-1.48980349e-01 5.14293373e-01 -5.61417341e-01 -1.54369578e-01
5.37903421e-03 -4.73468184e-01 -6.60196722e-01 2.65001565e-01
-1.56306654e-01 8.14010561e-01 3.67995575e-02 -1.03402984e+00
-4.34310347e-01 1.15656412e+00 7.86833584e-01 -2.00695440e-01
1.26483953e+00 1.34942442e-01 -3.82772774e-01 6.44770265e-01
1.25280952e+00 2.50079129e-02 -4.59601045e-01 1.99937031e-01
1.02336295e-01 -4.54698056e-02 1.94865212e-01 -7.19564497e-01
-6.20788872e-01 7.30898976e-01 4.92379516e-01 6.06798112e-01
1.06381595e+00 -4.78430986e-01 -6.34803250e-03 4.61108565e-01
2.76217639e-01 -9.36872482e-01 -3.85402888e-01 2.24432513e-01
1.20183671e+00 -9.85197306e-01 4.55451578e-01 -8.77325892e-01
-1.84188247e-01 1.63276982e+00 3.13119173e-01 -2.19442412e-01
8.63427401e-01 3.97234797e-01 -3.88664633e-01 -3.26387078e-01
-1.44262046e-01 -4.49142337e-01 4.43887800e-01 2.44549647e-01
4.53711987e-01 -2.31263921e-01 -1.29261315e+00 5.70314169e-01
-1.30045936e-01 3.18588555e-01 3.39313775e-01 7.24021196e-01
-2.87872285e-01 -1.11846983e+00 -7.41493940e-01 5.88338636e-02
-2.33278692e-01 8.97954851e-02 -4.15071577e-01 1.27494013e+00
5.49711347e-01 6.77116811e-01 -2.34089375e-01 6.79181218e-02
4.66233879e-01 1.22711770e-01 5.69069386e-01 -2.82789379e-01
-6.33126676e-01 -1.12476610e-01 1.88885823e-01 -2.20053136e-01
-6.88558161e-01 -7.90155351e-01 -1.03136992e+00 -2.05326527e-01
-7.25877643e-01 7.32439518e-01 1.27852154e+00 1.32547355e+00
-1.12145737e-01 4.37979698e-01 4.54620749e-01 -1.02418876e+00
-6.20574653e-01 -8.47170711e-01 -6.12664342e-01 -1.20458737e-01
2.15877011e-01 -1.05136037e+00 -6.23102307e-01 -2.24850431e-01] | [6.005269527435303, 3.5255823135375977] |
871f8762-c836-4a53-a4db-9d386561b2ed | panoptic-segmentation-of-satellite-image-time | 2107.07933 | null | https://arxiv.org/abs/2107.07933v4 | https://arxiv.org/pdf/2107.07933v4.pdf | Panoptic Segmentation of Satellite Image Time Series with Convolutional Temporal Attention Networks | Unprecedented access to multi-temporal satellite imagery has opened new perspectives for a variety of Earth observation tasks. Among them, pixel-precise panoptic segmentation of agricultural parcels has major economic and environmental implications. While researchers have explored this problem for single images, we argue that the complex temporal patterns of crop phenology are better addressed with temporal sequences of images. In this paper, we present the first end-to-end, single-stage method for panoptic segmentation of Satellite Image Time Series (SITS). This module can be combined with our novel image sequence encoding network which relies on temporal self-attention to extract rich and adaptive multi-scale spatio-temporal features. We also introduce PASTIS, the first open-access SITS dataset with panoptic annotations. We demonstrate the superiority of our encoder for semantic segmentation against multiple competing architectures, and set up the first state-of-the-art of panoptic segmentation of SITS. Our implementation and PASTIS are publicly available. | ['Loic Landrieu', 'Vivien Sainte Fare Garnot'] | 2021-07-16 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Garnot_Panoptic_Segmentation_of_Satellite_Image_Time_Series_With_Convolutional_Temporal_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Garnot_Panoptic_Segmentation_of_Satellite_Image_Time_Series_With_Convolutional_Temporal_ICCV_2021_paper.pdf | iccv-2021-1 | ['cloud-removal'] | ['computer-vision'] | [ 4.81520474e-01 -4.41432089e-01 -3.63978118e-01 -6.02463365e-01
-4.62452650e-01 -9.35839057e-01 4.66538757e-01 2.21124720e-02
-5.25511324e-01 4.94627923e-01 4.11515012e-02 -7.33645737e-01
-1.77538797e-01 -9.31352139e-01 -6.44442379e-01 -5.95275879e-01
-6.49892807e-01 3.28571498e-01 2.57944047e-01 -4.35647845e-01
-6.13269992e-02 5.58701634e-01 -1.57230818e+00 2.56000817e-01
9.38894391e-01 9.49687004e-01 7.00806141e-01 9.55458224e-01
-1.51409194e-01 1.95768058e-01 -2.24289015e-01 1.18449204e-01
4.70447063e-01 -2.55692124e-01 -9.23917234e-01 3.32040936e-01
5.25749803e-01 -4.61778551e-01 -1.87337056e-01 1.15250206e+00
1.90087691e-01 1.98943183e-01 3.37457865e-01 -9.60798979e-01
-5.39038360e-01 6.48429692e-01 -8.81468534e-01 7.08069146e-01
-4.89450395e-01 3.87868464e-01 1.29145992e+00 -1.52841955e-01
6.73628449e-01 1.22708392e+00 7.31287539e-01 9.35707428e-03
-1.37241495e+00 -5.82610250e-01 5.30935585e-01 -1.26436636e-01
-1.04475224e+00 -2.46612802e-01 1.62671849e-01 -5.76808155e-01
1.05258143e+00 4.55180198e-01 1.09024084e+00 5.41110396e-01
2.26303533e-01 1.11338949e+00 1.14453661e+00 6.54883832e-02
-6.63060546e-02 -6.90566957e-01 2.88066864e-01 4.74666059e-01
-9.52342525e-02 2.46593326e-01 4.42642942e-02 2.16867596e-01
8.58776391e-01 1.52462393e-01 -6.71863407e-02 5.56748882e-02
-1.25429809e+00 8.92350554e-01 9.01632130e-01 4.29823250e-01
-6.61758482e-01 2.53658682e-01 4.65077788e-01 2.29597330e-01
1.01960135e+00 5.53205550e-01 -9.57924843e-01 8.26294646e-02
-1.63606882e+00 2.56724387e-01 5.16209245e-01 6.87533081e-01
8.00365925e-01 -3.59778106e-02 1.09318390e-01 6.43393815e-01
1.92024797e-01 7.42003322e-01 2.62865901e-01 -1.01259339e+00
2.13385865e-01 2.91479498e-01 2.08071738e-01 -7.52632380e-01
-7.44550586e-01 -2.64640331e-01 -9.53532875e-01 9.28672850e-02
1.96512371e-01 -3.15639257e-01 -1.43102753e+00 1.61084700e+00
1.42652452e-01 3.00842166e-01 2.42988560e-02 1.21587503e+00
4.93176877e-01 1.33602774e+00 3.34578544e-01 -9.06645432e-02
1.58661759e+00 -9.29487705e-01 -5.94806671e-01 -6.69521093e-01
3.15365225e-01 -5.68114698e-01 8.12836468e-01 -2.16494665e-01
-5.14842629e-01 -5.23204565e-01 -9.36384737e-01 9.09023583e-02
-7.12456703e-01 1.08735196e-01 1.17278600e+00 1.28925130e-01
-1.19413054e+00 8.24624419e-01 -1.22468317e+00 -7.91516960e-01
5.50052881e-01 1.07930720e-01 -1.15141056e-01 4.14863318e-01
-1.28848588e+00 5.98844767e-01 7.33231485e-01 4.93135989e-01
-7.99078643e-01 -9.07238483e-01 -8.74025881e-01 1.48849964e-01
4.28381234e-01 -4.04104054e-01 1.30104518e+00 -1.27593279e+00
-1.26906812e+00 1.08750570e+00 -1.84988976e-01 -8.87032330e-01
1.35871485e-01 -4.42953527e-01 -5.12955904e-01 3.62936229e-01
4.89226371e-01 1.21269405e+00 6.62415326e-01 -8.61919940e-01
-9.60281610e-01 -4.30086583e-01 -1.33061141e-01 2.34859526e-01
-4.71787006e-02 8.32165331e-02 -2.67630190e-01 -7.82993436e-01
2.01572821e-01 -1.06976879e+00 -8.37766111e-01 2.00473428e-01
-1.54039785e-01 2.77402699e-01 7.82020509e-01 -9.26622629e-01
1.05532289e+00 -2.10584807e+00 1.30882666e-01 -1.31414458e-01
-6.30594033e-04 4.62062329e-01 -3.44076216e-01 3.09527665e-01
-1.32952869e-01 4.54673946e-01 -9.34298813e-01 1.44952133e-01
-1.79103792e-01 3.78287703e-01 -7.72904456e-01 3.46991479e-01
4.40553963e-01 1.05169475e+00 -9.45313156e-01 -3.21789920e-01
2.93392420e-01 5.80189005e-03 -2.65557021e-01 2.12986499e-01
-8.30931723e-01 5.78588784e-01 -4.83654737e-01 5.68266988e-01
9.75377917e-01 -3.53838533e-01 1.11360580e-01 -2.25138422e-02
-7.45130837e-01 1.18893877e-01 -6.10808730e-01 1.94092751e+00
-2.72357702e-01 9.16567981e-01 1.35673195e-01 -1.02852285e+00
6.02732360e-01 1.56993642e-01 7.17640460e-01 -7.43717194e-01
-3.76756698e-01 2.81771898e-01 -1.83097839e-01 -5.37648439e-01
1.00568330e+00 -1.11024175e-02 -1.59813151e-01 2.86652952e-01
-4.39775586e-02 -4.19385284e-01 3.90571207e-01 -1.38130561e-01
5.22081792e-01 6.08083487e-01 2.84822673e-01 -5.33039510e-01
7.99453482e-02 6.18141413e-01 4.09619540e-01 6.72345042e-01
-3.94891322e-01 5.89657784e-01 3.55322391e-01 -8.40334833e-01
-1.17950344e+00 -7.84376323e-01 -3.83751094e-01 1.20320082e+00
1.32155478e-01 -2.67885700e-02 -4.29324508e-01 -4.84438300e-01
2.03093454e-01 4.78968531e-01 -6.98807418e-01 6.51330113e-01
-6.15059793e-01 -1.43472636e+00 6.04129016e-01 3.92720729e-01
8.11463475e-01 -1.03289366e+00 -1.15415776e+00 5.93235135e-01
-2.80000567e-01 -1.21481609e+00 -3.45829517e-01 3.26213092e-01
-1.12609351e+00 -7.96454906e-01 -9.13590610e-01 -3.76949847e-01
5.51862903e-02 5.67083657e-01 1.31305385e+00 -4.10427809e-01
-4.55602467e-01 -7.38336965e-02 -3.48881721e-01 -1.31853878e-01
-9.18186083e-02 6.09163404e-01 -6.30195379e-01 -1.59045622e-01
3.06808472e-01 -7.17790663e-01 -8.96580637e-01 1.85659990e-01
-1.15094244e+00 4.50127602e-01 5.48426867e-01 5.57660818e-01
5.91135859e-01 -1.08060986e-01 1.88474059e-01 -7.43796349e-01
-9.29737762e-02 -7.73102880e-01 -1.24103343e+00 2.49163389e-01
-2.27302626e-01 1.68329477e-01 2.97833830e-01 -1.34139489e-02
-1.03047526e+00 2.73167998e-01 -3.07600915e-01 -1.37007192e-01
-4.46812630e-01 1.14976335e+00 3.68425906e-01 2.82896161e-01
2.89427310e-01 2.43629709e-01 -2.01703548e-01 -5.39224863e-01
7.14757383e-01 6.03761911e-01 7.63810277e-01 -1.75110370e-01
3.51827204e-01 8.11342895e-01 -1.93355158e-01 -1.17799020e+00
-1.18840265e+00 -8.85568023e-01 -8.73426139e-01 -4.64366265e-02
1.10409105e+00 -1.22274411e+00 -4.11031157e-01 7.21393049e-01
-1.08845651e+00 -7.25870013e-01 -2.29157627e-01 3.08955759e-01
-4.63207334e-01 2.72596866e-01 -5.00860572e-01 -6.23368323e-01
-4.33444649e-01 -9.33866560e-01 1.31072569e+00 2.79786706e-01
-1.69227626e-02 -8.55683029e-01 1.74295262e-01 -6.05631806e-03
5.11098742e-01 3.54759634e-01 5.08031070e-01 -1.92748100e-01
-7.40441620e-01 2.60405660e-01 -6.60198987e-01 5.40150218e-02
1.43842414e-01 1.73893183e-01 -9.40070510e-01 -2.30047941e-01
-2.52847970e-01 -2.07804665e-01 1.66202676e+00 1.07361436e+00
1.22708881e+00 -2.00919762e-01 -4.49923009e-01 1.19870555e+00
1.62837481e+00 1.78249568e-01 6.33041322e-01 5.15326679e-01
6.17695987e-01 6.51120484e-01 7.58912981e-01 4.24880981e-01
4.09926295e-01 3.36595684e-01 5.88040173e-01 -5.73031068e-01
1.81411266e-01 8.79367515e-02 9.74687636e-02 5.63127518e-01
5.39356731e-02 -5.64576566e-01 -1.22612047e+00 1.08606791e+00
-2.01980281e+00 -1.14019895e+00 -8.01120102e-02 1.77716637e+00
7.34798849e-01 -3.18575621e-01 -6.11657090e-02 -3.07071596e-01
5.87872207e-01 1.10901201e+00 -7.40633070e-01 -1.20718807e-01
-5.35972178e-01 2.95324147e-01 1.40617383e+00 4.96076226e-01
-2.08569360e+00 1.66101372e+00 6.53756046e+00 6.60545945e-01
-1.39590955e+00 5.77977262e-02 7.72359610e-01 2.86201060e-01
-1.50594383e-01 2.77107209e-01 -7.99048901e-01 3.66394788e-01
1.28462863e+00 2.08301961e-01 2.51084179e-01 3.34108204e-01
2.75056183e-01 -3.88167202e-01 -4.85954404e-01 3.69573116e-01
-6.09278500e-01 -1.46910775e+00 -8.30092281e-02 -6.22662641e-02
9.01402354e-01 8.68828177e-01 2.53059785e-03 -1.17791362e-01
7.71471620e-01 -1.00521505e+00 5.49007416e-01 4.44718689e-01
7.27669179e-01 -4.04109210e-01 3.90519410e-01 -5.29634319e-02
-1.67575228e+00 -3.78409252e-02 -2.01655105e-01 8.71637091e-02
3.12567264e-01 6.15349293e-01 -3.05282712e-01 7.71341562e-01
8.81231070e-01 1.40515256e+00 -4.49914008e-01 1.03012252e+00
6.43181652e-02 9.40554857e-01 -7.09493279e-01 5.00926137e-01
1.10166585e+00 -4.61917430e-01 5.64438343e-01 1.41216385e+00
3.98534209e-01 3.95912141e-01 3.86515409e-01 8.43602598e-01
1.66585699e-01 -1.74531490e-01 -6.58868372e-01 -6.37682557e-01
1.61627173e-01 1.09127128e+00 -1.00313008e+00 -4.46546257e-01
-3.64974231e-01 1.18367255e+00 -1.49498731e-01 5.09454489e-01
-6.68270588e-01 -1.46743685e-01 1.00804830e+00 -3.50359738e-01
6.61642551e-01 -5.36223769e-01 -2.28878081e-01 -1.27648139e+00
-3.05411786e-01 -6.53123379e-01 6.29414201e-01 -8.58754456e-01
-9.47888255e-01 4.85595107e-01 9.29755941e-02 -1.04061711e+00
-4.57165688e-02 -3.72191012e-01 -2.34341845e-01 9.39704835e-01
-2.13265681e+00 -1.53367317e+00 -2.81920969e-01 -5.07520996e-02
8.92355859e-01 2.39521086e-01 7.93240845e-01 9.55840051e-02
-3.42584193e-01 -4.06807601e-01 4.51312929e-01 6.65454641e-02
3.77112329e-01 -1.15557730e+00 1.20780981e+00 1.21812272e+00
-4.60184254e-02 -1.10102229e-01 6.40029430e-01 -6.26656950e-01
-1.07964981e+00 -1.55126619e+00 8.65785956e-01 2.31469154e-01
9.71181214e-01 -3.55406143e-02 -1.05594110e+00 9.22210872e-01
4.33181763e-01 -2.59085037e-02 2.87503839e-01 -5.32654151e-02
-1.14458337e-01 -2.03213230e-01 -6.84186161e-01 5.27124047e-01
8.61678779e-01 -4.10557836e-01 -3.32604170e-01 4.13809568e-01
9.25962031e-01 -2.81217694e-01 -8.30887735e-01 5.32989562e-01
6.25317872e-01 -7.32234418e-01 1.03806913e+00 -4.61970747e-01
6.93855047e-01 -3.58530253e-01 -2.67083466e-01 -1.43367910e+00
-4.81384307e-01 -5.82058311e-01 5.75448155e-01 7.10729241e-01
2.98627347e-01 -5.91958582e-01 4.40120339e-01 -1.15932919e-01
-3.29131514e-01 -2.04931498e-02 -6.79947078e-01 -7.19446599e-01
2.30578110e-01 -2.44059339e-01 7.47899234e-01 9.43918943e-01
-5.54080665e-01 -4.22216542e-02 -3.97225410e-01 7.04175055e-01
4.20488119e-01 8.51290345e-01 3.03399235e-01 -1.33703339e+00
3.64477374e-02 -6.78621411e-01 -1.32256165e-01 -1.49182427e+00
3.15901220e-01 -8.40177834e-01 3.60914022e-01 -1.42460704e+00
3.19036655e-02 -3.53265375e-01 -1.77626029e-01 6.61378682e-01
-2.98869759e-01 2.56565064e-01 5.08924313e-02 3.76078784e-01
-2.73010343e-01 5.34517944e-01 1.18631208e+00 -3.36818099e-01
-2.80445337e-01 -1.58893153e-01 -2.00194672e-01 2.99959958e-01
1.00833738e+00 -5.09312749e-01 -1.69510111e-01 -8.28126073e-01
1.67461768e-01 2.02508956e-01 5.34332752e-01 -7.90179431e-01
-4.85581420e-02 -5.67463994e-01 7.48035088e-02 -1.11963904e+00
1.23387175e-02 -6.27113402e-01 1.29674330e-01 4.24737364e-01
-3.20571929e-01 8.62368420e-02 8.85675669e-01 3.79721612e-01
-3.56849670e-01 8.68065283e-02 7.87388563e-01 -4.41937953e-01
-1.46908331e+00 6.48698509e-01 -6.99560285e-01 -7.25719705e-02
7.04119623e-01 9.15443078e-02 -2.83334106e-01 8.30546170e-02
-6.44654214e-01 7.43334651e-01 3.41515452e-01 6.19752169e-01
1.54199228e-01 -6.48226917e-01 -9.99976039e-01 2.85578907e-01
2.54790187e-02 6.61399886e-02 3.52107346e-01 7.37697542e-01
-9.80388820e-01 7.25215316e-01 -4.83679920e-01 -8.90358388e-01
-9.00312126e-01 3.28552842e-01 4.31181014e-01 -3.06727856e-01
-8.30699086e-01 7.38780916e-01 3.22897047e-01 -5.49198985e-01
-4.90783155e-01 -7.52539337e-01 -2.40440071e-01 5.30892491e-01
2.52287686e-01 -3.12259942e-01 -2.51290888e-01 -6.41496778e-01
-1.76262647e-01 5.10600626e-01 1.67710602e-01 -4.66996819e-01
1.76493573e+00 -3.83714139e-01 -3.79962474e-01 7.80423820e-01
9.69289780e-01 -9.03695941e-01 -1.70756352e+00 -3.03277016e-01
3.26963902e-01 -4.71423388e-01 4.67543542e-01 -8.35169256e-01
-1.39871442e+00 1.07047164e+00 8.05427969e-01 2.87186623e-01
1.35069740e+00 -2.74164468e-01 9.88759160e-01 2.62885392e-01
-1.68873854e-02 -8.52967978e-01 -6.46965504e-01 7.07213163e-01
7.02808619e-01 -1.60018015e+00 -2.26770411e-03 -2.24487990e-01
-5.27794063e-01 1.06363201e+00 1.84123218e-01 2.35291794e-02
8.06955695e-01 2.55578727e-01 1.27448127e-01 -3.14936846e-01
-6.74569786e-01 -8.77627313e-01 1.93383142e-01 2.22998783e-01
4.28668618e-01 4.93679136e-01 -2.04195138e-02 5.01692109e-03
-3.73626500e-02 2.28312105e-01 2.74605639e-02 9.27231729e-01
-5.80766141e-01 -7.00514078e-01 -2.15794176e-01 4.55244392e-01
-4.06233221e-01 -5.47898054e-01 1.84345916e-01 7.32002854e-01
-8.15168470e-02 4.86914933e-01 5.35377324e-01 3.13889310e-02
-1.68222114e-01 -5.27887829e-02 1.20258026e-01 -3.58475864e-01
-5.39220929e-01 2.51492888e-01 -2.15303414e-02 -5.40584326e-01
-9.26682889e-01 -9.52141464e-01 -9.43807423e-01 -2.61322826e-01
3.24469339e-03 5.09717688e-03 7.92007208e-01 8.91115725e-01
4.15477991e-01 4.91289735e-01 6.34709895e-01 -1.19909775e+00
-1.40359551e-01 -8.98411870e-01 -6.08780086e-01 1.15996026e-01
7.82854974e-01 -1.24027111e-01 -3.16068456e-02 4.26958442e-01] | [9.486859321594238, -1.461868166923523] |
a7bd8f60-fc94-4513-8d76-f72d1b258906 | privacy-enhancement-for-cloud-based-few-shot | 2205.07864 | null | https://arxiv.org/abs/2205.07864v2 | https://arxiv.org/pdf/2205.07864v2.pdf | Privacy Enhancement for Cloud-Based Few-Shot Learning | Requiring less data for accurate models, few-shot learning has shown robustness and generality in many application domains. However, deploying few-shot models in untrusted environments may inflict privacy concerns, e.g., attacks or adversaries that may breach the privacy of user-supplied data. This paper studies the privacy enhancement for the few-shot learning in an untrusted environment, e.g., the cloud, by establishing a novel privacy-preserved embedding space that preserves the privacy of data and maintains the accuracy of the model. We examine the impact of various image privacy methods such as blurring, pixelization, Gaussian noise, and differentially private pixelization (DP-Pix) on few-shot image classification and propose a method that learns privacy-preserved representation through the joint loss. The empirical results show how privacy-performance trade-off can be negotiated for privacy-enhanced few-shot learning. | ['Minwoo Lee', 'Liyue Fan', 'Muhammad Usama', 'Archit Parnami'] | 2022-05-10 | null | null | null | null | ['few-shot-image-classification'] | ['computer-vision'] | [ 1.11716120e-02 6.05643764e-02 -6.52331933e-02 -6.73993289e-01
-9.01620686e-01 -5.70408881e-01 3.66309077e-01 -1.76057011e-01
-3.51532429e-01 4.90298420e-01 -2.68901773e-02 -5.40964454e-02
-6.26791865e-02 -7.60249019e-01 -7.64391303e-01 -9.90817606e-01
5.19862585e-03 -5.26217043e-01 -2.45101243e-01 2.16783062e-01
1.06177717e-01 3.46990049e-01 -1.25284767e+00 2.73362488e-01
5.54106593e-01 1.17067051e+00 -2.04893604e-01 4.36836571e-01
3.75683963e-01 8.72365892e-01 -4.51616406e-01 -1.03835714e+00
1.08214366e+00 -1.85852125e-01 -3.45185935e-01 -1.75571561e-01
3.72149050e-01 -7.76144564e-01 -8.69666755e-01 1.47223318e+00
5.03892779e-01 1.68122411e-01 2.69356698e-01 -1.89584923e+00
-9.88536775e-01 1.67355075e-01 -6.00309551e-01 -6.72584251e-02
-1.78664953e-01 4.54847246e-01 5.89276910e-01 -5.13270140e-01
6.77605152e-01 7.71719038e-01 6.40240192e-01 8.08006287e-01
-1.04675257e+00 -9.50017095e-01 -3.15897465e-01 1.92943081e-01
-1.74788320e+00 -5.78207731e-01 5.82266569e-01 -3.00457895e-01
4.32773173e-01 6.96745872e-01 3.33434314e-01 1.11800182e+00
5.26205122e-01 5.55639088e-01 1.23186958e+00 -1.46432132e-01
6.26686156e-01 7.41206050e-01 7.06196055e-02 3.82906884e-01
7.08166540e-01 5.15075922e-01 -6.40110075e-01 -7.41250634e-01
2.72902489e-01 7.61003852e-01 -4.95320410e-01 -8.35088015e-01
-2.61769354e-01 8.70173991e-01 2.82678306e-01 -1.71671361e-01
-3.41744632e-01 2.76656508e-01 4.50929284e-01 5.81863940e-01
5.80515206e-01 3.02979827e-01 -4.08153445e-01 3.18595618e-02
-9.03413892e-01 1.89603344e-01 7.70655870e-01 1.31200874e+00
1.10516167e+00 -1.43514246e-01 -3.68088990e-01 4.57382858e-01
6.12723059e-04 1.30887210e-01 4.97603834e-01 -1.03960967e+00
2.82206744e-01 8.66430346e-03 1.85010761e-01 -1.10102403e+00
6.70811653e-01 6.05854690e-02 -8.31410646e-01 2.84752876e-01
-1.09969467e-01 -6.21504605e-01 -5.82934439e-01 1.89071655e+00
2.77493685e-01 4.50750709e-01 2.58048654e-01 9.60410833e-01
4.24980700e-01 4.84110653e-01 2.09902599e-01 -3.11172962e-01
1.23509085e+00 -8.22677553e-01 -9.17610347e-01 -1.96681932e-01
4.86430228e-01 -2.75866985e-01 1.01948595e+00 -1.19552463e-01
-8.41264606e-01 3.87285464e-02 -1.07677186e+00 -1.34044737e-01
-5.93213439e-01 -6.96016371e-01 7.32230246e-01 1.30357313e+00
-7.97645092e-01 7.00262785e-01 -7.33395040e-01 -1.94173530e-01
1.10137498e+00 3.19022119e-01 -7.49792874e-01 -3.57209593e-01
-1.28699875e+00 4.90283221e-01 1.47134990e-01 -4.07357633e-01
-9.14417326e-01 -1.10138869e+00 -1.06126261e+00 5.35208821e-01
4.69286561e-01 -8.45612466e-01 1.02908182e+00 -8.11084807e-01
-9.31657910e-01 1.08599198e+00 2.38015458e-01 -9.09544408e-01
8.90942037e-01 -1.45756036e-01 -3.24628085e-01 4.04776521e-02
-7.92121515e-02 3.40711892e-01 1.05989146e+00 -1.11673295e+00
-3.76477778e-01 -7.74404645e-01 3.70567590e-02 1.19288690e-01
-8.74897182e-01 4.99101281e-02 -1.83206663e-01 -6.28323555e-01
-4.06216085e-01 -8.22453916e-01 -4.75834697e-01 6.45193100e-01
-2.36738726e-01 5.24360657e-01 1.36405146e+00 -6.51550472e-01
1.01679254e+00 -2.81340051e+00 -5.59107423e-01 1.17320707e-03
3.59848499e-01 2.77035058e-01 6.18836991e-02 4.08220321e-01
2.17104897e-01 3.39366108e-01 -4.02699977e-01 -5.49146533e-01
2.06572637e-01 3.05444866e-01 -3.85447294e-01 8.18443477e-01
-2.51812637e-01 1.01509082e+00 -5.18773556e-01 -4.18114007e-01
2.40435541e-01 5.19071281e-01 -4.52963978e-01 3.36898685e-01
9.72568765e-02 -1.59438416e-01 -4.02366489e-01 5.87138236e-01
1.20235157e+00 -1.44297093e-01 -2.04029828e-01 3.18315357e-01
1.93616495e-01 -6.17269218e-01 -8.33153903e-01 1.55255723e+00
-2.42884412e-01 5.26863396e-01 3.29746842e-01 -3.37420374e-01
7.44255483e-01 3.75794768e-01 4.57181245e-01 -3.75013351e-01
3.35222542e-01 -1.60214648e-01 -4.67624545e-01 -4.90752637e-01
4.07832354e-01 -2.98911810e-01 -2.77854919e-01 6.60050035e-01
-4.89918515e-02 1.80439770e-01 -8.97061110e-01 2.25353569e-01
1.13429356e+00 -5.00373542e-01 5.27455866e-01 -7.20148161e-02
-1.16715372e-01 -3.54670912e-01 8.51215065e-01 9.27726507e-01
-8.99399102e-01 6.63925469e-01 1.81623533e-01 -5.12182176e-01
-1.06194270e+00 -7.88443625e-01 5.67196682e-02 1.01934397e+00
4.20315355e-01 -4.38549340e-01 -8.43530059e-01 -6.61277771e-01
5.52486598e-01 9.03802574e-01 -6.53441489e-01 -8.26378345e-01
1.78597301e-01 -6.46959603e-01 5.47402143e-01 1.43759012e-01
7.32264578e-01 -4.36928809e-01 -7.93470681e-01 -3.77260894e-01
3.03144872e-01 -9.92151678e-01 -7.34895289e-01 6.49475753e-02
-4.60598409e-01 -9.14458156e-01 -4.95755941e-01 -3.41974467e-01
6.51972294e-01 8.31967950e-01 2.24043131e-01 -3.72239053e-01
-4.53750700e-01 6.52940869e-01 -1.43681318e-01 -7.53312290e-01
-5.51722664e-03 -3.66227895e-01 -1.58567000e-02 4.37458038e-01
7.91651964e-01 -6.02722108e-01 -6.97537184e-01 5.25929183e-02
-1.09661806e+00 -3.17826420e-01 2.48740867e-01 7.60507762e-01
5.95526040e-01 3.78952920e-01 9.30884406e-02 -1.24561715e+00
8.00016940e-01 -6.70745850e-01 -3.61334741e-01 4.86072451e-01
-9.07694221e-01 -3.86011988e-01 6.96385145e-01 -4.85945761e-01
-1.20213294e+00 1.10088125e-01 3.94255698e-01 -1.22649884e+00
-1.24623381e-01 -2.35751420e-01 -4.81402785e-01 -7.86639750e-01
5.32387674e-01 3.84183198e-01 1.84259579e-01 -3.41170400e-01
7.27799416e-01 1.02306950e+00 4.12102908e-01 -1.74977481e-01
9.07665670e-01 6.40928149e-01 -1.68076247e-01 -7.46207476e-01
-4.68691736e-01 -4.91404951e-01 -1.49653673e-01 1.65287197e-01
6.95355833e-01 -1.15082109e+00 -5.57664037e-01 5.55890143e-01
-9.30150807e-01 3.26911509e-01 -6.20095551e-01 3.06806982e-01
-8.26729715e-01 4.63163704e-01 -5.16456366e-01 -9.73509014e-01
-6.89239442e-01 -8.02987933e-01 7.39730358e-01 2.59521574e-01
7.22821504e-02 -5.57729423e-01 -7.71204457e-02 5.00081480e-01
4.48749334e-01 4.81019765e-01 7.26168036e-01 -8.42273951e-01
-7.89942622e-01 -7.73922265e-01 -1.22457869e-01 4.36777890e-01
1.75427541e-01 -4.47565854e-01 -1.50067592e+00 -7.30901599e-01
7.55266905e-01 -3.88309181e-01 5.98176420e-01 2.26535022e-01
1.66388607e+00 -1.20103407e+00 -7.68237412e-02 1.28610778e+00
1.84902072e+00 1.49598360e-01 7.91296422e-01 7.81010017e-02
5.92684984e-01 5.63976407e-01 5.92414141e-01 8.96495104e-01
-2.94325221e-03 1.00570031e-01 5.58447123e-01 1.86232835e-01
5.17814398e-01 -4.97730583e-01 -7.18320301e-03 9.68493894e-02
4.10093665e-01 -1.26312703e-01 -4.43953604e-01 3.86319667e-01
-2.01353574e+00 -1.21544254e+00 5.30497432e-01 2.40001893e+00
5.71797252e-01 -3.73831272e-01 -1.95016563e-01 -3.67262036e-01
8.54752958e-01 6.91444218e-01 -8.30531478e-01 -5.51174641e-01
-5.08460589e-02 -1.13983519e-01 1.07227647e+00 7.01241056e-03
-1.11633015e+00 1.01493859e+00 6.28182507e+00 7.94487059e-01
-1.00789142e+00 5.76606154e-01 1.03332841e+00 -5.04649103e-01
-3.26965272e-01 1.74758226e-01 -4.51698631e-01 6.19449139e-01
7.41769373e-01 -1.04699850e+00 4.86247003e-01 1.42714608e+00
3.53430100e-02 3.67787212e-01 -1.01780915e+00 1.30546117e+00
2.22466081e-01 -1.47223318e+00 -1.49751663e-01 5.31551182e-01
8.17513347e-01 -2.34659702e-01 3.60002667e-01 8.90097097e-02
3.80864620e-01 -7.94733107e-01 3.47105950e-01 5.34136295e-01
9.89773273e-01 -9.19813514e-01 5.80326915e-01 5.44377625e-01
-5.32674909e-01 -4.07161862e-01 -8.10753286e-01 -1.66288391e-02
-1.55941337e-01 1.31545290e-01 -4.90365237e-01 2.56113797e-01
7.59985566e-01 3.81797582e-01 -2.65620410e-01 8.72171462e-01
9.04785544e-02 2.41952047e-01 3.79508249e-02 5.28449826e-02
2.37437591e-01 -2.60410875e-01 5.02155125e-01 6.99692070e-01
2.55771041e-01 4.51282084e-01 -2.89559394e-01 1.01301491e+00
-4.38736737e-01 1.80440247e-01 -1.41155708e+00 -1.52925953e-01
8.37901175e-01 1.03785527e+00 1.19686909e-01 -3.57612558e-02
-5.35022676e-01 1.33501244e+00 1.92490026e-01 3.27465534e-01
-5.84668159e-01 -4.90187556e-01 1.37953901e+00 1.13660425e-01
2.16628760e-01 1.30845144e-01 -5.11787117e-01 -1.17367542e+00
1.41815692e-01 -7.45791495e-01 5.71729720e-01 -4.90591735e-01
-1.49493182e+00 7.68684670e-02 -2.73741305e-01 -1.11615026e+00
1.07041158e-01 2.16447599e-02 -8.50005627e-01 7.27752864e-01
-1.27850699e+00 -1.17933738e+00 -1.04777589e-01 8.27435374e-01
1.39130875e-01 -4.81874496e-01 1.00559044e+00 8.30624998e-02
-5.17653108e-01 1.12020016e+00 5.30706346e-01 1.03612229e-01
6.94159806e-01 -7.31188536e-01 3.10970783e-01 9.77442801e-01
1.15398370e-01 9.18731093e-01 5.69926143e-01 -5.00346720e-01
-1.80699575e+00 -1.38574409e+00 6.38440132e-01 -6.81943670e-02
3.46408278e-01 -6.50210738e-01 -8.71696651e-01 9.39038455e-01
1.13496445e-01 6.47848964e-01 1.33285034e+00 -4.07157481e-01
-8.01753342e-01 -2.86109507e-01 -2.05211329e+00 5.49556315e-01
7.42597938e-01 -1.14381969e+00 -1.74477756e-01 2.62618244e-01
1.21432388e+00 7.71052763e-02 -6.69182241e-01 -1.15683086e-01
6.28537714e-01 -9.84765708e-01 9.08506811e-01 -1.14144492e+00
6.30025864e-02 2.11729094e-01 -7.18306661e-01 -9.62107062e-01
-3.37981761e-01 -1.07158852e+00 -3.48387361e-01 1.12402797e+00
-1.89956233e-01 -7.70956159e-01 1.03207862e+00 1.52728713e+00
5.18065333e-01 -3.40732902e-01 -1.05978978e+00 -9.79139268e-01
6.62763342e-02 -1.38023958e-01 1.08993304e+00 1.38513803e+00
-1.14208214e-01 -4.50940490e-01 -1.14089251e+00 4.12124336e-01
1.11972451e+00 -8.74991640e-02 8.70736539e-01 -7.33120561e-01
-1.45482093e-01 2.53886312e-01 -7.90772319e-01 -1.85608119e-01
3.49795401e-01 -6.01975858e-01 -3.78544748e-01 -7.10183978e-01
5.22698343e-01 -7.62433335e-02 -5.76312959e-01 5.80263674e-01
-2.53400151e-02 -1.49577498e-01 4.80292052e-01 3.89683336e-01
-4.85900015e-01 6.93885505e-01 8.59009445e-01 -9.97106954e-02
1.43546954e-01 1.13603391e-01 -1.21870029e+00 3.38843375e-01
8.45660031e-01 -7.90207982e-01 -7.15095103e-01 -1.40727192e-01
-3.95232111e-01 1.49256125e-01 4.48515087e-01 -8.64738286e-01
6.03570879e-01 -4.98176426e-01 9.14226810e-04 4.34513390e-02
4.36350673e-01 -1.38054287e+00 3.01390022e-01 4.98293936e-01
-4.35373217e-01 -4.01092976e-01 -1.77650243e-01 1.11073291e+00
-1.22515492e-01 -2.29975700e-01 1.03878653e+00 -3.25055212e-01
-7.82957375e-01 9.62274432e-01 2.00873196e-01 -6.19800538e-02
1.79790092e+00 -4.94087219e-01 -4.53042448e-01 -4.69967782e-01
-4.21994030e-01 2.07833007e-01 1.06620562e+00 3.05894405e-01
7.58771956e-01 -1.38669217e+00 -3.73975277e-01 4.37661976e-01
5.13335705e-01 -5.01317143e-01 5.77548265e-01 2.05848202e-01
-1.41218364e-01 -3.03274649e-03 -4.95422602e-01 1.36497453e-01
-1.60609472e+00 9.71123040e-01 3.11980873e-01 3.85019302e-01
-6.79828942e-01 9.05110776e-01 2.31013924e-01 -2.27904618e-01
4.00057614e-01 4.69349056e-01 6.14328086e-01 -3.50851983e-01
7.83698738e-01 5.74955702e-01 -3.14451456e-01 -2.89857566e-01
-2.65339881e-01 -6.99888319e-02 -3.06588829e-01 2.64391489e-02
1.33908916e+00 -3.71987700e-01 -3.79768084e-03 1.96181446e-01
1.71752572e+00 -3.82089913e-01 -1.53501403e+00 -2.80870467e-01
-3.07534814e-01 -1.51561022e+00 2.62491763e-01 -4.43519920e-01
-1.08804202e+00 8.82691741e-01 8.14688146e-01 6.23955354e-02
9.66900110e-01 -3.63127619e-01 1.01724660e+00 1.91161454e-01
7.21269906e-01 -1.14987338e+00 -4.47562844e-01 -2.23387375e-01
3.30203980e-01 -1.49169970e+00 -1.08805329e-01 -4.06953901e-01
-9.94024634e-01 4.37315673e-01 5.34225404e-01 5.75998276e-02
1.01578331e+00 3.19712043e-01 9.69300885e-03 -7.24941194e-02
-9.89205599e-01 5.06125629e-01 -2.60456532e-01 9.01700795e-01
-2.95245200e-01 4.62041557e-01 -1.84980318e-01 9.72180724e-01
-3.09909955e-02 1.42610148e-01 5.14536917e-01 1.27552974e+00
-2.13483974e-01 -7.82871664e-01 -2.59178489e-01 5.17771125e-01
-5.61946690e-01 5.36059327e-02 -5.95732987e-01 3.12260836e-01
1.75106898e-01 7.60137200e-01 -4.90253232e-02 -5.86228669e-01
5.60809709e-02 1.30825276e-02 4.43241075e-02 -5.39093614e-01
-6.45080447e-01 -3.84059429e-01 -3.25261980e-01 -7.25665808e-01
1.20021477e-01 -4.99373406e-01 -6.86733723e-01 -7.86141276e-01
-1.69006675e-01 1.99691698e-01 6.46296501e-01 5.59934199e-01
7.91969001e-01 -3.14810365e-01 1.21053660e+00 -1.01219406e-02
-1.27671695e+00 -2.82524794e-01 -1.28842556e+00 6.29740655e-01
3.93403560e-01 -2.82277912e-02 -6.53863609e-01 -2.26233974e-01] | [5.903683662414551, 6.811548233032227] |
1ce120f4-bf66-4db9-818e-11820baa33b6 | creating-annotated-dialogue-resources-cross | null | null | https://aclanthology.org/L16-1017 | https://aclanthology.org/L16-1017.pdf | Creating Annotated Dialogue Resources: Cross-domain Dialogue Act Classification | This paper describes a method to automatically create dialogue resources annotated with dialogue act information by reusing existing dialogue corpora. Numerous dialogue corpora are available for research purposes and many of them are annotated with dialogue act information that captures the intentions encoded in user utterances. Annotated dialogue resources, however, differ in various respects: data collection settings and modalities used, dialogue task domains and scenarios (if any) underlying the collection, number and roles of dialogue participants involved and dialogue act annotation schemes applied. The presented study encompasses three phases of data-driven investigation. We, first, assess the importance of various types of features and their combinations for effective cross-domain dialogue act classification. Second, we establish the best predictive model comparing various cross-corpora training settings. Finally, we specify models adaptation procedures and explore late fusion approaches to optimize the overall classification decision taking process. The proposed methodology accounts for empirically motivated and technically sound classification procedures that may reduce annotation and training costs significantly. | ['Volha Petukhova', 'Dilafruz Amanova', 'Dietrich Klakow'] | 2016-05-01 | creating-annotated-dialogue-resources-cross-1 | https://aclanthology.org/L16-1017 | https://aclanthology.org/L16-1017.pdf | lrec-2016-5 | ['sound-classification', 'dialogue-act-classification'] | ['audio', 'natural-language-processing'] | [ 5.75181484e-01 6.06645703e-01 4.46598511e-03 -7.71834552e-01
-6.34862304e-01 -7.75094867e-01 1.30970061e+00 3.45436424e-01
-5.97217143e-01 1.24389744e+00 6.57693386e-01 -4.23276844e-03
-1.38846934e-01 -3.97383124e-01 3.28929693e-01 -4.97665554e-01
2.70690054e-01 1.03675771e+00 8.49958956e-02 -5.86962461e-01
5.86500764e-01 3.57751071e-01 -1.29997444e+00 6.23014212e-01
9.08990562e-01 7.59316921e-01 1.12887181e-01 1.05170155e+00
-4.27643955e-01 1.19523191e+00 -9.56767976e-01 -5.79895973e-01
-6.81297481e-02 -8.85117054e-01 -1.59079504e+00 5.09451866e-01
-1.91648945e-01 -3.03156495e-01 -4.86486219e-02 6.15079045e-01
6.31830335e-01 4.31473166e-01 9.62319136e-01 -1.04403079e+00
1.13242052e-01 7.73846984e-01 1.57315403e-01 9.99867916e-02
9.63072956e-01 1.68476671e-01 9.06094015e-01 -4.99928355e-01
6.69349194e-01 1.21387327e+00 4.40407604e-01 1.00909019e+00
-1.17646968e+00 -1.04979120e-01 -4.87180054e-01 -8.76480713e-02
-7.11098909e-01 -9.50092971e-01 8.50724041e-01 -5.01342475e-01
1.18155098e+00 4.44870263e-01 5.07189214e-01 1.35370398e+00
-3.64872336e-01 9.18887794e-01 1.35380375e+00 -1.06757963e+00
9.33650136e-02 7.06690431e-01 4.93235022e-01 5.74843585e-01
-5.21064579e-01 -3.74971688e-01 -5.13662696e-01 -4.94354814e-01
4.97305334e-01 -8.32835793e-01 -2.48609707e-01 -6.91986158e-02
-1.00620472e+00 1.05343390e+00 -4.83844399e-01 8.90478849e-01
-3.26538235e-01 -7.89829910e-01 1.15552390e+00 3.80542517e-01
4.57474530e-01 7.59358466e-01 -7.87217736e-01 -6.70647621e-01
-4.57951516e-01 3.74325007e-01 1.40208948e+00 9.12367225e-01
6.00541234e-01 -2.25604132e-01 -3.69606555e-01 1.40671062e+00
2.60791868e-01 -1.82757348e-01 5.47560215e-01 -1.02296209e+00
7.55985439e-01 8.08174789e-01 2.05806002e-01 -5.39356172e-01
-6.91886425e-01 6.23241484e-01 -5.64625144e-01 -3.30454707e-01
7.92403877e-01 -6.22925460e-01 -1.31351605e-01 1.38437033e+00
3.94909054e-01 -6.91593707e-01 7.41005003e-01 4.22982454e-01
1.03886044e+00 5.17337739e-01 3.41849238e-01 -6.39396012e-01
1.66289353e+00 -7.56768227e-01 -1.19908571e+00 3.05209961e-02
9.54680979e-01 -9.06217933e-01 9.61387217e-01 1.86086193e-01
-1.29976749e+00 -4.93772209e-01 -6.48188651e-01 -8.29089209e-02
-4.14923817e-01 3.73136610e-01 5.60108364e-01 8.70969892e-01
-6.54584944e-01 3.70436043e-01 -2.68536985e-01 -6.31648600e-01
-3.00917894e-01 5.14107287e-01 -3.23394179e-01 5.60948968e-01
-1.35882509e+00 1.42886257e+00 8.43219876e-01 8.79145488e-02
-1.98275089e-01 1.59072410e-02 -9.29321527e-01 -1.92835718e-01
3.10774595e-01 -2.76366532e-01 1.59730387e+00 -1.08009863e+00
-1.98480546e+00 1.15035462e+00 1.17011398e-01 -5.94270885e-01
6.30508363e-01 2.29157340e-02 -3.11692804e-01 1.24973722e-01
-1.63249403e-01 4.74884689e-01 2.94667065e-01 -1.14846528e+00
-8.00814331e-01 -5.04007638e-01 3.68165880e-01 8.14765871e-01
-2.32404575e-01 3.72872293e-01 -1.90144435e-01 -2.59553879e-01
-3.71248782e-01 -7.73960471e-01 -1.19706273e-01 -7.18982518e-01
-3.08875203e-01 -5.76223850e-01 6.14739478e-01 -7.63268054e-01
1.16757798e+00 -1.80843747e+00 3.51832777e-01 2.80672032e-02
6.25326186e-02 4.19171453e-01 4.07595336e-01 8.48525047e-01
3.52224410e-01 -1.50764853e-01 -2.12208524e-01 -3.94243509e-01
1.93817586e-01 4.93306130e-01 7.44151399e-02 1.90327555e-01
-2.48946361e-02 4.02022749e-01 -5.28257489e-01 -9.28182483e-01
7.78390527e-01 1.45787045e-01 -2.89061368e-01 7.93888748e-01
-3.32217038e-01 7.80595481e-01 -7.23509192e-01 1.98464558e-01
6.71914965e-02 2.86666423e-01 6.35198355e-01 -2.50861228e-01
-2.41494849e-01 6.07199073e-01 -8.60722125e-01 1.57012463e+00
-6.90139771e-01 6.92439675e-01 2.48212054e-01 -1.14523351e+00
1.08194542e+00 8.59529436e-01 6.38795257e-01 -2.49300659e-01
4.62891519e-01 2.15639770e-01 2.13704675e-01 -8.20999622e-01
9.42917526e-01 -1.19051479e-01 -3.92887950e-01 4.32658702e-01
5.31771421e-01 -1.06016219e-01 4.95266408e-01 -1.67916104e-01
6.49918258e-01 1.60495877e-01 1.01816678e+00 -1.42848447e-01
1.14383161e+00 3.12540472e-01 2.38600805e-01 5.42119145e-01
-4.62760210e-01 -3.43212066e-03 6.77311599e-01 -4.60720450e-01
-9.76156771e-01 -1.76481664e-01 -3.33490431e-01 1.38507462e+00
-2.31336713e-01 -2.38679558e-01 -9.65443969e-01 -8.22081089e-01
-6.56758487e-01 7.89793313e-01 -3.16180766e-01 2.62914419e-01
-7.43152916e-01 -7.70908296e-01 1.03563011e+00 9.21373591e-02
7.77801752e-01 -1.52422738e+00 -7.57648349e-01 4.93404120e-01
-5.70634305e-01 -1.18829119e+00 8.51495042e-02 3.18381816e-01
-6.73397064e-01 -1.30469954e+00 -3.48096251e-01 -6.43810570e-01
4.07967895e-01 -3.83673787e-01 1.10677016e+00 -9.90677997e-02
-2.00837292e-02 6.44504964e-01 -7.97928512e-01 -2.76666254e-01
-1.22774768e+00 2.20002562e-01 -2.86129832e-01 -9.49392691e-02
6.44731402e-01 3.63223031e-02 1.28595978e-01 3.20169419e-01
-4.90688235e-01 1.55524090e-01 2.05816284e-01 1.11683977e+00
-2.47821331e-01 -6.69433624e-02 5.78551292e-01 -1.35807145e+00
1.42348301e+00 -2.56405145e-01 -1.94169983e-01 5.33311427e-01
-3.40189964e-01 1.80434495e-01 4.11058277e-01 -3.30046378e-02
-1.80296540e+00 -1.05000054e-02 -3.78253162e-01 3.90863121e-01
-9.26683903e-01 9.07202363e-02 -4.47329700e-01 1.34133697e-01
7.92763293e-01 6.81356937e-02 1.15069062e-01 -3.48263800e-01
1.81659982e-01 1.11510217e+00 1.31892800e-01 -9.18307960e-01
-7.59662241e-02 -2.41397858e-01 -4.75833744e-01 -1.20718515e+00
-4.31950390e-01 -7.02301502e-01 -1.20333385e+00 -5.82640529e-01
1.10759866e+00 -4.86573547e-01 -8.35285366e-01 4.83918011e-01
-1.36418712e+00 -1.30076304e-01 -1.84359983e-01 4.20885742e-01
-9.28534567e-01 4.16890562e-01 -7.44347513e-01 -1.40063345e+00
-3.73871684e-01 -1.07557714e+00 7.63045490e-01 2.70094067e-01
-8.45739424e-01 -1.39941859e+00 1.34937420e-01 8.64121437e-01
3.23924609e-03 6.76317289e-02 8.41140926e-01 -1.51645875e+00
3.82928878e-01 -8.73167813e-02 9.89436135e-02 2.67496169e-01
2.60283738e-01 -4.28928882e-02 -1.15584636e+00 9.72118601e-02
9.90800560e-02 -8.17241609e-01 1.01543523e-01 4.94472496e-02
4.25549209e-01 -7.50909269e-01 -8.35952088e-02 -4.25874770e-01
7.90832698e-01 7.09115386e-01 4.10870999e-01 1.28711239e-01
2.54714102e-01 1.44755244e+00 8.95794213e-01 7.31774271e-01
2.06433252e-01 9.40179765e-01 -2.51078248e-01 6.17119297e-02
2.37193674e-01 6.36223629e-02 2.84385800e-01 8.91873479e-01
-4.40573514e-01 -4.11836654e-01 -8.64099145e-01 3.34517151e-01
-1.83417714e+00 -1.14299905e+00 -1.39564916e-01 1.79260290e+00
1.20191121e+00 3.35455872e-03 6.01154268e-01 4.16700631e-01
7.96893775e-01 2.42146268e-01 -1.65894739e-02 -7.72710562e-01
-2.36415118e-03 -2.58214064e-02 1.20391585e-01 7.76534140e-01
-1.19519055e+00 7.28042543e-01 6.13887310e+00 4.89212364e-01
-4.52401400e-01 4.09698822e-02 4.38898116e-01 5.51287174e-01
1.78094164e-01 -6.14718795e-02 -7.77180791e-01 2.08280429e-01
1.17063308e+00 -1.95999265e-01 2.17521310e-01 5.62998295e-01
1.56716079e-01 -2.24931180e-01 -1.26029134e+00 5.33056498e-01
1.62474722e-01 -1.13537681e+00 -2.56212771e-01 1.30108401e-01
2.77411580e-01 -5.49683094e-01 -6.91031337e-01 4.35774952e-01
4.19563383e-01 -5.41974008e-01 2.34628633e-01 2.27980122e-01
3.33717674e-01 -5.05729496e-01 9.86559093e-01 5.38841724e-01
-7.93231428e-01 4.20848280e-02 -2.81423125e-02 -8.72761756e-02
1.85563713e-01 -1.98157057e-01 -1.46275055e+00 5.78447342e-01
6.97004572e-02 2.23560661e-01 -3.67915571e-01 3.97403389e-01
5.99130169e-02 3.84980887e-01 -9.66223478e-02 -4.62359369e-01
2.10439041e-01 -4.24329937e-01 5.08400261e-01 1.59869576e+00
-2.11382374e-01 3.55325073e-01 3.32099348e-01 3.67843091e-01
1.52420431e-01 6.33889019e-01 -8.06183040e-01 -6.52711168e-02
5.41113555e-01 1.16573310e+00 -5.91022611e-01 -2.64931500e-01
-5.13755262e-01 7.39024699e-01 2.55195610e-02 -4.98733111e-02
-4.45544988e-01 -2.00918078e-01 3.23421389e-01 -2.42512912e-01
-4.38304424e-01 7.97160566e-02 -1.90618157e-01 -1.10958326e+00
-2.27372676e-01 -1.12041116e+00 6.34736717e-01 -2.11912349e-01
-1.21163309e+00 8.12951326e-01 4.46905106e-01 -1.09228683e+00
-1.02632844e+00 -6.35907233e-01 -3.92360449e-01 7.49080122e-01
-7.08993554e-01 -1.14376426e+00 1.71791632e-02 5.78691363e-01
1.18127263e+00 -7.20325828e-01 1.43943930e+00 1.30672961e-01
-3.99524599e-01 2.51862288e-01 -1.48840651e-01 4.69688684e-01
6.66228175e-01 -1.27881312e+00 -3.33222985e-01 1.70450091e-01
-2.05202051e-03 3.94864619e-01 8.28796566e-01 -4.14322793e-01
-7.90361762e-01 -2.82153785e-01 1.16736674e+00 -2.50445068e-01
5.71203768e-01 -2.05998108e-01 -8.26082468e-01 4.35561061e-01
8.45744550e-01 -7.85712659e-01 1.08107126e+00 4.69277799e-01
2.09351435e-01 3.77472878e-01 -1.47743177e+00 4.33561385e-01
4.11599785e-01 -6.47165239e-01 -1.13015473e+00 3.64067793e-01
2.94294562e-02 -5.37547886e-01 -1.20707452e+00 2.06902072e-01
3.48215133e-01 -1.08671284e+00 5.61989903e-01 -7.26955235e-01
2.13202149e-01 3.53426009e-01 -9.35599282e-02 -1.04344881e+00
3.84047657e-01 -7.94653952e-01 1.47070646e-01 1.64116919e+00
5.24195194e-01 -3.15719068e-01 6.16757214e-01 1.16849303e+00
-2.04706751e-02 -2.69306779e-01 -8.24731112e-01 6.84599802e-02
-5.84363230e-02 -1.04319684e-01 2.40528032e-01 1.16263640e+00
7.58495390e-01 9.96240258e-01 -5.35188675e-01 -5.68754077e-01
2.06462845e-01 -1.80708259e-01 8.78467679e-01 -1.30507123e+00
-2.23683327e-01 -4.97620076e-01 -1.12226829e-01 -7.63124645e-01
3.92869622e-01 -3.47340465e-01 9.13852006e-02 -1.27675617e+00
-2.29527637e-01 -4.15949821e-01 5.26768565e-01 3.57350081e-01
-1.11428320e-01 -4.69524354e-01 1.62716731e-01 3.35713059e-01
-3.46021831e-01 4.87246305e-01 9.07428622e-01 4.23785374e-02
-5.46177804e-01 3.59091461e-01 -2.36580938e-01 8.94436479e-01
9.62507248e-01 9.90576074e-02 -5.63411653e-01 1.00339800e-01
-4.95279163e-01 7.64711618e-01 -1.57025829e-01 -6.19043469e-01
1.76225826e-01 -3.01026285e-01 8.73742551e-02 -3.80609244e-01
6.06768131e-01 -8.68116260e-01 -7.13091418e-02 1.36757374e-01
-1.06422925e+00 -1.12903371e-01 5.53831756e-02 2.41489246e-01
-4.13328141e-01 -9.67107832e-01 8.93727422e-01 -4.17268306e-01
-6.09148443e-01 -5.02108753e-01 -8.93766999e-01 2.25343719e-01
1.20695198e+00 -3.41825426e-01 -8.36023912e-02 -4.39906180e-01
-9.90080476e-01 3.86345014e-02 1.04877248e-01 2.23543629e-01
1.20152220e-01 -9.59825218e-01 -6.36879444e-01 -9.21617001e-02
1.14443667e-01 -2.45615274e-01 1.42455861e-01 4.80543792e-01
-4.00951743e-01 5.86155832e-01 -6.17561281e-01 -2.70866811e-01
-1.74193954e+00 -7.32238144e-02 4.00829673e-01 -8.32387567e-01
-2.43379414e-01 4.96624529e-01 -2.04226077e-01 -8.71842742e-01
3.86777103e-01 2.49099687e-01 -9.34617996e-01 5.76910973e-01
2.09162146e-01 5.09989679e-01 -1.88762341e-02 -1.00432050e+00
9.22018588e-02 -1.40893832e-01 -9.30381268e-02 -4.95992005e-01
1.02831614e+00 -4.02451605e-01 -2.77851149e-02 7.94451118e-01
8.28135431e-01 -3.35938305e-01 -8.96471679e-01 -3.56834650e-01
4.24521208e-01 -2.36164600e-01 -4.20428216e-01 -8.91715467e-01
-3.63124937e-01 7.10262418e-01 2.22313970e-01 8.63633394e-01
1.00613856e+00 -1.31364152e-01 3.03977072e-01 6.00059807e-01
3.38827789e-01 -1.71368635e+00 -1.41483083e-01 7.48791993e-01
9.39399302e-01 -1.37897849e+00 8.25007819e-03 -4.34881538e-01
-1.42798328e+00 1.37651300e+00 9.00024593e-01 4.11089450e-01
2.62313157e-01 2.30433688e-01 1.62756771e-01 -3.20794225e-01
-7.94790506e-01 -2.31371507e-01 2.81146076e-02 7.23961949e-01
9.88036633e-01 -2.70038154e-02 -9.58978415e-01 3.91622156e-01
-1.85294926e-01 -2.91266292e-01 5.25150120e-01 1.06627643e+00
-3.16427648e-01 -1.51305521e+00 -3.55817467e-01 4.43475693e-01
-5.16985178e-01 2.51672536e-01 -1.06542385e+00 1.02968049e+00
-1.18962020e-01 1.27455473e+00 -1.59060642e-01 -2.61440516e-01
4.66399610e-01 5.22496045e-01 4.68688577e-01 -7.23573148e-01
-1.32453644e+00 1.94335058e-01 1.20406508e+00 1.35501012e-01
-1.16936362e+00 -9.20297325e-01 -1.09248996e+00 -2.15783283e-01
-6.77456021e-01 7.35230029e-01 4.86913711e-01 1.37282169e+00
-2.42751285e-01 3.00856888e-01 6.03618562e-01 -8.05819690e-01
-5.57030618e-01 -1.60607672e+00 -3.47039640e-01 5.32361567e-01
-1.34334549e-01 -5.63320994e-01 -8.28454942e-02 4.03564781e-01] | [12.899742126464844, 7.927119731903076] |
b787ae5f-1184-469d-bedc-f29602af4313 | deep-coevolutionary-network-embedding-user | 1609.03675 | null | http://arxiv.org/abs/1609.03675v4 | http://arxiv.org/pdf/1609.03675v4.pdf | Deep Coevolutionary Network: Embedding User and Item Features for Recommendation | Recommender systems often use latent features to explain the behaviors of
users and capture the properties of items. As users interact with different
items over time, user and item features can influence each other, evolve and
co-evolve over time. The compatibility of user and item's feature further
influence the future interaction between users and items. Recently, point
process based models have been proposed in the literature aiming to capture the
temporally evolving nature of these latent features. However, these models
often make strong parametric assumptions about the evolution process of the
user and item latent features, which may not reflect the reality, and has
limited power in expressing the complex and nonlinear dynamics underlying these
processes. To address these limitations, we propose a novel deep coevolutionary
network model (DeepCoevolve), for learning user and item features based on
their interaction graph. DeepCoevolve use recurrent neural network (RNN) over
evolving networks to define the intensity function in point processes, which
allows the model to capture complex mutual influence between users and items,
and the feature evolution over time. We also develop an efficient procedure for
training the model parameters, and show that the learned models lead to
significant improvements in recommendation and activity prediction compared to
previous state-of-the-arts parametric models. | ['Yichen Wang', 'Hanjun Dai', 'Rakshit Trivedi', 'Le Song'] | 2016-09-13 | null | null | null | null | ['activity-prediction', 'activity-prediction'] | ['computer-vision', 'time-series'] | [-4.54403430e-01 -5.25216043e-01 -3.07336032e-01 -1.32505581e-01
6.24504566e-01 -6.16301656e-01 7.58791089e-01 -1.16115108e-01
8.96823332e-02 3.58970761e-01 3.56475890e-01 1.58667043e-01
-5.57283163e-01 -1.03860688e+00 -5.80221474e-01 -6.51295722e-01
-6.45951152e-01 5.20425737e-01 -6.80853333e-03 -4.89537448e-01
-1.38754118e-02 1.95684925e-01 -1.53531861e+00 1.88494865e-02
9.54952776e-01 6.93551838e-01 1.74032629e-01 6.99346900e-01
-2.68884689e-01 5.25335133e-01 -5.06285131e-01 -3.79317373e-01
1.69207484e-01 -5.55875778e-01 -2.01979443e-01 2.19635721e-02
-2.31064141e-01 -2.88688913e-02 -7.45664358e-01 7.71277130e-01
7.46887550e-02 4.14638638e-01 1.02942419e+00 -1.25232923e+00
-1.58150423e+00 7.64724851e-01 -1.81839824e-01 3.49167466e-01
1.26136914e-01 1.10662989e-01 1.20466840e+00 -6.33123100e-01
3.32310349e-01 1.06581581e+00 8.87631595e-01 4.31891084e-01
-1.10207891e+00 -5.25073171e-01 6.55068457e-01 2.06297159e-01
-1.09490347e+00 2.45513037e-01 9.32962239e-01 -5.22793174e-01
8.55411172e-01 1.51475415e-01 1.55112743e+00 1.64906538e+00
6.75687611e-01 9.85648334e-01 4.61585850e-01 -9.93905887e-02
-2.26156842e-02 2.12525293e-01 4.07557994e-01 4.19692278e-01
1.77816182e-01 3.30341816e-01 -3.86091590e-01 -3.73890787e-01
1.19268560e+00 7.34027207e-01 -1.82766676e-01 -3.10611904e-01
-6.57628477e-01 9.30723846e-01 4.31014955e-01 5.99546611e-01
-8.46781433e-01 1.28346547e-01 -9.65114534e-02 6.88928962e-01
6.49107516e-01 4.64268625e-01 -4.51065958e-01 -2.86717266e-01
-5.84167123e-01 1.51026681e-01 1.04513550e+00 6.88372016e-01
4.48250413e-01 5.68599850e-02 -2.03141719e-01 8.51611078e-01
5.26185453e-01 3.42888325e-01 7.04702914e-01 -3.02152187e-01
-1.60249725e-01 7.02261746e-01 2.57321626e-01 -1.26319051e+00
-4.12514091e-01 -9.80786264e-01 -1.10746372e+00 -4.30443078e-01
9.65344310e-02 -1.97360769e-01 -8.21977079e-01 1.75125229e+00
7.93653429e-02 7.68230140e-01 -1.29317820e-01 5.81612885e-01
5.07703543e-01 1.08233666e+00 6.53650016e-02 -4.42837507e-01
8.61669540e-01 -1.01043320e+00 -9.99740720e-01 2.20326453e-01
3.72140110e-01 -3.82659376e-01 9.57299590e-01 2.55670965e-01
-1.18921208e+00 -8.18510175e-01 -9.24649596e-01 4.71257567e-01
-4.07111734e-01 -2.79670954e-01 1.01326668e+00 3.87703240e-01
-1.05745399e+00 1.00707769e+00 -9.28419113e-01 -3.31355661e-01
1.79716244e-01 6.49042189e-01 4.13080037e-01 3.98175597e-01
-1.69429100e+00 5.65554261e-01 -1.08476385e-01 4.03292805e-01
-7.26055503e-01 -8.49297822e-01 -2.24944413e-01 5.30202568e-01
6.90121800e-02 -1.07154906e+00 1.08805454e+00 -1.27744544e+00
-1.84827018e+00 -1.66300192e-01 1.21337429e-01 -4.10238475e-01
3.12358528e-01 -4.03413475e-01 -9.44353640e-01 -6.03475988e-01
-6.58689380e-01 5.14288433e-02 8.78513277e-01 -8.84238183e-01
-6.58164203e-01 -1.32085010e-01 2.17124987e-02 5.35855055e-01
-8.05089533e-01 -3.25562119e-01 -6.35701597e-01 -6.21286571e-01
-3.59446079e-01 -1.11367869e+00 -2.14744687e-01 -3.12018394e-01
2.28464246e-01 -5.67376733e-01 8.37752879e-01 -1.51977643e-01
1.49637473e+00 -2.25509381e+00 4.08137769e-01 2.37020150e-01
2.60025978e-01 3.46170485e-01 -2.94036359e-01 6.56399488e-01
2.42233515e-01 -2.84869177e-03 3.55613649e-01 -2.45285198e-01
1.69334244e-02 5.87143004e-01 -3.08665723e-01 3.51701438e-01
-1.12167537e-01 1.38286901e+00 -1.00571716e+00 3.30905646e-01
-2.90017258e-02 9.73520696e-01 -6.67704999e-01 2.62969673e-01
-3.35747480e-01 3.63068312e-01 -5.81253529e-01 3.24458517e-02
3.08105588e-01 -7.28224933e-01 2.00381488e-01 1.28978372e-01
-2.57697236e-02 1.28556371e-01 -1.11197293e+00 1.22925055e+00
-4.80331898e-01 4.54677969e-01 -3.16157579e-01 -7.37848222e-01
8.87028754e-01 4.76025879e-01 7.32332110e-01 -6.61660731e-01
9.39132199e-02 1.89026091e-02 2.77552038e-01 -2.67438978e-01
4.83154655e-01 -3.24494243e-02 2.29794964e-01 6.57169163e-01
2.33892366e-01 3.62916559e-01 -1.07036233e-01 5.80452271e-02
9.12516534e-01 7.84407631e-02 -1.30202873e-02 -1.03762023e-01
3.17753524e-01 -6.88460529e-01 3.55831295e-01 8.10801804e-01
8.74899104e-02 1.29142674e-02 3.26154262e-01 -5.88150620e-01
-9.27125096e-01 -8.98040175e-01 7.00970590e-02 1.29053116e+00
2.55441457e-01 -3.14003348e-01 -1.54120862e-01 -4.09424096e-01
2.40460739e-01 7.56277621e-01 -1.10569823e+00 -6.20516241e-01
-3.86424452e-01 -1.00132644e+00 -9.01025608e-02 5.20811021e-01
1.02427416e-01 -1.23200691e+00 -3.37721370e-02 5.67582309e-01
1.56873375e-01 -3.42370719e-01 -8.17092896e-01 1.04269132e-01
-9.19725120e-01 -6.87466502e-01 -7.26336300e-01 -5.74811697e-01
4.67164099e-01 1.73205093e-01 1.02149665e+00 5.61925136e-02
1.36139346e-02 7.12882698e-01 -4.60396588e-01 -2.92946458e-01
-3.09421986e-01 8.07195231e-02 4.38108772e-01 4.66074675e-01
3.34137559e-01 -1.05804384e+00 -1.01543868e+00 5.37201226e-01
-1.03151858e+00 -1.36299223e-01 4.40725118e-01 8.65727305e-01
3.57836515e-01 2.88179904e-01 5.17676771e-01 -1.04787731e+00
1.12023199e+00 -9.54959512e-01 -2.23906472e-01 4.65709120e-01
-1.11466396e+00 1.08732872e-01 7.64959157e-01 -1.26131380e+00
-1.07200754e+00 -5.27798712e-01 9.80496928e-02 -7.20212102e-01
2.74607867e-01 7.75083125e-01 2.49212131e-01 3.15014571e-01
3.19137037e-01 3.79642814e-01 -2.40017116e-01 -3.88118654e-01
6.69349551e-01 3.67753208e-01 1.40179396e-01 -1.53944567e-01
7.53788650e-01 3.48708391e-01 -2.51724184e-01 -5.04620194e-01
-8.20327163e-01 -5.53074181e-01 -6.89247310e-01 -1.98484138e-01
3.47718149e-01 -6.60264194e-01 -7.36145318e-01 5.58912635e-01
-8.62098157e-01 -3.87328386e-01 -7.08492875e-01 6.32938266e-01
-4.15219158e-01 4.62703481e-02 -9.97774065e-01 -9.17261004e-01
-5.13925731e-01 -4.06861395e-01 3.66280496e-01 4.93834257e-01
-1.96154252e-01 -1.75036287e+00 5.33246994e-01 -4.86512005e-01
8.31858575e-01 -1.99042574e-01 9.58753645e-01 -7.18191087e-01
-3.87639523e-01 -3.72890562e-01 3.65886450e-01 9.54918042e-02
5.25111556e-01 3.05818647e-01 -1.55681640e-01 -3.79361719e-01
2.61001527e-01 4.57570463e-01 5.30951917e-01 7.64265776e-01
5.93310714e-01 -4.98820364e-01 -5.61178923e-01 6.67489231e-01
1.07947719e+00 4.55330849e-01 3.78542811e-01 -4.87683676e-02
7.29985297e-01 4.01165038e-01 1.87945411e-01 5.98223388e-01
2.56345540e-01 6.55171931e-01 1.79739162e-01 -9.75745469e-02
2.31961101e-01 -6.28579021e-01 5.43329298e-01 1.65701640e+00
-3.84418756e-01 -6.32753253e-01 -2.83686310e-01 2.84290999e-01
-2.30063796e+00 -1.14825141e+00 -2.54293531e-01 1.94309211e+00
3.75403494e-01 5.53417355e-02 3.75228047e-01 -5.13956368e-01
6.35385573e-01 -7.24184439e-02 -1.03857327e+00 -7.02106655e-02
-2.28968069e-01 -9.39212590e-02 2.09718093e-01 3.39115649e-01
-6.66322708e-01 7.56073952e-01 6.80883789e+00 3.34211379e-01
-1.19522262e+00 1.28354907e-01 2.99284041e-01 -4.00989920e-01
-5.67203462e-01 -1.96607336e-01 -6.95437670e-01 6.04358554e-01
1.28885627e+00 -4.63427275e-01 5.67184091e-01 4.34459150e-01
2.40437329e-01 6.58098757e-01 -1.03174198e+00 7.64307559e-01
-2.99191847e-02 -1.13521433e+00 2.04338387e-01 4.49149787e-01
1.00556314e+00 6.82005286e-02 7.94461489e-01 7.08740473e-01
7.56797254e-01 -8.54835570e-01 2.93513238e-01 1.24901354e+00
2.08546191e-01 -7.84112036e-01 5.64920664e-01 5.53662419e-01
-1.30735433e+00 -5.18464684e-01 -5.13180852e-01 -1.65793598e-01
3.22440624e-01 3.39601249e-01 -3.92285407e-01 4.32906836e-01
5.55799961e-01 1.24859226e+00 -2.80990064e-01 1.01859057e+00
-7.10332021e-02 1.07544136e+00 -2.26664945e-01 -4.85501200e-01
2.65005499e-01 -7.19875932e-01 6.89875305e-01 8.79802883e-01
7.23766923e-01 -1.60648033e-03 4.13360372e-02 1.03620827e+00
5.02893552e-02 3.56850326e-01 -4.16640759e-01 -4.68006045e-01
1.12398490e-01 1.22322738e+00 -5.64629257e-01 -2.38521487e-01
-5.44896781e-01 1.08261895e+00 3.40741962e-01 6.54986858e-01
-9.54214573e-01 2.08470255e-01 7.78592110e-01 3.51722419e-01
6.09335899e-01 -3.39897811e-01 2.93870240e-01 -1.45713365e+00
-3.01200122e-01 -4.64289665e-01 3.56911749e-01 -4.68606412e-01
-1.90264511e+00 6.74145579e-01 -3.43125343e-01 -1.20013666e+00
-3.07395875e-01 -3.26829672e-01 -8.29740882e-01 8.43660533e-01
-1.21092832e+00 -1.09464586e+00 6.25857562e-02 8.53325665e-01
4.75029260e-01 -3.06436151e-01 7.60893822e-01 6.76099136e-02
-5.59933245e-01 5.15649736e-01 6.49109364e-01 -3.68792266e-01
3.66267323e-01 -1.17272556e+00 5.64002275e-01 2.69478798e-01
6.21274233e-01 1.09147477e+00 7.35093117e-01 -8.83675873e-01
-1.57995319e+00 -1.00730014e+00 3.72561038e-01 -6.52370095e-01
9.82175469e-01 -4.40777034e-01 -1.15413129e+00 6.84698641e-01
2.02547863e-01 -9.55985785e-02 1.08032155e+00 7.36973822e-01
-6.48635551e-02 1.65883228e-01 -4.46841091e-01 8.57369125e-01
1.29152966e+00 -3.38272572e-01 -5.29635310e-01 2.03541070e-01
8.46923411e-01 1.78335801e-01 -8.88644636e-01 2.72974074e-01
8.40416729e-01 -6.37228549e-01 7.94159710e-01 -8.85211051e-01
7.31098726e-02 7.14147277e-03 2.66100883e-01 -1.79016399e+00
-1.30429888e+00 -9.14336920e-01 -1.07026947e+00 9.70505953e-01
5.06991148e-01 -7.19774485e-01 6.85684383e-01 6.15787148e-01
1.58818975e-01 -8.78885686e-01 -3.68317217e-01 -6.66937888e-01
3.17914970e-02 -1.26593187e-01 8.13277483e-01 9.84673679e-01
1.19394220e-01 6.44980311e-01 -8.28382194e-01 6.08435236e-02
8.74263346e-02 2.13122666e-01 3.91929746e-01 -1.78589833e+00
-8.38786364e-01 -7.01865792e-01 -1.92068979e-01 -1.30884492e+00
3.88351758e-03 -7.99569905e-01 -4.43706870e-01 -1.49618840e+00
2.75044054e-01 -3.43927443e-01 -9.76750791e-01 6.52205870e-02
-3.06493878e-01 -2.33502612e-01 5.58684617e-02 8.06111753e-01
-6.77309394e-01 9.53856707e-01 1.48617435e+00 6.57419786e-02
-8.39245021e-01 6.55448258e-01 -6.76599860e-01 3.83532286e-01
5.11706591e-01 -3.75984132e-01 -7.82795131e-01 -6.12800755e-02
7.63090134e-01 -1.28571969e-02 -2.25115195e-01 -7.14866579e-01
3.66373390e-01 -8.23481008e-02 5.05232334e-01 -2.99811304e-01
4.16504383e-01 -9.70012426e-01 6.24068022e-01 4.77925867e-01
-4.39072490e-01 2.34167740e-01 -1.21724747e-01 1.17080212e+00
-3.37573513e-02 1.38245185e-03 1.30392447e-01 2.84010656e-02
-1.69662148e-01 1.04569948e+00 -5.56631327e-01 -5.10750532e-01
9.02853310e-01 -2.83013284e-01 8.79537463e-02 -8.89468908e-01
-9.79427576e-01 2.64356852e-01 1.66543484e-01 9.35846746e-01
4.38628227e-01 -1.43994784e+00 -5.26017427e-01 3.36954772e-01
-1.71915591e-01 -6.82973087e-01 4.97906446e-01 6.65987611e-01
1.17824234e-01 3.85167956e-01 -2.07746044e-01 -4.22578454e-01
-8.80718112e-01 8.96298289e-01 3.79383653e-01 -7.39311397e-01
-5.51942706e-01 6.89867854e-01 2.96467990e-01 -3.75853658e-01
-2.90618371e-02 -1.75258011e-01 -9.03422058e-01 2.59990662e-01
2.94427812e-01 2.01954752e-01 -6.57781541e-01 -5.24187028e-01
2.80072361e-01 4.28754061e-01 -3.80480826e-01 4.93254252e-02
1.71366060e+00 -3.06027293e-01 1.49120063e-01 1.15008545e+00
9.94843364e-01 -1.36854306e-01 -1.37692845e+00 -5.82154334e-01
-2.89444476e-01 -7.90839791e-02 -1.59413777e-02 -6.50253057e-01
-1.04962993e+00 6.60745084e-01 6.13749862e-01 8.05386424e-01
8.52996469e-01 -1.20050527e-01 8.18597257e-01 3.83540019e-02
2.37133622e-01 -9.49120224e-01 3.42210233e-01 7.80562758e-01
6.50334179e-01 -5.45197964e-01 -2.35548839e-01 -1.63404807e-01
-5.04501164e-01 1.15171230e+00 5.06130338e-01 -4.19198692e-01
1.31207883e+00 -1.55441929e-02 -2.84993768e-01 -2.92906493e-01
-1.31593955e+00 1.29234463e-01 7.00734496e-01 3.65807384e-01
4.90296423e-01 1.75897792e-01 -3.32503617e-01 8.09432805e-01
-1.49175301e-01 1.76742822e-02 1.55599788e-01 5.02841771e-01
-8.89648870e-02 -1.21095335e+00 2.36250967e-01 6.46272182e-01
-2.93712050e-01 -8.47661048e-02 -6.63646758e-02 5.27543426e-01
3.93686481e-02 6.28626764e-01 3.55719328e-01 -5.66619039e-01
3.19210619e-01 6.51041344e-02 4.52487200e-01 -4.57828671e-01
-7.49483585e-01 2.13466465e-01 -5.02125382e-01 -9.90985632e-02
-2.40255296e-01 -6.81204319e-01 -8.00559223e-01 -2.32042938e-01
-6.50863886e-01 3.51174533e-01 4.00925219e-01 8.44714165e-01
6.36854768e-01 8.19847465e-01 9.86315489e-01 -7.19235063e-01
-5.80017269e-01 -1.12533915e+00 -8.34581852e-01 6.47226870e-01
7.98522960e-03 -7.79910922e-01 -3.91949147e-01 -2.25206897e-01] | [10.159870147705078, 5.604224681854248] |
4946403d-17ee-4679-8b2c-bce42483866d | pq-net-a-generative-part-seq2seq-network-for | 1911.10949 | null | https://arxiv.org/abs/1911.10949v3 | https://arxiv.org/pdf/1911.10949v3.pdf | PQ-NET: A Generative Part Seq2Seq Network for 3D Shapes | We introduce PQ-NET, a deep neural network which represents and generates 3D shapes via sequential part assembly. The input to our network is a 3D shape segmented into parts, where each part is first encoded into a feature representation using a part autoencoder. The core component of PQ-NET is a sequence-to-sequence or Seq2Seq autoencoder which encodes a sequence of part features into a latent vector of fixed size, and the decoder reconstructs the 3D shape, one part at a time, resulting in a sequential assembly. The latent space formed by the Seq2Seq encoder encodes both part structure and fine part geometry. The decoder can be adapted to perform several generative tasks including shape autoencoding, interpolation, novel shape generation, and single-view 3D reconstruction, where the generated shapes are all composed of meaningful parts. | ['Rundi Wu', 'Hao Zhang', 'Yixin Zhuang', 'Kai Xu', 'Baoquan Chen'] | 2019-11-25 | pq-net-a-generative-part-seq2seq-network-for-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Wu_PQ-NET_A_Generative_Part_Seq2Seq_Network_for_3D_Shapes_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Wu_PQ-NET_A_Generative_Part_Seq2Seq_Network_for_3D_Shapes_CVPR_2020_paper.pdf | cvpr-2020-6 | ['single-view-3d-reconstruction'] | ['computer-vision'] | [ 3.91497493e-01 5.10427415e-01 2.76009083e-01 -2.77776271e-01
-6.03891671e-01 -7.24865913e-01 4.37119514e-01 -5.17444611e-01
5.19768357e-01 5.53367078e-01 5.73559403e-01 6.79577366e-02
4.08576697e-01 -1.46894884e+00 -1.15998006e+00 -7.30558515e-01
1.59298062e-01 9.22701657e-01 -2.10301936e-01 -3.64047326e-02
-1.03773005e-01 9.56583440e-01 -1.46361554e+00 4.82163578e-01
3.84045362e-01 1.02777231e+00 2.79866070e-01 7.67531633e-01
-4.77511466e-01 2.98911810e-01 -5.48943877e-01 -4.79006648e-01
5.60609102e-01 -4.24550176e-01 -5.92239201e-01 5.88606238e-01
-1.45938843e-02 -7.95647383e-01 -3.26921582e-01 6.33317411e-01
5.84334195e-01 -4.28127795e-02 9.46409523e-01 -1.01055396e+00
-1.13268244e+00 6.97668374e-01 4.75987457e-02 -8.91893268e-01
3.47683340e-01 3.75840217e-01 8.26842189e-01 -9.87582564e-01
9.01502728e-01 1.43970919e+00 6.92207634e-01 5.15226126e-01
-1.60182416e+00 -3.30922604e-01 -3.66131783e-01 -6.58901811e-01
-1.12276089e+00 -3.14516991e-01 9.06290293e-01 -5.39120913e-01
1.21658361e+00 -1.07364424e-01 1.31227696e+00 1.09026551e+00
3.85423779e-01 1.02396894e+00 3.14314634e-01 9.53272879e-02
3.36281985e-01 -6.37941241e-01 -8.21375906e-01 8.26235414e-01
-1.00162975e-01 2.01079056e-01 9.86163914e-02 -2.55424380e-01
1.88128376e+00 3.64737809e-01 8.50774869e-02 -3.11653078e-01
-1.24851561e+00 8.79797578e-01 6.52217507e-01 -1.02424875e-01
-8.10752094e-01 3.10876489e-01 1.23867027e-01 9.80884880e-02
4.61079270e-01 4.08741742e-01 -6.30857706e-01 -1.72007252e-02
-6.45635724e-01 6.67604446e-01 8.04938912e-01 1.32220972e+00
8.81984472e-01 6.00921035e-01 -4.16607827e-01 8.13122392e-01
4.16294008e-01 8.07816625e-01 3.26880813e-01 -1.28942072e+00
1.40512094e-01 8.27209532e-01 -1.41164184e-01 -3.50597411e-01
1.45104304e-01 -1.48576543e-01 -1.18572342e+00 3.08358163e-01
-2.30703846e-01 -2.66815633e-01 -1.24496531e+00 1.62587368e+00
3.13592762e-01 2.20800135e-02 -8.11990816e-03 6.82205796e-01
1.02584398e+00 1.03938878e+00 -2.41650939e-01 3.14061195e-01
1.12941015e+00 -7.91285694e-01 -3.57671261e-01 1.49891138e-01
-2.24274788e-02 -8.33451033e-01 3.62394482e-01 -1.04501702e-01
-1.69724095e+00 -8.76485050e-01 -7.90280521e-01 -3.92820388e-01
-1.50260508e-01 6.10743165e-02 5.88425636e-01 -3.30396593e-02
-1.06326723e+00 8.90177548e-01 -5.97990274e-01 3.40147227e-01
7.18160093e-01 3.25364143e-01 -6.24571502e-01 4.40045781e-02
-7.70207584e-01 2.05238566e-01 2.89999485e-01 -5.35932258e-02
-1.12685859e+00 -6.97913408e-01 -1.41129041e+00 2.89616883e-01
-3.50749135e-01 -1.63290596e+00 1.33611083e+00 -8.78547609e-01
-1.90776312e+00 8.29271555e-01 -1.81077093e-01 -1.56716764e-01
8.73769075e-02 3.25858504e-01 -5.53435050e-02 -1.29691765e-01
2.96529353e-01 1.32799506e+00 1.39507794e+00 -1.39899671e+00
-1.29667800e-02 -1.98767066e-01 -2.48540923e-01 6.76087812e-02
7.21279085e-01 -4.23002362e-01 -4.86909389e-01 -1.15791976e+00
3.38299364e-01 -8.22247684e-01 -4.25805151e-01 1.92185506e-01
-6.54410541e-01 2.77607553e-02 5.51035166e-01 -6.55611932e-01
4.06941175e-01 -2.03844690e+00 5.37617207e-01 -3.65994289e-03
1.74799055e-01 -6.84981644e-02 -5.99269092e-01 7.29083538e-01
-3.53503019e-01 5.57000265e-02 -6.81690991e-01 -4.98152673e-01
7.22607225e-02 4.38365281e-01 -4.74717796e-01 -1.13072731e-01
5.59623480e-01 1.57661879e+00 -6.94140494e-01 8.40253290e-03
6.99962676e-02 8.35952222e-01 -8.23622882e-01 7.85585344e-01
-7.03430772e-01 4.64539915e-01 -4.81481075e-01 6.46129370e-01
8.66894126e-01 -3.19498271e-01 -2.48855412e-01 -1.94056332e-01
-6.82360381e-02 3.02657455e-01 -8.86565089e-01 1.90371478e+00
-3.63187194e-01 2.10392058e-01 -4.02761064e-02 -5.97101569e-01
1.15536976e+00 4.86982733e-01 5.65169394e-01 -3.28680933e-01
2.69491915e-02 1.26218975e-01 -3.89768720e-01 -2.38039151e-01
3.57688129e-01 -4.21789885e-01 -1.44162536e-01 7.66175449e-01
2.80103207e-01 -7.18006670e-01 -2.71359652e-01 -1.55570522e-01
8.66312981e-01 5.61746716e-01 1.86346814e-01 3.62806529e-01
3.91038507e-02 -1.98679194e-01 3.27078342e-01 2.96109438e-01
6.17933750e-01 1.03313076e+00 4.37320888e-01 -7.74394393e-01
-1.88374507e+00 -1.63951063e+00 5.76615781e-02 4.02458072e-01
-4.30175602e-01 -8.60510841e-02 -6.61272049e-01 -4.91029710e-01
4.68562245e-01 6.73239470e-01 -7.70961165e-01 -2.22075462e-01
-7.20646799e-01 5.21364855e-03 3.91274691e-01 7.40644157e-01
2.22096860e-01 -1.78411865e+00 -3.70104969e-01 5.64956903e-01
1.15686446e-01 -7.05335677e-01 -8.02494109e-01 2.07130671e-01
-1.34292531e+00 -7.17960417e-01 -1.07324290e+00 -1.04829109e+00
1.09735227e+00 -3.65752243e-02 1.32113349e+00 -5.41538522e-02
-2.78798908e-01 -9.02637020e-02 -1.08482607e-01 -1.91198215e-01
-9.47860897e-01 -1.88546374e-01 -2.16195747e-01 1.18861884e-01
-3.37668896e-01 -1.18448687e+00 -6.19334221e-01 -1.31862953e-01
-1.07308161e+00 6.00874186e-01 9.97270644e-01 8.35978389e-01
1.23493040e+00 -1.58634081e-01 3.16613942e-01 -6.63680792e-01
3.92500132e-01 -4.07684237e-01 -4.37881380e-01 -2.17032105e-01
1.46208242e-01 4.26131219e-01 8.78531337e-01 -2.11696491e-01
-1.04294956e+00 3.81674170e-01 -8.15367401e-01 -6.44338191e-01
-6.39765739e-01 -8.53379082e-04 -5.48784733e-01 4.36606377e-01
1.54669553e-01 9.07582641e-01 4.17667300e-01 -7.18522429e-01
7.23072350e-01 4.35492486e-01 4.59870905e-01 -4.14846212e-01
8.33333671e-01 3.42622727e-01 -2.47358996e-02 -6.60024405e-01
-5.68533897e-01 2.60257393e-01 -1.13201094e+00 3.99394259e-02
1.16544998e+00 -1.00101435e+00 -6.59401536e-01 7.32746184e-01
-1.51892638e+00 -6.35937333e-01 -1.07194769e+00 -2.41099998e-01
-1.07049024e+00 -2.09379308e-02 -8.19121778e-01 -4.65732187e-01
-6.92990124e-01 -1.15289152e+00 1.77270210e+00 1.03532858e-01
-1.44245312e-01 -5.66997588e-01 1.15313731e-01 4.00747880e-02
1.74730301e-01 3.88214380e-01 1.30600297e+00 -1.47620991e-01
-9.98205960e-01 -2.18214259e-01 -5.04419431e-02 3.18195462e-01
2.91131943e-01 -6.29839376e-02 -6.64735079e-01 -1.71598792e-01
-2.64269620e-01 -3.18160146e-01 6.33878291e-01 6.93389356e-01
1.48129427e+00 -7.07988083e-01 -1.74867526e-01 9.04096484e-01
1.30828667e+00 4.13175285e-01 6.92966044e-01 -5.40582120e-01
7.69330919e-01 1.41588449e-01 -3.91099453e-01 6.09147906e-01
4.24977332e-01 3.83526921e-01 5.11677325e-01 -1.40415709e-02
-4.38025892e-01 -8.80363882e-01 2.73944050e-01 1.11762452e+00
-1.33849293e-01 -1.32367581e-01 -4.50385422e-01 4.22971696e-01
-1.39442718e+00 -1.06890094e+00 1.66585058e-01 1.92978787e+00
7.32211828e-01 -1.16198495e-01 5.98462969e-02 -1.49950068e-02
3.66314054e-01 5.18831052e-02 -7.26188242e-01 -3.88817459e-01
-8.39427933e-02 4.35391814e-01 3.41402367e-02 3.85069251e-01
-6.03539169e-01 9.90681291e-01 6.97447395e+00 5.53849995e-01
-1.06368172e+00 -3.38880926e-01 4.99263495e-01 1.50427356e-01
-1.02322435e+00 -1.03865325e-01 -4.07378882e-01 2.97784209e-01
5.13626158e-01 3.11869860e-01 6.25374973e-01 8.28162074e-01
-1.99054942e-01 3.57210040e-01 -9.66950893e-01 8.74198377e-01
-4.12614346e-02 -1.73343849e+00 6.06801927e-01 2.17511192e-01
9.99685407e-01 -1.30545348e-02 -9.58660692e-02 2.44972602e-01
6.69814467e-01 -1.22746968e+00 9.76122200e-01 8.61787140e-01
1.19802248e+00 -8.31395149e-01 3.20933610e-01 5.92657208e-01
-1.25798106e+00 1.65539458e-01 -5.67496300e-01 6.79921359e-02
6.00298941e-01 6.28745079e-01 -8.50446582e-01 4.18124944e-01
2.49017000e-01 8.28944743e-01 1.18534356e-01 8.28499913e-01
-2.94510514e-01 2.95576125e-01 -1.66864067e-01 1.33467272e-01
3.09563689e-02 -4.51774567e-01 6.19654179e-01 8.02829802e-01
6.35085464e-01 3.37900639e-01 1.64901778e-01 1.61697102e+00
-4.04381990e-01 -3.91280770e-01 -1.02180612e+00 -3.25142920e-01
4.68957275e-01 9.74849343e-01 -5.29915750e-01 -5.86680055e-01
-7.69704953e-02 1.54983914e+00 2.98184872e-01 3.79726231e-01
-3.96907061e-01 -4.25689965e-01 8.58961403e-01 -5.36457263e-03
9.75604236e-01 -3.22986335e-01 -4.49854195e-01 -9.49053466e-01
-3.07862073e-01 -5.79106867e-01 -2.49570265e-01 -1.30258691e+00
-1.35509634e+00 7.97575831e-01 -3.39296460e-01 -1.24321699e+00
-4.28589076e-01 -4.72999960e-01 -6.73720121e-01 1.11851072e+00
-7.29023397e-01 -1.48948717e+00 -7.72489980e-02 2.42079496e-01
7.41898537e-01 -4.50294763e-01 1.25545108e+00 -7.09812120e-02
-1.85983285e-01 2.07471758e-01 -1.50105894e-01 4.47212398e-01
-6.34289086e-02 -1.17890668e+00 1.49926901e+00 2.23240569e-01
2.42747948e-01 4.48642075e-01 1.03999704e-01 -1.10318971e+00
-1.48696053e+00 -1.41409898e+00 9.43023920e-01 -4.66500163e-01
-1.21723145e-01 -5.61158419e-01 -6.89352095e-01 8.35153997e-01
5.77460341e-02 5.74546903e-02 6.20708108e-01 -5.06906331e-01
-3.61683279e-01 4.37829047e-01 -1.14829826e+00 5.96754432e-01
1.36023295e+00 -5.93679011e-01 -7.18078256e-01 -8.96803364e-02
1.12487435e+00 -8.01977038e-01 -1.16770720e+00 1.97319433e-01
7.82526672e-01 -8.66960287e-01 1.25112641e+00 -7.27736115e-01
1.05026758e+00 -1.65496290e-01 -2.53346056e-01 -1.47809398e+00
-9.23370361e-01 -3.28923970e-01 -3.81478816e-01 7.75220096e-01
4.09959912e-01 -1.54996172e-01 9.25722241e-01 1.85037881e-01
-5.63009560e-01 -1.09710002e+00 -6.65391982e-01 -4.73940015e-01
2.36601662e-02 -3.94577652e-01 1.56061018e+00 4.41131324e-01
-8.39681566e-01 2.79053748e-01 -1.73794508e-01 -2.58648396e-01
5.06451428e-01 9.88313794e-01 8.07922006e-01 -1.10997546e+00
-5.23509562e-01 -5.01241505e-01 -1.21663250e-01 -1.73669446e+00
1.21162213e-01 -1.31932068e+00 2.35556841e-01 -1.88944638e+00
6.80496842e-02 -2.70640403e-01 4.60469574e-01 6.03601754e-01
1.37362808e-01 1.60297930e-01 1.47122011e-01 2.88625490e-02
8.58087614e-02 1.25460315e+00 2.09403539e+00 -2.64807701e-01
-1.69794023e-01 2.40097731e-01 -4.94560510e-01 4.78836358e-01
3.77532959e-01 -3.65731835e-01 -1.27304897e-01 -6.83201015e-01
1.15521565e-01 6.85373545e-01 3.97358119e-01 -6.03156984e-01
-2.87236929e-01 -7.40013644e-02 1.09617972e+00 -9.77147996e-01
5.45382321e-01 -7.64602304e-01 7.63574839e-01 3.90182227e-01
-1.41131267e-01 1.90876320e-01 7.90302902e-02 3.61976147e-01
-1.07110634e-01 7.09639788e-02 4.54037875e-01 -6.06766939e-01
-1.71395868e-01 1.06356430e+00 -3.90256435e-01 -2.38857448e-01
5.35879195e-01 -5.77788591e-01 4.89595622e-01 -3.64210993e-01
-8.67498815e-01 -3.03095579e-01 5.76989889e-01 5.49261570e-01
9.80333269e-01 -2.18597865e+00 -8.51950347e-01 9.63079214e-01
-2.61440098e-01 7.48720407e-01 5.05732656e-01 -1.22899130e-01
-5.71386755e-01 2.99470007e-01 -5.89803457e-01 -4.55839723e-01
-5.19250631e-01 4.52536374e-01 3.47101420e-01 1.14104114e-01
-9.97066855e-01 1.08751392e+00 2.37311035e-01 -9.65196371e-01
-2.90874273e-01 -3.66786063e-01 2.32111782e-01 -2.74726361e-01
2.31487378e-01 -3.67332958e-02 -2.67525464e-01 -7.81185865e-01
1.82130769e-01 7.19836295e-01 5.55086613e-01 -1.29183725e-01
1.79274178e+00 4.46135312e-01 -3.99229437e-01 2.81267554e-01
1.53365409e+00 -4.60831858e-02 -1.75645769e+00 9.73878801e-02
-7.76429117e-01 -1.44394115e-01 -3.16659838e-01 -6.48037791e-01
-1.27565157e+00 6.66518092e-01 -1.06972508e-01 6.57778000e-03
9.28062439e-01 2.17079401e-01 1.46254945e+00 8.62089619e-02
3.87714356e-01 -5.68145394e-01 1.06883958e-01 9.16667998e-01
1.50352156e+00 -7.94567049e-01 -4.30631757e-01 -1.87997952e-01
-4.71393853e-01 1.21572042e+00 3.06080699e-01 -5.15651643e-01
9.09377694e-01 5.49385488e-01 -1.47560969e-01 -2.26192474e-01
-7.34166443e-01 -3.65640745e-02 4.93269891e-01 7.75300920e-01
2.62757897e-01 3.73572767e-01 6.31818593e-01 8.53163779e-01
-5.75858235e-01 3.25322263e-02 -3.57806124e-02 4.94872510e-01
-1.59824222e-01 -1.38246036e+00 -1.16514757e-01 5.28764486e-01
1.37706503e-01 6.83356375e-02 -3.67805839e-01 2.11028948e-01
5.95429540e-01 -1.54370457e-01 5.82327604e-01 -3.62788171e-01
4.35676277e-01 2.45743170e-01 6.38207018e-01 -8.49272609e-01
-5.66396058e-01 2.66724348e-01 -2.76844472e-01 -6.06740892e-01
1.81361660e-01 -5.98763049e-01 -1.40584719e+00 -9.89799052e-02
2.55307317e-01 -2.15505868e-01 5.78006744e-01 5.99951625e-01
7.63483465e-01 7.59129703e-01 8.12486529e-01 -1.48011768e+00
-2.83057272e-01 -7.91554570e-01 -5.89529514e-01 5.70358813e-01
4.02348727e-01 -2.28613526e-01 3.07124168e-01 3.45963776e-01] | [8.992735862731934, -3.6404168605804443] |
acee7c0d-10a8-4151-b47a-06d735eddec2 | neurohsmd-neuromorphic-hybrid-spiking-motion | 2112.06102 | null | https://arxiv.org/abs/2112.06102v5 | https://arxiv.org/pdf/2112.06102v5.pdf | NeuroHSMD: Neuromorphic Hybrid Spiking Motion Detector | Vertebrate retinas are highly-efficient in processing trivial visual tasks such as detecting moving objects, yet a complex challenges for modern computers. In vertebrates, the detection of object motion is performed by specialised retinal cells named Object Motion Sensitive Ganglion Cells (OMS-GC). OMS-GC process continuous visual signals and generate spike patterns that are post-processed by the Visual Cortex. Our previous Hybrid Sensitive Motion Detector (HSMD) algorithm was the first hybrid algorithm to enhance Background subtraction (BS) algorithms with a customised 3-layer Spiking Neural Network (SNN) that generates OMS-GC spiking-like responses. In this work, we present a Neuromorphic Hybrid Sensitive Motion Detector (NeuroHSMD) algorithm that accelerates our HSMD algorithm using Field-Programmable Gate Arrays (FPGAs). The NeuroHSMD was compared against the HSMD algorithm, using the same 2012 Change Detection (CDnet2012) and 2014 Change Detection (CDnet2014) benchmark datasets. When tested against the CDnet2012 and CDnet2014 datasets, NeuroHSMD performs object motion detection at 720x480 at 28.06 Frames Per Second (fps) and 720x480 at 28.71 fps, respectively, with no degradation of quality. Moreover, the NeuroHSMD proposed in this paper was completely implemented in Open Computer Language (OpenCL) and therefore is easily replicated in other devices such as Graphical Processing Units (GPUs) and clusters of Central Processing Units (CPUs). | ['T. M. McGinnity', 'Andreas Oikonomou', 'Joao Filipe Ferreira', 'Pedro Machado'] | 2021-12-12 | null | null | null | null | ['motion-detection'] | ['computer-vision'] | [ 4.52138335e-01 -7.20025778e-01 7.97703266e-01 2.05760926e-01
2.25127578e-01 -5.10761499e-01 4.90162134e-01 -7.31963515e-02
-1.29966557e+00 6.21272862e-01 -4.83977586e-01 -2.47437954e-01
5.09953678e-01 -6.27628684e-01 -8.04948509e-01 -8.48711491e-01
1.56434271e-02 -4.18053061e-01 1.38254631e+00 4.31894623e-02
4.86937732e-01 6.51462495e-01 -2.17831230e+00 5.10143161e-01
2.31823951e-01 1.08800733e+00 6.38952076e-01 1.30965579e+00
1.73265681e-01 7.20987141e-01 -5.91779232e-01 2.11134598e-01
5.43654084e-01 -3.76525164e-01 -9.46062505e-02 -4.09001112e-01
4.15121645e-01 2.97496747e-02 -2.53242731e-01 1.18687642e+00
8.03564489e-01 -4.23947796e-02 2.56934971e-01 -1.03462994e+00
-5.34324832e-02 -8.77594203e-03 -5.70155740e-01 1.17367923e+00
-4.34825756e-02 7.31168151e-01 -1.46275848e-01 -8.03032637e-01
7.42822528e-01 1.30494213e+00 6.66622043e-01 7.06423223e-01
-1.24104536e+00 -6.89798534e-01 -3.39155138e-01 2.75082320e-01
-1.19378817e+00 -4.08685923e-01 1.74752712e-01 -5.62533677e-01
1.45217919e+00 3.22522789e-01 9.10449922e-01 7.88989365e-01
7.64130950e-01 2.93741405e-01 1.14450252e+00 -8.59809220e-02
9.25896883e-01 -4.57113743e-01 2.72809565e-01 2.07428932e-01
9.06269014e-01 6.90159723e-02 -7.41827309e-01 1.18683554e-01
1.30488682e+00 -2.17145443e-01 -2.85999984e-01 3.54089528e-01
-1.42636836e+00 2.71100998e-01 3.68973911e-01 1.53946817e-01
-4.47270781e-01 6.01708472e-01 4.68820244e-01 1.72874123e-01
-3.95064801e-01 -1.16535746e-01 -3.24699253e-01 -1.52563244e-01
-8.70742619e-01 1.93438664e-01 6.14847779e-01 6.97484314e-01
6.13784492e-01 3.77938211e-01 -1.57421216e-01 2.66969353e-01
3.69019896e-01 7.30287313e-01 8.29402804e-01 -9.12195683e-01
-4.36166413e-02 6.39045000e-01 -4.58467752e-02 -8.03511441e-01
-6.48915291e-01 -1.41310826e-01 -1.01761019e+00 8.85976136e-01
4.12155658e-01 -1.25802606e-01 -1.26556909e+00 1.35391736e+00
4.18014862e-02 4.28373158e-01 2.14815661e-01 1.07958126e+00
1.02444279e+00 8.69229198e-01 2.72012532e-01 -1.48375094e-01
1.50607145e+00 -5.06778955e-01 -3.14042091e-01 -4.95979398e-01
2.47040242e-01 -6.54577553e-01 6.41263425e-01 2.44104490e-01
-1.01048970e+00 -7.82551527e-01 -1.25368834e+00 4.36411388e-02
-3.41327757e-01 1.06256016e-01 3.95397484e-01 6.38536870e-01
-1.52166414e+00 4.97889936e-01 -1.30559802e+00 -5.03878415e-01
4.74123329e-01 8.28947663e-01 -9.86904204e-02 5.39565146e-01
-5.52971303e-01 5.91954648e-01 4.60134715e-01 -7.14537278e-02
-8.78437579e-01 -4.56863016e-01 -3.40164542e-01 -2.92472541e-01
-3.46885711e-01 -7.41690397e-01 1.19546795e+00 -1.24099469e+00
-1.58639705e+00 1.21190381e+00 -1.18884049e-01 -9.78331506e-01
2.94367552e-01 2.83356905e-01 -4.81744260e-01 2.74291605e-01
-2.28973731e-01 1.07261097e+00 7.24510193e-01 -6.49522662e-01
-8.34411740e-01 -3.76474530e-01 -4.66407090e-01 -1.97180241e-01
2.89051443e-01 3.91686857e-01 -2.72505015e-01 -6.48878634e-01
2.03261122e-01 -9.53374505e-01 -3.27257216e-01 4.20378745e-01
1.74733862e-01 1.87891409e-01 9.47420418e-01 -2.29459420e-01
7.98003674e-01 -2.33460999e+00 -3.01813453e-01 -1.87916726e-01
-1.13892063e-01 1.14114571e+00 -1.50483191e-01 -1.87072888e-01
1.45972386e-01 -6.36396229e-01 -5.58888912e-01 1.61475152e-01
-5.95813036e-01 1.86180070e-01 -1.07314765e-01 6.42417490e-01
2.68771589e-01 7.32094109e-01 -7.00408459e-01 -1.00786075e-01
3.08231652e-01 5.22552967e-01 -4.69671130e-01 -3.63977790e-01
1.79641947e-01 5.03541052e-01 1.40955940e-01 7.84302175e-01
1.01336753e+00 9.12309450e-04 -2.49093890e-01 -1.12736665e-01
-8.79698932e-01 -1.19321294e-01 -1.33677948e+00 1.44472647e+00
1.09055333e-01 1.23541331e+00 1.01228274e-01 -4.96726394e-01
1.15549600e+00 -7.94371217e-02 1.56126255e-02 -8.62745225e-01
4.17081028e-01 4.43697780e-01 3.41218472e-01 -2.43456230e-01
2.60384470e-01 1.26906723e-01 4.10841137e-01 -1.27010807e-01
7.44723827e-02 1.66578695e-01 2.43074566e-01 -2.44783774e-01
1.49636221e+00 5.13343699e-02 3.93952787e-01 -5.39855063e-01
5.97331643e-01 9.48459506e-02 7.64437079e-01 6.24566734e-01
-7.34423339e-01 5.36729991e-01 5.34224771e-02 -5.32651424e-01
-8.65392983e-01 -1.28966105e+00 -2.74279892e-01 6.16811812e-01
3.71506393e-01 1.14461549e-01 -7.51533628e-01 4.07898545e-01
-1.61594465e-01 2.77616411e-01 -1.44198164e-01 -2.92716012e-03
-6.21094406e-01 -1.14312685e+00 7.55859196e-01 5.81576943e-01
9.37931001e-01 -1.50961983e+00 -2.05696368e+00 6.42063081e-01
5.52525818e-01 -1.15497899e+00 2.27746099e-01 4.13107425e-01
-9.34336007e-01 -8.67756426e-01 -7.08882391e-01 -1.01842034e+00
4.19618279e-01 5.42473614e-01 6.15339339e-01 -5.06168902e-01
-9.00924265e-01 4.55555283e-02 -3.90736721e-02 -7.54392803e-01
-4.26420905e-02 -7.74391055e-01 1.86569020e-01 -1.31524965e-01
7.66486704e-01 -7.17847824e-01 -9.75720584e-01 1.03201993e-01
-1.10072827e+00 1.73036233e-01 5.66015303e-01 2.65717208e-01
6.77729547e-01 -3.96690577e-01 1.65722683e-01 -2.84310937e-01
-2.15792879e-02 -1.35577455e-01 -1.23695147e+00 -3.26848507e-01
-1.45273088e-02 -1.86459094e-01 5.41884601e-01 -6.06956005e-01
-7.19902396e-01 5.87606549e-01 1.33416094e-02 -1.96046054e-01
-2.29785055e-01 -2.30645165e-01 3.23649913e-01 -6.85421586e-01
1.04652703e+00 5.16555011e-01 -2.88140446e-01 -1.47518978e-01
-4.74906862e-01 5.79096437e-01 1.28678274e+00 3.55410486e-01
3.43680441e-01 1.06467199e+00 3.06757450e-01 -1.08710921e+00
3.91041398e-01 -6.07716918e-01 -3.08274806e-01 -3.52002859e-01
9.41019416e-01 -1.15504754e+00 -9.54780102e-01 1.37118554e+00
-1.25077248e+00 -4.30570632e-01 9.14139003e-02 5.94084978e-01
-5.11004388e-01 1.60291985e-01 -6.42936289e-01 -9.09157991e-01
-5.93935907e-01 -8.87652099e-01 6.84940338e-01 9.36335981e-01
-4.24451903e-02 -3.08665723e-01 1.89556405e-01 -3.90944183e-01
6.62022173e-01 4.83897924e-01 2.90944368e-01 -4.42250520e-02
-8.18625212e-01 4.20009121e-02 -4.78692889e-01 1.54070079e-01
-1.40556410e-01 4.16600227e-01 -1.19464636e+00 -2.19738945e-01
1.13880180e-01 2.97357112e-01 1.21146309e+00 9.34207797e-01
7.04232693e-01 2.07820252e-01 -2.03670472e-01 8.69204640e-01
2.01817584e+00 7.49992490e-01 1.16360474e+00 7.87964821e-01
2.54691273e-01 1.85939744e-01 1.64590374e-01 6.75067604e-01
-4.67239283e-02 4.49222624e-01 6.71897709e-01 7.97340870e-02
-3.72086346e-01 4.56038594e-01 6.40824258e-01 4.74301189e-01
-3.44781637e-01 -1.37099642e-02 -9.78422105e-01 6.16297364e-01
-1.69201541e+00 -8.58868659e-01 -9.18397784e-01 2.40845275e+00
6.42651618e-01 2.38909394e-01 2.55635351e-01 2.48084322e-01
1.01179540e+00 -4.18362111e-01 -7.70865381e-01 -5.12416482e-01
-7.77964950e-01 7.48341084e-01 1.01832163e+00 -4.34349887e-02
-9.77356076e-01 8.11073720e-01 4.69663620e+00 2.76210308e-01
-1.60179842e+00 7.36474153e-03 7.49553069e-02 -1.41017213e-01
6.26550257e-01 2.95059588e-02 -1.02353227e+00 8.62747192e-01
1.23217082e+00 -2.30024338e-01 4.45838124e-01 5.71253836e-01
3.43595237e-01 -7.30925560e-01 -5.69876134e-01 1.40672982e+00
-4.04891014e-01 -1.43176234e+00 -3.01646590e-02 -2.40315825e-01
6.06905937e-01 4.27190602e-01 -1.18722841e-02 -2.95399558e-02
-6.92608953e-02 -6.84895992e-01 7.91393697e-01 5.27892113e-01
4.87161905e-01 -5.88963747e-01 7.05309212e-01 1.76643640e-01
-1.11082315e+00 -2.17174187e-01 -1.02741563e+00 -4.55202758e-01
-1.12548046e-01 4.84928787e-01 -4.04538631e-01 -1.45997852e-01
1.17501318e+00 4.97368455e-01 -7.66421139e-01 1.65948999e+00
3.17827940e-01 4.20618117e-01 -4.82867837e-01 -3.33713084e-01
3.62792373e-01 1.70933574e-01 9.00288880e-01 1.58908141e+00
5.10615230e-01 2.53457010e-01 -7.32857466e-01 8.19351554e-01
3.63412738e-01 -2.31778905e-01 -1.50103778e-01 1.03537507e-01
2.56544113e-01 1.29189050e+00 -1.36068010e+00 -4.01732504e-01
-3.17156643e-01 1.03097451e+00 -3.13339561e-01 3.31251957e-02
-6.75816417e-01 -7.69094169e-01 9.36760843e-01 3.81280519e-02
5.21724403e-01 -3.03444743e-01 -3.51575136e-01 -8.80256772e-01
-2.19416872e-01 -4.44958180e-01 1.11782208e-01 -8.27756524e-01
-7.77653694e-01 6.53009832e-01 -5.23082316e-01 -1.32700396e+00
1.86750814e-01 -1.17052341e+00 -6.63521826e-01 6.51402831e-01
-1.28876388e+00 -3.36024433e-01 -7.19836056e-01 8.14593256e-01
3.75515491e-01 -1.28932837e-02 4.95906591e-01 2.56377101e-01
-5.79611719e-01 7.25937635e-02 2.00158998e-01 1.34440726e-02
4.61329460e-01 -8.42285335e-01 9.18239534e-01 1.41526270e+00
-2.14736238e-01 3.66552770e-01 6.75362289e-01 -5.69718778e-01
-1.46419847e+00 -1.51908576e+00 4.28963900e-01 1.53722271e-01
3.69887501e-01 -2.42023721e-01 -1.04235137e+00 1.94041401e-01
-1.34952022e-02 3.04938734e-01 2.56887108e-01 -1.45703900e+00
2.13757474e-02 -6.74640834e-02 -1.28816998e+00 7.11985528e-01
1.08568370e+00 -9.33194906e-02 -3.37124228e-01 -1.72511593e-01
2.28318006e-01 -4.40867484e-01 -9.64855105e-02 3.60098541e-01
3.48759651e-01 -1.31668878e+00 7.48571873e-01 8.09131712e-02
-3.24880481e-02 -1.13536859e+00 -3.32922190e-01 -5.62810242e-01
-4.09676075e-01 -6.72576487e-01 4.52303588e-02 7.28253007e-01
-2.21689373e-01 -8.57751608e-01 5.28666735e-01 9.63970497e-02
-2.67458230e-01 -8.04389641e-02 -1.33749771e+00 -1.04921174e+00
-4.12685752e-01 -3.53404015e-01 3.16327624e-02 3.46553594e-01
-5.07369876e-01 -6.56718165e-02 2.44676068e-01 4.32122260e-01
7.08801985e-01 -1.80044249e-01 4.89754081e-01 -1.15173960e+00
-1.56396076e-01 -4.74951267e-01 -1.60650897e+00 -5.07143021e-01
-5.80766439e-01 -5.69504917e-01 1.05090998e-01 -1.27597737e+00
7.55200069e-03 4.14807826e-01 -4.76581514e-01 3.47397625e-01
7.94825405e-02 7.16109633e-01 1.36355862e-01 1.63811117e-01
-3.33003551e-01 -1.23930767e-01 6.51121318e-01 2.53102660e-01
-4.36754853e-01 4.36686836e-02 -9.56899393e-03 8.26095581e-01
6.94200158e-01 -4.59882110e-01 6.69398308e-02 -3.70356977e-01
-6.05078973e-02 -4.22324836e-01 1.00933933e+00 -2.07208419e+00
8.65221381e-01 1.85912505e-01 6.87583208e-01 -7.76566803e-01
8.48269463e-02 -3.77555460e-01 5.78185260e-01 1.34436262e+00
2.12553278e-01 1.81804314e-01 9.11584437e-01 5.14553428e-01
-2.85404567e-02 -6.48243818e-03 1.42503619e+00 -2.02843338e-01
-1.49841130e+00 -1.54037952e-01 -1.21362984e+00 -1.52948111e-01
1.25307930e+00 -8.70534718e-01 -5.57151079e-01 3.58777493e-01
-3.40587229e-01 -3.66797864e-01 6.13875926e-01 6.26792163e-02
7.86747038e-01 -9.41882372e-01 -5.40941298e-01 3.06105465e-01
-7.71599934e-02 -3.14197801e-02 2.93744713e-01 8.75714779e-01
-1.33869851e+00 5.83731294e-01 -1.15134668e+00 -9.55474854e-01
-1.38375545e+00 4.13728654e-01 2.00531036e-01 6.84033096e-01
-6.48389339e-01 1.21438634e+00 2.21787989e-01 4.21822637e-01
4.26936559e-02 -7.70815790e-01 -1.35615170e-01 -2.44514570e-01
1.00681543e+00 7.39619136e-01 1.79702401e-01 -5.08207798e-01
-8.49022388e-01 6.95785642e-01 4.59242880e-01 -1.88138738e-01
1.29748929e+00 5.79447076e-02 -2.43045464e-01 4.13636088e-01
7.69195437e-01 -4.74249691e-01 -1.52266669e+00 1.30432457e-01
1.47340029e-01 -2.62218148e-01 -5.24873361e-02 -4.86084342e-01
-8.19847286e-01 9.29643929e-01 1.26813698e+00 -3.03791523e-01
1.69475472e+00 -4.98172104e-01 5.84238052e-01 3.35188329e-01
5.14643967e-01 -9.37024772e-01 -2.20758021e-01 4.85521048e-01
4.87256080e-01 -6.88827693e-01 -3.57839257e-01 -2.52718508e-01
-5.53992212e-01 1.36137784e+00 7.65670478e-01 -5.53519189e-01
2.58764863e-01 6.92555606e-01 7.38562644e-02 1.74963966e-01
-9.36007321e-01 -5.50127327e-01 -1.61301255e-01 7.08572388e-01
3.63580063e-02 -3.00999340e-02 -2.75570244e-01 3.19636345e-01
9.23191607e-02 4.72135872e-01 8.16254377e-01 1.39196742e+00
-8.66779268e-01 -3.60775739e-01 -4.30496067e-01 1.38660207e-01
-4.16174054e-01 -1.56787544e-01 -3.30469549e-01 3.11034024e-01
3.82104784e-01 7.60107756e-01 5.28050005e-01 -3.00552011e-01
1.64182112e-01 -3.09070468e-01 5.77716470e-01 -4.66946244e-01
-8.25142503e-01 -1.76525488e-02 -5.50826252e-01 -6.13325715e-01
-5.82967818e-01 -5.37118912e-01 -1.79017603e+00 -3.02801937e-01
2.78333366e-01 -6.53994918e-01 1.07972205e+00 3.87904763e-01
5.86954474e-01 5.73662877e-01 1.38853192e-01 -9.50764656e-01
1.09109975e-01 -5.97302020e-01 -6.06370926e-01 -8.23174939e-02
3.04856092e-01 -3.64264548e-01 -3.68322730e-01 4.84820336e-01] | [8.650334358215332, -1.142829179763794] |
665ce979-c632-4d01-bb44-5e5054d8225e | degraphcs-embedding-variable-based-flow-graph | 2103.13020 | null | https://arxiv.org/abs/2103.13020v3 | https://arxiv.org/pdf/2103.13020v3.pdf | deGraphCS: Embedding Variable-based Flow Graph for Neural Code Search | With the rapid increase in the amount of public code repositories, developers maintain a great desire to retrieve precise code snippets by using natural language. Despite existing deep learning based approaches(e.g., DeepCS and MMAN) have provided the end-to-end solutions (i.e., accepts natural language as queries and shows related code fragments retrieved directly from code corpus), the accuracy of code search in the large-scale repositories is still limited by the code representation (e.g., AST) and modeling (e.g., directly fusing the features in the attention stage). In this paper, we propose a novel learnable deep Graph for Code Search (calleddeGraphCS), to transfer source code into variable-based flow graphs based on the intermediate representation technique, which can model code semantics more precisely compared to process the code as text directly or use the syntactic tree representation. Furthermore, we propose a well-designed graph optimization mechanism to refine the code representation, and apply an improved gated graph neural network to model variable-based flow graphs. To evaluate the effectiveness of deGraphCS, we collect a large-scale dataset from GitHub containing 41,152 code snippets written in C language, and reproduce several typical deep code search methods for comparison. Besides, we design a qualitative user study to verify the practical value of our approach. The experimental results have shown that deGraphCS can achieve state-of-the-art performances, and accurately retrieve code snippets satisfying the needs of the users. | ['Mingyang Geng', 'Xiangke Liao', 'Wei Dong', 'Bailin Xiao', 'Zhiming Wang', 'Xin Xia', 'Shanshan Li', 'Yue Yu', 'Chen Zeng'] | 2021-03-24 | null | null | null | null | ['code-search', 'code-search'] | ['computer-code', 'computer-vision'] | [-4.55917031e-01 -1.78633109e-01 -5.85405886e-01 -1.07893080e-01
-6.87638640e-01 -6.86577380e-01 3.46947908e-02 2.37838253e-01
6.20447919e-02 3.55613269e-02 9.67370644e-02 -7.11703420e-01
-2.16630816e-01 -9.20864940e-01 -7.32259631e-01 1.12110248e-03
-1.78506947e-03 6.15601540e-02 3.07636976e-01 -1.64460957e-01
6.66079402e-01 8.41798112e-02 -1.38039315e+00 3.21405351e-01
1.43202043e+00 8.23916972e-01 6.65403605e-01 3.72337103e-01
-8.67041886e-01 9.64260161e-01 -5.22571862e-01 -5.56798637e-01
-6.42995313e-02 -3.40938449e-01 -1.11772764e+00 -3.90374839e-01
1.92348182e-01 -2.75385350e-01 -4.78520572e-01 1.39776671e+00
1.95082426e-01 -1.63331926e-01 2.68164754e-01 -1.02726412e+00
-1.22971904e+00 8.75976980e-01 -7.11144805e-01 3.33448857e-01
5.58544397e-01 1.44281402e-01 1.25372839e+00 -8.71995151e-01
5.70957065e-01 1.13591015e+00 5.92585266e-01 4.99084860e-01
-9.50165868e-01 -7.10279286e-01 2.55193561e-01 3.26654136e-01
-1.48024452e+00 -2.01691628e-01 8.41575384e-01 -8.09728980e-01
1.30586934e+00 8.03184286e-02 5.88281035e-01 7.82494605e-01
2.96070248e-01 8.34021747e-01 7.86827579e-02 -3.67773294e-01
2.00571716e-02 -4.17102799e-02 4.31263566e-01 1.22488189e+00
3.78870070e-01 -2.48529583e-01 -2.23731294e-01 -4.72096175e-01
5.53265572e-01 2.70942718e-01 -6.45094037e-01 -3.73140007e-01
-7.76876748e-01 8.27953219e-01 8.10080230e-01 4.02754873e-01
-4.59651537e-02 5.48639476e-01 7.42487609e-01 2.59431630e-01
3.28231245e-01 3.96459877e-01 -5.14201224e-01 -3.13282549e-01
-9.99702215e-01 4.60391417e-02 7.68860340e-01 1.60962403e+00
1.15366530e+00 -5.58611788e-02 -3.32120240e-01 8.58288169e-01
8.37848365e-01 3.90254855e-01 8.43540728e-01 -5.05459964e-01
1.01370788e+00 1.28557181e+00 -3.97989959e-01 -1.28526282e+00
1.79786887e-02 -4.69985038e-01 -4.67216372e-01 -3.23856175e-01
-7.11184293e-02 1.79782227e-01 -6.87199354e-01 1.40624094e+00
-2.95975655e-01 -8.43914524e-02 -2.17181951e-01 7.91665673e-01
1.02625215e+00 5.49766123e-01 -9.81471464e-02 1.09909967e-01
1.12077248e+00 -1.25591886e+00 -3.53248566e-01 -5.74859679e-01
1.03997934e+00 -4.06226397e-01 1.45556426e+00 1.97817624e-01
-7.58575320e-01 -4.66113657e-01 -7.95190513e-01 -1.99688569e-01
-2.36663789e-01 1.76853791e-01 7.43137836e-01 6.07612371e-01
-1.17895424e+00 4.89891410e-01 -9.50574517e-01 -2.44642854e-01
5.52290380e-01 1.41188413e-01 -1.35260582e-01 -2.66410232e-01
-8.54812860e-01 2.02683628e-01 4.95768607e-01 8.76688138e-02
-1.05186880e+00 -6.62209928e-01 -1.05630195e+00 7.45600343e-01
3.13218027e-01 -4.86361742e-01 1.11073303e+00 -1.03815353e+00
-9.50576663e-01 7.15786815e-01 -1.11766540e-01 -2.49847770e-02
9.08200368e-02 -1.96661487e-01 -1.33442149e-01 -2.32412666e-02
3.18967551e-01 1.77937269e-01 4.28840667e-01 -9.83547270e-01
-4.11717206e-01 -1.99975967e-01 4.62562889e-01 -3.40423375e-01
-7.73889959e-01 3.05136442e-01 -9.82014179e-01 -5.69613755e-01
-9.33102444e-02 -6.49860382e-01 -1.11813851e-01 -4.72838581e-02
-3.25816751e-01 -5.18884480e-01 5.30854344e-01 -8.07173371e-01
2.04266143e+00 -2.38856030e+00 2.76456326e-01 2.92571813e-01
6.86530054e-01 2.56700397e-01 -2.83263803e-01 5.35650909e-01
-2.52052061e-02 7.67124355e-01 -3.24780941e-01 -1.41575709e-01
1.38230873e-02 -9.87154990e-03 -3.21750730e-01 2.33468965e-01
1.00797731e-02 1.27441764e+00 -1.21716177e+00 -6.67543352e-01
-3.74015719e-01 3.67138572e-02 -9.41859841e-01 3.37318718e-01
-4.29598868e-01 -2.77336061e-01 -1.10296226e+00 6.76716030e-01
3.70082587e-01 -7.63301611e-01 1.59021523e-02 2.03681141e-01
3.37597430e-02 3.49377960e-01 -4.78438079e-01 2.10770607e+00
-8.57380033e-01 7.41813779e-01 -2.20181420e-01 -6.71962798e-01
1.09085786e+00 2.26445030e-02 8.36553574e-02 -7.85251498e-01
-1.38860777e-01 3.74482155e-01 -2.52935827e-01 -9.12778437e-01
2.44196236e-01 6.04942322e-01 -3.01438749e-01 4.58942026e-01
-4.64934073e-02 1.72791690e-01 2.66667783e-01 4.82849509e-01
1.46822786e+00 2.23584592e-01 1.35067984e-01 -5.54063320e-01
8.02657008e-01 1.16086394e-01 4.70740050e-01 6.37652993e-01
2.69781291e-01 4.95619416e-01 8.58312905e-01 -4.17736053e-01
-4.94593352e-01 -4.59080964e-01 3.14064980e-01 1.06934869e+00
3.61698598e-01 -1.05201447e+00 -9.78616178e-01 -1.08508420e+00
-2.24621296e-02 4.79596525e-01 -4.69088614e-01 -5.74471056e-01
-7.41430879e-01 -2.83844978e-01 6.01281226e-01 6.73183501e-01
4.92900401e-01 -1.22531676e+00 -5.17749965e-01 2.20202535e-01
-2.29550898e-01 -6.30596340e-01 -9.20946896e-01 -1.33311749e-01
-6.77302063e-01 -1.34473610e+00 -4.41932172e-01 -1.06521940e+00
8.20329368e-01 2.70630658e-01 1.53518915e+00 1.13501894e+00
-1.64609373e-01 2.39196584e-01 -7.47203529e-01 2.15865359e-01
-4.08909976e-01 5.64153790e-01 -7.89611876e-01 -5.03225386e-01
3.93250495e-01 -4.16520178e-01 -6.88847840e-01 8.93853530e-02
-1.07337511e+00 -2.23779790e-02 4.59994167e-01 7.71625459e-01
2.25768849e-01 -2.10235372e-01 4.01007563e-01 -8.98401260e-01
1.02118230e+00 -8.09705257e-01 -9.66756940e-01 4.71606433e-01
-9.11033988e-01 3.46576840e-01 9.52217400e-01 -1.79079965e-01
-7.90922701e-01 -2.36110166e-01 -1.75120741e-01 -5.75219154e-01
3.45497578e-01 1.34311664e+00 -3.87084708e-02 -3.15955013e-01
8.12953413e-01 4.35127079e-01 -2.65613675e-01 -6.73343599e-01
1.14327654e-01 8.66810441e-01 2.13180587e-01 -8.06782246e-01
7.50075400e-01 -1.63132668e-01 -5.08418441e-01 -6.33435324e-02
-4.07156765e-01 -3.82375479e-01 -2.77057081e-01 -2.31100954e-02
6.85434103e-01 -6.75173461e-01 -5.70783556e-01 2.43481353e-01
-1.43213618e+00 -3.46540064e-01 3.50651681e-01 7.35927224e-02
-2.60215044e-01 6.86788201e-01 -7.11557448e-01 -4.19318974e-01
-7.43289828e-01 -1.54743445e+00 1.16118610e+00 1.71190128e-01
5.33077717e-02 -9.28287268e-01 2.02335477e-01 5.09289019e-02
5.60259938e-01 -1.36807561e-01 1.52402461e+00 -5.56614816e-01
-1.13117206e+00 -2.29253277e-01 -5.33908546e-01 8.06650743e-02
1.19206913e-01 1.52456209e-01 -6.18854702e-01 -5.31737030e-01
-3.01407963e-01 -1.32198989e-01 7.09421039e-01 -1.39425650e-01
1.62684810e+00 -4.27884161e-01 -6.11939490e-01 8.44229341e-01
1.84177125e+00 3.13075304e-01 7.98301876e-01 4.82958317e-01
1.05383825e+00 1.89833209e-01 2.93029070e-01 4.50993031e-01
5.54977357e-01 5.04241049e-01 8.04602683e-01 3.39795589e-01
-6.21341281e-02 -4.26155418e-01 2.81673551e-01 1.23590577e+00
2.31152609e-01 -4.28839684e-01 -1.41700768e+00 7.54163206e-01
-1.90661800e+00 -4.93041396e-01 -1.49867594e-01 2.01542473e+00
8.55342209e-01 1.52761433e-02 -3.64488184e-01 -4.16827977e-01
6.26564622e-01 2.60548070e-02 -4.79238451e-01 -1.76935837e-01
5.53974926e-01 1.34220898e-01 3.01111698e-01 3.11162591e-01
-4.62112874e-01 9.74352837e-01 5.02350187e+00 1.04005396e+00
-1.04918098e+00 7.39030614e-02 2.43409529e-01 3.97118211e-01
-8.37751687e-01 3.36557448e-01 -5.97586811e-01 7.95968473e-01
6.72873139e-01 -6.62487984e-01 8.30693781e-01 1.33980596e+00
-4.74916436e-02 4.20662582e-01 -1.23386180e+00 1.19419277e+00
4.80655627e-03 -1.50512695e+00 1.14237674e-01 -1.10100374e-01
5.51286399e-01 1.38737530e-01 -3.87319535e-01 8.22171628e-01
2.99417108e-01 -8.95240784e-01 7.83029437e-01 3.88143957e-01
8.30152512e-01 -3.78341943e-01 6.19803965e-01 3.97276640e-01
-1.56616342e+00 -3.10839266e-01 -4.67148840e-01 2.43251547e-01
-3.58106792e-01 3.86868030e-01 -4.46548522e-01 7.21218526e-01
7.56404698e-01 9.93355215e-01 -1.11356640e+00 1.38994920e+00
-2.37428725e-01 6.39036596e-01 1.13782465e-01 -3.79611433e-01
3.18899393e-01 -1.50348181e-02 1.92333251e-01 1.35400021e+00
4.98763800e-01 -3.32843781e-01 1.55325711e-01 1.68768704e+00
-4.08566803e-01 3.08124751e-01 -5.34522593e-01 -3.66045326e-01
5.18721700e-01 1.18468571e+00 -6.69014573e-01 -2.26053968e-01
-7.92292178e-01 6.21066749e-01 6.89873278e-01 6.00771666e-01
-8.23815644e-01 -9.51515794e-01 3.54254395e-01 8.92675593e-02
3.21068943e-01 -1.33769378e-01 6.84975088e-02 -1.35171771e+00
6.09925330e-01 -9.83060539e-01 3.46718788e-01 -8.23799253e-01
-9.49342012e-01 9.11460102e-01 -6.49041831e-02 -1.12229645e+00
-3.14108431e-01 -4.20693040e-01 -8.18941057e-01 1.01010966e+00
-1.40863085e+00 -7.96466231e-01 -6.95557535e-01 4.23528463e-01
6.37453318e-01 -2.99008399e-01 4.13857788e-01 6.78379536e-01
-5.09661019e-01 7.45617032e-01 -9.98827443e-02 4.54555631e-01
1.16173737e-01 -1.02753592e+00 8.57912600e-01 1.01889837e+00
1.41472325e-01 1.17310548e+00 1.22423105e-01 -7.49078155e-01
-1.81846201e+00 -1.16264319e+00 6.94936693e-01 -2.84605622e-01
8.68679643e-01 -5.89720666e-01 -1.32709920e+00 7.33672142e-01
3.95417809e-02 1.46387920e-01 2.18548760e-01 -1.16024539e-01
-4.49466377e-01 -3.29832472e-02 -6.88136935e-01 3.90069336e-01
1.44337308e+00 -8.55098665e-01 -3.77761573e-01 1.47533327e-01
1.15894961e+00 -4.02241975e-01 -6.58309281e-01 3.25737953e-01
1.77111506e-01 -7.94036746e-01 6.18111253e-01 -6.71719968e-01
8.30999970e-01 -2.95080364e-01 4.26156223e-02 -1.13687897e+00
-4.32751805e-01 -4.71368253e-01 -2.17404991e-01 1.34579527e+00
5.15493810e-01 -6.49669170e-01 7.71864057e-01 6.32937968e-01
-4.73892838e-01 -9.86981690e-01 -4.61131066e-01 -6.48426831e-01
-4.18524332e-02 -2.07271904e-01 9.12522495e-01 8.89771163e-01
1.71897084e-01 1.33417085e-01 2.30593264e-01 1.03042116e-02
1.08568177e-01 5.87211192e-01 5.87004423e-01 -1.14613736e+00
-6.00477993e-01 -7.83262968e-01 -4.24021393e-01 -1.48004806e+00
6.33370340e-01 -1.34352386e+00 5.24787493e-02 -1.82544446e+00
4.01409268e-01 -4.80454028e-01 -2.61953592e-01 6.61941826e-01
-3.82412076e-01 -5.96925795e-01 -5.50757013e-02 3.89582515e-01
-6.66438460e-01 6.15306258e-01 1.03922200e+00 -5.21586418e-01
-4.33174446e-02 -1.29633591e-01 -9.59773362e-01 3.00595433e-01
3.70336354e-01 -6.19125009e-01 -6.72464669e-01 -1.10314417e+00
8.49782884e-01 3.68847668e-01 1.64309770e-01 -6.01639450e-01
4.19646353e-01 -3.32020484e-02 -3.04360241e-01 -1.66138947e-01
-6.54246569e-01 -7.30767071e-01 5.07415272e-02 4.12317365e-01
-3.50503147e-01 4.30894732e-01 2.94841737e-01 5.65640211e-01
-4.27084446e-01 -8.74180973e-01 2.22686768e-01 -2.88690865e-01
-8.64468157e-01 5.52478313e-01 -1.17247880e-01 3.97542268e-01
5.02795994e-01 -6.49865903e-03 -6.65686488e-01 2.79537980e-02
-8.78543872e-03 4.63344753e-01 6.34667456e-01 7.52789617e-01
7.22031176e-01 -1.34240603e+00 -4.27278459e-01 1.77123010e-01
4.81946379e-01 5.45808449e-02 -1.64910838e-01 3.62576246e-01
-8.80755901e-01 3.24904203e-01 1.54351279e-01 -4.04692829e-01
-8.79143596e-01 8.28105628e-01 4.54898328e-01 -2.92773396e-01
-6.03513420e-01 7.07554340e-01 4.12959129e-01 -2.67587394e-01
3.70688029e-02 -6.07713461e-01 -2.01641336e-01 -3.64843905e-01
3.50226372e-01 2.87053687e-03 2.63396233e-01 -6.03022203e-02
-5.96370339e-01 6.23262465e-01 -1.34329975e-01 6.54051661e-01
1.19107819e+00 1.88542485e-01 -6.06343806e-01 -7.74407983e-02
1.55080914e+00 1.67752996e-01 -7.81590641e-01 -7.64480010e-02
3.32902312e-01 -8.20621550e-01 1.09931864e-01 -4.16600496e-01
-1.64499259e+00 7.86642611e-01 3.01883310e-01 3.93594444e-01
1.04890513e+00 2.51023591e-01 7.32731044e-01 5.18795848e-01
7.37365246e-01 -4.05110627e-01 1.17353059e-01 5.37283182e-01
8.41431618e-01 -9.32761133e-01 -2.88709402e-01 -4.50752467e-01
8.53955671e-02 1.29883993e+00 9.85217690e-01 -5.48986048e-02
3.85981143e-01 8.21628273e-02 -2.75591850e-01 -6.27567828e-01
-7.54050195e-01 1.13007225e-01 3.30991000e-01 2.85615116e-01
6.56817675e-01 -4.58321601e-01 -1.85706601e-01 7.40399241e-01
1.46811515e-01 1.47312889e-02 3.88644129e-01 9.20201957e-01
-3.42735738e-01 -1.01181281e+00 1.10251874e-01 7.29344249e-01
-3.35048914e-01 -6.61883712e-01 -2.33639598e-01 6.17633522e-01
-3.63151997e-01 7.99925387e-01 -2.80111611e-01 -4.35504198e-01
4.73657846e-01 -2.49671042e-01 1.13170281e-01 -9.79341269e-01
-7.95436680e-01 -4.75067735e-01 -2.45362878e-01 -7.72641063e-01
1.73387378e-01 -1.64043725e-01 -1.34048128e+00 -4.03624699e-02
-6.15345776e-01 4.22746509e-01 4.00241524e-01 6.31188989e-01
7.82117903e-01 6.33189499e-01 5.04587471e-01 -4.25607741e-01
-4.45962518e-01 -7.39380360e-01 -1.21065594e-01 3.77854884e-01
2.98544943e-01 -5.11345565e-01 -4.76786584e-01 5.27078360e-02] | [7.497451305389404, 8.050570487976074] |
cccf1726-49ca-43df-9201-00e104596005 | self-taught-cross-domain-few-shot-learning | 2109.01302 | null | https://arxiv.org/abs/2109.01302v1 | https://arxiv.org/pdf/2109.01302v1.pdf | Self-Taught Cross-Domain Few-Shot Learning with Weakly Supervised Object Localization and Task-Decomposition | The domain shift between the source and target domain is the main challenge in Cross-Domain Few-Shot Learning (CD-FSL). However, the target domain is absolutely unknown during the training on the source domain, which results in lacking directed guidance for target tasks. We observe that since there are similar backgrounds in target domains, it can apply self-labeled samples as prior tasks to transfer knowledge onto target tasks. To this end, we propose a task-expansion-decomposition framework for CD-FSL, called Self-Taught (ST) approach, which alleviates the problem of non-target guidance by constructing task-oriented metric spaces. Specifically, Weakly Supervised Object Localization (WSOL) and self-supervised technologies are employed to enrich task-oriented samples by exchanging and rotating the discriminative regions, which generates a more abundant task set. Then these tasks are decomposed into several tasks to finish the task of few-shot recognition and rotation classification. It helps to transfer the source knowledge onto the target tasks and focus on discriminative regions. We conduct extensive experiments under the cross-domain setting including 8 target domains: CUB, Cars, Places, Plantae, CropDieases, EuroSAT, ISIC, and ChestX. Experimental results demonstrate that the proposed ST approach is applicable to various metric-based models, and provides promising improvements in CD-FSL. | ['Zhongfei Zhang', 'Yanwei Pang', 'Zhong Ji', 'Xiyao Liu'] | 2021-09-03 | null | null | null | null | ['cross-domain-few-shot', 'cross-domain-few-shot-learning'] | ['computer-vision', 'computer-vision'] | [ 3.12445223e-01 -4.13775742e-01 -3.90699714e-01 -4.46327686e-01
-9.14849877e-01 -5.14932752e-01 6.54418826e-01 -3.63342971e-01
-1.31620690e-01 6.96494341e-01 1.17417276e-01 2.00324044e-01
-1.50810376e-01 -5.80465615e-01 -5.33794582e-01 -7.44108856e-01
2.86720872e-01 3.87393057e-01 5.14856577e-01 -4.07142788e-01
1.02299891e-01 1.94677442e-01 -1.37520945e+00 4.14523840e-01
1.05242622e+00 9.35213625e-01 6.33875430e-01 8.75175595e-02
-1.88266650e-01 8.19042563e-01 -4.56872851e-01 5.71028404e-02
2.78894454e-01 -5.15693724e-01 -5.92009604e-01 3.86509627e-01
1.57807246e-01 -2.34302253e-01 -1.75156787e-01 1.13067985e+00
4.51636076e-01 4.30926830e-01 9.22182143e-01 -1.37966204e+00
-9.16317105e-01 9.16500017e-02 -8.22544754e-01 1.92735612e-01
1.45220160e-01 4.73968089e-02 7.90023685e-01 -1.43498766e+00
7.25863934e-01 1.37840652e+00 4.17825788e-01 6.52956247e-01
-9.63415563e-01 -8.39021206e-01 2.61363268e-01 4.11666930e-01
-1.37789369e+00 -4.53522235e-01 1.11088145e+00 -5.37885666e-01
3.85385543e-01 -1.52448878e-01 1.33060113e-01 1.32340586e+00
-2.68299997e-01 1.07701719e+00 1.12982297e+00 -4.03860033e-01
3.61244380e-01 3.40052396e-01 2.71173775e-01 2.69476622e-01
1.50319666e-01 1.86261490e-01 -6.07367575e-01 6.04681410e-02
5.52422225e-01 2.62010694e-01 -3.18955034e-01 -9.66471493e-01
-1.33336031e+00 7.60238230e-01 3.06456834e-01 3.46432805e-01
-1.13105819e-01 -5.09187162e-01 6.05558634e-01 4.24511373e-01
7.89504528e-01 1.88861281e-01 -6.19430840e-01 5.84511496e-02
-6.04336441e-01 8.43658894e-02 5.21777153e-01 1.50496697e+00
1.02451730e+00 7.13280514e-02 -3.40452582e-01 1.29346371e+00
1.73420772e-01 6.01929724e-01 6.51022911e-01 -4.98377323e-01
8.26730847e-01 5.79180121e-01 1.68347031e-01 -8.40489268e-01
-1.13686673e-01 -5.01471400e-01 -7.43275166e-01 -2.75149625e-02
1.96593985e-01 -1.01815224e-01 -8.77003312e-01 1.61800230e+00
5.75535715e-01 4.69175339e-01 3.04106712e-01 9.54522967e-01
7.50476897e-01 6.31858945e-01 -2.68184580e-02 -1.05270058e-01
1.15240049e+00 -1.28366745e+00 -5.22019923e-01 -5.37646711e-01
9.54248905e-01 -6.59009457e-01 1.28965151e+00 1.18110254e-01
-3.25730532e-01 -1.10204649e+00 -1.19819152e+00 1.50299475e-01
-5.92418194e-01 3.56841087e-01 2.04557553e-01 3.22640568e-01
-2.56446272e-01 4.07538086e-01 -4.12105590e-01 -5.11277974e-01
6.82799637e-01 -3.06447595e-01 -3.88277829e-01 -7.72181630e-01
-1.34476781e+00 8.73775184e-01 8.78410041e-01 -2.39029542e-01
-1.32061052e+00 -8.60511661e-01 -1.19028950e+00 -7.48606250e-02
7.27136195e-01 -5.18960170e-02 1.02544880e+00 -9.12937164e-01
-1.20229840e+00 8.96336854e-01 1.53700300e-02 -1.84641123e-01
3.28220129e-01 -2.50319749e-01 -6.42229259e-01 -1.66726857e-01
5.97572327e-01 4.21784490e-01 1.09650159e+00 -1.36188567e+00
-1.02037597e+00 -5.34299314e-01 -5.94248660e-02 4.30848509e-01
-4.65512931e-01 -7.28882626e-02 -5.28845549e-01 -8.34564686e-01
-2.60107778e-02 -8.28169703e-01 1.12278640e-01 -2.12619249e-02
-3.26853618e-02 -2.78331608e-01 1.17464912e+00 -5.44823527e-01
9.66609538e-01 -2.48567390e+00 1.34029433e-01 -1.66232064e-01
-3.58413003e-04 3.44489813e-01 -4.80910093e-01 1.71262860e-01
-2.37373218e-01 -5.12125850e-01 -4.27778512e-01 -1.74992934e-01
-8.36461186e-02 3.90430212e-01 -4.37175006e-01 4.84932244e-01
3.78729343e-01 7.03996718e-01 -1.20041525e+00 -5.84774256e-01
2.97312081e-01 3.03102527e-02 -1.43196523e-01 3.65025401e-01
-1.30257234e-01 5.48587143e-01 -6.81169510e-01 8.89838278e-01
1.05504751e+00 -1.34248644e-01 -4.49704900e-02 -1.63600922e-01
9.29165110e-02 -8.03316981e-02 -1.36394596e+00 2.11530185e+00
-5.41765332e-01 2.66989529e-01 -6.47863522e-02 -1.39481807e+00
1.26874506e+00 -5.39095439e-02 4.75224644e-01 -7.76221752e-01
-1.16225757e-01 2.14627028e-01 -2.01852128e-01 -5.03394663e-01
9.97438356e-02 -2.00859740e-01 -2.15726882e-01 2.95014888e-01
4.60741878e-01 -8.05246979e-02 2.60604620e-01 7.10463673e-02
7.83878982e-01 5.23244560e-01 4.53785181e-01 -2.91657716e-01
7.12920427e-01 2.13357389e-01 9.41772938e-01 3.31052214e-01
-5.87238371e-01 6.03268325e-01 1.09711088e-01 -1.82526886e-01
-9.65439320e-01 -1.15775692e+00 -2.06582248e-02 1.61292911e+00
5.13126731e-01 -7.08518848e-02 -4.24776763e-01 -1.12130642e+00
4.00720313e-02 7.31490374e-01 -7.68432975e-01 -6.43901765e-01
-4.59647804e-01 -4.76754934e-01 1.16822951e-01 6.87270641e-01
6.79768085e-01 -1.00237381e+00 -3.87922913e-01 1.08868614e-01
-1.96796596e-01 -1.14461017e+00 -6.79998815e-01 3.87301177e-01
-7.92507946e-01 -1.22209263e+00 -1.22299492e+00 -1.11609352e+00
6.36169910e-01 8.95125389e-01 9.53037560e-01 -6.97221041e-01
-5.95695861e-02 3.28292459e-01 -6.64737344e-01 -3.83104771e-01
-8.62000287e-02 3.33497748e-02 1.85939223e-01 1.81383133e-01
7.05557168e-01 -4.12574857e-01 -3.60155106e-01 6.93535924e-01
-6.90620244e-01 -1.78995281e-01 7.61372626e-01 1.21285474e+00
6.09560788e-01 -9.45793316e-02 7.84815967e-01 -1.08375072e+00
4.55979854e-01 -8.14550042e-01 -4.87271160e-01 4.75780725e-01
-3.39945078e-01 -6.29692376e-02 7.43481755e-01 -7.17168212e-01
-1.48344409e+00 2.03460574e-01 5.44433713e-01 -8.29158664e-01
-3.55750114e-01 3.41319382e-01 -5.39073110e-01 6.38516853e-03
9.39242661e-01 5.48055708e-01 -6.16052523e-02 -6.29477680e-01
3.84940565e-01 7.53939927e-01 4.56123412e-01 -7.53483415e-01
9.78057325e-01 5.53050280e-01 -3.03819209e-01 -7.65719593e-01
-1.22393155e+00 -1.00984550e+00 -8.81962121e-01 -4.76486608e-03
6.28558695e-01 -1.13570440e+00 2.64812917e-01 4.31474417e-01
-1.07224643e+00 -1.91429615e-01 -5.90427876e-01 5.15611172e-01
-5.82689762e-01 4.30677056e-01 6.18238235e-03 -5.31352282e-01
-5.13676591e-02 -9.18377578e-01 1.19491041e+00 2.90358037e-01
1.98698699e-01 -9.15670037e-01 2.83502758e-01 1.41291499e-01
7.95946494e-02 -1.69847775e-02 7.62375116e-01 -9.76695299e-01
-1.19842194e-01 1.00330086e-02 -5.21723151e-01 6.84250891e-01
5.26860297e-01 -7.32521594e-01 -9.72775638e-01 -4.96559501e-01
2.59375393e-01 -8.31792653e-01 7.25091696e-01 5.22668585e-02
9.90984321e-01 3.13669115e-01 -4.59885061e-01 6.56106234e-01
1.24336958e+00 2.93409824e-01 2.89178252e-01 2.86994278e-01
7.15205610e-01 8.67087245e-01 1.63280439e+00 5.45050323e-01
2.16657951e-01 5.71777880e-01 1.90113842e-01 -6.87371641e-02
-3.11818063e-01 -2.44410396e-01 4.54853058e-01 7.87154019e-01
3.04512501e-01 2.47354418e-01 -8.78186822e-01 9.01121795e-01
-1.93086040e+00 -7.75559366e-01 2.69811958e-01 1.98710704e+00
7.76342571e-01 6.55285791e-02 1.32779945e-02 -6.37046322e-02
1.01589656e+00 3.60490501e-01 -1.00032258e+00 2.77297020e-01
-2.64004953e-02 3.63259651e-02 3.13972324e-01 -1.05488062e-01
-1.38901794e+00 1.04485047e+00 5.15788984e+00 1.28641450e+00
-8.38088810e-01 4.43048894e-01 2.18236908e-01 1.31363615e-01
2.20505744e-01 -3.66831347e-02 -9.29538965e-01 5.21247864e-01
3.03143114e-01 -3.25934649e-01 1.96357593e-01 1.47733402e+00
-1.40742183e-01 1.78032473e-01 -1.26828539e+00 1.17816162e+00
3.47444117e-01 -7.90650725e-01 -9.64272022e-02 -2.66976535e-01
8.36061835e-01 -1.45834126e-02 9.27940682e-02 9.89247441e-01
3.43650460e-01 -4.88289177e-01 3.92218947e-01 1.15458854e-01
1.16484416e+00 -5.92033088e-01 5.04390836e-01 6.26453400e-01
-1.48010814e+00 -2.77758420e-01 -1.01435471e+00 2.85410047e-01
-1.71509504e-01 4.17328238e-01 -7.37202644e-01 9.11743581e-01
6.83497608e-01 1.29353166e+00 -3.89721483e-01 7.65078962e-01
-5.00924923e-02 2.54110545e-01 1.97533801e-01 1.97515190e-01
2.43586943e-01 -2.68277884e-01 4.11776155e-01 9.88461673e-01
3.68788898e-01 9.34977829e-02 5.59009790e-01 8.05929601e-01
1.16871633e-01 2.38088705e-02 -8.25252354e-01 1.31161198e-01
5.58592379e-01 1.17124033e+00 -3.81500423e-01 -4.64924991e-01
-7.10101306e-01 1.12186289e+00 3.40334505e-01 4.72261429e-01
-9.42021072e-01 -7.20744252e-01 6.54565036e-01 -2.01505870e-01
4.73522931e-01 -9.31016803e-02 -1.67722832e-02 -1.35707557e+00
9.75246951e-02 -1.06042600e+00 5.84508657e-01 -6.20694697e-01
-1.66099930e+00 2.58808911e-01 8.27019364e-02 -1.89521587e+00
6.41477033e-02 -6.94888830e-01 -6.03240967e-01 8.88173997e-01
-1.90739691e+00 -1.38892245e+00 -6.41792715e-01 9.72378433e-01
1.14947104e+00 -5.75278997e-01 5.85710227e-01 4.40419883e-01
-4.75810558e-01 5.04641652e-01 5.23407519e-01 7.83547461e-02
1.28692853e+00 -1.02647161e+00 7.40351155e-02 8.15868974e-01
2.13773791e-02 2.96713352e-01 3.49599391e-01 -7.33646154e-01
-1.18421233e+00 -1.53383434e+00 2.77715832e-01 -2.88919747e-01
6.57519996e-01 -4.18731004e-01 -9.83253896e-01 5.62898934e-01
-1.97685942e-01 4.36273217e-01 5.90539932e-01 5.29416278e-02
-6.15143418e-01 -3.40744346e-01 -9.71014678e-01 2.08329141e-01
1.30736208e+00 -5.98394275e-01 -1.03092062e+00 5.40638447e-01
8.49818945e-01 -1.99225083e-01 -6.08714104e-01 4.14004087e-01
1.56469241e-01 -5.86793661e-01 1.03312695e+00 -8.10398817e-01
3.69057804e-01 -3.38630915e-01 -4.55566108e-01 -1.65662980e+00
-4.72263038e-01 7.81604648e-03 -1.40063375e-01 1.46784413e+00
1.29978228e-02 -4.28196996e-01 8.10218573e-01 9.73048713e-03
-3.24945658e-01 -4.28627104e-01 -6.85489416e-01 -1.25450969e+00
8.13618749e-02 -2.61250198e-01 5.37401319e-01 1.37314451e+00
-1.56383872e-01 5.19890726e-01 -4.63841438e-01 -1.38133001e-02
7.72315741e-01 4.54571635e-01 9.26436901e-01 -1.31797230e+00
-2.01025069e-01 9.53289494e-02 -1.32809982e-01 -1.35165620e+00
2.97583073e-01 -9.54059541e-01 2.74134099e-01 -1.23198688e+00
2.34874427e-01 -6.12097859e-01 -5.91225624e-01 3.94281238e-01
-2.16806516e-01 -1.50628358e-01 1.85203522e-01 3.84354085e-01
-9.19997454e-01 1.03436029e+00 1.37274396e+00 -4.03838158e-01
-1.00049481e-01 2.27193441e-03 -6.50573492e-01 6.19097888e-01
5.74787617e-01 -4.92489427e-01 -7.96785712e-01 -4.12216365e-01
-5.03955781e-01 -3.38047504e-01 1.35685399e-01 -1.14425981e+00
6.83035329e-02 -4.05760586e-01 3.63615215e-01 -7.45351315e-01
2.83857226e-01 -9.82735872e-01 -5.31146348e-01 3.88841510e-01
-1.30245864e-01 -6.22734189e-01 -1.18097346e-02 7.88458765e-01
-4.51550424e-01 -2.82437265e-01 1.02575660e+00 -7.17549697e-02
-1.45552039e+00 4.80551034e-01 2.69190073e-01 5.62066913e-01
1.37523055e+00 -2.32667387e-01 -3.82610261e-01 1.06397703e-01
-7.40572274e-01 5.97104013e-01 2.62815803e-01 6.55850232e-01
7.37393081e-01 -1.76790607e+00 -6.69903994e-01 1.70795545e-01
9.02103722e-01 2.46114984e-01 4.63970244e-01 6.45156682e-01
1.83832422e-02 3.62949908e-01 -5.43347359e-01 -7.93662250e-01
-1.03312981e+00 9.90048349e-01 4.08746973e-02 -3.45860451e-01
-5.02406299e-01 8.14226329e-01 8.01465809e-01 -7.03860462e-01
2.73941278e-01 -5.57758696e-02 -3.79799396e-01 3.31552029e-01
5.60169518e-01 4.97090101e-01 -6.15701377e-02 -5.29799819e-01
-4.17733043e-01 6.66251957e-01 -2.69311547e-01 2.20746025e-01
1.28286457e+00 -2.98316717e-01 3.22744101e-01 6.09173477e-01
1.28395593e+00 -4.31264848e-01 -1.52984607e+00 -8.66971195e-01
1.95083782e-01 -5.99855185e-01 -3.32382709e-01 -6.69425905e-01
-7.65235662e-01 1.20048833e+00 7.50431716e-01 -1.97072387e-01
1.11848152e+00 6.36699796e-02 7.82753885e-01 4.12631869e-01
5.20679414e-01 -1.55897808e+00 5.44507921e-01 7.63443708e-01
8.92972410e-01 -1.58959734e+00 -2.46853694e-01 -4.46897507e-01
-1.01252520e+00 7.87384927e-01 1.11819482e+00 -1.64577529e-01
6.12937033e-01 -1.16746508e-01 -2.28243843e-01 -4.21262011e-02
-5.06372929e-01 -4.61394042e-01 2.34841481e-01 1.10000575e+00
7.25709423e-02 -7.77062252e-02 3.43366079e-02 9.50809479e-01
4.93636578e-01 1.41518205e-01 7.08043901e-03 1.10973632e+00
-6.20095670e-01 -9.31182742e-01 -4.76581901e-01 2.97303826e-01
2.21221566e-01 7.40815997e-02 -2.42333546e-01 8.49595666e-01
4.10407990e-01 7.15336978e-01 -1.50280908e-01 -3.57679725e-01
6.31784320e-01 1.46328315e-01 3.38273227e-01 -1.11216545e+00
1.62512526e-01 2.21334919e-02 -1.76946484e-02 -4.83962953e-01
-2.94761807e-01 -5.28187931e-01 -9.71778750e-01 2.98721731e-01
-3.88036042e-01 1.14867471e-01 2.38382578e-01 8.92200291e-01
3.91613603e-01 6.09069884e-01 9.07926440e-01 -7.26388633e-01
-9.63623822e-01 -1.13993716e+00 -8.36500466e-01 6.41675532e-01
1.85589835e-01 -1.25865483e+00 -2.99582452e-01 1.54890135e-01] | [10.314334869384766, 2.9253389835357666] |
02039c4e-e672-4b6e-ba6a-9a4aa10edc9d | data-centric-domain-adaptation-for-historical | 2107.00927 | null | https://arxiv.org/abs/2107.00927v1 | https://arxiv.org/pdf/2107.00927v1.pdf | Data Centric Domain Adaptation for Historical Text with OCR Errors | We propose new methods for in-domain and cross-domain Named Entity Recognition (NER) on historical data for Dutch and French. For the cross-domain case, we address domain shift by integrating unsupervised in-domain data via contextualized string embeddings; and OCR errors by injecting synthetic OCR errors into the source domain and address data centric domain adaptation. We propose a general approach to imitate OCR errors in arbitrary input data. Our cross-domain as well as our in-domain results outperform several strong baselines and establish state-of-the-art results. We publish preprocessed versions of the French and Dutch Europeana NER corpora. | ['Hinrich Schütze', 'Benjamin Roth', 'Nina Poerner', 'Stefan Schweter', 'Luisa März'] | 2021-07-02 | null | null | null | null | ['cross-domain-named-entity-recognition'] | ['natural-language-processing'] | [ 7.78241903e-02 -2.26751402e-01 -1.01193562e-01 -6.09027803e-01
-1.40656400e+00 -1.42418277e+00 6.62890732e-01 2.29515880e-01
-1.07459748e+00 9.89751041e-01 4.10689116e-01 -2.21481323e-01
6.95090443e-02 -5.12882113e-01 -7.58658171e-01 -1.09590232e-01
2.93783814e-01 6.99282825e-01 2.85535574e-01 -2.89567292e-01
7.87926540e-02 7.24692881e-01 -7.28793621e-01 3.83835614e-01
1.05816126e+00 2.72049457e-01 -2.43152529e-01 7.07302630e-01
-3.72145176e-01 7.99386352e-02 -1.15032518e+00 -1.04775381e+00
1.47686794e-01 -9.00049582e-02 -1.07054484e+00 -2.98621327e-01
3.91515225e-01 2.49792993e-01 -3.17036450e-01 8.44414353e-01
8.31693172e-01 8.47904235e-02 7.48928964e-01 -8.72411907e-01
-1.20083714e+00 4.82023954e-01 -2.12654367e-01 4.22307760e-01
4.20889139e-01 -5.85155003e-02 8.74321282e-01 -9.37013924e-01
1.19444108e+00 1.04555368e+00 8.70305955e-01 9.61346686e-01
-1.49103296e+00 -4.86923337e-01 1.41619503e-01 -1.27179921e-01
-1.30497801e+00 -4.36478168e-01 5.05822361e-01 -1.91716254e-01
1.41604507e+00 -1.19344220e-01 -8.45054537e-02 1.76292503e+00
-2.56951183e-01 7.69714653e-01 8.68698359e-01 -5.84064007e-01
2.81824112e-01 1.14194743e-01 2.95713563e-02 9.01793763e-02
3.28897595e-01 2.55272985e-01 -2.98004091e-01 -3.65686744e-01
5.15006542e-01 -4.59033340e-01 4.26070951e-02 -1.82153396e-02
-1.38542831e+00 6.84107006e-01 7.55125657e-02 6.50251389e-01
-2.57158071e-01 -1.71437293e-01 7.07131326e-01 4.63461310e-01
6.38241351e-01 8.19226384e-01 -1.37806845e+00 -4.62567925e-01
-8.94873381e-01 1.55595347e-01 1.02624440e+00 1.16388381e+00
5.15679121e-01 5.21150231e-02 5.92139438e-02 1.30300653e+00
-1.83290526e-01 6.22218788e-01 6.46462083e-01 -6.15458310e-01
9.06418204e-01 3.37348938e-01 4.14985955e-01 -3.46787244e-01
-4.20945883e-01 -8.26075790e-04 -2.71720767e-01 -1.17307432e-01
7.29408264e-01 -5.66915393e-01 -1.25939524e+00 1.87836921e+00
2.34183043e-01 -1.55465350e-01 5.96765459e-01 3.75762194e-01
6.94950581e-01 7.81380773e-01 4.59730566e-01 1.90417960e-01
1.46454608e+00 -8.65876615e-01 -6.95694625e-01 -4.65580106e-01
9.08102930e-01 -7.52547562e-01 9.60276425e-01 1.71288297e-01
-7.92973638e-01 -6.10696495e-01 -7.78928697e-01 -2.98700392e-01
-1.07865894e+00 3.89352083e-01 -1.96421608e-01 7.97467411e-01
-7.15314865e-01 5.05214989e-01 -6.39467776e-01 -7.07899392e-01
1.39010638e-01 2.07097471e-01 -7.02743053e-01 -1.44901142e-01
-1.54218280e+00 1.14400029e+00 7.26070821e-01 -4.27840889e-01
-4.96247530e-01 -1.02071404e+00 -1.06199563e+00 -3.42831910e-01
-7.83448517e-02 -3.51887614e-01 1.35396361e+00 -8.27508330e-01
-1.26089180e+00 1.27645993e+00 -1.70741379e-01 -3.53555679e-01
3.23842525e-01 -5.17478228e-01 -1.48533010e+00 -1.43188238e-01
2.32298553e-01 6.68846846e-01 3.19600761e-01 -1.14479280e+00
-4.92274314e-01 -3.24024409e-01 -4.78651375e-01 -5.78973331e-02
-3.16006780e-01 4.38553423e-01 -2.04557389e-01 -1.20521939e+00
-4.76178378e-01 -8.30584228e-01 -1.64981365e-01 -4.20961916e-01
-3.27712834e-01 -3.39568585e-01 5.80994427e-01 -1.12421679e+00
1.38165760e+00 -2.36172152e+00 -3.06092370e-02 -1.63569897e-01
-3.55340570e-01 4.76950914e-01 -6.25660539e-01 5.32136202e-01
-2.17571378e-01 5.22009492e-01 -5.52873850e-01 -2.81803817e-01
3.21734428e-01 4.30025339e-01 -2.94841796e-01 3.03431690e-01
9.23785448e-01 8.59941840e-01 -1.01502287e+00 -4.43690270e-01
-1.50930852e-01 4.63708043e-01 -3.68570864e-01 2.56882161e-01
-1.70124367e-01 3.22228283e-01 -1.22873545e-01 5.75119138e-01
7.99496353e-01 3.25465053e-01 2.93598413e-01 5.99592291e-02
-1.98724285e-01 7.55898654e-01 -1.07978928e+00 2.10260105e+00
-5.59649289e-01 5.08091331e-01 -2.50395834e-01 -5.08505762e-01
1.14530659e+00 4.96735364e-01 -4.09263112e-02 -7.70183027e-01
-3.12978208e-01 5.99731803e-01 -4.47587132e-01 -2.75072902e-01
9.04257715e-01 -3.66598427e-01 -9.44755375e-01 1.02725171e-01
5.92503667e-01 -1.13799796e-01 2.38783628e-01 -1.52535275e-01
1.16992354e+00 3.91422123e-01 3.46401781e-01 -9.33575630e-02
2.32209802e-01 3.87482882e-01 7.49318361e-01 5.54951191e-01
-4.70624924e-01 7.68601656e-01 4.35011238e-01 -3.42947334e-01
-1.47865415e+00 -1.38619387e+00 -3.58783185e-01 1.35388231e+00
-2.95013309e-01 -5.48632741e-02 -7.09833980e-01 -1.34243977e+00
1.11261398e-01 9.65377271e-01 -6.57121420e-01 2.76178926e-01
-9.11454022e-01 -4.98201936e-01 1.45187294e+00 8.31660271e-01
2.12315828e-01 -1.10693216e+00 1.56082720e-01 4.88103181e-01
-1.26313850e-01 -1.52287984e+00 -7.09004045e-01 4.80491698e-01
-7.48586595e-01 -7.12944210e-01 -1.16075277e+00 -1.23623252e+00
2.50337392e-01 -6.49454236e-01 1.41961122e+00 -7.86125124e-01
-1.38791546e-01 2.44376436e-01 -5.12671709e-01 -2.42807239e-01
-7.10144520e-01 5.73180795e-01 1.30309165e-01 -4.79869872e-01
9.39156532e-01 -2.63785362e-01 -9.43498388e-02 3.34869146e-01
-1.03347301e+00 -8.67451668e-01 4.61219728e-01 8.95386577e-01
4.49197114e-01 -5.17358303e-01 7.19062328e-01 -1.12914002e+00
6.86460257e-01 -5.39598167e-01 -5.31536281e-01 4.81146872e-01
-2.33880058e-01 3.87592316e-01 8.04008067e-01 -5.14216244e-01
-1.42896414e+00 1.47796348e-01 -3.91121060e-01 -8.56331587e-02
-8.63615394e-01 3.23583893e-02 -4.26753193e-01 4.31064397e-01
1.07506001e+00 6.25030370e-03 -9.31166589e-01 -9.50396419e-01
8.08542252e-01 9.47801471e-01 8.49581540e-01 -8.08064461e-01
7.59545982e-01 8.29493627e-02 -9.23651814e-01 -6.96783662e-01
-5.04879117e-01 -5.81628442e-01 -1.09075856e+00 4.60695654e-01
1.16943944e+00 -1.06870925e+00 1.79717273e-01 4.22927022e-01
-1.78362274e+00 -1.78805903e-01 -5.00772476e-01 3.07433635e-01
-3.06426674e-01 2.96062201e-01 -7.06099689e-01 -4.70189810e-01
-6.88069835e-02 -6.22532845e-01 1.17018843e+00 2.08613783e-01
-4.15139437e-01 -1.25153685e+00 7.94884682e-01 -9.93209258e-02
4.78553884e-02 3.11245680e-01 7.92365253e-01 -1.55163074e+00
3.22237819e-01 -9.27665606e-02 -2.01077759e-01 4.17360604e-01
-4.24823165e-03 -1.26895308e-01 -9.50211704e-01 -1.47833109e-01
-6.01544201e-01 -2.65252233e-01 6.28756940e-01 -2.11901397e-01
3.99393767e-01 -5.32218367e-02 -3.71041656e-01 5.29223502e-01
1.54361796e+00 3.40256989e-01 7.26273894e-01 7.13897169e-01
4.10714030e-01 5.29980779e-01 4.79298234e-01 3.03542018e-01
2.52080590e-01 5.07695317e-01 -4.09651577e-01 -2.22161472e-01
-1.63518697e-01 -5.38475692e-01 3.96703660e-01 7.17880905e-01
4.14896905e-01 -6.30298555e-01 -1.21350837e+00 1.40528989e+00
-1.34291744e+00 -7.86831975e-01 5.04698642e-02 1.89887083e+00
1.17651498e+00 1.91012211e-02 2.26419911e-01 -2.46465042e-01
1.04102111e+00 4.24655341e-02 -6.19371355e-01 -8.77332091e-01
-5.51092982e-01 7.61308432e-01 7.86172807e-01 2.36171752e-01
-1.42426598e+00 1.31802237e+00 7.12542963e+00 6.14748597e-01
-6.55126989e-01 4.16217506e-01 2.58128077e-01 1.13061987e-01
-4.27218586e-01 -1.46065637e-01 -1.01865602e+00 3.79094273e-01
1.45673871e+00 -1.10925645e-01 1.81010440e-01 9.35521245e-01
-5.04868686e-01 4.78964299e-01 -1.12580407e+00 7.82478869e-01
-9.82357040e-02 -1.04566932e+00 -1.70575395e-01 -1.03626862e-01
1.00328588e+00 4.77236599e-01 -2.59154081e-01 4.98156130e-01
9.82246220e-01 -6.74397111e-01 4.82661158e-01 1.47459432e-01
1.27813125e+00 -8.32867980e-01 6.73358202e-01 -2.06657246e-01
-9.34256434e-01 3.04068029e-01 -4.63402033e-01 5.97055197e-01
3.49679738e-01 6.03219271e-01 -6.61922395e-01 7.21698761e-01
6.80763841e-01 6.39664769e-01 -5.53677976e-01 8.50191355e-01
-4.28444356e-01 6.18892312e-01 -4.03788418e-01 -2.54019946e-02
1.64797634e-01 2.66047329e-01 6.11569405e-01 2.05373788e+00
2.98258245e-01 -1.28991440e-01 -3.05339783e-01 6.30629003e-01
-5.37022769e-01 2.07389668e-01 -7.32468188e-01 -4.44861174e-01
7.95000911e-01 7.77696609e-01 -2.78837293e-01 -4.54668164e-01
-4.00312006e-01 1.69096434e+00 5.22531390e-01 6.34625137e-01
-8.19689393e-01 -1.28589356e+00 1.17096400e+00 -3.65509987e-01
9.10303295e-01 -3.96952301e-01 -4.31342691e-01 -1.50536799e+00
3.18402126e-02 -7.89729476e-01 8.91231596e-01 -5.00979900e-01
-1.87894487e+00 9.00083303e-01 -3.55812758e-01 -1.07808256e+00
-2.11708754e-01 -9.49036539e-01 -2.64070004e-01 9.40541685e-01
-1.71820176e+00 -8.18461776e-01 3.11326057e-01 5.87125063e-01
5.49301624e-01 -1.56956568e-01 1.16316283e+00 5.94244659e-01
-3.21359217e-01 1.02953768e+00 7.23802149e-01 8.46267164e-01
1.38769031e+00 -1.40146577e+00 1.45570493e+00 9.73862469e-01
3.19066793e-01 5.66432297e-01 4.45755124e-01 -7.89732456e-01
-8.38825703e-01 -1.07763624e+00 1.56044650e+00 -1.08053815e+00
9.56772387e-01 -6.75869107e-01 -9.34875846e-01 8.57313037e-01
2.84280121e-01 1.14208125e-01 8.29200804e-01 3.15219671e-01
-9.00550663e-01 2.41442978e-01 -1.45319164e+00 3.25495332e-01
1.15991139e+00 -8.02709103e-01 -1.42205465e+00 9.98558030e-02
1.05622542e+00 -2.91035295e-01 -1.22684121e+00 1.31913975e-01
2.87567556e-01 -3.77124280e-01 7.86232352e-01 -1.18200266e+00
2.41737992e-01 -2.91962624e-01 -2.38166124e-01 -1.66867173e+00
-2.09227249e-01 -5.14094353e-01 4.10747707e-01 1.73694241e+00
1.03569007e+00 -6.92536056e-01 4.49716419e-01 4.30510461e-01
-1.23754486e-01 3.02074552e-01 -1.20716786e+00 -1.21485078e+00
7.84941852e-01 -5.24219573e-01 6.40458763e-01 1.29471457e+00
6.62971884e-02 1.05676994e-01 -1.52280806e-02 4.62935448e-01
2.63976365e-01 -4.25369710e-01 3.65381986e-01 -1.12641358e+00
1.84708610e-02 -8.47554207e-02 -3.86080921e-01 -9.11424458e-01
3.22377563e-01 -7.18828201e-01 1.91935897e-01 -1.40594840e+00
-3.81366283e-01 -2.31256276e-01 -5.25867462e-01 5.00918150e-01
1.53273216e-03 1.58025593e-01 1.12206668e-01 -1.06512807e-01
-6.43687785e-01 2.30990201e-01 3.80437911e-01 5.94262667e-02
-2.89672971e-01 -6.35425448e-01 -4.75616217e-01 2.66600668e-01
6.87481701e-01 -1.04632151e+00 4.25340146e-01 -9.57059741e-01
7.44478554e-02 -2.54545003e-01 -7.28148222e-02 -9.89240646e-01
7.00630546e-02 -1.30736902e-01 5.74443281e-01 -3.88787955e-01
-3.27483974e-02 -6.24102771e-01 -2.33919874e-01 8.21564421e-02
-3.52623999e-01 4.76678461e-01 6.89910412e-01 6.34152353e-01
-2.91326344e-01 -1.88896850e-01 9.12177503e-01 -5.59645891e-02
-1.06150854e+00 -1.30015910e-01 -3.21623266e-01 9.27715719e-01
7.26207376e-01 -1.86186567e-01 -6.36167288e-01 2.83465832e-01
-9.42497194e-01 5.35033606e-02 7.09764242e-01 7.51596630e-01
1.90892756e-01 -1.41199481e+00 -9.14093435e-01 2.81091034e-01
6.23821974e-01 -4.72946644e-01 5.24474196e-02 1.27020972e-02
-4.74543720e-01 3.56209308e-01 -1.69063374e-01 -1.64795727e-01
-8.04072738e-01 6.28323078e-01 2.23533690e-01 -5.28090656e-01
1.80830173e-02 8.66235673e-01 -1.30425513e-01 -1.38636112e+00
2.56499015e-02 -1.94831818e-01 1.37207001e-01 9.52186137e-02
4.80386466e-01 3.93487960e-01 2.53817290e-01 -4.90240693e-01
-8.79693508e-01 5.34008563e-01 -5.67053221e-02 -5.81495285e-01
1.45186329e+00 -3.34394276e-02 4.64657664e-01 3.70585263e-01
1.53096378e+00 3.77376348e-01 -1.00784671e+00 -3.16621721e-01
6.13099337e-01 -5.45943379e-02 -4.26335424e-01 -1.53051651e+00
-3.07355881e-01 7.53690720e-01 5.05839050e-01 -7.96144754e-02
1.05599892e+00 2.24039122e-01 1.03383636e+00 4.41034496e-01
1.52729839e-01 -1.66570091e+00 -3.51195574e-01 1.01331723e+00
4.16048616e-01 -1.16481817e+00 -5.28618693e-01 1.00386225e-01
-9.37179863e-01 1.20570791e+00 3.32160562e-01 -1.31439120e-01
3.80501062e-01 4.43914682e-01 3.42066109e-01 1.92632765e-01
-4.48824048e-01 -2.67199576e-01 1.65922597e-01 1.08575797e+00
5.07199764e-01 -3.06139104e-02 -3.99144381e-01 8.54169488e-01
7.96055701e-03 1.32604456e-02 3.81900609e-01 9.89656210e-01
-2.41789483e-02 -1.59931636e+00 -4.07179683e-01 -2.30695650e-01
-7.67636240e-01 -4.82270241e-01 -7.32792318e-01 1.07769430e+00
1.46906689e-01 8.73113990e-01 8.97566080e-02 -3.16810161e-01
1.04930985e+00 7.46618211e-01 2.44894981e-01 -5.50593197e-01
-8.69484484e-01 -2.48038828e-01 6.88741148e-01 -1.89990044e-01
-1.76111758e-01 -7.36784220e-01 -1.25346804e+00 -7.80491680e-02
-1.40493887e-03 2.77504891e-01 8.91756058e-01 7.08236575e-01
7.54108548e-01 1.63177073e-01 3.86493981e-01 -2.82450289e-01
-4.33788151e-01 -9.63795304e-01 -4.72492516e-01 8.82457495e-01
2.37349555e-01 -2.00702101e-01 -4.76322442e-01 1.94921702e-01] | [9.837955474853516, 9.620129585266113] |
72c91543-ffc7-4117-bfa3-14491492eb21 | nonnegative-matrix-factorization-with | 1705.04193 | null | http://arxiv.org/abs/1705.04193v2 | http://arxiv.org/pdf/1705.04193v2.pdf | Nonnegative Matrix Factorization with Transform Learning | Traditional NMF-based signal decomposition relies on the factorization of
spectral data, which is typically computed by means of short-time frequency
transform. In this paper we propose to relax the choice of a pre-fixed
transform and learn a short-time orthogonal transform together with the
factorization. To this end, we formulate a regularized optimization problem
reminiscent of conventional NMF, yet with the transform as additional unknown
parameters, and design a novel block-descent algorithm enabling to find
stationary points of this objective function. The proposed joint transform
learning and factorization approach is tested for two audio signal processing
experiments, illustrating its conceptual and practical benefits. | ['Cédric Févotte', 'Dylan Fagot', 'Herwig Wendt'] | 2017-05-11 | null | null | null | null | ['audio-signal-processing'] | ['audio'] | [ 3.08296442e-01 -1.28179535e-01 1.57304779e-01 -1.29900917e-01
-9.20753002e-01 -6.25369728e-01 3.97053480e-01 7.17703328e-02
-6.41381443e-01 6.31990075e-01 3.38475794e-01 -2.38979891e-01
-5.08928418e-01 -2.38124996e-01 -6.41457200e-01 -8.72071385e-01
-9.67263803e-02 3.09456773e-02 -3.68429214e-01 -2.40657419e-01
-7.68511072e-02 2.14045689e-01 -1.21844411e+00 5.71761131e-02
8.40560794e-01 7.45854437e-01 1.49802104e-01 9.36567605e-01
3.60726178e-01 1.11736394e-01 -5.50031602e-01 -1.11671507e-01
5.54344416e-01 -4.80725884e-01 -5.99121809e-01 3.98047417e-01
3.06931317e-01 3.81130129e-02 -4.84649874e-02 1.01384699e+00
5.14805734e-01 6.44567192e-01 3.85308832e-01 -9.19016063e-01
-2.74396688e-01 4.70611483e-01 -4.73063022e-01 1.39150858e-01
4.65669185e-01 -2.30713308e-01 1.36992562e+00 -1.02827382e+00
5.68860769e-01 8.62857163e-01 9.26583111e-01 7.44027942e-02
-1.76376593e+00 -2.54532754e-01 1.71380669e-01 1.41383916e-01
-1.39279008e+00 -5.81907928e-01 1.07720995e+00 -5.12687683e-01
8.19762170e-01 2.47196406e-01 8.67087960e-01 8.39632094e-01
-6.66198060e-02 5.54263413e-01 1.03568351e+00 -7.17732131e-01
1.86749265e-01 -2.87269354e-01 1.02050252e-01 4.47731137e-01
-1.34157121e-01 -2.22117547e-02 -7.51880944e-01 -3.11612874e-01
4.83199596e-01 -1.22220889e-01 -6.51577652e-01 -5.16073346e-01
-1.25516009e+00 8.14490795e-01 1.67669982e-01 4.88163650e-01
-6.65150523e-01 3.04104269e-01 4.35525924e-01 5.84771156e-01
7.12868810e-01 3.62252504e-01 -4.88163620e-01 -2.21297100e-01
-1.34925234e+00 3.09801668e-01 7.47240961e-01 3.53174448e-01
7.52015114e-01 4.05017048e-01 -1.07223913e-02 6.98565781e-01
2.04494387e-01 3.35395008e-01 5.84401071e-01 -6.95161939e-01
3.02664489e-01 -2.97506042e-02 1.82530150e-01 -9.59492922e-01
-3.64171773e-01 -9.13890064e-01 -6.18397892e-01 -3.47447785e-04
5.46473563e-01 -3.29794854e-01 -6.67689979e-01 1.95128942e+00
4.87443149e-01 6.64202631e-01 -5.08246711e-03 9.53453183e-01
-5.49910180e-02 6.46160781e-01 -2.73168832e-01 -7.19944656e-01
1.00864506e+00 -6.50901437e-01 -1.00433576e+00 2.19768167e-01
4.26227689e-01 -1.03604150e+00 1.09507751e+00 8.49554896e-01
-1.15620816e+00 -6.55418217e-01 -1.08858967e+00 -1.49579847e-03
3.27523462e-02 5.00431299e-01 3.82343650e-01 6.26034319e-01
-1.00923681e+00 7.75119126e-01 -6.67809546e-01 1.43381476e-01
-2.26657599e-01 4.93397295e-01 -3.74258041e-01 1.54260740e-01
-1.22569203e+00 4.48631197e-01 3.58586401e-01 4.17866141e-01
-6.66331828e-01 -9.68828082e-01 -6.54046476e-01 -2.97692735e-02
3.35962743e-01 -9.19890404e-01 1.11634803e+00 -9.87438858e-01
-1.73005939e+00 5.23892701e-01 -4.45180237e-01 -6.50775611e-01
5.67512095e-01 -6.79316878e-01 -4.31829810e-01 2.61980534e-01
1.77819200e-03 -4.04723845e-02 1.90771294e+00 -9.32847857e-01
-3.24875623e-01 -1.29244745e-01 1.17416307e-01 9.00827423e-02
-6.36109114e-01 -2.72635728e-01 -7.21138269e-02 -1.14526439e+00
2.55305409e-01 -8.49014699e-01 -3.64966840e-01 -4.45329875e-01
-2.14214623e-01 1.21910714e-01 5.71124375e-01 -1.06266546e+00
1.51747310e+00 -2.28255558e+00 7.35629499e-01 3.91368210e-01
2.09812880e-01 4.84228656e-02 -1.33974224e-01 6.09553337e-01
-5.75364351e-01 -4.23650414e-01 -2.26491570e-01 -7.16029882e-01
-2.82149240e-02 9.95725244e-02 -4.92257267e-01 8.35398734e-01
2.35478003e-02 4.94011521e-01 -9.30151463e-01 5.13533168e-02
3.90550584e-01 7.25265503e-01 -8.01347256e-01 -1.92188956e-02
-8.03174451e-02 6.51935339e-01 -1.84147611e-01 6.45934567e-02
5.35378456e-01 1.81228518e-01 2.46258378e-01 -6.34367824e-01
-3.55713397e-01 4.03576791e-01 -1.53636491e+00 2.10103393e+00
-6.96732163e-01 5.10097325e-01 5.60565114e-01 -1.48886263e+00
7.91321635e-01 5.83501875e-01 1.07431066e+00 -1.97526231e-01
1.24084353e-01 3.34952593e-01 -1.55349374e-01 -3.73924136e-01
3.30559671e-01 -4.69227970e-01 1.84165776e-01 5.12005746e-01
3.24093789e-01 4.80904505e-02 3.49842846e-01 -8.78771394e-02
9.23750401e-01 2.78613418e-01 3.33028257e-01 -5.12431383e-01
9.66314137e-01 -4.58858430e-01 3.33776742e-01 3.62731695e-01
3.29244256e-01 5.38215280e-01 2.13183597e-01 -2.63543069e-01
-8.47929358e-01 -8.42229545e-01 -8.67901742e-02 1.11066830e+00
-4.47695971e-01 -9.13952053e-01 -8.64747047e-01 -4.81110245e-01
2.44985591e-03 3.60922486e-01 -5.32165468e-01 -8.78897011e-02
-8.11157703e-01 -5.22707224e-01 2.12405711e-01 1.17995098e-01
-1.05924830e-01 -3.23993921e-01 -4.85732585e-01 5.14909267e-01
-4.05146658e-01 -1.00880075e+00 -7.29271173e-01 6.43156111e-01
-9.73804772e-01 -7.63229370e-01 -7.67577648e-01 -4.28886771e-01
5.34622014e-01 3.53181988e-01 9.03346300e-01 -1.53142244e-01
-1.69583693e-01 5.55571675e-01 -4.65537906e-01 3.64243425e-02
-1.10759243e-01 1.42108560e-01 2.30913609e-01 6.56177640e-01
-1.00616224e-01 -1.23485970e+00 -3.77734154e-01 -5.35483211e-02
-1.09252059e+00 -9.36828181e-02 4.43195730e-01 1.12403893e+00
7.08295763e-01 2.46876881e-01 5.83453357e-01 -4.54944909e-01
1.02426147e+00 -2.52606720e-01 -5.54909706e-01 4.00634520e-02
-4.35604990e-01 2.99495518e-01 8.58343303e-01 -6.79582059e-01
-7.13610232e-01 3.89696836e-01 -2.24144116e-01 -6.37454689e-01
3.25280696e-01 8.71603429e-01 -1.18342705e-01 -4.96969819e-02
6.04236066e-01 2.79206425e-01 -1.99589521e-01 -9.04529572e-01
7.17165411e-01 2.48780683e-01 6.37342334e-01 -7.50343084e-01
1.17954874e+00 5.08387506e-01 7.07210675e-02 -1.04622030e+00
-8.66847813e-01 -7.68807888e-01 -7.88431108e-01 -3.10813606e-01
5.43485165e-01 -8.08534086e-01 -6.38952851e-01 -3.99829857e-02
-1.10393119e+00 -1.05558060e-01 -6.78794026e-01 9.19418216e-01
-7.46490717e-01 4.87139523e-01 -3.84795457e-01 -7.88583636e-01
-3.72598082e-01 -7.30132639e-01 1.05631709e+00 -4.67688292e-01
-3.47758919e-01 -1.10423613e+00 3.56455058e-01 4.64021638e-02
8.99066180e-02 1.67681977e-01 5.39251864e-01 -1.21920101e-01
-1.05585933e-01 -2.08228216e-01 2.33211547e-01 5.67382038e-01
3.89209628e-01 -2.60583937e-01 -1.02782786e+00 -6.32734895e-01
5.78922927e-01 1.14362940e-01 9.67847288e-01 7.49930739e-01
6.59077883e-01 -1.60169438e-01 6.61344156e-02 1.08432972e+00
1.36303246e+00 -1.01406977e-01 7.45152235e-02 5.43199442e-02
5.53827584e-01 3.96410316e-01 5.13803899e-01 7.20112026e-01
-1.80508077e-01 7.40239143e-01 5.93684241e-02 -8.81573465e-03
-9.26836431e-02 -2.29008168e-01 5.31878173e-01 1.24778843e+00
-3.38485122e-01 2.09334269e-01 -5.94126046e-01 4.88032579e-01
-1.99294317e+00 -7.59586155e-01 2.54325606e-02 2.29962873e+00
7.76311755e-01 -1.84498280e-02 2.71238804e-01 8.58411193e-01
3.91480267e-01 2.09047586e-01 -3.59939426e-01 -1.08249404e-01
-3.66304792e-03 6.35053098e-01 3.42872530e-01 8.94047141e-01
-1.17691839e+00 4.91601199e-01 6.83202267e+00 8.37391138e-01
-1.39366782e+00 3.54556918e-01 -1.34138599e-01 -1.51517406e-01
-3.82937461e-01 7.20145330e-02 -4.53714520e-01 1.38163894e-01
8.48960638e-01 -3.09637964e-01 8.58272612e-01 3.77597988e-01
6.15706801e-01 2.51786292e-01 -1.03139150e+00 1.11802089e+00
-2.40622554e-02 -9.14379179e-01 -2.30716914e-01 -5.92845045e-02
5.16890943e-01 -4.00318742e-01 1.50968611e-01 -7.92413652e-02
-3.00418288e-01 -7.95427740e-01 9.75937247e-01 3.69539887e-01
5.75404584e-01 -7.12910056e-01 1.76253185e-01 1.92130253e-01
-1.46582425e+00 -1.69873253e-01 -4.28037085e-02 -3.47061872e-01
3.89172792e-01 9.96576607e-01 -6.20512605e-01 8.02756310e-01
2.73911178e-01 9.09954548e-01 2.11934857e-02 7.96524704e-01
-3.63702625e-01 8.09218168e-01 -5.80557942e-01 5.60630143e-01
1.81988418e-01 -6.68564558e-01 9.38102901e-01 1.35908186e+00
5.54923058e-01 -2.00932488e-01 3.29676569e-01 6.14526391e-01
1.05636649e-01 2.77067333e-01 -1.67294070e-01 -1.79662257e-01
-8.55380818e-02 1.26105070e+00 -7.66200483e-01 1.01265766e-01
-4.48174298e-01 1.01549399e+00 1.12734228e-01 6.85236275e-01
-7.52239048e-01 -1.41251490e-01 8.03388417e-01 4.35405932e-02
5.68509817e-01 -7.03001380e-01 6.56341240e-02 -1.47087383e+00
4.13665175e-01 -9.85789478e-01 2.49652505e-01 -1.90214857e-01
-8.04068565e-01 4.92625028e-01 -1.81898832e-01 -1.44916832e+00
-5.92343569e-01 -3.71640652e-01 -3.62266749e-01 1.04240656e+00
-1.53574920e+00 -1.19192290e+00 2.27375761e-01 1.04166317e+00
4.32242870e-01 6.44457862e-02 8.01899612e-01 6.54300988e-01
-3.66419524e-01 4.02689427e-01 3.64670098e-01 -2.84532815e-01
7.26065040e-01 -1.38515842e+00 1.97884977e-01 1.02790606e+00
5.58055162e-01 9.64146912e-01 1.17003155e+00 -4.27237749e-01
-1.58361065e+00 -6.80362165e-01 7.91297436e-01 2.27995664e-02
9.89368916e-01 -4.93281543e-01 -6.45778477e-01 5.10777414e-01
1.26468346e-01 1.09876648e-01 7.53292739e-01 2.90758461e-01
-2.61196315e-01 -2.44335264e-01 -7.02071309e-01 3.06896448e-01
8.08530152e-01 -8.78160238e-01 -5.60209036e-01 2.49381050e-01
6.77968502e-01 -1.96864486e-01 -9.06472802e-01 1.34015337e-01
6.17668808e-01 -6.65222287e-01 1.24972582e+00 -6.83910072e-01
-7.70278126e-02 -5.91203094e-01 -3.05935264e-01 -1.66453397e+00
-3.76787156e-01 -1.41474903e+00 -3.65142465e-01 1.10099161e+00
1.22181244e-01 -5.22656083e-01 6.30814910e-01 -1.33069351e-01
-1.97698012e-01 -6.25375032e-01 -1.06460011e+00 -7.65506089e-01
-2.54164487e-01 -7.26127148e-01 1.72806546e-01 8.33347976e-01
4.82596904e-02 5.70392072e-01 -8.23817790e-01 2.33630046e-01
6.60110950e-01 2.41510093e-01 6.78550482e-01 -1.15851033e+00
-9.49939787e-01 -1.62613899e-01 -1.46253288e-01 -1.27526224e+00
1.88542947e-01 -6.89523160e-01 -3.71316895e-02 -9.09747541e-01
-4.44771022e-01 6.17186585e-03 -4.84227568e-01 -4.98111434e-02
-1.50666073e-01 1.56333163e-01 2.99874246e-01 9.22981203e-02
-1.00485452e-01 6.79252386e-01 1.04548943e+00 -1.05804503e-01
-4.75053370e-01 1.34557366e-01 -4.29586917e-01 5.56689382e-01
4.89576548e-01 -4.08866107e-01 -7.00209260e-01 -3.12564850e-01
5.33744991e-01 1.41392097e-01 7.81211108e-02 -8.97025943e-01
1.29864693e-01 1.40362695e-01 2.08020017e-01 -3.82040173e-01
6.82603896e-01 -9.39696074e-01 2.77175486e-01 3.83649230e-01
-2.85709649e-01 -3.07070650e-02 1.16618946e-01 6.21666551e-01
-5.42227745e-01 -3.25107783e-01 4.66745913e-01 2.50297487e-01
-2.52875268e-01 -4.83474210e-02 -3.22052360e-01 -1.94509462e-01
5.96774638e-01 -2.37303168e-01 5.75212538e-01 -4.53728557e-01
-1.27382815e+00 -2.54547715e-01 8.33343193e-02 7.43248016e-02
5.35240769e-01 -1.37520289e+00 -7.15623677e-01 4.11846131e-01
-4.18585509e-01 -4.54676956e-01 2.46328115e-01 1.23864794e+00
-1.08183071e-01 3.57802004e-01 -2.18525734e-02 -4.78900790e-01
-1.35918200e+00 6.08665347e-01 1.89733922e-01 -5.16352296e-01
-6.08292520e-01 7.09672391e-01 -4.24370170e-02 -1.92979619e-01
5.13911396e-02 -5.64948738e-01 1.05588511e-02 4.70753014e-01
3.90236348e-01 3.42383444e-01 3.33354592e-01 -7.33517110e-01
-5.96952736e-02 8.44056964e-01 3.91965806e-01 -5.67697644e-01
1.51045299e+00 -3.27341348e-01 -1.66300178e-01 5.31804860e-01
1.44252694e+00 4.64129239e-01 -1.02135098e+00 -3.31869751e-01
7.55619034e-02 -5.22443354e-01 3.44587415e-01 -7.95470998e-02
-9.72137809e-01 6.98628128e-01 4.64258015e-01 2.76890069e-01
1.51315057e+00 -6.06234372e-01 9.02051866e-01 4.03842270e-01
2.65407264e-01 -9.52611744e-01 6.56817257e-02 4.43997890e-01
9.47634399e-01 -6.42676115e-01 7.72754624e-02 -3.67173523e-01
2.08112106e-01 1.42851770e+00 -3.07058543e-01 -3.85123491e-01
8.85434628e-01 3.64509642e-01 -7.41578490e-02 1.27904534e-01
-5.16244292e-01 -2.75883377e-01 5.52740335e-01 4.51650977e-01
6.84072614e-01 3.88530232e-02 -6.97272301e-01 4.65847760e-01
-4.51461345e-01 1.61356181e-02 3.54810134e-02 7.68926501e-01
-1.16213225e-01 -1.55049324e+00 -5.96391380e-01 1.13966435e-01
-7.35747755e-01 -5.21861799e-02 -1.48327604e-01 3.89477909e-01
1.67910501e-01 9.20601606e-01 -4.21858341e-01 -4.33548868e-01
3.80362540e-01 2.87490427e-01 5.57474792e-01 -3.96037638e-01
-6.78960741e-01 7.48650491e-01 -1.00211538e-01 -5.16693473e-01
-6.31774485e-01 -9.26072240e-01 -8.13580751e-01 6.18169419e-02
-5.70435524e-01 4.66229707e-01 6.48922026e-01 8.65911067e-01
-7.45287091e-02 7.35564232e-01 7.08697975e-01 -1.12483191e+00
-7.96664298e-01 -7.66654193e-01 -6.47875786e-01 5.93116999e-01
9.21464026e-01 -5.49429655e-01 -3.34845871e-01 3.21041942e-01] | [15.463371276855469, 5.564912796020508] |
61052d57-bc9d-4ebe-a1c0-f974d08b8d61 | speech-driven-talking-face-generation-from-a | 2008.03592 | null | https://arxiv.org/abs/2008.03592v2 | https://arxiv.org/pdf/2008.03592v2.pdf | Speech Driven Talking Face Generation from a Single Image and an Emotion Condition | Visual emotion expression plays an important role in audiovisual speech communication. In this work, we propose a novel approach to rendering visual emotion expression in speech-driven talking face generation. Specifically, we design an end-to-end talking face generation system that takes a speech utterance, a single face image, and a categorical emotion label as input to render a talking face video synchronized with the speech and expressing the conditioned emotion. Objective evaluation on image quality, audiovisual synchronization, and visual emotion expression shows that the proposed system outperforms a state-of-the-art baseline system. Subjective evaluation of visual emotion expression and video realness also demonstrates the superiority of the proposed system. Furthermore, we conduct a human emotion recognition pilot study using generated videos with mismatched emotions among the audio and visual modalities. Results show that humans respond to the visual modality more significantly than the audio modality on this task. | ['Sefik Emre Eskimez', 'You Zhang', 'Zhiyao Duan'] | 2020-08-08 | null | null | null | null | ['talking-face-generation'] | ['computer-vision'] | [ 1.25505194e-01 2.75395423e-01 2.40339756e-01 -6.23730361e-01
-6.94341838e-01 -4.36365664e-01 7.39699602e-01 -6.40822768e-01
-8.17857236e-02 4.98555809e-01 5.00604331e-01 2.37688079e-01
6.47748947e-01 -1.29061863e-01 -6.19092107e-01 -7.19458759e-01
8.43434185e-02 -1.59082457e-01 -4.02073801e-01 -1.55358076e-01
7.41866380e-02 2.94260234e-01 -2.03397083e+00 8.33496630e-01
2.34303847e-01 1.38357031e+00 8.33509937e-02 1.00290287e+00
-8.01429898e-02 1.07451439e+00 -1.07322860e+00 -4.59312737e-01
-1.95686162e-01 -5.55543303e-01 -6.41886234e-01 5.18813491e-01
4.64093745e-01 -3.73690784e-01 -1.44284844e-01 7.21079350e-01
9.95370924e-01 1.38598114e-01 5.44787228e-01 -1.85861647e+00
-5.27313054e-01 1.57108486e-01 -4.23310399e-01 -8.94721076e-02
1.04234874e+00 1.75792068e-01 4.25219297e-01 -1.10244310e+00
9.58325684e-01 1.56262517e+00 1.25564635e-01 9.18981493e-01
-1.09369183e+00 -7.95889497e-01 8.52687061e-02 1.61636829e-01
-1.52672172e+00 -1.21326780e+00 9.62121010e-01 -4.78092909e-01
8.89611423e-01 2.97876805e-01 7.11689472e-01 1.37276828e+00
3.45487781e-02 4.86662775e-01 8.71796668e-01 -5.28094232e-01
3.04235250e-01 5.77453673e-01 -6.03027701e-01 4.95586365e-01
-8.96883488e-01 3.06598008e-01 -1.05529630e+00 -2.24797457e-01
3.48261505e-01 -5.61294019e-01 -4.33998704e-01 8.10823515e-02
-8.95758152e-01 6.10677958e-01 -7.46002123e-02 1.89442322e-01
-6.25825405e-01 3.51538628e-01 6.38807774e-01 3.53425115e-01
6.01121306e-01 -1.93618134e-01 2.98199117e-01 -5.08087218e-01
-1.05676901e+00 -2.43637145e-01 6.06720686e-01 7.47523606e-01
8.59164894e-02 5.50918162e-01 -5.51599205e-01 9.40794349e-01
4.05167878e-01 6.74835026e-01 1.78188741e-01 -1.22219563e+00
-1.17555626e-01 -1.25198722e-01 9.16075781e-02 -1.12705493e+00
-2.75974777e-02 3.81086856e-01 -4.68525410e-01 5.37514567e-01
-2.91639090e-01 -3.79238784e-01 -6.54011369e-01 1.98686695e+00
3.45707774e-01 2.38156095e-01 4.45437580e-01 1.08517253e+00
1.25763690e+00 1.04128194e+00 2.47125760e-01 -8.52644086e-01
1.38396931e+00 -6.06966972e-01 -1.41999936e+00 2.26198673e-01
1.51295125e-01 -9.72192168e-01 1.11992145e+00 5.37238836e-01
-1.27173030e+00 -7.67822444e-01 -8.11726630e-01 2.34911978e-01
2.26261482e-01 4.49746639e-01 3.54193091e-01 9.62123215e-01
-1.47055495e+00 -2.10917443e-01 -1.79131150e-01 -3.05330306e-01
1.57803103e-01 7.65893385e-02 -6.37051165e-01 2.58109301e-01
-9.86365139e-01 5.72253704e-01 -1.09488100e-01 -1.14899129e-01
-1.22541058e+00 -4.23812777e-01 -8.86542261e-01 1.16121516e-01
-3.73585187e-02 -2.94174463e-01 1.56884170e+00 -1.75973403e+00
-2.07277489e+00 1.00762665e+00 -6.23865187e-01 -4.80415858e-03
1.72955886e-01 1.38140365e-01 -8.08195114e-01 9.04732585e-01
-2.60845989e-01 1.15094233e+00 1.25838959e+00 -1.63908577e+00
-3.33812743e-01 -2.71707419e-02 -2.36781985e-01 3.44397396e-01
-3.22977662e-01 5.05470634e-01 -5.83178818e-01 -5.07120728e-01
-4.29788321e-01 -6.76335812e-01 5.68422914e-01 1.78669170e-01
-1.51588723e-01 -1.37017533e-01 1.16099393e+00 -4.95261729e-01
8.72818768e-01 -2.75356340e+00 -1.17867857e-01 1.82083279e-01
-2.97517348e-02 -1.06027918e-02 -2.28714809e-01 3.63135099e-01
-4.33445871e-01 7.50509351e-02 3.51997256e-01 -6.31800354e-01
1.53758954e-02 -1.19045578e-01 -3.53450835e-01 3.74040186e-01
2.27391303e-01 6.30972564e-01 -5.07313907e-01 -9.12685096e-01
1.69397414e-01 1.15931392e+00 -5.25145173e-01 5.54892480e-01
-7.47250319e-02 6.44128501e-01 3.61756682e-02 6.74790859e-01
5.33635199e-01 1.79140717e-01 2.55172104e-01 -3.72644871e-01
1.01998448e-01 -2.21860945e-01 -8.47247362e-01 1.43563008e+00
-5.96764743e-01 1.16016567e+00 4.50903207e-01 -3.81373644e-01
1.03876865e+00 1.10003340e+00 4.49462116e-01 -9.53610480e-01
4.30173308e-01 -3.43811035e-01 -3.53626668e-01 -9.47915733e-01
4.08217728e-01 -4.44257766e-01 1.18206985e-01 3.70983243e-01
1.90093473e-01 -3.03350419e-01 -1.12244627e-02 4.66164678e-01
5.60757697e-01 -1.08838014e-01 -7.22191259e-02 2.18690813e-01
4.94136095e-01 -5.01857698e-01 9.30254310e-02 7.06138387e-02
-3.85730714e-01 4.35529351e-01 7.28146076e-01 1.41478330e-01
-5.70021033e-01 -1.06928492e+00 1.77730232e-01 1.43781614e+00
5.72941639e-02 -5.07961810e-01 -9.54245687e-01 -1.76915944e-01
-5.44876099e-01 8.27706099e-01 -5.40960431e-01 -1.58417761e-01
2.47389041e-02 2.57204443e-01 6.94555581e-01 3.07793766e-01
3.11195076e-01 -1.44129479e+00 -5.73001325e-01 -2.08222896e-01
-6.71672046e-01 -1.37412870e+00 -6.09093904e-01 -6.82336867e-01
-9.54159629e-03 -5.79000473e-01 -8.53511870e-01 -1.01491189e+00
6.42310500e-01 6.64085075e-02 8.18402827e-01 -2.21668035e-01
-3.75828087e-01 9.93947327e-01 -3.59215736e-01 -3.62575799e-01
-5.60963750e-01 -9.25247192e-01 2.11677194e-01 5.61761081e-01
-8.45452771e-02 -2.53209829e-01 -4.98881817e-01 2.76034832e-01
-8.56203139e-01 8.85190349e-03 7.00997859e-02 6.35603428e-01
2.79545158e-01 -1.88860461e-01 6.26611233e-01 -4.58560698e-02
9.99703765e-01 -8.71103853e-02 -1.44289643e-01 1.55341029e-01
6.32191822e-02 -5.36416411e-01 2.69489735e-01 -7.24048913e-01
-1.60155320e+00 2.18244195e-01 -1.54854715e-01 -7.62135386e-01
-3.17755848e-01 1.94107801e-01 -4.12191212e-01 1.47119267e-02
5.31135499e-01 6.41850978e-02 3.94751817e-01 3.09677362e-01
5.85523725e-01 1.24916089e+00 8.52388263e-01 -4.03431147e-01
6.11117780e-02 5.29409409e-01 -4.27251935e-01 -1.32740712e+00
1.49021208e-01 -5.53461835e-02 1.12545140e-01 -1.13221073e+00
9.98887122e-01 -1.24383783e+00 -1.48158348e+00 3.94311368e-01
-1.34669840e+00 -2.53454626e-01 -2.37371363e-02 4.79367107e-01
-8.48650992e-01 2.38443092e-01 -5.39264917e-01 -1.36916173e+00
-4.44831371e-01 -1.36973488e+00 1.37519300e+00 1.76826626e-01
-4.09467727e-01 -4.74748731e-01 -1.10084169e-01 2.99794495e-01
4.57462281e-01 1.78380355e-01 4.74344105e-01 -9.29088071e-02
3.80310044e-03 -3.92638966e-02 -2.40725338e-01 2.76354104e-01
2.59947702e-02 5.31627715e-01 -1.39183795e+00 -1.80919215e-01
-5.65215126e-02 -7.86042988e-01 2.35548556e-01 5.15939355e-01
8.37325871e-01 -2.88016796e-01 -6.89939409e-02 3.51529300e-01
7.95479596e-01 7.02103794e-01 8.32309425e-01 -5.56633234e-01
8.57128799e-02 1.14637017e+00 8.32184613e-01 6.89584196e-01
4.26517762e-02 8.52269828e-01 3.30293477e-01 -3.76369834e-01
-1.16458893e-01 -1.87326774e-01 7.59098887e-01 5.30922174e-01
2.30814025e-01 -6.35364652e-01 -4.94373798e-01 2.62612939e-01
-1.30390882e+00 -1.27061415e+00 2.55526900e-01 1.87858260e+00
6.67039335e-01 -4.18561608e-01 1.90787107e-01 1.07225761e-01
9.19233680e-01 -3.05729453e-02 -1.10896796e-01 -7.84124374e-01
-2.13865563e-01 7.16527700e-02 -4.08414304e-01 5.86286724e-01
-6.54489398e-01 9.82851207e-01 6.84887409e+00 6.56872928e-01
-1.83403850e+00 6.70205578e-02 8.20414186e-01 -5.66308200e-01
-2.25607648e-01 -4.49895442e-01 -1.10547826e-01 3.52441430e-01
1.07926202e+00 -2.92049199e-01 3.49531591e-01 8.46072257e-01
7.47451246e-01 -1.89436838e-01 -1.08064795e+00 1.69231915e+00
6.55272841e-01 -1.15788865e+00 1.16588376e-01 -7.90126324e-02
2.58956313e-01 -6.91080928e-01 4.05934542e-01 1.17757414e-02
-3.82244289e-01 -1.31154370e+00 9.55169499e-01 5.13037860e-01
1.48924637e+00 -9.22007501e-01 4.22343045e-01 -2.29430810e-01
-1.17056477e+00 2.13870570e-01 1.50857821e-01 1.94653496e-01
6.87288344e-01 -3.56411822e-02 -8.83585036e-01 4.09025699e-02
7.87289739e-01 1.02095678e-01 -1.12236537e-01 3.76471519e-01
-1.18290074e-01 5.09636879e-01 -6.36109756e-03 7.91616142e-02
-3.01815748e-01 2.35293478e-01 4.25772458e-01 1.35613692e+00
2.95982003e-01 4.21402067e-01 -1.35531351e-01 5.83087385e-01
-1.35163277e-01 4.28871930e-01 -8.88695359e-01 -4.99486685e-01
3.92167926e-01 1.20825827e+00 -4.78531569e-01 -4.64096248e-01
-2.02819124e-01 1.19838333e+00 -3.19882214e-01 5.19877255e-01
-1.10574329e+00 -3.39609921e-01 6.60484672e-01 -2.34810516e-01
1.73799172e-01 2.05126777e-01 2.68245310e-01 -6.34216785e-01
4.90651876e-02 -1.06206596e+00 -1.26269966e-01 -1.58675873e+00
-7.43320704e-01 1.19067097e+00 -1.95186168e-01 -1.12190223e+00
-6.32222354e-01 -2.53613949e-01 -6.12803698e-01 6.90311790e-01
-9.60307777e-01 -8.87625217e-01 -5.30426741e-01 9.25087333e-01
5.86918771e-01 -2.46052027e-01 8.93766701e-01 3.15174997e-01
-1.96760282e-01 8.78117740e-01 -6.07634485e-01 -1.75128236e-01
1.12030113e+00 -3.45549673e-01 -2.80043870e-01 4.56704438e-01
-4.75337952e-02 2.11574540e-01 8.70723665e-01 -5.06500721e-01
-1.35783792e+00 -7.77827203e-01 6.17617190e-01 1.83377340e-01
7.17806146e-02 -5.12391627e-01 -4.66475338e-01 2.41316080e-01
8.47546756e-01 -1.61442459e-01 9.86248851e-01 -5.25886297e-01
-2.09855050e-01 -8.40134695e-02 -1.49232566e+00 5.97301066e-01
6.55755281e-01 -8.60690773e-01 -1.73788816e-01 1.07914492e-01
8.28234375e-01 -2.00455803e-02 -6.94177032e-01 2.40398705e-01
8.36948335e-01 -9.71915066e-01 7.10440755e-01 -3.29810381e-01
5.18540919e-01 -1.56558603e-01 -4.15538758e-01 -1.25053966e+00
2.99344778e-01 -9.19165432e-01 4.42857325e-01 1.66799724e+00
2.25062981e-01 -3.41170102e-01 4.57147777e-01 5.00343084e-01
1.51744917e-01 -7.89022148e-02 -9.56511080e-01 -4.54949111e-01
-6.25359297e-01 -4.30322289e-01 2.15557098e-01 9.10192490e-01
6.09510541e-01 6.00489080e-01 -7.61374831e-01 9.05675441e-02
1.34865850e-01 -2.57470578e-01 9.59893286e-01 -7.04914689e-01
-7.32433796e-02 -1.29581317e-01 -5.99827111e-01 -6.18433297e-01
4.51145172e-01 -5.17955244e-01 2.79435009e-01 -1.20930910e+00
-1.03087882e-02 6.02671742e-01 2.89064914e-01 2.82854050e-01
2.90864021e-01 3.38105679e-01 3.68955374e-01 -2.99730211e-01
-4.56481904e-01 9.43247914e-01 1.09165812e+00 -5.13273589e-02
-1.82287917e-01 -6.07392967e-01 -4.17056143e-01 4.66434270e-01
5.72844982e-01 -5.21178842e-02 -6.80144727e-01 3.45334373e-02
-1.87603250e-01 8.28741014e-01 3.53432655e-01 -7.52913713e-01
1.51373386e-01 1.95909198e-02 4.25500691e-01 -4.56454933e-01
1.13244641e+00 -7.51021624e-01 2.41566449e-01 1.75848857e-01
-4.72560316e-01 1.14640437e-01 4.76457626e-01 2.87662476e-01
-5.54288089e-01 4.09112990e-01 9.24654007e-01 3.28789502e-01
-4.64711577e-01 -1.05534233e-01 -9.20053363e-01 -8.94210413e-02
1.15117395e+00 -2.87395060e-01 -3.69707286e-01 -1.34725773e+00
-1.09218419e+00 -1.10988401e-01 3.29183698e-01 5.29314041e-01
1.15106523e+00 -1.45985186e+00 -7.11640418e-01 3.96972775e-01
1.95224568e-01 -9.34342980e-01 5.30341089e-01 8.31341922e-01
-1.77776307e-01 7.83832669e-02 -3.81638527e-01 -8.89472604e-01
-2.09163857e+00 6.01035714e-01 1.91999540e-01 6.19022429e-01
-1.83853105e-01 9.80091751e-01 5.09070575e-01 2.50016302e-01
6.24906301e-01 2.68656880e-01 -1.97126716e-01 2.02416450e-01
8.46728981e-01 3.18809375e-02 -1.70475051e-01 -1.17218113e+00
-4.09808218e-01 3.00992668e-01 3.81015778e-01 -8.73851061e-01
7.19860435e-01 -4.08694059e-01 1.90670893e-01 4.31521416e-01
1.34531355e+00 8.66355225e-02 -8.92323554e-01 2.03136742e-01
-7.66901255e-01 -5.47612667e-01 2.13374030e-02 -8.67980123e-01
-1.26703966e+00 1.06621444e+00 1.09582198e+00 1.99452955e-02
1.44523978e+00 9.18273851e-02 3.45649660e-01 -7.95281865e-03
1.46287099e-01 -1.12042975e+00 6.53313875e-01 -2.38282662e-02
1.41258276e+00 -1.10869515e+00 -6.04200959e-01 -6.58428073e-01
-1.37005031e+00 7.97306299e-01 6.19729996e-01 5.45357108e-01
5.39388180e-01 5.37459373e-01 6.87558353e-01 -2.36302644e-01
-1.16634631e+00 -8.75050798e-02 4.46771681e-02 8.95615637e-01
6.40611470e-01 -6.75920844e-02 -5.07737584e-02 3.52752149e-01
-3.67263287e-01 1.43962264e-01 3.39129597e-01 6.69180989e-01
-1.91196233e-01 -4.78522390e-01 -6.71121895e-01 -2.14797720e-01
-5.81836998e-01 5.36236390e-02 -8.41061711e-01 3.90400648e-01
-2.79358983e-01 1.60988367e+00 2.89903134e-01 -5.27771592e-01
3.76017362e-01 4.75769252e-01 4.24023062e-01 -1.10725842e-01
-5.25894523e-01 5.39934039e-01 2.93030500e-01 -7.82611310e-01
-6.02648854e-01 -2.44439155e-01 -1.40812337e+00 -2.37026647e-01
-8.69592186e-03 2.35041738e-01 1.02902198e+00 4.88603741e-01
6.50691450e-01 6.39362574e-01 9.64256823e-01 -1.19482934e+00
1.71070278e-01 -1.04540789e+00 -5.23079455e-01 5.94499230e-01
2.94922084e-01 -5.52632153e-01 -5.34243762e-01 5.70325077e-01] | [13.274232864379883, -0.40248703956604004] |
4d1add68-8d74-4e5d-8ae3-151af7451928 | continual-dialogue-state-tracking-via-example | 2305.13721 | null | https://arxiv.org/abs/2305.13721v1 | https://arxiv.org/pdf/2305.13721v1.pdf | Continual Dialogue State Tracking via Example-Guided Question Answering | Dialogue systems are frequently updated to accommodate new services, but naively updating them by continually training with data for new services in diminishing performance on previously learnt services. Motivated by the insight that dialogue state tracking (DST), a crucial component of dialogue systems that estimates the user's goal as a conversation proceeds, is a simple natural language understanding task, we propose reformulating it as a bundle of granular example-guided question answering tasks to minimize the task shift between services and thus benefit continual learning. Our approach alleviates service-specific memorization and teaches a model to contextualize the given question and example to extract the necessary information from the conversation. We find that a model with just 60M parameters can achieve a significant boost by learning to learn from in-context examples retrieved by a retriever trained to identify turns with similar dialogue state changes. Combining our method with dialogue-level memory replay, our approach attains state of the art performance on DST continual learning metrics without relying on any complex regularization or parameter expansion methods. | ['Chinnadhurai Sankar', 'Jonathan May', 'Jing Xu', 'Satwik Kottur', 'Khyathi Raghavi Chandu', 'Zhaojiang Lin', 'Andrea Madotto', 'Hyundong Cho'] | 2023-05-23 | null | null | null | null | ['memorization', 'dialogue-state-tracking'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.97071695e-01 6.72237933e-01 -3.17101210e-01 -6.86290205e-01
-1.03348768e+00 -6.62735939e-01 1.00592339e+00 1.96025431e-01
-3.66299272e-01 7.99978852e-01 6.72136664e-01 -6.07466936e-01
1.25811964e-01 -1.56273961e-01 -3.71971101e-01 -1.78056359e-01
-2.41067000e-02 1.20375156e+00 3.76500726e-01 -9.79584157e-01
4.61319119e-01 2.52030730e-01 -1.26858234e+00 6.87211037e-01
9.90540445e-01 7.88798571e-01 1.75981671e-01 9.52009320e-01
-7.75244355e-01 1.26275575e+00 -9.02790904e-01 -2.20718652e-01
-1.82349145e-01 -5.83615005e-01 -1.51750088e+00 2.90317297e-01
3.39083582e-01 -6.49605691e-01 -4.50197637e-01 2.63586640e-01
3.46581548e-01 4.52468514e-01 3.42187941e-01 -1.09801841e+00
-1.19933136e-01 6.31428659e-01 2.17863265e-02 3.54457080e-01
9.25542057e-01 4.66789991e-01 8.75747919e-01 -5.81139386e-01
5.01826108e-01 1.45599687e+00 5.08499265e-01 1.12123191e+00
-1.31470358e+00 -1.71118751e-01 5.07417381e-01 1.45803362e-01
-5.17071486e-01 -1.14284658e+00 4.33437079e-01 1.29359677e-01
1.55876386e+00 5.80741644e-01 4.58817542e-01 1.06064212e+00
-2.59982914e-01 1.10179734e+00 7.51870513e-01 -4.09640640e-01
1.27809212e-01 5.66043079e-01 3.00211340e-01 7.23334014e-01
-7.17191994e-01 -5.60275674e-01 -8.01926315e-01 -4.06175733e-01
9.38374996e-02 -1.40439585e-01 -2.76227713e-01 -1.77114412e-01
-7.24984527e-01 6.80638492e-01 -7.42317289e-02 1.90705359e-01
-1.10859752e-01 -2.17346042e-01 6.60532594e-01 8.41079950e-01
3.13820064e-01 8.40224802e-01 -1.00659299e+00 -9.33753014e-01
-4.34588403e-01 2.82937884e-01 1.73344910e+00 1.06664634e+00
7.85045922e-01 -2.52509326e-01 -3.28073233e-01 1.25137985e+00
1.03574153e-02 2.64950812e-01 6.09014988e-01 -1.31008756e+00
4.89591300e-01 7.75295138e-01 3.25362265e-01 -2.19641224e-01
-5.12164295e-01 1.12569243e-01 -2.54926761e-03 -5.22196114e-01
6.71624482e-01 -3.46861631e-01 -5.12274086e-01 1.73374581e+00
3.52616310e-01 -3.50148454e-02 3.03703815e-01 3.41338128e-01
6.31904066e-01 8.59267712e-01 3.62190641e-02 -5.30979037e-01
1.32773328e+00 -1.26090658e+00 -7.07784355e-01 -5.80246449e-01
9.09796834e-01 -6.41227961e-01 1.44605434e+00 2.85671175e-01
-1.11546278e+00 -1.16910882e-01 -7.48438835e-01 -1.66712075e-01
-2.30931506e-01 -6.90284908e-01 7.29587376e-01 5.35390079e-01
-1.14606094e+00 4.30920541e-01 -6.98779583e-01 -8.18143964e-01
-1.27919033e-01 1.73431799e-01 1.26910537e-01 -5.00583202e-02
-1.18620920e+00 1.33279967e+00 -1.10449247e-01 -1.75405785e-01
-8.52698147e-01 -6.33699536e-01 -7.05541790e-01 1.68855280e-01
7.10554361e-01 -3.39304715e-01 2.23863602e+00 -7.25089133e-01
-2.13343096e+00 5.60661376e-01 -5.34446239e-01 -5.33238649e-01
2.26249412e-01 -1.50054216e-01 -2.82665730e-01 5.29445373e-02
-2.47570395e-01 4.79993641e-01 7.66993940e-01 -1.16574419e+00
-8.10948670e-01 -2.06738755e-01 3.89741063e-01 5.78356385e-01
-4.58278000e-01 -1.79005697e-01 -6.21912539e-01 2.71023750e-01
2.38905743e-01 -9.83787835e-01 -8.24671462e-02 -4.21898216e-01
-5.19667082e-02 -6.88429713e-01 9.61792111e-01 -7.65162468e-01
1.24554908e+00 -1.53480542e+00 -2.59096399e-02 2.42622872e-03
8.72920603e-02 3.24068636e-01 -3.16984922e-01 8.52684855e-01
4.49480981e-01 -2.54064441e-01 -2.08909810e-02 -4.43188965e-01
1.58702940e-01 3.60208839e-01 -3.25640768e-01 -1.91388503e-01
4.75117080e-02 7.72269666e-01 -1.14402676e+00 -2.43734792e-01
6.29447848e-02 1.03117317e-01 -6.48192942e-01 7.37097144e-01
-8.31879973e-01 5.25805295e-01 -3.10568273e-01 3.53770047e-01
1.28178507e-01 -5.68865657e-01 6.72895014e-01 8.21502656e-02
1.92111298e-01 1.17676783e+00 -5.68769634e-01 1.89794433e+00
-1.03891695e+00 5.48044622e-01 2.79848844e-01 -8.55245769e-01
7.87945330e-01 4.30021614e-01 3.89467746e-01 -9.66615379e-01
-2.48903826e-01 -6.17047139e-02 -5.09033650e-02 -9.13798809e-01
6.45277917e-01 6.82351226e-03 -2.45837837e-01 1.01106179e+00
2.41494089e-01 -3.03133875e-01 6.70084879e-02 6.27321064e-01
1.34053564e+00 -1.21318042e-01 1.26076311e-01 7.74468407e-02
5.87813079e-01 1.23840593e-01 9.85161513e-02 9.80623007e-01
-2.25073293e-01 -2.58813411e-01 6.89736009e-01 -4.06044155e-01
-7.67550528e-01 -6.50311232e-01 1.17150791e-01 2.02107882e+00
-1.75455362e-02 -4.14537400e-01 -7.86593676e-01 -9.94967937e-01
7.74271488e-02 1.16853189e+00 -1.76836744e-01 -3.94223809e-01
-8.07047427e-01 -5.72766840e-01 2.61497140e-01 2.05473080e-01
5.08024335e-01 -1.21093261e+00 -3.38270068e-01 4.89965498e-01
-5.97376823e-01 -9.64010894e-01 -7.77280569e-01 2.74012446e-01
-9.04912591e-01 -9.28872645e-01 -3.81712377e-01 -7.70138562e-01
4.25221533e-01 3.26397359e-01 1.52933836e+00 2.33360007e-01
1.18754037e-01 9.63517725e-01 -2.54545122e-01 9.62244943e-02
-1.15329671e+00 4.96878445e-01 -1.45635279e-02 -1.80243522e-01
6.98858380e-01 -6.44276381e-01 -5.67270279e-01 5.05694926e-01
-4.47418422e-01 -7.27496669e-03 3.38535041e-01 9.48036075e-01
-4.24337775e-01 -6.08208954e-01 1.16385472e+00 -1.23328936e+00
1.32491064e+00 -5.08679628e-01 -2.25048661e-01 5.72249949e-01
-1.08412802e+00 4.52627331e-01 3.10425818e-01 -3.97024810e-01
-1.50350261e+00 -2.61494279e-01 -1.96596026e-01 4.20786917e-01
-3.29665728e-02 2.13382527e-01 1.33171156e-01 8.03204998e-02
9.48487639e-01 5.25891721e-01 3.63273621e-01 -5.66498458e-01
5.74099004e-01 1.05924284e+00 3.54043245e-01 -9.36429560e-01
1.78576753e-01 8.13685507e-02 -8.90562952e-01 -7.58918047e-01
-9.89422262e-01 -8.13693285e-01 -3.33151430e-01 -3.28752577e-01
3.07017177e-01 -5.66598594e-01 -1.33586109e+00 2.05159053e-01
-1.09994447e+00 -8.52494717e-01 -2.63620824e-01 -2.02783957e-01
-7.22100019e-01 4.84620601e-01 -8.99553299e-01 -1.04533470e+00
-5.16243696e-01 -8.96399319e-01 8.47193837e-01 2.99076259e-01
-6.50919378e-01 -9.79024172e-01 1.51454151e-01 6.66212320e-01
8.81329000e-01 -6.60560846e-01 1.08670866e+00 -1.19002664e+00
-5.42489290e-01 -2.93664575e-01 -2.39634607e-02 2.72730857e-01
3.99992019e-01 -6.78173602e-01 -1.10569882e+00 -5.08285403e-01
4.12797093e-01 -8.17501903e-01 4.84685272e-01 -2.83542663e-01
7.79844761e-01 -8.40373039e-01 -1.20509684e-01 -7.29413480e-02
6.80101991e-01 4.13995028e-01 2.35591650e-01 2.58675754e-01
2.34283388e-01 9.37894821e-01 4.87279564e-01 5.94864666e-01
8.43140244e-01 6.80324733e-01 3.95924971e-02 4.47361648e-01
-8.14322457e-02 -2.89295912e-01 5.02581060e-01 9.00936007e-01
4.30182844e-01 -1.04211003e-01 -9.15793061e-01 3.88332427e-01
-1.92152572e+00 -7.56694436e-01 6.22276008e-01 2.10056329e+00
1.50770557e+00 4.83748078e-01 2.47096032e-01 -4.25033003e-01
3.08237016e-01 3.36587518e-01 -1.12265408e+00 -6.61178350e-01
4.10086036e-01 -2.02064719e-02 6.19571917e-02 1.03263867e+00
-4.58847523e-01 1.10912693e+00 6.84703159e+00 2.93327808e-01
-8.56130898e-01 4.84230742e-02 5.11281312e-01 -9.16810613e-03
-3.10284197e-01 5.99538535e-02 -8.37294579e-01 2.82411754e-01
1.51696515e+00 -2.61089712e-01 9.21372235e-01 7.63965070e-01
2.26459831e-01 -2.80714869e-01 -1.46794796e+00 5.15970111e-01
1.55718997e-01 -1.28233302e+00 7.64132142e-02 -4.59835023e-01
3.65763307e-01 2.92397896e-03 -1.60824940e-01 1.00498700e+00
5.72281599e-01 -7.44783878e-01 -1.20698601e-01 5.27169168e-01
3.28305513e-01 -4.12931025e-01 4.22704786e-01 7.23165095e-01
-5.03860891e-01 -2.43293360e-01 5.44872284e-02 -8.54471698e-03
2.02001631e-01 -2.78417259e-01 -1.89478409e+00 -3.19015950e-01
2.53648579e-01 2.91811317e-01 -4.29476917e-01 5.49865663e-01
4.57377806e-02 7.11035550e-01 -3.09503019e-01 -4.26042110e-01
1.87198326e-01 4.24351953e-02 4.89326954e-01 1.25218689e+00
-2.80047953e-01 7.67409876e-02 2.86096305e-01 2.61337757e-01
-1.32269889e-01 3.52495424e-02 -3.76156628e-01 1.98291540e-02
6.93192661e-01 1.06998074e+00 -1.30583182e-01 -7.20998883e-01
-3.85792077e-01 1.10751176e+00 3.41360867e-01 4.23352778e-01
-2.69965500e-01 -7.55626187e-02 6.32807732e-01 1.75703034e-01
-5.73717915e-02 -4.61794026e-02 4.65850905e-02 -1.03550625e+00
-8.65652412e-02 -1.34727669e+00 4.07623976e-01 -6.29180014e-01
-9.94215965e-01 5.35312474e-01 -1.60349727e-01 -7.38285184e-01
-1.02093661e+00 -1.87960878e-01 -4.98205692e-01 7.60937572e-01
-1.58277786e+00 -6.74779594e-01 -2.22885817e-01 5.60814977e-01
1.23077798e+00 -2.35767260e-01 1.20204985e+00 3.25339474e-02
-3.38767022e-01 8.07537258e-01 1.14614926e-01 -1.83766410e-01
9.68511403e-01 -1.44091797e+00 5.19591570e-01 1.80359818e-02
-2.36027509e-01 8.03466082e-01 9.70001340e-01 -4.16551352e-01
-1.64158976e+00 -3.85615826e-01 1.10952735e+00 -9.35663521e-01
8.30318630e-01 -4.55102742e-01 -1.33391702e+00 8.96391094e-01
4.78907079e-01 -6.30434573e-01 6.71962917e-01 6.49485528e-01
-4.93291080e-01 -2.29230866e-01 -1.13100350e+00 7.05787420e-01
8.71753573e-01 -9.36858892e-01 -8.44129860e-01 7.14243650e-01
1.07179320e+00 -6.69099808e-01 -8.61194730e-01 -4.93311509e-02
4.49255854e-01 -7.12755382e-01 6.53466880e-01 -1.08919156e+00
-1.71862200e-01 3.96691859e-01 1.47310466e-01 -1.22797060e+00
2.66189694e-01 -1.37884140e+00 -6.79908514e-01 9.64427352e-01
7.11304784e-01 -6.55052304e-01 9.59238350e-01 1.34314346e+00
-2.29882583e-01 -7.32231677e-01 -7.42794394e-01 -3.12763780e-01
-2.30812326e-01 3.39827947e-02 6.27132773e-01 7.75498927e-01
8.46858025e-01 9.95143950e-01 -3.64905864e-01 -1.67727172e-01
1.60792634e-01 -4.07352261e-02 1.00731635e+00 -1.03403568e+00
-3.41079146e-01 -3.05077732e-01 4.57388252e-01 -1.92803931e+00
1.47419482e-01 -5.98553896e-01 4.19850051e-01 -1.29084146e+00
-4.62176576e-02 -4.31602806e-01 1.46858683e-02 4.78272647e-01
-8.74992758e-02 -6.81692839e-01 -6.70712888e-02 2.80986041e-01
-1.22318518e+00 4.54325557e-01 1.24239051e+00 -1.57007501e-01
-6.71623051e-01 5.79000473e-01 -9.34885621e-01 4.46605414e-01
5.88737249e-01 -2.08409682e-01 -7.77166724e-01 -1.14606686e-01
4.57013734e-02 7.30176210e-01 -2.64798850e-01 -5.18079340e-01
6.52762055e-01 -1.12170078e-01 -2.78520018e-01 -3.55224758e-01
7.42067873e-01 -6.44350767e-01 -6.51563585e-01 2.44141370e-01
-1.09485137e+00 2.33012829e-02 3.58589292e-01 6.15704954e-01
8.47984701e-02 -4.13775682e-01 4.08590376e-01 -4.39470470e-01
-7.95049727e-01 -6.79759160e-02 -8.08021426e-01 2.93757021e-01
2.96788067e-01 -2.92352177e-02 -5.66669762e-01 -1.19741964e+00
-9.85660672e-01 8.21393132e-01 1.27738193e-01 6.21402562e-01
3.23685378e-01 -7.62232900e-01 -4.90816861e-01 -2.24563088e-02
2.39173546e-01 -1.19934946e-01 2.38424793e-01 5.19820511e-01
-1.78621337e-02 6.82264328e-01 1.50524497e-01 -6.29883528e-01
-1.07299149e+00 1.16418429e-01 6.29101336e-01 -5.04486918e-01
-5.77889979e-01 7.98445344e-01 -3.65850508e-01 -9.41618562e-01
7.72291362e-01 -1.61703169e-01 -3.03168714e-01 2.24246219e-01
6.40041173e-01 2.08019674e-01 3.99005085e-01 1.21697918e-01
7.40133077e-02 -3.61831158e-01 -9.18012559e-01 -3.38603020e-01
1.23341691e+00 -6.19465709e-01 -1.18915692e-01 7.39519477e-01
1.14656687e+00 -3.24332744e-01 -1.58014095e+00 -8.42122436e-01
5.94951808e-01 -1.60193369e-01 -4.09111530e-01 -1.40086758e+00
-2.33976677e-01 5.10310829e-01 4.21210557e-01 7.77501106e-01
7.32861340e-01 1.86673790e-01 1.13040257e+00 1.32786179e+00
2.66919106e-01 -1.38165355e+00 6.75524592e-01 9.13245320e-01
9.12238598e-01 -1.45197439e+00 -2.46368945e-01 1.17395237e-01
-7.58726120e-01 1.00886202e+00 9.24808979e-01 4.44910049e-01
4.37129527e-01 1.78190749e-02 3.28670114e-01 -2.79883653e-01
-1.43730104e+00 8.98172557e-02 -1.04488194e-01 5.23406029e-01
4.01957870e-01 -3.45115185e-01 9.44920704e-02 8.49165469e-02
-9.19617414e-02 -3.28480363e-01 5.53758740e-01 1.13877881e+00
-8.35954130e-01 -1.12220383e+00 1.85557216e-01 7.07356751e-01
1.61481470e-01 -1.90161079e-01 -4.24968570e-01 5.74259579e-01
-1.02354956e+00 1.09334362e+00 5.86690418e-02 -2.29679480e-01
6.17470086e-01 9.50438321e-01 3.70631218e-01 -9.26905215e-01
-8.18077326e-01 -1.47378430e-01 7.49489307e-01 -6.40667260e-01
-1.24904349e-01 -7.88886487e-01 -1.22356570e+00 -3.84593517e-01
-2.17884183e-01 6.05723083e-01 6.34643495e-01 1.12821460e+00
4.78275537e-01 3.75493139e-01 9.35881257e-01 -4.92266744e-01
-1.13713193e+00 -1.36556101e+00 6.33468106e-02 3.33596766e-01
5.84772348e-01 -2.05425516e-01 -4.64311510e-01 -2.45095819e-01] | [12.863096237182617, 7.931657791137695] |
37299ea0-4f90-4c13-9fc6-8add64c52f74 | formnet-structural-encoding-beyond-sequential | 2203.08411 | null | https://arxiv.org/abs/2203.08411v2 | https://arxiv.org/pdf/2203.08411v2.pdf | FormNet: Structural Encoding beyond Sequential Modeling in Form Document Information Extraction | Sequence modeling has demonstrated state-of-the-art performance on natural language and document understanding tasks. However, it is challenging to correctly serialize tokens in form-like documents in practice due to their variety of layout patterns. We propose FormNet, a structure-aware sequence model to mitigate the suboptimal serialization of forms. First, we design Rich Attention that leverages the spatial relationship between tokens in a form for more precise attention score calculation. Second, we construct Super-Tokens for each word by embedding representations from their neighboring tokens through graph convolutions. FormNet therefore explicitly recovers local syntactic information that may have been lost during serialization. In experiments, FormNet outperforms existing methods with a more compact model size and less pre-training data, establishing new state-of-the-art performance on CORD, FUNSD and Payment benchmarks. | ['Tomas Pfister', 'Yasuhisa Fujii', 'Renshen Wang', 'Joshua Ainslie', 'Nan Hua', 'Guolong Su', 'Vincent Perot', 'Timothy Dozat', 'Chun-Liang Li', 'Chen-Yu Lee'] | 2022-03-16 | null | https://aclanthology.org/2022.acl-long.260 | https://aclanthology.org/2022.acl-long.260.pdf | acl-2022-5 | ['document-ai'] | ['natural-language-processing'] | [ 2.15392843e-01 -2.38465250e-01 -2.37743050e-01 -2.58049935e-01
-5.66882908e-01 -9.37833726e-01 6.56149328e-01 5.77658653e-01
-4.26431626e-01 4.71120149e-01 6.80032730e-01 -7.02500999e-01
5.71104623e-02 -7.18606412e-01 -7.91107357e-01 -8.92617777e-02
3.78015311e-03 3.04783911e-01 -5.78645170e-02 -1.77509129e-01
5.26603460e-01 4.43125308e-01 -1.01260221e+00 8.24984908e-01
9.27313030e-01 5.63068271e-01 5.94776690e-01 8.86655688e-01
-8.94046366e-01 8.01625192e-01 -7.70013571e-01 -5.63357770e-01
2.05957294e-01 -1.46858439e-01 -1.05261576e+00 -9.58101004e-02
8.63997102e-01 -6.80437982e-01 -7.16239750e-01 8.47336948e-01
3.79772842e-01 1.24810636e-01 5.45373857e-01 -6.20374441e-01
-1.32694924e+00 1.13584638e+00 -8.16421151e-01 2.54230082e-01
4.10567850e-01 3.62523228e-01 1.61573982e+00 -7.81713188e-01
7.54932225e-01 1.31729281e+00 6.33383036e-01 5.33786774e-01
-1.30591011e+00 -4.59182829e-01 4.45585161e-01 3.79407257e-01
-1.18181729e+00 -2.31781989e-01 4.28139031e-01 -1.93295509e-01
1.73315620e+00 5.33265360e-02 6.82699084e-01 9.09202993e-01
1.10061087e-01 1.31854784e+00 4.38028127e-01 -5.33149123e-01
-1.05287030e-01 -4.08400327e-01 6.80664122e-01 8.80735993e-01
4.53028142e-01 -3.83731693e-01 -4.64844108e-01 1.43059874e-02
6.20711505e-01 4.72698659e-02 -3.88430711e-03 -3.39947231e-02
-9.31640387e-01 6.44943118e-01 5.35397291e-01 4.46425498e-01
-3.62410903e-01 6.80108488e-01 6.15747750e-01 1.92883611e-01
1.83351934e-01 7.80165911e-01 -4.10349786e-01 -4.89585161e-01
-1.04615879e+00 4.93094742e-01 6.39721811e-01 1.22347128e+00
7.92228818e-01 8.09189454e-02 -6.42920256e-01 7.15551138e-01
7.47652575e-02 2.31364474e-01 4.95205075e-01 -4.62487459e-01
9.25459325e-01 6.94120109e-01 -9.50575694e-02 -6.33048356e-01
-1.90798908e-01 -5.44661701e-01 -5.52263379e-01 -3.06590319e-01
5.60562968e-01 5.46852201e-02 -1.18283021e+00 1.49896014e+00
-2.34488010e-01 2.11168438e-01 -3.31661314e-01 7.13009715e-01
5.77025056e-01 7.18277931e-01 7.46553540e-02 4.52878177e-01
1.35855770e+00 -1.26523495e+00 -5.44174373e-01 -4.65868950e-01
9.79456961e-01 -6.97975099e-01 1.40271854e+00 2.03911364e-01
-1.17489254e+00 -4.96265739e-01 -1.10064077e+00 -6.06285691e-01
-3.98533434e-01 2.25374149e-03 9.63063657e-01 6.30035222e-01
-1.09028852e+00 8.17653298e-01 -7.68012166e-01 -2.11964563e-01
5.91539085e-01 2.17856199e-01 -2.17423692e-01 -3.16795915e-01
-1.01338601e+00 6.10741854e-01 1.75753638e-01 -5.82073517e-02
-5.55309296e-01 -1.21605015e+00 -9.41728175e-01 7.03778088e-01
2.79388130e-01 -8.22130680e-01 1.45804358e+00 -5.48423707e-01
-1.11972427e+00 6.11841440e-01 -4.94049609e-01 -5.93761802e-01
2.02646106e-01 -4.65735376e-01 -2.44941041e-01 6.22343533e-02
7.43495449e-02 2.89425582e-01 6.22237802e-01 -9.59389508e-01
-3.27388167e-01 -3.80766928e-01 3.16064060e-02 4.56628725e-02
-4.93718028e-01 1.41357761e-02 -8.20344090e-01 -7.95986950e-01
-2.90346652e-01 -4.48011696e-01 -2.26709694e-01 -9.08340886e-02
-6.09344721e-01 -4.64769810e-01 4.80703086e-01 -7.78436303e-01
1.68501520e+00 -2.03877115e+00 -1.39966324e-01 2.14183703e-01
4.72468287e-01 6.50117874e-01 -6.92249477e-01 7.36548781e-01
1.14087544e-01 6.18169606e-01 6.18578754e-02 -6.78243339e-01
2.97211170e-01 1.39933467e-01 -4.02990282e-01 2.20938846e-02
2.79603779e-01 1.36629903e+00 -8.83682668e-01 -2.54716545e-01
5.41249476e-03 2.13201553e-01 -8.66937637e-01 9.40524340e-02
-5.58536172e-01 -4.17538524e-01 -5.19739628e-01 3.01170468e-01
7.54725158e-01 -5.51337361e-01 4.23327535e-01 6.72205240e-02
1.46023288e-01 1.00789225e+00 -9.25224781e-01 2.03638625e+00
-6.76951647e-01 6.47597551e-01 -1.78713530e-01 -7.55050898e-01
5.06537914e-01 -1.21932045e-01 1.26141712e-01 -7.97533691e-01
-2.90254235e-01 -9.43235680e-02 4.82182316e-02 -5.20082057e-01
1.26525962e+00 2.38476530e-01 -8.65365863e-02 8.68037641e-01
1.10493787e-01 4.89650406e-02 5.43342412e-01 8.09479892e-01
1.44469821e+00 -6.91532716e-02 1.52404636e-01 -1.48139587e-02
1.06461465e-01 -3.41522366e-01 2.78052181e-01 1.09663129e+00
1.99614272e-01 5.72206557e-01 6.27530813e-01 -3.40652198e-01
-1.13936150e+00 -1.13728881e+00 3.48882705e-01 1.22619665e+00
-7.11033642e-02 -9.98708010e-01 -6.35110795e-01 -8.50087643e-01
3.76031697e-01 8.88955712e-01 -4.96639132e-01 -9.71028432e-02
-1.05011463e+00 -2.86270797e-01 9.31428075e-01 1.06824160e+00
1.48917153e-01 -1.09590602e+00 -2.07047537e-01 5.83457172e-01
5.61518371e-02 -1.06055522e+00 -1.06691885e+00 1.87037215e-01
-6.83882535e-01 -8.52334261e-01 -6.35850430e-01 -7.71054566e-01
6.29112422e-01 4.58278447e-01 1.40196812e+00 4.98342007e-01
-3.81030738e-01 9.54244286e-02 -4.23943132e-01 -1.67446911e-01
-1.60217270e-01 5.04573882e-01 -5.46736240e-01 -2.40104660e-01
5.77190161e-01 -4.10205066e-01 -5.55565834e-01 -3.35603625e-01
-9.52451110e-01 -2.23298017e-02 7.12970734e-01 1.06382322e+00
1.30981132e-01 -2.69028068e-01 1.90464705e-01 -1.06701183e+00
1.04446280e+00 -1.27472997e-01 -4.39894885e-01 4.53519940e-01
-5.09985209e-01 4.20927346e-01 1.00310886e+00 -1.18776008e-01
-1.10572696e+00 -4.34100956e-01 6.00082055e-02 -1.47533819e-01
2.63289176e-02 6.40498936e-01 -4.61258553e-02 4.64217961e-01
4.27865714e-01 3.48262191e-01 -2.64111996e-01 -6.76916897e-01
6.80922627e-01 5.39359748e-01 4.54234689e-01 -8.58739138e-01
6.54712379e-01 1.91672564e-01 -2.92016357e-01 -7.82552421e-01
-6.07144356e-01 -6.59456313e-01 -5.14709115e-01 3.88938874e-01
6.40017569e-01 -6.08571708e-01 -6.61474466e-01 4.64734048e-01
-1.63012624e+00 -6.44247115e-01 -1.83071181e-01 2.83562653e-02
-9.95457098e-02 8.48174870e-01 -1.21095121e+00 -5.72975039e-01
-6.31859004e-01 -8.00754845e-01 1.05400395e+00 9.12549645e-02
-3.92140567e-01 -1.07024062e+00 -1.40845343e-01 2.16431558e-01
3.21212798e-01 -4.83500242e-01 1.40143895e+00 -5.23727596e-01
-9.02427435e-01 -1.36925057e-01 -6.45638883e-01 1.15488574e-01
5.30070812e-02 6.21394673e-03 -6.46050990e-01 -2.40994632e-01
-7.82099664e-01 -1.29033074e-01 1.37392962e+00 2.17057228e-01
1.38608503e+00 -5.91762900e-01 -4.19980496e-01 6.69893861e-01
1.42001259e+00 5.86212017e-02 7.23744154e-01 2.57805765e-01
9.93574798e-01 3.52775872e-01 9.05503482e-02 5.57317972e-01
3.48365337e-01 4.32531714e-01 1.31869614e-01 -8.01443532e-02
-4.99197096e-01 -7.56306350e-01 2.60275036e-01 6.04440987e-01
3.17623675e-01 -8.10640395e-01 -8.14878345e-01 9.13569570e-01
-1.77306235e+00 -1.13054979e+00 -2.59238750e-01 1.78713727e+00
8.38633060e-01 1.61977425e-01 2.49550473e-02 -9.63314921e-02
5.69514751e-01 3.60241085e-01 -3.87042314e-01 -6.85091436e-01
-1.38465330e-01 6.56749010e-01 7.78113484e-01 4.14016157e-01
-6.87754393e-01 1.34973896e+00 6.67295599e+00 1.01618481e+00
-9.54693556e-01 -1.36951998e-01 4.47286636e-01 -1.60559684e-01
-9.04576182e-01 -2.97646923e-03 -9.23460662e-01 5.03352940e-01
6.11589849e-01 -1.63574934e-01 6.45127714e-01 5.43613255e-01
1.90782115e-01 1.64656147e-01 -1.27854860e+00 8.19229186e-01
7.80302659e-02 -1.84075844e+00 4.81798172e-01 8.48527700e-02
6.88899398e-01 1.01116903e-01 8.00070614e-02 2.94472396e-01
7.52732813e-01 -1.28289235e+00 8.39343309e-01 1.96356028e-01
7.95977652e-01 -5.35464525e-01 4.70735699e-01 -1.17436880e-02
-1.28792655e+00 4.56392160e-03 -5.33876538e-01 -1.83088198e-01
3.32814902e-01 5.91530383e-01 -1.04738975e+00 6.28352582e-01
2.51685530e-01 8.08368683e-01 -6.38942540e-01 9.80694890e-01
-3.85101140e-01 7.90755093e-01 -2.21383750e-01 -6.00144684e-01
6.04527414e-01 -1.32039553e-02 2.23570392e-01 1.71012998e+00
1.99255452e-01 -8.02226812e-02 6.63384646e-02 1.17690623e+00
-4.45690662e-01 6.58654317e-04 -4.19067025e-01 -4.81586397e-01
6.34916902e-01 1.01227081e+00 -3.71846437e-01 -6.27906859e-01
-6.41577721e-01 1.07308018e+00 7.41108119e-01 5.67179620e-01
-5.77762485e-01 -7.23210633e-01 8.76211405e-01 2.94464320e-01
8.03550124e-01 -5.80861926e-01 -7.23181367e-01 -1.24252653e+00
2.66487122e-01 -7.93346226e-01 4.22543347e-01 -6.98721170e-01
-1.20033205e+00 4.01350677e-01 -3.97428900e-01 -6.57366216e-01
-1.78206623e-01 -7.61037827e-01 -8.42738628e-01 9.69164848e-01
-1.47673082e+00 -1.20742953e+00 -3.93363275e-02 2.73878485e-01
7.02250779e-01 -1.01006202e-01 6.43598139e-01 2.41929591e-01
-4.59920198e-01 1.19040358e+00 2.28512481e-01 4.56411004e-01
4.77092981e-01 -1.51094818e+00 1.18359077e+00 9.92483675e-01
3.43624443e-01 1.17600322e+00 3.07787031e-01 -6.37235701e-01
-1.56908715e+00 -9.71531928e-01 1.13710511e+00 -2.88393229e-01
8.48982990e-01 -6.38244629e-01 -1.01156926e+00 8.46923470e-01
4.05759275e-01 -3.47737432e-01 5.85550785e-01 2.52123296e-01
-9.61550891e-01 1.59343079e-01 -6.26863658e-01 8.53083909e-01
1.54169023e+00 -7.47817934e-01 -6.28602922e-01 1.84979439e-01
8.43605280e-01 -2.24008679e-01 -4.79918361e-01 -4.67732430e-01
6.87373400e-01 -7.15269804e-01 7.95849442e-01 -1.01727176e+00
7.09014893e-01 -8.48273805e-04 9.08372104e-02 -1.32958627e+00
-7.12974250e-01 -9.28457975e-01 -1.95509478e-01 1.35824907e+00
6.48346841e-01 -3.44026446e-01 1.11129320e+00 5.25533795e-01
-3.29690844e-01 -6.83407962e-01 -3.76703620e-01 -1.01912880e+00
2.99402684e-01 -4.79867041e-01 9.69376028e-01 7.22118020e-01
3.63077134e-01 4.65376019e-01 -1.72946528e-01 -3.73835832e-01
3.55529904e-01 2.55581379e-01 7.10880160e-01 -6.57529950e-01
-6.10764325e-01 -8.29666078e-01 -2.27371931e-01 -2.00405908e+00
5.13832510e-01 -1.33752584e+00 -1.34555370e-01 -1.91314507e+00
3.38047624e-01 -3.42424661e-01 -2.16730401e-01 6.06007993e-01
-4.57250416e-01 -1.47626758e-01 5.08675754e-01 -1.08688109e-01
-6.43350482e-01 5.68333209e-01 1.28655505e+00 -5.96465349e-01
-2.51375157e-02 -4.45290446e-01 -1.01985693e+00 8.89645964e-02
5.74763536e-01 -2.17434973e-01 -1.73476756e-01 -1.26096165e+00
2.31206015e-01 -1.62540421e-01 2.80807987e-02 -4.00398821e-01
1.94435164e-01 -2.56754942e-02 3.79466891e-01 -8.09339404e-01
3.32332030e-02 -3.85675758e-01 -6.12519443e-01 4.53552365e-01
-6.60228491e-01 3.41239512e-01 2.84270436e-01 5.35402775e-01
7.20192790e-02 -4.20847833e-01 2.49450549e-01 -2.58899897e-01
-8.98469090e-01 3.46108288e-01 -5.56039214e-01 2.41346031e-01
3.93662661e-01 -3.40465337e-01 -4.61814910e-01 -4.44162816e-01
-2.04932809e-01 2.23936066e-01 4.92345393e-01 4.51044142e-01
6.60533309e-01 -1.12147272e+00 -6.70718789e-01 2.75522679e-01
1.66508611e-02 8.66868347e-03 3.09685677e-01 2.83678472e-01
-7.73818374e-01 6.47999346e-01 1.50750559e-02 -1.52745858e-01
-1.30833530e+00 6.72023058e-01 -8.59931577e-03 -8.26817572e-01
-6.65503681e-01 8.89019966e-01 3.24256897e-01 -2.24206626e-01
8.59408826e-02 -6.19281888e-01 1.95981070e-01 -3.24164897e-01
7.66843021e-01 3.00281316e-01 3.33922982e-01 -2.90112440e-02
-2.48972043e-01 4.43432689e-01 -8.34447742e-01 -9.69406515e-02
1.36929750e+00 4.45845351e-02 -1.54158100e-01 -2.66789734e-01
1.47656322e+00 1.30294293e-01 -1.19859254e+00 -4.18495446e-01
3.47621053e-01 -5.56897938e-01 -1.43355504e-01 -7.50257552e-01
-8.68196249e-01 1.00217366e+00 -1.51881844e-01 -7.93666393e-03
6.50650084e-01 -2.50455379e-01 1.27935946e+00 5.60709298e-01
1.07406162e-01 -1.16760623e+00 3.92178446e-01 9.25348639e-01
7.59592354e-01 -5.74178338e-01 -6.08725585e-02 -4.90149796e-01
-3.56923640e-01 1.16299403e+00 6.29294872e-01 -1.55490071e-01
3.48976105e-01 5.48496783e-01 -1.92595005e-01 -6.52181059e-02
-9.62319016e-01 -9.26454365e-02 2.50186950e-01 6.41262233e-01
7.76610792e-01 4.29610945e-02 -2.52032548e-01 6.32048130e-01
-2.80921340e-01 -2.51769811e-01 5.12866437e-01 9.26254511e-01
-2.23136902e-01 -1.25824022e+00 3.80496122e-02 7.71778762e-01
-5.04904628e-01 -7.86028683e-01 -5.42753339e-01 5.25908828e-01
-3.51710320e-01 8.90954196e-01 4.09149170e-01 -2.71371096e-01
2.10431412e-01 9.45155472e-02 8.31368685e-01 -9.29604352e-01
-9.81247008e-01 -1.00666948e-01 3.18187952e-01 -5.63117385e-01
4.14794296e-01 -5.41502953e-01 -1.52792931e+00 -7.75782645e-01
-2.47705430e-01 -1.34557143e-01 3.98330390e-01 7.51495004e-01
7.77027130e-01 7.52301157e-01 1.24828175e-01 -5.34392059e-01
-8.80476058e-01 -1.05708158e+00 -6.26198471e-01 8.45508337e-01
2.08565131e-01 -1.05840981e-01 1.26246303e-01 -1.34516641e-01] | [10.855698585510254, 7.7542877197265625] |
5ed03802-ae5b-4864-a7fb-beb94140dfba | semantic-specialization-for-knowledge-based | 2304.11340 | null | https://arxiv.org/abs/2304.11340v1 | https://arxiv.org/pdf/2304.11340v1.pdf | Semantic Specialization for Knowledge-based Word Sense Disambiguation | A promising approach for knowledge-based Word Sense Disambiguation (WSD) is to select the sense whose contextualized embeddings computed for its definition sentence are closest to those computed for a target word in a given sentence. This approach relies on the similarity of the \textit{sense} and \textit{context} embeddings computed by a pre-trained language model. We propose a semantic specialization for WSD where contextualized embeddings are adapted to the WSD task using solely lexical knowledge. The key idea is, for a given sense, to bring semantically related senses and contexts closer and send different/unrelated senses farther away. We realize this idea as the joint optimization of the Attract-Repel objective for sense pairs and the self-training objective for context-sense pairs while controlling deviations from the original embeddings. The proposed method outperformed previous studies that adapt contextualized embeddings. It achieved state-of-the-art performance on knowledge-based WSD when combined with the reranking heuristic that uses the sense inventory. We found that the similarity characteristics of specialized embeddings conform to the key idea. We also found that the (dis)similarity of embeddings between the related/different/unrelated senses correlates well with the performance of WSD. | ['Naoaki Okazaki', 'Sakae Mizuki'] | 2023-04-22 | null | null | null | null | ['word-sense-disambiguation'] | ['natural-language-processing'] | [ 1.08282901e-01 -1.48388088e-01 -1.21968150e-01 -4.35367197e-01
-5.28826475e-01 -7.40589619e-01 7.54452884e-01 1.01691341e+00
-1.17678177e+00 4.53743219e-01 6.37168944e-01 -6.44877627e-02
-4.67140496e-01 -9.06797409e-01 -1.64444000e-02 -4.91081417e-01
1.31027400e-01 3.95569503e-01 4.77170229e-01 -8.57234836e-01
6.71126485e-01 2.54702598e-01 -1.59249294e+00 7.94634074e-02
1.06172538e+00 8.48197520e-01 7.97248423e-01 2.96847433e-01
-6.67059958e-01 -3.92048992e-02 -6.93949938e-01 -1.63477391e-01
8.65787640e-02 -2.91672081e-01 -1.11060131e+00 -3.72440249e-01
3.99082094e-01 6.19469821e-01 3.20285112e-02 1.17823577e+00
5.79692304e-01 6.40375793e-01 4.98888314e-01 -7.46972024e-01
-1.15058851e+00 5.11829317e-01 -7.93187991e-02 6.10330582e-01
7.36519516e-01 -3.82302076e-01 1.59149122e+00 -9.26834881e-01
7.50426769e-01 1.19120800e+00 4.84318852e-01 4.98035580e-01
-1.28095436e+00 -1.22002058e-01 4.26003039e-01 1.42619416e-01
-1.55649292e+00 5.76996692e-02 6.58633947e-01 -3.36786181e-01
1.52146554e+00 2.01703742e-01 7.91016042e-01 9.13933039e-01
-5.47901145e-04 5.19183576e-01 9.30886447e-01 -7.57836163e-01
5.40167332e-01 2.70885020e-01 6.88526750e-01 1.61943972e-01
5.48332155e-01 -2.54951686e-01 -6.01450205e-01 -3.27905774e-01
1.34823874e-01 3.80673818e-02 -2.04696566e-01 -2.99787462e-01
-1.13293314e+00 8.80652368e-01 3.56302500e-01 1.13529468e+00
-3.52897316e-01 -3.23917300e-01 4.24333632e-01 4.04042929e-01
5.72481632e-01 1.36133444e+00 -8.60471308e-01 -5.61450422e-02
-8.41562629e-01 4.32521909e-01 6.00528061e-01 6.40163541e-01
1.10444820e+00 -4.43954855e-01 -2.82249600e-01 9.66420472e-01
2.50517398e-01 4.12772954e-01 1.05007029e+00 -3.66922438e-01
2.63573766e-01 1.01547384e+00 4.28602211e-02 -1.27459347e+00
-4.43485349e-01 -3.36479336e-01 -3.89378637e-01 -3.40611517e-01
5.64875379e-02 7.98358321e-02 -8.04412246e-01 1.93014121e+00
3.28179240e-01 3.25501651e-01 3.94097060e-01 9.08450544e-01
6.41395390e-01 3.78528208e-01 3.83018345e-01 -1.48211226e-01
1.54277575e+00 -3.65927130e-01 -5.70743799e-01 -6.46308243e-01
7.71100104e-01 -9.39602733e-01 1.59741819e+00 -9.38204825e-02
-5.48336506e-01 -4.83629405e-01 -1.31387472e+00 -1.44850917e-03
-9.77165759e-01 -2.60575503e-01 3.60836834e-01 5.16378403e-01
-1.12819409e+00 6.57490253e-01 -2.74130940e-01 -9.53203440e-01
-1.61096096e-01 1.72848478e-01 -3.12850863e-01 2.14349449e-01
-1.85241628e+00 1.20608175e+00 8.63208055e-01 -5.67932487e-01
-1.73137501e-01 -7.53202558e-01 -9.99445200e-01 -9.98856500e-02
3.30106944e-01 -7.26389647e-01 7.26440191e-01 -6.84875846e-01
-9.00502801e-01 1.17671895e+00 -2.73091555e-01 -4.84419435e-01
-4.71246839e-01 -3.23163807e-01 -8.79382372e-01 -1.35073811e-02
6.12359524e-01 4.24808323e-01 5.21234035e-01 -1.06611145e+00
-7.17869163e-01 -4.48047101e-01 1.73824951e-01 4.07235235e-01
-7.94920743e-01 -2.32900843e-01 -1.87683657e-01 -8.64246070e-01
1.51781559e-01 -6.45279229e-01 -3.08565855e-01 -5.74298322e-01
-2.03869507e-01 -7.40961909e-01 7.00850129e-01 2.00150069e-02
1.90253627e+00 -2.17108703e+00 1.92228124e-01 2.87454575e-01
2.03711212e-01 5.12527347e-01 -4.86972481e-01 7.86527514e-01
-1.71396092e-01 4.19079840e-01 -1.49124175e-01 -4.09781747e-02
1.67698637e-01 4.04899895e-01 -2.69813061e-01 -7.94042721e-02
2.49173760e-01 4.26224977e-01 -1.49299502e+00 -4.45220530e-01
1.08444355e-01 1.97313771e-01 -5.61282218e-01 7.27394372e-02
-1.31982118e-01 -3.10238868e-01 -6.10460341e-01 9.55751464e-02
4.92502570e-01 -2.33709645e-02 4.66580510e-01 -2.90369958e-01
7.61269499e-03 6.99568748e-01 -1.31350517e+00 1.86230040e+00
-6.98728025e-01 3.14403474e-01 -4.91334170e-01 -8.65794182e-01
1.10978270e+00 1.13961443e-01 4.22146291e-01 -7.81551242e-01
-1.55813143e-01 3.69824797e-01 -2.22313061e-01 -5.57843745e-01
1.11483276e+00 -4.81018007e-01 -5.28760076e-01 3.97339642e-01
3.50344956e-01 -3.17523211e-01 3.74469727e-01 4.15375739e-01
1.00871980e+00 -3.02171648e-01 8.53680789e-01 -7.82929063e-01
4.72770154e-01 -3.31548683e-04 5.93775094e-01 7.36284494e-01
-1.32316634e-01 3.22519928e-01 3.14828306e-01 -3.24102342e-01
-7.03686059e-01 -1.41290557e+00 -1.49678901e-01 1.32728267e+00
5.93549550e-01 -9.25374269e-01 -6.30865157e-01 -8.20041776e-01
2.17972383e-01 1.10568213e+00 -7.72101641e-01 -3.41060907e-01
-2.96368599e-01 -5.10615349e-01 3.98684800e-01 3.83250356e-01
9.26059857e-02 -9.64819193e-01 -5.13772666e-01 3.37598681e-01
-5.20328507e-02 -8.36258411e-01 -5.94204545e-01 3.49611640e-01
-5.14383197e-01 -9.40908730e-01 -3.25956672e-01 -8.87923896e-01
3.97496819e-01 3.28157961e-01 1.36151874e+00 1.02331698e-01
-7.89420456e-02 4.17887032e-01 -9.14972782e-01 -3.61007780e-01
-8.42991695e-02 2.15457126e-01 5.88266432e-01 -1.15449578e-01
9.02725935e-01 -4.83381927e-01 -5.22873759e-01 1.00940242e-01
-1.44980609e+00 -7.41724730e-01 1.58878759e-01 9.28693473e-01
6.90570056e-01 -8.52674097e-02 5.57647288e-01 -5.62973619e-01
1.31620693e+00 -4.25631285e-01 -2.89714560e-02 4.11692441e-01
-1.05611694e+00 5.50609767e-01 3.50304872e-01 -5.29888988e-01
-7.57694900e-01 -2.65622020e-01 -2.28849530e-01 3.89462449e-02
-7.29036108e-02 5.44954181e-01 1.03751346e-02 4.15085942e-01
8.70019138e-01 9.22005549e-02 -5.78698874e-01 -4.02784377e-01
5.87668180e-01 6.67400539e-01 6.56986013e-02 -7.66590714e-01
4.39613044e-01 1.15856737e-01 -4.64011014e-01 -1.05584931e+00
-1.20845830e+00 -9.01552260e-01 -7.21894383e-01 2.87875712e-01
1.14108109e+00 -3.71216059e-01 -1.61430359e-01 -1.59477204e-01
-1.14869118e+00 3.37868422e-01 -6.60615444e-01 5.43146968e-01
-1.13455981e-01 4.58170325e-01 1.08900040e-01 -5.70043623e-01
-3.16244006e-01 -8.57262313e-01 1.17000473e+00 2.10609689e-01
-8.96566987e-01 -1.41983223e+00 4.16444510e-01 -2.25270361e-01
4.71115053e-01 -1.95226334e-02 1.14746714e+00 -1.43927360e+00
3.39915782e-01 -3.79738808e-02 8.14964920e-02 4.47374582e-01
5.52164793e-01 -4.32469189e-01 -7.82162070e-01 -1.51548132e-01
-9.78148654e-02 -5.64704351e-02 1.16522944e+00 5.94358221e-02
4.13595885e-01 -1.30411386e-01 -5.05331099e-01 6.78019896e-02
1.64262128e+00 1.09541327e-01 2.81368911e-01 5.21529913e-01
3.97109628e-01 3.48996490e-01 8.14731836e-01 4.55696166e-01
2.68352151e-01 5.77619612e-01 5.86406421e-03 2.53681570e-01
1.92447484e-01 -3.63444716e-01 1.59311786e-01 8.56823623e-01
4.22963977e-01 -3.44020873e-01 -9.91027236e-01 1.03064871e+00
-1.68170631e+00 -9.77954686e-01 2.29848906e-01 2.23322892e+00
9.80850697e-01 3.42591733e-01 -1.36109278e-01 9.83023345e-02
5.84903419e-01 4.85306323e-01 -8.65709484e-02 -7.77045667e-01
-3.05094928e-01 6.12559676e-01 1.97488204e-01 8.25703144e-01
-9.42183256e-01 1.33898580e+00 5.97292757e+00 9.84668970e-01
-8.89844894e-01 1.97358742e-01 -1.19405963e-01 1.00243539e-01
-8.44085038e-01 8.91403034e-02 -8.25188041e-01 3.61335337e-01
5.11257946e-01 -4.85034227e-01 1.19692631e-01 6.95458770e-01
1.50850299e-03 -2.47881755e-01 -1.10560346e+00 7.82548308e-01
2.51882017e-01 -1.31054974e+00 3.49813730e-01 -3.45009208e-01
6.46036863e-01 -1.16220452e-02 -1.40502915e-01 2.69743651e-01
2.70347476e-01 -5.88481843e-01 4.23021853e-01 5.15070081e-01
4.45137560e-01 -5.90004623e-01 7.89953828e-01 -4.28749882e-02
-1.42274094e+00 1.45502528e-02 -5.30578136e-01 -1.17283821e-01
1.65167227e-02 8.62204194e-01 -9.47763324e-01 5.83435655e-01
5.72884738e-01 7.22766101e-01 -5.80576777e-01 4.99461859e-01
-2.36953974e-01 3.06322426e-01 -1.27058640e-01 -4.87986237e-01
5.12422502e-01 -6.93426505e-02 8.97457182e-01 1.51101959e+00
3.07868540e-01 -1.98811619e-03 3.97939742e-01 6.63612604e-01
2.10001543e-01 4.16492999e-01 -6.96479201e-01 -1.39137298e-01
8.81409168e-01 9.35099721e-01 -5.86988628e-01 -4.78001416e-01
-1.32215276e-01 1.10406959e+00 3.17097694e-01 2.48759925e-01
-2.20590919e-01 -7.48637676e-01 1.43740118e+00 -4.80388030e-02
4.17072624e-01 -2.17751011e-01 -3.03562194e-01 -1.01930797e+00
7.73421526e-02 -2.42476285e-01 7.75013328e-01 -4.98364866e-01
-1.80650997e+00 7.81369925e-01 1.45155713e-01 -1.10414517e+00
-4.54763509e-02 -6.35515392e-01 -5.76303899e-01 8.41518581e-01
-1.51279020e+00 -6.46443188e-01 1.85481384e-01 4.92840141e-01
4.43164736e-01 -2.28385031e-01 1.21818948e+00 2.18194928e-02
-6.12514764e-02 5.36010087e-01 -3.51087786e-02 2.62219068e-02
8.26853871e-01 -1.60353422e+00 3.26502204e-01 7.37187624e-01
2.78033644e-01 1.10249877e+00 9.28464234e-01 -7.50417233e-01
-9.67009544e-01 -9.89754021e-01 1.71760333e+00 -3.81395400e-01
9.26231682e-01 -1.04786180e-01 -7.87995756e-01 1.14334799e-01
4.18848336e-01 -2.09863663e-01 9.57750022e-01 4.70681190e-01
-5.87907374e-01 -2.30435416e-01 -1.15371144e+00 6.25456750e-01
1.18205273e+00 -7.71300435e-01 -1.56712413e+00 2.74283681e-02
1.21391428e+00 1.13233380e-01 -8.36236000e-01 -1.14885177e-02
1.35520443e-01 -7.03410149e-01 1.04968953e+00 -8.65758419e-01
1.01105362e-01 -5.06526887e-01 -6.33580208e-01 -1.88849926e+00
-4.66827929e-01 -3.52056354e-01 3.60738695e-01 1.30596113e+00
6.85891092e-01 -7.05755055e-01 1.78772151e-01 1.39640063e-01
-7.58216009e-02 -6.47254169e-01 -1.06647396e+00 -1.00152433e+00
1.91365927e-01 -5.10768533e-01 7.88796127e-01 1.29919136e+00
3.63860965e-01 6.61063671e-01 3.65168631e-01 2.79003978e-01
2.17487544e-01 7.85600245e-02 -1.45908341e-01 -1.38733542e+00
6.86695725e-02 -5.40739954e-01 -8.65846276e-01 -8.48577023e-01
3.78463268e-01 -1.27749586e+00 -1.04074560e-01 -1.69235206e+00
-1.40025228e-01 -3.77338886e-01 -1.00541496e+00 4.64372098e-01
-5.92501104e-01 -1.51508912e-01 2.74518937e-01 -2.56252706e-01
-7.74863064e-01 4.55192000e-01 9.13533568e-01 -1.81841895e-01
-4.26582217e-01 -4.56017435e-01 -1.06412446e+00 5.55480003e-01
7.58488476e-01 -6.71793580e-01 -5.67561030e-01 -5.12690306e-01
6.25607491e-01 -4.80668902e-01 1.24040619e-01 -8.48177910e-01
2.11048499e-01 -3.23263794e-01 -5.12346402e-02 -2.05839753e-01
1.75610855e-01 -7.89127827e-01 -5.79329550e-01 2.20300332e-01
-6.21889889e-01 4.14220899e-01 4.72965688e-02 5.12947857e-01
-3.48582566e-01 -3.71822268e-01 5.15293360e-01 -1.21064864e-01
-1.29113507e+00 -2.22480491e-01 -3.54938090e-01 7.89914250e-01
9.52565312e-01 -2.97088891e-01 -1.45780314e-02 5.64905927e-02
-7.20179081e-01 9.18108150e-02 4.82397705e-01 9.56545770e-01
7.90100098e-01 -1.58334517e+00 -5.08152962e-01 8.60512555e-02
7.06796527e-01 -3.69327575e-01 -1.22133024e-01 3.63006651e-01
1.43257037e-01 3.11757028e-01 1.31857634e-01 -4.21317130e-01
-1.14272285e+00 6.39183581e-01 2.66174793e-01 -2.76227176e-01
-2.32450306e-01 9.49163079e-01 -1.32040069e-01 -6.29786015e-01
-1.70902342e-01 -6.19122922e-01 -2.92923778e-01 5.15983045e-01
6.38099015e-01 1.93416044e-01 2.93482065e-01 -5.54108381e-01
-8.30361485e-01 6.55232728e-01 -4.99696732e-02 -1.57429039e-01
1.24997830e+00 -2.29917169e-01 -2.11640775e-01 4.92055297e-01
1.42219985e+00 2.13417336e-01 -3.71272534e-01 -5.08867562e-01
5.94463050e-01 -5.02170920e-01 -2.75771115e-02 -8.82761240e-01
-5.49476445e-01 3.99706781e-01 6.44899905e-01 5.95428944e-01
1.10472441e+00 3.72157604e-01 8.37968767e-01 5.07274330e-01
3.15802544e-01 -1.53636742e+00 1.68935642e-01 1.01690912e+00
6.21818364e-01 -7.76418984e-01 -1.30404904e-01 -1.48145661e-01
-8.76886189e-01 9.70000386e-01 4.25098151e-01 -2.16400206e-01
8.43934298e-01 4.01408076e-02 2.07691282e-01 -3.39507848e-01
-4.68650222e-01 -7.21632063e-01 5.43685555e-01 7.25861967e-01
6.33745909e-01 2.25178048e-01 -8.12825918e-01 6.98306739e-01
-3.16276014e-01 -4.65715766e-01 1.17610015e-01 8.77548277e-01
-8.29871953e-01 -1.37224686e+00 -9.81293842e-02 2.99968451e-01
-5.45604490e-02 -4.74195033e-01 -7.09530950e-01 6.21302009e-01
5.04134178e-01 1.22316790e+00 1.59467041e-01 -7.51972735e-01
5.85431039e-01 3.24056387e-01 1.44495755e-01 -9.70680296e-01
-6.63909376e-01 -3.82290721e-01 7.22059309e-02 -5.00784516e-01
-5.18701911e-01 -5.02137840e-01 -1.45473814e+00 1.65059596e-01
-2.84186006e-01 5.20079851e-01 3.40396613e-01 1.28768587e+00
5.56906223e-01 3.10103148e-01 6.00030959e-01 -4.02229190e-01
-5.64055324e-01 -7.83196926e-01 -7.74717510e-01 1.00712371e+00
3.15981299e-01 -7.71078110e-01 -2.58714914e-01 -3.38840812e-01] | [10.314053535461426, 8.962039947509766] |
950a0cbe-328a-4b70-aa7f-84dfb453626d | mixture-model-for-designs-in-high-dimensional | 1210.4762 | null | https://arxiv.org/abs/1210.4762v2 | https://arxiv.org/pdf/1210.4762v2.pdf | Mixture model for designs in high dimensional regression and the LASSO | The LASSO is a recent technique for variable selection in the regression model \bean y & = & X\beta + z, \eean where $X\in \R^{n\times p}$ and $z$ is a centered gaussian i.i.d. noise vector $\mathcal N(0,\sigma^2I)$. The LASSO has been proved to achieve remarkable properties such as exact support recovery of sparse vectors when the columns are sufficently incoherent and low prediction error under even less stringent conditions. However, many matrices do not satisfy small coherence in practical applications and the LASSO estimator may thus suffer from what is known as the slow rate regime. The goal of the present paper is to study the LASSO from a slightly different perspective by proposing a mixture model for the design matrix which is able to capture in a natural way the potentially clustered nature of the columns in many practical situations. In this model, the columns of the design matrix are drawn from a Gaussian mixture model. Instead of requiring incoherence for the design matrix $X$, we only require incoherence of the much smaller matrix of the mixture's centers. Our main result states that $X\beta$ can be estimated with the same precision as for incoherent designs except for a correction term depending on the maximal variance in the mixture model. | ['Stéphane Chrétien', 'Mohamed Ibrahim Assoweh'] | 2012-10-17 | null | null | null | null | ['variable-selection'] | ['methodology'] | [ 1.30232185e-01 -5.88835850e-02 -1.07596606e-01 3.77277546e-02
-4.80402857e-01 -4.04257238e-01 2.84131825e-01 -3.74660641e-01
-2.13387191e-01 1.08014131e+00 -2.17945307e-01 -1.38597399e-01
-6.59159303e-01 -5.95898807e-01 -6.79427385e-01 -1.35413957e+00
-1.94825277e-01 4.67450738e-01 -5.88688791e-01 -1.33520141e-01
1.10263497e-01 1.66410744e-01 -1.38364089e+00 -4.06282306e-01
1.06928360e+00 9.77923870e-01 -7.30848312e-02 5.48961282e-01
1.16917521e-01 6.23804748e-01 -3.90504986e-01 -2.96661973e-01
6.10362411e-01 -8.09316754e-01 -1.78234562e-01 1.43923312e-01
1.56429932e-01 3.37977320e-01 -2.61888534e-01 1.61205077e+00
2.99341798e-01 2.51963258e-01 8.66016567e-01 -1.04112482e+00
-1.22955568e-01 8.14115047e-01 -8.20040762e-01 -5.45048639e-02
-4.43097502e-02 -1.75338909e-01 8.02869022e-01 -9.91413534e-01
5.18570125e-01 7.76106954e-01 5.86555660e-01 2.86199570e-01
-1.63765156e+00 -1.01103032e+00 -2.11204946e-01 -3.24174106e-01
-1.85477054e+00 -6.75338507e-01 7.21121252e-01 -5.96430540e-01
4.15521562e-01 6.28495812e-01 3.28480333e-01 9.17895138e-01
1.71289235e-01 2.32989311e-01 1.15766859e+00 -4.85396147e-01
4.91916418e-01 3.53983819e-01 2.67811596e-01 7.62667477e-01
7.10170567e-01 2.03050062e-01 -5.30804515e-01 -4.06573623e-01
7.90631771e-01 -7.26956204e-02 -3.34518284e-01 -5.92303097e-01
-1.11857522e+00 1.06922972e+00 6.62186136e-03 4.58078682e-01
-4.40280169e-01 3.20128083e-01 1.06130235e-01 3.55677456e-01
4.34639484e-01 4.33972210e-01 -2.25635767e-01 1.09798685e-01
-1.09497404e+00 3.41482460e-01 7.84510911e-01 1.18846583e+00
5.77937543e-01 5.51477551e-01 5.83229624e-02 7.80098975e-01
3.12022924e-01 9.87015307e-01 1.13507457e-01 -1.27223217e+00
5.88448465e-01 6.09826855e-02 3.55010778e-01 -9.62561429e-01
-2.63403833e-01 -8.50204289e-01 -1.50734866e+00 2.82189608e-01
6.20464444e-01 -5.20780504e-01 -4.66130912e-01 2.05093837e+00
1.65562332e-01 1.97561383e-01 -1.34869888e-01 7.62558877e-01
1.01999104e-01 4.99562532e-01 -2.09293723e-01 -1.00225127e+00
8.65895987e-01 -4.20128614e-01 -9.39615011e-01 -2.53822297e-01
5.28815329e-01 -7.78801084e-01 3.53791356e-01 7.32401729e-01
-1.26124942e+00 -3.38197649e-01 -9.90377784e-01 6.77978218e-01
2.77902424e-01 9.11389366e-02 5.57112753e-01 7.50484824e-01
-8.43274832e-01 4.89678353e-01 -7.13798702e-01 9.55837518e-02
1.71446025e-01 4.16337997e-01 -5.88116229e-01 -1.65014729e-01
-7.54578531e-01 5.91468573e-01 -6.70267418e-02 3.56924772e-01
-3.91042352e-01 -3.80984664e-01 -7.51495063e-01 1.17160253e-01
2.93777198e-01 -6.94500864e-01 5.58039665e-01 -9.92999434e-01
-1.09949648e+00 5.68003953e-01 -4.63291734e-01 -3.18450332e-01
4.64962244e-01 -8.81257579e-02 -2.47836709e-01 3.87620516e-02
3.60069156e-01 -3.24568123e-01 1.33079731e+00 -1.19854951e+00
-2.55379319e-01 -7.36124992e-01 -8.23960662e-01 -8.70794207e-02
1.55225739e-01 -1.19897261e-01 -1.25680774e-01 -8.80423665e-01
6.77286863e-01 -1.17329705e+00 -6.09846711e-01 -3.87622207e-01
-5.65106690e-01 3.72211754e-01 2.56147981e-01 -3.66765380e-01
1.49633682e+00 -2.44841647e+00 4.83451366e-01 6.95528150e-01
4.65750515e-01 -4.07186486e-02 7.56541640e-02 5.81618011e-01
-5.10850489e-01 -4.87736091e-02 -5.00974119e-01 -4.02455151e-01
-7.45084360e-02 -2.21990552e-02 -2.34836951e-01 1.14744747e+00
-2.38403693e-01 1.35355368e-01 -5.97669721e-01 -2.88639069e-01
-5.75158335e-02 3.03163737e-01 -7.01127827e-01 -5.31859556e-03
1.55094629e-02 6.98679566e-01 -5.27774215e-01 3.86291891e-01
8.44176471e-01 -3.49731058e-01 3.37338179e-01 1.70714080e-01
-1.68541119e-01 -4.22060490e-01 -1.85751712e+00 1.42174840e+00
-9.37767401e-02 4.17536825e-01 6.98679686e-01 -1.53940558e+00
8.56125295e-01 5.77434659e-01 7.98512101e-01 -2.10740954e-01
3.13064963e-01 5.92463613e-01 1.89877048e-01 -4.35459435e-01
2.04536051e-01 -6.31225169e-01 -1.14029482e-01 1.53029963e-01
-1.96060203e-02 3.30285020e-02 2.23740473e-01 6.91671446e-02
1.18392384e+00 -3.27975780e-01 4.55202401e-01 -6.30242288e-01
4.38681245e-01 -2.56780416e-01 8.52470338e-01 1.12896860e+00
-1.17059007e-01 5.82982957e-01 6.48671269e-01 1.46125937e-02
-1.06255496e+00 -1.01320779e+00 -5.30977786e-01 4.98783737e-01
-7.83765689e-02 -2.59181529e-01 -5.73694766e-01 -2.58033216e-01
-5.84094739e-03 6.62837446e-01 -7.31133461e-01 -4.27138573e-03
-6.30311489e-01 -1.18249404e+00 2.14150056e-01 1.90757390e-03
2.65059084e-01 -5.45541584e-01 -1.11821361e-01 2.30420217e-01
-2.68965214e-02 -7.73330331e-01 -3.24151725e-01 4.73934621e-01
-8.99489820e-01 -9.87149298e-01 -7.96318650e-01 -4.15310651e-01
6.88772917e-01 1.94535196e-01 9.12336409e-01 -2.81356096e-01
-1.12711765e-01 9.58451852e-02 -2.23380849e-01 -1.00559674e-01
-1.92518115e-01 -4.06668872e-01 5.37838757e-01 2.28185833e-01
2.22833261e-01 -8.38080525e-01 -5.50945938e-01 3.05848777e-01
-6.89216971e-01 -3.82545322e-01 2.38063753e-01 1.27444780e+00
5.71610928e-01 3.49456340e-01 3.76529157e-01 -9.14934754e-01
2.72642434e-01 -7.11542964e-01 -7.59818435e-01 -1.59671918e-01
-6.46581531e-01 1.63517073e-01 6.39715374e-01 -4.11320388e-01
-7.12215364e-01 -1.29640438e-02 -2.83616222e-02 -6.15516186e-01
2.46942177e-01 5.96955538e-01 -8.00762624e-02 5.59922308e-03
7.33959198e-01 1.46469906e-01 1.55754581e-01 -5.75222969e-01
1.82783484e-01 4.17666942e-01 3.14213485e-01 -4.55342710e-01
7.90843308e-01 5.44867456e-01 5.54511666e-01 -1.04405892e+00
-5.94999194e-01 -3.70338172e-01 -2.50584841e-01 3.28694880e-01
5.13158381e-01 -9.16123509e-01 -7.69839764e-01 6.13881648e-02
-6.85047865e-01 -1.05663329e-01 -7.06301093e-01 1.00727034e+00
-9.17662263e-01 2.62816370e-01 -4.03299928e-01 -1.19492662e+00
-1.51581034e-01 -1.02162480e+00 4.53745514e-01 -1.03292555e-01
-2.54649282e-01 -6.43340051e-01 2.86458880e-01 3.15991417e-02
3.18052292e-01 1.10252880e-01 9.02156293e-01 -5.59173882e-01
-3.56294304e-01 -4.23176348e-01 4.53598946e-02 3.93443078e-01
-1.99938074e-01 -9.86546949e-02 -5.91161549e-01 -2.95758814e-01
7.41176546e-01 2.51119196e-01 6.80768371e-01 9.63129997e-01
7.85822868e-01 -4.61459875e-01 -4.11546737e-01 7.82792509e-01
1.60969245e+00 1.71696931e-01 3.25333834e-01 -1.75783813e-01
4.28422272e-01 5.02532125e-01 4.20029402e-01 7.80353963e-01
-3.66820902e-01 6.60525262e-01 2.10376978e-01 1.60369799e-01
3.53692442e-01 1.76035002e-01 1.77293196e-01 9.75972295e-01
-7.34685212e-02 -8.75037611e-02 -6.39751434e-01 3.72411370e-01
-1.93600118e+00 -1.27401912e+00 -5.97445130e-01 2.66496444e+00
6.38821959e-01 -2.59930808e-02 -5.74653037e-02 2.01043487e-01
6.08027935e-01 9.08338502e-02 -2.72817016e-01 -1.89481765e-01
-4.03337240e-01 5.87125182e-01 7.17904627e-01 7.64118850e-01
-8.97008479e-01 3.43505532e-01 6.34965372e+00 8.84776890e-01
-8.20157290e-01 2.42573291e-01 3.57503474e-01 -2.79325873e-01
-2.05250353e-01 1.14841059e-01 -7.37822592e-01 7.14360356e-01
7.65149057e-01 -3.08052972e-02 2.99292207e-01 7.25246370e-01
3.73459339e-01 -3.88616771e-01 -8.88617754e-01 1.21713519e+00
2.16319002e-02 -1.02161288e+00 -6.12877846e-01 4.51691270e-01
9.38964605e-01 -5.64558767e-02 1.98679909e-01 1.89765245e-01
3.03713262e-01 -1.13476217e+00 4.58308667e-01 5.93761444e-01
9.58607793e-01 -7.49015987e-01 6.06381893e-01 6.40069842e-01
-8.26937675e-01 -2.46340707e-01 -4.01616156e-01 -1.65780365e-01
2.08316758e-01 1.10007370e+00 -2.02417254e-01 4.66007084e-01
4.61428821e-01 5.27588546e-01 8.18879753e-02 9.41622019e-01
2.66238481e-01 6.32813156e-01 -5.72509348e-01 2.82135755e-01
7.35874996e-02 -9.88224447e-01 1.00379086e+00 9.45284903e-01
9.02836442e-01 3.13558459e-01 -3.95217240e-02 7.49113202e-01
2.01536819e-01 2.36536548e-01 -9.32181954e-01 2.65763840e-03
3.63152295e-01 6.82998478e-01 -5.36781311e-01 -3.46990436e-01
-2.77030349e-01 7.34656811e-01 -2.27381214e-02 7.06477940e-01
-6.07038200e-01 -2.47163132e-01 7.62494266e-01 1.17985748e-01
5.49096704e-01 -3.70401084e-01 -6.09374642e-01 -1.51865494e+00
-1.01560377e-01 -1.07806981e+00 1.75100356e-01 -2.62475312e-01
-1.21857595e+00 4.29293603e-01 -1.75333664e-01 -1.41073847e+00
-3.50797355e-01 -4.15010273e-01 -2.32049718e-01 1.25410879e+00
-6.54816866e-01 -4.40115303e-01 4.30735260e-01 6.06612921e-01
2.11799353e-01 -3.87169331e-01 9.04295981e-01 2.53188521e-01
-8.67562354e-01 3.97184193e-01 9.65868115e-01 -1.86132580e-01
5.75037003e-01 -1.05654716e+00 -4.61831719e-01 9.54960942e-01
-1.06092598e-02 1.02213645e+00 1.40530252e+00 -7.58464932e-01
-1.39594746e+00 -6.15892708e-01 8.52238774e-01 -2.70572156e-01
6.73990250e-01 -2.05966413e-01 -7.51794815e-01 7.96377003e-01
-2.02648435e-02 3.41653854e-01 8.76115263e-01 2.48560041e-01
-2.56582558e-01 -5.13652265e-02 -1.02077210e+00 4.09808338e-01
7.07438469e-01 -1.22167870e-01 -1.12169750e-01 3.52846712e-01
7.73140788e-02 -1.39928833e-01 -8.68683159e-01 4.68035340e-01
5.54235816e-01 -1.12666154e+00 9.48366284e-01 -6.66221261e-01
2.22173914e-01 -2.59643286e-01 -5.62488616e-01 -9.88924623e-01
-4.53329980e-01 -7.46854246e-01 -2.18358308e-01 9.38248754e-01
3.03589672e-01 -5.68699062e-01 7.98563004e-01 4.29194987e-01
1.45903870e-01 -4.22489703e-01 -1.46634173e+00 -7.82137394e-01
1.21284835e-01 -4.35026497e-01 6.97819442e-02 1.05959630e+00
1.23220406e-01 4.49438602e-01 -6.93297267e-01 1.59721270e-01
1.00823331e+00 1.94109380e-01 6.65990829e-01 -1.19422090e+00
-7.03870654e-01 -5.32815397e-01 -8.94033536e-02 -8.14307809e-01
1.14219800e-01 -6.55686617e-01 7.21986592e-02 -6.24629378e-01
1.48753315e-01 -9.25445139e-01 -1.44666150e-01 -3.78404200e-01
-9.30744037e-02 7.00431466e-02 1.52644187e-01 3.34354699e-01
-2.04572976e-01 5.82886755e-01 9.72694337e-01 2.65632290e-03
-2.74574548e-01 4.76085246e-01 -5.87780535e-01 8.77992868e-01
4.14163291e-01 -6.24522328e-01 -3.00907761e-01 -1.49116263e-01
5.34835041e-01 6.97091043e-01 6.27941266e-02 -8.69243622e-01
3.47489156e-02 -1.61855161e-01 1.84278309e-01 -2.30824769e-01
4.58915502e-01 -8.96417201e-01 7.97108889e-01 3.74152064e-01
-1.17931016e-01 -9.90757197e-02 -3.24233264e-01 7.95153856e-01
-1.19201116e-01 -6.25338614e-01 9.44463551e-01 -1.40536427e-01
2.58752167e-01 2.70694584e-01 -3.87627989e-01 -1.10420175e-01
8.01871896e-01 6.73871860e-02 2.47631944e-03 -4.78801966e-01
-9.54125345e-01 -1.15899511e-01 4.01472807e-01 -4.00732845e-01
2.21251577e-01 -1.27303720e+00 -7.71551788e-01 4.31097537e-01
-2.44421527e-01 -1.23257741e-01 5.49102008e-01 1.34720016e+00
-2.77393818e-01 2.40696818e-01 1.73719481e-01 -6.32134497e-01
-8.50393116e-01 7.83410609e-01 2.72517532e-01 -2.73941696e-01
-4.65290308e-01 9.16360319e-01 2.50094473e-01 -3.67483981e-02
3.05869523e-02 2.25300103e-01 6.57842383e-02 9.96205658e-02
5.77579260e-01 4.88174438e-01 -2.30440065e-01 -8.21418107e-01
-1.81041360e-01 5.24794996e-01 2.53246576e-01 -3.67145717e-01
1.15269792e+00 -3.30573022e-01 -4.64100301e-01 5.50052941e-01
1.22054410e+00 6.46105528e-01 -9.10843432e-01 -2.37169579e-01
-3.04710060e-01 -5.14523029e-01 -2.10805953e-01 -1.48072392e-01
-9.77491677e-01 6.17292047e-01 6.00496471e-01 4.45006520e-01
9.17520106e-01 -1.79604679e-01 2.26673931e-01 8.48960057e-02
4.09032166e-01 -9.43601489e-01 -3.94081831e-01 3.35600257e-01
7.42082715e-01 -1.21096063e+00 1.36620536e-01 -4.79452044e-01
-3.19252729e-01 7.17970312e-01 5.51664345e-02 -5.14214396e-01
9.92416263e-01 2.23975807e-01 -2.16151938e-01 -2.08674580e-01
-3.88388574e-01 -6.28825203e-02 1.13100000e-02 3.83671671e-01
5.56778073e-01 1.82403028e-01 -8.39359879e-01 7.00274110e-01
-1.25974610e-01 -1.67744309e-01 3.48131061e-01 6.49289191e-01
-3.60738158e-01 -1.04597116e+00 -7.84292877e-01 6.34055316e-01
-6.27143979e-01 -5.61531223e-02 3.12517077e-01 6.19192660e-01
8.24179798e-02 1.00798202e+00 -2.65177667e-01 -1.00984611e-01
4.83060442e-02 2.16958746e-01 3.32618773e-01 -4.99088436e-01
-2.45809808e-01 6.91987634e-01 -1.20358065e-01 -5.33024132e-01
-2.23095700e-01 -9.53961551e-01 -7.19031215e-01 -4.38107222e-01
-4.15131629e-01 6.36632085e-01 5.60553730e-01 9.42420661e-01
9.96965989e-02 1.30161420e-01 8.00596952e-01 -5.62825680e-01
-8.65999401e-01 -1.04576790e+00 -1.46269715e+00 2.37076238e-01
4.20933276e-01 -5.68314552e-01 -6.82784915e-01 -4.62160259e-02] | [7.003347396850586, 4.49307107925415] |
848d7ae5-68c2-4242-a29f-81103f39869a | projection-free-online-convex-optimization | 2305.01333 | null | https://arxiv.org/abs/2305.01333v2 | https://arxiv.org/pdf/2305.01333v2.pdf | Projection-Free Online Convex Optimization with Stochastic Constraints | This paper develops projection-free algorithms for online convex optimization with stochastic constraints. We design an online primal-dual projection-free framework that can take any projection-free algorithms developed for online convex optimization with no long-term constraint. With this general template, we deduce sublinear regret and constraint violation bounds for various settings. Moreover, for the case where the loss and constraint functions are smooth, we develop a primal-dual conditional gradient method that achieves $O(\sqrt{T})$ regret and $O(T^{3/4})$ constraint violations. Furthermore, for the setting where the loss and constraint functions are stochastic and strong duality holds for the associated offline stochastic optimization problem, we prove that the constraint violation can be reduced to have the same asymptotic growth as the regret. | ['Dabeen Lee', 'Nam Ho-Nguyen', 'Duksang Lee'] | 2023-05-02 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [-2.00140134e-01 1.92623675e-01 -3.17108363e-01 -4.09516126e-01
-1.19531357e+00 -7.56765366e-01 -2.52524912e-01 9.58394781e-02
-5.74288309e-01 9.75551605e-01 1.13350525e-01 -6.80260420e-01
-4.53632474e-01 -5.22289276e-01 -1.01143599e+00 -7.34350502e-01
-3.05113465e-01 3.00157875e-01 -3.47144574e-01 -1.06257223e-01
8.66243094e-02 2.25616038e-01 -8.94224107e-01 -2.99908727e-01
8.98600399e-01 1.49754810e+00 -9.73941460e-02 6.10858858e-01
2.16731429e-02 3.62520754e-01 -4.21733372e-02 -6.63122714e-01
9.39945996e-01 -2.87305593e-01 -5.92834473e-01 3.70098472e-01
3.98522019e-01 -3.60375583e-01 -4.05902803e-01 1.23303342e+00
5.25143743e-01 3.22874725e-01 7.16846436e-02 -1.08739996e+00
-2.46938899e-01 2.93879777e-01 -8.07307422e-01 7.38312155e-02
2.80642211e-01 -6.54424503e-02 1.27064884e+00 -9.55517352e-01
4.80682850e-01 1.02403164e+00 5.80169857e-01 4.24908549e-01
-1.11028349e+00 -4.34110671e-01 5.98298073e-01 -1.80345356e-01
-1.11926401e+00 -2.88373500e-01 4.63632971e-01 -2.38532156e-01
4.94392991e-01 4.82203752e-01 4.47861463e-01 3.73776197e-01
-6.99400753e-02 8.53477240e-01 9.28251326e-01 -3.80112529e-01
2.87011236e-01 3.09230059e-01 2.37320364e-01 8.92269969e-01
3.28943700e-01 1.74198806e-01 -4.81059551e-01 -2.98659772e-01
3.46309721e-01 2.06263229e-01 -2.87681997e-01 -4.61015105e-01
-6.23863101e-01 1.03669405e+00 7.49457181e-02 -3.22849989e-01
-2.75224984e-01 3.04802001e-01 4.06013727e-01 5.24115741e-01
7.20070302e-01 -7.62816370e-02 -5.58276951e-01 -2.41761282e-01
-1.05245245e+00 2.86127418e-01 1.20822430e+00 1.37405944e+00
5.46002746e-01 1.10236160e-01 -4.06190932e-01 5.92703402e-01
1.81875229e-01 9.06998813e-01 -1.62157819e-01 -1.02083409e+00
1.14023483e+00 1.92266867e-01 7.13865459e-01 -7.24404991e-01
-2.34648272e-01 -7.13242531e-01 -3.69961739e-01 1.79225728e-01
5.95357597e-01 -7.33662128e-01 -3.41785848e-01 1.74924517e+00
4.46283251e-01 -4.19537216e-01 -1.62583098e-01 9.29326355e-01
-2.01334924e-01 6.29307389e-01 -5.41417897e-01 -8.26927483e-01
7.50560224e-01 -1.04866350e+00 -6.28894091e-01 -3.02529693e-01
6.96424425e-01 -4.68985677e-01 1.12043250e+00 3.52390915e-01
-1.49103963e+00 3.57641637e-01 -8.93561065e-01 9.89079401e-02
2.67208427e-01 -1.82478037e-02 8.15259099e-01 9.41599846e-01
-7.94130743e-01 6.10825181e-01 -7.21867263e-01 1.71561286e-01
2.66479194e-01 4.24676061e-01 -3.61853912e-02 -2.10825861e-01
-4.78806615e-01 5.31922758e-01 1.23198830e-01 4.16363358e-01
-9.04161692e-01 -9.82099593e-01 -5.31981170e-01 2.22909093e-01
7.55579174e-01 -5.15900850e-01 1.08950222e+00 -7.58478343e-01
-1.67590177e+00 7.21479237e-01 -2.91668415e-01 -4.00687695e-01
1.14721608e+00 -3.59289944e-01 2.95537561e-01 -2.21669108e-01
5.97850010e-02 -4.57661927e-01 6.79399669e-01 -8.48328769e-01
-6.89546347e-01 -6.47374332e-01 4.68892187e-01 5.61302662e-01
-5.14723420e-01 1.80215359e-01 -5.49316525e-01 -2.53726274e-01
-5.00425920e-02 -1.04626441e+00 -6.11540198e-01 4.08992231e-01
-2.12890387e-01 3.99071395e-01 9.93709713e-02 -6.83525324e-01
1.24906552e+00 -2.37418866e+00 8.38936567e-02 3.64402860e-01
-1.25951409e-01 -1.00107998e-01 1.67052537e-01 3.53679180e-01
2.00729981e-01 1.19026490e-01 -4.16818768e-01 -6.99566126e-01
4.35677499e-01 7.55503923e-02 -4.31010425e-01 8.84680331e-01
-5.94372451e-01 5.51580131e-01 -7.92176485e-01 -1.01843290e-01
-1.25476057e-02 -1.81458324e-01 -8.58242869e-01 9.61159822e-03
-8.86310637e-02 3.64871398e-02 -4.65115190e-01 4.20265794e-01
9.74395752e-01 -2.62942076e-01 4.53567892e-01 4.31145370e-01
-1.59524888e-01 7.67844729e-03 -1.52642655e+00 1.49018407e+00
-9.72359240e-01 4.64089513e-01 9.61270690e-01 -1.03762448e+00
5.05889058e-01 1.18920170e-01 5.71772754e-01 -4.43001121e-01
9.25246477e-02 3.01026702e-01 -6.31137133e-01 -1.41973644e-01
2.70580441e-01 -5.03585279e-01 5.49607612e-02 6.05240881e-01
-2.52796680e-01 3.15053701e-01 2.25626886e-01 1.55199111e-01
9.47519481e-01 -2.04357773e-01 -1.22518554e-01 -3.58289719e-01
3.78552854e-01 -3.10760796e-01 8.76177549e-01 9.67932761e-01
-2.19694927e-01 2.82646984e-01 8.43750834e-01 -2.15789229e-01
-9.54256237e-01 -1.12026417e+00 -8.47668797e-02 1.53254950e+00
9.08300951e-02 -1.21324107e-01 -2.84844697e-01 -6.40981495e-01
4.19877648e-01 6.07047498e-01 -4.67491567e-01 1.98235244e-01
-2.85922825e-01 -1.15149474e+00 -9.27384645e-02 2.69061446e-01
1.57480791e-01 -2.22234458e-01 -8.57324451e-02 2.54331052e-01
1.06049009e-01 -1.02805519e+00 -1.00007606e+00 1.49147913e-01
-1.07659984e+00 -8.69985223e-01 -6.45511568e-01 -4.56439584e-01
1.05585706e+00 2.30230719e-01 5.94611585e-01 -2.87432760e-01
-1.00071236e-01 5.89404166e-01 3.75019722e-02 -4.72043037e-01
4.01000082e-01 -2.84401834e-01 1.07083552e-01 2.33585149e-01
-2.90120542e-01 -4.08594042e-01 -6.12696171e-01 1.16384380e-01
-6.10849559e-01 -3.11027616e-01 4.54953276e-02 8.76632988e-01
6.87847257e-01 -1.45375878e-01 2.20933184e-01 -1.01696539e+00
6.93843842e-01 -5.17164111e-01 -1.38376629e+00 3.41704249e-01
-9.95701969e-01 1.87902644e-01 9.04688478e-01 -1.97480932e-01
-1.01993608e+00 4.69595492e-02 7.99077600e-02 -4.94817287e-01
7.89095283e-01 6.02587998e-01 -5.08370548e-02 -3.82471412e-01
4.90951777e-01 4.04573351e-01 -1.58706293e-01 -5.87842464e-01
4.59813505e-01 3.42044562e-01 1.78459451e-01 -9.04238284e-01
8.72980714e-01 9.17626858e-01 1.38608664e-01 -4.18652952e-01
-1.45373964e+00 -5.03766656e-01 -1.72099739e-01 -1.02505237e-02
2.66986996e-01 -8.25845063e-01 -1.13204896e+00 -1.33510813e-01
-8.19143653e-01 -2.91787028e-01 -5.80667198e-01 6.80161834e-01
-9.36724603e-01 4.51778382e-01 -4.24120992e-01 -1.55393136e+00
-4.39781040e-01 -8.17159712e-01 3.68998438e-01 4.43804637e-02
5.49895167e-01 -1.23262703e+00 -8.29768628e-02 3.25930953e-01
4.83612895e-01 2.75894642e-01 3.81072611e-01 -1.70498028e-01
-5.34386635e-01 -3.99632543e-01 -3.85406047e-01 5.29398978e-01
-1.69416055e-01 -3.53469938e-01 -3.40581983e-01 -7.01725543e-01
3.55553746e-01 -2.03999475e-01 5.72182178e-01 4.83165950e-01
1.19677103e+00 -1.03621018e+00 -4.87885736e-02 1.18293381e+00
1.78881741e+00 -2.12529629e-01 4.25099432e-02 2.02037811e-01
3.35510999e-01 2.40316078e-01 7.26054251e-01 1.21648884e+00
1.32951334e-01 3.35274339e-01 4.17463660e-01 1.10875741e-01
5.88897824e-01 -1.87709183e-01 6.07384145e-01 6.59821033e-01
1.93312600e-01 2.66749021e-02 -5.78078032e-01 7.23373532e-01
-2.15032220e+00 -8.23135734e-01 7.77094811e-02 2.67774034e+00
9.31586623e-01 8.14257637e-02 1.92631841e-01 -2.30446026e-01
6.31986439e-01 1.17985077e-01 -8.42345238e-01 -8.64039063e-01
-8.38583931e-02 2.62573987e-01 1.31947279e+00 8.36337149e-01
-7.57030070e-01 6.40180707e-01 7.23503876e+00 8.20620835e-01
-8.18476379e-01 3.90834332e-01 4.49869722e-01 -9.24867034e-01
-5.05152404e-01 2.35133588e-01 -7.74270236e-01 5.99541128e-01
9.13469911e-01 -7.89617419e-01 9.55056608e-01 1.17491817e+00
5.62439740e-01 2.38646790e-02 -1.03756428e+00 1.08067596e+00
-2.37814665e-01 -1.17668056e+00 -6.06076300e-01 4.01476771e-01
9.59269226e-01 1.97449350e-03 2.43662119e-01 1.12989031e-01
5.22399247e-01 -6.45450234e-01 6.85394168e-01 1.47965163e-01
9.16410089e-01 -9.75437105e-01 5.05990684e-01 4.84357178e-01
-9.62387204e-01 -5.83732545e-01 -4.94088471e-01 -1.84463099e-01
5.74687660e-01 9.41500604e-01 -2.75768429e-01 5.37799478e-01
4.53349054e-01 2.80511260e-01 3.30293387e-01 1.19437981e+00
-2.74893552e-01 5.82112491e-01 -8.23811471e-01 -2.01533511e-02
2.46759936e-01 -8.39435816e-01 7.40438282e-01 1.18243754e+00
2.26649135e-01 2.23667994e-01 3.88444453e-01 5.49538255e-01
-2.68705130e-01 5.01940668e-01 -2.94934928e-01 -3.28819491e-02
2.67680854e-01 9.36689675e-01 -2.44577006e-01 -2.85261218e-02
-4.93142486e-01 1.14489722e+00 5.81075847e-01 4.82000709e-01
-9.57584321e-01 -3.34721327e-01 8.51248085e-01 -7.06216246e-02
5.64202070e-01 -5.64574003e-01 -5.72270989e-01 -1.25921786e+00
7.73347914e-01 -2.81485021e-01 5.64563692e-01 5.34290001e-02
-1.19320118e+00 -1.34095907e-01 -1.89870194e-01 -7.86576927e-01
3.05642616e-02 -5.90874732e-01 -6.36670053e-01 8.23485136e-01
-1.53383708e+00 -6.44583702e-01 3.38415593e-01 7.32492805e-01
1.20719701e-01 5.50759844e-02 3.88649970e-01 5.14041543e-01
-7.36371517e-01 1.00034440e+00 9.52981889e-01 -1.58312187e-01
2.44576767e-01 -1.31790924e+00 -3.62489611e-01 9.72108066e-01
-1.68345943e-01 4.44412798e-01 7.11551607e-01 -1.39271617e-01
-1.95470011e+00 -9.21506941e-01 6.58220410e-01 -9.17186365e-02
1.05343688e+00 -4.55543369e-01 -2.01473445e-01 7.41671920e-01
-3.16388696e-01 3.99095982e-01 9.19742167e-01 4.87550229e-01
-2.64939904e-01 -4.09101665e-01 -1.40392828e+00 4.59816635e-01
1.25488412e+00 -6.37909710e-01 1.93371952e-01 1.00887156e+00
6.25748038e-01 -7.09555864e-01 -8.30812216e-01 -5.87881953e-02
5.86379051e-01 -6.75632060e-01 7.09199309e-01 -9.75269318e-01
-4.71041426e-02 9.79364514e-02 -4.18947011e-01 -8.47716808e-01
-6.22312613e-02 -1.27796388e+00 -4.31083262e-01 6.53205514e-01
7.31332481e-01 -8.23300302e-01 9.49788570e-01 1.04601991e+00
-1.50470715e-02 -8.91147733e-01 -1.28707993e+00 -1.29270399e+00
2.72267461e-01 -5.70095062e-01 9.20277834e-02 4.70829934e-01
2.66458333e-01 -1.46012738e-01 -7.34490395e-01 1.65426120e-01
8.49742889e-01 3.88035536e-01 5.58113396e-01 -5.07720232e-01
-8.26057673e-01 -1.82432607e-01 2.08422184e-01 -1.29333401e+00
9.16282609e-02 -1.06126463e+00 -5.47301061e-02 -1.13714409e+00
3.58913779e-01 -6.15673602e-01 -5.18711269e-01 2.66273230e-01
-9.60178822e-02 -3.21234256e-01 5.01031041e-01 -1.34470254e-01
-8.18949699e-01 8.74954104e-01 1.04975522e+00 8.34886283e-02
-3.25695336e-01 3.94990712e-01 -8.69146824e-01 4.49069947e-01
5.03479481e-01 -5.51238000e-01 -2.60308057e-01 -6.48269713e-01
7.46564806e-01 4.78927672e-01 -2.17093080e-01 -4.49001610e-01
1.88722894e-01 -5.98028362e-01 -3.23442757e-01 -3.44857037e-01
2.17778966e-01 -7.38825679e-01 -1.50600851e-01 4.70583528e-01
-3.22366506e-01 -1.40021399e-01 5.31302020e-02 9.13138568e-01
1.49193555e-01 -4.11495566e-01 8.57276678e-01 -4.48849835e-02
2.08286420e-01 8.34992290e-01 -1.23286732e-01 5.03713131e-01
1.06506646e+00 5.61882891e-02 5.56587391e-02 -7.12847769e-01
-9.77616370e-01 7.15006173e-01 1.34558961e-01 -2.42711827e-01
4.47353065e-01 -1.20048153e+00 -7.64656425e-01 -3.06740135e-01
-3.70428354e-01 -7.57142156e-02 2.37929806e-01 9.55703855e-01
-4.54487890e-01 3.71728778e-01 4.63849396e-01 -9.37081948e-02
-9.40826893e-01 6.64313376e-01 3.19540501e-01 -5.23521841e-01
-3.92869115e-01 1.00694168e+00 -3.46253067e-02 -2.98934877e-01
4.44583267e-01 1.00268289e-01 9.17337060e-01 -2.19917551e-01
4.50244188e-01 7.01502204e-01 -1.25133321e-01 1.37803033e-01
-3.45145702e-01 1.09092616e-01 -3.98778021e-02 -5.82601070e-01
1.50900602e+00 -4.73850846e-01 -5.43131791e-02 2.18938142e-01
1.46484017e+00 2.68340558e-01 -1.49859190e+00 -4.31373924e-01
-2.20331866e-02 -8.67669880e-01 8.35994184e-02 -5.28923810e-01
-1.17669094e+00 6.17812574e-01 5.00188231e-01 4.61248048e-02
1.10117459e+00 -4.40751493e-01 7.85507321e-01 4.02556986e-01
6.41861796e-01 -1.54214740e+00 -4.22793806e-01 8.13922048e-01
7.51765609e-01 -1.22758436e+00 1.79432213e-01 -4.36216205e-01
-3.23217928e-01 8.98513317e-01 5.33781946e-02 -4.17856991e-01
8.90133083e-01 1.40089616e-01 -5.83053231e-01 1.99770197e-01
-7.95585930e-01 -1.24928549e-01 6.92430884e-02 8.03965852e-02
1.96807340e-01 4.01405156e-01 -9.70404148e-01 5.74982584e-01
-9.32878330e-02 -8.80751684e-02 4.94529545e-01 9.83493745e-01
-4.39233899e-01 -1.12674141e+00 -1.23105496e-01 5.18557072e-01
-8.63187611e-01 -2.80765772e-01 -5.45924082e-02 3.43572408e-01
-4.85638857e-01 8.60595882e-01 -2.48956874e-01 1.65067211e-01
1.30686596e-01 -6.40433580e-02 6.74159050e-01 -5.98689556e-01
-5.77456772e-01 -6.51972592e-02 1.61052823e-01 -7.87804604e-01
2.56469995e-01 -6.78980112e-01 -1.19370270e+00 -6.97510600e-01
-4.97635871e-01 2.89058208e-01 9.41966236e-01 9.33824420e-01
3.03624481e-01 -6.58411682e-02 1.39055967e+00 -1.65171057e-01
-1.22458160e+00 -4.07884926e-01 -8.55285466e-01 2.18091160e-01
3.55670303e-01 -2.05191776e-01 -6.36006832e-01 -4.37034130e-01] | [6.152957439422607, 4.283018112182617] |
a5ed8bea-29b3-4e50-85d5-b9fe4ad4df91 | machine-learning-for-mention-head-detection | null | null | https://aclanthology.org/R13-1097 | https://aclanthology.org/R13-1097.pdf | Machine Learning for Mention Head Detection in Multilingual Coreference Resolution | null | ['ra', 'S K{\\"u}bler', 'Desislava Zhekova'] | 2013-09-01 | machine-learning-for-mention-head-detection-1 | https://aclanthology.org/R13-1097 | https://aclanthology.org/R13-1097.pdf | ranlp-2013-9 | ['head-detection'] | ['computer-vision'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.250907897949219, 3.764009952545166] |
c9c88ed8-f75f-4a9e-a22a-bc802d3845ce | structured-scene-memory-for-vision-language | 2103.03454 | null | https://arxiv.org/abs/2103.03454v1 | https://arxiv.org/pdf/2103.03454v1.pdf | Structured Scene Memory for Vision-Language Navigation | Recently, numerous algorithms have been developed to tackle the problem of vision-language navigation (VLN), i.e., entailing an agent to navigate 3D environments through following linguistic instructions. However, current VLN agents simply store their past experiences/observations as latent states in recurrent networks, failing to capture environment layouts and make long-term planning. To address these limitations, we propose a crucial architecture, called Structured Scene Memory (SSM). It is compartmentalized enough to accurately memorize the percepts during navigation. It also serves as a structured scene representation, which captures and disentangles visual and geometric cues in the environment. SSM has a collect-read controller that adaptively collects information for supporting current decision making and mimics iterative algorithms for long-range reasoning. As SSM provides a complete action space, i.e., all the navigable places on the map, a frontier-exploration based navigation decision making strategy is introduced to enable efficient and global planning. Experiment results on two VLN datasets (i.e., R2R and R4R) show that our method achieves state-of-the-art performance on several metrics. | ['Jianbing Shen', 'Caiming Xiong', 'Wei Liang', 'Wenguan Wang', 'Hanqing Wang'] | 2021-03-05 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Wang_Structured_Scene_Memory_for_Vision-Language_Navigation_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Wang_Structured_Scene_Memory_for_Vision-Language_Navigation_CVPR_2021_paper.pdf | cvpr-2021-1 | ['vision-language-navigation'] | ['computer-vision'] | [ 1.02890216e-01 -1.02452040e-01 -1.20201051e-01 -4.35744673e-01
-6.96303621e-02 -4.22754824e-01 8.61225843e-01 4.64842059e-02
-5.49626291e-01 3.12877566e-01 4.98929024e-01 -5.00942349e-01
-2.64667064e-01 -1.05958521e+00 -6.81975961e-01 -5.76107919e-01
-1.76824555e-01 5.39942086e-01 2.23270401e-01 -4.91307914e-01
5.53010762e-01 5.39874256e-01 -1.56709468e+00 9.05369967e-03
8.37607980e-01 7.70211875e-01 1.13100755e+00 6.01996601e-01
-4.13740963e-01 1.22448003e+00 -2.47752845e-01 3.72983426e-01
1.40290037e-01 -3.00245970e-01 -8.93686116e-01 -7.17471018e-02
7.35792890e-02 -3.31706226e-01 -7.60635972e-01 9.62702394e-01
2.31905416e-01 7.14219093e-01 4.54122782e-01 -8.82363021e-01
-7.71001577e-01 5.13139486e-01 -1.62658483e-01 1.27117619e-01
6.06909513e-01 5.39567947e-01 7.69504726e-01 -7.89771974e-01
7.08656669e-01 1.39193153e+00 1.52142391e-01 4.61116940e-01
-1.01999319e+00 -1.23092063e-01 7.44927049e-01 4.19782102e-01
-1.17440140e+00 -5.13217509e-01 8.11703086e-01 -2.70871371e-01
1.38839841e+00 5.19990176e-02 1.00449336e+00 1.25275421e+00
4.87381130e-01 9.69212770e-01 9.94539380e-01 -2.34215319e-01
6.33659542e-01 -4.09996629e-01 4.19594869e-02 1.01065218e+00
-3.19866650e-02 2.47890398e-01 -9.30039048e-01 2.16028407e-01
1.23362255e+00 2.60174662e-01 -3.64808530e-01 -6.83395684e-01
-1.47539055e+00 6.06063545e-01 8.76728237e-01 8.56579244e-02
-7.26874530e-01 1.18570447e-01 1.71506986e-01 6.73038885e-02
-2.97403246e-01 5.81467330e-01 -6.18293472e-02 -1.41183525e-01
-5.14187574e-01 1.75249428e-01 6.19087458e-01 9.99970257e-01
9.18881178e-01 1.10901862e-01 -1.76066831e-01 5.62855959e-01
4.61895525e-01 6.11418426e-01 5.28079569e-01 -1.30705655e+00
5.72662652e-01 7.79471159e-01 2.77825952e-01 -1.19350231e+00
-7.00941026e-01 -4.12284315e-01 -9.87572014e-01 2.68599421e-01
2.18604226e-02 2.72976696e-01 -1.37166476e+00 1.80358732e+00
-5.05965427e-02 2.04683580e-02 3.11327428e-01 1.07957590e+00
7.75095999e-01 8.53241384e-01 3.86532731e-02 1.15892418e-01
1.03088081e+00 -1.45737827e+00 -6.56709135e-01 -8.91594827e-01
2.57564217e-01 -1.14018295e-03 1.15176713e+00 2.45346919e-01
-1.03763855e+00 -7.42507756e-01 -9.93184686e-01 -2.61998326e-01
-5.32005489e-01 -8.42988193e-02 8.33984077e-01 -6.28464520e-02
-1.57587004e+00 2.61819154e-01 -1.08691442e+00 -4.47906792e-01
1.46223903e-01 -1.92832835e-02 -3.28999341e-01 -1.77795842e-01
-1.06655860e+00 1.02265036e+00 5.19347131e-01 6.75058722e-01
-1.37670398e+00 9.08802263e-03 -1.23918486e+00 -4.22270186e-02
6.66038215e-01 -9.26561892e-01 1.16578639e+00 -2.19097793e-01
-1.71367037e+00 5.85978270e-01 -5.94310164e-01 -4.94721800e-01
1.91605136e-01 -2.85650045e-01 -1.56071961e-01 -3.75473946e-02
2.61628717e-01 9.29312348e-01 3.57261002e-01 -1.33682513e+00
-6.83812320e-01 -4.36863869e-01 3.62890929e-01 7.09978819e-01
3.42042267e-01 -6.44191384e-01 -7.37909794e-01 -2.39590034e-01
9.19202983e-01 -9.45235908e-01 -9.01589036e-01 -1.11969158e-01
-4.65632319e-01 -5.10537885e-02 3.36065382e-01 -4.73695815e-01
1.08182430e+00 -1.80948746e+00 4.95815963e-01 2.63628513e-01
2.31765226e-01 -6.18273020e-02 -3.47669870e-01 6.29140377e-01
4.54211861e-01 -2.75761247e-01 2.65812613e-02 -6.10388935e-01
8.87148529e-02 7.05788612e-01 -6.93210423e-01 2.07568645e-01
-3.39858264e-01 1.38964140e+00 -1.12065601e+00 -1.16056271e-01
4.50749934e-01 2.61713713e-01 -4.38277453e-01 1.42708555e-01
-5.85668921e-01 5.86532593e-01 -7.42528737e-01 4.83334869e-01
3.77244383e-01 -4.64803904e-01 2.05606982e-01 3.05742264e-01
-5.80884695e-01 5.75315356e-01 -1.01213789e+00 2.55331349e+00
-6.73152804e-01 5.48315525e-01 -4.66262139e-02 -5.32283485e-01
1.01596344e+00 -2.91000128e-01 -1.72198594e-01 -1.33646810e+00
-1.17827006e-01 5.10978922e-02 -4.38183367e-01 -2.99221069e-01
8.20507050e-01 5.96508980e-01 -2.28971660e-01 2.78327793e-01
-2.16686562e-01 1.84427291e-01 -4.69969362e-02 2.04273194e-01
1.05883634e+00 4.15615976e-01 3.63336921e-01 -1.27926826e-01
4.47080255e-01 1.98936552e-01 4.94861782e-01 1.31021702e+00
-1.46338403e-01 1.60129055e-01 -1.70464814e-02 -7.94092357e-01
-5.43119431e-01 -1.27675700e+00 6.13291025e-01 1.16456318e+00
7.53914356e-01 -4.14081156e-01 -4.02153254e-01 -6.05366789e-02
-4.77310807e-01 1.06764793e+00 -6.62482977e-01 -2.19797850e-01
-6.84250057e-01 -1.68672517e-01 8.96530971e-02 6.29446447e-01
8.59603524e-01 -1.59924340e+00 -1.56143224e+00 3.24482411e-01
-4.58028704e-01 -8.11100006e-01 -2.66739219e-01 4.02121007e-01
-7.53143370e-01 -8.27674329e-01 -2.63158947e-01 -9.61174309e-01
7.21157789e-01 6.94864929e-01 1.20479929e+00 -5.31284921e-02
7.39295706e-02 3.05135757e-01 -2.85463005e-01 5.27638569e-02
-4.19111773e-02 -5.38558874e-04 1.29970536e-01 -2.87989318e-01
1.25012890e-01 -6.41790509e-01 -7.52023399e-01 3.41411352e-01
-4.36052382e-01 7.16995656e-01 7.72493660e-01 6.80449009e-01
9.95291710e-01 1.32907063e-01 6.27188832e-02 -4.15512979e-01
6.12828791e-01 -1.53822646e-01 -7.37214863e-01 4.14403886e-01
-4.63501096e-01 3.66677731e-01 5.97717047e-01 -1.04768395e-01
-1.15822661e+00 -2.74352003e-02 -8.98172855e-02 -2.88630217e-01
-4.79846299e-01 7.79732049e-01 -2.53489316e-01 1.50084794e-01
3.97724897e-01 8.99083376e-01 -2.65766799e-01 -4.19226944e-01
4.79743361e-01 1.20348923e-01 8.58944237e-01 -3.20606858e-01
3.84626091e-01 6.04339421e-01 -3.07841157e-03 -6.61122203e-01
-7.76350796e-01 -3.42388839e-01 -6.61974549e-01 -1.60327435e-01
9.04793024e-01 -8.86934638e-01 -9.32701588e-01 6.60522938e-01
-1.11874914e+00 -8.85432482e-01 -2.05061793e-01 3.18005770e-01
-9.16194260e-01 -5.64720593e-02 -4.95905131e-01 -9.19483721e-01
-4.69656736e-02 -1.21485388e+00 8.67852449e-01 3.72106850e-01
-1.80419475e-01 -9.56104457e-01 1.74644366e-01 -1.10796236e-01
5.20197809e-01 1.47987679e-01 8.66690338e-01 -6.29521757e-02
-1.18499601e+00 3.92727494e-01 -1.00911923e-01 -6.00847781e-01
6.81495070e-02 -6.52069330e-01 -6.57381415e-01 -2.02674106e-01
2.82371808e-02 -1.31137416e-01 1.16078377e+00 4.62811083e-01
1.00903916e+00 -2.97047645e-01 -5.96650243e-01 6.94668710e-01
1.33111787e+00 6.41233981e-01 6.34978473e-01 6.79942906e-01
6.28436089e-01 6.63663268e-01 6.35061264e-01 3.48674744e-01
1.00828946e+00 5.68795979e-01 7.50459433e-01 2.42390096e-01
1.95111446e-02 -8.95423949e-01 4.07476008e-01 6.65424168e-01
1.02379262e-01 -4.16850001e-01 -1.21604037e+00 3.10668200e-01
-2.23226833e+00 -9.26556230e-01 3.03662747e-01 1.79836690e+00
3.44318181e-01 3.47784489e-01 -4.63012934e-01 -4.07652467e-01
1.84135646e-01 6.76589489e-01 -1.09190428e+00 -2.02229202e-01
-1.33432716e-01 -3.89142454e-01 3.95541549e-01 8.38714004e-01
-8.25444520e-01 1.35888195e+00 6.03072309e+00 4.17431146e-01
-9.45110261e-01 -2.70859808e-01 2.42345795e-01 -2.32062656e-02
-4.32079703e-01 -4.94096763e-02 -6.74768984e-01 6.55911714e-02
6.62225544e-01 1.22171342e-01 9.89153087e-01 7.91894078e-01
3.33128244e-01 -5.62943995e-01 -9.69940960e-01 1.16025841e+00
-7.47712515e-03 -1.53342319e+00 2.79178649e-01 2.91075986e-02
5.24243355e-01 3.81829351e-01 2.88851529e-01 5.15176237e-01
6.46402001e-01 -1.21001923e+00 1.05498874e+00 9.97711420e-01
5.19323468e-01 -6.79534197e-01 3.51038963e-01 9.75457311e-01
-1.39436817e+00 -3.72073412e-01 -4.09644783e-01 -3.68293792e-01
5.45114219e-01 -5.19511523e-03 -5.17878354e-01 5.25408864e-01
7.09266543e-01 7.65489399e-01 -6.39853656e-01 8.86589944e-01
-4.69082117e-01 -6.39768392e-02 -1.84889227e-01 -2.28080228e-01
7.69034207e-01 -4.86970514e-01 6.20789468e-01 6.56398594e-01
2.22631261e-01 1.06507443e-01 5.26437521e-01 9.44757700e-01
4.74846065e-01 -3.90528142e-01 -7.38222122e-01 2.63865501e-01
5.77003360e-01 6.44040704e-01 -9.27602112e-01 -2.09805802e-01
-5.93100162e-03 1.08731973e+00 6.75006211e-01 7.80349791e-01
-5.11806667e-01 -8.50940123e-02 6.64256513e-01 -2.58056819e-01
2.44979575e-01 -8.84138644e-01 -2.23461241e-01 -9.71358836e-01
-3.87801528e-02 -4.90058452e-01 -1.18964091e-02 -1.04564369e+00
-5.93070269e-01 8.89026999e-01 -1.69899032e-01 -8.66806746e-01
-4.63555038e-01 -4.60175186e-01 -6.21904314e-01 9.89013195e-01
-1.68911731e+00 -1.01657224e+00 -5.66311359e-01 5.64878702e-01
8.12534392e-01 -1.63214087e-01 9.56319809e-01 -4.43530232e-01
-4.55747873e-01 -4.80321273e-02 -2.04800755e-01 -6.07000552e-02
-1.16489278e-02 -1.11083984e+00 9.30326164e-01 9.77168739e-01
3.88581336e-01 1.07809484e+00 6.73600435e-01 -8.29542041e-01
-1.79098701e+00 -9.26802635e-01 6.98584080e-01 -4.33552027e-01
3.23735327e-01 -5.20167947e-01 -8.14861834e-01 8.25759470e-01
6.08151406e-02 -4.37279880e-01 2.87385166e-01 1.56485796e-01
-2.31022149e-01 2.52204657e-01 -5.69729507e-01 1.19618821e+00
1.79648149e+00 -6.36685312e-01 -7.79292822e-01 4.58635204e-02
9.49297249e-01 -7.38119602e-01 7.92442411e-02 8.92556533e-02
3.98798406e-01 -1.24840117e+00 9.72346604e-01 -3.13994408e-01
9.92981493e-02 -5.71794093e-01 -4.08198982e-01 -1.34926748e+00
-7.09967613e-01 -6.05366230e-01 -4.52104777e-01 5.46550691e-01
3.11271727e-01 -7.12780356e-01 8.10079992e-01 4.99081075e-01
-3.08844656e-01 -8.81237745e-01 -8.82973671e-01 -5.33904552e-01
-6.68363631e-01 -5.81342578e-01 6.79133177e-01 3.77082735e-01
-4.11614448e-01 3.57588738e-01 -2.13295609e-01 4.74354535e-01
5.38403511e-01 5.90087235e-01 5.77376723e-01 -9.53646839e-01
4.68341745e-02 -5.88281274e-01 -1.47388160e-01 -2.02000642e+00
3.32812041e-01 -7.91884542e-01 3.59609455e-01 -2.35131121e+00
-8.81651416e-02 -4.57736939e-01 -3.73525947e-01 5.08233309e-01
2.80352801e-01 -4.97354537e-01 2.40607649e-01 3.70356411e-01
-1.11770666e+00 7.92726994e-01 1.46049047e+00 -5.41475117e-02
-7.03325093e-01 -1.34175658e-01 -7.10702360e-01 8.53424847e-01
6.71723187e-01 1.08183995e-01 -7.98196137e-01 -8.67987514e-01
4.80941713e-01 3.29777122e-01 6.67934179e-01 -1.11999488e+00
8.97201717e-01 -3.64892423e-01 3.90612096e-01 -1.23420429e+00
6.10377669e-01 -6.98115468e-01 1.03295743e-01 4.98931438e-01
-5.03385365e-01 3.63924086e-01 -4.77059819e-02 9.54072535e-01
-1.83389112e-01 1.08079568e-01 1.48715705e-01 -5.20002067e-01
-1.75076175e+00 4.06666994e-01 -5.96577525e-01 -1.83129057e-01
6.65556729e-01 -4.46236789e-01 -2.78052837e-01 -4.98539835e-01
-6.41612530e-01 9.52159822e-01 6.56331778e-01 6.73639059e-01
1.12109935e+00 -1.33293974e+00 -1.54446065e-01 5.08048952e-01
7.58887827e-02 5.47296703e-01 6.50873363e-01 3.86378109e-01
-5.15047371e-01 8.65498900e-01 -3.34266692e-01 -5.51285744e-01
-5.80718815e-01 6.69677675e-01 3.67114335e-01 -3.36565822e-01
-9.78170156e-01 9.58267987e-01 4.62949514e-01 -6.18481874e-01
4.61099595e-01 -5.31016171e-01 -5.59568226e-01 -1.19418874e-01
7.00499475e-01 2.44449854e-01 -3.82561803e-01 -4.96615261e-01
-3.86663735e-01 5.08693576e-01 4.69887443e-02 -3.73405755e-01
1.14453316e+00 -6.40979409e-01 -2.14340910e-01 7.72084296e-01
6.44068360e-01 -2.92582631e-01 -1.64456511e+00 -3.99435073e-01
5.39191365e-02 -3.11297417e-01 9.95891318e-02 -9.59279299e-01
-5.91673315e-01 8.64965916e-01 2.74829715e-01 -7.54426867e-02
1.04142308e+00 -2.13537775e-02 6.43744648e-01 1.05888963e+00
1.08177125e+00 -9.64788496e-01 1.86120257e-01 1.22774231e+00
1.06316149e+00 -1.02696061e+00 -4.60434467e-01 1.21823326e-01
-7.68017650e-01 8.58983815e-01 8.06726336e-01 9.18811336e-02
2.82060117e-01 -1.68473035e-01 8.78457800e-02 -3.76230806e-01
-9.25624192e-01 -2.84227699e-01 1.66536108e-01 5.86030722e-01
-3.33993554e-01 9.78158191e-02 6.29185975e-01 4.71839488e-01
-4.93019521e-01 -2.68744916e-01 2.72740990e-01 1.13114369e+00
-7.99799621e-01 -5.02419055e-01 -3.69429812e-02 2.20787451e-01
6.09840512e-01 1.15697328e-02 -1.25238061e-01 4.63554502e-01
-1.17040940e-01 9.16941106e-01 1.50462836e-01 -4.31901842e-01
4.02672917e-01 -1.83855802e-01 2.74999410e-01 -6.22304022e-01
2.62498558e-02 -2.36838207e-01 -1.93712607e-01 -1.27857459e+00
-2.12565467e-01 -4.98533905e-01 -1.62966526e+00 -2.58341804e-02
2.82154381e-01 9.17830039e-03 5.04458666e-01 9.28786278e-01
5.00492811e-01 7.86861837e-01 1.97886690e-01 -1.12903595e+00
-2.79986709e-01 -4.84845072e-01 -3.16899419e-01 -1.44063994e-01
7.04200685e-01 -8.84580612e-01 9.51181278e-02 -4.83561337e-01] | [4.5190958976745605, 0.5307801365852356] |
60b5ae3e-6b8b-4287-88dc-ed05acde9e6c | preservation-of-the-global-knowledge-by-not | 2106.03097 | null | https://arxiv.org/abs/2106.03097v5 | https://arxiv.org/pdf/2106.03097v5.pdf | Preservation of the Global Knowledge by Not-True Distillation in Federated Learning | In federated learning, a strong global model is collaboratively learned by aggregating clients' locally trained models. Although this precludes the need to access clients' data directly, the global model's convergence often suffers from data heterogeneity. This study starts from an analogy to continual learning and suggests that forgetting could be the bottleneck of federated learning. We observe that the global model forgets the knowledge from previous rounds, and the local training induces forgetting the knowledge outside of the local distribution. Based on our findings, we hypothesize that tackling down forgetting will relieve the data heterogeneity problem. To this end, we propose a novel and effective algorithm, Federated Not-True Distillation (FedNTD), which preserves the global perspective on locally available data only for the not-true classes. In the experiments, FedNTD shows state-of-the-art performance on various setups without compromising data privacy or incurring additional communication costs. | ['Sangmin Bae', 'Yongjin Shin', 'Se-Young Yun', 'Minchan Jeong', 'Gihun Lee'] | 2021-06-06 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [-4.88339424e-01 2.67325699e-01 -3.62654239e-01 -6.02196813e-01
-9.38736141e-01 -7.18425393e-01 4.78487343e-01 -1.02232508e-01
-3.54145736e-01 1.01543069e+00 2.49676019e-01 -2.28671193e-01
-2.75222421e-01 -7.87158310e-01 -8.54542971e-01 -1.08185017e+00
-3.91889885e-02 6.46114707e-01 5.81462309e-02 4.10999656e-01
-2.69887239e-01 5.01021862e-01 -1.43553793e+00 6.17392004e-01
7.24284351e-01 8.35369229e-01 -2.94308633e-01 3.26545596e-01
-3.45826268e-01 1.26845884e+00 -5.61916292e-01 -8.31508756e-01
6.09219849e-01 -2.39784434e-01 -9.02297437e-01 -1.51504427e-01
7.07244694e-01 -7.97362924e-01 -5.53054988e-01 9.19904828e-01
3.21831316e-01 -6.13869261e-03 -7.83454180e-02 -1.56103921e+00
-6.10553026e-01 8.95365596e-01 -1.47430286e-01 -1.26677126e-01
-1.34816661e-01 2.82281816e-01 8.21494639e-01 -7.29994416e-01
8.37265253e-01 9.48629797e-01 7.82508492e-01 6.94025040e-01
-1.37453139e+00 -7.84634531e-01 5.04129112e-01 9.52669680e-02
-1.24736202e+00 -7.60824144e-01 4.19519484e-01 2.45608967e-02
6.07383072e-01 5.43980181e-01 2.41652712e-01 1.18665779e+00
-8.85419995e-02 9.35844541e-01 1.30106044e+00 -1.29988477e-01
6.34605706e-01 6.21037960e-01 6.30874038e-02 6.45811737e-01
5.36885619e-01 5.61301149e-02 -1.01234734e+00 -8.04201901e-01
3.63779992e-01 7.04968750e-01 -3.41758996e-01 -7.98280060e-01
-5.25998294e-01 7.07515299e-01 2.05885038e-01 1.85671657e-01
-4.43199068e-01 6.02303892e-02 3.53457212e-01 8.39359045e-01
5.64706981e-01 -1.49377704e-01 -9.33886528e-01 -1.77006483e-01
-9.08608556e-01 1.19539141e-03 1.16836500e+00 9.07562852e-01
1.29901516e+00 -4.95003164e-01 -1.39548913e-01 4.75939184e-01
9.96177197e-02 2.70729363e-01 3.28623205e-01 -1.24769557e+00
4.06700879e-01 7.49587834e-01 1.70671567e-01 -3.18553299e-01
1.64893091e-01 -4.54739809e-01 -6.28644764e-01 3.64933461e-01
5.65960228e-01 -4.74317372e-01 -5.31318009e-01 2.09971404e+00
6.91116691e-01 1.02246858e-01 5.84255420e-02 7.38147318e-01
8.74697268e-02 6.29644766e-02 -1.55377612e-01 -4.21727836e-01
7.08659232e-01 -1.19608736e+00 -7.04454362e-01 3.04730773e-01
4.80375469e-01 -2.69248635e-01 9.35914159e-01 5.40326238e-01
-8.16631556e-01 1.15443155e-01 -5.50271273e-01 8.08607973e-03
-3.87338251e-01 -4.38301086e-01 8.89539719e-01 7.99749851e-01
-1.27625811e+00 6.60578012e-01 -1.06585324e+00 -4.76727486e-01
9.12058949e-01 4.01454955e-01 -7.53135383e-01 -4.05783176e-01
-7.37720311e-01 6.10983193e-01 -1.63115174e-01 -4.78684127e-01
-1.24027395e+00 -1.02276361e+00 -7.25460052e-02 3.19440573e-01
5.75278759e-01 -9.59925950e-01 1.47816145e+00 -8.68737400e-01
-1.41626143e+00 5.20667732e-01 -3.35746109e-01 -5.98563969e-01
1.20831943e+00 -2.76542962e-01 -2.66333342e-01 -2.34084696e-01
-3.27902496e-01 -1.67378038e-01 8.39594662e-01 -1.54261649e+00
-9.36011136e-01 -8.59483898e-01 7.50576034e-02 -8.86226818e-02
-7.35424161e-01 -4.79257643e-01 -4.14892972e-01 -1.91146478e-01
-1.45860165e-01 -6.26264274e-01 -9.22153518e-02 5.10086119e-01
-1.11143179e-01 -2.33537361e-01 1.31300235e+00 -3.46401691e-01
1.16647506e+00 -2.16328049e+00 -4.32470709e-01 2.79276401e-01
6.55636549e-01 2.34639272e-01 -1.91746712e-01 7.79387176e-01
4.11858946e-01 2.13539489e-02 9.26981121e-02 -1.06379199e+00
1.65621370e-01 7.13781834e-01 -5.32587767e-01 6.27231836e-01
-7.26002812e-01 6.33137167e-01 -8.51042569e-01 -1.90575346e-01
-1.07634947e-01 5.89494646e-01 -7.78479278e-01 3.57949257e-01
-2.84165770e-01 3.73058170e-01 -2.97486126e-01 5.56839943e-01
1.11626971e+00 -3.84249419e-01 5.77093601e-01 1.94496393e-01
2.46289298e-02 4.03942466e-01 -1.07133138e+00 1.74026215e+00
-4.22414631e-01 1.00050591e-01 8.45028162e-01 -5.36706388e-01
5.26673436e-01 4.64287549e-01 6.38374150e-01 -5.48562467e-01
-2.75965720e-01 2.25768015e-01 -6.63884044e-01 -4.10997212e-01
1.03790006e-02 -1.17457574e-02 4.24925089e-01 1.25787723e+00
2.75300592e-01 8.28149080e-01 -5.38649738e-01 4.17841643e-01
1.42025948e+00 -2.08279133e-01 -1.36395738e-01 -1.19378723e-01
-6.67115720e-03 -3.17342579e-01 7.88956225e-01 1.21757817e+00
-3.45757842e-01 1.11285932e-01 2.91731447e-01 -9.08539772e-01
-5.90864182e-01 -1.22701526e+00 3.07205290e-01 1.59633052e+00
-4.57101800e-02 -6.29363716e-01 -7.42411256e-01 -1.41066980e+00
6.10761821e-01 7.60043144e-01 -7.73742080e-01 -3.69395269e-03
-2.41735473e-01 -4.03928190e-01 5.86158335e-01 6.58914372e-02
4.96473670e-01 -4.18490499e-01 -3.62598002e-01 5.54805025e-02
-4.51280959e-02 -4.28699017e-01 -6.87461913e-01 3.27317148e-01
-9.65680420e-01 -1.29040885e+00 -1.73499554e-01 -1.64696768e-01
6.63150668e-01 6.44334853e-01 9.00385022e-01 1.32342249e-01
-4.38658241e-03 5.89391589e-01 -3.03266905e-02 -3.12018663e-01
-2.35095769e-01 3.01865399e-01 1.54792070e-01 5.05109787e-01
5.84616959e-01 -1.17798746e+00 -8.19987118e-01 2.71608263e-01
-9.96712804e-01 -2.83001870e-01 4.17024106e-01 7.62900114e-01
3.28385025e-01 1.29513964e-01 6.12353861e-01 -1.52777565e+00
5.53319156e-01 -7.60820389e-01 -2.83810139e-01 5.86843193e-01
-1.23059535e+00 9.05395001e-02 8.76206458e-01 -4.57396507e-01
-1.47791672e+00 -1.09414794e-01 3.60875100e-01 -7.47709870e-01
-1.40076861e-01 -4.46891561e-02 -2.64328688e-01 -1.98327228e-01
6.52677357e-01 2.86288977e-01 3.57057631e-01 -1.20316374e+00
7.45842814e-01 7.90154934e-01 3.79135668e-01 -7.60342360e-01
6.12139523e-01 1.07271159e+00 -5.48865318e-01 -1.45091608e-01
-6.70407653e-01 -2.09426731e-01 -2.64169365e-01 3.15843560e-02
-4.77535985e-02 -8.78996849e-01 -1.11472476e+00 4.84939396e-01
-8.80052745e-01 -5.09492218e-01 -8.47200871e-01 1.49472922e-01
-3.10006648e-01 2.10010931e-01 -6.92667723e-01 -8.97240281e-01
-6.45358384e-01 -5.18185318e-01 3.74783814e-01 1.87922791e-01
6.72867298e-02 -9.18170035e-01 4.93122786e-01 4.94223744e-01
1.02438259e+00 -1.19539551e-01 5.91150343e-01 -1.01437986e+00
-9.24480557e-01 -2.68865675e-01 -5.27277477e-02 3.12422901e-01
5.18760979e-01 -3.41155469e-01 -1.47360325e+00 -8.77930522e-01
3.40014577e-01 -3.71753365e-01 8.05031717e-01 -3.36370349e-01
1.16386759e+00 -1.24979055e+00 -2.59745032e-01 6.25442803e-01
1.61300540e+00 -2.15045348e-01 2.58827120e-01 5.95303401e-02
4.73068804e-01 4.06084180e-01 2.92806718e-02 1.05958867e+00
5.89009464e-01 2.41829500e-01 6.59809709e-01 8.60181004e-02
-2.70181656e-01 -7.59041488e-01 3.60933483e-01 4.61572081e-01
2.27408394e-01 -7.32534658e-03 -4.07378674e-01 7.08696842e-01
-2.17279649e+00 -1.04550111e+00 3.22584242e-01 2.47305226e+00
1.13845468e+00 -4.46223289e-01 1.53413221e-01 -3.44290107e-01
4.89663124e-01 1.58828497e-02 -1.05059111e+00 -1.42311081e-01
1.38165420e-02 1.07393309e-01 7.85710752e-01 5.27293265e-01
-6.12859488e-01 6.35871708e-01 6.23945713e+00 6.38863444e-01
-1.12306607e+00 8.02178264e-01 5.67936540e-01 -7.90561736e-01
-6.03243649e-01 2.18351379e-01 -6.39770329e-01 4.85452712e-01
9.25999463e-01 -2.97328085e-01 1.11765468e+00 1.15796101e+00
-1.62781611e-01 2.86590261e-03 -1.32700002e+00 7.40875721e-01
-1.82861716e-01 -1.38460791e+00 7.58887529e-02 3.20096850e-01
9.80793118e-01 5.25064051e-01 1.06513493e-01 3.00158679e-01
1.01085889e+00 -6.49870813e-01 5.09702682e-01 8.27651501e-01
5.48822880e-01 -7.85001516e-01 4.39477146e-01 7.82310009e-01
-5.58906138e-01 -4.40810084e-01 -2.96195239e-01 1.24169312e-01
-2.53471881e-01 6.88314199e-01 -7.78443933e-01 5.24424314e-01
9.03315842e-01 9.09128934e-02 -5.54350674e-01 9.13672149e-01
1.35997072e-01 7.11548388e-01 -6.02839410e-01 5.04509568e-01
-1.07028596e-01 -6.54916242e-02 2.40846708e-01 9.10492241e-01
1.58947617e-01 -1.35377347e-01 6.36997446e-02 7.62026846e-01
-6.98365510e-01 -2.48431135e-02 -5.96204698e-01 2.83781022e-01
8.73646855e-01 1.05796468e+00 1.35982737e-01 -4.64221805e-01
-4.99808341e-01 1.02849758e+00 9.01741683e-01 5.27505696e-01
-4.14621353e-01 -3.06928176e-02 1.23432577e+00 1.97328582e-01
3.38940740e-01 1.69213071e-01 -2.67714888e-01 -1.41736209e+00
2.12359816e-01 -9.09795463e-01 9.38068926e-01 -6.89270869e-02
-1.81743228e+00 3.30059022e-01 -5.51932693e-01 -5.99220634e-01
5.89101240e-02 1.92864522e-01 -5.77843010e-01 6.40916824e-01
-1.46227670e+00 -1.13641572e+00 -1.76516116e-01 1.24240041e+00
8.47870857e-02 -1.10803731e-01 1.04336846e+00 5.65617144e-01
-4.47843760e-01 1.27124393e+00 6.54627621e-01 -2.18093991e-01
1.19381261e+00 -8.32079291e-01 -8.83010402e-02 5.60513437e-01
2.84114569e-01 1.03043413e+00 4.85617071e-01 -4.61237490e-01
-1.73674917e+00 -1.27692461e+00 1.02159369e+00 -6.48903608e-01
3.37842822e-01 -4.23605174e-01 -1.16605449e+00 1.09253871e+00
2.90640324e-01 4.82623875e-01 9.04576480e-01 2.75106192e-01
-1.00915003e+00 -7.35691965e-01 -1.93038976e+00 1.85595557e-01
1.20885408e+00 -8.16339612e-01 1.97189189e-02 3.74824107e-01
7.04992175e-01 2.38541048e-02 -8.33894849e-01 -1.87610492e-01
5.96464694e-01 -1.34244204e+00 4.20769930e-01 -1.10192919e+00
-5.50030529e-01 -3.74616869e-02 -3.16725284e-01 -1.02205110e+00
-2.40335628e-01 -1.20270753e+00 -8.83813441e-01 1.46234238e+00
1.22293867e-01 -1.24786115e+00 1.22003150e+00 1.30971503e+00
4.34990108e-01 -5.22921741e-01 -1.35399079e+00 -8.03046048e-01
3.94708328e-02 6.86417753e-03 1.18414056e+00 1.29252422e+00
4.75180037e-02 -3.06890577e-01 -5.70316613e-01 1.73877254e-01
1.04279768e+00 2.93257177e-01 1.11053669e+00 -1.16594076e+00
-4.68711227e-01 -2.04559811e-03 3.25062722e-01 -8.87606859e-01
-4.11931537e-02 -7.96619534e-01 -3.60727608e-01 -1.17045534e+00
4.58545715e-01 -7.10142672e-01 -7.32864916e-01 1.14229059e+00
2.06106588e-01 -2.28199482e-01 2.25856692e-01 5.62597334e-01
-1.01095164e+00 5.51254034e-01 9.33745801e-01 1.68225430e-02
-1.92366183e-01 2.42643714e-01 -1.09772205e+00 2.00439975e-01
7.28641510e-01 -7.73161769e-01 -5.97883880e-01 -5.26569068e-01
-1.14055030e-01 -2.51242697e-01 2.16925934e-01 -5.99075675e-01
7.49262750e-01 -3.72539431e-01 2.44412031e-02 -1.33791178e-01
8.35115388e-02 -1.21155059e+00 6.02722704e-01 3.12626332e-01
-3.79087627e-01 -2.50919640e-01 -3.02507848e-01 8.25579047e-01
7.92031288e-02 3.09237868e-01 6.95935011e-01 -2.14375630e-01
3.26186791e-02 6.90251172e-01 -1.75489001e-02 -3.71827409e-02
9.99341011e-01 2.29500681e-01 -8.04704666e-01 -5.76325297e-01
-7.25621939e-01 3.18239331e-01 9.54480767e-01 2.04795793e-01
1.65991381e-01 -1.22183299e+00 -4.15396363e-01 5.17464161e-01
-1.24416374e-01 -9.66254100e-02 4.62974370e-01 8.17552209e-01
5.02494909e-02 2.51473010e-01 1.49677768e-01 -1.49009943e-01
-1.15038824e+00 8.35934341e-01 3.52455109e-01 -2.21464500e-01
-6.29034817e-01 7.57366955e-01 -3.79069559e-02 -8.42487454e-01
8.39750409e-01 2.76859268e-03 7.74659932e-01 -5.33609837e-02
7.81240165e-01 6.91446066e-01 3.50114197e-01 5.02459519e-02
-3.20135742e-01 -3.39083523e-01 -5.35196304e-01 5.86275198e-02
1.62811506e+00 -5.14256537e-01 -2.93108255e-01 3.29539657e-01
1.26149058e+00 2.88388908e-01 -1.76181412e+00 -8.20444047e-01
-1.80482060e-01 -8.78008544e-01 -6.20820262e-02 -1.08515465e+00
-1.48534071e+00 4.91140693e-01 6.19610488e-01 7.70664513e-02
1.02163053e+00 4.30869758e-02 8.84259522e-01 4.95736748e-01
1.00145125e+00 -1.01855087e+00 -3.43156993e-01 1.06520131e-01
4.35981721e-01 -9.73435581e-01 -1.33107141e-01 1.41524702e-01
-4.51007873e-01 7.93676794e-01 6.71000659e-01 2.21083149e-01
8.35147142e-01 2.40556091e-01 2.30417907e-01 -8.91804174e-02
-1.63044572e+00 3.72850478e-01 -6.73371494e-01 5.04148245e-01
-2.68390775e-01 8.40836242e-02 1.11472465e-01 7.23888874e-01
2.10525990e-01 2.58313954e-01 1.51691347e-01 1.23680747e+00
-2.28678361e-01 -1.40129316e+00 -3.12349647e-01 4.08987761e-01
-5.13323545e-01 3.74192208e-01 -5.38308859e-01 2.98470289e-01
2.57026762e-01 9.02773917e-01 -1.54412761e-01 -3.02583069e-01
1.01620734e-01 4.65925336e-01 1.57537058e-01 -2.95690030e-01
-8.39835286e-01 -2.20258310e-01 -1.04875281e-01 -9.51830626e-01
-3.11933160e-02 -5.41035056e-01 -9.42474544e-01 -1.03738630e+00
-2.95002848e-01 4.90263969e-01 6.04500175e-01 5.27775884e-01
1.08771849e+00 -1.55452162e-01 1.05047011e+00 -9.12977755e-02
-1.15736377e+00 -6.35673404e-01 -7.63125956e-01 4.78288472e-01
6.85119689e-01 -1.83684409e-01 -6.24362767e-01 -1.72343001e-01] | [5.86270809173584, 6.384459018707275] |
0b0e3b84-c2dc-45d2-b891-4f02329f7f74 | comprehensive-movie-recommendation-system | 2112.12463 | null | https://arxiv.org/abs/2112.12463v1 | https://arxiv.org/pdf/2112.12463v1.pdf | Comprehensive Movie Recommendation System | A recommender system, also known as a recommendation system, is a type of information filtering system that attempts to forecast a user's rating or preference for an item. This article designs and implements a complete movie recommendation system prototype based on the Genre, Pearson Correlation Coefficient, Cosine Similarity, KNN-Based, Content-Based Filtering using TFIDF and SVD, Collaborative Filtering using TFIDF and SVD, Surprise Library based recommendation system technology. Apart from that in this paper, we present a novel idea that applies machine learning techniques to construct a cluster for the movie based on genres and then observes the inertia value number of clusters were defined. The constraints of the approaches discussed in this work have been described, as well as how one strategy overcomes the disadvantages of another. The whole work has been done on the dataset Movie Lens present at the group lens website which contains 100836 ratings and 3683 tag applications across 9742 movies. These data were created by 610 users between March 29, 1996, and September 24, 2018. | ['Jaydip Sen', 'Ananda Chatterjee', 'Hrisav Bhowmick'] | 2021-12-23 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [-5.40214658e-01 -4.86218095e-01 -3.87942731e-01 -4.61463898e-01
2.04053119e-01 -8.61320436e-01 6.51757061e-01 2.78374702e-01
-4.29816425e-01 3.24679255e-01 8.76423836e-01 -2.50303209e-01
-1.03768432e+00 -7.88400471e-01 -9.06568244e-02 -2.77733803e-01
-1.98449403e-01 2.56229609e-01 3.45718741e-01 -5.23014247e-01
1.17670786e+00 2.96929985e-01 -2.12785101e+00 9.21650052e-01
4.48008955e-01 1.19262493e+00 3.60838681e-01 7.48865902e-01
-2.15705976e-01 5.79315960e-01 -2.97650099e-01 -5.15899777e-01
2.77369589e-01 -2.05671862e-01 -5.38472235e-01 -1.86662391e-01
2.99564868e-01 -2.15093076e-01 -1.03750199e-01 8.33222151e-01
2.96606183e-01 6.24682128e-01 9.92255092e-01 -7.60468066e-01
-9.05964255e-01 7.74622262e-01 8.57166946e-03 3.46573472e-01
8.04448187e-01 -9.43476200e-01 8.66986275e-01 -1.14752734e+00
8.24845254e-01 8.58780503e-01 7.52432942e-01 1.58315882e-01
-5.49158990e-01 -2.04257667e-01 1.31630704e-01 5.13437152e-01
-1.20459962e+00 8.05949345e-02 2.48176977e-01 -8.42048585e-01
8.71979415e-01 5.29264867e-01 8.31120670e-01 4.07056272e-01
2.90057123e-01 1.52262021e-02 1.00578904e+00 -6.86058640e-01
3.84729922e-01 7.70730317e-01 4.90512162e-01 2.61059016e-01
-1.71860028e-02 -1.49533793e-01 -3.24489564e-01 -4.33596879e-01
4.56093103e-01 5.95519245e-01 -3.35670169e-03 -1.88330200e-03
-9.27255869e-01 8.89462829e-01 -1.05608022e-02 6.81341231e-01
-2.76670396e-01 -8.75835299e-01 3.98172617e-01 8.68856013e-01
5.27254224e-01 7.02948928e-01 -6.71667159e-01 5.86552396e-02
-8.27587724e-01 2.05110997e-01 1.14646900e+00 9.90485251e-01
1.83972731e-01 -4.78193074e-01 -1.01880491e-01 8.42160046e-01
7.42281199e-01 6.04088232e-02 9.05101001e-01 -9.35495555e-01
-2.04831615e-01 6.08716071e-01 3.28212380e-01 -1.35224843e+00
-3.14909220e-01 -2.02212974e-01 -2.42982566e-01 1.82774514e-01
2.67444313e-01 -1.86624408e-01 -3.85627419e-01 8.50802600e-01
3.74237001e-01 -1.74014002e-01 -1.30769923e-01 1.05004907e+00
1.20908999e+00 5.32445610e-01 -3.38342428e-01 -5.92152894e-01
1.38683605e+00 -9.65594053e-01 -8.55915546e-01 9.29731905e-01
1.76598802e-01 -1.40660262e+00 6.05298877e-01 1.31916106e+00
-9.69953120e-01 -9.80984867e-01 -8.73011589e-01 3.80011857e-01
-8.68779361e-01 2.20551118e-01 8.16296279e-01 8.19690704e-01
-1.08236539e+00 1.04897141e+00 -8.79048090e-03 -8.00140381e-01
-5.13097048e-01 4.27550256e-01 -7.76973739e-02 1.04175888e-01
-1.06631947e+00 8.55876267e-01 8.19961261e-03 -5.02668381e-01
-4.73955780e-01 -4.83768672e-01 9.19790566e-02 -1.79489762e-01
9.52271074e-02 -4.12427396e-01 7.75300384e-01 -9.67193186e-01
-1.55818427e+00 2.60268480e-01 1.35795206e-01 5.27564669e-03
4.15106304e-02 -1.35745585e-01 -1.22418177e+00 -2.63436764e-01
-1.96727112e-01 -1.78632706e-01 5.37766635e-01 -8.33358645e-01
-1.12278140e+00 -5.05428493e-01 3.40847790e-01 2.46708333e-01
-9.17235017e-01 6.29614592e-01 -4.84162092e-01 -7.11128354e-01
9.31978673e-02 -6.91429496e-01 -2.38741547e-01 -7.08256125e-01
3.56858283e-01 -6.40133440e-01 6.20203093e-02 -4.22116339e-01
1.92439318e+00 -1.82241035e+00 9.54177827e-02 5.06987095e-01
-6.27484173e-02 2.83514913e-02 3.14055264e-01 1.10684323e+00
2.95030981e-01 -9.47015360e-02 8.52955222e-01 2.11292833e-01
-7.66076446e-02 -4.74495739e-02 -2.97881901e-01 2.87156343e-01
-1.06060934e+00 -1.65185347e-01 -6.15798414e-01 -3.05135161e-01
1.24914892e-01 3.57864320e-01 -7.69710898e-01 3.06293935e-01
1.15778252e-01 1.08329229e-01 -3.62436295e-01 4.53566045e-01
4.47795242e-01 -1.41867874e-02 4.00763988e-01 -5.99800229e-01
-6.76857829e-01 2.63051301e-01 -1.59139276e+00 1.72685480e+00
-2.86227733e-01 8.64055008e-02 -1.28056288e-01 -7.95464098e-01
1.10407126e+00 4.57186818e-01 7.68324792e-01 -1.98895097e-01
3.85507464e-01 1.58454359e-01 -7.73506612e-02 -1.02689075e+00
8.25677872e-01 2.74326652e-01 2.02686742e-01 5.73393762e-01
2.85407275e-01 5.13592243e-01 5.82369506e-01 3.51969093e-01
8.77963483e-01 1.54988900e-01 2.87662119e-01 -5.95528364e-01
6.42997384e-01 1.97178021e-01 2.10966930e-01 7.21068919e-01
3.82071912e-01 4.75852281e-01 -3.43663335e-01 -5.64138949e-01
-1.01596653e+00 -6.35760665e-01 -3.55254918e-01 1.54360330e+00
1.25804111e-01 -1.10698903e+00 -4.88540649e-01 -4.31942225e-01
1.38400495e-01 3.46212238e-01 -5.28026819e-01 1.45719677e-01
-6.95901830e-03 -4.04594064e-01 -2.33658671e-01 2.25768834e-02
-1.42375231e-01 -7.23460078e-01 1.82120919e-01 3.16745847e-01
2.03929499e-01 -3.15182775e-01 -5.27199268e-01 -2.37124279e-01
-9.13833082e-01 -8.40515733e-01 -4.67749625e-01 -6.48843706e-01
5.95538259e-01 5.17198205e-01 1.13149524e+00 2.23402575e-01
1.08883843e-01 5.22097349e-01 -1.29398048e+00 -2.20294029e-01
1.70151740e-02 -1.80933684e-01 6.00640476e-01 1.03272200e-01
8.56625438e-01 -5.89500248e-01 -8.05617929e-01 7.17589080e-01
-4.69661593e-01 -5.70430338e-01 1.81173682e-01 3.17370355e-01
3.24981362e-01 2.87259430e-01 5.23727059e-01 -1.08997118e+00
1.06387126e+00 -9.45884883e-01 -2.03411341e-01 2.25148976e-01
-1.44427407e+00 -5.49994469e-01 8.92443657e-01 -4.65397388e-01
-9.74393189e-01 -1.42779425e-01 -2.09292933e-01 9.08745602e-02
-2.20279455e-01 6.01473749e-01 2.41378143e-01 -3.51724327e-02
1.04731846e+00 -1.46655321e-01 -1.55099347e-01 -1.25869060e+00
5.14907122e-01 1.40411699e+00 2.03252852e-01 -1.54309869e-01
2.25683331e-01 2.84113377e-01 -5.33699214e-01 -3.04460883e-01
-6.72201335e-01 -1.13516903e+00 -8.01764071e-01 -8.24359357e-01
6.55415893e-01 -7.97265589e-01 -7.79549003e-01 -8.36313441e-02
-5.19824624e-01 5.97964764e-01 9.64491442e-02 1.17596352e+00
-1.59294114e-01 2.86125928e-01 -6.69544101e-01 -8.24159384e-01
-4.28413272e-01 -5.76153338e-01 -2.78222430e-02 4.41652149e-01
-3.32020432e-01 -8.71617496e-01 2.27643043e-01 4.92806166e-01
7.33904123e-01 -3.78604859e-01 5.61986685e-01 -1.00509644e+00
1.78036660e-01 -3.64147127e-01 2.86595613e-01 3.82389337e-01
1.78703636e-01 2.47448280e-01 -3.42881918e-01 -4.30066586e-01
6.40637055e-02 2.56685138e-01 4.30609226e-01 3.28602195e-01
8.92488778e-01 -3.36879820e-01 -2.39700228e-01 2.93943346e-01
1.71622217e+00 7.76562750e-01 3.82064492e-01 4.31816429e-01
3.65540951e-01 4.52675253e-01 1.07070005e+00 9.10638928e-01
3.14856857e-01 6.40386701e-01 7.76729882e-02 5.28819978e-01
1.46822125e-01 -3.54492068e-02 2.66235054e-01 1.42292225e+00
-9.00111973e-01 -1.97489575e-01 -1.77045912e-01 6.76542073e-02
-1.86787844e+00 -1.04427111e+00 -3.95750761e-01 2.47166491e+00
4.26105291e-01 4.17125002e-02 6.30317330e-01 2.60382026e-01
6.27312899e-01 -5.90161383e-01 -3.08975931e-02 -9.09088254e-01
2.20529079e-01 -1.53208494e-01 5.43873131e-01 3.92114818e-01
-9.00170743e-01 5.85223615e-01 6.26468992e+00 7.91055620e-01
-7.53142655e-01 2.02487364e-01 -4.63466905e-03 -1.90112010e-01
-3.53985220e-01 5.59478439e-02 -9.87146378e-01 7.92411268e-01
1.01982582e+00 -2.19452575e-01 8.67104352e-01 1.01510894e+00
4.12548244e-01 -1.98248193e-01 -6.21091723e-01 6.84452593e-01
6.34633422e-01 -1.43295562e+00 -5.85556589e-02 9.39867944e-02
8.93059969e-01 -8.57787430e-02 -1.61778644e-01 1.80352390e-01
3.78543615e-01 -4.46576685e-01 4.41820621e-01 1.09056079e+00
2.13282555e-01 -6.20070875e-01 7.43765593e-01 3.76113385e-01
-8.80268514e-01 -4.40013468e-01 -8.53510797e-01 -6.76807880e-01
-7.03015104e-02 5.84206641e-01 -3.30751777e-01 6.60403192e-01
1.14629972e+00 1.01594722e+00 -3.84176582e-01 1.42329371e+00
3.12593013e-01 6.50136769e-01 -4.03318331e-02 -5.16395330e-01
-1.13200478e-01 -7.18026936e-01 2.79519647e-01 1.16105878e+00
7.52072215e-01 3.63605946e-01 1.27796158e-01 -9.58516002e-02
5.23073494e-01 9.80130434e-01 -2.58437574e-01 3.85555774e-01
7.05796540e-01 1.65371239e+00 -7.96087980e-01 -4.57712203e-01
-5.72243094e-01 5.57974398e-01 -2.69365191e-01 -7.18643218e-02
-5.58448553e-01 -4.12053376e-01 2.76355058e-01 5.01279771e-01
4.02451366e-01 1.69904996e-02 -1.28217116e-01 -1.11249232e+00
-5.61668813e-01 -9.67774451e-01 6.93431318e-01 -6.96415484e-01
-1.56431293e+00 7.19877481e-01 4.39659357e-02 -1.71609616e+00
2.30057281e-03 -7.07143307e-01 -5.25130093e-01 7.61060297e-01
-4.15569425e-01 -6.36255562e-01 -1.92241833e-01 8.39167297e-01
5.11237323e-01 -9.93067682e-01 8.65407825e-01 8.96696985e-01
-7.19312951e-02 2.44782731e-01 9.58594084e-01 -4.48126465e-01
1.06362450e+00 -1.33097267e+00 -4.50726658e-01 2.18779936e-01
2.00272009e-01 1.04311478e+00 8.29885900e-01 -5.79596639e-01
-1.44581115e+00 -3.78102064e-01 1.19131720e+00 -4.47467268e-01
6.22243047e-01 -4.80468683e-02 -3.07538033e-01 1.66635618e-01
4.05847937e-01 -5.81115246e-01 1.23710227e+00 5.63195646e-01
-8.56214836e-02 -4.79234457e-01 -1.20734119e+00 2.01878637e-01
9.42239285e-01 -7.46613517e-02 -6.58711016e-01 7.87599683e-01
3.12769622e-01 8.98511037e-02 -1.85733950e+00 1.63253814e-01
1.05808902e+00 -1.09835756e+00 1.06775916e+00 -9.29554164e-01
3.00182492e-01 -5.80509007e-01 -4.56463099e-01 -1.26687598e+00
-1.20330167e+00 -1.55116618e-01 6.69799373e-03 1.47995448e+00
6.57672107e-01 -2.28623793e-01 5.54909706e-01 3.69514078e-01
-2.00921193e-01 -7.29156792e-01 -3.29721868e-01 -4.66481596e-01
-1.90905333e-01 -1.71830431e-01 4.69237745e-01 1.06483376e+00
6.16656780e-01 4.11959350e-01 -7.37311959e-01 -2.48787433e-01
2.21919045e-01 2.70876884e-01 5.20957470e-01 -1.47538710e+00
-5.39600372e-01 -3.12893808e-01 -2.03869864e-01 -8.57117891e-01
-7.35961139e-01 -7.88352907e-01 -6.01251841e-01 -1.49552441e+00
8.24190229e-02 -5.13281763e-01 -7.84915566e-01 -2.23655954e-01
3.86915058e-01 3.90524328e-01 5.21922633e-02 7.63576686e-01
-7.69602358e-01 -5.39345801e-01 1.14407730e+00 2.83402920e-01
-7.78018162e-02 4.56282556e-01 -8.51776898e-01 7.04008400e-01
5.38844109e-01 -4.53250349e-01 -6.12589478e-01 -9.33522582e-02
8.94138336e-01 -2.38694072e-01 -7.04559565e-01 -7.92652309e-01
6.25963271e-01 -1.72276437e-01 5.74015141e-01 -7.16944277e-01
7.00044408e-02 -9.95949149e-01 7.19211996e-01 8.24098140e-02
-5.51566005e-01 2.98351765e-01 -6.11319482e-01 5.73099554e-01
-9.92252752e-02 -7.49462962e-01 1.94819242e-01 -3.72443557e-01
-7.97123909e-01 1.00529976e-01 -7.96818614e-01 -7.48552799e-01
8.45132947e-01 -2.26590753e-01 -1.03427976e-01 -4.07074571e-01
-1.27630758e+00 -2.54148722e-01 3.89259875e-01 8.56879354e-01
3.16068381e-01 -1.45820868e+00 -5.07616818e-01 -1.10120192e-01
1.36106178e-01 -1.26825035e+00 3.15516561e-01 7.97107041e-01
-3.22076529e-01 2.96524018e-01 -4.79503214e-01 2.31891274e-01
-1.58563375e+00 1.13567221e+00 -2.16563523e-01 6.80627003e-02
-2.26469815e-01 7.53704846e-01 -6.71024859e-01 -2.57700861e-01
3.44928503e-01 1.54704258e-01 -1.77602768e+00 5.58086872e-01
8.38711798e-01 6.52089655e-01 2.45546624e-01 -5.82125783e-01
-1.75788626e-01 7.64678597e-01 -2.16434419e-01 1.27869314e-02
1.31047416e+00 -5.64913750e-01 -3.63826394e-01 4.96746749e-01
9.81090784e-01 3.19695652e-01 -2.59680748e-01 -2.85998821e-01
1.19543762e-03 -8.16274226e-01 2.19680950e-01 -1.13603556e+00
-9.15929019e-01 8.49476382e-02 9.39252198e-01 1.01413465e+00
1.15026414e+00 -1.80622369e-01 4.55054551e-01 1.89915732e-01
4.51653451e-01 -1.67075229e+00 -3.43764096e-01 4.03667152e-01
6.55035853e-01 -1.03580964e+00 2.54391164e-01 -3.57492328e-01
-5.09810209e-01 1.33737004e+00 3.28538418e-01 -4.28531229e-01
1.59561431e+00 1.33984149e-01 -6.78466111e-02 -4.99068722e-02
-9.52871501e-01 -7.82338716e-03 6.84261918e-01 3.40438783e-01
1.01068330e+00 1.15864657e-01 -1.61016178e+00 1.26847506e+00
-2.96323240e-01 3.76567870e-01 4.18964148e-01 7.62971580e-01
-9.52385366e-01 -1.28769815e+00 -3.54292005e-01 9.42708552e-01
-8.40327084e-01 -1.99359030e-01 -3.53428155e-01 1.79218668e-02
7.93080628e-01 1.47939849e+00 -1.63692404e-02 -1.17225349e+00
3.75956744e-01 -4.81098443e-01 3.17303896e-01 -6.23972476e-01
-1.08206868e+00 1.44073904e-01 3.99555713e-01 -1.73820421e-01
-5.69196284e-01 -7.54428864e-01 -6.52586639e-01 -4.63378400e-01
-5.49017489e-01 8.78675699e-01 1.17528582e+00 6.73296630e-01
4.83986050e-01 3.84107307e-02 1.07827938e+00 -6.21949852e-01
-4.24469858e-01 -1.24861646e+00 -8.91083300e-01 5.23250639e-01
-5.24596572e-01 -7.63640702e-01 -5.51711321e-01 2.54029453e-01] | [10.017128944396973, 5.887871265411377] |
52b8edf8-29bc-4841-953c-af4594cda075 | logical-message-passing-networks-with-one-hop | 2301.08859 | null | https://arxiv.org/abs/2301.08859v3 | https://arxiv.org/pdf/2301.08859v3.pdf | Logical Message Passing Networks with One-hop Inference on Atomic Formulas | Complex Query Answering (CQA) over Knowledge Graphs (KGs) has attracted a lot of attention to potentially support many applications. Given that KGs are usually incomplete, neural models are proposed to answer the logical queries by parameterizing set operators with complex neural networks. However, such methods usually train neural set operators with a large number of entity and relation embeddings from the zero, where whether and how the embeddings or the neural set operators contribute to the performance remains not clear. In this paper, we propose a simple framework for complex query answering that decomposes the KG embeddings from neural set operators. We propose to represent the complex queries into the query graph. On top of the query graph, we propose the Logical Message Passing Neural Network (LMPNN) that connects the local one-hop inferences on atomic formulas to the global logical reasoning for complex query answering. We leverage existing effective KG embeddings to conduct one-hop inferences on atomic formulas, the results of which are regarded as the messages passed in LMPNN. The reasoning process over the overall logical formulas is turned into the forward pass of LMPNN that incrementally aggregates local information to finally predict the answers' embeddings. The complex logical inference across different types of queries will then be learned from training examples based on the LMPNN architecture. Theoretically, our query-graph represenation is more general than the prevailing operator-tree formulation, so our approach applies to a broader range of complex KG queries. Empirically, our approach yields the new state-of-the-art neural CQA model. Our research bridges the gap between complex KG query answering tasks and the long-standing achievements of knowledge graph representation learning. | ['Simon See', 'Ginny Y. Wong', 'Yangqiu Song', 'ZiHao Wang'] | 2023-01-21 | null | null | null | null | ['complex-query-answering', 'logical-reasoning'] | ['knowledge-base', 'reasoning'] | [-9.74450856e-02 5.51235259e-01 -3.82147402e-01 -5.73916078e-01
-4.01767880e-01 -3.36626202e-01 5.01510501e-01 5.40534258e-01
-2.47854963e-01 3.57070446e-01 1.56256542e-01 -6.42784119e-01
-4.70237523e-01 -1.77767694e+00 -9.84812796e-01 -1.20815292e-01
-7.05495328e-02 8.20392132e-01 4.99480963e-01 -5.41289270e-01
-1.25562295e-01 2.86939740e-01 -1.51871514e+00 5.26357234e-01
8.86602879e-01 1.38023734e+00 -3.63052398e-01 6.05603278e-01
-7.36908674e-01 1.54607165e+00 -4.49424535e-01 -9.30576801e-01
-2.12741494e-01 -2.44107962e-01 -1.26110351e+00 -7.17064261e-01
4.43016857e-01 -3.79412353e-01 -4.22865421e-01 9.80785966e-01
2.32171580e-01 6.73568621e-02 3.74638170e-01 -1.38425696e+00
-1.08169067e+00 1.03615022e+00 2.80944377e-01 -2.12566666e-02
5.84119856e-01 -9.83305927e-03 1.74988413e+00 -7.83767581e-01
4.93975610e-01 1.52082205e+00 5.84945560e-01 3.08479548e-01
-9.70685005e-01 -4.14906502e-01 2.18301192e-01 7.54258573e-01
-1.49208772e+00 -8.06743205e-02 7.28198707e-01 -1.37995541e-01
1.31485140e+00 2.25463565e-02 7.94105470e-01 3.94470930e-01
3.30544375e-02 1.11342132e+00 3.31878066e-01 -3.63034517e-01
3.24703157e-01 6.44033998e-02 5.39590120e-01 1.13799417e+00
4.14187759e-01 -2.99054831e-01 -6.14645779e-01 -3.62002075e-01
3.53194863e-01 1.02860525e-01 -2.47643605e-01 -4.91136044e-01
-8.71396720e-01 1.25741303e+00 7.49402702e-01 6.11388125e-02
-3.98047686e-01 6.32499874e-01 5.63511848e-01 7.14067400e-01
1.90662757e-01 5.85871339e-01 -7.38183975e-01 2.92040378e-01
-3.44224900e-01 4.16781396e-01 1.35378969e+00 8.92202795e-01
1.08884811e+00 -2.68012583e-01 -3.11351299e-01 5.97391069e-01
5.54224968e-01 4.27977115e-01 -8.88715759e-02 -8.37332189e-01
5.21874130e-01 1.23450112e+00 -2.14558631e-01 -1.25810146e+00
-2.04887137e-01 -2.74304390e-01 -5.27446330e-01 -3.83026481e-01
1.71669662e-01 1.21736988e-01 -3.99106115e-01 1.74246955e+00
4.22926068e-01 2.14639038e-01 4.17899162e-01 5.09370327e-01
1.00233769e+00 6.00242972e-01 2.86131427e-02 1.49960384e-01
1.42692387e+00 -8.99122655e-01 -7.03970373e-01 1.82667766e-02
1.21925318e+00 -2.99282297e-02 1.08100188e+00 2.16132432e-01
-9.91372883e-01 -4.93831158e-01 -1.04650331e+00 -5.20272613e-01
-9.05175030e-01 -2.55600214e-01 1.05999804e+00 2.39116207e-01
-1.25043511e+00 3.06937784e-01 -5.12541831e-01 -3.18713933e-01
5.52758515e-01 6.21977210e-01 -6.93067955e-03 -6.79946601e-01
-1.82391953e+00 8.82562280e-01 8.65840197e-01 4.39410210e-01
-5.95145583e-01 -6.72971189e-01 -1.27711272e+00 3.75044197e-01
1.00624478e+00 -1.27871919e+00 1.28312492e+00 -2.34351635e-01
-1.32265365e+00 4.18799251e-01 -2.67638057e-01 -5.77386439e-01
-1.53588101e-01 -2.44259417e-01 -4.94284689e-01 1.84995785e-01
-2.42853276e-02 5.49013853e-01 5.33780515e-01 -9.81474459e-01
-5.59395492e-01 -3.61765891e-01 9.64674413e-01 1.00483589e-01
-3.09635550e-01 -3.43421131e-01 -5.49292862e-01 7.49992905e-03
1.61736868e-02 -3.57811302e-01 8.32230523e-02 -2.24747546e-02
-3.71581346e-01 -1.08238435e+00 6.61568820e-01 -1.13301106e-01
1.62595487e+00 -1.87340271e+00 1.44112200e-01 3.22660208e-01
5.38538575e-01 1.51387557e-01 -1.37598068e-01 7.20287502e-01
1.92877561e-01 1.21290825e-01 1.71293858e-02 1.56887714e-02
6.04727685e-01 6.77359521e-01 -6.85732365e-01 -1.67559355e-01
5.19596696e-01 1.60324597e+00 -1.09223747e+00 -5.75735688e-01
-1.63694605e-01 1.06689781e-01 -9.49115098e-01 3.25982928e-01
-1.02094746e+00 -4.94988382e-01 -7.34691858e-01 7.58010089e-01
5.65654457e-01 -4.85775113e-01 2.06676647e-01 -5.59076905e-01
5.16419709e-01 4.78527188e-01 -9.90351498e-01 1.67388773e+00
-3.81910473e-01 1.63868308e-01 -1.53595462e-01 -1.34921563e+00
7.55918145e-01 1.71354145e-01 7.00163543e-02 -7.27258146e-01
-1.75945610e-01 1.51459038e-01 -3.17648984e-02 -7.94010162e-01
5.63639820e-01 -2.91806102e-01 -3.17784846e-01 4.64674324e-01
2.51479477e-01 -4.15442526e-01 3.10437322e-01 6.77037299e-01
1.12184441e+00 -2.57021915e-02 1.83497146e-01 1.52450264e-01
7.75771439e-01 8.47260505e-02 2.99784839e-01 9.79989767e-01
9.26276222e-02 -2.94219136e-01 9.58754897e-01 -6.65759921e-01
-2.72445410e-01 -1.16645348e+00 2.76242256e-01 1.37917948e+00
6.84167966e-02 -7.35694706e-01 -3.77715349e-01 -8.54094028e-01
3.13193798e-01 7.12486863e-01 -4.30204809e-01 -6.60798252e-01
-6.34672821e-01 -2.03190863e-01 9.21010494e-01 6.81688726e-01
5.87159634e-01 -1.36492634e+00 -1.44333825e-01 1.99076548e-01
3.28571163e-02 -1.15977454e+00 -6.77905604e-02 4.84491587e-02
-7.34373808e-01 -1.38010895e+00 2.15798378e-01 -7.99062610e-01
3.86021048e-01 -1.92923084e-01 1.49328721e+00 4.40610528e-01
1.08769223e-01 6.95113003e-01 -2.43329793e-01 -3.81446868e-01
-2.49538958e-01 9.24366415e-02 -1.64034776e-02 5.86734824e-02
7.96711028e-01 -3.94169152e-01 -4.54634607e-01 -1.68860063e-01
-1.24087703e+00 -2.24593565e-01 4.83943701e-01 6.51523054e-01
6.16882622e-01 4.48489696e-01 7.03827679e-01 -1.31838632e+00
1.08195806e+00 -6.52611077e-01 -6.11704648e-01 7.19086885e-01
-6.04027629e-01 5.21158457e-01 9.22026753e-01 -2.13147968e-01
-9.42595959e-01 -5.25595129e-01 -1.16966404e-01 -4.08483565e-01
2.03702450e-01 1.27221704e+00 -2.49756724e-01 7.81023502e-02
4.10106689e-01 1.76674873e-01 -3.72280389e-01 -1.83963496e-03
1.06825495e+00 1.04993485e-01 4.11632925e-01 -1.09333456e+00
8.94221544e-01 2.74835616e-01 9.22508091e-02 -3.90016049e-01
-1.13981092e+00 -2.86404043e-01 -3.74639601e-01 1.81250080e-01
8.58757436e-01 -5.02545476e-01 -1.21276224e+00 1.07537024e-01
-1.41235816e+00 -2.93752491e-01 -5.80255926e-01 2.17987478e-01
-4.23550367e-01 2.14665115e-01 -7.19104886e-01 -7.27637649e-01
-3.17465484e-01 -1.00341690e+00 9.56211030e-01 5.64036369e-02
-1.37582242e-01 -1.32998776e+00 5.30125871e-02 2.90530920e-01
4.71912473e-01 -1.80743083e-01 1.84569442e+00 -8.66308570e-01
-1.12595654e+00 -3.03745210e-01 -4.46107060e-01 3.43570799e-01
-9.42948982e-02 -3.36978614e-01 -9.33676541e-01 6.57074377e-02
-9.80421528e-02 -7.88573861e-01 8.44822049e-01 2.91053150e-02
1.12904668e+00 -4.57996279e-01 -1.92346945e-01 4.77992177e-01
1.47523940e+00 5.49383787e-03 5.55798769e-01 -7.68088847e-02
8.15554857e-01 4.48942095e-01 2.61144459e-01 2.83205975e-02
9.94917870e-01 1.62814349e-01 5.97000360e-01 3.18728417e-01
1.93709373e-01 -7.32802093e-01 2.48502403e-01 9.89865065e-01
2.55829036e-01 -2.66278654e-01 -9.04013932e-01 3.48062634e-01
-1.79754329e+00 -9.41619396e-01 7.69325495e-02 1.84499288e+00
9.37074125e-01 1.85341969e-01 -3.99112880e-01 8.47887993e-02
1.60144776e-01 3.63858134e-01 -7.08328485e-01 -6.26135349e-01
4.97750752e-02 5.85904777e-01 -7.71868229e-02 6.82362020e-01
-7.72251964e-01 1.07530880e+00 5.57430744e+00 6.04328096e-01
-7.98170686e-01 -1.14476927e-01 1.23243649e-02 2.27829397e-01
-7.29590893e-01 2.16238365e-01 -8.32688451e-01 -1.03820026e-01
9.68188286e-01 -1.95228666e-01 6.33903027e-01 7.36465216e-01
-3.16857189e-01 7.80349746e-02 -1.73500156e+00 7.63238072e-01
4.78153564e-02 -1.64765930e+00 6.41967893e-01 -2.22267881e-01
3.86313647e-01 -1.07369423e-01 -2.31256440e-01 1.23243058e+00
5.97184062e-01 -1.09160233e+00 3.57672423e-01 8.63262236e-01
5.15164077e-01 -4.75163430e-01 8.69786501e-01 3.48587394e-01
-1.52612138e+00 -3.29519778e-01 -4.60685074e-01 -7.35725313e-02
5.04115894e-02 4.25192386e-01 -9.13212538e-01 1.05079782e+00
4.40920383e-01 7.03099072e-01 -6.39467061e-01 5.36746919e-01
-5.34575224e-01 4.92769390e-01 -3.87928963e-01 -2.76358843e-01
2.32784331e-01 -1.60932347e-01 -1.85687125e-01 9.43135977e-01
9.36546363e-03 2.20600501e-01 1.38807178e-01 1.30245054e+00
-5.04465640e-01 2.08186079e-03 -7.92082727e-01 -3.50512654e-01
5.52724719e-01 8.26157212e-01 -9.57172886e-02 -7.97930062e-01
-7.27748573e-01 6.72157764e-01 8.83170247e-01 7.15277553e-01
-5.28140545e-01 -5.72808385e-01 4.37207401e-01 -1.34892121e-01
4.45914239e-01 8.93756822e-02 9.86773372e-02 -1.25242043e+00
3.59848857e-01 -7.23727942e-01 1.06917191e+00 -8.44205678e-01
-1.46816838e+00 2.27273807e-01 2.35029802e-01 -4.84788120e-01
-2.77186424e-01 -7.86421299e-01 -8.40073705e-01 8.30309868e-01
-1.86097836e+00 -1.18885458e+00 -1.54454634e-01 7.38177299e-01
9.46797431e-03 1.74239725e-01 1.03720450e+00 2.36219540e-01
-3.86844099e-01 5.83761871e-01 -5.71777225e-01 3.57649654e-01
1.56584606e-01 -1.30637217e+00 5.84150199e-03 3.98146808e-01
1.63790837e-01 1.13985872e+00 7.04812557e-02 -3.79476666e-01
-1.93825305e+00 -1.10711646e+00 1.23892093e+00 -5.48619211e-01
1.07127309e+00 -2.82886803e-01 -1.19216979e+00 9.88819838e-01
1.15107633e-01 3.45648706e-01 7.07913339e-01 4.90958601e-01
-7.25610554e-01 -4.61893380e-01 -8.19706976e-01 5.14175534e-01
9.29403484e-01 -1.15846598e+00 -9.38705027e-01 2.54370481e-01
1.31260312e+00 9.48300585e-03 -1.19900525e+00 4.82476264e-01
3.28857243e-01 -6.50439739e-01 9.00040984e-01 -1.11676741e+00
3.78460884e-01 -5.00659406e-01 -3.77891600e-01 -9.49718714e-01
-2.99199708e-02 -1.92635179e-01 -8.13792408e-01 8.38846207e-01
4.91109371e-01 -9.42053616e-01 7.50116110e-01 6.57978058e-01
-1.74306676e-01 -1.26449740e+00 -8.20194066e-01 -2.86469281e-01
4.92303818e-02 -9.03493583e-01 8.89323235e-01 9.46373045e-01
2.35150605e-01 8.07069242e-01 3.71386170e-01 4.45224077e-01
4.27803248e-01 3.94912958e-01 7.38975406e-01 -1.37123251e+00
-4.88293260e-01 -4.20990855e-01 -4.70438629e-01 -1.41036701e+00
5.15254676e-01 -1.37412691e+00 -3.40085804e-01 -2.01822758e+00
-1.03517860e-01 -5.06254077e-01 -4.14540023e-01 6.87792540e-01
-1.84032276e-01 -5.12693644e-01 -6.76286756e-04 -1.48529276e-01
-9.81728971e-01 7.46333659e-01 1.27687204e+00 -6.27741277e-01
6.97770864e-02 -1.44591898e-01 -8.24286819e-01 5.41737795e-01
4.70796466e-01 -2.03148186e-01 -1.14934611e+00 -7.02750504e-01
1.12395251e+00 1.31292656e-01 4.83156651e-01 -6.50742233e-01
7.61567712e-01 -2.70393826e-02 -2.02414081e-01 -5.30506611e-01
3.94190192e-01 -7.86790073e-01 -3.71813506e-01 2.81227350e-01
-4.69639748e-01 1.00383498e-01 -9.16049723e-03 7.31418371e-01
-6.27473652e-01 -1.41922131e-01 1.41285747e-01 -1.84163466e-01
-8.26449513e-01 5.71808159e-01 1.09489553e-01 4.46726590e-01
5.50204158e-01 9.90668908e-02 -4.75844473e-01 -4.70693916e-01
-7.45176375e-01 6.87957823e-01 -1.38781264e-01 1.76704317e-01
8.60165894e-01 -1.39214337e+00 -2.90310711e-01 2.31539831e-01
3.46733242e-01 7.63679147e-01 1.90540906e-02 8.16914916e-01
-4.55683738e-01 6.80442035e-01 5.98186076e-01 -3.36348057e-01
-5.29176652e-01 7.44694769e-01 5.85809112e-01 -6.12307668e-01
-3.02527159e-01 9.54527557e-01 1.76987231e-01 -8.53877664e-01
3.85263234e-01 -1.01705575e+00 -1.94145292e-01 -8.21712762e-02
4.07210350e-01 1.07223339e-01 -1.19501576e-01 -9.19619203e-02
-2.87197560e-01 3.31838667e-01 -1.42953008e-01 3.40114594e-01
1.11629164e+00 3.74514908e-01 -6.70934856e-01 4.03502882e-01
1.38521588e+00 -2.45188296e-01 -2.71479636e-01 -7.35381246e-01
2.37867355e-01 4.29830737e-02 -3.04349899e-01 -5.87815762e-01
-9.26425874e-01 9.83698905e-01 -7.21100420e-02 4.64164287e-01
9.75064158e-01 5.51948547e-01 7.36365497e-01 1.30950761e+00
3.65796089e-01 -8.19010437e-01 1.23227634e-01 8.28919888e-01
6.71287477e-01 -1.03899205e+00 -1.31130397e-01 -3.87512714e-01
-1.23728909e-01 1.20209157e+00 6.67688131e-01 -2.47794930e-02
6.77488863e-01 1.63734204e-03 -1.69995725e-01 -9.40982461e-01
-1.04602802e+00 -1.61317766e-01 1.18405744e-01 4.08757180e-01
6.59686923e-02 6.39698431e-02 9.09958035e-02 7.66739190e-01
-2.47456208e-01 3.91091138e-01 7.03789592e-02 8.80261779e-01
-4.51246768e-01 -1.09010935e+00 9.00309086e-02 6.10518217e-01
-2.09065471e-02 -3.14531475e-01 -3.16891491e-01 8.45436752e-01
2.02813491e-01 8.32747936e-01 -1.81079164e-01 -4.36926544e-01
6.07623518e-01 5.54919720e-01 3.77988487e-01 -7.65910327e-01
-2.93106139e-01 -8.54189217e-01 3.51367354e-01 -6.93206310e-01
-3.35263640e-01 -4.26597819e-02 -1.66302729e+00 -3.08136880e-01
-2.59022325e-01 4.28545326e-01 4.06337120e-02 1.15591443e+00
4.30525035e-01 5.07195890e-01 1.09978475e-01 8.93289149e-02
-6.70108140e-01 -7.61965513e-01 -4.80746835e-01 4.62784916e-01
2.79928535e-01 -4.21544045e-01 -1.79971978e-01 -2.69259691e-01] | [9.222637176513672, 7.684553623199463] |
f3cdb86c-609a-49fe-9cf3-ef2da552ad02 | input-convex-neural-networks | 1609.07152 | null | http://arxiv.org/abs/1609.07152v3 | http://arxiv.org/pdf/1609.07152v3.pdf | Input Convex Neural Networks | This paper presents the input convex neural network architecture. These are
scalar-valued (potentially deep) neural networks with constraints on the
network parameters such that the output of the network is a convex function of
(some of) the inputs. The networks allow for efficient inference via
optimization over some inputs to the network given others, and can be applied
to settings including structured prediction, data imputation, reinforcement
learning, and others. In this paper we lay the basic groundwork for these
models, proposing methods for inference, optimization and learning, and analyze
their representational power. We show that many existing neural network
architectures can be made input-convex with a minor modification, and develop
specialized optimization algorithms tailored to this setting. Finally, we
highlight the performance of the methods on multi-label prediction, image
completion, and reinforcement learning problems, where we show improvement over
the existing state of the art in many cases. | ['Brandon Amos', 'Lei Xu', 'J. Zico Kolter'] | 2016-09-22 | input-convex-neural-networks-1 | https://icml.cc/Conferences/2017/Schedule?showEvent=835 | http://proceedings.mlr.press/v70/amos17b/amos17b.pdf | icml-2017-8 | ['inference-optimization'] | ['audio'] | [ 7.44121492e-01 5.50635099e-01 -6.36142313e-01 -7.39347339e-01
-6.92185819e-01 -2.32405350e-01 2.19697535e-01 -2.02270657e-01
-4.72817838e-01 9.77938831e-01 3.02699506e-01 -3.26459169e-01
-5.57658792e-01 -6.37962282e-01 -1.29147494e+00 -9.11156952e-01
-2.80962251e-02 7.44176030e-01 -8.07252944e-01 8.55286792e-02
-1.09088220e-01 4.45941925e-01 -1.23662794e+00 5.29287040e-01
4.94018137e-01 1.08358002e+00 -3.14269327e-02 6.93228900e-01
2.97172397e-01 1.21343887e+00 -8.79823640e-02 -4.67195809e-01
4.03442889e-01 7.04820007e-02 -7.71026254e-01 6.86151534e-02
6.81441605e-01 -5.58200538e-01 -2.84800380e-01 1.04615390e+00
3.27291697e-01 3.55654836e-01 7.74517834e-01 -1.51436579e+00
-9.22855139e-01 7.56826162e-01 -1.26069292e-01 -4.52220082e-01
-2.43872508e-01 -1.46752015e-01 1.17325401e+00 -9.07255709e-01
5.29595256e-01 1.25321507e+00 1.13705480e+00 7.24093974e-01
-1.53131938e+00 -3.58814061e-01 2.53750712e-01 7.99493119e-02
-7.42355883e-01 -7.23885596e-01 4.73736912e-01 -3.97925109e-01
9.58535790e-01 1.76983222e-01 2.59032518e-01 1.29902935e+00
-6.69947416e-02 1.01054573e+00 1.03924227e+00 -4.49812442e-01
2.28385821e-01 5.47610857e-02 3.30056012e-01 9.89246786e-01
-1.82744488e-01 3.20164621e-01 -3.73077899e-01 -3.02044809e-01
7.78833866e-01 2.60603935e-01 -1.72900721e-01 -5.29958308e-01
-1.17418265e+00 1.09767282e+00 2.99393535e-01 -3.69075447e-01
-4.80873734e-01 5.69050550e-01 3.00406277e-01 4.21754360e-01
6.26601100e-01 3.87386739e-01 -8.75970960e-01 2.48582542e-01
-9.05920863e-01 3.05058658e-01 9.54746842e-01 8.71087551e-01
8.61466467e-01 2.38211617e-01 -4.53343630e-01 1.08734906e+00
2.80899286e-01 4.71406937e-01 2.00607538e-01 -1.61824989e+00
6.24352038e-01 1.17971219e-01 1.11667611e-01 -7.93788552e-01
-7.48526514e-01 -4.89976674e-01 -1.28935587e+00 2.73888409e-01
4.74701524e-01 -5.31374872e-01 -7.73311675e-01 1.98647964e+00
2.56137662e-02 3.71070385e-01 7.74670988e-02 4.76038009e-01
5.97975016e-01 5.80024123e-01 2.96836756e-02 -2.83161163e-01
8.81225228e-01 -1.09135854e+00 -5.49164474e-01 -2.53039211e-01
6.80587351e-01 -1.87457129e-01 9.18687761e-01 4.26810205e-01
-1.31373644e+00 -2.88651317e-01 -7.37709045e-01 -3.73242915e-01
-1.71192303e-01 4.69279736e-01 9.24581468e-01 5.15082240e-01
-1.31811845e+00 1.12124586e+00 -9.79750037e-01 5.58726043e-02
5.41476667e-01 8.11848760e-01 -2.08227798e-01 -1.27612144e-01
-9.58733261e-01 8.28908563e-01 4.47441876e-01 4.32760268e-01
-9.58767474e-01 -9.63009655e-01 -7.94767141e-01 1.84390634e-01
2.36290067e-01 -1.01387858e+00 1.51051879e+00 -1.27399409e+00
-1.47660351e+00 8.12802315e-01 -1.25986531e-01 -7.31236279e-01
4.77039963e-01 -1.06571391e-01 8.66808891e-02 -1.76568598e-01
-1.76798671e-01 9.02191222e-01 7.84076273e-01 -9.95480359e-01
-4.36527371e-01 -2.95320243e-01 2.50429511e-01 1.69912875e-01
-3.50951552e-01 -8.48262757e-02 -1.54858604e-01 -6.17743790e-01
-2.43101582e-01 -1.06875694e+00 -5.63671649e-01 2.65652657e-01
-6.62791848e-01 -2.83284754e-01 1.89869970e-01 -7.55039036e-01
6.68247283e-01 -1.91966689e+00 4.69016105e-01 1.88082501e-01
2.15930119e-01 -1.71940595e-01 -5.82862675e-01 1.67614400e-01
-3.25726449e-01 3.51533666e-02 -4.72403884e-01 -9.00367975e-01
3.29180151e-01 6.04083002e-01 -4.08453882e-01 6.76888347e-01
8.73096883e-02 1.17289484e+00 -6.24708474e-01 -1.71747997e-01
1.01227716e-01 3.68058473e-01 -7.77914762e-01 2.23616689e-01
-5.06473482e-01 2.61754394e-01 -1.84033662e-01 5.57173610e-01
4.64242399e-01 -5.07082045e-01 2.73826778e-01 -1.47532254e-01
2.05435261e-01 1.36736527e-01 -1.18848777e+00 1.52539504e+00
-6.52962446e-01 5.80170095e-01 3.44962090e-01 -1.53497791e+00
4.27331150e-01 3.75021309e-01 7.32597411e-01 -2.47296289e-01
-1.15942828e-01 -5.26016541e-02 -2.34488294e-01 -4.85350430e-01
2.37927675e-01 -1.97553799e-01 3.25024575e-01 7.69248307e-01
1.54531032e-01 4.22731370e-01 6.72155246e-02 -6.66536987e-02
7.23540246e-01 1.94280520e-01 1.35234147e-01 2.71904469e-02
1.67021796e-01 -2.38554761e-01 6.63340271e-01 1.19966900e+00
2.06183523e-01 5.43467104e-01 6.41143441e-01 -5.48340857e-01
-1.35345387e+00 -7.29639947e-01 -2.15822637e-01 1.66749406e+00
-7.02529907e-01 1.54067902e-02 -7.79200494e-01 -6.01490557e-01
2.13811025e-01 6.36872113e-01 -9.81760502e-01 9.78163183e-02
-5.91859519e-01 -1.14500141e+00 6.57617271e-01 9.00234282e-01
2.86250412e-01 -1.25482440e+00 -5.42172901e-02 1.16903238e-01
-1.77781001e-01 -9.82926846e-01 -1.67020932e-01 4.74647999e-01
-1.20134175e+00 -9.27238464e-01 -4.71667409e-01 -8.15121770e-01
8.16580713e-01 -1.57331735e-01 1.26157534e+00 2.12131962e-01
1.96474135e-01 4.74302024e-01 3.15319687e-01 -2.98171490e-01
-5.14538407e-01 2.61289060e-01 1.99760020e-01 7.33271688e-02
-1.37724832e-01 -6.82828069e-01 -2.61110157e-01 -3.62428501e-02
-8.54607463e-01 3.40926617e-01 5.68809807e-01 1.24673510e+00
8.08824599e-01 -3.91369969e-01 6.49293423e-01 -1.51505518e+00
6.35960996e-01 -6.61893606e-01 -6.08439147e-01 5.94490886e-01
-8.19738746e-01 4.24406081e-01 9.18497205e-01 -4.65781629e-01
-9.39343274e-01 4.41431969e-01 -1.89504534e-01 -5.39480627e-01
-1.45084277e-01 6.55504584e-01 -1.04205430e-01 -9.46813822e-02
5.90075791e-01 6.13199733e-03 1.76514432e-01 -5.95298350e-01
6.50477111e-01 4.00142074e-01 4.93558466e-01 -8.36736023e-01
5.47844291e-01 5.01074970e-01 3.66558403e-01 -3.37231398e-01
-1.31947398e+00 9.03759673e-02 -7.89506555e-01 1.16923861e-01
6.36253595e-01 -9.19695020e-01 -1.02420449e+00 2.41850585e-01
-1.14923406e+00 -7.71241248e-01 -5.37987113e-01 4.00605440e-01
-1.03971303e+00 6.72671497e-02 -8.69240046e-01 -7.21053720e-01
-4.05955493e-01 -1.15107727e+00 8.10757518e-01 -5.48855476e-02
1.18896656e-01 -1.54565752e+00 -2.08688038e-03 4.61782455e-01
3.47037792e-01 8.37292969e-02 1.19253087e+00 -7.74318695e-01
-6.06041729e-01 8.07658955e-02 -2.93618619e-01 7.23193049e-01
-3.18387300e-01 -1.29169047e-01 -1.05520308e+00 -3.12836617e-01
-2.90204547e-02 -7.57100403e-01 1.17973757e+00 8.76828671e-01
1.72890317e+00 -9.53426123e-01 -1.02826245e-01 1.16622174e+00
1.28694797e+00 -3.60271126e-01 2.48239264e-01 2.40102112e-01
8.36679697e-01 5.22427261e-01 6.26111701e-02 3.99111181e-01
3.69866937e-01 3.92007232e-01 7.66469121e-01 -2.27495059e-01
2.39639029e-01 -1.62226856e-01 3.85373533e-01 5.43709159e-01
-2.14155748e-01 -3.84007506e-02 -4.85180169e-01 3.59639317e-01
-2.29854655e+00 -1.06504631e+00 1.17998071e-01 2.01556134e+00
9.43503261e-01 -4.02788639e-01 6.72437996e-02 -2.42693081e-01
4.89907593e-01 1.17481954e-01 -1.10850012e+00 -5.29701829e-01
-2.59919941e-01 4.04248178e-01 8.48919332e-01 6.19891942e-01
-1.23363042e+00 4.69992787e-01 7.75477409e+00 5.43555200e-01
-6.59619749e-01 2.07928419e-01 1.04794884e+00 -3.10820609e-01
-4.19735223e-01 -2.52516299e-01 -7.33386576e-01 6.26805946e-02
1.24813008e+00 3.19626629e-01 1.17426491e+00 9.15656388e-01
1.82957083e-01 4.33687896e-01 -1.68876040e+00 7.39199042e-01
4.45738621e-02 -1.68124866e+00 -2.45168000e-01 2.00108662e-01
9.84125078e-01 3.39053839e-01 5.17199576e-01 5.06611228e-01
8.34723234e-01 -1.52902901e+00 2.16123238e-01 6.94342792e-01
8.51959884e-01 -6.84794843e-01 5.12069464e-01 5.37964821e-01
-3.29962403e-01 -4.44815069e-01 -6.72663093e-01 -2.03125134e-01
-1.47734955e-01 5.14932811e-01 -6.10969603e-01 2.72386134e-01
3.47555876e-01 1.01025236e+00 -2.50501543e-01 7.09186196e-01
-2.27604315e-01 7.81038404e-01 -4.21205997e-01 2.73375995e-02
2.87264973e-01 -4.66117024e-01 2.15126127e-01 1.11251259e+00
4.98963185e-02 -1.43643916e-01 3.36434186e-01 9.51357841e-01
-6.56235516e-01 6.30192878e-03 -5.79674304e-01 2.36487702e-01
1.53321594e-01 1.29846179e+00 2.42160000e-02 -2.25484684e-01
-4.69604373e-01 6.80467546e-01 9.92359400e-01 7.61356413e-01
-7.15172231e-01 1.01304380e-03 7.57283866e-01 -4.09857899e-01
1.04835555e-01 1.75552577e-01 -5.57758808e-01 -1.22065639e+00
-8.63589197e-02 -9.90844131e-01 4.66078013e-01 -7.27685392e-01
-1.50448215e+00 -2.21948139e-02 -7.70470500e-02 -6.21711552e-01
-5.09877980e-01 -1.06180620e+00 -3.73383254e-01 7.91511595e-01
-1.45571935e+00 -1.14408183e+00 2.58289754e-01 6.39663339e-01
2.03615442e-01 -3.31590682e-01 1.01437020e+00 2.20506325e-01
-7.93475509e-01 6.84565127e-01 8.46498251e-01 2.55125761e-01
3.22823763e-01 -1.45330918e+00 7.84306005e-02 4.71040338e-01
4.35844332e-01 6.04431987e-01 3.98324490e-01 -2.49686852e-01
-1.46067619e+00 -1.40000165e+00 7.85568178e-01 -3.84049565e-01
6.08297288e-01 -2.88557738e-01 -6.48316681e-01 1.44364417e+00
1.79605827e-01 7.37391040e-02 8.12419534e-01 7.11421430e-01
-3.28542084e-01 -1.32067576e-01 -1.17022872e+00 5.38224220e-01
8.85407448e-01 -4.43923265e-01 -1.13473192e-01 9.89515960e-01
6.53022707e-01 -5.77124178e-01 -9.94657576e-01 3.07387859e-01
7.43890405e-01 -7.52565145e-01 1.23454154e+00 -1.52003396e+00
8.31351280e-01 2.50298679e-01 -3.12440187e-01 -1.36326790e+00
-5.73913157e-01 -5.87925136e-01 -6.55279636e-01 7.01322794e-01
5.28627574e-01 -4.77281988e-01 1.01657331e+00 1.02045393e+00
-1.75261810e-01 -1.09278500e+00 -9.97239172e-01 -4.02667224e-01
5.03337681e-01 -4.97185737e-01 4.21191484e-01 9.61049736e-01
-4.27565128e-01 2.61466444e-01 -9.42420304e-01 1.36804134e-01
8.56559098e-01 7.97473043e-02 4.80919868e-01 -1.23105955e+00
-6.49206877e-01 -3.01876187e-01 2.32537836e-01 -1.21812117e+00
8.78406703e-01 -1.34603667e+00 -1.39035851e-01 -1.46827865e+00
4.87774223e-01 -3.97895336e-01 -5.06622672e-01 1.01123714e+00
1.16452612e-02 1.18422098e-01 2.93821335e-01 8.35085139e-02
-5.79935431e-01 4.32782590e-01 1.09003484e+00 -2.89195836e-01
4.44806665e-02 4.99180436e-01 -8.54793549e-01 7.52163351e-01
8.47698867e-01 -6.14898026e-01 -2.35711887e-01 -8.45745504e-01
6.14134073e-01 2.73094475e-01 6.44006491e-01 -4.89597738e-01
2.91387826e-01 -4.08997178e-01 6.38602376e-01 -4.37142253e-01
4.78822887e-01 -8.97232413e-01 2.32602544e-02 3.51565719e-01
-1.12064648e+00 9.40588862e-02 -1.61685243e-01 6.36846721e-01
3.10286850e-01 -5.74020147e-01 7.75817096e-01 -1.96540490e-01
-7.58488327e-02 6.06937349e-01 -1.37898177e-01 2.24840716e-01
4.67437834e-01 7.57091269e-02 -2.09215268e-01 -5.44455945e-01
-9.67080355e-01 4.23569053e-01 4.32327278e-02 -5.57298660e-02
4.61891085e-01 -1.38982248e+00 -9.18518245e-01 2.84706187e-02
-3.96871597e-01 -2.86690947e-02 9.60634202e-02 8.92246485e-01
3.14966962e-03 3.52214694e-01 -1.43706977e-01 -3.81554365e-01
-9.40552771e-01 5.20717025e-01 6.88387454e-01 -5.49956322e-01
-3.60236585e-01 6.18809402e-01 2.73079574e-02 -1.10667205e+00
6.54089630e-01 -4.46802735e-01 -1.11576147e-01 -9.60960612e-02
2.56628573e-01 5.33939481e-01 -7.47943968e-02 -2.70960152e-01
1.10070549e-01 1.80799142e-01 -2.13393271e-02 -4.12367247e-02
1.79083359e+00 7.17544034e-02 -3.32465589e-01 5.40561616e-01
1.45501697e+00 -5.10454416e-01 -1.62632966e+00 -6.18224859e-01
-1.93237156e-01 1.56985089e-01 -5.04118064e-03 -8.70691419e-01
-1.40763259e+00 1.02282429e+00 3.77801865e-01 -8.53050947e-02
1.04172909e+00 -3.52970898e-01 5.63198447e-01 1.12367511e+00
-1.69538245e-01 -1.15755796e+00 -2.94003844e-01 7.42527425e-01
6.68846488e-01 -1.28083074e+00 1.90842971e-01 7.75652528e-02
-3.87015879e-01 1.34284317e+00 2.76773363e-01 -2.77269334e-01
7.79746473e-01 1.72965392e-01 -3.98822814e-01 1.21910393e-01
-1.04174960e+00 1.63184687e-01 3.67684215e-01 7.67099679e-01
3.42956752e-01 7.40999356e-02 2.90144593e-01 5.79227090e-01
-8.11708644e-02 1.36843145e-01 4.27189499e-01 3.58864039e-01
-2.49477148e-01 -1.06176805e+00 -1.42758980e-01 8.20310950e-01
-6.53144658e-01 -4.39738303e-01 -1.18901059e-01 5.11479676e-01
-4.48514894e-02 7.33060181e-01 -1.25311082e-03 4.37855497e-02
1.60485461e-01 2.49152586e-01 4.95379210e-01 -5.02150953e-01
-8.30513954e-01 -3.91044527e-01 1.00662261e-01 -5.01678109e-01
-5.90731561e-01 -7.44120359e-01 -8.90596092e-01 -3.20209205e-01
-8.40661395e-03 -2.64818221e-01 5.41849315e-01 1.10339272e+00
3.01144272e-01 2.84824878e-01 6.05559349e-01 -8.28838706e-01
-1.19841039e+00 -9.57842410e-01 -3.74509275e-01 4.54766750e-01
6.65509105e-01 -3.16276282e-01 -1.96943849e-01 2.91548580e-01] | [8.266666412353516, 4.3725810050964355] |
09e8ea34-5ffd-4bee-93f3-86866a1a35ac | detecting-causes-of-stock-price-rise-and | null | null | https://aclanthology.org/2022.fnp-1.4 | https://aclanthology.org/2022.fnp-1.4.pdf | Detecting Causes of Stock Price Rise and Decline by Machine Reading Comprehension with BERT | In this paper, we focused on news reported when stock prices fluctuate significantly. The news reported when stock prices change is a very useful source of information on what factors cause stock prices to change. However, because it is manually produced, not all events that cause stock prices to change are necessarily reported. Thus, in order to provide investors with information on those causes of stock price changes, it is necessary to develop a system to collect information on events that could be closely related to the stock price changes of certain companies from the Internet. As the first step towards developing such a system, this paper takes an approach of employing a BERT-based machine reading comprehension model, which extracts causes of stock price rise and decline from news reports on stock price changes. In the evaluation, the approach of using the title of the article as the question of machine reading comprehension performs well. It is shown that the fine-tuned machine reading comprehension model successfully detects additional causes of stock price rise and decline other than those stated in the title of the article. | ['Takehito Utsuro', 'Gakuto Tsutsumi'] | null | null | null | null | fnp-lrec-2022-6 | ['machine-reading-comprehension'] | ['natural-language-processing'] | [ 1.01683557e-01 2.17519283e-01 -3.25019777e-01 -3.89500320e-01
-7.12463796e-01 -7.60800302e-01 6.64221227e-01 9.45886314e-01
-2.84773201e-01 8.02339733e-01 3.70453835e-01 -7.57673919e-01
7.46323839e-02 -1.41121078e+00 -8.08518589e-01 -1.17598839e-01
4.85952407e-01 3.22906435e-01 3.47383142e-01 -6.58486307e-01
1.04438078e+00 2.29240507e-01 -1.76244283e+00 2.10508313e-02
4.32527065e-01 9.27703559e-01 4.22721446e-01 5.89643776e-01
-7.88188279e-01 1.02830708e+00 -1.04799724e+00 -3.19245636e-01
1.23318657e-01 -4.99618828e-01 -7.16205955e-01 -2.07662489e-02
1.93318322e-01 -3.70631039e-01 4.40613866e-01 1.25959170e+00
-1.72867671e-01 -5.66528141e-01 4.86994714e-01 -9.48322892e-01
-4.47823703e-01 1.11643887e+00 -3.77336502e-01 1.01327252e+00
9.30274069e-01 -2.01864347e-01 1.31026947e+00 -3.03038120e-01
6.07709885e-01 8.16363633e-01 1.67720616e-01 -2.15384483e-01
-8.29088986e-01 -5.59131503e-01 2.67547816e-01 4.35965806e-02
-7.20286846e-01 -4.72378768e-02 7.89604127e-01 -6.43496096e-01
1.03078091e+00 5.04426718e-01 6.43186629e-01 8.09412777e-01
8.83250058e-01 3.61336261e-01 1.24639153e+00 -7.64073849e-01
2.87079155e-01 4.53566849e-01 5.85948944e-01 1.16127491e-01
7.75170803e-01 -1.56491518e-01 -6.65615797e-01 -1.65141732e-01
4.81807917e-01 -1.18737467e-01 -3.02898765e-01 5.57019353e-01
-1.05915034e+00 1.01231313e+00 -2.30130851e-01 7.32237101e-01
-6.93310201e-01 -1.39252931e-01 1.76714599e-01 8.01972985e-01
6.58784449e-01 7.11844683e-01 -8.44590068e-01 -7.63778090e-01
-9.26521659e-01 4.66065496e-01 1.56517065e+00 7.48117924e-01
4.81085330e-01 -1.62550494e-01 2.56993711e-01 3.04406844e-02
3.65686893e-01 5.32844007e-01 4.89791214e-01 -3.65613818e-01
7.65069664e-01 7.87231803e-01 5.07861495e-01 -1.45508075e+00
-3.35090011e-01 -6.66650176e-01 -6.51273802e-02 1.66859284e-01
3.70693475e-01 -8.23027045e-02 -3.78513426e-01 1.35283232e+00
-1.24501050e-01 -3.97437721e-01 2.24503815e-01 4.67850834e-01
6.99941933e-01 1.09771752e+00 -1.36252284e-01 -8.05475056e-01
1.60136688e+00 -2.09242672e-01 -1.26102257e+00 1.29236802e-02
4.55221981e-01 -1.05587471e+00 8.41563106e-01 5.66721976e-01
-1.20216334e+00 -4.80273694e-01 -1.08713472e+00 3.59974355e-01
-8.35126817e-01 -4.31473672e-01 2.72614419e-01 3.73296887e-01
-6.23711348e-01 5.97305715e-01 -4.37534302e-01 -1.70378804e-01
-5.63289940e-01 -2.11879656e-01 4.13527101e-01 6.73162937e-01
-1.64530790e+00 1.27060688e+00 5.49172640e-01 -3.79352301e-01
7.76532888e-02 -6.06467783e-01 -6.92963839e-01 3.85077626e-01
5.26426435e-01 -8.04287121e-02 1.75151861e+00 -8.31131220e-01
-1.09589744e+00 7.44573355e-01 -1.22855432e-01 -7.20446467e-01
4.88412529e-01 -2.00825423e-01 -6.76879227e-01 5.70268789e-03
2.96069443e-01 -2.94855207e-01 7.18390107e-01 -7.06236482e-01
-1.15371752e+00 -2.03443959e-01 6.24301359e-02 -1.63208142e-01
1.84961408e-01 3.94498765e-01 4.19903547e-02 -7.00564742e-01
-6.15544356e-02 -3.44967753e-01 2.19472632e-01 -1.00937128e+00
-1.74301252e-01 -2.75242448e-01 4.57134843e-01 -9.45507467e-01
1.57008493e+00 -1.79093015e+00 -3.63450140e-01 4.52516675e-01
7.23407045e-02 -3.65511775e-01 5.11025190e-01 7.97388911e-01
-2.66471088e-01 4.48941916e-01 1.87222242e-01 4.88916993e-01
2.87328392e-01 7.01592416e-02 -1.06799400e+00 -1.71895340e-01
1.59393802e-01 6.54588461e-01 -6.47247910e-01 1.48093253e-01
-6.19255118e-02 -3.06618422e-01 2.20817670e-01 -1.27788857e-01
-6.50234222e-01 -1.12466782e-01 -5.79662681e-01 4.21500087e-01
3.53579015e-01 -2.37395227e-01 -4.02283221e-01 2.42459580e-01
-7.76982307e-01 8.96891713e-01 -9.85258102e-01 7.26730943e-01
-1.87736511e-01 6.64074540e-01 -5.15129149e-01 -6.15948021e-01
1.28726590e+00 3.32362592e-01 1.91041619e-01 -9.67929363e-01
1.47400960e-01 3.40544581e-01 -1.25001907e-01 -4.72352684e-01
7.37403750e-01 -2.66344398e-01 -3.53208929e-01 7.74714828e-01
-4.98780996e-01 -2.75605589e-01 6.89917982e-01 -4.15256917e-02
1.08797991e+00 -3.76808375e-01 8.66363823e-01 -4.06625867e-02
5.82801878e-01 3.17305148e-01 3.19340080e-01 5.90513527e-01
2.81519353e-01 1.46210834e-01 1.10781479e+00 -4.73091274e-01
-9.10674751e-01 -7.73756981e-01 -7.89189935e-02 7.96248496e-01
-1.57598689e-01 -4.35563326e-01 -7.63657272e-01 -2.75941163e-01
-1.68479696e-01 1.47259462e+00 -5.08115709e-01 2.13464051e-01
-1.52711347e-01 -7.71345854e-01 -9.22930241e-02 1.77388802e-01
6.57055080e-01 -9.59797442e-01 -1.31693125e+00 7.47388303e-01
-2.76254386e-01 -8.91211033e-01 -3.65665928e-02 2.79508382e-01
-7.20393002e-01 -1.18723083e+00 -5.56581736e-01 -3.75317067e-01
3.27264726e-01 -1.95289642e-01 1.27385998e+00 -9.78858247e-02
9.89384130e-02 4.26607072e-01 -6.46966755e-01 -1.41986096e+00
-8.90341640e-01 2.19055668e-01 -5.13058245e-01 -3.51677269e-01
6.98330045e-01 -2.07386151e-01 1.87415898e-01 1.04592919e-01
-1.05770171e+00 -1.52765989e-01 5.65709591e-01 -1.45766120e-02
2.57350981e-01 6.86545849e-01 7.07615972e-01 -1.03018558e+00
1.15816033e+00 -6.29087150e-01 -1.15974247e+00 3.30409914e-01
-9.76208806e-01 1.85426831e-01 2.75423229e-01 -1.30642867e-02
-1.10373259e+00 -4.04894441e-01 -1.23512231e-01 5.94721854e-01
-4.00128067e-01 1.30295193e+00 2.50652105e-01 5.92281640e-01
3.75023872e-01 5.11016846e-01 -3.83386295e-03 -4.09558654e-01
-3.49081606e-01 4.55819994e-01 3.55605036e-01 1.35302812e-01
7.82602549e-01 -1.07958503e-01 -3.72723579e-01 -7.54495382e-01
-1.04534936e+00 -6.07166529e-01 -3.44905525e-01 -3.45468789e-01
6.44872427e-01 -8.08178127e-01 -1.33230314e-01 5.60184836e-01
-1.07552850e+00 3.04404020e-01 -2.61216909e-01 4.20264781e-01
-4.32830930e-01 -4.46110219e-02 -3.14070612e-01 -8.62407506e-01
-2.33871758e-01 -6.74305499e-01 5.48713446e-01 3.97514135e-01
-8.58471513e-01 -1.11739731e+00 2.30166614e-01 2.80745059e-01
6.35499358e-01 4.49247360e-01 1.11411691e+00 -1.13383257e+00
-5.83243310e-01 -6.49993002e-01 1.19728908e-01 3.45513895e-02
3.91563624e-01 2.98010409e-01 -3.84287804e-01 2.21873119e-01
6.61291003e-01 3.38414520e-01 3.91890675e-01 3.67157489e-01
2.76812434e-01 -4.42858934e-01 1.69860780e-01 -2.43789822e-01
1.41390860e+00 7.41760552e-01 6.37535036e-01 1.01840496e+00
-1.62781432e-01 6.08884335e-01 7.46099234e-01 6.66036725e-01
4.65746939e-01 3.33244145e-01 2.13098571e-01 3.27871799e-01
5.38180292e-01 -1.61275953e-01 6.50844395e-01 9.80610251e-01
-6.60934150e-02 -7.49117434e-02 -9.36217248e-01 3.45701873e-01
-1.59677756e+00 -1.25032938e+00 -3.13604981e-01 1.82151222e+00
6.21748626e-01 9.87279534e-01 1.11489594e-01 3.39078724e-01
5.43088078e-01 3.28032404e-01 -2.56381035e-01 -5.93320847e-01
3.76535393e-03 -5.32730035e-02 6.73622668e-01 4.23334360e-01
-7.08125114e-01 3.50871027e-01 6.66739988e+00 6.04711510e-02
-1.01772118e+00 -2.83195108e-01 5.13252676e-01 4.78512406e-01
-5.91265619e-01 -1.89568371e-01 -1.16565359e+00 6.84081674e-01
1.40085936e+00 -7.73647726e-01 -1.09170169e-01 8.13247979e-01
7.44302273e-01 -7.21967161e-01 -8.69092405e-01 5.54847777e-01
7.33146146e-02 -1.35969365e+00 4.42497954e-02 1.53185651e-01
2.80604482e-01 -4.51089561e-01 -1.37108073e-01 7.65839890e-02
1.44713163e-01 -4.23119217e-01 1.14373124e+00 7.80582428e-01
-1.71868250e-01 -8.48255813e-01 1.26134324e+00 6.58270717e-01
-8.78575146e-01 8.78383219e-02 -4.18607265e-01 -6.76713228e-01
2.56991714e-01 9.21999872e-01 -1.05488729e+00 4.59685534e-01
5.71463823e-01 4.67617780e-01 -6.03810608e-01 7.74359345e-01
-3.99440408e-01 8.74804139e-01 -2.22085580e-01 -6.31938994e-01
1.70312464e-01 -2.74494529e-01 4.88563508e-01 1.03571534e+00
6.17567182e-01 2.93873012e-01 -3.15589428e-01 1.03726411e+00
1.90642759e-01 2.38312110e-01 -8.04209113e-01 -2.90827334e-01
1.67062402e-01 5.80702543e-01 -9.63256717e-01 -4.07715589e-01
-6.34459257e-01 3.51086557e-01 -4.86212313e-01 1.29926547e-01
-6.77797616e-01 -5.95873177e-01 2.61580765e-01 4.15754676e-01
3.46975058e-01 -3.78171802e-02 -4.60950077e-01 -1.02581716e+00
2.86987633e-01 -8.23049009e-01 2.08519518e-01 -8.25622439e-01
-9.44351077e-01 3.50110561e-01 3.53078574e-01 -9.72715199e-01
-5.65509439e-01 -3.66790831e-01 -9.09897804e-01 8.94026160e-01
-1.47943819e+00 -1.61785647e-01 -3.14861611e-02 -4.18548025e-02
5.95933139e-01 -2.94365972e-01 5.43233991e-01 -4.02628839e-01
-1.41431868e-01 -2.82409638e-01 -2.36867860e-01 1.37037382e-01
2.26389349e-01 -1.49893177e+00 5.70147574e-01 9.48080838e-01
2.64489025e-01 4.29241776e-01 1.25704098e+00 -1.04538763e+00
-7.90488183e-01 -5.56997240e-01 1.60776508e+00 -4.26326007e-01
1.18862903e+00 6.82059601e-02 -1.02357125e+00 6.10997021e-01
5.09405792e-01 -1.30518138e+00 6.84668601e-01 -1.42788261e-01
6.97419280e-03 -1.20156571e-01 -8.11971962e-01 3.18168938e-01
9.03609619e-02 -3.09552699e-01 -1.70265555e+00 1.95417523e-01
7.63805270e-01 -3.62337649e-01 -9.49316621e-01 -2.39080206e-01
8.18613637e-03 -8.97762418e-01 3.35015565e-01 -2.57688910e-01
5.14655113e-01 -1.09559432e-01 -2.10822448e-02 -1.43533242e+00
-1.02702864e-01 -6.10593438e-01 3.33573848e-01 1.35451818e+00
1.04320979e+00 -1.13320875e+00 2.84168363e-01 9.92123902e-01
3.16861212e-01 -6.23338670e-02 -5.60776651e-01 -5.21340966e-01
-6.37409613e-02 -5.78177989e-01 8.40448916e-01 7.56714165e-01
2.02749744e-01 2.25344077e-01 2.65784651e-01 2.14049220e-01
2.91151404e-01 3.80857795e-01 6.31316781e-01 -1.43925023e+00
-8.50099549e-02 -5.88294566e-01 -3.46173972e-01 -6.08092666e-01
-1.90603465e-01 -1.88818008e-01 -2.00749531e-01 -1.56344438e+00
-1.94997847e-01 3.19479942e-01 -2.76822746e-01 1.45277753e-01
1.08404815e-01 -7.50643671e-01 3.90637636e-01 2.20877573e-01
-1.22907544e-02 -1.29840925e-01 9.52877045e-01 -3.07115377e-04
-2.91103959e-01 5.72791100e-01 -8.87296319e-01 9.36937571e-01
1.18377471e+00 -4.54736024e-01 -2.50204742e-01 2.32067525e-01
1.05513716e+00 2.95128763e-01 2.08414957e-01 -6.93175018e-01
3.12517017e-01 -2.28470922e-01 2.25994006e-01 -1.14883959e+00
-3.76031250e-01 -8.63348007e-01 3.89271468e-01 6.85526669e-01
-5.59399128e-01 8.75966966e-01 1.08404696e-01 4.03431952e-01
-8.11046243e-01 -7.67344773e-01 1.79670915e-01 -3.56488347e-01
-7.84169197e-01 -6.18958235e-01 -1.13902152e+00 -1.00496918e-01
1.15723908e+00 -2.05197409e-01 -3.77558976e-01 -8.68156016e-01
-5.41475952e-01 4.99771014e-02 2.19176769e-01 5.94253659e-01
4.43293154e-01 -6.51380599e-01 -6.84772611e-01 6.68811947e-02
-7.95014128e-02 -3.71095002e-01 -4.33511496e-01 6.48449302e-01
-8.90016794e-01 8.55047047e-01 -2.42880747e-01 -2.50931252e-02
-1.07800806e+00 4.26564932e-01 -5.32415044e-03 -5.71122527e-01
-5.72359324e-01 3.82745236e-01 -4.59040642e-01 2.21216023e-01
2.35773131e-01 -1.24440157e+00 -6.91334903e-01 8.84248435e-01
1.05023742e+00 2.66629875e-01 3.66304845e-01 -2.91368872e-01
5.51550882e-03 5.65458953e-01 -1.88456476e-01 -2.55757689e-01
1.47990680e+00 -3.87723833e-01 -2.58564591e-01 1.23405743e+00
7.30703890e-01 9.50751454e-02 -5.71708202e-01 -1.44444555e-01
8.48515213e-01 -2.40497246e-01 1.41424522e-01 -1.02926505e+00
-7.61363626e-01 2.40463331e-01 4.85293679e-02 1.25515938e+00
1.07580507e+00 8.07302892e-02 6.54149890e-01 5.36499619e-01
5.42142332e-01 -1.33367777e+00 -5.25594175e-01 5.89670658e-01
1.16377687e+00 -1.08423364e+00 -7.08036572e-02 -4.24237609e-01
-4.59578276e-01 1.62218976e+00 -3.87868136e-02 2.15305746e-01
7.74727941e-01 3.84511501e-01 3.11592042e-01 -5.57826281e-01
-8.96094263e-01 -6.23742826e-02 6.89647049e-02 4.17584330e-02
2.70135880e-01 7.36126378e-02 -7.77722359e-01 5.41766465e-01
-9.00303841e-01 2.01515645e-01 1.20871842e+00 1.20224452e+00
-9.92531598e-01 -8.11720431e-01 -8.32265973e-01 6.85257316e-01
-9.75255191e-01 7.80997872e-02 -7.59998858e-01 1.19162869e+00
-2.76331663e-01 9.21587825e-01 4.24683511e-01 -3.66530903e-02
8.11228156e-01 2.36213431e-01 -1.67047143e-01 -5.65455139e-01
-7.35754788e-01 1.14609636e-01 4.34746981e-01 -3.19902539e-01
-5.26501596e-01 -1.07460690e+00 -1.27750301e+00 -3.71775597e-01
-8.68654028e-02 3.80631000e-01 8.15503180e-01 1.21999431e+00
-2.09635228e-01 5.70248067e-01 5.80922008e-01 -7.82317594e-02
-8.19862127e-01 -7.93188274e-01 -8.14799011e-01 2.08936334e-01
3.84585023e-01 -4.08538461e-01 -5.97573102e-01 4.12370741e-01] | [4.500942230224609, 4.357499122619629] |
1a2c24be-a3c7-47fd-aedc-abaa46413c42 | mastering-the-game-of-stratego-with-model | 2206.15378 | null | https://arxiv.org/abs/2206.15378v1 | https://arxiv.org/pdf/2206.15378v1.pdf | Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning | We introduce DeepNash, an autonomous agent capable of learning to play the imperfect information game Stratego from scratch, up to a human expert level. Stratego is one of the few iconic board games that Artificial Intelligence (AI) has not yet mastered. This popular game has an enormous game tree on the order of $10^{535}$ nodes, i.e., $10^{175}$ times larger than that of Go. It has the additional complexity of requiring decision-making under imperfect information, similar to Texas hold'em poker, which has a significantly smaller game tree (on the order of $10^{164}$ nodes). Decisions in Stratego are made over a large number of discrete actions with no obvious link between action and outcome. Episodes are long, with often hundreds of moves before a player wins, and situations in Stratego can not easily be broken down into manageably-sized sub-problems as in poker. For these reasons, Stratego has been a grand challenge for the field of AI for decades, and existing AI methods barely reach an amateur level of play. DeepNash uses a game-theoretic, model-free deep reinforcement learning method, without search, that learns to master Stratego via self-play. The Regularised Nash Dynamics (R-NaD) algorithm, a key component of DeepNash, converges to an approximate Nash equilibrium, instead of 'cycling' around it, by directly modifying the underlying multi-agent learning dynamics. DeepNash beats existing state-of-the-art AI methods in Stratego and achieved a yearly (2022) and all-time top-3 rank on the Gravon games platform, competing with human expert players. | ['Karl Tuyls', 'Demis Hassabis', 'Satinder Singh', 'David Silver', 'Remi Munos', 'Nathalie Beauguerlange', 'Laurent SIfre', 'Edward Lockhart', 'Shayegan Omidshafiei', 'Bilal Piot', 'Jean-Baptiste Lespiau', 'Marc Lanctot', 'Mark Rowland', 'Tom Eccles', 'Toby Pohlen', 'Finbarr Timbers', 'Sherjil Ozair', 'Mina Khan', 'Aleksandra Malysheva', 'Audrunas Gruslys', 'Zhe Wang', 'Sarah H. Cen', 'Romuald Elie', 'Stephen Mcaleer', 'Thomas Anthony', 'Neil Burch', 'Jerome T. Connor', 'Paul Muller', 'Vincent de Boer', 'Florian Strub', 'Eugene Tarassov', 'Daniel Hennes', 'Bart De Vylder', 'Julien Perolat'] | 2022-06-30 | null | null | null | null | ['board-games'] | ['playing-games'] | [-2.46059015e-01 3.85604203e-01 5.85264973e-02 4.87237751e-01
-6.15256846e-01 -8.04991663e-01 3.15533400e-01 -6.65296987e-02
-7.43839920e-01 1.02858818e+00 -2.54879683e-01 -6.10075235e-01
-5.97384632e-01 -1.21467936e+00 -5.47786415e-01 -6.14912331e-01
-5.32964230e-01 1.11694312e+00 5.40671051e-01 -1.17247331e+00
1.84983194e-01 -2.40231916e-01 -1.45949423e+00 -2.12625578e-01
6.72582448e-01 8.43942404e-01 -1.93588156e-02 9.65730369e-01
2.50931889e-01 1.29321682e+00 -6.25288844e-01 -4.95997190e-01
7.89408982e-01 -6.85175896e-01 -8.69832873e-01 -4.86327708e-01
-3.98918509e-01 -2.83951521e-01 -5.94042242e-01 1.03624809e+00
6.08995974e-01 1.83092341e-01 2.54749537e-01 -1.23339856e+00
-9.68487710e-02 1.01768327e+00 -5.00928760e-01 2.12487906e-01
2.66950399e-01 6.72171295e-01 1.50541782e+00 -3.68567277e-03
5.89948714e-01 9.67950702e-01 6.70578301e-01 6.57194912e-01
-8.29726696e-01 -6.88098609e-01 9.53420550e-02 2.73128569e-01
-1.16650224e+00 1.26651436e-01 3.14341128e-01 -1.53492287e-01
1.16927660e+00 2.18576938e-01 1.39827907e+00 6.23860657e-01
4.29595172e-01 7.92348146e-01 1.23341453e+00 -2.22019866e-01
7.82960713e-01 -8.51548314e-01 -6.09533727e-01 7.19010293e-01
3.19233924e-01 6.12773538e-01 -4.72531945e-01 -5.12037054e-02
9.00865972e-01 -3.56862366e-01 2.85997331e-01 -3.11698437e-01
-8.50357115e-01 1.10747540e+00 2.88200021e-01 1.46769658e-01
-4.37910527e-01 5.86858988e-01 2.62302160e-01 7.47828543e-01
-6.68301508e-02 1.12586355e+00 -5.27724445e-01 -8.97049904e-01
-6.52079344e-01 9.30498242e-01 1.19345963e+00 4.56143767e-01
7.59953678e-01 2.77947575e-01 4.40455407e-01 3.73140723e-01
-1.65944353e-01 2.16018781e-01 5.37796080e-01 -1.37492919e+00
2.21997470e-01 6.74999356e-01 2.77363956e-01 -8.01279783e-01
-6.04155958e-01 -3.90241683e-01 -7.87045956e-01 9.22687769e-01
6.33520246e-01 -6.98376656e-01 -6.65391147e-01 1.73339999e+00
1.48829684e-01 1.54601529e-01 1.21719033e-01 7.81829596e-01
2.83924609e-01 6.87339783e-01 -2.80427635e-01 -4.45969962e-02
1.15467989e+00 -8.12718809e-01 -1.48736641e-01 -7.96613753e-01
5.18994272e-01 -1.86603174e-01 7.11052001e-01 7.81761885e-01
-1.66330850e+00 2.43743602e-02 -1.26284313e+00 5.21216571e-01
-2.76990622e-01 -1.02499890e+00 1.14562595e+00 6.77125692e-01
-1.24530673e+00 7.35070646e-01 -8.04188550e-01 2.79850252e-02
4.52016801e-01 7.74820685e-01 -2.06544980e-01 -1.72731355e-01
-1.64870477e+00 1.08870506e+00 6.16551101e-01 -1.43005252e-01
-1.20667481e+00 -6.41824305e-01 -6.82232857e-01 2.15542197e-01
1.35782623e+00 -7.23644793e-01 1.62856019e+00 -7.78580010e-01
-1.88766861e+00 7.45131969e-01 6.61238909e-01 -7.01206684e-01
8.08113098e-01 1.97405010e-01 2.90989857e-02 -1.09029196e-01
2.58284152e-01 4.41640824e-01 3.59914184e-01 -7.81547129e-01
-1.08135974e+00 -1.83684260e-01 8.37902308e-01 7.98737526e-01
2.92972594e-01 -8.45792666e-02 -2.13947520e-01 -2.73483396e-01
3.92749086e-02 -9.61157024e-01 -1.01744771e+00 -6.00594699e-01
4.50719297e-02 -4.42927301e-01 -1.79972351e-01 -9.03754160e-02
1.26356411e+00 -1.59462392e+00 3.66414130e-01 2.16829002e-01
6.14812851e-01 1.33137181e-01 -1.16466984e-01 7.42681444e-01
2.71951020e-01 2.46545956e-01 -6.32887110e-02 4.08810794e-01
3.63161355e-01 4.27688986e-01 3.97532582e-01 1.36440694e-01
-2.65305459e-01 1.20262074e+00 -1.34652591e+00 7.33500794e-02
-8.65705907e-02 -5.78296721e-01 -9.79542971e-01 -7.78563023e-02
-4.02994573e-01 -1.93367794e-01 -4.88941938e-01 3.59999537e-01
2.49326348e-01 -2.08404988e-01 6.24099851e-01 8.11995566e-01
-2.67505884e-01 2.51875371e-01 -1.50371289e+00 1.51660943e+00
-1.80417180e-01 3.42950165e-01 2.27043346e-01 -1.25724626e+00
5.83976030e-01 1.56747162e-01 7.29183137e-01 -9.39988077e-01
3.85472447e-01 4.91850257e-01 7.56534636e-01 3.59610915e-02
4.00523245e-01 -5.07855177e-01 -7.57416725e-01 6.67875886e-01
-7.61058107e-02 -8.04518402e-01 6.16777718e-01 2.15454906e-01
2.03094149e+00 -2.07433775e-01 5.92075109e-01 -2.43213385e-01
-4.68125977e-02 5.29835522e-01 8.94473374e-01 1.45378494e+00
-3.09018433e-01 8.04743692e-02 1.01518333e+00 -8.14008951e-01
-1.06971359e+00 -1.03190613e+00 6.65138483e-01 1.26815069e+00
3.66580993e-01 -6.11820161e-01 -8.28116953e-01 -1.10178843e-01
-1.23629663e-02 2.33542621e-01 -7.33866274e-01 -3.06516081e-01
-6.86544359e-01 -8.54121685e-01 6.44761920e-01 2.80664206e-01
8.41142058e-01 -1.47512770e+00 -8.25140357e-01 8.47146690e-01
1.04840703e-01 -6.62582695e-01 1.83012802e-02 5.35988629e-01
-2.94238895e-01 -1.36730015e+00 -2.56670147e-01 -6.94985330e-01
-8.18056911e-02 -1.03768572e-01 1.46318436e+00 2.47997746e-01
-3.18280071e-01 1.24093033e-01 -3.50941539e-01 -5.84764183e-01
-4.39659834e-01 1.94838926e-01 2.28822708e-01 -7.69512594e-01
2.57312417e-01 -7.92768955e-01 -6.61003292e-01 3.13640654e-01
-7.38230109e-01 6.74439520e-02 6.36945188e-01 1.12445664e+00
1.46767780e-01 6.77168667e-01 6.04556918e-01 -6.99493110e-01
8.24851513e-01 -4.84081060e-01 -8.68351460e-01 -1.30167067e-01
-3.56934577e-01 -1.68182105e-01 8.12751234e-01 -2.31104240e-01
-4.48394179e-01 -1.95178479e-01 -9.13359597e-02 1.48896426e-01
3.71929765e-01 7.69928336e-01 1.99966729e-02 -1.10019013e-01
8.48129511e-01 2.12616414e-01 1.59082860e-01 1.18326172e-01
2.26755977e-01 2.36926869e-01 4.23121184e-01 -6.17490470e-01
9.08955216e-01 1.44452900e-01 6.43340824e-03 -3.36232275e-01
-6.97768211e-01 -4.92296368e-02 -1.83724925e-01 -3.53573203e-01
7.02662647e-01 -8.47360134e-01 -1.59527206e+00 8.73627126e-01
-4.37730849e-01 -9.77662504e-01 -6.39847815e-01 1.40546829e-01
-1.04023564e+00 -2.44676650e-01 -8.25507343e-01 -5.76215982e-01
2.81516239e-02 -8.78614426e-01 1.42495394e-01 5.42579830e-01
-3.50263380e-02 -7.33807743e-01 4.07053888e-01 5.51590860e-01
5.11517346e-01 2.44827971e-01 8.16176474e-01 -6.02837801e-01
-7.44559824e-01 -3.06390792e-01 1.57342866e-01 2.55516712e-02
-1.39624953e-01 -3.95546526e-01 -7.73706213e-02 -4.27356035e-01
-1.52610093e-01 -6.42946124e-01 4.27196175e-01 4.48774785e-01
5.23712218e-01 -2.47647971e-01 4.91771065e-02 1.98621437e-01
1.44650340e+00 7.20171988e-01 7.22150862e-01 1.01426852e+00
2.83410817e-01 1.47094071e-01 4.98079777e-01 6.95426881e-01
5.98615289e-01 4.63274777e-01 9.76837218e-01 3.53104502e-01
4.23419982e-01 -2.63889998e-01 3.24759126e-01 7.09387898e-01
-5.43328345e-01 -2.23936304e-01 -1.36482882e+00 4.38597769e-01
-2.10463381e+00 -1.28234541e+00 4.47319478e-01 1.88328946e+00
8.88658226e-01 8.48960996e-01 4.37296450e-01 2.33683810e-01
4.37028289e-01 1.50032535e-01 -9.12411630e-01 -6.94620490e-01
-1.05959319e-01 6.85722947e-01 6.63476288e-01 5.57798564e-01
-7.77391016e-01 1.52990794e+00 6.69492149e+00 1.17985499e+00
-6.98711812e-01 -1.58701554e-01 5.41261554e-01 -4.01797116e-01
-2.06706822e-02 1.03068426e-01 -2.30225161e-01 3.91175777e-01
8.11762214e-01 -7.26799786e-01 1.25031233e+00 8.25260222e-01
1.59286261e-01 -4.01254773e-01 -7.05531895e-01 9.39671338e-01
-4.20511425e-01 -1.79274464e+00 -4.47826058e-01 4.25387800e-01
1.01985729e+00 2.57868141e-01 -4.71574366e-02 8.44317675e-01
1.57485354e+00 -1.53974652e+00 7.10124195e-01 -1.21547177e-01
4.27628875e-01 -1.27381217e+00 7.80842066e-01 7.79610097e-01
-1.17106962e+00 -7.07289755e-01 -3.35176885e-01 -1.07170236e+00
-1.81703456e-02 -4.43902873e-02 -4.46044654e-01 4.73115593e-01
7.99036086e-01 2.98444092e-01 -4.68334779e-02 1.00660360e+00
-2.81189829e-01 5.12316942e-01 -5.90771139e-01 -4.93857592e-01
9.18265998e-01 -2.72611350e-01 5.36503971e-01 1.93221405e-01
7.94897079e-02 7.74386346e-01 3.88224572e-01 3.65791053e-01
-1.93976954e-01 -1.47842199e-01 -4.21204120e-01 -2.16417834e-01
5.21812201e-01 8.51024210e-01 -8.10032606e-01 -6.34527206e-02
-7.95533359e-02 7.18988299e-01 3.06491345e-01 -1.72266558e-01
-6.41348541e-01 -3.54127645e-01 8.94189119e-01 1.29081635e-02
2.91801006e-01 -2.05851018e-01 -2.63697863e-01 -8.72053564e-01
-2.46706024e-01 -1.55858028e+00 3.67438138e-01 -5.15626967e-01
-1.15346980e+00 3.75894725e-01 -3.92662406e-01 -9.69122887e-01
-6.27427042e-01 -6.93978727e-01 -8.55081320e-01 4.50276226e-01
-8.62187982e-01 -6.11760557e-01 8.52130949e-02 4.97127503e-01
3.90213192e-01 -4.76678401e-01 7.09618747e-01 -6.69157431e-02
-4.31658924e-01 3.31815630e-01 3.73113930e-01 1.43162578e-01
-9.16712284e-02 -1.46224535e+00 6.92291439e-01 4.07159001e-01
-1.23580337e-01 9.81512200e-03 6.82354391e-01 -4.25244659e-01
-1.52197433e+00 -2.47359499e-01 2.30042443e-01 -3.95071507e-01
1.15811479e+00 -3.21746349e-01 -2.72199929e-01 3.52208704e-01
8.01660717e-02 -1.60776794e-01 3.15541059e-01 1.60625577e-01
1.05274156e-01 -6.33018687e-02 -9.00984406e-01 1.08718407e+00
1.27040160e+00 -2.67181873e-01 -5.60552359e-01 -2.00430658e-02
4.58969772e-01 -7.42898643e-01 -5.02427042e-01 1.62312478e-01
6.83346450e-01 -1.20341194e+00 8.29010606e-01 -8.86623502e-01
5.32021999e-01 -1.28057361e-01 1.00535139e-01 -1.61310279e+00
-4.99419063e-01 -1.34046054e+00 1.65453658e-01 3.42642099e-01
3.70054871e-01 -5.32065511e-01 1.51055205e+00 3.99039358e-01
5.49231879e-02 -1.17455864e+00 -1.24884713e+00 -8.74326885e-01
6.99639559e-01 -4.08471853e-01 7.16721117e-01 7.41762280e-01
3.27079803e-01 2.28114918e-01 -5.37596643e-01 -3.19687992e-01
4.98423934e-01 -1.25506580e-01 6.80741310e-01 -1.21777523e+00
-8.48282576e-01 -8.47742260e-01 -6.85016394e-01 -8.94448340e-01
-2.16183230e-01 -5.03303826e-01 1.27950743e-01 -1.39939487e+00
5.29400706e-02 -6.17189348e-01 -3.21667403e-01 5.78924835e-01
-6.84762821e-02 3.39433908e-01 5.32219708e-01 -1.09504700e-01
-1.00555253e+00 2.50794649e-01 1.44799280e+00 -2.11145699e-01
-2.94059336e-01 1.07522428e-01 -1.04754436e+00 9.38849270e-01
1.05377507e+00 -4.33827549e-01 -3.33939165e-01 -1.05731077e-01
1.14484227e+00 4.94037271e-01 -5.07437550e-02 -1.20594406e+00
4.29426134e-01 -6.66936994e-01 -2.25087374e-01 5.53908795e-02
2.67045319e-01 -2.98705578e-01 1.66479379e-01 1.00837731e+00
-2.27589998e-02 3.98643762e-01 1.68886259e-01 3.58246624e-01
-1.37235865e-01 -3.16848874e-01 6.78735137e-01 -7.70455420e-01
-9.06536818e-01 3.56118500e-01 -1.12167645e+00 7.31333256e-01
1.29533041e+00 -3.94922674e-01 -2.83401698e-01 -8.40196431e-01
-6.58828795e-01 6.44561470e-01 4.20216352e-01 5.75171188e-02
2.15415075e-01 -9.47552502e-01 -7.90527165e-01 -4.08226043e-01
-3.83023143e-01 2.72095025e-01 4.72724199e-01 2.57477462e-01
-9.67021763e-01 -5.10791279e-02 -7.15076566e-01 1.22292995e-01
-7.68149853e-01 1.29296467e-01 7.58779347e-01 -1.08237541e+00
-4.76535439e-01 9.61105406e-01 -9.57506076e-02 -6.75342619e-01
-1.38455018e-01 1.40075102e-01 2.33251005e-02 -2.93081831e-02
4.76192087e-01 4.23913509e-01 -1.25934079e-01 3.71305309e-02
-3.32524091e-01 1.01919651e-01 5.72844557e-02 -4.40879583e-01
1.63755679e+00 2.85694420e-01 -2.96476353e-02 -4.78977105e-03
3.03514719e-01 -3.33194435e-01 -1.32450080e+00 9.85362306e-02
-1.00016244e-01 -1.63851276e-01 -1.18742615e-01 -1.06669354e+00
-8.37554157e-01 5.34018040e-01 6.80161193e-02 6.26679361e-01
8.62264156e-01 -1.94229335e-01 7.20029116e-01 7.44852245e-01
9.77864563e-01 -1.47298312e+00 4.19900149e-01 1.21443737e+00
3.91801775e-01 -1.03925145e+00 -4.18057367e-02 2.30795249e-01
-6.21396482e-01 7.88132250e-01 8.23631823e-01 -5.73962271e-01
4.70129877e-01 4.33189422e-01 -2.13021971e-02 -2.87388146e-01
-1.15425014e+00 -4.80327785e-01 -5.18563509e-01 6.38747752e-01
-3.69058162e-01 2.97030568e-01 -2.28637144e-01 7.76381195e-01
-8.30137730e-01 3.57538387e-02 8.27483118e-01 1.20073795e+00
-7.89984345e-01 -1.25520957e+00 -7.86519498e-02 5.35208583e-01
-4.96796727e-01 -1.67795926e-01 -2.48404801e-01 9.82653081e-01
1.82444036e-01 8.51361454e-01 -1.59637053e-02 -5.30610025e-01
2.75565058e-01 -3.59862417e-01 5.46473920e-01 -5.39491117e-01
-8.21124315e-01 -3.72937322e-01 1.18532412e-01 -8.15118253e-01
1.27371758e-01 -5.77317715e-01 -1.22041571e+00 -1.21696627e+00
1.94783378e-02 4.62502241e-01 1.40006810e-01 1.16450810e+00
1.68949144e-03 3.44048828e-01 5.62395990e-01 -9.11814988e-01
-7.38508999e-01 -3.49618912e-01 -8.70140791e-01 -2.69022621e-02
-1.61388427e-01 -5.58238328e-01 -2.42780864e-01 -8.94124269e-01] | [3.525688886642456, 1.5252134799957275] |
7f8ee8b8-0c55-4d12-9686-f0744bd6ce24 | a-neural-comprehensive-ranker-ncr-for-open | 1709.10204 | null | http://arxiv.org/abs/1709.10204v2 | http://arxiv.org/pdf/1709.10204v2.pdf | A Neural Comprehensive Ranker (NCR) for Open-Domain Question Answering | This paper proposes a novel neural machine reading model for open-domain
question answering at scale. Existing machine comprehension models typically
assume that a short piece of relevant text containing answers is already
identified and given to the models, from which the models are designed to
extract answers. This assumption, however, is not realistic for building a
large-scale open-domain question answering system which requires both deep text
understanding and identifying relevant text from corpus simultaneously.
In this paper, we introduce Neural Comprehensive Ranker (NCR) that integrates
both passage ranking and answer extraction in one single framework. A Q&A
system based on this framework allows users to issue an open-domain question
without needing to provide a piece of text that must contain the answer.
Experiments show that the unified NCR model is able to outperform the
states-of-the-art in both retrieval of relevant text and answer extraction. | ['Hao Ma', 'Bin Bi'] | 2017-09-29 | null | null | null | null | ['passage-ranking'] | ['natural-language-processing'] | [ 2.46499866e-01 4.20514822e-01 1.56757593e-01 -4.03747052e-01
-1.53643799e+00 -6.98891044e-01 4.07386214e-01 8.30000520e-01
-6.47407889e-01 6.74889266e-01 4.27488178e-01 -5.74897408e-01
-4.32295173e-01 -1.04096127e+00 -6.42387629e-01 2.42145240e-01
4.16827261e-01 8.62454355e-01 6.58071041e-01 -7.69003868e-01
5.85929871e-01 -2.92834520e-01 -1.58375049e+00 6.82466030e-01
1.38269651e+00 9.29560781e-01 5.47762215e-01 1.14463353e+00
-8.13647151e-01 1.30032277e+00 -6.89373255e-01 -4.38453853e-01
-4.99964841e-02 -5.94897509e-01 -1.67017782e+00 -6.66956842e-01
7.79842079e-01 -7.06370950e-01 -2.47335672e-01 7.64500022e-01
4.94566202e-01 3.53590280e-01 7.34777927e-01 -4.25562561e-01
-9.15492296e-01 5.77553689e-01 7.75529668e-02 5.21759748e-01
8.73407781e-01 -1.62025660e-01 1.56222308e+00 -8.25593233e-01
5.65645635e-01 9.90153253e-01 5.75765930e-02 8.10130000e-01
-8.01975191e-01 -2.75990982e-02 1.18221208e-01 3.82992774e-01
-8.17240536e-01 -3.15595597e-01 6.49029315e-01 -6.25232980e-02
1.23675323e+00 3.93389404e-01 1.08681463e-01 7.61016488e-01
6.66560307e-02 1.01512897e+00 7.67894983e-01 -7.45646894e-01
3.29012603e-01 6.98691830e-02 1.11960733e+00 5.09891629e-01
1.49037302e-01 -4.54963684e-01 -5.19103944e-01 -1.05713621e-01
1.50316194e-01 -3.16073418e-01 -4.44253862e-01 1.11029066e-01
-6.84054434e-01 9.12126005e-01 3.51101875e-01 2.44310394e-01
-4.63847458e-01 -2.30419740e-01 4.53216553e-01 7.66570687e-01
4.15722191e-01 9.71832812e-01 -7.01934457e-01 -1.09733358e-01
-1.06114423e+00 5.87004662e-01 1.44490921e+00 7.80486345e-01
6.19196773e-01 -7.57277489e-01 -6.44836962e-01 9.01243567e-01
2.75654227e-01 3.66581619e-01 4.68599677e-01 -9.48751032e-01
7.59693444e-01 9.83067572e-01 2.66225547e-01 -9.29656804e-01
-2.86324054e-01 -5.36266923e-01 -2.89052308e-01 -5.20829678e-01
6.03413045e-01 -2.03353584e-01 -6.15333557e-01 1.46843076e+00
1.97876871e-01 -3.61323059e-01 4.87816662e-01 9.40595627e-01
1.50551474e+00 9.36425447e-01 -1.29220381e-01 -9.95737687e-03
1.59621632e+00 -1.07932305e+00 -8.19302857e-01 -4.24745440e-01
5.34738839e-01 -7.02788591e-01 1.33209229e+00 2.42351487e-01
-1.25899327e+00 -6.01559579e-01 -1.00255311e+00 -8.34477305e-01
-4.76501822e-01 3.26114781e-02 4.64812629e-02 2.18550563e-01
-1.00663745e+00 2.59277433e-01 -3.59992832e-02 -4.47060108e-01
1.16133034e-01 1.01686813e-01 1.13126785e-01 -5.61482847e-01
-1.51342928e+00 1.03700244e+00 1.19149968e-01 -1.85740933e-01
-8.54010701e-01 -4.27366376e-01 -6.45936549e-01 5.48258603e-01
6.24377191e-01 -1.14552653e+00 1.92486811e+00 -6.38944387e-01
-1.35409272e+00 7.39299297e-01 -5.30484080e-01 -4.63706791e-01
7.20503228e-03 -7.85799444e-01 -1.85850158e-01 7.44788170e-01
2.53961623e-01 5.00863969e-01 7.87426531e-01 -1.05474198e+00
-5.88324070e-01 -5.71457267e-01 7.14841783e-01 4.24183875e-01
-3.64266396e-01 2.13190258e-01 -3.74611408e-01 -7.48146996e-02
1.57097057e-01 -1.93187028e-01 -8.24093372e-02 -4.40331548e-01
-2.13930622e-01 -8.48234892e-01 3.91396761e-01 -1.08526266e+00
1.53292370e+00 -1.36348116e+00 2.31919270e-02 3.88524719e-02
5.76029003e-01 2.43389264e-01 -6.51760519e-01 8.75110567e-01
3.55417609e-01 9.92917269e-02 -3.76580399e-03 6.59564286e-02
2.88482189e-01 -5.66316135e-02 -7.72187293e-01 -3.26970309e-01
4.23746146e-02 1.15398121e+00 -1.09055197e+00 -4.93672103e-01
-3.56704056e-01 -8.87328982e-02 -5.33633590e-01 6.80806458e-01
-9.14520442e-01 6.48913980e-02 -8.18167686e-01 4.64512259e-01
4.11754221e-01 -5.04833579e-01 -2.67567307e-01 5.05673945e-01
3.05696607e-01 1.01176929e+00 -7.75027692e-01 1.61318862e+00
-5.09037316e-01 4.91368234e-01 9.16042998e-02 -7.72288978e-01
9.23713446e-01 3.44128937e-01 -1.12579077e-01 -1.03692174e+00
1.38382554e-01 3.65464211e-01 -9.99991819e-02 -9.42198694e-01
8.30386758e-01 1.36548474e-01 -4.32078615e-02 8.60508502e-01
2.50720024e-01 -2.26346672e-01 5.20475388e-01 7.55606413e-01
1.49791038e+00 3.87288886e-03 1.61252305e-01 -2.56996512e-01
9.22347605e-01 2.33599097e-01 -1.62787214e-01 1.24697328e+00
-4.49302010e-02 4.18947250e-01 5.88742614e-01 -1.28470600e-01
-8.06683481e-01 -9.66392279e-01 3.01393509e-01 1.49750507e+00
6.08073268e-03 -5.31414390e-01 -9.96198773e-01 -8.59295607e-01
-3.21498156e-01 9.00498152e-01 -4.11525309e-01 -2.02113576e-02
-6.32347107e-01 1.19996272e-01 2.69140929e-01 3.46642435e-01
4.31520432e-01 -1.09918189e+00 -5.82430780e-01 3.77679557e-01
-8.00635338e-01 -7.50425696e-01 -3.19995373e-01 1.48771927e-01
-9.58498418e-01 -1.17292070e+00 -8.15926373e-01 -9.82793927e-01
4.74384725e-01 3.41748536e-01 1.79111063e+00 4.86516595e-01
3.09328228e-01 6.65729284e-01 -7.85822690e-01 -5.21330059e-01
-3.85777712e-01 6.53314829e-01 -5.09903789e-01 -3.44636112e-01
9.18990731e-01 -2.61167288e-01 -6.73933327e-01 2.13060696e-02
-9.58392024e-01 -2.07135916e-01 4.18888271e-01 6.64848387e-01
2.37374768e-01 -3.09873641e-01 1.37956893e+00 -7.12276816e-01
1.52487040e+00 -6.95095122e-01 -4.74175870e-01 6.65919602e-01
-3.51015836e-01 1.88678458e-01 7.52866447e-01 -5.03057316e-02
-1.17749178e+00 -4.48814571e-01 -5.49895525e-01 3.13887447e-01
-2.61190385e-01 9.61038053e-01 -4.11082730e-02 3.90852541e-01
9.80185866e-01 2.70136416e-01 -1.76758200e-01 -6.88895762e-01
5.33644259e-01 9.44688439e-01 3.49563330e-01 -6.10875666e-01
6.29985690e-01 -8.55952352e-02 -5.49099505e-01 -7.89573908e-01
-1.57631421e+00 -1.01271772e+00 -6.22835338e-01 -2.07541555e-01
8.62463474e-01 -6.54726923e-01 -4.43167150e-01 7.97913596e-02
-1.52478302e+00 -6.45766640e-03 -3.69467646e-01 2.19981223e-02
-3.63742262e-01 3.33773255e-01 -6.71619415e-01 -7.45018065e-01
-7.95137107e-01 -6.33583128e-01 8.26816201e-01 4.59358811e-01
-4.76384610e-01 -8.81542683e-01 2.78194398e-01 1.07015789e+00
5.67655265e-01 -5.94641507e-01 1.22160625e+00 -1.27180791e+00
-7.43155837e-01 -4.93159175e-01 -2.41329536e-01 3.03954512e-01
-3.87286425e-01 -6.08580351e-01 -1.04470813e+00 4.75308448e-02
3.01265270e-01 -8.10259759e-01 1.10972607e+00 1.24065235e-01
9.43592906e-01 -3.47373843e-01 9.47789103e-02 -2.95924723e-01
1.19243670e+00 -1.80553839e-01 6.53331995e-01 2.53206670e-01
2.16093808e-01 1.19026506e+00 4.96649712e-01 2.72758435e-02
8.78781438e-01 5.23367450e-02 3.36255103e-01 2.79087871e-01
9.63260531e-02 -4.57695365e-01 2.36912981e-01 1.17680728e+00
3.74815613e-01 -3.76256377e-01 -8.20864201e-01 8.05148721e-01
-1.71269608e+00 -8.38554323e-01 -2.99256116e-01 1.92191172e+00
1.03735459e+00 7.91678205e-02 -2.33936742e-01 9.88827925e-03
2.19279557e-01 6.61995336e-02 -5.32656968e-01 -5.38583457e-01
1.68869615e-01 7.51615524e-01 -2.04407081e-01 7.22541988e-01
-7.75099337e-01 8.32433462e-01 5.70131779e+00 7.43180037e-01
-3.13990325e-01 1.10770367e-01 3.36479068e-01 1.71295986e-01
-6.14151359e-01 1.51657149e-01 -9.25429344e-01 -1.41335860e-01
1.26534259e+00 -1.60673037e-01 2.23620936e-01 5.51460683e-01
2.42358595e-02 -4.81756538e-01 -1.28127611e+00 4.28636014e-01
4.26590383e-01 -1.13044393e+00 3.36134523e-01 -3.90194327e-01
4.74714309e-01 -1.33770511e-01 -3.40181202e-01 8.26515436e-01
2.42100969e-01 -9.91689503e-01 2.75930941e-01 7.98056185e-01
1.12538502e-01 -5.94633520e-01 9.50317383e-01 9.92473304e-01
-6.96860611e-01 -2.71750629e-01 -5.79900682e-01 -4.59297657e-01
1.17771536e-01 3.01807284e-01 -4.86512631e-01 4.94892329e-01
5.52539885e-01 1.24451324e-01 -8.89260650e-01 9.49430585e-01
-6.49426937e-01 7.76949823e-01 -2.07046106e-01 -7.27765560e-01
3.29288244e-01 8.93620998e-02 3.96504015e-01 6.80568039e-01
1.02035433e-01 4.77928579e-01 -1.16322562e-02 1.00999618e+00
-4.51273590e-01 6.15373135e-01 -4.14411992e-01 -5.17907459e-03
3.05329770e-01 1.06037343e+00 -3.56981009e-01 -5.77404857e-01
-6.59483552e-01 1.01806748e+00 6.54568613e-01 3.98891181e-01
-2.75643378e-01 -7.05107927e-01 2.77835131e-03 9.28507820e-02
2.27991343e-01 -4.92001213e-02 -2.27739617e-01 -1.27132308e+00
3.84252727e-01 -1.13191628e+00 7.14866340e-01 -9.52348351e-01
-1.44262743e+00 5.78695774e-01 -2.42077067e-01 -8.21461022e-01
-4.42576706e-01 -5.18833101e-01 -6.00821912e-01 1.17408526e+00
-2.00790071e+00 -8.50866914e-01 -2.49364942e-01 4.66029227e-01
8.61938119e-01 5.11949882e-02 1.00291955e+00 1.64302364e-01
-2.12245613e-01 3.11594367e-01 6.70294985e-02 1.42363384e-01
6.40792608e-01 -1.43263686e+00 2.17401356e-01 8.16875577e-01
1.75453275e-01 9.96658683e-01 5.35640359e-01 -5.90953767e-01
-1.36068463e+00 -6.90199435e-01 1.63010883e+00 -9.68928754e-01
4.95712370e-01 -1.33368358e-01 -1.22403193e+00 3.43794137e-01
8.46449077e-01 -6.01770043e-01 8.87001812e-01 2.86648422e-01
-2.38241479e-01 -2.80125588e-02 -8.19292545e-01 5.59505105e-01
4.17977065e-01 -9.19589639e-01 -1.62834156e+00 3.34312648e-01
1.07178152e+00 -1.19989373e-01 -6.67729735e-01 1.96633548e-01
2.01158419e-01 -6.00120008e-01 8.92660916e-01 -9.27135229e-01
9.19164658e-01 -1.06502295e-01 -1.69476718e-01 -1.05004716e+00
7.44635090e-02 -4.17982459e-01 -5.99364102e-01 9.92321730e-01
7.11794436e-01 -3.32589924e-01 4.47084665e-01 8.11873794e-01
-9.82401222e-02 -8.06482792e-01 -8.71060729e-01 -2.42885888e-01
6.30902350e-01 -3.59564632e-01 5.14186561e-01 4.76551354e-01
3.30085754e-01 1.28014851e+00 2.16779888e-01 4.33138423e-02
3.16798389e-01 2.61451900e-01 7.42492497e-01 -1.42996407e+00
-5.81255928e-02 -3.67817283e-01 5.14050841e-01 -1.85645318e+00
1.46429777e-01 -8.81928504e-01 9.66778845e-02 -2.30296850e+00
2.71847248e-01 1.49643272e-01 -2.66388327e-01 -3.79284732e-02
-5.97082317e-01 -3.16711903e-01 -3.71877593e-03 9.28330347e-02
-1.23222375e+00 5.27984858e-01 1.38659441e+00 -1.99069425e-01
-6.59473240e-02 1.85251966e-01 -1.13742495e+00 7.21605778e-01
7.03529954e-01 -3.68248492e-01 -6.38882756e-01 -6.52538180e-01
6.75965071e-01 5.45043528e-01 2.50521839e-01 -8.09084356e-01
6.46342516e-01 1.97001055e-01 2.03830913e-01 -8.57796311e-01
1.40275717e-01 -6.32055879e-01 -1.14654863e+00 7.11768046e-02
-1.05738378e+00 -1.32275966e-03 1.60437841e-02 5.41560173e-01
-3.78161222e-01 -1.01371872e+00 3.04752856e-01 -5.10539174e-01
-5.75439215e-01 -7.62451962e-02 -6.02319539e-01 6.23336554e-01
2.91811675e-01 3.45693529e-02 -6.60371482e-01 -9.43978906e-01
-4.92010564e-01 7.26058006e-01 -1.61985546e-01 5.57489097e-01
1.02305257e+00 -7.77149141e-01 -1.00389075e+00 -2.98757970e-01
3.63435805e-01 7.28711393e-03 4.70301211e-01 3.36527228e-01
-4.22346085e-01 1.00024426e+00 2.70221353e-01 -5.15551120e-02
-1.10386407e+00 5.32268405e-01 2.85188019e-01 -8.92651081e-01
-3.92724961e-01 8.96400392e-01 -1.14018366e-01 -8.58411491e-01
4.36431617e-01 -5.20329058e-01 -1.07735896e+00 2.11422011e-01
9.71888840e-01 2.13548213e-01 3.12663168e-01 -8.24212730e-02
2.28141040e-01 2.48111933e-01 -2.42073581e-01 -2.19829321e-01
1.06125712e+00 -4.57432866e-01 -4.01927471e-01 3.01768064e-01
9.58909869e-01 -8.32530707e-02 -5.54537833e-01 -5.67168653e-01
3.94521147e-01 -3.86798531e-02 5.87143051e-03 -1.15254676e+00
-1.65379271e-01 1.16445720e+00 1.88962102e-01 3.82420927e-01
1.05860019e+00 2.08248034e-01 1.33807814e+00 1.14986956e+00
1.19016826e-01 -1.26266038e+00 3.94326806e-01 1.16558301e+00
9.82435226e-01 -1.32214010e+00 -2.62712240e-01 -5.64429862e-03
-2.23421887e-01 1.15075207e+00 9.19714928e-01 2.66356822e-02
3.95754516e-01 -3.98619622e-01 2.01970011e-01 -6.87437832e-01
-1.15600753e+00 -4.84749615e-01 8.25290799e-01 2.87787139e-01
5.07539749e-01 -2.75223851e-01 -4.94268447e-01 1.02015853e+00
-3.22950304e-01 -1.05728157e-01 6.09690309e-01 1.06120288e+00
-1.13883686e+00 -9.39434946e-01 -1.97295070e-01 7.64849484e-01
-6.31250679e-01 -5.69079280e-01 -7.33739972e-01 2.05991447e-01
-4.73091841e-01 1.51382399e+00 -1.31934002e-01 -4.41019237e-02
4.82531518e-01 4.81176347e-01 2.35834718e-01 -1.01056969e+00
-9.55983043e-01 -4.04047251e-01 3.75294030e-01 -3.01999927e-01
-1.44643158e-01 -1.79000467e-01 -1.10602498e+00 1.12884328e-01
-5.67457378e-01 5.63660204e-01 2.85698414e-01 1.13207221e+00
4.10345018e-01 4.61417168e-01 2.55766630e-01 7.39387944e-02
-9.46228445e-01 -1.29231036e+00 -1.99984044e-01 6.20218217e-02
6.08581126e-01 -1.64013822e-02 -1.57221839e-01 -1.28500372e-01] | [11.33544635772705, 7.987379550933838] |
ea33b670-4125-4e17-b7a5-97fab85195da | learning-to-compose-with-professional | 1702.00503 | null | http://arxiv.org/abs/1702.00503v2 | http://arxiv.org/pdf/1702.00503v2.pdf | Learning to Compose with Professional Photographs on the Web | Photo composition is an important factor affecting the aesthetics in
photography. However, it is a highly challenging task to model the aesthetic
properties of good compositions due to the lack of globally applicable rules to
the wide variety of photographic styles. Inspired by the thinking process of
photo taking, we formulate the photo composition problem as a view finding
process which successively examines pairs of views and determines their
aesthetic preferences. We further exploit the rich professional photographs on
the web to mine unlimited high-quality ranking samples and demonstrate that an
aesthetics-aware deep ranking network can be trained without explicitly
modeling any photographic rules. The resulting model is simple and effective in
terms of its architectural design and data sampling method. It is also generic
since it naturally learns any photographic rules implicitly encoded in
professional photographs. The experiments show that the proposed view finding
network achieves state-of-the-art performance with sliding window search
strategy on two image cropping datasets. | ['Kwan-Liu Ma', 'Shao-Yi Chien', 'Yi-Ling Chen', 'Jan Klopp', 'Min Sun'] | 2017-02-01 | null | null | null | null | ['image-cropping'] | ['computer-vision'] | [ 3.57503444e-01 -1.75530314e-01 -4.92068604e-02 -7.33709216e-01
-5.84458888e-01 -7.68443465e-01 4.72810686e-01 -3.63415152e-01
2.04560161e-01 2.30326634e-02 4.43589658e-01 -7.95245990e-02
-4.57154453e-01 -7.48615623e-01 -8.78864050e-01 -4.29735392e-01
3.28857780e-01 2.27375194e-01 -3.63208711e-01 -3.67289573e-01
4.33686316e-01 4.85092849e-01 -2.04759669e+00 7.50066042e-01
8.73129129e-01 1.47174919e+00 2.56155908e-01 6.38524354e-01
-2.21408289e-02 7.64104187e-01 -3.00209820e-01 -9.84094620e-01
7.92577565e-01 -3.56170654e-01 -8.20538878e-01 6.31628513e-01
1.07562244e+00 -5.15973449e-01 -8.82214010e-02 1.11311042e+00
4.10801530e-01 1.53680354e-01 7.40828753e-01 -1.07624471e+00
-1.46808290e+00 3.65626276e-01 -5.97798645e-01 -5.01098216e-01
2.11560011e-01 1.94509581e-01 1.60109770e+00 -7.60476530e-01
6.89241946e-01 1.20454538e+00 2.68063635e-01 4.51725364e-01
-8.14243019e-01 -3.69748682e-01 2.88424760e-01 4.89330918e-01
-9.57059622e-01 -3.65213722e-01 1.10152829e+00 -2.74641067e-01
4.85315830e-01 5.68458498e-01 1.10385358e+00 9.87305820e-01
-5.92186116e-02 9.35919046e-01 1.30116963e+00 -3.48004192e-01
3.31906527e-01 6.24934807e-02 -3.65091085e-01 6.23564839e-01
-1.42532036e-01 -5.86168841e-02 -5.94747663e-01 1.95752114e-01
8.34400773e-01 2.61066705e-01 -2.08176538e-01 -7.30770350e-01
-7.84693420e-01 6.02844536e-01 6.81351781e-01 -2.48738259e-01
-2.77105451e-01 2.02015430e-01 2.10996255e-01 5.16247332e-01
1.55627266e-01 9.33160961e-01 -3.48336577e-01 1.74553737e-01
-6.45447493e-01 -1.67478155e-02 4.76656497e-01 1.22768104e+00
8.02968562e-01 -1.10247768e-01 1.64079845e-01 1.27290618e+00
4.13516760e-02 2.93179631e-01 8.70630667e-02 -1.36956692e+00
2.21941531e-01 8.91535163e-01 6.14750795e-02 -1.19468355e+00
2.47557629e-02 -2.42347822e-01 -9.72707629e-01 5.99759936e-01
2.93228682e-02 3.88494968e-01 -7.92602956e-01 1.49823713e+00
1.62277073e-01 -6.69452846e-01 -1.04187094e-01 1.05520225e+00
8.83744657e-01 4.45329159e-01 -7.54948109e-02 -9.67633631e-03
1.45328033e+00 -1.19526780e+00 -3.46413910e-01 -3.32126111e-01
1.30112264e-02 -1.09921980e+00 1.69030595e+00 6.95767403e-01
-1.27210402e+00 -6.07760787e-01 -1.23691988e+00 -5.53068995e-01
-3.86577904e-01 6.40667558e-01 8.53944719e-01 5.58142722e-01
-1.15687764e+00 6.18058145e-01 -4.70485687e-02 -5.19658685e-01
7.66725183e-01 2.60879964e-01 -3.35038632e-01 -3.58125627e-01
-7.84836769e-01 6.05660200e-01 -1.09823801e-01 1.19373433e-01
-7.59043992e-01 -7.31373072e-01 -5.36413848e-01 1.70121685e-01
4.51486111e-01 -1.05996561e+00 1.13978565e+00 -1.79827392e+00
-1.78758764e+00 1.05989480e+00 1.29036739e-01 1.38437763e-01
5.29577017e-01 -2.01661304e-01 -4.01446283e-01 2.67701060e-01
-2.31653064e-01 7.61264265e-01 1.07782352e+00 -1.61103475e+00
-5.56363821e-01 -2.38555670e-01 7.99982071e-01 5.18785894e-01
-8.44406545e-01 -1.56109869e-01 -5.54064274e-01 -6.61109388e-01
-7.62507319e-02 -9.58625317e-01 -1.56590179e-01 6.06683552e-01
-3.41107309e-01 1.21871931e-02 2.20990106e-01 -4.84847933e-01
1.06104553e+00 -1.95206439e+00 -5.02030998e-02 1.75977170e-01
9.63822976e-02 -1.68268561e-01 -4.51036185e-01 6.36982977e-01
-6.76864460e-02 2.16835946e-01 1.98625863e-01 -8.47947001e-02
2.82779008e-01 -2.15874892e-02 -3.28703731e-01 -3.53424437e-03
9.47834775e-02 8.45530808e-01 -8.40185881e-01 -3.09200257e-01
2.29547590e-01 2.42670298e-01 -5.37585914e-01 3.29078794e-01
-3.44008267e-01 -1.75688997e-01 -3.43384206e-01 8.82535696e-01
7.93690979e-01 -5.49675226e-01 4.62035328e-01 -6.88082635e-01
8.60606730e-02 3.69112454e-02 -9.87996459e-01 1.96298432e+00
-9.95868385e-01 7.35242486e-01 -3.44116062e-01 -3.81533980e-01
9.39492762e-01 -1.48134336e-01 2.10678160e-01 -7.37967670e-01
1.67591479e-02 -7.55308345e-02 -2.27563560e-01 -7.81157315e-01
7.17195570e-01 -9.76760611e-02 7.11138919e-02 7.19739139e-01
-3.34385723e-01 -2.46265426e-01 -5.87953366e-02 2.98878718e-02
8.79731953e-01 1.62410423e-01 4.53303039e-01 -2.55812556e-01
3.48658770e-01 -9.99656096e-02 4.00845081e-01 3.43728006e-01
1.46068066e-01 9.11428630e-01 2.94599146e-01 -9.56309617e-01
-1.25642085e+00 -1.19100201e+00 3.27352792e-01 1.27260375e+00
4.52247322e-01 -3.92780811e-01 -5.94453156e-01 -4.62102801e-01
-3.82303536e-01 5.19365132e-01 -8.53809237e-01 -4.02555987e-02
-3.00723284e-01 -3.00626129e-01 -1.79137856e-01 3.79673928e-01
6.84223831e-01 -1.09344137e+00 -6.68473542e-01 -3.60750824e-01
-1.53399035e-01 -9.12692487e-01 -8.05533826e-01 -3.45799118e-01
-6.57343268e-01 -1.22761393e+00 -5.27442753e-01 -8.37145567e-01
9.65447545e-01 7.23337293e-01 1.60518026e+00 1.62245229e-01
-3.14989209e-01 4.58757788e-01 -4.70147401e-01 -2.65273392e-01
-3.58351201e-01 -9.57388133e-02 -3.22084367e-01 3.37689251e-01
2.17453957e-01 -8.21700573e-01 -1.37052393e+00 3.81039143e-01
-1.16892326e+00 3.77513260e-01 1.01792562e+00 6.10099256e-01
6.11797214e-01 2.43790060e-01 1.60080284e-01 -8.89743090e-01
6.78770959e-01 8.08189735e-02 -3.56326759e-01 9.94050682e-01
-6.18571997e-01 -4.38761478e-03 7.45629072e-01 -3.56825083e-01
-1.16582191e+00 1.72051862e-01 4.17531580e-01 -5.95823705e-01
-9.75778922e-02 5.69256097e-02 -6.05341315e-01 -1.45665482e-01
4.40813035e-01 1.15969526e-02 -1.37874395e-01 -3.59045267e-01
7.43075669e-01 3.88948739e-01 3.40590745e-01 -5.73462427e-01
6.91070378e-01 7.19313204e-01 -8.60815719e-02 -5.62142849e-01
-9.98664260e-01 -1.65097386e-01 -4.83118802e-01 -5.89991629e-01
7.91341186e-01 -7.99266815e-01 -8.71859729e-01 2.98316181e-01
-1.00262392e+00 -2.89436746e-02 -2.99474835e-01 -1.46851569e-01
-7.37463176e-01 4.80938256e-01 -2.20893368e-01 -7.38864839e-01
-4.67746764e-01 -1.08181417e+00 1.09356916e+00 1.97955981e-01
-1.23262137e-01 -6.43947721e-01 -2.62563020e-01 5.88444054e-01
1.44376859e-01 2.53864914e-01 1.34799719e+00 2.32263416e-01
-1.06494272e+00 -2.78155487e-02 -4.24642384e-01 5.21333337e-01
2.14468315e-01 3.69941831e-01 -1.10511875e+00 1.10876277e-01
3.62153575e-02 -5.82454503e-01 6.63598597e-01 3.48348200e-01
1.59105074e+00 -5.84710300e-01 2.86734670e-01 7.68973589e-01
1.85353076e+00 3.55016626e-02 8.22187483e-01 4.63751286e-01
7.35538423e-01 1.10659695e+00 4.13324147e-01 4.07303095e-01
5.15778422e-01 5.49174190e-01 6.36765599e-01 -1.85376585e-01
-2.10602343e-01 -4.48445559e-01 2.59592563e-01 8.07325244e-01
-1.34961277e-01 -2.44690701e-01 -5.28811991e-01 4.92365181e-01
-1.74372518e+00 -1.07848191e+00 1.75758421e-01 2.05184507e+00
6.97987497e-01 -2.20067799e-01 8.91094506e-02 -1.70372128e-01
6.76580846e-01 2.54979342e-01 -7.05623746e-01 -8.21794450e-01
-3.31125885e-01 -4.55583669e-02 3.38537425e-01 9.88935605e-02
-1.05997515e+00 7.37350881e-01 6.42480087e+00 8.87695491e-01
-7.38203347e-01 -2.25972444e-01 1.12538385e+00 -3.91291499e-01
-7.24926710e-01 -8.56795311e-02 -1.92540720e-01 2.48941898e-01
1.27390549e-01 -3.42537765e-03 9.88671780e-01 9.29426551e-01
1.87988281e-01 6.79228380e-02 -1.29493213e+00 1.35450447e+00
2.97255933e-01 -1.48780310e+00 6.44749343e-01 -2.81266049e-02
1.17820501e+00 -5.70871294e-01 5.53296149e-01 -3.85537714e-01
5.18436253e-01 -1.15288651e+00 8.63415897e-01 5.14844120e-01
9.97803390e-01 -8.46855938e-01 5.74467592e-02 -1.85559154e-01
-1.11748588e+00 -5.05067408e-01 -6.13001347e-01 -1.19653016e-01
2.57591158e-02 4.19131517e-01 -3.77094418e-01 4.63115782e-01
9.45059240e-01 8.13488543e-01 -7.48819947e-01 9.09941256e-01
-2.51037955e-01 -6.95592389e-02 1.78440530e-02 -1.36795074e-01
1.59024149e-01 -3.77495438e-01 -3.07159834e-02 5.09646475e-01
6.08953118e-01 8.10955763e-02 -2.75783479e-01 8.54070187e-01
-4.77441877e-01 2.14939564e-01 -5.29351413e-01 1.45305634e-01
2.57589430e-01 1.50315952e+00 -6.61175787e-01 -2.19818465e-02
-3.41320008e-01 1.08841240e+00 4.25049841e-01 2.83701122e-01
-5.46548486e-01 1.20729968e-01 8.30689609e-01 6.90067932e-02
4.76540715e-01 2.36809567e-01 -6.71848714e-01 -1.05176985e+00
3.24408352e-01 -1.20761812e+00 2.68909693e-01 -1.33590412e+00
-1.95541787e+00 8.75682950e-01 -2.87520736e-01 -1.83755636e+00
2.04604775e-01 -7.53940344e-01 -8.17530155e-01 4.42997158e-01
-1.53775084e+00 -1.57371843e+00 -6.93623662e-01 3.83594990e-01
8.88682485e-01 -2.07375586e-01 6.56236291e-01 6.97213365e-03
-1.54079407e-01 5.92325926e-01 1.51502952e-01 -2.11853474e-01
7.74605572e-01 -1.33385897e+00 4.27513659e-01 6.23745680e-01
2.24607974e-01 6.09995902e-01 5.78901708e-01 -1.27439886e-01
-1.46321273e+00 -8.40954304e-01 6.54879093e-01 -3.63304585e-01
4.95998681e-01 -3.58560115e-01 -2.72087395e-01 3.89246829e-02
6.76792741e-01 -4.79443401e-01 9.29442525e-01 9.61995423e-02
-7.22546577e-01 -5.53532004e-01 -7.99644411e-01 9.86403704e-01
1.57379413e+00 -4.33248997e-01 -3.09918463e-01 5.07642269e-01
4.72905278e-01 6.96359277e-02 -7.43108928e-01 9.88453925e-02
9.71765995e-01 -1.46215653e+00 1.26021886e+00 -5.50077200e-01
1.19455612e+00 -7.74084106e-02 -2.61010617e-01 -1.26134229e+00
-4.23249036e-01 -7.02623248e-01 3.85779262e-01 1.32593071e+00
2.64141053e-01 -1.57162830e-01 8.74276280e-01 9.39794183e-01
1.96409851e-01 -1.04262030e+00 -5.42545021e-01 -5.03952920e-01
-2.31644064e-01 -2.07442358e-01 1.04719234e+00 6.72552109e-01
-4.25178736e-01 2.59423316e-01 -6.75217152e-01 -5.38861342e-02
5.63740790e-01 9.13939178e-01 8.60431552e-01 -1.17514408e+00
-4.77682173e-01 -6.85443640e-01 -6.98234364e-02 -1.05696094e+00
-1.41233414e-01 -5.99901795e-01 3.81557606e-02 -1.58869243e+00
3.97419065e-01 -4.08656985e-01 -2.57599622e-01 -2.48928778e-02
-5.73568493e-02 3.34870785e-01 3.53300810e-01 3.29637468e-01
-7.95420766e-01 6.35365725e-01 1.68199980e+00 -3.25228781e-01
1.26405358e-01 -1.68980435e-01 -1.30665064e+00 8.36171508e-01
6.81361914e-01 3.22913751e-02 -1.00063574e+00 -9.29310501e-01
1.00413668e+00 -3.31399202e-01 4.06427830e-01 -7.20623493e-01
-4.20542317e-04 -4.56225812e-01 5.97884834e-01 -3.54907691e-01
4.74254996e-01 -1.10722792e+00 3.13735932e-01 -2.06084803e-01
-7.23256588e-01 4.64729160e-01 -3.16710323e-01 7.52720952e-01
-3.37696932e-02 -1.19535439e-01 6.08891368e-01 -4.94207054e-01
-9.59772944e-01 4.33960110e-01 8.95295814e-02 2.69206297e-02
8.05011034e-01 -4.95512605e-01 -4.23441827e-01 -6.24123096e-01
-2.83513576e-01 3.37313339e-02 1.13342202e+00 5.99437356e-01
8.78349245e-01 -1.62529635e+00 -2.83977449e-01 6.13507703e-02
4.81155276e-01 -2.06654996e-01 6.74406350e-01 1.63428724e-01
-7.41995275e-01 -1.85913816e-01 -5.56032419e-01 -2.65017599e-01
-1.31293344e+00 7.05423474e-01 2.04576835e-01 -1.01123869e-01
-3.48169506e-01 8.90908301e-01 7.03622162e-01 -2.20294356e-01
1.59639046e-01 -2.16506854e-01 -1.55851975e-01 -4.67310213e-02
4.48720813e-01 3.49252045e-01 -1.48847312e-01 -3.19461912e-01
-1.67221529e-03 9.83554125e-01 -1.91768864e-03 1.38628185e-01
1.64728665e+00 -4.17993307e-01 -2.18011439e-01 2.05067307e-01
1.26952112e+00 -2.15177566e-01 -1.44650495e+00 8.38972777e-02
-3.40688437e-01 -8.83127630e-01 -1.41157448e-01 -1.06439996e+00
-1.24988985e+00 9.36677337e-01 5.02586722e-01 1.19161576e-01
1.71971416e+00 -1.82075784e-01 8.72447312e-01 4.39075291e-01
1.88307598e-01 -1.54712760e+00 4.38774139e-01 -1.20840929e-01
1.37867093e+00 -1.34515500e+00 2.12750584e-01 -5.19145608e-01
-9.71387565e-01 1.35641575e+00 6.97874248e-01 -2.70727456e-01
5.52735746e-01 -3.65094393e-01 3.76089096e-01 -3.84719998e-01
-8.88312697e-01 2.31415089e-02 7.35880733e-01 5.93696952e-01
1.94540173e-01 3.05594075e-02 -3.41591090e-02 4.85297531e-01
-4.47434902e-01 -1.72450677e-01 5.05567908e-01 3.89050722e-01
-1.14587069e-01 -1.10892880e+00 7.69237950e-02 4.39571112e-01
-3.27800363e-02 -2.68698901e-01 -9.43055332e-01 5.23957193e-01
1.82107583e-01 8.41279685e-01 -1.86625510e-01 -5.89218080e-01
3.00409377e-01 -2.86753654e-01 5.66789269e-01 -2.39080727e-01
-6.85985684e-01 -1.87411875e-01 5.01744747e-02 -7.46277153e-01
-6.81170464e-01 -6.03110254e-01 -3.46740723e-01 -5.49573004e-01
6.62721768e-02 -4.07723933e-01 5.18602788e-01 4.45727170e-01
5.83943963e-01 3.24682146e-01 9.74088967e-01 -5.85080385e-01
-4.92612988e-01 -3.59823883e-01 -5.78712165e-01 8.17567170e-01
9.67067182e-02 -4.45672661e-01 -1.51002541e-01 3.36759299e-01] | [11.500914573669434, -0.9595103859901428] |
c64d51cd-38bf-4944-9d30-1f04271c5bc0 | convolutional-neural-networks-based-focal | 2001.03329 | null | https://arxiv.org/abs/2001.03329v1 | https://arxiv.org/pdf/2001.03329v1.pdf | Convolutional Neural Networks based Focal Loss for Class Imbalance Problem: A Case Study of Canine Red Blood Cells Morphology Classification | Morphologies of red blood cells are normally interpreted by a pathologist. It is time-consuming and laborious. Furthermore, a misclassified red blood cell morphology will lead to false disease diagnosis and improper treatment. Thus, a decent pathologist must truly be an expert in classifying red blood cell morphology. In the past decade, many approaches have been proposed for classifying human red blood cell morphology. However, those approaches have not addressed the class imbalance problem in classification. A class imbalance problem---a problem where the numbers of samples in classes are very different---is one of the problems that can lead to a biased model towards the majority class. Due to the rarity of every type of abnormal blood cell morphology, the data from the collection process are usually imbalanced. In this study, we aimed to solve this problem specifically for classification of dog red blood cell morphology by using a Convolutional Neural Network (CNN)---a well-known deep learning technique---in conjunction with a focal loss function, adept at handling class imbalance problem. The proposed technique was conducted on a well-designed framework: two different CNNs were used to verify the effectiveness of the focal loss function and the optimal hyper-parameters were determined by 5-fold cross-validation. The experimental results show that both CNNs models augmented with the focal loss function achieved higher $F_{1}$-scores, compared to the models augmented with a conventional cross-entropy loss function that does not address class imbalance problem. In other words, the focal loss function truly enabled the CNNs models to be less biased towards the majority class than the cross-entropy did in the classification task of imbalanced dog red blood cell data. | ['Supawit Vatathanavaro', 'Suchat Tungjitnob', 'Kitsuchart Pasupa'] | 2020-01-10 | null | null | null | null | ['morphology-classification'] | ['computer-vision'] | [-2.86770552e-01 -2.71162577e-02 1.04144901e-01 -4.99036700e-01
-5.38826548e-02 -1.49096893e-02 -1.72043428e-01 5.03327072e-01
-5.15551329e-01 8.86605382e-01 -3.16318363e-01 -8.65890533e-02
-1.34290263e-01 -1.10570133e+00 -3.94946128e-01 -9.83612061e-01
-8.71122107e-02 5.71399152e-01 -1.36597678e-01 -2.14747995e-01
2.76795596e-01 6.90929592e-01 -1.53763354e+00 2.54983693e-01
1.10315919e+00 1.10845053e+00 -4.35833216e-01 3.48680347e-01
-3.35553646e-01 9.38425660e-01 -8.05406332e-01 -4.99977231e-01
3.69935870e-01 -5.82713366e-01 -4.90003049e-01 -2.22937554e-01
3.17275703e-01 -3.00376505e-01 -1.48015052e-01 1.06829357e+00
7.61192560e-01 -1.35213673e-01 8.18139732e-01 -1.31176138e+00
-3.96377742e-01 3.47804695e-01 -5.47210515e-01 4.82522547e-01
-8.31689537e-02 2.76778966e-01 6.89101279e-01 -3.29953730e-01
2.96811432e-01 1.25313783e+00 8.44250500e-01 4.37869340e-01
-1.09551907e+00 -8.72459471e-01 -2.85335630e-01 2.12383837e-01
-1.25298274e+00 2.01290585e-02 8.46010745e-01 -5.44459522e-01
7.28357255e-01 1.30728245e-01 1.02573979e+00 3.62144440e-01
6.43692017e-01 5.38329184e-01 9.36797917e-01 -1.01011962e-01
9.15577263e-02 2.64324903e-01 2.53866196e-01 6.02544010e-01
8.25975537e-01 2.27344930e-01 -5.25522418e-03 7.83811137e-02
5.19769251e-01 8.30826387e-02 -2.20981047e-01 -3.16178292e-01
-7.02175140e-01 8.58985662e-01 5.07297873e-01 5.50249219e-01
-2.11482584e-01 -1.63458928e-01 7.53602445e-01 4.52342778e-01
3.82069081e-01 7.91303813e-01 -4.24435675e-01 1.69642076e-01
-6.68173909e-01 3.82744670e-01 1.07311296e+00 2.51566768e-01
5.72499275e-01 2.98683316e-01 -1.42837465e-01 8.81475151e-01
1.00486614e-01 1.95508078e-01 7.38729358e-01 -3.80779803e-01
7.83968344e-02 1.10751081e+00 -2.71256179e-01 -1.49988246e+00
-6.69795334e-01 -6.22715890e-01 -1.24512029e+00 4.07769412e-01
7.50317633e-01 7.69515485e-02 -8.38062763e-01 1.71315885e+00
4.54090923e-01 -3.15222085e-01 1.50798813e-01 9.99399066e-01
9.49195385e-01 3.06972653e-01 1.36111796e-01 -1.35796830e-01
1.33777392e+00 -3.77522767e-01 -9.30007756e-01 2.52317846e-01
7.55827129e-01 -5.27285218e-01 8.42486739e-01 3.73049796e-01
-9.64036345e-01 -4.97109324e-01 -1.37069821e+00 1.33779377e-01
-5.23657203e-01 9.01156813e-02 4.30486828e-01 6.47152841e-01
-7.13959277e-01 7.05415666e-01 -5.32959044e-01 -3.12707514e-01
7.65598834e-01 4.72602099e-01 -5.45985937e-01 1.19811170e-01
-1.23667932e+00 1.09336400e+00 5.39279759e-01 2.77726173e-01
-4.10384178e-01 -9.30225551e-01 -6.77279651e-01 1.11950792e-01
-2.76314050e-01 -5.36118627e-01 7.63944447e-01 -1.04691660e+00
-1.05153358e+00 1.03146851e+00 3.78377169e-01 -4.63675320e-01
7.27231503e-01 1.33212388e-01 -3.45028520e-01 -6.65747821e-02
-1.68111786e-01 3.91687721e-01 2.35344961e-01 -1.09244752e+00
-4.15204674e-01 -6.42398655e-01 -2.63773710e-01 -5.96937165e-02
-8.84792358e-02 -3.98309022e-01 4.09387648e-01 -6.92749381e-01
1.59568593e-01 -5.34846842e-01 6.22660704e-02 1.62367359e-01
-3.59036058e-01 -1.37485683e-01 8.09522510e-01 -5.83116055e-01
1.07704508e+00 -2.08174062e+00 -2.90798485e-01 1.25237003e-01
3.98758292e-01 6.89237595e-01 2.01530978e-01 -5.21309897e-02
-3.97722334e-01 2.49559090e-01 -2.40085185e-01 1.61323696e-01
-1.96600869e-01 1.01231888e-01 1.95966557e-01 6.87565327e-01
3.76837522e-01 5.34362197e-01 -7.16855347e-01 -6.24917924e-01
2.92601377e-01 6.35651410e-01 -3.89084905e-01 4.16538388e-01
6.69376971e-03 2.62561470e-01 -1.60627171e-01 6.94294274e-01
1.07168984e+00 -9.13947374e-02 2.71587912e-02 -4.98174101e-01
2.41681844e-01 -2.74183363e-01 -9.53096807e-01 8.68459821e-01
-1.63036570e-01 6.05521739e-01 -1.54983133e-01 -1.26840377e+00
1.22368717e+00 1.92856789e-01 5.62525630e-01 -8.72632384e-01
4.89459366e-01 4.52104151e-01 5.62571585e-01 -8.35822463e-01
1.11826867e-01 -6.37396038e-01 2.46473417e-01 1.41929239e-01
-1.36347979e-01 1.38806477e-01 3.27894956e-01 -3.60552311e-01
8.64222407e-01 -3.47476691e-01 3.51109058e-01 -3.98670495e-01
7.89151728e-01 8.65069628e-02 1.09377587e+00 3.07024896e-01
-6.58895016e-01 4.48500812e-01 7.56473660e-01 -1.03314221e+00
-1.06285369e+00 -7.97626495e-01 -4.55451220e-01 2.68769920e-01
8.32739025e-02 4.65548694e-01 -7.24590778e-01 -6.62160456e-01
4.09631342e-01 4.98116702e-01 -7.03614056e-01 -4.41361725e-01
-6.38403952e-01 -1.25700712e+00 9.51699734e-01 3.47757667e-01
9.45576429e-01 -1.09435129e+00 -7.78914094e-01 1.54615149e-01
-4.09663916e-02 -6.48574769e-01 -2.21821386e-03 2.42417499e-01
-9.65579033e-01 -1.56497014e+00 -5.35944998e-01 -8.15983176e-01
8.46391082e-01 -2.69875616e-01 1.06061506e+00 5.68779230e-01
-8.02128673e-01 -4.62247491e-01 -3.24399292e-01 -5.73480904e-01
-6.35907054e-01 -2.50837952e-02 -1.16025709e-01 1.28020695e-03
8.42464745e-01 -2.72532463e-01 -8.37848961e-01 2.35894293e-01
-1.11043203e+00 -3.44162583e-01 6.58501089e-01 1.07699251e+00
3.88666809e-01 4.01160598e-01 1.05599523e+00 -1.03369594e+00
3.69644850e-01 -3.45178783e-01 -4.88294333e-01 1.47007585e-01
-7.09501147e-01 -3.03003155e-02 1.05926108e+00 -2.92288095e-01
-7.91339815e-01 -2.97503054e-01 -2.79118448e-01 -2.18489245e-01
-3.32647711e-01 2.61101037e-01 -3.36133629e-01 6.31734449e-03
5.68409085e-01 1.01308003e-01 5.73868215e-01 -1.88433319e-01
-4.39738214e-01 6.10852599e-01 2.19306946e-01 -1.63066253e-01
3.01119715e-01 4.33081806e-01 3.05797577e-01 -5.58976293e-01
-5.75772166e-01 -1.53686047e-01 -2.32660785e-01 -2.35626444e-01
8.39922249e-01 -6.18825495e-01 -9.24114823e-01 1.05640638e+00
-9.70439911e-01 -8.38973150e-02 -1.70092419e-01 5.18964946e-01
-2.80042797e-01 3.94899845e-01 -6.88627779e-01 -6.62495255e-01
-6.46429121e-01 -1.12178981e+00 2.69959748e-01 5.31801522e-01
-1.32891014e-01 -1.13171315e+00 6.99325427e-02 3.57309639e-01
5.99161923e-01 6.72431111e-01 1.47667432e+00 -1.10100257e+00
-4.78465706e-02 -4.78666544e-01 -3.48442733e-01 5.43899715e-01
3.11623156e-01 2.10888490e-01 -9.89923596e-01 -3.72956485e-01
2.81511068e-01 -3.36295784e-01 7.31037676e-01 5.67024112e-01
1.14816344e+00 -3.97735298e-01 -6.99379519e-02 7.50953913e-01
1.65883160e+00 6.53475940e-01 8.63913417e-01 3.66527259e-01
4.07502443e-01 6.80084169e-01 5.30961752e-01 4.37592417e-01
2.18357638e-01 3.00992072e-01 3.88890803e-01 -4.90121245e-01
-5.21598831e-02 1.82165466e-02 -2.65892506e-01 6.57003820e-01
1.85371771e-01 -3.75955373e-01 -7.71259069e-01 5.82495391e-01
-1.45402610e+00 -9.69340920e-01 -1.54838294e-01 2.12020326e+00
9.15726066e-01 -4.28234786e-02 9.87812728e-02 5.31620443e-01
7.43505657e-01 -1.82539359e-01 -7.91744828e-01 -5.60487568e-01
-2.62126178e-01 4.52438332e-02 3.94620001e-01 1.56085044e-01
-9.77483869e-01 4.15722758e-01 5.87746096e+00 6.43773079e-01
-1.47632432e+00 -4.25249815e-01 1.12432361e+00 -8.21481869e-02
3.21089067e-02 -4.40176338e-01 -6.57848835e-01 7.94076979e-01
3.98660004e-01 -1.12375572e-01 -1.09254137e-01 6.94567800e-01
1.43119514e-01 -2.51204908e-01 -1.10492170e+00 8.14450324e-01
1.41422331e-01 -9.93025064e-01 6.42863438e-02 -1.49993980e-02
3.77414912e-01 -4.86839414e-01 -6.17939755e-02 2.96564907e-01
1.09288275e-01 -1.20484638e+00 2.23895401e-01 4.14459020e-01
5.90291977e-01 -8.07348788e-01 1.50746858e+00 1.89057618e-01
-8.27103436e-01 -1.41418457e-01 -5.09572744e-01 -1.19611518e-02
-1.56073108e-01 9.73628104e-01 -8.29674006e-01 3.73894215e-01
7.87035346e-01 3.16969216e-01 -3.66713166e-01 1.24066901e+00
3.38180542e-01 2.32165635e-01 -1.23103134e-01 -2.21303597e-01
-2.48633504e-01 -8.67013112e-02 3.43893260e-01 1.14513445e+00
2.01108500e-01 1.11647122e-01 -1.65298566e-01 9.67839718e-01
-3.45977582e-02 2.67751455e-01 -5.75124919e-01 1.94371641e-01
3.73822451e-01 1.12373042e+00 -8.12111199e-01 -1.34753719e-01
4.49166149e-02 1.80100903e-01 2.14229971e-01 3.51879932e-02
-8.72037888e-01 -7.93635249e-01 6.55743003e-01 2.27619976e-01
-4.52138670e-02 6.10331833e-01 -5.91762722e-01 -9.07183290e-01
-7.01652542e-02 -7.71774709e-01 6.18198097e-01 -2.50408590e-01
-1.64547610e+00 6.69357717e-01 -2.16743127e-01 -1.22421181e+00
5.32518215e-02 -6.64793015e-01 -6.05765045e-01 7.83441663e-01
-1.72524273e+00 -7.09837437e-01 -6.91147923e-01 1.94076270e-01
1.06890447e-01 -3.87484044e-01 5.42739451e-01 6.34490073e-01
-8.48436534e-01 8.55866194e-01 -1.13335244e-01 4.96380627e-01
6.03633046e-01 -1.17684925e+00 -4.64154571e-01 5.18572927e-01
-8.04899395e-01 3.92535061e-01 7.49753952e-01 -5.63600183e-01
-8.96620750e-01 -1.05688894e+00 8.08108270e-01 2.01510623e-01
3.57656814e-02 1.54721275e-01 -1.20702112e+00 1.68059111e-01
-9.06784236e-02 2.08970770e-01 1.02601707e+00 -3.84416282e-01
-1.41857490e-01 -6.39052987e-01 -1.89322090e+00 1.14129461e-01
3.65260303e-01 3.42428088e-02 -6.19244814e-01 1.77445710e-02
2.73425251e-01 -3.70823443e-01 -1.00889015e+00 7.88832486e-01
6.72150850e-01 -1.25497091e+00 8.15489650e-01 -6.21826112e-01
5.78549087e-01 -4.02031422e-01 9.80900154e-02 -1.35157716e+00
-1.00935265e-01 2.75592476e-01 2.20391721e-01 1.03134501e+00
9.32252035e-02 -9.53865767e-01 7.98700631e-01 4.57305223e-01
-8.60735625e-02 -1.17356372e+00 -9.44049537e-01 -6.49048567e-01
4.61069763e-01 2.77581692e-01 8.54232728e-01 1.03460801e+00
-3.02903969e-02 -1.40418828e-01 2.83095930e-02 -3.37939709e-01
5.66778719e-01 -7.47635514e-02 7.24651814e-01 -1.42197192e+00
1.05796084e-01 -6.81454122e-01 -9.30402994e-01 -1.55865178e-01
1.09345242e-01 -7.73388863e-01 4.07189243e-02 -1.31008995e+00
1.96001023e-01 -7.13331938e-01 -3.41131210e-01 4.19510633e-01
-2.49319434e-01 1.50633126e-01 -5.39841093e-02 9.89294797e-03
-3.60790826e-02 3.85725796e-01 1.40586460e+00 -2.36270234e-01
-8.43223333e-02 -8.88866410e-02 -7.34311342e-01 5.60686648e-01
1.01595306e+00 -5.57500422e-01 -2.95068502e-01 -9.78843048e-02
2.21720770e-01 2.98830420e-02 2.93790936e-01 -1.14892137e+00
1.05728410e-01 -6.33137748e-02 6.33936286e-01 -6.51022971e-01
-2.23036539e-02 -7.88743198e-01 2.81041950e-01 9.43046272e-01
-2.22199365e-01 -1.72381699e-02 2.97515482e-01 1.60800636e-01
-5.33475757e-01 -1.77393675e-01 1.32624543e+00 -2.50877202e-01
-1.63046524e-01 2.78933853e-01 -2.71372557e-01 5.89639060e-02
1.33591974e+00 -5.72716475e-01 -5.82144499e-01 -4.05056700e-02
-3.79810095e-01 2.46024504e-01 5.39663672e-01 -5.73531550e-04
6.45150304e-01 -1.21051657e+00 -8.56949091e-01 5.01357198e-01
9.51141790e-02 2.29221046e-01 5.40058792e-01 9.06227052e-01
-1.17185855e+00 2.69379973e-01 -7.19576240e-01 -6.23524189e-01
-1.18539083e+00 3.54024231e-01 9.76657867e-01 -5.32715261e-01
-3.54443610e-01 7.63188541e-01 9.12740901e-02 -5.71941018e-01
1.81797251e-01 -4.06405956e-01 -4.44388598e-01 2.04365447e-01
3.77800584e-01 5.67258537e-01 3.05017740e-01 -5.35832286e-01
-2.75906593e-01 5.32333672e-01 -1.34166449e-01 6.77976608e-01
1.36015189e+00 2.56891012e-01 -4.38072026e-01 3.35027933e-01
1.41227782e+00 -5.41704118e-01 -7.89094925e-01 1.99630037e-01
-3.21044594e-01 -6.08464777e-01 -7.35131949e-02 -9.36864734e-01
-1.72108018e+00 9.52328205e-01 8.92796993e-01 3.62402648e-01
1.20855212e+00 -5.81070185e-01 7.88894236e-01 1.93360418e-01
2.00268120e-01 -1.16836965e+00 -8.45029112e-03 1.72788993e-01
5.38214207e-01 -1.34431815e+00 2.12486591e-02 -2.38857076e-01
-4.48492259e-01 1.31429029e+00 1.08806551e+00 -1.30759507e-01
7.88868308e-01 3.21401685e-01 3.52352053e-01 -3.59848678e-01
-5.10464847e-01 2.80328214e-01 -1.55962572e-01 5.36731958e-01
6.84873164e-01 -1.33519378e-02 -5.71294785e-01 5.86429000e-01
-2.38125265e-01 4.34500635e-01 5.17135203e-01 7.88305342e-01
-1.68386012e-01 -7.48224914e-01 -3.52920413e-01 8.98724377e-01
-6.27784789e-01 4.60271269e-01 -1.37576267e-01 8.63495946e-01
5.56255162e-01 8.50374341e-01 4.94530559e-01 -5.42705357e-01
3.48060071e-01 -2.03803163e-02 2.53712982e-01 -4.25914861e-02
-8.53641033e-01 -4.06470329e-01 -2.58190960e-01 -3.15288991e-01
-3.78811508e-01 -4.47775781e-01 -1.31782138e+00 -7.32975841e-01
-5.28961360e-01 1.33340493e-01 3.08900356e-01 7.53056467e-01
7.05349818e-02 6.24870896e-01 6.52185619e-01 -1.65433422e-01
-6.70998752e-01 -1.10735881e+00 -8.19647729e-01 6.96593165e-01
4.61879939e-01 -7.77744830e-01 -8.02278161e-01 -2.25366771e-01] | [15.115007400512695, -2.8220016956329346] |
75d11110-54ce-49db-8255-444705dcb1a1 | generative-domain-migration-hashing-for | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Jingyi_Zhang_Generative_Domain-Migration_Hashing_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Jingyi_Zhang_Generative_Domain-Migration_Hashing_ECCV_2018_paper.pdf | Generative Domain-Migration Hashing for Sketch-to-Image Retrieval | Due to the succinct nature of free-hand sketch drawings, sketch-based image retrieval (SBIR) has abundant practical use cases in consumer electronics. However, SBIR remains a long-standing unsolved problem mainly due to the significant discrepancy between the sketch domain and the image domain. In this work, we propose a Generative Domain-migration Hashing (GDH) approach, which for the first time generates hashing codes from synthetic natural images that are migrated from sketches. The generative model learns a mapping that the distributions of sketches can be indistinguishable from the distribution of natural images using an adversarial loss, and simultaneously learns an inverse mapping based on the cycle consistency loss in order to enhance the indistinguishability. With the robust mapping learned from the generative model, GDH can migrate sketches to their indistinguishable image counterparts while preserving the domain-invariant information of sketches. With an end-to-end multi-task learning framework, the generative model and binarized hashing codes can be jointly optimized. Comprehensive experiments of both category-level and fine-grained SBIR on multiple large-scale datasets demonstrate the consistently balanced superiority of GDH in terms of efficiency, memory costs and effectiveness. | ['Luc van Gool', 'Heng Tao Shen', 'Mengyang Yu', 'Fan Zhu', 'Li Liu', 'Jingyi Zhang', 'Fumin Shen', 'Ling Shao'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 1.22512549e-01 -4.20754075e-01 -2.41747439e-01 -3.09616655e-01
-1.15278459e+00 -8.29469323e-01 5.18465459e-01 -3.55965108e-01
-7.67416283e-02 5.44816613e-01 -3.84798087e-02 1.48730427e-01
-2.46358022e-01 -7.67714083e-01 -6.92007840e-01 -8.83280396e-01
1.92480922e-01 6.31345093e-01 1.66377276e-01 -1.01673126e-01
2.66269535e-01 4.26773459e-01 -1.43481159e+00 2.36670226e-01
4.76801187e-01 1.08164024e+00 2.57899821e-01 4.93994564e-01
-1.06048770e-01 3.76107037e-01 -4.03882742e-01 -6.20540142e-01
5.25473654e-01 -3.76088411e-01 -3.73752117e-01 -5.62281124e-02
7.70076156e-01 -7.71793723e-01 -9.81367052e-01 1.12756288e+00
7.15393960e-01 -1.92904435e-02 5.32619953e-01 -1.79551268e+00
-1.24615586e+00 2.02085644e-01 -5.70527136e-01 -3.41519445e-01
2.17370525e-01 3.99121605e-02 1.17646945e+00 -9.35587525e-01
7.92386949e-01 1.52137983e+00 3.30589443e-01 5.82416296e-01
-1.35667956e+00 -1.24992037e+00 2.01524459e-02 1.71386406e-01
-1.94533181e+00 -1.52156636e-01 9.85917568e-01 -1.84486702e-01
1.82629138e-01 -9.25952476e-03 2.41450578e-01 1.03594375e+00
-2.79773362e-02 9.95285988e-01 8.75150681e-01 -3.50445658e-02
2.21418649e-01 2.38393456e-01 -4.03445661e-01 7.51288533e-01
2.05414340e-01 2.33140707e-01 -7.08424151e-01 -3.75335008e-01
1.07300651e+00 5.03868163e-01 -1.14753470e-01 -8.00939858e-01
-1.24078906e+00 9.73051727e-01 5.84474564e-01 1.20852798e-01
-2.15620294e-01 3.70168120e-01 4.18983936e-01 3.60214084e-01
8.25566277e-02 4.33150232e-02 6.76187128e-02 3.34887981e-01
-9.48469400e-01 5.94885290e-01 5.03340900e-01 1.31963921e+00
9.17139769e-01 -3.11503083e-01 -3.79078984e-01 8.96645904e-01
2.02923939e-01 1.05348372e+00 3.72980684e-01 -5.07692456e-01
4.16027725e-01 2.98252672e-01 4.33760993e-02 -1.38354003e+00
5.59149146e-01 4.44380846e-03 -1.07034540e+00 5.21850735e-02
2.04280168e-01 5.05651057e-01 -6.57200336e-01 1.90813315e+00
2.47281209e-01 6.78481534e-02 -1.20962001e-01 1.14362955e+00
6.16434872e-01 8.01859796e-01 -6.02292679e-02 3.43922079e-01
1.28998303e+00 -9.09669995e-01 -6.25162005e-01 -1.44144267e-01
-1.28678367e-01 -7.77691483e-01 1.17172384e+00 1.91535726e-01
-9.95589256e-01 -7.86895812e-01 -1.27838051e+00 -3.28912944e-01
-2.37112641e-01 1.39806286e-01 3.72049451e-01 4.88133371e-01
-8.09705436e-01 3.95871639e-01 -5.17983854e-01 -6.04134090e-02
5.37472248e-01 3.00419301e-01 -5.12839854e-01 -6.18646204e-01
-1.14054883e+00 1.76303938e-01 4.21950132e-01 -2.32935086e-01
-1.03365064e+00 -7.26707816e-01 -8.57983410e-01 1.61228225e-01
1.54864952e-01 -5.09851515e-01 9.70131814e-01 -6.90080225e-01
-1.27398789e+00 8.80315304e-01 9.62989181e-02 -1.10429361e-01
6.36577606e-01 2.82474816e-01 -2.54989475e-01 4.23166037e-01
2.29758650e-01 9.37398553e-01 1.36833930e+00 -1.16942859e+00
-7.09743023e-01 -4.89291519e-01 -1.75462320e-01 -3.52156698e-03
-5.16416430e-01 -4.72450972e-01 -8.09733927e-01 -9.11632657e-01
3.54279019e-02 -1.17673850e+00 2.69386858e-01 8.09094667e-01
-2.22992599e-01 -8.04379284e-02 1.08699024e+00 -6.91849411e-01
1.12295616e+00 -2.45278502e+00 3.75037700e-01 1.51905447e-01
1.81238905e-01 1.64302569e-02 -5.12229145e-01 5.31172395e-01
1.38132647e-01 -2.20578790e-01 -2.31484339e-01 -3.35352451e-01
3.47393483e-01 1.48285151e-01 -7.88656354e-01 4.96382594e-01
1.90749899e-01 1.11342347e+00 -1.03125906e+00 -5.16112328e-01
1.35834783e-01 5.85732937e-01 -4.52781290e-01 4.67809200e-01
1.96082909e-02 9.80560258e-02 -4.16903377e-01 7.99269497e-01
1.20872223e+00 -2.58640409e-01 9.17579085e-02 -2.88362771e-01
4.28856939e-01 -2.83340812e-01 -1.09728205e+00 2.07122493e+00
-4.76260841e-01 4.73909289e-01 6.21824479e-03 -6.85413957e-01
1.09912491e+00 3.67015563e-02 1.48177490e-01 -1.05902886e+00
-4.25668746e-01 3.88536423e-01 -5.08606434e-01 1.37048930e-01
4.77144539e-01 -3.83220732e-01 -4.86618787e-01 6.31646335e-01
1.08139351e-01 -2.65816152e-01 -2.42165864e-01 3.69616777e-01
7.55400002e-01 -1.82734188e-02 -1.00765023e-02 -1.25448868e-01
5.24932683e-01 -3.83274853e-01 3.08175832e-01 6.60984695e-01
-8.45240280e-02 9.27588046e-01 1.14723019e-01 -6.29613817e-01
-1.38053501e+00 -1.52702236e+00 5.20135145e-05 9.48541582e-01
7.19001174e-01 -1.45826429e-01 -4.37871009e-01 -6.93591714e-01
5.30870616e-01 1.66498080e-01 -5.15574992e-01 -4.91741091e-01
-3.87463599e-01 -6.42314106e-02 6.28292501e-01 3.54370683e-01
7.16115296e-01 -1.01539326e+00 -1.02767691e-01 1.11133464e-01
-1.22800626e-01 -1.04226184e+00 -1.10814810e+00 -4.77350563e-01
-4.88256395e-01 -7.33326852e-01 -1.19845891e+00 -1.01674616e+00
6.90697193e-01 7.19043374e-01 9.16073143e-01 1.10295653e-01
-7.57734537e-01 3.90383571e-01 -1.17221750e-01 1.81574095e-02
-3.30893546e-01 2.52065714e-02 -2.01213658e-01 2.18337908e-01
1.73907876e-01 -4.97574806e-01 -8.77918005e-01 5.39489090e-01
-1.41694462e+00 -1.42121568e-01 8.42551768e-01 1.14281499e+00
7.75556386e-01 2.14595452e-01 4.21194494e-01 -5.08452415e-01
4.17481989e-01 -4.25109208e-01 -7.53536224e-01 5.11699677e-01
-5.35697520e-01 3.44805390e-01 6.20568931e-01 -7.33403862e-01
-8.15966845e-01 4.21493389e-02 1.58289909e-01 -9.67425704e-01
2.47106761e-01 -6.67414889e-02 -3.32908958e-01 -2.51141638e-01
1.40783921e-01 6.56741083e-01 1.79089680e-01 -1.91056833e-01
6.01544619e-01 6.54096305e-01 7.74312317e-01 -8.46230567e-01
1.27677882e+00 7.94108212e-01 -6.25134408e-02 -6.90358341e-01
-4.53237206e-01 -4.91504848e-01 -2.43609011e-01 1.00329675e-01
5.94433069e-01 -1.10357630e+00 -6.20646179e-01 5.94319463e-01
-1.08169568e+00 1.19186975e-01 -9.75133106e-02 5.80518916e-02
-5.54903269e-01 6.17982447e-01 -5.39383173e-01 -6.64649069e-01
-5.42756498e-01 -1.03474748e+00 1.64981461e+00 1.67835519e-01
2.13429183e-01 -7.06951857e-01 1.50045738e-01 6.62203655e-02
3.92544597e-01 1.96585879e-02 1.20297801e+00 -1.36172667e-01
-1.15440536e+00 -3.85152131e-01 -5.94258130e-01 3.44067156e-01
1.08779207e-01 -3.58232409e-01 -8.84795666e-01 -7.68594801e-01
-3.54323357e-01 -6.17111504e-01 7.84082890e-01 -1.31622240e-01
1.27777755e+00 -4.57984775e-01 -2.41635859e-01 6.37727082e-01
1.71547091e+00 1.02405317e-01 8.56403470e-01 4.54046428e-02
6.28980994e-01 3.46318990e-01 7.37961650e-01 5.87390780e-01
2.25704893e-01 7.49984562e-01 2.65833348e-01 -2.71488298e-02
-3.46613944e-01 -9.96933043e-01 9.66348574e-02 5.60634553e-01
7.39691854e-01 -1.11319780e-01 -3.43501121e-01 6.62389934e-01
-1.76512325e+00 -1.07981861e+00 7.17062414e-01 2.49710846e+00
7.24237680e-01 -1.41157329e-01 1.13234133e-01 -8.06110129e-02
1.06858671e+00 3.42418760e-01 -7.68504322e-01 1.96733773e-02
-1.01347215e-01 1.74043134e-01 4.39495534e-01 2.40522772e-01
-1.03025186e+00 7.55622566e-01 5.08673477e+00 1.37881100e+00
-1.16390455e+00 2.55698133e-02 5.03003955e-01 6.22693673e-02
-6.13600612e-01 -4.18099836e-02 -6.44458294e-01 8.24210882e-01
2.85423487e-01 -3.88906538e-01 8.31540585e-01 1.13802993e+00
-6.43389046e-01 4.06947672e-01 -1.16879833e+00 1.45197880e+00
1.21901140e-01 -1.33241856e+00 4.55747157e-01 2.01644614e-01
8.00540149e-01 -4.34553355e-01 4.85132724e-01 4.04898494e-01
3.12502682e-02 -8.55510235e-01 9.88244593e-01 2.95620382e-01
1.40417302e+00 -8.37273598e-01 4.75373894e-01 1.16995491e-01
-1.53395569e+00 -8.90482813e-02 -6.94429338e-01 4.29214597e-01
-1.75014675e-01 2.11948514e-01 -6.32934570e-01 4.97555703e-01
5.62146723e-01 5.40902436e-01 -3.54391754e-01 6.97470486e-01
4.59686741e-02 -1.13080136e-01 -1.60548076e-01 1.05671827e-02
2.32667416e-01 -3.29800963e-01 2.63337910e-01 1.03104842e+00
4.47743416e-01 -8.10924545e-02 1.23958260e-01 1.19700837e+00
-3.16919595e-01 -2.07343370e-01 -8.57748091e-01 -3.64831686e-01
8.83939624e-01 1.29290450e+00 -5.14985263e-01 -2.04512000e-01
-1.33013830e-01 1.59846401e+00 2.18909204e-01 2.47455075e-01
-8.38105142e-01 -7.84102976e-01 9.36393440e-01 -4.56612110e-02
7.53084660e-01 -8.28081146e-02 -9.25017241e-03 -1.10107970e+00
2.80926555e-01 -8.11450601e-01 2.94163316e-01 -6.54320359e-01
-1.72726750e+00 5.01151443e-01 -2.14868575e-01 -1.46242440e+00
-4.67877015e-02 -3.89097035e-01 -3.66790533e-01 8.85433972e-01
-1.72063279e+00 -1.38472795e+00 -5.05823016e-01 7.78005362e-01
3.44606876e-01 -2.09045127e-01 8.29598486e-01 5.31182289e-01
-1.19160436e-01 1.18195474e+00 4.73585963e-01 2.12686673e-01
9.75639582e-01 -1.03536582e+00 5.05168915e-01 4.28659797e-01
1.13648072e-01 6.38693929e-01 2.56295592e-01 -4.28993821e-01
-1.84732318e+00 -1.22098923e+00 5.48837602e-01 -6.91920444e-02
4.95261461e-01 -7.33680367e-01 -9.33594465e-01 2.37057716e-01
-1.62960246e-01 3.15361142e-01 4.34212834e-01 -4.72873986e-01
-1.15781415e+00 -4.96860206e-01 -1.31468642e+00 4.41321850e-01
9.72540557e-01 -1.20270026e+00 -2.62524128e-01 2.33675361e-01
6.54416561e-01 -2.26041242e-01 -8.92141521e-01 -7.75576606e-02
1.04405451e+00 -7.30514824e-01 1.40536416e+00 -3.48211646e-01
5.53737581e-01 -4.29818153e-01 -3.99925709e-01 -8.67856920e-01
-4.09389287e-01 -5.16058445e-01 1.63579777e-01 1.30712652e+00
-1.22708574e-01 -3.82160544e-01 7.40094066e-01 5.02474487e-01
5.36285758e-01 -3.52321804e-01 -8.94568741e-01 -1.12505507e+00
4.40598503e-02 1.97342321e-01 8.91107023e-01 6.63464189e-01
-5.50392270e-01 1.72672033e-01 -6.67238057e-01 2.26756364e-01
1.02068269e+00 7.70027041e-01 8.91110897e-01 -1.08099997e+00
-3.28785300e-01 -2.70682603e-01 -6.36825860e-01 -1.19934583e+00
3.59359205e-01 -9.09272194e-01 1.78983256e-01 -1.08011413e+00
6.10229731e-01 -6.75639689e-01 -3.78693491e-01 2.31435433e-01
-1.66481569e-01 5.05060136e-01 4.35521722e-01 4.42592889e-01
-7.37141907e-01 9.05562520e-01 1.19699490e+00 -5.08928120e-01
3.19628447e-01 -4.59738195e-01 -5.98159611e-01 -3.32525112e-02
3.12109411e-01 -5.78774929e-01 -6.40975237e-01 -4.27358866e-01
-8.33409950e-02 1.69921249e-01 5.80561876e-01 -7.11636126e-01
2.48452976e-01 -4.83678021e-02 3.09130222e-01 -6.04730785e-01
3.69639665e-01 -8.96283507e-01 2.56152034e-01 4.71859455e-01
-3.96749854e-01 -5.76435328e-02 6.59792647e-02 9.65028405e-01
-3.73586535e-01 1.09978661e-01 9.36246455e-01 1.60475671e-02
-7.24779904e-01 7.82842457e-01 4.16351229e-01 -2.08201669e-02
9.87645030e-01 -8.04512724e-02 -2.22343102e-01 -4.81557965e-01
2.14381553e-02 1.57262515e-02 7.60215998e-01 6.15098476e-01
9.17251110e-01 -1.91086328e+00 -6.37835979e-01 5.91416955e-01
5.11717677e-01 -6.35809675e-02 6.87101662e-01 1.86353438e-02
-3.46006781e-01 4.24369812e-01 -5.14660180e-01 -4.96331096e-01
-9.99398291e-01 9.76343632e-01 -8.07297304e-02 -2.08733350e-01
-5.48057139e-01 7.46472239e-01 6.55944347e-01 -4.70995367e-01
3.45824808e-01 1.57407328e-01 7.55244613e-01 -2.61346757e-01
8.40826988e-01 1.91606149e-01 -8.99616927e-02 -2.84754008e-01
-3.68675083e-01 6.61994457e-01 -3.93698245e-01 -1.63085431e-01
1.03591752e+00 -6.13898225e-02 5.33641465e-02 -2.83440053e-02
1.84528601e+00 -2.04889104e-01 -1.42119658e+00 -4.63359565e-01
-2.76763737e-01 -1.13105738e+00 -4.21227096e-03 -6.12773061e-01
-1.16673863e+00 9.16071415e-01 7.19172537e-01 -2.03028858e-01
1.14266860e+00 1.84490886e-02 1.29002655e+00 3.46279770e-01
6.52103901e-01 -7.96637058e-01 5.09659350e-01 -1.16930142e-01
1.03366053e+00 -1.19771552e+00 -2.10092589e-01 -9.45215747e-02
-6.40912235e-01 9.64758337e-01 2.55690485e-01 -4.53180224e-01
4.37588811e-01 6.52216226e-02 -2.57887065e-01 2.60139331e-02
-4.23249692e-01 2.61189699e-01 2.09171847e-01 6.02256477e-01
-2.57744163e-01 6.63866922e-02 4.10093814e-02 6.00409746e-01
2.40346149e-01 9.50651616e-02 -1.31094232e-01 8.00471783e-01
-6.43688813e-02 -1.23441899e+00 -3.77670199e-01 -9.86795872e-03
1.92787480e-02 1.06019555e-02 -3.02962244e-01 6.72868729e-01
-1.10489495e-01 4.60848093e-01 1.12797633e-01 -4.30259049e-01
1.86077252e-01 -2.39069298e-01 6.47582889e-01 -1.79616675e-01
-1.67196814e-03 -1.02538168e-01 -7.17164218e-01 -4.62176740e-01
1.24916717e-01 -3.94461274e-01 -9.30352032e-01 -5.13504922e-01
-1.00874789e-01 9.86421034e-02 7.84489632e-01 3.54963094e-01
6.17319882e-01 -5.65517433e-02 1.06597185e+00 -8.21170390e-01
-1.10084629e+00 -3.36564690e-01 -9.79005456e-01 8.12632143e-01
4.11184728e-01 -6.62170887e-01 -3.34588796e-01 -5.74877560e-02] | [11.626060485839844, 0.6967300176620483] |
6eeb9fbf-96f1-4897-a8d3-b144cd386936 | transifc-invariant-cues-aware-feature | null | null | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10023961 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10023961 | TransIFC: Invariant Cues-aware Feature Concentration Learning for Efficient Fine-grained Bird Image Classification | Fine-grained bird image classification (FBIC) is not
only meaningful for endangered bird observation and protection
but also a prevalent task for image classification in multimedia
processing and computer vision. However, FBIC suffers from
several challenges, such as bird molting, complex background, and
arbitrary bird posture. To effectively tackle these challenges, we
present a novel invariant cues-aware feature concentration
Transformer (TransIFC), which learns invariant and core
information in bird images. To this end, two novel modules are
proposed to leverage the characteristics of bird images, namely,
the hierarchy stage feature aggregation (HSFA) module and the
feature in feature abstraction (FFA) module. The HSFA module
aggregates the multiscale information of bird images by
concatenating multilayer features. The FFA module extracts the
invariant cues of birds through feature selection based on
discrimination scores. Transformer is employed as the backbone
to reveal the long-dependent semantic relationships in bird
images. Moreover, abundant visualizations are provided to prove
the interpretability of the HSFA and FFA modules in TransIFC.
Comprehensive experiments demonstrate that TransIFC can
achieve state-of-the-art performance on the CUB-200-2011 dataset
(91.0%) and the NABirds dataset (90.9%). Finally, extended
experiments have been conducted on the Stanford Cars dataset to
suggest the potential of generalizing our method on other finegrained visual classification tasks. | ['You-Fu Li', 'Zhaoli Zhang', 'Tingting Liu', 'Bochen Xie', 'Yongjian Deng', 'Cheng Zhang', 'Hai Liu'] | 2022-12-31 | null | null | null | tip-2022-12 | ['fine-grained-image-classification'] | ['computer-vision'] | [ 4.25770506e-02 -6.57570720e-01 1.94283992e-01 -3.82677704e-01
2.45724842e-02 -5.85100472e-01 4.52686340e-01 1.33429572e-01
-4.42616433e-01 3.07121962e-01 1.17659517e-01 2.54483908e-01
-2.14114457e-01 -8.19955647e-01 -5.95775902e-01 -8.17490518e-01
-4.25710529e-01 -4.71622765e-01 3.10006410e-01 -4.68311846e-01
1.28633574e-01 5.64201534e-01 -2.07492352e+00 5.47931612e-01
8.60424936e-01 1.27672338e+00 1.90304473e-01 4.34658080e-01
1.42168999e-01 5.70795536e-01 -4.56866562e-01 -3.33648354e-01
3.95612419e-01 -1.41503662e-01 -4.29466099e-01 2.10771412e-01
6.58709824e-01 -4.25195724e-01 -3.30398232e-02 1.13836455e+00
3.01286757e-01 2.55893528e-01 6.33196115e-01 -1.29276156e+00
-6.65687680e-01 2.80785233e-01 -7.46307790e-01 5.07287323e-01
1.45642206e-01 3.48977566e-01 1.02980530e+00 -1.10506999e+00
3.00649017e-01 1.14196229e+00 6.65656805e-01 2.60010839e-01
-1.05315828e+00 -8.10864627e-01 5.10526180e-01 4.47337180e-01
-1.41096818e+00 -2.02386007e-01 6.21104538e-01 -7.79310226e-01
7.33649313e-01 5.51004350e-01 1.03278780e+00 4.05174464e-01
7.44472891e-02 6.21310830e-01 1.13408089e+00 -1.24010146e-01
-1.32177472e-01 -5.14898263e-02 3.13250780e-01 1.23879695e+00
1.90799266e-01 5.00569820e-01 -8.47220540e-01 1.51518762e-01
5.97657800e-01 4.15830463e-01 -2.84495920e-01 -1.94323719e-01
-1.22886407e+00 7.58547127e-01 1.10785222e+00 1.33805387e-02
-2.19882324e-01 -1.38916001e-01 3.69549721e-01 2.22490445e-01
4.38497305e-01 4.16194439e-01 -2.18776748e-01 5.25252044e-01
-9.06256616e-01 2.15005949e-01 4.10217404e-01 8.58844936e-01
9.64397609e-01 2.04895481e-01 -4.28970605e-01 1.00908363e+00
3.67157042e-01 4.12997752e-01 2.88345277e-01 -5.73081732e-01
3.65709290e-02 8.16974878e-01 -3.10554683e-01 -1.39691818e+00
-5.54343760e-01 -8.59314501e-01 -8.60194862e-01 1.06371805e-01
1.00597285e-01 2.49167398e-01 -8.23058426e-01 1.67065585e+00
6.10032260e-01 3.00453901e-02 -4.36604619e-01 1.40136051e+00
1.44524837e+00 5.78776836e-01 2.45190173e-01 7.95757622e-02
1.63389397e+00 -1.11764145e+00 -1.45586103e-01 -9.65333357e-02
3.12796608e-02 -4.28236544e-01 1.07513487e+00 1.78650320e-01
-6.67435110e-01 -8.01912546e-01 -1.24246287e+00 -8.09715316e-02
-7.00052261e-01 3.93564463e-01 7.24066496e-01 4.11452204e-01
-9.40853238e-01 1.67171225e-01 -4.25570965e-01 -3.27859730e-01
4.79871333e-01 2.31997967e-01 -5.04977047e-01 6.21765666e-02
-1.00049543e+00 3.97126645e-01 2.28310525e-01 4.42605048e-01
-1.22630012e+00 -9.52371538e-01 -1.00283539e+00 1.58152491e-01
1.04962595e-01 -6.79942489e-01 6.50051594e-01 -9.22671497e-01
-1.07194328e+00 9.15541351e-01 9.37882587e-02 -4.09013361e-01
1.62405401e-01 -4.93704081e-02 -3.96450490e-01 4.01291877e-01
2.70386338e-01 9.00711954e-01 1.17545235e+00 -1.05875611e+00
-1.03046298e+00 -4.92542207e-01 2.42884085e-01 2.30488077e-01
-4.28203434e-01 2.45510694e-02 -2.31886834e-01 -1.02527523e+00
-1.00620918e-01 -6.16107106e-01 -1.14019662e-01 3.46436083e-01
-4.56732977e-03 2.45104060e-02 6.97310686e-01 -5.24647534e-01
1.31789351e+00 -2.33016968e+00 3.85881752e-01 4.74144109e-02
4.82738674e-01 1.53357103e-01 -1.05450995e-01 2.80591875e-01
1.80613339e-01 4.19175960e-02 -2.92788297e-01 2.36150339e-01
-4.40533608e-02 -1.19235963e-01 -3.31915677e-01 5.62989473e-01
4.41667944e-01 8.39916945e-01 -7.96228349e-01 -6.02422595e-01
3.89403582e-01 3.06547612e-01 -7.93972373e-01 1.99088395e-01
1.12510398e-01 3.14792991e-01 -3.68439704e-01 1.06017780e+00
7.19411910e-01 8.71952623e-02 -2.76378602e-01 -6.24424934e-01
-4.77283239e-01 -5.20910859e-01 -7.30242610e-01 1.43336701e+00
-1.40126795e-01 3.98887932e-01 2.47919858e-01 -7.23058760e-01
7.63513923e-01 -3.43239367e-01 8.28495920e-02 -6.42157435e-01
3.77214253e-02 -1.09285697e-01 2.21256609e-03 -5.07584214e-01
5.65258622e-01 -3.57993990e-01 -4.19665724e-01 -1.09217875e-02
4.03275192e-01 -4.49329428e-02 3.63738507e-01 1.91670537e-01
5.71031272e-01 -1.05697401e-01 3.34924459e-01 -6.88103139e-01
7.04114079e-01 1.38152279e-02 6.08770490e-01 5.63832223e-01
-5.27560294e-01 2.16235429e-01 -6.25540793e-04 -7.79191971e-01
-4.65431303e-01 -1.05225670e+00 -2.77913153e-01 1.54968309e+00
5.13437092e-01 -6.00000918e-01 -7.24643409e-01 -7.72817612e-01
2.05098480e-01 1.41338855e-01 -9.08263922e-01 -3.09250861e-01
-1.87736869e-01 -8.86640728e-01 5.33726394e-01 4.50699359e-01
9.20036256e-01 -9.56383944e-01 -9.62029159e-01 -9.07597244e-02
-1.36303350e-01 -9.00469124e-01 -6.47075236e-01 3.24941948e-02
-4.84153777e-01 -1.08250558e+00 -3.17869395e-01 -6.56115592e-01
6.97652519e-01 9.39961255e-01 8.97531986e-01 5.01912594e-01
-8.63434911e-01 4.29504335e-01 -5.34217119e-01 -1.91407636e-01
3.17890048e-01 -1.02981031e-01 9.98271629e-02 2.76517868e-01
1.43925309e-01 -3.55257839e-01 -7.56651938e-01 6.54365182e-01
-9.73952115e-01 1.83167264e-01 6.93154454e-01 1.25038528e+00
7.15643942e-01 2.29686260e-01 3.74273181e-01 -4.89631027e-01
2.04847664e-01 -4.00941104e-01 -6.30583584e-01 2.63058305e-01
-1.90708742e-01 -1.75331593e-01 8.01662266e-01 -1.91180736e-01
-9.06366229e-01 4.55198623e-02 -6.08636662e-02 -2.61621177e-01
-3.82153600e-01 4.89218444e-01 -2.12744549e-01 -4.25806731e-01
4.14081842e-01 2.78790176e-01 -2.92166531e-01 -4.18109447e-01
2.68025756e-01 3.79290104e-01 6.18254125e-01 -5.00988305e-01
8.55718076e-01 5.89744866e-01 3.69119309e-02 -1.15411639e+00
-1.03726304e+00 -4.57692474e-01 -7.56999969e-01 -5.61059237e-01
9.73281145e-01 -1.12589216e+00 -7.42669702e-01 5.50526023e-01
-4.98778641e-01 -1.62228234e-02 -1.12027779e-01 1.96234226e-01
-1.57926425e-01 2.09409192e-01 -4.70298588e-01 -5.45828104e-01
-3.16636503e-01 -1.14519763e+00 1.22614133e+00 6.19772553e-01
2.73783535e-01 -6.28883123e-01 -3.05656642e-01 5.10153353e-01
2.83008188e-01 4.01362091e-01 9.50661659e-01 -1.83215648e-01
-6.49591386e-01 1.13326155e-01 -5.02217710e-01 1.71312615e-01
-4.64953035e-02 1.97172791e-01 -9.99863029e-01 -5.49835324e-01
-3.39074016e-01 -2.71863073e-01 1.35424995e+00 1.73735425e-01
1.24836051e+00 -2.91535825e-01 -7.81860501e-02 1.10658741e+00
1.35774195e+00 -2.46796664e-03 1.43339587e-02 2.35898763e-01
7.32579410e-01 8.19627643e-01 8.14358830e-01 4.91575509e-01
5.39530456e-01 7.34018862e-01 6.03254437e-01 -2.81317234e-01
-2.61988550e-01 -5.51572084e-01 4.73011971e-01 5.86178839e-01
-1.15446366e-01 2.76514679e-01 -6.38920367e-01 3.91239822e-01
-1.64844859e+00 -1.10338008e+00 -9.89034325e-02 1.71609187e+00
4.73089248e-01 -2.63797045e-01 3.79024088e-01 5.89992143e-02
6.74127042e-01 2.61359990e-01 -5.35682619e-01 -1.96416155e-01
-2.26681367e-01 -5.31556681e-02 2.73273140e-01 1.67258516e-01
-1.54606068e+00 1.12040126e+00 5.36598158e+00 9.91441607e-01
-1.22835660e+00 5.51749393e-02 4.30726349e-01 -1.65293843e-01
9.12853628e-02 -2.57343829e-01 -8.43777180e-01 4.71673131e-01
2.86788285e-01 4.14698943e-02 5.58307886e-01 6.72408521e-01
-8.36962909e-02 -1.02772087e-01 -7.02352583e-01 9.45489287e-01
2.38680884e-01 -1.18902993e+00 3.66379648e-01 -1.86715335e-01
4.06587362e-01 -2.97685415e-01 3.27592582e-01 3.51414919e-01
-5.61875813e-02 -1.03854632e+00 1.13866949e+00 5.27738094e-01
1.01951778e+00 -8.16259027e-01 4.87687826e-01 3.24001312e-01
-1.85309792e+00 -3.79753292e-01 -6.19955063e-01 -2.58676738e-01
-3.60950351e-01 3.33232582e-01 -4.90707666e-01 7.11221933e-01
1.32074487e+00 1.14117622e+00 -1.14728093e+00 1.21201181e+00
-1.46944532e-02 4.74873513e-01 -1.78060994e-01 1.23100514e-02
3.74065310e-01 -1.30459309e-01 6.02619946e-01 1.48166704e+00
1.29741244e-02 2.61361659e-01 3.35441738e-01 7.68823385e-01
6.24347851e-02 1.72681570e-01 -5.10040522e-01 1.29540816e-01
1.64276883e-01 1.80191028e+00 -8.76329839e-01 -1.86620131e-02
-2.68998325e-01 7.47420788e-01 4.13519800e-01 2.12300614e-01
-8.88665795e-01 -3.97791892e-01 8.61149430e-01 -2.81174239e-02
5.71874559e-01 -1.66692719e-01 4.44171624e-03 -1.47643793e+00
-1.17356651e-01 -8.85861218e-01 7.17701972e-01 -3.67419779e-01
-1.39904702e+00 7.49785662e-01 1.30303986e-02 -1.53138411e+00
4.12981331e-01 -6.91150069e-01 -4.72653866e-01 5.81761897e-01
-1.70355570e+00 -1.61749732e+00 -7.95004129e-01 7.65565753e-01
6.83769703e-01 -2.61239260e-01 5.66772223e-01 3.17946792e-01
-8.36962938e-01 6.47270560e-01 -2.61276841e-01 1.29661798e-01
4.45128411e-01 -1.08988714e+00 -2.92722993e-02 1.11330044e+00
2.95837998e-01 5.33459425e-01 3.91144395e-01 -3.58263671e-01
-1.54035962e+00 -1.40972853e+00 2.57178936e-02 -1.32486388e-01
5.72118998e-01 -4.66792375e-01 -6.51845396e-01 1.08037718e-01
-1.90861598e-01 3.27163547e-01 8.89288664e-01 -2.00025424e-01
-6.78412616e-01 -4.83368725e-01 -1.04848790e+00 4.14531767e-01
1.10449612e+00 -5.45961499e-01 -4.54398274e-01 -8.75539407e-02
6.30970478e-01 -1.94841757e-01 -8.05742145e-01 5.36510527e-01
8.66883874e-01 -1.29099548e+00 1.19551980e+00 -6.41303360e-01
3.64893526e-01 -9.37090695e-01 -5.11759639e-01 -1.30175412e+00
-7.25256741e-01 -4.17674743e-02 1.79827645e-01 1.26816022e+00
6.75773323e-02 -4.11941528e-01 1.13634609e-01 3.64023186e-02
-2.58630246e-01 -6.95396423e-01 -8.91678989e-01 -5.11663735e-01
-2.09666163e-01 -2.18764856e-01 7.86320925e-01 6.66350126e-01
-2.92277873e-01 2.02424258e-01 -4.57121491e-01 4.01251376e-01
7.93629587e-01 7.78748214e-01 6.47444785e-01 -1.29800880e+00
-6.81730807e-02 -5.56288421e-01 -6.68970823e-01 -9.00856316e-01
1.12549245e-01 -1.04650795e+00 1.62810922e-01 -1.13182247e+00
3.19482803e-01 -2.37059027e-01 -4.78830785e-01 4.67466533e-01
-3.10316950e-01 7.40868866e-01 4.92841065e-01 1.02215275e-01
-8.44630837e-01 7.51836717e-01 1.17242312e+00 -3.63359541e-01
1.77770987e-01 -1.39544293e-01 -8.98989081e-01 7.84024179e-01
4.47508186e-01 -1.13449894e-01 -1.43344969e-01 -3.75035524e-01
-1.89785048e-01 -2.42896035e-01 9.48095143e-01 -9.01672900e-01
2.62617499e-01 -1.57367259e-01 8.34644139e-01 -6.36055410e-01
1.73462898e-01 -8.54600370e-01 8.27537775e-02 5.20367146e-01
-1.05205186e-01 -6.91348463e-02 3.43708843e-01 6.72507823e-01
-5.24628103e-01 1.38449728e-01 9.68918562e-01 -7.52832144e-02
-1.22764957e+00 5.01372039e-01 -2.38633946e-01 2.82733254e-02
1.09162676e+00 -1.17598280e-01 -4.69470471e-01 1.65305316e-01
-4.00543600e-01 4.88015175e-01 4.78873789e-01 4.96609777e-01
9.45802391e-01 -1.22090280e+00 -7.68770874e-01 8.42038691e-01
4.84570265e-01 -1.71477914e-01 6.84121788e-01 8.82577240e-01
-5.61966002e-01 2.95716524e-01 -6.05341494e-01 -8.03322434e-01
-1.39577913e+00 5.74605882e-01 4.01970804e-01 4.94942106e-02
-4.78913069e-01 1.07923663e+00 9.07149255e-01 -2.35709518e-01
4.43721227e-02 -3.82064700e-01 -4.29810673e-01 4.25920427e-01
7.38551855e-01 3.37744474e-01 -3.46790962e-02 -1.03609562e+00
-5.06938875e-01 8.60335827e-01 -3.83499190e-02 3.88297498e-01
1.39312363e+00 -2.90305376e-01 -2.76618868e-01 2.10066751e-01
8.10806632e-01 -7.90509284e-02 -1.40705240e+00 -9.33780298e-02
-1.45537972e-01 -7.99907625e-01 7.12276548e-02 -8.95412683e-01
-1.22105193e+00 1.17245257e+00 6.26349747e-01 2.44645685e-01
1.51109362e+00 -1.61726937e-01 5.70259869e-01 1.48255289e-01
5.72754025e-01 -6.59388602e-01 -9.42650214e-02 2.66880333e-01
1.17103028e+00 -1.13175642e+00 1.10533029e-01 -4.04216141e-01
-7.62580574e-01 1.06762958e+00 8.21582198e-01 -3.77513245e-02
6.94561958e-01 1.22554585e-01 -1.69349350e-02 -2.90786147e-01
-7.93907821e-01 -5.55548966e-01 7.95238554e-01 4.73412305e-01
2.29567125e-01 2.92034507e-01 -5.66679128e-02 1.11850691e+00
-2.70790547e-01 -4.02592570e-01 9.41371843e-02 5.75710833e-01
-5.64601839e-01 -4.80314910e-01 -2.68778354e-01 5.81445932e-01
-2.59048253e-01 -3.04016352e-01 -5.33253491e-01 6.01676762e-01
6.14071846e-01 1.03599620e+00 -1.96936145e-01 -8.14453900e-01
4.13479716e-01 -3.45744789e-01 3.32444936e-01 -6.19255066e-01
-1.15257406e+00 7.35028759e-02 -1.74972281e-01 -6.36950433e-01
-5.55184245e-01 -4.33808237e-01 -8.86232793e-01 -2.04312041e-01
-2.53787935e-01 1.47268638e-01 2.70096570e-01 6.11372113e-01
2.64898181e-01 4.98846024e-01 6.66884720e-01 -1.03883398e+00
-1.20890088e-01 -5.48609972e-01 -6.97397709e-01 2.28191555e-01
6.10771537e-01 -1.15787649e+00 -4.18468207e-01 4.09973972e-02] | [9.625448226928711, 2.027846574783325] |
46299c88-ee63-4aef-b4a2-849fd756c717 | compositional-generalization-by-factorizing | null | null | https://aclanthology.org/2020.acl-srw.42 | https://aclanthology.org/2020.acl-srw.42.pdf | Compositional Generalization by Factorizing Alignment and Translation | Standard methods in deep learning for natural language processing fail to capture the compositional structure of human language that allows for systematic generalization outside of the training distribution. However, human learners readily generalize in this way, e.g. by applying known grammatical rules to novel words. Inspired by work in cognitive science suggesting a functional distinction between systems for syntactic and semantic processing, we implement a modification to an existing approach in neural machine translation, imposing an analogous separation between alignment and translation. The resulting architecture substantially outperforms standard recurrent networks on the SCAN dataset, a compositional generalization task, without any additional supervision. Our work suggests that learning to align and to translate in separate modules may be a useful heuristic for capturing compositional structure. | ['all', "R O{'}Reilly", 'Jason Jo', 'Jacob Russin', 'Yoshua Bengio'] | 2020-07-01 | null | null | null | acl-2020-6 | ['systematic-generalization'] | ['reasoning'] | [ 6.18594229e-01 5.08366823e-01 -1.95868060e-01 -8.04086983e-01
-3.50032836e-01 -8.87746334e-01 9.54917908e-01 2.41410151e-01
-4.29262280e-01 5.12037933e-01 5.54069221e-01 -7.87003994e-01
1.88531622e-01 -9.11312044e-01 -1.01382494e+00 -3.48081738e-01
2.02015266e-01 7.76635587e-01 -2.20160764e-02 -5.72907269e-01
2.51739413e-01 2.88944095e-01 -1.09488773e+00 7.46249318e-01
6.58572495e-01 2.78463006e-01 2.83889353e-01 4.05851692e-01
-5.19341409e-01 5.07287681e-01 -2.95561075e-01 -5.00721335e-01
2.70105004e-01 -1.10414612e+00 -1.29782844e+00 -1.08336501e-01
6.25332952e-01 -1.49162218e-01 -7.61154890e-02 9.59721148e-01
1.58351466e-01 2.41708696e-01 5.21831810e-01 -5.38870156e-01
-1.44034660e+00 1.23328888e+00 8.79692361e-02 3.68935645e-01
4.50999409e-01 -3.54416370e-02 1.19897461e+00 -8.42806220e-01
7.11891294e-01 1.42176735e+00 8.44695985e-01 1.03338301e+00
-1.52269518e+00 -3.55891138e-01 3.21769387e-01 -2.18944252e-01
-7.51508892e-01 -4.60120708e-01 5.03423154e-01 -3.02999347e-01
1.50674069e+00 8.85038003e-02 5.98809063e-01 1.28370011e+00
3.59667480e-01 7.77206779e-01 1.14742732e+00 -6.65749907e-01
3.86700034e-02 -1.10317655e-01 1.43905923e-01 7.22435474e-01
2.23139331e-01 2.84497738e-01 -5.83754182e-01 -1.13173723e-02
6.08946204e-01 3.56067382e-02 1.29094899e-01 -2.33035773e-01
-1.34154487e+00 8.19171786e-01 6.05763257e-01 4.79660362e-01
-9.08435807e-02 3.42192352e-01 6.67536080e-01 8.69514346e-01
3.29111099e-01 9.94133234e-01 -6.23874009e-01 2.52104044e-01
-8.58519852e-01 2.12803856e-01 7.31186211e-01 9.01046515e-01
8.58563900e-01 1.82161853e-01 3.62212732e-02 6.03733063e-01
1.27650753e-01 2.39990920e-01 1.02463341e+00 -1.01163840e+00
2.84205317e-01 5.15756130e-01 -5.34802496e-01 -4.20931011e-01
-3.57423306e-01 -4.81659472e-01 -4.82218444e-01 -1.67057738e-01
3.63624454e-01 -1.50420040e-01 -8.69208217e-01 2.14823413e+00
-1.75508052e-01 -2.97399789e-01 2.38229036e-01 6.77110672e-01
4.71541911e-01 5.76673210e-01 2.41545051e-01 -5.27559556e-02
1.04876888e+00 -8.48684967e-01 -1.93719044e-01 -5.81123590e-01
8.72931719e-01 -6.55643344e-01 1.48794007e+00 1.26107872e-01
-1.39646888e+00 -3.63522381e-01 -8.77657235e-01 -3.94032121e-01
-5.40620506e-01 -4.02476549e-01 9.87906516e-01 5.21947145e-01
-1.47272038e+00 8.10194075e-01 -7.13255525e-01 -8.43475044e-01
3.02589089e-01 3.44821423e-01 -2.48571992e-01 1.52208000e-01
-1.26301491e+00 1.23798311e+00 5.08952141e-01 -1.15530873e-02
-6.93765342e-01 -5.40299177e-01 -9.24732089e-01 1.75654758e-02
-6.98373094e-02 -1.26660120e+00 1.79519951e+00 -1.83414137e+00
-1.54210091e+00 1.26084828e+00 -3.63075227e-01 -7.69290149e-01
1.19619491e-02 -2.06522450e-01 3.00632715e-02 -1.25893623e-01
4.07309458e-02 8.73891354e-01 7.88687229e-01 -8.89505744e-01
-2.65155733e-01 -4.16705281e-01 2.10070148e-01 4.00744587e-01
-6.72231689e-02 2.74044037e-01 5.39856970e-01 -8.57062161e-01
2.22586930e-01 -9.27737951e-01 -1.85773179e-01 -2.66192079e-01
-5.71505278e-02 -5.31669319e-01 2.28333130e-01 -5.33961594e-01
8.23126018e-01 -1.61609972e+00 3.80079508e-01 7.46365264e-02
2.18732022e-02 1.19205162e-01 -2.96185791e-01 7.14679658e-01
-2.78890818e-01 2.96195656e-01 -4.80508924e-01 -1.92859128e-01
2.53956109e-01 5.36773801e-01 -7.64651835e-01 1.74371943e-01
3.69082421e-01 1.40128636e+00 -1.06355548e+00 1.11863136e-01
-3.27193826e-01 -2.32952330e-02 -6.83881581e-01 -1.43791810e-01
-6.08096123e-01 1.78452626e-01 -9.61724445e-02 3.06784540e-01
-7.50361383e-02 -1.87615193e-02 4.52648163e-01 5.28808057e-01
-4.92463112e-02 1.26150858e+00 -3.55733007e-01 2.11705589e+00
-6.74511731e-01 6.81209266e-01 -2.82555491e-01 -1.48706400e+00
8.49596262e-01 4.99637008e-01 -1.94905221e-01 -7.06600726e-01
7.35933185e-02 4.20453370e-01 5.46785712e-01 -2.18921214e-01
2.77642727e-01 -7.33780622e-01 -2.34038711e-01 1.18118119e+00
3.50314170e-01 -2.43325919e-01 1.22040883e-01 6.48798719e-02
1.01425397e+00 3.44730228e-01 3.48927230e-01 -6.06551707e-01
3.08733135e-01 4.28870134e-02 3.42541277e-01 7.75034189e-01
1.41738445e-01 3.47180039e-01 2.04526141e-01 -8.10427427e-01
-1.31236863e+00 -1.26044989e+00 1.11887254e-01 1.71082199e+00
-4.00659174e-01 -3.08768690e-01 -8.46472085e-01 -6.34242058e-01
-2.56888539e-01 9.60697055e-01 -5.58787405e-01 -5.41134179e-01
-1.09189892e+00 -6.04005218e-01 7.64695287e-01 7.93062210e-01
2.67784834e-01 -1.52942169e+00 -6.90950394e-01 2.39984781e-01
1.85687751e-01 -5.92387378e-01 -4.41819519e-01 5.07815361e-01
-1.30526686e+00 -5.32898724e-01 -2.03653246e-01 -1.27871227e+00
6.48596168e-01 1.33752078e-01 1.45227134e+00 3.03893268e-01
1.02722578e-01 3.06949735e-01 -1.27696425e-01 -2.69430965e-01
-9.05189097e-01 4.54778880e-01 6.60296753e-02 -6.30557716e-01
6.21420801e-01 -8.39869261e-01 -2.40520716e-01 -5.75072616e-02
-9.32246864e-01 5.42248785e-02 6.30182624e-01 1.06926429e+00
1.73879966e-01 -6.19114339e-01 8.83558452e-01 -1.14711976e+00
1.02665436e+00 -5.65375865e-01 -1.31821990e-01 2.22574294e-01
-6.74674869e-01 5.68270862e-01 1.00064957e+00 -4.33870494e-01
-9.49943006e-01 6.45550191e-02 -1.88584656e-01 1.45258188e-01
-4.03302073e-01 6.19442940e-01 5.61425500e-02 1.32041633e-01
8.49803567e-01 5.61826110e-01 7.80524015e-02 -2.49678537e-01
6.81495011e-01 1.53292224e-01 4.74126011e-01 -1.00087500e+00
7.86165953e-01 1.79008886e-01 -4.02837358e-02 -6.00948215e-01
-8.96235883e-01 6.23019226e-02 -8.98902059e-01 5.60546935e-01
7.48159945e-01 -6.03507042e-01 -3.41708452e-01 -1.25189587e-01
-1.35202575e+00 -5.44447541e-01 -6.03321731e-01 3.18853378e-01
-9.30515707e-01 1.97637677e-01 -8.43723774e-01 -1.58294246e-01
-4.65242505e-01 -7.60865152e-01 1.01117706e+00 -6.66453913e-02
-8.72598946e-01 -1.35270166e+00 2.35410452e-01 -4.76580262e-02
7.73424447e-01 -1.88506201e-01 1.38147283e+00 -9.82839763e-01
-3.60155255e-01 1.69819534e-01 2.16714680e-01 3.63090992e-01
1.45605385e-01 -3.25104505e-01 -6.07235491e-01 -1.21596634e-01
2.49795184e-01 -4.80008483e-01 1.08560073e+00 7.40138888e-02
6.99869096e-01 -5.33962548e-01 -1.51251867e-01 6.40189886e-01
9.34647143e-01 4.32396010e-02 3.20413232e-01 3.88931602e-01
4.06731099e-01 7.37398624e-01 -2.81502038e-01 -4.00534958e-01
3.52406800e-01 2.24335596e-01 -1.09861739e-01 6.59355968e-02
-2.01212913e-01 -5.72465003e-01 7.31376708e-01 9.47599471e-01
4.76365797e-02 1.55399233e-01 -1.04413199e+00 5.55551708e-01
-1.60307729e+00 -1.14481294e+00 3.12554330e-01 1.94716406e+00
1.28705978e+00 1.74291417e-01 4.25104015e-02 -2.11610302e-01
4.92995828e-01 5.93021996e-02 -4.23218310e-01 -1.18488026e+00
-1.59981072e-01 8.83075416e-01 3.54883075e-01 6.51025772e-01
-6.58517718e-01 1.43987894e+00 7.74404669e+00 1.54280558e-01
-1.14861822e+00 2.44768366e-01 2.94079095e-01 -1.21453116e-02
-7.48497248e-01 2.53336996e-01 -5.78714430e-01 8.56232643e-02
1.26450634e+00 -2.44639918e-01 8.91651928e-01 3.11696082e-01
-1.23648904e-02 4.03433144e-01 -1.71890485e+00 3.54494601e-01
3.09522092e-01 -1.45326889e+00 5.68987846e-01 -2.93984503e-01
7.48777092e-01 1.90078124e-01 1.66406110e-01 3.48868996e-01
7.00702727e-01 -1.32687962e+00 7.02465177e-01 2.68596202e-01
4.29867327e-01 -4.13461447e-01 2.91342437e-01 5.17920673e-01
-6.99032366e-01 -8.51187035e-02 -6.00669086e-01 -5.99125087e-01
6.95278347e-02 -9.43887383e-02 -7.91518152e-01 4.05283049e-02
9.62084308e-02 6.42940164e-01 -7.33635485e-01 3.96122813e-01
-7.03779042e-01 6.74146056e-01 -1.47998646e-01 -1.80588886e-01
5.31165838e-01 -1.92915067e-01 4.70808893e-01 1.37585068e+00
2.33184099e-01 -1.26158878e-01 3.34863216e-02 1.03675747e+00
-3.85048062e-01 9.66265127e-02 -9.88652587e-01 -1.42573193e-01
3.32321703e-01 6.29511476e-01 -5.96044064e-01 -5.07871687e-01
-4.95500118e-01 1.08705366e+00 5.41523278e-01 4.58420902e-01
-2.92262584e-01 7.95539916e-02 4.71337497e-01 1.65925384e-01
1.90401465e-01 -5.08255303e-01 -6.25509322e-01 -1.28760207e+00
6.58752248e-02 -1.12385416e+00 2.84482896e-01 -6.43788934e-01
-1.24744475e+00 5.47273695e-01 -9.03477799e-03 -5.94594419e-01
-7.37933815e-01 -7.29128897e-01 -8.63182366e-01 1.04049563e+00
-9.80072260e-01 -1.29641414e+00 5.06527841e-01 4.71481264e-01
1.00526142e+00 -3.17678660e-01 1.16446209e+00 -2.30677873e-01
-1.15233667e-01 6.98735595e-01 -1.04645826e-01 5.38195670e-02
5.63960075e-01 -1.47556102e+00 1.13460350e+00 9.65319276e-01
6.78312123e-01 1.41560018e+00 7.52909780e-01 -4.53754932e-01
-1.36860466e+00 -9.74645019e-01 1.39392924e+00 -6.41079187e-01
7.86638737e-01 -6.76357150e-01 -1.01397765e+00 1.23671865e+00
8.53957891e-01 -4.72394675e-01 7.76066065e-01 3.30776513e-01
-8.52194130e-01 1.81741983e-01 -6.89348280e-01 7.14713097e-01
1.53521979e+00 -1.00462484e+00 -1.53388715e+00 4.21387285e-01
1.02103782e+00 -7.35880435e-02 -3.55996490e-01 1.55557781e-01
5.55607915e-01 -6.63204193e-01 7.07185268e-01 -1.30245721e+00
7.33791947e-01 -1.04514159e-01 -2.00769886e-01 -1.75607705e+00
-4.60652202e-01 -6.89765453e-01 1.37378797e-01 8.25881124e-01
7.73215830e-01 -8.23981941e-01 5.76161981e-01 4.37872797e-01
-6.17620051e-01 -6.40697062e-01 -7.02897847e-01 -8.30778062e-01
9.68214869e-01 -2.98496395e-01 6.55549645e-01 1.16945124e+00
3.89091432e-01 8.72930408e-01 1.78991541e-01 -3.47610801e-01
2.33232468e-01 3.26732159e-01 4.19940978e-01 -1.13890898e+00
-5.65744400e-01 -1.01481998e+00 -2.04481840e-01 -1.22794116e+00
8.19638014e-01 -1.73932695e+00 -5.59863858e-02 -1.56312728e+00
7.31660575e-02 2.90900804e-02 -9.58586782e-02 5.38968563e-01
1.21538028e-01 1.44825801e-01 3.48057188e-02 2.53517687e-01
-2.62013465e-01 3.45794022e-01 9.10828531e-01 -1.48939535e-01
4.67061177e-02 -1.63940758e-01 -1.16166544e+00 7.13930607e-01
1.08253372e+00 -4.32549447e-01 -5.18137276e-01 -9.98569071e-01
7.43859291e-01 -4.23938155e-01 1.87505990e-01 -3.91343623e-01
1.51470751e-01 -1.50751039e-01 3.25238258e-01 -5.45949861e-02
-2.45757833e-01 -5.54653823e-01 -1.51695937e-01 6.21691585e-01
-9.09741580e-01 6.36111915e-01 1.58453479e-01 2.16170236e-01
-1.75612450e-01 -3.74989033e-01 5.33407927e-01 -6.98378861e-01
-5.68484485e-01 -1.56081215e-01 -7.70844817e-01 5.81100471e-02
4.64761585e-01 -2.64650494e-01 -5.60752213e-01 -1.40807092e-01
-9.51307595e-01 -7.72335008e-03 6.93530083e-01 5.23627460e-01
4.70742196e-01 -1.30794311e+00 -6.91913307e-01 2.39574328e-01
-1.15266934e-01 -1.46036699e-01 -4.72467095e-01 6.50022149e-01
-5.54671943e-01 4.24863964e-01 -3.22429508e-01 -3.30114037e-01
-6.08455360e-01 6.71157479e-01 6.04733706e-01 5.39908819e-02
-6.05669975e-01 7.91683137e-01 3.81714731e-01 -8.69805038e-01
-1.62288561e-01 -6.47013485e-01 2.57451862e-01 -3.02134812e-01
3.28263432e-01 -2.41319761e-01 1.39268354e-01 -5.34817636e-01
-1.77511275e-01 3.59191746e-01 -2.02233359e-01 -1.47681668e-01
1.26287830e+00 -2.11764455e-01 -6.41554534e-01 8.72844040e-01
1.07720113e+00 -1.06712647e-01 -6.23587251e-01 -4.11541790e-01
6.14266217e-01 1.06282182e-01 -6.21284187e-01 -7.45855510e-01
-3.98449272e-01 1.13633490e+00 1.02046952e-02 2.47563720e-01
8.49034429e-01 2.67265648e-01 9.01836872e-01 9.49043930e-01
1.59190610e-01 -9.51204121e-01 -1.35214418e-01 1.13788223e+00
9.00131702e-01 -9.52182114e-01 -3.00313741e-01 1.66280672e-01
-3.68941158e-01 1.45846748e+00 4.75816935e-01 -5.63449979e-01
2.97024637e-01 8.43119696e-02 6.95928410e-02 -2.72427917e-01
-1.17147219e+00 9.06607285e-02 2.08317995e-01 4.78500634e-01
1.18340242e+00 1.42387658e-01 -5.50433695e-01 3.18342596e-01
-6.86496735e-01 -4.40196812e-01 5.61687708e-01 8.64039958e-01
-6.44942343e-01 -1.31421161e+00 -8.92697498e-02 3.47058028e-01
-3.56350243e-01 -5.70552289e-01 -9.65501070e-01 4.82276142e-01
3.88582498e-02 4.49104905e-01 1.90925658e-01 -8.44257921e-02
1.71569690e-01 8.59989166e-01 8.23801816e-01 -1.28743720e+00
-9.96147573e-01 -3.22084814e-01 8.76556411e-02 -5.01215696e-01
-4.92166132e-01 -9.08569336e-01 -1.38892472e+00 -1.91005737e-01
2.00937346e-01 1.98660716e-02 4.10959691e-01 1.19468856e+00
1.59491792e-01 2.95992762e-01 2.04401508e-01 -6.57609403e-01
-8.37411523e-01 -9.96083081e-01 1.69556905e-02 5.10925829e-01
3.55927944e-01 -1.86615095e-01 -7.55035132e-02 3.23145241e-01] | [10.673141479492188, 9.047115325927734] |
aac1fc0b-3b40-42a5-b914-3388ee60e516 | memory-time-span-in-lstms-for-multi-speaker | 1808.08097 | null | http://arxiv.org/abs/1808.08097v1 | http://arxiv.org/pdf/1808.08097v1.pdf | Memory Time Span in LSTMs for Multi-Speaker Source Separation | With deep learning approaches becoming state-of-the-art in many speech (as
well as non-speech) related machine learning tasks, efforts are being taken to
delve into the neural networks which are often considered as a black box. In
this paper it is analyzed how recurrent neural network (RNNs) cope with
temporal dependencies by determining the relevant memory time span in a long
short-term memory (LSTM) cell. This is done by leaking the state variable with
a controlled lifetime and evaluating the task performance. This technique can
be used for any task to estimate the time span the LSTM exploits in that
specific scenario. The focus in this paper is on the task of separating
speakers from overlapping speech. We discern two effects: A long term effect,
probably due to speaker characterization and a short term effect, probably
exploiting phone-size formant tracks. | ['Hugo Van hamme', 'Jeroen Zegers'] | 2018-08-24 | null | null | null | null | ['multi-speaker-source-separation'] | ['speech'] | [ 2.07149506e-01 1.31851122e-01 -1.01891518e-01 -3.12944144e-01
-5.14895856e-01 -3.99067372e-01 8.49342346e-01 2.67374758e-02
-6.04717791e-01 6.21308744e-01 2.56416887e-01 -5.89180887e-01
-1.23883724e-01 -2.92150557e-01 -4.60118026e-01 -1.01175547e+00
-2.04382285e-01 4.83803540e-01 2.17319742e-01 -3.66908237e-02
2.46734962e-01 7.38108695e-01 -1.60941815e+00 2.76968062e-01
4.25537378e-01 7.56448746e-01 2.57115960e-01 9.05080080e-01
-2.93196470e-01 9.79593992e-01 -1.00538945e+00 3.20567330e-03
-1.97782815e-01 -5.39292693e-01 -9.22305882e-01 -5.13133146e-02
-8.74526054e-02 2.87910193e-01 -2.40587488e-01 8.68962884e-01
6.16363108e-01 4.61577147e-01 4.83401388e-01 -6.88119471e-01
3.52334797e-01 7.43258238e-01 -5.93542457e-02 8.04285944e-01
-9.77453515e-02 -1.62181377e-01 5.17039180e-01 -6.63754284e-01
4.07656074e-01 1.23612010e+00 4.79525477e-01 5.18804729e-01
-9.77561712e-01 -4.59283918e-01 2.37381220e-01 2.63136834e-01
-1.00874710e+00 -8.87005389e-01 7.95578957e-01 -6.84753180e-01
1.39850545e+00 1.67179838e-01 2.70337462e-01 1.26975060e+00
3.47518802e-01 6.75419450e-01 1.15190768e+00 -7.78542101e-01
2.53939450e-01 2.00197607e-01 4.43240941e-01 2.75282085e-01
-5.12500525e-01 3.85229886e-01 -5.77936769e-01 1.07887417e-01
5.32672167e-01 -3.10765296e-01 -1.17427446e-01 2.65877247e-01
-6.82328761e-01 5.82621455e-01 2.17696261e-02 9.32682753e-01
-4.19617474e-01 1.11050241e-01 8.24412584e-01 6.35228336e-01
8.40229094e-01 8.84050056e-02 -5.45438826e-01 -4.15905595e-01
-1.13343441e+00 -2.74110548e-02 8.52816999e-01 1.77326396e-01
4.73760784e-01 4.76987571e-01 -2.60585785e-01 1.12056494e+00
1.09971218e-01 -1.11409262e-01 9.41567838e-01 -4.41186816e-01
2.83334672e-01 6.44256314e-03 -1.30285084e-01 -3.86111468e-01
-4.64663327e-01 -6.98970795e-01 -6.44197047e-01 3.32867593e-01
5.50363719e-01 -2.45772779e-01 -9.59614336e-01 1.96823573e+00
1.15802243e-01 2.92330772e-01 7.71148279e-02 5.31518161e-01
3.19216251e-01 9.19319153e-01 9.21568200e-02 -6.50115967e-01
1.32155907e+00 -9.90472436e-01 -9.28172290e-01 -3.54832262e-01
6.34976447e-01 -7.74097681e-01 5.47846079e-01 5.01211405e-01
-9.78386641e-01 -6.95881128e-01 -8.76359522e-01 1.72506824e-01
-7.06360042e-01 8.03452730e-02 1.67793527e-01 8.11585844e-01
-1.01313794e+00 9.02059376e-01 -9.94898796e-01 -3.53609979e-01
-2.64265776e-01 4.43230450e-01 7.83344284e-02 5.63264906e-01
-1.47988212e+00 1.19343722e+00 3.45663816e-01 4.80203897e-01
-9.59874570e-01 -2.96184719e-01 -5.03853381e-01 5.98071069e-02
2.10259736e-01 -2.54733354e-01 1.45844197e+00 -1.32801211e+00
-1.83611822e+00 9.19554293e-01 -4.49990571e-01 -9.35291886e-01
4.82409298e-01 -4.41845544e-02 -7.20644772e-01 -1.90616563e-01
-4.06445235e-01 1.69230446e-01 1.24809182e+00 -8.60650003e-01
-5.02349675e-01 -4.31027442e-01 -5.08372247e-01 2.47112326e-02
-2.23229840e-01 4.53441918e-01 -6.76691309e-02 -8.04776609e-01
-6.63185641e-02 -1.08211958e+00 2.10034335e-03 -6.85127974e-01
-2.88538188e-01 -5.43408275e-01 8.29148114e-01 -8.37411523e-01
1.49146855e+00 -2.03045249e+00 2.27022544e-01 3.98599952e-02
-2.27447644e-01 7.61263430e-01 7.80714750e-02 5.21318734e-01
-3.77991408e-01 -1.01525690e-02 -1.49442747e-01 -7.57357538e-01
-2.06111327e-01 2.63850868e-01 -7.66164601e-01 4.92230862e-01
6.89120069e-02 3.88189137e-01 -5.11370897e-01 -1.75915480e-01
2.54129827e-01 7.34352291e-01 3.65536690e-01 2.80925035e-01
-1.76609769e-01 6.03939235e-01 -1.96214721e-01 9.91040543e-02
2.10090548e-01 5.04952550e-01 1.09700590e-01 1.99734628e-01
-5.34985781e-01 7.36389399e-01 -7.67488837e-01 1.39152002e+00
-7.19792604e-01 1.02105463e+00 -6.98078889e-03 -1.07188237e+00
1.21157062e+00 8.44157577e-01 2.07644805e-01 -6.48471236e-01
2.45689332e-01 4.60774720e-01 2.49349684e-01 -4.63035554e-01
3.32010657e-01 -4.84619349e-01 2.94213653e-01 4.19216186e-01
1.57240704e-01 5.09252548e-01 8.90832953e-03 -4.90924567e-01
9.97845054e-01 -1.51880264e-01 7.58462995e-02 -3.47413778e-01
7.19208479e-01 -6.11502886e-01 3.42452973e-01 3.91615123e-01
-2.80453652e-01 2.64100552e-01 5.53756177e-01 -3.95526111e-01
-8.08641076e-01 -5.83655894e-01 -1.48702160e-01 1.13707364e+00
-5.82766652e-01 -1.41973943e-01 -7.06011593e-01 -3.77090335e-01
-4.50149179e-01 7.56124675e-01 -6.00399911e-01 -4.61145014e-01
-1.03694236e+00 -5.70966542e-01 8.44211221e-01 4.58203971e-01
1.07525080e-01 -1.48663664e+00 -8.23018014e-01 5.35240412e-01
1.11498497e-01 -8.71187210e-01 -1.97282970e-01 9.30013239e-01
-1.13470376e+00 -3.36530596e-01 -9.33077037e-01 -6.01776361e-01
-3.76691744e-02 -2.13245541e-01 9.93299663e-01 -2.85864621e-01
6.82246983e-02 1.02995127e-01 -1.86285079e-01 -4.65701669e-01
-8.30887377e-01 4.15624797e-01 4.61064205e-02 1.70852810e-01
3.97192508e-01 -6.11204505e-01 -2.70152867e-01 2.42330328e-01
-7.93060482e-01 -3.10813457e-01 4.96471852e-01 5.99311888e-01
3.30779761e-01 2.10868180e-01 6.37075186e-01 -7.60300577e-01
6.60848677e-01 -3.62785697e-01 -6.25257254e-01 2.19934553e-01
-6.54796302e-01 3.69244844e-01 4.51253712e-01 -5.31779110e-01
-1.19006205e+00 -6.50926232e-02 -5.19677401e-01 -5.11417806e-01
-2.99154490e-01 6.63637102e-01 -8.39371011e-02 2.51662761e-01
2.68547058e-01 2.58691341e-01 -1.25528619e-01 -7.39076018e-01
-1.41669273e-01 7.50608265e-01 2.48825252e-01 -3.88563991e-01
1.00243144e-01 7.68503621e-02 -5.82008585e-02 -1.18811429e+00
-6.86333120e-01 -5.93850195e-01 -6.57122195e-01 -4.98242170e-01
7.60549366e-01 -4.68390435e-01 -4.01913524e-01 5.87248266e-01
-1.38379467e+00 -6.13714159e-01 -2.80251533e-01 5.15329897e-01
-4.65184510e-01 -2.28284061e-01 -7.21584141e-01 -1.44050634e+00
-2.97854811e-01 -1.06200564e+00 6.74725533e-01 5.72897829e-02
-2.13248506e-01 -1.31380165e+00 5.95279157e-01 2.78794821e-02
3.70197028e-01 -2.01092705e-01 8.14361334e-01 -9.47317481e-01
-1.87324896e-01 -1.68684721e-01 2.26435661e-01 6.81308329e-01
-2.22330287e-01 1.82960695e-03 -1.61599076e+00 -3.12815756e-01
5.67441463e-01 1.22421116e-01 1.23854685e+00 7.56799459e-01
8.92931879e-01 -1.71718314e-01 -2.77561843e-01 3.02758336e-01
1.15619326e+00 5.52061021e-01 7.20388472e-01 2.45478675e-01
3.80484849e-01 1.06661785e+00 3.62565637e-01 3.69219445e-02
-3.77219647e-01 7.95270205e-01 2.32930705e-01 1.46772400e-01
-1.81797609e-01 -6.06704839e-02 8.46627653e-01 1.06673622e+00
6.07517213e-02 -4.08368587e-01 -1.00425923e+00 7.01739073e-01
-1.86941874e+00 -9.99765217e-01 -9.23627317e-02 2.52392340e+00
6.29626215e-01 5.24980485e-01 2.10785821e-01 5.67798257e-01
9.13203180e-01 5.43941557e-01 -4.17455226e-01 -8.80090356e-01
-3.00146528e-02 1.98499337e-01 3.34201485e-01 7.20874250e-01
-8.64188135e-01 7.46235609e-01 5.87296486e+00 1.01997364e+00
-1.69722891e+00 2.96048939e-01 7.54262388e-01 -1.92243457e-02
7.25084022e-02 -1.30716590e-02 -9.32389855e-01 4.73682404e-01
1.89588594e+00 2.64503986e-01 2.00099990e-01 3.90015244e-01
5.26628733e-01 -1.71070009e-01 -9.84360874e-01 8.53779793e-01
-2.66925525e-02 -8.31447423e-01 -2.48049721e-01 1.71761304e-01
2.69044906e-01 2.12932244e-01 1.97285697e-01 4.25900936e-01
-2.92334765e-01 -1.13670599e+00 9.66884613e-01 5.76151788e-01
5.85498750e-01 -7.22440362e-01 5.96921504e-01 4.81543183e-01
-1.23470187e+00 -1.05255887e-01 -4.40769978e-02 -1.40046209e-01
3.60602617e-01 8.70424211e-01 -8.70067537e-01 1.30031839e-01
4.69163924e-01 4.29212362e-01 -2.87442595e-01 1.01851046e+00
-8.04700553e-02 1.05581212e+00 -1.68924078e-01 7.07322881e-02
4.36401367e-01 -4.81194742e-02 8.87912989e-01 1.50308490e+00
2.86390364e-01 -5.10048270e-01 -2.61175334e-01 6.41258657e-01
5.04509322e-02 -1.61262587e-01 -4.69068170e-01 -2.81684667e-01
3.25494945e-01 9.48033094e-01 -1.00584900e+00 -2.48206556e-01
7.80371111e-03 9.98264074e-01 2.24848360e-01 4.22725379e-01
-5.10119498e-01 -3.27470094e-01 4.75005180e-01 -6.79837689e-02
3.11902881e-01 -2.56418258e-01 -5.19775413e-02 -7.86637723e-01
-5.92704415e-02 -5.00965118e-01 2.34260306e-01 -5.13271332e-01
-7.70608664e-01 9.53791201e-01 -3.60758565e-02 -7.88250506e-01
-8.31674159e-01 -6.29471540e-01 -8.36384535e-01 1.15976965e+00
-1.53488302e+00 -8.44940186e-01 4.02780056e-01 3.54392767e-01
9.09341753e-01 -1.58919334e-01 7.58140922e-01 2.80668318e-01
-8.98825049e-01 2.62107521e-01 1.21599391e-01 -1.32680297e-01
5.58721125e-01 -1.19342816e+00 2.98316747e-01 8.41084421e-01
3.12925786e-01 5.45086384e-01 1.16884851e+00 -3.13352227e-01
-7.25742877e-01 -7.88271844e-01 1.47635794e+00 -2.30195150e-01
7.17558444e-01 -4.09692883e-01 -1.17120326e+00 4.83472615e-01
3.52747947e-01 -6.90515265e-02 3.89177114e-01 1.38253450e-01
-1.88590050e-01 -1.41689017e-01 -6.38440728e-01 1.68430179e-01
4.32293028e-01 -1.08583939e+00 -5.88071942e-01 -3.12657245e-02
5.42033970e-01 -1.65301412e-01 -4.45030689e-01 1.30339012e-01
5.04905999e-01 -1.16422021e+00 5.97159147e-01 -4.18944567e-01
-1.29662290e-01 3.54608856e-02 9.25425589e-02 -1.29046464e+00
1.84775814e-01 -8.85899544e-01 -3.13111991e-01 1.42214811e+00
4.00841713e-01 -7.99953222e-01 6.09242797e-01 2.19212666e-01
-2.84246475e-01 -7.08693743e-01 -1.25625598e+00 -9.75014448e-01
9.26740840e-02 -5.00428379e-01 1.51043519e-01 5.96667707e-01
-2.72439927e-01 6.19385123e-01 -4.26036656e-01 5.72565161e-02
9.30146081e-04 -1.03415012e-01 6.83510602e-02 -1.21615553e+00
-1.66830197e-01 -7.85188437e-01 -1.04018472e-01 -8.54440749e-01
5.40060282e-01 -4.65342432e-01 2.14667559e-01 -9.95071173e-01
-3.66570979e-01 -1.84430316e-01 -7.73031414e-01 1.05536200e-01
2.59155929e-01 -2.57764041e-01 -9.47818719e-03 1.19826190e-01
-1.00315444e-01 3.23725015e-01 5.16208351e-01 5.59940599e-02
-3.73419940e-01 6.90325916e-01 1.86642155e-01 5.52647114e-01
9.46155787e-01 -6.42323494e-01 -4.06469434e-01 -2.27916777e-01
1.88031018e-01 4.75940198e-01 3.88197452e-01 -1.09451807e+00
3.29934865e-01 2.49146715e-01 -1.43806681e-01 -6.13075197e-01
5.97561240e-01 -6.16420567e-01 1.91891998e-01 6.56928241e-01
-6.86026692e-01 9.67222229e-02 4.02193069e-01 5.22046328e-01
-5.68066835e-01 -7.02899337e-01 8.89584363e-01 -2.20178932e-01
-5.20795107e-01 -7.33306333e-02 -7.45242178e-01 -3.50957841e-01
6.27635896e-01 -1.13499425e-01 1.03104182e-01 -3.94314587e-01
-1.01638210e+00 -3.74491870e-01 -1.19298764e-01 3.61156940e-01
1.75820455e-01 -8.36911976e-01 -4.02211398e-01 1.52930439e-01
-2.14270815e-01 -5.14690816e-01 5.17188907e-01 9.96194303e-01
-1.51188606e-02 8.15207422e-01 5.07789962e-02 -4.05296981e-01
-1.49313402e+00 5.65643311e-01 7.19393790e-01 -5.96929789e-01
-3.03121358e-01 8.92983973e-01 -8.33891109e-02 -3.50365415e-02
6.25563741e-01 -2.92740226e-01 -5.76238215e-01 6.09736264e-01
4.97476101e-01 5.52472770e-01 5.39168954e-01 -8.75900626e-01
-2.48219058e-01 4.00382340e-01 -3.18342149e-02 -5.41820586e-01
1.17118025e+00 -3.70245069e-01 -1.44599855e-01 1.38644791e+00
1.35793698e+00 -1.11025617e-01 -1.10271764e+00 -1.47545919e-01
4.89148915e-01 2.86008745e-01 5.16495109e-01 -7.75868595e-01
-8.39000285e-01 1.40197492e+00 9.45558012e-01 6.65946782e-01
1.00092471e+00 -2.41666272e-01 8.01060259e-01 2.82844380e-02
-1.45017775e-02 -1.18028545e+00 -2.23951772e-01 8.32606554e-01
8.93450975e-01 -9.20840681e-01 -5.05024314e-01 1.33740425e-01
-2.74005532e-01 1.43978596e+00 3.38640250e-02 -2.66872346e-02
8.56061518e-01 2.41834641e-01 8.47401693e-02 -2.44062953e-02
-1.06501329e+00 -3.50864500e-01 2.84056276e-01 2.74655342e-01
6.77853048e-01 6.81910589e-02 -1.95507571e-01 2.57284671e-01
-2.40215603e-02 -1.93339363e-01 2.83959895e-01 7.92427242e-01
-4.66424644e-01 -1.42149103e+00 -4.84297007e-01 2.94150174e-01
-7.39402950e-01 -9.19822529e-02 -2.28603750e-01 4.52332675e-01
1.07081712e-03 1.01471031e+00 1.56699806e-01 -1.99715272e-01
2.01302484e-01 6.04557991e-01 2.28406683e-01 -6.64082706e-01
-1.09364617e+00 4.14279759e-01 2.15797052e-01 -2.32579648e-01
-5.20613611e-01 -7.04797328e-01 -1.04449284e+00 8.53176191e-02
-4.70770746e-01 2.65254319e-01 1.04689074e+00 1.24936402e+00
8.23990628e-02 8.44926775e-01 3.60277832e-01 -1.10557318e+00
-5.44562578e-01 -1.16881299e+00 -6.17051125e-01 -9.65127200e-02
8.84045541e-01 -5.19881964e-01 -5.97709119e-01 -1.01945484e-02] | [14.513855934143066, 6.205336570739746] |
b3b18038-4903-4d69-83a7-c3dda8d19d07 | video-moment-retrieval-from-text-queries-via | 2204.09409 | null | https://arxiv.org/abs/2204.09409v3 | https://arxiv.org/pdf/2204.09409v3.pdf | Video Moment Retrieval from Text Queries via Single Frame Annotation | Video moment retrieval aims at finding the start and end timestamps of a moment (part of a video) described by a given natural language query. Fully supervised methods need complete temporal boundary annotations to achieve promising results, which is costly since the annotator needs to watch the whole moment. Weakly supervised methods only rely on the paired video and query, but the performance is relatively poor. In this paper, we look closer into the annotation process and propose a new paradigm called "glance annotation". This paradigm requires the timestamp of only one single random frame, which we refer to as a "glance", within the temporal boundary of the fully supervised counterpart. We argue this is beneficial because comparing to weak supervision, trivial cost is added yet more potential in performance is provided. Under the glance annotation setting, we propose a method named as Video moment retrieval via Glance Annotation (ViGA) based on contrastive learning. ViGA cuts the input video into clips and contrasts between clips and queries, in which glance guided Gaussian distributed weights are assigned to all clips. Our extensive experiments indicate that ViGA achieves better results than the state-of-the-art weakly supervised methods by a large margin, even comparable to fully supervised methods in some cases. | ['Xiaowei Guo', 'Yu-Gang Jiang', 'Huyang Sun', 'Jingjing Chen', 'Elena Daskalaki', 'Pai Peng', 'Tianwen Qian', 'Ran Cui'] | 2022-04-20 | null | null | null | null | ['moment-retrieval'] | ['computer-vision'] | [-5.08265495e-02 -4.46783490e-02 -5.17847180e-01 -3.81271005e-01
-1.19946623e+00 -7.42685795e-01 8.10671985e-01 2.61977106e-01
-6.57310069e-01 3.47326487e-01 2.93737650e-01 2.62274027e-01
3.05136796e-02 -1.90825090e-01 -8.45291793e-01 -7.75376439e-01
-3.55663806e-01 9.77757052e-02 4.22017902e-01 2.08268419e-01
8.39446709e-02 7.42835030e-02 -1.39686489e+00 2.10751817e-01
2.62239605e-01 1.07201397e+00 2.46070698e-01 6.35821939e-01
1.93132870e-02 1.06049263e+00 -3.81691247e-01 -4.62032974e-01
2.73199677e-01 -3.24154764e-01 -8.07161093e-01 2.80455947e-01
7.27812052e-01 -5.24052799e-01 -2.99457073e-01 1.05404902e+00
3.17834705e-01 4.26428914e-01 3.26724350e-01 -1.39782488e+00
-3.15285593e-01 6.19609773e-01 -6.85274661e-01 4.14494962e-01
8.31857443e-01 -8.93368199e-02 1.42158914e+00 -9.93799388e-01
1.06508505e+00 9.30810690e-01 5.82779408e-01 4.49907392e-01
-1.04727328e+00 -2.53528625e-01 4.48105454e-01 3.60710204e-01
-1.47204578e+00 -4.50016558e-01 9.85594273e-01 -3.98334026e-01
7.14960277e-01 2.39774972e-01 4.67640162e-01 1.04042864e+00
-2.12376952e-01 1.15952802e+00 7.52725482e-01 -4.40038472e-01
3.09875995e-01 -1.79873426e-02 1.34589314e-01 8.14654171e-01
-4.29043442e-01 -2.70232499e-01 -9.16847706e-01 -2.78211653e-01
6.05436444e-01 1.79392949e-01 -4.69488531e-01 -4.38732624e-01
-1.71247959e+00 6.81174695e-01 1.92843497e-01 4.75917459e-01
-3.28890115e-01 1.07829593e-01 5.92024803e-01 4.43800867e-01
9.07259643e-01 2.38314256e-01 -2.41729915e-01 -4.73103434e-01
-1.39478576e+00 8.35406035e-02 7.86624849e-01 9.57592487e-01
7.85985053e-01 -4.37751293e-01 -2.97532618e-01 5.37349999e-01
7.24370331e-02 2.05228895e-01 3.80977392e-01 -1.25132716e+00
4.54924822e-01 2.35225648e-01 2.38991037e-01 -9.82503831e-01
-1.82977095e-01 -2.01008752e-01 -4.48932588e-01 -1.34744078e-01
5.33608973e-01 5.91638461e-02 -6.04527295e-01 1.69511747e+00
2.64920443e-01 5.40887892e-01 -1.47230268e-01 1.26049829e+00
6.22082651e-01 8.68399858e-01 3.28962468e-02 -6.64270937e-01
1.22761118e+00 -1.39703083e+00 -7.84052372e-01 -7.05152377e-02
6.82925701e-01 -6.91084266e-01 1.31658471e+00 4.18341607e-01
-1.23467648e+00 -4.15390253e-01 -9.23060715e-01 -2.24825576e-01
-1.40229791e-01 -1.31371347e-02 6.81386709e-01 8.78247246e-02
-1.01778114e+00 5.02225757e-01 -9.75827754e-01 -4.33430195e-01
-4.05357592e-02 -7.01732337e-02 -6.13472164e-01 -9.61376131e-02
-1.06541526e+00 3.81597072e-01 2.13600516e-01 -2.31874660e-02
-8.92360389e-01 -6.05744123e-01 -1.02253103e+00 -9.75801200e-02
8.20406795e-01 -4.18575794e-01 1.47339880e+00 -1.04472256e+00
-1.25626266e+00 1.16922879e+00 -4.38607723e-01 -5.72516203e-01
8.21555793e-01 -5.92895329e-01 -1.78157076e-01 7.71254063e-01
2.38890201e-01 7.72792041e-01 1.02470946e+00 -1.03734994e+00
-7.11332798e-01 -4.07471582e-02 5.71462333e-01 2.71513253e-01
-3.57111067e-01 3.59173596e-01 -1.20540428e+00 -7.73611605e-01
2.74772704e-01 -9.91778553e-01 1.86738204e-02 1.27279490e-01
-1.11843273e-01 -4.96545315e-01 9.69521523e-01 -4.66656417e-01
1.46720743e+00 -2.36712766e+00 5.65405414e-02 -1.03075661e-01
1.99452966e-01 -2.17666104e-01 -2.14544442e-02 5.48873782e-01
-7.91459009e-02 -7.38170817e-02 7.17755109e-02 -6.38397098e-01
7.29556903e-02 9.49133933e-02 -6.30297959e-01 8.07141900e-01
-1.23533361e-01 8.43147457e-01 -1.35677660e+00 -8.14092100e-01
1.47866249e-01 3.16716582e-01 -5.63545585e-01 4.21354800e-01
-4.87727046e-01 4.16732252e-01 -4.06055272e-01 7.83029318e-01
1.92245513e-01 -4.49668020e-01 -1.74379036e-01 -3.70573550e-01
-1.61999986e-01 -3.59186754e-02 -1.00223720e+00 2.19094849e+00
-1.90173820e-01 7.98950791e-01 -1.50401101e-01 -8.11791956e-01
6.47469997e-01 5.62498808e-01 6.33507133e-01 -6.04928255e-01
-2.30057389e-01 3.51859592e-02 -6.63533866e-01 -7.99591899e-01
6.24409258e-01 5.99196069e-02 -2.16410458e-01 5.39386570e-01
6.88903779e-02 1.89587891e-01 4.30897355e-01 6.18506670e-01
1.07294166e+00 5.40655017e-01 2.34982118e-01 5.49266487e-02
4.04898018e-01 -1.45298302e-01 6.40713990e-01 8.94728959e-01
-4.97797936e-01 8.10683787e-01 7.01587200e-01 -3.83080244e-01
-9.97974038e-01 -7.58815348e-01 2.25259706e-01 1.54696488e+00
3.94627362e-01 -8.33831429e-01 -7.38786221e-01 -8.02757978e-01
-3.81278932e-01 2.49710575e-01 -5.53587675e-01 9.77225974e-02
-6.42736256e-01 -1.78962439e-01 3.27141702e-01 4.24431682e-01
5.22407174e-01 -8.96187007e-01 -7.52694845e-01 -8.22947398e-02
-4.21645939e-01 -1.51240456e+00 -8.39315355e-01 -5.27323149e-02
-6.42374992e-01 -8.86803329e-01 -9.98456776e-01 -7.63452649e-01
3.89444441e-01 5.10578156e-01 1.06786501e+00 4.33375575e-02
9.28711146e-02 6.49955928e-01 -7.47433722e-01 1.28575057e-01
1.49612099e-01 -9.63258371e-02 1.33870065e-01 2.36872882e-01
3.16418737e-01 -4.23409224e-01 -7.06735730e-01 2.67591089e-01
-1.00756073e+00 6.17396198e-02 3.17854196e-01 6.61618292e-01
8.52963388e-01 -1.12078510e-01 1.52222902e-01 -8.09773862e-01
1.82095826e-01 -5.03379583e-01 -4.41080421e-01 2.63280630e-01
-3.42890531e-01 -1.44393802e-01 4.61014509e-01 -6.26121879e-01
-8.31835866e-01 1.31513149e-01 2.37054557e-01 -9.55927789e-01
-6.04211837e-02 7.62192965e-01 7.33894631e-02 2.45364234e-01
4.92990434e-01 8.65963548e-02 -3.34257573e-01 -4.82324094e-01
4.17613417e-01 3.29994917e-01 8.92198086e-01 -4.92946804e-01
7.23370135e-01 7.87511766e-01 -2.45924890e-01 -7.57903695e-01
-1.29912198e+00 -1.04231989e+00 -6.33224607e-01 -6.04628146e-01
8.35098743e-01 -1.18041575e+00 -5.97273111e-01 1.71828344e-01
-1.03094602e+00 -4.85433072e-01 -2.04071969e-01 6.41757071e-01
-7.33825445e-01 7.11022079e-01 -7.17833161e-01 -7.06887603e-01
-2.70153303e-02 -8.06336939e-01 1.51234472e+00 -2.31191423e-02
-2.09173992e-01 -9.97526109e-01 1.15280285e-01 3.11087579e-01
2.48726998e-02 3.82329732e-01 3.06700766e-01 -7.59897470e-01
-5.32147586e-01 -5.93592584e-01 -1.55261248e-01 1.41807362e-01
-1.32678658e-01 3.99121456e-02 -9.25880551e-01 -3.14499795e-01
9.73185897e-02 -4.62748915e-01 7.92416394e-01 2.41245478e-01
1.06283259e+00 -3.90183061e-01 -2.53412008e-01 5.49661934e-01
1.32642043e+00 -1.90810755e-01 3.58589798e-01 4.87881601e-01
7.72074521e-01 6.21585488e-01 1.12417424e+00 5.84722340e-01
3.23228955e-01 8.63221705e-01 2.35290468e-01 -5.74487410e-02
7.92130381e-02 -4.64037359e-01 6.91534221e-01 7.56713271e-01
1.99653171e-02 -3.06184858e-01 -7.31736422e-01 4.95718211e-01
-2.27580810e+00 -1.27555037e+00 1.17518604e-01 2.09995699e+00
7.91483819e-01 2.26882905e-01 2.40997300e-01 1.83090910e-01
7.25583613e-01 7.55330503e-01 -3.12873423e-01 6.99935183e-02
-1.82544485e-01 -5.12515962e-01 2.46142372e-01 5.21834373e-01
-1.37145710e+00 9.24192250e-01 6.31192160e+00 9.28304017e-01
-1.19443595e+00 3.11427176e-01 5.20033002e-01 -3.90588939e-01
-1.66864656e-02 1.55953869e-01 -5.95967114e-01 4.41786706e-01
9.13542509e-01 -2.28236660e-01 1.95540860e-01 9.70757008e-01
6.01067424e-01 -3.83784264e-01 -1.55834532e+00 1.20720088e+00
5.49208045e-01 -1.20040190e+00 -1.26968920e-01 -3.09972078e-01
8.15311849e-01 -4.63803124e-04 -1.86119020e-01 2.04079524e-01
-3.50925922e-01 -4.94471580e-01 1.07580233e+00 6.24265850e-01
6.89880669e-01 -4.43035901e-01 6.24785185e-01 4.09410805e-01
-1.26081216e+00 1.76352844e-01 -1.44382656e-01 -3.83648984e-02
4.97004420e-01 7.08016455e-01 -4.52455878e-01 3.95340800e-01
8.74498427e-01 1.01152766e+00 -5.93004346e-01 1.10591567e+00
-2.35114604e-01 6.45305157e-01 -2.70413429e-01 3.23769778e-01
5.12449503e-01 1.44897671e-02 6.85573399e-01 1.43244052e+00
1.53177008e-01 4.11717035e-02 6.71246946e-01 3.10734928e-01
-1.85922727e-01 1.77805737e-01 -5.29328942e-01 3.08468863e-02
2.39592686e-01 1.32199812e+00 -8.75458121e-01 -6.09076858e-01
-5.86965024e-01 1.11381698e+00 2.41603673e-01 5.50449491e-01
-9.28548694e-01 -1.70336857e-01 2.62633953e-02 -2.45601870e-02
2.51576483e-01 -2.40745053e-01 3.37796301e-01 -1.43611658e+00
4.47077274e-01 -6.20506048e-01 7.78343201e-01 -1.13892078e+00
-1.09848857e+00 4.94695157e-01 2.56705254e-01 -1.63692367e+00
-4.96404141e-01 -1.18793897e-01 -4.28839296e-01 2.03034356e-01
-1.64652383e+00 -1.12316680e+00 -4.34250653e-01 8.77167106e-01
8.65386069e-01 2.61748552e-01 4.59820718e-01 4.11017030e-01
-2.66417980e-01 4.23967272e-01 -1.52822465e-01 1.76421136e-01
1.19702315e+00 -1.34788871e+00 -2.17616767e-01 9.62708533e-01
5.56866109e-01 4.48757112e-01 9.16084111e-01 -4.43129480e-01
-1.40924037e+00 -9.32270885e-01 1.21315002e+00 -4.75744575e-01
9.43118215e-01 -3.22625041e-01 -9.97201324e-01 8.80186737e-01
1.20540850e-01 4.29791957e-01 5.69595337e-01 3.23803052e-02
-5.84170580e-01 -2.32016258e-02 -6.74541771e-01 5.56183934e-01
9.10191774e-01 -9.82727945e-01 -7.41704524e-01 7.01902330e-01
9.67892468e-01 -4.98322159e-01 -7.14934587e-01 2.97788560e-01
4.17692512e-01 -1.03353643e+00 7.43964314e-01 -2.70316958e-01
3.55203122e-01 -5.39538920e-01 -1.24918275e-01 -6.97913229e-01
9.99416411e-03 -1.01742363e+00 -4.31865782e-01 1.33361804e+00
2.14169562e-01 2.64325272e-02 7.48867869e-01 6.12819374e-01
-4.46735919e-02 -5.28690398e-01 -7.79622614e-01 -1.00319338e+00
-5.40897250e-01 -7.65425682e-01 9.94601175e-02 1.15747964e+00
4.30810362e-01 2.20688358e-01 -5.81011653e-01 2.28998765e-01
6.07634425e-01 2.33367011e-01 6.53626919e-01 -9.40453708e-01
-3.30402583e-01 -2.75753558e-01 -4.48280185e-01 -1.29349673e+00
4.15712714e-01 -5.91649652e-01 3.26400906e-01 -1.10766518e+00
5.46794713e-01 2.81907362e-03 -3.79875720e-01 5.03894091e-01
-1.70514613e-01 5.79425633e-01 1.54873535e-01 6.87023342e-01
-1.53968859e+00 2.00293049e-01 8.24641883e-01 -1.05719119e-01
-2.73160070e-01 -2.22010210e-01 -1.49635270e-01 9.21135187e-01
2.49619752e-01 -2.76759565e-01 -3.52232307e-01 -5.55368483e-01
3.85454237e-01 2.90641785e-01 4.48962390e-01 -8.82561147e-01
6.97668672e-01 4.92898412e-02 -1.22078046e-01 -5.01942098e-01
2.96522766e-01 -7.49814391e-01 4.14473861e-02 -7.84456432e-02
-7.44344354e-01 1.45493865e-01 -2.77445227e-01 7.92930305e-01
-7.14541495e-01 -1.90504491e-01 2.98802257e-01 1.82083680e-03
-1.09902346e+00 4.85908329e-01 -7.25631714e-02 -7.58940950e-02
1.04085565e+00 -1.88263386e-01 -1.19050555e-01 -7.40678549e-01
-9.35755551e-01 1.80679321e-01 5.92183590e-01 4.95780259e-01
4.08156186e-01 -1.36047339e+00 -4.68956351e-01 -3.59658569e-01
4.74598646e-01 -6.19929992e-02 1.12167433e-01 1.13479340e+00
-3.52968693e-01 3.70215863e-01 4.39450115e-01 -8.38181257e-01
-1.36314893e+00 8.87785554e-01 -1.78296626e-01 -2.09281787e-01
-8.78336310e-01 6.53263509e-01 2.77437180e-01 3.55509341e-01
7.95640588e-01 -2.11403280e-01 -1.63045809e-01 3.99120003e-01
7.52914310e-01 8.91715512e-02 -1.42512843e-01 -8.44194651e-01
-2.61243999e-01 5.74286461e-01 -3.21261771e-02 -2.47461811e-01
1.18601084e+00 -4.76234853e-01 -3.12843360e-02 8.89911950e-01
1.50874674e+00 1.55384511e-01 -1.59095597e+00 -5.62723577e-01
3.25368255e-01 -5.42670667e-01 4.98948097e-02 -4.05017465e-01
-9.84659851e-01 5.70053339e-01 3.18272203e-01 4.08850253e-01
1.13696921e+00 4.31727827e-01 8.23136568e-01 5.03218114e-01
3.57936621e-01 -1.07188952e+00 3.15595835e-01 3.37902606e-01
6.79699302e-01 -1.41723466e+00 -3.93687077e-02 -3.08065951e-01
-7.16547191e-01 9.67124879e-01 3.17632407e-01 -1.56262830e-01
3.75573963e-01 -5.76106161e-02 1.48734555e-01 -1.99693263e-01
-9.07425284e-01 -2.96111941e-01 3.45690519e-01 1.24745466e-01
4.04827893e-01 -3.94169092e-01 -2.65688509e-01 3.72032911e-01
1.41686812e-01 2.38014311e-02 2.31402859e-01 1.10320330e+00
-3.07384998e-01 -7.29227781e-01 -2.61611640e-01 -4.00773212e-02
-7.87643909e-01 3.39735881e-03 -1.56493127e-01 6.20000958e-01
-1.59237236e-01 9.34165180e-01 1.68215767e-01 -1.20941795e-01
1.37193963e-01 1.57542646e-01 2.31593162e-01 -5.05785704e-01
-3.44879687e-01 5.33486724e-01 2.22840328e-02 -9.98716474e-01
-1.11518764e+00 -6.18788600e-01 -1.19139278e+00 -1.82769820e-02
-2.14711323e-01 4.54185575e-01 3.26267332e-01 1.13569450e+00
9.15706307e-02 -7.91091472e-03 7.74563670e-01 -1.13284969e+00
-7.54084066e-02 -6.75881624e-01 -5.02796590e-01 7.93818891e-01
5.99632800e-01 -3.69104594e-01 -7.57237017e-01 6.60401464e-01] | [10.105945587158203, 0.696228563785553] |
365e8643-5830-438f-8840-5a599ccaeeed | domain-embedded-multi-model-generative | 2002.02909 | null | https://arxiv.org/abs/2002.02909v2 | https://arxiv.org/pdf/2002.02909v2.pdf | Domain Embedded Multi-model Generative Adversarial Networks for Image-based Face Inpainting | Prior knowledge of face shape and structure plays an important role in face inpainting. However, traditional face inpainting methods mainly focus on the generated image resolution of the missing portion without consideration of the special particularities of the human face explicitly and generally produce discordant facial parts. To solve this problem, we present a domain embedded multi-model generative adversarial model for inpainting of face images with large cropped regions. We firstly represent only face regions using the latent variable as the domain knowledge and combine it with the non-face parts textures to generate high-quality face images with plausible contents. Two adversarial discriminators are finally used to judge whether the generated distribution is close to the real distribution or not. It can not only synthesize novel image structures but also explicitly utilize the embedded face domain knowledge to generate better predictions with consistency on structures and appearance. Experiments on both CelebA and CelebA-HQ face datasets demonstrate that our proposed approach achieved state-of-the-art performance and generates higher quality inpainting results than existing ones. | ['Qi Song', 'Bin Kong', 'Siwei Lyu', 'Jiancheng Lv', 'Canghong Shi', 'Youbing Yin', 'Xian Zhang', 'Xin Wang', 'Xiaojie Li'] | 2020-02-05 | null | null | null | null | ['facial-inpainting'] | ['computer-vision'] | [ 2.82731801e-01 3.28281343e-01 4.95438762e-02 -3.65241289e-01
-7.03413904e-01 -4.70762014e-01 2.94561177e-01 -9.91242766e-01
2.39655226e-01 1.09445739e+00 8.96079689e-02 2.94066936e-01
3.42100769e-01 -8.42124701e-01 -1.00495434e+00 -8.59366059e-01
4.80410039e-01 5.54538727e-01 -3.55392456e-01 -2.01411992e-01
-1.60773292e-01 7.05703735e-01 -1.59239125e+00 7.05002129e-01
9.55276787e-01 9.79293942e-01 1.04860282e-02 4.39581752e-01
-1.41268328e-01 7.48539925e-01 -7.97179461e-01 -7.02658534e-01
5.96249282e-01 -8.34140420e-01 -4.45064336e-01 6.30281508e-01
8.14039171e-01 -6.82407439e-01 -3.39057416e-01 1.21189106e+00
4.22483772e-01 -2.48840913e-01 9.63957310e-01 -1.19665110e+00
-1.10929036e+00 3.31181847e-02 -1.02918077e+00 -1.50286272e-01
4.96045023e-01 1.78332493e-01 2.52601057e-01 -9.66269016e-01
8.25037539e-01 1.71432781e+00 4.70257968e-01 9.10392046e-01
-1.46936512e+00 -9.74586904e-01 -1.21958032e-01 -6.34878427e-02
-1.40914178e+00 -7.74206460e-01 1.27071190e+00 -2.18773574e-01
2.48806641e-01 2.50324100e-01 2.84416556e-01 1.24936032e+00
3.12307715e-01 4.26160783e-01 1.35463071e+00 -4.42894399e-01
-7.95521587e-02 1.75875768e-01 -9.17323291e-01 8.54687691e-01
-6.27010167e-02 3.02603155e-01 -4.00594324e-01 -3.35923612e-01
1.15513885e+00 -2.10514404e-02 -3.00184011e-01 -4.40637395e-02
-7.38864601e-01 7.83041060e-01 2.58878529e-01 -1.52164260e-02
-6.50919437e-01 -8.72498304e-02 4.80649062e-03 3.90162230e-01
6.67509496e-01 1.90225631e-01 -1.61522314e-01 4.46980000e-01
-1.14252484e+00 4.55909342e-01 4.95881349e-01 1.05202913e+00
8.79878759e-01 4.67559665e-01 -2.46780396e-01 1.20512724e+00
1.38788268e-01 6.60757124e-01 2.33930320e-01 -1.25386763e+00
3.69364589e-01 3.47205907e-01 2.19075173e-01 -1.17434537e+00
4.66275603e-01 6.36048149e-03 -7.92541564e-01 5.05457401e-01
2.42766351e-01 -2.05358401e-01 -1.14799678e+00 1.80381751e+00
5.42893469e-01 3.90044868e-01 -3.74940038e-02 7.79142678e-01
8.91113639e-01 7.83461511e-01 -1.08433649e-01 -4.56568629e-01
1.26453519e+00 -8.77223909e-01 -1.04201639e+00 -2.93615013e-01
-3.73841673e-01 -1.25240123e+00 7.30686545e-01 2.45477140e-01
-1.44286430e+00 -8.81830037e-01 -8.20208907e-01 -1.37177315e-02
2.02542067e-01 2.58878142e-01 3.14493120e-01 6.45701885e-01
-1.05839217e+00 4.45601314e-01 -3.90556663e-01 1.75206378e-01
8.30937505e-01 2.11410016e-01 -7.19962478e-01 -2.82947510e-01
-9.92724240e-01 6.96802557e-01 1.64224222e-01 6.78723007e-02
-1.15111935e+00 -7.95646071e-01 -8.70676160e-01 -1.08122669e-01
2.00718537e-01 -6.45883799e-01 9.28037405e-01 -1.74884093e+00
-1.66453063e+00 1.06595075e+00 -2.18400523e-01 3.78613025e-02
7.80427337e-01 5.64314388e-02 -4.29818243e-01 2.90410936e-01
1.00823544e-01 1.01114476e+00 1.61801469e+00 -1.72522986e+00
-1.94403023e-01 -3.40274274e-01 -3.50022942e-01 4.25375402e-02
-4.64349799e-02 5.76686785e-02 -4.03982282e-01 -1.13912189e+00
-1.16657853e-01 -6.77099288e-01 2.28790045e-01 7.21855044e-01
-1.27227366e-01 2.54567564e-01 1.19183958e+00 -1.23042011e+00
5.82942247e-01 -2.09931517e+00 2.58163869e-01 -9.80068892e-02
9.53569412e-02 2.93340921e-01 -4.22185183e-01 1.91779912e-01
-2.62612551e-01 -1.05763875e-01 -3.09612244e-01 -5.14483750e-01
-4.32268113e-01 4.69779164e-01 -5.14135063e-01 5.60557187e-01
5.68361282e-01 7.63407171e-01 -6.46332383e-01 -7.39841223e-01
1.64705422e-02 8.25951099e-01 -6.79629147e-01 5.77506661e-01
-4.13055778e-01 7.89413929e-01 -3.32951516e-01 9.21499014e-01
1.32340229e+00 2.44445831e-01 -3.83260334e-03 -3.26678902e-01
4.28956240e-01 -5.21690726e-01 -9.33957815e-01 1.57612872e+00
-3.77766013e-01 5.04459679e-01 4.78854269e-01 -7.04403162e-01
1.09725702e+00 3.61112386e-01 2.69701988e-01 -5.40728033e-01
1.76382661e-01 5.32384403e-02 -1.26144171e-01 -5.32935560e-01
8.51798952e-02 -5.50800085e-01 4.85161901e-01 3.09406787e-01
1.12711594e-01 -2.45871887e-01 -2.28621334e-01 -2.06295118e-01
4.41845596e-01 4.27544057e-01 3.64040807e-02 -5.60093820e-02
5.37079930e-01 -4.73318934e-01 7.65684605e-01 7.27337301e-02
-3.93022113e-02 1.18284822e+00 6.20011985e-01 -3.63603711e-01
-1.26931334e+00 -1.13227129e+00 -1.61209665e-02 5.37686586e-01
-5.10448515e-02 2.01433942e-01 -1.05760384e+00 -7.03530490e-01
1.57625489e-02 5.38891912e-01 -1.02663827e+00 -2.61258572e-01
-5.46929657e-01 -3.29594582e-01 5.35525858e-01 3.94225866e-01
7.31834412e-01 -1.37952268e+00 -2.75742374e-02 7.57261142e-02
-4.32441294e-01 -1.09833121e+00 -7.56123841e-01 -5.79903483e-01
-7.51330376e-01 -9.56695437e-01 -1.07565510e+00 -1.12154055e+00
1.16462755e+00 -5.81530668e-02 1.18292773e+00 1.46017030e-01
-4.88349497e-01 -1.40825674e-01 -1.54753879e-01 -2.26290256e-01
-9.47489262e-01 -6.49105132e-01 -1.01572573e-01 4.73306805e-01
-1.92057163e-01 -6.59927487e-01 -6.24687076e-01 3.06798756e-01
-1.06097162e+00 2.14638650e-01 5.83358049e-01 1.15775943e+00
5.09204447e-01 2.27330148e-01 6.43040121e-01 -1.04959047e+00
4.92244571e-01 -3.90706688e-01 -4.48613435e-01 3.09042603e-01
-2.25797966e-01 6.83991384e-05 7.29998589e-01 -8.22515607e-01
-1.53999102e+00 8.30577612e-02 -1.87974334e-01 -1.03102934e+00
-2.07738370e-01 -8.18694383e-02 -6.08724296e-01 -2.53915578e-01
5.37134290e-01 5.69810212e-01 4.73216653e-01 -4.04132903e-01
2.48947382e-01 4.46769565e-01 7.02041328e-01 -9.53889906e-01
9.80944455e-01 6.61543667e-01 6.47168905e-02 -6.65077627e-01
-3.01643521e-01 4.74074244e-01 -4.91569877e-01 -1.94334194e-01
6.86475217e-01 -9.61137891e-01 -3.47606361e-01 6.50174201e-01
-1.36142540e+00 -5.77750839e-02 -2.40045264e-01 -9.02474970e-02
-6.34121716e-01 3.61333013e-01 -6.90453887e-01 -7.07026601e-01
-2.28791788e-01 -1.22615135e+00 1.25773335e+00 2.60982752e-01
1.78351060e-01 -7.08475649e-01 -2.01828539e-01 4.05935407e-01
1.74713135e-01 7.49404788e-01 9.87452805e-01 1.27272680e-01
-4.39557701e-01 4.28907154e-03 -2.23015532e-01 6.73648179e-01
6.66837633e-01 1.90027297e-01 -1.10614979e+00 -3.75123411e-01
2.56573737e-01 -3.57520252e-01 4.53015655e-01 2.28882119e-01
1.21179545e+00 -6.68913722e-01 -1.13233387e-01 7.44699657e-01
1.39474607e+00 1.51673779e-01 1.12056613e+00 -3.71333927e-01
5.58708310e-01 9.34160531e-01 7.26532221e-01 4.66703802e-01
-1.74897254e-01 6.44550383e-01 4.63033438e-01 -2.78694987e-01
-4.58922982e-01 -5.67913413e-01 4.31070149e-01 2.17984125e-01
-8.89367312e-02 -1.88399658e-01 -2.48165250e-01 4.91243422e-01
-1.38451087e+00 -1.14340436e+00 3.43179077e-01 1.86766565e+00
1.32442820e+00 -3.88705879e-01 -9.05436724e-02 -7.36392662e-02
1.03311551e+00 4.78232317e-02 -5.72418928e-01 -2.71617472e-01
-1.37196958e-01 4.96177107e-01 9.02772024e-02 4.83873069e-01
-6.99388742e-01 1.00437117e+00 6.07771826e+00 1.15787101e+00
-1.05043101e+00 2.03629225e-01 1.19656789e+00 -3.83160338e-02
-3.54524970e-01 -2.36134887e-01 -5.52614093e-01 8.41487944e-01
4.15024996e-01 1.31310225e-01 6.57254577e-01 7.22116709e-01
-9.14492831e-03 1.54544502e-01 -9.63310361e-01 1.19681036e+00
2.83042789e-01 -1.31646240e+00 4.64419812e-01 6.70052394e-02
1.26770639e+00 -1.00341964e+00 3.87196153e-01 4.45648059e-02
-8.58192965e-02 -1.43263543e+00 8.90731931e-01 7.89324701e-01
1.30859232e+00 -1.01002526e+00 4.54668790e-01 2.83139974e-01
-8.95369411e-01 2.25782126e-01 -6.54579043e-01 1.88488767e-01
-3.38280015e-02 3.96809131e-01 -6.32003903e-01 4.75403845e-01
4.78255212e-01 3.61946255e-01 -4.11469430e-01 4.48365271e-01
-2.45084971e-01 3.46274167e-01 -1.25712976e-01 7.42201030e-01
-3.45941663e-01 -3.01979959e-01 4.13878053e-01 5.45102835e-01
4.57031965e-01 2.00702339e-01 -7.02637359e-02 1.35280478e+00
-3.85106504e-01 -1.23058334e-02 -6.83320820e-01 -7.65600950e-02
4.47095335e-01 1.10492826e+00 -4.29007471e-01 -2.60525733e-01
-4.41564530e-01 1.33141696e+00 2.19461247e-01 3.68781567e-01
-1.12656283e+00 -9.00825560e-02 7.61054277e-01 3.54373008e-01
4.66750115e-01 2.62254477e-01 -2.10587293e-01 -1.00917482e+00
1.08828649e-01 -1.32295108e+00 -4.82519828e-02 -9.26147461e-01
-1.49355042e+00 9.65395808e-01 -1.26324981e-01 -1.28517270e+00
-3.82207006e-01 -3.94238204e-01 -6.37378812e-01 1.29611492e+00
-1.38901615e+00 -1.69201720e+00 -3.75239313e-01 7.63002694e-01
7.30770230e-01 -4.97626841e-01 7.50105500e-01 3.24010313e-01
-2.51900405e-01 8.87127280e-01 -6.54183254e-02 5.64294495e-02
1.07987130e+00 -6.38816833e-01 2.29939431e-01 7.67994404e-01
-1.71740130e-01 4.24598634e-01 7.94721186e-01 -8.77190650e-01
-1.24186218e+00 -1.33920634e+00 4.36209828e-01 -2.72131294e-01
-1.28022835e-01 -2.26301402e-01 -9.52972054e-01 5.59823513e-01
4.42630798e-01 3.10751706e-01 3.87623131e-01 -6.73083901e-01
-3.69715989e-01 -1.04777999e-01 -1.90560019e+00 5.01157224e-01
9.33891356e-01 -3.90667588e-01 -4.48061824e-01 1.56121418e-01
4.19196457e-01 -4.91654247e-01 -6.50037289e-01 4.22497869e-01
4.87773240e-01 -1.02178633e+00 1.13048816e+00 -5.36639273e-01
8.93344223e-01 -3.16836059e-01 -5.80671504e-02 -1.16474903e+00
-2.56453484e-01 -6.48530483e-01 -5.84883541e-02 1.61859560e+00
-1.02141939e-01 -5.29818952e-01 9.50802982e-01 4.15830940e-01
2.32936829e-01 -6.83483899e-01 -6.84313655e-01 -5.00895083e-01
2.01079056e-01 3.09189737e-01 8.88765156e-01 8.65476191e-01
-8.93357635e-01 -1.47627518e-01 -8.11627805e-01 1.64331779e-01
7.93653369e-01 2.99365222e-01 7.99510300e-01 -9.75302279e-01
-3.91109258e-01 8.49235244e-03 -2.16064125e-01 -6.37203217e-01
6.38681948e-01 -5.64342082e-01 4.01373068e-03 -9.39053893e-01
2.46565402e-01 -1.87981799e-01 2.12842301e-01 4.95358229e-01
-3.23909104e-01 7.57354498e-01 4.18391870e-03 1.47577450e-01
2.53590643e-01 7.79219210e-01 1.84759033e+00 -1.97801590e-01
1.75104991e-01 -1.66444167e-01 -8.68359089e-01 5.34092546e-01
4.95705068e-01 -4.48329866e-01 -5.74952602e-01 -4.07779485e-01
-3.27863187e-01 4.42078561e-01 4.98037398e-01 -8.59634697e-01
-3.04955751e-01 -5.03212810e-01 1.10927880e+00 -2.33590990e-01
6.00256920e-01 -9.19312835e-01 6.39358222e-01 2.84892410e-01
-1.06396511e-01 -4.81629111e-02 4.24605101e-01 4.48599070e-01
-4.17225063e-01 -1.42053112e-01 1.40729511e+00 -2.74959266e-01
-3.90185744e-01 4.87360686e-01 1.54952765e-01 -4.91852164e-02
1.16991234e+00 -2.55998999e-01 -7.10809454e-02 -6.10041797e-01
-6.44349813e-01 -2.93981135e-01 7.66731977e-01 4.61206824e-01
8.29948843e-01 -1.57419014e+00 -1.12233388e+00 8.48613441e-01
-2.47120425e-01 -3.62230614e-02 4.44393009e-01 1.40185863e-01
-7.17015624e-01 -1.41553044e-01 -8.43083680e-01 -3.59582514e-01
-1.43874288e+00 6.23570323e-01 2.97808349e-01 8.15288052e-02
-2.44117185e-01 9.50458467e-01 7.13345647e-01 -1.31006792e-01
-1.70534119e-01 2.30493441e-01 1.70733303e-01 -1.90568954e-01
5.73748469e-01 6.01031594e-02 -2.21986830e-01 -9.37632740e-01
-1.23976521e-01 6.71814799e-01 -1.84879944e-01 5.38372993e-03
1.20262206e+00 6.65872172e-02 -2.63768703e-01 -2.59914190e-01
1.16039491e+00 2.36278981e-01 -1.87096453e+00 -1.45198330e-02
-8.67058575e-01 -1.09486222e+00 -1.69230223e-01 -8.43775213e-01
-1.60101366e+00 7.02595413e-01 6.79798603e-01 -5.18253982e-01
1.48466551e+00 -1.64722070e-01 7.49436259e-01 -4.23177540e-01
4.44457024e-01 -6.26024723e-01 3.46037835e-01 -8.13811496e-02
1.45904338e+00 -1.18182242e+00 4.75784615e-02 -7.61825323e-01
-6.60447598e-01 1.07052553e+00 8.77996922e-01 -4.23553348e-01
4.50991750e-01 3.26280028e-01 1.29117653e-01 1.92913696e-01
-5.97856998e-01 4.34654087e-01 3.16888571e-01 9.37457561e-01
2.46329159e-01 -2.70599812e-01 -1.00548625e-01 4.49862242e-01
-5.85152172e-02 4.72569540e-02 1.62394524e-01 7.24063635e-01
1.60333849e-02 -1.33581352e+00 -7.46609449e-01 2.12548509e-01
-6.52689934e-01 5.83638698e-02 -2.76602536e-01 6.17132962e-01
5.16521811e-01 7.39905596e-01 -3.37387510e-02 -1.46597743e-01
5.59202582e-02 5.31917699e-02 1.03712237e+00 -5.43030798e-01
-2.60716289e-01 2.36186370e-01 -2.97457993e-01 -3.61563712e-01
-1.31161973e-01 -4.01593119e-01 -8.27163100e-01 -5.23937762e-01
-1.19665779e-01 -5.55481464e-02 3.57581079e-01 6.24875486e-01
5.27078688e-01 3.88058573e-01 8.65925610e-01 -9.93485749e-01
-5.39717019e-01 -9.80363786e-01 -6.82500541e-01 6.48909092e-01
3.65264893e-01 -7.48268783e-01 -1.99738741e-01 6.47750556e-01] | [12.628503799438477, -0.17896310985088348] |
71206cb6-afb3-4b19-bdbc-865a6edf492f | comparison-analysis-of-tree-based-and | 2010.14921 | null | https://arxiv.org/abs/2010.14921v1 | https://arxiv.org/pdf/2010.14921v1.pdf | Comparison Analysis of Tree Based and Ensembled Regression Algorithms for Traffic Accident Severity Prediction | Rapid increase of traffic volume on urban roads over time has changed the traffic scenario globally. It has also increased the ratio of road accidents that can be severe and fatal in the worst case. To improve traffic safety and its management on urban roads, there is a need for prediction of severity level of accidents. Various machine learning models are being used for accident prediction. In this study, tree based ensemble models (Random Forest, AdaBoost, Extra Tree, and Gradient Boosting) and ensemble of two statistical models (Logistic Regression Stochastic Gradient Descent) as voting classifiers are compared for prediction of road accident severity. Significant features that are strongly correlated with the accident severity are identified by Random Forest. Analysis proved Random Forest as the best performing model with highest classification results with 0.974 accuracy, 0.954 precision, 0.930 recall and 0.942 F-score using 20 most significant features as compared to other techniques classification of road accidents severity. | ['Hamza Ahmad Madni', 'Najia Saher', 'Saleem Ullah', 'Abid Ishaq', 'Saima Sadiq', 'Muhammad Umer'] | 2020-10-27 | null | null | null | null | ['severity-prediction'] | ['computer-vision'] | [-2.29819536e-01 -2.67049432e-01 -2.60369450e-01 -3.46766323e-01
-2.70220041e-01 -1.11636661e-01 5.35485864e-01 2.28753150e-01
-6.18733823e-01 1.19105136e+00 3.74871224e-01 -9.66174304e-01
-4.75162894e-01 -1.13137686e+00 -1.31971896e-01 -6.93600237e-01
2.59588748e-01 3.28476936e-01 5.85380852e-01 -4.68075514e-01
5.44034481e-01 7.94194877e-01 -1.92269003e+00 3.51362944e-01
1.05002511e+00 7.23357975e-01 3.06456894e-01 7.02524662e-01
-3.38091590e-02 7.91009486e-01 -2.06555188e-01 -3.26440543e-01
1.90390363e-01 -2.72420552e-02 -6.45395339e-01 -5.52602232e-01
-1.56780258e-01 4.60285246e-02 -5.65814488e-02 4.08194691e-01
4.95127290e-01 3.02565068e-01 1.21810782e+00 -1.23887670e+00
1.74041510e-01 -1.39601773e-03 -5.61445534e-01 5.04873335e-01
1.21121429e-01 5.18825166e-02 2.71435887e-01 -8.48376453e-01
1.63618043e-01 1.12438738e+00 6.38819456e-01 1.42786518e-01
-8.93254936e-01 -1.06245673e+00 -3.73931646e-01 9.70141172e-01
-1.06239939e+00 -1.26237899e-01 2.25417137e-01 -7.30488479e-01
1.17246759e+00 4.45060879e-01 6.57039642e-01 3.84965897e-01
6.34128690e-01 -2.32821214e-03 1.66152370e+00 -4.22234684e-01
-4.95986342e-02 3.30474585e-01 4.62874651e-01 4.87196535e-01
5.38644373e-01 5.10317624e-01 -2.30683237e-02 6.38931245e-03
1.61860157e-02 2.34977946e-01 4.83301967e-01 1.50860697e-01
-7.03816593e-01 8.38631332e-01 5.37540853e-01 3.24160427e-01
-8.55879128e-01 -1.94292709e-01 5.81376076e-01 2.12254107e-01
1.47166893e-01 -2.06653595e-01 -7.20031917e-01 -2.66559571e-01
-4.65205193e-01 3.21649194e-01 3.70605916e-01 1.76766887e-01
6.83552921e-01 -6.29130527e-02 -1.78796813e-01 9.72665966e-01
1.89647987e-01 7.41957426e-01 2.64103889e-01 -5.41210949e-01
5.34875691e-01 7.56742954e-01 -1.02550395e-01 -9.49084222e-01
-7.21104860e-01 -3.39336157e-01 -9.34709132e-01 7.01874316e-01
3.58902842e-01 -3.20227325e-01 -9.52961445e-01 1.07202971e+00
3.00917327e-01 -1.30177930e-01 -6.03401773e-02 2.31887162e-01
5.65562785e-01 6.80176437e-01 7.76280463e-01 -1.54042259e-01
1.19677019e+00 -5.49045801e-01 -4.79502231e-01 5.85783832e-02
7.76804924e-01 -8.70049536e-01 5.17260969e-01 3.09446305e-01
-5.17633200e-01 -7.06434488e-01 -3.88406515e-01 5.56404412e-01
-7.57798970e-01 -3.12591456e-02 2.89881229e-01 9.44407165e-01
-6.10410333e-01 2.48406187e-01 -2.47964963e-01 -5.38841367e-01
4.15329784e-01 5.44781983e-01 -5.90104580e-01 -1.74881294e-01
-1.00327539e+00 1.68972397e+00 2.29866877e-01 -1.52132198e-01
-4.45165522e-02 -4.23446774e-01 -3.74369383e-01 -4.49356705e-01
-9.28779468e-02 -6.71740949e-01 8.00424516e-01 -5.02080619e-01
-8.04335892e-01 5.95187604e-01 -6.24820650e-01 -3.96918625e-01
4.95862156e-01 -8.33498836e-02 -7.88365185e-01 -5.17616451e-01
3.19882512e-01 3.21715921e-01 1.99108765e-01 -1.06462252e+00
-1.42674387e+00 -6.84077799e-01 -4.31115240e-01 2.76169684e-02
3.17327291e-01 3.90184373e-01 6.65773988e-01 -6.70176074e-02
-1.14423275e-01 -1.01607430e+00 -5.71193874e-01 -8.04950058e-01
-1.62318632e-01 -6.72525465e-01 1.02757835e+00 -8.81269038e-01
1.45328093e+00 -1.40636408e+00 -6.61612332e-01 5.90580523e-01
-1.11264534e-01 6.85416281e-01 5.83983243e-01 6.03955686e-01
-3.52647394e-01 7.79898539e-02 9.18249488e-02 5.53504586e-01
-5.67936182e-01 3.51465315e-01 1.61564037e-01 -5.14649749e-02
-6.83036596e-02 5.76338589e-01 -6.72468364e-01 -6.54863358e-01
9.10329759e-01 4.11512166e-01 -7.78188184e-02 1.13864519e-01
8.57075989e-01 4.83841211e-01 -4.02576685e-01 2.28418961e-01
6.22903943e-01 6.37625992e-01 -2.99518794e-01 -8.04467276e-02
-7.14423895e-01 2.58914500e-01 -9.74792302e-01 3.25825244e-01
-8.92211199e-01 6.94725037e-01 -5.57744265e-01 -1.22095478e+00
1.22096574e+00 5.04722536e-01 2.53004342e-01 -6.85891151e-01
8.03781524e-02 3.49236876e-01 2.30035290e-01 -1.16270924e+00
-1.02109695e-02 -1.55657515e-01 2.52844065e-01 -3.43078263e-02
-5.61248183e-01 2.59575725e-01 2.20015183e-01 -1.42501518e-01
9.82709706e-01 -2.90290564e-01 6.45426989e-01 -2.18313098e-01
9.29746389e-01 1.94599345e-01 1.84139058e-01 5.15588760e-01
-3.01279753e-01 3.35159637e-02 2.78142124e-01 -6.29399717e-01
-8.71684730e-01 -9.22054291e-01 -4.39399868e-01 9.91653740e-01
-2.56289631e-01 2.48758420e-01 -4.09295827e-01 -5.49997151e-01
-5.49882231e-03 1.10737789e+00 -3.99770468e-01 -1.75307214e-01
-6.58579648e-01 -8.26466620e-01 6.48385212e-02 4.69843656e-01
6.69510067e-01 -1.04702222e+00 -4.62072283e-01 1.78102970e-01
-3.70941758e-02 -6.45428598e-01 5.36721766e-01 4.24910374e-02
-8.58543694e-01 -1.17954433e+00 -3.71302754e-01 -6.52556539e-01
3.20293099e-01 5.62833011e-01 7.96879947e-01 9.79354754e-02
-3.55429053e-01 -1.77326128e-01 -2.51765579e-01 -8.82496357e-01
-5.46167970e-01 1.02004133e-01 -1.66894034e-01 -2.43703276e-02
5.33162236e-01 -7.69922733e-01 -6.08867764e-01 3.61956120e-01
-7.96977803e-02 -6.58673048e-02 7.76506424e-01 3.99032265e-01
-4.38919105e-02 3.98391157e-01 9.71890867e-01 -8.09176624e-01
5.41975915e-01 -1.06658578e+00 -8.54625776e-02 -7.98120052e-02
-8.85245740e-01 -2.48196885e-01 3.06126624e-01 2.04201147e-01
-1.17403662e+00 -4.51076515e-02 -4.76597041e-01 5.33844292e-01
-8.29346716e-01 2.05348566e-01 1.37768790e-01 1.64345339e-01
7.91750371e-01 -1.06984586e-01 -5.74466214e-03 -4.22032684e-01
-1.73981011e-01 1.04274559e+00 1.62069112e-01 -5.20919971e-02
6.33594453e-01 1.01832628e-01 5.90345800e-01 -9.80558097e-01
-6.21825039e-01 -8.77425015e-01 -8.50783288e-01 -9.38917816e-01
1.30272365e+00 -5.52025437e-01 -7.03160465e-01 4.28774208e-01
-9.84176636e-01 1.47864446e-01 4.56370682e-01 8.27508926e-01
-2.38173008e-01 -9.46766958e-02 3.02769780e-01 -1.42608798e+00
-3.09712559e-01 -9.30866182e-01 2.27097824e-01 1.35277212e-01
-5.82478285e-01 -7.05618441e-01 8.43320414e-02 6.17949963e-01
6.87790930e-01 5.65741599e-01 1.12270927e+00 -3.83973032e-01
-1.86095890e-02 -6.10825837e-01 -3.65439802e-01 3.17260951e-01
2.64286101e-01 3.81852239e-01 -7.23052323e-01 6.05289757e-01
-6.08334541e-01 1.96492806e-01 8.95670474e-01 7.74267197e-01
8.82830679e-01 -1.15601562e-01 -7.53341973e-01 -1.36905998e-01
1.36908066e+00 7.36246407e-01 8.11686575e-01 6.28280342e-01
4.34036165e-01 1.04432857e+00 7.65920460e-01 -5.00850566e-02
5.23524225e-01 6.90630138e-01 2.39787862e-01 -2.71612909e-02
-2.63557643e-01 -1.91690803e-01 1.19354896e-01 2.44401738e-01
-7.66233981e-01 1.57505155e-01 -1.36096907e+00 4.48120922e-01
-1.85790956e+00 -1.59792149e+00 -1.28200102e+00 2.36454749e+00
1.64766744e-01 3.44177336e-01 8.07704449e-01 1.00346744e+00
1.00109136e+00 -4.06767935e-01 1.23474039e-01 -1.07633507e+00
2.53182143e-01 3.47079754e-01 9.35490429e-01 4.99146074e-01
-9.75136518e-01 4.94201064e-01 5.86345148e+00 6.44167125e-01
-9.97642815e-01 1.21348880e-01 7.08813012e-01 3.74785662e-01
2.00207934e-01 9.45928097e-02 -7.61436939e-01 7.29829490e-01
1.18668830e+00 -7.05418289e-02 -8.42778683e-02 6.58837378e-01
8.30554903e-01 -6.14919722e-01 4.89065461e-02 5.65215468e-01
-6.53448880e-01 -9.79720533e-01 -1.36207640e-01 1.87126637e-01
7.11838543e-01 1.48202822e-01 -1.74710110e-01 3.31418663e-01
5.72736800e-01 -1.08783615e+00 2.05300927e-01 8.11334789e-01
4.74256903e-01 -1.17496073e+00 1.04544032e+00 5.64176083e-01
-9.92102921e-01 -4.47656482e-01 3.26419957e-02 -5.49717247e-01
4.36578304e-01 7.44011819e-01 -1.06189334e+00 4.14176375e-01
9.96006250e-01 2.73706943e-01 -5.59084475e-01 1.22225726e+00
-2.45582417e-01 1.03169143e+00 -2.89136171e-01 -2.74055094e-01
1.58842906e-01 -9.51141939e-02 2.41479039e-01 1.22851813e+00
2.50586182e-01 6.50100261e-02 -1.89125553e-01 -1.99298546e-01
7.60657072e-01 4.22861248e-01 -1.07361615e+00 1.14193296e+00
5.78010023e-01 9.51579869e-01 -4.33784217e-01 -3.01493973e-01
-5.24279833e-01 6.20456599e-02 7.54453987e-02 -5.27802221e-02
-6.91554725e-01 -4.11400080e-01 4.15192842e-01 6.65312111e-01
-1.72721401e-01 -1.22047365e-01 -8.35240185e-01 -1.00522406e-01
-3.36257249e-01 -1.45847216e-01 3.58369708e-01 -7.47657478e-01
-9.86904204e-01 4.92549330e-01 4.11589235e-01 -9.04715538e-01
-1.64888397e-01 -4.78775293e-01 -1.13541281e+00 1.09814990e+00
-1.13373435e+00 -9.93595839e-01 -2.60702878e-01 4.64983195e-01
5.93211114e-01 -3.52035433e-01 5.97224653e-01 2.81106323e-01
-7.04410255e-01 1.49423704e-01 -5.08357175e-02 -1.96126223e-01
3.59019190e-01 -9.22193825e-01 -4.66724224e-02 3.60229969e-01
-7.18706310e-01 -2.53099017e-03 6.69702053e-01 -7.43765056e-01
-3.35464329e-01 -1.15078962e+00 1.59117830e+00 -3.77354175e-01
4.13356394e-01 4.84227329e-01 -4.53466892e-01 1.27497748e-01
-7.90219605e-02 -4.97766137e-01 6.91672385e-01 2.22699285e-01
7.69347697e-03 -4.73299593e-01 -1.35049093e+00 4.37866658e-01
6.69314206e-01 -2.32022200e-02 -4.16388482e-01 1.87585160e-01
-1.21051736e-01 5.12745082e-01 -7.38065541e-01 6.06940925e-01
1.02545440e+00 -1.27909851e+00 9.32105541e-01 -7.41824389e-01
2.38115355e-01 -2.33695507e-01 3.84894311e-02 -9.66739476e-01
-6.21232688e-01 9.88574326e-03 5.56710064e-01 1.05418873e+00
9.91673470e-01 -8.88443232e-01 6.27277076e-01 7.96379268e-01
-1.48365632e-01 -1.06208944e+00 -9.70694840e-01 -5.38687766e-01
2.49731973e-01 -6.88343346e-01 3.91790122e-01 4.73055214e-01
-2.89359123e-01 5.72886765e-01 -2.50542581e-01 -2.47572780e-01
4.40426230e-01 -4.68575478e-01 9.70278919e-01 -1.49040413e+00
5.37523687e-01 -5.19976199e-01 -8.53295386e-01 1.67535339e-02
3.61730717e-02 -7.42888629e-01 -5.54894745e-01 -1.82596111e+00
1.59331664e-01 -7.68876553e-01 -2.04077676e-01 3.39826345e-01
-4.66996104e-01 4.22514640e-02 -1.38315499e-01 8.90916884e-02
3.54689240e-01 -1.63252637e-01 9.00754511e-01 3.40250254e-01
-1.60028160e-01 9.34932709e-01 -4.15995240e-01 6.62087381e-01
1.54063034e+00 -6.17201030e-01 -3.50014120e-01 1.17671095e-01
9.11130682e-02 1.73048630e-01 3.43568414e-01 -1.24084842e+00
2.69361343e-02 -5.28831542e-01 3.29699129e-01 -9.22981679e-01
-2.27709115e-03 -1.01722276e+00 4.71474260e-01 9.37151611e-01
-2.30358645e-01 3.96580130e-01 -6.28166720e-02 4.18325692e-01
2.27986332e-02 -2.27857530e-01 9.12837982e-01 2.58837074e-01
-7.34749496e-01 2.28819996e-02 -1.06146014e+00 -6.68547571e-01
1.50057280e+00 -6.77546501e-01 2.65228632e-03 -2.89621860e-01
-8.25634897e-01 8.67346823e-02 -6.75384924e-02 5.14071643e-01
1.83041751e-01 -1.12393963e+00 -8.86323512e-01 -2.77151465e-02
3.15958820e-02 -5.54428577e-01 3.31941992e-01 1.12834084e+00
-5.65946758e-01 5.81921339e-01 -6.08082116e-01 -4.86122519e-02
-1.74588776e+00 4.04525846e-01 2.16511726e-01 -3.94571394e-01
-3.01253527e-01 4.79175061e-01 -3.39235634e-01 -2.35279366e-01
-2.42037669e-01 2.20506474e-01 -9.35451329e-01 -6.22831285e-02
2.23947242e-01 1.60092318e+00 2.16931283e-01 -1.12737739e+00
-4.62577671e-01 9.28938925e-01 2.59948701e-01 7.35445693e-02
1.06587422e+00 2.49806270e-02 -2.17993915e-01 5.52310228e-01
1.12071240e+00 -8.63773003e-02 -3.32282960e-01 2.75590628e-01
1.55338153e-01 -3.32204193e-01 2.17525020e-01 -1.18018234e+00
-7.04955518e-01 8.23543489e-01 9.11121666e-01 3.68299127e-01
1.10936224e+00 -3.25130820e-01 8.27511549e-01 2.79322505e-01
2.62474626e-01 -1.01920068e+00 -1.04708660e+00 3.18695873e-01
6.77386105e-01 -1.48694861e+00 -8.06229115e-02 -5.51571548e-01
-6.75275743e-01 1.30847311e+00 4.31539208e-01 -1.63635805e-01
1.37251794e+00 -2.69253626e-02 -3.65936309e-02 -1.36270383e-02
-8.57561588e-01 -6.16024673e-01 1.89138427e-01 9.76745605e-01
6.41024411e-01 5.98390937e-01 -1.11878383e+00 6.00927584e-02
-4.66264248e-01 1.64182201e-01 -7.46486187e-02 6.75862849e-01
-1.05523407e+00 -1.15358353e+00 -4.53952551e-01 1.11553609e+00
-5.73869467e-01 2.23772988e-01 -1.68949172e-01 8.42996776e-01
5.99422157e-01 1.61255944e+00 -2.57125478e-02 -6.36844099e-01
6.50527716e-01 1.52236775e-01 -6.13923231e-03 -1.11670651e-01
-7.29707479e-01 -8.03877831e-01 6.65411770e-01 -1.46801740e-01
-3.53911161e-01 -8.34769905e-01 -1.14880502e+00 -7.16164649e-01
-3.59557122e-01 3.85166407e-01 1.18131757e+00 1.02817070e+00
1.31750271e-01 2.74684757e-01 1.00694215e+00 -5.06115973e-01
1.52004704e-01 -1.08164155e+00 -2.08746850e-01 1.48955598e-01
1.42296687e-01 -1.00106370e+00 -2.40542784e-01 -8.07325095e-02] | [6.830014228820801, 2.0767405033111572] |
dc600694-e99d-4545-9fde-cf04eac23fff | bert-based-distractor-generation-for-swedish | 2108.03973 | null | https://arxiv.org/abs/2108.03973v1 | https://arxiv.org/pdf/2108.03973v1.pdf | BERT-based distractor generation for Swedish reading comprehension questions using a small-scale dataset | An important part when constructing multiple-choice questions (MCQs) for reading comprehension assessment are the distractors, the incorrect but preferably plausible answer options. In this paper, we present a new BERT-based method for automatically generating distractors using only a small-scale dataset. We also release a new such dataset of Swedish MCQs (used for training the model), and propose a methodology for assessing the generated distractors. Evaluation shows that from a student's perspective, our method generated one or more plausible distractors for more than 50% of the MCQs in our test set. From a teacher's perspective, about 50% of the generated distractors were deemed appropriate. We also do a thorough analysis of the results. | ['Johan Boye', 'Dmytro Kalpakchi'] | 2021-08-09 | null | https://aclanthology.org/2021.inlg-1.43 | https://aclanthology.org/2021.inlg-1.43.pdf | inlg-acl-2021-8 | ['distractor-generation'] | ['natural-language-processing'] | [-1.37978822e-01 6.33911431e-01 2.12075785e-01 -4.86209422e-01
-1.25066674e+00 -9.50662732e-01 2.98546731e-01 4.62717861e-01
-4.64184344e-01 1.03609824e+00 4.03815925e-01 -8.17905486e-01
-2.89154142e-01 -4.65338171e-01 -4.16772127e-01 -1.17492154e-01
8.10222447e-01 6.44289374e-01 6.74246550e-01 -4.12007600e-01
7.49673128e-01 2.54642755e-01 -2.07231712e+00 3.72095913e-01
1.75107503e+00 3.64059150e-01 3.86649817e-01 8.30493212e-01
-6.13847971e-01 1.28430951e+00 -1.18671346e+00 -6.48740530e-01
-2.67564088e-01 -8.26129675e-01 -1.46663570e+00 -2.53800601e-01
9.09486473e-01 -3.30856919e-01 3.17898810e-01 8.29117358e-01
5.96751809e-01 4.12090689e-01 9.86896098e-01 -8.91349614e-01
-8.71663868e-01 8.75822127e-01 3.08034509e-01 4.31390762e-01
1.19321132e+00 1.84862971e-01 8.90264094e-01 -1.04387021e+00
4.19045955e-01 1.21046972e+00 8.24459866e-02 7.41627514e-01
-9.31091487e-01 -3.84242058e-01 -1.08755223e-01 6.27352238e-01
-1.20600998e+00 -3.13206196e-01 4.86558378e-01 -4.97128904e-01
6.99610889e-01 5.84330022e-01 4.47883874e-01 1.13945246e+00
1.22756779e-01 7.90685594e-01 1.66906583e+00 -9.69745278e-01
2.37850368e-01 5.00702024e-01 5.92103481e-01 -4.00709510e-02
2.59738952e-01 -1.01567827e-01 -2.35964805e-01 8.74950550e-03
4.40417707e-01 -8.44165742e-01 -3.66667330e-01 1.99464470e-01
-8.06232452e-01 9.14620936e-01 -1.40605256e-01 4.13213849e-01
-3.64281312e-02 -6.11929059e-01 -3.61107498e-01 6.04605138e-01
6.84352592e-02 9.24170256e-01 -5.58945060e-01 -5.25927126e-01
-6.63672090e-01 3.83068174e-01 1.07528913e+00 1.07061732e+00
2.64269501e-01 -5.96145913e-02 -7.30225623e-01 1.04881835e+00
1.87348008e-01 5.93768835e-01 7.85021067e-01 -9.22866702e-01
4.28571969e-01 6.39805555e-01 5.70883334e-01 -8.22372377e-01
-1.83649004e-01 -3.49399835e-01 -5.14194369e-02 1.92291245e-01
8.68884802e-01 8.66356716e-02 -6.48399711e-01 1.51967216e+00
2.23412991e-01 -5.17498016e-01 -1.58498168e-01 6.62580788e-01
1.15192330e+00 2.63733596e-01 2.05985665e-01 -3.19878578e-01
1.43040061e+00 -8.07696223e-01 -1.06469357e+00 9.83629492e-04
8.31674039e-01 -1.09038186e+00 1.87256968e+00 8.48452151e-01
-1.52266777e+00 -7.15792358e-01 -7.35367537e-01 -1.24936402e-01
-1.84197977e-01 -8.26970488e-03 -1.58926070e-01 6.78941131e-01
-8.79912913e-01 1.90175712e-01 2.31844038e-01 -2.10863445e-02
-1.73222244e-01 -1.20367393e-01 1.28173336e-01 -1.54100031e-01
-1.42204928e+00 1.50872743e+00 1.41691983e-01 -6.11058772e-01
-6.95104897e-01 -5.35024643e-01 -7.34898984e-01 2.76298225e-01
4.13814932e-01 -3.74171823e-01 1.79764915e+00 -9.55368340e-01
-1.64363027e+00 8.00468504e-01 -1.28819153e-01 2.05260411e-01
6.11640871e-01 -3.23751986e-01 -5.03094673e-01 2.13596880e-01
3.06259811e-01 4.00346279e-01 6.22906804e-01 -9.96432543e-01
-6.39088750e-01 1.01774558e-01 2.13873968e-01 3.58575106e-01
-1.00574614e-02 2.36169547e-01 2.82850415e-01 -7.50211000e-01
9.31237116e-02 -5.66377223e-01 1.10717431e-01 -5.23907185e-01
-1.18423961e-01 -1.03878474e+00 -1.55027598e-01 -5.07100761e-01
1.78155947e+00 -1.63402760e+00 -9.11628678e-02 1.25696972e-01
2.93543220e-01 4.73391593e-01 -2.65730113e-01 5.20352840e-01
-4.03590560e-01 4.40058261e-01 4.41266447e-02 2.30341971e-01
1.78154916e-01 -1.14419922e-01 -1.88237953e-03 -2.20093176e-01
1.20277792e-01 5.74306607e-01 -1.08729231e+00 -7.24793911e-01
1.67188048e-01 -2.65752792e-01 -3.15251172e-01 7.43153334e-01
-3.35312277e-01 2.01065630e-01 -1.22407377e-01 4.14274603e-01
7.84088969e-01 1.25716299e-01 -3.58547837e-01 5.34122109e-01
-2.84557473e-02 1.00034344e+00 -1.29529119e+00 1.19401479e+00
-4.28515881e-01 4.58279282e-01 -6.00420654e-01 -2.23184496e-01
8.66405487e-01 2.20878035e-01 -3.86520803e-01 -8.86299372e-01
6.08516298e-02 5.17906368e-01 5.43698788e-01 -1.07195413e+00
5.34367085e-01 -2.25280970e-01 -3.97221111e-02 7.03788102e-01
1.41155601e-01 -5.23039937e-01 5.68245530e-01 3.17229301e-01
1.02976644e+00 -7.68561149e-03 6.36955559e-01 -5.34828007e-01
9.59805608e-01 2.20185667e-02 -3.34914476e-02 9.20086741e-01
-3.50879073e-01 6.83223128e-01 4.41171527e-01 -8.23519528e-02
-6.35578275e-01 -1.13009536e+00 -1.32774085e-01 1.22579396e+00
-4.51226294e-01 -4.29308176e-01 -1.00585794e+00 -9.12044883e-01
-5.44927180e-01 1.74684620e+00 -2.82311678e-01 -5.83847798e-02
-4.05505389e-01 4.02727611e-02 5.34782290e-01 3.04492056e-01
4.61850539e-02 -1.15048981e+00 -5.10621130e-01 2.79245526e-01
-7.75824130e-01 -7.03401506e-01 -3.72086197e-01 -3.41615416e-02
-5.98135829e-01 -1.22800052e+00 -3.70423317e-01 -7.61341035e-01
5.75656950e-01 2.72582233e-01 1.85890758e+00 5.83746254e-01
2.77527630e-01 2.58547008e-01 -7.91548729e-01 -6.63461149e-01
-8.58507693e-01 -1.00077078e-01 -2.19554007e-01 -8.34377825e-01
9.44047749e-01 -2.08649024e-01 -1.67908460e-01 3.84811491e-01
-9.34048831e-01 -1.99257255e-01 3.74514490e-01 5.40627062e-01
1.59787789e-01 -2.12888628e-01 8.18653822e-01 -9.04507816e-01
1.47658610e+00 -6.08263850e-01 -3.93184394e-01 3.27596128e-01
-5.34829676e-01 1.65413037e-01 7.69110382e-01 -6.55486763e-01
-9.62057531e-01 -4.56503451e-01 -6.21925771e-01 -9.67948139e-03
-7.90366054e-01 4.61742640e-01 -2.53920496e-01 2.76598185e-02
1.31169629e+00 3.42932865e-02 -1.53229982e-01 -4.33874607e-01
2.77186543e-01 7.70203233e-01 1.41470715e-01 -8.74430835e-01
6.73864782e-01 -8.50404561e-01 -3.26508731e-01 -6.37441635e-01
-1.05447280e+00 -1.96499079e-01 -6.01306558e-01 -4.97584134e-01
3.73790652e-01 -6.39279664e-01 -6.99716449e-01 1.82785213e-01
-1.01106393e+00 -2.05993444e-01 -3.29833984e-01 6.01197541e-01
-4.38681185e-01 2.65915126e-01 -2.28968263e-01 -9.74534810e-01
1.41120315e-01 -1.21097124e+00 2.91365385e-01 5.39165139e-01
-9.98538971e-01 -9.72135007e-01 6.26984537e-02 7.82144666e-01
4.70693588e-01 -2.12069035e-01 1.42726314e+00 -1.25682282e+00
-2.17037871e-01 5.06785586e-02 2.49937326e-01 4.30773497e-01
-7.53823966e-02 2.23257795e-01 -8.07816803e-01 1.39413893e-01
3.42496157e-01 -7.00633824e-01 1.81485102e-01 1.25984073e-01
1.24548388e+00 -4.70587403e-01 4.47206318e-01 -1.68007076e-01
1.26957011e+00 7.78627619e-02 8.21541905e-01 1.91484749e-01
3.91406387e-01 9.50592756e-01 1.00581706e+00 1.49919644e-01
8.77274692e-01 2.85627514e-01 3.34634572e-01 6.87111616e-01
-3.10266137e-01 -4.96703595e-01 5.36143303e-01 1.31317031e+00
3.31297040e-01 -4.90688443e-01 -1.03470325e+00 8.04868698e-01
-1.06620824e+00 -9.11396384e-01 -8.93220544e-01 2.31010628e+00
1.17906010e+00 1.58966914e-01 2.14495659e-01 4.61444587e-01
4.68832701e-01 -4.45910245e-01 -4.23433371e-02 -9.83314276e-01
-5.40118255e-02 9.05183315e-01 -2.82845318e-01 1.07652235e+00
-1.91004083e-01 8.82572055e-01 7.34063005e+00 1.09222269e+00
-4.64303762e-01 -1.88720539e-01 4.38729525e-01 7.67308474e-02
-1.03682470e+00 -7.68686756e-02 -7.53210187e-01 5.71253955e-01
1.36656570e+00 -2.09779412e-01 2.25244179e-01 6.43042028e-01
3.44632357e-01 -8.93150151e-01 -1.00879848e+00 5.68933189e-01
2.41604716e-01 -5.77432394e-01 2.60160804e-01 -3.59083146e-01
5.24625421e-01 -7.96676219e-01 -8.01735446e-02 5.40603876e-01
5.72776973e-01 -1.33597934e+00 8.30486357e-01 6.44549489e-01
5.87285101e-01 -9.89679337e-01 7.29416609e-01 9.04476523e-01
-2.81509429e-01 6.22788561e-04 -4.71253872e-01 -6.31837368e-01
-4.86279726e-01 5.15618980e-01 -9.33671772e-01 2.74227262e-01
5.23480117e-01 -1.95997715e-01 -1.40690136e+00 9.11669135e-01
-9.96288538e-01 1.04763699e+00 -1.64856255e-01 -6.54823244e-01
-1.32504981e-02 -1.09410793e-01 3.51563573e-01 8.19832087e-01
5.24352491e-01 4.78469849e-01 -3.31152737e-01 8.35597157e-01
2.26592094e-01 5.50856709e-01 -3.81116331e-01 3.08442354e-01
9.53656912e-01 9.37779546e-01 -1.66890174e-01 -4.15895820e-01
-2.20139369e-01 5.74059188e-01 6.47678316e-01 7.66417831e-02
-6.76783085e-01 -5.03179669e-01 2.06062779e-01 2.50097185e-01
-2.61996955e-01 3.33246112e-01 -5.63801885e-01 -1.05488372e+00
8.72821435e-02 -1.64629960e+00 4.02871877e-01 -1.33343422e+00
-1.38771021e+00 5.19342124e-01 2.59766221e-01 -1.27122438e+00
-5.60818076e-01 -5.46032190e-01 -7.92494297e-01 1.44200039e+00
-1.15791905e+00 -2.82795191e-01 -5.35775840e-01 4.94742274e-01
7.37436175e-01 1.23467147e-01 8.00087988e-01 -3.71273980e-02
-1.59467652e-01 6.43352509e-01 -4.27197695e-01 -3.12635332e-01
1.12500906e+00 -1.74120355e+00 6.73240796e-03 8.98760796e-01
-1.63398430e-01 6.70678496e-01 1.10765958e+00 -4.23005700e-01
-6.07981563e-01 -3.51967126e-01 1.72726917e+00 -8.79405260e-01
3.86732519e-01 1.77209347e-01 -1.32129753e+00 3.14091027e-01
7.42885053e-01 -7.16985404e-01 1.18425858e+00 -9.23320353e-02
1.74859818e-02 4.89278436e-01 -1.39612651e+00 6.41486645e-01
7.47140586e-01 -2.41636753e-01 -1.59193802e+00 1.15002960e-01
4.62883294e-01 -6.54018402e-01 -7.70807505e-01 1.90482028e-02
1.84432089e-01 -1.22354770e+00 5.77291310e-01 -8.10535192e-01
8.84149849e-01 -1.92126006e-01 7.32717961e-02 -1.92045116e+00
-5.51357090e-01 -3.74347150e-01 1.06144184e-02 1.34224522e+00
4.45793718e-01 -3.13884467e-01 3.08216512e-01 8.85296702e-01
-2.35960484e-01 -6.67284846e-01 -8.08495224e-01 -4.77175474e-01
8.84188533e-01 -4.06554312e-01 6.79657757e-01 8.69204760e-01
9.32364240e-02 5.23262560e-01 1.81454420e-01 -1.97304085e-01
3.48244905e-01 -2.96392620e-01 5.69081426e-01 -1.17379892e+00
8.04301817e-03 -4.54341620e-01 2.24801213e-01 -1.05146861e+00
8.69973078e-02 -5.99439263e-01 -1.18921627e-03 -1.46824598e+00
-9.61794779e-02 -1.07777998e-01 8.82957801e-02 9.80385244e-02
-9.52000439e-01 -3.55891854e-01 1.94769308e-01 -2.52374470e-01
-2.76489824e-01 2.23397225e-01 1.65766490e+00 3.59518796e-01
6.21455461e-02 2.60469437e-01 -1.04547656e+00 7.17820346e-01
1.00694752e+00 -6.56782031e-01 -7.61213660e-01 -5.12012720e-01
4.59248990e-01 1.47004992e-01 -7.63488710e-02 -8.84776294e-01
-1.60469301e-02 -6.81486249e-01 4.59956288e-01 -6.98882878e-01
-3.03267866e-01 -6.49629176e-01 -2.15039447e-01 2.56869256e-01
-7.10633218e-01 5.82854927e-01 1.33211076e-01 -2.37649754e-01
-2.34723911e-01 -1.16797590e+00 8.07036936e-01 -3.46103430e-01
-2.00662211e-01 -4.79645073e-01 -9.71367955e-01 7.57850409e-01
9.21156228e-01 -2.31101841e-01 -4.49839026e-01 -8.04325819e-01
-6.86085045e-01 3.19557935e-01 3.74508709e-01 4.48732764e-01
7.41558909e-01 -1.33712935e+00 -9.36925709e-01 4.00312766e-02
3.72491926e-01 -4.58995141e-02 7.73276342e-03 4.95322526e-01
-6.03997886e-01 6.34480536e-01 -3.17919344e-01 -2.80042350e-01
-1.25690508e+00 3.78014982e-01 2.21143171e-01 -1.12879157e-01
1.36550277e-01 9.09077048e-01 -4.03386235e-01 -4.46684867e-01
5.02916388e-02 -6.00541770e-01 -9.93296385e-01 3.59097362e-01
7.45959997e-01 6.01967990e-01 2.20117450e-01 -5.72983384e-01
9.54145491e-02 2.73415595e-01 1.92334931e-02 -2.29550913e-01
6.19642913e-01 -3.40438515e-01 -3.87628190e-02 7.88859963e-01
4.90013838e-01 4.78541315e-01 -4.82936054e-01 2.33196542e-02
2.42437840e-01 -7.45610476e-01 -3.65709335e-01 -1.41803753e+00
-2.06496671e-01 8.46956372e-01 1.19990930e-01 4.82634068e-01
1.15552533e+00 -1.66301996e-01 1.75949961e-01 3.53485525e-01
3.54086637e-01 -1.22535014e+00 2.61868358e-01 8.22817087e-01
1.05794120e+00 -1.16693342e+00 -3.05065781e-01 -5.81949174e-01
-7.85040319e-01 1.32806289e+00 1.25318146e+00 1.77938089e-01
1.04192078e-01 -3.25614482e-01 4.68762398e-01 9.53819901e-02
-1.24544144e+00 -1.45116940e-01 4.68338728e-01 7.82707274e-01
1.07591140e+00 1.51146621e-01 -1.17294061e+00 9.99135792e-01
-8.10940862e-01 -2.67167479e-01 1.37391675e+00 7.90838838e-01
-8.79926205e-01 -1.22866774e+00 -7.30308354e-01 6.07228041e-01
-3.25652003e-01 -3.82229209e-01 -9.10786867e-01 6.03601754e-01
-4.74342220e-02 1.61492300e+00 -2.86764681e-01 -2.16274485e-01
6.40148878e-01 5.41530669e-01 7.59788036e-01 -9.35686588e-01
-1.03076339e+00 -3.97765189e-01 3.11815530e-01 -2.68170029e-01
-1.67159110e-01 -6.40585124e-01 -8.64639401e-01 -5.33764243e-01
-5.23755193e-01 4.55505073e-01 2.22456247e-01 1.21937454e+00
-7.48049691e-02 4.50973928e-01 7.10704207e-01 1.63664728e-01
-1.11049831e+00 -1.65927732e+00 -1.76253244e-01 5.17368376e-01
1.02979094e-01 -6.71841860e-01 -7.99069524e-01 -3.24163973e-01] | [11.48919677734375, 8.199422836303711] |
44f83efa-a644-4ceb-b4a1-472187adb171 | deim-an-effective-deep-encoding-and | 2203.10482 | null | https://arxiv.org/abs/2203.10482v1 | https://arxiv.org/pdf/2203.10482v1.pdf | DEIM: An effective deep encoding and interaction model for sentence matching | Natural language sentence matching is the task of comparing two sentences and identifying the relationship between them.It has a wide range of applications in natural language processing tasks such as reading comprehension, question and answer systems. The main approach is to compute the interaction between text representations and sentence pairs through an attention mechanism, which can extract the semantic information between sentence pairs well. However,this kind of method can not gain satisfactory results when dealing with complex semantic features. To solve this problem, we propose a sentence matching method based on deep encoding and interaction to extract deep semantic information. In the encoder layer,we refer to the information of another sentence in the process of encoding a single sentence, and later use a heuristic algorithm to fuse the information. In the interaction layer, we use a bidirectional attention mechanism and a self-attention mechanism to obtain deep semantic information.Finally, we perform a pooling operation and input it to the MLP for classification. we evaluate our model on three tasks: recognizing textual entailment, paraphrase recognition, and answer selection. We conducted experiments on the SNLI and SciTail datasets for the recognizing textual entailment task, the Quora dataset for the paraphrase recognition task, and the WikiQA dataset for the answer selection task. The experimental results show that the proposed algorithm can effectively extract deep semantic features that verify the effectiveness of the algorithm on sentence matching tasks. | ['Zhenguo Zhang', 'Rongyi Cui', 'Yahui Zhao', 'Kexin Jiang'] | 2022-03-20 | null | null | null | null | ['answer-selection'] | ['natural-language-processing'] | [ 5.59627831e-01 -2.47357965e-01 1.94312632e-01 -6.78698003e-01
-5.54097295e-01 -1.96569577e-01 4.54807192e-01 3.76349449e-01
-7.29352176e-01 2.82481551e-01 3.63130033e-01 -4.62536216e-01
-1.26891464e-01 -1.04222822e+00 -6.04057670e-01 -3.09206694e-01
7.32414484e-01 1.56846419e-01 3.46010774e-01 -3.18048775e-01
7.64412761e-01 -3.42793800e-02 -1.60345066e+00 8.95526469e-01
1.15870309e+00 1.24237001e+00 5.01387000e-01 4.54980612e-01
-7.95233548e-01 1.19165611e+00 -6.70588076e-01 -5.74620664e-01
-2.13857889e-01 -9.82658029e-01 -1.34915411e+00 -1.82166338e-01
-2.09917333e-02 -1.90548152e-01 -3.27085346e-01 1.18025005e+00
5.58618188e-01 1.91072106e-01 3.61155987e-01 -9.13156390e-01
-8.41903329e-01 5.21474838e-01 -2.73913771e-01 5.12743533e-01
9.02673900e-01 -6.15796782e-02 1.02260363e+00 -9.55642045e-01
2.30843872e-01 1.35983050e+00 3.78258467e-01 3.85684967e-01
-7.17088640e-01 -4.73126739e-01 -1.02547042e-01 8.82103741e-01
-9.65271235e-01 -4.87651318e-01 8.84088159e-01 -1.84446335e-01
1.10100675e+00 3.82415503e-01 3.72363925e-01 6.71501696e-01
2.75774181e-01 1.12499416e+00 9.01514053e-01 -5.17912924e-01
1.88596416e-02 1.05808243e-01 7.62815952e-01 6.90008342e-01
-3.32487166e-01 -2.91276544e-01 -3.07072401e-01 1.50833009e-02
3.64977233e-02 2.95192122e-01 -4.41748053e-01 4.98713076e-01
-1.00781953e+00 1.00820220e+00 8.44376206e-01 5.91665745e-01
-2.37423494e-01 -2.25910127e-01 6.95862055e-01 8.84064078e-01
3.60859036e-01 4.59356457e-01 -1.47906587e-01 1.33564264e-01
-7.23866880e-01 1.34789288e-01 7.45988965e-01 6.35192513e-01
8.11214566e-01 -7.30181336e-01 -6.38003290e-01 1.07759249e+00
2.22524554e-01 2.53214687e-01 1.02537572e+00 -6.50422037e-01
8.83024037e-01 1.11587453e+00 -3.00117999e-01 -1.38682342e+00
-3.32380921e-01 -9.18719545e-02 -8.67424488e-01 -3.49611193e-01
7.97349662e-02 9.64733586e-02 -4.61210519e-01 1.51486754e+00
-6.72263512e-03 -6.90946430e-02 3.91782850e-01 1.08824992e+00
1.43191767e+00 7.49623954e-01 -1.69613570e-01 -1.37078822e-01
1.64554763e+00 -1.16073120e+00 -7.91533172e-01 -4.60449219e-01
7.64312387e-01 -8.03385675e-01 1.25091612e+00 -2.62618721e-01
-1.12064016e+00 -7.95553148e-01 -1.09444439e+00 -5.88799179e-01
-4.11770642e-01 1.89043716e-01 2.06978053e-01 -6.09082468e-02
-5.03391564e-01 7.41354287e-01 -3.10170710e-01 -3.69657636e-01
4.20569032e-01 1.80020884e-01 -2.24207997e-01 -3.73417407e-01
-1.76747549e+00 1.11073649e+00 5.61037481e-01 2.87758023e-01
8.17981884e-02 -2.77705967e-01 -1.08483338e+00 6.23056352e-01
7.08625615e-02 -9.95085895e-01 1.08759761e+00 -1.36310542e+00
-1.32547009e+00 1.12702918e+00 -4.89795417e-01 -4.58463669e-01
3.77636775e-02 2.50839721e-02 -3.35311890e-01 2.40519568e-01
2.00159356e-01 2.60296911e-01 5.76795578e-01 -3.18049163e-01
-3.29446197e-01 -7.16988385e-01 4.08024430e-01 4.84220862e-01
-3.36201280e-01 4.09473836e-01 -1.13654852e-01 -3.48188579e-01
4.75449711e-01 -4.19437230e-01 4.09797430e-02 -2.56618559e-01
-3.34660679e-01 -6.02280796e-01 5.95494449e-01 -1.01333511e+00
1.28287089e+00 -2.18441463e+00 3.96402836e-01 -1.25440463e-01
-1.42340111e-02 2.46176958e-01 -2.03780055e-01 3.87547880e-01
-9.11397487e-02 -1.01437084e-01 -5.32114744e-01 -2.05309376e-01
-2.33629160e-02 -3.09111942e-02 -4.35103476e-01 -6.96844980e-03
3.67695242e-01 1.21036005e+00 -7.90870547e-01 -5.07962644e-01
-1.26043990e-01 -2.67486513e-01 -3.93029094e-01 6.93360388e-01
-1.30541012e-01 -3.47033925e-02 -4.59923834e-01 1.23588040e-01
6.41514421e-01 -2.30403990e-01 -1.68118536e-01 -2.32343644e-01
2.92209178e-01 8.52043271e-01 -8.31898689e-01 1.85819209e+00
-6.93312049e-01 6.33047879e-01 -2.00860918e-01 -1.37589693e+00
1.13534057e+00 9.63515118e-02 -2.17635229e-01 -1.21877217e+00
3.77139807e-01 4.48747724e-01 1.60495207e-01 -1.20399761e+00
3.05300057e-01 -3.77253652e-01 -2.46173859e-01 8.58872771e-01
9.53252427e-03 -6.92126453e-02 1.21525116e-01 2.21851781e-01
1.09391904e+00 -3.49692583e-01 2.51625806e-01 -1.96318358e-01
1.27834117e+00 -1.53632909e-01 1.82110444e-01 4.16444004e-01
3.25628854e-02 4.34492677e-01 5.35527468e-01 -4.41678077e-01
-5.80514669e-01 -6.38682783e-01 4.91329283e-02 8.65124583e-01
4.39734966e-01 -2.28280053e-01 -8.84403527e-01 -8.01082194e-01
-7.91642815e-02 6.34672105e-01 -4.62176502e-01 -8.72709870e-01
-5.63152015e-01 -4.63802606e-01 3.24909538e-01 5.42827606e-01
1.02768707e+00 -1.64421344e+00 -5.10341287e-01 1.41745105e-01
-6.17646992e-01 -9.39402938e-01 -5.92037201e-01 -1.19746305e-01
-5.90168357e-01 -1.26800251e+00 -3.68438393e-01 -1.24225473e+00
6.66512966e-01 3.68877023e-01 9.93050158e-01 5.86382329e-01
2.03296170e-02 -1.18866697e-01 -3.11928868e-01 -2.16865679e-03
-4.48359549e-01 1.67612866e-01 -4.72279966e-01 2.39570618e-01
7.43595600e-01 -2.68365443e-01 -6.26416624e-01 8.49049091e-02
-8.90327990e-01 2.62508184e-01 4.44197923e-01 1.15621245e+00
5.69424406e-03 -2.50731707e-01 9.36431646e-01 -7.87046313e-01
1.25120270e+00 -5.61887681e-01 -8.10240433e-02 4.77872163e-01
-1.87369362e-01 3.53159040e-01 9.13579524e-01 -1.25110624e-02
-1.16694283e+00 -3.40661079e-01 -6.40785992e-01 -4.11032066e-02
-3.85494083e-02 8.51316154e-01 -3.54793429e-01 2.62908757e-01
3.73613387e-01 5.20047963e-01 1.87161684e-01 -4.17881399e-01
6.34478107e-02 1.35196352e+00 5.07987559e-01 -3.21452469e-01
1.81526676e-01 9.48722661e-02 -4.13280994e-01 -3.41619462e-01
-1.24886441e+00 -3.97252023e-01 -4.62329566e-01 1.05502985e-01
1.00183725e+00 -3.95449221e-01 -8.69356453e-01 5.23426235e-01
-1.73473036e+00 2.86327511e-01 3.43617648e-02 3.10591310e-01
-4.39497709e-01 5.79847455e-01 -7.17489183e-01 -4.22049373e-01
-7.79928207e-01 -1.16277301e+00 8.63213718e-01 4.33721364e-01
-1.08137466e-01 -9.30380106e-01 -3.12761813e-02 6.62068546e-01
3.77022445e-01 -3.64275545e-01 1.26095688e+00 -9.91625249e-01
-2.02752709e-01 -2.35855654e-01 -5.57040572e-01 4.29621428e-01
-4.74600587e-03 -5.13263643e-01 -8.51861835e-01 -9.54199508e-02
5.84914088e-01 -6.46611154e-01 1.19878757e+00 -1.47813454e-01
1.45266771e+00 -3.95237803e-01 9.49683115e-02 3.66596729e-01
1.08003521e+00 8.11189264e-02 7.81319141e-01 2.04457641e-01
4.33057308e-01 9.35825944e-01 5.39644837e-01 -2.50419490e-02
4.41310734e-01 5.29135644e-01 1.66473880e-01 1.48905829e-01
-2.81141717e-02 -2.61860758e-01 1.57746971e-01 1.19575167e+00
5.47699392e-01 3.08942108e-04 -7.32404828e-01 2.65505135e-01
-2.15570855e+00 -1.19050634e+00 -2.20255733e-01 1.99994469e+00
9.81946170e-01 1.51743531e-01 -4.16500360e-01 3.38398695e-01
7.69445419e-01 4.66858633e-02 -5.28124511e-01 -7.46082902e-01
-8.30945559e-03 3.33193123e-01 -4.09577698e-01 5.23418069e-01
-7.87450135e-01 7.71682501e-01 4.71930838e+00 8.64622176e-01
-9.24429417e-01 7.22258314e-02 6.00299358e-01 3.07099193e-01
-4.47870582e-01 -5.41290361e-03 -3.90466571e-01 7.91910589e-01
7.60842025e-01 -4.18471128e-01 4.12140459e-01 3.55950326e-01
2.21039355e-02 -1.89263195e-01 -1.20952392e+00 9.85610306e-01
5.58184981e-01 -1.45559001e+00 2.13267177e-01 -7.59250998e-01
2.15671524e-01 -3.08456510e-01 -4.73581612e-01 5.55679202e-01
-4.72813547e-01 -1.13781714e+00 3.12516928e-01 8.08881342e-01
2.39382237e-01 -6.95901692e-01 1.14163435e+00 7.84881055e-01
-1.09372807e+00 -3.19389760e-01 -6.63695753e-01 -5.99116027e-01
-3.30905542e-02 3.20770383e-01 -2.49477103e-01 7.60299146e-01
4.12917882e-01 8.38777602e-01 -7.08596051e-01 1.03691924e+00
-5.43471456e-01 1.21390507e-01 -3.47026886e-04 -4.53277647e-01
1.28885493e-01 -2.15945110e-01 7.14084506e-02 1.04371977e+00
7.52430558e-02 8.54906291e-02 -1.11786455e-01 1.08487952e+00
-4.24739808e-01 3.29849929e-01 -4.87108082e-01 2.32032716e-01
4.84089494e-01 1.02883542e+00 -3.27317238e-01 -5.27288437e-01
-4.47083950e-01 1.45438385e+00 8.28442156e-01 1.81737483e-01
-6.64852440e-01 -1.10564792e+00 1.82228699e-01 -2.82357126e-01
-1.47439614e-01 3.62485468e-01 -4.14371550e-01 -1.32402217e+00
4.90781099e-01 -7.80957162e-01 5.29390574e-01 -9.46858585e-01
-1.40003264e+00 5.32212079e-01 -4.20831144e-01 -9.16727543e-01
-1.49979904e-01 -5.18563330e-01 -1.24735737e+00 1.26293230e+00
-1.49845862e+00 -8.21996272e-01 -4.44487363e-01 5.69369912e-01
7.79651701e-01 -5.33995740e-02 7.96000242e-01 3.79101127e-01
-6.37918591e-01 4.68078166e-01 -1.67244136e-01 3.42660695e-01
3.96196485e-01 -9.02319670e-01 3.07270557e-01 6.85269892e-01
-1.33969914e-02 6.65826797e-01 1.82341263e-01 -2.04858065e-01
-1.18147373e+00 -8.21673214e-01 1.46709466e+00 -6.60676435e-02
4.46447670e-01 -2.56634176e-01 -1.31834996e+00 3.73590082e-01
5.00060678e-01 -2.89814711e-01 6.02445543e-01 -1.14565186e-01
-1.59052610e-01 -6.54183775e-02 -1.02990878e+00 3.41718763e-01
9.06817913e-01 -9.63213027e-01 -1.35058177e+00 4.12620038e-01
1.10103369e+00 -2.49671996e-01 -5.45010984e-01 2.79073298e-01
3.33115280e-01 -8.92882705e-01 6.80674434e-01 -9.98216271e-01
1.05805957e+00 -1.19595625e-01 -2.38357950e-02 -1.24618435e+00
-1.97763413e-01 1.02745883e-01 2.11645022e-01 1.35477805e+00
3.14927757e-01 -6.06790960e-01 3.22615027e-01 5.27035713e-01
-2.16059670e-01 -1.02898240e+00 -8.80914807e-01 -4.13082331e-01
8.58001709e-02 -5.20555116e-02 8.04557264e-01 9.29079473e-01
4.97656107e-01 9.80473697e-01 2.09944248e-01 -3.13748270e-01
9.44301337e-02 9.01199877e-01 4.11407262e-01 -9.42587137e-01
-2.90525079e-01 -6.51809931e-01 -2.58779705e-01 -1.21786237e+00
8.32562029e-01 -1.44169962e+00 7.09820120e-03 -1.76677930e+00
6.81216419e-01 1.24777779e-01 -2.25825936e-01 1.77171528e-01
-7.23217189e-01 -1.66652367e-01 2.12364092e-01 1.44600585e-01
-6.35783315e-01 9.02559876e-01 1.19182301e+00 -4.79772002e-01
1.84903234e-01 1.64442345e-01 -6.86794460e-01 6.57185137e-01
8.09273005e-01 -3.06174070e-01 -2.91842818e-01 -7.95162380e-01
2.65111119e-01 2.37666205e-01 5.57455361e-01 -6.99952304e-01
5.09655595e-01 6.72635287e-02 1.86603829e-01 -4.38528329e-01
6.59626815e-03 -6.51727557e-01 -5.78402698e-01 7.17393756e-01
-7.56312728e-01 2.26511151e-01 -2.15276163e-02 1.78351909e-01
-6.36989176e-01 -9.22433019e-01 6.14849687e-01 -1.46767408e-01
-6.41205609e-01 3.69848236e-02 -1.80926800e-01 3.78847003e-01
7.45033503e-01 6.21963739e-02 -5.16282678e-01 -2.13415042e-01
-4.92398798e-01 7.12703884e-01 1.68685690e-01 6.22823536e-01
1.11162448e+00 -1.50713563e+00 -7.84922421e-01 5.10171175e-01
2.40869939e-01 -2.08911613e-01 2.08179653e-01 7.23608613e-01
-4.35526788e-01 2.98586279e-01 -1.29329592e-01 -3.19161147e-01
-1.31022954e+00 5.34670532e-01 6.74815774e-01 -2.40468502e-01
-2.22171277e-01 7.63649225e-01 5.41432835e-02 -7.00180829e-01
1.50204916e-02 -2.00545117e-01 -5.99007547e-01 3.99289317e-02
7.46002614e-01 6.44598454e-02 2.41135359e-01 -4.16765124e-01
-4.67758179e-01 5.14043152e-01 -8.81936178e-02 2.69688398e-01
9.28780913e-01 -1.51385859e-01 -6.59161210e-01 2.55813032e-01
1.69546819e+00 -4.76263463e-01 -3.16427857e-01 -4.55848575e-01
3.32217157e-01 -4.71231490e-01 -2.53387421e-01 -4.84700412e-01
-8.40056121e-01 1.37172735e+00 1.85148805e-01 2.99315572e-01
1.11787105e+00 2.08753809e-01 1.25347543e+00 6.74714267e-01
-1.61577314e-01 -8.17071915e-01 2.28229150e-01 8.46920192e-01
1.18421268e+00 -1.27103055e+00 -3.42450917e-01 -1.60639226e-01
-3.49539310e-01 1.28106868e+00 8.59313428e-01 -1.92777053e-01
3.60262841e-01 -1.69723958e-01 -2.36187011e-01 -2.48913094e-01
-7.76399791e-01 -2.59197146e-01 3.21938992e-01 -1.35081962e-01
5.37234187e-01 -3.46191585e-01 -8.34703505e-01 1.08966827e+00
-1.03783317e-01 -1.48249030e-01 1.42437607e-01 7.36720204e-01
-7.68080533e-01 -9.00192678e-01 -4.00100239e-02 6.80337787e-01
-1.68845743e-01 -3.98208380e-01 -5.97709894e-01 -6.06258139e-02
-1.35059908e-01 1.09899771e+00 2.88961112e-01 -4.53669429e-01
6.15366340e-01 2.85615176e-01 4.31475192e-01 -6.54181778e-01
-9.85295951e-01 -8.41193795e-01 1.20571963e-01 -4.52469766e-01
-2.45125100e-01 -3.84228647e-01 -1.58944273e+00 -2.58996934e-01
-3.24239314e-01 3.82664829e-01 3.52623075e-01 1.53468204e+00
3.88152868e-01 6.35701895e-01 8.75706851e-01 -2.03052074e-01
-7.91145742e-01 -9.66730773e-01 -4.42522345e-03 9.96744156e-01
1.27612635e-01 -1.64754570e-01 -4.24043119e-01 -3.08984011e-01] | [11.070534706115723, 8.27021312713623] |
e67d052b-c805-47bf-8131-6d500eedb64d | a-large-scale-evaluation-of-neural-machine | null | null | https://aclanthology.org/2021.eacl-main.303 | https://aclanthology.org/2021.eacl-main.303.pdf | A Large-scale Evaluation of Neural Machine Transliteration for Indic Languages | We take up the task of large-scale evaluation of neural machine transliteration between English and Indic languages, with a focus on multilingual transliteration to utilize orthographic similarity between Indian languages. We create a corpus of 600K word pairs mined from parallel translation corpora and monolingual corpora, which is the largest transliteration corpora for Indian languages mined from public sources. We perform a detailed analysis of multilingual transliteration and propose an improved multilingual training recipe for Indic languages. We analyze various factors affecting transliteration quality like language family, transliteration direction and word origin. | ['Rahul Kejriwal', 'Siddharth Jain', 'Anoop Kunchukuttan'] | 2021-04-01 | null | null | null | eacl-2021-2 | ['transliteration'] | ['natural-language-processing'] | [-1.89870466e-02 -5.30849338e-01 -6.76990867e-01 -4.58029211e-01
-9.24050033e-01 -1.06982660e+00 3.86799335e-01 -2.37329841e-01
-7.60193229e-01 9.81229246e-01 4.69022989e-01 -1.30649662e+00
2.87828475e-01 -5.13893843e-01 -8.53362322e-01 1.67907663e-02
6.25408232e-01 8.04662168e-01 -6.71169698e-01 -6.13426149e-01
1.62377864e-01 2.02045172e-01 -4.10322696e-01 2.23699436e-01
1.69846284e+00 -6.86214417e-02 4.23893869e-01 5.51935613e-01
-2.38420233e-01 2.60011479e-02 -4.75033104e-01 -1.03021872e+00
3.66774917e-01 -9.02565539e-01 -9.89317358e-01 -2.02569261e-01
6.40818477e-01 -1.59631193e-01 -9.68055502e-02 8.73039901e-01
6.80086076e-01 -4.84220713e-01 5.61485827e-01 -4.15935189e-01
-1.60807788e+00 1.29137194e+00 -2.17186674e-01 7.51365781e-01
7.90333897e-02 1.29240468e-01 1.32905102e+00 -1.31953442e+00
9.86302495e-01 1.19800019e+00 5.45432031e-01 4.12849009e-01
-1.09249163e+00 -6.74509883e-01 -1.96197718e-01 -6.64880499e-02
-1.33435118e+00 -4.51253474e-01 3.16982210e-01 -2.11582258e-01
1.43975377e+00 1.12944655e-02 4.06008273e-01 1.13420546e+00
5.09367704e-01 7.02359617e-01 1.46786594e+00 -9.15294290e-01
-3.84908617e-01 2.71696955e-01 1.99063808e-01 7.18484104e-01
1.73059791e-01 1.51619181e-01 -5.03639042e-01 1.02483958e-01
7.83558726e-01 -5.66512585e-01 8.36969316e-02 5.38121462e-01
-1.63691986e+00 9.13913250e-01 7.16858953e-02 5.25462329e-01
-2.12133437e-01 -2.79593974e-01 4.62563962e-01 8.66223276e-01
4.59238917e-01 7.48861074e-01 -7.39758551e-01 -1.94007903e-01
-5.20174682e-01 -2.86941111e-01 6.18784189e-01 1.04179978e+00
8.83146226e-01 3.18836868e-01 1.03357814e-01 1.29551387e+00
-4.37850878e-02 1.17482483e+00 1.04855263e+00 -3.32658380e-01
9.20359492e-01 1.84743464e-01 -4.67788845e-01 -4.67068732e-01
-4.15997989e-02 -2.76849985e-01 -4.19201255e-01 -6.65441990e-01
5.35353184e-01 -4.96275157e-01 -6.31219089e-01 1.65901446e+00
3.62574533e-02 -9.07171249e-01 3.52348626e-01 5.61405361e-01
5.47235548e-01 7.75267780e-01 -1.48656130e-01 3.95488478e-02
1.20902610e+00 -1.22242951e+00 -5.96233487e-01 -3.75269920e-01
1.05265665e+00 -1.38081276e+00 1.51546264e+00 -2.20194124e-02
-1.27122903e+00 -7.21545875e-01 -7.89985478e-01 -2.88507968e-01
-3.80509734e-01 7.03314722e-01 9.13028657e-01 8.13962638e-01
-9.72082257e-01 2.78919995e-01 -8.38573933e-01 -6.12500727e-01
-1.28783882e-01 1.92430973e-01 -5.32417834e-01 4.99581434e-02
-1.33544302e+00 1.38260555e+00 4.01999444e-01 5.05220657e-03
-5.05840421e-01 -3.73770267e-01 -7.92992651e-01 -4.49415177e-01
-3.40732664e-01 -4.68239009e-01 1.07582569e+00 -1.61902535e+00
-1.63091898e+00 1.27588725e+00 -3.62370193e-01 -5.85716330e-02
7.55883306e-02 -5.25998920e-02 -8.22048068e-01 -3.77912492e-01
1.28225833e-01 5.38587749e-01 1.99105099e-01 -5.25432229e-01
-8.30192864e-01 -2.40671232e-01 -4.25624579e-01 4.51077223e-01
-4.41097736e-01 7.96205103e-01 -3.64559144e-01 -9.99576390e-01
-1.34181291e-01 -1.13338816e+00 -1.74041033e-01 -5.78312397e-01
-2.99685240e-01 -2.31589794e-01 -1.21778689e-01 -1.40404344e+00
1.12920690e+00 -1.53459692e+00 2.61821330e-01 1.60873294e-01
-3.39510262e-01 2.33995900e-01 -5.53376615e-01 4.70136076e-01
9.49109253e-03 3.83362681e-01 -1.73369597e-03 -1.82814270e-01
6.41531721e-02 5.89281082e-01 -3.06740552e-01 4.84278560e-01
4.28018123e-01 1.49427342e+00 -7.42983520e-01 -4.14756089e-01
-2.47463614e-01 2.35754609e-01 -3.88659775e-01 2.86790375e-02
2.85817444e-01 5.79701781e-01 -6.81000501e-02 9.08450902e-01
4.12703663e-01 3.13740760e-01 2.55463690e-01 3.29649299e-01
-4.72826958e-01 1.17148018e+00 -4.00429517e-01 1.97971559e+00
-9.54085886e-01 6.21608257e-01 -3.00531745e-01 -4.73885387e-01
1.02882075e+00 1.75607398e-01 -2.38057956e-01 -9.81136143e-01
2.04556644e-01 9.89780009e-01 5.39049208e-01 -1.16141498e-01
1.08585405e+00 -1.17890976e-01 -4.08900499e-01 9.02471900e-01
4.10501331e-01 -2.51878411e-01 4.48667675e-01 -8.68561193e-02
5.76671898e-01 1.71722472e-01 3.91929567e-01 -8.37718129e-01
3.07283938e-01 2.78602928e-01 5.40683627e-01 4.28343177e-01
5.20232581e-02 4.63621825e-01 1.83898993e-02 -6.19440675e-01
-1.45639646e+00 -1.29101110e+00 -2.09660053e-01 1.43645513e+00
-4.65416849e-01 -2.75765747e-01 -7.29360104e-01 -7.23040223e-01
-4.00581121e-01 5.47853887e-01 -2.52936989e-01 1.65624425e-01
-1.33389151e+00 -1.01324224e+00 1.23614693e+00 2.58357644e-01
1.14704557e-01 -1.09035587e+00 3.86892110e-01 5.13161495e-02
-5.74548602e-01 -1.06207132e+00 -1.16104591e+00 2.61527836e-01
-8.61335635e-01 -4.47512329e-01 -9.36767578e-01 -1.40002763e+00
7.55154729e-01 -6.02790266e-02 1.44961357e+00 -1.04946814e-01
8.17659497e-02 -3.05667400e-01 -2.73936838e-01 -3.79363239e-01
-1.01927423e+00 8.08909535e-01 4.23607081e-01 -5.02598405e-01
5.40418923e-01 -3.65959644e-01 -5.71895316e-02 3.25075835e-01
-5.24400175e-01 5.90053312e-02 7.24632323e-01 7.51227021e-01
5.18947423e-01 -7.59411514e-01 5.21547973e-01 -7.82072008e-01
8.38339567e-01 -4.61499214e-01 -6.60133123e-01 5.63724995e-01
-6.63047373e-01 1.60033330e-01 1.01465118e+00 -6.30029917e-01
-1.01475990e+00 -2.40818024e-01 -4.04668152e-01 4.95512247e-01
3.09446961e-01 6.05805278e-01 -7.66712651e-02 -3.38979624e-03
1.01249897e+00 3.64248544e-01 -5.34630358e-01 -4.59132075e-01
7.08935678e-01 8.83012176e-01 7.77262807e-01 -8.35920155e-01
8.29788148e-01 -1.82943344e-01 -7.38812923e-01 -6.22482896e-01
-1.32692307e-01 -8.24864507e-02 -1.04424763e+00 3.31772149e-01
6.61809027e-01 -1.04875517e+00 -1.01515204e-01 3.61727655e-01
-1.50122881e+00 -3.25891048e-01 -1.46544024e-01 8.99932623e-01
-1.95400760e-01 3.32671106e-02 -1.18391871e+00 -4.59873304e-02
-8.22513402e-01 -1.25678790e+00 8.11815560e-01 -6.49471655e-02
-5.16274989e-01 -1.22541118e+00 7.57594764e-01 2.54487067e-01
5.34574032e-01 -4.48932856e-01 1.17373049e+00 -6.04856193e-01
-2.26961210e-01 2.67044455e-01 -5.38082719e-02 5.49451649e-01
4.22122806e-01 -9.53679234e-02 -1.94876045e-01 -2.54797459e-01
-4.57544923e-01 -5.50924659e-01 4.62875009e-01 3.77192236e-02
-7.39094093e-02 -2.35254854e-01 2.31109858e-01 8.60957623e-01
1.06307566e+00 8.50852802e-02 2.74028450e-01 4.42670584e-01
7.20578372e-01 5.36468029e-01 2.51300067e-01 -1.93883091e-01
7.44209111e-01 3.70044231e-01 -5.79345226e-01 -5.65775156e-01
-3.23579520e-01 -5.72561681e-01 1.04517043e+00 1.98593581e+00
1.03496060e-01 -1.25094503e-01 -1.29967916e+00 9.20153856e-01
-1.27907097e+00 -1.90239981e-01 -9.75516364e-02 2.13520670e+00
1.29295766e+00 -2.51095265e-01 -3.16648475e-05 -5.12688816e-01
7.37149417e-01 -2.06354380e-01 -3.05559784e-01 -1.28007174e+00
-7.46437311e-01 7.56798923e-01 9.63848531e-01 1.01384389e+00
-3.62268806e-01 1.92907417e+00 7.50328016e+00 8.36163521e-01
-1.24681735e+00 3.20968539e-01 7.12721169e-01 2.86944628e-01
-8.48357499e-01 1.30751908e-01 -1.08155251e+00 1.72022924e-01
1.31051981e+00 -4.68708485e-01 8.23770642e-01 3.79292190e-01
1.91968381e-01 1.65491670e-01 -1.19402373e+00 6.24015152e-01
2.39104986e-01 -1.06290257e+00 2.22532526e-01 -1.23029813e-01
1.28030849e+00 7.22148955e-01 1.84881195e-01 3.05726975e-01
9.56340134e-01 -1.04075527e+00 7.01946616e-01 -8.85464028e-02
1.10060525e+00 -9.15222645e-01 2.81279773e-01 1.49686843e-01
-8.49323392e-01 5.30842483e-01 -5.73456347e-01 -1.71077400e-01
2.86461115e-01 4.13836181e-01 -9.55503404e-01 4.07685935e-01
2.72321463e-01 5.54156244e-01 -7.77887106e-01 1.61469772e-01
-7.71671653e-01 9.80559886e-01 -3.05264533e-01 -1.68265387e-01
3.53655338e-01 -4.72831428e-01 2.78985143e-01 1.51731813e+00
5.93161404e-01 -6.51067734e-01 1.81305945e-01 6.77788854e-01
-3.69965285e-01 9.72330511e-01 -5.71121693e-01 -5.81739664e-01
2.90594131e-01 1.15172744e+00 -6.37140930e-01 -3.12094867e-01
-3.87758225e-01 1.59553146e+00 7.43515253e-01 4.37314212e-01
-5.34822047e-01 -4.40017283e-01 7.53544331e-01 -3.50016505e-01
-2.17168361e-01 -6.52917862e-01 -3.46298158e-01 -1.49668968e+00
1.50138587e-01 -1.45493269e+00 -2.52101831e-02 -4.77016658e-01
-1.18970716e+00 1.06098342e+00 -4.59899962e-01 -7.70108998e-01
-5.41937768e-01 -8.85804653e-01 -6.14400387e-01 1.67052984e+00
-1.27827859e+00 -1.44669068e+00 8.04259062e-01 4.99874145e-01
6.81878150e-01 -5.75600505e-01 1.07705367e+00 7.18266189e-01
-6.47350848e-01 1.16403830e+00 2.68064886e-01 2.94132054e-01
1.10819137e+00 -1.01489782e+00 1.40611720e+00 1.12148428e+00
6.37421727e-01 1.10078442e+00 2.86933541e-01 -7.42067575e-01
-1.37175894e+00 -1.08998895e+00 1.96968257e+00 -7.49996960e-01
9.40721750e-01 -5.98049521e-01 -3.21847469e-01 1.10818756e+00
5.84160805e-01 -5.62014639e-01 8.85036170e-01 4.39039826e-01
-3.69272709e-01 2.92577446e-01 -5.26695788e-01 1.04969800e+00
1.11981058e+00 -9.44120824e-01 -6.29281700e-01 6.08589709e-01
9.53942895e-01 -1.49786100e-01 -9.43384528e-01 8.06299075e-02
6.82971716e-01 -1.21511266e-01 6.50956035e-01 -7.59560168e-01
6.55459523e-01 -8.71500447e-02 -3.86812463e-02 -1.96907079e+00
-4.06291693e-01 -8.92625093e-01 7.26105690e-01 1.28429365e+00
1.19278252e+00 -7.64073849e-01 3.38243902e-01 -1.49333999e-01
-2.50204474e-01 -3.71751279e-01 -8.71391356e-01 -9.21155334e-01
8.55863631e-01 -1.34753183e-01 4.67717081e-01 1.19676220e+00
2.81565368e-01 1.00168490e+00 -4.63327646e-01 -2.07541585e-01
-1.71290249e-01 5.65403886e-02 6.36426806e-01 -4.07788813e-01
-5.62746882e-01 -3.99962157e-01 1.19347349e-01 -1.05619121e+00
5.90001881e-01 -1.81284451e+00 -1.08815849e-01 -1.20699286e+00
2.02317789e-01 -3.01915199e-01 -5.29556647e-02 4.96600717e-01
-4.16247487e-01 6.45130813e-01 -8.15636292e-02 3.06335479e-01
-1.24065511e-01 3.30806464e-01 1.27914476e+00 7.60206431e-02
-4.19095397e-01 -3.46282631e-01 -7.95891106e-01 4.91615832e-01
1.05961263e+00 -4.68042850e-01 -7.07150921e-02 -1.38603914e+00
5.53388357e-01 -1.32662252e-01 -7.87996948e-01 -2.81579167e-01
-7.24108964e-02 -4.12390381e-01 2.53219783e-01 -2.49753058e-01
-3.14459205e-01 -2.22491339e-01 -2.08775595e-01 4.75813359e-01
-4.69390750e-01 1.28596187e+00 1.90762922e-01 -4.11454886e-01
-1.42728165e-01 4.22275886e-02 6.56862080e-01 -3.18790913e-01
-4.46713269e-01 2.41509527e-01 -7.42942512e-01 9.33923945e-02
3.32810342e-01 1.04530223e-01 -3.04860055e-01 9.81486961e-02
-2.27510870e-01 -2.20370591e-02 5.61853051e-01 7.83011615e-01
1.54716477e-01 -1.50708830e+00 -1.30578101e+00 4.83005494e-01
2.47770399e-01 -8.92879546e-01 -6.38792634e-01 6.57966435e-01
-9.33015525e-01 5.25593519e-01 -5.90968251e-01 -1.08334765e-01
-1.00447607e+00 2.46270522e-01 1.39314830e-01 -5.66937268e-01
1.52734490e-02 7.05455065e-01 -1.48889378e-01 -1.56662154e+00
-3.15896749e-01 -2.81076282e-01 2.28192478e-01 -3.95727485e-01
3.37343037e-01 3.04048836e-01 3.24016482e-01 -1.15346682e+00
-9.07825381e-02 5.92903197e-01 -2.68017590e-01 -6.38903439e-01
7.09273338e-01 -4.15668488e-01 -5.38304090e-01 4.53166693e-01
1.21497703e+00 7.42587745e-01 -9.27165374e-02 -5.20229816e-01
7.40731955e-02 -2.28388339e-01 -5.17117798e-01 -9.24435794e-01
-5.73396981e-01 8.46681654e-01 2.67938852e-01 -9.13451314e-01
7.32210577e-01 -2.37600252e-01 1.23266304e+00 9.12850082e-01
2.66085476e-01 -1.57940626e+00 -4.00014460e-01 1.44175828e+00
5.76893449e-01 -1.38810253e+00 -5.14201999e-01 -6.18899614e-02
-5.74602008e-01 9.20669377e-01 5.34305811e-01 -5.44518121e-02
2.67151326e-01 1.88834324e-01 5.93687117e-01 2.34718233e-01
-3.74098986e-01 1.46990931e-02 4.13680613e-01 3.66813719e-01
1.02088630e+00 4.04214114e-01 -1.26776922e+00 3.04922968e-01
-1.05285406e+00 -4.87860769e-01 2.25309551e-01 5.74214995e-01
-1.14419125e-01 -1.83366680e+00 -4.69661146e-01 9.83892381e-02
-8.41066003e-01 -1.00308180e+00 -8.09910417e-01 3.62209767e-01
1.78536475e-01 8.72646213e-01 3.92888784e-02 -3.69850874e-01
2.52913516e-02 1.07417122e-01 5.08639812e-01 -6.29063487e-01
-1.04380918e+00 -1.34703606e-01 3.04354817e-01 -3.29015888e-02
1.91941127e-01 -4.48309034e-01 -9.93548870e-01 -6.28375053e-01
-2.84632802e-01 2.98052907e-01 1.03465354e+00 1.01357579e+00
-3.50490175e-02 3.05482298e-01 7.15564966e-01 -4.86267239e-01
-1.76238552e-01 -1.08138156e+00 -1.75773010e-01 1.52369455e-01
-1.10685900e-01 4.96167839e-01 1.71495299e-03 7.95122087e-02] | [11.518038749694824, 10.375542640686035] |
4942f881-ff3c-4a8a-ad6d-50445169cc15 | cross-platform-and-cross-domain-abusive | 2211.06452 | null | https://arxiv.org/abs/2211.06452v1 | https://arxiv.org/pdf/2211.06452v1.pdf | Cross-Platform and Cross-Domain Abusive Language Detection with Supervised Contrastive Learning | The prevalence of abusive language on different online platforms has been a major concern that raises the need for automated cross-platform abusive language detection. However, prior works focus on concatenating data from multiple platforms, inherently adopting Empirical Risk Minimization (ERM) method. In this work, we address this challenge from the perspective of domain generalization objective. We design SCL-Fish, a supervised contrastive learning integrated meta-learning algorithm to detect abusive language on unseen platforms. Our experimental analysis shows that SCL-Fish achieves better performance over ERM and the existing state-of-the-art models. We also show that SCL-Fish is data-efficient and achieves comparable performance with the large-scale pre-trained models upon finetuning for the abusive language detection task. | ['Laks V. S. Lakshmanan', 'Muhammad Abdul-Mageed', 'Md Tawkat Islam Khondaker'] | 2022-11-11 | null | null | null | null | ['abusive-language'] | ['natural-language-processing'] | [-1.56396821e-01 -5.31471014e-01 -5.67297578e-01 -1.23047754e-01
-1.40018618e+00 -7.89131641e-01 5.35134852e-01 1.59713224e-01
-4.11324173e-01 5.24882793e-01 7.59932175e-02 -2.59339005e-01
1.83346838e-01 -1.49380907e-01 -4.16155517e-01 -2.21549228e-01
-5.12759201e-02 1.94231957e-01 -1.58948693e-02 -3.04972172e-01
4.47538286e-01 5.43124914e-01 -9.31534410e-01 4.20747757e-01
8.32171619e-01 6.92004561e-01 -5.43602943e-01 4.28856671e-01
-2.42349103e-01 8.22236240e-01 -9.91704822e-01 -1.06085563e+00
1.50492027e-01 -3.02675724e-01 -7.06374466e-01 -5.35265692e-02
7.98449636e-01 -4.42614645e-01 -1.32127792e-01 1.31478751e+00
5.67636609e-01 -1.68393061e-01 5.65050423e-01 -1.38526666e+00
-5.39502263e-01 8.02416325e-01 -1.10627735e+00 9.77223754e-01
3.22521359e-01 1.68599293e-01 8.77911747e-01 -7.51189828e-01
4.02318388e-01 1.57866704e+00 6.55038714e-01 6.90818310e-01
-1.30896676e+00 -1.26126945e+00 3.97047400e-01 -1.54466793e-01
-1.52237165e+00 -5.35148144e-01 9.88222122e-01 -8.96203995e-01
9.20961738e-01 -3.76457279e-03 8.69696960e-02 1.93205571e+00
1.22123443e-01 9.64718223e-01 9.73625302e-01 -2.92296767e-01
6.78566322e-02 4.27159935e-01 4.07108873e-01 6.07861638e-01
5.56855798e-01 -2.39469379e-01 -9.22340989e-01 -8.57038319e-01
1.68552734e-02 -2.27190703e-01 2.86103278e-01 2.55414069e-01
-3.82033646e-01 1.08639681e+00 -5.38982376e-02 4.41007793e-01
-7.01254010e-02 -1.02107443e-01 9.18605328e-01 2.77192354e-01
8.59116375e-01 6.61144078e-01 -2.15369821e-01 -2.27975637e-01
-1.04335153e+00 1.46676630e-01 6.02183819e-01 7.54426956e-01
3.34529936e-01 2.85634696e-02 2.38484181e-02 9.40727234e-01
3.59846264e-01 2.83046693e-01 8.42060864e-01 -5.57414651e-01
6.95271969e-01 7.35410571e-01 -1.81860309e-02 -1.00449240e+00
-1.81560829e-01 -4.43476349e-01 -2.82012999e-01 -1.05598591e-01
1.67382926e-01 -4.00332928e-01 -3.54924619e-01 1.73889232e+00
2.36283690e-01 3.08593929e-01 -2.02178553e-01 4.68521535e-01
3.89775872e-01 4.45951462e-01 6.04056537e-01 -4.30746019e-01
1.19569039e+00 -8.88136089e-01 -9.43459630e-01 -5.63814342e-01
1.00015819e+00 -7.07044244e-01 1.53289354e+00 6.82796836e-01
-7.88988590e-01 6.66754926e-03 -1.35194409e+00 -9.04617533e-02
-4.99737203e-01 -2.60511696e-01 3.34551066e-01 1.27410865e+00
-4.33482945e-01 2.59541303e-01 -5.75804591e-01 -1.33149952e-01
6.02517307e-01 -6.85044974e-02 -1.72845170e-01 2.95167625e-01
-1.23694944e+00 9.53937173e-01 9.25290883e-02 -4.88159984e-01
-1.33280647e+00 -9.22997117e-01 -8.30286026e-01 -3.62462670e-01
3.75934392e-01 -8.54063481e-02 1.42352962e+00 -1.11669922e+00
-1.45180058e+00 1.35381722e+00 9.78004001e-03 -4.18764979e-01
5.70618808e-01 -7.69839764e-01 -8.21769357e-01 -1.27353519e-01
2.56940901e-01 5.34265935e-02 1.19960082e+00 -1.15950871e+00
-3.37706864e-01 -5.23866892e-01 -1.47087976e-01 -8.02330226e-02
-1.01745725e+00 7.63487160e-01 1.30464792e-01 -7.32249022e-01
-6.66220486e-01 -6.03162646e-01 3.03630412e-01 -3.94491643e-01
-5.41763186e-01 -3.80203187e-01 1.09275877e+00 -8.83853912e-01
1.79516149e+00 -2.44502401e+00 1.34843569e-02 -5.43312542e-02
2.92460680e-01 6.62214220e-01 -8.49398524e-02 2.37267494e-01
8.90841484e-02 6.42644227e-01 -2.02075556e-01 -9.52283025e-01
5.09578222e-03 -2.17216402e-01 -6.00496113e-01 8.25131297e-01
3.36975634e-01 2.97258705e-01 -8.52317631e-01 -7.81367123e-01
-2.68016040e-01 2.02189922e-01 -4.93125916e-01 4.13463533e-01
-8.07089955e-02 9.83355939e-02 -3.12496513e-01 8.92365158e-01
6.93515122e-01 1.03554703e-01 1.74981102e-01 3.09709460e-01
2.18713079e-02 4.80284870e-01 -6.33413911e-01 1.62188280e+00
-4.91517931e-01 4.53336239e-01 1.83370158e-01 -4.74123091e-01
8.92966092e-01 3.26072089e-02 1.58144146e-01 -5.61489880e-01
4.33267355e-01 2.19571874e-01 -2.05081463e-01 -5.86741209e-01
4.28579152e-01 -4.14306819e-01 -5.33398151e-01 4.57516283e-01
6.72815293e-02 3.10268342e-01 -9.00078192e-02 4.57853973e-01
1.17226505e+00 -2.62643069e-01 6.13805950e-01 -2.33212233e-01
7.61456788e-01 -7.96985179e-02 4.77818400e-01 7.08297253e-01
-9.17802513e-01 1.45301819e-02 7.24291146e-01 -1.39822513e-01
-6.66121125e-01 -1.07075715e+00 -1.26746535e-01 1.72730446e+00
1.19124027e-02 -4.11576748e-01 -9.28803623e-01 -1.20432746e+00
1.67937592e-01 9.11124825e-01 -5.64090371e-01 -3.82506073e-01
-5.22699833e-01 -8.21752250e-01 1.20320630e+00 9.98076648e-02
5.08033931e-01 -7.08019614e-01 -1.29064530e-01 1.28327057e-01
-2.39057466e-01 -1.25983357e+00 -7.27653861e-01 -1.99647188e-01
-6.04498565e-01 -8.63950908e-01 -7.08155334e-02 -3.83727133e-01
-1.36320964e-01 1.85502842e-01 1.10599637e+00 6.42960221e-02
-3.36827964e-01 4.09341268e-02 -2.31289595e-01 -4.94502068e-01
-8.77352238e-01 4.40234840e-01 4.04142141e-01 2.86185086e-01
8.96086693e-01 -5.09074390e-01 -3.01170498e-01 1.52925536e-01
-7.98903346e-01 -7.96389401e-01 2.56923527e-01 3.41954648e-01
2.18218416e-01 -2.83921868e-01 9.88617778e-01 -1.00752950e+00
1.09760821e+00 -9.44426179e-01 -4.37502444e-01 6.26556873e-02
-3.58146846e-01 -3.36008489e-01 4.21240985e-01 -9.02851462e-01
-1.13999867e+00 -2.67991602e-01 -1.59295931e-01 -5.45082867e-01
-1.23836063e-01 2.70751268e-01 -3.85068029e-01 7.24785700e-02
1.04605937e+00 -4.05284911e-02 -5.07169962e-02 -7.47539401e-01
2.56593019e-01 1.24394381e+00 2.17354387e-01 -4.32134300e-01
5.32070220e-01 5.41632116e-01 -5.71853995e-01 -9.16091442e-01
-1.29854417e+00 -8.19212079e-01 -5.68298161e-01 -3.93698849e-02
7.12731838e-01 -1.23094237e+00 -5.28069496e-01 3.83150697e-01
-1.02026010e+00 -3.45402569e-01 5.21558404e-01 -1.50596052e-01
-3.13335210e-01 7.02423573e-01 -9.08754349e-01 -1.22692800e+00
-5.87625861e-01 -9.91842747e-01 1.11328924e+00 -2.18345448e-01
-6.30424738e-01 -9.62236047e-01 4.85520840e-01 6.43356681e-01
7.04020262e-02 3.16575944e-01 7.24071503e-01 -1.13519728e+00
1.78436965e-01 -2.92789161e-01 -3.36581394e-02 4.75439131e-01
7.55070597e-02 1.15095183e-01 -1.17372882e+00 -5.85058928e-01
7.11500943e-02 -6.16040647e-01 5.97029209e-01 -2.30295539e-01
1.09997487e+00 -5.32436728e-01 -3.01422536e-01 2.20271781e-01
1.23025930e+00 -1.51394129e-01 3.07582200e-01 5.66550970e-01
6.39715791e-01 4.87866104e-01 5.68820596e-01 6.66186512e-01
1.57855041e-02 6.76467419e-01 3.43925893e-01 3.37413251e-01
3.49863656e-02 -4.61886466e-01 1.01832652e+00 3.56410444e-01
9.06592667e-01 -1.35228917e-01 -1.09978807e+00 7.29316950e-01
-1.59778404e+00 -9.91931558e-01 9.48872119e-02 1.94470823e+00
9.35478091e-01 2.77126372e-01 8.52444291e-01 -3.95137668e-02
8.69130135e-01 7.64345005e-02 -6.58899069e-01 -8.24551761e-01
-1.52972490e-01 -1.90203756e-01 9.48760062e-02 6.58758759e-01
-1.40531611e+00 1.15759575e+00 7.12171841e+00 1.13106012e+00
-1.14093423e+00 8.00085247e-01 6.19471371e-01 -5.64430118e-01
1.66731119e-01 -5.58797240e-01 -1.20485914e+00 7.52364874e-01
1.20279145e+00 -2.55154938e-01 2.67380178e-01 1.20579457e+00
8.26494023e-02 2.08880231e-01 -1.02625215e+00 9.53458965e-01
5.45684457e-01 -7.95607328e-01 1.05931520e-01 1.92031831e-01
5.69242299e-01 1.99960798e-01 4.71161604e-01 6.13759398e-01
3.00682068e-01 -9.53423858e-01 7.89991796e-01 -1.91663787e-01
4.68097061e-01 -8.31216633e-01 5.04595876e-01 7.14632392e-01
-3.74213099e-01 -5.27446091e-01 -1.22444175e-01 -7.08553940e-02
7.03053698e-02 5.61962783e-01 -6.95104003e-01 -4.08241674e-02
8.03704739e-01 6.49007976e-01 -7.74991333e-01 5.64380348e-01
-3.55826691e-02 1.02158618e+00 -1.50266841e-01 -8.38074014e-02
2.70775944e-01 3.21624279e-01 7.86503971e-01 1.74595666e+00
-1.20494403e-01 -3.82167161e-01 2.36406043e-01 8.08848739e-01
-4.44763482e-01 2.69617975e-01 -8.22914481e-01 -3.09760898e-01
5.72079301e-01 1.28861248e+00 2.83188559e-02 -2.76747435e-01
-4.53796387e-01 8.57043266e-01 8.41601551e-01 -2.04406381e-01
-7.81026542e-01 2.15843208e-02 1.24013460e+00 2.95332044e-01
-2.61268020e-01 -1.84972882e-01 -4.12712157e-01 -1.16713977e+00
-1.31197304e-01 -1.23560321e+00 1.13486934e+00 -1.85098857e-01
-2.05010295e+00 6.34138942e-01 8.64958912e-02 -1.01104212e+00
-2.00044841e-01 -5.98710656e-01 -4.13191944e-01 6.50424957e-01
-1.48221612e+00 -1.22301841e+00 2.87730638e-02 4.82113183e-01
6.36565804e-01 -4.99202251e-01 5.30625463e-01 3.78011495e-01
-1.16700470e+00 1.17450297e+00 -2.78657079e-01 2.59060532e-01
9.54171956e-01 -8.09100151e-01 1.07641175e-01 1.00828886e+00
-2.33958334e-01 8.00036728e-01 7.42585957e-01 -9.46576238e-01
-1.06177902e+00 -1.20948350e+00 6.20946586e-01 -9.96807873e-01
1.31939149e+00 -6.51572227e-01 -1.10975587e+00 8.30266297e-01
3.60618174e-01 -2.08799839e-01 1.47953510e+00 2.92073637e-01
-1.15874577e+00 1.10788777e-01 -1.50714779e+00 6.08751893e-01
1.02512884e+00 -7.10754156e-01 -6.66829705e-01 3.48776728e-01
6.86143517e-01 -6.91975057e-02 -6.20293498e-01 -6.18968830e-02
3.08108717e-01 -8.62845302e-01 8.01381707e-01 -1.25162780e+00
5.59184551e-01 2.08020210e-01 -2.48387963e-01 -9.66456413e-01
-2.07320362e-01 -8.74170005e-01 -3.79320711e-01 1.46354640e+00
4.49686289e-01 -5.13835073e-01 4.77469504e-01 3.28271329e-01
2.68418938e-01 -2.45237157e-01 -1.22630763e+00 -1.09052002e+00
7.29410291e-01 -3.07652712e-01 2.72435188e-01 1.39277112e+00
3.61969948e-01 5.71943939e-01 -5.51470101e-01 4.33470726e-01
7.89976001e-01 -2.19218656e-01 7.07632661e-01 -9.13245797e-01
-2.27726370e-01 -6.78862095e-01 -2.01970130e-01 -5.31993806e-01
7.80319691e-01 -7.72076964e-01 -3.74135554e-01 -4.07588392e-01
6.45758450e-01 3.13793048e-02 -2.23434359e-01 4.11159813e-01
-3.77696455e-01 5.18151760e-01 1.12639740e-01 2.02927306e-01
-9.25931156e-01 2.94205189e-01 4.84861940e-01 -2.85827219e-01
-3.46631229e-01 -3.55623245e-01 -1.00897765e+00 8.43845367e-01
1.01373768e+00 -7.92044520e-01 -9.61357206e-02 -2.02331498e-01
2.87370294e-01 -4.25351858e-01 7.62824863e-02 -5.36424518e-01
-3.01780313e-01 -1.87803239e-01 -1.51520938e-01 -3.10686171e-01
2.63969004e-01 -4.73465592e-01 -4.44759548e-01 4.18629318e-01
-5.42026162e-01 2.70030975e-01 4.70250428e-01 5.66287637e-01
1.06099889e-01 -4.29081678e-01 1.10015512e+00 -5.52161522e-02
-3.32064956e-01 1.75677180e-01 -6.46926999e-01 4.21825588e-01
1.01241183e+00 2.05911130e-01 -5.24563015e-01 -2.46298701e-01
-5.53628325e-01 6.09743372e-02 2.36494631e-01 6.64353549e-01
3.65174741e-01 -1.04696870e+00 -8.44025791e-01 -1.49297640e-01
5.24305701e-01 -7.04308629e-01 1.37970466e-02 5.87936997e-01
-1.70863301e-01 8.12427476e-02 9.66286734e-02 -4.28117603e-01
-1.62925911e+00 9.10382092e-01 5.18302083e-01 -4.89688784e-01
-1.03323147e-01 7.15228081e-01 6.07530884e-02 -3.63862902e-01
1.54712722e-01 6.53120220e-01 2.96936687e-02 1.67033046e-01
1.01406980e+00 4.93579209e-01 5.87746277e-02 -7.92539120e-01
-6.55207038e-01 -2.56771781e-02 -6.33229494e-01 -8.44353586e-02
9.54742253e-01 -1.33526787e-01 -1.55949771e-01 5.99177122e-01
1.43575287e+00 4.61030304e-01 -8.50123286e-01 -3.61517012e-01
4.37266886e-01 -7.28355825e-01 9.00023207e-02 -9.09854889e-01
-4.80423838e-01 8.38566661e-01 5.31918168e-01 2.28514045e-01
7.86953330e-01 1.73061356e-01 6.83670402e-01 8.46404880e-02
3.38531107e-01 -1.06633162e+00 8.91313910e-01 3.11666399e-01
9.73728955e-01 -1.37279058e+00 9.41822082e-02 -6.94948971e-01
-6.80566669e-01 5.48220992e-01 1.03301823e+00 -1.12286188e-01
5.96289814e-01 4.07429695e-01 3.54328334e-01 -1.24691144e-01
-5.57542443e-01 7.51452148e-02 5.90476282e-02 4.15437222e-01
4.81249571e-01 6.27674535e-02 -2.47425318e-01 9.59319413e-01
-5.32084703e-02 -4.18817908e-01 4.45449173e-01 9.12382603e-01
-2.87283361e-01 -1.07058668e+00 -4.26533043e-01 9.95189101e-02
-9.41162288e-01 -3.89074013e-02 -1.18887556e+00 7.70211339e-01
-8.07085186e-02 1.04404199e+00 -3.50163095e-02 -4.65763479e-01
2.60748804e-01 3.45748603e-01 2.88824528e-01 -6.83109403e-01
-1.08244097e+00 1.09342679e-01 2.38435313e-01 -3.94509703e-01
-6.52492931e-03 -8.30825627e-01 -6.45557046e-01 -5.90610683e-01
-2.73002476e-01 -2.30523735e-01 4.34110254e-01 8.49965870e-01
6.78619683e-01 -4.43790443e-02 7.80537784e-01 -4.47095871e-01
-1.03353667e+00 -1.22080445e+00 -5.41647434e-01 7.92705238e-01
5.59886098e-01 -7.78408229e-01 -7.08856285e-01 -3.33141297e-01] | [8.813624382019043, 10.530917167663574] |
55b990d4-9a31-44bb-aaac-69ecf67b7c55 | parkinsons-disease-emg-signal-prediction | null | null | https://ieeexplore.ieee.org/abstract/document/8914553 | https://ieeexplore.ieee.org/abstract/document/8914553 | Parkinson’s Disease EMG Signal Prediction Using Neural Networks | This paper proposes a comparison between different neural network models, using multilayer perceptron (MLPs) and recurrent neural network (RNN) models, for predicting Parkinson's disease electromyography (EMG) signals, to anticipate resulting resting tremor patterns. The experimental results indicate that the proposed models can adapt to different frequencies and amplitudes of tremor, and provide reasonable predictions for both EMG envelopes and EMG raw signals. Therefore, one could use these models as input for a control strategy for functional electrical stimulation (FES) devices used on tremor suppression, by dynamically predicting and improving FES control parameters based on tremor forecast. | ['Maria Claudia Ferrari de Castro', 'Esther Luna Colombini', 'Rafael Anicet Zanini'] | 2019-10-06 | null | null | null | null | ['emg-signal-prediction', 'electromyography-emg'] | ['medical', 'medical'] | [ 6.38018131e-01 3.20714951e-01 -4.78675455e-01 -8.80862251e-02
-3.69713083e-02 3.49538386e-01 6.26069382e-02 -8.74164045e-01
-3.20965052e-01 1.00777912e+00 7.29999840e-01 -2.73167580e-01
-7.80238271e-01 -3.60553890e-01 -6.69177920e-02 -3.96645963e-01
-2.40102097e-01 3.09044719e-01 2.52893772e-02 -2.79542178e-01
3.14076960e-01 6.18158281e-01 -1.48921657e+00 5.31101525e-01
7.93799877e-01 5.91220438e-01 7.87692845e-01 6.94114566e-01
3.62762690e-01 8.76336217e-01 -5.50127268e-01 6.01190805e-01
-9.65770036e-02 -2.48909131e-01 -7.14850843e-01 5.49265668e-02
-9.40323651e-01 -2.55929023e-01 -8.21700096e-01 7.83922434e-01
7.27121174e-01 -9.74132214e-03 7.25164890e-01 -6.13990545e-01
-4.38315988e-01 8.34417105e-01 3.33934650e-02 2.65790850e-01
4.19301897e-01 3.99779886e-01 1.27599433e-01 -3.13707560e-01
5.41987836e-01 8.33326221e-01 1.04990244e+00 9.90367949e-01
-1.14207232e+00 -3.26338321e-01 -2.80171841e-01 7.86717772e-01
-5.73164761e-01 -2.11929277e-01 8.13838124e-01 -4.23390925e-01
1.28949976e+00 4.37306166e-01 1.16588008e+00 1.58215308e+00
1.09099543e+00 7.78449535e-01 1.06932163e+00 -4.08958346e-01
1.15626179e-01 -2.16137841e-01 2.73511231e-01 -7.33320788e-02
9.50571671e-02 5.91792643e-01 -2.60571450e-01 -1.06040023e-01
9.92346048e-01 1.21905610e-01 -7.03942060e-01 5.44271529e-01
-1.00258446e+00 4.74586606e-01 2.65524507e-01 7.07478762e-01
-1.50129628e+00 1.27082080e-01 5.87305188e-01 4.46146727e-01
2.20021680e-02 5.14706731e-01 -5.62182426e-01 -7.82148242e-01
-5.16120970e-01 2.39486948e-01 7.73666799e-01 2.99470097e-01
-3.41828316e-01 5.43140113e-01 -4.48858202e-01 9.03190017e-01
3.72953057e-01 2.79411614e-01 1.38329232e+00 -8.98261130e-01
2.23071307e-01 7.36542881e-01 1.69337288e-01 -6.50663435e-01
-1.04616618e+00 -1.86596066e-01 -9.71939802e-01 2.57937670e-01
-1.98666304e-01 -6.20450616e-01 -9.00500178e-01 1.18805528e+00
-3.27442795e-01 -1.41172320e-01 1.83484569e-01 8.71825039e-01
4.56929922e-01 4.31296408e-01 2.27113233e-05 -4.87988234e-01
1.05858576e+00 -5.00479579e-01 -1.24499333e+00 -4.62639272e-01
5.40192783e-01 -2.52429336e-01 8.28836083e-01 7.38732636e-01
-9.71941948e-01 -4.54127133e-01 -7.30977535e-01 7.03464806e-01
3.68138850e-02 6.00050807e-01 3.75323564e-01 2.94837713e-01
-7.64684796e-01 1.56266165e+00 -1.44518399e+00 -2.89301127e-01
-1.69089124e-01 9.67065156e-01 2.02224165e-01 7.36806929e-01
-1.45552099e+00 1.43261874e+00 4.87116307e-01 8.91717076e-01
-1.53652877e-01 1.30483359e-02 -2.14431047e-01 -1.46634027e-01
-4.60624188e-01 -7.15359807e-01 1.20853651e+00 -5.13377309e-01
-1.98901021e+00 8.25894028e-02 1.38274729e-01 -7.22433925e-01
2.96436429e-01 -3.79129618e-01 -6.67835653e-01 -9.84227881e-02
-5.87828159e-01 2.55334973e-01 6.71009898e-01 -5.89358747e-01
-2.02725455e-01 -3.89599293e-01 -6.25693738e-01 1.60953298e-01
-2.80986160e-01 3.91589925e-02 4.51793909e-01 -6.99767947e-01
2.44471118e-01 -9.90027785e-01 -3.53908598e-01 -4.22031701e-01
-5.92331052e-01 -3.54434878e-01 6.86560035e-01 -1.21357727e+00
1.36418355e+00 -1.68596458e+00 5.30736089e-01 2.62208134e-01
-4.01845902e-01 4.13903654e-01 1.34313270e-01 4.96528268e-01
-2.65401751e-01 -4.94482398e-01 -3.83376982e-03 4.42512602e-01
-1.35605916e-01 6.46298885e-01 9.16358605e-02 2.35431895e-01
2.50998363e-02 8.25747013e-01 -4.57256079e-01 1.66718990e-01
4.59758580e-01 5.64520538e-01 1.24196438e-02 3.69836539e-01
-2.47627661e-01 7.13094831e-01 -4.42291319e-01 5.70759952e-01
-1.51636690e-01 3.95482555e-02 4.15499836e-01 -4.28272843e-01
-2.12738961e-02 3.76736313e-01 -9.06154454e-01 9.75158215e-01
-3.80871743e-01 6.37168467e-01 -4.92394529e-02 -9.74978566e-01
1.26358891e+00 9.57821190e-01 7.29375601e-01 -4.51992571e-01
7.01801002e-01 5.99631488e-01 2.49288216e-01 -1.43467188e+00
-1.18742660e-01 -1.37929350e-01 4.74499464e-01 2.99308658e-01
-1.26771405e-01 4.18909222e-01 -2.16802824e-02 -1.08505869e+00
1.32969928e+00 4.24488604e-01 -1.51925245e-02 4.01974507e-02
5.05613148e-01 -7.16697723e-02 5.53164601e-01 3.27787489e-01
-3.35496515e-02 1.84119433e-01 1.07971936e-01 -3.04211736e-01
-7.92636275e-01 -6.82918966e-01 1.90545376e-02 5.07060826e-01
-2.94699252e-01 2.06741989e-01 -6.05407298e-01 1.24836899e-01
-1.69680730e-01 8.56543541e-01 -3.21490973e-01 -5.67359805e-01
-9.49590921e-01 -8.28484714e-01 4.56456304e-01 9.36245024e-01
3.98204215e-02 -1.75984943e+00 -9.60480988e-01 1.04984617e+00
-2.70852447e-01 -3.69158506e-01 1.08529359e-01 7.54315019e-01
-1.65422487e+00 -1.02132893e+00 -1.00129080e+00 -1.00060511e+00
2.41217822e-01 -5.51101327e-01 1.39169559e-01 -4.50290680e-01
-1.46611392e-01 3.33758563e-01 -3.94568235e-01 -2.61897862e-01
-7.44821668e-01 -2.07912818e-01 4.61032301e-01 -3.14202219e-01
3.59636098e-01 -1.26657152e+00 -5.59308350e-01 1.63381457e-01
-5.03245533e-01 1.45865679e-01 1.18420708e+00 1.08274245e+00
5.74065030e-01 -3.39920484e-02 8.46288741e-01 -4.24415290e-01
1.64914680e+00 -4.21827406e-01 -4.23444360e-02 2.54216015e-01
-8.60865593e-01 5.91835799e-03 8.76943767e-01 -1.03540170e+00
-1.03925407e+00 9.28317159e-02 -4.73458469e-01 -6.38400614e-01
-1.99766412e-01 6.67576015e-01 3.37982357e-01 3.10151689e-02
9.65814412e-01 7.70147562e-01 3.28833967e-01 -8.26466560e-01
-1.22068338e-01 1.36680949e+00 9.33314860e-01 1.10561512e-01
-4.82255965e-02 -1.69681132e-01 -2.38168463e-01 -6.71474278e-01
1.58366174e-01 -1.58594161e-01 -5.64112902e-01 -6.26324415e-01
5.42788982e-01 -3.87665451e-01 -7.88298070e-01 5.35721421e-01
-1.29336452e+00 -2.69896537e-01 -2.70248532e-01 1.44752276e+00
-1.26638722e+00 1.67606100e-01 -1.30013919e+00 -1.24483824e+00
-9.64339733e-01 -1.09822297e+00 4.51784343e-01 1.91143975e-01
-8.34674418e-01 -7.28242278e-01 2.76038259e-01 4.33397256e-02
6.43463314e-01 2.87399739e-01 9.35324252e-01 -4.34977233e-01
1.50670663e-01 -4.02975082e-01 4.18046176e-01 6.29124641e-01
3.93008053e-01 -3.99872631e-01 -4.66142625e-01 3.15535702e-02
7.85181642e-01 6.51104227e-02 3.54411155e-01 8.91530156e-01
9.03162479e-01 -5.51022768e-01 -4.41652030e-01 1.16147347e-01
1.14978147e+00 8.66326869e-01 1.11629927e+00 5.25735497e-01
2.46512383e-01 4.21995938e-01 3.02804053e-01 4.39790905e-01
-3.45281750e-01 3.51188481e-01 2.48606190e-01 2.31060043e-01
-8.29180703e-02 6.51346967e-02 5.68699956e-01 1.20622170e+00
-8.83350968e-01 1.17203131e-01 -4.85260189e-01 2.67438799e-01
-1.97546709e+00 -8.01273763e-01 -2.32316598e-01 1.61026120e+00
7.72131264e-01 1.84091359e-01 2.46061265e-01 5.98658204e-01
9.81441617e-01 -4.14663970e-01 -9.53603983e-01 -4.43252623e-01
2.30552033e-01 2.56294698e-01 6.43345952e-01 -6.69635236e-02
-4.26039338e-01 3.72660458e-01 7.44822741e+00 4.45734292e-01
-1.28861582e+00 -8.34219605e-02 -2.80044615e-01 -1.27343059e-01
1.57148167e-01 -5.65057099e-01 -1.35768458e-01 8.07858944e-01
1.50593936e+00 -7.77389109e-02 8.03518832e-01 6.04397535e-01
9.90707159e-01 3.59116316e-01 -7.60754108e-01 9.32848334e-01
-5.14867067e-01 -1.35083318e+00 -3.12843442e-01 -1.42679498e-01
2.05804169e-01 3.95827144e-01 -4.04467434e-01 1.42645106e-01
-3.21376741e-01 -7.50456750e-01 3.18211049e-01 1.59021032e+00
2.35668749e-01 -3.97353679e-01 1.11501932e+00 4.27326858e-01
-7.46262074e-01 -4.11710024e-01 -1.77719966e-01 -1.71689332e-01
5.11773229e-01 1.47707447e-01 -8.02303433e-01 2.09539637e-01
4.53808397e-01 7.50852883e-01 -6.01676777e-02 1.18319511e+00
-2.59794116e-01 7.58706570e-01 -4.21313852e-01 -7.85034657e-01
-7.41123259e-02 -1.62334993e-01 9.32341278e-01 8.23896706e-01
4.94676322e-01 -5.45066707e-02 -4.39063579e-01 1.09579420e+00
7.12363839e-01 -2.56853849e-01 -3.85690600e-01 -3.34961504e-01
3.19947213e-01 5.81635654e-01 -2.44887874e-01 3.66517380e-02
2.32098922e-02 8.10370266e-01 -2.83776373e-01 3.38427305e-01
-5.36340177e-01 -5.62084734e-01 1.61518797e-01 5.83243482e-02
-2.71011889e-01 1.45647690e-01 -4.80176896e-01 -6.76987588e-01
3.66259456e-01 -7.60697842e-01 -2.02656295e-02 -1.15933001e+00
-1.26299453e+00 6.95698500e-01 1.72620919e-02 -1.60829735e+00
-9.14548934e-01 -8.06587279e-01 -9.16744351e-01 7.76714265e-01
-7.98522472e-01 -5.74794352e-01 3.70578825e-01 5.66711605e-01
4.79876399e-01 -3.49781424e-01 9.91324127e-01 1.88447550e-01
-4.11414355e-01 -1.87119231e-01 3.57049555e-01 -3.73543501e-01
1.54859528e-01 -9.47513759e-01 -1.40851989e-01 2.10213244e-01
-5.47792554e-01 3.65300238e-01 1.08772349e+00 -1.05124164e+00
-1.50263023e+00 -1.00923145e+00 8.89236927e-01 3.44146907e-01
8.02564383e-01 5.42058647e-01 -1.05445790e+00 6.12988830e-01
-1.79320015e-02 -8.66206050e-01 2.29497761e-01 -2.10861862e-01
9.72784340e-01 2.00858161e-01 -1.16758275e+00 8.86820555e-01
8.79793584e-01 -2.40381151e-01 -1.32382500e+00 7.09909320e-01
2.81151682e-01 -3.13662052e-01 -1.47694087e+00 8.16397786e-01
9.36591208e-01 -4.92818624e-01 7.37081170e-01 -5.48707545e-01
1.54045597e-02 2.24410698e-01 1.59594119e-01 -1.66367829e+00
-6.32307231e-01 -7.09405839e-01 -6.92243636e-01 2.82397747e-01
3.26356649e-01 -7.85449803e-01 9.28679943e-01 3.90485674e-01
-5.87428570e-01 -1.20699704e+00 -1.06788409e+00 -9.32530880e-01
-5.19266367e-01 -4.81985807e-01 1.68015748e-01 2.83831775e-01
7.98980713e-01 2.28451595e-01 -9.44486678e-01 -8.26490447e-02
7.11997300e-02 -3.77004385e-01 -1.22617714e-01 -1.33280206e+00
-4.25182968e-01 -5.05852699e-01 -6.01202428e-01 -6.56327605e-01
-8.24493095e-02 -6.93911970e-01 2.39100844e-01 -2.04716659e+00
-4.90832657e-01 3.03206205e-01 -4.65970427e-01 7.14958251e-01
3.23901057e-01 -9.81322303e-02 -4.71824743e-02 5.68478703e-01
4.38081056e-01 7.03597307e-01 1.22846448e+00 -2.48592839e-01
-1.05327713e+00 7.90439069e-01 -2.61653692e-01 9.02103841e-01
1.11392629e+00 -5.98962486e-01 -1.87541395e-01 8.13591760e-03
-1.89486906e-01 7.20117927e-01 1.61698550e-01 -1.46499610e+00
3.82565230e-01 -1.19396724e-01 5.10870993e-01 -8.89308333e-01
3.20267171e-01 -9.24793243e-01 7.41273940e-01 1.35239017e+00
-4.53340799e-01 -1.35542199e-01 1.42179921e-01 5.15099406e-01
-6.71402961e-02 -5.15518486e-01 4.47195500e-01 1.04153939e-02
-3.48134756e-01 -1.69482082e-01 -1.33725703e+00 -1.00032818e+00
6.40144706e-01 -8.56510222e-01 1.12711959e-01 -1.45150259e-01
-1.72207320e+00 1.14615234e-02 -4.82893974e-01 5.09322882e-01
9.49189007e-01 -1.40732253e+00 -4.60494876e-01 3.26559037e-01
-4.60222095e-01 -9.13747191e-01 4.52401191e-01 1.17882204e+00
-3.21740210e-01 6.89100206e-01 -9.70727205e-01 -4.20954049e-01
-1.03312862e+00 -7.49680698e-02 6.39352381e-01 -3.47501099e-01
-1.21368730e+00 1.81418523e-01 -1.11143017e+00 -4.90834683e-01
5.39463758e-01 -4.31609094e-01 -8.65552723e-01 -5.73230267e-01
3.82914007e-01 7.40100622e-01 8.19943696e-02 -2.16400385e-01
8.56874213e-02 2.25844711e-01 3.17879081e-01 7.33755827e-02
1.70312023e+00 3.15800682e-02 -5.93104586e-02 6.21147275e-01
4.65299129e-01 -7.69903719e-01 -1.00368929e+00 2.06451222e-01
3.54861736e-01 2.12494552e-01 2.57089794e-01 -1.24138129e+00
-9.74230707e-01 3.28420043e-01 1.40275156e+00 7.04085231e-02
1.47848797e+00 -4.09900218e-01 1.18365991e+00 8.71907055e-01
4.13419813e-01 -1.57173228e+00 -3.95853758e-01 2.08709210e-01
1.28771567e+00 -2.39003852e-01 -2.56470859e-01 -1.26180071e-02
-4.38657075e-01 1.88307106e+00 2.44036153e-01 -6.39322579e-01
7.02116787e-01 2.20411554e-01 -7.84917995e-02 -1.53810129e-01
-8.49614978e-01 1.78071126e-01 2.40351886e-01 9.60400343e-01
3.30848634e-01 2.29924783e-01 -9.99714315e-01 1.03237045e+00
-2.68332683e-03 1.02198172e+00 3.89065534e-01 1.06514466e+00
-7.73303449e-01 -9.82458711e-01 -7.05089927e-01 1.28783333e+00
-1.26549438e-01 3.17359149e-01 -2.74439871e-01 7.75215626e-01
1.08229637e-01 8.39407742e-01 -1.45335600e-01 -1.00561821e+00
8.06500375e-01 2.13409811e-01 6.56193197e-01 -3.96869153e-01
-8.03135931e-01 3.24210078e-01 2.84743756e-01 -4.40729678e-01
-4.93813455e-01 -5.95552921e-01 -1.54014409e+00 2.22205579e-01
-5.76550603e-01 -6.87804595e-02 7.26708233e-01 1.10308588e+00
2.35213965e-01 1.14693725e+00 6.33853078e-01 -1.17211378e+00
-1.05423725e+00 -1.99865913e+00 -9.66491163e-01 1.38859320e-02
-1.57942981e-01 -8.07837009e-01 -4.39062566e-01 -9.10041258e-02] | [6.8661370277404785, 0.2295713722705841] |
04da1f79-0bfd-4e91-9f11-71d589b0a5a9 | medical-image-harmonization-using-deep | 2010.05355 | null | https://arxiv.org/abs/2010.05355v1 | https://arxiv.org/pdf/2010.05355v1.pdf | Medical Image Harmonization Using Deep Learning Based Canonical Mapping: Toward Robust and Generalizable Learning in Imaging | Conventional and deep learning-based methods have shown great potential in the medical imaging domain, as means for deriving diagnostic, prognostic, and predictive biomarkers, and by contributing to precision medicine. However, these methods have yet to see widespread clinical adoption, in part due to limited generalization performance across various imaging devices, acquisition protocols, and patient populations. In this work, we propose a new paradigm in which data from a diverse range of acquisition conditions are "harmonized" to a common reference domain, where accurate model learning and prediction can take place. By learning an unsupervised image to image canonical mapping from diverse datasets to a reference domain using generative deep learning models, we aim to reduce confounding data variation while preserving semantic information, thereby rendering the learning task easier in the reference domain. We test this approach on two example problems, namely MRI-based brain age prediction and classification of schizophrenia, leveraging pooled cohorts of neuroimaging MRI data spanning 9 sites and 9701 subjects. Our results indicate a substantial improvement in these tasks in out-of-sample data, even when training is restricted to a single site. | ['Christos Davatzikos', 'Ilya M. Nasrallah', 'Haochang Shou', 'David A. Wolk', 'R. Nick Bryan', 'Susan M. Resnick', 'Hans J. Grabe', 'Marilyn S. Albert', 'John C. Morris', 'Ruben C. Gur', 'Raquel E. Gur', 'Daniel H. Wolf', 'Theodore D. Satterthwaite', 'Nikolaos Koutsouleris', 'Jurgen Fripp', 'Sterling C. Johnson', 'Henry Völzke', 'Chuanjun Zhuo', 'Paul Maruff', 'Colin L. Masters', 'Yong Fan', 'Mohamad Habes', 'Ahmed Abdulkadir', 'Dhivya Srinivasan', 'Guray Erus', 'Jimit Doshi', 'Vishnu M. Bashyam'] | 2020-10-11 | null | null | null | null | ['image-harmonization'] | ['computer-vision'] | [ 4.85642731e-01 -8.73113573e-02 -7.23161697e-02 -8.28301549e-01
-9.79742646e-01 -1.88230574e-01 6.19402945e-01 2.66337335e-01
-6.27287686e-01 6.58821881e-01 3.50756794e-01 -6.08589128e-02
-4.32161808e-01 -4.98025537e-01 -3.79423618e-01 -7.19289422e-01
-1.47499725e-01 8.76901329e-01 -1.82813704e-01 4.04853404e-01
7.17007443e-02 4.13837284e-01 -9.95424330e-01 1.97465003e-01
8.46797824e-01 7.55138516e-01 3.23238760e-01 2.22605675e-01
2.86497951e-01 2.51355231e-01 -3.38442415e-01 -3.48258346e-01
6.93359375e-02 -3.29364240e-01 -6.36268258e-01 4.55908999e-02
3.66442174e-01 -3.36736411e-01 -2.26511434e-01 9.37266350e-01
9.96094286e-01 -2.90231168e-01 8.99111688e-01 -8.93214524e-01
-7.29895055e-01 5.90624034e-01 -6.34247243e-01 1.24904469e-01
-2.12267891e-01 1.53379455e-01 6.84275508e-01 -4.38048065e-01
9.20579135e-01 9.03068781e-01 7.78742731e-01 6.41244173e-01
-1.67450488e+00 -7.29686618e-01 -1.28873691e-01 2.56031781e-01
-1.09855628e+00 -5.09171486e-01 5.11478782e-01 -1.02047360e+00
6.34026885e-01 -1.66696176e-01 6.50379300e-01 1.26805937e+00
3.97074401e-01 2.42309377e-01 1.40152252e+00 -1.24685116e-01
2.63485372e-01 -1.78579241e-01 9.11402032e-02 4.86171544e-01
2.41580397e-01 -6.81472123e-02 -3.86914670e-01 -2.75328040e-01
5.25087953e-01 3.35730202e-02 -1.45526096e-01 -4.87311751e-01
-1.19327211e+00 7.84015179e-01 3.99080515e-01 4.56286043e-01
-5.60011446e-01 -2.32583791e-01 3.70820940e-01 5.27947396e-02
6.58031762e-01 4.11942869e-01 -4.64660645e-01 2.06340998e-01
-1.22196805e+00 2.59876579e-01 2.71428347e-01 2.73386538e-01
2.85740972e-01 3.06014903e-03 -2.54528248e-03 9.67396796e-01
2.18374938e-01 5.32591105e-01 9.57199931e-01 -7.70806611e-01
1.89834878e-01 4.43717301e-01 -2.63112664e-01 -7.07558572e-01
-1.03971064e+00 -6.20404005e-01 -9.94253457e-01 2.31845945e-01
4.79273170e-01 -1.17888808e-01 -1.09806812e+00 2.02870369e+00
1.54410988e-01 -1.18842207e-01 -2.11691231e-01 8.68149042e-01
4.67561156e-01 -2.38020215e-02 4.60522652e-01 -1.04502998e-01
1.37783718e+00 -2.54047662e-01 -8.90678167e-02 -3.54279131e-01
8.21516037e-01 -3.86076748e-01 7.43186653e-01 5.83019376e-01
-9.41658437e-01 -3.18200201e-01 -8.55909169e-01 -1.09820828e-01
-2.36280382e-01 1.79516882e-01 5.19386351e-01 5.88577032e-01
-1.05381155e+00 5.58219194e-01 -1.00465882e+00 -5.42269886e-01
8.14926863e-01 4.83635783e-01 -6.50885403e-01 -1.82049289e-01
-8.61153007e-01 1.06339788e+00 3.96035075e-01 -6.24629073e-02
-8.15060735e-01 -1.23449123e+00 -4.89950418e-01 -1.33100361e-01
-1.50999889e-01 -1.04483557e+00 8.30996037e-01 -6.74160302e-01
-9.51314747e-01 1.27670825e+00 1.41855283e-02 -7.09395647e-01
5.17486632e-01 -9.75898057e-02 -3.94969702e-01 9.63253379e-02
2.82976091e-01 7.24114656e-01 9.72315490e-01 -9.18852389e-01
-7.18379840e-02 -1.15164006e+00 -5.61762035e-01 -5.73272631e-02
-1.24819815e-01 7.62991160e-02 1.45942450e-01 -4.82983649e-01
3.34228963e-01 -9.14554656e-01 -3.26921314e-01 -1.29801938e-02
-2.86314964e-01 2.51107305e-01 3.88058662e-01 -1.18091834e+00
7.89572954e-01 -2.01199603e+00 3.45717996e-01 3.16110492e-01
7.66480267e-01 1.99782729e-01 2.71910615e-02 -5.48710488e-02
-4.13230062e-01 9.13833156e-02 -3.48796219e-01 -2.52547055e-01
-1.94235623e-01 -1.88548267e-01 1.26791447e-01 7.31823623e-01
1.68718636e-01 8.64256442e-01 -6.64601445e-01 -4.05273408e-01
1.98172748e-01 4.97186691e-01 -5.00586510e-01 -1.37752350e-02
7.49339014e-02 9.43457365e-01 -3.96308094e-01 4.32434857e-01
4.52114701e-01 -3.48483264e-01 4.45468247e-01 -1.95239186e-01
2.83099443e-01 1.76099092e-02 -4.73074675e-01 1.93534112e+00
-5.52408278e-01 4.24585044e-01 -1.57620702e-02 -1.34669447e+00
1.01347351e+00 1.48233205e-01 8.15993607e-01 -8.90970647e-01
3.04402977e-01 3.01841289e-01 5.08872330e-01 -4.98527676e-01
-2.52580941e-01 -5.26430249e-01 1.67043209e-01 4.80918944e-01
1.70670867e-01 4.32431884e-02 -7.60230795e-03 -9.69743133e-02
1.12857270e+00 -1.97549179e-01 1.38832375e-01 -2.68311292e-01
2.81898320e-01 -3.07763577e-01 5.01227796e-01 5.02852678e-01
-3.41478437e-01 7.16986001e-01 5.71513176e-01 -3.08764130e-01
-1.36561859e+00 -1.28222322e+00 -6.17921233e-01 5.86636424e-01
-6.11637235e-01 -8.28546360e-02 -6.51710868e-01 -3.53739560e-01
3.09230983e-02 4.89553899e-01 -8.19388866e-01 -4.26758498e-01
-2.29942128e-01 -1.34555447e+00 4.64727581e-01 5.63969910e-01
2.69061416e-01 -6.74828887e-01 -5.57155311e-01 2.63062268e-01
1.26644656e-01 -1.04041195e+00 -6.12631924e-02 1.00382134e-01
-1.15122485e+00 -1.20194185e+00 -1.04472733e+00 -4.34602886e-01
5.69012046e-01 -2.56268531e-01 1.02859831e+00 -2.00217262e-01
-7.18132257e-01 2.29491100e-01 -1.04939518e-02 -4.31370318e-01
-4.80153501e-01 2.49524280e-01 1.50421828e-01 -3.82101461e-02
3.20836484e-01 -8.30130696e-01 -8.98645937e-01 -3.09707839e-02
-5.99081099e-01 4.95702326e-02 8.68149757e-01 9.39741075e-01
5.87462187e-01 -4.01369780e-01 8.27896953e-01 -9.06903207e-01
3.59208822e-01 -6.87951744e-01 -4.09322113e-01 1.66025490e-01
-1.05621612e+00 1.38824299e-01 2.33771250e-01 -2.42583916e-01
-1.03560328e+00 -1.21171169e-01 -1.65579617e-01 4.21071276e-02
-2.35124260e-01 7.09213018e-01 -1.34736523e-01 -1.02481490e-03
7.41874397e-01 -3.03040259e-02 4.32463408e-01 -6.05226696e-01
1.73648030e-01 5.63709617e-01 5.15551627e-01 -5.08657873e-01
4.60437655e-01 4.98965144e-01 3.67650300e-01 -7.76428342e-01
-7.46714652e-01 -1.56733662e-01 -9.95741665e-01 4.90564890e-02
1.03431463e+00 -8.71373177e-01 -2.64286458e-01 6.41463935e-01
-7.00054467e-01 -3.55645508e-01 -3.03438697e-02 7.41971791e-01
-5.59420466e-01 2.76188910e-01 -3.86015236e-01 -3.09428841e-01
-5.88867843e-01 -1.28049493e+00 1.11355186e+00 1.12770170e-01
-3.97660315e-01 -1.09730148e+00 1.38871774e-01 4.67485607e-01
4.12675142e-01 3.86386812e-01 1.48275328e+00 -9.19869781e-01
-1.18198201e-01 -2.87845552e-01 -3.79508227e-01 3.86433601e-01
2.22908720e-01 -3.70371163e-01 -8.30269635e-01 -2.81812757e-01
5.48221767e-02 -2.57939607e-01 7.79484868e-01 7.82902837e-01
1.38300991e+00 4.08202887e-01 -3.76627922e-01 5.86068273e-01
1.17817152e+00 2.08189279e-01 6.22073233e-01 2.65275836e-01
5.93644857e-01 6.25118554e-01 -1.04568690e-01 2.10637510e-01
3.54637742e-01 6.90382183e-01 -1.07143307e-02 -1.17165856e-01
-1.80343762e-01 8.51292908e-02 -1.36224210e-01 4.87178624e-01
-1.19452395e-01 3.13037753e-01 -1.20381796e+00 4.54362303e-01
-1.43967688e+00 -8.46587420e-01 -7.56353065e-02 2.41834331e+00
6.92392826e-01 1.26759395e-01 3.22387934e-01 -1.19479097e-01
6.76132977e-01 -3.87395024e-02 -7.04707742e-01 -3.82538401e-02
-1.81799456e-01 2.95130879e-01 3.38252753e-01 -3.60005461e-02
-7.89852440e-01 3.62131000e-01 6.67000675e+00 2.70908982e-01
-1.58058512e+00 5.24149179e-01 1.10568643e+00 -1.44929066e-01
-7.33629316e-02 -3.24905336e-01 -2.75026381e-01 3.12630326e-01
1.23424542e+00 -1.86919987e-01 3.21735620e-01 6.11722171e-01
4.06407833e-01 -1.08603485e-01 -1.16054404e+00 1.07398355e+00
6.51393235e-02 -1.08777893e+00 -2.79977262e-01 1.35501161e-01
4.67810392e-01 2.92362064e-01 2.32695371e-01 -4.70578745e-02
-2.39772871e-01 -1.23367286e+00 4.49608386e-01 7.46738255e-01
9.66482759e-01 -4.38955188e-01 6.11139536e-01 -2.48695910e-02
-4.75696772e-01 2.76939720e-02 -1.09418586e-01 2.30950445e-01
9.12055671e-02 8.17134500e-01 -1.08272386e+00 4.59540397e-01
6.16155505e-01 5.73749781e-01 -7.51734436e-01 1.13947463e+00
1.45498946e-01 5.15068352e-01 3.72870476e-03 5.66073120e-01
-7.64042586e-02 -2.76625842e-01 2.34071106e-01 8.48191142e-01
4.72578138e-01 -9.78784487e-02 -1.34938598e-01 1.04939973e+00
-7.35762939e-02 1.06752448e-01 -3.94843280e-01 -2.37573326e-01
2.04006657e-01 1.30502248e+00 -8.67631376e-01 -5.73009439e-02
-3.95942897e-01 5.22961974e-01 3.24040204e-01 7.38329217e-02
-6.23854458e-01 -3.32602374e-02 5.14062464e-01 3.70433480e-01
-1.50516421e-01 -3.10665309e-01 -6.27537131e-01 -9.95263755e-01
-1.03133194e-01 -8.34252298e-01 2.81933814e-01 -7.06395209e-01
-1.43341517e+00 4.00495470e-01 1.55784503e-01 -6.68228924e-01
-3.61699194e-01 -7.12518394e-01 -4.53722894e-01 1.15804195e+00
-1.11409712e+00 -1.16941762e+00 -1.50534064e-01 2.97224820e-01
7.73618743e-02 -4.16896462e-01 8.51858139e-01 5.31046867e-01
-4.81364042e-01 3.17156017e-01 3.61553311e-01 -1.78410076e-02
1.00031114e+00 -1.13372755e+00 6.24901727e-02 6.67848349e-01
-4.58861468e-03 5.81372797e-01 5.07102549e-01 -6.28306091e-01
-1.06739187e+00 -9.75136459e-01 6.61136746e-01 -4.45665061e-01
7.82705128e-01 -3.51075470e-01 -1.02800775e+00 5.47806859e-01
-4.13237542e-01 -1.40281945e-01 9.70650017e-01 5.02323031e-01
-4.98325735e-01 -2.30608419e-01 -1.27216887e+00 4.86436963e-01
1.02827370e+00 -7.05651820e-01 -4.83849913e-01 4.15721714e-01
2.48816460e-01 -2.18357205e-01 -1.37102246e+00 4.10168052e-01
7.98207700e-01 -8.65935326e-01 1.01908767e+00 -8.18829954e-01
6.01575017e-01 1.08744577e-01 3.61391157e-02 -1.42948282e+00
-3.62038076e-01 -1.59064680e-01 8.71436000e-02 1.14222491e+00
4.00205672e-01 -5.43658853e-01 7.50723600e-01 8.46373856e-01
-1.97822496e-01 -8.27789605e-01 -9.33427572e-01 -6.14572644e-01
4.91054088e-01 -4.52342629e-01 5.96563637e-01 6.69929087e-01
-3.63231272e-01 3.03984404e-01 -1.48859590e-01 8.61583352e-02
7.35268772e-01 -2.45744195e-02 3.28495830e-01 -1.55648160e+00
-2.31342658e-01 -5.98432600e-01 -7.00680375e-01 3.27455029e-02
4.01048094e-01 -1.14473033e+00 -1.87216341e-01 -1.51338494e+00
6.47241831e-01 -6.88734770e-01 -3.87296349e-01 2.92392701e-01
-1.09300129e-01 3.06053519e-01 -1.50258737e-02 1.37701273e-01
-1.27876252e-01 1.99484095e-01 1.07611740e+00 -2.66400129e-01
4.68798615e-02 -1.75654948e-01 -8.40159118e-01 6.11127377e-01
8.36281180e-01 -4.66212064e-01 -5.78292072e-01 -4.94768828e-01
-1.28815025e-01 -7.56892264e-02 6.10590398e-01 -1.21010983e+00
-2.70284474e-01 1.12986714e-01 8.85082841e-01 6.93833595e-03
1.89150795e-01 -5.02919912e-01 4.60353613e-01 6.40042067e-01
-3.66403848e-01 -4.43374254e-02 -8.28104317e-02 4.09615040e-01
2.16714069e-01 -2.57218599e-01 9.13337290e-01 1.62850469e-02
-5.79263985e-01 5.51581025e-01 -4.30969149e-02 3.11453819e-01
6.95354164e-01 2.28353143e-02 -3.31223696e-01 -1.46989450e-01
-1.17749751e+00 -1.73569217e-01 4.19690549e-01 4.27626133e-01
3.38170826e-01 -1.04487801e+00 -8.72292817e-01 5.67230433e-02
1.78595603e-01 -2.54009992e-01 5.99505723e-01 1.21303976e+00
-1.49614811e-01 4.54150081e-01 -5.72338760e-01 -7.45759308e-01
-1.04062378e+00 4.59231049e-01 4.28386927e-01 -2.05291227e-01
-6.98650420e-01 4.80887741e-01 3.18385482e-01 -2.63472855e-01
-1.21350139e-01 -1.21458977e-01 -2.83249021e-02 1.63998440e-01
5.12602687e-01 2.05775440e-01 4.92554069e-01 -6.62198126e-01
-3.32420617e-01 3.04051876e-01 -2.57208705e-01 -3.82957682e-02
1.63500738e+00 -9.50188860e-02 -3.51232067e-02 3.67656678e-01
1.17439198e+00 -3.86611789e-01 -9.26594138e-01 -2.03873396e-01
1.87557682e-01 -2.73647845e-01 1.13423444e-01 -9.11457121e-01
-1.17045975e+00 9.01338756e-01 1.10940123e+00 -1.61380157e-01
1.08483708e+00 2.22345665e-01 5.80891490e-01 -9.79975238e-02
4.99846488e-01 -9.55913484e-01 -1.29668802e-01 1.26134619e-01
7.81735182e-01 -1.30722189e+00 8.94770175e-02 6.23193607e-02
-5.31736135e-01 9.46083307e-01 2.91074425e-01 1.06310673e-01
6.08557761e-01 -9.49622504e-03 -2.17993017e-02 -3.14893425e-01
-3.95363688e-01 5.41586056e-02 3.97230387e-01 1.08289206e+00
6.83794916e-01 2.02948958e-01 -4.83579010e-01 9.67608511e-01
-2.21609131e-01 2.45943472e-01 3.23299944e-01 6.22442007e-01
-1.04218416e-01 -1.36125910e+00 -1.49340108e-01 1.21525013e+00
-5.79633117e-01 -3.68637126e-03 -5.35760149e-02 6.40654564e-01
2.75867015e-01 3.99732888e-01 4.92485538e-02 -1.17115945e-01
9.83179808e-02 4.42932129e-01 7.31764615e-01 -6.73868179e-01
-1.48378477e-01 -1.39435112e-01 4.85253334e-02 -3.61918569e-01
-4.42760944e-01 -1.03507721e+00 -1.07854879e+00 7.63250813e-02
1.14786617e-01 -2.03601226e-01 8.51653337e-01 1.24373162e+00
5.89881122e-01 6.59144044e-01 2.05702767e-01 -8.29509497e-01
-7.48716354e-01 -8.69409561e-01 -7.54728854e-01 5.08915603e-01
-4.55988534e-02 -8.56287897e-01 1.78343564e-01 1.52175218e-01] | [14.251005172729492, -1.7838460206985474] |
aa517278-4f1f-4f50-8c5a-a83ee674260f | chinese-grammatical-error-diagnosis-using-2 | null | null | https://aclanthology.org/W16-4920 | https://aclanthology.org/W16-4920.pdf | Chinese Grammatical Error Diagnosis Using Single Word Embedding | Abstract Automatic grammatical error detection for Chinese has been a big challenge for NLP researchers. Due to the formal and strict grammar rules in Chinese, it is hard for foreign students to master Chinese. A computer-assisted learning tool which can automatically detect and correct Chinese grammatical errors is necessary for those foreign students. Some of the previous works have sought to identify Chinese grammatical errors using template- and learning-based methods. In contrast, this study introduced convolutional neural network (CNN) and long-short term memory (LSTM) for the shared task of Chinese Grammatical Error Diagnosis (CGED). Different from traditional word-based embedding, single word embedding was used as input of CNN and LSTM. The proposed single word embedding can capture both semantic and syntactic information to detect those four type grammatical error. In experimental evaluation, the recall and f1-score of our submitted results Run1 of the TOCFL testing data ranked the fourth place in all submissions in detection-level. | ['Xue-jie Zhang', 'Bo Peng', 'Jixian Zhang', 'Jinnan Yang', 'Jin Wang'] | 2016-12-01 | null | null | null | ws-2016-12 | ['grammatical-error-detection'] | ['natural-language-processing'] | [-1.37467504e-01 -7.52599388e-02 3.56365323e-01 -4.89612192e-01
-6.37395024e-01 -1.34977445e-01 -1.73790097e-01 6.92714214e-01
-7.97125518e-01 6.87431574e-01 9.84607041e-02 -5.76421797e-01
1.50785118e-01 -8.87105048e-01 -6.68099463e-01 -1.95907205e-01
1.62167341e-01 2.40866795e-01 1.27138823e-01 -3.06607336e-01
6.06887519e-01 9.11420882e-02 -1.27618003e+00 4.62570846e-01
1.72486687e+00 7.21619189e-01 7.14252412e-01 8.42212737e-01
-4.95221436e-01 5.71563900e-01 -8.57964993e-01 -5.77428699e-01
-5.20858467e-01 -5.25456250e-01 -1.08422792e+00 -6.17730260e-01
3.54014814e-01 6.82441564e-03 -2.91881952e-02 1.52115262e+00
8.06491315e-01 1.24412246e-01 3.23163003e-01 -5.41298866e-01
-1.36087954e+00 7.27363825e-01 2.51224101e-01 4.27627206e-01
3.65877777e-01 -2.44910568e-01 8.49113643e-01 -1.19330490e+00
4.18207318e-01 1.09708095e+00 7.60194302e-01 8.62354517e-01
-3.50302875e-01 -6.26633942e-01 1.22321106e-01 5.60290039e-01
-1.02121294e+00 7.99899474e-02 4.50536311e-01 -1.89989403e-01
1.60872757e+00 8.03135335e-02 5.02458394e-01 7.40482032e-01
6.08553588e-01 8.17625523e-01 7.79384792e-01 -1.14947677e+00
-2.42262393e-01 2.53349990e-01 5.76009393e-01 9.30300474e-01
4.53219712e-01 -7.18946457e-02 -3.93070310e-01 3.87488931e-01
4.19612229e-02 8.44108239e-02 -2.23499551e-01 8.02314758e-01
-7.31137574e-01 7.47619033e-01 3.42967182e-01 8.29517126e-01
1.32108688e-01 2.74771333e-01 3.89307618e-01 6.40697122e-01
5.88621259e-01 5.69878042e-01 -7.58507550e-01 -4.64231223e-01
-7.25646675e-01 1.95975885e-01 6.48860395e-01 1.05466986e+00
4.03432429e-01 2.48718828e-01 -3.17684591e-01 8.44665825e-01
4.43597287e-01 2.82429397e-01 1.07157040e+00 -1.87186137e-01
6.29885137e-01 9.06311214e-01 -3.53906482e-01 -1.19232345e+00
-2.73685932e-01 -5.96349597e-01 -5.87572694e-01 -2.47895457e-02
2.22810388e-01 -2.71656334e-01 -8.09580445e-01 1.39037693e+00
-1.88394278e-01 1.32969901e-01 7.30606467e-02 6.00471437e-01
1.37094438e+00 6.78260982e-01 2.51263857e-01 2.53667116e-01
1.21417010e+00 -1.05404103e+00 -9.70467091e-01 -1.29910365e-01
1.46809435e+00 -1.13374138e+00 1.10675931e+00 2.87918329e-01
-1.02197373e+00 -6.31791234e-01 -9.86152172e-01 -4.59509492e-01
-8.45123470e-01 4.26623434e-01 1.35664374e-01 8.63559186e-01
-9.06003654e-01 8.34345579e-01 -6.31147087e-01 -3.86256874e-01
1.00490071e-01 2.41461098e-01 -2.86300361e-01 -4.02331769e-01
-1.40398812e+00 1.07117367e+00 4.83665198e-01 3.14923495e-01
-2.54151762e-01 -6.66038215e-01 -9.76725280e-01 3.18487793e-01
-1.88932493e-01 -4.08531241e-02 1.16456938e+00 -8.62081349e-01
-1.27436435e+00 8.99807155e-01 -3.34756196e-01 -2.94858068e-01
5.72223179e-02 -3.74439329e-01 -6.01188719e-01 -3.75414163e-01
-1.32586405e-01 3.66430402e-01 2.03650922e-01 -5.30022025e-01
-6.89245701e-01 -4.13994849e-01 -4.20924991e-01 7.67144114e-02
-6.71206653e-01 4.81082529e-01 -1.03390560e-01 -7.49810874e-01
2.06052482e-01 -5.76865733e-01 5.93693256e-02 -5.45550346e-01
-1.85796544e-01 -1.03235209e+00 7.74437487e-01 -1.24330854e+00
1.89354861e+00 -2.00468826e+00 -9.67188105e-02 -1.02976546e-01
-5.73105812e-02 7.76401937e-01 -1.55610934e-01 3.49908590e-01
-4.83263619e-02 6.48142755e-01 4.98384982e-02 -4.76265579e-01
1.72733203e-01 1.17858402e-01 1.56907320e-01 9.82401296e-02
4.34882492e-01 8.82569313e-01 -1.01633513e+00 -4.77697790e-01
1.65397134e-02 2.23112032e-01 -5.52128911e-01 3.05063874e-01
-3.17330435e-02 1.17194660e-01 -2.77855933e-01 6.83486462e-01
8.04973185e-01 1.81276500e-01 3.10583250e-03 4.36536998e-01
-2.01524973e-01 7.27886260e-01 -1.08313251e+00 1.71525776e+00
-7.12460876e-01 5.53380668e-01 -4.78113145e-01 -8.70513856e-01
1.17190182e+00 5.82571208e-01 -3.18006307e-01 -9.16324139e-01
9.16357264e-02 7.24944353e-01 3.01596940e-01 -8.07599306e-01
5.70469379e-01 1.56703904e-01 -1.95675492e-01 2.14575008e-01
4.20216411e-01 1.96930513e-01 1.52248517e-01 7.70775676e-02
1.19348705e+00 9.81124565e-02 -8.76843557e-02 -2.70046204e-01
7.96307802e-01 -1.37220934e-01 8.84173930e-01 5.16877294e-01
-2.88799286e-01 6.81715488e-01 2.36861333e-01 -7.24065304e-01
-6.85566783e-01 -4.57324117e-01 -7.73305520e-02 1.02214587e+00
-3.07328820e-01 -5.73495805e-01 -8.60524535e-01 -8.52916718e-01
-7.93338418e-02 6.76648676e-01 -3.71791124e-01 -3.17173362e-01
-8.56930971e-01 -4.72861230e-01 8.44838738e-01 4.57162350e-01
6.10844493e-01 -1.50551498e+00 -2.94213027e-01 6.40038252e-01
-1.28803074e-01 -8.21678638e-01 -5.89871764e-01 1.98093921e-01
-7.74997532e-01 -1.11717176e+00 -5.32304704e-01 -1.61849833e+00
7.85227716e-01 -3.50858182e-01 1.16030025e+00 9.22742903e-01
-4.28860962e-01 -1.50734574e-01 -9.79426384e-01 -8.05108309e-01
-4.10206616e-01 1.61325663e-01 -3.05141956e-01 -4.62494910e-01
9.51058149e-01 -2.29489431e-02 -2.38427103e-01 -3.62570405e-01
-5.46539247e-01 -3.10128689e-01 3.51286680e-01 9.45094049e-01
2.64964432e-01 -8.57358575e-02 6.09113514e-01 -9.29217875e-01
1.00653827e+00 -2.15568349e-01 -4.52970296e-01 5.35986662e-01
-7.36874878e-01 1.22878589e-01 6.37755752e-01 1.11479193e-01
-9.92762804e-01 -3.50113064e-01 -7.12077022e-01 2.18007028e-01
-1.98411003e-01 7.93084264e-01 -8.23772177e-02 -1.46938324e-01
2.71281958e-01 2.75465012e-01 -3.56116354e-01 -8.59284937e-01
-3.06782126e-01 9.47172344e-01 5.71440272e-02 -6.12692237e-01
1.03913283e-03 -6.72087431e-01 -3.95623296e-01 -6.70473039e-01
-7.92507648e-01 -1.27451122e-01 -6.49213374e-01 -5.90999946e-02
1.14589596e+00 -7.60468125e-01 -6.05944276e-01 8.07899177e-01
-1.76828122e+00 3.57070155e-02 2.23239198e-01 5.46674907e-01
2.56935030e-01 4.71234381e-01 -8.47273827e-01 -6.14925921e-01
-6.19310319e-01 -1.20451057e+00 7.41820097e-01 2.77333826e-01
-1.00725271e-01 -1.20875216e+00 1.82441905e-01 5.81622906e-02
5.52195728e-01 4.04572114e-02 1.16659129e+00 -7.17629910e-01
-2.51722872e-01 -2.94927508e-01 -1.52358174e-01 8.78538013e-01
-2.49389604e-01 -8.75432342e-02 -7.27416158e-01 -1.26619935e-01
-1.48718715e-01 -2.83825308e-01 1.06699836e+00 2.76021987e-01
1.32718050e+00 -2.89065272e-01 -6.83080778e-02 5.68917036e-01
1.72407269e+00 3.13167155e-01 6.35640323e-01 4.44728613e-01
7.92480648e-01 3.26685816e-01 3.50515217e-01 -2.08312925e-03
4.17471200e-01 4.56293702e-01 8.03700238e-02 4.16279733e-01
-2.26228699e-01 -3.38602871e-01 5.05719900e-01 1.42017972e+00
6.97132200e-02 -4.25066888e-01 -1.20782220e+00 8.34330797e-01
-1.62585545e+00 -6.45299375e-01 -6.42687976e-01 1.89366448e+00
9.71014380e-01 -3.57840359e-02 -6.66260958e-01 2.51036435e-01
7.27588415e-01 -3.64925086e-01 2.08877251e-01 -1.40828717e+00
-4.36453931e-02 9.51261401e-01 2.23813951e-01 5.20475328e-01
-9.83784318e-01 1.20350122e+00 5.17220545e+00 9.68481421e-01
-1.09374452e+00 5.39275825e-01 3.63954782e-01 4.23792988e-01
-3.19684267e-01 -1.52619213e-01 -1.26527321e+00 8.09060991e-01
1.26051545e+00 1.13158822e-01 -5.12759648e-02 6.25921667e-01
7.43288249e-02 -1.02767898e-02 -7.20372736e-01 6.49859548e-01
2.11180329e-01 -1.30113673e+00 -2.79711038e-02 -2.75756031e-01
7.78426051e-01 -8.63078013e-02 -8.49027038e-02 6.55867815e-01
3.10830772e-01 -1.42197561e+00 3.77678901e-01 4.81079400e-01
6.48238599e-01 -9.71432984e-01 1.36663866e+00 4.03705895e-01
-9.85397816e-01 -1.65668294e-01 -8.69427025e-01 -5.33357859e-01
-2.53343999e-01 7.02608049e-01 -3.92198533e-01 4.23703671e-01
9.49880302e-01 8.54584932e-01 -8.10732841e-01 1.05140781e+00
-6.89610600e-01 8.54518056e-01 1.05854444e-01 -6.44175649e-01
2.90829271e-01 -2.27260496e-02 2.52739221e-01 1.60013008e+00
7.98229992e-01 -6.90258965e-02 8.54556262e-03 9.15844262e-01
-3.19127172e-01 3.96375507e-01 -5.50153136e-01 -1.54384792e-01
4.07023996e-01 8.97830367e-01 -2.83131152e-01 -7.20562786e-02
-3.16942841e-01 1.03327119e+00 7.44576573e-01 -8.47851411e-02
-4.37491834e-01 -1.19393051e+00 4.23359901e-01 -4.04818714e-01
3.09786648e-01 -3.30385715e-01 -5.16013026e-01 -1.25060737e+00
1.93073079e-01 -7.58078575e-01 2.89829820e-01 -3.53558362e-01
-1.06965697e+00 6.78808272e-01 -7.05145359e-01 -9.91905987e-01
2.41080940e-01 -9.88987923e-01 -1.10248280e+00 1.37446594e+00
-1.71795869e+00 -1.02540267e+00 -1.22717910e-01 2.71185279e-01
8.37458491e-01 -4.38798308e-01 1.20346570e+00 5.87024450e-01
-1.00275946e+00 1.16606033e+00 2.75970995e-01 6.46764278e-01
7.32101440e-01 -1.50565946e+00 1.72383875e-01 1.14883471e+00
6.68879598e-02 6.35869086e-01 2.96710432e-01 -7.52758443e-01
-1.04467225e+00 -1.15712011e+00 2.21188641e+00 -2.76852071e-01
1.59174427e-01 -1.74423218e-01 -1.08923709e+00 5.22641957e-01
4.25152689e-01 1.20851444e-02 4.71393079e-01 1.21579774e-01
1.09390631e-01 -1.17845077e-03 -1.02365220e+00 1.04376562e-02
6.78717792e-01 -2.20872253e-01 -7.56475925e-01 4.33456570e-01
9.95776594e-01 -7.02888429e-01 -8.40374708e-01 2.33344227e-01
1.24666251e-01 -6.96071565e-01 3.26015681e-01 -8.19051445e-01
8.23181987e-01 -3.94972012e-04 4.85151187e-02 -1.54505718e+00
-2.27882847e-01 -3.49558964e-02 2.66118705e-01 1.36196589e+00
6.47268176e-01 -4.42551672e-01 4.77525055e-01 3.94014210e-01
-9.90521729e-01 -9.01690125e-01 -9.17158484e-01 -7.63716578e-01
6.49846554e-01 -5.03856182e-01 4.93747622e-01 9.59103346e-01
8.26823935e-02 -1.41251922e-01 -1.44645050e-02 6.53870478e-02
3.94340698e-03 -2.86893249e-01 -3.02839302e-03 -1.03151178e+00
1.13921255e-01 -3.95200223e-01 -6.14046156e-01 -5.02269268e-01
4.24290240e-01 -1.34633994e+00 1.78570628e-01 -1.60875785e+00
-1.31305754e-01 -5.40981710e-01 -4.72241700e-01 4.73735839e-01
-5.35373271e-01 -1.69625223e-01 7.42350668e-02 -5.63644230e-01
-4.66356009e-01 4.88811731e-01 1.07003367e+00 -7.27474317e-02
1.59997314e-01 -2.50167251e-01 -5.27054429e-01 4.80988950e-01
7.95201898e-01 -8.69451821e-01 2.15986148e-01 -1.17949355e+00
6.75194383e-01 -3.64593595e-01 2.38400884e-02 -1.09970665e+00
3.61470819e-01 2.36966595e-01 4.27900463e-01 -4.62333441e-01
-5.51188648e-01 -5.97629309e-01 -6.75648570e-01 8.33319068e-01
-3.32130909e-01 6.85878575e-01 2.97107667e-01 2.41753206e-01
-6.46882057e-01 -8.92013729e-01 6.38303101e-01 -4.76152152e-01
-8.29231441e-01 5.79587482e-02 -4.88193750e-01 2.58283257e-01
6.73412740e-01 -7.72967041e-02 -2.49020576e-01 1.38867125e-02
-6.01162076e-01 2.92901158e-01 -1.56687796e-01 5.15564144e-01
1.17345905e+00 -1.29262650e+00 -9.61747766e-01 4.35057849e-01
1.80768609e-01 -5.10490686e-02 3.68490964e-01 6.68510854e-01
-1.07685399e+00 5.94338417e-01 -1.77037299e-01 -2.45221704e-01
-1.44025218e+00 4.15814221e-02 4.84757274e-01 -6.14073455e-01
-3.39990467e-01 1.49802721e+00 -8.36927593e-01 -1.06742060e+00
4.02173579e-01 -6.02983594e-01 -5.56546032e-01 -3.90049487e-01
6.54056549e-01 3.44640017e-01 6.99298143e-01 -3.13264608e-01
-3.22207779e-01 5.13841689e-01 -7.47108161e-02 3.46541762e-01
1.42496634e+00 1.10685125e-01 -4.81764048e-01 2.65312701e-01
1.31601393e+00 -1.44853201e-02 -9.42523777e-03 -1.95837229e-01
4.21500236e-01 -3.23412240e-01 1.06116548e-01 -9.93870974e-01
-8.59735012e-01 1.56018388e+00 8.70523930e-01 -2.68841296e-01
7.10768402e-01 -5.64690232e-01 1.25275791e+00 5.14734685e-01
1.54561698e-01 -1.43270612e+00 -1.03875935e-01 1.26793194e+00
7.16978788e-01 -1.52515805e+00 -5.35539329e-01 -1.49335265e-01
-2.27860004e-01 1.68874836e+00 1.11584079e+00 -3.10594827e-01
9.16073680e-01 -1.30936867e-02 -1.49414148e-02 -1.44224511e-02
-6.13184869e-01 2.54333392e-02 6.15660787e-01 2.30932549e-01
1.23751378e+00 6.98562637e-02 -9.99755383e-01 1.03781605e+00
-1.42644778e-01 -1.02309860e-01 6.00783110e-01 1.15464330e+00
-7.79400408e-01 -1.61134052e+00 -1.39854150e-03 4.31360543e-01
-6.96353495e-01 -4.90746528e-01 -2.73979902e-01 4.98854578e-01
6.65218413e-01 1.07567263e+00 1.71203181e-01 -4.33717936e-01
3.02222610e-01 5.10137618e-01 5.00410676e-01 -1.16924667e+00
-1.20525265e+00 -6.33982420e-01 -9.07204077e-02 -3.56823355e-01
6.13291282e-03 -4.15550947e-01 -1.46067035e+00 -2.60751009e-01
-5.54999530e-01 2.35511124e-01 8.18639934e-01 1.18206882e+00
3.81317258e-01 7.87161589e-01 1.50594950e-01 5.31284548e-02
-3.53963047e-01 -1.47104728e+00 -4.58274305e-01 4.09864932e-01
5.41337840e-02 -1.92579508e-01 -2.14963838e-01 -2.99361616e-01] | [11.046547889709473, 10.792856216430664] |
eabb4373-5a00-4a39-b046-495b4baff3a5 | fakeretouch-evading-deepfakes-detection-via | 2009.09213 | null | https://arxiv.org/abs/2009.09213v4 | https://arxiv.org/pdf/2009.09213v4.pdf | Dodging DeepFake Detection via Implicit Spatial-Domain Notch Filtering | The current high-fidelity generation and high-precision detection of DeepFake images are at an arms race. We believe that producing DeepFakes that are highly realistic and 'detection evasive' can serve the ultimate goal of improving future generation DeepFake detection capabilities. In this paper, we propose a simple yet powerful pipeline to reduce the artifact patterns of fake images without hurting image quality by performing implicit spatial-domain notch filtering. We first demonstrate that frequency-domain notch filtering, although famously shown to be effective in removing periodic noise in the spatial domain, is infeasible for our task at hand due to the manual designs required for the notch filters. We, therefore, resort to a learning-based approach to reproduce the notch filtering effects, but solely in the spatial domain. We adopt a combination of adding overwhelming spatial noise for breaking the periodic noise pattern and deep image filtering to reconstruct the noise-free fake images, and we name our method DeepNotch. Deep image filtering provides a specialized filter for each pixel in the noisy image, producing filtered images with high fidelity compared to their DeepFake counterparts. Moreover, we also use the semantic information of the image to generate an adversarial guidance map to add noise intelligently. Our large-scale evaluation on 3 representative state-of-the-art DeepFake detection methods (tested on 16 types of DeepFakes) has demonstrated that our technique significantly reduces the accuracy of these 3 fake image detection methods, 36.79% on average and up to 97.02% in the best case. | ['Geguang Pu', 'Yang Liu', 'Felix Juefei-Xu', 'Yihao Huang', 'Qing Guo'] | 2020-09-19 | null | null | null | null | ['fake-image-detection'] | ['computer-vision'] | [ 3.36881459e-01 -4.50153127e-02 2.89262980e-01 8.61930996e-02
-8.12240005e-01 -7.69611716e-01 6.85104489e-01 -5.15621662e-01
-3.14047545e-01 4.26166296e-01 3.40851098e-02 -4.69875187e-01
1.70347571e-01 -7.63878465e-01 -9.56693411e-01 -7.22897947e-01
1.95763618e-01 -1.61371186e-01 5.05988955e-01 -5.41957617e-01
3.29381078e-01 4.64811116e-01 -1.47339213e+00 5.72979093e-01
9.73657131e-01 9.90494549e-01 -6.14157692e-02 8.26048255e-01
2.98395991e-01 7.00050712e-01 -1.13456380e+00 -7.70878732e-01
6.17818654e-01 -3.44906718e-01 -4.44000304e-01 -3.34125638e-01
7.20928013e-01 -9.56320941e-01 -6.58666372e-01 1.21237457e+00
4.87624824e-01 -3.19107741e-01 5.17367423e-01 -1.11333692e+00
-5.74542880e-01 2.43992642e-01 -6.68075860e-01 1.64185360e-01
8.23905244e-02 6.42224312e-01 3.42124164e-01 -8.35692465e-01
4.90090340e-01 1.24555409e+00 1.16121840e+00 6.41656399e-01
-1.22276866e+00 -9.46727693e-01 -5.41070163e-01 9.90404263e-02
-1.24075854e+00 -5.04246175e-01 8.49169195e-01 -4.10761863e-01
7.04051912e-01 2.24437550e-01 4.96739775e-01 1.50520444e+00
3.56205821e-01 5.44317484e-01 1.25435424e+00 -4.77449596e-01
-6.80368170e-02 -3.54682691e-02 -3.14040989e-01 7.25055218e-01
4.39141661e-01 6.59061670e-01 -3.71249318e-01 -2.35660836e-01
8.37011218e-01 -3.59280080e-01 -2.81094551e-01 7.21802264e-02
-1.07289457e+00 9.39533174e-01 7.96702921e-01 1.65594667e-01
-2.15968654e-01 7.97533989e-01 3.06424528e-01 3.20280254e-01
3.59849691e-01 7.73893058e-01 -4.38263565e-02 1.95453223e-02
-1.14460945e+00 4.48643178e-01 4.83642578e-01 5.43885648e-01
6.12954080e-01 2.88830668e-01 -3.06641251e-01 5.81270814e-01
-9.50366072e-03 6.61691785e-01 1.70929074e-01 -1.00339937e+00
4.01083052e-01 9.49642956e-02 4.05657381e-01 -1.36715567e+00
-1.52563363e-01 -5.39732218e-01 -8.26585412e-01 6.39184713e-01
6.90246284e-01 -2.11649597e-01 -1.16218114e+00 1.43277824e+00
7.12098405e-02 4.39358771e-01 -8.61374065e-02 1.10283804e+00
4.89886701e-01 4.58529323e-01 -2.88442492e-01 2.13892400e-01
1.33374429e+00 -7.90304840e-01 -7.73905098e-01 -3.19778115e-01
5.59177458e-01 -1.02655995e+00 1.18347883e+00 6.41541600e-01
-7.04659879e-01 -4.60613877e-01 -1.38818371e+00 -1.28225952e-01
-3.40286881e-01 2.54679382e-01 4.96414751e-01 1.20492184e+00
-8.91760826e-01 6.38285935e-01 -4.47404414e-01 2.78128028e-01
8.86504531e-01 8.47244710e-02 -2.36230657e-01 -2.54668985e-02
-1.36867094e+00 1.01618683e+00 -1.39378048e-02 1.25383317e-01
-1.08202386e+00 -6.84119582e-01 -4.22343135e-01 -7.14581162e-02
4.01830614e-01 -5.31535983e-01 1.03609800e+00 -1.01283491e+00
-1.35756481e+00 9.04521585e-01 1.52186483e-01 -7.80730426e-01
1.07160282e+00 -2.96138674e-01 -4.29341197e-01 3.37815017e-01
-2.03186702e-02 5.68055570e-01 1.65965331e+00 -1.47553706e+00
-3.29132468e-01 -5.57418242e-02 2.54321452e-02 -2.94250816e-01
-3.76816183e-01 -1.17655188e-01 -2.58177787e-01 -1.17564607e+00
-5.50769605e-02 -9.08798575e-01 -9.39140618e-02 7.49315098e-02
-4.45695072e-01 4.77410436e-01 1.35573375e+00 -8.20522249e-01
1.14380407e+00 -2.31148839e+00 -4.44320142e-01 1.40971392e-01
5.04676461e-01 7.07874179e-01 -2.46004298e-01 2.26292670e-01
2.21725941e-01 2.28525132e-01 -2.51119822e-01 -3.61876190e-01
-1.80311769e-01 -8.27519223e-02 -7.80146897e-01 6.38268471e-01
2.67309457e-01 1.12946212e+00 -1.00036287e+00 1.42810103e-02
3.53257447e-01 5.32212138e-01 -7.34678745e-01 -1.20170549e-01
-1.18861184e-01 3.14742327e-01 -5.41460924e-02 3.87819886e-01
1.19133222e+00 1.14187002e-01 -2.00427890e-01 -5.24484575e-01
1.06057055e-01 2.99636405e-02 -7.47704804e-01 1.37561560e+00
-4.07181472e-01 8.99990797e-01 2.62704015e-01 -6.72803283e-01
7.89853454e-01 -8.25905427e-02 2.41176523e-02 -8.96546364e-01
2.89007992e-01 6.04172230e-01 1.67322606e-01 -4.36910421e-01
5.73407233e-01 -1.45707056e-01 -3.02303493e-01 3.19914848e-01
-8.50523338e-02 -5.43489873e-01 -4.25424993e-01 2.76845723e-01
1.26690197e+00 -8.43751803e-02 -3.51116091e-01 -2.85358042e-01
2.15863287e-01 1.74461901e-01 1.26805454e-01 1.40370810e+00
-3.92714441e-01 8.04711759e-01 5.15047371e-01 -5.47395110e-01
-1.25896811e+00 -1.04139137e+00 1.07955307e-01 5.60927570e-01
5.05753219e-01 -3.11413914e-01 -1.10203481e+00 -9.33460295e-01
1.62454143e-01 5.81836104e-01 -5.31714499e-01 -5.26507914e-01
-6.89802051e-01 -6.79602921e-01 1.44648457e+00 1.40227437e-01
9.73979712e-01 -6.96298480e-01 -5.99054456e-01 1.54916406e-01
-2.30927944e-01 -1.30335152e+00 -4.76631403e-01 -3.35868411e-02
-3.50833505e-01 -1.06344199e+00 -6.46313369e-01 -5.01598656e-01
3.72074991e-01 5.38669169e-01 8.00138235e-01 5.38646817e-01
-5.85937500e-01 -3.58826071e-01 -3.54145348e-01 -2.91245371e-01
-7.98793256e-01 -3.53850484e-01 -7.64698535e-02 -2.68261470e-02
-8.66197888e-03 -2.91723758e-01 -8.78905475e-01 4.54270899e-01
-1.03916085e+00 -4.78977710e-02 6.54565275e-01 1.19096220e+00
9.78862718e-02 3.81824285e-01 5.94833493e-01 -7.62709856e-01
7.77514040e-01 1.52844563e-03 -7.02439308e-01 -2.26040632e-01
-1.51536331e-01 -1.10689335e-01 6.89393938e-01 -6.28425419e-01
-9.18058157e-01 9.17298440e-03 -3.07534695e-01 -6.29820526e-01
8.86221901e-02 3.82806081e-03 4.81331088e-02 -6.64944112e-01
1.30684519e+00 2.31194943e-01 1.69716448e-01 -2.26439893e-01
4.91402149e-01 7.27484107e-01 9.69548464e-01 -2.94496447e-01
1.21404850e+00 9.59743619e-01 1.14554681e-01 -6.95266545e-01
-6.20770574e-01 -1.20618023e-01 -5.61353602e-02 -1.59357429e-01
5.61222911e-01 -8.96793783e-01 -7.03065217e-01 1.19139934e+00
-1.30169213e+00 -5.47501922e-01 5.87372594e-02 3.66352685e-02
-3.38036895e-01 6.13952756e-01 -6.65213764e-01 -5.82144797e-01
-2.72276968e-01 -1.38231885e+00 1.19182241e+00 -2.49491677e-01
1.02673613e-01 -5.22304654e-01 -4.01817948e-01 4.33339417e-01
6.81374848e-01 2.94943988e-01 5.56268156e-01 4.81320582e-02
-6.88275218e-01 -3.57242346e-01 -6.03327036e-01 6.62498832e-01
4.48990576e-02 -1.26212850e-01 -1.33847833e+00 -2.13378713e-01
4.60111462e-02 -2.81184673e-01 1.24610996e+00 2.32176602e-01
1.25621009e+00 -3.80635142e-01 -2.79655993e-01 7.66284704e-01
1.32295132e+00 -1.71385467e-01 1.24433398e+00 2.77887166e-01
8.71669531e-01 3.99206161e-01 5.50229490e-01 1.49880558e-01
-2.10937545e-01 1.01701260e+00 6.29894674e-01 -2.23002404e-01
-6.86711669e-01 -3.09542656e-01 2.85277814e-01 -1.32155448e-01
2.61107206e-01 -4.07844871e-01 -6.98167443e-01 4.65865076e-01
-1.59678149e+00 -1.12809384e+00 -5.66572189e-01 2.01061606e+00
6.22956753e-01 2.90444642e-01 7.10047176e-03 1.86336175e-01
6.16717577e-01 1.16292432e-01 -2.58810848e-01 -3.81095439e-01
-3.37018788e-01 2.92055756e-01 1.06991482e+00 6.16832674e-01
-1.36575198e+00 1.38136506e+00 6.43056154e+00 1.31861734e+00
-1.26110351e+00 2.56424278e-01 5.89934945e-01 4.21235301e-02
-2.64028728e-01 -1.74005389e-01 -4.28465217e-01 8.07438791e-01
7.03630328e-01 4.31354195e-01 6.06453955e-01 5.19706070e-01
4.90253001e-01 -1.09380469e-01 -4.16879445e-01 1.13050807e+00
8.15167557e-03 -1.78946733e+00 3.67629118e-02 1.95495650e-01
7.57075608e-01 -1.88478753e-01 4.17578399e-01 1.45581495e-02
2.47068048e-01 -1.20977187e+00 1.02042115e+00 2.51115292e-01
1.04867220e+00 -8.11649561e-01 7.51455486e-01 4.31535453e-01
-5.62783182e-01 -1.41079023e-01 -1.85698643e-01 -9.78897735e-02
9.17224139e-02 9.68417764e-01 -8.88426483e-01 2.00124830e-01
8.25235665e-01 1.98177934e-01 -3.94956529e-01 8.45595956e-01
-4.72721636e-01 5.74931383e-01 -4.21162665e-01 2.90179163e-01
2.67969936e-01 2.96012402e-01 5.76326311e-01 1.23507822e+00
4.36612874e-01 -1.60826072e-01 -2.39541307e-01 9.40914273e-01
-2.39563733e-01 -6.53675973e-01 -6.75883293e-01 7.57220089e-02
5.64716995e-01 1.01001322e+00 -7.14111924e-01 -2.38645047e-01
3.92654948e-02 1.42051208e+00 -9.39576924e-02 1.55578688e-01
-1.21479452e+00 -5.13084173e-01 9.12578464e-01 2.58119851e-01
3.81534666e-01 -1.46648765e-01 -7.18742967e-01 -1.16105556e+00
3.38699184e-02 -1.10222530e+00 -2.44523939e-02 -7.16703892e-01
-1.18339443e+00 4.41234261e-01 -3.86319041e-01 -1.23521173e+00
-6.14295416e-02 -7.17051446e-01 -4.25163597e-01 8.92220497e-01
-1.50211561e+00 -1.31250787e+00 -4.97080058e-01 5.60400903e-01
1.98068395e-01 8.73426092e-04 5.79387009e-01 4.09033269e-01
-7.78952390e-02 8.37577343e-01 3.60954665e-02 2.11168408e-01
8.32091510e-01 -8.46346021e-01 8.45860839e-01 1.14809561e+00
-5.57504073e-02 5.02431095e-01 8.52107882e-01 -8.73626590e-01
-1.47337067e+00 -1.00093305e+00 4.32075590e-01 -5.65628350e-01
7.00329840e-01 -7.22933471e-01 -8.65845323e-01 1.56915620e-01
-1.20871000e-01 2.56263256e-01 -7.37040713e-02 -6.70078397e-01
-7.06152141e-01 6.82825595e-02 -1.60014117e+00 6.46073520e-01
1.07695103e+00 -6.59860611e-01 -2.85116673e-01 1.66413724e-01
7.00347304e-01 -5.87304831e-01 -1.75175205e-01 4.40817684e-01
5.47308624e-01 -1.33212435e+00 1.32628751e+00 -8.25056359e-02
4.99144197e-01 -6.55234337e-01 1.55503824e-01 -1.35166109e+00
-2.76271343e-01 -8.43648314e-01 -9.87980217e-02 8.09965670e-01
2.28927620e-02 -6.87900424e-01 9.79423821e-01 -2.23227337e-01
-2.34655216e-01 -2.72342592e-01 -9.61160660e-01 -8.25126171e-01
3.28941755e-02 -5.40710449e-01 5.71840286e-01 9.09768820e-01
-5.49147606e-01 -1.81583643e-01 -8.02255213e-01 5.11218488e-01
7.23438859e-01 -2.25247785e-01 1.02759886e+00 -7.40948796e-01
-4.79574412e-01 -4.62456137e-01 -5.63413620e-01 -1.14199150e+00
-1.04758374e-01 -3.01261872e-01 1.80929407e-01 -9.45262790e-01
-2.01320827e-01 -4.01435316e-01 2.09544435e-01 3.75114024e-01
-2.00866088e-01 1.14519191e+00 8.39616060e-02 3.02240252e-01
8.16550553e-02 1.92438170e-01 1.31847584e+00 -3.32271099e-01
1.37975514e-01 -3.23657453e-01 -6.52095854e-01 7.67698824e-01
6.22106075e-01 -5.05534470e-01 -2.36663148e-01 -6.06376648e-01
6.27682507e-02 -4.05446500e-01 1.07166219e+00 -1.09199154e+00
-1.13849685e-01 1.12416476e-01 4.58799690e-01 -3.50663394e-01
4.23974335e-01 -6.00342929e-01 7.61028379e-02 8.13667655e-01
1.30439609e-01 -5.18562675e-01 3.90384346e-01 4.54733640e-01
-6.90167695e-02 -1.71443924e-01 1.18935084e+00 3.56420092e-02
-7.91577518e-01 -2.07029700e-01 -4.57812935e-01 -1.46646872e-01
8.25402200e-01 -3.17331612e-01 -9.50300395e-01 -6.08186364e-01
-2.59072393e-01 -4.15559739e-01 7.37018347e-01 3.19191277e-01
7.33301163e-01 -1.07042873e+00 -5.71364582e-01 4.79289860e-01
-6.08576052e-02 -4.08667654e-01 4.79756325e-01 5.70326030e-01
-9.64560330e-01 -6.75375164e-02 -1.35744765e-01 -4.58402902e-01
-1.09154820e+00 5.18655241e-01 5.05635440e-01 -1.15988269e-01
-7.93096066e-01 1.02777863e+00 1.91801608e-01 -3.34538221e-02
-1.27043232e-01 -1.00635357e-01 3.16781402e-01 -2.27903485e-01
6.18780911e-01 2.42070302e-01 3.50724518e-01 -6.38917089e-01
-3.08237791e-01 3.12043816e-01 3.52307074e-02 -5.62448911e-02
9.91979003e-01 6.87412396e-02 1.78963929e-01 -5.60468853e-01
1.22608626e+00 4.58218455e-01 -1.44460678e+00 1.99369520e-01
-2.67459422e-01 -9.65696871e-01 4.32814866e-01 -1.09108841e+00
-1.03278124e+00 7.59932041e-01 7.16463625e-01 2.62943447e-01
1.23529792e+00 -2.11744219e-01 1.06554043e+00 -1.68022644e-02
3.79507512e-01 -7.37516642e-01 4.83754426e-01 4.04996932e-01
8.05078804e-01 -9.78633225e-01 -1.21348463e-01 -7.47104704e-01
-1.64863780e-01 8.97054255e-01 3.76838893e-01 -4.86591339e-01
2.78240621e-01 5.68513989e-01 3.35397899e-01 -1.42229751e-01
-1.21667705e-01 5.28114401e-02 5.90359047e-02 7.47827709e-01
-2.97786206e-01 -5.21116599e-04 -1.52911514e-01 3.80485356e-01
-4.25310791e-01 -1.47262663e-01 7.26712167e-01 6.11996949e-01
-4.04424667e-01 -9.20368850e-01 -7.72153616e-01 3.35114002e-01
-6.17223501e-01 -1.71835959e-01 -5.78303218e-01 6.27906919e-01
4.03128088e-01 1.15379786e+00 -1.55247375e-01 -6.16427004e-01
3.27364385e-01 -5.09695172e-01 5.70781946e-01 -2.05391675e-01
-6.45178556e-01 7.38912448e-02 9.78518277e-02 -8.34981203e-01
1.45454049e-01 -1.56275585e-01 -8.75804543e-01 -6.77383363e-01
-4.57646430e-01 -3.23634923e-01 7.07269430e-01 7.87121177e-01
4.50840145e-01 3.80616933e-01 5.39933383e-01 -1.15133393e+00
-6.96365952e-01 -6.56429827e-01 -2.63299465e-01 6.47291541e-01
7.13683128e-01 -7.24657655e-01 -8.17240834e-01 -1.56906500e-01] | [12.46097183227539, 1.0605998039245605] |
fd3c6d48-f149-4434-98d1-107d285cf9b6 | unsupervised-collaborative-learning-of | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Sheng_Unsupervised_Collaborative_Learning_of_Keyframe_Detection_and_Visual_Odometry_Towards_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Sheng_Unsupervised_Collaborative_Learning_of_Keyframe_Detection_and_Visual_Odometry_Towards_ICCV_2019_paper.pdf | Unsupervised Collaborative Learning of Keyframe Detection and Visual Odometry Towards Monocular Deep SLAM | In this paper we tackle the joint learning problem of keyframe detection and visual odometry towards monocular visual SLAM systems. As an important task in visual SLAM, keyframe selection helps efficient camera relocalization and effective augmentation of visual odometry. To benefit from it, we first present a deep network design for the keyframe selection, which is able to reliably detect keyframes and localize new frames, then an end-to-end unsupervised deep framework further proposed for simultaneously learning the keyframe selection and the visual odometry tasks. As far as we know, it is the first work to jointly optimize these two complementary tasks in a single deep framework. To make the two tasks facilitate each other in the learning, a collaborative optimization loss based on both geometric and visual metrics is proposed. Extensive experiments on publicly available datasets (i.e. KITTI raw dataset and its odometry split) clearly demonstrate the effectiveness of the proposed approach, and new state-of-the-art results are established on the unsupervised depth and pose estimation from monocular videos.
| [' Xiaogang Wang', ' Wanli Ouyang', ' Dan Xu', 'Lu Sheng'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['camera-relocalization'] | ['computer-vision'] | [-2.88159758e-01 -2.38625444e-02 -3.53206307e-01 -3.06745082e-01
-4.74391818e-01 -2.46055052e-01 5.64469099e-01 -1.50202528e-01
-6.36375844e-01 7.05741942e-01 1.24191016e-01 4.00890261e-02
5.84610552e-02 -2.75839299e-01 -9.23275888e-01 -5.45153737e-01
-1.18739247e-01 5.07606626e-01 9.54501778e-02 7.85715356e-02
3.72822970e-01 6.13089561e-01 -1.55986154e+00 -5.34402430e-01
6.73449218e-01 1.00305116e+00 6.41221523e-01 6.01407766e-01
3.18981856e-01 1.22275734e+00 -8.49001929e-02 -2.48319516e-03
5.12578547e-01 -5.22119217e-02 -7.07762301e-01 4.18958247e-01
8.60448837e-01 -5.74682295e-01 -7.94127941e-01 1.03152561e+00
6.75913870e-01 3.47219020e-01 3.20693612e-01 -1.38217819e+00
-2.52444059e-01 1.55118644e-01 -4.41849798e-01 -1.32586202e-02
5.84709525e-01 1.99364558e-01 1.11015379e+00 -1.18707192e+00
9.54809725e-01 1.07420254e+00 4.88418519e-01 1.01071216e-01
-8.85769486e-01 -5.57084143e-01 2.82408327e-01 3.21092039e-01
-1.56581414e+00 -5.96595764e-01 8.32220614e-01 -6.31920516e-01
9.91471410e-01 -4.31826651e-01 9.12095129e-01 6.92046106e-01
2.95228183e-01 9.50711250e-01 7.54524708e-01 -3.26550007e-01
-2.17275135e-02 -2.78412234e-02 -3.23194265e-01 1.11960793e+00
1.56359911e-01 2.52141118e-01 -9.05430615e-01 3.41318607e-01
8.59816790e-01 1.84546575e-01 -4.73163635e-01 -1.28861296e+00
-1.47745311e+00 8.06030989e-01 8.00413072e-01 -2.29325369e-01
-3.06169420e-01 5.70690572e-01 2.55527496e-01 2.13459939e-01
3.97370994e-01 3.55978072e-01 -2.39264786e-01 -5.95625602e-02
-9.45478857e-01 2.75817007e-01 5.14832616e-01 1.16957498e+00
1.21816337e+00 8.26424733e-02 2.05539212e-01 4.52533573e-01
6.67957485e-01 6.48727715e-01 2.60341853e-01 -1.06863236e+00
4.86146092e-01 5.25330901e-01 2.80353338e-01 -1.01530170e+00
-6.73892140e-01 -2.94980168e-01 -6.12389982e-01 2.48978689e-01
9.35758576e-02 -1.88178699e-02 -9.30821061e-01 1.64760780e+00
4.65710431e-01 4.41786081e-01 -9.49266478e-02 1.29368174e+00
7.86119401e-01 2.91678995e-01 -3.02766502e-01 -1.09261274e-01
1.01197326e+00 -1.18005538e+00 -7.84878194e-01 -6.31769598e-01
4.98277545e-01 -8.70191038e-01 5.21742761e-01 1.64513111e-01
-9.85799491e-01 -6.93721175e-01 -1.49553919e+00 -5.80676377e-01
-4.13831472e-02 3.31126660e-01 8.43748152e-01 -7.35641196e-02
-1.18424344e+00 1.21468335e-01 -1.08445513e+00 -5.28798342e-01
2.64761716e-01 4.61829215e-01 -5.30039966e-01 -2.55849779e-01
-9.92991030e-01 1.06292045e+00 5.79075396e-01 8.35601687e-02
-1.28205240e+00 -2.44235277e-01 -1.37846732e+00 -2.77405471e-01
5.11511922e-01 -1.16456378e+00 1.13450420e+00 -5.24658501e-01
-1.52226305e+00 1.08193839e+00 -2.94947743e-01 -7.54958034e-01
5.87324321e-01 -6.02703154e-01 1.83734953e-01 2.22272635e-01
3.54319602e-01 1.02946496e+00 9.15327311e-01 -1.30027461e+00
-1.01216292e+00 -4.12773073e-01 1.97137415e-01 6.46285951e-01
1.00597650e-01 -4.35836613e-01 -8.41049492e-01 -2.89879829e-01
5.85911095e-01 -1.19218612e+00 -2.30139762e-01 2.06470162e-01
-4.72232521e-01 6.19072542e-02 5.39783597e-01 -4.00829792e-01
8.55826378e-01 -1.79463923e+00 6.51875138e-01 -1.23776302e-01
4.88375038e-01 -1.11860506e-01 2.95441359e-01 2.06528455e-01
2.28005484e-01 -7.28039205e-01 1.72586232e-01 -1.04092503e+00
5.55078238e-02 1.80277437e-01 -1.55523002e-01 1.19945443e+00
-4.90048937e-02 8.78252685e-01 -1.02883971e+00 -3.12043160e-01
7.22706974e-01 4.13162619e-01 -4.67634529e-01 4.31652188e-01
-8.71980265e-02 6.55434787e-01 -4.37366851e-02 8.31087053e-01
6.18198872e-01 -2.95611799e-01 -8.06108788e-02 -3.44794154e-01
-2.68968612e-01 2.94364899e-01 -1.27384877e+00 2.51202917e+00
-4.16091412e-01 9.25614238e-01 3.29514071e-02 -7.64529288e-01
8.20575297e-01 -7.23695830e-02 5.70821524e-01 -5.78326881e-01
3.34984481e-01 3.79755527e-01 -4.35883969e-01 -2.76614726e-01
9.95354176e-01 7.98285529e-02 -2.70127729e-02 7.13188052e-02
4.86283571e-01 -3.32659006e-01 1.16734065e-01 2.42008671e-01
7.04667926e-01 4.37719375e-01 7.48178244e-01 -7.34545514e-02
6.54080391e-01 -4.69350219e-02 5.37146330e-01 5.53186715e-01
-3.28326494e-01 5.15742302e-01 2.27708910e-02 -6.58286214e-01
-1.09252274e+00 -9.79313374e-01 7.22430050e-02 6.40106201e-01
8.64198267e-01 -3.57002318e-01 -1.40999243e-01 -5.51451147e-01
2.33906597e-01 6.88758194e-02 -3.30035478e-01 -6.02106899e-02
-5.04847646e-01 -1.05326153e-01 1.62183329e-01 3.35099310e-01
4.42139804e-01 -6.79017305e-01 -7.34820485e-01 1.11684240e-01
-2.30024010e-01 -1.59707212e+00 -4.78556097e-01 4.02852267e-01
-5.43712676e-01 -1.21091270e+00 -6.51258230e-01 -9.71338451e-01
5.10063291e-01 7.36914814e-01 1.05705059e+00 -3.14294137e-02
-1.33164495e-01 3.77335906e-01 -1.74097970e-01 -1.80560172e-01
4.15563304e-03 1.80348545e-01 3.39047611e-01 -7.96276107e-02
2.33308882e-01 -5.69605768e-01 -7.83727586e-01 1.55558571e-01
-5.00995517e-01 2.95651644e-01 5.50089478e-01 7.74457693e-01
6.40952349e-01 -5.54526865e-01 -1.29744306e-01 -2.81451851e-01
-9.26850364e-02 -2.50615180e-01 -1.00452209e+00 -2.02340990e-01
-5.58470190e-01 1.00555971e-01 2.07700148e-01 1.39415205e-01
-6.19643390e-01 6.08770192e-01 4.62155044e-03 -7.23140240e-01
9.66487303e-02 3.80997002e-01 -3.84163558e-02 -5.74170887e-01
3.71873915e-01 3.17671388e-01 2.45502088e-02 -2.05137059e-01
5.16678095e-01 4.27344799e-01 7.48414278e-01 -4.37540077e-02
1.07723486e+00 9.87446725e-01 2.13282049e-01 -7.81822979e-01
-8.80198658e-01 -1.01122081e+00 -9.66071308e-01 -1.14674814e-01
1.06160021e+00 -1.75299847e+00 -8.24082136e-01 5.69420636e-01
-1.33210051e+00 -1.40195668e-01 2.88761370e-02 8.21349800e-01
-9.15852189e-01 6.42529190e-01 -3.68601978e-01 -6.97070658e-01
-1.83445513e-01 -1.45396984e+00 1.55288911e+00 8.49024355e-02
9.46872234e-02 -1.07105052e+00 3.37795585e-01 1.27361283e-01
-2.40304157e-01 2.15146437e-01 1.62101030e-01 -1.18481673e-01
-1.38142526e+00 -2.54107006e-02 -2.19700009e-01 2.08098397e-01
-9.84111801e-02 -4.27125722e-01 -9.37913537e-01 -7.53021955e-01
-1.56835746e-02 -5.51630318e-01 1.10680413e+00 3.48321855e-01
4.16862994e-01 1.50884897e-01 -3.75619560e-01 1.41679883e+00
1.67980909e+00 7.53617659e-02 4.84347254e-01 8.06486964e-01
1.26847267e+00 2.01008126e-01 1.01673961e+00 6.14876091e-01
9.15342093e-01 1.02605712e+00 9.47976649e-01 -2.19873160e-01
-4.24091183e-02 -4.19192672e-01 3.95749539e-01 7.48374879e-01
2.83298314e-01 -1.35984004e-01 -7.95526505e-01 6.72341347e-01
-2.09656692e+00 -5.85707188e-01 2.56302267e-01 2.06126738e+00
3.80232066e-01 4.52572778e-02 -2.22612709e-01 -1.85846061e-01
5.19836962e-01 6.32014751e-01 -5.51784098e-01 2.72437990e-01
-1.88936427e-01 -2.64127791e-01 8.17897022e-01 8.38393033e-01
-1.55943227e+00 1.32948768e+00 4.92599106e+00 2.67894328e-01
-1.16370940e+00 5.36484569e-02 -1.34618860e-02 -1.55093044e-01
3.23260464e-02 3.07851225e-01 -1.05773938e+00 1.58166975e-01
1.98397130e-01 1.31085321e-01 5.06769896e-01 9.23175871e-01
1.91449925e-01 -3.49096924e-01 -1.19643700e+00 1.62192404e+00
3.37030351e-01 -1.38153708e+00 -2.30917633e-01 1.44917443e-01
9.58881855e-01 5.76707780e-01 -1.63023695e-01 1.94163080e-02
3.30067992e-01 -7.26096511e-01 1.08816600e+00 4.53816980e-01
6.84076965e-01 -7.88354456e-01 8.02780986e-01 2.21956894e-01
-1.52487040e+00 -1.15948528e-01 -5.17459929e-01 -2.75801450e-01
6.34076893e-01 5.72595894e-01 -8.47746670e-01 8.14122438e-01
5.43831348e-01 1.33061266e+00 -4.68797356e-01 1.23407638e+00
-6.59417570e-01 -2.95844436e-01 -2.52061844e-01 3.39319795e-01
3.73967826e-01 -1.25408836e-03 7.88639963e-01 7.72790194e-01
4.22729194e-01 -4.08766240e-01 6.39036894e-01 6.02723658e-01
-4.15503569e-02 -1.71004862e-01 -8.51272583e-01 2.82587796e-01
5.19669592e-01 1.21954572e+00 -4.73996699e-01 -2.08341673e-01
-4.49194342e-01 1.21871996e+00 5.50567150e-01 3.21626157e-01
-7.56494105e-01 -2.10424215e-01 1.03059006e+00 -1.52260035e-01
3.46195221e-01 -6.92278445e-01 -1.08659137e-02 -1.54332066e+00
2.59088632e-02 -5.82376719e-01 5.30381575e-02 -8.61123860e-01
-7.72409320e-01 5.34854650e-01 -2.71381855e-01 -1.54298329e+00
-4.93963987e-01 -5.32992840e-01 -6.44120527e-03 7.73885667e-01
-1.90854597e+00 -1.30856669e+00 -7.95505106e-01 5.73396683e-01
4.95925039e-01 -2.38413692e-01 2.06507146e-01 4.01121616e-01
-1.36523396e-01 2.73349136e-01 9.96714383e-02 1.47940889e-02
9.33629572e-01 -1.33303607e+00 6.15351021e-01 1.19989741e+00
4.02628601e-01 4.82671142e-01 6.94811165e-01 -5.81205726e-01
-1.66317868e+00 -1.02034891e+00 9.29797292e-01 -5.77109635e-01
5.26990950e-01 -4.59246129e-01 -3.21095288e-01 7.97515273e-01
-1.96097773e-02 6.74078986e-02 -2.77802385e-02 -3.05580974e-01
1.89104423e-01 -7.84674734e-02 -5.61716437e-01 5.09779692e-01
1.33727503e+00 -7.76241064e-01 -3.09191972e-01 5.33173025e-01
8.43674183e-01 -1.07649875e+00 -5.22278309e-01 3.92048627e-01
4.43279654e-01 -1.21289694e+00 1.18923366e+00 -4.05178145e-02
1.01487510e-01 -7.10413754e-01 -3.24443310e-01 -1.19364130e+00
-2.16257706e-01 -7.53501952e-01 -4.37690735e-01 8.30718040e-01
-2.05676064e-01 -3.15990090e-01 8.58989656e-01 -1.96350157e-01
-3.61428291e-01 -5.97920179e-01 -1.01839077e+00 -5.09016812e-01
-6.97210371e-01 -4.75562453e-01 2.96106011e-01 6.91653609e-01
-3.48893642e-01 5.34969926e-01 -9.96597648e-01 4.88553911e-01
7.44047761e-01 1.28776163e-01 1.40105057e+00 -1.09734070e+00
-1.94426879e-01 -4.99056242e-02 -9.66300607e-01 -1.87133527e+00
3.64593208e-01 -7.90844858e-01 3.87901783e-01 -1.62160969e+00
3.84398438e-02 -1.38656512e-01 2.43231449e-02 4.06773053e-02
-1.84489846e-01 2.03058764e-01 4.20892566e-01 5.08399844e-01
-9.40306783e-01 8.98257792e-01 1.16578376e+00 -5.69444224e-02
-2.41758078e-01 -1.81620836e-01 -2.09742352e-01 8.43556941e-01
1.14145488e-01 -3.73660862e-01 -5.30861557e-01 -6.53740942e-01
3.27039331e-01 2.09997162e-01 6.39736831e-01 -1.26077473e+00
5.17987251e-01 5.73041514e-02 3.01431030e-01 -8.81495595e-01
5.90060174e-01 -8.94500732e-01 -1.47524223e-01 5.28321326e-01
9.34014693e-02 1.86136618e-01 -1.51996613e-01 8.24272096e-01
-4.51917917e-01 1.62875608e-01 7.97192454e-01 -1.82953820e-01
-1.51120424e+00 7.15251803e-01 4.72746380e-02 -4.38650958e-02
9.61019337e-01 -1.90874472e-01 -1.70926630e-01 -6.56468332e-01
-4.74295646e-01 4.36863780e-01 8.67389023e-01 5.58419466e-01
8.00191164e-01 -1.44445956e+00 -3.43081027e-01 3.15851718e-01
4.22608018e-01 3.81199211e-01 1.42102703e-01 1.18568957e+00
-1.03475893e+00 6.80819929e-01 -2.48573631e-01 -1.10543430e+00
-1.07126713e+00 5.11243463e-01 3.65429670e-01 -1.57222643e-01
-5.79998851e-01 1.02327049e+00 3.93674195e-01 -4.71899718e-01
5.29271781e-01 -3.89796495e-01 -1.37421176e-01 -3.96953784e-02
3.05415899e-01 2.73927897e-01 -1.05588436e-02 -9.73672330e-01
-5.50138116e-01 7.81490266e-01 1.04270540e-01 -1.26159593e-01
1.21507478e+00 -8.67293358e-01 -1.07384965e-01 3.97108346e-01
1.41334176e+00 -1.39060795e-01 -1.76143599e+00 -5.19706607e-01
-1.71243295e-01 -7.94461548e-01 1.13590680e-01 -9.19618681e-02
-8.00616264e-01 8.24043691e-01 7.23866284e-01 -4.00094420e-01
9.37068880e-01 -3.23144533e-02 8.53839815e-01 6.59203887e-01
7.16251135e-01 -9.12712157e-01 3.03953648e-01 9.34404314e-01
5.87582648e-01 -1.65304852e+00 2.77718186e-01 -1.90554857e-01
-4.61397797e-01 9.90801811e-01 6.81897044e-01 -2.55606532e-01
2.81041265e-01 -4.24787290e-02 1.74436405e-01 -2.23871708e-01
-5.42177618e-01 -6.37120962e-01 3.51897836e-01 5.07551968e-01
8.50583762e-02 -3.19268048e-01 4.28824648e-02 -1.25445247e-01
-1.63029090e-01 -7.19076917e-02 3.99345547e-01 1.00584364e+00
-6.33871675e-01 -8.01230311e-01 -1.96893662e-01 -1.72782525e-01
-1.26382887e-01 -5.76742962e-02 -3.21622461e-01 1.03929555e+00
1.90575540e-01 5.82678556e-01 -1.27597330e-02 -4.63169932e-01
1.56086043e-01 -2.82545120e-01 6.27276361e-01 -5.65317929e-01
-3.78612317e-02 4.71675768e-02 -8.70145112e-02 -8.49265099e-01
-6.39007866e-01 -3.63563031e-01 -1.17925227e+00 -2.88837165e-01
-2.34104052e-01 -2.59140670e-01 6.68378294e-01 1.08122194e+00
1.47592738e-01 3.78697395e-01 5.28268874e-01 -1.52759707e+00
-4.03815567e-01 -6.56531990e-01 -5.92146337e-01 3.38991314e-01
8.52474868e-01 -1.00205493e+00 -3.46610248e-01 -3.69151086e-02] | [8.019814491271973, -2.17936635017395] |
a71325ef-932b-4acd-b0fc-586bcf9e05cb | generating-image-descriptions-using | null | null | https://aclanthology.org/W17-4750 | https://aclanthology.org/W17-4750.pdf | Generating Image Descriptions using Multilingual Data | null | ['Alan Jaffe'] | 2017-09-01 | null | null | null | ws-2017-9 | ['multimodal-machine-translation'] | ['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] | [-7.255665302276611, 3.7011587619781494] |
9455420e-0aa0-4676-a787-0640e7e243a1 | lrelu-piece-wise-linear-activation-functions | 1910.12259 | null | https://arxiv.org/abs/1910.12259v1 | https://arxiv.org/pdf/1910.12259v1.pdf | L*ReLU: Piece-wise Linear Activation Functions for Deep Fine-grained Visual Categorization | Deep neural networks paved the way for significant improvements in image visual categorization during the last years. However, even though the tasks are highly varying, differing in complexity and difficulty, existing solutions mostly build on the same architectural decisions. This also applies to the selection of activation functions (AFs), where most approaches build on Rectified Linear Units (ReLUs). In this paper, however, we show that the choice of a proper AF has a significant impact on the classification accuracy, in particular, if fine, subtle details are of relevance. Therefore, we propose to model the degree of absence and the presence of features via the AF by using piece-wise linear functions, which we refer to as L*ReLU. In this way, we can ensure the required properties, while still inheriting the benefits in terms of computational efficiency from ReLUs. We demonstrate our approach for the task of Fine-grained Visual Categorization (FGVC), running experiments on seven different benchmark datasets. The results do not only demonstrate superior results but also that for different tasks, having different characteristics, different AFs are selected. | ['Peter M. Roth', 'Mina Basirat'] | 2019-10-27 | null | null | null | null | ['fine-grained-visual-categorization'] | ['computer-vision'] | [ 1.81581363e-01 -3.37833256e-01 -5.94005994e-02 -4.32019234e-01
-4.97738980e-02 -7.06432998e-01 8.47012758e-01 3.51431847e-01
-5.39044023e-01 5.12352526e-01 -7.48491287e-03 -1.96280554e-01
-5.09368539e-01 -7.70758629e-01 -6.51716650e-01 -8.74478459e-01
-3.23501788e-02 -3.67189124e-02 3.87164772e-01 -2.96656281e-01
5.33018112e-01 7.86971569e-01 -1.94899142e+00 3.37583780e-01
7.00954854e-01 1.29013646e+00 1.00523837e-01 3.57733190e-01
-2.32309505e-01 8.52023125e-01 -6.24538839e-01 -1.84563652e-01
2.38055766e-01 -1.50982395e-01 -8.27911675e-01 -7.12035298e-02
3.62418085e-01 5.84433554e-03 2.59557478e-02 9.25288737e-01
3.76810044e-01 1.60126220e-02 9.50677633e-01 -1.21000540e+00
-5.83783209e-01 5.49802899e-01 -4.76209670e-01 2.14133263e-01
1.63008329e-02 -2.71583404e-02 1.00474322e+00 -7.31734335e-01
4.80010837e-01 1.16352367e+00 6.13208234e-01 3.05716485e-01
-1.31691158e+00 -3.62556875e-01 5.64105034e-01 3.99498373e-01
-1.59544516e+00 -3.71737659e-01 5.08412600e-01 -6.82525456e-01
7.12973356e-01 1.90021127e-01 2.68176436e-01 8.10761034e-01
1.23707488e-01 3.58169556e-01 1.28558671e+00 -5.69822311e-01
2.36948416e-01 3.46848279e-01 4.03758436e-01 4.37968045e-01
3.45563054e-01 -1.99595064e-01 -1.13138817e-01 1.05077289e-01
4.57167834e-01 1.00502120e-02 -5.44608295e-01 -7.26313055e-01
-9.76032436e-01 8.19590569e-01 7.11087644e-01 8.14892054e-01
-2.08925381e-01 -3.26575488e-02 4.90586758e-01 4.45015997e-01
1.54763207e-01 4.10928071e-01 -2.23832190e-01 2.60911614e-01
-6.01889193e-01 -2.86701624e-03 4.23894256e-01 3.38009089e-01
7.99831808e-01 -1.51130661e-01 -4.77926254e-01 7.92234361e-01
2.24308874e-02 -2.92909741e-02 7.52264142e-01 -5.25985599e-01
9.66348425e-02 8.44388723e-01 9.74192992e-02 -1.17070067e+00
-6.20499253e-01 -7.22598255e-01 -9.82665718e-01 6.42733455e-01
4.60790902e-01 1.49942577e-01 -1.01723766e+00 1.92798746e+00
5.78020737e-02 -2.34479710e-01 -9.59245488e-02 9.30092931e-01
6.57514989e-01 4.27541971e-01 2.77379960e-01 6.54716864e-02
1.44988334e+00 -7.50045002e-01 -4.55891460e-01 -1.68516368e-01
4.39119101e-01 -6.56281173e-01 1.28428364e+00 5.10978997e-01
-8.02768469e-01 -8.78093004e-01 -1.16570425e+00 3.00911348e-02
-7.90437937e-01 3.32012206e-01 4.61817116e-01 6.50461078e-01
-1.25187850e+00 7.35771358e-01 -4.14953083e-01 -4.36027557e-01
2.26059332e-01 4.72687304e-01 -5.89974344e-01 1.78105250e-01
-9.65741515e-01 9.74375188e-01 4.98078644e-01 2.00720072e-01
-6.57388389e-01 -4.88632590e-01 -4.82409418e-01 4.91353542e-01
3.02525580e-01 -5.12696207e-01 7.96100140e-01 -1.30024672e+00
-1.13608968e+00 7.52203524e-01 1.74543168e-02 -5.64669490e-01
7.55523264e-01 -9.40598994e-02 -3.82009037e-02 -6.63388893e-02
-2.06104994e-01 7.38272429e-01 1.02701902e+00 -1.40325761e+00
-7.69750178e-01 -2.39019588e-01 5.42966425e-01 -1.42873973e-01
-6.49944961e-01 -9.42558199e-02 -2.05696076e-01 -5.32306790e-01
-1.75884157e-01 -8.28577042e-01 -1.13539889e-01 1.43299568e-02
-2.37423643e-01 -4.94600862e-01 6.69434190e-01 -2.89506465e-01
1.21998096e+00 -2.44498253e+00 3.39730203e-01 1.85440034e-01
2.50517249e-01 4.38861489e-01 -2.65434384e-02 2.92782724e-01
-8.82224366e-02 2.27997109e-01 -3.04994881e-01 -1.34738997e-01
7.86319077e-02 1.91598926e-02 -2.07861230e-01 4.08215940e-01
2.79390752e-01 7.37625480e-01 -3.61132175e-01 -3.37371618e-01
3.09220314e-01 5.88433623e-01 -4.19610590e-01 1.20073393e-01
5.22314608e-02 2.70072520e-01 -3.83287698e-01 1.57881454e-01
6.59787297e-01 -3.12066376e-01 1.13382474e-01 -3.41134161e-01
-3.03867519e-01 -1.18247040e-01 -1.23348963e+00 1.28169024e+00
-6.71896577e-01 5.08768439e-01 1.09798841e-01 -1.26573944e+00
9.19077814e-01 2.96154432e-02 2.26515532e-01 -8.46774399e-01
3.86604488e-01 2.51136333e-01 3.63049954e-01 -1.30907357e-01
4.65567231e-01 1.07207946e-01 4.00354676e-02 4.14564997e-01
6.92228507e-03 3.72615963e-01 4.68376726e-01 -6.62012026e-02
7.54668534e-01 1.35486871e-01 5.25657237e-01 -5.79270005e-01
9.63198781e-01 -1.49417624e-01 1.34075657e-01 7.69603252e-01
-2.27741092e-01 6.54727101e-01 6.61000311e-01 -3.57177526e-01
-8.19977462e-01 -7.66161323e-01 -3.28338355e-01 1.17744660e+00
2.84518361e-01 -2.23982319e-01 -8.61700177e-01 -6.62795603e-01
-1.43818883e-02 3.57795715e-01 -9.36113596e-01 -1.91941813e-01
-6.16411269e-01 -8.55401397e-01 3.91932636e-01 4.57877398e-01
3.13684076e-01 -1.25243950e+00 -9.23400044e-01 2.15671845e-02
1.32640079e-01 -1.20255458e+00 -1.28924474e-01 5.65411389e-01
-5.32882035e-01 -1.06713235e+00 -8.66111636e-01 -6.86576188e-01
7.04886436e-01 5.94605982e-01 1.09860253e+00 3.23797077e-01
-2.88477182e-01 1.99735746e-01 -5.31897128e-01 -1.49690345e-01
-2.83792168e-01 3.71899843e-01 -2.33570084e-01 4.13148761e-01
6.68745264e-02 -5.63616931e-01 -7.35168099e-01 3.81587595e-01
-1.08768785e+00 -1.54739857e-01 7.30451882e-01 8.23512077e-01
4.62622076e-01 1.23724513e-01 6.39147818e-01 -8.36942017e-01
7.01083958e-01 -1.89483747e-01 -5.43198466e-01 4.37806547e-01
-6.99538052e-01 2.95677751e-01 1.11208940e+00 -3.45382273e-01
-8.69635463e-01 -1.58676147e-01 -2.32953727e-01 -2.60496348e-01
-5.07891119e-01 2.71566600e-01 -2.91795935e-02 -2.81939030e-01
6.71161413e-01 6.65719435e-02 -3.41373414e-01 -4.77041364e-01
3.49561155e-01 6.52110040e-01 1.88490316e-01 -2.82507688e-01
5.01535833e-01 4.91642922e-01 -3.21667679e-02 -5.68157136e-01
-5.11146486e-01 -1.96921512e-01 -6.59811497e-01 -4.87887226e-02
8.58385563e-01 -5.33387005e-01 -7.79634893e-01 3.95032138e-01
-8.65849316e-01 -1.92613795e-01 -2.37276033e-01 1.43192157e-01
-2.94527382e-01 4.01976377e-01 -3.28503728e-01 -6.75094545e-01
-2.78896511e-01 -1.42407024e+00 8.02797198e-01 1.72516033e-01
-9.99501869e-02 -7.67524958e-01 -3.72683883e-01 -4.03172933e-02
6.85857892e-01 1.56158581e-01 1.42397320e+00 -4.68299896e-01
-4.38946366e-01 7.64793064e-03 -6.21404827e-01 5.55491388e-01
2.12918773e-01 -6.01485260e-02 -1.10916901e+00 -3.44293147e-01
-1.87309295e-01 -1.34428680e-01 1.20463538e+00 4.57215071e-01
1.34187853e+00 -2.87139770e-02 -1.56877980e-01 5.50497890e-01
1.67004454e+00 8.44072104e-02 6.04911447e-01 6.05756402e-01
5.99478424e-01 8.18601072e-01 4.17598307e-01 2.26730257e-01
9.62242633e-02 9.28182781e-01 6.37818873e-01 -2.13642314e-01
-3.57050508e-01 1.45932078e-01 2.45760977e-01 3.51690143e-01
-2.20266819e-01 -3.36802006e-01 -6.74287796e-01 4.38452959e-01
-1.77657759e+00 -5.88582635e-01 1.23264529e-02 2.33006358e+00
4.91856754e-01 2.80071944e-01 1.36131138e-01 5.15467405e-01
6.57845855e-01 2.29393914e-01 -3.29094976e-01 -6.86564505e-01
-2.46561855e-01 1.75971091e-01 3.93594354e-01 3.69717121e-01
-1.14608479e+00 7.32340515e-01 5.46996307e+00 1.06221581e+00
-1.59398782e+00 1.65601261e-02 7.02673733e-01 2.36117840e-01
-1.08179264e-01 -2.86607534e-01 -7.25186944e-01 4.15916711e-01
6.00129008e-01 2.19567209e-01 5.14656723e-01 7.62226820e-01
3.70385498e-02 -7.81474635e-04 -8.11192870e-01 7.80064285e-01
7.66186342e-02 -9.18389082e-01 1.87662765e-01 -2.02228427e-01
3.96585852e-01 -2.12799579e-01 1.02385143e-02 2.24988952e-01
-1.43206000e-01 -1.20479572e+00 1.00940800e+00 2.73162812e-01
5.91341853e-01 -7.24058688e-01 8.12235236e-01 2.61365622e-01
-1.08912098e+00 -3.40752095e-01 -4.16629612e-01 8.25926214e-02
-2.58206874e-01 4.32327420e-01 -4.32338119e-01 6.20579958e-01
7.55535841e-01 3.38565797e-01 -8.68817151e-01 9.85265434e-01
-2.97885716e-01 2.91128546e-01 1.98933575e-02 -9.20535177e-02
2.33923480e-01 -1.04733869e-01 8.38219076e-02 1.33729935e+00
2.99553692e-01 -4.50177282e-01 -6.75265267e-02 7.41294205e-01
9.40856561e-02 4.28616017e-01 -3.98143858e-01 1.69940665e-01
1.71156242e-01 1.48760855e+00 -1.09120619e+00 -1.28419638e-01
-3.07275414e-01 8.30454648e-01 4.78107333e-01 2.24401772e-01
-7.69185245e-01 -4.06863809e-01 5.70890307e-01 1.99507207e-01
6.19561374e-01 -1.18582256e-01 -2.76867568e-01 -8.52070510e-01
1.49771661e-01 -9.84086514e-01 3.45321596e-01 -4.11163479e-01
-1.19070232e+00 1.08131385e+00 -8.07030424e-02 -1.19669735e+00
-9.80680436e-02 -9.60745394e-01 -3.69478852e-01 7.34817863e-01
-1.79483271e+00 -1.11034000e+00 -5.90825737e-01 5.85166216e-01
4.18864042e-01 1.56910539e-01 6.78034961e-01 4.91958886e-01
-5.10742784e-01 6.32442951e-01 -4.28670980e-02 -1.40280694e-01
7.45078921e-01 -1.28166652e+00 1.75841421e-01 6.84563994e-01
1.85409799e-01 6.47469461e-01 6.07965231e-01 4.93626148e-02
-1.02040565e+00 -9.48027909e-01 6.86398804e-01 1.62944887e-02
4.63667691e-01 -4.18495685e-01 -9.85662520e-01 2.23324537e-01
2.60736048e-01 6.10720515e-02 3.08106393e-01 1.88907161e-01
-4.74796414e-01 -2.41035581e-01 -1.07271636e+00 5.42783141e-01
9.91568208e-01 -2.96146065e-01 -3.71384621e-01 -9.42651629e-02
4.21175808e-01 1.71822578e-01 -6.97426140e-01 6.26055956e-01
6.92955136e-01 -1.36529267e+00 1.09297800e+00 -4.42661613e-01
2.48009846e-01 -6.22087598e-01 -1.13782465e-01 -1.31251955e+00
-5.97140193e-01 1.60164788e-01 2.22169638e-01 1.31474257e+00
3.19919169e-01 -7.45451093e-01 5.10297298e-01 1.96252689e-01
2.63574552e-02 -8.34103405e-01 -7.81717539e-01 -7.27269173e-01
1.14416011e-01 -1.33427665e-01 6.13470733e-01 6.71319485e-01
-4.82594341e-01 4.06841308e-01 -4.06209588e-01 -3.08833700e-02
1.79628834e-01 4.42146808e-01 6.27671301e-01 -1.40901506e+00
-3.67011994e-01 -8.73997688e-01 -5.60961664e-01 -7.33103096e-01
-2.62280852e-02 -7.62809217e-01 -7.48445988e-02 -1.45310497e+00
1.70847610e-01 -6.62367582e-01 -7.01692939e-01 5.96250415e-01
-2.88578659e-01 6.75955772e-01 4.13716525e-01 2.55044460e-01
-3.61947417e-01 2.71588773e-01 9.58438039e-01 -1.83308646e-01
-7.87253901e-02 -1.63939193e-01 -8.14364970e-01 7.62902260e-01
7.87325859e-01 -1.76410079e-01 -3.77851695e-01 -4.39116895e-01
8.95354450e-02 -5.15593708e-01 3.39605838e-01 -1.13746119e+00
-5.26034227e-03 -6.81542279e-03 3.00391376e-01 -1.80553541e-01
2.97855288e-01 -8.70420814e-01 7.28227198e-02 5.57788193e-01
-4.28829074e-01 -2.50460245e-02 1.17689453e-01 2.33606517e-01
-4.37041283e-01 -4.75230247e-01 9.57316041e-01 -7.19887912e-02
-1.05936837e+00 8.46410692e-02 -2.59352535e-01 -3.34073186e-01
1.00629461e+00 -2.49815077e-01 -3.99823099e-01 -1.04029559e-01
-5.69469690e-01 -2.20110744e-01 4.67780828e-01 5.92590213e-01
1.53794214e-01 -9.51527417e-01 -5.91720223e-01 1.95987612e-01
3.75016689e-01 -4.22737420e-01 3.29037130e-01 8.97099257e-01
-3.59805942e-01 5.86329341e-01 -5.03136337e-01 -6.03215218e-01
-1.22587276e+00 9.21723485e-01 3.31843197e-01 -2.83119291e-01
-5.13372064e-01 6.00904405e-01 6.42514646e-01 -5.50791770e-02
3.59972388e-01 -3.73303771e-01 -7.90231347e-01 4.14870888e-01
4.60166425e-01 1.00236468e-01 4.86769438e-01 -6.48379564e-01
-3.77164304e-01 8.59052896e-01 -3.47947069e-02 3.92953932e-01
1.15916526e+00 -1.79946870e-01 -1.58161838e-02 3.32208246e-01
1.05485296e+00 1.53938979e-01 -1.08167946e+00 1.04582915e-02
8.76721367e-02 -3.73350173e-01 4.91293855e-02 -7.82775998e-01
-1.19937420e+00 1.00453699e+00 9.78063285e-01 5.69006503e-01
1.35532331e+00 -1.79449156e-01 2.34174088e-01 9.91128460e-02
5.49403906e-01 -5.80597341e-01 -3.11859846e-01 2.26809427e-01
8.24319541e-01 -1.02939010e+00 -1.63577572e-01 -4.58473474e-01
-4.55235332e-01 1.08447468e+00 4.60798681e-01 -2.71672189e-01
4.42030549e-01 2.04061851e-01 1.95395395e-01 7.61619434e-02
-5.18646836e-01 -4.21470821e-01 2.58171469e-01 3.64685565e-01
6.01000786e-01 3.46784554e-02 -7.45274782e-01 4.61389214e-01
-2.40712315e-02 -2.87003368e-02 3.16841364e-01 6.81247473e-01
-4.56599176e-01 -1.33586037e+00 -3.40555161e-01 3.34733278e-01
-4.46721286e-01 1.29068205e-02 -4.47279692e-01 8.92922223e-01
6.12509906e-01 8.70504260e-01 -2.65466385e-02 -3.96626323e-01
3.98500919e-01 -1.27925903e-01 4.99914408e-01 -2.31142998e-01
-7.67230213e-01 -2.91951269e-01 -1.84863564e-02 -3.16363901e-01
-4.08277005e-01 -3.03205371e-01 -8.29570234e-01 -2.04111055e-01
-2.76228249e-01 2.52319545e-01 6.35845602e-01 8.73904765e-01
2.97896087e-01 7.24220395e-01 6.27843440e-01 -7.50909984e-01
-6.08603776e-01 -1.02842128e+00 -5.11959612e-01 5.09242833e-01
3.46415460e-01 -1.00656176e+00 -4.27668720e-01 -1.72521204e-01] | [9.518637657165527, 2.2033040523529053] |
5ae1750f-b891-4c74-9636-57514789996f | an-encryption-method-of-convmixer-models | 2207.11939 | null | https://arxiv.org/abs/2207.11939v1 | https://arxiv.org/pdf/2207.11939v1.pdf | An Encryption Method of ConvMixer Models without Performance Degradation | In this paper, we propose an encryption method for ConvMixer models with a secret key. Encryption methods for DNN models have been studied to achieve adversarial defense, model protection and privacy-preserving image classification. However, the use of conventional encryption methods degrades the performance of models compared with that of plain models. Accordingly, we propose a novel method for encrypting ConvMixer models. The method is carried out on the basis of an embedding architecture that ConvMixer has, and models encrypted with the method can have the same performance as models trained with plain images only when using test images encrypted with a secret key. In addition, the proposed method does not require any specially prepared data for model training or network modification. In an experiment, the effectiveness of the proposed method is evaluated in terms of classification accuracy and model protection in an image classification task on the CIFAR10 dataset. | ['Hitoshi Kiya', 'Ryota Iijima'] | 2022-07-25 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 3.58870298e-01 -9.96551514e-02 1.99341372e-01 -4.10839885e-01
-1.16220564e-01 -7.04674959e-01 7.57443190e-01 -1.84591994e-01
-8.80529881e-01 5.70856571e-01 -3.59117270e-01 -2.80410737e-01
8.95761028e-02 -7.20504701e-01 -6.06744945e-01 -9.14783955e-01
3.45493108e-01 -4.98459525e-02 -1.43079847e-01 4.11289521e-02
1.55962825e-01 7.49792635e-01 -1.33985472e+00 4.37083811e-01
6.87616169e-01 1.25845182e+00 2.53036410e-01 4.71117824e-01
2.44177356e-01 6.82811797e-01 -9.29628253e-01 -8.95058632e-01
9.11782920e-01 -2.43311942e-01 -4.93466824e-01 -2.05723703e-01
4.98260021e-01 -7.27861702e-01 -8.75233352e-01 1.44428241e+00
2.19965518e-01 -1.05815403e-01 4.89303172e-01 -1.71197677e+00
-3.40942085e-01 4.77063596e-01 1.02603823e-01 -8.32582340e-02
-5.20484507e-01 -6.55201729e-03 4.40794140e-01 -2.00046480e-01
2.55571336e-01 9.96391535e-01 3.76943350e-01 9.48708355e-01
-1.28558874e+00 -1.17650437e+00 -3.04440148e-02 3.22242707e-01
-1.35207808e+00 -3.67701322e-01 6.65533245e-01 -7.16906562e-02
6.11458838e-01 7.37396657e-01 5.98285019e-01 1.09949088e+00
4.48578596e-01 5.61750948e-01 1.37468624e+00 -1.50642902e-01
1.81415558e-01 7.50804067e-01 3.20956439e-01 2.74502188e-01
6.06068313e-01 3.33275735e-01 -1.13224555e-02 -3.88135791e-01
3.88756275e-01 2.55465478e-01 -5.12576163e-01 -2.43998900e-01
-8.86670768e-01 7.12116003e-01 3.81037325e-01 3.18053246e-01
-1.21295691e-01 2.62761325e-01 4.94659066e-01 4.80746955e-01
2.53386170e-01 4.09241319e-01 -3.25390965e-01 5.43999672e-01
-8.62192810e-01 2.93880522e-01 7.10397184e-01 9.30583596e-01
3.41100872e-01 1.62453055e-01 1.83786064e-01 4.22716677e-01
4.68679190e-01 5.93946457e-01 7.02232718e-01 -6.27838969e-01
5.01856029e-01 3.96470398e-01 -1.36427417e-01 -9.54914749e-01
2.03778133e-01 -2.64696032e-01 -1.12811840e+00 6.02218091e-01
1.48642823e-01 -1.84241563e-01 -9.60222483e-01 1.77681708e+00
3.90620669e-03 2.66620755e-01 9.12539184e-01 3.25841039e-01
7.93220282e-01 9.20413256e-01 1.42363966e-01 -3.22811119e-03
1.23488402e+00 -9.26008523e-01 -8.84468138e-01 9.53536704e-02
4.27285016e-01 -6.92392945e-01 4.19033051e-01 5.49110830e-01
-7.33105600e-01 -5.78504741e-01 -1.34260082e+00 9.71782655e-02
-5.24798810e-01 1.75388545e-01 5.23178518e-01 1.10262156e+00
-9.27039921e-01 6.25552177e-01 -7.48514235e-01 5.59987649e-02
5.81947684e-01 7.88305998e-01 -8.13683867e-01 8.36530477e-02
-1.23776865e+00 9.17192638e-01 8.51431489e-01 2.82480359e-01
-1.15729189e+00 -6.41812086e-01 -9.05856371e-01 3.80218863e-01
-3.50547880e-01 -3.41553301e-01 8.10492158e-01 -1.27812028e+00
-1.21626675e+00 5.89403152e-01 3.46602619e-01 -1.05468488e+00
7.68140316e-01 4.19550575e-02 -6.32562339e-01 2.34541535e-01
-6.23891592e-01 7.33241022e-01 1.08740735e+00 -1.45424116e+00
-5.07183075e-01 -2.41915151e-01 3.02821517e-01 -1.13669954e-01
-7.83766925e-01 -5.69764292e-03 1.68549612e-01 -7.89961159e-01
-8.71274620e-02 -8.77091289e-01 -2.66306639e-01 1.43014893e-01
-4.30193961e-01 4.00946289e-01 1.46393919e+00 -7.47623742e-01
8.86373997e-01 -2.45378327e+00 -2.82832801e-01 5.84622502e-01
1.84637770e-01 1.02582896e+00 -1.91171885e-01 4.74171579e-01
-2.86013156e-01 4.43057060e-01 -2.19979450e-01 -5.48187971e-01
-5.96911199e-02 2.17994779e-01 -5.79836965e-01 3.45629960e-01
-2.89576679e-01 6.55037463e-01 -2.20243692e-01 -3.90736341e-01
3.28496516e-01 8.44503701e-01 -3.64306509e-01 2.18396604e-01
1.15397356e-01 2.21387506e-01 -3.29355925e-01 1.78905845e-01
1.26239181e+00 1.36369258e-01 3.62668455e-01 -4.31230485e-01
1.85607314e-01 -2.45360687e-01 -1.00899446e+00 9.98416841e-01
-2.92907387e-01 6.72356248e-01 1.61853284e-01 -9.02186692e-01
8.83445501e-01 6.83883131e-01 1.12998933e-01 -2.57287502e-01
4.43832666e-01 6.37755319e-02 2.85157740e-01 -1.81679159e-01
2.51892358e-01 6.71677887e-02 2.99885899e-01 3.64543438e-01
-2.24012863e-02 -8.81245136e-02 -2.85533249e-01 7.62733296e-02
8.12493980e-01 -4.22821701e-01 7.03841299e-02 -2.87696451e-01
8.21688771e-01 -2.99388528e-01 4.39968973e-01 5.09213805e-01
-8.85539800e-02 1.18982784e-01 2.67462641e-01 -5.17205834e-01
-1.02348816e+00 -7.72707283e-01 -3.75187635e-01 3.21818218e-02
2.75263667e-01 -4.79386061e-01 -1.13118064e+00 -9.56143498e-01
-8.05835947e-02 6.75435841e-01 -5.53253174e-01 -5.24325192e-01
-4.35967565e-01 -6.62704051e-01 1.03751910e+00 2.10516766e-01
1.28059113e+00 -9.92619455e-01 -3.56873661e-01 -2.25281209e-01
-4.56088521e-02 -1.22043502e+00 -2.72809446e-01 -5.29957470e-03
-1.06098306e+00 -1.06554258e+00 -4.29985017e-01 -9.68412220e-01
1.07862318e+00 2.11151898e-01 2.78193563e-01 4.72439408e-01
8.36525336e-02 1.13936283e-01 -1.66296735e-01 -6.05204999e-01
-8.53757858e-01 -9.73721594e-02 1.02677949e-01 3.49344194e-01
9.71295014e-02 -4.61817801e-01 -4.68951881e-01 2.53137261e-01
-1.43338084e+00 4.63810302e-02 5.17260373e-01 8.76299381e-01
3.90358359e-01 5.49574435e-01 4.33151603e-01 -9.71348822e-01
5.68301082e-01 6.41759411e-02 -9.54876661e-01 5.65630734e-01
-8.22873592e-01 -5.62266931e-02 9.56293583e-01 -6.02823138e-01
-9.89723623e-01 4.54609543e-02 -3.01927596e-01 -6.59659982e-01
-1.90317243e-01 4.28729057e-02 -7.36245394e-01 -6.24300599e-01
1.97765604e-01 4.55981523e-01 2.66872972e-01 -5.15809655e-01
-1.38650313e-01 8.88859808e-01 3.37651074e-01 -9.34275314e-02
1.08503580e+00 5.86764395e-01 2.23542765e-01 -6.43705547e-01
-2.83017177e-02 3.06001157e-01 -3.54534417e-01 1.78307027e-01
1.05926418e+00 -9.03789401e-01 -7.85318434e-01 1.13464808e+00
-1.08627331e+00 4.27626967e-02 1.83160715e-02 7.85009325e-01
-2.51906991e-01 6.47936702e-01 -6.19005024e-01 -6.71418130e-01
-7.51953304e-01 -1.36921954e+00 1.71863765e-01 1.09178521e-01
2.90465564e-01 -9.73841488e-01 -4.46890801e-01 4.04242575e-01
4.94799912e-01 3.11212361e-01 9.90360975e-01 -1.10928905e+00
-7.47466207e-01 -6.17931128e-01 9.80463997e-02 1.37822008e+00
1.52299494e-01 -9.71994326e-02 -1.30843878e+00 -5.43806374e-01
7.07657218e-01 -1.88783213e-01 8.07701170e-01 4.92549464e-02
1.66711783e+00 -7.66379774e-01 -3.22913259e-01 1.04435861e+00
1.53011262e+00 7.25463867e-01 1.05523610e+00 4.00856972e-01
6.11439109e-01 3.86638165e-01 4.10932571e-01 5.20822667e-02
-1.48034453e-01 2.41203830e-01 7.75812030e-01 -2.14313120e-01
5.83740830e-01 -1.58103466e-01 2.33068481e-01 5.04749060e-01
7.16269985e-02 -5.66168189e-01 -3.95250052e-01 8.27088580e-02
-1.63184130e+00 -1.06052327e+00 8.30017552e-02 2.19904041e+00
4.26376075e-01 -1.07775163e-02 -6.37253404e-01 3.47650558e-01
6.64830565e-01 2.24929079e-01 -2.74294227e-01 -5.02275944e-01
-1.33751318e-01 2.59184837e-01 8.36139441e-01 2.96773165e-01
-1.28414440e+00 8.61312032e-01 6.13928509e+00 9.43073809e-01
-1.29900193e+00 6.92072208e-04 7.63999999e-01 1.01541422e-01
-2.73300201e-01 -1.52468802e-02 -6.44788980e-01 5.82231581e-01
7.46021748e-01 -4.47827950e-02 4.31303859e-01 9.40861464e-01
-1.48019388e-01 3.21596980e-01 -1.09965789e+00 1.15266299e+00
1.46238357e-01 -1.31064343e+00 3.50824267e-01 3.37128729e-01
4.19681519e-01 -4.58283931e-01 3.60301882e-01 1.46660162e-02
-1.56057760e-01 -1.21296179e+00 4.95403856e-01 2.71420717e-01
4.83094305e-01 -1.07055390e+00 9.86644983e-01 5.23608685e-01
-7.30518997e-01 -4.64645326e-02 -6.46942735e-01 2.80553341e-01
-1.23526566e-01 -1.75668165e-01 -6.48813367e-01 6.30581617e-01
5.50750017e-01 3.39969486e-01 -5.32293677e-01 7.45841801e-01
-1.71773389e-01 5.54429829e-01 -2.16752768e-01 5.39965779e-02
2.12855771e-01 -3.20400596e-01 4.52431887e-01 6.98651850e-01
1.67727172e-01 1.14702828e-01 -2.88303375e-01 6.10027671e-01
-3.39765131e-01 5.41681908e-02 -8.66196513e-01 -2.51439005e-01
3.41735363e-01 1.31306577e+00 -4.76496994e-01 -5.44118285e-01
-1.64702937e-01 1.00058913e+00 -1.61528707e-01 3.92018706e-01
-8.56562436e-01 -3.15545499e-01 8.38683069e-01 -1.07955299e-01
2.74695069e-01 7.41250515e-02 -5.35896122e-02 -1.18036330e+00
5.28781749e-02 -1.13600516e+00 1.56228498e-01 -3.89619112e-01
-1.20769334e+00 1.16690683e+00 2.26602420e-01 -1.35818481e+00
1.39198169e-01 -7.06034780e-01 -6.79090858e-01 8.62802327e-01
-1.35361648e+00 -1.41146731e+00 -6.60192966e-02 8.85801494e-01
-2.38466375e-02 -6.69050038e-01 1.13602400e+00 4.05061841e-01
-4.31502223e-01 9.77054060e-01 2.27895141e-01 3.98936152e-01
3.42593461e-01 -7.11105764e-01 4.46589947e-01 1.06004179e+00
1.67270809e-01 8.08008969e-01 4.13800448e-01 -5.37825167e-01
-9.20503616e-01 -1.21352410e+00 7.80845225e-01 1.02967903e-01
-1.11975946e-01 -7.51372457e-01 -7.23172247e-01 9.59466636e-01
5.84332407e-01 6.06097840e-02 8.82035136e-01 -7.67366648e-01
-3.58018458e-01 -4.77763742e-01 -1.93140554e+00 5.45030653e-01
4.54927683e-01 -5.23021877e-01 -4.46133226e-01 2.90957540e-01
7.51953840e-01 -1.44992724e-01 -9.06991184e-01 4.11024034e-01
6.39792085e-01 -6.74478710e-01 1.01400936e+00 -7.30405569e-01
1.45644126e-02 -3.81402105e-01 -4.61094022e-01 -1.04688728e+00
1.33637086e-01 -4.82424438e-01 1.43747687e-01 1.28542328e+00
2.67186671e-01 -1.23906314e+00 7.46510804e-01 9.55563128e-01
2.18948916e-01 -4.47634786e-01 -8.88067842e-01 -8.72185707e-01
-9.28598717e-02 -1.04662374e-01 1.04768252e+00 1.03850520e+00
-5.94568729e-01 -2.41395906e-01 -7.62144268e-01 5.18651664e-01
8.24645281e-01 -2.26316437e-01 9.48112905e-01 -9.07648146e-01
-1.72441840e-01 1.78167060e-01 -9.27173555e-01 -6.19043529e-01
5.07471025e-01 -1.01122606e+00 -5.80807269e-01 -1.06024659e+00
6.59120008e-02 -5.97285926e-01 -4.67452198e-01 5.30966163e-01
1.20050043e-01 3.44664037e-01 6.25945389e-01 1.94259748e-01
2.84202009e-01 5.06573260e-01 1.04697907e+00 -3.62063259e-01
2.49317512e-01 3.67071122e-01 -5.69262445e-01 6.07150674e-01
1.24381208e+00 -8.08707654e-01 -9.31614041e-01 -4.33736563e-01
-4.28690583e-01 -3.30128580e-01 6.13535941e-01 -1.25770724e+00
7.02872500e-02 4.02778313e-02 5.35782695e-01 -4.92952496e-01
7.27651298e-01 -1.68799794e+00 7.50023127e-01 9.41186130e-01
-3.31949979e-01 1.88043833e-01 3.53600413e-01 4.09215599e-01
-3.60808641e-01 -7.23340571e-01 1.03046799e+00 4.70064878e-02
-5.34242451e-01 4.48525935e-01 -2.03778863e-01 -5.97246826e-01
1.45811939e+00 -2.43014425e-01 -3.77125680e-01 -1.82690606e-01
-5.60852885e-01 -5.14460802e-02 6.59060597e-01 3.32288384e-01
8.06435645e-01 -1.29775178e+00 -4.15418684e-01 6.17045581e-01
-1.08044915e-01 -3.28538448e-01 2.55230546e-01 1.47385180e-01
-9.72484648e-01 2.69097686e-01 -5.80932558e-01 -2.01095641e-01
-1.85293806e+00 9.00786996e-01 5.20796895e-01 -3.38300973e-01
-4.25367087e-01 5.05970359e-01 5.43995380e-01 -3.21064144e-01
2.59251088e-01 -4.78626966e-01 -2.03877673e-01 -3.88527691e-01
6.76848531e-01 9.44482684e-02 3.77995241e-03 -6.49262249e-01
-1.95429027e-01 1.30809456e-01 -3.90541583e-01 -2.19545111e-01
1.16693354e+00 2.15414643e-01 -2.99596906e-01 -2.37762704e-01
1.40494359e+00 -2.62679458e-02 -1.00757849e+00 1.62038893e-01
-5.49850643e-01 -8.07340801e-01 9.94071830e-03 -5.61661899e-01
-1.46024644e+00 8.67070735e-01 1.03805733e+00 7.43156374e-02
1.20869350e+00 -7.43138433e-01 9.04199421e-01 5.13196886e-01
2.44054645e-01 -7.05979407e-01 -4.52646196e-01 5.66559918e-02
6.74030662e-01 -1.15717244e+00 -9.44712162e-02 -3.99286002e-01
-6.49709523e-01 1.04261148e+00 7.19533026e-01 -1.51173651e-01
1.15927505e+00 2.56757110e-01 3.96134883e-01 1.35283604e-01
-6.33869827e-01 5.62065661e-01 5.99180385e-02 7.08782911e-01
-3.78666878e-01 -3.16512063e-02 -3.95912439e-01 5.36397934e-01
-3.15428585e-01 -2.36947089e-01 4.15453285e-01 9.73348379e-01
5.30077592e-02 -1.56566453e+00 -2.72358656e-01 3.63362074e-01
-9.24566805e-01 -2.40485370e-01 -3.10844928e-01 9.67411518e-01
2.89181828e-01 9.02671397e-01 4.55064140e-02 -7.39922166e-01
1.02418683e-01 -5.28138615e-02 2.92694867e-01 -1.10750511e-01
-1.09751236e+00 -3.68012816e-01 -1.52677551e-01 -1.38254181e-01
-3.05891007e-01 -3.19214433e-01 -9.79279578e-01 -5.32491386e-01
-5.28234482e-01 3.00130814e-01 1.02225566e+00 6.70750499e-01
1.52135700e-01 2.19270691e-01 7.69212902e-01 -3.67014945e-01
-8.07462692e-01 -9.16249394e-01 -6.50042593e-01 3.96527588e-01
2.56217360e-01 -1.56569660e-01 -3.62329215e-01 3.42929699e-02] | [5.525876522064209, 7.369512557983398] |
fe26acbc-c9a1-48d7-a704-9a9859682ed4 | online-nonnegative-tensor-factorization-and | 2009.07612 | null | https://arxiv.org/abs/2009.07612v4 | https://arxiv.org/pdf/2009.07612v4.pdf | Online nonnegative CP-dictionary learning for Markovian data | Online Tensor Factorization (OTF) is a fundamental tool in learning low-dimensional interpretable features from streaming multi-modal data. While various algorithmic and theoretical aspects of OTF have been investigated recently, a general convergence guarantee to stationary points of the objective function without any incoherence or sparsity assumptions is still lacking even for the i.i.d. case. In this work, we introduce a novel algorithm that learns a CANDECOMP/PARAFAC (CP) basis from a given stream of tensor-valued data under general constraints, including nonnegativity constraints that induce interpretability of the learned CP basis. We prove that our algorithm converges almost surely to the set of stationary points of the objective function under the hypothesis that the sequence of data tensors is generated by an underlying Markov chain. Our setting covers the classical i.i.d. case as well as a wide range of application contexts including data streams generated by independent or MCMC sampling. Our result closes a gap between OTF and Online Matrix Factorization in global convergence analysis \commHL{for CP-decompositions}. Experimentally, we show that our algorithm converges much faster than standard algorithms for nonnegative tensor factorization tasks on both synthetic and real-world data. Also, we demonstrate the utility of our algorithm on a diverse set of examples from image, video, and time-series data, illustrating how one may learn qualitatively different CP-dictionaries from the same tensor data by exploiting the tensor structure in multiple ways. | ['Christopher Strohmeier', 'Deanna Needell', 'Hanbaek Lyu'] | 2020-09-16 | null | null | null | null | ['online-nonnegative-cp-decomposition'] | ['methodology'] | [ 3.40470336e-02 -3.89914781e-01 -1.33969411e-01 -1.12150922e-01
-4.75793749e-01 -8.62076283e-01 3.27153891e-01 2.69216318e-02
-1.12483203e-01 4.10248101e-01 3.99501026e-01 -2.67977148e-01
-5.87729394e-01 -2.95499295e-01 -8.25686276e-01 -1.10992730e+00
-5.90364575e-01 6.53854132e-01 -4.95937169e-01 -1.73051387e-01
-8.25809240e-02 4.03770179e-01 -1.36878181e+00 5.19319594e-01
4.52429831e-01 9.65413272e-01 -1.21121153e-01 9.00854826e-01
-2.72135623e-03 5.56279004e-01 -1.76010087e-01 -2.83509433e-01
6.37269318e-01 -1.92356650e-02 -8.20616782e-01 6.03109717e-01
2.76888877e-01 -1.15750939e-01 -2.69205689e-01 1.02170944e+00
9.42581445e-02 1.24517106e-01 5.68851352e-01 -1.58273244e+00
-4.33313370e-01 5.42655528e-01 -7.12821603e-01 3.89730453e-01
2.52075970e-01 -1.44501537e-01 1.38829076e+00 -1.09475756e+00
4.60242420e-01 1.15349877e+00 4.47576106e-01 1.00897990e-01
-1.51195085e+00 -3.18576932e-01 1.16613254e-01 1.47383317e-01
-1.05880940e+00 -2.15566814e-01 6.74094439e-01 -7.32811928e-01
4.30680245e-01 4.48888361e-01 8.31213117e-01 1.11461091e+00
6.02119230e-02 1.00015581e+00 9.25093949e-01 -3.84367377e-01
3.39747697e-01 -3.08367610e-01 3.67739737e-01 5.51870823e-01
4.31581974e-01 -7.42521286e-02 -1.03831267e+00 -7.60404766e-01
6.75807178e-01 3.90873611e-01 -2.75452048e-01 -5.21383584e-01
-1.74485302e+00 8.92472684e-01 -1.24108598e-01 3.17692637e-01
-6.52077794e-01 2.08586261e-01 5.32890737e-01 5.36919653e-01
4.85255927e-01 7.27952719e-02 -5.06658077e-01 -1.82351232e-01
-5.37956953e-01 2.66506433e-01 8.82002413e-01 8.98748577e-01
7.37734616e-01 2.03260422e-01 2.08629772e-01 4.23014969e-01
2.34744474e-01 8.64797711e-01 2.80307114e-01 -1.12399876e+00
6.28910303e-01 1.43340841e-01 3.10667038e-01 -1.08965671e+00
-3.34609658e-01 -5.25807500e-01 -1.08112943e+00 -4.60058004e-01
6.30754948e-01 -8.76949877e-02 -3.85588497e-01 1.69481218e+00
4.15967107e-01 4.61589098e-01 -9.47775468e-02 1.05622911e+00
2.38639186e-03 6.79910660e-01 -5.13275743e-01 -5.45092940e-01
1.25051129e+00 -4.34938461e-01 -7.93584287e-01 2.07903937e-01
5.42174399e-01 -7.22002804e-01 7.70272195e-01 9.18640971e-01
-8.74205530e-01 -1.63377479e-01 -6.17930353e-01 1.92097381e-01
2.60061055e-01 1.95646826e-02 1.13075304e+00 4.21754867e-01
-8.58623385e-01 3.49324703e-01 -1.10277951e+00 -1.99999571e-01
7.72292838e-02 2.39169270e-01 -5.16933382e-01 -3.44980687e-01
-8.09936523e-01 -7.02854097e-02 1.65738434e-01 3.94454718e-01
-1.03449976e+00 -8.19814503e-01 -3.45834523e-01 -8.49577710e-02
2.86677569e-01 -9.72277164e-01 1.10494888e+00 -1.16008663e+00
-1.11172986e+00 3.70899320e-01 -4.66819227e-01 -4.12292182e-01
3.06318969e-01 -3.48572373e-01 -2.72489071e-01 3.80541891e-01
2.59067893e-01 -1.67965159e-01 1.49452364e+00 -1.09996247e+00
-4.29877639e-01 -4.82878983e-01 2.58769631e-01 -5.31134717e-02
-5.16871810e-01 -1.11977696e-01 -2.27317959e-02 -7.87050605e-01
4.05840456e-01 -1.20691717e+00 -2.76501149e-01 -1.18850298e-01
-2.92263180e-01 -1.60506383e-01 8.01838160e-01 -3.79951209e-01
1.05022383e+00 -2.18518090e+00 6.15676582e-01 3.93768162e-01
3.81319642e-01 -2.78431028e-01 -2.01637954e-01 8.98557007e-01
-5.31109631e-01 4.02452536e-02 -2.18886688e-01 -3.78188789e-01
-6.73778206e-02 5.47944844e-01 -8.38462830e-01 9.42502856e-01
-1.37039140e-01 4.31390107e-01 -1.13081861e+00 1.95851251e-02
-1.16123669e-01 2.22850174e-01 -8.70271921e-01 -1.64908841e-02
-2.02433839e-01 6.05699718e-01 -4.02447551e-01 4.78736639e-01
5.49921453e-01 -6.37800694e-01 2.31547147e-01 -3.40467781e-01
2.98927836e-02 -2.58618921e-01 -1.57184732e+00 1.84706962e+00
-4.36296135e-01 5.53534687e-01 3.23715538e-01 -1.38966417e+00
2.26631925e-01 6.93808675e-01 1.04162920e+00 5.27599789e-02
-4.11798712e-03 2.61189431e-01 7.83726871e-02 -6.61876619e-01
3.34675729e-01 -5.46620965e-01 8.97230208e-02 9.21776175e-01
2.47974306e-01 3.69913042e-01 3.60085875e-01 5.63453853e-01
1.07007575e+00 -4.24664021e-01 -4.28751968e-02 -4.55074251e-01
3.95891219e-01 -1.73487857e-01 5.72786450e-01 7.45990038e-01
2.26849362e-01 4.85199958e-01 5.94673276e-01 -6.59607768e-01
-1.08433104e+00 -1.09337616e+00 -7.11200535e-02 1.01850462e+00
-2.01779068e-01 -7.68626153e-01 -2.60646820e-01 -3.71809632e-01
9.09755081e-02 1.97613895e-01 -7.50485837e-01 2.34095290e-01
-3.34472835e-01 -1.02401698e+00 7.08650798e-02 2.20110208e-01
-9.59311351e-02 -5.57952672e-02 -1.47064209e-01 2.19835162e-01
-5.89632452e-01 -1.39020133e+00 -4.99706089e-01 -1.19484812e-01
-1.22468281e+00 -1.10619926e+00 -3.49683315e-01 -1.11538127e-01
8.35144222e-01 8.80127549e-01 9.68733847e-01 -2.72447288e-01
-9.15339738e-02 1.06771517e+00 -3.49304855e-01 -1.61985144e-01
-4.87142988e-02 -2.81624794e-01 6.18086100e-01 9.52844203e-01
-1.03099577e-01 -6.76025927e-01 -5.26259780e-01 2.83218890e-01
-1.37014377e+00 -3.44375707e-02 1.34430945e-01 9.55463111e-01
5.57609022e-01 2.64898032e-01 1.12392336e-01 -5.21625161e-01
5.55894196e-01 -8.87660623e-01 -5.09777963e-01 2.99152378e-02
-1.56243339e-01 3.82812709e-01 7.24147260e-01 -5.68840981e-01
-6.71989083e-01 -9.54409782e-03 5.13252854e-01 -9.83496845e-01
3.51306438e-01 9.22635376e-01 1.34069651e-01 2.23589182e-01
5.06081045e-01 2.85942733e-01 1.34927586e-01 -5.92109680e-01
3.51337433e-01 3.23926330e-01 2.50025421e-01 -1.20444489e+00
1.03542995e+00 1.23448503e+00 3.67150813e-01 -1.26109302e+00
-1.08069444e+00 -8.24157596e-01 -6.58810019e-01 -2.95927167e-01
4.58905786e-01 -1.11020219e+00 -1.06276023e+00 1.84225261e-01
-1.21910310e+00 1.34820685e-01 -3.79962415e-01 7.91099191e-01
-6.31554008e-01 5.86905777e-01 -4.32326287e-01 -9.66529071e-01
-3.75400335e-02 -7.94512928e-01 1.08855247e+00 -4.32128280e-01
3.29219885e-02 -1.10467100e+00 3.39592159e-01 4.93145376e-01
-4.42510247e-02 2.69296050e-01 7.83361733e-01 -2.86048979e-01
-6.82579756e-01 -2.26749286e-01 2.27924690e-01 5.19358754e-01
4.45536040e-02 1.86324298e-01 -7.33396649e-01 -5.23063600e-01
3.37952822e-01 -4.95633297e-02 7.33348250e-01 3.74797553e-01
9.91148353e-01 -7.26110220e-01 9.75612029e-02 5.71354449e-01
1.45942330e+00 -3.72662306e-01 -2.38895901e-02 -1.31606415e-01
8.23901832e-01 5.41699827e-01 4.24566120e-01 9.65500653e-01
2.72427917e-01 4.88516867e-01 5.23808122e-01 4.30835545e-01
5.09457469e-01 -4.93834689e-02 6.62618339e-01 1.23319244e+00
-4.47571725e-01 -7.00346231e-02 -9.01824772e-01 7.19649613e-01
-2.14846158e+00 -1.15286791e+00 -4.51484799e-01 2.37217903e+00
4.60015386e-01 -3.45014066e-01 4.87289220e-01 3.25345010e-01
5.61815441e-01 9.01756510e-02 -4.41123426e-01 -1.60463944e-01
-2.49453545e-01 1.14431478e-01 4.16596383e-01 4.55396533e-01
-9.51673329e-01 2.26185858e-01 5.91808128e+00 3.81602675e-01
-1.13914037e+00 4.06054765e-01 1.47543848e-01 -2.54723042e-01
-5.37493587e-01 1.14532270e-01 -3.45100224e-01 1.80241734e-01
9.65009272e-01 -3.90802354e-01 6.84597731e-01 5.78198612e-01
4.47186023e-01 1.55306488e-01 -1.20256829e+00 1.32788193e+00
-1.00416616e-01 -1.31394291e+00 7.23716989e-02 3.71995807e-01
9.75764155e-01 1.95112824e-01 1.72917709e-01 -2.95134991e-01
2.37674043e-01 -5.74064314e-01 9.01462615e-01 5.07233799e-01
3.95639539e-01 -5.86464286e-01 3.09342772e-01 3.97814989e-01
-1.10276520e+00 -3.04026455e-01 -2.80594409e-01 -3.48580480e-01
1.67231888e-01 1.21045995e+00 -6.65139139e-01 8.68810356e-01
5.94775975e-01 1.02975702e+00 -1.40949473e-01 7.91716933e-01
1.53265685e-01 9.76141810e-01 -7.73669422e-01 3.73164058e-01
2.95343697e-01 -6.07898295e-01 8.36578310e-01 9.00926769e-01
3.40022862e-01 1.99643061e-01 1.44280076e-01 4.06940907e-01
1.79456249e-01 4.19557393e-02 -6.01379693e-01 -3.96109611e-01
-2.36565415e-02 1.19229221e+00 -5.96650958e-01 -1.77108943e-01
-5.17832756e-01 8.60670805e-01 1.76844858e-02 6.41430378e-01
-6.52081728e-01 3.37289065e-01 1.09853733e+00 1.08464502e-01
4.48462546e-01 -8.55935693e-01 -1.44217983e-01 -1.99223852e+00
2.38837093e-01 -1.01640654e+00 6.10306978e-01 -3.35658818e-01
-1.56578302e+00 2.21315473e-01 1.58165902e-01 -1.42381561e+00
-2.77487427e-01 -5.87343752e-01 -3.55147660e-01 4.58794594e-01
-1.01941514e+00 -9.13441658e-01 5.95640875e-02 1.07150710e+00
2.78105944e-01 1.36897201e-02 5.91577232e-01 2.41292953e-01
-5.29259920e-01 -3.92865874e-02 6.29115760e-01 1.25480630e-02
3.14983547e-01 -1.21666133e+00 -5.55870719e-02 1.04027522e+00
4.92003173e-01 8.87743056e-01 9.02423799e-01 -2.79993027e-01
-2.31330466e+00 -8.40066433e-01 4.25269186e-01 -3.83845150e-01
1.35846353e+00 -5.00487924e-01 -5.68918943e-01 9.22394931e-01
-1.07888468e-01 5.12909114e-01 1.07788765e+00 3.75230819e-01
-5.76448143e-01 -2.92758197e-01 -7.20497251e-01 3.42530042e-01
1.00954664e+00 -6.03862822e-01 -1.41960919e-01 8.99110913e-01
5.39784253e-01 -1.03517041e-01 -1.18381321e+00 2.65241452e-02
5.90241790e-01 -7.67554164e-01 1.08142412e+00 -1.24099112e+00
3.61681879e-01 -5.72766125e-01 -6.35112166e-01 -1.11249352e+00
-3.62151444e-01 -1.09547901e+00 -5.48468053e-01 7.14209318e-01
1.01032719e-01 -6.15288258e-01 4.38324690e-01 3.39853585e-01
5.53802773e-02 -5.37130952e-01 -1.09047055e+00 -9.72781599e-01
-1.33574039e-01 -9.82055783e-01 3.22468966e-01 1.00134397e+00
5.96841536e-02 3.90881330e-01 -5.53433836e-01 5.99917412e-01
1.13457513e+00 4.40907866e-01 8.64581287e-01 -1.14836395e+00
-7.02955067e-01 4.39872406e-02 -3.32048893e-01 -9.51474190e-01
2.29612783e-01 -9.85309899e-01 -4.04629648e-01 -8.58819485e-01
2.42481545e-01 -3.79901379e-01 -9.10497084e-02 1.56108394e-01
2.29134604e-01 -4.86530513e-02 4.85467762e-01 5.12785733e-01
-5.58602273e-01 5.59696496e-01 1.20414567e+00 -1.03906967e-01
6.65622652e-02 1.72536746e-01 -5.19981205e-01 5.81216812e-01
3.57849121e-01 -4.67666328e-01 -6.21371150e-01 -6.56271815e-01
8.62260580e-01 3.73855233e-01 3.52164865e-01 -5.59041381e-01
1.48529693e-01 -3.93265724e-01 -2.00821385e-01 -4.38593745e-01
2.53510863e-01 -9.52620804e-01 4.29751694e-01 3.87995839e-01
-2.82104518e-02 2.86489844e-01 -1.00400858e-01 1.14183760e+00
-1.33321121e-01 2.51166914e-02 4.00013626e-01 3.74268033e-02
-1.85540840e-01 6.71221316e-01 -4.01344627e-01 2.55148143e-01
6.47939265e-01 1.90895230e-01 6.29886752e-03 -5.50023019e-01
-1.02208078e+00 4.29596938e-03 1.07379951e-01 2.97391683e-01
5.32272577e-01 -1.43203485e+00 -8.53729427e-01 3.94003212e-01
2.83600725e-02 -4.47278693e-02 3.52359265e-01 1.28947198e+00
-1.84159532e-01 4.49986428e-01 2.43194044e-01 -1.05507112e+00
-8.84724617e-01 6.31700695e-01 -4.21033688e-02 -2.59085357e-01
-4.15694773e-01 4.79638189e-01 2.93012440e-01 -1.91892639e-01
-1.48023754e-01 -4.73597437e-01 4.40989316e-01 2.76939481e-01
6.08269870e-01 5.18318176e-01 1.01488642e-01 -6.94501162e-01
-1.97684988e-01 3.39215577e-01 5.11362031e-02 -4.30005729e-01
1.60027409e+00 -2.52460450e-01 -5.68259239e-01 9.29359853e-01
1.46731699e+00 -1.84657183e-04 -9.85133886e-01 -4.44916427e-01
-9.67679620e-02 -6.39713228e-01 -1.24771655e-01 -2.67218594e-02
-1.22177863e+00 7.98099756e-01 1.96584612e-01 5.19067228e-01
1.06429505e+00 5.77241443e-02 6.39008462e-01 5.34595728e-01
6.63919806e-01 -6.70723021e-01 3.59657377e-01 6.34530902e-01
9.85512018e-01 -9.70467985e-01 9.99600627e-03 -2.47106776e-01
-4.50354010e-01 1.38702965e+00 -2.73964882e-01 -3.20552975e-01
1.12232029e+00 5.60035631e-02 -4.65002060e-01 -2.81338245e-01
-1.03181887e+00 4.80652042e-02 4.85809371e-02 2.89341688e-01
1.73751116e-01 2.79589504e-01 -1.47446543e-01 3.35894525e-01
-1.81722209e-01 -5.73668070e-02 8.65125597e-01 7.64098704e-01
2.16630306e-02 -1.07456565e+00 -7.49111712e-01 3.62099230e-01
-5.47502697e-01 -4.23246212e-02 1.37467265e-01 2.98830688e-01
-1.60130933e-01 7.50765860e-01 -8.06189552e-02 -3.11536103e-01
-1.45822108e-01 5.46435080e-02 6.60543263e-01 -5.44479311e-01
-1.98483527e-01 2.57402390e-01 -2.08505794e-01 -6.27649784e-01
-8.18409324e-01 -1.41317213e+00 -9.41081703e-01 -5.33481777e-01
-2.46077389e-01 2.82618910e-01 6.81557059e-01 1.07633054e+00
3.15138072e-01 -1.30649582e-02 1.01806927e+00 -7.79972136e-01
-5.96876144e-01 -6.02984726e-01 -8.92057896e-01 4.30971175e-01
7.01289952e-01 -6.35593116e-01 -6.80927575e-01 4.42930967e-01] | [7.177590370178223, 4.660567760467529] |
33b026be-9af6-4b9e-b367-0a7e1c9fb6f9 | local-post-hoc-explanations-for-predictive | 2009.10513 | null | https://arxiv.org/abs/2009.10513v2 | https://arxiv.org/pdf/2009.10513v2.pdf | Local Post-Hoc Explanations for Predictive Process Monitoring in Manufacturing | This study proposes an innovative explainable predictive quality analytics solution to facilitate data-driven decision-making for process planning in manufacturing by combining process mining, machine learning, and explainable artificial intelligence (XAI) methods. For this purpose, after integrating the top-floor and shop-floor data obtained from various enterprise information systems, a deep learning model was applied to predict the process outcomes. Since this study aims to operationalize the delivered predictive insights by embedding them into decision-making processes, it is essential to generate relevant explanations for domain experts. To this end, two complementary local post-hoc explanation approaches, Shapley values and Individual Conditional Expectation (ICE) plots are adopted, which are expected to enhance the decision-making capabilities by enabling experts to examine explanations from different perspectives. After assessing the predictive strength of the applied deep neural network with relevant binary classification evaluation measures, a discussion of the generated explanations is provided. | ['Nijat Mehdiyev', 'Peter Fettke'] | 2020-09-22 | null | null | null | null | ['predictive-process-monitoring'] | ['time-series'] | [ 2.05493972e-01 7.28678346e-01 -4.04886782e-01 -5.49217522e-01
-5.65809682e-02 4.52913940e-02 4.52617556e-01 5.91628850e-01
4.02331382e-01 7.82548308e-01 1.57474771e-01 -8.53757381e-01
-1.10018206e+00 -1.01850927e+00 -3.31462204e-01 -3.48559588e-01
-3.40775192e-01 6.22437418e-01 -1.11259747e+00 -6.84000924e-02
7.20051944e-01 6.45502031e-01 -1.60621738e+00 7.20359623e-01
1.39889657e+00 1.34354019e+00 6.16914816e-02 3.17114174e-01
-2.46643633e-01 1.11213207e+00 -2.32848555e-01 -5.43975055e-01
5.46369143e-02 -3.53253573e-01 -7.45053828e-01 6.02862313e-02
-4.06109363e-01 -2.82517195e-01 2.79199988e-01 6.57119930e-01
-6.45467490e-02 1.23941176e-01 8.33709061e-01 -1.63668728e+00
-1.37787509e+00 8.88126194e-01 1.11038268e-01 -1.94223344e-01
3.21938246e-01 5.55405021e-01 1.23905385e+00 -7.93114066e-01
2.91095614e-01 1.28034079e+00 2.32823655e-01 6.64704666e-02
-9.18652236e-01 -4.95090306e-01 1.84429154e-01 5.22946715e-01
-5.11146486e-01 3.61226290e-01 9.87543106e-01 -6.11224115e-01
8.86883616e-01 2.40371555e-01 9.08400655e-01 9.41326499e-01
9.06473100e-01 6.76857650e-01 1.31614482e+00 -4.91863281e-01
6.36374652e-01 2.19513461e-01 3.14747810e-01 3.18791568e-01
6.72117293e-01 5.00728846e-01 -3.75970721e-01 4.13130373e-01
5.62290847e-01 5.86551428e-01 1.89906120e-01 -3.10712278e-01
-1.09349883e+00 1.12089527e+00 6.64097846e-01 4.22716051e-01
-1.15965223e+00 -1.21133663e-01 1.12260960e-01 4.93500054e-01
1.78482369e-01 1.20331061e+00 -7.23661661e-01 1.79321378e-01
-5.24315774e-01 1.80297017e-01 7.89655626e-01 7.89534032e-01
7.13466465e-01 4.47965294e-01 -5.56384206e-01 -1.35772198e-01
6.72586501e-01 2.53721744e-01 2.28330866e-01 -1.13889825e+00
3.06660414e-01 1.28914011e+00 3.04024905e-01 -1.13381946e+00
-4.71952349e-01 -6.33883417e-01 -9.01913822e-01 6.69969738e-01
-5.67611419e-02 -1.78058431e-01 -8.48182201e-01 8.95395517e-01
-1.54374450e-01 -4.49822426e-01 2.84496248e-01 9.79237199e-01
5.63591003e-01 4.72590774e-01 4.50584233e-01 -1.64362386e-01
1.29208326e+00 -9.69122827e-01 -1.26618230e+00 -7.88781419e-02
3.90085042e-01 -3.27394903e-02 1.02438939e+00 7.53493249e-01
-9.46387231e-01 -9.62668598e-01 -1.30428874e+00 3.38594168e-01
-5.11840999e-01 8.99956748e-02 1.04895413e+00 2.65817225e-01
-6.31818116e-01 1.04928076e+00 -4.89322692e-01 5.34477085e-02
4.94606197e-01 4.08733279e-01 -1.94993809e-01 -1.95659213e-02
-1.32777750e+00 1.12969315e+00 6.63959622e-01 1.92534938e-01
-7.77855098e-01 -7.00110257e-01 -8.16352844e-01 6.82374120e-01
2.73414344e-01 -9.45724428e-01 9.77091014e-01 -1.13720083e+00
-1.36738718e+00 9.21149622e-04 1.84751347e-01 -7.15266347e-01
4.39253837e-01 -3.29339564e-01 -6.14001215e-01 -5.19000702e-02
-1.41838919e-02 1.95805460e-01 3.55329394e-01 -1.66239393e+00
-6.77097321e-01 -7.42697001e-01 -2.72833253e-03 -5.17119952e-02
6.63786680e-02 -4.59454358e-01 4.69110727e-01 -2.53736943e-01
2.42258295e-01 -4.23412859e-01 -4.77007478e-01 -5.45306981e-01
-4.35241878e-01 -2.07278460e-01 4.04196709e-01 -8.76549006e-01
1.04251289e+00 -1.73626614e+00 -9.11126882e-02 5.81548870e-01
4.97504681e-01 -8.45555961e-02 1.22092776e-01 4.90846336e-01
-1.52842477e-01 3.28394920e-01 1.21635996e-01 -6.61744550e-02
4.96998280e-01 2.04483286e-01 -5.07567227e-02 -2.39178807e-01
7.31229186e-01 1.04084277e+00 -7.15830982e-01 1.95111353e-02
8.10709357e-01 1.09336585e-01 -4.51467067e-01 3.31595182e-01
-4.07766998e-01 5.83612859e-01 -6.47858500e-01 9.91074622e-01
5.35149515e-01 -2.10936084e-01 3.87273073e-01 -9.66109186e-02
-1.49186969e-01 2.24290758e-01 -1.06334996e+00 1.16748178e+00
-5.88546038e-01 4.02379334e-01 -3.02275151e-01 -9.65994954e-01
1.39210975e+00 3.40174258e-01 4.35370982e-01 -9.95908797e-01
3.36688429e-01 1.10005043e-01 2.19288066e-01 -5.92536032e-01
3.86850864e-01 -2.43201748e-01 1.18527144e-01 3.22814494e-01
-1.06855027e-01 2.35623851e-01 -3.11103351e-02 -4.25356656e-01
9.26735282e-01 1.14036128e-01 5.69856167e-01 -1.69900935e-02
5.67177236e-01 2.59075284e-01 5.14998257e-01 4.46909994e-01
-2.91369744e-02 8.76361206e-02 5.57072282e-01 -1.04214644e+00
-8.89778316e-01 -9.08461630e-01 1.45095274e-01 5.69251359e-01
1.22724220e-01 -1.16625227e-01 -3.77662838e-01 -6.21203542e-01
2.81705290e-01 1.61564946e+00 -7.04195619e-01 -2.66783208e-01
2.32856229e-01 -3.09788920e-02 -2.60373354e-01 7.87154317e-01
4.06847745e-01 -1.41961026e+00 -8.32162857e-01 3.03236693e-01
3.00934076e-01 -5.26918232e-01 8.58977497e-01 5.79394817e-01
-1.00472391e+00 -1.09861231e+00 5.59037365e-02 -7.25630745e-02
3.31259757e-01 -7.87020177e-02 1.19248152e+00 4.92508970e-02
9.12501514e-02 -5.45966066e-02 -3.73448670e-01 -8.81213665e-01
-6.62782669e-01 -5.35832494e-02 5.80534302e-02 -3.28657687e-01
7.58960783e-01 -4.35276538e-01 -5.53429544e-01 1.48831487e-01
-5.53580463e-01 1.07422143e-01 1.16572297e+00 7.24324524e-01
4.59129840e-01 3.91098946e-01 8.76361847e-01 -5.86191237e-01
1.13734508e+00 -6.24833405e-01 -3.91740471e-01 3.71104896e-01
-1.59925640e+00 3.24351341e-01 6.26728833e-01 8.70302171e-02
-1.41880333e+00 -5.36455750e-01 1.30087286e-01 -3.14466983e-01
-7.02835858e-01 9.84836102e-01 -4.94549215e-01 5.42366445e-01
3.31736416e-01 -1.58379599e-01 1.66572794e-01 -8.20426196e-02
5.37359715e-01 4.80814725e-01 8.34016874e-02 -1.70146048e-01
5.59554517e-01 -1.03647411e-01 1.46643266e-01 2.06497312e-01
-5.63367724e-01 7.08533973e-02 -6.58959985e-01 -4.93107885e-01
1.19564068e+00 -5.12208521e-01 -1.18949544e+00 -4.71648186e-01
-1.05597734e+00 7.12362006e-02 -5.86756468e-01 8.82869482e-01
-8.69935095e-01 -3.52866143e-01 -2.86164284e-01 -1.11437070e+00
-3.91274542e-01 -1.23557067e+00 5.14516890e-01 3.32054555e-01
-7.26492167e-01 -9.34700251e-01 -4.19219881e-01 7.39944041e-01
4.07264888e-01 5.08642256e-01 1.45172405e+00 -1.28058624e+00
-9.16150987e-01 -4.81152415e-01 -7.22864643e-02 3.58805627e-01
1.78297251e-01 3.81948985e-02 -7.36992300e-01 3.27799618e-01
1.63354352e-01 1.68794040e-02 2.13117182e-01 5.90667665e-01
1.16427553e+00 -5.25113642e-01 -8.05412754e-02 2.21531615e-01
1.34502888e+00 8.79037917e-01 7.21704721e-01 8.33820283e-01
3.35529208e-01 1.13512325e+00 1.20193601e+00 8.79517257e-01
4.01709586e-01 3.05895284e-02 8.82001102e-01 5.43593206e-02
4.15568769e-01 -2.08901659e-01 6.99316412e-02 3.16533059e-01
-4.59708214e-01 -5.15872240e-02 -8.56951356e-01 1.34328246e-01
-2.06293583e+00 -1.14257014e+00 -4.50360596e-01 1.69635975e+00
-2.36197412e-01 3.59625608e-01 -3.77023607e-01 4.47935551e-01
3.55598986e-01 -3.95292521e-01 -6.27149642e-01 -9.50877726e-01
1.41739354e-01 1.32605225e-01 1.75536484e-01 2.31329054e-01
-7.24578261e-01 4.57879841e-01 5.73525858e+00 3.78141627e-02
-5.67417085e-01 -3.77033263e-01 9.08516943e-01 1.19447753e-01
-8.99502456e-01 3.02831791e-02 -1.08712569e-01 2.04806864e-01
1.08635128e+00 -2.82396257e-01 3.52128774e-01 1.25378001e+00
7.62935281e-01 1.83137357e-01 -1.44306612e+00 3.61021131e-01
-4.83876526e-01 -1.60956228e+00 3.03195834e-01 3.79661441e-01
9.06023443e-01 -7.90412426e-01 2.20713496e-01 3.96841228e-01
6.07204437e-01 -1.47064281e+00 5.74871659e-01 1.12813735e+00
3.25942859e-02 -1.13155615e+00 1.43661118e+00 6.26021251e-02
-7.21140802e-01 -1.03016317e+00 -2.69464701e-01 -6.65915489e-01
3.91338170e-02 6.25444174e-01 -1.06028938e+00 1.26693165e+00
5.42915344e-01 4.50306028e-01 -2.82097578e-01 4.58165556e-01
-4.66381639e-01 3.83196115e-01 5.47734797e-01 -1.05871491e-01
2.43890822e-01 -5.11176467e-01 5.34736644e-03 4.93012011e-01
7.07450390e-01 -1.14059709e-01 -3.45502585e-01 1.79704487e+00
3.41521800e-01 -2.15281677e-02 -9.35602367e-01 -7.90307447e-02
4.66127217e-01 1.19861519e+00 -4.31940198e-01 -1.77409902e-01
-1.92559466e-01 4.69632596e-01 -3.79625857e-02 3.38566452e-01
-6.49674296e-01 -1.14123374e-01 7.75315404e-01 9.73209590e-02
-7.76011404e-03 -8.74130204e-02 -1.49885106e+00 -4.59183276e-01
-2.33348995e-01 -9.22334433e-01 9.73252431e-02 -1.17384136e+00
-1.35083878e+00 3.34136307e-01 -7.79489502e-02 -1.10819471e+00
-6.07938111e-01 -7.95528591e-01 -8.96003067e-01 9.75815296e-01
-1.43027866e+00 -1.22192454e+00 -6.44774616e-01 7.47736096e-02
4.41340327e-01 -5.21099567e-01 7.24937916e-01 -2.87530363e-01
-5.11452615e-01 -4.62419763e-02 -7.20911706e-03 -3.57151896e-01
6.49928227e-02 -1.43011606e+00 2.96260361e-02 6.15571618e-01
-3.11172724e-01 6.02974594e-01 7.75772750e-01 -9.03731227e-01
-1.26632392e+00 -8.14390659e-01 9.89764035e-01 -1.95453912e-01
5.00643730e-01 4.24232066e-01 -9.26846147e-01 7.19303489e-01
5.31433702e-01 -7.81961083e-01 1.05249035e+00 2.73007840e-01
4.73238289e-01 -2.29930967e-01 -1.25187099e+00 5.75856030e-01
6.42871022e-01 1.83907861e-04 -8.30122828e-01 -1.08612306e-01
8.05883825e-01 4.11224037e-01 -1.41109502e+00 5.37516713e-01
3.34326178e-01 -1.22046340e+00 6.31660223e-01 -1.03516185e+00
1.15828443e+00 -1.45114884e-01 -1.35426402e-01 -1.36901021e+00
-7.12021232e-01 -1.91165090e-01 -4.20137107e-01 1.13234985e+00
5.97170711e-01 -5.22675812e-01 7.30812848e-01 1.29360175e+00
-5.08443654e-01 -1.05915034e+00 -5.99685729e-01 -2.80249774e-01
-1.59152970e-01 -6.15258992e-01 1.35973644e+00 9.96649921e-01
1.64365068e-01 1.92894518e-01 -1.48265317e-01 2.72134393e-01
4.65530187e-01 2.91101515e-01 8.46642435e-01 -1.80031502e+00
2.68777430e-01 -5.29826880e-01 -4.53014165e-01 -2.50569224e-01
7.36380741e-02 -8.34226191e-01 -4.43658441e-01 -2.21638680e+00
-2.31413826e-01 -1.47803426e-01 -7.32547939e-01 4.12731409e-01
-1.30233780e-01 -7.99526691e-01 4.77145582e-01 5.79176731e-02
-1.50022894e-01 9.35920060e-01 1.54363942e+00 -1.46705136e-01
-2.60362864e-01 2.49647349e-01 -1.21266568e+00 6.90189004e-01
1.03404737e+00 -1.33386388e-01 -4.90084201e-01 2.44970471e-02
4.20819581e-01 3.13720137e-01 3.85340452e-01 -1.14688790e+00
-9.69435200e-02 -4.77219671e-01 8.13804567e-01 -6.81941271e-01
-1.36178806e-01 -1.24334502e+00 4.60141897e-01 7.70791352e-01
-7.53249466e-01 4.10412669e-01 -3.25995348e-02 4.68111217e-01
-4.77624893e-01 -1.61120430e-01 1.14825383e-01 -4.48713936e-02
-8.34914267e-01 -5.55526577e-02 -5.24634242e-01 -9.57477570e-01
1.34557164e+00 -5.23451626e-01 -9.44063589e-02 -4.33323950e-01
-9.85607743e-01 4.17543113e-01 5.14247939e-02 3.80872607e-01
8.82927120e-01 -1.48089719e+00 -4.92936432e-01 3.12437832e-01
1.21129639e-01 -6.07183464e-02 3.94580096e-01 6.04060829e-01
-5.51797092e-01 7.34580100e-01 -8.46055210e-01 -1.80495381e-01
-5.41148126e-01 8.88266504e-01 2.48397738e-02 -6.44518256e-01
-2.89198130e-01 6.93195760e-02 -1.50781095e-01 -5.48641682e-01
-2.22269502e-02 -7.79299557e-01 -5.69459498e-01 8.97077471e-02
8.41454193e-02 7.78743505e-01 3.22439075e-02 1.27423674e-01
-2.18257289e-02 -1.10240988e-01 3.38704824e-01 7.06397444e-02
1.44830525e+00 -1.54646918e-01 1.38756588e-01 5.25907159e-01
3.81377429e-01 -5.41663110e-01 -1.30821478e+00 2.85744458e-01
5.10910451e-01 -4.21429098e-01 9.04650148e-03 -1.26136518e+00
-8.18367004e-01 1.08875251e+00 5.25981843e-01 5.18688440e-01
1.08354211e+00 -1.31927341e-01 2.26555973e-01 5.68256140e-01
1.56139195e-01 -1.17691147e+00 2.20613018e-01 7.82538429e-02
1.08216012e+00 -1.39873970e+00 -1.97075158e-02 -2.36833081e-01
-1.10252368e+00 1.45340979e+00 8.18152726e-01 1.88333258e-01
3.83348376e-01 -1.77681804e-01 1.86983831e-02 -6.45173013e-01
-7.55828023e-01 1.43061401e-02 4.86010849e-01 7.80356467e-01
4.85874355e-01 4.64483678e-01 -1.27440095e-01 1.39938951e+00
-4.19359207e-01 3.57877403e-01 2.46764436e-01 6.09996676e-01
-3.72940809e-01 -6.62064970e-01 -4.44701970e-01 6.82911932e-01
-1.87144965e-01 1.28870115e-01 -4.02083814e-01 1.13879824e+00
2.30392545e-01 1.24554193e+00 1.38382778e-01 -6.60773873e-01
5.72475195e-01 1.57894075e-01 -1.49878711e-01 -2.17224821e-01
-6.43888891e-01 -4.73219782e-01 3.08898278e-02 -5.42150676e-01
-2.14749828e-01 -4.03189361e-01 -1.40339124e+00 -4.60245371e-01
-1.36292502e-01 -1.84862260e-02 8.62846613e-01 1.34446967e+00
5.75707972e-01 1.51745296e+00 6.16647422e-01 -5.76747656e-01
-6.90955937e-01 -1.01259184e+00 -4.83721703e-01 5.00806212e-01
-2.07791284e-01 -7.51544297e-01 -1.74821436e-01 -3.21527869e-01] | [8.662771224975586, 5.939888000488281] |
db57d7b3-d428-4351-bd10-73a099aa7b9e | automl-in-the-wild-obstacles-workarounds-and | 2302.10827 | null | https://arxiv.org/abs/2302.10827v1 | https://arxiv.org/pdf/2302.10827v1.pdf | AutoML in The Wild: Obstacles, Workarounds, and Expectations | Automated machine learning (AutoML) is envisioned to make ML techniques accessible to ordinary users. Recent work has investigated the role of humans in enhancing AutoML functionality throughout a standard ML workflow. However, it is also critical to understand how users adopt existing AutoML solutions in complex, real-world settings from a holistic perspective. To fill this gap, this study conducted semi-structured interviews of AutoML users (N = 19) focusing on understanding (1) the limitations of AutoML encountered by users in their real-world practices, (2) the strategies users adopt to cope with such limitations, and (3) how the limitations and workarounds impact their use of AutoML. Our findings reveal that users actively exercise user agency to overcome three major challenges arising from customizability, transparency, and privacy. Furthermore, users make cautious decisions about whether and how to apply AutoML on a case-by-case basis. Finally, we derive design implications for developing future AutoML solutions. | ['Ting Wang', 'Fenglong Ma', 'Xinning Gui', 'Qiurong Song', 'Yuan Sun'] | 2023-02-21 | null | null | null | null | ['automl'] | ['methodology'] | [ 1.51764760e-02 5.86665213e-01 -1.08835243e-01 -7.28012681e-01
-6.99974000e-01 -9.41628039e-01 3.56121451e-01 6.26956522e-01
-6.71701252e-01 4.06054765e-01 6.09799743e-01 -8.83219719e-01
3.80472057e-02 -9.53190494e-03 -1.19430512e-01 9.59517956e-02
7.53744900e-01 1.11235254e-01 -6.22742236e-01 3.19235355e-01
5.06501436e-01 8.75495851e-01 -8.10382068e-01 5.23146927e-01
8.86336327e-01 1.99517161e-01 -1.60793468e-01 4.04024929e-01
-4.31740552e-01 1.12186706e+00 -6.42236769e-01 -4.98512715e-01
4.57737237e-01 -2.26989061e-01 -8.96345854e-01 3.84673059e-01
6.38851762e-01 -8.72827888e-01 3.39853495e-01 7.37778008e-01
3.29795837e-01 3.16359736e-02 2.24527717e-01 -1.47474635e+00
-6.12122655e-01 4.08224612e-01 -1.23846114e-01 -2.33451501e-01
7.27063537e-01 8.68755996e-01 6.78630114e-01 -4.99215394e-01
5.89006007e-01 1.12980449e+00 8.17278266e-01 4.16411817e-01
-1.59298289e+00 -7.81578004e-01 7.25429356e-02 -4.13300455e-01
-1.24295771e+00 -8.51414919e-01 3.23586941e-01 -9.29263115e-01
1.01653945e+00 3.10322672e-01 7.59478927e-01 1.02899826e+00
9.40184072e-02 4.51595187e-01 1.59519696e+00 -7.00781465e-01
2.31606364e-01 1.20183468e+00 4.20357972e-01 5.33678532e-01
7.23734915e-01 -7.33592272e-01 -3.60523790e-01 -7.11927831e-01
4.25821841e-01 4.13471088e-02 -1.85082123e-01 -6.60930499e-02
-9.71963584e-01 7.16519773e-01 -1.83818128e-03 5.03554702e-01
-4.25224066e-01 -3.82104576e-01 1.81057185e-01 1.04378276e-01
3.98789123e-02 1.22564542e+00 -4.43209141e-01 -6.56451821e-01
-5.70324898e-01 3.89539674e-02 1.14802289e+00 1.08030307e+00
8.24681282e-01 -3.71806920e-01 -1.57086446e-03 5.07569671e-01
3.44581693e-01 9.13614631e-02 -1.04305716e-02 -1.14635897e+00
2.99682438e-01 1.07076049e+00 6.39737248e-01 -7.89891183e-01
-1.25136718e-01 -2.31263831e-01 -9.09220800e-02 2.54944891e-01
8.19521785e-01 -3.97878289e-01 -2.41165921e-01 1.21056676e+00
2.69853193e-02 -6.10399365e-01 -2.01479614e-01 5.31067848e-01
6.00777939e-02 1.95123598e-01 6.97207153e-01 -3.08380991e-01
1.15950274e+00 -1.97812408e-01 -9.71803188e-01 -7.40166485e-01
1.09942818e+00 -1.12796760e+00 1.61591256e+00 3.52817625e-02
-1.04789436e+00 -2.93879122e-01 -6.97845757e-01 -2.46756315e-01
-2.53420591e-01 2.88566887e-01 6.05154753e-01 1.27242517e+00
-6.51507854e-01 3.07969719e-01 -8.88712227e-01 -1.19682753e+00
4.06388551e-01 1.15437090e-01 -3.23673934e-01 -1.18555427e-01
-5.88527560e-01 1.00961173e+00 1.08565815e-01 3.72217536e-01
4.72921617e-02 -7.42483258e-01 -8.46157670e-01 -8.23329911e-02
3.99768978e-01 -4.22680020e-01 1.95635104e+00 -8.02985311e-01
-7.70518541e-01 6.57503724e-01 -2.51758881e-02 -2.81456746e-02
7.60786474e-01 -3.08089316e-01 -1.83622420e-01 -5.09698272e-01
3.36370766e-01 2.26638749e-01 6.53895661e-02 -1.34569514e+00
-8.27323139e-01 -4.08854961e-01 1.15876265e-01 1.71109572e-01
-4.18606460e-01 4.11185920e-02 3.96893770e-02 -6.00265265e-02
-1.44349113e-01 -9.12268281e-01 -2.19295338e-01 2.20936954e-01
-2.99568087e-01 5.45150377e-02 7.06954479e-01 -7.60904491e-01
1.49914062e+00 -2.21919656e+00 -7.39669323e-01 2.69059360e-01
5.57996869e-01 4.48771834e-01 2.09653497e-01 1.23292470e+00
5.56212842e-01 7.22982407e-01 1.27269417e-01 -2.05706120e-01
1.01399302e-01 -6.20435402e-02 1.42366767e-01 4.34282422e-01
-2.46793330e-01 5.60296893e-01 -9.04853225e-01 -5.99755645e-01
2.01449856e-01 4.95437235e-01 -1.28225699e-01 3.22940797e-01
2.09896430e-01 3.13381463e-01 -4.62730438e-01 8.72555971e-01
6.96731091e-01 -1.60845160e-01 5.99439561e-01 -1.70771569e-01
-7.12204218e-01 2.90885895e-01 -9.90351021e-01 1.05887294e+00
-6.02422118e-01 7.05946088e-01 5.93086779e-01 6.42166734e-02
5.09312630e-01 2.61789441e-01 -1.04534850e-02 -1.68981478e-01
6.83917254e-02 2.84279346e-01 1.42320514e-01 -8.49799871e-01
1.66080669e-01 1.27543092e-01 1.19131029e-01 9.59484458e-01
-4.22410101e-01 -1.78868510e-02 -2.15845510e-01 1.31918177e-01
1.00608504e+00 5.56333847e-02 1.05110776e+00 -8.63108486e-02
4.22403775e-02 1.11302212e-01 5.65136671e-01 1.09188879e+00
-7.08029509e-01 -1.10108934e-01 4.09121990e-01 -5.14300227e-01
-7.38088608e-01 -4.97874528e-01 2.98412055e-01 7.18475580e-01
-4.50976014e-01 -6.96305633e-01 -9.26916659e-01 -8.46517742e-01
2.66313493e-01 1.48086488e+00 -6.73739389e-02 1.33584633e-01
-6.99662343e-02 -2.65144296e-02 3.54629666e-01 2.84581661e-01
7.96617866e-01 -8.85587335e-01 -1.42573738e+00 4.78375375e-01
-3.50719124e-01 -1.27576125e+00 -6.30900264e-01 -5.51446199e-01
-5.80809355e-01 -9.96422648e-01 1.40133113e-01 -6.54178560e-02
7.79884279e-01 4.61921245e-01 6.70940578e-01 1.98259518e-01
-6.11192405e-01 8.69312704e-01 -1.13968462e-01 -5.40233433e-01
-7.17614651e-01 1.30644858e-01 -3.30637097e-01 -1.54454708e-01
9.90471542e-01 -1.67406783e-01 -3.14039111e-01 1.67100251e-01
-7.31393099e-01 3.72107178e-01 8.87866914e-01 9.88997221e-02
-2.91658163e-01 -1.90759331e-01 4.78741974e-01 -1.51343191e+00
1.39032626e+00 -1.91128790e-01 -2.83915818e-01 5.57543516e-01
-1.15568423e+00 -4.50685471e-01 5.56483030e-01 -1.63278952e-01
-1.22300053e+00 6.55735284e-02 2.24686429e-01 2.82412946e-01
-7.04055130e-01 4.90335792e-01 -2.49969065e-01 -1.94324687e-01
9.52160060e-01 -4.39313203e-01 4.39451396e-01 -3.13407987e-01
1.01856254e-01 1.29364431e+00 -1.76973268e-01 -5.06204367e-01
7.22860217e-01 1.59574717e-01 -5.82696915e-01 -9.17203784e-01
-7.64987469e-01 -4.30170655e-01 -8.76111448e-01 -2.50984102e-01
4.95780945e-01 -6.10908687e-01 -1.18185008e+00 6.78876340e-02
-8.18756342e-01 -7.84982681e-01 -6.96148798e-02 6.90796793e-01
-1.83821529e-01 7.40620419e-02 -2.70772904e-01 -1.41908431e+00
-1.90746352e-01 -1.06633425e+00 4.24086660e-01 2.62406498e-01
-1.46499145e+00 -8.07299793e-01 -4.48212534e-01 9.15996730e-01
7.36917496e-01 3.71560514e-01 1.04826808e+00 -6.85310721e-01
-6.83877349e-01 -5.14515400e-01 -2.76440620e-01 1.61435276e-01
7.49500036e-01 8.15964863e-02 -1.00162852e+00 -4.25129890e-01
8.81127343e-02 -1.42585829e-01 -4.14625853e-01 -2.41958767e-01
6.90910399e-01 -1.00773227e+00 -2.56357044e-01 6.34129047e-02
1.43353093e+00 4.69785035e-01 2.67625377e-02 6.10210657e-01
5.61273754e-01 8.14443290e-01 8.40978920e-01 5.10638177e-01
3.18247885e-01 6.31081536e-02 -1.88973904e-01 -3.57617810e-03
4.85998780e-01 -4.18285549e-01 -2.49926634e-02 -1.04387496e-02
2.70735115e-01 1.15594395e-01 -1.33879757e+00 3.01970482e-01
-1.95371842e+00 -5.69627285e-01 1.37721181e-01 2.30722117e+00
4.65654671e-01 3.15946221e-01 1.64179683e-01 -2.62831479e-01
3.47338229e-01 -9.57345143e-02 -5.00582695e-01 -7.33854115e-01
7.51041055e-01 -3.48615170e-01 5.07893085e-01 4.67481971e-01
-5.96263528e-01 6.98279679e-01 5.73919868e+00 -3.01946968e-01
-9.65108275e-01 1.10197373e-01 6.44892454e-01 -4.17493731e-02
-3.99940610e-01 4.99705613e-01 -6.90734386e-01 1.42351851e-01
6.88375175e-01 -3.45134169e-01 6.05641365e-01 9.50067103e-01
5.70313632e-01 -9.00025517e-02 -1.38225865e+00 7.38563538e-01
-1.68391839e-01 -1.17685986e+00 -3.76942366e-01 3.62159699e-01
1.51492879e-01 -5.00217855e-01 -1.01744093e-01 2.85701394e-01
1.34390429e-01 -1.04675913e+00 4.93582547e-01 3.09417576e-01
6.64559841e-01 -5.48336804e-01 6.39698088e-01 6.73645914e-01
-6.86324835e-01 -2.70213395e-01 2.00889930e-01 -6.19352043e-01
1.22163370e-01 6.54974729e-02 -1.77493477e+00 -2.24563450e-01
4.54681009e-01 1.44237742e-01 -6.19802594e-01 6.83525741e-01
-1.33629397e-01 5.66083848e-01 -1.31168514e-01 1.75575446e-02
4.25371915e-01 -2.38965094e-01 7.29300827e-02 1.65185905e+00
-4.80334088e-02 5.31612821e-02 7.78965726e-02 8.33159506e-01
2.58659899e-01 1.41212046e-01 -9.47442174e-01 -8.09257507e-01
1.07532179e+00 1.50541615e+00 -6.28170133e-01 2.59352159e-02
-7.02146173e-01 7.24134147e-01 2.58568078e-01 6.44926608e-01
-6.40898794e-02 -2.75134385e-01 8.08212936e-01 8.62608433e-01
-6.26147807e-01 -5.10789275e-01 -4.12429094e-01 -5.57125926e-01
8.16135630e-02 -1.43501461e+00 1.30080387e-01 -5.78996778e-01
-1.10922396e+00 1.82595655e-01 4.40569967e-02 -8.63430977e-01
-2.17628613e-01 -2.13014930e-01 -2.32043982e-01 1.08476949e+00
-7.52983451e-01 -1.37102246e+00 -6.32570446e-01 -9.91945565e-02
3.25820476e-01 1.60104588e-01 1.00414383e+00 -3.81663330e-02
-5.66092610e-01 5.07700026e-01 -2.92454630e-01 -1.26497909e-01
1.12899685e+00 -1.03015172e+00 5.29888630e-01 8.13453197e-01
-4.50428486e-01 1.67270195e+00 5.98044693e-01 -9.50102627e-01
-1.60952663e+00 -8.87999713e-01 1.28596997e+00 -6.43553197e-01
1.24776885e-01 -6.75040066e-01 -9.75280464e-01 1.51842427e+00
3.69694471e-01 -4.82025027e-01 1.33812630e+00 3.12462866e-01
-6.96739331e-02 1.82605740e-02 -1.86159039e+00 1.02876568e+00
8.89191985e-01 -8.66047800e-01 -2.21469119e-01 2.41589949e-01
2.79793888e-01 -2.33548433e-01 -9.00039911e-01 -7.47036710e-02
9.64845717e-01 -1.14382315e+00 2.13496864e-01 -4.80935007e-01
-1.20713472e-01 -1.91843361e-01 2.52959162e-01 -8.98015261e-01
-3.66780221e-01 -1.10395241e+00 5.87506354e-01 1.56863022e+00
2.34070629e-01 -9.75502074e-01 4.68336642e-01 2.10898423e+00
2.22868741e-01 -5.54421365e-01 -2.80287087e-01 4.56135757e-02
-5.34688413e-01 -5.34923673e-01 7.23663807e-01 1.44733119e+00
2.63522834e-01 1.22757768e-03 -2.14975342e-01 2.71767139e-01
3.30973119e-01 -4.39670801e-01 1.16894066e+00 -1.01694930e+00
-1.48751840e-01 -1.20035194e-01 -8.42041001e-02 -4.94058162e-01
-2.49613091e-01 -4.91788745e-01 -2.82805711e-01 -1.91155279e+00
1.63327530e-01 -3.93735766e-01 2.86454469e-01 8.04509759e-01
-9.46804732e-02 -6.95803285e-01 8.03856075e-01 5.21485567e-01
-5.27772345e-02 -3.08220297e-01 5.15387118e-01 4.82464105e-01
-7.61735201e-01 1.48443118e-01 -1.57297444e+00 8.13614130e-01
8.83159816e-01 -2.56380796e-01 -6.79669142e-01 -2.61828065e-01
1.82046518e-01 -2.08160132e-01 9.61003825e-02 -7.87869453e-01
4.54888821e-01 -8.01655829e-01 3.69826555e-01 -2.74278760e-01
5.58052817e-03 -1.31283069e+00 4.01074827e-01 3.70855272e-01
-4.16344911e-01 -3.90041359e-02 4.56724197e-01 -2.52562374e-01
3.32375616e-01 -5.27043104e-01 5.56948304e-01 -3.61745536e-01
-1.07378714e-01 -3.24283034e-01 -1.08662152e+00 -1.39227539e-01
1.27463555e+00 -5.03648162e-01 -3.88410866e-01 -8.02789211e-01
-2.67786056e-01 3.37180138e-01 9.25716281e-01 1.08145839e-02
3.02187502e-01 -7.14773476e-01 -1.44061148e-01 4.21046048e-01
3.53282064e-01 -7.76400566e-02 2.69972775e-02 7.58435190e-01
-7.56410897e-01 5.01709878e-01 -2.85172313e-01 5.86942881e-02
-1.29891002e+00 1.16236787e-02 3.50514472e-01 2.89936721e-01
-5.26766777e-01 3.90617281e-01 -1.29119948e-01 -5.68830371e-01
4.71648574e-01 -2.05498040e-01 2.91579872e-01 3.64029408e-02
3.69240552e-01 5.60502708e-01 -1.70495823e-01 -2.06200212e-01
-3.73371691e-01 1.21902183e-01 -5.42655408e-01 -4.54362094e-01
8.49207580e-01 -2.63060987e-01 -6.46802038e-02 7.43443251e-01
8.28614652e-01 3.53423983e-01 -1.24848700e+00 -1.51881203e-01
2.25395903e-01 -1.07680976e+00 -4.94708329e-01 -1.04268229e+00
-8.25867429e-02 5.41778564e-01 1.07989475e-01 2.42035329e-01
8.11729074e-01 -2.49584481e-01 4.37670350e-01 8.09399307e-01
1.94401756e-01 -1.20936179e+00 -3.46072197e-01 -1.77902341e-01
9.21386361e-01 -1.07274950e+00 1.32260993e-01 -4.62263793e-01
-1.04367661e+00 1.18670952e+00 7.71472752e-01 9.32206631e-01
4.74473268e-01 1.05476320e-01 5.67620218e-01 -8.46736580e-02
-5.07443964e-01 4.42962795e-01 -4.20873284e-01 5.01705587e-01
1.05584586e+00 1.69128299e-01 -6.81775928e-01 2.92820007e-01
-4.72869836e-02 3.95504683e-01 8.41530204e-01 1.63723683e+00
-2.18694851e-01 -1.10349131e+00 -5.04882991e-01 6.64538264e-01
-2.76247770e-01 3.10490459e-01 -8.92394483e-01 1.14427245e+00
1.21175990e-01 1.25746787e+00 -2.98043162e-01 -1.98499471e-01
6.33448958e-01 5.15484095e-01 -5.55097982e-02 -1.04445338e+00
-8.97468030e-01 5.85530605e-03 4.36717033e-01 -3.05887550e-01
2.08932713e-01 -9.62435961e-01 -8.72439265e-01 -1.89677894e-01
1.68633446e-01 -8.12874436e-02 8.73290777e-01 1.00630462e+00
7.42210448e-01 -9.67605859e-02 2.24284291e-01 -9.02298093e-02
-7.41770089e-01 -1.12979305e+00 -2.57730812e-01 4.35323656e-01
3.22111994e-01 4.32593264e-02 -3.29583734e-01 -1.17563471e-01] | [9.707473754882812, 7.109508991241455] |
21fc7e68-aba2-4b01-8570-f4011ae9561c | sawu-net-spatial-attention-weighted-unmixing | 2304.11320 | null | https://arxiv.org/abs/2304.11320v1 | https://arxiv.org/pdf/2304.11320v1.pdf | SAWU-Net: Spatial Attention Weighted Unmixing Network for Hyperspectral Images | Hyperspectral unmixing is a critical yet challenging task in hyperspectral image interpretation. Recently, great efforts have been made to solve the hyperspectral unmixing task via deep autoencoders. However, existing networks mainly focus on extracting spectral features from mixed pixels, and the employment of spatial feature prior knowledge is still insufficient. To this end, we put forward a spatial attention weighted unmixing network, dubbed as SAWU-Net, which learns a spatial attention network and a weighted unmixing network in an end-to-end manner for better spatial feature exploitation. In particular, we design a spatial attention module, which consists of a pixel attention block and a window attention block to efficiently model pixel-based spectral information and patch-based spatial information, respectively. While in the weighted unmixing framework, the central pixel abundance is dynamically weighted by the coarse-grained abundances of surrounding pixels. In addition, SAWU-Net generates dynamically adaptive spatial weights through the spatial attention mechanism, so as to dynamically integrate surrounding pixels more effectively. Experimental results on real and synthetic datasets demonstrate the better accuracy and superiority of SAWU-Net, which reflects the effectiveness of the proposed spatial attention mechanism. | ['Xinbo Gao', 'Junyu Dong', 'Feng Gao', 'Xuewen Qin', 'Lin Qi'] | 2023-04-22 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 5.26300371e-01 -4.44344312e-01 1.06710389e-01 -1.18072525e-01
-2.80788779e-01 -1.34773880e-01 4.28473532e-01 -1.14210851e-01
-3.04215044e-01 4.90467936e-01 2.85253108e-01 -1.43699810e-01
-4.23463643e-01 -9.60131824e-01 -5.92720628e-01 -1.26996505e+00
2.64776260e-01 -1.71547934e-01 -9.29885507e-02 -1.82341650e-01
-2.56100781e-02 2.35424772e-01 -1.38104188e+00 -9.33745801e-02
1.30214643e+00 1.12122381e+00 7.78000116e-01 3.60727221e-01
-6.45999759e-02 4.54537272e-01 -1.67055622e-01 4.24482137e-01
3.84782583e-01 -4.11518246e-01 -2.00426966e-01 5.47564745e-01
3.66118491e-01 -4.15837497e-01 -6.09374821e-01 1.45798349e+00
4.79310662e-01 4.53542501e-01 4.27414805e-01 -9.45906997e-01
-9.48565245e-01 6.81579649e-01 -1.05549276e+00 3.33686650e-01
-4.99744505e-01 4.63193774e-01 8.57410014e-01 -7.03668416e-01
-7.90911019e-02 9.88396704e-01 4.41078216e-01 -7.01775178e-02
-1.08935153e+00 -8.02052796e-01 5.80436766e-01 1.38060227e-01
-1.59899378e+00 -1.35507897e-01 9.30009425e-01 -4.32869703e-01
6.48714244e-01 2.36669034e-01 8.71917963e-01 3.51084679e-01
-9.92010310e-02 7.50759304e-01 8.71874332e-01 -2.37908289e-01
-2.45660599e-02 -1.48456544e-01 3.37902270e-02 3.05705547e-01
3.06802928e-01 2.16323569e-01 -1.62427336e-01 9.57662389e-02
9.61989880e-01 6.54971182e-01 -6.73240483e-01 -2.17759162e-01
-1.24915254e+00 6.17427111e-01 1.20430374e+00 3.60950619e-01
-8.70205820e-01 1.76517870e-02 -7.83455223e-02 -4.78342921e-02
7.97140777e-01 2.50185996e-01 -2.75868684e-01 5.17653644e-01
-1.06036198e+00 1.27853274e-01 -1.97994992e-01 5.00971735e-01
1.11980700e+00 1.62574679e-01 -1.02241166e-01 1.09314954e+00
5.15308857e-01 6.09641850e-01 6.97261095e-01 -4.96684968e-01
4.65319246e-01 8.63045931e-01 1.07700966e-01 -1.09009683e+00
-3.82172316e-01 -6.75246239e-01 -1.17626929e+00 2.21404210e-01
-1.95194155e-01 -3.08273822e-01 -1.03013456e+00 1.68073618e+00
1.70421481e-01 5.49471796e-01 7.07040653e-02 1.29905438e+00
5.66510737e-01 1.13531280e+00 3.03219169e-01 -3.00976306e-01
1.28242469e+00 -1.16274488e+00 -7.90720463e-01 -5.53829074e-01
2.05285013e-01 -4.60286856e-01 8.01470518e-01 -3.69327031e-02
-8.88443470e-01 -5.43994129e-01 -1.26484191e+00 1.15820639e-01
-5.01920760e-01 2.85460263e-01 6.65242195e-01 1.74664304e-01
-6.61258399e-01 4.58404601e-01 -6.06663346e-01 -1.54587984e-01
7.52715886e-01 2.41382122e-01 -1.47356540e-01 -1.69707593e-02
-1.18429720e+00 4.64864880e-01 7.85344303e-01 6.49188101e-01
-6.88118279e-01 -6.15563869e-01 -8.45098376e-01 4.17830914e-01
3.21037829e-01 -4.61425900e-01 9.05199230e-01 -1.33803868e+00
-1.31897390e+00 2.37648994e-01 -5.57481274e-02 -4.71029058e-02
1.27057061e-02 5.11889718e-02 -6.05101109e-01 1.24152929e-01
5.67336120e-02 7.09410429e-01 8.89720976e-01 -1.23524678e+00
-9.46383953e-01 -5.65117240e-01 -6.00172952e-02 6.19276285e-01
-8.26559901e-01 -2.91091353e-01 -1.98671415e-01 -1.02677631e+00
4.01587099e-01 -5.62816858e-01 -3.90433192e-01 -3.12669985e-02
-1.74438149e-01 1.70546666e-01 9.42657828e-01 -8.14169705e-01
1.33981049e+00 -2.25536203e+00 4.00746047e-01 1.83107018e-01
3.04293364e-01 5.21298289e-01 -4.20252919e-01 -2.34527644e-02
-5.60095489e-01 -6.22156914e-03 -8.28530669e-01 1.67609259e-01
-1.58324093e-01 -7.40648359e-02 -3.04516435e-01 4.56406206e-01
4.00458783e-01 8.69380474e-01 -9.85961616e-01 -5.62119521e-02
4.12371933e-01 5.26447475e-01 -3.41165394e-01 2.38324344e-01
-4.75400150e-01 2.96863765e-01 -4.56568956e-01 8.03091407e-01
1.12139308e+00 -3.33735198e-01 -4.82703783e-02 -4.40148234e-01
-5.09519994e-01 -3.40899825e-01 -1.24359035e+00 1.63452232e+00
-3.72012556e-01 3.47928256e-01 4.17743832e-01 -9.41652775e-01
7.65246868e-01 1.91308439e-01 4.48217720e-01 -7.49016285e-01
1.41814947e-01 2.81986296e-01 1.65321417e-02 -5.88073671e-01
5.85049629e-01 -4.60175246e-01 5.42296290e-01 4.82716888e-01
-1.87828258e-01 6.04422949e-02 -2.09975660e-01 -2.85127282e-01
4.49215412e-01 1.13031343e-01 1.65577054e-01 -3.87077123e-01
6.70399547e-01 8.11967105e-02 4.43359613e-01 1.82598427e-01
-1.67667612e-01 5.52548885e-01 -7.27726445e-02 -3.64384174e-01
-8.38826358e-01 -6.30820274e-01 -1.69928744e-01 1.19070411e+00
5.82171381e-01 1.93647951e-01 -5.20740867e-01 -2.61703163e-01
-3.63482386e-02 2.76773334e-01 -6.96150839e-01 -2.61277229e-01
-2.88111925e-01 -1.38481534e+00 1.66729033e-01 6.18939221e-01
8.77127528e-01 -1.17994964e+00 -3.50295126e-01 4.39396292e-01
-1.17922634e-01 -3.95820320e-01 -6.27870023e-01 1.67390466e-01
-8.17669570e-01 -8.78908396e-01 -1.11275721e+00 -7.40642369e-01
6.66659653e-01 9.64425445e-01 4.15859580e-01 -6.52194843e-02
-2.50018477e-01 -3.27843785e-01 -2.76432574e-01 -4.95532840e-01
3.37761939e-01 2.58599818e-01 -2.97507286e-01 5.57040751e-01
4.01205391e-01 -7.89192975e-01 -9.26712036e-01 2.24592790e-01
-1.64154124e+00 1.43723011e-01 9.24725115e-01 9.22391176e-01
6.02566302e-01 4.46371704e-01 3.93056512e-01 -5.52868843e-01
4.56859738e-01 -7.40187943e-01 -6.91211820e-01 2.37956345e-01
-3.67339134e-01 -8.93891603e-02 3.82839322e-01 -4.81327176e-01
-1.19032693e+00 4.82138135e-02 -7.39111826e-02 -5.10219574e-01
4.88410220e-02 9.24618781e-01 -5.94850123e-01 -1.16062976e-01
3.95218492e-01 5.81897974e-01 1.30280808e-01 -3.43059421e-01
2.64566302e-01 1.06908357e+00 4.67274755e-01 -1.63543060e-01
8.18748891e-01 5.61375439e-01 -3.56575489e-01 -8.55071008e-01
-7.97413707e-01 -5.57190835e-01 -3.63287628e-01 -3.48839387e-02
9.55873668e-01 -1.28963280e+00 -3.74337703e-01 8.55612159e-01
-6.66780412e-01 -5.42150438e-01 -1.03205323e-01 4.43656534e-01
-9.44465995e-02 3.94034505e-01 -2.72990406e-01 -7.24746048e-01
-3.82585675e-01 -1.13930058e+00 1.07648945e+00 6.13654792e-01
4.22684848e-01 -9.72647786e-01 -1.02029108e-01 2.24763632e-01
5.86871088e-01 1.68243587e-01 7.95108318e-01 6.37025386e-03
-6.07350230e-01 -4.82510105e-02 -8.78371954e-01 3.62937480e-01
3.99208575e-01 -1.12066135e-01 -1.09555542e+00 -1.69752091e-01
-4.66287631e-04 -5.41156568e-02 1.27569103e+00 6.38269365e-01
1.48508024e+00 -2.27553502e-01 -2.68079281e-01 9.49542582e-01
1.69839823e+00 3.53390515e-01 6.43897951e-01 4.82564807e-01
9.69583631e-01 5.49847186e-01 4.19454664e-01 4.73013252e-01
1.09136224e-01 2.63104230e-01 9.93768513e-01 -4.14257199e-01
1.79046899e-01 -5.99747822e-02 -9.16008055e-02 5.54033041e-01
-2.80464143e-02 -1.29038140e-01 -6.90050542e-01 5.55423498e-01
-2.14515162e+00 -1.04604137e+00 -1.57321170e-02 2.01841760e+00
5.88217676e-01 -4.53241080e-01 -1.64159223e-01 5.16551957e-02
9.40264642e-01 7.90457606e-01 -8.21863830e-01 4.86558676e-01
-5.35842180e-01 3.44669484e-02 6.01942480e-01 4.01830047e-01
-1.33716452e+00 7.97847211e-01 5.10884714e+00 9.18127000e-01
-1.44314194e+00 1.44387588e-01 5.99701822e-01 -1.42713621e-01
-4.75110769e-01 -1.18136823e-01 -2.41533637e-01 6.31352007e-01
2.79812723e-01 1.57716960e-01 7.55097449e-01 4.63655442e-01
4.04106557e-01 1.19096629e-01 -1.96710363e-01 9.73061502e-01
7.08114430e-02 -1.17755842e+00 1.04865506e-01 2.29862496e-01
8.40574801e-01 5.82782328e-02 2.54372120e-01 9.78296846e-02
1.36818245e-01 -1.03535652e+00 5.33920407e-01 6.23604655e-01
6.18815660e-01 -7.63831496e-01 7.48834133e-01 2.94174433e-01
-1.46888971e+00 -4.08595353e-01 -6.70205176e-01 -7.97131956e-02
-6.57963827e-02 6.50090158e-01 7.15115219e-02 7.33924925e-01
4.48138148e-01 1.02108729e+00 -3.76226515e-01 1.08906317e+00
-9.58563909e-02 3.87247562e-01 -2.35762253e-01 2.71965533e-01
7.28508294e-01 -7.60963142e-01 2.39127547e-01 9.15020406e-01
5.14619231e-01 4.61696833e-01 1.87498212e-01 8.69281411e-01
-1.02887079e-01 1.73221335e-01 -2.86146939e-01 -2.03210413e-01
3.23855191e-01 1.45733511e+00 -4.02836055e-01 -2.40259811e-01
-2.93498665e-01 9.09334302e-01 9.39277485e-02 5.79175651e-01
-8.21146071e-01 -4.63344246e-01 9.28862929e-01 2.65368279e-02
4.19483542e-01 3.12341657e-02 -1.85327962e-01 -1.21096647e+00
-1.85004160e-01 -8.31716359e-01 2.89892703e-01 -1.21738052e+00
-1.16210461e+00 5.51393986e-01 -2.70959795e-01 -1.24893534e+00
5.64543903e-01 -7.35330403e-01 -9.18983877e-01 1.36925435e+00
-1.94497406e+00 -1.19390190e+00 -9.98118401e-01 3.95549834e-01
4.72312570e-01 -7.11761490e-02 4.61302042e-01 2.73831904e-01
-1.10600853e+00 -1.22858904e-01 3.76466662e-01 -1.08112462e-01
5.11137128e-01 -1.07672513e+00 -1.75423101e-01 9.47626054e-01
-2.66724169e-01 5.39681673e-01 2.90374905e-01 -3.94411772e-01
-1.08195150e+00 -1.75566781e+00 2.61760741e-01 2.68408865e-01
7.65759051e-01 3.06455731e-01 -1.13470852e+00 5.43942451e-01
3.69211376e-01 2.65954077e-01 7.52759874e-01 -2.63852209e-01
-1.21294372e-01 -4.67163831e-01 -8.02217245e-01 6.38338268e-01
8.14071417e-01 -3.71153563e-01 -3.32769066e-01 5.69225430e-01
7.29308307e-01 -1.65421262e-01 -6.81818724e-01 5.70422947e-01
4.43164527e-01 -7.08782613e-01 8.76096845e-01 -2.24540859e-01
6.45504177e-01 -7.80809879e-01 -1.39269203e-01 -1.65947104e+00
-8.27327609e-01 -6.68688640e-02 7.52502009e-02 9.58595216e-01
3.51510733e-01 -7.71514475e-01 5.73811352e-01 1.66961253e-01
-2.63238847e-01 -6.02179170e-01 -3.76115054e-01 -3.48747104e-01
-5.69466688e-02 -9.18317214e-02 1.11395562e+00 1.05957568e+00
-3.88523519e-01 1.76581994e-01 -3.10774684e-01 6.29273355e-01
3.83222759e-01 2.91561782e-01 2.99768567e-01 -1.10608304e+00
-1.51191711e-01 -9.62562859e-01 4.01023403e-02 -1.07750666e+00
1.18854225e-01 -7.26059556e-01 2.00860083e-01 -1.49916518e+00
3.08988482e-01 -3.01430315e-01 -7.17955351e-01 4.57142353e-01
-8.56497228e-01 4.47468340e-01 -6.74554706e-02 4.04692620e-01
-1.75512910e-01 9.46163595e-01 1.25660574e+00 -6.61222219e-01
-3.36983174e-01 -2.77244657e-01 -9.81973469e-01 5.36958277e-01
9.59372044e-01 -1.09671697e-01 -4.39128190e-01 -8.43202829e-01
8.02657828e-02 -1.00115970e-01 3.71531725e-01 -9.71879959e-01
1.59006953e-01 -5.03767252e-01 5.25652468e-01 -5.98456502e-01
1.85522512e-01 -9.45515394e-01 1.52247891e-01 2.06192181e-01
-6.73180073e-02 -3.61761302e-01 3.22015792e-01 6.12307310e-01
-4.56263632e-01 -2.36150503e-01 6.77200615e-01 -1.89793110e-01
-1.04937637e+00 9.23663676e-01 -1.85251310e-01 -5.35587609e-01
1.04241896e+00 -1.66406602e-01 -2.62859285e-01 9.40959156e-02
-4.30972070e-01 5.09211302e-01 4.06586766e-01 1.80905044e-01
5.81459880e-01 -1.31983960e+00 -6.94628239e-01 4.51227993e-01
4.51254487e-01 2.98217058e-01 8.37652624e-01 7.19594419e-01
-8.24677944e-01 1.76871687e-01 -2.21320033e-01 -4.08155590e-01
-8.29306245e-01 6.81913495e-01 6.67682409e-01 -1.05756316e-02
-2.19705433e-01 9.43687141e-01 5.23429632e-01 -4.41968352e-01
-1.28363624e-01 -3.34643155e-01 -4.25299942e-01 2.37766609e-01
7.69737303e-01 1.69355452e-01 -1.40642151e-01 -6.73008978e-01
-1.46968905e-02 5.69582760e-01 3.96765977e-01 8.34852457e-02
1.53385270e+00 -3.00133765e-01 -4.86256361e-01 1.92442805e-01
8.33770454e-01 -3.10292929e-01 -1.66913068e+00 -5.88234365e-01
-5.24009645e-01 -5.93034089e-01 7.48058379e-01 -6.77398682e-01
-1.40811837e+00 1.04792225e+00 6.64737642e-01 2.98492759e-01
1.53374803e+00 -4.42162126e-01 7.17661083e-01 2.31708735e-02
-2.72849888e-01 -7.82018185e-01 -1.13431208e-01 2.89354563e-01
7.68648505e-01 -1.43287802e+00 -4.25806530e-02 -2.49877810e-01
-4.49204177e-01 8.83093476e-01 8.01026285e-01 1.42788766e-02
7.41496086e-01 -1.42999053e-01 1.75615221e-01 -1.06180817e-01
-1.73918039e-01 -5.16116798e-01 2.39708573e-01 3.81585866e-01
2.40961403e-01 2.90839702e-01 1.47792056e-01 6.04535043e-01
2.90580899e-01 -1.20108388e-01 -6.82449490e-02 6.57176435e-01
-8.09650004e-01 -6.48211122e-01 -6.09468579e-01 5.28462172e-01
-5.64250071e-03 -3.68900329e-01 -6.43660827e-03 3.92280817e-01
4.81348366e-01 8.26610625e-01 2.96667427e-01 -4.53072995e-01
1.39907673e-01 -1.17583521e-01 -9.31836106e-03 -6.11169755e-01
-3.78132463e-01 3.20827335e-01 -6.81821406e-01 -6.21186523e-03
-6.60012960e-01 -3.90182436e-01 -7.31805980e-01 -1.18245110e-01
-5.00432968e-01 7.98500031e-02 5.62910140e-01 8.79295290e-01
2.02340081e-01 9.03478622e-01 7.21373379e-01 -1.23702073e+00
-4.46096569e-01 -1.47685504e+00 -1.02745986e+00 1.09364152e-01
5.02840579e-01 -4.53668773e-01 -2.21926987e-01 -1.38293654e-01] | [10.004490852355957, -1.741968035697937] |
0566a5bd-45e4-4a4a-9656-8878d50a92af | view-n-gram-network-for-3d-object-retrieval | 1908.01958 | null | https://arxiv.org/abs/1908.01958v2 | https://arxiv.org/pdf/1908.01958v2.pdf | View N-gram Network for 3D Object Retrieval | How to aggregate multi-view representations of a 3D object into an informative and discriminative one remains a key challenge for multi-view 3D object retrieval. Existing methods either use view-wise pooling strategies which neglect the spatial information across different views or employ recurrent neural networks which may face the efficiency problem. To address these issues, we propose an effective and efficient framework called View N-gram Network (VNN). Inspired by n-gram models in natural language processing, VNN divides the view sequence into a set of visual n-grams, which involve overlapping consecutive view sub-sequences. By doing so, spatial information across multiple views is captured, which helps to learn a discriminative global embedding for each 3D object. Experiments on 3D shape retrieval benchmarks, including ModelNet10, ModelNet40 and ShapeNetCore55 datasets, demonstrate the superiority of our proposed method. | ['Tengteng Huang', 'Song Bai', 'Xinwei He', 'Xiang Bai'] | 2019-08-06 | view-n-gram-network-for-3d-object-retrieval-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/He_View_N-Gram_Network_for_3D_Object_Retrieval_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/He_View_N-Gram_Network_for_3D_Object_Retrieval_ICCV_2019_paper.pdf | iccv-2019-10 | ['3d-shape-retrieval', '3d-object-retrieval'] | ['computer-vision', 'computer-vision'] | [-3.24335247e-01 -7.04453111e-01 -2.80594021e-01 -4.86759514e-01
-8.73274386e-01 -8.37417364e-01 6.99913204e-01 -1.20821245e-01
1.21603526e-01 -4.40450162e-02 7.84886479e-01 7.22937435e-02
-5.98912537e-02 -8.33126783e-01 -3.09387714e-01 -8.06065083e-01
2.57262528e-01 2.15910867e-01 -1.39346523e-02 1.44094825e-01
4.09692705e-01 9.21384752e-01 -1.25321138e+00 4.74651694e-01
9.17939767e-02 1.13131380e+00 3.46871704e-01 3.11088860e-01
-4.02160645e-01 4.04643416e-01 -4.46941137e-01 -1.23541020e-01
2.94166058e-01 -2.74726227e-02 -4.60428655e-01 1.32070154e-01
6.89916551e-01 -7.36897767e-01 -5.74982882e-01 1.09987724e+00
1.02450633e+00 1.61520660e-01 7.85300314e-01 -8.78458738e-01
-1.21395051e+00 1.41627401e-01 -8.26306462e-01 2.24812746e-01
3.34099501e-01 5.30781737e-03 1.44611311e+00 -1.47559822e+00
7.47675121e-01 1.75781083e+00 1.68034762e-01 4.02672201e-01
-1.02887845e+00 -5.02044261e-01 4.97002542e-01 3.94648425e-02
-1.13097918e+00 -1.69076622e-01 1.07247150e+00 -3.75413030e-01
1.26641536e+00 1.37581721e-01 8.63759041e-01 1.09934914e+00
3.18383276e-01 1.35522199e+00 8.16671431e-01 3.00375760e-01
-1.31524056e-01 -3.09541553e-01 4.69976738e-02 5.42400956e-01
1.85890451e-01 -1.32249370e-01 -5.87371826e-01 -4.55788136e-01
9.13540840e-01 7.16160893e-01 -2.12441608e-01 -9.09431756e-01
-1.30985117e+00 8.39935720e-01 5.98015964e-01 1.98335588e-01
-5.22651255e-01 3.67172658e-02 8.41879904e-01 2.62592018e-01
8.30852509e-01 6.12202426e-03 -2.42667913e-01 2.34314993e-01
-5.74230969e-01 2.38113701e-01 3.77785206e-01 1.10708356e+00
5.47313035e-01 -1.82914995e-02 -2.80953497e-01 1.09410882e+00
4.50196952e-01 8.23234677e-01 3.90918940e-01 -5.51128745e-01
7.17065096e-01 1.00679231e+00 -2.18744248e-01 -1.22900712e+00
-1.22552902e-01 -1.22390449e-01 -1.07528484e+00 -7.06731156e-02
-3.23689044e-01 4.61359262e-01 -1.02395284e+00 1.52645457e+00
2.07159996e-01 -3.41943175e-01 -9.78940427e-02 1.17143202e+00
1.31437862e+00 7.45527089e-01 -8.88396129e-02 1.26101002e-01
1.25639236e+00 -1.02514815e+00 -3.81114453e-01 -1.25022590e-01
2.12203547e-01 -9.91166294e-01 7.61863291e-01 -3.15644639e-03
-1.25882006e+00 -5.59920132e-01 -1.07611370e+00 -3.86426061e-01
-6.14644229e-01 1.77609175e-01 4.96857554e-01 1.99045897e-01
-1.01097715e+00 1.75270483e-01 -6.10869527e-01 -2.39470407e-01
5.55293560e-01 1.50180370e-01 -6.83665812e-01 -4.90098953e-01
-8.54436696e-01 6.10739350e-01 1.40775084e-01 8.95025358e-02
-1.04282784e+00 -4.09733921e-01 -1.07695997e+00 1.74774349e-01
2.32793778e-01 -8.31066728e-01 8.04564595e-01 -3.55592847e-01
-1.19266045e+00 1.08845854e+00 -3.99816751e-01 2.23862126e-01
1.94831476e-01 -3.13882828e-01 -1.55285791e-01 3.49102259e-01
1.38783753e-01 5.15106916e-01 9.96309519e-01 -1.33739161e+00
-1.60234958e-01 -8.46499324e-01 2.15864837e-01 5.44517219e-01
-2.11817116e-01 -3.58053409e-02 -7.48976290e-01 -7.45497048e-01
6.33768916e-01 -9.32225883e-01 -1.43173739e-01 9.41974372e-02
-3.15736592e-01 -5.96567214e-01 1.12427378e+00 -5.54405451e-01
8.89131725e-01 -2.12504888e+00 6.82308376e-01 1.42726272e-01
4.26622331e-01 2.06619054e-01 -4.49926615e-01 5.92568338e-01
-1.38400167e-01 -6.16592402e-03 1.63261905e-01 -4.23891962e-01
-7.92991966e-02 2.55637854e-01 -5.09813905e-01 4.47672874e-01
2.02094167e-01 1.28613329e+00 -8.54826152e-01 -3.52239966e-01
4.27607387e-01 5.67097247e-01 -5.15415192e-01 3.54134470e-01
-7.82541484e-02 2.81374175e-02 -9.63901162e-01 9.93269265e-01
8.72803926e-01 -6.71166062e-01 -8.47324729e-02 -4.11259145e-01
6.78955317e-02 2.33050659e-01 -7.26231694e-01 2.25592685e+00
-5.16892076e-01 3.29261482e-01 -1.50489360e-01 -1.00734091e+00
9.91492391e-01 2.19065592e-01 5.61676979e-01 -6.15163267e-01
-6.28976449e-02 6.32659122e-02 -7.21824944e-01 -6.12029791e-01
5.65461993e-01 -6.61371201e-02 -1.97853863e-01 7.45973229e-01
2.96990246e-01 -2.20390067e-01 -3.66939038e-01 1.73180357e-01
7.01740324e-01 2.15882033e-01 5.08733451e-01 -9.36679449e-03
6.94282830e-01 -6.30341530e-01 2.35221818e-01 5.91657639e-01
-1.70831114e-01 9.88745749e-01 3.95694971e-01 -7.78159440e-01
-1.09907210e+00 -1.22643912e+00 2.95851409e-01 9.53870714e-01
3.14035565e-01 -4.03524697e-01 1.33105619e-02 -9.12831903e-01
1.75900504e-01 1.46528095e-01 -5.98653853e-01 -1.92082718e-01
-7.01720059e-01 -3.06579858e-01 1.12224124e-01 6.42170131e-01
4.79836494e-01 -9.90259111e-01 -5.26237190e-01 8.49717855e-02
-6.27255291e-02 -9.97587383e-01 -9.10863340e-01 -1.46474719e-01
-1.07529485e+00 -7.69082785e-01 -1.29874778e+00 -9.73539352e-01
7.04201937e-01 1.03439367e+00 1.33790767e+00 -2.04911307e-01
-2.27357268e-01 6.99900985e-01 -3.80210757e-01 -1.39987186e-01
1.88993633e-01 1.56187966e-01 -1.21216476e-01 -1.69275656e-01
6.75615668e-01 -8.02527964e-01 -8.95315289e-01 2.31181040e-01
-9.86753106e-01 -1.86491519e-01 8.52699220e-01 9.36824560e-01
1.03188205e+00 -7.59801686e-01 3.27990055e-01 -5.60503662e-01
6.24622941e-01 -4.57595229e-01 -4.34577376e-01 6.49307132e-01
-1.44754767e-01 1.30513050e-02 4.36449975e-01 -3.69646311e-01
-7.72775769e-01 -6.56448379e-02 4.08977754e-02 -1.13258219e+00
2.15366352e-02 5.31180024e-01 -4.21921819e-01 -1.58331007e-01
8.03633109e-02 8.03467810e-01 3.43861505e-02 -5.51642656e-01
5.47877669e-01 4.92844135e-01 -1.84855208e-01 -3.80511194e-01
6.04062676e-01 7.08557546e-01 -6.28410652e-02 -7.75062442e-01
-9.83252943e-01 -6.94061220e-01 -5.72456062e-01 -7.38389939e-02
1.04106224e+00 -1.15273952e+00 -7.66536415e-01 3.62817675e-01
-1.35591519e+00 5.26923895e-01 4.43958156e-02 3.46456259e-01
-5.48022985e-01 5.27369738e-01 -5.96379399e-01 -5.37308574e-01
-6.97552562e-01 -1.13864338e+00 1.67893469e+00 2.32251599e-01
1.24172315e-01 -7.39720225e-01 5.36847301e-02 1.69501543e-01
3.97309899e-01 -5.79603426e-02 1.25595391e+00 -5.90419829e-01
-9.17871952e-01 -3.77802700e-01 -5.70858955e-01 4.13021028e-01
4.10144448e-01 -3.05340618e-01 -8.56678963e-01 -5.78702331e-01
2.53281057e-01 -4.76703048e-01 1.12876940e+00 3.98774236e-01
1.44961560e+00 -7.46461824e-02 -2.80566573e-01 4.93911445e-01
1.56102288e+00 4.99904342e-02 3.17127496e-01 -2.03696713e-01
8.37373972e-01 3.31760436e-01 3.10035706e-01 4.34578598e-01
3.41762275e-01 4.69374180e-01 5.00686169e-01 2.18724996e-01
1.24591872e-01 -4.61112767e-01 1.04829237e-01 1.35445952e+00
-5.48254214e-02 -3.76618117e-01 -5.07483244e-01 6.27862036e-01
-1.62297702e+00 -9.73235071e-01 6.37511909e-01 1.91943753e+00
1.35535419e-01 -5.59172221e-02 -8.99242014e-02 -4.62200105e-01
4.77988452e-01 1.06082964e+00 -7.67054677e-01 -1.53452829e-01
-3.39282006e-01 -9.05577317e-02 1.33097723e-01 1.04134552e-01
-1.10561764e+00 6.94432139e-01 5.71474171e+00 8.92375112e-01
-1.17776442e+00 -7.89363459e-02 4.23545033e-01 -3.77895117e-01
-7.28340685e-01 -2.53472328e-01 -6.97436213e-01 -8.91697034e-03
1.09044068e-01 -3.17361504e-02 1.13432527e-01 8.85763407e-01
-1.45288348e-01 2.75616884e-01 -1.01383209e+00 1.42575169e+00
6.60713434e-01 -1.37600446e+00 9.69282448e-01 8.17165822e-02
7.16906667e-01 2.87734479e-01 2.79092669e-01 1.26832828e-01
5.05957566e-02 -9.03057814e-01 3.96562040e-01 5.85197747e-01
5.29470146e-01 -7.63700426e-01 5.36283076e-01 1.13755777e-01
-1.59288752e+00 8.95238966e-02 -6.99640572e-01 5.26885927e-01
3.49874198e-01 4.19042826e-01 -1.70433566e-01 9.27857995e-01
7.27507591e-01 1.41079104e+00 -5.17215431e-01 8.34537148e-01
1.04294814e-01 -1.72518328e-01 -1.89362645e-01 -3.12556714e-01
5.74710071e-01 -4.29208457e-01 9.33477640e-01 7.46766448e-01
4.33359772e-01 1.29729629e-01 1.23512484e-01 9.03253615e-01
-1.08801566e-01 5.53096309e-02 -1.18795407e+00 -2.82410592e-01
1.18323117e-01 1.15250921e+00 -5.21147192e-01 -3.41859609e-01
-7.61610568e-01 1.17624784e+00 3.75250340e-01 6.52671874e-01
-2.31073990e-01 -1.82285905e-01 8.42178285e-01 -3.39665055e-01
7.80922771e-01 -3.85494083e-01 -8.89026821e-02 -1.30909789e+00
3.70205641e-01 -7.33812332e-01 4.71022666e-01 -9.88032818e-01
-1.79354846e+00 6.78516984e-01 -3.49370725e-02 -1.76656592e+00
-1.84870828e-02 -6.50889754e-01 -1.74797669e-01 1.02142608e+00
-1.42419231e+00 -1.47334623e+00 -9.00562331e-02 4.88909364e-01
8.60937536e-01 -3.86218667e-01 8.49691808e-01 2.84973890e-01
3.60310376e-02 2.68395215e-01 -1.68004371e-02 7.93851689e-02
6.70774996e-01 -9.59810317e-01 6.30499125e-01 1.32110059e-01
3.04980576e-01 9.24690127e-01 5.61720543e-02 -4.66045171e-01
-1.73341978e+00 -8.82451534e-01 9.51371014e-01 -4.36087430e-01
3.72420996e-01 -3.29959363e-01 -9.48510647e-01 5.41747153e-01
2.30013013e-01 2.42719784e-01 8.14757764e-01 -3.51075530e-02
-7.43415594e-01 9.70851034e-02 -7.59175181e-01 5.70175767e-01
1.26114225e+00 -1.09706283e+00 -7.19317436e-01 1.09814413e-01
8.60961556e-01 -5.45760691e-01 -7.91202724e-01 4.94996935e-01
8.00767779e-01 -1.11097169e+00 1.45108616e+00 -9.72993493e-01
7.73406267e-01 -1.34061528e-02 -6.24284744e-01 -1.30844462e+00
-3.27459395e-01 2.24583712e-03 -5.54815084e-02 9.21440423e-01
8.11252221e-02 -5.83157301e-01 6.64092064e-01 1.16094887e-01
7.75014013e-02 -1.11300659e+00 -9.42322612e-01 -4.82115239e-01
1.38303861e-01 -3.56375985e-02 6.88950896e-01 7.43357420e-01
-5.95386088e-01 4.02933329e-01 -4.06486899e-01 8.52655396e-02
6.19349480e-01 9.40754592e-01 5.88284314e-01 -9.94503975e-01
-7.31650461e-03 -7.07689345e-01 -6.44427359e-01 -1.67510712e+00
2.99857825e-01 -9.83813107e-01 -3.21570218e-01 -1.58221829e+00
6.39472187e-01 -2.39201374e-02 -4.97327596e-01 1.27168134e-01
-1.53362155e-01 1.43707722e-01 2.89577723e-01 2.29596287e-01
-7.08883405e-01 1.01994038e+00 1.62866473e+00 -4.21922505e-01
6.65551201e-02 -5.84706329e-02 -4.56971556e-01 6.54643357e-01
2.81894028e-01 -1.85726359e-01 -5.01976728e-01 -8.10897410e-01
2.98857361e-01 2.84477562e-01 4.99530733e-01 -3.07812214e-01
9.33822617e-02 -1.02345860e-02 6.53345764e-01 -1.33973682e+00
6.63934052e-01 -9.43066418e-01 -1.91025272e-01 1.52831376e-01
-3.31456602e-01 4.65150654e-01 1.06414236e-01 8.34091365e-01
-6.62672222e-01 1.25260815e-01 3.55675220e-01 -5.58583796e-01
-6.90473199e-01 8.84739101e-01 8.84626657e-02 -1.52216867e-01
6.19311988e-01 -1.25874430e-01 -2.16035128e-01 -4.83233690e-01
-5.47962844e-01 3.83152753e-01 3.62198621e-01 9.21766579e-01
1.12875640e+00 -1.93094182e+00 -4.48272318e-01 4.03613925e-01
4.98292655e-01 -9.31405462e-03 6.76457703e-01 3.50736529e-01
-1.88803464e-01 6.95999563e-01 -1.88440636e-01 -8.15245211e-01
-1.39692938e+00 5.45796871e-01 1.88096732e-01 -2.58393139e-01
-6.36054158e-01 9.75129724e-01 5.20721793e-01 -6.02583885e-01
6.61248863e-02 -2.31181607e-01 -3.69965464e-01 4.76795733e-01
4.97198492e-01 -3.77494395e-02 -1.53337508e-01 -7.77663410e-01
-3.88006091e-01 1.14643812e+00 -3.59593719e-01 9.43451598e-02
1.77976120e+00 -3.16800207e-01 -4.70764160e-01 8.21256280e-01
1.84241378e+00 -3.03162754e-01 -9.44401443e-01 -6.77487195e-01
-3.43919367e-01 -6.85279489e-01 -2.13324338e-01 -1.92363545e-01
-1.32013714e+00 1.20071208e+00 5.16431868e-01 1.51232585e-01
9.86934245e-01 2.60730028e-01 9.13135588e-01 5.33509135e-01
2.84761280e-01 -7.11600721e-01 4.47453201e-01 6.30783439e-01
1.33625805e+00 -1.26806712e+00 9.35045034e-02 -7.99600333e-02
-5.59659004e-01 1.02042854e+00 5.10394156e-01 -4.28937823e-01
9.24695790e-01 -4.95467752e-01 -1.81233566e-02 -6.23009205e-01
-7.36377418e-01 -1.18822809e-02 6.11780882e-01 1.65783599e-01
2.17009500e-01 -4.18494642e-02 7.81301185e-02 4.12802994e-01
2.92902857e-01 -3.20783168e-01 -2.18538612e-01 9.55909252e-01
-1.59038231e-01 -9.23525512e-01 4.37119752e-02 5.46904981e-01
-3.24436069e-01 -6.26949444e-02 -6.10102594e-01 4.10672337e-01
-4.53005522e-01 1.94733351e-01 -5.93285225e-02 -2.87893295e-01
5.11329353e-01 1.41459987e-01 4.58197266e-01 -4.71014738e-01
-3.41891408e-01 2.77286261e-01 -4.32503998e-01 -7.25481987e-01
-7.64755309e-01 -4.61643547e-01 -6.32825017e-01 -2.54857820e-02
-1.17034979e-01 -3.53581101e-01 3.54497045e-01 7.67687440e-01
4.54808503e-01 2.69658178e-01 8.20580721e-01 -1.17454123e+00
-7.62580931e-01 -6.86919332e-01 -6.19136453e-01 4.74500060e-01
5.22086263e-01 -7.02108681e-01 -3.63709450e-01 -4.19837743e-01] | [8.140265464782715, -3.8591115474700928] |
f023ae73-8889-4815-90fb-11a3130ecef2 | spatiotemporal-multi-scale-bilateral-motion | 2209.12364 | null | https://arxiv.org/abs/2209.12364v1 | https://arxiv.org/pdf/2209.12364v1.pdf | Spatiotemporal Multi-scale Bilateral Motion Network for Gait Recognition | The critical goal of gait recognition is to acquire the inter-frame walking habit representation from the gait sequences. The relations between frames, however, have not received adequate attention in comparison to the intra-frame features. In this paper, motivated by optical flow, the bilateral motion-oriented features are proposed, which can allow the classic convolutional structure to have the capability to directly portray gait movement patterns at the feature level. Based on such features, we develop a set of multi-scale temporal representations that force the motion context to be richly described at various levels of temporal resolution. Furthermore, a correction block is devised to eliminate the segmentation noise of silhouettes for getting more precise gait information. Subsequently, the temporal feature set and the spatial features are combined to comprehensively characterize gait processes. Extensive experiments are conducted on CASIA-B and OU-MVLP datasets, and the results achieve an outstanding identification performance, which has demonstrated the effectiveness of the proposed approach. | ['Kejun Wang', 'Yu Zhang', 'Shan Du', 'Xinnan Ding'] | 2022-09-26 | null | null | null | null | ['gait-recognition'] | ['computer-vision'] | [-3.63202430e-02 -7.69471943e-01 -6.42010793e-02 -3.54640573e-01
-9.16593298e-02 1.94735564e-02 3.17174017e-01 -2.69304663e-01
-1.94011152e-01 7.40969479e-01 2.93060690e-01 4.11417723e-01
-1.02312215e-01 -9.16664898e-01 -1.77319780e-01 -9.36066806e-01
-4.87650901e-01 -4.90559429e-01 3.68672788e-01 -3.05287689e-01
1.18358470e-01 2.16114089e-01 -1.61168945e+00 -2.23528240e-02
8.29089165e-01 1.01191831e+00 3.22115600e-01 4.82470691e-01
5.52244745e-02 7.13336527e-01 -3.62294495e-01 -3.33278850e-02
-2.05106996e-02 -5.00720620e-01 -4.46454287e-01 4.65856493e-01
1.01587981e-01 -4.62706417e-01 -7.97583222e-01 9.39416289e-01
4.12872702e-01 2.17069298e-01 4.25726533e-01 -1.11238515e+00
-5.37033141e-01 6.10271143e-03 -6.89739168e-01 6.21794641e-01
4.81721163e-01 4.57854211e-01 9.88961399e-01 -7.56018221e-01
4.74931836e-01 1.43135929e+00 4.96358365e-01 1.88394293e-01
-8.26686740e-01 -5.59907973e-01 2.35606313e-01 7.45447636e-01
-1.29028237e+00 -2.07203209e-01 1.00687242e+00 -3.83444130e-01
6.02715373e-01 9.41401049e-02 1.04730952e+00 1.04890239e+00
2.31443375e-01 9.29813683e-01 7.82076180e-01 -1.73189938e-01
-1.89756647e-01 -7.41638005e-01 1.65405065e-01 8.55427980e-01
3.99780482e-01 1.42723277e-01 -4.93911386e-01 3.72004211e-01
1.09445059e+00 1.89892352e-01 -6.27902746e-01 -1.96132287e-01
-1.36703444e+00 5.56640565e-01 6.31798148e-01 6.20316327e-01
-5.08162558e-01 1.38914913e-01 6.50238931e-01 -1.26677647e-01
2.82403111e-01 -1.63287625e-01 8.06846097e-02 -3.71889174e-01
-8.10615003e-01 1.22882940e-01 4.36077416e-02 7.44680703e-01
8.31058025e-01 4.27429587e-01 -3.03985178e-01 7.59214640e-01
3.74331474e-01 5.57894528e-01 6.14818931e-01 -9.41025972e-01
4.70163286e-01 6.79503024e-01 1.70671437e-02 -1.69477379e+00
-2.72605985e-01 -1.63220733e-01 -1.16154611e+00 -1.85735673e-01
3.75031382e-01 1.23763837e-01 -7.26648748e-01 1.58050871e+00
1.28322035e-01 6.20358586e-01 -1.00789696e-01 1.16555297e+00
5.90103626e-01 6.61046386e-01 2.46632516e-01 -2.98651934e-01
1.44793475e+00 -5.88515282e-01 -1.07479560e+00 -9.27073732e-02
3.46686482e-01 -4.94063377e-01 9.65453148e-01 -9.52074602e-02
-8.23971629e-01 -1.13364673e+00 -1.29405117e+00 1.13021761e-01
-2.11982414e-01 2.69338220e-01 6.13909841e-01 2.09297016e-01
-7.01233447e-01 6.53677166e-01 -1.09989452e+00 -5.16345322e-01
4.33176994e-01 1.26428917e-01 -4.12781090e-01 -9.49792638e-02
-1.60860968e+00 6.07336402e-01 6.74685001e-01 6.76216781e-01
-4.70408142e-01 -2.00440601e-01 -1.16348422e+00 -1.09038306e-02
1.92799550e-02 -6.13813102e-01 6.46206260e-01 -5.74901938e-01
-1.25458574e+00 4.75178689e-01 -2.89538324e-01 -2.77516365e-01
6.80514157e-01 -9.27719921e-02 -6.20879233e-01 5.87354064e-01
3.26387405e-01 4.19571072e-01 8.31064820e-01 -9.03937161e-01
-8.92803431e-01 -2.65569210e-01 -8.17715470e-03 2.58556515e-01
-4.79346275e-01 -2.67788708e-01 -6.76144838e-01 -9.25035715e-01
1.01105437e-01 -5.12653947e-01 -1.04322404e-01 -4.27243412e-02
-2.41144925e-01 -8.08741618e-03 1.10862577e+00 -8.07353616e-01
1.53625822e+00 -2.19794536e+00 1.87119037e-01 7.91301653e-02
1.58553436e-01 4.65170264e-01 6.77214414e-02 2.33063027e-01
-7.26401731e-02 -5.28963134e-02 -3.59978229e-01 -3.84915769e-02
-1.63068518e-01 4.50173497e-01 -1.33464798e-01 4.53655750e-01
4.26544100e-01 8.83664966e-01 -8.75174880e-01 -8.15319538e-01
6.26432657e-01 4.43071276e-01 -3.35179657e-01 2.13036418e-01
3.52880329e-01 5.68009675e-01 -8.79812121e-01 6.00566506e-01
6.59403265e-01 -1.55549303e-01 2.05040481e-02 -5.94530702e-01
1.12245269e-02 -1.58886254e-01 -1.19377494e+00 1.72629178e+00
-9.36505497e-02 4.21106726e-01 -1.37856334e-01 -1.20743752e+00
1.03377807e+00 1.87611938e-01 8.69529605e-01 -8.41234565e-01
1.02992713e-01 4.83247526e-02 8.15184042e-03 -8.11426401e-01
3.94931465e-01 1.61059484e-01 -8.55055898e-02 -5.56701310e-02
-1.86347738e-02 4.59572822e-01 6.25468433e-01 -2.56404847e-01
7.94855118e-01 3.79449546e-01 2.94190705e-01 -2.50103563e-01
1.12075412e+00 -1.43725619e-01 1.08366287e+00 2.63799012e-01
-6.52772486e-01 5.12449384e-01 -2.14668307e-02 -6.86991334e-01
-8.99682879e-01 -9.67767835e-01 -4.08845022e-02 3.99322540e-01
6.21380568e-01 -4.66782004e-01 -6.43391669e-01 -3.89406711e-01
-8.90375972e-02 -1.21207498e-01 -5.54964662e-01 -3.51652235e-01
-1.07900739e+00 -8.60493898e-01 4.73942697e-01 8.05864096e-01
1.29224789e+00 -1.07400334e+00 -8.59578371e-01 5.18826663e-01
-6.88811123e-01 -1.28795826e+00 -6.40777886e-01 -4.52067554e-01
-8.18977416e-01 -1.15702164e+00 -9.95128334e-01 -9.31930125e-01
2.84893483e-01 3.06653082e-01 6.22342467e-01 3.59626532e-01
-6.04982495e-01 8.81178677e-02 -4.17396754e-01 2.91832298e-01
2.51895458e-01 -1.70723200e-01 1.59261972e-01 3.39726686e-01
4.99431819e-01 -8.25440824e-01 -8.30412209e-01 3.40628475e-01
-7.08343565e-01 1.16770722e-01 5.26194930e-01 1.06090772e+00
3.64048690e-01 7.94990137e-02 6.18324518e-01 -8.88785347e-02
4.64336723e-01 -1.69211909e-01 -7.99405128e-02 1.31076857e-01
-1.47223864e-02 -2.27480993e-01 5.96454740e-01 -2.94816047e-01
-1.06343615e+00 -1.46785736e-01 -1.00763373e-01 -6.26243651e-01
-2.39339054e-01 5.91793180e-01 -4.76029158e-01 1.57995269e-01
7.80474991e-02 6.59048140e-01 5.14089353e-02 -3.24223250e-01
2.05827147e-01 4.01938051e-01 8.28925729e-01 -7.27574646e-01
7.91286647e-01 5.59597611e-01 3.88082340e-02 -8.73513997e-01
-4.42097157e-01 -4.49540377e-01 -9.78147924e-01 -4.12373841e-01
1.11891770e+00 -8.61968338e-01 -7.64657319e-01 8.31793964e-01
-9.90928710e-01 1.68932974e-01 -1.52932554e-01 6.44047618e-01
-7.96507418e-01 9.28551316e-01 -1.04193175e+00 -5.69847047e-01
-1.01094484e-01 -1.22329795e+00 9.14801300e-01 5.16110420e-01
-8.59552920e-02 -1.00703514e+00 -1.39964655e-01 2.50177085e-02
1.58586249e-01 4.82818514e-01 7.90764987e-01 -3.17969732e-02
-5.80764234e-01 2.40422562e-02 -2.90349245e-01 1.51651949e-01
6.97153807e-01 2.76424825e-01 -7.20067501e-01 -1.91387832e-01
-1.39748514e-01 -1.10908803e-02 8.62327874e-01 5.15877366e-01
1.01178420e+00 5.05412146e-02 -4.73941356e-01 7.09439278e-01
1.15042102e+00 3.93541843e-01 8.18499804e-01 5.18608630e-01
8.81001532e-01 5.74792504e-01 9.90882754e-01 5.55951834e-01
3.96694839e-01 7.37872720e-01 2.32888907e-01 -9.71102417e-02
-1.23953529e-01 -3.21967274e-01 2.61655271e-01 1.24926710e+00
-4.69005078e-01 1.41874149e-01 -6.98163211e-01 5.73217392e-01
-2.06096148e+00 -1.25923789e+00 -1.94451407e-01 1.74152374e+00
4.44963753e-01 2.10384540e-02 1.43534109e-01 4.31902707e-01
1.02779579e+00 6.12558365e-01 -4.87412453e-01 2.01520890e-01
-2.40653589e-01 -7.78385475e-02 8.50222036e-02 2.02884302e-01
-1.52341533e+00 7.77653575e-01 6.15805912e+00 8.70135665e-01
-9.98701513e-01 -3.37474525e-01 4.79011565e-01 4.66047615e-01
2.88114399e-01 -2.55376697e-01 -5.10814548e-01 8.17273259e-01
4.85817701e-01 -3.48384887e-01 2.50892434e-02 5.68809092e-01
5.74260056e-01 1.28289893e-01 -7.25807846e-01 1.28078723e+00
-2.75908083e-01 -1.18521190e+00 1.35552347e-01 1.21975183e-01
3.66662413e-01 -6.97214842e-01 -6.32305222e-04 1.59180239e-01
1.19964697e-03 -8.33394766e-01 5.03317952e-01 7.92783976e-01
7.34961390e-01 -9.09949243e-01 6.74891829e-01 2.12017477e-01
-2.17004323e+00 -1.70301154e-01 -4.57376868e-01 -2.51737654e-01
4.93976027e-01 4.20403838e-01 -1.60033777e-01 1.19163382e+00
9.06979561e-01 1.67247760e+00 -5.12446582e-01 1.05166125e+00
-1.24201588e-01 5.06967604e-01 -6.49930164e-02 1.89869493e-01
3.09329569e-01 -4.38212276e-01 3.88973504e-01 1.28048420e+00
5.12341976e-01 2.19298333e-01 4.29880321e-01 8.51727128e-01
3.94838452e-01 9.48222652e-02 -4.75759238e-01 -7.26570785e-02
2.90842324e-01 1.21467197e+00 -5.70375025e-01 -3.30774307e-01
-4.66631234e-01 9.47806895e-01 5.24864979e-02 3.87155831e-01
-7.67262697e-01 -3.78534108e-01 1.02448213e+00 -2.37661436e-01
3.91483724e-01 -4.68248844e-01 7.61075616e-02 -1.58765244e+00
2.16512844e-01 -5.68765998e-01 4.31387186e-01 -6.19751692e-01
-1.49426043e+00 3.04307014e-01 -6.89535961e-02 -1.67398691e+00
-2.96884745e-01 -4.98977542e-01 -8.73579979e-01 1.02896214e+00
-1.49407768e+00 -1.16104233e+00 -5.89521050e-01 8.01538169e-01
5.62516630e-01 5.95565476e-02 5.45805216e-01 6.94200277e-01
-8.77626538e-01 3.11604470e-01 -1.28398210e-01 6.51087523e-01
4.99888003e-01 -9.21021163e-01 3.83227766e-01 1.11705518e+00
-1.86495721e-01 6.21473432e-01 4.74355221e-01 -6.95002019e-01
-1.19841659e+00 -1.15794802e+00 4.32225645e-01 3.75807509e-02
6.19115829e-01 4.97918911e-02 -1.08070695e+00 4.84732628e-01
-1.80647582e-01 3.57514620e-01 3.87124717e-01 -4.10715461e-01
7.12769479e-02 -1.85222700e-01 -7.82302618e-01 5.10694504e-01
1.26534474e+00 -4.88668203e-01 -8.37684095e-01 -2.73409784e-01
3.18878144e-01 -1.70253187e-01 -1.01815116e+00 6.52975619e-01
6.00988686e-01 -8.74602497e-01 1.25733018e+00 -4.13768709e-01
5.58917880e-01 -7.26673961e-01 -2.33813107e-01 -1.13605309e+00
-5.53120971e-01 -2.15490624e-01 -1.22265324e-01 1.42388666e+00
-5.36767900e-01 -4.32230413e-01 6.74015284e-01 -3.76607142e-02
-9.33868214e-02 -6.25925481e-01 -8.32023919e-01 -6.94066048e-01
-2.56634533e-01 -2.48750895e-01 6.43800199e-01 9.35454547e-01
7.79351220e-02 -6.77236319e-02 -7.01744735e-01 1.07002690e-01
5.97356260e-01 3.12384874e-01 5.57483494e-01 -1.12830961e+00
-3.57505574e-04 -4.43679214e-01 -1.06211603e+00 -1.21417129e+00
1.37530148e-01 -6.27174199e-01 -4.89361677e-03 -1.54616833e+00
5.11769168e-02 -1.07278042e-01 -4.88695294e-01 2.20531002e-01
-5.86052001e-01 2.83855855e-01 1.90208599e-01 4.65115339e-01
-4.02678818e-01 9.53992307e-01 1.74582434e+00 -2.46356368e-01
1.33333892e-01 -2.61863738e-01 -2.21409753e-01 8.08793545e-01
6.17924690e-01 2.73451716e-01 -4.50384021e-01 -1.90676808e-01
-7.20421195e-01 2.42480576e-01 3.94987553e-01 -1.35926247e+00
5.80278449e-02 -1.54346913e-01 7.03718901e-01 -6.83323741e-01
5.15283048e-01 -6.42572582e-01 -2.61257999e-02 6.29000127e-01
-2.08415464e-02 2.13541418e-01 -8.81294161e-02 7.82011271e-01
-7.06851482e-01 3.66273046e-01 6.24499500e-01 -1.67623565e-01
-1.44574308e+00 7.32959747e-01 -3.49056423e-01 -3.51200104e-02
1.08784556e+00 -6.59475327e-01 -1.22601427e-01 -1.70846850e-01
-8.81778061e-01 3.72990042e-01 3.17137599e-01 5.31522512e-01
8.01778853e-01 -1.85653389e+00 -4.83127207e-01 4.36501294e-01
1.51883975e-01 -1.95778519e-01 6.39514387e-01 8.54412556e-01
-6.33399546e-01 4.03433204e-01 -7.80347884e-01 -8.93623769e-01
-9.58010018e-01 4.39953238e-01 4.58697349e-01 -2.18809769e-01
-1.00039208e+00 2.77012378e-01 2.81747967e-01 1.06266454e-01
-1.91037729e-01 -3.72893572e-01 -5.97640574e-01 1.68511778e-01
6.84650481e-01 3.39525223e-01 -2.89693028e-01 -1.19007516e+00
-4.26194966e-01 9.71841276e-01 1.20032564e-01 1.83981046e-01
1.15280187e+00 -5.93208849e-01 7.69684762e-02 4.59635079e-01
1.11844301e+00 -5.55997491e-01 -1.68100107e+00 -2.39121914e-01
9.54014659e-02 -8.82648766e-01 -2.81934023e-01 -1.35828063e-01
-1.31659293e+00 1.31277037e+00 7.25886464e-01 2.49390025e-02
1.20302510e+00 -6.33759975e-01 1.07019925e+00 -1.99010260e-02
5.73464692e-01 -9.37095761e-01 2.42400959e-01 3.76282364e-01
4.91491288e-01 -1.01179612e+00 -2.10567698e-01 -5.52532077e-01
-4.85686004e-01 1.42647636e+00 8.03288102e-01 -3.33779693e-01
4.45499718e-01 1.00828998e-01 -1.07883431e-01 5.43318130e-02
-2.95392931e-01 -4.51600701e-01 2.60375261e-01 8.25603127e-01
2.65670389e-01 -3.06686591e-02 -3.78663599e-01 5.32212079e-01
7.81870559e-02 1.74262166e-01 1.92034483e-01 9.53939319e-01
-5.05343199e-01 -8.98623765e-01 -2.82220125e-01 2.04225406e-01
-3.11425388e-01 3.29823852e-01 2.24996671e-01 9.36962724e-01
1.49101883e-01 9.23341036e-01 6.16338998e-02 -6.02744639e-01
3.29450250e-01 -1.25597745e-01 2.61430770e-01 -2.24036187e-01
3.69515009e-02 1.67733982e-01 -6.55261427e-02 -7.59950161e-01
-7.91016281e-01 -6.34217918e-01 -1.30763364e+00 -3.54487568e-01
3.79162319e-02 7.55641684e-02 -1.51582062e-01 1.07592452e+00
1.38971895e-01 7.68316269e-01 5.34150839e-01 -1.08947730e+00
-9.79673341e-02 -8.29945147e-01 -8.35105598e-01 7.65275776e-01
2.98166841e-01 -1.04980707e+00 -1.31023660e-01 3.57805520e-01] | [14.283258438110352, 1.4251500368118286] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.